diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,98033 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9619348632678302, + "eval_steps": 500, + "global_step": 14000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 6.87096330905593e-05, + "grad_norm": 7.6275835037231445, + "learning_rate": 3.4340659340659344e-08, + "loss": 2.8946, + "step": 1 + }, + { + "epoch": 0.0001374192661811186, + "grad_norm": 7.034836769104004, + "learning_rate": 6.868131868131869e-08, + "loss": 2.9183, + "step": 2 + }, + { + "epoch": 0.00020612889927167788, + "grad_norm": 6.99338960647583, + "learning_rate": 1.0302197802197802e-07, + "loss": 2.8106, + "step": 3 + }, + { + "epoch": 0.0002748385323622372, + "grad_norm": 7.4205002784729, + "learning_rate": 1.3736263736263737e-07, + "loss": 2.9096, + "step": 4 + }, + { + "epoch": 0.0003435481654527965, + "grad_norm": 7.320061683654785, + "learning_rate": 1.717032967032967e-07, + "loss": 2.8319, + "step": 5 + }, + { + "epoch": 0.00041225779854335576, + "grad_norm": 7.4554290771484375, + "learning_rate": 2.0604395604395605e-07, + "loss": 2.9632, + "step": 6 + }, + { + "epoch": 0.0004809674316339151, + "grad_norm": 7.403331756591797, + "learning_rate": 2.403846153846154e-07, + "loss": 2.9145, + "step": 7 + }, + { + "epoch": 0.0005496770647244744, + "grad_norm": 7.208906650543213, + "learning_rate": 2.7472527472527475e-07, + "loss": 2.9067, + "step": 8 + }, + { + "epoch": 0.0006183866978150337, + "grad_norm": 7.182309150695801, + "learning_rate": 3.0906593406593407e-07, + "loss": 2.9476, + "step": 9 + }, + { + "epoch": 0.000687096330905593, + "grad_norm": 7.336084365844727, + "learning_rate": 3.434065934065934e-07, + "loss": 2.9436, + "step": 10 + }, + { + "epoch": 0.0007558059639961522, + "grad_norm": 7.305691719055176, + "learning_rate": 3.7774725274725277e-07, + "loss": 2.8734, + "step": 11 + }, + { + "epoch": 0.0008245155970867115, + "grad_norm": 6.958437919616699, + "learning_rate": 4.120879120879121e-07, + "loss": 2.8335, + "step": 12 + }, + { + "epoch": 0.0008932252301772709, + "grad_norm": 7.204424858093262, + "learning_rate": 4.464285714285714e-07, + "loss": 2.8405, + "step": 13 + }, + { + "epoch": 0.0009619348632678302, + "grad_norm": 7.613925933837891, + "learning_rate": 4.807692307692308e-07, + "loss": 2.8819, + "step": 14 + }, + { + "epoch": 0.0010306444963583896, + "grad_norm": 7.293791770935059, + "learning_rate": 5.151098901098902e-07, + "loss": 2.9231, + "step": 15 + }, + { + "epoch": 0.0010993541294489488, + "grad_norm": 7.24744176864624, + "learning_rate": 5.494505494505495e-07, + "loss": 2.8779, + "step": 16 + }, + { + "epoch": 0.001168063762539508, + "grad_norm": 7.415509223937988, + "learning_rate": 5.837912087912088e-07, + "loss": 2.9154, + "step": 17 + }, + { + "epoch": 0.0012367733956300674, + "grad_norm": 7.4035139083862305, + "learning_rate": 6.181318681318681e-07, + "loss": 2.8272, + "step": 18 + }, + { + "epoch": 0.0013054830287206266, + "grad_norm": 7.272813320159912, + "learning_rate": 6.524725274725276e-07, + "loss": 2.8737, + "step": 19 + }, + { + "epoch": 0.001374192661811186, + "grad_norm": 7.2420196533203125, + "learning_rate": 6.868131868131868e-07, + "loss": 2.9522, + "step": 20 + }, + { + "epoch": 0.0014429022949017452, + "grad_norm": 7.265440940856934, + "learning_rate": 7.211538461538462e-07, + "loss": 2.8594, + "step": 21 + }, + { + "epoch": 0.0015116119279923045, + "grad_norm": 7.195631504058838, + "learning_rate": 7.554945054945055e-07, + "loss": 2.821, + "step": 22 + }, + { + "epoch": 0.0015803215610828637, + "grad_norm": 7.36093807220459, + "learning_rate": 7.898351648351649e-07, + "loss": 2.9065, + "step": 23 + }, + { + "epoch": 0.001649031194173423, + "grad_norm": 6.964420318603516, + "learning_rate": 8.241758241758242e-07, + "loss": 2.8252, + "step": 24 + }, + { + "epoch": 0.0017177408272639825, + "grad_norm": 7.27252197265625, + "learning_rate": 8.585164835164836e-07, + "loss": 2.8766, + "step": 25 + }, + { + "epoch": 0.0017864504603545418, + "grad_norm": 6.662872314453125, + "learning_rate": 8.928571428571428e-07, + "loss": 2.7526, + "step": 26 + }, + { + "epoch": 0.001855160093445101, + "grad_norm": 7.0713210105896, + "learning_rate": 9.271978021978023e-07, + "loss": 2.7311, + "step": 27 + }, + { + "epoch": 0.0019238697265356603, + "grad_norm": 7.623928070068359, + "learning_rate": 9.615384615384617e-07, + "loss": 2.8237, + "step": 28 + }, + { + "epoch": 0.0019925793596262194, + "grad_norm": 7.193939208984375, + "learning_rate": 9.958791208791208e-07, + "loss": 2.8225, + "step": 29 + }, + { + "epoch": 0.002061288992716779, + "grad_norm": 6.923072338104248, + "learning_rate": 1.0302197802197803e-06, + "loss": 2.6622, + "step": 30 + }, + { + "epoch": 0.0021299986258073384, + "grad_norm": 7.405568599700928, + "learning_rate": 1.0645604395604397e-06, + "loss": 2.7985, + "step": 31 + }, + { + "epoch": 0.0021987082588978977, + "grad_norm": 6.8562541007995605, + "learning_rate": 1.098901098901099e-06, + "loss": 2.7624, + "step": 32 + }, + { + "epoch": 0.002267417891988457, + "grad_norm": 7.247007369995117, + "learning_rate": 1.1332417582417583e-06, + "loss": 2.8215, + "step": 33 + }, + { + "epoch": 0.002336127525079016, + "grad_norm": 7.152578830718994, + "learning_rate": 1.1675824175824176e-06, + "loss": 2.7058, + "step": 34 + }, + { + "epoch": 0.0024048371581695755, + "grad_norm": 6.76224946975708, + "learning_rate": 1.201923076923077e-06, + "loss": 2.6896, + "step": 35 + }, + { + "epoch": 0.0024735467912601347, + "grad_norm": 6.8790364265441895, + "learning_rate": 1.2362637362637363e-06, + "loss": 2.6491, + "step": 36 + }, + { + "epoch": 0.002542256424350694, + "grad_norm": 6.814023494720459, + "learning_rate": 1.2706043956043956e-06, + "loss": 2.5844, + "step": 37 + }, + { + "epoch": 0.0026109660574412533, + "grad_norm": 6.636133670806885, + "learning_rate": 1.3049450549450551e-06, + "loss": 2.6174, + "step": 38 + }, + { + "epoch": 0.0026796756905318126, + "grad_norm": 6.30716609954834, + "learning_rate": 1.3392857142857143e-06, + "loss": 2.4916, + "step": 39 + }, + { + "epoch": 0.002748385323622372, + "grad_norm": 6.623904228210449, + "learning_rate": 1.3736263736263736e-06, + "loss": 2.5194, + "step": 40 + }, + { + "epoch": 0.002817094956712931, + "grad_norm": 6.534739971160889, + "learning_rate": 1.4079670329670331e-06, + "loss": 2.4827, + "step": 41 + }, + { + "epoch": 0.0028858045898034904, + "grad_norm": 6.654708385467529, + "learning_rate": 1.4423076923076924e-06, + "loss": 2.4983, + "step": 42 + }, + { + "epoch": 0.0029545142228940497, + "grad_norm": 6.537683486938477, + "learning_rate": 1.4766483516483518e-06, + "loss": 2.5157, + "step": 43 + }, + { + "epoch": 0.003023223855984609, + "grad_norm": 6.166847229003906, + "learning_rate": 1.510989010989011e-06, + "loss": 2.444, + "step": 44 + }, + { + "epoch": 0.0030919334890751682, + "grad_norm": 6.101149082183838, + "learning_rate": 1.5453296703296704e-06, + "loss": 2.3854, + "step": 45 + }, + { + "epoch": 0.0031606431221657275, + "grad_norm": 6.729682922363281, + "learning_rate": 1.5796703296703297e-06, + "loss": 2.4451, + "step": 46 + }, + { + "epoch": 0.0032293527552562868, + "grad_norm": 6.204443454742432, + "learning_rate": 1.6140109890109893e-06, + "loss": 2.4385, + "step": 47 + }, + { + "epoch": 0.003298062388346846, + "grad_norm": 5.8173298835754395, + "learning_rate": 1.6483516483516484e-06, + "loss": 2.3583, + "step": 48 + }, + { + "epoch": 0.0033667720214374053, + "grad_norm": 5.775567531585693, + "learning_rate": 1.6826923076923077e-06, + "loss": 2.2726, + "step": 49 + }, + { + "epoch": 0.003435481654527965, + "grad_norm": 5.820885181427002, + "learning_rate": 1.7170329670329672e-06, + "loss": 2.3267, + "step": 50 + }, + { + "epoch": 0.0035041912876185243, + "grad_norm": 5.634654998779297, + "learning_rate": 1.7513736263736266e-06, + "loss": 2.2839, + "step": 51 + }, + { + "epoch": 0.0035729009207090836, + "grad_norm": 5.820955276489258, + "learning_rate": 1.7857142857142857e-06, + "loss": 2.2362, + "step": 52 + }, + { + "epoch": 0.003641610553799643, + "grad_norm": 5.6372389793396, + "learning_rate": 1.8200549450549452e-06, + "loss": 2.2835, + "step": 53 + }, + { + "epoch": 0.003710320186890202, + "grad_norm": 5.373723983764648, + "learning_rate": 1.8543956043956045e-06, + "loss": 2.1533, + "step": 54 + }, + { + "epoch": 0.0037790298199807614, + "grad_norm": 5.111402988433838, + "learning_rate": 1.8887362637362636e-06, + "loss": 2.2055, + "step": 55 + }, + { + "epoch": 0.0038477394530713207, + "grad_norm": 4.865420818328857, + "learning_rate": 1.9230769230769234e-06, + "loss": 2.1924, + "step": 56 + }, + { + "epoch": 0.0039164490861618795, + "grad_norm": 4.456362724304199, + "learning_rate": 1.9574175824175823e-06, + "loss": 2.1377, + "step": 57 + }, + { + "epoch": 0.003985158719252439, + "grad_norm": 4.122003078460693, + "learning_rate": 1.9917582417582416e-06, + "loss": 2.0542, + "step": 58 + }, + { + "epoch": 0.004053868352342998, + "grad_norm": 3.8278002738952637, + "learning_rate": 2.0260989010989014e-06, + "loss": 2.0219, + "step": 59 + }, + { + "epoch": 0.004122577985433558, + "grad_norm": 3.7971367835998535, + "learning_rate": 2.0604395604395607e-06, + "loss": 2.0652, + "step": 60 + }, + { + "epoch": 0.0041912876185241175, + "grad_norm": 3.7625250816345215, + "learning_rate": 2.0947802197802196e-06, + "loss": 2.0637, + "step": 61 + }, + { + "epoch": 0.004259997251614677, + "grad_norm": 3.3542184829711914, + "learning_rate": 2.1291208791208793e-06, + "loss": 2.002, + "step": 62 + }, + { + "epoch": 0.004328706884705236, + "grad_norm": 3.2720158100128174, + "learning_rate": 2.1634615384615387e-06, + "loss": 1.9667, + "step": 63 + }, + { + "epoch": 0.004397416517795795, + "grad_norm": 3.4174327850341797, + "learning_rate": 2.197802197802198e-06, + "loss": 2.0516, + "step": 64 + }, + { + "epoch": 0.004466126150886355, + "grad_norm": 3.233170986175537, + "learning_rate": 2.2321428571428573e-06, + "loss": 2.0174, + "step": 65 + }, + { + "epoch": 0.004534835783976914, + "grad_norm": 3.1018035411834717, + "learning_rate": 2.2664835164835166e-06, + "loss": 1.9514, + "step": 66 + }, + { + "epoch": 0.004603545417067473, + "grad_norm": 3.185926675796509, + "learning_rate": 2.300824175824176e-06, + "loss": 1.9775, + "step": 67 + }, + { + "epoch": 0.004672255050158032, + "grad_norm": 3.1349759101867676, + "learning_rate": 2.3351648351648353e-06, + "loss": 1.9103, + "step": 68 + }, + { + "epoch": 0.004740964683248592, + "grad_norm": 3.2016353607177734, + "learning_rate": 2.3695054945054946e-06, + "loss": 1.8872, + "step": 69 + }, + { + "epoch": 0.004809674316339151, + "grad_norm": 2.82181453704834, + "learning_rate": 2.403846153846154e-06, + "loss": 1.8656, + "step": 70 + }, + { + "epoch": 0.00487838394942971, + "grad_norm": 2.9412529468536377, + "learning_rate": 2.4381868131868133e-06, + "loss": 1.8396, + "step": 71 + }, + { + "epoch": 0.0049470935825202695, + "grad_norm": 2.7847046852111816, + "learning_rate": 2.4725274725274726e-06, + "loss": 1.9568, + "step": 72 + }, + { + "epoch": 0.005015803215610829, + "grad_norm": 2.9551219940185547, + "learning_rate": 2.506868131868132e-06, + "loss": 1.843, + "step": 73 + }, + { + "epoch": 0.005084512848701388, + "grad_norm": 2.636174440383911, + "learning_rate": 2.5412087912087912e-06, + "loss": 1.8051, + "step": 74 + }, + { + "epoch": 0.005153222481791947, + "grad_norm": 2.5625147819519043, + "learning_rate": 2.5755494505494505e-06, + "loss": 1.8249, + "step": 75 + }, + { + "epoch": 0.005221932114882507, + "grad_norm": 2.8047335147857666, + "learning_rate": 2.6098901098901103e-06, + "loss": 1.8585, + "step": 76 + }, + { + "epoch": 0.005290641747973066, + "grad_norm": 2.790449619293213, + "learning_rate": 2.644230769230769e-06, + "loss": 1.8639, + "step": 77 + }, + { + "epoch": 0.005359351381063625, + "grad_norm": 2.593555450439453, + "learning_rate": 2.6785714285714285e-06, + "loss": 1.7297, + "step": 78 + }, + { + "epoch": 0.005428061014154184, + "grad_norm": 2.4444713592529297, + "learning_rate": 2.7129120879120883e-06, + "loss": 1.7266, + "step": 79 + }, + { + "epoch": 0.005496770647244744, + "grad_norm": 2.5306320190429688, + "learning_rate": 2.747252747252747e-06, + "loss": 1.7997, + "step": 80 + }, + { + "epoch": 0.005565480280335303, + "grad_norm": 2.3893392086029053, + "learning_rate": 2.7815934065934065e-06, + "loss": 1.8138, + "step": 81 + }, + { + "epoch": 0.005634189913425862, + "grad_norm": 2.4648749828338623, + "learning_rate": 2.8159340659340662e-06, + "loss": 1.736, + "step": 82 + }, + { + "epoch": 0.0057028995465164215, + "grad_norm": 2.3402011394500732, + "learning_rate": 2.850274725274725e-06, + "loss": 1.7413, + "step": 83 + }, + { + "epoch": 0.005771609179606981, + "grad_norm": 2.384143352508545, + "learning_rate": 2.884615384615385e-06, + "loss": 1.7028, + "step": 84 + }, + { + "epoch": 0.00584031881269754, + "grad_norm": 2.4274344444274902, + "learning_rate": 2.918956043956044e-06, + "loss": 1.7012, + "step": 85 + }, + { + "epoch": 0.005909028445788099, + "grad_norm": 2.3654894828796387, + "learning_rate": 2.9532967032967035e-06, + "loss": 1.6972, + "step": 86 + }, + { + "epoch": 0.005977738078878659, + "grad_norm": 2.293001413345337, + "learning_rate": 2.987637362637363e-06, + "loss": 1.6364, + "step": 87 + }, + { + "epoch": 0.006046447711969218, + "grad_norm": 2.3870837688446045, + "learning_rate": 3.021978021978022e-06, + "loss": 1.6952, + "step": 88 + }, + { + "epoch": 0.006115157345059777, + "grad_norm": 2.3209028244018555, + "learning_rate": 3.0563186813186815e-06, + "loss": 1.6679, + "step": 89 + }, + { + "epoch": 0.0061838669781503364, + "grad_norm": 2.3537614345550537, + "learning_rate": 3.090659340659341e-06, + "loss": 1.6645, + "step": 90 + }, + { + "epoch": 0.006252576611240896, + "grad_norm": 2.4990780353546143, + "learning_rate": 3.125e-06, + "loss": 1.7019, + "step": 91 + }, + { + "epoch": 0.006321286244331455, + "grad_norm": 2.2731547355651855, + "learning_rate": 3.1593406593406595e-06, + "loss": 1.6265, + "step": 92 + }, + { + "epoch": 0.006389995877422014, + "grad_norm": 2.4347567558288574, + "learning_rate": 3.193681318681319e-06, + "loss": 1.6253, + "step": 93 + }, + { + "epoch": 0.0064587055105125735, + "grad_norm": 2.36327862739563, + "learning_rate": 3.2280219780219785e-06, + "loss": 1.6279, + "step": 94 + }, + { + "epoch": 0.006527415143603133, + "grad_norm": 2.1690657138824463, + "learning_rate": 3.262362637362638e-06, + "loss": 1.6204, + "step": 95 + }, + { + "epoch": 0.006596124776693692, + "grad_norm": 2.274386167526245, + "learning_rate": 3.2967032967032968e-06, + "loss": 1.6423, + "step": 96 + }, + { + "epoch": 0.006664834409784251, + "grad_norm": 2.2845969200134277, + "learning_rate": 3.331043956043956e-06, + "loss": 1.5863, + "step": 97 + }, + { + "epoch": 0.006733544042874811, + "grad_norm": 2.2257418632507324, + "learning_rate": 3.3653846153846154e-06, + "loss": 1.6801, + "step": 98 + }, + { + "epoch": 0.006802253675965371, + "grad_norm": 2.1797356605529785, + "learning_rate": 3.3997252747252747e-06, + "loss": 1.6156, + "step": 99 + }, + { + "epoch": 0.00687096330905593, + "grad_norm": 2.4152626991271973, + "learning_rate": 3.4340659340659345e-06, + "loss": 1.5866, + "step": 100 + }, + { + "epoch": 0.006939672942146489, + "grad_norm": 2.4223873615264893, + "learning_rate": 3.468406593406594e-06, + "loss": 1.538, + "step": 101 + }, + { + "epoch": 0.007008382575237049, + "grad_norm": 2.350939989089966, + "learning_rate": 3.502747252747253e-06, + "loss": 1.6015, + "step": 102 + }, + { + "epoch": 0.007077092208327608, + "grad_norm": 2.2195897102355957, + "learning_rate": 3.537087912087912e-06, + "loss": 1.5623, + "step": 103 + }, + { + "epoch": 0.007145801841418167, + "grad_norm": 2.3513755798339844, + "learning_rate": 3.5714285714285714e-06, + "loss": 1.6151, + "step": 104 + }, + { + "epoch": 0.007214511474508726, + "grad_norm": 2.2731540203094482, + "learning_rate": 3.6057692307692307e-06, + "loss": 1.5388, + "step": 105 + }, + { + "epoch": 0.007283221107599286, + "grad_norm": 2.3015456199645996, + "learning_rate": 3.6401098901098904e-06, + "loss": 1.5513, + "step": 106 + }, + { + "epoch": 0.007351930740689845, + "grad_norm": 2.541792154312134, + "learning_rate": 3.6744505494505498e-06, + "loss": 1.5543, + "step": 107 + }, + { + "epoch": 0.007420640373780404, + "grad_norm": 2.2383248805999756, + "learning_rate": 3.708791208791209e-06, + "loss": 1.5505, + "step": 108 + }, + { + "epoch": 0.0074893500068709635, + "grad_norm": 2.3158953189849854, + "learning_rate": 3.7431318681318684e-06, + "loss": 1.5215, + "step": 109 + }, + { + "epoch": 0.007558059639961523, + "grad_norm": 2.24996280670166, + "learning_rate": 3.7774725274725273e-06, + "loss": 1.5197, + "step": 110 + }, + { + "epoch": 0.007626769273052082, + "grad_norm": 2.339571237564087, + "learning_rate": 3.8118131868131875e-06, + "loss": 1.5122, + "step": 111 + }, + { + "epoch": 0.007695478906142641, + "grad_norm": 2.3899333477020264, + "learning_rate": 3.846153846153847e-06, + "loss": 1.5806, + "step": 112 + }, + { + "epoch": 0.007764188539233201, + "grad_norm": 2.2707226276397705, + "learning_rate": 3.880494505494506e-06, + "loss": 1.4875, + "step": 113 + }, + { + "epoch": 0.007832898172323759, + "grad_norm": 2.4128918647766113, + "learning_rate": 3.914835164835165e-06, + "loss": 1.5629, + "step": 114 + }, + { + "epoch": 0.007901607805414318, + "grad_norm": 2.2566280364990234, + "learning_rate": 3.949175824175824e-06, + "loss": 1.492, + "step": 115 + }, + { + "epoch": 0.007970317438504878, + "grad_norm": 2.4572346210479736, + "learning_rate": 3.983516483516483e-06, + "loss": 1.5491, + "step": 116 + }, + { + "epoch": 0.008039027071595437, + "grad_norm": 2.351505994796753, + "learning_rate": 4.017857142857143e-06, + "loss": 1.4869, + "step": 117 + }, + { + "epoch": 0.008107736704685996, + "grad_norm": 2.407135009765625, + "learning_rate": 4.052197802197803e-06, + "loss": 1.4832, + "step": 118 + }, + { + "epoch": 0.008176446337776557, + "grad_norm": 2.2351458072662354, + "learning_rate": 4.086538461538462e-06, + "loss": 1.4449, + "step": 119 + }, + { + "epoch": 0.008245155970867116, + "grad_norm": 2.301877975463867, + "learning_rate": 4.120879120879121e-06, + "loss": 1.435, + "step": 120 + }, + { + "epoch": 0.008313865603957676, + "grad_norm": 2.291778564453125, + "learning_rate": 4.15521978021978e-06, + "loss": 1.4471, + "step": 121 + }, + { + "epoch": 0.008382575237048235, + "grad_norm": 2.071810483932495, + "learning_rate": 4.189560439560439e-06, + "loss": 1.4251, + "step": 122 + }, + { + "epoch": 0.008451284870138794, + "grad_norm": 1.9987494945526123, + "learning_rate": 4.2239010989011e-06, + "loss": 1.4582, + "step": 123 + }, + { + "epoch": 0.008519994503229354, + "grad_norm": 1.9596329927444458, + "learning_rate": 4.258241758241759e-06, + "loss": 1.4143, + "step": 124 + }, + { + "epoch": 0.008588704136319913, + "grad_norm": 1.8985395431518555, + "learning_rate": 4.292582417582418e-06, + "loss": 1.4898, + "step": 125 + }, + { + "epoch": 0.008657413769410472, + "grad_norm": 1.9674025774002075, + "learning_rate": 4.326923076923077e-06, + "loss": 1.4418, + "step": 126 + }, + { + "epoch": 0.008726123402501031, + "grad_norm": 1.9395891427993774, + "learning_rate": 4.361263736263736e-06, + "loss": 1.4103, + "step": 127 + }, + { + "epoch": 0.00879483303559159, + "grad_norm": 1.9848326444625854, + "learning_rate": 4.395604395604396e-06, + "loss": 1.3934, + "step": 128 + }, + { + "epoch": 0.00886354266868215, + "grad_norm": 2.5897297859191895, + "learning_rate": 4.429945054945056e-06, + "loss": 1.3345, + "step": 129 + }, + { + "epoch": 0.00893225230177271, + "grad_norm": 2.467331647872925, + "learning_rate": 4.464285714285715e-06, + "loss": 1.3537, + "step": 130 + }, + { + "epoch": 0.009000961934863268, + "grad_norm": 2.2849247455596924, + "learning_rate": 4.4986263736263735e-06, + "loss": 1.3731, + "step": 131 + }, + { + "epoch": 0.009069671567953828, + "grad_norm": 2.3435752391815186, + "learning_rate": 4.532967032967033e-06, + "loss": 1.4271, + "step": 132 + }, + { + "epoch": 0.009138381201044387, + "grad_norm": 1.9016616344451904, + "learning_rate": 4.567307692307692e-06, + "loss": 1.3948, + "step": 133 + }, + { + "epoch": 0.009207090834134946, + "grad_norm": 1.9436383247375488, + "learning_rate": 4.601648351648352e-06, + "loss": 1.3657, + "step": 134 + }, + { + "epoch": 0.009275800467225506, + "grad_norm": 1.8262126445770264, + "learning_rate": 4.635989010989012e-06, + "loss": 1.3794, + "step": 135 + }, + { + "epoch": 0.009344510100316065, + "grad_norm": 1.9528446197509766, + "learning_rate": 4.6703296703296706e-06, + "loss": 1.4398, + "step": 136 + }, + { + "epoch": 0.009413219733406624, + "grad_norm": 1.868327021598816, + "learning_rate": 4.7046703296703295e-06, + "loss": 1.3879, + "step": 137 + }, + { + "epoch": 0.009481929366497183, + "grad_norm": 1.9784241914749146, + "learning_rate": 4.739010989010989e-06, + "loss": 1.4418, + "step": 138 + }, + { + "epoch": 0.009550638999587743, + "grad_norm": 1.8264296054840088, + "learning_rate": 4.773351648351648e-06, + "loss": 1.4034, + "step": 139 + }, + { + "epoch": 0.009619348632678302, + "grad_norm": 1.8559573888778687, + "learning_rate": 4.807692307692308e-06, + "loss": 1.4029, + "step": 140 + }, + { + "epoch": 0.009688058265768861, + "grad_norm": 1.7852126359939575, + "learning_rate": 4.842032967032968e-06, + "loss": 1.4524, + "step": 141 + }, + { + "epoch": 0.00975676789885942, + "grad_norm": 1.877097725868225, + "learning_rate": 4.8763736263736265e-06, + "loss": 1.3647, + "step": 142 + }, + { + "epoch": 0.00982547753194998, + "grad_norm": 1.8444008827209473, + "learning_rate": 4.910714285714286e-06, + "loss": 1.3675, + "step": 143 + }, + { + "epoch": 0.009894187165040539, + "grad_norm": 1.8507466316223145, + "learning_rate": 4.945054945054945e-06, + "loss": 1.3751, + "step": 144 + }, + { + "epoch": 0.009962896798131098, + "grad_norm": 1.7089701890945435, + "learning_rate": 4.979395604395604e-06, + "loss": 1.3548, + "step": 145 + }, + { + "epoch": 0.010031606431221658, + "grad_norm": 1.8487894535064697, + "learning_rate": 5.013736263736264e-06, + "loss": 1.3926, + "step": 146 + }, + { + "epoch": 0.010100316064312217, + "grad_norm": 1.7713396549224854, + "learning_rate": 5.0480769230769235e-06, + "loss": 1.3107, + "step": 147 + }, + { + "epoch": 0.010169025697402776, + "grad_norm": 1.950148344039917, + "learning_rate": 5.0824175824175824e-06, + "loss": 1.3245, + "step": 148 + }, + { + "epoch": 0.010237735330493335, + "grad_norm": 1.8153810501098633, + "learning_rate": 5.116758241758242e-06, + "loss": 1.3405, + "step": 149 + }, + { + "epoch": 0.010306444963583895, + "grad_norm": 1.7532216310501099, + "learning_rate": 5.151098901098901e-06, + "loss": 1.3503, + "step": 150 + }, + { + "epoch": 0.010375154596674454, + "grad_norm": 1.7673825025558472, + "learning_rate": 5.185439560439561e-06, + "loss": 1.3175, + "step": 151 + }, + { + "epoch": 0.010443864229765013, + "grad_norm": 1.9625242948532104, + "learning_rate": 5.219780219780221e-06, + "loss": 1.3695, + "step": 152 + }, + { + "epoch": 0.010512573862855572, + "grad_norm": 1.8338172435760498, + "learning_rate": 5.2541208791208795e-06, + "loss": 1.3307, + "step": 153 + }, + { + "epoch": 0.010581283495946132, + "grad_norm": 1.8488733768463135, + "learning_rate": 5.288461538461538e-06, + "loss": 1.3603, + "step": 154 + }, + { + "epoch": 0.010649993129036691, + "grad_norm": 1.7387433052062988, + "learning_rate": 5.322802197802198e-06, + "loss": 1.3519, + "step": 155 + }, + { + "epoch": 0.01071870276212725, + "grad_norm": 1.8353899717330933, + "learning_rate": 5.357142857142857e-06, + "loss": 1.3633, + "step": 156 + }, + { + "epoch": 0.01078741239521781, + "grad_norm": 1.8494092226028442, + "learning_rate": 5.391483516483517e-06, + "loss": 1.3698, + "step": 157 + }, + { + "epoch": 0.010856122028308369, + "grad_norm": 1.6814545392990112, + "learning_rate": 5.4258241758241765e-06, + "loss": 1.3351, + "step": 158 + }, + { + "epoch": 0.010924831661398928, + "grad_norm": 1.776051640510559, + "learning_rate": 5.4601648351648354e-06, + "loss": 1.3166, + "step": 159 + }, + { + "epoch": 0.010993541294489487, + "grad_norm": 1.714782953262329, + "learning_rate": 5.494505494505494e-06, + "loss": 1.3375, + "step": 160 + }, + { + "epoch": 0.011062250927580047, + "grad_norm": 1.8178660869598389, + "learning_rate": 5.528846153846154e-06, + "loss": 1.2968, + "step": 161 + }, + { + "epoch": 0.011130960560670606, + "grad_norm": 1.8437119722366333, + "learning_rate": 5.563186813186813e-06, + "loss": 1.3047, + "step": 162 + }, + { + "epoch": 0.011199670193761165, + "grad_norm": 1.729608178138733, + "learning_rate": 5.597527472527473e-06, + "loss": 1.3025, + "step": 163 + }, + { + "epoch": 0.011268379826851724, + "grad_norm": 1.9262492656707764, + "learning_rate": 5.6318681318681325e-06, + "loss": 1.3427, + "step": 164 + }, + { + "epoch": 0.011337089459942284, + "grad_norm": 1.7662638425827026, + "learning_rate": 5.666208791208791e-06, + "loss": 1.2914, + "step": 165 + }, + { + "epoch": 0.011405799093032843, + "grad_norm": 1.7553950548171997, + "learning_rate": 5.70054945054945e-06, + "loss": 1.3458, + "step": 166 + }, + { + "epoch": 0.011474508726123402, + "grad_norm": 1.7810910940170288, + "learning_rate": 5.73489010989011e-06, + "loss": 1.3387, + "step": 167 + }, + { + "epoch": 0.011543218359213962, + "grad_norm": 1.875128149986267, + "learning_rate": 5.76923076923077e-06, + "loss": 1.3337, + "step": 168 + }, + { + "epoch": 0.01161192799230452, + "grad_norm": 1.884323000907898, + "learning_rate": 5.803571428571429e-06, + "loss": 1.3702, + "step": 169 + }, + { + "epoch": 0.01168063762539508, + "grad_norm": 1.7443432807922363, + "learning_rate": 5.837912087912088e-06, + "loss": 1.2652, + "step": 170 + }, + { + "epoch": 0.01174934725848564, + "grad_norm": 1.8351597785949707, + "learning_rate": 5.872252747252747e-06, + "loss": 1.3235, + "step": 171 + }, + { + "epoch": 0.011818056891576199, + "grad_norm": 1.8752599954605103, + "learning_rate": 5.906593406593407e-06, + "loss": 1.3595, + "step": 172 + }, + { + "epoch": 0.011886766524666758, + "grad_norm": 1.7817935943603516, + "learning_rate": 5.940934065934066e-06, + "loss": 1.3247, + "step": 173 + }, + { + "epoch": 0.011955476157757317, + "grad_norm": 1.719714641571045, + "learning_rate": 5.975274725274726e-06, + "loss": 1.2907, + "step": 174 + }, + { + "epoch": 0.012024185790847877, + "grad_norm": 1.8358261585235596, + "learning_rate": 6.0096153846153855e-06, + "loss": 1.2979, + "step": 175 + }, + { + "epoch": 0.012092895423938436, + "grad_norm": 1.8906748294830322, + "learning_rate": 6.043956043956044e-06, + "loss": 1.2685, + "step": 176 + }, + { + "epoch": 0.012161605057028995, + "grad_norm": 1.7719372510910034, + "learning_rate": 6.078296703296703e-06, + "loss": 1.2832, + "step": 177 + }, + { + "epoch": 0.012230314690119554, + "grad_norm": 1.7777578830718994, + "learning_rate": 6.112637362637363e-06, + "loss": 1.3081, + "step": 178 + }, + { + "epoch": 0.012299024323210114, + "grad_norm": 1.744153380393982, + "learning_rate": 6.146978021978022e-06, + "loss": 1.2954, + "step": 179 + }, + { + "epoch": 0.012367733956300673, + "grad_norm": 1.789657711982727, + "learning_rate": 6.181318681318682e-06, + "loss": 1.3028, + "step": 180 + }, + { + "epoch": 0.012436443589391232, + "grad_norm": 1.8172622919082642, + "learning_rate": 6.215659340659341e-06, + "loss": 1.2935, + "step": 181 + }, + { + "epoch": 0.012505153222481791, + "grad_norm": 1.786291480064392, + "learning_rate": 6.25e-06, + "loss": 1.3341, + "step": 182 + }, + { + "epoch": 0.01257386285557235, + "grad_norm": 1.7313944101333618, + "learning_rate": 6.284340659340659e-06, + "loss": 1.2926, + "step": 183 + }, + { + "epoch": 0.01264257248866291, + "grad_norm": 1.815619945526123, + "learning_rate": 6.318681318681319e-06, + "loss": 1.2983, + "step": 184 + }, + { + "epoch": 0.01271128212175347, + "grad_norm": 1.7984061241149902, + "learning_rate": 6.353021978021978e-06, + "loss": 1.3171, + "step": 185 + }, + { + "epoch": 0.012779991754844029, + "grad_norm": 1.8223800659179688, + "learning_rate": 6.387362637362638e-06, + "loss": 1.3016, + "step": 186 + }, + { + "epoch": 0.012848701387934588, + "grad_norm": 1.884972333908081, + "learning_rate": 6.4217032967032965e-06, + "loss": 1.225, + "step": 187 + }, + { + "epoch": 0.012917411021025147, + "grad_norm": 1.7358647584915161, + "learning_rate": 6.456043956043957e-06, + "loss": 1.2578, + "step": 188 + }, + { + "epoch": 0.012986120654115706, + "grad_norm": 1.8495426177978516, + "learning_rate": 6.490384615384616e-06, + "loss": 1.3054, + "step": 189 + }, + { + "epoch": 0.013054830287206266, + "grad_norm": 1.8365671634674072, + "learning_rate": 6.524725274725276e-06, + "loss": 1.3704, + "step": 190 + }, + { + "epoch": 0.013123539920296825, + "grad_norm": 1.7551157474517822, + "learning_rate": 6.559065934065935e-06, + "loss": 1.2577, + "step": 191 + }, + { + "epoch": 0.013192249553387384, + "grad_norm": 2.048203706741333, + "learning_rate": 6.5934065934065935e-06, + "loss": 1.3455, + "step": 192 + }, + { + "epoch": 0.013260959186477943, + "grad_norm": 1.842388391494751, + "learning_rate": 6.627747252747253e-06, + "loss": 1.3461, + "step": 193 + }, + { + "epoch": 0.013329668819568503, + "grad_norm": 1.729679822921753, + "learning_rate": 6.662087912087912e-06, + "loss": 1.2194, + "step": 194 + }, + { + "epoch": 0.013398378452659062, + "grad_norm": 1.7447823286056519, + "learning_rate": 6.696428571428572e-06, + "loss": 1.2711, + "step": 195 + }, + { + "epoch": 0.013467088085749621, + "grad_norm": 1.7975962162017822, + "learning_rate": 6.730769230769231e-06, + "loss": 1.3434, + "step": 196 + }, + { + "epoch": 0.01353579771884018, + "grad_norm": 1.8379902839660645, + "learning_rate": 6.76510989010989e-06, + "loss": 1.2949, + "step": 197 + }, + { + "epoch": 0.013604507351930742, + "grad_norm": 1.8645070791244507, + "learning_rate": 6.7994505494505495e-06, + "loss": 1.3122, + "step": 198 + }, + { + "epoch": 0.0136732169850213, + "grad_norm": 1.805772304534912, + "learning_rate": 6.833791208791208e-06, + "loss": 1.2728, + "step": 199 + }, + { + "epoch": 0.01374192661811186, + "grad_norm": 1.8442202806472778, + "learning_rate": 6.868131868131869e-06, + "loss": 1.2768, + "step": 200 + }, + { + "epoch": 0.01381063625120242, + "grad_norm": 1.8673653602600098, + "learning_rate": 6.902472527472528e-06, + "loss": 1.2525, + "step": 201 + }, + { + "epoch": 0.013879345884292979, + "grad_norm": 1.783299446105957, + "learning_rate": 6.936813186813188e-06, + "loss": 1.2552, + "step": 202 + }, + { + "epoch": 0.013948055517383538, + "grad_norm": 1.699723243713379, + "learning_rate": 6.9711538461538465e-06, + "loss": 1.3176, + "step": 203 + }, + { + "epoch": 0.014016765150474097, + "grad_norm": 1.776279091835022, + "learning_rate": 7.005494505494506e-06, + "loss": 1.2669, + "step": 204 + }, + { + "epoch": 0.014085474783564656, + "grad_norm": 1.808342456817627, + "learning_rate": 7.039835164835165e-06, + "loss": 1.2792, + "step": 205 + }, + { + "epoch": 0.014154184416655216, + "grad_norm": 1.9380438327789307, + "learning_rate": 7.074175824175824e-06, + "loss": 1.3084, + "step": 206 + }, + { + "epoch": 0.014222894049745775, + "grad_norm": 1.791932463645935, + "learning_rate": 7.108516483516484e-06, + "loss": 1.2564, + "step": 207 + }, + { + "epoch": 0.014291603682836334, + "grad_norm": 1.865995168685913, + "learning_rate": 7.142857142857143e-06, + "loss": 1.2537, + "step": 208 + }, + { + "epoch": 0.014360313315926894, + "grad_norm": 1.8747751712799072, + "learning_rate": 7.177197802197802e-06, + "loss": 1.234, + "step": 209 + }, + { + "epoch": 0.014429022949017453, + "grad_norm": 1.6733052730560303, + "learning_rate": 7.211538461538461e-06, + "loss": 1.237, + "step": 210 + }, + { + "epoch": 0.014497732582108012, + "grad_norm": 1.8283827304840088, + "learning_rate": 7.245879120879122e-06, + "loss": 1.2965, + "step": 211 + }, + { + "epoch": 0.014566442215198571, + "grad_norm": 2.0461442470550537, + "learning_rate": 7.280219780219781e-06, + "loss": 1.2776, + "step": 212 + }, + { + "epoch": 0.01463515184828913, + "grad_norm": 1.9120794534683228, + "learning_rate": 7.314560439560441e-06, + "loss": 1.2874, + "step": 213 + }, + { + "epoch": 0.01470386148137969, + "grad_norm": 1.8318010568618774, + "learning_rate": 7.3489010989010995e-06, + "loss": 1.2561, + "step": 214 + }, + { + "epoch": 0.01477257111447025, + "grad_norm": 1.7767771482467651, + "learning_rate": 7.383241758241758e-06, + "loss": 1.2218, + "step": 215 + }, + { + "epoch": 0.014841280747560808, + "grad_norm": 1.725032091140747, + "learning_rate": 7.417582417582418e-06, + "loss": 1.2684, + "step": 216 + }, + { + "epoch": 0.014909990380651368, + "grad_norm": 1.8620021343231201, + "learning_rate": 7.451923076923077e-06, + "loss": 1.2966, + "step": 217 + }, + { + "epoch": 0.014978700013741927, + "grad_norm": 1.807593584060669, + "learning_rate": 7.486263736263737e-06, + "loss": 1.2494, + "step": 218 + }, + { + "epoch": 0.015047409646832486, + "grad_norm": 1.901322364807129, + "learning_rate": 7.520604395604396e-06, + "loss": 1.2957, + "step": 219 + }, + { + "epoch": 0.015116119279923046, + "grad_norm": 1.7582428455352783, + "learning_rate": 7.554945054945055e-06, + "loss": 1.2443, + "step": 220 + }, + { + "epoch": 0.015184828913013605, + "grad_norm": 1.7627885341644287, + "learning_rate": 7.589285714285714e-06, + "loss": 1.2433, + "step": 221 + }, + { + "epoch": 0.015253538546104164, + "grad_norm": 2.0233709812164307, + "learning_rate": 7.623626373626375e-06, + "loss": 1.2507, + "step": 222 + }, + { + "epoch": 0.015322248179194723, + "grad_norm": 1.8181531429290771, + "learning_rate": 7.657967032967034e-06, + "loss": 1.2237, + "step": 223 + }, + { + "epoch": 0.015390957812285283, + "grad_norm": 1.7916814088821411, + "learning_rate": 7.692307692307694e-06, + "loss": 1.2904, + "step": 224 + }, + { + "epoch": 0.015459667445375842, + "grad_norm": 1.6912288665771484, + "learning_rate": 7.726648351648352e-06, + "loss": 1.2788, + "step": 225 + }, + { + "epoch": 0.015528377078466401, + "grad_norm": 1.745093584060669, + "learning_rate": 7.760989010989011e-06, + "loss": 1.2848, + "step": 226 + }, + { + "epoch": 0.01559708671155696, + "grad_norm": 1.7356537580490112, + "learning_rate": 7.795329670329671e-06, + "loss": 1.2961, + "step": 227 + }, + { + "epoch": 0.015665796344647518, + "grad_norm": 1.8689312934875488, + "learning_rate": 7.82967032967033e-06, + "loss": 1.241, + "step": 228 + }, + { + "epoch": 0.015734505977738077, + "grad_norm": 1.926622748374939, + "learning_rate": 7.864010989010989e-06, + "loss": 1.2791, + "step": 229 + }, + { + "epoch": 0.015803215610828637, + "grad_norm": 1.7743676900863647, + "learning_rate": 7.898351648351649e-06, + "loss": 1.2209, + "step": 230 + }, + { + "epoch": 0.015871925243919196, + "grad_norm": 1.7702791690826416, + "learning_rate": 7.932692307692308e-06, + "loss": 1.2796, + "step": 231 + }, + { + "epoch": 0.015940634877009755, + "grad_norm": 1.8199025392532349, + "learning_rate": 7.967032967032966e-06, + "loss": 1.2037, + "step": 232 + }, + { + "epoch": 0.016009344510100314, + "grad_norm": 1.8574044704437256, + "learning_rate": 8.001373626373626e-06, + "loss": 1.2816, + "step": 233 + }, + { + "epoch": 0.016078054143190874, + "grad_norm": 1.77090322971344, + "learning_rate": 8.035714285714286e-06, + "loss": 1.2349, + "step": 234 + }, + { + "epoch": 0.016146763776281433, + "grad_norm": 1.6844888925552368, + "learning_rate": 8.070054945054946e-06, + "loss": 1.208, + "step": 235 + }, + { + "epoch": 0.016215473409371992, + "grad_norm": 1.8097244501113892, + "learning_rate": 8.104395604395605e-06, + "loss": 1.298, + "step": 236 + }, + { + "epoch": 0.01628418304246255, + "grad_norm": 1.836336374282837, + "learning_rate": 8.138736263736264e-06, + "loss": 1.2622, + "step": 237 + }, + { + "epoch": 0.016352892675553114, + "grad_norm": 1.9476301670074463, + "learning_rate": 8.173076923076923e-06, + "loss": 1.284, + "step": 238 + }, + { + "epoch": 0.016421602308643674, + "grad_norm": 1.7375494241714478, + "learning_rate": 8.207417582417583e-06, + "loss": 1.2117, + "step": 239 + }, + { + "epoch": 0.016490311941734233, + "grad_norm": 1.8387936353683472, + "learning_rate": 8.241758241758243e-06, + "loss": 1.2941, + "step": 240 + }, + { + "epoch": 0.016559021574824792, + "grad_norm": 1.71334707736969, + "learning_rate": 8.2760989010989e-06, + "loss": 1.3045, + "step": 241 + }, + { + "epoch": 0.01662773120791535, + "grad_norm": 1.9597705602645874, + "learning_rate": 8.31043956043956e-06, + "loss": 1.2983, + "step": 242 + }, + { + "epoch": 0.01669644084100591, + "grad_norm": 1.7602078914642334, + "learning_rate": 8.34478021978022e-06, + "loss": 1.2751, + "step": 243 + }, + { + "epoch": 0.01676515047409647, + "grad_norm": 1.8743059635162354, + "learning_rate": 8.379120879120878e-06, + "loss": 1.2431, + "step": 244 + }, + { + "epoch": 0.01683386010718703, + "grad_norm": 1.8574447631835938, + "learning_rate": 8.41346153846154e-06, + "loss": 1.2816, + "step": 245 + }, + { + "epoch": 0.01690256974027759, + "grad_norm": 1.6902004480361938, + "learning_rate": 8.4478021978022e-06, + "loss": 1.1887, + "step": 246 + }, + { + "epoch": 0.016971279373368148, + "grad_norm": 2.046617269515991, + "learning_rate": 8.482142857142858e-06, + "loss": 1.2346, + "step": 247 + }, + { + "epoch": 0.017039989006458707, + "grad_norm": 1.8492083549499512, + "learning_rate": 8.516483516483517e-06, + "loss": 1.2345, + "step": 248 + }, + { + "epoch": 0.017108698639549266, + "grad_norm": 1.8801002502441406, + "learning_rate": 8.550824175824177e-06, + "loss": 1.2139, + "step": 249 + }, + { + "epoch": 0.017177408272639826, + "grad_norm": 1.8670740127563477, + "learning_rate": 8.585164835164835e-06, + "loss": 1.3071, + "step": 250 + }, + { + "epoch": 0.017246117905730385, + "grad_norm": 2.1648364067077637, + "learning_rate": 8.619505494505495e-06, + "loss": 1.2647, + "step": 251 + }, + { + "epoch": 0.017314827538820944, + "grad_norm": 1.8854186534881592, + "learning_rate": 8.653846153846155e-06, + "loss": 1.2521, + "step": 252 + }, + { + "epoch": 0.017383537171911503, + "grad_norm": 1.9196877479553223, + "learning_rate": 8.688186813186813e-06, + "loss": 1.2519, + "step": 253 + }, + { + "epoch": 0.017452246805002063, + "grad_norm": 2.0414044857025146, + "learning_rate": 8.722527472527472e-06, + "loss": 1.2677, + "step": 254 + }, + { + "epoch": 0.017520956438092622, + "grad_norm": 2.1110246181488037, + "learning_rate": 8.756868131868132e-06, + "loss": 1.2524, + "step": 255 + }, + { + "epoch": 0.01758966607118318, + "grad_norm": 1.9257311820983887, + "learning_rate": 8.791208791208792e-06, + "loss": 1.2374, + "step": 256 + }, + { + "epoch": 0.01765837570427374, + "grad_norm": 1.8006879091262817, + "learning_rate": 8.825549450549452e-06, + "loss": 1.2457, + "step": 257 + }, + { + "epoch": 0.0177270853373643, + "grad_norm": 1.912639856338501, + "learning_rate": 8.859890109890111e-06, + "loss": 1.269, + "step": 258 + }, + { + "epoch": 0.01779579497045486, + "grad_norm": 1.8918613195419312, + "learning_rate": 8.89423076923077e-06, + "loss": 1.1954, + "step": 259 + }, + { + "epoch": 0.01786450460354542, + "grad_norm": 1.8851152658462524, + "learning_rate": 8.92857142857143e-06, + "loss": 1.2295, + "step": 260 + }, + { + "epoch": 0.017933214236635978, + "grad_norm": 1.964800477027893, + "learning_rate": 8.962912087912089e-06, + "loss": 1.3144, + "step": 261 + }, + { + "epoch": 0.018001923869726537, + "grad_norm": 1.9323924779891968, + "learning_rate": 8.997252747252747e-06, + "loss": 1.2897, + "step": 262 + }, + { + "epoch": 0.018070633502817096, + "grad_norm": 1.8984012603759766, + "learning_rate": 9.031593406593407e-06, + "loss": 1.2486, + "step": 263 + }, + { + "epoch": 0.018139343135907655, + "grad_norm": 1.8609565496444702, + "learning_rate": 9.065934065934067e-06, + "loss": 1.1762, + "step": 264 + }, + { + "epoch": 0.018208052768998215, + "grad_norm": 1.9598054885864258, + "learning_rate": 9.100274725274725e-06, + "loss": 1.1733, + "step": 265 + }, + { + "epoch": 0.018276762402088774, + "grad_norm": 1.752849817276001, + "learning_rate": 9.134615384615384e-06, + "loss": 1.2241, + "step": 266 + }, + { + "epoch": 0.018345472035179333, + "grad_norm": 1.999859094619751, + "learning_rate": 9.168956043956044e-06, + "loss": 1.2115, + "step": 267 + }, + { + "epoch": 0.018414181668269893, + "grad_norm": 2.041778564453125, + "learning_rate": 9.203296703296704e-06, + "loss": 1.2424, + "step": 268 + }, + { + "epoch": 0.018482891301360452, + "grad_norm": 1.8538613319396973, + "learning_rate": 9.237637362637364e-06, + "loss": 1.2484, + "step": 269 + }, + { + "epoch": 0.01855160093445101, + "grad_norm": 1.7494012117385864, + "learning_rate": 9.271978021978023e-06, + "loss": 1.2607, + "step": 270 + }, + { + "epoch": 0.01862031056754157, + "grad_norm": 1.8772673606872559, + "learning_rate": 9.306318681318681e-06, + "loss": 1.2494, + "step": 271 + }, + { + "epoch": 0.01868902020063213, + "grad_norm": 1.8677517175674438, + "learning_rate": 9.340659340659341e-06, + "loss": 1.2064, + "step": 272 + }, + { + "epoch": 0.01875772983372269, + "grad_norm": 1.8427382707595825, + "learning_rate": 9.375000000000001e-06, + "loss": 1.2351, + "step": 273 + }, + { + "epoch": 0.018826439466813248, + "grad_norm": 1.778281331062317, + "learning_rate": 9.409340659340659e-06, + "loss": 1.2408, + "step": 274 + }, + { + "epoch": 0.018895149099903807, + "grad_norm": 1.8262327909469604, + "learning_rate": 9.443681318681319e-06, + "loss": 1.2042, + "step": 275 + }, + { + "epoch": 0.018963858732994367, + "grad_norm": 1.875033974647522, + "learning_rate": 9.478021978021978e-06, + "loss": 1.2058, + "step": 276 + }, + { + "epoch": 0.019032568366084926, + "grad_norm": 1.6990453004837036, + "learning_rate": 9.512362637362638e-06, + "loss": 1.2071, + "step": 277 + }, + { + "epoch": 0.019101277999175485, + "grad_norm": 1.8255594968795776, + "learning_rate": 9.546703296703296e-06, + "loss": 1.1919, + "step": 278 + }, + { + "epoch": 0.019169987632266045, + "grad_norm": 1.7334634065628052, + "learning_rate": 9.581043956043958e-06, + "loss": 1.2269, + "step": 279 + }, + { + "epoch": 0.019238697265356604, + "grad_norm": 1.7248928546905518, + "learning_rate": 9.615384615384616e-06, + "loss": 1.2308, + "step": 280 + }, + { + "epoch": 0.019307406898447163, + "grad_norm": 1.8505756855010986, + "learning_rate": 9.649725274725275e-06, + "loss": 1.2556, + "step": 281 + }, + { + "epoch": 0.019376116531537722, + "grad_norm": 1.8849050998687744, + "learning_rate": 9.684065934065935e-06, + "loss": 1.2186, + "step": 282 + }, + { + "epoch": 0.01944482616462828, + "grad_norm": 1.8148585557937622, + "learning_rate": 9.718406593406593e-06, + "loss": 1.2253, + "step": 283 + }, + { + "epoch": 0.01951353579771884, + "grad_norm": 1.9047515392303467, + "learning_rate": 9.752747252747253e-06, + "loss": 1.2401, + "step": 284 + }, + { + "epoch": 0.0195822454308094, + "grad_norm": 1.7177579402923584, + "learning_rate": 9.787087912087913e-06, + "loss": 1.2543, + "step": 285 + }, + { + "epoch": 0.01965095506389996, + "grad_norm": 1.8326857089996338, + "learning_rate": 9.821428571428573e-06, + "loss": 1.2494, + "step": 286 + }, + { + "epoch": 0.01971966469699052, + "grad_norm": 1.8222200870513916, + "learning_rate": 9.85576923076923e-06, + "loss": 1.2755, + "step": 287 + }, + { + "epoch": 0.019788374330081078, + "grad_norm": 1.7772092819213867, + "learning_rate": 9.89010989010989e-06, + "loss": 1.2246, + "step": 288 + }, + { + "epoch": 0.019857083963171637, + "grad_norm": 1.8634166717529297, + "learning_rate": 9.92445054945055e-06, + "loss": 1.2298, + "step": 289 + }, + { + "epoch": 0.019925793596262197, + "grad_norm": 1.7279053926467896, + "learning_rate": 9.958791208791208e-06, + "loss": 1.2034, + "step": 290 + }, + { + "epoch": 0.019994503229352756, + "grad_norm": 1.8501166105270386, + "learning_rate": 9.99313186813187e-06, + "loss": 1.2231, + "step": 291 + }, + { + "epoch": 0.020063212862443315, + "grad_norm": 2.0626301765441895, + "learning_rate": 1.0027472527472528e-05, + "loss": 1.1975, + "step": 292 + }, + { + "epoch": 0.020131922495533874, + "grad_norm": 1.8844376802444458, + "learning_rate": 1.0061813186813187e-05, + "loss": 1.1887, + "step": 293 + }, + { + "epoch": 0.020200632128624434, + "grad_norm": 1.730533242225647, + "learning_rate": 1.0096153846153847e-05, + "loss": 1.1784, + "step": 294 + }, + { + "epoch": 0.020269341761714993, + "grad_norm": 1.8917726278305054, + "learning_rate": 1.0130494505494507e-05, + "loss": 1.1778, + "step": 295 + }, + { + "epoch": 0.020338051394805552, + "grad_norm": 1.8432811498641968, + "learning_rate": 1.0164835164835165e-05, + "loss": 1.2488, + "step": 296 + }, + { + "epoch": 0.02040676102789611, + "grad_norm": 1.8013173341751099, + "learning_rate": 1.0199175824175825e-05, + "loss": 1.1407, + "step": 297 + }, + { + "epoch": 0.02047547066098667, + "grad_norm": 1.8322722911834717, + "learning_rate": 1.0233516483516484e-05, + "loss": 1.2415, + "step": 298 + }, + { + "epoch": 0.02054418029407723, + "grad_norm": 1.7555615901947021, + "learning_rate": 1.0267857142857142e-05, + "loss": 1.1632, + "step": 299 + }, + { + "epoch": 0.02061288992716779, + "grad_norm": 2.021744728088379, + "learning_rate": 1.0302197802197802e-05, + "loss": 1.2568, + "step": 300 + }, + { + "epoch": 0.02068159956025835, + "grad_norm": 1.7449944019317627, + "learning_rate": 1.0336538461538462e-05, + "loss": 1.2325, + "step": 301 + }, + { + "epoch": 0.020750309193348908, + "grad_norm": 1.8534226417541504, + "learning_rate": 1.0370879120879122e-05, + "loss": 1.2655, + "step": 302 + }, + { + "epoch": 0.020819018826439467, + "grad_norm": 1.8116271495819092, + "learning_rate": 1.0405219780219781e-05, + "loss": 1.2343, + "step": 303 + }, + { + "epoch": 0.020887728459530026, + "grad_norm": 1.8660705089569092, + "learning_rate": 1.0439560439560441e-05, + "loss": 1.1751, + "step": 304 + }, + { + "epoch": 0.020956438092620586, + "grad_norm": 1.8714817762374878, + "learning_rate": 1.04739010989011e-05, + "loss": 1.2103, + "step": 305 + }, + { + "epoch": 0.021025147725711145, + "grad_norm": 1.9253548383712769, + "learning_rate": 1.0508241758241759e-05, + "loss": 1.2089, + "step": 306 + }, + { + "epoch": 0.021093857358801704, + "grad_norm": 1.805106520652771, + "learning_rate": 1.0542582417582419e-05, + "loss": 1.1798, + "step": 307 + }, + { + "epoch": 0.021162566991892263, + "grad_norm": 1.7228858470916748, + "learning_rate": 1.0576923076923077e-05, + "loss": 1.2012, + "step": 308 + }, + { + "epoch": 0.021231276624982823, + "grad_norm": 1.8198844194412231, + "learning_rate": 1.0611263736263737e-05, + "loss": 1.2146, + "step": 309 + }, + { + "epoch": 0.021299986258073382, + "grad_norm": 1.871071219444275, + "learning_rate": 1.0645604395604396e-05, + "loss": 1.2307, + "step": 310 + }, + { + "epoch": 0.02136869589116394, + "grad_norm": 1.8119062185287476, + "learning_rate": 1.0679945054945054e-05, + "loss": 1.1919, + "step": 311 + }, + { + "epoch": 0.0214374055242545, + "grad_norm": 1.9573986530303955, + "learning_rate": 1.0714285714285714e-05, + "loss": 1.1784, + "step": 312 + }, + { + "epoch": 0.02150611515734506, + "grad_norm": 1.754317045211792, + "learning_rate": 1.0748626373626376e-05, + "loss": 1.1841, + "step": 313 + }, + { + "epoch": 0.02157482479043562, + "grad_norm": 1.7678745985031128, + "learning_rate": 1.0782967032967034e-05, + "loss": 1.1457, + "step": 314 + }, + { + "epoch": 0.02164353442352618, + "grad_norm": 1.8167684078216553, + "learning_rate": 1.0817307692307693e-05, + "loss": 1.1853, + "step": 315 + }, + { + "epoch": 0.021712244056616738, + "grad_norm": 2.001152992248535, + "learning_rate": 1.0851648351648353e-05, + "loss": 1.2262, + "step": 316 + }, + { + "epoch": 0.021780953689707297, + "grad_norm": 1.7427979707717896, + "learning_rate": 1.0885989010989011e-05, + "loss": 1.2244, + "step": 317 + }, + { + "epoch": 0.021849663322797856, + "grad_norm": 1.7057188749313354, + "learning_rate": 1.0920329670329671e-05, + "loss": 1.1894, + "step": 318 + }, + { + "epoch": 0.021918372955888415, + "grad_norm": 1.7295200824737549, + "learning_rate": 1.095467032967033e-05, + "loss": 1.1737, + "step": 319 + }, + { + "epoch": 0.021987082588978975, + "grad_norm": 1.7670615911483765, + "learning_rate": 1.0989010989010989e-05, + "loss": 1.2619, + "step": 320 + }, + { + "epoch": 0.022055792222069534, + "grad_norm": 1.8475568294525146, + "learning_rate": 1.1023351648351648e-05, + "loss": 1.2017, + "step": 321 + }, + { + "epoch": 0.022124501855160093, + "grad_norm": 1.7367092370986938, + "learning_rate": 1.1057692307692308e-05, + "loss": 1.1847, + "step": 322 + }, + { + "epoch": 0.022193211488250653, + "grad_norm": 1.7827174663543701, + "learning_rate": 1.1092032967032966e-05, + "loss": 1.2143, + "step": 323 + }, + { + "epoch": 0.022261921121341212, + "grad_norm": 1.8809077739715576, + "learning_rate": 1.1126373626373626e-05, + "loss": 1.2584, + "step": 324 + }, + { + "epoch": 0.02233063075443177, + "grad_norm": 1.8600494861602783, + "learning_rate": 1.1160714285714287e-05, + "loss": 1.2311, + "step": 325 + }, + { + "epoch": 0.02239934038752233, + "grad_norm": 1.6830791234970093, + "learning_rate": 1.1195054945054945e-05, + "loss": 1.2496, + "step": 326 + }, + { + "epoch": 0.02246805002061289, + "grad_norm": 1.662007212638855, + "learning_rate": 1.1229395604395605e-05, + "loss": 1.2077, + "step": 327 + }, + { + "epoch": 0.02253675965370345, + "grad_norm": 1.7917267084121704, + "learning_rate": 1.1263736263736265e-05, + "loss": 1.2083, + "step": 328 + }, + { + "epoch": 0.022605469286794008, + "grad_norm": 1.7548619508743286, + "learning_rate": 1.1298076923076923e-05, + "loss": 1.1923, + "step": 329 + }, + { + "epoch": 0.022674178919884568, + "grad_norm": 1.8426109552383423, + "learning_rate": 1.1332417582417583e-05, + "loss": 1.1896, + "step": 330 + }, + { + "epoch": 0.022742888552975127, + "grad_norm": 1.9298467636108398, + "learning_rate": 1.1366758241758242e-05, + "loss": 1.1604, + "step": 331 + }, + { + "epoch": 0.022811598186065686, + "grad_norm": 1.9566547870635986, + "learning_rate": 1.14010989010989e-05, + "loss": 1.1904, + "step": 332 + }, + { + "epoch": 0.022880307819156245, + "grad_norm": 1.8469277620315552, + "learning_rate": 1.143543956043956e-05, + "loss": 1.1808, + "step": 333 + }, + { + "epoch": 0.022949017452246805, + "grad_norm": 2.0074307918548584, + "learning_rate": 1.146978021978022e-05, + "loss": 1.2141, + "step": 334 + }, + { + "epoch": 0.023017727085337364, + "grad_norm": 1.857134222984314, + "learning_rate": 1.150412087912088e-05, + "loss": 1.231, + "step": 335 + }, + { + "epoch": 0.023086436718427923, + "grad_norm": 1.906341791152954, + "learning_rate": 1.153846153846154e-05, + "loss": 1.1558, + "step": 336 + }, + { + "epoch": 0.023155146351518482, + "grad_norm": 1.7417521476745605, + "learning_rate": 1.15728021978022e-05, + "loss": 1.1244, + "step": 337 + }, + { + "epoch": 0.02322385598460904, + "grad_norm": 1.8375133275985718, + "learning_rate": 1.1607142857142857e-05, + "loss": 1.2057, + "step": 338 + }, + { + "epoch": 0.0232925656176996, + "grad_norm": 1.7321984767913818, + "learning_rate": 1.1641483516483517e-05, + "loss": 1.1893, + "step": 339 + }, + { + "epoch": 0.02336127525079016, + "grad_norm": 1.8157168626785278, + "learning_rate": 1.1675824175824177e-05, + "loss": 1.1542, + "step": 340 + }, + { + "epoch": 0.02342998488388072, + "grad_norm": 1.7845510244369507, + "learning_rate": 1.1710164835164837e-05, + "loss": 1.1912, + "step": 341 + }, + { + "epoch": 0.02349869451697128, + "grad_norm": 1.8669679164886475, + "learning_rate": 1.1744505494505495e-05, + "loss": 1.1883, + "step": 342 + }, + { + "epoch": 0.023567404150061838, + "grad_norm": 1.7617942094802856, + "learning_rate": 1.1778846153846154e-05, + "loss": 1.1774, + "step": 343 + }, + { + "epoch": 0.023636113783152397, + "grad_norm": 1.8350175619125366, + "learning_rate": 1.1813186813186814e-05, + "loss": 1.1903, + "step": 344 + }, + { + "epoch": 0.023704823416242957, + "grad_norm": 1.9152894020080566, + "learning_rate": 1.1847527472527472e-05, + "loss": 1.1903, + "step": 345 + }, + { + "epoch": 0.023773533049333516, + "grad_norm": 1.799851894378662, + "learning_rate": 1.1881868131868132e-05, + "loss": 1.1737, + "step": 346 + }, + { + "epoch": 0.023842242682424075, + "grad_norm": 1.7395038604736328, + "learning_rate": 1.1916208791208792e-05, + "loss": 1.224, + "step": 347 + }, + { + "epoch": 0.023910952315514634, + "grad_norm": 1.7343039512634277, + "learning_rate": 1.1950549450549451e-05, + "loss": 1.203, + "step": 348 + }, + { + "epoch": 0.023979661948605194, + "grad_norm": 1.8492523431777954, + "learning_rate": 1.1984890109890111e-05, + "loss": 1.1985, + "step": 349 + }, + { + "epoch": 0.024048371581695753, + "grad_norm": 1.8523385524749756, + "learning_rate": 1.2019230769230771e-05, + "loss": 1.1841, + "step": 350 + }, + { + "epoch": 0.024117081214786312, + "grad_norm": 1.7179025411605835, + "learning_rate": 1.2053571428571429e-05, + "loss": 1.1506, + "step": 351 + }, + { + "epoch": 0.02418579084787687, + "grad_norm": 1.682279109954834, + "learning_rate": 1.2087912087912089e-05, + "loss": 1.1157, + "step": 352 + }, + { + "epoch": 0.02425450048096743, + "grad_norm": 1.7604466676712036, + "learning_rate": 1.2122252747252748e-05, + "loss": 1.1788, + "step": 353 + }, + { + "epoch": 0.02432321011405799, + "grad_norm": 1.8254975080490112, + "learning_rate": 1.2156593406593407e-05, + "loss": 1.1973, + "step": 354 + }, + { + "epoch": 0.02439191974714855, + "grad_norm": 1.7576887607574463, + "learning_rate": 1.2190934065934066e-05, + "loss": 1.1646, + "step": 355 + }, + { + "epoch": 0.02446062938023911, + "grad_norm": 1.7221802473068237, + "learning_rate": 1.2225274725274726e-05, + "loss": 1.2016, + "step": 356 + }, + { + "epoch": 0.024529339013329668, + "grad_norm": 1.8835794925689697, + "learning_rate": 1.2259615384615384e-05, + "loss": 1.1952, + "step": 357 + }, + { + "epoch": 0.024598048646420227, + "grad_norm": 1.6638824939727783, + "learning_rate": 1.2293956043956044e-05, + "loss": 1.2074, + "step": 358 + }, + { + "epoch": 0.024666758279510786, + "grad_norm": 1.684415340423584, + "learning_rate": 1.2328296703296705e-05, + "loss": 1.2205, + "step": 359 + }, + { + "epoch": 0.024735467912601346, + "grad_norm": 1.7900677919387817, + "learning_rate": 1.2362637362637363e-05, + "loss": 1.1849, + "step": 360 + }, + { + "epoch": 0.024804177545691905, + "grad_norm": 1.897916555404663, + "learning_rate": 1.2396978021978023e-05, + "loss": 1.2029, + "step": 361 + }, + { + "epoch": 0.024872887178782464, + "grad_norm": 1.7303158044815063, + "learning_rate": 1.2431318681318683e-05, + "loss": 1.199, + "step": 362 + }, + { + "epoch": 0.024941596811873024, + "grad_norm": 1.7396302223205566, + "learning_rate": 1.2465659340659341e-05, + "loss": 1.1786, + "step": 363 + }, + { + "epoch": 0.025010306444963583, + "grad_norm": 1.7236690521240234, + "learning_rate": 1.25e-05, + "loss": 1.1686, + "step": 364 + }, + { + "epoch": 0.025079016078054142, + "grad_norm": 1.790769100189209, + "learning_rate": 1.253434065934066e-05, + "loss": 1.181, + "step": 365 + }, + { + "epoch": 0.0251477257111447, + "grad_norm": 1.752195119857788, + "learning_rate": 1.2568681318681318e-05, + "loss": 1.2312, + "step": 366 + }, + { + "epoch": 0.02521643534423526, + "grad_norm": 1.849685788154602, + "learning_rate": 1.2603021978021978e-05, + "loss": 1.1762, + "step": 367 + }, + { + "epoch": 0.02528514497732582, + "grad_norm": 1.8445295095443726, + "learning_rate": 1.2637362637362638e-05, + "loss": 1.2112, + "step": 368 + }, + { + "epoch": 0.02535385461041638, + "grad_norm": 1.6746313571929932, + "learning_rate": 1.2671703296703296e-05, + "loss": 1.1647, + "step": 369 + }, + { + "epoch": 0.02542256424350694, + "grad_norm": 1.776367425918579, + "learning_rate": 1.2706043956043956e-05, + "loss": 1.2223, + "step": 370 + }, + { + "epoch": 0.025491273876597498, + "grad_norm": 1.783318281173706, + "learning_rate": 1.2740384615384615e-05, + "loss": 1.2163, + "step": 371 + }, + { + "epoch": 0.025559983509688057, + "grad_norm": 1.6832436323165894, + "learning_rate": 1.2774725274725275e-05, + "loss": 1.1984, + "step": 372 + }, + { + "epoch": 0.025628693142778616, + "grad_norm": 1.755056619644165, + "learning_rate": 1.2809065934065933e-05, + "loss": 1.1882, + "step": 373 + }, + { + "epoch": 0.025697402775869176, + "grad_norm": 1.913102149963379, + "learning_rate": 1.2843406593406593e-05, + "loss": 1.1917, + "step": 374 + }, + { + "epoch": 0.025766112408959735, + "grad_norm": 1.8405261039733887, + "learning_rate": 1.2877747252747253e-05, + "loss": 1.1697, + "step": 375 + }, + { + "epoch": 0.025834822042050294, + "grad_norm": 1.6997301578521729, + "learning_rate": 1.2912087912087914e-05, + "loss": 1.2135, + "step": 376 + }, + { + "epoch": 0.025903531675140853, + "grad_norm": 1.9041036367416382, + "learning_rate": 1.2946428571428574e-05, + "loss": 1.146, + "step": 377 + }, + { + "epoch": 0.025972241308231413, + "grad_norm": 1.900850772857666, + "learning_rate": 1.2980769230769232e-05, + "loss": 1.2076, + "step": 378 + }, + { + "epoch": 0.026040950941321972, + "grad_norm": 1.8306934833526611, + "learning_rate": 1.3015109890109892e-05, + "loss": 1.2269, + "step": 379 + }, + { + "epoch": 0.02610966057441253, + "grad_norm": 1.8734626770019531, + "learning_rate": 1.3049450549450551e-05, + "loss": 1.2295, + "step": 380 + }, + { + "epoch": 0.02617837020750309, + "grad_norm": 1.7055972814559937, + "learning_rate": 1.308379120879121e-05, + "loss": 1.1682, + "step": 381 + }, + { + "epoch": 0.02624707984059365, + "grad_norm": 1.8813316822052002, + "learning_rate": 1.311813186813187e-05, + "loss": 1.2098, + "step": 382 + }, + { + "epoch": 0.02631578947368421, + "grad_norm": 1.8031595945358276, + "learning_rate": 1.3152472527472529e-05, + "loss": 1.176, + "step": 383 + }, + { + "epoch": 0.02638449910677477, + "grad_norm": 1.7897268533706665, + "learning_rate": 1.3186813186813187e-05, + "loss": 1.1607, + "step": 384 + }, + { + "epoch": 0.026453208739865328, + "grad_norm": 1.7428929805755615, + "learning_rate": 1.3221153846153847e-05, + "loss": 1.1805, + "step": 385 + }, + { + "epoch": 0.026521918372955887, + "grad_norm": 1.7763307094573975, + "learning_rate": 1.3255494505494507e-05, + "loss": 1.1896, + "step": 386 + }, + { + "epoch": 0.026590628006046446, + "grad_norm": 2.0113863945007324, + "learning_rate": 1.3289835164835165e-05, + "loss": 1.2106, + "step": 387 + }, + { + "epoch": 0.026659337639137005, + "grad_norm": 1.7478519678115845, + "learning_rate": 1.3324175824175824e-05, + "loss": 1.1249, + "step": 388 + }, + { + "epoch": 0.026728047272227565, + "grad_norm": 1.8563778400421143, + "learning_rate": 1.3358516483516484e-05, + "loss": 1.2306, + "step": 389 + }, + { + "epoch": 0.026796756905318124, + "grad_norm": 1.731191873550415, + "learning_rate": 1.3392857142857144e-05, + "loss": 1.1987, + "step": 390 + }, + { + "epoch": 0.026865466538408683, + "grad_norm": 1.642269253730774, + "learning_rate": 1.3427197802197802e-05, + "loss": 1.1517, + "step": 391 + }, + { + "epoch": 0.026934176171499243, + "grad_norm": 1.7173988819122314, + "learning_rate": 1.3461538461538462e-05, + "loss": 1.1805, + "step": 392 + }, + { + "epoch": 0.027002885804589802, + "grad_norm": 1.785833477973938, + "learning_rate": 1.3495879120879121e-05, + "loss": 1.1684, + "step": 393 + }, + { + "epoch": 0.02707159543768036, + "grad_norm": 1.7967194318771362, + "learning_rate": 1.353021978021978e-05, + "loss": 1.1489, + "step": 394 + }, + { + "epoch": 0.02714030507077092, + "grad_norm": 1.6991841793060303, + "learning_rate": 1.356456043956044e-05, + "loss": 1.1133, + "step": 395 + }, + { + "epoch": 0.027209014703861483, + "grad_norm": 1.874516487121582, + "learning_rate": 1.3598901098901099e-05, + "loss": 1.1382, + "step": 396 + }, + { + "epoch": 0.027277724336952042, + "grad_norm": 1.7772109508514404, + "learning_rate": 1.3633241758241757e-05, + "loss": 1.1943, + "step": 397 + }, + { + "epoch": 0.0273464339700426, + "grad_norm": 1.8556393384933472, + "learning_rate": 1.3667582417582417e-05, + "loss": 1.2484, + "step": 398 + }, + { + "epoch": 0.02741514360313316, + "grad_norm": 1.7514337301254272, + "learning_rate": 1.3701923076923078e-05, + "loss": 1.1326, + "step": 399 + }, + { + "epoch": 0.02748385323622372, + "grad_norm": 1.754815936088562, + "learning_rate": 1.3736263736263738e-05, + "loss": 1.129, + "step": 400 + }, + { + "epoch": 0.02755256286931428, + "grad_norm": 1.7418264150619507, + "learning_rate": 1.3770604395604398e-05, + "loss": 1.2041, + "step": 401 + }, + { + "epoch": 0.02762127250240484, + "grad_norm": 1.9370641708374023, + "learning_rate": 1.3804945054945056e-05, + "loss": 1.154, + "step": 402 + }, + { + "epoch": 0.027689982135495398, + "grad_norm": 1.7114958763122559, + "learning_rate": 1.3839285714285715e-05, + "loss": 1.1845, + "step": 403 + }, + { + "epoch": 0.027758691768585957, + "grad_norm": 1.8793137073516846, + "learning_rate": 1.3873626373626375e-05, + "loss": 1.1751, + "step": 404 + }, + { + "epoch": 0.027827401401676517, + "grad_norm": 1.6935629844665527, + "learning_rate": 1.3907967032967035e-05, + "loss": 1.1538, + "step": 405 + }, + { + "epoch": 0.027896111034767076, + "grad_norm": 1.9229997396469116, + "learning_rate": 1.3942307692307693e-05, + "loss": 1.1987, + "step": 406 + }, + { + "epoch": 0.027964820667857635, + "grad_norm": 1.923025369644165, + "learning_rate": 1.3976648351648353e-05, + "loss": 1.1887, + "step": 407 + }, + { + "epoch": 0.028033530300948194, + "grad_norm": 1.7327035665512085, + "learning_rate": 1.4010989010989013e-05, + "loss": 1.1121, + "step": 408 + }, + { + "epoch": 0.028102239934038754, + "grad_norm": 1.7460893392562866, + "learning_rate": 1.404532967032967e-05, + "loss": 1.1529, + "step": 409 + }, + { + "epoch": 0.028170949567129313, + "grad_norm": 1.8924154043197632, + "learning_rate": 1.407967032967033e-05, + "loss": 1.1601, + "step": 410 + }, + { + "epoch": 0.028239659200219872, + "grad_norm": 1.7994405031204224, + "learning_rate": 1.411401098901099e-05, + "loss": 1.1898, + "step": 411 + }, + { + "epoch": 0.02830836883331043, + "grad_norm": 1.7748185396194458, + "learning_rate": 1.4148351648351648e-05, + "loss": 1.2062, + "step": 412 + }, + { + "epoch": 0.02837707846640099, + "grad_norm": 1.8718863725662231, + "learning_rate": 1.4182692307692308e-05, + "loss": 1.149, + "step": 413 + }, + { + "epoch": 0.02844578809949155, + "grad_norm": 1.7027894258499146, + "learning_rate": 1.4217032967032968e-05, + "loss": 1.1757, + "step": 414 + }, + { + "epoch": 0.02851449773258211, + "grad_norm": 1.826253890991211, + "learning_rate": 1.4251373626373626e-05, + "loss": 1.2326, + "step": 415 + }, + { + "epoch": 0.02858320736567267, + "grad_norm": 1.6773431301116943, + "learning_rate": 1.4285714285714285e-05, + "loss": 1.1911, + "step": 416 + }, + { + "epoch": 0.028651916998763228, + "grad_norm": 1.8553942441940308, + "learning_rate": 1.4320054945054945e-05, + "loss": 1.2104, + "step": 417 + }, + { + "epoch": 0.028720626631853787, + "grad_norm": 1.8479082584381104, + "learning_rate": 1.4354395604395603e-05, + "loss": 1.1884, + "step": 418 + }, + { + "epoch": 0.028789336264944346, + "grad_norm": 1.781855583190918, + "learning_rate": 1.4388736263736263e-05, + "loss": 1.2427, + "step": 419 + }, + { + "epoch": 0.028858045898034906, + "grad_norm": 1.7435884475708008, + "learning_rate": 1.4423076923076923e-05, + "loss": 1.2562, + "step": 420 + }, + { + "epoch": 0.028926755531125465, + "grad_norm": 1.7677274942398071, + "learning_rate": 1.4457417582417584e-05, + "loss": 1.1504, + "step": 421 + }, + { + "epoch": 0.028995465164216024, + "grad_norm": 1.8199288845062256, + "learning_rate": 1.4491758241758244e-05, + "loss": 1.1939, + "step": 422 + }, + { + "epoch": 0.029064174797306584, + "grad_norm": 1.7487211227416992, + "learning_rate": 1.4526098901098904e-05, + "loss": 1.2137, + "step": 423 + }, + { + "epoch": 0.029132884430397143, + "grad_norm": 1.7172019481658936, + "learning_rate": 1.4560439560439562e-05, + "loss": 1.1819, + "step": 424 + }, + { + "epoch": 0.029201594063487702, + "grad_norm": 1.7076219320297241, + "learning_rate": 1.4594780219780221e-05, + "loss": 1.1657, + "step": 425 + }, + { + "epoch": 0.02927030369657826, + "grad_norm": 1.7704896926879883, + "learning_rate": 1.4629120879120881e-05, + "loss": 1.1455, + "step": 426 + }, + { + "epoch": 0.02933901332966882, + "grad_norm": 1.6982651948928833, + "learning_rate": 1.466346153846154e-05, + "loss": 1.2136, + "step": 427 + }, + { + "epoch": 0.02940772296275938, + "grad_norm": 1.7250289916992188, + "learning_rate": 1.4697802197802199e-05, + "loss": 1.1527, + "step": 428 + }, + { + "epoch": 0.02947643259584994, + "grad_norm": 1.8329815864562988, + "learning_rate": 1.4732142857142859e-05, + "loss": 1.2187, + "step": 429 + }, + { + "epoch": 0.0295451422289405, + "grad_norm": 1.6966925859451294, + "learning_rate": 1.4766483516483517e-05, + "loss": 1.1629, + "step": 430 + }, + { + "epoch": 0.029613851862031058, + "grad_norm": 1.6227772235870361, + "learning_rate": 1.4800824175824177e-05, + "loss": 1.1771, + "step": 431 + }, + { + "epoch": 0.029682561495121617, + "grad_norm": 1.6905862092971802, + "learning_rate": 1.4835164835164836e-05, + "loss": 1.1699, + "step": 432 + }, + { + "epoch": 0.029751271128212176, + "grad_norm": 1.708912968635559, + "learning_rate": 1.4869505494505494e-05, + "loss": 1.1593, + "step": 433 + }, + { + "epoch": 0.029819980761302736, + "grad_norm": 1.813547968864441, + "learning_rate": 1.4903846153846154e-05, + "loss": 1.1643, + "step": 434 + }, + { + "epoch": 0.029888690394393295, + "grad_norm": 1.9627575874328613, + "learning_rate": 1.4938186813186814e-05, + "loss": 1.1677, + "step": 435 + }, + { + "epoch": 0.029957400027483854, + "grad_norm": 1.7052664756774902, + "learning_rate": 1.4972527472527474e-05, + "loss": 1.193, + "step": 436 + }, + { + "epoch": 0.030026109660574413, + "grad_norm": 1.97419273853302, + "learning_rate": 1.5006868131868132e-05, + "loss": 1.1591, + "step": 437 + }, + { + "epoch": 0.030094819293664973, + "grad_norm": 1.760110855102539, + "learning_rate": 1.5041208791208791e-05, + "loss": 1.209, + "step": 438 + }, + { + "epoch": 0.030163528926755532, + "grad_norm": 1.658682942390442, + "learning_rate": 1.5075549450549451e-05, + "loss": 1.1586, + "step": 439 + }, + { + "epoch": 0.03023223855984609, + "grad_norm": 1.7949769496917725, + "learning_rate": 1.510989010989011e-05, + "loss": 1.1841, + "step": 440 + }, + { + "epoch": 0.03030094819293665, + "grad_norm": 1.7680435180664062, + "learning_rate": 1.5144230769230769e-05, + "loss": 1.1188, + "step": 441 + }, + { + "epoch": 0.03036965782602721, + "grad_norm": 1.6473820209503174, + "learning_rate": 1.5178571428571429e-05, + "loss": 1.1651, + "step": 442 + }, + { + "epoch": 0.03043836745911777, + "grad_norm": 1.6909507513046265, + "learning_rate": 1.5212912087912087e-05, + "loss": 1.1127, + "step": 443 + }, + { + "epoch": 0.03050707709220833, + "grad_norm": 1.6468228101730347, + "learning_rate": 1.524725274725275e-05, + "loss": 1.1326, + "step": 444 + }, + { + "epoch": 0.030575786725298888, + "grad_norm": 1.72689688205719, + "learning_rate": 1.528159340659341e-05, + "loss": 1.1243, + "step": 445 + }, + { + "epoch": 0.030644496358389447, + "grad_norm": 1.6930038928985596, + "learning_rate": 1.5315934065934068e-05, + "loss": 1.1497, + "step": 446 + }, + { + "epoch": 0.030713205991480006, + "grad_norm": 1.6721019744873047, + "learning_rate": 1.5350274725274726e-05, + "loss": 1.1406, + "step": 447 + }, + { + "epoch": 0.030781915624570565, + "grad_norm": 1.6625045537948608, + "learning_rate": 1.5384615384615387e-05, + "loss": 1.1625, + "step": 448 + }, + { + "epoch": 0.030850625257661125, + "grad_norm": 1.7111830711364746, + "learning_rate": 1.5418956043956045e-05, + "loss": 1.1466, + "step": 449 + }, + { + "epoch": 0.030919334890751684, + "grad_norm": 1.6702873706817627, + "learning_rate": 1.5453296703296703e-05, + "loss": 1.2002, + "step": 450 + }, + { + "epoch": 0.030988044523842243, + "grad_norm": 1.8359884023666382, + "learning_rate": 1.5487637362637365e-05, + "loss": 1.182, + "step": 451 + }, + { + "epoch": 0.031056754156932802, + "grad_norm": 1.8337528705596924, + "learning_rate": 1.5521978021978023e-05, + "loss": 1.196, + "step": 452 + }, + { + "epoch": 0.031125463790023362, + "grad_norm": 1.7453434467315674, + "learning_rate": 1.555631868131868e-05, + "loss": 1.1364, + "step": 453 + }, + { + "epoch": 0.03119417342311392, + "grad_norm": 1.6364496946334839, + "learning_rate": 1.5590659340659342e-05, + "loss": 1.1706, + "step": 454 + }, + { + "epoch": 0.03126288305620448, + "grad_norm": 1.6656005382537842, + "learning_rate": 1.5625e-05, + "loss": 1.1314, + "step": 455 + }, + { + "epoch": 0.031331592689295036, + "grad_norm": 1.8373081684112549, + "learning_rate": 1.565934065934066e-05, + "loss": 1.19, + "step": 456 + }, + { + "epoch": 0.031400302322385595, + "grad_norm": 1.6809718608856201, + "learning_rate": 1.569368131868132e-05, + "loss": 1.1735, + "step": 457 + }, + { + "epoch": 0.031469011955476155, + "grad_norm": 1.7215808629989624, + "learning_rate": 1.5728021978021978e-05, + "loss": 1.173, + "step": 458 + }, + { + "epoch": 0.031537721588566714, + "grad_norm": 1.5807435512542725, + "learning_rate": 1.576236263736264e-05, + "loss": 1.1216, + "step": 459 + }, + { + "epoch": 0.03160643122165727, + "grad_norm": 1.6677583456039429, + "learning_rate": 1.5796703296703297e-05, + "loss": 1.1252, + "step": 460 + }, + { + "epoch": 0.03167514085474783, + "grad_norm": 1.7506805658340454, + "learning_rate": 1.5831043956043955e-05, + "loss": 1.1899, + "step": 461 + }, + { + "epoch": 0.03174385048783839, + "grad_norm": 1.7662265300750732, + "learning_rate": 1.5865384615384617e-05, + "loss": 1.1668, + "step": 462 + }, + { + "epoch": 0.03181256012092895, + "grad_norm": 1.6827303171157837, + "learning_rate": 1.5899725274725275e-05, + "loss": 1.1956, + "step": 463 + }, + { + "epoch": 0.03188126975401951, + "grad_norm": 1.748664140701294, + "learning_rate": 1.5934065934065933e-05, + "loss": 1.1483, + "step": 464 + }, + { + "epoch": 0.03194997938711007, + "grad_norm": 1.7393075227737427, + "learning_rate": 1.5968406593406594e-05, + "loss": 1.1331, + "step": 465 + }, + { + "epoch": 0.03201868902020063, + "grad_norm": 1.7697886228561401, + "learning_rate": 1.6002747252747252e-05, + "loss": 1.1412, + "step": 466 + }, + { + "epoch": 0.03208739865329119, + "grad_norm": 1.787345051765442, + "learning_rate": 1.6037087912087914e-05, + "loss": 1.1623, + "step": 467 + }, + { + "epoch": 0.03215610828638175, + "grad_norm": 1.6121190786361694, + "learning_rate": 1.6071428571428572e-05, + "loss": 1.09, + "step": 468 + }, + { + "epoch": 0.03222481791947231, + "grad_norm": 1.6945720911026, + "learning_rate": 1.6105769230769233e-05, + "loss": 1.1617, + "step": 469 + }, + { + "epoch": 0.032293527552562866, + "grad_norm": 1.7717190980911255, + "learning_rate": 1.614010989010989e-05, + "loss": 1.1549, + "step": 470 + }, + { + "epoch": 0.032362237185653425, + "grad_norm": 1.7646262645721436, + "learning_rate": 1.617445054945055e-05, + "loss": 1.129, + "step": 471 + }, + { + "epoch": 0.032430946818743984, + "grad_norm": 1.7824926376342773, + "learning_rate": 1.620879120879121e-05, + "loss": 1.2007, + "step": 472 + }, + { + "epoch": 0.032499656451834544, + "grad_norm": 1.7755070924758911, + "learning_rate": 1.624313186813187e-05, + "loss": 1.1694, + "step": 473 + }, + { + "epoch": 0.0325683660849251, + "grad_norm": 1.7027180194854736, + "learning_rate": 1.6277472527472527e-05, + "loss": 1.1455, + "step": 474 + }, + { + "epoch": 0.03263707571801567, + "grad_norm": 1.682457447052002, + "learning_rate": 1.631181318681319e-05, + "loss": 1.1902, + "step": 475 + }, + { + "epoch": 0.03270578535110623, + "grad_norm": 1.758440613746643, + "learning_rate": 1.6346153846153847e-05, + "loss": 1.1948, + "step": 476 + }, + { + "epoch": 0.03277449498419679, + "grad_norm": 1.722205400466919, + "learning_rate": 1.6380494505494508e-05, + "loss": 1.1929, + "step": 477 + }, + { + "epoch": 0.03284320461728735, + "grad_norm": 1.674831509590149, + "learning_rate": 1.6414835164835166e-05, + "loss": 1.1434, + "step": 478 + }, + { + "epoch": 0.032911914250377906, + "grad_norm": 1.7555466890335083, + "learning_rate": 1.6449175824175824e-05, + "loss": 1.1218, + "step": 479 + }, + { + "epoch": 0.032980623883468466, + "grad_norm": 1.8458892107009888, + "learning_rate": 1.6483516483516486e-05, + "loss": 1.1315, + "step": 480 + }, + { + "epoch": 0.033049333516559025, + "grad_norm": 1.6962281465530396, + "learning_rate": 1.6517857142857144e-05, + "loss": 1.1574, + "step": 481 + }, + { + "epoch": 0.033118043149649584, + "grad_norm": 1.6085964441299438, + "learning_rate": 1.65521978021978e-05, + "loss": 1.1334, + "step": 482 + }, + { + "epoch": 0.03318675278274014, + "grad_norm": 1.7329930067062378, + "learning_rate": 1.6586538461538463e-05, + "loss": 1.1931, + "step": 483 + }, + { + "epoch": 0.0332554624158307, + "grad_norm": 1.8329919576644897, + "learning_rate": 1.662087912087912e-05, + "loss": 1.2145, + "step": 484 + }, + { + "epoch": 0.03332417204892126, + "grad_norm": 1.8985154628753662, + "learning_rate": 1.665521978021978e-05, + "loss": 1.1462, + "step": 485 + }, + { + "epoch": 0.03339288168201182, + "grad_norm": 1.7815524339675903, + "learning_rate": 1.668956043956044e-05, + "loss": 1.1624, + "step": 486 + }, + { + "epoch": 0.03346159131510238, + "grad_norm": 1.6886197328567505, + "learning_rate": 1.67239010989011e-05, + "loss": 1.2356, + "step": 487 + }, + { + "epoch": 0.03353030094819294, + "grad_norm": 1.7008507251739502, + "learning_rate": 1.6758241758241757e-05, + "loss": 1.1603, + "step": 488 + }, + { + "epoch": 0.0335990105812835, + "grad_norm": 1.6761882305145264, + "learning_rate": 1.6792582417582418e-05, + "loss": 1.1397, + "step": 489 + }, + { + "epoch": 0.03366772021437406, + "grad_norm": 1.7582684755325317, + "learning_rate": 1.682692307692308e-05, + "loss": 1.1686, + "step": 490 + }, + { + "epoch": 0.03373642984746462, + "grad_norm": 1.7318756580352783, + "learning_rate": 1.6861263736263738e-05, + "loss": 1.0909, + "step": 491 + }, + { + "epoch": 0.03380513948055518, + "grad_norm": 1.70333731174469, + "learning_rate": 1.68956043956044e-05, + "loss": 1.1877, + "step": 492 + }, + { + "epoch": 0.033873849113645736, + "grad_norm": 1.7727246284484863, + "learning_rate": 1.6929945054945057e-05, + "loss": 1.2018, + "step": 493 + }, + { + "epoch": 0.033942558746736295, + "grad_norm": 1.6856415271759033, + "learning_rate": 1.6964285714285715e-05, + "loss": 1.1593, + "step": 494 + }, + { + "epoch": 0.034011268379826855, + "grad_norm": 1.8097364902496338, + "learning_rate": 1.6998626373626377e-05, + "loss": 1.1625, + "step": 495 + }, + { + "epoch": 0.034079978012917414, + "grad_norm": 1.7924458980560303, + "learning_rate": 1.7032967032967035e-05, + "loss": 1.17, + "step": 496 + }, + { + "epoch": 0.03414868764600797, + "grad_norm": 1.6794463396072388, + "learning_rate": 1.7067307692307693e-05, + "loss": 1.1672, + "step": 497 + }, + { + "epoch": 0.03421739727909853, + "grad_norm": 1.7365984916687012, + "learning_rate": 1.7101648351648354e-05, + "loss": 1.1798, + "step": 498 + }, + { + "epoch": 0.03428610691218909, + "grad_norm": 1.7460614442825317, + "learning_rate": 1.7135989010989012e-05, + "loss": 1.156, + "step": 499 + }, + { + "epoch": 0.03435481654527965, + "grad_norm": 1.7511144876480103, + "learning_rate": 1.717032967032967e-05, + "loss": 1.1679, + "step": 500 + }, + { + "epoch": 0.03442352617837021, + "grad_norm": 1.662729263305664, + "learning_rate": 1.7204670329670332e-05, + "loss": 1.1791, + "step": 501 + }, + { + "epoch": 0.03449223581146077, + "grad_norm": 1.7757298946380615, + "learning_rate": 1.723901098901099e-05, + "loss": 1.1987, + "step": 502 + }, + { + "epoch": 0.03456094544455133, + "grad_norm": 1.6984843015670776, + "learning_rate": 1.7273351648351648e-05, + "loss": 1.1425, + "step": 503 + }, + { + "epoch": 0.03462965507764189, + "grad_norm": 1.6390759944915771, + "learning_rate": 1.730769230769231e-05, + "loss": 1.1737, + "step": 504 + }, + { + "epoch": 0.03469836471073245, + "grad_norm": 1.8220621347427368, + "learning_rate": 1.7342032967032967e-05, + "loss": 1.1631, + "step": 505 + }, + { + "epoch": 0.03476707434382301, + "grad_norm": 1.885154128074646, + "learning_rate": 1.7376373626373625e-05, + "loss": 1.1379, + "step": 506 + }, + { + "epoch": 0.034835783976913566, + "grad_norm": 1.7664802074432373, + "learning_rate": 1.7410714285714287e-05, + "loss": 1.1472, + "step": 507 + }, + { + "epoch": 0.034904493610004125, + "grad_norm": 1.6190131902694702, + "learning_rate": 1.7445054945054945e-05, + "loss": 1.2116, + "step": 508 + }, + { + "epoch": 0.034973203243094685, + "grad_norm": 1.585085391998291, + "learning_rate": 1.7479395604395603e-05, + "loss": 1.1387, + "step": 509 + }, + { + "epoch": 0.035041912876185244, + "grad_norm": 1.7882020473480225, + "learning_rate": 1.7513736263736264e-05, + "loss": 1.1029, + "step": 510 + }, + { + "epoch": 0.0351106225092758, + "grad_norm": 1.9038236141204834, + "learning_rate": 1.7548076923076922e-05, + "loss": 1.1546, + "step": 511 + }, + { + "epoch": 0.03517933214236636, + "grad_norm": 1.7224149703979492, + "learning_rate": 1.7582417582417584e-05, + "loss": 1.2019, + "step": 512 + }, + { + "epoch": 0.03524804177545692, + "grad_norm": 1.6775593757629395, + "learning_rate": 1.7616758241758245e-05, + "loss": 1.1445, + "step": 513 + }, + { + "epoch": 0.03531675140854748, + "grad_norm": 1.650673508644104, + "learning_rate": 1.7651098901098903e-05, + "loss": 1.2096, + "step": 514 + }, + { + "epoch": 0.03538546104163804, + "grad_norm": 1.8677386045455933, + "learning_rate": 1.768543956043956e-05, + "loss": 1.1303, + "step": 515 + }, + { + "epoch": 0.0354541706747286, + "grad_norm": 1.7717978954315186, + "learning_rate": 1.7719780219780223e-05, + "loss": 1.1319, + "step": 516 + }, + { + "epoch": 0.03552288030781916, + "grad_norm": 1.6628707647323608, + "learning_rate": 1.775412087912088e-05, + "loss": 1.1787, + "step": 517 + }, + { + "epoch": 0.03559158994090972, + "grad_norm": 1.6762495040893555, + "learning_rate": 1.778846153846154e-05, + "loss": 1.1333, + "step": 518 + }, + { + "epoch": 0.03566029957400028, + "grad_norm": 1.679317593574524, + "learning_rate": 1.78228021978022e-05, + "loss": 1.1904, + "step": 519 + }, + { + "epoch": 0.03572900920709084, + "grad_norm": 1.6273823976516724, + "learning_rate": 1.785714285714286e-05, + "loss": 1.1496, + "step": 520 + }, + { + "epoch": 0.035797718840181396, + "grad_norm": 1.7513161897659302, + "learning_rate": 1.7891483516483517e-05, + "loss": 1.1161, + "step": 521 + }, + { + "epoch": 0.035866428473271955, + "grad_norm": 1.6488507986068726, + "learning_rate": 1.7925824175824178e-05, + "loss": 1.1353, + "step": 522 + }, + { + "epoch": 0.035935138106362514, + "grad_norm": 1.6639676094055176, + "learning_rate": 1.7960164835164836e-05, + "loss": 1.1503, + "step": 523 + }, + { + "epoch": 0.036003847739453074, + "grad_norm": 1.6695992946624756, + "learning_rate": 1.7994505494505494e-05, + "loss": 1.2539, + "step": 524 + }, + { + "epoch": 0.03607255737254363, + "grad_norm": 1.641670823097229, + "learning_rate": 1.8028846153846156e-05, + "loss": 1.1772, + "step": 525 + }, + { + "epoch": 0.03614126700563419, + "grad_norm": 1.6594593524932861, + "learning_rate": 1.8063186813186814e-05, + "loss": 1.1489, + "step": 526 + }, + { + "epoch": 0.03620997663872475, + "grad_norm": 1.6270357370376587, + "learning_rate": 1.809752747252747e-05, + "loss": 1.1083, + "step": 527 + }, + { + "epoch": 0.03627868627181531, + "grad_norm": 1.7334877252578735, + "learning_rate": 1.8131868131868133e-05, + "loss": 1.1204, + "step": 528 + }, + { + "epoch": 0.03634739590490587, + "grad_norm": 1.6656981706619263, + "learning_rate": 1.816620879120879e-05, + "loss": 1.1692, + "step": 529 + }, + { + "epoch": 0.03641610553799643, + "grad_norm": 1.6904979944229126, + "learning_rate": 1.820054945054945e-05, + "loss": 1.1922, + "step": 530 + }, + { + "epoch": 0.03648481517108699, + "grad_norm": 1.6186573505401611, + "learning_rate": 1.823489010989011e-05, + "loss": 1.1633, + "step": 531 + }, + { + "epoch": 0.03655352480417755, + "grad_norm": 1.784148931503296, + "learning_rate": 1.826923076923077e-05, + "loss": 1.1046, + "step": 532 + }, + { + "epoch": 0.03662223443726811, + "grad_norm": 1.736232876777649, + "learning_rate": 1.8303571428571427e-05, + "loss": 1.1113, + "step": 533 + }, + { + "epoch": 0.036690944070358666, + "grad_norm": 1.6031584739685059, + "learning_rate": 1.8337912087912088e-05, + "loss": 1.1833, + "step": 534 + }, + { + "epoch": 0.036759653703449226, + "grad_norm": 1.6498686075210571, + "learning_rate": 1.837225274725275e-05, + "loss": 1.1512, + "step": 535 + }, + { + "epoch": 0.036828363336539785, + "grad_norm": 1.660404086112976, + "learning_rate": 1.8406593406593408e-05, + "loss": 1.1672, + "step": 536 + }, + { + "epoch": 0.036897072969630344, + "grad_norm": 1.7683199644088745, + "learning_rate": 1.844093406593407e-05, + "loss": 1.1466, + "step": 537 + }, + { + "epoch": 0.036965782602720904, + "grad_norm": 1.6840310096740723, + "learning_rate": 1.8475274725274727e-05, + "loss": 1.091, + "step": 538 + }, + { + "epoch": 0.03703449223581146, + "grad_norm": 1.6686112880706787, + "learning_rate": 1.8509615384615385e-05, + "loss": 1.1636, + "step": 539 + }, + { + "epoch": 0.03710320186890202, + "grad_norm": 1.522059679031372, + "learning_rate": 1.8543956043956047e-05, + "loss": 1.0856, + "step": 540 + }, + { + "epoch": 0.03717191150199258, + "grad_norm": 1.6786359548568726, + "learning_rate": 1.8578296703296705e-05, + "loss": 1.1994, + "step": 541 + }, + { + "epoch": 0.03724062113508314, + "grad_norm": 1.5737775564193726, + "learning_rate": 1.8612637362637363e-05, + "loss": 1.1588, + "step": 542 + }, + { + "epoch": 0.0373093307681737, + "grad_norm": 1.6301608085632324, + "learning_rate": 1.8646978021978024e-05, + "loss": 1.1051, + "step": 543 + }, + { + "epoch": 0.03737804040126426, + "grad_norm": 1.5849961042404175, + "learning_rate": 1.8681318681318682e-05, + "loss": 1.1414, + "step": 544 + }, + { + "epoch": 0.03744675003435482, + "grad_norm": 1.553804636001587, + "learning_rate": 1.871565934065934e-05, + "loss": 1.1455, + "step": 545 + }, + { + "epoch": 0.03751545966744538, + "grad_norm": 1.5821083784103394, + "learning_rate": 1.8750000000000002e-05, + "loss": 1.1538, + "step": 546 + }, + { + "epoch": 0.03758416930053594, + "grad_norm": 1.6158208847045898, + "learning_rate": 1.878434065934066e-05, + "loss": 1.1207, + "step": 547 + }, + { + "epoch": 0.037652878933626496, + "grad_norm": 1.5815308094024658, + "learning_rate": 1.8818681318681318e-05, + "loss": 1.1127, + "step": 548 + }, + { + "epoch": 0.037721588566717056, + "grad_norm": 1.5653152465820312, + "learning_rate": 1.885302197802198e-05, + "loss": 1.1646, + "step": 549 + }, + { + "epoch": 0.037790298199807615, + "grad_norm": 1.615859031677246, + "learning_rate": 1.8887362637362637e-05, + "loss": 1.1481, + "step": 550 + }, + { + "epoch": 0.037859007832898174, + "grad_norm": 1.7190217971801758, + "learning_rate": 1.8921703296703295e-05, + "loss": 1.1551, + "step": 551 + }, + { + "epoch": 0.03792771746598873, + "grad_norm": 1.8301186561584473, + "learning_rate": 1.8956043956043957e-05, + "loss": 1.1494, + "step": 552 + }, + { + "epoch": 0.03799642709907929, + "grad_norm": 1.567068338394165, + "learning_rate": 1.8990384615384615e-05, + "loss": 1.1509, + "step": 553 + }, + { + "epoch": 0.03806513673216985, + "grad_norm": 1.563622236251831, + "learning_rate": 1.9024725274725276e-05, + "loss": 1.1457, + "step": 554 + }, + { + "epoch": 0.03813384636526041, + "grad_norm": 1.7564185857772827, + "learning_rate": 1.9059065934065934e-05, + "loss": 1.1381, + "step": 555 + }, + { + "epoch": 0.03820255599835097, + "grad_norm": 1.7551841735839844, + "learning_rate": 1.9093406593406592e-05, + "loss": 1.1786, + "step": 556 + }, + { + "epoch": 0.03827126563144153, + "grad_norm": 1.66153883934021, + "learning_rate": 1.9127747252747254e-05, + "loss": 1.1238, + "step": 557 + }, + { + "epoch": 0.03833997526453209, + "grad_norm": 1.6048210859298706, + "learning_rate": 1.9162087912087915e-05, + "loss": 1.1796, + "step": 558 + }, + { + "epoch": 0.03840868489762265, + "grad_norm": 1.5770543813705444, + "learning_rate": 1.9196428571428573e-05, + "loss": 1.1577, + "step": 559 + }, + { + "epoch": 0.03847739453071321, + "grad_norm": 1.7265191078186035, + "learning_rate": 1.923076923076923e-05, + "loss": 1.1673, + "step": 560 + }, + { + "epoch": 0.03854610416380377, + "grad_norm": 1.585929036140442, + "learning_rate": 1.9265109890109893e-05, + "loss": 1.1579, + "step": 561 + }, + { + "epoch": 0.038614813796894326, + "grad_norm": 1.6755855083465576, + "learning_rate": 1.929945054945055e-05, + "loss": 1.1445, + "step": 562 + }, + { + "epoch": 0.038683523429984885, + "grad_norm": 1.4991428852081299, + "learning_rate": 1.933379120879121e-05, + "loss": 1.0672, + "step": 563 + }, + { + "epoch": 0.038752233063075445, + "grad_norm": 1.6351678371429443, + "learning_rate": 1.936813186813187e-05, + "loss": 1.0936, + "step": 564 + }, + { + "epoch": 0.038820942696166004, + "grad_norm": 1.6292706727981567, + "learning_rate": 1.940247252747253e-05, + "loss": 1.1484, + "step": 565 + }, + { + "epoch": 0.03888965232925656, + "grad_norm": 1.7358602285385132, + "learning_rate": 1.9436813186813187e-05, + "loss": 1.1204, + "step": 566 + }, + { + "epoch": 0.03895836196234712, + "grad_norm": 1.565197229385376, + "learning_rate": 1.9471153846153848e-05, + "loss": 1.0896, + "step": 567 + }, + { + "epoch": 0.03902707159543768, + "grad_norm": 1.7518967390060425, + "learning_rate": 1.9505494505494506e-05, + "loss": 1.1718, + "step": 568 + }, + { + "epoch": 0.03909578122852824, + "grad_norm": 1.5328354835510254, + "learning_rate": 1.9539835164835164e-05, + "loss": 1.1217, + "step": 569 + }, + { + "epoch": 0.0391644908616188, + "grad_norm": 1.6033594608306885, + "learning_rate": 1.9574175824175826e-05, + "loss": 1.1414, + "step": 570 + }, + { + "epoch": 0.03923320049470936, + "grad_norm": 1.6320630311965942, + "learning_rate": 1.9608516483516484e-05, + "loss": 1.1762, + "step": 571 + }, + { + "epoch": 0.03930191012779992, + "grad_norm": 1.6371948719024658, + "learning_rate": 1.9642857142857145e-05, + "loss": 1.1363, + "step": 572 + }, + { + "epoch": 0.03937061976089048, + "grad_norm": 1.6304577589035034, + "learning_rate": 1.9677197802197803e-05, + "loss": 1.1422, + "step": 573 + }, + { + "epoch": 0.03943932939398104, + "grad_norm": 1.754946231842041, + "learning_rate": 1.971153846153846e-05, + "loss": 1.1987, + "step": 574 + }, + { + "epoch": 0.0395080390270716, + "grad_norm": 1.546567678451538, + "learning_rate": 1.9745879120879123e-05, + "loss": 1.1183, + "step": 575 + }, + { + "epoch": 0.039576748660162156, + "grad_norm": 1.5899689197540283, + "learning_rate": 1.978021978021978e-05, + "loss": 1.164, + "step": 576 + }, + { + "epoch": 0.039645458293252715, + "grad_norm": 1.619973063468933, + "learning_rate": 1.981456043956044e-05, + "loss": 1.1066, + "step": 577 + }, + { + "epoch": 0.039714167926343275, + "grad_norm": 1.5166184902191162, + "learning_rate": 1.98489010989011e-05, + "loss": 1.0588, + "step": 578 + }, + { + "epoch": 0.039782877559433834, + "grad_norm": 1.620619535446167, + "learning_rate": 1.9883241758241758e-05, + "loss": 1.1344, + "step": 579 + }, + { + "epoch": 0.03985158719252439, + "grad_norm": 1.5495107173919678, + "learning_rate": 1.9917582417582416e-05, + "loss": 1.103, + "step": 580 + }, + { + "epoch": 0.03992029682561495, + "grad_norm": 1.6892927885055542, + "learning_rate": 1.9951923076923078e-05, + "loss": 1.1201, + "step": 581 + }, + { + "epoch": 0.03998900645870551, + "grad_norm": 1.5335522890090942, + "learning_rate": 1.998626373626374e-05, + "loss": 1.0989, + "step": 582 + }, + { + "epoch": 0.04005771609179607, + "grad_norm": 1.609611988067627, + "learning_rate": 2.0020604395604397e-05, + "loss": 1.1107, + "step": 583 + }, + { + "epoch": 0.04012642572488663, + "grad_norm": 1.6187851428985596, + "learning_rate": 2.0054945054945055e-05, + "loss": 1.1543, + "step": 584 + }, + { + "epoch": 0.04019513535797719, + "grad_norm": 1.5373953580856323, + "learning_rate": 2.0089285714285717e-05, + "loss": 1.0769, + "step": 585 + }, + { + "epoch": 0.04026384499106775, + "grad_norm": 1.456292986869812, + "learning_rate": 2.0123626373626375e-05, + "loss": 1.1157, + "step": 586 + }, + { + "epoch": 0.04033255462415831, + "grad_norm": 1.4465841054916382, + "learning_rate": 2.0157967032967036e-05, + "loss": 1.0778, + "step": 587 + }, + { + "epoch": 0.04040126425724887, + "grad_norm": 1.5776021480560303, + "learning_rate": 2.0192307692307694e-05, + "loss": 1.1189, + "step": 588 + }, + { + "epoch": 0.04046997389033943, + "grad_norm": 1.558456540107727, + "learning_rate": 2.0226648351648352e-05, + "loss": 1.16, + "step": 589 + }, + { + "epoch": 0.040538683523429986, + "grad_norm": 1.580700159072876, + "learning_rate": 2.0260989010989014e-05, + "loss": 1.1453, + "step": 590 + }, + { + "epoch": 0.040607393156520545, + "grad_norm": 1.6480629444122314, + "learning_rate": 2.0295329670329672e-05, + "loss": 1.1296, + "step": 591 + }, + { + "epoch": 0.040676102789611104, + "grad_norm": 1.609710931777954, + "learning_rate": 2.032967032967033e-05, + "loss": 1.1112, + "step": 592 + }, + { + "epoch": 0.040744812422701664, + "grad_norm": 1.5200614929199219, + "learning_rate": 2.036401098901099e-05, + "loss": 1.1221, + "step": 593 + }, + { + "epoch": 0.04081352205579222, + "grad_norm": 1.6501137018203735, + "learning_rate": 2.039835164835165e-05, + "loss": 1.1325, + "step": 594 + }, + { + "epoch": 0.04088223168888278, + "grad_norm": 1.5481053590774536, + "learning_rate": 2.0432692307692307e-05, + "loss": 1.1045, + "step": 595 + }, + { + "epoch": 0.04095094132197334, + "grad_norm": 1.5386232137680054, + "learning_rate": 2.046703296703297e-05, + "loss": 1.1852, + "step": 596 + }, + { + "epoch": 0.0410196509550639, + "grad_norm": 1.5110889673233032, + "learning_rate": 2.0501373626373627e-05, + "loss": 1.1613, + "step": 597 + }, + { + "epoch": 0.04108836058815446, + "grad_norm": 1.6547242403030396, + "learning_rate": 2.0535714285714285e-05, + "loss": 1.1597, + "step": 598 + }, + { + "epoch": 0.04115707022124502, + "grad_norm": 1.671565294265747, + "learning_rate": 2.0570054945054946e-05, + "loss": 1.1675, + "step": 599 + }, + { + "epoch": 0.04122577985433558, + "grad_norm": 1.793827772140503, + "learning_rate": 2.0604395604395604e-05, + "loss": 1.1577, + "step": 600 + }, + { + "epoch": 0.04129448948742614, + "grad_norm": 1.5763945579528809, + "learning_rate": 2.0638736263736262e-05, + "loss": 1.1711, + "step": 601 + }, + { + "epoch": 0.0413631991205167, + "grad_norm": 1.465631127357483, + "learning_rate": 2.0673076923076924e-05, + "loss": 1.0787, + "step": 602 + }, + { + "epoch": 0.041431908753607256, + "grad_norm": 1.5874712467193604, + "learning_rate": 2.0707417582417585e-05, + "loss": 1.1457, + "step": 603 + }, + { + "epoch": 0.041500618386697816, + "grad_norm": 1.5705673694610596, + "learning_rate": 2.0741758241758243e-05, + "loss": 1.157, + "step": 604 + }, + { + "epoch": 0.041569328019788375, + "grad_norm": 1.6986249685287476, + "learning_rate": 2.0776098901098905e-05, + "loss": 1.1515, + "step": 605 + }, + { + "epoch": 0.041638037652878934, + "grad_norm": 1.574314832687378, + "learning_rate": 2.0810439560439563e-05, + "loss": 1.1062, + "step": 606 + }, + { + "epoch": 0.041706747285969493, + "grad_norm": 1.5826187133789062, + "learning_rate": 2.084478021978022e-05, + "loss": 1.1786, + "step": 607 + }, + { + "epoch": 0.04177545691906005, + "grad_norm": 1.5744619369506836, + "learning_rate": 2.0879120879120882e-05, + "loss": 1.1469, + "step": 608 + }, + { + "epoch": 0.04184416655215061, + "grad_norm": 1.6228045225143433, + "learning_rate": 2.091346153846154e-05, + "loss": 1.1313, + "step": 609 + }, + { + "epoch": 0.04191287618524117, + "grad_norm": 1.5369139909744263, + "learning_rate": 2.09478021978022e-05, + "loss": 1.1169, + "step": 610 + }, + { + "epoch": 0.04198158581833173, + "grad_norm": 1.529010534286499, + "learning_rate": 2.098214285714286e-05, + "loss": 1.1129, + "step": 611 + }, + { + "epoch": 0.04205029545142229, + "grad_norm": 1.580208420753479, + "learning_rate": 2.1016483516483518e-05, + "loss": 1.1199, + "step": 612 + }, + { + "epoch": 0.04211900508451285, + "grad_norm": 1.6764243841171265, + "learning_rate": 2.1050824175824176e-05, + "loss": 1.1343, + "step": 613 + }, + { + "epoch": 0.04218771471760341, + "grad_norm": 1.5353320837020874, + "learning_rate": 2.1085164835164837e-05, + "loss": 1.1219, + "step": 614 + }, + { + "epoch": 0.04225642435069397, + "grad_norm": 1.5774692296981812, + "learning_rate": 2.1119505494505496e-05, + "loss": 1.1534, + "step": 615 + }, + { + "epoch": 0.04232513398378453, + "grad_norm": 1.6071498394012451, + "learning_rate": 2.1153846153846154e-05, + "loss": 1.1494, + "step": 616 + }, + { + "epoch": 0.042393843616875086, + "grad_norm": 1.5281569957733154, + "learning_rate": 2.1188186813186815e-05, + "loss": 1.1025, + "step": 617 + }, + { + "epoch": 0.042462553249965646, + "grad_norm": 1.6154829263687134, + "learning_rate": 2.1222527472527473e-05, + "loss": 1.1664, + "step": 618 + }, + { + "epoch": 0.042531262883056205, + "grad_norm": 1.6135022640228271, + "learning_rate": 2.125686813186813e-05, + "loss": 1.1398, + "step": 619 + }, + { + "epoch": 0.042599972516146764, + "grad_norm": 1.61142098903656, + "learning_rate": 2.1291208791208793e-05, + "loss": 1.1276, + "step": 620 + }, + { + "epoch": 0.04266868214923732, + "grad_norm": 1.581107258796692, + "learning_rate": 2.132554945054945e-05, + "loss": 1.107, + "step": 621 + }, + { + "epoch": 0.04273739178232788, + "grad_norm": 1.4848287105560303, + "learning_rate": 2.135989010989011e-05, + "loss": 1.1223, + "step": 622 + }, + { + "epoch": 0.04280610141541844, + "grad_norm": 1.56912100315094, + "learning_rate": 2.139423076923077e-05, + "loss": 1.1346, + "step": 623 + }, + { + "epoch": 0.042874811048509, + "grad_norm": 1.6470636129379272, + "learning_rate": 2.1428571428571428e-05, + "loss": 1.1328, + "step": 624 + }, + { + "epoch": 0.04294352068159956, + "grad_norm": 1.6345257759094238, + "learning_rate": 2.1462912087912086e-05, + "loss": 1.1308, + "step": 625 + }, + { + "epoch": 0.04301223031469012, + "grad_norm": 1.5969098806381226, + "learning_rate": 2.149725274725275e-05, + "loss": 1.0734, + "step": 626 + }, + { + "epoch": 0.04308093994778068, + "grad_norm": 1.680349588394165, + "learning_rate": 2.153159340659341e-05, + "loss": 1.1296, + "step": 627 + }, + { + "epoch": 0.04314964958087124, + "grad_norm": 1.5738202333450317, + "learning_rate": 2.1565934065934067e-05, + "loss": 1.1542, + "step": 628 + }, + { + "epoch": 0.0432183592139618, + "grad_norm": 1.6026647090911865, + "learning_rate": 2.160027472527473e-05, + "loss": 1.149, + "step": 629 + }, + { + "epoch": 0.04328706884705236, + "grad_norm": 1.5017374753952026, + "learning_rate": 2.1634615384615387e-05, + "loss": 1.1741, + "step": 630 + }, + { + "epoch": 0.043355778480142916, + "grad_norm": 1.6968938112258911, + "learning_rate": 2.1668956043956045e-05, + "loss": 1.1161, + "step": 631 + }, + { + "epoch": 0.043424488113233475, + "grad_norm": 1.5485678911209106, + "learning_rate": 2.1703296703296706e-05, + "loss": 1.1812, + "step": 632 + }, + { + "epoch": 0.043493197746324035, + "grad_norm": 1.556099772453308, + "learning_rate": 2.1737637362637364e-05, + "loss": 1.1261, + "step": 633 + }, + { + "epoch": 0.043561907379414594, + "grad_norm": 1.5704066753387451, + "learning_rate": 2.1771978021978022e-05, + "loss": 1.1462, + "step": 634 + }, + { + "epoch": 0.04363061701250515, + "grad_norm": 1.6116307973861694, + "learning_rate": 2.1806318681318684e-05, + "loss": 1.134, + "step": 635 + }, + { + "epoch": 0.04369932664559571, + "grad_norm": 1.41005539894104, + "learning_rate": 2.1840659340659342e-05, + "loss": 1.1115, + "step": 636 + }, + { + "epoch": 0.04376803627868627, + "grad_norm": 1.6039692163467407, + "learning_rate": 2.1875e-05, + "loss": 1.1162, + "step": 637 + }, + { + "epoch": 0.04383674591177683, + "grad_norm": 1.6972017288208008, + "learning_rate": 2.190934065934066e-05, + "loss": 1.1619, + "step": 638 + }, + { + "epoch": 0.04390545554486739, + "grad_norm": 1.79932701587677, + "learning_rate": 2.194368131868132e-05, + "loss": 1.1446, + "step": 639 + }, + { + "epoch": 0.04397416517795795, + "grad_norm": 1.7467010021209717, + "learning_rate": 2.1978021978021977e-05, + "loss": 1.1285, + "step": 640 + }, + { + "epoch": 0.04404287481104851, + "grad_norm": 1.5828945636749268, + "learning_rate": 2.201236263736264e-05, + "loss": 1.1374, + "step": 641 + }, + { + "epoch": 0.04411158444413907, + "grad_norm": 1.563348412513733, + "learning_rate": 2.2046703296703297e-05, + "loss": 1.0929, + "step": 642 + }, + { + "epoch": 0.04418029407722963, + "grad_norm": 1.5947496891021729, + "learning_rate": 2.2081043956043955e-05, + "loss": 1.1427, + "step": 643 + }, + { + "epoch": 0.04424900371032019, + "grad_norm": 1.603566288948059, + "learning_rate": 2.2115384615384616e-05, + "loss": 1.1338, + "step": 644 + }, + { + "epoch": 0.044317713343410746, + "grad_norm": 1.7433652877807617, + "learning_rate": 2.2149725274725274e-05, + "loss": 1.1285, + "step": 645 + }, + { + "epoch": 0.044386422976501305, + "grad_norm": 1.7295756340026855, + "learning_rate": 2.2184065934065932e-05, + "loss": 1.1258, + "step": 646 + }, + { + "epoch": 0.044455132609591864, + "grad_norm": 1.620483160018921, + "learning_rate": 2.2218406593406594e-05, + "loss": 1.1511, + "step": 647 + }, + { + "epoch": 0.044523842242682424, + "grad_norm": 1.6002135276794434, + "learning_rate": 2.2252747252747252e-05, + "loss": 1.1474, + "step": 648 + }, + { + "epoch": 0.04459255187577298, + "grad_norm": 1.5935968160629272, + "learning_rate": 2.2287087912087913e-05, + "loss": 1.1229, + "step": 649 + }, + { + "epoch": 0.04466126150886354, + "grad_norm": 1.5148497819900513, + "learning_rate": 2.2321428571428575e-05, + "loss": 1.1281, + "step": 650 + }, + { + "epoch": 0.0447299711419541, + "grad_norm": 1.5224000215530396, + "learning_rate": 2.2355769230769233e-05, + "loss": 1.106, + "step": 651 + }, + { + "epoch": 0.04479868077504466, + "grad_norm": 1.4542063474655151, + "learning_rate": 2.239010989010989e-05, + "loss": 1.1183, + "step": 652 + }, + { + "epoch": 0.04486739040813522, + "grad_norm": 1.5977063179016113, + "learning_rate": 2.2424450549450552e-05, + "loss": 1.1007, + "step": 653 + }, + { + "epoch": 0.04493610004122578, + "grad_norm": 1.6027153730392456, + "learning_rate": 2.245879120879121e-05, + "loss": 1.1024, + "step": 654 + }, + { + "epoch": 0.04500480967431634, + "grad_norm": 1.5036693811416626, + "learning_rate": 2.249313186813187e-05, + "loss": 1.1425, + "step": 655 + }, + { + "epoch": 0.0450735193074069, + "grad_norm": 1.583584189414978, + "learning_rate": 2.252747252747253e-05, + "loss": 1.1229, + "step": 656 + }, + { + "epoch": 0.04514222894049746, + "grad_norm": 1.5424867868423462, + "learning_rate": 2.2561813186813188e-05, + "loss": 1.1588, + "step": 657 + }, + { + "epoch": 0.045210938573588016, + "grad_norm": 1.6314481496810913, + "learning_rate": 2.2596153846153846e-05, + "loss": 1.1733, + "step": 658 + }, + { + "epoch": 0.045279648206678576, + "grad_norm": 1.5366076231002808, + "learning_rate": 2.2630494505494507e-05, + "loss": 1.1172, + "step": 659 + }, + { + "epoch": 0.045348357839769135, + "grad_norm": 1.6070585250854492, + "learning_rate": 2.2664835164835165e-05, + "loss": 1.1247, + "step": 660 + }, + { + "epoch": 0.045417067472859694, + "grad_norm": 1.5759259462356567, + "learning_rate": 2.2699175824175824e-05, + "loss": 1.1263, + "step": 661 + }, + { + "epoch": 0.045485777105950254, + "grad_norm": 1.4769904613494873, + "learning_rate": 2.2733516483516485e-05, + "loss": 1.0934, + "step": 662 + }, + { + "epoch": 0.04555448673904081, + "grad_norm": 1.4979413747787476, + "learning_rate": 2.2767857142857143e-05, + "loss": 1.1345, + "step": 663 + }, + { + "epoch": 0.04562319637213137, + "grad_norm": 1.4468190670013428, + "learning_rate": 2.28021978021978e-05, + "loss": 1.0741, + "step": 664 + }, + { + "epoch": 0.04569190600522193, + "grad_norm": 1.4540092945098877, + "learning_rate": 2.2836538461538463e-05, + "loss": 1.1457, + "step": 665 + }, + { + "epoch": 0.04576061563831249, + "grad_norm": 1.6263999938964844, + "learning_rate": 2.287087912087912e-05, + "loss": 1.1576, + "step": 666 + }, + { + "epoch": 0.04582932527140305, + "grad_norm": 1.5335606336593628, + "learning_rate": 2.2905219780219782e-05, + "loss": 1.1147, + "step": 667 + }, + { + "epoch": 0.04589803490449361, + "grad_norm": 1.651685118675232, + "learning_rate": 2.293956043956044e-05, + "loss": 1.1628, + "step": 668 + }, + { + "epoch": 0.04596674453758417, + "grad_norm": 1.4871405363082886, + "learning_rate": 2.2973901098901098e-05, + "loss": 1.1121, + "step": 669 + }, + { + "epoch": 0.04603545417067473, + "grad_norm": 1.536476731300354, + "learning_rate": 2.300824175824176e-05, + "loss": 1.1081, + "step": 670 + }, + { + "epoch": 0.04610416380376529, + "grad_norm": 1.5245531797409058, + "learning_rate": 2.3042582417582418e-05, + "loss": 1.1844, + "step": 671 + }, + { + "epoch": 0.046172873436855846, + "grad_norm": 1.444306492805481, + "learning_rate": 2.307692307692308e-05, + "loss": 1.1184, + "step": 672 + }, + { + "epoch": 0.046241583069946406, + "grad_norm": 1.5055241584777832, + "learning_rate": 2.3111263736263737e-05, + "loss": 1.0907, + "step": 673 + }, + { + "epoch": 0.046310292703036965, + "grad_norm": 1.5358582735061646, + "learning_rate": 2.31456043956044e-05, + "loss": 1.171, + "step": 674 + }, + { + "epoch": 0.046379002336127524, + "grad_norm": 1.6040992736816406, + "learning_rate": 2.3179945054945057e-05, + "loss": 1.1531, + "step": 675 + }, + { + "epoch": 0.04644771196921808, + "grad_norm": 1.4809134006500244, + "learning_rate": 2.3214285714285715e-05, + "loss": 1.1556, + "step": 676 + }, + { + "epoch": 0.04651642160230864, + "grad_norm": 1.5698521137237549, + "learning_rate": 2.3248626373626376e-05, + "loss": 1.1211, + "step": 677 + }, + { + "epoch": 0.0465851312353992, + "grad_norm": 1.6523971557617188, + "learning_rate": 2.3282967032967034e-05, + "loss": 1.1748, + "step": 678 + }, + { + "epoch": 0.04665384086848976, + "grad_norm": 1.5106384754180908, + "learning_rate": 2.3317307692307692e-05, + "loss": 1.0961, + "step": 679 + }, + { + "epoch": 0.04672255050158032, + "grad_norm": 1.437982201576233, + "learning_rate": 2.3351648351648354e-05, + "loss": 1.1012, + "step": 680 + }, + { + "epoch": 0.04679126013467088, + "grad_norm": 1.531221628189087, + "learning_rate": 2.3385989010989012e-05, + "loss": 1.12, + "step": 681 + }, + { + "epoch": 0.04685996976776144, + "grad_norm": 1.552534580230713, + "learning_rate": 2.3420329670329673e-05, + "loss": 1.124, + "step": 682 + }, + { + "epoch": 0.046928679400852, + "grad_norm": 1.5104304552078247, + "learning_rate": 2.345467032967033e-05, + "loss": 1.1388, + "step": 683 + }, + { + "epoch": 0.04699738903394256, + "grad_norm": 1.437806487083435, + "learning_rate": 2.348901098901099e-05, + "loss": 1.0647, + "step": 684 + }, + { + "epoch": 0.04706609866703312, + "grad_norm": 1.4210172891616821, + "learning_rate": 2.352335164835165e-05, + "loss": 1.0862, + "step": 685 + }, + { + "epoch": 0.047134808300123676, + "grad_norm": 1.6331843137741089, + "learning_rate": 2.355769230769231e-05, + "loss": 1.1362, + "step": 686 + }, + { + "epoch": 0.047203517933214235, + "grad_norm": 1.52473783493042, + "learning_rate": 2.3592032967032967e-05, + "loss": 1.0895, + "step": 687 + }, + { + "epoch": 0.047272227566304795, + "grad_norm": 1.6656391620635986, + "learning_rate": 2.3626373626373628e-05, + "loss": 1.2214, + "step": 688 + }, + { + "epoch": 0.047340937199395354, + "grad_norm": 1.4407799243927002, + "learning_rate": 2.3660714285714286e-05, + "loss": 1.135, + "step": 689 + }, + { + "epoch": 0.04740964683248591, + "grad_norm": 1.4530463218688965, + "learning_rate": 2.3695054945054944e-05, + "loss": 1.1282, + "step": 690 + }, + { + "epoch": 0.04747835646557647, + "grad_norm": 1.4991074800491333, + "learning_rate": 2.3729395604395606e-05, + "loss": 1.1334, + "step": 691 + }, + { + "epoch": 0.04754706609866703, + "grad_norm": 1.5602877140045166, + "learning_rate": 2.3763736263736264e-05, + "loss": 1.1359, + "step": 692 + }, + { + "epoch": 0.04761577573175759, + "grad_norm": 1.5010954141616821, + "learning_rate": 2.3798076923076922e-05, + "loss": 1.1025, + "step": 693 + }, + { + "epoch": 0.04768448536484815, + "grad_norm": 1.513083815574646, + "learning_rate": 2.3832417582417583e-05, + "loss": 1.1052, + "step": 694 + }, + { + "epoch": 0.04775319499793871, + "grad_norm": 1.5073212385177612, + "learning_rate": 2.3866758241758245e-05, + "loss": 1.1529, + "step": 695 + }, + { + "epoch": 0.04782190463102927, + "grad_norm": 1.7182540893554688, + "learning_rate": 2.3901098901098903e-05, + "loss": 1.1607, + "step": 696 + }, + { + "epoch": 0.04789061426411983, + "grad_norm": 1.5094006061553955, + "learning_rate": 2.3935439560439564e-05, + "loss": 1.0929, + "step": 697 + }, + { + "epoch": 0.04795932389721039, + "grad_norm": 1.533164381980896, + "learning_rate": 2.3969780219780222e-05, + "loss": 1.1256, + "step": 698 + }, + { + "epoch": 0.04802803353030095, + "grad_norm": 1.5570908784866333, + "learning_rate": 2.400412087912088e-05, + "loss": 1.1535, + "step": 699 + }, + { + "epoch": 0.048096743163391506, + "grad_norm": 1.5087084770202637, + "learning_rate": 2.4038461538461542e-05, + "loss": 1.1123, + "step": 700 + }, + { + "epoch": 0.048165452796482065, + "grad_norm": 1.4968030452728271, + "learning_rate": 2.40728021978022e-05, + "loss": 1.1138, + "step": 701 + }, + { + "epoch": 0.048234162429572625, + "grad_norm": 1.7165296077728271, + "learning_rate": 2.4107142857142858e-05, + "loss": 1.1372, + "step": 702 + }, + { + "epoch": 0.048302872062663184, + "grad_norm": 1.4747172594070435, + "learning_rate": 2.414148351648352e-05, + "loss": 1.1268, + "step": 703 + }, + { + "epoch": 0.04837158169575374, + "grad_norm": 1.4719194173812866, + "learning_rate": 2.4175824175824177e-05, + "loss": 1.1064, + "step": 704 + }, + { + "epoch": 0.0484402913288443, + "grad_norm": 1.4546678066253662, + "learning_rate": 2.4210164835164835e-05, + "loss": 1.1195, + "step": 705 + }, + { + "epoch": 0.04850900096193486, + "grad_norm": 1.5038517713546753, + "learning_rate": 2.4244505494505497e-05, + "loss": 1.0949, + "step": 706 + }, + { + "epoch": 0.04857771059502542, + "grad_norm": 1.5176599025726318, + "learning_rate": 2.4278846153846155e-05, + "loss": 1.1468, + "step": 707 + }, + { + "epoch": 0.04864642022811598, + "grad_norm": 1.5845799446105957, + "learning_rate": 2.4313186813186813e-05, + "loss": 1.1302, + "step": 708 + }, + { + "epoch": 0.04871512986120654, + "grad_norm": 1.519037127494812, + "learning_rate": 2.4347527472527474e-05, + "loss": 1.0711, + "step": 709 + }, + { + "epoch": 0.0487838394942971, + "grad_norm": 1.488869309425354, + "learning_rate": 2.4381868131868133e-05, + "loss": 1.1176, + "step": 710 + }, + { + "epoch": 0.04885254912738766, + "grad_norm": 1.514968991279602, + "learning_rate": 2.441620879120879e-05, + "loss": 1.1592, + "step": 711 + }, + { + "epoch": 0.04892125876047822, + "grad_norm": 1.3722155094146729, + "learning_rate": 2.4450549450549452e-05, + "loss": 1.1104, + "step": 712 + }, + { + "epoch": 0.04898996839356878, + "grad_norm": 1.6659224033355713, + "learning_rate": 2.448489010989011e-05, + "loss": 1.1262, + "step": 713 + }, + { + "epoch": 0.049058678026659336, + "grad_norm": 1.4159419536590576, + "learning_rate": 2.4519230769230768e-05, + "loss": 1.0783, + "step": 714 + }, + { + "epoch": 0.049127387659749895, + "grad_norm": 1.5085432529449463, + "learning_rate": 2.455357142857143e-05, + "loss": 1.1848, + "step": 715 + }, + { + "epoch": 0.049196097292840454, + "grad_norm": 1.4752134084701538, + "learning_rate": 2.4587912087912088e-05, + "loss": 1.0854, + "step": 716 + }, + { + "epoch": 0.049264806925931014, + "grad_norm": 1.4816161394119263, + "learning_rate": 2.462225274725275e-05, + "loss": 1.1171, + "step": 717 + }, + { + "epoch": 0.04933351655902157, + "grad_norm": 1.5188652276992798, + "learning_rate": 2.465659340659341e-05, + "loss": 1.1285, + "step": 718 + }, + { + "epoch": 0.04940222619211213, + "grad_norm": 1.4633381366729736, + "learning_rate": 2.469093406593407e-05, + "loss": 1.1564, + "step": 719 + }, + { + "epoch": 0.04947093582520269, + "grad_norm": 1.4717392921447754, + "learning_rate": 2.4725274725274727e-05, + "loss": 1.1437, + "step": 720 + }, + { + "epoch": 0.04953964545829325, + "grad_norm": 1.533626914024353, + "learning_rate": 2.4759615384615388e-05, + "loss": 1.1026, + "step": 721 + }, + { + "epoch": 0.04960835509138381, + "grad_norm": 1.5112859010696411, + "learning_rate": 2.4793956043956046e-05, + "loss": 1.1445, + "step": 722 + }, + { + "epoch": 0.04967706472447437, + "grad_norm": 1.5426957607269287, + "learning_rate": 2.4828296703296704e-05, + "loss": 1.0984, + "step": 723 + }, + { + "epoch": 0.04974577435756493, + "grad_norm": 1.4626768827438354, + "learning_rate": 2.4862637362637366e-05, + "loss": 1.1308, + "step": 724 + }, + { + "epoch": 0.04981448399065549, + "grad_norm": 1.4310908317565918, + "learning_rate": 2.4896978021978024e-05, + "loss": 1.1373, + "step": 725 + }, + { + "epoch": 0.04988319362374605, + "grad_norm": 1.406058430671692, + "learning_rate": 2.4931318681318682e-05, + "loss": 1.0952, + "step": 726 + }, + { + "epoch": 0.049951903256836606, + "grad_norm": 1.595439076423645, + "learning_rate": 2.4965659340659343e-05, + "loss": 1.0736, + "step": 727 + }, + { + "epoch": 0.050020612889927166, + "grad_norm": 1.4899888038635254, + "learning_rate": 2.5e-05, + "loss": 1.0924, + "step": 728 + }, + { + "epoch": 0.050089322523017725, + "grad_norm": 1.5029473304748535, + "learning_rate": 2.503434065934066e-05, + "loss": 1.1316, + "step": 729 + }, + { + "epoch": 0.050158032156108284, + "grad_norm": 1.5585927963256836, + "learning_rate": 2.506868131868132e-05, + "loss": 1.1859, + "step": 730 + }, + { + "epoch": 0.050226741789198844, + "grad_norm": 1.5078426599502563, + "learning_rate": 2.510302197802198e-05, + "loss": 1.1245, + "step": 731 + }, + { + "epoch": 0.0502954514222894, + "grad_norm": 1.5172957181930542, + "learning_rate": 2.5137362637362637e-05, + "loss": 1.1126, + "step": 732 + }, + { + "epoch": 0.05036416105537996, + "grad_norm": 1.4753841161727905, + "learning_rate": 2.5171703296703298e-05, + "loss": 1.1425, + "step": 733 + }, + { + "epoch": 0.05043287068847052, + "grad_norm": 1.4766228199005127, + "learning_rate": 2.5206043956043956e-05, + "loss": 1.1283, + "step": 734 + }, + { + "epoch": 0.05050158032156108, + "grad_norm": 1.5686410665512085, + "learning_rate": 2.5240384615384614e-05, + "loss": 1.0978, + "step": 735 + }, + { + "epoch": 0.05057028995465164, + "grad_norm": 1.525753140449524, + "learning_rate": 2.5274725274725276e-05, + "loss": 1.0957, + "step": 736 + }, + { + "epoch": 0.0506389995877422, + "grad_norm": 1.6094753742218018, + "learning_rate": 2.5309065934065934e-05, + "loss": 1.1678, + "step": 737 + }, + { + "epoch": 0.05070770922083276, + "grad_norm": 1.635509967803955, + "learning_rate": 2.5343406593406592e-05, + "loss": 1.133, + "step": 738 + }, + { + "epoch": 0.05077641885392332, + "grad_norm": 1.4611396789550781, + "learning_rate": 2.5377747252747253e-05, + "loss": 1.1268, + "step": 739 + }, + { + "epoch": 0.05084512848701388, + "grad_norm": 1.5076327323913574, + "learning_rate": 2.541208791208791e-05, + "loss": 1.1022, + "step": 740 + }, + { + "epoch": 0.050913838120104436, + "grad_norm": 1.5467561483383179, + "learning_rate": 2.544642857142857e-05, + "loss": 1.1255, + "step": 741 + }, + { + "epoch": 0.050982547753194996, + "grad_norm": 1.5929625034332275, + "learning_rate": 2.548076923076923e-05, + "loss": 1.1256, + "step": 742 + }, + { + "epoch": 0.051051257386285555, + "grad_norm": 1.4621307849884033, + "learning_rate": 2.551510989010989e-05, + "loss": 1.1383, + "step": 743 + }, + { + "epoch": 0.051119967019376114, + "grad_norm": 1.4789154529571533, + "learning_rate": 2.554945054945055e-05, + "loss": 1.0889, + "step": 744 + }, + { + "epoch": 0.05118867665246667, + "grad_norm": 1.4541867971420288, + "learning_rate": 2.558379120879121e-05, + "loss": 1.1005, + "step": 745 + }, + { + "epoch": 0.05125738628555723, + "grad_norm": 1.4250370264053345, + "learning_rate": 2.5618131868131866e-05, + "loss": 1.1322, + "step": 746 + }, + { + "epoch": 0.05132609591864779, + "grad_norm": 1.5041378736495972, + "learning_rate": 2.5652472527472528e-05, + "loss": 1.1002, + "step": 747 + }, + { + "epoch": 0.05139480555173835, + "grad_norm": 1.5743403434753418, + "learning_rate": 2.5686813186813186e-05, + "loss": 1.1179, + "step": 748 + }, + { + "epoch": 0.05146351518482891, + "grad_norm": 1.516469120979309, + "learning_rate": 2.5721153846153844e-05, + "loss": 1.1436, + "step": 749 + }, + { + "epoch": 0.05153222481791947, + "grad_norm": 1.6065118312835693, + "learning_rate": 2.5755494505494505e-05, + "loss": 1.1597, + "step": 750 + }, + { + "epoch": 0.05160093445101003, + "grad_norm": 1.5636159181594849, + "learning_rate": 2.578983516483517e-05, + "loss": 1.142, + "step": 751 + }, + { + "epoch": 0.05166964408410059, + "grad_norm": 1.4468353986740112, + "learning_rate": 2.582417582417583e-05, + "loss": 1.1312, + "step": 752 + }, + { + "epoch": 0.05173835371719115, + "grad_norm": 1.4934002161026, + "learning_rate": 2.5858516483516486e-05, + "loss": 1.1092, + "step": 753 + }, + { + "epoch": 0.05180706335028171, + "grad_norm": 1.5502829551696777, + "learning_rate": 2.5892857142857148e-05, + "loss": 1.1589, + "step": 754 + }, + { + "epoch": 0.051875772983372266, + "grad_norm": 1.538283348083496, + "learning_rate": 2.5927197802197806e-05, + "loss": 1.1248, + "step": 755 + }, + { + "epoch": 0.051944482616462825, + "grad_norm": 1.3884409666061401, + "learning_rate": 2.5961538461538464e-05, + "loss": 1.1338, + "step": 756 + }, + { + "epoch": 0.052013192249553385, + "grad_norm": 1.482311725616455, + "learning_rate": 2.5995879120879125e-05, + "loss": 1.1565, + "step": 757 + }, + { + "epoch": 0.052081901882643944, + "grad_norm": 1.7636150121688843, + "learning_rate": 2.6030219780219783e-05, + "loss": 1.1862, + "step": 758 + }, + { + "epoch": 0.0521506115157345, + "grad_norm": 1.6587697267532349, + "learning_rate": 2.606456043956044e-05, + "loss": 1.0984, + "step": 759 + }, + { + "epoch": 0.05221932114882506, + "grad_norm": 1.527251124382019, + "learning_rate": 2.6098901098901103e-05, + "loss": 1.1197, + "step": 760 + }, + { + "epoch": 0.05228803078191562, + "grad_norm": 1.6263151168823242, + "learning_rate": 2.613324175824176e-05, + "loss": 1.122, + "step": 761 + }, + { + "epoch": 0.05235674041500618, + "grad_norm": 1.4537711143493652, + "learning_rate": 2.616758241758242e-05, + "loss": 1.0984, + "step": 762 + }, + { + "epoch": 0.05242545004809674, + "grad_norm": 1.79281747341156, + "learning_rate": 2.620192307692308e-05, + "loss": 1.1184, + "step": 763 + }, + { + "epoch": 0.0524941596811873, + "grad_norm": 1.4136320352554321, + "learning_rate": 2.623626373626374e-05, + "loss": 1.121, + "step": 764 + }, + { + "epoch": 0.05256286931427786, + "grad_norm": 1.4549843072891235, + "learning_rate": 2.6270604395604397e-05, + "loss": 1.0694, + "step": 765 + }, + { + "epoch": 0.05263157894736842, + "grad_norm": 1.5701510906219482, + "learning_rate": 2.6304945054945058e-05, + "loss": 1.1197, + "step": 766 + }, + { + "epoch": 0.05270028858045898, + "grad_norm": 1.4749747514724731, + "learning_rate": 2.6339285714285716e-05, + "loss": 1.1308, + "step": 767 + }, + { + "epoch": 0.05276899821354954, + "grad_norm": 1.5145008563995361, + "learning_rate": 2.6373626373626374e-05, + "loss": 1.1321, + "step": 768 + }, + { + "epoch": 0.052837707846640096, + "grad_norm": 1.5834907293319702, + "learning_rate": 2.6407967032967036e-05, + "loss": 1.1043, + "step": 769 + }, + { + "epoch": 0.052906417479730655, + "grad_norm": 1.4773167371749878, + "learning_rate": 2.6442307692307694e-05, + "loss": 1.0981, + "step": 770 + }, + { + "epoch": 0.052975127112821215, + "grad_norm": 1.4857847690582275, + "learning_rate": 2.647664835164835e-05, + "loss": 1.0984, + "step": 771 + }, + { + "epoch": 0.053043836745911774, + "grad_norm": 1.396246314048767, + "learning_rate": 2.6510989010989013e-05, + "loss": 1.0935, + "step": 772 + }, + { + "epoch": 0.05311254637900233, + "grad_norm": 1.4466161727905273, + "learning_rate": 2.654532967032967e-05, + "loss": 1.1154, + "step": 773 + }, + { + "epoch": 0.05318125601209289, + "grad_norm": 1.5052876472473145, + "learning_rate": 2.657967032967033e-05, + "loss": 1.1087, + "step": 774 + }, + { + "epoch": 0.05324996564518345, + "grad_norm": 1.493972897529602, + "learning_rate": 2.661401098901099e-05, + "loss": 1.0476, + "step": 775 + }, + { + "epoch": 0.05331867527827401, + "grad_norm": 1.5093737840652466, + "learning_rate": 2.664835164835165e-05, + "loss": 1.1144, + "step": 776 + }, + { + "epoch": 0.05338738491136457, + "grad_norm": 1.480675458908081, + "learning_rate": 2.668269230769231e-05, + "loss": 1.1252, + "step": 777 + }, + { + "epoch": 0.05345609454445513, + "grad_norm": 1.4171292781829834, + "learning_rate": 2.6717032967032968e-05, + "loss": 1.1191, + "step": 778 + }, + { + "epoch": 0.05352480417754569, + "grad_norm": 1.5520141124725342, + "learning_rate": 2.6751373626373626e-05, + "loss": 1.1548, + "step": 779 + }, + { + "epoch": 0.05359351381063625, + "grad_norm": 1.4213820695877075, + "learning_rate": 2.6785714285714288e-05, + "loss": 1.1111, + "step": 780 + }, + { + "epoch": 0.05366222344372681, + "grad_norm": 1.4651081562042236, + "learning_rate": 2.6820054945054946e-05, + "loss": 1.1144, + "step": 781 + }, + { + "epoch": 0.053730933076817367, + "grad_norm": 1.5782885551452637, + "learning_rate": 2.6854395604395604e-05, + "loss": 1.1435, + "step": 782 + }, + { + "epoch": 0.053799642709907926, + "grad_norm": 1.3264998197555542, + "learning_rate": 2.6888736263736265e-05, + "loss": 1.0854, + "step": 783 + }, + { + "epoch": 0.053868352342998485, + "grad_norm": 1.4412413835525513, + "learning_rate": 2.6923076923076923e-05, + "loss": 1.123, + "step": 784 + }, + { + "epoch": 0.053937061976089044, + "grad_norm": 1.4012126922607422, + "learning_rate": 2.695741758241758e-05, + "loss": 1.1074, + "step": 785 + }, + { + "epoch": 0.054005771609179604, + "grad_norm": 1.5122679471969604, + "learning_rate": 2.6991758241758243e-05, + "loss": 1.0876, + "step": 786 + }, + { + "epoch": 0.05407448124227016, + "grad_norm": 1.5978994369506836, + "learning_rate": 2.70260989010989e-05, + "loss": 1.1578, + "step": 787 + }, + { + "epoch": 0.05414319087536072, + "grad_norm": 1.5070650577545166, + "learning_rate": 2.706043956043956e-05, + "loss": 1.1275, + "step": 788 + }, + { + "epoch": 0.05421190050845128, + "grad_norm": 1.4350719451904297, + "learning_rate": 2.709478021978022e-05, + "loss": 1.0959, + "step": 789 + }, + { + "epoch": 0.05428061014154184, + "grad_norm": 1.4921575784683228, + "learning_rate": 2.712912087912088e-05, + "loss": 1.1501, + "step": 790 + }, + { + "epoch": 0.05434931977463241, + "grad_norm": 1.5572041273117065, + "learning_rate": 2.7163461538461536e-05, + "loss": 1.1596, + "step": 791 + }, + { + "epoch": 0.054418029407722966, + "grad_norm": 1.4882301092147827, + "learning_rate": 2.7197802197802198e-05, + "loss": 1.1015, + "step": 792 + }, + { + "epoch": 0.054486739040813525, + "grad_norm": 1.3584814071655273, + "learning_rate": 2.7232142857142856e-05, + "loss": 1.1197, + "step": 793 + }, + { + "epoch": 0.054555448673904085, + "grad_norm": 1.493331789970398, + "learning_rate": 2.7266483516483514e-05, + "loss": 1.137, + "step": 794 + }, + { + "epoch": 0.054624158306994644, + "grad_norm": 1.4200912714004517, + "learning_rate": 2.7300824175824175e-05, + "loss": 1.1291, + "step": 795 + }, + { + "epoch": 0.0546928679400852, + "grad_norm": 1.5342891216278076, + "learning_rate": 2.7335164835164834e-05, + "loss": 1.1579, + "step": 796 + }, + { + "epoch": 0.05476157757317576, + "grad_norm": 1.5382280349731445, + "learning_rate": 2.73695054945055e-05, + "loss": 1.1806, + "step": 797 + }, + { + "epoch": 0.05483028720626632, + "grad_norm": 1.483425259590149, + "learning_rate": 2.7403846153846156e-05, + "loss": 1.1553, + "step": 798 + }, + { + "epoch": 0.05489899683935688, + "grad_norm": 1.4174275398254395, + "learning_rate": 2.7438186813186818e-05, + "loss": 1.1223, + "step": 799 + }, + { + "epoch": 0.05496770647244744, + "grad_norm": 1.4763333797454834, + "learning_rate": 2.7472527472527476e-05, + "loss": 1.1072, + "step": 800 + }, + { + "epoch": 0.055036416105538, + "grad_norm": 1.4077016115188599, + "learning_rate": 2.7506868131868134e-05, + "loss": 1.1039, + "step": 801 + }, + { + "epoch": 0.05510512573862856, + "grad_norm": 1.5620847940444946, + "learning_rate": 2.7541208791208795e-05, + "loss": 1.1294, + "step": 802 + }, + { + "epoch": 0.05517383537171912, + "grad_norm": 1.536561131477356, + "learning_rate": 2.7575549450549453e-05, + "loss": 1.1388, + "step": 803 + }, + { + "epoch": 0.05524254500480968, + "grad_norm": 1.3974305391311646, + "learning_rate": 2.760989010989011e-05, + "loss": 1.0808, + "step": 804 + }, + { + "epoch": 0.05531125463790024, + "grad_norm": 1.4657763242721558, + "learning_rate": 2.7644230769230773e-05, + "loss": 1.1543, + "step": 805 + }, + { + "epoch": 0.055379964270990796, + "grad_norm": 1.5054908990859985, + "learning_rate": 2.767857142857143e-05, + "loss": 1.0502, + "step": 806 + }, + { + "epoch": 0.055448673904081355, + "grad_norm": 1.4191210269927979, + "learning_rate": 2.771291208791209e-05, + "loss": 1.0931, + "step": 807 + }, + { + "epoch": 0.055517383537171915, + "grad_norm": 1.4759290218353271, + "learning_rate": 2.774725274725275e-05, + "loss": 1.0685, + "step": 808 + }, + { + "epoch": 0.055586093170262474, + "grad_norm": 1.4544363021850586, + "learning_rate": 2.778159340659341e-05, + "loss": 1.1639, + "step": 809 + }, + { + "epoch": 0.05565480280335303, + "grad_norm": 1.3902570009231567, + "learning_rate": 2.781593406593407e-05, + "loss": 1.1108, + "step": 810 + }, + { + "epoch": 0.05572351243644359, + "grad_norm": 1.4537017345428467, + "learning_rate": 2.7850274725274728e-05, + "loss": 1.1404, + "step": 811 + }, + { + "epoch": 0.05579222206953415, + "grad_norm": 1.4528374671936035, + "learning_rate": 2.7884615384615386e-05, + "loss": 1.1038, + "step": 812 + }, + { + "epoch": 0.05586093170262471, + "grad_norm": 1.4246433973312378, + "learning_rate": 2.7918956043956048e-05, + "loss": 1.1084, + "step": 813 + }, + { + "epoch": 0.05592964133571527, + "grad_norm": 1.4498121738433838, + "learning_rate": 2.7953296703296706e-05, + "loss": 1.1655, + "step": 814 + }, + { + "epoch": 0.05599835096880583, + "grad_norm": 1.4160346984863281, + "learning_rate": 2.7987637362637364e-05, + "loss": 1.1415, + "step": 815 + }, + { + "epoch": 0.05606706060189639, + "grad_norm": 1.5895782709121704, + "learning_rate": 2.8021978021978025e-05, + "loss": 1.1495, + "step": 816 + }, + { + "epoch": 0.05613577023498695, + "grad_norm": 1.4746981859207153, + "learning_rate": 2.8056318681318683e-05, + "loss": 1.0675, + "step": 817 + }, + { + "epoch": 0.05620447986807751, + "grad_norm": 1.4312578439712524, + "learning_rate": 2.809065934065934e-05, + "loss": 1.0769, + "step": 818 + }, + { + "epoch": 0.05627318950116807, + "grad_norm": 1.416415810585022, + "learning_rate": 2.8125000000000003e-05, + "loss": 1.1179, + "step": 819 + }, + { + "epoch": 0.056341899134258626, + "grad_norm": 1.4548060894012451, + "learning_rate": 2.815934065934066e-05, + "loss": 1.1022, + "step": 820 + }, + { + "epoch": 0.056410608767349185, + "grad_norm": 1.4552110433578491, + "learning_rate": 2.819368131868132e-05, + "loss": 1.1328, + "step": 821 + }, + { + "epoch": 0.056479318400439744, + "grad_norm": 1.5629949569702148, + "learning_rate": 2.822802197802198e-05, + "loss": 1.1008, + "step": 822 + }, + { + "epoch": 0.056548028033530304, + "grad_norm": 1.4156683683395386, + "learning_rate": 2.8262362637362638e-05, + "loss": 1.0929, + "step": 823 + }, + { + "epoch": 0.05661673766662086, + "grad_norm": 1.4376745223999023, + "learning_rate": 2.8296703296703296e-05, + "loss": 1.1322, + "step": 824 + }, + { + "epoch": 0.05668544729971142, + "grad_norm": 1.4272054433822632, + "learning_rate": 2.8331043956043958e-05, + "loss": 1.1284, + "step": 825 + }, + { + "epoch": 0.05675415693280198, + "grad_norm": 1.4059700965881348, + "learning_rate": 2.8365384615384616e-05, + "loss": 1.1357, + "step": 826 + }, + { + "epoch": 0.05682286656589254, + "grad_norm": 1.4463239908218384, + "learning_rate": 2.8399725274725274e-05, + "loss": 1.1025, + "step": 827 + }, + { + "epoch": 0.0568915761989831, + "grad_norm": 1.3251619338989258, + "learning_rate": 2.8434065934065935e-05, + "loss": 1.0829, + "step": 828 + }, + { + "epoch": 0.05696028583207366, + "grad_norm": 1.3279505968093872, + "learning_rate": 2.8468406593406593e-05, + "loss": 1.0662, + "step": 829 + }, + { + "epoch": 0.05702899546516422, + "grad_norm": 1.4810810089111328, + "learning_rate": 2.850274725274725e-05, + "loss": 1.0903, + "step": 830 + }, + { + "epoch": 0.05709770509825478, + "grad_norm": 1.345834493637085, + "learning_rate": 2.8537087912087913e-05, + "loss": 1.1272, + "step": 831 + }, + { + "epoch": 0.05716641473134534, + "grad_norm": 1.4236658811569214, + "learning_rate": 2.857142857142857e-05, + "loss": 1.1296, + "step": 832 + }, + { + "epoch": 0.057235124364435896, + "grad_norm": 1.408970594406128, + "learning_rate": 2.860576923076923e-05, + "loss": 1.067, + "step": 833 + }, + { + "epoch": 0.057303833997526456, + "grad_norm": 1.5834919214248657, + "learning_rate": 2.864010989010989e-05, + "loss": 1.1255, + "step": 834 + }, + { + "epoch": 0.057372543630617015, + "grad_norm": 1.386441707611084, + "learning_rate": 2.867445054945055e-05, + "loss": 1.0891, + "step": 835 + }, + { + "epoch": 0.057441253263707574, + "grad_norm": 1.4311716556549072, + "learning_rate": 2.8708791208791206e-05, + "loss": 1.1451, + "step": 836 + }, + { + "epoch": 0.057509962896798134, + "grad_norm": 1.604468822479248, + "learning_rate": 2.8743131868131868e-05, + "loss": 1.133, + "step": 837 + }, + { + "epoch": 0.05757867252988869, + "grad_norm": 1.4489911794662476, + "learning_rate": 2.8777472527472526e-05, + "loss": 1.0764, + "step": 838 + }, + { + "epoch": 0.05764738216297925, + "grad_norm": 1.4453843832015991, + "learning_rate": 2.8811813186813187e-05, + "loss": 1.124, + "step": 839 + }, + { + "epoch": 0.05771609179606981, + "grad_norm": 1.4595693349838257, + "learning_rate": 2.8846153846153845e-05, + "loss": 1.1147, + "step": 840 + }, + { + "epoch": 0.05778480142916037, + "grad_norm": 1.469038724899292, + "learning_rate": 2.8880494505494504e-05, + "loss": 1.076, + "step": 841 + }, + { + "epoch": 0.05785351106225093, + "grad_norm": 1.4464813470840454, + "learning_rate": 2.891483516483517e-05, + "loss": 1.1113, + "step": 842 + }, + { + "epoch": 0.05792222069534149, + "grad_norm": 1.4939638376235962, + "learning_rate": 2.894917582417583e-05, + "loss": 1.1331, + "step": 843 + }, + { + "epoch": 0.05799093032843205, + "grad_norm": 1.4726017713546753, + "learning_rate": 2.8983516483516488e-05, + "loss": 1.113, + "step": 844 + }, + { + "epoch": 0.05805963996152261, + "grad_norm": 1.3222732543945312, + "learning_rate": 2.9017857142857146e-05, + "loss": 1.1, + "step": 845 + }, + { + "epoch": 0.05812834959461317, + "grad_norm": 1.5053627490997314, + "learning_rate": 2.9052197802197807e-05, + "loss": 1.1357, + "step": 846 + }, + { + "epoch": 0.058197059227703726, + "grad_norm": 1.4681017398834229, + "learning_rate": 2.9086538461538465e-05, + "loss": 1.0848, + "step": 847 + }, + { + "epoch": 0.058265768860794286, + "grad_norm": 1.4774632453918457, + "learning_rate": 2.9120879120879123e-05, + "loss": 1.1093, + "step": 848 + }, + { + "epoch": 0.058334478493884845, + "grad_norm": 1.4523344039916992, + "learning_rate": 2.9155219780219785e-05, + "loss": 1.1047, + "step": 849 + }, + { + "epoch": 0.058403188126975404, + "grad_norm": 1.433306336402893, + "learning_rate": 2.9189560439560443e-05, + "loss": 1.1378, + "step": 850 + }, + { + "epoch": 0.05847189776006596, + "grad_norm": 1.4135252237319946, + "learning_rate": 2.92239010989011e-05, + "loss": 1.0792, + "step": 851 + }, + { + "epoch": 0.05854060739315652, + "grad_norm": 1.4103024005889893, + "learning_rate": 2.9258241758241762e-05, + "loss": 1.1032, + "step": 852 + }, + { + "epoch": 0.05860931702624708, + "grad_norm": 1.5433036088943481, + "learning_rate": 2.929258241758242e-05, + "loss": 1.1475, + "step": 853 + }, + { + "epoch": 0.05867802665933764, + "grad_norm": 1.536545991897583, + "learning_rate": 2.932692307692308e-05, + "loss": 1.0551, + "step": 854 + }, + { + "epoch": 0.0587467362924282, + "grad_norm": 1.3682740926742554, + "learning_rate": 2.936126373626374e-05, + "loss": 1.106, + "step": 855 + }, + { + "epoch": 0.05881544592551876, + "grad_norm": 1.4626885652542114, + "learning_rate": 2.9395604395604398e-05, + "loss": 1.1194, + "step": 856 + }, + { + "epoch": 0.05888415555860932, + "grad_norm": 1.4672825336456299, + "learning_rate": 2.9429945054945056e-05, + "loss": 1.1136, + "step": 857 + }, + { + "epoch": 0.05895286519169988, + "grad_norm": 1.4086077213287354, + "learning_rate": 2.9464285714285718e-05, + "loss": 1.0754, + "step": 858 + }, + { + "epoch": 0.05902157482479044, + "grad_norm": 1.5354646444320679, + "learning_rate": 2.9498626373626376e-05, + "loss": 1.1056, + "step": 859 + }, + { + "epoch": 0.059090284457881, + "grad_norm": 1.4741487503051758, + "learning_rate": 2.9532967032967034e-05, + "loss": 1.1119, + "step": 860 + }, + { + "epoch": 0.059158994090971556, + "grad_norm": 1.4652661085128784, + "learning_rate": 2.9567307692307695e-05, + "loss": 1.0953, + "step": 861 + }, + { + "epoch": 0.059227703724062115, + "grad_norm": 1.5189121961593628, + "learning_rate": 2.9601648351648353e-05, + "loss": 1.1103, + "step": 862 + }, + { + "epoch": 0.059296413357152675, + "grad_norm": 1.353331446647644, + "learning_rate": 2.963598901098901e-05, + "loss": 1.1234, + "step": 863 + }, + { + "epoch": 0.059365122990243234, + "grad_norm": 1.503118634223938, + "learning_rate": 2.9670329670329673e-05, + "loss": 1.0878, + "step": 864 + }, + { + "epoch": 0.05943383262333379, + "grad_norm": 1.4193084239959717, + "learning_rate": 2.970467032967033e-05, + "loss": 1.1248, + "step": 865 + }, + { + "epoch": 0.05950254225642435, + "grad_norm": 1.4480834007263184, + "learning_rate": 2.973901098901099e-05, + "loss": 1.112, + "step": 866 + }, + { + "epoch": 0.05957125188951491, + "grad_norm": 1.3487073183059692, + "learning_rate": 2.977335164835165e-05, + "loss": 1.1106, + "step": 867 + }, + { + "epoch": 0.05963996152260547, + "grad_norm": 1.4660007953643799, + "learning_rate": 2.9807692307692308e-05, + "loss": 1.1367, + "step": 868 + }, + { + "epoch": 0.05970867115569603, + "grad_norm": 1.4333559274673462, + "learning_rate": 2.9842032967032966e-05, + "loss": 1.0367, + "step": 869 + }, + { + "epoch": 0.05977738078878659, + "grad_norm": 1.4390825033187866, + "learning_rate": 2.9876373626373628e-05, + "loss": 1.1559, + "step": 870 + }, + { + "epoch": 0.05984609042187715, + "grad_norm": 1.4467182159423828, + "learning_rate": 2.9910714285714286e-05, + "loss": 1.0673, + "step": 871 + }, + { + "epoch": 0.05991480005496771, + "grad_norm": 1.4348983764648438, + "learning_rate": 2.9945054945054947e-05, + "loss": 1.0941, + "step": 872 + }, + { + "epoch": 0.05998350968805827, + "grad_norm": 1.528368353843689, + "learning_rate": 2.9979395604395605e-05, + "loss": 1.1307, + "step": 873 + }, + { + "epoch": 0.06005221932114883, + "grad_norm": 1.4063431024551392, + "learning_rate": 3.0013736263736263e-05, + "loss": 1.1402, + "step": 874 + }, + { + "epoch": 0.060120928954239386, + "grad_norm": 1.3995405435562134, + "learning_rate": 3.0048076923076925e-05, + "loss": 1.1025, + "step": 875 + }, + { + "epoch": 0.060189638587329945, + "grad_norm": 1.4533464908599854, + "learning_rate": 3.0082417582417583e-05, + "loss": 1.1444, + "step": 876 + }, + { + "epoch": 0.060258348220420505, + "grad_norm": 1.4461671113967896, + "learning_rate": 3.011675824175824e-05, + "loss": 1.1026, + "step": 877 + }, + { + "epoch": 0.060327057853511064, + "grad_norm": 1.446152925491333, + "learning_rate": 3.0151098901098902e-05, + "loss": 1.097, + "step": 878 + }, + { + "epoch": 0.06039576748660162, + "grad_norm": 1.4517766237258911, + "learning_rate": 3.018543956043956e-05, + "loss": 1.1706, + "step": 879 + }, + { + "epoch": 0.06046447711969218, + "grad_norm": 1.5013878345489502, + "learning_rate": 3.021978021978022e-05, + "loss": 1.1359, + "step": 880 + }, + { + "epoch": 0.06053318675278274, + "grad_norm": 1.4268689155578613, + "learning_rate": 3.025412087912088e-05, + "loss": 1.101, + "step": 881 + }, + { + "epoch": 0.0606018963858733, + "grad_norm": 1.4781041145324707, + "learning_rate": 3.0288461538461538e-05, + "loss": 1.1156, + "step": 882 + }, + { + "epoch": 0.06067060601896386, + "grad_norm": 1.4854310750961304, + "learning_rate": 3.0322802197802196e-05, + "loss": 1.1145, + "step": 883 + }, + { + "epoch": 0.06073931565205442, + "grad_norm": 1.381654977798462, + "learning_rate": 3.0357142857142857e-05, + "loss": 1.0669, + "step": 884 + }, + { + "epoch": 0.06080802528514498, + "grad_norm": 1.364824891090393, + "learning_rate": 3.0391483516483515e-05, + "loss": 1.1188, + "step": 885 + }, + { + "epoch": 0.06087673491823554, + "grad_norm": 1.3482393026351929, + "learning_rate": 3.0425824175824173e-05, + "loss": 1.0997, + "step": 886 + }, + { + "epoch": 0.0609454445513261, + "grad_norm": 1.4068613052368164, + "learning_rate": 3.0460164835164835e-05, + "loss": 1.1267, + "step": 887 + }, + { + "epoch": 0.06101415418441666, + "grad_norm": 1.431233286857605, + "learning_rate": 3.04945054945055e-05, + "loss": 1.0971, + "step": 888 + }, + { + "epoch": 0.061082863817507216, + "grad_norm": 1.4806586503982544, + "learning_rate": 3.052884615384616e-05, + "loss": 1.1725, + "step": 889 + }, + { + "epoch": 0.061151573450597775, + "grad_norm": 1.4754273891448975, + "learning_rate": 3.056318681318682e-05, + "loss": 1.1392, + "step": 890 + }, + { + "epoch": 0.061220283083688334, + "grad_norm": 1.4816205501556396, + "learning_rate": 3.0597527472527474e-05, + "loss": 1.1166, + "step": 891 + }, + { + "epoch": 0.061288992716778894, + "grad_norm": 1.399486780166626, + "learning_rate": 3.0631868131868135e-05, + "loss": 1.1288, + "step": 892 + }, + { + "epoch": 0.06135770234986945, + "grad_norm": 1.3433797359466553, + "learning_rate": 3.06662087912088e-05, + "loss": 1.0733, + "step": 893 + }, + { + "epoch": 0.06142641198296001, + "grad_norm": 1.2972005605697632, + "learning_rate": 3.070054945054945e-05, + "loss": 1.1394, + "step": 894 + }, + { + "epoch": 0.06149512161605057, + "grad_norm": 1.4313923120498657, + "learning_rate": 3.073489010989011e-05, + "loss": 1.0861, + "step": 895 + }, + { + "epoch": 0.06156383124914113, + "grad_norm": 1.447175145149231, + "learning_rate": 3.0769230769230774e-05, + "loss": 1.0655, + "step": 896 + }, + { + "epoch": 0.06163254088223169, + "grad_norm": 1.4825847148895264, + "learning_rate": 3.080357142857143e-05, + "loss": 1.1264, + "step": 897 + }, + { + "epoch": 0.06170125051532225, + "grad_norm": 1.353255033493042, + "learning_rate": 3.083791208791209e-05, + "loss": 1.1506, + "step": 898 + }, + { + "epoch": 0.06176996014841281, + "grad_norm": 1.3968778848648071, + "learning_rate": 3.087225274725275e-05, + "loss": 1.1065, + "step": 899 + }, + { + "epoch": 0.06183866978150337, + "grad_norm": 1.5221953392028809, + "learning_rate": 3.0906593406593407e-05, + "loss": 1.0727, + "step": 900 + }, + { + "epoch": 0.06190737941459393, + "grad_norm": 1.562593936920166, + "learning_rate": 3.094093406593407e-05, + "loss": 1.1074, + "step": 901 + }, + { + "epoch": 0.061976089047684486, + "grad_norm": 1.581390142440796, + "learning_rate": 3.097527472527473e-05, + "loss": 1.1454, + "step": 902 + }, + { + "epoch": 0.062044798680775046, + "grad_norm": 1.2937110662460327, + "learning_rate": 3.1009615384615384e-05, + "loss": 1.1128, + "step": 903 + }, + { + "epoch": 0.062113508313865605, + "grad_norm": 1.3846871852874756, + "learning_rate": 3.1043956043956046e-05, + "loss": 1.0948, + "step": 904 + }, + { + "epoch": 0.062182217946956164, + "grad_norm": 1.4935693740844727, + "learning_rate": 3.107829670329671e-05, + "loss": 1.0705, + "step": 905 + }, + { + "epoch": 0.062250927580046723, + "grad_norm": 1.449673056602478, + "learning_rate": 3.111263736263736e-05, + "loss": 1.0797, + "step": 906 + }, + { + "epoch": 0.06231963721313728, + "grad_norm": 1.4412567615509033, + "learning_rate": 3.114697802197802e-05, + "loss": 1.0772, + "step": 907 + }, + { + "epoch": 0.06238834684622784, + "grad_norm": 1.3524343967437744, + "learning_rate": 3.1181318681318685e-05, + "loss": 1.0832, + "step": 908 + }, + { + "epoch": 0.0624570564793184, + "grad_norm": 1.4268931150436401, + "learning_rate": 3.121565934065934e-05, + "loss": 1.1533, + "step": 909 + }, + { + "epoch": 0.06252576611240895, + "grad_norm": 1.336808443069458, + "learning_rate": 3.125e-05, + "loss": 1.0934, + "step": 910 + }, + { + "epoch": 0.06259447574549952, + "grad_norm": 1.4370853900909424, + "learning_rate": 3.128434065934066e-05, + "loss": 1.0968, + "step": 911 + }, + { + "epoch": 0.06266318537859007, + "grad_norm": 1.3944498300552368, + "learning_rate": 3.131868131868132e-05, + "loss": 1.0908, + "step": 912 + }, + { + "epoch": 0.06273189501168064, + "grad_norm": 1.3697738647460938, + "learning_rate": 3.135302197802198e-05, + "loss": 1.1081, + "step": 913 + }, + { + "epoch": 0.06280060464477119, + "grad_norm": 1.4474891424179077, + "learning_rate": 3.138736263736264e-05, + "loss": 1.105, + "step": 914 + }, + { + "epoch": 0.06286931427786176, + "grad_norm": 1.3978513479232788, + "learning_rate": 3.1421703296703294e-05, + "loss": 1.0995, + "step": 915 + }, + { + "epoch": 0.06293802391095231, + "grad_norm": 1.4603374004364014, + "learning_rate": 3.1456043956043956e-05, + "loss": 1.136, + "step": 916 + }, + { + "epoch": 0.06300673354404288, + "grad_norm": 1.3741391897201538, + "learning_rate": 3.149038461538462e-05, + "loss": 1.1272, + "step": 917 + }, + { + "epoch": 0.06307544317713343, + "grad_norm": 1.3966724872589111, + "learning_rate": 3.152472527472528e-05, + "loss": 1.0745, + "step": 918 + }, + { + "epoch": 0.063144152810224, + "grad_norm": 1.4083391427993774, + "learning_rate": 3.155906593406593e-05, + "loss": 1.0786, + "step": 919 + }, + { + "epoch": 0.06321286244331455, + "grad_norm": 1.4186851978302002, + "learning_rate": 3.1593406593406595e-05, + "loss": 1.0906, + "step": 920 + }, + { + "epoch": 0.06328157207640511, + "grad_norm": 1.4101675748825073, + "learning_rate": 3.1627747252747256e-05, + "loss": 1.1351, + "step": 921 + }, + { + "epoch": 0.06335028170949566, + "grad_norm": 1.481859803199768, + "learning_rate": 3.166208791208791e-05, + "loss": 1.1259, + "step": 922 + }, + { + "epoch": 0.06341899134258623, + "grad_norm": 1.4539313316345215, + "learning_rate": 3.169642857142857e-05, + "loss": 1.1108, + "step": 923 + }, + { + "epoch": 0.06348770097567678, + "grad_norm": 1.3604670763015747, + "learning_rate": 3.1730769230769234e-05, + "loss": 1.102, + "step": 924 + }, + { + "epoch": 0.06355641060876735, + "grad_norm": 1.3818186521530151, + "learning_rate": 3.176510989010989e-05, + "loss": 1.1683, + "step": 925 + }, + { + "epoch": 0.0636251202418579, + "grad_norm": 1.4193683862686157, + "learning_rate": 3.179945054945055e-05, + "loss": 1.0728, + "step": 926 + }, + { + "epoch": 0.06369382987494847, + "grad_norm": 1.3873271942138672, + "learning_rate": 3.183379120879121e-05, + "loss": 1.1275, + "step": 927 + }, + { + "epoch": 0.06376253950803902, + "grad_norm": 1.4585365056991577, + "learning_rate": 3.1868131868131866e-05, + "loss": 1.1124, + "step": 928 + }, + { + "epoch": 0.06383124914112959, + "grad_norm": 1.345398187637329, + "learning_rate": 3.190247252747253e-05, + "loss": 1.1046, + "step": 929 + }, + { + "epoch": 0.06389995877422014, + "grad_norm": 1.4639475345611572, + "learning_rate": 3.193681318681319e-05, + "loss": 1.1047, + "step": 930 + }, + { + "epoch": 0.0639686684073107, + "grad_norm": 1.4185409545898438, + "learning_rate": 3.1971153846153843e-05, + "loss": 1.1003, + "step": 931 + }, + { + "epoch": 0.06403737804040126, + "grad_norm": 1.373375654220581, + "learning_rate": 3.2005494505494505e-05, + "loss": 1.109, + "step": 932 + }, + { + "epoch": 0.06410608767349182, + "grad_norm": 1.3224807977676392, + "learning_rate": 3.2039835164835166e-05, + "loss": 1.0851, + "step": 933 + }, + { + "epoch": 0.06417479730658238, + "grad_norm": 1.407913088798523, + "learning_rate": 3.207417582417583e-05, + "loss": 1.0811, + "step": 934 + }, + { + "epoch": 0.06424350693967294, + "grad_norm": 1.4234535694122314, + "learning_rate": 3.210851648351649e-05, + "loss": 1.1211, + "step": 935 + }, + { + "epoch": 0.0643122165727635, + "grad_norm": 1.4126135110855103, + "learning_rate": 3.2142857142857144e-05, + "loss": 1.1312, + "step": 936 + }, + { + "epoch": 0.06438092620585406, + "grad_norm": 1.4735673666000366, + "learning_rate": 3.2177197802197805e-05, + "loss": 1.1239, + "step": 937 + }, + { + "epoch": 0.06444963583894461, + "grad_norm": 1.3803472518920898, + "learning_rate": 3.221153846153847e-05, + "loss": 1.0811, + "step": 938 + }, + { + "epoch": 0.06451834547203518, + "grad_norm": 1.4406859874725342, + "learning_rate": 3.224587912087912e-05, + "loss": 1.0899, + "step": 939 + }, + { + "epoch": 0.06458705510512573, + "grad_norm": 1.4692795276641846, + "learning_rate": 3.228021978021978e-05, + "loss": 1.0667, + "step": 940 + }, + { + "epoch": 0.0646557647382163, + "grad_norm": 1.4070768356323242, + "learning_rate": 3.2314560439560444e-05, + "loss": 1.1192, + "step": 941 + }, + { + "epoch": 0.06472447437130685, + "grad_norm": 1.4568644762039185, + "learning_rate": 3.23489010989011e-05, + "loss": 1.0928, + "step": 942 + }, + { + "epoch": 0.06479318400439742, + "grad_norm": 1.5557680130004883, + "learning_rate": 3.238324175824176e-05, + "loss": 1.0911, + "step": 943 + }, + { + "epoch": 0.06486189363748797, + "grad_norm": 1.5251449346542358, + "learning_rate": 3.241758241758242e-05, + "loss": 1.0977, + "step": 944 + }, + { + "epoch": 0.06493060327057854, + "grad_norm": 1.4782894849777222, + "learning_rate": 3.2451923076923077e-05, + "loss": 1.1122, + "step": 945 + }, + { + "epoch": 0.06499931290366909, + "grad_norm": 1.4046821594238281, + "learning_rate": 3.248626373626374e-05, + "loss": 1.1192, + "step": 946 + }, + { + "epoch": 0.06506802253675965, + "grad_norm": 1.490224838256836, + "learning_rate": 3.25206043956044e-05, + "loss": 1.0992, + "step": 947 + }, + { + "epoch": 0.0651367321698502, + "grad_norm": 1.4943729639053345, + "learning_rate": 3.2554945054945054e-05, + "loss": 1.1313, + "step": 948 + }, + { + "epoch": 0.06520544180294077, + "grad_norm": 1.5264720916748047, + "learning_rate": 3.2589285714285716e-05, + "loss": 1.1729, + "step": 949 + }, + { + "epoch": 0.06527415143603134, + "grad_norm": 1.4615525007247925, + "learning_rate": 3.262362637362638e-05, + "loss": 1.1272, + "step": 950 + }, + { + "epoch": 0.06534286106912189, + "grad_norm": 1.4637812376022339, + "learning_rate": 3.265796703296704e-05, + "loss": 1.0971, + "step": 951 + }, + { + "epoch": 0.06541157070221246, + "grad_norm": 1.3105591535568237, + "learning_rate": 3.269230769230769e-05, + "loss": 1.0881, + "step": 952 + }, + { + "epoch": 0.06548028033530301, + "grad_norm": 1.343328833580017, + "learning_rate": 3.2726648351648355e-05, + "loss": 1.0928, + "step": 953 + }, + { + "epoch": 0.06554898996839358, + "grad_norm": 1.358388066291809, + "learning_rate": 3.2760989010989016e-05, + "loss": 1.0683, + "step": 954 + }, + { + "epoch": 0.06561769960148413, + "grad_norm": 1.4012616872787476, + "learning_rate": 3.279532967032967e-05, + "loss": 1.0848, + "step": 955 + }, + { + "epoch": 0.0656864092345747, + "grad_norm": 1.3092725276947021, + "learning_rate": 3.282967032967033e-05, + "loss": 1.14, + "step": 956 + }, + { + "epoch": 0.06575511886766525, + "grad_norm": 1.4069507122039795, + "learning_rate": 3.2864010989010994e-05, + "loss": 1.0657, + "step": 957 + }, + { + "epoch": 0.06582382850075581, + "grad_norm": 1.4943088293075562, + "learning_rate": 3.289835164835165e-05, + "loss": 1.123, + "step": 958 + }, + { + "epoch": 0.06589253813384637, + "grad_norm": 1.3950778245925903, + "learning_rate": 3.293269230769231e-05, + "loss": 1.102, + "step": 959 + }, + { + "epoch": 0.06596124776693693, + "grad_norm": 1.3530895709991455, + "learning_rate": 3.296703296703297e-05, + "loss": 1.0751, + "step": 960 + }, + { + "epoch": 0.06602995740002748, + "grad_norm": 1.4943287372589111, + "learning_rate": 3.3001373626373626e-05, + "loss": 1.0981, + "step": 961 + }, + { + "epoch": 0.06609866703311805, + "grad_norm": 1.3274608850479126, + "learning_rate": 3.303571428571429e-05, + "loss": 1.0741, + "step": 962 + }, + { + "epoch": 0.0661673766662086, + "grad_norm": 1.3561444282531738, + "learning_rate": 3.307005494505495e-05, + "loss": 1.0386, + "step": 963 + }, + { + "epoch": 0.06623608629929917, + "grad_norm": 1.4107874631881714, + "learning_rate": 3.31043956043956e-05, + "loss": 1.117, + "step": 964 + }, + { + "epoch": 0.06630479593238972, + "grad_norm": 1.4804577827453613, + "learning_rate": 3.3138736263736265e-05, + "loss": 1.1458, + "step": 965 + }, + { + "epoch": 0.06637350556548029, + "grad_norm": 1.454041600227356, + "learning_rate": 3.3173076923076926e-05, + "loss": 1.0538, + "step": 966 + }, + { + "epoch": 0.06644221519857084, + "grad_norm": 1.3747029304504395, + "learning_rate": 3.320741758241758e-05, + "loss": 1.0933, + "step": 967 + }, + { + "epoch": 0.0665109248316614, + "grad_norm": 1.4516454935073853, + "learning_rate": 3.324175824175824e-05, + "loss": 1.0631, + "step": 968 + }, + { + "epoch": 0.06657963446475196, + "grad_norm": 1.5046333074569702, + "learning_rate": 3.3276098901098904e-05, + "loss": 1.1387, + "step": 969 + }, + { + "epoch": 0.06664834409784252, + "grad_norm": 1.3914655447006226, + "learning_rate": 3.331043956043956e-05, + "loss": 1.0982, + "step": 970 + }, + { + "epoch": 0.06671705373093308, + "grad_norm": 1.3381242752075195, + "learning_rate": 3.334478021978022e-05, + "loss": 1.0885, + "step": 971 + }, + { + "epoch": 0.06678576336402364, + "grad_norm": 1.3480498790740967, + "learning_rate": 3.337912087912088e-05, + "loss": 1.0754, + "step": 972 + }, + { + "epoch": 0.0668544729971142, + "grad_norm": 1.5746835470199585, + "learning_rate": 3.3413461538461536e-05, + "loss": 1.062, + "step": 973 + }, + { + "epoch": 0.06692318263020476, + "grad_norm": 1.4684290885925293, + "learning_rate": 3.34478021978022e-05, + "loss": 1.1405, + "step": 974 + }, + { + "epoch": 0.06699189226329531, + "grad_norm": 1.3673005104064941, + "learning_rate": 3.348214285714286e-05, + "loss": 1.0914, + "step": 975 + }, + { + "epoch": 0.06706060189638588, + "grad_norm": 1.3029637336730957, + "learning_rate": 3.3516483516483513e-05, + "loss": 1.0765, + "step": 976 + }, + { + "epoch": 0.06712931152947643, + "grad_norm": 1.398276448249817, + "learning_rate": 3.3550824175824175e-05, + "loss": 1.0901, + "step": 977 + }, + { + "epoch": 0.067198021162567, + "grad_norm": 1.454602599143982, + "learning_rate": 3.3585164835164836e-05, + "loss": 1.0885, + "step": 978 + }, + { + "epoch": 0.06726673079565755, + "grad_norm": 1.4805094003677368, + "learning_rate": 3.36195054945055e-05, + "loss": 1.1184, + "step": 979 + }, + { + "epoch": 0.06733544042874812, + "grad_norm": 1.370829463005066, + "learning_rate": 3.365384615384616e-05, + "loss": 1.1083, + "step": 980 + }, + { + "epoch": 0.06740415006183867, + "grad_norm": 1.377810001373291, + "learning_rate": 3.3688186813186814e-05, + "loss": 1.0347, + "step": 981 + }, + { + "epoch": 0.06747285969492924, + "grad_norm": 1.411838173866272, + "learning_rate": 3.3722527472527475e-05, + "loss": 1.1283, + "step": 982 + }, + { + "epoch": 0.06754156932801979, + "grad_norm": 1.3672724962234497, + "learning_rate": 3.375686813186814e-05, + "loss": 1.1075, + "step": 983 + }, + { + "epoch": 0.06761027896111035, + "grad_norm": 1.3151394128799438, + "learning_rate": 3.37912087912088e-05, + "loss": 1.0681, + "step": 984 + }, + { + "epoch": 0.0676789885942009, + "grad_norm": 1.3258355855941772, + "learning_rate": 3.382554945054945e-05, + "loss": 1.1102, + "step": 985 + }, + { + "epoch": 0.06774769822729147, + "grad_norm": 1.408223271369934, + "learning_rate": 3.3859890109890114e-05, + "loss": 1.092, + "step": 986 + }, + { + "epoch": 0.06781640786038202, + "grad_norm": 1.3944084644317627, + "learning_rate": 3.3894230769230776e-05, + "loss": 1.1241, + "step": 987 + }, + { + "epoch": 0.06788511749347259, + "grad_norm": 1.3261446952819824, + "learning_rate": 3.392857142857143e-05, + "loss": 1.1093, + "step": 988 + }, + { + "epoch": 0.06795382712656314, + "grad_norm": 1.4190901517868042, + "learning_rate": 3.396291208791209e-05, + "loss": 1.1624, + "step": 989 + }, + { + "epoch": 0.06802253675965371, + "grad_norm": 1.3850319385528564, + "learning_rate": 3.399725274725275e-05, + "loss": 1.1063, + "step": 990 + }, + { + "epoch": 0.06809124639274426, + "grad_norm": 1.3559582233428955, + "learning_rate": 3.403159340659341e-05, + "loss": 1.1159, + "step": 991 + }, + { + "epoch": 0.06815995602583483, + "grad_norm": 1.3187414407730103, + "learning_rate": 3.406593406593407e-05, + "loss": 1.0957, + "step": 992 + }, + { + "epoch": 0.06822866565892538, + "grad_norm": 1.423840045928955, + "learning_rate": 3.410027472527473e-05, + "loss": 1.1303, + "step": 993 + }, + { + "epoch": 0.06829737529201595, + "grad_norm": 1.4675886631011963, + "learning_rate": 3.4134615384615386e-05, + "loss": 1.0818, + "step": 994 + }, + { + "epoch": 0.0683660849251065, + "grad_norm": 1.5575145483016968, + "learning_rate": 3.416895604395605e-05, + "loss": 1.1292, + "step": 995 + }, + { + "epoch": 0.06843479455819707, + "grad_norm": 1.397610068321228, + "learning_rate": 3.420329670329671e-05, + "loss": 1.0087, + "step": 996 + }, + { + "epoch": 0.06850350419128762, + "grad_norm": 1.33454430103302, + "learning_rate": 3.423763736263736e-05, + "loss": 1.0591, + "step": 997 + }, + { + "epoch": 0.06857221382437818, + "grad_norm": 1.3827753067016602, + "learning_rate": 3.4271978021978025e-05, + "loss": 1.1427, + "step": 998 + }, + { + "epoch": 0.06864092345746874, + "grad_norm": 1.423107624053955, + "learning_rate": 3.4306318681318686e-05, + "loss": 1.1207, + "step": 999 + }, + { + "epoch": 0.0687096330905593, + "grad_norm": 1.3693747520446777, + "learning_rate": 3.434065934065934e-05, + "loss": 1.0998, + "step": 1000 + }, + { + "epoch": 0.06877834272364985, + "grad_norm": 1.3276665210723877, + "learning_rate": 3.4375e-05, + "loss": 1.1174, + "step": 1001 + }, + { + "epoch": 0.06884705235674042, + "grad_norm": 1.3595703840255737, + "learning_rate": 3.4409340659340664e-05, + "loss": 1.0841, + "step": 1002 + }, + { + "epoch": 0.06891576198983097, + "grad_norm": 1.35532546043396, + "learning_rate": 3.444368131868132e-05, + "loss": 1.0664, + "step": 1003 + }, + { + "epoch": 0.06898447162292154, + "grad_norm": 1.3515594005584717, + "learning_rate": 3.447802197802198e-05, + "loss": 1.0631, + "step": 1004 + }, + { + "epoch": 0.06905318125601209, + "grad_norm": 1.3367459774017334, + "learning_rate": 3.451236263736264e-05, + "loss": 1.0466, + "step": 1005 + }, + { + "epoch": 0.06912189088910266, + "grad_norm": 1.3244125843048096, + "learning_rate": 3.4546703296703296e-05, + "loss": 1.1086, + "step": 1006 + }, + { + "epoch": 0.06919060052219321, + "grad_norm": 1.317345142364502, + "learning_rate": 3.458104395604396e-05, + "loss": 1.0964, + "step": 1007 + }, + { + "epoch": 0.06925931015528378, + "grad_norm": 1.331430435180664, + "learning_rate": 3.461538461538462e-05, + "loss": 1.0858, + "step": 1008 + }, + { + "epoch": 0.06932801978837433, + "grad_norm": 1.4278026819229126, + "learning_rate": 3.464972527472527e-05, + "loss": 1.0832, + "step": 1009 + }, + { + "epoch": 0.0693967294214649, + "grad_norm": 1.3889950513839722, + "learning_rate": 3.4684065934065935e-05, + "loss": 1.1333, + "step": 1010 + }, + { + "epoch": 0.06946543905455545, + "grad_norm": 1.3458361625671387, + "learning_rate": 3.4718406593406596e-05, + "loss": 1.0922, + "step": 1011 + }, + { + "epoch": 0.06953414868764601, + "grad_norm": 1.3423494100570679, + "learning_rate": 3.475274725274725e-05, + "loss": 1.1058, + "step": 1012 + }, + { + "epoch": 0.06960285832073657, + "grad_norm": 1.4420130252838135, + "learning_rate": 3.478708791208791e-05, + "loss": 1.1318, + "step": 1013 + }, + { + "epoch": 0.06967156795382713, + "grad_norm": 1.3939435482025146, + "learning_rate": 3.4821428571428574e-05, + "loss": 1.0953, + "step": 1014 + }, + { + "epoch": 0.06974027758691768, + "grad_norm": 1.3866348266601562, + "learning_rate": 3.485576923076923e-05, + "loss": 1.1069, + "step": 1015 + }, + { + "epoch": 0.06980898722000825, + "grad_norm": 1.3328083753585815, + "learning_rate": 3.489010989010989e-05, + "loss": 1.0834, + "step": 1016 + }, + { + "epoch": 0.0698776968530988, + "grad_norm": 1.370126724243164, + "learning_rate": 3.492445054945055e-05, + "loss": 1.0905, + "step": 1017 + }, + { + "epoch": 0.06994640648618937, + "grad_norm": 1.4053303003311157, + "learning_rate": 3.4958791208791206e-05, + "loss": 1.1298, + "step": 1018 + }, + { + "epoch": 0.07001511611927992, + "grad_norm": 1.4576952457427979, + "learning_rate": 3.499313186813187e-05, + "loss": 1.1418, + "step": 1019 + }, + { + "epoch": 0.07008382575237049, + "grad_norm": 1.4194189310073853, + "learning_rate": 3.502747252747253e-05, + "loss": 1.1039, + "step": 1020 + }, + { + "epoch": 0.07015253538546104, + "grad_norm": 1.329639196395874, + "learning_rate": 3.5061813186813183e-05, + "loss": 1.082, + "step": 1021 + }, + { + "epoch": 0.0702212450185516, + "grad_norm": 1.4481053352355957, + "learning_rate": 3.5096153846153845e-05, + "loss": 1.1464, + "step": 1022 + }, + { + "epoch": 0.07028995465164216, + "grad_norm": 1.388555645942688, + "learning_rate": 3.5130494505494506e-05, + "loss": 1.1095, + "step": 1023 + }, + { + "epoch": 0.07035866428473272, + "grad_norm": 1.3954246044158936, + "learning_rate": 3.516483516483517e-05, + "loss": 1.0998, + "step": 1024 + }, + { + "epoch": 0.07042737391782328, + "grad_norm": 1.36551034450531, + "learning_rate": 3.519917582417583e-05, + "loss": 1.0845, + "step": 1025 + }, + { + "epoch": 0.07049608355091384, + "grad_norm": 1.336656093597412, + "learning_rate": 3.523351648351649e-05, + "loss": 1.1362, + "step": 1026 + }, + { + "epoch": 0.0705647931840044, + "grad_norm": 1.3815699815750122, + "learning_rate": 3.5267857142857145e-05, + "loss": 1.0637, + "step": 1027 + }, + { + "epoch": 0.07063350281709496, + "grad_norm": 1.3232109546661377, + "learning_rate": 3.530219780219781e-05, + "loss": 1.0793, + "step": 1028 + }, + { + "epoch": 0.07070221245018551, + "grad_norm": 1.3314350843429565, + "learning_rate": 3.533653846153847e-05, + "loss": 1.097, + "step": 1029 + }, + { + "epoch": 0.07077092208327608, + "grad_norm": 1.3498375415802002, + "learning_rate": 3.537087912087912e-05, + "loss": 1.1043, + "step": 1030 + }, + { + "epoch": 0.07083963171636663, + "grad_norm": 1.4220689535140991, + "learning_rate": 3.5405219780219784e-05, + "loss": 1.1095, + "step": 1031 + }, + { + "epoch": 0.0709083413494572, + "grad_norm": 1.2913086414337158, + "learning_rate": 3.5439560439560446e-05, + "loss": 1.0425, + "step": 1032 + }, + { + "epoch": 0.07097705098254775, + "grad_norm": 1.3750488758087158, + "learning_rate": 3.54739010989011e-05, + "loss": 1.0838, + "step": 1033 + }, + { + "epoch": 0.07104576061563832, + "grad_norm": 1.3314766883850098, + "learning_rate": 3.550824175824176e-05, + "loss": 1.0958, + "step": 1034 + }, + { + "epoch": 0.07111447024872887, + "grad_norm": 1.3499672412872314, + "learning_rate": 3.554258241758242e-05, + "loss": 1.069, + "step": 1035 + }, + { + "epoch": 0.07118317988181944, + "grad_norm": 1.3745057582855225, + "learning_rate": 3.557692307692308e-05, + "loss": 1.0646, + "step": 1036 + }, + { + "epoch": 0.07125188951490999, + "grad_norm": 1.3761942386627197, + "learning_rate": 3.561126373626374e-05, + "loss": 1.1143, + "step": 1037 + }, + { + "epoch": 0.07132059914800055, + "grad_norm": 1.2487132549285889, + "learning_rate": 3.56456043956044e-05, + "loss": 1.1018, + "step": 1038 + }, + { + "epoch": 0.07138930878109111, + "grad_norm": 1.3698650598526, + "learning_rate": 3.5679945054945056e-05, + "loss": 1.0822, + "step": 1039 + }, + { + "epoch": 0.07145801841418167, + "grad_norm": 1.354215145111084, + "learning_rate": 3.571428571428572e-05, + "loss": 1.0682, + "step": 1040 + }, + { + "epoch": 0.07152672804727223, + "grad_norm": 1.3804718255996704, + "learning_rate": 3.574862637362638e-05, + "loss": 1.0959, + "step": 1041 + }, + { + "epoch": 0.07159543768036279, + "grad_norm": 1.2439203262329102, + "learning_rate": 3.578296703296703e-05, + "loss": 1.0378, + "step": 1042 + }, + { + "epoch": 0.07166414731345334, + "grad_norm": 1.3683545589447021, + "learning_rate": 3.5817307692307695e-05, + "loss": 1.0941, + "step": 1043 + }, + { + "epoch": 0.07173285694654391, + "grad_norm": 1.4034461975097656, + "learning_rate": 3.5851648351648356e-05, + "loss": 1.109, + "step": 1044 + }, + { + "epoch": 0.07180156657963446, + "grad_norm": 1.4001030921936035, + "learning_rate": 3.588598901098901e-05, + "loss": 1.0768, + "step": 1045 + }, + { + "epoch": 0.07187027621272503, + "grad_norm": 1.3104106187820435, + "learning_rate": 3.592032967032967e-05, + "loss": 1.1118, + "step": 1046 + }, + { + "epoch": 0.07193898584581558, + "grad_norm": 1.3062206506729126, + "learning_rate": 3.5954670329670334e-05, + "loss": 1.0966, + "step": 1047 + }, + { + "epoch": 0.07200769547890615, + "grad_norm": 1.5144338607788086, + "learning_rate": 3.598901098901099e-05, + "loss": 1.1036, + "step": 1048 + }, + { + "epoch": 0.0720764051119967, + "grad_norm": 1.444446086883545, + "learning_rate": 3.602335164835165e-05, + "loss": 1.1024, + "step": 1049 + }, + { + "epoch": 0.07214511474508727, + "grad_norm": 1.43593168258667, + "learning_rate": 3.605769230769231e-05, + "loss": 1.0926, + "step": 1050 + }, + { + "epoch": 0.07221382437817782, + "grad_norm": 1.387202262878418, + "learning_rate": 3.6092032967032966e-05, + "loss": 1.0334, + "step": 1051 + }, + { + "epoch": 0.07228253401126838, + "grad_norm": 1.4713658094406128, + "learning_rate": 3.612637362637363e-05, + "loss": 1.1145, + "step": 1052 + }, + { + "epoch": 0.07235124364435894, + "grad_norm": 1.2860442399978638, + "learning_rate": 3.616071428571429e-05, + "loss": 1.0563, + "step": 1053 + }, + { + "epoch": 0.0724199532774495, + "grad_norm": 1.320730209350586, + "learning_rate": 3.619505494505494e-05, + "loss": 1.1643, + "step": 1054 + }, + { + "epoch": 0.07248866291054006, + "grad_norm": 1.2672808170318604, + "learning_rate": 3.6229395604395605e-05, + "loss": 1.0459, + "step": 1055 + }, + { + "epoch": 0.07255737254363062, + "grad_norm": 1.2695953845977783, + "learning_rate": 3.6263736263736266e-05, + "loss": 1.0706, + "step": 1056 + }, + { + "epoch": 0.07262608217672117, + "grad_norm": 1.39935302734375, + "learning_rate": 3.629807692307692e-05, + "loss": 1.1053, + "step": 1057 + }, + { + "epoch": 0.07269479180981174, + "grad_norm": 1.3790128231048584, + "learning_rate": 3.633241758241758e-05, + "loss": 1.0796, + "step": 1058 + }, + { + "epoch": 0.07276350144290229, + "grad_norm": 1.3270244598388672, + "learning_rate": 3.6366758241758244e-05, + "loss": 1.1296, + "step": 1059 + }, + { + "epoch": 0.07283221107599286, + "grad_norm": 1.3071601390838623, + "learning_rate": 3.64010989010989e-05, + "loss": 1.0699, + "step": 1060 + }, + { + "epoch": 0.07290092070908341, + "grad_norm": 1.4042185544967651, + "learning_rate": 3.643543956043956e-05, + "loss": 1.1092, + "step": 1061 + }, + { + "epoch": 0.07296963034217398, + "grad_norm": 1.3847103118896484, + "learning_rate": 3.646978021978022e-05, + "loss": 1.1109, + "step": 1062 + }, + { + "epoch": 0.07303833997526453, + "grad_norm": 1.317400336265564, + "learning_rate": 3.6504120879120876e-05, + "loss": 1.0953, + "step": 1063 + }, + { + "epoch": 0.0731070496083551, + "grad_norm": 1.4477953910827637, + "learning_rate": 3.653846153846154e-05, + "loss": 1.0937, + "step": 1064 + }, + { + "epoch": 0.07317575924144565, + "grad_norm": 1.305516242980957, + "learning_rate": 3.65728021978022e-05, + "loss": 1.0301, + "step": 1065 + }, + { + "epoch": 0.07324446887453621, + "grad_norm": 1.3344005346298218, + "learning_rate": 3.6607142857142853e-05, + "loss": 1.1283, + "step": 1066 + }, + { + "epoch": 0.07331317850762677, + "grad_norm": 1.2268102169036865, + "learning_rate": 3.6641483516483515e-05, + "loss": 1.0545, + "step": 1067 + }, + { + "epoch": 0.07338188814071733, + "grad_norm": 1.2860010862350464, + "learning_rate": 3.6675824175824176e-05, + "loss": 1.0807, + "step": 1068 + }, + { + "epoch": 0.07345059777380789, + "grad_norm": 1.3504037857055664, + "learning_rate": 3.671016483516483e-05, + "loss": 1.0864, + "step": 1069 + }, + { + "epoch": 0.07351930740689845, + "grad_norm": 1.3357665538787842, + "learning_rate": 3.67445054945055e-05, + "loss": 1.1047, + "step": 1070 + }, + { + "epoch": 0.073588017039989, + "grad_norm": 1.2876487970352173, + "learning_rate": 3.677884615384616e-05, + "loss": 1.1054, + "step": 1071 + }, + { + "epoch": 0.07365672667307957, + "grad_norm": 1.3798362016677856, + "learning_rate": 3.6813186813186815e-05, + "loss": 1.121, + "step": 1072 + }, + { + "epoch": 0.07372543630617012, + "grad_norm": 1.381758451461792, + "learning_rate": 3.684752747252748e-05, + "loss": 1.0791, + "step": 1073 + }, + { + "epoch": 0.07379414593926069, + "grad_norm": 1.37455153465271, + "learning_rate": 3.688186813186814e-05, + "loss": 1.0861, + "step": 1074 + }, + { + "epoch": 0.07386285557235124, + "grad_norm": 1.4117423295974731, + "learning_rate": 3.691620879120879e-05, + "loss": 1.1455, + "step": 1075 + }, + { + "epoch": 0.07393156520544181, + "grad_norm": 1.3898241519927979, + "learning_rate": 3.6950549450549454e-05, + "loss": 1.0772, + "step": 1076 + }, + { + "epoch": 0.07400027483853236, + "grad_norm": 1.3814170360565186, + "learning_rate": 3.6984890109890116e-05, + "loss": 1.0969, + "step": 1077 + }, + { + "epoch": 0.07406898447162293, + "grad_norm": 1.301477074623108, + "learning_rate": 3.701923076923077e-05, + "loss": 1.0723, + "step": 1078 + }, + { + "epoch": 0.07413769410471348, + "grad_norm": 1.3751341104507446, + "learning_rate": 3.705357142857143e-05, + "loss": 1.1048, + "step": 1079 + }, + { + "epoch": 0.07420640373780404, + "grad_norm": 1.4143025875091553, + "learning_rate": 3.708791208791209e-05, + "loss": 1.1075, + "step": 1080 + }, + { + "epoch": 0.0742751133708946, + "grad_norm": 1.3410093784332275, + "learning_rate": 3.712225274725275e-05, + "loss": 1.0506, + "step": 1081 + }, + { + "epoch": 0.07434382300398516, + "grad_norm": 1.3675014972686768, + "learning_rate": 3.715659340659341e-05, + "loss": 1.097, + "step": 1082 + }, + { + "epoch": 0.07441253263707572, + "grad_norm": 1.2829482555389404, + "learning_rate": 3.719093406593407e-05, + "loss": 1.1028, + "step": 1083 + }, + { + "epoch": 0.07448124227016628, + "grad_norm": 1.2998284101486206, + "learning_rate": 3.7225274725274726e-05, + "loss": 1.064, + "step": 1084 + }, + { + "epoch": 0.07454995190325683, + "grad_norm": 1.3494220972061157, + "learning_rate": 3.725961538461539e-05, + "loss": 1.0988, + "step": 1085 + }, + { + "epoch": 0.0746186615363474, + "grad_norm": 1.3361632823944092, + "learning_rate": 3.729395604395605e-05, + "loss": 1.1033, + "step": 1086 + }, + { + "epoch": 0.07468737116943795, + "grad_norm": 1.2651543617248535, + "learning_rate": 3.73282967032967e-05, + "loss": 1.0811, + "step": 1087 + }, + { + "epoch": 0.07475608080252852, + "grad_norm": 1.2843908071517944, + "learning_rate": 3.7362637362637365e-05, + "loss": 1.1192, + "step": 1088 + }, + { + "epoch": 0.07482479043561907, + "grad_norm": 1.341558814048767, + "learning_rate": 3.7396978021978026e-05, + "loss": 1.0726, + "step": 1089 + }, + { + "epoch": 0.07489350006870964, + "grad_norm": 1.331427812576294, + "learning_rate": 3.743131868131868e-05, + "loss": 1.119, + "step": 1090 + }, + { + "epoch": 0.07496220970180019, + "grad_norm": 1.2691376209259033, + "learning_rate": 3.746565934065934e-05, + "loss": 1.0691, + "step": 1091 + }, + { + "epoch": 0.07503091933489076, + "grad_norm": 1.3998358249664307, + "learning_rate": 3.7500000000000003e-05, + "loss": 1.0922, + "step": 1092 + }, + { + "epoch": 0.07509962896798131, + "grad_norm": 1.4654971361160278, + "learning_rate": 3.753434065934066e-05, + "loss": 1.1458, + "step": 1093 + }, + { + "epoch": 0.07516833860107187, + "grad_norm": 1.3083269596099854, + "learning_rate": 3.756868131868132e-05, + "loss": 1.0999, + "step": 1094 + }, + { + "epoch": 0.07523704823416243, + "grad_norm": 1.325014352798462, + "learning_rate": 3.760302197802198e-05, + "loss": 1.105, + "step": 1095 + }, + { + "epoch": 0.07530575786725299, + "grad_norm": 1.3088865280151367, + "learning_rate": 3.7637362637362636e-05, + "loss": 1.0965, + "step": 1096 + }, + { + "epoch": 0.07537446750034354, + "grad_norm": 1.289223074913025, + "learning_rate": 3.76717032967033e-05, + "loss": 1.0891, + "step": 1097 + }, + { + "epoch": 0.07544317713343411, + "grad_norm": 1.370656967163086, + "learning_rate": 3.770604395604396e-05, + "loss": 1.1097, + "step": 1098 + }, + { + "epoch": 0.07551188676652466, + "grad_norm": 1.310691237449646, + "learning_rate": 3.774038461538461e-05, + "loss": 1.1084, + "step": 1099 + }, + { + "epoch": 0.07558059639961523, + "grad_norm": 1.3133823871612549, + "learning_rate": 3.7774725274725275e-05, + "loss": 1.0927, + "step": 1100 + }, + { + "epoch": 0.07564930603270578, + "grad_norm": 1.2718353271484375, + "learning_rate": 3.7809065934065936e-05, + "loss": 1.0828, + "step": 1101 + }, + { + "epoch": 0.07571801566579635, + "grad_norm": 1.4147344827651978, + "learning_rate": 3.784340659340659e-05, + "loss": 1.0755, + "step": 1102 + }, + { + "epoch": 0.0757867252988869, + "grad_norm": 1.318552851676941, + "learning_rate": 3.787774725274725e-05, + "loss": 1.0607, + "step": 1103 + }, + { + "epoch": 0.07585543493197747, + "grad_norm": 1.3469411134719849, + "learning_rate": 3.7912087912087914e-05, + "loss": 1.1471, + "step": 1104 + }, + { + "epoch": 0.07592414456506802, + "grad_norm": 1.4155298471450806, + "learning_rate": 3.794642857142857e-05, + "loss": 1.1054, + "step": 1105 + }, + { + "epoch": 0.07599285419815859, + "grad_norm": 1.4345290660858154, + "learning_rate": 3.798076923076923e-05, + "loss": 1.0657, + "step": 1106 + }, + { + "epoch": 0.07606156383124914, + "grad_norm": 1.3097498416900635, + "learning_rate": 3.801510989010989e-05, + "loss": 1.0722, + "step": 1107 + }, + { + "epoch": 0.0761302734643397, + "grad_norm": 1.3550337553024292, + "learning_rate": 3.804945054945055e-05, + "loss": 1.065, + "step": 1108 + }, + { + "epoch": 0.07619898309743026, + "grad_norm": 1.2843133211135864, + "learning_rate": 3.808379120879121e-05, + "loss": 1.0892, + "step": 1109 + }, + { + "epoch": 0.07626769273052082, + "grad_norm": 1.3690263032913208, + "learning_rate": 3.811813186813187e-05, + "loss": 1.0793, + "step": 1110 + }, + { + "epoch": 0.07633640236361137, + "grad_norm": 1.3067830801010132, + "learning_rate": 3.815247252747253e-05, + "loss": 1.0951, + "step": 1111 + }, + { + "epoch": 0.07640511199670194, + "grad_norm": 1.3722710609436035, + "learning_rate": 3.8186813186813185e-05, + "loss": 1.1416, + "step": 1112 + }, + { + "epoch": 0.0764738216297925, + "grad_norm": 1.3796422481536865, + "learning_rate": 3.8221153846153846e-05, + "loss": 1.0674, + "step": 1113 + }, + { + "epoch": 0.07654253126288306, + "grad_norm": 1.4271140098571777, + "learning_rate": 3.825549450549451e-05, + "loss": 1.1155, + "step": 1114 + }, + { + "epoch": 0.07661124089597361, + "grad_norm": 1.4239282608032227, + "learning_rate": 3.828983516483517e-05, + "loss": 1.0556, + "step": 1115 + }, + { + "epoch": 0.07667995052906418, + "grad_norm": 1.3916305303573608, + "learning_rate": 3.832417582417583e-05, + "loss": 1.125, + "step": 1116 + }, + { + "epoch": 0.07674866016215473, + "grad_norm": 1.3042205572128296, + "learning_rate": 3.8358516483516485e-05, + "loss": 1.0809, + "step": 1117 + }, + { + "epoch": 0.0768173697952453, + "grad_norm": 1.3731143474578857, + "learning_rate": 3.839285714285715e-05, + "loss": 1.0984, + "step": 1118 + }, + { + "epoch": 0.07688607942833585, + "grad_norm": 1.357211947441101, + "learning_rate": 3.842719780219781e-05, + "loss": 1.0733, + "step": 1119 + }, + { + "epoch": 0.07695478906142642, + "grad_norm": 1.347878336906433, + "learning_rate": 3.846153846153846e-05, + "loss": 1.0768, + "step": 1120 + }, + { + "epoch": 0.07702349869451697, + "grad_norm": 1.219785451889038, + "learning_rate": 3.8495879120879124e-05, + "loss": 1.0497, + "step": 1121 + }, + { + "epoch": 0.07709220832760753, + "grad_norm": 1.2074612379074097, + "learning_rate": 3.8530219780219786e-05, + "loss": 1.0555, + "step": 1122 + }, + { + "epoch": 0.07716091796069809, + "grad_norm": 1.3359200954437256, + "learning_rate": 3.856456043956044e-05, + "loss": 1.0765, + "step": 1123 + }, + { + "epoch": 0.07722962759378865, + "grad_norm": 1.264155626296997, + "learning_rate": 3.85989010989011e-05, + "loss": 1.0784, + "step": 1124 + }, + { + "epoch": 0.0772983372268792, + "grad_norm": 1.3957231044769287, + "learning_rate": 3.863324175824176e-05, + "loss": 1.0917, + "step": 1125 + }, + { + "epoch": 0.07736704685996977, + "grad_norm": 1.2831783294677734, + "learning_rate": 3.866758241758242e-05, + "loss": 1.0677, + "step": 1126 + }, + { + "epoch": 0.07743575649306032, + "grad_norm": 1.33376944065094, + "learning_rate": 3.870192307692308e-05, + "loss": 1.0967, + "step": 1127 + }, + { + "epoch": 0.07750446612615089, + "grad_norm": 1.2568466663360596, + "learning_rate": 3.873626373626374e-05, + "loss": 1.0446, + "step": 1128 + }, + { + "epoch": 0.07757317575924144, + "grad_norm": 1.416704773902893, + "learning_rate": 3.8770604395604396e-05, + "loss": 1.1147, + "step": 1129 + }, + { + "epoch": 0.07764188539233201, + "grad_norm": 1.341920018196106, + "learning_rate": 3.880494505494506e-05, + "loss": 1.0486, + "step": 1130 + }, + { + "epoch": 0.07771059502542256, + "grad_norm": 1.3020137548446655, + "learning_rate": 3.883928571428572e-05, + "loss": 1.0661, + "step": 1131 + }, + { + "epoch": 0.07777930465851313, + "grad_norm": 1.3581492900848389, + "learning_rate": 3.887362637362637e-05, + "loss": 1.1054, + "step": 1132 + }, + { + "epoch": 0.07784801429160368, + "grad_norm": 1.4023548364639282, + "learning_rate": 3.8907967032967034e-05, + "loss": 1.082, + "step": 1133 + }, + { + "epoch": 0.07791672392469425, + "grad_norm": 1.392538070678711, + "learning_rate": 3.8942307692307696e-05, + "loss": 1.0449, + "step": 1134 + }, + { + "epoch": 0.0779854335577848, + "grad_norm": 1.3158538341522217, + "learning_rate": 3.897664835164835e-05, + "loss": 1.0978, + "step": 1135 + }, + { + "epoch": 0.07805414319087536, + "grad_norm": 1.3221285343170166, + "learning_rate": 3.901098901098901e-05, + "loss": 1.1533, + "step": 1136 + }, + { + "epoch": 0.07812285282396592, + "grad_norm": 1.491398572921753, + "learning_rate": 3.9045329670329673e-05, + "loss": 1.1413, + "step": 1137 + }, + { + "epoch": 0.07819156245705648, + "grad_norm": 1.4224355220794678, + "learning_rate": 3.907967032967033e-05, + "loss": 1.091, + "step": 1138 + }, + { + "epoch": 0.07826027209014703, + "grad_norm": 1.406309723854065, + "learning_rate": 3.911401098901099e-05, + "loss": 1.1059, + "step": 1139 + }, + { + "epoch": 0.0783289817232376, + "grad_norm": 1.3449573516845703, + "learning_rate": 3.914835164835165e-05, + "loss": 1.0911, + "step": 1140 + }, + { + "epoch": 0.07839769135632815, + "grad_norm": 1.3929564952850342, + "learning_rate": 3.918269230769231e-05, + "loss": 1.1252, + "step": 1141 + }, + { + "epoch": 0.07846640098941872, + "grad_norm": 1.3525325059890747, + "learning_rate": 3.921703296703297e-05, + "loss": 1.0669, + "step": 1142 + }, + { + "epoch": 0.07853511062250927, + "grad_norm": 1.309613823890686, + "learning_rate": 3.925137362637363e-05, + "loss": 1.0785, + "step": 1143 + }, + { + "epoch": 0.07860382025559984, + "grad_norm": 1.314847707748413, + "learning_rate": 3.928571428571429e-05, + "loss": 1.0818, + "step": 1144 + }, + { + "epoch": 0.07867252988869039, + "grad_norm": 1.342573642730713, + "learning_rate": 3.9320054945054945e-05, + "loss": 1.1005, + "step": 1145 + }, + { + "epoch": 0.07874123952178096, + "grad_norm": 1.3224034309387207, + "learning_rate": 3.9354395604395606e-05, + "loss": 1.1366, + "step": 1146 + }, + { + "epoch": 0.07880994915487151, + "grad_norm": 1.3424675464630127, + "learning_rate": 3.938873626373627e-05, + "loss": 1.1331, + "step": 1147 + }, + { + "epoch": 0.07887865878796207, + "grad_norm": 1.2758612632751465, + "learning_rate": 3.942307692307692e-05, + "loss": 1.1159, + "step": 1148 + }, + { + "epoch": 0.07894736842105263, + "grad_norm": 1.4892581701278687, + "learning_rate": 3.9457417582417584e-05, + "loss": 1.1217, + "step": 1149 + }, + { + "epoch": 0.0790160780541432, + "grad_norm": 1.3329577445983887, + "learning_rate": 3.9491758241758245e-05, + "loss": 1.0593, + "step": 1150 + }, + { + "epoch": 0.07908478768723375, + "grad_norm": 1.4000244140625, + "learning_rate": 3.95260989010989e-05, + "loss": 1.1391, + "step": 1151 + }, + { + "epoch": 0.07915349732032431, + "grad_norm": 1.2686454057693481, + "learning_rate": 3.956043956043956e-05, + "loss": 1.0737, + "step": 1152 + }, + { + "epoch": 0.07922220695341486, + "grad_norm": 1.302539587020874, + "learning_rate": 3.959478021978022e-05, + "loss": 1.136, + "step": 1153 + }, + { + "epoch": 0.07929091658650543, + "grad_norm": 1.3268086910247803, + "learning_rate": 3.962912087912088e-05, + "loss": 1.0648, + "step": 1154 + }, + { + "epoch": 0.07935962621959598, + "grad_norm": 1.3139164447784424, + "learning_rate": 3.966346153846154e-05, + "loss": 1.1029, + "step": 1155 + }, + { + "epoch": 0.07942833585268655, + "grad_norm": 1.361253023147583, + "learning_rate": 3.96978021978022e-05, + "loss": 1.0613, + "step": 1156 + }, + { + "epoch": 0.0794970454857771, + "grad_norm": 1.375410556793213, + "learning_rate": 3.9732142857142855e-05, + "loss": 1.1224, + "step": 1157 + }, + { + "epoch": 0.07956575511886767, + "grad_norm": 1.2509381771087646, + "learning_rate": 3.9766483516483516e-05, + "loss": 1.0604, + "step": 1158 + }, + { + "epoch": 0.07963446475195822, + "grad_norm": 1.2788052558898926, + "learning_rate": 3.980082417582418e-05, + "loss": 1.0684, + "step": 1159 + }, + { + "epoch": 0.07970317438504879, + "grad_norm": 1.3083988428115845, + "learning_rate": 3.983516483516483e-05, + "loss": 1.0811, + "step": 1160 + }, + { + "epoch": 0.07977188401813934, + "grad_norm": 1.3417307138442993, + "learning_rate": 3.98695054945055e-05, + "loss": 1.0431, + "step": 1161 + }, + { + "epoch": 0.0798405936512299, + "grad_norm": 1.3022977113723755, + "learning_rate": 3.9903846153846155e-05, + "loss": 1.0653, + "step": 1162 + }, + { + "epoch": 0.07990930328432046, + "grad_norm": 1.3030470609664917, + "learning_rate": 3.993818681318682e-05, + "loss": 1.0522, + "step": 1163 + }, + { + "epoch": 0.07997801291741102, + "grad_norm": 1.3041048049926758, + "learning_rate": 3.997252747252748e-05, + "loss": 1.129, + "step": 1164 + }, + { + "epoch": 0.08004672255050158, + "grad_norm": 1.3287690877914429, + "learning_rate": 4.000686813186813e-05, + "loss": 1.1161, + "step": 1165 + }, + { + "epoch": 0.08011543218359214, + "grad_norm": 1.4290844202041626, + "learning_rate": 4.0041208791208794e-05, + "loss": 1.1227, + "step": 1166 + }, + { + "epoch": 0.0801841418166827, + "grad_norm": 1.3530848026275635, + "learning_rate": 4.0075549450549456e-05, + "loss": 1.08, + "step": 1167 + }, + { + "epoch": 0.08025285144977326, + "grad_norm": 1.3624489307403564, + "learning_rate": 4.010989010989011e-05, + "loss": 1.0965, + "step": 1168 + }, + { + "epoch": 0.08032156108286381, + "grad_norm": 1.394946575164795, + "learning_rate": 4.014423076923077e-05, + "loss": 1.1351, + "step": 1169 + }, + { + "epoch": 0.08039027071595438, + "grad_norm": 1.372463345527649, + "learning_rate": 4.017857142857143e-05, + "loss": 1.0921, + "step": 1170 + }, + { + "epoch": 0.08045898034904493, + "grad_norm": 1.2642098665237427, + "learning_rate": 4.021291208791209e-05, + "loss": 1.1291, + "step": 1171 + }, + { + "epoch": 0.0805276899821355, + "grad_norm": 1.302855372428894, + "learning_rate": 4.024725274725275e-05, + "loss": 1.0662, + "step": 1172 + }, + { + "epoch": 0.08059639961522605, + "grad_norm": 1.3996840715408325, + "learning_rate": 4.028159340659341e-05, + "loss": 1.1268, + "step": 1173 + }, + { + "epoch": 0.08066510924831662, + "grad_norm": 1.2783498764038086, + "learning_rate": 4.031593406593407e-05, + "loss": 1.0764, + "step": 1174 + }, + { + "epoch": 0.08073381888140717, + "grad_norm": 1.242422342300415, + "learning_rate": 4.035027472527473e-05, + "loss": 1.0848, + "step": 1175 + }, + { + "epoch": 0.08080252851449773, + "grad_norm": 1.4106597900390625, + "learning_rate": 4.038461538461539e-05, + "loss": 1.1063, + "step": 1176 + }, + { + "epoch": 0.08087123814758829, + "grad_norm": 1.39316725730896, + "learning_rate": 4.041895604395605e-05, + "loss": 1.0974, + "step": 1177 + }, + { + "epoch": 0.08093994778067885, + "grad_norm": 1.273749828338623, + "learning_rate": 4.0453296703296704e-05, + "loss": 1.1124, + "step": 1178 + }, + { + "epoch": 0.0810086574137694, + "grad_norm": 1.4128731489181519, + "learning_rate": 4.0487637362637366e-05, + "loss": 1.1104, + "step": 1179 + }, + { + "epoch": 0.08107736704685997, + "grad_norm": 1.4279683828353882, + "learning_rate": 4.052197802197803e-05, + "loss": 1.1478, + "step": 1180 + }, + { + "epoch": 0.08114607667995052, + "grad_norm": 1.336701512336731, + "learning_rate": 4.055631868131868e-05, + "loss": 1.0546, + "step": 1181 + }, + { + "epoch": 0.08121478631304109, + "grad_norm": 1.3396425247192383, + "learning_rate": 4.0590659340659343e-05, + "loss": 1.0772, + "step": 1182 + }, + { + "epoch": 0.08128349594613164, + "grad_norm": 1.3063268661499023, + "learning_rate": 4.0625000000000005e-05, + "loss": 1.0962, + "step": 1183 + }, + { + "epoch": 0.08135220557922221, + "grad_norm": 1.4072928428649902, + "learning_rate": 4.065934065934066e-05, + "loss": 1.0658, + "step": 1184 + }, + { + "epoch": 0.08142091521231276, + "grad_norm": 1.2798244953155518, + "learning_rate": 4.069368131868132e-05, + "loss": 1.0464, + "step": 1185 + }, + { + "epoch": 0.08148962484540333, + "grad_norm": 1.235996127128601, + "learning_rate": 4.072802197802198e-05, + "loss": 1.1133, + "step": 1186 + }, + { + "epoch": 0.08155833447849388, + "grad_norm": 1.2953678369522095, + "learning_rate": 4.076236263736264e-05, + "loss": 1.1237, + "step": 1187 + }, + { + "epoch": 0.08162704411158445, + "grad_norm": 1.262385606765747, + "learning_rate": 4.07967032967033e-05, + "loss": 1.0712, + "step": 1188 + }, + { + "epoch": 0.081695753744675, + "grad_norm": 1.3755881786346436, + "learning_rate": 4.083104395604396e-05, + "loss": 1.0769, + "step": 1189 + }, + { + "epoch": 0.08176446337776556, + "grad_norm": 1.2263274192810059, + "learning_rate": 4.0865384615384615e-05, + "loss": 1.0423, + "step": 1190 + }, + { + "epoch": 0.08183317301085612, + "grad_norm": 1.4322693347930908, + "learning_rate": 4.0899725274725276e-05, + "loss": 1.119, + "step": 1191 + }, + { + "epoch": 0.08190188264394668, + "grad_norm": 1.2918741703033447, + "learning_rate": 4.093406593406594e-05, + "loss": 1.0541, + "step": 1192 + }, + { + "epoch": 0.08197059227703724, + "grad_norm": 1.3548181056976318, + "learning_rate": 4.096840659340659e-05, + "loss": 1.0499, + "step": 1193 + }, + { + "epoch": 0.0820393019101278, + "grad_norm": 1.392324686050415, + "learning_rate": 4.1002747252747254e-05, + "loss": 1.1363, + "step": 1194 + }, + { + "epoch": 0.08210801154321835, + "grad_norm": 1.3255046606063843, + "learning_rate": 4.1037087912087915e-05, + "loss": 1.0969, + "step": 1195 + }, + { + "epoch": 0.08217672117630892, + "grad_norm": 1.3309195041656494, + "learning_rate": 4.107142857142857e-05, + "loss": 1.1081, + "step": 1196 + }, + { + "epoch": 0.08224543080939947, + "grad_norm": 1.2782918214797974, + "learning_rate": 4.110576923076923e-05, + "loss": 1.027, + "step": 1197 + }, + { + "epoch": 0.08231414044249004, + "grad_norm": 1.3891754150390625, + "learning_rate": 4.114010989010989e-05, + "loss": 1.091, + "step": 1198 + }, + { + "epoch": 0.08238285007558059, + "grad_norm": 1.4215075969696045, + "learning_rate": 4.117445054945055e-05, + "loss": 1.108, + "step": 1199 + }, + { + "epoch": 0.08245155970867116, + "grad_norm": 1.2634252309799194, + "learning_rate": 4.120879120879121e-05, + "loss": 1.1124, + "step": 1200 + }, + { + "epoch": 0.08252026934176171, + "grad_norm": 1.299891471862793, + "learning_rate": 4.124313186813187e-05, + "loss": 1.0721, + "step": 1201 + }, + { + "epoch": 0.08258897897485228, + "grad_norm": 1.3797742128372192, + "learning_rate": 4.1277472527472525e-05, + "loss": 1.0613, + "step": 1202 + }, + { + "epoch": 0.08265768860794283, + "grad_norm": 1.3862735033035278, + "learning_rate": 4.1311813186813186e-05, + "loss": 1.1246, + "step": 1203 + }, + { + "epoch": 0.0827263982410334, + "grad_norm": 1.3099116086959839, + "learning_rate": 4.134615384615385e-05, + "loss": 1.0937, + "step": 1204 + }, + { + "epoch": 0.08279510787412395, + "grad_norm": 1.3757355213165283, + "learning_rate": 4.13804945054945e-05, + "loss": 1.0773, + "step": 1205 + }, + { + "epoch": 0.08286381750721451, + "grad_norm": 1.2889293432235718, + "learning_rate": 4.141483516483517e-05, + "loss": 1.0576, + "step": 1206 + }, + { + "epoch": 0.08293252714030507, + "grad_norm": 1.3284356594085693, + "learning_rate": 4.144917582417583e-05, + "loss": 1.0301, + "step": 1207 + }, + { + "epoch": 0.08300123677339563, + "grad_norm": 1.5088787078857422, + "learning_rate": 4.148351648351649e-05, + "loss": 1.1072, + "step": 1208 + }, + { + "epoch": 0.08306994640648618, + "grad_norm": 1.3909233808517456, + "learning_rate": 4.151785714285715e-05, + "loss": 1.107, + "step": 1209 + }, + { + "epoch": 0.08313865603957675, + "grad_norm": 1.366409182548523, + "learning_rate": 4.155219780219781e-05, + "loss": 1.0891, + "step": 1210 + }, + { + "epoch": 0.0832073656726673, + "grad_norm": 1.319533109664917, + "learning_rate": 4.1586538461538464e-05, + "loss": 1.0801, + "step": 1211 + }, + { + "epoch": 0.08327607530575787, + "grad_norm": 1.3336902856826782, + "learning_rate": 4.1620879120879126e-05, + "loss": 1.0925, + "step": 1212 + }, + { + "epoch": 0.08334478493884842, + "grad_norm": 1.2513045072555542, + "learning_rate": 4.165521978021979e-05, + "loss": 1.0691, + "step": 1213 + }, + { + "epoch": 0.08341349457193899, + "grad_norm": 1.3528050184249878, + "learning_rate": 4.168956043956044e-05, + "loss": 1.1339, + "step": 1214 + }, + { + "epoch": 0.08348220420502954, + "grad_norm": 1.340057611465454, + "learning_rate": 4.17239010989011e-05, + "loss": 1.0666, + "step": 1215 + }, + { + "epoch": 0.0835509138381201, + "grad_norm": 1.3060529232025146, + "learning_rate": 4.1758241758241765e-05, + "loss": 1.0487, + "step": 1216 + }, + { + "epoch": 0.08361962347121066, + "grad_norm": 1.264041543006897, + "learning_rate": 4.179258241758242e-05, + "loss": 1.0534, + "step": 1217 + }, + { + "epoch": 0.08368833310430122, + "grad_norm": 1.234014630317688, + "learning_rate": 4.182692307692308e-05, + "loss": 1.0676, + "step": 1218 + }, + { + "epoch": 0.08375704273739178, + "grad_norm": 1.3258403539657593, + "learning_rate": 4.186126373626374e-05, + "loss": 1.0885, + "step": 1219 + }, + { + "epoch": 0.08382575237048234, + "grad_norm": 1.3439202308654785, + "learning_rate": 4.18956043956044e-05, + "loss": 1.1498, + "step": 1220 + }, + { + "epoch": 0.0838944620035729, + "grad_norm": 1.3548206090927124, + "learning_rate": 4.192994505494506e-05, + "loss": 1.0752, + "step": 1221 + }, + { + "epoch": 0.08396317163666346, + "grad_norm": 1.4965053796768188, + "learning_rate": 4.196428571428572e-05, + "loss": 1.093, + "step": 1222 + }, + { + "epoch": 0.08403188126975401, + "grad_norm": 1.3878140449523926, + "learning_rate": 4.1998626373626374e-05, + "loss": 1.0655, + "step": 1223 + }, + { + "epoch": 0.08410059090284458, + "grad_norm": 1.3366225957870483, + "learning_rate": 4.2032967032967036e-05, + "loss": 1.073, + "step": 1224 + }, + { + "epoch": 0.08416930053593513, + "grad_norm": 1.4231667518615723, + "learning_rate": 4.20673076923077e-05, + "loss": 1.149, + "step": 1225 + }, + { + "epoch": 0.0842380101690257, + "grad_norm": 1.3695783615112305, + "learning_rate": 4.210164835164835e-05, + "loss": 1.1051, + "step": 1226 + }, + { + "epoch": 0.08430671980211625, + "grad_norm": 1.4094688892364502, + "learning_rate": 4.2135989010989013e-05, + "loss": 1.0942, + "step": 1227 + }, + { + "epoch": 0.08437542943520682, + "grad_norm": 1.3024629354476929, + "learning_rate": 4.2170329670329675e-05, + "loss": 1.1377, + "step": 1228 + }, + { + "epoch": 0.08444413906829737, + "grad_norm": 1.3711402416229248, + "learning_rate": 4.220467032967033e-05, + "loss": 1.0889, + "step": 1229 + }, + { + "epoch": 0.08451284870138794, + "grad_norm": 1.5331076383590698, + "learning_rate": 4.223901098901099e-05, + "loss": 1.1238, + "step": 1230 + }, + { + "epoch": 0.08458155833447849, + "grad_norm": 1.4338552951812744, + "learning_rate": 4.227335164835165e-05, + "loss": 1.1359, + "step": 1231 + }, + { + "epoch": 0.08465026796756905, + "grad_norm": 1.3040436506271362, + "learning_rate": 4.230769230769231e-05, + "loss": 1.1042, + "step": 1232 + }, + { + "epoch": 0.0847189776006596, + "grad_norm": 1.2828335762023926, + "learning_rate": 4.234203296703297e-05, + "loss": 1.0944, + "step": 1233 + }, + { + "epoch": 0.08478768723375017, + "grad_norm": 1.3610950708389282, + "learning_rate": 4.237637362637363e-05, + "loss": 1.0904, + "step": 1234 + }, + { + "epoch": 0.08485639686684072, + "grad_norm": 1.3233088254928589, + "learning_rate": 4.2410714285714285e-05, + "loss": 1.0968, + "step": 1235 + }, + { + "epoch": 0.08492510649993129, + "grad_norm": 1.2663334608078003, + "learning_rate": 4.2445054945054946e-05, + "loss": 1.0702, + "step": 1236 + }, + { + "epoch": 0.08499381613302184, + "grad_norm": 1.3334208726882935, + "learning_rate": 4.247939560439561e-05, + "loss": 1.0919, + "step": 1237 + }, + { + "epoch": 0.08506252576611241, + "grad_norm": 1.2943518161773682, + "learning_rate": 4.251373626373626e-05, + "loss": 1.0626, + "step": 1238 + }, + { + "epoch": 0.08513123539920296, + "grad_norm": 1.3223865032196045, + "learning_rate": 4.2548076923076924e-05, + "loss": 1.0713, + "step": 1239 + }, + { + "epoch": 0.08519994503229353, + "grad_norm": 1.1943161487579346, + "learning_rate": 4.2582417582417585e-05, + "loss": 1.0564, + "step": 1240 + }, + { + "epoch": 0.08526865466538408, + "grad_norm": 1.3394056558609009, + "learning_rate": 4.261675824175824e-05, + "loss": 1.0874, + "step": 1241 + }, + { + "epoch": 0.08533736429847465, + "grad_norm": 1.2182786464691162, + "learning_rate": 4.26510989010989e-05, + "loss": 1.0305, + "step": 1242 + }, + { + "epoch": 0.0854060739315652, + "grad_norm": 1.4370388984680176, + "learning_rate": 4.268543956043956e-05, + "loss": 1.1064, + "step": 1243 + }, + { + "epoch": 0.08547478356465577, + "grad_norm": 1.3128165006637573, + "learning_rate": 4.271978021978022e-05, + "loss": 1.1064, + "step": 1244 + }, + { + "epoch": 0.08554349319774632, + "grad_norm": 1.3112328052520752, + "learning_rate": 4.275412087912088e-05, + "loss": 1.1127, + "step": 1245 + }, + { + "epoch": 0.08561220283083688, + "grad_norm": 1.255954384803772, + "learning_rate": 4.278846153846154e-05, + "loss": 1.1068, + "step": 1246 + }, + { + "epoch": 0.08568091246392744, + "grad_norm": 1.2900466918945312, + "learning_rate": 4.2822802197802195e-05, + "loss": 1.1224, + "step": 1247 + }, + { + "epoch": 0.085749622097018, + "grad_norm": 1.3149195909500122, + "learning_rate": 4.2857142857142856e-05, + "loss": 1.037, + "step": 1248 + }, + { + "epoch": 0.08581833173010855, + "grad_norm": 1.3482764959335327, + "learning_rate": 4.289148351648352e-05, + "loss": 1.04, + "step": 1249 + }, + { + "epoch": 0.08588704136319912, + "grad_norm": 1.2861526012420654, + "learning_rate": 4.292582417582417e-05, + "loss": 1.0688, + "step": 1250 + }, + { + "epoch": 0.08595575099628967, + "grad_norm": 1.426600456237793, + "learning_rate": 4.2960164835164834e-05, + "loss": 1.1126, + "step": 1251 + }, + { + "epoch": 0.08602446062938024, + "grad_norm": 1.2635215520858765, + "learning_rate": 4.29945054945055e-05, + "loss": 1.0834, + "step": 1252 + }, + { + "epoch": 0.08609317026247079, + "grad_norm": 1.4234157800674438, + "learning_rate": 4.302884615384616e-05, + "loss": 1.0982, + "step": 1253 + }, + { + "epoch": 0.08616187989556136, + "grad_norm": 1.4602895975112915, + "learning_rate": 4.306318681318682e-05, + "loss": 1.0521, + "step": 1254 + }, + { + "epoch": 0.08623058952865191, + "grad_norm": 1.2740758657455444, + "learning_rate": 4.309752747252748e-05, + "loss": 1.1056, + "step": 1255 + }, + { + "epoch": 0.08629929916174248, + "grad_norm": 1.5043905973434448, + "learning_rate": 4.3131868131868134e-05, + "loss": 1.117, + "step": 1256 + }, + { + "epoch": 0.08636800879483303, + "grad_norm": 1.3626179695129395, + "learning_rate": 4.3166208791208796e-05, + "loss": 1.0788, + "step": 1257 + }, + { + "epoch": 0.0864367184279236, + "grad_norm": 1.4024180173873901, + "learning_rate": 4.320054945054946e-05, + "loss": 1.0924, + "step": 1258 + }, + { + "epoch": 0.08650542806101415, + "grad_norm": 1.2590590715408325, + "learning_rate": 4.323489010989011e-05, + "loss": 1.0639, + "step": 1259 + }, + { + "epoch": 0.08657413769410471, + "grad_norm": 1.32333242893219, + "learning_rate": 4.326923076923077e-05, + "loss": 1.0645, + "step": 1260 + }, + { + "epoch": 0.08664284732719527, + "grad_norm": 1.4754812717437744, + "learning_rate": 4.3303571428571435e-05, + "loss": 1.0967, + "step": 1261 + }, + { + "epoch": 0.08671155696028583, + "grad_norm": 1.322131872177124, + "learning_rate": 4.333791208791209e-05, + "loss": 1.0271, + "step": 1262 + }, + { + "epoch": 0.08678026659337638, + "grad_norm": 1.3303189277648926, + "learning_rate": 4.337225274725275e-05, + "loss": 1.0765, + "step": 1263 + }, + { + "epoch": 0.08684897622646695, + "grad_norm": 1.4703439474105835, + "learning_rate": 4.340659340659341e-05, + "loss": 1.0871, + "step": 1264 + }, + { + "epoch": 0.0869176858595575, + "grad_norm": 1.425412654876709, + "learning_rate": 4.344093406593407e-05, + "loss": 1.0904, + "step": 1265 + }, + { + "epoch": 0.08698639549264807, + "grad_norm": 1.3442906141281128, + "learning_rate": 4.347527472527473e-05, + "loss": 1.1211, + "step": 1266 + }, + { + "epoch": 0.08705510512573864, + "grad_norm": 1.3670949935913086, + "learning_rate": 4.350961538461539e-05, + "loss": 1.0778, + "step": 1267 + }, + { + "epoch": 0.08712381475882919, + "grad_norm": 1.3429619073867798, + "learning_rate": 4.3543956043956044e-05, + "loss": 1.0714, + "step": 1268 + }, + { + "epoch": 0.08719252439191975, + "grad_norm": 1.2746083736419678, + "learning_rate": 4.3578296703296706e-05, + "loss": 1.0553, + "step": 1269 + }, + { + "epoch": 0.0872612340250103, + "grad_norm": 1.2878304719924927, + "learning_rate": 4.361263736263737e-05, + "loss": 1.0782, + "step": 1270 + }, + { + "epoch": 0.08732994365810087, + "grad_norm": 1.3116893768310547, + "learning_rate": 4.364697802197802e-05, + "loss": 1.0506, + "step": 1271 + }, + { + "epoch": 0.08739865329119142, + "grad_norm": 1.270467758178711, + "learning_rate": 4.3681318681318683e-05, + "loss": 1.0817, + "step": 1272 + }, + { + "epoch": 0.08746736292428199, + "grad_norm": 1.2343814373016357, + "learning_rate": 4.3715659340659345e-05, + "loss": 1.0787, + "step": 1273 + }, + { + "epoch": 0.08753607255737254, + "grad_norm": 1.3938156366348267, + "learning_rate": 4.375e-05, + "loss": 1.0904, + "step": 1274 + }, + { + "epoch": 0.08760478219046311, + "grad_norm": 1.2517480850219727, + "learning_rate": 4.378434065934066e-05, + "loss": 1.0743, + "step": 1275 + }, + { + "epoch": 0.08767349182355366, + "grad_norm": 1.3994803428649902, + "learning_rate": 4.381868131868132e-05, + "loss": 1.1441, + "step": 1276 + }, + { + "epoch": 0.08774220145664423, + "grad_norm": 1.3263864517211914, + "learning_rate": 4.385302197802198e-05, + "loss": 1.0933, + "step": 1277 + }, + { + "epoch": 0.08781091108973478, + "grad_norm": 1.2801116704940796, + "learning_rate": 4.388736263736264e-05, + "loss": 1.0891, + "step": 1278 + }, + { + "epoch": 0.08787962072282535, + "grad_norm": 1.2551467418670654, + "learning_rate": 4.39217032967033e-05, + "loss": 1.0853, + "step": 1279 + }, + { + "epoch": 0.0879483303559159, + "grad_norm": 1.3759182691574097, + "learning_rate": 4.3956043956043955e-05, + "loss": 1.099, + "step": 1280 + }, + { + "epoch": 0.08801703998900647, + "grad_norm": 1.2929344177246094, + "learning_rate": 4.3990384615384616e-05, + "loss": 1.0648, + "step": 1281 + }, + { + "epoch": 0.08808574962209702, + "grad_norm": 1.2807961702346802, + "learning_rate": 4.402472527472528e-05, + "loss": 1.0437, + "step": 1282 + }, + { + "epoch": 0.08815445925518758, + "grad_norm": 1.3230775594711304, + "learning_rate": 4.405906593406593e-05, + "loss": 1.0718, + "step": 1283 + }, + { + "epoch": 0.08822316888827814, + "grad_norm": 1.314412236213684, + "learning_rate": 4.4093406593406594e-05, + "loss": 1.114, + "step": 1284 + }, + { + "epoch": 0.0882918785213687, + "grad_norm": 1.3139026165008545, + "learning_rate": 4.4127747252747255e-05, + "loss": 1.0981, + "step": 1285 + }, + { + "epoch": 0.08836058815445925, + "grad_norm": 1.3282109498977661, + "learning_rate": 4.416208791208791e-05, + "loss": 1.0319, + "step": 1286 + }, + { + "epoch": 0.08842929778754982, + "grad_norm": 1.4454092979431152, + "learning_rate": 4.419642857142857e-05, + "loss": 1.0404, + "step": 1287 + }, + { + "epoch": 0.08849800742064037, + "grad_norm": 1.3889720439910889, + "learning_rate": 4.423076923076923e-05, + "loss": 1.074, + "step": 1288 + }, + { + "epoch": 0.08856671705373094, + "grad_norm": 1.3127012252807617, + "learning_rate": 4.426510989010989e-05, + "loss": 1.0433, + "step": 1289 + }, + { + "epoch": 0.08863542668682149, + "grad_norm": 1.3362138271331787, + "learning_rate": 4.429945054945055e-05, + "loss": 1.0846, + "step": 1290 + }, + { + "epoch": 0.08870413631991206, + "grad_norm": 1.2667760848999023, + "learning_rate": 4.433379120879121e-05, + "loss": 1.0513, + "step": 1291 + }, + { + "epoch": 0.08877284595300261, + "grad_norm": 1.29586923122406, + "learning_rate": 4.4368131868131865e-05, + "loss": 1.0663, + "step": 1292 + }, + { + "epoch": 0.08884155558609318, + "grad_norm": 1.3126155138015747, + "learning_rate": 4.4402472527472526e-05, + "loss": 1.1122, + "step": 1293 + }, + { + "epoch": 0.08891026521918373, + "grad_norm": 1.352506399154663, + "learning_rate": 4.443681318681319e-05, + "loss": 1.0536, + "step": 1294 + }, + { + "epoch": 0.0889789748522743, + "grad_norm": 1.3374871015548706, + "learning_rate": 4.447115384615384e-05, + "loss": 1.0813, + "step": 1295 + }, + { + "epoch": 0.08904768448536485, + "grad_norm": 1.3383516073226929, + "learning_rate": 4.4505494505494504e-05, + "loss": 1.0921, + "step": 1296 + }, + { + "epoch": 0.08911639411845541, + "grad_norm": 1.3239001035690308, + "learning_rate": 4.453983516483517e-05, + "loss": 1.0578, + "step": 1297 + }, + { + "epoch": 0.08918510375154597, + "grad_norm": 1.3340544700622559, + "learning_rate": 4.457417582417583e-05, + "loss": 1.1472, + "step": 1298 + }, + { + "epoch": 0.08925381338463653, + "grad_norm": 1.2395721673965454, + "learning_rate": 4.460851648351649e-05, + "loss": 1.073, + "step": 1299 + }, + { + "epoch": 0.08932252301772708, + "grad_norm": 1.2405164241790771, + "learning_rate": 4.464285714285715e-05, + "loss": 1.057, + "step": 1300 + }, + { + "epoch": 0.08939123265081765, + "grad_norm": 1.2760014533996582, + "learning_rate": 4.4677197802197804e-05, + "loss": 1.0375, + "step": 1301 + }, + { + "epoch": 0.0894599422839082, + "grad_norm": 1.4255304336547852, + "learning_rate": 4.4711538461538466e-05, + "loss": 1.0943, + "step": 1302 + }, + { + "epoch": 0.08952865191699877, + "grad_norm": 1.3272331953048706, + "learning_rate": 4.474587912087913e-05, + "loss": 1.0258, + "step": 1303 + }, + { + "epoch": 0.08959736155008932, + "grad_norm": 1.3840243816375732, + "learning_rate": 4.478021978021978e-05, + "loss": 1.0918, + "step": 1304 + }, + { + "epoch": 0.08966607118317989, + "grad_norm": 1.3899037837982178, + "learning_rate": 4.481456043956044e-05, + "loss": 1.0636, + "step": 1305 + }, + { + "epoch": 0.08973478081627044, + "grad_norm": 1.255737066268921, + "learning_rate": 4.4848901098901105e-05, + "loss": 1.0865, + "step": 1306 + }, + { + "epoch": 0.089803490449361, + "grad_norm": 1.2357064485549927, + "learning_rate": 4.488324175824176e-05, + "loss": 1.1004, + "step": 1307 + }, + { + "epoch": 0.08987220008245156, + "grad_norm": 1.354512095451355, + "learning_rate": 4.491758241758242e-05, + "loss": 1.0808, + "step": 1308 + }, + { + "epoch": 0.08994090971554212, + "grad_norm": 1.388146162033081, + "learning_rate": 4.495192307692308e-05, + "loss": 1.1332, + "step": 1309 + }, + { + "epoch": 0.09000961934863268, + "grad_norm": 1.3441296815872192, + "learning_rate": 4.498626373626374e-05, + "loss": 1.0788, + "step": 1310 + }, + { + "epoch": 0.09007832898172324, + "grad_norm": 1.3285506963729858, + "learning_rate": 4.50206043956044e-05, + "loss": 1.0655, + "step": 1311 + }, + { + "epoch": 0.0901470386148138, + "grad_norm": 1.2609862089157104, + "learning_rate": 4.505494505494506e-05, + "loss": 1.0868, + "step": 1312 + }, + { + "epoch": 0.09021574824790436, + "grad_norm": 1.3951236009597778, + "learning_rate": 4.5089285714285714e-05, + "loss": 1.075, + "step": 1313 + }, + { + "epoch": 0.09028445788099491, + "grad_norm": 1.3515429496765137, + "learning_rate": 4.5123626373626376e-05, + "loss": 1.0846, + "step": 1314 + }, + { + "epoch": 0.09035316751408548, + "grad_norm": 1.2488977909088135, + "learning_rate": 4.515796703296704e-05, + "loss": 1.0423, + "step": 1315 + }, + { + "epoch": 0.09042187714717603, + "grad_norm": 1.296078085899353, + "learning_rate": 4.519230769230769e-05, + "loss": 1.089, + "step": 1316 + }, + { + "epoch": 0.0904905867802666, + "grad_norm": 1.3046743869781494, + "learning_rate": 4.5226648351648353e-05, + "loss": 1.0898, + "step": 1317 + }, + { + "epoch": 0.09055929641335715, + "grad_norm": 1.3286888599395752, + "learning_rate": 4.5260989010989015e-05, + "loss": 1.0755, + "step": 1318 + }, + { + "epoch": 0.09062800604644772, + "grad_norm": 1.329409122467041, + "learning_rate": 4.529532967032967e-05, + "loss": 1.1003, + "step": 1319 + }, + { + "epoch": 0.09069671567953827, + "grad_norm": 1.2935664653778076, + "learning_rate": 4.532967032967033e-05, + "loss": 1.0413, + "step": 1320 + }, + { + "epoch": 0.09076542531262884, + "grad_norm": 1.385083794593811, + "learning_rate": 4.536401098901099e-05, + "loss": 1.0932, + "step": 1321 + }, + { + "epoch": 0.09083413494571939, + "grad_norm": 1.3373879194259644, + "learning_rate": 4.539835164835165e-05, + "loss": 1.0968, + "step": 1322 + }, + { + "epoch": 0.09090284457880995, + "grad_norm": 1.253228783607483, + "learning_rate": 4.543269230769231e-05, + "loss": 1.1192, + "step": 1323 + }, + { + "epoch": 0.09097155421190051, + "grad_norm": 1.344900369644165, + "learning_rate": 4.546703296703297e-05, + "loss": 1.0801, + "step": 1324 + }, + { + "epoch": 0.09104026384499107, + "grad_norm": 1.2347816228866577, + "learning_rate": 4.5501373626373625e-05, + "loss": 1.0659, + "step": 1325 + }, + { + "epoch": 0.09110897347808163, + "grad_norm": 1.4204108715057373, + "learning_rate": 4.5535714285714286e-05, + "loss": 1.1092, + "step": 1326 + }, + { + "epoch": 0.09117768311117219, + "grad_norm": 1.502670168876648, + "learning_rate": 4.557005494505495e-05, + "loss": 1.1082, + "step": 1327 + }, + { + "epoch": 0.09124639274426274, + "grad_norm": 1.3358792066574097, + "learning_rate": 4.56043956043956e-05, + "loss": 1.0576, + "step": 1328 + }, + { + "epoch": 0.09131510237735331, + "grad_norm": 1.3090710639953613, + "learning_rate": 4.5638736263736264e-05, + "loss": 1.0824, + "step": 1329 + }, + { + "epoch": 0.09138381201044386, + "grad_norm": 1.3204933404922485, + "learning_rate": 4.5673076923076925e-05, + "loss": 1.087, + "step": 1330 + }, + { + "epoch": 0.09145252164353443, + "grad_norm": 1.2756474018096924, + "learning_rate": 4.5707417582417587e-05, + "loss": 1.0759, + "step": 1331 + }, + { + "epoch": 0.09152123127662498, + "grad_norm": 1.3266096115112305, + "learning_rate": 4.574175824175824e-05, + "loss": 1.0507, + "step": 1332 + }, + { + "epoch": 0.09158994090971555, + "grad_norm": 1.29004967212677, + "learning_rate": 4.57760989010989e-05, + "loss": 1.1035, + "step": 1333 + }, + { + "epoch": 0.0916586505428061, + "grad_norm": 1.3374402523040771, + "learning_rate": 4.5810439560439564e-05, + "loss": 1.0665, + "step": 1334 + }, + { + "epoch": 0.09172736017589667, + "grad_norm": 1.3258659839630127, + "learning_rate": 4.584478021978022e-05, + "loss": 1.1414, + "step": 1335 + }, + { + "epoch": 0.09179606980898722, + "grad_norm": 1.3526602983474731, + "learning_rate": 4.587912087912088e-05, + "loss": 1.1019, + "step": 1336 + }, + { + "epoch": 0.09186477944207778, + "grad_norm": 1.353269338607788, + "learning_rate": 4.591346153846154e-05, + "loss": 1.0596, + "step": 1337 + }, + { + "epoch": 0.09193348907516834, + "grad_norm": 1.2739591598510742, + "learning_rate": 4.5947802197802196e-05, + "loss": 1.1306, + "step": 1338 + }, + { + "epoch": 0.0920021987082589, + "grad_norm": 1.22578763961792, + "learning_rate": 4.598214285714286e-05, + "loss": 1.0939, + "step": 1339 + }, + { + "epoch": 0.09207090834134946, + "grad_norm": 1.4014140367507935, + "learning_rate": 4.601648351648352e-05, + "loss": 1.0759, + "step": 1340 + }, + { + "epoch": 0.09213961797444002, + "grad_norm": 1.2945795059204102, + "learning_rate": 4.6050824175824174e-05, + "loss": 1.0729, + "step": 1341 + }, + { + "epoch": 0.09220832760753057, + "grad_norm": 1.3551573753356934, + "learning_rate": 4.6085164835164835e-05, + "loss": 1.0676, + "step": 1342 + }, + { + "epoch": 0.09227703724062114, + "grad_norm": 1.2853291034698486, + "learning_rate": 4.61195054945055e-05, + "loss": 1.0877, + "step": 1343 + }, + { + "epoch": 0.09234574687371169, + "grad_norm": 1.401591181755066, + "learning_rate": 4.615384615384616e-05, + "loss": 1.0555, + "step": 1344 + }, + { + "epoch": 0.09241445650680226, + "grad_norm": 1.2113499641418457, + "learning_rate": 4.618818681318682e-05, + "loss": 1.0418, + "step": 1345 + }, + { + "epoch": 0.09248316613989281, + "grad_norm": 1.38390052318573, + "learning_rate": 4.6222527472527474e-05, + "loss": 1.098, + "step": 1346 + }, + { + "epoch": 0.09255187577298338, + "grad_norm": 1.472365140914917, + "learning_rate": 4.6256868131868136e-05, + "loss": 1.1014, + "step": 1347 + }, + { + "epoch": 0.09262058540607393, + "grad_norm": 1.2844189405441284, + "learning_rate": 4.62912087912088e-05, + "loss": 1.0799, + "step": 1348 + }, + { + "epoch": 0.0926892950391645, + "grad_norm": 1.3180406093597412, + "learning_rate": 4.632554945054945e-05, + "loss": 1.0679, + "step": 1349 + }, + { + "epoch": 0.09275800467225505, + "grad_norm": 1.2763539552688599, + "learning_rate": 4.635989010989011e-05, + "loss": 1.0867, + "step": 1350 + }, + { + "epoch": 0.09282671430534561, + "grad_norm": 1.2917757034301758, + "learning_rate": 4.6394230769230775e-05, + "loss": 1.1188, + "step": 1351 + }, + { + "epoch": 0.09289542393843617, + "grad_norm": 1.317297339439392, + "learning_rate": 4.642857142857143e-05, + "loss": 1.1161, + "step": 1352 + }, + { + "epoch": 0.09296413357152673, + "grad_norm": 1.2065463066101074, + "learning_rate": 4.646291208791209e-05, + "loss": 1.0442, + "step": 1353 + }, + { + "epoch": 0.09303284320461729, + "grad_norm": 1.3690966367721558, + "learning_rate": 4.649725274725275e-05, + "loss": 1.1071, + "step": 1354 + }, + { + "epoch": 0.09310155283770785, + "grad_norm": 1.2739450931549072, + "learning_rate": 4.653159340659341e-05, + "loss": 1.0391, + "step": 1355 + }, + { + "epoch": 0.0931702624707984, + "grad_norm": 1.306848168373108, + "learning_rate": 4.656593406593407e-05, + "loss": 1.0796, + "step": 1356 + }, + { + "epoch": 0.09323897210388897, + "grad_norm": 1.2748074531555176, + "learning_rate": 4.660027472527473e-05, + "loss": 1.0877, + "step": 1357 + }, + { + "epoch": 0.09330768173697952, + "grad_norm": 1.2940200567245483, + "learning_rate": 4.6634615384615384e-05, + "loss": 1.1087, + "step": 1358 + }, + { + "epoch": 0.09337639137007009, + "grad_norm": 1.2744495868682861, + "learning_rate": 4.6668956043956046e-05, + "loss": 1.0778, + "step": 1359 + }, + { + "epoch": 0.09344510100316064, + "grad_norm": 1.2396055459976196, + "learning_rate": 4.670329670329671e-05, + "loss": 1.0686, + "step": 1360 + }, + { + "epoch": 0.09351381063625121, + "grad_norm": 1.3092446327209473, + "learning_rate": 4.673763736263736e-05, + "loss": 1.1016, + "step": 1361 + }, + { + "epoch": 0.09358252026934176, + "grad_norm": 1.1852563619613647, + "learning_rate": 4.6771978021978023e-05, + "loss": 1.04, + "step": 1362 + }, + { + "epoch": 0.09365122990243233, + "grad_norm": 1.3224940299987793, + "learning_rate": 4.6806318681318685e-05, + "loss": 1.0635, + "step": 1363 + }, + { + "epoch": 0.09371993953552288, + "grad_norm": 1.2308567762374878, + "learning_rate": 4.6840659340659346e-05, + "loss": 1.0866, + "step": 1364 + }, + { + "epoch": 0.09378864916861344, + "grad_norm": 1.3290839195251465, + "learning_rate": 4.6875e-05, + "loss": 1.0822, + "step": 1365 + }, + { + "epoch": 0.093857358801704, + "grad_norm": 1.3022931814193726, + "learning_rate": 4.690934065934066e-05, + "loss": 1.1015, + "step": 1366 + }, + { + "epoch": 0.09392606843479456, + "grad_norm": 1.3068950176239014, + "learning_rate": 4.6943681318681324e-05, + "loss": 0.9975, + "step": 1367 + }, + { + "epoch": 0.09399477806788512, + "grad_norm": 1.3149222135543823, + "learning_rate": 4.697802197802198e-05, + "loss": 1.1065, + "step": 1368 + }, + { + "epoch": 0.09406348770097568, + "grad_norm": 1.3517099618911743, + "learning_rate": 4.701236263736264e-05, + "loss": 1.0627, + "step": 1369 + }, + { + "epoch": 0.09413219733406623, + "grad_norm": 1.2750579118728638, + "learning_rate": 4.70467032967033e-05, + "loss": 1.085, + "step": 1370 + }, + { + "epoch": 0.0942009069671568, + "grad_norm": 1.2657418251037598, + "learning_rate": 4.7081043956043956e-05, + "loss": 1.0467, + "step": 1371 + }, + { + "epoch": 0.09426961660024735, + "grad_norm": 1.306312084197998, + "learning_rate": 4.711538461538462e-05, + "loss": 1.0517, + "step": 1372 + }, + { + "epoch": 0.09433832623333792, + "grad_norm": 1.2761298418045044, + "learning_rate": 4.714972527472528e-05, + "loss": 1.0864, + "step": 1373 + }, + { + "epoch": 0.09440703586642847, + "grad_norm": 1.3750017881393433, + "learning_rate": 4.7184065934065934e-05, + "loss": 1.13, + "step": 1374 + }, + { + "epoch": 0.09447574549951904, + "grad_norm": 1.307848572731018, + "learning_rate": 4.7218406593406595e-05, + "loss": 1.0694, + "step": 1375 + }, + { + "epoch": 0.09454445513260959, + "grad_norm": 1.2171320915222168, + "learning_rate": 4.7252747252747257e-05, + "loss": 1.0982, + "step": 1376 + }, + { + "epoch": 0.09461316476570016, + "grad_norm": 1.1790937185287476, + "learning_rate": 4.728708791208791e-05, + "loss": 1.0218, + "step": 1377 + }, + { + "epoch": 0.09468187439879071, + "grad_norm": 1.3411155939102173, + "learning_rate": 4.732142857142857e-05, + "loss": 1.082, + "step": 1378 + }, + { + "epoch": 0.09475058403188127, + "grad_norm": 1.2269355058670044, + "learning_rate": 4.7355769230769234e-05, + "loss": 1.1092, + "step": 1379 + }, + { + "epoch": 0.09481929366497183, + "grad_norm": 1.2654460668563843, + "learning_rate": 4.739010989010989e-05, + "loss": 1.0649, + "step": 1380 + }, + { + "epoch": 0.09488800329806239, + "grad_norm": 1.1747267246246338, + "learning_rate": 4.742445054945055e-05, + "loss": 1.0624, + "step": 1381 + }, + { + "epoch": 0.09495671293115295, + "grad_norm": 1.2789942026138306, + "learning_rate": 4.745879120879121e-05, + "loss": 1.011, + "step": 1382 + }, + { + "epoch": 0.09502542256424351, + "grad_norm": 1.2980546951293945, + "learning_rate": 4.7493131868131866e-05, + "loss": 1.0831, + "step": 1383 + }, + { + "epoch": 0.09509413219733406, + "grad_norm": 1.3734074831008911, + "learning_rate": 4.752747252747253e-05, + "loss": 1.0683, + "step": 1384 + }, + { + "epoch": 0.09516284183042463, + "grad_norm": 1.364248275756836, + "learning_rate": 4.756181318681319e-05, + "loss": 1.0214, + "step": 1385 + }, + { + "epoch": 0.09523155146351518, + "grad_norm": 1.2376508712768555, + "learning_rate": 4.7596153846153844e-05, + "loss": 1.0791, + "step": 1386 + }, + { + "epoch": 0.09530026109660575, + "grad_norm": 1.3386377096176147, + "learning_rate": 4.7630494505494505e-05, + "loss": 1.0668, + "step": 1387 + }, + { + "epoch": 0.0953689707296963, + "grad_norm": 1.223351001739502, + "learning_rate": 4.766483516483517e-05, + "loss": 1.0689, + "step": 1388 + }, + { + "epoch": 0.09543768036278687, + "grad_norm": 1.2665002346038818, + "learning_rate": 4.769917582417583e-05, + "loss": 1.0599, + "step": 1389 + }, + { + "epoch": 0.09550638999587742, + "grad_norm": 1.3234366178512573, + "learning_rate": 4.773351648351649e-05, + "loss": 1.0456, + "step": 1390 + }, + { + "epoch": 0.09557509962896799, + "grad_norm": 1.2248855829238892, + "learning_rate": 4.7767857142857144e-05, + "loss": 1.0534, + "step": 1391 + }, + { + "epoch": 0.09564380926205854, + "grad_norm": 1.2552475929260254, + "learning_rate": 4.7802197802197806e-05, + "loss": 1.0596, + "step": 1392 + }, + { + "epoch": 0.0957125188951491, + "grad_norm": 1.1932016611099243, + "learning_rate": 4.783653846153847e-05, + "loss": 1.0444, + "step": 1393 + }, + { + "epoch": 0.09578122852823966, + "grad_norm": 1.2426620721817017, + "learning_rate": 4.787087912087913e-05, + "loss": 1.093, + "step": 1394 + }, + { + "epoch": 0.09584993816133022, + "grad_norm": 1.2062158584594727, + "learning_rate": 4.790521978021978e-05, + "loss": 1.0668, + "step": 1395 + }, + { + "epoch": 0.09591864779442077, + "grad_norm": 1.2813827991485596, + "learning_rate": 4.7939560439560445e-05, + "loss": 1.0839, + "step": 1396 + }, + { + "epoch": 0.09598735742751134, + "grad_norm": 1.2124255895614624, + "learning_rate": 4.7973901098901106e-05, + "loss": 1.0836, + "step": 1397 + }, + { + "epoch": 0.0960560670606019, + "grad_norm": 1.2640279531478882, + "learning_rate": 4.800824175824176e-05, + "loss": 1.0874, + "step": 1398 + }, + { + "epoch": 0.09612477669369246, + "grad_norm": 1.321155071258545, + "learning_rate": 4.804258241758242e-05, + "loss": 1.0864, + "step": 1399 + }, + { + "epoch": 0.09619348632678301, + "grad_norm": 1.3115631341934204, + "learning_rate": 4.8076923076923084e-05, + "loss": 1.0706, + "step": 1400 + }, + { + "epoch": 0.09626219595987358, + "grad_norm": 1.2679675817489624, + "learning_rate": 4.811126373626374e-05, + "loss": 1.0936, + "step": 1401 + }, + { + "epoch": 0.09633090559296413, + "grad_norm": 1.2640880346298218, + "learning_rate": 4.81456043956044e-05, + "loss": 1.1094, + "step": 1402 + }, + { + "epoch": 0.0963996152260547, + "grad_norm": 1.2915562391281128, + "learning_rate": 4.817994505494506e-05, + "loss": 1.0681, + "step": 1403 + }, + { + "epoch": 0.09646832485914525, + "grad_norm": 1.2992143630981445, + "learning_rate": 4.8214285714285716e-05, + "loss": 1.1055, + "step": 1404 + }, + { + "epoch": 0.09653703449223582, + "grad_norm": 1.282727599143982, + "learning_rate": 4.824862637362638e-05, + "loss": 1.0947, + "step": 1405 + }, + { + "epoch": 0.09660574412532637, + "grad_norm": 1.2701010704040527, + "learning_rate": 4.828296703296704e-05, + "loss": 1.1203, + "step": 1406 + }, + { + "epoch": 0.09667445375841693, + "grad_norm": 1.361436128616333, + "learning_rate": 4.8317307692307693e-05, + "loss": 1.1238, + "step": 1407 + }, + { + "epoch": 0.09674316339150749, + "grad_norm": 1.206047773361206, + "learning_rate": 4.8351648351648355e-05, + "loss": 1.0844, + "step": 1408 + }, + { + "epoch": 0.09681187302459805, + "grad_norm": 1.2738022804260254, + "learning_rate": 4.8385989010989016e-05, + "loss": 1.0727, + "step": 1409 + }, + { + "epoch": 0.0968805826576886, + "grad_norm": 1.2727105617523193, + "learning_rate": 4.842032967032967e-05, + "loss": 1.1318, + "step": 1410 + }, + { + "epoch": 0.09694929229077917, + "grad_norm": 1.3065145015716553, + "learning_rate": 4.845467032967033e-05, + "loss": 1.008, + "step": 1411 + }, + { + "epoch": 0.09701800192386972, + "grad_norm": 1.250906229019165, + "learning_rate": 4.8489010989010994e-05, + "loss": 1.0792, + "step": 1412 + }, + { + "epoch": 0.09708671155696029, + "grad_norm": 1.3009296655654907, + "learning_rate": 4.852335164835165e-05, + "loss": 1.0879, + "step": 1413 + }, + { + "epoch": 0.09715542119005084, + "grad_norm": 1.2521177530288696, + "learning_rate": 4.855769230769231e-05, + "loss": 1.1193, + "step": 1414 + }, + { + "epoch": 0.09722413082314141, + "grad_norm": 1.3665554523468018, + "learning_rate": 4.859203296703297e-05, + "loss": 1.1365, + "step": 1415 + }, + { + "epoch": 0.09729284045623196, + "grad_norm": 1.3133963346481323, + "learning_rate": 4.8626373626373626e-05, + "loss": 1.0454, + "step": 1416 + }, + { + "epoch": 0.09736155008932253, + "grad_norm": 1.2715580463409424, + "learning_rate": 4.866071428571429e-05, + "loss": 1.0368, + "step": 1417 + }, + { + "epoch": 0.09743025972241308, + "grad_norm": 1.3201946020126343, + "learning_rate": 4.869505494505495e-05, + "loss": 1.0668, + "step": 1418 + }, + { + "epoch": 0.09749896935550365, + "grad_norm": 1.4186021089553833, + "learning_rate": 4.8729395604395604e-05, + "loss": 1.047, + "step": 1419 + }, + { + "epoch": 0.0975676789885942, + "grad_norm": 1.2789784669876099, + "learning_rate": 4.8763736263736265e-05, + "loss": 1.0704, + "step": 1420 + }, + { + "epoch": 0.09763638862168476, + "grad_norm": 1.1966907978057861, + "learning_rate": 4.8798076923076926e-05, + "loss": 1.0764, + "step": 1421 + }, + { + "epoch": 0.09770509825477532, + "grad_norm": 1.3008646965026855, + "learning_rate": 4.883241758241758e-05, + "loss": 1.0309, + "step": 1422 + }, + { + "epoch": 0.09777380788786588, + "grad_norm": 1.3532465696334839, + "learning_rate": 4.886675824175824e-05, + "loss": 1.1002, + "step": 1423 + }, + { + "epoch": 0.09784251752095643, + "grad_norm": 1.2983850240707397, + "learning_rate": 4.8901098901098904e-05, + "loss": 1.092, + "step": 1424 + }, + { + "epoch": 0.097911227154047, + "grad_norm": 1.252535104751587, + "learning_rate": 4.893543956043956e-05, + "loss": 1.0752, + "step": 1425 + }, + { + "epoch": 0.09797993678713755, + "grad_norm": 1.2596217393875122, + "learning_rate": 4.896978021978022e-05, + "loss": 1.0374, + "step": 1426 + }, + { + "epoch": 0.09804864642022812, + "grad_norm": 1.3025498390197754, + "learning_rate": 4.900412087912088e-05, + "loss": 1.0725, + "step": 1427 + }, + { + "epoch": 0.09811735605331867, + "grad_norm": 1.2887141704559326, + "learning_rate": 4.9038461538461536e-05, + "loss": 1.0529, + "step": 1428 + }, + { + "epoch": 0.09818606568640924, + "grad_norm": 1.3206496238708496, + "learning_rate": 4.90728021978022e-05, + "loss": 1.051, + "step": 1429 + }, + { + "epoch": 0.09825477531949979, + "grad_norm": 1.2431976795196533, + "learning_rate": 4.910714285714286e-05, + "loss": 1.0832, + "step": 1430 + }, + { + "epoch": 0.09832348495259036, + "grad_norm": 1.2815593481063843, + "learning_rate": 4.9141483516483514e-05, + "loss": 1.0673, + "step": 1431 + }, + { + "epoch": 0.09839219458568091, + "grad_norm": 1.2862306833267212, + "learning_rate": 4.9175824175824175e-05, + "loss": 1.0563, + "step": 1432 + }, + { + "epoch": 0.09846090421877148, + "grad_norm": 1.2998971939086914, + "learning_rate": 4.921016483516484e-05, + "loss": 1.0921, + "step": 1433 + }, + { + "epoch": 0.09852961385186203, + "grad_norm": 1.2313848733901978, + "learning_rate": 4.92445054945055e-05, + "loss": 1.0985, + "step": 1434 + }, + { + "epoch": 0.0985983234849526, + "grad_norm": 1.2480573654174805, + "learning_rate": 4.927884615384616e-05, + "loss": 1.0641, + "step": 1435 + }, + { + "epoch": 0.09866703311804315, + "grad_norm": 1.4184104204177856, + "learning_rate": 4.931318681318682e-05, + "loss": 1.1146, + "step": 1436 + }, + { + "epoch": 0.09873574275113371, + "grad_norm": 1.3841133117675781, + "learning_rate": 4.9347527472527476e-05, + "loss": 1.076, + "step": 1437 + }, + { + "epoch": 0.09880445238422426, + "grad_norm": 1.2942969799041748, + "learning_rate": 4.938186813186814e-05, + "loss": 1.0519, + "step": 1438 + }, + { + "epoch": 0.09887316201731483, + "grad_norm": 1.2936192750930786, + "learning_rate": 4.94162087912088e-05, + "loss": 1.0716, + "step": 1439 + }, + { + "epoch": 0.09894187165040538, + "grad_norm": 1.3379738330841064, + "learning_rate": 4.945054945054945e-05, + "loss": 1.0968, + "step": 1440 + }, + { + "epoch": 0.09901058128349595, + "grad_norm": 1.3171402215957642, + "learning_rate": 4.9484890109890115e-05, + "loss": 1.053, + "step": 1441 + }, + { + "epoch": 0.0990792909165865, + "grad_norm": 1.2431278228759766, + "learning_rate": 4.9519230769230776e-05, + "loss": 1.0514, + "step": 1442 + }, + { + "epoch": 0.09914800054967707, + "grad_norm": 1.258926510810852, + "learning_rate": 4.955357142857143e-05, + "loss": 1.0392, + "step": 1443 + }, + { + "epoch": 0.09921671018276762, + "grad_norm": 1.2499626874923706, + "learning_rate": 4.958791208791209e-05, + "loss": 1.0911, + "step": 1444 + }, + { + "epoch": 0.09928541981585819, + "grad_norm": 1.3242855072021484, + "learning_rate": 4.9622252747252754e-05, + "loss": 1.1057, + "step": 1445 + }, + { + "epoch": 0.09935412944894874, + "grad_norm": 1.361946940422058, + "learning_rate": 4.965659340659341e-05, + "loss": 1.073, + "step": 1446 + }, + { + "epoch": 0.0994228390820393, + "grad_norm": 1.3291690349578857, + "learning_rate": 4.969093406593407e-05, + "loss": 1.0615, + "step": 1447 + }, + { + "epoch": 0.09949154871512986, + "grad_norm": 1.2458477020263672, + "learning_rate": 4.972527472527473e-05, + "loss": 1.0912, + "step": 1448 + }, + { + "epoch": 0.09956025834822042, + "grad_norm": 1.2715243101119995, + "learning_rate": 4.9759615384615386e-05, + "loss": 1.0311, + "step": 1449 + }, + { + "epoch": 0.09962896798131098, + "grad_norm": 1.3295639753341675, + "learning_rate": 4.979395604395605e-05, + "loss": 1.0734, + "step": 1450 + }, + { + "epoch": 0.09969767761440154, + "grad_norm": 1.2607442140579224, + "learning_rate": 4.982829670329671e-05, + "loss": 1.0956, + "step": 1451 + }, + { + "epoch": 0.0997663872474921, + "grad_norm": 1.3046011924743652, + "learning_rate": 4.9862637362637363e-05, + "loss": 1.0824, + "step": 1452 + }, + { + "epoch": 0.09983509688058266, + "grad_norm": 1.1655594110488892, + "learning_rate": 4.9896978021978025e-05, + "loss": 1.0793, + "step": 1453 + }, + { + "epoch": 0.09990380651367321, + "grad_norm": 1.3410218954086304, + "learning_rate": 4.9931318681318686e-05, + "loss": 1.0852, + "step": 1454 + }, + { + "epoch": 0.09997251614676378, + "grad_norm": 1.3166815042495728, + "learning_rate": 4.996565934065934e-05, + "loss": 1.0768, + "step": 1455 + }, + { + "epoch": 0.10004122577985433, + "grad_norm": 1.272876501083374, + "learning_rate": 5e-05, + "loss": 1.1186, + "step": 1456 + }, + { + "epoch": 0.1001099354129449, + "grad_norm": 1.2234960794448853, + "learning_rate": 4.999999928088262e-05, + "loss": 1.0821, + "step": 1457 + }, + { + "epoch": 0.10017864504603545, + "grad_norm": 1.2698850631713867, + "learning_rate": 4.999999712353053e-05, + "loss": 1.0738, + "step": 1458 + }, + { + "epoch": 0.10024735467912602, + "grad_norm": 1.2999560832977295, + "learning_rate": 4.999999352794384e-05, + "loss": 1.0307, + "step": 1459 + }, + { + "epoch": 0.10031606431221657, + "grad_norm": 1.26005220413208, + "learning_rate": 4.999998849412277e-05, + "loss": 1.0696, + "step": 1460 + }, + { + "epoch": 0.10038477394530713, + "grad_norm": 1.276008129119873, + "learning_rate": 4.9999982022067614e-05, + "loss": 1.1142, + "step": 1461 + }, + { + "epoch": 0.10045348357839769, + "grad_norm": 1.1824654340744019, + "learning_rate": 4.999997411177873e-05, + "loss": 1.0401, + "step": 1462 + }, + { + "epoch": 0.10052219321148825, + "grad_norm": 1.2764183282852173, + "learning_rate": 4.9999964763256575e-05, + "loss": 1.1088, + "step": 1463 + }, + { + "epoch": 0.1005909028445788, + "grad_norm": 1.2581321001052856, + "learning_rate": 4.999995397650169e-05, + "loss": 1.0713, + "step": 1464 + }, + { + "epoch": 0.10065961247766937, + "grad_norm": 1.288217306137085, + "learning_rate": 4.99999417515147e-05, + "loss": 1.0818, + "step": 1465 + }, + { + "epoch": 0.10072832211075992, + "grad_norm": 1.3536568880081177, + "learning_rate": 4.99999280882963e-05, + "loss": 1.08, + "step": 1466 + }, + { + "epoch": 0.10079703174385049, + "grad_norm": 1.2842787504196167, + "learning_rate": 4.9999912986847284e-05, + "loss": 1.1286, + "step": 1467 + }, + { + "epoch": 0.10086574137694104, + "grad_norm": 1.291973352432251, + "learning_rate": 4.999989644716852e-05, + "loss": 1.0551, + "step": 1468 + }, + { + "epoch": 0.10093445101003161, + "grad_norm": 1.2063801288604736, + "learning_rate": 4.9999878469260944e-05, + "loss": 1.1141, + "step": 1469 + }, + { + "epoch": 0.10100316064312216, + "grad_norm": 1.3906915187835693, + "learning_rate": 4.999985905312561e-05, + "loss": 1.0795, + "step": 1470 + }, + { + "epoch": 0.10107187027621273, + "grad_norm": 1.2617043256759644, + "learning_rate": 4.999983819876363e-05, + "loss": 1.0672, + "step": 1471 + }, + { + "epoch": 0.10114057990930328, + "grad_norm": 1.3072538375854492, + "learning_rate": 4.9999815906176206e-05, + "loss": 1.1008, + "step": 1472 + }, + { + "epoch": 0.10120928954239385, + "grad_norm": 1.3387470245361328, + "learning_rate": 4.9999792175364606e-05, + "loss": 1.0484, + "step": 1473 + }, + { + "epoch": 0.1012779991754844, + "grad_norm": 1.3512241840362549, + "learning_rate": 4.9999767006330214e-05, + "loss": 1.0818, + "step": 1474 + }, + { + "epoch": 0.10134670880857496, + "grad_norm": 1.2933716773986816, + "learning_rate": 4.999974039907446e-05, + "loss": 1.0771, + "step": 1475 + }, + { + "epoch": 0.10141541844166552, + "grad_norm": 1.2845193147659302, + "learning_rate": 4.999971235359889e-05, + "loss": 1.0778, + "step": 1476 + }, + { + "epoch": 0.10148412807475608, + "grad_norm": 1.264211893081665, + "learning_rate": 4.999968286990512e-05, + "loss": 1.1241, + "step": 1477 + }, + { + "epoch": 0.10155283770784664, + "grad_norm": 1.3372876644134521, + "learning_rate": 4.9999651947994824e-05, + "loss": 1.0677, + "step": 1478 + }, + { + "epoch": 0.1016215473409372, + "grad_norm": 1.2458248138427734, + "learning_rate": 4.9999619587869793e-05, + "loss": 1.1084, + "step": 1479 + }, + { + "epoch": 0.10169025697402775, + "grad_norm": 1.280226707458496, + "learning_rate": 4.99995857895319e-05, + "loss": 1.0812, + "step": 1480 + }, + { + "epoch": 0.10175896660711832, + "grad_norm": 1.2756706476211548, + "learning_rate": 4.999955055298308e-05, + "loss": 1.0685, + "step": 1481 + }, + { + "epoch": 0.10182767624020887, + "grad_norm": 1.2809486389160156, + "learning_rate": 4.999951387822536e-05, + "loss": 1.0467, + "step": 1482 + }, + { + "epoch": 0.10189638587329944, + "grad_norm": 1.3525607585906982, + "learning_rate": 4.999947576526084e-05, + "loss": 1.095, + "step": 1483 + }, + { + "epoch": 0.10196509550638999, + "grad_norm": 1.2279099225997925, + "learning_rate": 4.9999436214091724e-05, + "loss": 1.0765, + "step": 1484 + }, + { + "epoch": 0.10203380513948056, + "grad_norm": 1.283117651939392, + "learning_rate": 4.999939522472029e-05, + "loss": 1.0395, + "step": 1485 + }, + { + "epoch": 0.10210251477257111, + "grad_norm": 1.2845925092697144, + "learning_rate": 4.999935279714889e-05, + "loss": 1.0851, + "step": 1486 + }, + { + "epoch": 0.10217122440566168, + "grad_norm": 1.2507386207580566, + "learning_rate": 4.9999308931379976e-05, + "loss": 1.0449, + "step": 1487 + }, + { + "epoch": 0.10223993403875223, + "grad_norm": 1.2447125911712646, + "learning_rate": 4.999926362741605e-05, + "loss": 1.0649, + "step": 1488 + }, + { + "epoch": 0.1023086436718428, + "grad_norm": 1.1920353174209595, + "learning_rate": 4.999921688525973e-05, + "loss": 1.0477, + "step": 1489 + }, + { + "epoch": 0.10237735330493335, + "grad_norm": 1.2895169258117676, + "learning_rate": 4.9999168704913714e-05, + "loss": 1.102, + "step": 1490 + }, + { + "epoch": 0.10244606293802391, + "grad_norm": 1.3395603895187378, + "learning_rate": 4.999911908638076e-05, + "loss": 1.0662, + "step": 1491 + }, + { + "epoch": 0.10251477257111447, + "grad_norm": 1.1883882284164429, + "learning_rate": 4.9999068029663724e-05, + "loss": 1.0895, + "step": 1492 + }, + { + "epoch": 0.10258348220420503, + "grad_norm": 1.3088696002960205, + "learning_rate": 4.999901553476555e-05, + "loss": 1.0914, + "step": 1493 + }, + { + "epoch": 0.10265219183729558, + "grad_norm": 1.2613099813461304, + "learning_rate": 4.999896160168926e-05, + "loss": 1.059, + "step": 1494 + }, + { + "epoch": 0.10272090147038615, + "grad_norm": 1.199046015739441, + "learning_rate": 4.9998906230437946e-05, + "loss": 1.0776, + "step": 1495 + }, + { + "epoch": 0.1027896111034767, + "grad_norm": 1.2550249099731445, + "learning_rate": 4.9998849421014806e-05, + "loss": 1.0421, + "step": 1496 + }, + { + "epoch": 0.10285832073656727, + "grad_norm": 1.2493432760238647, + "learning_rate": 4.999879117342309e-05, + "loss": 1.0785, + "step": 1497 + }, + { + "epoch": 0.10292703036965782, + "grad_norm": 1.3821302652359009, + "learning_rate": 4.999873148766617e-05, + "loss": 1.0756, + "step": 1498 + }, + { + "epoch": 0.10299574000274839, + "grad_norm": 1.2333570718765259, + "learning_rate": 4.999867036374747e-05, + "loss": 1.0646, + "step": 1499 + }, + { + "epoch": 0.10306444963583894, + "grad_norm": 1.2179633378982544, + "learning_rate": 4.99986078016705e-05, + "loss": 1.0624, + "step": 1500 + }, + { + "epoch": 0.1031331592689295, + "grad_norm": 1.243916392326355, + "learning_rate": 4.999854380143887e-05, + "loss": 1.0349, + "step": 1501 + }, + { + "epoch": 0.10320186890202006, + "grad_norm": 1.3510416746139526, + "learning_rate": 4.999847836305626e-05, + "loss": 1.094, + "step": 1502 + }, + { + "epoch": 0.10327057853511062, + "grad_norm": 1.3110347986221313, + "learning_rate": 4.999841148652643e-05, + "loss": 1.0654, + "step": 1503 + }, + { + "epoch": 0.10333928816820118, + "grad_norm": 1.2657229900360107, + "learning_rate": 4.9998343171853223e-05, + "loss": 1.1107, + "step": 1504 + }, + { + "epoch": 0.10340799780129174, + "grad_norm": 1.3069169521331787, + "learning_rate": 4.999827341904059e-05, + "loss": 1.1023, + "step": 1505 + }, + { + "epoch": 0.1034767074343823, + "grad_norm": 1.3091444969177246, + "learning_rate": 4.999820222809252e-05, + "loss": 1.0856, + "step": 1506 + }, + { + "epoch": 0.10354541706747286, + "grad_norm": 1.3859574794769287, + "learning_rate": 4.999812959901311e-05, + "loss": 1.0859, + "step": 1507 + }, + { + "epoch": 0.10361412670056341, + "grad_norm": 1.200980544090271, + "learning_rate": 4.999805553180656e-05, + "loss": 1.0512, + "step": 1508 + }, + { + "epoch": 0.10368283633365398, + "grad_norm": 1.2746727466583252, + "learning_rate": 4.999798002647711e-05, + "loss": 1.0652, + "step": 1509 + }, + { + "epoch": 0.10375154596674453, + "grad_norm": 1.237650752067566, + "learning_rate": 4.9997903083029117e-05, + "loss": 1.089, + "step": 1510 + }, + { + "epoch": 0.1038202555998351, + "grad_norm": 1.2778815031051636, + "learning_rate": 4.9997824701467e-05, + "loss": 1.0482, + "step": 1511 + }, + { + "epoch": 0.10388896523292565, + "grad_norm": 1.253006100654602, + "learning_rate": 4.999774488179526e-05, + "loss": 1.048, + "step": 1512 + }, + { + "epoch": 0.10395767486601622, + "grad_norm": 1.3484008312225342, + "learning_rate": 4.9997663624018506e-05, + "loss": 1.0597, + "step": 1513 + }, + { + "epoch": 0.10402638449910677, + "grad_norm": 1.3262109756469727, + "learning_rate": 4.9997580928141416e-05, + "loss": 1.0913, + "step": 1514 + }, + { + "epoch": 0.10409509413219734, + "grad_norm": 1.2588205337524414, + "learning_rate": 4.9997496794168726e-05, + "loss": 1.0708, + "step": 1515 + }, + { + "epoch": 0.10416380376528789, + "grad_norm": 1.2651864290237427, + "learning_rate": 4.9997411222105287e-05, + "loss": 1.0922, + "step": 1516 + }, + { + "epoch": 0.10423251339837845, + "grad_norm": 1.280624270439148, + "learning_rate": 4.999732421195602e-05, + "loss": 1.065, + "step": 1517 + }, + { + "epoch": 0.104301223031469, + "grad_norm": 1.2742594480514526, + "learning_rate": 4.999723576372594e-05, + "loss": 1.0872, + "step": 1518 + }, + { + "epoch": 0.10436993266455957, + "grad_norm": 1.2743715047836304, + "learning_rate": 4.9997145877420125e-05, + "loss": 1.085, + "step": 1519 + }, + { + "epoch": 0.10443864229765012, + "grad_norm": 1.3351998329162598, + "learning_rate": 4.999705455304375e-05, + "loss": 1.0574, + "step": 1520 + }, + { + "epoch": 0.10450735193074069, + "grad_norm": 1.2589765787124634, + "learning_rate": 4.999696179060207e-05, + "loss": 1.0919, + "step": 1521 + }, + { + "epoch": 0.10457606156383124, + "grad_norm": 1.327614665031433, + "learning_rate": 4.9996867590100416e-05, + "loss": 1.0554, + "step": 1522 + }, + { + "epoch": 0.10464477119692181, + "grad_norm": 1.3307448625564575, + "learning_rate": 4.9996771951544205e-05, + "loss": 1.0824, + "step": 1523 + }, + { + "epoch": 0.10471348083001236, + "grad_norm": 1.2914636135101318, + "learning_rate": 4.999667487493896e-05, + "loss": 1.0676, + "step": 1524 + }, + { + "epoch": 0.10478219046310293, + "grad_norm": 1.3306961059570312, + "learning_rate": 4.9996576360290244e-05, + "loss": 1.0739, + "step": 1525 + }, + { + "epoch": 0.10485090009619348, + "grad_norm": 1.1610969305038452, + "learning_rate": 4.999647640760373e-05, + "loss": 1.0423, + "step": 1526 + }, + { + "epoch": 0.10491960972928405, + "grad_norm": 1.349721074104309, + "learning_rate": 4.999637501688517e-05, + "loss": 1.0911, + "step": 1527 + }, + { + "epoch": 0.1049883193623746, + "grad_norm": 1.22369384765625, + "learning_rate": 4.9996272188140395e-05, + "loss": 1.0377, + "step": 1528 + }, + { + "epoch": 0.10505702899546517, + "grad_norm": 1.3024455308914185, + "learning_rate": 4.9996167921375325e-05, + "loss": 1.1153, + "step": 1529 + }, + { + "epoch": 0.10512573862855572, + "grad_norm": 1.1852585077285767, + "learning_rate": 4.999606221659595e-05, + "loss": 1.0915, + "step": 1530 + }, + { + "epoch": 0.10519444826164628, + "grad_norm": 1.257179617881775, + "learning_rate": 4.999595507380836e-05, + "loss": 1.0812, + "step": 1531 + }, + { + "epoch": 0.10526315789473684, + "grad_norm": 1.3378115892410278, + "learning_rate": 4.999584649301872e-05, + "loss": 1.0836, + "step": 1532 + }, + { + "epoch": 0.1053318675278274, + "grad_norm": 1.2561935186386108, + "learning_rate": 4.999573647423326e-05, + "loss": 1.0622, + "step": 1533 + }, + { + "epoch": 0.10540057716091795, + "grad_norm": 1.31657874584198, + "learning_rate": 4.999562501745833e-05, + "loss": 1.1149, + "step": 1534 + }, + { + "epoch": 0.10546928679400852, + "grad_norm": 1.218398094177246, + "learning_rate": 4.999551212270033e-05, + "loss": 1.0827, + "step": 1535 + }, + { + "epoch": 0.10553799642709907, + "grad_norm": 1.2894662618637085, + "learning_rate": 4.999539778996577e-05, + "loss": 1.0662, + "step": 1536 + }, + { + "epoch": 0.10560670606018964, + "grad_norm": 1.2247436046600342, + "learning_rate": 4.99952820192612e-05, + "loss": 1.0815, + "step": 1537 + }, + { + "epoch": 0.10567541569328019, + "grad_norm": 1.2220253944396973, + "learning_rate": 4.9995164810593305e-05, + "loss": 1.0772, + "step": 1538 + }, + { + "epoch": 0.10574412532637076, + "grad_norm": 1.1616848707199097, + "learning_rate": 4.999504616396882e-05, + "loss": 1.1052, + "step": 1539 + }, + { + "epoch": 0.10581283495946131, + "grad_norm": 1.2436491250991821, + "learning_rate": 4.999492607939456e-05, + "loss": 1.0559, + "step": 1540 + }, + { + "epoch": 0.10588154459255188, + "grad_norm": 1.2103008031845093, + "learning_rate": 4.999480455687745e-05, + "loss": 1.0885, + "step": 1541 + }, + { + "epoch": 0.10595025422564243, + "grad_norm": 1.2059721946716309, + "learning_rate": 4.999468159642447e-05, + "loss": 1.0822, + "step": 1542 + }, + { + "epoch": 0.106018963858733, + "grad_norm": 1.1944838762283325, + "learning_rate": 4.99945571980427e-05, + "loss": 1.0379, + "step": 1543 + }, + { + "epoch": 0.10608767349182355, + "grad_norm": 1.2740137577056885, + "learning_rate": 4.999443136173929e-05, + "loss": 1.0992, + "step": 1544 + }, + { + "epoch": 0.10615638312491411, + "grad_norm": 1.2140851020812988, + "learning_rate": 4.9994304087521496e-05, + "loss": 1.1126, + "step": 1545 + }, + { + "epoch": 0.10622509275800467, + "grad_norm": 1.2254486083984375, + "learning_rate": 4.9994175375396615e-05, + "loss": 1.1063, + "step": 1546 + }, + { + "epoch": 0.10629380239109523, + "grad_norm": 1.2097018957138062, + "learning_rate": 4.9994045225372075e-05, + "loss": 1.1109, + "step": 1547 + }, + { + "epoch": 0.10636251202418578, + "grad_norm": 1.2839246988296509, + "learning_rate": 4.999391363745535e-05, + "loss": 1.0533, + "step": 1548 + }, + { + "epoch": 0.10643122165727635, + "grad_norm": 1.1923298835754395, + "learning_rate": 4.9993780611654005e-05, + "loss": 1.0497, + "step": 1549 + }, + { + "epoch": 0.1064999312903669, + "grad_norm": 1.146784782409668, + "learning_rate": 4.99936461479757e-05, + "loss": 1.072, + "step": 1550 + }, + { + "epoch": 0.10656864092345747, + "grad_norm": 1.2163045406341553, + "learning_rate": 4.9993510246428185e-05, + "loss": 1.0599, + "step": 1551 + }, + { + "epoch": 0.10663735055654802, + "grad_norm": 1.1379610300064087, + "learning_rate": 4.9993372907019256e-05, + "loss": 1.0149, + "step": 1552 + }, + { + "epoch": 0.10670606018963859, + "grad_norm": 1.2762796878814697, + "learning_rate": 4.999323412975683e-05, + "loss": 1.0623, + "step": 1553 + }, + { + "epoch": 0.10677476982272914, + "grad_norm": 1.3049037456512451, + "learning_rate": 4.999309391464887e-05, + "loss": 1.0446, + "step": 1554 + }, + { + "epoch": 0.1068434794558197, + "grad_norm": 1.2086464166641235, + "learning_rate": 4.999295226170347e-05, + "loss": 1.0737, + "step": 1555 + }, + { + "epoch": 0.10691218908891026, + "grad_norm": 1.3631081581115723, + "learning_rate": 4.999280917092875e-05, + "loss": 1.087, + "step": 1556 + }, + { + "epoch": 0.10698089872200083, + "grad_norm": 1.2668733596801758, + "learning_rate": 4.9992664642332976e-05, + "loss": 1.0698, + "step": 1557 + }, + { + "epoch": 0.10704960835509138, + "grad_norm": 1.210692286491394, + "learning_rate": 4.9992518675924434e-05, + "loss": 1.0744, + "step": 1558 + }, + { + "epoch": 0.10711831798818194, + "grad_norm": 1.25871741771698, + "learning_rate": 4.9992371271711534e-05, + "loss": 1.0688, + "step": 1559 + }, + { + "epoch": 0.1071870276212725, + "grad_norm": 1.099435567855835, + "learning_rate": 4.9992222429702754e-05, + "loss": 1.0081, + "step": 1560 + }, + { + "epoch": 0.10725573725436306, + "grad_norm": 1.214288592338562, + "learning_rate": 4.999207214990665e-05, + "loss": 1.0587, + "step": 1561 + }, + { + "epoch": 0.10732444688745361, + "grad_norm": 1.2272498607635498, + "learning_rate": 4.999192043233188e-05, + "loss": 1.0695, + "step": 1562 + }, + { + "epoch": 0.10739315652054418, + "grad_norm": 1.1519272327423096, + "learning_rate": 4.999176727698716e-05, + "loss": 1.0542, + "step": 1563 + }, + { + "epoch": 0.10746186615363473, + "grad_norm": 1.266766905784607, + "learning_rate": 4.9991612683881314e-05, + "loss": 1.0759, + "step": 1564 + }, + { + "epoch": 0.1075305757867253, + "grad_norm": 1.2163397073745728, + "learning_rate": 4.9991456653023226e-05, + "loss": 1.0664, + "step": 1565 + }, + { + "epoch": 0.10759928541981585, + "grad_norm": 1.3226927518844604, + "learning_rate": 4.999129918442187e-05, + "loss": 1.0247, + "step": 1566 + }, + { + "epoch": 0.10766799505290642, + "grad_norm": 1.249815583229065, + "learning_rate": 4.9991140278086316e-05, + "loss": 1.0632, + "step": 1567 + }, + { + "epoch": 0.10773670468599697, + "grad_norm": 1.2963767051696777, + "learning_rate": 4.9990979934025696e-05, + "loss": 1.1355, + "step": 1568 + }, + { + "epoch": 0.10780541431908754, + "grad_norm": 1.2136685848236084, + "learning_rate": 4.9990818152249244e-05, + "loss": 1.0864, + "step": 1569 + }, + { + "epoch": 0.10787412395217809, + "grad_norm": 1.212208867073059, + "learning_rate": 4.999065493276626e-05, + "loss": 1.0299, + "step": 1570 + }, + { + "epoch": 0.10794283358526865, + "grad_norm": 1.185073733329773, + "learning_rate": 4.9990490275586135e-05, + "loss": 1.0758, + "step": 1571 + }, + { + "epoch": 0.10801154321835921, + "grad_norm": 1.2976653575897217, + "learning_rate": 4.9990324180718335e-05, + "loss": 1.0908, + "step": 1572 + }, + { + "epoch": 0.10808025285144977, + "grad_norm": 1.3215500116348267, + "learning_rate": 4.999015664817243e-05, + "loss": 1.0731, + "step": 1573 + }, + { + "epoch": 0.10814896248454033, + "grad_norm": 1.196366786956787, + "learning_rate": 4.998998767795805e-05, + "loss": 1.0646, + "step": 1574 + }, + { + "epoch": 0.10821767211763089, + "grad_norm": 1.2911126613616943, + "learning_rate": 4.9989817270084905e-05, + "loss": 1.0704, + "step": 1575 + }, + { + "epoch": 0.10828638175072144, + "grad_norm": 1.2613106966018677, + "learning_rate": 4.998964542456282e-05, + "loss": 1.0877, + "step": 1576 + }, + { + "epoch": 0.10835509138381201, + "grad_norm": 1.2894078493118286, + "learning_rate": 4.998947214140167e-05, + "loss": 1.0475, + "step": 1577 + }, + { + "epoch": 0.10842380101690256, + "grad_norm": 1.2662047147750854, + "learning_rate": 4.9989297420611426e-05, + "loss": 1.0926, + "step": 1578 + }, + { + "epoch": 0.10849251064999313, + "grad_norm": 1.2152013778686523, + "learning_rate": 4.9989121262202135e-05, + "loss": 1.0613, + "step": 1579 + }, + { + "epoch": 0.10856122028308368, + "grad_norm": 1.2184345722198486, + "learning_rate": 4.998894366618393e-05, + "loss": 1.0283, + "step": 1580 + }, + { + "epoch": 0.10862992991617425, + "grad_norm": 1.2261587381362915, + "learning_rate": 4.9988764632567044e-05, + "loss": 1.0481, + "step": 1581 + }, + { + "epoch": 0.10869863954926481, + "grad_norm": 1.3080343008041382, + "learning_rate": 4.998858416136175e-05, + "loss": 1.0743, + "step": 1582 + }, + { + "epoch": 0.10876734918235537, + "grad_norm": 1.2150989770889282, + "learning_rate": 4.998840225257846e-05, + "loss": 1.0346, + "step": 1583 + }, + { + "epoch": 0.10883605881544593, + "grad_norm": 1.2218056917190552, + "learning_rate": 4.998821890622762e-05, + "loss": 1.0326, + "step": 1584 + }, + { + "epoch": 0.10890476844853648, + "grad_norm": 1.2550586462020874, + "learning_rate": 4.998803412231978e-05, + "loss": 1.0773, + "step": 1585 + }, + { + "epoch": 0.10897347808162705, + "grad_norm": 1.340316653251648, + "learning_rate": 4.998784790086557e-05, + "loss": 1.0732, + "step": 1586 + }, + { + "epoch": 0.1090421877147176, + "grad_norm": 1.2334810495376587, + "learning_rate": 4.998766024187571e-05, + "loss": 1.0812, + "step": 1587 + }, + { + "epoch": 0.10911089734780817, + "grad_norm": 1.3193124532699585, + "learning_rate": 4.9987471145361e-05, + "loss": 1.0589, + "step": 1588 + }, + { + "epoch": 0.10917960698089872, + "grad_norm": 1.2762266397476196, + "learning_rate": 4.99872806113323e-05, + "loss": 1.0912, + "step": 1589 + }, + { + "epoch": 0.10924831661398929, + "grad_norm": 1.3713821172714233, + "learning_rate": 4.9987088639800585e-05, + "loss": 1.0339, + "step": 1590 + }, + { + "epoch": 0.10931702624707984, + "grad_norm": 1.2835079431533813, + "learning_rate": 4.99868952307769e-05, + "loss": 1.0561, + "step": 1591 + }, + { + "epoch": 0.1093857358801704, + "grad_norm": 1.3130748271942139, + "learning_rate": 4.998670038427236e-05, + "loss": 1.0938, + "step": 1592 + }, + { + "epoch": 0.10945444551326096, + "grad_norm": 1.2071493864059448, + "learning_rate": 4.9986504100298185e-05, + "loss": 1.0608, + "step": 1593 + }, + { + "epoch": 0.10952315514635153, + "grad_norm": 1.2077372074127197, + "learning_rate": 4.998630637886566e-05, + "loss": 1.0573, + "step": 1594 + }, + { + "epoch": 0.10959186477944208, + "grad_norm": 1.333223581314087, + "learning_rate": 4.998610721998618e-05, + "loss": 1.1094, + "step": 1595 + }, + { + "epoch": 0.10966057441253264, + "grad_norm": 1.232903003692627, + "learning_rate": 4.9985906623671174e-05, + "loss": 1.06, + "step": 1596 + }, + { + "epoch": 0.1097292840456232, + "grad_norm": 1.3169827461242676, + "learning_rate": 4.99857045899322e-05, + "loss": 1.068, + "step": 1597 + }, + { + "epoch": 0.10979799367871376, + "grad_norm": 1.2244253158569336, + "learning_rate": 4.998550111878087e-05, + "loss": 1.0439, + "step": 1598 + }, + { + "epoch": 0.10986670331180431, + "grad_norm": 1.3144006729125977, + "learning_rate": 4.9985296210228895e-05, + "loss": 1.1212, + "step": 1599 + }, + { + "epoch": 0.10993541294489488, + "grad_norm": 1.2804158926010132, + "learning_rate": 4.998508986428807e-05, + "loss": 1.117, + "step": 1600 + }, + { + "epoch": 0.11000412257798543, + "grad_norm": 1.2195775508880615, + "learning_rate": 4.998488208097025e-05, + "loss": 1.06, + "step": 1601 + }, + { + "epoch": 0.110072832211076, + "grad_norm": 1.218427300453186, + "learning_rate": 4.9984672860287406e-05, + "loss": 1.0702, + "step": 1602 + }, + { + "epoch": 0.11014154184416655, + "grad_norm": 1.3716959953308105, + "learning_rate": 4.998446220225156e-05, + "loss": 1.0743, + "step": 1603 + }, + { + "epoch": 0.11021025147725712, + "grad_norm": 1.1780202388763428, + "learning_rate": 4.998425010687484e-05, + "loss": 1.0978, + "step": 1604 + }, + { + "epoch": 0.11027896111034767, + "grad_norm": 1.2341092824935913, + "learning_rate": 4.998403657416945e-05, + "loss": 1.0688, + "step": 1605 + }, + { + "epoch": 0.11034767074343824, + "grad_norm": 1.2479525804519653, + "learning_rate": 4.9983821604147664e-05, + "loss": 1.1017, + "step": 1606 + }, + { + "epoch": 0.11041638037652879, + "grad_norm": 1.1940895318984985, + "learning_rate": 4.998360519682186e-05, + "loss": 1.1089, + "step": 1607 + }, + { + "epoch": 0.11048509000961936, + "grad_norm": 1.2112818956375122, + "learning_rate": 4.9983387352204476e-05, + "loss": 1.0854, + "step": 1608 + }, + { + "epoch": 0.11055379964270991, + "grad_norm": 1.2113357782363892, + "learning_rate": 4.998316807030805e-05, + "loss": 1.0944, + "step": 1609 + }, + { + "epoch": 0.11062250927580047, + "grad_norm": 1.218792200088501, + "learning_rate": 4.998294735114521e-05, + "loss": 1.115, + "step": 1610 + }, + { + "epoch": 0.11069121890889103, + "grad_norm": 1.202117681503296, + "learning_rate": 4.9982725194728626e-05, + "loss": 1.0875, + "step": 1611 + }, + { + "epoch": 0.11075992854198159, + "grad_norm": 1.2665085792541504, + "learning_rate": 4.9982501601071094e-05, + "loss": 1.0667, + "step": 1612 + }, + { + "epoch": 0.11082863817507214, + "grad_norm": 1.214227557182312, + "learning_rate": 4.998227657018549e-05, + "loss": 1.0241, + "step": 1613 + }, + { + "epoch": 0.11089734780816271, + "grad_norm": 1.270468831062317, + "learning_rate": 4.998205010208474e-05, + "loss": 1.0479, + "step": 1614 + }, + { + "epoch": 0.11096605744125326, + "grad_norm": 1.2347943782806396, + "learning_rate": 4.9981822196781884e-05, + "loss": 1.0699, + "step": 1615 + }, + { + "epoch": 0.11103476707434383, + "grad_norm": 1.2413369417190552, + "learning_rate": 4.998159285429003e-05, + "loss": 1.0731, + "step": 1616 + }, + { + "epoch": 0.11110347670743438, + "grad_norm": 1.2350839376449585, + "learning_rate": 4.998136207462237e-05, + "loss": 1.0411, + "step": 1617 + }, + { + "epoch": 0.11117218634052495, + "grad_norm": 1.2937819957733154, + "learning_rate": 4.998112985779218e-05, + "loss": 1.1144, + "step": 1618 + }, + { + "epoch": 0.1112408959736155, + "grad_norm": 1.349845051765442, + "learning_rate": 4.998089620381282e-05, + "loss": 1.0738, + "step": 1619 + }, + { + "epoch": 0.11130960560670607, + "grad_norm": 1.2375743389129639, + "learning_rate": 4.9980661112697734e-05, + "loss": 1.0638, + "step": 1620 + }, + { + "epoch": 0.11137831523979662, + "grad_norm": 1.3472522497177124, + "learning_rate": 4.9980424584460444e-05, + "loss": 1.1206, + "step": 1621 + }, + { + "epoch": 0.11144702487288718, + "grad_norm": 1.2837010622024536, + "learning_rate": 4.998018661911456e-05, + "loss": 1.0622, + "step": 1622 + }, + { + "epoch": 0.11151573450597774, + "grad_norm": 1.2701892852783203, + "learning_rate": 4.997994721667377e-05, + "loss": 1.0579, + "step": 1623 + }, + { + "epoch": 0.1115844441390683, + "grad_norm": 1.3427847623825073, + "learning_rate": 4.997970637715186e-05, + "loss": 1.0242, + "step": 1624 + }, + { + "epoch": 0.11165315377215886, + "grad_norm": 1.362738847732544, + "learning_rate": 4.997946410056266e-05, + "loss": 1.0711, + "step": 1625 + }, + { + "epoch": 0.11172186340524942, + "grad_norm": 1.2108924388885498, + "learning_rate": 4.997922038692012e-05, + "loss": 1.001, + "step": 1626 + }, + { + "epoch": 0.11179057303833997, + "grad_norm": 1.3250902891159058, + "learning_rate": 4.997897523623827e-05, + "loss": 0.9944, + "step": 1627 + }, + { + "epoch": 0.11185928267143054, + "grad_norm": 1.213483214378357, + "learning_rate": 4.99787286485312e-05, + "loss": 1.0643, + "step": 1628 + }, + { + "epoch": 0.11192799230452109, + "grad_norm": 1.247254490852356, + "learning_rate": 4.997848062381311e-05, + "loss": 1.0803, + "step": 1629 + }, + { + "epoch": 0.11199670193761166, + "grad_norm": 1.269310712814331, + "learning_rate": 4.997823116209824e-05, + "loss": 1.0515, + "step": 1630 + }, + { + "epoch": 0.11206541157070221, + "grad_norm": 1.2360167503356934, + "learning_rate": 4.997798026340098e-05, + "loss": 1.0533, + "step": 1631 + }, + { + "epoch": 0.11213412120379278, + "grad_norm": 1.227670669555664, + "learning_rate": 4.997772792773574e-05, + "loss": 1.0649, + "step": 1632 + }, + { + "epoch": 0.11220283083688333, + "grad_norm": 1.2066998481750488, + "learning_rate": 4.9977474155117045e-05, + "loss": 1.0785, + "step": 1633 + }, + { + "epoch": 0.1122715404699739, + "grad_norm": 1.218161702156067, + "learning_rate": 4.9977218945559477e-05, + "loss": 1.0404, + "step": 1634 + }, + { + "epoch": 0.11234025010306445, + "grad_norm": 1.1746302843093872, + "learning_rate": 4.997696229907775e-05, + "loss": 1.072, + "step": 1635 + }, + { + "epoch": 0.11240895973615501, + "grad_norm": 1.219980001449585, + "learning_rate": 4.9976704215686596e-05, + "loss": 1.0792, + "step": 1636 + }, + { + "epoch": 0.11247766936924557, + "grad_norm": 1.2770551443099976, + "learning_rate": 4.9976444695400895e-05, + "loss": 1.0727, + "step": 1637 + }, + { + "epoch": 0.11254637900233613, + "grad_norm": 1.3186520338058472, + "learning_rate": 4.997618373823555e-05, + "loss": 1.0351, + "step": 1638 + }, + { + "epoch": 0.11261508863542669, + "grad_norm": 1.2329301834106445, + "learning_rate": 4.997592134420558e-05, + "loss": 1.0329, + "step": 1639 + }, + { + "epoch": 0.11268379826851725, + "grad_norm": 1.1994075775146484, + "learning_rate": 4.997565751332609e-05, + "loss": 1.0697, + "step": 1640 + }, + { + "epoch": 0.1127525079016078, + "grad_norm": 1.3777343034744263, + "learning_rate": 4.9975392245612254e-05, + "loss": 1.1002, + "step": 1641 + }, + { + "epoch": 0.11282121753469837, + "grad_norm": 1.3160675764083862, + "learning_rate": 4.997512554107932e-05, + "loss": 1.06, + "step": 1642 + }, + { + "epoch": 0.11288992716778892, + "grad_norm": 1.2431226968765259, + "learning_rate": 4.997485739974265e-05, + "loss": 1.0671, + "step": 1643 + }, + { + "epoch": 0.11295863680087949, + "grad_norm": 1.2577093839645386, + "learning_rate": 4.9974587821617655e-05, + "loss": 1.0416, + "step": 1644 + }, + { + "epoch": 0.11302734643397004, + "grad_norm": 1.3104867935180664, + "learning_rate": 4.9974316806719855e-05, + "loss": 1.0721, + "step": 1645 + }, + { + "epoch": 0.11309605606706061, + "grad_norm": 1.1997019052505493, + "learning_rate": 4.997404435506484e-05, + "loss": 1.0303, + "step": 1646 + }, + { + "epoch": 0.11316476570015116, + "grad_norm": 1.2837848663330078, + "learning_rate": 4.997377046666828e-05, + "loss": 1.1174, + "step": 1647 + }, + { + "epoch": 0.11323347533324173, + "grad_norm": 1.2262263298034668, + "learning_rate": 4.9973495141545926e-05, + "loss": 1.0825, + "step": 1648 + }, + { + "epoch": 0.11330218496633228, + "grad_norm": 1.2141212224960327, + "learning_rate": 4.997321837971363e-05, + "loss": 1.109, + "step": 1649 + }, + { + "epoch": 0.11337089459942284, + "grad_norm": 1.3640062808990479, + "learning_rate": 4.99729401811873e-05, + "loss": 1.0768, + "step": 1650 + }, + { + "epoch": 0.1134396042325134, + "grad_norm": 1.142034649848938, + "learning_rate": 4.997266054598295e-05, + "loss": 1.0541, + "step": 1651 + }, + { + "epoch": 0.11350831386560396, + "grad_norm": 1.1967718601226807, + "learning_rate": 4.997237947411667e-05, + "loss": 1.0406, + "step": 1652 + }, + { + "epoch": 0.11357702349869452, + "grad_norm": 1.2290706634521484, + "learning_rate": 4.997209696560463e-05, + "loss": 1.0667, + "step": 1653 + }, + { + "epoch": 0.11364573313178508, + "grad_norm": 1.2722803354263306, + "learning_rate": 4.9971813020463066e-05, + "loss": 1.0674, + "step": 1654 + }, + { + "epoch": 0.11371444276487563, + "grad_norm": 1.2533915042877197, + "learning_rate": 4.997152763870834e-05, + "loss": 1.0814, + "step": 1655 + }, + { + "epoch": 0.1137831523979662, + "grad_norm": 1.276506781578064, + "learning_rate": 4.997124082035684e-05, + "loss": 1.0715, + "step": 1656 + }, + { + "epoch": 0.11385186203105675, + "grad_norm": 1.1275882720947266, + "learning_rate": 4.997095256542509e-05, + "loss": 1.0782, + "step": 1657 + }, + { + "epoch": 0.11392057166414732, + "grad_norm": 1.1819758415222168, + "learning_rate": 4.997066287392966e-05, + "loss": 1.0328, + "step": 1658 + }, + { + "epoch": 0.11398928129723787, + "grad_norm": 1.2214068174362183, + "learning_rate": 4.9970371745887224e-05, + "loss": 1.0385, + "step": 1659 + }, + { + "epoch": 0.11405799093032844, + "grad_norm": 1.1601923704147339, + "learning_rate": 4.9970079181314525e-05, + "loss": 1.109, + "step": 1660 + }, + { + "epoch": 0.11412670056341899, + "grad_norm": 1.2607474327087402, + "learning_rate": 4.996978518022839e-05, + "loss": 1.0975, + "step": 1661 + }, + { + "epoch": 0.11419541019650956, + "grad_norm": 1.1646887063980103, + "learning_rate": 4.9969489742645745e-05, + "loss": 1.0613, + "step": 1662 + }, + { + "epoch": 0.11426411982960011, + "grad_norm": 1.2083619832992554, + "learning_rate": 4.996919286858358e-05, + "loss": 1.0465, + "step": 1663 + }, + { + "epoch": 0.11433282946269067, + "grad_norm": 1.231943130493164, + "learning_rate": 4.9968894558058976e-05, + "loss": 1.0768, + "step": 1664 + }, + { + "epoch": 0.11440153909578123, + "grad_norm": 1.2569210529327393, + "learning_rate": 4.996859481108909e-05, + "loss": 1.0538, + "step": 1665 + }, + { + "epoch": 0.11447024872887179, + "grad_norm": 1.1652899980545044, + "learning_rate": 4.9968293627691165e-05, + "loss": 1.0619, + "step": 1666 + }, + { + "epoch": 0.11453895836196235, + "grad_norm": 1.1531192064285278, + "learning_rate": 4.996799100788253e-05, + "loss": 1.0216, + "step": 1667 + }, + { + "epoch": 0.11460766799505291, + "grad_norm": 1.2421181201934814, + "learning_rate": 4.99676869516806e-05, + "loss": 1.0988, + "step": 1668 + }, + { + "epoch": 0.11467637762814346, + "grad_norm": 1.1997101306915283, + "learning_rate": 4.996738145910287e-05, + "loss": 1.0406, + "step": 1669 + }, + { + "epoch": 0.11474508726123403, + "grad_norm": 1.196838617324829, + "learning_rate": 4.99670745301669e-05, + "loss": 1.0478, + "step": 1670 + }, + { + "epoch": 0.11481379689432458, + "grad_norm": 1.2698692083358765, + "learning_rate": 4.9966766164890357e-05, + "loss": 1.065, + "step": 1671 + }, + { + "epoch": 0.11488250652741515, + "grad_norm": 1.2871986627578735, + "learning_rate": 4.9966456363290984e-05, + "loss": 1.0295, + "step": 1672 + }, + { + "epoch": 0.1149512161605057, + "grad_norm": 1.2037335634231567, + "learning_rate": 4.9966145125386584e-05, + "loss": 1.0341, + "step": 1673 + }, + { + "epoch": 0.11501992579359627, + "grad_norm": 1.2422738075256348, + "learning_rate": 4.9965832451195094e-05, + "loss": 1.0537, + "step": 1674 + }, + { + "epoch": 0.11508863542668682, + "grad_norm": 1.2841640710830688, + "learning_rate": 4.9965518340734484e-05, + "loss": 1.0631, + "step": 1675 + }, + { + "epoch": 0.11515734505977739, + "grad_norm": 1.450535535812378, + "learning_rate": 4.9965202794022826e-05, + "loss": 1.0867, + "step": 1676 + }, + { + "epoch": 0.11522605469286794, + "grad_norm": 1.2018766403198242, + "learning_rate": 4.996488581107827e-05, + "loss": 1.0395, + "step": 1677 + }, + { + "epoch": 0.1152947643259585, + "grad_norm": 1.2518428564071655, + "learning_rate": 4.996456739191905e-05, + "loss": 1.0829, + "step": 1678 + }, + { + "epoch": 0.11536347395904906, + "grad_norm": 1.2460041046142578, + "learning_rate": 4.99642475365635e-05, + "loss": 1.0784, + "step": 1679 + }, + { + "epoch": 0.11543218359213962, + "grad_norm": 1.2336058616638184, + "learning_rate": 4.9963926245030005e-05, + "loss": 1.0469, + "step": 1680 + }, + { + "epoch": 0.11550089322523018, + "grad_norm": 1.23346745967865, + "learning_rate": 4.9963603517337053e-05, + "loss": 1.0856, + "step": 1681 + }, + { + "epoch": 0.11556960285832074, + "grad_norm": 1.3015215396881104, + "learning_rate": 4.996327935350321e-05, + "loss": 1.014, + "step": 1682 + }, + { + "epoch": 0.1156383124914113, + "grad_norm": 1.280089020729065, + "learning_rate": 4.9962953753547134e-05, + "loss": 1.0593, + "step": 1683 + }, + { + "epoch": 0.11570702212450186, + "grad_norm": 1.4753174781799316, + "learning_rate": 4.996262671748755e-05, + "loss": 1.066, + "step": 1684 + }, + { + "epoch": 0.11577573175759241, + "grad_norm": 1.261409044265747, + "learning_rate": 4.9962298245343265e-05, + "loss": 1.044, + "step": 1685 + }, + { + "epoch": 0.11584444139068298, + "grad_norm": 1.236315369606018, + "learning_rate": 4.996196833713318e-05, + "loss": 1.083, + "step": 1686 + }, + { + "epoch": 0.11591315102377353, + "grad_norm": 1.3093464374542236, + "learning_rate": 4.9961636992876284e-05, + "loss": 1.1397, + "step": 1687 + }, + { + "epoch": 0.1159818606568641, + "grad_norm": 1.2032585144042969, + "learning_rate": 4.996130421259163e-05, + "loss": 1.0452, + "step": 1688 + }, + { + "epoch": 0.11605057028995465, + "grad_norm": 1.258208990097046, + "learning_rate": 4.996096999629836e-05, + "loss": 1.058, + "step": 1689 + }, + { + "epoch": 0.11611927992304522, + "grad_norm": 1.256055474281311, + "learning_rate": 4.996063434401571e-05, + "loss": 1.0552, + "step": 1690 + }, + { + "epoch": 0.11618798955613577, + "grad_norm": 1.2940309047698975, + "learning_rate": 4.996029725576299e-05, + "loss": 1.0335, + "step": 1691 + }, + { + "epoch": 0.11625669918922633, + "grad_norm": 1.278304934501648, + "learning_rate": 4.995995873155958e-05, + "loss": 1.061, + "step": 1692 + }, + { + "epoch": 0.11632540882231689, + "grad_norm": 1.2653254270553589, + "learning_rate": 4.995961877142497e-05, + "loss": 1.0702, + "step": 1693 + }, + { + "epoch": 0.11639411845540745, + "grad_norm": 1.244340181350708, + "learning_rate": 4.99592773753787e-05, + "loss": 1.0291, + "step": 1694 + }, + { + "epoch": 0.116462828088498, + "grad_norm": 1.310774803161621, + "learning_rate": 4.995893454344043e-05, + "loss": 1.0751, + "step": 1695 + }, + { + "epoch": 0.11653153772158857, + "grad_norm": 1.2421014308929443, + "learning_rate": 4.9958590275629867e-05, + "loss": 1.0916, + "step": 1696 + }, + { + "epoch": 0.11660024735467912, + "grad_norm": 1.2996034622192383, + "learning_rate": 4.995824457196683e-05, + "loss": 1.1206, + "step": 1697 + }, + { + "epoch": 0.11666895698776969, + "grad_norm": 1.1796554327011108, + "learning_rate": 4.99578974324712e-05, + "loss": 1.0798, + "step": 1698 + }, + { + "epoch": 0.11673766662086024, + "grad_norm": 1.2090938091278076, + "learning_rate": 4.995754885716295e-05, + "loss": 1.0637, + "step": 1699 + }, + { + "epoch": 0.11680637625395081, + "grad_norm": 1.3023666143417358, + "learning_rate": 4.995719884606212e-05, + "loss": 1.0501, + "step": 1700 + }, + { + "epoch": 0.11687508588704136, + "grad_norm": 1.2479311227798462, + "learning_rate": 4.995684739918888e-05, + "loss": 1.0972, + "step": 1701 + }, + { + "epoch": 0.11694379552013193, + "grad_norm": 1.215667486190796, + "learning_rate": 4.995649451656341e-05, + "loss": 1.0309, + "step": 1702 + }, + { + "epoch": 0.11701250515322248, + "grad_norm": 1.3372145891189575, + "learning_rate": 4.9956140198206025e-05, + "loss": 1.0582, + "step": 1703 + }, + { + "epoch": 0.11708121478631305, + "grad_norm": 1.2506862878799438, + "learning_rate": 4.995578444413712e-05, + "loss": 1.0853, + "step": 1704 + }, + { + "epoch": 0.1171499244194036, + "grad_norm": 1.1827460527420044, + "learning_rate": 4.995542725437715e-05, + "loss": 1.0685, + "step": 1705 + }, + { + "epoch": 0.11721863405249416, + "grad_norm": 1.2823127508163452, + "learning_rate": 4.995506862894666e-05, + "loss": 1.0928, + "step": 1706 + }, + { + "epoch": 0.11728734368558472, + "grad_norm": 1.1919749975204468, + "learning_rate": 4.995470856786629e-05, + "loss": 1.0537, + "step": 1707 + }, + { + "epoch": 0.11735605331867528, + "grad_norm": 1.243523359298706, + "learning_rate": 4.995434707115676e-05, + "loss": 1.0812, + "step": 1708 + }, + { + "epoch": 0.11742476295176583, + "grad_norm": 1.2187387943267822, + "learning_rate": 4.995398413883885e-05, + "loss": 1.0461, + "step": 1709 + }, + { + "epoch": 0.1174934725848564, + "grad_norm": 1.1900115013122559, + "learning_rate": 4.995361977093345e-05, + "loss": 1.0314, + "step": 1710 + }, + { + "epoch": 0.11756218221794695, + "grad_norm": 1.2555694580078125, + "learning_rate": 4.995325396746152e-05, + "loss": 1.0669, + "step": 1711 + }, + { + "epoch": 0.11763089185103752, + "grad_norm": 1.2465200424194336, + "learning_rate": 4.99528867284441e-05, + "loss": 1.1047, + "step": 1712 + }, + { + "epoch": 0.11769960148412807, + "grad_norm": 1.2392805814743042, + "learning_rate": 4.995251805390233e-05, + "loss": 1.0693, + "step": 1713 + }, + { + "epoch": 0.11776831111721864, + "grad_norm": 1.2611545324325562, + "learning_rate": 4.99521479438574e-05, + "loss": 1.1111, + "step": 1714 + }, + { + "epoch": 0.11783702075030919, + "grad_norm": 1.2106997966766357, + "learning_rate": 4.995177639833062e-05, + "loss": 1.0924, + "step": 1715 + }, + { + "epoch": 0.11790573038339976, + "grad_norm": 1.2217506170272827, + "learning_rate": 4.995140341734336e-05, + "loss": 1.071, + "step": 1716 + }, + { + "epoch": 0.11797444001649031, + "grad_norm": 1.1610147953033447, + "learning_rate": 4.995102900091707e-05, + "loss": 1.0675, + "step": 1717 + }, + { + "epoch": 0.11804314964958088, + "grad_norm": 1.3721065521240234, + "learning_rate": 4.995065314907329e-05, + "loss": 1.013, + "step": 1718 + }, + { + "epoch": 0.11811185928267143, + "grad_norm": 1.2121676206588745, + "learning_rate": 4.995027586183366e-05, + "loss": 1.0463, + "step": 1719 + }, + { + "epoch": 0.118180568915762, + "grad_norm": 1.1983064413070679, + "learning_rate": 4.994989713921987e-05, + "loss": 1.0925, + "step": 1720 + }, + { + "epoch": 0.11824927854885255, + "grad_norm": 1.2356934547424316, + "learning_rate": 4.994951698125371e-05, + "loss": 1.0447, + "step": 1721 + }, + { + "epoch": 0.11831798818194311, + "grad_norm": 1.2637883424758911, + "learning_rate": 4.994913538795704e-05, + "loss": 1.0489, + "step": 1722 + }, + { + "epoch": 0.11838669781503366, + "grad_norm": 1.22151780128479, + "learning_rate": 4.994875235935184e-05, + "loss": 1.0877, + "step": 1723 + }, + { + "epoch": 0.11845540744812423, + "grad_norm": 1.2834657430648804, + "learning_rate": 4.994836789546011e-05, + "loss": 1.0919, + "step": 1724 + }, + { + "epoch": 0.11852411708121478, + "grad_norm": 1.2662606239318848, + "learning_rate": 4.9947981996303996e-05, + "loss": 1.0788, + "step": 1725 + }, + { + "epoch": 0.11859282671430535, + "grad_norm": 1.2773579359054565, + "learning_rate": 4.99475946619057e-05, + "loss": 1.0286, + "step": 1726 + }, + { + "epoch": 0.1186615363473959, + "grad_norm": 1.2114742994308472, + "learning_rate": 4.994720589228748e-05, + "loss": 1.0476, + "step": 1727 + }, + { + "epoch": 0.11873024598048647, + "grad_norm": 1.2318038940429688, + "learning_rate": 4.9946815687471715e-05, + "loss": 1.0596, + "step": 1728 + }, + { + "epoch": 0.11879895561357702, + "grad_norm": 1.2128164768218994, + "learning_rate": 4.9946424047480865e-05, + "loss": 1.0806, + "step": 1729 + }, + { + "epoch": 0.11886766524666759, + "grad_norm": 1.2321903705596924, + "learning_rate": 4.9946030972337446e-05, + "loss": 1.0812, + "step": 1730 + }, + { + "epoch": 0.11893637487975814, + "grad_norm": 1.1765241622924805, + "learning_rate": 4.994563646206407e-05, + "loss": 1.0219, + "step": 1731 + }, + { + "epoch": 0.1190050845128487, + "grad_norm": 1.1955032348632812, + "learning_rate": 4.9945240516683444e-05, + "loss": 1.0735, + "step": 1732 + }, + { + "epoch": 0.11907379414593926, + "grad_norm": 1.2644084692001343, + "learning_rate": 4.994484313621834e-05, + "loss": 1.0789, + "step": 1733 + }, + { + "epoch": 0.11914250377902982, + "grad_norm": 1.272268533706665, + "learning_rate": 4.994444432069162e-05, + "loss": 1.0713, + "step": 1734 + }, + { + "epoch": 0.11921121341212038, + "grad_norm": 1.2473607063293457, + "learning_rate": 4.994404407012623e-05, + "loss": 1.0747, + "step": 1735 + }, + { + "epoch": 0.11927992304521094, + "grad_norm": 1.1575753688812256, + "learning_rate": 4.9943642384545195e-05, + "loss": 1.0733, + "step": 1736 + }, + { + "epoch": 0.1193486326783015, + "grad_norm": 1.2175729274749756, + "learning_rate": 4.994323926397162e-05, + "loss": 1.0623, + "step": 1737 + }, + { + "epoch": 0.11941734231139206, + "grad_norm": 1.1797293424606323, + "learning_rate": 4.9942834708428696e-05, + "loss": 1.0512, + "step": 1738 + }, + { + "epoch": 0.11948605194448261, + "grad_norm": 1.2085540294647217, + "learning_rate": 4.9942428717939706e-05, + "loss": 1.0663, + "step": 1739 + }, + { + "epoch": 0.11955476157757318, + "grad_norm": 1.1278406381607056, + "learning_rate": 4.994202129252799e-05, + "loss": 1.0644, + "step": 1740 + }, + { + "epoch": 0.11962347121066373, + "grad_norm": 1.2739521265029907, + "learning_rate": 4.994161243221701e-05, + "loss": 1.0583, + "step": 1741 + }, + { + "epoch": 0.1196921808437543, + "grad_norm": 1.2373552322387695, + "learning_rate": 4.994120213703026e-05, + "loss": 1.0753, + "step": 1742 + }, + { + "epoch": 0.11976089047684485, + "grad_norm": 1.2334778308868408, + "learning_rate": 4.994079040699137e-05, + "loss": 1.0596, + "step": 1743 + }, + { + "epoch": 0.11982960010993542, + "grad_norm": 1.2245686054229736, + "learning_rate": 4.994037724212401e-05, + "loss": 1.0748, + "step": 1744 + }, + { + "epoch": 0.11989830974302597, + "grad_norm": 1.2519580125808716, + "learning_rate": 4.993996264245196e-05, + "loss": 1.0617, + "step": 1745 + }, + { + "epoch": 0.11996701937611653, + "grad_norm": 1.1554702520370483, + "learning_rate": 4.993954660799907e-05, + "loss": 1.0388, + "step": 1746 + }, + { + "epoch": 0.12003572900920709, + "grad_norm": 1.3334646224975586, + "learning_rate": 4.993912913878925e-05, + "loss": 1.0484, + "step": 1747 + }, + { + "epoch": 0.12010443864229765, + "grad_norm": 1.2477765083312988, + "learning_rate": 4.993871023484655e-05, + "loss": 1.0609, + "step": 1748 + }, + { + "epoch": 0.1201731482753882, + "grad_norm": 1.1983277797698975, + "learning_rate": 4.993828989619506e-05, + "loss": 1.0759, + "step": 1749 + }, + { + "epoch": 0.12024185790847877, + "grad_norm": 1.258815884590149, + "learning_rate": 4.993786812285895e-05, + "loss": 1.0741, + "step": 1750 + }, + { + "epoch": 0.12031056754156932, + "grad_norm": 1.303263783454895, + "learning_rate": 4.99374449148625e-05, + "loss": 1.0398, + "step": 1751 + }, + { + "epoch": 0.12037927717465989, + "grad_norm": 1.266694188117981, + "learning_rate": 4.993702027223004e-05, + "loss": 1.0949, + "step": 1752 + }, + { + "epoch": 0.12044798680775044, + "grad_norm": 1.2113178968429565, + "learning_rate": 4.993659419498601e-05, + "loss": 1.1044, + "step": 1753 + }, + { + "epoch": 0.12051669644084101, + "grad_norm": 1.1757824420928955, + "learning_rate": 4.993616668315493e-05, + "loss": 1.0185, + "step": 1754 + }, + { + "epoch": 0.12058540607393156, + "grad_norm": 1.269137978553772, + "learning_rate": 4.993573773676138e-05, + "loss": 1.1114, + "step": 1755 + }, + { + "epoch": 0.12065411570702213, + "grad_norm": 1.212325096130371, + "learning_rate": 4.993530735583003e-05, + "loss": 1.0437, + "step": 1756 + }, + { + "epoch": 0.12072282534011268, + "grad_norm": 1.1939510107040405, + "learning_rate": 4.993487554038567e-05, + "loss": 1.0491, + "step": 1757 + }, + { + "epoch": 0.12079153497320325, + "grad_norm": 1.1697989702224731, + "learning_rate": 4.9934442290453116e-05, + "loss": 1.0296, + "step": 1758 + }, + { + "epoch": 0.1208602446062938, + "grad_norm": 1.2638152837753296, + "learning_rate": 4.99340076060573e-05, + "loss": 1.0569, + "step": 1759 + }, + { + "epoch": 0.12092895423938436, + "grad_norm": 1.2323623895645142, + "learning_rate": 4.993357148722323e-05, + "loss": 1.0557, + "step": 1760 + }, + { + "epoch": 0.12099766387247492, + "grad_norm": 1.2282782793045044, + "learning_rate": 4.9933133933975995e-05, + "loss": 1.0838, + "step": 1761 + }, + { + "epoch": 0.12106637350556548, + "grad_norm": 1.2190513610839844, + "learning_rate": 4.993269494634077e-05, + "loss": 1.0527, + "step": 1762 + }, + { + "epoch": 0.12113508313865604, + "grad_norm": 1.2713148593902588, + "learning_rate": 4.99322545243428e-05, + "loss": 1.0711, + "step": 1763 + }, + { + "epoch": 0.1212037927717466, + "grad_norm": 1.3641690015792847, + "learning_rate": 4.993181266800744e-05, + "loss": 1.0839, + "step": 1764 + }, + { + "epoch": 0.12127250240483715, + "grad_norm": 1.2832139730453491, + "learning_rate": 4.993136937736009e-05, + "loss": 1.0744, + "step": 1765 + }, + { + "epoch": 0.12134121203792772, + "grad_norm": 1.1629066467285156, + "learning_rate": 4.9930924652426266e-05, + "loss": 1.0125, + "step": 1766 + }, + { + "epoch": 0.12140992167101827, + "grad_norm": 1.1535333395004272, + "learning_rate": 4.9930478493231546e-05, + "loss": 1.0259, + "step": 1767 + }, + { + "epoch": 0.12147863130410884, + "grad_norm": 1.2653679847717285, + "learning_rate": 4.9930030899801596e-05, + "loss": 1.0442, + "step": 1768 + }, + { + "epoch": 0.12154734093719939, + "grad_norm": 1.2368016242980957, + "learning_rate": 4.992958187216217e-05, + "loss": 1.0526, + "step": 1769 + }, + { + "epoch": 0.12161605057028996, + "grad_norm": 1.2480616569519043, + "learning_rate": 4.99291314103391e-05, + "loss": 1.0724, + "step": 1770 + }, + { + "epoch": 0.12168476020338051, + "grad_norm": 1.3095299005508423, + "learning_rate": 4.992867951435831e-05, + "loss": 1.0731, + "step": 1771 + }, + { + "epoch": 0.12175346983647108, + "grad_norm": 1.2594797611236572, + "learning_rate": 4.992822618424577e-05, + "loss": 1.0556, + "step": 1772 + }, + { + "epoch": 0.12182217946956163, + "grad_norm": 1.1636569499969482, + "learning_rate": 4.992777142002759e-05, + "loss": 1.0615, + "step": 1773 + }, + { + "epoch": 0.1218908891026522, + "grad_norm": 1.3029924631118774, + "learning_rate": 4.992731522172992e-05, + "loss": 1.0935, + "step": 1774 + }, + { + "epoch": 0.12195959873574275, + "grad_norm": 1.23307466506958, + "learning_rate": 4.9926857589378996e-05, + "loss": 1.0617, + "step": 1775 + }, + { + "epoch": 0.12202830836883331, + "grad_norm": 1.1474740505218506, + "learning_rate": 4.992639852300116e-05, + "loss": 1.016, + "step": 1776 + }, + { + "epoch": 0.12209701800192387, + "grad_norm": 1.2328764200210571, + "learning_rate": 4.992593802262281e-05, + "loss": 1.0364, + "step": 1777 + }, + { + "epoch": 0.12216572763501443, + "grad_norm": 1.2362756729125977, + "learning_rate": 4.992547608827045e-05, + "loss": 1.0576, + "step": 1778 + }, + { + "epoch": 0.12223443726810498, + "grad_norm": 1.183736801147461, + "learning_rate": 4.9925012719970646e-05, + "loss": 1.1234, + "step": 1779 + }, + { + "epoch": 0.12230314690119555, + "grad_norm": 1.2077425718307495, + "learning_rate": 4.9924547917750056e-05, + "loss": 1.0949, + "step": 1780 + }, + { + "epoch": 0.1223718565342861, + "grad_norm": 1.1995948553085327, + "learning_rate": 4.9924081681635426e-05, + "loss": 1.0084, + "step": 1781 + }, + { + "epoch": 0.12244056616737667, + "grad_norm": 1.0904030799865723, + "learning_rate": 4.992361401165357e-05, + "loss": 1.0457, + "step": 1782 + }, + { + "epoch": 0.12250927580046722, + "grad_norm": 1.3229080438613892, + "learning_rate": 4.99231449078314e-05, + "loss": 1.0845, + "step": 1783 + }, + { + "epoch": 0.12257798543355779, + "grad_norm": 1.2534668445587158, + "learning_rate": 4.9922674370195896e-05, + "loss": 1.026, + "step": 1784 + }, + { + "epoch": 0.12264669506664834, + "grad_norm": 1.1735820770263672, + "learning_rate": 4.992220239877413e-05, + "loss": 1.0441, + "step": 1785 + }, + { + "epoch": 0.1227154046997389, + "grad_norm": 1.2226263284683228, + "learning_rate": 4.992172899359326e-05, + "loss": 1.0769, + "step": 1786 + }, + { + "epoch": 0.12278411433282946, + "grad_norm": 1.1919715404510498, + "learning_rate": 4.9921254154680515e-05, + "loss": 1.0597, + "step": 1787 + }, + { + "epoch": 0.12285282396592002, + "grad_norm": 1.222571611404419, + "learning_rate": 4.9920777882063216e-05, + "loss": 1.081, + "step": 1788 + }, + { + "epoch": 0.12292153359901058, + "grad_norm": 1.1579211950302124, + "learning_rate": 4.992030017576875e-05, + "loss": 0.9928, + "step": 1789 + }, + { + "epoch": 0.12299024323210114, + "grad_norm": 1.2816745042800903, + "learning_rate": 4.991982103582462e-05, + "loss": 1.0348, + "step": 1790 + }, + { + "epoch": 0.1230589528651917, + "grad_norm": 1.21284019947052, + "learning_rate": 4.991934046225838e-05, + "loss": 1.0492, + "step": 1791 + }, + { + "epoch": 0.12312766249828226, + "grad_norm": 1.2169233560562134, + "learning_rate": 4.9918858455097676e-05, + "loss": 1.0327, + "step": 1792 + }, + { + "epoch": 0.12319637213137281, + "grad_norm": 1.191023588180542, + "learning_rate": 4.9918375014370244e-05, + "loss": 1.0872, + "step": 1793 + }, + { + "epoch": 0.12326508176446338, + "grad_norm": 1.2515642642974854, + "learning_rate": 4.991789014010388e-05, + "loss": 1.0978, + "step": 1794 + }, + { + "epoch": 0.12333379139755393, + "grad_norm": 1.1892155408859253, + "learning_rate": 4.9917403832326496e-05, + "loss": 1.0887, + "step": 1795 + }, + { + "epoch": 0.1234025010306445, + "grad_norm": 1.2209995985031128, + "learning_rate": 4.991691609106607e-05, + "loss": 1.0464, + "step": 1796 + }, + { + "epoch": 0.12347121066373505, + "grad_norm": 1.2275320291519165, + "learning_rate": 4.991642691635064e-05, + "loss": 1.0463, + "step": 1797 + }, + { + "epoch": 0.12353992029682562, + "grad_norm": 1.3610484600067139, + "learning_rate": 4.991593630820837e-05, + "loss": 1.0856, + "step": 1798 + }, + { + "epoch": 0.12360862992991617, + "grad_norm": 1.2642292976379395, + "learning_rate": 4.991544426666747e-05, + "loss": 1.0654, + "step": 1799 + }, + { + "epoch": 0.12367733956300674, + "grad_norm": 1.2563198804855347, + "learning_rate": 4.991495079175626e-05, + "loss": 1.1109, + "step": 1800 + }, + { + "epoch": 0.12374604919609729, + "grad_norm": 1.2132833003997803, + "learning_rate": 4.991445588350312e-05, + "loss": 1.021, + "step": 1801 + }, + { + "epoch": 0.12381475882918785, + "grad_norm": 1.3016748428344727, + "learning_rate": 4.991395954193652e-05, + "loss": 0.9731, + "step": 1802 + }, + { + "epoch": 0.1238834684622784, + "grad_norm": 1.373178243637085, + "learning_rate": 4.991346176708503e-05, + "loss": 1.0803, + "step": 1803 + }, + { + "epoch": 0.12395217809536897, + "grad_norm": 1.1918234825134277, + "learning_rate": 4.991296255897726e-05, + "loss": 1.0378, + "step": 1804 + }, + { + "epoch": 0.12402088772845953, + "grad_norm": 1.2061179876327515, + "learning_rate": 4.991246191764195e-05, + "loss": 1.0219, + "step": 1805 + }, + { + "epoch": 0.12408959736155009, + "grad_norm": 1.2493854761123657, + "learning_rate": 4.9911959843107905e-05, + "loss": 1.0929, + "step": 1806 + }, + { + "epoch": 0.12415830699464064, + "grad_norm": 1.2413263320922852, + "learning_rate": 4.9911456335403984e-05, + "loss": 1.0203, + "step": 1807 + }, + { + "epoch": 0.12422701662773121, + "grad_norm": 1.1752562522888184, + "learning_rate": 4.991095139455918e-05, + "loss": 1.0188, + "step": 1808 + }, + { + "epoch": 0.12429572626082176, + "grad_norm": 1.280829906463623, + "learning_rate": 4.9910445020602523e-05, + "loss": 1.0951, + "step": 1809 + }, + { + "epoch": 0.12436443589391233, + "grad_norm": 1.3041975498199463, + "learning_rate": 4.9909937213563165e-05, + "loss": 1.0675, + "step": 1810 + }, + { + "epoch": 0.12443314552700288, + "grad_norm": 1.30000901222229, + "learning_rate": 4.99094279734703e-05, + "loss": 1.0476, + "step": 1811 + }, + { + "epoch": 0.12450185516009345, + "grad_norm": 1.2183818817138672, + "learning_rate": 4.9908917300353234e-05, + "loss": 1.0783, + "step": 1812 + }, + { + "epoch": 0.124570564793184, + "grad_norm": 1.2389949560165405, + "learning_rate": 4.990840519424134e-05, + "loss": 1.0677, + "step": 1813 + }, + { + "epoch": 0.12463927442627457, + "grad_norm": 1.185084581375122, + "learning_rate": 4.990789165516409e-05, + "loss": 1.0417, + "step": 1814 + }, + { + "epoch": 0.12470798405936512, + "grad_norm": 1.1933190822601318, + "learning_rate": 4.990737668315101e-05, + "loss": 1.0279, + "step": 1815 + }, + { + "epoch": 0.12477669369245568, + "grad_norm": 1.2084872722625732, + "learning_rate": 4.990686027823174e-05, + "loss": 1.0505, + "step": 1816 + }, + { + "epoch": 0.12484540332554624, + "grad_norm": 1.1990752220153809, + "learning_rate": 4.990634244043599e-05, + "loss": 1.0324, + "step": 1817 + }, + { + "epoch": 0.1249141129586368, + "grad_norm": 1.1692872047424316, + "learning_rate": 4.990582316979354e-05, + "loss": 1.0141, + "step": 1818 + }, + { + "epoch": 0.12498282259172735, + "grad_norm": 1.197450876235962, + "learning_rate": 4.990530246633427e-05, + "loss": 1.0746, + "step": 1819 + }, + { + "epoch": 0.1250515322248179, + "grad_norm": 1.3050987720489502, + "learning_rate": 4.990478033008814e-05, + "loss": 1.0803, + "step": 1820 + }, + { + "epoch": 0.1251202418579085, + "grad_norm": 1.2882078886032104, + "learning_rate": 4.990425676108518e-05, + "loss": 1.04, + "step": 1821 + }, + { + "epoch": 0.12518895149099904, + "grad_norm": 1.1703356504440308, + "learning_rate": 4.990373175935551e-05, + "loss": 1.0429, + "step": 1822 + }, + { + "epoch": 0.1252576611240896, + "grad_norm": 1.2651312351226807, + "learning_rate": 4.9903205324929334e-05, + "loss": 1.0468, + "step": 1823 + }, + { + "epoch": 0.12532637075718014, + "grad_norm": 1.155109167098999, + "learning_rate": 4.9902677457836944e-05, + "loss": 1.034, + "step": 1824 + }, + { + "epoch": 0.12539508039027072, + "grad_norm": 1.2163504362106323, + "learning_rate": 4.990214815810871e-05, + "loss": 1.097, + "step": 1825 + }, + { + "epoch": 0.12546379002336128, + "grad_norm": 1.132635474205017, + "learning_rate": 4.9901617425775067e-05, + "loss": 1.0463, + "step": 1826 + }, + { + "epoch": 0.12553249965645183, + "grad_norm": 1.2280560731887817, + "learning_rate": 4.990108526086656e-05, + "loss": 1.0667, + "step": 1827 + }, + { + "epoch": 0.12560120928954238, + "grad_norm": 1.2402983903884888, + "learning_rate": 4.9900551663413796e-05, + "loss": 1.0797, + "step": 1828 + }, + { + "epoch": 0.12566991892263296, + "grad_norm": 1.2338190078735352, + "learning_rate": 4.990001663344749e-05, + "loss": 1.0693, + "step": 1829 + }, + { + "epoch": 0.12573862855572351, + "grad_norm": 1.167102336883545, + "learning_rate": 4.9899480170998406e-05, + "loss": 1.0718, + "step": 1830 + }, + { + "epoch": 0.12580733818881407, + "grad_norm": 1.1901339292526245, + "learning_rate": 4.98989422760974e-05, + "loss": 1.0357, + "step": 1831 + }, + { + "epoch": 0.12587604782190462, + "grad_norm": 1.2092560529708862, + "learning_rate": 4.989840294877543e-05, + "loss": 1.0605, + "step": 1832 + }, + { + "epoch": 0.1259447574549952, + "grad_norm": 1.204717993736267, + "learning_rate": 4.9897862189063525e-05, + "loss": 1.0647, + "step": 1833 + }, + { + "epoch": 0.12601346708808575, + "grad_norm": 1.1581673622131348, + "learning_rate": 4.9897319996992785e-05, + "loss": 1.0223, + "step": 1834 + }, + { + "epoch": 0.1260821767211763, + "grad_norm": 1.2164273262023926, + "learning_rate": 4.9896776372594405e-05, + "loss": 1.0308, + "step": 1835 + }, + { + "epoch": 0.12615088635426686, + "grad_norm": 1.399481177330017, + "learning_rate": 4.989623131589967e-05, + "loss": 1.115, + "step": 1836 + }, + { + "epoch": 0.12621959598735744, + "grad_norm": 1.1030441522598267, + "learning_rate": 4.9895684826939917e-05, + "loss": 1.044, + "step": 1837 + }, + { + "epoch": 0.126288305620448, + "grad_norm": 1.2337533235549927, + "learning_rate": 4.9895136905746596e-05, + "loss": 1.0578, + "step": 1838 + }, + { + "epoch": 0.12635701525353854, + "grad_norm": 1.2009621858596802, + "learning_rate": 4.9894587552351235e-05, + "loss": 1.037, + "step": 1839 + }, + { + "epoch": 0.1264257248866291, + "grad_norm": 1.199291706085205, + "learning_rate": 4.9894036766785424e-05, + "loss": 1.0417, + "step": 1840 + }, + { + "epoch": 0.12649443451971967, + "grad_norm": 1.2953300476074219, + "learning_rate": 4.989348454908086e-05, + "loss": 1.0476, + "step": 1841 + }, + { + "epoch": 0.12656314415281023, + "grad_norm": 1.2713030576705933, + "learning_rate": 4.989293089926931e-05, + "loss": 1.0951, + "step": 1842 + }, + { + "epoch": 0.12663185378590078, + "grad_norm": 1.2420439720153809, + "learning_rate": 4.989237581738262e-05, + "loss": 1.0693, + "step": 1843 + }, + { + "epoch": 0.12670056341899133, + "grad_norm": 1.1967136859893799, + "learning_rate": 4.989181930345274e-05, + "loss": 1.0287, + "step": 1844 + }, + { + "epoch": 0.1267692730520819, + "grad_norm": 1.1865209341049194, + "learning_rate": 4.9891261357511656e-05, + "loss": 1.0667, + "step": 1845 + }, + { + "epoch": 0.12683798268517246, + "grad_norm": 1.2069655656814575, + "learning_rate": 4.9890701979591485e-05, + "loss": 1.0795, + "step": 1846 + }, + { + "epoch": 0.12690669231826301, + "grad_norm": 1.210248589515686, + "learning_rate": 4.9890141169724414e-05, + "loss": 1.0502, + "step": 1847 + }, + { + "epoch": 0.12697540195135357, + "grad_norm": 1.2541717290878296, + "learning_rate": 4.988957892794269e-05, + "loss": 1.0913, + "step": 1848 + }, + { + "epoch": 0.12704411158444415, + "grad_norm": 1.1920695304870605, + "learning_rate": 4.988901525427867e-05, + "loss": 1.0549, + "step": 1849 + }, + { + "epoch": 0.1271128212175347, + "grad_norm": 1.2397167682647705, + "learning_rate": 4.9888450148764774e-05, + "loss": 1.0716, + "step": 1850 + }, + { + "epoch": 0.12718153085062525, + "grad_norm": 1.2130851745605469, + "learning_rate": 4.9887883611433526e-05, + "loss": 1.0637, + "step": 1851 + }, + { + "epoch": 0.1272502404837158, + "grad_norm": 1.169157862663269, + "learning_rate": 4.98873156423175e-05, + "loss": 1.0406, + "step": 1852 + }, + { + "epoch": 0.12731895011680638, + "grad_norm": 1.2007516622543335, + "learning_rate": 4.988674624144939e-05, + "loss": 1.0541, + "step": 1853 + }, + { + "epoch": 0.12738765974989694, + "grad_norm": 1.1627428531646729, + "learning_rate": 4.988617540886194e-05, + "loss": 1.032, + "step": 1854 + }, + { + "epoch": 0.1274563693829875, + "grad_norm": 1.255191445350647, + "learning_rate": 4.988560314458798e-05, + "loss": 1.093, + "step": 1855 + }, + { + "epoch": 0.12752507901607804, + "grad_norm": 1.222791314125061, + "learning_rate": 4.988502944866045e-05, + "loss": 1.0496, + "step": 1856 + }, + { + "epoch": 0.12759378864916862, + "grad_norm": 1.2753757238388062, + "learning_rate": 4.988445432111235e-05, + "loss": 1.062, + "step": 1857 + }, + { + "epoch": 0.12766249828225917, + "grad_norm": 1.2476376295089722, + "learning_rate": 4.9883877761976774e-05, + "loss": 1.052, + "step": 1858 + }, + { + "epoch": 0.12773120791534973, + "grad_norm": 1.2657197713851929, + "learning_rate": 4.988329977128687e-05, + "loss": 1.056, + "step": 1859 + }, + { + "epoch": 0.12779991754844028, + "grad_norm": 1.2328625917434692, + "learning_rate": 4.9882720349075915e-05, + "loss": 1.0747, + "step": 1860 + }, + { + "epoch": 0.12786862718153086, + "grad_norm": 1.20888352394104, + "learning_rate": 4.9882139495377226e-05, + "loss": 1.0671, + "step": 1861 + }, + { + "epoch": 0.1279373368146214, + "grad_norm": 1.2792268991470337, + "learning_rate": 4.9881557210224215e-05, + "loss": 1.0723, + "step": 1862 + }, + { + "epoch": 0.12800604644771196, + "grad_norm": 1.336855173110962, + "learning_rate": 4.9880973493650394e-05, + "loss": 1.0391, + "step": 1863 + }, + { + "epoch": 0.12807475608080252, + "grad_norm": 1.24221932888031, + "learning_rate": 4.988038834568934e-05, + "loss": 1.0751, + "step": 1864 + }, + { + "epoch": 0.1281434657138931, + "grad_norm": 1.249198079109192, + "learning_rate": 4.987980176637472e-05, + "loss": 1.0573, + "step": 1865 + }, + { + "epoch": 0.12821217534698365, + "grad_norm": 1.1561717987060547, + "learning_rate": 4.987921375574027e-05, + "loss": 1.0637, + "step": 1866 + }, + { + "epoch": 0.1282808849800742, + "grad_norm": 1.199523687362671, + "learning_rate": 4.9878624313819825e-05, + "loss": 1.0138, + "step": 1867 + }, + { + "epoch": 0.12834959461316475, + "grad_norm": 1.2832188606262207, + "learning_rate": 4.9878033440647284e-05, + "loss": 1.0448, + "step": 1868 + }, + { + "epoch": 0.12841830424625533, + "grad_norm": 1.1581268310546875, + "learning_rate": 4.987744113625665e-05, + "loss": 1.0578, + "step": 1869 + }, + { + "epoch": 0.12848701387934588, + "grad_norm": 1.202241063117981, + "learning_rate": 4.9876847400682e-05, + "loss": 1.0654, + "step": 1870 + }, + { + "epoch": 0.12855572351243644, + "grad_norm": 1.2851862907409668, + "learning_rate": 4.987625223395749e-05, + "loss": 1.0484, + "step": 1871 + }, + { + "epoch": 0.128624433145527, + "grad_norm": 1.2772659063339233, + "learning_rate": 4.987565563611735e-05, + "loss": 1.0762, + "step": 1872 + }, + { + "epoch": 0.12869314277861757, + "grad_norm": 1.1434767246246338, + "learning_rate": 4.9875057607195916e-05, + "loss": 1.0382, + "step": 1873 + }, + { + "epoch": 0.12876185241170812, + "grad_norm": 1.1921664476394653, + "learning_rate": 4.987445814722758e-05, + "loss": 1.0739, + "step": 1874 + }, + { + "epoch": 0.12883056204479867, + "grad_norm": 1.2452815771102905, + "learning_rate": 4.987385725624684e-05, + "loss": 1.0883, + "step": 1875 + }, + { + "epoch": 0.12889927167788923, + "grad_norm": 1.1308833360671997, + "learning_rate": 4.987325493428825e-05, + "loss": 1.0483, + "step": 1876 + }, + { + "epoch": 0.1289679813109798, + "grad_norm": 1.1240451335906982, + "learning_rate": 4.987265118138647e-05, + "loss": 1.0622, + "step": 1877 + }, + { + "epoch": 0.12903669094407036, + "grad_norm": 1.2100175619125366, + "learning_rate": 4.987204599757623e-05, + "loss": 1.0943, + "step": 1878 + }, + { + "epoch": 0.1291054005771609, + "grad_norm": 1.1979787349700928, + "learning_rate": 4.987143938289236e-05, + "loss": 1.0626, + "step": 1879 + }, + { + "epoch": 0.12917411021025146, + "grad_norm": 1.1582800149917603, + "learning_rate": 4.987083133736974e-05, + "loss": 1.0764, + "step": 1880 + }, + { + "epoch": 0.12924281984334204, + "grad_norm": 1.1735774278640747, + "learning_rate": 4.987022186104337e-05, + "loss": 1.0685, + "step": 1881 + }, + { + "epoch": 0.1293115294764326, + "grad_norm": 1.1239666938781738, + "learning_rate": 4.986961095394829e-05, + "loss": 1.077, + "step": 1882 + }, + { + "epoch": 0.12938023910952315, + "grad_norm": 1.171852469444275, + "learning_rate": 4.986899861611965e-05, + "loss": 1.0331, + "step": 1883 + }, + { + "epoch": 0.1294489487426137, + "grad_norm": 1.1573134660720825, + "learning_rate": 4.986838484759269e-05, + "loss": 1.0111, + "step": 1884 + }, + { + "epoch": 0.12951765837570428, + "grad_norm": 1.2507376670837402, + "learning_rate": 4.986776964840271e-05, + "loss": 1.0282, + "step": 1885 + }, + { + "epoch": 0.12958636800879483, + "grad_norm": 1.268500804901123, + "learning_rate": 4.986715301858511e-05, + "loss": 1.1051, + "step": 1886 + }, + { + "epoch": 0.12965507764188539, + "grad_norm": 1.1859909296035767, + "learning_rate": 4.986653495817535e-05, + "loss": 1.0693, + "step": 1887 + }, + { + "epoch": 0.12972378727497594, + "grad_norm": 1.2765653133392334, + "learning_rate": 4.986591546720901e-05, + "loss": 1.0837, + "step": 1888 + }, + { + "epoch": 0.12979249690806652, + "grad_norm": 1.2257330417633057, + "learning_rate": 4.98652945457217e-05, + "loss": 1.036, + "step": 1889 + }, + { + "epoch": 0.12986120654115707, + "grad_norm": 1.226369023323059, + "learning_rate": 4.9864672193749165e-05, + "loss": 1.0961, + "step": 1890 + }, + { + "epoch": 0.12992991617424762, + "grad_norm": 1.3136779069900513, + "learning_rate": 4.98640484113272e-05, + "loss": 1.0874, + "step": 1891 + }, + { + "epoch": 0.12999862580733818, + "grad_norm": 1.237450122833252, + "learning_rate": 4.9863423198491686e-05, + "loss": 1.0758, + "step": 1892 + }, + { + "epoch": 0.13006733544042876, + "grad_norm": 1.2138400077819824, + "learning_rate": 4.98627965552786e-05, + "loss": 1.0709, + "step": 1893 + }, + { + "epoch": 0.1301360450735193, + "grad_norm": 1.2274196147918701, + "learning_rate": 4.986216848172398e-05, + "loss": 0.9998, + "step": 1894 + }, + { + "epoch": 0.13020475470660986, + "grad_norm": 1.2252181768417358, + "learning_rate": 4.986153897786398e-05, + "loss": 1.0415, + "step": 1895 + }, + { + "epoch": 0.1302734643397004, + "grad_norm": 1.2525200843811035, + "learning_rate": 4.986090804373479e-05, + "loss": 1.0805, + "step": 1896 + }, + { + "epoch": 0.130342173972791, + "grad_norm": 1.2349554300308228, + "learning_rate": 4.9860275679372715e-05, + "loss": 1.0252, + "step": 1897 + }, + { + "epoch": 0.13041088360588154, + "grad_norm": 1.2854909896850586, + "learning_rate": 4.9859641884814145e-05, + "loss": 1.0649, + "step": 1898 + }, + { + "epoch": 0.1304795932389721, + "grad_norm": 1.2805274724960327, + "learning_rate": 4.9859006660095535e-05, + "loss": 1.0391, + "step": 1899 + }, + { + "epoch": 0.13054830287206268, + "grad_norm": 1.158811330795288, + "learning_rate": 4.985837000525343e-05, + "loss": 1.0779, + "step": 1900 + }, + { + "epoch": 0.13061701250515323, + "grad_norm": 1.175549030303955, + "learning_rate": 4.985773192032446e-05, + "loss": 1.0419, + "step": 1901 + }, + { + "epoch": 0.13068572213824378, + "grad_norm": 1.248637080192566, + "learning_rate": 4.985709240534532e-05, + "loss": 1.0449, + "step": 1902 + }, + { + "epoch": 0.13075443177133433, + "grad_norm": 1.330094814300537, + "learning_rate": 4.985645146035282e-05, + "loss": 1.0294, + "step": 1903 + }, + { + "epoch": 0.13082314140442491, + "grad_norm": 1.2334578037261963, + "learning_rate": 4.985580908538381e-05, + "loss": 1.0568, + "step": 1904 + }, + { + "epoch": 0.13089185103751547, + "grad_norm": 1.205999732017517, + "learning_rate": 4.985516528047527e-05, + "loss": 1.0224, + "step": 1905 + }, + { + "epoch": 0.13096056067060602, + "grad_norm": 1.209873080253601, + "learning_rate": 4.985452004566422e-05, + "loss": 1.0829, + "step": 1906 + }, + { + "epoch": 0.13102927030369657, + "grad_norm": 1.1964229345321655, + "learning_rate": 4.985387338098779e-05, + "loss": 1.0709, + "step": 1907 + }, + { + "epoch": 0.13109797993678715, + "grad_norm": 1.1765846014022827, + "learning_rate": 4.985322528648317e-05, + "loss": 1.0059, + "step": 1908 + }, + { + "epoch": 0.1311666895698777, + "grad_norm": 1.1809709072113037, + "learning_rate": 4.985257576218767e-05, + "loss": 1.0875, + "step": 1909 + }, + { + "epoch": 0.13123539920296826, + "grad_norm": 1.235024333000183, + "learning_rate": 4.9851924808138626e-05, + "loss": 1.0451, + "step": 1910 + }, + { + "epoch": 0.1313041088360588, + "grad_norm": 1.174133062362671, + "learning_rate": 4.985127242437351e-05, + "loss": 1.0499, + "step": 1911 + }, + { + "epoch": 0.1313728184691494, + "grad_norm": 1.1540931463241577, + "learning_rate": 4.9850618610929834e-05, + "loss": 1.0175, + "step": 1912 + }, + { + "epoch": 0.13144152810223994, + "grad_norm": 1.2096532583236694, + "learning_rate": 4.984996336784523e-05, + "loss": 1.0283, + "step": 1913 + }, + { + "epoch": 0.1315102377353305, + "grad_norm": 1.1982718706130981, + "learning_rate": 4.9849306695157386e-05, + "loss": 1.0486, + "step": 1914 + }, + { + "epoch": 0.13157894736842105, + "grad_norm": 1.2571481466293335, + "learning_rate": 4.984864859290407e-05, + "loss": 1.0826, + "step": 1915 + }, + { + "epoch": 0.13164765700151163, + "grad_norm": 1.2067022323608398, + "learning_rate": 4.9847989061123156e-05, + "loss": 1.0381, + "step": 1916 + }, + { + "epoch": 0.13171636663460218, + "grad_norm": 1.2622921466827393, + "learning_rate": 4.9847328099852584e-05, + "loss": 1.0234, + "step": 1917 + }, + { + "epoch": 0.13178507626769273, + "grad_norm": 1.1235628128051758, + "learning_rate": 4.9846665709130375e-05, + "loss": 1.0512, + "step": 1918 + }, + { + "epoch": 0.13185378590078328, + "grad_norm": 1.22051203250885, + "learning_rate": 4.984600188899464e-05, + "loss": 1.0608, + "step": 1919 + }, + { + "epoch": 0.13192249553387386, + "grad_norm": 1.193264365196228, + "learning_rate": 4.9845336639483563e-05, + "loss": 1.0671, + "step": 1920 + }, + { + "epoch": 0.13199120516696441, + "grad_norm": 1.23177969455719, + "learning_rate": 4.9844669960635416e-05, + "loss": 1.0634, + "step": 1921 + }, + { + "epoch": 0.13205991480005497, + "grad_norm": 1.171226978302002, + "learning_rate": 4.9844001852488566e-05, + "loss": 1.0237, + "step": 1922 + }, + { + "epoch": 0.13212862443314552, + "grad_norm": 1.2039718627929688, + "learning_rate": 4.984333231508143e-05, + "loss": 1.0198, + "step": 1923 + }, + { + "epoch": 0.1321973340662361, + "grad_norm": 1.2184741497039795, + "learning_rate": 4.9842661348452525e-05, + "loss": 1.0584, + "step": 1924 + }, + { + "epoch": 0.13226604369932665, + "grad_norm": 1.1749235391616821, + "learning_rate": 4.9841988952640475e-05, + "loss": 1.0442, + "step": 1925 + }, + { + "epoch": 0.1323347533324172, + "grad_norm": 1.2462007999420166, + "learning_rate": 4.9841315127683936e-05, + "loss": 1.0829, + "step": 1926 + }, + { + "epoch": 0.13240346296550776, + "grad_norm": 1.223894476890564, + "learning_rate": 4.984063987362169e-05, + "loss": 1.0668, + "step": 1927 + }, + { + "epoch": 0.13247217259859834, + "grad_norm": 1.134151577949524, + "learning_rate": 4.9839963190492576e-05, + "loss": 1.076, + "step": 1928 + }, + { + "epoch": 0.1325408822316889, + "grad_norm": 1.2133049964904785, + "learning_rate": 4.9839285078335526e-05, + "loss": 1.0621, + "step": 1929 + }, + { + "epoch": 0.13260959186477944, + "grad_norm": 1.2139127254486084, + "learning_rate": 4.983860553718955e-05, + "loss": 1.0644, + "step": 1930 + }, + { + "epoch": 0.13267830149787, + "grad_norm": 1.163832187652588, + "learning_rate": 4.983792456709375e-05, + "loss": 1.06, + "step": 1931 + }, + { + "epoch": 0.13274701113096057, + "grad_norm": 1.1919041872024536, + "learning_rate": 4.9837242168087284e-05, + "loss": 1.0247, + "step": 1932 + }, + { + "epoch": 0.13281572076405113, + "grad_norm": 1.1922816038131714, + "learning_rate": 4.983655834020943e-05, + "loss": 1.0619, + "step": 1933 + }, + { + "epoch": 0.13288443039714168, + "grad_norm": 1.2068791389465332, + "learning_rate": 4.9835873083499506e-05, + "loss": 1.0853, + "step": 1934 + }, + { + "epoch": 0.13295314003023223, + "grad_norm": 1.2028738260269165, + "learning_rate": 4.983518639799696e-05, + "loss": 1.0495, + "step": 1935 + }, + { + "epoch": 0.1330218496633228, + "grad_norm": 1.2497295141220093, + "learning_rate": 4.983449828374127e-05, + "loss": 1.0602, + "step": 1936 + }, + { + "epoch": 0.13309055929641336, + "grad_norm": 1.1984483003616333, + "learning_rate": 4.983380874077204e-05, + "loss": 1.0791, + "step": 1937 + }, + { + "epoch": 0.13315926892950392, + "grad_norm": 1.21039617061615, + "learning_rate": 4.983311776912894e-05, + "loss": 1.0662, + "step": 1938 + }, + { + "epoch": 0.13322797856259447, + "grad_norm": 1.1718097925186157, + "learning_rate": 4.9832425368851714e-05, + "loss": 1.0466, + "step": 1939 + }, + { + "epoch": 0.13329668819568505, + "grad_norm": 1.1896586418151855, + "learning_rate": 4.9831731539980194e-05, + "loss": 1.0612, + "step": 1940 + }, + { + "epoch": 0.1333653978287756, + "grad_norm": 1.1792829036712646, + "learning_rate": 4.983103628255431e-05, + "loss": 1.0562, + "step": 1941 + }, + { + "epoch": 0.13343410746186615, + "grad_norm": 1.2090370655059814, + "learning_rate": 4.983033959661404e-05, + "loss": 1.0797, + "step": 1942 + }, + { + "epoch": 0.1335028170949567, + "grad_norm": 1.2147518396377563, + "learning_rate": 4.982964148219947e-05, + "loss": 1.0527, + "step": 1943 + }, + { + "epoch": 0.13357152672804729, + "grad_norm": 1.2557188272476196, + "learning_rate": 4.9828941939350776e-05, + "loss": 1.064, + "step": 1944 + }, + { + "epoch": 0.13364023636113784, + "grad_norm": 1.153956651687622, + "learning_rate": 4.9828240968108184e-05, + "loss": 1.078, + "step": 1945 + }, + { + "epoch": 0.1337089459942284, + "grad_norm": 1.160117506980896, + "learning_rate": 4.9827538568512035e-05, + "loss": 1.0523, + "step": 1946 + }, + { + "epoch": 0.13377765562731894, + "grad_norm": 1.1819616556167603, + "learning_rate": 4.982683474060272e-05, + "loss": 1.0682, + "step": 1947 + }, + { + "epoch": 0.13384636526040952, + "grad_norm": 1.2715705633163452, + "learning_rate": 4.9826129484420744e-05, + "loss": 1.106, + "step": 1948 + }, + { + "epoch": 0.13391507489350007, + "grad_norm": 1.207221508026123, + "learning_rate": 4.982542280000667e-05, + "loss": 1.0711, + "step": 1949 + }, + { + "epoch": 0.13398378452659063, + "grad_norm": 1.398022174835205, + "learning_rate": 4.9824714687401175e-05, + "loss": 1.0149, + "step": 1950 + }, + { + "epoch": 0.13405249415968118, + "grad_norm": 1.1603754758834839, + "learning_rate": 4.982400514664497e-05, + "loss": 1.0544, + "step": 1951 + }, + { + "epoch": 0.13412120379277176, + "grad_norm": 1.2351586818695068, + "learning_rate": 4.9823294177778885e-05, + "loss": 1.0847, + "step": 1952 + }, + { + "epoch": 0.1341899134258623, + "grad_norm": 1.2711869478225708, + "learning_rate": 4.982258178084382e-05, + "loss": 1.0978, + "step": 1953 + }, + { + "epoch": 0.13425862305895286, + "grad_norm": 1.2681734561920166, + "learning_rate": 4.982186795588076e-05, + "loss": 1.0599, + "step": 1954 + }, + { + "epoch": 0.13432733269204342, + "grad_norm": 1.2139811515808105, + "learning_rate": 4.982115270293078e-05, + "loss": 1.0233, + "step": 1955 + }, + { + "epoch": 0.134396042325134, + "grad_norm": 1.2115709781646729, + "learning_rate": 4.982043602203501e-05, + "loss": 1.0354, + "step": 1956 + }, + { + "epoch": 0.13446475195822455, + "grad_norm": 1.128997564315796, + "learning_rate": 4.981971791323469e-05, + "loss": 1.0439, + "step": 1957 + }, + { + "epoch": 0.1345334615913151, + "grad_norm": 1.1622636318206787, + "learning_rate": 4.9818998376571136e-05, + "loss": 1.0348, + "step": 1958 + }, + { + "epoch": 0.13460217122440565, + "grad_norm": 1.1685645580291748, + "learning_rate": 4.981827741208574e-05, + "loss": 1.0674, + "step": 1959 + }, + { + "epoch": 0.13467088085749623, + "grad_norm": 1.1986509561538696, + "learning_rate": 4.9817555019819974e-05, + "loss": 1.0718, + "step": 1960 + }, + { + "epoch": 0.13473959049058679, + "grad_norm": 1.2427603006362915, + "learning_rate": 4.98168311998154e-05, + "loss": 1.0787, + "step": 1961 + }, + { + "epoch": 0.13480830012367734, + "grad_norm": 1.2256584167480469, + "learning_rate": 4.981610595211366e-05, + "loss": 1.0581, + "step": 1962 + }, + { + "epoch": 0.1348770097567679, + "grad_norm": 1.2192955017089844, + "learning_rate": 4.981537927675647e-05, + "loss": 1.0797, + "step": 1963 + }, + { + "epoch": 0.13494571938985847, + "grad_norm": 1.2081060409545898, + "learning_rate": 4.981465117378565e-05, + "loss": 1.0753, + "step": 1964 + }, + { + "epoch": 0.13501442902294902, + "grad_norm": 1.1172372102737427, + "learning_rate": 4.981392164324307e-05, + "loss": 1.0393, + "step": 1965 + }, + { + "epoch": 0.13508313865603958, + "grad_norm": 1.2591646909713745, + "learning_rate": 4.9813190685170716e-05, + "loss": 1.0511, + "step": 1966 + }, + { + "epoch": 0.13515184828913013, + "grad_norm": 1.2499399185180664, + "learning_rate": 4.981245829961062e-05, + "loss": 1.1162, + "step": 1967 + }, + { + "epoch": 0.1352205579222207, + "grad_norm": 1.1161015033721924, + "learning_rate": 4.981172448660494e-05, + "loss": 1.0109, + "step": 1968 + }, + { + "epoch": 0.13528926755531126, + "grad_norm": 1.2825241088867188, + "learning_rate": 4.981098924619587e-05, + "loss": 1.0513, + "step": 1969 + }, + { + "epoch": 0.1353579771884018, + "grad_norm": 1.2661597728729248, + "learning_rate": 4.981025257842572e-05, + "loss": 1.0411, + "step": 1970 + }, + { + "epoch": 0.13542668682149236, + "grad_norm": 1.2152210474014282, + "learning_rate": 4.9809514483336864e-05, + "loss": 1.0165, + "step": 1971 + }, + { + "epoch": 0.13549539645458294, + "grad_norm": 1.1764500141143799, + "learning_rate": 4.980877496097177e-05, + "loss": 1.0575, + "step": 1972 + }, + { + "epoch": 0.1355641060876735, + "grad_norm": 1.118719220161438, + "learning_rate": 4.980803401137298e-05, + "loss": 1.0352, + "step": 1973 + }, + { + "epoch": 0.13563281572076405, + "grad_norm": 1.1825342178344727, + "learning_rate": 4.980729163458312e-05, + "loss": 1.0117, + "step": 1974 + }, + { + "epoch": 0.1357015253538546, + "grad_norm": 1.1681084632873535, + "learning_rate": 4.980654783064489e-05, + "loss": 1.0864, + "step": 1975 + }, + { + "epoch": 0.13577023498694518, + "grad_norm": 1.2994136810302734, + "learning_rate": 4.9805802599601093e-05, + "loss": 1.0128, + "step": 1976 + }, + { + "epoch": 0.13583894462003573, + "grad_norm": 1.1843369007110596, + "learning_rate": 4.98050559414946e-05, + "loss": 1.0636, + "step": 1977 + }, + { + "epoch": 0.1359076542531263, + "grad_norm": 1.207662582397461, + "learning_rate": 4.980430785636837e-05, + "loss": 1.0484, + "step": 1978 + }, + { + "epoch": 0.13597636388621684, + "grad_norm": 1.2368249893188477, + "learning_rate": 4.980355834426542e-05, + "loss": 1.0581, + "step": 1979 + }, + { + "epoch": 0.13604507351930742, + "grad_norm": 1.2835763692855835, + "learning_rate": 4.9802807405228886e-05, + "loss": 1.0923, + "step": 1980 + }, + { + "epoch": 0.13611378315239797, + "grad_norm": 1.1676721572875977, + "learning_rate": 4.980205503930196e-05, + "loss": 1.0428, + "step": 1981 + }, + { + "epoch": 0.13618249278548852, + "grad_norm": 1.218184232711792, + "learning_rate": 4.980130124652794e-05, + "loss": 1.0759, + "step": 1982 + }, + { + "epoch": 0.13625120241857908, + "grad_norm": 1.2737833261489868, + "learning_rate": 4.980054602695018e-05, + "loss": 1.0623, + "step": 1983 + }, + { + "epoch": 0.13631991205166966, + "grad_norm": 1.2925821542739868, + "learning_rate": 4.979978938061213e-05, + "loss": 1.0735, + "step": 1984 + }, + { + "epoch": 0.1363886216847602, + "grad_norm": 1.1856029033660889, + "learning_rate": 4.979903130755731e-05, + "loss": 1.0252, + "step": 1985 + }, + { + "epoch": 0.13645733131785076, + "grad_norm": 1.235740065574646, + "learning_rate": 4.979827180782934e-05, + "loss": 1.0912, + "step": 1986 + }, + { + "epoch": 0.1365260409509413, + "grad_norm": 1.1967555284500122, + "learning_rate": 4.979751088147192e-05, + "loss": 1.0855, + "step": 1987 + }, + { + "epoch": 0.1365947505840319, + "grad_norm": 1.1469731330871582, + "learning_rate": 4.979674852852882e-05, + "loss": 1.0292, + "step": 1988 + }, + { + "epoch": 0.13666346021712245, + "grad_norm": 1.1653364896774292, + "learning_rate": 4.979598474904389e-05, + "loss": 1.0478, + "step": 1989 + }, + { + "epoch": 0.136732169850213, + "grad_norm": 1.234846591949463, + "learning_rate": 4.9795219543061085e-05, + "loss": 1.0875, + "step": 1990 + }, + { + "epoch": 0.13680087948330355, + "grad_norm": 1.1835161447525024, + "learning_rate": 4.9794452910624414e-05, + "loss": 1.0118, + "step": 1991 + }, + { + "epoch": 0.13686958911639413, + "grad_norm": 1.1914703845977783, + "learning_rate": 4.979368485177798e-05, + "loss": 1.0527, + "step": 1992 + }, + { + "epoch": 0.13693829874948468, + "grad_norm": 1.1221058368682861, + "learning_rate": 4.979291536656598e-05, + "loss": 1.0468, + "step": 1993 + }, + { + "epoch": 0.13700700838257523, + "grad_norm": 1.2495726346969604, + "learning_rate": 4.979214445503267e-05, + "loss": 1.039, + "step": 1994 + }, + { + "epoch": 0.1370757180156658, + "grad_norm": 1.1671959161758423, + "learning_rate": 4.979137211722241e-05, + "loss": 1.0644, + "step": 1995 + }, + { + "epoch": 0.13714442764875637, + "grad_norm": 1.1525121927261353, + "learning_rate": 4.979059835317963e-05, + "loss": 1.0275, + "step": 1996 + }, + { + "epoch": 0.13721313728184692, + "grad_norm": 1.152010202407837, + "learning_rate": 4.978982316294884e-05, + "loss": 1.0531, + "step": 1997 + }, + { + "epoch": 0.13728184691493747, + "grad_norm": 1.149851679801941, + "learning_rate": 4.978904654657463e-05, + "loss": 1.0737, + "step": 1998 + }, + { + "epoch": 0.13735055654802802, + "grad_norm": 1.1729182004928589, + "learning_rate": 4.97882685041017e-05, + "loss": 1.0374, + "step": 1999 + }, + { + "epoch": 0.1374192661811186, + "grad_norm": 1.1922054290771484, + "learning_rate": 4.9787489035574795e-05, + "loss": 1.0022, + "step": 2000 + }, + { + "epoch": 0.13748797581420916, + "grad_norm": 1.242296576499939, + "learning_rate": 4.978670814103876e-05, + "loss": 1.0527, + "step": 2001 + }, + { + "epoch": 0.1375566854472997, + "grad_norm": 1.212737798690796, + "learning_rate": 4.978592582053852e-05, + "loss": 1.092, + "step": 2002 + }, + { + "epoch": 0.13762539508039026, + "grad_norm": 1.1689212322235107, + "learning_rate": 4.9785142074119076e-05, + "loss": 1.0104, + "step": 2003 + }, + { + "epoch": 0.13769410471348084, + "grad_norm": 1.228391408920288, + "learning_rate": 4.978435690182552e-05, + "loss": 1.056, + "step": 2004 + }, + { + "epoch": 0.1377628143465714, + "grad_norm": 1.2652239799499512, + "learning_rate": 4.9783570303703026e-05, + "loss": 1.1074, + "step": 2005 + }, + { + "epoch": 0.13783152397966195, + "grad_norm": 1.2459733486175537, + "learning_rate": 4.978278227979684e-05, + "loss": 1.0693, + "step": 2006 + }, + { + "epoch": 0.1379002336127525, + "grad_norm": 1.103789210319519, + "learning_rate": 4.97819928301523e-05, + "loss": 1.0104, + "step": 2007 + }, + { + "epoch": 0.13796894324584308, + "grad_norm": 1.1873407363891602, + "learning_rate": 4.978120195481483e-05, + "loss": 1.0062, + "step": 2008 + }, + { + "epoch": 0.13803765287893363, + "grad_norm": 1.1711655855178833, + "learning_rate": 4.9780409653829916e-05, + "loss": 1.0101, + "step": 2009 + }, + { + "epoch": 0.13810636251202418, + "grad_norm": 1.3015793561935425, + "learning_rate": 4.9779615927243154e-05, + "loss": 1.0741, + "step": 2010 + }, + { + "epoch": 0.13817507214511474, + "grad_norm": 1.316821575164795, + "learning_rate": 4.9778820775100185e-05, + "loss": 1.0921, + "step": 2011 + }, + { + "epoch": 0.13824378177820532, + "grad_norm": 1.2683734893798828, + "learning_rate": 4.9778024197446764e-05, + "loss": 1.0409, + "step": 2012 + }, + { + "epoch": 0.13831249141129587, + "grad_norm": 1.194350004196167, + "learning_rate": 4.9777226194328724e-05, + "loss": 1.0948, + "step": 2013 + }, + { + "epoch": 0.13838120104438642, + "grad_norm": 1.204685926437378, + "learning_rate": 4.977642676579197e-05, + "loss": 1.0635, + "step": 2014 + }, + { + "epoch": 0.13844991067747697, + "grad_norm": 1.2427505254745483, + "learning_rate": 4.977562591188248e-05, + "loss": 1.0585, + "step": 2015 + }, + { + "epoch": 0.13851862031056755, + "grad_norm": 1.2027688026428223, + "learning_rate": 4.977482363264635e-05, + "loss": 1.0446, + "step": 2016 + }, + { + "epoch": 0.1385873299436581, + "grad_norm": 1.2124366760253906, + "learning_rate": 4.9774019928129725e-05, + "loss": 1.0585, + "step": 2017 + }, + { + "epoch": 0.13865603957674866, + "grad_norm": 1.1484651565551758, + "learning_rate": 4.977321479837883e-05, + "loss": 1.0472, + "step": 2018 + }, + { + "epoch": 0.1387247492098392, + "grad_norm": 1.1761763095855713, + "learning_rate": 4.977240824343999e-05, + "loss": 1.0408, + "step": 2019 + }, + { + "epoch": 0.1387934588429298, + "grad_norm": 1.1738284826278687, + "learning_rate": 4.9771600263359616e-05, + "loss": 1.0314, + "step": 2020 + }, + { + "epoch": 0.13886216847602034, + "grad_norm": 1.2546887397766113, + "learning_rate": 4.9770790858184184e-05, + "loss": 1.0627, + "step": 2021 + }, + { + "epoch": 0.1389308781091109, + "grad_norm": 1.177005648612976, + "learning_rate": 4.9769980027960264e-05, + "loss": 1.0545, + "step": 2022 + }, + { + "epoch": 0.13899958774220145, + "grad_norm": 1.1530095338821411, + "learning_rate": 4.9769167772734484e-05, + "loss": 1.0515, + "step": 2023 + }, + { + "epoch": 0.13906829737529203, + "grad_norm": 1.1721638441085815, + "learning_rate": 4.976835409255358e-05, + "loss": 1.075, + "step": 2024 + }, + { + "epoch": 0.13913700700838258, + "grad_norm": 1.1503596305847168, + "learning_rate": 4.976753898746438e-05, + "loss": 0.9884, + "step": 2025 + }, + { + "epoch": 0.13920571664147313, + "grad_norm": 1.1560215950012207, + "learning_rate": 4.976672245751376e-05, + "loss": 1.0514, + "step": 2026 + }, + { + "epoch": 0.13927442627456368, + "grad_norm": 1.2295459508895874, + "learning_rate": 4.9765904502748696e-05, + "loss": 1.0477, + "step": 2027 + }, + { + "epoch": 0.13934313590765426, + "grad_norm": 1.3454172611236572, + "learning_rate": 4.9765085123216244e-05, + "loss": 1.0965, + "step": 2028 + }, + { + "epoch": 0.13941184554074482, + "grad_norm": 1.191519856452942, + "learning_rate": 4.976426431896355e-05, + "loss": 1.0211, + "step": 2029 + }, + { + "epoch": 0.13948055517383537, + "grad_norm": 1.2312183380126953, + "learning_rate": 4.976344209003783e-05, + "loss": 1.0478, + "step": 2030 + }, + { + "epoch": 0.13954926480692592, + "grad_norm": 1.1668120622634888, + "learning_rate": 4.976261843648638e-05, + "loss": 1.0231, + "step": 2031 + }, + { + "epoch": 0.1396179744400165, + "grad_norm": 1.1942435503005981, + "learning_rate": 4.976179335835659e-05, + "loss": 1.038, + "step": 2032 + }, + { + "epoch": 0.13968668407310705, + "grad_norm": 1.2331167459487915, + "learning_rate": 4.9760966855695926e-05, + "loss": 1.0412, + "step": 2033 + }, + { + "epoch": 0.1397553937061976, + "grad_norm": 1.0895633697509766, + "learning_rate": 4.976013892855194e-05, + "loss": 1.053, + "step": 2034 + }, + { + "epoch": 0.13982410333928816, + "grad_norm": 1.223996639251709, + "learning_rate": 4.9759309576972255e-05, + "loss": 1.0844, + "step": 2035 + }, + { + "epoch": 0.13989281297237874, + "grad_norm": 1.127711534500122, + "learning_rate": 4.9758478801004595e-05, + "loss": 1.0374, + "step": 2036 + }, + { + "epoch": 0.1399615226054693, + "grad_norm": 1.2011858224868774, + "learning_rate": 4.975764660069674e-05, + "loss": 1.0316, + "step": 2037 + }, + { + "epoch": 0.14003023223855984, + "grad_norm": 1.1893682479858398, + "learning_rate": 4.975681297609657e-05, + "loss": 1.0897, + "step": 2038 + }, + { + "epoch": 0.1400989418716504, + "grad_norm": 1.1334235668182373, + "learning_rate": 4.975597792725205e-05, + "loss": 1.0662, + "step": 2039 + }, + { + "epoch": 0.14016765150474098, + "grad_norm": 1.2280901670455933, + "learning_rate": 4.9755141454211206e-05, + "loss": 1.0583, + "step": 2040 + }, + { + "epoch": 0.14023636113783153, + "grad_norm": 1.1949288845062256, + "learning_rate": 4.9754303557022175e-05, + "loss": 1.0768, + "step": 2041 + }, + { + "epoch": 0.14030507077092208, + "grad_norm": 1.2306663990020752, + "learning_rate": 4.975346423573315e-05, + "loss": 1.0622, + "step": 2042 + }, + { + "epoch": 0.14037378040401263, + "grad_norm": 1.254250407218933, + "learning_rate": 4.975262349039243e-05, + "loss": 1.0718, + "step": 2043 + }, + { + "epoch": 0.1404424900371032, + "grad_norm": 1.1876864433288574, + "learning_rate": 4.975178132104836e-05, + "loss": 1.0875, + "step": 2044 + }, + { + "epoch": 0.14051119967019376, + "grad_norm": 1.2631012201309204, + "learning_rate": 4.97509377277494e-05, + "loss": 1.0642, + "step": 2045 + }, + { + "epoch": 0.14057990930328432, + "grad_norm": 1.2306398153305054, + "learning_rate": 4.975009271054409e-05, + "loss": 1.0756, + "step": 2046 + }, + { + "epoch": 0.14064861893637487, + "grad_norm": 1.2833131551742554, + "learning_rate": 4.9749246269481046e-05, + "loss": 1.0401, + "step": 2047 + }, + { + "epoch": 0.14071732856946545, + "grad_norm": 1.153713583946228, + "learning_rate": 4.974839840460895e-05, + "loss": 1.0647, + "step": 2048 + }, + { + "epoch": 0.140786038202556, + "grad_norm": 1.2047966718673706, + "learning_rate": 4.9747549115976576e-05, + "loss": 1.0556, + "step": 2049 + }, + { + "epoch": 0.14085474783564655, + "grad_norm": 1.1953734159469604, + "learning_rate": 4.97466984036328e-05, + "loss": 1.0582, + "step": 2050 + }, + { + "epoch": 0.1409234574687371, + "grad_norm": 1.1580188274383545, + "learning_rate": 4.974584626762655e-05, + "loss": 1.0336, + "step": 2051 + }, + { + "epoch": 0.1409921671018277, + "grad_norm": 1.255658507347107, + "learning_rate": 4.974499270800685e-05, + "loss": 1.0625, + "step": 2052 + }, + { + "epoch": 0.14106087673491824, + "grad_norm": 1.1891038417816162, + "learning_rate": 4.9744137724822815e-05, + "loss": 1.0567, + "step": 2053 + }, + { + "epoch": 0.1411295863680088, + "grad_norm": 1.182938814163208, + "learning_rate": 4.9743281318123616e-05, + "loss": 1.0372, + "step": 2054 + }, + { + "epoch": 0.14119829600109934, + "grad_norm": 1.133491039276123, + "learning_rate": 4.974242348795853e-05, + "loss": 1.0371, + "step": 2055 + }, + { + "epoch": 0.14126700563418992, + "grad_norm": 1.1503316164016724, + "learning_rate": 4.9741564234376917e-05, + "loss": 1.0137, + "step": 2056 + }, + { + "epoch": 0.14133571526728048, + "grad_norm": 1.1781196594238281, + "learning_rate": 4.9740703557428186e-05, + "loss": 1.041, + "step": 2057 + }, + { + "epoch": 0.14140442490037103, + "grad_norm": 1.1357308626174927, + "learning_rate": 4.9739841457161876e-05, + "loss": 1.0157, + "step": 2058 + }, + { + "epoch": 0.14147313453346158, + "grad_norm": 1.1727654933929443, + "learning_rate": 4.973897793362756e-05, + "loss": 1.0343, + "step": 2059 + }, + { + "epoch": 0.14154184416655216, + "grad_norm": 1.1439012289047241, + "learning_rate": 4.973811298687493e-05, + "loss": 1.0408, + "step": 2060 + }, + { + "epoch": 0.1416105537996427, + "grad_norm": 1.2140663862228394, + "learning_rate": 4.973724661695375e-05, + "loss": 1.043, + "step": 2061 + }, + { + "epoch": 0.14167926343273327, + "grad_norm": 1.2179067134857178, + "learning_rate": 4.9736378823913854e-05, + "loss": 1.0438, + "step": 2062 + }, + { + "epoch": 0.14174797306582382, + "grad_norm": 1.1680006980895996, + "learning_rate": 4.9735509607805164e-05, + "loss": 1.04, + "step": 2063 + }, + { + "epoch": 0.1418166826989144, + "grad_norm": 1.077454924583435, + "learning_rate": 4.9734638968677694e-05, + "loss": 1.0644, + "step": 2064 + }, + { + "epoch": 0.14188539233200495, + "grad_norm": 1.219865322113037, + "learning_rate": 4.973376690658151e-05, + "loss": 1.0323, + "step": 2065 + }, + { + "epoch": 0.1419541019650955, + "grad_norm": 1.2254400253295898, + "learning_rate": 4.9732893421566814e-05, + "loss": 1.0581, + "step": 2066 + }, + { + "epoch": 0.14202281159818606, + "grad_norm": 1.233520746231079, + "learning_rate": 4.973201851368383e-05, + "loss": 1.0636, + "step": 2067 + }, + { + "epoch": 0.14209152123127664, + "grad_norm": 1.1421204805374146, + "learning_rate": 4.973114218298291e-05, + "loss": 1.0243, + "step": 2068 + }, + { + "epoch": 0.1421602308643672, + "grad_norm": 1.1251956224441528, + "learning_rate": 4.973026442951445e-05, + "loss": 1.0485, + "step": 2069 + }, + { + "epoch": 0.14222894049745774, + "grad_norm": 1.2014940977096558, + "learning_rate": 4.9729385253328965e-05, + "loss": 1.0699, + "step": 2070 + }, + { + "epoch": 0.1422976501305483, + "grad_norm": 1.132133960723877, + "learning_rate": 4.972850465447702e-05, + "loss": 1.0599, + "step": 2071 + }, + { + "epoch": 0.14236635976363887, + "grad_norm": 1.2576725482940674, + "learning_rate": 4.972762263300928e-05, + "loss": 1.0626, + "step": 2072 + }, + { + "epoch": 0.14243506939672942, + "grad_norm": 1.2905255556106567, + "learning_rate": 4.972673918897649e-05, + "loss": 1.077, + "step": 2073 + }, + { + "epoch": 0.14250377902981998, + "grad_norm": 1.2956457138061523, + "learning_rate": 4.9725854322429466e-05, + "loss": 1.0649, + "step": 2074 + }, + { + "epoch": 0.14257248866291053, + "grad_norm": 1.310103178024292, + "learning_rate": 4.9724968033419124e-05, + "loss": 1.0938, + "step": 2075 + }, + { + "epoch": 0.1426411982960011, + "grad_norm": 1.3549299240112305, + "learning_rate": 4.972408032199645e-05, + "loss": 1.0535, + "step": 2076 + }, + { + "epoch": 0.14270990792909166, + "grad_norm": 1.2673665285110474, + "learning_rate": 4.97231911882125e-05, + "loss": 1.0297, + "step": 2077 + }, + { + "epoch": 0.14277861756218221, + "grad_norm": 1.147138237953186, + "learning_rate": 4.972230063211843e-05, + "loss": 1.0511, + "step": 2078 + }, + { + "epoch": 0.14284732719527277, + "grad_norm": 1.1755274534225464, + "learning_rate": 4.97214086537655e-05, + "loss": 1.0485, + "step": 2079 + }, + { + "epoch": 0.14291603682836335, + "grad_norm": 1.1978373527526855, + "learning_rate": 4.972051525320499e-05, + "loss": 1.0308, + "step": 2080 + }, + { + "epoch": 0.1429847464614539, + "grad_norm": 1.275174856185913, + "learning_rate": 4.971962043048831e-05, + "loss": 1.0367, + "step": 2081 + }, + { + "epoch": 0.14305345609454445, + "grad_norm": 1.430925726890564, + "learning_rate": 4.971872418566694e-05, + "loss": 1.017, + "step": 2082 + }, + { + "epoch": 0.143122165727635, + "grad_norm": 1.1988416910171509, + "learning_rate": 4.9717826518792445e-05, + "loss": 1.0648, + "step": 2083 + }, + { + "epoch": 0.14319087536072558, + "grad_norm": 1.1960046291351318, + "learning_rate": 4.971692742991646e-05, + "loss": 1.0619, + "step": 2084 + }, + { + "epoch": 0.14325958499381614, + "grad_norm": 1.2281731367111206, + "learning_rate": 4.9716026919090705e-05, + "loss": 1.0661, + "step": 2085 + }, + { + "epoch": 0.1433282946269067, + "grad_norm": 1.195803165435791, + "learning_rate": 4.9715124986366994e-05, + "loss": 1.0157, + "step": 2086 + }, + { + "epoch": 0.14339700425999724, + "grad_norm": 1.2007070779800415, + "learning_rate": 4.9714221631797217e-05, + "loss": 1.0661, + "step": 2087 + }, + { + "epoch": 0.14346571389308782, + "grad_norm": 1.1287513971328735, + "learning_rate": 4.971331685543333e-05, + "loss": 1.0707, + "step": 2088 + }, + { + "epoch": 0.14353442352617837, + "grad_norm": 1.2721507549285889, + "learning_rate": 4.97124106573274e-05, + "loss": 1.0781, + "step": 2089 + }, + { + "epoch": 0.14360313315926893, + "grad_norm": 1.1666544675827026, + "learning_rate": 4.971150303753155e-05, + "loss": 1.0374, + "step": 2090 + }, + { + "epoch": 0.14367184279235948, + "grad_norm": 1.1177866458892822, + "learning_rate": 4.9710593996098e-05, + "loss": 1.0139, + "step": 2091 + }, + { + "epoch": 0.14374055242545006, + "grad_norm": 1.2890223264694214, + "learning_rate": 4.970968353307904e-05, + "loss": 1.0467, + "step": 2092 + }, + { + "epoch": 0.1438092620585406, + "grad_norm": 1.214280366897583, + "learning_rate": 4.970877164852705e-05, + "loss": 1.0512, + "step": 2093 + }, + { + "epoch": 0.14387797169163116, + "grad_norm": 1.2654207944869995, + "learning_rate": 4.97078583424945e-05, + "loss": 1.0387, + "step": 2094 + }, + { + "epoch": 0.14394668132472171, + "grad_norm": 1.2588666677474976, + "learning_rate": 4.9706943615033924e-05, + "loss": 1.063, + "step": 2095 + }, + { + "epoch": 0.1440153909578123, + "grad_norm": 1.1724265813827515, + "learning_rate": 4.9706027466197944e-05, + "loss": 1.0411, + "step": 2096 + }, + { + "epoch": 0.14408410059090285, + "grad_norm": 1.1346619129180908, + "learning_rate": 4.970510989603927e-05, + "loss": 1.0216, + "step": 2097 + }, + { + "epoch": 0.1441528102239934, + "grad_norm": 1.2387630939483643, + "learning_rate": 4.9704190904610686e-05, + "loss": 1.0671, + "step": 2098 + }, + { + "epoch": 0.14422151985708395, + "grad_norm": 1.2822259664535522, + "learning_rate": 4.9703270491965064e-05, + "loss": 1.0872, + "step": 2099 + }, + { + "epoch": 0.14429022949017453, + "grad_norm": 1.1275029182434082, + "learning_rate": 4.970234865815535e-05, + "loss": 1.0615, + "step": 2100 + }, + { + "epoch": 0.14435893912326508, + "grad_norm": 1.2970936298370361, + "learning_rate": 4.970142540323458e-05, + "loss": 1.045, + "step": 2101 + }, + { + "epoch": 0.14442764875635564, + "grad_norm": 1.2319996356964111, + "learning_rate": 4.970050072725587e-05, + "loss": 1.0544, + "step": 2102 + }, + { + "epoch": 0.1444963583894462, + "grad_norm": 1.2130082845687866, + "learning_rate": 4.9699574630272415e-05, + "loss": 1.0581, + "step": 2103 + }, + { + "epoch": 0.14456506802253677, + "grad_norm": 1.2142375707626343, + "learning_rate": 4.969864711233749e-05, + "loss": 1.0954, + "step": 2104 + }, + { + "epoch": 0.14463377765562732, + "grad_norm": 1.2757967710494995, + "learning_rate": 4.969771817350445e-05, + "loss": 1.0362, + "step": 2105 + }, + { + "epoch": 0.14470248728871787, + "grad_norm": 1.2378095388412476, + "learning_rate": 4.969678781382675e-05, + "loss": 1.0462, + "step": 2106 + }, + { + "epoch": 0.14477119692180843, + "grad_norm": 1.154741883277893, + "learning_rate": 4.96958560333579e-05, + "loss": 1.0293, + "step": 2107 + }, + { + "epoch": 0.144839906554899, + "grad_norm": 1.1773836612701416, + "learning_rate": 4.969492283215152e-05, + "loss": 1.0273, + "step": 2108 + }, + { + "epoch": 0.14490861618798956, + "grad_norm": 1.2216695547103882, + "learning_rate": 4.9693988210261275e-05, + "loss": 1.072, + "step": 2109 + }, + { + "epoch": 0.1449773258210801, + "grad_norm": 1.2240233421325684, + "learning_rate": 4.969305216774095e-05, + "loss": 1.0786, + "step": 2110 + }, + { + "epoch": 0.14504603545417066, + "grad_norm": 1.0969043970108032, + "learning_rate": 4.96921147046444e-05, + "loss": 1.0225, + "step": 2111 + }, + { + "epoch": 0.14511474508726124, + "grad_norm": 1.2231452465057373, + "learning_rate": 4.9691175821025534e-05, + "loss": 1.032, + "step": 2112 + }, + { + "epoch": 0.1451834547203518, + "grad_norm": 1.1397098302841187, + "learning_rate": 4.9690235516938377e-05, + "loss": 1.0255, + "step": 2113 + }, + { + "epoch": 0.14525216435344235, + "grad_norm": 1.2263953685760498, + "learning_rate": 4.968929379243703e-05, + "loss": 1.0407, + "step": 2114 + }, + { + "epoch": 0.1453208739865329, + "grad_norm": 1.21125328540802, + "learning_rate": 4.968835064757567e-05, + "loss": 1.0343, + "step": 2115 + }, + { + "epoch": 0.14538958361962348, + "grad_norm": 1.1043307781219482, + "learning_rate": 4.968740608240854e-05, + "loss": 1.0539, + "step": 2116 + }, + { + "epoch": 0.14545829325271403, + "grad_norm": 1.1908918619155884, + "learning_rate": 4.968646009698999e-05, + "loss": 1.0595, + "step": 2117 + }, + { + "epoch": 0.14552700288580458, + "grad_norm": 1.2230149507522583, + "learning_rate": 4.968551269137446e-05, + "loss": 1.0986, + "step": 2118 + }, + { + "epoch": 0.14559571251889514, + "grad_norm": 1.190597653388977, + "learning_rate": 4.968456386561642e-05, + "loss": 1.0479, + "step": 2119 + }, + { + "epoch": 0.14566442215198572, + "grad_norm": 1.2132030725479126, + "learning_rate": 4.968361361977047e-05, + "loss": 0.9983, + "step": 2120 + }, + { + "epoch": 0.14573313178507627, + "grad_norm": 1.4255651235580444, + "learning_rate": 4.9682661953891295e-05, + "loss": 1.0595, + "step": 2121 + }, + { + "epoch": 0.14580184141816682, + "grad_norm": 1.2591257095336914, + "learning_rate": 4.9681708868033616e-05, + "loss": 1.0593, + "step": 2122 + }, + { + "epoch": 0.14587055105125737, + "grad_norm": 1.1797186136245728, + "learning_rate": 4.968075436225228e-05, + "loss": 1.0531, + "step": 2123 + }, + { + "epoch": 0.14593926068434795, + "grad_norm": 1.1804002523422241, + "learning_rate": 4.967979843660219e-05, + "loss": 1.0031, + "step": 2124 + }, + { + "epoch": 0.1460079703174385, + "grad_norm": 1.2565667629241943, + "learning_rate": 4.967884109113835e-05, + "loss": 1.0589, + "step": 2125 + }, + { + "epoch": 0.14607667995052906, + "grad_norm": 1.1874576807022095, + "learning_rate": 4.967788232591582e-05, + "loss": 1.0723, + "step": 2126 + }, + { + "epoch": 0.1461453895836196, + "grad_norm": 1.1643282175064087, + "learning_rate": 4.967692214098977e-05, + "loss": 1.0819, + "step": 2127 + }, + { + "epoch": 0.1462140992167102, + "grad_norm": 1.2469050884246826, + "learning_rate": 4.9675960536415436e-05, + "loss": 1.0731, + "step": 2128 + }, + { + "epoch": 0.14628280884980074, + "grad_norm": 1.21713387966156, + "learning_rate": 4.967499751224814e-05, + "loss": 1.0528, + "step": 2129 + }, + { + "epoch": 0.1463515184828913, + "grad_norm": 1.2657346725463867, + "learning_rate": 4.9674033068543284e-05, + "loss": 1.0381, + "step": 2130 + }, + { + "epoch": 0.14642022811598185, + "grad_norm": 1.1718063354492188, + "learning_rate": 4.967306720535635e-05, + "loss": 1.0349, + "step": 2131 + }, + { + "epoch": 0.14648893774907243, + "grad_norm": 1.2917616367340088, + "learning_rate": 4.96720999227429e-05, + "loss": 1.056, + "step": 2132 + }, + { + "epoch": 0.14655764738216298, + "grad_norm": 1.1560237407684326, + "learning_rate": 4.967113122075859e-05, + "loss": 1.0367, + "step": 2133 + }, + { + "epoch": 0.14662635701525353, + "grad_norm": 1.234586477279663, + "learning_rate": 4.9670161099459135e-05, + "loss": 1.0432, + "step": 2134 + }, + { + "epoch": 0.14669506664834409, + "grad_norm": 1.2013591527938843, + "learning_rate": 4.9669189558900364e-05, + "loss": 1.0855, + "step": 2135 + }, + { + "epoch": 0.14676377628143467, + "grad_norm": 1.1843430995941162, + "learning_rate": 4.9668216599138154e-05, + "loss": 1.0312, + "step": 2136 + }, + { + "epoch": 0.14683248591452522, + "grad_norm": 1.271765112876892, + "learning_rate": 4.9667242220228484e-05, + "loss": 1.0956, + "step": 2137 + }, + { + "epoch": 0.14690119554761577, + "grad_norm": 1.2001084089279175, + "learning_rate": 4.966626642222741e-05, + "loss": 1.0624, + "step": 2138 + }, + { + "epoch": 0.14696990518070632, + "grad_norm": 1.2367550134658813, + "learning_rate": 4.9665289205191065e-05, + "loss": 1.0445, + "step": 2139 + }, + { + "epoch": 0.1470386148137969, + "grad_norm": 1.2834820747375488, + "learning_rate": 4.966431056917567e-05, + "loss": 1.0246, + "step": 2140 + }, + { + "epoch": 0.14710732444688746, + "grad_norm": 1.198280930519104, + "learning_rate": 4.9663330514237536e-05, + "loss": 1.0165, + "step": 2141 + }, + { + "epoch": 0.147176034079978, + "grad_norm": 1.1812949180603027, + "learning_rate": 4.966234904043303e-05, + "loss": 1.049, + "step": 2142 + }, + { + "epoch": 0.14724474371306856, + "grad_norm": 1.2257894277572632, + "learning_rate": 4.966136614781862e-05, + "loss": 1.0827, + "step": 2143 + }, + { + "epoch": 0.14731345334615914, + "grad_norm": 1.2529746294021606, + "learning_rate": 4.966038183645087e-05, + "loss": 1.0235, + "step": 2144 + }, + { + "epoch": 0.1473821629792497, + "grad_norm": 1.1491693258285522, + "learning_rate": 4.9659396106386366e-05, + "loss": 1.0022, + "step": 2145 + }, + { + "epoch": 0.14745087261234024, + "grad_norm": 1.2559800148010254, + "learning_rate": 4.9658408957681844e-05, + "loss": 1.1155, + "step": 2146 + }, + { + "epoch": 0.1475195822454308, + "grad_norm": 1.2633482217788696, + "learning_rate": 4.9657420390394095e-05, + "loss": 1.0282, + "step": 2147 + }, + { + "epoch": 0.14758829187852138, + "grad_norm": 1.203756332397461, + "learning_rate": 4.9656430404579985e-05, + "loss": 1.0694, + "step": 2148 + }, + { + "epoch": 0.14765700151161193, + "grad_norm": 1.1831135749816895, + "learning_rate": 4.965543900029647e-05, + "loss": 1.0188, + "step": 2149 + }, + { + "epoch": 0.14772571114470248, + "grad_norm": 1.2110562324523926, + "learning_rate": 4.965444617760058e-05, + "loss": 1.0591, + "step": 2150 + }, + { + "epoch": 0.14779442077779303, + "grad_norm": 1.2604767084121704, + "learning_rate": 4.965345193654943e-05, + "loss": 1.0257, + "step": 2151 + }, + { + "epoch": 0.14786313041088361, + "grad_norm": 1.2588189840316772, + "learning_rate": 4.965245627720022e-05, + "loss": 1.0371, + "step": 2152 + }, + { + "epoch": 0.14793184004397417, + "grad_norm": 1.2023391723632812, + "learning_rate": 4.965145919961024e-05, + "loss": 0.9974, + "step": 2153 + }, + { + "epoch": 0.14800054967706472, + "grad_norm": 1.2382168769836426, + "learning_rate": 4.9650460703836836e-05, + "loss": 1.1043, + "step": 2154 + }, + { + "epoch": 0.14806925931015527, + "grad_norm": 1.1742361783981323, + "learning_rate": 4.964946078993746e-05, + "loss": 1.0664, + "step": 2155 + }, + { + "epoch": 0.14813796894324585, + "grad_norm": 1.184643268585205, + "learning_rate": 4.9648459457969646e-05, + "loss": 1.0636, + "step": 2156 + }, + { + "epoch": 0.1482066785763364, + "grad_norm": 1.2309761047363281, + "learning_rate": 4.9647456707990966e-05, + "loss": 1.0644, + "step": 2157 + }, + { + "epoch": 0.14827538820942696, + "grad_norm": 1.1865662336349487, + "learning_rate": 4.964645254005915e-05, + "loss": 1.0368, + "step": 2158 + }, + { + "epoch": 0.1483440978425175, + "grad_norm": 1.138304591178894, + "learning_rate": 4.9645446954231936e-05, + "loss": 1.0233, + "step": 2159 + }, + { + "epoch": 0.1484128074756081, + "grad_norm": 1.103031039237976, + "learning_rate": 4.964443995056719e-05, + "loss": 1.0606, + "step": 2160 + }, + { + "epoch": 0.14848151710869864, + "grad_norm": 1.1865016222000122, + "learning_rate": 4.964343152912284e-05, + "loss": 1.0273, + "step": 2161 + }, + { + "epoch": 0.1485502267417892, + "grad_norm": 1.2575809955596924, + "learning_rate": 4.9642421689956896e-05, + "loss": 1.0607, + "step": 2162 + }, + { + "epoch": 0.14861893637487975, + "grad_norm": 1.136902928352356, + "learning_rate": 4.964141043312747e-05, + "loss": 1.0375, + "step": 2163 + }, + { + "epoch": 0.14868764600797033, + "grad_norm": 1.1528828144073486, + "learning_rate": 4.9640397758692715e-05, + "loss": 1.0475, + "step": 2164 + }, + { + "epoch": 0.14875635564106088, + "grad_norm": 1.1507638692855835, + "learning_rate": 4.9639383666710906e-05, + "loss": 0.9897, + "step": 2165 + }, + { + "epoch": 0.14882506527415143, + "grad_norm": 1.1984466314315796, + "learning_rate": 4.963836815724037e-05, + "loss": 1.004, + "step": 2166 + }, + { + "epoch": 0.14889377490724198, + "grad_norm": 1.2011932134628296, + "learning_rate": 4.9637351230339544e-05, + "loss": 1.0413, + "step": 2167 + }, + { + "epoch": 0.14896248454033256, + "grad_norm": 1.14418625831604, + "learning_rate": 4.963633288606692e-05, + "loss": 1.0533, + "step": 2168 + }, + { + "epoch": 0.14903119417342311, + "grad_norm": 1.2219151258468628, + "learning_rate": 4.963531312448109e-05, + "loss": 1.0488, + "step": 2169 + }, + { + "epoch": 0.14909990380651367, + "grad_norm": 1.212564468383789, + "learning_rate": 4.9634291945640724e-05, + "loss": 1.069, + "step": 2170 + }, + { + "epoch": 0.14916861343960422, + "grad_norm": 1.1682974100112915, + "learning_rate": 4.963326934960455e-05, + "loss": 1.0128, + "step": 2171 + }, + { + "epoch": 0.1492373230726948, + "grad_norm": 1.242405891418457, + "learning_rate": 4.9632245336431416e-05, + "loss": 1.0238, + "step": 2172 + }, + { + "epoch": 0.14930603270578535, + "grad_norm": 1.1936802864074707, + "learning_rate": 4.963121990618023e-05, + "loss": 1.0645, + "step": 2173 + }, + { + "epoch": 0.1493747423388759, + "grad_norm": 1.1716408729553223, + "learning_rate": 4.963019305890997e-05, + "loss": 1.0205, + "step": 2174 + }, + { + "epoch": 0.14944345197196646, + "grad_norm": 1.1128424406051636, + "learning_rate": 4.962916479467973e-05, + "loss": 1.0195, + "step": 2175 + }, + { + "epoch": 0.14951216160505704, + "grad_norm": 1.1198680400848389, + "learning_rate": 4.962813511354865e-05, + "loss": 1.0471, + "step": 2176 + }, + { + "epoch": 0.1495808712381476, + "grad_norm": 1.1038738489151, + "learning_rate": 4.9627104015575976e-05, + "loss": 1.0527, + "step": 2177 + }, + { + "epoch": 0.14964958087123814, + "grad_norm": 1.1449074745178223, + "learning_rate": 4.9626071500821034e-05, + "loss": 1.0362, + "step": 2178 + }, + { + "epoch": 0.1497182905043287, + "grad_norm": 1.1520577669143677, + "learning_rate": 4.9625037569343205e-05, + "loss": 1.0313, + "step": 2179 + }, + { + "epoch": 0.14978700013741927, + "grad_norm": 1.1937421560287476, + "learning_rate": 4.962400222120197e-05, + "loss": 0.9942, + "step": 2180 + }, + { + "epoch": 0.14985570977050983, + "grad_norm": 1.1935521364212036, + "learning_rate": 4.9622965456456914e-05, + "loss": 1.0562, + "step": 2181 + }, + { + "epoch": 0.14992441940360038, + "grad_norm": 1.1235543489456177, + "learning_rate": 4.9621927275167666e-05, + "loss": 1.0183, + "step": 2182 + }, + { + "epoch": 0.14999312903669093, + "grad_norm": 1.1562926769256592, + "learning_rate": 4.9620887677393955e-05, + "loss": 1.0318, + "step": 2183 + }, + { + "epoch": 0.1500618386697815, + "grad_norm": 1.1421089172363281, + "learning_rate": 4.961984666319559e-05, + "loss": 1.0265, + "step": 2184 + }, + { + "epoch": 0.15013054830287206, + "grad_norm": 1.1838215589523315, + "learning_rate": 4.961880423263246e-05, + "loss": 1.0195, + "step": 2185 + }, + { + "epoch": 0.15019925793596262, + "grad_norm": 1.1987862586975098, + "learning_rate": 4.961776038576453e-05, + "loss": 1.033, + "step": 2186 + }, + { + "epoch": 0.15026796756905317, + "grad_norm": 1.1488758325576782, + "learning_rate": 4.961671512265185e-05, + "loss": 1.0377, + "step": 2187 + }, + { + "epoch": 0.15033667720214375, + "grad_norm": 1.2281266450881958, + "learning_rate": 4.961566844335457e-05, + "loss": 1.0164, + "step": 2188 + }, + { + "epoch": 0.1504053868352343, + "grad_norm": 1.1961249113082886, + "learning_rate": 4.9614620347932883e-05, + "loss": 1.0674, + "step": 2189 + }, + { + "epoch": 0.15047409646832485, + "grad_norm": 1.181390404701233, + "learning_rate": 4.9613570836447095e-05, + "loss": 1.0882, + "step": 2190 + }, + { + "epoch": 0.1505428061014154, + "grad_norm": 1.2038159370422363, + "learning_rate": 4.961251990895759e-05, + "loss": 1.0642, + "step": 2191 + }, + { + "epoch": 0.15061151573450599, + "grad_norm": 1.2474219799041748, + "learning_rate": 4.961146756552483e-05, + "loss": 1.0663, + "step": 2192 + }, + { + "epoch": 0.15068022536759654, + "grad_norm": 1.2773725986480713, + "learning_rate": 4.961041380620933e-05, + "loss": 1.0314, + "step": 2193 + }, + { + "epoch": 0.1507489350006871, + "grad_norm": 1.2255923748016357, + "learning_rate": 4.960935863107174e-05, + "loss": 1.0443, + "step": 2194 + }, + { + "epoch": 0.15081764463377764, + "grad_norm": 1.1450048685073853, + "learning_rate": 4.960830204017275e-05, + "loss": 1.0292, + "step": 2195 + }, + { + "epoch": 0.15088635426686822, + "grad_norm": 1.1890852451324463, + "learning_rate": 4.9607244033573156e-05, + "loss": 1.0268, + "step": 2196 + }, + { + "epoch": 0.15095506389995877, + "grad_norm": 1.0777084827423096, + "learning_rate": 4.960618461133381e-05, + "loss": 1.0304, + "step": 2197 + }, + { + "epoch": 0.15102377353304933, + "grad_norm": 1.2112178802490234, + "learning_rate": 4.960512377351566e-05, + "loss": 1.0668, + "step": 2198 + }, + { + "epoch": 0.15109248316613988, + "grad_norm": 1.1969114542007446, + "learning_rate": 4.9604061520179746e-05, + "loss": 1.0732, + "step": 2199 + }, + { + "epoch": 0.15116119279923046, + "grad_norm": 1.2260370254516602, + "learning_rate": 4.9602997851387175e-05, + "loss": 1.0757, + "step": 2200 + }, + { + "epoch": 0.151229902432321, + "grad_norm": 1.2114545106887817, + "learning_rate": 4.960193276719914e-05, + "loss": 1.0392, + "step": 2201 + }, + { + "epoch": 0.15129861206541156, + "grad_norm": 1.1992963552474976, + "learning_rate": 4.9600866267676914e-05, + "loss": 1.0502, + "step": 2202 + }, + { + "epoch": 0.15136732169850212, + "grad_norm": 1.1586086750030518, + "learning_rate": 4.9599798352881854e-05, + "loss": 1.0104, + "step": 2203 + }, + { + "epoch": 0.1514360313315927, + "grad_norm": 1.1397020816802979, + "learning_rate": 4.959872902287539e-05, + "loss": 1.0607, + "step": 2204 + }, + { + "epoch": 0.15150474096468325, + "grad_norm": 1.1274487972259521, + "learning_rate": 4.959765827771904e-05, + "loss": 1.0301, + "step": 2205 + }, + { + "epoch": 0.1515734505977738, + "grad_norm": 1.1879081726074219, + "learning_rate": 4.959658611747442e-05, + "loss": 1.031, + "step": 2206 + }, + { + "epoch": 0.15164216023086435, + "grad_norm": 1.2725635766983032, + "learning_rate": 4.959551254220319e-05, + "loss": 1.0844, + "step": 2207 + }, + { + "epoch": 0.15171086986395493, + "grad_norm": 1.153200387954712, + "learning_rate": 4.959443755196712e-05, + "loss": 0.9964, + "step": 2208 + }, + { + "epoch": 0.15177957949704549, + "grad_norm": 1.1883119344711304, + "learning_rate": 4.959336114682806e-05, + "loss": 1.0656, + "step": 2209 + }, + { + "epoch": 0.15184828913013604, + "grad_norm": 1.2131496667861938, + "learning_rate": 4.959228332684791e-05, + "loss": 1.0625, + "step": 2210 + }, + { + "epoch": 0.1519169987632266, + "grad_norm": 1.1903493404388428, + "learning_rate": 4.9591204092088715e-05, + "loss": 1.0325, + "step": 2211 + }, + { + "epoch": 0.15198570839631717, + "grad_norm": 1.187154769897461, + "learning_rate": 4.9590123442612534e-05, + "loss": 1.0407, + "step": 2212 + }, + { + "epoch": 0.15205441802940772, + "grad_norm": 1.1666339635849, + "learning_rate": 4.958904137848155e-05, + "loss": 1.0058, + "step": 2213 + }, + { + "epoch": 0.15212312766249828, + "grad_norm": 1.1444131135940552, + "learning_rate": 4.9587957899758e-05, + "loss": 1.0103, + "step": 2214 + }, + { + "epoch": 0.15219183729558886, + "grad_norm": 1.2680962085723877, + "learning_rate": 4.958687300650423e-05, + "loss": 1.0874, + "step": 2215 + }, + { + "epoch": 0.1522605469286794, + "grad_norm": 1.2456238269805908, + "learning_rate": 4.958578669878265e-05, + "loss": 1.0668, + "step": 2216 + }, + { + "epoch": 0.15232925656176996, + "grad_norm": 1.1602126359939575, + "learning_rate": 4.958469897665575e-05, + "loss": 1.0194, + "step": 2217 + }, + { + "epoch": 0.1523979661948605, + "grad_norm": 1.1087301969528198, + "learning_rate": 4.9583609840186104e-05, + "loss": 0.9942, + "step": 2218 + }, + { + "epoch": 0.1524666758279511, + "grad_norm": 1.1833585500717163, + "learning_rate": 4.9582519289436376e-05, + "loss": 1.0343, + "step": 2219 + }, + { + "epoch": 0.15253538546104164, + "grad_norm": 1.1989959478378296, + "learning_rate": 4.95814273244693e-05, + "loss": 1.053, + "step": 2220 + }, + { + "epoch": 0.1526040950941322, + "grad_norm": 1.1625863313674927, + "learning_rate": 4.95803339453477e-05, + "loss": 1.0784, + "step": 2221 + }, + { + "epoch": 0.15267280472722275, + "grad_norm": 1.151932716369629, + "learning_rate": 4.957923915213447e-05, + "loss": 1.0704, + "step": 2222 + }, + { + "epoch": 0.15274151436031333, + "grad_norm": 1.1790019273757935, + "learning_rate": 4.957814294489261e-05, + "loss": 1.0787, + "step": 2223 + }, + { + "epoch": 0.15281022399340388, + "grad_norm": 1.214274287223816, + "learning_rate": 4.957704532368517e-05, + "loss": 1.0541, + "step": 2224 + }, + { + "epoch": 0.15287893362649443, + "grad_norm": 1.2650436162948608, + "learning_rate": 4.9575946288575296e-05, + "loss": 1.0597, + "step": 2225 + }, + { + "epoch": 0.152947643259585, + "grad_norm": 1.1213693618774414, + "learning_rate": 4.957484583962621e-05, + "loss": 1.021, + "step": 2226 + }, + { + "epoch": 0.15301635289267557, + "grad_norm": 1.276288390159607, + "learning_rate": 4.957374397690123e-05, + "loss": 1.0777, + "step": 2227 + }, + { + "epoch": 0.15308506252576612, + "grad_norm": 1.198522686958313, + "learning_rate": 4.9572640700463745e-05, + "loss": 1.0466, + "step": 2228 + }, + { + "epoch": 0.15315377215885667, + "grad_norm": 1.2206180095672607, + "learning_rate": 4.957153601037722e-05, + "loss": 1.057, + "step": 2229 + }, + { + "epoch": 0.15322248179194722, + "grad_norm": 1.1689445972442627, + "learning_rate": 4.9570429906705214e-05, + "loss": 1.0434, + "step": 2230 + }, + { + "epoch": 0.1532911914250378, + "grad_norm": 1.1838473081588745, + "learning_rate": 4.9569322389511353e-05, + "loss": 1.0439, + "step": 2231 + }, + { + "epoch": 0.15335990105812836, + "grad_norm": 1.2598369121551514, + "learning_rate": 4.956821345885936e-05, + "loss": 1.0406, + "step": 2232 + }, + { + "epoch": 0.1534286106912189, + "grad_norm": 1.2793020009994507, + "learning_rate": 4.956710311481303e-05, + "loss": 1.0473, + "step": 2233 + }, + { + "epoch": 0.15349732032430946, + "grad_norm": 1.2067623138427734, + "learning_rate": 4.956599135743622e-05, + "loss": 1.052, + "step": 2234 + }, + { + "epoch": 0.15356602995740004, + "grad_norm": 1.1950831413269043, + "learning_rate": 4.956487818679293e-05, + "loss": 1.0605, + "step": 2235 + }, + { + "epoch": 0.1536347395904906, + "grad_norm": 1.2416167259216309, + "learning_rate": 4.956376360294716e-05, + "loss": 0.9872, + "step": 2236 + }, + { + "epoch": 0.15370344922358115, + "grad_norm": 1.1258955001831055, + "learning_rate": 4.9562647605963055e-05, + "loss": 1.0126, + "step": 2237 + }, + { + "epoch": 0.1537721588566717, + "grad_norm": 1.177380919456482, + "learning_rate": 4.95615301959048e-05, + "loss": 1.1056, + "step": 2238 + }, + { + "epoch": 0.15384086848976228, + "grad_norm": 1.1367013454437256, + "learning_rate": 4.95604113728367e-05, + "loss": 1.0267, + "step": 2239 + }, + { + "epoch": 0.15390957812285283, + "grad_norm": 1.14414644241333, + "learning_rate": 4.95592911368231e-05, + "loss": 1.0468, + "step": 2240 + }, + { + "epoch": 0.15397828775594338, + "grad_norm": 1.2070467472076416, + "learning_rate": 4.955816948792846e-05, + "loss": 1.0494, + "step": 2241 + }, + { + "epoch": 0.15404699738903394, + "grad_norm": 1.2368484735488892, + "learning_rate": 4.95570464262173e-05, + "loss": 1.0637, + "step": 2242 + }, + { + "epoch": 0.15411570702212452, + "grad_norm": 1.2614622116088867, + "learning_rate": 4.955592195175424e-05, + "loss": 1.064, + "step": 2243 + }, + { + "epoch": 0.15418441665521507, + "grad_norm": 1.1434587240219116, + "learning_rate": 4.955479606460396e-05, + "loss": 1.032, + "step": 2244 + }, + { + "epoch": 0.15425312628830562, + "grad_norm": 1.13898503780365, + "learning_rate": 4.955366876483123e-05, + "loss": 0.976, + "step": 2245 + }, + { + "epoch": 0.15432183592139617, + "grad_norm": 1.1515989303588867, + "learning_rate": 4.9552540052500904e-05, + "loss": 1.0454, + "step": 2246 + }, + { + "epoch": 0.15439054555448675, + "grad_norm": 1.1650410890579224, + "learning_rate": 4.955140992767793e-05, + "loss": 1.0399, + "step": 2247 + }, + { + "epoch": 0.1544592551875773, + "grad_norm": 1.2016187906265259, + "learning_rate": 4.9550278390427306e-05, + "loss": 1.0158, + "step": 2248 + }, + { + "epoch": 0.15452796482066786, + "grad_norm": 1.2061265707015991, + "learning_rate": 4.954914544081414e-05, + "loss": 1.0841, + "step": 2249 + }, + { + "epoch": 0.1545966744537584, + "grad_norm": 1.139814019203186, + "learning_rate": 4.95480110789036e-05, + "loss": 1.0644, + "step": 2250 + }, + { + "epoch": 0.154665384086849, + "grad_norm": 1.0898452997207642, + "learning_rate": 4.954687530476095e-05, + "loss": 0.9849, + "step": 2251 + }, + { + "epoch": 0.15473409371993954, + "grad_norm": 1.244497537612915, + "learning_rate": 4.954573811845153e-05, + "loss": 1.0375, + "step": 2252 + }, + { + "epoch": 0.1548028033530301, + "grad_norm": 1.1185200214385986, + "learning_rate": 4.954459952004077e-05, + "loss": 0.9996, + "step": 2253 + }, + { + "epoch": 0.15487151298612065, + "grad_norm": 1.171263575553894, + "learning_rate": 4.9543459509594156e-05, + "loss": 1.071, + "step": 2254 + }, + { + "epoch": 0.15494022261921123, + "grad_norm": 1.116504430770874, + "learning_rate": 4.954231808717729e-05, + "loss": 1.0315, + "step": 2255 + }, + { + "epoch": 0.15500893225230178, + "grad_norm": 1.1282061338424683, + "learning_rate": 4.9541175252855826e-05, + "loss": 1.0079, + "step": 2256 + }, + { + "epoch": 0.15507764188539233, + "grad_norm": 1.1424884796142578, + "learning_rate": 4.954003100669551e-05, + "loss": 1.0461, + "step": 2257 + }, + { + "epoch": 0.15514635151848288, + "grad_norm": 1.2760591506958008, + "learning_rate": 4.953888534876218e-05, + "loss": 1.0266, + "step": 2258 + }, + { + "epoch": 0.15521506115157346, + "grad_norm": 1.12129807472229, + "learning_rate": 4.953773827912173e-05, + "loss": 1.012, + "step": 2259 + }, + { + "epoch": 0.15528377078466402, + "grad_norm": 1.1659220457077026, + "learning_rate": 4.953658979784016e-05, + "loss": 1.0353, + "step": 2260 + }, + { + "epoch": 0.15535248041775457, + "grad_norm": 1.1857622861862183, + "learning_rate": 4.953543990498355e-05, + "loss": 1.0439, + "step": 2261 + }, + { + "epoch": 0.15542119005084512, + "grad_norm": 1.3241257667541504, + "learning_rate": 4.953428860061803e-05, + "loss": 1.0414, + "step": 2262 + }, + { + "epoch": 0.1554898996839357, + "grad_norm": 1.1870286464691162, + "learning_rate": 4.953313588480984e-05, + "loss": 1.0156, + "step": 2263 + }, + { + "epoch": 0.15555860931702625, + "grad_norm": 1.209124207496643, + "learning_rate": 4.953198175762532e-05, + "loss": 1.0877, + "step": 2264 + }, + { + "epoch": 0.1556273189501168, + "grad_norm": 1.131463646888733, + "learning_rate": 4.953082621913083e-05, + "loss": 0.9949, + "step": 2265 + }, + { + "epoch": 0.15569602858320736, + "grad_norm": 1.1830835342407227, + "learning_rate": 4.9529669269392876e-05, + "loss": 1.0116, + "step": 2266 + }, + { + "epoch": 0.15576473821629794, + "grad_norm": 1.2839008569717407, + "learning_rate": 4.952851090847801e-05, + "loss": 1.0611, + "step": 2267 + }, + { + "epoch": 0.1558334478493885, + "grad_norm": 1.138162612915039, + "learning_rate": 4.952735113645286e-05, + "loss": 0.9941, + "step": 2268 + }, + { + "epoch": 0.15590215748247904, + "grad_norm": 1.1669154167175293, + "learning_rate": 4.952618995338415e-05, + "loss": 0.9865, + "step": 2269 + }, + { + "epoch": 0.1559708671155696, + "grad_norm": 1.14726984500885, + "learning_rate": 4.9525027359338696e-05, + "loss": 1.0439, + "step": 2270 + }, + { + "epoch": 0.15603957674866017, + "grad_norm": 1.1396747827529907, + "learning_rate": 4.952386335438336e-05, + "loss": 1.0385, + "step": 2271 + }, + { + "epoch": 0.15610828638175073, + "grad_norm": 1.26766836643219, + "learning_rate": 4.9522697938585125e-05, + "loss": 1.0954, + "step": 2272 + }, + { + "epoch": 0.15617699601484128, + "grad_norm": 1.1580278873443604, + "learning_rate": 4.9521531112011035e-05, + "loss": 1.0291, + "step": 2273 + }, + { + "epoch": 0.15624570564793183, + "grad_norm": 1.2258729934692383, + "learning_rate": 4.95203628747282e-05, + "loss": 1.0559, + "step": 2274 + }, + { + "epoch": 0.1563144152810224, + "grad_norm": 1.2302390336990356, + "learning_rate": 4.951919322680385e-05, + "loss": 1.0095, + "step": 2275 + }, + { + "epoch": 0.15638312491411296, + "grad_norm": 1.204282522201538, + "learning_rate": 4.951802216830525e-05, + "loss": 1.0432, + "step": 2276 + }, + { + "epoch": 0.15645183454720352, + "grad_norm": 1.1639041900634766, + "learning_rate": 4.95168496992998e-05, + "loss": 1.0094, + "step": 2277 + }, + { + "epoch": 0.15652054418029407, + "grad_norm": 1.1636120080947876, + "learning_rate": 4.951567581985493e-05, + "loss": 1.0537, + "step": 2278 + }, + { + "epoch": 0.15658925381338465, + "grad_norm": 1.1300266981124878, + "learning_rate": 4.951450053003818e-05, + "loss": 1.0538, + "step": 2279 + }, + { + "epoch": 0.1566579634464752, + "grad_norm": 1.270743489265442, + "learning_rate": 4.951332382991716e-05, + "loss": 1.0236, + "step": 2280 + }, + { + "epoch": 0.15672667307956575, + "grad_norm": 1.2583415508270264, + "learning_rate": 4.9512145719559564e-05, + "loss": 1.0774, + "step": 2281 + }, + { + "epoch": 0.1567953827126563, + "grad_norm": 1.2174373865127563, + "learning_rate": 4.9510966199033174e-05, + "loss": 1.059, + "step": 2282 + }, + { + "epoch": 0.15686409234574689, + "grad_norm": 1.2209687232971191, + "learning_rate": 4.950978526840584e-05, + "loss": 1.0906, + "step": 2283 + }, + { + "epoch": 0.15693280197883744, + "grad_norm": 1.114472508430481, + "learning_rate": 4.9508602927745505e-05, + "loss": 1.0212, + "step": 2284 + }, + { + "epoch": 0.157001511611928, + "grad_norm": 1.1647052764892578, + "learning_rate": 4.9507419177120185e-05, + "loss": 1.0308, + "step": 2285 + }, + { + "epoch": 0.15707022124501854, + "grad_norm": 1.210378885269165, + "learning_rate": 4.950623401659799e-05, + "loss": 1.0253, + "step": 2286 + }, + { + "epoch": 0.15713893087810912, + "grad_norm": 1.1792949438095093, + "learning_rate": 4.950504744624709e-05, + "loss": 1.029, + "step": 2287 + }, + { + "epoch": 0.15720764051119968, + "grad_norm": 1.1649631261825562, + "learning_rate": 4.950385946613575e-05, + "loss": 1.0691, + "step": 2288 + }, + { + "epoch": 0.15727635014429023, + "grad_norm": 1.2161858081817627, + "learning_rate": 4.950267007633232e-05, + "loss": 1.0408, + "step": 2289 + }, + { + "epoch": 0.15734505977738078, + "grad_norm": 1.208359956741333, + "learning_rate": 4.950147927690522e-05, + "loss": 1.0429, + "step": 2290 + }, + { + "epoch": 0.15741376941047136, + "grad_norm": 1.1887210607528687, + "learning_rate": 4.950028706792296e-05, + "loss": 1.0413, + "step": 2291 + }, + { + "epoch": 0.1574824790435619, + "grad_norm": 1.2434903383255005, + "learning_rate": 4.9499093449454116e-05, + "loss": 1.0805, + "step": 2292 + }, + { + "epoch": 0.15755118867665246, + "grad_norm": 1.2707335948944092, + "learning_rate": 4.949789842156736e-05, + "loss": 1.0368, + "step": 2293 + }, + { + "epoch": 0.15761989830974302, + "grad_norm": 1.1541576385498047, + "learning_rate": 4.949670198433146e-05, + "loss": 1.0634, + "step": 2294 + }, + { + "epoch": 0.1576886079428336, + "grad_norm": 1.1523042917251587, + "learning_rate": 4.949550413781522e-05, + "loss": 1.0399, + "step": 2295 + }, + { + "epoch": 0.15775731757592415, + "grad_norm": 1.1627463102340698, + "learning_rate": 4.9494304882087564e-05, + "loss": 1.0318, + "step": 2296 + }, + { + "epoch": 0.1578260272090147, + "grad_norm": 1.1945165395736694, + "learning_rate": 4.9493104217217486e-05, + "loss": 1.0766, + "step": 2297 + }, + { + "epoch": 0.15789473684210525, + "grad_norm": 1.1293286085128784, + "learning_rate": 4.9491902143274056e-05, + "loss": 1.0474, + "step": 2298 + }, + { + "epoch": 0.15796344647519583, + "grad_norm": 1.2915675640106201, + "learning_rate": 4.9490698660326426e-05, + "loss": 1.0772, + "step": 2299 + }, + { + "epoch": 0.1580321561082864, + "grad_norm": 1.2581186294555664, + "learning_rate": 4.9489493768443836e-05, + "loss": 1.0482, + "step": 2300 + }, + { + "epoch": 0.15810086574137694, + "grad_norm": 1.1289277076721191, + "learning_rate": 4.94882874676956e-05, + "loss": 1.0208, + "step": 2301 + }, + { + "epoch": 0.1581695753744675, + "grad_norm": 1.1776288747787476, + "learning_rate": 4.948707975815112e-05, + "loss": 1.0499, + "step": 2302 + }, + { + "epoch": 0.15823828500755807, + "grad_norm": 1.134703516960144, + "learning_rate": 4.948587063987988e-05, + "loss": 1.022, + "step": 2303 + }, + { + "epoch": 0.15830699464064862, + "grad_norm": 1.2057275772094727, + "learning_rate": 4.948466011295142e-05, + "loss": 1.0551, + "step": 2304 + }, + { + "epoch": 0.15837570427373918, + "grad_norm": 1.17746901512146, + "learning_rate": 4.94834481774354e-05, + "loss": 1.02, + "step": 2305 + }, + { + "epoch": 0.15844441390682973, + "grad_norm": 1.2121551036834717, + "learning_rate": 4.948223483340153e-05, + "loss": 1.0354, + "step": 2306 + }, + { + "epoch": 0.1585131235399203, + "grad_norm": 1.1366276741027832, + "learning_rate": 4.948102008091962e-05, + "loss": 1.0486, + "step": 2307 + }, + { + "epoch": 0.15858183317301086, + "grad_norm": 1.2407221794128418, + "learning_rate": 4.9479803920059556e-05, + "loss": 1.0348, + "step": 2308 + }, + { + "epoch": 0.1586505428061014, + "grad_norm": 1.1924757957458496, + "learning_rate": 4.9478586350891285e-05, + "loss": 1.023, + "step": 2309 + }, + { + "epoch": 0.15871925243919197, + "grad_norm": 1.2066031694412231, + "learning_rate": 4.947736737348488e-05, + "loss": 1.0377, + "step": 2310 + }, + { + "epoch": 0.15878796207228255, + "grad_norm": 1.2681770324707031, + "learning_rate": 4.947614698791045e-05, + "loss": 1.0675, + "step": 2311 + }, + { + "epoch": 0.1588566717053731, + "grad_norm": 1.2432379722595215, + "learning_rate": 4.9474925194238205e-05, + "loss": 1.0348, + "step": 2312 + }, + { + "epoch": 0.15892538133846365, + "grad_norm": 1.1806225776672363, + "learning_rate": 4.947370199253844e-05, + "loss": 1.0399, + "step": 2313 + }, + { + "epoch": 0.1589940909715542, + "grad_norm": 1.1315042972564697, + "learning_rate": 4.947247738288152e-05, + "loss": 1.0253, + "step": 2314 + }, + { + "epoch": 0.15906280060464478, + "grad_norm": 1.1946943998336792, + "learning_rate": 4.94712513653379e-05, + "loss": 1.0224, + "step": 2315 + }, + { + "epoch": 0.15913151023773534, + "grad_norm": 1.1887450218200684, + "learning_rate": 4.947002393997811e-05, + "loss": 1.0269, + "step": 2316 + }, + { + "epoch": 0.1592002198708259, + "grad_norm": 1.2467516660690308, + "learning_rate": 4.9468795106872765e-05, + "loss": 1.0864, + "step": 2317 + }, + { + "epoch": 0.15926892950391644, + "grad_norm": 1.1473792791366577, + "learning_rate": 4.946756486609255e-05, + "loss": 0.9868, + "step": 2318 + }, + { + "epoch": 0.15933763913700702, + "grad_norm": 1.184155821800232, + "learning_rate": 4.946633321770825e-05, + "loss": 1.0405, + "step": 2319 + }, + { + "epoch": 0.15940634877009757, + "grad_norm": 1.2065727710723877, + "learning_rate": 4.946510016179072e-05, + "loss": 1.0626, + "step": 2320 + }, + { + "epoch": 0.15947505840318812, + "grad_norm": 1.1485800743103027, + "learning_rate": 4.9463865698410896e-05, + "loss": 1.024, + "step": 2321 + }, + { + "epoch": 0.15954376803627868, + "grad_norm": 1.1607940196990967, + "learning_rate": 4.946262982763979e-05, + "loss": 1.0272, + "step": 2322 + }, + { + "epoch": 0.15961247766936926, + "grad_norm": 1.2902666330337524, + "learning_rate": 4.946139254954851e-05, + "loss": 1.047, + "step": 2323 + }, + { + "epoch": 0.1596811873024598, + "grad_norm": 1.1535648107528687, + "learning_rate": 4.946015386420823e-05, + "loss": 1.0425, + "step": 2324 + }, + { + "epoch": 0.15974989693555036, + "grad_norm": 1.2084240913391113, + "learning_rate": 4.9458913771690204e-05, + "loss": 1.0656, + "step": 2325 + }, + { + "epoch": 0.15981860656864091, + "grad_norm": 1.164624571800232, + "learning_rate": 4.94576722720658e-05, + "loss": 1.0295, + "step": 2326 + }, + { + "epoch": 0.1598873162017315, + "grad_norm": 1.2290780544281006, + "learning_rate": 4.945642936540641e-05, + "loss": 1.0698, + "step": 2327 + }, + { + "epoch": 0.15995602583482205, + "grad_norm": 1.1799705028533936, + "learning_rate": 4.9455185051783555e-05, + "loss": 1.0429, + "step": 2328 + }, + { + "epoch": 0.1600247354679126, + "grad_norm": 1.195796251296997, + "learning_rate": 4.945393933126881e-05, + "loss": 1.0521, + "step": 2329 + }, + { + "epoch": 0.16009344510100315, + "grad_norm": 1.1394927501678467, + "learning_rate": 4.945269220393385e-05, + "loss": 1.0336, + "step": 2330 + }, + { + "epoch": 0.16016215473409373, + "grad_norm": 1.066078543663025, + "learning_rate": 4.9451443669850416e-05, + "loss": 0.9916, + "step": 2331 + }, + { + "epoch": 0.16023086436718428, + "grad_norm": 1.2452435493469238, + "learning_rate": 4.9450193729090335e-05, + "loss": 1.0512, + "step": 2332 + }, + { + "epoch": 0.16029957400027484, + "grad_norm": 1.2253321409225464, + "learning_rate": 4.944894238172553e-05, + "loss": 1.0946, + "step": 2333 + }, + { + "epoch": 0.1603682836333654, + "grad_norm": 1.1744149923324585, + "learning_rate": 4.944768962782796e-05, + "loss": 1.0419, + "step": 2334 + }, + { + "epoch": 0.16043699326645597, + "grad_norm": 1.1668726205825806, + "learning_rate": 4.944643546746972e-05, + "loss": 1.0259, + "step": 2335 + }, + { + "epoch": 0.16050570289954652, + "grad_norm": 1.1528998613357544, + "learning_rate": 4.944517990072295e-05, + "loss": 1.0542, + "step": 2336 + }, + { + "epoch": 0.16057441253263707, + "grad_norm": 1.1393232345581055, + "learning_rate": 4.944392292765989e-05, + "loss": 1.0356, + "step": 2337 + }, + { + "epoch": 0.16064312216572763, + "grad_norm": 1.176249623298645, + "learning_rate": 4.944266454835284e-05, + "loss": 1.0791, + "step": 2338 + }, + { + "epoch": 0.1607118317988182, + "grad_norm": 1.1942336559295654, + "learning_rate": 4.944140476287421e-05, + "loss": 1.0546, + "step": 2339 + }, + { + "epoch": 0.16078054143190876, + "grad_norm": 1.1445409059524536, + "learning_rate": 4.944014357129646e-05, + "loss": 1.0247, + "step": 2340 + }, + { + "epoch": 0.1608492510649993, + "grad_norm": 1.1738768815994263, + "learning_rate": 4.943888097369216e-05, + "loss": 1.0436, + "step": 2341 + }, + { + "epoch": 0.16091796069808986, + "grad_norm": 1.0678972005844116, + "learning_rate": 4.943761697013393e-05, + "loss": 1.0061, + "step": 2342 + }, + { + "epoch": 0.16098667033118044, + "grad_norm": 1.130382776260376, + "learning_rate": 4.94363515606945e-05, + "loss": 1.0323, + "step": 2343 + }, + { + "epoch": 0.161055379964271, + "grad_norm": 1.1597270965576172, + "learning_rate": 4.9435084745446666e-05, + "loss": 1.0615, + "step": 2344 + }, + { + "epoch": 0.16112408959736155, + "grad_norm": 1.1440789699554443, + "learning_rate": 4.9433816524463304e-05, + "loss": 1.0153, + "step": 2345 + }, + { + "epoch": 0.1611927992304521, + "grad_norm": 1.1291786432266235, + "learning_rate": 4.943254689781738e-05, + "loss": 1.0335, + "step": 2346 + }, + { + "epoch": 0.16126150886354268, + "grad_norm": 1.140394687652588, + "learning_rate": 4.9431275865581926e-05, + "loss": 1.0305, + "step": 2347 + }, + { + "epoch": 0.16133021849663323, + "grad_norm": 1.1540391445159912, + "learning_rate": 4.943000342783007e-05, + "loss": 1.026, + "step": 2348 + }, + { + "epoch": 0.16139892812972378, + "grad_norm": 1.2391756772994995, + "learning_rate": 4.942872958463501e-05, + "loss": 1.0612, + "step": 2349 + }, + { + "epoch": 0.16146763776281434, + "grad_norm": 1.2748925685882568, + "learning_rate": 4.9427454336070034e-05, + "loss": 1.0107, + "step": 2350 + }, + { + "epoch": 0.16153634739590492, + "grad_norm": 1.111071228981018, + "learning_rate": 4.94261776822085e-05, + "loss": 1.0171, + "step": 2351 + }, + { + "epoch": 0.16160505702899547, + "grad_norm": 1.2121661901474, + "learning_rate": 4.9424899623123866e-05, + "loss": 1.041, + "step": 2352 + }, + { + "epoch": 0.16167376666208602, + "grad_norm": 1.15183687210083, + "learning_rate": 4.942362015888964e-05, + "loss": 0.9994, + "step": 2353 + }, + { + "epoch": 0.16174247629517657, + "grad_norm": 1.1844203472137451, + "learning_rate": 4.9422339289579446e-05, + "loss": 1.0483, + "step": 2354 + }, + { + "epoch": 0.16181118592826715, + "grad_norm": 1.2220244407653809, + "learning_rate": 4.942105701526696e-05, + "loss": 1.081, + "step": 2355 + }, + { + "epoch": 0.1618798955613577, + "grad_norm": 1.1834291219711304, + "learning_rate": 4.9419773336025956e-05, + "loss": 1.0273, + "step": 2356 + }, + { + "epoch": 0.16194860519444826, + "grad_norm": 1.090090274810791, + "learning_rate": 4.941848825193028e-05, + "loss": 1.0361, + "step": 2357 + }, + { + "epoch": 0.1620173148275388, + "grad_norm": 1.150828242301941, + "learning_rate": 4.941720176305386e-05, + "loss": 1.033, + "step": 2358 + }, + { + "epoch": 0.1620860244606294, + "grad_norm": 1.2889703512191772, + "learning_rate": 4.941591386947071e-05, + "loss": 1.0684, + "step": 2359 + }, + { + "epoch": 0.16215473409371994, + "grad_norm": 1.210449457168579, + "learning_rate": 4.941462457125493e-05, + "loss": 1.0313, + "step": 2360 + }, + { + "epoch": 0.1622234437268105, + "grad_norm": 1.2151168584823608, + "learning_rate": 4.941333386848068e-05, + "loss": 1.0301, + "step": 2361 + }, + { + "epoch": 0.16229215335990105, + "grad_norm": 1.1718803644180298, + "learning_rate": 4.941204176122222e-05, + "loss": 1.0285, + "step": 2362 + }, + { + "epoch": 0.16236086299299163, + "grad_norm": 1.3058661222457886, + "learning_rate": 4.9410748249553886e-05, + "loss": 1.0943, + "step": 2363 + }, + { + "epoch": 0.16242957262608218, + "grad_norm": 1.124323844909668, + "learning_rate": 4.940945333355008e-05, + "loss": 1.0215, + "step": 2364 + }, + { + "epoch": 0.16249828225917273, + "grad_norm": 1.1826387643814087, + "learning_rate": 4.9408157013285313e-05, + "loss": 1.0946, + "step": 2365 + }, + { + "epoch": 0.16256699189226329, + "grad_norm": 1.2777916193008423, + "learning_rate": 4.940685928883415e-05, + "loss": 1.0454, + "step": 2366 + }, + { + "epoch": 0.16263570152535387, + "grad_norm": 1.2365453243255615, + "learning_rate": 4.940556016027126e-05, + "loss": 0.9986, + "step": 2367 + }, + { + "epoch": 0.16270441115844442, + "grad_norm": 1.1323450803756714, + "learning_rate": 4.940425962767138e-05, + "loss": 1.0203, + "step": 2368 + }, + { + "epoch": 0.16277312079153497, + "grad_norm": 1.2462658882141113, + "learning_rate": 4.940295769110932e-05, + "loss": 0.9911, + "step": 2369 + }, + { + "epoch": 0.16284183042462552, + "grad_norm": 1.1563185453414917, + "learning_rate": 4.940165435065998e-05, + "loss": 1.0766, + "step": 2370 + }, + { + "epoch": 0.1629105400577161, + "grad_norm": 1.1422048807144165, + "learning_rate": 4.940034960639834e-05, + "loss": 1.0354, + "step": 2371 + }, + { + "epoch": 0.16297924969080665, + "grad_norm": 1.2947746515274048, + "learning_rate": 4.939904345839946e-05, + "loss": 1.0731, + "step": 2372 + }, + { + "epoch": 0.1630479593238972, + "grad_norm": 1.2166059017181396, + "learning_rate": 4.93977359067385e-05, + "loss": 1.0272, + "step": 2373 + }, + { + "epoch": 0.16311666895698776, + "grad_norm": 1.1052334308624268, + "learning_rate": 4.939642695149067e-05, + "loss": 1.046, + "step": 2374 + }, + { + "epoch": 0.16318537859007834, + "grad_norm": 1.2061291933059692, + "learning_rate": 4.939511659273126e-05, + "loss": 1.0118, + "step": 2375 + }, + { + "epoch": 0.1632540882231689, + "grad_norm": 1.1958491802215576, + "learning_rate": 4.939380483053568e-05, + "loss": 1.0508, + "step": 2376 + }, + { + "epoch": 0.16332279785625944, + "grad_norm": 1.181037425994873, + "learning_rate": 4.939249166497937e-05, + "loss": 1.0565, + "step": 2377 + }, + { + "epoch": 0.16339150748935, + "grad_norm": 1.1875698566436768, + "learning_rate": 4.93911770961379e-05, + "loss": 1.073, + "step": 2378 + }, + { + "epoch": 0.16346021712244058, + "grad_norm": 1.1772834062576294, + "learning_rate": 4.938986112408688e-05, + "loss": 1.0355, + "step": 2379 + }, + { + "epoch": 0.16352892675553113, + "grad_norm": 1.171299695968628, + "learning_rate": 4.938854374890202e-05, + "loss": 1.038, + "step": 2380 + }, + { + "epoch": 0.16359763638862168, + "grad_norm": 1.1699671745300293, + "learning_rate": 4.93872249706591e-05, + "loss": 0.9961, + "step": 2381 + }, + { + "epoch": 0.16366634602171223, + "grad_norm": 1.151211142539978, + "learning_rate": 4.9385904789434004e-05, + "loss": 1.0469, + "step": 2382 + }, + { + "epoch": 0.1637350556548028, + "grad_norm": 1.1677616834640503, + "learning_rate": 4.9384583205302675e-05, + "loss": 1.0268, + "step": 2383 + }, + { + "epoch": 0.16380376528789337, + "grad_norm": 1.1547976732254028, + "learning_rate": 4.938326021834114e-05, + "loss": 1.009, + "step": 2384 + }, + { + "epoch": 0.16387247492098392, + "grad_norm": 1.2309494018554688, + "learning_rate": 4.938193582862552e-05, + "loss": 1.0604, + "step": 2385 + }, + { + "epoch": 0.16394118455407447, + "grad_norm": 1.2136471271514893, + "learning_rate": 4.938061003623199e-05, + "loss": 1.094, + "step": 2386 + }, + { + "epoch": 0.16400989418716505, + "grad_norm": 1.1671332120895386, + "learning_rate": 4.937928284123683e-05, + "loss": 1.0153, + "step": 2387 + }, + { + "epoch": 0.1640786038202556, + "grad_norm": 1.1996675729751587, + "learning_rate": 4.93779542437164e-05, + "loss": 1.0445, + "step": 2388 + }, + { + "epoch": 0.16414731345334616, + "grad_norm": 1.112322449684143, + "learning_rate": 4.937662424374712e-05, + "loss": 1.0328, + "step": 2389 + }, + { + "epoch": 0.1642160230864367, + "grad_norm": 1.0915433168411255, + "learning_rate": 4.9375292841405504e-05, + "loss": 1.001, + "step": 2390 + }, + { + "epoch": 0.1642847327195273, + "grad_norm": 1.2343658208847046, + "learning_rate": 4.9373960036768165e-05, + "loss": 1.0464, + "step": 2391 + }, + { + "epoch": 0.16435344235261784, + "grad_norm": 1.0603348016738892, + "learning_rate": 4.937262582991176e-05, + "loss": 0.9998, + "step": 2392 + }, + { + "epoch": 0.1644221519857084, + "grad_norm": 1.200046181678772, + "learning_rate": 4.9371290220913056e-05, + "loss": 1.0264, + "step": 2393 + }, + { + "epoch": 0.16449086161879894, + "grad_norm": 1.1193976402282715, + "learning_rate": 4.9369953209848884e-05, + "loss": 1.0119, + "step": 2394 + }, + { + "epoch": 0.16455957125188952, + "grad_norm": 1.2449499368667603, + "learning_rate": 4.936861479679617e-05, + "loss": 1.0413, + "step": 2395 + }, + { + "epoch": 0.16462828088498008, + "grad_norm": 1.1432477235794067, + "learning_rate": 4.936727498183189e-05, + "loss": 1.0164, + "step": 2396 + }, + { + "epoch": 0.16469699051807063, + "grad_norm": 1.184070348739624, + "learning_rate": 4.936593376503315e-05, + "loss": 1.0436, + "step": 2397 + }, + { + "epoch": 0.16476570015116118, + "grad_norm": 1.1424431800842285, + "learning_rate": 4.936459114647709e-05, + "loss": 1.0073, + "step": 2398 + }, + { + "epoch": 0.16483440978425176, + "grad_norm": 1.2368580102920532, + "learning_rate": 4.936324712624096e-05, + "loss": 1.0479, + "step": 2399 + }, + { + "epoch": 0.16490311941734231, + "grad_norm": 1.186348557472229, + "learning_rate": 4.936190170440208e-05, + "loss": 0.9863, + "step": 2400 + }, + { + "epoch": 0.16497182905043287, + "grad_norm": 1.2168073654174805, + "learning_rate": 4.9360554881037855e-05, + "loss": 1.0416, + "step": 2401 + }, + { + "epoch": 0.16504053868352342, + "grad_norm": 1.1231858730316162, + "learning_rate": 4.9359206656225755e-05, + "loss": 1.0321, + "step": 2402 + }, + { + "epoch": 0.165109248316614, + "grad_norm": 1.135892391204834, + "learning_rate": 4.9357857030043343e-05, + "loss": 1.0058, + "step": 2403 + }, + { + "epoch": 0.16517795794970455, + "grad_norm": 1.1612383127212524, + "learning_rate": 4.935650600256827e-05, + "loss": 1.019, + "step": 2404 + }, + { + "epoch": 0.1652466675827951, + "grad_norm": 1.215786099433899, + "learning_rate": 4.935515357387827e-05, + "loss": 1.0474, + "step": 2405 + }, + { + "epoch": 0.16531537721588566, + "grad_norm": 1.2504756450653076, + "learning_rate": 4.935379974405112e-05, + "loss": 1.0355, + "step": 2406 + }, + { + "epoch": 0.16538408684897624, + "grad_norm": 1.1821438074111938, + "learning_rate": 4.935244451316472e-05, + "loss": 1.0054, + "step": 2407 + }, + { + "epoch": 0.1654527964820668, + "grad_norm": 1.1680669784545898, + "learning_rate": 4.935108788129704e-05, + "loss": 1.0182, + "step": 2408 + }, + { + "epoch": 0.16552150611515734, + "grad_norm": 1.202965497970581, + "learning_rate": 4.934972984852613e-05, + "loss": 1.0449, + "step": 2409 + }, + { + "epoch": 0.1655902157482479, + "grad_norm": 1.1380828619003296, + "learning_rate": 4.93483704149301e-05, + "loss": 1.0385, + "step": 2410 + }, + { + "epoch": 0.16565892538133847, + "grad_norm": 1.193235993385315, + "learning_rate": 4.9347009580587165e-05, + "loss": 0.9942, + "step": 2411 + }, + { + "epoch": 0.16572763501442903, + "grad_norm": 1.1878880262374878, + "learning_rate": 4.934564734557562e-05, + "loss": 1.0202, + "step": 2412 + }, + { + "epoch": 0.16579634464751958, + "grad_norm": 1.2026424407958984, + "learning_rate": 4.934428370997382e-05, + "loss": 1.0995, + "step": 2413 + }, + { + "epoch": 0.16586505428061013, + "grad_norm": 1.1557793617248535, + "learning_rate": 4.9342918673860226e-05, + "loss": 1.0259, + "step": 2414 + }, + { + "epoch": 0.1659337639137007, + "grad_norm": 1.2235156297683716, + "learning_rate": 4.934155223731336e-05, + "loss": 1.0672, + "step": 2415 + }, + { + "epoch": 0.16600247354679126, + "grad_norm": 1.3160674571990967, + "learning_rate": 4.9340184400411836e-05, + "loss": 1.0377, + "step": 2416 + }, + { + "epoch": 0.16607118317988181, + "grad_norm": 1.3907958269119263, + "learning_rate": 4.933881516323435e-05, + "loss": 0.9764, + "step": 2417 + }, + { + "epoch": 0.16613989281297237, + "grad_norm": 1.2111878395080566, + "learning_rate": 4.933744452585966e-05, + "loss": 1.0318, + "step": 2418 + }, + { + "epoch": 0.16620860244606295, + "grad_norm": 1.2016631364822388, + "learning_rate": 4.933607248836664e-05, + "loss": 1.0566, + "step": 2419 + }, + { + "epoch": 0.1662773120791535, + "grad_norm": 1.169152855873108, + "learning_rate": 4.93346990508342e-05, + "loss": 1.0732, + "step": 2420 + }, + { + "epoch": 0.16634602171224405, + "grad_norm": 1.1430100202560425, + "learning_rate": 4.933332421334136e-05, + "loss": 1.0281, + "step": 2421 + }, + { + "epoch": 0.1664147313453346, + "grad_norm": 1.1076127290725708, + "learning_rate": 4.9331947975967215e-05, + "loss": 1.0564, + "step": 2422 + }, + { + "epoch": 0.16648344097842518, + "grad_norm": 1.2955708503723145, + "learning_rate": 4.9330570338790935e-05, + "loss": 1.072, + "step": 2423 + }, + { + "epoch": 0.16655215061151574, + "grad_norm": 1.154833197593689, + "learning_rate": 4.932919130189179e-05, + "loss": 1.0253, + "step": 2424 + }, + { + "epoch": 0.1666208602446063, + "grad_norm": 1.129295825958252, + "learning_rate": 4.9327810865349086e-05, + "loss": 0.9904, + "step": 2425 + }, + { + "epoch": 0.16668956987769684, + "grad_norm": 1.158825159072876, + "learning_rate": 4.932642902924227e-05, + "loss": 1.0265, + "step": 2426 + }, + { + "epoch": 0.16675827951078742, + "grad_norm": 1.2039096355438232, + "learning_rate": 4.9325045793650826e-05, + "loss": 1.0765, + "step": 2427 + }, + { + "epoch": 0.16682698914387797, + "grad_norm": 1.0715053081512451, + "learning_rate": 4.932366115865432e-05, + "loss": 1.012, + "step": 2428 + }, + { + "epoch": 0.16689569877696853, + "grad_norm": 1.1150124073028564, + "learning_rate": 4.932227512433242e-05, + "loss": 1.0095, + "step": 2429 + }, + { + "epoch": 0.16696440841005908, + "grad_norm": 1.219042181968689, + "learning_rate": 4.932088769076487e-05, + "loss": 1.0646, + "step": 2430 + }, + { + "epoch": 0.16703311804314966, + "grad_norm": 1.1471165418624878, + "learning_rate": 4.9319498858031476e-05, + "loss": 1.0107, + "step": 2431 + }, + { + "epoch": 0.1671018276762402, + "grad_norm": 1.1477656364440918, + "learning_rate": 4.931810862621214e-05, + "loss": 0.9919, + "step": 2432 + }, + { + "epoch": 0.16717053730933076, + "grad_norm": 1.0903956890106201, + "learning_rate": 4.931671699538685e-05, + "loss": 0.9394, + "step": 2433 + }, + { + "epoch": 0.16723924694242132, + "grad_norm": 1.2415586709976196, + "learning_rate": 4.931532396563564e-05, + "loss": 1.0879, + "step": 2434 + }, + { + "epoch": 0.1673079565755119, + "grad_norm": 1.180834174156189, + "learning_rate": 4.9313929537038684e-05, + "loss": 1.0286, + "step": 2435 + }, + { + "epoch": 0.16737666620860245, + "grad_norm": 1.182312250137329, + "learning_rate": 4.9312533709676177e-05, + "loss": 1.0077, + "step": 2436 + }, + { + "epoch": 0.167445375841693, + "grad_norm": 1.1164681911468506, + "learning_rate": 4.931113648362844e-05, + "loss": 1.058, + "step": 2437 + }, + { + "epoch": 0.16751408547478355, + "grad_norm": 1.2352262735366821, + "learning_rate": 4.9309737858975835e-05, + "loss": 1.0702, + "step": 2438 + }, + { + "epoch": 0.16758279510787413, + "grad_norm": 1.141087293624878, + "learning_rate": 4.930833783579884e-05, + "loss": 1.0244, + "step": 2439 + }, + { + "epoch": 0.16765150474096469, + "grad_norm": 1.2437409162521362, + "learning_rate": 4.930693641417798e-05, + "loss": 1.079, + "step": 2440 + }, + { + "epoch": 0.16772021437405524, + "grad_norm": 1.2158547639846802, + "learning_rate": 4.930553359419391e-05, + "loss": 1.0411, + "step": 2441 + }, + { + "epoch": 0.1677889240071458, + "grad_norm": 1.1137773990631104, + "learning_rate": 4.93041293759273e-05, + "loss": 1.0365, + "step": 2442 + }, + { + "epoch": 0.16785763364023637, + "grad_norm": 1.210998773574829, + "learning_rate": 4.9302723759458944e-05, + "loss": 1.0588, + "step": 2443 + }, + { + "epoch": 0.16792634327332692, + "grad_norm": 1.262130618095398, + "learning_rate": 4.930131674486971e-05, + "loss": 1.0396, + "step": 2444 + }, + { + "epoch": 0.16799505290641747, + "grad_norm": 1.2382268905639648, + "learning_rate": 4.929990833224054e-05, + "loss": 1.0362, + "step": 2445 + }, + { + "epoch": 0.16806376253950803, + "grad_norm": 1.2779937982559204, + "learning_rate": 4.929849852165246e-05, + "loss": 1.0777, + "step": 2446 + }, + { + "epoch": 0.1681324721725986, + "grad_norm": 1.1249923706054688, + "learning_rate": 4.9297087313186574e-05, + "loss": 1.0689, + "step": 2447 + }, + { + "epoch": 0.16820118180568916, + "grad_norm": 1.3164280652999878, + "learning_rate": 4.9295674706924074e-05, + "loss": 1.0874, + "step": 2448 + }, + { + "epoch": 0.1682698914387797, + "grad_norm": 1.156427264213562, + "learning_rate": 4.9294260702946216e-05, + "loss": 1.0431, + "step": 2449 + }, + { + "epoch": 0.16833860107187026, + "grad_norm": 1.1743465662002563, + "learning_rate": 4.9292845301334356e-05, + "loss": 1.0688, + "step": 2450 + }, + { + "epoch": 0.16840731070496084, + "grad_norm": 1.3264979124069214, + "learning_rate": 4.9291428502169924e-05, + "loss": 1.0233, + "step": 2451 + }, + { + "epoch": 0.1684760203380514, + "grad_norm": 1.2407602071762085, + "learning_rate": 4.929001030553442e-05, + "loss": 1.0374, + "step": 2452 + }, + { + "epoch": 0.16854472997114195, + "grad_norm": 1.1223618984222412, + "learning_rate": 4.9288590711509424e-05, + "loss": 1.022, + "step": 2453 + }, + { + "epoch": 0.1686134396042325, + "grad_norm": 1.1658709049224854, + "learning_rate": 4.928716972017662e-05, + "loss": 1.0218, + "step": 2454 + }, + { + "epoch": 0.16868214923732308, + "grad_norm": 1.281746745109558, + "learning_rate": 4.9285747331617746e-05, + "loss": 1.0372, + "step": 2455 + }, + { + "epoch": 0.16875085887041363, + "grad_norm": 1.2516580820083618, + "learning_rate": 4.928432354591465e-05, + "loss": 1.0358, + "step": 2456 + }, + { + "epoch": 0.16881956850350419, + "grad_norm": 1.2058254480361938, + "learning_rate": 4.928289836314921e-05, + "loss": 1.0736, + "step": 2457 + }, + { + "epoch": 0.16888827813659474, + "grad_norm": 1.238680362701416, + "learning_rate": 4.9281471783403446e-05, + "loss": 1.0282, + "step": 2458 + }, + { + "epoch": 0.16895698776968532, + "grad_norm": 1.1617881059646606, + "learning_rate": 4.928004380675941e-05, + "loss": 1.0122, + "step": 2459 + }, + { + "epoch": 0.16902569740277587, + "grad_norm": 1.1509184837341309, + "learning_rate": 4.927861443329926e-05, + "loss": 1.0487, + "step": 2460 + }, + { + "epoch": 0.16909440703586642, + "grad_norm": 1.152822732925415, + "learning_rate": 4.927718366310522e-05, + "loss": 1.0545, + "step": 2461 + }, + { + "epoch": 0.16916311666895698, + "grad_norm": 1.225714921951294, + "learning_rate": 4.9275751496259615e-05, + "loss": 1.0648, + "step": 2462 + }, + { + "epoch": 0.16923182630204756, + "grad_norm": 1.188731074333191, + "learning_rate": 4.927431793284482e-05, + "loss": 0.9759, + "step": 2463 + }, + { + "epoch": 0.1693005359351381, + "grad_norm": 1.1608762741088867, + "learning_rate": 4.9272882972943315e-05, + "loss": 1.0076, + "step": 2464 + }, + { + "epoch": 0.16936924556822866, + "grad_norm": 1.2058448791503906, + "learning_rate": 4.9271446616637664e-05, + "loss": 1.0049, + "step": 2465 + }, + { + "epoch": 0.1694379552013192, + "grad_norm": 1.167988896369934, + "learning_rate": 4.927000886401047e-05, + "loss": 1.0418, + "step": 2466 + }, + { + "epoch": 0.1695066648344098, + "grad_norm": 1.224555253982544, + "learning_rate": 4.926856971514448e-05, + "loss": 1.0604, + "step": 2467 + }, + { + "epoch": 0.16957537446750034, + "grad_norm": 1.210327386856079, + "learning_rate": 4.926712917012247e-05, + "loss": 1.0554, + "step": 2468 + }, + { + "epoch": 0.1696440841005909, + "grad_norm": 1.1680412292480469, + "learning_rate": 4.926568722902731e-05, + "loss": 1.0743, + "step": 2469 + }, + { + "epoch": 0.16971279373368145, + "grad_norm": 1.234989881515503, + "learning_rate": 4.926424389194196e-05, + "loss": 1.067, + "step": 2470 + }, + { + "epoch": 0.16978150336677203, + "grad_norm": 1.223240613937378, + "learning_rate": 4.9262799158949455e-05, + "loss": 1.0714, + "step": 2471 + }, + { + "epoch": 0.16985021299986258, + "grad_norm": 1.1628639698028564, + "learning_rate": 4.9261353030132906e-05, + "loss": 1.0285, + "step": 2472 + }, + { + "epoch": 0.16991892263295313, + "grad_norm": 1.2318525314331055, + "learning_rate": 4.925990550557551e-05, + "loss": 1.0271, + "step": 2473 + }, + { + "epoch": 0.1699876322660437, + "grad_norm": 1.1493088006973267, + "learning_rate": 4.9258456585360545e-05, + "loss": 1.0428, + "step": 2474 + }, + { + "epoch": 0.17005634189913427, + "grad_norm": 1.1843628883361816, + "learning_rate": 4.925700626957136e-05, + "loss": 1.0379, + "step": 2475 + }, + { + "epoch": 0.17012505153222482, + "grad_norm": 1.1758619546890259, + "learning_rate": 4.925555455829141e-05, + "loss": 1.0298, + "step": 2476 + }, + { + "epoch": 0.17019376116531537, + "grad_norm": 1.2262814044952393, + "learning_rate": 4.925410145160417e-05, + "loss": 1.0347, + "step": 2477 + }, + { + "epoch": 0.17026247079840592, + "grad_norm": 1.18234121799469, + "learning_rate": 4.9252646949593285e-05, + "loss": 1.0065, + "step": 2478 + }, + { + "epoch": 0.1703311804314965, + "grad_norm": 1.233423113822937, + "learning_rate": 4.92511910523424e-05, + "loss": 1.0093, + "step": 2479 + }, + { + "epoch": 0.17039989006458706, + "grad_norm": 1.1605323553085327, + "learning_rate": 4.924973375993528e-05, + "loss": 1.026, + "step": 2480 + }, + { + "epoch": 0.1704685996976776, + "grad_norm": 1.1920280456542969, + "learning_rate": 4.924827507245576e-05, + "loss": 1.0781, + "step": 2481 + }, + { + "epoch": 0.17053730933076816, + "grad_norm": 1.1938587427139282, + "learning_rate": 4.924681498998776e-05, + "loss": 1.0303, + "step": 2482 + }, + { + "epoch": 0.17060601896385874, + "grad_norm": 1.2513880729675293, + "learning_rate": 4.924535351261528e-05, + "loss": 1.0604, + "step": 2483 + }, + { + "epoch": 0.1706747285969493, + "grad_norm": 1.110047698020935, + "learning_rate": 4.92438906404224e-05, + "loss": 0.983, + "step": 2484 + }, + { + "epoch": 0.17074343823003985, + "grad_norm": 1.3020715713500977, + "learning_rate": 4.924242637349327e-05, + "loss": 1.0625, + "step": 2485 + }, + { + "epoch": 0.1708121478631304, + "grad_norm": 1.0953869819641113, + "learning_rate": 4.9240960711912126e-05, + "loss": 0.995, + "step": 2486 + }, + { + "epoch": 0.17088085749622098, + "grad_norm": 1.2379974126815796, + "learning_rate": 4.92394936557633e-05, + "loss": 0.985, + "step": 2487 + }, + { + "epoch": 0.17094956712931153, + "grad_norm": 1.2187168598175049, + "learning_rate": 4.923802520513118e-05, + "loss": 1.0138, + "step": 2488 + }, + { + "epoch": 0.17101827676240208, + "grad_norm": 1.184126853942871, + "learning_rate": 4.923655536010025e-05, + "loss": 1.0215, + "step": 2489 + }, + { + "epoch": 0.17108698639549264, + "grad_norm": 1.1892060041427612, + "learning_rate": 4.923508412075507e-05, + "loss": 1.0489, + "step": 2490 + }, + { + "epoch": 0.17115569602858322, + "grad_norm": 1.1624865531921387, + "learning_rate": 4.9233611487180275e-05, + "loss": 1.0447, + "step": 2491 + }, + { + "epoch": 0.17122440566167377, + "grad_norm": 1.2135241031646729, + "learning_rate": 4.923213745946059e-05, + "loss": 1.0484, + "step": 2492 + }, + { + "epoch": 0.17129311529476432, + "grad_norm": 1.201427936553955, + "learning_rate": 4.923066203768081e-05, + "loss": 1.0362, + "step": 2493 + }, + { + "epoch": 0.17136182492785487, + "grad_norm": 1.1488018035888672, + "learning_rate": 4.9229185221925816e-05, + "loss": 1.0478, + "step": 2494 + }, + { + "epoch": 0.17143053456094545, + "grad_norm": 1.1793416738510132, + "learning_rate": 4.922770701228057e-05, + "loss": 1.0163, + "step": 2495 + }, + { + "epoch": 0.171499244194036, + "grad_norm": 1.129070520401001, + "learning_rate": 4.9226227408830115e-05, + "loss": 1.076, + "step": 2496 + }, + { + "epoch": 0.17156795382712656, + "grad_norm": 1.1948144435882568, + "learning_rate": 4.922474641165956e-05, + "loss": 1.0173, + "step": 2497 + }, + { + "epoch": 0.1716366634602171, + "grad_norm": 1.1547240018844604, + "learning_rate": 4.922326402085413e-05, + "loss": 1.0441, + "step": 2498 + }, + { + "epoch": 0.1717053730933077, + "grad_norm": 1.1643320322036743, + "learning_rate": 4.9221780236499075e-05, + "loss": 1.0055, + "step": 2499 + }, + { + "epoch": 0.17177408272639824, + "grad_norm": 1.147326111793518, + "learning_rate": 4.922029505867978e-05, + "loss": 1.0303, + "step": 2500 + }, + { + "epoch": 0.1718427923594888, + "grad_norm": 1.158223032951355, + "learning_rate": 4.9218808487481684e-05, + "loss": 1.0395, + "step": 2501 + }, + { + "epoch": 0.17191150199257935, + "grad_norm": 1.2219926118850708, + "learning_rate": 4.921732052299029e-05, + "loss": 1.1303, + "step": 2502 + }, + { + "epoch": 0.17198021162566993, + "grad_norm": 1.1510004997253418, + "learning_rate": 4.921583116529122e-05, + "loss": 1.0356, + "step": 2503 + }, + { + "epoch": 0.17204892125876048, + "grad_norm": 1.1996551752090454, + "learning_rate": 4.9214340414470145e-05, + "loss": 1.0726, + "step": 2504 + }, + { + "epoch": 0.17211763089185103, + "grad_norm": 1.180259108543396, + "learning_rate": 4.921284827061283e-05, + "loss": 1.0789, + "step": 2505 + }, + { + "epoch": 0.17218634052494158, + "grad_norm": 1.1664396524429321, + "learning_rate": 4.9211354733805115e-05, + "loss": 1.0218, + "step": 2506 + }, + { + "epoch": 0.17225505015803216, + "grad_norm": 1.2411658763885498, + "learning_rate": 4.920985980413293e-05, + "loss": 1.0643, + "step": 2507 + }, + { + "epoch": 0.17232375979112272, + "grad_norm": 1.0913450717926025, + "learning_rate": 4.9208363481682274e-05, + "loss": 1.0221, + "step": 2508 + }, + { + "epoch": 0.17239246942421327, + "grad_norm": 1.269060492515564, + "learning_rate": 4.9206865766539226e-05, + "loss": 1.034, + "step": 2509 + }, + { + "epoch": 0.17246117905730382, + "grad_norm": 1.134227991104126, + "learning_rate": 4.920536665878994e-05, + "loss": 1.0695, + "step": 2510 + }, + { + "epoch": 0.1725298886903944, + "grad_norm": 1.1295498609542847, + "learning_rate": 4.920386615852068e-05, + "loss": 0.9937, + "step": 2511 + }, + { + "epoch": 0.17259859832348495, + "grad_norm": 1.1971131563186646, + "learning_rate": 4.920236426581776e-05, + "loss": 1.049, + "step": 2512 + }, + { + "epoch": 0.1726673079565755, + "grad_norm": 1.2320427894592285, + "learning_rate": 4.920086098076757e-05, + "loss": 1.0171, + "step": 2513 + }, + { + "epoch": 0.17273601758966606, + "grad_norm": 1.134353518486023, + "learning_rate": 4.919935630345661e-05, + "loss": 0.9779, + "step": 2514 + }, + { + "epoch": 0.17280472722275664, + "grad_norm": 1.2535995244979858, + "learning_rate": 4.919785023397144e-05, + "loss": 1.0914, + "step": 2515 + }, + { + "epoch": 0.1728734368558472, + "grad_norm": 1.1486432552337646, + "learning_rate": 4.9196342772398696e-05, + "loss": 1.0275, + "step": 2516 + }, + { + "epoch": 0.17294214648893774, + "grad_norm": 1.1415603160858154, + "learning_rate": 4.9194833918825104e-05, + "loss": 1.093, + "step": 2517 + }, + { + "epoch": 0.1730108561220283, + "grad_norm": 1.1935666799545288, + "learning_rate": 4.9193323673337476e-05, + "loss": 1.0214, + "step": 2518 + }, + { + "epoch": 0.17307956575511887, + "grad_norm": 1.089578628540039, + "learning_rate": 4.9191812036022677e-05, + "loss": 1.0042, + "step": 2519 + }, + { + "epoch": 0.17314827538820943, + "grad_norm": 1.1445072889328003, + "learning_rate": 4.919029900696769e-05, + "loss": 1.0568, + "step": 2520 + }, + { + "epoch": 0.17321698502129998, + "grad_norm": 1.173996925354004, + "learning_rate": 4.918878458625955e-05, + "loss": 1.0098, + "step": 2521 + }, + { + "epoch": 0.17328569465439053, + "grad_norm": 1.165274977684021, + "learning_rate": 4.918726877398538e-05, + "loss": 1.0552, + "step": 2522 + }, + { + "epoch": 0.1733544042874811, + "grad_norm": 1.2313742637634277, + "learning_rate": 4.918575157023239e-05, + "loss": 1.0407, + "step": 2523 + }, + { + "epoch": 0.17342311392057166, + "grad_norm": 1.1750680208206177, + "learning_rate": 4.918423297508786e-05, + "loss": 1.0643, + "step": 2524 + }, + { + "epoch": 0.17349182355366222, + "grad_norm": 1.1957526206970215, + "learning_rate": 4.918271298863915e-05, + "loss": 1.0327, + "step": 2525 + }, + { + "epoch": 0.17356053318675277, + "grad_norm": 1.0779818296432495, + "learning_rate": 4.91811916109737e-05, + "loss": 0.9935, + "step": 2526 + }, + { + "epoch": 0.17362924281984335, + "grad_norm": 1.0965957641601562, + "learning_rate": 4.917966884217905e-05, + "loss": 1.0474, + "step": 2527 + }, + { + "epoch": 0.1736979524529339, + "grad_norm": 1.1485090255737305, + "learning_rate": 4.917814468234279e-05, + "loss": 1.0062, + "step": 2528 + }, + { + "epoch": 0.17376666208602445, + "grad_norm": 1.1923962831497192, + "learning_rate": 4.9176619131552604e-05, + "loss": 1.0378, + "step": 2529 + }, + { + "epoch": 0.173835371719115, + "grad_norm": 1.1562131643295288, + "learning_rate": 4.917509218989626e-05, + "loss": 1.1027, + "step": 2530 + }, + { + "epoch": 0.1739040813522056, + "grad_norm": 1.1479378938674927, + "learning_rate": 4.9173563857461615e-05, + "loss": 1.0501, + "step": 2531 + }, + { + "epoch": 0.17397279098529614, + "grad_norm": 1.0935546159744263, + "learning_rate": 4.9172034134336575e-05, + "loss": 1.0137, + "step": 2532 + }, + { + "epoch": 0.1740415006183867, + "grad_norm": 1.1612389087677002, + "learning_rate": 4.9170503020609146e-05, + "loss": 1.0256, + "step": 2533 + }, + { + "epoch": 0.17411021025147727, + "grad_norm": 1.1546604633331299, + "learning_rate": 4.916897051636742e-05, + "loss": 1.0218, + "step": 2534 + }, + { + "epoch": 0.17417891988456782, + "grad_norm": 1.1445949077606201, + "learning_rate": 4.9167436621699555e-05, + "loss": 1.0798, + "step": 2535 + }, + { + "epoch": 0.17424762951765838, + "grad_norm": 1.1489583253860474, + "learning_rate": 4.91659013366938e-05, + "loss": 1.0424, + "step": 2536 + }, + { + "epoch": 0.17431633915074893, + "grad_norm": 1.1168129444122314, + "learning_rate": 4.9164364661438475e-05, + "loss": 1.0362, + "step": 2537 + }, + { + "epoch": 0.1743850487838395, + "grad_norm": 1.2008044719696045, + "learning_rate": 4.9162826596021985e-05, + "loss": 1.0317, + "step": 2538 + }, + { + "epoch": 0.17445375841693006, + "grad_norm": 1.2183582782745361, + "learning_rate": 4.9161287140532806e-05, + "loss": 1.026, + "step": 2539 + }, + { + "epoch": 0.1745224680500206, + "grad_norm": 1.1555736064910889, + "learning_rate": 4.915974629505952e-05, + "loss": 1.0085, + "step": 2540 + }, + { + "epoch": 0.17459117768311117, + "grad_norm": 1.1668058633804321, + "learning_rate": 4.9158204059690755e-05, + "loss": 0.99, + "step": 2541 + }, + { + "epoch": 0.17465988731620175, + "grad_norm": 1.2594177722930908, + "learning_rate": 4.915666043451524e-05, + "loss": 1.0558, + "step": 2542 + }, + { + "epoch": 0.1747285969492923, + "grad_norm": 1.192557692527771, + "learning_rate": 4.9155115419621776e-05, + "loss": 1.0148, + "step": 2543 + }, + { + "epoch": 0.17479730658238285, + "grad_norm": 1.1260164976119995, + "learning_rate": 4.915356901509926e-05, + "loss": 1.0377, + "step": 2544 + }, + { + "epoch": 0.1748660162154734, + "grad_norm": 1.1604959964752197, + "learning_rate": 4.915202122103664e-05, + "loss": 1.0336, + "step": 2545 + }, + { + "epoch": 0.17493472584856398, + "grad_norm": 1.260061264038086, + "learning_rate": 4.915047203752296e-05, + "loss": 1.0441, + "step": 2546 + }, + { + "epoch": 0.17500343548165453, + "grad_norm": 1.0943950414657593, + "learning_rate": 4.914892146464736e-05, + "loss": 1.0133, + "step": 2547 + }, + { + "epoch": 0.1750721451147451, + "grad_norm": 1.1829614639282227, + "learning_rate": 4.914736950249903e-05, + "loss": 0.9936, + "step": 2548 + }, + { + "epoch": 0.17514085474783564, + "grad_norm": 1.1259663105010986, + "learning_rate": 4.9145816151167256e-05, + "loss": 0.9934, + "step": 2549 + }, + { + "epoch": 0.17520956438092622, + "grad_norm": 1.2301037311553955, + "learning_rate": 4.91442614107414e-05, + "loss": 1.0277, + "step": 2550 + }, + { + "epoch": 0.17527827401401677, + "grad_norm": 1.2051640748977661, + "learning_rate": 4.9142705281310904e-05, + "loss": 1.0558, + "step": 2551 + }, + { + "epoch": 0.17534698364710732, + "grad_norm": 1.1832793951034546, + "learning_rate": 4.91411477629653e-05, + "loss": 1.0208, + "step": 2552 + }, + { + "epoch": 0.17541569328019788, + "grad_norm": 1.1527769565582275, + "learning_rate": 4.9139588855794175e-05, + "loss": 1.0247, + "step": 2553 + }, + { + "epoch": 0.17548440291328846, + "grad_norm": 1.1508708000183105, + "learning_rate": 4.9138028559887226e-05, + "loss": 1.036, + "step": 2554 + }, + { + "epoch": 0.175553112546379, + "grad_norm": 1.2155877351760864, + "learning_rate": 4.9136466875334215e-05, + "loss": 1.0547, + "step": 2555 + }, + { + "epoch": 0.17562182217946956, + "grad_norm": 1.2257353067398071, + "learning_rate": 4.9134903802224977e-05, + "loss": 1.0513, + "step": 2556 + }, + { + "epoch": 0.1756905318125601, + "grad_norm": 1.128828763961792, + "learning_rate": 4.9133339340649444e-05, + "loss": 1.0066, + "step": 2557 + }, + { + "epoch": 0.1757592414456507, + "grad_norm": 1.1581653356552124, + "learning_rate": 4.913177349069761e-05, + "loss": 1.0003, + "step": 2558 + }, + { + "epoch": 0.17582795107874125, + "grad_norm": 1.1616777181625366, + "learning_rate": 4.913020625245957e-05, + "loss": 1.0363, + "step": 2559 + }, + { + "epoch": 0.1758966607118318, + "grad_norm": 1.2358766794204712, + "learning_rate": 4.912863762602547e-05, + "loss": 1.0511, + "step": 2560 + }, + { + "epoch": 0.17596537034492235, + "grad_norm": 1.2049992084503174, + "learning_rate": 4.9127067611485555e-05, + "loss": 1.0301, + "step": 2561 + }, + { + "epoch": 0.17603407997801293, + "grad_norm": 1.186498999595642, + "learning_rate": 4.9125496208930155e-05, + "loss": 1.0591, + "step": 2562 + }, + { + "epoch": 0.17610278961110348, + "grad_norm": 1.2115790843963623, + "learning_rate": 4.912392341844967e-05, + "loss": 1.07, + "step": 2563 + }, + { + "epoch": 0.17617149924419404, + "grad_norm": 1.1552784442901611, + "learning_rate": 4.9122349240134575e-05, + "loss": 1.0177, + "step": 2564 + }, + { + "epoch": 0.1762402088772846, + "grad_norm": 1.127769112586975, + "learning_rate": 4.912077367407544e-05, + "loss": 1.0583, + "step": 2565 + }, + { + "epoch": 0.17630891851037517, + "grad_norm": 1.173713207244873, + "learning_rate": 4.91191967203629e-05, + "loss": 1.0401, + "step": 2566 + }, + { + "epoch": 0.17637762814346572, + "grad_norm": 1.1179141998291016, + "learning_rate": 4.911761837908768e-05, + "loss": 1.0395, + "step": 2567 + }, + { + "epoch": 0.17644633777655627, + "grad_norm": 1.1389559507369995, + "learning_rate": 4.9116038650340587e-05, + "loss": 1.0098, + "step": 2568 + }, + { + "epoch": 0.17651504740964682, + "grad_norm": 1.1085821390151978, + "learning_rate": 4.911445753421249e-05, + "loss": 1.0158, + "step": 2569 + }, + { + "epoch": 0.1765837570427374, + "grad_norm": 1.149715542793274, + "learning_rate": 4.911287503079435e-05, + "loss": 1.0625, + "step": 2570 + }, + { + "epoch": 0.17665246667582796, + "grad_norm": 1.1724092960357666, + "learning_rate": 4.911129114017722e-05, + "loss": 1.0351, + "step": 2571 + }, + { + "epoch": 0.1767211763089185, + "grad_norm": 1.1968817710876465, + "learning_rate": 4.91097058624522e-05, + "loss": 1.0517, + "step": 2572 + }, + { + "epoch": 0.17678988594200906, + "grad_norm": 1.134566068649292, + "learning_rate": 4.910811919771051e-05, + "loss": 1.0373, + "step": 2573 + }, + { + "epoch": 0.17685859557509964, + "grad_norm": 1.1678001880645752, + "learning_rate": 4.910653114604342e-05, + "loss": 1.0294, + "step": 2574 + }, + { + "epoch": 0.1769273052081902, + "grad_norm": 1.1686134338378906, + "learning_rate": 4.91049417075423e-05, + "loss": 1.0418, + "step": 2575 + }, + { + "epoch": 0.17699601484128075, + "grad_norm": 1.147993564605713, + "learning_rate": 4.910335088229857e-05, + "loss": 1.0535, + "step": 2576 + }, + { + "epoch": 0.1770647244743713, + "grad_norm": 1.1467921733856201, + "learning_rate": 4.910175867040377e-05, + "loss": 0.986, + "step": 2577 + }, + { + "epoch": 0.17713343410746188, + "grad_norm": 1.2073546648025513, + "learning_rate": 4.9100165071949485e-05, + "loss": 1.0247, + "step": 2578 + }, + { + "epoch": 0.17720214374055243, + "grad_norm": 1.167632818222046, + "learning_rate": 4.909857008702739e-05, + "loss": 1.0635, + "step": 2579 + }, + { + "epoch": 0.17727085337364298, + "grad_norm": 1.134186029434204, + "learning_rate": 4.909697371572927e-05, + "loss": 1.0257, + "step": 2580 + }, + { + "epoch": 0.17733956300673354, + "grad_norm": 1.1560479402542114, + "learning_rate": 4.909537595814693e-05, + "loss": 0.9869, + "step": 2581 + }, + { + "epoch": 0.17740827263982412, + "grad_norm": 1.1629999876022339, + "learning_rate": 4.909377681437231e-05, + "loss": 1.0474, + "step": 2582 + }, + { + "epoch": 0.17747698227291467, + "grad_norm": 1.2074073553085327, + "learning_rate": 4.909217628449739e-05, + "loss": 1.0672, + "step": 2583 + }, + { + "epoch": 0.17754569190600522, + "grad_norm": 1.199908971786499, + "learning_rate": 4.9090574368614274e-05, + "loss": 1.0684, + "step": 2584 + }, + { + "epoch": 0.17761440153909577, + "grad_norm": 1.198516845703125, + "learning_rate": 4.9088971066815095e-05, + "loss": 1.0121, + "step": 2585 + }, + { + "epoch": 0.17768311117218635, + "grad_norm": 1.1474460363388062, + "learning_rate": 4.9087366379192094e-05, + "loss": 1.0263, + "step": 2586 + }, + { + "epoch": 0.1777518208052769, + "grad_norm": 1.1441298723220825, + "learning_rate": 4.90857603058376e-05, + "loss": 1.0589, + "step": 2587 + }, + { + "epoch": 0.17782053043836746, + "grad_norm": 1.1410282850265503, + "learning_rate": 4.908415284684399e-05, + "loss": 1.0113, + "step": 2588 + }, + { + "epoch": 0.177889240071458, + "grad_norm": 1.1289117336273193, + "learning_rate": 4.9082544002303764e-05, + "loss": 0.9757, + "step": 2589 + }, + { + "epoch": 0.1779579497045486, + "grad_norm": 1.1493924856185913, + "learning_rate": 4.908093377230946e-05, + "loss": 1.0026, + "step": 2590 + }, + { + "epoch": 0.17802665933763914, + "grad_norm": 1.1671161651611328, + "learning_rate": 4.9079322156953724e-05, + "loss": 1.0161, + "step": 2591 + }, + { + "epoch": 0.1780953689707297, + "grad_norm": 1.249498963356018, + "learning_rate": 4.907770915632926e-05, + "loss": 1.0353, + "step": 2592 + }, + { + "epoch": 0.17816407860382025, + "grad_norm": 1.1041359901428223, + "learning_rate": 4.907609477052887e-05, + "loss": 0.9913, + "step": 2593 + }, + { + "epoch": 0.17823278823691083, + "grad_norm": 1.165157675743103, + "learning_rate": 4.9074478999645434e-05, + "loss": 1.0612, + "step": 2594 + }, + { + "epoch": 0.17830149787000138, + "grad_norm": 1.1946908235549927, + "learning_rate": 4.907286184377189e-05, + "loss": 1.0112, + "step": 2595 + }, + { + "epoch": 0.17837020750309193, + "grad_norm": 1.1254702806472778, + "learning_rate": 4.907124330300129e-05, + "loss": 0.9713, + "step": 2596 + }, + { + "epoch": 0.17843891713618248, + "grad_norm": 1.182671308517456, + "learning_rate": 4.906962337742674e-05, + "loss": 1.0218, + "step": 2597 + }, + { + "epoch": 0.17850762676927306, + "grad_norm": 1.2198752164840698, + "learning_rate": 4.9068002067141426e-05, + "loss": 1.0355, + "step": 2598 + }, + { + "epoch": 0.17857633640236362, + "grad_norm": 1.1340889930725098, + "learning_rate": 4.9066379372238635e-05, + "loss": 1.037, + "step": 2599 + }, + { + "epoch": 0.17864504603545417, + "grad_norm": 1.172663688659668, + "learning_rate": 4.9064755292811716e-05, + "loss": 1.0267, + "step": 2600 + }, + { + "epoch": 0.17871375566854472, + "grad_norm": 1.2031970024108887, + "learning_rate": 4.9063129828954093e-05, + "loss": 1.0419, + "step": 2601 + }, + { + "epoch": 0.1787824653016353, + "grad_norm": 1.2651196718215942, + "learning_rate": 4.906150298075928e-05, + "loss": 1.0395, + "step": 2602 + }, + { + "epoch": 0.17885117493472585, + "grad_norm": 1.153951644897461, + "learning_rate": 4.9059874748320876e-05, + "loss": 1.038, + "step": 2603 + }, + { + "epoch": 0.1789198845678164, + "grad_norm": 1.171531081199646, + "learning_rate": 4.905824513173254e-05, + "loss": 1.039, + "step": 2604 + }, + { + "epoch": 0.17898859420090696, + "grad_norm": 1.1353882551193237, + "learning_rate": 4.9056614131088044e-05, + "loss": 1.0237, + "step": 2605 + }, + { + "epoch": 0.17905730383399754, + "grad_norm": 1.10830819606781, + "learning_rate": 4.90549817464812e-05, + "loss": 0.9965, + "step": 2606 + }, + { + "epoch": 0.1791260134670881, + "grad_norm": 1.1931005716323853, + "learning_rate": 4.9053347978005913e-05, + "loss": 1.0435, + "step": 2607 + }, + { + "epoch": 0.17919472310017864, + "grad_norm": 1.1498774290084839, + "learning_rate": 4.9051712825756194e-05, + "loss": 1.019, + "step": 2608 + }, + { + "epoch": 0.1792634327332692, + "grad_norm": 1.1863746643066406, + "learning_rate": 4.90500762898261e-05, + "loss": 1.0512, + "step": 2609 + }, + { + "epoch": 0.17933214236635978, + "grad_norm": 1.1282869577407837, + "learning_rate": 4.904843837030978e-05, + "loss": 1.025, + "step": 2610 + }, + { + "epoch": 0.17940085199945033, + "grad_norm": 1.11941659450531, + "learning_rate": 4.904679906730146e-05, + "loss": 1.0051, + "step": 2611 + }, + { + "epoch": 0.17946956163254088, + "grad_norm": 1.177125334739685, + "learning_rate": 4.904515838089546e-05, + "loss": 1.0427, + "step": 2612 + }, + { + "epoch": 0.17953827126563143, + "grad_norm": 1.2704150676727295, + "learning_rate": 4.904351631118615e-05, + "loss": 1.0651, + "step": 2613 + }, + { + "epoch": 0.179606980898722, + "grad_norm": 1.1897532939910889, + "learning_rate": 4.904187285826801e-05, + "loss": 1.0779, + "step": 2614 + }, + { + "epoch": 0.17967569053181257, + "grad_norm": 1.212453007698059, + "learning_rate": 4.904022802223559e-05, + "loss": 1.0318, + "step": 2615 + }, + { + "epoch": 0.17974440016490312, + "grad_norm": 1.180811882019043, + "learning_rate": 4.9038581803183505e-05, + "loss": 1.009, + "step": 2616 + }, + { + "epoch": 0.17981310979799367, + "grad_norm": 1.1375998258590698, + "learning_rate": 4.903693420120646e-05, + "loss": 0.9897, + "step": 2617 + }, + { + "epoch": 0.17988181943108425, + "grad_norm": 1.1110987663269043, + "learning_rate": 4.903528521639926e-05, + "loss": 1.0307, + "step": 2618 + }, + { + "epoch": 0.1799505290641748, + "grad_norm": 1.0765316486358643, + "learning_rate": 4.903363484885675e-05, + "loss": 1.0159, + "step": 2619 + }, + { + "epoch": 0.18001923869726535, + "grad_norm": 1.2153493165969849, + "learning_rate": 4.903198309867387e-05, + "loss": 1.0514, + "step": 2620 + }, + { + "epoch": 0.1800879483303559, + "grad_norm": 1.2321637868881226, + "learning_rate": 4.9030329965945677e-05, + "loss": 1.0907, + "step": 2621 + }, + { + "epoch": 0.1801566579634465, + "grad_norm": 1.1270240545272827, + "learning_rate": 4.9028675450767237e-05, + "loss": 1.0265, + "step": 2622 + }, + { + "epoch": 0.18022536759653704, + "grad_norm": 1.149049997329712, + "learning_rate": 4.902701955323376e-05, + "loss": 1.0155, + "step": 2623 + }, + { + "epoch": 0.1802940772296276, + "grad_norm": 1.2440898418426514, + "learning_rate": 4.9025362273440487e-05, + "loss": 1.0399, + "step": 2624 + }, + { + "epoch": 0.18036278686271814, + "grad_norm": 1.123850703239441, + "learning_rate": 4.902370361148277e-05, + "loss": 1.0387, + "step": 2625 + }, + { + "epoch": 0.18043149649580872, + "grad_norm": 1.1072287559509277, + "learning_rate": 4.9022043567456045e-05, + "loss": 1.0298, + "step": 2626 + }, + { + "epoch": 0.18050020612889928, + "grad_norm": 1.198527216911316, + "learning_rate": 4.9020382141455796e-05, + "loss": 1.0351, + "step": 2627 + }, + { + "epoch": 0.18056891576198983, + "grad_norm": 1.203123927116394, + "learning_rate": 4.9018719333577605e-05, + "loss": 1.0638, + "step": 2628 + }, + { + "epoch": 0.18063762539508038, + "grad_norm": 1.26703679561615, + "learning_rate": 4.901705514391714e-05, + "loss": 1.0453, + "step": 2629 + }, + { + "epoch": 0.18070633502817096, + "grad_norm": 1.1025997400283813, + "learning_rate": 4.901538957257014e-05, + "loss": 1.068, + "step": 2630 + }, + { + "epoch": 0.1807750446612615, + "grad_norm": 1.159144401550293, + "learning_rate": 4.9013722619632405e-05, + "loss": 1.032, + "step": 2631 + }, + { + "epoch": 0.18084375429435207, + "grad_norm": 1.196587085723877, + "learning_rate": 4.901205428519987e-05, + "loss": 1.0762, + "step": 2632 + }, + { + "epoch": 0.18091246392744262, + "grad_norm": 1.135947585105896, + "learning_rate": 4.9010384569368477e-05, + "loss": 1.0196, + "step": 2633 + }, + { + "epoch": 0.1809811735605332, + "grad_norm": 1.1677244901657104, + "learning_rate": 4.9008713472234305e-05, + "loss": 1.0161, + "step": 2634 + }, + { + "epoch": 0.18104988319362375, + "grad_norm": 1.1928921937942505, + "learning_rate": 4.900704099389349e-05, + "loss": 1.0472, + "step": 2635 + }, + { + "epoch": 0.1811185928267143, + "grad_norm": 1.2328462600708008, + "learning_rate": 4.9005367134442235e-05, + "loss": 1.0265, + "step": 2636 + }, + { + "epoch": 0.18118730245980486, + "grad_norm": 1.1292465925216675, + "learning_rate": 4.900369189397686e-05, + "loss": 1.0228, + "step": 2637 + }, + { + "epoch": 0.18125601209289544, + "grad_norm": 1.1940200328826904, + "learning_rate": 4.900201527259371e-05, + "loss": 1.0207, + "step": 2638 + }, + { + "epoch": 0.181324721725986, + "grad_norm": 1.2172197103500366, + "learning_rate": 4.900033727038927e-05, + "loss": 1.0503, + "step": 2639 + }, + { + "epoch": 0.18139343135907654, + "grad_norm": 1.2315748929977417, + "learning_rate": 4.899865788746005e-05, + "loss": 1.0567, + "step": 2640 + }, + { + "epoch": 0.1814621409921671, + "grad_norm": 1.2242588996887207, + "learning_rate": 4.899697712390268e-05, + "loss": 1.0707, + "step": 2641 + }, + { + "epoch": 0.18153085062525767, + "grad_norm": 1.1356219053268433, + "learning_rate": 4.899529497981385e-05, + "loss": 1.0172, + "step": 2642 + }, + { + "epoch": 0.18159956025834822, + "grad_norm": 1.1502424478530884, + "learning_rate": 4.899361145529032e-05, + "loss": 1.0464, + "step": 2643 + }, + { + "epoch": 0.18166826989143878, + "grad_norm": 1.1078358888626099, + "learning_rate": 4.899192655042897e-05, + "loss": 1.0206, + "step": 2644 + }, + { + "epoch": 0.18173697952452933, + "grad_norm": 1.108736515045166, + "learning_rate": 4.89902402653267e-05, + "loss": 1.0174, + "step": 2645 + }, + { + "epoch": 0.1818056891576199, + "grad_norm": 1.151834487915039, + "learning_rate": 4.898855260008054e-05, + "loss": 1.0164, + "step": 2646 + }, + { + "epoch": 0.18187439879071046, + "grad_norm": 1.2008581161499023, + "learning_rate": 4.8986863554787575e-05, + "loss": 1.0477, + "step": 2647 + }, + { + "epoch": 0.18194310842380101, + "grad_norm": 1.1987035274505615, + "learning_rate": 4.8985173129544974e-05, + "loss": 1.0603, + "step": 2648 + }, + { + "epoch": 0.18201181805689157, + "grad_norm": 1.191982626914978, + "learning_rate": 4.8983481324449985e-05, + "loss": 1.0576, + "step": 2649 + }, + { + "epoch": 0.18208052768998215, + "grad_norm": 1.121823787689209, + "learning_rate": 4.898178813959995e-05, + "loss": 1.0269, + "step": 2650 + }, + { + "epoch": 0.1821492373230727, + "grad_norm": 1.0701229572296143, + "learning_rate": 4.898009357509226e-05, + "loss": 1.0399, + "step": 2651 + }, + { + "epoch": 0.18221794695616325, + "grad_norm": 1.1147541999816895, + "learning_rate": 4.8978397631024404e-05, + "loss": 1.0075, + "step": 2652 + }, + { + "epoch": 0.1822866565892538, + "grad_norm": 1.1782760620117188, + "learning_rate": 4.8976700307493954e-05, + "loss": 1.0499, + "step": 2653 + }, + { + "epoch": 0.18235536622234438, + "grad_norm": 1.1500838994979858, + "learning_rate": 4.897500160459856e-05, + "loss": 1.014, + "step": 2654 + }, + { + "epoch": 0.18242407585543494, + "grad_norm": 1.1602284908294678, + "learning_rate": 4.897330152243593e-05, + "loss": 1.0367, + "step": 2655 + }, + { + "epoch": 0.1824927854885255, + "grad_norm": 1.2321763038635254, + "learning_rate": 4.89716000611039e-05, + "loss": 1.0671, + "step": 2656 + }, + { + "epoch": 0.18256149512161604, + "grad_norm": 1.2622344493865967, + "learning_rate": 4.8969897220700315e-05, + "loss": 1.0709, + "step": 2657 + }, + { + "epoch": 0.18263020475470662, + "grad_norm": 1.223636507987976, + "learning_rate": 4.896819300132317e-05, + "loss": 1.0779, + "step": 2658 + }, + { + "epoch": 0.18269891438779717, + "grad_norm": 1.1982489824295044, + "learning_rate": 4.896648740307049e-05, + "loss": 1.0553, + "step": 2659 + }, + { + "epoch": 0.18276762402088773, + "grad_norm": 1.1243752241134644, + "learning_rate": 4.896478042604041e-05, + "loss": 1.0439, + "step": 2660 + }, + { + "epoch": 0.18283633365397828, + "grad_norm": 1.2952488660812378, + "learning_rate": 4.896307207033112e-05, + "loss": 1.0167, + "step": 2661 + }, + { + "epoch": 0.18290504328706886, + "grad_norm": 1.1374566555023193, + "learning_rate": 4.89613623360409e-05, + "loss": 1.0697, + "step": 2662 + }, + { + "epoch": 0.1829737529201594, + "grad_norm": 1.2730886936187744, + "learning_rate": 4.8959651223268125e-05, + "loss": 1.0625, + "step": 2663 + }, + { + "epoch": 0.18304246255324996, + "grad_norm": 1.2338356971740723, + "learning_rate": 4.8957938732111215e-05, + "loss": 1.0344, + "step": 2664 + }, + { + "epoch": 0.18311117218634052, + "grad_norm": 1.2149360179901123, + "learning_rate": 4.89562248626687e-05, + "loss": 1.0417, + "step": 2665 + }, + { + "epoch": 0.1831798818194311, + "grad_norm": 1.1822634935379028, + "learning_rate": 4.895450961503918e-05, + "loss": 1.0492, + "step": 2666 + }, + { + "epoch": 0.18324859145252165, + "grad_norm": 1.135701298713684, + "learning_rate": 4.895279298932132e-05, + "loss": 1.0167, + "step": 2667 + }, + { + "epoch": 0.1833173010856122, + "grad_norm": 1.1557788848876953, + "learning_rate": 4.895107498561389e-05, + "loss": 1.0385, + "step": 2668 + }, + { + "epoch": 0.18338601071870275, + "grad_norm": 1.1946042776107788, + "learning_rate": 4.894935560401572e-05, + "loss": 1.0431, + "step": 2669 + }, + { + "epoch": 0.18345472035179333, + "grad_norm": 1.2030574083328247, + "learning_rate": 4.8947634844625725e-05, + "loss": 1.0687, + "step": 2670 + }, + { + "epoch": 0.18352342998488388, + "grad_norm": 1.3229066133499146, + "learning_rate": 4.89459127075429e-05, + "loss": 1.0223, + "step": 2671 + }, + { + "epoch": 0.18359213961797444, + "grad_norm": 1.1561930179595947, + "learning_rate": 4.894418919286631e-05, + "loss": 1.0594, + "step": 2672 + }, + { + "epoch": 0.183660849251065, + "grad_norm": 1.0933449268341064, + "learning_rate": 4.894246430069512e-05, + "loss": 1.0027, + "step": 2673 + }, + { + "epoch": 0.18372955888415557, + "grad_norm": 1.1393638849258423, + "learning_rate": 4.894073803112855e-05, + "loss": 1.0016, + "step": 2674 + }, + { + "epoch": 0.18379826851724612, + "grad_norm": 1.2067824602127075, + "learning_rate": 4.893901038426593e-05, + "loss": 1.0418, + "step": 2675 + }, + { + "epoch": 0.18386697815033667, + "grad_norm": 1.1197309494018555, + "learning_rate": 4.8937281360206634e-05, + "loss": 1.0417, + "step": 2676 + }, + { + "epoch": 0.18393568778342723, + "grad_norm": 1.1192930936813354, + "learning_rate": 4.893555095905014e-05, + "loss": 1.0021, + "step": 2677 + }, + { + "epoch": 0.1840043974165178, + "grad_norm": 1.2642565965652466, + "learning_rate": 4.893381918089599e-05, + "loss": 1.0347, + "step": 2678 + }, + { + "epoch": 0.18407310704960836, + "grad_norm": 1.2680344581604004, + "learning_rate": 4.8932086025843816e-05, + "loss": 1.0645, + "step": 2679 + }, + { + "epoch": 0.1841418166826989, + "grad_norm": 1.0623328685760498, + "learning_rate": 4.893035149399333e-05, + "loss": 0.9681, + "step": 2680 + }, + { + "epoch": 0.18421052631578946, + "grad_norm": 1.1694368124008179, + "learning_rate": 4.89286155854443e-05, + "loss": 0.9959, + "step": 2681 + }, + { + "epoch": 0.18427923594888004, + "grad_norm": 1.2637172937393188, + "learning_rate": 4.892687830029662e-05, + "loss": 1.0417, + "step": 2682 + }, + { + "epoch": 0.1843479455819706, + "grad_norm": 1.1817055940628052, + "learning_rate": 4.892513963865022e-05, + "loss": 1.0255, + "step": 2683 + }, + { + "epoch": 0.18441665521506115, + "grad_norm": 1.1548393964767456, + "learning_rate": 4.892339960060512e-05, + "loss": 1.0245, + "step": 2684 + }, + { + "epoch": 0.1844853648481517, + "grad_norm": 1.2325905561447144, + "learning_rate": 4.892165818626143e-05, + "loss": 1.0584, + "step": 2685 + }, + { + "epoch": 0.18455407448124228, + "grad_norm": 1.19758939743042, + "learning_rate": 4.8919915395719326e-05, + "loss": 1.0556, + "step": 2686 + }, + { + "epoch": 0.18462278411433283, + "grad_norm": 1.2257832288742065, + "learning_rate": 4.891817122907908e-05, + "loss": 1.0354, + "step": 2687 + }, + { + "epoch": 0.18469149374742339, + "grad_norm": 1.0986528396606445, + "learning_rate": 4.8916425686441016e-05, + "loss": 1.0218, + "step": 2688 + }, + { + "epoch": 0.18476020338051394, + "grad_norm": 1.121129035949707, + "learning_rate": 4.8914678767905576e-05, + "loss": 1.0405, + "step": 2689 + }, + { + "epoch": 0.18482891301360452, + "grad_norm": 1.1327835321426392, + "learning_rate": 4.891293047357324e-05, + "loss": 0.9914, + "step": 2690 + }, + { + "epoch": 0.18489762264669507, + "grad_norm": 1.2078337669372559, + "learning_rate": 4.89111808035446e-05, + "loss": 1.0551, + "step": 2691 + }, + { + "epoch": 0.18496633227978562, + "grad_norm": 1.1123764514923096, + "learning_rate": 4.890942975792031e-05, + "loss": 1.0358, + "step": 2692 + }, + { + "epoch": 0.18503504191287617, + "grad_norm": 1.1208966970443726, + "learning_rate": 4.89076773368011e-05, + "loss": 1.064, + "step": 2693 + }, + { + "epoch": 0.18510375154596675, + "grad_norm": 1.1944881677627563, + "learning_rate": 4.890592354028779e-05, + "loss": 1.0476, + "step": 2694 + }, + { + "epoch": 0.1851724611790573, + "grad_norm": 1.0824190378189087, + "learning_rate": 4.890416836848127e-05, + "loss": 1.0188, + "step": 2695 + }, + { + "epoch": 0.18524117081214786, + "grad_norm": 1.0908145904541016, + "learning_rate": 4.890241182148253e-05, + "loss": 1.0658, + "step": 2696 + }, + { + "epoch": 0.1853098804452384, + "grad_norm": 1.2722406387329102, + "learning_rate": 4.890065389939261e-05, + "loss": 1.0796, + "step": 2697 + }, + { + "epoch": 0.185378590078329, + "grad_norm": 1.1867938041687012, + "learning_rate": 4.889889460231264e-05, + "loss": 1.0402, + "step": 2698 + }, + { + "epoch": 0.18544729971141954, + "grad_norm": 1.1408790349960327, + "learning_rate": 4.8897133930343835e-05, + "loss": 0.9625, + "step": 2699 + }, + { + "epoch": 0.1855160093445101, + "grad_norm": 1.165608525276184, + "learning_rate": 4.889537188358749e-05, + "loss": 1.0004, + "step": 2700 + }, + { + "epoch": 0.18558471897760065, + "grad_norm": 1.1412962675094604, + "learning_rate": 4.889360846214497e-05, + "loss": 1.0212, + "step": 2701 + }, + { + "epoch": 0.18565342861069123, + "grad_norm": 1.161191463470459, + "learning_rate": 4.889184366611772e-05, + "loss": 1.0485, + "step": 2702 + }, + { + "epoch": 0.18572213824378178, + "grad_norm": 1.2153743505477905, + "learning_rate": 4.889007749560727e-05, + "loss": 1.0643, + "step": 2703 + }, + { + "epoch": 0.18579084787687233, + "grad_norm": 1.1028077602386475, + "learning_rate": 4.888830995071523e-05, + "loss": 1.0, + "step": 2704 + }, + { + "epoch": 0.18585955750996289, + "grad_norm": 1.2296732664108276, + "learning_rate": 4.888654103154329e-05, + "loss": 1.0579, + "step": 2705 + }, + { + "epoch": 0.18592826714305347, + "grad_norm": 1.160088300704956, + "learning_rate": 4.88847707381932e-05, + "loss": 1.0309, + "step": 2706 + }, + { + "epoch": 0.18599697677614402, + "grad_norm": 1.2124089002609253, + "learning_rate": 4.888299907076682e-05, + "loss": 1.0406, + "step": 2707 + }, + { + "epoch": 0.18606568640923457, + "grad_norm": 1.1093512773513794, + "learning_rate": 4.888122602936606e-05, + "loss": 0.9973, + "step": 2708 + }, + { + "epoch": 0.18613439604232512, + "grad_norm": 1.1842323541641235, + "learning_rate": 4.887945161409293e-05, + "loss": 1.066, + "step": 2709 + }, + { + "epoch": 0.1862031056754157, + "grad_norm": 1.131941795349121, + "learning_rate": 4.887767582504951e-05, + "loss": 1.0512, + "step": 2710 + }, + { + "epoch": 0.18627181530850626, + "grad_norm": 1.1598401069641113, + "learning_rate": 4.887589866233796e-05, + "loss": 1.0332, + "step": 2711 + }, + { + "epoch": 0.1863405249415968, + "grad_norm": 1.1502795219421387, + "learning_rate": 4.8874120126060513e-05, + "loss": 1.0367, + "step": 2712 + }, + { + "epoch": 0.18640923457468736, + "grad_norm": 1.1472777128219604, + "learning_rate": 4.8872340216319495e-05, + "loss": 1.0389, + "step": 2713 + }, + { + "epoch": 0.18647794420777794, + "grad_norm": 1.1574219465255737, + "learning_rate": 4.88705589332173e-05, + "loss": 0.9855, + "step": 2714 + }, + { + "epoch": 0.1865466538408685, + "grad_norm": 1.0340094566345215, + "learning_rate": 4.88687762768564e-05, + "loss": 0.9939, + "step": 2715 + }, + { + "epoch": 0.18661536347395905, + "grad_norm": 1.1482551097869873, + "learning_rate": 4.886699224733936e-05, + "loss": 1.0528, + "step": 2716 + }, + { + "epoch": 0.1866840731070496, + "grad_norm": 1.1326172351837158, + "learning_rate": 4.8865206844768805e-05, + "loss": 1.009, + "step": 2717 + }, + { + "epoch": 0.18675278274014018, + "grad_norm": 1.1286405324935913, + "learning_rate": 4.8863420069247456e-05, + "loss": 1.0369, + "step": 2718 + }, + { + "epoch": 0.18682149237323073, + "grad_norm": 1.2488667964935303, + "learning_rate": 4.8861631920878104e-05, + "loss": 1.0599, + "step": 2719 + }, + { + "epoch": 0.18689020200632128, + "grad_norm": 1.0749056339263916, + "learning_rate": 4.8859842399763614e-05, + "loss": 0.9871, + "step": 2720 + }, + { + "epoch": 0.18695891163941183, + "grad_norm": 1.1804087162017822, + "learning_rate": 4.885805150600693e-05, + "loss": 1.1101, + "step": 2721 + }, + { + "epoch": 0.18702762127250241, + "grad_norm": 1.211484670639038, + "learning_rate": 4.88562592397111e-05, + "loss": 1.0289, + "step": 2722 + }, + { + "epoch": 0.18709633090559297, + "grad_norm": 1.1476444005966187, + "learning_rate": 4.8854465600979216e-05, + "loss": 1.0088, + "step": 2723 + }, + { + "epoch": 0.18716504053868352, + "grad_norm": 1.214740514755249, + "learning_rate": 4.8852670589914475e-05, + "loss": 1.0926, + "step": 2724 + }, + { + "epoch": 0.18723375017177407, + "grad_norm": 1.0797841548919678, + "learning_rate": 4.8850874206620134e-05, + "loss": 0.9935, + "step": 2725 + }, + { + "epoch": 0.18730245980486465, + "grad_norm": 1.1021164655685425, + "learning_rate": 4.8849076451199544e-05, + "loss": 1.0098, + "step": 2726 + }, + { + "epoch": 0.1873711694379552, + "grad_norm": 1.1362757682800293, + "learning_rate": 4.884727732375613e-05, + "loss": 1.009, + "step": 2727 + }, + { + "epoch": 0.18743987907104576, + "grad_norm": 1.1530351638793945, + "learning_rate": 4.884547682439339e-05, + "loss": 1.0397, + "step": 2728 + }, + { + "epoch": 0.1875085887041363, + "grad_norm": 1.0977703332901, + "learning_rate": 4.884367495321491e-05, + "loss": 1.0438, + "step": 2729 + }, + { + "epoch": 0.1875772983372269, + "grad_norm": 1.1632192134857178, + "learning_rate": 4.884187171032434e-05, + "loss": 1.0344, + "step": 2730 + }, + { + "epoch": 0.18764600797031744, + "grad_norm": 1.169043779373169, + "learning_rate": 4.884006709582544e-05, + "loss": 1.0554, + "step": 2731 + }, + { + "epoch": 0.187714717603408, + "grad_norm": 1.1728906631469727, + "learning_rate": 4.8838261109822006e-05, + "loss": 1.028, + "step": 2732 + }, + { + "epoch": 0.18778342723649855, + "grad_norm": 1.2064781188964844, + "learning_rate": 4.8836453752417955e-05, + "loss": 0.9905, + "step": 2733 + }, + { + "epoch": 0.18785213686958913, + "grad_norm": 1.1843647956848145, + "learning_rate": 4.883464502371724e-05, + "loss": 1.0684, + "step": 2734 + }, + { + "epoch": 0.18792084650267968, + "grad_norm": 1.0816813707351685, + "learning_rate": 4.883283492382394e-05, + "loss": 0.9908, + "step": 2735 + }, + { + "epoch": 0.18798955613577023, + "grad_norm": 1.1325968503952026, + "learning_rate": 4.883102345284218e-05, + "loss": 1.0239, + "step": 2736 + }, + { + "epoch": 0.18805826576886078, + "grad_norm": 1.1772798299789429, + "learning_rate": 4.882921061087616e-05, + "loss": 0.9915, + "step": 2737 + }, + { + "epoch": 0.18812697540195136, + "grad_norm": 1.0947883129119873, + "learning_rate": 4.8827396398030196e-05, + "loss": 1.0155, + "step": 2738 + }, + { + "epoch": 0.18819568503504192, + "grad_norm": 1.204433560371399, + "learning_rate": 4.882558081440864e-05, + "loss": 1.0739, + "step": 2739 + }, + { + "epoch": 0.18826439466813247, + "grad_norm": 1.175732135772705, + "learning_rate": 4.8823763860115944e-05, + "loss": 1.0306, + "step": 2740 + }, + { + "epoch": 0.18833310430122302, + "grad_norm": 1.0902233123779297, + "learning_rate": 4.8821945535256644e-05, + "loss": 1.0021, + "step": 2741 + }, + { + "epoch": 0.1884018139343136, + "grad_norm": 1.3316148519515991, + "learning_rate": 4.882012583993534e-05, + "loss": 1.0753, + "step": 2742 + }, + { + "epoch": 0.18847052356740415, + "grad_norm": 1.167851448059082, + "learning_rate": 4.881830477425672e-05, + "loss": 1.0591, + "step": 2743 + }, + { + "epoch": 0.1885392332004947, + "grad_norm": 1.124693751335144, + "learning_rate": 4.881648233832555e-05, + "loss": 1.0725, + "step": 2744 + }, + { + "epoch": 0.18860794283358526, + "grad_norm": 1.1950538158416748, + "learning_rate": 4.881465853224667e-05, + "loss": 0.9972, + "step": 2745 + }, + { + "epoch": 0.18867665246667584, + "grad_norm": 1.2331063747406006, + "learning_rate": 4.881283335612501e-05, + "loss": 1.0239, + "step": 2746 + }, + { + "epoch": 0.1887453620997664, + "grad_norm": 1.0918314456939697, + "learning_rate": 4.8811006810065564e-05, + "loss": 1.0207, + "step": 2747 + }, + { + "epoch": 0.18881407173285694, + "grad_norm": 1.240299105644226, + "learning_rate": 4.880917889417342e-05, + "loss": 1.0116, + "step": 2748 + }, + { + "epoch": 0.1888827813659475, + "grad_norm": 1.2081960439682007, + "learning_rate": 4.880734960855372e-05, + "loss": 0.9896, + "step": 2749 + }, + { + "epoch": 0.18895149099903807, + "grad_norm": 1.0632967948913574, + "learning_rate": 4.8805518953311724e-05, + "loss": 1.0139, + "step": 2750 + }, + { + "epoch": 0.18902020063212863, + "grad_norm": 1.1283340454101562, + "learning_rate": 4.8803686928552736e-05, + "loss": 1.0028, + "step": 2751 + }, + { + "epoch": 0.18908891026521918, + "grad_norm": 1.144422173500061, + "learning_rate": 4.8801853534382145e-05, + "loss": 1.0267, + "step": 2752 + }, + { + "epoch": 0.18915761989830973, + "grad_norm": 1.1197495460510254, + "learning_rate": 4.8800018770905444e-05, + "loss": 1.0292, + "step": 2753 + }, + { + "epoch": 0.1892263295314003, + "grad_norm": 1.20218026638031, + "learning_rate": 4.8798182638228166e-05, + "loss": 1.0466, + "step": 2754 + }, + { + "epoch": 0.18929503916449086, + "grad_norm": 1.1555067300796509, + "learning_rate": 4.879634513645596e-05, + "loss": 1.0131, + "step": 2755 + }, + { + "epoch": 0.18936374879758142, + "grad_norm": 1.1793274879455566, + "learning_rate": 4.8794506265694515e-05, + "loss": 1.0465, + "step": 2756 + }, + { + "epoch": 0.18943245843067197, + "grad_norm": 1.2320914268493652, + "learning_rate": 4.8792666026049636e-05, + "loss": 1.0313, + "step": 2757 + }, + { + "epoch": 0.18950116806376255, + "grad_norm": 1.240899682044983, + "learning_rate": 4.8790824417627195e-05, + "loss": 1.0617, + "step": 2758 + }, + { + "epoch": 0.1895698776968531, + "grad_norm": 1.1653006076812744, + "learning_rate": 4.8788981440533126e-05, + "loss": 1.0479, + "step": 2759 + }, + { + "epoch": 0.18963858732994365, + "grad_norm": 1.1640629768371582, + "learning_rate": 4.8787137094873464e-05, + "loss": 1.0497, + "step": 2760 + }, + { + "epoch": 0.1897072969630342, + "grad_norm": 1.222497582435608, + "learning_rate": 4.8785291380754304e-05, + "loss": 1.0247, + "step": 2761 + }, + { + "epoch": 0.18977600659612479, + "grad_norm": 1.1650795936584473, + "learning_rate": 4.878344429828184e-05, + "loss": 1.0098, + "step": 2762 + }, + { + "epoch": 0.18984471622921534, + "grad_norm": 1.1005711555480957, + "learning_rate": 4.878159584756232e-05, + "loss": 1.0219, + "step": 2763 + }, + { + "epoch": 0.1899134258623059, + "grad_norm": 1.1121512651443481, + "learning_rate": 4.877974602870209e-05, + "loss": 0.9899, + "step": 2764 + }, + { + "epoch": 0.18998213549539644, + "grad_norm": 1.1171183586120605, + "learning_rate": 4.8777894841807584e-05, + "loss": 1.0245, + "step": 2765 + }, + { + "epoch": 0.19005084512848702, + "grad_norm": 1.1539112329483032, + "learning_rate": 4.877604228698527e-05, + "loss": 1.0457, + "step": 2766 + }, + { + "epoch": 0.19011955476157757, + "grad_norm": 1.085932970046997, + "learning_rate": 4.877418836434174e-05, + "loss": 1.0313, + "step": 2767 + }, + { + "epoch": 0.19018826439466813, + "grad_norm": 1.109019160270691, + "learning_rate": 4.877233307398367e-05, + "loss": 0.9979, + "step": 2768 + }, + { + "epoch": 0.19025697402775868, + "grad_norm": 1.1723535060882568, + "learning_rate": 4.877047641601775e-05, + "loss": 1.0602, + "step": 2769 + }, + { + "epoch": 0.19032568366084926, + "grad_norm": 1.1960244178771973, + "learning_rate": 4.876861839055083e-05, + "loss": 1.0264, + "step": 2770 + }, + { + "epoch": 0.1903943932939398, + "grad_norm": 1.187696099281311, + "learning_rate": 4.876675899768978e-05, + "loss": 1.0105, + "step": 2771 + }, + { + "epoch": 0.19046310292703036, + "grad_norm": 1.092822551727295, + "learning_rate": 4.876489823754158e-05, + "loss": 1.0287, + "step": 2772 + }, + { + "epoch": 0.19053181256012092, + "grad_norm": 1.2227849960327148, + "learning_rate": 4.876303611021327e-05, + "loss": 1.0743, + "step": 2773 + }, + { + "epoch": 0.1906005221932115, + "grad_norm": 1.1735219955444336, + "learning_rate": 4.876117261581198e-05, + "loss": 1.0333, + "step": 2774 + }, + { + "epoch": 0.19066923182630205, + "grad_norm": 1.1534497737884521, + "learning_rate": 4.875930775444492e-05, + "loss": 1.0252, + "step": 2775 + }, + { + "epoch": 0.1907379414593926, + "grad_norm": 1.213480830192566, + "learning_rate": 4.8757441526219375e-05, + "loss": 1.0373, + "step": 2776 + }, + { + "epoch": 0.19080665109248315, + "grad_norm": 1.2066891193389893, + "learning_rate": 4.8755573931242704e-05, + "loss": 1.0161, + "step": 2777 + }, + { + "epoch": 0.19087536072557373, + "grad_norm": 1.2316193580627441, + "learning_rate": 4.875370496962234e-05, + "loss": 1.0259, + "step": 2778 + }, + { + "epoch": 0.1909440703586643, + "grad_norm": 1.1809360980987549, + "learning_rate": 4.875183464146582e-05, + "loss": 1.0817, + "step": 2779 + }, + { + "epoch": 0.19101277999175484, + "grad_norm": 1.1079050302505493, + "learning_rate": 4.874996294688074e-05, + "loss": 1.0178, + "step": 2780 + }, + { + "epoch": 0.1910814896248454, + "grad_norm": 1.1080548763275146, + "learning_rate": 4.874808988597477e-05, + "loss": 1.0064, + "step": 2781 + }, + { + "epoch": 0.19115019925793597, + "grad_norm": 1.228547215461731, + "learning_rate": 4.874621545885566e-05, + "loss": 1.0607, + "step": 2782 + }, + { + "epoch": 0.19121890889102652, + "grad_norm": 1.2200863361358643, + "learning_rate": 4.874433966563127e-05, + "loss": 1.0537, + "step": 2783 + }, + { + "epoch": 0.19128761852411708, + "grad_norm": 1.1762925386428833, + "learning_rate": 4.874246250640948e-05, + "loss": 0.9773, + "step": 2784 + }, + { + "epoch": 0.19135632815720763, + "grad_norm": 1.0731611251831055, + "learning_rate": 4.874058398129831e-05, + "loss": 0.9889, + "step": 2785 + }, + { + "epoch": 0.1914250377902982, + "grad_norm": 1.083801031112671, + "learning_rate": 4.873870409040581e-05, + "loss": 1.0569, + "step": 2786 + }, + { + "epoch": 0.19149374742338876, + "grad_norm": 1.1964901685714722, + "learning_rate": 4.873682283384014e-05, + "loss": 1.0425, + "step": 2787 + }, + { + "epoch": 0.1915624570564793, + "grad_norm": 1.2355222702026367, + "learning_rate": 4.873494021170953e-05, + "loss": 1.0339, + "step": 2788 + }, + { + "epoch": 0.19163116668956987, + "grad_norm": 1.2363396883010864, + "learning_rate": 4.8733056224122284e-05, + "loss": 1.0853, + "step": 2789 + }, + { + "epoch": 0.19169987632266045, + "grad_norm": 1.1626554727554321, + "learning_rate": 4.873117087118678e-05, + "loss": 1.011, + "step": 2790 + }, + { + "epoch": 0.191768585955751, + "grad_norm": 1.1757115125656128, + "learning_rate": 4.872928415301149e-05, + "loss": 1.0089, + "step": 2791 + }, + { + "epoch": 0.19183729558884155, + "grad_norm": 1.1835823059082031, + "learning_rate": 4.872739606970494e-05, + "loss": 1.0534, + "step": 2792 + }, + { + "epoch": 0.1919060052219321, + "grad_norm": 1.153192162513733, + "learning_rate": 4.872550662137577e-05, + "loss": 0.9837, + "step": 2793 + }, + { + "epoch": 0.19197471485502268, + "grad_norm": 1.1496317386627197, + "learning_rate": 4.872361580813267e-05, + "loss": 1.0007, + "step": 2794 + }, + { + "epoch": 0.19204342448811323, + "grad_norm": 1.1560559272766113, + "learning_rate": 4.872172363008441e-05, + "loss": 1.0268, + "step": 2795 + }, + { + "epoch": 0.1921121341212038, + "grad_norm": 1.2283042669296265, + "learning_rate": 4.871983008733987e-05, + "loss": 1.0275, + "step": 2796 + }, + { + "epoch": 0.19218084375429434, + "grad_norm": 1.1835542917251587, + "learning_rate": 4.871793518000796e-05, + "loss": 1.0407, + "step": 2797 + }, + { + "epoch": 0.19224955338738492, + "grad_norm": 1.2196686267852783, + "learning_rate": 4.8716038908197694e-05, + "loss": 1.0804, + "step": 2798 + }, + { + "epoch": 0.19231826302047547, + "grad_norm": 1.1109960079193115, + "learning_rate": 4.8714141272018174e-05, + "loss": 0.996, + "step": 2799 + }, + { + "epoch": 0.19238697265356602, + "grad_norm": 1.1496388912200928, + "learning_rate": 4.871224227157857e-05, + "loss": 1.0071, + "step": 2800 + }, + { + "epoch": 0.19245568228665658, + "grad_norm": 1.155119776725769, + "learning_rate": 4.8710341906988113e-05, + "loss": 1.0224, + "step": 2801 + }, + { + "epoch": 0.19252439191974716, + "grad_norm": 1.1068319082260132, + "learning_rate": 4.8708440178356155e-05, + "loss": 1.0444, + "step": 2802 + }, + { + "epoch": 0.1925931015528377, + "grad_norm": 1.1548848152160645, + "learning_rate": 4.870653708579209e-05, + "loss": 1.001, + "step": 2803 + }, + { + "epoch": 0.19266181118592826, + "grad_norm": 1.179634928703308, + "learning_rate": 4.87046326294054e-05, + "loss": 1.0029, + "step": 2804 + }, + { + "epoch": 0.1927305208190188, + "grad_norm": 1.1946998834609985, + "learning_rate": 4.870272680930564e-05, + "loss": 1.0051, + "step": 2805 + }, + { + "epoch": 0.1927992304521094, + "grad_norm": 1.2603615522384644, + "learning_rate": 4.8700819625602467e-05, + "loss": 1.0723, + "step": 2806 + }, + { + "epoch": 0.19286794008519995, + "grad_norm": 1.1071715354919434, + "learning_rate": 4.86989110784056e-05, + "loss": 1.0267, + "step": 2807 + }, + { + "epoch": 0.1929366497182905, + "grad_norm": 1.2459923028945923, + "learning_rate": 4.869700116782481e-05, + "loss": 1.028, + "step": 2808 + }, + { + "epoch": 0.19300535935138105, + "grad_norm": 1.1276764869689941, + "learning_rate": 4.869508989397e-05, + "loss": 1.0008, + "step": 2809 + }, + { + "epoch": 0.19307406898447163, + "grad_norm": 1.1580579280853271, + "learning_rate": 4.869317725695112e-05, + "loss": 1.076, + "step": 2810 + }, + { + "epoch": 0.19314277861756218, + "grad_norm": 1.1694746017456055, + "learning_rate": 4.8691263256878196e-05, + "loss": 1.0494, + "step": 2811 + }, + { + "epoch": 0.19321148825065274, + "grad_norm": 1.1122609376907349, + "learning_rate": 4.868934789386134e-05, + "loss": 0.9959, + "step": 2812 + }, + { + "epoch": 0.1932801978837433, + "grad_norm": 1.1557917594909668, + "learning_rate": 4.868743116801074e-05, + "loss": 1.0608, + "step": 2813 + }, + { + "epoch": 0.19334890751683387, + "grad_norm": 1.153130054473877, + "learning_rate": 4.868551307943668e-05, + "loss": 1.0363, + "step": 2814 + }, + { + "epoch": 0.19341761714992442, + "grad_norm": 1.0876350402832031, + "learning_rate": 4.868359362824949e-05, + "loss": 1.007, + "step": 2815 + }, + { + "epoch": 0.19348632678301497, + "grad_norm": 1.10475754737854, + "learning_rate": 4.8681672814559597e-05, + "loss": 0.9883, + "step": 2816 + }, + { + "epoch": 0.19355503641610552, + "grad_norm": 1.2718753814697266, + "learning_rate": 4.867975063847751e-05, + "loss": 1.0945, + "step": 2817 + }, + { + "epoch": 0.1936237460491961, + "grad_norm": 1.1154457330703735, + "learning_rate": 4.8677827100113796e-05, + "loss": 0.9711, + "step": 2818 + }, + { + "epoch": 0.19369245568228666, + "grad_norm": 1.1154392957687378, + "learning_rate": 4.8675902199579134e-05, + "loss": 1.0078, + "step": 2819 + }, + { + "epoch": 0.1937611653153772, + "grad_norm": 1.1651206016540527, + "learning_rate": 4.867397593698426e-05, + "loss": 1.0494, + "step": 2820 + }, + { + "epoch": 0.19382987494846776, + "grad_norm": 1.1507261991500854, + "learning_rate": 4.867204831243998e-05, + "loss": 1.0143, + "step": 2821 + }, + { + "epoch": 0.19389858458155834, + "grad_norm": 1.1626824140548706, + "learning_rate": 4.8670119326057193e-05, + "loss": 1.038, + "step": 2822 + }, + { + "epoch": 0.1939672942146489, + "grad_norm": 1.2161059379577637, + "learning_rate": 4.866818897794687e-05, + "loss": 0.9918, + "step": 2823 + }, + { + "epoch": 0.19403600384773945, + "grad_norm": 1.1840440034866333, + "learning_rate": 4.866625726822007e-05, + "loss": 1.0509, + "step": 2824 + }, + { + "epoch": 0.19410471348083, + "grad_norm": 1.1685023307800293, + "learning_rate": 4.866432419698792e-05, + "loss": 1.042, + "step": 2825 + }, + { + "epoch": 0.19417342311392058, + "grad_norm": 1.133290410041809, + "learning_rate": 4.866238976436163e-05, + "loss": 1.022, + "step": 2826 + }, + { + "epoch": 0.19424213274701113, + "grad_norm": 1.0342984199523926, + "learning_rate": 4.866045397045248e-05, + "loss": 0.9841, + "step": 2827 + }, + { + "epoch": 0.19431084238010168, + "grad_norm": 1.1483962535858154, + "learning_rate": 4.8658516815371844e-05, + "loss": 0.9952, + "step": 2828 + }, + { + "epoch": 0.19437955201319224, + "grad_norm": 1.0960545539855957, + "learning_rate": 4.865657829923116e-05, + "loss": 1.0444, + "step": 2829 + }, + { + "epoch": 0.19444826164628282, + "grad_norm": 1.133612036705017, + "learning_rate": 4.865463842214195e-05, + "loss": 1.0298, + "step": 2830 + }, + { + "epoch": 0.19451697127937337, + "grad_norm": 1.1421371698379517, + "learning_rate": 4.8652697184215816e-05, + "loss": 1.0183, + "step": 2831 + }, + { + "epoch": 0.19458568091246392, + "grad_norm": 1.105244755744934, + "learning_rate": 4.865075458556444e-05, + "loss": 1.0117, + "step": 2832 + }, + { + "epoch": 0.19465439054555447, + "grad_norm": 1.1536036729812622, + "learning_rate": 4.8648810626299564e-05, + "loss": 0.9928, + "step": 2833 + }, + { + "epoch": 0.19472310017864505, + "grad_norm": 1.0787522792816162, + "learning_rate": 4.864686530653303e-05, + "loss": 0.9867, + "step": 2834 + }, + { + "epoch": 0.1947918098117356, + "grad_norm": 1.146125316619873, + "learning_rate": 4.864491862637676e-05, + "loss": 1.0395, + "step": 2835 + }, + { + "epoch": 0.19486051944482616, + "grad_norm": 1.18100106716156, + "learning_rate": 4.864297058594274e-05, + "loss": 1.0535, + "step": 2836 + }, + { + "epoch": 0.1949292290779167, + "grad_norm": 1.1942782402038574, + "learning_rate": 4.8641021185343035e-05, + "loss": 1.0643, + "step": 2837 + }, + { + "epoch": 0.1949979387110073, + "grad_norm": 1.1069035530090332, + "learning_rate": 4.8639070424689795e-05, + "loss": 1.0219, + "step": 2838 + }, + { + "epoch": 0.19506664834409784, + "grad_norm": 1.1614689826965332, + "learning_rate": 4.863711830409525e-05, + "loss": 1.0684, + "step": 2839 + }, + { + "epoch": 0.1951353579771884, + "grad_norm": 1.0856672525405884, + "learning_rate": 4.86351648236717e-05, + "loss": 1.0318, + "step": 2840 + }, + { + "epoch": 0.19520406761027895, + "grad_norm": 1.1302341222763062, + "learning_rate": 4.8633209983531526e-05, + "loss": 1.0378, + "step": 2841 + }, + { + "epoch": 0.19527277724336953, + "grad_norm": 1.0803455114364624, + "learning_rate": 4.863125378378719e-05, + "loss": 1.0276, + "step": 2842 + }, + { + "epoch": 0.19534148687646008, + "grad_norm": 1.1720216274261475, + "learning_rate": 4.8629296224551234e-05, + "loss": 1.106, + "step": 2843 + }, + { + "epoch": 0.19541019650955063, + "grad_norm": 1.0796507596969604, + "learning_rate": 4.8627337305936275e-05, + "loss": 1.0126, + "step": 2844 + }, + { + "epoch": 0.19547890614264118, + "grad_norm": 1.1778688430786133, + "learning_rate": 4.8625377028055006e-05, + "loss": 1.037, + "step": 2845 + }, + { + "epoch": 0.19554761577573176, + "grad_norm": 1.0902340412139893, + "learning_rate": 4.862341539102021e-05, + "loss": 1.0349, + "step": 2846 + }, + { + "epoch": 0.19561632540882232, + "grad_norm": 1.1670628786087036, + "learning_rate": 4.862145239494472e-05, + "loss": 0.9966, + "step": 2847 + }, + { + "epoch": 0.19568503504191287, + "grad_norm": 1.173034429550171, + "learning_rate": 4.8619488039941484e-05, + "loss": 1.0635, + "step": 2848 + }, + { + "epoch": 0.19575374467500345, + "grad_norm": 1.095612645149231, + "learning_rate": 4.86175223261235e-05, + "loss": 1.0161, + "step": 2849 + }, + { + "epoch": 0.195822454308094, + "grad_norm": 1.1629751920700073, + "learning_rate": 4.861555525360386e-05, + "loss": 1.0196, + "step": 2850 + }, + { + "epoch": 0.19589116394118455, + "grad_norm": 1.139648675918579, + "learning_rate": 4.861358682249572e-05, + "loss": 1.0345, + "step": 2851 + }, + { + "epoch": 0.1959598735742751, + "grad_norm": 1.11143159866333, + "learning_rate": 4.8611617032912326e-05, + "loss": 1.0079, + "step": 2852 + }, + { + "epoch": 0.1960285832073657, + "grad_norm": 1.2658207416534424, + "learning_rate": 4.8609645884967005e-05, + "loss": 1.0183, + "step": 2853 + }, + { + "epoch": 0.19609729284045624, + "grad_norm": 1.0817304849624634, + "learning_rate": 4.8607673378773156e-05, + "loss": 1.0107, + "step": 2854 + }, + { + "epoch": 0.1961660024735468, + "grad_norm": 1.1624689102172852, + "learning_rate": 4.860569951444424e-05, + "loss": 0.9975, + "step": 2855 + }, + { + "epoch": 0.19623471210663734, + "grad_norm": 1.2008650302886963, + "learning_rate": 4.860372429209383e-05, + "loss": 1.065, + "step": 2856 + }, + { + "epoch": 0.19630342173972792, + "grad_norm": 1.137695550918579, + "learning_rate": 4.860174771183556e-05, + "loss": 1.0499, + "step": 2857 + }, + { + "epoch": 0.19637213137281848, + "grad_norm": 1.2243564128875732, + "learning_rate": 4.8599769773783134e-05, + "loss": 1.0361, + "step": 2858 + }, + { + "epoch": 0.19644084100590903, + "grad_norm": 1.1663775444030762, + "learning_rate": 4.859779047805033e-05, + "loss": 1.0418, + "step": 2859 + }, + { + "epoch": 0.19650955063899958, + "grad_norm": 1.1342673301696777, + "learning_rate": 4.859580982475104e-05, + "loss": 1.018, + "step": 2860 + }, + { + "epoch": 0.19657826027209016, + "grad_norm": 1.1241106986999512, + "learning_rate": 4.8593827813999194e-05, + "loss": 1.0095, + "step": 2861 + }, + { + "epoch": 0.1966469699051807, + "grad_norm": 1.164384365081787, + "learning_rate": 4.859184444590882e-05, + "loss": 1.0003, + "step": 2862 + }, + { + "epoch": 0.19671567953827127, + "grad_norm": 1.1002206802368164, + "learning_rate": 4.858985972059402e-05, + "loss": 1.018, + "step": 2863 + }, + { + "epoch": 0.19678438917136182, + "grad_norm": 1.1314283609390259, + "learning_rate": 4.858787363816898e-05, + "loss": 1.0008, + "step": 2864 + }, + { + "epoch": 0.1968530988044524, + "grad_norm": 1.1572357416152954, + "learning_rate": 4.858588619874794e-05, + "loss": 1.0568, + "step": 2865 + }, + { + "epoch": 0.19692180843754295, + "grad_norm": 1.1417245864868164, + "learning_rate": 4.8583897402445254e-05, + "loss": 1.0095, + "step": 2866 + }, + { + "epoch": 0.1969905180706335, + "grad_norm": 1.0932618379592896, + "learning_rate": 4.858190724937533e-05, + "loss": 0.9963, + "step": 2867 + }, + { + "epoch": 0.19705922770372405, + "grad_norm": 1.1719475984573364, + "learning_rate": 4.857991573965266e-05, + "loss": 1.0132, + "step": 2868 + }, + { + "epoch": 0.19712793733681463, + "grad_norm": 1.149912714958191, + "learning_rate": 4.8577922873391814e-05, + "loss": 1.0594, + "step": 2869 + }, + { + "epoch": 0.1971966469699052, + "grad_norm": 1.1564011573791504, + "learning_rate": 4.857592865070745e-05, + "loss": 1.0318, + "step": 2870 + }, + { + "epoch": 0.19726535660299574, + "grad_norm": 1.141700029373169, + "learning_rate": 4.857393307171427e-05, + "loss": 1.0126, + "step": 2871 + }, + { + "epoch": 0.1973340662360863, + "grad_norm": 1.2078033685684204, + "learning_rate": 4.857193613652711e-05, + "loss": 1.0334, + "step": 2872 + }, + { + "epoch": 0.19740277586917687, + "grad_norm": 1.1191651821136475, + "learning_rate": 4.8569937845260824e-05, + "loss": 1.0207, + "step": 2873 + }, + { + "epoch": 0.19747148550226742, + "grad_norm": 1.1767507791519165, + "learning_rate": 4.856793819803039e-05, + "loss": 1.0225, + "step": 2874 + }, + { + "epoch": 0.19754019513535798, + "grad_norm": 1.2308937311172485, + "learning_rate": 4.856593719495084e-05, + "loss": 1.0253, + "step": 2875 + }, + { + "epoch": 0.19760890476844853, + "grad_norm": 1.2305316925048828, + "learning_rate": 4.85639348361373e-05, + "loss": 1.0328, + "step": 2876 + }, + { + "epoch": 0.1976776144015391, + "grad_norm": 1.131380558013916, + "learning_rate": 4.856193112170495e-05, + "loss": 1.0135, + "step": 2877 + }, + { + "epoch": 0.19774632403462966, + "grad_norm": 1.1656721830368042, + "learning_rate": 4.855992605176907e-05, + "loss": 1.0763, + "step": 2878 + }, + { + "epoch": 0.1978150336677202, + "grad_norm": 1.146160364151001, + "learning_rate": 4.855791962644501e-05, + "loss": 1.0922, + "step": 2879 + }, + { + "epoch": 0.19788374330081077, + "grad_norm": 1.1927412748336792, + "learning_rate": 4.85559118458482e-05, + "loss": 1.0838, + "step": 2880 + }, + { + "epoch": 0.19795245293390135, + "grad_norm": 1.1864092350006104, + "learning_rate": 4.855390271009414e-05, + "loss": 1.0507, + "step": 2881 + }, + { + "epoch": 0.1980211625669919, + "grad_norm": 1.1308996677398682, + "learning_rate": 4.8551892219298426e-05, + "loss": 1.0172, + "step": 2882 + }, + { + "epoch": 0.19808987220008245, + "grad_norm": 1.157084345817566, + "learning_rate": 4.85498803735767e-05, + "loss": 1.0536, + "step": 2883 + }, + { + "epoch": 0.198158581833173, + "grad_norm": 1.173508882522583, + "learning_rate": 4.8547867173044724e-05, + "loss": 1.047, + "step": 2884 + }, + { + "epoch": 0.19822729146626358, + "grad_norm": 1.1931278705596924, + "learning_rate": 4.854585261781831e-05, + "loss": 1.0236, + "step": 2885 + }, + { + "epoch": 0.19829600109935414, + "grad_norm": 1.1004407405853271, + "learning_rate": 4.8543836708013346e-05, + "loss": 1.018, + "step": 2886 + }, + { + "epoch": 0.1983647107324447, + "grad_norm": 1.1369121074676514, + "learning_rate": 4.854181944374581e-05, + "loss": 1.0065, + "step": 2887 + }, + { + "epoch": 0.19843342036553524, + "grad_norm": 1.0863412618637085, + "learning_rate": 4.8539800825131764e-05, + "loss": 1.0511, + "step": 2888 + }, + { + "epoch": 0.19850212999862582, + "grad_norm": 1.1656053066253662, + "learning_rate": 4.8537780852287316e-05, + "loss": 1.035, + "step": 2889 + }, + { + "epoch": 0.19857083963171637, + "grad_norm": 1.1472002267837524, + "learning_rate": 4.85357595253287e-05, + "loss": 1.0443, + "step": 2890 + }, + { + "epoch": 0.19863954926480692, + "grad_norm": 1.2241532802581787, + "learning_rate": 4.8533736844372184e-05, + "loss": 1.0307, + "step": 2891 + }, + { + "epoch": 0.19870825889789748, + "grad_norm": 1.254367709159851, + "learning_rate": 4.853171280953414e-05, + "loss": 1.043, + "step": 2892 + }, + { + "epoch": 0.19877696853098806, + "grad_norm": 1.1269181966781616, + "learning_rate": 4.8529687420931006e-05, + "loss": 0.9847, + "step": 2893 + }, + { + "epoch": 0.1988456781640786, + "grad_norm": 1.0743377208709717, + "learning_rate": 4.85276606786793e-05, + "loss": 1.0103, + "step": 2894 + }, + { + "epoch": 0.19891438779716916, + "grad_norm": 1.166887640953064, + "learning_rate": 4.852563258289562e-05, + "loss": 1.0535, + "step": 2895 + }, + { + "epoch": 0.19898309743025971, + "grad_norm": 1.1123852729797363, + "learning_rate": 4.8523603133696635e-05, + "loss": 1.0511, + "step": 2896 + }, + { + "epoch": 0.1990518070633503, + "grad_norm": 1.238600730895996, + "learning_rate": 4.8521572331199115e-05, + "loss": 1.0417, + "step": 2897 + }, + { + "epoch": 0.19912051669644085, + "grad_norm": 1.246733546257019, + "learning_rate": 4.851954017551988e-05, + "loss": 1.0773, + "step": 2898 + }, + { + "epoch": 0.1991892263295314, + "grad_norm": 1.2074594497680664, + "learning_rate": 4.851750666677584e-05, + "loss": 1.0202, + "step": 2899 + }, + { + "epoch": 0.19925793596262195, + "grad_norm": 1.27336847782135, + "learning_rate": 4.851547180508397e-05, + "loss": 1.0432, + "step": 2900 + }, + { + "epoch": 0.19932664559571253, + "grad_norm": 1.2648990154266357, + "learning_rate": 4.851343559056135e-05, + "loss": 1.0054, + "step": 2901 + }, + { + "epoch": 0.19939535522880308, + "grad_norm": 1.154901385307312, + "learning_rate": 4.8511398023325125e-05, + "loss": 1.0393, + "step": 2902 + }, + { + "epoch": 0.19946406486189364, + "grad_norm": 1.1975780725479126, + "learning_rate": 4.85093591034925e-05, + "loss": 1.0168, + "step": 2903 + }, + { + "epoch": 0.1995327744949842, + "grad_norm": 1.1483519077301025, + "learning_rate": 4.850731883118078e-05, + "loss": 1.0271, + "step": 2904 + }, + { + "epoch": 0.19960148412807477, + "grad_norm": 1.1329030990600586, + "learning_rate": 4.8505277206507346e-05, + "loss": 0.9835, + "step": 2905 + }, + { + "epoch": 0.19967019376116532, + "grad_norm": 1.166006088256836, + "learning_rate": 4.850323422958964e-05, + "loss": 1.0337, + "step": 2906 + }, + { + "epoch": 0.19973890339425587, + "grad_norm": 1.1600127220153809, + "learning_rate": 4.850118990054521e-05, + "loss": 1.0504, + "step": 2907 + }, + { + "epoch": 0.19980761302734643, + "grad_norm": 1.1990917921066284, + "learning_rate": 4.8499144219491646e-05, + "loss": 1.0574, + "step": 2908 + }, + { + "epoch": 0.199876322660437, + "grad_norm": 1.1139283180236816, + "learning_rate": 4.8497097186546645e-05, + "loss": 0.9862, + "step": 2909 + }, + { + "epoch": 0.19994503229352756, + "grad_norm": 1.1637606620788574, + "learning_rate": 4.849504880182797e-05, + "loss": 0.9979, + "step": 2910 + }, + { + "epoch": 0.2000137419266181, + "grad_norm": 1.208289623260498, + "learning_rate": 4.849299906545346e-05, + "loss": 1.0389, + "step": 2911 + }, + { + "epoch": 0.20008245155970866, + "grad_norm": 1.1238768100738525, + "learning_rate": 4.849094797754105e-05, + "loss": 1.0707, + "step": 2912 + }, + { + "epoch": 0.20015116119279924, + "grad_norm": 1.184517502784729, + "learning_rate": 4.8488895538208715e-05, + "loss": 1.0647, + "step": 2913 + }, + { + "epoch": 0.2002198708258898, + "grad_norm": 1.111506700515747, + "learning_rate": 4.848684174757454e-05, + "loss": 1.0343, + "step": 2914 + }, + { + "epoch": 0.20028858045898035, + "grad_norm": 1.2307666540145874, + "learning_rate": 4.8484786605756685e-05, + "loss": 1.0061, + "step": 2915 + }, + { + "epoch": 0.2003572900920709, + "grad_norm": 1.1550651788711548, + "learning_rate": 4.848273011287338e-05, + "loss": 1.0292, + "step": 2916 + }, + { + "epoch": 0.20042599972516148, + "grad_norm": 1.1062744855880737, + "learning_rate": 4.848067226904292e-05, + "loss": 1.0403, + "step": 2917 + }, + { + "epoch": 0.20049470935825203, + "grad_norm": 1.042489767074585, + "learning_rate": 4.8478613074383705e-05, + "loss": 0.9932, + "step": 2918 + }, + { + "epoch": 0.20056341899134258, + "grad_norm": 1.2445154190063477, + "learning_rate": 4.8476552529014195e-05, + "loss": 1.0525, + "step": 2919 + }, + { + "epoch": 0.20063212862443314, + "grad_norm": 1.1089403629302979, + "learning_rate": 4.847449063305294e-05, + "loss": 1.0294, + "step": 2920 + }, + { + "epoch": 0.20070083825752372, + "grad_norm": 1.1617282629013062, + "learning_rate": 4.847242738661855e-05, + "loss": 0.997, + "step": 2921 + }, + { + "epoch": 0.20076954789061427, + "grad_norm": 1.132398009300232, + "learning_rate": 4.8470362789829714e-05, + "loss": 1.0609, + "step": 2922 + }, + { + "epoch": 0.20083825752370482, + "grad_norm": 1.1309963464736938, + "learning_rate": 4.846829684280523e-05, + "loss": 0.9995, + "step": 2923 + }, + { + "epoch": 0.20090696715679537, + "grad_norm": 1.1238200664520264, + "learning_rate": 4.846622954566392e-05, + "loss": 1.0175, + "step": 2924 + }, + { + "epoch": 0.20097567678988595, + "grad_norm": 1.2122045755386353, + "learning_rate": 4.8464160898524744e-05, + "loss": 1.0935, + "step": 2925 + }, + { + "epoch": 0.2010443864229765, + "grad_norm": 1.1438333988189697, + "learning_rate": 4.8462090901506704e-05, + "loss": 0.9948, + "step": 2926 + }, + { + "epoch": 0.20111309605606706, + "grad_norm": 1.159052848815918, + "learning_rate": 4.8460019554728876e-05, + "loss": 1.0464, + "step": 2927 + }, + { + "epoch": 0.2011818056891576, + "grad_norm": 1.0827853679656982, + "learning_rate": 4.845794685831042e-05, + "loss": 1.0529, + "step": 2928 + }, + { + "epoch": 0.2012505153222482, + "grad_norm": 1.1707926988601685, + "learning_rate": 4.845587281237059e-05, + "loss": 1.0298, + "step": 2929 + }, + { + "epoch": 0.20131922495533874, + "grad_norm": 1.0884902477264404, + "learning_rate": 4.84537974170287e-05, + "loss": 1.0108, + "step": 2930 + }, + { + "epoch": 0.2013879345884293, + "grad_norm": 1.1716365814208984, + "learning_rate": 4.845172067240415e-05, + "loss": 0.997, + "step": 2931 + }, + { + "epoch": 0.20145664422151985, + "grad_norm": 1.2145098447799683, + "learning_rate": 4.8449642578616405e-05, + "loss": 1.0513, + "step": 2932 + }, + { + "epoch": 0.20152535385461043, + "grad_norm": 1.093224048614502, + "learning_rate": 4.844756313578501e-05, + "loss": 0.9891, + "step": 2933 + }, + { + "epoch": 0.20159406348770098, + "grad_norm": 1.1573151350021362, + "learning_rate": 4.844548234402962e-05, + "loss": 1.0115, + "step": 2934 + }, + { + "epoch": 0.20166277312079153, + "grad_norm": 1.112059473991394, + "learning_rate": 4.8443400203469915e-05, + "loss": 1.0014, + "step": 2935 + }, + { + "epoch": 0.20173148275388209, + "grad_norm": 1.0464282035827637, + "learning_rate": 4.84413167142257e-05, + "loss": 1.0409, + "step": 2936 + }, + { + "epoch": 0.20180019238697267, + "grad_norm": 1.1431435346603394, + "learning_rate": 4.8439231876416825e-05, + "loss": 1.0505, + "step": 2937 + }, + { + "epoch": 0.20186890202006322, + "grad_norm": 1.2363916635513306, + "learning_rate": 4.8437145690163225e-05, + "loss": 1.0538, + "step": 2938 + }, + { + "epoch": 0.20193761165315377, + "grad_norm": 1.1746454238891602, + "learning_rate": 4.843505815558493e-05, + "loss": 1.086, + "step": 2939 + }, + { + "epoch": 0.20200632128624432, + "grad_norm": 1.1659003496170044, + "learning_rate": 4.843296927280203e-05, + "loss": 1.0185, + "step": 2940 + }, + { + "epoch": 0.2020750309193349, + "grad_norm": 1.1310045719146729, + "learning_rate": 4.8430879041934684e-05, + "loss": 1.0639, + "step": 2941 + }, + { + "epoch": 0.20214374055242545, + "grad_norm": 1.1462703943252563, + "learning_rate": 4.8428787463103166e-05, + "loss": 1.0577, + "step": 2942 + }, + { + "epoch": 0.202212450185516, + "grad_norm": 1.1103769540786743, + "learning_rate": 4.842669453642779e-05, + "loss": 1.0009, + "step": 2943 + }, + { + "epoch": 0.20228115981860656, + "grad_norm": 1.1762871742248535, + "learning_rate": 4.8424600262028955e-05, + "loss": 1.0626, + "step": 2944 + }, + { + "epoch": 0.20234986945169714, + "grad_norm": 1.1194217205047607, + "learning_rate": 4.842250464002715e-05, + "loss": 1.0085, + "step": 2945 + }, + { + "epoch": 0.2024185790847877, + "grad_norm": 1.1482247114181519, + "learning_rate": 4.8420407670542934e-05, + "loss": 1.0191, + "step": 2946 + }, + { + "epoch": 0.20248728871787824, + "grad_norm": 1.1108320951461792, + "learning_rate": 4.841830935369695e-05, + "loss": 1.0355, + "step": 2947 + }, + { + "epoch": 0.2025559983509688, + "grad_norm": 1.1823300123214722, + "learning_rate": 4.8416209689609906e-05, + "loss": 1.0219, + "step": 2948 + }, + { + "epoch": 0.20262470798405938, + "grad_norm": 1.194290041923523, + "learning_rate": 4.8414108678402595e-05, + "loss": 1.0292, + "step": 2949 + }, + { + "epoch": 0.20269341761714993, + "grad_norm": 1.1673853397369385, + "learning_rate": 4.841200632019589e-05, + "loss": 1.0354, + "step": 2950 + }, + { + "epoch": 0.20276212725024048, + "grad_norm": 1.2086540460586548, + "learning_rate": 4.8409902615110736e-05, + "loss": 1.0469, + "step": 2951 + }, + { + "epoch": 0.20283083688333103, + "grad_norm": 1.1900922060012817, + "learning_rate": 4.8407797563268154e-05, + "loss": 1.0492, + "step": 2952 + }, + { + "epoch": 0.2028995465164216, + "grad_norm": 1.1924707889556885, + "learning_rate": 4.8405691164789255e-05, + "loss": 1.0022, + "step": 2953 + }, + { + "epoch": 0.20296825614951217, + "grad_norm": 1.2250827550888062, + "learning_rate": 4.840358341979522e-05, + "loss": 1.0207, + "step": 2954 + }, + { + "epoch": 0.20303696578260272, + "grad_norm": 1.0617984533309937, + "learning_rate": 4.8401474328407294e-05, + "loss": 0.9714, + "step": 2955 + }, + { + "epoch": 0.20310567541569327, + "grad_norm": 1.182363510131836, + "learning_rate": 4.8399363890746815e-05, + "loss": 1.0548, + "step": 2956 + }, + { + "epoch": 0.20317438504878385, + "grad_norm": 1.1219449043273926, + "learning_rate": 4.83972521069352e-05, + "loss": 1.0021, + "step": 2957 + }, + { + "epoch": 0.2032430946818744, + "grad_norm": 1.0955256223678589, + "learning_rate": 4.8395138977093946e-05, + "loss": 0.9842, + "step": 2958 + }, + { + "epoch": 0.20331180431496496, + "grad_norm": 1.2314808368682861, + "learning_rate": 4.83930245013446e-05, + "loss": 1.036, + "step": 2959 + }, + { + "epoch": 0.2033805139480555, + "grad_norm": 1.1447805166244507, + "learning_rate": 4.839090867980882e-05, + "loss": 0.9825, + "step": 2960 + }, + { + "epoch": 0.2034492235811461, + "grad_norm": 1.1807277202606201, + "learning_rate": 4.8388791512608336e-05, + "loss": 1.081, + "step": 2961 + }, + { + "epoch": 0.20351793321423664, + "grad_norm": 1.149202823638916, + "learning_rate": 4.838667299986493e-05, + "loss": 1.0502, + "step": 2962 + }, + { + "epoch": 0.2035866428473272, + "grad_norm": 1.1148569583892822, + "learning_rate": 4.838455314170048e-05, + "loss": 1.0041, + "step": 2963 + }, + { + "epoch": 0.20365535248041775, + "grad_norm": 1.1517611742019653, + "learning_rate": 4.838243193823696e-05, + "loss": 1.0203, + "step": 2964 + }, + { + "epoch": 0.20372406211350833, + "grad_norm": 1.163144826889038, + "learning_rate": 4.838030938959638e-05, + "loss": 1.0617, + "step": 2965 + }, + { + "epoch": 0.20379277174659888, + "grad_norm": 1.129289150238037, + "learning_rate": 4.837818549590085e-05, + "loss": 1.0494, + "step": 2966 + }, + { + "epoch": 0.20386148137968943, + "grad_norm": 1.176582932472229, + "learning_rate": 4.837606025727257e-05, + "loss": 1.0404, + "step": 2967 + }, + { + "epoch": 0.20393019101277998, + "grad_norm": 1.1253105401992798, + "learning_rate": 4.83739336738338e-05, + "loss": 1.0053, + "step": 2968 + }, + { + "epoch": 0.20399890064587056, + "grad_norm": 1.1237226724624634, + "learning_rate": 4.837180574570688e-05, + "loss": 1.0124, + "step": 2969 + }, + { + "epoch": 0.20406761027896111, + "grad_norm": 1.2509299516677856, + "learning_rate": 4.836967647301421e-05, + "loss": 1.0291, + "step": 2970 + }, + { + "epoch": 0.20413631991205167, + "grad_norm": 1.2765991687774658, + "learning_rate": 4.836754585587831e-05, + "loss": 1.0312, + "step": 2971 + }, + { + "epoch": 0.20420502954514222, + "grad_norm": 1.095746636390686, + "learning_rate": 4.836541389442175e-05, + "loss": 1.0084, + "step": 2972 + }, + { + "epoch": 0.2042737391782328, + "grad_norm": 1.263426423072815, + "learning_rate": 4.836328058876717e-05, + "loss": 1.0169, + "step": 2973 + }, + { + "epoch": 0.20434244881132335, + "grad_norm": 1.1580357551574707, + "learning_rate": 4.836114593903731e-05, + "loss": 1.026, + "step": 2974 + }, + { + "epoch": 0.2044111584444139, + "grad_norm": 1.1766085624694824, + "learning_rate": 4.835900994535496e-05, + "loss": 1.0536, + "step": 2975 + }, + { + "epoch": 0.20447986807750446, + "grad_norm": 1.1606192588806152, + "learning_rate": 4.835687260784302e-05, + "loss": 1.0636, + "step": 2976 + }, + { + "epoch": 0.20454857771059504, + "grad_norm": 1.1301348209381104, + "learning_rate": 4.835473392662443e-05, + "loss": 1.0234, + "step": 2977 + }, + { + "epoch": 0.2046172873436856, + "grad_norm": 1.1949025392532349, + "learning_rate": 4.835259390182224e-05, + "loss": 1.0304, + "step": 2978 + }, + { + "epoch": 0.20468599697677614, + "grad_norm": 1.03449547290802, + "learning_rate": 4.8350452533559565e-05, + "loss": 1.0091, + "step": 2979 + }, + { + "epoch": 0.2047547066098667, + "grad_norm": 1.2120225429534912, + "learning_rate": 4.834830982195959e-05, + "loss": 1.0328, + "step": 2980 + }, + { + "epoch": 0.20482341624295727, + "grad_norm": 1.2480100393295288, + "learning_rate": 4.834616576714559e-05, + "loss": 1.0261, + "step": 2981 + }, + { + "epoch": 0.20489212587604783, + "grad_norm": 1.092877984046936, + "learning_rate": 4.8344020369240904e-05, + "loss": 1.018, + "step": 2982 + }, + { + "epoch": 0.20496083550913838, + "grad_norm": 1.1245217323303223, + "learning_rate": 4.8341873628368964e-05, + "loss": 0.9781, + "step": 2983 + }, + { + "epoch": 0.20502954514222893, + "grad_norm": 1.1306263208389282, + "learning_rate": 4.833972554465327e-05, + "loss": 1.0043, + "step": 2984 + }, + { + "epoch": 0.2050982547753195, + "grad_norm": 1.157559871673584, + "learning_rate": 4.8337576118217384e-05, + "loss": 1.0016, + "step": 2985 + }, + { + "epoch": 0.20516696440841006, + "grad_norm": 1.1523889303207397, + "learning_rate": 4.833542534918498e-05, + "loss": 1.0597, + "step": 2986 + }, + { + "epoch": 0.20523567404150062, + "grad_norm": 1.1610020399093628, + "learning_rate": 4.833327323767979e-05, + "loss": 1.0169, + "step": 2987 + }, + { + "epoch": 0.20530438367459117, + "grad_norm": 1.1575754880905151, + "learning_rate": 4.8331119783825615e-05, + "loss": 1.0311, + "step": 2988 + }, + { + "epoch": 0.20537309330768175, + "grad_norm": 1.1660561561584473, + "learning_rate": 4.832896498774634e-05, + "loss": 1.0781, + "step": 2989 + }, + { + "epoch": 0.2054418029407723, + "grad_norm": 1.0637627840042114, + "learning_rate": 4.8326808849565936e-05, + "loss": 0.9938, + "step": 2990 + }, + { + "epoch": 0.20551051257386285, + "grad_norm": 1.1113004684448242, + "learning_rate": 4.8324651369408444e-05, + "loss": 1.0078, + "step": 2991 + }, + { + "epoch": 0.2055792222069534, + "grad_norm": 1.141141414642334, + "learning_rate": 4.832249254739798e-05, + "loss": 1.0302, + "step": 2992 + }, + { + "epoch": 0.20564793184004398, + "grad_norm": 1.0615458488464355, + "learning_rate": 4.832033238365874e-05, + "loss": 1.0258, + "step": 2993 + }, + { + "epoch": 0.20571664147313454, + "grad_norm": 1.1848410367965698, + "learning_rate": 4.8318170878315e-05, + "loss": 1.0433, + "step": 2994 + }, + { + "epoch": 0.2057853511062251, + "grad_norm": 1.1447581052780151, + "learning_rate": 4.83160080314911e-05, + "loss": 1.0428, + "step": 2995 + }, + { + "epoch": 0.20585406073931564, + "grad_norm": 1.1212306022644043, + "learning_rate": 4.831384384331147e-05, + "loss": 1.0995, + "step": 2996 + }, + { + "epoch": 0.20592277037240622, + "grad_norm": 1.1352659463882446, + "learning_rate": 4.831167831390063e-05, + "loss": 1.0666, + "step": 2997 + }, + { + "epoch": 0.20599148000549677, + "grad_norm": 1.0611259937286377, + "learning_rate": 4.830951144338315e-05, + "loss": 0.9868, + "step": 2998 + }, + { + "epoch": 0.20606018963858733, + "grad_norm": 1.1632184982299805, + "learning_rate": 4.830734323188369e-05, + "loss": 1.0287, + "step": 2999 + }, + { + "epoch": 0.20612889927167788, + "grad_norm": 1.0761624574661255, + "learning_rate": 4.830517367952697e-05, + "loss": 0.9897, + "step": 3000 + }, + { + "epoch": 0.20619760890476846, + "grad_norm": 1.1177080869674683, + "learning_rate": 4.830300278643784e-05, + "loss": 1.0593, + "step": 3001 + }, + { + "epoch": 0.206266318537859, + "grad_norm": 1.1772363185882568, + "learning_rate": 4.830083055274115e-05, + "loss": 1.0504, + "step": 3002 + }, + { + "epoch": 0.20633502817094956, + "grad_norm": 1.1568341255187988, + "learning_rate": 4.8298656978561896e-05, + "loss": 1.0658, + "step": 3003 + }, + { + "epoch": 0.20640373780404012, + "grad_norm": 1.0953537225723267, + "learning_rate": 4.829648206402511e-05, + "loss": 1.0159, + "step": 3004 + }, + { + "epoch": 0.2064724474371307, + "grad_norm": 1.1347849369049072, + "learning_rate": 4.829430580925591e-05, + "loss": 1.0272, + "step": 3005 + }, + { + "epoch": 0.20654115707022125, + "grad_norm": 1.1372421979904175, + "learning_rate": 4.8292128214379514e-05, + "loss": 1.0206, + "step": 3006 + }, + { + "epoch": 0.2066098667033118, + "grad_norm": 1.1520689725875854, + "learning_rate": 4.828994927952117e-05, + "loss": 0.9769, + "step": 3007 + }, + { + "epoch": 0.20667857633640235, + "grad_norm": 1.199216365814209, + "learning_rate": 4.828776900480625e-05, + "loss": 1.029, + "step": 3008 + }, + { + "epoch": 0.20674728596949293, + "grad_norm": 1.1467456817626953, + "learning_rate": 4.828558739036018e-05, + "loss": 1.0025, + "step": 3009 + }, + { + "epoch": 0.20681599560258349, + "grad_norm": 1.1408166885375977, + "learning_rate": 4.8283404436308464e-05, + "loss": 1.0189, + "step": 3010 + }, + { + "epoch": 0.20688470523567404, + "grad_norm": 1.2053720951080322, + "learning_rate": 4.828122014277668e-05, + "loss": 1.0277, + "step": 3011 + }, + { + "epoch": 0.2069534148687646, + "grad_norm": 1.0819803476333618, + "learning_rate": 4.8279034509890506e-05, + "loss": 1.0111, + "step": 3012 + }, + { + "epoch": 0.20702212450185517, + "grad_norm": 1.1038028001785278, + "learning_rate": 4.827684753777567e-05, + "loss": 1.0047, + "step": 3013 + }, + { + "epoch": 0.20709083413494572, + "grad_norm": 1.1948883533477783, + "learning_rate": 4.827465922655799e-05, + "loss": 1.0464, + "step": 3014 + }, + { + "epoch": 0.20715954376803628, + "grad_norm": 1.198475956916809, + "learning_rate": 4.827246957636335e-05, + "loss": 1.0117, + "step": 3015 + }, + { + "epoch": 0.20722825340112683, + "grad_norm": 1.175293207168579, + "learning_rate": 4.827027858731773e-05, + "loss": 1.0281, + "step": 3016 + }, + { + "epoch": 0.2072969630342174, + "grad_norm": 1.1211017370224, + "learning_rate": 4.8268086259547166e-05, + "loss": 0.9918, + "step": 3017 + }, + { + "epoch": 0.20736567266730796, + "grad_norm": 1.0990709066390991, + "learning_rate": 4.826589259317779e-05, + "loss": 1.0134, + "step": 3018 + }, + { + "epoch": 0.2074343823003985, + "grad_norm": 1.0625267028808594, + "learning_rate": 4.82636975883358e-05, + "loss": 0.9836, + "step": 3019 + }, + { + "epoch": 0.20750309193348906, + "grad_norm": 1.1184723377227783, + "learning_rate": 4.8261501245147464e-05, + "loss": 1.021, + "step": 3020 + }, + { + "epoch": 0.20757180156657964, + "grad_norm": 1.0870373249053955, + "learning_rate": 4.8259303563739155e-05, + "loss": 1.0422, + "step": 3021 + }, + { + "epoch": 0.2076405111996702, + "grad_norm": 1.1491317749023438, + "learning_rate": 4.82571045442373e-05, + "loss": 1.0082, + "step": 3022 + }, + { + "epoch": 0.20770922083276075, + "grad_norm": 1.0781608819961548, + "learning_rate": 4.825490418676838e-05, + "loss": 1.0272, + "step": 3023 + }, + { + "epoch": 0.2077779304658513, + "grad_norm": 1.179448127746582, + "learning_rate": 4.825270249145902e-05, + "loss": 1.0604, + "step": 3024 + }, + { + "epoch": 0.20784664009894188, + "grad_norm": 1.1133766174316406, + "learning_rate": 4.825049945843585e-05, + "loss": 1.0247, + "step": 3025 + }, + { + "epoch": 0.20791534973203243, + "grad_norm": 1.1526366472244263, + "learning_rate": 4.824829508782564e-05, + "loss": 1.0703, + "step": 3026 + }, + { + "epoch": 0.207984059365123, + "grad_norm": 1.1202317476272583, + "learning_rate": 4.8246089379755176e-05, + "loss": 1.0009, + "step": 3027 + }, + { + "epoch": 0.20805276899821354, + "grad_norm": 1.1596745252609253, + "learning_rate": 4.8243882334351367e-05, + "loss": 1.0133, + "step": 3028 + }, + { + "epoch": 0.20812147863130412, + "grad_norm": 1.2655020952224731, + "learning_rate": 4.824167395174117e-05, + "loss": 1.0288, + "step": 3029 + }, + { + "epoch": 0.20819018826439467, + "grad_norm": 1.1491897106170654, + "learning_rate": 4.8239464232051655e-05, + "loss": 1.0264, + "step": 3030 + }, + { + "epoch": 0.20825889789748522, + "grad_norm": 1.2638355493545532, + "learning_rate": 4.823725317540993e-05, + "loss": 1.0621, + "step": 3031 + }, + { + "epoch": 0.20832760753057578, + "grad_norm": 1.155471920967102, + "learning_rate": 4.82350407819432e-05, + "loss": 1.0723, + "step": 3032 + }, + { + "epoch": 0.20839631716366636, + "grad_norm": 1.0799894332885742, + "learning_rate": 4.823282705177874e-05, + "loss": 0.979, + "step": 3033 + }, + { + "epoch": 0.2084650267967569, + "grad_norm": 1.1328576803207397, + "learning_rate": 4.82306119850439e-05, + "loss": 1.0193, + "step": 3034 + }, + { + "epoch": 0.20853373642984746, + "grad_norm": 1.2413586378097534, + "learning_rate": 4.822839558186612e-05, + "loss": 1.014, + "step": 3035 + }, + { + "epoch": 0.208602446062938, + "grad_norm": 1.099092721939087, + "learning_rate": 4.8226177842372904e-05, + "loss": 1.0312, + "step": 3036 + }, + { + "epoch": 0.2086711556960286, + "grad_norm": 1.1605263948440552, + "learning_rate": 4.822395876669185e-05, + "loss": 1.0214, + "step": 3037 + }, + { + "epoch": 0.20873986532911915, + "grad_norm": 1.0977890491485596, + "learning_rate": 4.822173835495059e-05, + "loss": 1.0244, + "step": 3038 + }, + { + "epoch": 0.2088085749622097, + "grad_norm": 1.151098370552063, + "learning_rate": 4.8219516607276895e-05, + "loss": 1.0474, + "step": 3039 + }, + { + "epoch": 0.20887728459530025, + "grad_norm": 1.186753749847412, + "learning_rate": 4.821729352379856e-05, + "loss": 1.0521, + "step": 3040 + }, + { + "epoch": 0.20894599422839083, + "grad_norm": 1.1510847806930542, + "learning_rate": 4.821506910464349e-05, + "loss": 1.0332, + "step": 3041 + }, + { + "epoch": 0.20901470386148138, + "grad_norm": 1.1816688776016235, + "learning_rate": 4.821284334993964e-05, + "loss": 0.9894, + "step": 3042 + }, + { + "epoch": 0.20908341349457193, + "grad_norm": 1.1177700757980347, + "learning_rate": 4.821061625981508e-05, + "loss": 1.0293, + "step": 3043 + }, + { + "epoch": 0.2091521231276625, + "grad_norm": 1.138933777809143, + "learning_rate": 4.820838783439791e-05, + "loss": 1.06, + "step": 3044 + }, + { + "epoch": 0.20922083276075307, + "grad_norm": 1.1809439659118652, + "learning_rate": 4.820615807381634e-05, + "loss": 1.0182, + "step": 3045 + }, + { + "epoch": 0.20928954239384362, + "grad_norm": 1.1097043752670288, + "learning_rate": 4.820392697819864e-05, + "loss": 1.0278, + "step": 3046 + }, + { + "epoch": 0.20935825202693417, + "grad_norm": 1.2235523462295532, + "learning_rate": 4.820169454767318e-05, + "loss": 1.0328, + "step": 3047 + }, + { + "epoch": 0.20942696166002472, + "grad_norm": 1.1064817905426025, + "learning_rate": 4.819946078236838e-05, + "loss": 0.9943, + "step": 3048 + }, + { + "epoch": 0.2094956712931153, + "grad_norm": 1.182559609413147, + "learning_rate": 4.819722568241274e-05, + "loss": 1.024, + "step": 3049 + }, + { + "epoch": 0.20956438092620586, + "grad_norm": 1.1377662420272827, + "learning_rate": 4.819498924793484e-05, + "loss": 1.0339, + "step": 3050 + }, + { + "epoch": 0.2096330905592964, + "grad_norm": 1.1595994234085083, + "learning_rate": 4.819275147906337e-05, + "loss": 1.0695, + "step": 3051 + }, + { + "epoch": 0.20970180019238696, + "grad_norm": 1.074355959892273, + "learning_rate": 4.8190512375927044e-05, + "loss": 1.0371, + "step": 3052 + }, + { + "epoch": 0.20977050982547754, + "grad_norm": 1.0819453001022339, + "learning_rate": 4.818827193865468e-05, + "loss": 1.0216, + "step": 3053 + }, + { + "epoch": 0.2098392194585681, + "grad_norm": 1.171368956565857, + "learning_rate": 4.8186030167375165e-05, + "loss": 1.0036, + "step": 3054 + }, + { + "epoch": 0.20990792909165865, + "grad_norm": 1.1060798168182373, + "learning_rate": 4.818378706221748e-05, + "loss": 0.9939, + "step": 3055 + }, + { + "epoch": 0.2099766387247492, + "grad_norm": 1.173113465309143, + "learning_rate": 4.8181542623310656e-05, + "loss": 1.0316, + "step": 3056 + }, + { + "epoch": 0.21004534835783978, + "grad_norm": 1.1729856729507446, + "learning_rate": 4.8179296850783816e-05, + "loss": 1.021, + "step": 3057 + }, + { + "epoch": 0.21011405799093033, + "grad_norm": 1.0794404745101929, + "learning_rate": 4.817704974476617e-05, + "loss": 0.9797, + "step": 3058 + }, + { + "epoch": 0.21018276762402088, + "grad_norm": 1.090517282485962, + "learning_rate": 4.8174801305386976e-05, + "loss": 1.0072, + "step": 3059 + }, + { + "epoch": 0.21025147725711144, + "grad_norm": 1.2336070537567139, + "learning_rate": 4.81725515327756e-05, + "loss": 1.0254, + "step": 3060 + }, + { + "epoch": 0.21032018689020202, + "grad_norm": 1.1413195133209229, + "learning_rate": 4.817030042706146e-05, + "loss": 1.0141, + "step": 3061 + }, + { + "epoch": 0.21038889652329257, + "grad_norm": 1.1786434650421143, + "learning_rate": 4.816804798837407e-05, + "loss": 0.9819, + "step": 3062 + }, + { + "epoch": 0.21045760615638312, + "grad_norm": 1.1430208683013916, + "learning_rate": 4.8165794216843e-05, + "loss": 1.0529, + "step": 3063 + }, + { + "epoch": 0.21052631578947367, + "grad_norm": 1.1099228858947754, + "learning_rate": 4.816353911259791e-05, + "loss": 0.995, + "step": 3064 + }, + { + "epoch": 0.21059502542256425, + "grad_norm": 1.2377585172653198, + "learning_rate": 4.8161282675768543e-05, + "loss": 0.9863, + "step": 3065 + }, + { + "epoch": 0.2106637350556548, + "grad_norm": 1.1764110326766968, + "learning_rate": 4.815902490648471e-05, + "loss": 0.9826, + "step": 3066 + }, + { + "epoch": 0.21073244468874536, + "grad_norm": 1.1192649602890015, + "learning_rate": 4.8156765804876284e-05, + "loss": 1.0244, + "step": 3067 + }, + { + "epoch": 0.2108011543218359, + "grad_norm": 1.1673181056976318, + "learning_rate": 4.815450537107324e-05, + "loss": 1.0263, + "step": 3068 + }, + { + "epoch": 0.2108698639549265, + "grad_norm": 1.1772756576538086, + "learning_rate": 4.815224360520564e-05, + "loss": 1.1055, + "step": 3069 + }, + { + "epoch": 0.21093857358801704, + "grad_norm": 1.1742030382156372, + "learning_rate": 4.8149980507403564e-05, + "loss": 1.0214, + "step": 3070 + }, + { + "epoch": 0.2110072832211076, + "grad_norm": 1.093216896057129, + "learning_rate": 4.814771607779723e-05, + "loss": 1.0163, + "step": 3071 + }, + { + "epoch": 0.21107599285419815, + "grad_norm": 1.119563102722168, + "learning_rate": 4.8145450316516905e-05, + "loss": 0.9742, + "step": 3072 + }, + { + "epoch": 0.21114470248728873, + "grad_norm": 1.184365153312683, + "learning_rate": 4.8143183223692935e-05, + "loss": 0.9938, + "step": 3073 + }, + { + "epoch": 0.21121341212037928, + "grad_norm": 1.1613339185714722, + "learning_rate": 4.814091479945574e-05, + "loss": 0.9885, + "step": 3074 + }, + { + "epoch": 0.21128212175346983, + "grad_norm": 1.1699374914169312, + "learning_rate": 4.813864504393584e-05, + "loss": 1.0259, + "step": 3075 + }, + { + "epoch": 0.21135083138656038, + "grad_norm": 1.2174092531204224, + "learning_rate": 4.8136373957263784e-05, + "loss": 1.0801, + "step": 3076 + }, + { + "epoch": 0.21141954101965096, + "grad_norm": 1.1764581203460693, + "learning_rate": 4.8134101539570247e-05, + "loss": 1.0193, + "step": 3077 + }, + { + "epoch": 0.21148825065274152, + "grad_norm": 1.1091383695602417, + "learning_rate": 4.8131827790985956e-05, + "loss": 1.01, + "step": 3078 + }, + { + "epoch": 0.21155696028583207, + "grad_norm": 1.1703519821166992, + "learning_rate": 4.812955271164171e-05, + "loss": 1.0175, + "step": 3079 + }, + { + "epoch": 0.21162566991892262, + "grad_norm": 1.0992181301116943, + "learning_rate": 4.812727630166841e-05, + "loss": 1.03, + "step": 3080 + }, + { + "epoch": 0.2116943795520132, + "grad_norm": 1.1348872184753418, + "learning_rate": 4.8124998561197e-05, + "loss": 1.0432, + "step": 3081 + }, + { + "epoch": 0.21176308918510375, + "grad_norm": 1.1647982597351074, + "learning_rate": 4.812271949035852e-05, + "loss": 1.0319, + "step": 3082 + }, + { + "epoch": 0.2118317988181943, + "grad_norm": 1.1235355138778687, + "learning_rate": 4.812043908928409e-05, + "loss": 1.0147, + "step": 3083 + }, + { + "epoch": 0.21190050845128486, + "grad_norm": 1.1889777183532715, + "learning_rate": 4.81181573581049e-05, + "loss": 1.0242, + "step": 3084 + }, + { + "epoch": 0.21196921808437544, + "grad_norm": 1.1930363178253174, + "learning_rate": 4.81158742969522e-05, + "loss": 1.005, + "step": 3085 + }, + { + "epoch": 0.212037927717466, + "grad_norm": 1.2399543523788452, + "learning_rate": 4.811358990595736e-05, + "loss": 1.0326, + "step": 3086 + }, + { + "epoch": 0.21210663735055654, + "grad_norm": 1.0744733810424805, + "learning_rate": 4.811130418525178e-05, + "loss": 0.9991, + "step": 3087 + }, + { + "epoch": 0.2121753469836471, + "grad_norm": 1.066217303276062, + "learning_rate": 4.8109017134966965e-05, + "loss": 0.9993, + "step": 3088 + }, + { + "epoch": 0.21224405661673768, + "grad_norm": 1.1291695833206177, + "learning_rate": 4.810672875523448e-05, + "loss": 1.0289, + "step": 3089 + }, + { + "epoch": 0.21231276624982823, + "grad_norm": 1.108022689819336, + "learning_rate": 4.810443904618598e-05, + "loss": 1.0175, + "step": 3090 + }, + { + "epoch": 0.21238147588291878, + "grad_norm": 1.1180157661437988, + "learning_rate": 4.81021480079532e-05, + "loss": 0.9677, + "step": 3091 + }, + { + "epoch": 0.21245018551600933, + "grad_norm": 1.115347146987915, + "learning_rate": 4.809985564066791e-05, + "loss": 1.0371, + "step": 3092 + }, + { + "epoch": 0.2125188951490999, + "grad_norm": 1.1771798133850098, + "learning_rate": 4.809756194446203e-05, + "loss": 1.0122, + "step": 3093 + }, + { + "epoch": 0.21258760478219046, + "grad_norm": 1.1510835886001587, + "learning_rate": 4.8095266919467485e-05, + "loss": 1.0382, + "step": 3094 + }, + { + "epoch": 0.21265631441528102, + "grad_norm": 1.1000373363494873, + "learning_rate": 4.8092970565816314e-05, + "loss": 1.0302, + "step": 3095 + }, + { + "epoch": 0.21272502404837157, + "grad_norm": 1.159906268119812, + "learning_rate": 4.809067288364063e-05, + "loss": 1.0185, + "step": 3096 + }, + { + "epoch": 0.21279373368146215, + "grad_norm": 1.1418437957763672, + "learning_rate": 4.8088373873072615e-05, + "loss": 1.0084, + "step": 3097 + }, + { + "epoch": 0.2128624433145527, + "grad_norm": 1.120036244392395, + "learning_rate": 4.808607353424452e-05, + "loss": 1.0294, + "step": 3098 + }, + { + "epoch": 0.21293115294764325, + "grad_norm": 1.1418977975845337, + "learning_rate": 4.8083771867288706e-05, + "loss": 1.0138, + "step": 3099 + }, + { + "epoch": 0.2129998625807338, + "grad_norm": 1.1397868394851685, + "learning_rate": 4.8081468872337565e-05, + "loss": 1.014, + "step": 3100 + }, + { + "epoch": 0.2130685722138244, + "grad_norm": 1.1869159936904907, + "learning_rate": 4.807916454952359e-05, + "loss": 0.9849, + "step": 3101 + }, + { + "epoch": 0.21313728184691494, + "grad_norm": 1.247646689414978, + "learning_rate": 4.807685889897936e-05, + "loss": 1.0453, + "step": 3102 + }, + { + "epoch": 0.2132059914800055, + "grad_norm": 1.2072675228118896, + "learning_rate": 4.80745519208375e-05, + "loss": 1.043, + "step": 3103 + }, + { + "epoch": 0.21327470111309604, + "grad_norm": 1.114959955215454, + "learning_rate": 4.807224361523074e-05, + "loss": 1.0114, + "step": 3104 + }, + { + "epoch": 0.21334341074618662, + "grad_norm": 1.0958518981933594, + "learning_rate": 4.806993398229187e-05, + "loss": 1.0212, + "step": 3105 + }, + { + "epoch": 0.21341212037927718, + "grad_norm": 1.0836964845657349, + "learning_rate": 4.8067623022153764e-05, + "loss": 0.995, + "step": 3106 + }, + { + "epoch": 0.21348083001236773, + "grad_norm": 1.2151447534561157, + "learning_rate": 4.8065310734949376e-05, + "loss": 1.0466, + "step": 3107 + }, + { + "epoch": 0.21354953964545828, + "grad_norm": 1.1305865049362183, + "learning_rate": 4.806299712081172e-05, + "loss": 1.0025, + "step": 3108 + }, + { + "epoch": 0.21361824927854886, + "grad_norm": 1.191773533821106, + "learning_rate": 4.806068217987391e-05, + "loss": 1.0213, + "step": 3109 + }, + { + "epoch": 0.2136869589116394, + "grad_norm": 1.1816233396530151, + "learning_rate": 4.80583659122691e-05, + "loss": 0.9964, + "step": 3110 + }, + { + "epoch": 0.21375566854472997, + "grad_norm": 1.111752986907959, + "learning_rate": 4.805604831813057e-05, + "loss": 1.0186, + "step": 3111 + }, + { + "epoch": 0.21382437817782052, + "grad_norm": 1.1746792793273926, + "learning_rate": 4.805372939759164e-05, + "loss": 1.0501, + "step": 3112 + }, + { + "epoch": 0.2138930878109111, + "grad_norm": 1.1741808652877808, + "learning_rate": 4.8051409150785705e-05, + "loss": 1.0347, + "step": 3113 + }, + { + "epoch": 0.21396179744400165, + "grad_norm": 1.0842182636260986, + "learning_rate": 4.804908757784626e-05, + "loss": 1.0155, + "step": 3114 + }, + { + "epoch": 0.2140305070770922, + "grad_norm": 1.1508294343948364, + "learning_rate": 4.8046764678906865e-05, + "loss": 0.9688, + "step": 3115 + }, + { + "epoch": 0.21409921671018275, + "grad_norm": 1.1840550899505615, + "learning_rate": 4.8044440454101146e-05, + "loss": 1.0111, + "step": 3116 + }, + { + "epoch": 0.21416792634327333, + "grad_norm": 1.1793817281723022, + "learning_rate": 4.804211490356282e-05, + "loss": 1.0645, + "step": 3117 + }, + { + "epoch": 0.2142366359763639, + "grad_norm": 1.2108005285263062, + "learning_rate": 4.803978802742567e-05, + "loss": 1.035, + "step": 3118 + }, + { + "epoch": 0.21430534560945444, + "grad_norm": 1.0647437572479248, + "learning_rate": 4.803745982582357e-05, + "loss": 1.0308, + "step": 3119 + }, + { + "epoch": 0.214374055242545, + "grad_norm": 1.173326849937439, + "learning_rate": 4.803513029889045e-05, + "loss": 0.997, + "step": 3120 + }, + { + "epoch": 0.21444276487563557, + "grad_norm": 1.1561483144760132, + "learning_rate": 4.803279944676032e-05, + "loss": 1.0444, + "step": 3121 + }, + { + "epoch": 0.21451147450872612, + "grad_norm": 1.1935383081436157, + "learning_rate": 4.80304672695673e-05, + "loss": 1.0597, + "step": 3122 + }, + { + "epoch": 0.21458018414181668, + "grad_norm": 1.104387879371643, + "learning_rate": 4.802813376744553e-05, + "loss": 0.9892, + "step": 3123 + }, + { + "epoch": 0.21464889377490723, + "grad_norm": 1.129299521446228, + "learning_rate": 4.802579894052926e-05, + "loss": 1.0546, + "step": 3124 + }, + { + "epoch": 0.2147176034079978, + "grad_norm": 1.3284999132156372, + "learning_rate": 4.802346278895282e-05, + "loss": 1.0341, + "step": 3125 + }, + { + "epoch": 0.21478631304108836, + "grad_norm": 1.142834186553955, + "learning_rate": 4.802112531285061e-05, + "loss": 1.0166, + "step": 3126 + }, + { + "epoch": 0.2148550226741789, + "grad_norm": 1.0980380773544312, + "learning_rate": 4.8018786512357094e-05, + "loss": 1.0169, + "step": 3127 + }, + { + "epoch": 0.21492373230726947, + "grad_norm": 1.171985149383545, + "learning_rate": 4.8016446387606826e-05, + "loss": 1.021, + "step": 3128 + }, + { + "epoch": 0.21499244194036005, + "grad_norm": 1.1761832237243652, + "learning_rate": 4.801410493873443e-05, + "loss": 1.0724, + "step": 3129 + }, + { + "epoch": 0.2150611515734506, + "grad_norm": 1.2480534315109253, + "learning_rate": 4.801176216587461e-05, + "loss": 1.0336, + "step": 3130 + }, + { + "epoch": 0.21512986120654115, + "grad_norm": 1.1236897706985474, + "learning_rate": 4.800941806916214e-05, + "loss": 0.992, + "step": 3131 + }, + { + "epoch": 0.2151985708396317, + "grad_norm": 1.092087984085083, + "learning_rate": 4.800707264873189e-05, + "loss": 0.9766, + "step": 3132 + }, + { + "epoch": 0.21526728047272228, + "grad_norm": 1.1256508827209473, + "learning_rate": 4.8004725904718764e-05, + "loss": 0.9917, + "step": 3133 + }, + { + "epoch": 0.21533599010581284, + "grad_norm": 1.2742228507995605, + "learning_rate": 4.80023778372578e-05, + "loss": 1.0854, + "step": 3134 + }, + { + "epoch": 0.2154046997389034, + "grad_norm": 1.156865119934082, + "learning_rate": 4.800002844648405e-05, + "loss": 1.0257, + "step": 3135 + }, + { + "epoch": 0.21547340937199394, + "grad_norm": 1.1459258794784546, + "learning_rate": 4.799767773253269e-05, + "loss": 1.0141, + "step": 3136 + }, + { + "epoch": 0.21554211900508452, + "grad_norm": 1.2157299518585205, + "learning_rate": 4.799532569553895e-05, + "loss": 1.0269, + "step": 3137 + }, + { + "epoch": 0.21561082863817507, + "grad_norm": 1.123813509941101, + "learning_rate": 4.799297233563814e-05, + "loss": 1.0111, + "step": 3138 + }, + { + "epoch": 0.21567953827126563, + "grad_norm": 1.1000994443893433, + "learning_rate": 4.7990617652965665e-05, + "loss": 1.0118, + "step": 3139 + }, + { + "epoch": 0.21574824790435618, + "grad_norm": 1.2246023416519165, + "learning_rate": 4.7988261647656966e-05, + "loss": 1.0263, + "step": 3140 + }, + { + "epoch": 0.21581695753744676, + "grad_norm": 1.1690865755081177, + "learning_rate": 4.79859043198476e-05, + "loss": 0.9925, + "step": 3141 + }, + { + "epoch": 0.2158856671705373, + "grad_norm": 1.1050697565078735, + "learning_rate": 4.798354566967316e-05, + "loss": 1.0438, + "step": 3142 + }, + { + "epoch": 0.21595437680362786, + "grad_norm": 1.071176290512085, + "learning_rate": 4.7981185697269356e-05, + "loss": 1.0009, + "step": 3143 + }, + { + "epoch": 0.21602308643671841, + "grad_norm": 1.3047049045562744, + "learning_rate": 4.797882440277196e-05, + "loss": 1.0555, + "step": 3144 + }, + { + "epoch": 0.216091796069809, + "grad_norm": 1.157649278640747, + "learning_rate": 4.7976461786316796e-05, + "loss": 1.0625, + "step": 3145 + }, + { + "epoch": 0.21616050570289955, + "grad_norm": 1.1516375541687012, + "learning_rate": 4.79740978480398e-05, + "loss": 0.98, + "step": 3146 + }, + { + "epoch": 0.2162292153359901, + "grad_norm": 1.1480772495269775, + "learning_rate": 4.7971732588076965e-05, + "loss": 1.0485, + "step": 3147 + }, + { + "epoch": 0.21629792496908065, + "grad_norm": 1.1362390518188477, + "learning_rate": 4.796936600656435e-05, + "loss": 1.0111, + "step": 3148 + }, + { + "epoch": 0.21636663460217123, + "grad_norm": 1.1726707220077515, + "learning_rate": 4.7966998103638125e-05, + "loss": 1.0279, + "step": 3149 + }, + { + "epoch": 0.21643534423526178, + "grad_norm": 1.1729474067687988, + "learning_rate": 4.79646288794345e-05, + "loss": 0.9851, + "step": 3150 + }, + { + "epoch": 0.21650405386835234, + "grad_norm": 1.0974047183990479, + "learning_rate": 4.796225833408978e-05, + "loss": 1.0365, + "step": 3151 + }, + { + "epoch": 0.2165727635014429, + "grad_norm": 1.0712699890136719, + "learning_rate": 4.795988646774034e-05, + "loss": 0.9993, + "step": 3152 + }, + { + "epoch": 0.21664147313453347, + "grad_norm": 1.2003037929534912, + "learning_rate": 4.7957513280522626e-05, + "loss": 1.033, + "step": 3153 + }, + { + "epoch": 0.21671018276762402, + "grad_norm": 1.0635714530944824, + "learning_rate": 4.795513877257317e-05, + "loss": 1.019, + "step": 3154 + }, + { + "epoch": 0.21677889240071457, + "grad_norm": 1.1575738191604614, + "learning_rate": 4.795276294402858e-05, + "loss": 1.0199, + "step": 3155 + }, + { + "epoch": 0.21684760203380513, + "grad_norm": 1.0975351333618164, + "learning_rate": 4.795038579502553e-05, + "loss": 0.9984, + "step": 3156 + }, + { + "epoch": 0.2169163116668957, + "grad_norm": 1.1902508735656738, + "learning_rate": 4.7948007325700785e-05, + "loss": 1.0189, + "step": 3157 + }, + { + "epoch": 0.21698502129998626, + "grad_norm": 1.1665300130844116, + "learning_rate": 4.7945627536191166e-05, + "loss": 1.0068, + "step": 3158 + }, + { + "epoch": 0.2170537309330768, + "grad_norm": 1.0959250926971436, + "learning_rate": 4.794324642663359e-05, + "loss": 1.0595, + "step": 3159 + }, + { + "epoch": 0.21712244056616736, + "grad_norm": 1.080974817276001, + "learning_rate": 4.794086399716503e-05, + "loss": 1.0332, + "step": 3160 + }, + { + "epoch": 0.21719115019925794, + "grad_norm": 1.2380194664001465, + "learning_rate": 4.793848024792256e-05, + "loss": 1.0301, + "step": 3161 + }, + { + "epoch": 0.2172598598323485, + "grad_norm": 1.1573166847229004, + "learning_rate": 4.793609517904329e-05, + "loss": 1.0057, + "step": 3162 + }, + { + "epoch": 0.21732856946543905, + "grad_norm": 1.1578127145767212, + "learning_rate": 4.793370879066447e-05, + "loss": 1.0036, + "step": 3163 + }, + { + "epoch": 0.21739727909852963, + "grad_norm": 1.1802363395690918, + "learning_rate": 4.793132108292336e-05, + "loss": 1.0161, + "step": 3164 + }, + { + "epoch": 0.21746598873162018, + "grad_norm": 1.155503273010254, + "learning_rate": 4.792893205595733e-05, + "loss": 1.0342, + "step": 3165 + }, + { + "epoch": 0.21753469836471073, + "grad_norm": 1.1501637697219849, + "learning_rate": 4.792654170990382e-05, + "loss": 0.9716, + "step": 3166 + }, + { + "epoch": 0.21760340799780128, + "grad_norm": 1.1331449747085571, + "learning_rate": 4.792415004490034e-05, + "loss": 1.013, + "step": 3167 + }, + { + "epoch": 0.21767211763089186, + "grad_norm": 1.1885665655136108, + "learning_rate": 4.792175706108448e-05, + "loss": 1.018, + "step": 3168 + }, + { + "epoch": 0.21774082726398242, + "grad_norm": 1.1685718297958374, + "learning_rate": 4.791936275859392e-05, + "loss": 1.0555, + "step": 3169 + }, + { + "epoch": 0.21780953689707297, + "grad_norm": 1.196363925933838, + "learning_rate": 4.79169671375664e-05, + "loss": 1.0363, + "step": 3170 + }, + { + "epoch": 0.21787824653016352, + "grad_norm": 1.0797984600067139, + "learning_rate": 4.791457019813973e-05, + "loss": 0.9983, + "step": 3171 + }, + { + "epoch": 0.2179469561632541, + "grad_norm": 1.1710695028305054, + "learning_rate": 4.791217194045181e-05, + "loss": 0.9992, + "step": 3172 + }, + { + "epoch": 0.21801566579634465, + "grad_norm": 1.1916300058364868, + "learning_rate": 4.7909772364640606e-05, + "loss": 1.0198, + "step": 3173 + }, + { + "epoch": 0.2180843754294352, + "grad_norm": 1.1788506507873535, + "learning_rate": 4.7907371470844164e-05, + "loss": 1.0268, + "step": 3174 + }, + { + "epoch": 0.21815308506252576, + "grad_norm": 1.0758082866668701, + "learning_rate": 4.790496925920062e-05, + "loss": 0.9748, + "step": 3175 + }, + { + "epoch": 0.21822179469561634, + "grad_norm": 1.0601714849472046, + "learning_rate": 4.790256572984815e-05, + "loss": 0.98, + "step": 3176 + }, + { + "epoch": 0.2182905043287069, + "grad_norm": 1.2459365129470825, + "learning_rate": 4.7900160882925036e-05, + "loss": 1.0527, + "step": 3177 + }, + { + "epoch": 0.21835921396179744, + "grad_norm": 1.3127707242965698, + "learning_rate": 4.7897754718569635e-05, + "loss": 1.018, + "step": 3178 + }, + { + "epoch": 0.218427923594888, + "grad_norm": 1.1335817575454712, + "learning_rate": 4.7895347236920364e-05, + "loss": 1.0144, + "step": 3179 + }, + { + "epoch": 0.21849663322797858, + "grad_norm": 1.165594458580017, + "learning_rate": 4.7892938438115724e-05, + "loss": 1.0206, + "step": 3180 + }, + { + "epoch": 0.21856534286106913, + "grad_norm": 1.052741289138794, + "learning_rate": 4.7890528322294297e-05, + "loss": 0.9844, + "step": 3181 + }, + { + "epoch": 0.21863405249415968, + "grad_norm": 1.1185566186904907, + "learning_rate": 4.788811688959474e-05, + "loss": 1.0262, + "step": 3182 + }, + { + "epoch": 0.21870276212725023, + "grad_norm": 1.244473934173584, + "learning_rate": 4.788570414015576e-05, + "loss": 1.0473, + "step": 3183 + }, + { + "epoch": 0.2187714717603408, + "grad_norm": 1.128389596939087, + "learning_rate": 4.7883290074116185e-05, + "loss": 1.0072, + "step": 3184 + }, + { + "epoch": 0.21884018139343137, + "grad_norm": 1.2067015171051025, + "learning_rate": 4.788087469161488e-05, + "loss": 0.9287, + "step": 3185 + }, + { + "epoch": 0.21890889102652192, + "grad_norm": 1.0836436748504639, + "learning_rate": 4.787845799279081e-05, + "loss": 1.0462, + "step": 3186 + }, + { + "epoch": 0.21897760065961247, + "grad_norm": 1.1340205669403076, + "learning_rate": 4.7876039977783e-05, + "loss": 0.9931, + "step": 3187 + }, + { + "epoch": 0.21904631029270305, + "grad_norm": 1.1320629119873047, + "learning_rate": 4.7873620646730557e-05, + "loss": 0.9897, + "step": 3188 + }, + { + "epoch": 0.2191150199257936, + "grad_norm": 1.1445207595825195, + "learning_rate": 4.7871199999772667e-05, + "loss": 1.0005, + "step": 3189 + }, + { + "epoch": 0.21918372955888415, + "grad_norm": 1.116685390472412, + "learning_rate": 4.7868778037048575e-05, + "loss": 1.0031, + "step": 3190 + }, + { + "epoch": 0.2192524391919747, + "grad_norm": 1.0851261615753174, + "learning_rate": 4.7866354758697644e-05, + "loss": 1.0218, + "step": 3191 + }, + { + "epoch": 0.2193211488250653, + "grad_norm": 1.102839708328247, + "learning_rate": 4.786393016485925e-05, + "loss": 0.9908, + "step": 3192 + }, + { + "epoch": 0.21938985845815584, + "grad_norm": 1.1198983192443848, + "learning_rate": 4.7861504255672904e-05, + "loss": 1.0255, + "step": 3193 + }, + { + "epoch": 0.2194585680912464, + "grad_norm": 1.1323689222335815, + "learning_rate": 4.785907703127816e-05, + "loss": 1.0552, + "step": 3194 + }, + { + "epoch": 0.21952727772433694, + "grad_norm": 1.2012600898742676, + "learning_rate": 4.785664849181465e-05, + "loss": 1.0365, + "step": 3195 + }, + { + "epoch": 0.21959598735742752, + "grad_norm": 1.0865232944488525, + "learning_rate": 4.785421863742209e-05, + "loss": 0.9788, + "step": 3196 + }, + { + "epoch": 0.21966469699051808, + "grad_norm": 1.17808997631073, + "learning_rate": 4.785178746824026e-05, + "loss": 1.0315, + "step": 3197 + }, + { + "epoch": 0.21973340662360863, + "grad_norm": 1.174149513244629, + "learning_rate": 4.784935498440904e-05, + "loss": 1.0448, + "step": 3198 + }, + { + "epoch": 0.21980211625669918, + "grad_norm": 1.097187876701355, + "learning_rate": 4.784692118606835e-05, + "loss": 1.0005, + "step": 3199 + }, + { + "epoch": 0.21987082588978976, + "grad_norm": 1.1362563371658325, + "learning_rate": 4.784448607335822e-05, + "loss": 1.0516, + "step": 3200 + }, + { + "epoch": 0.21993953552288031, + "grad_norm": 1.1554080247879028, + "learning_rate": 4.784204964641873e-05, + "loss": 1.0388, + "step": 3201 + }, + { + "epoch": 0.22000824515597087, + "grad_norm": 1.177138090133667, + "learning_rate": 4.783961190539006e-05, + "loss": 1.0209, + "step": 3202 + }, + { + "epoch": 0.22007695478906142, + "grad_norm": 1.1966804265975952, + "learning_rate": 4.783717285041244e-05, + "loss": 1.0149, + "step": 3203 + }, + { + "epoch": 0.220145664422152, + "grad_norm": 1.1454241275787354, + "learning_rate": 4.7834732481626185e-05, + "loss": 1.0153, + "step": 3204 + }, + { + "epoch": 0.22021437405524255, + "grad_norm": 1.2469284534454346, + "learning_rate": 4.7832290799171695e-05, + "loss": 1.004, + "step": 3205 + }, + { + "epoch": 0.2202830836883331, + "grad_norm": 1.150277853012085, + "learning_rate": 4.782984780318944e-05, + "loss": 1.003, + "step": 3206 + }, + { + "epoch": 0.22035179332142366, + "grad_norm": 1.147217869758606, + "learning_rate": 4.782740349381996e-05, + "loss": 0.9704, + "step": 3207 + }, + { + "epoch": 0.22042050295451424, + "grad_norm": 1.1518118381500244, + "learning_rate": 4.782495787120388e-05, + "loss": 1.007, + "step": 3208 + }, + { + "epoch": 0.2204892125876048, + "grad_norm": 1.3332154750823975, + "learning_rate": 4.782251093548188e-05, + "loss": 1.0195, + "step": 3209 + }, + { + "epoch": 0.22055792222069534, + "grad_norm": 1.0944666862487793, + "learning_rate": 4.782006268679475e-05, + "loss": 0.9964, + "step": 3210 + }, + { + "epoch": 0.2206266318537859, + "grad_norm": 1.2364284992218018, + "learning_rate": 4.7817613125283325e-05, + "loss": 1.0489, + "step": 3211 + }, + { + "epoch": 0.22069534148687647, + "grad_norm": 1.135055422782898, + "learning_rate": 4.781516225108853e-05, + "loss": 0.9927, + "step": 3212 + }, + { + "epoch": 0.22076405111996703, + "grad_norm": 1.1333591938018799, + "learning_rate": 4.781271006435136e-05, + "loss": 1.0434, + "step": 3213 + }, + { + "epoch": 0.22083276075305758, + "grad_norm": 1.2254260778427124, + "learning_rate": 4.7810256565212893e-05, + "loss": 1.0451, + "step": 3214 + }, + { + "epoch": 0.22090147038614813, + "grad_norm": 1.1829962730407715, + "learning_rate": 4.780780175381428e-05, + "loss": 1.0381, + "step": 3215 + }, + { + "epoch": 0.2209701800192387, + "grad_norm": 1.1520094871520996, + "learning_rate": 4.780534563029673e-05, + "loss": 1.0074, + "step": 3216 + }, + { + "epoch": 0.22103888965232926, + "grad_norm": 1.149367094039917, + "learning_rate": 4.7802888194801555e-05, + "loss": 1.0225, + "step": 3217 + }, + { + "epoch": 0.22110759928541981, + "grad_norm": 1.2218292951583862, + "learning_rate": 4.780042944747012e-05, + "loss": 1.0478, + "step": 3218 + }, + { + "epoch": 0.22117630891851037, + "grad_norm": 1.164026141166687, + "learning_rate": 4.779796938844389e-05, + "loss": 1.006, + "step": 3219 + }, + { + "epoch": 0.22124501855160095, + "grad_norm": 1.1485296487808228, + "learning_rate": 4.7795508017864374e-05, + "loss": 1.011, + "step": 3220 + }, + { + "epoch": 0.2213137281846915, + "grad_norm": 1.1299031972885132, + "learning_rate": 4.779304533587318e-05, + "loss": 1.0229, + "step": 3221 + }, + { + "epoch": 0.22138243781778205, + "grad_norm": 1.083575963973999, + "learning_rate": 4.7790581342612e-05, + "loss": 0.9927, + "step": 3222 + }, + { + "epoch": 0.2214511474508726, + "grad_norm": 1.1331331729888916, + "learning_rate": 4.778811603822255e-05, + "loss": 1.0481, + "step": 3223 + }, + { + "epoch": 0.22151985708396318, + "grad_norm": 1.1445324420928955, + "learning_rate": 4.778564942284669e-05, + "loss": 1.0138, + "step": 3224 + }, + { + "epoch": 0.22158856671705374, + "grad_norm": 1.1769160032272339, + "learning_rate": 4.7783181496626315e-05, + "loss": 1.0192, + "step": 3225 + }, + { + "epoch": 0.2216572763501443, + "grad_norm": 1.0747736692428589, + "learning_rate": 4.77807122597034e-05, + "loss": 0.9962, + "step": 3226 + }, + { + "epoch": 0.22172598598323484, + "grad_norm": 1.156272530555725, + "learning_rate": 4.777824171221999e-05, + "loss": 1.0313, + "step": 3227 + }, + { + "epoch": 0.22179469561632542, + "grad_norm": 1.1339908838272095, + "learning_rate": 4.777576985431823e-05, + "loss": 1.0626, + "step": 3228 + }, + { + "epoch": 0.22186340524941597, + "grad_norm": 1.22012197971344, + "learning_rate": 4.7773296686140314e-05, + "loss": 1.0396, + "step": 3229 + }, + { + "epoch": 0.22193211488250653, + "grad_norm": 1.1058682203292847, + "learning_rate": 4.7770822207828526e-05, + "loss": 1.03, + "step": 3230 + }, + { + "epoch": 0.22200082451559708, + "grad_norm": 1.2369714975357056, + "learning_rate": 4.776834641952522e-05, + "loss": 1.0813, + "step": 3231 + }, + { + "epoch": 0.22206953414868766, + "grad_norm": 1.1876083612442017, + "learning_rate": 4.7765869321372836e-05, + "loss": 1.0574, + "step": 3232 + }, + { + "epoch": 0.2221382437817782, + "grad_norm": 1.0794286727905273, + "learning_rate": 4.776339091351386e-05, + "loss": 1.0015, + "step": 3233 + }, + { + "epoch": 0.22220695341486876, + "grad_norm": 1.1502162218093872, + "learning_rate": 4.776091119609089e-05, + "loss": 1.0446, + "step": 3234 + }, + { + "epoch": 0.22227566304795932, + "grad_norm": 1.1282237768173218, + "learning_rate": 4.775843016924657e-05, + "loss": 1.0204, + "step": 3235 + }, + { + "epoch": 0.2223443726810499, + "grad_norm": 1.2108439207077026, + "learning_rate": 4.7755947833123646e-05, + "loss": 1.0522, + "step": 3236 + }, + { + "epoch": 0.22241308231414045, + "grad_norm": 1.191667079925537, + "learning_rate": 4.775346418786492e-05, + "loss": 1.0339, + "step": 3237 + }, + { + "epoch": 0.222481791947231, + "grad_norm": 1.1290026903152466, + "learning_rate": 4.7750979233613266e-05, + "loss": 1.0593, + "step": 3238 + }, + { + "epoch": 0.22255050158032155, + "grad_norm": 1.1858729124069214, + "learning_rate": 4.774849297051165e-05, + "loss": 1.0561, + "step": 3239 + }, + { + "epoch": 0.22261921121341213, + "grad_norm": 1.1507152318954468, + "learning_rate": 4.774600539870311e-05, + "loss": 1.0464, + "step": 3240 + }, + { + "epoch": 0.22268792084650268, + "grad_norm": 1.2032849788665771, + "learning_rate": 4.7743516518330745e-05, + "loss": 1.0205, + "step": 3241 + }, + { + "epoch": 0.22275663047959324, + "grad_norm": 1.0966898202896118, + "learning_rate": 4.774102632953774e-05, + "loss": 1.037, + "step": 3242 + }, + { + "epoch": 0.2228253401126838, + "grad_norm": 1.116865634918213, + "learning_rate": 4.773853483246735e-05, + "loss": 0.9618, + "step": 3243 + }, + { + "epoch": 0.22289404974577437, + "grad_norm": 1.174569010734558, + "learning_rate": 4.773604202726293e-05, + "loss": 1.0135, + "step": 3244 + }, + { + "epoch": 0.22296275937886492, + "grad_norm": 1.1436381340026855, + "learning_rate": 4.773354791406787e-05, + "loss": 1.0261, + "step": 3245 + }, + { + "epoch": 0.22303146901195547, + "grad_norm": 1.1677361726760864, + "learning_rate": 4.7731052493025654e-05, + "loss": 1.0117, + "step": 3246 + }, + { + "epoch": 0.22310017864504603, + "grad_norm": 1.0908598899841309, + "learning_rate": 4.7728555764279854e-05, + "loss": 1.0242, + "step": 3247 + }, + { + "epoch": 0.2231688882781366, + "grad_norm": 1.1311904191970825, + "learning_rate": 4.77260577279741e-05, + "loss": 1.0124, + "step": 3248 + }, + { + "epoch": 0.22323759791122716, + "grad_norm": 1.0437517166137695, + "learning_rate": 4.7723558384252095e-05, + "loss": 0.9906, + "step": 3249 + }, + { + "epoch": 0.2233063075443177, + "grad_norm": 1.2501864433288574, + "learning_rate": 4.7721057733257634e-05, + "loss": 1.0543, + "step": 3250 + }, + { + "epoch": 0.22337501717740826, + "grad_norm": 1.1345535516738892, + "learning_rate": 4.771855577513458e-05, + "loss": 1.0183, + "step": 3251 + }, + { + "epoch": 0.22344372681049884, + "grad_norm": 1.157211422920227, + "learning_rate": 4.771605251002685e-05, + "loss": 1.0166, + "step": 3252 + }, + { + "epoch": 0.2235124364435894, + "grad_norm": 1.1283575296401978, + "learning_rate": 4.7713547938078484e-05, + "loss": 0.9759, + "step": 3253 + }, + { + "epoch": 0.22358114607667995, + "grad_norm": 1.1328320503234863, + "learning_rate": 4.771104205943355e-05, + "loss": 0.9986, + "step": 3254 + }, + { + "epoch": 0.2236498557097705, + "grad_norm": 1.1290028095245361, + "learning_rate": 4.770853487423622e-05, + "loss": 1.0179, + "step": 3255 + }, + { + "epoch": 0.22371856534286108, + "grad_norm": 1.1934947967529297, + "learning_rate": 4.770602638263072e-05, + "loss": 1.0048, + "step": 3256 + }, + { + "epoch": 0.22378727497595163, + "grad_norm": 1.1162176132202148, + "learning_rate": 4.7703516584761364e-05, + "loss": 1.0481, + "step": 3257 + }, + { + "epoch": 0.22385598460904219, + "grad_norm": 1.0806913375854492, + "learning_rate": 4.770100548077255e-05, + "loss": 0.9938, + "step": 3258 + }, + { + "epoch": 0.22392469424213274, + "grad_norm": 1.1243282556533813, + "learning_rate": 4.769849307080873e-05, + "loss": 1.0243, + "step": 3259 + }, + { + "epoch": 0.22399340387522332, + "grad_norm": 1.157659649848938, + "learning_rate": 4.769597935501444e-05, + "loss": 1.0295, + "step": 3260 + }, + { + "epoch": 0.22406211350831387, + "grad_norm": 1.1219416856765747, + "learning_rate": 4.7693464333534295e-05, + "loss": 0.9844, + "step": 3261 + }, + { + "epoch": 0.22413082314140442, + "grad_norm": 1.1286581754684448, + "learning_rate": 4.7690948006512984e-05, + "loss": 1.0144, + "step": 3262 + }, + { + "epoch": 0.22419953277449498, + "grad_norm": 1.1293777227401733, + "learning_rate": 4.768843037409527e-05, + "loss": 0.9974, + "step": 3263 + }, + { + "epoch": 0.22426824240758556, + "grad_norm": 1.0896445512771606, + "learning_rate": 4.7685911436425995e-05, + "loss": 1.0042, + "step": 3264 + }, + { + "epoch": 0.2243369520406761, + "grad_norm": 1.1748597621917725, + "learning_rate": 4.768339119365006e-05, + "loss": 1.0148, + "step": 3265 + }, + { + "epoch": 0.22440566167376666, + "grad_norm": 1.2038769721984863, + "learning_rate": 4.768086964591248e-05, + "loss": 1.0447, + "step": 3266 + }, + { + "epoch": 0.2244743713068572, + "grad_norm": 1.121333360671997, + "learning_rate": 4.767834679335828e-05, + "loss": 1.0218, + "step": 3267 + }, + { + "epoch": 0.2245430809399478, + "grad_norm": 1.1308776140213013, + "learning_rate": 4.7675822636132615e-05, + "loss": 1.0169, + "step": 3268 + }, + { + "epoch": 0.22461179057303834, + "grad_norm": 1.1283351182937622, + "learning_rate": 4.767329717438071e-05, + "loss": 1.004, + "step": 3269 + }, + { + "epoch": 0.2246805002061289, + "grad_norm": 1.098657488822937, + "learning_rate": 4.767077040824783e-05, + "loss": 1.0162, + "step": 3270 + }, + { + "epoch": 0.22474920983921945, + "grad_norm": 1.05263352394104, + "learning_rate": 4.766824233787936e-05, + "loss": 1.0269, + "step": 3271 + }, + { + "epoch": 0.22481791947231003, + "grad_norm": 1.0879019498825073, + "learning_rate": 4.766571296342073e-05, + "loss": 1.0103, + "step": 3272 + }, + { + "epoch": 0.22488662910540058, + "grad_norm": 1.0851633548736572, + "learning_rate": 4.766318228501746e-05, + "loss": 1.0088, + "step": 3273 + }, + { + "epoch": 0.22495533873849113, + "grad_norm": 1.109468698501587, + "learning_rate": 4.766065030281512e-05, + "loss": 1.0306, + "step": 3274 + }, + { + "epoch": 0.2250240483715817, + "grad_norm": 1.1224440336227417, + "learning_rate": 4.765811701695939e-05, + "loss": 1.0362, + "step": 3275 + }, + { + "epoch": 0.22509275800467227, + "grad_norm": 1.1149373054504395, + "learning_rate": 4.7655582427596004e-05, + "loss": 1.0026, + "step": 3276 + }, + { + "epoch": 0.22516146763776282, + "grad_norm": 1.159224033355713, + "learning_rate": 4.765304653487077e-05, + "loss": 1.0579, + "step": 3277 + }, + { + "epoch": 0.22523017727085337, + "grad_norm": 1.159855604171753, + "learning_rate": 4.765050933892958e-05, + "loss": 1.0559, + "step": 3278 + }, + { + "epoch": 0.22529888690394392, + "grad_norm": 1.2146260738372803, + "learning_rate": 4.764797083991841e-05, + "loss": 1.0656, + "step": 3279 + }, + { + "epoch": 0.2253675965370345, + "grad_norm": 1.11897611618042, + "learning_rate": 4.7645431037983276e-05, + "loss": 1.0237, + "step": 3280 + }, + { + "epoch": 0.22543630617012506, + "grad_norm": 1.222722053527832, + "learning_rate": 4.764288993327031e-05, + "loss": 1.0535, + "step": 3281 + }, + { + "epoch": 0.2255050158032156, + "grad_norm": 1.1685045957565308, + "learning_rate": 4.764034752592569e-05, + "loss": 1.0276, + "step": 3282 + }, + { + "epoch": 0.22557372543630616, + "grad_norm": 1.1500009298324585, + "learning_rate": 4.763780381609568e-05, + "loss": 1.0294, + "step": 3283 + }, + { + "epoch": 0.22564243506939674, + "grad_norm": 1.1845608949661255, + "learning_rate": 4.763525880392662e-05, + "loss": 1.0187, + "step": 3284 + }, + { + "epoch": 0.2257111447024873, + "grad_norm": 1.1076769828796387, + "learning_rate": 4.7632712489564926e-05, + "loss": 1.0096, + "step": 3285 + }, + { + "epoch": 0.22577985433557785, + "grad_norm": 1.0610506534576416, + "learning_rate": 4.763016487315707e-05, + "loss": 0.9804, + "step": 3286 + }, + { + "epoch": 0.2258485639686684, + "grad_norm": 1.0821211338043213, + "learning_rate": 4.7627615954849635e-05, + "loss": 0.976, + "step": 3287 + }, + { + "epoch": 0.22591727360175898, + "grad_norm": 1.0953022241592407, + "learning_rate": 4.7625065734789254e-05, + "loss": 1.0028, + "step": 3288 + }, + { + "epoch": 0.22598598323484953, + "grad_norm": 1.163912296295166, + "learning_rate": 4.7622514213122636e-05, + "loss": 1.0131, + "step": 3289 + }, + { + "epoch": 0.22605469286794008, + "grad_norm": 1.1579954624176025, + "learning_rate": 4.761996138999656e-05, + "loss": 1.038, + "step": 3290 + }, + { + "epoch": 0.22612340250103063, + "grad_norm": 1.2296828031539917, + "learning_rate": 4.761740726555791e-05, + "loss": 1.0346, + "step": 3291 + }, + { + "epoch": 0.22619211213412121, + "grad_norm": 1.1198142766952515, + "learning_rate": 4.76148518399536e-05, + "loss": 1.0208, + "step": 3292 + }, + { + "epoch": 0.22626082176721177, + "grad_norm": 1.1462255716323853, + "learning_rate": 4.761229511333066e-05, + "loss": 1.0606, + "step": 3293 + }, + { + "epoch": 0.22632953140030232, + "grad_norm": 1.0823636054992676, + "learning_rate": 4.760973708583617e-05, + "loss": 1.0201, + "step": 3294 + }, + { + "epoch": 0.22639824103339287, + "grad_norm": 1.1247832775115967, + "learning_rate": 4.760717775761729e-05, + "loss": 0.971, + "step": 3295 + }, + { + "epoch": 0.22646695066648345, + "grad_norm": 1.1773936748504639, + "learning_rate": 4.760461712882126e-05, + "loss": 1.0164, + "step": 3296 + }, + { + "epoch": 0.226535660299574, + "grad_norm": 1.2096678018569946, + "learning_rate": 4.760205519959538e-05, + "loss": 1.051, + "step": 3297 + }, + { + "epoch": 0.22660436993266456, + "grad_norm": 1.1092524528503418, + "learning_rate": 4.7599491970087054e-05, + "loss": 0.9577, + "step": 3298 + }, + { + "epoch": 0.2266730795657551, + "grad_norm": 1.0959184169769287, + "learning_rate": 4.759692744044373e-05, + "loss": 0.9763, + "step": 3299 + }, + { + "epoch": 0.2267417891988457, + "grad_norm": 1.138248324394226, + "learning_rate": 4.759436161081295e-05, + "loss": 1.0299, + "step": 3300 + }, + { + "epoch": 0.22681049883193624, + "grad_norm": 1.111831545829773, + "learning_rate": 4.759179448134232e-05, + "loss": 1.0023, + "step": 3301 + }, + { + "epoch": 0.2268792084650268, + "grad_norm": 1.253044843673706, + "learning_rate": 4.758922605217954e-05, + "loss": 1.051, + "step": 3302 + }, + { + "epoch": 0.22694791809811735, + "grad_norm": 1.134804129600525, + "learning_rate": 4.7586656323472344e-05, + "loss": 1.0186, + "step": 3303 + }, + { + "epoch": 0.22701662773120793, + "grad_norm": 1.1415207386016846, + "learning_rate": 4.758408529536859e-05, + "loss": 1.0207, + "step": 3304 + }, + { + "epoch": 0.22708533736429848, + "grad_norm": 1.1245133876800537, + "learning_rate": 4.758151296801618e-05, + "loss": 1.0294, + "step": 3305 + }, + { + "epoch": 0.22715404699738903, + "grad_norm": 1.1455057859420776, + "learning_rate": 4.7578939341563095e-05, + "loss": 0.9988, + "step": 3306 + }, + { + "epoch": 0.22722275663047958, + "grad_norm": 1.1596249341964722, + "learning_rate": 4.757636441615739e-05, + "loss": 1.0546, + "step": 3307 + }, + { + "epoch": 0.22729146626357016, + "grad_norm": 1.1042977571487427, + "learning_rate": 4.757378819194722e-05, + "loss": 1.0507, + "step": 3308 + }, + { + "epoch": 0.22736017589666072, + "grad_norm": 1.2806419134140015, + "learning_rate": 4.757121066908077e-05, + "loss": 1.0505, + "step": 3309 + }, + { + "epoch": 0.22742888552975127, + "grad_norm": 1.1435866355895996, + "learning_rate": 4.7568631847706334e-05, + "loss": 1.0379, + "step": 3310 + }, + { + "epoch": 0.22749759516284182, + "grad_norm": 1.127596139907837, + "learning_rate": 4.756605172797227e-05, + "loss": 1.064, + "step": 3311 + }, + { + "epoch": 0.2275663047959324, + "grad_norm": 1.150359869003296, + "learning_rate": 4.7563470310027006e-05, + "loss": 1.038, + "step": 3312 + }, + { + "epoch": 0.22763501442902295, + "grad_norm": 1.1244453191757202, + "learning_rate": 4.7560887594019054e-05, + "loss": 0.9853, + "step": 3313 + }, + { + "epoch": 0.2277037240621135, + "grad_norm": 1.1048247814178467, + "learning_rate": 4.7558303580096995e-05, + "loss": 1.0452, + "step": 3314 + }, + { + "epoch": 0.22777243369520406, + "grad_norm": 1.2627230882644653, + "learning_rate": 4.7555718268409486e-05, + "loss": 1.0535, + "step": 3315 + }, + { + "epoch": 0.22784114332829464, + "grad_norm": 1.1936025619506836, + "learning_rate": 4.7553131659105266e-05, + "loss": 1.0485, + "step": 3316 + }, + { + "epoch": 0.2279098529613852, + "grad_norm": 1.1373018026351929, + "learning_rate": 4.755054375233312e-05, + "loss": 1.0292, + "step": 3317 + }, + { + "epoch": 0.22797856259447574, + "grad_norm": 1.1747688055038452, + "learning_rate": 4.7547954548241953e-05, + "loss": 1.0643, + "step": 3318 + }, + { + "epoch": 0.2280472722275663, + "grad_norm": 1.069026231765747, + "learning_rate": 4.75453640469807e-05, + "loss": 0.9851, + "step": 3319 + }, + { + "epoch": 0.22811598186065687, + "grad_norm": 1.110581398010254, + "learning_rate": 4.7542772248698406e-05, + "loss": 0.998, + "step": 3320 + }, + { + "epoch": 0.22818469149374743, + "grad_norm": 1.1464874744415283, + "learning_rate": 4.754017915354416e-05, + "loss": 1.0172, + "step": 3321 + }, + { + "epoch": 0.22825340112683798, + "grad_norm": 1.1533128023147583, + "learning_rate": 4.7537584761667164e-05, + "loss": 1.0249, + "step": 3322 + }, + { + "epoch": 0.22832211075992853, + "grad_norm": 1.0610815286636353, + "learning_rate": 4.753498907321665e-05, + "loss": 1.0212, + "step": 3323 + }, + { + "epoch": 0.2283908203930191, + "grad_norm": 1.207985520362854, + "learning_rate": 4.753239208834196e-05, + "loss": 1.0034, + "step": 3324 + }, + { + "epoch": 0.22845953002610966, + "grad_norm": 1.1361886262893677, + "learning_rate": 4.752979380719249e-05, + "loss": 1.0386, + "step": 3325 + }, + { + "epoch": 0.22852823965920022, + "grad_norm": 1.1489421129226685, + "learning_rate": 4.752719422991771e-05, + "loss": 1.0724, + "step": 3326 + }, + { + "epoch": 0.22859694929229077, + "grad_norm": 1.1816935539245605, + "learning_rate": 4.75245933566672e-05, + "loss": 1.0041, + "step": 3327 + }, + { + "epoch": 0.22866565892538135, + "grad_norm": 1.1821033954620361, + "learning_rate": 4.7521991187590556e-05, + "loss": 1.0398, + "step": 3328 + }, + { + "epoch": 0.2287343685584719, + "grad_norm": 1.1020628213882446, + "learning_rate": 4.7519387722837496e-05, + "loss": 1.0109, + "step": 3329 + }, + { + "epoch": 0.22880307819156245, + "grad_norm": 1.1392911672592163, + "learning_rate": 4.751678296255779e-05, + "loss": 0.9771, + "step": 3330 + }, + { + "epoch": 0.228871787824653, + "grad_norm": 1.1691656112670898, + "learning_rate": 4.7514176906901284e-05, + "loss": 0.9768, + "step": 3331 + }, + { + "epoch": 0.22894049745774359, + "grad_norm": 1.1972426176071167, + "learning_rate": 4.7511569556017915e-05, + "loss": 1.0028, + "step": 3332 + }, + { + "epoch": 0.22900920709083414, + "grad_norm": 1.037060260772705, + "learning_rate": 4.750896091005767e-05, + "loss": 1.0138, + "step": 3333 + }, + { + "epoch": 0.2290779167239247, + "grad_norm": 1.1646255254745483, + "learning_rate": 4.750635096917063e-05, + "loss": 1.0189, + "step": 3334 + }, + { + "epoch": 0.22914662635701524, + "grad_norm": 1.0955933332443237, + "learning_rate": 4.750373973350695e-05, + "loss": 1.0327, + "step": 3335 + }, + { + "epoch": 0.22921533599010582, + "grad_norm": 1.0895107984542847, + "learning_rate": 4.750112720321683e-05, + "loss": 1.0155, + "step": 3336 + }, + { + "epoch": 0.22928404562319638, + "grad_norm": 1.0784705877304077, + "learning_rate": 4.7498513378450595e-05, + "loss": 0.9805, + "step": 3337 + }, + { + "epoch": 0.22935275525628693, + "grad_norm": 1.2070258855819702, + "learning_rate": 4.7495898259358597e-05, + "loss": 1.0625, + "step": 3338 + }, + { + "epoch": 0.22942146488937748, + "grad_norm": 1.180679202079773, + "learning_rate": 4.749328184609129e-05, + "loss": 1.0468, + "step": 3339 + }, + { + "epoch": 0.22949017452246806, + "grad_norm": 1.1550397872924805, + "learning_rate": 4.7490664138799194e-05, + "loss": 1.0386, + "step": 3340 + }, + { + "epoch": 0.2295588841555586, + "grad_norm": 1.1426230669021606, + "learning_rate": 4.74880451376329e-05, + "loss": 0.9931, + "step": 3341 + }, + { + "epoch": 0.22962759378864916, + "grad_norm": 1.0945508480072021, + "learning_rate": 4.748542484274308e-05, + "loss": 0.9813, + "step": 3342 + }, + { + "epoch": 0.22969630342173972, + "grad_norm": 1.1748844385147095, + "learning_rate": 4.7482803254280484e-05, + "loss": 1.0314, + "step": 3343 + }, + { + "epoch": 0.2297650130548303, + "grad_norm": 1.189815640449524, + "learning_rate": 4.748018037239592e-05, + "loss": 1.0314, + "step": 3344 + }, + { + "epoch": 0.22983372268792085, + "grad_norm": 1.1495659351348877, + "learning_rate": 4.747755619724029e-05, + "loss": 0.9909, + "step": 3345 + }, + { + "epoch": 0.2299024323210114, + "grad_norm": 1.1831938028335571, + "learning_rate": 4.747493072896456e-05, + "loss": 1.0521, + "step": 3346 + }, + { + "epoch": 0.22997114195410195, + "grad_norm": 1.1350034475326538, + "learning_rate": 4.7472303967719765e-05, + "loss": 0.9903, + "step": 3347 + }, + { + "epoch": 0.23003985158719253, + "grad_norm": 1.176360845565796, + "learning_rate": 4.746967591365702e-05, + "loss": 1.0453, + "step": 3348 + }, + { + "epoch": 0.2301085612202831, + "grad_norm": 1.141459584236145, + "learning_rate": 4.746704656692753e-05, + "loss": 0.986, + "step": 3349 + }, + { + "epoch": 0.23017727085337364, + "grad_norm": 1.2204965353012085, + "learning_rate": 4.7464415927682546e-05, + "loss": 1.0454, + "step": 3350 + }, + { + "epoch": 0.2302459804864642, + "grad_norm": 1.1276541948318481, + "learning_rate": 4.74617839960734e-05, + "loss": 1.0375, + "step": 3351 + }, + { + "epoch": 0.23031469011955477, + "grad_norm": 1.0448012351989746, + "learning_rate": 4.745915077225154e-05, + "loss": 0.9809, + "step": 3352 + }, + { + "epoch": 0.23038339975264532, + "grad_norm": 1.1680525541305542, + "learning_rate": 4.745651625636841e-05, + "loss": 1.0299, + "step": 3353 + }, + { + "epoch": 0.23045210938573588, + "grad_norm": 1.1151220798492432, + "learning_rate": 4.74538804485756e-05, + "loss": 1.0214, + "step": 3354 + }, + { + "epoch": 0.23052081901882643, + "grad_norm": 1.189978003501892, + "learning_rate": 4.745124334902474e-05, + "loss": 1.0381, + "step": 3355 + }, + { + "epoch": 0.230589528651917, + "grad_norm": 1.1485155820846558, + "learning_rate": 4.7448604957867536e-05, + "loss": 1.0273, + "step": 3356 + }, + { + "epoch": 0.23065823828500756, + "grad_norm": 1.161517858505249, + "learning_rate": 4.7445965275255774e-05, + "loss": 1.0712, + "step": 3357 + }, + { + "epoch": 0.2307269479180981, + "grad_norm": 1.0985502004623413, + "learning_rate": 4.744332430134132e-05, + "loss": 1.0185, + "step": 3358 + }, + { + "epoch": 0.23079565755118867, + "grad_norm": 1.1654293537139893, + "learning_rate": 4.744068203627611e-05, + "loss": 1.0091, + "step": 3359 + }, + { + "epoch": 0.23086436718427925, + "grad_norm": 1.1390657424926758, + "learning_rate": 4.7438038480212135e-05, + "loss": 1.03, + "step": 3360 + }, + { + "epoch": 0.2309330768173698, + "grad_norm": 1.1290026903152466, + "learning_rate": 4.743539363330149e-05, + "loss": 1.0617, + "step": 3361 + }, + { + "epoch": 0.23100178645046035, + "grad_norm": 1.1878429651260376, + "learning_rate": 4.743274749569633e-05, + "loss": 1.0311, + "step": 3362 + }, + { + "epoch": 0.2310704960835509, + "grad_norm": 1.180542230606079, + "learning_rate": 4.743010006754888e-05, + "loss": 1.0278, + "step": 3363 + }, + { + "epoch": 0.23113920571664148, + "grad_norm": 1.1523452997207642, + "learning_rate": 4.7427451349011454e-05, + "loss": 0.99, + "step": 3364 + }, + { + "epoch": 0.23120791534973203, + "grad_norm": 1.1059411764144897, + "learning_rate": 4.742480134023643e-05, + "loss": 0.9957, + "step": 3365 + }, + { + "epoch": 0.2312766249828226, + "grad_norm": 1.1783188581466675, + "learning_rate": 4.742215004137624e-05, + "loss": 1.0832, + "step": 3366 + }, + { + "epoch": 0.23134533461591314, + "grad_norm": 1.1253081560134888, + "learning_rate": 4.7419497452583453e-05, + "loss": 0.9897, + "step": 3367 + }, + { + "epoch": 0.23141404424900372, + "grad_norm": 1.1291029453277588, + "learning_rate": 4.741684357401064e-05, + "loss": 1.0533, + "step": 3368 + }, + { + "epoch": 0.23148275388209427, + "grad_norm": 1.1689512729644775, + "learning_rate": 4.741418840581048e-05, + "loss": 1.016, + "step": 3369 + }, + { + "epoch": 0.23155146351518482, + "grad_norm": 1.0916651487350464, + "learning_rate": 4.741153194813573e-05, + "loss": 1.0023, + "step": 3370 + }, + { + "epoch": 0.23162017314827538, + "grad_norm": 1.0594606399536133, + "learning_rate": 4.740887420113921e-05, + "loss": 1.0165, + "step": 3371 + }, + { + "epoch": 0.23168888278136596, + "grad_norm": 1.1326566934585571, + "learning_rate": 4.740621516497382e-05, + "loss": 1.0415, + "step": 3372 + }, + { + "epoch": 0.2317575924144565, + "grad_norm": 1.185113549232483, + "learning_rate": 4.740355483979254e-05, + "loss": 1.0092, + "step": 3373 + }, + { + "epoch": 0.23182630204754706, + "grad_norm": 1.069326639175415, + "learning_rate": 4.740089322574841e-05, + "loss": 0.9823, + "step": 3374 + }, + { + "epoch": 0.2318950116806376, + "grad_norm": 1.0736101865768433, + "learning_rate": 4.739823032299454e-05, + "loss": 1.0176, + "step": 3375 + }, + { + "epoch": 0.2319637213137282, + "grad_norm": 1.2618082761764526, + "learning_rate": 4.739556613168415e-05, + "loss": 1.0551, + "step": 3376 + }, + { + "epoch": 0.23203243094681875, + "grad_norm": 1.1997977495193481, + "learning_rate": 4.739290065197049e-05, + "loss": 0.9837, + "step": 3377 + }, + { + "epoch": 0.2321011405799093, + "grad_norm": 1.171280860900879, + "learning_rate": 4.739023388400691e-05, + "loss": 1.0264, + "step": 3378 + }, + { + "epoch": 0.23216985021299985, + "grad_norm": 1.1332341432571411, + "learning_rate": 4.738756582794682e-05, + "loss": 0.9983, + "step": 3379 + }, + { + "epoch": 0.23223855984609043, + "grad_norm": 1.2016035318374634, + "learning_rate": 4.738489648394373e-05, + "loss": 0.9964, + "step": 3380 + }, + { + "epoch": 0.23230726947918098, + "grad_norm": 1.2123304605484009, + "learning_rate": 4.7382225852151185e-05, + "loss": 1.0429, + "step": 3381 + }, + { + "epoch": 0.23237597911227154, + "grad_norm": 1.100932002067566, + "learning_rate": 4.737955393272284e-05, + "loss": 1.0159, + "step": 3382 + }, + { + "epoch": 0.2324446887453621, + "grad_norm": 1.2727307081222534, + "learning_rate": 4.73768807258124e-05, + "loss": 1.0362, + "step": 3383 + }, + { + "epoch": 0.23251339837845267, + "grad_norm": 1.1806437969207764, + "learning_rate": 4.737420623157366e-05, + "loss": 0.9919, + "step": 3384 + }, + { + "epoch": 0.23258210801154322, + "grad_norm": 1.1803326606750488, + "learning_rate": 4.737153045016047e-05, + "loss": 1.0399, + "step": 3385 + }, + { + "epoch": 0.23265081764463377, + "grad_norm": 1.151228904724121, + "learning_rate": 4.736885338172677e-05, + "loss": 1.0258, + "step": 3386 + }, + { + "epoch": 0.23271952727772433, + "grad_norm": 1.1296710968017578, + "learning_rate": 4.7366175026426587e-05, + "loss": 1.0143, + "step": 3387 + }, + { + "epoch": 0.2327882369108149, + "grad_norm": 1.24259352684021, + "learning_rate": 4.736349538441399e-05, + "loss": 1.0307, + "step": 3388 + }, + { + "epoch": 0.23285694654390546, + "grad_norm": 1.1160824298858643, + "learning_rate": 4.7360814455843134e-05, + "loss": 1.0381, + "step": 3389 + }, + { + "epoch": 0.232925656176996, + "grad_norm": 1.1595182418823242, + "learning_rate": 4.7358132240868256e-05, + "loss": 1.0426, + "step": 3390 + }, + { + "epoch": 0.23299436581008656, + "grad_norm": 1.1708194017410278, + "learning_rate": 4.735544873964367e-05, + "loss": 1.0166, + "step": 3391 + }, + { + "epoch": 0.23306307544317714, + "grad_norm": 1.1713849306106567, + "learning_rate": 4.735276395232374e-05, + "loss": 1.0351, + "step": 3392 + }, + { + "epoch": 0.2331317850762677, + "grad_norm": 1.0950297117233276, + "learning_rate": 4.735007787906294e-05, + "loss": 1.0208, + "step": 3393 + }, + { + "epoch": 0.23320049470935825, + "grad_norm": 1.19597589969635, + "learning_rate": 4.734739052001578e-05, + "loss": 1.0311, + "step": 3394 + }, + { + "epoch": 0.2332692043424488, + "grad_norm": 1.1202747821807861, + "learning_rate": 4.734470187533686e-05, + "loss": 1.0115, + "step": 3395 + }, + { + "epoch": 0.23333791397553938, + "grad_norm": 1.10140860080719, + "learning_rate": 4.7342011945180876e-05, + "loss": 0.9983, + "step": 3396 + }, + { + "epoch": 0.23340662360862993, + "grad_norm": 1.318435788154602, + "learning_rate": 4.733932072970257e-05, + "loss": 1.0647, + "step": 3397 + }, + { + "epoch": 0.23347533324172048, + "grad_norm": 1.161156415939331, + "learning_rate": 4.733662822905676e-05, + "loss": 0.9896, + "step": 3398 + }, + { + "epoch": 0.23354404287481104, + "grad_norm": 1.1992530822753906, + "learning_rate": 4.733393444339834e-05, + "loss": 1.0711, + "step": 3399 + }, + { + "epoch": 0.23361275250790162, + "grad_norm": 1.1098456382751465, + "learning_rate": 4.73312393728823e-05, + "loss": 0.9995, + "step": 3400 + }, + { + "epoch": 0.23368146214099217, + "grad_norm": 1.1422414779663086, + "learning_rate": 4.7328543017663676e-05, + "loss": 1.016, + "step": 3401 + }, + { + "epoch": 0.23375017177408272, + "grad_norm": 1.219099998474121, + "learning_rate": 4.732584537789758e-05, + "loss": 1.0142, + "step": 3402 + }, + { + "epoch": 0.23381888140717327, + "grad_norm": 1.16793692111969, + "learning_rate": 4.732314645373921e-05, + "loss": 1.0452, + "step": 3403 + }, + { + "epoch": 0.23388759104026385, + "grad_norm": 1.214918613433838, + "learning_rate": 4.732044624534384e-05, + "loss": 0.9953, + "step": 3404 + }, + { + "epoch": 0.2339563006733544, + "grad_norm": 1.200857162475586, + "learning_rate": 4.7317744752866814e-05, + "loss": 1.0449, + "step": 3405 + }, + { + "epoch": 0.23402501030644496, + "grad_norm": 1.1950905323028564, + "learning_rate": 4.731504197646353e-05, + "loss": 1.0025, + "step": 3406 + }, + { + "epoch": 0.2340937199395355, + "grad_norm": 1.178755760192871, + "learning_rate": 4.731233791628949e-05, + "loss": 1.0414, + "step": 3407 + }, + { + "epoch": 0.2341624295726261, + "grad_norm": 1.1843129396438599, + "learning_rate": 4.730963257250026e-05, + "loss": 1.0318, + "step": 3408 + }, + { + "epoch": 0.23423113920571664, + "grad_norm": 1.1444075107574463, + "learning_rate": 4.730692594525146e-05, + "loss": 1.0389, + "step": 3409 + }, + { + "epoch": 0.2342998488388072, + "grad_norm": 1.1752268075942993, + "learning_rate": 4.730421803469882e-05, + "loss": 1.0199, + "step": 3410 + }, + { + "epoch": 0.23436855847189775, + "grad_norm": 1.1322426795959473, + "learning_rate": 4.730150884099811e-05, + "loss": 1.0189, + "step": 3411 + }, + { + "epoch": 0.23443726810498833, + "grad_norm": 1.1669723987579346, + "learning_rate": 4.7298798364305194e-05, + "loss": 1.0097, + "step": 3412 + }, + { + "epoch": 0.23450597773807888, + "grad_norm": 1.2233357429504395, + "learning_rate": 4.729608660477601e-05, + "loss": 1.0226, + "step": 3413 + }, + { + "epoch": 0.23457468737116943, + "grad_norm": 1.0708606243133545, + "learning_rate": 4.7293373562566564e-05, + "loss": 1.007, + "step": 3414 + }, + { + "epoch": 0.23464339700425998, + "grad_norm": 1.1095895767211914, + "learning_rate": 4.7290659237832925e-05, + "loss": 1.0291, + "step": 3415 + }, + { + "epoch": 0.23471210663735056, + "grad_norm": 1.1031001806259155, + "learning_rate": 4.728794363073125e-05, + "loss": 0.9979, + "step": 3416 + }, + { + "epoch": 0.23478081627044112, + "grad_norm": 1.106162667274475, + "learning_rate": 4.728522674141776e-05, + "loss": 1.0064, + "step": 3417 + }, + { + "epoch": 0.23484952590353167, + "grad_norm": 1.1228348016738892, + "learning_rate": 4.7282508570048764e-05, + "loss": 1.0339, + "step": 3418 + }, + { + "epoch": 0.23491823553662222, + "grad_norm": 1.1749321222305298, + "learning_rate": 4.727978911678065e-05, + "loss": 1.0175, + "step": 3419 + }, + { + "epoch": 0.2349869451697128, + "grad_norm": 1.1430530548095703, + "learning_rate": 4.7277068381769846e-05, + "loss": 1.0093, + "step": 3420 + }, + { + "epoch": 0.23505565480280335, + "grad_norm": 1.1576958894729614, + "learning_rate": 4.727434636517288e-05, + "loss": 0.9827, + "step": 3421 + }, + { + "epoch": 0.2351243644358939, + "grad_norm": 1.1838417053222656, + "learning_rate": 4.727162306714635e-05, + "loss": 1.0064, + "step": 3422 + }, + { + "epoch": 0.23519307406898446, + "grad_norm": 1.1111273765563965, + "learning_rate": 4.7268898487846924e-05, + "loss": 0.9893, + "step": 3423 + }, + { + "epoch": 0.23526178370207504, + "grad_norm": 1.1445159912109375, + "learning_rate": 4.726617262743134e-05, + "loss": 1.0424, + "step": 3424 + }, + { + "epoch": 0.2353304933351656, + "grad_norm": 1.1157315969467163, + "learning_rate": 4.7263445486056435e-05, + "loss": 1.0117, + "step": 3425 + }, + { + "epoch": 0.23539920296825614, + "grad_norm": 1.1123884916305542, + "learning_rate": 4.726071706387907e-05, + "loss": 1.0091, + "step": 3426 + }, + { + "epoch": 0.2354679126013467, + "grad_norm": 1.1457291841506958, + "learning_rate": 4.7257987361056235e-05, + "loss": 1.0368, + "step": 3427 + }, + { + "epoch": 0.23553662223443728, + "grad_norm": 1.1903389692306519, + "learning_rate": 4.725525637774496e-05, + "loss": 0.9747, + "step": 3428 + }, + { + "epoch": 0.23560533186752783, + "grad_norm": 1.1070842742919922, + "learning_rate": 4.725252411410235e-05, + "loss": 1.0323, + "step": 3429 + }, + { + "epoch": 0.23567404150061838, + "grad_norm": 1.1644798517227173, + "learning_rate": 4.72497905702856e-05, + "loss": 1.0179, + "step": 3430 + }, + { + "epoch": 0.23574275113370893, + "grad_norm": 1.0899325609207153, + "learning_rate": 4.724705574645196e-05, + "loss": 0.9864, + "step": 3431 + }, + { + "epoch": 0.2358114607667995, + "grad_norm": 1.1422884464263916, + "learning_rate": 4.724431964275877e-05, + "loss": 1.005, + "step": 3432 + }, + { + "epoch": 0.23588017039989007, + "grad_norm": 1.1026530265808105, + "learning_rate": 4.7241582259363436e-05, + "loss": 0.989, + "step": 3433 + }, + { + "epoch": 0.23594888003298062, + "grad_norm": 1.0988291501998901, + "learning_rate": 4.723884359642344e-05, + "loss": 1.0241, + "step": 3434 + }, + { + "epoch": 0.23601758966607117, + "grad_norm": 1.1088424921035767, + "learning_rate": 4.7236103654096326e-05, + "loss": 0.9924, + "step": 3435 + }, + { + "epoch": 0.23608629929916175, + "grad_norm": 1.1632977724075317, + "learning_rate": 4.723336243253973e-05, + "loss": 1.0408, + "step": 3436 + }, + { + "epoch": 0.2361550089322523, + "grad_norm": 1.11815345287323, + "learning_rate": 4.723061993191135e-05, + "loss": 1.0271, + "step": 3437 + }, + { + "epoch": 0.23622371856534286, + "grad_norm": 1.1921499967575073, + "learning_rate": 4.722787615236895e-05, + "loss": 1.0502, + "step": 3438 + }, + { + "epoch": 0.2362924281984334, + "grad_norm": 1.2654168605804443, + "learning_rate": 4.7225131094070396e-05, + "loss": 1.0435, + "step": 3439 + }, + { + "epoch": 0.236361137831524, + "grad_norm": 1.2608063220977783, + "learning_rate": 4.72223847571736e-05, + "loss": 1.0233, + "step": 3440 + }, + { + "epoch": 0.23642984746461454, + "grad_norm": 1.1156870126724243, + "learning_rate": 4.721963714183656e-05, + "loss": 0.973, + "step": 3441 + }, + { + "epoch": 0.2364985570977051, + "grad_norm": 1.1044732332229614, + "learning_rate": 4.721688824821735e-05, + "loss": 0.9834, + "step": 3442 + }, + { + "epoch": 0.23656726673079564, + "grad_norm": 1.1069051027297974, + "learning_rate": 4.721413807647409e-05, + "loss": 1.0146, + "step": 3443 + }, + { + "epoch": 0.23663597636388622, + "grad_norm": 1.1015149354934692, + "learning_rate": 4.7211386626765026e-05, + "loss": 1.0073, + "step": 3444 + }, + { + "epoch": 0.23670468599697678, + "grad_norm": 1.0958709716796875, + "learning_rate": 4.720863389924842e-05, + "loss": 1.0129, + "step": 3445 + }, + { + "epoch": 0.23677339563006733, + "grad_norm": 1.1267751455307007, + "learning_rate": 4.720587989408266e-05, + "loss": 1.0045, + "step": 3446 + }, + { + "epoch": 0.23684210526315788, + "grad_norm": 1.0895271301269531, + "learning_rate": 4.720312461142616e-05, + "loss": 1.0371, + "step": 3447 + }, + { + "epoch": 0.23691081489624846, + "grad_norm": 1.1594698429107666, + "learning_rate": 4.720036805143744e-05, + "loss": 0.9992, + "step": 3448 + }, + { + "epoch": 0.23697952452933901, + "grad_norm": 1.1214191913604736, + "learning_rate": 4.719761021427508e-05, + "loss": 0.9925, + "step": 3449 + }, + { + "epoch": 0.23704823416242957, + "grad_norm": 1.1397995948791504, + "learning_rate": 4.719485110009775e-05, + "loss": 1.0087, + "step": 3450 + }, + { + "epoch": 0.23711694379552012, + "grad_norm": 1.1275653839111328, + "learning_rate": 4.719209070906416e-05, + "loss": 0.9646, + "step": 3451 + }, + { + "epoch": 0.2371856534286107, + "grad_norm": 1.1409411430358887, + "learning_rate": 4.718932904133312e-05, + "loss": 1.0537, + "step": 3452 + }, + { + "epoch": 0.23725436306170125, + "grad_norm": 1.195157527923584, + "learning_rate": 4.718656609706352e-05, + "loss": 1.0484, + "step": 3453 + }, + { + "epoch": 0.2373230726947918, + "grad_norm": 1.1316587924957275, + "learning_rate": 4.7183801876414294e-05, + "loss": 0.9638, + "step": 3454 + }, + { + "epoch": 0.23739178232788236, + "grad_norm": 1.1232661008834839, + "learning_rate": 4.718103637954447e-05, + "loss": 1.0259, + "step": 3455 + }, + { + "epoch": 0.23746049196097294, + "grad_norm": 1.1389747858047485, + "learning_rate": 4.717826960661315e-05, + "loss": 0.9955, + "step": 3456 + }, + { + "epoch": 0.2375292015940635, + "grad_norm": 1.1659549474716187, + "learning_rate": 4.717550155777951e-05, + "loss": 1.0372, + "step": 3457 + }, + { + "epoch": 0.23759791122715404, + "grad_norm": 1.1677463054656982, + "learning_rate": 4.717273223320278e-05, + "loss": 1.006, + "step": 3458 + }, + { + "epoch": 0.2376666208602446, + "grad_norm": 1.123334527015686, + "learning_rate": 4.7169961633042274e-05, + "loss": 1.0604, + "step": 3459 + }, + { + "epoch": 0.23773533049333517, + "grad_norm": 1.1368478536605835, + "learning_rate": 4.7167189757457405e-05, + "loss": 1.0317, + "step": 3460 + }, + { + "epoch": 0.23780404012642573, + "grad_norm": 1.1178184747695923, + "learning_rate": 4.716441660660762e-05, + "loss": 1.0099, + "step": 3461 + }, + { + "epoch": 0.23787274975951628, + "grad_norm": 1.1214712858200073, + "learning_rate": 4.7161642180652464e-05, + "loss": 0.9634, + "step": 3462 + }, + { + "epoch": 0.23794145939260683, + "grad_norm": 1.1865304708480835, + "learning_rate": 4.7158866479751554e-05, + "loss": 1.0405, + "step": 3463 + }, + { + "epoch": 0.2380101690256974, + "grad_norm": 1.0964176654815674, + "learning_rate": 4.715608950406456e-05, + "loss": 0.9924, + "step": 3464 + }, + { + "epoch": 0.23807887865878796, + "grad_norm": 1.1473205089569092, + "learning_rate": 4.715331125375124e-05, + "loss": 1.0117, + "step": 3465 + }, + { + "epoch": 0.23814758829187851, + "grad_norm": 1.1916953325271606, + "learning_rate": 4.7150531728971435e-05, + "loss": 1.0255, + "step": 3466 + }, + { + "epoch": 0.23821629792496907, + "grad_norm": 1.0699963569641113, + "learning_rate": 4.714775092988505e-05, + "loss": 1.0123, + "step": 3467 + }, + { + "epoch": 0.23828500755805965, + "grad_norm": 1.1273651123046875, + "learning_rate": 4.714496885665205e-05, + "loss": 1.0223, + "step": 3468 + }, + { + "epoch": 0.2383537171911502, + "grad_norm": 1.2608393430709839, + "learning_rate": 4.7142185509432506e-05, + "loss": 1.0418, + "step": 3469 + }, + { + "epoch": 0.23842242682424075, + "grad_norm": 1.1837735176086426, + "learning_rate": 4.7139400888386524e-05, + "loss": 1.0388, + "step": 3470 + }, + { + "epoch": 0.2384911364573313, + "grad_norm": 1.1119495630264282, + "learning_rate": 4.713661499367431e-05, + "loss": 0.9843, + "step": 3471 + }, + { + "epoch": 0.23855984609042188, + "grad_norm": 1.1836422681808472, + "learning_rate": 4.713382782545613e-05, + "loss": 1.0308, + "step": 3472 + }, + { + "epoch": 0.23862855572351244, + "grad_norm": 1.1455488204956055, + "learning_rate": 4.7131039383892326e-05, + "loss": 1.0225, + "step": 3473 + }, + { + "epoch": 0.238697265356603, + "grad_norm": 1.1152423620224, + "learning_rate": 4.712824966914333e-05, + "loss": 1.0166, + "step": 3474 + }, + { + "epoch": 0.23876597498969354, + "grad_norm": 1.1509424448013306, + "learning_rate": 4.712545868136961e-05, + "loss": 0.9989, + "step": 3475 + }, + { + "epoch": 0.23883468462278412, + "grad_norm": 1.2360386848449707, + "learning_rate": 4.7122666420731764e-05, + "loss": 1.0432, + "step": 3476 + }, + { + "epoch": 0.23890339425587467, + "grad_norm": 1.163522481918335, + "learning_rate": 4.7119872887390394e-05, + "loss": 1.0626, + "step": 3477 + }, + { + "epoch": 0.23897210388896523, + "grad_norm": 1.1312577724456787, + "learning_rate": 4.711707808150623e-05, + "loss": 1.0465, + "step": 3478 + }, + { + "epoch": 0.23904081352205578, + "grad_norm": 1.0834884643554688, + "learning_rate": 4.711428200324004e-05, + "loss": 1.0134, + "step": 3479 + }, + { + "epoch": 0.23910952315514636, + "grad_norm": 1.0966874361038208, + "learning_rate": 4.7111484652752694e-05, + "loss": 1.0306, + "step": 3480 + }, + { + "epoch": 0.2391782327882369, + "grad_norm": 1.1224082708358765, + "learning_rate": 4.710868603020512e-05, + "loss": 0.9954, + "step": 3481 + }, + { + "epoch": 0.23924694242132746, + "grad_norm": 1.1690632104873657, + "learning_rate": 4.7105886135758315e-05, + "loss": 0.9962, + "step": 3482 + }, + { + "epoch": 0.23931565205441804, + "grad_norm": 1.1375782489776611, + "learning_rate": 4.710308496957336e-05, + "loss": 1.0329, + "step": 3483 + }, + { + "epoch": 0.2393843616875086, + "grad_norm": 1.1552517414093018, + "learning_rate": 4.7100282531811416e-05, + "loss": 1.0176, + "step": 3484 + }, + { + "epoch": 0.23945307132059915, + "grad_norm": 1.122229814529419, + "learning_rate": 4.7097478822633676e-05, + "loss": 1.0209, + "step": 3485 + }, + { + "epoch": 0.2395217809536897, + "grad_norm": 1.1433213949203491, + "learning_rate": 4.7094673842201466e-05, + "loss": 1.0143, + "step": 3486 + }, + { + "epoch": 0.23959049058678028, + "grad_norm": 1.0533760786056519, + "learning_rate": 4.7091867590676134e-05, + "loss": 1.0212, + "step": 3487 + }, + { + "epoch": 0.23965920021987083, + "grad_norm": 1.2042303085327148, + "learning_rate": 4.7089060068219126e-05, + "loss": 0.9841, + "step": 3488 + }, + { + "epoch": 0.23972790985296138, + "grad_norm": 1.1289857625961304, + "learning_rate": 4.708625127499197e-05, + "loss": 0.9906, + "step": 3489 + }, + { + "epoch": 0.23979661948605194, + "grad_norm": 1.1087746620178223, + "learning_rate": 4.708344121115624e-05, + "loss": 1.0201, + "step": 3490 + }, + { + "epoch": 0.23986532911914252, + "grad_norm": 1.098288893699646, + "learning_rate": 4.70806298768736e-05, + "loss": 1.0283, + "step": 3491 + }, + { + "epoch": 0.23993403875223307, + "grad_norm": 1.1618781089782715, + "learning_rate": 4.707781727230579e-05, + "loss": 0.9986, + "step": 3492 + }, + { + "epoch": 0.24000274838532362, + "grad_norm": 1.1511929035186768, + "learning_rate": 4.707500339761461e-05, + "loss": 1.0091, + "step": 3493 + }, + { + "epoch": 0.24007145801841417, + "grad_norm": 1.1471246480941772, + "learning_rate": 4.707218825296196e-05, + "loss": 1.0064, + "step": 3494 + }, + { + "epoch": 0.24014016765150475, + "grad_norm": 1.203221082687378, + "learning_rate": 4.7069371838509764e-05, + "loss": 1.0463, + "step": 3495 + }, + { + "epoch": 0.2402088772845953, + "grad_norm": 1.1506860256195068, + "learning_rate": 4.7066554154420064e-05, + "loss": 0.9786, + "step": 3496 + }, + { + "epoch": 0.24027758691768586, + "grad_norm": 1.1906689405441284, + "learning_rate": 4.706373520085496e-05, + "loss": 1.0076, + "step": 3497 + }, + { + "epoch": 0.2403462965507764, + "grad_norm": 1.1411652565002441, + "learning_rate": 4.7060914977976625e-05, + "loss": 0.9767, + "step": 3498 + }, + { + "epoch": 0.240415006183867, + "grad_norm": 1.0928215980529785, + "learning_rate": 4.70580934859473e-05, + "loss": 1.0016, + "step": 3499 + }, + { + "epoch": 0.24048371581695754, + "grad_norm": 1.2238017320632935, + "learning_rate": 4.70552707249293e-05, + "loss": 0.9999, + "step": 3500 + }, + { + "epoch": 0.2405524254500481, + "grad_norm": 1.1149200201034546, + "learning_rate": 4.705244669508503e-05, + "loss": 0.9889, + "step": 3501 + }, + { + "epoch": 0.24062113508313865, + "grad_norm": 1.055656909942627, + "learning_rate": 4.7049621396576945e-05, + "loss": 0.9912, + "step": 3502 + }, + { + "epoch": 0.24068984471622923, + "grad_norm": 1.121321678161621, + "learning_rate": 4.7046794829567584e-05, + "loss": 1.0362, + "step": 3503 + }, + { + "epoch": 0.24075855434931978, + "grad_norm": 1.1550426483154297, + "learning_rate": 4.7043966994219565e-05, + "loss": 1.0597, + "step": 3504 + }, + { + "epoch": 0.24082726398241033, + "grad_norm": 1.1211438179016113, + "learning_rate": 4.7041137890695565e-05, + "loss": 1.0108, + "step": 3505 + }, + { + "epoch": 0.24089597361550089, + "grad_norm": 1.2229881286621094, + "learning_rate": 4.703830751915833e-05, + "loss": 1.0184, + "step": 3506 + }, + { + "epoch": 0.24096468324859147, + "grad_norm": 1.1442443132400513, + "learning_rate": 4.7035475879770715e-05, + "loss": 1.0342, + "step": 3507 + }, + { + "epoch": 0.24103339288168202, + "grad_norm": 1.0980870723724365, + "learning_rate": 4.703264297269561e-05, + "loss": 1.0504, + "step": 3508 + }, + { + "epoch": 0.24110210251477257, + "grad_norm": 1.091330885887146, + "learning_rate": 4.7029808798095975e-05, + "loss": 0.9936, + "step": 3509 + }, + { + "epoch": 0.24117081214786312, + "grad_norm": 1.1397145986557007, + "learning_rate": 4.702697335613488e-05, + "loss": 1.0339, + "step": 3510 + }, + { + "epoch": 0.2412395217809537, + "grad_norm": 1.2264565229415894, + "learning_rate": 4.7024136646975435e-05, + "loss": 1.0377, + "step": 3511 + }, + { + "epoch": 0.24130823141404426, + "grad_norm": 1.151474118232727, + "learning_rate": 4.7021298670780845e-05, + "loss": 1.031, + "step": 3512 + }, + { + "epoch": 0.2413769410471348, + "grad_norm": 1.1745935678482056, + "learning_rate": 4.701845942771436e-05, + "loss": 0.9825, + "step": 3513 + }, + { + "epoch": 0.24144565068022536, + "grad_norm": 1.1904493570327759, + "learning_rate": 4.7015618917939336e-05, + "loss": 1.0052, + "step": 3514 + }, + { + "epoch": 0.24151436031331594, + "grad_norm": 1.133575439453125, + "learning_rate": 4.701277714161918e-05, + "loss": 1.053, + "step": 3515 + }, + { + "epoch": 0.2415830699464065, + "grad_norm": 1.1105564832687378, + "learning_rate": 4.700993409891737e-05, + "loss": 0.9884, + "step": 3516 + }, + { + "epoch": 0.24165177957949704, + "grad_norm": 1.1061204671859741, + "learning_rate": 4.7007089789997485e-05, + "loss": 0.9831, + "step": 3517 + }, + { + "epoch": 0.2417204892125876, + "grad_norm": 1.1206663846969604, + "learning_rate": 4.700424421502313e-05, + "loss": 0.9719, + "step": 3518 + }, + { + "epoch": 0.24178919884567818, + "grad_norm": 1.170954942703247, + "learning_rate": 4.7001397374158026e-05, + "loss": 0.9948, + "step": 3519 + }, + { + "epoch": 0.24185790847876873, + "grad_norm": 1.0986850261688232, + "learning_rate": 4.6998549267565946e-05, + "loss": 1.009, + "step": 3520 + }, + { + "epoch": 0.24192661811185928, + "grad_norm": 1.1422178745269775, + "learning_rate": 4.699569989541074e-05, + "loss": 1.0193, + "step": 3521 + }, + { + "epoch": 0.24199532774494983, + "grad_norm": 1.0805256366729736, + "learning_rate": 4.6992849257856334e-05, + "loss": 0.9951, + "step": 3522 + }, + { + "epoch": 0.24206403737804041, + "grad_norm": 1.0789103507995605, + "learning_rate": 4.698999735506671e-05, + "loss": 1.0296, + "step": 3523 + }, + { + "epoch": 0.24213274701113097, + "grad_norm": 1.1235674619674683, + "learning_rate": 4.698714418720595e-05, + "loss": 1.0378, + "step": 3524 + }, + { + "epoch": 0.24220145664422152, + "grad_norm": 1.1342463493347168, + "learning_rate": 4.698428975443819e-05, + "loss": 0.9916, + "step": 3525 + }, + { + "epoch": 0.24227016627731207, + "grad_norm": 1.0979574918746948, + "learning_rate": 4.698143405692765e-05, + "loss": 1.0071, + "step": 3526 + }, + { + "epoch": 0.24233887591040265, + "grad_norm": 1.1012547016143799, + "learning_rate": 4.6978577094838615e-05, + "loss": 0.9849, + "step": 3527 + }, + { + "epoch": 0.2424075855434932, + "grad_norm": 1.1477662324905396, + "learning_rate": 4.697571886833544e-05, + "loss": 0.988, + "step": 3528 + }, + { + "epoch": 0.24247629517658376, + "grad_norm": 1.145579218864441, + "learning_rate": 4.697285937758255e-05, + "loss": 1.0559, + "step": 3529 + }, + { + "epoch": 0.2425450048096743, + "grad_norm": 1.2189531326293945, + "learning_rate": 4.696999862274446e-05, + "loss": 1.0141, + "step": 3530 + }, + { + "epoch": 0.2426137144427649, + "grad_norm": 1.1038439273834229, + "learning_rate": 4.696713660398575e-05, + "loss": 0.9435, + "step": 3531 + }, + { + "epoch": 0.24268242407585544, + "grad_norm": 1.085964560508728, + "learning_rate": 4.696427332147106e-05, + "loss": 1.0287, + "step": 3532 + }, + { + "epoch": 0.242751133708946, + "grad_norm": 1.1679177284240723, + "learning_rate": 4.696140877536512e-05, + "loss": 0.9619, + "step": 3533 + }, + { + "epoch": 0.24281984334203655, + "grad_norm": 1.1798161268234253, + "learning_rate": 4.695854296583272e-05, + "loss": 1.0045, + "step": 3534 + }, + { + "epoch": 0.24288855297512713, + "grad_norm": 1.0980225801467896, + "learning_rate": 4.695567589303874e-05, + "loss": 1.0331, + "step": 3535 + }, + { + "epoch": 0.24295726260821768, + "grad_norm": 1.1130189895629883, + "learning_rate": 4.6952807557148106e-05, + "loss": 0.989, + "step": 3536 + }, + { + "epoch": 0.24302597224130823, + "grad_norm": 1.0908037424087524, + "learning_rate": 4.694993795832584e-05, + "loss": 0.9803, + "step": 3537 + }, + { + "epoch": 0.24309468187439878, + "grad_norm": 1.2695716619491577, + "learning_rate": 4.694706709673703e-05, + "loss": 1.0309, + "step": 3538 + }, + { + "epoch": 0.24316339150748936, + "grad_norm": 1.1312544345855713, + "learning_rate": 4.694419497254683e-05, + "loss": 0.9858, + "step": 3539 + }, + { + "epoch": 0.24323210114057991, + "grad_norm": 1.1654963493347168, + "learning_rate": 4.6941321585920474e-05, + "loss": 0.9981, + "step": 3540 + }, + { + "epoch": 0.24330081077367047, + "grad_norm": 1.0997287034988403, + "learning_rate": 4.693844693702326e-05, + "loss": 0.9887, + "step": 3541 + }, + { + "epoch": 0.24336952040676102, + "grad_norm": 1.1961556673049927, + "learning_rate": 4.693557102602057e-05, + "loss": 1.0506, + "step": 3542 + }, + { + "epoch": 0.2434382300398516, + "grad_norm": 1.1125487089157104, + "learning_rate": 4.693269385307787e-05, + "loss": 1.0419, + "step": 3543 + }, + { + "epoch": 0.24350693967294215, + "grad_norm": 1.0952529907226562, + "learning_rate": 4.692981541836065e-05, + "loss": 1.0031, + "step": 3544 + }, + { + "epoch": 0.2435756493060327, + "grad_norm": 1.1478259563446045, + "learning_rate": 4.692693572203453e-05, + "loss": 1.0297, + "step": 3545 + }, + { + "epoch": 0.24364435893912326, + "grad_norm": 1.0876431465148926, + "learning_rate": 4.692405476426516e-05, + "loss": 1.016, + "step": 3546 + }, + { + "epoch": 0.24371306857221384, + "grad_norm": 1.1878355741500854, + "learning_rate": 4.692117254521829e-05, + "loss": 1.0318, + "step": 3547 + }, + { + "epoch": 0.2437817782053044, + "grad_norm": 1.1028344631195068, + "learning_rate": 4.691828906505973e-05, + "loss": 1.0286, + "step": 3548 + }, + { + "epoch": 0.24385048783839494, + "grad_norm": 1.0768591165542603, + "learning_rate": 4.691540432395536e-05, + "loss": 0.9859, + "step": 3549 + }, + { + "epoch": 0.2439191974714855, + "grad_norm": 1.1063746213912964, + "learning_rate": 4.691251832207115e-05, + "loss": 1.003, + "step": 3550 + }, + { + "epoch": 0.24398790710457607, + "grad_norm": 1.0916955471038818, + "learning_rate": 4.690963105957311e-05, + "loss": 1.0309, + "step": 3551 + }, + { + "epoch": 0.24405661673766663, + "grad_norm": 1.2106469869613647, + "learning_rate": 4.690674253662736e-05, + "loss": 1.0098, + "step": 3552 + }, + { + "epoch": 0.24412532637075718, + "grad_norm": 1.1716567277908325, + "learning_rate": 4.6903852753400074e-05, + "loss": 0.9995, + "step": 3553 + }, + { + "epoch": 0.24419403600384773, + "grad_norm": 1.127650499343872, + "learning_rate": 4.690096171005749e-05, + "loss": 1.0533, + "step": 3554 + }, + { + "epoch": 0.2442627456369383, + "grad_norm": 1.1901713609695435, + "learning_rate": 4.689806940676593e-05, + "loss": 1.083, + "step": 3555 + }, + { + "epoch": 0.24433145527002886, + "grad_norm": 1.1793956756591797, + "learning_rate": 4.68951758436918e-05, + "loss": 1.0004, + "step": 3556 + }, + { + "epoch": 0.24440016490311942, + "grad_norm": 1.2148644924163818, + "learning_rate": 4.6892281021001545e-05, + "loss": 1.0793, + "step": 3557 + }, + { + "epoch": 0.24446887453620997, + "grad_norm": 1.1025614738464355, + "learning_rate": 4.6889384938861715e-05, + "loss": 1.0251, + "step": 3558 + }, + { + "epoch": 0.24453758416930055, + "grad_norm": 1.1302939653396606, + "learning_rate": 4.688648759743891e-05, + "loss": 1.0013, + "step": 3559 + }, + { + "epoch": 0.2446062938023911, + "grad_norm": 1.2375346422195435, + "learning_rate": 4.688358899689982e-05, + "loss": 1.0129, + "step": 3560 + }, + { + "epoch": 0.24467500343548165, + "grad_norm": 1.0923891067504883, + "learning_rate": 4.6880689137411204e-05, + "loss": 0.9746, + "step": 3561 + }, + { + "epoch": 0.2447437130685722, + "grad_norm": 1.2910594940185547, + "learning_rate": 4.6877788019139876e-05, + "loss": 1.0641, + "step": 3562 + }, + { + "epoch": 0.24481242270166279, + "grad_norm": 1.1430705785751343, + "learning_rate": 4.687488564225275e-05, + "loss": 1.0208, + "step": 3563 + }, + { + "epoch": 0.24488113233475334, + "grad_norm": 1.212027668952942, + "learning_rate": 4.687198200691679e-05, + "loss": 1.0206, + "step": 3564 + }, + { + "epoch": 0.2449498419678439, + "grad_norm": 1.2682629823684692, + "learning_rate": 4.686907711329903e-05, + "loss": 1.0802, + "step": 3565 + }, + { + "epoch": 0.24501855160093444, + "grad_norm": 1.176578402519226, + "learning_rate": 4.6866170961566615e-05, + "loss": 1.068, + "step": 3566 + }, + { + "epoch": 0.24508726123402502, + "grad_norm": 1.1625372171401978, + "learning_rate": 4.68632635518867e-05, + "loss": 1.0152, + "step": 3567 + }, + { + "epoch": 0.24515597086711557, + "grad_norm": 1.1086121797561646, + "learning_rate": 4.6860354884426576e-05, + "loss": 1.0506, + "step": 3568 + }, + { + "epoch": 0.24522468050020613, + "grad_norm": 1.147650122642517, + "learning_rate": 4.685744495935356e-05, + "loss": 1.0874, + "step": 3569 + }, + { + "epoch": 0.24529339013329668, + "grad_norm": 1.1340112686157227, + "learning_rate": 4.6854533776835065e-05, + "loss": 1.0285, + "step": 3570 + }, + { + "epoch": 0.24536209976638726, + "grad_norm": 1.1492149829864502, + "learning_rate": 4.6851621337038565e-05, + "loss": 0.9953, + "step": 3571 + }, + { + "epoch": 0.2454308093994778, + "grad_norm": 1.131468653678894, + "learning_rate": 4.6848707640131614e-05, + "loss": 1.0127, + "step": 3572 + }, + { + "epoch": 0.24549951903256836, + "grad_norm": 1.1343345642089844, + "learning_rate": 4.684579268628184e-05, + "loss": 1.0314, + "step": 3573 + }, + { + "epoch": 0.24556822866565892, + "grad_norm": 1.2388594150543213, + "learning_rate": 4.684287647565692e-05, + "loss": 0.9866, + "step": 3574 + }, + { + "epoch": 0.2456369382987495, + "grad_norm": 1.1231132745742798, + "learning_rate": 4.6839959008424644e-05, + "loss": 0.9604, + "step": 3575 + }, + { + "epoch": 0.24570564793184005, + "grad_norm": 1.1209851503372192, + "learning_rate": 4.683704028475284e-05, + "loss": 1.0142, + "step": 3576 + }, + { + "epoch": 0.2457743575649306, + "grad_norm": 1.1838798522949219, + "learning_rate": 4.683412030480943e-05, + "loss": 1.0333, + "step": 3577 + }, + { + "epoch": 0.24584306719802115, + "grad_norm": 1.1089930534362793, + "learning_rate": 4.6831199068762376e-05, + "loss": 1.0108, + "step": 3578 + }, + { + "epoch": 0.24591177683111173, + "grad_norm": 1.1850488185882568, + "learning_rate": 4.682827657677976e-05, + "loss": 1.0327, + "step": 3579 + }, + { + "epoch": 0.24598048646420229, + "grad_norm": 1.2055379152297974, + "learning_rate": 4.6825352829029705e-05, + "loss": 1.0477, + "step": 3580 + }, + { + "epoch": 0.24604919609729284, + "grad_norm": 1.2587542533874512, + "learning_rate": 4.682242782568041e-05, + "loss": 1.0414, + "step": 3581 + }, + { + "epoch": 0.2461179057303834, + "grad_norm": 1.2069756984710693, + "learning_rate": 4.681950156690014e-05, + "loss": 1.0027, + "step": 3582 + }, + { + "epoch": 0.24618661536347397, + "grad_norm": 1.1150243282318115, + "learning_rate": 4.681657405285725e-05, + "loss": 0.9773, + "step": 3583 + }, + { + "epoch": 0.24625532499656452, + "grad_norm": 1.1958894729614258, + "learning_rate": 4.681364528372016e-05, + "loss": 1.032, + "step": 3584 + }, + { + "epoch": 0.24632403462965508, + "grad_norm": 1.0612375736236572, + "learning_rate": 4.681071525965737e-05, + "loss": 0.9787, + "step": 3585 + }, + { + "epoch": 0.24639274426274563, + "grad_norm": 1.1327003240585327, + "learning_rate": 4.680778398083742e-05, + "loss": 1.0144, + "step": 3586 + }, + { + "epoch": 0.2464614538958362, + "grad_norm": 1.1935666799545288, + "learning_rate": 4.680485144742894e-05, + "loss": 1.0232, + "step": 3587 + }, + { + "epoch": 0.24653016352892676, + "grad_norm": 1.1032577753067017, + "learning_rate": 4.6801917659600666e-05, + "loss": 1.0173, + "step": 3588 + }, + { + "epoch": 0.2465988731620173, + "grad_norm": 1.174096703529358, + "learning_rate": 4.679898261752136e-05, + "loss": 0.9987, + "step": 3589 + }, + { + "epoch": 0.24666758279510786, + "grad_norm": 1.110048770904541, + "learning_rate": 4.679604632135988e-05, + "loss": 1.008, + "step": 3590 + }, + { + "epoch": 0.24673629242819844, + "grad_norm": 1.1811264753341675, + "learning_rate": 4.679310877128514e-05, + "loss": 0.9984, + "step": 3591 + }, + { + "epoch": 0.246805002061289, + "grad_norm": 1.2335671186447144, + "learning_rate": 4.679016996746614e-05, + "loss": 1.0057, + "step": 3592 + }, + { + "epoch": 0.24687371169437955, + "grad_norm": 1.1860700845718384, + "learning_rate": 4.678722991007195e-05, + "loss": 1.0625, + "step": 3593 + }, + { + "epoch": 0.2469424213274701, + "grad_norm": 1.174341082572937, + "learning_rate": 4.678428859927171e-05, + "loss": 0.9911, + "step": 3594 + }, + { + "epoch": 0.24701113096056068, + "grad_norm": 1.1314308643341064, + "learning_rate": 4.678134603523462e-05, + "loss": 0.9922, + "step": 3595 + }, + { + "epoch": 0.24707984059365123, + "grad_norm": 1.1150416135787964, + "learning_rate": 4.677840221812998e-05, + "loss": 1.0098, + "step": 3596 + }, + { + "epoch": 0.2471485502267418, + "grad_norm": 1.0581655502319336, + "learning_rate": 4.6775457148127145e-05, + "loss": 0.9927, + "step": 3597 + }, + { + "epoch": 0.24721725985983234, + "grad_norm": 1.0915625095367432, + "learning_rate": 4.6772510825395525e-05, + "loss": 0.9869, + "step": 3598 + }, + { + "epoch": 0.24728596949292292, + "grad_norm": 1.1190921068191528, + "learning_rate": 4.6769563250104644e-05, + "loss": 1.0004, + "step": 3599 + }, + { + "epoch": 0.24735467912601347, + "grad_norm": 1.0542621612548828, + "learning_rate": 4.676661442242406e-05, + "loss": 0.9981, + "step": 3600 + }, + { + "epoch": 0.24742338875910402, + "grad_norm": 1.1339040994644165, + "learning_rate": 4.6763664342523416e-05, + "loss": 1.0557, + "step": 3601 + }, + { + "epoch": 0.24749209839219458, + "grad_norm": 1.1027885675430298, + "learning_rate": 4.676071301057243e-05, + "loss": 1.0354, + "step": 3602 + }, + { + "epoch": 0.24756080802528516, + "grad_norm": 1.114315390586853, + "learning_rate": 4.6757760426740896e-05, + "loss": 1.0091, + "step": 3603 + }, + { + "epoch": 0.2476295176583757, + "grad_norm": 1.1564373970031738, + "learning_rate": 4.675480659119867e-05, + "loss": 1.0401, + "step": 3604 + }, + { + "epoch": 0.24769822729146626, + "grad_norm": 1.1314568519592285, + "learning_rate": 4.675185150411569e-05, + "loss": 1.0035, + "step": 3605 + }, + { + "epoch": 0.2477669369245568, + "grad_norm": 1.1057785749435425, + "learning_rate": 4.6748895165661954e-05, + "loss": 1.0078, + "step": 3606 + }, + { + "epoch": 0.2478356465576474, + "grad_norm": 1.0976202487945557, + "learning_rate": 4.674593757600754e-05, + "loss": 1.0008, + "step": 3607 + }, + { + "epoch": 0.24790435619073795, + "grad_norm": 1.103329062461853, + "learning_rate": 4.674297873532259e-05, + "loss": 0.9822, + "step": 3608 + }, + { + "epoch": 0.2479730658238285, + "grad_norm": 1.1081674098968506, + "learning_rate": 4.674001864377734e-05, + "loss": 1.0575, + "step": 3609 + }, + { + "epoch": 0.24804177545691905, + "grad_norm": 1.104522943496704, + "learning_rate": 4.6737057301542075e-05, + "loss": 1.0202, + "step": 3610 + }, + { + "epoch": 0.24811048509000963, + "grad_norm": 1.1192930936813354, + "learning_rate": 4.673409470878715e-05, + "loss": 1.0363, + "step": 3611 + }, + { + "epoch": 0.24817919472310018, + "grad_norm": 1.1603219509124756, + "learning_rate": 4.673113086568301e-05, + "loss": 1.0024, + "step": 3612 + }, + { + "epoch": 0.24824790435619074, + "grad_norm": 1.131610631942749, + "learning_rate": 4.6728165772400165e-05, + "loss": 0.9667, + "step": 3613 + }, + { + "epoch": 0.2483166139892813, + "grad_norm": 1.2747893333435059, + "learning_rate": 4.672519942910919e-05, + "loss": 1.0097, + "step": 3614 + }, + { + "epoch": 0.24838532362237187, + "grad_norm": 1.2132744789123535, + "learning_rate": 4.6722231835980735e-05, + "loss": 1.0247, + "step": 3615 + }, + { + "epoch": 0.24845403325546242, + "grad_norm": 1.124321460723877, + "learning_rate": 4.671926299318553e-05, + "loss": 1.0257, + "step": 3616 + }, + { + "epoch": 0.24852274288855297, + "grad_norm": 1.2275136709213257, + "learning_rate": 4.6716292900894366e-05, + "loss": 1.027, + "step": 3617 + }, + { + "epoch": 0.24859145252164352, + "grad_norm": 1.1087286472320557, + "learning_rate": 4.671332155927811e-05, + "loss": 1.005, + "step": 3618 + }, + { + "epoch": 0.2486601621547341, + "grad_norm": 1.1529003381729126, + "learning_rate": 4.6710348968507715e-05, + "loss": 0.9937, + "step": 3619 + }, + { + "epoch": 0.24872887178782466, + "grad_norm": 1.1182626485824585, + "learning_rate": 4.670737512875417e-05, + "loss": 1.0042, + "step": 3620 + }, + { + "epoch": 0.2487975814209152, + "grad_norm": 1.1437933444976807, + "learning_rate": 4.6704400040188576e-05, + "loss": 1.003, + "step": 3621 + }, + { + "epoch": 0.24886629105400576, + "grad_norm": 1.1817138195037842, + "learning_rate": 4.6701423702982075e-05, + "loss": 1.0067, + "step": 3622 + }, + { + "epoch": 0.24893500068709634, + "grad_norm": 1.1199740171432495, + "learning_rate": 4.6698446117305904e-05, + "loss": 1.0132, + "step": 3623 + }, + { + "epoch": 0.2490037103201869, + "grad_norm": 1.168595314025879, + "learning_rate": 4.669546728333136e-05, + "loss": 1.057, + "step": 3624 + }, + { + "epoch": 0.24907241995327745, + "grad_norm": 1.1999930143356323, + "learning_rate": 4.6692487201229805e-05, + "loss": 1.044, + "step": 3625 + }, + { + "epoch": 0.249141129586368, + "grad_norm": 1.1382957696914673, + "learning_rate": 4.66895058711727e-05, + "loss": 1.0212, + "step": 3626 + }, + { + "epoch": 0.24920983921945858, + "grad_norm": 1.1386473178863525, + "learning_rate": 4.668652329333154e-05, + "loss": 0.983, + "step": 3627 + }, + { + "epoch": 0.24927854885254913, + "grad_norm": 1.0683811902999878, + "learning_rate": 4.6683539467877915e-05, + "loss": 0.9989, + "step": 3628 + }, + { + "epoch": 0.24934725848563968, + "grad_norm": 1.1915745735168457, + "learning_rate": 4.668055439498349e-05, + "loss": 1.0281, + "step": 3629 + }, + { + "epoch": 0.24941596811873024, + "grad_norm": 1.0534465312957764, + "learning_rate": 4.6677568074819986e-05, + "loss": 0.9632, + "step": 3630 + }, + { + "epoch": 0.24948467775182082, + "grad_norm": 1.09433913230896, + "learning_rate": 4.667458050755921e-05, + "loss": 1.0121, + "step": 3631 + }, + { + "epoch": 0.24955338738491137, + "grad_norm": 1.0982122421264648, + "learning_rate": 4.6671591693373043e-05, + "loss": 0.9888, + "step": 3632 + }, + { + "epoch": 0.24962209701800192, + "grad_norm": 1.100877285003662, + "learning_rate": 4.666860163243341e-05, + "loss": 1.034, + "step": 3633 + }, + { + "epoch": 0.24969080665109247, + "grad_norm": 1.1925259828567505, + "learning_rate": 4.6665610324912335e-05, + "loss": 1.027, + "step": 3634 + }, + { + "epoch": 0.24975951628418305, + "grad_norm": 1.0914720296859741, + "learning_rate": 4.666261777098191e-05, + "loss": 0.986, + "step": 3635 + }, + { + "epoch": 0.2498282259172736, + "grad_norm": 1.1598222255706787, + "learning_rate": 4.665962397081429e-05, + "loss": 1.0447, + "step": 3636 + }, + { + "epoch": 0.24989693555036416, + "grad_norm": 1.1084721088409424, + "learning_rate": 4.665662892458172e-05, + "loss": 1.004, + "step": 3637 + }, + { + "epoch": 0.2499656451834547, + "grad_norm": 1.0900660753250122, + "learning_rate": 4.665363263245649e-05, + "loss": 1.0036, + "step": 3638 + }, + { + "epoch": 0.2500343548165453, + "grad_norm": 1.104061245918274, + "learning_rate": 4.665063509461097e-05, + "loss": 1.0304, + "step": 3639 + }, + { + "epoch": 0.2501030644496358, + "grad_norm": 1.1117284297943115, + "learning_rate": 4.6647636311217613e-05, + "loss": 0.9716, + "step": 3640 + }, + { + "epoch": 0.2501717740827264, + "grad_norm": 1.1447994709014893, + "learning_rate": 4.664463628244895e-05, + "loss": 1.0394, + "step": 3641 + }, + { + "epoch": 0.250240483715817, + "grad_norm": 1.2110528945922852, + "learning_rate": 4.664163500847755e-05, + "loss": 1.0498, + "step": 3642 + }, + { + "epoch": 0.2503091933489075, + "grad_norm": 1.160157561302185, + "learning_rate": 4.6638632489476086e-05, + "loss": 0.9971, + "step": 3643 + }, + { + "epoch": 0.2503779029819981, + "grad_norm": 1.1038448810577393, + "learning_rate": 4.663562872561729e-05, + "loss": 1.059, + "step": 3644 + }, + { + "epoch": 0.25044661261508866, + "grad_norm": 1.1368685960769653, + "learning_rate": 4.6632623717073964e-05, + "loss": 0.9874, + "step": 3645 + }, + { + "epoch": 0.2505153222481792, + "grad_norm": 1.1318825483322144, + "learning_rate": 4.6629617464018985e-05, + "loss": 1.0524, + "step": 3646 + }, + { + "epoch": 0.25058403188126976, + "grad_norm": 1.1810511350631714, + "learning_rate": 4.66266099666253e-05, + "loss": 1.0033, + "step": 3647 + }, + { + "epoch": 0.2506527415143603, + "grad_norm": 1.2213659286499023, + "learning_rate": 4.662360122506594e-05, + "loss": 0.9953, + "step": 3648 + }, + { + "epoch": 0.25072145114745087, + "grad_norm": 1.1475211381912231, + "learning_rate": 4.6620591239513976e-05, + "loss": 1.026, + "step": 3649 + }, + { + "epoch": 0.25079016078054145, + "grad_norm": 1.1088424921035767, + "learning_rate": 4.661758001014258e-05, + "loss": 0.9828, + "step": 3650 + }, + { + "epoch": 0.250858870413632, + "grad_norm": 1.2615996599197388, + "learning_rate": 4.6614567537124994e-05, + "loss": 1.0048, + "step": 3651 + }, + { + "epoch": 0.25092758004672255, + "grad_norm": 1.1392616033554077, + "learning_rate": 4.661155382063451e-05, + "loss": 1.0192, + "step": 3652 + }, + { + "epoch": 0.25099628967981313, + "grad_norm": 1.2329214811325073, + "learning_rate": 4.660853886084452e-05, + "loss": 1.0381, + "step": 3653 + }, + { + "epoch": 0.25106499931290366, + "grad_norm": 1.0756943225860596, + "learning_rate": 4.660552265792846e-05, + "loss": 0.979, + "step": 3654 + }, + { + "epoch": 0.25113370894599424, + "grad_norm": 1.1312164068222046, + "learning_rate": 4.6602505212059855e-05, + "loss": 1.0748, + "step": 3655 + }, + { + "epoch": 0.25120241857908476, + "grad_norm": 1.138766884803772, + "learning_rate": 4.6599486523412295e-05, + "loss": 1.0368, + "step": 3656 + }, + { + "epoch": 0.25127112821217534, + "grad_norm": 1.079370379447937, + "learning_rate": 4.6596466592159446e-05, + "loss": 0.9809, + "step": 3657 + }, + { + "epoch": 0.2513398378452659, + "grad_norm": 1.1702358722686768, + "learning_rate": 4.6593445418475045e-05, + "loss": 0.9965, + "step": 3658 + }, + { + "epoch": 0.25140854747835645, + "grad_norm": 1.1692465543746948, + "learning_rate": 4.659042300253289e-05, + "loss": 1.0624, + "step": 3659 + }, + { + "epoch": 0.25147725711144703, + "grad_norm": 1.1960861682891846, + "learning_rate": 4.658739934450687e-05, + "loss": 1.0325, + "step": 3660 + }, + { + "epoch": 0.2515459667445376, + "grad_norm": 1.0887835025787354, + "learning_rate": 4.658437444457093e-05, + "loss": 1.0545, + "step": 3661 + }, + { + "epoch": 0.25161467637762813, + "grad_norm": 1.1567937135696411, + "learning_rate": 4.658134830289909e-05, + "loss": 0.9922, + "step": 3662 + }, + { + "epoch": 0.2516833860107187, + "grad_norm": 1.2081875801086426, + "learning_rate": 4.657832091966543e-05, + "loss": 1.0325, + "step": 3663 + }, + { + "epoch": 0.25175209564380924, + "grad_norm": 1.1212272644042969, + "learning_rate": 4.6575292295044126e-05, + "loss": 0.9834, + "step": 3664 + }, + { + "epoch": 0.2518208052768998, + "grad_norm": 1.126142144203186, + "learning_rate": 4.657226242920942e-05, + "loss": 1.0401, + "step": 3665 + }, + { + "epoch": 0.2518895149099904, + "grad_norm": 1.1239615678787231, + "learning_rate": 4.65692313223356e-05, + "loss": 1.0295, + "step": 3666 + }, + { + "epoch": 0.2519582245430809, + "grad_norm": 1.1923370361328125, + "learning_rate": 4.656619897459707e-05, + "loss": 1.0406, + "step": 3667 + }, + { + "epoch": 0.2520269341761715, + "grad_norm": 1.149319052696228, + "learning_rate": 4.656316538616824e-05, + "loss": 0.9811, + "step": 3668 + }, + { + "epoch": 0.2520956438092621, + "grad_norm": 1.097008228302002, + "learning_rate": 4.656013055722367e-05, + "loss": 1.016, + "step": 3669 + }, + { + "epoch": 0.2521643534423526, + "grad_norm": 1.033589243888855, + "learning_rate": 4.655709448793792e-05, + "loss": 0.9851, + "step": 3670 + }, + { + "epoch": 0.2522330630754432, + "grad_norm": 1.1520308256149292, + "learning_rate": 4.655405717848569e-05, + "loss": 1.0472, + "step": 3671 + }, + { + "epoch": 0.2523017727085337, + "grad_norm": 1.1103944778442383, + "learning_rate": 4.655101862904168e-05, + "loss": 0.9885, + "step": 3672 + }, + { + "epoch": 0.2523704823416243, + "grad_norm": 1.209547996520996, + "learning_rate": 4.65479788397807e-05, + "loss": 1.0154, + "step": 3673 + }, + { + "epoch": 0.25243919197471487, + "grad_norm": 1.1123695373535156, + "learning_rate": 4.6544937810877654e-05, + "loss": 1.0028, + "step": 3674 + }, + { + "epoch": 0.2525079016078054, + "grad_norm": 1.1362240314483643, + "learning_rate": 4.654189554250746e-05, + "loss": 0.9893, + "step": 3675 + }, + { + "epoch": 0.252576611240896, + "grad_norm": 1.1061813831329346, + "learning_rate": 4.653885203484515e-05, + "loss": 1.0266, + "step": 3676 + }, + { + "epoch": 0.25264532087398656, + "grad_norm": 1.1106137037277222, + "learning_rate": 4.6535807288065824e-05, + "loss": 1.0083, + "step": 3677 + }, + { + "epoch": 0.2527140305070771, + "grad_norm": 1.119681477546692, + "learning_rate": 4.653276130234463e-05, + "loss": 1.0236, + "step": 3678 + }, + { + "epoch": 0.25278274014016766, + "grad_norm": 1.1129450798034668, + "learning_rate": 4.652971407785681e-05, + "loss": 1.0079, + "step": 3679 + }, + { + "epoch": 0.2528514497732582, + "grad_norm": 1.1385815143585205, + "learning_rate": 4.6526665614777666e-05, + "loss": 0.9978, + "step": 3680 + }, + { + "epoch": 0.25292015940634877, + "grad_norm": 1.174641728401184, + "learning_rate": 4.652361591328257e-05, + "loss": 1.0187, + "step": 3681 + }, + { + "epoch": 0.25298886903943935, + "grad_norm": 1.0959815979003906, + "learning_rate": 4.652056497354698e-05, + "loss": 0.9698, + "step": 3682 + }, + { + "epoch": 0.25305757867252987, + "grad_norm": 1.1914905309677124, + "learning_rate": 4.651751279574641e-05, + "loss": 0.9971, + "step": 3683 + }, + { + "epoch": 0.25312628830562045, + "grad_norm": 1.0774507522583008, + "learning_rate": 4.6514459380056444e-05, + "loss": 0.9791, + "step": 3684 + }, + { + "epoch": 0.25319499793871103, + "grad_norm": 1.0792535543441772, + "learning_rate": 4.6511404726652754e-05, + "loss": 0.9997, + "step": 3685 + }, + { + "epoch": 0.25326370757180156, + "grad_norm": 1.1409186124801636, + "learning_rate": 4.650834883571106e-05, + "loss": 1.0047, + "step": 3686 + }, + { + "epoch": 0.25333241720489214, + "grad_norm": 1.0744630098342896, + "learning_rate": 4.6505291707407174e-05, + "loss": 1.0417, + "step": 3687 + }, + { + "epoch": 0.25340112683798266, + "grad_norm": 1.1779234409332275, + "learning_rate": 4.6502233341916965e-05, + "loss": 0.994, + "step": 3688 + }, + { + "epoch": 0.25346983647107324, + "grad_norm": 1.1735395193099976, + "learning_rate": 4.6499173739416384e-05, + "loss": 1.0491, + "step": 3689 + }, + { + "epoch": 0.2535385461041638, + "grad_norm": 1.1759095191955566, + "learning_rate": 4.649611290008145e-05, + "loss": 1.0253, + "step": 3690 + }, + { + "epoch": 0.25360725573725434, + "grad_norm": 1.1668455600738525, + "learning_rate": 4.649305082408825e-05, + "loss": 1.0068, + "step": 3691 + }, + { + "epoch": 0.2536759653703449, + "grad_norm": 1.1750859022140503, + "learning_rate": 4.648998751161293e-05, + "loss": 1.0246, + "step": 3692 + }, + { + "epoch": 0.2537446750034355, + "grad_norm": 1.169073462486267, + "learning_rate": 4.648692296283174e-05, + "loss": 1.0492, + "step": 3693 + }, + { + "epoch": 0.25381338463652603, + "grad_norm": 1.1056652069091797, + "learning_rate": 4.648385717792098e-05, + "loss": 1.0394, + "step": 3694 + }, + { + "epoch": 0.2538820942696166, + "grad_norm": 1.0938621759414673, + "learning_rate": 4.6480790157056995e-05, + "loss": 1.0149, + "step": 3695 + }, + { + "epoch": 0.25395080390270713, + "grad_norm": 1.0677684545516968, + "learning_rate": 4.647772190041627e-05, + "loss": 0.9854, + "step": 3696 + }, + { + "epoch": 0.2540195135357977, + "grad_norm": 1.1318833827972412, + "learning_rate": 4.647465240817529e-05, + "loss": 1.023, + "step": 3697 + }, + { + "epoch": 0.2540882231688883, + "grad_norm": 1.1680909395217896, + "learning_rate": 4.647158168051066e-05, + "loss": 1.0034, + "step": 3698 + }, + { + "epoch": 0.2541569328019788, + "grad_norm": 1.1309857368469238, + "learning_rate": 4.646850971759902e-05, + "loss": 0.9927, + "step": 3699 + }, + { + "epoch": 0.2542256424350694, + "grad_norm": 1.1563429832458496, + "learning_rate": 4.64654365196171e-05, + "loss": 0.9973, + "step": 3700 + }, + { + "epoch": 0.25429435206816, + "grad_norm": 1.0642952919006348, + "learning_rate": 4.646236208674172e-05, + "loss": 0.994, + "step": 3701 + }, + { + "epoch": 0.2543630617012505, + "grad_norm": 1.0615206956863403, + "learning_rate": 4.645928641914972e-05, + "loss": 0.9651, + "step": 3702 + }, + { + "epoch": 0.2544317713343411, + "grad_norm": 1.1360700130462646, + "learning_rate": 4.645620951701807e-05, + "loss": 0.9885, + "step": 3703 + }, + { + "epoch": 0.2545004809674316, + "grad_norm": 1.2268877029418945, + "learning_rate": 4.645313138052376e-05, + "loss": 1.0394, + "step": 3704 + }, + { + "epoch": 0.2545691906005222, + "grad_norm": 1.0715532302856445, + "learning_rate": 4.6450052009843885e-05, + "loss": 0.9851, + "step": 3705 + }, + { + "epoch": 0.25463790023361277, + "grad_norm": 1.103995442390442, + "learning_rate": 4.6446971405155596e-05, + "loss": 0.9942, + "step": 3706 + }, + { + "epoch": 0.2547066098667033, + "grad_norm": 1.1432462930679321, + "learning_rate": 4.644388956663612e-05, + "loss": 1.0187, + "step": 3707 + }, + { + "epoch": 0.2547753194997939, + "grad_norm": 1.1062805652618408, + "learning_rate": 4.6440806494462755e-05, + "loss": 0.997, + "step": 3708 + }, + { + "epoch": 0.25484402913288445, + "grad_norm": 1.1612145900726318, + "learning_rate": 4.6437722188812864e-05, + "loss": 1.0263, + "step": 3709 + }, + { + "epoch": 0.254912738765975, + "grad_norm": 1.231427788734436, + "learning_rate": 4.643463664986388e-05, + "loss": 1.0791, + "step": 3710 + }, + { + "epoch": 0.25498144839906556, + "grad_norm": 1.1217185258865356, + "learning_rate": 4.643154987779333e-05, + "loss": 0.9899, + "step": 3711 + }, + { + "epoch": 0.2550501580321561, + "grad_norm": 1.0816409587860107, + "learning_rate": 4.6428461872778776e-05, + "loss": 1.0505, + "step": 3712 + }, + { + "epoch": 0.25511886766524666, + "grad_norm": 1.1667671203613281, + "learning_rate": 4.642537263499788e-05, + "loss": 1.0122, + "step": 3713 + }, + { + "epoch": 0.25518757729833724, + "grad_norm": 1.1080068349838257, + "learning_rate": 4.642228216462836e-05, + "loss": 0.9708, + "step": 3714 + }, + { + "epoch": 0.25525628693142777, + "grad_norm": 1.1219269037246704, + "learning_rate": 4.6419190461848005e-05, + "loss": 0.9996, + "step": 3715 + }, + { + "epoch": 0.25532499656451835, + "grad_norm": 1.1081352233886719, + "learning_rate": 4.641609752683469e-05, + "loss": 0.9979, + "step": 3716 + }, + { + "epoch": 0.2553937061976089, + "grad_norm": 1.143578052520752, + "learning_rate": 4.641300335976634e-05, + "loss": 0.9822, + "step": 3717 + }, + { + "epoch": 0.25546241583069945, + "grad_norm": 1.181684136390686, + "learning_rate": 4.640990796082095e-05, + "loss": 0.9982, + "step": 3718 + }, + { + "epoch": 0.25553112546379003, + "grad_norm": 1.0945219993591309, + "learning_rate": 4.640681133017662e-05, + "loss": 1.0021, + "step": 3719 + }, + { + "epoch": 0.25559983509688056, + "grad_norm": 1.1793862581253052, + "learning_rate": 4.640371346801149e-05, + "loss": 1.0302, + "step": 3720 + }, + { + "epoch": 0.25566854472997114, + "grad_norm": 1.1305969953536987, + "learning_rate": 4.640061437450377e-05, + "loss": 0.9725, + "step": 3721 + }, + { + "epoch": 0.2557372543630617, + "grad_norm": 1.0996021032333374, + "learning_rate": 4.639751404983176e-05, + "loss": 0.9873, + "step": 3722 + }, + { + "epoch": 0.25580596399615224, + "grad_norm": 1.1743967533111572, + "learning_rate": 4.6394412494173804e-05, + "loss": 1.0375, + "step": 3723 + }, + { + "epoch": 0.2558746736292428, + "grad_norm": 1.126446008682251, + "learning_rate": 4.639130970770835e-05, + "loss": 1.0214, + "step": 3724 + }, + { + "epoch": 0.2559433832623334, + "grad_norm": 1.0905985832214355, + "learning_rate": 4.6388205690613885e-05, + "loss": 0.9901, + "step": 3725 + }, + { + "epoch": 0.2560120928954239, + "grad_norm": 1.0787701606750488, + "learning_rate": 4.638510044306899e-05, + "loss": 0.9559, + "step": 3726 + }, + { + "epoch": 0.2560808025285145, + "grad_norm": 1.254197359085083, + "learning_rate": 4.6381993965252304e-05, + "loss": 1.0155, + "step": 3727 + }, + { + "epoch": 0.25614951216160503, + "grad_norm": 1.0974581241607666, + "learning_rate": 4.637888625734255e-05, + "loss": 0.9933, + "step": 3728 + }, + { + "epoch": 0.2562182217946956, + "grad_norm": 1.1428461074829102, + "learning_rate": 4.63757773195185e-05, + "loss": 0.996, + "step": 3729 + }, + { + "epoch": 0.2562869314277862, + "grad_norm": 1.16291081905365, + "learning_rate": 4.637266715195901e-05, + "loss": 1.064, + "step": 3730 + }, + { + "epoch": 0.2563556410608767, + "grad_norm": 1.058424711227417, + "learning_rate": 4.6369555754843014e-05, + "loss": 1.0071, + "step": 3731 + }, + { + "epoch": 0.2564243506939673, + "grad_norm": 1.1671509742736816, + "learning_rate": 4.636644312834951e-05, + "loss": 1.033, + "step": 3732 + }, + { + "epoch": 0.2564930603270579, + "grad_norm": 1.176649570465088, + "learning_rate": 4.636332927265755e-05, + "loss": 1.0096, + "step": 3733 + }, + { + "epoch": 0.2565617699601484, + "grad_norm": 1.1102588176727295, + "learning_rate": 4.636021418794629e-05, + "loss": 0.9855, + "step": 3734 + }, + { + "epoch": 0.256630479593239, + "grad_norm": 1.11301851272583, + "learning_rate": 4.635709787439493e-05, + "loss": 1.0194, + "step": 3735 + }, + { + "epoch": 0.2566991892263295, + "grad_norm": 1.0906847715377808, + "learning_rate": 4.6353980332182756e-05, + "loss": 1.0187, + "step": 3736 + }, + { + "epoch": 0.2567678988594201, + "grad_norm": 1.151524305343628, + "learning_rate": 4.635086156148911e-05, + "loss": 1.0219, + "step": 3737 + }, + { + "epoch": 0.25683660849251067, + "grad_norm": 1.1333705186843872, + "learning_rate": 4.634774156249342e-05, + "loss": 1.0345, + "step": 3738 + }, + { + "epoch": 0.2569053181256012, + "grad_norm": 1.1113635301589966, + "learning_rate": 4.6344620335375165e-05, + "loss": 1.0004, + "step": 3739 + }, + { + "epoch": 0.25697402775869177, + "grad_norm": 1.1291612386703491, + "learning_rate": 4.6341497880313924e-05, + "loss": 1.0078, + "step": 3740 + }, + { + "epoch": 0.25704273739178235, + "grad_norm": 1.0951992273330688, + "learning_rate": 4.633837419748932e-05, + "loss": 0.9908, + "step": 3741 + }, + { + "epoch": 0.2571114470248729, + "grad_norm": 1.1462346315383911, + "learning_rate": 4.633524928708106e-05, + "loss": 1.0058, + "step": 3742 + }, + { + "epoch": 0.25718015665796345, + "grad_norm": 1.1701215505599976, + "learning_rate": 4.6332123149268916e-05, + "loss": 1.0476, + "step": 3743 + }, + { + "epoch": 0.257248866291054, + "grad_norm": 1.174586296081543, + "learning_rate": 4.632899578423274e-05, + "loss": 1.0273, + "step": 3744 + }, + { + "epoch": 0.25731757592414456, + "grad_norm": 1.1220732927322388, + "learning_rate": 4.632586719215244e-05, + "loss": 1.0281, + "step": 3745 + }, + { + "epoch": 0.25738628555723514, + "grad_norm": 1.1630891561508179, + "learning_rate": 4.6322737373208e-05, + "loss": 1.002, + "step": 3746 + }, + { + "epoch": 0.25745499519032566, + "grad_norm": 1.1217522621154785, + "learning_rate": 4.631960632757948e-05, + "loss": 1.0221, + "step": 3747 + }, + { + "epoch": 0.25752370482341624, + "grad_norm": 1.1181656122207642, + "learning_rate": 4.6316474055447005e-05, + "loss": 1.0187, + "step": 3748 + }, + { + "epoch": 0.2575924144565068, + "grad_norm": 1.1725819110870361, + "learning_rate": 4.631334055699078e-05, + "loss": 1.0346, + "step": 3749 + }, + { + "epoch": 0.25766112408959735, + "grad_norm": 1.0560280084609985, + "learning_rate": 4.631020583239107e-05, + "loss": 1.0231, + "step": 3750 + }, + { + "epoch": 0.25772983372268793, + "grad_norm": 1.2495685815811157, + "learning_rate": 4.6307069881828205e-05, + "loss": 1.0548, + "step": 3751 + }, + { + "epoch": 0.25779854335577845, + "grad_norm": 1.1612706184387207, + "learning_rate": 4.63039327054826e-05, + "loss": 1.0223, + "step": 3752 + }, + { + "epoch": 0.25786725298886903, + "grad_norm": 1.1648447513580322, + "learning_rate": 4.6300794303534744e-05, + "loss": 1.0473, + "step": 3753 + }, + { + "epoch": 0.2579359626219596, + "grad_norm": 1.0671497583389282, + "learning_rate": 4.629765467616518e-05, + "loss": 0.9979, + "step": 3754 + }, + { + "epoch": 0.25800467225505014, + "grad_norm": 1.1751644611358643, + "learning_rate": 4.629451382355452e-05, + "loss": 1.0679, + "step": 3755 + }, + { + "epoch": 0.2580733818881407, + "grad_norm": 1.037760615348816, + "learning_rate": 4.6291371745883474e-05, + "loss": 0.9979, + "step": 3756 + }, + { + "epoch": 0.2581420915212313, + "grad_norm": 1.1435027122497559, + "learning_rate": 4.628822844333278e-05, + "loss": 0.9814, + "step": 3757 + }, + { + "epoch": 0.2582108011543218, + "grad_norm": 1.086582064628601, + "learning_rate": 4.628508391608329e-05, + "loss": 1.0032, + "step": 3758 + }, + { + "epoch": 0.2582795107874124, + "grad_norm": 1.1078418493270874, + "learning_rate": 4.628193816431591e-05, + "loss": 1.0205, + "step": 3759 + }, + { + "epoch": 0.2583482204205029, + "grad_norm": 1.150707721710205, + "learning_rate": 4.62787911882116e-05, + "loss": 1.0314, + "step": 3760 + }, + { + "epoch": 0.2584169300535935, + "grad_norm": 1.063823938369751, + "learning_rate": 4.6275642987951396e-05, + "loss": 1.0148, + "step": 3761 + }, + { + "epoch": 0.2584856396866841, + "grad_norm": 1.1260151863098145, + "learning_rate": 4.6272493563716434e-05, + "loss": 1.0434, + "step": 3762 + }, + { + "epoch": 0.2585543493197746, + "grad_norm": 1.1629862785339355, + "learning_rate": 4.6269342915687884e-05, + "loss": 0.994, + "step": 3763 + }, + { + "epoch": 0.2586230589528652, + "grad_norm": 1.1506469249725342, + "learning_rate": 4.6266191044047004e-05, + "loss": 1.0637, + "step": 3764 + }, + { + "epoch": 0.2586917685859558, + "grad_norm": 1.1266156435012817, + "learning_rate": 4.626303794897512e-05, + "loss": 1.0211, + "step": 3765 + }, + { + "epoch": 0.2587604782190463, + "grad_norm": 1.0643614530563354, + "learning_rate": 4.625988363065363e-05, + "loss": 0.9917, + "step": 3766 + }, + { + "epoch": 0.2588291878521369, + "grad_norm": 1.089905858039856, + "learning_rate": 4.6256728089264e-05, + "loss": 1.0046, + "step": 3767 + }, + { + "epoch": 0.2588978974852274, + "grad_norm": 1.2007625102996826, + "learning_rate": 4.625357132498775e-05, + "loss": 1.0307, + "step": 3768 + }, + { + "epoch": 0.258966607118318, + "grad_norm": 1.1548783779144287, + "learning_rate": 4.6250413338006515e-05, + "loss": 1.0196, + "step": 3769 + }, + { + "epoch": 0.25903531675140856, + "grad_norm": 1.1784335374832153, + "learning_rate": 4.624725412850196e-05, + "loss": 1.007, + "step": 3770 + }, + { + "epoch": 0.2591040263844991, + "grad_norm": 1.138848900794983, + "learning_rate": 4.624409369665582e-05, + "loss": 0.9772, + "step": 3771 + }, + { + "epoch": 0.25917273601758967, + "grad_norm": 1.1635223627090454, + "learning_rate": 4.6240932042649926e-05, + "loss": 1.0771, + "step": 3772 + }, + { + "epoch": 0.25924144565068025, + "grad_norm": 1.123651146888733, + "learning_rate": 4.623776916666617e-05, + "loss": 1.0449, + "step": 3773 + }, + { + "epoch": 0.25931015528377077, + "grad_norm": 1.1077024936676025, + "learning_rate": 4.623460506888649e-05, + "loss": 1.0087, + "step": 3774 + }, + { + "epoch": 0.25937886491686135, + "grad_norm": 1.1568564176559448, + "learning_rate": 4.6231439749492935e-05, + "loss": 1.0205, + "step": 3775 + }, + { + "epoch": 0.2594475745499519, + "grad_norm": 1.1845672130584717, + "learning_rate": 4.6228273208667595e-05, + "loss": 1.0121, + "step": 3776 + }, + { + "epoch": 0.25951628418304246, + "grad_norm": 1.098218560218811, + "learning_rate": 4.6225105446592646e-05, + "loss": 1.0181, + "step": 3777 + }, + { + "epoch": 0.25958499381613304, + "grad_norm": 1.151556134223938, + "learning_rate": 4.6221936463450313e-05, + "loss": 1.0467, + "step": 3778 + }, + { + "epoch": 0.25965370344922356, + "grad_norm": 1.1599406003952026, + "learning_rate": 4.621876625942292e-05, + "loss": 1.0252, + "step": 3779 + }, + { + "epoch": 0.25972241308231414, + "grad_norm": 1.143224835395813, + "learning_rate": 4.6215594834692845e-05, + "loss": 1.002, + "step": 3780 + }, + { + "epoch": 0.2597911227154047, + "grad_norm": 1.1161472797393799, + "learning_rate": 4.6212422189442536e-05, + "loss": 1.0194, + "step": 3781 + }, + { + "epoch": 0.25985983234849525, + "grad_norm": 1.1720703840255737, + "learning_rate": 4.6209248323854505e-05, + "loss": 0.997, + "step": 3782 + }, + { + "epoch": 0.2599285419815858, + "grad_norm": 1.1950087547302246, + "learning_rate": 4.620607323811136e-05, + "loss": 1.0269, + "step": 3783 + }, + { + "epoch": 0.25999725161467635, + "grad_norm": 1.109749436378479, + "learning_rate": 4.620289693239574e-05, + "loss": 1.0192, + "step": 3784 + }, + { + "epoch": 0.26006596124776693, + "grad_norm": 1.1006176471710205, + "learning_rate": 4.61997194068904e-05, + "loss": 1.0308, + "step": 3785 + }, + { + "epoch": 0.2601346708808575, + "grad_norm": 1.1648743152618408, + "learning_rate": 4.6196540661778114e-05, + "loss": 1.0498, + "step": 3786 + }, + { + "epoch": 0.26020338051394803, + "grad_norm": 1.1445064544677734, + "learning_rate": 4.619336069724177e-05, + "loss": 0.9922, + "step": 3787 + }, + { + "epoch": 0.2602720901470386, + "grad_norm": 0.9917647838592529, + "learning_rate": 4.619017951346432e-05, + "loss": 0.9451, + "step": 3788 + }, + { + "epoch": 0.2603407997801292, + "grad_norm": 1.0601063966751099, + "learning_rate": 4.618699711062875e-05, + "loss": 0.9879, + "step": 3789 + }, + { + "epoch": 0.2604095094132197, + "grad_norm": 1.0271998643875122, + "learning_rate": 4.618381348891816e-05, + "loss": 0.9953, + "step": 3790 + }, + { + "epoch": 0.2604782190463103, + "grad_norm": 1.106622338294983, + "learning_rate": 4.61806286485157e-05, + "loss": 1.0319, + "step": 3791 + }, + { + "epoch": 0.2605469286794008, + "grad_norm": 1.1277573108673096, + "learning_rate": 4.617744258960459e-05, + "loss": 1.0439, + "step": 3792 + }, + { + "epoch": 0.2606156383124914, + "grad_norm": 1.0709083080291748, + "learning_rate": 4.617425531236811e-05, + "loss": 1.0521, + "step": 3793 + }, + { + "epoch": 0.260684347945582, + "grad_norm": 1.1136687994003296, + "learning_rate": 4.617106681698964e-05, + "loss": 0.9839, + "step": 3794 + }, + { + "epoch": 0.2607530575786725, + "grad_norm": 1.0893486738204956, + "learning_rate": 4.61678771036526e-05, + "loss": 0.9771, + "step": 3795 + }, + { + "epoch": 0.2608217672117631, + "grad_norm": 1.1539204120635986, + "learning_rate": 4.616468617254051e-05, + "loss": 1.0227, + "step": 3796 + }, + { + "epoch": 0.26089047684485367, + "grad_norm": 1.1145641803741455, + "learning_rate": 4.6161494023836915e-05, + "loss": 1.0177, + "step": 3797 + }, + { + "epoch": 0.2609591864779442, + "grad_norm": 1.118013620376587, + "learning_rate": 4.6158300657725476e-05, + "loss": 1.014, + "step": 3798 + }, + { + "epoch": 0.2610278961110348, + "grad_norm": 1.119649887084961, + "learning_rate": 4.61551060743899e-05, + "loss": 1.0483, + "step": 3799 + }, + { + "epoch": 0.26109660574412535, + "grad_norm": 1.2208480834960938, + "learning_rate": 4.615191027401397e-05, + "loss": 1.0961, + "step": 3800 + }, + { + "epoch": 0.2611653153772159, + "grad_norm": 1.1167482137680054, + "learning_rate": 4.614871325678154e-05, + "loss": 1.0373, + "step": 3801 + }, + { + "epoch": 0.26123402501030646, + "grad_norm": 1.1197617053985596, + "learning_rate": 4.614551502287653e-05, + "loss": 1.0378, + "step": 3802 + }, + { + "epoch": 0.261302734643397, + "grad_norm": 1.077839970588684, + "learning_rate": 4.614231557248294e-05, + "loss": 1.0014, + "step": 3803 + }, + { + "epoch": 0.26137144427648756, + "grad_norm": 1.165567398071289, + "learning_rate": 4.613911490578482e-05, + "loss": 1.0283, + "step": 3804 + }, + { + "epoch": 0.26144015390957814, + "grad_norm": 1.1173679828643799, + "learning_rate": 4.613591302296632e-05, + "loss": 0.9707, + "step": 3805 + }, + { + "epoch": 0.26150886354266867, + "grad_norm": 1.1741645336151123, + "learning_rate": 4.613270992421161e-05, + "loss": 1.0138, + "step": 3806 + }, + { + "epoch": 0.26157757317575925, + "grad_norm": 1.1205273866653442, + "learning_rate": 4.6129505609705e-05, + "loss": 1.0487, + "step": 3807 + }, + { + "epoch": 0.26164628280884983, + "grad_norm": 1.0978105068206787, + "learning_rate": 4.612630007963081e-05, + "loss": 1.0281, + "step": 3808 + }, + { + "epoch": 0.26171499244194035, + "grad_norm": 1.1851376295089722, + "learning_rate": 4.6123093334173464e-05, + "loss": 1.0377, + "step": 3809 + }, + { + "epoch": 0.26178370207503093, + "grad_norm": 1.1376523971557617, + "learning_rate": 4.6119885373517424e-05, + "loss": 1.0256, + "step": 3810 + }, + { + "epoch": 0.26185241170812146, + "grad_norm": 1.1419093608856201, + "learning_rate": 4.6116676197847265e-05, + "loss": 1.0081, + "step": 3811 + }, + { + "epoch": 0.26192112134121204, + "grad_norm": 1.192095160484314, + "learning_rate": 4.61134658073476e-05, + "loss": 1.0594, + "step": 3812 + }, + { + "epoch": 0.2619898309743026, + "grad_norm": 1.1356091499328613, + "learning_rate": 4.611025420220312e-05, + "loss": 1.0209, + "step": 3813 + }, + { + "epoch": 0.26205854060739314, + "grad_norm": 1.0959616899490356, + "learning_rate": 4.610704138259859e-05, + "loss": 0.9788, + "step": 3814 + }, + { + "epoch": 0.2621272502404837, + "grad_norm": 1.082021951675415, + "learning_rate": 4.6103827348718834e-05, + "loss": 1.0076, + "step": 3815 + }, + { + "epoch": 0.2621959598735743, + "grad_norm": 1.137729287147522, + "learning_rate": 4.6100612100748765e-05, + "loss": 1.0202, + "step": 3816 + }, + { + "epoch": 0.2622646695066648, + "grad_norm": 1.156856894493103, + "learning_rate": 4.609739563887334e-05, + "loss": 1.0239, + "step": 3817 + }, + { + "epoch": 0.2623333791397554, + "grad_norm": 1.0934593677520752, + "learning_rate": 4.609417796327761e-05, + "loss": 0.9971, + "step": 3818 + }, + { + "epoch": 0.26240208877284593, + "grad_norm": 1.0405113697052002, + "learning_rate": 4.609095907414669e-05, + "loss": 0.9826, + "step": 3819 + }, + { + "epoch": 0.2624707984059365, + "grad_norm": 1.0390129089355469, + "learning_rate": 4.608773897166574e-05, + "loss": 1.0305, + "step": 3820 + }, + { + "epoch": 0.2625395080390271, + "grad_norm": 1.0936105251312256, + "learning_rate": 4.608451765602004e-05, + "loss": 0.9993, + "step": 3821 + }, + { + "epoch": 0.2626082176721176, + "grad_norm": 1.1191329956054688, + "learning_rate": 4.608129512739489e-05, + "loss": 1.004, + "step": 3822 + }, + { + "epoch": 0.2626769273052082, + "grad_norm": 1.1352405548095703, + "learning_rate": 4.607807138597568e-05, + "loss": 0.9822, + "step": 3823 + }, + { + "epoch": 0.2627456369382988, + "grad_norm": 1.0998789072036743, + "learning_rate": 4.607484643194788e-05, + "loss": 1.0064, + "step": 3824 + }, + { + "epoch": 0.2628143465713893, + "grad_norm": 1.1811574697494507, + "learning_rate": 4.607162026549701e-05, + "loss": 0.9957, + "step": 3825 + }, + { + "epoch": 0.2628830562044799, + "grad_norm": 1.1120418310165405, + "learning_rate": 4.606839288680868e-05, + "loss": 1.0124, + "step": 3826 + }, + { + "epoch": 0.2629517658375704, + "grad_norm": 1.1521347761154175, + "learning_rate": 4.606516429606855e-05, + "loss": 1.0484, + "step": 3827 + }, + { + "epoch": 0.263020475470661, + "grad_norm": 1.1933423280715942, + "learning_rate": 4.606193449346237e-05, + "loss": 1.018, + "step": 3828 + }, + { + "epoch": 0.26308918510375157, + "grad_norm": 1.184470295906067, + "learning_rate": 4.605870347917594e-05, + "loss": 0.9907, + "step": 3829 + }, + { + "epoch": 0.2631578947368421, + "grad_norm": 1.0993112325668335, + "learning_rate": 4.6055471253395124e-05, + "loss": 0.9759, + "step": 3830 + }, + { + "epoch": 0.26322660436993267, + "grad_norm": 1.0783884525299072, + "learning_rate": 4.605223781630589e-05, + "loss": 1.0059, + "step": 3831 + }, + { + "epoch": 0.26329531400302325, + "grad_norm": 1.196721076965332, + "learning_rate": 4.6049003168094265e-05, + "loss": 1.029, + "step": 3832 + }, + { + "epoch": 0.2633640236361138, + "grad_norm": 1.1461446285247803, + "learning_rate": 4.604576730894631e-05, + "loss": 0.9864, + "step": 3833 + }, + { + "epoch": 0.26343273326920436, + "grad_norm": 1.1134722232818604, + "learning_rate": 4.6042530239048195e-05, + "loss": 0.9333, + "step": 3834 + }, + { + "epoch": 0.2635014429022949, + "grad_norm": 1.2054920196533203, + "learning_rate": 4.603929195858615e-05, + "loss": 1.008, + "step": 3835 + }, + { + "epoch": 0.26357015253538546, + "grad_norm": 1.1642042398452759, + "learning_rate": 4.603605246774646e-05, + "loss": 1.0114, + "step": 3836 + }, + { + "epoch": 0.26363886216847604, + "grad_norm": 1.1759494543075562, + "learning_rate": 4.60328117667155e-05, + "loss": 1.0477, + "step": 3837 + }, + { + "epoch": 0.26370757180156656, + "grad_norm": 1.065551996231079, + "learning_rate": 4.602956985567971e-05, + "loss": 0.9976, + "step": 3838 + }, + { + "epoch": 0.26377628143465714, + "grad_norm": 1.1070902347564697, + "learning_rate": 4.602632673482558e-05, + "loss": 0.9916, + "step": 3839 + }, + { + "epoch": 0.2638449910677477, + "grad_norm": 1.0822365283966064, + "learning_rate": 4.60230824043397e-05, + "loss": 1.0345, + "step": 3840 + }, + { + "epoch": 0.26391370070083825, + "grad_norm": 1.0635693073272705, + "learning_rate": 4.6019836864408705e-05, + "loss": 0.9783, + "step": 3841 + }, + { + "epoch": 0.26398241033392883, + "grad_norm": 1.0859986543655396, + "learning_rate": 4.601659011521931e-05, + "loss": 0.9688, + "step": 3842 + }, + { + "epoch": 0.26405111996701935, + "grad_norm": 1.0574982166290283, + "learning_rate": 4.60133421569583e-05, + "loss": 0.9631, + "step": 3843 + }, + { + "epoch": 0.26411982960010993, + "grad_norm": 1.092757225036621, + "learning_rate": 4.6010092989812534e-05, + "loss": 0.9947, + "step": 3844 + }, + { + "epoch": 0.2641885392332005, + "grad_norm": 1.1404521465301514, + "learning_rate": 4.600684261396893e-05, + "loss": 1.0211, + "step": 3845 + }, + { + "epoch": 0.26425724886629104, + "grad_norm": 1.1739387512207031, + "learning_rate": 4.600359102961447e-05, + "loss": 1.016, + "step": 3846 + }, + { + "epoch": 0.2643259584993816, + "grad_norm": 1.1225405931472778, + "learning_rate": 4.600033823693622e-05, + "loss": 1.0085, + "step": 3847 + }, + { + "epoch": 0.2643946681324722, + "grad_norm": 1.1795785427093506, + "learning_rate": 4.599708423612133e-05, + "loss": 1.0057, + "step": 3848 + }, + { + "epoch": 0.2644633777655627, + "grad_norm": 1.1819217205047607, + "learning_rate": 4.599382902735698e-05, + "loss": 1.0059, + "step": 3849 + }, + { + "epoch": 0.2645320873986533, + "grad_norm": 1.1086320877075195, + "learning_rate": 4.5990572610830454e-05, + "loss": 0.9866, + "step": 3850 + }, + { + "epoch": 0.26460079703174383, + "grad_norm": 1.1040706634521484, + "learning_rate": 4.598731498672907e-05, + "loss": 0.9857, + "step": 3851 + }, + { + "epoch": 0.2646695066648344, + "grad_norm": 1.161350965499878, + "learning_rate": 4.598405615524026e-05, + "loss": 0.9975, + "step": 3852 + }, + { + "epoch": 0.264738216297925, + "grad_norm": 1.1325291395187378, + "learning_rate": 4.59807961165515e-05, + "loss": 0.987, + "step": 3853 + }, + { + "epoch": 0.2648069259310155, + "grad_norm": 1.1202956438064575, + "learning_rate": 4.5977534870850326e-05, + "loss": 1.0689, + "step": 3854 + }, + { + "epoch": 0.2648756355641061, + "grad_norm": 1.119558572769165, + "learning_rate": 4.597427241832435e-05, + "loss": 0.992, + "step": 3855 + }, + { + "epoch": 0.2649443451971967, + "grad_norm": 1.1754555702209473, + "learning_rate": 4.597100875916129e-05, + "loss": 1.0283, + "step": 3856 + }, + { + "epoch": 0.2650130548302872, + "grad_norm": 1.09481942653656, + "learning_rate": 4.5967743893548874e-05, + "loss": 1.0101, + "step": 3857 + }, + { + "epoch": 0.2650817644633778, + "grad_norm": 1.108686923980713, + "learning_rate": 4.5964477821674936e-05, + "loss": 1.0196, + "step": 3858 + }, + { + "epoch": 0.2651504740964683, + "grad_norm": 1.0953853130340576, + "learning_rate": 4.596121054372737e-05, + "loss": 1.0014, + "step": 3859 + }, + { + "epoch": 0.2652191837295589, + "grad_norm": 1.0751186609268188, + "learning_rate": 4.5957942059894144e-05, + "loss": 1.0175, + "step": 3860 + }, + { + "epoch": 0.26528789336264946, + "grad_norm": 1.1819438934326172, + "learning_rate": 4.595467237036329e-05, + "loss": 1.0095, + "step": 3861 + }, + { + "epoch": 0.26535660299574, + "grad_norm": 1.2734198570251465, + "learning_rate": 4.595140147532292e-05, + "loss": 1.0269, + "step": 3862 + }, + { + "epoch": 0.26542531262883057, + "grad_norm": 1.0737522840499878, + "learning_rate": 4.594812937496119e-05, + "loss": 0.9784, + "step": 3863 + }, + { + "epoch": 0.26549402226192115, + "grad_norm": 1.133978247642517, + "learning_rate": 4.5944856069466345e-05, + "loss": 1.0326, + "step": 3864 + }, + { + "epoch": 0.26556273189501167, + "grad_norm": 1.085374355316162, + "learning_rate": 4.5941581559026704e-05, + "loss": 1.0038, + "step": 3865 + }, + { + "epoch": 0.26563144152810225, + "grad_norm": 1.1626068353652954, + "learning_rate": 4.593830584383065e-05, + "loss": 1.0719, + "step": 3866 + }, + { + "epoch": 0.2657001511611928, + "grad_norm": 1.2080262899398804, + "learning_rate": 4.5935028924066626e-05, + "loss": 0.9943, + "step": 3867 + }, + { + "epoch": 0.26576886079428336, + "grad_norm": 1.1458263397216797, + "learning_rate": 4.593175079992315e-05, + "loss": 0.9818, + "step": 3868 + }, + { + "epoch": 0.26583757042737394, + "grad_norm": 1.0450705289840698, + "learning_rate": 4.592847147158881e-05, + "loss": 1.0079, + "step": 3869 + }, + { + "epoch": 0.26590628006046446, + "grad_norm": 1.149052619934082, + "learning_rate": 4.5925190939252275e-05, + "loss": 1.0523, + "step": 3870 + }, + { + "epoch": 0.26597498969355504, + "grad_norm": 1.0741902589797974, + "learning_rate": 4.592190920310226e-05, + "loss": 0.9371, + "step": 3871 + }, + { + "epoch": 0.2660436993266456, + "grad_norm": 1.091770052909851, + "learning_rate": 4.591862626332757e-05, + "loss": 0.9862, + "step": 3872 + }, + { + "epoch": 0.26611240895973615, + "grad_norm": 1.1118284463882446, + "learning_rate": 4.591534212011706e-05, + "loss": 1.0226, + "step": 3873 + }, + { + "epoch": 0.2661811185928267, + "grad_norm": 1.102608561515808, + "learning_rate": 4.591205677365967e-05, + "loss": 1.0169, + "step": 3874 + }, + { + "epoch": 0.26624982822591725, + "grad_norm": 1.0672353506088257, + "learning_rate": 4.59087702241444e-05, + "loss": 0.9925, + "step": 3875 + }, + { + "epoch": 0.26631853785900783, + "grad_norm": 1.1060415506362915, + "learning_rate": 4.590548247176034e-05, + "loss": 1.0104, + "step": 3876 + }, + { + "epoch": 0.2663872474920984, + "grad_norm": 1.1715725660324097, + "learning_rate": 4.590219351669661e-05, + "loss": 1.0206, + "step": 3877 + }, + { + "epoch": 0.26645595712518894, + "grad_norm": 1.1699213981628418, + "learning_rate": 4.589890335914244e-05, + "loss": 1.0201, + "step": 3878 + }, + { + "epoch": 0.2665246667582795, + "grad_norm": 1.1148661375045776, + "learning_rate": 4.5895611999287104e-05, + "loss": 0.996, + "step": 3879 + }, + { + "epoch": 0.2665933763913701, + "grad_norm": 1.1090632677078247, + "learning_rate": 4.589231943731994e-05, + "loss": 1.0384, + "step": 3880 + }, + { + "epoch": 0.2666620860244606, + "grad_norm": 1.1451574563980103, + "learning_rate": 4.588902567343038e-05, + "loss": 1.021, + "step": 3881 + }, + { + "epoch": 0.2667307956575512, + "grad_norm": 1.0571731328964233, + "learning_rate": 4.588573070780792e-05, + "loss": 0.9938, + "step": 3882 + }, + { + "epoch": 0.2667995052906417, + "grad_norm": 1.1585485935211182, + "learning_rate": 4.5882434540642095e-05, + "loss": 1.0361, + "step": 3883 + }, + { + "epoch": 0.2668682149237323, + "grad_norm": 1.010402798652649, + "learning_rate": 4.587913717212255e-05, + "loss": 0.9616, + "step": 3884 + }, + { + "epoch": 0.2669369245568229, + "grad_norm": 1.157664179801941, + "learning_rate": 4.587583860243897e-05, + "loss": 0.9876, + "step": 3885 + }, + { + "epoch": 0.2670056341899134, + "grad_norm": 1.1481939554214478, + "learning_rate": 4.587253883178113e-05, + "loss": 1.0161, + "step": 3886 + }, + { + "epoch": 0.267074343823004, + "grad_norm": 1.1248313188552856, + "learning_rate": 4.586923786033885e-05, + "loss": 0.993, + "step": 3887 + }, + { + "epoch": 0.26714305345609457, + "grad_norm": 1.1278746128082275, + "learning_rate": 4.586593568830205e-05, + "loss": 0.9622, + "step": 3888 + }, + { + "epoch": 0.2672117630891851, + "grad_norm": 1.1932767629623413, + "learning_rate": 4.5862632315860684e-05, + "loss": 0.9887, + "step": 3889 + }, + { + "epoch": 0.2672804727222757, + "grad_norm": 1.0585094690322876, + "learning_rate": 4.5859327743204804e-05, + "loss": 0.9921, + "step": 3890 + }, + { + "epoch": 0.2673491823553662, + "grad_norm": 1.0391663312911987, + "learning_rate": 4.5856021970524524e-05, + "loss": 0.9849, + "step": 3891 + }, + { + "epoch": 0.2674178919884568, + "grad_norm": 1.1364507675170898, + "learning_rate": 4.5852714998010004e-05, + "loss": 1.0009, + "step": 3892 + }, + { + "epoch": 0.26748660162154736, + "grad_norm": 1.125081181526184, + "learning_rate": 4.584940682585152e-05, + "loss": 0.9955, + "step": 3893 + }, + { + "epoch": 0.2675553112546379, + "grad_norm": 1.1399728059768677, + "learning_rate": 4.5846097454239355e-05, + "loss": 1.024, + "step": 3894 + }, + { + "epoch": 0.26762402088772846, + "grad_norm": 1.373766541481018, + "learning_rate": 4.584278688336393e-05, + "loss": 1.0513, + "step": 3895 + }, + { + "epoch": 0.26769273052081904, + "grad_norm": 1.0344990491867065, + "learning_rate": 4.583947511341568e-05, + "loss": 1.0105, + "step": 3896 + }, + { + "epoch": 0.26776144015390957, + "grad_norm": 1.0772488117218018, + "learning_rate": 4.5836162144585125e-05, + "loss": 0.994, + "step": 3897 + }, + { + "epoch": 0.26783014978700015, + "grad_norm": 1.1495451927185059, + "learning_rate": 4.5832847977062874e-05, + "loss": 1.0318, + "step": 3898 + }, + { + "epoch": 0.2678988594200907, + "grad_norm": 1.124179482460022, + "learning_rate": 4.582953261103958e-05, + "loss": 1.0092, + "step": 3899 + }, + { + "epoch": 0.26796756905318125, + "grad_norm": 1.1371355056762695, + "learning_rate": 4.582621604670598e-05, + "loss": 1.0135, + "step": 3900 + }, + { + "epoch": 0.26803627868627183, + "grad_norm": 1.112733006477356, + "learning_rate": 4.5822898284252865e-05, + "loss": 1.0271, + "step": 3901 + }, + { + "epoch": 0.26810498831936236, + "grad_norm": 1.1518539190292358, + "learning_rate": 4.581957932387111e-05, + "loss": 1.0244, + "step": 3902 + }, + { + "epoch": 0.26817369795245294, + "grad_norm": 1.1509110927581787, + "learning_rate": 4.5816259165751654e-05, + "loss": 1.0171, + "step": 3903 + }, + { + "epoch": 0.2682424075855435, + "grad_norm": 1.1147494316101074, + "learning_rate": 4.581293781008549e-05, + "loss": 1.0257, + "step": 3904 + }, + { + "epoch": 0.26831111721863404, + "grad_norm": 1.1361757516860962, + "learning_rate": 4.580961525706371e-05, + "loss": 1.0195, + "step": 3905 + }, + { + "epoch": 0.2683798268517246, + "grad_norm": 1.1038579940795898, + "learning_rate": 4.580629150687745e-05, + "loss": 1.0131, + "step": 3906 + }, + { + "epoch": 0.26844853648481515, + "grad_norm": 1.1247353553771973, + "learning_rate": 4.580296655971793e-05, + "loss": 1.0136, + "step": 3907 + }, + { + "epoch": 0.26851724611790573, + "grad_norm": 1.0608495473861694, + "learning_rate": 4.5799640415776426e-05, + "loss": 1.0032, + "step": 3908 + }, + { + "epoch": 0.2685859557509963, + "grad_norm": 1.068076491355896, + "learning_rate": 4.579631307524429e-05, + "loss": 0.9815, + "step": 3909 + }, + { + "epoch": 0.26865466538408683, + "grad_norm": 1.1208820343017578, + "learning_rate": 4.579298453831294e-05, + "loss": 0.9909, + "step": 3910 + }, + { + "epoch": 0.2687233750171774, + "grad_norm": 1.0381370782852173, + "learning_rate": 4.5789654805173874e-05, + "loss": 0.9646, + "step": 3911 + }, + { + "epoch": 0.268792084650268, + "grad_norm": 1.126429796218872, + "learning_rate": 4.578632387601865e-05, + "loss": 1.0393, + "step": 3912 + }, + { + "epoch": 0.2688607942833585, + "grad_norm": 1.147524118423462, + "learning_rate": 4.578299175103887e-05, + "loss": 1.0206, + "step": 3913 + }, + { + "epoch": 0.2689295039164491, + "grad_norm": 1.1172568798065186, + "learning_rate": 4.577965843042626e-05, + "loss": 1.0371, + "step": 3914 + }, + { + "epoch": 0.2689982135495396, + "grad_norm": 1.1509349346160889, + "learning_rate": 4.577632391437257e-05, + "loss": 0.9821, + "step": 3915 + }, + { + "epoch": 0.2690669231826302, + "grad_norm": 1.1370787620544434, + "learning_rate": 4.577298820306962e-05, + "loss": 1.0416, + "step": 3916 + }, + { + "epoch": 0.2691356328157208, + "grad_norm": 1.0798755884170532, + "learning_rate": 4.5769651296709335e-05, + "loss": 1.0377, + "step": 3917 + }, + { + "epoch": 0.2692043424488113, + "grad_norm": 1.1240307092666626, + "learning_rate": 4.5766313195483676e-05, + "loss": 1.0343, + "step": 3918 + }, + { + "epoch": 0.2692730520819019, + "grad_norm": 1.1378536224365234, + "learning_rate": 4.576297389958468e-05, + "loss": 0.9886, + "step": 3919 + }, + { + "epoch": 0.26934176171499247, + "grad_norm": 1.1024564504623413, + "learning_rate": 4.575963340920445e-05, + "loss": 0.9758, + "step": 3920 + }, + { + "epoch": 0.269410471348083, + "grad_norm": 1.2042298316955566, + "learning_rate": 4.5756291724535174e-05, + "loss": 1.0279, + "step": 3921 + }, + { + "epoch": 0.26947918098117357, + "grad_norm": 1.1709504127502441, + "learning_rate": 4.575294884576909e-05, + "loss": 1.0081, + "step": 3922 + }, + { + "epoch": 0.2695478906142641, + "grad_norm": 1.1177353858947754, + "learning_rate": 4.574960477309851e-05, + "loss": 1.0077, + "step": 3923 + }, + { + "epoch": 0.2696166002473547, + "grad_norm": 1.1453508138656616, + "learning_rate": 4.5746259506715825e-05, + "loss": 1.0211, + "step": 3924 + }, + { + "epoch": 0.26968530988044526, + "grad_norm": 1.1485880613327026, + "learning_rate": 4.574291304681347e-05, + "loss": 0.9978, + "step": 3925 + }, + { + "epoch": 0.2697540195135358, + "grad_norm": 1.0593396425247192, + "learning_rate": 4.5739565393583985e-05, + "loss": 0.997, + "step": 3926 + }, + { + "epoch": 0.26982272914662636, + "grad_norm": 1.1134872436523438, + "learning_rate": 4.5736216547219945e-05, + "loss": 1.0238, + "step": 3927 + }, + { + "epoch": 0.26989143877971694, + "grad_norm": 1.2028940916061401, + "learning_rate": 4.573286650791401e-05, + "loss": 1.0516, + "step": 3928 + }, + { + "epoch": 0.26996014841280747, + "grad_norm": 1.195779800415039, + "learning_rate": 4.57295152758589e-05, + "loss": 1.0087, + "step": 3929 + }, + { + "epoch": 0.27002885804589805, + "grad_norm": 1.1627577543258667, + "learning_rate": 4.572616285124742e-05, + "loss": 1.0351, + "step": 3930 + }, + { + "epoch": 0.27009756767898857, + "grad_norm": 1.046594262123108, + "learning_rate": 4.5722809234272434e-05, + "loss": 0.9924, + "step": 3931 + }, + { + "epoch": 0.27016627731207915, + "grad_norm": 1.1015704870224, + "learning_rate": 4.571945442512686e-05, + "loss": 0.9504, + "step": 3932 + }, + { + "epoch": 0.27023498694516973, + "grad_norm": 1.1681492328643799, + "learning_rate": 4.5716098424003716e-05, + "loss": 1.0082, + "step": 3933 + }, + { + "epoch": 0.27030369657826026, + "grad_norm": 1.0959315299987793, + "learning_rate": 4.571274123109606e-05, + "loss": 1.0066, + "step": 3934 + }, + { + "epoch": 0.27037240621135084, + "grad_norm": 1.0989785194396973, + "learning_rate": 4.570938284659702e-05, + "loss": 0.9831, + "step": 3935 + }, + { + "epoch": 0.2704411158444414, + "grad_norm": 1.2660428285598755, + "learning_rate": 4.570602327069981e-05, + "loss": 1.0074, + "step": 3936 + }, + { + "epoch": 0.27050982547753194, + "grad_norm": 1.170395016670227, + "learning_rate": 4.570266250359773e-05, + "loss": 0.9898, + "step": 3937 + }, + { + "epoch": 0.2705785351106225, + "grad_norm": 1.1537004709243774, + "learning_rate": 4.569930054548408e-05, + "loss": 1.0162, + "step": 3938 + }, + { + "epoch": 0.27064724474371304, + "grad_norm": 1.1113784313201904, + "learning_rate": 4.569593739655229e-05, + "loss": 0.9997, + "step": 3939 + }, + { + "epoch": 0.2707159543768036, + "grad_norm": 1.1355582475662231, + "learning_rate": 4.569257305699585e-05, + "loss": 1.0122, + "step": 3940 + }, + { + "epoch": 0.2707846640098942, + "grad_norm": 1.1423890590667725, + "learning_rate": 4.5689207527008296e-05, + "loss": 0.9913, + "step": 3941 + }, + { + "epoch": 0.27085337364298473, + "grad_norm": 1.1467210054397583, + "learning_rate": 4.568584080678324e-05, + "loss": 0.9997, + "step": 3942 + }, + { + "epoch": 0.2709220832760753, + "grad_norm": 1.1101340055465698, + "learning_rate": 4.568247289651438e-05, + "loss": 1.0412, + "step": 3943 + }, + { + "epoch": 0.2709907929091659, + "grad_norm": 1.1356219053268433, + "learning_rate": 4.567910379639547e-05, + "loss": 0.9998, + "step": 3944 + }, + { + "epoch": 0.2710595025422564, + "grad_norm": 1.1596505641937256, + "learning_rate": 4.5675733506620324e-05, + "loss": 1.0106, + "step": 3945 + }, + { + "epoch": 0.271128212175347, + "grad_norm": 1.127724289894104, + "learning_rate": 4.567236202738283e-05, + "loss": 0.9838, + "step": 3946 + }, + { + "epoch": 0.2711969218084375, + "grad_norm": 1.1878604888916016, + "learning_rate": 4.566898935887696e-05, + "loss": 0.9922, + "step": 3947 + }, + { + "epoch": 0.2712656314415281, + "grad_norm": 1.1560958623886108, + "learning_rate": 4.566561550129673e-05, + "loss": 0.9903, + "step": 3948 + }, + { + "epoch": 0.2713343410746187, + "grad_norm": 1.1889435052871704, + "learning_rate": 4.566224045483625e-05, + "loss": 0.983, + "step": 3949 + }, + { + "epoch": 0.2714030507077092, + "grad_norm": 1.1813679933547974, + "learning_rate": 4.565886421968967e-05, + "loss": 1.0226, + "step": 3950 + }, + { + "epoch": 0.2714717603407998, + "grad_norm": 1.2311233282089233, + "learning_rate": 4.5655486796051225e-05, + "loss": 0.9982, + "step": 3951 + }, + { + "epoch": 0.27154046997389036, + "grad_norm": 1.1707204580307007, + "learning_rate": 4.565210818411522e-05, + "loss": 1.0563, + "step": 3952 + }, + { + "epoch": 0.2716091796069809, + "grad_norm": 1.097758173942566, + "learning_rate": 4.564872838407602e-05, + "loss": 1.0041, + "step": 3953 + }, + { + "epoch": 0.27167788924007147, + "grad_norm": 1.194537878036499, + "learning_rate": 4.564534739612807e-05, + "loss": 1.0102, + "step": 3954 + }, + { + "epoch": 0.271746598873162, + "grad_norm": 1.1750242710113525, + "learning_rate": 4.5641965220465876e-05, + "loss": 1.027, + "step": 3955 + }, + { + "epoch": 0.2718153085062526, + "grad_norm": 1.2508273124694824, + "learning_rate": 4.5638581857284006e-05, + "loss": 1.0215, + "step": 3956 + }, + { + "epoch": 0.27188401813934315, + "grad_norm": 1.213907241821289, + "learning_rate": 4.563519730677711e-05, + "loss": 1.0384, + "step": 3957 + }, + { + "epoch": 0.2719527277724337, + "grad_norm": 1.1155874729156494, + "learning_rate": 4.5631811569139885e-05, + "loss": 1.0179, + "step": 3958 + }, + { + "epoch": 0.27202143740552426, + "grad_norm": 1.0450189113616943, + "learning_rate": 4.5628424644567134e-05, + "loss": 0.9873, + "step": 3959 + }, + { + "epoch": 0.27209014703861484, + "grad_norm": 1.0691815614700317, + "learning_rate": 4.562503653325368e-05, + "loss": 1.012, + "step": 3960 + }, + { + "epoch": 0.27215885667170536, + "grad_norm": 1.0852689743041992, + "learning_rate": 4.5621647235394446e-05, + "loss": 0.9953, + "step": 3961 + }, + { + "epoch": 0.27222756630479594, + "grad_norm": 1.1491062641143799, + "learning_rate": 4.561825675118443e-05, + "loss": 1.02, + "step": 3962 + }, + { + "epoch": 0.27229627593788647, + "grad_norm": 1.1015039682388306, + "learning_rate": 4.561486508081867e-05, + "loss": 0.9931, + "step": 3963 + }, + { + "epoch": 0.27236498557097705, + "grad_norm": 1.0884208679199219, + "learning_rate": 4.5611472224492305e-05, + "loss": 0.9545, + "step": 3964 + }, + { + "epoch": 0.2724336952040676, + "grad_norm": 1.1184202432632446, + "learning_rate": 4.5608078182400496e-05, + "loss": 1.0108, + "step": 3965 + }, + { + "epoch": 0.27250240483715815, + "grad_norm": 1.112719178199768, + "learning_rate": 4.560468295473852e-05, + "loss": 1.0063, + "step": 3966 + }, + { + "epoch": 0.27257111447024873, + "grad_norm": 1.1922105550765991, + "learning_rate": 4.560128654170169e-05, + "loss": 0.9867, + "step": 3967 + }, + { + "epoch": 0.2726398241033393, + "grad_norm": 1.1673957109451294, + "learning_rate": 4.559788894348541e-05, + "loss": 1.0396, + "step": 3968 + }, + { + "epoch": 0.27270853373642984, + "grad_norm": 1.078062891960144, + "learning_rate": 4.559449016028514e-05, + "loss": 1.0256, + "step": 3969 + }, + { + "epoch": 0.2727772433695204, + "grad_norm": 1.1616220474243164, + "learning_rate": 4.5591090192296405e-05, + "loss": 1.0419, + "step": 3970 + }, + { + "epoch": 0.27284595300261094, + "grad_norm": 1.1636219024658203, + "learning_rate": 4.558768903971481e-05, + "loss": 1.0283, + "step": 3971 + }, + { + "epoch": 0.2729146626357015, + "grad_norm": 1.0999870300292969, + "learning_rate": 4.558428670273601e-05, + "loss": 1.0363, + "step": 3972 + }, + { + "epoch": 0.2729833722687921, + "grad_norm": 1.0723776817321777, + "learning_rate": 4.5580883181555755e-05, + "loss": 0.9655, + "step": 3973 + }, + { + "epoch": 0.2730520819018826, + "grad_norm": 1.1668058633804321, + "learning_rate": 4.557747847636984e-05, + "loss": 1.039, + "step": 3974 + }, + { + "epoch": 0.2731207915349732, + "grad_norm": 1.0962657928466797, + "learning_rate": 4.557407258737414e-05, + "loss": 1.0071, + "step": 3975 + }, + { + "epoch": 0.2731895011680638, + "grad_norm": 1.111406683921814, + "learning_rate": 4.557066551476458e-05, + "loss": 1.0153, + "step": 3976 + }, + { + "epoch": 0.2732582108011543, + "grad_norm": 1.1827276945114136, + "learning_rate": 4.556725725873717e-05, + "loss": 1.0737, + "step": 3977 + }, + { + "epoch": 0.2733269204342449, + "grad_norm": 1.1476627588272095, + "learning_rate": 4.5563847819488006e-05, + "loss": 1.0023, + "step": 3978 + }, + { + "epoch": 0.2733956300673354, + "grad_norm": 1.0366765260696411, + "learning_rate": 4.5560437197213203e-05, + "loss": 0.9756, + "step": 3979 + }, + { + "epoch": 0.273464339700426, + "grad_norm": 1.2177717685699463, + "learning_rate": 4.5557025392108986e-05, + "loss": 1.0571, + "step": 3980 + }, + { + "epoch": 0.2735330493335166, + "grad_norm": 1.1580822467803955, + "learning_rate": 4.5553612404371635e-05, + "loss": 1.0352, + "step": 3981 + }, + { + "epoch": 0.2736017589666071, + "grad_norm": 1.1734728813171387, + "learning_rate": 4.555019823419749e-05, + "loss": 1.0249, + "step": 3982 + }, + { + "epoch": 0.2736704685996977, + "grad_norm": 1.1359363794326782, + "learning_rate": 4.554678288178298e-05, + "loss": 1.0329, + "step": 3983 + }, + { + "epoch": 0.27373917823278826, + "grad_norm": 1.1077136993408203, + "learning_rate": 4.554336634732458e-05, + "loss": 1.0163, + "step": 3984 + }, + { + "epoch": 0.2738078878658788, + "grad_norm": 1.091048002243042, + "learning_rate": 4.553994863101882e-05, + "loss": 0.9987, + "step": 3985 + }, + { + "epoch": 0.27387659749896937, + "grad_norm": 1.0927504301071167, + "learning_rate": 4.553652973306236e-05, + "loss": 1.0241, + "step": 3986 + }, + { + "epoch": 0.2739453071320599, + "grad_norm": 1.1223857402801514, + "learning_rate": 4.553310965365186e-05, + "loss": 0.9982, + "step": 3987 + }, + { + "epoch": 0.27401401676515047, + "grad_norm": 1.1371592283248901, + "learning_rate": 4.552968839298408e-05, + "loss": 0.9959, + "step": 3988 + }, + { + "epoch": 0.27408272639824105, + "grad_norm": 1.088913083076477, + "learning_rate": 4.552626595125585e-05, + "loss": 1.0078, + "step": 3989 + }, + { + "epoch": 0.2741514360313316, + "grad_norm": 1.099880576133728, + "learning_rate": 4.552284232866405e-05, + "loss": 1.0052, + "step": 3990 + }, + { + "epoch": 0.27422014566442215, + "grad_norm": 1.0739957094192505, + "learning_rate": 4.5519417525405636e-05, + "loss": 0.968, + "step": 3991 + }, + { + "epoch": 0.27428885529751273, + "grad_norm": 1.1182997226715088, + "learning_rate": 4.551599154167766e-05, + "loss": 1.0521, + "step": 3992 + }, + { + "epoch": 0.27435756493060326, + "grad_norm": 1.1150577068328857, + "learning_rate": 4.551256437767719e-05, + "loss": 1.0125, + "step": 3993 + }, + { + "epoch": 0.27442627456369384, + "grad_norm": 1.196151614189148, + "learning_rate": 4.5509136033601404e-05, + "loss": 1.0603, + "step": 3994 + }, + { + "epoch": 0.27449498419678436, + "grad_norm": 1.0349725484848022, + "learning_rate": 4.550570650964752e-05, + "loss": 0.9604, + "step": 3995 + }, + { + "epoch": 0.27456369382987494, + "grad_norm": 1.1536740064620972, + "learning_rate": 4.550227580601285e-05, + "loss": 0.9787, + "step": 3996 + }, + { + "epoch": 0.2746324034629655, + "grad_norm": 1.0916804075241089, + "learning_rate": 4.5498843922894746e-05, + "loss": 0.9773, + "step": 3997 + }, + { + "epoch": 0.27470111309605605, + "grad_norm": 1.1054619550704956, + "learning_rate": 4.549541086049065e-05, + "loss": 0.999, + "step": 3998 + }, + { + "epoch": 0.27476982272914663, + "grad_norm": 1.1631674766540527, + "learning_rate": 4.5491976618998076e-05, + "loss": 1.0034, + "step": 3999 + }, + { + "epoch": 0.2748385323622372, + "grad_norm": 1.0515494346618652, + "learning_rate": 4.548854119861458e-05, + "loss": 1.0266, + "step": 4000 + }, + { + "epoch": 0.27490724199532773, + "grad_norm": 1.188470721244812, + "learning_rate": 4.548510459953779e-05, + "loss": 1.0156, + "step": 4001 + }, + { + "epoch": 0.2749759516284183, + "grad_norm": 1.0443915128707886, + "learning_rate": 4.548166682196543e-05, + "loss": 0.9996, + "step": 4002 + }, + { + "epoch": 0.27504466126150884, + "grad_norm": 1.135895013809204, + "learning_rate": 4.547822786609526e-05, + "loss": 1.0203, + "step": 4003 + }, + { + "epoch": 0.2751133708945994, + "grad_norm": 1.060324788093567, + "learning_rate": 4.547478773212513e-05, + "loss": 0.9655, + "step": 4004 + }, + { + "epoch": 0.27518208052769, + "grad_norm": 1.0425927639007568, + "learning_rate": 4.5471346420252956e-05, + "loss": 0.9827, + "step": 4005 + }, + { + "epoch": 0.2752507901607805, + "grad_norm": 1.1224939823150635, + "learning_rate": 4.546790393067669e-05, + "loss": 0.9892, + "step": 4006 + }, + { + "epoch": 0.2753194997938711, + "grad_norm": 1.0672608613967896, + "learning_rate": 4.54644602635944e-05, + "loss": 0.9797, + "step": 4007 + }, + { + "epoch": 0.2753882094269617, + "grad_norm": 1.1077567338943481, + "learning_rate": 4.5461015419204187e-05, + "loss": 0.9898, + "step": 4008 + }, + { + "epoch": 0.2754569190600522, + "grad_norm": 1.165834903717041, + "learning_rate": 4.545756939770423e-05, + "loss": 1.0302, + "step": 4009 + }, + { + "epoch": 0.2755256286931428, + "grad_norm": 1.0996370315551758, + "learning_rate": 4.545412219929278e-05, + "loss": 0.9951, + "step": 4010 + }, + { + "epoch": 0.2755943383262333, + "grad_norm": 1.1430490016937256, + "learning_rate": 4.5450673824168155e-05, + "loss": 1.0301, + "step": 4011 + }, + { + "epoch": 0.2756630479593239, + "grad_norm": 1.1339198350906372, + "learning_rate": 4.544722427252873e-05, + "loss": 1.0174, + "step": 4012 + }, + { + "epoch": 0.2757317575924145, + "grad_norm": 1.0856895446777344, + "learning_rate": 4.544377354457296e-05, + "loss": 0.9963, + "step": 4013 + }, + { + "epoch": 0.275800467225505, + "grad_norm": 1.1991052627563477, + "learning_rate": 4.544032164049936e-05, + "loss": 1.0327, + "step": 4014 + }, + { + "epoch": 0.2758691768585956, + "grad_norm": 1.0942062139511108, + "learning_rate": 4.543686856050653e-05, + "loss": 1.0079, + "step": 4015 + }, + { + "epoch": 0.27593788649168616, + "grad_norm": 1.114415168762207, + "learning_rate": 4.5433414304793106e-05, + "loss": 0.9981, + "step": 4016 + }, + { + "epoch": 0.2760065961247767, + "grad_norm": 1.1528691053390503, + "learning_rate": 4.542995887355782e-05, + "loss": 1.0286, + "step": 4017 + }, + { + "epoch": 0.27607530575786726, + "grad_norm": 1.18036949634552, + "learning_rate": 4.542650226699945e-05, + "loss": 1.0307, + "step": 4018 + }, + { + "epoch": 0.2761440153909578, + "grad_norm": 1.1506669521331787, + "learning_rate": 4.542304448531686e-05, + "loss": 1.0359, + "step": 4019 + }, + { + "epoch": 0.27621272502404837, + "grad_norm": 1.0748343467712402, + "learning_rate": 4.5419585528708974e-05, + "loss": 0.9863, + "step": 4020 + }, + { + "epoch": 0.27628143465713895, + "grad_norm": 1.0937004089355469, + "learning_rate": 4.541612539737479e-05, + "loss": 0.9692, + "step": 4021 + }, + { + "epoch": 0.27635014429022947, + "grad_norm": 1.0815670490264893, + "learning_rate": 4.541266409151336e-05, + "loss": 1.0119, + "step": 4022 + }, + { + "epoch": 0.27641885392332005, + "grad_norm": 1.2244293689727783, + "learning_rate": 4.5409201611323805e-05, + "loss": 0.9935, + "step": 4023 + }, + { + "epoch": 0.27648756355641063, + "grad_norm": 1.1890537738800049, + "learning_rate": 4.540573795700532e-05, + "loss": 0.9915, + "step": 4024 + }, + { + "epoch": 0.27655627318950116, + "grad_norm": 1.076858639717102, + "learning_rate": 4.5402273128757184e-05, + "loss": 0.9711, + "step": 4025 + }, + { + "epoch": 0.27662498282259174, + "grad_norm": 1.1862539052963257, + "learning_rate": 4.539880712677871e-05, + "loss": 1.0282, + "step": 4026 + }, + { + "epoch": 0.27669369245568226, + "grad_norm": 1.1314882040023804, + "learning_rate": 4.53953399512693e-05, + "loss": 1.0114, + "step": 4027 + }, + { + "epoch": 0.27676240208877284, + "grad_norm": 1.1191670894622803, + "learning_rate": 4.5391871602428415e-05, + "loss": 0.9628, + "step": 4028 + }, + { + "epoch": 0.2768311117218634, + "grad_norm": 1.0893094539642334, + "learning_rate": 4.538840208045559e-05, + "loss": 1.011, + "step": 4029 + }, + { + "epoch": 0.27689982135495395, + "grad_norm": 1.1589319705963135, + "learning_rate": 4.5384931385550435e-05, + "loss": 0.995, + "step": 4030 + }, + { + "epoch": 0.2769685309880445, + "grad_norm": 1.106321930885315, + "learning_rate": 4.53814595179126e-05, + "loss": 1.0039, + "step": 4031 + }, + { + "epoch": 0.2770372406211351, + "grad_norm": 1.1529529094696045, + "learning_rate": 4.537798647774182e-05, + "loss": 1.0122, + "step": 4032 + }, + { + "epoch": 0.27710595025422563, + "grad_norm": 1.1112743616104126, + "learning_rate": 4.537451226523791e-05, + "loss": 1.0043, + "step": 4033 + }, + { + "epoch": 0.2771746598873162, + "grad_norm": 1.093901515007019, + "learning_rate": 4.5371036880600736e-05, + "loss": 1.0057, + "step": 4034 + }, + { + "epoch": 0.27724336952040673, + "grad_norm": 1.0613317489624023, + "learning_rate": 4.536756032403023e-05, + "loss": 1.0165, + "step": 4035 + }, + { + "epoch": 0.2773120791534973, + "grad_norm": 1.1421502828598022, + "learning_rate": 4.5364082595726396e-05, + "loss": 1.0221, + "step": 4036 + }, + { + "epoch": 0.2773807887865879, + "grad_norm": 1.0926967859268188, + "learning_rate": 4.5360603695889305e-05, + "loss": 0.9885, + "step": 4037 + }, + { + "epoch": 0.2774494984196784, + "grad_norm": 1.1018859148025513, + "learning_rate": 4.53571236247191e-05, + "loss": 0.9848, + "step": 4038 + }, + { + "epoch": 0.277518208052769, + "grad_norm": 1.0962320566177368, + "learning_rate": 4.5353642382415976e-05, + "loss": 1.0183, + "step": 4039 + }, + { + "epoch": 0.2775869176858596, + "grad_norm": 1.0815845727920532, + "learning_rate": 4.535015996918023e-05, + "loss": 0.998, + "step": 4040 + }, + { + "epoch": 0.2776556273189501, + "grad_norm": 1.1513900756835938, + "learning_rate": 4.534667638521218e-05, + "loss": 1.02, + "step": 4041 + }, + { + "epoch": 0.2777243369520407, + "grad_norm": 1.1575769186019897, + "learning_rate": 4.5343191630712244e-05, + "loss": 1.0359, + "step": 4042 + }, + { + "epoch": 0.2777930465851312, + "grad_norm": 1.089937448501587, + "learning_rate": 4.53397057058809e-05, + "loss": 1.02, + "step": 4043 + }, + { + "epoch": 0.2778617562182218, + "grad_norm": 1.1535406112670898, + "learning_rate": 4.533621861091869e-05, + "loss": 1.0149, + "step": 4044 + }, + { + "epoch": 0.27793046585131237, + "grad_norm": 1.1231255531311035, + "learning_rate": 4.533273034602622e-05, + "loss": 0.9924, + "step": 4045 + }, + { + "epoch": 0.2779991754844029, + "grad_norm": 1.0998717546463013, + "learning_rate": 4.532924091140417e-05, + "loss": 1.0234, + "step": 4046 + }, + { + "epoch": 0.2780678851174935, + "grad_norm": 1.150254249572754, + "learning_rate": 4.532575030725329e-05, + "loss": 0.9997, + "step": 4047 + }, + { + "epoch": 0.27813659475058405, + "grad_norm": 1.1257789134979248, + "learning_rate": 4.532225853377439e-05, + "loss": 0.9701, + "step": 4048 + }, + { + "epoch": 0.2782053043836746, + "grad_norm": 1.1631195545196533, + "learning_rate": 4.531876559116834e-05, + "loss": 0.9743, + "step": 4049 + }, + { + "epoch": 0.27827401401676516, + "grad_norm": 1.1170179843902588, + "learning_rate": 4.5315271479636096e-05, + "loss": 1.0561, + "step": 4050 + }, + { + "epoch": 0.2783427236498557, + "grad_norm": 1.0560503005981445, + "learning_rate": 4.531177619937868e-05, + "loss": 1.0115, + "step": 4051 + }, + { + "epoch": 0.27841143328294626, + "grad_norm": 1.1039681434631348, + "learning_rate": 4.530827975059715e-05, + "loss": 1.0428, + "step": 4052 + }, + { + "epoch": 0.27848014291603684, + "grad_norm": 1.1364246606826782, + "learning_rate": 4.5304782133492675e-05, + "loss": 1.0017, + "step": 4053 + }, + { + "epoch": 0.27854885254912737, + "grad_norm": 1.1189976930618286, + "learning_rate": 4.5301283348266466e-05, + "loss": 0.9875, + "step": 4054 + }, + { + "epoch": 0.27861756218221795, + "grad_norm": 1.1663086414337158, + "learning_rate": 4.5297783395119794e-05, + "loss": 0.9948, + "step": 4055 + }, + { + "epoch": 0.27868627181530853, + "grad_norm": 1.1309847831726074, + "learning_rate": 4.529428227425404e-05, + "loss": 0.9944, + "step": 4056 + }, + { + "epoch": 0.27875498144839905, + "grad_norm": 1.1632472276687622, + "learning_rate": 4.5290779985870585e-05, + "loss": 1.0175, + "step": 4057 + }, + { + "epoch": 0.27882369108148963, + "grad_norm": 1.111895203590393, + "learning_rate": 4.5287276530170936e-05, + "loss": 1.0177, + "step": 4058 + }, + { + "epoch": 0.27889240071458016, + "grad_norm": 1.1579540967941284, + "learning_rate": 4.528377190735664e-05, + "loss": 1.0257, + "step": 4059 + }, + { + "epoch": 0.27896111034767074, + "grad_norm": 1.1775087118148804, + "learning_rate": 4.52802661176293e-05, + "loss": 1.009, + "step": 4060 + }, + { + "epoch": 0.2790298199807613, + "grad_norm": 1.2572624683380127, + "learning_rate": 4.527675916119063e-05, + "loss": 0.9877, + "step": 4061 + }, + { + "epoch": 0.27909852961385184, + "grad_norm": 1.1554021835327148, + "learning_rate": 4.5273251038242364e-05, + "loss": 1.038, + "step": 4062 + }, + { + "epoch": 0.2791672392469424, + "grad_norm": 1.0864394903182983, + "learning_rate": 4.5269741748986326e-05, + "loss": 1.0126, + "step": 4063 + }, + { + "epoch": 0.279235948880033, + "grad_norm": 1.1241368055343628, + "learning_rate": 4.526623129362441e-05, + "loss": 1.0073, + "step": 4064 + }, + { + "epoch": 0.2793046585131235, + "grad_norm": 1.1844024658203125, + "learning_rate": 4.526271967235855e-05, + "loss": 1.037, + "step": 4065 + }, + { + "epoch": 0.2793733681462141, + "grad_norm": 1.0612763166427612, + "learning_rate": 4.5259206885390805e-05, + "loss": 0.9833, + "step": 4066 + }, + { + "epoch": 0.27944207777930463, + "grad_norm": 1.0525753498077393, + "learning_rate": 4.525569293292322e-05, + "loss": 0.991, + "step": 4067 + }, + { + "epoch": 0.2795107874123952, + "grad_norm": 1.139453411102295, + "learning_rate": 4.525217781515798e-05, + "loss": 0.9873, + "step": 4068 + }, + { + "epoch": 0.2795794970454858, + "grad_norm": 1.0900031328201294, + "learning_rate": 4.5248661532297296e-05, + "loss": 0.9939, + "step": 4069 + }, + { + "epoch": 0.2796482066785763, + "grad_norm": 1.1510246992111206, + "learning_rate": 4.524514408454347e-05, + "loss": 0.9758, + "step": 4070 + }, + { + "epoch": 0.2797169163116669, + "grad_norm": 1.0697077512741089, + "learning_rate": 4.524162547209883e-05, + "loss": 1.0098, + "step": 4071 + }, + { + "epoch": 0.2797856259447575, + "grad_norm": 1.085476040840149, + "learning_rate": 4.523810569516583e-05, + "loss": 0.969, + "step": 4072 + }, + { + "epoch": 0.279854335577848, + "grad_norm": 1.2024980783462524, + "learning_rate": 4.523458475394696e-05, + "loss": 0.9945, + "step": 4073 + }, + { + "epoch": 0.2799230452109386, + "grad_norm": 1.0604040622711182, + "learning_rate": 4.523106264864475e-05, + "loss": 1.0154, + "step": 4074 + }, + { + "epoch": 0.2799917548440291, + "grad_norm": 1.080061912536621, + "learning_rate": 4.522753937946185e-05, + "loss": 0.9884, + "step": 4075 + }, + { + "epoch": 0.2800604644771197, + "grad_norm": 1.0556787252426147, + "learning_rate": 4.522401494660094e-05, + "loss": 1.0143, + "step": 4076 + }, + { + "epoch": 0.28012917411021027, + "grad_norm": 1.065567135810852, + "learning_rate": 4.522048935026478e-05, + "loss": 0.9907, + "step": 4077 + }, + { + "epoch": 0.2801978837433008, + "grad_norm": 1.1319260597229004, + "learning_rate": 4.52169625906562e-05, + "loss": 0.9965, + "step": 4078 + }, + { + "epoch": 0.28026659337639137, + "grad_norm": 1.1151578426361084, + "learning_rate": 4.5213434667978086e-05, + "loss": 1.0159, + "step": 4079 + }, + { + "epoch": 0.28033530300948195, + "grad_norm": 1.1048837900161743, + "learning_rate": 4.5209905582433396e-05, + "loss": 1.059, + "step": 4080 + }, + { + "epoch": 0.2804040126425725, + "grad_norm": 1.1091341972351074, + "learning_rate": 4.5206375334225174e-05, + "loss": 0.9547, + "step": 4081 + }, + { + "epoch": 0.28047272227566306, + "grad_norm": 1.0529595613479614, + "learning_rate": 4.520284392355649e-05, + "loss": 1.0114, + "step": 4082 + }, + { + "epoch": 0.2805414319087536, + "grad_norm": 1.047479510307312, + "learning_rate": 4.519931135063051e-05, + "loss": 0.9254, + "step": 4083 + }, + { + "epoch": 0.28061014154184416, + "grad_norm": 1.1701987981796265, + "learning_rate": 4.519577761565047e-05, + "loss": 1.0133, + "step": 4084 + }, + { + "epoch": 0.28067885117493474, + "grad_norm": 1.157114028930664, + "learning_rate": 4.519224271881966e-05, + "loss": 1.0065, + "step": 4085 + }, + { + "epoch": 0.28074756080802526, + "grad_norm": 1.1425409317016602, + "learning_rate": 4.518870666034144e-05, + "loss": 1.0078, + "step": 4086 + }, + { + "epoch": 0.28081627044111585, + "grad_norm": 1.1210449934005737, + "learning_rate": 4.518516944041923e-05, + "loss": 1.0052, + "step": 4087 + }, + { + "epoch": 0.2808849800742064, + "grad_norm": 1.1314133405685425, + "learning_rate": 4.518163105925653e-05, + "loss": 1.0273, + "step": 4088 + }, + { + "epoch": 0.28095368970729695, + "grad_norm": 1.088564395904541, + "learning_rate": 4.51780915170569e-05, + "loss": 0.9943, + "step": 4089 + }, + { + "epoch": 0.28102239934038753, + "grad_norm": 1.105067253112793, + "learning_rate": 4.517455081402396e-05, + "loss": 0.9947, + "step": 4090 + }, + { + "epoch": 0.28109110897347805, + "grad_norm": 1.1295734643936157, + "learning_rate": 4.5171008950361426e-05, + "loss": 1.0179, + "step": 4091 + }, + { + "epoch": 0.28115981860656863, + "grad_norm": 1.0503149032592773, + "learning_rate": 4.5167465926273044e-05, + "loss": 1.0147, + "step": 4092 + }, + { + "epoch": 0.2812285282396592, + "grad_norm": 1.1069998741149902, + "learning_rate": 4.5163921741962644e-05, + "loss": 0.9722, + "step": 4093 + }, + { + "epoch": 0.28129723787274974, + "grad_norm": 1.2195279598236084, + "learning_rate": 4.516037639763412e-05, + "loss": 1.0199, + "step": 4094 + }, + { + "epoch": 0.2813659475058403, + "grad_norm": 1.0830111503601074, + "learning_rate": 4.515682989349143e-05, + "loss": 0.9629, + "step": 4095 + }, + { + "epoch": 0.2814346571389309, + "grad_norm": 1.1794121265411377, + "learning_rate": 4.5153282229738616e-05, + "loss": 1.0167, + "step": 4096 + }, + { + "epoch": 0.2815033667720214, + "grad_norm": 1.209815263748169, + "learning_rate": 4.5149733406579754e-05, + "loss": 0.9548, + "step": 4097 + }, + { + "epoch": 0.281572076405112, + "grad_norm": 1.095791220664978, + "learning_rate": 4.5146183424219015e-05, + "loss": 0.9869, + "step": 4098 + }, + { + "epoch": 0.28164078603820253, + "grad_norm": 1.0876537561416626, + "learning_rate": 4.514263228286063e-05, + "loss": 0.987, + "step": 4099 + }, + { + "epoch": 0.2817094956712931, + "grad_norm": 1.111743450164795, + "learning_rate": 4.51390799827089e-05, + "loss": 0.9803, + "step": 4100 + }, + { + "epoch": 0.2817782053043837, + "grad_norm": 1.075046181678772, + "learning_rate": 4.5135526523968164e-05, + "loss": 0.9914, + "step": 4101 + }, + { + "epoch": 0.2818469149374742, + "grad_norm": 1.1164677143096924, + "learning_rate": 4.5131971906842874e-05, + "loss": 0.9969, + "step": 4102 + }, + { + "epoch": 0.2819156245705648, + "grad_norm": 1.1512174606323242, + "learning_rate": 4.512841613153751e-05, + "loss": 1.0078, + "step": 4103 + }, + { + "epoch": 0.2819843342036554, + "grad_norm": 1.0557527542114258, + "learning_rate": 4.512485919825664e-05, + "loss": 1.0018, + "step": 4104 + }, + { + "epoch": 0.2820530438367459, + "grad_norm": 1.0195260047912598, + "learning_rate": 4.5121301107204895e-05, + "loss": 0.9732, + "step": 4105 + }, + { + "epoch": 0.2821217534698365, + "grad_norm": 1.0687204599380493, + "learning_rate": 4.511774185858696e-05, + "loss": 1.0246, + "step": 4106 + }, + { + "epoch": 0.282190463102927, + "grad_norm": 1.2311443090438843, + "learning_rate": 4.511418145260761e-05, + "loss": 1.0317, + "step": 4107 + }, + { + "epoch": 0.2822591727360176, + "grad_norm": 1.1516624689102173, + "learning_rate": 4.5110619889471656e-05, + "loss": 1.0401, + "step": 4108 + }, + { + "epoch": 0.28232788236910816, + "grad_norm": 1.0329241752624512, + "learning_rate": 4.510705716938401e-05, + "loss": 0.9568, + "step": 4109 + }, + { + "epoch": 0.2823965920021987, + "grad_norm": 1.108432650566101, + "learning_rate": 4.5103493292549615e-05, + "loss": 0.9962, + "step": 4110 + }, + { + "epoch": 0.28246530163528927, + "grad_norm": 1.0812965631484985, + "learning_rate": 4.5099928259173516e-05, + "loss": 1.0093, + "step": 4111 + }, + { + "epoch": 0.28253401126837985, + "grad_norm": 1.0794544219970703, + "learning_rate": 4.5096362069460795e-05, + "loss": 0.9944, + "step": 4112 + }, + { + "epoch": 0.2826027209014704, + "grad_norm": 1.133633017539978, + "learning_rate": 4.509279472361663e-05, + "loss": 0.9827, + "step": 4113 + }, + { + "epoch": 0.28267143053456095, + "grad_norm": 1.0940347909927368, + "learning_rate": 4.508922622184622e-05, + "loss": 1.0344, + "step": 4114 + }, + { + "epoch": 0.28274014016765153, + "grad_norm": 1.1390225887298584, + "learning_rate": 4.5085656564354885e-05, + "loss": 1.0012, + "step": 4115 + }, + { + "epoch": 0.28280884980074206, + "grad_norm": 1.074709415435791, + "learning_rate": 4.508208575134797e-05, + "loss": 0.968, + "step": 4116 + }, + { + "epoch": 0.28287755943383264, + "grad_norm": 1.1907309293746948, + "learning_rate": 4.507851378303091e-05, + "loss": 1.0105, + "step": 4117 + }, + { + "epoch": 0.28294626906692316, + "grad_norm": 1.063205361366272, + "learning_rate": 4.507494065960919e-05, + "loss": 0.9919, + "step": 4118 + }, + { + "epoch": 0.28301497870001374, + "grad_norm": 1.0945051908493042, + "learning_rate": 4.5071366381288375e-05, + "loss": 0.9815, + "step": 4119 + }, + { + "epoch": 0.2830836883331043, + "grad_norm": 1.118027687072754, + "learning_rate": 4.5067790948274094e-05, + "loss": 0.9958, + "step": 4120 + }, + { + "epoch": 0.28315239796619485, + "grad_norm": 1.1101188659667969, + "learning_rate": 4.506421436077203e-05, + "loss": 0.9755, + "step": 4121 + }, + { + "epoch": 0.2832211075992854, + "grad_norm": 1.1141915321350098, + "learning_rate": 4.506063661898795e-05, + "loss": 1.0157, + "step": 4122 + }, + { + "epoch": 0.283289817232376, + "grad_norm": 1.1089519262313843, + "learning_rate": 4.505705772312768e-05, + "loss": 0.9902, + "step": 4123 + }, + { + "epoch": 0.28335852686546653, + "grad_norm": 1.11955726146698, + "learning_rate": 4.505347767339711e-05, + "loss": 1.0246, + "step": 4124 + }, + { + "epoch": 0.2834272364985571, + "grad_norm": 1.065928339958191, + "learning_rate": 4.50498964700022e-05, + "loss": 1.0078, + "step": 4125 + }, + { + "epoch": 0.28349594613164764, + "grad_norm": 1.0710283517837524, + "learning_rate": 4.504631411314896e-05, + "loss": 0.9659, + "step": 4126 + }, + { + "epoch": 0.2835646557647382, + "grad_norm": 1.065206527709961, + "learning_rate": 4.504273060304349e-05, + "loss": 1.0079, + "step": 4127 + }, + { + "epoch": 0.2836333653978288, + "grad_norm": 1.0925087928771973, + "learning_rate": 4.503914593989196e-05, + "loss": 1.0124, + "step": 4128 + }, + { + "epoch": 0.2837020750309193, + "grad_norm": 1.0552291870117188, + "learning_rate": 4.5035560123900575e-05, + "loss": 1.0027, + "step": 4129 + }, + { + "epoch": 0.2837707846640099, + "grad_norm": 1.0638177394866943, + "learning_rate": 4.5031973155275645e-05, + "loss": 0.9847, + "step": 4130 + }, + { + "epoch": 0.2838394942971005, + "grad_norm": 1.1584256887435913, + "learning_rate": 4.5028385034223495e-05, + "loss": 1.0304, + "step": 4131 + }, + { + "epoch": 0.283908203930191, + "grad_norm": 1.0626139640808105, + "learning_rate": 4.502479576095058e-05, + "loss": 0.9709, + "step": 4132 + }, + { + "epoch": 0.2839769135632816, + "grad_norm": 1.0976110696792603, + "learning_rate": 4.502120533566337e-05, + "loss": 0.9573, + "step": 4133 + }, + { + "epoch": 0.2840456231963721, + "grad_norm": 1.1014277935028076, + "learning_rate": 4.501761375856842e-05, + "loss": 1.0032, + "step": 4134 + }, + { + "epoch": 0.2841143328294627, + "grad_norm": 1.0749835968017578, + "learning_rate": 4.501402102987237e-05, + "loss": 1.0127, + "step": 4135 + }, + { + "epoch": 0.28418304246255327, + "grad_norm": 1.082747459411621, + "learning_rate": 4.501042714978187e-05, + "loss": 0.9853, + "step": 4136 + }, + { + "epoch": 0.2842517520956438, + "grad_norm": 1.1676241159439087, + "learning_rate": 4.5006832118503716e-05, + "loss": 1.0249, + "step": 4137 + }, + { + "epoch": 0.2843204617287344, + "grad_norm": 1.1284387111663818, + "learning_rate": 4.5003235936244706e-05, + "loss": 0.9732, + "step": 4138 + }, + { + "epoch": 0.28438917136182496, + "grad_norm": 1.1603871583938599, + "learning_rate": 4.499963860321173e-05, + "loss": 1.0176, + "step": 4139 + }, + { + "epoch": 0.2844578809949155, + "grad_norm": 1.1577892303466797, + "learning_rate": 4.4996040119611735e-05, + "loss": 1.0227, + "step": 4140 + }, + { + "epoch": 0.28452659062800606, + "grad_norm": 1.082183837890625, + "learning_rate": 4.499244048565175e-05, + "loss": 1.0292, + "step": 4141 + }, + { + "epoch": 0.2845953002610966, + "grad_norm": 1.154434084892273, + "learning_rate": 4.498883970153885e-05, + "loss": 1.0114, + "step": 4142 + }, + { + "epoch": 0.28466400989418716, + "grad_norm": 1.125007152557373, + "learning_rate": 4.498523776748019e-05, + "loss": 0.9898, + "step": 4143 + }, + { + "epoch": 0.28473271952727774, + "grad_norm": 1.1017076969146729, + "learning_rate": 4.4981634683682985e-05, + "loss": 1.0224, + "step": 4144 + }, + { + "epoch": 0.28480142916036827, + "grad_norm": 1.051255464553833, + "learning_rate": 4.497803045035452e-05, + "loss": 1.0003, + "step": 4145 + }, + { + "epoch": 0.28487013879345885, + "grad_norm": 1.2016234397888184, + "learning_rate": 4.497442506770214e-05, + "loss": 1.0363, + "step": 4146 + }, + { + "epoch": 0.28493884842654943, + "grad_norm": 1.1330174207687378, + "learning_rate": 4.497081853593328e-05, + "loss": 1.0134, + "step": 4147 + }, + { + "epoch": 0.28500755805963995, + "grad_norm": 1.1299785375595093, + "learning_rate": 4.4967210855255394e-05, + "loss": 1.0174, + "step": 4148 + }, + { + "epoch": 0.28507626769273053, + "grad_norm": 1.1047329902648926, + "learning_rate": 4.4963602025876047e-05, + "loss": 0.9922, + "step": 4149 + }, + { + "epoch": 0.28514497732582106, + "grad_norm": 1.112443447113037, + "learning_rate": 4.495999204800285e-05, + "loss": 0.9947, + "step": 4150 + }, + { + "epoch": 0.28521368695891164, + "grad_norm": 1.1152596473693848, + "learning_rate": 4.495638092184348e-05, + "loss": 1.0245, + "step": 4151 + }, + { + "epoch": 0.2852823965920022, + "grad_norm": 1.0625451803207397, + "learning_rate": 4.4952768647605674e-05, + "loss": 0.9899, + "step": 4152 + }, + { + "epoch": 0.28535110622509274, + "grad_norm": 1.0922232866287231, + "learning_rate": 4.4949155225497264e-05, + "loss": 1.024, + "step": 4153 + }, + { + "epoch": 0.2854198158581833, + "grad_norm": 1.1093908548355103, + "learning_rate": 4.494554065572611e-05, + "loss": 1.0052, + "step": 4154 + }, + { + "epoch": 0.2854885254912739, + "grad_norm": 1.07260262966156, + "learning_rate": 4.494192493850017e-05, + "loss": 0.9667, + "step": 4155 + }, + { + "epoch": 0.28555723512436443, + "grad_norm": 1.0966236591339111, + "learning_rate": 4.4938308074027446e-05, + "loss": 0.9797, + "step": 4156 + }, + { + "epoch": 0.285625944757455, + "grad_norm": 1.0334473848342896, + "learning_rate": 4.493469006251601e-05, + "loss": 1.0024, + "step": 4157 + }, + { + "epoch": 0.28569465439054553, + "grad_norm": 1.0887854099273682, + "learning_rate": 4.493107090417402e-05, + "loss": 0.9665, + "step": 4158 + }, + { + "epoch": 0.2857633640236361, + "grad_norm": 1.1673898696899414, + "learning_rate": 4.492745059920966e-05, + "loss": 0.9951, + "step": 4159 + }, + { + "epoch": 0.2858320736567267, + "grad_norm": 1.0727537870407104, + "learning_rate": 4.492382914783123e-05, + "loss": 1.0234, + "step": 4160 + }, + { + "epoch": 0.2859007832898172, + "grad_norm": 1.0601615905761719, + "learning_rate": 4.492020655024706e-05, + "loss": 0.99, + "step": 4161 + }, + { + "epoch": 0.2859694929229078, + "grad_norm": 1.1537638902664185, + "learning_rate": 4.491658280666554e-05, + "loss": 0.9855, + "step": 4162 + }, + { + "epoch": 0.2860382025559984, + "grad_norm": 1.1851884126663208, + "learning_rate": 4.491295791729516e-05, + "loss": 1.02, + "step": 4163 + }, + { + "epoch": 0.2861069121890889, + "grad_norm": 1.3083772659301758, + "learning_rate": 4.4909331882344455e-05, + "loss": 1.0468, + "step": 4164 + }, + { + "epoch": 0.2861756218221795, + "grad_norm": 1.175761342048645, + "learning_rate": 4.490570470202202e-05, + "loss": 1.0209, + "step": 4165 + }, + { + "epoch": 0.28624433145527, + "grad_norm": 1.154751181602478, + "learning_rate": 4.490207637653654e-05, + "loss": 0.9578, + "step": 4166 + }, + { + "epoch": 0.2863130410883606, + "grad_norm": 1.3282426595687866, + "learning_rate": 4.4898446906096734e-05, + "loss": 1.0214, + "step": 4167 + }, + { + "epoch": 0.28638175072145117, + "grad_norm": 1.0871336460113525, + "learning_rate": 4.489481629091141e-05, + "loss": 0.9806, + "step": 4168 + }, + { + "epoch": 0.2864504603545417, + "grad_norm": 1.2090984582901, + "learning_rate": 4.489118453118944e-05, + "loss": 1.0142, + "step": 4169 + }, + { + "epoch": 0.28651916998763227, + "grad_norm": 1.1065027713775635, + "learning_rate": 4.488755162713975e-05, + "loss": 1.0239, + "step": 4170 + }, + { + "epoch": 0.28658787962072285, + "grad_norm": 1.1150673627853394, + "learning_rate": 4.4883917578971345e-05, + "loss": 1.0213, + "step": 4171 + }, + { + "epoch": 0.2866565892538134, + "grad_norm": 1.106276035308838, + "learning_rate": 4.488028238689328e-05, + "loss": 1.0242, + "step": 4172 + }, + { + "epoch": 0.28672529888690396, + "grad_norm": 1.1969269514083862, + "learning_rate": 4.4876646051114694e-05, + "loss": 1.0799, + "step": 4173 + }, + { + "epoch": 0.2867940085199945, + "grad_norm": 1.0372074842453003, + "learning_rate": 4.4873008571844774e-05, + "loss": 0.9632, + "step": 4174 + }, + { + "epoch": 0.28686271815308506, + "grad_norm": 1.131528377532959, + "learning_rate": 4.486936994929279e-05, + "loss": 0.9993, + "step": 4175 + }, + { + "epoch": 0.28693142778617564, + "grad_norm": 1.1060341596603394, + "learning_rate": 4.486573018366808e-05, + "loss": 1.0332, + "step": 4176 + }, + { + "epoch": 0.28700013741926617, + "grad_norm": 1.17742121219635, + "learning_rate": 4.486208927518001e-05, + "loss": 1.0012, + "step": 4177 + }, + { + "epoch": 0.28706884705235675, + "grad_norm": 1.063137412071228, + "learning_rate": 4.485844722403806e-05, + "loss": 0.9701, + "step": 4178 + }, + { + "epoch": 0.2871375566854473, + "grad_norm": 1.1273547410964966, + "learning_rate": 4.485480403045175e-05, + "loss": 1.0374, + "step": 4179 + }, + { + "epoch": 0.28720626631853785, + "grad_norm": 1.144516110420227, + "learning_rate": 4.4851159694630676e-05, + "loss": 1.0066, + "step": 4180 + }, + { + "epoch": 0.28727497595162843, + "grad_norm": 1.1688176393508911, + "learning_rate": 4.484751421678448e-05, + "loss": 1.0209, + "step": 4181 + }, + { + "epoch": 0.28734368558471896, + "grad_norm": 1.155029058456421, + "learning_rate": 4.484386759712289e-05, + "loss": 0.9881, + "step": 4182 + }, + { + "epoch": 0.28741239521780954, + "grad_norm": 1.1125619411468506, + "learning_rate": 4.484021983585571e-05, + "loss": 0.9758, + "step": 4183 + }, + { + "epoch": 0.2874811048509001, + "grad_norm": 1.1705204248428345, + "learning_rate": 4.483657093319277e-05, + "loss": 1.0513, + "step": 4184 + }, + { + "epoch": 0.28754981448399064, + "grad_norm": 1.1857069730758667, + "learning_rate": 4.4832920889344e-05, + "loss": 1.0384, + "step": 4185 + }, + { + "epoch": 0.2876185241170812, + "grad_norm": 1.1383851766586304, + "learning_rate": 4.4829269704519386e-05, + "loss": 0.9989, + "step": 4186 + }, + { + "epoch": 0.2876872337501718, + "grad_norm": 1.1481760740280151, + "learning_rate": 4.482561737892898e-05, + "loss": 0.9738, + "step": 4187 + }, + { + "epoch": 0.2877559433832623, + "grad_norm": 1.1552090644836426, + "learning_rate": 4.482196391278289e-05, + "loss": 1.0525, + "step": 4188 + }, + { + "epoch": 0.2878246530163529, + "grad_norm": 1.1319018602371216, + "learning_rate": 4.4818309306291304e-05, + "loss": 1.0502, + "step": 4189 + }, + { + "epoch": 0.28789336264944343, + "grad_norm": 1.1363571882247925, + "learning_rate": 4.481465355966446e-05, + "loss": 0.9885, + "step": 4190 + }, + { + "epoch": 0.287962072282534, + "grad_norm": 1.1719588041305542, + "learning_rate": 4.4810996673112686e-05, + "loss": 1.0564, + "step": 4191 + }, + { + "epoch": 0.2880307819156246, + "grad_norm": 1.1119521856307983, + "learning_rate": 4.480733864684635e-05, + "loss": 1.0268, + "step": 4192 + }, + { + "epoch": 0.2880994915487151, + "grad_norm": 1.0732192993164062, + "learning_rate": 4.48036794810759e-05, + "loss": 0.9966, + "step": 4193 + }, + { + "epoch": 0.2881682011818057, + "grad_norm": 1.1937979459762573, + "learning_rate": 4.480001917601185e-05, + "loss": 1.0146, + "step": 4194 + }, + { + "epoch": 0.2882369108148963, + "grad_norm": 1.1034915447235107, + "learning_rate": 4.4796357731864757e-05, + "loss": 0.9781, + "step": 4195 + }, + { + "epoch": 0.2883056204479868, + "grad_norm": 1.1268372535705566, + "learning_rate": 4.479269514884529e-05, + "loss": 1.0128, + "step": 4196 + }, + { + "epoch": 0.2883743300810774, + "grad_norm": 1.095152497291565, + "learning_rate": 4.478903142716413e-05, + "loss": 0.9528, + "step": 4197 + }, + { + "epoch": 0.2884430397141679, + "grad_norm": 1.112098217010498, + "learning_rate": 4.478536656703206e-05, + "loss": 0.9914, + "step": 4198 + }, + { + "epoch": 0.2885117493472585, + "grad_norm": 1.0238338708877563, + "learning_rate": 4.478170056865991e-05, + "loss": 0.9952, + "step": 4199 + }, + { + "epoch": 0.28858045898034906, + "grad_norm": 1.1795321702957153, + "learning_rate": 4.477803343225859e-05, + "loss": 1.0344, + "step": 4200 + }, + { + "epoch": 0.2886491686134396, + "grad_norm": 1.117147445678711, + "learning_rate": 4.477436515803907e-05, + "loss": 1.0169, + "step": 4201 + }, + { + "epoch": 0.28871787824653017, + "grad_norm": 1.1725022792816162, + "learning_rate": 4.477069574621238e-05, + "loss": 1.0043, + "step": 4202 + }, + { + "epoch": 0.28878658787962075, + "grad_norm": 1.0954171419143677, + "learning_rate": 4.476702519698961e-05, + "loss": 1.0309, + "step": 4203 + }, + { + "epoch": 0.2888552975127113, + "grad_norm": 1.1775529384613037, + "learning_rate": 4.476335351058194e-05, + "loss": 1.0213, + "step": 4204 + }, + { + "epoch": 0.28892400714580185, + "grad_norm": 1.093569278717041, + "learning_rate": 4.475968068720059e-05, + "loss": 1.0252, + "step": 4205 + }, + { + "epoch": 0.2889927167788924, + "grad_norm": 1.0889227390289307, + "learning_rate": 4.475600672705686e-05, + "loss": 0.9827, + "step": 4206 + }, + { + "epoch": 0.28906142641198296, + "grad_norm": 1.0935287475585938, + "learning_rate": 4.475233163036211e-05, + "loss": 1.0182, + "step": 4207 + }, + { + "epoch": 0.28913013604507354, + "grad_norm": 1.180214524269104, + "learning_rate": 4.4748655397327764e-05, + "loss": 0.9857, + "step": 4208 + }, + { + "epoch": 0.28919884567816406, + "grad_norm": 1.1144886016845703, + "learning_rate": 4.474497802816532e-05, + "loss": 0.9765, + "step": 4209 + }, + { + "epoch": 0.28926755531125464, + "grad_norm": 1.079519510269165, + "learning_rate": 4.4741299523086323e-05, + "loss": 0.9735, + "step": 4210 + }, + { + "epoch": 0.2893362649443452, + "grad_norm": 1.145896315574646, + "learning_rate": 4.47376198823024e-05, + "loss": 1.053, + "step": 4211 + }, + { + "epoch": 0.28940497457743575, + "grad_norm": 1.1376031637191772, + "learning_rate": 4.4733939106025244e-05, + "loss": 1.043, + "step": 4212 + }, + { + "epoch": 0.2894736842105263, + "grad_norm": 1.0716915130615234, + "learning_rate": 4.47302571944666e-05, + "loss": 0.9846, + "step": 4213 + }, + { + "epoch": 0.28954239384361685, + "grad_norm": 1.121251106262207, + "learning_rate": 4.472657414783829e-05, + "loss": 1.0067, + "step": 4214 + }, + { + "epoch": 0.28961110347670743, + "grad_norm": 1.063905119895935, + "learning_rate": 4.4722889966352197e-05, + "loss": 0.984, + "step": 4215 + }, + { + "epoch": 0.289679813109798, + "grad_norm": 1.074385643005371, + "learning_rate": 4.471920465022027e-05, + "loss": 0.9853, + "step": 4216 + }, + { + "epoch": 0.28974852274288854, + "grad_norm": 1.1462568044662476, + "learning_rate": 4.471551819965453e-05, + "loss": 1.0733, + "step": 4217 + }, + { + "epoch": 0.2898172323759791, + "grad_norm": 1.0978622436523438, + "learning_rate": 4.471183061486704e-05, + "loss": 1.0574, + "step": 4218 + }, + { + "epoch": 0.2898859420090697, + "grad_norm": 1.102986454963684, + "learning_rate": 4.470814189606995e-05, + "loss": 1.0179, + "step": 4219 + }, + { + "epoch": 0.2899546516421602, + "grad_norm": 1.2101283073425293, + "learning_rate": 4.4704452043475485e-05, + "loss": 1.038, + "step": 4220 + }, + { + "epoch": 0.2900233612752508, + "grad_norm": 1.027475118637085, + "learning_rate": 4.47007610572959e-05, + "loss": 0.9508, + "step": 4221 + }, + { + "epoch": 0.2900920709083413, + "grad_norm": 1.1581770181655884, + "learning_rate": 4.469706893774356e-05, + "loss": 1.0264, + "step": 4222 + }, + { + "epoch": 0.2901607805414319, + "grad_norm": 1.0908185243606567, + "learning_rate": 4.4693375685030835e-05, + "loss": 1.0467, + "step": 4223 + }, + { + "epoch": 0.2902294901745225, + "grad_norm": 1.0499666929244995, + "learning_rate": 4.4689681299370224e-05, + "loss": 0.9982, + "step": 4224 + }, + { + "epoch": 0.290298199807613, + "grad_norm": 1.1103383302688599, + "learning_rate": 4.468598578097425e-05, + "loss": 1.0089, + "step": 4225 + }, + { + "epoch": 0.2903669094407036, + "grad_norm": 1.0724600553512573, + "learning_rate": 4.468228913005551e-05, + "loss": 0.9751, + "step": 4226 + }, + { + "epoch": 0.29043561907379417, + "grad_norm": 1.1937885284423828, + "learning_rate": 4.467859134682669e-05, + "loss": 1.0517, + "step": 4227 + }, + { + "epoch": 0.2905043287068847, + "grad_norm": 1.1466789245605469, + "learning_rate": 4.4674892431500504e-05, + "loss": 1.0171, + "step": 4228 + }, + { + "epoch": 0.2905730383399753, + "grad_norm": 1.0984835624694824, + "learning_rate": 4.467119238428975e-05, + "loss": 1.007, + "step": 4229 + }, + { + "epoch": 0.2906417479730658, + "grad_norm": 1.101624846458435, + "learning_rate": 4.46674912054073e-05, + "loss": 1.0545, + "step": 4230 + }, + { + "epoch": 0.2907104576061564, + "grad_norm": 1.0800073146820068, + "learning_rate": 4.466378889506607e-05, + "loss": 1.0397, + "step": 4231 + }, + { + "epoch": 0.29077916723924696, + "grad_norm": 1.1678065061569214, + "learning_rate": 4.4660085453479046e-05, + "loss": 0.9872, + "step": 4232 + }, + { + "epoch": 0.2908478768723375, + "grad_norm": 1.158882975578308, + "learning_rate": 4.46563808808593e-05, + "loss": 1.0545, + "step": 4233 + }, + { + "epoch": 0.29091658650542807, + "grad_norm": 1.0705369710922241, + "learning_rate": 4.4652675177419947e-05, + "loss": 0.9535, + "step": 4234 + }, + { + "epoch": 0.29098529613851865, + "grad_norm": 1.0900273323059082, + "learning_rate": 4.464896834337418e-05, + "loss": 1.0157, + "step": 4235 + }, + { + "epoch": 0.29105400577160917, + "grad_norm": 1.1000492572784424, + "learning_rate": 4.464526037893523e-05, + "loss": 1.0325, + "step": 4236 + }, + { + "epoch": 0.29112271540469975, + "grad_norm": 1.117544412612915, + "learning_rate": 4.4641551284316443e-05, + "loss": 1.0111, + "step": 4237 + }, + { + "epoch": 0.2911914250377903, + "grad_norm": 1.1328917741775513, + "learning_rate": 4.463784105973118e-05, + "loss": 0.9651, + "step": 4238 + }, + { + "epoch": 0.29126013467088085, + "grad_norm": 1.0799626111984253, + "learning_rate": 4.463412970539289e-05, + "loss": 1.0065, + "step": 4239 + }, + { + "epoch": 0.29132884430397143, + "grad_norm": 1.1403864622116089, + "learning_rate": 4.46304172215151e-05, + "loss": 0.9863, + "step": 4240 + }, + { + "epoch": 0.29139755393706196, + "grad_norm": 1.182726502418518, + "learning_rate": 4.462670360831137e-05, + "loss": 1.0177, + "step": 4241 + }, + { + "epoch": 0.29146626357015254, + "grad_norm": 1.0760945081710815, + "learning_rate": 4.462298886599535e-05, + "loss": 0.9926, + "step": 4242 + }, + { + "epoch": 0.2915349732032431, + "grad_norm": 1.2148855924606323, + "learning_rate": 4.461927299478074e-05, + "loss": 1.0167, + "step": 4243 + }, + { + "epoch": 0.29160368283633364, + "grad_norm": 1.0774763822555542, + "learning_rate": 4.461555599488132e-05, + "loss": 1.0189, + "step": 4244 + }, + { + "epoch": 0.2916723924694242, + "grad_norm": 1.0960031747817993, + "learning_rate": 4.461183786651093e-05, + "loss": 1.0096, + "step": 4245 + }, + { + "epoch": 0.29174110210251475, + "grad_norm": 1.1550471782684326, + "learning_rate": 4.460811860988346e-05, + "loss": 0.9941, + "step": 4246 + }, + { + "epoch": 0.29180981173560533, + "grad_norm": 1.1092621088027954, + "learning_rate": 4.4604398225212885e-05, + "loss": 0.9902, + "step": 4247 + }, + { + "epoch": 0.2918785213686959, + "grad_norm": 1.1091234683990479, + "learning_rate": 4.460067671271323e-05, + "loss": 0.9945, + "step": 4248 + }, + { + "epoch": 0.29194723100178643, + "grad_norm": 1.04619562625885, + "learning_rate": 4.45969540725986e-05, + "loss": 0.9798, + "step": 4249 + }, + { + "epoch": 0.292015940634877, + "grad_norm": 1.070877194404602, + "learning_rate": 4.459323030508314e-05, + "loss": 0.9937, + "step": 4250 + }, + { + "epoch": 0.2920846502679676, + "grad_norm": 1.0271501541137695, + "learning_rate": 4.4589505410381104e-05, + "loss": 0.9491, + "step": 4251 + }, + { + "epoch": 0.2921533599010581, + "grad_norm": 1.1524114608764648, + "learning_rate": 4.458577938870676e-05, + "loss": 1.0181, + "step": 4252 + }, + { + "epoch": 0.2922220695341487, + "grad_norm": 1.2039520740509033, + "learning_rate": 4.4582052240274465e-05, + "loss": 1.0119, + "step": 4253 + }, + { + "epoch": 0.2922907791672392, + "grad_norm": 1.09976065158844, + "learning_rate": 4.4578323965298645e-05, + "loss": 0.9781, + "step": 4254 + }, + { + "epoch": 0.2923594888003298, + "grad_norm": 1.2249151468276978, + "learning_rate": 4.4574594563993786e-05, + "loss": 1.0548, + "step": 4255 + }, + { + "epoch": 0.2924281984334204, + "grad_norm": 1.1694742441177368, + "learning_rate": 4.457086403657444e-05, + "loss": 0.9874, + "step": 4256 + }, + { + "epoch": 0.2924969080665109, + "grad_norm": 1.1710134744644165, + "learning_rate": 4.456713238325522e-05, + "loss": 1.0007, + "step": 4257 + }, + { + "epoch": 0.2925656176996015, + "grad_norm": 1.123995304107666, + "learning_rate": 4.45633996042508e-05, + "loss": 1.0115, + "step": 4258 + }, + { + "epoch": 0.29263432733269207, + "grad_norm": 1.2278462648391724, + "learning_rate": 4.4559665699775935e-05, + "loss": 1.0354, + "step": 4259 + }, + { + "epoch": 0.2927030369657826, + "grad_norm": 1.138338565826416, + "learning_rate": 4.455593067004542e-05, + "loss": 1.0029, + "step": 4260 + }, + { + "epoch": 0.2927717465988732, + "grad_norm": 1.1174495220184326, + "learning_rate": 4.455219451527414e-05, + "loss": 0.9954, + "step": 4261 + }, + { + "epoch": 0.2928404562319637, + "grad_norm": 1.0478988885879517, + "learning_rate": 4.4548457235677036e-05, + "loss": 0.9901, + "step": 4262 + }, + { + "epoch": 0.2929091658650543, + "grad_norm": 1.0883114337921143, + "learning_rate": 4.45447188314691e-05, + "loss": 0.958, + "step": 4263 + }, + { + "epoch": 0.29297787549814486, + "grad_norm": 1.1907122135162354, + "learning_rate": 4.454097930286541e-05, + "loss": 0.9758, + "step": 4264 + }, + { + "epoch": 0.2930465851312354, + "grad_norm": 1.0990794897079468, + "learning_rate": 4.453723865008109e-05, + "loss": 0.9759, + "step": 4265 + }, + { + "epoch": 0.29311529476432596, + "grad_norm": 1.1079870462417603, + "learning_rate": 4.4533496873331354e-05, + "loss": 1.0088, + "step": 4266 + }, + { + "epoch": 0.29318400439741654, + "grad_norm": 1.1602200269699097, + "learning_rate": 4.452975397283145e-05, + "loss": 0.9818, + "step": 4267 + }, + { + "epoch": 0.29325271403050707, + "grad_norm": 1.070995807647705, + "learning_rate": 4.4526009948796703e-05, + "loss": 0.983, + "step": 4268 + }, + { + "epoch": 0.29332142366359765, + "grad_norm": 1.131775975227356, + "learning_rate": 4.452226480144252e-05, + "loss": 0.9741, + "step": 4269 + }, + { + "epoch": 0.29339013329668817, + "grad_norm": 1.1642851829528809, + "learning_rate": 4.4518518530984334e-05, + "loss": 0.972, + "step": 4270 + }, + { + "epoch": 0.29345884292977875, + "grad_norm": 1.117569088935852, + "learning_rate": 4.451477113763768e-05, + "loss": 1.0201, + "step": 4271 + }, + { + "epoch": 0.29352755256286933, + "grad_norm": 1.0757036209106445, + "learning_rate": 4.451102262161814e-05, + "loss": 1.0125, + "step": 4272 + }, + { + "epoch": 0.29359626219595986, + "grad_norm": 1.0517531633377075, + "learning_rate": 4.450727298314137e-05, + "loss": 1.0076, + "step": 4273 + }, + { + "epoch": 0.29366497182905044, + "grad_norm": 1.1263177394866943, + "learning_rate": 4.450352222242308e-05, + "loss": 1.0062, + "step": 4274 + }, + { + "epoch": 0.293733681462141, + "grad_norm": 1.1413146257400513, + "learning_rate": 4.4499770339679045e-05, + "loss": 1.0047, + "step": 4275 + }, + { + "epoch": 0.29380239109523154, + "grad_norm": 1.1400480270385742, + "learning_rate": 4.449601733512511e-05, + "loss": 1.0123, + "step": 4276 + }, + { + "epoch": 0.2938711007283221, + "grad_norm": 1.1950517892837524, + "learning_rate": 4.449226320897719e-05, + "loss": 0.9819, + "step": 4277 + }, + { + "epoch": 0.29393981036141265, + "grad_norm": 1.1479028463363647, + "learning_rate": 4.4488507961451254e-05, + "loss": 1.0375, + "step": 4278 + }, + { + "epoch": 0.2940085199945032, + "grad_norm": 1.1412104368209839, + "learning_rate": 4.448475159276333e-05, + "loss": 1.0239, + "step": 4279 + }, + { + "epoch": 0.2940772296275938, + "grad_norm": 1.0680317878723145, + "learning_rate": 4.4480994103129534e-05, + "loss": 0.9876, + "step": 4280 + }, + { + "epoch": 0.29414593926068433, + "grad_norm": 1.1539324522018433, + "learning_rate": 4.447723549276602e-05, + "loss": 1.0302, + "step": 4281 + }, + { + "epoch": 0.2942146488937749, + "grad_norm": 1.1167304515838623, + "learning_rate": 4.447347576188903e-05, + "loss": 1.0172, + "step": 4282 + }, + { + "epoch": 0.2942833585268655, + "grad_norm": 1.1188160181045532, + "learning_rate": 4.446971491071485e-05, + "loss": 1.0269, + "step": 4283 + }, + { + "epoch": 0.294352068159956, + "grad_norm": 1.1115628480911255, + "learning_rate": 4.446595293945985e-05, + "loss": 1.0028, + "step": 4284 + }, + { + "epoch": 0.2944207777930466, + "grad_norm": 1.2011451721191406, + "learning_rate": 4.446218984834044e-05, + "loss": 0.9994, + "step": 4285 + }, + { + "epoch": 0.2944894874261371, + "grad_norm": 1.0603604316711426, + "learning_rate": 4.445842563757311e-05, + "loss": 0.9742, + "step": 4286 + }, + { + "epoch": 0.2945581970592277, + "grad_norm": 1.0863456726074219, + "learning_rate": 4.445466030737442e-05, + "loss": 1.0042, + "step": 4287 + }, + { + "epoch": 0.2946269066923183, + "grad_norm": 1.1623698472976685, + "learning_rate": 4.445089385796099e-05, + "loss": 1.0304, + "step": 4288 + }, + { + "epoch": 0.2946956163254088, + "grad_norm": 1.1246927976608276, + "learning_rate": 4.444712628954949e-05, + "loss": 0.9744, + "step": 4289 + }, + { + "epoch": 0.2947643259584994, + "grad_norm": 1.057605504989624, + "learning_rate": 4.444335760235668e-05, + "loss": 0.9815, + "step": 4290 + }, + { + "epoch": 0.29483303559158996, + "grad_norm": 1.0854514837265015, + "learning_rate": 4.443958779659936e-05, + "loss": 0.9751, + "step": 4291 + }, + { + "epoch": 0.2949017452246805, + "grad_norm": 1.1201062202453613, + "learning_rate": 4.44358168724944e-05, + "loss": 0.9658, + "step": 4292 + }, + { + "epoch": 0.29497045485777107, + "grad_norm": 1.0743790864944458, + "learning_rate": 4.4432044830258745e-05, + "loss": 0.9942, + "step": 4293 + }, + { + "epoch": 0.2950391644908616, + "grad_norm": 1.1318970918655396, + "learning_rate": 4.442827167010941e-05, + "loss": 0.9884, + "step": 4294 + }, + { + "epoch": 0.2951078741239522, + "grad_norm": 1.2502444982528687, + "learning_rate": 4.4424497392263445e-05, + "loss": 1.0482, + "step": 4295 + }, + { + "epoch": 0.29517658375704275, + "grad_norm": 1.2035428285598755, + "learning_rate": 4.4420721996937996e-05, + "loss": 0.9843, + "step": 4296 + }, + { + "epoch": 0.2952452933901333, + "grad_norm": 1.1266535520553589, + "learning_rate": 4.4416945484350246e-05, + "loss": 1.0463, + "step": 4297 + }, + { + "epoch": 0.29531400302322386, + "grad_norm": 1.1896246671676636, + "learning_rate": 4.441316785471746e-05, + "loss": 0.9859, + "step": 4298 + }, + { + "epoch": 0.29538271265631444, + "grad_norm": 1.0470001697540283, + "learning_rate": 4.440938910825696e-05, + "loss": 1.0031, + "step": 4299 + }, + { + "epoch": 0.29545142228940496, + "grad_norm": 1.1885900497436523, + "learning_rate": 4.440560924518615e-05, + "loss": 1.0139, + "step": 4300 + }, + { + "epoch": 0.29552013192249554, + "grad_norm": 1.07623291015625, + "learning_rate": 4.4401828265722464e-05, + "loss": 1.0116, + "step": 4301 + }, + { + "epoch": 0.29558884155558607, + "grad_norm": 1.1129039525985718, + "learning_rate": 4.439804617008343e-05, + "loss": 1.0163, + "step": 4302 + }, + { + "epoch": 0.29565755118867665, + "grad_norm": 1.091982364654541, + "learning_rate": 4.439426295848663e-05, + "loss": 1.0036, + "step": 4303 + }, + { + "epoch": 0.29572626082176723, + "grad_norm": 1.1326571702957153, + "learning_rate": 4.4390478631149715e-05, + "loss": 1.0453, + "step": 4304 + }, + { + "epoch": 0.29579497045485775, + "grad_norm": 1.1498955488204956, + "learning_rate": 4.4386693188290376e-05, + "loss": 0.9958, + "step": 4305 + }, + { + "epoch": 0.29586368008794833, + "grad_norm": 1.067696452140808, + "learning_rate": 4.43829066301264e-05, + "loss": 0.9797, + "step": 4306 + }, + { + "epoch": 0.2959323897210389, + "grad_norm": 1.0764408111572266, + "learning_rate": 4.4379118956875623e-05, + "loss": 0.9628, + "step": 4307 + }, + { + "epoch": 0.29600109935412944, + "grad_norm": 1.0922244787216187, + "learning_rate": 4.437533016875594e-05, + "loss": 0.9805, + "step": 4308 + }, + { + "epoch": 0.29606980898722, + "grad_norm": 1.233925223350525, + "learning_rate": 4.4371540265985343e-05, + "loss": 0.9985, + "step": 4309 + }, + { + "epoch": 0.29613851862031054, + "grad_norm": 1.0937960147857666, + "learning_rate": 4.436774924878184e-05, + "loss": 0.9678, + "step": 4310 + }, + { + "epoch": 0.2962072282534011, + "grad_norm": 1.0517139434814453, + "learning_rate": 4.436395711736353e-05, + "loss": 1.0419, + "step": 4311 + }, + { + "epoch": 0.2962759378864917, + "grad_norm": 1.1244386434555054, + "learning_rate": 4.436016387194858e-05, + "loss": 1.0441, + "step": 4312 + }, + { + "epoch": 0.2963446475195822, + "grad_norm": 1.1179715394973755, + "learning_rate": 4.4356369512755206e-05, + "loss": 1.0173, + "step": 4313 + }, + { + "epoch": 0.2964133571526728, + "grad_norm": 1.031566858291626, + "learning_rate": 4.4352574040001695e-05, + "loss": 0.9879, + "step": 4314 + }, + { + "epoch": 0.2964820667857634, + "grad_norm": 1.064401388168335, + "learning_rate": 4.434877745390639e-05, + "loss": 1.0085, + "step": 4315 + }, + { + "epoch": 0.2965507764188539, + "grad_norm": 1.0711297988891602, + "learning_rate": 4.434497975468772e-05, + "loss": 0.9857, + "step": 4316 + }, + { + "epoch": 0.2966194860519445, + "grad_norm": 1.072522521018982, + "learning_rate": 4.434118094256416e-05, + "loss": 0.9678, + "step": 4317 + }, + { + "epoch": 0.296688195685035, + "grad_norm": 1.1045984029769897, + "learning_rate": 4.4337381017754266e-05, + "loss": 0.9966, + "step": 4318 + }, + { + "epoch": 0.2967569053181256, + "grad_norm": 1.174133539199829, + "learning_rate": 4.433357998047662e-05, + "loss": 1.002, + "step": 4319 + }, + { + "epoch": 0.2968256149512162, + "grad_norm": 1.033514142036438, + "learning_rate": 4.4329777830949907e-05, + "loss": 0.9644, + "step": 4320 + }, + { + "epoch": 0.2968943245843067, + "grad_norm": 1.0926586389541626, + "learning_rate": 4.4325974569392865e-05, + "loss": 1.0085, + "step": 4321 + }, + { + "epoch": 0.2969630342173973, + "grad_norm": 1.0854896306991577, + "learning_rate": 4.432217019602429e-05, + "loss": 1.0011, + "step": 4322 + }, + { + "epoch": 0.29703174385048786, + "grad_norm": 1.1199150085449219, + "learning_rate": 4.4318364711063045e-05, + "loss": 1.0117, + "step": 4323 + }, + { + "epoch": 0.2971004534835784, + "grad_norm": 1.1099777221679688, + "learning_rate": 4.431455811472805e-05, + "loss": 0.9809, + "step": 4324 + }, + { + "epoch": 0.29716916311666897, + "grad_norm": 1.1231279373168945, + "learning_rate": 4.431075040723831e-05, + "loss": 1.0213, + "step": 4325 + }, + { + "epoch": 0.2972378727497595, + "grad_norm": 1.2173620462417603, + "learning_rate": 4.4306941588812866e-05, + "loss": 1.0421, + "step": 4326 + }, + { + "epoch": 0.29730658238285007, + "grad_norm": 1.1030960083007812, + "learning_rate": 4.4303131659670856e-05, + "loss": 0.9978, + "step": 4327 + }, + { + "epoch": 0.29737529201594065, + "grad_norm": 1.071695327758789, + "learning_rate": 4.4299320620031444e-05, + "loss": 1.007, + "step": 4328 + }, + { + "epoch": 0.2974440016490312, + "grad_norm": 1.0695005655288696, + "learning_rate": 4.429550847011389e-05, + "loss": 1.0094, + "step": 4329 + }, + { + "epoch": 0.29751271128212176, + "grad_norm": 1.2727233171463013, + "learning_rate": 4.429169521013749e-05, + "loss": 1.0511, + "step": 4330 + }, + { + "epoch": 0.29758142091521234, + "grad_norm": 1.0809416770935059, + "learning_rate": 4.428788084032163e-05, + "loss": 0.9988, + "step": 4331 + }, + { + "epoch": 0.29765013054830286, + "grad_norm": 1.1121559143066406, + "learning_rate": 4.428406536088575e-05, + "loss": 1.0338, + "step": 4332 + }, + { + "epoch": 0.29771884018139344, + "grad_norm": 1.0767353773117065, + "learning_rate": 4.428024877204935e-05, + "loss": 0.9703, + "step": 4333 + }, + { + "epoch": 0.29778754981448396, + "grad_norm": 1.1393219232559204, + "learning_rate": 4.427643107403199e-05, + "loss": 1.0288, + "step": 4334 + }, + { + "epoch": 0.29785625944757455, + "grad_norm": 1.1601051092147827, + "learning_rate": 4.427261226705329e-05, + "loss": 0.9765, + "step": 4335 + }, + { + "epoch": 0.2979249690806651, + "grad_norm": 1.0704066753387451, + "learning_rate": 4.426879235133298e-05, + "loss": 1.0288, + "step": 4336 + }, + { + "epoch": 0.29799367871375565, + "grad_norm": 1.09932541847229, + "learning_rate": 4.426497132709079e-05, + "loss": 1.0153, + "step": 4337 + }, + { + "epoch": 0.29806238834684623, + "grad_norm": 1.0890167951583862, + "learning_rate": 4.426114919454654e-05, + "loss": 1.0175, + "step": 4338 + }, + { + "epoch": 0.2981310979799368, + "grad_norm": 1.1625932455062866, + "learning_rate": 4.425732595392011e-05, + "loss": 1.0055, + "step": 4339 + }, + { + "epoch": 0.29819980761302733, + "grad_norm": 1.1051733493804932, + "learning_rate": 4.425350160543148e-05, + "loss": 1.0001, + "step": 4340 + }, + { + "epoch": 0.2982685172461179, + "grad_norm": 1.1292754411697388, + "learning_rate": 4.4249676149300635e-05, + "loss": 0.985, + "step": 4341 + }, + { + "epoch": 0.29833722687920844, + "grad_norm": 1.081939935684204, + "learning_rate": 4.4245849585747654e-05, + "loss": 0.9726, + "step": 4342 + }, + { + "epoch": 0.298405936512299, + "grad_norm": 1.0513708591461182, + "learning_rate": 4.4242021914992696e-05, + "loss": 0.9875, + "step": 4343 + }, + { + "epoch": 0.2984746461453896, + "grad_norm": 1.079567551612854, + "learning_rate": 4.4238193137255936e-05, + "loss": 0.9551, + "step": 4344 + }, + { + "epoch": 0.2985433557784801, + "grad_norm": 1.2208142280578613, + "learning_rate": 4.423436325275766e-05, + "loss": 1.0378, + "step": 4345 + }, + { + "epoch": 0.2986120654115707, + "grad_norm": 1.122745156288147, + "learning_rate": 4.423053226171818e-05, + "loss": 1.0073, + "step": 4346 + }, + { + "epoch": 0.2986807750446613, + "grad_norm": 1.0428242683410645, + "learning_rate": 4.422670016435792e-05, + "loss": 0.9813, + "step": 4347 + }, + { + "epoch": 0.2987494846777518, + "grad_norm": 1.1255532503128052, + "learning_rate": 4.4222866960897316e-05, + "loss": 0.9959, + "step": 4348 + }, + { + "epoch": 0.2988181943108424, + "grad_norm": 1.1551661491394043, + "learning_rate": 4.4219032651556905e-05, + "loss": 1.0157, + "step": 4349 + }, + { + "epoch": 0.2988869039439329, + "grad_norm": 1.0616037845611572, + "learning_rate": 4.421519723655726e-05, + "loss": 0.996, + "step": 4350 + }, + { + "epoch": 0.2989556135770235, + "grad_norm": 1.0464738607406616, + "learning_rate": 4.421136071611903e-05, + "loss": 1.0133, + "step": 4351 + }, + { + "epoch": 0.2990243232101141, + "grad_norm": 1.098384976387024, + "learning_rate": 4.420752309046293e-05, + "loss": 1.0069, + "step": 4352 + }, + { + "epoch": 0.2990930328432046, + "grad_norm": 1.1459531784057617, + "learning_rate": 4.4203684359809745e-05, + "loss": 1.0331, + "step": 4353 + }, + { + "epoch": 0.2991617424762952, + "grad_norm": 1.0429404973983765, + "learning_rate": 4.419984452438031e-05, + "loss": 0.9551, + "step": 4354 + }, + { + "epoch": 0.29923045210938576, + "grad_norm": 1.179228663444519, + "learning_rate": 4.419600358439552e-05, + "loss": 0.9657, + "step": 4355 + }, + { + "epoch": 0.2992991617424763, + "grad_norm": 1.1241989135742188, + "learning_rate": 4.4192161540076346e-05, + "loss": 0.9929, + "step": 4356 + }, + { + "epoch": 0.29936787137556686, + "grad_norm": 1.0949689149856567, + "learning_rate": 4.4188318391643824e-05, + "loss": 0.9914, + "step": 4357 + }, + { + "epoch": 0.2994365810086574, + "grad_norm": 1.125244379043579, + "learning_rate": 4.418447413931904e-05, + "loss": 1.0236, + "step": 4358 + }, + { + "epoch": 0.29950529064174797, + "grad_norm": 1.1155617237091064, + "learning_rate": 4.4180628783323166e-05, + "loss": 1.0617, + "step": 4359 + }, + { + "epoch": 0.29957400027483855, + "grad_norm": 1.1309312582015991, + "learning_rate": 4.417678232387741e-05, + "loss": 1.0127, + "step": 4360 + }, + { + "epoch": 0.2996427099079291, + "grad_norm": 1.1427360773086548, + "learning_rate": 4.417293476120306e-05, + "loss": 1.0076, + "step": 4361 + }, + { + "epoch": 0.29971141954101965, + "grad_norm": 1.1204801797866821, + "learning_rate": 4.416908609552146e-05, + "loss": 1.006, + "step": 4362 + }, + { + "epoch": 0.29978012917411023, + "grad_norm": 1.221006989479065, + "learning_rate": 4.4165236327054024e-05, + "loss": 0.9673, + "step": 4363 + }, + { + "epoch": 0.29984883880720076, + "grad_norm": 1.2193233966827393, + "learning_rate": 4.416138545602223e-05, + "loss": 1.0366, + "step": 4364 + }, + { + "epoch": 0.29991754844029134, + "grad_norm": 1.1186870336532593, + "learning_rate": 4.4157533482647616e-05, + "loss": 0.9877, + "step": 4365 + }, + { + "epoch": 0.29998625807338186, + "grad_norm": 1.119498372077942, + "learning_rate": 4.415368040715178e-05, + "loss": 0.9908, + "step": 4366 + }, + { + "epoch": 0.30005496770647244, + "grad_norm": 1.1241462230682373, + "learning_rate": 4.414982622975639e-05, + "loss": 1.0119, + "step": 4367 + }, + { + "epoch": 0.300123677339563, + "grad_norm": 1.1049811840057373, + "learning_rate": 4.4145970950683175e-05, + "loss": 0.9939, + "step": 4368 + }, + { + "epoch": 0.30019238697265355, + "grad_norm": 1.25480055809021, + "learning_rate": 4.414211457015392e-05, + "loss": 1.0044, + "step": 4369 + }, + { + "epoch": 0.3002610966057441, + "grad_norm": 1.0926555395126343, + "learning_rate": 4.413825708839049e-05, + "loss": 0.9992, + "step": 4370 + }, + { + "epoch": 0.3003298062388347, + "grad_norm": 1.144036889076233, + "learning_rate": 4.413439850561479e-05, + "loss": 0.9856, + "step": 4371 + }, + { + "epoch": 0.30039851587192523, + "grad_norm": 1.1016008853912354, + "learning_rate": 4.413053882204882e-05, + "loss": 1.0222, + "step": 4372 + }, + { + "epoch": 0.3004672255050158, + "grad_norm": 1.1184933185577393, + "learning_rate": 4.4126678037914616e-05, + "loss": 1.0136, + "step": 4373 + }, + { + "epoch": 0.30053593513810634, + "grad_norm": 1.1029267311096191, + "learning_rate": 4.4122816153434287e-05, + "loss": 0.9791, + "step": 4374 + }, + { + "epoch": 0.3006046447711969, + "grad_norm": 1.1668769121170044, + "learning_rate": 4.4118953168830004e-05, + "loss": 0.9906, + "step": 4375 + }, + { + "epoch": 0.3006733544042875, + "grad_norm": 1.0773462057113647, + "learning_rate": 4.4115089084324e-05, + "loss": 0.9954, + "step": 4376 + }, + { + "epoch": 0.300742064037378, + "grad_norm": 1.1030287742614746, + "learning_rate": 4.4111223900138574e-05, + "loss": 1.0062, + "step": 4377 + }, + { + "epoch": 0.3008107736704686, + "grad_norm": 1.0970124006271362, + "learning_rate": 4.410735761649609e-05, + "loss": 0.9627, + "step": 4378 + }, + { + "epoch": 0.3008794833035592, + "grad_norm": 1.1080117225646973, + "learning_rate": 4.410349023361898e-05, + "loss": 1.0038, + "step": 4379 + }, + { + "epoch": 0.3009481929366497, + "grad_norm": 1.0679881572723389, + "learning_rate": 4.409962175172971e-05, + "loss": 1.0074, + "step": 4380 + }, + { + "epoch": 0.3010169025697403, + "grad_norm": 1.0405353307724, + "learning_rate": 4.4095752171050864e-05, + "loss": 0.9901, + "step": 4381 + }, + { + "epoch": 0.3010856122028308, + "grad_norm": 1.1691608428955078, + "learning_rate": 4.4091881491805034e-05, + "loss": 1.0043, + "step": 4382 + }, + { + "epoch": 0.3011543218359214, + "grad_norm": 1.1078599691390991, + "learning_rate": 4.4088009714214895e-05, + "loss": 1.0239, + "step": 4383 + }, + { + "epoch": 0.30122303146901197, + "grad_norm": 1.1463556289672852, + "learning_rate": 4.4084136838503195e-05, + "loss": 0.9964, + "step": 4384 + }, + { + "epoch": 0.3012917411021025, + "grad_norm": 1.0736892223358154, + "learning_rate": 4.4080262864892754e-05, + "loss": 1.0203, + "step": 4385 + }, + { + "epoch": 0.3013604507351931, + "grad_norm": 1.0984688997268677, + "learning_rate": 4.407638779360641e-05, + "loss": 0.9566, + "step": 4386 + }, + { + "epoch": 0.30142916036828366, + "grad_norm": 1.1075599193572998, + "learning_rate": 4.4072511624867114e-05, + "loss": 1.0406, + "step": 4387 + }, + { + "epoch": 0.3014978700013742, + "grad_norm": 1.0742888450622559, + "learning_rate": 4.4068634358897855e-05, + "loss": 0.9651, + "step": 4388 + }, + { + "epoch": 0.30156657963446476, + "grad_norm": 1.156599760055542, + "learning_rate": 4.4064755995921684e-05, + "loss": 1.0515, + "step": 4389 + }, + { + "epoch": 0.3016352892675553, + "grad_norm": 1.1158761978149414, + "learning_rate": 4.406087653616172e-05, + "loss": 0.9684, + "step": 4390 + }, + { + "epoch": 0.30170399890064586, + "grad_norm": 1.2457016706466675, + "learning_rate": 4.405699597984116e-05, + "loss": 0.9846, + "step": 4391 + }, + { + "epoch": 0.30177270853373644, + "grad_norm": 1.2011405229568481, + "learning_rate": 4.405311432718324e-05, + "loss": 1.0319, + "step": 4392 + }, + { + "epoch": 0.30184141816682697, + "grad_norm": 1.1409780979156494, + "learning_rate": 4.4049231578411264e-05, + "loss": 0.9915, + "step": 4393 + }, + { + "epoch": 0.30191012779991755, + "grad_norm": 1.1269605159759521, + "learning_rate": 4.404534773374862e-05, + "loss": 1.0204, + "step": 4394 + }, + { + "epoch": 0.30197883743300813, + "grad_norm": 1.1486742496490479, + "learning_rate": 4.404146279341872e-05, + "loss": 1.0157, + "step": 4395 + }, + { + "epoch": 0.30204754706609865, + "grad_norm": 1.1133739948272705, + "learning_rate": 4.403757675764508e-05, + "loss": 0.9976, + "step": 4396 + }, + { + "epoch": 0.30211625669918923, + "grad_norm": 1.231611728668213, + "learning_rate": 4.4033689626651266e-05, + "loss": 1.0448, + "step": 4397 + }, + { + "epoch": 0.30218496633227976, + "grad_norm": 1.1058270931243896, + "learning_rate": 4.402980140066089e-05, + "loss": 0.9908, + "step": 4398 + }, + { + "epoch": 0.30225367596537034, + "grad_norm": 1.2407169342041016, + "learning_rate": 4.402591207989764e-05, + "loss": 1.0371, + "step": 4399 + }, + { + "epoch": 0.3023223855984609, + "grad_norm": 1.0504019260406494, + "learning_rate": 4.4022021664585266e-05, + "loss": 1.0243, + "step": 4400 + }, + { + "epoch": 0.30239109523155144, + "grad_norm": 1.0961852073669434, + "learning_rate": 4.401813015494759e-05, + "loss": 1.0392, + "step": 4401 + }, + { + "epoch": 0.302459804864642, + "grad_norm": 1.800099492073059, + "learning_rate": 4.401423755120848e-05, + "loss": 1.0087, + "step": 4402 + }, + { + "epoch": 0.3025285144977326, + "grad_norm": 1.0932551622390747, + "learning_rate": 4.4010343853591874e-05, + "loss": 1.0086, + "step": 4403 + }, + { + "epoch": 0.30259722413082313, + "grad_norm": 1.1678630113601685, + "learning_rate": 4.400644906232178e-05, + "loss": 0.9915, + "step": 4404 + }, + { + "epoch": 0.3026659337639137, + "grad_norm": 1.1026129722595215, + "learning_rate": 4.4002553177622256e-05, + "loss": 1.0463, + "step": 4405 + }, + { + "epoch": 0.30273464339700423, + "grad_norm": 1.0983833074569702, + "learning_rate": 4.3998656199717435e-05, + "loss": 0.9857, + "step": 4406 + }, + { + "epoch": 0.3028033530300948, + "grad_norm": 1.1003013849258423, + "learning_rate": 4.3994758128831515e-05, + "loss": 0.997, + "step": 4407 + }, + { + "epoch": 0.3028720626631854, + "grad_norm": 1.1052610874176025, + "learning_rate": 4.399085896518874e-05, + "loss": 0.964, + "step": 4408 + }, + { + "epoch": 0.3029407722962759, + "grad_norm": 1.0908063650131226, + "learning_rate": 4.3986958709013424e-05, + "loss": 0.9917, + "step": 4409 + }, + { + "epoch": 0.3030094819293665, + "grad_norm": 1.1131672859191895, + "learning_rate": 4.398305736052994e-05, + "loss": 0.9739, + "step": 4410 + }, + { + "epoch": 0.3030781915624571, + "grad_norm": 1.0878393650054932, + "learning_rate": 4.397915491996276e-05, + "loss": 0.9958, + "step": 4411 + }, + { + "epoch": 0.3031469011955476, + "grad_norm": 1.061259388923645, + "learning_rate": 4.397525138753636e-05, + "loss": 1.006, + "step": 4412 + }, + { + "epoch": 0.3032156108286382, + "grad_norm": 1.0571848154067993, + "learning_rate": 4.397134676347532e-05, + "loss": 0.9779, + "step": 4413 + }, + { + "epoch": 0.3032843204617287, + "grad_norm": 1.108086109161377, + "learning_rate": 4.3967441048004265e-05, + "loss": 0.9718, + "step": 4414 + }, + { + "epoch": 0.3033530300948193, + "grad_norm": 1.1695809364318848, + "learning_rate": 4.396353424134789e-05, + "loss": 0.9856, + "step": 4415 + }, + { + "epoch": 0.30342173972790987, + "grad_norm": 1.0841913223266602, + "learning_rate": 4.395962634373097e-05, + "loss": 0.9719, + "step": 4416 + }, + { + "epoch": 0.3034904493610004, + "grad_norm": 1.0961748361587524, + "learning_rate": 4.395571735537829e-05, + "loss": 0.9983, + "step": 4417 + }, + { + "epoch": 0.30355915899409097, + "grad_norm": 1.1671165227890015, + "learning_rate": 4.3951807276514747e-05, + "loss": 0.9779, + "step": 4418 + }, + { + "epoch": 0.30362786862718155, + "grad_norm": 1.0999315977096558, + "learning_rate": 4.394789610736529e-05, + "loss": 1.0101, + "step": 4419 + }, + { + "epoch": 0.3036965782602721, + "grad_norm": 1.1093662977218628, + "learning_rate": 4.3943983848154925e-05, + "loss": 0.9918, + "step": 4420 + }, + { + "epoch": 0.30376528789336266, + "grad_norm": 1.1657485961914062, + "learning_rate": 4.3940070499108724e-05, + "loss": 1.0101, + "step": 4421 + }, + { + "epoch": 0.3038339975264532, + "grad_norm": 1.0687209367752075, + "learning_rate": 4.393615606045181e-05, + "loss": 1.0522, + "step": 4422 + }, + { + "epoch": 0.30390270715954376, + "grad_norm": 1.1294934749603271, + "learning_rate": 4.3932240532409385e-05, + "loss": 1.0025, + "step": 4423 + }, + { + "epoch": 0.30397141679263434, + "grad_norm": 1.125839352607727, + "learning_rate": 4.3928323915206706e-05, + "loss": 0.9817, + "step": 4424 + }, + { + "epoch": 0.30404012642572487, + "grad_norm": 1.0904982089996338, + "learning_rate": 4.392440620906909e-05, + "loss": 0.9908, + "step": 4425 + }, + { + "epoch": 0.30410883605881545, + "grad_norm": 1.091186285018921, + "learning_rate": 4.392048741422193e-05, + "loss": 1.0471, + "step": 4426 + }, + { + "epoch": 0.304177545691906, + "grad_norm": 1.1685359477996826, + "learning_rate": 4.391656753089066e-05, + "loss": 0.9784, + "step": 4427 + }, + { + "epoch": 0.30424625532499655, + "grad_norm": 1.0950195789337158, + "learning_rate": 4.39126465593008e-05, + "loss": 0.9951, + "step": 4428 + }, + { + "epoch": 0.30431496495808713, + "grad_norm": 1.1247971057891846, + "learning_rate": 4.390872449967791e-05, + "loss": 1.0018, + "step": 4429 + }, + { + "epoch": 0.3043836745911777, + "grad_norm": 1.1625536680221558, + "learning_rate": 4.3904801352247626e-05, + "loss": 0.9994, + "step": 4430 + }, + { + "epoch": 0.30445238422426824, + "grad_norm": 1.0624538660049438, + "learning_rate": 4.390087711723566e-05, + "loss": 0.9831, + "step": 4431 + }, + { + "epoch": 0.3045210938573588, + "grad_norm": 1.2048816680908203, + "learning_rate": 4.389695179486775e-05, + "loss": 1.0136, + "step": 4432 + }, + { + "epoch": 0.30458980349044934, + "grad_norm": 1.152484655380249, + "learning_rate": 4.389302538536972e-05, + "loss": 1.0249, + "step": 4433 + }, + { + "epoch": 0.3046585131235399, + "grad_norm": 1.076192855834961, + "learning_rate": 4.388909788896746e-05, + "loss": 0.9791, + "step": 4434 + }, + { + "epoch": 0.3047272227566305, + "grad_norm": 1.0937080383300781, + "learning_rate": 4.388516930588692e-05, + "loss": 1.0053, + "step": 4435 + }, + { + "epoch": 0.304795932389721, + "grad_norm": 1.0718775987625122, + "learning_rate": 4.388123963635411e-05, + "loss": 0.9948, + "step": 4436 + }, + { + "epoch": 0.3048646420228116, + "grad_norm": 1.1107969284057617, + "learning_rate": 4.387730888059509e-05, + "loss": 1.0094, + "step": 4437 + }, + { + "epoch": 0.3049333516559022, + "grad_norm": 1.0509315729141235, + "learning_rate": 4.3873377038836e-05, + "loss": 0.9924, + "step": 4438 + }, + { + "epoch": 0.3050020612889927, + "grad_norm": 1.0770039558410645, + "learning_rate": 4.386944411130304e-05, + "loss": 1.0163, + "step": 4439 + }, + { + "epoch": 0.3050707709220833, + "grad_norm": 1.182923674583435, + "learning_rate": 4.3865510098222465e-05, + "loss": 0.9944, + "step": 4440 + }, + { + "epoch": 0.3051394805551738, + "grad_norm": 1.0021651983261108, + "learning_rate": 4.386157499982059e-05, + "loss": 0.9839, + "step": 4441 + }, + { + "epoch": 0.3052081901882644, + "grad_norm": 1.0752249956130981, + "learning_rate": 4.385763881632381e-05, + "loss": 1.0085, + "step": 4442 + }, + { + "epoch": 0.305276899821355, + "grad_norm": 1.123220443725586, + "learning_rate": 4.385370154795857e-05, + "loss": 1.0099, + "step": 4443 + }, + { + "epoch": 0.3053456094544455, + "grad_norm": 1.0554437637329102, + "learning_rate": 4.3849763194951376e-05, + "loss": 1.0178, + "step": 4444 + }, + { + "epoch": 0.3054143190875361, + "grad_norm": 1.0826609134674072, + "learning_rate": 4.3845823757528794e-05, + "loss": 1.0308, + "step": 4445 + }, + { + "epoch": 0.30548302872062666, + "grad_norm": 1.125235676765442, + "learning_rate": 4.3841883235917464e-05, + "loss": 0.995, + "step": 4446 + }, + { + "epoch": 0.3055517383537172, + "grad_norm": 1.0874426364898682, + "learning_rate": 4.383794163034408e-05, + "loss": 1.0168, + "step": 4447 + }, + { + "epoch": 0.30562044798680776, + "grad_norm": 1.105880618095398, + "learning_rate": 4.3833998941035405e-05, + "loss": 1.0369, + "step": 4448 + }, + { + "epoch": 0.3056891576198983, + "grad_norm": 1.0896271467208862, + "learning_rate": 4.3830055168218246e-05, + "loss": 0.9692, + "step": 4449 + }, + { + "epoch": 0.30575786725298887, + "grad_norm": 1.1271538734436035, + "learning_rate": 4.38261103121195e-05, + "loss": 0.9936, + "step": 4450 + }, + { + "epoch": 0.30582657688607945, + "grad_norm": 1.0362807512283325, + "learning_rate": 4.38221643729661e-05, + "loss": 0.9578, + "step": 4451 + }, + { + "epoch": 0.30589528651917, + "grad_norm": 1.0473884344100952, + "learning_rate": 4.381821735098507e-05, + "loss": 0.997, + "step": 4452 + }, + { + "epoch": 0.30596399615226055, + "grad_norm": 1.1718385219573975, + "learning_rate": 4.381426924640346e-05, + "loss": 1.0112, + "step": 4453 + }, + { + "epoch": 0.30603270578535113, + "grad_norm": 1.148314118385315, + "learning_rate": 4.381032005944842e-05, + "loss": 1.0227, + "step": 4454 + }, + { + "epoch": 0.30610141541844166, + "grad_norm": 1.0951579809188843, + "learning_rate": 4.3806369790347134e-05, + "loss": 0.9863, + "step": 4455 + }, + { + "epoch": 0.30617012505153224, + "grad_norm": 1.1331350803375244, + "learning_rate": 4.380241843932687e-05, + "loss": 1.037, + "step": 4456 + }, + { + "epoch": 0.30623883468462276, + "grad_norm": 1.0917091369628906, + "learning_rate": 4.3798466006614925e-05, + "loss": 1.0055, + "step": 4457 + }, + { + "epoch": 0.30630754431771334, + "grad_norm": 1.1063848733901978, + "learning_rate": 4.3794512492438686e-05, + "loss": 1.0002, + "step": 4458 + }, + { + "epoch": 0.3063762539508039, + "grad_norm": 1.1537724733352661, + "learning_rate": 4.379055789702562e-05, + "loss": 1.0288, + "step": 4459 + }, + { + "epoch": 0.30644496358389445, + "grad_norm": 1.0382872819900513, + "learning_rate": 4.378660222060321e-05, + "loss": 1.0179, + "step": 4460 + }, + { + "epoch": 0.306513673216985, + "grad_norm": 1.0692129135131836, + "learning_rate": 4.3782645463399026e-05, + "loss": 1.0109, + "step": 4461 + }, + { + "epoch": 0.3065823828500756, + "grad_norm": 1.1581696271896362, + "learning_rate": 4.37786876256407e-05, + "loss": 1.0214, + "step": 4462 + }, + { + "epoch": 0.30665109248316613, + "grad_norm": 1.0501604080200195, + "learning_rate": 4.3774728707555924e-05, + "loss": 0.958, + "step": 4463 + }, + { + "epoch": 0.3067198021162567, + "grad_norm": 1.1486424207687378, + "learning_rate": 4.377076870937246e-05, + "loss": 0.984, + "step": 4464 + }, + { + "epoch": 0.30678851174934724, + "grad_norm": 1.1525593996047974, + "learning_rate": 4.3766807631318106e-05, + "loss": 1.0173, + "step": 4465 + }, + { + "epoch": 0.3068572213824378, + "grad_norm": 1.1736600399017334, + "learning_rate": 4.376284547362076e-05, + "loss": 1.0094, + "step": 4466 + }, + { + "epoch": 0.3069259310155284, + "grad_norm": 1.1405155658721924, + "learning_rate": 4.375888223650836e-05, + "loss": 0.9719, + "step": 4467 + }, + { + "epoch": 0.3069946406486189, + "grad_norm": 1.1094951629638672, + "learning_rate": 4.375491792020889e-05, + "loss": 0.9981, + "step": 4468 + }, + { + "epoch": 0.3070633502817095, + "grad_norm": 1.1382876634597778, + "learning_rate": 4.375095252495044e-05, + "loss": 0.9885, + "step": 4469 + }, + { + "epoch": 0.3071320599148001, + "grad_norm": 1.1502288579940796, + "learning_rate": 4.3746986050961115e-05, + "loss": 1.0671, + "step": 4470 + }, + { + "epoch": 0.3072007695478906, + "grad_norm": 1.0942810773849487, + "learning_rate": 4.374301849846911e-05, + "loss": 0.9959, + "step": 4471 + }, + { + "epoch": 0.3072694791809812, + "grad_norm": 1.1166936159133911, + "learning_rate": 4.373904986770269e-05, + "loss": 1.041, + "step": 4472 + }, + { + "epoch": 0.3073381888140717, + "grad_norm": 1.0656864643096924, + "learning_rate": 4.373508015889015e-05, + "loss": 0.9683, + "step": 4473 + }, + { + "epoch": 0.3074068984471623, + "grad_norm": 1.1142293214797974, + "learning_rate": 4.3731109372259884e-05, + "loss": 1.0201, + "step": 4474 + }, + { + "epoch": 0.30747560808025287, + "grad_norm": 1.124493956565857, + "learning_rate": 4.37271375080403e-05, + "loss": 0.9911, + "step": 4475 + }, + { + "epoch": 0.3075443177133434, + "grad_norm": 1.1147059202194214, + "learning_rate": 4.372316456645992e-05, + "loss": 1.0066, + "step": 4476 + }, + { + "epoch": 0.307613027346434, + "grad_norm": 1.1200028657913208, + "learning_rate": 4.37191905477473e-05, + "loss": 0.9827, + "step": 4477 + }, + { + "epoch": 0.30768173697952456, + "grad_norm": 1.1323390007019043, + "learning_rate": 4.371521545213106e-05, + "loss": 1.035, + "step": 4478 + }, + { + "epoch": 0.3077504466126151, + "grad_norm": 1.0734928846359253, + "learning_rate": 4.371123927983989e-05, + "loss": 1.0176, + "step": 4479 + }, + { + "epoch": 0.30781915624570566, + "grad_norm": 1.1397267580032349, + "learning_rate": 4.3707262031102525e-05, + "loss": 1.088, + "step": 4480 + }, + { + "epoch": 0.3078878658787962, + "grad_norm": 1.135842204093933, + "learning_rate": 4.370328370614779e-05, + "loss": 1.0158, + "step": 4481 + }, + { + "epoch": 0.30795657551188677, + "grad_norm": 1.067326545715332, + "learning_rate": 4.369930430520454e-05, + "loss": 0.9444, + "step": 4482 + }, + { + "epoch": 0.30802528514497735, + "grad_norm": 1.2025519609451294, + "learning_rate": 4.369532382850172e-05, + "loss": 1.0588, + "step": 4483 + }, + { + "epoch": 0.30809399477806787, + "grad_norm": 1.1548088788986206, + "learning_rate": 4.369134227626832e-05, + "loss": 1.0089, + "step": 4484 + }, + { + "epoch": 0.30816270441115845, + "grad_norm": 1.116045355796814, + "learning_rate": 4.368735964873339e-05, + "loss": 0.9968, + "step": 4485 + }, + { + "epoch": 0.30823141404424903, + "grad_norm": 1.1063706874847412, + "learning_rate": 4.368337594612606e-05, + "loss": 1.0059, + "step": 4486 + }, + { + "epoch": 0.30830012367733955, + "grad_norm": 1.0684515237808228, + "learning_rate": 4.36793911686755e-05, + "loss": 1.0173, + "step": 4487 + }, + { + "epoch": 0.30836883331043013, + "grad_norm": 1.1462935209274292, + "learning_rate": 4.367540531661096e-05, + "loss": 1.0084, + "step": 4488 + }, + { + "epoch": 0.30843754294352066, + "grad_norm": 1.1567169427871704, + "learning_rate": 4.367141839016173e-05, + "loss": 1.0504, + "step": 4489 + }, + { + "epoch": 0.30850625257661124, + "grad_norm": 1.1520153284072876, + "learning_rate": 4.36674303895572e-05, + "loss": 1.0618, + "step": 4490 + }, + { + "epoch": 0.3085749622097018, + "grad_norm": 1.0832356214523315, + "learning_rate": 4.3663441315026764e-05, + "loss": 1.0117, + "step": 4491 + }, + { + "epoch": 0.30864367184279234, + "grad_norm": 1.1439915895462036, + "learning_rate": 4.365945116679994e-05, + "loss": 0.9929, + "step": 4492 + }, + { + "epoch": 0.3087123814758829, + "grad_norm": 1.1171730756759644, + "learning_rate": 4.365545994510627e-05, + "loss": 1.0071, + "step": 4493 + }, + { + "epoch": 0.3087810911089735, + "grad_norm": 1.1279429197311401, + "learning_rate": 4.365146765017535e-05, + "loss": 0.9417, + "step": 4494 + }, + { + "epoch": 0.30884980074206403, + "grad_norm": 1.0805449485778809, + "learning_rate": 4.364747428223688e-05, + "loss": 0.9882, + "step": 4495 + }, + { + "epoch": 0.3089185103751546, + "grad_norm": 1.0944775342941284, + "learning_rate": 4.3643479841520585e-05, + "loss": 0.9686, + "step": 4496 + }, + { + "epoch": 0.30898722000824513, + "grad_norm": 1.118577480316162, + "learning_rate": 4.3639484328256266e-05, + "loss": 1.0132, + "step": 4497 + }, + { + "epoch": 0.3090559296413357, + "grad_norm": 1.1718355417251587, + "learning_rate": 4.363548774267376e-05, + "loss": 1.0011, + "step": 4498 + }, + { + "epoch": 0.3091246392744263, + "grad_norm": 1.1704692840576172, + "learning_rate": 4.3631490085003026e-05, + "loss": 0.9969, + "step": 4499 + }, + { + "epoch": 0.3091933489075168, + "grad_norm": 1.037871241569519, + "learning_rate": 4.3627491355474016e-05, + "loss": 0.9975, + "step": 4500 + }, + { + "epoch": 0.3092620585406074, + "grad_norm": 1.0956875085830688, + "learning_rate": 4.36234915543168e-05, + "loss": 0.9774, + "step": 4501 + }, + { + "epoch": 0.309330768173698, + "grad_norm": 1.060884714126587, + "learning_rate": 4.3619490681761455e-05, + "loss": 0.9977, + "step": 4502 + }, + { + "epoch": 0.3093994778067885, + "grad_norm": 1.0414823293685913, + "learning_rate": 4.361548873803817e-05, + "loss": 0.9876, + "step": 4503 + }, + { + "epoch": 0.3094681874398791, + "grad_norm": 1.0760776996612549, + "learning_rate": 4.361148572337717e-05, + "loss": 0.9519, + "step": 4504 + }, + { + "epoch": 0.3095368970729696, + "grad_norm": 1.0708603858947754, + "learning_rate": 4.360748163800874e-05, + "loss": 1.0127, + "step": 4505 + }, + { + "epoch": 0.3096056067060602, + "grad_norm": 1.1513644456863403, + "learning_rate": 4.3603476482163244e-05, + "loss": 0.9823, + "step": 4506 + }, + { + "epoch": 0.30967431633915077, + "grad_norm": 1.116601824760437, + "learning_rate": 4.359947025607108e-05, + "loss": 1.0194, + "step": 4507 + }, + { + "epoch": 0.3097430259722413, + "grad_norm": 1.1892155408859253, + "learning_rate": 4.359546295996274e-05, + "loss": 1.0294, + "step": 4508 + }, + { + "epoch": 0.3098117356053319, + "grad_norm": 1.2123680114746094, + "learning_rate": 4.3591454594068745e-05, + "loss": 1.0451, + "step": 4509 + }, + { + "epoch": 0.30988044523842245, + "grad_norm": 1.0434831380844116, + "learning_rate": 4.358744515861972e-05, + "loss": 0.9646, + "step": 4510 + }, + { + "epoch": 0.309949154871513, + "grad_norm": 1.0364089012145996, + "learning_rate": 4.3583434653846294e-05, + "loss": 0.9402, + "step": 4511 + }, + { + "epoch": 0.31001786450460356, + "grad_norm": 1.0677298307418823, + "learning_rate": 4.357942307997921e-05, + "loss": 1.0019, + "step": 4512 + }, + { + "epoch": 0.3100865741376941, + "grad_norm": 1.1298764944076538, + "learning_rate": 4.357541043724924e-05, + "loss": 0.96, + "step": 4513 + }, + { + "epoch": 0.31015528377078466, + "grad_norm": 1.1541541814804077, + "learning_rate": 4.3571396725887236e-05, + "loss": 1.0159, + "step": 4514 + }, + { + "epoch": 0.31022399340387524, + "grad_norm": 1.0362440347671509, + "learning_rate": 4.35673819461241e-05, + "loss": 0.9776, + "step": 4515 + }, + { + "epoch": 0.31029270303696577, + "grad_norm": 1.116459608078003, + "learning_rate": 4.356336609819081e-05, + "loss": 1.0394, + "step": 4516 + }, + { + "epoch": 0.31036141267005635, + "grad_norm": 1.1794750690460205, + "learning_rate": 4.3559349182318376e-05, + "loss": 1.039, + "step": 4517 + }, + { + "epoch": 0.3104301223031469, + "grad_norm": 1.0509165525436401, + "learning_rate": 4.35553311987379e-05, + "loss": 1.0029, + "step": 4518 + }, + { + "epoch": 0.31049883193623745, + "grad_norm": 1.0034308433532715, + "learning_rate": 4.3551312147680544e-05, + "loss": 0.9786, + "step": 4519 + }, + { + "epoch": 0.31056754156932803, + "grad_norm": 1.1466026306152344, + "learning_rate": 4.354729202937751e-05, + "loss": 1.0006, + "step": 4520 + }, + { + "epoch": 0.31063625120241856, + "grad_norm": 1.0794247388839722, + "learning_rate": 4.354327084406007e-05, + "loss": 0.9967, + "step": 4521 + }, + { + "epoch": 0.31070496083550914, + "grad_norm": 1.1342120170593262, + "learning_rate": 4.3539248591959556e-05, + "loss": 0.956, + "step": 4522 + }, + { + "epoch": 0.3107736704685997, + "grad_norm": 1.0310946702957153, + "learning_rate": 4.353522527330739e-05, + "loss": 0.9834, + "step": 4523 + }, + { + "epoch": 0.31084238010169024, + "grad_norm": 1.1012287139892578, + "learning_rate": 4.353120088833501e-05, + "loss": 1.0001, + "step": 4524 + }, + { + "epoch": 0.3109110897347808, + "grad_norm": 1.0634123086929321, + "learning_rate": 4.352717543727394e-05, + "loss": 0.9783, + "step": 4525 + }, + { + "epoch": 0.3109797993678714, + "grad_norm": 1.0877821445465088, + "learning_rate": 4.352314892035576e-05, + "loss": 0.9981, + "step": 4526 + }, + { + "epoch": 0.3110485090009619, + "grad_norm": 1.2160449028015137, + "learning_rate": 4.351912133781213e-05, + "loss": 1.0073, + "step": 4527 + }, + { + "epoch": 0.3111172186340525, + "grad_norm": 1.055301308631897, + "learning_rate": 4.351509268987472e-05, + "loss": 1.0014, + "step": 4528 + }, + { + "epoch": 0.31118592826714303, + "grad_norm": 1.1644766330718994, + "learning_rate": 4.351106297677534e-05, + "loss": 0.9779, + "step": 4529 + }, + { + "epoch": 0.3112546379002336, + "grad_norm": 1.1613224744796753, + "learning_rate": 4.350703219874578e-05, + "loss": 0.9936, + "step": 4530 + }, + { + "epoch": 0.3113233475333242, + "grad_norm": 1.0253714323043823, + "learning_rate": 4.350300035601794e-05, + "loss": 0.9811, + "step": 4531 + }, + { + "epoch": 0.3113920571664147, + "grad_norm": 1.144373893737793, + "learning_rate": 4.349896744882378e-05, + "loss": 1.0118, + "step": 4532 + }, + { + "epoch": 0.3114607667995053, + "grad_norm": 1.1052558422088623, + "learning_rate": 4.3494933477395304e-05, + "loss": 1.038, + "step": 4533 + }, + { + "epoch": 0.3115294764325959, + "grad_norm": 1.0618072748184204, + "learning_rate": 4.3490898441964575e-05, + "loss": 0.9969, + "step": 4534 + }, + { + "epoch": 0.3115981860656864, + "grad_norm": 1.1767808198928833, + "learning_rate": 4.3486862342763735e-05, + "loss": 1.0052, + "step": 4535 + }, + { + "epoch": 0.311666895698777, + "grad_norm": 1.0949881076812744, + "learning_rate": 4.348282518002498e-05, + "loss": 1.0536, + "step": 4536 + }, + { + "epoch": 0.3117356053318675, + "grad_norm": 1.1210392713546753, + "learning_rate": 4.347878695398056e-05, + "loss": 1.0161, + "step": 4537 + }, + { + "epoch": 0.3118043149649581, + "grad_norm": 1.1291422843933105, + "learning_rate": 4.347474766486279e-05, + "loss": 0.9682, + "step": 4538 + }, + { + "epoch": 0.31187302459804866, + "grad_norm": 1.111934781074524, + "learning_rate": 4.347070731290405e-05, + "loss": 1.0176, + "step": 4539 + }, + { + "epoch": 0.3119417342311392, + "grad_norm": 1.1750084161758423, + "learning_rate": 4.3466665898336784e-05, + "loss": 1.049, + "step": 4540 + }, + { + "epoch": 0.31201044386422977, + "grad_norm": 1.099775791168213, + "learning_rate": 4.3462623421393486e-05, + "loss": 0.9892, + "step": 4541 + }, + { + "epoch": 0.31207915349732035, + "grad_norm": 1.1578716039657593, + "learning_rate": 4.3458579882306726e-05, + "loss": 1.0092, + "step": 4542 + }, + { + "epoch": 0.3121478631304109, + "grad_norm": 1.076208233833313, + "learning_rate": 4.345453528130912e-05, + "loss": 1.0036, + "step": 4543 + }, + { + "epoch": 0.31221657276350145, + "grad_norm": 1.1665081977844238, + "learning_rate": 4.345048961863334e-05, + "loss": 1.0498, + "step": 4544 + }, + { + "epoch": 0.312285282396592, + "grad_norm": 1.1566935777664185, + "learning_rate": 4.344644289451216e-05, + "loss": 0.966, + "step": 4545 + }, + { + "epoch": 0.31235399202968256, + "grad_norm": 1.1281824111938477, + "learning_rate": 4.3442395109178355e-05, + "loss": 1.0447, + "step": 4546 + }, + { + "epoch": 0.31242270166277314, + "grad_norm": 1.109261155128479, + "learning_rate": 4.34383462628648e-05, + "loss": 1.0157, + "step": 4547 + }, + { + "epoch": 0.31249141129586366, + "grad_norm": 1.1134412288665771, + "learning_rate": 4.343429635580444e-05, + "loss": 0.993, + "step": 4548 + }, + { + "epoch": 0.31256012092895424, + "grad_norm": 1.1417465209960938, + "learning_rate": 4.3430245388230246e-05, + "loss": 0.9835, + "step": 4549 + }, + { + "epoch": 0.3126288305620448, + "grad_norm": 1.1219110488891602, + "learning_rate": 4.342619336037527e-05, + "loss": 0.9911, + "step": 4550 + }, + { + "epoch": 0.31269754019513535, + "grad_norm": 1.0872726440429688, + "learning_rate": 4.3422140272472625e-05, + "loss": 0.9796, + "step": 4551 + }, + { + "epoch": 0.31276624982822593, + "grad_norm": 1.1222203969955444, + "learning_rate": 4.3418086124755483e-05, + "loss": 0.9683, + "step": 4552 + }, + { + "epoch": 0.31283495946131645, + "grad_norm": 1.091464638710022, + "learning_rate": 4.341403091745708e-05, + "loss": 0.9829, + "step": 4553 + }, + { + "epoch": 0.31290366909440703, + "grad_norm": 1.0972874164581299, + "learning_rate": 4.3409974650810706e-05, + "loss": 0.9917, + "step": 4554 + }, + { + "epoch": 0.3129723787274976, + "grad_norm": 1.0739209651947021, + "learning_rate": 4.340591732504971e-05, + "loss": 0.9661, + "step": 4555 + }, + { + "epoch": 0.31304108836058814, + "grad_norm": 1.214332938194275, + "learning_rate": 4.340185894040752e-05, + "loss": 1.0247, + "step": 4556 + }, + { + "epoch": 0.3131097979936787, + "grad_norm": 1.1517688035964966, + "learning_rate": 4.3397799497117596e-05, + "loss": 1.0096, + "step": 4557 + }, + { + "epoch": 0.3131785076267693, + "grad_norm": 1.134806752204895, + "learning_rate": 4.33937389954135e-05, + "loss": 1.0237, + "step": 4558 + }, + { + "epoch": 0.3132472172598598, + "grad_norm": 1.0922162532806396, + "learning_rate": 4.3389677435528796e-05, + "loss": 0.9999, + "step": 4559 + }, + { + "epoch": 0.3133159268929504, + "grad_norm": 1.149766445159912, + "learning_rate": 4.338561481769717e-05, + "loss": 0.9967, + "step": 4560 + }, + { + "epoch": 0.3133846365260409, + "grad_norm": 1.0437816381454468, + "learning_rate": 4.338155114215234e-05, + "loss": 0.9902, + "step": 4561 + }, + { + "epoch": 0.3134533461591315, + "grad_norm": 1.0378859043121338, + "learning_rate": 4.337748640912808e-05, + "loss": 0.9816, + "step": 4562 + }, + { + "epoch": 0.3135220557922221, + "grad_norm": 1.1157746315002441, + "learning_rate": 4.337342061885822e-05, + "loss": 0.9999, + "step": 4563 + }, + { + "epoch": 0.3135907654253126, + "grad_norm": 1.102311372756958, + "learning_rate": 4.336935377157668e-05, + "loss": 1.0074, + "step": 4564 + }, + { + "epoch": 0.3136594750584032, + "grad_norm": 1.1318608522415161, + "learning_rate": 4.3365285867517425e-05, + "loss": 0.9883, + "step": 4565 + }, + { + "epoch": 0.31372818469149377, + "grad_norm": 1.0873181819915771, + "learning_rate": 4.336121690691447e-05, + "loss": 1.0661, + "step": 4566 + }, + { + "epoch": 0.3137968943245843, + "grad_norm": 1.088470697402954, + "learning_rate": 4.3357146890001896e-05, + "loss": 0.984, + "step": 4567 + }, + { + "epoch": 0.3138656039576749, + "grad_norm": 1.1094187498092651, + "learning_rate": 4.335307581701386e-05, + "loss": 1.0325, + "step": 4568 + }, + { + "epoch": 0.3139343135907654, + "grad_norm": 1.04935622215271, + "learning_rate": 4.334900368818456e-05, + "loss": 0.9812, + "step": 4569 + }, + { + "epoch": 0.314003023223856, + "grad_norm": 1.070473313331604, + "learning_rate": 4.334493050374826e-05, + "loss": 0.9976, + "step": 4570 + }, + { + "epoch": 0.31407173285694656, + "grad_norm": 1.0275218486785889, + "learning_rate": 4.33408562639393e-05, + "loss": 0.9648, + "step": 4571 + }, + { + "epoch": 0.3141404424900371, + "grad_norm": 1.0505112409591675, + "learning_rate": 4.3336780968992065e-05, + "loss": 0.9915, + "step": 4572 + }, + { + "epoch": 0.31420915212312767, + "grad_norm": 1.0940184593200684, + "learning_rate": 4.333270461914099e-05, + "loss": 0.9821, + "step": 4573 + }, + { + "epoch": 0.31427786175621825, + "grad_norm": 1.128340482711792, + "learning_rate": 4.332862721462061e-05, + "loss": 1.0056, + "step": 4574 + }, + { + "epoch": 0.31434657138930877, + "grad_norm": 1.191727638244629, + "learning_rate": 4.332454875566548e-05, + "loss": 0.9985, + "step": 4575 + }, + { + "epoch": 0.31441528102239935, + "grad_norm": 1.0631831884384155, + "learning_rate": 4.332046924251023e-05, + "loss": 1.0167, + "step": 4576 + }, + { + "epoch": 0.3144839906554899, + "grad_norm": 1.0685317516326904, + "learning_rate": 4.331638867538955e-05, + "loss": 0.9813, + "step": 4577 + }, + { + "epoch": 0.31455270028858046, + "grad_norm": 1.1501680612564087, + "learning_rate": 4.33123070545382e-05, + "loss": 0.9885, + "step": 4578 + }, + { + "epoch": 0.31462140992167104, + "grad_norm": 1.107223391532898, + "learning_rate": 4.3308224380191e-05, + "loss": 1.0158, + "step": 4579 + }, + { + "epoch": 0.31469011955476156, + "grad_norm": 1.1556978225708008, + "learning_rate": 4.3304140652582806e-05, + "loss": 0.9882, + "step": 4580 + }, + { + "epoch": 0.31475882918785214, + "grad_norm": 1.1220952272415161, + "learning_rate": 4.330005587194856e-05, + "loss": 1.0134, + "step": 4581 + }, + { + "epoch": 0.3148275388209427, + "grad_norm": 1.1255192756652832, + "learning_rate": 4.329597003852326e-05, + "loss": 1.0202, + "step": 4582 + }, + { + "epoch": 0.31489624845403325, + "grad_norm": 1.1085788011550903, + "learning_rate": 4.329188315254196e-05, + "loss": 0.9737, + "step": 4583 + }, + { + "epoch": 0.3149649580871238, + "grad_norm": 1.087589144706726, + "learning_rate": 4.328779521423978e-05, + "loss": 0.9934, + "step": 4584 + }, + { + "epoch": 0.31503366772021435, + "grad_norm": 1.1207101345062256, + "learning_rate": 4.3283706223851885e-05, + "loss": 0.9786, + "step": 4585 + }, + { + "epoch": 0.31510237735330493, + "grad_norm": 1.059441089630127, + "learning_rate": 4.327961618161352e-05, + "loss": 0.9778, + "step": 4586 + }, + { + "epoch": 0.3151710869863955, + "grad_norm": 1.1184173822402954, + "learning_rate": 4.3275525087759984e-05, + "loss": 1.0109, + "step": 4587 + }, + { + "epoch": 0.31523979661948603, + "grad_norm": 1.0898525714874268, + "learning_rate": 4.327143294252663e-05, + "loss": 0.9875, + "step": 4588 + }, + { + "epoch": 0.3153085062525766, + "grad_norm": 1.1958571672439575, + "learning_rate": 4.326733974614888e-05, + "loss": 1.012, + "step": 4589 + }, + { + "epoch": 0.3153772158856672, + "grad_norm": 1.062178611755371, + "learning_rate": 4.3263245498862215e-05, + "loss": 1.0062, + "step": 4590 + }, + { + "epoch": 0.3154459255187577, + "grad_norm": 1.160032868385315, + "learning_rate": 4.325915020090217e-05, + "loss": 1.0255, + "step": 4591 + }, + { + "epoch": 0.3155146351518483, + "grad_norm": 1.096483826637268, + "learning_rate": 4.325505385250434e-05, + "loss": 0.9894, + "step": 4592 + }, + { + "epoch": 0.3155833447849388, + "grad_norm": 1.1719257831573486, + "learning_rate": 4.3250956453904414e-05, + "loss": 0.9987, + "step": 4593 + }, + { + "epoch": 0.3156520544180294, + "grad_norm": 1.1080000400543213, + "learning_rate": 4.3246858005338076e-05, + "loss": 0.9946, + "step": 4594 + }, + { + "epoch": 0.31572076405112, + "grad_norm": 1.1530729532241821, + "learning_rate": 4.324275850704112e-05, + "loss": 0.9985, + "step": 4595 + }, + { + "epoch": 0.3157894736842105, + "grad_norm": 1.0431824922561646, + "learning_rate": 4.3238657959249396e-05, + "loss": 1.0081, + "step": 4596 + }, + { + "epoch": 0.3158581833173011, + "grad_norm": 1.116015911102295, + "learning_rate": 4.3234556362198804e-05, + "loss": 1.0062, + "step": 4597 + }, + { + "epoch": 0.31592689295039167, + "grad_norm": 1.1070990562438965, + "learning_rate": 4.3230453716125296e-05, + "loss": 0.9852, + "step": 4598 + }, + { + "epoch": 0.3159956025834822, + "grad_norm": 1.2231581211090088, + "learning_rate": 4.3226350021264904e-05, + "loss": 1.0094, + "step": 4599 + }, + { + "epoch": 0.3160643122165728, + "grad_norm": 1.1435961723327637, + "learning_rate": 4.3222245277853706e-05, + "loss": 1.0131, + "step": 4600 + }, + { + "epoch": 0.3161330218496633, + "grad_norm": 1.1149728298187256, + "learning_rate": 4.3218139486127854e-05, + "loss": 0.9954, + "step": 4601 + }, + { + "epoch": 0.3162017314827539, + "grad_norm": 1.1296374797821045, + "learning_rate": 4.321403264632354e-05, + "loss": 0.9905, + "step": 4602 + }, + { + "epoch": 0.31627044111584446, + "grad_norm": 1.112879991531372, + "learning_rate": 4.3209924758677036e-05, + "loss": 1.0133, + "step": 4603 + }, + { + "epoch": 0.316339150748935, + "grad_norm": 1.1101688146591187, + "learning_rate": 4.320581582342466e-05, + "loss": 1.0377, + "step": 4604 + }, + { + "epoch": 0.31640786038202556, + "grad_norm": 1.0840023756027222, + "learning_rate": 4.32017058408028e-05, + "loss": 1.0015, + "step": 4605 + }, + { + "epoch": 0.31647657001511614, + "grad_norm": 1.1265166997909546, + "learning_rate": 4.31975948110479e-05, + "loss": 0.9971, + "step": 4606 + }, + { + "epoch": 0.31654527964820667, + "grad_norm": 1.124611735343933, + "learning_rate": 4.319348273439648e-05, + "loss": 0.9687, + "step": 4607 + }, + { + "epoch": 0.31661398928129725, + "grad_norm": 1.2135210037231445, + "learning_rate": 4.318936961108508e-05, + "loss": 0.9954, + "step": 4608 + }, + { + "epoch": 0.3166826989143878, + "grad_norm": 1.1312733888626099, + "learning_rate": 4.318525544135034e-05, + "loss": 1.0425, + "step": 4609 + }, + { + "epoch": 0.31675140854747835, + "grad_norm": 1.1698867082595825, + "learning_rate": 4.318114022542894e-05, + "loss": 1.0266, + "step": 4610 + }, + { + "epoch": 0.31682011818056893, + "grad_norm": 1.107397437095642, + "learning_rate": 4.317702396355763e-05, + "loss": 0.9926, + "step": 4611 + }, + { + "epoch": 0.31688882781365946, + "grad_norm": 1.125187873840332, + "learning_rate": 4.317290665597322e-05, + "loss": 0.9837, + "step": 4612 + }, + { + "epoch": 0.31695753744675004, + "grad_norm": 1.095511555671692, + "learning_rate": 4.316878830291256e-05, + "loss": 0.9878, + "step": 4613 + }, + { + "epoch": 0.3170262470798406, + "grad_norm": 1.0631835460662842, + "learning_rate": 4.316466890461259e-05, + "loss": 0.9747, + "step": 4614 + }, + { + "epoch": 0.31709495671293114, + "grad_norm": 1.1718144416809082, + "learning_rate": 4.3160548461310304e-05, + "loss": 0.9745, + "step": 4615 + }, + { + "epoch": 0.3171636663460217, + "grad_norm": 1.102500081062317, + "learning_rate": 4.3156426973242723e-05, + "loss": 0.9959, + "step": 4616 + }, + { + "epoch": 0.31723237597911225, + "grad_norm": 1.2082544565200806, + "learning_rate": 4.315230444064698e-05, + "loss": 1.0232, + "step": 4617 + }, + { + "epoch": 0.3173010856122028, + "grad_norm": 1.0823177099227905, + "learning_rate": 4.314818086376022e-05, + "loss": 0.9458, + "step": 4618 + }, + { + "epoch": 0.3173697952452934, + "grad_norm": 1.1124037504196167, + "learning_rate": 4.3144056242819686e-05, + "loss": 1.0333, + "step": 4619 + }, + { + "epoch": 0.31743850487838393, + "grad_norm": 1.1479939222335815, + "learning_rate": 4.313993057806266e-05, + "loss": 1.0381, + "step": 4620 + }, + { + "epoch": 0.3175072145114745, + "grad_norm": 1.1566247940063477, + "learning_rate": 4.3135803869726494e-05, + "loss": 1.0114, + "step": 4621 + }, + { + "epoch": 0.3175759241445651, + "grad_norm": 1.092981219291687, + "learning_rate": 4.313167611804858e-05, + "loss": 0.9765, + "step": 4622 + }, + { + "epoch": 0.3176446337776556, + "grad_norm": 1.1538292169570923, + "learning_rate": 4.31275473232664e-05, + "loss": 1.03, + "step": 4623 + }, + { + "epoch": 0.3177133434107462, + "grad_norm": 1.1301254034042358, + "learning_rate": 4.3123417485617474e-05, + "loss": 0.9811, + "step": 4624 + }, + { + "epoch": 0.3177820530438367, + "grad_norm": 1.0797032117843628, + "learning_rate": 4.3119286605339394e-05, + "loss": 1.0004, + "step": 4625 + }, + { + "epoch": 0.3178507626769273, + "grad_norm": 1.15577232837677, + "learning_rate": 4.31151546826698e-05, + "loss": 0.9667, + "step": 4626 + }, + { + "epoch": 0.3179194723100179, + "grad_norm": 1.1705820560455322, + "learning_rate": 4.31110217178464e-05, + "loss": 1.0088, + "step": 4627 + }, + { + "epoch": 0.3179881819431084, + "grad_norm": 1.1647160053253174, + "learning_rate": 4.310688771110697e-05, + "loss": 0.9959, + "step": 4628 + }, + { + "epoch": 0.318056891576199, + "grad_norm": 1.1046810150146484, + "learning_rate": 4.3102752662689324e-05, + "loss": 0.9995, + "step": 4629 + }, + { + "epoch": 0.31812560120928957, + "grad_norm": 1.1230285167694092, + "learning_rate": 4.309861657283135e-05, + "loss": 0.9836, + "step": 4630 + }, + { + "epoch": 0.3181943108423801, + "grad_norm": 1.0583515167236328, + "learning_rate": 4.309447944177101e-05, + "loss": 0.9432, + "step": 4631 + }, + { + "epoch": 0.31826302047547067, + "grad_norm": 1.074642539024353, + "learning_rate": 4.3090341269746304e-05, + "loss": 0.9847, + "step": 4632 + }, + { + "epoch": 0.3183317301085612, + "grad_norm": 1.1129075288772583, + "learning_rate": 4.308620205699528e-05, + "loss": 0.9795, + "step": 4633 + }, + { + "epoch": 0.3184004397416518, + "grad_norm": 1.0940414667129517, + "learning_rate": 4.308206180375609e-05, + "loss": 1.0006, + "step": 4634 + }, + { + "epoch": 0.31846914937474236, + "grad_norm": 1.1217859983444214, + "learning_rate": 4.307792051026691e-05, + "loss": 1.0229, + "step": 4635 + }, + { + "epoch": 0.3185378590078329, + "grad_norm": 1.152350664138794, + "learning_rate": 4.3073778176765985e-05, + "loss": 1.0083, + "step": 4636 + }, + { + "epoch": 0.31860656864092346, + "grad_norm": 1.121545433998108, + "learning_rate": 4.306963480349162e-05, + "loss": 1.0423, + "step": 4637 + }, + { + "epoch": 0.31867527827401404, + "grad_norm": 1.0708571672439575, + "learning_rate": 4.3065490390682186e-05, + "loss": 0.9917, + "step": 4638 + }, + { + "epoch": 0.31874398790710456, + "grad_norm": 1.1251477003097534, + "learning_rate": 4.3061344938576095e-05, + "loss": 0.9741, + "step": 4639 + }, + { + "epoch": 0.31881269754019514, + "grad_norm": 1.0715409517288208, + "learning_rate": 4.305719844741185e-05, + "loss": 0.9664, + "step": 4640 + }, + { + "epoch": 0.31888140717328567, + "grad_norm": 1.0629026889801025, + "learning_rate": 4.305305091742799e-05, + "loss": 0.9474, + "step": 4641 + }, + { + "epoch": 0.31895011680637625, + "grad_norm": 1.0613250732421875, + "learning_rate": 4.3048902348863116e-05, + "loss": 0.953, + "step": 4642 + }, + { + "epoch": 0.31901882643946683, + "grad_norm": 1.1585242748260498, + "learning_rate": 4.304475274195589e-05, + "loss": 1.0047, + "step": 4643 + }, + { + "epoch": 0.31908753607255735, + "grad_norm": 1.1101102828979492, + "learning_rate": 4.304060209694505e-05, + "loss": 0.996, + "step": 4644 + }, + { + "epoch": 0.31915624570564793, + "grad_norm": 1.0695534944534302, + "learning_rate": 4.303645041406936e-05, + "loss": 0.9531, + "step": 4645 + }, + { + "epoch": 0.3192249553387385, + "grad_norm": 1.0946297645568848, + "learning_rate": 4.303229769356769e-05, + "loss": 1.0315, + "step": 4646 + }, + { + "epoch": 0.31929366497182904, + "grad_norm": 1.1143556833267212, + "learning_rate": 4.3028143935678916e-05, + "loss": 1.0345, + "step": 4647 + }, + { + "epoch": 0.3193623746049196, + "grad_norm": 1.1350165605545044, + "learning_rate": 4.302398914064202e-05, + "loss": 1.0103, + "step": 4648 + }, + { + "epoch": 0.31943108423801014, + "grad_norm": 1.0903728008270264, + "learning_rate": 4.3019833308696016e-05, + "loss": 0.9851, + "step": 4649 + }, + { + "epoch": 0.3194997938711007, + "grad_norm": 1.122125506401062, + "learning_rate": 4.301567644007999e-05, + "loss": 0.9892, + "step": 4650 + }, + { + "epoch": 0.3195685035041913, + "grad_norm": 1.1665635108947754, + "learning_rate": 4.3011518535033087e-05, + "loss": 1.0101, + "step": 4651 + }, + { + "epoch": 0.31963721313728183, + "grad_norm": 1.10844886302948, + "learning_rate": 4.3007359593794496e-05, + "loss": 0.9885, + "step": 4652 + }, + { + "epoch": 0.3197059227703724, + "grad_norm": 1.1498616933822632, + "learning_rate": 4.3003199616603495e-05, + "loss": 1.0441, + "step": 4653 + }, + { + "epoch": 0.319774632403463, + "grad_norm": 1.0542629957199097, + "learning_rate": 4.299903860369939e-05, + "loss": 0.9841, + "step": 4654 + }, + { + "epoch": 0.3198433420365535, + "grad_norm": 1.1470531225204468, + "learning_rate": 4.299487655532158e-05, + "loss": 0.9939, + "step": 4655 + }, + { + "epoch": 0.3199120516696441, + "grad_norm": 1.1123583316802979, + "learning_rate": 4.299071347170949e-05, + "loss": 1.0453, + "step": 4656 + }, + { + "epoch": 0.3199807613027346, + "grad_norm": 1.0829827785491943, + "learning_rate": 4.298654935310261e-05, + "loss": 0.994, + "step": 4657 + }, + { + "epoch": 0.3200494709358252, + "grad_norm": 1.1078524589538574, + "learning_rate": 4.2982384199740526e-05, + "loss": 0.9507, + "step": 4658 + }, + { + "epoch": 0.3201181805689158, + "grad_norm": 1.034843921661377, + "learning_rate": 4.297821801186284e-05, + "loss": 0.9529, + "step": 4659 + }, + { + "epoch": 0.3201868902020063, + "grad_norm": 1.0854064226150513, + "learning_rate": 4.297405078970924e-05, + "loss": 0.9533, + "step": 4660 + }, + { + "epoch": 0.3202555998350969, + "grad_norm": 1.1395134925842285, + "learning_rate": 4.2969882533519444e-05, + "loss": 1.0616, + "step": 4661 + }, + { + "epoch": 0.32032430946818746, + "grad_norm": 1.1067450046539307, + "learning_rate": 4.296571324353327e-05, + "loss": 1.0394, + "step": 4662 + }, + { + "epoch": 0.320393019101278, + "grad_norm": 1.14835786819458, + "learning_rate": 4.296154291999057e-05, + "loss": 0.9858, + "step": 4663 + }, + { + "epoch": 0.32046172873436857, + "grad_norm": 1.1638574600219727, + "learning_rate": 4.295737156313126e-05, + "loss": 1.0002, + "step": 4664 + }, + { + "epoch": 0.3205304383674591, + "grad_norm": 1.0767050981521606, + "learning_rate": 4.2953199173195305e-05, + "loss": 0.9896, + "step": 4665 + }, + { + "epoch": 0.32059914800054967, + "grad_norm": 1.1680127382278442, + "learning_rate": 4.2949025750422755e-05, + "loss": 0.9926, + "step": 4666 + }, + { + "epoch": 0.32066785763364025, + "grad_norm": 1.1209317445755005, + "learning_rate": 4.294485129505369e-05, + "loss": 0.9994, + "step": 4667 + }, + { + "epoch": 0.3207365672667308, + "grad_norm": 1.1034845113754272, + "learning_rate": 4.2940675807328276e-05, + "loss": 1.0246, + "step": 4668 + }, + { + "epoch": 0.32080527689982136, + "grad_norm": 1.078668475151062, + "learning_rate": 4.293649928748673e-05, + "loss": 0.9739, + "step": 4669 + }, + { + "epoch": 0.32087398653291194, + "grad_norm": 1.067981481552124, + "learning_rate": 4.29323217357693e-05, + "loss": 0.9972, + "step": 4670 + }, + { + "epoch": 0.32094269616600246, + "grad_norm": 1.1239792108535767, + "learning_rate": 4.2928143152416346e-05, + "loss": 0.9698, + "step": 4671 + }, + { + "epoch": 0.32101140579909304, + "grad_norm": 1.1457738876342773, + "learning_rate": 4.292396353766824e-05, + "loss": 0.9861, + "step": 4672 + }, + { + "epoch": 0.32108011543218357, + "grad_norm": 1.1030749082565308, + "learning_rate": 4.291978289176545e-05, + "loss": 1.0093, + "step": 4673 + }, + { + "epoch": 0.32114882506527415, + "grad_norm": 1.1338485479354858, + "learning_rate": 4.291560121494848e-05, + "loss": 1.0086, + "step": 4674 + }, + { + "epoch": 0.3212175346983647, + "grad_norm": 1.1290864944458008, + "learning_rate": 4.291141850745788e-05, + "loss": 1.0055, + "step": 4675 + }, + { + "epoch": 0.32128624433145525, + "grad_norm": 1.0861927270889282, + "learning_rate": 4.29072347695343e-05, + "loss": 1.0025, + "step": 4676 + }, + { + "epoch": 0.32135495396454583, + "grad_norm": 1.0866354703903198, + "learning_rate": 4.2903050001418424e-05, + "loss": 0.9988, + "step": 4677 + }, + { + "epoch": 0.3214236635976364, + "grad_norm": 1.1438888311386108, + "learning_rate": 4.2898864203351e-05, + "loss": 1.0224, + "step": 4678 + }, + { + "epoch": 0.32149237323072694, + "grad_norm": 1.056691288948059, + "learning_rate": 4.2894677375572835e-05, + "loss": 0.9858, + "step": 4679 + }, + { + "epoch": 0.3215610828638175, + "grad_norm": 1.1276191473007202, + "learning_rate": 4.2890489518324785e-05, + "loss": 0.9979, + "step": 4680 + }, + { + "epoch": 0.32162979249690804, + "grad_norm": 1.0246254205703735, + "learning_rate": 4.288630063184778e-05, + "loss": 0.9891, + "step": 4681 + }, + { + "epoch": 0.3216985021299986, + "grad_norm": 1.132500410079956, + "learning_rate": 4.2882110716382806e-05, + "loss": 0.9412, + "step": 4682 + }, + { + "epoch": 0.3217672117630892, + "grad_norm": 1.0650873184204102, + "learning_rate": 4.287791977217091e-05, + "loss": 0.9733, + "step": 4683 + }, + { + "epoch": 0.3218359213961797, + "grad_norm": 1.161197543144226, + "learning_rate": 4.287372779945319e-05, + "loss": 1.0244, + "step": 4684 + }, + { + "epoch": 0.3219046310292703, + "grad_norm": 1.0938704013824463, + "learning_rate": 4.28695347984708e-05, + "loss": 0.9913, + "step": 4685 + }, + { + "epoch": 0.3219733406623609, + "grad_norm": 1.063938021659851, + "learning_rate": 4.286534076946498e-05, + "loss": 0.9932, + "step": 4686 + }, + { + "epoch": 0.3220420502954514, + "grad_norm": 1.1960246562957764, + "learning_rate": 4.2861145712677e-05, + "loss": 0.998, + "step": 4687 + }, + { + "epoch": 0.322110759928542, + "grad_norm": 1.1957292556762695, + "learning_rate": 4.2856949628348184e-05, + "loss": 1.0355, + "step": 4688 + }, + { + "epoch": 0.3221794695616325, + "grad_norm": 1.0833055973052979, + "learning_rate": 4.2852752516719953e-05, + "loss": 0.9896, + "step": 4689 + }, + { + "epoch": 0.3222481791947231, + "grad_norm": 1.0790338516235352, + "learning_rate": 4.2848554378033755e-05, + "loss": 0.957, + "step": 4690 + }, + { + "epoch": 0.3223168888278137, + "grad_norm": 1.076296091079712, + "learning_rate": 4.28443552125311e-05, + "loss": 0.9829, + "step": 4691 + }, + { + "epoch": 0.3223855984609042, + "grad_norm": 1.0516005754470825, + "learning_rate": 4.2840155020453576e-05, + "loss": 0.9809, + "step": 4692 + }, + { + "epoch": 0.3224543080939948, + "grad_norm": 1.09862220287323, + "learning_rate": 4.283595380204282e-05, + "loss": 0.9797, + "step": 4693 + }, + { + "epoch": 0.32252301772708536, + "grad_norm": 1.0823651552200317, + "learning_rate": 4.28317515575405e-05, + "loss": 0.9984, + "step": 4694 + }, + { + "epoch": 0.3225917273601759, + "grad_norm": 1.0220662355422974, + "learning_rate": 4.282754828718839e-05, + "loss": 0.9583, + "step": 4695 + }, + { + "epoch": 0.32266043699326646, + "grad_norm": 1.041246771812439, + "learning_rate": 4.2823343991228304e-05, + "loss": 1.0042, + "step": 4696 + }, + { + "epoch": 0.322729146626357, + "grad_norm": 1.1460217237472534, + "learning_rate": 4.28191386699021e-05, + "loss": 1.0379, + "step": 4697 + }, + { + "epoch": 0.32279785625944757, + "grad_norm": 1.0696783065795898, + "learning_rate": 4.281493232345171e-05, + "loss": 0.995, + "step": 4698 + }, + { + "epoch": 0.32286656589253815, + "grad_norm": 1.1381726264953613, + "learning_rate": 4.281072495211913e-05, + "loss": 0.987, + "step": 4699 + }, + { + "epoch": 0.3229352755256287, + "grad_norm": 1.1346697807312012, + "learning_rate": 4.2806516556146405e-05, + "loss": 0.978, + "step": 4700 + }, + { + "epoch": 0.32300398515871925, + "grad_norm": 1.1485037803649902, + "learning_rate": 4.280230713577564e-05, + "loss": 0.9841, + "step": 4701 + }, + { + "epoch": 0.32307269479180983, + "grad_norm": 1.0991997718811035, + "learning_rate": 4.2798096691249e-05, + "loss": 0.9773, + "step": 4702 + }, + { + "epoch": 0.32314140442490036, + "grad_norm": 1.0619418621063232, + "learning_rate": 4.2793885222808706e-05, + "loss": 0.9489, + "step": 4703 + }, + { + "epoch": 0.32321011405799094, + "grad_norm": 1.0625897645950317, + "learning_rate": 4.278967273069705e-05, + "loss": 0.9981, + "step": 4704 + }, + { + "epoch": 0.32327882369108146, + "grad_norm": 1.1011927127838135, + "learning_rate": 4.2785459215156364e-05, + "loss": 1.0011, + "step": 4705 + }, + { + "epoch": 0.32334753332417204, + "grad_norm": 1.190456509590149, + "learning_rate": 4.2781244676429056e-05, + "loss": 1.0125, + "step": 4706 + }, + { + "epoch": 0.3234162429572626, + "grad_norm": 1.1073412895202637, + "learning_rate": 4.277702911475758e-05, + "loss": 1.0355, + "step": 4707 + }, + { + "epoch": 0.32348495259035315, + "grad_norm": 1.1136884689331055, + "learning_rate": 4.2772812530384457e-05, + "loss": 1.0145, + "step": 4708 + }, + { + "epoch": 0.3235536622234437, + "grad_norm": 1.1451362371444702, + "learning_rate": 4.276859492355228e-05, + "loss": 1.0206, + "step": 4709 + }, + { + "epoch": 0.3236223718565343, + "grad_norm": 1.140942096710205, + "learning_rate": 4.276437629450365e-05, + "loss": 1.0297, + "step": 4710 + }, + { + "epoch": 0.32369108148962483, + "grad_norm": 1.0269604921340942, + "learning_rate": 4.27601566434813e-05, + "loss": 0.9822, + "step": 4711 + }, + { + "epoch": 0.3237597911227154, + "grad_norm": 1.0508328676223755, + "learning_rate": 4.275593597072796e-05, + "loss": 0.9635, + "step": 4712 + }, + { + "epoch": 0.32382850075580594, + "grad_norm": 1.0657600164413452, + "learning_rate": 4.275171427648645e-05, + "loss": 1.0396, + "step": 4713 + }, + { + "epoch": 0.3238972103888965, + "grad_norm": 1.0645363330841064, + "learning_rate": 4.2747491560999644e-05, + "loss": 0.979, + "step": 4714 + }, + { + "epoch": 0.3239659200219871, + "grad_norm": 1.150546669960022, + "learning_rate": 4.274326782451047e-05, + "loss": 1.0315, + "step": 4715 + }, + { + "epoch": 0.3240346296550776, + "grad_norm": 1.0557105541229248, + "learning_rate": 4.2739043067261913e-05, + "loss": 0.9472, + "step": 4716 + }, + { + "epoch": 0.3241033392881682, + "grad_norm": 1.0410869121551514, + "learning_rate": 4.273481728949703e-05, + "loss": 0.9558, + "step": 4717 + }, + { + "epoch": 0.3241720489212588, + "grad_norm": 1.0328891277313232, + "learning_rate": 4.273059049145892e-05, + "loss": 0.988, + "step": 4718 + }, + { + "epoch": 0.3242407585543493, + "grad_norm": 1.169754981994629, + "learning_rate": 4.272636267339074e-05, + "loss": 1.006, + "step": 4719 + }, + { + "epoch": 0.3243094681874399, + "grad_norm": 1.1208182573318481, + "learning_rate": 4.272213383553574e-05, + "loss": 1.0313, + "step": 4720 + }, + { + "epoch": 0.3243781778205304, + "grad_norm": 1.0847489833831787, + "learning_rate": 4.2717903978137176e-05, + "loss": 1.0232, + "step": 4721 + }, + { + "epoch": 0.324446887453621, + "grad_norm": 1.1201448440551758, + "learning_rate": 4.27136731014384e-05, + "loss": 0.9637, + "step": 4722 + }, + { + "epoch": 0.32451559708671157, + "grad_norm": 1.0845615863800049, + "learning_rate": 4.2709441205682815e-05, + "loss": 0.9865, + "step": 4723 + }, + { + "epoch": 0.3245843067198021, + "grad_norm": 1.0875004529953003, + "learning_rate": 4.2705208291113875e-05, + "loss": 0.9616, + "step": 4724 + }, + { + "epoch": 0.3246530163528927, + "grad_norm": 1.023398756980896, + "learning_rate": 4.2700974357975096e-05, + "loss": 0.9469, + "step": 4725 + }, + { + "epoch": 0.32472172598598326, + "grad_norm": 1.0545376539230347, + "learning_rate": 4.269673940651005e-05, + "loss": 0.9894, + "step": 4726 + }, + { + "epoch": 0.3247904356190738, + "grad_norm": 1.089643120765686, + "learning_rate": 4.269250343696238e-05, + "loss": 1.0335, + "step": 4727 + }, + { + "epoch": 0.32485914525216436, + "grad_norm": 1.1092860698699951, + "learning_rate": 4.2688266449575784e-05, + "loss": 1.0365, + "step": 4728 + }, + { + "epoch": 0.3249278548852549, + "grad_norm": 1.1308962106704712, + "learning_rate": 4.268402844459399e-05, + "loss": 1.017, + "step": 4729 + }, + { + "epoch": 0.32499656451834547, + "grad_norm": 1.1664996147155762, + "learning_rate": 4.267978942226083e-05, + "loss": 0.9733, + "step": 4730 + }, + { + "epoch": 0.32506527415143605, + "grad_norm": 1.0371121168136597, + "learning_rate": 4.267554938282016e-05, + "loss": 0.9678, + "step": 4731 + }, + { + "epoch": 0.32513398378452657, + "grad_norm": 1.0567210912704468, + "learning_rate": 4.267130832651592e-05, + "loss": 1.0084, + "step": 4732 + }, + { + "epoch": 0.32520269341761715, + "grad_norm": 1.005855917930603, + "learning_rate": 4.266706625359208e-05, + "loss": 0.9699, + "step": 4733 + }, + { + "epoch": 0.32527140305070773, + "grad_norm": 1.1423777341842651, + "learning_rate": 4.266282316429269e-05, + "loss": 1.0213, + "step": 4734 + }, + { + "epoch": 0.32534011268379825, + "grad_norm": 1.0777506828308105, + "learning_rate": 4.265857905886186e-05, + "loss": 1.0086, + "step": 4735 + }, + { + "epoch": 0.32540882231688883, + "grad_norm": 1.175920009613037, + "learning_rate": 4.2654333937543736e-05, + "loss": 0.995, + "step": 4736 + }, + { + "epoch": 0.32547753194997936, + "grad_norm": 1.0674996376037598, + "learning_rate": 4.265008780058255e-05, + "loss": 0.9971, + "step": 4737 + }, + { + "epoch": 0.32554624158306994, + "grad_norm": 1.0581142902374268, + "learning_rate": 4.2645840648222565e-05, + "loss": 0.9746, + "step": 4738 + }, + { + "epoch": 0.3256149512161605, + "grad_norm": 1.129329800605774, + "learning_rate": 4.264159248070814e-05, + "loss": 0.9901, + "step": 4739 + }, + { + "epoch": 0.32568366084925104, + "grad_norm": 1.063292384147644, + "learning_rate": 4.263734329828365e-05, + "loss": 0.9373, + "step": 4740 + }, + { + "epoch": 0.3257523704823416, + "grad_norm": 1.0612082481384277, + "learning_rate": 4.2633093101193547e-05, + "loss": 0.9631, + "step": 4741 + }, + { + "epoch": 0.3258210801154322, + "grad_norm": 1.0930711030960083, + "learning_rate": 4.262884188968236e-05, + "loss": 0.9971, + "step": 4742 + }, + { + "epoch": 0.32588978974852273, + "grad_norm": 1.1406854391098022, + "learning_rate": 4.2624589663994645e-05, + "loss": 0.9957, + "step": 4743 + }, + { + "epoch": 0.3259584993816133, + "grad_norm": 1.094336986541748, + "learning_rate": 4.262033642437503e-05, + "loss": 0.9918, + "step": 4744 + }, + { + "epoch": 0.32602720901470383, + "grad_norm": 1.16280198097229, + "learning_rate": 4.261608217106821e-05, + "loss": 1.023, + "step": 4745 + }, + { + "epoch": 0.3260959186477944, + "grad_norm": 1.044719934463501, + "learning_rate": 4.261182690431892e-05, + "loss": 0.9863, + "step": 4746 + }, + { + "epoch": 0.326164628280885, + "grad_norm": 1.084218144416809, + "learning_rate": 4.260757062437196e-05, + "loss": 1.0066, + "step": 4747 + }, + { + "epoch": 0.3262333379139755, + "grad_norm": 1.1104875802993774, + "learning_rate": 4.260331333147222e-05, + "loss": 1.019, + "step": 4748 + }, + { + "epoch": 0.3263020475470661, + "grad_norm": 1.126665711402893, + "learning_rate": 4.259905502586458e-05, + "loss": 1.0094, + "step": 4749 + }, + { + "epoch": 0.3263707571801567, + "grad_norm": 1.10634446144104, + "learning_rate": 4.259479570779403e-05, + "loss": 0.9856, + "step": 4750 + }, + { + "epoch": 0.3264394668132472, + "grad_norm": 1.0737894773483276, + "learning_rate": 4.259053537750562e-05, + "loss": 1.0052, + "step": 4751 + }, + { + "epoch": 0.3265081764463378, + "grad_norm": 1.1108721494674683, + "learning_rate": 4.2586274035244436e-05, + "loss": 1.0244, + "step": 4752 + }, + { + "epoch": 0.32657688607942836, + "grad_norm": 1.2207231521606445, + "learning_rate": 4.2582011681255626e-05, + "loss": 1.0704, + "step": 4753 + }, + { + "epoch": 0.3266455957125189, + "grad_norm": 1.0993447303771973, + "learning_rate": 4.25777483157844e-05, + "loss": 0.972, + "step": 4754 + }, + { + "epoch": 0.32671430534560947, + "grad_norm": 1.122963547706604, + "learning_rate": 4.257348393907604e-05, + "loss": 0.9917, + "step": 4755 + }, + { + "epoch": 0.3267830149787, + "grad_norm": 1.0360808372497559, + "learning_rate": 4.256921855137587e-05, + "loss": 0.971, + "step": 4756 + }, + { + "epoch": 0.3268517246117906, + "grad_norm": 1.116472840309143, + "learning_rate": 4.2564952152929256e-05, + "loss": 0.9809, + "step": 4757 + }, + { + "epoch": 0.32692043424488115, + "grad_norm": 1.1112686395645142, + "learning_rate": 4.256068474398166e-05, + "loss": 1.0212, + "step": 4758 + }, + { + "epoch": 0.3269891438779717, + "grad_norm": 1.1710503101348877, + "learning_rate": 4.2556416324778577e-05, + "loss": 1.0275, + "step": 4759 + }, + { + "epoch": 0.32705785351106226, + "grad_norm": 1.105372667312622, + "learning_rate": 4.255214689556557e-05, + "loss": 0.9855, + "step": 4760 + }, + { + "epoch": 0.32712656314415284, + "grad_norm": 1.0986448526382446, + "learning_rate": 4.2547876456588246e-05, + "loss": 0.9967, + "step": 4761 + }, + { + "epoch": 0.32719527277724336, + "grad_norm": 1.1257189512252808, + "learning_rate": 4.254360500809229e-05, + "loss": 1.0774, + "step": 4762 + }, + { + "epoch": 0.32726398241033394, + "grad_norm": 1.114073395729065, + "learning_rate": 4.253933255032345e-05, + "loss": 0.9739, + "step": 4763 + }, + { + "epoch": 0.32733269204342447, + "grad_norm": 1.0208971500396729, + "learning_rate": 4.253505908352749e-05, + "loss": 1.0006, + "step": 4764 + }, + { + "epoch": 0.32740140167651505, + "grad_norm": 1.1465214490890503, + "learning_rate": 4.2530784607950274e-05, + "loss": 1.0233, + "step": 4765 + }, + { + "epoch": 0.3274701113096056, + "grad_norm": 1.1296143531799316, + "learning_rate": 4.252650912383771e-05, + "loss": 1.0487, + "step": 4766 + }, + { + "epoch": 0.32753882094269615, + "grad_norm": 1.065604567527771, + "learning_rate": 4.252223263143576e-05, + "loss": 1.0166, + "step": 4767 + }, + { + "epoch": 0.32760753057578673, + "grad_norm": 1.1312202215194702, + "learning_rate": 4.251795513099045e-05, + "loss": 1.019, + "step": 4768 + }, + { + "epoch": 0.3276762402088773, + "grad_norm": 1.062427043914795, + "learning_rate": 4.2513676622747866e-05, + "loss": 1.0116, + "step": 4769 + }, + { + "epoch": 0.32774494984196784, + "grad_norm": 1.0744071006774902, + "learning_rate": 4.250939710695415e-05, + "loss": 0.984, + "step": 4770 + }, + { + "epoch": 0.3278136594750584, + "grad_norm": 1.2116323709487915, + "learning_rate": 4.2505116583855485e-05, + "loss": 1.0089, + "step": 4771 + }, + { + "epoch": 0.32788236910814894, + "grad_norm": 1.1192516088485718, + "learning_rate": 4.250083505369814e-05, + "loss": 1.0065, + "step": 4772 + }, + { + "epoch": 0.3279510787412395, + "grad_norm": 1.1120299100875854, + "learning_rate": 4.249655251672843e-05, + "loss": 1.0095, + "step": 4773 + }, + { + "epoch": 0.3280197883743301, + "grad_norm": 1.112690806388855, + "learning_rate": 4.249226897319271e-05, + "loss": 1.0225, + "step": 4774 + }, + { + "epoch": 0.3280884980074206, + "grad_norm": 1.230918526649475, + "learning_rate": 4.248798442333743e-05, + "loss": 1.0111, + "step": 4775 + }, + { + "epoch": 0.3281572076405112, + "grad_norm": 1.1359354257583618, + "learning_rate": 4.248369886740907e-05, + "loss": 1.03, + "step": 4776 + }, + { + "epoch": 0.3282259172736018, + "grad_norm": 1.0319499969482422, + "learning_rate": 4.247941230565417e-05, + "loss": 0.9694, + "step": 4777 + }, + { + "epoch": 0.3282946269066923, + "grad_norm": 1.0889101028442383, + "learning_rate": 4.247512473831934e-05, + "loss": 0.9823, + "step": 4778 + }, + { + "epoch": 0.3283633365397829, + "grad_norm": 1.0808792114257812, + "learning_rate": 4.247083616565124e-05, + "loss": 1.009, + "step": 4779 + }, + { + "epoch": 0.3284320461728734, + "grad_norm": 1.1098527908325195, + "learning_rate": 4.246654658789658e-05, + "loss": 0.9897, + "step": 4780 + }, + { + "epoch": 0.328500755805964, + "grad_norm": 1.1149108409881592, + "learning_rate": 4.2462256005302157e-05, + "loss": 0.9667, + "step": 4781 + }, + { + "epoch": 0.3285694654390546, + "grad_norm": 1.0974942445755005, + "learning_rate": 4.245796441811479e-05, + "loss": 0.9908, + "step": 4782 + }, + { + "epoch": 0.3286381750721451, + "grad_norm": 1.0324515104293823, + "learning_rate": 4.2453671826581374e-05, + "loss": 0.953, + "step": 4783 + }, + { + "epoch": 0.3287068847052357, + "grad_norm": 1.2215732336044312, + "learning_rate": 4.244937823094885e-05, + "loss": 1.0044, + "step": 4784 + }, + { + "epoch": 0.32877559433832626, + "grad_norm": 1.1111106872558594, + "learning_rate": 4.244508363146425e-05, + "loss": 1.0159, + "step": 4785 + }, + { + "epoch": 0.3288443039714168, + "grad_norm": 1.1595323085784912, + "learning_rate": 4.2440788028374624e-05, + "loss": 1.0253, + "step": 4786 + }, + { + "epoch": 0.32891301360450736, + "grad_norm": 1.0509164333343506, + "learning_rate": 4.243649142192709e-05, + "loss": 0.9959, + "step": 4787 + }, + { + "epoch": 0.3289817232375979, + "grad_norm": 1.1242308616638184, + "learning_rate": 4.243219381236884e-05, + "loss": 0.9619, + "step": 4788 + }, + { + "epoch": 0.32905043287068847, + "grad_norm": 1.1492544412612915, + "learning_rate": 4.242789519994711e-05, + "loss": 0.9973, + "step": 4789 + }, + { + "epoch": 0.32911914250377905, + "grad_norm": 1.1480127573013306, + "learning_rate": 4.242359558490919e-05, + "loss": 1.0463, + "step": 4790 + }, + { + "epoch": 0.3291878521368696, + "grad_norm": 1.1790097951889038, + "learning_rate": 4.2419294967502445e-05, + "loss": 1.0006, + "step": 4791 + }, + { + "epoch": 0.32925656176996015, + "grad_norm": 1.1200451850891113, + "learning_rate": 4.2414993347974274e-05, + "loss": 1.0011, + "step": 4792 + }, + { + "epoch": 0.32932527140305073, + "grad_norm": 1.109589695930481, + "learning_rate": 4.2410690726572166e-05, + "loss": 1.0222, + "step": 4793 + }, + { + "epoch": 0.32939398103614126, + "grad_norm": 1.1510181427001953, + "learning_rate": 4.240638710354363e-05, + "loss": 1.0268, + "step": 4794 + }, + { + "epoch": 0.32946269066923184, + "grad_norm": 1.1625566482543945, + "learning_rate": 4.240208247913625e-05, + "loss": 1.0053, + "step": 4795 + }, + { + "epoch": 0.32953140030232236, + "grad_norm": 1.0356121063232422, + "learning_rate": 4.239777685359769e-05, + "loss": 0.9551, + "step": 4796 + }, + { + "epoch": 0.32960010993541294, + "grad_norm": 1.053629755973816, + "learning_rate": 4.2393470227175623e-05, + "loss": 1.0185, + "step": 4797 + }, + { + "epoch": 0.3296688195685035, + "grad_norm": 1.1042481660842896, + "learning_rate": 4.2389162600117836e-05, + "loss": 0.9908, + "step": 4798 + }, + { + "epoch": 0.32973752920159405, + "grad_norm": 1.086660385131836, + "learning_rate": 4.238485397267211e-05, + "loss": 1.0025, + "step": 4799 + }, + { + "epoch": 0.32980623883468463, + "grad_norm": 1.0564237833023071, + "learning_rate": 4.238054434508634e-05, + "loss": 0.9885, + "step": 4800 + }, + { + "epoch": 0.3298749484677752, + "grad_norm": 1.070617914199829, + "learning_rate": 4.2376233717608454e-05, + "loss": 0.9686, + "step": 4801 + }, + { + "epoch": 0.32994365810086573, + "grad_norm": 1.090632438659668, + "learning_rate": 4.2371922090486436e-05, + "loss": 1.0132, + "step": 4802 + }, + { + "epoch": 0.3300123677339563, + "grad_norm": 1.0695592164993286, + "learning_rate": 4.236760946396833e-05, + "loss": 1.0411, + "step": 4803 + }, + { + "epoch": 0.33008107736704684, + "grad_norm": 1.1045160293579102, + "learning_rate": 4.236329583830224e-05, + "loss": 1.0076, + "step": 4804 + }, + { + "epoch": 0.3301497870001374, + "grad_norm": 1.14304518699646, + "learning_rate": 4.235898121373634e-05, + "loss": 0.9885, + "step": 4805 + }, + { + "epoch": 0.330218496633228, + "grad_norm": 1.185180902481079, + "learning_rate": 4.2354665590518825e-05, + "loss": 0.9561, + "step": 4806 + }, + { + "epoch": 0.3302872062663185, + "grad_norm": 1.104315161705017, + "learning_rate": 4.235034896889798e-05, + "loss": 1.0025, + "step": 4807 + }, + { + "epoch": 0.3303559158994091, + "grad_norm": 1.0800213813781738, + "learning_rate": 4.234603134912214e-05, + "loss": 1.0166, + "step": 4808 + }, + { + "epoch": 0.3304246255324997, + "grad_norm": 1.051841378211975, + "learning_rate": 4.234171273143968e-05, + "loss": 0.9923, + "step": 4809 + }, + { + "epoch": 0.3304933351655902, + "grad_norm": 1.0894834995269775, + "learning_rate": 4.233739311609908e-05, + "loss": 1.0103, + "step": 4810 + }, + { + "epoch": 0.3305620447986808, + "grad_norm": 1.0932170152664185, + "learning_rate": 4.233307250334882e-05, + "loss": 0.9682, + "step": 4811 + }, + { + "epoch": 0.3306307544317713, + "grad_norm": 1.0945360660552979, + "learning_rate": 4.232875089343746e-05, + "loss": 1.0303, + "step": 4812 + }, + { + "epoch": 0.3306994640648619, + "grad_norm": 1.0975521802902222, + "learning_rate": 4.2324428286613636e-05, + "loss": 1.0041, + "step": 4813 + }, + { + "epoch": 0.3307681736979525, + "grad_norm": 1.0654882192611694, + "learning_rate": 4.232010468312601e-05, + "loss": 1.002, + "step": 4814 + }, + { + "epoch": 0.330836883331043, + "grad_norm": 1.066735029220581, + "learning_rate": 4.231578008322332e-05, + "loss": 0.9495, + "step": 4815 + }, + { + "epoch": 0.3309055929641336, + "grad_norm": 1.1439697742462158, + "learning_rate": 4.2311454487154375e-05, + "loss": 0.9567, + "step": 4816 + }, + { + "epoch": 0.33097430259722416, + "grad_norm": 0.9923565983772278, + "learning_rate": 4.2307127895167994e-05, + "loss": 0.9634, + "step": 4817 + }, + { + "epoch": 0.3310430122303147, + "grad_norm": 1.0808674097061157, + "learning_rate": 4.23028003075131e-05, + "loss": 0.9945, + "step": 4818 + }, + { + "epoch": 0.33111172186340526, + "grad_norm": 1.0947425365447998, + "learning_rate": 4.229847172443866e-05, + "loss": 0.9966, + "step": 4819 + }, + { + "epoch": 0.3311804314964958, + "grad_norm": 1.0799939632415771, + "learning_rate": 4.2294142146193686e-05, + "loss": 1.0135, + "step": 4820 + }, + { + "epoch": 0.33124914112958637, + "grad_norm": 1.0946165323257446, + "learning_rate": 4.228981157302726e-05, + "loss": 1.04, + "step": 4821 + }, + { + "epoch": 0.33131785076267695, + "grad_norm": 1.0931029319763184, + "learning_rate": 4.228548000518851e-05, + "loss": 1.0143, + "step": 4822 + }, + { + "epoch": 0.33138656039576747, + "grad_norm": 1.1608744859695435, + "learning_rate": 4.228114744292664e-05, + "loss": 1.0126, + "step": 4823 + }, + { + "epoch": 0.33145527002885805, + "grad_norm": 1.0812007188796997, + "learning_rate": 4.22768138864909e-05, + "loss": 0.9948, + "step": 4824 + }, + { + "epoch": 0.33152397966194863, + "grad_norm": 1.0460643768310547, + "learning_rate": 4.227247933613058e-05, + "loss": 0.9901, + "step": 4825 + }, + { + "epoch": 0.33159268929503916, + "grad_norm": 1.051575779914856, + "learning_rate": 4.226814379209506e-05, + "loss": 1.0199, + "step": 4826 + }, + { + "epoch": 0.33166139892812974, + "grad_norm": 1.0848979949951172, + "learning_rate": 4.2263807254633756e-05, + "loss": 0.9878, + "step": 4827 + }, + { + "epoch": 0.33173010856122026, + "grad_norm": 1.0845032930374146, + "learning_rate": 4.2259469723996146e-05, + "loss": 0.9724, + "step": 4828 + }, + { + "epoch": 0.33179881819431084, + "grad_norm": 1.0510774850845337, + "learning_rate": 4.225513120043177e-05, + "loss": 0.974, + "step": 4829 + }, + { + "epoch": 0.3318675278274014, + "grad_norm": 1.1318233013153076, + "learning_rate": 4.2250791684190213e-05, + "loss": 0.9883, + "step": 4830 + }, + { + "epoch": 0.33193623746049195, + "grad_norm": 1.0574135780334473, + "learning_rate": 4.224645117552113e-05, + "loss": 0.977, + "step": 4831 + }, + { + "epoch": 0.3320049470935825, + "grad_norm": 1.041151523590088, + "learning_rate": 4.2242109674674236e-05, + "loss": 0.9385, + "step": 4832 + }, + { + "epoch": 0.3320736567266731, + "grad_norm": 1.1329985857009888, + "learning_rate": 4.223776718189928e-05, + "loss": 1.0167, + "step": 4833 + }, + { + "epoch": 0.33214236635976363, + "grad_norm": 1.121299147605896, + "learning_rate": 4.2233423697446086e-05, + "loss": 0.9701, + "step": 4834 + }, + { + "epoch": 0.3322110759928542, + "grad_norm": 1.1185898780822754, + "learning_rate": 4.2229079221564535e-05, + "loss": 1.0098, + "step": 4835 + }, + { + "epoch": 0.33227978562594473, + "grad_norm": 1.051711082458496, + "learning_rate": 4.222473375450455e-05, + "loss": 1.0036, + "step": 4836 + }, + { + "epoch": 0.3323484952590353, + "grad_norm": 1.074851393699646, + "learning_rate": 4.222038729651615e-05, + "loss": 0.9794, + "step": 4837 + }, + { + "epoch": 0.3324172048921259, + "grad_norm": 1.087506651878357, + "learning_rate": 4.221603984784936e-05, + "loss": 1.0071, + "step": 4838 + }, + { + "epoch": 0.3324859145252164, + "grad_norm": 1.0723857879638672, + "learning_rate": 4.2211691408754305e-05, + "loss": 0.9937, + "step": 4839 + }, + { + "epoch": 0.332554624158307, + "grad_norm": 1.1382544040679932, + "learning_rate": 4.220734197948113e-05, + "loss": 1.0248, + "step": 4840 + }, + { + "epoch": 0.3326233337913976, + "grad_norm": 1.1363027095794678, + "learning_rate": 4.220299156028007e-05, + "loss": 1.0153, + "step": 4841 + }, + { + "epoch": 0.3326920434244881, + "grad_norm": 1.1630833148956299, + "learning_rate": 4.219864015140139e-05, + "loss": 1.033, + "step": 4842 + }, + { + "epoch": 0.3327607530575787, + "grad_norm": 1.08159601688385, + "learning_rate": 4.219428775309543e-05, + "loss": 0.9807, + "step": 4843 + }, + { + "epoch": 0.3328294626906692, + "grad_norm": 1.010703444480896, + "learning_rate": 4.218993436561258e-05, + "loss": 0.9938, + "step": 4844 + }, + { + "epoch": 0.3328981723237598, + "grad_norm": 1.129599928855896, + "learning_rate": 4.218557998920328e-05, + "loss": 1.0244, + "step": 4845 + }, + { + "epoch": 0.33296688195685037, + "grad_norm": 1.0727709531784058, + "learning_rate": 4.218122462411804e-05, + "loss": 0.9707, + "step": 4846 + }, + { + "epoch": 0.3330355915899409, + "grad_norm": 1.0579936504364014, + "learning_rate": 4.217686827060744e-05, + "loss": 0.9646, + "step": 4847 + }, + { + "epoch": 0.3331043012230315, + "grad_norm": 1.1781926155090332, + "learning_rate": 4.217251092892207e-05, + "loss": 0.9787, + "step": 4848 + }, + { + "epoch": 0.33317301085612205, + "grad_norm": 1.1089544296264648, + "learning_rate": 4.2168152599312615e-05, + "loss": 0.999, + "step": 4849 + }, + { + "epoch": 0.3332417204892126, + "grad_norm": 1.150490164756775, + "learning_rate": 4.216379328202982e-05, + "loss": 1.0301, + "step": 4850 + }, + { + "epoch": 0.33331043012230316, + "grad_norm": 1.1702429056167603, + "learning_rate": 4.2159432977324445e-05, + "loss": 1.0235, + "step": 4851 + }, + { + "epoch": 0.3333791397553937, + "grad_norm": 1.0537000894546509, + "learning_rate": 4.2155071685447365e-05, + "loss": 0.9865, + "step": 4852 + }, + { + "epoch": 0.33344784938848426, + "grad_norm": 1.14378023147583, + "learning_rate": 4.2150709406649473e-05, + "loss": 0.9746, + "step": 4853 + }, + { + "epoch": 0.33351655902157484, + "grad_norm": 1.0676696300506592, + "learning_rate": 4.214634614118172e-05, + "loss": 0.9804, + "step": 4854 + }, + { + "epoch": 0.33358526865466537, + "grad_norm": 1.1732559204101562, + "learning_rate": 4.214198188929513e-05, + "loss": 0.9597, + "step": 4855 + }, + { + "epoch": 0.33365397828775595, + "grad_norm": 1.0395355224609375, + "learning_rate": 4.213761665124078e-05, + "loss": 0.945, + "step": 4856 + }, + { + "epoch": 0.33372268792084653, + "grad_norm": 1.1396015882492065, + "learning_rate": 4.213325042726979e-05, + "loss": 0.9947, + "step": 4857 + }, + { + "epoch": 0.33379139755393705, + "grad_norm": 1.101180076599121, + "learning_rate": 4.2128883217633355e-05, + "loss": 0.9898, + "step": 4858 + }, + { + "epoch": 0.33386010718702763, + "grad_norm": 1.0529305934906006, + "learning_rate": 4.21245150225827e-05, + "loss": 0.9723, + "step": 4859 + }, + { + "epoch": 0.33392881682011816, + "grad_norm": 1.0616697072982788, + "learning_rate": 4.212014584236914e-05, + "loss": 0.9949, + "step": 4860 + }, + { + "epoch": 0.33399752645320874, + "grad_norm": 1.1057884693145752, + "learning_rate": 4.2115775677244027e-05, + "loss": 0.981, + "step": 4861 + }, + { + "epoch": 0.3340662360862993, + "grad_norm": 1.0686067342758179, + "learning_rate": 4.211140452745879e-05, + "loss": 0.9867, + "step": 4862 + }, + { + "epoch": 0.33413494571938984, + "grad_norm": 1.1240755319595337, + "learning_rate": 4.2107032393264864e-05, + "loss": 0.9581, + "step": 4863 + }, + { + "epoch": 0.3342036553524804, + "grad_norm": 1.0785539150238037, + "learning_rate": 4.2102659274913813e-05, + "loss": 0.9939, + "step": 4864 + }, + { + "epoch": 0.334272364985571, + "grad_norm": 1.0462696552276611, + "learning_rate": 4.209828517265719e-05, + "loss": 1.0067, + "step": 4865 + }, + { + "epoch": 0.3343410746186615, + "grad_norm": 1.0456403493881226, + "learning_rate": 4.209391008674665e-05, + "loss": 0.9513, + "step": 4866 + }, + { + "epoch": 0.3344097842517521, + "grad_norm": 1.215147852897644, + "learning_rate": 4.2089534017433893e-05, + "loss": 1.0304, + "step": 4867 + }, + { + "epoch": 0.33447849388484263, + "grad_norm": 1.1115868091583252, + "learning_rate": 4.208515696497065e-05, + "loss": 1.0318, + "step": 4868 + }, + { + "epoch": 0.3345472035179332, + "grad_norm": 1.0785233974456787, + "learning_rate": 4.208077892960876e-05, + "loss": 0.9978, + "step": 4869 + }, + { + "epoch": 0.3346159131510238, + "grad_norm": 1.1226460933685303, + "learning_rate": 4.207639991160006e-05, + "loss": 0.9613, + "step": 4870 + }, + { + "epoch": 0.3346846227841143, + "grad_norm": 1.1248666048049927, + "learning_rate": 4.2072019911196484e-05, + "loss": 1.0044, + "step": 4871 + }, + { + "epoch": 0.3347533324172049, + "grad_norm": 1.0929466485977173, + "learning_rate": 4.206763892865002e-05, + "loss": 0.9867, + "step": 4872 + }, + { + "epoch": 0.3348220420502955, + "grad_norm": 1.0990355014801025, + "learning_rate": 4.206325696421269e-05, + "loss": 0.9945, + "step": 4873 + }, + { + "epoch": 0.334890751683386, + "grad_norm": 1.1515254974365234, + "learning_rate": 4.205887401813659e-05, + "loss": 1.016, + "step": 4874 + }, + { + "epoch": 0.3349594613164766, + "grad_norm": 1.058943271636963, + "learning_rate": 4.2054490090673875e-05, + "loss": 0.9958, + "step": 4875 + }, + { + "epoch": 0.3350281709495671, + "grad_norm": 1.0781149864196777, + "learning_rate": 4.205010518207674e-05, + "loss": 0.9958, + "step": 4876 + }, + { + "epoch": 0.3350968805826577, + "grad_norm": 1.0941909551620483, + "learning_rate": 4.204571929259745e-05, + "loss": 0.9867, + "step": 4877 + }, + { + "epoch": 0.33516559021574827, + "grad_norm": 1.1730581521987915, + "learning_rate": 4.204133242248832e-05, + "loss": 1.0713, + "step": 4878 + }, + { + "epoch": 0.3352342998488388, + "grad_norm": 1.0805975198745728, + "learning_rate": 4.203694457200173e-05, + "loss": 1.025, + "step": 4879 + }, + { + "epoch": 0.33530300948192937, + "grad_norm": 1.0981208086013794, + "learning_rate": 4.203255574139011e-05, + "loss": 0.9951, + "step": 4880 + }, + { + "epoch": 0.33537171911501995, + "grad_norm": 1.1204429864883423, + "learning_rate": 4.2028165930905936e-05, + "loss": 1.0032, + "step": 4881 + }, + { + "epoch": 0.3354404287481105, + "grad_norm": 1.0851795673370361, + "learning_rate": 4.2023775140801766e-05, + "loss": 0.9931, + "step": 4882 + }, + { + "epoch": 0.33550913838120106, + "grad_norm": 1.0711805820465088, + "learning_rate": 4.201938337133019e-05, + "loss": 0.9479, + "step": 4883 + }, + { + "epoch": 0.3355778480142916, + "grad_norm": 1.0827966928482056, + "learning_rate": 4.2014990622743864e-05, + "loss": 0.9951, + "step": 4884 + }, + { + "epoch": 0.33564655764738216, + "grad_norm": 1.0780737400054932, + "learning_rate": 4.20105968952955e-05, + "loss": 0.9866, + "step": 4885 + }, + { + "epoch": 0.33571526728047274, + "grad_norm": 1.1153512001037598, + "learning_rate": 4.200620218923787e-05, + "loss": 0.9666, + "step": 4886 + }, + { + "epoch": 0.33578397691356326, + "grad_norm": 1.2192682027816772, + "learning_rate": 4.20018065048238e-05, + "loss": 0.9972, + "step": 4887 + }, + { + "epoch": 0.33585268654665384, + "grad_norm": 1.061566948890686, + "learning_rate": 4.199740984230617e-05, + "loss": 0.9713, + "step": 4888 + }, + { + "epoch": 0.3359213961797444, + "grad_norm": 1.0936275720596313, + "learning_rate": 4.199301220193791e-05, + "loss": 1.0245, + "step": 4889 + }, + { + "epoch": 0.33599010581283495, + "grad_norm": 1.1056740283966064, + "learning_rate": 4.198861358397203e-05, + "loss": 0.9952, + "step": 4890 + }, + { + "epoch": 0.33605881544592553, + "grad_norm": 1.066258192062378, + "learning_rate": 4.1984213988661544e-05, + "loss": 0.9812, + "step": 4891 + }, + { + "epoch": 0.33612752507901605, + "grad_norm": 1.1655532121658325, + "learning_rate": 4.197981341625961e-05, + "loss": 0.9981, + "step": 4892 + }, + { + "epoch": 0.33619623471210663, + "grad_norm": 1.0962108373641968, + "learning_rate": 4.197541186701935e-05, + "loss": 1.0059, + "step": 4893 + }, + { + "epoch": 0.3362649443451972, + "grad_norm": 1.0914340019226074, + "learning_rate": 4.1971009341194e-05, + "loss": 0.9742, + "step": 4894 + }, + { + "epoch": 0.33633365397828774, + "grad_norm": 1.1467957496643066, + "learning_rate": 4.1966605839036834e-05, + "loss": 1.0268, + "step": 4895 + }, + { + "epoch": 0.3364023636113783, + "grad_norm": 1.052859902381897, + "learning_rate": 4.196220136080117e-05, + "loss": 0.9606, + "step": 4896 + }, + { + "epoch": 0.3364710732444689, + "grad_norm": 1.090179443359375, + "learning_rate": 4.195779590674041e-05, + "loss": 0.9925, + "step": 4897 + }, + { + "epoch": 0.3365397828775594, + "grad_norm": 1.0731751918792725, + "learning_rate": 4.1953389477107984e-05, + "loss": 0.9934, + "step": 4898 + }, + { + "epoch": 0.33660849251065, + "grad_norm": 1.1658220291137695, + "learning_rate": 4.194898207215741e-05, + "loss": 1.015, + "step": 4899 + }, + { + "epoch": 0.33667720214374053, + "grad_norm": 1.0854649543762207, + "learning_rate": 4.1944573692142217e-05, + "loss": 0.9677, + "step": 4900 + }, + { + "epoch": 0.3367459117768311, + "grad_norm": 1.0858501195907593, + "learning_rate": 4.194016433731605e-05, + "loss": 1.0099, + "step": 4901 + }, + { + "epoch": 0.3368146214099217, + "grad_norm": 1.1249938011169434, + "learning_rate": 4.193575400793254e-05, + "loss": 0.9821, + "step": 4902 + }, + { + "epoch": 0.3368833310430122, + "grad_norm": 1.0913915634155273, + "learning_rate": 4.1931342704245437e-05, + "loss": 1.0064, + "step": 4903 + }, + { + "epoch": 0.3369520406761028, + "grad_norm": 1.1256881952285767, + "learning_rate": 4.192693042650851e-05, + "loss": 1.01, + "step": 4904 + }, + { + "epoch": 0.3370207503091934, + "grad_norm": 1.0780035257339478, + "learning_rate": 4.19225171749756e-05, + "loss": 1.0009, + "step": 4905 + }, + { + "epoch": 0.3370894599422839, + "grad_norm": 1.0448402166366577, + "learning_rate": 4.191810294990059e-05, + "loss": 0.9786, + "step": 4906 + }, + { + "epoch": 0.3371581695753745, + "grad_norm": 1.0468318462371826, + "learning_rate": 4.191368775153743e-05, + "loss": 0.9506, + "step": 4907 + }, + { + "epoch": 0.337226879208465, + "grad_norm": 1.0410579442977905, + "learning_rate": 4.190927158014013e-05, + "loss": 0.9712, + "step": 4908 + }, + { + "epoch": 0.3372955888415556, + "grad_norm": 1.0675652027130127, + "learning_rate": 4.1904854435962755e-05, + "loss": 0.9573, + "step": 4909 + }, + { + "epoch": 0.33736429847464616, + "grad_norm": 1.0680217742919922, + "learning_rate": 4.19004363192594e-05, + "loss": 0.976, + "step": 4910 + }, + { + "epoch": 0.3374330081077367, + "grad_norm": 1.247410774230957, + "learning_rate": 4.189601723028426e-05, + "loss": 1.0293, + "step": 4911 + }, + { + "epoch": 0.33750171774082727, + "grad_norm": 1.1469614505767822, + "learning_rate": 4.189159716929154e-05, + "loss": 0.9727, + "step": 4912 + }, + { + "epoch": 0.33757042737391785, + "grad_norm": 1.2015377283096313, + "learning_rate": 4.188717613653554e-05, + "loss": 1.0257, + "step": 4913 + }, + { + "epoch": 0.33763913700700837, + "grad_norm": 1.1377078294754028, + "learning_rate": 4.1882754132270584e-05, + "loss": 0.9948, + "step": 4914 + }, + { + "epoch": 0.33770784664009895, + "grad_norm": 1.0506882667541504, + "learning_rate": 4.1878331156751084e-05, + "loss": 0.9512, + "step": 4915 + }, + { + "epoch": 0.3377765562731895, + "grad_norm": 1.1382899284362793, + "learning_rate": 4.187390721023149e-05, + "loss": 1.0191, + "step": 4916 + }, + { + "epoch": 0.33784526590628006, + "grad_norm": 1.249981164932251, + "learning_rate": 4.1869482292966294e-05, + "loss": 1.0095, + "step": 4917 + }, + { + "epoch": 0.33791397553937064, + "grad_norm": 1.060076355934143, + "learning_rate": 4.1865056405210065e-05, + "loss": 0.9815, + "step": 4918 + }, + { + "epoch": 0.33798268517246116, + "grad_norm": 1.1625521183013916, + "learning_rate": 4.186062954721743e-05, + "loss": 0.9994, + "step": 4919 + }, + { + "epoch": 0.33805139480555174, + "grad_norm": 1.1672813892364502, + "learning_rate": 4.185620171924306e-05, + "loss": 0.9969, + "step": 4920 + }, + { + "epoch": 0.3381201044386423, + "grad_norm": 1.1121249198913574, + "learning_rate": 4.185177292154168e-05, + "loss": 0.9624, + "step": 4921 + }, + { + "epoch": 0.33818881407173285, + "grad_norm": 1.0986491441726685, + "learning_rate": 4.184734315436808e-05, + "loss": 1.016, + "step": 4922 + }, + { + "epoch": 0.3382575237048234, + "grad_norm": 1.079426646232605, + "learning_rate": 4.18429124179771e-05, + "loss": 0.9678, + "step": 4923 + }, + { + "epoch": 0.33832623333791395, + "grad_norm": 1.3173456192016602, + "learning_rate": 4.183848071262364e-05, + "loss": 1.0078, + "step": 4924 + }, + { + "epoch": 0.33839494297100453, + "grad_norm": 1.1816589832305908, + "learning_rate": 4.183404803856264e-05, + "loss": 0.9923, + "step": 4925 + }, + { + "epoch": 0.3384636526040951, + "grad_norm": 1.1215671300888062, + "learning_rate": 4.1829614396049135e-05, + "loss": 1.0486, + "step": 4926 + }, + { + "epoch": 0.33853236223718564, + "grad_norm": 1.165967583656311, + "learning_rate": 4.182517978533817e-05, + "loss": 1.025, + "step": 4927 + }, + { + "epoch": 0.3386010718702762, + "grad_norm": 1.1291385889053345, + "learning_rate": 4.1820744206684867e-05, + "loss": 0.9835, + "step": 4928 + }, + { + "epoch": 0.3386697815033668, + "grad_norm": 1.072044014930725, + "learning_rate": 4.181630766034441e-05, + "loss": 0.9664, + "step": 4929 + }, + { + "epoch": 0.3387384911364573, + "grad_norm": 1.086524486541748, + "learning_rate": 4.181187014657203e-05, + "loss": 1.0081, + "step": 4930 + }, + { + "epoch": 0.3388072007695479, + "grad_norm": 1.025126338005066, + "learning_rate": 4.180743166562301e-05, + "loss": 0.9828, + "step": 4931 + }, + { + "epoch": 0.3388759104026384, + "grad_norm": 1.0880427360534668, + "learning_rate": 4.180299221775269e-05, + "loss": 1.0145, + "step": 4932 + }, + { + "epoch": 0.338944620035729, + "grad_norm": 1.1212215423583984, + "learning_rate": 4.179855180321648e-05, + "loss": 1.0348, + "step": 4933 + }, + { + "epoch": 0.3390133296688196, + "grad_norm": 1.1505630016326904, + "learning_rate": 4.1794110422269825e-05, + "loss": 1.0051, + "step": 4934 + }, + { + "epoch": 0.3390820393019101, + "grad_norm": 1.0800995826721191, + "learning_rate": 4.1789668075168235e-05, + "loss": 1.0206, + "step": 4935 + }, + { + "epoch": 0.3391507489350007, + "grad_norm": 1.1571727991104126, + "learning_rate": 4.178522476216727e-05, + "loss": 1.0139, + "step": 4936 + }, + { + "epoch": 0.33921945856809127, + "grad_norm": 1.121020793914795, + "learning_rate": 4.1780780483522575e-05, + "loss": 1.0416, + "step": 4937 + }, + { + "epoch": 0.3392881682011818, + "grad_norm": 1.1045589447021484, + "learning_rate": 4.1776335239489795e-05, + "loss": 0.9295, + "step": 4938 + }, + { + "epoch": 0.3393568778342724, + "grad_norm": 1.0894348621368408, + "learning_rate": 4.177188903032469e-05, + "loss": 1.0246, + "step": 4939 + }, + { + "epoch": 0.3394255874673629, + "grad_norm": 1.0786265134811401, + "learning_rate": 4.176744185628303e-05, + "loss": 0.9627, + "step": 4940 + }, + { + "epoch": 0.3394942971004535, + "grad_norm": 1.0491188764572144, + "learning_rate": 4.1762993717620666e-05, + "loss": 0.9807, + "step": 4941 + }, + { + "epoch": 0.33956300673354406, + "grad_norm": 1.0689561367034912, + "learning_rate": 4.175854461459349e-05, + "loss": 0.9861, + "step": 4942 + }, + { + "epoch": 0.3396317163666346, + "grad_norm": 1.0500197410583496, + "learning_rate": 4.175409454745746e-05, + "loss": 0.9656, + "step": 4943 + }, + { + "epoch": 0.33970042599972516, + "grad_norm": 1.133461833000183, + "learning_rate": 4.17496435164686e-05, + "loss": 0.9987, + "step": 4944 + }, + { + "epoch": 0.33976913563281574, + "grad_norm": 1.1308006048202515, + "learning_rate": 4.174519152188294e-05, + "loss": 1.0026, + "step": 4945 + }, + { + "epoch": 0.33983784526590627, + "grad_norm": 1.0875728130340576, + "learning_rate": 4.174073856395664e-05, + "loss": 1.0397, + "step": 4946 + }, + { + "epoch": 0.33990655489899685, + "grad_norm": 1.128118634223938, + "learning_rate": 4.173628464294585e-05, + "loss": 0.9585, + "step": 4947 + }, + { + "epoch": 0.3399752645320874, + "grad_norm": 1.2008875608444214, + "learning_rate": 4.173182975910681e-05, + "loss": 1.0291, + "step": 4948 + }, + { + "epoch": 0.34004397416517795, + "grad_norm": 1.1366002559661865, + "learning_rate": 4.1727373912695804e-05, + "loss": 0.9754, + "step": 4949 + }, + { + "epoch": 0.34011268379826853, + "grad_norm": 1.1922169923782349, + "learning_rate": 4.1722917103969174e-05, + "loss": 0.9831, + "step": 4950 + }, + { + "epoch": 0.34018139343135906, + "grad_norm": 1.0557749271392822, + "learning_rate": 4.171845933318332e-05, + "loss": 0.9541, + "step": 4951 + }, + { + "epoch": 0.34025010306444964, + "grad_norm": 1.0931445360183716, + "learning_rate": 4.17140006005947e-05, + "loss": 1.0329, + "step": 4952 + }, + { + "epoch": 0.3403188126975402, + "grad_norm": 1.1593152284622192, + "learning_rate": 4.1709540906459824e-05, + "loss": 0.9643, + "step": 4953 + }, + { + "epoch": 0.34038752233063074, + "grad_norm": 1.0515482425689697, + "learning_rate": 4.1705080251035236e-05, + "loss": 0.9964, + "step": 4954 + }, + { + "epoch": 0.3404562319637213, + "grad_norm": 1.1200385093688965, + "learning_rate": 4.170061863457756e-05, + "loss": 0.9987, + "step": 4955 + }, + { + "epoch": 0.34052494159681185, + "grad_norm": 1.019829273223877, + "learning_rate": 4.1696156057343496e-05, + "loss": 0.9885, + "step": 4956 + }, + { + "epoch": 0.3405936512299024, + "grad_norm": 1.2008854150772095, + "learning_rate": 4.169169251958974e-05, + "loss": 0.9842, + "step": 4957 + }, + { + "epoch": 0.340662360862993, + "grad_norm": 1.1665782928466797, + "learning_rate": 4.168722802157309e-05, + "loss": 0.9787, + "step": 4958 + }, + { + "epoch": 0.34073107049608353, + "grad_norm": 1.1742783784866333, + "learning_rate": 4.1682762563550396e-05, + "loss": 0.9957, + "step": 4959 + }, + { + "epoch": 0.3407997801291741, + "grad_norm": 1.0635825395584106, + "learning_rate": 4.167829614577854e-05, + "loss": 0.9986, + "step": 4960 + }, + { + "epoch": 0.3408684897622647, + "grad_norm": 1.1295632123947144, + "learning_rate": 4.167382876851448e-05, + "loss": 1.0082, + "step": 4961 + }, + { + "epoch": 0.3409371993953552, + "grad_norm": 1.1158968210220337, + "learning_rate": 4.1669360432015204e-05, + "loss": 1.0151, + "step": 4962 + }, + { + "epoch": 0.3410059090284458, + "grad_norm": 1.1011978387832642, + "learning_rate": 4.16648911365378e-05, + "loss": 0.9959, + "step": 4963 + }, + { + "epoch": 0.3410746186615363, + "grad_norm": 1.0793920755386353, + "learning_rate": 4.166042088233936e-05, + "loss": 1.0297, + "step": 4964 + }, + { + "epoch": 0.3411433282946269, + "grad_norm": 1.0724666118621826, + "learning_rate": 4.165594966967707e-05, + "loss": 0.9944, + "step": 4965 + }, + { + "epoch": 0.3412120379277175, + "grad_norm": 1.117058277130127, + "learning_rate": 4.165147749880816e-05, + "loss": 1.0382, + "step": 4966 + }, + { + "epoch": 0.341280747560808, + "grad_norm": 1.144649863243103, + "learning_rate": 4.164700436998988e-05, + "loss": 0.9705, + "step": 4967 + }, + { + "epoch": 0.3413494571938986, + "grad_norm": 1.1164155006408691, + "learning_rate": 4.164253028347961e-05, + "loss": 0.9708, + "step": 4968 + }, + { + "epoch": 0.34141816682698917, + "grad_norm": 1.174672245979309, + "learning_rate": 4.163805523953471e-05, + "loss": 1.0047, + "step": 4969 + }, + { + "epoch": 0.3414868764600797, + "grad_norm": 1.1343364715576172, + "learning_rate": 4.163357923841264e-05, + "loss": 0.981, + "step": 4970 + }, + { + "epoch": 0.34155558609317027, + "grad_norm": 1.193847417831421, + "learning_rate": 4.1629102280370904e-05, + "loss": 1.0218, + "step": 4971 + }, + { + "epoch": 0.3416242957262608, + "grad_norm": 1.0173300504684448, + "learning_rate": 4.1624624365667044e-05, + "loss": 0.9753, + "step": 4972 + }, + { + "epoch": 0.3416930053593514, + "grad_norm": 1.1121450662612915, + "learning_rate": 4.162014549455868e-05, + "loss": 0.9816, + "step": 4973 + }, + { + "epoch": 0.34176171499244196, + "grad_norm": 1.0831068754196167, + "learning_rate": 4.1615665667303486e-05, + "loss": 0.9957, + "step": 4974 + }, + { + "epoch": 0.3418304246255325, + "grad_norm": 1.1657699346542358, + "learning_rate": 4.161118488415917e-05, + "loss": 1.0702, + "step": 4975 + }, + { + "epoch": 0.34189913425862306, + "grad_norm": 1.178100824356079, + "learning_rate": 4.160670314538353e-05, + "loss": 1.0386, + "step": 4976 + }, + { + "epoch": 0.34196784389171364, + "grad_norm": 1.0711357593536377, + "learning_rate": 4.1602220451234374e-05, + "loss": 0.9668, + "step": 4977 + }, + { + "epoch": 0.34203655352480417, + "grad_norm": 1.0401140451431274, + "learning_rate": 4.15977368019696e-05, + "loss": 0.9522, + "step": 4978 + }, + { + "epoch": 0.34210526315789475, + "grad_norm": 1.1357511281967163, + "learning_rate": 4.1593252197847146e-05, + "loss": 0.9842, + "step": 4979 + }, + { + "epoch": 0.34217397279098527, + "grad_norm": 1.0811764001846313, + "learning_rate": 4.158876663912501e-05, + "loss": 0.9638, + "step": 4980 + }, + { + "epoch": 0.34224268242407585, + "grad_norm": 1.0520213842391968, + "learning_rate": 4.158428012606124e-05, + "loss": 1.0063, + "step": 4981 + }, + { + "epoch": 0.34231139205716643, + "grad_norm": 1.1088266372680664, + "learning_rate": 4.157979265891395e-05, + "loss": 1.0009, + "step": 4982 + }, + { + "epoch": 0.34238010169025695, + "grad_norm": 1.1397068500518799, + "learning_rate": 4.1575304237941305e-05, + "loss": 0.9955, + "step": 4983 + }, + { + "epoch": 0.34244881132334754, + "grad_norm": 1.0671511888504028, + "learning_rate": 4.157081486340151e-05, + "loss": 0.9848, + "step": 4984 + }, + { + "epoch": 0.3425175209564381, + "grad_norm": 1.0757451057434082, + "learning_rate": 4.156632453555284e-05, + "loss": 0.9768, + "step": 4985 + }, + { + "epoch": 0.34258623058952864, + "grad_norm": 1.1473791599273682, + "learning_rate": 4.1561833254653614e-05, + "loss": 1.0081, + "step": 4986 + }, + { + "epoch": 0.3426549402226192, + "grad_norm": 1.0758559703826904, + "learning_rate": 4.1557341020962224e-05, + "loss": 0.9986, + "step": 4987 + }, + { + "epoch": 0.34272364985570974, + "grad_norm": 1.061622977256775, + "learning_rate": 4.155284783473709e-05, + "loss": 1.0023, + "step": 4988 + }, + { + "epoch": 0.3427923594888003, + "grad_norm": 1.20256769657135, + "learning_rate": 4.154835369623673e-05, + "loss": 1.0067, + "step": 4989 + }, + { + "epoch": 0.3428610691218909, + "grad_norm": 1.0844454765319824, + "learning_rate": 4.154385860571966e-05, + "loss": 1.0159, + "step": 4990 + }, + { + "epoch": 0.34292977875498143, + "grad_norm": 1.0490301847457886, + "learning_rate": 4.153936256344451e-05, + "loss": 0.9794, + "step": 4991 + }, + { + "epoch": 0.342998488388072, + "grad_norm": 1.1244186162948608, + "learning_rate": 4.15348655696699e-05, + "loss": 0.9935, + "step": 4992 + }, + { + "epoch": 0.3430671980211626, + "grad_norm": 1.0593855381011963, + "learning_rate": 4.153036762465455e-05, + "loss": 0.9919, + "step": 4993 + }, + { + "epoch": 0.3431359076542531, + "grad_norm": 1.1091234683990479, + "learning_rate": 4.152586872865725e-05, + "loss": 0.9771, + "step": 4994 + }, + { + "epoch": 0.3432046172873437, + "grad_norm": 1.1078581809997559, + "learning_rate": 4.152136888193679e-05, + "loss": 1.0099, + "step": 4995 + }, + { + "epoch": 0.3432733269204342, + "grad_norm": 1.0891157388687134, + "learning_rate": 4.151686808475204e-05, + "loss": 1.0006, + "step": 4996 + }, + { + "epoch": 0.3433420365535248, + "grad_norm": 1.0878918170928955, + "learning_rate": 4.1512366337361953e-05, + "loss": 1.0125, + "step": 4997 + }, + { + "epoch": 0.3434107461866154, + "grad_norm": 1.1682813167572021, + "learning_rate": 4.150786364002549e-05, + "loss": 0.9649, + "step": 4998 + }, + { + "epoch": 0.3434794558197059, + "grad_norm": 1.0236252546310425, + "learning_rate": 4.15033599930017e-05, + "loss": 1.0122, + "step": 4999 + }, + { + "epoch": 0.3435481654527965, + "grad_norm": 1.1011345386505127, + "learning_rate": 4.1498855396549676e-05, + "loss": 0.9608, + "step": 5000 + }, + { + "epoch": 0.34361687508588706, + "grad_norm": 1.1376631259918213, + "learning_rate": 4.1494349850928554e-05, + "loss": 1.0127, + "step": 5001 + }, + { + "epoch": 0.3436855847189776, + "grad_norm": 1.1115657091140747, + "learning_rate": 4.148984335639755e-05, + "loss": 1.0339, + "step": 5002 + }, + { + "epoch": 0.34375429435206817, + "grad_norm": 1.0772751569747925, + "learning_rate": 4.1485335913215904e-05, + "loss": 0.9844, + "step": 5003 + }, + { + "epoch": 0.3438230039851587, + "grad_norm": 1.1074310541152954, + "learning_rate": 4.148082752164294e-05, + "loss": 1.0018, + "step": 5004 + }, + { + "epoch": 0.3438917136182493, + "grad_norm": 1.1922780275344849, + "learning_rate": 4.1476318181938026e-05, + "loss": 0.9953, + "step": 5005 + }, + { + "epoch": 0.34396042325133985, + "grad_norm": 1.1247490644454956, + "learning_rate": 4.147180789436056e-05, + "loss": 1.0177, + "step": 5006 + }, + { + "epoch": 0.3440291328844304, + "grad_norm": 1.1208957433700562, + "learning_rate": 4.146729665917003e-05, + "loss": 0.989, + "step": 5007 + }, + { + "epoch": 0.34409784251752096, + "grad_norm": 1.110831618309021, + "learning_rate": 4.146278447662597e-05, + "loss": 0.998, + "step": 5008 + }, + { + "epoch": 0.34416655215061154, + "grad_norm": 1.1110050678253174, + "learning_rate": 4.145827134698795e-05, + "loss": 0.977, + "step": 5009 + }, + { + "epoch": 0.34423526178370206, + "grad_norm": 1.0822672843933105, + "learning_rate": 4.145375727051563e-05, + "loss": 0.9742, + "step": 5010 + }, + { + "epoch": 0.34430397141679264, + "grad_norm": 1.0918858051300049, + "learning_rate": 4.144924224746867e-05, + "loss": 1.0008, + "step": 5011 + }, + { + "epoch": 0.34437268104988317, + "grad_norm": 1.146870732307434, + "learning_rate": 4.144472627810684e-05, + "loss": 1.0426, + "step": 5012 + }, + { + "epoch": 0.34444139068297375, + "grad_norm": 1.1232277154922485, + "learning_rate": 4.144020936268993e-05, + "loss": 1.0127, + "step": 5013 + }, + { + "epoch": 0.3445101003160643, + "grad_norm": 1.020625352859497, + "learning_rate": 4.14356915014778e-05, + "loss": 0.9848, + "step": 5014 + }, + { + "epoch": 0.34457880994915485, + "grad_norm": 1.0696405172348022, + "learning_rate": 4.1431172694730366e-05, + "loss": 0.9849, + "step": 5015 + }, + { + "epoch": 0.34464751958224543, + "grad_norm": 1.090226650238037, + "learning_rate": 4.142665294270758e-05, + "loss": 1.0316, + "step": 5016 + }, + { + "epoch": 0.344716229215336, + "grad_norm": 1.0753260850906372, + "learning_rate": 4.1422132245669464e-05, + "loss": 0.9871, + "step": 5017 + }, + { + "epoch": 0.34478493884842654, + "grad_norm": 1.1032053232192993, + "learning_rate": 4.1417610603876096e-05, + "loss": 0.9934, + "step": 5018 + }, + { + "epoch": 0.3448536484815171, + "grad_norm": 1.0949996709823608, + "learning_rate": 4.14130880175876e-05, + "loss": 0.9914, + "step": 5019 + }, + { + "epoch": 0.34492235811460764, + "grad_norm": 1.1271018981933594, + "learning_rate": 4.140856448706416e-05, + "loss": 1.043, + "step": 5020 + }, + { + "epoch": 0.3449910677476982, + "grad_norm": 1.0260335206985474, + "learning_rate": 4.1404040012566006e-05, + "loss": 1.0142, + "step": 5021 + }, + { + "epoch": 0.3450597773807888, + "grad_norm": 1.1100003719329834, + "learning_rate": 4.139951459435343e-05, + "loss": 0.9701, + "step": 5022 + }, + { + "epoch": 0.3451284870138793, + "grad_norm": 1.0835341215133667, + "learning_rate": 4.139498823268679e-05, + "loss": 0.9704, + "step": 5023 + }, + { + "epoch": 0.3451971966469699, + "grad_norm": 1.073480486869812, + "learning_rate": 4.139046092782647e-05, + "loss": 0.9655, + "step": 5024 + }, + { + "epoch": 0.3452659062800605, + "grad_norm": 1.0004281997680664, + "learning_rate": 4.138593268003292e-05, + "loss": 0.9613, + "step": 5025 + }, + { + "epoch": 0.345334615913151, + "grad_norm": 1.1760194301605225, + "learning_rate": 4.138140348956666e-05, + "loss": 1.0226, + "step": 5026 + }, + { + "epoch": 0.3454033255462416, + "grad_norm": 1.16511869430542, + "learning_rate": 4.137687335668824e-05, + "loss": 0.9567, + "step": 5027 + }, + { + "epoch": 0.3454720351793321, + "grad_norm": 1.0612366199493408, + "learning_rate": 4.137234228165828e-05, + "loss": 0.9778, + "step": 5028 + }, + { + "epoch": 0.3455407448124227, + "grad_norm": 1.0306007862091064, + "learning_rate": 4.136781026473746e-05, + "loss": 0.976, + "step": 5029 + }, + { + "epoch": 0.3456094544455133, + "grad_norm": 1.0441416501998901, + "learning_rate": 4.136327730618649e-05, + "loss": 0.9855, + "step": 5030 + }, + { + "epoch": 0.3456781640786038, + "grad_norm": 1.1520190238952637, + "learning_rate": 4.135874340626615e-05, + "loss": 1.0188, + "step": 5031 + }, + { + "epoch": 0.3457468737116944, + "grad_norm": 1.1633042097091675, + "learning_rate": 4.135420856523728e-05, + "loss": 1.0044, + "step": 5032 + }, + { + "epoch": 0.34581558334478496, + "grad_norm": 1.0584999322891235, + "learning_rate": 4.1349672783360774e-05, + "loss": 0.9933, + "step": 5033 + }, + { + "epoch": 0.3458842929778755, + "grad_norm": 1.0655416250228882, + "learning_rate": 4.134513606089755e-05, + "loss": 0.9918, + "step": 5034 + }, + { + "epoch": 0.34595300261096606, + "grad_norm": 1.111200213432312, + "learning_rate": 4.134059839810861e-05, + "loss": 1.0161, + "step": 5035 + }, + { + "epoch": 0.3460217122440566, + "grad_norm": 1.2279688119888306, + "learning_rate": 4.1336059795255016e-05, + "loss": 0.9584, + "step": 5036 + }, + { + "epoch": 0.34609042187714717, + "grad_norm": 1.09981107711792, + "learning_rate": 4.133152025259786e-05, + "loss": 1.0158, + "step": 5037 + }, + { + "epoch": 0.34615913151023775, + "grad_norm": 1.103541374206543, + "learning_rate": 4.132697977039831e-05, + "loss": 0.9766, + "step": 5038 + }, + { + "epoch": 0.3462278411433283, + "grad_norm": 1.042873501777649, + "learning_rate": 4.132243834891755e-05, + "loss": 0.9638, + "step": 5039 + }, + { + "epoch": 0.34629655077641885, + "grad_norm": 1.0966391563415527, + "learning_rate": 4.1317895988416874e-05, + "loss": 1.0112, + "step": 5040 + }, + { + "epoch": 0.34636526040950943, + "grad_norm": 1.1003446578979492, + "learning_rate": 4.13133526891576e-05, + "loss": 0.9714, + "step": 5041 + }, + { + "epoch": 0.34643397004259996, + "grad_norm": 1.1252511739730835, + "learning_rate": 4.1308808451401085e-05, + "loss": 1.0358, + "step": 5042 + }, + { + "epoch": 0.34650267967569054, + "grad_norm": 1.132210373878479, + "learning_rate": 4.130426327540876e-05, + "loss": 0.9589, + "step": 5043 + }, + { + "epoch": 0.34657138930878106, + "grad_norm": 1.1257137060165405, + "learning_rate": 4.129971716144212e-05, + "loss": 1.0412, + "step": 5044 + }, + { + "epoch": 0.34664009894187164, + "grad_norm": 1.043306827545166, + "learning_rate": 4.1295170109762685e-05, + "loss": 0.9413, + "step": 5045 + }, + { + "epoch": 0.3467088085749622, + "grad_norm": 1.1278913021087646, + "learning_rate": 4.129062212063204e-05, + "loss": 1.0598, + "step": 5046 + }, + { + "epoch": 0.34677751820805275, + "grad_norm": 1.0898112058639526, + "learning_rate": 4.128607319431185e-05, + "loss": 0.9914, + "step": 5047 + }, + { + "epoch": 0.34684622784114333, + "grad_norm": 1.1640783548355103, + "learning_rate": 4.12815233310638e-05, + "loss": 1.034, + "step": 5048 + }, + { + "epoch": 0.3469149374742339, + "grad_norm": 1.1598256826400757, + "learning_rate": 4.127697253114963e-05, + "loss": 1.0166, + "step": 5049 + }, + { + "epoch": 0.34698364710732443, + "grad_norm": 1.1315739154815674, + "learning_rate": 4.1272420794831157e-05, + "loss": 0.997, + "step": 5050 + }, + { + "epoch": 0.347052356740415, + "grad_norm": 1.10471773147583, + "learning_rate": 4.1267868122370244e-05, + "loss": 1.0298, + "step": 5051 + }, + { + "epoch": 0.34712106637350554, + "grad_norm": 1.0827481746673584, + "learning_rate": 4.1263314514028795e-05, + "loss": 1.0189, + "step": 5052 + }, + { + "epoch": 0.3471897760065961, + "grad_norm": 1.0999703407287598, + "learning_rate": 4.125875997006877e-05, + "loss": 0.993, + "step": 5053 + }, + { + "epoch": 0.3472584856396867, + "grad_norm": 1.117574691772461, + "learning_rate": 4.1254204490752214e-05, + "loss": 1.011, + "step": 5054 + }, + { + "epoch": 0.3473271952727772, + "grad_norm": 1.1566798686981201, + "learning_rate": 4.1249648076341165e-05, + "loss": 1.0238, + "step": 5055 + }, + { + "epoch": 0.3473959049058678, + "grad_norm": 1.1496683359146118, + "learning_rate": 4.1245090727097787e-05, + "loss": 1.0186, + "step": 5056 + }, + { + "epoch": 0.3474646145389584, + "grad_norm": 1.1128636598587036, + "learning_rate": 4.124053244328423e-05, + "loss": 1.0181, + "step": 5057 + }, + { + "epoch": 0.3475333241720489, + "grad_norm": 1.1086119413375854, + "learning_rate": 4.123597322516276e-05, + "loss": 1.0, + "step": 5058 + }, + { + "epoch": 0.3476020338051395, + "grad_norm": 1.1416085958480835, + "learning_rate": 4.123141307299564e-05, + "loss": 1.0072, + "step": 5059 + }, + { + "epoch": 0.34767074343823, + "grad_norm": 1.1197890043258667, + "learning_rate": 4.1226851987045226e-05, + "loss": 1.013, + "step": 5060 + }, + { + "epoch": 0.3477394530713206, + "grad_norm": 1.0723230838775635, + "learning_rate": 4.1222289967573904e-05, + "loss": 0.9913, + "step": 5061 + }, + { + "epoch": 0.3478081627044112, + "grad_norm": 1.1853208541870117, + "learning_rate": 4.1217727014844134e-05, + "loss": 1.0172, + "step": 5062 + }, + { + "epoch": 0.3478768723375017, + "grad_norm": 1.1590245962142944, + "learning_rate": 4.121316312911843e-05, + "loss": 1.012, + "step": 5063 + }, + { + "epoch": 0.3479455819705923, + "grad_norm": 1.192644476890564, + "learning_rate": 4.120859831065932e-05, + "loss": 0.9742, + "step": 5064 + }, + { + "epoch": 0.34801429160368286, + "grad_norm": 1.0964192152023315, + "learning_rate": 4.1204032559729436e-05, + "loss": 1.0043, + "step": 5065 + }, + { + "epoch": 0.3480830012367734, + "grad_norm": 1.1114280223846436, + "learning_rate": 4.1199465876591444e-05, + "loss": 0.9808, + "step": 5066 + }, + { + "epoch": 0.34815171086986396, + "grad_norm": 1.0695126056671143, + "learning_rate": 4.1194898261508055e-05, + "loss": 0.9893, + "step": 5067 + }, + { + "epoch": 0.34822042050295454, + "grad_norm": 1.1031042337417603, + "learning_rate": 4.119032971474205e-05, + "loss": 0.9875, + "step": 5068 + }, + { + "epoch": 0.34828913013604507, + "grad_norm": 1.0991020202636719, + "learning_rate": 4.1185760236556236e-05, + "loss": 0.9962, + "step": 5069 + }, + { + "epoch": 0.34835783976913565, + "grad_norm": 1.099916934967041, + "learning_rate": 4.118118982721351e-05, + "loss": 0.9799, + "step": 5070 + }, + { + "epoch": 0.34842654940222617, + "grad_norm": 1.1014224290847778, + "learning_rate": 4.11766184869768e-05, + "loss": 0.9363, + "step": 5071 + }, + { + "epoch": 0.34849525903531675, + "grad_norm": 1.0733423233032227, + "learning_rate": 4.11720462161091e-05, + "loss": 0.9894, + "step": 5072 + }, + { + "epoch": 0.34856396866840733, + "grad_norm": 1.1597157716751099, + "learning_rate": 4.116747301487343e-05, + "loss": 1.0427, + "step": 5073 + }, + { + "epoch": 0.34863267830149786, + "grad_norm": 1.14413583278656, + "learning_rate": 4.11628988835329e-05, + "loss": 1.0537, + "step": 5074 + }, + { + "epoch": 0.34870138793458844, + "grad_norm": 1.1336241960525513, + "learning_rate": 4.115832382235065e-05, + "loss": 1.0331, + "step": 5075 + }, + { + "epoch": 0.348770097567679, + "grad_norm": 1.079771637916565, + "learning_rate": 4.1153747831589884e-05, + "loss": 1.0235, + "step": 5076 + }, + { + "epoch": 0.34883880720076954, + "grad_norm": 1.0840595960617065, + "learning_rate": 4.114917091151386e-05, + "loss": 1.0037, + "step": 5077 + }, + { + "epoch": 0.3489075168338601, + "grad_norm": 1.2221276760101318, + "learning_rate": 4.1144593062385874e-05, + "loss": 1.0117, + "step": 5078 + }, + { + "epoch": 0.34897622646695065, + "grad_norm": 1.0744929313659668, + "learning_rate": 4.1140014284469294e-05, + "loss": 1.011, + "step": 5079 + }, + { + "epoch": 0.3490449361000412, + "grad_norm": 1.0683865547180176, + "learning_rate": 4.1135434578027534e-05, + "loss": 1.0006, + "step": 5080 + }, + { + "epoch": 0.3491136457331318, + "grad_norm": 1.0914896726608276, + "learning_rate": 4.1130853943324053e-05, + "loss": 0.9958, + "step": 5081 + }, + { + "epoch": 0.34918235536622233, + "grad_norm": 1.1052489280700684, + "learning_rate": 4.112627238062239e-05, + "loss": 0.9743, + "step": 5082 + }, + { + "epoch": 0.3492510649993129, + "grad_norm": 1.107586145401001, + "learning_rate": 4.112168989018611e-05, + "loss": 0.9852, + "step": 5083 + }, + { + "epoch": 0.3493197746324035, + "grad_norm": 1.0164456367492676, + "learning_rate": 4.1117106472278835e-05, + "loss": 0.9482, + "step": 5084 + }, + { + "epoch": 0.349388484265494, + "grad_norm": 1.0787712335586548, + "learning_rate": 4.1112522127164253e-05, + "loss": 0.9877, + "step": 5085 + }, + { + "epoch": 0.3494571938985846, + "grad_norm": 1.0664088726043701, + "learning_rate": 4.1107936855106105e-05, + "loss": 0.99, + "step": 5086 + }, + { + "epoch": 0.3495259035316751, + "grad_norm": 1.0927847623825073, + "learning_rate": 4.110335065636817e-05, + "loss": 0.9828, + "step": 5087 + }, + { + "epoch": 0.3495946131647657, + "grad_norm": 1.0623801946640015, + "learning_rate": 4.109876353121428e-05, + "loss": 0.9813, + "step": 5088 + }, + { + "epoch": 0.3496633227978563, + "grad_norm": 1.0851101875305176, + "learning_rate": 4.1094175479908345e-05, + "loss": 0.9678, + "step": 5089 + }, + { + "epoch": 0.3497320324309468, + "grad_norm": 1.1195943355560303, + "learning_rate": 4.108958650271431e-05, + "loss": 1.0091, + "step": 5090 + }, + { + "epoch": 0.3498007420640374, + "grad_norm": 1.1060795783996582, + "learning_rate": 4.108499659989618e-05, + "loss": 0.9969, + "step": 5091 + }, + { + "epoch": 0.34986945169712796, + "grad_norm": 1.1154677867889404, + "learning_rate": 4.108040577171799e-05, + "loss": 0.9822, + "step": 5092 + }, + { + "epoch": 0.3499381613302185, + "grad_norm": 1.1213204860687256, + "learning_rate": 4.107581401844386e-05, + "loss": 0.9876, + "step": 5093 + }, + { + "epoch": 0.35000687096330907, + "grad_norm": 1.248352289199829, + "learning_rate": 4.1071221340337965e-05, + "loss": 0.9677, + "step": 5094 + }, + { + "epoch": 0.3500755805963996, + "grad_norm": 1.136888861656189, + "learning_rate": 4.10666277376645e-05, + "loss": 1.0062, + "step": 5095 + }, + { + "epoch": 0.3501442902294902, + "grad_norm": 1.0601065158843994, + "learning_rate": 4.1062033210687734e-05, + "loss": 1.0071, + "step": 5096 + }, + { + "epoch": 0.35021299986258075, + "grad_norm": 1.0254418849945068, + "learning_rate": 4.105743775967199e-05, + "loss": 0.9607, + "step": 5097 + }, + { + "epoch": 0.3502817094956713, + "grad_norm": 1.1244062185287476, + "learning_rate": 4.105284138488165e-05, + "loss": 0.9856, + "step": 5098 + }, + { + "epoch": 0.35035041912876186, + "grad_norm": 1.108063817024231, + "learning_rate": 4.1048244086581124e-05, + "loss": 0.9654, + "step": 5099 + }, + { + "epoch": 0.35041912876185244, + "grad_norm": 1.1432061195373535, + "learning_rate": 4.10436458650349e-05, + "loss": 0.9802, + "step": 5100 + }, + { + "epoch": 0.35048783839494296, + "grad_norm": 1.1080119609832764, + "learning_rate": 4.1039046720507527e-05, + "loss": 0.9803, + "step": 5101 + }, + { + "epoch": 0.35055654802803354, + "grad_norm": 1.175736665725708, + "learning_rate": 4.1034446653263556e-05, + "loss": 0.9916, + "step": 5102 + }, + { + "epoch": 0.35062525766112407, + "grad_norm": 1.0354318618774414, + "learning_rate": 4.102984566356766e-05, + "loss": 0.9831, + "step": 5103 + }, + { + "epoch": 0.35069396729421465, + "grad_norm": 1.0837827920913696, + "learning_rate": 4.1025243751684514e-05, + "loss": 1.0231, + "step": 5104 + }, + { + "epoch": 0.35076267692730523, + "grad_norm": 1.093741774559021, + "learning_rate": 4.102064091787886e-05, + "loss": 0.9667, + "step": 5105 + }, + { + "epoch": 0.35083138656039575, + "grad_norm": 1.1360137462615967, + "learning_rate": 4.101603716241551e-05, + "loss": 0.9798, + "step": 5106 + }, + { + "epoch": 0.35090009619348633, + "grad_norm": 1.084713101387024, + "learning_rate": 4.1011432485559305e-05, + "loss": 0.9909, + "step": 5107 + }, + { + "epoch": 0.3509688058265769, + "grad_norm": 1.0947210788726807, + "learning_rate": 4.100682688757514e-05, + "loss": 0.9719, + "step": 5108 + }, + { + "epoch": 0.35103751545966744, + "grad_norm": 1.204687476158142, + "learning_rate": 4.1002220368728004e-05, + "loss": 0.9912, + "step": 5109 + }, + { + "epoch": 0.351106225092758, + "grad_norm": 1.0677053928375244, + "learning_rate": 4.0997612929282884e-05, + "loss": 1.0173, + "step": 5110 + }, + { + "epoch": 0.35117493472584854, + "grad_norm": 1.050344705581665, + "learning_rate": 4.0993004569504835e-05, + "loss": 0.9694, + "step": 5111 + }, + { + "epoch": 0.3512436443589391, + "grad_norm": 1.1228307485580444, + "learning_rate": 4.0988395289658985e-05, + "loss": 0.9806, + "step": 5112 + }, + { + "epoch": 0.3513123539920297, + "grad_norm": 1.1306467056274414, + "learning_rate": 4.098378509001051e-05, + "loss": 1.0002, + "step": 5113 + }, + { + "epoch": 0.3513810636251202, + "grad_norm": 1.089831829071045, + "learning_rate": 4.0979173970824626e-05, + "loss": 0.9742, + "step": 5114 + }, + { + "epoch": 0.3514497732582108, + "grad_norm": 1.0890663862228394, + "learning_rate": 4.097456193236661e-05, + "loss": 1.0237, + "step": 5115 + }, + { + "epoch": 0.3515184828913014, + "grad_norm": 1.1265004873275757, + "learning_rate": 4.096994897490177e-05, + "loss": 0.9912, + "step": 5116 + }, + { + "epoch": 0.3515871925243919, + "grad_norm": 1.0150630474090576, + "learning_rate": 4.0965335098695515e-05, + "loss": 0.9809, + "step": 5117 + }, + { + "epoch": 0.3516559021574825, + "grad_norm": 1.160783290863037, + "learning_rate": 4.096072030401326e-05, + "loss": 0.9834, + "step": 5118 + }, + { + "epoch": 0.351724611790573, + "grad_norm": 1.1744464635849, + "learning_rate": 4.0956104591120506e-05, + "loss": 0.95, + "step": 5119 + }, + { + "epoch": 0.3517933214236636, + "grad_norm": 1.145297884941101, + "learning_rate": 4.095148796028278e-05, + "loss": 1.0207, + "step": 5120 + }, + { + "epoch": 0.3518620310567542, + "grad_norm": 1.1139271259307861, + "learning_rate": 4.094687041176567e-05, + "loss": 1.0281, + "step": 5121 + }, + { + "epoch": 0.3519307406898447, + "grad_norm": 1.131093144416809, + "learning_rate": 4.094225194583484e-05, + "loss": 1.0082, + "step": 5122 + }, + { + "epoch": 0.3519994503229353, + "grad_norm": 1.0418843030929565, + "learning_rate": 4.0937632562755964e-05, + "loss": 0.9802, + "step": 5123 + }, + { + "epoch": 0.35206815995602586, + "grad_norm": 1.0315998792648315, + "learning_rate": 4.093301226279482e-05, + "loss": 0.954, + "step": 5124 + }, + { + "epoch": 0.3521368695891164, + "grad_norm": 1.0428663492202759, + "learning_rate": 4.092839104621718e-05, + "loss": 0.9496, + "step": 5125 + }, + { + "epoch": 0.35220557922220697, + "grad_norm": 1.1105706691741943, + "learning_rate": 4.092376891328893e-05, + "loss": 1.0028, + "step": 5126 + }, + { + "epoch": 0.3522742888552975, + "grad_norm": 1.1527431011199951, + "learning_rate": 4.091914586427595e-05, + "loss": 0.9704, + "step": 5127 + }, + { + "epoch": 0.35234299848838807, + "grad_norm": 1.177968144416809, + "learning_rate": 4.091452189944422e-05, + "loss": 0.9916, + "step": 5128 + }, + { + "epoch": 0.35241170812147865, + "grad_norm": 1.0759522914886475, + "learning_rate": 4.0909897019059764e-05, + "loss": 0.9926, + "step": 5129 + }, + { + "epoch": 0.3524804177545692, + "grad_norm": 1.0959984064102173, + "learning_rate": 4.0905271223388606e-05, + "loss": 0.9943, + "step": 5130 + }, + { + "epoch": 0.35254912738765976, + "grad_norm": 1.093056082725525, + "learning_rate": 4.090064451269691e-05, + "loss": 0.9706, + "step": 5131 + }, + { + "epoch": 0.35261783702075034, + "grad_norm": 1.0593098402023315, + "learning_rate": 4.089601688725083e-05, + "loss": 0.9732, + "step": 5132 + }, + { + "epoch": 0.35268654665384086, + "grad_norm": 1.163281798362732, + "learning_rate": 4.089138834731658e-05, + "loss": 0.9712, + "step": 5133 + }, + { + "epoch": 0.35275525628693144, + "grad_norm": 1.0508283376693726, + "learning_rate": 4.088675889316046e-05, + "loss": 0.9833, + "step": 5134 + }, + { + "epoch": 0.35282396592002196, + "grad_norm": 1.1255351305007935, + "learning_rate": 4.088212852504878e-05, + "loss": 0.9909, + "step": 5135 + }, + { + "epoch": 0.35289267555311254, + "grad_norm": 1.0701682567596436, + "learning_rate": 4.087749724324793e-05, + "loss": 1.0239, + "step": 5136 + }, + { + "epoch": 0.3529613851862031, + "grad_norm": 1.2047299146652222, + "learning_rate": 4.087286504802435e-05, + "loss": 0.9871, + "step": 5137 + }, + { + "epoch": 0.35303009481929365, + "grad_norm": 1.073714256286621, + "learning_rate": 4.086823193964452e-05, + "loss": 1.0222, + "step": 5138 + }, + { + "epoch": 0.35309880445238423, + "grad_norm": 1.12971830368042, + "learning_rate": 4.086359791837497e-05, + "loss": 1.0141, + "step": 5139 + }, + { + "epoch": 0.3531675140854748, + "grad_norm": 1.0583996772766113, + "learning_rate": 4.0858962984482324e-05, + "loss": 0.971, + "step": 5140 + }, + { + "epoch": 0.35323622371856533, + "grad_norm": 1.0391980409622192, + "learning_rate": 4.0854327138233195e-05, + "loss": 0.9789, + "step": 5141 + }, + { + "epoch": 0.3533049333516559, + "grad_norm": 1.113708257675171, + "learning_rate": 4.08496903798943e-05, + "loss": 1.0078, + "step": 5142 + }, + { + "epoch": 0.35337364298474644, + "grad_norm": 1.1047321557998657, + "learning_rate": 4.084505270973238e-05, + "loss": 1.001, + "step": 5143 + }, + { + "epoch": 0.353442352617837, + "grad_norm": 1.1650997400283813, + "learning_rate": 4.084041412801423e-05, + "loss": 1.0128, + "step": 5144 + }, + { + "epoch": 0.3535110622509276, + "grad_norm": 1.1425820589065552, + "learning_rate": 4.0835774635006725e-05, + "loss": 1.0135, + "step": 5145 + }, + { + "epoch": 0.3535797718840181, + "grad_norm": 1.0713392496109009, + "learning_rate": 4.083113423097675e-05, + "loss": 0.9681, + "step": 5146 + }, + { + "epoch": 0.3536484815171087, + "grad_norm": 1.1126039028167725, + "learning_rate": 4.082649291619128e-05, + "loss": 1.0035, + "step": 5147 + }, + { + "epoch": 0.3537171911501993, + "grad_norm": 1.0840659141540527, + "learning_rate": 4.082185069091733e-05, + "loss": 0.999, + "step": 5148 + }, + { + "epoch": 0.3537859007832898, + "grad_norm": 1.089412808418274, + "learning_rate": 4.0817207555421945e-05, + "loss": 0.9966, + "step": 5149 + }, + { + "epoch": 0.3538546104163804, + "grad_norm": 1.1179373264312744, + "learning_rate": 4.081256350997226e-05, + "loss": 0.9866, + "step": 5150 + }, + { + "epoch": 0.3539233200494709, + "grad_norm": 1.123998761177063, + "learning_rate": 4.0807918554835436e-05, + "loss": 0.9703, + "step": 5151 + }, + { + "epoch": 0.3539920296825615, + "grad_norm": 1.0560760498046875, + "learning_rate": 4.0803272690278705e-05, + "loss": 1.0048, + "step": 5152 + }, + { + "epoch": 0.3540607393156521, + "grad_norm": 1.122294545173645, + "learning_rate": 4.079862591656932e-05, + "loss": 0.9706, + "step": 5153 + }, + { + "epoch": 0.3541294489487426, + "grad_norm": 1.1424895524978638, + "learning_rate": 4.079397823397462e-05, + "loss": 1.0032, + "step": 5154 + }, + { + "epoch": 0.3541981585818332, + "grad_norm": 1.073663353919983, + "learning_rate": 4.0789329642761986e-05, + "loss": 0.9374, + "step": 5155 + }, + { + "epoch": 0.35426686821492376, + "grad_norm": 1.1384689807891846, + "learning_rate": 4.0784680143198836e-05, + "loss": 1.0378, + "step": 5156 + }, + { + "epoch": 0.3543355778480143, + "grad_norm": 1.1172828674316406, + "learning_rate": 4.078002973555268e-05, + "loss": 0.9959, + "step": 5157 + }, + { + "epoch": 0.35440428748110486, + "grad_norm": 1.1628412008285522, + "learning_rate": 4.077537842009102e-05, + "loss": 1.0179, + "step": 5158 + }, + { + "epoch": 0.3544729971141954, + "grad_norm": 1.1164822578430176, + "learning_rate": 4.0770726197081463e-05, + "loss": 1.0481, + "step": 5159 + }, + { + "epoch": 0.35454170674728597, + "grad_norm": 0.9994493126869202, + "learning_rate": 4.076607306679165e-05, + "loss": 0.9735, + "step": 5160 + }, + { + "epoch": 0.35461041638037655, + "grad_norm": 1.0926886796951294, + "learning_rate": 4.076141902948925e-05, + "loss": 1.0352, + "step": 5161 + }, + { + "epoch": 0.35467912601346707, + "grad_norm": 1.059385061264038, + "learning_rate": 4.0756764085442034e-05, + "loss": 1.0222, + "step": 5162 + }, + { + "epoch": 0.35474783564655765, + "grad_norm": 1.0932021141052246, + "learning_rate": 4.07521082349178e-05, + "loss": 1.0162, + "step": 5163 + }, + { + "epoch": 0.35481654527964823, + "grad_norm": 1.1385524272918701, + "learning_rate": 4.074745147818436e-05, + "loss": 0.9807, + "step": 5164 + }, + { + "epoch": 0.35488525491273876, + "grad_norm": 1.1767692565917969, + "learning_rate": 4.074279381550965e-05, + "loss": 0.9953, + "step": 5165 + }, + { + "epoch": 0.35495396454582934, + "grad_norm": 1.1445776224136353, + "learning_rate": 4.073813524716161e-05, + "loss": 1.0014, + "step": 5166 + }, + { + "epoch": 0.35502267417891986, + "grad_norm": 1.1241748332977295, + "learning_rate": 4.0733475773408244e-05, + "loss": 1.0167, + "step": 5167 + }, + { + "epoch": 0.35509138381201044, + "grad_norm": 1.0650845766067505, + "learning_rate": 4.072881539451762e-05, + "loss": 0.9806, + "step": 5168 + }, + { + "epoch": 0.355160093445101, + "grad_norm": 1.0276296138763428, + "learning_rate": 4.0724154110757825e-05, + "loss": 0.9763, + "step": 5169 + }, + { + "epoch": 0.35522880307819155, + "grad_norm": 1.0939847230911255, + "learning_rate": 4.071949192239704e-05, + "loss": 1.0093, + "step": 5170 + }, + { + "epoch": 0.3552975127112821, + "grad_norm": 1.0533984899520874, + "learning_rate": 4.0714828829703455e-05, + "loss": 0.9513, + "step": 5171 + }, + { + "epoch": 0.3553662223443727, + "grad_norm": 1.09126615524292, + "learning_rate": 4.071016483294536e-05, + "loss": 1.0554, + "step": 5172 + }, + { + "epoch": 0.35543493197746323, + "grad_norm": 1.1489840745925903, + "learning_rate": 4.070549993239106e-05, + "loss": 0.9771, + "step": 5173 + }, + { + "epoch": 0.3555036416105538, + "grad_norm": 1.1862045526504517, + "learning_rate": 4.070083412830893e-05, + "loss": 1.0105, + "step": 5174 + }, + { + "epoch": 0.35557235124364434, + "grad_norm": 1.109049677848816, + "learning_rate": 4.069616742096738e-05, + "loss": 0.9813, + "step": 5175 + }, + { + "epoch": 0.3556410608767349, + "grad_norm": 1.0487141609191895, + "learning_rate": 4.069149981063488e-05, + "loss": 0.9833, + "step": 5176 + }, + { + "epoch": 0.3557097705098255, + "grad_norm": 1.1144639253616333, + "learning_rate": 4.068683129757998e-05, + "loss": 0.9895, + "step": 5177 + }, + { + "epoch": 0.355778480142916, + "grad_norm": 1.1216516494750977, + "learning_rate": 4.068216188207123e-05, + "loss": 0.9585, + "step": 5178 + }, + { + "epoch": 0.3558471897760066, + "grad_norm": 1.0937154293060303, + "learning_rate": 4.0677491564377274e-05, + "loss": 1.0117, + "step": 5179 + }, + { + "epoch": 0.3559158994090972, + "grad_norm": 1.0477290153503418, + "learning_rate": 4.067282034476678e-05, + "loss": 0.9855, + "step": 5180 + }, + { + "epoch": 0.3559846090421877, + "grad_norm": 1.155320644378662, + "learning_rate": 4.06681482235085e-05, + "loss": 0.9927, + "step": 5181 + }, + { + "epoch": 0.3560533186752783, + "grad_norm": 1.0603339672088623, + "learning_rate": 4.066347520087119e-05, + "loss": 0.9673, + "step": 5182 + }, + { + "epoch": 0.3561220283083688, + "grad_norm": 1.0458152294158936, + "learning_rate": 4.0658801277123716e-05, + "loss": 0.9407, + "step": 5183 + }, + { + "epoch": 0.3561907379414594, + "grad_norm": 1.1021931171417236, + "learning_rate": 4.065412645253495e-05, + "loss": 1.0182, + "step": 5184 + }, + { + "epoch": 0.35625944757454997, + "grad_norm": 1.201931118965149, + "learning_rate": 4.064945072737383e-05, + "loss": 1.0024, + "step": 5185 + }, + { + "epoch": 0.3563281572076405, + "grad_norm": 1.1297192573547363, + "learning_rate": 4.0644774101909356e-05, + "loss": 0.978, + "step": 5186 + }, + { + "epoch": 0.3563968668407311, + "grad_norm": 1.2335238456726074, + "learning_rate": 4.0640096576410555e-05, + "loss": 0.9977, + "step": 5187 + }, + { + "epoch": 0.35646557647382165, + "grad_norm": 1.1013797521591187, + "learning_rate": 4.0635418151146555e-05, + "loss": 1.0247, + "step": 5188 + }, + { + "epoch": 0.3565342861069122, + "grad_norm": 1.0726635456085205, + "learning_rate": 4.063073882638646e-05, + "loss": 0.9468, + "step": 5189 + }, + { + "epoch": 0.35660299574000276, + "grad_norm": 1.052034854888916, + "learning_rate": 4.062605860239951e-05, + "loss": 0.9766, + "step": 5190 + }, + { + "epoch": 0.3566717053730933, + "grad_norm": 1.0406616926193237, + "learning_rate": 4.062137747945492e-05, + "loss": 0.9706, + "step": 5191 + }, + { + "epoch": 0.35674041500618386, + "grad_norm": 1.1812790632247925, + "learning_rate": 4.061669545782202e-05, + "loss": 0.9942, + "step": 5192 + }, + { + "epoch": 0.35680912463927444, + "grad_norm": 1.1191446781158447, + "learning_rate": 4.061201253777015e-05, + "loss": 1.0269, + "step": 5193 + }, + { + "epoch": 0.35687783427236497, + "grad_norm": 1.0577480792999268, + "learning_rate": 4.0607328719568715e-05, + "loss": 1.0196, + "step": 5194 + }, + { + "epoch": 0.35694654390545555, + "grad_norm": 1.1477493047714233, + "learning_rate": 4.060264400348718e-05, + "loss": 1.0287, + "step": 5195 + }, + { + "epoch": 0.35701525353854613, + "grad_norm": 1.1139540672302246, + "learning_rate": 4.059795838979504e-05, + "loss": 0.9984, + "step": 5196 + }, + { + "epoch": 0.35708396317163665, + "grad_norm": 1.1041510105133057, + "learning_rate": 4.059327187876188e-05, + "loss": 1.0029, + "step": 5197 + }, + { + "epoch": 0.35715267280472723, + "grad_norm": 1.0548995733261108, + "learning_rate": 4.058858447065729e-05, + "loss": 0.9753, + "step": 5198 + }, + { + "epoch": 0.35722138243781776, + "grad_norm": 1.1524467468261719, + "learning_rate": 4.058389616575093e-05, + "loss": 1.0456, + "step": 5199 + }, + { + "epoch": 0.35729009207090834, + "grad_norm": 1.105651617050171, + "learning_rate": 4.057920696431253e-05, + "loss": 0.9588, + "step": 5200 + }, + { + "epoch": 0.3573588017039989, + "grad_norm": 1.0739067792892456, + "learning_rate": 4.0574516866611856e-05, + "loss": 0.9511, + "step": 5201 + }, + { + "epoch": 0.35742751133708944, + "grad_norm": 1.1421278715133667, + "learning_rate": 4.056982587291872e-05, + "loss": 0.9828, + "step": 5202 + }, + { + "epoch": 0.35749622097018, + "grad_norm": 1.1155128479003906, + "learning_rate": 4.0565133983503e-05, + "loss": 0.9665, + "step": 5203 + }, + { + "epoch": 0.3575649306032706, + "grad_norm": 1.0838340520858765, + "learning_rate": 4.056044119863461e-05, + "loss": 0.9629, + "step": 5204 + }, + { + "epoch": 0.35763364023636113, + "grad_norm": 1.0393528938293457, + "learning_rate": 4.055574751858352e-05, + "loss": 0.9437, + "step": 5205 + }, + { + "epoch": 0.3577023498694517, + "grad_norm": 1.0268888473510742, + "learning_rate": 4.055105294361977e-05, + "loss": 1.0193, + "step": 5206 + }, + { + "epoch": 0.35777105950254223, + "grad_norm": 1.079647421836853, + "learning_rate": 4.054635747401342e-05, + "loss": 0.9471, + "step": 5207 + }, + { + "epoch": 0.3578397691356328, + "grad_norm": 1.0607900619506836, + "learning_rate": 4.05416611100346e-05, + "loss": 0.994, + "step": 5208 + }, + { + "epoch": 0.3579084787687234, + "grad_norm": 1.0501071214675903, + "learning_rate": 4.0536963851953504e-05, + "loss": 0.974, + "step": 5209 + }, + { + "epoch": 0.3579771884018139, + "grad_norm": 1.0804823637008667, + "learning_rate": 4.0532265700040336e-05, + "loss": 0.9828, + "step": 5210 + }, + { + "epoch": 0.3580458980349045, + "grad_norm": 1.1525408029556274, + "learning_rate": 4.0527566654565406e-05, + "loss": 0.9785, + "step": 5211 + }, + { + "epoch": 0.3581146076679951, + "grad_norm": 1.1590752601623535, + "learning_rate": 4.052286671579903e-05, + "loss": 0.9566, + "step": 5212 + }, + { + "epoch": 0.3581833173010856, + "grad_norm": 1.1208384037017822, + "learning_rate": 4.051816588401159e-05, + "loss": 1.0329, + "step": 5213 + }, + { + "epoch": 0.3582520269341762, + "grad_norm": 1.0811676979064941, + "learning_rate": 4.051346415947354e-05, + "loss": 0.9605, + "step": 5214 + }, + { + "epoch": 0.3583207365672667, + "grad_norm": 1.0366268157958984, + "learning_rate": 4.050876154245535e-05, + "loss": 0.9559, + "step": 5215 + }, + { + "epoch": 0.3583894462003573, + "grad_norm": 1.1405938863754272, + "learning_rate": 4.050405803322757e-05, + "loss": 1.0156, + "step": 5216 + }, + { + "epoch": 0.35845815583344787, + "grad_norm": 1.1146671772003174, + "learning_rate": 4.0499353632060775e-05, + "loss": 0.9409, + "step": 5217 + }, + { + "epoch": 0.3585268654665384, + "grad_norm": 1.1061322689056396, + "learning_rate": 4.049464833922562e-05, + "loss": 1.0221, + "step": 5218 + }, + { + "epoch": 0.35859557509962897, + "grad_norm": 1.1417900323867798, + "learning_rate": 4.04899421549928e-05, + "loss": 0.9651, + "step": 5219 + }, + { + "epoch": 0.35866428473271955, + "grad_norm": 1.1340961456298828, + "learning_rate": 4.048523507963304e-05, + "loss": 0.9962, + "step": 5220 + }, + { + "epoch": 0.3587329943658101, + "grad_norm": 1.1168649196624756, + "learning_rate": 4.048052711341716e-05, + "loss": 1.0108, + "step": 5221 + }, + { + "epoch": 0.35880170399890066, + "grad_norm": 1.1344764232635498, + "learning_rate": 4.047581825661599e-05, + "loss": 0.9881, + "step": 5222 + }, + { + "epoch": 0.3588704136319912, + "grad_norm": 1.013463020324707, + "learning_rate": 4.047110850950043e-05, + "loss": 0.9828, + "step": 5223 + }, + { + "epoch": 0.35893912326508176, + "grad_norm": 1.104137659072876, + "learning_rate": 4.046639787234143e-05, + "loss": 0.9949, + "step": 5224 + }, + { + "epoch": 0.35900783289817234, + "grad_norm": 1.1156740188598633, + "learning_rate": 4.046168634540999e-05, + "loss": 0.9639, + "step": 5225 + }, + { + "epoch": 0.35907654253126287, + "grad_norm": 1.085581660270691, + "learning_rate": 4.045697392897717e-05, + "loss": 0.9925, + "step": 5226 + }, + { + "epoch": 0.35914525216435345, + "grad_norm": 1.1051347255706787, + "learning_rate": 4.045226062331405e-05, + "loss": 1.0186, + "step": 5227 + }, + { + "epoch": 0.359213961797444, + "grad_norm": 1.042074203491211, + "learning_rate": 4.0447546428691805e-05, + "loss": 0.972, + "step": 5228 + }, + { + "epoch": 0.35928267143053455, + "grad_norm": 1.1127150058746338, + "learning_rate": 4.0442831345381645e-05, + "loss": 0.9578, + "step": 5229 + }, + { + "epoch": 0.35935138106362513, + "grad_norm": 1.0994895696640015, + "learning_rate": 4.04381153736548e-05, + "loss": 0.967, + "step": 5230 + }, + { + "epoch": 0.35942009069671566, + "grad_norm": 1.1338666677474976, + "learning_rate": 4.043339851378259e-05, + "loss": 0.9957, + "step": 5231 + }, + { + "epoch": 0.35948880032980624, + "grad_norm": 1.0272928476333618, + "learning_rate": 4.0428680766036384e-05, + "loss": 0.9826, + "step": 5232 + }, + { + "epoch": 0.3595575099628968, + "grad_norm": 1.1071184873580933, + "learning_rate": 4.0423962130687576e-05, + "loss": 1.0088, + "step": 5233 + }, + { + "epoch": 0.35962621959598734, + "grad_norm": 1.0673938989639282, + "learning_rate": 4.041924260800763e-05, + "loss": 0.9448, + "step": 5234 + }, + { + "epoch": 0.3596949292290779, + "grad_norm": 1.1007784605026245, + "learning_rate": 4.0414522198268056e-05, + "loss": 0.934, + "step": 5235 + }, + { + "epoch": 0.3597636388621685, + "grad_norm": 1.0739024877548218, + "learning_rate": 4.040980090174043e-05, + "loss": 0.9495, + "step": 5236 + }, + { + "epoch": 0.359832348495259, + "grad_norm": 1.0941917896270752, + "learning_rate": 4.040507871869633e-05, + "loss": 0.9943, + "step": 5237 + }, + { + "epoch": 0.3599010581283496, + "grad_norm": 1.0761632919311523, + "learning_rate": 4.040035564940746e-05, + "loss": 0.9702, + "step": 5238 + }, + { + "epoch": 0.35996976776144013, + "grad_norm": 1.0442396402359009, + "learning_rate": 4.039563169414553e-05, + "loss": 1.0001, + "step": 5239 + }, + { + "epoch": 0.3600384773945307, + "grad_norm": 1.131272792816162, + "learning_rate": 4.0390906853182274e-05, + "loss": 0.9926, + "step": 5240 + }, + { + "epoch": 0.3601071870276213, + "grad_norm": 1.049099087715149, + "learning_rate": 4.038618112678955e-05, + "loss": 0.9684, + "step": 5241 + }, + { + "epoch": 0.3601758966607118, + "grad_norm": 1.1140066385269165, + "learning_rate": 4.038145451523919e-05, + "loss": 0.992, + "step": 5242 + }, + { + "epoch": 0.3602446062938024, + "grad_norm": 1.1345529556274414, + "learning_rate": 4.0376727018803145e-05, + "loss": 0.9862, + "step": 5243 + }, + { + "epoch": 0.360313315926893, + "grad_norm": 1.0728485584259033, + "learning_rate": 4.037199863775336e-05, + "loss": 1.0108, + "step": 5244 + }, + { + "epoch": 0.3603820255599835, + "grad_norm": 1.022675633430481, + "learning_rate": 4.0367269372361875e-05, + "loss": 0.9945, + "step": 5245 + }, + { + "epoch": 0.3604507351930741, + "grad_norm": 1.1559593677520752, + "learning_rate": 4.0362539222900744e-05, + "loss": 0.9958, + "step": 5246 + }, + { + "epoch": 0.3605194448261646, + "grad_norm": 1.0538747310638428, + "learning_rate": 4.035780818964211e-05, + "loss": 0.9891, + "step": 5247 + }, + { + "epoch": 0.3605881544592552, + "grad_norm": 1.089695930480957, + "learning_rate": 4.0353076272858124e-05, + "loss": 0.99, + "step": 5248 + }, + { + "epoch": 0.36065686409234576, + "grad_norm": 1.0826351642608643, + "learning_rate": 4.034834347282103e-05, + "loss": 0.9852, + "step": 5249 + }, + { + "epoch": 0.3607255737254363, + "grad_norm": 1.2572968006134033, + "learning_rate": 4.034360978980309e-05, + "loss": 1.0267, + "step": 5250 + }, + { + "epoch": 0.36079428335852687, + "grad_norm": 1.1624269485473633, + "learning_rate": 4.0338875224076636e-05, + "loss": 0.9798, + "step": 5251 + }, + { + "epoch": 0.36086299299161745, + "grad_norm": 1.1663967370986938, + "learning_rate": 4.033413977591404e-05, + "loss": 0.999, + "step": 5252 + }, + { + "epoch": 0.360931702624708, + "grad_norm": 1.1764107942581177, + "learning_rate": 4.0329403445587735e-05, + "loss": 1.014, + "step": 5253 + }, + { + "epoch": 0.36100041225779855, + "grad_norm": 1.2538641691207886, + "learning_rate": 4.03246662333702e-05, + "loss": 1.0663, + "step": 5254 + }, + { + "epoch": 0.3610691218908891, + "grad_norm": 1.0460015535354614, + "learning_rate": 4.031992813953395e-05, + "loss": 0.9889, + "step": 5255 + }, + { + "epoch": 0.36113783152397966, + "grad_norm": 1.0959489345550537, + "learning_rate": 4.031518916435158e-05, + "loss": 1.0237, + "step": 5256 + }, + { + "epoch": 0.36120654115707024, + "grad_norm": 1.1291438341140747, + "learning_rate": 4.031044930809572e-05, + "loss": 0.9607, + "step": 5257 + }, + { + "epoch": 0.36127525079016076, + "grad_norm": 1.0556457042694092, + "learning_rate": 4.0305708571039036e-05, + "loss": 0.9364, + "step": 5258 + }, + { + "epoch": 0.36134396042325134, + "grad_norm": 1.1064507961273193, + "learning_rate": 4.030096695345428e-05, + "loss": 1.0268, + "step": 5259 + }, + { + "epoch": 0.3614126700563419, + "grad_norm": 1.1562494039535522, + "learning_rate": 4.0296224455614226e-05, + "loss": 1.0096, + "step": 5260 + }, + { + "epoch": 0.36148137968943245, + "grad_norm": 1.1618168354034424, + "learning_rate": 4.0291481077791696e-05, + "loss": 1.0272, + "step": 5261 + }, + { + "epoch": 0.361550089322523, + "grad_norm": 1.1630067825317383, + "learning_rate": 4.02867368202596e-05, + "loss": 0.9817, + "step": 5262 + }, + { + "epoch": 0.36161879895561355, + "grad_norm": 1.0058321952819824, + "learning_rate": 4.028199168329084e-05, + "loss": 0.9834, + "step": 5263 + }, + { + "epoch": 0.36168750858870413, + "grad_norm": 1.0613391399383545, + "learning_rate": 4.0277245667158424e-05, + "loss": 1.0062, + "step": 5264 + }, + { + "epoch": 0.3617562182217947, + "grad_norm": 1.1205861568450928, + "learning_rate": 4.027249877213537e-05, + "loss": 0.9935, + "step": 5265 + }, + { + "epoch": 0.36182492785488524, + "grad_norm": 1.0937259197235107, + "learning_rate": 4.026775099849478e-05, + "loss": 1.0047, + "step": 5266 + }, + { + "epoch": 0.3618936374879758, + "grad_norm": 1.084435224533081, + "learning_rate": 4.026300234650979e-05, + "loss": 0.9666, + "step": 5267 + }, + { + "epoch": 0.3619623471210664, + "grad_norm": 1.095532774925232, + "learning_rate": 4.0258252816453576e-05, + "loss": 1.0228, + "step": 5268 + }, + { + "epoch": 0.3620310567541569, + "grad_norm": 1.0439544916152954, + "learning_rate": 4.025350240859937e-05, + "loss": 0.9843, + "step": 5269 + }, + { + "epoch": 0.3620997663872475, + "grad_norm": 1.026584506034851, + "learning_rate": 4.02487511232205e-05, + "loss": 0.9865, + "step": 5270 + }, + { + "epoch": 0.362168476020338, + "grad_norm": 1.107568383216858, + "learning_rate": 4.024399896059025e-05, + "loss": 1.0243, + "step": 5271 + }, + { + "epoch": 0.3622371856534286, + "grad_norm": 1.0481411218643188, + "learning_rate": 4.023924592098204e-05, + "loss": 0.9671, + "step": 5272 + }, + { + "epoch": 0.3623058952865192, + "grad_norm": 1.049780249595642, + "learning_rate": 4.0234492004669314e-05, + "loss": 0.9796, + "step": 5273 + }, + { + "epoch": 0.3623746049196097, + "grad_norm": 1.1673166751861572, + "learning_rate": 4.0229737211925536e-05, + "loss": 0.9582, + "step": 5274 + }, + { + "epoch": 0.3624433145527003, + "grad_norm": 1.1454499959945679, + "learning_rate": 4.0224981543024274e-05, + "loss": 1.03, + "step": 5275 + }, + { + "epoch": 0.36251202418579087, + "grad_norm": 1.0970247983932495, + "learning_rate": 4.0220224998239107e-05, + "loss": 1.0141, + "step": 5276 + }, + { + "epoch": 0.3625807338188814, + "grad_norm": 1.0680837631225586, + "learning_rate": 4.0215467577843665e-05, + "loss": 0.9594, + "step": 5277 + }, + { + "epoch": 0.362649443451972, + "grad_norm": 1.0774484872817993, + "learning_rate": 4.0210709282111664e-05, + "loss": 0.9627, + "step": 5278 + }, + { + "epoch": 0.3627181530850625, + "grad_norm": 1.1396852731704712, + "learning_rate": 4.0205950111316823e-05, + "loss": 0.9649, + "step": 5279 + }, + { + "epoch": 0.3627868627181531, + "grad_norm": 1.121199369430542, + "learning_rate": 4.020119006573295e-05, + "loss": 1.0251, + "step": 5280 + }, + { + "epoch": 0.36285557235124366, + "grad_norm": 1.1505086421966553, + "learning_rate": 4.0196429145633884e-05, + "loss": 1.0045, + "step": 5281 + }, + { + "epoch": 0.3629242819843342, + "grad_norm": 1.0215157270431519, + "learning_rate": 4.0191667351293504e-05, + "loss": 0.9815, + "step": 5282 + }, + { + "epoch": 0.36299299161742477, + "grad_norm": 1.1172053813934326, + "learning_rate": 4.018690468298577e-05, + "loss": 0.9615, + "step": 5283 + }, + { + "epoch": 0.36306170125051535, + "grad_norm": 1.115746021270752, + "learning_rate": 4.018214114098467e-05, + "loss": 0.9755, + "step": 5284 + }, + { + "epoch": 0.36313041088360587, + "grad_norm": 1.042150616645813, + "learning_rate": 4.0177376725564255e-05, + "loss": 0.9666, + "step": 5285 + }, + { + "epoch": 0.36319912051669645, + "grad_norm": 1.0979843139648438, + "learning_rate": 4.0172611436998594e-05, + "loss": 0.9986, + "step": 5286 + }, + { + "epoch": 0.363267830149787, + "grad_norm": 1.1661038398742676, + "learning_rate": 4.016784527556187e-05, + "loss": 1.0124, + "step": 5287 + }, + { + "epoch": 0.36333653978287755, + "grad_norm": 1.089908242225647, + "learning_rate": 4.0163078241528235e-05, + "loss": 1.0009, + "step": 5288 + }, + { + "epoch": 0.36340524941596813, + "grad_norm": 1.0321866273880005, + "learning_rate": 4.015831033517196e-05, + "loss": 0.9563, + "step": 5289 + }, + { + "epoch": 0.36347395904905866, + "grad_norm": 1.2240160703659058, + "learning_rate": 4.015354155676734e-05, + "loss": 0.9809, + "step": 5290 + }, + { + "epoch": 0.36354266868214924, + "grad_norm": 1.0834475755691528, + "learning_rate": 4.0148771906588706e-05, + "loss": 1.0206, + "step": 5291 + }, + { + "epoch": 0.3636113783152398, + "grad_norm": 1.1160221099853516, + "learning_rate": 4.014400138491047e-05, + "loss": 1.0144, + "step": 5292 + }, + { + "epoch": 0.36368008794833034, + "grad_norm": 1.1219310760498047, + "learning_rate": 4.013922999200706e-05, + "loss": 0.9571, + "step": 5293 + }, + { + "epoch": 0.3637487975814209, + "grad_norm": 1.139559030532837, + "learning_rate": 4.013445772815298e-05, + "loss": 0.989, + "step": 5294 + }, + { + "epoch": 0.36381750721451145, + "grad_norm": 1.0946044921875, + "learning_rate": 4.012968459362278e-05, + "loss": 1.0168, + "step": 5295 + }, + { + "epoch": 0.36388621684760203, + "grad_norm": 1.1236326694488525, + "learning_rate": 4.012491058869104e-05, + "loss": 0.9863, + "step": 5296 + }, + { + "epoch": 0.3639549264806926, + "grad_norm": 1.0957008600234985, + "learning_rate": 4.0120135713632425e-05, + "loss": 1.0076, + "step": 5297 + }, + { + "epoch": 0.36402363611378313, + "grad_norm": 1.095502495765686, + "learning_rate": 4.0115359968721624e-05, + "loss": 1.0183, + "step": 5298 + }, + { + "epoch": 0.3640923457468737, + "grad_norm": 1.1260255575180054, + "learning_rate": 4.0110583354233375e-05, + "loss": 1.0046, + "step": 5299 + }, + { + "epoch": 0.3641610553799643, + "grad_norm": 1.1175310611724854, + "learning_rate": 4.010580587044248e-05, + "loss": 0.9807, + "step": 5300 + }, + { + "epoch": 0.3642297650130548, + "grad_norm": 1.2104113101959229, + "learning_rate": 4.010102751762379e-05, + "loss": 0.9694, + "step": 5301 + }, + { + "epoch": 0.3642984746461454, + "grad_norm": 1.1471190452575684, + "learning_rate": 4.0096248296052186e-05, + "loss": 0.9995, + "step": 5302 + }, + { + "epoch": 0.3643671842792359, + "grad_norm": 1.0741945505142212, + "learning_rate": 4.009146820600263e-05, + "loss": 1.0035, + "step": 5303 + }, + { + "epoch": 0.3644358939123265, + "grad_norm": 1.0507900714874268, + "learning_rate": 4.00866872477501e-05, + "loss": 0.9716, + "step": 5304 + }, + { + "epoch": 0.3645046035454171, + "grad_norm": 1.1214643716812134, + "learning_rate": 4.008190542156965e-05, + "loss": 1.0165, + "step": 5305 + }, + { + "epoch": 0.3645733131785076, + "grad_norm": 1.113943338394165, + "learning_rate": 4.0077122727736396e-05, + "loss": 0.9879, + "step": 5306 + }, + { + "epoch": 0.3646420228115982, + "grad_norm": 1.0854878425598145, + "learning_rate": 4.007233916652544e-05, + "loss": 1.0068, + "step": 5307 + }, + { + "epoch": 0.36471073244468877, + "grad_norm": 1.0423377752304077, + "learning_rate": 4.006755473821202e-05, + "loss": 0.9544, + "step": 5308 + }, + { + "epoch": 0.3647794420777793, + "grad_norm": 1.0629942417144775, + "learning_rate": 4.0062769443071355e-05, + "loss": 0.9769, + "step": 5309 + }, + { + "epoch": 0.3648481517108699, + "grad_norm": 1.0966823101043701, + "learning_rate": 4.005798328137875e-05, + "loss": 0.9843, + "step": 5310 + }, + { + "epoch": 0.3649168613439604, + "grad_norm": 1.0926939249038696, + "learning_rate": 4.005319625340955e-05, + "loss": 0.9882, + "step": 5311 + }, + { + "epoch": 0.364985570977051, + "grad_norm": 1.0954961776733398, + "learning_rate": 4.004840835943915e-05, + "loss": 1.0196, + "step": 5312 + }, + { + "epoch": 0.36505428061014156, + "grad_norm": 1.0287151336669922, + "learning_rate": 4.004361959974299e-05, + "loss": 0.9692, + "step": 5313 + }, + { + "epoch": 0.3651229902432321, + "grad_norm": 1.1142406463623047, + "learning_rate": 4.0038829974596556e-05, + "loss": 0.9855, + "step": 5314 + }, + { + "epoch": 0.36519169987632266, + "grad_norm": 1.1182336807250977, + "learning_rate": 4.003403948427542e-05, + "loss": 0.9898, + "step": 5315 + }, + { + "epoch": 0.36526040950941324, + "grad_norm": 1.03201425075531, + "learning_rate": 4.002924812905515e-05, + "loss": 0.9889, + "step": 5316 + }, + { + "epoch": 0.36532911914250377, + "grad_norm": 1.0531588792800903, + "learning_rate": 4.00244559092114e-05, + "loss": 0.9967, + "step": 5317 + }, + { + "epoch": 0.36539782877559435, + "grad_norm": 1.0959587097167969, + "learning_rate": 4.0019662825019864e-05, + "loss": 1.0119, + "step": 5318 + }, + { + "epoch": 0.36546653840868487, + "grad_norm": 1.0344326496124268, + "learning_rate": 4.001486887675628e-05, + "loss": 1.0101, + "step": 5319 + }, + { + "epoch": 0.36553524804177545, + "grad_norm": 1.0795083045959473, + "learning_rate": 4.001007406469646e-05, + "loss": 0.9934, + "step": 5320 + }, + { + "epoch": 0.36560395767486603, + "grad_norm": 1.1326823234558105, + "learning_rate": 4.000527838911622e-05, + "loss": 1.0085, + "step": 5321 + }, + { + "epoch": 0.36567266730795656, + "grad_norm": 1.1322611570358276, + "learning_rate": 4.0000481850291455e-05, + "loss": 0.9791, + "step": 5322 + }, + { + "epoch": 0.36574137694104714, + "grad_norm": 1.1029685735702515, + "learning_rate": 3.999568444849813e-05, + "loss": 1.0075, + "step": 5323 + }, + { + "epoch": 0.3658100865741377, + "grad_norm": 1.1296465396881104, + "learning_rate": 3.999088618401221e-05, + "loss": 0.9631, + "step": 5324 + }, + { + "epoch": 0.36587879620722824, + "grad_norm": 1.0889583826065063, + "learning_rate": 3.998608705710975e-05, + "loss": 0.9968, + "step": 5325 + }, + { + "epoch": 0.3659475058403188, + "grad_norm": 1.0837490558624268, + "learning_rate": 3.998128706806684e-05, + "loss": 0.9616, + "step": 5326 + }, + { + "epoch": 0.36601621547340935, + "grad_norm": 1.0880788564682007, + "learning_rate": 3.9976486217159625e-05, + "loss": 0.9764, + "step": 5327 + }, + { + "epoch": 0.3660849251064999, + "grad_norm": 1.0540189743041992, + "learning_rate": 3.9971684504664295e-05, + "loss": 1.0068, + "step": 5328 + }, + { + "epoch": 0.3661536347395905, + "grad_norm": 1.01442289352417, + "learning_rate": 3.996688193085707e-05, + "loss": 0.9423, + "step": 5329 + }, + { + "epoch": 0.36622234437268103, + "grad_norm": 1.1028790473937988, + "learning_rate": 3.9962078496014255e-05, + "loss": 1.0136, + "step": 5330 + }, + { + "epoch": 0.3662910540057716, + "grad_norm": 1.1256109476089478, + "learning_rate": 3.9957274200412197e-05, + "loss": 1.03, + "step": 5331 + }, + { + "epoch": 0.3663597636388622, + "grad_norm": 1.053491473197937, + "learning_rate": 3.995246904432727e-05, + "loss": 0.9995, + "step": 5332 + }, + { + "epoch": 0.3664284732719527, + "grad_norm": 1.1394803524017334, + "learning_rate": 3.9947663028035905e-05, + "loss": 1.0486, + "step": 5333 + }, + { + "epoch": 0.3664971829050433, + "grad_norm": 1.0399543046951294, + "learning_rate": 3.9942856151814616e-05, + "loss": 0.976, + "step": 5334 + }, + { + "epoch": 0.3665658925381338, + "grad_norm": 1.1557612419128418, + "learning_rate": 3.9938048415939914e-05, + "loss": 1.0191, + "step": 5335 + }, + { + "epoch": 0.3666346021712244, + "grad_norm": 1.146078109741211, + "learning_rate": 3.99332398206884e-05, + "loss": 0.9805, + "step": 5336 + }, + { + "epoch": 0.366703311804315, + "grad_norm": 1.8104833364486694, + "learning_rate": 3.99284303663367e-05, + "loss": 1.0303, + "step": 5337 + }, + { + "epoch": 0.3667720214374055, + "grad_norm": 1.057808518409729, + "learning_rate": 3.9923620053161515e-05, + "loss": 0.9625, + "step": 5338 + }, + { + "epoch": 0.3668407310704961, + "grad_norm": 1.092546820640564, + "learning_rate": 3.991880888143955e-05, + "loss": 0.988, + "step": 5339 + }, + { + "epoch": 0.36690944070358666, + "grad_norm": 1.117030382156372, + "learning_rate": 3.9913996851447615e-05, + "loss": 0.9741, + "step": 5340 + }, + { + "epoch": 0.3669781503366772, + "grad_norm": 1.1627845764160156, + "learning_rate": 3.990918396346254e-05, + "loss": 1.0141, + "step": 5341 + }, + { + "epoch": 0.36704685996976777, + "grad_norm": 1.172778844833374, + "learning_rate": 3.9904370217761195e-05, + "loss": 1.0014, + "step": 5342 + }, + { + "epoch": 0.3671155696028583, + "grad_norm": 1.1329686641693115, + "learning_rate": 3.989955561462052e-05, + "loss": 0.9873, + "step": 5343 + }, + { + "epoch": 0.3671842792359489, + "grad_norm": 1.1235606670379639, + "learning_rate": 3.98947401543175e-05, + "loss": 0.9988, + "step": 5344 + }, + { + "epoch": 0.36725298886903945, + "grad_norm": 1.1024506092071533, + "learning_rate": 3.988992383712915e-05, + "loss": 1.0147, + "step": 5345 + }, + { + "epoch": 0.36732169850213, + "grad_norm": 1.0850709676742554, + "learning_rate": 3.9885106663332574e-05, + "loss": 0.9913, + "step": 5346 + }, + { + "epoch": 0.36739040813522056, + "grad_norm": 1.0816618204116821, + "learning_rate": 3.988028863320488e-05, + "loss": 1.0095, + "step": 5347 + }, + { + "epoch": 0.36745911776831114, + "grad_norm": 1.0861314535140991, + "learning_rate": 3.9875469747023255e-05, + "loss": 0.9839, + "step": 5348 + }, + { + "epoch": 0.36752782740140166, + "grad_norm": 1.098950982093811, + "learning_rate": 3.9870650005064926e-05, + "loss": 0.9888, + "step": 5349 + }, + { + "epoch": 0.36759653703449224, + "grad_norm": 1.1840109825134277, + "learning_rate": 3.986582940760717e-05, + "loss": 0.962, + "step": 5350 + }, + { + "epoch": 0.36766524666758277, + "grad_norm": 1.1143990755081177, + "learning_rate": 3.986100795492731e-05, + "loss": 1.0312, + "step": 5351 + }, + { + "epoch": 0.36773395630067335, + "grad_norm": 1.0306155681610107, + "learning_rate": 3.9856185647302726e-05, + "loss": 0.9789, + "step": 5352 + }, + { + "epoch": 0.36780266593376393, + "grad_norm": 1.1300791501998901, + "learning_rate": 3.9851362485010834e-05, + "loss": 1.0264, + "step": 5353 + }, + { + "epoch": 0.36787137556685445, + "grad_norm": 1.1478816270828247, + "learning_rate": 3.984653846832912e-05, + "loss": 0.9918, + "step": 5354 + }, + { + "epoch": 0.36794008519994503, + "grad_norm": 1.0858619213104248, + "learning_rate": 3.98417135975351e-05, + "loss": 1.0086, + "step": 5355 + }, + { + "epoch": 0.3680087948330356, + "grad_norm": 1.0312639474868774, + "learning_rate": 3.983688787290635e-05, + "loss": 1.0066, + "step": 5356 + }, + { + "epoch": 0.36807750446612614, + "grad_norm": 1.0862618684768677, + "learning_rate": 3.983206129472048e-05, + "loss": 1.0228, + "step": 5357 + }, + { + "epoch": 0.3681462140992167, + "grad_norm": 1.1603033542633057, + "learning_rate": 3.982723386325517e-05, + "loss": 1.0199, + "step": 5358 + }, + { + "epoch": 0.36821492373230724, + "grad_norm": 1.1152985095977783, + "learning_rate": 3.982240557878814e-05, + "loss": 0.9154, + "step": 5359 + }, + { + "epoch": 0.3682836333653978, + "grad_norm": 1.1507495641708374, + "learning_rate": 3.981757644159714e-05, + "loss": 0.9796, + "step": 5360 + }, + { + "epoch": 0.3683523429984884, + "grad_norm": 1.0995047092437744, + "learning_rate": 3.981274645196002e-05, + "loss": 1.0232, + "step": 5361 + }, + { + "epoch": 0.3684210526315789, + "grad_norm": 1.0763981342315674, + "learning_rate": 3.9807915610154623e-05, + "loss": 1.0224, + "step": 5362 + }, + { + "epoch": 0.3684897622646695, + "grad_norm": 1.0824306011199951, + "learning_rate": 3.980308391645886e-05, + "loss": 0.94, + "step": 5363 + }, + { + "epoch": 0.3685584718977601, + "grad_norm": 1.102342963218689, + "learning_rate": 3.9798251371150716e-05, + "loss": 0.9923, + "step": 5364 + }, + { + "epoch": 0.3686271815308506, + "grad_norm": 1.1050269603729248, + "learning_rate": 3.97934179745082e-05, + "loss": 1.0043, + "step": 5365 + }, + { + "epoch": 0.3686958911639412, + "grad_norm": 1.0590091943740845, + "learning_rate": 3.9788583726809354e-05, + "loss": 1.0344, + "step": 5366 + }, + { + "epoch": 0.3687646007970317, + "grad_norm": 1.0545026063919067, + "learning_rate": 3.978374862833231e-05, + "loss": 0.9878, + "step": 5367 + }, + { + "epoch": 0.3688333104301223, + "grad_norm": 1.1290544271469116, + "learning_rate": 3.977891267935522e-05, + "loss": 1.0018, + "step": 5368 + }, + { + "epoch": 0.3689020200632129, + "grad_norm": 1.0676848888397217, + "learning_rate": 3.9774075880156293e-05, + "loss": 0.9862, + "step": 5369 + }, + { + "epoch": 0.3689707296963034, + "grad_norm": 1.1550586223602295, + "learning_rate": 3.976923823101379e-05, + "loss": 1.0016, + "step": 5370 + }, + { + "epoch": 0.369039439329394, + "grad_norm": 1.045777678489685, + "learning_rate": 3.9764399732206016e-05, + "loss": 1.0061, + "step": 5371 + }, + { + "epoch": 0.36910814896248456, + "grad_norm": 1.0654953718185425, + "learning_rate": 3.9759560384011326e-05, + "loss": 0.9962, + "step": 5372 + }, + { + "epoch": 0.3691768585955751, + "grad_norm": 1.101372480392456, + "learning_rate": 3.975472018670813e-05, + "loss": 0.983, + "step": 5373 + }, + { + "epoch": 0.36924556822866567, + "grad_norm": 1.0573173761367798, + "learning_rate": 3.974987914057488e-05, + "loss": 0.9866, + "step": 5374 + }, + { + "epoch": 0.3693142778617562, + "grad_norm": 1.187264323234558, + "learning_rate": 3.9745037245890074e-05, + "loss": 1.0154, + "step": 5375 + }, + { + "epoch": 0.36938298749484677, + "grad_norm": 1.1361008882522583, + "learning_rate": 3.974019450293227e-05, + "loss": 0.9981, + "step": 5376 + }, + { + "epoch": 0.36945169712793735, + "grad_norm": 1.048796534538269, + "learning_rate": 3.9735350911980055e-05, + "loss": 1.0047, + "step": 5377 + }, + { + "epoch": 0.3695204067610279, + "grad_norm": 1.19966721534729, + "learning_rate": 3.973050647331209e-05, + "loss": 1.0127, + "step": 5378 + }, + { + "epoch": 0.36958911639411846, + "grad_norm": 1.2475837469100952, + "learning_rate": 3.972566118720707e-05, + "loss": 1.0108, + "step": 5379 + }, + { + "epoch": 0.36965782602720904, + "grad_norm": 1.137022852897644, + "learning_rate": 3.972081505394375e-05, + "loss": 0.9857, + "step": 5380 + }, + { + "epoch": 0.36972653566029956, + "grad_norm": 1.1570111513137817, + "learning_rate": 3.97159680738009e-05, + "loss": 1.0308, + "step": 5381 + }, + { + "epoch": 0.36979524529339014, + "grad_norm": 1.1456255912780762, + "learning_rate": 3.971112024705739e-05, + "loss": 0.9802, + "step": 5382 + }, + { + "epoch": 0.3698639549264807, + "grad_norm": 1.098300814628601, + "learning_rate": 3.970627157399209e-05, + "loss": 0.9694, + "step": 5383 + }, + { + "epoch": 0.36993266455957124, + "grad_norm": 1.1061497926712036, + "learning_rate": 3.9701422054883973e-05, + "loss": 1.0317, + "step": 5384 + }, + { + "epoch": 0.3700013741926618, + "grad_norm": 1.2172231674194336, + "learning_rate": 3.9696571690011995e-05, + "loss": 0.995, + "step": 5385 + }, + { + "epoch": 0.37007008382575235, + "grad_norm": 1.1376475095748901, + "learning_rate": 3.9691720479655215e-05, + "loss": 1.0073, + "step": 5386 + }, + { + "epoch": 0.37013879345884293, + "grad_norm": 1.044442892074585, + "learning_rate": 3.968686842409271e-05, + "loss": 0.9947, + "step": 5387 + }, + { + "epoch": 0.3702075030919335, + "grad_norm": 1.1830315589904785, + "learning_rate": 3.968201552360361e-05, + "loss": 0.9944, + "step": 5388 + }, + { + "epoch": 0.37027621272502403, + "grad_norm": 1.0615675449371338, + "learning_rate": 3.967716177846713e-05, + "loss": 0.9567, + "step": 5389 + }, + { + "epoch": 0.3703449223581146, + "grad_norm": 1.1049731969833374, + "learning_rate": 3.967230718896247e-05, + "loss": 1.0071, + "step": 5390 + }, + { + "epoch": 0.3704136319912052, + "grad_norm": 1.0690903663635254, + "learning_rate": 3.9667451755368924e-05, + "loss": 0.9647, + "step": 5391 + }, + { + "epoch": 0.3704823416242957, + "grad_norm": 1.0637274980545044, + "learning_rate": 3.966259547796583e-05, + "loss": 0.9766, + "step": 5392 + }, + { + "epoch": 0.3705510512573863, + "grad_norm": 1.1282966136932373, + "learning_rate": 3.965773835703254e-05, + "loss": 0.9863, + "step": 5393 + }, + { + "epoch": 0.3706197608904768, + "grad_norm": 1.0629407167434692, + "learning_rate": 3.9652880392848514e-05, + "loss": 0.9805, + "step": 5394 + }, + { + "epoch": 0.3706884705235674, + "grad_norm": 1.084092378616333, + "learning_rate": 3.9648021585693205e-05, + "loss": 0.9929, + "step": 5395 + }, + { + "epoch": 0.370757180156658, + "grad_norm": 1.0660361051559448, + "learning_rate": 3.9643161935846154e-05, + "loss": 0.9996, + "step": 5396 + }, + { + "epoch": 0.3708258897897485, + "grad_norm": 1.1267366409301758, + "learning_rate": 3.9638301443586925e-05, + "loss": 0.9785, + "step": 5397 + }, + { + "epoch": 0.3708945994228391, + "grad_norm": 1.1098339557647705, + "learning_rate": 3.963344010919513e-05, + "loss": 1.005, + "step": 5398 + }, + { + "epoch": 0.37096330905592967, + "grad_norm": 1.1280295848846436, + "learning_rate": 3.962857793295046e-05, + "loss": 0.9735, + "step": 5399 + }, + { + "epoch": 0.3710320186890202, + "grad_norm": 1.1493481397628784, + "learning_rate": 3.962371491513262e-05, + "loss": 1.0428, + "step": 5400 + }, + { + "epoch": 0.3711007283221108, + "grad_norm": 1.0941275358200073, + "learning_rate": 3.9618851056021364e-05, + "loss": 0.9859, + "step": 5401 + }, + { + "epoch": 0.3711694379552013, + "grad_norm": 1.1683557033538818, + "learning_rate": 3.961398635589653e-05, + "loss": 0.9386, + "step": 5402 + }, + { + "epoch": 0.3712381475882919, + "grad_norm": 1.0912718772888184, + "learning_rate": 3.960912081503797e-05, + "loss": 1.0021, + "step": 5403 + }, + { + "epoch": 0.37130685722138246, + "grad_norm": 1.1875755786895752, + "learning_rate": 3.96042544337256e-05, + "loss": 1.0008, + "step": 5404 + }, + { + "epoch": 0.371375566854473, + "grad_norm": 1.0314416885375977, + "learning_rate": 3.9599387212239375e-05, + "loss": 0.9693, + "step": 5405 + }, + { + "epoch": 0.37144427648756356, + "grad_norm": 1.1037518978118896, + "learning_rate": 3.9594519150859306e-05, + "loss": 0.9664, + "step": 5406 + }, + { + "epoch": 0.37151298612065414, + "grad_norm": 1.0835572481155396, + "learning_rate": 3.958965024986545e-05, + "loss": 0.9884, + "step": 5407 + }, + { + "epoch": 0.37158169575374467, + "grad_norm": 1.0921335220336914, + "learning_rate": 3.958478050953791e-05, + "loss": 1.0105, + "step": 5408 + }, + { + "epoch": 0.37165040538683525, + "grad_norm": 1.1386774778366089, + "learning_rate": 3.957990993015683e-05, + "loss": 0.9861, + "step": 5409 + }, + { + "epoch": 0.37171911501992577, + "grad_norm": 1.1200408935546875, + "learning_rate": 3.957503851200244e-05, + "loss": 1.022, + "step": 5410 + }, + { + "epoch": 0.37178782465301635, + "grad_norm": 1.0937930345535278, + "learning_rate": 3.957016625535496e-05, + "loss": 1.0103, + "step": 5411 + }, + { + "epoch": 0.37185653428610693, + "grad_norm": 1.0684657096862793, + "learning_rate": 3.9565293160494705e-05, + "loss": 0.9795, + "step": 5412 + }, + { + "epoch": 0.37192524391919746, + "grad_norm": 1.0633888244628906, + "learning_rate": 3.9560419227702006e-05, + "loss": 0.9922, + "step": 5413 + }, + { + "epoch": 0.37199395355228804, + "grad_norm": 1.0481617450714111, + "learning_rate": 3.955554445725728e-05, + "loss": 1.0316, + "step": 5414 + }, + { + "epoch": 0.3720626631853786, + "grad_norm": 1.1549104452133179, + "learning_rate": 3.955066884944094e-05, + "loss": 0.9666, + "step": 5415 + }, + { + "epoch": 0.37213137281846914, + "grad_norm": 1.1443121433258057, + "learning_rate": 3.954579240453351e-05, + "loss": 1.0406, + "step": 5416 + }, + { + "epoch": 0.3722000824515597, + "grad_norm": 1.165172815322876, + "learning_rate": 3.95409151228155e-05, + "loss": 0.9981, + "step": 5417 + }, + { + "epoch": 0.37226879208465025, + "grad_norm": 1.1133649349212646, + "learning_rate": 3.953603700456752e-05, + "loss": 0.9696, + "step": 5418 + }, + { + "epoch": 0.3723375017177408, + "grad_norm": 1.0722463130950928, + "learning_rate": 3.953115805007018e-05, + "loss": 0.9587, + "step": 5419 + }, + { + "epoch": 0.3724062113508314, + "grad_norm": 1.0837022066116333, + "learning_rate": 3.952627825960419e-05, + "loss": 0.9209, + "step": 5420 + }, + { + "epoch": 0.37247492098392193, + "grad_norm": 1.0874396562576294, + "learning_rate": 3.952139763345026e-05, + "loss": 0.9581, + "step": 5421 + }, + { + "epoch": 0.3725436306170125, + "grad_norm": 1.111142635345459, + "learning_rate": 3.9516516171889186e-05, + "loss": 0.9867, + "step": 5422 + }, + { + "epoch": 0.3726123402501031, + "grad_norm": 1.095029354095459, + "learning_rate": 3.951163387520178e-05, + "loss": 1.0339, + "step": 5423 + }, + { + "epoch": 0.3726810498831936, + "grad_norm": 1.118064522743225, + "learning_rate": 3.950675074366893e-05, + "loss": 0.9895, + "step": 5424 + }, + { + "epoch": 0.3727497595162842, + "grad_norm": 1.0700663328170776, + "learning_rate": 3.950186677757156e-05, + "loss": 1.0339, + "step": 5425 + }, + { + "epoch": 0.3728184691493747, + "grad_norm": 1.0598098039627075, + "learning_rate": 3.9496981977190625e-05, + "loss": 0.9744, + "step": 5426 + }, + { + "epoch": 0.3728871787824653, + "grad_norm": 1.0479406118392944, + "learning_rate": 3.949209634280716e-05, + "loss": 0.9891, + "step": 5427 + }, + { + "epoch": 0.3729558884155559, + "grad_norm": 1.1761751174926758, + "learning_rate": 3.9487209874702236e-05, + "loss": 0.9813, + "step": 5428 + }, + { + "epoch": 0.3730245980486464, + "grad_norm": 1.1329879760742188, + "learning_rate": 3.948232257315695e-05, + "loss": 0.9809, + "step": 5429 + }, + { + "epoch": 0.373093307681737, + "grad_norm": 1.0623250007629395, + "learning_rate": 3.9477434438452486e-05, + "loss": 0.9735, + "step": 5430 + }, + { + "epoch": 0.37316201731482757, + "grad_norm": 1.1361876726150513, + "learning_rate": 3.947254547087004e-05, + "loss": 0.9669, + "step": 5431 + }, + { + "epoch": 0.3732307269479181, + "grad_norm": 1.070910096168518, + "learning_rate": 3.9467655670690884e-05, + "loss": 0.9794, + "step": 5432 + }, + { + "epoch": 0.37329943658100867, + "grad_norm": 1.0840610265731812, + "learning_rate": 3.9462765038196316e-05, + "loss": 0.9801, + "step": 5433 + }, + { + "epoch": 0.3733681462140992, + "grad_norm": 1.0262733697891235, + "learning_rate": 3.9457873573667693e-05, + "loss": 0.973, + "step": 5434 + }, + { + "epoch": 0.3734368558471898, + "grad_norm": 1.139829158782959, + "learning_rate": 3.945298127738642e-05, + "loss": 0.9754, + "step": 5435 + }, + { + "epoch": 0.37350556548028035, + "grad_norm": 1.115731120109558, + "learning_rate": 3.944808814963395e-05, + "loss": 1.0125, + "step": 5436 + }, + { + "epoch": 0.3735742751133709, + "grad_norm": 1.0814082622528076, + "learning_rate": 3.944319419069177e-05, + "loss": 0.9679, + "step": 5437 + }, + { + "epoch": 0.37364298474646146, + "grad_norm": 1.0799790620803833, + "learning_rate": 3.943829940084144e-05, + "loss": 0.9979, + "step": 5438 + }, + { + "epoch": 0.37371169437955204, + "grad_norm": 1.1432479619979858, + "learning_rate": 3.9433403780364545e-05, + "loss": 1.0125, + "step": 5439 + }, + { + "epoch": 0.37378040401264256, + "grad_norm": 1.1186528205871582, + "learning_rate": 3.942850732954273e-05, + "loss": 0.9787, + "step": 5440 + }, + { + "epoch": 0.37384911364573314, + "grad_norm": 1.0444098711013794, + "learning_rate": 3.9423610048657686e-05, + "loss": 0.9503, + "step": 5441 + }, + { + "epoch": 0.37391782327882367, + "grad_norm": 1.0729337930679321, + "learning_rate": 3.941871193799115e-05, + "loss": 0.9858, + "step": 5442 + }, + { + "epoch": 0.37398653291191425, + "grad_norm": 1.0378741025924683, + "learning_rate": 3.9413812997824915e-05, + "loss": 0.9743, + "step": 5443 + }, + { + "epoch": 0.37405524254500483, + "grad_norm": 1.0870440006256104, + "learning_rate": 3.9408913228440805e-05, + "loss": 0.9765, + "step": 5444 + }, + { + "epoch": 0.37412395217809535, + "grad_norm": 1.1052837371826172, + "learning_rate": 3.94040126301207e-05, + "loss": 0.9953, + "step": 5445 + }, + { + "epoch": 0.37419266181118593, + "grad_norm": 1.102900505065918, + "learning_rate": 3.939911120314653e-05, + "loss": 0.967, + "step": 5446 + }, + { + "epoch": 0.3742613714442765, + "grad_norm": 1.02222740650177, + "learning_rate": 3.9394208947800274e-05, + "loss": 0.97, + "step": 5447 + }, + { + "epoch": 0.37433008107736704, + "grad_norm": 1.0711361169815063, + "learning_rate": 3.938930586436396e-05, + "loss": 0.9755, + "step": 5448 + }, + { + "epoch": 0.3743987907104576, + "grad_norm": 1.0523061752319336, + "learning_rate": 3.938440195311964e-05, + "loss": 0.988, + "step": 5449 + }, + { + "epoch": 0.37446750034354814, + "grad_norm": 1.0767430067062378, + "learning_rate": 3.9379497214349456e-05, + "loss": 0.9774, + "step": 5450 + }, + { + "epoch": 0.3745362099766387, + "grad_norm": 1.08974289894104, + "learning_rate": 3.9374591648335555e-05, + "loss": 0.9568, + "step": 5451 + }, + { + "epoch": 0.3746049196097293, + "grad_norm": 1.0768944025039673, + "learning_rate": 3.9369685255360175e-05, + "loss": 0.967, + "step": 5452 + }, + { + "epoch": 0.37467362924281983, + "grad_norm": 1.1054508686065674, + "learning_rate": 3.9364778035705555e-05, + "loss": 0.971, + "step": 5453 + }, + { + "epoch": 0.3747423388759104, + "grad_norm": 1.140887975692749, + "learning_rate": 3.935986998965402e-05, + "loss": 0.9967, + "step": 5454 + }, + { + "epoch": 0.374811048509001, + "grad_norm": 1.0531951189041138, + "learning_rate": 3.9354961117487913e-05, + "loss": 1.001, + "step": 5455 + }, + { + "epoch": 0.3748797581420915, + "grad_norm": 1.0094832181930542, + "learning_rate": 3.9350051419489646e-05, + "loss": 0.9325, + "step": 5456 + }, + { + "epoch": 0.3749484677751821, + "grad_norm": 1.0865617990493774, + "learning_rate": 3.934514089594167e-05, + "loss": 0.9615, + "step": 5457 + }, + { + "epoch": 0.3750171774082726, + "grad_norm": 1.1320942640304565, + "learning_rate": 3.93402295471265e-05, + "loss": 0.9506, + "step": 5458 + }, + { + "epoch": 0.3750858870413632, + "grad_norm": 1.0983774662017822, + "learning_rate": 3.933531737332664e-05, + "loss": 0.967, + "step": 5459 + }, + { + "epoch": 0.3751545966744538, + "grad_norm": 1.075051188468933, + "learning_rate": 3.9330404374824733e-05, + "loss": 0.9801, + "step": 5460 + }, + { + "epoch": 0.3752233063075443, + "grad_norm": 1.092058539390564, + "learning_rate": 3.93254905519034e-05, + "loss": 1.0227, + "step": 5461 + }, + { + "epoch": 0.3752920159406349, + "grad_norm": 1.0608515739440918, + "learning_rate": 3.932057590484532e-05, + "loss": 0.996, + "step": 5462 + }, + { + "epoch": 0.37536072557372546, + "grad_norm": 1.1941457986831665, + "learning_rate": 3.931566043393324e-05, + "loss": 0.9885, + "step": 5463 + }, + { + "epoch": 0.375429435206816, + "grad_norm": 1.058799147605896, + "learning_rate": 3.931074413944996e-05, + "loss": 0.9557, + "step": 5464 + }, + { + "epoch": 0.37549814483990657, + "grad_norm": 1.0843806266784668, + "learning_rate": 3.9305827021678276e-05, + "loss": 0.9857, + "step": 5465 + }, + { + "epoch": 0.3755668544729971, + "grad_norm": 1.0919262170791626, + "learning_rate": 3.93009090809011e-05, + "loss": 0.9657, + "step": 5466 + }, + { + "epoch": 0.37563556410608767, + "grad_norm": 1.0999634265899658, + "learning_rate": 3.9295990317401335e-05, + "loss": 0.972, + "step": 5467 + }, + { + "epoch": 0.37570427373917825, + "grad_norm": 1.120017647743225, + "learning_rate": 3.929107073146197e-05, + "loss": 0.9942, + "step": 5468 + }, + { + "epoch": 0.3757729833722688, + "grad_norm": 1.1115822792053223, + "learning_rate": 3.928615032336602e-05, + "loss": 0.9265, + "step": 5469 + }, + { + "epoch": 0.37584169300535936, + "grad_norm": 1.0996865034103394, + "learning_rate": 3.928122909339654e-05, + "loss": 0.9761, + "step": 5470 + }, + { + "epoch": 0.37591040263844994, + "grad_norm": 1.0702660083770752, + "learning_rate": 3.927630704183668e-05, + "loss": 1.0132, + "step": 5471 + }, + { + "epoch": 0.37597911227154046, + "grad_norm": 1.1525394916534424, + "learning_rate": 3.927138416896956e-05, + "loss": 0.9918, + "step": 5472 + }, + { + "epoch": 0.37604782190463104, + "grad_norm": 1.1255327463150024, + "learning_rate": 3.926646047507842e-05, + "loss": 1.022, + "step": 5473 + }, + { + "epoch": 0.37611653153772157, + "grad_norm": 1.1007704734802246, + "learning_rate": 3.9261535960446504e-05, + "loss": 0.9884, + "step": 5474 + }, + { + "epoch": 0.37618524117081215, + "grad_norm": 1.075990915298462, + "learning_rate": 3.925661062535713e-05, + "loss": 1.0304, + "step": 5475 + }, + { + "epoch": 0.3762539508039027, + "grad_norm": 1.0322952270507812, + "learning_rate": 3.925168447009363e-05, + "loss": 0.9471, + "step": 5476 + }, + { + "epoch": 0.37632266043699325, + "grad_norm": 1.1118923425674438, + "learning_rate": 3.924675749493941e-05, + "loss": 1.002, + "step": 5477 + }, + { + "epoch": 0.37639137007008383, + "grad_norm": 1.1857810020446777, + "learning_rate": 3.9241829700177926e-05, + "loss": 1.0004, + "step": 5478 + }, + { + "epoch": 0.3764600797031744, + "grad_norm": 1.0674686431884766, + "learning_rate": 3.923690108609266e-05, + "loss": 0.9583, + "step": 5479 + }, + { + "epoch": 0.37652878933626494, + "grad_norm": 1.0291436910629272, + "learning_rate": 3.923197165296715e-05, + "loss": 0.9902, + "step": 5480 + }, + { + "epoch": 0.3765974989693555, + "grad_norm": 1.0337222814559937, + "learning_rate": 3.9227041401084996e-05, + "loss": 0.9218, + "step": 5481 + }, + { + "epoch": 0.37666620860244604, + "grad_norm": 1.0899581909179688, + "learning_rate": 3.922211033072982e-05, + "loss": 0.9925, + "step": 5482 + }, + { + "epoch": 0.3767349182355366, + "grad_norm": 1.0547640323638916, + "learning_rate": 3.921717844218532e-05, + "loss": 0.9888, + "step": 5483 + }, + { + "epoch": 0.3768036278686272, + "grad_norm": 1.132128357887268, + "learning_rate": 3.92122457357352e-05, + "loss": 0.9756, + "step": 5484 + }, + { + "epoch": 0.3768723375017177, + "grad_norm": 1.173034429550171, + "learning_rate": 3.9207312211663246e-05, + "loss": 1.0281, + "step": 5485 + }, + { + "epoch": 0.3769410471348083, + "grad_norm": 1.090606451034546, + "learning_rate": 3.920237787025329e-05, + "loss": 0.9774, + "step": 5486 + }, + { + "epoch": 0.3770097567678989, + "grad_norm": 1.113521933555603, + "learning_rate": 3.919744271178919e-05, + "loss": 1.0256, + "step": 5487 + }, + { + "epoch": 0.3770784664009894, + "grad_norm": 1.1096783876419067, + "learning_rate": 3.919250673655487e-05, + "loss": 0.9708, + "step": 5488 + }, + { + "epoch": 0.37714717603408, + "grad_norm": 1.0999666452407837, + "learning_rate": 3.91875699448343e-05, + "loss": 0.9721, + "step": 5489 + }, + { + "epoch": 0.3772158856671705, + "grad_norm": 1.0549956560134888, + "learning_rate": 3.918263233691147e-05, + "loss": 0.9542, + "step": 5490 + }, + { + "epoch": 0.3772845953002611, + "grad_norm": 1.0144022703170776, + "learning_rate": 3.917769391307046e-05, + "loss": 0.9332, + "step": 5491 + }, + { + "epoch": 0.3773533049333517, + "grad_norm": 1.1124346256256104, + "learning_rate": 3.917275467359535e-05, + "loss": 1.0043, + "step": 5492 + }, + { + "epoch": 0.3774220145664422, + "grad_norm": 1.1304668188095093, + "learning_rate": 3.916781461877031e-05, + "loss": 1.0334, + "step": 5493 + }, + { + "epoch": 0.3774907241995328, + "grad_norm": 1.0842883586883545, + "learning_rate": 3.916287374887954e-05, + "loss": 0.9829, + "step": 5494 + }, + { + "epoch": 0.37755943383262336, + "grad_norm": 1.0339770317077637, + "learning_rate": 3.915793206420727e-05, + "loss": 0.963, + "step": 5495 + }, + { + "epoch": 0.3776281434657139, + "grad_norm": 1.0954805612564087, + "learning_rate": 3.915298956503781e-05, + "loss": 0.9868, + "step": 5496 + }, + { + "epoch": 0.37769685309880446, + "grad_norm": 1.0280332565307617, + "learning_rate": 3.914804625165548e-05, + "loss": 0.977, + "step": 5497 + }, + { + "epoch": 0.377765562731895, + "grad_norm": 1.1700321435928345, + "learning_rate": 3.9143102124344674e-05, + "loss": 0.9914, + "step": 5498 + }, + { + "epoch": 0.37783427236498557, + "grad_norm": 1.122685432434082, + "learning_rate": 3.913815718338984e-05, + "loss": 1.0109, + "step": 5499 + }, + { + "epoch": 0.37790298199807615, + "grad_norm": 1.0780036449432373, + "learning_rate": 3.913321142907543e-05, + "loss": 0.991, + "step": 5500 + }, + { + "epoch": 0.3779716916311667, + "grad_norm": 1.0919932126998901, + "learning_rate": 3.9128264861686005e-05, + "loss": 1.0231, + "step": 5501 + }, + { + "epoch": 0.37804040126425725, + "grad_norm": 1.0647143125534058, + "learning_rate": 3.9123317481506096e-05, + "loss": 0.9291, + "step": 5502 + }, + { + "epoch": 0.37810911089734783, + "grad_norm": 1.104554295539856, + "learning_rate": 3.911836928882034e-05, + "loss": 0.996, + "step": 5503 + }, + { + "epoch": 0.37817782053043836, + "grad_norm": 1.1892527341842651, + "learning_rate": 3.911342028391343e-05, + "loss": 1.0396, + "step": 5504 + }, + { + "epoch": 0.37824653016352894, + "grad_norm": 1.1137919425964355, + "learning_rate": 3.9108470467070045e-05, + "loss": 1.0003, + "step": 5505 + }, + { + "epoch": 0.37831523979661946, + "grad_norm": 1.0987330675125122, + "learning_rate": 3.910351983857495e-05, + "loss": 0.9714, + "step": 5506 + }, + { + "epoch": 0.37838394942971004, + "grad_norm": 1.1035772562026978, + "learning_rate": 3.909856839871296e-05, + "loss": 0.9737, + "step": 5507 + }, + { + "epoch": 0.3784526590628006, + "grad_norm": 1.1109801530838013, + "learning_rate": 3.9093616147768935e-05, + "loss": 0.9688, + "step": 5508 + }, + { + "epoch": 0.37852136869589115, + "grad_norm": 1.039899230003357, + "learning_rate": 3.9088663086027766e-05, + "loss": 0.9468, + "step": 5509 + }, + { + "epoch": 0.3785900783289817, + "grad_norm": 1.1545213460922241, + "learning_rate": 3.908370921377439e-05, + "loss": 0.986, + "step": 5510 + }, + { + "epoch": 0.3786587879620723, + "grad_norm": 1.1167323589324951, + "learning_rate": 3.9078754531293823e-05, + "loss": 1.016, + "step": 5511 + }, + { + "epoch": 0.37872749759516283, + "grad_norm": 1.0573707818984985, + "learning_rate": 3.907379903887108e-05, + "loss": 0.9582, + "step": 5512 + }, + { + "epoch": 0.3787962072282534, + "grad_norm": 1.0822904109954834, + "learning_rate": 3.9068842736791266e-05, + "loss": 0.9674, + "step": 5513 + }, + { + "epoch": 0.37886491686134394, + "grad_norm": 1.016116976737976, + "learning_rate": 3.9063885625339505e-05, + "loss": 0.9719, + "step": 5514 + }, + { + "epoch": 0.3789336264944345, + "grad_norm": 1.1274278163909912, + "learning_rate": 3.905892770480098e-05, + "loss": 1.0167, + "step": 5515 + }, + { + "epoch": 0.3790023361275251, + "grad_norm": 1.0670660734176636, + "learning_rate": 3.905396897546092e-05, + "loss": 0.9663, + "step": 5516 + }, + { + "epoch": 0.3790710457606156, + "grad_norm": 1.1197783946990967, + "learning_rate": 3.9049009437604586e-05, + "loss": 1.0149, + "step": 5517 + }, + { + "epoch": 0.3791397553937062, + "grad_norm": 1.0930112600326538, + "learning_rate": 3.904404909151731e-05, + "loss": 0.9787, + "step": 5518 + }, + { + "epoch": 0.3792084650267968, + "grad_norm": 1.1183120012283325, + "learning_rate": 3.903908793748445e-05, + "loss": 1.0117, + "step": 5519 + }, + { + "epoch": 0.3792771746598873, + "grad_norm": 1.1326152086257935, + "learning_rate": 3.903412597579142e-05, + "loss": 0.9641, + "step": 5520 + }, + { + "epoch": 0.3793458842929779, + "grad_norm": 1.1315248012542725, + "learning_rate": 3.9029163206723684e-05, + "loss": 1.0464, + "step": 5521 + }, + { + "epoch": 0.3794145939260684, + "grad_norm": 1.0469330549240112, + "learning_rate": 3.902419963056675e-05, + "loss": 1.0017, + "step": 5522 + }, + { + "epoch": 0.379483303559159, + "grad_norm": 1.0800776481628418, + "learning_rate": 3.901923524760616e-05, + "loss": 0.9798, + "step": 5523 + }, + { + "epoch": 0.37955201319224957, + "grad_norm": 1.0338501930236816, + "learning_rate": 3.90142700581275e-05, + "loss": 0.9822, + "step": 5524 + }, + { + "epoch": 0.3796207228253401, + "grad_norm": 1.049815058708191, + "learning_rate": 3.900930406241644e-05, + "loss": 0.9846, + "step": 5525 + }, + { + "epoch": 0.3796894324584307, + "grad_norm": 1.067252516746521, + "learning_rate": 3.900433726075865e-05, + "loss": 0.9777, + "step": 5526 + }, + { + "epoch": 0.37975814209152126, + "grad_norm": 1.0859345197677612, + "learning_rate": 3.899936965343989e-05, + "loss": 0.9816, + "step": 5527 + }, + { + "epoch": 0.3798268517246118, + "grad_norm": 1.061845302581787, + "learning_rate": 3.8994401240745923e-05, + "loss": 0.987, + "step": 5528 + }, + { + "epoch": 0.37989556135770236, + "grad_norm": 1.1229759454727173, + "learning_rate": 3.898943202296258e-05, + "loss": 1.0062, + "step": 5529 + }, + { + "epoch": 0.3799642709907929, + "grad_norm": 1.1279900074005127, + "learning_rate": 3.8984462000375746e-05, + "loss": 0.9556, + "step": 5530 + }, + { + "epoch": 0.38003298062388347, + "grad_norm": 1.113208293914795, + "learning_rate": 3.897949117327134e-05, + "loss": 1.0106, + "step": 5531 + }, + { + "epoch": 0.38010169025697405, + "grad_norm": 1.132212519645691, + "learning_rate": 3.897451954193534e-05, + "loss": 0.9942, + "step": 5532 + }, + { + "epoch": 0.38017039989006457, + "grad_norm": 1.1754510402679443, + "learning_rate": 3.8969547106653736e-05, + "loss": 1.0087, + "step": 5533 + }, + { + "epoch": 0.38023910952315515, + "grad_norm": 1.0501375198364258, + "learning_rate": 3.8964573867712615e-05, + "loss": 0.9534, + "step": 5534 + }, + { + "epoch": 0.38030781915624573, + "grad_norm": 1.186414361000061, + "learning_rate": 3.8959599825398075e-05, + "loss": 1.0041, + "step": 5535 + }, + { + "epoch": 0.38037652878933625, + "grad_norm": 1.0925910472869873, + "learning_rate": 3.8954624979996256e-05, + "loss": 0.9914, + "step": 5536 + }, + { + "epoch": 0.38044523842242683, + "grad_norm": 1.084606647491455, + "learning_rate": 3.894964933179339e-05, + "loss": 1.0021, + "step": 5537 + }, + { + "epoch": 0.38051394805551736, + "grad_norm": 1.0615266561508179, + "learning_rate": 3.8944672881075686e-05, + "loss": 0.9572, + "step": 5538 + }, + { + "epoch": 0.38058265768860794, + "grad_norm": 1.1176384687423706, + "learning_rate": 3.893969562812946e-05, + "loss": 1.0059, + "step": 5539 + }, + { + "epoch": 0.3806513673216985, + "grad_norm": 1.0925004482269287, + "learning_rate": 3.8934717573241045e-05, + "loss": 1.0012, + "step": 5540 + }, + { + "epoch": 0.38072007695478904, + "grad_norm": 1.084755778312683, + "learning_rate": 3.892973871669682e-05, + "loss": 0.9954, + "step": 5541 + }, + { + "epoch": 0.3807887865878796, + "grad_norm": 1.173672080039978, + "learning_rate": 3.892475905878323e-05, + "loss": 0.9845, + "step": 5542 + }, + { + "epoch": 0.3808574962209702, + "grad_norm": 1.025591492652893, + "learning_rate": 3.891977859978673e-05, + "loss": 0.9918, + "step": 5543 + }, + { + "epoch": 0.38092620585406073, + "grad_norm": 1.0986157655715942, + "learning_rate": 3.8914797339993855e-05, + "loss": 1.0056, + "step": 5544 + }, + { + "epoch": 0.3809949154871513, + "grad_norm": 1.108893632888794, + "learning_rate": 3.8909815279691186e-05, + "loss": 1.0301, + "step": 5545 + }, + { + "epoch": 0.38106362512024183, + "grad_norm": 1.0335181951522827, + "learning_rate": 3.8904832419165314e-05, + "loss": 1.0013, + "step": 5546 + }, + { + "epoch": 0.3811323347533324, + "grad_norm": 1.1210073232650757, + "learning_rate": 3.8899848758702916e-05, + "loss": 1.0055, + "step": 5547 + }, + { + "epoch": 0.381201044386423, + "grad_norm": 1.0893235206604004, + "learning_rate": 3.8894864298590695e-05, + "loss": 0.9791, + "step": 5548 + }, + { + "epoch": 0.3812697540195135, + "grad_norm": 1.1363742351531982, + "learning_rate": 3.88898790391154e-05, + "loss": 0.9316, + "step": 5549 + }, + { + "epoch": 0.3813384636526041, + "grad_norm": 1.059350609779358, + "learning_rate": 3.888489298056384e-05, + "loss": 0.9858, + "step": 5550 + }, + { + "epoch": 0.3814071732856947, + "grad_norm": 1.0838184356689453, + "learning_rate": 3.887990612322284e-05, + "loss": 0.9633, + "step": 5551 + }, + { + "epoch": 0.3814758829187852, + "grad_norm": 1.1263841390609741, + "learning_rate": 3.8874918467379315e-05, + "loss": 1.0412, + "step": 5552 + }, + { + "epoch": 0.3815445925518758, + "grad_norm": 1.1053980588912964, + "learning_rate": 3.886993001332019e-05, + "loss": 1.0233, + "step": 5553 + }, + { + "epoch": 0.3816133021849663, + "grad_norm": 1.064596176147461, + "learning_rate": 3.8864940761332434e-05, + "loss": 0.9729, + "step": 5554 + }, + { + "epoch": 0.3816820118180569, + "grad_norm": 1.1734148263931274, + "learning_rate": 3.885995071170311e-05, + "loss": 1.0084, + "step": 5555 + }, + { + "epoch": 0.38175072145114747, + "grad_norm": 1.1661723852157593, + "learning_rate": 3.885495986471926e-05, + "loss": 0.9837, + "step": 5556 + }, + { + "epoch": 0.381819431084238, + "grad_norm": 1.0658705234527588, + "learning_rate": 3.884996822066802e-05, + "loss": 0.9911, + "step": 5557 + }, + { + "epoch": 0.3818881407173286, + "grad_norm": 1.1348721981048584, + "learning_rate": 3.884497577983655e-05, + "loss": 0.9772, + "step": 5558 + }, + { + "epoch": 0.38195685035041915, + "grad_norm": 1.016069769859314, + "learning_rate": 3.8839982542512065e-05, + "loss": 0.9468, + "step": 5559 + }, + { + "epoch": 0.3820255599835097, + "grad_norm": 1.0761911869049072, + "learning_rate": 3.883498850898183e-05, + "loss": 0.9785, + "step": 5560 + }, + { + "epoch": 0.38209426961660026, + "grad_norm": 1.059739112854004, + "learning_rate": 3.882999367953314e-05, + "loss": 0.9533, + "step": 5561 + }, + { + "epoch": 0.3821629792496908, + "grad_norm": 1.0593022108078003, + "learning_rate": 3.8824998054453346e-05, + "loss": 0.9765, + "step": 5562 + }, + { + "epoch": 0.38223168888278136, + "grad_norm": 1.134031057357788, + "learning_rate": 3.882000163402984e-05, + "loss": 0.9674, + "step": 5563 + }, + { + "epoch": 0.38230039851587194, + "grad_norm": 1.0913571119308472, + "learning_rate": 3.881500441855006e-05, + "loss": 1.0021, + "step": 5564 + }, + { + "epoch": 0.38236910814896247, + "grad_norm": 1.1024694442749023, + "learning_rate": 3.881000640830151e-05, + "loss": 1.0146, + "step": 5565 + }, + { + "epoch": 0.38243781778205305, + "grad_norm": 1.099445104598999, + "learning_rate": 3.880500760357171e-05, + "loss": 0.9798, + "step": 5566 + }, + { + "epoch": 0.3825065274151436, + "grad_norm": 1.091576337814331, + "learning_rate": 3.880000800464824e-05, + "loss": 0.9642, + "step": 5567 + }, + { + "epoch": 0.38257523704823415, + "grad_norm": 1.1303130388259888, + "learning_rate": 3.8795007611818724e-05, + "loss": 0.9839, + "step": 5568 + }, + { + "epoch": 0.38264394668132473, + "grad_norm": 1.0396980047225952, + "learning_rate": 3.879000642537083e-05, + "loss": 0.9576, + "step": 5569 + }, + { + "epoch": 0.38271265631441526, + "grad_norm": 1.2005733251571655, + "learning_rate": 3.878500444559228e-05, + "loss": 1.0144, + "step": 5570 + }, + { + "epoch": 0.38278136594750584, + "grad_norm": 1.1183348894119263, + "learning_rate": 3.878000167277081e-05, + "loss": 0.9558, + "step": 5571 + }, + { + "epoch": 0.3828500755805964, + "grad_norm": 1.0711731910705566, + "learning_rate": 3.8774998107194264e-05, + "loss": 0.9617, + "step": 5572 + }, + { + "epoch": 0.38291878521368694, + "grad_norm": 1.112058401107788, + "learning_rate": 3.8769993749150474e-05, + "loss": 1.0124, + "step": 5573 + }, + { + "epoch": 0.3829874948467775, + "grad_norm": 1.0372811555862427, + "learning_rate": 3.876498859892734e-05, + "loss": 0.9728, + "step": 5574 + }, + { + "epoch": 0.3830562044798681, + "grad_norm": 1.1157848834991455, + "learning_rate": 3.87599826568128e-05, + "loss": 1.019, + "step": 5575 + }, + { + "epoch": 0.3831249141129586, + "grad_norm": 1.1206037998199463, + "learning_rate": 3.875497592309485e-05, + "loss": 0.9982, + "step": 5576 + }, + { + "epoch": 0.3831936237460492, + "grad_norm": 1.1563348770141602, + "learning_rate": 3.874996839806152e-05, + "loss": 0.949, + "step": 5577 + }, + { + "epoch": 0.38326233337913973, + "grad_norm": 1.1212838888168335, + "learning_rate": 3.87449600820009e-05, + "loss": 0.9683, + "step": 5578 + }, + { + "epoch": 0.3833310430122303, + "grad_norm": 1.0793803930282593, + "learning_rate": 3.8739950975201096e-05, + "loss": 1.0026, + "step": 5579 + }, + { + "epoch": 0.3833997526453209, + "grad_norm": 1.1199443340301514, + "learning_rate": 3.87349410779503e-05, + "loss": 1.0023, + "step": 5580 + }, + { + "epoch": 0.3834684622784114, + "grad_norm": 1.0745865106582642, + "learning_rate": 3.872993039053671e-05, + "loss": 1.0009, + "step": 5581 + }, + { + "epoch": 0.383537171911502, + "grad_norm": 1.0670493841171265, + "learning_rate": 3.87249189132486e-05, + "loss": 0.9171, + "step": 5582 + }, + { + "epoch": 0.3836058815445926, + "grad_norm": 1.1139715909957886, + "learning_rate": 3.871990664637427e-05, + "loss": 1.0247, + "step": 5583 + }, + { + "epoch": 0.3836745911776831, + "grad_norm": 1.1298879384994507, + "learning_rate": 3.871489359020209e-05, + "loss": 1.0343, + "step": 5584 + }, + { + "epoch": 0.3837433008107737, + "grad_norm": 1.0709081888198853, + "learning_rate": 3.870987974502043e-05, + "loss": 1.017, + "step": 5585 + }, + { + "epoch": 0.3838120104438642, + "grad_norm": 1.1238380670547485, + "learning_rate": 3.8704865111117746e-05, + "loss": 1.0079, + "step": 5586 + }, + { + "epoch": 0.3838807200769548, + "grad_norm": 1.111555814743042, + "learning_rate": 3.869984968878253e-05, + "loss": 0.9815, + "step": 5587 + }, + { + "epoch": 0.38394942971004536, + "grad_norm": 1.0528141260147095, + "learning_rate": 3.869483347830333e-05, + "loss": 0.9939, + "step": 5588 + }, + { + "epoch": 0.3840181393431359, + "grad_norm": 1.0914226770401, + "learning_rate": 3.8689816479968686e-05, + "loss": 1.008, + "step": 5589 + }, + { + "epoch": 0.38408684897622647, + "grad_norm": 1.0337122678756714, + "learning_rate": 3.868479869406727e-05, + "loss": 0.9787, + "step": 5590 + }, + { + "epoch": 0.38415555860931705, + "grad_norm": 1.0717660188674927, + "learning_rate": 3.8679780120887716e-05, + "loss": 1.0167, + "step": 5591 + }, + { + "epoch": 0.3842242682424076, + "grad_norm": 1.1491254568099976, + "learning_rate": 3.8674760760718753e-05, + "loss": 1.0273, + "step": 5592 + }, + { + "epoch": 0.38429297787549815, + "grad_norm": 1.074352741241455, + "learning_rate": 3.866974061384915e-05, + "loss": 1.0019, + "step": 5593 + }, + { + "epoch": 0.3843616875085887, + "grad_norm": 1.109535813331604, + "learning_rate": 3.86647196805677e-05, + "loss": 0.9591, + "step": 5594 + }, + { + "epoch": 0.38443039714167926, + "grad_norm": 1.0856236219406128, + "learning_rate": 3.8659697961163254e-05, + "loss": 0.9283, + "step": 5595 + }, + { + "epoch": 0.38449910677476984, + "grad_norm": 1.101631999015808, + "learning_rate": 3.865467545592472e-05, + "loss": 1.0071, + "step": 5596 + }, + { + "epoch": 0.38456781640786036, + "grad_norm": 1.0073438882827759, + "learning_rate": 3.864965216514101e-05, + "loss": 0.9505, + "step": 5597 + }, + { + "epoch": 0.38463652604095094, + "grad_norm": 1.1098411083221436, + "learning_rate": 3.864462808910117e-05, + "loss": 0.9805, + "step": 5598 + }, + { + "epoch": 0.3847052356740415, + "grad_norm": 1.0849436521530151, + "learning_rate": 3.8639603228094165e-05, + "loss": 1.0103, + "step": 5599 + }, + { + "epoch": 0.38477394530713205, + "grad_norm": 1.1050077676773071, + "learning_rate": 3.863457758240912e-05, + "loss": 1.0224, + "step": 5600 + }, + { + "epoch": 0.38484265494022263, + "grad_norm": 1.0883253812789917, + "learning_rate": 3.862955115233514e-05, + "loss": 0.942, + "step": 5601 + }, + { + "epoch": 0.38491136457331315, + "grad_norm": 1.1536240577697754, + "learning_rate": 3.8624523938161386e-05, + "loss": 1.0173, + "step": 5602 + }, + { + "epoch": 0.38498007420640373, + "grad_norm": 1.1141860485076904, + "learning_rate": 3.861949594017708e-05, + "loss": 0.9748, + "step": 5603 + }, + { + "epoch": 0.3850487838394943, + "grad_norm": 1.046034812927246, + "learning_rate": 3.8614467158671474e-05, + "loss": 0.9983, + "step": 5604 + }, + { + "epoch": 0.38511749347258484, + "grad_norm": 1.1051552295684814, + "learning_rate": 3.860943759393387e-05, + "loss": 0.9756, + "step": 5605 + }, + { + "epoch": 0.3851862031056754, + "grad_norm": 1.1498479843139648, + "learning_rate": 3.860440724625364e-05, + "loss": 1.0092, + "step": 5606 + }, + { + "epoch": 0.385254912738766, + "grad_norm": 1.1035650968551636, + "learning_rate": 3.8599376115920134e-05, + "loss": 0.9766, + "step": 5607 + }, + { + "epoch": 0.3853236223718565, + "grad_norm": 1.150899887084961, + "learning_rate": 3.859434420322282e-05, + "loss": 0.9686, + "step": 5608 + }, + { + "epoch": 0.3853923320049471, + "grad_norm": 1.0526421070098877, + "learning_rate": 3.858931150845117e-05, + "loss": 0.988, + "step": 5609 + }, + { + "epoch": 0.3854610416380376, + "grad_norm": 1.0493223667144775, + "learning_rate": 3.8584278031894716e-05, + "loss": 0.9554, + "step": 5610 + }, + { + "epoch": 0.3855297512711282, + "grad_norm": 1.1323914527893066, + "learning_rate": 3.8579243773843035e-05, + "loss": 1.0329, + "step": 5611 + }, + { + "epoch": 0.3855984609042188, + "grad_norm": 1.1903271675109863, + "learning_rate": 3.8574208734585735e-05, + "loss": 1.0369, + "step": 5612 + }, + { + "epoch": 0.3856671705373093, + "grad_norm": 1.1243685483932495, + "learning_rate": 3.856917291441249e-05, + "loss": 1.0175, + "step": 5613 + }, + { + "epoch": 0.3857358801703999, + "grad_norm": 1.0774121284484863, + "learning_rate": 3.8564136313612984e-05, + "loss": 0.9616, + "step": 5614 + }, + { + "epoch": 0.38580458980349047, + "grad_norm": 1.0458035469055176, + "learning_rate": 3.8559098932477e-05, + "loss": 0.9931, + "step": 5615 + }, + { + "epoch": 0.385873299436581, + "grad_norm": 1.0930330753326416, + "learning_rate": 3.855406077129432e-05, + "loss": 0.9777, + "step": 5616 + }, + { + "epoch": 0.3859420090696716, + "grad_norm": 1.0920132398605347, + "learning_rate": 3.8549021830354784e-05, + "loss": 0.9578, + "step": 5617 + }, + { + "epoch": 0.3860107187027621, + "grad_norm": 1.1373833417892456, + "learning_rate": 3.8543982109948285e-05, + "loss": 0.968, + "step": 5618 + }, + { + "epoch": 0.3860794283358527, + "grad_norm": 1.1047126054763794, + "learning_rate": 3.8538941610364756e-05, + "loss": 0.9847, + "step": 5619 + }, + { + "epoch": 0.38614813796894326, + "grad_norm": 1.0271366834640503, + "learning_rate": 3.853390033189417e-05, + "loss": 0.9513, + "step": 5620 + }, + { + "epoch": 0.3862168476020338, + "grad_norm": 1.0876086950302124, + "learning_rate": 3.852885827482655e-05, + "loss": 0.9529, + "step": 5621 + }, + { + "epoch": 0.38628555723512437, + "grad_norm": 1.048245906829834, + "learning_rate": 3.852381543945196e-05, + "loss": 0.9743, + "step": 5622 + }, + { + "epoch": 0.38635426686821495, + "grad_norm": 1.0699949264526367, + "learning_rate": 3.8518771826060515e-05, + "loss": 0.9531, + "step": 5623 + }, + { + "epoch": 0.38642297650130547, + "grad_norm": 1.1266541481018066, + "learning_rate": 3.851372743494237e-05, + "loss": 1.0042, + "step": 5624 + }, + { + "epoch": 0.38649168613439605, + "grad_norm": 1.0364165306091309, + "learning_rate": 3.850868226638773e-05, + "loss": 0.9416, + "step": 5625 + }, + { + "epoch": 0.3865603957674866, + "grad_norm": 1.1523932218551636, + "learning_rate": 3.850363632068683e-05, + "loss": 1.0279, + "step": 5626 + }, + { + "epoch": 0.38662910540057716, + "grad_norm": 1.1152054071426392, + "learning_rate": 3.849858959812997e-05, + "loss": 1.0091, + "step": 5627 + }, + { + "epoch": 0.38669781503366774, + "grad_norm": 1.1420842409133911, + "learning_rate": 3.8493542099007474e-05, + "loss": 1.0058, + "step": 5628 + }, + { + "epoch": 0.38676652466675826, + "grad_norm": 1.1500840187072754, + "learning_rate": 3.848849382360974e-05, + "loss": 1.0068, + "step": 5629 + }, + { + "epoch": 0.38683523429984884, + "grad_norm": 1.137353777885437, + "learning_rate": 3.848344477222717e-05, + "loss": 0.9962, + "step": 5630 + }, + { + "epoch": 0.3869039439329394, + "grad_norm": 1.089689016342163, + "learning_rate": 3.8478394945150246e-05, + "loss": 0.9927, + "step": 5631 + }, + { + "epoch": 0.38697265356602994, + "grad_norm": 1.105492353439331, + "learning_rate": 3.847334434266948e-05, + "loss": 0.9632, + "step": 5632 + }, + { + "epoch": 0.3870413631991205, + "grad_norm": 1.1250905990600586, + "learning_rate": 3.846829296507543e-05, + "loss": 1.0107, + "step": 5633 + }, + { + "epoch": 0.38711007283221105, + "grad_norm": 1.1114941835403442, + "learning_rate": 3.8463240812658694e-05, + "loss": 0.9949, + "step": 5634 + }, + { + "epoch": 0.38717878246530163, + "grad_norm": 1.0324835777282715, + "learning_rate": 3.845818788570993e-05, + "loss": 0.9796, + "step": 5635 + }, + { + "epoch": 0.3872474920983922, + "grad_norm": 1.0939141511917114, + "learning_rate": 3.8453134184519825e-05, + "loss": 0.9797, + "step": 5636 + }, + { + "epoch": 0.38731620173148273, + "grad_norm": 1.0878592729568481, + "learning_rate": 3.84480797093791e-05, + "loss": 0.9637, + "step": 5637 + }, + { + "epoch": 0.3873849113645733, + "grad_norm": 1.1379303932189941, + "learning_rate": 3.844302446057856e-05, + "loss": 1.0081, + "step": 5638 + }, + { + "epoch": 0.3874536209976639, + "grad_norm": 1.1696242094039917, + "learning_rate": 3.843796843840902e-05, + "loss": 1.0614, + "step": 5639 + }, + { + "epoch": 0.3875223306307544, + "grad_norm": 1.0945446491241455, + "learning_rate": 3.843291164316134e-05, + "loss": 1.0037, + "step": 5640 + }, + { + "epoch": 0.387591040263845, + "grad_norm": 1.078055500984192, + "learning_rate": 3.842785407512645e-05, + "loss": 0.9758, + "step": 5641 + }, + { + "epoch": 0.3876597498969355, + "grad_norm": 1.036315679550171, + "learning_rate": 3.84227957345953e-05, + "loss": 0.9597, + "step": 5642 + }, + { + "epoch": 0.3877284595300261, + "grad_norm": 1.1169334650039673, + "learning_rate": 3.8417736621858896e-05, + "loss": 0.9699, + "step": 5643 + }, + { + "epoch": 0.3877971691631167, + "grad_norm": 1.1241955757141113, + "learning_rate": 3.841267673720829e-05, + "loss": 0.9983, + "step": 5644 + }, + { + "epoch": 0.3878658787962072, + "grad_norm": 1.131903052330017, + "learning_rate": 3.840761608093456e-05, + "loss": 1.0007, + "step": 5645 + }, + { + "epoch": 0.3879345884292978, + "grad_norm": 1.1048929691314697, + "learning_rate": 3.840255465332886e-05, + "loss": 0.9953, + "step": 5646 + }, + { + "epoch": 0.38800329806238837, + "grad_norm": 1.191329002380371, + "learning_rate": 3.839749245468236e-05, + "loss": 1.0188, + "step": 5647 + }, + { + "epoch": 0.3880720076954789, + "grad_norm": 1.0559113025665283, + "learning_rate": 3.8392429485286285e-05, + "loss": 0.9698, + "step": 5648 + }, + { + "epoch": 0.3881407173285695, + "grad_norm": 1.1380003690719604, + "learning_rate": 3.8387365745431917e-05, + "loss": 1.0273, + "step": 5649 + }, + { + "epoch": 0.38820942696166, + "grad_norm": 1.0818957090377808, + "learning_rate": 3.838230123541056e-05, + "loss": 1.0182, + "step": 5650 + }, + { + "epoch": 0.3882781365947506, + "grad_norm": 1.061636209487915, + "learning_rate": 3.8377235955513566e-05, + "loss": 0.9785, + "step": 5651 + }, + { + "epoch": 0.38834684622784116, + "grad_norm": 1.092851996421814, + "learning_rate": 3.837216990603234e-05, + "loss": 0.98, + "step": 5652 + }, + { + "epoch": 0.3884155558609317, + "grad_norm": 1.0638821125030518, + "learning_rate": 3.836710308725834e-05, + "loss": 0.9302, + "step": 5653 + }, + { + "epoch": 0.38848426549402226, + "grad_norm": 1.1624352931976318, + "learning_rate": 3.836203549948305e-05, + "loss": 1.0095, + "step": 5654 + }, + { + "epoch": 0.38855297512711284, + "grad_norm": 1.0741020441055298, + "learning_rate": 3.8356967142998e-05, + "loss": 0.9854, + "step": 5655 + }, + { + "epoch": 0.38862168476020337, + "grad_norm": 1.0887985229492188, + "learning_rate": 3.835189801809478e-05, + "loss": 1.0202, + "step": 5656 + }, + { + "epoch": 0.38869039439329395, + "grad_norm": 1.0627368688583374, + "learning_rate": 3.8346828125065004e-05, + "loss": 0.9929, + "step": 5657 + }, + { + "epoch": 0.38875910402638447, + "grad_norm": 1.1624583005905151, + "learning_rate": 3.834175746420034e-05, + "loss": 0.987, + "step": 5658 + }, + { + "epoch": 0.38882781365947505, + "grad_norm": 1.1255810260772705, + "learning_rate": 3.833668603579251e-05, + "loss": 0.9826, + "step": 5659 + }, + { + "epoch": 0.38889652329256563, + "grad_norm": 1.1069749593734741, + "learning_rate": 3.833161384013326e-05, + "loss": 0.9421, + "step": 5660 + }, + { + "epoch": 0.38896523292565616, + "grad_norm": 1.079607367515564, + "learning_rate": 3.8326540877514406e-05, + "loss": 0.9644, + "step": 5661 + }, + { + "epoch": 0.38903394255874674, + "grad_norm": 1.0997358560562134, + "learning_rate": 3.8321467148227776e-05, + "loss": 0.9961, + "step": 5662 + }, + { + "epoch": 0.3891026521918373, + "grad_norm": 1.1171740293502808, + "learning_rate": 3.831639265256526e-05, + "loss": 0.969, + "step": 5663 + }, + { + "epoch": 0.38917136182492784, + "grad_norm": 1.0898092985153198, + "learning_rate": 3.831131739081879e-05, + "loss": 0.9798, + "step": 5664 + }, + { + "epoch": 0.3892400714580184, + "grad_norm": 1.1490517854690552, + "learning_rate": 3.830624136328036e-05, + "loss": 1.009, + "step": 5665 + }, + { + "epoch": 0.38930878109110895, + "grad_norm": 1.1233901977539062, + "learning_rate": 3.830116457024196e-05, + "loss": 1.0268, + "step": 5666 + }, + { + "epoch": 0.3893774907241995, + "grad_norm": 1.0976533889770508, + "learning_rate": 3.8296087011995686e-05, + "loss": 1.0213, + "step": 5667 + }, + { + "epoch": 0.3894462003572901, + "grad_norm": 1.0825155973434448, + "learning_rate": 3.8291008688833625e-05, + "loss": 0.9568, + "step": 5668 + }, + { + "epoch": 0.38951490999038063, + "grad_norm": 1.0903940200805664, + "learning_rate": 3.8285929601047944e-05, + "loss": 0.9484, + "step": 5669 + }, + { + "epoch": 0.3895836196234712, + "grad_norm": 1.1894041299819946, + "learning_rate": 3.8280849748930826e-05, + "loss": 1.0065, + "step": 5670 + }, + { + "epoch": 0.3896523292565618, + "grad_norm": 1.0686790943145752, + "learning_rate": 3.827576913277453e-05, + "loss": 0.9917, + "step": 5671 + }, + { + "epoch": 0.3897210388896523, + "grad_norm": 1.0499036312103271, + "learning_rate": 3.827068775287133e-05, + "loss": 0.9649, + "step": 5672 + }, + { + "epoch": 0.3897897485227429, + "grad_norm": 1.1400930881500244, + "learning_rate": 3.826560560951355e-05, + "loss": 1.0106, + "step": 5673 + }, + { + "epoch": 0.3898584581558334, + "grad_norm": 1.1243349313735962, + "learning_rate": 3.826052270299356e-05, + "loss": 0.9714, + "step": 5674 + }, + { + "epoch": 0.389927167788924, + "grad_norm": 1.1146824359893799, + "learning_rate": 3.82554390336038e-05, + "loss": 0.9986, + "step": 5675 + }, + { + "epoch": 0.3899958774220146, + "grad_norm": 1.039708137512207, + "learning_rate": 3.8250354601636704e-05, + "loss": 0.9981, + "step": 5676 + }, + { + "epoch": 0.3900645870551051, + "grad_norm": 1.1424564123153687, + "learning_rate": 3.82452694073848e-05, + "loss": 0.9694, + "step": 5677 + }, + { + "epoch": 0.3901332966881957, + "grad_norm": 1.1246552467346191, + "learning_rate": 3.824018345114061e-05, + "loss": 1.0274, + "step": 5678 + }, + { + "epoch": 0.39020200632128627, + "grad_norm": 1.0528095960617065, + "learning_rate": 3.823509673319674e-05, + "loss": 0.976, + "step": 5679 + }, + { + "epoch": 0.3902707159543768, + "grad_norm": 1.0414913892745972, + "learning_rate": 3.823000925384583e-05, + "loss": 0.9914, + "step": 5680 + }, + { + "epoch": 0.39033942558746737, + "grad_norm": 1.0630582571029663, + "learning_rate": 3.8224921013380544e-05, + "loss": 0.9783, + "step": 5681 + }, + { + "epoch": 0.3904081352205579, + "grad_norm": 1.1065373420715332, + "learning_rate": 3.821983201209363e-05, + "loss": 1.0278, + "step": 5682 + }, + { + "epoch": 0.3904768448536485, + "grad_norm": 1.1483781337738037, + "learning_rate": 3.821474225027783e-05, + "loss": 1.0129, + "step": 5683 + }, + { + "epoch": 0.39054555448673905, + "grad_norm": 1.1003226041793823, + "learning_rate": 3.820965172822597e-05, + "loss": 0.9473, + "step": 5684 + }, + { + "epoch": 0.3906142641198296, + "grad_norm": 1.0768723487854004, + "learning_rate": 3.82045604462309e-05, + "loss": 1.0102, + "step": 5685 + }, + { + "epoch": 0.39068297375292016, + "grad_norm": 1.07964289188385, + "learning_rate": 3.819946840458552e-05, + "loss": 0.9843, + "step": 5686 + }, + { + "epoch": 0.39075168338601074, + "grad_norm": 1.0646305084228516, + "learning_rate": 3.8194375603582776e-05, + "loss": 0.9846, + "step": 5687 + }, + { + "epoch": 0.39082039301910126, + "grad_norm": 1.1399022340774536, + "learning_rate": 3.818928204351564e-05, + "loss": 1.0355, + "step": 5688 + }, + { + "epoch": 0.39088910265219184, + "grad_norm": 1.063452124595642, + "learning_rate": 3.818418772467716e-05, + "loss": 1.0187, + "step": 5689 + }, + { + "epoch": 0.39095781228528237, + "grad_norm": 1.1169315576553345, + "learning_rate": 3.817909264736039e-05, + "loss": 1.0446, + "step": 5690 + }, + { + "epoch": 0.39102652191837295, + "grad_norm": 1.0885385274887085, + "learning_rate": 3.8173996811858454e-05, + "loss": 0.9826, + "step": 5691 + }, + { + "epoch": 0.39109523155146353, + "grad_norm": 1.091800570487976, + "learning_rate": 3.816890021846452e-05, + "loss": 0.9834, + "step": 5692 + }, + { + "epoch": 0.39116394118455405, + "grad_norm": 1.113247036933899, + "learning_rate": 3.816380286747178e-05, + "loss": 0.9651, + "step": 5693 + }, + { + "epoch": 0.39123265081764463, + "grad_norm": 1.124884009361267, + "learning_rate": 3.8158704759173495e-05, + "loss": 0.9921, + "step": 5694 + }, + { + "epoch": 0.3913013604507352, + "grad_norm": 1.1186156272888184, + "learning_rate": 3.815360589386294e-05, + "loss": 0.9805, + "step": 5695 + }, + { + "epoch": 0.39137007008382574, + "grad_norm": 1.1232470273971558, + "learning_rate": 3.8148506271833465e-05, + "loss": 0.9849, + "step": 5696 + }, + { + "epoch": 0.3914387797169163, + "grad_norm": 1.0651450157165527, + "learning_rate": 3.814340589337844e-05, + "loss": 0.9611, + "step": 5697 + }, + { + "epoch": 0.3915074893500069, + "grad_norm": 1.0818992853164673, + "learning_rate": 3.813830475879129e-05, + "loss": 0.9836, + "step": 5698 + }, + { + "epoch": 0.3915761989830974, + "grad_norm": 1.0819411277770996, + "learning_rate": 3.8133202868365474e-05, + "loss": 0.9865, + "step": 5699 + }, + { + "epoch": 0.391644908616188, + "grad_norm": 1.0772135257720947, + "learning_rate": 3.812810022239451e-05, + "loss": 0.9864, + "step": 5700 + }, + { + "epoch": 0.39171361824927853, + "grad_norm": 1.1848393678665161, + "learning_rate": 3.812299682117194e-05, + "loss": 0.9684, + "step": 5701 + }, + { + "epoch": 0.3917823278823691, + "grad_norm": 1.0869873762130737, + "learning_rate": 3.811789266499136e-05, + "loss": 1.0, + "step": 5702 + }, + { + "epoch": 0.3918510375154597, + "grad_norm": 1.1256242990493774, + "learning_rate": 3.8112787754146416e-05, + "loss": 1.0084, + "step": 5703 + }, + { + "epoch": 0.3919197471485502, + "grad_norm": 1.0875388383865356, + "learning_rate": 3.8107682088930794e-05, + "loss": 1.0243, + "step": 5704 + }, + { + "epoch": 0.3919884567816408, + "grad_norm": 1.1133211851119995, + "learning_rate": 3.810257566963821e-05, + "loss": 0.9635, + "step": 5705 + }, + { + "epoch": 0.3920571664147314, + "grad_norm": 1.1518645286560059, + "learning_rate": 3.809746849656244e-05, + "loss": 0.9677, + "step": 5706 + }, + { + "epoch": 0.3921258760478219, + "grad_norm": 1.0861876010894775, + "learning_rate": 3.809236056999729e-05, + "loss": 0.989, + "step": 5707 + }, + { + "epoch": 0.3921945856809125, + "grad_norm": 1.078652262687683, + "learning_rate": 3.808725189023662e-05, + "loss": 0.9884, + "step": 5708 + }, + { + "epoch": 0.392263295314003, + "grad_norm": 1.1109918355941772, + "learning_rate": 3.808214245757433e-05, + "loss": 1.0085, + "step": 5709 + }, + { + "epoch": 0.3923320049470936, + "grad_norm": 1.0590310096740723, + "learning_rate": 3.807703227230435e-05, + "loss": 0.9671, + "step": 5710 + }, + { + "epoch": 0.39240071458018416, + "grad_norm": 1.1398009061813354, + "learning_rate": 3.8071921334720696e-05, + "loss": 0.9938, + "step": 5711 + }, + { + "epoch": 0.3924694242132747, + "grad_norm": 1.115168809890747, + "learning_rate": 3.806680964511736e-05, + "loss": 1.0102, + "step": 5712 + }, + { + "epoch": 0.39253813384636527, + "grad_norm": 1.086202621459961, + "learning_rate": 3.806169720378844e-05, + "loss": 0.9706, + "step": 5713 + }, + { + "epoch": 0.39260684347945585, + "grad_norm": 1.0518600940704346, + "learning_rate": 3.8056584011028044e-05, + "loss": 0.9969, + "step": 5714 + }, + { + "epoch": 0.39267555311254637, + "grad_norm": 1.1433491706848145, + "learning_rate": 3.8051470067130334e-05, + "loss": 0.9486, + "step": 5715 + }, + { + "epoch": 0.39274426274563695, + "grad_norm": 1.0650503635406494, + "learning_rate": 3.804635537238951e-05, + "loss": 1.0164, + "step": 5716 + }, + { + "epoch": 0.3928129723787275, + "grad_norm": 1.077404260635376, + "learning_rate": 3.8041239927099805e-05, + "loss": 1.0164, + "step": 5717 + }, + { + "epoch": 0.39288168201181806, + "grad_norm": 1.0718801021575928, + "learning_rate": 3.8036123731555525e-05, + "loss": 0.9928, + "step": 5718 + }, + { + "epoch": 0.39295039164490864, + "grad_norm": 1.000999927520752, + "learning_rate": 3.803100678605099e-05, + "loss": 0.9598, + "step": 5719 + }, + { + "epoch": 0.39301910127799916, + "grad_norm": 1.1028951406478882, + "learning_rate": 3.802588909088059e-05, + "loss": 0.9937, + "step": 5720 + }, + { + "epoch": 0.39308781091108974, + "grad_norm": 1.0276685953140259, + "learning_rate": 3.8020770646338734e-05, + "loss": 0.922, + "step": 5721 + }, + { + "epoch": 0.3931565205441803, + "grad_norm": 1.073103666305542, + "learning_rate": 3.801565145271988e-05, + "loss": 0.975, + "step": 5722 + }, + { + "epoch": 0.39322523017727085, + "grad_norm": 1.0523924827575684, + "learning_rate": 3.801053151031853e-05, + "loss": 0.9856, + "step": 5723 + }, + { + "epoch": 0.3932939398103614, + "grad_norm": 1.065126895904541, + "learning_rate": 3.8005410819429236e-05, + "loss": 0.9882, + "step": 5724 + }, + { + "epoch": 0.39336264944345195, + "grad_norm": 1.1985548734664917, + "learning_rate": 3.8000289380346585e-05, + "loss": 1.0535, + "step": 5725 + }, + { + "epoch": 0.39343135907654253, + "grad_norm": 1.060925841331482, + "learning_rate": 3.7995167193365216e-05, + "loss": 0.9757, + "step": 5726 + }, + { + "epoch": 0.3935000687096331, + "grad_norm": 1.1418098211288452, + "learning_rate": 3.79900442587798e-05, + "loss": 1.0331, + "step": 5727 + }, + { + "epoch": 0.39356877834272364, + "grad_norm": 1.096435785293579, + "learning_rate": 3.798492057688506e-05, + "loss": 1.0053, + "step": 5728 + }, + { + "epoch": 0.3936374879758142, + "grad_norm": 1.043055534362793, + "learning_rate": 3.7979796147975756e-05, + "loss": 0.9515, + "step": 5729 + }, + { + "epoch": 0.3937061976089048, + "grad_norm": 1.0383057594299316, + "learning_rate": 3.797467097234669e-05, + "loss": 0.9646, + "step": 5730 + }, + { + "epoch": 0.3937749072419953, + "grad_norm": 1.1805171966552734, + "learning_rate": 3.796954505029272e-05, + "loss": 0.9994, + "step": 5731 + }, + { + "epoch": 0.3938436168750859, + "grad_norm": 1.0695627927780151, + "learning_rate": 3.796441838210873e-05, + "loss": 0.985, + "step": 5732 + }, + { + "epoch": 0.3939123265081764, + "grad_norm": 1.100736141204834, + "learning_rate": 3.7959290968089656e-05, + "loss": 0.9836, + "step": 5733 + }, + { + "epoch": 0.393981036141267, + "grad_norm": 1.0474913120269775, + "learning_rate": 3.7954162808530474e-05, + "loss": 0.9604, + "step": 5734 + }, + { + "epoch": 0.3940497457743576, + "grad_norm": 1.1507400274276733, + "learning_rate": 3.7949033903726206e-05, + "loss": 0.9953, + "step": 5735 + }, + { + "epoch": 0.3941184554074481, + "grad_norm": 1.043353796005249, + "learning_rate": 3.7943904253971905e-05, + "loss": 0.9608, + "step": 5736 + }, + { + "epoch": 0.3941871650405387, + "grad_norm": 0.9694053530693054, + "learning_rate": 3.7938773859562695e-05, + "loss": 0.9507, + "step": 5737 + }, + { + "epoch": 0.39425587467362927, + "grad_norm": 1.1024960279464722, + "learning_rate": 3.7933642720793705e-05, + "loss": 0.9935, + "step": 5738 + }, + { + "epoch": 0.3943245843067198, + "grad_norm": 1.058049201965332, + "learning_rate": 3.792851083796014e-05, + "loss": 0.9757, + "step": 5739 + }, + { + "epoch": 0.3943932939398104, + "grad_norm": 1.1500279903411865, + "learning_rate": 3.792337821135723e-05, + "loss": 0.9946, + "step": 5740 + }, + { + "epoch": 0.3944620035729009, + "grad_norm": 1.077372670173645, + "learning_rate": 3.791824484128025e-05, + "loss": 0.9681, + "step": 5741 + }, + { + "epoch": 0.3945307132059915, + "grad_norm": 1.167116641998291, + "learning_rate": 3.7913110728024515e-05, + "loss": 1.0145, + "step": 5742 + }, + { + "epoch": 0.39459942283908206, + "grad_norm": 1.124370813369751, + "learning_rate": 3.7907975871885404e-05, + "loss": 0.9647, + "step": 5743 + }, + { + "epoch": 0.3946681324721726, + "grad_norm": 1.2510695457458496, + "learning_rate": 3.79028402731583e-05, + "loss": 0.9955, + "step": 5744 + }, + { + "epoch": 0.39473684210526316, + "grad_norm": 1.0840632915496826, + "learning_rate": 3.7897703932138674e-05, + "loss": 0.9552, + "step": 5745 + }, + { + "epoch": 0.39480555173835374, + "grad_norm": 1.108481526374817, + "learning_rate": 3.7892566849122e-05, + "loss": 0.9564, + "step": 5746 + }, + { + "epoch": 0.39487426137144427, + "grad_norm": 1.1007859706878662, + "learning_rate": 3.788742902440382e-05, + "loss": 0.9548, + "step": 5747 + }, + { + "epoch": 0.39494297100453485, + "grad_norm": 1.1037265062332153, + "learning_rate": 3.78822904582797e-05, + "loss": 0.9408, + "step": 5748 + }, + { + "epoch": 0.3950116806376254, + "grad_norm": 1.0891962051391602, + "learning_rate": 3.787715115104526e-05, + "loss": 1.011, + "step": 5749 + }, + { + "epoch": 0.39508039027071595, + "grad_norm": 1.126690149307251, + "learning_rate": 3.787201110299617e-05, + "loss": 1.0122, + "step": 5750 + }, + { + "epoch": 0.39514909990380653, + "grad_norm": 1.0201104879379272, + "learning_rate": 3.786687031442814e-05, + "loss": 0.9784, + "step": 5751 + }, + { + "epoch": 0.39521780953689706, + "grad_norm": 1.1077384948730469, + "learning_rate": 3.7861728785636896e-05, + "loss": 0.9889, + "step": 5752 + }, + { + "epoch": 0.39528651916998764, + "grad_norm": 1.083074927330017, + "learning_rate": 3.785658651691824e-05, + "loss": 0.9986, + "step": 5753 + }, + { + "epoch": 0.3953552288030782, + "grad_norm": 1.139318823814392, + "learning_rate": 3.7851443508568e-05, + "loss": 0.9722, + "step": 5754 + }, + { + "epoch": 0.39542393843616874, + "grad_norm": 1.172850489616394, + "learning_rate": 3.784629976088205e-05, + "loss": 1.0231, + "step": 5755 + }, + { + "epoch": 0.3954926480692593, + "grad_norm": 1.052956223487854, + "learning_rate": 3.784115527415631e-05, + "loss": 0.9989, + "step": 5756 + }, + { + "epoch": 0.39556135770234985, + "grad_norm": 1.0878543853759766, + "learning_rate": 3.783601004868674e-05, + "loss": 1.0026, + "step": 5757 + }, + { + "epoch": 0.3956300673354404, + "grad_norm": 1.07817542552948, + "learning_rate": 3.7830864084769335e-05, + "loss": 0.9403, + "step": 5758 + }, + { + "epoch": 0.395698776968531, + "grad_norm": 1.033249855041504, + "learning_rate": 3.7825717382700144e-05, + "loss": 0.9644, + "step": 5759 + }, + { + "epoch": 0.39576748660162153, + "grad_norm": 1.0924935340881348, + "learning_rate": 3.782056994277524e-05, + "loss": 0.9842, + "step": 5760 + }, + { + "epoch": 0.3958361962347121, + "grad_norm": 1.0330169200897217, + "learning_rate": 3.781542176529078e-05, + "loss": 0.9647, + "step": 5761 + }, + { + "epoch": 0.3959049058678027, + "grad_norm": 1.0830612182617188, + "learning_rate": 3.781027285054292e-05, + "loss": 0.9893, + "step": 5762 + }, + { + "epoch": 0.3959736155008932, + "grad_norm": 1.1117358207702637, + "learning_rate": 3.7805123198827857e-05, + "loss": 0.9982, + "step": 5763 + }, + { + "epoch": 0.3960423251339838, + "grad_norm": 1.1599229574203491, + "learning_rate": 3.779997281044188e-05, + "loss": 0.97, + "step": 5764 + }, + { + "epoch": 0.3961110347670743, + "grad_norm": 1.0987645387649536, + "learning_rate": 3.7794821685681265e-05, + "loss": 0.9557, + "step": 5765 + }, + { + "epoch": 0.3961797444001649, + "grad_norm": 1.1178549528121948, + "learning_rate": 3.778966982484237e-05, + "loss": 1.0402, + "step": 5766 + }, + { + "epoch": 0.3962484540332555, + "grad_norm": 1.084671139717102, + "learning_rate": 3.778451722822155e-05, + "loss": 1.0006, + "step": 5767 + }, + { + "epoch": 0.396317163666346, + "grad_norm": 1.0257023572921753, + "learning_rate": 3.777936389611526e-05, + "loss": 0.9372, + "step": 5768 + }, + { + "epoch": 0.3963858732994366, + "grad_norm": 1.1416611671447754, + "learning_rate": 3.777420982881996e-05, + "loss": 0.9641, + "step": 5769 + }, + { + "epoch": 0.39645458293252717, + "grad_norm": 1.0732784271240234, + "learning_rate": 3.7769055026632147e-05, + "loss": 0.9458, + "step": 5770 + }, + { + "epoch": 0.3965232925656177, + "grad_norm": 1.1437686681747437, + "learning_rate": 3.77638994898484e-05, + "loss": 0.9595, + "step": 5771 + }, + { + "epoch": 0.39659200219870827, + "grad_norm": 1.1129440069198608, + "learning_rate": 3.775874321876528e-05, + "loss": 0.9707, + "step": 5772 + }, + { + "epoch": 0.3966607118317988, + "grad_norm": 0.9890673160552979, + "learning_rate": 3.775358621367945e-05, + "loss": 0.9359, + "step": 5773 + }, + { + "epoch": 0.3967294214648894, + "grad_norm": 1.0155553817749023, + "learning_rate": 3.774842847488759e-05, + "loss": 0.9616, + "step": 5774 + }, + { + "epoch": 0.39679813109797996, + "grad_norm": 1.1061081886291504, + "learning_rate": 3.7743270002686395e-05, + "loss": 0.9673, + "step": 5775 + }, + { + "epoch": 0.3968668407310705, + "grad_norm": 1.1549142599105835, + "learning_rate": 3.7738110797372664e-05, + "loss": 1.0372, + "step": 5776 + }, + { + "epoch": 0.39693555036416106, + "grad_norm": 1.1501612663269043, + "learning_rate": 3.7732950859243174e-05, + "loss": 0.9704, + "step": 5777 + }, + { + "epoch": 0.39700425999725164, + "grad_norm": 1.067508339881897, + "learning_rate": 3.772779018859479e-05, + "loss": 0.9699, + "step": 5778 + }, + { + "epoch": 0.39707296963034217, + "grad_norm": 1.088582992553711, + "learning_rate": 3.77226287857244e-05, + "loss": 0.9611, + "step": 5779 + }, + { + "epoch": 0.39714167926343275, + "grad_norm": 1.0608315467834473, + "learning_rate": 3.771746665092892e-05, + "loss": 0.9571, + "step": 5780 + }, + { + "epoch": 0.39721038889652327, + "grad_norm": 1.0853066444396973, + "learning_rate": 3.771230378450535e-05, + "loss": 0.9681, + "step": 5781 + }, + { + "epoch": 0.39727909852961385, + "grad_norm": 1.0344982147216797, + "learning_rate": 3.77071401867507e-05, + "loss": 0.9898, + "step": 5782 + }, + { + "epoch": 0.39734780816270443, + "grad_norm": 1.1049457788467407, + "learning_rate": 3.770197585796201e-05, + "loss": 1.0044, + "step": 5783 + }, + { + "epoch": 0.39741651779579495, + "grad_norm": 1.0171616077423096, + "learning_rate": 3.76968107984364e-05, + "loss": 0.9475, + "step": 5784 + }, + { + "epoch": 0.39748522742888553, + "grad_norm": 1.1043403148651123, + "learning_rate": 3.7691645008471e-05, + "loss": 0.9762, + "step": 5785 + }, + { + "epoch": 0.3975539370619761, + "grad_norm": 1.0807322263717651, + "learning_rate": 3.768647848836301e-05, + "loss": 0.9944, + "step": 5786 + }, + { + "epoch": 0.39762264669506664, + "grad_norm": 1.0461219549179077, + "learning_rate": 3.7681311238409645e-05, + "loss": 0.9467, + "step": 5787 + }, + { + "epoch": 0.3976913563281572, + "grad_norm": 1.1062642335891724, + "learning_rate": 3.767614325890817e-05, + "loss": 0.9692, + "step": 5788 + }, + { + "epoch": 0.39776006596124774, + "grad_norm": 1.093839406967163, + "learning_rate": 3.767097455015591e-05, + "loss": 1.0058, + "step": 5789 + }, + { + "epoch": 0.3978287755943383, + "grad_norm": 1.0500301122665405, + "learning_rate": 3.7665805112450205e-05, + "loss": 0.9578, + "step": 5790 + }, + { + "epoch": 0.3978974852274289, + "grad_norm": 1.0743367671966553, + "learning_rate": 3.766063494608846e-05, + "loss": 0.9651, + "step": 5791 + }, + { + "epoch": 0.39796619486051943, + "grad_norm": 1.0362612009048462, + "learning_rate": 3.7655464051368103e-05, + "loss": 0.99, + "step": 5792 + }, + { + "epoch": 0.39803490449361, + "grad_norm": 1.0773053169250488, + "learning_rate": 3.765029242858662e-05, + "loss": 0.9748, + "step": 5793 + }, + { + "epoch": 0.3981036141267006, + "grad_norm": 1.071006178855896, + "learning_rate": 3.764512007804152e-05, + "loss": 0.9444, + "step": 5794 + }, + { + "epoch": 0.3981723237597911, + "grad_norm": 1.216206669807434, + "learning_rate": 3.763994700003037e-05, + "loss": 0.996, + "step": 5795 + }, + { + "epoch": 0.3982410333928817, + "grad_norm": 1.0663137435913086, + "learning_rate": 3.763477319485079e-05, + "loss": 0.9752, + "step": 5796 + }, + { + "epoch": 0.3983097430259722, + "grad_norm": 1.0766496658325195, + "learning_rate": 3.7629598662800406e-05, + "loss": 0.9899, + "step": 5797 + }, + { + "epoch": 0.3983784526590628, + "grad_norm": 1.1340177059173584, + "learning_rate": 3.7624423404176904e-05, + "loss": 0.9609, + "step": 5798 + }, + { + "epoch": 0.3984471622921534, + "grad_norm": 1.0686352252960205, + "learning_rate": 3.7619247419278025e-05, + "loss": 0.9879, + "step": 5799 + }, + { + "epoch": 0.3985158719252439, + "grad_norm": 1.1124228239059448, + "learning_rate": 3.761407070840154e-05, + "loss": 1.0151, + "step": 5800 + }, + { + "epoch": 0.3985845815583345, + "grad_norm": 1.0558593273162842, + "learning_rate": 3.760889327184525e-05, + "loss": 0.9547, + "step": 5801 + }, + { + "epoch": 0.39865329119142506, + "grad_norm": 1.1026517152786255, + "learning_rate": 3.7603715109907026e-05, + "loss": 0.9998, + "step": 5802 + }, + { + "epoch": 0.3987220008245156, + "grad_norm": 1.1791142225265503, + "learning_rate": 3.759853622288475e-05, + "loss": 1.0322, + "step": 5803 + }, + { + "epoch": 0.39879071045760617, + "grad_norm": 1.1022378206253052, + "learning_rate": 3.759335661107638e-05, + "loss": 0.9788, + "step": 5804 + }, + { + "epoch": 0.3988594200906967, + "grad_norm": 1.0805397033691406, + "learning_rate": 3.7588176274779875e-05, + "loss": 0.943, + "step": 5805 + }, + { + "epoch": 0.3989281297237873, + "grad_norm": 1.0036572217941284, + "learning_rate": 3.758299521429326e-05, + "loss": 0.9446, + "step": 5806 + }, + { + "epoch": 0.39899683935687785, + "grad_norm": 1.0495610237121582, + "learning_rate": 3.757781342991461e-05, + "loss": 0.9702, + "step": 5807 + }, + { + "epoch": 0.3990655489899684, + "grad_norm": 1.092386245727539, + "learning_rate": 3.7572630921942016e-05, + "loss": 0.9927, + "step": 5808 + }, + { + "epoch": 0.39913425862305896, + "grad_norm": 1.0937310457229614, + "learning_rate": 3.756744769067363e-05, + "loss": 1.0178, + "step": 5809 + }, + { + "epoch": 0.39920296825614954, + "grad_norm": 1.1279832124710083, + "learning_rate": 3.756226373640764e-05, + "loss": 0.9967, + "step": 5810 + }, + { + "epoch": 0.39927167788924006, + "grad_norm": 1.1364672183990479, + "learning_rate": 3.755707905944229e-05, + "loss": 0.9917, + "step": 5811 + }, + { + "epoch": 0.39934038752233064, + "grad_norm": 1.0612208843231201, + "learning_rate": 3.7551893660075825e-05, + "loss": 1.0016, + "step": 5812 + }, + { + "epoch": 0.39940909715542117, + "grad_norm": 1.1268988847732544, + "learning_rate": 3.754670753860657e-05, + "loss": 1.0059, + "step": 5813 + }, + { + "epoch": 0.39947780678851175, + "grad_norm": 1.106921672821045, + "learning_rate": 3.754152069533289e-05, + "loss": 0.9627, + "step": 5814 + }, + { + "epoch": 0.3995465164216023, + "grad_norm": 1.0773272514343262, + "learning_rate": 3.7536333130553166e-05, + "loss": 0.9899, + "step": 5815 + }, + { + "epoch": 0.39961522605469285, + "grad_norm": 1.0747060775756836, + "learning_rate": 3.753114484456585e-05, + "loss": 1.0125, + "step": 5816 + }, + { + "epoch": 0.39968393568778343, + "grad_norm": 1.0322579145431519, + "learning_rate": 3.75259558376694e-05, + "loss": 0.9278, + "step": 5817 + }, + { + "epoch": 0.399752645320874, + "grad_norm": 1.0248669385910034, + "learning_rate": 3.7520766110162356e-05, + "loss": 0.9451, + "step": 5818 + }, + { + "epoch": 0.39982135495396454, + "grad_norm": 1.0670384168624878, + "learning_rate": 3.751557566234326e-05, + "loss": 0.9736, + "step": 5819 + }, + { + "epoch": 0.3998900645870551, + "grad_norm": 1.096880316734314, + "learning_rate": 3.751038449451074e-05, + "loss": 0.986, + "step": 5820 + }, + { + "epoch": 0.39995877422014564, + "grad_norm": 1.0169364213943481, + "learning_rate": 3.7505192606963426e-05, + "loss": 0.9897, + "step": 5821 + }, + { + "epoch": 0.4000274838532362, + "grad_norm": 1.1205395460128784, + "learning_rate": 3.7500000000000003e-05, + "loss": 0.9824, + "step": 5822 + }, + { + "epoch": 0.4000961934863268, + "grad_norm": 1.0481783151626587, + "learning_rate": 3.74948066739192e-05, + "loss": 0.9825, + "step": 5823 + }, + { + "epoch": 0.4001649031194173, + "grad_norm": 1.0398527383804321, + "learning_rate": 3.7489612629019786e-05, + "loss": 0.9885, + "step": 5824 + }, + { + "epoch": 0.4002336127525079, + "grad_norm": 1.1351145505905151, + "learning_rate": 3.748441786560058e-05, + "loss": 0.9982, + "step": 5825 + }, + { + "epoch": 0.4003023223855985, + "grad_norm": 1.1263140439987183, + "learning_rate": 3.747922238396042e-05, + "loss": 1.0152, + "step": 5826 + }, + { + "epoch": 0.400371032018689, + "grad_norm": 1.1154539585113525, + "learning_rate": 3.747402618439822e-05, + "loss": 0.9421, + "step": 5827 + }, + { + "epoch": 0.4004397416517796, + "grad_norm": 1.052106499671936, + "learning_rate": 3.746882926721288e-05, + "loss": 0.9803, + "step": 5828 + }, + { + "epoch": 0.4005084512848701, + "grad_norm": 1.1942203044891357, + "learning_rate": 3.7463631632703406e-05, + "loss": 0.976, + "step": 5829 + }, + { + "epoch": 0.4005771609179607, + "grad_norm": 1.0773166418075562, + "learning_rate": 3.74584332811688e-05, + "loss": 0.9873, + "step": 5830 + }, + { + "epoch": 0.4006458705510513, + "grad_norm": 1.08528470993042, + "learning_rate": 3.7453234212908123e-05, + "loss": 1.0098, + "step": 5831 + }, + { + "epoch": 0.4007145801841418, + "grad_norm": 1.0687590837478638, + "learning_rate": 3.744803442822048e-05, + "loss": 0.9686, + "step": 5832 + }, + { + "epoch": 0.4007832898172324, + "grad_norm": 1.0699225664138794, + "learning_rate": 3.7442833927404994e-05, + "loss": 0.9617, + "step": 5833 + }, + { + "epoch": 0.40085199945032296, + "grad_norm": 1.0513315200805664, + "learning_rate": 3.743763271076087e-05, + "loss": 1.0022, + "step": 5834 + }, + { + "epoch": 0.4009207090834135, + "grad_norm": 1.0936998128890991, + "learning_rate": 3.743243077858732e-05, + "loss": 0.9931, + "step": 5835 + }, + { + "epoch": 0.40098941871650406, + "grad_norm": 1.1599897146224976, + "learning_rate": 3.74272281311836e-05, + "loss": 0.9809, + "step": 5836 + }, + { + "epoch": 0.4010581283495946, + "grad_norm": 1.0877082347869873, + "learning_rate": 3.742202476884903e-05, + "loss": 1.0197, + "step": 5837 + }, + { + "epoch": 0.40112683798268517, + "grad_norm": 1.1497151851654053, + "learning_rate": 3.741682069188295e-05, + "loss": 1.0267, + "step": 5838 + }, + { + "epoch": 0.40119554761577575, + "grad_norm": 1.186575174331665, + "learning_rate": 3.741161590058474e-05, + "loss": 1.0292, + "step": 5839 + }, + { + "epoch": 0.4012642572488663, + "grad_norm": 1.0355021953582764, + "learning_rate": 3.740641039525384e-05, + "loss": 0.9756, + "step": 5840 + }, + { + "epoch": 0.40133296688195685, + "grad_norm": 1.0789352655410767, + "learning_rate": 3.740120417618971e-05, + "loss": 0.9808, + "step": 5841 + }, + { + "epoch": 0.40140167651504743, + "grad_norm": 1.1060246229171753, + "learning_rate": 3.739599724369187e-05, + "loss": 1.0002, + "step": 5842 + }, + { + "epoch": 0.40147038614813796, + "grad_norm": 1.014970064163208, + "learning_rate": 3.739078959805987e-05, + "loss": 0.9296, + "step": 5843 + }, + { + "epoch": 0.40153909578122854, + "grad_norm": 1.14531409740448, + "learning_rate": 3.738558123959329e-05, + "loss": 0.9756, + "step": 5844 + }, + { + "epoch": 0.40160780541431906, + "grad_norm": 1.1391386985778809, + "learning_rate": 3.738037216859177e-05, + "loss": 1.0004, + "step": 5845 + }, + { + "epoch": 0.40167651504740964, + "grad_norm": 1.0701968669891357, + "learning_rate": 3.7375162385355e-05, + "loss": 0.9571, + "step": 5846 + }, + { + "epoch": 0.4017452246805002, + "grad_norm": 1.0322779417037964, + "learning_rate": 3.736995189018268e-05, + "loss": 0.9989, + "step": 5847 + }, + { + "epoch": 0.40181393431359075, + "grad_norm": 1.0978516340255737, + "learning_rate": 3.736474068337457e-05, + "loss": 0.9765, + "step": 5848 + }, + { + "epoch": 0.40188264394668133, + "grad_norm": 1.169094204902649, + "learning_rate": 3.7359528765230466e-05, + "loss": 0.9973, + "step": 5849 + }, + { + "epoch": 0.4019513535797719, + "grad_norm": 1.098265290260315, + "learning_rate": 3.735431613605021e-05, + "loss": 0.9862, + "step": 5850 + }, + { + "epoch": 0.40202006321286243, + "grad_norm": 1.106567144393921, + "learning_rate": 3.734910279613367e-05, + "loss": 0.9575, + "step": 5851 + }, + { + "epoch": 0.402088772845953, + "grad_norm": 1.0946850776672363, + "learning_rate": 3.734388874578079e-05, + "loss": 0.9927, + "step": 5852 + }, + { + "epoch": 0.40215748247904354, + "grad_norm": 1.0441590547561646, + "learning_rate": 3.7338673985291514e-05, + "loss": 0.9423, + "step": 5853 + }, + { + "epoch": 0.4022261921121341, + "grad_norm": 1.079594373703003, + "learning_rate": 3.7333458514965846e-05, + "loss": 1.0068, + "step": 5854 + }, + { + "epoch": 0.4022949017452247, + "grad_norm": 1.1437907218933105, + "learning_rate": 3.7328242335103826e-05, + "loss": 0.9651, + "step": 5855 + }, + { + "epoch": 0.4023636113783152, + "grad_norm": 1.0664628744125366, + "learning_rate": 3.732302544600555e-05, + "loss": 0.9633, + "step": 5856 + }, + { + "epoch": 0.4024323210114058, + "grad_norm": 1.0761957168579102, + "learning_rate": 3.731780784797112e-05, + "loss": 0.9513, + "step": 5857 + }, + { + "epoch": 0.4025010306444964, + "grad_norm": 1.1059809923171997, + "learning_rate": 3.7312589541300724e-05, + "loss": 0.994, + "step": 5858 + }, + { + "epoch": 0.4025697402775869, + "grad_norm": 1.1548784971237183, + "learning_rate": 3.7307370526294556e-05, + "loss": 1.0567, + "step": 5859 + }, + { + "epoch": 0.4026384499106775, + "grad_norm": 1.1076059341430664, + "learning_rate": 3.7302150803252863e-05, + "loss": 0.9938, + "step": 5860 + }, + { + "epoch": 0.402707159543768, + "grad_norm": 1.0281000137329102, + "learning_rate": 3.729693037247594e-05, + "loss": 0.9736, + "step": 5861 + }, + { + "epoch": 0.4027758691768586, + "grad_norm": 1.1450904607772827, + "learning_rate": 3.7291709234264104e-05, + "loss": 0.986, + "step": 5862 + }, + { + "epoch": 0.40284457880994917, + "grad_norm": 1.1724339723587036, + "learning_rate": 3.7286487388917747e-05, + "loss": 0.9837, + "step": 5863 + }, + { + "epoch": 0.4029132884430397, + "grad_norm": 1.0621658563613892, + "learning_rate": 3.728126483673725e-05, + "loss": 0.9875, + "step": 5864 + }, + { + "epoch": 0.4029819980761303, + "grad_norm": 1.0541019439697266, + "learning_rate": 3.727604157802308e-05, + "loss": 0.9705, + "step": 5865 + }, + { + "epoch": 0.40305070770922086, + "grad_norm": 1.0250160694122314, + "learning_rate": 3.7270817613075714e-05, + "loss": 0.9725, + "step": 5866 + }, + { + "epoch": 0.4031194173423114, + "grad_norm": 1.005388617515564, + "learning_rate": 3.72655929421957e-05, + "loss": 0.9635, + "step": 5867 + }, + { + "epoch": 0.40318812697540196, + "grad_norm": 1.0178115367889404, + "learning_rate": 3.7260367565683605e-05, + "loss": 0.9669, + "step": 5868 + }, + { + "epoch": 0.4032568366084925, + "grad_norm": 1.0695453882217407, + "learning_rate": 3.7255141483840026e-05, + "loss": 0.9963, + "step": 5869 + }, + { + "epoch": 0.40332554624158307, + "grad_norm": 1.037446141242981, + "learning_rate": 3.724991469696565e-05, + "loss": 0.9397, + "step": 5870 + }, + { + "epoch": 0.40339425587467365, + "grad_norm": 1.0225123167037964, + "learning_rate": 3.7244687205361136e-05, + "loss": 0.9605, + "step": 5871 + }, + { + "epoch": 0.40346296550776417, + "grad_norm": 1.0919169187545776, + "learning_rate": 3.723945900932723e-05, + "loss": 0.9501, + "step": 5872 + }, + { + "epoch": 0.40353167514085475, + "grad_norm": 1.0696632862091064, + "learning_rate": 3.7234230109164724e-05, + "loss": 0.9705, + "step": 5873 + }, + { + "epoch": 0.40360038477394533, + "grad_norm": 1.1683512926101685, + "learning_rate": 3.722900050517441e-05, + "loss": 1.0002, + "step": 5874 + }, + { + "epoch": 0.40366909440703586, + "grad_norm": 1.0866297483444214, + "learning_rate": 3.7223770197657156e-05, + "loss": 0.9787, + "step": 5875 + }, + { + "epoch": 0.40373780404012644, + "grad_norm": 1.1190030574798584, + "learning_rate": 3.721853918691386e-05, + "loss": 1.0072, + "step": 5876 + }, + { + "epoch": 0.40380651367321696, + "grad_norm": 1.0998541116714478, + "learning_rate": 3.721330747324544e-05, + "loss": 0.9473, + "step": 5877 + }, + { + "epoch": 0.40387522330630754, + "grad_norm": 1.1375139951705933, + "learning_rate": 3.720807505695291e-05, + "loss": 0.9604, + "step": 5878 + }, + { + "epoch": 0.4039439329393981, + "grad_norm": 1.0132173299789429, + "learning_rate": 3.720284193833725e-05, + "loss": 0.9784, + "step": 5879 + }, + { + "epoch": 0.40401264257248864, + "grad_norm": 1.1212657690048218, + "learning_rate": 3.719760811769953e-05, + "loss": 0.9857, + "step": 5880 + }, + { + "epoch": 0.4040813522055792, + "grad_norm": 1.1428474187850952, + "learning_rate": 3.719237359534087e-05, + "loss": 0.9584, + "step": 5881 + }, + { + "epoch": 0.4041500618386698, + "grad_norm": 1.0906352996826172, + "learning_rate": 3.718713837156237e-05, + "loss": 0.9833, + "step": 5882 + }, + { + "epoch": 0.40421877147176033, + "grad_norm": 1.0940278768539429, + "learning_rate": 3.718190244666525e-05, + "loss": 1.0029, + "step": 5883 + }, + { + "epoch": 0.4042874811048509, + "grad_norm": 1.1124032735824585, + "learning_rate": 3.71766658209507e-05, + "loss": 0.9984, + "step": 5884 + }, + { + "epoch": 0.40435619073794143, + "grad_norm": 1.048907995223999, + "learning_rate": 3.717142849471999e-05, + "loss": 0.965, + "step": 5885 + }, + { + "epoch": 0.404424900371032, + "grad_norm": 1.146571397781372, + "learning_rate": 3.716619046827442e-05, + "loss": 1.0322, + "step": 5886 + }, + { + "epoch": 0.4044936100041226, + "grad_norm": 1.0573322772979736, + "learning_rate": 3.716095174191534e-05, + "loss": 0.9934, + "step": 5887 + }, + { + "epoch": 0.4045623196372131, + "grad_norm": 1.1048049926757812, + "learning_rate": 3.715571231594411e-05, + "loss": 0.9774, + "step": 5888 + }, + { + "epoch": 0.4046310292703037, + "grad_norm": 1.1047827005386353, + "learning_rate": 3.715047219066217e-05, + "loss": 0.9668, + "step": 5889 + }, + { + "epoch": 0.4046997389033943, + "grad_norm": 1.0458101034164429, + "learning_rate": 3.714523136637096e-05, + "loss": 0.9739, + "step": 5890 + }, + { + "epoch": 0.4047684485364848, + "grad_norm": 1.08072829246521, + "learning_rate": 3.7139989843372007e-05, + "loss": 0.9972, + "step": 5891 + }, + { + "epoch": 0.4048371581695754, + "grad_norm": 1.1458656787872314, + "learning_rate": 3.713474762196684e-05, + "loss": 0.9933, + "step": 5892 + }, + { + "epoch": 0.4049058678026659, + "grad_norm": 1.0955108404159546, + "learning_rate": 3.7129504702457036e-05, + "loss": 0.9554, + "step": 5893 + }, + { + "epoch": 0.4049745774357565, + "grad_norm": 1.1553370952606201, + "learning_rate": 3.712426108514422e-05, + "loss": 0.9818, + "step": 5894 + }, + { + "epoch": 0.40504328706884707, + "grad_norm": 1.1239478588104248, + "learning_rate": 3.711901677033006e-05, + "loss": 0.9583, + "step": 5895 + }, + { + "epoch": 0.4051119967019376, + "grad_norm": 1.0304676294326782, + "learning_rate": 3.711377175831626e-05, + "loss": 0.9702, + "step": 5896 + }, + { + "epoch": 0.4051807063350282, + "grad_norm": 1.1466331481933594, + "learning_rate": 3.710852604940455e-05, + "loss": 0.9898, + "step": 5897 + }, + { + "epoch": 0.40524941596811875, + "grad_norm": 1.096445918083191, + "learning_rate": 3.7103279643896723e-05, + "loss": 1.0138, + "step": 5898 + }, + { + "epoch": 0.4053181256012093, + "grad_norm": 1.1763026714324951, + "learning_rate": 3.709803254209459e-05, + "loss": 0.9673, + "step": 5899 + }, + { + "epoch": 0.40538683523429986, + "grad_norm": 1.0639947652816772, + "learning_rate": 3.7092784744300036e-05, + "loss": 0.9639, + "step": 5900 + }, + { + "epoch": 0.4054555448673904, + "grad_norm": 1.1092987060546875, + "learning_rate": 3.708753625081494e-05, + "loss": 0.9612, + "step": 5901 + }, + { + "epoch": 0.40552425450048096, + "grad_norm": 1.0504441261291504, + "learning_rate": 3.708228706194125e-05, + "loss": 0.9841, + "step": 5902 + }, + { + "epoch": 0.40559296413357154, + "grad_norm": 1.1915996074676514, + "learning_rate": 3.7077037177980954e-05, + "loss": 0.9854, + "step": 5903 + }, + { + "epoch": 0.40566167376666207, + "grad_norm": 1.1175752878189087, + "learning_rate": 3.707178659923608e-05, + "loss": 1.0216, + "step": 5904 + }, + { + "epoch": 0.40573038339975265, + "grad_norm": 1.0364826917648315, + "learning_rate": 3.706653532600868e-05, + "loss": 0.9355, + "step": 5905 + }, + { + "epoch": 0.4057990930328432, + "grad_norm": 1.1044539213180542, + "learning_rate": 3.706128335860086e-05, + "loss": 0.9624, + "step": 5906 + }, + { + "epoch": 0.40586780266593375, + "grad_norm": 1.1061867475509644, + "learning_rate": 3.705603069731476e-05, + "loss": 0.9686, + "step": 5907 + }, + { + "epoch": 0.40593651229902433, + "grad_norm": 1.1608693599700928, + "learning_rate": 3.7050777342452566e-05, + "loss": 1.0089, + "step": 5908 + }, + { + "epoch": 0.40600522193211486, + "grad_norm": 1.1803675889968872, + "learning_rate": 3.70455232943165e-05, + "loss": 0.9784, + "step": 5909 + }, + { + "epoch": 0.40607393156520544, + "grad_norm": 1.0986841917037964, + "learning_rate": 3.704026855320882e-05, + "loss": 0.9302, + "step": 5910 + }, + { + "epoch": 0.406142641198296, + "grad_norm": 1.1389946937561035, + "learning_rate": 3.703501311943184e-05, + "loss": 1.034, + "step": 5911 + }, + { + "epoch": 0.40621135083138654, + "grad_norm": 1.1222599744796753, + "learning_rate": 3.702975699328789e-05, + "loss": 1.0019, + "step": 5912 + }, + { + "epoch": 0.4062800604644771, + "grad_norm": 1.0354435443878174, + "learning_rate": 3.702450017507935e-05, + "loss": 0.9677, + "step": 5913 + }, + { + "epoch": 0.4063487700975677, + "grad_norm": 0.9940625429153442, + "learning_rate": 3.7019242665108655e-05, + "loss": 0.9287, + "step": 5914 + }, + { + "epoch": 0.4064174797306582, + "grad_norm": 1.0982099771499634, + "learning_rate": 3.701398446367825e-05, + "loss": 0.9817, + "step": 5915 + }, + { + "epoch": 0.4064861893637488, + "grad_norm": 1.1984351873397827, + "learning_rate": 3.700872557109065e-05, + "loss": 0.9815, + "step": 5916 + }, + { + "epoch": 0.40655489899683933, + "grad_norm": 1.0121192932128906, + "learning_rate": 3.700346598764838e-05, + "loss": 0.991, + "step": 5917 + }, + { + "epoch": 0.4066236086299299, + "grad_norm": 1.056689739227295, + "learning_rate": 3.699820571365404e-05, + "loss": 1.0025, + "step": 5918 + }, + { + "epoch": 0.4066923182630205, + "grad_norm": 1.0454752445220947, + "learning_rate": 3.699294474941024e-05, + "loss": 0.9553, + "step": 5919 + }, + { + "epoch": 0.406761027896111, + "grad_norm": 1.1050666570663452, + "learning_rate": 3.698768309521964e-05, + "loss": 1.0008, + "step": 5920 + }, + { + "epoch": 0.4068297375292016, + "grad_norm": 1.1624643802642822, + "learning_rate": 3.698242075138494e-05, + "loss": 0.9774, + "step": 5921 + }, + { + "epoch": 0.4068984471622922, + "grad_norm": 1.1383414268493652, + "learning_rate": 3.697715771820888e-05, + "loss": 1.0261, + "step": 5922 + }, + { + "epoch": 0.4069671567953827, + "grad_norm": 1.0928876399993896, + "learning_rate": 3.697189399599425e-05, + "loss": 1.0078, + "step": 5923 + }, + { + "epoch": 0.4070358664284733, + "grad_norm": 1.1146560907363892, + "learning_rate": 3.6966629585043846e-05, + "loss": 0.9799, + "step": 5924 + }, + { + "epoch": 0.4071045760615638, + "grad_norm": 1.1360365152359009, + "learning_rate": 3.696136448566054e-05, + "loss": 0.9968, + "step": 5925 + }, + { + "epoch": 0.4071732856946544, + "grad_norm": 1.04770028591156, + "learning_rate": 3.6956098698147226e-05, + "loss": 0.9822, + "step": 5926 + }, + { + "epoch": 0.40724199532774497, + "grad_norm": 1.114266037940979, + "learning_rate": 3.6950832222806846e-05, + "loss": 0.9999, + "step": 5927 + }, + { + "epoch": 0.4073107049608355, + "grad_norm": 1.1061071157455444, + "learning_rate": 3.694556505994238e-05, + "loss": 1.0046, + "step": 5928 + }, + { + "epoch": 0.40737941459392607, + "grad_norm": 1.0510965585708618, + "learning_rate": 3.694029720985683e-05, + "loss": 0.9879, + "step": 5929 + }, + { + "epoch": 0.40744812422701665, + "grad_norm": 1.1150283813476562, + "learning_rate": 3.693502867285327e-05, + "loss": 0.9888, + "step": 5930 + }, + { + "epoch": 0.4075168338601072, + "grad_norm": 1.1717113256454468, + "learning_rate": 3.692975944923478e-05, + "loss": 0.9793, + "step": 5931 + }, + { + "epoch": 0.40758554349319776, + "grad_norm": 1.0710705518722534, + "learning_rate": 3.6924489539304514e-05, + "loss": 0.997, + "step": 5932 + }, + { + "epoch": 0.4076542531262883, + "grad_norm": 1.089956283569336, + "learning_rate": 3.691921894336563e-05, + "loss": 0.9706, + "step": 5933 + }, + { + "epoch": 0.40772296275937886, + "grad_norm": 1.0879583358764648, + "learning_rate": 3.691394766172135e-05, + "loss": 0.9734, + "step": 5934 + }, + { + "epoch": 0.40779167239246944, + "grad_norm": 1.0899569988250732, + "learning_rate": 3.690867569467492e-05, + "loss": 0.9899, + "step": 5935 + }, + { + "epoch": 0.40786038202555996, + "grad_norm": 1.0643121004104614, + "learning_rate": 3.690340304252964e-05, + "loss": 0.9661, + "step": 5936 + }, + { + "epoch": 0.40792909165865054, + "grad_norm": 1.1292835474014282, + "learning_rate": 3.6898129705588844e-05, + "loss": 1.0009, + "step": 5937 + }, + { + "epoch": 0.4079978012917411, + "grad_norm": 1.0558017492294312, + "learning_rate": 3.6892855684155895e-05, + "loss": 0.9805, + "step": 5938 + }, + { + "epoch": 0.40806651092483165, + "grad_norm": 1.0569862127304077, + "learning_rate": 3.688758097853422e-05, + "loss": 1.0015, + "step": 5939 + }, + { + "epoch": 0.40813522055792223, + "grad_norm": 1.0419104099273682, + "learning_rate": 3.688230558902725e-05, + "loss": 0.9487, + "step": 5940 + }, + { + "epoch": 0.40820393019101275, + "grad_norm": 1.0583701133728027, + "learning_rate": 3.687702951593849e-05, + "loss": 0.9853, + "step": 5941 + }, + { + "epoch": 0.40827263982410333, + "grad_norm": 0.9715238809585571, + "learning_rate": 3.687175275957146e-05, + "loss": 0.9582, + "step": 5942 + }, + { + "epoch": 0.4083413494571939, + "grad_norm": 1.0489120483398438, + "learning_rate": 3.686647532022973e-05, + "loss": 0.9802, + "step": 5943 + }, + { + "epoch": 0.40841005909028444, + "grad_norm": 1.0989128351211548, + "learning_rate": 3.6861197198216915e-05, + "loss": 0.9457, + "step": 5944 + }, + { + "epoch": 0.408478768723375, + "grad_norm": 1.1937135457992554, + "learning_rate": 3.685591839383666e-05, + "loss": 1.0172, + "step": 5945 + }, + { + "epoch": 0.4085474783564656, + "grad_norm": 1.0925402641296387, + "learning_rate": 3.685063890739265e-05, + "loss": 0.9499, + "step": 5946 + }, + { + "epoch": 0.4086161879895561, + "grad_norm": 1.1614634990692139, + "learning_rate": 3.68453587391886e-05, + "loss": 1.0361, + "step": 5947 + }, + { + "epoch": 0.4086848976226467, + "grad_norm": 1.046694278717041, + "learning_rate": 3.684007788952829e-05, + "loss": 0.9744, + "step": 5948 + }, + { + "epoch": 0.40875360725573723, + "grad_norm": 1.1606483459472656, + "learning_rate": 3.6834796358715516e-05, + "loss": 1.0034, + "step": 5949 + }, + { + "epoch": 0.4088223168888278, + "grad_norm": 1.124338984489441, + "learning_rate": 3.682951414705413e-05, + "loss": 0.9743, + "step": 5950 + }, + { + "epoch": 0.4088910265219184, + "grad_norm": 1.0753884315490723, + "learning_rate": 3.6824231254848e-05, + "loss": 0.967, + "step": 5951 + }, + { + "epoch": 0.4089597361550089, + "grad_norm": 1.103418231010437, + "learning_rate": 3.681894768240106e-05, + "loss": 1.0042, + "step": 5952 + }, + { + "epoch": 0.4090284457880995, + "grad_norm": 1.1428083181381226, + "learning_rate": 3.6813663430017263e-05, + "loss": 0.999, + "step": 5953 + }, + { + "epoch": 0.4090971554211901, + "grad_norm": 1.0821481943130493, + "learning_rate": 3.680837849800062e-05, + "loss": 0.9808, + "step": 5954 + }, + { + "epoch": 0.4091658650542806, + "grad_norm": 1.1984364986419678, + "learning_rate": 3.680309288665515e-05, + "loss": 0.9655, + "step": 5955 + }, + { + "epoch": 0.4092345746873712, + "grad_norm": 1.0942744016647339, + "learning_rate": 3.679780659628494e-05, + "loss": 0.9928, + "step": 5956 + }, + { + "epoch": 0.4093032843204617, + "grad_norm": 1.185929298400879, + "learning_rate": 3.679251962719412e-05, + "loss": 1.0123, + "step": 5957 + }, + { + "epoch": 0.4093719939535523, + "grad_norm": 1.130793571472168, + "learning_rate": 3.6787231979686835e-05, + "loss": 0.9693, + "step": 5958 + }, + { + "epoch": 0.40944070358664286, + "grad_norm": 1.0812442302703857, + "learning_rate": 3.678194365406729e-05, + "loss": 0.964, + "step": 5959 + }, + { + "epoch": 0.4095094132197334, + "grad_norm": 1.1599228382110596, + "learning_rate": 3.6776654650639696e-05, + "loss": 0.9614, + "step": 5960 + }, + { + "epoch": 0.40957812285282397, + "grad_norm": 1.1406663656234741, + "learning_rate": 3.677136496970835e-05, + "loss": 1.0245, + "step": 5961 + }, + { + "epoch": 0.40964683248591455, + "grad_norm": 1.0813205242156982, + "learning_rate": 3.676607461157755e-05, + "loss": 0.9614, + "step": 5962 + }, + { + "epoch": 0.40971554211900507, + "grad_norm": 1.1340867280960083, + "learning_rate": 3.6760783576551646e-05, + "loss": 0.982, + "step": 5963 + }, + { + "epoch": 0.40978425175209565, + "grad_norm": 1.067282795906067, + "learning_rate": 3.675549186493504e-05, + "loss": 0.9789, + "step": 5964 + }, + { + "epoch": 0.4098529613851862, + "grad_norm": 1.0850096940994263, + "learning_rate": 3.675019947703216e-05, + "loss": 0.9836, + "step": 5965 + }, + { + "epoch": 0.40992167101827676, + "grad_norm": 1.1610101461410522, + "learning_rate": 3.674490641314746e-05, + "loss": 1.0081, + "step": 5966 + }, + { + "epoch": 0.40999038065136734, + "grad_norm": 1.1682567596435547, + "learning_rate": 3.6739612673585464e-05, + "loss": 1.0075, + "step": 5967 + }, + { + "epoch": 0.41005909028445786, + "grad_norm": 1.1305862665176392, + "learning_rate": 3.67343182586507e-05, + "loss": 1.0442, + "step": 5968 + }, + { + "epoch": 0.41012779991754844, + "grad_norm": 1.0502678155899048, + "learning_rate": 3.6729023168647766e-05, + "loss": 0.9594, + "step": 5969 + }, + { + "epoch": 0.410196509550639, + "grad_norm": 1.1033300161361694, + "learning_rate": 3.6723727403881284e-05, + "loss": 0.9912, + "step": 5970 + }, + { + "epoch": 0.41026521918372955, + "grad_norm": 1.1753127574920654, + "learning_rate": 3.6718430964655906e-05, + "loss": 1.0322, + "step": 5971 + }, + { + "epoch": 0.4103339288168201, + "grad_norm": 1.082200288772583, + "learning_rate": 3.671313385127634e-05, + "loss": 0.9815, + "step": 5972 + }, + { + "epoch": 0.41040263844991065, + "grad_norm": 1.0238240957260132, + "learning_rate": 3.670783606404733e-05, + "loss": 0.9698, + "step": 5973 + }, + { + "epoch": 0.41047134808300123, + "grad_norm": 1.0419608354568481, + "learning_rate": 3.670253760327365e-05, + "loss": 0.9413, + "step": 5974 + }, + { + "epoch": 0.4105400577160918, + "grad_norm": 1.110153079032898, + "learning_rate": 3.669723846926012e-05, + "loss": 0.9969, + "step": 5975 + }, + { + "epoch": 0.41060876734918234, + "grad_norm": 1.1631489992141724, + "learning_rate": 3.669193866231157e-05, + "loss": 0.9448, + "step": 5976 + }, + { + "epoch": 0.4106774769822729, + "grad_norm": 1.0414544343948364, + "learning_rate": 3.6686638182732945e-05, + "loss": 0.9534, + "step": 5977 + }, + { + "epoch": 0.4107461866153635, + "grad_norm": 1.111603021621704, + "learning_rate": 3.6681337030829145e-05, + "loss": 0.9727, + "step": 5978 + }, + { + "epoch": 0.410814896248454, + "grad_norm": 1.1119691133499146, + "learning_rate": 3.667603520690514e-05, + "loss": 1.0512, + "step": 5979 + }, + { + "epoch": 0.4108836058815446, + "grad_norm": 1.177992582321167, + "learning_rate": 3.6670732711265955e-05, + "loss": 0.9925, + "step": 5980 + }, + { + "epoch": 0.4109523155146351, + "grad_norm": 1.1607675552368164, + "learning_rate": 3.666542954421663e-05, + "loss": 0.9982, + "step": 5981 + }, + { + "epoch": 0.4110210251477257, + "grad_norm": 1.0752785205841064, + "learning_rate": 3.666012570606226e-05, + "loss": 1.0162, + "step": 5982 + }, + { + "epoch": 0.4110897347808163, + "grad_norm": 1.0901529788970947, + "learning_rate": 3.6654821197107966e-05, + "loss": 0.974, + "step": 5983 + }, + { + "epoch": 0.4111584444139068, + "grad_norm": 1.0387572050094604, + "learning_rate": 3.6649516017658915e-05, + "loss": 1.0158, + "step": 5984 + }, + { + "epoch": 0.4112271540469974, + "grad_norm": 1.0566489696502686, + "learning_rate": 3.664421016802031e-05, + "loss": 0.9954, + "step": 5985 + }, + { + "epoch": 0.41129586368008797, + "grad_norm": 1.0930472612380981, + "learning_rate": 3.663890364849739e-05, + "loss": 0.979, + "step": 5986 + }, + { + "epoch": 0.4113645733131785, + "grad_norm": 1.0127832889556885, + "learning_rate": 3.6633596459395444e-05, + "loss": 0.9784, + "step": 5987 + }, + { + "epoch": 0.4114332829462691, + "grad_norm": 1.0685750246047974, + "learning_rate": 3.662828860101979e-05, + "loss": 0.967, + "step": 5988 + }, + { + "epoch": 0.4115019925793596, + "grad_norm": 1.0809916257858276, + "learning_rate": 3.6622980073675776e-05, + "loss": 1.005, + "step": 5989 + }, + { + "epoch": 0.4115707022124502, + "grad_norm": 1.1432400941848755, + "learning_rate": 3.6617670877668806e-05, + "loss": 0.9936, + "step": 5990 + }, + { + "epoch": 0.41163941184554076, + "grad_norm": 1.0711630582809448, + "learning_rate": 3.661236101330432e-05, + "loss": 0.9474, + "step": 5991 + }, + { + "epoch": 0.4117081214786313, + "grad_norm": 1.044480562210083, + "learning_rate": 3.660705048088778e-05, + "loss": 0.9569, + "step": 5992 + }, + { + "epoch": 0.41177683111172186, + "grad_norm": 1.087786316871643, + "learning_rate": 3.660173928072471e-05, + "loss": 0.9688, + "step": 5993 + }, + { + "epoch": 0.41184554074481244, + "grad_norm": 1.0357427597045898, + "learning_rate": 3.659642741312065e-05, + "loss": 0.9578, + "step": 5994 + }, + { + "epoch": 0.41191425037790297, + "grad_norm": 1.0901039838790894, + "learning_rate": 3.659111487838119e-05, + "loss": 0.9895, + "step": 5995 + }, + { + "epoch": 0.41198296001099355, + "grad_norm": 1.1309070587158203, + "learning_rate": 3.658580167681196e-05, + "loss": 0.9667, + "step": 5996 + }, + { + "epoch": 0.4120516696440841, + "grad_norm": 1.099904179573059, + "learning_rate": 3.6580487808718616e-05, + "loss": 1.0052, + "step": 5997 + }, + { + "epoch": 0.41212037927717465, + "grad_norm": 1.0865029096603394, + "learning_rate": 3.657517327440688e-05, + "loss": 0.9738, + "step": 5998 + }, + { + "epoch": 0.41218908891026523, + "grad_norm": 1.1122322082519531, + "learning_rate": 3.656985807418248e-05, + "loss": 1.0018, + "step": 5999 + }, + { + "epoch": 0.41225779854335576, + "grad_norm": 1.2506239414215088, + "learning_rate": 3.65645422083512e-05, + "loss": 1.0025, + "step": 6000 + }, + { + "epoch": 0.41232650817644634, + "grad_norm": 1.072919249534607, + "learning_rate": 3.6559225677218856e-05, + "loss": 0.9705, + "step": 6001 + }, + { + "epoch": 0.4123952178095369, + "grad_norm": 1.127630352973938, + "learning_rate": 3.6553908481091306e-05, + "loss": 0.9821, + "step": 6002 + }, + { + "epoch": 0.41246392744262744, + "grad_norm": 1.0975620746612549, + "learning_rate": 3.6548590620274445e-05, + "loss": 0.9595, + "step": 6003 + }, + { + "epoch": 0.412532637075718, + "grad_norm": 1.0340241193771362, + "learning_rate": 3.654327209507421e-05, + "loss": 0.9664, + "step": 6004 + }, + { + "epoch": 0.41260134670880855, + "grad_norm": 1.1191208362579346, + "learning_rate": 3.6537952905796566e-05, + "loss": 1.0113, + "step": 6005 + }, + { + "epoch": 0.4126700563418991, + "grad_norm": 1.1432888507843018, + "learning_rate": 3.6532633052747534e-05, + "loss": 1.0269, + "step": 6006 + }, + { + "epoch": 0.4127387659749897, + "grad_norm": 1.238326072692871, + "learning_rate": 3.652731253623315e-05, + "loss": 1.0111, + "step": 6007 + }, + { + "epoch": 0.41280747560808023, + "grad_norm": 1.0875637531280518, + "learning_rate": 3.6521991356559504e-05, + "loss": 0.9437, + "step": 6008 + }, + { + "epoch": 0.4128761852411708, + "grad_norm": 1.063961386680603, + "learning_rate": 3.651666951403272e-05, + "loss": 0.9833, + "step": 6009 + }, + { + "epoch": 0.4129448948742614, + "grad_norm": 1.099212408065796, + "learning_rate": 3.6511347008958965e-05, + "loss": 0.946, + "step": 6010 + }, + { + "epoch": 0.4130136045073519, + "grad_norm": 1.0693004131317139, + "learning_rate": 3.650602384164443e-05, + "loss": 1.0007, + "step": 6011 + }, + { + "epoch": 0.4130823141404425, + "grad_norm": 1.132544755935669, + "learning_rate": 3.650070001239536e-05, + "loss": 1.0113, + "step": 6012 + }, + { + "epoch": 0.4131510237735331, + "grad_norm": 1.083000659942627, + "learning_rate": 3.649537552151803e-05, + "loss": 0.9815, + "step": 6013 + }, + { + "epoch": 0.4132197334066236, + "grad_norm": 1.130203366279602, + "learning_rate": 3.6490050369318753e-05, + "loss": 0.9908, + "step": 6014 + }, + { + "epoch": 0.4132884430397142, + "grad_norm": 1.064681887626648, + "learning_rate": 3.648472455610389e-05, + "loss": 0.946, + "step": 6015 + }, + { + "epoch": 0.4133571526728047, + "grad_norm": 1.062926173210144, + "learning_rate": 3.647939808217983e-05, + "loss": 1.0038, + "step": 6016 + }, + { + "epoch": 0.4134258623058953, + "grad_norm": 1.100295901298523, + "learning_rate": 3.647407094785299e-05, + "loss": 1.0073, + "step": 6017 + }, + { + "epoch": 0.41349457193898587, + "grad_norm": 1.0129601955413818, + "learning_rate": 3.6468743153429843e-05, + "loss": 0.9815, + "step": 6018 + }, + { + "epoch": 0.4135632815720764, + "grad_norm": 1.0893741846084595, + "learning_rate": 3.64634146992169e-05, + "loss": 0.9563, + "step": 6019 + }, + { + "epoch": 0.41363199120516697, + "grad_norm": 1.1345889568328857, + "learning_rate": 3.64580855855207e-05, + "loss": 0.9683, + "step": 6020 + }, + { + "epoch": 0.41370070083825755, + "grad_norm": 1.0619518756866455, + "learning_rate": 3.645275581264782e-05, + "loss": 0.9714, + "step": 6021 + }, + { + "epoch": 0.4137694104713481, + "grad_norm": 1.012769103050232, + "learning_rate": 3.644742538090488e-05, + "loss": 0.9729, + "step": 6022 + }, + { + "epoch": 0.41383812010443866, + "grad_norm": 1.10735285282135, + "learning_rate": 3.6442094290598544e-05, + "loss": 0.9701, + "step": 6023 + }, + { + "epoch": 0.4139068297375292, + "grad_norm": 1.0908575057983398, + "learning_rate": 3.64367625420355e-05, + "loss": 1.0257, + "step": 6024 + }, + { + "epoch": 0.41397553937061976, + "grad_norm": 1.1260226964950562, + "learning_rate": 3.6431430135522466e-05, + "loss": 0.9644, + "step": 6025 + }, + { + "epoch": 0.41404424900371034, + "grad_norm": 1.1308634281158447, + "learning_rate": 3.642609707136624e-05, + "loss": 0.9929, + "step": 6026 + }, + { + "epoch": 0.41411295863680087, + "grad_norm": 1.1252515316009521, + "learning_rate": 3.642076334987361e-05, + "loss": 0.9884, + "step": 6027 + }, + { + "epoch": 0.41418166826989145, + "grad_norm": 1.013276219367981, + "learning_rate": 3.641542897135143e-05, + "loss": 0.9454, + "step": 6028 + }, + { + "epoch": 0.414250377902982, + "grad_norm": 1.0239388942718506, + "learning_rate": 3.641009393610658e-05, + "loss": 0.9815, + "step": 6029 + }, + { + "epoch": 0.41431908753607255, + "grad_norm": 1.117706298828125, + "learning_rate": 3.640475824444599e-05, + "loss": 0.9928, + "step": 6030 + }, + { + "epoch": 0.41438779716916313, + "grad_norm": 1.2017848491668701, + "learning_rate": 3.639942189667661e-05, + "loss": 0.9936, + "step": 6031 + }, + { + "epoch": 0.41445650680225365, + "grad_norm": 1.1529871225357056, + "learning_rate": 3.6394084893105435e-05, + "loss": 0.9638, + "step": 6032 + }, + { + "epoch": 0.41452521643534423, + "grad_norm": 1.0919896364212036, + "learning_rate": 3.638874723403951e-05, + "loss": 0.9574, + "step": 6033 + }, + { + "epoch": 0.4145939260684348, + "grad_norm": 1.061051845550537, + "learning_rate": 3.638340891978589e-05, + "loss": 1.0056, + "step": 6034 + }, + { + "epoch": 0.41466263570152534, + "grad_norm": 1.0963695049285889, + "learning_rate": 3.63780699506517e-05, + "loss": 0.923, + "step": 6035 + }, + { + "epoch": 0.4147313453346159, + "grad_norm": 1.1079349517822266, + "learning_rate": 3.637273032694408e-05, + "loss": 0.9759, + "step": 6036 + }, + { + "epoch": 0.4148000549677065, + "grad_norm": 1.0964831113815308, + "learning_rate": 3.636739004897023e-05, + "loss": 0.9814, + "step": 6037 + }, + { + "epoch": 0.414868764600797, + "grad_norm": 1.1327465772628784, + "learning_rate": 3.6362049117037343e-05, + "loss": 1.0284, + "step": 6038 + }, + { + "epoch": 0.4149374742338876, + "grad_norm": 1.0865339040756226, + "learning_rate": 3.635670753145272e-05, + "loss": 1.0015, + "step": 6039 + }, + { + "epoch": 0.41500618386697813, + "grad_norm": 1.0861560106277466, + "learning_rate": 3.635136529252362e-05, + "loss": 0.9734, + "step": 6040 + }, + { + "epoch": 0.4150748935000687, + "grad_norm": 1.0644497871398926, + "learning_rate": 3.63460224005574e-05, + "loss": 1.0037, + "step": 6041 + }, + { + "epoch": 0.4151436031331593, + "grad_norm": 1.110547423362732, + "learning_rate": 3.6340678855861445e-05, + "loss": 0.9287, + "step": 6042 + }, + { + "epoch": 0.4152123127662498, + "grad_norm": 1.0563433170318604, + "learning_rate": 3.633533465874313e-05, + "loss": 0.9528, + "step": 6043 + }, + { + "epoch": 0.4152810223993404, + "grad_norm": 1.0800138711929321, + "learning_rate": 3.632998980950993e-05, + "loss": 0.9777, + "step": 6044 + }, + { + "epoch": 0.415349732032431, + "grad_norm": 1.104297399520874, + "learning_rate": 3.632464430846933e-05, + "loss": 0.9626, + "step": 6045 + }, + { + "epoch": 0.4154184416655215, + "grad_norm": 1.082276701927185, + "learning_rate": 3.631929815592884e-05, + "loss": 0.9603, + "step": 6046 + }, + { + "epoch": 0.4154871512986121, + "grad_norm": 1.080325961112976, + "learning_rate": 3.631395135219603e-05, + "loss": 0.9996, + "step": 6047 + }, + { + "epoch": 0.4155558609317026, + "grad_norm": 1.2146286964416504, + "learning_rate": 3.63086038975785e-05, + "loss": 1.0346, + "step": 6048 + }, + { + "epoch": 0.4156245705647932, + "grad_norm": 1.15412175655365, + "learning_rate": 3.630325579238389e-05, + "loss": 1.0124, + "step": 6049 + }, + { + "epoch": 0.41569328019788376, + "grad_norm": 1.0954006910324097, + "learning_rate": 3.629790703691985e-05, + "loss": 0.9983, + "step": 6050 + }, + { + "epoch": 0.4157619898309743, + "grad_norm": 1.02727210521698, + "learning_rate": 3.6292557631494126e-05, + "loss": 0.945, + "step": 6051 + }, + { + "epoch": 0.41583069946406487, + "grad_norm": 1.066541314125061, + "learning_rate": 3.628720757641443e-05, + "loss": 0.9899, + "step": 6052 + }, + { + "epoch": 0.41589940909715545, + "grad_norm": 1.2094612121582031, + "learning_rate": 3.628185687198857e-05, + "loss": 1.0064, + "step": 6053 + }, + { + "epoch": 0.415968118730246, + "grad_norm": 1.124634861946106, + "learning_rate": 3.6276505518524376e-05, + "loss": 0.9887, + "step": 6054 + }, + { + "epoch": 0.41603682836333655, + "grad_norm": 1.1254218816757202, + "learning_rate": 3.627115351632967e-05, + "loss": 0.9883, + "step": 6055 + }, + { + "epoch": 0.4161055379964271, + "grad_norm": 1.1126874685287476, + "learning_rate": 3.6265800865712396e-05, + "loss": 0.9662, + "step": 6056 + }, + { + "epoch": 0.41617424762951766, + "grad_norm": 1.101195216178894, + "learning_rate": 3.6260447566980454e-05, + "loss": 0.9935, + "step": 6057 + }, + { + "epoch": 0.41624295726260824, + "grad_norm": 1.1830614805221558, + "learning_rate": 3.6255093620441834e-05, + "loss": 1.0123, + "step": 6058 + }, + { + "epoch": 0.41631166689569876, + "grad_norm": 1.0882591009140015, + "learning_rate": 3.6249739026404544e-05, + "loss": 1.0058, + "step": 6059 + }, + { + "epoch": 0.41638037652878934, + "grad_norm": 1.0531905889511108, + "learning_rate": 3.624438378517662e-05, + "loss": 0.9347, + "step": 6060 + }, + { + "epoch": 0.4164490861618799, + "grad_norm": 1.030240774154663, + "learning_rate": 3.623902789706616e-05, + "loss": 0.9652, + "step": 6061 + }, + { + "epoch": 0.41651779579497045, + "grad_norm": 1.123207688331604, + "learning_rate": 3.623367136238127e-05, + "loss": 0.966, + "step": 6062 + }, + { + "epoch": 0.416586505428061, + "grad_norm": 1.125841736793518, + "learning_rate": 3.6228314181430126e-05, + "loss": 1.0314, + "step": 6063 + }, + { + "epoch": 0.41665521506115155, + "grad_norm": 1.0235817432403564, + "learning_rate": 3.622295635452091e-05, + "loss": 0.9525, + "step": 6064 + }, + { + "epoch": 0.41672392469424213, + "grad_norm": 1.0680993795394897, + "learning_rate": 3.621759788196186e-05, + "loss": 0.944, + "step": 6065 + }, + { + "epoch": 0.4167926343273327, + "grad_norm": 1.1308207511901855, + "learning_rate": 3.6212238764061234e-05, + "loss": 0.9923, + "step": 6066 + }, + { + "epoch": 0.41686134396042324, + "grad_norm": 1.096101999282837, + "learning_rate": 3.6206879001127357e-05, + "loss": 0.9636, + "step": 6067 + }, + { + "epoch": 0.4169300535935138, + "grad_norm": 1.109940767288208, + "learning_rate": 3.620151859346856e-05, + "loss": 1.0235, + "step": 6068 + }, + { + "epoch": 0.4169987632266044, + "grad_norm": 1.098802089691162, + "learning_rate": 3.619615754139323e-05, + "loss": 0.9368, + "step": 6069 + }, + { + "epoch": 0.4170674728596949, + "grad_norm": 1.0938299894332886, + "learning_rate": 3.619079584520978e-05, + "loss": 0.9647, + "step": 6070 + }, + { + "epoch": 0.4171361824927855, + "grad_norm": 1.120007872581482, + "learning_rate": 3.618543350522667e-05, + "loss": 1.0026, + "step": 6071 + }, + { + "epoch": 0.417204892125876, + "grad_norm": 1.0958913564682007, + "learning_rate": 3.6180070521752394e-05, + "loss": 0.9965, + "step": 6072 + }, + { + "epoch": 0.4172736017589666, + "grad_norm": 1.0853767395019531, + "learning_rate": 3.617470689509548e-05, + "loss": 0.9645, + "step": 6073 + }, + { + "epoch": 0.4173423113920572, + "grad_norm": 1.0735288858413696, + "learning_rate": 3.616934262556448e-05, + "loss": 0.9676, + "step": 6074 + }, + { + "epoch": 0.4174110210251477, + "grad_norm": 1.0988023281097412, + "learning_rate": 3.616397771346802e-05, + "loss": 0.9729, + "step": 6075 + }, + { + "epoch": 0.4174797306582383, + "grad_norm": 1.0369043350219727, + "learning_rate": 3.615861215911472e-05, + "loss": 0.9409, + "step": 6076 + }, + { + "epoch": 0.41754844029132887, + "grad_norm": 1.084027647972107, + "learning_rate": 3.615324596281328e-05, + "loss": 0.9844, + "step": 6077 + }, + { + "epoch": 0.4176171499244194, + "grad_norm": 1.037423849105835, + "learning_rate": 3.61478791248724e-05, + "loss": 0.9637, + "step": 6078 + }, + { + "epoch": 0.41768585955751, + "grad_norm": 1.0560184717178345, + "learning_rate": 3.614251164560082e-05, + "loss": 0.9745, + "step": 6079 + }, + { + "epoch": 0.4177545691906005, + "grad_norm": 1.0677518844604492, + "learning_rate": 3.6137143525307344e-05, + "loss": 1.0126, + "step": 6080 + }, + { + "epoch": 0.4178232788236911, + "grad_norm": 1.1614817380905151, + "learning_rate": 3.613177476430079e-05, + "loss": 0.9678, + "step": 6081 + }, + { + "epoch": 0.41789198845678166, + "grad_norm": 1.0713070631027222, + "learning_rate": 3.612640536289002e-05, + "loss": 0.9728, + "step": 6082 + }, + { + "epoch": 0.4179606980898722, + "grad_norm": 1.0608179569244385, + "learning_rate": 3.6121035321383935e-05, + "loss": 1.0253, + "step": 6083 + }, + { + "epoch": 0.41802940772296276, + "grad_norm": 1.1800175905227661, + "learning_rate": 3.6115664640091474e-05, + "loss": 1.031, + "step": 6084 + }, + { + "epoch": 0.41809811735605334, + "grad_norm": 1.1008601188659668, + "learning_rate": 3.6110293319321595e-05, + "loss": 0.9494, + "step": 6085 + }, + { + "epoch": 0.41816682698914387, + "grad_norm": 1.043919324874878, + "learning_rate": 3.610492135938332e-05, + "loss": 0.9944, + "step": 6086 + }, + { + "epoch": 0.41823553662223445, + "grad_norm": 1.1422995328903198, + "learning_rate": 3.6099548760585686e-05, + "loss": 0.9793, + "step": 6087 + }, + { + "epoch": 0.418304246255325, + "grad_norm": 1.1025114059448242, + "learning_rate": 3.609417552323778e-05, + "loss": 0.9727, + "step": 6088 + }, + { + "epoch": 0.41837295588841555, + "grad_norm": 1.1464221477508545, + "learning_rate": 3.608880164764873e-05, + "loss": 0.9904, + "step": 6089 + }, + { + "epoch": 0.41844166552150613, + "grad_norm": 1.1049261093139648, + "learning_rate": 3.6083427134127675e-05, + "loss": 0.9514, + "step": 6090 + }, + { + "epoch": 0.41851037515459666, + "grad_norm": 1.0912660360336304, + "learning_rate": 3.607805198298382e-05, + "loss": 1.0055, + "step": 6091 + }, + { + "epoch": 0.41857908478768724, + "grad_norm": 1.080983281135559, + "learning_rate": 3.607267619452638e-05, + "loss": 0.989, + "step": 6092 + }, + { + "epoch": 0.4186477944207778, + "grad_norm": 1.0364291667938232, + "learning_rate": 3.606729976906464e-05, + "loss": 0.9367, + "step": 6093 + }, + { + "epoch": 0.41871650405386834, + "grad_norm": 1.0703353881835938, + "learning_rate": 3.606192270690788e-05, + "loss": 0.934, + "step": 6094 + }, + { + "epoch": 0.4187852136869589, + "grad_norm": 1.1553685665130615, + "learning_rate": 3.6056545008365466e-05, + "loss": 1.0256, + "step": 6095 + }, + { + "epoch": 0.41885392332004945, + "grad_norm": 1.1333813667297363, + "learning_rate": 3.605116667374675e-05, + "loss": 1.0106, + "step": 6096 + }, + { + "epoch": 0.41892263295314003, + "grad_norm": 1.0859729051589966, + "learning_rate": 3.6045787703361154e-05, + "loss": 0.9954, + "step": 6097 + }, + { + "epoch": 0.4189913425862306, + "grad_norm": 1.0924965143203735, + "learning_rate": 3.604040809751813e-05, + "loss": 0.9962, + "step": 6098 + }, + { + "epoch": 0.41906005221932113, + "grad_norm": 1.1844233274459839, + "learning_rate": 3.603502785652716e-05, + "loss": 1.0102, + "step": 6099 + }, + { + "epoch": 0.4191287618524117, + "grad_norm": 1.0253279209136963, + "learning_rate": 3.602964698069776e-05, + "loss": 0.9806, + "step": 6100 + }, + { + "epoch": 0.4191974714855023, + "grad_norm": 1.2167679071426392, + "learning_rate": 3.6024265470339504e-05, + "loss": 1.0396, + "step": 6101 + }, + { + "epoch": 0.4192661811185928, + "grad_norm": 1.027276873588562, + "learning_rate": 3.6018883325761975e-05, + "loss": 0.988, + "step": 6102 + }, + { + "epoch": 0.4193348907516834, + "grad_norm": 1.1046133041381836, + "learning_rate": 3.60135005472748e-05, + "loss": 1.0022, + "step": 6103 + }, + { + "epoch": 0.4194036003847739, + "grad_norm": 1.067599892616272, + "learning_rate": 3.6008117135187656e-05, + "loss": 0.9427, + "step": 6104 + }, + { + "epoch": 0.4194723100178645, + "grad_norm": 1.1333386898040771, + "learning_rate": 3.600273308981025e-05, + "loss": 0.9739, + "step": 6105 + }, + { + "epoch": 0.4195410196509551, + "grad_norm": 1.0629535913467407, + "learning_rate": 3.599734841145231e-05, + "loss": 1.0068, + "step": 6106 + }, + { + "epoch": 0.4196097292840456, + "grad_norm": 1.0694295167922974, + "learning_rate": 3.599196310042363e-05, + "loss": 0.9477, + "step": 6107 + }, + { + "epoch": 0.4196784389171362, + "grad_norm": 1.0682631731033325, + "learning_rate": 3.598657715703401e-05, + "loss": 0.9479, + "step": 6108 + }, + { + "epoch": 0.41974714855022677, + "grad_norm": 1.072402000427246, + "learning_rate": 3.5981190581593306e-05, + "loss": 0.9659, + "step": 6109 + }, + { + "epoch": 0.4198158581833173, + "grad_norm": 1.118033528327942, + "learning_rate": 3.5975803374411396e-05, + "loss": 0.9819, + "step": 6110 + }, + { + "epoch": 0.41988456781640787, + "grad_norm": 1.0591927766799927, + "learning_rate": 3.597041553579822e-05, + "loss": 0.9682, + "step": 6111 + }, + { + "epoch": 0.4199532774494984, + "grad_norm": 1.1167811155319214, + "learning_rate": 3.596502706606373e-05, + "loss": 0.9919, + "step": 6112 + }, + { + "epoch": 0.420021987082589, + "grad_norm": 1.0649945735931396, + "learning_rate": 3.5959637965517917e-05, + "loss": 0.9795, + "step": 6113 + }, + { + "epoch": 0.42009069671567956, + "grad_norm": 1.1006487607955933, + "learning_rate": 3.595424823447081e-05, + "loss": 0.9483, + "step": 6114 + }, + { + "epoch": 0.4201594063487701, + "grad_norm": 1.0754603147506714, + "learning_rate": 3.594885787323248e-05, + "loss": 1.0037, + "step": 6115 + }, + { + "epoch": 0.42022811598186066, + "grad_norm": 1.0790232419967651, + "learning_rate": 3.594346688211303e-05, + "loss": 0.9775, + "step": 6116 + }, + { + "epoch": 0.42029682561495124, + "grad_norm": 1.1823327541351318, + "learning_rate": 3.593807526142261e-05, + "loss": 0.9865, + "step": 6117 + }, + { + "epoch": 0.42036553524804177, + "grad_norm": 1.0441560745239258, + "learning_rate": 3.593268301147139e-05, + "loss": 0.9839, + "step": 6118 + }, + { + "epoch": 0.42043424488113235, + "grad_norm": 1.0815716981887817, + "learning_rate": 3.592729013256958e-05, + "loss": 0.9922, + "step": 6119 + }, + { + "epoch": 0.42050295451422287, + "grad_norm": 1.0130969285964966, + "learning_rate": 3.592189662502743e-05, + "loss": 1.0014, + "step": 6120 + }, + { + "epoch": 0.42057166414731345, + "grad_norm": 1.0780633687973022, + "learning_rate": 3.5916502489155225e-05, + "loss": 0.9831, + "step": 6121 + }, + { + "epoch": 0.42064037378040403, + "grad_norm": 1.1102144718170166, + "learning_rate": 3.591110772526329e-05, + "loss": 0.971, + "step": 6122 + }, + { + "epoch": 0.42070908341349456, + "grad_norm": 1.0670403242111206, + "learning_rate": 3.5905712333661975e-05, + "loss": 0.9966, + "step": 6123 + }, + { + "epoch": 0.42077779304658514, + "grad_norm": 1.1473647356033325, + "learning_rate": 3.590031631466169e-05, + "loss": 1.0219, + "step": 6124 + }, + { + "epoch": 0.4208465026796757, + "grad_norm": 1.0155792236328125, + "learning_rate": 3.589491966857285e-05, + "loss": 0.9919, + "step": 6125 + }, + { + "epoch": 0.42091521231276624, + "grad_norm": 1.1017817258834839, + "learning_rate": 3.588952239570592e-05, + "loss": 0.9879, + "step": 6126 + }, + { + "epoch": 0.4209839219458568, + "grad_norm": 1.099192500114441, + "learning_rate": 3.588412449637141e-05, + "loss": 1.0113, + "step": 6127 + }, + { + "epoch": 0.42105263157894735, + "grad_norm": 1.1297956705093384, + "learning_rate": 3.587872597087986e-05, + "loss": 0.9795, + "step": 6128 + }, + { + "epoch": 0.4211213412120379, + "grad_norm": 1.2044613361358643, + "learning_rate": 3.587332681954183e-05, + "loss": 0.9817, + "step": 6129 + }, + { + "epoch": 0.4211900508451285, + "grad_norm": 1.1686432361602783, + "learning_rate": 3.586792704266794e-05, + "loss": 1.0039, + "step": 6130 + }, + { + "epoch": 0.42125876047821903, + "grad_norm": 1.1271470785140991, + "learning_rate": 3.5862526640568824e-05, + "loss": 0.9893, + "step": 6131 + }, + { + "epoch": 0.4213274701113096, + "grad_norm": 1.0711157321929932, + "learning_rate": 3.585712561355519e-05, + "loss": 1.0041, + "step": 6132 + }, + { + "epoch": 0.4213961797444002, + "grad_norm": 1.0278412103652954, + "learning_rate": 3.585172396193773e-05, + "loss": 0.9694, + "step": 6133 + }, + { + "epoch": 0.4214648893774907, + "grad_norm": 1.0952260494232178, + "learning_rate": 3.5846321686027204e-05, + "loss": 0.9525, + "step": 6134 + }, + { + "epoch": 0.4215335990105813, + "grad_norm": 1.1388616561889648, + "learning_rate": 3.5840918786134416e-05, + "loss": 0.984, + "step": 6135 + }, + { + "epoch": 0.4216023086436718, + "grad_norm": 1.020870566368103, + "learning_rate": 3.583551526257017e-05, + "loss": 0.983, + "step": 6136 + }, + { + "epoch": 0.4216710182767624, + "grad_norm": 1.1178871393203735, + "learning_rate": 3.5830111115645345e-05, + "loss": 0.9667, + "step": 6137 + }, + { + "epoch": 0.421739727909853, + "grad_norm": 1.094468355178833, + "learning_rate": 3.5824706345670825e-05, + "loss": 0.9896, + "step": 6138 + }, + { + "epoch": 0.4218084375429435, + "grad_norm": 1.0744693279266357, + "learning_rate": 3.581930095295755e-05, + "loss": 0.969, + "step": 6139 + }, + { + "epoch": 0.4218771471760341, + "grad_norm": 1.038970947265625, + "learning_rate": 3.581389493781649e-05, + "loss": 0.9564, + "step": 6140 + }, + { + "epoch": 0.42194585680912466, + "grad_norm": 1.1846939325332642, + "learning_rate": 3.580848830055865e-05, + "loss": 1.0137, + "step": 6141 + }, + { + "epoch": 0.4220145664422152, + "grad_norm": 1.1062902212142944, + "learning_rate": 3.580308104149507e-05, + "loss": 0.9636, + "step": 6142 + }, + { + "epoch": 0.42208327607530577, + "grad_norm": 1.0923384428024292, + "learning_rate": 3.579767316093682e-05, + "loss": 0.9678, + "step": 6143 + }, + { + "epoch": 0.4221519857083963, + "grad_norm": 1.0527158975601196, + "learning_rate": 3.579226465919501e-05, + "loss": 0.9553, + "step": 6144 + }, + { + "epoch": 0.4222206953414869, + "grad_norm": 1.1179213523864746, + "learning_rate": 3.578685553658081e-05, + "loss": 1.0045, + "step": 6145 + }, + { + "epoch": 0.42228940497457745, + "grad_norm": 1.1116585731506348, + "learning_rate": 3.578144579340538e-05, + "loss": 0.9545, + "step": 6146 + }, + { + "epoch": 0.422358114607668, + "grad_norm": 1.028160572052002, + "learning_rate": 3.5776035429979945e-05, + "loss": 0.9442, + "step": 6147 + }, + { + "epoch": 0.42242682424075856, + "grad_norm": 1.1488438844680786, + "learning_rate": 3.5770624446615765e-05, + "loss": 0.9931, + "step": 6148 + }, + { + "epoch": 0.42249553387384914, + "grad_norm": 1.0542049407958984, + "learning_rate": 3.576521284362413e-05, + "loss": 1.004, + "step": 6149 + }, + { + "epoch": 0.42256424350693966, + "grad_norm": 1.1871914863586426, + "learning_rate": 3.5759800621316356e-05, + "loss": 1.0182, + "step": 6150 + }, + { + "epoch": 0.42263295314003024, + "grad_norm": 1.1061383485794067, + "learning_rate": 3.5754387780003824e-05, + "loss": 0.9574, + "step": 6151 + }, + { + "epoch": 0.42270166277312077, + "grad_norm": 1.0546103715896606, + "learning_rate": 3.5748974319997916e-05, + "loss": 0.9522, + "step": 6152 + }, + { + "epoch": 0.42277037240621135, + "grad_norm": 1.0223159790039062, + "learning_rate": 3.574356024161007e-05, + "loss": 0.9707, + "step": 6153 + }, + { + "epoch": 0.42283908203930193, + "grad_norm": 0.9917662143707275, + "learning_rate": 3.5738145545151746e-05, + "loss": 0.9782, + "step": 6154 + }, + { + "epoch": 0.42290779167239245, + "grad_norm": 1.097856879234314, + "learning_rate": 3.5732730230934466e-05, + "loss": 0.9858, + "step": 6155 + }, + { + "epoch": 0.42297650130548303, + "grad_norm": 1.1313655376434326, + "learning_rate": 3.5727314299269755e-05, + "loss": 0.9815, + "step": 6156 + }, + { + "epoch": 0.4230452109385736, + "grad_norm": 1.0676463842391968, + "learning_rate": 3.572189775046919e-05, + "loss": 1.0065, + "step": 6157 + }, + { + "epoch": 0.42311392057166414, + "grad_norm": 1.0772939920425415, + "learning_rate": 3.571648058484439e-05, + "loss": 1.0111, + "step": 6158 + }, + { + "epoch": 0.4231826302047547, + "grad_norm": 1.0629637241363525, + "learning_rate": 3.5711062802706995e-05, + "loss": 0.989, + "step": 6159 + }, + { + "epoch": 0.42325133983784524, + "grad_norm": 1.1085423231124878, + "learning_rate": 3.570564440436869e-05, + "loss": 1.0412, + "step": 6160 + }, + { + "epoch": 0.4233200494709358, + "grad_norm": 1.0274909734725952, + "learning_rate": 3.5700225390141196e-05, + "loss": 0.969, + "step": 6161 + }, + { + "epoch": 0.4233887591040264, + "grad_norm": 1.085405707359314, + "learning_rate": 3.569480576033625e-05, + "loss": 0.9479, + "step": 6162 + }, + { + "epoch": 0.4234574687371169, + "grad_norm": 1.0589176416397095, + "learning_rate": 3.568938551526565e-05, + "loss": 1.0263, + "step": 6163 + }, + { + "epoch": 0.4235261783702075, + "grad_norm": 1.0781489610671997, + "learning_rate": 3.568396465524122e-05, + "loss": 0.9696, + "step": 6164 + }, + { + "epoch": 0.4235948880032981, + "grad_norm": 1.1026661396026611, + "learning_rate": 3.567854318057481e-05, + "loss": 0.9688, + "step": 6165 + }, + { + "epoch": 0.4236635976363886, + "grad_norm": 1.0467414855957031, + "learning_rate": 3.5673121091578336e-05, + "loss": 0.9652, + "step": 6166 + }, + { + "epoch": 0.4237323072694792, + "grad_norm": 1.0183144807815552, + "learning_rate": 3.5667698388563705e-05, + "loss": 0.9702, + "step": 6167 + }, + { + "epoch": 0.4238010169025697, + "grad_norm": 1.0060534477233887, + "learning_rate": 3.56622750718429e-05, + "loss": 0.9637, + "step": 6168 + }, + { + "epoch": 0.4238697265356603, + "grad_norm": 1.0880088806152344, + "learning_rate": 3.56568511417279e-05, + "loss": 0.9764, + "step": 6169 + }, + { + "epoch": 0.4239384361687509, + "grad_norm": 1.1004441976547241, + "learning_rate": 3.565142659853075e-05, + "loss": 0.968, + "step": 6170 + }, + { + "epoch": 0.4240071458018414, + "grad_norm": 1.0190173387527466, + "learning_rate": 3.5646001442563534e-05, + "loss": 0.9521, + "step": 6171 + }, + { + "epoch": 0.424075855434932, + "grad_norm": 1.1300297975540161, + "learning_rate": 3.564057567413834e-05, + "loss": 0.9881, + "step": 6172 + }, + { + "epoch": 0.42414456506802256, + "grad_norm": 1.0802944898605347, + "learning_rate": 3.5635149293567316e-05, + "loss": 0.9819, + "step": 6173 + }, + { + "epoch": 0.4242132747011131, + "grad_norm": 1.0866843461990356, + "learning_rate": 3.562972230116264e-05, + "loss": 1.025, + "step": 6174 + }, + { + "epoch": 0.42428198433420367, + "grad_norm": 1.0016698837280273, + "learning_rate": 3.562429469723651e-05, + "loss": 0.9377, + "step": 6175 + }, + { + "epoch": 0.4243506939672942, + "grad_norm": 1.097575306892395, + "learning_rate": 3.56188664821012e-05, + "loss": 0.9365, + "step": 6176 + }, + { + "epoch": 0.42441940360038477, + "grad_norm": 1.0930150747299194, + "learning_rate": 3.5613437656068964e-05, + "loss": 0.9883, + "step": 6177 + }, + { + "epoch": 0.42448811323347535, + "grad_norm": 1.0600684881210327, + "learning_rate": 3.560800821945214e-05, + "loss": 0.9798, + "step": 6178 + }, + { + "epoch": 0.4245568228665659, + "grad_norm": 1.1010329723358154, + "learning_rate": 3.5602578172563065e-05, + "loss": 0.9698, + "step": 6179 + }, + { + "epoch": 0.42462553249965646, + "grad_norm": 1.0741534233093262, + "learning_rate": 3.559714751571413e-05, + "loss": 0.9711, + "step": 6180 + }, + { + "epoch": 0.42469424213274704, + "grad_norm": 1.0091314315795898, + "learning_rate": 3.5591716249217765e-05, + "loss": 0.9479, + "step": 6181 + }, + { + "epoch": 0.42476295176583756, + "grad_norm": 1.0967844724655151, + "learning_rate": 3.5586284373386426e-05, + "loss": 0.9388, + "step": 6182 + }, + { + "epoch": 0.42483166139892814, + "grad_norm": 1.0369800329208374, + "learning_rate": 3.55808518885326e-05, + "loss": 1.0293, + "step": 6183 + }, + { + "epoch": 0.42490037103201866, + "grad_norm": 1.1233566999435425, + "learning_rate": 3.557541879496881e-05, + "loss": 1.0095, + "step": 6184 + }, + { + "epoch": 0.42496908066510924, + "grad_norm": 1.2163654565811157, + "learning_rate": 3.556998509300763e-05, + "loss": 0.9965, + "step": 6185 + }, + { + "epoch": 0.4250377902981998, + "grad_norm": 1.0316663980484009, + "learning_rate": 3.556455078296165e-05, + "loss": 0.975, + "step": 6186 + }, + { + "epoch": 0.42510649993129035, + "grad_norm": 1.0386861562728882, + "learning_rate": 3.5559115865143505e-05, + "loss": 0.9463, + "step": 6187 + }, + { + "epoch": 0.42517520956438093, + "grad_norm": 1.096934199333191, + "learning_rate": 3.555368033986586e-05, + "loss": 0.977, + "step": 6188 + }, + { + "epoch": 0.4252439191974715, + "grad_norm": 1.1220917701721191, + "learning_rate": 3.5548244207441425e-05, + "loss": 1.0179, + "step": 6189 + }, + { + "epoch": 0.42531262883056203, + "grad_norm": 1.1465216875076294, + "learning_rate": 3.5542807468182934e-05, + "loss": 1.0057, + "step": 6190 + }, + { + "epoch": 0.4253813384636526, + "grad_norm": 1.069791555404663, + "learning_rate": 3.553737012240315e-05, + "loss": 0.9731, + "step": 6191 + }, + { + "epoch": 0.42545004809674314, + "grad_norm": 1.1743125915527344, + "learning_rate": 3.5531932170414896e-05, + "loss": 0.9565, + "step": 6192 + }, + { + "epoch": 0.4255187577298337, + "grad_norm": 1.0324777364730835, + "learning_rate": 3.5526493612530995e-05, + "loss": 0.9457, + "step": 6193 + }, + { + "epoch": 0.4255874673629243, + "grad_norm": 1.0829493999481201, + "learning_rate": 3.552105444906435e-05, + "loss": 0.9853, + "step": 6194 + }, + { + "epoch": 0.4256561769960148, + "grad_norm": 1.0261976718902588, + "learning_rate": 3.5515614680327845e-05, + "loss": 0.9673, + "step": 6195 + }, + { + "epoch": 0.4257248866291054, + "grad_norm": 1.0106616020202637, + "learning_rate": 3.5510174306634444e-05, + "loss": 0.9863, + "step": 6196 + }, + { + "epoch": 0.425793596262196, + "grad_norm": 1.0648868083953857, + "learning_rate": 3.550473332829713e-05, + "loss": 0.9823, + "step": 6197 + }, + { + "epoch": 0.4258623058952865, + "grad_norm": 1.1037708520889282, + "learning_rate": 3.54992917456289e-05, + "loss": 0.9833, + "step": 6198 + }, + { + "epoch": 0.4259310155283771, + "grad_norm": 1.0830854177474976, + "learning_rate": 3.549384955894283e-05, + "loss": 0.9458, + "step": 6199 + }, + { + "epoch": 0.4259997251614676, + "grad_norm": 1.068798303604126, + "learning_rate": 3.5488406768551984e-05, + "loss": 0.9631, + "step": 6200 + }, + { + "epoch": 0.4260684347945582, + "grad_norm": 1.1303075551986694, + "learning_rate": 3.5482963374769495e-05, + "loss": 0.955, + "step": 6201 + }, + { + "epoch": 0.4261371444276488, + "grad_norm": 1.1623117923736572, + "learning_rate": 3.547751937790851e-05, + "loss": 1.0107, + "step": 6202 + }, + { + "epoch": 0.4262058540607393, + "grad_norm": 1.0531898736953735, + "learning_rate": 3.547207477828223e-05, + "loss": 0.9389, + "step": 6203 + }, + { + "epoch": 0.4262745636938299, + "grad_norm": 1.1001943349838257, + "learning_rate": 3.546662957620387e-05, + "loss": 0.9618, + "step": 6204 + }, + { + "epoch": 0.42634327332692046, + "grad_norm": 1.088127613067627, + "learning_rate": 3.5461183771986694e-05, + "loss": 0.9224, + "step": 6205 + }, + { + "epoch": 0.426411982960011, + "grad_norm": 1.16457200050354, + "learning_rate": 3.545573736594399e-05, + "loss": 1.0261, + "step": 6206 + }, + { + "epoch": 0.42648069259310156, + "grad_norm": 1.1342041492462158, + "learning_rate": 3.54502903583891e-05, + "loss": 0.9763, + "step": 6207 + }, + { + "epoch": 0.4265494022261921, + "grad_norm": 1.0569438934326172, + "learning_rate": 3.544484274963538e-05, + "loss": 0.9691, + "step": 6208 + }, + { + "epoch": 0.42661811185928267, + "grad_norm": 1.081949234008789, + "learning_rate": 3.543939453999621e-05, + "loss": 0.9808, + "step": 6209 + }, + { + "epoch": 0.42668682149237325, + "grad_norm": 1.0458087921142578, + "learning_rate": 3.5433945729785044e-05, + "loss": 0.9282, + "step": 6210 + }, + { + "epoch": 0.42675553112546377, + "grad_norm": 1.0016939640045166, + "learning_rate": 3.542849631931534e-05, + "loss": 0.9617, + "step": 6211 + }, + { + "epoch": 0.42682424075855435, + "grad_norm": 1.0335866212844849, + "learning_rate": 3.542304630890061e-05, + "loss": 0.9609, + "step": 6212 + }, + { + "epoch": 0.42689295039164493, + "grad_norm": 1.076625108718872, + "learning_rate": 3.541759569885437e-05, + "loss": 0.9772, + "step": 6213 + }, + { + "epoch": 0.42696166002473546, + "grad_norm": 1.0505797863006592, + "learning_rate": 3.541214448949021e-05, + "loss": 0.9608, + "step": 6214 + }, + { + "epoch": 0.42703036965782604, + "grad_norm": 0.9684593677520752, + "learning_rate": 3.540669268112172e-05, + "loss": 0.9176, + "step": 6215 + }, + { + "epoch": 0.42709907929091656, + "grad_norm": 1.045006513595581, + "learning_rate": 3.540124027406255e-05, + "loss": 0.964, + "step": 6216 + }, + { + "epoch": 0.42716778892400714, + "grad_norm": 1.0981398820877075, + "learning_rate": 3.539578726862637e-05, + "loss": 0.9724, + "step": 6217 + }, + { + "epoch": 0.4272364985570977, + "grad_norm": 1.0932353734970093, + "learning_rate": 3.539033366512687e-05, + "loss": 0.9323, + "step": 6218 + }, + { + "epoch": 0.42730520819018825, + "grad_norm": 1.0513256788253784, + "learning_rate": 3.5384879463877834e-05, + "loss": 0.9871, + "step": 6219 + }, + { + "epoch": 0.4273739178232788, + "grad_norm": 1.1234941482543945, + "learning_rate": 3.5379424665193006e-05, + "loss": 1.0463, + "step": 6220 + }, + { + "epoch": 0.4274426274563694, + "grad_norm": 1.0431244373321533, + "learning_rate": 3.53739692693862e-05, + "loss": 0.9846, + "step": 6221 + }, + { + "epoch": 0.42751133708945993, + "grad_norm": 1.0628645420074463, + "learning_rate": 3.5368513276771276e-05, + "loss": 0.9522, + "step": 6222 + }, + { + "epoch": 0.4275800467225505, + "grad_norm": 1.142438292503357, + "learning_rate": 3.53630566876621e-05, + "loss": 0.989, + "step": 6223 + }, + { + "epoch": 0.42764875635564104, + "grad_norm": 1.1292489767074585, + "learning_rate": 3.535759950237259e-05, + "loss": 1.0234, + "step": 6224 + }, + { + "epoch": 0.4277174659887316, + "grad_norm": 1.0968966484069824, + "learning_rate": 3.5352141721216695e-05, + "loss": 1.0218, + "step": 6225 + }, + { + "epoch": 0.4277861756218222, + "grad_norm": 1.1641583442687988, + "learning_rate": 3.53466833445084e-05, + "loss": 0.9681, + "step": 6226 + }, + { + "epoch": 0.4278548852549127, + "grad_norm": 1.1642780303955078, + "learning_rate": 3.534122437256173e-05, + "loss": 1.0049, + "step": 6227 + }, + { + "epoch": 0.4279235948880033, + "grad_norm": 1.1285420656204224, + "learning_rate": 3.5335764805690715e-05, + "loss": 1.022, + "step": 6228 + }, + { + "epoch": 0.4279923045210939, + "grad_norm": 1.1125426292419434, + "learning_rate": 3.533030464420946e-05, + "loss": 1.0265, + "step": 6229 + }, + { + "epoch": 0.4280610141541844, + "grad_norm": 1.113362193107605, + "learning_rate": 3.532484388843207e-05, + "loss": 0.9714, + "step": 6230 + }, + { + "epoch": 0.428129723787275, + "grad_norm": 1.0607067346572876, + "learning_rate": 3.531938253867272e-05, + "loss": 1.0031, + "step": 6231 + }, + { + "epoch": 0.4281984334203655, + "grad_norm": 1.1044435501098633, + "learning_rate": 3.531392059524557e-05, + "loss": 0.9615, + "step": 6232 + }, + { + "epoch": 0.4282671430534561, + "grad_norm": 1.1056221723556519, + "learning_rate": 3.5308458058464864e-05, + "loss": 0.973, + "step": 6233 + }, + { + "epoch": 0.42833585268654667, + "grad_norm": 1.0719102621078491, + "learning_rate": 3.530299492864485e-05, + "loss": 0.9937, + "step": 6234 + }, + { + "epoch": 0.4284045623196372, + "grad_norm": 1.1508020162582397, + "learning_rate": 3.529753120609982e-05, + "loss": 1.0031, + "step": 6235 + }, + { + "epoch": 0.4284732719527278, + "grad_norm": 1.1168593168258667, + "learning_rate": 3.52920668911441e-05, + "loss": 0.9679, + "step": 6236 + }, + { + "epoch": 0.42854198158581835, + "grad_norm": 1.111945390701294, + "learning_rate": 3.528660198409205e-05, + "loss": 0.9769, + "step": 6237 + }, + { + "epoch": 0.4286106912189089, + "grad_norm": 1.121957540512085, + "learning_rate": 3.528113648525805e-05, + "loss": 1.0199, + "step": 6238 + }, + { + "epoch": 0.42867940085199946, + "grad_norm": 1.0874910354614258, + "learning_rate": 3.527567039495654e-05, + "loss": 1.003, + "step": 6239 + }, + { + "epoch": 0.42874811048509, + "grad_norm": 1.1704007387161255, + "learning_rate": 3.527020371350198e-05, + "loss": 0.9672, + "step": 6240 + }, + { + "epoch": 0.42881682011818056, + "grad_norm": 1.0609411001205444, + "learning_rate": 3.5264736441208864e-05, + "loss": 0.9411, + "step": 6241 + }, + { + "epoch": 0.42888552975127114, + "grad_norm": 1.0496739149093628, + "learning_rate": 3.525926857839172e-05, + "loss": 0.9778, + "step": 6242 + }, + { + "epoch": 0.42895423938436167, + "grad_norm": 1.0769106149673462, + "learning_rate": 3.52538001253651e-05, + "loss": 0.9778, + "step": 6243 + }, + { + "epoch": 0.42902294901745225, + "grad_norm": 1.0717382431030273, + "learning_rate": 3.5248331082443625e-05, + "loss": 0.9721, + "step": 6244 + }, + { + "epoch": 0.42909165865054283, + "grad_norm": 1.1359022855758667, + "learning_rate": 3.524286144994191e-05, + "loss": 1.017, + "step": 6245 + }, + { + "epoch": 0.42916036828363335, + "grad_norm": 1.1508814096450806, + "learning_rate": 3.523739122817461e-05, + "loss": 0.9774, + "step": 6246 + }, + { + "epoch": 0.42922907791672393, + "grad_norm": 1.0055263042449951, + "learning_rate": 3.5231920417456446e-05, + "loss": 0.9523, + "step": 6247 + }, + { + "epoch": 0.42929778754981446, + "grad_norm": 1.1405835151672363, + "learning_rate": 3.522644901810213e-05, + "loss": 1.0036, + "step": 6248 + }, + { + "epoch": 0.42936649718290504, + "grad_norm": 1.0966780185699463, + "learning_rate": 3.5220977030426444e-05, + "loss": 1.0115, + "step": 6249 + }, + { + "epoch": 0.4294352068159956, + "grad_norm": 1.1422765254974365, + "learning_rate": 3.521550445474418e-05, + "loss": 0.9997, + "step": 6250 + }, + { + "epoch": 0.42950391644908614, + "grad_norm": 1.114567756652832, + "learning_rate": 3.5210031291370174e-05, + "loss": 0.9802, + "step": 6251 + }, + { + "epoch": 0.4295726260821767, + "grad_norm": 1.0855478048324585, + "learning_rate": 3.5204557540619294e-05, + "loss": 0.9544, + "step": 6252 + }, + { + "epoch": 0.4296413357152673, + "grad_norm": 1.107183814048767, + "learning_rate": 3.5199083202806433e-05, + "loss": 0.9596, + "step": 6253 + }, + { + "epoch": 0.4297100453483578, + "grad_norm": 1.0974757671356201, + "learning_rate": 3.519360827824654e-05, + "loss": 0.9828, + "step": 6254 + }, + { + "epoch": 0.4297787549814484, + "grad_norm": 1.0674840211868286, + "learning_rate": 3.518813276725458e-05, + "loss": 0.9505, + "step": 6255 + }, + { + "epoch": 0.42984746461453893, + "grad_norm": 1.0888313055038452, + "learning_rate": 3.518265667014555e-05, + "loss": 0.9317, + "step": 6256 + }, + { + "epoch": 0.4299161742476295, + "grad_norm": 1.1822656393051147, + "learning_rate": 3.51771799872345e-05, + "loss": 0.9675, + "step": 6257 + }, + { + "epoch": 0.4299848838807201, + "grad_norm": 1.161669373512268, + "learning_rate": 3.517170271883648e-05, + "loss": 1.0132, + "step": 6258 + }, + { + "epoch": 0.4300535935138106, + "grad_norm": 1.055323839187622, + "learning_rate": 3.5166224865266607e-05, + "loss": 0.9624, + "step": 6259 + }, + { + "epoch": 0.4301223031469012, + "grad_norm": 1.0903027057647705, + "learning_rate": 3.516074642684002e-05, + "loss": 0.997, + "step": 6260 + }, + { + "epoch": 0.4301910127799918, + "grad_norm": 1.1360114812850952, + "learning_rate": 3.515526740387188e-05, + "loss": 0.957, + "step": 6261 + }, + { + "epoch": 0.4302597224130823, + "grad_norm": 1.0854167938232422, + "learning_rate": 3.51497877966774e-05, + "loss": 0.9562, + "step": 6262 + }, + { + "epoch": 0.4303284320461729, + "grad_norm": 1.0349466800689697, + "learning_rate": 3.5144307605571814e-05, + "loss": 0.9512, + "step": 6263 + }, + { + "epoch": 0.4303971416792634, + "grad_norm": 1.0604593753814697, + "learning_rate": 3.51388268308704e-05, + "loss": 0.9605, + "step": 6264 + }, + { + "epoch": 0.430465851312354, + "grad_norm": 1.125087857246399, + "learning_rate": 3.5133345472888467e-05, + "loss": 0.9986, + "step": 6265 + }, + { + "epoch": 0.43053456094544457, + "grad_norm": 1.1516560316085815, + "learning_rate": 3.512786353194134e-05, + "loss": 0.992, + "step": 6266 + }, + { + "epoch": 0.4306032705785351, + "grad_norm": 1.1518136262893677, + "learning_rate": 3.51223810083444e-05, + "loss": 0.9588, + "step": 6267 + }, + { + "epoch": 0.43067198021162567, + "grad_norm": 1.0460563898086548, + "learning_rate": 3.511689790241306e-05, + "loss": 0.993, + "step": 6268 + }, + { + "epoch": 0.43074068984471625, + "grad_norm": 1.0949833393096924, + "learning_rate": 3.511141421446274e-05, + "loss": 0.9649, + "step": 6269 + }, + { + "epoch": 0.4308093994778068, + "grad_norm": 1.0038316249847412, + "learning_rate": 3.510592994480893e-05, + "loss": 0.9464, + "step": 6270 + }, + { + "epoch": 0.43087810911089736, + "grad_norm": 1.1106761693954468, + "learning_rate": 3.510044509376714e-05, + "loss": 0.9927, + "step": 6271 + }, + { + "epoch": 0.4309468187439879, + "grad_norm": 1.055987000465393, + "learning_rate": 3.509495966165289e-05, + "loss": 1.0057, + "step": 6272 + }, + { + "epoch": 0.43101552837707846, + "grad_norm": 1.0154019594192505, + "learning_rate": 3.508947364878178e-05, + "loss": 0.9527, + "step": 6273 + }, + { + "epoch": 0.43108423801016904, + "grad_norm": 1.013479232788086, + "learning_rate": 3.508398705546939e-05, + "loss": 0.9509, + "step": 6274 + }, + { + "epoch": 0.43115294764325957, + "grad_norm": 1.1051311492919922, + "learning_rate": 3.5078499882031377e-05, + "loss": 0.9946, + "step": 6275 + }, + { + "epoch": 0.43122165727635015, + "grad_norm": 1.017910122871399, + "learning_rate": 3.507301212878341e-05, + "loss": 0.9629, + "step": 6276 + }, + { + "epoch": 0.4312903669094407, + "grad_norm": 1.1048197746276855, + "learning_rate": 3.5067523796041204e-05, + "loss": 0.9636, + "step": 6277 + }, + { + "epoch": 0.43135907654253125, + "grad_norm": 1.1165714263916016, + "learning_rate": 3.5062034884120485e-05, + "loss": 1.0008, + "step": 6278 + }, + { + "epoch": 0.43142778617562183, + "grad_norm": 1.113180160522461, + "learning_rate": 3.505654539333704e-05, + "loss": 0.9897, + "step": 6279 + }, + { + "epoch": 0.43149649580871235, + "grad_norm": 1.0874340534210205, + "learning_rate": 3.505105532400667e-05, + "loss": 0.9711, + "step": 6280 + }, + { + "epoch": 0.43156520544180293, + "grad_norm": 1.073140025138855, + "learning_rate": 3.504556467644521e-05, + "loss": 0.935, + "step": 6281 + }, + { + "epoch": 0.4316339150748935, + "grad_norm": 1.0665889978408813, + "learning_rate": 3.5040073450968543e-05, + "loss": 0.9819, + "step": 6282 + }, + { + "epoch": 0.43170262470798404, + "grad_norm": 1.0368585586547852, + "learning_rate": 3.5034581647892574e-05, + "loss": 0.9738, + "step": 6283 + }, + { + "epoch": 0.4317713343410746, + "grad_norm": 1.1205185651779175, + "learning_rate": 3.502908926753323e-05, + "loss": 1.0213, + "step": 6284 + }, + { + "epoch": 0.4318400439741652, + "grad_norm": 1.0700206756591797, + "learning_rate": 3.502359631020651e-05, + "loss": 0.943, + "step": 6285 + }, + { + "epoch": 0.4319087536072557, + "grad_norm": 1.0750290155410767, + "learning_rate": 3.5018102776228394e-05, + "loss": 0.981, + "step": 6286 + }, + { + "epoch": 0.4319774632403463, + "grad_norm": 1.0964843034744263, + "learning_rate": 3.501260866591494e-05, + "loss": 0.968, + "step": 6287 + }, + { + "epoch": 0.43204617287343683, + "grad_norm": 1.1505918502807617, + "learning_rate": 3.500711397958222e-05, + "loss": 0.9906, + "step": 6288 + }, + { + "epoch": 0.4321148825065274, + "grad_norm": 1.0608009099960327, + "learning_rate": 3.5001618717546326e-05, + "loss": 0.9714, + "step": 6289 + }, + { + "epoch": 0.432183592139618, + "grad_norm": 1.1005897521972656, + "learning_rate": 3.49961228801234e-05, + "loss": 0.977, + "step": 6290 + }, + { + "epoch": 0.4322523017727085, + "grad_norm": 1.1263436079025269, + "learning_rate": 3.499062646762964e-05, + "loss": 1.0331, + "step": 6291 + }, + { + "epoch": 0.4323210114057991, + "grad_norm": 1.1469361782073975, + "learning_rate": 3.4985129480381215e-05, + "loss": 0.9833, + "step": 6292 + }, + { + "epoch": 0.4323897210388897, + "grad_norm": 1.1020153760910034, + "learning_rate": 3.497963191869439e-05, + "loss": 1.0122, + "step": 6293 + }, + { + "epoch": 0.4324584306719802, + "grad_norm": 1.0617284774780273, + "learning_rate": 3.497413378288541e-05, + "loss": 0.9675, + "step": 6294 + }, + { + "epoch": 0.4325271403050708, + "grad_norm": 1.166558027267456, + "learning_rate": 3.4968635073270605e-05, + "loss": 0.9746, + "step": 6295 + }, + { + "epoch": 0.4325958499381613, + "grad_norm": 1.0238300561904907, + "learning_rate": 3.49631357901663e-05, + "loss": 0.9686, + "step": 6296 + }, + { + "epoch": 0.4326645595712519, + "grad_norm": 1.0422804355621338, + "learning_rate": 3.4957635933888866e-05, + "loss": 0.9983, + "step": 6297 + }, + { + "epoch": 0.43273326920434246, + "grad_norm": 1.0907074213027954, + "learning_rate": 3.4952135504754715e-05, + "loss": 0.9759, + "step": 6298 + }, + { + "epoch": 0.432801978837433, + "grad_norm": 1.1163179874420166, + "learning_rate": 3.494663450308027e-05, + "loss": 0.9774, + "step": 6299 + }, + { + "epoch": 0.43287068847052357, + "grad_norm": 1.1916849613189697, + "learning_rate": 3.494113292918201e-05, + "loss": 1.0157, + "step": 6300 + }, + { + "epoch": 0.43293939810361415, + "grad_norm": 1.1073100566864014, + "learning_rate": 3.4935630783376436e-05, + "loss": 1.0117, + "step": 6301 + }, + { + "epoch": 0.4330081077367047, + "grad_norm": 1.2050859928131104, + "learning_rate": 3.4930128065980075e-05, + "loss": 0.9744, + "step": 6302 + }, + { + "epoch": 0.43307681736979525, + "grad_norm": 1.1062475442886353, + "learning_rate": 3.4924624777309505e-05, + "loss": 1.0062, + "step": 6303 + }, + { + "epoch": 0.4331455270028858, + "grad_norm": 1.0876041650772095, + "learning_rate": 3.4919120917681327e-05, + "loss": 1.0008, + "step": 6304 + }, + { + "epoch": 0.43321423663597636, + "grad_norm": 1.0674160718917847, + "learning_rate": 3.491361648741217e-05, + "loss": 1.0042, + "step": 6305 + }, + { + "epoch": 0.43328294626906694, + "grad_norm": 1.100002646446228, + "learning_rate": 3.49081114868187e-05, + "loss": 0.9682, + "step": 6306 + }, + { + "epoch": 0.43335165590215746, + "grad_norm": 1.0788488388061523, + "learning_rate": 3.490260591621761e-05, + "loss": 0.9894, + "step": 6307 + }, + { + "epoch": 0.43342036553524804, + "grad_norm": 1.0892252922058105, + "learning_rate": 3.489709977592565e-05, + "loss": 1.0216, + "step": 6308 + }, + { + "epoch": 0.4334890751683386, + "grad_norm": 1.0924304723739624, + "learning_rate": 3.4891593066259556e-05, + "loss": 0.9979, + "step": 6309 + }, + { + "epoch": 0.43355778480142915, + "grad_norm": 1.0553431510925293, + "learning_rate": 3.488608578753616e-05, + "loss": 0.9682, + "step": 6310 + }, + { + "epoch": 0.4336264944345197, + "grad_norm": 1.0872797966003418, + "learning_rate": 3.4880577940072276e-05, + "loss": 1.0099, + "step": 6311 + }, + { + "epoch": 0.43369520406761025, + "grad_norm": 1.1108355522155762, + "learning_rate": 3.487506952418476e-05, + "loss": 0.9744, + "step": 6312 + }, + { + "epoch": 0.43376391370070083, + "grad_norm": 1.117989420890808, + "learning_rate": 3.486956054019052e-05, + "loss": 0.9605, + "step": 6313 + }, + { + "epoch": 0.4338326233337914, + "grad_norm": 1.1054103374481201, + "learning_rate": 3.486405098840648e-05, + "loss": 0.9872, + "step": 6314 + }, + { + "epoch": 0.43390133296688194, + "grad_norm": 1.1396907567977905, + "learning_rate": 3.485854086914961e-05, + "loss": 0.9829, + "step": 6315 + }, + { + "epoch": 0.4339700425999725, + "grad_norm": 0.974848210811615, + "learning_rate": 3.485303018273689e-05, + "loss": 0.9514, + "step": 6316 + }, + { + "epoch": 0.4340387522330631, + "grad_norm": 1.06144380569458, + "learning_rate": 3.484751892948535e-05, + "loss": 0.9724, + "step": 6317 + }, + { + "epoch": 0.4341074618661536, + "grad_norm": 1.1086996793746948, + "learning_rate": 3.484200710971204e-05, + "loss": 1.0292, + "step": 6318 + }, + { + "epoch": 0.4341761714992442, + "grad_norm": 1.0975617170333862, + "learning_rate": 3.483649472373407e-05, + "loss": 0.9816, + "step": 6319 + }, + { + "epoch": 0.4342448811323347, + "grad_norm": 1.0613621473312378, + "learning_rate": 3.483098177186855e-05, + "loss": 0.9822, + "step": 6320 + }, + { + "epoch": 0.4343135907654253, + "grad_norm": 1.148586392402649, + "learning_rate": 3.482546825443266e-05, + "loss": 0.9855, + "step": 6321 + }, + { + "epoch": 0.4343823003985159, + "grad_norm": 1.0573641061782837, + "learning_rate": 3.4819954171743556e-05, + "loss": 0.9333, + "step": 6322 + }, + { + "epoch": 0.4344510100316064, + "grad_norm": 1.0303696393966675, + "learning_rate": 3.481443952411848e-05, + "loss": 0.978, + "step": 6323 + }, + { + "epoch": 0.434519719664697, + "grad_norm": 1.0938752889633179, + "learning_rate": 3.480892431187468e-05, + "loss": 0.9976, + "step": 6324 + }, + { + "epoch": 0.43458842929778757, + "grad_norm": 1.0656867027282715, + "learning_rate": 3.480340853532945e-05, + "loss": 0.9726, + "step": 6325 + }, + { + "epoch": 0.4346571389308781, + "grad_norm": 1.047560214996338, + "learning_rate": 3.479789219480011e-05, + "loss": 0.9594, + "step": 6326 + }, + { + "epoch": 0.4347258485639687, + "grad_norm": 1.0783953666687012, + "learning_rate": 3.4792375290604e-05, + "loss": 0.9861, + "step": 6327 + }, + { + "epoch": 0.43479455819705926, + "grad_norm": 1.126306414604187, + "learning_rate": 3.478685782305851e-05, + "loss": 0.989, + "step": 6328 + }, + { + "epoch": 0.4348632678301498, + "grad_norm": 1.0888265371322632, + "learning_rate": 3.478133979248106e-05, + "loss": 0.9867, + "step": 6329 + }, + { + "epoch": 0.43493197746324036, + "grad_norm": 1.0833510160446167, + "learning_rate": 3.477582119918909e-05, + "loss": 0.9606, + "step": 6330 + }, + { + "epoch": 0.4350006870963309, + "grad_norm": 1.1357604265213013, + "learning_rate": 3.477030204350009e-05, + "loss": 0.9945, + "step": 6331 + }, + { + "epoch": 0.43506939672942146, + "grad_norm": 1.1036194562911987, + "learning_rate": 3.476478232573157e-05, + "loss": 0.995, + "step": 6332 + }, + { + "epoch": 0.43513810636251204, + "grad_norm": 1.1042715311050415, + "learning_rate": 3.4759262046201074e-05, + "loss": 0.9693, + "step": 6333 + }, + { + "epoch": 0.43520681599560257, + "grad_norm": 1.0930207967758179, + "learning_rate": 3.475374120522619e-05, + "loss": 0.9616, + "step": 6334 + }, + { + "epoch": 0.43527552562869315, + "grad_norm": 1.152984857559204, + "learning_rate": 3.4748219803124513e-05, + "loss": 0.9852, + "step": 6335 + }, + { + "epoch": 0.43534423526178373, + "grad_norm": 1.1261905431747437, + "learning_rate": 3.47426978402137e-05, + "loss": 1.038, + "step": 6336 + }, + { + "epoch": 0.43541294489487425, + "grad_norm": 1.169884443283081, + "learning_rate": 3.473717531681142e-05, + "loss": 1.0022, + "step": 6337 + }, + { + "epoch": 0.43548165452796483, + "grad_norm": 1.1240395307540894, + "learning_rate": 3.473165223323539e-05, + "loss": 0.9668, + "step": 6338 + }, + { + "epoch": 0.43555036416105536, + "grad_norm": 1.0652244091033936, + "learning_rate": 3.472612858980333e-05, + "loss": 0.9271, + "step": 6339 + }, + { + "epoch": 0.43561907379414594, + "grad_norm": 1.0677202939987183, + "learning_rate": 3.472060438683302e-05, + "loss": 0.979, + "step": 6340 + }, + { + "epoch": 0.4356877834272365, + "grad_norm": 1.112514615058899, + "learning_rate": 3.471507962464226e-05, + "loss": 0.9601, + "step": 6341 + }, + { + "epoch": 0.43575649306032704, + "grad_norm": 1.101981282234192, + "learning_rate": 3.470955430354891e-05, + "loss": 0.968, + "step": 6342 + }, + { + "epoch": 0.4358252026934176, + "grad_norm": 1.1252692937850952, + "learning_rate": 3.470402842387081e-05, + "loss": 1.0071, + "step": 6343 + }, + { + "epoch": 0.4358939123265082, + "grad_norm": 1.147780179977417, + "learning_rate": 3.469850198592588e-05, + "loss": 0.9984, + "step": 6344 + }, + { + "epoch": 0.43596262195959873, + "grad_norm": 1.0479329824447632, + "learning_rate": 3.469297499003205e-05, + "loss": 1.0039, + "step": 6345 + }, + { + "epoch": 0.4360313315926893, + "grad_norm": 1.0661522150039673, + "learning_rate": 3.468744743650726e-05, + "loss": 0.9798, + "step": 6346 + }, + { + "epoch": 0.43610004122577983, + "grad_norm": 1.0943946838378906, + "learning_rate": 3.468191932566954e-05, + "loss": 0.9596, + "step": 6347 + }, + { + "epoch": 0.4361687508588704, + "grad_norm": 1.1967391967773438, + "learning_rate": 3.46763906578369e-05, + "loss": 0.9944, + "step": 6348 + }, + { + "epoch": 0.436237460491961, + "grad_norm": 1.1453502178192139, + "learning_rate": 3.4670861433327406e-05, + "loss": 0.9872, + "step": 6349 + }, + { + "epoch": 0.4363061701250515, + "grad_norm": 1.060978651046753, + "learning_rate": 3.466533165245915e-05, + "loss": 0.9773, + "step": 6350 + }, + { + "epoch": 0.4363748797581421, + "grad_norm": 1.1048130989074707, + "learning_rate": 3.465980131555026e-05, + "loss": 0.9802, + "step": 6351 + }, + { + "epoch": 0.4364435893912327, + "grad_norm": 1.1008096933364868, + "learning_rate": 3.465427042291889e-05, + "loss": 1.0044, + "step": 6352 + }, + { + "epoch": 0.4365122990243232, + "grad_norm": 1.13772451877594, + "learning_rate": 3.464873897488322e-05, + "loss": 0.957, + "step": 6353 + }, + { + "epoch": 0.4365810086574138, + "grad_norm": 1.0741448402404785, + "learning_rate": 3.4643206971761487e-05, + "loss": 0.9569, + "step": 6354 + }, + { + "epoch": 0.4366497182905043, + "grad_norm": 1.054861307144165, + "learning_rate": 3.463767441387194e-05, + "loss": 0.9925, + "step": 6355 + }, + { + "epoch": 0.4367184279235949, + "grad_norm": 1.0707284212112427, + "learning_rate": 3.463214130153285e-05, + "loss": 0.9913, + "step": 6356 + }, + { + "epoch": 0.43678713755668547, + "grad_norm": 1.0438766479492188, + "learning_rate": 3.462660763506255e-05, + "loss": 0.9565, + "step": 6357 + }, + { + "epoch": 0.436855847189776, + "grad_norm": 1.1531680822372437, + "learning_rate": 3.462107341477938e-05, + "loss": 0.9869, + "step": 6358 + }, + { + "epoch": 0.43692455682286657, + "grad_norm": 1.057746410369873, + "learning_rate": 3.4615538641001735e-05, + "loss": 1.0326, + "step": 6359 + }, + { + "epoch": 0.43699326645595715, + "grad_norm": 1.067171573638916, + "learning_rate": 3.4610003314048e-05, + "loss": 0.9788, + "step": 6360 + }, + { + "epoch": 0.4370619760890477, + "grad_norm": 0.9882630109786987, + "learning_rate": 3.4604467434236645e-05, + "loss": 0.9205, + "step": 6361 + }, + { + "epoch": 0.43713068572213826, + "grad_norm": 1.1034008264541626, + "learning_rate": 3.459893100188613e-05, + "loss": 0.9921, + "step": 6362 + }, + { + "epoch": 0.4371993953552288, + "grad_norm": 1.1099258661270142, + "learning_rate": 3.459339401731496e-05, + "loss": 1.0082, + "step": 6363 + }, + { + "epoch": 0.43726810498831936, + "grad_norm": 1.1104501485824585, + "learning_rate": 3.45878564808417e-05, + "loss": 0.9831, + "step": 6364 + }, + { + "epoch": 0.43733681462140994, + "grad_norm": 1.1302664279937744, + "learning_rate": 3.4582318392784883e-05, + "loss": 0.988, + "step": 6365 + }, + { + "epoch": 0.43740552425450047, + "grad_norm": 1.089514136314392, + "learning_rate": 3.4576779753463135e-05, + "loss": 0.9546, + "step": 6366 + }, + { + "epoch": 0.43747423388759105, + "grad_norm": 1.1141722202301025, + "learning_rate": 3.4571240563195095e-05, + "loss": 1.0033, + "step": 6367 + }, + { + "epoch": 0.4375429435206816, + "grad_norm": 1.0897518396377563, + "learning_rate": 3.4565700822299416e-05, + "loss": 0.9433, + "step": 6368 + }, + { + "epoch": 0.43761165315377215, + "grad_norm": 1.07350754737854, + "learning_rate": 3.45601605310948e-05, + "loss": 0.9765, + "step": 6369 + }, + { + "epoch": 0.43768036278686273, + "grad_norm": 1.1064928770065308, + "learning_rate": 3.455461968989998e-05, + "loss": 0.9356, + "step": 6370 + }, + { + "epoch": 0.43774907241995326, + "grad_norm": 1.0881178379058838, + "learning_rate": 3.454907829903372e-05, + "loss": 0.9789, + "step": 6371 + }, + { + "epoch": 0.43781778205304384, + "grad_norm": 1.0975112915039062, + "learning_rate": 3.4543536358814795e-05, + "loss": 0.9437, + "step": 6372 + }, + { + "epoch": 0.4378864916861344, + "grad_norm": 1.1000205278396606, + "learning_rate": 3.453799386956206e-05, + "loss": 0.9615, + "step": 6373 + }, + { + "epoch": 0.43795520131922494, + "grad_norm": 1.0912915468215942, + "learning_rate": 3.453245083159434e-05, + "loss": 0.9668, + "step": 6374 + }, + { + "epoch": 0.4380239109523155, + "grad_norm": 1.0513150691986084, + "learning_rate": 3.452690724523054e-05, + "loss": 0.9607, + "step": 6375 + }, + { + "epoch": 0.4380926205854061, + "grad_norm": 1.0963389873504639, + "learning_rate": 3.4521363110789575e-05, + "loss": 0.9825, + "step": 6376 + }, + { + "epoch": 0.4381613302184966, + "grad_norm": 1.0487178564071655, + "learning_rate": 3.45158184285904e-05, + "loss": 0.9742, + "step": 6377 + }, + { + "epoch": 0.4382300398515872, + "grad_norm": 1.0990850925445557, + "learning_rate": 3.4510273198951986e-05, + "loss": 0.9743, + "step": 6378 + }, + { + "epoch": 0.43829874948467773, + "grad_norm": 1.0091177225112915, + "learning_rate": 3.450472742219335e-05, + "loss": 0.9705, + "step": 6379 + }, + { + "epoch": 0.4383674591177683, + "grad_norm": 1.02817964553833, + "learning_rate": 3.4499181098633556e-05, + "loss": 0.9719, + "step": 6380 + }, + { + "epoch": 0.4384361687508589, + "grad_norm": 1.032530665397644, + "learning_rate": 3.4493634228591645e-05, + "loss": 0.9573, + "step": 6381 + }, + { + "epoch": 0.4385048783839494, + "grad_norm": 1.0531635284423828, + "learning_rate": 3.4488086812386766e-05, + "loss": 0.9655, + "step": 6382 + }, + { + "epoch": 0.43857358801704, + "grad_norm": 1.066216230392456, + "learning_rate": 3.448253885033803e-05, + "loss": 0.9689, + "step": 6383 + }, + { + "epoch": 0.4386422976501306, + "grad_norm": 1.0880088806152344, + "learning_rate": 3.44769903427646e-05, + "loss": 0.9831, + "step": 6384 + }, + { + "epoch": 0.4387110072832211, + "grad_norm": 1.1352722644805908, + "learning_rate": 3.4471441289985716e-05, + "loss": 0.985, + "step": 6385 + }, + { + "epoch": 0.4387797169163117, + "grad_norm": 1.125238060951233, + "learning_rate": 3.4465891692320584e-05, + "loss": 0.9653, + "step": 6386 + }, + { + "epoch": 0.4388484265494022, + "grad_norm": 1.1115961074829102, + "learning_rate": 3.4460341550088476e-05, + "loss": 1.0014, + "step": 6387 + }, + { + "epoch": 0.4389171361824928, + "grad_norm": 1.175912618637085, + "learning_rate": 3.4454790863608684e-05, + "loss": 1.0013, + "step": 6388 + }, + { + "epoch": 0.43898584581558336, + "grad_norm": 1.1310540437698364, + "learning_rate": 3.444923963320054e-05, + "loss": 1.0002, + "step": 6389 + }, + { + "epoch": 0.4390545554486739, + "grad_norm": 0.9966714382171631, + "learning_rate": 3.444368785918339e-05, + "loss": 0.9467, + "step": 6390 + }, + { + "epoch": 0.43912326508176447, + "grad_norm": 1.1152222156524658, + "learning_rate": 3.443813554187665e-05, + "loss": 0.9505, + "step": 6391 + }, + { + "epoch": 0.43919197471485505, + "grad_norm": 1.0495952367782593, + "learning_rate": 3.4432582681599725e-05, + "loss": 1.0054, + "step": 6392 + }, + { + "epoch": 0.4392606843479456, + "grad_norm": 1.0901929140090942, + "learning_rate": 3.4427029278672064e-05, + "loss": 0.9787, + "step": 6393 + }, + { + "epoch": 0.43932939398103615, + "grad_norm": 1.0837503671646118, + "learning_rate": 3.4421475333413155e-05, + "loss": 0.9933, + "step": 6394 + }, + { + "epoch": 0.4393981036141267, + "grad_norm": 1.0929458141326904, + "learning_rate": 3.4415920846142524e-05, + "loss": 1.0174, + "step": 6395 + }, + { + "epoch": 0.43946681324721726, + "grad_norm": 1.1211087703704834, + "learning_rate": 3.4410365817179694e-05, + "loss": 0.9706, + "step": 6396 + }, + { + "epoch": 0.43953552288030784, + "grad_norm": 1.1394002437591553, + "learning_rate": 3.440481024684427e-05, + "loss": 0.9749, + "step": 6397 + }, + { + "epoch": 0.43960423251339836, + "grad_norm": 1.0180697441101074, + "learning_rate": 3.439925413545585e-05, + "loss": 0.9875, + "step": 6398 + }, + { + "epoch": 0.43967294214648894, + "grad_norm": 1.0595653057098389, + "learning_rate": 3.4393697483334056e-05, + "loss": 0.9472, + "step": 6399 + }, + { + "epoch": 0.4397416517795795, + "grad_norm": 1.055827021598816, + "learning_rate": 3.438814029079858e-05, + "loss": 0.9554, + "step": 6400 + }, + { + "epoch": 0.43981036141267005, + "grad_norm": 1.1310606002807617, + "learning_rate": 3.438258255816912e-05, + "loss": 1.0051, + "step": 6401 + }, + { + "epoch": 0.43987907104576063, + "grad_norm": 1.1417979001998901, + "learning_rate": 3.43770242857654e-05, + "loss": 0.9893, + "step": 6402 + }, + { + "epoch": 0.43994778067885115, + "grad_norm": 1.1386256217956543, + "learning_rate": 3.4371465473907194e-05, + "loss": 0.9773, + "step": 6403 + }, + { + "epoch": 0.44001649031194173, + "grad_norm": 1.0725144147872925, + "learning_rate": 3.436590612291429e-05, + "loss": 0.9844, + "step": 6404 + }, + { + "epoch": 0.4400851999450323, + "grad_norm": 1.0440349578857422, + "learning_rate": 3.4360346233106526e-05, + "loss": 0.9665, + "step": 6405 + }, + { + "epoch": 0.44015390957812284, + "grad_norm": 1.071652889251709, + "learning_rate": 3.4354785804803744e-05, + "loss": 1.0079, + "step": 6406 + }, + { + "epoch": 0.4402226192112134, + "grad_norm": 1.0756633281707764, + "learning_rate": 3.434922483832584e-05, + "loss": 0.9561, + "step": 6407 + }, + { + "epoch": 0.440291328844304, + "grad_norm": 1.0684877634048462, + "learning_rate": 3.4343663333992735e-05, + "loss": 0.9485, + "step": 6408 + }, + { + "epoch": 0.4403600384773945, + "grad_norm": 1.0306178331375122, + "learning_rate": 3.433810129212437e-05, + "loss": 0.9538, + "step": 6409 + }, + { + "epoch": 0.4404287481104851, + "grad_norm": 1.0862905979156494, + "learning_rate": 3.433253871304073e-05, + "loss": 1.0043, + "step": 6410 + }, + { + "epoch": 0.4404974577435756, + "grad_norm": 1.0811127424240112, + "learning_rate": 3.432697559706184e-05, + "loss": 1.0248, + "step": 6411 + }, + { + "epoch": 0.4405661673766662, + "grad_norm": 1.1458100080490112, + "learning_rate": 3.432141194450772e-05, + "loss": 0.9723, + "step": 6412 + }, + { + "epoch": 0.4406348770097568, + "grad_norm": 1.0508977174758911, + "learning_rate": 3.431584775569846e-05, + "loss": 0.9517, + "step": 6413 + }, + { + "epoch": 0.4407035866428473, + "grad_norm": 1.132312536239624, + "learning_rate": 3.431028303095415e-05, + "loss": 0.9914, + "step": 6414 + }, + { + "epoch": 0.4407722962759379, + "grad_norm": 1.080151915550232, + "learning_rate": 3.430471777059494e-05, + "loss": 1.0343, + "step": 6415 + }, + { + "epoch": 0.44084100590902847, + "grad_norm": 1.024770975112915, + "learning_rate": 3.429915197494099e-05, + "loss": 0.9593, + "step": 6416 + }, + { + "epoch": 0.440909715542119, + "grad_norm": 1.0020415782928467, + "learning_rate": 3.4293585644312496e-05, + "loss": 0.9465, + "step": 6417 + }, + { + "epoch": 0.4409784251752096, + "grad_norm": 1.0740349292755127, + "learning_rate": 3.428801877902969e-05, + "loss": 0.9782, + "step": 6418 + }, + { + "epoch": 0.4410471348083001, + "grad_norm": 1.0910271406173706, + "learning_rate": 3.4282451379412825e-05, + "loss": 0.9804, + "step": 6419 + }, + { + "epoch": 0.4411158444413907, + "grad_norm": 1.0592575073242188, + "learning_rate": 3.427688344578219e-05, + "loss": 0.9763, + "step": 6420 + }, + { + "epoch": 0.44118455407448126, + "grad_norm": 1.030966877937317, + "learning_rate": 3.42713149784581e-05, + "loss": 0.9492, + "step": 6421 + }, + { + "epoch": 0.4412532637075718, + "grad_norm": 1.0479410886764526, + "learning_rate": 3.4265745977760925e-05, + "loss": 0.9168, + "step": 6422 + }, + { + "epoch": 0.44132197334066237, + "grad_norm": 1.0780950784683228, + "learning_rate": 3.4260176444011034e-05, + "loss": 1.0104, + "step": 6423 + }, + { + "epoch": 0.44139068297375295, + "grad_norm": 1.0641522407531738, + "learning_rate": 3.425460637752883e-05, + "loss": 0.9934, + "step": 6424 + }, + { + "epoch": 0.44145939260684347, + "grad_norm": 1.0434374809265137, + "learning_rate": 3.4249035778634765e-05, + "loss": 0.9722, + "step": 6425 + }, + { + "epoch": 0.44152810223993405, + "grad_norm": 1.1079216003417969, + "learning_rate": 3.424346464764931e-05, + "loss": 1.0008, + "step": 6426 + }, + { + "epoch": 0.4415968118730246, + "grad_norm": 1.0516263246536255, + "learning_rate": 3.423789298489298e-05, + "loss": 0.9666, + "step": 6427 + }, + { + "epoch": 0.44166552150611516, + "grad_norm": 1.0021138191223145, + "learning_rate": 3.423232079068629e-05, + "loss": 0.9241, + "step": 6428 + }, + { + "epoch": 0.44173423113920574, + "grad_norm": 1.1129564046859741, + "learning_rate": 3.422674806534981e-05, + "loss": 0.975, + "step": 6429 + }, + { + "epoch": 0.44180294077229626, + "grad_norm": 1.0567470788955688, + "learning_rate": 3.422117480920415e-05, + "loss": 0.9787, + "step": 6430 + }, + { + "epoch": 0.44187165040538684, + "grad_norm": 1.1427494287490845, + "learning_rate": 3.421560102256992e-05, + "loss": 0.9849, + "step": 6431 + }, + { + "epoch": 0.4419403600384774, + "grad_norm": 1.081299901008606, + "learning_rate": 3.4210026705767785e-05, + "loss": 0.982, + "step": 6432 + }, + { + "epoch": 0.44200906967156794, + "grad_norm": 1.0893539190292358, + "learning_rate": 3.420445185911843e-05, + "loss": 1.0065, + "step": 6433 + }, + { + "epoch": 0.4420777793046585, + "grad_norm": 1.0919442176818848, + "learning_rate": 3.4198876482942574e-05, + "loss": 0.9906, + "step": 6434 + }, + { + "epoch": 0.44214648893774905, + "grad_norm": 1.1240202188491821, + "learning_rate": 3.419330057756095e-05, + "loss": 0.9753, + "step": 6435 + }, + { + "epoch": 0.44221519857083963, + "grad_norm": 1.1218538284301758, + "learning_rate": 3.418772414329436e-05, + "loss": 0.9653, + "step": 6436 + }, + { + "epoch": 0.4422839082039302, + "grad_norm": 1.0402075052261353, + "learning_rate": 3.41821471804636e-05, + "loss": 0.9627, + "step": 6437 + }, + { + "epoch": 0.44235261783702073, + "grad_norm": 1.0506569147109985, + "learning_rate": 3.417656968938952e-05, + "loss": 0.9854, + "step": 6438 + }, + { + "epoch": 0.4424213274701113, + "grad_norm": 1.1450051069259644, + "learning_rate": 3.417099167039297e-05, + "loss": 0.9497, + "step": 6439 + }, + { + "epoch": 0.4424900371032019, + "grad_norm": 1.1374880075454712, + "learning_rate": 3.416541312379486e-05, + "loss": 0.9959, + "step": 6440 + }, + { + "epoch": 0.4425587467362924, + "grad_norm": 1.110426425933838, + "learning_rate": 3.415983404991613e-05, + "loss": 0.9728, + "step": 6441 + }, + { + "epoch": 0.442627456369383, + "grad_norm": 1.0690906047821045, + "learning_rate": 3.415425444907772e-05, + "loss": 0.9423, + "step": 6442 + }, + { + "epoch": 0.4426961660024735, + "grad_norm": 1.0343137979507446, + "learning_rate": 3.4148674321600645e-05, + "loss": 0.9544, + "step": 6443 + }, + { + "epoch": 0.4427648756355641, + "grad_norm": 1.0930882692337036, + "learning_rate": 3.414309366780591e-05, + "loss": 0.9822, + "step": 6444 + }, + { + "epoch": 0.4428335852686547, + "grad_norm": 1.0651592016220093, + "learning_rate": 3.413751248801457e-05, + "loss": 0.9955, + "step": 6445 + }, + { + "epoch": 0.4429022949017452, + "grad_norm": 1.0326844453811646, + "learning_rate": 3.413193078254772e-05, + "loss": 0.9867, + "step": 6446 + }, + { + "epoch": 0.4429710045348358, + "grad_norm": 1.1302982568740845, + "learning_rate": 3.412634855172644e-05, + "loss": 0.9241, + "step": 6447 + }, + { + "epoch": 0.44303971416792637, + "grad_norm": 1.0780829191207886, + "learning_rate": 3.4120765795871904e-05, + "loss": 0.9748, + "step": 6448 + }, + { + "epoch": 0.4431084238010169, + "grad_norm": 1.077981948852539, + "learning_rate": 3.4115182515305266e-05, + "loss": 0.9724, + "step": 6449 + }, + { + "epoch": 0.4431771334341075, + "grad_norm": 1.1514074802398682, + "learning_rate": 3.410959871034774e-05, + "loss": 0.9712, + "step": 6450 + }, + { + "epoch": 0.443245843067198, + "grad_norm": 1.079561471939087, + "learning_rate": 3.410401438132056e-05, + "loss": 0.9716, + "step": 6451 + }, + { + "epoch": 0.4433145527002886, + "grad_norm": 1.1541938781738281, + "learning_rate": 3.409842952854497e-05, + "loss": 1.0289, + "step": 6452 + }, + { + "epoch": 0.44338326233337916, + "grad_norm": 1.116607904434204, + "learning_rate": 3.4092844152342284e-05, + "loss": 0.9674, + "step": 6453 + }, + { + "epoch": 0.4434519719664697, + "grad_norm": 1.1121764183044434, + "learning_rate": 3.408725825303383e-05, + "loss": 0.9687, + "step": 6454 + }, + { + "epoch": 0.44352068159956026, + "grad_norm": 1.0978680849075317, + "learning_rate": 3.408167183094093e-05, + "loss": 0.9407, + "step": 6455 + }, + { + "epoch": 0.44358939123265084, + "grad_norm": 1.0842430591583252, + "learning_rate": 3.4076084886385e-05, + "loss": 0.976, + "step": 6456 + }, + { + "epoch": 0.44365810086574137, + "grad_norm": 1.0674655437469482, + "learning_rate": 3.407049741968744e-05, + "loss": 0.9605, + "step": 6457 + }, + { + "epoch": 0.44372681049883195, + "grad_norm": 1.1453361511230469, + "learning_rate": 3.406490943116969e-05, + "loss": 0.9651, + "step": 6458 + }, + { + "epoch": 0.44379552013192247, + "grad_norm": 1.0840528011322021, + "learning_rate": 3.405932092115323e-05, + "loss": 0.9949, + "step": 6459 + }, + { + "epoch": 0.44386422976501305, + "grad_norm": 1.0812768936157227, + "learning_rate": 3.4053731889959554e-05, + "loss": 0.9855, + "step": 6460 + }, + { + "epoch": 0.44393293939810363, + "grad_norm": 1.1290957927703857, + "learning_rate": 3.404814233791021e-05, + "loss": 0.9587, + "step": 6461 + }, + { + "epoch": 0.44400164903119416, + "grad_norm": 1.0745679140090942, + "learning_rate": 3.4042552265326756e-05, + "loss": 0.9918, + "step": 6462 + }, + { + "epoch": 0.44407035866428474, + "grad_norm": 1.1324443817138672, + "learning_rate": 3.4036961672530777e-05, + "loss": 0.979, + "step": 6463 + }, + { + "epoch": 0.4441390682973753, + "grad_norm": 1.1264296770095825, + "learning_rate": 3.4031370559843904e-05, + "loss": 0.9901, + "step": 6464 + }, + { + "epoch": 0.44420777793046584, + "grad_norm": 1.1696099042892456, + "learning_rate": 3.4025778927587795e-05, + "loss": 0.9907, + "step": 6465 + }, + { + "epoch": 0.4442764875635564, + "grad_norm": 1.106164813041687, + "learning_rate": 3.4020186776084124e-05, + "loss": 1.0066, + "step": 6466 + }, + { + "epoch": 0.44434519719664695, + "grad_norm": 0.9998903274536133, + "learning_rate": 3.401459410565461e-05, + "loss": 0.9593, + "step": 6467 + }, + { + "epoch": 0.4444139068297375, + "grad_norm": 1.1102120876312256, + "learning_rate": 3.400900091662099e-05, + "loss": 0.9637, + "step": 6468 + }, + { + "epoch": 0.4444826164628281, + "grad_norm": 1.0915948152542114, + "learning_rate": 3.400340720930504e-05, + "loss": 0.994, + "step": 6469 + }, + { + "epoch": 0.44455132609591863, + "grad_norm": 1.0347658395767212, + "learning_rate": 3.399781298402857e-05, + "loss": 0.9855, + "step": 6470 + }, + { + "epoch": 0.4446200357290092, + "grad_norm": 1.0559908151626587, + "learning_rate": 3.39922182411134e-05, + "loss": 0.9681, + "step": 6471 + }, + { + "epoch": 0.4446887453620998, + "grad_norm": 1.1239455938339233, + "learning_rate": 3.3986622980881396e-05, + "loss": 0.9761, + "step": 6472 + }, + { + "epoch": 0.4447574549951903, + "grad_norm": 1.1626782417297363, + "learning_rate": 3.398102720365445e-05, + "loss": 0.9756, + "step": 6473 + }, + { + "epoch": 0.4448261646282809, + "grad_norm": 1.0445022583007812, + "learning_rate": 3.397543090975449e-05, + "loss": 0.9694, + "step": 6474 + }, + { + "epoch": 0.4448948742613714, + "grad_norm": 1.055953025817871, + "learning_rate": 3.3969834099503455e-05, + "loss": 0.9552, + "step": 6475 + }, + { + "epoch": 0.444963583894462, + "grad_norm": 1.0266311168670654, + "learning_rate": 3.3964236773223344e-05, + "loss": 0.9699, + "step": 6476 + }, + { + "epoch": 0.4450322935275526, + "grad_norm": 1.0779510736465454, + "learning_rate": 3.395863893123615e-05, + "loss": 0.9447, + "step": 6477 + }, + { + "epoch": 0.4451010031606431, + "grad_norm": 1.107114553451538, + "learning_rate": 3.395304057386392e-05, + "loss": 0.9968, + "step": 6478 + }, + { + "epoch": 0.4451697127937337, + "grad_norm": 1.0397162437438965, + "learning_rate": 3.394744170142873e-05, + "loss": 0.9727, + "step": 6479 + }, + { + "epoch": 0.44523842242682427, + "grad_norm": 1.0363045930862427, + "learning_rate": 3.394184231425267e-05, + "loss": 0.9461, + "step": 6480 + }, + { + "epoch": 0.4453071320599148, + "grad_norm": 1.1534690856933594, + "learning_rate": 3.3936242412657874e-05, + "loss": 1.0114, + "step": 6481 + }, + { + "epoch": 0.44537584169300537, + "grad_norm": 1.0929317474365234, + "learning_rate": 3.393064199696651e-05, + "loss": 0.9454, + "step": 6482 + }, + { + "epoch": 0.4454445513260959, + "grad_norm": 1.0983211994171143, + "learning_rate": 3.3925041067500743e-05, + "loss": 0.9395, + "step": 6483 + }, + { + "epoch": 0.4455132609591865, + "grad_norm": 1.061780571937561, + "learning_rate": 3.3919439624582816e-05, + "loss": 0.9828, + "step": 6484 + }, + { + "epoch": 0.44558197059227705, + "grad_norm": 0.994208037853241, + "learning_rate": 3.391383766853496e-05, + "loss": 0.9693, + "step": 6485 + }, + { + "epoch": 0.4456506802253676, + "grad_norm": 1.0780326128005981, + "learning_rate": 3.3908235199679456e-05, + "loss": 0.9484, + "step": 6486 + }, + { + "epoch": 0.44571938985845816, + "grad_norm": 1.131791353225708, + "learning_rate": 3.390263221833862e-05, + "loss": 0.981, + "step": 6487 + }, + { + "epoch": 0.44578809949154874, + "grad_norm": 1.1080663204193115, + "learning_rate": 3.389702872483477e-05, + "loss": 1.0016, + "step": 6488 + }, + { + "epoch": 0.44585680912463926, + "grad_norm": 1.081605315208435, + "learning_rate": 3.3891424719490295e-05, + "loss": 0.9819, + "step": 6489 + }, + { + "epoch": 0.44592551875772984, + "grad_norm": 1.0715110301971436, + "learning_rate": 3.388582020262757e-05, + "loss": 0.9659, + "step": 6490 + }, + { + "epoch": 0.44599422839082037, + "grad_norm": 1.051609992980957, + "learning_rate": 3.388021517456903e-05, + "loss": 1.0107, + "step": 6491 + }, + { + "epoch": 0.44606293802391095, + "grad_norm": 1.040592074394226, + "learning_rate": 3.387460963563712e-05, + "loss": 0.9871, + "step": 6492 + }, + { + "epoch": 0.44613164765700153, + "grad_norm": 1.0570937395095825, + "learning_rate": 3.386900358615433e-05, + "loss": 0.9333, + "step": 6493 + }, + { + "epoch": 0.44620035729009205, + "grad_norm": 1.0889278650283813, + "learning_rate": 3.3863397026443186e-05, + "loss": 1.003, + "step": 6494 + }, + { + "epoch": 0.44626906692318263, + "grad_norm": 1.1677439212799072, + "learning_rate": 3.38577899568262e-05, + "loss": 0.994, + "step": 6495 + }, + { + "epoch": 0.4463377765562732, + "grad_norm": 1.064171552658081, + "learning_rate": 3.385218237762596e-05, + "loss": 0.9467, + "step": 6496 + }, + { + "epoch": 0.44640648618936374, + "grad_norm": 1.2001352310180664, + "learning_rate": 3.3846574289165065e-05, + "loss": 0.9946, + "step": 6497 + }, + { + "epoch": 0.4464751958224543, + "grad_norm": 1.0991301536560059, + "learning_rate": 3.384096569176615e-05, + "loss": 0.9592, + "step": 6498 + }, + { + "epoch": 0.44654390545554484, + "grad_norm": 1.1090854406356812, + "learning_rate": 3.383535658575187e-05, + "loss": 0.9744, + "step": 6499 + }, + { + "epoch": 0.4466126150886354, + "grad_norm": 1.072676420211792, + "learning_rate": 3.382974697144491e-05, + "loss": 0.987, + "step": 6500 + }, + { + "epoch": 0.446681324721726, + "grad_norm": 1.0427416563034058, + "learning_rate": 3.3824136849167984e-05, + "loss": 0.938, + "step": 6501 + }, + { + "epoch": 0.4467500343548165, + "grad_norm": 1.0507146120071411, + "learning_rate": 3.381852621924386e-05, + "loss": 0.9216, + "step": 6502 + }, + { + "epoch": 0.4468187439879071, + "grad_norm": 1.0295398235321045, + "learning_rate": 3.381291508199528e-05, + "loss": 0.9777, + "step": 6503 + }, + { + "epoch": 0.4468874536209977, + "grad_norm": 1.0870238542556763, + "learning_rate": 3.380730343774509e-05, + "loss": 0.9899, + "step": 6504 + }, + { + "epoch": 0.4469561632540882, + "grad_norm": 1.088460087776184, + "learning_rate": 3.38016912868161e-05, + "loss": 0.9626, + "step": 6505 + }, + { + "epoch": 0.4470248728871788, + "grad_norm": 1.098455786705017, + "learning_rate": 3.379607862953116e-05, + "loss": 0.9418, + "step": 6506 + }, + { + "epoch": 0.4470935825202693, + "grad_norm": 1.0483225584030151, + "learning_rate": 3.3790465466213194e-05, + "loss": 1.0186, + "step": 6507 + }, + { + "epoch": 0.4471622921533599, + "grad_norm": 1.1580700874328613, + "learning_rate": 3.37848517971851e-05, + "loss": 1.0211, + "step": 6508 + }, + { + "epoch": 0.4472310017864505, + "grad_norm": 1.0900499820709229, + "learning_rate": 3.3779237622769846e-05, + "loss": 0.9701, + "step": 6509 + }, + { + "epoch": 0.447299711419541, + "grad_norm": 1.1116232872009277, + "learning_rate": 3.3773622943290407e-05, + "loss": 0.9497, + "step": 6510 + }, + { + "epoch": 0.4473684210526316, + "grad_norm": 1.1390396356582642, + "learning_rate": 3.376800775906978e-05, + "loss": 0.9982, + "step": 6511 + }, + { + "epoch": 0.44743713068572216, + "grad_norm": 1.078375220298767, + "learning_rate": 3.376239207043102e-05, + "loss": 0.9683, + "step": 6512 + }, + { + "epoch": 0.4475058403188127, + "grad_norm": 1.1716989278793335, + "learning_rate": 3.375677587769718e-05, + "loss": 1.0048, + "step": 6513 + }, + { + "epoch": 0.44757454995190327, + "grad_norm": 1.1379435062408447, + "learning_rate": 3.375115918119137e-05, + "loss": 0.9857, + "step": 6514 + }, + { + "epoch": 0.4476432595849938, + "grad_norm": 1.0341289043426514, + "learning_rate": 3.374554198123671e-05, + "loss": 0.9597, + "step": 6515 + }, + { + "epoch": 0.44771196921808437, + "grad_norm": 1.0647879838943481, + "learning_rate": 3.3739924278156354e-05, + "loss": 0.9314, + "step": 6516 + }, + { + "epoch": 0.44778067885117495, + "grad_norm": 1.0231614112854004, + "learning_rate": 3.373430607227347e-05, + "loss": 0.9429, + "step": 6517 + }, + { + "epoch": 0.4478493884842655, + "grad_norm": 1.0650173425674438, + "learning_rate": 3.372868736391129e-05, + "loss": 0.9641, + "step": 6518 + }, + { + "epoch": 0.44791809811735606, + "grad_norm": 1.156617283821106, + "learning_rate": 3.3723068153393054e-05, + "loss": 0.9583, + "step": 6519 + }, + { + "epoch": 0.44798680775044664, + "grad_norm": 1.0588672161102295, + "learning_rate": 3.371744844104202e-05, + "loss": 0.9919, + "step": 6520 + }, + { + "epoch": 0.44805551738353716, + "grad_norm": 1.100574254989624, + "learning_rate": 3.3711828227181495e-05, + "loss": 0.9799, + "step": 6521 + }, + { + "epoch": 0.44812422701662774, + "grad_norm": 1.0830533504486084, + "learning_rate": 3.37062075121348e-05, + "loss": 1.0002, + "step": 6522 + }, + { + "epoch": 0.44819293664971827, + "grad_norm": 1.0934745073318481, + "learning_rate": 3.37005862962253e-05, + "loss": 0.945, + "step": 6523 + }, + { + "epoch": 0.44826164628280885, + "grad_norm": 1.0882713794708252, + "learning_rate": 3.369496457977637e-05, + "loss": 0.944, + "step": 6524 + }, + { + "epoch": 0.4483303559158994, + "grad_norm": 1.0376667976379395, + "learning_rate": 3.3689342363111435e-05, + "loss": 0.9594, + "step": 6525 + }, + { + "epoch": 0.44839906554898995, + "grad_norm": 1.0858511924743652, + "learning_rate": 3.368371964655393e-05, + "loss": 1.0016, + "step": 6526 + }, + { + "epoch": 0.44846777518208053, + "grad_norm": 0.9950829744338989, + "learning_rate": 3.367809643042733e-05, + "loss": 0.9533, + "step": 6527 + }, + { + "epoch": 0.4485364848151711, + "grad_norm": 1.067159652709961, + "learning_rate": 3.3672472715055134e-05, + "loss": 0.9652, + "step": 6528 + }, + { + "epoch": 0.44860519444826163, + "grad_norm": 1.0453457832336426, + "learning_rate": 3.366684850076087e-05, + "loss": 1.0037, + "step": 6529 + }, + { + "epoch": 0.4486739040813522, + "grad_norm": 1.0851848125457764, + "learning_rate": 3.3661223787868094e-05, + "loss": 0.9842, + "step": 6530 + }, + { + "epoch": 0.44874261371444274, + "grad_norm": 1.0432816743850708, + "learning_rate": 3.3655598576700395e-05, + "loss": 0.9646, + "step": 6531 + }, + { + "epoch": 0.4488113233475333, + "grad_norm": 1.0508923530578613, + "learning_rate": 3.3649972867581385e-05, + "loss": 0.9436, + "step": 6532 + }, + { + "epoch": 0.4488800329806239, + "grad_norm": 1.136314034461975, + "learning_rate": 3.3644346660834714e-05, + "loss": 0.978, + "step": 6533 + }, + { + "epoch": 0.4489487426137144, + "grad_norm": 1.0878750085830688, + "learning_rate": 3.363871995678405e-05, + "loss": 0.985, + "step": 6534 + }, + { + "epoch": 0.449017452246805, + "grad_norm": 1.137404441833496, + "learning_rate": 3.363309275575309e-05, + "loss": 0.9727, + "step": 6535 + }, + { + "epoch": 0.4490861618798956, + "grad_norm": 1.1036536693572998, + "learning_rate": 3.362746505806557e-05, + "loss": 0.9584, + "step": 6536 + }, + { + "epoch": 0.4491548715129861, + "grad_norm": 1.1035064458847046, + "learning_rate": 3.362183686404524e-05, + "loss": 0.9945, + "step": 6537 + }, + { + "epoch": 0.4492235811460767, + "grad_norm": 1.0558137893676758, + "learning_rate": 3.3616208174015906e-05, + "loss": 1.0022, + "step": 6538 + }, + { + "epoch": 0.4492922907791672, + "grad_norm": 1.1074285507202148, + "learning_rate": 3.361057898830136e-05, + "loss": 0.9983, + "step": 6539 + }, + { + "epoch": 0.4493610004122578, + "grad_norm": 1.0797072649002075, + "learning_rate": 3.360494930722546e-05, + "loss": 0.9687, + "step": 6540 + }, + { + "epoch": 0.4494297100453484, + "grad_norm": 1.1084452867507935, + "learning_rate": 3.3599319131112075e-05, + "loss": 0.9963, + "step": 6541 + }, + { + "epoch": 0.4494984196784389, + "grad_norm": 1.0371533632278442, + "learning_rate": 3.3593688460285093e-05, + "loss": 0.9409, + "step": 6542 + }, + { + "epoch": 0.4495671293115295, + "grad_norm": 1.0498509407043457, + "learning_rate": 3.358805729506847e-05, + "loss": 0.9533, + "step": 6543 + }, + { + "epoch": 0.44963583894462006, + "grad_norm": 1.0766799449920654, + "learning_rate": 3.358242563578613e-05, + "loss": 0.9514, + "step": 6544 + }, + { + "epoch": 0.4497045485777106, + "grad_norm": 1.200533151626587, + "learning_rate": 3.357679348276208e-05, + "loss": 0.9682, + "step": 6545 + }, + { + "epoch": 0.44977325821080116, + "grad_norm": 1.1226569414138794, + "learning_rate": 3.357116083632034e-05, + "loss": 0.9358, + "step": 6546 + }, + { + "epoch": 0.4498419678438917, + "grad_norm": 1.1241285800933838, + "learning_rate": 3.3565527696784925e-05, + "loss": 0.9617, + "step": 6547 + }, + { + "epoch": 0.44991067747698227, + "grad_norm": 1.1333893537521362, + "learning_rate": 3.355989406447994e-05, + "loss": 0.9875, + "step": 6548 + }, + { + "epoch": 0.44997938711007285, + "grad_norm": 1.0765358209609985, + "learning_rate": 3.355425993972947e-05, + "loss": 1.0179, + "step": 6549 + }, + { + "epoch": 0.4500480967431634, + "grad_norm": 1.0711925029754639, + "learning_rate": 3.3548625322857634e-05, + "loss": 0.9616, + "step": 6550 + }, + { + "epoch": 0.45011680637625395, + "grad_norm": 1.1725494861602783, + "learning_rate": 3.35429902141886e-05, + "loss": 0.9772, + "step": 6551 + }, + { + "epoch": 0.45018551600934453, + "grad_norm": 1.005982518196106, + "learning_rate": 3.353735461404654e-05, + "loss": 0.9692, + "step": 6552 + }, + { + "epoch": 0.45025422564243506, + "grad_norm": 1.089609146118164, + "learning_rate": 3.353171852275569e-05, + "loss": 0.9717, + "step": 6553 + }, + { + "epoch": 0.45032293527552564, + "grad_norm": 1.0415890216827393, + "learning_rate": 3.3526081940640265e-05, + "loss": 0.9769, + "step": 6554 + }, + { + "epoch": 0.45039164490861616, + "grad_norm": 1.0306977033615112, + "learning_rate": 3.352044486802455e-05, + "loss": 0.9634, + "step": 6555 + }, + { + "epoch": 0.45046035454170674, + "grad_norm": 1.0921220779418945, + "learning_rate": 3.351480730523283e-05, + "loss": 0.9674, + "step": 6556 + }, + { + "epoch": 0.4505290641747973, + "grad_norm": 1.0609883069992065, + "learning_rate": 3.3509169252589436e-05, + "loss": 0.9786, + "step": 6557 + }, + { + "epoch": 0.45059777380788785, + "grad_norm": 1.0549893379211426, + "learning_rate": 3.350353071041872e-05, + "loss": 0.9525, + "step": 6558 + }, + { + "epoch": 0.4506664834409784, + "grad_norm": 1.0758063793182373, + "learning_rate": 3.349789167904508e-05, + "loss": 0.9868, + "step": 6559 + }, + { + "epoch": 0.450735193074069, + "grad_norm": 1.1299067735671997, + "learning_rate": 3.349225215879291e-05, + "loss": 0.9624, + "step": 6560 + }, + { + "epoch": 0.45080390270715953, + "grad_norm": 1.0982056856155396, + "learning_rate": 3.348661214998664e-05, + "loss": 0.9945, + "step": 6561 + }, + { + "epoch": 0.4508726123402501, + "grad_norm": 1.0934799909591675, + "learning_rate": 3.348097165295076e-05, + "loss": 0.9643, + "step": 6562 + }, + { + "epoch": 0.45094132197334064, + "grad_norm": 1.0448367595672607, + "learning_rate": 3.3475330668009745e-05, + "loss": 0.9252, + "step": 6563 + }, + { + "epoch": 0.4510100316064312, + "grad_norm": 1.0688891410827637, + "learning_rate": 3.346968919548813e-05, + "loss": 0.9563, + "step": 6564 + }, + { + "epoch": 0.4510787412395218, + "grad_norm": 1.0837700366973877, + "learning_rate": 3.3464047235710464e-05, + "loss": 0.9746, + "step": 6565 + }, + { + "epoch": 0.4511474508726123, + "grad_norm": 1.0413225889205933, + "learning_rate": 3.3458404789001306e-05, + "loss": 0.9516, + "step": 6566 + }, + { + "epoch": 0.4512161605057029, + "grad_norm": 1.041175365447998, + "learning_rate": 3.3452761855685286e-05, + "loss": 0.9821, + "step": 6567 + }, + { + "epoch": 0.4512848701387935, + "grad_norm": 1.1214364767074585, + "learning_rate": 3.3447118436087035e-05, + "loss": 0.9978, + "step": 6568 + }, + { + "epoch": 0.451353579771884, + "grad_norm": 1.1091502904891968, + "learning_rate": 3.34414745305312e-05, + "loss": 1.0042, + "step": 6569 + }, + { + "epoch": 0.4514222894049746, + "grad_norm": 1.039304256439209, + "learning_rate": 3.3435830139342495e-05, + "loss": 0.9585, + "step": 6570 + }, + { + "epoch": 0.4514909990380651, + "grad_norm": 1.1066467761993408, + "learning_rate": 3.343018526284562e-05, + "loss": 0.9651, + "step": 6571 + }, + { + "epoch": 0.4515597086711557, + "grad_norm": 1.0490354299545288, + "learning_rate": 3.3424539901365325e-05, + "loss": 0.9712, + "step": 6572 + }, + { + "epoch": 0.45162841830424627, + "grad_norm": 1.0512511730194092, + "learning_rate": 3.341889405522639e-05, + "loss": 0.9816, + "step": 6573 + }, + { + "epoch": 0.4516971279373368, + "grad_norm": 1.0088461637496948, + "learning_rate": 3.3413247724753606e-05, + "loss": 0.9402, + "step": 6574 + }, + { + "epoch": 0.4517658375704274, + "grad_norm": 1.05424964427948, + "learning_rate": 3.340760091027182e-05, + "loss": 0.9715, + "step": 6575 + }, + { + "epoch": 0.45183454720351796, + "grad_norm": 1.0885263681411743, + "learning_rate": 3.340195361210588e-05, + "loss": 0.9658, + "step": 6576 + }, + { + "epoch": 0.4519032568366085, + "grad_norm": 1.0187594890594482, + "learning_rate": 3.3396305830580665e-05, + "loss": 0.9591, + "step": 6577 + }, + { + "epoch": 0.45197196646969906, + "grad_norm": 1.132533311843872, + "learning_rate": 3.33906575660211e-05, + "loss": 0.9589, + "step": 6578 + }, + { + "epoch": 0.4520406761027896, + "grad_norm": 1.0676264762878418, + "learning_rate": 3.338500881875212e-05, + "loss": 0.9876, + "step": 6579 + }, + { + "epoch": 0.45210938573588016, + "grad_norm": 1.1530014276504517, + "learning_rate": 3.3379359589098684e-05, + "loss": 0.9844, + "step": 6580 + }, + { + "epoch": 0.45217809536897074, + "grad_norm": 1.1091017723083496, + "learning_rate": 3.337370987738581e-05, + "loss": 0.9662, + "step": 6581 + }, + { + "epoch": 0.45224680500206127, + "grad_norm": 1.0892490148544312, + "learning_rate": 3.336805968393851e-05, + "loss": 0.9902, + "step": 6582 + }, + { + "epoch": 0.45231551463515185, + "grad_norm": 1.0753928422927856, + "learning_rate": 3.336240900908185e-05, + "loss": 0.9908, + "step": 6583 + }, + { + "epoch": 0.45238422426824243, + "grad_norm": 1.0778125524520874, + "learning_rate": 3.335675785314088e-05, + "loss": 0.9645, + "step": 6584 + }, + { + "epoch": 0.45245293390133295, + "grad_norm": 1.0227055549621582, + "learning_rate": 3.335110621644073e-05, + "loss": 0.98, + "step": 6585 + }, + { + "epoch": 0.45252164353442353, + "grad_norm": 1.0451778173446655, + "learning_rate": 3.3345454099306545e-05, + "loss": 0.9861, + "step": 6586 + }, + { + "epoch": 0.45259035316751406, + "grad_norm": 1.068804383277893, + "learning_rate": 3.333980150206347e-05, + "loss": 0.9794, + "step": 6587 + }, + { + "epoch": 0.45265906280060464, + "grad_norm": 1.0848418474197388, + "learning_rate": 3.3334148425036696e-05, + "loss": 0.9327, + "step": 6588 + }, + { + "epoch": 0.4527277724336952, + "grad_norm": 1.0876373052597046, + "learning_rate": 3.332849486855144e-05, + "loss": 0.9952, + "step": 6589 + }, + { + "epoch": 0.45279648206678574, + "grad_norm": 1.0575705766677856, + "learning_rate": 3.3322840832932954e-05, + "loss": 0.9909, + "step": 6590 + }, + { + "epoch": 0.4528651916998763, + "grad_norm": 1.1748416423797607, + "learning_rate": 3.331718631850652e-05, + "loss": 0.9883, + "step": 6591 + }, + { + "epoch": 0.4529339013329669, + "grad_norm": 1.0562033653259277, + "learning_rate": 3.331153132559742e-05, + "loss": 0.9688, + "step": 6592 + }, + { + "epoch": 0.45300261096605743, + "grad_norm": 1.0306323766708374, + "learning_rate": 3.3305875854530985e-05, + "loss": 0.9614, + "step": 6593 + }, + { + "epoch": 0.453071320599148, + "grad_norm": 1.0812180042266846, + "learning_rate": 3.330021990563259e-05, + "loss": 0.9813, + "step": 6594 + }, + { + "epoch": 0.45314003023223853, + "grad_norm": 1.0467826128005981, + "learning_rate": 3.329456347922759e-05, + "loss": 0.9633, + "step": 6595 + }, + { + "epoch": 0.4532087398653291, + "grad_norm": 1.0834425687789917, + "learning_rate": 3.328890657564142e-05, + "loss": 0.9332, + "step": 6596 + }, + { + "epoch": 0.4532774494984197, + "grad_norm": 0.9957287311553955, + "learning_rate": 3.328324919519951e-05, + "loss": 0.9822, + "step": 6597 + }, + { + "epoch": 0.4533461591315102, + "grad_norm": 1.1309577226638794, + "learning_rate": 3.327759133822731e-05, + "loss": 0.9412, + "step": 6598 + }, + { + "epoch": 0.4534148687646008, + "grad_norm": 1.0599305629730225, + "learning_rate": 3.327193300505035e-05, + "loss": 0.9932, + "step": 6599 + }, + { + "epoch": 0.4534835783976914, + "grad_norm": 1.0354361534118652, + "learning_rate": 3.326627419599412e-05, + "loss": 0.9402, + "step": 6600 + }, + { + "epoch": 0.4535522880307819, + "grad_norm": 1.036450982093811, + "learning_rate": 3.326061491138417e-05, + "loss": 0.9826, + "step": 6601 + }, + { + "epoch": 0.4536209976638725, + "grad_norm": 1.111031174659729, + "learning_rate": 3.325495515154608e-05, + "loss": 0.9745, + "step": 6602 + }, + { + "epoch": 0.453689707296963, + "grad_norm": 1.1014206409454346, + "learning_rate": 3.3249294916805454e-05, + "loss": 0.9627, + "step": 6603 + }, + { + "epoch": 0.4537584169300536, + "grad_norm": 1.1640063524246216, + "learning_rate": 3.324363420748793e-05, + "loss": 1.0054, + "step": 6604 + }, + { + "epoch": 0.45382712656314417, + "grad_norm": 1.0530911684036255, + "learning_rate": 3.323797302391914e-05, + "loss": 1.0092, + "step": 6605 + }, + { + "epoch": 0.4538958361962347, + "grad_norm": 1.1500601768493652, + "learning_rate": 3.323231136642481e-05, + "loss": 1.0314, + "step": 6606 + }, + { + "epoch": 0.45396454582932527, + "grad_norm": 1.0854499340057373, + "learning_rate": 3.322664923533061e-05, + "loss": 0.9529, + "step": 6607 + }, + { + "epoch": 0.45403325546241585, + "grad_norm": 1.1024099588394165, + "learning_rate": 3.322098663096229e-05, + "loss": 0.9991, + "step": 6608 + }, + { + "epoch": 0.4541019650955064, + "grad_norm": 1.1042280197143555, + "learning_rate": 3.321532355364563e-05, + "loss": 0.974, + "step": 6609 + }, + { + "epoch": 0.45417067472859696, + "grad_norm": 1.0677436590194702, + "learning_rate": 3.320966000370641e-05, + "loss": 0.9765, + "step": 6610 + }, + { + "epoch": 0.4542393843616875, + "grad_norm": 1.0384461879730225, + "learning_rate": 3.320399598147047e-05, + "loss": 0.9457, + "step": 6611 + }, + { + "epoch": 0.45430809399477806, + "grad_norm": 1.0818812847137451, + "learning_rate": 3.319833148726363e-05, + "loss": 0.9769, + "step": 6612 + }, + { + "epoch": 0.45437680362786864, + "grad_norm": 1.1988089084625244, + "learning_rate": 3.319266652141178e-05, + "loss": 0.9988, + "step": 6613 + }, + { + "epoch": 0.45444551326095917, + "grad_norm": 1.0197821855545044, + "learning_rate": 3.3187001084240824e-05, + "loss": 0.997, + "step": 6614 + }, + { + "epoch": 0.45451422289404975, + "grad_norm": 0.9889854788780212, + "learning_rate": 3.3181335176076685e-05, + "loss": 0.9644, + "step": 6615 + }, + { + "epoch": 0.4545829325271403, + "grad_norm": 1.2144639492034912, + "learning_rate": 3.317566879724532e-05, + "loss": 1.0023, + "step": 6616 + }, + { + "epoch": 0.45465164216023085, + "grad_norm": 1.0714679956436157, + "learning_rate": 3.3170001948072714e-05, + "loss": 0.9712, + "step": 6617 + }, + { + "epoch": 0.45472035179332143, + "grad_norm": 1.0656789541244507, + "learning_rate": 3.316433462888488e-05, + "loss": 0.9799, + "step": 6618 + }, + { + "epoch": 0.45478906142641196, + "grad_norm": 1.046878457069397, + "learning_rate": 3.315866684000786e-05, + "loss": 0.9489, + "step": 6619 + }, + { + "epoch": 0.45485777105950254, + "grad_norm": 1.0921236276626587, + "learning_rate": 3.3152998581767694e-05, + "loss": 0.9752, + "step": 6620 + }, + { + "epoch": 0.4549264806925931, + "grad_norm": 1.0802826881408691, + "learning_rate": 3.31473298544905e-05, + "loss": 1.0052, + "step": 6621 + }, + { + "epoch": 0.45499519032568364, + "grad_norm": 1.0298963785171509, + "learning_rate": 3.314166065850239e-05, + "loss": 0.9625, + "step": 6622 + }, + { + "epoch": 0.4550638999587742, + "grad_norm": 1.083872675895691, + "learning_rate": 3.313599099412951e-05, + "loss": 0.9558, + "step": 6623 + }, + { + "epoch": 0.4551326095918648, + "grad_norm": 1.1035561561584473, + "learning_rate": 3.313032086169803e-05, + "loss": 0.9692, + "step": 6624 + }, + { + "epoch": 0.4552013192249553, + "grad_norm": 1.0644339323043823, + "learning_rate": 3.3124650261534136e-05, + "loss": 0.9548, + "step": 6625 + }, + { + "epoch": 0.4552700288580459, + "grad_norm": 1.0560461282730103, + "learning_rate": 3.311897919396407e-05, + "loss": 0.974, + "step": 6626 + }, + { + "epoch": 0.45533873849113643, + "grad_norm": 1.130772352218628, + "learning_rate": 3.311330765931409e-05, + "loss": 0.9272, + "step": 6627 + }, + { + "epoch": 0.455407448124227, + "grad_norm": 1.109862208366394, + "learning_rate": 3.310763565791045e-05, + "loss": 0.9704, + "step": 6628 + }, + { + "epoch": 0.4554761577573176, + "grad_norm": 1.0279146432876587, + "learning_rate": 3.310196319007949e-05, + "loss": 0.97, + "step": 6629 + }, + { + "epoch": 0.4555448673904081, + "grad_norm": 1.099230170249939, + "learning_rate": 3.309629025614752e-05, + "loss": 0.9849, + "step": 6630 + }, + { + "epoch": 0.4556135770234987, + "grad_norm": 1.1409600973129272, + "learning_rate": 3.309061685644091e-05, + "loss": 0.9758, + "step": 6631 + }, + { + "epoch": 0.4556822866565893, + "grad_norm": 1.094799518585205, + "learning_rate": 3.308494299128605e-05, + "loss": 1.0064, + "step": 6632 + }, + { + "epoch": 0.4557509962896798, + "grad_norm": 1.1197497844696045, + "learning_rate": 3.307926866100934e-05, + "loss": 0.9939, + "step": 6633 + }, + { + "epoch": 0.4558197059227704, + "grad_norm": 1.1090930700302124, + "learning_rate": 3.3073593865937234e-05, + "loss": 0.9866, + "step": 6634 + }, + { + "epoch": 0.4558884155558609, + "grad_norm": 1.0779615640640259, + "learning_rate": 3.3067918606396204e-05, + "loss": 0.9533, + "step": 6635 + }, + { + "epoch": 0.4559571251889515, + "grad_norm": 1.1003831624984741, + "learning_rate": 3.3062242882712724e-05, + "loss": 0.9507, + "step": 6636 + }, + { + "epoch": 0.45602583482204206, + "grad_norm": 1.0424506664276123, + "learning_rate": 3.3056566695213335e-05, + "loss": 0.9996, + "step": 6637 + }, + { + "epoch": 0.4560945444551326, + "grad_norm": 1.1116901636123657, + "learning_rate": 3.305089004422457e-05, + "loss": 0.9801, + "step": 6638 + }, + { + "epoch": 0.45616325408822317, + "grad_norm": 1.0519964694976807, + "learning_rate": 3.3045212930073014e-05, + "loss": 0.9716, + "step": 6639 + }, + { + "epoch": 0.45623196372131375, + "grad_norm": 1.1403875350952148, + "learning_rate": 3.303953535308526e-05, + "loss": 1.0433, + "step": 6640 + }, + { + "epoch": 0.4563006733544043, + "grad_norm": 1.019309163093567, + "learning_rate": 3.303385731358794e-05, + "loss": 0.9349, + "step": 6641 + }, + { + "epoch": 0.45636938298749485, + "grad_norm": 1.129399299621582, + "learning_rate": 3.3028178811907714e-05, + "loss": 1.0021, + "step": 6642 + }, + { + "epoch": 0.4564380926205854, + "grad_norm": 1.106167197227478, + "learning_rate": 3.302249984837125e-05, + "loss": 0.9521, + "step": 6643 + }, + { + "epoch": 0.45650680225367596, + "grad_norm": 1.117174506187439, + "learning_rate": 3.301682042330526e-05, + "loss": 1.0224, + "step": 6644 + }, + { + "epoch": 0.45657551188676654, + "grad_norm": 1.1681851148605347, + "learning_rate": 3.3011140537036476e-05, + "loss": 1.028, + "step": 6645 + }, + { + "epoch": 0.45664422151985706, + "grad_norm": 1.0896358489990234, + "learning_rate": 3.300546018989167e-05, + "loss": 0.9753, + "step": 6646 + }, + { + "epoch": 0.45671293115294764, + "grad_norm": 1.073257565498352, + "learning_rate": 3.2999779382197615e-05, + "loss": 0.9799, + "step": 6647 + }, + { + "epoch": 0.4567816407860382, + "grad_norm": 1.025431513786316, + "learning_rate": 3.2994098114281134e-05, + "loss": 0.9392, + "step": 6648 + }, + { + "epoch": 0.45685035041912875, + "grad_norm": 1.0618044137954712, + "learning_rate": 3.298841638646906e-05, + "loss": 0.9461, + "step": 6649 + }, + { + "epoch": 0.45691906005221933, + "grad_norm": 1.0861711502075195, + "learning_rate": 3.298273419908827e-05, + "loss": 0.9949, + "step": 6650 + }, + { + "epoch": 0.4569877696853099, + "grad_norm": 1.0383074283599854, + "learning_rate": 3.297705155246563e-05, + "loss": 1.0141, + "step": 6651 + }, + { + "epoch": 0.45705647931840043, + "grad_norm": 1.096114993095398, + "learning_rate": 3.29713684469281e-05, + "loss": 0.9823, + "step": 6652 + }, + { + "epoch": 0.457125188951491, + "grad_norm": 1.1534353494644165, + "learning_rate": 3.296568488280259e-05, + "loss": 0.9777, + "step": 6653 + }, + { + "epoch": 0.45719389858458154, + "grad_norm": 1.0555105209350586, + "learning_rate": 3.2960000860416084e-05, + "loss": 0.9662, + "step": 6654 + }, + { + "epoch": 0.4572626082176721, + "grad_norm": 1.0459089279174805, + "learning_rate": 3.295431638009559e-05, + "loss": 0.9948, + "step": 6655 + }, + { + "epoch": 0.4573313178507627, + "grad_norm": 1.0555084943771362, + "learning_rate": 3.294863144216812e-05, + "loss": 0.9515, + "step": 6656 + }, + { + "epoch": 0.4574000274838532, + "grad_norm": 1.0638128519058228, + "learning_rate": 3.294294604696073e-05, + "loss": 0.9741, + "step": 6657 + }, + { + "epoch": 0.4574687371169438, + "grad_norm": 1.0848950147628784, + "learning_rate": 3.29372601948005e-05, + "loss": 0.9942, + "step": 6658 + }, + { + "epoch": 0.4575374467500344, + "grad_norm": 1.0522549152374268, + "learning_rate": 3.2931573886014536e-05, + "loss": 1.0213, + "step": 6659 + }, + { + "epoch": 0.4576061563831249, + "grad_norm": 1.0633609294891357, + "learning_rate": 3.292588712092995e-05, + "loss": 1.0009, + "step": 6660 + }, + { + "epoch": 0.4576748660162155, + "grad_norm": 1.1294642686843872, + "learning_rate": 3.292019989987391e-05, + "loss": 1.0005, + "step": 6661 + }, + { + "epoch": 0.457743575649306, + "grad_norm": 1.1292229890823364, + "learning_rate": 3.2914512223173616e-05, + "loss": 1.0049, + "step": 6662 + }, + { + "epoch": 0.4578122852823966, + "grad_norm": 1.0754066705703735, + "learning_rate": 3.290882409115624e-05, + "loss": 1.004, + "step": 6663 + }, + { + "epoch": 0.45788099491548717, + "grad_norm": 1.041571021080017, + "learning_rate": 3.2903135504149046e-05, + "loss": 0.9939, + "step": 6664 + }, + { + "epoch": 0.4579497045485777, + "grad_norm": 1.1317527294158936, + "learning_rate": 3.2897446462479285e-05, + "loss": 0.9722, + "step": 6665 + }, + { + "epoch": 0.4580184141816683, + "grad_norm": 1.14284086227417, + "learning_rate": 3.289175696647424e-05, + "loss": 0.9601, + "step": 6666 + }, + { + "epoch": 0.45808712381475886, + "grad_norm": 1.074235439300537, + "learning_rate": 3.288606701646124e-05, + "loss": 0.9765, + "step": 6667 + }, + { + "epoch": 0.4581558334478494, + "grad_norm": 1.1169724464416504, + "learning_rate": 3.28803766127676e-05, + "loss": 0.9923, + "step": 6668 + }, + { + "epoch": 0.45822454308093996, + "grad_norm": 1.089281439781189, + "learning_rate": 3.28746857557207e-05, + "loss": 0.9814, + "step": 6669 + }, + { + "epoch": 0.4582932527140305, + "grad_norm": 1.1072981357574463, + "learning_rate": 3.286899444564794e-05, + "loss": 0.9422, + "step": 6670 + }, + { + "epoch": 0.45836196234712107, + "grad_norm": 1.0823099613189697, + "learning_rate": 3.2863302682876715e-05, + "loss": 0.9534, + "step": 6671 + }, + { + "epoch": 0.45843067198021165, + "grad_norm": 1.0982455015182495, + "learning_rate": 3.285761046773449e-05, + "loss": 0.9947, + "step": 6672 + }, + { + "epoch": 0.45849938161330217, + "grad_norm": 1.1417341232299805, + "learning_rate": 3.2851917800548724e-05, + "loss": 0.9763, + "step": 6673 + }, + { + "epoch": 0.45856809124639275, + "grad_norm": 1.0343687534332275, + "learning_rate": 3.284622468164691e-05, + "loss": 0.9573, + "step": 6674 + }, + { + "epoch": 0.45863680087948333, + "grad_norm": 1.119604229927063, + "learning_rate": 3.284053111135658e-05, + "loss": 1.0087, + "step": 6675 + }, + { + "epoch": 0.45870551051257386, + "grad_norm": 1.099144697189331, + "learning_rate": 3.2834837090005274e-05, + "loss": 0.9288, + "step": 6676 + }, + { + "epoch": 0.45877422014566444, + "grad_norm": 1.043051838874817, + "learning_rate": 3.282914261792056e-05, + "loss": 0.9881, + "step": 6677 + }, + { + "epoch": 0.45884292977875496, + "grad_norm": 1.0614495277404785, + "learning_rate": 3.282344769543005e-05, + "loss": 0.9718, + "step": 6678 + }, + { + "epoch": 0.45891163941184554, + "grad_norm": 1.1007343530654907, + "learning_rate": 3.281775232286137e-05, + "loss": 0.988, + "step": 6679 + }, + { + "epoch": 0.4589803490449361, + "grad_norm": 1.1040444374084473, + "learning_rate": 3.2812056500542164e-05, + "loss": 0.974, + "step": 6680 + }, + { + "epoch": 0.45904905867802664, + "grad_norm": 1.1129146814346313, + "learning_rate": 3.28063602288001e-05, + "loss": 0.9626, + "step": 6681 + }, + { + "epoch": 0.4591177683111172, + "grad_norm": 1.1083784103393555, + "learning_rate": 3.280066350796291e-05, + "loss": 0.9928, + "step": 6682 + }, + { + "epoch": 0.4591864779442078, + "grad_norm": 1.131178855895996, + "learning_rate": 3.2794966338358296e-05, + "loss": 0.9524, + "step": 6683 + }, + { + "epoch": 0.45925518757729833, + "grad_norm": 1.0870827436447144, + "learning_rate": 3.2789268720314013e-05, + "loss": 1.0194, + "step": 6684 + }, + { + "epoch": 0.4593238972103889, + "grad_norm": 1.1141996383666992, + "learning_rate": 3.278357065415787e-05, + "loss": 1.017, + "step": 6685 + }, + { + "epoch": 0.45939260684347943, + "grad_norm": 1.1885651350021362, + "learning_rate": 3.277787214021764e-05, + "loss": 0.9749, + "step": 6686 + }, + { + "epoch": 0.45946131647657, + "grad_norm": 1.0381147861480713, + "learning_rate": 3.277217317882116e-05, + "loss": 0.9071, + "step": 6687 + }, + { + "epoch": 0.4595300261096606, + "grad_norm": 1.0754002332687378, + "learning_rate": 3.27664737702963e-05, + "loss": 0.9398, + "step": 6688 + }, + { + "epoch": 0.4595987357427511, + "grad_norm": 1.0476593971252441, + "learning_rate": 3.2760773914970945e-05, + "loss": 0.9635, + "step": 6689 + }, + { + "epoch": 0.4596674453758417, + "grad_norm": 1.0408769845962524, + "learning_rate": 3.2755073613173006e-05, + "loss": 0.9656, + "step": 6690 + }, + { + "epoch": 0.4597361550089323, + "grad_norm": 1.0114295482635498, + "learning_rate": 3.27493728652304e-05, + "loss": 0.9392, + "step": 6691 + }, + { + "epoch": 0.4598048646420228, + "grad_norm": 1.1024879217147827, + "learning_rate": 3.2743671671471095e-05, + "loss": 1.0077, + "step": 6692 + }, + { + "epoch": 0.4598735742751134, + "grad_norm": 1.1032278537750244, + "learning_rate": 3.273797003222309e-05, + "loss": 0.93, + "step": 6693 + }, + { + "epoch": 0.4599422839082039, + "grad_norm": 1.0897806882858276, + "learning_rate": 3.2732267947814386e-05, + "loss": 0.979, + "step": 6694 + }, + { + "epoch": 0.4600109935412945, + "grad_norm": 1.093967080116272, + "learning_rate": 3.272656541857301e-05, + "loss": 0.9855, + "step": 6695 + }, + { + "epoch": 0.46007970317438507, + "grad_norm": 1.0973268747329712, + "learning_rate": 3.2720862444827046e-05, + "loss": 0.9636, + "step": 6696 + }, + { + "epoch": 0.4601484128074756, + "grad_norm": 1.0614945888519287, + "learning_rate": 3.2715159026904574e-05, + "loss": 0.9624, + "step": 6697 + }, + { + "epoch": 0.4602171224405662, + "grad_norm": 1.015808343887329, + "learning_rate": 3.270945516513371e-05, + "loss": 0.9701, + "step": 6698 + }, + { + "epoch": 0.46028583207365675, + "grad_norm": 1.034421682357788, + "learning_rate": 3.270375085984258e-05, + "loss": 0.9662, + "step": 6699 + }, + { + "epoch": 0.4603545417067473, + "grad_norm": 1.079347014427185, + "learning_rate": 3.2698046111359364e-05, + "loss": 0.9864, + "step": 6700 + }, + { + "epoch": 0.46042325133983786, + "grad_norm": 1.069149136543274, + "learning_rate": 3.269234092001226e-05, + "loss": 0.968, + "step": 6701 + }, + { + "epoch": 0.4604919609729284, + "grad_norm": 1.0081762075424194, + "learning_rate": 3.268663528612946e-05, + "loss": 0.965, + "step": 6702 + }, + { + "epoch": 0.46056067060601896, + "grad_norm": 1.013427972793579, + "learning_rate": 3.268092921003922e-05, + "loss": 0.959, + "step": 6703 + }, + { + "epoch": 0.46062938023910954, + "grad_norm": 1.0796793699264526, + "learning_rate": 3.267522269206981e-05, + "loss": 0.9392, + "step": 6704 + }, + { + "epoch": 0.46069808987220007, + "grad_norm": 1.0749032497406006, + "learning_rate": 3.266951573254952e-05, + "loss": 0.9583, + "step": 6705 + }, + { + "epoch": 0.46076679950529065, + "grad_norm": 1.1895140409469604, + "learning_rate": 3.2663808331806664e-05, + "loss": 1.02, + "step": 6706 + }, + { + "epoch": 0.4608355091383812, + "grad_norm": 1.0832619667053223, + "learning_rate": 3.265810049016959e-05, + "loss": 0.9895, + "step": 6707 + }, + { + "epoch": 0.46090421877147175, + "grad_norm": 1.040293574333191, + "learning_rate": 3.2652392207966664e-05, + "loss": 0.9823, + "step": 6708 + }, + { + "epoch": 0.46097292840456233, + "grad_norm": 1.0706592798233032, + "learning_rate": 3.264668348552627e-05, + "loss": 0.9658, + "step": 6709 + }, + { + "epoch": 0.46104163803765286, + "grad_norm": 1.044285774230957, + "learning_rate": 3.2640974323176846e-05, + "loss": 0.9634, + "step": 6710 + }, + { + "epoch": 0.46111034767074344, + "grad_norm": 1.0471235513687134, + "learning_rate": 3.263526472124683e-05, + "loss": 0.9871, + "step": 6711 + }, + { + "epoch": 0.461179057303834, + "grad_norm": 1.1662826538085938, + "learning_rate": 3.262955468006469e-05, + "loss": 0.9961, + "step": 6712 + }, + { + "epoch": 0.46124776693692454, + "grad_norm": 1.0290005207061768, + "learning_rate": 3.262384419995892e-05, + "loss": 0.9334, + "step": 6713 + }, + { + "epoch": 0.4613164765700151, + "grad_norm": 1.0337541103363037, + "learning_rate": 3.261813328125804e-05, + "loss": 0.9664, + "step": 6714 + }, + { + "epoch": 0.4613851862031057, + "grad_norm": 1.1055171489715576, + "learning_rate": 3.2612421924290596e-05, + "loss": 0.9446, + "step": 6715 + }, + { + "epoch": 0.4614538958361962, + "grad_norm": 1.1057571172714233, + "learning_rate": 3.2606710129385165e-05, + "loss": 0.9784, + "step": 6716 + }, + { + "epoch": 0.4615226054692868, + "grad_norm": 1.0795115232467651, + "learning_rate": 3.260099789687033e-05, + "loss": 0.9905, + "step": 6717 + }, + { + "epoch": 0.46159131510237733, + "grad_norm": 1.075189471244812, + "learning_rate": 3.259528522707473e-05, + "loss": 0.9946, + "step": 6718 + }, + { + "epoch": 0.4616600247354679, + "grad_norm": 1.0425522327423096, + "learning_rate": 3.2589572120327e-05, + "loss": 0.9591, + "step": 6719 + }, + { + "epoch": 0.4617287343685585, + "grad_norm": 1.0586262941360474, + "learning_rate": 3.258385857695581e-05, + "loss": 0.9769, + "step": 6720 + }, + { + "epoch": 0.461797444001649, + "grad_norm": 1.0738418102264404, + "learning_rate": 3.257814459728987e-05, + "loss": 0.9926, + "step": 6721 + }, + { + "epoch": 0.4618661536347396, + "grad_norm": 1.0714919567108154, + "learning_rate": 3.257243018165787e-05, + "loss": 0.9818, + "step": 6722 + }, + { + "epoch": 0.4619348632678302, + "grad_norm": 1.100171685218811, + "learning_rate": 3.2566715330388605e-05, + "loss": 1.0069, + "step": 6723 + }, + { + "epoch": 0.4620035729009207, + "grad_norm": 1.0086168050765991, + "learning_rate": 3.25610000438108e-05, + "loss": 0.9404, + "step": 6724 + }, + { + "epoch": 0.4620722825340113, + "grad_norm": 1.0603337287902832, + "learning_rate": 3.2555284322253284e-05, + "loss": 0.9333, + "step": 6725 + }, + { + "epoch": 0.4621409921671018, + "grad_norm": 1.1212586164474487, + "learning_rate": 3.254956816604486e-05, + "loss": 1.0019, + "step": 6726 + }, + { + "epoch": 0.4622097018001924, + "grad_norm": 1.0510433912277222, + "learning_rate": 3.2543851575514386e-05, + "loss": 0.9935, + "step": 6727 + }, + { + "epoch": 0.46227841143328297, + "grad_norm": 1.0472825765609741, + "learning_rate": 3.253813455099074e-05, + "loss": 0.9568, + "step": 6728 + }, + { + "epoch": 0.4623471210663735, + "grad_norm": 1.0694239139556885, + "learning_rate": 3.25324170928028e-05, + "loss": 0.9437, + "step": 6729 + }, + { + "epoch": 0.46241583069946407, + "grad_norm": 1.063399314880371, + "learning_rate": 3.2526699201279496e-05, + "loss": 0.9946, + "step": 6730 + }, + { + "epoch": 0.46248454033255465, + "grad_norm": 1.1217241287231445, + "learning_rate": 3.2520980876749785e-05, + "loss": 0.9799, + "step": 6731 + }, + { + "epoch": 0.4625532499656452, + "grad_norm": 1.0153027772903442, + "learning_rate": 3.251526211954262e-05, + "loss": 0.9643, + "step": 6732 + }, + { + "epoch": 0.46262195959873575, + "grad_norm": 1.0563148260116577, + "learning_rate": 3.2509542929987017e-05, + "loss": 0.9537, + "step": 6733 + }, + { + "epoch": 0.4626906692318263, + "grad_norm": 1.0958130359649658, + "learning_rate": 3.250382330841198e-05, + "loss": 0.941, + "step": 6734 + }, + { + "epoch": 0.46275937886491686, + "grad_norm": 1.0259239673614502, + "learning_rate": 3.249810325514656e-05, + "loss": 0.9498, + "step": 6735 + }, + { + "epoch": 0.46282808849800744, + "grad_norm": 1.1360687017440796, + "learning_rate": 3.249238277051984e-05, + "loss": 0.9839, + "step": 6736 + }, + { + "epoch": 0.46289679813109796, + "grad_norm": 1.1213382482528687, + "learning_rate": 3.248666185486091e-05, + "loss": 0.977, + "step": 6737 + }, + { + "epoch": 0.46296550776418854, + "grad_norm": 1.147067666053772, + "learning_rate": 3.2480940508498875e-05, + "loss": 0.9988, + "step": 6738 + }, + { + "epoch": 0.4630342173972791, + "grad_norm": 1.0878669023513794, + "learning_rate": 3.247521873176291e-05, + "loss": 0.9375, + "step": 6739 + }, + { + "epoch": 0.46310292703036965, + "grad_norm": 1.1358462572097778, + "learning_rate": 3.246949652498216e-05, + "loss": 0.983, + "step": 6740 + }, + { + "epoch": 0.46317163666346023, + "grad_norm": 1.0654540061950684, + "learning_rate": 3.2463773888485823e-05, + "loss": 0.9586, + "step": 6741 + }, + { + "epoch": 0.46324034629655075, + "grad_norm": 1.1137415170669556, + "learning_rate": 3.2458050822603145e-05, + "loss": 1.0017, + "step": 6742 + }, + { + "epoch": 0.46330905592964133, + "grad_norm": 1.0922024250030518, + "learning_rate": 3.245232732766333e-05, + "loss": 0.9837, + "step": 6743 + }, + { + "epoch": 0.4633777655627319, + "grad_norm": 1.055667757987976, + "learning_rate": 3.244660340399568e-05, + "loss": 1.0132, + "step": 6744 + }, + { + "epoch": 0.46344647519582244, + "grad_norm": 1.1739027500152588, + "learning_rate": 3.2440879051929475e-05, + "loss": 0.9873, + "step": 6745 + }, + { + "epoch": 0.463515184828913, + "grad_norm": 1.069902777671814, + "learning_rate": 3.243515427179403e-05, + "loss": 0.9662, + "step": 6746 + }, + { + "epoch": 0.4635838944620036, + "grad_norm": 1.0424573421478271, + "learning_rate": 3.2429429063918696e-05, + "loss": 0.9507, + "step": 6747 + }, + { + "epoch": 0.4636526040950941, + "grad_norm": 1.1582660675048828, + "learning_rate": 3.242370342863284e-05, + "loss": 0.9991, + "step": 6748 + }, + { + "epoch": 0.4637213137281847, + "grad_norm": 1.1470659971237183, + "learning_rate": 3.241797736626585e-05, + "loss": 1.0004, + "step": 6749 + }, + { + "epoch": 0.4637900233612752, + "grad_norm": 1.0695390701293945, + "learning_rate": 3.241225087714714e-05, + "loss": 0.9549, + "step": 6750 + }, + { + "epoch": 0.4638587329943658, + "grad_norm": 1.0169528722763062, + "learning_rate": 3.240652396160617e-05, + "loss": 0.9522, + "step": 6751 + }, + { + "epoch": 0.4639274426274564, + "grad_norm": 1.0632281303405762, + "learning_rate": 3.240079661997238e-05, + "loss": 0.9365, + "step": 6752 + }, + { + "epoch": 0.4639961522605469, + "grad_norm": 1.0581711530685425, + "learning_rate": 3.2395068852575284e-05, + "loss": 0.9708, + "step": 6753 + }, + { + "epoch": 0.4640648618936375, + "grad_norm": 1.1772300004959106, + "learning_rate": 3.2389340659744375e-05, + "loss": 0.9862, + "step": 6754 + }, + { + "epoch": 0.4641335715267281, + "grad_norm": 1.0632858276367188, + "learning_rate": 3.238361204180921e-05, + "loss": 1.0065, + "step": 6755 + }, + { + "epoch": 0.4642022811598186, + "grad_norm": 1.0892151594161987, + "learning_rate": 3.237788299909934e-05, + "loss": 0.9419, + "step": 6756 + }, + { + "epoch": 0.4642709907929092, + "grad_norm": 1.0896213054656982, + "learning_rate": 3.2372153531944375e-05, + "loss": 0.9704, + "step": 6757 + }, + { + "epoch": 0.4643397004259997, + "grad_norm": 1.034374713897705, + "learning_rate": 3.2366423640673894e-05, + "loss": 0.9969, + "step": 6758 + }, + { + "epoch": 0.4644084100590903, + "grad_norm": 1.0345654487609863, + "learning_rate": 3.236069332561756e-05, + "loss": 0.9489, + "step": 6759 + }, + { + "epoch": 0.46447711969218086, + "grad_norm": 1.253265380859375, + "learning_rate": 3.2354962587105025e-05, + "loss": 1.0131, + "step": 6760 + }, + { + "epoch": 0.4645458293252714, + "grad_norm": 1.0422464609146118, + "learning_rate": 3.2349231425465974e-05, + "loss": 0.9879, + "step": 6761 + }, + { + "epoch": 0.46461453895836197, + "grad_norm": 1.1369699239730835, + "learning_rate": 3.234349984103014e-05, + "loss": 0.9935, + "step": 6762 + }, + { + "epoch": 0.46468324859145255, + "grad_norm": 1.021132469177246, + "learning_rate": 3.233776783412721e-05, + "loss": 0.9549, + "step": 6763 + }, + { + "epoch": 0.46475195822454307, + "grad_norm": 1.0447628498077393, + "learning_rate": 3.2332035405086985e-05, + "loss": 0.9867, + "step": 6764 + }, + { + "epoch": 0.46482066785763365, + "grad_norm": 1.043485164642334, + "learning_rate": 3.232630255423923e-05, + "loss": 0.9487, + "step": 6765 + }, + { + "epoch": 0.4648893774907242, + "grad_norm": 1.0781848430633545, + "learning_rate": 3.232056928191376e-05, + "loss": 0.9598, + "step": 6766 + }, + { + "epoch": 0.46495808712381476, + "grad_norm": 1.0171785354614258, + "learning_rate": 3.231483558844041e-05, + "loss": 0.9625, + "step": 6767 + }, + { + "epoch": 0.46502679675690534, + "grad_norm": 1.0894280672073364, + "learning_rate": 3.2309101474149015e-05, + "loss": 0.9458, + "step": 6768 + }, + { + "epoch": 0.46509550638999586, + "grad_norm": 1.1751562356948853, + "learning_rate": 3.230336693936947e-05, + "loss": 0.9993, + "step": 6769 + }, + { + "epoch": 0.46516421602308644, + "grad_norm": 1.0915789604187012, + "learning_rate": 3.229763198443168e-05, + "loss": 0.9787, + "step": 6770 + }, + { + "epoch": 0.465232925656177, + "grad_norm": 1.1320760250091553, + "learning_rate": 3.2291896609665575e-05, + "loss": 0.9679, + "step": 6771 + }, + { + "epoch": 0.46530163528926755, + "grad_norm": 1.0984923839569092, + "learning_rate": 3.22861608154011e-05, + "loss": 0.992, + "step": 6772 + }, + { + "epoch": 0.4653703449223581, + "grad_norm": 1.0862516164779663, + "learning_rate": 3.228042460196824e-05, + "loss": 1.0049, + "step": 6773 + }, + { + "epoch": 0.46543905455544865, + "grad_norm": 1.0706868171691895, + "learning_rate": 3.2274687969696995e-05, + "loss": 0.9533, + "step": 6774 + }, + { + "epoch": 0.46550776418853923, + "grad_norm": 1.0953491926193237, + "learning_rate": 3.2268950918917376e-05, + "loss": 0.9807, + "step": 6775 + }, + { + "epoch": 0.4655764738216298, + "grad_norm": 1.0780001878738403, + "learning_rate": 3.226321344995945e-05, + "loss": 0.9539, + "step": 6776 + }, + { + "epoch": 0.46564518345472034, + "grad_norm": 1.0565428733825684, + "learning_rate": 3.2257475563153276e-05, + "loss": 0.9888, + "step": 6777 + }, + { + "epoch": 0.4657138930878109, + "grad_norm": 1.0444800853729248, + "learning_rate": 3.2251737258828965e-05, + "loss": 0.9731, + "step": 6778 + }, + { + "epoch": 0.4657826027209015, + "grad_norm": 1.0464046001434326, + "learning_rate": 3.2245998537316635e-05, + "loss": 0.9698, + "step": 6779 + }, + { + "epoch": 0.465851312353992, + "grad_norm": 1.0756593942642212, + "learning_rate": 3.224025939894642e-05, + "loss": 0.9621, + "step": 6780 + }, + { + "epoch": 0.4659200219870826, + "grad_norm": 1.0486458539962769, + "learning_rate": 3.2234519844048494e-05, + "loss": 0.9573, + "step": 6781 + }, + { + "epoch": 0.4659887316201731, + "grad_norm": 1.149466872215271, + "learning_rate": 3.2228779872953065e-05, + "loss": 0.9799, + "step": 6782 + }, + { + "epoch": 0.4660574412532637, + "grad_norm": 1.042209506034851, + "learning_rate": 3.222303948599032e-05, + "loss": 0.9414, + "step": 6783 + }, + { + "epoch": 0.4661261508863543, + "grad_norm": 1.0921306610107422, + "learning_rate": 3.2217298683490525e-05, + "loss": 1.0055, + "step": 6784 + }, + { + "epoch": 0.4661948605194448, + "grad_norm": 1.1541039943695068, + "learning_rate": 3.2211557465783944e-05, + "loss": 0.9794, + "step": 6785 + }, + { + "epoch": 0.4662635701525354, + "grad_norm": 1.0933316946029663, + "learning_rate": 3.220581583320085e-05, + "loss": 0.9691, + "step": 6786 + }, + { + "epoch": 0.46633227978562597, + "grad_norm": 1.1274737119674683, + "learning_rate": 3.220007378607157e-05, + "loss": 0.9239, + "step": 6787 + }, + { + "epoch": 0.4664009894187165, + "grad_norm": 1.1327219009399414, + "learning_rate": 3.2194331324726434e-05, + "loss": 0.9866, + "step": 6788 + }, + { + "epoch": 0.4664696990518071, + "grad_norm": 1.0425777435302734, + "learning_rate": 3.218858844949581e-05, + "loss": 0.97, + "step": 6789 + }, + { + "epoch": 0.4665384086848976, + "grad_norm": 1.1779794692993164, + "learning_rate": 3.2182845160710064e-05, + "loss": 1.0091, + "step": 6790 + }, + { + "epoch": 0.4666071183179882, + "grad_norm": 1.1059526205062866, + "learning_rate": 3.2177101458699624e-05, + "loss": 0.9808, + "step": 6791 + }, + { + "epoch": 0.46667582795107876, + "grad_norm": 1.0584979057312012, + "learning_rate": 3.21713573437949e-05, + "loss": 0.9894, + "step": 6792 + }, + { + "epoch": 0.4667445375841693, + "grad_norm": 1.1513476371765137, + "learning_rate": 3.216561281632637e-05, + "loss": 1.0, + "step": 6793 + }, + { + "epoch": 0.46681324721725986, + "grad_norm": 1.0268548727035522, + "learning_rate": 3.2159867876624504e-05, + "loss": 0.9218, + "step": 6794 + }, + { + "epoch": 0.46688195685035044, + "grad_norm": 1.065699815750122, + "learning_rate": 3.215412252501981e-05, + "loss": 0.9486, + "step": 6795 + }, + { + "epoch": 0.46695066648344097, + "grad_norm": 1.0509895086288452, + "learning_rate": 3.2148376761842805e-05, + "loss": 0.9586, + "step": 6796 + }, + { + "epoch": 0.46701937611653155, + "grad_norm": 1.1118853092193604, + "learning_rate": 3.2142630587424045e-05, + "loss": 0.9706, + "step": 6797 + }, + { + "epoch": 0.4670880857496221, + "grad_norm": 1.2042099237442017, + "learning_rate": 3.213688400209409e-05, + "loss": 0.9813, + "step": 6798 + }, + { + "epoch": 0.46715679538271265, + "grad_norm": 1.0926086902618408, + "learning_rate": 3.213113700618357e-05, + "loss": 1.0213, + "step": 6799 + }, + { + "epoch": 0.46722550501580323, + "grad_norm": 1.0971020460128784, + "learning_rate": 3.212538960002307e-05, + "loss": 0.9676, + "step": 6800 + }, + { + "epoch": 0.46729421464889376, + "grad_norm": 1.060280442237854, + "learning_rate": 3.211964178394326e-05, + "loss": 0.9647, + "step": 6801 + }, + { + "epoch": 0.46736292428198434, + "grad_norm": 1.1684585809707642, + "learning_rate": 3.21138935582748e-05, + "loss": 0.9924, + "step": 6802 + }, + { + "epoch": 0.4674316339150749, + "grad_norm": 1.0634584426879883, + "learning_rate": 3.210814492334837e-05, + "loss": 0.9483, + "step": 6803 + }, + { + "epoch": 0.46750034354816544, + "grad_norm": 1.0737053155899048, + "learning_rate": 3.2102395879494705e-05, + "loss": 0.9593, + "step": 6804 + }, + { + "epoch": 0.467569053181256, + "grad_norm": 1.0812420845031738, + "learning_rate": 3.209664642704454e-05, + "loss": 0.9891, + "step": 6805 + }, + { + "epoch": 0.46763776281434655, + "grad_norm": 1.0331324338912964, + "learning_rate": 3.2090896566328624e-05, + "loss": 0.9047, + "step": 6806 + }, + { + "epoch": 0.4677064724474371, + "grad_norm": 1.165490984916687, + "learning_rate": 3.208514629767776e-05, + "loss": 0.9851, + "step": 6807 + }, + { + "epoch": 0.4677751820805277, + "grad_norm": 1.0420506000518799, + "learning_rate": 3.207939562142274e-05, + "loss": 0.9568, + "step": 6808 + }, + { + "epoch": 0.46784389171361823, + "grad_norm": 1.1534576416015625, + "learning_rate": 3.2073644537894407e-05, + "loss": 1.007, + "step": 6809 + }, + { + "epoch": 0.4679126013467088, + "grad_norm": 1.16647207736969, + "learning_rate": 3.2067893047423626e-05, + "loss": 0.9779, + "step": 6810 + }, + { + "epoch": 0.4679813109797994, + "grad_norm": 1.1115508079528809, + "learning_rate": 3.2062141150341265e-05, + "loss": 0.9979, + "step": 6811 + }, + { + "epoch": 0.4680500206128899, + "grad_norm": 1.1094621419906616, + "learning_rate": 3.205638884697822e-05, + "loss": 0.9582, + "step": 6812 + }, + { + "epoch": 0.4681187302459805, + "grad_norm": 1.101922869682312, + "learning_rate": 3.205063613766543e-05, + "loss": 1.0129, + "step": 6813 + }, + { + "epoch": 0.468187439879071, + "grad_norm": 1.121812343597412, + "learning_rate": 3.204488302273386e-05, + "loss": 0.9888, + "step": 6814 + }, + { + "epoch": 0.4682561495121616, + "grad_norm": 1.0201653242111206, + "learning_rate": 3.2039129502514455e-05, + "loss": 0.9714, + "step": 6815 + }, + { + "epoch": 0.4683248591452522, + "grad_norm": 1.0899693965911865, + "learning_rate": 3.2033375577338215e-05, + "loss": 0.9375, + "step": 6816 + }, + { + "epoch": 0.4683935687783427, + "grad_norm": 1.0842530727386475, + "learning_rate": 3.2027621247536165e-05, + "loss": 0.995, + "step": 6817 + }, + { + "epoch": 0.4684622784114333, + "grad_norm": 1.1424603462219238, + "learning_rate": 3.202186651343936e-05, + "loss": 1.0195, + "step": 6818 + }, + { + "epoch": 0.46853098804452387, + "grad_norm": 1.047404408454895, + "learning_rate": 3.201611137537885e-05, + "loss": 0.9577, + "step": 6819 + }, + { + "epoch": 0.4685996976776144, + "grad_norm": 1.147681474685669, + "learning_rate": 3.201035583368573e-05, + "loss": 0.9606, + "step": 6820 + }, + { + "epoch": 0.46866840731070497, + "grad_norm": 1.0161904096603394, + "learning_rate": 3.200459988869111e-05, + "loss": 0.9527, + "step": 6821 + }, + { + "epoch": 0.4687371169437955, + "grad_norm": 1.1316505670547485, + "learning_rate": 3.1998843540726135e-05, + "loss": 0.9767, + "step": 6822 + }, + { + "epoch": 0.4688058265768861, + "grad_norm": 1.0719739198684692, + "learning_rate": 3.199308679012196e-05, + "loss": 0.9267, + "step": 6823 + }, + { + "epoch": 0.46887453620997666, + "grad_norm": 1.0413422584533691, + "learning_rate": 3.1987329637209765e-05, + "loss": 0.9652, + "step": 6824 + }, + { + "epoch": 0.4689432458430672, + "grad_norm": 1.061252474784851, + "learning_rate": 3.1981572082320756e-05, + "loss": 0.9641, + "step": 6825 + }, + { + "epoch": 0.46901195547615776, + "grad_norm": 1.1117939949035645, + "learning_rate": 3.1975814125786166e-05, + "loss": 1.0034, + "step": 6826 + }, + { + "epoch": 0.46908066510924834, + "grad_norm": 1.0653941631317139, + "learning_rate": 3.197005576793723e-05, + "loss": 0.9619, + "step": 6827 + }, + { + "epoch": 0.46914937474233886, + "grad_norm": 1.11330246925354, + "learning_rate": 3.196429700910525e-05, + "loss": 0.9995, + "step": 6828 + }, + { + "epoch": 0.46921808437542945, + "grad_norm": 1.1230071783065796, + "learning_rate": 3.19585378496215e-05, + "loss": 0.9663, + "step": 6829 + }, + { + "epoch": 0.46928679400851997, + "grad_norm": 1.1047347784042358, + "learning_rate": 3.195277828981731e-05, + "loss": 1.0003, + "step": 6830 + }, + { + "epoch": 0.46935550364161055, + "grad_norm": 1.0991228818893433, + "learning_rate": 3.1947018330024026e-05, + "loss": 0.9995, + "step": 6831 + }, + { + "epoch": 0.46942421327470113, + "grad_norm": 1.072390079498291, + "learning_rate": 3.1941257970573015e-05, + "loss": 0.9709, + "step": 6832 + }, + { + "epoch": 0.46949292290779165, + "grad_norm": 1.070662260055542, + "learning_rate": 3.193549721179566e-05, + "loss": 0.9697, + "step": 6833 + }, + { + "epoch": 0.46956163254088223, + "grad_norm": 1.0023475885391235, + "learning_rate": 3.192973605402339e-05, + "loss": 0.9208, + "step": 6834 + }, + { + "epoch": 0.4696303421739728, + "grad_norm": 1.066493272781372, + "learning_rate": 3.1923974497587625e-05, + "loss": 0.9337, + "step": 6835 + }, + { + "epoch": 0.46969905180706334, + "grad_norm": 1.1033072471618652, + "learning_rate": 3.1918212542819826e-05, + "loss": 0.9752, + "step": 6836 + }, + { + "epoch": 0.4697677614401539, + "grad_norm": 1.0308811664581299, + "learning_rate": 3.191245019005148e-05, + "loss": 0.9762, + "step": 6837 + }, + { + "epoch": 0.46983647107324444, + "grad_norm": 1.0100877285003662, + "learning_rate": 3.190668743961409e-05, + "loss": 0.9488, + "step": 6838 + }, + { + "epoch": 0.469905180706335, + "grad_norm": 1.051715612411499, + "learning_rate": 3.190092429183918e-05, + "loss": 0.9512, + "step": 6839 + }, + { + "epoch": 0.4699738903394256, + "grad_norm": 1.1360456943511963, + "learning_rate": 3.189516074705831e-05, + "loss": 1.0267, + "step": 6840 + }, + { + "epoch": 0.47004259997251613, + "grad_norm": 1.0745872259140015, + "learning_rate": 3.188939680560304e-05, + "loss": 0.9691, + "step": 6841 + }, + { + "epoch": 0.4701113096056067, + "grad_norm": 1.070771336555481, + "learning_rate": 3.188363246780497e-05, + "loss": 0.9715, + "step": 6842 + }, + { + "epoch": 0.4701800192386973, + "grad_norm": 1.1049072742462158, + "learning_rate": 3.187786773399573e-05, + "loss": 0.9514, + "step": 6843 + }, + { + "epoch": 0.4702487288717878, + "grad_norm": 1.1183357238769531, + "learning_rate": 3.1872102604506945e-05, + "loss": 0.9639, + "step": 6844 + }, + { + "epoch": 0.4703174385048784, + "grad_norm": 1.066472053527832, + "learning_rate": 3.1866337079670284e-05, + "loss": 0.9938, + "step": 6845 + }, + { + "epoch": 0.4703861481379689, + "grad_norm": 1.1055848598480225, + "learning_rate": 3.186057115981745e-05, + "loss": 1.0204, + "step": 6846 + }, + { + "epoch": 0.4704548577710595, + "grad_norm": 1.0609352588653564, + "learning_rate": 3.185480484528013e-05, + "loss": 0.9673, + "step": 6847 + }, + { + "epoch": 0.4705235674041501, + "grad_norm": 1.0876115560531616, + "learning_rate": 3.1849038136390075e-05, + "loss": 1.0084, + "step": 6848 + }, + { + "epoch": 0.4705922770372406, + "grad_norm": 1.1362733840942383, + "learning_rate": 3.1843271033479026e-05, + "loss": 0.9964, + "step": 6849 + }, + { + "epoch": 0.4706609866703312, + "grad_norm": 1.0404099225997925, + "learning_rate": 3.183750353687877e-05, + "loss": 0.996, + "step": 6850 + }, + { + "epoch": 0.47072969630342176, + "grad_norm": 1.060478687286377, + "learning_rate": 3.1831735646921106e-05, + "loss": 0.9712, + "step": 6851 + }, + { + "epoch": 0.4707984059365123, + "grad_norm": 0.9812084436416626, + "learning_rate": 3.1825967363937856e-05, + "loss": 0.9693, + "step": 6852 + }, + { + "epoch": 0.47086711556960287, + "grad_norm": 1.0615413188934326, + "learning_rate": 3.182019868826086e-05, + "loss": 0.9496, + "step": 6853 + }, + { + "epoch": 0.4709358252026934, + "grad_norm": 1.080780267715454, + "learning_rate": 3.181442962022199e-05, + "loss": 0.942, + "step": 6854 + }, + { + "epoch": 0.471004534835784, + "grad_norm": 1.0275993347167969, + "learning_rate": 3.180866016015315e-05, + "loss": 0.9762, + "step": 6855 + }, + { + "epoch": 0.47107324446887455, + "grad_norm": 1.0958956480026245, + "learning_rate": 3.180289030838624e-05, + "loss": 0.9691, + "step": 6856 + }, + { + "epoch": 0.4711419541019651, + "grad_norm": 1.1414638757705688, + "learning_rate": 3.1797120065253184e-05, + "loss": 0.9729, + "step": 6857 + }, + { + "epoch": 0.47121066373505566, + "grad_norm": 1.079981803894043, + "learning_rate": 3.179134943108597e-05, + "loss": 0.9966, + "step": 6858 + }, + { + "epoch": 0.47127937336814624, + "grad_norm": 1.0415412187576294, + "learning_rate": 3.178557840621656e-05, + "loss": 0.9654, + "step": 6859 + }, + { + "epoch": 0.47134808300123676, + "grad_norm": 1.13535737991333, + "learning_rate": 3.177980699097696e-05, + "loss": 0.9346, + "step": 6860 + }, + { + "epoch": 0.47141679263432734, + "grad_norm": 1.1362208127975464, + "learning_rate": 3.17740351856992e-05, + "loss": 0.9409, + "step": 6861 + }, + { + "epoch": 0.47148550226741787, + "grad_norm": 1.107728362083435, + "learning_rate": 3.176826299071533e-05, + "loss": 0.9577, + "step": 6862 + }, + { + "epoch": 0.47155421190050845, + "grad_norm": 1.049482822418213, + "learning_rate": 3.1762490406357415e-05, + "loss": 0.9782, + "step": 6863 + }, + { + "epoch": 0.471622921533599, + "grad_norm": 1.0336154699325562, + "learning_rate": 3.1756717432957544e-05, + "loss": 0.9809, + "step": 6864 + }, + { + "epoch": 0.47169163116668955, + "grad_norm": 1.056215524673462, + "learning_rate": 3.175094407084784e-05, + "loss": 0.9362, + "step": 6865 + }, + { + "epoch": 0.47176034079978013, + "grad_norm": 1.0441627502441406, + "learning_rate": 3.1745170320360444e-05, + "loss": 0.9502, + "step": 6866 + }, + { + "epoch": 0.4718290504328707, + "grad_norm": 1.132157564163208, + "learning_rate": 3.1739396181827506e-05, + "loss": 0.98, + "step": 6867 + }, + { + "epoch": 0.47189776006596124, + "grad_norm": 1.1467186212539673, + "learning_rate": 3.173362165558122e-05, + "loss": 0.9643, + "step": 6868 + }, + { + "epoch": 0.4719664696990518, + "grad_norm": 1.095137596130371, + "learning_rate": 3.172784674195379e-05, + "loss": 0.9943, + "step": 6869 + }, + { + "epoch": 0.47203517933214234, + "grad_norm": 1.0592131614685059, + "learning_rate": 3.172207144127743e-05, + "loss": 0.9711, + "step": 6870 + }, + { + "epoch": 0.4721038889652329, + "grad_norm": 1.0115705728530884, + "learning_rate": 3.171629575388441e-05, + "loss": 0.9756, + "step": 6871 + }, + { + "epoch": 0.4721725985983235, + "grad_norm": 1.0694818496704102, + "learning_rate": 3.1710519680106985e-05, + "loss": 0.987, + "step": 6872 + }, + { + "epoch": 0.472241308231414, + "grad_norm": 1.0876014232635498, + "learning_rate": 3.170474322027745e-05, + "loss": 0.9784, + "step": 6873 + }, + { + "epoch": 0.4723100178645046, + "grad_norm": 1.1098145246505737, + "learning_rate": 3.169896637472813e-05, + "loss": 0.9645, + "step": 6874 + }, + { + "epoch": 0.4723787274975952, + "grad_norm": 1.0461699962615967, + "learning_rate": 3.169318914379136e-05, + "loss": 0.9492, + "step": 6875 + }, + { + "epoch": 0.4724474371306857, + "grad_norm": 1.0443247556686401, + "learning_rate": 3.16874115277995e-05, + "loss": 0.9676, + "step": 6876 + }, + { + "epoch": 0.4725161467637763, + "grad_norm": 1.1485347747802734, + "learning_rate": 3.168163352708492e-05, + "loss": 0.9816, + "step": 6877 + }, + { + "epoch": 0.4725848563968668, + "grad_norm": 1.0536645650863647, + "learning_rate": 3.167585514198005e-05, + "loss": 0.9889, + "step": 6878 + }, + { + "epoch": 0.4726535660299574, + "grad_norm": 1.0549578666687012, + "learning_rate": 3.1670076372817304e-05, + "loss": 0.9376, + "step": 6879 + }, + { + "epoch": 0.472722275663048, + "grad_norm": 1.0654493570327759, + "learning_rate": 3.166429721992912e-05, + "loss": 0.9373, + "step": 6880 + }, + { + "epoch": 0.4727909852961385, + "grad_norm": 1.057033896446228, + "learning_rate": 3.1658517683648e-05, + "loss": 0.9421, + "step": 6881 + }, + { + "epoch": 0.4728596949292291, + "grad_norm": 1.071701169013977, + "learning_rate": 3.16527377643064e-05, + "loss": 0.9799, + "step": 6882 + }, + { + "epoch": 0.47292840456231966, + "grad_norm": 1.0390890836715698, + "learning_rate": 3.1646957462236856e-05, + "loss": 0.9713, + "step": 6883 + }, + { + "epoch": 0.4729971141954102, + "grad_norm": 1.0896351337432861, + "learning_rate": 3.164117677777191e-05, + "loss": 0.9811, + "step": 6884 + }, + { + "epoch": 0.47306582382850076, + "grad_norm": 1.12496018409729, + "learning_rate": 3.163539571124411e-05, + "loss": 1.0021, + "step": 6885 + }, + { + "epoch": 0.4731345334615913, + "grad_norm": 1.1655055284500122, + "learning_rate": 3.162961426298604e-05, + "loss": 0.9948, + "step": 6886 + }, + { + "epoch": 0.47320324309468187, + "grad_norm": 1.0777440071105957, + "learning_rate": 3.16238324333303e-05, + "loss": 1.0028, + "step": 6887 + }, + { + "epoch": 0.47327195272777245, + "grad_norm": 1.1369719505310059, + "learning_rate": 3.161805022260952e-05, + "loss": 0.9676, + "step": 6888 + }, + { + "epoch": 0.473340662360863, + "grad_norm": 1.08415949344635, + "learning_rate": 3.1612267631156354e-05, + "loss": 0.9783, + "step": 6889 + }, + { + "epoch": 0.47340937199395355, + "grad_norm": 1.0549311637878418, + "learning_rate": 3.1606484659303456e-05, + "loss": 0.9639, + "step": 6890 + }, + { + "epoch": 0.47347808162704413, + "grad_norm": 1.0864264965057373, + "learning_rate": 3.160070130738352e-05, + "loss": 0.9616, + "step": 6891 + }, + { + "epoch": 0.47354679126013466, + "grad_norm": 1.1189779043197632, + "learning_rate": 3.159491757572927e-05, + "loss": 0.9628, + "step": 6892 + }, + { + "epoch": 0.47361550089322524, + "grad_norm": 1.052039623260498, + "learning_rate": 3.158913346467343e-05, + "loss": 0.9568, + "step": 6893 + }, + { + "epoch": 0.47368421052631576, + "grad_norm": 1.0762981176376343, + "learning_rate": 3.1583348974548754e-05, + "loss": 0.9391, + "step": 6894 + }, + { + "epoch": 0.47375292015940634, + "grad_norm": 1.016418695449829, + "learning_rate": 3.157756410568803e-05, + "loss": 0.9613, + "step": 6895 + }, + { + "epoch": 0.4738216297924969, + "grad_norm": 1.0634615421295166, + "learning_rate": 3.1571778858424065e-05, + "loss": 0.9859, + "step": 6896 + }, + { + "epoch": 0.47389033942558745, + "grad_norm": 1.070502519607544, + "learning_rate": 3.1565993233089654e-05, + "loss": 0.9635, + "step": 6897 + }, + { + "epoch": 0.47395904905867803, + "grad_norm": 1.0715309381484985, + "learning_rate": 3.156020723001766e-05, + "loss": 0.9517, + "step": 6898 + }, + { + "epoch": 0.4740277586917686, + "grad_norm": 1.1845786571502686, + "learning_rate": 3.155442084954094e-05, + "loss": 0.9725, + "step": 6899 + }, + { + "epoch": 0.47409646832485913, + "grad_norm": 1.060873031616211, + "learning_rate": 3.154863409199239e-05, + "loss": 0.9201, + "step": 6900 + }, + { + "epoch": 0.4741651779579497, + "grad_norm": 1.049660563468933, + "learning_rate": 3.1542846957704915e-05, + "loss": 0.9307, + "step": 6901 + }, + { + "epoch": 0.47423388759104024, + "grad_norm": 1.0278011560440063, + "learning_rate": 3.153705944701144e-05, + "loss": 1.0008, + "step": 6902 + }, + { + "epoch": 0.4743025972241308, + "grad_norm": 1.0652233362197876, + "learning_rate": 3.1531271560244926e-05, + "loss": 0.9522, + "step": 6903 + }, + { + "epoch": 0.4743713068572214, + "grad_norm": 1.1312768459320068, + "learning_rate": 3.152548329773834e-05, + "loss": 0.9901, + "step": 6904 + }, + { + "epoch": 0.4744400164903119, + "grad_norm": 1.0664856433868408, + "learning_rate": 3.151969465982467e-05, + "loss": 0.9406, + "step": 6905 + }, + { + "epoch": 0.4745087261234025, + "grad_norm": 1.0250213146209717, + "learning_rate": 3.151390564683695e-05, + "loss": 0.9882, + "step": 6906 + }, + { + "epoch": 0.4745774357564931, + "grad_norm": 1.051744818687439, + "learning_rate": 3.150811625910821e-05, + "loss": 0.9657, + "step": 6907 + }, + { + "epoch": 0.4746461453895836, + "grad_norm": 1.0181074142456055, + "learning_rate": 3.150232649697151e-05, + "loss": 0.9358, + "step": 6908 + }, + { + "epoch": 0.4747148550226742, + "grad_norm": 1.0507185459136963, + "learning_rate": 3.149653636075993e-05, + "loss": 0.9999, + "step": 6909 + }, + { + "epoch": 0.4747835646557647, + "grad_norm": 1.0608599185943604, + "learning_rate": 3.149074585080658e-05, + "loss": 0.9359, + "step": 6910 + }, + { + "epoch": 0.4748522742888553, + "grad_norm": 1.0900009870529175, + "learning_rate": 3.1484954967444564e-05, + "loss": 0.9832, + "step": 6911 + }, + { + "epoch": 0.47492098392194587, + "grad_norm": 1.0607144832611084, + "learning_rate": 3.147916371100705e-05, + "loss": 0.9586, + "step": 6912 + }, + { + "epoch": 0.4749896935550364, + "grad_norm": 1.073495626449585, + "learning_rate": 3.1473372081827204e-05, + "loss": 1.0171, + "step": 6913 + }, + { + "epoch": 0.475058403188127, + "grad_norm": 1.1190077066421509, + "learning_rate": 3.146758008023821e-05, + "loss": 0.9785, + "step": 6914 + }, + { + "epoch": 0.47512711282121756, + "grad_norm": 1.1067053079605103, + "learning_rate": 3.1461787706573264e-05, + "loss": 0.9521, + "step": 6915 + }, + { + "epoch": 0.4751958224543081, + "grad_norm": 1.14242684841156, + "learning_rate": 3.145599496116562e-05, + "loss": 1.005, + "step": 6916 + }, + { + "epoch": 0.47526453208739866, + "grad_norm": 1.085811734199524, + "learning_rate": 3.1450201844348526e-05, + "loss": 1.0202, + "step": 6917 + }, + { + "epoch": 0.4753332417204892, + "grad_norm": 1.0784368515014648, + "learning_rate": 3.1444408356455244e-05, + "loss": 0.9847, + "step": 6918 + }, + { + "epoch": 0.47540195135357977, + "grad_norm": 1.061821699142456, + "learning_rate": 3.1438614497819084e-05, + "loss": 0.9905, + "step": 6919 + }, + { + "epoch": 0.47547066098667035, + "grad_norm": 1.0555979013442993, + "learning_rate": 3.143282026877336e-05, + "loss": 0.9793, + "step": 6920 + }, + { + "epoch": 0.47553937061976087, + "grad_norm": 1.1447501182556152, + "learning_rate": 3.1427025669651405e-05, + "loss": 0.9868, + "step": 6921 + }, + { + "epoch": 0.47560808025285145, + "grad_norm": 1.0307408571243286, + "learning_rate": 3.142123070078659e-05, + "loss": 0.9249, + "step": 6922 + }, + { + "epoch": 0.47567678988594203, + "grad_norm": 1.1257050037384033, + "learning_rate": 3.141543536251227e-05, + "loss": 1.0003, + "step": 6923 + }, + { + "epoch": 0.47574549951903256, + "grad_norm": 1.1247177124023438, + "learning_rate": 3.140963965516188e-05, + "loss": 0.9586, + "step": 6924 + }, + { + "epoch": 0.47581420915212314, + "grad_norm": 1.0670456886291504, + "learning_rate": 3.1403843579068826e-05, + "loss": 0.9181, + "step": 6925 + }, + { + "epoch": 0.47588291878521366, + "grad_norm": 1.235668659210205, + "learning_rate": 3.139804713456655e-05, + "loss": 0.9747, + "step": 6926 + }, + { + "epoch": 0.47595162841830424, + "grad_norm": 1.0801067352294922, + "learning_rate": 3.139225032198853e-05, + "loss": 0.976, + "step": 6927 + }, + { + "epoch": 0.4760203380513948, + "grad_norm": 1.151387095451355, + "learning_rate": 3.138645314166824e-05, + "loss": 0.9817, + "step": 6928 + }, + { + "epoch": 0.47608904768448534, + "grad_norm": 1.0172878503799438, + "learning_rate": 3.13806555939392e-05, + "loss": 0.9708, + "step": 6929 + }, + { + "epoch": 0.4761577573175759, + "grad_norm": 0.9939367771148682, + "learning_rate": 3.1374857679134936e-05, + "loss": 0.9608, + "step": 6930 + }, + { + "epoch": 0.4762264669506665, + "grad_norm": 1.043352723121643, + "learning_rate": 3.1369059397589e-05, + "loss": 0.9522, + "step": 6931 + }, + { + "epoch": 0.47629517658375703, + "grad_norm": 1.1050224304199219, + "learning_rate": 3.136326074963494e-05, + "loss": 0.9864, + "step": 6932 + }, + { + "epoch": 0.4763638862168476, + "grad_norm": 1.0680886507034302, + "learning_rate": 3.135746173560639e-05, + "loss": 0.9804, + "step": 6933 + }, + { + "epoch": 0.47643259584993813, + "grad_norm": 1.1012009382247925, + "learning_rate": 3.135166235583693e-05, + "loss": 0.982, + "step": 6934 + }, + { + "epoch": 0.4765013054830287, + "grad_norm": 1.0166845321655273, + "learning_rate": 3.134586261066022e-05, + "loss": 0.9452, + "step": 6935 + }, + { + "epoch": 0.4765700151161193, + "grad_norm": 1.0791476964950562, + "learning_rate": 3.134006250040989e-05, + "loss": 0.976, + "step": 6936 + }, + { + "epoch": 0.4766387247492098, + "grad_norm": 1.115661382675171, + "learning_rate": 3.133426202541964e-05, + "loss": 1.0281, + "step": 6937 + }, + { + "epoch": 0.4767074343823004, + "grad_norm": 1.1610840559005737, + "learning_rate": 3.132846118602315e-05, + "loss": 0.9827, + "step": 6938 + }, + { + "epoch": 0.476776144015391, + "grad_norm": 1.0985969305038452, + "learning_rate": 3.132265998255415e-05, + "loss": 0.9717, + "step": 6939 + }, + { + "epoch": 0.4768448536484815, + "grad_norm": 1.1198351383209229, + "learning_rate": 3.1316858415346384e-05, + "loss": 0.9817, + "step": 6940 + }, + { + "epoch": 0.4769135632815721, + "grad_norm": 1.0313310623168945, + "learning_rate": 3.131105648473359e-05, + "loss": 0.9628, + "step": 6941 + }, + { + "epoch": 0.4769822729146626, + "grad_norm": 1.1009756326675415, + "learning_rate": 3.130525419104958e-05, + "loss": 1.0005, + "step": 6942 + }, + { + "epoch": 0.4770509825477532, + "grad_norm": 1.0933960676193237, + "learning_rate": 3.1299451534628135e-05, + "loss": 0.9393, + "step": 6943 + }, + { + "epoch": 0.47711969218084377, + "grad_norm": 1.0726895332336426, + "learning_rate": 3.1293648515803084e-05, + "loss": 0.9589, + "step": 6944 + }, + { + "epoch": 0.4771884018139343, + "grad_norm": 1.1011050939559937, + "learning_rate": 3.1287845134908264e-05, + "loss": 0.9901, + "step": 6945 + }, + { + "epoch": 0.4772571114470249, + "grad_norm": 1.0067453384399414, + "learning_rate": 3.1282041392277555e-05, + "loss": 0.913, + "step": 6946 + }, + { + "epoch": 0.47732582108011545, + "grad_norm": 1.065934181213379, + "learning_rate": 3.1276237288244836e-05, + "loss": 0.9807, + "step": 6947 + }, + { + "epoch": 0.477394530713206, + "grad_norm": 1.0847958326339722, + "learning_rate": 3.1270432823144015e-05, + "loss": 0.9608, + "step": 6948 + }, + { + "epoch": 0.47746324034629656, + "grad_norm": 0.9805562496185303, + "learning_rate": 3.1264627997309006e-05, + "loss": 0.9179, + "step": 6949 + }, + { + "epoch": 0.4775319499793871, + "grad_norm": 1.1159974336624146, + "learning_rate": 3.1258822811073776e-05, + "loss": 0.9425, + "step": 6950 + }, + { + "epoch": 0.47760065961247766, + "grad_norm": 1.0579921007156372, + "learning_rate": 3.125301726477228e-05, + "loss": 0.9761, + "step": 6951 + }, + { + "epoch": 0.47766936924556824, + "grad_norm": 1.0527983903884888, + "learning_rate": 3.1247211358738516e-05, + "loss": 0.9383, + "step": 6952 + }, + { + "epoch": 0.47773807887865877, + "grad_norm": 1.0648223161697388, + "learning_rate": 3.1241405093306496e-05, + "loss": 0.9458, + "step": 6953 + }, + { + "epoch": 0.47780678851174935, + "grad_norm": 1.0432538986206055, + "learning_rate": 3.1235598468810237e-05, + "loss": 0.9788, + "step": 6954 + }, + { + "epoch": 0.4778754981448399, + "grad_norm": 1.1753772497177124, + "learning_rate": 3.1229791485583815e-05, + "loss": 0.9905, + "step": 6955 + }, + { + "epoch": 0.47794420777793045, + "grad_norm": 1.0808818340301514, + "learning_rate": 3.122398414396128e-05, + "loss": 0.9565, + "step": 6956 + }, + { + "epoch": 0.47801291741102103, + "grad_norm": 1.0283722877502441, + "learning_rate": 3.121817644427673e-05, + "loss": 0.9776, + "step": 6957 + }, + { + "epoch": 0.47808162704411156, + "grad_norm": 1.0377682447433472, + "learning_rate": 3.1212368386864286e-05, + "loss": 0.9515, + "step": 6958 + }, + { + "epoch": 0.47815033667720214, + "grad_norm": 1.0523762702941895, + "learning_rate": 3.120655997205807e-05, + "loss": 0.9942, + "step": 6959 + }, + { + "epoch": 0.4782190463102927, + "grad_norm": 1.0380586385726929, + "learning_rate": 3.1200751200192246e-05, + "loss": 0.944, + "step": 6960 + }, + { + "epoch": 0.47828775594338324, + "grad_norm": 1.0602562427520752, + "learning_rate": 3.1194942071600986e-05, + "loss": 0.943, + "step": 6961 + }, + { + "epoch": 0.4783564655764738, + "grad_norm": 1.0996675491333008, + "learning_rate": 3.1189132586618485e-05, + "loss": 0.9518, + "step": 6962 + }, + { + "epoch": 0.4784251752095644, + "grad_norm": 1.1051490306854248, + "learning_rate": 3.118332274557896e-05, + "loss": 0.9701, + "step": 6963 + }, + { + "epoch": 0.4784938848426549, + "grad_norm": 1.10733163356781, + "learning_rate": 3.1177512548816645e-05, + "loss": 0.966, + "step": 6964 + }, + { + "epoch": 0.4785625944757455, + "grad_norm": 1.1344575881958008, + "learning_rate": 3.117170199666581e-05, + "loss": 1.0163, + "step": 6965 + }, + { + "epoch": 0.4786313041088361, + "grad_norm": 1.0768029689788818, + "learning_rate": 3.1165891089460714e-05, + "loss": 0.9535, + "step": 6966 + }, + { + "epoch": 0.4787000137419266, + "grad_norm": 1.0898317098617554, + "learning_rate": 3.116007982753567e-05, + "loss": 0.9804, + "step": 6967 + }, + { + "epoch": 0.4787687233750172, + "grad_norm": 1.067244529724121, + "learning_rate": 3.115426821122498e-05, + "loss": 0.9763, + "step": 6968 + }, + { + "epoch": 0.4788374330081077, + "grad_norm": 1.0663257837295532, + "learning_rate": 3.1148456240862995e-05, + "loss": 0.9457, + "step": 6969 + }, + { + "epoch": 0.4789061426411983, + "grad_norm": 1.1488968133926392, + "learning_rate": 3.114264391678408e-05, + "loss": 0.9836, + "step": 6970 + }, + { + "epoch": 0.4789748522742889, + "grad_norm": 1.0664358139038086, + "learning_rate": 3.1136831239322594e-05, + "loss": 0.9603, + "step": 6971 + }, + { + "epoch": 0.4790435619073794, + "grad_norm": 1.0573756694793701, + "learning_rate": 3.1131018208812954e-05, + "loss": 0.9508, + "step": 6972 + }, + { + "epoch": 0.47911227154047, + "grad_norm": 1.0646573305130005, + "learning_rate": 3.112520482558957e-05, + "loss": 0.995, + "step": 6973 + }, + { + "epoch": 0.47918098117356056, + "grad_norm": 1.0532808303833008, + "learning_rate": 3.111939108998689e-05, + "loss": 0.9576, + "step": 6974 + }, + { + "epoch": 0.4792496908066511, + "grad_norm": 1.0840481519699097, + "learning_rate": 3.111357700233937e-05, + "loss": 0.9372, + "step": 6975 + }, + { + "epoch": 0.47931840043974167, + "grad_norm": 1.101271152496338, + "learning_rate": 3.11077625629815e-05, + "loss": 0.9892, + "step": 6976 + }, + { + "epoch": 0.4793871100728322, + "grad_norm": 1.1709266901016235, + "learning_rate": 3.110194777224777e-05, + "loss": 0.985, + "step": 6977 + }, + { + "epoch": 0.47945581970592277, + "grad_norm": 1.0789943933486938, + "learning_rate": 3.109613263047271e-05, + "loss": 0.9536, + "step": 6978 + }, + { + "epoch": 0.47952452933901335, + "grad_norm": 1.0321195125579834, + "learning_rate": 3.109031713799085e-05, + "loss": 0.935, + "step": 6979 + }, + { + "epoch": 0.4795932389721039, + "grad_norm": 1.0693066120147705, + "learning_rate": 3.108450129513676e-05, + "loss": 0.987, + "step": 6980 + }, + { + "epoch": 0.47966194860519445, + "grad_norm": 1.0362294912338257, + "learning_rate": 3.107868510224502e-05, + "loss": 0.9682, + "step": 6981 + }, + { + "epoch": 0.47973065823828503, + "grad_norm": 1.0563825368881226, + "learning_rate": 3.107286855965023e-05, + "loss": 0.9696, + "step": 6982 + }, + { + "epoch": 0.47979936787137556, + "grad_norm": 1.064205288887024, + "learning_rate": 3.106705166768702e-05, + "loss": 0.99, + "step": 6983 + }, + { + "epoch": 0.47986807750446614, + "grad_norm": 1.1433203220367432, + "learning_rate": 3.106123442669002e-05, + "loss": 1.0034, + "step": 6984 + }, + { + "epoch": 0.47993678713755666, + "grad_norm": 1.067988634109497, + "learning_rate": 3.1055416836993904e-05, + "loss": 0.9063, + "step": 6985 + }, + { + "epoch": 0.48000549677064724, + "grad_norm": 1.1109745502471924, + "learning_rate": 3.104959889893335e-05, + "loss": 0.9431, + "step": 6986 + }, + { + "epoch": 0.4800742064037378, + "grad_norm": 1.088895320892334, + "learning_rate": 3.1043780612843057e-05, + "loss": 0.9919, + "step": 6987 + }, + { + "epoch": 0.48014291603682835, + "grad_norm": 1.1167415380477905, + "learning_rate": 3.1037961979057753e-05, + "loss": 0.985, + "step": 6988 + }, + { + "epoch": 0.48021162566991893, + "grad_norm": 1.0805920362472534, + "learning_rate": 3.103214299791218e-05, + "loss": 0.9649, + "step": 6989 + }, + { + "epoch": 0.4802803353030095, + "grad_norm": 1.1070891618728638, + "learning_rate": 3.10263236697411e-05, + "loss": 0.9654, + "step": 6990 + }, + { + "epoch": 0.48034904493610003, + "grad_norm": 1.0639833211898804, + "learning_rate": 3.102050399487928e-05, + "loss": 0.9628, + "step": 6991 + }, + { + "epoch": 0.4804177545691906, + "grad_norm": 1.037883996963501, + "learning_rate": 3.1014683973661554e-05, + "loss": 0.9598, + "step": 6992 + }, + { + "epoch": 0.48048646420228114, + "grad_norm": 1.0470589399337769, + "learning_rate": 3.1008863606422714e-05, + "loss": 0.9879, + "step": 6993 + }, + { + "epoch": 0.4805551738353717, + "grad_norm": 1.0365351438522339, + "learning_rate": 3.100304289349762e-05, + "loss": 1.0117, + "step": 6994 + }, + { + "epoch": 0.4806238834684623, + "grad_norm": 1.1083648204803467, + "learning_rate": 3.0997221835221126e-05, + "loss": 0.9918, + "step": 6995 + }, + { + "epoch": 0.4806925931015528, + "grad_norm": 1.0617012977600098, + "learning_rate": 3.099140043192812e-05, + "loss": 0.9496, + "step": 6996 + }, + { + "epoch": 0.4807613027346434, + "grad_norm": 1.0677226781845093, + "learning_rate": 3.0985578683953494e-05, + "loss": 0.9598, + "step": 6997 + }, + { + "epoch": 0.480830012367734, + "grad_norm": 1.0439879894256592, + "learning_rate": 3.097975659163218e-05, + "loss": 0.9859, + "step": 6998 + }, + { + "epoch": 0.4808987220008245, + "grad_norm": 1.027805209159851, + "learning_rate": 3.097393415529912e-05, + "loss": 0.9788, + "step": 6999 + }, + { + "epoch": 0.4809674316339151, + "grad_norm": 1.0639952421188354, + "learning_rate": 3.096811137528926e-05, + "loss": 0.9457, + "step": 7000 + }, + { + "epoch": 0.4810361412670056, + "grad_norm": 1.0611807107925415, + "learning_rate": 3.09622882519376e-05, + "loss": 0.9738, + "step": 7001 + }, + { + "epoch": 0.4811048509000962, + "grad_norm": 1.0858782529830933, + "learning_rate": 3.0956464785579124e-05, + "loss": 0.9663, + "step": 7002 + }, + { + "epoch": 0.4811735605331868, + "grad_norm": 1.0683393478393555, + "learning_rate": 3.095064097654887e-05, + "loss": 1.0025, + "step": 7003 + }, + { + "epoch": 0.4812422701662773, + "grad_norm": 1.0534794330596924, + "learning_rate": 3.094481682518187e-05, + "loss": 1.0038, + "step": 7004 + }, + { + "epoch": 0.4813109797993679, + "grad_norm": 1.0287145376205444, + "learning_rate": 3.093899233181317e-05, + "loss": 0.9505, + "step": 7005 + }, + { + "epoch": 0.48137968943245846, + "grad_norm": 1.0371081829071045, + "learning_rate": 3.093316749677788e-05, + "loss": 0.9664, + "step": 7006 + }, + { + "epoch": 0.481448399065549, + "grad_norm": 1.0933619737625122, + "learning_rate": 3.092734232041107e-05, + "loss": 0.9307, + "step": 7007 + }, + { + "epoch": 0.48151710869863956, + "grad_norm": 1.137099027633667, + "learning_rate": 3.092151680304787e-05, + "loss": 0.9311, + "step": 7008 + }, + { + "epoch": 0.4815858183317301, + "grad_norm": 1.0392111539840698, + "learning_rate": 3.0915690945023416e-05, + "loss": 0.9378, + "step": 7009 + }, + { + "epoch": 0.48165452796482067, + "grad_norm": 1.093446135520935, + "learning_rate": 3.0909864746672874e-05, + "loss": 0.9351, + "step": 7010 + }, + { + "epoch": 0.48172323759791125, + "grad_norm": 1.0807859897613525, + "learning_rate": 3.090403820833142e-05, + "loss": 0.9702, + "step": 7011 + }, + { + "epoch": 0.48179194723100177, + "grad_norm": 1.01943838596344, + "learning_rate": 3.089821133033424e-05, + "loss": 0.9595, + "step": 7012 + }, + { + "epoch": 0.48186065686409235, + "grad_norm": 1.0562231540679932, + "learning_rate": 3.089238411301656e-05, + "loss": 0.9798, + "step": 7013 + }, + { + "epoch": 0.48192936649718293, + "grad_norm": 1.1038283109664917, + "learning_rate": 3.088655655671362e-05, + "loss": 1.01, + "step": 7014 + }, + { + "epoch": 0.48199807613027346, + "grad_norm": 1.1074597835540771, + "learning_rate": 3.088072866176067e-05, + "loss": 0.947, + "step": 7015 + }, + { + "epoch": 0.48206678576336404, + "grad_norm": 1.057008981704712, + "learning_rate": 3.0874900428492994e-05, + "loss": 1.0008, + "step": 7016 + }, + { + "epoch": 0.48213549539645456, + "grad_norm": 1.1481953859329224, + "learning_rate": 3.086907185724586e-05, + "loss": 0.9638, + "step": 7017 + }, + { + "epoch": 0.48220420502954514, + "grad_norm": 1.0491546392440796, + "learning_rate": 3.086324294835461e-05, + "loss": 0.9722, + "step": 7018 + }, + { + "epoch": 0.4822729146626357, + "grad_norm": 1.1010950803756714, + "learning_rate": 3.085741370215457e-05, + "loss": 1.0108, + "step": 7019 + }, + { + "epoch": 0.48234162429572625, + "grad_norm": 1.09512460231781, + "learning_rate": 3.085158411898108e-05, + "loss": 0.9475, + "step": 7020 + }, + { + "epoch": 0.4824103339288168, + "grad_norm": 1.0529519319534302, + "learning_rate": 3.084575419916954e-05, + "loss": 0.9382, + "step": 7021 + }, + { + "epoch": 0.4824790435619074, + "grad_norm": 1.0962811708450317, + "learning_rate": 3.083992394305532e-05, + "loss": 0.991, + "step": 7022 + }, + { + "epoch": 0.48254775319499793, + "grad_norm": 1.065176248550415, + "learning_rate": 3.0834093350973825e-05, + "loss": 0.9812, + "step": 7023 + }, + { + "epoch": 0.4826164628280885, + "grad_norm": 1.0261316299438477, + "learning_rate": 3.082826242326051e-05, + "loss": 0.9621, + "step": 7024 + }, + { + "epoch": 0.48268517246117904, + "grad_norm": 1.0982991456985474, + "learning_rate": 3.0822431160250797e-05, + "loss": 0.981, + "step": 7025 + }, + { + "epoch": 0.4827538820942696, + "grad_norm": 1.0491456985473633, + "learning_rate": 3.0816599562280186e-05, + "loss": 0.9636, + "step": 7026 + }, + { + "epoch": 0.4828225917273602, + "grad_norm": 1.0945523977279663, + "learning_rate": 3.081076762968415e-05, + "loss": 0.9609, + "step": 7027 + }, + { + "epoch": 0.4828913013604507, + "grad_norm": 1.0561527013778687, + "learning_rate": 3.080493536279818e-05, + "loss": 0.9465, + "step": 7028 + }, + { + "epoch": 0.4829600109935413, + "grad_norm": 1.026342511177063, + "learning_rate": 3.0799102761957825e-05, + "loss": 0.9816, + "step": 7029 + }, + { + "epoch": 0.4830287206266319, + "grad_norm": 1.0910576581954956, + "learning_rate": 3.079326982749862e-05, + "loss": 0.9996, + "step": 7030 + }, + { + "epoch": 0.4830974302597224, + "grad_norm": 1.152040719985962, + "learning_rate": 3.078743655975614e-05, + "loss": 0.9994, + "step": 7031 + }, + { + "epoch": 0.483166139892813, + "grad_norm": 1.0909795761108398, + "learning_rate": 3.0781602959065966e-05, + "loss": 0.995, + "step": 7032 + }, + { + "epoch": 0.4832348495259035, + "grad_norm": 0.9777387380599976, + "learning_rate": 3.077576902576369e-05, + "loss": 0.9139, + "step": 7033 + }, + { + "epoch": 0.4833035591589941, + "grad_norm": 1.060328722000122, + "learning_rate": 3.0769934760184956e-05, + "loss": 0.9892, + "step": 7034 + }, + { + "epoch": 0.48337226879208467, + "grad_norm": 1.124165654182434, + "learning_rate": 3.076410016266538e-05, + "loss": 0.9862, + "step": 7035 + }, + { + "epoch": 0.4834409784251752, + "grad_norm": 1.0879566669464111, + "learning_rate": 3.075826523354064e-05, + "loss": 0.9606, + "step": 7036 + }, + { + "epoch": 0.4835096880582658, + "grad_norm": 1.0349317789077759, + "learning_rate": 3.0752429973146415e-05, + "loss": 0.9891, + "step": 7037 + }, + { + "epoch": 0.48357839769135635, + "grad_norm": 1.0852564573287964, + "learning_rate": 3.074659438181841e-05, + "loss": 0.9931, + "step": 7038 + }, + { + "epoch": 0.4836471073244469, + "grad_norm": 1.0189584493637085, + "learning_rate": 3.074075845989232e-05, + "loss": 0.9085, + "step": 7039 + }, + { + "epoch": 0.48371581695753746, + "grad_norm": 1.0500168800354004, + "learning_rate": 3.07349222077039e-05, + "loss": 0.9796, + "step": 7040 + }, + { + "epoch": 0.483784526590628, + "grad_norm": 1.1330512762069702, + "learning_rate": 3.07290856255889e-05, + "loss": 0.9811, + "step": 7041 + }, + { + "epoch": 0.48385323622371856, + "grad_norm": 1.0956968069076538, + "learning_rate": 3.0723248713883104e-05, + "loss": 0.9623, + "step": 7042 + }, + { + "epoch": 0.48392194585680914, + "grad_norm": 1.076766848564148, + "learning_rate": 3.0717411472922294e-05, + "loss": 0.9542, + "step": 7043 + }, + { + "epoch": 0.48399065548989967, + "grad_norm": 1.0405075550079346, + "learning_rate": 3.071157390304229e-05, + "loss": 0.9252, + "step": 7044 + }, + { + "epoch": 0.48405936512299025, + "grad_norm": 1.0491349697113037, + "learning_rate": 3.0705736004578916e-05, + "loss": 0.9775, + "step": 7045 + }, + { + "epoch": 0.48412807475608083, + "grad_norm": 1.0944358110427856, + "learning_rate": 3.069989777786803e-05, + "loss": 0.961, + "step": 7046 + }, + { + "epoch": 0.48419678438917135, + "grad_norm": 1.1267101764678955, + "learning_rate": 3.069405922324551e-05, + "loss": 0.982, + "step": 7047 + }, + { + "epoch": 0.48426549402226193, + "grad_norm": 1.0135509967803955, + "learning_rate": 3.068822034104722e-05, + "loss": 0.9501, + "step": 7048 + }, + { + "epoch": 0.48433420365535246, + "grad_norm": 1.133386492729187, + "learning_rate": 3.0682381131609104e-05, + "loss": 1.0242, + "step": 7049 + }, + { + "epoch": 0.48440291328844304, + "grad_norm": 1.0446410179138184, + "learning_rate": 3.067654159526704e-05, + "loss": 0.9285, + "step": 7050 + }, + { + "epoch": 0.4844716229215336, + "grad_norm": 1.140494704246521, + "learning_rate": 3.067070173235702e-05, + "loss": 0.9659, + "step": 7051 + }, + { + "epoch": 0.48454033255462414, + "grad_norm": 1.1884857416152954, + "learning_rate": 3.066486154321499e-05, + "loss": 1.0025, + "step": 7052 + }, + { + "epoch": 0.4846090421877147, + "grad_norm": 1.1447515487670898, + "learning_rate": 3.0659021028176907e-05, + "loss": 0.9707, + "step": 7053 + }, + { + "epoch": 0.4846777518208053, + "grad_norm": 1.048417568206787, + "learning_rate": 3.065318018757881e-05, + "loss": 0.9462, + "step": 7054 + }, + { + "epoch": 0.4847464614538958, + "grad_norm": 1.2272781133651733, + "learning_rate": 3.06473390217567e-05, + "loss": 0.9726, + "step": 7055 + }, + { + "epoch": 0.4848151710869864, + "grad_norm": 1.0775541067123413, + "learning_rate": 3.0641497531046615e-05, + "loss": 0.941, + "step": 7056 + }, + { + "epoch": 0.48488388072007693, + "grad_norm": 1.085113763809204, + "learning_rate": 3.063565571578462e-05, + "loss": 0.9334, + "step": 7057 + }, + { + "epoch": 0.4849525903531675, + "grad_norm": 1.1262723207473755, + "learning_rate": 3.0629813576306796e-05, + "loss": 0.9907, + "step": 7058 + }, + { + "epoch": 0.4850212999862581, + "grad_norm": 1.0730607509613037, + "learning_rate": 3.062397111294921e-05, + "loss": 0.9529, + "step": 7059 + }, + { + "epoch": 0.4850900096193486, + "grad_norm": 1.0306133031845093, + "learning_rate": 3.061812832604802e-05, + "loss": 0.9219, + "step": 7060 + }, + { + "epoch": 0.4851587192524392, + "grad_norm": 1.105135202407837, + "learning_rate": 3.061228521593931e-05, + "loss": 1.0045, + "step": 7061 + }, + { + "epoch": 0.4852274288855298, + "grad_norm": 1.0544326305389404, + "learning_rate": 3.060644178295927e-05, + "loss": 0.978, + "step": 7062 + }, + { + "epoch": 0.4852961385186203, + "grad_norm": 1.0661650896072388, + "learning_rate": 3.060059802744405e-05, + "loss": 0.9569, + "step": 7063 + }, + { + "epoch": 0.4853648481517109, + "grad_norm": 1.0365160703659058, + "learning_rate": 3.0594753949729836e-05, + "loss": 0.9847, + "step": 7064 + }, + { + "epoch": 0.4854335577848014, + "grad_norm": 1.0128273963928223, + "learning_rate": 3.0588909550152845e-05, + "loss": 0.9358, + "step": 7065 + }, + { + "epoch": 0.485502267417892, + "grad_norm": 1.0805107355117798, + "learning_rate": 3.058306482904928e-05, + "loss": 0.9792, + "step": 7066 + }, + { + "epoch": 0.48557097705098257, + "grad_norm": 1.085618019104004, + "learning_rate": 3.057721978675542e-05, + "loss": 0.9921, + "step": 7067 + }, + { + "epoch": 0.4856396866840731, + "grad_norm": 1.0459215641021729, + "learning_rate": 3.057137442360749e-05, + "loss": 0.9368, + "step": 7068 + }, + { + "epoch": 0.48570839631716367, + "grad_norm": 1.0661747455596924, + "learning_rate": 3.056552873994179e-05, + "loss": 0.9909, + "step": 7069 + }, + { + "epoch": 0.48577710595025425, + "grad_norm": 1.0669937133789062, + "learning_rate": 3.0559682736094625e-05, + "loss": 0.967, + "step": 7070 + }, + { + "epoch": 0.4858458155833448, + "grad_norm": 1.0197681188583374, + "learning_rate": 3.0553836412402294e-05, + "loss": 0.9553, + "step": 7071 + }, + { + "epoch": 0.48591452521643536, + "grad_norm": 1.0705618858337402, + "learning_rate": 3.054798976920114e-05, + "loss": 0.9736, + "step": 7072 + }, + { + "epoch": 0.4859832348495259, + "grad_norm": 1.0477852821350098, + "learning_rate": 3.054214280682752e-05, + "loss": 1.0254, + "step": 7073 + }, + { + "epoch": 0.48605194448261646, + "grad_norm": 1.119655966758728, + "learning_rate": 3.0536295525617805e-05, + "loss": 0.9851, + "step": 7074 + }, + { + "epoch": 0.48612065411570704, + "grad_norm": 1.1039361953735352, + "learning_rate": 3.053044792590838e-05, + "loss": 0.9791, + "step": 7075 + }, + { + "epoch": 0.48618936374879757, + "grad_norm": 1.0359349250793457, + "learning_rate": 3.052460000803566e-05, + "loss": 0.9901, + "step": 7076 + }, + { + "epoch": 0.48625807338188815, + "grad_norm": 1.0426154136657715, + "learning_rate": 3.051875177233607e-05, + "loss": 0.9584, + "step": 7077 + }, + { + "epoch": 0.4863267830149787, + "grad_norm": 1.09795081615448, + "learning_rate": 3.051290321914605e-05, + "loss": 0.9593, + "step": 7078 + }, + { + "epoch": 0.48639549264806925, + "grad_norm": 1.121289610862732, + "learning_rate": 3.050705434880207e-05, + "loss": 0.9648, + "step": 7079 + }, + { + "epoch": 0.48646420228115983, + "grad_norm": 0.9670535326004028, + "learning_rate": 3.050120516164062e-05, + "loss": 0.9288, + "step": 7080 + }, + { + "epoch": 0.48653291191425035, + "grad_norm": 1.0191830396652222, + "learning_rate": 3.049535565799819e-05, + "loss": 0.9537, + "step": 7081 + }, + { + "epoch": 0.48660162154734093, + "grad_norm": 1.1372934579849243, + "learning_rate": 3.0489505838211296e-05, + "loss": 0.9985, + "step": 7082 + }, + { + "epoch": 0.4866703311804315, + "grad_norm": 1.0879929065704346, + "learning_rate": 3.0483655702616477e-05, + "loss": 0.9987, + "step": 7083 + }, + { + "epoch": 0.48673904081352204, + "grad_norm": 1.039386510848999, + "learning_rate": 3.0477805251550285e-05, + "loss": 0.9528, + "step": 7084 + }, + { + "epoch": 0.4868077504466126, + "grad_norm": 1.099795937538147, + "learning_rate": 3.047195448534931e-05, + "loss": 0.9256, + "step": 7085 + }, + { + "epoch": 0.4868764600797032, + "grad_norm": 1.0959020853042603, + "learning_rate": 3.046610340435012e-05, + "loss": 1.0024, + "step": 7086 + }, + { + "epoch": 0.4869451697127937, + "grad_norm": 1.1638904809951782, + "learning_rate": 3.0460252008889335e-05, + "loss": 0.9968, + "step": 7087 + }, + { + "epoch": 0.4870138793458843, + "grad_norm": 1.1334702968597412, + "learning_rate": 3.0454400299303582e-05, + "loss": 1.0001, + "step": 7088 + }, + { + "epoch": 0.48708258897897483, + "grad_norm": 1.1000741720199585, + "learning_rate": 3.0448548275929507e-05, + "loss": 0.9674, + "step": 7089 + }, + { + "epoch": 0.4871512986120654, + "grad_norm": 1.0872515439987183, + "learning_rate": 3.044269593910377e-05, + "loss": 0.9682, + "step": 7090 + }, + { + "epoch": 0.487220008245156, + "grad_norm": 1.080994725227356, + "learning_rate": 3.0436843289163052e-05, + "loss": 0.9645, + "step": 7091 + }, + { + "epoch": 0.4872887178782465, + "grad_norm": 1.0549638271331787, + "learning_rate": 3.0430990326444058e-05, + "loss": 0.9869, + "step": 7092 + }, + { + "epoch": 0.4873574275113371, + "grad_norm": 1.0724300146102905, + "learning_rate": 3.0425137051283504e-05, + "loss": 0.9617, + "step": 7093 + }, + { + "epoch": 0.4874261371444277, + "grad_norm": 1.1487228870391846, + "learning_rate": 3.041928346401812e-05, + "loss": 0.9393, + "step": 7094 + }, + { + "epoch": 0.4874948467775182, + "grad_norm": 1.0658326148986816, + "learning_rate": 3.0413429564984662e-05, + "loss": 0.9446, + "step": 7095 + }, + { + "epoch": 0.4875635564106088, + "grad_norm": 1.0636787414550781, + "learning_rate": 3.0407575354519902e-05, + "loss": 0.9375, + "step": 7096 + }, + { + "epoch": 0.4876322660436993, + "grad_norm": 1.100066065788269, + "learning_rate": 3.0401720832960635e-05, + "loss": 0.9764, + "step": 7097 + }, + { + "epoch": 0.4877009756767899, + "grad_norm": 1.164029836654663, + "learning_rate": 3.039586600064366e-05, + "loss": 1.006, + "step": 7098 + }, + { + "epoch": 0.48776968530988046, + "grad_norm": 1.0850636959075928, + "learning_rate": 3.0390010857905797e-05, + "loss": 0.9903, + "step": 7099 + }, + { + "epoch": 0.487838394942971, + "grad_norm": 1.1491996049880981, + "learning_rate": 3.03841554050839e-05, + "loss": 0.975, + "step": 7100 + }, + { + "epoch": 0.48790710457606157, + "grad_norm": 1.145714521408081, + "learning_rate": 3.037829964251483e-05, + "loss": 0.9829, + "step": 7101 + }, + { + "epoch": 0.48797581420915215, + "grad_norm": 1.0586673021316528, + "learning_rate": 3.0372443570535452e-05, + "loss": 0.9571, + "step": 7102 + }, + { + "epoch": 0.4880445238422427, + "grad_norm": 1.0517127513885498, + "learning_rate": 3.0366587189482683e-05, + "loss": 0.9491, + "step": 7103 + }, + { + "epoch": 0.48811323347533325, + "grad_norm": 1.0822972059249878, + "learning_rate": 3.0360730499693414e-05, + "loss": 0.9721, + "step": 7104 + }, + { + "epoch": 0.4881819431084238, + "grad_norm": 1.0888279676437378, + "learning_rate": 3.0354873501504595e-05, + "loss": 0.9658, + "step": 7105 + }, + { + "epoch": 0.48825065274151436, + "grad_norm": 1.0408871173858643, + "learning_rate": 3.034901619525316e-05, + "loss": 0.9678, + "step": 7106 + }, + { + "epoch": 0.48831936237460494, + "grad_norm": 1.0863457918167114, + "learning_rate": 3.034315858127609e-05, + "loss": 0.9796, + "step": 7107 + }, + { + "epoch": 0.48838807200769546, + "grad_norm": 1.0100480318069458, + "learning_rate": 3.0337300659910367e-05, + "loss": 0.952, + "step": 7108 + }, + { + "epoch": 0.48845678164078604, + "grad_norm": 1.0750024318695068, + "learning_rate": 3.033144243149299e-05, + "loss": 0.9277, + "step": 7109 + }, + { + "epoch": 0.4885254912738766, + "grad_norm": 1.1258959770202637, + "learning_rate": 3.0325583896360975e-05, + "loss": 0.954, + "step": 7110 + }, + { + "epoch": 0.48859420090696715, + "grad_norm": 1.1274597644805908, + "learning_rate": 3.031972505485138e-05, + "loss": 0.946, + "step": 7111 + }, + { + "epoch": 0.4886629105400577, + "grad_norm": 1.078186273574829, + "learning_rate": 3.031386590730123e-05, + "loss": 0.9713, + "step": 7112 + }, + { + "epoch": 0.48873162017314825, + "grad_norm": 1.1285840272903442, + "learning_rate": 3.030800645404763e-05, + "loss": 0.9482, + "step": 7113 + }, + { + "epoch": 0.48880032980623883, + "grad_norm": 1.0307674407958984, + "learning_rate": 3.030214669542764e-05, + "loss": 0.9931, + "step": 7114 + }, + { + "epoch": 0.4888690394393294, + "grad_norm": 1.1208925247192383, + "learning_rate": 3.0296286631778387e-05, + "loss": 0.9524, + "step": 7115 + }, + { + "epoch": 0.48893774907241994, + "grad_norm": 1.0499573945999146, + "learning_rate": 3.0290426263437e-05, + "loss": 0.953, + "step": 7116 + }, + { + "epoch": 0.4890064587055105, + "grad_norm": 1.0836200714111328, + "learning_rate": 3.028456559074061e-05, + "loss": 0.9752, + "step": 7117 + }, + { + "epoch": 0.4890751683386011, + "grad_norm": 1.0532556772232056, + "learning_rate": 3.0278704614026387e-05, + "loss": 1.0006, + "step": 7118 + }, + { + "epoch": 0.4891438779716916, + "grad_norm": 1.0427420139312744, + "learning_rate": 3.0272843333631502e-05, + "loss": 0.9329, + "step": 7119 + }, + { + "epoch": 0.4892125876047822, + "grad_norm": 1.107961654663086, + "learning_rate": 3.0266981749893157e-05, + "loss": 0.9901, + "step": 7120 + }, + { + "epoch": 0.4892812972378727, + "grad_norm": 1.1490191221237183, + "learning_rate": 3.0261119863148568e-05, + "loss": 1.014, + "step": 7121 + }, + { + "epoch": 0.4893500068709633, + "grad_norm": 1.062988042831421, + "learning_rate": 3.025525767373496e-05, + "loss": 0.9901, + "step": 7122 + }, + { + "epoch": 0.4894187165040539, + "grad_norm": 1.0675058364868164, + "learning_rate": 3.024939518198958e-05, + "loss": 1.0134, + "step": 7123 + }, + { + "epoch": 0.4894874261371444, + "grad_norm": 1.0301798582077026, + "learning_rate": 3.0243532388249695e-05, + "loss": 0.9718, + "step": 7124 + }, + { + "epoch": 0.489556135770235, + "grad_norm": 1.0776267051696777, + "learning_rate": 3.0237669292852587e-05, + "loss": 0.9437, + "step": 7125 + }, + { + "epoch": 0.48962484540332557, + "grad_norm": 1.0972278118133545, + "learning_rate": 3.0231805896135564e-05, + "loss": 0.9902, + "step": 7126 + }, + { + "epoch": 0.4896935550364161, + "grad_norm": 1.0771214962005615, + "learning_rate": 3.022594219843593e-05, + "loss": 0.9675, + "step": 7127 + }, + { + "epoch": 0.4897622646695067, + "grad_norm": 0.9923551082611084, + "learning_rate": 3.022007820009104e-05, + "loss": 0.9146, + "step": 7128 + }, + { + "epoch": 0.4898309743025972, + "grad_norm": 1.0783509016036987, + "learning_rate": 3.021421390143822e-05, + "loss": 0.9756, + "step": 7129 + }, + { + "epoch": 0.4898996839356878, + "grad_norm": 1.0690889358520508, + "learning_rate": 3.020834930281486e-05, + "loss": 0.9893, + "step": 7130 + }, + { + "epoch": 0.48996839356877836, + "grad_norm": 1.1088862419128418, + "learning_rate": 3.020248440455834e-05, + "loss": 0.9605, + "step": 7131 + }, + { + "epoch": 0.4900371032018689, + "grad_norm": 1.0766079425811768, + "learning_rate": 3.019661920700606e-05, + "loss": 0.9916, + "step": 7132 + }, + { + "epoch": 0.49010581283495946, + "grad_norm": 1.1048054695129395, + "learning_rate": 3.0190753710495455e-05, + "loss": 1.0062, + "step": 7133 + }, + { + "epoch": 0.49017452246805004, + "grad_norm": 1.0188488960266113, + "learning_rate": 3.0184887915363953e-05, + "loss": 0.9547, + "step": 7134 + }, + { + "epoch": 0.49024323210114057, + "grad_norm": 1.0643341541290283, + "learning_rate": 3.0179021821949006e-05, + "loss": 0.9549, + "step": 7135 + }, + { + "epoch": 0.49031194173423115, + "grad_norm": 1.050413966178894, + "learning_rate": 3.0173155430588096e-05, + "loss": 0.9679, + "step": 7136 + }, + { + "epoch": 0.4903806513673217, + "grad_norm": 1.0662033557891846, + "learning_rate": 3.01672887416187e-05, + "loss": 0.969, + "step": 7137 + }, + { + "epoch": 0.49044936100041225, + "grad_norm": 1.0812968015670776, + "learning_rate": 3.0161421755378337e-05, + "loss": 0.9686, + "step": 7138 + }, + { + "epoch": 0.49051807063350283, + "grad_norm": 1.0742363929748535, + "learning_rate": 3.0155554472204528e-05, + "loss": 0.9774, + "step": 7139 + }, + { + "epoch": 0.49058678026659336, + "grad_norm": 1.1026798486709595, + "learning_rate": 3.0149686892434813e-05, + "loss": 1.015, + "step": 7140 + }, + { + "epoch": 0.49065548989968394, + "grad_norm": 1.0892374515533447, + "learning_rate": 3.0143819016406753e-05, + "loss": 0.9699, + "step": 7141 + }, + { + "epoch": 0.4907241995327745, + "grad_norm": 1.086172103881836, + "learning_rate": 3.0137950844457918e-05, + "loss": 0.9679, + "step": 7142 + }, + { + "epoch": 0.49079290916586504, + "grad_norm": 1.0804412364959717, + "learning_rate": 3.0132082376925907e-05, + "loss": 1.0089, + "step": 7143 + }, + { + "epoch": 0.4908616187989556, + "grad_norm": 1.0198841094970703, + "learning_rate": 3.0126213614148325e-05, + "loss": 0.9647, + "step": 7144 + }, + { + "epoch": 0.49093032843204615, + "grad_norm": 1.1164727210998535, + "learning_rate": 3.0120344556462805e-05, + "loss": 0.9721, + "step": 7145 + }, + { + "epoch": 0.49099903806513673, + "grad_norm": 1.1040637493133545, + "learning_rate": 3.011447520420698e-05, + "loss": 0.9686, + "step": 7146 + }, + { + "epoch": 0.4910677476982273, + "grad_norm": 1.0725185871124268, + "learning_rate": 3.0108605557718512e-05, + "loss": 0.9876, + "step": 7147 + }, + { + "epoch": 0.49113645733131783, + "grad_norm": 1.0851709842681885, + "learning_rate": 3.0102735617335083e-05, + "loss": 0.9796, + "step": 7148 + }, + { + "epoch": 0.4912051669644084, + "grad_norm": 1.0499250888824463, + "learning_rate": 3.009686538339439e-05, + "loss": 0.9746, + "step": 7149 + }, + { + "epoch": 0.491273876597499, + "grad_norm": 1.0791391134262085, + "learning_rate": 3.0090994856234128e-05, + "loss": 0.964, + "step": 7150 + }, + { + "epoch": 0.4913425862305895, + "grad_norm": 1.0247498750686646, + "learning_rate": 3.0085124036192047e-05, + "loss": 0.9682, + "step": 7151 + }, + { + "epoch": 0.4914112958636801, + "grad_norm": 1.1282098293304443, + "learning_rate": 3.007925292360587e-05, + "loss": 0.9539, + "step": 7152 + }, + { + "epoch": 0.4914800054967706, + "grad_norm": 1.0944180488586426, + "learning_rate": 3.0073381518813376e-05, + "loss": 0.968, + "step": 7153 + }, + { + "epoch": 0.4915487151298612, + "grad_norm": 1.0444988012313843, + "learning_rate": 3.006750982215234e-05, + "loss": 1.0038, + "step": 7154 + }, + { + "epoch": 0.4916174247629518, + "grad_norm": 1.0384210348129272, + "learning_rate": 3.0061637833960547e-05, + "loss": 0.9447, + "step": 7155 + }, + { + "epoch": 0.4916861343960423, + "grad_norm": 1.0536249876022339, + "learning_rate": 3.0055765554575827e-05, + "loss": 0.9375, + "step": 7156 + }, + { + "epoch": 0.4917548440291329, + "grad_norm": 1.1377421617507935, + "learning_rate": 3.0049892984335992e-05, + "loss": 1.0108, + "step": 7157 + }, + { + "epoch": 0.49182355366222347, + "grad_norm": 1.103411316871643, + "learning_rate": 3.0044020123578885e-05, + "loss": 0.9716, + "step": 7158 + }, + { + "epoch": 0.491892263295314, + "grad_norm": 1.0698812007904053, + "learning_rate": 3.0038146972642382e-05, + "loss": 0.9531, + "step": 7159 + }, + { + "epoch": 0.49196097292840457, + "grad_norm": 1.0567713975906372, + "learning_rate": 3.003227353186435e-05, + "loss": 0.9584, + "step": 7160 + }, + { + "epoch": 0.4920296825614951, + "grad_norm": 1.0514847040176392, + "learning_rate": 3.00263998015827e-05, + "loss": 0.9353, + "step": 7161 + }, + { + "epoch": 0.4920983921945857, + "grad_norm": 1.0179085731506348, + "learning_rate": 3.002052578213533e-05, + "loss": 0.9794, + "step": 7162 + }, + { + "epoch": 0.49216710182767626, + "grad_norm": 1.0867615938186646, + "learning_rate": 3.0014651473860167e-05, + "loss": 0.99, + "step": 7163 + }, + { + "epoch": 0.4922358114607668, + "grad_norm": 1.0014809370040894, + "learning_rate": 3.0008776877095167e-05, + "loss": 0.9878, + "step": 7164 + }, + { + "epoch": 0.49230452109385736, + "grad_norm": 1.1365934610366821, + "learning_rate": 3.000290199217829e-05, + "loss": 0.9645, + "step": 7165 + }, + { + "epoch": 0.49237323072694794, + "grad_norm": 1.0920748710632324, + "learning_rate": 2.9997026819447505e-05, + "loss": 0.9608, + "step": 7166 + }, + { + "epoch": 0.49244194036003847, + "grad_norm": 1.0991820096969604, + "learning_rate": 2.999115135924082e-05, + "loss": 0.9935, + "step": 7167 + }, + { + "epoch": 0.49251064999312905, + "grad_norm": 1.0828454494476318, + "learning_rate": 2.998527561189624e-05, + "loss": 0.9992, + "step": 7168 + }, + { + "epoch": 0.49257935962621957, + "grad_norm": 1.0568950176239014, + "learning_rate": 2.997939957775179e-05, + "loss": 0.947, + "step": 7169 + }, + { + "epoch": 0.49264806925931015, + "grad_norm": 1.0627243518829346, + "learning_rate": 2.9973523257145525e-05, + "loss": 0.9476, + "step": 7170 + }, + { + "epoch": 0.49271677889240073, + "grad_norm": 1.0908358097076416, + "learning_rate": 2.9967646650415492e-05, + "loss": 1.0048, + "step": 7171 + }, + { + "epoch": 0.49278548852549126, + "grad_norm": 1.0224741697311401, + "learning_rate": 2.9961769757899782e-05, + "loss": 0.9467, + "step": 7172 + }, + { + "epoch": 0.49285419815858184, + "grad_norm": 1.0108363628387451, + "learning_rate": 2.995589257993648e-05, + "loss": 0.9459, + "step": 7173 + }, + { + "epoch": 0.4929229077916724, + "grad_norm": 1.0579197406768799, + "learning_rate": 2.99500151168637e-05, + "loss": 0.9432, + "step": 7174 + }, + { + "epoch": 0.49299161742476294, + "grad_norm": 1.0480115413665771, + "learning_rate": 2.9944137369019564e-05, + "loss": 0.9546, + "step": 7175 + }, + { + "epoch": 0.4930603270578535, + "grad_norm": 1.0560007095336914, + "learning_rate": 2.9938259336742226e-05, + "loss": 0.9641, + "step": 7176 + }, + { + "epoch": 0.49312903669094404, + "grad_norm": 1.0161799192428589, + "learning_rate": 2.9932381020369838e-05, + "loss": 0.9577, + "step": 7177 + }, + { + "epoch": 0.4931977463240346, + "grad_norm": 1.0069198608398438, + "learning_rate": 2.9926502420240575e-05, + "loss": 0.9373, + "step": 7178 + }, + { + "epoch": 0.4932664559571252, + "grad_norm": 1.0741323232650757, + "learning_rate": 2.9920623536692638e-05, + "loss": 0.9927, + "step": 7179 + }, + { + "epoch": 0.49333516559021573, + "grad_norm": 1.0708730220794678, + "learning_rate": 2.9914744370064223e-05, + "loss": 0.9753, + "step": 7180 + }, + { + "epoch": 0.4934038752233063, + "grad_norm": 1.071999430656433, + "learning_rate": 2.9908864920693564e-05, + "loss": 0.9348, + "step": 7181 + }, + { + "epoch": 0.4934725848563969, + "grad_norm": 1.0914338827133179, + "learning_rate": 2.9902985188918902e-05, + "loss": 0.9641, + "step": 7182 + }, + { + "epoch": 0.4935412944894874, + "grad_norm": 0.9872948527336121, + "learning_rate": 2.9897105175078487e-05, + "loss": 0.9289, + "step": 7183 + }, + { + "epoch": 0.493610004122578, + "grad_norm": 1.0685169696807861, + "learning_rate": 2.9891224879510598e-05, + "loss": 0.9877, + "step": 7184 + }, + { + "epoch": 0.4936787137556685, + "grad_norm": 1.0393192768096924, + "learning_rate": 2.9885344302553525e-05, + "loss": 0.9684, + "step": 7185 + }, + { + "epoch": 0.4937474233887591, + "grad_norm": 1.0059562921524048, + "learning_rate": 2.987946344454557e-05, + "loss": 0.9687, + "step": 7186 + }, + { + "epoch": 0.4938161330218497, + "grad_norm": 1.1193290948867798, + "learning_rate": 2.987358230582506e-05, + "loss": 1.0074, + "step": 7187 + }, + { + "epoch": 0.4938848426549402, + "grad_norm": 1.0441182851791382, + "learning_rate": 2.986770088673033e-05, + "loss": 0.9713, + "step": 7188 + }, + { + "epoch": 0.4939535522880308, + "grad_norm": 1.1001067161560059, + "learning_rate": 2.9861819187599744e-05, + "loss": 0.9536, + "step": 7189 + }, + { + "epoch": 0.49402226192112136, + "grad_norm": 1.140536904335022, + "learning_rate": 2.9855937208771657e-05, + "loss": 0.986, + "step": 7190 + }, + { + "epoch": 0.4940909715542119, + "grad_norm": 1.0392941236495972, + "learning_rate": 2.985005495058446e-05, + "loss": 0.9822, + "step": 7191 + }, + { + "epoch": 0.49415968118730247, + "grad_norm": 1.1161117553710938, + "learning_rate": 2.9844172413376566e-05, + "loss": 0.9899, + "step": 7192 + }, + { + "epoch": 0.494228390820393, + "grad_norm": 1.0920461416244507, + "learning_rate": 2.9838289597486384e-05, + "loss": 0.9849, + "step": 7193 + }, + { + "epoch": 0.4942971004534836, + "grad_norm": 1.0664819478988647, + "learning_rate": 2.9832406503252353e-05, + "loss": 0.9575, + "step": 7194 + }, + { + "epoch": 0.49436581008657415, + "grad_norm": 1.010867953300476, + "learning_rate": 2.9826523131012923e-05, + "loss": 0.9454, + "step": 7195 + }, + { + "epoch": 0.4944345197196647, + "grad_norm": 1.1016920804977417, + "learning_rate": 2.9820639481106556e-05, + "loss": 0.9879, + "step": 7196 + }, + { + "epoch": 0.49450322935275526, + "grad_norm": 1.0399259328842163, + "learning_rate": 2.9814755553871742e-05, + "loss": 0.9625, + "step": 7197 + }, + { + "epoch": 0.49457193898584584, + "grad_norm": 1.0955535173416138, + "learning_rate": 2.9808871349646973e-05, + "loss": 1.0064, + "step": 7198 + }, + { + "epoch": 0.49464064861893636, + "grad_norm": 1.0689445734024048, + "learning_rate": 2.980298686877077e-05, + "loss": 1.0032, + "step": 7199 + }, + { + "epoch": 0.49470935825202694, + "grad_norm": 1.009669542312622, + "learning_rate": 2.9797102111581665e-05, + "loss": 0.931, + "step": 7200 + }, + { + "epoch": 0.49477806788511747, + "grad_norm": 1.0733777284622192, + "learning_rate": 2.9791217078418192e-05, + "loss": 0.9685, + "step": 7201 + }, + { + "epoch": 0.49484677751820805, + "grad_norm": 1.0643854141235352, + "learning_rate": 2.978533176961893e-05, + "loss": 0.9704, + "step": 7202 + }, + { + "epoch": 0.4949154871512986, + "grad_norm": 1.0690712928771973, + "learning_rate": 2.9779446185522442e-05, + "loss": 0.9841, + "step": 7203 + }, + { + "epoch": 0.49498419678438915, + "grad_norm": 1.0078834295272827, + "learning_rate": 2.9773560326467338e-05, + "loss": 0.9496, + "step": 7204 + }, + { + "epoch": 0.49505290641747973, + "grad_norm": 1.1249891519546509, + "learning_rate": 2.9767674192792216e-05, + "loss": 0.9959, + "step": 7205 + }, + { + "epoch": 0.4951216160505703, + "grad_norm": 1.0379009246826172, + "learning_rate": 2.97617877848357e-05, + "loss": 1.0314, + "step": 7206 + }, + { + "epoch": 0.49519032568366084, + "grad_norm": 1.01010262966156, + "learning_rate": 2.975590110293644e-05, + "loss": 0.945, + "step": 7207 + }, + { + "epoch": 0.4952590353167514, + "grad_norm": 1.071494460105896, + "learning_rate": 2.9750014147433087e-05, + "loss": 0.9848, + "step": 7208 + }, + { + "epoch": 0.49532774494984194, + "grad_norm": 1.0295915603637695, + "learning_rate": 2.9744126918664318e-05, + "loss": 0.9777, + "step": 7209 + }, + { + "epoch": 0.4953964545829325, + "grad_norm": 1.0343587398529053, + "learning_rate": 2.9738239416968817e-05, + "loss": 0.9434, + "step": 7210 + }, + { + "epoch": 0.4954651642160231, + "grad_norm": 1.1026290655136108, + "learning_rate": 2.9732351642685293e-05, + "loss": 0.9847, + "step": 7211 + }, + { + "epoch": 0.4955338738491136, + "grad_norm": 1.0049492120742798, + "learning_rate": 2.9726463596152464e-05, + "loss": 0.9361, + "step": 7212 + }, + { + "epoch": 0.4956025834822042, + "grad_norm": 1.059207558631897, + "learning_rate": 2.972057527770907e-05, + "loss": 0.9662, + "step": 7213 + }, + { + "epoch": 0.4956712931152948, + "grad_norm": 1.0805084705352783, + "learning_rate": 2.9714686687693853e-05, + "loss": 0.9195, + "step": 7214 + }, + { + "epoch": 0.4957400027483853, + "grad_norm": 0.980968177318573, + "learning_rate": 2.970879782644559e-05, + "loss": 0.9279, + "step": 7215 + }, + { + "epoch": 0.4958087123814759, + "grad_norm": 0.9925081133842468, + "learning_rate": 2.9702908694303063e-05, + "loss": 0.9551, + "step": 7216 + }, + { + "epoch": 0.4958774220145664, + "grad_norm": 1.1033062934875488, + "learning_rate": 2.9697019291605054e-05, + "loss": 0.9867, + "step": 7217 + }, + { + "epoch": 0.495946131647657, + "grad_norm": 1.050562858581543, + "learning_rate": 2.9691129618690405e-05, + "loss": 0.9702, + "step": 7218 + }, + { + "epoch": 0.4960148412807476, + "grad_norm": 1.081762433052063, + "learning_rate": 2.9685239675897914e-05, + "loss": 0.9884, + "step": 7219 + }, + { + "epoch": 0.4960835509138381, + "grad_norm": 1.1267284154891968, + "learning_rate": 2.9679349463566454e-05, + "loss": 0.9732, + "step": 7220 + }, + { + "epoch": 0.4961522605469287, + "grad_norm": 1.11149001121521, + "learning_rate": 2.9673458982034862e-05, + "loss": 0.9492, + "step": 7221 + }, + { + "epoch": 0.49622097018001926, + "grad_norm": 1.108323335647583, + "learning_rate": 2.9667568231642023e-05, + "loss": 1.006, + "step": 7222 + }, + { + "epoch": 0.4962896798131098, + "grad_norm": 1.0482523441314697, + "learning_rate": 2.966167721272684e-05, + "loss": 0.9888, + "step": 7223 + }, + { + "epoch": 0.49635838944620037, + "grad_norm": 1.1044590473175049, + "learning_rate": 2.9655785925628203e-05, + "loss": 0.969, + "step": 7224 + }, + { + "epoch": 0.4964270990792909, + "grad_norm": 1.043988585472107, + "learning_rate": 2.9649894370685046e-05, + "loss": 0.9677, + "step": 7225 + }, + { + "epoch": 0.49649580871238147, + "grad_norm": 1.2007534503936768, + "learning_rate": 2.9644002548236293e-05, + "loss": 0.9769, + "step": 7226 + }, + { + "epoch": 0.49656451834547205, + "grad_norm": 1.094581127166748, + "learning_rate": 2.963811045862091e-05, + "loss": 0.9809, + "step": 7227 + }, + { + "epoch": 0.4966332279785626, + "grad_norm": 1.0310256481170654, + "learning_rate": 2.9632218102177862e-05, + "loss": 0.96, + "step": 7228 + }, + { + "epoch": 0.49670193761165315, + "grad_norm": 1.0188568830490112, + "learning_rate": 2.9626325479246132e-05, + "loss": 0.9787, + "step": 7229 + }, + { + "epoch": 0.49677064724474373, + "grad_norm": 1.0949138402938843, + "learning_rate": 2.9620432590164717e-05, + "loss": 1.0055, + "step": 7230 + }, + { + "epoch": 0.49683935687783426, + "grad_norm": 1.095057725906372, + "learning_rate": 2.9614539435272626e-05, + "loss": 0.9812, + "step": 7231 + }, + { + "epoch": 0.49690806651092484, + "grad_norm": 1.1648569107055664, + "learning_rate": 2.9608646014908893e-05, + "loss": 0.9594, + "step": 7232 + }, + { + "epoch": 0.49697677614401536, + "grad_norm": 1.0674409866333008, + "learning_rate": 2.9602752329412576e-05, + "loss": 0.9561, + "step": 7233 + }, + { + "epoch": 0.49704548577710594, + "grad_norm": 1.0760630369186401, + "learning_rate": 2.9596858379122718e-05, + "loss": 1.0048, + "step": 7234 + }, + { + "epoch": 0.4971141954101965, + "grad_norm": 1.0303515195846558, + "learning_rate": 2.95909641643784e-05, + "loss": 0.9503, + "step": 7235 + }, + { + "epoch": 0.49718290504328705, + "grad_norm": 1.1617141962051392, + "learning_rate": 2.9585069685518714e-05, + "loss": 0.9918, + "step": 7236 + }, + { + "epoch": 0.49725161467637763, + "grad_norm": 1.1705117225646973, + "learning_rate": 2.957917494288276e-05, + "loss": 1.0091, + "step": 7237 + }, + { + "epoch": 0.4973203243094682, + "grad_norm": 0.997304379940033, + "learning_rate": 2.9573279936809667e-05, + "loss": 0.9747, + "step": 7238 + }, + { + "epoch": 0.49738903394255873, + "grad_norm": 1.0629680156707764, + "learning_rate": 2.956738466763857e-05, + "loss": 0.9566, + "step": 7239 + }, + { + "epoch": 0.4974577435756493, + "grad_norm": 1.0115952491760254, + "learning_rate": 2.956148913570862e-05, + "loss": 0.908, + "step": 7240 + }, + { + "epoch": 0.49752645320873984, + "grad_norm": 1.1166189908981323, + "learning_rate": 2.9555593341358977e-05, + "loss": 1.0019, + "step": 7241 + }, + { + "epoch": 0.4975951628418304, + "grad_norm": 1.147104024887085, + "learning_rate": 2.9549697284928823e-05, + "loss": 0.9909, + "step": 7242 + }, + { + "epoch": 0.497663872474921, + "grad_norm": 1.0290166139602661, + "learning_rate": 2.9543800966757368e-05, + "loss": 0.9153, + "step": 7243 + }, + { + "epoch": 0.4977325821080115, + "grad_norm": 1.07168710231781, + "learning_rate": 2.9537904387183807e-05, + "loss": 0.9557, + "step": 7244 + }, + { + "epoch": 0.4978012917411021, + "grad_norm": 1.0981512069702148, + "learning_rate": 2.953200754654738e-05, + "loss": 0.9826, + "step": 7245 + }, + { + "epoch": 0.4978700013741927, + "grad_norm": 1.0624533891677856, + "learning_rate": 2.9526110445187316e-05, + "loss": 1.0214, + "step": 7246 + }, + { + "epoch": 0.4979387110072832, + "grad_norm": 1.0962226390838623, + "learning_rate": 2.952021308344287e-05, + "loss": 0.9544, + "step": 7247 + }, + { + "epoch": 0.4980074206403738, + "grad_norm": 0.9911406636238098, + "learning_rate": 2.9514315461653337e-05, + "loss": 0.9649, + "step": 7248 + }, + { + "epoch": 0.4980761302734643, + "grad_norm": 0.9791325926780701, + "learning_rate": 2.9508417580157975e-05, + "loss": 0.9076, + "step": 7249 + }, + { + "epoch": 0.4981448399065549, + "grad_norm": 1.022562026977539, + "learning_rate": 2.9502519439296107e-05, + "loss": 0.9641, + "step": 7250 + }, + { + "epoch": 0.4982135495396455, + "grad_norm": 1.0405131578445435, + "learning_rate": 2.949662103940704e-05, + "loss": 0.9168, + "step": 7251 + }, + { + "epoch": 0.498282259172736, + "grad_norm": 1.1761159896850586, + "learning_rate": 2.949072238083011e-05, + "loss": 0.9816, + "step": 7252 + }, + { + "epoch": 0.4983509688058266, + "grad_norm": 1.1181639432907104, + "learning_rate": 2.948482346390466e-05, + "loss": 1.0179, + "step": 7253 + }, + { + "epoch": 0.49841967843891716, + "grad_norm": 1.126877784729004, + "learning_rate": 2.9478924288970046e-05, + "loss": 1.0412, + "step": 7254 + }, + { + "epoch": 0.4984883880720077, + "grad_norm": 1.1686887741088867, + "learning_rate": 2.9473024856365644e-05, + "loss": 1.0, + "step": 7255 + }, + { + "epoch": 0.49855709770509826, + "grad_norm": 1.2219276428222656, + "learning_rate": 2.9467125166430854e-05, + "loss": 0.9935, + "step": 7256 + }, + { + "epoch": 0.4986258073381888, + "grad_norm": 1.2182501554489136, + "learning_rate": 2.946122521950507e-05, + "loss": 1.0404, + "step": 7257 + }, + { + "epoch": 0.49869451697127937, + "grad_norm": 1.0666805505752563, + "learning_rate": 2.945532501592773e-05, + "loss": 0.9043, + "step": 7258 + }, + { + "epoch": 0.49876322660436995, + "grad_norm": 1.0100593566894531, + "learning_rate": 2.9449424556038256e-05, + "loss": 0.9673, + "step": 7259 + }, + { + "epoch": 0.49883193623746047, + "grad_norm": 1.131700873374939, + "learning_rate": 2.9443523840176097e-05, + "loss": 0.9652, + "step": 7260 + }, + { + "epoch": 0.49890064587055105, + "grad_norm": 0.9933229684829712, + "learning_rate": 2.9437622868680725e-05, + "loss": 0.9533, + "step": 7261 + }, + { + "epoch": 0.49896935550364163, + "grad_norm": 1.0775938034057617, + "learning_rate": 2.943172164189161e-05, + "loss": 0.9231, + "step": 7262 + }, + { + "epoch": 0.49903806513673216, + "grad_norm": 1.1018834114074707, + "learning_rate": 2.9425820160148255e-05, + "loss": 0.9807, + "step": 7263 + }, + { + "epoch": 0.49910677476982274, + "grad_norm": 1.059148907661438, + "learning_rate": 2.9419918423790165e-05, + "loss": 0.9439, + "step": 7264 + }, + { + "epoch": 0.49917548440291326, + "grad_norm": 0.9982153177261353, + "learning_rate": 2.941401643315686e-05, + "loss": 0.9664, + "step": 7265 + }, + { + "epoch": 0.49924419403600384, + "grad_norm": 1.0509912967681885, + "learning_rate": 2.9408114188587887e-05, + "loss": 0.9458, + "step": 7266 + }, + { + "epoch": 0.4993129036690944, + "grad_norm": 1.0779180526733398, + "learning_rate": 2.9402211690422787e-05, + "loss": 0.9955, + "step": 7267 + }, + { + "epoch": 0.49938161330218495, + "grad_norm": 1.081323504447937, + "learning_rate": 2.9396308939001132e-05, + "loss": 0.9427, + "step": 7268 + }, + { + "epoch": 0.4994503229352755, + "grad_norm": 1.086982011795044, + "learning_rate": 2.9390405934662506e-05, + "loss": 0.9382, + "step": 7269 + }, + { + "epoch": 0.4995190325683661, + "grad_norm": 1.0510154962539673, + "learning_rate": 2.93845026777465e-05, + "loss": 0.9895, + "step": 7270 + }, + { + "epoch": 0.49958774220145663, + "grad_norm": 1.11043119430542, + "learning_rate": 2.9378599168592734e-05, + "loss": 0.9805, + "step": 7271 + }, + { + "epoch": 0.4996564518345472, + "grad_norm": 1.0846375226974487, + "learning_rate": 2.9372695407540823e-05, + "loss": 1.0134, + "step": 7272 + }, + { + "epoch": 0.49972516146763774, + "grad_norm": 1.0987173318862915, + "learning_rate": 2.9366791394930416e-05, + "loss": 1.0311, + "step": 7273 + }, + { + "epoch": 0.4997938711007283, + "grad_norm": 1.0897959470748901, + "learning_rate": 2.936088713110116e-05, + "loss": 0.9538, + "step": 7274 + }, + { + "epoch": 0.4998625807338189, + "grad_norm": 1.08627188205719, + "learning_rate": 2.935498261639273e-05, + "loss": 0.9787, + "step": 7275 + }, + { + "epoch": 0.4999312903669094, + "grad_norm": 1.1821637153625488, + "learning_rate": 2.93490778511448e-05, + "loss": 0.9477, + "step": 7276 + }, + { + "epoch": 0.5, + "grad_norm": 1.072405219078064, + "learning_rate": 2.9343172835697075e-05, + "loss": 0.967, + "step": 7277 + }, + { + "epoch": 0.5000687096330906, + "grad_norm": 1.0431324243545532, + "learning_rate": 2.9337267570389267e-05, + "loss": 0.9552, + "step": 7278 + }, + { + "epoch": 0.5001374192661812, + "grad_norm": 1.1076757907867432, + "learning_rate": 2.93313620555611e-05, + "loss": 0.9775, + "step": 7279 + }, + { + "epoch": 0.5002061288992716, + "grad_norm": 1.0678085088729858, + "learning_rate": 2.932545629155231e-05, + "loss": 0.9476, + "step": 7280 + }, + { + "epoch": 0.5002748385323622, + "grad_norm": 1.0481605529785156, + "learning_rate": 2.9319550278702662e-05, + "loss": 0.94, + "step": 7281 + }, + { + "epoch": 0.5003435481654528, + "grad_norm": 1.0288376808166504, + "learning_rate": 2.9313644017351914e-05, + "loss": 0.9425, + "step": 7282 + }, + { + "epoch": 0.5004122577985434, + "grad_norm": 1.1319606304168701, + "learning_rate": 2.9307737507839862e-05, + "loss": 0.9894, + "step": 7283 + }, + { + "epoch": 0.500480967431634, + "grad_norm": 1.0867879390716553, + "learning_rate": 2.93018307505063e-05, + "loss": 0.9446, + "step": 7284 + }, + { + "epoch": 0.5005496770647245, + "grad_norm": 1.0595815181732178, + "learning_rate": 2.929592374569103e-05, + "loss": 0.9693, + "step": 7285 + }, + { + "epoch": 0.500618386697815, + "grad_norm": 0.9974536895751953, + "learning_rate": 2.9290016493733892e-05, + "loss": 0.945, + "step": 7286 + }, + { + "epoch": 0.5006870963309056, + "grad_norm": 1.0901464223861694, + "learning_rate": 2.9284108994974717e-05, + "loss": 1.0059, + "step": 7287 + }, + { + "epoch": 0.5007558059639962, + "grad_norm": 1.0895167589187622, + "learning_rate": 2.9278201249753366e-05, + "loss": 0.9726, + "step": 7288 + }, + { + "epoch": 0.5008245155970867, + "grad_norm": 1.1003408432006836, + "learning_rate": 2.927229325840971e-05, + "loss": 0.923, + "step": 7289 + }, + { + "epoch": 0.5008932252301773, + "grad_norm": 1.0202510356903076, + "learning_rate": 2.926638502128362e-05, + "loss": 0.9881, + "step": 7290 + }, + { + "epoch": 0.5009619348632678, + "grad_norm": 1.1000051498413086, + "learning_rate": 2.926047653871501e-05, + "loss": 0.9742, + "step": 7291 + }, + { + "epoch": 0.5010306444963584, + "grad_norm": 1.0754607915878296, + "learning_rate": 2.9254567811043777e-05, + "loss": 1.0105, + "step": 7292 + }, + { + "epoch": 0.501099354129449, + "grad_norm": 0.9918239116668701, + "learning_rate": 2.9248658838609843e-05, + "loss": 0.9607, + "step": 7293 + }, + { + "epoch": 0.5011680637625395, + "grad_norm": 1.0785366296768188, + "learning_rate": 2.9242749621753172e-05, + "loss": 0.9548, + "step": 7294 + }, + { + "epoch": 0.5012367733956301, + "grad_norm": 1.068550705909729, + "learning_rate": 2.9236840160813695e-05, + "loss": 0.9497, + "step": 7295 + }, + { + "epoch": 0.5013054830287206, + "grad_norm": 0.9998784065246582, + "learning_rate": 2.923093045613139e-05, + "loss": 0.9404, + "step": 7296 + }, + { + "epoch": 0.5013741926618112, + "grad_norm": 1.09646737575531, + "learning_rate": 2.9225020508046232e-05, + "loss": 0.9446, + "step": 7297 + }, + { + "epoch": 0.5014429022949017, + "grad_norm": 1.092983603477478, + "learning_rate": 2.9219110316898214e-05, + "loss": 0.9361, + "step": 7298 + }, + { + "epoch": 0.5015116119279923, + "grad_norm": 1.013128399848938, + "learning_rate": 2.9213199883027366e-05, + "loss": 0.9441, + "step": 7299 + }, + { + "epoch": 0.5015803215610829, + "grad_norm": 1.131780982017517, + "learning_rate": 2.9207289206773692e-05, + "loss": 0.989, + "step": 7300 + }, + { + "epoch": 0.5016490311941735, + "grad_norm": 1.1409029960632324, + "learning_rate": 2.9201378288477237e-05, + "loss": 0.9449, + "step": 7301 + }, + { + "epoch": 0.501717740827264, + "grad_norm": 1.0572909116744995, + "learning_rate": 2.9195467128478044e-05, + "loss": 0.9579, + "step": 7302 + }, + { + "epoch": 0.5017864504603545, + "grad_norm": 1.0834685564041138, + "learning_rate": 2.9189555727116185e-05, + "loss": 0.9456, + "step": 7303 + }, + { + "epoch": 0.5018551600934451, + "grad_norm": 1.1034884452819824, + "learning_rate": 2.9183644084731748e-05, + "loss": 0.985, + "step": 7304 + }, + { + "epoch": 0.5019238697265357, + "grad_norm": 1.093835711479187, + "learning_rate": 2.9177732201664814e-05, + "loss": 0.969, + "step": 7305 + }, + { + "epoch": 0.5019925793596263, + "grad_norm": 1.0401171445846558, + "learning_rate": 2.917182007825549e-05, + "loss": 0.9744, + "step": 7306 + }, + { + "epoch": 0.5020612889927167, + "grad_norm": 1.0661449432373047, + "learning_rate": 2.9165907714843905e-05, + "loss": 0.9276, + "step": 7307 + }, + { + "epoch": 0.5021299986258073, + "grad_norm": 1.0954790115356445, + "learning_rate": 2.915999511177019e-05, + "loss": 0.9668, + "step": 7308 + }, + { + "epoch": 0.5021987082588979, + "grad_norm": 1.0766305923461914, + "learning_rate": 2.915408226937449e-05, + "loss": 0.9624, + "step": 7309 + }, + { + "epoch": 0.5022674178919885, + "grad_norm": 1.1351209878921509, + "learning_rate": 2.9148169187996965e-05, + "loss": 0.9672, + "step": 7310 + }, + { + "epoch": 0.5023361275250791, + "grad_norm": 1.0149521827697754, + "learning_rate": 2.914225586797781e-05, + "loss": 0.9421, + "step": 7311 + }, + { + "epoch": 0.5024048371581695, + "grad_norm": 1.1194168329238892, + "learning_rate": 2.9136342309657193e-05, + "loss": 0.992, + "step": 7312 + }, + { + "epoch": 0.5024735467912601, + "grad_norm": 1.1229437589645386, + "learning_rate": 2.9130428513375324e-05, + "loss": 0.9548, + "step": 7313 + }, + { + "epoch": 0.5025422564243507, + "grad_norm": 1.0364174842834473, + "learning_rate": 2.912451447947242e-05, + "loss": 0.9202, + "step": 7314 + }, + { + "epoch": 0.5026109660574413, + "grad_norm": 1.0807201862335205, + "learning_rate": 2.9118600208288715e-05, + "loss": 0.9759, + "step": 7315 + }, + { + "epoch": 0.5026796756905318, + "grad_norm": 1.0578818321228027, + "learning_rate": 2.9112685700164443e-05, + "loss": 0.9422, + "step": 7316 + }, + { + "epoch": 0.5027483853236224, + "grad_norm": 1.1039384603500366, + "learning_rate": 2.910677095543988e-05, + "loss": 0.9739, + "step": 7317 + }, + { + "epoch": 0.5028170949567129, + "grad_norm": 1.1557139158248901, + "learning_rate": 2.9100855974455286e-05, + "loss": 0.9758, + "step": 7318 + }, + { + "epoch": 0.5028858045898035, + "grad_norm": 1.169468641281128, + "learning_rate": 2.9094940757550955e-05, + "loss": 0.9749, + "step": 7319 + }, + { + "epoch": 0.5029545142228941, + "grad_norm": 1.1305807828903198, + "learning_rate": 2.9089025305067163e-05, + "loss": 0.9783, + "step": 7320 + }, + { + "epoch": 0.5030232238559846, + "grad_norm": 1.1032986640930176, + "learning_rate": 2.908310961734425e-05, + "loss": 0.9627, + "step": 7321 + }, + { + "epoch": 0.5030919334890752, + "grad_norm": 1.0794436931610107, + "learning_rate": 2.9077193694722526e-05, + "loss": 0.9798, + "step": 7322 + }, + { + "epoch": 0.5031606431221657, + "grad_norm": 1.0324206352233887, + "learning_rate": 2.907127753754234e-05, + "loss": 0.9671, + "step": 7323 + }, + { + "epoch": 0.5032293527552563, + "grad_norm": 1.0317940711975098, + "learning_rate": 2.9065361146144044e-05, + "loss": 0.9408, + "step": 7324 + }, + { + "epoch": 0.5032980623883468, + "grad_norm": 1.0784310102462769, + "learning_rate": 2.9059444520867985e-05, + "loss": 0.9601, + "step": 7325 + }, + { + "epoch": 0.5033667720214374, + "grad_norm": 1.1490144729614258, + "learning_rate": 2.9053527662054565e-05, + "loss": 0.9951, + "step": 7326 + }, + { + "epoch": 0.503435481654528, + "grad_norm": 1.0230766534805298, + "learning_rate": 2.904761057004417e-05, + "loss": 0.9802, + "step": 7327 + }, + { + "epoch": 0.5035041912876185, + "grad_norm": 1.0580589771270752, + "learning_rate": 2.9041693245177204e-05, + "loss": 0.9504, + "step": 7328 + }, + { + "epoch": 0.503572900920709, + "grad_norm": 1.0055389404296875, + "learning_rate": 2.9035775687794096e-05, + "loss": 0.9418, + "step": 7329 + }, + { + "epoch": 0.5036416105537996, + "grad_norm": 1.1137909889221191, + "learning_rate": 2.902985789823527e-05, + "loss": 0.926, + "step": 7330 + }, + { + "epoch": 0.5037103201868902, + "grad_norm": 1.0990859270095825, + "learning_rate": 2.9023939876841177e-05, + "loss": 0.9821, + "step": 7331 + }, + { + "epoch": 0.5037790298199808, + "grad_norm": 1.1719112396240234, + "learning_rate": 2.9018021623952275e-05, + "loss": 1.0258, + "step": 7332 + }, + { + "epoch": 0.5038477394530714, + "grad_norm": 1.0481983423233032, + "learning_rate": 2.901210313990904e-05, + "loss": 0.9266, + "step": 7333 + }, + { + "epoch": 0.5039164490861618, + "grad_norm": 1.0836517810821533, + "learning_rate": 2.9006184425051958e-05, + "loss": 0.9531, + "step": 7334 + }, + { + "epoch": 0.5039851587192524, + "grad_norm": 1.0456571578979492, + "learning_rate": 2.9000265479721526e-05, + "loss": 0.9277, + "step": 7335 + }, + { + "epoch": 0.504053868352343, + "grad_norm": 1.1293396949768066, + "learning_rate": 2.8994346304258267e-05, + "loss": 0.9741, + "step": 7336 + }, + { + "epoch": 0.5041225779854336, + "grad_norm": 1.0013254880905151, + "learning_rate": 2.8988426899002697e-05, + "loss": 0.9017, + "step": 7337 + }, + { + "epoch": 0.5041912876185242, + "grad_norm": 1.0806059837341309, + "learning_rate": 2.8982507264295355e-05, + "loss": 0.9396, + "step": 7338 + }, + { + "epoch": 0.5042599972516146, + "grad_norm": 1.0383386611938477, + "learning_rate": 2.8976587400476802e-05, + "loss": 0.9443, + "step": 7339 + }, + { + "epoch": 0.5043287068847052, + "grad_norm": 1.1290233135223389, + "learning_rate": 2.8970667307887606e-05, + "loss": 1.008, + "step": 7340 + }, + { + "epoch": 0.5043974165177958, + "grad_norm": 1.0182485580444336, + "learning_rate": 2.896474698686833e-05, + "loss": 0.9501, + "step": 7341 + }, + { + "epoch": 0.5044661261508864, + "grad_norm": 1.0665053129196167, + "learning_rate": 2.8958826437759586e-05, + "loss": 0.9526, + "step": 7342 + }, + { + "epoch": 0.504534835783977, + "grad_norm": 1.0773673057556152, + "learning_rate": 2.8952905660901962e-05, + "loss": 0.983, + "step": 7343 + }, + { + "epoch": 0.5046035454170674, + "grad_norm": 1.0798490047454834, + "learning_rate": 2.894698465663609e-05, + "loss": 0.9398, + "step": 7344 + }, + { + "epoch": 0.504672255050158, + "grad_norm": 1.060197114944458, + "learning_rate": 2.89410634253026e-05, + "loss": 0.9931, + "step": 7345 + }, + { + "epoch": 0.5047409646832486, + "grad_norm": 1.1089712381362915, + "learning_rate": 2.893514196724213e-05, + "loss": 0.9344, + "step": 7346 + }, + { + "epoch": 0.5048096743163392, + "grad_norm": 1.0942414999008179, + "learning_rate": 2.8929220282795348e-05, + "loss": 0.974, + "step": 7347 + }, + { + "epoch": 0.5048783839494297, + "grad_norm": 1.146021842956543, + "learning_rate": 2.892329837230292e-05, + "loss": 0.9808, + "step": 7348 + }, + { + "epoch": 0.5049470935825203, + "grad_norm": 1.1528167724609375, + "learning_rate": 2.891737623610552e-05, + "loss": 0.9557, + "step": 7349 + }, + { + "epoch": 0.5050158032156108, + "grad_norm": 1.092326283454895, + "learning_rate": 2.891145387454386e-05, + "loss": 0.9781, + "step": 7350 + }, + { + "epoch": 0.5050845128487014, + "grad_norm": 1.0560706853866577, + "learning_rate": 2.8905531287958636e-05, + "loss": 0.9713, + "step": 7351 + }, + { + "epoch": 0.505153222481792, + "grad_norm": 1.1170462369918823, + "learning_rate": 2.889960847669059e-05, + "loss": 0.9721, + "step": 7352 + }, + { + "epoch": 0.5052219321148825, + "grad_norm": 1.1261924505233765, + "learning_rate": 2.8893685441080437e-05, + "loss": 0.9788, + "step": 7353 + }, + { + "epoch": 0.5052906417479731, + "grad_norm": 1.0351604223251343, + "learning_rate": 2.8887762181468936e-05, + "loss": 0.973, + "step": 7354 + }, + { + "epoch": 0.5053593513810636, + "grad_norm": 1.0589497089385986, + "learning_rate": 2.888183869819685e-05, + "loss": 0.9811, + "step": 7355 + }, + { + "epoch": 0.5054280610141542, + "grad_norm": 1.1533671617507935, + "learning_rate": 2.8875914991604948e-05, + "loss": 0.9786, + "step": 7356 + }, + { + "epoch": 0.5054967706472447, + "grad_norm": 1.13767671585083, + "learning_rate": 2.8869991062034025e-05, + "loss": 0.988, + "step": 7357 + }, + { + "epoch": 0.5055654802803353, + "grad_norm": 1.0646300315856934, + "learning_rate": 2.8864066909824866e-05, + "loss": 0.9788, + "step": 7358 + }, + { + "epoch": 0.5056341899134259, + "grad_norm": 1.1401029825210571, + "learning_rate": 2.88581425353183e-05, + "loss": 0.9299, + "step": 7359 + }, + { + "epoch": 0.5057028995465164, + "grad_norm": 1.0848826169967651, + "learning_rate": 2.8852217938855152e-05, + "loss": 0.9822, + "step": 7360 + }, + { + "epoch": 0.505771609179607, + "grad_norm": 1.0324338674545288, + "learning_rate": 2.8846293120776245e-05, + "loss": 0.9807, + "step": 7361 + }, + { + "epoch": 0.5058403188126975, + "grad_norm": 1.0535567998886108, + "learning_rate": 2.8840368081422443e-05, + "loss": 0.9429, + "step": 7362 + }, + { + "epoch": 0.5059090284457881, + "grad_norm": 1.0699615478515625, + "learning_rate": 2.883444282113461e-05, + "loss": 0.981, + "step": 7363 + }, + { + "epoch": 0.5059777380788787, + "grad_norm": 1.104823112487793, + "learning_rate": 2.8828517340253613e-05, + "loss": 0.9875, + "step": 7364 + }, + { + "epoch": 0.5060464477119693, + "grad_norm": 1.008460521697998, + "learning_rate": 2.8822591639120355e-05, + "loss": 0.9806, + "step": 7365 + }, + { + "epoch": 0.5061151573450597, + "grad_norm": 1.0932495594024658, + "learning_rate": 2.8816665718075724e-05, + "loss": 0.9648, + "step": 7366 + }, + { + "epoch": 0.5061838669781503, + "grad_norm": 1.0857678651809692, + "learning_rate": 2.8810739577460644e-05, + "loss": 0.9565, + "step": 7367 + }, + { + "epoch": 0.5062525766112409, + "grad_norm": 1.1443966627120972, + "learning_rate": 2.8804813217616044e-05, + "loss": 0.9691, + "step": 7368 + }, + { + "epoch": 0.5063212862443315, + "grad_norm": 1.1255162954330444, + "learning_rate": 2.8798886638882855e-05, + "loss": 0.9901, + "step": 7369 + }, + { + "epoch": 0.5063899958774221, + "grad_norm": 1.1002426147460938, + "learning_rate": 2.8792959841602034e-05, + "loss": 0.9682, + "step": 7370 + }, + { + "epoch": 0.5064587055105125, + "grad_norm": 1.0662792921066284, + "learning_rate": 2.8787032826114552e-05, + "loss": 0.949, + "step": 7371 + }, + { + "epoch": 0.5065274151436031, + "grad_norm": 1.0770518779754639, + "learning_rate": 2.878110559276137e-05, + "loss": 1.0057, + "step": 7372 + }, + { + "epoch": 0.5065961247766937, + "grad_norm": 1.1123597621917725, + "learning_rate": 2.8775178141883502e-05, + "loss": 0.9449, + "step": 7373 + }, + { + "epoch": 0.5066648344097843, + "grad_norm": 1.0712920427322388, + "learning_rate": 2.8769250473821925e-05, + "loss": 0.9825, + "step": 7374 + }, + { + "epoch": 0.5067335440428749, + "grad_norm": 1.0986428260803223, + "learning_rate": 2.8763322588917674e-05, + "loss": 1.0313, + "step": 7375 + }, + { + "epoch": 0.5068022536759653, + "grad_norm": 1.0819892883300781, + "learning_rate": 2.875739448751176e-05, + "loss": 0.9716, + "step": 7376 + }, + { + "epoch": 0.5068709633090559, + "grad_norm": 1.0339531898498535, + "learning_rate": 2.875146616994524e-05, + "loss": 0.9533, + "step": 7377 + }, + { + "epoch": 0.5069396729421465, + "grad_norm": 1.1681475639343262, + "learning_rate": 2.874553763655916e-05, + "loss": 0.9841, + "step": 7378 + }, + { + "epoch": 0.5070083825752371, + "grad_norm": 1.0391077995300293, + "learning_rate": 2.873960888769458e-05, + "loss": 0.9553, + "step": 7379 + }, + { + "epoch": 0.5070770922083276, + "grad_norm": 1.0456880331039429, + "learning_rate": 2.873367992369258e-05, + "loss": 0.9594, + "step": 7380 + }, + { + "epoch": 0.5071458018414182, + "grad_norm": 1.0082639455795288, + "learning_rate": 2.8727750744894248e-05, + "loss": 0.9477, + "step": 7381 + }, + { + "epoch": 0.5072145114745087, + "grad_norm": 1.101845622062683, + "learning_rate": 2.8721821351640697e-05, + "loss": 0.965, + "step": 7382 + }, + { + "epoch": 0.5072832211075993, + "grad_norm": 1.0190818309783936, + "learning_rate": 2.8715891744273034e-05, + "loss": 0.9366, + "step": 7383 + }, + { + "epoch": 0.5073519307406898, + "grad_norm": 1.039903163909912, + "learning_rate": 2.8709961923132377e-05, + "loss": 0.9494, + "step": 7384 + }, + { + "epoch": 0.5074206403737804, + "grad_norm": 1.0878939628601074, + "learning_rate": 2.870403188855988e-05, + "loss": 0.9892, + "step": 7385 + }, + { + "epoch": 0.507489350006871, + "grad_norm": 1.0146416425704956, + "learning_rate": 2.8698101640896673e-05, + "loss": 0.9606, + "step": 7386 + }, + { + "epoch": 0.5075580596399615, + "grad_norm": 1.0940032005310059, + "learning_rate": 2.869217118048394e-05, + "loss": 0.9667, + "step": 7387 + }, + { + "epoch": 0.5076267692730521, + "grad_norm": 1.0812238454818726, + "learning_rate": 2.868624050766286e-05, + "loss": 0.9694, + "step": 7388 + }, + { + "epoch": 0.5076954789061426, + "grad_norm": 1.085328459739685, + "learning_rate": 2.86803096227746e-05, + "loss": 0.9771, + "step": 7389 + }, + { + "epoch": 0.5077641885392332, + "grad_norm": 1.10030198097229, + "learning_rate": 2.867437852616037e-05, + "loss": 0.9743, + "step": 7390 + }, + { + "epoch": 0.5078328981723238, + "grad_norm": 1.1582491397857666, + "learning_rate": 2.8668447218161393e-05, + "loss": 0.9827, + "step": 7391 + }, + { + "epoch": 0.5079016078054143, + "grad_norm": 1.0815428495407104, + "learning_rate": 2.866251569911888e-05, + "loss": 0.972, + "step": 7392 + }, + { + "epoch": 0.5079703174385048, + "grad_norm": 1.036820411682129, + "learning_rate": 2.8656583969374073e-05, + "loss": 0.9798, + "step": 7393 + }, + { + "epoch": 0.5080390270715954, + "grad_norm": 1.0190948247909546, + "learning_rate": 2.8650652029268223e-05, + "loss": 0.9899, + "step": 7394 + }, + { + "epoch": 0.508107736704686, + "grad_norm": 1.1068756580352783, + "learning_rate": 2.864471987914258e-05, + "loss": 1.0138, + "step": 7395 + }, + { + "epoch": 0.5081764463377766, + "grad_norm": 1.0745384693145752, + "learning_rate": 2.863878751933843e-05, + "loss": 0.9348, + "step": 7396 + }, + { + "epoch": 0.5082451559708672, + "grad_norm": 1.0815799236297607, + "learning_rate": 2.8632854950197053e-05, + "loss": 0.9771, + "step": 7397 + }, + { + "epoch": 0.5083138656039576, + "grad_norm": 1.0432552099227905, + "learning_rate": 2.8626922172059746e-05, + "loss": 0.9506, + "step": 7398 + }, + { + "epoch": 0.5083825752370482, + "grad_norm": 1.1905966997146606, + "learning_rate": 2.862098918526781e-05, + "loss": 0.9636, + "step": 7399 + }, + { + "epoch": 0.5084512848701388, + "grad_norm": 1.1358150243759155, + "learning_rate": 2.8615055990162576e-05, + "loss": 1.0112, + "step": 7400 + }, + { + "epoch": 0.5085199945032294, + "grad_norm": 1.1335431337356567, + "learning_rate": 2.8609122587085384e-05, + "loss": 1.0194, + "step": 7401 + }, + { + "epoch": 0.50858870413632, + "grad_norm": 1.0695489645004272, + "learning_rate": 2.860318897637756e-05, + "loss": 0.9776, + "step": 7402 + }, + { + "epoch": 0.5086574137694104, + "grad_norm": 1.0791170597076416, + "learning_rate": 2.8597255158380482e-05, + "loss": 0.9688, + "step": 7403 + }, + { + "epoch": 0.508726123402501, + "grad_norm": 1.0665687322616577, + "learning_rate": 2.85913211334355e-05, + "loss": 0.948, + "step": 7404 + }, + { + "epoch": 0.5087948330355916, + "grad_norm": 1.0285413265228271, + "learning_rate": 2.8585386901884003e-05, + "loss": 0.925, + "step": 7405 + }, + { + "epoch": 0.5088635426686822, + "grad_norm": 1.0370707511901855, + "learning_rate": 2.8579452464067385e-05, + "loss": 0.8995, + "step": 7406 + }, + { + "epoch": 0.5089322523017727, + "grad_norm": 1.1169564723968506, + "learning_rate": 2.8573517820327055e-05, + "loss": 0.9618, + "step": 7407 + }, + { + "epoch": 0.5090009619348632, + "grad_norm": 1.0267781019210815, + "learning_rate": 2.856758297100442e-05, + "loss": 0.9266, + "step": 7408 + }, + { + "epoch": 0.5090696715679538, + "grad_norm": 1.051588773727417, + "learning_rate": 2.856164791644091e-05, + "loss": 0.951, + "step": 7409 + }, + { + "epoch": 0.5091383812010444, + "grad_norm": 1.1064505577087402, + "learning_rate": 2.8555712656977962e-05, + "loss": 0.9782, + "step": 7410 + }, + { + "epoch": 0.509207090834135, + "grad_norm": 1.0249520540237427, + "learning_rate": 2.8549777192957046e-05, + "loss": 0.9331, + "step": 7411 + }, + { + "epoch": 0.5092758004672255, + "grad_norm": 1.1087316274642944, + "learning_rate": 2.85438415247196e-05, + "loss": 0.9258, + "step": 7412 + }, + { + "epoch": 0.5093445101003161, + "grad_norm": 1.0732601881027222, + "learning_rate": 2.8537905652607122e-05, + "loss": 0.9579, + "step": 7413 + }, + { + "epoch": 0.5094132197334066, + "grad_norm": 1.0780125856399536, + "learning_rate": 2.8531969576961082e-05, + "loss": 0.9768, + "step": 7414 + }, + { + "epoch": 0.5094819293664972, + "grad_norm": 1.1190794706344604, + "learning_rate": 2.8526033298122985e-05, + "loss": 0.9704, + "step": 7415 + }, + { + "epoch": 0.5095506389995877, + "grad_norm": 1.07879638671875, + "learning_rate": 2.8520096816434343e-05, + "loss": 0.9944, + "step": 7416 + }, + { + "epoch": 0.5096193486326783, + "grad_norm": 1.0422766208648682, + "learning_rate": 2.851416013223668e-05, + "loss": 0.9444, + "step": 7417 + }, + { + "epoch": 0.5096880582657689, + "grad_norm": 1.1411949396133423, + "learning_rate": 2.850822324587153e-05, + "loss": 0.9705, + "step": 7418 + }, + { + "epoch": 0.5097567678988594, + "grad_norm": 1.0387260913848877, + "learning_rate": 2.8502286157680436e-05, + "loss": 0.9722, + "step": 7419 + }, + { + "epoch": 0.50982547753195, + "grad_norm": 1.0857200622558594, + "learning_rate": 2.8496348868004946e-05, + "loss": 0.9161, + "step": 7420 + }, + { + "epoch": 0.5098941871650405, + "grad_norm": 1.0358450412750244, + "learning_rate": 2.849041137718665e-05, + "loss": 0.9773, + "step": 7421 + }, + { + "epoch": 0.5099628967981311, + "grad_norm": 1.0487775802612305, + "learning_rate": 2.8484473685567103e-05, + "loss": 0.9811, + "step": 7422 + }, + { + "epoch": 0.5100316064312217, + "grad_norm": 1.0488477945327759, + "learning_rate": 2.8478535793487916e-05, + "loss": 0.9983, + "step": 7423 + }, + { + "epoch": 0.5101003160643122, + "grad_norm": 1.0800267457962036, + "learning_rate": 2.847259770129068e-05, + "loss": 0.9567, + "step": 7424 + }, + { + "epoch": 0.5101690256974027, + "grad_norm": 1.0608975887298584, + "learning_rate": 2.846665940931702e-05, + "loss": 0.9238, + "step": 7425 + }, + { + "epoch": 0.5102377353304933, + "grad_norm": 1.1213032007217407, + "learning_rate": 2.8460720917908563e-05, + "loss": 0.9389, + "step": 7426 + }, + { + "epoch": 0.5103064449635839, + "grad_norm": 1.0988212823867798, + "learning_rate": 2.8454782227406935e-05, + "loss": 0.9431, + "step": 7427 + }, + { + "epoch": 0.5103751545966745, + "grad_norm": 1.1149734258651733, + "learning_rate": 2.844884333815379e-05, + "loss": 0.9747, + "step": 7428 + }, + { + "epoch": 0.5104438642297651, + "grad_norm": 1.140520691871643, + "learning_rate": 2.84429042504908e-05, + "loss": 0.9873, + "step": 7429 + }, + { + "epoch": 0.5105125738628555, + "grad_norm": 1.0979593992233276, + "learning_rate": 2.8436964964759628e-05, + "loss": 0.9611, + "step": 7430 + }, + { + "epoch": 0.5105812834959461, + "grad_norm": 1.0506914854049683, + "learning_rate": 2.8431025481301947e-05, + "loss": 0.9746, + "step": 7431 + }, + { + "epoch": 0.5106499931290367, + "grad_norm": 1.0779087543487549, + "learning_rate": 2.8425085800459466e-05, + "loss": 0.9895, + "step": 7432 + }, + { + "epoch": 0.5107187027621273, + "grad_norm": 1.1063730716705322, + "learning_rate": 2.8419145922573887e-05, + "loss": 0.9869, + "step": 7433 + }, + { + "epoch": 0.5107874123952179, + "grad_norm": 1.0668694972991943, + "learning_rate": 2.8413205847986935e-05, + "loss": 0.9588, + "step": 7434 + }, + { + "epoch": 0.5108561220283083, + "grad_norm": 1.025736927986145, + "learning_rate": 2.840726557704032e-05, + "loss": 0.9103, + "step": 7435 + }, + { + "epoch": 0.5109248316613989, + "grad_norm": 1.1486419439315796, + "learning_rate": 2.8401325110075798e-05, + "loss": 0.9796, + "step": 7436 + }, + { + "epoch": 0.5109935412944895, + "grad_norm": 1.0959032773971558, + "learning_rate": 2.839538444743512e-05, + "loss": 0.9449, + "step": 7437 + }, + { + "epoch": 0.5110622509275801, + "grad_norm": 1.1012616157531738, + "learning_rate": 2.838944358946004e-05, + "loss": 0.9596, + "step": 7438 + }, + { + "epoch": 0.5111309605606706, + "grad_norm": 1.055791974067688, + "learning_rate": 2.838350253649234e-05, + "loss": 0.9213, + "step": 7439 + }, + { + "epoch": 0.5111996701937611, + "grad_norm": 1.0994250774383545, + "learning_rate": 2.83775612888738e-05, + "loss": 1.0161, + "step": 7440 + }, + { + "epoch": 0.5112683798268517, + "grad_norm": 1.000446081161499, + "learning_rate": 2.8371619846946224e-05, + "loss": 0.9783, + "step": 7441 + }, + { + "epoch": 0.5113370894599423, + "grad_norm": 1.1207865476608276, + "learning_rate": 2.836567821105141e-05, + "loss": 0.9831, + "step": 7442 + }, + { + "epoch": 0.5114057990930329, + "grad_norm": 1.080034852027893, + "learning_rate": 2.8359736381531182e-05, + "loss": 0.9785, + "step": 7443 + }, + { + "epoch": 0.5114745087261234, + "grad_norm": 1.1569472551345825, + "learning_rate": 2.835379435872737e-05, + "loss": 1.0039, + "step": 7444 + }, + { + "epoch": 0.511543218359214, + "grad_norm": 1.0874218940734863, + "learning_rate": 2.8347852142981806e-05, + "loss": 0.9723, + "step": 7445 + }, + { + "epoch": 0.5116119279923045, + "grad_norm": 1.079333782196045, + "learning_rate": 2.8341909734636353e-05, + "loss": 0.9748, + "step": 7446 + }, + { + "epoch": 0.5116806376253951, + "grad_norm": 1.141187310218811, + "learning_rate": 2.8335967134032875e-05, + "loss": 0.9688, + "step": 7447 + }, + { + "epoch": 0.5117493472584856, + "grad_norm": 1.0453555583953857, + "learning_rate": 2.833002434151324e-05, + "loss": 0.969, + "step": 7448 + }, + { + "epoch": 0.5118180568915762, + "grad_norm": 1.0907783508300781, + "learning_rate": 2.8324081357419335e-05, + "loss": 0.9222, + "step": 7449 + }, + { + "epoch": 0.5118867665246668, + "grad_norm": 1.1052857637405396, + "learning_rate": 2.8318138182093052e-05, + "loss": 0.9571, + "step": 7450 + }, + { + "epoch": 0.5119554761577573, + "grad_norm": 1.0123026371002197, + "learning_rate": 2.8312194815876304e-05, + "loss": 0.9546, + "step": 7451 + }, + { + "epoch": 0.5120241857908479, + "grad_norm": 1.0222671031951904, + "learning_rate": 2.830625125911101e-05, + "loss": 0.9251, + "step": 7452 + }, + { + "epoch": 0.5120928954239384, + "grad_norm": 1.072178840637207, + "learning_rate": 2.83003075121391e-05, + "loss": 0.9449, + "step": 7453 + }, + { + "epoch": 0.512161605057029, + "grad_norm": 1.091517686843872, + "learning_rate": 2.8294363575302508e-05, + "loss": 0.9584, + "step": 7454 + }, + { + "epoch": 0.5122303146901196, + "grad_norm": 1.0678493976593018, + "learning_rate": 2.8288419448943187e-05, + "loss": 0.9655, + "step": 7455 + }, + { + "epoch": 0.5122990243232101, + "grad_norm": 1.034224033355713, + "learning_rate": 2.8282475133403096e-05, + "loss": 0.9444, + "step": 7456 + }, + { + "epoch": 0.5123677339563006, + "grad_norm": 1.040105938911438, + "learning_rate": 2.8276530629024222e-05, + "loss": 0.9489, + "step": 7457 + }, + { + "epoch": 0.5124364435893912, + "grad_norm": 1.1082971096038818, + "learning_rate": 2.827058593614853e-05, + "loss": 0.9552, + "step": 7458 + }, + { + "epoch": 0.5125051532224818, + "grad_norm": 1.0729174613952637, + "learning_rate": 2.826464105511803e-05, + "loss": 0.9436, + "step": 7459 + }, + { + "epoch": 0.5125738628555724, + "grad_norm": 1.100282907485962, + "learning_rate": 2.825869598627472e-05, + "loss": 0.9374, + "step": 7460 + }, + { + "epoch": 0.512642572488663, + "grad_norm": 1.0253169536590576, + "learning_rate": 2.825275072996061e-05, + "loss": 0.9337, + "step": 7461 + }, + { + "epoch": 0.5127112821217534, + "grad_norm": 1.0843260288238525, + "learning_rate": 2.824680528651774e-05, + "loss": 0.9231, + "step": 7462 + }, + { + "epoch": 0.512779991754844, + "grad_norm": 1.043258547782898, + "learning_rate": 2.8240859656288138e-05, + "loss": 0.9656, + "step": 7463 + }, + { + "epoch": 0.5128487013879346, + "grad_norm": 1.1348823308944702, + "learning_rate": 2.8234913839613862e-05, + "loss": 1.0421, + "step": 7464 + }, + { + "epoch": 0.5129174110210252, + "grad_norm": 1.0470694303512573, + "learning_rate": 2.8228967836836958e-05, + "loss": 0.9762, + "step": 7465 + }, + { + "epoch": 0.5129861206541158, + "grad_norm": 1.0451632738113403, + "learning_rate": 2.8223021648299507e-05, + "loss": 0.9385, + "step": 7466 + }, + { + "epoch": 0.5130548302872062, + "grad_norm": 1.0921064615249634, + "learning_rate": 2.821707527434359e-05, + "loss": 0.9631, + "step": 7467 + }, + { + "epoch": 0.5131235399202968, + "grad_norm": 1.0280896425247192, + "learning_rate": 2.8211128715311286e-05, + "loss": 0.8945, + "step": 7468 + }, + { + "epoch": 0.5131922495533874, + "grad_norm": 1.069305419921875, + "learning_rate": 2.8205181971544704e-05, + "loss": 0.9509, + "step": 7469 + }, + { + "epoch": 0.513260959186478, + "grad_norm": 1.0661295652389526, + "learning_rate": 2.8199235043385964e-05, + "loss": 0.9811, + "step": 7470 + }, + { + "epoch": 0.5133296688195685, + "grad_norm": 1.0946505069732666, + "learning_rate": 2.819328793117718e-05, + "loss": 0.972, + "step": 7471 + }, + { + "epoch": 0.513398378452659, + "grad_norm": 1.0766737461090088, + "learning_rate": 2.818734063526049e-05, + "loss": 0.9752, + "step": 7472 + }, + { + "epoch": 0.5134670880857496, + "grad_norm": 1.079322338104248, + "learning_rate": 2.8181393155978032e-05, + "loss": 0.9317, + "step": 7473 + }, + { + "epoch": 0.5135357977188402, + "grad_norm": 1.0166569948196411, + "learning_rate": 2.8175445493671972e-05, + "loss": 0.9768, + "step": 7474 + }, + { + "epoch": 0.5136045073519308, + "grad_norm": 1.0977566242218018, + "learning_rate": 2.8169497648684462e-05, + "loss": 0.9601, + "step": 7475 + }, + { + "epoch": 0.5136732169850213, + "grad_norm": 1.0676777362823486, + "learning_rate": 2.8163549621357688e-05, + "loss": 0.9681, + "step": 7476 + }, + { + "epoch": 0.5137419266181119, + "grad_norm": 1.1402910947799683, + "learning_rate": 2.8157601412033836e-05, + "loss": 0.9909, + "step": 7477 + }, + { + "epoch": 0.5138106362512024, + "grad_norm": 1.0496255159378052, + "learning_rate": 2.8151653021055103e-05, + "loss": 0.9814, + "step": 7478 + }, + { + "epoch": 0.513879345884293, + "grad_norm": 1.0426263809204102, + "learning_rate": 2.814570444876368e-05, + "loss": 0.9415, + "step": 7479 + }, + { + "epoch": 0.5139480555173835, + "grad_norm": 1.059940218925476, + "learning_rate": 2.8139755695501814e-05, + "loss": 0.9747, + "step": 7480 + }, + { + "epoch": 0.5140167651504741, + "grad_norm": 1.0789495706558228, + "learning_rate": 2.8133806761611702e-05, + "loss": 0.9571, + "step": 7481 + }, + { + "epoch": 0.5140854747835647, + "grad_norm": 1.1006040573120117, + "learning_rate": 2.812785764743561e-05, + "loss": 0.975, + "step": 7482 + }, + { + "epoch": 0.5141541844166552, + "grad_norm": 1.0077242851257324, + "learning_rate": 2.8121908353315766e-05, + "loss": 0.9627, + "step": 7483 + }, + { + "epoch": 0.5142228940497457, + "grad_norm": 1.0238744020462036, + "learning_rate": 2.8115958879594444e-05, + "loss": 0.9737, + "step": 7484 + }, + { + "epoch": 0.5142916036828363, + "grad_norm": 1.1355983018875122, + "learning_rate": 2.811000922661391e-05, + "loss": 0.9853, + "step": 7485 + }, + { + "epoch": 0.5143603133159269, + "grad_norm": 1.0579218864440918, + "learning_rate": 2.810405939471643e-05, + "loss": 0.9737, + "step": 7486 + }, + { + "epoch": 0.5144290229490175, + "grad_norm": 1.0291119813919067, + "learning_rate": 2.809810938424432e-05, + "loss": 0.9896, + "step": 7487 + }, + { + "epoch": 0.514497732582108, + "grad_norm": 1.0928064584732056, + "learning_rate": 2.8092159195539853e-05, + "loss": 1.002, + "step": 7488 + }, + { + "epoch": 0.5145664422151985, + "grad_norm": 1.0246381759643555, + "learning_rate": 2.808620882894536e-05, + "loss": 0.938, + "step": 7489 + }, + { + "epoch": 0.5146351518482891, + "grad_norm": 1.1433031558990479, + "learning_rate": 2.8080258284803158e-05, + "loss": 0.8985, + "step": 7490 + }, + { + "epoch": 0.5147038614813797, + "grad_norm": 1.0804330110549927, + "learning_rate": 2.8074307563455572e-05, + "loss": 0.9819, + "step": 7491 + }, + { + "epoch": 0.5147725711144703, + "grad_norm": 1.0559426546096802, + "learning_rate": 2.8068356665244945e-05, + "loss": 0.9769, + "step": 7492 + }, + { + "epoch": 0.5148412807475609, + "grad_norm": 1.0428307056427002, + "learning_rate": 2.8062405590513625e-05, + "loss": 0.9671, + "step": 7493 + }, + { + "epoch": 0.5149099903806513, + "grad_norm": 1.0959681272506714, + "learning_rate": 2.805645433960398e-05, + "loss": 0.9455, + "step": 7494 + }, + { + "epoch": 0.5149787000137419, + "grad_norm": 1.0521819591522217, + "learning_rate": 2.805050291285839e-05, + "loss": 0.9485, + "step": 7495 + }, + { + "epoch": 0.5150474096468325, + "grad_norm": 1.0535569190979004, + "learning_rate": 2.8044551310619215e-05, + "loss": 0.9269, + "step": 7496 + }, + { + "epoch": 0.5151161192799231, + "grad_norm": 1.0671330690383911, + "learning_rate": 2.8038599533228867e-05, + "loss": 0.9574, + "step": 7497 + }, + { + "epoch": 0.5151848289130136, + "grad_norm": 1.0582090616226196, + "learning_rate": 2.8032647581029743e-05, + "loss": 0.9796, + "step": 7498 + }, + { + "epoch": 0.5152535385461041, + "grad_norm": 1.0656901597976685, + "learning_rate": 2.8026695454364245e-05, + "loss": 0.9694, + "step": 7499 + }, + { + "epoch": 0.5153222481791947, + "grad_norm": 1.057648777961731, + "learning_rate": 2.8020743153574807e-05, + "loss": 0.9636, + "step": 7500 + }, + { + "epoch": 0.5153909578122853, + "grad_norm": 1.1232863664627075, + "learning_rate": 2.801479067900386e-05, + "loss": 0.9676, + "step": 7501 + }, + { + "epoch": 0.5154596674453759, + "grad_norm": 1.0750466585159302, + "learning_rate": 2.800883803099384e-05, + "loss": 0.9391, + "step": 7502 + }, + { + "epoch": 0.5155283770784664, + "grad_norm": 1.0299639701843262, + "learning_rate": 2.8002885209887207e-05, + "loss": 0.9481, + "step": 7503 + }, + { + "epoch": 0.5155970867115569, + "grad_norm": 1.0155854225158691, + "learning_rate": 2.7996932216026413e-05, + "loss": 0.9657, + "step": 7504 + }, + { + "epoch": 0.5156657963446475, + "grad_norm": 1.030240774154663, + "learning_rate": 2.799097904975394e-05, + "loss": 0.9658, + "step": 7505 + }, + { + "epoch": 0.5157345059777381, + "grad_norm": 1.1565173864364624, + "learning_rate": 2.798502571141226e-05, + "loss": 1.0189, + "step": 7506 + }, + { + "epoch": 0.5158032156108286, + "grad_norm": 1.11015784740448, + "learning_rate": 2.7979072201343877e-05, + "loss": 0.9904, + "step": 7507 + }, + { + "epoch": 0.5158719252439192, + "grad_norm": 1.0056989192962646, + "learning_rate": 2.7973118519891283e-05, + "loss": 0.9212, + "step": 7508 + }, + { + "epoch": 0.5159406348770098, + "grad_norm": 1.108483076095581, + "learning_rate": 2.7967164667396995e-05, + "loss": 0.9651, + "step": 7509 + }, + { + "epoch": 0.5160093445101003, + "grad_norm": 1.0895555019378662, + "learning_rate": 2.796121064420354e-05, + "loss": 0.9906, + "step": 7510 + }, + { + "epoch": 0.5160780541431909, + "grad_norm": 1.0651962757110596, + "learning_rate": 2.7955256450653427e-05, + "loss": 0.9522, + "step": 7511 + }, + { + "epoch": 0.5161467637762814, + "grad_norm": 1.0372774600982666, + "learning_rate": 2.7949302087089223e-05, + "loss": 0.9498, + "step": 7512 + }, + { + "epoch": 0.516215473409372, + "grad_norm": 1.110507607460022, + "learning_rate": 2.7943347553853468e-05, + "loss": 0.9774, + "step": 7513 + }, + { + "epoch": 0.5162841830424626, + "grad_norm": 1.1223771572113037, + "learning_rate": 2.7937392851288725e-05, + "loss": 0.9848, + "step": 7514 + }, + { + "epoch": 0.5163528926755531, + "grad_norm": 1.053696870803833, + "learning_rate": 2.7931437979737558e-05, + "loss": 0.9642, + "step": 7515 + }, + { + "epoch": 0.5164216023086436, + "grad_norm": 1.0687577724456787, + "learning_rate": 2.7925482939542548e-05, + "loss": 0.9931, + "step": 7516 + }, + { + "epoch": 0.5164903119417342, + "grad_norm": 1.0445836782455444, + "learning_rate": 2.791952773104629e-05, + "loss": 0.9078, + "step": 7517 + }, + { + "epoch": 0.5165590215748248, + "grad_norm": 1.0886160135269165, + "learning_rate": 2.7913572354591383e-05, + "loss": 1.013, + "step": 7518 + }, + { + "epoch": 0.5166277312079154, + "grad_norm": 1.0303419828414917, + "learning_rate": 2.7907616810520437e-05, + "loss": 0.9561, + "step": 7519 + }, + { + "epoch": 0.5166964408410059, + "grad_norm": 1.1543431282043457, + "learning_rate": 2.790166109917607e-05, + "loss": 0.9987, + "step": 7520 + }, + { + "epoch": 0.5167651504740964, + "grad_norm": 1.160888910293579, + "learning_rate": 2.7895705220900902e-05, + "loss": 0.9437, + "step": 7521 + }, + { + "epoch": 0.516833860107187, + "grad_norm": 1.039186954498291, + "learning_rate": 2.7889749176037576e-05, + "loss": 0.9551, + "step": 7522 + }, + { + "epoch": 0.5169025697402776, + "grad_norm": 1.0223156213760376, + "learning_rate": 2.7883792964928756e-05, + "loss": 0.9503, + "step": 7523 + }, + { + "epoch": 0.5169712793733682, + "grad_norm": 1.0852652788162231, + "learning_rate": 2.7877836587917072e-05, + "loss": 0.9643, + "step": 7524 + }, + { + "epoch": 0.5170399890064588, + "grad_norm": 1.086021065711975, + "learning_rate": 2.7871880045345213e-05, + "loss": 0.9442, + "step": 7525 + }, + { + "epoch": 0.5171086986395492, + "grad_norm": 1.0907257795333862, + "learning_rate": 2.7865923337555843e-05, + "loss": 0.9704, + "step": 7526 + }, + { + "epoch": 0.5171774082726398, + "grad_norm": 1.0635186433792114, + "learning_rate": 2.7859966464891652e-05, + "loss": 0.9291, + "step": 7527 + }, + { + "epoch": 0.5172461179057304, + "grad_norm": 1.0812859535217285, + "learning_rate": 2.7854009427695334e-05, + "loss": 0.9173, + "step": 7528 + }, + { + "epoch": 0.517314827538821, + "grad_norm": 1.0403494834899902, + "learning_rate": 2.784805222630959e-05, + "loss": 0.9389, + "step": 7529 + }, + { + "epoch": 0.5173835371719115, + "grad_norm": 1.096888542175293, + "learning_rate": 2.784209486107714e-05, + "loss": 0.9405, + "step": 7530 + }, + { + "epoch": 0.517452246805002, + "grad_norm": 1.0462820529937744, + "learning_rate": 2.7836137332340706e-05, + "loss": 0.9464, + "step": 7531 + }, + { + "epoch": 0.5175209564380926, + "grad_norm": 1.145092248916626, + "learning_rate": 2.7830179640443017e-05, + "loss": 0.9771, + "step": 7532 + }, + { + "epoch": 0.5175896660711832, + "grad_norm": 1.0173358917236328, + "learning_rate": 2.782422178572682e-05, + "loss": 0.9412, + "step": 7533 + }, + { + "epoch": 0.5176583757042738, + "grad_norm": 1.1239732503890991, + "learning_rate": 2.781826376853487e-05, + "loss": 1.0199, + "step": 7534 + }, + { + "epoch": 0.5177270853373643, + "grad_norm": 1.0327202081680298, + "learning_rate": 2.781230558920992e-05, + "loss": 0.9202, + "step": 7535 + }, + { + "epoch": 0.5177957949704548, + "grad_norm": 1.1344575881958008, + "learning_rate": 2.780634724809475e-05, + "loss": 0.9692, + "step": 7536 + }, + { + "epoch": 0.5178645046035454, + "grad_norm": 1.0883558988571167, + "learning_rate": 2.780038874553213e-05, + "loss": 0.9254, + "step": 7537 + }, + { + "epoch": 0.517933214236636, + "grad_norm": 1.0378873348236084, + "learning_rate": 2.779443008186485e-05, + "loss": 0.9407, + "step": 7538 + }, + { + "epoch": 0.5180019238697265, + "grad_norm": 1.0068684816360474, + "learning_rate": 2.7788471257435716e-05, + "loss": 0.9619, + "step": 7539 + }, + { + "epoch": 0.5180706335028171, + "grad_norm": 1.136466383934021, + "learning_rate": 2.7782512272587525e-05, + "loss": 1.0182, + "step": 7540 + }, + { + "epoch": 0.5181393431359077, + "grad_norm": 1.1144490242004395, + "learning_rate": 2.7776553127663107e-05, + "loss": 1.0106, + "step": 7541 + }, + { + "epoch": 0.5182080527689982, + "grad_norm": 1.1038891077041626, + "learning_rate": 2.7770593823005274e-05, + "loss": 0.9877, + "step": 7542 + }, + { + "epoch": 0.5182767624020888, + "grad_norm": 1.0711861848831177, + "learning_rate": 2.7764634358956877e-05, + "loss": 1.0003, + "step": 7543 + }, + { + "epoch": 0.5183454720351793, + "grad_norm": 1.1250014305114746, + "learning_rate": 2.775867473586074e-05, + "loss": 0.9971, + "step": 7544 + }, + { + "epoch": 0.5184141816682699, + "grad_norm": 1.1496942043304443, + "learning_rate": 2.775271495405973e-05, + "loss": 1.0239, + "step": 7545 + }, + { + "epoch": 0.5184828913013605, + "grad_norm": 1.0942723751068115, + "learning_rate": 2.7746755013896714e-05, + "loss": 0.9511, + "step": 7546 + }, + { + "epoch": 0.518551600934451, + "grad_norm": 1.1106055974960327, + "learning_rate": 2.7740794915714547e-05, + "loss": 0.9741, + "step": 7547 + }, + { + "epoch": 0.5186203105675415, + "grad_norm": 1.1031348705291748, + "learning_rate": 2.773483465985613e-05, + "loss": 0.945, + "step": 7548 + }, + { + "epoch": 0.5186890202006321, + "grad_norm": 1.0921345949172974, + "learning_rate": 2.772887424666434e-05, + "loss": 0.9482, + "step": 7549 + }, + { + "epoch": 0.5187577298337227, + "grad_norm": 1.0164765119552612, + "learning_rate": 2.772291367648207e-05, + "loss": 0.9438, + "step": 7550 + }, + { + "epoch": 0.5188264394668133, + "grad_norm": 1.0695929527282715, + "learning_rate": 2.7716952949652247e-05, + "loss": 1.0137, + "step": 7551 + }, + { + "epoch": 0.5188951490999038, + "grad_norm": 1.1154944896697998, + "learning_rate": 2.7710992066517766e-05, + "loss": 0.9927, + "step": 7552 + }, + { + "epoch": 0.5189638587329943, + "grad_norm": 1.0319559574127197, + "learning_rate": 2.7705031027421568e-05, + "loss": 0.8943, + "step": 7553 + }, + { + "epoch": 0.5190325683660849, + "grad_norm": 1.0873311758041382, + "learning_rate": 2.7699069832706594e-05, + "loss": 1.0144, + "step": 7554 + }, + { + "epoch": 0.5191012779991755, + "grad_norm": 1.0773563385009766, + "learning_rate": 2.769310848271577e-05, + "loss": 0.9383, + "step": 7555 + }, + { + "epoch": 0.5191699876322661, + "grad_norm": 1.0267608165740967, + "learning_rate": 2.7687146977792056e-05, + "loss": 0.9412, + "step": 7556 + }, + { + "epoch": 0.5192386972653567, + "grad_norm": 1.063597559928894, + "learning_rate": 2.7681185318278418e-05, + "loss": 0.9556, + "step": 7557 + }, + { + "epoch": 0.5193074068984471, + "grad_norm": 1.0335419178009033, + "learning_rate": 2.7675223504517823e-05, + "loss": 0.9535, + "step": 7558 + }, + { + "epoch": 0.5193761165315377, + "grad_norm": 1.1167609691619873, + "learning_rate": 2.7669261536853252e-05, + "loss": 1.0098, + "step": 7559 + }, + { + "epoch": 0.5194448261646283, + "grad_norm": 1.108703374862671, + "learning_rate": 2.766329941562769e-05, + "loss": 0.9911, + "step": 7560 + }, + { + "epoch": 0.5195135357977189, + "grad_norm": 1.1277403831481934, + "learning_rate": 2.7657337141184138e-05, + "loss": 0.9652, + "step": 7561 + }, + { + "epoch": 0.5195822454308094, + "grad_norm": 1.068960428237915, + "learning_rate": 2.7651374713865602e-05, + "loss": 0.9955, + "step": 7562 + }, + { + "epoch": 0.5196509550638999, + "grad_norm": 1.0726022720336914, + "learning_rate": 2.764541213401509e-05, + "loss": 0.9969, + "step": 7563 + }, + { + "epoch": 0.5197196646969905, + "grad_norm": 1.1193702220916748, + "learning_rate": 2.7639449401975635e-05, + "loss": 1.0007, + "step": 7564 + }, + { + "epoch": 0.5197883743300811, + "grad_norm": 1.1419661045074463, + "learning_rate": 2.763348651809026e-05, + "loss": 0.9606, + "step": 7565 + }, + { + "epoch": 0.5198570839631717, + "grad_norm": 1.073000192642212, + "learning_rate": 2.7627523482702022e-05, + "loss": 0.9554, + "step": 7566 + }, + { + "epoch": 0.5199257935962622, + "grad_norm": 1.0326755046844482, + "learning_rate": 2.7621560296153954e-05, + "loss": 0.9461, + "step": 7567 + }, + { + "epoch": 0.5199945032293527, + "grad_norm": 1.0848389863967896, + "learning_rate": 2.761559695878912e-05, + "loss": 0.9406, + "step": 7568 + }, + { + "epoch": 0.5200632128624433, + "grad_norm": 1.0298765897750854, + "learning_rate": 2.760963347095059e-05, + "loss": 0.9447, + "step": 7569 + }, + { + "epoch": 0.5201319224955339, + "grad_norm": 1.0076769590377808, + "learning_rate": 2.7603669832981433e-05, + "loss": 0.9379, + "step": 7570 + }, + { + "epoch": 0.5202006321286244, + "grad_norm": 1.0367310047149658, + "learning_rate": 2.759770604522474e-05, + "loss": 0.9759, + "step": 7571 + }, + { + "epoch": 0.520269341761715, + "grad_norm": 1.010219931602478, + "learning_rate": 2.7591742108023606e-05, + "loss": 0.9924, + "step": 7572 + }, + { + "epoch": 0.5203380513948056, + "grad_norm": 1.0188348293304443, + "learning_rate": 2.7585778021721126e-05, + "loss": 0.9476, + "step": 7573 + }, + { + "epoch": 0.5204067610278961, + "grad_norm": 1.0577579736709595, + "learning_rate": 2.757981378666042e-05, + "loss": 0.9609, + "step": 7574 + }, + { + "epoch": 0.5204754706609866, + "grad_norm": 1.0875767469406128, + "learning_rate": 2.7573849403184587e-05, + "loss": 0.9893, + "step": 7575 + }, + { + "epoch": 0.5205441802940772, + "grad_norm": 1.109561800956726, + "learning_rate": 2.7567884871636778e-05, + "loss": 0.9736, + "step": 7576 + }, + { + "epoch": 0.5206128899271678, + "grad_norm": 1.0284112691879272, + "learning_rate": 2.7561920192360107e-05, + "loss": 0.9632, + "step": 7577 + }, + { + "epoch": 0.5206815995602584, + "grad_norm": 1.043914556503296, + "learning_rate": 2.7555955365697732e-05, + "loss": 0.934, + "step": 7578 + }, + { + "epoch": 0.5207503091933489, + "grad_norm": 1.05775785446167, + "learning_rate": 2.7549990391992812e-05, + "loss": 0.9799, + "step": 7579 + }, + { + "epoch": 0.5208190188264394, + "grad_norm": 1.058459758758545, + "learning_rate": 2.754402527158849e-05, + "loss": 0.9429, + "step": 7580 + }, + { + "epoch": 0.52088772845953, + "grad_norm": 1.0916706323623657, + "learning_rate": 2.753806000482795e-05, + "loss": 1.0228, + "step": 7581 + }, + { + "epoch": 0.5209564380926206, + "grad_norm": 1.024909257888794, + "learning_rate": 2.753209459205437e-05, + "loss": 0.9481, + "step": 7582 + }, + { + "epoch": 0.5210251477257112, + "grad_norm": 1.099428415298462, + "learning_rate": 2.752612903361092e-05, + "loss": 0.9658, + "step": 7583 + }, + { + "epoch": 0.5210938573588016, + "grad_norm": 1.0265462398529053, + "learning_rate": 2.752016332984082e-05, + "loss": 0.9342, + "step": 7584 + }, + { + "epoch": 0.5211625669918922, + "grad_norm": 1.1320050954818726, + "learning_rate": 2.7514197481087257e-05, + "loss": 0.9815, + "step": 7585 + }, + { + "epoch": 0.5212312766249828, + "grad_norm": 1.056722640991211, + "learning_rate": 2.7508231487693443e-05, + "loss": 0.964, + "step": 7586 + }, + { + "epoch": 0.5212999862580734, + "grad_norm": 1.0590628385543823, + "learning_rate": 2.7502265350002603e-05, + "loss": 0.9661, + "step": 7587 + }, + { + "epoch": 0.521368695891164, + "grad_norm": 1.0753024816513062, + "learning_rate": 2.7496299068357955e-05, + "loss": 0.964, + "step": 7588 + }, + { + "epoch": 0.5214374055242545, + "grad_norm": 1.0005970001220703, + "learning_rate": 2.7490332643102756e-05, + "loss": 0.9261, + "step": 7589 + }, + { + "epoch": 0.521506115157345, + "grad_norm": 1.0780255794525146, + "learning_rate": 2.7484366074580232e-05, + "loss": 0.9625, + "step": 7590 + }, + { + "epoch": 0.5215748247904356, + "grad_norm": 1.0579122304916382, + "learning_rate": 2.7478399363133638e-05, + "loss": 0.9256, + "step": 7591 + }, + { + "epoch": 0.5216435344235262, + "grad_norm": 1.1233505010604858, + "learning_rate": 2.7472432509106248e-05, + "loss": 0.9877, + "step": 7592 + }, + { + "epoch": 0.5217122440566168, + "grad_norm": 1.0629284381866455, + "learning_rate": 2.7466465512841315e-05, + "loss": 0.9731, + "step": 7593 + }, + { + "epoch": 0.5217809536897073, + "grad_norm": 1.0400846004486084, + "learning_rate": 2.7460498374682138e-05, + "loss": 0.9383, + "step": 7594 + }, + { + "epoch": 0.5218496633227978, + "grad_norm": 1.0418832302093506, + "learning_rate": 2.7454531094971974e-05, + "loss": 0.9479, + "step": 7595 + }, + { + "epoch": 0.5219183729558884, + "grad_norm": 1.0336310863494873, + "learning_rate": 2.7448563674054144e-05, + "loss": 0.9578, + "step": 7596 + }, + { + "epoch": 0.521987082588979, + "grad_norm": 1.027681827545166, + "learning_rate": 2.744259611227194e-05, + "loss": 0.9305, + "step": 7597 + }, + { + "epoch": 0.5220557922220695, + "grad_norm": 1.052764892578125, + "learning_rate": 2.7436628409968664e-05, + "loss": 0.9389, + "step": 7598 + }, + { + "epoch": 0.5221245018551601, + "grad_norm": 1.0343621969223022, + "learning_rate": 2.743066056748765e-05, + "loss": 0.9619, + "step": 7599 + }, + { + "epoch": 0.5221932114882507, + "grad_norm": 1.0922186374664307, + "learning_rate": 2.74246925851722e-05, + "loss": 0.963, + "step": 7600 + }, + { + "epoch": 0.5222619211213412, + "grad_norm": 1.0794005393981934, + "learning_rate": 2.7418724463365675e-05, + "loss": 1.0028, + "step": 7601 + }, + { + "epoch": 0.5223306307544318, + "grad_norm": 1.1324797868728638, + "learning_rate": 2.7412756202411404e-05, + "loss": 0.9629, + "step": 7602 + }, + { + "epoch": 0.5223993403875223, + "grad_norm": 1.0543409585952759, + "learning_rate": 2.740678780265274e-05, + "loss": 0.9581, + "step": 7603 + }, + { + "epoch": 0.5224680500206129, + "grad_norm": 1.039689064025879, + "learning_rate": 2.7400819264433043e-05, + "loss": 0.8985, + "step": 7604 + }, + { + "epoch": 0.5225367596537035, + "grad_norm": 1.0436605215072632, + "learning_rate": 2.7394850588095677e-05, + "loss": 0.943, + "step": 7605 + }, + { + "epoch": 0.522605469286794, + "grad_norm": 1.0792533159255981, + "learning_rate": 2.738888177398401e-05, + "loss": 0.927, + "step": 7606 + }, + { + "epoch": 0.5226741789198845, + "grad_norm": 1.060821533203125, + "learning_rate": 2.738291282244144e-05, + "loss": 0.9453, + "step": 7607 + }, + { + "epoch": 0.5227428885529751, + "grad_norm": 1.1144564151763916, + "learning_rate": 2.737694373381135e-05, + "loss": 0.9775, + "step": 7608 + }, + { + "epoch": 0.5228115981860657, + "grad_norm": 1.096540093421936, + "learning_rate": 2.7370974508437132e-05, + "loss": 0.9742, + "step": 7609 + }, + { + "epoch": 0.5228803078191563, + "grad_norm": 1.017017126083374, + "learning_rate": 2.7365005146662193e-05, + "loss": 0.9468, + "step": 7610 + }, + { + "epoch": 0.5229490174522468, + "grad_norm": 1.0505653619766235, + "learning_rate": 2.735903564882995e-05, + "loss": 0.9392, + "step": 7611 + }, + { + "epoch": 0.5230177270853373, + "grad_norm": 1.0249953269958496, + "learning_rate": 2.7353066015283833e-05, + "loss": 0.998, + "step": 7612 + }, + { + "epoch": 0.5230864367184279, + "grad_norm": 1.0708519220352173, + "learning_rate": 2.7347096246367254e-05, + "loss": 0.9525, + "step": 7613 + }, + { + "epoch": 0.5231551463515185, + "grad_norm": 1.1096824407577515, + "learning_rate": 2.7341126342423663e-05, + "loss": 1.0117, + "step": 7614 + }, + { + "epoch": 0.5232238559846091, + "grad_norm": 1.026315450668335, + "learning_rate": 2.73351563037965e-05, + "loss": 0.9221, + "step": 7615 + }, + { + "epoch": 0.5232925656176997, + "grad_norm": 1.0876820087432861, + "learning_rate": 2.7329186130829217e-05, + "loss": 1.0135, + "step": 7616 + }, + { + "epoch": 0.5233612752507901, + "grad_norm": 1.1069923639297485, + "learning_rate": 2.732321582386528e-05, + "loss": 0.9633, + "step": 7617 + }, + { + "epoch": 0.5234299848838807, + "grad_norm": 1.0487842559814453, + "learning_rate": 2.7317245383248148e-05, + "loss": 0.9282, + "step": 7618 + }, + { + "epoch": 0.5234986945169713, + "grad_norm": 1.0277786254882812, + "learning_rate": 2.731127480932131e-05, + "loss": 0.9366, + "step": 7619 + }, + { + "epoch": 0.5235674041500619, + "grad_norm": 1.0435962677001953, + "learning_rate": 2.730530410242824e-05, + "loss": 0.9429, + "step": 7620 + }, + { + "epoch": 0.5236361137831524, + "grad_norm": 1.1026619672775269, + "learning_rate": 2.7299333262912425e-05, + "loss": 0.9848, + "step": 7621 + }, + { + "epoch": 0.5237048234162429, + "grad_norm": 1.0580893754959106, + "learning_rate": 2.7293362291117375e-05, + "loss": 0.9209, + "step": 7622 + }, + { + "epoch": 0.5237735330493335, + "grad_norm": 1.0974929332733154, + "learning_rate": 2.7287391187386586e-05, + "loss": 0.9382, + "step": 7623 + }, + { + "epoch": 0.5238422426824241, + "grad_norm": 1.0230038166046143, + "learning_rate": 2.7281419952063574e-05, + "loss": 0.9736, + "step": 7624 + }, + { + "epoch": 0.5239109523155147, + "grad_norm": 1.081688404083252, + "learning_rate": 2.7275448585491868e-05, + "loss": 0.9517, + "step": 7625 + }, + { + "epoch": 0.5239796619486052, + "grad_norm": 1.072648525238037, + "learning_rate": 2.7269477088014988e-05, + "loss": 0.9783, + "step": 7626 + }, + { + "epoch": 0.5240483715816957, + "grad_norm": 1.1692286729812622, + "learning_rate": 2.7263505459976486e-05, + "loss": 0.9732, + "step": 7627 + }, + { + "epoch": 0.5241170812147863, + "grad_norm": 1.041699767112732, + "learning_rate": 2.7257533701719883e-05, + "loss": 0.9781, + "step": 7628 + }, + { + "epoch": 0.5241857908478769, + "grad_norm": 1.0671781301498413, + "learning_rate": 2.7251561813588744e-05, + "loss": 0.9771, + "step": 7629 + }, + { + "epoch": 0.5242545004809674, + "grad_norm": 1.1202960014343262, + "learning_rate": 2.7245589795926634e-05, + "loss": 0.9569, + "step": 7630 + }, + { + "epoch": 0.524323210114058, + "grad_norm": 1.039629578590393, + "learning_rate": 2.7239617649077108e-05, + "loss": 0.9187, + "step": 7631 + }, + { + "epoch": 0.5243919197471486, + "grad_norm": 1.0562766790390015, + "learning_rate": 2.7233645373383744e-05, + "loss": 0.9344, + "step": 7632 + }, + { + "epoch": 0.5244606293802391, + "grad_norm": 1.0135926008224487, + "learning_rate": 2.7227672969190128e-05, + "loss": 0.9598, + "step": 7633 + }, + { + "epoch": 0.5245293390133297, + "grad_norm": 1.0459957122802734, + "learning_rate": 2.7221700436839835e-05, + "loss": 0.9614, + "step": 7634 + }, + { + "epoch": 0.5245980486464202, + "grad_norm": 1.0747126340866089, + "learning_rate": 2.721572777667648e-05, + "loss": 0.979, + "step": 7635 + }, + { + "epoch": 0.5246667582795108, + "grad_norm": 1.0977506637573242, + "learning_rate": 2.720975498904365e-05, + "loss": 0.9639, + "step": 7636 + }, + { + "epoch": 0.5247354679126014, + "grad_norm": 1.069119930267334, + "learning_rate": 2.720378207428496e-05, + "loss": 0.9578, + "step": 7637 + }, + { + "epoch": 0.5248041775456919, + "grad_norm": 1.0980448722839355, + "learning_rate": 2.7197809032744044e-05, + "loss": 0.9976, + "step": 7638 + }, + { + "epoch": 0.5248728871787824, + "grad_norm": 1.080407977104187, + "learning_rate": 2.7191835864764503e-05, + "loss": 0.9366, + "step": 7639 + }, + { + "epoch": 0.524941596811873, + "grad_norm": 1.0292913913726807, + "learning_rate": 2.718586257068999e-05, + "loss": 0.9581, + "step": 7640 + }, + { + "epoch": 0.5250103064449636, + "grad_norm": 1.0664111375808716, + "learning_rate": 2.7179889150864124e-05, + "loss": 0.9538, + "step": 7641 + }, + { + "epoch": 0.5250790160780542, + "grad_norm": 1.0769983530044556, + "learning_rate": 2.7173915605630572e-05, + "loss": 0.9818, + "step": 7642 + }, + { + "epoch": 0.5251477257111447, + "grad_norm": 1.0943056344985962, + "learning_rate": 2.7167941935332984e-05, + "loss": 0.9625, + "step": 7643 + }, + { + "epoch": 0.5252164353442352, + "grad_norm": 1.0047563314437866, + "learning_rate": 2.716196814031502e-05, + "loss": 0.949, + "step": 7644 + }, + { + "epoch": 0.5252851449773258, + "grad_norm": 1.1019190549850464, + "learning_rate": 2.7155994220920337e-05, + "loss": 1.0226, + "step": 7645 + }, + { + "epoch": 0.5253538546104164, + "grad_norm": 1.098685383796692, + "learning_rate": 2.715002017749263e-05, + "loss": 0.9793, + "step": 7646 + }, + { + "epoch": 0.525422564243507, + "grad_norm": 0.9895489811897278, + "learning_rate": 2.7144046010375563e-05, + "loss": 0.9433, + "step": 7647 + }, + { + "epoch": 0.5254912738765976, + "grad_norm": 1.1046817302703857, + "learning_rate": 2.7138071719912846e-05, + "loss": 0.9501, + "step": 7648 + }, + { + "epoch": 0.525559983509688, + "grad_norm": 1.0405646562576294, + "learning_rate": 2.713209730644816e-05, + "loss": 0.9512, + "step": 7649 + }, + { + "epoch": 0.5256286931427786, + "grad_norm": 1.1114330291748047, + "learning_rate": 2.7126122770325225e-05, + "loss": 0.9661, + "step": 7650 + }, + { + "epoch": 0.5256974027758692, + "grad_norm": 1.1149758100509644, + "learning_rate": 2.7120148111887732e-05, + "loss": 0.9774, + "step": 7651 + }, + { + "epoch": 0.5257661124089598, + "grad_norm": 1.0256524085998535, + "learning_rate": 2.7114173331479413e-05, + "loss": 0.9616, + "step": 7652 + }, + { + "epoch": 0.5258348220420503, + "grad_norm": 1.0942522287368774, + "learning_rate": 2.7108198429443997e-05, + "loss": 1.0088, + "step": 7653 + }, + { + "epoch": 0.5259035316751408, + "grad_norm": 1.0245615243911743, + "learning_rate": 2.7102223406125206e-05, + "loss": 0.9848, + "step": 7654 + }, + { + "epoch": 0.5259722413082314, + "grad_norm": 1.1330559253692627, + "learning_rate": 2.7096248261866793e-05, + "loss": 0.9782, + "step": 7655 + }, + { + "epoch": 0.526040950941322, + "grad_norm": 1.1345901489257812, + "learning_rate": 2.7090272997012488e-05, + "loss": 0.9912, + "step": 7656 + }, + { + "epoch": 0.5261096605744126, + "grad_norm": 1.0512765645980835, + "learning_rate": 2.708429761190605e-05, + "loss": 0.9055, + "step": 7657 + }, + { + "epoch": 0.5261783702075031, + "grad_norm": 1.062636375427246, + "learning_rate": 2.7078322106891246e-05, + "loss": 0.9933, + "step": 7658 + }, + { + "epoch": 0.5262470798405936, + "grad_norm": 1.0470505952835083, + "learning_rate": 2.707234648231184e-05, + "loss": 0.9211, + "step": 7659 + }, + { + "epoch": 0.5263157894736842, + "grad_norm": 1.0417665243148804, + "learning_rate": 2.7066370738511598e-05, + "loss": 0.9397, + "step": 7660 + }, + { + "epoch": 0.5263844991067748, + "grad_norm": 1.103659987449646, + "learning_rate": 2.706039487583431e-05, + "loss": 1.0032, + "step": 7661 + }, + { + "epoch": 0.5264532087398653, + "grad_norm": 1.0437109470367432, + "learning_rate": 2.7054418894623757e-05, + "loss": 0.945, + "step": 7662 + }, + { + "epoch": 0.5265219183729559, + "grad_norm": 1.042004942893982, + "learning_rate": 2.704844279522375e-05, + "loss": 0.9553, + "step": 7663 + }, + { + "epoch": 0.5265906280060465, + "grad_norm": 1.0803011655807495, + "learning_rate": 2.7042466577978066e-05, + "loss": 0.9568, + "step": 7664 + }, + { + "epoch": 0.526659337639137, + "grad_norm": 1.014916181564331, + "learning_rate": 2.703649024323053e-05, + "loss": 0.9187, + "step": 7665 + }, + { + "epoch": 0.5267280472722276, + "grad_norm": 1.07234525680542, + "learning_rate": 2.7030513791324953e-05, + "loss": 0.9486, + "step": 7666 + }, + { + "epoch": 0.5267967569053181, + "grad_norm": 1.1357519626617432, + "learning_rate": 2.7024537222605158e-05, + "loss": 0.9735, + "step": 7667 + }, + { + "epoch": 0.5268654665384087, + "grad_norm": 1.0731593370437622, + "learning_rate": 2.7018560537414968e-05, + "loss": 0.9644, + "step": 7668 + }, + { + "epoch": 0.5269341761714993, + "grad_norm": 1.0714004039764404, + "learning_rate": 2.7012583736098218e-05, + "loss": 0.9375, + "step": 7669 + }, + { + "epoch": 0.5270028858045898, + "grad_norm": 1.0715786218643188, + "learning_rate": 2.7006606818998748e-05, + "loss": 0.9587, + "step": 7670 + }, + { + "epoch": 0.5270715954376803, + "grad_norm": 1.0582139492034912, + "learning_rate": 2.7000629786460425e-05, + "loss": 0.9542, + "step": 7671 + }, + { + "epoch": 0.5271403050707709, + "grad_norm": 1.095890998840332, + "learning_rate": 2.6994652638827078e-05, + "loss": 0.9751, + "step": 7672 + }, + { + "epoch": 0.5272090147038615, + "grad_norm": 1.1243690252304077, + "learning_rate": 2.698867537644259e-05, + "loss": 0.9819, + "step": 7673 + }, + { + "epoch": 0.5272777243369521, + "grad_norm": 1.0648304224014282, + "learning_rate": 2.6982697999650812e-05, + "loss": 0.9549, + "step": 7674 + }, + { + "epoch": 0.5273464339700425, + "grad_norm": 1.1024994850158691, + "learning_rate": 2.6976720508795632e-05, + "loss": 0.9361, + "step": 7675 + }, + { + "epoch": 0.5274151436031331, + "grad_norm": 1.0418369770050049, + "learning_rate": 2.6970742904220926e-05, + "loss": 0.9647, + "step": 7676 + }, + { + "epoch": 0.5274838532362237, + "grad_norm": 1.094261646270752, + "learning_rate": 2.696476518627058e-05, + "loss": 0.9738, + "step": 7677 + }, + { + "epoch": 0.5275525628693143, + "grad_norm": 1.0629663467407227, + "learning_rate": 2.6958787355288496e-05, + "loss": 0.9342, + "step": 7678 + }, + { + "epoch": 0.5276212725024049, + "grad_norm": 1.0278708934783936, + "learning_rate": 2.695280941161857e-05, + "loss": 0.9341, + "step": 7679 + }, + { + "epoch": 0.5276899821354955, + "grad_norm": 1.0861517190933228, + "learning_rate": 2.6946831355604708e-05, + "loss": 0.9591, + "step": 7680 + }, + { + "epoch": 0.5277586917685859, + "grad_norm": 1.0421713590621948, + "learning_rate": 2.694085318759082e-05, + "loss": 0.924, + "step": 7681 + }, + { + "epoch": 0.5278274014016765, + "grad_norm": 1.0835312604904175, + "learning_rate": 2.693487490792083e-05, + "loss": 0.9641, + "step": 7682 + }, + { + "epoch": 0.5278961110347671, + "grad_norm": 1.087754726409912, + "learning_rate": 2.692889651693868e-05, + "loss": 0.9397, + "step": 7683 + }, + { + "epoch": 0.5279648206678577, + "grad_norm": 1.1175190210342407, + "learning_rate": 2.6922918014988274e-05, + "loss": 0.9777, + "step": 7684 + }, + { + "epoch": 0.5280335303009482, + "grad_norm": 1.1015219688415527, + "learning_rate": 2.6916939402413572e-05, + "loss": 0.9452, + "step": 7685 + }, + { + "epoch": 0.5281022399340387, + "grad_norm": 1.0609591007232666, + "learning_rate": 2.6910960679558522e-05, + "loss": 0.9715, + "step": 7686 + }, + { + "epoch": 0.5281709495671293, + "grad_norm": 1.0783839225769043, + "learning_rate": 2.690498184676706e-05, + "loss": 0.938, + "step": 7687 + }, + { + "epoch": 0.5282396592002199, + "grad_norm": 1.1161913871765137, + "learning_rate": 2.689900290438316e-05, + "loss": 0.9899, + "step": 7688 + }, + { + "epoch": 0.5283083688333104, + "grad_norm": 1.1273884773254395, + "learning_rate": 2.6893023852750777e-05, + "loss": 0.9653, + "step": 7689 + }, + { + "epoch": 0.528377078466401, + "grad_norm": 1.0777629613876343, + "learning_rate": 2.688704469221389e-05, + "loss": 0.9444, + "step": 7690 + }, + { + "epoch": 0.5284457880994915, + "grad_norm": 1.0522956848144531, + "learning_rate": 2.688106542311647e-05, + "loss": 0.9497, + "step": 7691 + }, + { + "epoch": 0.5285144977325821, + "grad_norm": 1.0472899675369263, + "learning_rate": 2.6875086045802505e-05, + "loss": 0.9001, + "step": 7692 + }, + { + "epoch": 0.5285832073656727, + "grad_norm": 0.9767271876335144, + "learning_rate": 2.6869106560615976e-05, + "loss": 0.9193, + "step": 7693 + }, + { + "epoch": 0.5286519169987632, + "grad_norm": 1.0805951356887817, + "learning_rate": 2.686312696790089e-05, + "loss": 0.9934, + "step": 7694 + }, + { + "epoch": 0.5287206266318538, + "grad_norm": 0.990277111530304, + "learning_rate": 2.6857147268001248e-05, + "loss": 0.9253, + "step": 7695 + }, + { + "epoch": 0.5287893362649444, + "grad_norm": 1.079158067703247, + "learning_rate": 2.6851167461261062e-05, + "loss": 0.9341, + "step": 7696 + }, + { + "epoch": 0.5288580458980349, + "grad_norm": 1.0894132852554321, + "learning_rate": 2.684518754802433e-05, + "loss": 0.9438, + "step": 7697 + }, + { + "epoch": 0.5289267555311254, + "grad_norm": 1.0770548582077026, + "learning_rate": 2.6839207528635086e-05, + "loss": 0.9564, + "step": 7698 + }, + { + "epoch": 0.528995465164216, + "grad_norm": 1.0039749145507812, + "learning_rate": 2.6833227403437354e-05, + "loss": 0.9555, + "step": 7699 + }, + { + "epoch": 0.5290641747973066, + "grad_norm": 1.085056185722351, + "learning_rate": 2.6827247172775166e-05, + "loss": 0.9835, + "step": 7700 + }, + { + "epoch": 0.5291328844303972, + "grad_norm": 1.0799392461776733, + "learning_rate": 2.682126683699257e-05, + "loss": 0.9598, + "step": 7701 + }, + { + "epoch": 0.5292015940634877, + "grad_norm": 1.0605113506317139, + "learning_rate": 2.6815286396433594e-05, + "loss": 0.9394, + "step": 7702 + }, + { + "epoch": 0.5292703036965782, + "grad_norm": 1.036690354347229, + "learning_rate": 2.6809305851442307e-05, + "loss": 0.9564, + "step": 7703 + }, + { + "epoch": 0.5293390133296688, + "grad_norm": 1.093849778175354, + "learning_rate": 2.6803325202362757e-05, + "loss": 0.972, + "step": 7704 + }, + { + "epoch": 0.5294077229627594, + "grad_norm": 1.0928246974945068, + "learning_rate": 2.6797344449539e-05, + "loss": 0.935, + "step": 7705 + }, + { + "epoch": 0.52947643259585, + "grad_norm": 1.0430952310562134, + "learning_rate": 2.6791363593315122e-05, + "loss": 0.9719, + "step": 7706 + }, + { + "epoch": 0.5295451422289404, + "grad_norm": 1.084763765335083, + "learning_rate": 2.6785382634035176e-05, + "loss": 0.971, + "step": 7707 + }, + { + "epoch": 0.529613851862031, + "grad_norm": 1.1128875017166138, + "learning_rate": 2.677940157204326e-05, + "loss": 0.958, + "step": 7708 + }, + { + "epoch": 0.5296825614951216, + "grad_norm": 1.0787893533706665, + "learning_rate": 2.6773420407683464e-05, + "loss": 0.96, + "step": 7709 + }, + { + "epoch": 0.5297512711282122, + "grad_norm": 1.0786528587341309, + "learning_rate": 2.6767439141299865e-05, + "loss": 0.9737, + "step": 7710 + }, + { + "epoch": 0.5298199807613028, + "grad_norm": 1.1005933284759521, + "learning_rate": 2.676145777323658e-05, + "loss": 0.9608, + "step": 7711 + }, + { + "epoch": 0.5298886903943933, + "grad_norm": 1.1208574771881104, + "learning_rate": 2.6755476303837697e-05, + "loss": 0.9659, + "step": 7712 + }, + { + "epoch": 0.5299574000274838, + "grad_norm": 1.0700502395629883, + "learning_rate": 2.674949473344733e-05, + "loss": 0.9261, + "step": 7713 + }, + { + "epoch": 0.5300261096605744, + "grad_norm": 1.0506492853164673, + "learning_rate": 2.67435130624096e-05, + "loss": 0.9827, + "step": 7714 + }, + { + "epoch": 0.530094819293665, + "grad_norm": 1.1159145832061768, + "learning_rate": 2.673753129106863e-05, + "loss": 1.0285, + "step": 7715 + }, + { + "epoch": 0.5301635289267556, + "grad_norm": 1.009621500968933, + "learning_rate": 2.6731549419768543e-05, + "loss": 0.9759, + "step": 7716 + }, + { + "epoch": 0.5302322385598461, + "grad_norm": 1.0891143083572388, + "learning_rate": 2.6725567448853472e-05, + "loss": 0.9262, + "step": 7717 + }, + { + "epoch": 0.5303009481929366, + "grad_norm": 1.011208176612854, + "learning_rate": 2.6719585378667556e-05, + "loss": 0.9375, + "step": 7718 + }, + { + "epoch": 0.5303696578260272, + "grad_norm": 1.0544613599777222, + "learning_rate": 2.6713603209554944e-05, + "loss": 0.9848, + "step": 7719 + }, + { + "epoch": 0.5304383674591178, + "grad_norm": 1.0384726524353027, + "learning_rate": 2.670762094185978e-05, + "loss": 0.9284, + "step": 7720 + }, + { + "epoch": 0.5305070770922083, + "grad_norm": 1.058451533317566, + "learning_rate": 2.6701638575926225e-05, + "loss": 0.9783, + "step": 7721 + }, + { + "epoch": 0.5305757867252989, + "grad_norm": 1.0682909488677979, + "learning_rate": 2.669565611209845e-05, + "loss": 0.9816, + "step": 7722 + }, + { + "epoch": 0.5306444963583894, + "grad_norm": 1.0059316158294678, + "learning_rate": 2.66896735507206e-05, + "loss": 0.9193, + "step": 7723 + }, + { + "epoch": 0.53071320599148, + "grad_norm": 1.116761565208435, + "learning_rate": 2.668369089213687e-05, + "loss": 0.9613, + "step": 7724 + }, + { + "epoch": 0.5307819156245706, + "grad_norm": 1.0714699029922485, + "learning_rate": 2.6677708136691425e-05, + "loss": 0.9435, + "step": 7725 + }, + { + "epoch": 0.5308506252576611, + "grad_norm": 1.0775960683822632, + "learning_rate": 2.6671725284728454e-05, + "loss": 0.9817, + "step": 7726 + }, + { + "epoch": 0.5309193348907517, + "grad_norm": 1.0477327108383179, + "learning_rate": 2.6665742336592153e-05, + "loss": 0.9591, + "step": 7727 + }, + { + "epoch": 0.5309880445238423, + "grad_norm": 1.0017237663269043, + "learning_rate": 2.6659759292626708e-05, + "loss": 0.9333, + "step": 7728 + }, + { + "epoch": 0.5310567541569328, + "grad_norm": 0.9825414419174194, + "learning_rate": 2.665377615317632e-05, + "loss": 0.9139, + "step": 7729 + }, + { + "epoch": 0.5311254637900233, + "grad_norm": 1.0277495384216309, + "learning_rate": 2.6647792918585196e-05, + "loss": 0.9238, + "step": 7730 + }, + { + "epoch": 0.5311941734231139, + "grad_norm": 1.118532419204712, + "learning_rate": 2.6641809589197552e-05, + "loss": 0.9493, + "step": 7731 + }, + { + "epoch": 0.5312628830562045, + "grad_norm": 1.0940215587615967, + "learning_rate": 2.6635826165357607e-05, + "loss": 0.9343, + "step": 7732 + }, + { + "epoch": 0.5313315926892951, + "grad_norm": 1.103316068649292, + "learning_rate": 2.6629842647409574e-05, + "loss": 0.9989, + "step": 7733 + }, + { + "epoch": 0.5314003023223856, + "grad_norm": 1.0618817806243896, + "learning_rate": 2.6623859035697694e-05, + "loss": 0.9539, + "step": 7734 + }, + { + "epoch": 0.5314690119554761, + "grad_norm": 1.1222566366195679, + "learning_rate": 2.6617875330566188e-05, + "loss": 0.979, + "step": 7735 + }, + { + "epoch": 0.5315377215885667, + "grad_norm": 1.094098448753357, + "learning_rate": 2.66118915323593e-05, + "loss": 0.9852, + "step": 7736 + }, + { + "epoch": 0.5316064312216573, + "grad_norm": 1.105728268623352, + "learning_rate": 2.6605907641421284e-05, + "loss": 0.9642, + "step": 7737 + }, + { + "epoch": 0.5316751408547479, + "grad_norm": 1.0288678407669067, + "learning_rate": 2.6599923658096382e-05, + "loss": 0.9446, + "step": 7738 + }, + { + "epoch": 0.5317438504878383, + "grad_norm": 1.0243034362792969, + "learning_rate": 2.659393958272883e-05, + "loss": 0.9286, + "step": 7739 + }, + { + "epoch": 0.5318125601209289, + "grad_norm": 1.1310614347457886, + "learning_rate": 2.6587955415662924e-05, + "loss": 0.9658, + "step": 7740 + }, + { + "epoch": 0.5318812697540195, + "grad_norm": 1.0354957580566406, + "learning_rate": 2.6581971157242903e-05, + "loss": 0.9802, + "step": 7741 + }, + { + "epoch": 0.5319499793871101, + "grad_norm": 1.074121356010437, + "learning_rate": 2.6575986807813048e-05, + "loss": 0.9281, + "step": 7742 + }, + { + "epoch": 0.5320186890202007, + "grad_norm": 1.0490665435791016, + "learning_rate": 2.657000236771763e-05, + "loss": 0.9439, + "step": 7743 + }, + { + "epoch": 0.5320873986532912, + "grad_norm": 1.181753158569336, + "learning_rate": 2.656401783730093e-05, + "loss": 0.936, + "step": 7744 + }, + { + "epoch": 0.5321561082863817, + "grad_norm": 1.0627638101577759, + "learning_rate": 2.6558033216907248e-05, + "loss": 0.9372, + "step": 7745 + }, + { + "epoch": 0.5322248179194723, + "grad_norm": 1.0784693956375122, + "learning_rate": 2.655204850688085e-05, + "loss": 0.927, + "step": 7746 + }, + { + "epoch": 0.5322935275525629, + "grad_norm": 1.0092204809188843, + "learning_rate": 2.654606370756606e-05, + "loss": 0.9216, + "step": 7747 + }, + { + "epoch": 0.5323622371856535, + "grad_norm": 1.0798243284225464, + "learning_rate": 2.6540078819307164e-05, + "loss": 0.9704, + "step": 7748 + }, + { + "epoch": 0.532430946818744, + "grad_norm": 1.1286587715148926, + "learning_rate": 2.6534093842448478e-05, + "loss": 0.9447, + "step": 7749 + }, + { + "epoch": 0.5324996564518345, + "grad_norm": 1.194719910621643, + "learning_rate": 2.65281087773343e-05, + "loss": 0.9575, + "step": 7750 + }, + { + "epoch": 0.5325683660849251, + "grad_norm": 1.0109882354736328, + "learning_rate": 2.652212362430896e-05, + "loss": 0.9532, + "step": 7751 + }, + { + "epoch": 0.5326370757180157, + "grad_norm": 1.1025292873382568, + "learning_rate": 2.6516138383716776e-05, + "loss": 0.9254, + "step": 7752 + }, + { + "epoch": 0.5327057853511062, + "grad_norm": 1.0404773950576782, + "learning_rate": 2.651015305590207e-05, + "loss": 0.9835, + "step": 7753 + }, + { + "epoch": 0.5327744949841968, + "grad_norm": 1.0854620933532715, + "learning_rate": 2.650416764120918e-05, + "loss": 0.9207, + "step": 7754 + }, + { + "epoch": 0.5328432046172873, + "grad_norm": 1.033673882484436, + "learning_rate": 2.649818213998245e-05, + "loss": 0.9535, + "step": 7755 + }, + { + "epoch": 0.5329119142503779, + "grad_norm": 1.023894190788269, + "learning_rate": 2.649219655256621e-05, + "loss": 0.9283, + "step": 7756 + }, + { + "epoch": 0.5329806238834685, + "grad_norm": 1.0691570043563843, + "learning_rate": 2.648621087930482e-05, + "loss": 0.9558, + "step": 7757 + }, + { + "epoch": 0.533049333516559, + "grad_norm": 1.1084740161895752, + "learning_rate": 2.6480225120542613e-05, + "loss": 1.0188, + "step": 7758 + }, + { + "epoch": 0.5331180431496496, + "grad_norm": 1.1225485801696777, + "learning_rate": 2.6474239276623964e-05, + "loss": 0.9966, + "step": 7759 + }, + { + "epoch": 0.5331867527827402, + "grad_norm": 1.019452452659607, + "learning_rate": 2.646825334789323e-05, + "loss": 0.9626, + "step": 7760 + }, + { + "epoch": 0.5332554624158307, + "grad_norm": 1.0532203912734985, + "learning_rate": 2.6462267334694768e-05, + "loss": 0.9418, + "step": 7761 + }, + { + "epoch": 0.5333241720489212, + "grad_norm": 1.0950291156768799, + "learning_rate": 2.6456281237372965e-05, + "loss": 0.9753, + "step": 7762 + }, + { + "epoch": 0.5333928816820118, + "grad_norm": 1.1029257774353027, + "learning_rate": 2.645029505627219e-05, + "loss": 0.9963, + "step": 7763 + }, + { + "epoch": 0.5334615913151024, + "grad_norm": 1.09083890914917, + "learning_rate": 2.6444308791736823e-05, + "loss": 0.9679, + "step": 7764 + }, + { + "epoch": 0.533530300948193, + "grad_norm": 1.1026487350463867, + "learning_rate": 2.643832244411125e-05, + "loss": 0.9692, + "step": 7765 + }, + { + "epoch": 0.5335990105812835, + "grad_norm": 1.0196683406829834, + "learning_rate": 2.6432336013739867e-05, + "loss": 0.9259, + "step": 7766 + }, + { + "epoch": 0.533667720214374, + "grad_norm": 1.037759780883789, + "learning_rate": 2.642634950096706e-05, + "loss": 0.9733, + "step": 7767 + }, + { + "epoch": 0.5337364298474646, + "grad_norm": 1.0665781497955322, + "learning_rate": 2.642036290613724e-05, + "loss": 0.9421, + "step": 7768 + }, + { + "epoch": 0.5338051394805552, + "grad_norm": 1.0417782068252563, + "learning_rate": 2.6414376229594813e-05, + "loss": 0.9534, + "step": 7769 + }, + { + "epoch": 0.5338738491136458, + "grad_norm": 1.1131356954574585, + "learning_rate": 2.640838947168418e-05, + "loss": 0.9769, + "step": 7770 + }, + { + "epoch": 0.5339425587467362, + "grad_norm": 1.0640997886657715, + "learning_rate": 2.6402402632749758e-05, + "loss": 0.9675, + "step": 7771 + }, + { + "epoch": 0.5340112683798268, + "grad_norm": 1.0814881324768066, + "learning_rate": 2.6396415713135962e-05, + "loss": 0.9857, + "step": 7772 + }, + { + "epoch": 0.5340799780129174, + "grad_norm": 1.0800472497940063, + "learning_rate": 2.639042871318723e-05, + "loss": 0.9422, + "step": 7773 + }, + { + "epoch": 0.534148687646008, + "grad_norm": 1.0790518522262573, + "learning_rate": 2.6384441633247987e-05, + "loss": 0.9415, + "step": 7774 + }, + { + "epoch": 0.5342173972790986, + "grad_norm": 1.0664736032485962, + "learning_rate": 2.6378454473662656e-05, + "loss": 0.9789, + "step": 7775 + }, + { + "epoch": 0.5342861069121891, + "grad_norm": 1.1090229749679565, + "learning_rate": 2.637246723477567e-05, + "loss": 1.0055, + "step": 7776 + }, + { + "epoch": 0.5343548165452796, + "grad_norm": 1.0873401165008545, + "learning_rate": 2.636647991693149e-05, + "loss": 0.9448, + "step": 7777 + }, + { + "epoch": 0.5344235261783702, + "grad_norm": 1.0732457637786865, + "learning_rate": 2.6360492520474557e-05, + "loss": 0.9434, + "step": 7778 + }, + { + "epoch": 0.5344922358114608, + "grad_norm": 1.1385325193405151, + "learning_rate": 2.635450504574931e-05, + "loss": 0.9694, + "step": 7779 + }, + { + "epoch": 0.5345609454445513, + "grad_norm": 1.103662371635437, + "learning_rate": 2.6348517493100224e-05, + "loss": 0.9732, + "step": 7780 + }, + { + "epoch": 0.5346296550776419, + "grad_norm": 1.0951426029205322, + "learning_rate": 2.6342529862871735e-05, + "loss": 0.9665, + "step": 7781 + }, + { + "epoch": 0.5346983647107324, + "grad_norm": 1.1141077280044556, + "learning_rate": 2.633654215540833e-05, + "loss": 0.971, + "step": 7782 + }, + { + "epoch": 0.534767074343823, + "grad_norm": 1.1767851114273071, + "learning_rate": 2.6330554371054467e-05, + "loss": 0.9583, + "step": 7783 + }, + { + "epoch": 0.5348357839769136, + "grad_norm": 1.0810933113098145, + "learning_rate": 2.6324566510154624e-05, + "loss": 0.9782, + "step": 7784 + }, + { + "epoch": 0.5349044936100041, + "grad_norm": 1.1599361896514893, + "learning_rate": 2.6318578573053275e-05, + "loss": 0.9903, + "step": 7785 + }, + { + "epoch": 0.5349732032430947, + "grad_norm": 1.0553919076919556, + "learning_rate": 2.6312590560094914e-05, + "loss": 0.9539, + "step": 7786 + }, + { + "epoch": 0.5350419128761852, + "grad_norm": 1.0797666311264038, + "learning_rate": 2.630660247162401e-05, + "loss": 0.9208, + "step": 7787 + }, + { + "epoch": 0.5351106225092758, + "grad_norm": 1.0784900188446045, + "learning_rate": 2.6300614307985068e-05, + "loss": 0.9815, + "step": 7788 + }, + { + "epoch": 0.5351793321423663, + "grad_norm": 1.1619542837142944, + "learning_rate": 2.6294626069522572e-05, + "loss": 0.9644, + "step": 7789 + }, + { + "epoch": 0.5352480417754569, + "grad_norm": 1.1636730432510376, + "learning_rate": 2.6288637756581032e-05, + "loss": 1.0167, + "step": 7790 + }, + { + "epoch": 0.5353167514085475, + "grad_norm": 1.1590620279312134, + "learning_rate": 2.6282649369504942e-05, + "loss": 0.9813, + "step": 7791 + }, + { + "epoch": 0.5353854610416381, + "grad_norm": 1.1042591333389282, + "learning_rate": 2.627666090863881e-05, + "loss": 0.9702, + "step": 7792 + }, + { + "epoch": 0.5354541706747286, + "grad_norm": 1.1121799945831299, + "learning_rate": 2.6270672374327172e-05, + "loss": 0.9441, + "step": 7793 + }, + { + "epoch": 0.5355228803078191, + "grad_norm": 0.9940848350524902, + "learning_rate": 2.626468376691451e-05, + "loss": 0.9657, + "step": 7794 + }, + { + "epoch": 0.5355915899409097, + "grad_norm": 1.0795494318008423, + "learning_rate": 2.625869508674538e-05, + "loss": 1.0087, + "step": 7795 + }, + { + "epoch": 0.5356602995740003, + "grad_norm": 1.0529690980911255, + "learning_rate": 2.6252706334164272e-05, + "loss": 0.9376, + "step": 7796 + }, + { + "epoch": 0.5357290092070909, + "grad_norm": 1.0626888275146484, + "learning_rate": 2.6246717509515746e-05, + "loss": 0.9475, + "step": 7797 + }, + { + "epoch": 0.5357977188401813, + "grad_norm": 1.060115933418274, + "learning_rate": 2.6240728613144312e-05, + "loss": 0.9206, + "step": 7798 + }, + { + "epoch": 0.5358664284732719, + "grad_norm": 1.063529372215271, + "learning_rate": 2.6234739645394523e-05, + "loss": 0.9217, + "step": 7799 + }, + { + "epoch": 0.5359351381063625, + "grad_norm": 1.0662938356399536, + "learning_rate": 2.6228750606610907e-05, + "loss": 0.9486, + "step": 7800 + }, + { + "epoch": 0.5360038477394531, + "grad_norm": 1.0641816854476929, + "learning_rate": 2.6222761497138032e-05, + "loss": 0.9696, + "step": 7801 + }, + { + "epoch": 0.5360725573725437, + "grad_norm": 1.0602766275405884, + "learning_rate": 2.6216772317320422e-05, + "loss": 0.9369, + "step": 7802 + }, + { + "epoch": 0.5361412670056341, + "grad_norm": 1.0803234577178955, + "learning_rate": 2.6210783067502647e-05, + "loss": 0.9768, + "step": 7803 + }, + { + "epoch": 0.5362099766387247, + "grad_norm": 1.1225618124008179, + "learning_rate": 2.6204793748029254e-05, + "loss": 0.968, + "step": 7804 + }, + { + "epoch": 0.5362786862718153, + "grad_norm": 1.0130479335784912, + "learning_rate": 2.6198804359244815e-05, + "loss": 0.9523, + "step": 7805 + }, + { + "epoch": 0.5363473959049059, + "grad_norm": 1.0545486211776733, + "learning_rate": 2.6192814901493902e-05, + "loss": 0.9749, + "step": 7806 + }, + { + "epoch": 0.5364161055379965, + "grad_norm": 1.0222861766815186, + "learning_rate": 2.618682537512106e-05, + "loss": 0.9331, + "step": 7807 + }, + { + "epoch": 0.536484815171087, + "grad_norm": 1.039752721786499, + "learning_rate": 2.6180835780470887e-05, + "loss": 0.9813, + "step": 7808 + }, + { + "epoch": 0.5365535248041775, + "grad_norm": 1.05781888961792, + "learning_rate": 2.6174846117887946e-05, + "loss": 0.9323, + "step": 7809 + }, + { + "epoch": 0.5366222344372681, + "grad_norm": 1.121062994003296, + "learning_rate": 2.6168856387716835e-05, + "loss": 0.9704, + "step": 7810 + }, + { + "epoch": 0.5366909440703587, + "grad_norm": 1.1141690015792847, + "learning_rate": 2.616286659030212e-05, + "loss": 0.9913, + "step": 7811 + }, + { + "epoch": 0.5367596537034492, + "grad_norm": 1.0040727853775024, + "learning_rate": 2.61568767259884e-05, + "loss": 0.9735, + "step": 7812 + }, + { + "epoch": 0.5368283633365398, + "grad_norm": 1.103376030921936, + "learning_rate": 2.6150886795120272e-05, + "loss": 0.9459, + "step": 7813 + }, + { + "epoch": 0.5368970729696303, + "grad_norm": 1.0474053621292114, + "learning_rate": 2.6144896798042317e-05, + "loss": 0.9762, + "step": 7814 + }, + { + "epoch": 0.5369657826027209, + "grad_norm": 1.0653916597366333, + "learning_rate": 2.6138906735099157e-05, + "loss": 0.9808, + "step": 7815 + }, + { + "epoch": 0.5370344922358115, + "grad_norm": 1.053084373474121, + "learning_rate": 2.6132916606635384e-05, + "loss": 0.9364, + "step": 7816 + }, + { + "epoch": 0.537103201868902, + "grad_norm": 1.0490862131118774, + "learning_rate": 2.6126926412995606e-05, + "loss": 0.9645, + "step": 7817 + }, + { + "epoch": 0.5371719115019926, + "grad_norm": 1.118878960609436, + "learning_rate": 2.6120936154524445e-05, + "loss": 0.9488, + "step": 7818 + }, + { + "epoch": 0.5372406211350831, + "grad_norm": 1.1072825193405151, + "learning_rate": 2.6114945831566508e-05, + "loss": 0.9747, + "step": 7819 + }, + { + "epoch": 0.5373093307681737, + "grad_norm": 1.0708039999008179, + "learning_rate": 2.610895544446641e-05, + "loss": 0.9943, + "step": 7820 + }, + { + "epoch": 0.5373780404012642, + "grad_norm": 1.0457358360290527, + "learning_rate": 2.610296499356879e-05, + "loss": 0.9693, + "step": 7821 + }, + { + "epoch": 0.5374467500343548, + "grad_norm": 1.131820559501648, + "learning_rate": 2.609697447921827e-05, + "loss": 0.961, + "step": 7822 + }, + { + "epoch": 0.5375154596674454, + "grad_norm": 1.1263444423675537, + "learning_rate": 2.6090983901759477e-05, + "loss": 0.9974, + "step": 7823 + }, + { + "epoch": 0.537584169300536, + "grad_norm": 1.042392373085022, + "learning_rate": 2.6084993261537034e-05, + "loss": 0.9782, + "step": 7824 + }, + { + "epoch": 0.5376528789336265, + "grad_norm": 1.0255076885223389, + "learning_rate": 2.6079002558895592e-05, + "loss": 0.9665, + "step": 7825 + }, + { + "epoch": 0.537721588566717, + "grad_norm": 1.0478283166885376, + "learning_rate": 2.60730117941798e-05, + "loss": 0.9193, + "step": 7826 + }, + { + "epoch": 0.5377902981998076, + "grad_norm": 1.096530795097351, + "learning_rate": 2.6067020967734286e-05, + "loss": 0.9536, + "step": 7827 + }, + { + "epoch": 0.5378590078328982, + "grad_norm": 1.1092392206192017, + "learning_rate": 2.606103007990371e-05, + "loss": 0.9732, + "step": 7828 + }, + { + "epoch": 0.5379277174659888, + "grad_norm": 1.1311153173446655, + "learning_rate": 2.605503913103272e-05, + "loss": 0.9805, + "step": 7829 + }, + { + "epoch": 0.5379964270990792, + "grad_norm": 1.0626578330993652, + "learning_rate": 2.604904812146597e-05, + "loss": 0.9621, + "step": 7830 + }, + { + "epoch": 0.5380651367321698, + "grad_norm": 1.0488696098327637, + "learning_rate": 2.6043057051548125e-05, + "loss": 0.9415, + "step": 7831 + }, + { + "epoch": 0.5381338463652604, + "grad_norm": 1.062219500541687, + "learning_rate": 2.6037065921623837e-05, + "loss": 0.9838, + "step": 7832 + }, + { + "epoch": 0.538202555998351, + "grad_norm": 0.9971074461936951, + "learning_rate": 2.6031074732037787e-05, + "loss": 0.9424, + "step": 7833 + }, + { + "epoch": 0.5382712656314416, + "grad_norm": 1.0346873998641968, + "learning_rate": 2.6025083483134636e-05, + "loss": 0.9626, + "step": 7834 + }, + { + "epoch": 0.538339975264532, + "grad_norm": 1.1627087593078613, + "learning_rate": 2.6019092175259053e-05, + "loss": 0.9864, + "step": 7835 + }, + { + "epoch": 0.5384086848976226, + "grad_norm": 1.084943413734436, + "learning_rate": 2.6013100808755718e-05, + "loss": 0.9612, + "step": 7836 + }, + { + "epoch": 0.5384773945307132, + "grad_norm": 1.0665427446365356, + "learning_rate": 2.6007109383969313e-05, + "loss": 0.98, + "step": 7837 + }, + { + "epoch": 0.5385461041638038, + "grad_norm": 1.0504928827285767, + "learning_rate": 2.6001117901244513e-05, + "loss": 0.9576, + "step": 7838 + }, + { + "epoch": 0.5386148137968944, + "grad_norm": 1.1398065090179443, + "learning_rate": 2.5995126360926027e-05, + "loss": 1.0194, + "step": 7839 + }, + { + "epoch": 0.5386835234299849, + "grad_norm": 1.058422327041626, + "learning_rate": 2.5989134763358514e-05, + "loss": 0.9674, + "step": 7840 + }, + { + "epoch": 0.5387522330630754, + "grad_norm": 1.0884413719177246, + "learning_rate": 2.5983143108886688e-05, + "loss": 0.9431, + "step": 7841 + }, + { + "epoch": 0.538820942696166, + "grad_norm": 1.064111590385437, + "learning_rate": 2.5977151397855233e-05, + "loss": 0.9751, + "step": 7842 + }, + { + "epoch": 0.5388896523292566, + "grad_norm": 1.0701220035552979, + "learning_rate": 2.597115963060885e-05, + "loss": 0.9787, + "step": 7843 + }, + { + "epoch": 0.5389583619623471, + "grad_norm": 1.0372371673583984, + "learning_rate": 2.5965167807492256e-05, + "loss": 0.9554, + "step": 7844 + }, + { + "epoch": 0.5390270715954377, + "grad_norm": 1.0600497722625732, + "learning_rate": 2.5959175928850142e-05, + "loss": 0.941, + "step": 7845 + }, + { + "epoch": 0.5390957812285282, + "grad_norm": 1.0440127849578857, + "learning_rate": 2.5953183995027226e-05, + "loss": 0.9525, + "step": 7846 + }, + { + "epoch": 0.5391644908616188, + "grad_norm": 1.0474653244018555, + "learning_rate": 2.594719200636821e-05, + "loss": 1.0054, + "step": 7847 + }, + { + "epoch": 0.5392332004947094, + "grad_norm": 1.0385091304779053, + "learning_rate": 2.5941199963217816e-05, + "loss": 0.9452, + "step": 7848 + }, + { + "epoch": 0.5393019101277999, + "grad_norm": 1.0930731296539307, + "learning_rate": 2.5935207865920763e-05, + "loss": 0.9851, + "step": 7849 + }, + { + "epoch": 0.5393706197608905, + "grad_norm": 1.0258129835128784, + "learning_rate": 2.592921571482177e-05, + "loss": 0.9368, + "step": 7850 + }, + { + "epoch": 0.539439329393981, + "grad_norm": 1.0816915035247803, + "learning_rate": 2.5923223510265572e-05, + "loss": 0.9778, + "step": 7851 + }, + { + "epoch": 0.5395080390270716, + "grad_norm": 1.1414014101028442, + "learning_rate": 2.591723125259688e-05, + "loss": 0.9953, + "step": 7852 + }, + { + "epoch": 0.5395767486601621, + "grad_norm": 1.116042971611023, + "learning_rate": 2.5911238942160433e-05, + "loss": 0.9705, + "step": 7853 + }, + { + "epoch": 0.5396454582932527, + "grad_norm": 1.1102471351623535, + "learning_rate": 2.5905246579300967e-05, + "loss": 0.9719, + "step": 7854 + }, + { + "epoch": 0.5397141679263433, + "grad_norm": 1.0873228311538696, + "learning_rate": 2.589925416436322e-05, + "loss": 1.0058, + "step": 7855 + }, + { + "epoch": 0.5397828775594339, + "grad_norm": 1.109985589981079, + "learning_rate": 2.589326169769193e-05, + "loss": 0.9543, + "step": 7856 + }, + { + "epoch": 0.5398515871925244, + "grad_norm": 1.1463313102722168, + "learning_rate": 2.588726917963183e-05, + "loss": 1.0142, + "step": 7857 + }, + { + "epoch": 0.5399202968256149, + "grad_norm": 1.0568314790725708, + "learning_rate": 2.5881276610527684e-05, + "loss": 0.9584, + "step": 7858 + }, + { + "epoch": 0.5399890064587055, + "grad_norm": 1.0317444801330566, + "learning_rate": 2.587528399072423e-05, + "loss": 0.9583, + "step": 7859 + }, + { + "epoch": 0.5400577160917961, + "grad_norm": 1.0741522312164307, + "learning_rate": 2.5869291320566217e-05, + "loss": 1.0073, + "step": 7860 + }, + { + "epoch": 0.5401264257248867, + "grad_norm": 1.049225091934204, + "learning_rate": 2.586329860039841e-05, + "loss": 0.9432, + "step": 7861 + }, + { + "epoch": 0.5401951353579771, + "grad_norm": 1.0776565074920654, + "learning_rate": 2.585730583056556e-05, + "loss": 1.0027, + "step": 7862 + }, + { + "epoch": 0.5402638449910677, + "grad_norm": 1.028512716293335, + "learning_rate": 2.5851313011412427e-05, + "loss": 0.9541, + "step": 7863 + }, + { + "epoch": 0.5403325546241583, + "grad_norm": 1.0653527975082397, + "learning_rate": 2.584532014328378e-05, + "loss": 0.9803, + "step": 7864 + }, + { + "epoch": 0.5404012642572489, + "grad_norm": 1.0413897037506104, + "learning_rate": 2.5839327226524375e-05, + "loss": 0.9472, + "step": 7865 + }, + { + "epoch": 0.5404699738903395, + "grad_norm": 1.0416209697723389, + "learning_rate": 2.583333426147898e-05, + "loss": 0.9738, + "step": 7866 + }, + { + "epoch": 0.5405386835234299, + "grad_norm": 1.06528902053833, + "learning_rate": 2.5827341248492387e-05, + "loss": 0.9656, + "step": 7867 + }, + { + "epoch": 0.5406073931565205, + "grad_norm": 1.120166301727295, + "learning_rate": 2.5821348187909344e-05, + "loss": 0.9795, + "step": 7868 + }, + { + "epoch": 0.5406761027896111, + "grad_norm": 1.0554132461547852, + "learning_rate": 2.5815355080074653e-05, + "loss": 0.9347, + "step": 7869 + }, + { + "epoch": 0.5407448124227017, + "grad_norm": 1.0898772478103638, + "learning_rate": 2.5809361925333077e-05, + "loss": 0.9495, + "step": 7870 + }, + { + "epoch": 0.5408135220557923, + "grad_norm": 1.0241730213165283, + "learning_rate": 2.5803368724029398e-05, + "loss": 0.9942, + "step": 7871 + }, + { + "epoch": 0.5408822316888828, + "grad_norm": 1.0971513986587524, + "learning_rate": 2.5797375476508413e-05, + "loss": 0.9791, + "step": 7872 + }, + { + "epoch": 0.5409509413219733, + "grad_norm": 1.134979009628296, + "learning_rate": 2.57913821831149e-05, + "loss": 0.9738, + "step": 7873 + }, + { + "epoch": 0.5410196509550639, + "grad_norm": 1.0387283563613892, + "learning_rate": 2.578538884419365e-05, + "loss": 0.9677, + "step": 7874 + }, + { + "epoch": 0.5410883605881545, + "grad_norm": 1.1160820722579956, + "learning_rate": 2.577939546008946e-05, + "loss": 1.0031, + "step": 7875 + }, + { + "epoch": 0.541157070221245, + "grad_norm": 1.038243293762207, + "learning_rate": 2.5773402031147124e-05, + "loss": 0.9405, + "step": 7876 + }, + { + "epoch": 0.5412257798543356, + "grad_norm": 1.0519771575927734, + "learning_rate": 2.576740855771144e-05, + "loss": 0.9389, + "step": 7877 + }, + { + "epoch": 0.5412944894874261, + "grad_norm": 1.0473638772964478, + "learning_rate": 2.576141504012721e-05, + "loss": 0.9491, + "step": 7878 + }, + { + "epoch": 0.5413631991205167, + "grad_norm": 1.0059518814086914, + "learning_rate": 2.5755421478739246e-05, + "loss": 0.9322, + "step": 7879 + }, + { + "epoch": 0.5414319087536072, + "grad_norm": 1.0900956392288208, + "learning_rate": 2.5749427873892333e-05, + "loss": 0.9514, + "step": 7880 + }, + { + "epoch": 0.5415006183866978, + "grad_norm": 1.037406086921692, + "learning_rate": 2.57434342259313e-05, + "loss": 0.9624, + "step": 7881 + }, + { + "epoch": 0.5415693280197884, + "grad_norm": 1.020850658416748, + "learning_rate": 2.5737440535200947e-05, + "loss": 0.9758, + "step": 7882 + }, + { + "epoch": 0.5416380376528789, + "grad_norm": 1.0677319765090942, + "learning_rate": 2.5731446802046084e-05, + "loss": 0.9394, + "step": 7883 + }, + { + "epoch": 0.5417067472859695, + "grad_norm": 1.0158965587615967, + "learning_rate": 2.5725453026811535e-05, + "loss": 0.931, + "step": 7884 + }, + { + "epoch": 0.54177545691906, + "grad_norm": 1.0561074018478394, + "learning_rate": 2.5719459209842117e-05, + "loss": 0.9734, + "step": 7885 + }, + { + "epoch": 0.5418441665521506, + "grad_norm": 0.9895002245903015, + "learning_rate": 2.5713465351482645e-05, + "loss": 0.9135, + "step": 7886 + }, + { + "epoch": 0.5419128761852412, + "grad_norm": 1.0696877241134644, + "learning_rate": 2.570747145207796e-05, + "loss": 0.9701, + "step": 7887 + }, + { + "epoch": 0.5419815858183318, + "grad_norm": 1.127126693725586, + "learning_rate": 2.570147751197286e-05, + "loss": 0.9324, + "step": 7888 + }, + { + "epoch": 0.5420502954514222, + "grad_norm": 1.0751227140426636, + "learning_rate": 2.5695483531512187e-05, + "loss": 0.9705, + "step": 7889 + }, + { + "epoch": 0.5421190050845128, + "grad_norm": 0.966996967792511, + "learning_rate": 2.5689489511040776e-05, + "loss": 0.9169, + "step": 7890 + }, + { + "epoch": 0.5421877147176034, + "grad_norm": 1.0667564868927002, + "learning_rate": 2.568349545090345e-05, + "loss": 0.9521, + "step": 7891 + }, + { + "epoch": 0.542256424350694, + "grad_norm": 1.1414191722869873, + "learning_rate": 2.5677501351445055e-05, + "loss": 0.9284, + "step": 7892 + }, + { + "epoch": 0.5423251339837846, + "grad_norm": 1.0091010332107544, + "learning_rate": 2.5671507213010415e-05, + "loss": 0.9197, + "step": 7893 + }, + { + "epoch": 0.542393843616875, + "grad_norm": 1.0720371007919312, + "learning_rate": 2.566551303594437e-05, + "loss": 0.9855, + "step": 7894 + }, + { + "epoch": 0.5424625532499656, + "grad_norm": 1.1083875894546509, + "learning_rate": 2.5659518820591777e-05, + "loss": 1.0052, + "step": 7895 + }, + { + "epoch": 0.5425312628830562, + "grad_norm": 1.080137014389038, + "learning_rate": 2.5653524567297453e-05, + "loss": 0.9757, + "step": 7896 + }, + { + "epoch": 0.5425999725161468, + "grad_norm": 1.0964627265930176, + "learning_rate": 2.564753027640627e-05, + "loss": 0.9597, + "step": 7897 + }, + { + "epoch": 0.5426686821492374, + "grad_norm": 1.0700609683990479, + "learning_rate": 2.5641535948263062e-05, + "loss": 0.9638, + "step": 7898 + }, + { + "epoch": 0.5427373917823278, + "grad_norm": 1.0576286315917969, + "learning_rate": 2.563554158321268e-05, + "loss": 0.9329, + "step": 7899 + }, + { + "epoch": 0.5428061014154184, + "grad_norm": 1.0242704153060913, + "learning_rate": 2.562954718159999e-05, + "loss": 0.9617, + "step": 7900 + }, + { + "epoch": 0.542874811048509, + "grad_norm": 1.0728349685668945, + "learning_rate": 2.5623552743769817e-05, + "loss": 0.9269, + "step": 7901 + }, + { + "epoch": 0.5429435206815996, + "grad_norm": 1.032022476196289, + "learning_rate": 2.561755827006705e-05, + "loss": 0.9645, + "step": 7902 + }, + { + "epoch": 0.5430122303146901, + "grad_norm": 1.0703140497207642, + "learning_rate": 2.5611563760836523e-05, + "loss": 0.938, + "step": 7903 + }, + { + "epoch": 0.5430809399477807, + "grad_norm": 1.0384986400604248, + "learning_rate": 2.5605569216423108e-05, + "loss": 0.96, + "step": 7904 + }, + { + "epoch": 0.5431496495808712, + "grad_norm": 1.1309444904327393, + "learning_rate": 2.5599574637171668e-05, + "loss": 0.9573, + "step": 7905 + }, + { + "epoch": 0.5432183592139618, + "grad_norm": 1.127309799194336, + "learning_rate": 2.559358002342706e-05, + "loss": 0.9719, + "step": 7906 + }, + { + "epoch": 0.5432870688470524, + "grad_norm": 1.0888175964355469, + "learning_rate": 2.5587585375534162e-05, + "loss": 0.9806, + "step": 7907 + }, + { + "epoch": 0.5433557784801429, + "grad_norm": 1.03667151927948, + "learning_rate": 2.5581590693837825e-05, + "loss": 0.9509, + "step": 7908 + }, + { + "epoch": 0.5434244881132335, + "grad_norm": 1.1388224363327026, + "learning_rate": 2.5575595978682937e-05, + "loss": 0.9805, + "step": 7909 + }, + { + "epoch": 0.543493197746324, + "grad_norm": 1.038603663444519, + "learning_rate": 2.5569601230414364e-05, + "loss": 0.9179, + "step": 7910 + }, + { + "epoch": 0.5435619073794146, + "grad_norm": 1.0710561275482178, + "learning_rate": 2.5563606449376976e-05, + "loss": 0.9607, + "step": 7911 + }, + { + "epoch": 0.5436306170125051, + "grad_norm": 1.0686718225479126, + "learning_rate": 2.555761163591565e-05, + "loss": 0.9252, + "step": 7912 + }, + { + "epoch": 0.5436993266455957, + "grad_norm": 1.0682555437088013, + "learning_rate": 2.5551616790375278e-05, + "loss": 0.9763, + "step": 7913 + }, + { + "epoch": 0.5437680362786863, + "grad_norm": 1.07375168800354, + "learning_rate": 2.5545621913100713e-05, + "loss": 0.9928, + "step": 7914 + }, + { + "epoch": 0.5438367459117769, + "grad_norm": 1.0803871154785156, + "learning_rate": 2.5539627004436866e-05, + "loss": 0.9164, + "step": 7915 + }, + { + "epoch": 0.5439054555448674, + "grad_norm": 1.0182394981384277, + "learning_rate": 2.55336320647286e-05, + "loss": 0.9259, + "step": 7916 + }, + { + "epoch": 0.5439741651779579, + "grad_norm": 1.0596462488174438, + "learning_rate": 2.5527637094320812e-05, + "loss": 0.9625, + "step": 7917 + }, + { + "epoch": 0.5440428748110485, + "grad_norm": 1.1357917785644531, + "learning_rate": 2.552164209355838e-05, + "loss": 0.984, + "step": 7918 + }, + { + "epoch": 0.5441115844441391, + "grad_norm": 1.1132042407989502, + "learning_rate": 2.551564706278619e-05, + "loss": 0.9752, + "step": 7919 + }, + { + "epoch": 0.5441802940772297, + "grad_norm": 1.07838773727417, + "learning_rate": 2.550965200234915e-05, + "loss": 0.9806, + "step": 7920 + }, + { + "epoch": 0.5442490037103201, + "grad_norm": 1.0504357814788818, + "learning_rate": 2.5503656912592133e-05, + "loss": 0.9864, + "step": 7921 + }, + { + "epoch": 0.5443177133434107, + "grad_norm": 1.047164797782898, + "learning_rate": 2.549766179386004e-05, + "loss": 0.9761, + "step": 7922 + }, + { + "epoch": 0.5443864229765013, + "grad_norm": 1.0413298606872559, + "learning_rate": 2.5491666646497774e-05, + "loss": 0.942, + "step": 7923 + }, + { + "epoch": 0.5444551326095919, + "grad_norm": 1.0401281118392944, + "learning_rate": 2.548567147085022e-05, + "loss": 0.9383, + "step": 7924 + }, + { + "epoch": 0.5445238422426825, + "grad_norm": 1.1297262907028198, + "learning_rate": 2.5479676267262287e-05, + "loss": 1.0235, + "step": 7925 + }, + { + "epoch": 0.5445925518757729, + "grad_norm": 1.081335186958313, + "learning_rate": 2.5473681036078868e-05, + "loss": 0.9705, + "step": 7926 + }, + { + "epoch": 0.5446612615088635, + "grad_norm": 1.1007254123687744, + "learning_rate": 2.546768577764487e-05, + "loss": 0.9562, + "step": 7927 + }, + { + "epoch": 0.5447299711419541, + "grad_norm": 1.12466299533844, + "learning_rate": 2.546169049230519e-05, + "loss": 0.9635, + "step": 7928 + }, + { + "epoch": 0.5447986807750447, + "grad_norm": 1.2298433780670166, + "learning_rate": 2.5455695180404744e-05, + "loss": 0.9576, + "step": 7929 + }, + { + "epoch": 0.5448673904081353, + "grad_norm": 0.9995604753494263, + "learning_rate": 2.5449699842288433e-05, + "loss": 0.9468, + "step": 7930 + }, + { + "epoch": 0.5449361000412258, + "grad_norm": 1.0788490772247314, + "learning_rate": 2.5443704478301154e-05, + "loss": 0.9436, + "step": 7931 + }, + { + "epoch": 0.5450048096743163, + "grad_norm": 1.1096291542053223, + "learning_rate": 2.543770908878783e-05, + "loss": 0.9173, + "step": 7932 + }, + { + "epoch": 0.5450735193074069, + "grad_norm": 1.0544739961624146, + "learning_rate": 2.5431713674093378e-05, + "loss": 0.969, + "step": 7933 + }, + { + "epoch": 0.5451422289404975, + "grad_norm": 1.0757200717926025, + "learning_rate": 2.5425718234562697e-05, + "loss": 0.9698, + "step": 7934 + }, + { + "epoch": 0.545210938573588, + "grad_norm": 1.029861330986023, + "learning_rate": 2.5419722770540705e-05, + "loss": 0.9554, + "step": 7935 + }, + { + "epoch": 0.5452796482066786, + "grad_norm": 0.9947589039802551, + "learning_rate": 2.541372728237231e-05, + "loss": 0.9417, + "step": 7936 + }, + { + "epoch": 0.5453483578397691, + "grad_norm": 1.0979501008987427, + "learning_rate": 2.5407731770402445e-05, + "loss": 0.9855, + "step": 7937 + }, + { + "epoch": 0.5454170674728597, + "grad_norm": 1.164117455482483, + "learning_rate": 2.540173623497602e-05, + "loss": 0.9668, + "step": 7938 + }, + { + "epoch": 0.5454857771059503, + "grad_norm": 1.1355178356170654, + "learning_rate": 2.539574067643795e-05, + "loss": 0.9585, + "step": 7939 + }, + { + "epoch": 0.5455544867390408, + "grad_norm": 1.055469036102295, + "learning_rate": 2.5389745095133165e-05, + "loss": 0.9496, + "step": 7940 + }, + { + "epoch": 0.5456231963721314, + "grad_norm": 1.024946928024292, + "learning_rate": 2.5383749491406584e-05, + "loss": 0.9787, + "step": 7941 + }, + { + "epoch": 0.5456919060052219, + "grad_norm": 1.0524029731750488, + "learning_rate": 2.5377753865603122e-05, + "loss": 0.9646, + "step": 7942 + }, + { + "epoch": 0.5457606156383125, + "grad_norm": 1.0321294069290161, + "learning_rate": 2.5371758218067716e-05, + "loss": 0.9623, + "step": 7943 + }, + { + "epoch": 0.545829325271403, + "grad_norm": 1.0817097425460815, + "learning_rate": 2.5365762549145283e-05, + "loss": 0.986, + "step": 7944 + }, + { + "epoch": 0.5458980349044936, + "grad_norm": 1.0385783910751343, + "learning_rate": 2.535976685918075e-05, + "loss": 0.9294, + "step": 7945 + }, + { + "epoch": 0.5459667445375842, + "grad_norm": 1.0969388484954834, + "learning_rate": 2.5353771148519057e-05, + "loss": 0.9506, + "step": 7946 + }, + { + "epoch": 0.5460354541706748, + "grad_norm": 1.0436185598373413, + "learning_rate": 2.534777541750512e-05, + "loss": 0.9181, + "step": 7947 + }, + { + "epoch": 0.5461041638037653, + "grad_norm": 1.0724081993103027, + "learning_rate": 2.5341779666483878e-05, + "loss": 0.9891, + "step": 7948 + }, + { + "epoch": 0.5461728734368558, + "grad_norm": 1.0230896472930908, + "learning_rate": 2.533578389580026e-05, + "loss": 0.9747, + "step": 7949 + }, + { + "epoch": 0.5462415830699464, + "grad_norm": 1.0273053646087646, + "learning_rate": 2.532978810579919e-05, + "loss": 0.9329, + "step": 7950 + }, + { + "epoch": 0.546310292703037, + "grad_norm": 1.094933032989502, + "learning_rate": 2.5323792296825628e-05, + "loss": 0.9658, + "step": 7951 + }, + { + "epoch": 0.5463790023361276, + "grad_norm": 1.0721641778945923, + "learning_rate": 2.5317796469224485e-05, + "loss": 0.9356, + "step": 7952 + }, + { + "epoch": 0.546447711969218, + "grad_norm": 1.0873091220855713, + "learning_rate": 2.5311800623340708e-05, + "loss": 0.9614, + "step": 7953 + }, + { + "epoch": 0.5465164216023086, + "grad_norm": 1.103251338005066, + "learning_rate": 2.5305804759519226e-05, + "loss": 0.9961, + "step": 7954 + }, + { + "epoch": 0.5465851312353992, + "grad_norm": 1.0280898809432983, + "learning_rate": 2.5299808878104982e-05, + "loss": 0.995, + "step": 7955 + }, + { + "epoch": 0.5466538408684898, + "grad_norm": 1.0806275606155396, + "learning_rate": 2.529381297944292e-05, + "loss": 0.973, + "step": 7956 + }, + { + "epoch": 0.5467225505015804, + "grad_norm": 1.0740324258804321, + "learning_rate": 2.5287817063877972e-05, + "loss": 0.9588, + "step": 7957 + }, + { + "epoch": 0.5467912601346708, + "grad_norm": 1.0251491069793701, + "learning_rate": 2.528182113175509e-05, + "loss": 0.9617, + "step": 7958 + }, + { + "epoch": 0.5468599697677614, + "grad_norm": 1.084995985031128, + "learning_rate": 2.5275825183419206e-05, + "loss": 0.9462, + "step": 7959 + }, + { + "epoch": 0.546928679400852, + "grad_norm": 1.1187002658843994, + "learning_rate": 2.5269829219215263e-05, + "loss": 0.9493, + "step": 7960 + }, + { + "epoch": 0.5469973890339426, + "grad_norm": 1.0157877206802368, + "learning_rate": 2.5263833239488216e-05, + "loss": 0.9568, + "step": 7961 + }, + { + "epoch": 0.5470660986670332, + "grad_norm": 1.0047460794448853, + "learning_rate": 2.5257837244582998e-05, + "loss": 0.9184, + "step": 7962 + }, + { + "epoch": 0.5471348083001237, + "grad_norm": 1.0628859996795654, + "learning_rate": 2.525184123484457e-05, + "loss": 0.9811, + "step": 7963 + }, + { + "epoch": 0.5472035179332142, + "grad_norm": 1.1022135019302368, + "learning_rate": 2.5245845210617867e-05, + "loss": 0.9557, + "step": 7964 + }, + { + "epoch": 0.5472722275663048, + "grad_norm": 1.0559957027435303, + "learning_rate": 2.5239849172247832e-05, + "loss": 0.9352, + "step": 7965 + }, + { + "epoch": 0.5473409371993954, + "grad_norm": 1.1872501373291016, + "learning_rate": 2.523385312007943e-05, + "loss": 1.0301, + "step": 7966 + }, + { + "epoch": 0.5474096468324859, + "grad_norm": 1.002002477645874, + "learning_rate": 2.5227857054457593e-05, + "loss": 0.9438, + "step": 7967 + }, + { + "epoch": 0.5474783564655765, + "grad_norm": 1.0417814254760742, + "learning_rate": 2.5221860975727275e-05, + "loss": 0.95, + "step": 7968 + }, + { + "epoch": 0.547547066098667, + "grad_norm": 1.0291081666946411, + "learning_rate": 2.5215864884233443e-05, + "loss": 0.9903, + "step": 7969 + }, + { + "epoch": 0.5476157757317576, + "grad_norm": 1.031184434890747, + "learning_rate": 2.5209868780321022e-05, + "loss": 0.9641, + "step": 7970 + }, + { + "epoch": 0.5476844853648482, + "grad_norm": 1.0026473999023438, + "learning_rate": 2.520387266433499e-05, + "loss": 0.9609, + "step": 7971 + }, + { + "epoch": 0.5477531949979387, + "grad_norm": 1.0193266868591309, + "learning_rate": 2.519787653662028e-05, + "loss": 0.9619, + "step": 7972 + }, + { + "epoch": 0.5478219046310293, + "grad_norm": 1.0452187061309814, + "learning_rate": 2.5191880397521856e-05, + "loss": 0.9566, + "step": 7973 + }, + { + "epoch": 0.5478906142641198, + "grad_norm": 1.0487756729125977, + "learning_rate": 2.518588424738467e-05, + "loss": 0.95, + "step": 7974 + }, + { + "epoch": 0.5479593238972104, + "grad_norm": 1.0455374717712402, + "learning_rate": 2.5179888086553675e-05, + "loss": 0.931, + "step": 7975 + }, + { + "epoch": 0.5480280335303009, + "grad_norm": 1.1735355854034424, + "learning_rate": 2.517389191537383e-05, + "loss": 0.9472, + "step": 7976 + }, + { + "epoch": 0.5480967431633915, + "grad_norm": 1.1072808504104614, + "learning_rate": 2.5167895734190093e-05, + "loss": 0.973, + "step": 7977 + }, + { + "epoch": 0.5481654527964821, + "grad_norm": 1.0972752571105957, + "learning_rate": 2.516189954334741e-05, + "loss": 0.9815, + "step": 7978 + }, + { + "epoch": 0.5482341624295727, + "grad_norm": 1.040696144104004, + "learning_rate": 2.5155903343190753e-05, + "loss": 0.9492, + "step": 7979 + }, + { + "epoch": 0.5483028720626631, + "grad_norm": 1.0469499826431274, + "learning_rate": 2.5149907134065058e-05, + "loss": 0.9558, + "step": 7980 + }, + { + "epoch": 0.5483715816957537, + "grad_norm": 1.1251680850982666, + "learning_rate": 2.514391091631531e-05, + "loss": 0.9594, + "step": 7981 + }, + { + "epoch": 0.5484402913288443, + "grad_norm": 1.0098590850830078, + "learning_rate": 2.5137914690286445e-05, + "loss": 0.9648, + "step": 7982 + }, + { + "epoch": 0.5485090009619349, + "grad_norm": 1.0395210981369019, + "learning_rate": 2.513191845632344e-05, + "loss": 0.9844, + "step": 7983 + }, + { + "epoch": 0.5485777105950255, + "grad_norm": 1.131784439086914, + "learning_rate": 2.512592221477124e-05, + "loss": 0.9868, + "step": 7984 + }, + { + "epoch": 0.5486464202281159, + "grad_norm": 1.0675370693206787, + "learning_rate": 2.5119925965974816e-05, + "loss": 0.9939, + "step": 7985 + }, + { + "epoch": 0.5487151298612065, + "grad_norm": 1.171054482460022, + "learning_rate": 2.511392971027913e-05, + "loss": 1.0051, + "step": 7986 + }, + { + "epoch": 0.5487838394942971, + "grad_norm": 1.0730468034744263, + "learning_rate": 2.5107933448029124e-05, + "loss": 0.9395, + "step": 7987 + }, + { + "epoch": 0.5488525491273877, + "grad_norm": 1.1507817506790161, + "learning_rate": 2.5101937179569784e-05, + "loss": 0.9521, + "step": 7988 + }, + { + "epoch": 0.5489212587604783, + "grad_norm": 1.013590931892395, + "learning_rate": 2.5095940905246056e-05, + "loss": 0.961, + "step": 7989 + }, + { + "epoch": 0.5489899683935687, + "grad_norm": 1.1101058721542358, + "learning_rate": 2.5089944625402906e-05, + "loss": 0.9225, + "step": 7990 + }, + { + "epoch": 0.5490586780266593, + "grad_norm": 1.0935776233673096, + "learning_rate": 2.50839483403853e-05, + "loss": 0.9789, + "step": 7991 + }, + { + "epoch": 0.5491273876597499, + "grad_norm": 1.089259147644043, + "learning_rate": 2.507795205053819e-05, + "loss": 1.0002, + "step": 7992 + }, + { + "epoch": 0.5491960972928405, + "grad_norm": 1.0995001792907715, + "learning_rate": 2.5071955756206543e-05, + "loss": 1.0282, + "step": 7993 + }, + { + "epoch": 0.549264806925931, + "grad_norm": 1.0174835920333862, + "learning_rate": 2.506595945773534e-05, + "loss": 0.9431, + "step": 7994 + }, + { + "epoch": 0.5493335165590216, + "grad_norm": 1.096086025238037, + "learning_rate": 2.505996315546952e-05, + "loss": 0.9698, + "step": 7995 + }, + { + "epoch": 0.5494022261921121, + "grad_norm": 1.070104718208313, + "learning_rate": 2.5053966849754052e-05, + "loss": 0.9812, + "step": 7996 + }, + { + "epoch": 0.5494709358252027, + "grad_norm": 1.051385760307312, + "learning_rate": 2.504797054093391e-05, + "loss": 0.9476, + "step": 7997 + }, + { + "epoch": 0.5495396454582933, + "grad_norm": 1.1355637311935425, + "learning_rate": 2.504197422935405e-05, + "loss": 0.9581, + "step": 7998 + }, + { + "epoch": 0.5496083550913838, + "grad_norm": 1.1386737823486328, + "learning_rate": 2.5035977915359442e-05, + "loss": 0.9649, + "step": 7999 + }, + { + "epoch": 0.5496770647244744, + "grad_norm": 1.0011066198349, + "learning_rate": 2.502998159929505e-05, + "loss": 0.9598, + "step": 8000 + }, + { + "epoch": 0.5497457743575649, + "grad_norm": 1.1100945472717285, + "learning_rate": 2.5023985281505825e-05, + "loss": 0.9739, + "step": 8001 + }, + { + "epoch": 0.5498144839906555, + "grad_norm": 1.0222301483154297, + "learning_rate": 2.5017988962336748e-05, + "loss": 0.9579, + "step": 8002 + }, + { + "epoch": 0.549883193623746, + "grad_norm": 1.0474375486373901, + "learning_rate": 2.5011992642132776e-05, + "loss": 0.9769, + "step": 8003 + }, + { + "epoch": 0.5499519032568366, + "grad_norm": 1.0752254724502563, + "learning_rate": 2.500599632123887e-05, + "loss": 0.9612, + "step": 8004 + }, + { + "epoch": 0.5500206128899272, + "grad_norm": 1.0363268852233887, + "learning_rate": 2.5e-05, + "loss": 0.9469, + "step": 8005 + }, + { + "epoch": 0.5500893225230177, + "grad_norm": 1.1066510677337646, + "learning_rate": 2.499400367876113e-05, + "loss": 0.9998, + "step": 8006 + }, + { + "epoch": 0.5501580321561083, + "grad_norm": 1.1089556217193604, + "learning_rate": 2.4988007357867227e-05, + "loss": 0.9841, + "step": 8007 + }, + { + "epoch": 0.5502267417891988, + "grad_norm": 1.0614635944366455, + "learning_rate": 2.498201103766326e-05, + "loss": 0.9906, + "step": 8008 + }, + { + "epoch": 0.5502954514222894, + "grad_norm": 1.010819673538208, + "learning_rate": 2.497601471849418e-05, + "loss": 0.9286, + "step": 8009 + }, + { + "epoch": 0.55036416105538, + "grad_norm": 1.0575289726257324, + "learning_rate": 2.4970018400704957e-05, + "loss": 0.9283, + "step": 8010 + }, + { + "epoch": 0.5504328706884706, + "grad_norm": 1.080979824066162, + "learning_rate": 2.496402208464056e-05, + "loss": 0.9605, + "step": 8011 + }, + { + "epoch": 0.550501580321561, + "grad_norm": 1.088341236114502, + "learning_rate": 2.495802577064595e-05, + "loss": 0.9501, + "step": 8012 + }, + { + "epoch": 0.5505702899546516, + "grad_norm": 1.0734748840332031, + "learning_rate": 2.495202945906609e-05, + "loss": 0.9729, + "step": 8013 + }, + { + "epoch": 0.5506389995877422, + "grad_norm": 1.0474892854690552, + "learning_rate": 2.4946033150245954e-05, + "loss": 0.9297, + "step": 8014 + }, + { + "epoch": 0.5507077092208328, + "grad_norm": 0.9979798793792725, + "learning_rate": 2.494003684453049e-05, + "loss": 0.9513, + "step": 8015 + }, + { + "epoch": 0.5507764188539234, + "grad_norm": 1.1464966535568237, + "learning_rate": 2.4934040542264667e-05, + "loss": 0.959, + "step": 8016 + }, + { + "epoch": 0.5508451284870138, + "grad_norm": 1.0894944667816162, + "learning_rate": 2.4928044243793456e-05, + "loss": 0.965, + "step": 8017 + }, + { + "epoch": 0.5509138381201044, + "grad_norm": 1.0862141847610474, + "learning_rate": 2.492204794946182e-05, + "loss": 0.9819, + "step": 8018 + }, + { + "epoch": 0.550982547753195, + "grad_norm": 1.0843409299850464, + "learning_rate": 2.4916051659614713e-05, + "loss": 0.948, + "step": 8019 + }, + { + "epoch": 0.5510512573862856, + "grad_norm": 1.1293820142745972, + "learning_rate": 2.4910055374597103e-05, + "loss": 0.9691, + "step": 8020 + }, + { + "epoch": 0.5511199670193762, + "grad_norm": 1.0495827198028564, + "learning_rate": 2.490405909475395e-05, + "loss": 0.9245, + "step": 8021 + }, + { + "epoch": 0.5511886766524666, + "grad_norm": 1.0695171356201172, + "learning_rate": 2.4898062820430225e-05, + "loss": 0.9572, + "step": 8022 + }, + { + "epoch": 0.5512573862855572, + "grad_norm": 1.1084959506988525, + "learning_rate": 2.489206655197088e-05, + "loss": 0.9433, + "step": 8023 + }, + { + "epoch": 0.5513260959186478, + "grad_norm": 1.0474777221679688, + "learning_rate": 2.4886070289720877e-05, + "loss": 0.9506, + "step": 8024 + }, + { + "epoch": 0.5513948055517384, + "grad_norm": 1.1119284629821777, + "learning_rate": 2.488007403402519e-05, + "loss": 0.9904, + "step": 8025 + }, + { + "epoch": 0.551463515184829, + "grad_norm": 1.0419389009475708, + "learning_rate": 2.487407778522876e-05, + "loss": 0.9335, + "step": 8026 + }, + { + "epoch": 0.5515322248179195, + "grad_norm": 1.0716807842254639, + "learning_rate": 2.4868081543676562e-05, + "loss": 0.977, + "step": 8027 + }, + { + "epoch": 0.55160093445101, + "grad_norm": 1.056767463684082, + "learning_rate": 2.486208530971356e-05, + "loss": 0.9698, + "step": 8028 + }, + { + "epoch": 0.5516696440841006, + "grad_norm": 1.1173882484436035, + "learning_rate": 2.48560890836847e-05, + "loss": 0.9523, + "step": 8029 + }, + { + "epoch": 0.5517383537171912, + "grad_norm": 1.012874722480774, + "learning_rate": 2.4850092865934944e-05, + "loss": 0.9603, + "step": 8030 + }, + { + "epoch": 0.5518070633502817, + "grad_norm": 1.0863533020019531, + "learning_rate": 2.4844096656809257e-05, + "loss": 0.9536, + "step": 8031 + }, + { + "epoch": 0.5518757729833723, + "grad_norm": 1.0249234437942505, + "learning_rate": 2.483810045665259e-05, + "loss": 0.9144, + "step": 8032 + }, + { + "epoch": 0.5519444826164628, + "grad_norm": 1.0021467208862305, + "learning_rate": 2.4832104265809913e-05, + "loss": 0.946, + "step": 8033 + }, + { + "epoch": 0.5520131922495534, + "grad_norm": 1.0363941192626953, + "learning_rate": 2.4826108084626172e-05, + "loss": 0.9553, + "step": 8034 + }, + { + "epoch": 0.5520819018826439, + "grad_norm": 1.067721962928772, + "learning_rate": 2.4820111913446327e-05, + "loss": 0.9626, + "step": 8035 + }, + { + "epoch": 0.5521506115157345, + "grad_norm": 1.049018383026123, + "learning_rate": 2.4814115752615334e-05, + "loss": 0.9506, + "step": 8036 + }, + { + "epoch": 0.5522193211488251, + "grad_norm": 1.1049176454544067, + "learning_rate": 2.4808119602478146e-05, + "loss": 0.9806, + "step": 8037 + }, + { + "epoch": 0.5522880307819156, + "grad_norm": 1.1125751733779907, + "learning_rate": 2.480212346337972e-05, + "loss": 0.9732, + "step": 8038 + }, + { + "epoch": 0.5523567404150062, + "grad_norm": 1.0177476406097412, + "learning_rate": 2.4796127335665024e-05, + "loss": 0.9405, + "step": 8039 + }, + { + "epoch": 0.5524254500480967, + "grad_norm": 0.9661063551902771, + "learning_rate": 2.4790131219678984e-05, + "loss": 0.9388, + "step": 8040 + }, + { + "epoch": 0.5524941596811873, + "grad_norm": 1.0834698677062988, + "learning_rate": 2.4784135115766566e-05, + "loss": 0.9422, + "step": 8041 + }, + { + "epoch": 0.5525628693142779, + "grad_norm": 0.9743319153785706, + "learning_rate": 2.4778139024272724e-05, + "loss": 0.9215, + "step": 8042 + }, + { + "epoch": 0.5526315789473685, + "grad_norm": 1.0611083507537842, + "learning_rate": 2.477214294554241e-05, + "loss": 0.9486, + "step": 8043 + }, + { + "epoch": 0.5527002885804589, + "grad_norm": 1.024606466293335, + "learning_rate": 2.4766146879920583e-05, + "loss": 0.9237, + "step": 8044 + }, + { + "epoch": 0.5527689982135495, + "grad_norm": 1.1137018203735352, + "learning_rate": 2.476015082775217e-05, + "loss": 0.9618, + "step": 8045 + }, + { + "epoch": 0.5528377078466401, + "grad_norm": 1.065101981163025, + "learning_rate": 2.475415478938214e-05, + "loss": 0.9669, + "step": 8046 + }, + { + "epoch": 0.5529064174797307, + "grad_norm": 0.9718230962753296, + "learning_rate": 2.4748158765155436e-05, + "loss": 0.958, + "step": 8047 + }, + { + "epoch": 0.5529751271128213, + "grad_norm": 1.0767009258270264, + "learning_rate": 2.4742162755416997e-05, + "loss": 0.9825, + "step": 8048 + }, + { + "epoch": 0.5530438367459117, + "grad_norm": 1.135481595993042, + "learning_rate": 2.4736166760511794e-05, + "loss": 0.9771, + "step": 8049 + }, + { + "epoch": 0.5531125463790023, + "grad_norm": 1.0930986404418945, + "learning_rate": 2.4730170780784742e-05, + "loss": 1.0024, + "step": 8050 + }, + { + "epoch": 0.5531812560120929, + "grad_norm": 0.9940704703330994, + "learning_rate": 2.4724174816580803e-05, + "loss": 0.9113, + "step": 8051 + }, + { + "epoch": 0.5532499656451835, + "grad_norm": 1.073751449584961, + "learning_rate": 2.4718178868244915e-05, + "loss": 0.9226, + "step": 8052 + }, + { + "epoch": 0.553318675278274, + "grad_norm": 1.1208500862121582, + "learning_rate": 2.4712182936122027e-05, + "loss": 0.9736, + "step": 8053 + }, + { + "epoch": 0.5533873849113645, + "grad_norm": 1.1896405220031738, + "learning_rate": 2.470618702055709e-05, + "loss": 0.9586, + "step": 8054 + }, + { + "epoch": 0.5534560945444551, + "grad_norm": 1.0814722776412964, + "learning_rate": 2.4700191121895024e-05, + "loss": 0.9263, + "step": 8055 + }, + { + "epoch": 0.5535248041775457, + "grad_norm": 1.168330192565918, + "learning_rate": 2.469419524048078e-05, + "loss": 0.9658, + "step": 8056 + }, + { + "epoch": 0.5535935138106363, + "grad_norm": 1.1330550909042358, + "learning_rate": 2.4688199376659298e-05, + "loss": 0.9787, + "step": 8057 + }, + { + "epoch": 0.5536622234437268, + "grad_norm": 1.0366926193237305, + "learning_rate": 2.4682203530775518e-05, + "loss": 0.9088, + "step": 8058 + }, + { + "epoch": 0.5537309330768174, + "grad_norm": 1.1310526132583618, + "learning_rate": 2.4676207703174378e-05, + "loss": 0.9877, + "step": 8059 + }, + { + "epoch": 0.5537996427099079, + "grad_norm": 1.1160835027694702, + "learning_rate": 2.4670211894200814e-05, + "loss": 0.9638, + "step": 8060 + }, + { + "epoch": 0.5538683523429985, + "grad_norm": 1.0817314386367798, + "learning_rate": 2.4664216104199747e-05, + "loss": 0.9826, + "step": 8061 + }, + { + "epoch": 0.553937061976089, + "grad_norm": 1.1063231229782104, + "learning_rate": 2.4658220333516128e-05, + "loss": 0.9636, + "step": 8062 + }, + { + "epoch": 0.5540057716091796, + "grad_norm": 1.0531383752822876, + "learning_rate": 2.465222458249488e-05, + "loss": 0.9437, + "step": 8063 + }, + { + "epoch": 0.5540744812422702, + "grad_norm": 1.081064224243164, + "learning_rate": 2.4646228851480956e-05, + "loss": 0.9736, + "step": 8064 + }, + { + "epoch": 0.5541431908753607, + "grad_norm": 1.0922236442565918, + "learning_rate": 2.4640233140819257e-05, + "loss": 0.9537, + "step": 8065 + }, + { + "epoch": 0.5542119005084513, + "grad_norm": 1.118567943572998, + "learning_rate": 2.4634237450854726e-05, + "loss": 0.9541, + "step": 8066 + }, + { + "epoch": 0.5542806101415418, + "grad_norm": 1.1150243282318115, + "learning_rate": 2.462824178193229e-05, + "loss": 0.9975, + "step": 8067 + }, + { + "epoch": 0.5543493197746324, + "grad_norm": 1.1148229837417603, + "learning_rate": 2.462224613439688e-05, + "loss": 0.9857, + "step": 8068 + }, + { + "epoch": 0.554418029407723, + "grad_norm": 1.0545207262039185, + "learning_rate": 2.461625050859343e-05, + "loss": 0.9836, + "step": 8069 + }, + { + "epoch": 0.5544867390408135, + "grad_norm": 1.0999482870101929, + "learning_rate": 2.461025490486684e-05, + "loss": 1.0008, + "step": 8070 + }, + { + "epoch": 0.554555448673904, + "grad_norm": 1.1728122234344482, + "learning_rate": 2.4604259323562058e-05, + "loss": 0.9892, + "step": 8071 + }, + { + "epoch": 0.5546241583069946, + "grad_norm": 1.094092607498169, + "learning_rate": 2.4598263765023986e-05, + "loss": 0.9862, + "step": 8072 + }, + { + "epoch": 0.5546928679400852, + "grad_norm": 1.053130865097046, + "learning_rate": 2.4592268229597558e-05, + "loss": 0.9487, + "step": 8073 + }, + { + "epoch": 0.5547615775731758, + "grad_norm": 1.0294878482818604, + "learning_rate": 2.4586272717627698e-05, + "loss": 0.9451, + "step": 8074 + }, + { + "epoch": 0.5548302872062664, + "grad_norm": 1.1768956184387207, + "learning_rate": 2.4580277229459307e-05, + "loss": 0.9578, + "step": 8075 + }, + { + "epoch": 0.5548989968393568, + "grad_norm": 1.0130599737167358, + "learning_rate": 2.4574281765437312e-05, + "loss": 0.9372, + "step": 8076 + }, + { + "epoch": 0.5549677064724474, + "grad_norm": 1.0243488550186157, + "learning_rate": 2.4568286325906628e-05, + "loss": 0.964, + "step": 8077 + }, + { + "epoch": 0.555036416105538, + "grad_norm": 1.0336716175079346, + "learning_rate": 2.4562290911212166e-05, + "loss": 0.939, + "step": 8078 + }, + { + "epoch": 0.5551051257386286, + "grad_norm": 1.0565416812896729, + "learning_rate": 2.4556295521698855e-05, + "loss": 0.9723, + "step": 8079 + }, + { + "epoch": 0.5551738353717192, + "grad_norm": 1.072805404663086, + "learning_rate": 2.455030015771158e-05, + "loss": 0.9726, + "step": 8080 + }, + { + "epoch": 0.5552425450048096, + "grad_norm": 0.976097583770752, + "learning_rate": 2.454430481959526e-05, + "loss": 0.9338, + "step": 8081 + }, + { + "epoch": 0.5553112546379002, + "grad_norm": 1.0676982402801514, + "learning_rate": 2.453830950769481e-05, + "loss": 0.9484, + "step": 8082 + }, + { + "epoch": 0.5553799642709908, + "grad_norm": 1.016520619392395, + "learning_rate": 2.4532314222355135e-05, + "loss": 0.9471, + "step": 8083 + }, + { + "epoch": 0.5554486739040814, + "grad_norm": 1.0645785331726074, + "learning_rate": 2.4526318963921138e-05, + "loss": 0.9199, + "step": 8084 + }, + { + "epoch": 0.555517383537172, + "grad_norm": 1.0880485773086548, + "learning_rate": 2.4520323732737722e-05, + "loss": 0.9909, + "step": 8085 + }, + { + "epoch": 0.5555860931702624, + "grad_norm": 1.0387498140335083, + "learning_rate": 2.4514328529149786e-05, + "loss": 0.9381, + "step": 8086 + }, + { + "epoch": 0.555654802803353, + "grad_norm": 1.117708444595337, + "learning_rate": 2.4508333353502232e-05, + "loss": 0.9736, + "step": 8087 + }, + { + "epoch": 0.5557235124364436, + "grad_norm": 1.0348232984542847, + "learning_rate": 2.4502338206139957e-05, + "loss": 0.9787, + "step": 8088 + }, + { + "epoch": 0.5557922220695342, + "grad_norm": 1.0199726819992065, + "learning_rate": 2.4496343087407866e-05, + "loss": 0.9167, + "step": 8089 + }, + { + "epoch": 0.5558609317026247, + "grad_norm": 0.9877827763557434, + "learning_rate": 2.4490347997650862e-05, + "loss": 0.9365, + "step": 8090 + }, + { + "epoch": 0.5559296413357153, + "grad_norm": 1.1361820697784424, + "learning_rate": 2.4484352937213815e-05, + "loss": 0.9684, + "step": 8091 + }, + { + "epoch": 0.5559983509688058, + "grad_norm": 1.0887975692749023, + "learning_rate": 2.4478357906441627e-05, + "loss": 1.0008, + "step": 8092 + }, + { + "epoch": 0.5560670606018964, + "grad_norm": 1.0536137819290161, + "learning_rate": 2.4472362905679197e-05, + "loss": 0.9676, + "step": 8093 + }, + { + "epoch": 0.556135770234987, + "grad_norm": 1.0435787439346313, + "learning_rate": 2.44663679352714e-05, + "loss": 0.8991, + "step": 8094 + }, + { + "epoch": 0.5562044798680775, + "grad_norm": 1.0883809328079224, + "learning_rate": 2.446037299556314e-05, + "loss": 0.9489, + "step": 8095 + }, + { + "epoch": 0.5562731895011681, + "grad_norm": 1.121354579925537, + "learning_rate": 2.445437808689929e-05, + "loss": 0.9687, + "step": 8096 + }, + { + "epoch": 0.5563418991342586, + "grad_norm": 1.0936778783798218, + "learning_rate": 2.444838320962473e-05, + "loss": 1.0018, + "step": 8097 + }, + { + "epoch": 0.5564106087673492, + "grad_norm": 1.076822280883789, + "learning_rate": 2.4442388364084348e-05, + "loss": 0.9532, + "step": 8098 + }, + { + "epoch": 0.5564793184004397, + "grad_norm": 1.0799297094345093, + "learning_rate": 2.4436393550623023e-05, + "loss": 0.9461, + "step": 8099 + }, + { + "epoch": 0.5565480280335303, + "grad_norm": 1.0415747165679932, + "learning_rate": 2.4430398769585645e-05, + "loss": 0.9744, + "step": 8100 + }, + { + "epoch": 0.5566167376666209, + "grad_norm": 1.04105544090271, + "learning_rate": 2.442440402131707e-05, + "loss": 0.9833, + "step": 8101 + }, + { + "epoch": 0.5566854472997114, + "grad_norm": 1.0877386331558228, + "learning_rate": 2.4418409306162178e-05, + "loss": 0.933, + "step": 8102 + }, + { + "epoch": 0.556754156932802, + "grad_norm": 1.1091816425323486, + "learning_rate": 2.4412414624465843e-05, + "loss": 0.9417, + "step": 8103 + }, + { + "epoch": 0.5568228665658925, + "grad_norm": 1.0757808685302734, + "learning_rate": 2.4406419976572944e-05, + "loss": 1.0007, + "step": 8104 + }, + { + "epoch": 0.5568915761989831, + "grad_norm": 1.1144886016845703, + "learning_rate": 2.440042536282834e-05, + "loss": 0.9614, + "step": 8105 + }, + { + "epoch": 0.5569602858320737, + "grad_norm": 1.0520880222320557, + "learning_rate": 2.4394430783576898e-05, + "loss": 0.9198, + "step": 8106 + }, + { + "epoch": 0.5570289954651643, + "grad_norm": 1.2122749090194702, + "learning_rate": 2.4388436239163483e-05, + "loss": 1.0076, + "step": 8107 + }, + { + "epoch": 0.5570977050982547, + "grad_norm": 1.0533114671707153, + "learning_rate": 2.4382441729932957e-05, + "loss": 0.9701, + "step": 8108 + }, + { + "epoch": 0.5571664147313453, + "grad_norm": 1.0799119472503662, + "learning_rate": 2.437644725623018e-05, + "loss": 0.957, + "step": 8109 + }, + { + "epoch": 0.5572351243644359, + "grad_norm": 1.0445353984832764, + "learning_rate": 2.4370452818400024e-05, + "loss": 0.9512, + "step": 8110 + }, + { + "epoch": 0.5573038339975265, + "grad_norm": 1.0561144351959229, + "learning_rate": 2.4364458416787324e-05, + "loss": 0.9638, + "step": 8111 + }, + { + "epoch": 0.5573725436306171, + "grad_norm": 1.0578727722167969, + "learning_rate": 2.435846405173694e-05, + "loss": 0.9417, + "step": 8112 + }, + { + "epoch": 0.5574412532637075, + "grad_norm": 1.105652093887329, + "learning_rate": 2.435246972359373e-05, + "loss": 0.9733, + "step": 8113 + }, + { + "epoch": 0.5575099628967981, + "grad_norm": 1.1279033422470093, + "learning_rate": 2.4346475432702542e-05, + "loss": 0.9817, + "step": 8114 + }, + { + "epoch": 0.5575786725298887, + "grad_norm": 1.0050054788589478, + "learning_rate": 2.4340481179408236e-05, + "loss": 0.9886, + "step": 8115 + }, + { + "epoch": 0.5576473821629793, + "grad_norm": 1.0108942985534668, + "learning_rate": 2.433448696405563e-05, + "loss": 0.94, + "step": 8116 + }, + { + "epoch": 0.5577160917960698, + "grad_norm": 1.0638384819030762, + "learning_rate": 2.432849278698959e-05, + "loss": 0.9713, + "step": 8117 + }, + { + "epoch": 0.5577848014291603, + "grad_norm": 1.0707703828811646, + "learning_rate": 2.4322498648554955e-05, + "loss": 0.9801, + "step": 8118 + }, + { + "epoch": 0.5578535110622509, + "grad_norm": 1.0833581686019897, + "learning_rate": 2.431650454909655e-05, + "loss": 0.9703, + "step": 8119 + }, + { + "epoch": 0.5579222206953415, + "grad_norm": 1.0555773973464966, + "learning_rate": 2.4310510488959233e-05, + "loss": 0.9461, + "step": 8120 + }, + { + "epoch": 0.557990930328432, + "grad_norm": 1.006712555885315, + "learning_rate": 2.430451646848782e-05, + "loss": 0.9555, + "step": 8121 + }, + { + "epoch": 0.5580596399615226, + "grad_norm": 1.0764027833938599, + "learning_rate": 2.4298522488027148e-05, + "loss": 0.9526, + "step": 8122 + }, + { + "epoch": 0.5581283495946132, + "grad_norm": 1.0061030387878418, + "learning_rate": 2.429252854792205e-05, + "loss": 0.9381, + "step": 8123 + }, + { + "epoch": 0.5581970592277037, + "grad_norm": 1.0468506813049316, + "learning_rate": 2.428653464851735e-05, + "loss": 0.9653, + "step": 8124 + }, + { + "epoch": 0.5582657688607943, + "grad_norm": 1.1586140394210815, + "learning_rate": 2.4280540790157892e-05, + "loss": 0.9622, + "step": 8125 + }, + { + "epoch": 0.5583344784938848, + "grad_norm": 1.0630632638931274, + "learning_rate": 2.427454697318847e-05, + "loss": 0.9293, + "step": 8126 + }, + { + "epoch": 0.5584031881269754, + "grad_norm": 1.094009518623352, + "learning_rate": 2.4268553197953922e-05, + "loss": 0.9667, + "step": 8127 + }, + { + "epoch": 0.558471897760066, + "grad_norm": 1.1208690404891968, + "learning_rate": 2.426255946479906e-05, + "loss": 0.9709, + "step": 8128 + }, + { + "epoch": 0.5585406073931565, + "grad_norm": 1.1344331502914429, + "learning_rate": 2.4256565774068708e-05, + "loss": 0.977, + "step": 8129 + }, + { + "epoch": 0.558609317026247, + "grad_norm": 1.0961017608642578, + "learning_rate": 2.4250572126107673e-05, + "loss": 0.966, + "step": 8130 + }, + { + "epoch": 0.5586780266593376, + "grad_norm": 1.1083478927612305, + "learning_rate": 2.4244578521260767e-05, + "loss": 0.984, + "step": 8131 + }, + { + "epoch": 0.5587467362924282, + "grad_norm": 1.0758135318756104, + "learning_rate": 2.4238584959872794e-05, + "loss": 0.9743, + "step": 8132 + }, + { + "epoch": 0.5588154459255188, + "grad_norm": 1.1373931169509888, + "learning_rate": 2.423259144228856e-05, + "loss": 0.9239, + "step": 8133 + }, + { + "epoch": 0.5588841555586093, + "grad_norm": 1.0251411199569702, + "learning_rate": 2.4226597968852878e-05, + "loss": 0.9783, + "step": 8134 + }, + { + "epoch": 0.5589528651916998, + "grad_norm": 1.0151883363723755, + "learning_rate": 2.4220604539910548e-05, + "loss": 0.909, + "step": 8135 + }, + { + "epoch": 0.5590215748247904, + "grad_norm": 1.1652894020080566, + "learning_rate": 2.4214611155806358e-05, + "loss": 0.9703, + "step": 8136 + }, + { + "epoch": 0.559090284457881, + "grad_norm": 1.0623139142990112, + "learning_rate": 2.4208617816885107e-05, + "loss": 0.963, + "step": 8137 + }, + { + "epoch": 0.5591589940909716, + "grad_norm": 1.081092119216919, + "learning_rate": 2.420262452349159e-05, + "loss": 0.9878, + "step": 8138 + }, + { + "epoch": 0.5592277037240622, + "grad_norm": 1.054573893547058, + "learning_rate": 2.4196631275970598e-05, + "loss": 0.971, + "step": 8139 + }, + { + "epoch": 0.5592964133571526, + "grad_norm": 1.037084937095642, + "learning_rate": 2.4190638074666933e-05, + "loss": 0.9465, + "step": 8140 + }, + { + "epoch": 0.5593651229902432, + "grad_norm": 1.1339244842529297, + "learning_rate": 2.4184644919925353e-05, + "loss": 0.9683, + "step": 8141 + }, + { + "epoch": 0.5594338326233338, + "grad_norm": 1.0496783256530762, + "learning_rate": 2.417865181209066e-05, + "loss": 0.9541, + "step": 8142 + }, + { + "epoch": 0.5595025422564244, + "grad_norm": 1.0292783975601196, + "learning_rate": 2.4172658751507622e-05, + "loss": 0.9566, + "step": 8143 + }, + { + "epoch": 0.559571251889515, + "grad_norm": 1.1177914142608643, + "learning_rate": 2.4166665738521018e-05, + "loss": 0.965, + "step": 8144 + }, + { + "epoch": 0.5596399615226054, + "grad_norm": 1.1095287799835205, + "learning_rate": 2.4160672773475628e-05, + "loss": 0.9693, + "step": 8145 + }, + { + "epoch": 0.559708671155696, + "grad_norm": 1.0822004079818726, + "learning_rate": 2.4154679856716235e-05, + "loss": 0.9728, + "step": 8146 + }, + { + "epoch": 0.5597773807887866, + "grad_norm": 1.0106550455093384, + "learning_rate": 2.414868698858758e-05, + "loss": 0.9366, + "step": 8147 + }, + { + "epoch": 0.5598460904218772, + "grad_norm": 1.0722202062606812, + "learning_rate": 2.4142694169434445e-05, + "loss": 0.9837, + "step": 8148 + }, + { + "epoch": 0.5599148000549677, + "grad_norm": 1.030537724494934, + "learning_rate": 2.413670139960159e-05, + "loss": 0.9479, + "step": 8149 + }, + { + "epoch": 0.5599835096880582, + "grad_norm": 1.0537415742874146, + "learning_rate": 2.413070867943378e-05, + "loss": 0.9386, + "step": 8150 + }, + { + "epoch": 0.5600522193211488, + "grad_norm": 1.0460091829299927, + "learning_rate": 2.412471600927578e-05, + "loss": 0.9535, + "step": 8151 + }, + { + "epoch": 0.5601209289542394, + "grad_norm": 1.0303634405136108, + "learning_rate": 2.4118723389472322e-05, + "loss": 0.9748, + "step": 8152 + }, + { + "epoch": 0.56018963858733, + "grad_norm": 1.089834213256836, + "learning_rate": 2.4112730820368176e-05, + "loss": 0.9745, + "step": 8153 + }, + { + "epoch": 0.5602583482204205, + "grad_norm": 1.037977933883667, + "learning_rate": 2.410673830230808e-05, + "loss": 0.9443, + "step": 8154 + }, + { + "epoch": 0.5603270578535111, + "grad_norm": 1.0867059230804443, + "learning_rate": 2.4100745835636784e-05, + "loss": 0.9504, + "step": 8155 + }, + { + "epoch": 0.5603957674866016, + "grad_norm": 1.059528112411499, + "learning_rate": 2.4094753420699042e-05, + "loss": 0.921, + "step": 8156 + }, + { + "epoch": 0.5604644771196922, + "grad_norm": 1.1040730476379395, + "learning_rate": 2.4088761057839576e-05, + "loss": 0.9643, + "step": 8157 + }, + { + "epoch": 0.5605331867527827, + "grad_norm": 1.0722639560699463, + "learning_rate": 2.4082768747403128e-05, + "loss": 0.9545, + "step": 8158 + }, + { + "epoch": 0.5606018963858733, + "grad_norm": 1.0466264486312866, + "learning_rate": 2.4076776489734434e-05, + "loss": 0.982, + "step": 8159 + }, + { + "epoch": 0.5606706060189639, + "grad_norm": 1.0231505632400513, + "learning_rate": 2.407078428517823e-05, + "loss": 0.9333, + "step": 8160 + }, + { + "epoch": 0.5607393156520544, + "grad_norm": 1.0716018676757812, + "learning_rate": 2.4064792134079246e-05, + "loss": 0.915, + "step": 8161 + }, + { + "epoch": 0.560808025285145, + "grad_norm": 1.0310906171798706, + "learning_rate": 2.405880003678219e-05, + "loss": 0.9407, + "step": 8162 + }, + { + "epoch": 0.5608767349182355, + "grad_norm": 1.0825937986373901, + "learning_rate": 2.4052807993631796e-05, + "loss": 0.9777, + "step": 8163 + }, + { + "epoch": 0.5609454445513261, + "grad_norm": 1.0572245121002197, + "learning_rate": 2.4046816004972777e-05, + "loss": 0.96, + "step": 8164 + }, + { + "epoch": 0.5610141541844167, + "grad_norm": 1.0903772115707397, + "learning_rate": 2.404082407114986e-05, + "loss": 0.9469, + "step": 8165 + }, + { + "epoch": 0.5610828638175072, + "grad_norm": 1.071311593055725, + "learning_rate": 2.403483219250775e-05, + "loss": 0.9522, + "step": 8166 + }, + { + "epoch": 0.5611515734505977, + "grad_norm": 1.1685760021209717, + "learning_rate": 2.4028840369391155e-05, + "loss": 0.9828, + "step": 8167 + }, + { + "epoch": 0.5612202830836883, + "grad_norm": 1.1439348459243774, + "learning_rate": 2.4022848602144776e-05, + "loss": 1.0152, + "step": 8168 + }, + { + "epoch": 0.5612889927167789, + "grad_norm": 1.038857340812683, + "learning_rate": 2.401685689111332e-05, + "loss": 0.9441, + "step": 8169 + }, + { + "epoch": 0.5613577023498695, + "grad_norm": 1.0485121011734009, + "learning_rate": 2.4010865236641488e-05, + "loss": 0.966, + "step": 8170 + }, + { + "epoch": 0.5614264119829601, + "grad_norm": 1.0293594598770142, + "learning_rate": 2.400487363907399e-05, + "loss": 0.9606, + "step": 8171 + }, + { + "epoch": 0.5614951216160505, + "grad_norm": 1.004468560218811, + "learning_rate": 2.399888209875549e-05, + "loss": 0.9242, + "step": 8172 + }, + { + "epoch": 0.5615638312491411, + "grad_norm": 1.03374183177948, + "learning_rate": 2.3992890616030692e-05, + "loss": 0.9448, + "step": 8173 + }, + { + "epoch": 0.5616325408822317, + "grad_norm": 1.040827751159668, + "learning_rate": 2.3986899191244284e-05, + "loss": 0.9533, + "step": 8174 + }, + { + "epoch": 0.5617012505153223, + "grad_norm": 1.0427873134613037, + "learning_rate": 2.3980907824740946e-05, + "loss": 0.9383, + "step": 8175 + }, + { + "epoch": 0.5617699601484129, + "grad_norm": 1.0685869455337524, + "learning_rate": 2.3974916516865376e-05, + "loss": 0.9751, + "step": 8176 + }, + { + "epoch": 0.5618386697815033, + "grad_norm": 0.9738472104072571, + "learning_rate": 2.3968925267962215e-05, + "loss": 0.9184, + "step": 8177 + }, + { + "epoch": 0.5619073794145939, + "grad_norm": 1.0434461832046509, + "learning_rate": 2.396293407837617e-05, + "loss": 0.9799, + "step": 8178 + }, + { + "epoch": 0.5619760890476845, + "grad_norm": 1.0850834846496582, + "learning_rate": 2.3956942948451884e-05, + "loss": 0.9352, + "step": 8179 + }, + { + "epoch": 0.5620447986807751, + "grad_norm": 1.0543718338012695, + "learning_rate": 2.3950951878534032e-05, + "loss": 0.9782, + "step": 8180 + }, + { + "epoch": 0.5621135083138656, + "grad_norm": 1.0852339267730713, + "learning_rate": 2.394496086896729e-05, + "loss": 0.9501, + "step": 8181 + }, + { + "epoch": 0.5621822179469561, + "grad_norm": 1.044064998626709, + "learning_rate": 2.39389699200963e-05, + "loss": 0.9496, + "step": 8182 + }, + { + "epoch": 0.5622509275800467, + "grad_norm": 1.1145339012145996, + "learning_rate": 2.3932979032265716e-05, + "loss": 0.9739, + "step": 8183 + }, + { + "epoch": 0.5623196372131373, + "grad_norm": 1.1097880601882935, + "learning_rate": 2.3926988205820204e-05, + "loss": 0.9782, + "step": 8184 + }, + { + "epoch": 0.5623883468462278, + "grad_norm": 1.0128103494644165, + "learning_rate": 2.3920997441104407e-05, + "loss": 0.9303, + "step": 8185 + }, + { + "epoch": 0.5624570564793184, + "grad_norm": 1.0686123371124268, + "learning_rate": 2.3915006738462975e-05, + "loss": 0.9458, + "step": 8186 + }, + { + "epoch": 0.562525766112409, + "grad_norm": 1.0777051448822021, + "learning_rate": 2.3909016098240536e-05, + "loss": 0.9116, + "step": 8187 + }, + { + "epoch": 0.5625944757454995, + "grad_norm": 1.0630851984024048, + "learning_rate": 2.3903025520781734e-05, + "loss": 0.9797, + "step": 8188 + }, + { + "epoch": 0.5626631853785901, + "grad_norm": 1.1154659986495972, + "learning_rate": 2.3897035006431214e-05, + "loss": 0.9809, + "step": 8189 + }, + { + "epoch": 0.5627318950116806, + "grad_norm": 1.0571075677871704, + "learning_rate": 2.3891044555533588e-05, + "loss": 0.9555, + "step": 8190 + }, + { + "epoch": 0.5628006046447712, + "grad_norm": 0.9950230121612549, + "learning_rate": 2.38850541684335e-05, + "loss": 0.9099, + "step": 8191 + }, + { + "epoch": 0.5628693142778618, + "grad_norm": 1.0704114437103271, + "learning_rate": 2.3879063845475567e-05, + "loss": 0.9276, + "step": 8192 + }, + { + "epoch": 0.5629380239109523, + "grad_norm": 1.041064739227295, + "learning_rate": 2.38730735870044e-05, + "loss": 0.9802, + "step": 8193 + }, + { + "epoch": 0.5630067335440428, + "grad_norm": 1.0470445156097412, + "learning_rate": 2.3867083393364622e-05, + "loss": 0.9125, + "step": 8194 + }, + { + "epoch": 0.5630754431771334, + "grad_norm": 1.067085862159729, + "learning_rate": 2.3861093264900845e-05, + "loss": 0.9415, + "step": 8195 + }, + { + "epoch": 0.563144152810224, + "grad_norm": 1.1292307376861572, + "learning_rate": 2.3855103201957692e-05, + "loss": 0.9949, + "step": 8196 + }, + { + "epoch": 0.5632128624433146, + "grad_norm": 1.0737261772155762, + "learning_rate": 2.384911320487974e-05, + "loss": 0.9775, + "step": 8197 + }, + { + "epoch": 0.5632815720764051, + "grad_norm": 1.1423487663269043, + "learning_rate": 2.3843123274011606e-05, + "loss": 0.9951, + "step": 8198 + }, + { + "epoch": 0.5633502817094956, + "grad_norm": 1.2046170234680176, + "learning_rate": 2.3837133409697883e-05, + "loss": 0.9775, + "step": 8199 + }, + { + "epoch": 0.5634189913425862, + "grad_norm": 1.0653892755508423, + "learning_rate": 2.3831143612283174e-05, + "loss": 0.9698, + "step": 8200 + }, + { + "epoch": 0.5634877009756768, + "grad_norm": 1.1164482831954956, + "learning_rate": 2.3825153882112053e-05, + "loss": 0.956, + "step": 8201 + }, + { + "epoch": 0.5635564106087674, + "grad_norm": 1.1007084846496582, + "learning_rate": 2.3819164219529116e-05, + "loss": 0.9371, + "step": 8202 + }, + { + "epoch": 0.563625120241858, + "grad_norm": 1.069344162940979, + "learning_rate": 2.3813174624878946e-05, + "loss": 0.9616, + "step": 8203 + }, + { + "epoch": 0.5636938298749484, + "grad_norm": 1.0416959524154663, + "learning_rate": 2.3807185098506107e-05, + "loss": 0.9799, + "step": 8204 + }, + { + "epoch": 0.563762539508039, + "grad_norm": 1.1118665933609009, + "learning_rate": 2.3801195640755184e-05, + "loss": 0.9854, + "step": 8205 + }, + { + "epoch": 0.5638312491411296, + "grad_norm": 1.0449762344360352, + "learning_rate": 2.379520625197074e-05, + "loss": 0.9637, + "step": 8206 + }, + { + "epoch": 0.5638999587742202, + "grad_norm": 1.092163324356079, + "learning_rate": 2.378921693249736e-05, + "loss": 0.936, + "step": 8207 + }, + { + "epoch": 0.5639686684073107, + "grad_norm": 1.0535589456558228, + "learning_rate": 2.3783227682679584e-05, + "loss": 0.9402, + "step": 8208 + }, + { + "epoch": 0.5640373780404012, + "grad_norm": 1.007072925567627, + "learning_rate": 2.3777238502861977e-05, + "loss": 0.965, + "step": 8209 + }, + { + "epoch": 0.5641060876734918, + "grad_norm": 1.0507606267929077, + "learning_rate": 2.3771249393389088e-05, + "loss": 0.947, + "step": 8210 + }, + { + "epoch": 0.5641747973065824, + "grad_norm": 1.075488805770874, + "learning_rate": 2.3765260354605483e-05, + "loss": 1.0001, + "step": 8211 + }, + { + "epoch": 0.564243506939673, + "grad_norm": 1.0376499891281128, + "learning_rate": 2.3759271386855694e-05, + "loss": 0.9482, + "step": 8212 + }, + { + "epoch": 0.5643122165727635, + "grad_norm": 1.0577564239501953, + "learning_rate": 2.3753282490484263e-05, + "loss": 0.936, + "step": 8213 + }, + { + "epoch": 0.564380926205854, + "grad_norm": 1.0571439266204834, + "learning_rate": 2.3747293665835733e-05, + "loss": 0.9735, + "step": 8214 + }, + { + "epoch": 0.5644496358389446, + "grad_norm": 1.0209808349609375, + "learning_rate": 2.374130491325463e-05, + "loss": 0.9028, + "step": 8215 + }, + { + "epoch": 0.5645183454720352, + "grad_norm": 1.0783004760742188, + "learning_rate": 2.3735316233085486e-05, + "loss": 0.9316, + "step": 8216 + }, + { + "epoch": 0.5645870551051257, + "grad_norm": 1.152384877204895, + "learning_rate": 2.3729327625672844e-05, + "loss": 0.9956, + "step": 8217 + }, + { + "epoch": 0.5646557647382163, + "grad_norm": 1.1121267080307007, + "learning_rate": 2.372333909136119e-05, + "loss": 0.9437, + "step": 8218 + }, + { + "epoch": 0.5647244743713069, + "grad_norm": 1.0828849077224731, + "learning_rate": 2.3717350630495067e-05, + "loss": 0.9374, + "step": 8219 + }, + { + "epoch": 0.5647931840043974, + "grad_norm": 1.0694189071655273, + "learning_rate": 2.3711362243418977e-05, + "loss": 0.9488, + "step": 8220 + }, + { + "epoch": 0.564861893637488, + "grad_norm": 1.0938282012939453, + "learning_rate": 2.370537393047743e-05, + "loss": 0.9245, + "step": 8221 + }, + { + "epoch": 0.5649306032705785, + "grad_norm": 1.0084342956542969, + "learning_rate": 2.3699385692014945e-05, + "loss": 0.9185, + "step": 8222 + }, + { + "epoch": 0.5649993129036691, + "grad_norm": 1.033704400062561, + "learning_rate": 2.3693397528375996e-05, + "loss": 0.9448, + "step": 8223 + }, + { + "epoch": 0.5650680225367597, + "grad_norm": 1.0529592037200928, + "learning_rate": 2.368740943990509e-05, + "loss": 0.9879, + "step": 8224 + }, + { + "epoch": 0.5651367321698502, + "grad_norm": 0.9760762453079224, + "learning_rate": 2.3681421426946727e-05, + "loss": 0.946, + "step": 8225 + }, + { + "epoch": 0.5652054418029407, + "grad_norm": 1.0866272449493408, + "learning_rate": 2.3675433489845378e-05, + "loss": 0.9755, + "step": 8226 + }, + { + "epoch": 0.5652741514360313, + "grad_norm": 1.0656784772872925, + "learning_rate": 2.3669445628945542e-05, + "loss": 0.9594, + "step": 8227 + }, + { + "epoch": 0.5653428610691219, + "grad_norm": 1.073761224746704, + "learning_rate": 2.366345784459168e-05, + "loss": 0.9811, + "step": 8228 + }, + { + "epoch": 0.5654115707022125, + "grad_norm": 1.0290896892547607, + "learning_rate": 2.365747013712827e-05, + "loss": 0.9562, + "step": 8229 + }, + { + "epoch": 0.5654802803353031, + "grad_norm": 1.0164068937301636, + "learning_rate": 2.3651482506899786e-05, + "loss": 0.9241, + "step": 8230 + }, + { + "epoch": 0.5655489899683935, + "grad_norm": 1.0990172624588013, + "learning_rate": 2.3645494954250692e-05, + "loss": 0.9788, + "step": 8231 + }, + { + "epoch": 0.5656176996014841, + "grad_norm": 1.141615867614746, + "learning_rate": 2.3639507479525455e-05, + "loss": 0.974, + "step": 8232 + }, + { + "epoch": 0.5656864092345747, + "grad_norm": 1.0695083141326904, + "learning_rate": 2.3633520083068516e-05, + "loss": 0.9503, + "step": 8233 + }, + { + "epoch": 0.5657551188676653, + "grad_norm": 1.147944688796997, + "learning_rate": 2.362753276522433e-05, + "loss": 0.9745, + "step": 8234 + }, + { + "epoch": 0.5658238285007559, + "grad_norm": 1.0997486114501953, + "learning_rate": 2.362154552633735e-05, + "loss": 0.9569, + "step": 8235 + }, + { + "epoch": 0.5658925381338463, + "grad_norm": 1.039139747619629, + "learning_rate": 2.361555836675202e-05, + "loss": 0.9517, + "step": 8236 + }, + { + "epoch": 0.5659612477669369, + "grad_norm": 1.067442536354065, + "learning_rate": 2.360957128681277e-05, + "loss": 0.9782, + "step": 8237 + }, + { + "epoch": 0.5660299574000275, + "grad_norm": 1.043867826461792, + "learning_rate": 2.360358428686404e-05, + "loss": 0.9298, + "step": 8238 + }, + { + "epoch": 0.5660986670331181, + "grad_norm": 1.0990631580352783, + "learning_rate": 2.359759736725025e-05, + "loss": 1.0035, + "step": 8239 + }, + { + "epoch": 0.5661673766662086, + "grad_norm": 1.1794650554656982, + "learning_rate": 2.359161052831583e-05, + "loss": 1.0122, + "step": 8240 + }, + { + "epoch": 0.5662360862992991, + "grad_norm": 1.0387215614318848, + "learning_rate": 2.358562377040519e-05, + "loss": 0.9688, + "step": 8241 + }, + { + "epoch": 0.5663047959323897, + "grad_norm": 1.1177064180374146, + "learning_rate": 2.3579637093862765e-05, + "loss": 1.0034, + "step": 8242 + }, + { + "epoch": 0.5663735055654803, + "grad_norm": 0.9707015752792358, + "learning_rate": 2.3573650499032945e-05, + "loss": 0.9337, + "step": 8243 + }, + { + "epoch": 0.5664422151985709, + "grad_norm": 1.0212297439575195, + "learning_rate": 2.3567663986260143e-05, + "loss": 0.9591, + "step": 8244 + }, + { + "epoch": 0.5665109248316614, + "grad_norm": 1.1251147985458374, + "learning_rate": 2.3561677555888752e-05, + "loss": 0.9935, + "step": 8245 + }, + { + "epoch": 0.566579634464752, + "grad_norm": 1.0507246255874634, + "learning_rate": 2.355569120826318e-05, + "loss": 0.9364, + "step": 8246 + }, + { + "epoch": 0.5666483440978425, + "grad_norm": 1.0454250574111938, + "learning_rate": 2.354970494372782e-05, + "loss": 0.9111, + "step": 8247 + }, + { + "epoch": 0.5667170537309331, + "grad_norm": 1.0304062366485596, + "learning_rate": 2.354371876262704e-05, + "loss": 0.9522, + "step": 8248 + }, + { + "epoch": 0.5667857633640236, + "grad_norm": 1.0388457775115967, + "learning_rate": 2.353773266530524e-05, + "loss": 0.9427, + "step": 8249 + }, + { + "epoch": 0.5668544729971142, + "grad_norm": 1.0937696695327759, + "learning_rate": 2.353174665210678e-05, + "loss": 0.9482, + "step": 8250 + }, + { + "epoch": 0.5669231826302048, + "grad_norm": 1.0868942737579346, + "learning_rate": 2.352576072337604e-05, + "loss": 0.9648, + "step": 8251 + }, + { + "epoch": 0.5669918922632953, + "grad_norm": 1.0043271780014038, + "learning_rate": 2.3519774879457396e-05, + "loss": 0.9435, + "step": 8252 + }, + { + "epoch": 0.5670606018963859, + "grad_norm": 1.139588475227356, + "learning_rate": 2.3513789120695193e-05, + "loss": 1.0116, + "step": 8253 + }, + { + "epoch": 0.5671293115294764, + "grad_norm": 1.0718952417373657, + "learning_rate": 2.3507803447433794e-05, + "loss": 0.9573, + "step": 8254 + }, + { + "epoch": 0.567198021162567, + "grad_norm": 1.073157787322998, + "learning_rate": 2.350181786001755e-05, + "loss": 0.9613, + "step": 8255 + }, + { + "epoch": 0.5672667307956576, + "grad_norm": 1.100189208984375, + "learning_rate": 2.3495832358790814e-05, + "loss": 0.9862, + "step": 8256 + }, + { + "epoch": 0.5673354404287481, + "grad_norm": 1.083743691444397, + "learning_rate": 2.3489846944097925e-05, + "loss": 0.9468, + "step": 8257 + }, + { + "epoch": 0.5674041500618386, + "grad_norm": 1.126269817352295, + "learning_rate": 2.3483861616283233e-05, + "loss": 0.9683, + "step": 8258 + }, + { + "epoch": 0.5674728596949292, + "grad_norm": 1.0917832851409912, + "learning_rate": 2.3477876375691047e-05, + "loss": 0.9411, + "step": 8259 + }, + { + "epoch": 0.5675415693280198, + "grad_norm": 1.1787593364715576, + "learning_rate": 2.3471891222665707e-05, + "loss": 0.9564, + "step": 8260 + }, + { + "epoch": 0.5676102789611104, + "grad_norm": 1.0611505508422852, + "learning_rate": 2.346590615755153e-05, + "loss": 0.9542, + "step": 8261 + }, + { + "epoch": 0.567678988594201, + "grad_norm": 1.002738356590271, + "learning_rate": 2.3459921180692835e-05, + "loss": 0.9342, + "step": 8262 + }, + { + "epoch": 0.5677476982272914, + "grad_norm": 1.0048048496246338, + "learning_rate": 2.345393629243395e-05, + "loss": 0.9421, + "step": 8263 + }, + { + "epoch": 0.567816407860382, + "grad_norm": 1.0114601850509644, + "learning_rate": 2.3447951493119152e-05, + "loss": 0.9644, + "step": 8264 + }, + { + "epoch": 0.5678851174934726, + "grad_norm": 1.0609689950942993, + "learning_rate": 2.344196678309276e-05, + "loss": 0.9761, + "step": 8265 + }, + { + "epoch": 0.5679538271265632, + "grad_norm": 1.0761195421218872, + "learning_rate": 2.3435982162699072e-05, + "loss": 0.9246, + "step": 8266 + }, + { + "epoch": 0.5680225367596538, + "grad_norm": 1.0754047632217407, + "learning_rate": 2.3429997632282373e-05, + "loss": 0.9901, + "step": 8267 + }, + { + "epoch": 0.5680912463927442, + "grad_norm": 1.0595753192901611, + "learning_rate": 2.3424013192186965e-05, + "loss": 0.9479, + "step": 8268 + }, + { + "epoch": 0.5681599560258348, + "grad_norm": 0.9998588562011719, + "learning_rate": 2.3418028842757106e-05, + "loss": 0.9163, + "step": 8269 + }, + { + "epoch": 0.5682286656589254, + "grad_norm": 1.1361825466156006, + "learning_rate": 2.3412044584337082e-05, + "loss": 0.982, + "step": 8270 + }, + { + "epoch": 0.568297375292016, + "grad_norm": 1.008378505706787, + "learning_rate": 2.340606041727117e-05, + "loss": 0.9459, + "step": 8271 + }, + { + "epoch": 0.5683660849251065, + "grad_norm": 1.0667859315872192, + "learning_rate": 2.3400076341903627e-05, + "loss": 0.9455, + "step": 8272 + }, + { + "epoch": 0.568434794558197, + "grad_norm": 1.1055630445480347, + "learning_rate": 2.339409235857872e-05, + "loss": 0.9603, + "step": 8273 + }, + { + "epoch": 0.5685035041912876, + "grad_norm": 1.0928668975830078, + "learning_rate": 2.3388108467640703e-05, + "loss": 0.925, + "step": 8274 + }, + { + "epoch": 0.5685722138243782, + "grad_norm": 1.0208711624145508, + "learning_rate": 2.3382124669433814e-05, + "loss": 0.992, + "step": 8275 + }, + { + "epoch": 0.5686409234574687, + "grad_norm": 1.1288968324661255, + "learning_rate": 2.3376140964302308e-05, + "loss": 0.9508, + "step": 8276 + }, + { + "epoch": 0.5687096330905593, + "grad_norm": 1.0308289527893066, + "learning_rate": 2.3370157352590425e-05, + "loss": 0.9311, + "step": 8277 + }, + { + "epoch": 0.5687783427236499, + "grad_norm": 1.0779651403427124, + "learning_rate": 2.3364173834642402e-05, + "loss": 0.9517, + "step": 8278 + }, + { + "epoch": 0.5688470523567404, + "grad_norm": 0.9942987561225891, + "learning_rate": 2.3358190410802454e-05, + "loss": 0.9384, + "step": 8279 + }, + { + "epoch": 0.568915761989831, + "grad_norm": 1.0196291208267212, + "learning_rate": 2.3352207081414807e-05, + "loss": 0.9394, + "step": 8280 + }, + { + "epoch": 0.5689844716229215, + "grad_norm": 1.0729445219039917, + "learning_rate": 2.3346223846823682e-05, + "loss": 0.9686, + "step": 8281 + }, + { + "epoch": 0.5690531812560121, + "grad_norm": 1.0495648384094238, + "learning_rate": 2.33402407073733e-05, + "loss": 0.9691, + "step": 8282 + }, + { + "epoch": 0.5691218908891027, + "grad_norm": 1.0347814559936523, + "learning_rate": 2.3334257663407856e-05, + "loss": 0.9681, + "step": 8283 + }, + { + "epoch": 0.5691906005221932, + "grad_norm": 1.100172519683838, + "learning_rate": 2.332827471527155e-05, + "loss": 0.9795, + "step": 8284 + }, + { + "epoch": 0.5692593101552837, + "grad_norm": 1.0817933082580566, + "learning_rate": 2.332229186330858e-05, + "loss": 0.9673, + "step": 8285 + }, + { + "epoch": 0.5693280197883743, + "grad_norm": 1.0720397233963013, + "learning_rate": 2.3316309107863134e-05, + "loss": 0.9626, + "step": 8286 + }, + { + "epoch": 0.5693967294214649, + "grad_norm": 1.0736145973205566, + "learning_rate": 2.3310326449279398e-05, + "loss": 0.9551, + "step": 8287 + }, + { + "epoch": 0.5694654390545555, + "grad_norm": 1.1094971895217896, + "learning_rate": 2.3304343887901563e-05, + "loss": 0.9787, + "step": 8288 + }, + { + "epoch": 0.569534148687646, + "grad_norm": 1.0280027389526367, + "learning_rate": 2.3298361424073778e-05, + "loss": 0.9673, + "step": 8289 + }, + { + "epoch": 0.5696028583207365, + "grad_norm": 1.09266996383667, + "learning_rate": 2.3292379058140223e-05, + "loss": 1.0118, + "step": 8290 + }, + { + "epoch": 0.5696715679538271, + "grad_norm": 1.0636800527572632, + "learning_rate": 2.328639679044506e-05, + "loss": 0.9823, + "step": 8291 + }, + { + "epoch": 0.5697402775869177, + "grad_norm": 1.0403416156768799, + "learning_rate": 2.3280414621332446e-05, + "loss": 0.9344, + "step": 8292 + }, + { + "epoch": 0.5698089872200083, + "grad_norm": 1.0879924297332764, + "learning_rate": 2.327443255114654e-05, + "loss": 0.939, + "step": 8293 + }, + { + "epoch": 0.5698776968530989, + "grad_norm": 1.0373705625534058, + "learning_rate": 2.3268450580231466e-05, + "loss": 0.969, + "step": 8294 + }, + { + "epoch": 0.5699464064861893, + "grad_norm": 1.054081678390503, + "learning_rate": 2.3262468708931376e-05, + "loss": 0.9262, + "step": 8295 + }, + { + "epoch": 0.5700151161192799, + "grad_norm": 1.079153060913086, + "learning_rate": 2.3256486937590404e-05, + "loss": 0.9969, + "step": 8296 + }, + { + "epoch": 0.5700838257523705, + "grad_norm": 1.0820362567901611, + "learning_rate": 2.325050526655267e-05, + "loss": 0.9559, + "step": 8297 + }, + { + "epoch": 0.5701525353854611, + "grad_norm": 1.05488121509552, + "learning_rate": 2.3244523696162312e-05, + "loss": 0.9535, + "step": 8298 + }, + { + "epoch": 0.5702212450185516, + "grad_norm": 0.9982771873474121, + "learning_rate": 2.323854222676343e-05, + "loss": 0.9417, + "step": 8299 + }, + { + "epoch": 0.5702899546516421, + "grad_norm": 1.0992047786712646, + "learning_rate": 2.3232560858700137e-05, + "loss": 0.9364, + "step": 8300 + }, + { + "epoch": 0.5703586642847327, + "grad_norm": 1.0366100072860718, + "learning_rate": 2.3226579592316538e-05, + "loss": 0.9717, + "step": 8301 + }, + { + "epoch": 0.5704273739178233, + "grad_norm": 1.0678000450134277, + "learning_rate": 2.3220598427956736e-05, + "loss": 0.9451, + "step": 8302 + }, + { + "epoch": 0.5704960835509139, + "grad_norm": 1.0439618825912476, + "learning_rate": 2.3214617365964833e-05, + "loss": 0.9375, + "step": 8303 + }, + { + "epoch": 0.5705647931840044, + "grad_norm": 1.006255865097046, + "learning_rate": 2.320863640668489e-05, + "loss": 0.9315, + "step": 8304 + }, + { + "epoch": 0.5706335028170949, + "grad_norm": 1.0809487104415894, + "learning_rate": 2.3202655550461004e-05, + "loss": 0.995, + "step": 8305 + }, + { + "epoch": 0.5707022124501855, + "grad_norm": 1.1241159439086914, + "learning_rate": 2.319667479763725e-05, + "loss": 1.0097, + "step": 8306 + }, + { + "epoch": 0.5707709220832761, + "grad_norm": 1.028143048286438, + "learning_rate": 2.31906941485577e-05, + "loss": 0.9704, + "step": 8307 + }, + { + "epoch": 0.5708396317163666, + "grad_norm": 1.086441159248352, + "learning_rate": 2.3184713603566405e-05, + "loss": 0.9329, + "step": 8308 + }, + { + "epoch": 0.5709083413494572, + "grad_norm": 1.0788978338241577, + "learning_rate": 2.3178733163007438e-05, + "loss": 0.9803, + "step": 8309 + }, + { + "epoch": 0.5709770509825478, + "grad_norm": 1.0863155126571655, + "learning_rate": 2.317275282722484e-05, + "loss": 0.9881, + "step": 8310 + }, + { + "epoch": 0.5710457606156383, + "grad_norm": 1.0650447607040405, + "learning_rate": 2.316677259656265e-05, + "loss": 0.9057, + "step": 8311 + }, + { + "epoch": 0.5711144702487289, + "grad_norm": 1.0672796964645386, + "learning_rate": 2.3160792471364916e-05, + "loss": 0.9411, + "step": 8312 + }, + { + "epoch": 0.5711831798818194, + "grad_norm": 1.046020269393921, + "learning_rate": 2.3154812451975673e-05, + "loss": 0.9359, + "step": 8313 + }, + { + "epoch": 0.57125188951491, + "grad_norm": 1.0400211811065674, + "learning_rate": 2.314883253873895e-05, + "loss": 0.943, + "step": 8314 + }, + { + "epoch": 0.5713205991480006, + "grad_norm": 1.0890060663223267, + "learning_rate": 2.3142852731998758e-05, + "loss": 0.9909, + "step": 8315 + }, + { + "epoch": 0.5713893087810911, + "grad_norm": 1.0167665481567383, + "learning_rate": 2.313687303209911e-05, + "loss": 0.9441, + "step": 8316 + }, + { + "epoch": 0.5714580184141816, + "grad_norm": 1.0768375396728516, + "learning_rate": 2.3130893439384023e-05, + "loss": 0.935, + "step": 8317 + }, + { + "epoch": 0.5715267280472722, + "grad_norm": 1.0862486362457275, + "learning_rate": 2.31249139541975e-05, + "loss": 0.9831, + "step": 8318 + }, + { + "epoch": 0.5715954376803628, + "grad_norm": 1.0744599103927612, + "learning_rate": 2.3118934576883538e-05, + "loss": 0.945, + "step": 8319 + }, + { + "epoch": 0.5716641473134534, + "grad_norm": 1.120355486869812, + "learning_rate": 2.3112955307786115e-05, + "loss": 0.9623, + "step": 8320 + }, + { + "epoch": 0.5717328569465439, + "grad_norm": 1.1068307161331177, + "learning_rate": 2.310697614724923e-05, + "loss": 1.0016, + "step": 8321 + }, + { + "epoch": 0.5718015665796344, + "grad_norm": 1.052675485610962, + "learning_rate": 2.3100997095616843e-05, + "loss": 0.961, + "step": 8322 + }, + { + "epoch": 0.571870276212725, + "grad_norm": 1.0197556018829346, + "learning_rate": 2.309501815323294e-05, + "loss": 0.9676, + "step": 8323 + }, + { + "epoch": 0.5719389858458156, + "grad_norm": 1.065348505973816, + "learning_rate": 2.308903932044149e-05, + "loss": 0.9273, + "step": 8324 + }, + { + "epoch": 0.5720076954789062, + "grad_norm": 1.0048727989196777, + "learning_rate": 2.308306059758643e-05, + "loss": 0.9012, + "step": 8325 + }, + { + "epoch": 0.5720764051119968, + "grad_norm": 1.043336272239685, + "learning_rate": 2.307708198501173e-05, + "loss": 0.9541, + "step": 8326 + }, + { + "epoch": 0.5721451147450872, + "grad_norm": 1.0459842681884766, + "learning_rate": 2.3071103483061326e-05, + "loss": 0.9186, + "step": 8327 + }, + { + "epoch": 0.5722138243781778, + "grad_norm": 1.0245212316513062, + "learning_rate": 2.3065125092079165e-05, + "loss": 0.9445, + "step": 8328 + }, + { + "epoch": 0.5722825340112684, + "grad_norm": 0.9917733669281006, + "learning_rate": 2.3059146812409188e-05, + "loss": 0.8958, + "step": 8329 + }, + { + "epoch": 0.572351243644359, + "grad_norm": 1.1275783777236938, + "learning_rate": 2.30531686443953e-05, + "loss": 0.9489, + "step": 8330 + }, + { + "epoch": 0.5724199532774495, + "grad_norm": 1.0372157096862793, + "learning_rate": 2.3047190588381433e-05, + "loss": 0.9513, + "step": 8331 + }, + { + "epoch": 0.57248866291054, + "grad_norm": 1.0490700006484985, + "learning_rate": 2.3041212644711506e-05, + "loss": 0.9124, + "step": 8332 + }, + { + "epoch": 0.5725573725436306, + "grad_norm": 1.048523187637329, + "learning_rate": 2.303523481372942e-05, + "loss": 0.9617, + "step": 8333 + }, + { + "epoch": 0.5726260821767212, + "grad_norm": 1.022863507270813, + "learning_rate": 2.3029257095779083e-05, + "loss": 0.9486, + "step": 8334 + }, + { + "epoch": 0.5726947918098118, + "grad_norm": 1.0403293371200562, + "learning_rate": 2.3023279491204373e-05, + "loss": 0.9425, + "step": 8335 + }, + { + "epoch": 0.5727635014429023, + "grad_norm": 1.0268193483352661, + "learning_rate": 2.301730200034919e-05, + "loss": 0.9575, + "step": 8336 + }, + { + "epoch": 0.5728322110759928, + "grad_norm": 1.096199631690979, + "learning_rate": 2.3011324623557416e-05, + "loss": 0.9838, + "step": 8337 + }, + { + "epoch": 0.5729009207090834, + "grad_norm": 1.0216106176376343, + "learning_rate": 2.300534736117292e-05, + "loss": 0.9592, + "step": 8338 + }, + { + "epoch": 0.572969630342174, + "grad_norm": 1.1091972589492798, + "learning_rate": 2.2999370213539588e-05, + "loss": 0.9507, + "step": 8339 + }, + { + "epoch": 0.5730383399752645, + "grad_norm": 1.1628891229629517, + "learning_rate": 2.2993393181001254e-05, + "loss": 0.9812, + "step": 8340 + }, + { + "epoch": 0.5731070496083551, + "grad_norm": 1.0708550214767456, + "learning_rate": 2.2987416263901788e-05, + "loss": 0.9513, + "step": 8341 + }, + { + "epoch": 0.5731757592414457, + "grad_norm": 1.0849964618682861, + "learning_rate": 2.2981439462585038e-05, + "loss": 0.9711, + "step": 8342 + }, + { + "epoch": 0.5732444688745362, + "grad_norm": 1.0757108926773071, + "learning_rate": 2.297546277739485e-05, + "loss": 0.9539, + "step": 8343 + }, + { + "epoch": 0.5733131785076268, + "grad_norm": 1.035889983177185, + "learning_rate": 2.2969486208675056e-05, + "loss": 0.9711, + "step": 8344 + }, + { + "epoch": 0.5733818881407173, + "grad_norm": 1.040533185005188, + "learning_rate": 2.296350975676948e-05, + "loss": 0.9694, + "step": 8345 + }, + { + "epoch": 0.5734505977738079, + "grad_norm": 1.0738838911056519, + "learning_rate": 2.295753342202194e-05, + "loss": 1.0137, + "step": 8346 + }, + { + "epoch": 0.5735193074068985, + "grad_norm": 1.1162959337234497, + "learning_rate": 2.2951557204776257e-05, + "loss": 0.9509, + "step": 8347 + }, + { + "epoch": 0.573588017039989, + "grad_norm": 1.0537275075912476, + "learning_rate": 2.294558110537624e-05, + "loss": 0.9101, + "step": 8348 + }, + { + "epoch": 0.5736567266730795, + "grad_norm": 1.0646467208862305, + "learning_rate": 2.29396051241657e-05, + "loss": 0.9491, + "step": 8349 + }, + { + "epoch": 0.5737254363061701, + "grad_norm": 1.0404292345046997, + "learning_rate": 2.2933629261488408e-05, + "loss": 0.9537, + "step": 8350 + }, + { + "epoch": 0.5737941459392607, + "grad_norm": 1.0992194414138794, + "learning_rate": 2.292765351768817e-05, + "loss": 0.9885, + "step": 8351 + }, + { + "epoch": 0.5738628555723513, + "grad_norm": 1.011911153793335, + "learning_rate": 2.2921677893108756e-05, + "loss": 0.9346, + "step": 8352 + }, + { + "epoch": 0.5739315652054418, + "grad_norm": 1.0499540567398071, + "learning_rate": 2.291570238809395e-05, + "loss": 0.9758, + "step": 8353 + }, + { + "epoch": 0.5740002748385323, + "grad_norm": 1.068367600440979, + "learning_rate": 2.290972700298752e-05, + "loss": 0.9394, + "step": 8354 + }, + { + "epoch": 0.5740689844716229, + "grad_norm": 1.093922734260559, + "learning_rate": 2.2903751738133216e-05, + "loss": 0.9249, + "step": 8355 + }, + { + "epoch": 0.5741376941047135, + "grad_norm": 1.0597398281097412, + "learning_rate": 2.2897776593874797e-05, + "loss": 0.9583, + "step": 8356 + }, + { + "epoch": 0.5742064037378041, + "grad_norm": 1.0818170309066772, + "learning_rate": 2.2891801570556005e-05, + "loss": 0.9961, + "step": 8357 + }, + { + "epoch": 0.5742751133708947, + "grad_norm": 1.0329452753067017, + "learning_rate": 2.288582666852059e-05, + "loss": 0.9576, + "step": 8358 + }, + { + "epoch": 0.5743438230039851, + "grad_norm": 1.0607715845108032, + "learning_rate": 2.287985188811228e-05, + "loss": 0.9722, + "step": 8359 + }, + { + "epoch": 0.5744125326370757, + "grad_norm": 1.1379250288009644, + "learning_rate": 2.2873877229674788e-05, + "loss": 0.9669, + "step": 8360 + }, + { + "epoch": 0.5744812422701663, + "grad_norm": 1.0440137386322021, + "learning_rate": 2.2867902693551845e-05, + "loss": 0.9636, + "step": 8361 + }, + { + "epoch": 0.5745499519032569, + "grad_norm": 1.0839521884918213, + "learning_rate": 2.2861928280087157e-05, + "loss": 0.9933, + "step": 8362 + }, + { + "epoch": 0.5746186615363474, + "grad_norm": 1.067549705505371, + "learning_rate": 2.2855953989624436e-05, + "loss": 0.9429, + "step": 8363 + }, + { + "epoch": 0.5746873711694379, + "grad_norm": 1.1607228517532349, + "learning_rate": 2.284997982250737e-05, + "loss": 1.0149, + "step": 8364 + }, + { + "epoch": 0.5747560808025285, + "grad_norm": 1.1034960746765137, + "learning_rate": 2.284400577907967e-05, + "loss": 0.9462, + "step": 8365 + }, + { + "epoch": 0.5748247904356191, + "grad_norm": 1.049457311630249, + "learning_rate": 2.283803185968499e-05, + "loss": 0.9787, + "step": 8366 + }, + { + "epoch": 0.5748935000687097, + "grad_norm": 1.0581917762756348, + "learning_rate": 2.283205806466702e-05, + "loss": 0.969, + "step": 8367 + }, + { + "epoch": 0.5749622097018002, + "grad_norm": 1.0988500118255615, + "learning_rate": 2.282608439436943e-05, + "loss": 0.9541, + "step": 8368 + }, + { + "epoch": 0.5750309193348907, + "grad_norm": 1.0279496908187866, + "learning_rate": 2.2820110849135872e-05, + "loss": 0.9555, + "step": 8369 + }, + { + "epoch": 0.5750996289679813, + "grad_norm": 1.071113109588623, + "learning_rate": 2.2814137429310023e-05, + "loss": 0.9649, + "step": 8370 + }, + { + "epoch": 0.5751683386010719, + "grad_norm": 1.0819363594055176, + "learning_rate": 2.2808164135235503e-05, + "loss": 0.9575, + "step": 8371 + }, + { + "epoch": 0.5752370482341624, + "grad_norm": 1.0521152019500732, + "learning_rate": 2.280219096725596e-05, + "loss": 0.9896, + "step": 8372 + }, + { + "epoch": 0.575305757867253, + "grad_norm": 1.1571683883666992, + "learning_rate": 2.2796217925715034e-05, + "loss": 0.9598, + "step": 8373 + }, + { + "epoch": 0.5753744675003436, + "grad_norm": 1.1033477783203125, + "learning_rate": 2.279024501095635e-05, + "loss": 0.9562, + "step": 8374 + }, + { + "epoch": 0.5754431771334341, + "grad_norm": 1.025890827178955, + "learning_rate": 2.278427222332353e-05, + "loss": 0.9636, + "step": 8375 + }, + { + "epoch": 0.5755118867665246, + "grad_norm": 1.1662184000015259, + "learning_rate": 2.2778299563160167e-05, + "loss": 0.9861, + "step": 8376 + }, + { + "epoch": 0.5755805963996152, + "grad_norm": 0.991990864276886, + "learning_rate": 2.2772327030809877e-05, + "loss": 0.9542, + "step": 8377 + }, + { + "epoch": 0.5756493060327058, + "grad_norm": 1.0755976438522339, + "learning_rate": 2.2766354626616262e-05, + "loss": 0.9281, + "step": 8378 + }, + { + "epoch": 0.5757180156657964, + "grad_norm": 1.119813323020935, + "learning_rate": 2.2760382350922894e-05, + "loss": 1.0155, + "step": 8379 + }, + { + "epoch": 0.5757867252988869, + "grad_norm": 1.119657278060913, + "learning_rate": 2.2754410204073372e-05, + "loss": 0.9594, + "step": 8380 + }, + { + "epoch": 0.5758554349319774, + "grad_norm": 1.1034772396087646, + "learning_rate": 2.274843818641126e-05, + "loss": 1.0001, + "step": 8381 + }, + { + "epoch": 0.575924144565068, + "grad_norm": 1.0127402544021606, + "learning_rate": 2.2742466298280123e-05, + "loss": 0.9242, + "step": 8382 + }, + { + "epoch": 0.5759928541981586, + "grad_norm": 1.0667884349822998, + "learning_rate": 2.2736494540023523e-05, + "loss": 0.9685, + "step": 8383 + }, + { + "epoch": 0.5760615638312492, + "grad_norm": 1.056803584098816, + "learning_rate": 2.273052291198501e-05, + "loss": 0.9798, + "step": 8384 + }, + { + "epoch": 0.5761302734643396, + "grad_norm": 1.099452018737793, + "learning_rate": 2.272455141450814e-05, + "loss": 0.937, + "step": 8385 + }, + { + "epoch": 0.5761989830974302, + "grad_norm": 1.0914260149002075, + "learning_rate": 2.2718580047936432e-05, + "loss": 0.9847, + "step": 8386 + }, + { + "epoch": 0.5762676927305208, + "grad_norm": 1.0913265943527222, + "learning_rate": 2.2712608812613423e-05, + "loss": 0.9711, + "step": 8387 + }, + { + "epoch": 0.5763364023636114, + "grad_norm": 1.078989028930664, + "learning_rate": 2.2706637708882634e-05, + "loss": 0.9477, + "step": 8388 + }, + { + "epoch": 0.576405111996702, + "grad_norm": 1.0095754861831665, + "learning_rate": 2.270066673708758e-05, + "loss": 0.9258, + "step": 8389 + }, + { + "epoch": 0.5764738216297925, + "grad_norm": 1.0434049367904663, + "learning_rate": 2.2694695897571772e-05, + "loss": 0.9577, + "step": 8390 + }, + { + "epoch": 0.576542531262883, + "grad_norm": 1.0373715162277222, + "learning_rate": 2.2688725190678694e-05, + "loss": 0.9397, + "step": 8391 + }, + { + "epoch": 0.5766112408959736, + "grad_norm": 1.0986424684524536, + "learning_rate": 2.2682754616751855e-05, + "loss": 0.9507, + "step": 8392 + }, + { + "epoch": 0.5766799505290642, + "grad_norm": 1.053407907485962, + "learning_rate": 2.2676784176134724e-05, + "loss": 0.946, + "step": 8393 + }, + { + "epoch": 0.5767486601621548, + "grad_norm": 1.0622971057891846, + "learning_rate": 2.2670813869170782e-05, + "loss": 0.9463, + "step": 8394 + }, + { + "epoch": 0.5768173697952453, + "grad_norm": 1.0725042819976807, + "learning_rate": 2.266484369620351e-05, + "loss": 0.9753, + "step": 8395 + }, + { + "epoch": 0.5768860794283358, + "grad_norm": 1.0902683734893799, + "learning_rate": 2.2658873657576346e-05, + "loss": 0.9686, + "step": 8396 + }, + { + "epoch": 0.5769547890614264, + "grad_norm": 1.0978312492370605, + "learning_rate": 2.265290375363275e-05, + "loss": 0.9437, + "step": 8397 + }, + { + "epoch": 0.577023498694517, + "grad_norm": 1.0408087968826294, + "learning_rate": 2.2646933984716173e-05, + "loss": 0.9618, + "step": 8398 + }, + { + "epoch": 0.5770922083276075, + "grad_norm": 1.014440894126892, + "learning_rate": 2.2640964351170045e-05, + "loss": 0.9506, + "step": 8399 + }, + { + "epoch": 0.5771609179606981, + "grad_norm": 1.0379761457443237, + "learning_rate": 2.2634994853337816e-05, + "loss": 0.9532, + "step": 8400 + }, + { + "epoch": 0.5772296275937886, + "grad_norm": 1.0882869958877563, + "learning_rate": 2.2629025491562877e-05, + "loss": 0.9957, + "step": 8401 + }, + { + "epoch": 0.5772983372268792, + "grad_norm": 1.0447174310684204, + "learning_rate": 2.2623056266188655e-05, + "loss": 0.9433, + "step": 8402 + }, + { + "epoch": 0.5773670468599698, + "grad_norm": 1.0023030042648315, + "learning_rate": 2.2617087177558565e-05, + "loss": 0.9147, + "step": 8403 + }, + { + "epoch": 0.5774357564930603, + "grad_norm": 1.1128251552581787, + "learning_rate": 2.261111822601599e-05, + "loss": 0.937, + "step": 8404 + }, + { + "epoch": 0.5775044661261509, + "grad_norm": 1.1056294441223145, + "learning_rate": 2.2605149411904335e-05, + "loss": 0.9601, + "step": 8405 + }, + { + "epoch": 0.5775731757592415, + "grad_norm": 1.0648149251937866, + "learning_rate": 2.2599180735566966e-05, + "loss": 0.9604, + "step": 8406 + }, + { + "epoch": 0.577641885392332, + "grad_norm": 1.0791388750076294, + "learning_rate": 2.2593212197347267e-05, + "loss": 0.9413, + "step": 8407 + }, + { + "epoch": 0.5777105950254225, + "grad_norm": 1.0881011486053467, + "learning_rate": 2.25872437975886e-05, + "loss": 0.9833, + "step": 8408 + }, + { + "epoch": 0.5777793046585131, + "grad_norm": 1.106813669204712, + "learning_rate": 2.2581275536634327e-05, + "loss": 0.9667, + "step": 8409 + }, + { + "epoch": 0.5778480142916037, + "grad_norm": 0.9978519082069397, + "learning_rate": 2.257530741482781e-05, + "loss": 0.9356, + "step": 8410 + }, + { + "epoch": 0.5779167239246943, + "grad_norm": 1.1261736154556274, + "learning_rate": 2.2569339432512364e-05, + "loss": 0.9751, + "step": 8411 + }, + { + "epoch": 0.5779854335577848, + "grad_norm": 1.0374795198440552, + "learning_rate": 2.256337159003134e-05, + "loss": 0.9614, + "step": 8412 + }, + { + "epoch": 0.5780541431908753, + "grad_norm": 1.0852032899856567, + "learning_rate": 2.2557403887728066e-05, + "loss": 0.9405, + "step": 8413 + }, + { + "epoch": 0.5781228528239659, + "grad_norm": 1.0523642301559448, + "learning_rate": 2.255143632594586e-05, + "loss": 0.9607, + "step": 8414 + }, + { + "epoch": 0.5781915624570565, + "grad_norm": 1.1139925718307495, + "learning_rate": 2.2545468905028028e-05, + "loss": 0.9564, + "step": 8415 + }, + { + "epoch": 0.5782602720901471, + "grad_norm": 1.1114760637283325, + "learning_rate": 2.2539501625317878e-05, + "loss": 0.9799, + "step": 8416 + }, + { + "epoch": 0.5783289817232375, + "grad_norm": 1.0397248268127441, + "learning_rate": 2.253353448715869e-05, + "loss": 0.9455, + "step": 8417 + }, + { + "epoch": 0.5783976913563281, + "grad_norm": 1.1236552000045776, + "learning_rate": 2.2527567490893758e-05, + "loss": 0.9858, + "step": 8418 + }, + { + "epoch": 0.5784664009894187, + "grad_norm": 1.1075940132141113, + "learning_rate": 2.2521600636866365e-05, + "loss": 0.9316, + "step": 8419 + }, + { + "epoch": 0.5785351106225093, + "grad_norm": 1.0421911478042603, + "learning_rate": 2.251563392541977e-05, + "loss": 0.9455, + "step": 8420 + }, + { + "epoch": 0.5786038202555999, + "grad_norm": 1.0496299266815186, + "learning_rate": 2.2509667356897253e-05, + "loss": 0.931, + "step": 8421 + }, + { + "epoch": 0.5786725298886904, + "grad_norm": 1.024817943572998, + "learning_rate": 2.2503700931642048e-05, + "loss": 0.9358, + "step": 8422 + }, + { + "epoch": 0.5787412395217809, + "grad_norm": 1.0652806758880615, + "learning_rate": 2.2497734649997403e-05, + "loss": 0.996, + "step": 8423 + }, + { + "epoch": 0.5788099491548715, + "grad_norm": 1.0680073499679565, + "learning_rate": 2.249176851230656e-05, + "loss": 0.9405, + "step": 8424 + }, + { + "epoch": 0.5788786587879621, + "grad_norm": 1.0671273469924927, + "learning_rate": 2.248580251891275e-05, + "loss": 0.9272, + "step": 8425 + }, + { + "epoch": 0.5789473684210527, + "grad_norm": 1.0433149337768555, + "learning_rate": 2.2479836670159186e-05, + "loss": 0.9586, + "step": 8426 + }, + { + "epoch": 0.5790160780541432, + "grad_norm": 1.100906252861023, + "learning_rate": 2.247387096638908e-05, + "loss": 0.9603, + "step": 8427 + }, + { + "epoch": 0.5790847876872337, + "grad_norm": 1.0841255187988281, + "learning_rate": 2.246790540794564e-05, + "loss": 0.9683, + "step": 8428 + }, + { + "epoch": 0.5791534973203243, + "grad_norm": 1.1661382913589478, + "learning_rate": 2.246193999517205e-05, + "loss": 0.966, + "step": 8429 + }, + { + "epoch": 0.5792222069534149, + "grad_norm": 1.0185332298278809, + "learning_rate": 2.2455974728411508e-05, + "loss": 0.9219, + "step": 8430 + }, + { + "epoch": 0.5792909165865054, + "grad_norm": 1.011412501335144, + "learning_rate": 2.2450009608007197e-05, + "loss": 0.9317, + "step": 8431 + }, + { + "epoch": 0.579359626219596, + "grad_norm": 1.024886131286621, + "learning_rate": 2.244404463430227e-05, + "loss": 0.9644, + "step": 8432 + }, + { + "epoch": 0.5794283358526865, + "grad_norm": 1.0336936712265015, + "learning_rate": 2.24380798076399e-05, + "loss": 0.948, + "step": 8433 + }, + { + "epoch": 0.5794970454857771, + "grad_norm": 1.0569086074829102, + "learning_rate": 2.2432115128363228e-05, + "loss": 0.9466, + "step": 8434 + }, + { + "epoch": 0.5795657551188677, + "grad_norm": 1.0360333919525146, + "learning_rate": 2.2426150596815412e-05, + "loss": 0.9748, + "step": 8435 + }, + { + "epoch": 0.5796344647519582, + "grad_norm": 1.1435158252716064, + "learning_rate": 2.2420186213339594e-05, + "loss": 0.9702, + "step": 8436 + }, + { + "epoch": 0.5797031743850488, + "grad_norm": 1.0430688858032227, + "learning_rate": 2.241422197827888e-05, + "loss": 0.9267, + "step": 8437 + }, + { + "epoch": 0.5797718840181394, + "grad_norm": 1.0625081062316895, + "learning_rate": 2.2408257891976403e-05, + "loss": 0.9646, + "step": 8438 + }, + { + "epoch": 0.5798405936512299, + "grad_norm": 1.0820462703704834, + "learning_rate": 2.2402293954775262e-05, + "loss": 0.9798, + "step": 8439 + }, + { + "epoch": 0.5799093032843204, + "grad_norm": 1.1133283376693726, + "learning_rate": 2.2396330167018566e-05, + "loss": 0.9346, + "step": 8440 + }, + { + "epoch": 0.579978012917411, + "grad_norm": 1.087127447128296, + "learning_rate": 2.2390366529049422e-05, + "loss": 0.9661, + "step": 8441 + }, + { + "epoch": 0.5800467225505016, + "grad_norm": 1.1171742677688599, + "learning_rate": 2.2384403041210887e-05, + "loss": 0.9573, + "step": 8442 + }, + { + "epoch": 0.5801154321835922, + "grad_norm": 1.122660756111145, + "learning_rate": 2.2378439703846052e-05, + "loss": 1.0233, + "step": 8443 + }, + { + "epoch": 0.5801841418166827, + "grad_norm": 1.097005009651184, + "learning_rate": 2.2372476517297984e-05, + "loss": 1.0152, + "step": 8444 + }, + { + "epoch": 0.5802528514497732, + "grad_norm": 1.0759735107421875, + "learning_rate": 2.2366513481909735e-05, + "loss": 0.9348, + "step": 8445 + }, + { + "epoch": 0.5803215610828638, + "grad_norm": 1.0397167205810547, + "learning_rate": 2.2360550598024374e-05, + "loss": 0.9417, + "step": 8446 + }, + { + "epoch": 0.5803902707159544, + "grad_norm": 1.1270006895065308, + "learning_rate": 2.2354587865984915e-05, + "loss": 1.0019, + "step": 8447 + }, + { + "epoch": 0.580458980349045, + "grad_norm": 1.0479638576507568, + "learning_rate": 2.2348625286134404e-05, + "loss": 0.9547, + "step": 8448 + }, + { + "epoch": 0.5805276899821354, + "grad_norm": 1.045440912246704, + "learning_rate": 2.2342662858815868e-05, + "loss": 0.9779, + "step": 8449 + }, + { + "epoch": 0.580596399615226, + "grad_norm": 1.0561045408248901, + "learning_rate": 2.2336700584372315e-05, + "loss": 0.9263, + "step": 8450 + }, + { + "epoch": 0.5806651092483166, + "grad_norm": 1.0168118476867676, + "learning_rate": 2.2330738463146757e-05, + "loss": 0.938, + "step": 8451 + }, + { + "epoch": 0.5807338188814072, + "grad_norm": 1.0615235567092896, + "learning_rate": 2.2324776495482186e-05, + "loss": 0.9486, + "step": 8452 + }, + { + "epoch": 0.5808025285144978, + "grad_norm": 0.9890648722648621, + "learning_rate": 2.2318814681721588e-05, + "loss": 0.9294, + "step": 8453 + }, + { + "epoch": 0.5808712381475883, + "grad_norm": 1.1622400283813477, + "learning_rate": 2.2312853022207946e-05, + "loss": 1.0143, + "step": 8454 + }, + { + "epoch": 0.5809399477806788, + "grad_norm": 1.030155062675476, + "learning_rate": 2.2306891517284234e-05, + "loss": 0.8894, + "step": 8455 + }, + { + "epoch": 0.5810086574137694, + "grad_norm": 1.1021932363510132, + "learning_rate": 2.230093016729342e-05, + "loss": 0.9804, + "step": 8456 + }, + { + "epoch": 0.58107736704686, + "grad_norm": 1.0067055225372314, + "learning_rate": 2.2294968972578434e-05, + "loss": 0.9353, + "step": 8457 + }, + { + "epoch": 0.5811460766799506, + "grad_norm": 1.0309773683547974, + "learning_rate": 2.2289007933482236e-05, + "loss": 0.9393, + "step": 8458 + }, + { + "epoch": 0.5812147863130411, + "grad_norm": 1.0439413785934448, + "learning_rate": 2.2283047050347762e-05, + "loss": 0.9668, + "step": 8459 + }, + { + "epoch": 0.5812834959461316, + "grad_norm": 1.0531890392303467, + "learning_rate": 2.2277086323517928e-05, + "loss": 0.9458, + "step": 8460 + }, + { + "epoch": 0.5813522055792222, + "grad_norm": 1.055365800857544, + "learning_rate": 2.2271125753335676e-05, + "loss": 0.9521, + "step": 8461 + }, + { + "epoch": 0.5814209152123128, + "grad_norm": 1.0945158004760742, + "learning_rate": 2.226516534014388e-05, + "loss": 1.0012, + "step": 8462 + }, + { + "epoch": 0.5814896248454033, + "grad_norm": 1.1345115900039673, + "learning_rate": 2.225920508428546e-05, + "loss": 0.9458, + "step": 8463 + }, + { + "epoch": 0.5815583344784939, + "grad_norm": 1.0509583950042725, + "learning_rate": 2.2253244986103295e-05, + "loss": 0.9221, + "step": 8464 + }, + { + "epoch": 0.5816270441115844, + "grad_norm": 1.1030950546264648, + "learning_rate": 2.224728504594027e-05, + "loss": 0.9449, + "step": 8465 + }, + { + "epoch": 0.581695753744675, + "grad_norm": 0.9997944235801697, + "learning_rate": 2.224132526413927e-05, + "loss": 0.9432, + "step": 8466 + }, + { + "epoch": 0.5817644633777656, + "grad_norm": 1.1590739488601685, + "learning_rate": 2.223536564104314e-05, + "loss": 0.9582, + "step": 8467 + }, + { + "epoch": 0.5818331730108561, + "grad_norm": 1.170854926109314, + "learning_rate": 2.222940617699473e-05, + "loss": 0.95, + "step": 8468 + }, + { + "epoch": 0.5819018826439467, + "grad_norm": 1.0899665355682373, + "learning_rate": 2.22234468723369e-05, + "loss": 0.9806, + "step": 8469 + }, + { + "epoch": 0.5819705922770373, + "grad_norm": 1.1382347345352173, + "learning_rate": 2.2217487727412474e-05, + "loss": 0.984, + "step": 8470 + }, + { + "epoch": 0.5820393019101278, + "grad_norm": 1.0138827562332153, + "learning_rate": 2.2211528742564293e-05, + "loss": 0.9462, + "step": 8471 + }, + { + "epoch": 0.5821080115432183, + "grad_norm": 1.0799050331115723, + "learning_rate": 2.2205569918135154e-05, + "loss": 0.9519, + "step": 8472 + }, + { + "epoch": 0.5821767211763089, + "grad_norm": 1.1326273679733276, + "learning_rate": 2.2199611254467874e-05, + "loss": 0.9807, + "step": 8473 + }, + { + "epoch": 0.5822454308093995, + "grad_norm": 1.0765947103500366, + "learning_rate": 2.2193652751905258e-05, + "loss": 0.9888, + "step": 8474 + }, + { + "epoch": 0.5823141404424901, + "grad_norm": 1.0438101291656494, + "learning_rate": 2.218769441079008e-05, + "loss": 0.9628, + "step": 8475 + }, + { + "epoch": 0.5823828500755805, + "grad_norm": 1.0837615728378296, + "learning_rate": 2.218173623146513e-05, + "loss": 0.9422, + "step": 8476 + }, + { + "epoch": 0.5824515597086711, + "grad_norm": 1.036830186843872, + "learning_rate": 2.2175778214273185e-05, + "loss": 0.9461, + "step": 8477 + }, + { + "epoch": 0.5825202693417617, + "grad_norm": 1.001599907875061, + "learning_rate": 2.216982035955699e-05, + "loss": 0.9587, + "step": 8478 + }, + { + "epoch": 0.5825889789748523, + "grad_norm": 1.0017989873886108, + "learning_rate": 2.2163862667659303e-05, + "loss": 0.926, + "step": 8479 + }, + { + "epoch": 0.5826576886079429, + "grad_norm": 1.0642625093460083, + "learning_rate": 2.2157905138922863e-05, + "loss": 0.9939, + "step": 8480 + }, + { + "epoch": 0.5827263982410333, + "grad_norm": 1.05978262424469, + "learning_rate": 2.2151947773690413e-05, + "loss": 0.942, + "step": 8481 + }, + { + "epoch": 0.5827951078741239, + "grad_norm": 1.0412431955337524, + "learning_rate": 2.214599057230468e-05, + "loss": 0.9643, + "step": 8482 + }, + { + "epoch": 0.5828638175072145, + "grad_norm": 1.105751872062683, + "learning_rate": 2.2140033535108357e-05, + "loss": 0.9565, + "step": 8483 + }, + { + "epoch": 0.5829325271403051, + "grad_norm": 1.0413792133331299, + "learning_rate": 2.213407666244416e-05, + "loss": 0.9415, + "step": 8484 + }, + { + "epoch": 0.5830012367733957, + "grad_norm": 1.0753458738327026, + "learning_rate": 2.212811995465479e-05, + "loss": 0.9551, + "step": 8485 + }, + { + "epoch": 0.5830699464064862, + "grad_norm": 1.0506718158721924, + "learning_rate": 2.2122163412082927e-05, + "loss": 0.9369, + "step": 8486 + }, + { + "epoch": 0.5831386560395767, + "grad_norm": 1.0045757293701172, + "learning_rate": 2.2116207035071253e-05, + "loss": 0.9447, + "step": 8487 + }, + { + "epoch": 0.5832073656726673, + "grad_norm": 1.1126084327697754, + "learning_rate": 2.2110250823962427e-05, + "loss": 0.9675, + "step": 8488 + }, + { + "epoch": 0.5832760753057579, + "grad_norm": 1.1001956462860107, + "learning_rate": 2.2104294779099104e-05, + "loss": 0.9279, + "step": 8489 + }, + { + "epoch": 0.5833447849388484, + "grad_norm": 1.0937919616699219, + "learning_rate": 2.2098338900823936e-05, + "loss": 0.9976, + "step": 8490 + }, + { + "epoch": 0.583413494571939, + "grad_norm": 1.0465415716171265, + "learning_rate": 2.2092383189479565e-05, + "loss": 0.9651, + "step": 8491 + }, + { + "epoch": 0.5834822042050295, + "grad_norm": 1.0820320844650269, + "learning_rate": 2.2086427645408626e-05, + "loss": 0.966, + "step": 8492 + }, + { + "epoch": 0.5835509138381201, + "grad_norm": 1.046000599861145, + "learning_rate": 2.2080472268953716e-05, + "loss": 0.9519, + "step": 8493 + }, + { + "epoch": 0.5836196234712107, + "grad_norm": 1.096548080444336, + "learning_rate": 2.2074517060457458e-05, + "loss": 0.9528, + "step": 8494 + }, + { + "epoch": 0.5836883331043012, + "grad_norm": 1.094602346420288, + "learning_rate": 2.2068562020262448e-05, + "loss": 0.9272, + "step": 8495 + }, + { + "epoch": 0.5837570427373918, + "grad_norm": 1.0147782564163208, + "learning_rate": 2.2062607148711284e-05, + "loss": 0.9666, + "step": 8496 + }, + { + "epoch": 0.5838257523704823, + "grad_norm": 1.0128471851348877, + "learning_rate": 2.2056652446146538e-05, + "loss": 0.9351, + "step": 8497 + }, + { + "epoch": 0.5838944620035729, + "grad_norm": 1.0936768054962158, + "learning_rate": 2.205069791291078e-05, + "loss": 0.959, + "step": 8498 + }, + { + "epoch": 0.5839631716366634, + "grad_norm": 1.107866883277893, + "learning_rate": 2.2044743549346576e-05, + "loss": 0.9704, + "step": 8499 + }, + { + "epoch": 0.584031881269754, + "grad_norm": 1.0585589408874512, + "learning_rate": 2.2038789355796467e-05, + "loss": 0.9479, + "step": 8500 + }, + { + "epoch": 0.5841005909028446, + "grad_norm": 1.0268148183822632, + "learning_rate": 2.2032835332603004e-05, + "loss": 0.9104, + "step": 8501 + }, + { + "epoch": 0.5841693005359352, + "grad_norm": 1.035850167274475, + "learning_rate": 2.2026881480108726e-05, + "loss": 0.9437, + "step": 8502 + }, + { + "epoch": 0.5842380101690257, + "grad_norm": 1.0255519151687622, + "learning_rate": 2.202092779865613e-05, + "loss": 0.9632, + "step": 8503 + }, + { + "epoch": 0.5843067198021162, + "grad_norm": 1.1010828018188477, + "learning_rate": 2.201497428858774e-05, + "loss": 0.9813, + "step": 8504 + }, + { + "epoch": 0.5843754294352068, + "grad_norm": 1.0553983449935913, + "learning_rate": 2.2009020950246064e-05, + "loss": 0.9485, + "step": 8505 + }, + { + "epoch": 0.5844441390682974, + "grad_norm": 1.0633714199066162, + "learning_rate": 2.2003067783973586e-05, + "loss": 0.9646, + "step": 8506 + }, + { + "epoch": 0.584512848701388, + "grad_norm": 1.0748955011367798, + "learning_rate": 2.1997114790112806e-05, + "loss": 0.9141, + "step": 8507 + }, + { + "epoch": 0.5845815583344784, + "grad_norm": 1.0068353414535522, + "learning_rate": 2.1991161969006163e-05, + "loss": 0.9461, + "step": 8508 + }, + { + "epoch": 0.584650267967569, + "grad_norm": 1.0884753465652466, + "learning_rate": 2.198520932099614e-05, + "loss": 0.9662, + "step": 8509 + }, + { + "epoch": 0.5847189776006596, + "grad_norm": 1.0245354175567627, + "learning_rate": 2.1979256846425195e-05, + "loss": 0.9347, + "step": 8510 + }, + { + "epoch": 0.5847876872337502, + "grad_norm": 1.1236556768417358, + "learning_rate": 2.1973304545635758e-05, + "loss": 0.9632, + "step": 8511 + }, + { + "epoch": 0.5848563968668408, + "grad_norm": 1.041322946548462, + "learning_rate": 2.196735241897027e-05, + "loss": 0.9419, + "step": 8512 + }, + { + "epoch": 0.5849251064999312, + "grad_norm": 1.031422734260559, + "learning_rate": 2.196140046677114e-05, + "loss": 0.9839, + "step": 8513 + }, + { + "epoch": 0.5849938161330218, + "grad_norm": 1.0003737211227417, + "learning_rate": 2.1955448689380788e-05, + "loss": 0.9262, + "step": 8514 + }, + { + "epoch": 0.5850625257661124, + "grad_norm": 1.1246957778930664, + "learning_rate": 2.1949497087141617e-05, + "loss": 0.9903, + "step": 8515 + }, + { + "epoch": 0.585131235399203, + "grad_norm": 1.0781903266906738, + "learning_rate": 2.1943545660396018e-05, + "loss": 0.9986, + "step": 8516 + }, + { + "epoch": 0.5851999450322936, + "grad_norm": 1.0954687595367432, + "learning_rate": 2.1937594409486384e-05, + "loss": 0.9203, + "step": 8517 + }, + { + "epoch": 0.5852686546653841, + "grad_norm": 1.0100404024124146, + "learning_rate": 2.1931643334755067e-05, + "loss": 0.9522, + "step": 8518 + }, + { + "epoch": 0.5853373642984746, + "grad_norm": 1.053091287612915, + "learning_rate": 2.1925692436544437e-05, + "loss": 0.9617, + "step": 8519 + }, + { + "epoch": 0.5854060739315652, + "grad_norm": 1.0681917667388916, + "learning_rate": 2.1919741715196848e-05, + "loss": 0.9672, + "step": 8520 + }, + { + "epoch": 0.5854747835646558, + "grad_norm": 1.1389161348342896, + "learning_rate": 2.1913791171054643e-05, + "loss": 0.9768, + "step": 8521 + }, + { + "epoch": 0.5855434931977463, + "grad_norm": 1.0725005865097046, + "learning_rate": 2.1907840804460153e-05, + "loss": 0.9648, + "step": 8522 + }, + { + "epoch": 0.5856122028308369, + "grad_norm": 1.035865068435669, + "learning_rate": 2.1901890615755694e-05, + "loss": 0.9381, + "step": 8523 + }, + { + "epoch": 0.5856809124639274, + "grad_norm": 1.0525115728378296, + "learning_rate": 2.1895940605283575e-05, + "loss": 0.9273, + "step": 8524 + }, + { + "epoch": 0.585749622097018, + "grad_norm": 1.1024340391159058, + "learning_rate": 2.1889990773386097e-05, + "loss": 0.968, + "step": 8525 + }, + { + "epoch": 0.5858183317301086, + "grad_norm": 1.0370898246765137, + "learning_rate": 2.188404112040556e-05, + "loss": 0.9752, + "step": 8526 + }, + { + "epoch": 0.5858870413631991, + "grad_norm": 1.058131456375122, + "learning_rate": 2.187809164668424e-05, + "loss": 0.9542, + "step": 8527 + }, + { + "epoch": 0.5859557509962897, + "grad_norm": 1.0513874292373657, + "learning_rate": 2.1872142352564398e-05, + "loss": 0.9798, + "step": 8528 + }, + { + "epoch": 0.5860244606293802, + "grad_norm": 1.0171164274215698, + "learning_rate": 2.18661932383883e-05, + "loss": 0.9574, + "step": 8529 + }, + { + "epoch": 0.5860931702624708, + "grad_norm": 1.0213735103607178, + "learning_rate": 2.1860244304498192e-05, + "loss": 0.9316, + "step": 8530 + }, + { + "epoch": 0.5861618798955613, + "grad_norm": 1.1056795120239258, + "learning_rate": 2.1854295551236314e-05, + "loss": 0.9877, + "step": 8531 + }, + { + "epoch": 0.5862305895286519, + "grad_norm": 1.0545852184295654, + "learning_rate": 2.1848346978944906e-05, + "loss": 0.9851, + "step": 8532 + }, + { + "epoch": 0.5862992991617425, + "grad_norm": 1.100354552268982, + "learning_rate": 2.1842398587966166e-05, + "loss": 0.9358, + "step": 8533 + }, + { + "epoch": 0.5863680087948331, + "grad_norm": 1.0196435451507568, + "learning_rate": 2.1836450378642318e-05, + "loss": 0.9538, + "step": 8534 + }, + { + "epoch": 0.5864367184279236, + "grad_norm": 1.0472290515899658, + "learning_rate": 2.183050235131554e-05, + "loss": 0.957, + "step": 8535 + }, + { + "epoch": 0.5865054280610141, + "grad_norm": 1.0605465173721313, + "learning_rate": 2.182455450632803e-05, + "loss": 0.9283, + "step": 8536 + }, + { + "epoch": 0.5865741376941047, + "grad_norm": 1.0707612037658691, + "learning_rate": 2.1818606844021967e-05, + "loss": 0.9461, + "step": 8537 + }, + { + "epoch": 0.5866428473271953, + "grad_norm": 1.101098895072937, + "learning_rate": 2.1812659364739522e-05, + "loss": 0.9629, + "step": 8538 + }, + { + "epoch": 0.5867115569602859, + "grad_norm": 1.147814393043518, + "learning_rate": 2.1806712068822828e-05, + "loss": 0.9861, + "step": 8539 + }, + { + "epoch": 0.5867802665933763, + "grad_norm": 1.0925902128219604, + "learning_rate": 2.1800764956614038e-05, + "loss": 0.9598, + "step": 8540 + }, + { + "epoch": 0.5868489762264669, + "grad_norm": 1.1295390129089355, + "learning_rate": 2.1794818028455295e-05, + "loss": 0.9555, + "step": 8541 + }, + { + "epoch": 0.5869176858595575, + "grad_norm": 1.0939204692840576, + "learning_rate": 2.1788871284688713e-05, + "loss": 0.9248, + "step": 8542 + }, + { + "epoch": 0.5869863954926481, + "grad_norm": 1.063940167427063, + "learning_rate": 2.1782924725656423e-05, + "loss": 0.9794, + "step": 8543 + }, + { + "epoch": 0.5870551051257387, + "grad_norm": 1.1348495483398438, + "learning_rate": 2.17769783517005e-05, + "loss": 0.9564, + "step": 8544 + }, + { + "epoch": 0.5871238147588292, + "grad_norm": 1.0466439723968506, + "learning_rate": 2.1771032163163048e-05, + "loss": 0.9141, + "step": 8545 + }, + { + "epoch": 0.5871925243919197, + "grad_norm": 1.0337218046188354, + "learning_rate": 2.1765086160386144e-05, + "loss": 0.9217, + "step": 8546 + }, + { + "epoch": 0.5872612340250103, + "grad_norm": 1.0501742362976074, + "learning_rate": 2.175914034371186e-05, + "loss": 0.9705, + "step": 8547 + }, + { + "epoch": 0.5873299436581009, + "grad_norm": 1.028437614440918, + "learning_rate": 2.1753194713482273e-05, + "loss": 0.948, + "step": 8548 + }, + { + "epoch": 0.5873986532911915, + "grad_norm": 1.0832816362380981, + "learning_rate": 2.1747249270039397e-05, + "loss": 0.9564, + "step": 8549 + }, + { + "epoch": 0.587467362924282, + "grad_norm": 1.016579031944275, + "learning_rate": 2.174130401372529e-05, + "loss": 0.9189, + "step": 8550 + }, + { + "epoch": 0.5875360725573725, + "grad_norm": 1.071571946144104, + "learning_rate": 2.1735358944881974e-05, + "loss": 0.9444, + "step": 8551 + }, + { + "epoch": 0.5876047821904631, + "grad_norm": 1.0632028579711914, + "learning_rate": 2.1729414063851465e-05, + "loss": 0.9531, + "step": 8552 + }, + { + "epoch": 0.5876734918235537, + "grad_norm": 1.0279667377471924, + "learning_rate": 2.1723469370975787e-05, + "loss": 0.9712, + "step": 8553 + }, + { + "epoch": 0.5877422014566442, + "grad_norm": 1.0034009218215942, + "learning_rate": 2.1717524866596907e-05, + "loss": 0.922, + "step": 8554 + }, + { + "epoch": 0.5878109110897348, + "grad_norm": 1.0205683708190918, + "learning_rate": 2.171158055105682e-05, + "loss": 0.9578, + "step": 8555 + }, + { + "epoch": 0.5878796207228253, + "grad_norm": 1.0349361896514893, + "learning_rate": 2.17056364246975e-05, + "loss": 0.9532, + "step": 8556 + }, + { + "epoch": 0.5879483303559159, + "grad_norm": 1.0845686197280884, + "learning_rate": 2.1699692487860908e-05, + "loss": 0.9461, + "step": 8557 + }, + { + "epoch": 0.5880170399890065, + "grad_norm": 1.0411958694458008, + "learning_rate": 2.1693748740888996e-05, + "loss": 0.9468, + "step": 8558 + }, + { + "epoch": 0.588085749622097, + "grad_norm": 1.097540259361267, + "learning_rate": 2.1687805184123705e-05, + "loss": 0.9405, + "step": 8559 + }, + { + "epoch": 0.5881544592551876, + "grad_norm": 1.2767246961593628, + "learning_rate": 2.1681861817906954e-05, + "loss": 0.9931, + "step": 8560 + }, + { + "epoch": 0.5882231688882782, + "grad_norm": 1.0969653129577637, + "learning_rate": 2.1675918642580674e-05, + "loss": 0.9608, + "step": 8561 + }, + { + "epoch": 0.5882918785213687, + "grad_norm": 1.080195426940918, + "learning_rate": 2.1669975658486763e-05, + "loss": 0.9712, + "step": 8562 + }, + { + "epoch": 0.5883605881544592, + "grad_norm": 1.0246658325195312, + "learning_rate": 2.1664032865967134e-05, + "loss": 0.952, + "step": 8563 + }, + { + "epoch": 0.5884292977875498, + "grad_norm": 1.075478434562683, + "learning_rate": 2.165809026536365e-05, + "loss": 0.9694, + "step": 8564 + }, + { + "epoch": 0.5884980074206404, + "grad_norm": 1.095471978187561, + "learning_rate": 2.16521478570182e-05, + "loss": 0.9648, + "step": 8565 + }, + { + "epoch": 0.588566717053731, + "grad_norm": 1.1008292436599731, + "learning_rate": 2.164620564127264e-05, + "loss": 0.9506, + "step": 8566 + }, + { + "epoch": 0.5886354266868215, + "grad_norm": 1.0269302129745483, + "learning_rate": 2.1640263618468827e-05, + "loss": 0.968, + "step": 8567 + }, + { + "epoch": 0.588704136319912, + "grad_norm": 1.0772895812988281, + "learning_rate": 2.1634321788948596e-05, + "loss": 0.9529, + "step": 8568 + }, + { + "epoch": 0.5887728459530026, + "grad_norm": 1.0319945812225342, + "learning_rate": 2.162838015305378e-05, + "loss": 0.9438, + "step": 8569 + }, + { + "epoch": 0.5888415555860932, + "grad_norm": 1.01304292678833, + "learning_rate": 2.1622438711126205e-05, + "loss": 0.9021, + "step": 8570 + }, + { + "epoch": 0.5889102652191838, + "grad_norm": 1.0363833904266357, + "learning_rate": 2.1616497463507664e-05, + "loss": 0.9773, + "step": 8571 + }, + { + "epoch": 0.5889789748522742, + "grad_norm": 1.0589823722839355, + "learning_rate": 2.1610556410539962e-05, + "loss": 0.9351, + "step": 8572 + }, + { + "epoch": 0.5890476844853648, + "grad_norm": 1.05962336063385, + "learning_rate": 2.160461555256489e-05, + "loss": 0.942, + "step": 8573 + }, + { + "epoch": 0.5891163941184554, + "grad_norm": 1.0837469100952148, + "learning_rate": 2.1598674889924208e-05, + "loss": 0.9833, + "step": 8574 + }, + { + "epoch": 0.589185103751546, + "grad_norm": 1.0395944118499756, + "learning_rate": 2.159273442295969e-05, + "loss": 0.9096, + "step": 8575 + }, + { + "epoch": 0.5892538133846366, + "grad_norm": 1.0631132125854492, + "learning_rate": 2.1586794152013074e-05, + "loss": 0.9413, + "step": 8576 + }, + { + "epoch": 0.5893225230177271, + "grad_norm": 0.9581295251846313, + "learning_rate": 2.1580854077426112e-05, + "loss": 0.8987, + "step": 8577 + }, + { + "epoch": 0.5893912326508176, + "grad_norm": 1.0244121551513672, + "learning_rate": 2.1574914199540543e-05, + "loss": 0.9137, + "step": 8578 + }, + { + "epoch": 0.5894599422839082, + "grad_norm": 1.0887916088104248, + "learning_rate": 2.156897451869806e-05, + "loss": 0.944, + "step": 8579 + }, + { + "epoch": 0.5895286519169988, + "grad_norm": 0.994799792766571, + "learning_rate": 2.156303503524038e-05, + "loss": 0.9347, + "step": 8580 + }, + { + "epoch": 0.5895973615500893, + "grad_norm": 1.1391170024871826, + "learning_rate": 2.1557095749509206e-05, + "loss": 0.9416, + "step": 8581 + }, + { + "epoch": 0.5896660711831799, + "grad_norm": 1.02945876121521, + "learning_rate": 2.155115666184621e-05, + "loss": 0.9667, + "step": 8582 + }, + { + "epoch": 0.5897347808162704, + "grad_norm": 1.1608954668045044, + "learning_rate": 2.1545217772593064e-05, + "loss": 0.9597, + "step": 8583 + }, + { + "epoch": 0.589803490449361, + "grad_norm": 1.1002767086029053, + "learning_rate": 2.1539279082091446e-05, + "loss": 0.9187, + "step": 8584 + }, + { + "epoch": 0.5898722000824516, + "grad_norm": 1.0786247253417969, + "learning_rate": 2.153334059068298e-05, + "loss": 0.9994, + "step": 8585 + }, + { + "epoch": 0.5899409097155421, + "grad_norm": 1.0707590579986572, + "learning_rate": 2.1527402298709322e-05, + "loss": 0.9549, + "step": 8586 + }, + { + "epoch": 0.5900096193486327, + "grad_norm": 1.1057919263839722, + "learning_rate": 2.1521464206512086e-05, + "loss": 0.9649, + "step": 8587 + }, + { + "epoch": 0.5900783289817232, + "grad_norm": 1.03874933719635, + "learning_rate": 2.1515526314432896e-05, + "loss": 0.9547, + "step": 8588 + }, + { + "epoch": 0.5901470386148138, + "grad_norm": 1.130325198173523, + "learning_rate": 2.1509588622813364e-05, + "loss": 0.9423, + "step": 8589 + }, + { + "epoch": 0.5902157482479043, + "grad_norm": 1.0928221940994263, + "learning_rate": 2.1503651131995057e-05, + "loss": 0.9682, + "step": 8590 + }, + { + "epoch": 0.5902844578809949, + "grad_norm": 1.0224369764328003, + "learning_rate": 2.149771384231957e-05, + "loss": 0.9349, + "step": 8591 + }, + { + "epoch": 0.5903531675140855, + "grad_norm": 1.0394842624664307, + "learning_rate": 2.1491776754128477e-05, + "loss": 0.933, + "step": 8592 + }, + { + "epoch": 0.5904218771471761, + "grad_norm": 1.028569221496582, + "learning_rate": 2.148583986776332e-05, + "loss": 0.9437, + "step": 8593 + }, + { + "epoch": 0.5904905867802666, + "grad_norm": 1.0793079137802124, + "learning_rate": 2.147990318356566e-05, + "loss": 0.9581, + "step": 8594 + }, + { + "epoch": 0.5905592964133571, + "grad_norm": 1.028691053390503, + "learning_rate": 2.1473966701877025e-05, + "loss": 0.9494, + "step": 8595 + }, + { + "epoch": 0.5906280060464477, + "grad_norm": 1.0644124746322632, + "learning_rate": 2.1468030423038927e-05, + "loss": 0.9517, + "step": 8596 + }, + { + "epoch": 0.5906967156795383, + "grad_norm": 1.0215489864349365, + "learning_rate": 2.1462094347392887e-05, + "loss": 0.9108, + "step": 8597 + }, + { + "epoch": 0.5907654253126289, + "grad_norm": 1.0419279336929321, + "learning_rate": 2.14561584752804e-05, + "loss": 0.9508, + "step": 8598 + }, + { + "epoch": 0.5908341349457193, + "grad_norm": 1.0336662530899048, + "learning_rate": 2.145022280704297e-05, + "loss": 0.9507, + "step": 8599 + }, + { + "epoch": 0.5909028445788099, + "grad_norm": 0.9965399503707886, + "learning_rate": 2.144428734302204e-05, + "loss": 0.9084, + "step": 8600 + }, + { + "epoch": 0.5909715542119005, + "grad_norm": 1.0320615768432617, + "learning_rate": 2.1438352083559095e-05, + "loss": 0.9508, + "step": 8601 + }, + { + "epoch": 0.5910402638449911, + "grad_norm": 0.9905860424041748, + "learning_rate": 2.1432417028995583e-05, + "loss": 0.9329, + "step": 8602 + }, + { + "epoch": 0.5911089734780817, + "grad_norm": 1.1317156553268433, + "learning_rate": 2.142648217967295e-05, + "loss": 0.9758, + "step": 8603 + }, + { + "epoch": 0.5911776831111721, + "grad_norm": 1.0861601829528809, + "learning_rate": 2.142054753593262e-05, + "loss": 0.9555, + "step": 8604 + }, + { + "epoch": 0.5912463927442627, + "grad_norm": 1.0616008043289185, + "learning_rate": 2.1414613098116e-05, + "loss": 0.9406, + "step": 8605 + }, + { + "epoch": 0.5913151023773533, + "grad_norm": 1.0346499681472778, + "learning_rate": 2.1408678866564508e-05, + "loss": 0.9707, + "step": 8606 + }, + { + "epoch": 0.5913838120104439, + "grad_norm": 1.0594913959503174, + "learning_rate": 2.1402744841619527e-05, + "loss": 0.9969, + "step": 8607 + }, + { + "epoch": 0.5914525216435345, + "grad_norm": 1.099685788154602, + "learning_rate": 2.1396811023622436e-05, + "loss": 0.9529, + "step": 8608 + }, + { + "epoch": 0.591521231276625, + "grad_norm": 1.0637857913970947, + "learning_rate": 2.139087741291463e-05, + "loss": 0.9327, + "step": 8609 + }, + { + "epoch": 0.5915899409097155, + "grad_norm": 1.074530839920044, + "learning_rate": 2.1384944009837427e-05, + "loss": 0.9368, + "step": 8610 + }, + { + "epoch": 0.5916586505428061, + "grad_norm": 1.0846824645996094, + "learning_rate": 2.1379010814732193e-05, + "loss": 1.0063, + "step": 8611 + }, + { + "epoch": 0.5917273601758967, + "grad_norm": 0.9979630708694458, + "learning_rate": 2.1373077827940263e-05, + "loss": 0.9279, + "step": 8612 + }, + { + "epoch": 0.5917960698089872, + "grad_norm": 1.0485748052597046, + "learning_rate": 2.136714504980295e-05, + "loss": 0.9515, + "step": 8613 + }, + { + "epoch": 0.5918647794420778, + "grad_norm": 1.0415478944778442, + "learning_rate": 2.136121248066158e-05, + "loss": 0.976, + "step": 8614 + }, + { + "epoch": 0.5919334890751683, + "grad_norm": 1.0514096021652222, + "learning_rate": 2.1355280120857423e-05, + "loss": 0.9309, + "step": 8615 + }, + { + "epoch": 0.5920021987082589, + "grad_norm": 1.0980095863342285, + "learning_rate": 2.1349347970731783e-05, + "loss": 0.9341, + "step": 8616 + }, + { + "epoch": 0.5920709083413495, + "grad_norm": 1.0384389162063599, + "learning_rate": 2.1343416030625933e-05, + "loss": 0.9438, + "step": 8617 + }, + { + "epoch": 0.59213961797444, + "grad_norm": 1.0700944662094116, + "learning_rate": 2.1337484300881122e-05, + "loss": 0.9832, + "step": 8618 + }, + { + "epoch": 0.5922083276075306, + "grad_norm": 1.0609846115112305, + "learning_rate": 2.1331552781838617e-05, + "loss": 0.9831, + "step": 8619 + }, + { + "epoch": 0.5922770372406211, + "grad_norm": 1.0838510990142822, + "learning_rate": 2.1325621473839635e-05, + "loss": 0.9488, + "step": 8620 + }, + { + "epoch": 0.5923457468737117, + "grad_norm": 1.065243124961853, + "learning_rate": 2.1319690377225406e-05, + "loss": 0.9279, + "step": 8621 + }, + { + "epoch": 0.5924144565068022, + "grad_norm": 1.0658084154129028, + "learning_rate": 2.131375949233715e-05, + "loss": 0.8967, + "step": 8622 + }, + { + "epoch": 0.5924831661398928, + "grad_norm": 0.9949260354042053, + "learning_rate": 2.1307828819516058e-05, + "loss": 0.9132, + "step": 8623 + }, + { + "epoch": 0.5925518757729834, + "grad_norm": 1.0524113178253174, + "learning_rate": 2.1301898359103333e-05, + "loss": 0.9025, + "step": 8624 + }, + { + "epoch": 0.592620585406074, + "grad_norm": 1.0569570064544678, + "learning_rate": 2.1295968111440134e-05, + "loss": 0.9689, + "step": 8625 + }, + { + "epoch": 0.5926892950391645, + "grad_norm": 1.051104187965393, + "learning_rate": 2.129003807686763e-05, + "loss": 0.9332, + "step": 8626 + }, + { + "epoch": 0.592758004672255, + "grad_norm": 0.9804357886314392, + "learning_rate": 2.1284108255726972e-05, + "loss": 0.9205, + "step": 8627 + }, + { + "epoch": 0.5928267143053456, + "grad_norm": 1.0934354066848755, + "learning_rate": 2.1278178648359305e-05, + "loss": 0.986, + "step": 8628 + }, + { + "epoch": 0.5928954239384362, + "grad_norm": 1.172425389289856, + "learning_rate": 2.1272249255105754e-05, + "loss": 0.9342, + "step": 8629 + }, + { + "epoch": 0.5929641335715268, + "grad_norm": 1.0433841943740845, + "learning_rate": 2.1266320076307428e-05, + "loss": 0.9295, + "step": 8630 + }, + { + "epoch": 0.5930328432046172, + "grad_norm": 0.9865366816520691, + "learning_rate": 2.1260391112305428e-05, + "loss": 0.9006, + "step": 8631 + }, + { + "epoch": 0.5931015528377078, + "grad_norm": 1.165756344795227, + "learning_rate": 2.1254462363440845e-05, + "loss": 0.9141, + "step": 8632 + }, + { + "epoch": 0.5931702624707984, + "grad_norm": 1.0892424583435059, + "learning_rate": 2.124853383005476e-05, + "loss": 0.9284, + "step": 8633 + }, + { + "epoch": 0.593238972103889, + "grad_norm": 1.0994964838027954, + "learning_rate": 2.1242605512488248e-05, + "loss": 0.9512, + "step": 8634 + }, + { + "epoch": 0.5933076817369796, + "grad_norm": 1.07211172580719, + "learning_rate": 2.1236677411082338e-05, + "loss": 0.9488, + "step": 8635 + }, + { + "epoch": 0.59337639137007, + "grad_norm": 1.0607553720474243, + "learning_rate": 2.1230749526178077e-05, + "loss": 0.9142, + "step": 8636 + }, + { + "epoch": 0.5934451010031606, + "grad_norm": 1.0675928592681885, + "learning_rate": 2.1224821858116507e-05, + "loss": 0.9697, + "step": 8637 + }, + { + "epoch": 0.5935138106362512, + "grad_norm": 1.093902349472046, + "learning_rate": 2.1218894407238626e-05, + "loss": 0.994, + "step": 8638 + }, + { + "epoch": 0.5935825202693418, + "grad_norm": 1.0545483827590942, + "learning_rate": 2.1212967173885454e-05, + "loss": 0.9705, + "step": 8639 + }, + { + "epoch": 0.5936512299024324, + "grad_norm": 1.0644662380218506, + "learning_rate": 2.120704015839797e-05, + "loss": 0.9767, + "step": 8640 + }, + { + "epoch": 0.5937199395355229, + "grad_norm": 1.0906296968460083, + "learning_rate": 2.1201113361117154e-05, + "loss": 0.9807, + "step": 8641 + }, + { + "epoch": 0.5937886491686134, + "grad_norm": 1.0428147315979004, + "learning_rate": 2.1195186782383962e-05, + "loss": 0.9525, + "step": 8642 + }, + { + "epoch": 0.593857358801704, + "grad_norm": 1.0613865852355957, + "learning_rate": 2.1189260422539355e-05, + "loss": 0.9638, + "step": 8643 + }, + { + "epoch": 0.5939260684347946, + "grad_norm": 1.0446687936782837, + "learning_rate": 2.1183334281924275e-05, + "loss": 0.9156, + "step": 8644 + }, + { + "epoch": 0.5939947780678851, + "grad_norm": 1.1123692989349365, + "learning_rate": 2.1177408360879654e-05, + "loss": 0.9641, + "step": 8645 + }, + { + "epoch": 0.5940634877009757, + "grad_norm": 0.9995151162147522, + "learning_rate": 2.1171482659746393e-05, + "loss": 0.9146, + "step": 8646 + }, + { + "epoch": 0.5941321973340662, + "grad_norm": 1.070706844329834, + "learning_rate": 2.1165557178865398e-05, + "loss": 0.9489, + "step": 8647 + }, + { + "epoch": 0.5942009069671568, + "grad_norm": 1.0651544332504272, + "learning_rate": 2.115963191857756e-05, + "loss": 0.8959, + "step": 8648 + }, + { + "epoch": 0.5942696166002474, + "grad_norm": 1.0384211540222168, + "learning_rate": 2.1153706879223757e-05, + "loss": 0.9558, + "step": 8649 + }, + { + "epoch": 0.5943383262333379, + "grad_norm": 1.0756454467773438, + "learning_rate": 2.114778206114486e-05, + "loss": 0.9616, + "step": 8650 + }, + { + "epoch": 0.5944070358664285, + "grad_norm": 1.1007044315338135, + "learning_rate": 2.1141857464681703e-05, + "loss": 0.9448, + "step": 8651 + }, + { + "epoch": 0.594475745499519, + "grad_norm": 1.0633275508880615, + "learning_rate": 2.113593309017514e-05, + "loss": 0.9669, + "step": 8652 + }, + { + "epoch": 0.5945444551326096, + "grad_norm": 1.067934513092041, + "learning_rate": 2.1130008937965984e-05, + "loss": 0.9693, + "step": 8653 + }, + { + "epoch": 0.5946131647657001, + "grad_norm": 1.0844757556915283, + "learning_rate": 2.1124085008395054e-05, + "loss": 0.9254, + "step": 8654 + }, + { + "epoch": 0.5946818743987907, + "grad_norm": 1.0446580648422241, + "learning_rate": 2.111816130180316e-05, + "loss": 0.9873, + "step": 8655 + }, + { + "epoch": 0.5947505840318813, + "grad_norm": 1.0759583711624146, + "learning_rate": 2.1112237818531073e-05, + "loss": 0.9352, + "step": 8656 + }, + { + "epoch": 0.5948192936649719, + "grad_norm": 1.0747852325439453, + "learning_rate": 2.110631455891957e-05, + "loss": 0.9486, + "step": 8657 + }, + { + "epoch": 0.5948880032980624, + "grad_norm": 1.1056722402572632, + "learning_rate": 2.1100391523309414e-05, + "loss": 0.955, + "step": 8658 + }, + { + "epoch": 0.5949567129311529, + "grad_norm": 1.0622384548187256, + "learning_rate": 2.109446871204136e-05, + "loss": 0.9161, + "step": 8659 + }, + { + "epoch": 0.5950254225642435, + "grad_norm": 1.116897702217102, + "learning_rate": 2.108854612545615e-05, + "loss": 0.9726, + "step": 8660 + }, + { + "epoch": 0.5950941321973341, + "grad_norm": 1.1349583864212036, + "learning_rate": 2.1082623763894485e-05, + "loss": 0.9741, + "step": 8661 + }, + { + "epoch": 0.5951628418304247, + "grad_norm": 1.064561128616333, + "learning_rate": 2.1076701627697086e-05, + "loss": 0.9632, + "step": 8662 + }, + { + "epoch": 0.5952315514635151, + "grad_norm": 1.0724797248840332, + "learning_rate": 2.1070779717204657e-05, + "loss": 0.9596, + "step": 8663 + }, + { + "epoch": 0.5953002610966057, + "grad_norm": 1.0613282918930054, + "learning_rate": 2.1064858032757868e-05, + "loss": 0.9737, + "step": 8664 + }, + { + "epoch": 0.5953689707296963, + "grad_norm": 1.087071418762207, + "learning_rate": 2.1058936574697405e-05, + "loss": 0.931, + "step": 8665 + }, + { + "epoch": 0.5954376803627869, + "grad_norm": 1.1153513193130493, + "learning_rate": 2.1053015343363912e-05, + "loss": 0.9611, + "step": 8666 + }, + { + "epoch": 0.5955063899958775, + "grad_norm": 1.1457053422927856, + "learning_rate": 2.104709433909804e-05, + "loss": 0.9756, + "step": 8667 + }, + { + "epoch": 0.5955750996289679, + "grad_norm": 1.0790817737579346, + "learning_rate": 2.1041173562240417e-05, + "loss": 0.9397, + "step": 8668 + }, + { + "epoch": 0.5956438092620585, + "grad_norm": 1.0754468441009521, + "learning_rate": 2.103525301313167e-05, + "loss": 0.9743, + "step": 8669 + }, + { + "epoch": 0.5957125188951491, + "grad_norm": 1.0722955465316772, + "learning_rate": 2.1029332692112406e-05, + "loss": 0.9328, + "step": 8670 + }, + { + "epoch": 0.5957812285282397, + "grad_norm": 1.077365517616272, + "learning_rate": 2.1023412599523204e-05, + "loss": 0.9511, + "step": 8671 + }, + { + "epoch": 0.5958499381613303, + "grad_norm": 1.0776400566101074, + "learning_rate": 2.1017492735704647e-05, + "loss": 0.9743, + "step": 8672 + }, + { + "epoch": 0.5959186477944208, + "grad_norm": 1.124054193496704, + "learning_rate": 2.1011573100997305e-05, + "loss": 0.9533, + "step": 8673 + }, + { + "epoch": 0.5959873574275113, + "grad_norm": 1.0153175592422485, + "learning_rate": 2.100565369574174e-05, + "loss": 0.943, + "step": 8674 + }, + { + "epoch": 0.5960560670606019, + "grad_norm": 1.017107367515564, + "learning_rate": 2.0999734520278476e-05, + "loss": 0.9355, + "step": 8675 + }, + { + "epoch": 0.5961247766936925, + "grad_norm": 1.0977834463119507, + "learning_rate": 2.0993815574948045e-05, + "loss": 0.9434, + "step": 8676 + }, + { + "epoch": 0.596193486326783, + "grad_norm": 1.0541061162948608, + "learning_rate": 2.0987896860090965e-05, + "loss": 0.9583, + "step": 8677 + }, + { + "epoch": 0.5962621959598736, + "grad_norm": 1.0843658447265625, + "learning_rate": 2.0981978376047727e-05, + "loss": 0.9618, + "step": 8678 + }, + { + "epoch": 0.5963309055929641, + "grad_norm": 1.0566246509552002, + "learning_rate": 2.0976060123158826e-05, + "loss": 0.931, + "step": 8679 + }, + { + "epoch": 0.5963996152260547, + "grad_norm": 1.123156189918518, + "learning_rate": 2.097014210176474e-05, + "loss": 0.9388, + "step": 8680 + }, + { + "epoch": 0.5964683248591452, + "grad_norm": 1.0416887998580933, + "learning_rate": 2.0964224312205913e-05, + "loss": 0.9612, + "step": 8681 + }, + { + "epoch": 0.5965370344922358, + "grad_norm": 1.0637091398239136, + "learning_rate": 2.0958306754822798e-05, + "loss": 0.9688, + "step": 8682 + }, + { + "epoch": 0.5966057441253264, + "grad_norm": 1.0039037466049194, + "learning_rate": 2.095238942995583e-05, + "loss": 0.9053, + "step": 8683 + }, + { + "epoch": 0.5966744537584169, + "grad_norm": 1.070523977279663, + "learning_rate": 2.0946472337945434e-05, + "loss": 0.9603, + "step": 8684 + }, + { + "epoch": 0.5967431633915075, + "grad_norm": 1.0902800559997559, + "learning_rate": 2.0940555479132024e-05, + "loss": 0.9869, + "step": 8685 + }, + { + "epoch": 0.596811873024598, + "grad_norm": 1.0427241325378418, + "learning_rate": 2.093463885385597e-05, + "loss": 0.97, + "step": 8686 + }, + { + "epoch": 0.5968805826576886, + "grad_norm": 1.0260999202728271, + "learning_rate": 2.0928722462457663e-05, + "loss": 0.9873, + "step": 8687 + }, + { + "epoch": 0.5969492922907792, + "grad_norm": 1.0407408475875854, + "learning_rate": 2.0922806305277476e-05, + "loss": 0.9608, + "step": 8688 + }, + { + "epoch": 0.5970180019238698, + "grad_norm": 1.125340461730957, + "learning_rate": 2.0916890382655752e-05, + "loss": 0.9573, + "step": 8689 + }, + { + "epoch": 0.5970867115569602, + "grad_norm": 1.0343691110610962, + "learning_rate": 2.0910974694932846e-05, + "loss": 0.9334, + "step": 8690 + }, + { + "epoch": 0.5971554211900508, + "grad_norm": 1.0379054546356201, + "learning_rate": 2.090505924244906e-05, + "loss": 0.9217, + "step": 8691 + }, + { + "epoch": 0.5972241308231414, + "grad_norm": 1.076706886291504, + "learning_rate": 2.0899144025544716e-05, + "loss": 0.9168, + "step": 8692 + }, + { + "epoch": 0.597292840456232, + "grad_norm": 1.1087309122085571, + "learning_rate": 2.089322904456012e-05, + "loss": 1.0035, + "step": 8693 + }, + { + "epoch": 0.5973615500893226, + "grad_norm": 1.0288314819335938, + "learning_rate": 2.0887314299835552e-05, + "loss": 0.9639, + "step": 8694 + }, + { + "epoch": 0.597430259722413, + "grad_norm": 1.0415581464767456, + "learning_rate": 2.0881399791711288e-05, + "loss": 0.946, + "step": 8695 + }, + { + "epoch": 0.5974989693555036, + "grad_norm": 1.073193907737732, + "learning_rate": 2.0875485520527587e-05, + "loss": 0.9459, + "step": 8696 + }, + { + "epoch": 0.5975676789885942, + "grad_norm": 1.0454269647598267, + "learning_rate": 2.0869571486624685e-05, + "loss": 0.9444, + "step": 8697 + }, + { + "epoch": 0.5976363886216848, + "grad_norm": 1.055361032485962, + "learning_rate": 2.0863657690342813e-05, + "loss": 0.9428, + "step": 8698 + }, + { + "epoch": 0.5977050982547754, + "grad_norm": 1.0348161458969116, + "learning_rate": 2.0857744132022196e-05, + "loss": 0.9227, + "step": 8699 + }, + { + "epoch": 0.5977738078878658, + "grad_norm": 1.060080885887146, + "learning_rate": 2.085183081200303e-05, + "loss": 0.9388, + "step": 8700 + }, + { + "epoch": 0.5978425175209564, + "grad_norm": 1.1550278663635254, + "learning_rate": 2.084591773062552e-05, + "loss": 0.9749, + "step": 8701 + }, + { + "epoch": 0.597911227154047, + "grad_norm": 1.0204272270202637, + "learning_rate": 2.084000488822982e-05, + "loss": 0.9573, + "step": 8702 + }, + { + "epoch": 0.5979799367871376, + "grad_norm": 1.1552367210388184, + "learning_rate": 2.0834092285156097e-05, + "loss": 0.9345, + "step": 8703 + }, + { + "epoch": 0.5980486464202281, + "grad_norm": 0.9998540878295898, + "learning_rate": 2.0828179921744513e-05, + "loss": 0.9424, + "step": 8704 + }, + { + "epoch": 0.5981173560533187, + "grad_norm": 1.0775669813156128, + "learning_rate": 2.082226779833519e-05, + "loss": 0.9425, + "step": 8705 + }, + { + "epoch": 0.5981860656864092, + "grad_norm": 1.0738190412521362, + "learning_rate": 2.081635591526826e-05, + "loss": 0.9442, + "step": 8706 + }, + { + "epoch": 0.5982547753194998, + "grad_norm": 1.02934992313385, + "learning_rate": 2.0810444272883817e-05, + "loss": 0.9552, + "step": 8707 + }, + { + "epoch": 0.5983234849525904, + "grad_norm": 1.0609931945800781, + "learning_rate": 2.080453287152196e-05, + "loss": 0.9521, + "step": 8708 + }, + { + "epoch": 0.5983921945856809, + "grad_norm": 1.010019302368164, + "learning_rate": 2.0798621711522768e-05, + "loss": 0.8751, + "step": 8709 + }, + { + "epoch": 0.5984609042187715, + "grad_norm": 1.055459976196289, + "learning_rate": 2.0792710793226313e-05, + "loss": 0.9067, + "step": 8710 + }, + { + "epoch": 0.598529613851862, + "grad_norm": 1.0549994707107544, + "learning_rate": 2.0786800116972643e-05, + "loss": 0.9667, + "step": 8711 + }, + { + "epoch": 0.5985983234849526, + "grad_norm": 1.0478159189224243, + "learning_rate": 2.0780889683101788e-05, + "loss": 0.944, + "step": 8712 + }, + { + "epoch": 0.5986670331180431, + "grad_norm": 1.073844313621521, + "learning_rate": 2.0774979491953777e-05, + "loss": 0.9567, + "step": 8713 + }, + { + "epoch": 0.5987357427511337, + "grad_norm": 1.0329818725585938, + "learning_rate": 2.076906954386862e-05, + "loss": 0.9379, + "step": 8714 + }, + { + "epoch": 0.5988044523842243, + "grad_norm": 1.0791375637054443, + "learning_rate": 2.0763159839186308e-05, + "loss": 1.0061, + "step": 8715 + }, + { + "epoch": 0.5988731620173148, + "grad_norm": 1.073974370956421, + "learning_rate": 2.075725037824684e-05, + "loss": 0.9637, + "step": 8716 + }, + { + "epoch": 0.5989418716504054, + "grad_norm": 1.0549582242965698, + "learning_rate": 2.075134116139016e-05, + "loss": 0.9726, + "step": 8717 + }, + { + "epoch": 0.5990105812834959, + "grad_norm": 1.1246488094329834, + "learning_rate": 2.0745432188956232e-05, + "loss": 0.9559, + "step": 8718 + }, + { + "epoch": 0.5990792909165865, + "grad_norm": 1.0625227689743042, + "learning_rate": 2.0739523461284995e-05, + "loss": 0.9575, + "step": 8719 + }, + { + "epoch": 0.5991480005496771, + "grad_norm": 1.0378797054290771, + "learning_rate": 2.0733614978716378e-05, + "loss": 0.935, + "step": 8720 + }, + { + "epoch": 0.5992167101827677, + "grad_norm": 1.0627779960632324, + "learning_rate": 2.07277067415903e-05, + "loss": 0.9406, + "step": 8721 + }, + { + "epoch": 0.5992854198158581, + "grad_norm": 0.9959418177604675, + "learning_rate": 2.0721798750246636e-05, + "loss": 0.9588, + "step": 8722 + }, + { + "epoch": 0.5993541294489487, + "grad_norm": 1.1032923460006714, + "learning_rate": 2.071589100502529e-05, + "loss": 0.9831, + "step": 8723 + }, + { + "epoch": 0.5994228390820393, + "grad_norm": 1.04517662525177, + "learning_rate": 2.070998350626611e-05, + "loss": 0.9361, + "step": 8724 + }, + { + "epoch": 0.5994915487151299, + "grad_norm": 1.1122007369995117, + "learning_rate": 2.0704076254308972e-05, + "loss": 1.0017, + "step": 8725 + }, + { + "epoch": 0.5995602583482205, + "grad_norm": 1.0198721885681152, + "learning_rate": 2.0698169249493714e-05, + "loss": 0.944, + "step": 8726 + }, + { + "epoch": 0.5996289679813109, + "grad_norm": 1.083479404449463, + "learning_rate": 2.0692262492160143e-05, + "loss": 0.9592, + "step": 8727 + }, + { + "epoch": 0.5996976776144015, + "grad_norm": 1.0317647457122803, + "learning_rate": 2.0686355982648088e-05, + "loss": 0.9572, + "step": 8728 + }, + { + "epoch": 0.5997663872474921, + "grad_norm": 1.0734529495239258, + "learning_rate": 2.0680449721297344e-05, + "loss": 0.9306, + "step": 8729 + }, + { + "epoch": 0.5998350968805827, + "grad_norm": 1.0666356086730957, + "learning_rate": 2.0674543708447692e-05, + "loss": 0.9845, + "step": 8730 + }, + { + "epoch": 0.5999038065136733, + "grad_norm": 1.030983328819275, + "learning_rate": 2.0668637944438912e-05, + "loss": 0.9618, + "step": 8731 + }, + { + "epoch": 0.5999725161467637, + "grad_norm": 1.0081413984298706, + "learning_rate": 2.066273242961074e-05, + "loss": 0.9308, + "step": 8732 + }, + { + "epoch": 0.6000412257798543, + "grad_norm": 1.0417424440383911, + "learning_rate": 2.0656827164302927e-05, + "loss": 0.9012, + "step": 8733 + }, + { + "epoch": 0.6001099354129449, + "grad_norm": 1.0331727266311646, + "learning_rate": 2.0650922148855208e-05, + "loss": 0.9402, + "step": 8734 + }, + { + "epoch": 0.6001786450460355, + "grad_norm": 1.062338948249817, + "learning_rate": 2.0645017383607276e-05, + "loss": 0.9698, + "step": 8735 + }, + { + "epoch": 0.600247354679126, + "grad_norm": 1.1157009601593018, + "learning_rate": 2.0639112868898845e-05, + "loss": 0.9546, + "step": 8736 + }, + { + "epoch": 0.6003160643122166, + "grad_norm": 1.0517983436584473, + "learning_rate": 2.0633208605069593e-05, + "loss": 0.9485, + "step": 8737 + }, + { + "epoch": 0.6003847739453071, + "grad_norm": 1.0826389789581299, + "learning_rate": 2.0627304592459183e-05, + "loss": 0.9568, + "step": 8738 + }, + { + "epoch": 0.6004534835783977, + "grad_norm": 1.049261450767517, + "learning_rate": 2.062140083140727e-05, + "loss": 0.9618, + "step": 8739 + }, + { + "epoch": 0.6005221932114883, + "grad_norm": 1.0119612216949463, + "learning_rate": 2.06154973222535e-05, + "loss": 0.9249, + "step": 8740 + }, + { + "epoch": 0.6005909028445788, + "grad_norm": 1.0490816831588745, + "learning_rate": 2.0609594065337506e-05, + "loss": 0.9658, + "step": 8741 + }, + { + "epoch": 0.6006596124776694, + "grad_norm": 1.0926406383514404, + "learning_rate": 2.0603691060998874e-05, + "loss": 0.9436, + "step": 8742 + }, + { + "epoch": 0.6007283221107599, + "grad_norm": 1.0500818490982056, + "learning_rate": 2.0597788309577222e-05, + "loss": 0.9628, + "step": 8743 + }, + { + "epoch": 0.6007970317438505, + "grad_norm": 0.9725491404533386, + "learning_rate": 2.059188581141212e-05, + "loss": 0.9095, + "step": 8744 + }, + { + "epoch": 0.600865741376941, + "grad_norm": 1.0330381393432617, + "learning_rate": 2.0585983566843145e-05, + "loss": 0.9293, + "step": 8745 + }, + { + "epoch": 0.6009344510100316, + "grad_norm": 0.9649110436439514, + "learning_rate": 2.058008157620984e-05, + "loss": 0.8861, + "step": 8746 + }, + { + "epoch": 0.6010031606431222, + "grad_norm": 1.0719375610351562, + "learning_rate": 2.0574179839851747e-05, + "loss": 0.9478, + "step": 8747 + }, + { + "epoch": 0.6010718702762127, + "grad_norm": 1.1144722700119019, + "learning_rate": 2.0568278358108394e-05, + "loss": 0.9232, + "step": 8748 + }, + { + "epoch": 0.6011405799093033, + "grad_norm": 1.061164140701294, + "learning_rate": 2.0562377131319284e-05, + "loss": 0.9762, + "step": 8749 + }, + { + "epoch": 0.6012092895423938, + "grad_norm": 1.0359314680099487, + "learning_rate": 2.0556476159823902e-05, + "loss": 0.9594, + "step": 8750 + }, + { + "epoch": 0.6012779991754844, + "grad_norm": 1.0498729944229126, + "learning_rate": 2.0550575443961746e-05, + "loss": 0.928, + "step": 8751 + }, + { + "epoch": 0.601346708808575, + "grad_norm": 1.102211594581604, + "learning_rate": 2.0544674984072276e-05, + "loss": 0.9519, + "step": 8752 + }, + { + "epoch": 0.6014154184416656, + "grad_norm": 1.0825600624084473, + "learning_rate": 2.053877478049493e-05, + "loss": 0.9629, + "step": 8753 + }, + { + "epoch": 0.601484128074756, + "grad_norm": 1.0574942827224731, + "learning_rate": 2.053287483356915e-05, + "loss": 0.9626, + "step": 8754 + }, + { + "epoch": 0.6015528377078466, + "grad_norm": 1.069841980934143, + "learning_rate": 2.052697514363436e-05, + "loss": 0.9385, + "step": 8755 + }, + { + "epoch": 0.6016215473409372, + "grad_norm": 1.0652754306793213, + "learning_rate": 2.0521075711029963e-05, + "loss": 0.9377, + "step": 8756 + }, + { + "epoch": 0.6016902569740278, + "grad_norm": 1.0442314147949219, + "learning_rate": 2.0515176536095352e-05, + "loss": 0.9447, + "step": 8757 + }, + { + "epoch": 0.6017589666071184, + "grad_norm": 1.0733866691589355, + "learning_rate": 2.0509277619169894e-05, + "loss": 0.9278, + "step": 8758 + }, + { + "epoch": 0.6018276762402088, + "grad_norm": 1.079131007194519, + "learning_rate": 2.050337896059296e-05, + "loss": 0.9816, + "step": 8759 + }, + { + "epoch": 0.6018963858732994, + "grad_norm": 1.0635396242141724, + "learning_rate": 2.049748056070389e-05, + "loss": 0.9472, + "step": 8760 + }, + { + "epoch": 0.60196509550639, + "grad_norm": 1.0335851907730103, + "learning_rate": 2.049158241984202e-05, + "loss": 0.9329, + "step": 8761 + }, + { + "epoch": 0.6020338051394806, + "grad_norm": 1.0518336296081543, + "learning_rate": 2.0485684538346673e-05, + "loss": 0.929, + "step": 8762 + }, + { + "epoch": 0.6021025147725712, + "grad_norm": 1.0316444635391235, + "learning_rate": 2.0479786916557132e-05, + "loss": 0.9432, + "step": 8763 + }, + { + "epoch": 0.6021712244056616, + "grad_norm": 1.138946533203125, + "learning_rate": 2.0473889554812693e-05, + "loss": 0.9775, + "step": 8764 + }, + { + "epoch": 0.6022399340387522, + "grad_norm": 1.0342596769332886, + "learning_rate": 2.0467992453452626e-05, + "loss": 0.9462, + "step": 8765 + }, + { + "epoch": 0.6023086436718428, + "grad_norm": 1.0647504329681396, + "learning_rate": 2.0462095612816192e-05, + "loss": 0.966, + "step": 8766 + }, + { + "epoch": 0.6023773533049334, + "grad_norm": 1.1190409660339355, + "learning_rate": 2.0456199033242645e-05, + "loss": 0.9569, + "step": 8767 + }, + { + "epoch": 0.6024460629380239, + "grad_norm": 1.0492653846740723, + "learning_rate": 2.045030271507118e-05, + "loss": 0.9318, + "step": 8768 + }, + { + "epoch": 0.6025147725711145, + "grad_norm": 1.0946234464645386, + "learning_rate": 2.0444406658641026e-05, + "loss": 0.9537, + "step": 8769 + }, + { + "epoch": 0.602583482204205, + "grad_norm": 1.009356141090393, + "learning_rate": 2.0438510864291387e-05, + "loss": 0.9335, + "step": 8770 + }, + { + "epoch": 0.6026521918372956, + "grad_norm": 1.0607550144195557, + "learning_rate": 2.0432615332361433e-05, + "loss": 0.9288, + "step": 8771 + }, + { + "epoch": 0.6027209014703862, + "grad_norm": 1.0455714464187622, + "learning_rate": 2.0426720063190335e-05, + "loss": 0.9161, + "step": 8772 + }, + { + "epoch": 0.6027896111034767, + "grad_norm": 1.0657906532287598, + "learning_rate": 2.042082505711725e-05, + "loss": 0.9927, + "step": 8773 + }, + { + "epoch": 0.6028583207365673, + "grad_norm": 1.0836628675460815, + "learning_rate": 2.0414930314481296e-05, + "loss": 0.9653, + "step": 8774 + }, + { + "epoch": 0.6029270303696578, + "grad_norm": 1.0946381092071533, + "learning_rate": 2.0409035835621607e-05, + "loss": 0.9819, + "step": 8775 + }, + { + "epoch": 0.6029957400027484, + "grad_norm": 1.05648672580719, + "learning_rate": 2.0403141620877285e-05, + "loss": 0.9488, + "step": 8776 + }, + { + "epoch": 0.6030644496358389, + "grad_norm": 1.1172930002212524, + "learning_rate": 2.039724767058744e-05, + "loss": 1.0082, + "step": 8777 + }, + { + "epoch": 0.6031331592689295, + "grad_norm": 1.0359981060028076, + "learning_rate": 2.039135398509111e-05, + "loss": 0.9407, + "step": 8778 + }, + { + "epoch": 0.6032018689020201, + "grad_norm": 1.0897821187973022, + "learning_rate": 2.0385460564727383e-05, + "loss": 0.9499, + "step": 8779 + }, + { + "epoch": 0.6032705785351106, + "grad_norm": 1.0935345888137817, + "learning_rate": 2.0379567409835293e-05, + "loss": 0.9457, + "step": 8780 + }, + { + "epoch": 0.6033392881682011, + "grad_norm": 1.1423364877700806, + "learning_rate": 2.0373674520753877e-05, + "loss": 0.9714, + "step": 8781 + }, + { + "epoch": 0.6034079978012917, + "grad_norm": 1.0500028133392334, + "learning_rate": 2.0367781897822147e-05, + "loss": 0.9458, + "step": 8782 + }, + { + "epoch": 0.6034767074343823, + "grad_norm": 1.0201109647750854, + "learning_rate": 2.036188954137909e-05, + "loss": 0.9183, + "step": 8783 + }, + { + "epoch": 0.6035454170674729, + "grad_norm": 1.0678023099899292, + "learning_rate": 2.0355997451763713e-05, + "loss": 0.935, + "step": 8784 + }, + { + "epoch": 0.6036141267005635, + "grad_norm": 1.0518240928649902, + "learning_rate": 2.035010562931496e-05, + "loss": 0.9502, + "step": 8785 + }, + { + "epoch": 0.6036828363336539, + "grad_norm": 1.1021757125854492, + "learning_rate": 2.03442140743718e-05, + "loss": 0.9651, + "step": 8786 + }, + { + "epoch": 0.6037515459667445, + "grad_norm": 1.0707334280014038, + "learning_rate": 2.0338322787273172e-05, + "loss": 0.9917, + "step": 8787 + }, + { + "epoch": 0.6038202555998351, + "grad_norm": 1.0630691051483154, + "learning_rate": 2.033243176835798e-05, + "loss": 0.9669, + "step": 8788 + }, + { + "epoch": 0.6038889652329257, + "grad_norm": 1.055376410484314, + "learning_rate": 2.0326541017965143e-05, + "loss": 0.9147, + "step": 8789 + }, + { + "epoch": 0.6039576748660163, + "grad_norm": 1.1224926710128784, + "learning_rate": 2.032065053643355e-05, + "loss": 1.0121, + "step": 8790 + }, + { + "epoch": 0.6040263844991067, + "grad_norm": 1.0542795658111572, + "learning_rate": 2.0314760324102085e-05, + "loss": 0.9458, + "step": 8791 + }, + { + "epoch": 0.6040950941321973, + "grad_norm": 1.0442794561386108, + "learning_rate": 2.0308870381309607e-05, + "loss": 0.9699, + "step": 8792 + }, + { + "epoch": 0.6041638037652879, + "grad_norm": 1.0638765096664429, + "learning_rate": 2.030298070839495e-05, + "loss": 0.9669, + "step": 8793 + }, + { + "epoch": 0.6042325133983785, + "grad_norm": 1.046014428138733, + "learning_rate": 2.0297091305696943e-05, + "loss": 0.947, + "step": 8794 + }, + { + "epoch": 0.604301223031469, + "grad_norm": 0.9870065450668335, + "learning_rate": 2.0291202173554414e-05, + "loss": 0.9526, + "step": 8795 + }, + { + "epoch": 0.6043699326645595, + "grad_norm": 1.0540623664855957, + "learning_rate": 2.028531331230615e-05, + "loss": 0.9511, + "step": 8796 + }, + { + "epoch": 0.6044386422976501, + "grad_norm": 1.088005781173706, + "learning_rate": 2.027942472229094e-05, + "loss": 0.9737, + "step": 8797 + }, + { + "epoch": 0.6045073519307407, + "grad_norm": 1.1167956590652466, + "learning_rate": 2.0273536403847542e-05, + "loss": 0.9998, + "step": 8798 + }, + { + "epoch": 0.6045760615638313, + "grad_norm": 1.0825997591018677, + "learning_rate": 2.026764835731471e-05, + "loss": 0.9655, + "step": 8799 + }, + { + "epoch": 0.6046447711969218, + "grad_norm": 1.0790698528289795, + "learning_rate": 2.0261760583031186e-05, + "loss": 0.9433, + "step": 8800 + }, + { + "epoch": 0.6047134808300124, + "grad_norm": 1.055609107017517, + "learning_rate": 2.0255873081335685e-05, + "loss": 0.8995, + "step": 8801 + }, + { + "epoch": 0.6047821904631029, + "grad_norm": 1.0028069019317627, + "learning_rate": 2.0249985852566925e-05, + "loss": 0.9297, + "step": 8802 + }, + { + "epoch": 0.6048509000961935, + "grad_norm": 1.0794270038604736, + "learning_rate": 2.0244098897063568e-05, + "loss": 1.0057, + "step": 8803 + }, + { + "epoch": 0.604919609729284, + "grad_norm": 1.1042420864105225, + "learning_rate": 2.0238212215164307e-05, + "loss": 0.9714, + "step": 8804 + }, + { + "epoch": 0.6049883193623746, + "grad_norm": 1.064664602279663, + "learning_rate": 2.023232580720779e-05, + "loss": 0.9392, + "step": 8805 + }, + { + "epoch": 0.6050570289954652, + "grad_norm": 1.0340029001235962, + "learning_rate": 2.0226439673532668e-05, + "loss": 0.9639, + "step": 8806 + }, + { + "epoch": 0.6051257386285557, + "grad_norm": 1.0376945734024048, + "learning_rate": 2.0220553814477557e-05, + "loss": 0.9508, + "step": 8807 + }, + { + "epoch": 0.6051944482616463, + "grad_norm": 1.0821436643600464, + "learning_rate": 2.021466823038108e-05, + "loss": 0.9579, + "step": 8808 + }, + { + "epoch": 0.6052631578947368, + "grad_norm": 1.1002475023269653, + "learning_rate": 2.020878292158181e-05, + "loss": 0.9694, + "step": 8809 + }, + { + "epoch": 0.6053318675278274, + "grad_norm": 1.053646445274353, + "learning_rate": 2.020289788841834e-05, + "loss": 0.9544, + "step": 8810 + }, + { + "epoch": 0.605400577160918, + "grad_norm": 1.0188701152801514, + "learning_rate": 2.019701313122923e-05, + "loss": 0.9172, + "step": 8811 + }, + { + "epoch": 0.6054692867940085, + "grad_norm": 1.1106960773468018, + "learning_rate": 2.0191128650353026e-05, + "loss": 0.9791, + "step": 8812 + }, + { + "epoch": 0.605537996427099, + "grad_norm": 1.1262898445129395, + "learning_rate": 2.0185244446128267e-05, + "loss": 0.9615, + "step": 8813 + }, + { + "epoch": 0.6056067060601896, + "grad_norm": 1.0314570665359497, + "learning_rate": 2.017936051889345e-05, + "loss": 0.9326, + "step": 8814 + }, + { + "epoch": 0.6056754156932802, + "grad_norm": 1.0685030221939087, + "learning_rate": 2.0173476868987086e-05, + "loss": 0.9567, + "step": 8815 + }, + { + "epoch": 0.6057441253263708, + "grad_norm": 1.065748691558838, + "learning_rate": 2.016759349674765e-05, + "loss": 0.9795, + "step": 8816 + }, + { + "epoch": 0.6058128349594614, + "grad_norm": 1.0742942094802856, + "learning_rate": 2.016171040251362e-05, + "loss": 0.9837, + "step": 8817 + }, + { + "epoch": 0.6058815445925518, + "grad_norm": 1.003937005996704, + "learning_rate": 2.015582758662344e-05, + "loss": 0.9564, + "step": 8818 + }, + { + "epoch": 0.6059502542256424, + "grad_norm": 1.0198386907577515, + "learning_rate": 2.0149945049415548e-05, + "loss": 0.9059, + "step": 8819 + }, + { + "epoch": 0.606018963858733, + "grad_norm": 1.0754082202911377, + "learning_rate": 2.0144062791228352e-05, + "loss": 0.9571, + "step": 8820 + }, + { + "epoch": 0.6060876734918236, + "grad_norm": 1.0518404245376587, + "learning_rate": 2.0138180812400265e-05, + "loss": 0.9781, + "step": 8821 + }, + { + "epoch": 0.6061563831249142, + "grad_norm": 1.0800223350524902, + "learning_rate": 2.013229911326967e-05, + "loss": 0.9786, + "step": 8822 + }, + { + "epoch": 0.6062250927580046, + "grad_norm": 1.1307330131530762, + "learning_rate": 2.012641769417495e-05, + "loss": 0.9745, + "step": 8823 + }, + { + "epoch": 0.6062938023910952, + "grad_norm": 1.072396993637085, + "learning_rate": 2.012053655545444e-05, + "loss": 0.9651, + "step": 8824 + }, + { + "epoch": 0.6063625120241858, + "grad_norm": 1.0248361825942993, + "learning_rate": 2.0114655697446484e-05, + "loss": 0.9047, + "step": 8825 + }, + { + "epoch": 0.6064312216572764, + "grad_norm": 1.0842866897583008, + "learning_rate": 2.0108775120489405e-05, + "loss": 0.9585, + "step": 8826 + }, + { + "epoch": 0.606499931290367, + "grad_norm": 1.0478782653808594, + "learning_rate": 2.0102894824921515e-05, + "loss": 0.966, + "step": 8827 + }, + { + "epoch": 0.6065686409234574, + "grad_norm": 1.0886378288269043, + "learning_rate": 2.0097014811081107e-05, + "loss": 0.9899, + "step": 8828 + }, + { + "epoch": 0.606637350556548, + "grad_norm": 1.1186963319778442, + "learning_rate": 2.009113507930644e-05, + "loss": 0.9552, + "step": 8829 + }, + { + "epoch": 0.6067060601896386, + "grad_norm": 1.0079166889190674, + "learning_rate": 2.0085255629935783e-05, + "loss": 0.9508, + "step": 8830 + }, + { + "epoch": 0.6067747698227292, + "grad_norm": 1.049862027168274, + "learning_rate": 2.0079376463307368e-05, + "loss": 0.9663, + "step": 8831 + }, + { + "epoch": 0.6068434794558197, + "grad_norm": 0.9785298109054565, + "learning_rate": 2.0073497579759424e-05, + "loss": 0.9117, + "step": 8832 + }, + { + "epoch": 0.6069121890889103, + "grad_norm": 1.047663927078247, + "learning_rate": 2.006761897963017e-05, + "loss": 0.941, + "step": 8833 + }, + { + "epoch": 0.6069808987220008, + "grad_norm": 1.0222848653793335, + "learning_rate": 2.006174066325778e-05, + "loss": 0.9006, + "step": 8834 + }, + { + "epoch": 0.6070496083550914, + "grad_norm": 1.0338807106018066, + "learning_rate": 2.0055862630980438e-05, + "loss": 0.9481, + "step": 8835 + }, + { + "epoch": 0.6071183179881819, + "grad_norm": 1.0719283819198608, + "learning_rate": 2.0049984883136306e-05, + "loss": 0.9309, + "step": 8836 + }, + { + "epoch": 0.6071870276212725, + "grad_norm": 1.0228372812271118, + "learning_rate": 2.004410742006352e-05, + "loss": 0.9581, + "step": 8837 + }, + { + "epoch": 0.6072557372543631, + "grad_norm": 1.0632877349853516, + "learning_rate": 2.0038230242100227e-05, + "loss": 0.9735, + "step": 8838 + }, + { + "epoch": 0.6073244468874536, + "grad_norm": 1.064202070236206, + "learning_rate": 2.0032353349584514e-05, + "loss": 0.9494, + "step": 8839 + }, + { + "epoch": 0.6073931565205442, + "grad_norm": 1.0298054218292236, + "learning_rate": 2.002647674285448e-05, + "loss": 0.9746, + "step": 8840 + }, + { + "epoch": 0.6074618661536347, + "grad_norm": 1.082353115081787, + "learning_rate": 2.0020600422248214e-05, + "loss": 0.9342, + "step": 8841 + }, + { + "epoch": 0.6075305757867253, + "grad_norm": 1.0630178451538086, + "learning_rate": 2.0014724388103763e-05, + "loss": 0.9584, + "step": 8842 + }, + { + "epoch": 0.6075992854198159, + "grad_norm": 1.0731428861618042, + "learning_rate": 2.0008848640759186e-05, + "loss": 0.9481, + "step": 8843 + }, + { + "epoch": 0.6076679950529064, + "grad_norm": 1.0500524044036865, + "learning_rate": 2.0002973180552504e-05, + "loss": 0.9561, + "step": 8844 + }, + { + "epoch": 0.6077367046859969, + "grad_norm": 1.0358256101608276, + "learning_rate": 1.999709800782172e-05, + "loss": 0.9065, + "step": 8845 + }, + { + "epoch": 0.6078054143190875, + "grad_norm": 1.0853990316390991, + "learning_rate": 1.9991223122904835e-05, + "loss": 0.9713, + "step": 8846 + }, + { + "epoch": 0.6078741239521781, + "grad_norm": 1.0735313892364502, + "learning_rate": 1.9985348526139832e-05, + "loss": 0.9168, + "step": 8847 + }, + { + "epoch": 0.6079428335852687, + "grad_norm": 1.0692994594573975, + "learning_rate": 1.9979474217864683e-05, + "loss": 0.9804, + "step": 8848 + }, + { + "epoch": 0.6080115432183593, + "grad_norm": 1.051141381263733, + "learning_rate": 1.9973600198417307e-05, + "loss": 0.9372, + "step": 8849 + }, + { + "epoch": 0.6080802528514497, + "grad_norm": 1.074462652206421, + "learning_rate": 1.996772646813565e-05, + "loss": 0.9826, + "step": 8850 + }, + { + "epoch": 0.6081489624845403, + "grad_norm": 1.0802773237228394, + "learning_rate": 1.996185302735762e-05, + "loss": 0.9218, + "step": 8851 + }, + { + "epoch": 0.6082176721176309, + "grad_norm": 1.062247633934021, + "learning_rate": 1.995597987642112e-05, + "loss": 0.9562, + "step": 8852 + }, + { + "epoch": 0.6082863817507215, + "grad_norm": 1.0591323375701904, + "learning_rate": 1.995010701566402e-05, + "loss": 0.91, + "step": 8853 + }, + { + "epoch": 0.608355091383812, + "grad_norm": 1.1370303630828857, + "learning_rate": 1.9944234445424182e-05, + "loss": 0.9644, + "step": 8854 + }, + { + "epoch": 0.6084238010169025, + "grad_norm": 1.0487895011901855, + "learning_rate": 1.9938362166039455e-05, + "loss": 0.9827, + "step": 8855 + }, + { + "epoch": 0.6084925106499931, + "grad_norm": 1.1328208446502686, + "learning_rate": 1.993249017784766e-05, + "loss": 0.9577, + "step": 8856 + }, + { + "epoch": 0.6085612202830837, + "grad_norm": 1.0983558893203735, + "learning_rate": 1.992661848118662e-05, + "loss": 0.9549, + "step": 8857 + }, + { + "epoch": 0.6086299299161743, + "grad_norm": 1.0783991813659668, + "learning_rate": 1.9920747076394138e-05, + "loss": 0.9472, + "step": 8858 + }, + { + "epoch": 0.6086986395492648, + "grad_norm": 1.0081523656845093, + "learning_rate": 1.9914875963807965e-05, + "loss": 0.9335, + "step": 8859 + }, + { + "epoch": 0.6087673491823554, + "grad_norm": 1.0321587324142456, + "learning_rate": 1.9909005143765874e-05, + "loss": 0.9136, + "step": 8860 + }, + { + "epoch": 0.6088360588154459, + "grad_norm": 1.069655179977417, + "learning_rate": 1.990313461660562e-05, + "loss": 0.9901, + "step": 8861 + }, + { + "epoch": 0.6089047684485365, + "grad_norm": 1.020999550819397, + "learning_rate": 1.9897264382664916e-05, + "loss": 0.933, + "step": 8862 + }, + { + "epoch": 0.608973478081627, + "grad_norm": 1.0407865047454834, + "learning_rate": 1.989139444228149e-05, + "loss": 0.9596, + "step": 8863 + }, + { + "epoch": 0.6090421877147176, + "grad_norm": 1.1384217739105225, + "learning_rate": 1.988552479579303e-05, + "loss": 0.9987, + "step": 8864 + }, + { + "epoch": 0.6091108973478082, + "grad_norm": 1.0868481397628784, + "learning_rate": 1.98796554435372e-05, + "loss": 0.9131, + "step": 8865 + }, + { + "epoch": 0.6091796069808987, + "grad_norm": 1.0782986879348755, + "learning_rate": 1.9873786385851677e-05, + "loss": 0.9508, + "step": 8866 + }, + { + "epoch": 0.6092483166139893, + "grad_norm": 1.0819249153137207, + "learning_rate": 1.9867917623074095e-05, + "loss": 0.9629, + "step": 8867 + }, + { + "epoch": 0.6093170262470798, + "grad_norm": 1.1280854940414429, + "learning_rate": 1.9862049155542078e-05, + "loss": 0.9482, + "step": 8868 + }, + { + "epoch": 0.6093857358801704, + "grad_norm": 1.108609914779663, + "learning_rate": 1.9856180983593256e-05, + "loss": 0.984, + "step": 8869 + }, + { + "epoch": 0.609454445513261, + "grad_norm": 1.0523027181625366, + "learning_rate": 1.9850313107565193e-05, + "loss": 0.899, + "step": 8870 + }, + { + "epoch": 0.6095231551463515, + "grad_norm": 1.0976531505584717, + "learning_rate": 1.9844445527795475e-05, + "loss": 0.9671, + "step": 8871 + }, + { + "epoch": 0.609591864779442, + "grad_norm": 1.0578479766845703, + "learning_rate": 1.9838578244621665e-05, + "loss": 0.911, + "step": 8872 + }, + { + "epoch": 0.6096605744125326, + "grad_norm": 1.0431283712387085, + "learning_rate": 1.98327112583813e-05, + "loss": 0.9349, + "step": 8873 + }, + { + "epoch": 0.6097292840456232, + "grad_norm": 1.1828677654266357, + "learning_rate": 1.9826844569411917e-05, + "loss": 1.0355, + "step": 8874 + }, + { + "epoch": 0.6097979936787138, + "grad_norm": 1.0955836772918701, + "learning_rate": 1.9820978178051e-05, + "loss": 0.948, + "step": 8875 + }, + { + "epoch": 0.6098667033118044, + "grad_norm": 1.1135927438735962, + "learning_rate": 1.981511208463605e-05, + "loss": 0.9663, + "step": 8876 + }, + { + "epoch": 0.6099354129448948, + "grad_norm": 1.0787875652313232, + "learning_rate": 1.9809246289504548e-05, + "loss": 0.9757, + "step": 8877 + }, + { + "epoch": 0.6100041225779854, + "grad_norm": 1.0165647268295288, + "learning_rate": 1.9803380792993935e-05, + "loss": 0.958, + "step": 8878 + }, + { + "epoch": 0.610072832211076, + "grad_norm": 1.053451657295227, + "learning_rate": 1.9797515595441667e-05, + "loss": 0.9659, + "step": 8879 + }, + { + "epoch": 0.6101415418441666, + "grad_norm": 1.0766078233718872, + "learning_rate": 1.9791650697185147e-05, + "loss": 0.9818, + "step": 8880 + }, + { + "epoch": 0.6102102514772572, + "grad_norm": 1.0503193140029907, + "learning_rate": 1.9785786098561785e-05, + "loss": 0.9627, + "step": 8881 + }, + { + "epoch": 0.6102789611103476, + "grad_norm": 1.0670135021209717, + "learning_rate": 1.9779921799908968e-05, + "loss": 0.9332, + "step": 8882 + }, + { + "epoch": 0.6103476707434382, + "grad_norm": 1.0751127004623413, + "learning_rate": 1.9774057801564066e-05, + "loss": 0.984, + "step": 8883 + }, + { + "epoch": 0.6104163803765288, + "grad_norm": 1.0707913637161255, + "learning_rate": 1.9768194103864446e-05, + "loss": 0.9623, + "step": 8884 + }, + { + "epoch": 0.6104850900096194, + "grad_norm": 1.1064833402633667, + "learning_rate": 1.976233070714742e-05, + "loss": 0.9522, + "step": 8885 + }, + { + "epoch": 0.61055379964271, + "grad_norm": 0.9976239800453186, + "learning_rate": 1.975646761175031e-05, + "loss": 0.9109, + "step": 8886 + }, + { + "epoch": 0.6106225092758004, + "grad_norm": 1.014216661453247, + "learning_rate": 1.9750604818010425e-05, + "loss": 0.9698, + "step": 8887 + }, + { + "epoch": 0.610691218908891, + "grad_norm": 1.050604224205017, + "learning_rate": 1.974474232626505e-05, + "loss": 0.9254, + "step": 8888 + }, + { + "epoch": 0.6107599285419816, + "grad_norm": 1.0048243999481201, + "learning_rate": 1.973888013685144e-05, + "loss": 0.9364, + "step": 8889 + }, + { + "epoch": 0.6108286381750722, + "grad_norm": 1.0655728578567505, + "learning_rate": 1.973301825010685e-05, + "loss": 0.9508, + "step": 8890 + }, + { + "epoch": 0.6108973478081627, + "grad_norm": 1.1956934928894043, + "learning_rate": 1.97271566663685e-05, + "loss": 0.9356, + "step": 8891 + }, + { + "epoch": 0.6109660574412533, + "grad_norm": 1.0939905643463135, + "learning_rate": 1.972129538597362e-05, + "loss": 0.9693, + "step": 8892 + }, + { + "epoch": 0.6110347670743438, + "grad_norm": 1.092108964920044, + "learning_rate": 1.9715434409259392e-05, + "loss": 0.9607, + "step": 8893 + }, + { + "epoch": 0.6111034767074344, + "grad_norm": 1.087136149406433, + "learning_rate": 1.970957373656301e-05, + "loss": 0.9596, + "step": 8894 + }, + { + "epoch": 0.611172186340525, + "grad_norm": 1.1015759706497192, + "learning_rate": 1.970371336822162e-05, + "loss": 0.9859, + "step": 8895 + }, + { + "epoch": 0.6112408959736155, + "grad_norm": 1.0377388000488281, + "learning_rate": 1.9697853304572368e-05, + "loss": 0.9273, + "step": 8896 + }, + { + "epoch": 0.6113096056067061, + "grad_norm": 1.0143009424209595, + "learning_rate": 1.969199354595238e-05, + "loss": 0.9523, + "step": 8897 + }, + { + "epoch": 0.6113783152397966, + "grad_norm": 1.0801914930343628, + "learning_rate": 1.9686134092698768e-05, + "loss": 0.9572, + "step": 8898 + }, + { + "epoch": 0.6114470248728872, + "grad_norm": 1.0849347114562988, + "learning_rate": 1.9680274945148633e-05, + "loss": 0.9517, + "step": 8899 + }, + { + "epoch": 0.6115157345059777, + "grad_norm": 1.0440565347671509, + "learning_rate": 1.9674416103639028e-05, + "loss": 0.9786, + "step": 8900 + }, + { + "epoch": 0.6115844441390683, + "grad_norm": 1.05801522731781, + "learning_rate": 1.9668557568507016e-05, + "loss": 0.9276, + "step": 8901 + }, + { + "epoch": 0.6116531537721589, + "grad_norm": 1.0999804735183716, + "learning_rate": 1.9662699340089642e-05, + "loss": 0.9447, + "step": 8902 + }, + { + "epoch": 0.6117218634052494, + "grad_norm": 1.0126975774765015, + "learning_rate": 1.965684141872391e-05, + "loss": 0.9341, + "step": 8903 + }, + { + "epoch": 0.61179057303834, + "grad_norm": 1.0927927494049072, + "learning_rate": 1.965098380474685e-05, + "loss": 0.9678, + "step": 8904 + }, + { + "epoch": 0.6118592826714305, + "grad_norm": 1.0450774431228638, + "learning_rate": 1.964512649849542e-05, + "loss": 0.9514, + "step": 8905 + }, + { + "epoch": 0.6119279923045211, + "grad_norm": 1.0487409830093384, + "learning_rate": 1.9639269500306595e-05, + "loss": 0.9525, + "step": 8906 + }, + { + "epoch": 0.6119967019376117, + "grad_norm": 1.028035283088684, + "learning_rate": 1.9633412810517327e-05, + "loss": 0.9474, + "step": 8907 + }, + { + "epoch": 0.6120654115707023, + "grad_norm": 1.0683684349060059, + "learning_rate": 1.9627556429464547e-05, + "loss": 0.9593, + "step": 8908 + }, + { + "epoch": 0.6121341212037927, + "grad_norm": 1.0722213983535767, + "learning_rate": 1.9621700357485182e-05, + "loss": 0.9294, + "step": 8909 + }, + { + "epoch": 0.6122028308368833, + "grad_norm": 1.0120867490768433, + "learning_rate": 1.9615844594916104e-05, + "loss": 0.9824, + "step": 8910 + }, + { + "epoch": 0.6122715404699739, + "grad_norm": 1.0639451742172241, + "learning_rate": 1.9609989142094205e-05, + "loss": 0.9312, + "step": 8911 + }, + { + "epoch": 0.6123402501030645, + "grad_norm": 1.1845481395721436, + "learning_rate": 1.9604133999356344e-05, + "loss": 0.9752, + "step": 8912 + }, + { + "epoch": 0.6124089597361551, + "grad_norm": 1.0840847492218018, + "learning_rate": 1.959827916703937e-05, + "loss": 0.9389, + "step": 8913 + }, + { + "epoch": 0.6124776693692455, + "grad_norm": 1.0349054336547852, + "learning_rate": 1.9592424645480097e-05, + "loss": 0.9184, + "step": 8914 + }, + { + "epoch": 0.6125463790023361, + "grad_norm": 1.019416093826294, + "learning_rate": 1.9586570435015347e-05, + "loss": 0.9656, + "step": 8915 + }, + { + "epoch": 0.6126150886354267, + "grad_norm": 1.0471709966659546, + "learning_rate": 1.9580716535981884e-05, + "loss": 0.9371, + "step": 8916 + }, + { + "epoch": 0.6126837982685173, + "grad_norm": 0.989752471446991, + "learning_rate": 1.9574862948716502e-05, + "loss": 0.9219, + "step": 8917 + }, + { + "epoch": 0.6127525079016078, + "grad_norm": 1.016549825668335, + "learning_rate": 1.9569009673555944e-05, + "loss": 0.964, + "step": 8918 + }, + { + "epoch": 0.6128212175346983, + "grad_norm": 1.050855040550232, + "learning_rate": 1.9563156710836943e-05, + "loss": 0.9714, + "step": 8919 + }, + { + "epoch": 0.6128899271677889, + "grad_norm": 1.0833916664123535, + "learning_rate": 1.955730406089624e-05, + "loss": 0.9649, + "step": 8920 + }, + { + "epoch": 0.6129586368008795, + "grad_norm": 1.077046513557434, + "learning_rate": 1.9551451724070502e-05, + "loss": 0.9687, + "step": 8921 + }, + { + "epoch": 0.61302734643397, + "grad_norm": 1.1039164066314697, + "learning_rate": 1.954559970069642e-05, + "loss": 0.9537, + "step": 8922 + }, + { + "epoch": 0.6130960560670606, + "grad_norm": 1.0650179386138916, + "learning_rate": 1.9539747991110667e-05, + "loss": 0.9572, + "step": 8923 + }, + { + "epoch": 0.6131647657001512, + "grad_norm": 1.1204454898834229, + "learning_rate": 1.9533896595649888e-05, + "loss": 0.9789, + "step": 8924 + }, + { + "epoch": 0.6132334753332417, + "grad_norm": 1.1048787832260132, + "learning_rate": 1.95280455146507e-05, + "loss": 0.9613, + "step": 8925 + }, + { + "epoch": 0.6133021849663323, + "grad_norm": 1.0397615432739258, + "learning_rate": 1.952219474844972e-05, + "loss": 0.9495, + "step": 8926 + }, + { + "epoch": 0.6133708945994228, + "grad_norm": 1.1404024362564087, + "learning_rate": 1.9516344297383532e-05, + "loss": 1.0149, + "step": 8927 + }, + { + "epoch": 0.6134396042325134, + "grad_norm": 1.0516622066497803, + "learning_rate": 1.9510494161788713e-05, + "loss": 0.946, + "step": 8928 + }, + { + "epoch": 0.613508313865604, + "grad_norm": 1.05170476436615, + "learning_rate": 1.9504644342001814e-05, + "loss": 0.9395, + "step": 8929 + }, + { + "epoch": 0.6135770234986945, + "grad_norm": 1.0261989831924438, + "learning_rate": 1.949879483835939e-05, + "loss": 0.9161, + "step": 8930 + }, + { + "epoch": 0.613645733131785, + "grad_norm": 0.9961035847663879, + "learning_rate": 1.9492945651197935e-05, + "loss": 0.9381, + "step": 8931 + }, + { + "epoch": 0.6137144427648756, + "grad_norm": 1.0561436414718628, + "learning_rate": 1.9487096780853952e-05, + "loss": 0.9225, + "step": 8932 + }, + { + "epoch": 0.6137831523979662, + "grad_norm": 1.2435988187789917, + "learning_rate": 1.9481248227663937e-05, + "loss": 0.9878, + "step": 8933 + }, + { + "epoch": 0.6138518620310568, + "grad_norm": 1.1060045957565308, + "learning_rate": 1.9475399991964344e-05, + "loss": 0.9656, + "step": 8934 + }, + { + "epoch": 0.6139205716641473, + "grad_norm": 1.0322651863098145, + "learning_rate": 1.9469552074091628e-05, + "loss": 0.9524, + "step": 8935 + }, + { + "epoch": 0.6139892812972378, + "grad_norm": 1.0370045900344849, + "learning_rate": 1.9463704474382204e-05, + "loss": 0.9294, + "step": 8936 + }, + { + "epoch": 0.6140579909303284, + "grad_norm": 1.0807446241378784, + "learning_rate": 1.9457857193172492e-05, + "loss": 0.9732, + "step": 8937 + }, + { + "epoch": 0.614126700563419, + "grad_norm": 1.027463436126709, + "learning_rate": 1.9452010230798867e-05, + "loss": 0.9503, + "step": 8938 + }, + { + "epoch": 0.6141954101965096, + "grad_norm": 1.1124794483184814, + "learning_rate": 1.944616358759771e-05, + "loss": 0.9716, + "step": 8939 + }, + { + "epoch": 0.6142641198296002, + "grad_norm": 1.0386343002319336, + "learning_rate": 1.9440317263905387e-05, + "loss": 0.9545, + "step": 8940 + }, + { + "epoch": 0.6143328294626906, + "grad_norm": 1.0512182712554932, + "learning_rate": 1.9434471260058213e-05, + "loss": 0.9321, + "step": 8941 + }, + { + "epoch": 0.6144015390957812, + "grad_norm": 1.1085201501846313, + "learning_rate": 1.9428625576392514e-05, + "loss": 0.9469, + "step": 8942 + }, + { + "epoch": 0.6144702487288718, + "grad_norm": 1.1288979053497314, + "learning_rate": 1.9422780213244588e-05, + "loss": 0.9605, + "step": 8943 + }, + { + "epoch": 0.6145389583619624, + "grad_norm": 1.072414755821228, + "learning_rate": 1.9416935170950713e-05, + "loss": 0.9505, + "step": 8944 + }, + { + "epoch": 0.614607667995053, + "grad_norm": 1.0575779676437378, + "learning_rate": 1.9411090449847168e-05, + "loss": 0.9163, + "step": 8945 + }, + { + "epoch": 0.6146763776281434, + "grad_norm": 1.0560709238052368, + "learning_rate": 1.940524605027017e-05, + "loss": 0.9702, + "step": 8946 + }, + { + "epoch": 0.614745087261234, + "grad_norm": 1.0343621969223022, + "learning_rate": 1.9399401972555954e-05, + "loss": 0.9418, + "step": 8947 + }, + { + "epoch": 0.6148137968943246, + "grad_norm": 1.0426849126815796, + "learning_rate": 1.9393558217040733e-05, + "loss": 0.9286, + "step": 8948 + }, + { + "epoch": 0.6148825065274152, + "grad_norm": 1.0848028659820557, + "learning_rate": 1.9387714784060685e-05, + "loss": 0.948, + "step": 8949 + }, + { + "epoch": 0.6149512161605057, + "grad_norm": 1.0470858812332153, + "learning_rate": 1.938187167395199e-05, + "loss": 0.93, + "step": 8950 + }, + { + "epoch": 0.6150199257935962, + "grad_norm": 1.0958181619644165, + "learning_rate": 1.937602888705079e-05, + "loss": 0.9719, + "step": 8951 + }, + { + "epoch": 0.6150886354266868, + "grad_norm": 1.1548666954040527, + "learning_rate": 1.9370186423693214e-05, + "loss": 1.0059, + "step": 8952 + }, + { + "epoch": 0.6151573450597774, + "grad_norm": 1.0567728281021118, + "learning_rate": 1.936434428421538e-05, + "loss": 0.9492, + "step": 8953 + }, + { + "epoch": 0.615226054692868, + "grad_norm": 1.0223729610443115, + "learning_rate": 1.9358502468953384e-05, + "loss": 0.9136, + "step": 8954 + }, + { + "epoch": 0.6152947643259585, + "grad_norm": 1.058180332183838, + "learning_rate": 1.9352660978243314e-05, + "loss": 0.9275, + "step": 8955 + }, + { + "epoch": 0.6153634739590491, + "grad_norm": 1.089864730834961, + "learning_rate": 1.9346819812421204e-05, + "loss": 0.9523, + "step": 8956 + }, + { + "epoch": 0.6154321835921396, + "grad_norm": 1.0074336528778076, + "learning_rate": 1.93409789718231e-05, + "loss": 0.9031, + "step": 8957 + }, + { + "epoch": 0.6155008932252302, + "grad_norm": 1.0101666450500488, + "learning_rate": 1.9335138456785022e-05, + "loss": 0.9774, + "step": 8958 + }, + { + "epoch": 0.6155696028583207, + "grad_norm": 1.0737794637680054, + "learning_rate": 1.9329298267642986e-05, + "loss": 0.9536, + "step": 8959 + }, + { + "epoch": 0.6156383124914113, + "grad_norm": 1.0801277160644531, + "learning_rate": 1.932345840473296e-05, + "loss": 0.9555, + "step": 8960 + }, + { + "epoch": 0.6157070221245019, + "grad_norm": 1.0129233598709106, + "learning_rate": 1.931761886839091e-05, + "loss": 0.9328, + "step": 8961 + }, + { + "epoch": 0.6157757317575924, + "grad_norm": 1.032218098640442, + "learning_rate": 1.931177965895278e-05, + "loss": 0.9098, + "step": 8962 + }, + { + "epoch": 0.615844441390683, + "grad_norm": 1.1168755292892456, + "learning_rate": 1.93059407767545e-05, + "loss": 0.9983, + "step": 8963 + }, + { + "epoch": 0.6159131510237735, + "grad_norm": 1.0799732208251953, + "learning_rate": 1.9300102222131965e-05, + "loss": 0.9717, + "step": 8964 + }, + { + "epoch": 0.6159818606568641, + "grad_norm": 1.1005586385726929, + "learning_rate": 1.9294263995421093e-05, + "loss": 0.9287, + "step": 8965 + }, + { + "epoch": 0.6160505702899547, + "grad_norm": 1.0767056941986084, + "learning_rate": 1.928842609695772e-05, + "loss": 0.9456, + "step": 8966 + }, + { + "epoch": 0.6161192799230452, + "grad_norm": 1.0756791830062866, + "learning_rate": 1.9282588527077715e-05, + "loss": 0.9625, + "step": 8967 + }, + { + "epoch": 0.6161879895561357, + "grad_norm": 1.1125558614730835, + "learning_rate": 1.9276751286116905e-05, + "loss": 0.9924, + "step": 8968 + }, + { + "epoch": 0.6162566991892263, + "grad_norm": 1.0447282791137695, + "learning_rate": 1.92709143744111e-05, + "loss": 0.934, + "step": 8969 + }, + { + "epoch": 0.6163254088223169, + "grad_norm": 1.130477786064148, + "learning_rate": 1.9265077792296106e-05, + "loss": 0.9724, + "step": 8970 + }, + { + "epoch": 0.6163941184554075, + "grad_norm": 1.017505168914795, + "learning_rate": 1.9259241540107688e-05, + "loss": 0.9086, + "step": 8971 + }, + { + "epoch": 0.6164628280884981, + "grad_norm": 1.1184678077697754, + "learning_rate": 1.9253405618181604e-05, + "loss": 0.939, + "step": 8972 + }, + { + "epoch": 0.6165315377215885, + "grad_norm": 1.1000231504440308, + "learning_rate": 1.924757002685359e-05, + "loss": 0.9267, + "step": 8973 + }, + { + "epoch": 0.6166002473546791, + "grad_norm": 1.068152666091919, + "learning_rate": 1.924173476645936e-05, + "loss": 0.9677, + "step": 8974 + }, + { + "epoch": 0.6166689569877697, + "grad_norm": 1.099643349647522, + "learning_rate": 1.923589983733462e-05, + "loss": 0.9852, + "step": 8975 + }, + { + "epoch": 0.6167376666208603, + "grad_norm": 1.0824006795883179, + "learning_rate": 1.923006523981506e-05, + "loss": 0.9772, + "step": 8976 + }, + { + "epoch": 0.6168063762539508, + "grad_norm": 1.0228116512298584, + "learning_rate": 1.9224230974236317e-05, + "loss": 0.9278, + "step": 8977 + }, + { + "epoch": 0.6168750858870413, + "grad_norm": 1.0349684953689575, + "learning_rate": 1.921839704093404e-05, + "loss": 0.9433, + "step": 8978 + }, + { + "epoch": 0.6169437955201319, + "grad_norm": 1.0614444017410278, + "learning_rate": 1.921256344024386e-05, + "loss": 0.9411, + "step": 8979 + }, + { + "epoch": 0.6170125051532225, + "grad_norm": 1.223217487335205, + "learning_rate": 1.9206730172501376e-05, + "loss": 0.9515, + "step": 8980 + }, + { + "epoch": 0.6170812147863131, + "grad_norm": 1.0178459882736206, + "learning_rate": 1.9200897238042184e-05, + "loss": 0.9528, + "step": 8981 + }, + { + "epoch": 0.6171499244194036, + "grad_norm": 1.0350674390792847, + "learning_rate": 1.9195064637201826e-05, + "loss": 0.9185, + "step": 8982 + }, + { + "epoch": 0.6172186340524941, + "grad_norm": 1.0816013813018799, + "learning_rate": 1.9189232370315862e-05, + "loss": 0.9815, + "step": 8983 + }, + { + "epoch": 0.6172873436855847, + "grad_norm": 1.0902273654937744, + "learning_rate": 1.918340043771982e-05, + "loss": 1.0206, + "step": 8984 + }, + { + "epoch": 0.6173560533186753, + "grad_norm": 1.083923101425171, + "learning_rate": 1.91775688397492e-05, + "loss": 0.9529, + "step": 8985 + }, + { + "epoch": 0.6174247629517658, + "grad_norm": 1.040716290473938, + "learning_rate": 1.9171737576739505e-05, + "loss": 0.9638, + "step": 8986 + }, + { + "epoch": 0.6174934725848564, + "grad_norm": 1.1500492095947266, + "learning_rate": 1.9165906649026184e-05, + "loss": 0.9404, + "step": 8987 + }, + { + "epoch": 0.617562182217947, + "grad_norm": 1.049118995666504, + "learning_rate": 1.916007605694469e-05, + "loss": 0.9294, + "step": 8988 + }, + { + "epoch": 0.6176308918510375, + "grad_norm": 1.1242703199386597, + "learning_rate": 1.9154245800830468e-05, + "loss": 1.012, + "step": 8989 + }, + { + "epoch": 0.6176996014841281, + "grad_norm": 1.0387428998947144, + "learning_rate": 1.9148415881018917e-05, + "loss": 0.9195, + "step": 8990 + }, + { + "epoch": 0.6177683111172186, + "grad_norm": 1.0765241384506226, + "learning_rate": 1.9142586297845445e-05, + "loss": 0.9366, + "step": 8991 + }, + { + "epoch": 0.6178370207503092, + "grad_norm": 1.0488216876983643, + "learning_rate": 1.91367570516454e-05, + "loss": 0.9367, + "step": 8992 + }, + { + "epoch": 0.6179057303833998, + "grad_norm": 1.1013802289962769, + "learning_rate": 1.9130928142754146e-05, + "loss": 0.9225, + "step": 8993 + }, + { + "epoch": 0.6179744400164903, + "grad_norm": 1.0807584524154663, + "learning_rate": 1.9125099571507016e-05, + "loss": 0.9494, + "step": 8994 + }, + { + "epoch": 0.6180431496495808, + "grad_norm": 0.9822477698326111, + "learning_rate": 1.9119271338239333e-05, + "loss": 0.9271, + "step": 8995 + }, + { + "epoch": 0.6181118592826714, + "grad_norm": 1.153709053993225, + "learning_rate": 1.9113443443286382e-05, + "loss": 1.031, + "step": 8996 + }, + { + "epoch": 0.618180568915762, + "grad_norm": 1.031466007232666, + "learning_rate": 1.9107615886983443e-05, + "loss": 0.9461, + "step": 8997 + }, + { + "epoch": 0.6182492785488526, + "grad_norm": 1.0528377294540405, + "learning_rate": 1.9101788669665767e-05, + "loss": 0.9302, + "step": 8998 + }, + { + "epoch": 0.6183179881819431, + "grad_norm": 1.0635265111923218, + "learning_rate": 1.9095961791668588e-05, + "loss": 0.9384, + "step": 8999 + }, + { + "epoch": 0.6183866978150336, + "grad_norm": 1.1276512145996094, + "learning_rate": 1.9090135253327125e-05, + "loss": 1.0036, + "step": 9000 + }, + { + "epoch": 0.6184554074481242, + "grad_norm": 1.0495657920837402, + "learning_rate": 1.9084309054976593e-05, + "loss": 0.9526, + "step": 9001 + }, + { + "epoch": 0.6185241170812148, + "grad_norm": 1.0945966243743896, + "learning_rate": 1.9078483196952142e-05, + "loss": 0.909, + "step": 9002 + }, + { + "epoch": 0.6185928267143054, + "grad_norm": 1.1098166704177856, + "learning_rate": 1.907265767958894e-05, + "loss": 0.9582, + "step": 9003 + }, + { + "epoch": 0.618661536347396, + "grad_norm": 1.1283363103866577, + "learning_rate": 1.9066832503222128e-05, + "loss": 0.9606, + "step": 9004 + }, + { + "epoch": 0.6187302459804864, + "grad_norm": 1.0797359943389893, + "learning_rate": 1.9061007668186825e-05, + "loss": 0.9666, + "step": 9005 + }, + { + "epoch": 0.618798955613577, + "grad_norm": 1.027330756187439, + "learning_rate": 1.905518317481814e-05, + "loss": 0.9723, + "step": 9006 + }, + { + "epoch": 0.6188676652466676, + "grad_norm": 1.0657832622528076, + "learning_rate": 1.9049359023451136e-05, + "loss": 0.9591, + "step": 9007 + }, + { + "epoch": 0.6189363748797582, + "grad_norm": 1.03392493724823, + "learning_rate": 1.904353521442088e-05, + "loss": 0.9395, + "step": 9008 + }, + { + "epoch": 0.6190050845128487, + "grad_norm": 1.0770541429519653, + "learning_rate": 1.9037711748062408e-05, + "loss": 0.9504, + "step": 9009 + }, + { + "epoch": 0.6190737941459392, + "grad_norm": 1.0874862670898438, + "learning_rate": 1.9031888624710743e-05, + "loss": 0.975, + "step": 9010 + }, + { + "epoch": 0.6191425037790298, + "grad_norm": 1.0372254848480225, + "learning_rate": 1.9026065844700898e-05, + "loss": 0.9318, + "step": 9011 + }, + { + "epoch": 0.6192112134121204, + "grad_norm": 1.1310477256774902, + "learning_rate": 1.9020243408367827e-05, + "loss": 0.9414, + "step": 9012 + }, + { + "epoch": 0.619279923045211, + "grad_norm": 1.110448956489563, + "learning_rate": 1.901442131604651e-05, + "loss": 0.9913, + "step": 9013 + }, + { + "epoch": 0.6193486326783015, + "grad_norm": 1.1016700267791748, + "learning_rate": 1.9008599568071886e-05, + "loss": 0.9223, + "step": 9014 + }, + { + "epoch": 0.619417342311392, + "grad_norm": 1.0431089401245117, + "learning_rate": 1.9002778164778876e-05, + "loss": 0.9369, + "step": 9015 + }, + { + "epoch": 0.6194860519444826, + "grad_norm": 1.0748600959777832, + "learning_rate": 1.8996957106502393e-05, + "loss": 0.9691, + "step": 9016 + }, + { + "epoch": 0.6195547615775732, + "grad_norm": 1.0799819231033325, + "learning_rate": 1.899113639357729e-05, + "loss": 0.9463, + "step": 9017 + }, + { + "epoch": 0.6196234712106637, + "grad_norm": 1.0681363344192505, + "learning_rate": 1.8985316026338455e-05, + "loss": 0.9001, + "step": 9018 + }, + { + "epoch": 0.6196921808437543, + "grad_norm": 1.0155775547027588, + "learning_rate": 1.8979496005120723e-05, + "loss": 0.9014, + "step": 9019 + }, + { + "epoch": 0.6197608904768449, + "grad_norm": 1.0422254800796509, + "learning_rate": 1.897367633025891e-05, + "loss": 0.969, + "step": 9020 + }, + { + "epoch": 0.6198296001099354, + "grad_norm": 1.0791019201278687, + "learning_rate": 1.896785700208783e-05, + "loss": 0.9233, + "step": 9021 + }, + { + "epoch": 0.619898309743026, + "grad_norm": 1.0454026460647583, + "learning_rate": 1.8962038020942256e-05, + "loss": 0.9421, + "step": 9022 + }, + { + "epoch": 0.6199670193761165, + "grad_norm": 1.0365991592407227, + "learning_rate": 1.895621938715695e-05, + "loss": 0.9535, + "step": 9023 + }, + { + "epoch": 0.6200357290092071, + "grad_norm": 1.0211907625198364, + "learning_rate": 1.8950401101066653e-05, + "loss": 0.9028, + "step": 9024 + }, + { + "epoch": 0.6201044386422977, + "grad_norm": 1.0260953903198242, + "learning_rate": 1.8944583163006095e-05, + "loss": 0.9227, + "step": 9025 + }, + { + "epoch": 0.6201731482753882, + "grad_norm": 1.023830533027649, + "learning_rate": 1.8938765573309978e-05, + "loss": 0.9541, + "step": 9026 + }, + { + "epoch": 0.6202418579084787, + "grad_norm": 1.0328949689865112, + "learning_rate": 1.8932948332312988e-05, + "loss": 0.9197, + "step": 9027 + }, + { + "epoch": 0.6203105675415693, + "grad_norm": 1.1093244552612305, + "learning_rate": 1.8927131440349775e-05, + "loss": 0.971, + "step": 9028 + }, + { + "epoch": 0.6203792771746599, + "grad_norm": 1.0376067161560059, + "learning_rate": 1.8921314897754984e-05, + "loss": 0.9318, + "step": 9029 + }, + { + "epoch": 0.6204479868077505, + "grad_norm": 1.0011990070343018, + "learning_rate": 1.8915498704863248e-05, + "loss": 0.9147, + "step": 9030 + }, + { + "epoch": 0.620516696440841, + "grad_norm": 1.0434844493865967, + "learning_rate": 1.8909682862009158e-05, + "loss": 0.9307, + "step": 9031 + }, + { + "epoch": 0.6205854060739315, + "grad_norm": 1.0609160661697388, + "learning_rate": 1.8903867369527302e-05, + "loss": 0.9091, + "step": 9032 + }, + { + "epoch": 0.6206541157070221, + "grad_norm": 1.0302327871322632, + "learning_rate": 1.8898052227752238e-05, + "loss": 0.9691, + "step": 9033 + }, + { + "epoch": 0.6207228253401127, + "grad_norm": 1.148792028427124, + "learning_rate": 1.8892237437018507e-05, + "loss": 0.9486, + "step": 9034 + }, + { + "epoch": 0.6207915349732033, + "grad_norm": 1.0662809610366821, + "learning_rate": 1.8886422997660626e-05, + "loss": 0.9342, + "step": 9035 + }, + { + "epoch": 0.6208602446062939, + "grad_norm": 1.0796377658843994, + "learning_rate": 1.8880608910013108e-05, + "loss": 0.9656, + "step": 9036 + }, + { + "epoch": 0.6209289542393843, + "grad_norm": 1.0295075178146362, + "learning_rate": 1.8874795174410437e-05, + "loss": 0.9427, + "step": 9037 + }, + { + "epoch": 0.6209976638724749, + "grad_norm": 0.9811196327209473, + "learning_rate": 1.886898179118706e-05, + "loss": 0.9676, + "step": 9038 + }, + { + "epoch": 0.6210663735055655, + "grad_norm": 1.0778324604034424, + "learning_rate": 1.8863168760677415e-05, + "loss": 0.9459, + "step": 9039 + }, + { + "epoch": 0.6211350831386561, + "grad_norm": 1.0842944383621216, + "learning_rate": 1.8857356083215928e-05, + "loss": 0.9563, + "step": 9040 + }, + { + "epoch": 0.6212037927717466, + "grad_norm": 1.063260793685913, + "learning_rate": 1.8851543759137007e-05, + "loss": 0.9812, + "step": 9041 + }, + { + "epoch": 0.6212725024048371, + "grad_norm": 1.0038669109344482, + "learning_rate": 1.8845731788775027e-05, + "loss": 0.8947, + "step": 9042 + }, + { + "epoch": 0.6213412120379277, + "grad_norm": 1.0516093969345093, + "learning_rate": 1.883992017246434e-05, + "loss": 0.9424, + "step": 9043 + }, + { + "epoch": 0.6214099216710183, + "grad_norm": 1.1503475904464722, + "learning_rate": 1.8834108910539295e-05, + "loss": 0.9562, + "step": 9044 + }, + { + "epoch": 0.6214786313041089, + "grad_norm": 1.0527318716049194, + "learning_rate": 1.8828298003334198e-05, + "loss": 0.9452, + "step": 9045 + }, + { + "epoch": 0.6215473409371994, + "grad_norm": 1.0730291604995728, + "learning_rate": 1.882248745118335e-05, + "loss": 0.9779, + "step": 9046 + }, + { + "epoch": 0.6216160505702899, + "grad_norm": 1.0739465951919556, + "learning_rate": 1.881667725442105e-05, + "loss": 0.9413, + "step": 9047 + }, + { + "epoch": 0.6216847602033805, + "grad_norm": 1.1163065433502197, + "learning_rate": 1.881086741338152e-05, + "loss": 0.9481, + "step": 9048 + }, + { + "epoch": 0.6217534698364711, + "grad_norm": 1.120238184928894, + "learning_rate": 1.880505792839902e-05, + "loss": 0.9892, + "step": 9049 + }, + { + "epoch": 0.6218221794695616, + "grad_norm": 1.0183731317520142, + "learning_rate": 1.8799248799807756e-05, + "loss": 0.924, + "step": 9050 + }, + { + "epoch": 0.6218908891026522, + "grad_norm": 1.0498465299606323, + "learning_rate": 1.879344002794193e-05, + "loss": 0.9193, + "step": 9051 + }, + { + "epoch": 0.6219595987357428, + "grad_norm": 1.1405115127563477, + "learning_rate": 1.8787631613135727e-05, + "loss": 0.9413, + "step": 9052 + }, + { + "epoch": 0.6220283083688333, + "grad_norm": 1.068801999092102, + "learning_rate": 1.8781823555723273e-05, + "loss": 0.9236, + "step": 9053 + }, + { + "epoch": 0.6220970180019239, + "grad_norm": 1.1396245956420898, + "learning_rate": 1.8776015856038722e-05, + "loss": 0.9545, + "step": 9054 + }, + { + "epoch": 0.6221657276350144, + "grad_norm": 1.0251078605651855, + "learning_rate": 1.877020851441619e-05, + "loss": 0.9532, + "step": 9055 + }, + { + "epoch": 0.622234437268105, + "grad_norm": 1.0407533645629883, + "learning_rate": 1.8764401531189756e-05, + "loss": 0.9554, + "step": 9056 + }, + { + "epoch": 0.6223031469011956, + "grad_norm": 1.002779483795166, + "learning_rate": 1.8758594906693507e-05, + "loss": 0.9652, + "step": 9057 + }, + { + "epoch": 0.6223718565342861, + "grad_norm": 1.019196629524231, + "learning_rate": 1.8752788641261487e-05, + "loss": 0.9429, + "step": 9058 + }, + { + "epoch": 0.6224405661673766, + "grad_norm": 1.0521376132965088, + "learning_rate": 1.8746982735227725e-05, + "loss": 0.9642, + "step": 9059 + }, + { + "epoch": 0.6225092758004672, + "grad_norm": 1.0409250259399414, + "learning_rate": 1.874117718892623e-05, + "loss": 0.9648, + "step": 9060 + }, + { + "epoch": 0.6225779854335578, + "grad_norm": 1.0817558765411377, + "learning_rate": 1.8735372002690997e-05, + "loss": 0.954, + "step": 9061 + }, + { + "epoch": 0.6226466950666484, + "grad_norm": 1.1342278718948364, + "learning_rate": 1.8729567176856e-05, + "loss": 0.9749, + "step": 9062 + }, + { + "epoch": 0.6227154046997389, + "grad_norm": 1.0211663246154785, + "learning_rate": 1.8723762711755173e-05, + "loss": 0.9731, + "step": 9063 + }, + { + "epoch": 0.6227841143328294, + "grad_norm": 1.075421929359436, + "learning_rate": 1.8717958607722448e-05, + "loss": 0.9572, + "step": 9064 + }, + { + "epoch": 0.62285282396592, + "grad_norm": 1.071903109550476, + "learning_rate": 1.8712154865091735e-05, + "loss": 0.9194, + "step": 9065 + }, + { + "epoch": 0.6229215335990106, + "grad_norm": 1.1329820156097412, + "learning_rate": 1.8706351484196925e-05, + "loss": 0.9612, + "step": 9066 + }, + { + "epoch": 0.6229902432321012, + "grad_norm": 1.0739246606826782, + "learning_rate": 1.8700548465371874e-05, + "loss": 0.9677, + "step": 9067 + }, + { + "epoch": 0.6230589528651918, + "grad_norm": 1.0790859460830688, + "learning_rate": 1.8694745808950426e-05, + "loss": 0.9751, + "step": 9068 + }, + { + "epoch": 0.6231276624982822, + "grad_norm": 1.0480377674102783, + "learning_rate": 1.868894351526641e-05, + "loss": 0.9251, + "step": 9069 + }, + { + "epoch": 0.6231963721313728, + "grad_norm": 1.1012450456619263, + "learning_rate": 1.8683141584653626e-05, + "loss": 0.9529, + "step": 9070 + }, + { + "epoch": 0.6232650817644634, + "grad_norm": 1.0805083513259888, + "learning_rate": 1.867734001744585e-05, + "loss": 0.9746, + "step": 9071 + }, + { + "epoch": 0.623333791397554, + "grad_norm": 1.0526015758514404, + "learning_rate": 1.867153881397686e-05, + "loss": 0.9753, + "step": 9072 + }, + { + "epoch": 0.6234025010306445, + "grad_norm": 0.9975852370262146, + "learning_rate": 1.866573797458037e-05, + "loss": 0.9534, + "step": 9073 + }, + { + "epoch": 0.623471210663735, + "grad_norm": 1.0814950466156006, + "learning_rate": 1.8659937499590114e-05, + "loss": 0.9502, + "step": 9074 + }, + { + "epoch": 0.6235399202968256, + "grad_norm": 1.0051078796386719, + "learning_rate": 1.8654137389339788e-05, + "loss": 0.8792, + "step": 9075 + }, + { + "epoch": 0.6236086299299162, + "grad_norm": 1.0649853944778442, + "learning_rate": 1.8648337644163068e-05, + "loss": 0.9508, + "step": 9076 + }, + { + "epoch": 0.6236773395630067, + "grad_norm": 1.1427644491195679, + "learning_rate": 1.8642538264393624e-05, + "loss": 0.9642, + "step": 9077 + }, + { + "epoch": 0.6237460491960973, + "grad_norm": 1.0590921640396118, + "learning_rate": 1.863673925036506e-05, + "loss": 0.9647, + "step": 9078 + }, + { + "epoch": 0.6238147588291878, + "grad_norm": 1.047491192817688, + "learning_rate": 1.8630940602411012e-05, + "loss": 0.9297, + "step": 9079 + }, + { + "epoch": 0.6238834684622784, + "grad_norm": 1.050012230873108, + "learning_rate": 1.8625142320865073e-05, + "loss": 0.9746, + "step": 9080 + }, + { + "epoch": 0.623952178095369, + "grad_norm": 1.0480506420135498, + "learning_rate": 1.8619344406060805e-05, + "loss": 0.9306, + "step": 9081 + }, + { + "epoch": 0.6240208877284595, + "grad_norm": 1.065933346748352, + "learning_rate": 1.861354685833176e-05, + "loss": 0.969, + "step": 9082 + }, + { + "epoch": 0.6240895973615501, + "grad_norm": 1.027335524559021, + "learning_rate": 1.8607749678011482e-05, + "loss": 0.9396, + "step": 9083 + }, + { + "epoch": 0.6241583069946407, + "grad_norm": 1.0398236513137817, + "learning_rate": 1.860195286543346e-05, + "loss": 0.9318, + "step": 9084 + }, + { + "epoch": 0.6242270166277312, + "grad_norm": 1.0970174074172974, + "learning_rate": 1.8596156420931183e-05, + "loss": 0.9418, + "step": 9085 + }, + { + "epoch": 0.6242957262608217, + "grad_norm": 1.0414962768554688, + "learning_rate": 1.8590360344838126e-05, + "loss": 0.9203, + "step": 9086 + }, + { + "epoch": 0.6243644358939123, + "grad_norm": 1.0987855195999146, + "learning_rate": 1.858456463748773e-05, + "loss": 0.9764, + "step": 9087 + }, + { + "epoch": 0.6244331455270029, + "grad_norm": 1.1104713678359985, + "learning_rate": 1.8578769299213427e-05, + "loss": 0.9841, + "step": 9088 + }, + { + "epoch": 0.6245018551600935, + "grad_norm": 0.9939348101615906, + "learning_rate": 1.85729743303486e-05, + "loss": 0.9488, + "step": 9089 + }, + { + "epoch": 0.624570564793184, + "grad_norm": 1.0954521894454956, + "learning_rate": 1.856717973122664e-05, + "loss": 0.9871, + "step": 9090 + }, + { + "epoch": 0.6246392744262745, + "grad_norm": 1.023397445678711, + "learning_rate": 1.856138550218092e-05, + "loss": 0.9174, + "step": 9091 + }, + { + "epoch": 0.6247079840593651, + "grad_norm": 1.1052006483078003, + "learning_rate": 1.8555591643544755e-05, + "loss": 0.9754, + "step": 9092 + }, + { + "epoch": 0.6247766936924557, + "grad_norm": 1.0384336709976196, + "learning_rate": 1.8549798155651487e-05, + "loss": 0.9438, + "step": 9093 + }, + { + "epoch": 0.6248454033255463, + "grad_norm": 1.0523169040679932, + "learning_rate": 1.8544005038834385e-05, + "loss": 0.9176, + "step": 9094 + }, + { + "epoch": 0.6249141129586367, + "grad_norm": 1.047298789024353, + "learning_rate": 1.853821229342674e-05, + "loss": 0.9678, + "step": 9095 + }, + { + "epoch": 0.6249828225917273, + "grad_norm": 1.0447338819503784, + "learning_rate": 1.8532419919761803e-05, + "loss": 0.9274, + "step": 9096 + }, + { + "epoch": 0.6250515322248179, + "grad_norm": 1.0172284841537476, + "learning_rate": 1.85266279181728e-05, + "loss": 0.9364, + "step": 9097 + }, + { + "epoch": 0.6251202418579085, + "grad_norm": 1.0587470531463623, + "learning_rate": 1.8520836288992953e-05, + "loss": 0.9477, + "step": 9098 + }, + { + "epoch": 0.6251889514909991, + "grad_norm": 1.074280023574829, + "learning_rate": 1.8515045032555438e-05, + "loss": 0.9512, + "step": 9099 + }, + { + "epoch": 0.6252576611240896, + "grad_norm": 1.0918114185333252, + "learning_rate": 1.8509254149193432e-05, + "loss": 0.9889, + "step": 9100 + }, + { + "epoch": 0.6253263707571801, + "grad_norm": 1.039754867553711, + "learning_rate": 1.850346363924007e-05, + "loss": 0.9614, + "step": 9101 + }, + { + "epoch": 0.6253950803902707, + "grad_norm": 1.0107409954071045, + "learning_rate": 1.8497673503028497e-05, + "loss": 0.9394, + "step": 9102 + }, + { + "epoch": 0.6254637900233613, + "grad_norm": 1.0671528577804565, + "learning_rate": 1.8491883740891796e-05, + "loss": 0.964, + "step": 9103 + }, + { + "epoch": 0.6255324996564519, + "grad_norm": 1.1264888048171997, + "learning_rate": 1.8486094353163058e-05, + "loss": 0.9467, + "step": 9104 + }, + { + "epoch": 0.6256012092895424, + "grad_norm": 1.0313621759414673, + "learning_rate": 1.8480305340175333e-05, + "loss": 0.9292, + "step": 9105 + }, + { + "epoch": 0.6256699189226329, + "grad_norm": 1.0486934185028076, + "learning_rate": 1.847451670226167e-05, + "loss": 0.9696, + "step": 9106 + }, + { + "epoch": 0.6257386285557235, + "grad_norm": 1.095382571220398, + "learning_rate": 1.8468728439755076e-05, + "loss": 0.9591, + "step": 9107 + }, + { + "epoch": 0.6258073381888141, + "grad_norm": 1.0282671451568604, + "learning_rate": 1.846294055298857e-05, + "loss": 0.9462, + "step": 9108 + }, + { + "epoch": 0.6258760478219046, + "grad_norm": 1.1355316638946533, + "learning_rate": 1.845715304229509e-05, + "loss": 0.9151, + "step": 9109 + }, + { + "epoch": 0.6259447574549952, + "grad_norm": 1.0899393558502197, + "learning_rate": 1.8451365908007612e-05, + "loss": 0.9801, + "step": 9110 + }, + { + "epoch": 0.6260134670880857, + "grad_norm": 1.0332672595977783, + "learning_rate": 1.844557915045906e-05, + "loss": 0.9296, + "step": 9111 + }, + { + "epoch": 0.6260821767211763, + "grad_norm": 1.1107244491577148, + "learning_rate": 1.843979276998234e-05, + "loss": 0.956, + "step": 9112 + }, + { + "epoch": 0.6261508863542669, + "grad_norm": 1.060687780380249, + "learning_rate": 1.8434006766910356e-05, + "loss": 0.9705, + "step": 9113 + }, + { + "epoch": 0.6262195959873574, + "grad_norm": 1.1289398670196533, + "learning_rate": 1.8428221141575948e-05, + "loss": 0.9385, + "step": 9114 + }, + { + "epoch": 0.626288305620448, + "grad_norm": 1.1018658876419067, + "learning_rate": 1.8422435894311972e-05, + "loss": 0.9433, + "step": 9115 + }, + { + "epoch": 0.6263570152535386, + "grad_norm": 1.1150054931640625, + "learning_rate": 1.8416651025451248e-05, + "loss": 0.9951, + "step": 9116 + }, + { + "epoch": 0.6264257248866291, + "grad_norm": 1.0912315845489502, + "learning_rate": 1.8410866535326575e-05, + "loss": 0.9668, + "step": 9117 + }, + { + "epoch": 0.6264944345197196, + "grad_norm": 1.0953936576843262, + "learning_rate": 1.840508242427074e-05, + "loss": 0.9406, + "step": 9118 + }, + { + "epoch": 0.6265631441528102, + "grad_norm": 1.0806362628936768, + "learning_rate": 1.8399298692616486e-05, + "loss": 0.9818, + "step": 9119 + }, + { + "epoch": 0.6266318537859008, + "grad_norm": 1.0411418676376343, + "learning_rate": 1.8393515340696554e-05, + "loss": 0.934, + "step": 9120 + }, + { + "epoch": 0.6267005634189914, + "grad_norm": 1.0685508251190186, + "learning_rate": 1.8387732368843652e-05, + "loss": 0.9708, + "step": 9121 + }, + { + "epoch": 0.6267692730520819, + "grad_norm": 1.0789681673049927, + "learning_rate": 1.838194977739048e-05, + "loss": 0.9057, + "step": 9122 + }, + { + "epoch": 0.6268379826851724, + "grad_norm": 1.1633884906768799, + "learning_rate": 1.837616756666971e-05, + "loss": 0.9745, + "step": 9123 + }, + { + "epoch": 0.626906692318263, + "grad_norm": 1.163059949874878, + "learning_rate": 1.8370385737013967e-05, + "loss": 0.943, + "step": 9124 + }, + { + "epoch": 0.6269754019513536, + "grad_norm": 1.0980881452560425, + "learning_rate": 1.8364604288755895e-05, + "loss": 0.9794, + "step": 9125 + }, + { + "epoch": 0.6270441115844442, + "grad_norm": 1.058536171913147, + "learning_rate": 1.8358823222228097e-05, + "loss": 0.9319, + "step": 9126 + }, + { + "epoch": 0.6271128212175346, + "grad_norm": 1.0397175550460815, + "learning_rate": 1.8353042537763143e-05, + "loss": 0.96, + "step": 9127 + }, + { + "epoch": 0.6271815308506252, + "grad_norm": 1.0388480424880981, + "learning_rate": 1.8347262235693605e-05, + "loss": 0.9094, + "step": 9128 + }, + { + "epoch": 0.6272502404837158, + "grad_norm": 1.0292664766311646, + "learning_rate": 1.8341482316352014e-05, + "loss": 0.939, + "step": 9129 + }, + { + "epoch": 0.6273189501168064, + "grad_norm": 1.100785493850708, + "learning_rate": 1.833570278007088e-05, + "loss": 0.9827, + "step": 9130 + }, + { + "epoch": 0.627387659749897, + "grad_norm": 1.0818531513214111, + "learning_rate": 1.83299236271827e-05, + "loss": 0.9601, + "step": 9131 + }, + { + "epoch": 0.6274563693829875, + "grad_norm": 1.0359406471252441, + "learning_rate": 1.832414485801995e-05, + "loss": 0.9219, + "step": 9132 + }, + { + "epoch": 0.627525079016078, + "grad_norm": 1.0754178762435913, + "learning_rate": 1.8318366472915087e-05, + "loss": 0.9691, + "step": 9133 + }, + { + "epoch": 0.6275937886491686, + "grad_norm": 1.0224851369857788, + "learning_rate": 1.8312588472200514e-05, + "loss": 1.0012, + "step": 9134 + }, + { + "epoch": 0.6276624982822592, + "grad_norm": 1.0003687143325806, + "learning_rate": 1.830681085620865e-05, + "loss": 0.9254, + "step": 9135 + }, + { + "epoch": 0.6277312079153498, + "grad_norm": 1.0227795839309692, + "learning_rate": 1.8301033625271875e-05, + "loss": 0.9248, + "step": 9136 + }, + { + "epoch": 0.6277999175484403, + "grad_norm": 1.1365183591842651, + "learning_rate": 1.8295256779722556e-05, + "loss": 1.0002, + "step": 9137 + }, + { + "epoch": 0.6278686271815308, + "grad_norm": 1.116054892539978, + "learning_rate": 1.828948031989302e-05, + "loss": 0.963, + "step": 9138 + }, + { + "epoch": 0.6279373368146214, + "grad_norm": 1.150278925895691, + "learning_rate": 1.8283704246115597e-05, + "loss": 0.987, + "step": 9139 + }, + { + "epoch": 0.628006046447712, + "grad_norm": 1.0721158981323242, + "learning_rate": 1.8277928558722573e-05, + "loss": 0.9632, + "step": 9140 + }, + { + "epoch": 0.6280747560808025, + "grad_norm": 1.1102919578552246, + "learning_rate": 1.8272153258046215e-05, + "loss": 0.9206, + "step": 9141 + }, + { + "epoch": 0.6281434657138931, + "grad_norm": 1.0681004524230957, + "learning_rate": 1.8266378344418777e-05, + "loss": 0.9547, + "step": 9142 + }, + { + "epoch": 0.6282121753469836, + "grad_norm": 1.0106093883514404, + "learning_rate": 1.826060381817249e-05, + "loss": 0.9208, + "step": 9143 + }, + { + "epoch": 0.6282808849800742, + "grad_norm": 1.0209025144577026, + "learning_rate": 1.8254829679639565e-05, + "loss": 0.9522, + "step": 9144 + }, + { + "epoch": 0.6283495946131648, + "grad_norm": 1.0883195400238037, + "learning_rate": 1.8249055929152166e-05, + "loss": 0.9636, + "step": 9145 + }, + { + "epoch": 0.6284183042462553, + "grad_norm": 1.062874674797058, + "learning_rate": 1.8243282567042462e-05, + "loss": 0.9639, + "step": 9146 + }, + { + "epoch": 0.6284870138793459, + "grad_norm": 1.1217015981674194, + "learning_rate": 1.8237509593642595e-05, + "loss": 0.9391, + "step": 9147 + }, + { + "epoch": 0.6285557235124365, + "grad_norm": 1.0708799362182617, + "learning_rate": 1.823173700928468e-05, + "loss": 0.9422, + "step": 9148 + }, + { + "epoch": 0.628624433145527, + "grad_norm": 1.0440629720687866, + "learning_rate": 1.8225964814300805e-05, + "loss": 0.8899, + "step": 9149 + }, + { + "epoch": 0.6286931427786175, + "grad_norm": 1.123105764389038, + "learning_rate": 1.822019300902304e-05, + "loss": 0.9257, + "step": 9150 + }, + { + "epoch": 0.6287618524117081, + "grad_norm": 1.0420767068862915, + "learning_rate": 1.821442159378345e-05, + "loss": 0.9584, + "step": 9151 + }, + { + "epoch": 0.6288305620447987, + "grad_norm": 1.098388910293579, + "learning_rate": 1.8208650568914033e-05, + "loss": 0.9898, + "step": 9152 + }, + { + "epoch": 0.6288992716778893, + "grad_norm": 1.083321213722229, + "learning_rate": 1.8202879934746815e-05, + "loss": 0.9874, + "step": 9153 + }, + { + "epoch": 0.6289679813109798, + "grad_norm": 1.1128212213516235, + "learning_rate": 1.819710969161378e-05, + "loss": 0.9663, + "step": 9154 + }, + { + "epoch": 0.6290366909440703, + "grad_norm": 1.0307596921920776, + "learning_rate": 1.8191339839846862e-05, + "loss": 0.9543, + "step": 9155 + }, + { + "epoch": 0.6291054005771609, + "grad_norm": 1.087129831314087, + "learning_rate": 1.8185570379778012e-05, + "loss": 0.952, + "step": 9156 + }, + { + "epoch": 0.6291741102102515, + "grad_norm": 1.0820673704147339, + "learning_rate": 1.8179801311739144e-05, + "loss": 0.9534, + "step": 9157 + }, + { + "epoch": 0.6292428198433421, + "grad_norm": 1.0516927242279053, + "learning_rate": 1.817403263606215e-05, + "loss": 0.9393, + "step": 9158 + }, + { + "epoch": 0.6293115294764325, + "grad_norm": 1.0788766145706177, + "learning_rate": 1.8168264353078903e-05, + "loss": 0.9272, + "step": 9159 + }, + { + "epoch": 0.6293802391095231, + "grad_norm": 1.1118569374084473, + "learning_rate": 1.8162496463121237e-05, + "loss": 0.9631, + "step": 9160 + }, + { + "epoch": 0.6294489487426137, + "grad_norm": 1.0135287046432495, + "learning_rate": 1.8156728966520976e-05, + "loss": 0.922, + "step": 9161 + }, + { + "epoch": 0.6295176583757043, + "grad_norm": 1.0523607730865479, + "learning_rate": 1.8150961863609934e-05, + "loss": 0.953, + "step": 9162 + }, + { + "epoch": 0.6295863680087949, + "grad_norm": 1.0412825345993042, + "learning_rate": 1.8145195154719868e-05, + "loss": 0.9637, + "step": 9163 + }, + { + "epoch": 0.6296550776418854, + "grad_norm": 1.0307477712631226, + "learning_rate": 1.8139428840182563e-05, + "loss": 0.9312, + "step": 9164 + }, + { + "epoch": 0.6297237872749759, + "grad_norm": 1.0610665082931519, + "learning_rate": 1.813366292032972e-05, + "loss": 0.9862, + "step": 9165 + }, + { + "epoch": 0.6297924969080665, + "grad_norm": 1.0349655151367188, + "learning_rate": 1.812789739549306e-05, + "loss": 0.9623, + "step": 9166 + }, + { + "epoch": 0.6298612065411571, + "grad_norm": 1.027820110321045, + "learning_rate": 1.8122132266004278e-05, + "loss": 0.9661, + "step": 9167 + }, + { + "epoch": 0.6299299161742477, + "grad_norm": 1.0045640468597412, + "learning_rate": 1.8116367532195034e-05, + "loss": 0.9565, + "step": 9168 + }, + { + "epoch": 0.6299986258073382, + "grad_norm": 1.0554882287979126, + "learning_rate": 1.8110603194396976e-05, + "loss": 0.9561, + "step": 9169 + }, + { + "epoch": 0.6300673354404287, + "grad_norm": 1.0545332431793213, + "learning_rate": 1.8104839252941704e-05, + "loss": 0.9392, + "step": 9170 + }, + { + "epoch": 0.6301360450735193, + "grad_norm": 1.0933996438980103, + "learning_rate": 1.8099075708160824e-05, + "loss": 0.9948, + "step": 9171 + }, + { + "epoch": 0.6302047547066099, + "grad_norm": 1.0471994876861572, + "learning_rate": 1.8093312560385913e-05, + "loss": 0.9395, + "step": 9172 + }, + { + "epoch": 0.6302734643397004, + "grad_norm": 1.0770156383514404, + "learning_rate": 1.8087549809948527e-05, + "loss": 0.9519, + "step": 9173 + }, + { + "epoch": 0.630342173972791, + "grad_norm": 1.0808600187301636, + "learning_rate": 1.8081787457180183e-05, + "loss": 0.9137, + "step": 9174 + }, + { + "epoch": 0.6304108836058815, + "grad_norm": 1.1284575462341309, + "learning_rate": 1.8076025502412385e-05, + "loss": 0.9264, + "step": 9175 + }, + { + "epoch": 0.6304795932389721, + "grad_norm": 1.0494335889816284, + "learning_rate": 1.807026394597662e-05, + "loss": 0.9791, + "step": 9176 + }, + { + "epoch": 0.6305483028720626, + "grad_norm": 1.1133614778518677, + "learning_rate": 1.806450278820434e-05, + "loss": 0.9547, + "step": 9177 + }, + { + "epoch": 0.6306170125051532, + "grad_norm": 1.0526719093322754, + "learning_rate": 1.805874202942699e-05, + "loss": 0.9564, + "step": 9178 + }, + { + "epoch": 0.6306857221382438, + "grad_norm": 1.0895261764526367, + "learning_rate": 1.8052981669975983e-05, + "loss": 0.9478, + "step": 9179 + }, + { + "epoch": 0.6307544317713344, + "grad_norm": 1.101974368095398, + "learning_rate": 1.80472217101827e-05, + "loss": 0.9504, + "step": 9180 + }, + { + "epoch": 0.6308231414044249, + "grad_norm": 1.0989142656326294, + "learning_rate": 1.804146215037851e-05, + "loss": 0.9235, + "step": 9181 + }, + { + "epoch": 0.6308918510375154, + "grad_norm": 1.0972615480422974, + "learning_rate": 1.803570299089476e-05, + "loss": 0.9649, + "step": 9182 + }, + { + "epoch": 0.630960560670606, + "grad_norm": 1.0630693435668945, + "learning_rate": 1.8029944232062767e-05, + "loss": 0.9426, + "step": 9183 + }, + { + "epoch": 0.6310292703036966, + "grad_norm": 1.1107209920883179, + "learning_rate": 1.802418587421385e-05, + "loss": 0.9778, + "step": 9184 + }, + { + "epoch": 0.6310979799367872, + "grad_norm": 1.0215140581130981, + "learning_rate": 1.801842791767925e-05, + "loss": 0.9775, + "step": 9185 + }, + { + "epoch": 0.6311666895698776, + "grad_norm": 1.1157827377319336, + "learning_rate": 1.8012670362790245e-05, + "loss": 0.9288, + "step": 9186 + }, + { + "epoch": 0.6312353992029682, + "grad_norm": 1.0364948511123657, + "learning_rate": 1.8006913209878045e-05, + "loss": 0.9568, + "step": 9187 + }, + { + "epoch": 0.6313041088360588, + "grad_norm": 1.0815441608428955, + "learning_rate": 1.8001156459273867e-05, + "loss": 0.9707, + "step": 9188 + }, + { + "epoch": 0.6313728184691494, + "grad_norm": 1.0289883613586426, + "learning_rate": 1.7995400111308885e-05, + "loss": 0.9495, + "step": 9189 + }, + { + "epoch": 0.63144152810224, + "grad_norm": 1.051255226135254, + "learning_rate": 1.798964416631428e-05, + "loss": 0.9464, + "step": 9190 + }, + { + "epoch": 0.6315102377353305, + "grad_norm": 1.0622718334197998, + "learning_rate": 1.798388862462116e-05, + "loss": 0.9825, + "step": 9191 + }, + { + "epoch": 0.631578947368421, + "grad_norm": 1.0324333906173706, + "learning_rate": 1.7978133486560645e-05, + "loss": 0.9453, + "step": 9192 + }, + { + "epoch": 0.6316476570015116, + "grad_norm": 1.0632507801055908, + "learning_rate": 1.7972378752463834e-05, + "loss": 0.9509, + "step": 9193 + }, + { + "epoch": 0.6317163666346022, + "grad_norm": 1.0150349140167236, + "learning_rate": 1.7966624422661788e-05, + "loss": 0.9554, + "step": 9194 + }, + { + "epoch": 0.6317850762676928, + "grad_norm": 1.0778230428695679, + "learning_rate": 1.7960870497485558e-05, + "loss": 0.9291, + "step": 9195 + }, + { + "epoch": 0.6318537859007833, + "grad_norm": 1.031546711921692, + "learning_rate": 1.7955116977266144e-05, + "loss": 0.953, + "step": 9196 + }, + { + "epoch": 0.6319224955338738, + "grad_norm": 1.0647623538970947, + "learning_rate": 1.7949363862334567e-05, + "loss": 0.9936, + "step": 9197 + }, + { + "epoch": 0.6319912051669644, + "grad_norm": 1.0607178211212158, + "learning_rate": 1.7943611153021776e-05, + "loss": 0.9874, + "step": 9198 + }, + { + "epoch": 0.632059914800055, + "grad_norm": 1.1125621795654297, + "learning_rate": 1.793785884965874e-05, + "loss": 0.9703, + "step": 9199 + }, + { + "epoch": 0.6321286244331455, + "grad_norm": 1.0832568407058716, + "learning_rate": 1.7932106952576384e-05, + "loss": 0.9629, + "step": 9200 + }, + { + "epoch": 0.6321973340662361, + "grad_norm": 1.1405233144760132, + "learning_rate": 1.7926355462105596e-05, + "loss": 0.9401, + "step": 9201 + }, + { + "epoch": 0.6322660436993266, + "grad_norm": 1.075221061706543, + "learning_rate": 1.7920604378577266e-05, + "loss": 0.9447, + "step": 9202 + }, + { + "epoch": 0.6323347533324172, + "grad_norm": 1.089547872543335, + "learning_rate": 1.791485370232225e-05, + "loss": 0.9502, + "step": 9203 + }, + { + "epoch": 0.6324034629655078, + "grad_norm": 1.0736066102981567, + "learning_rate": 1.7909103433671375e-05, + "loss": 0.9607, + "step": 9204 + }, + { + "epoch": 0.6324721725985983, + "grad_norm": 0.976852297782898, + "learning_rate": 1.790335357295547e-05, + "loss": 0.9057, + "step": 9205 + }, + { + "epoch": 0.6325408822316889, + "grad_norm": 1.048709750175476, + "learning_rate": 1.78976041205053e-05, + "loss": 0.9455, + "step": 9206 + }, + { + "epoch": 0.6326095918647795, + "grad_norm": 1.0408943891525269, + "learning_rate": 1.789185507665163e-05, + "loss": 0.9491, + "step": 9207 + }, + { + "epoch": 0.63267830149787, + "grad_norm": 1.0124558210372925, + "learning_rate": 1.7886106441725206e-05, + "loss": 0.9302, + "step": 9208 + }, + { + "epoch": 0.6327470111309605, + "grad_norm": 1.047102451324463, + "learning_rate": 1.7880358216056744e-05, + "loss": 0.9097, + "step": 9209 + }, + { + "epoch": 0.6328157207640511, + "grad_norm": 1.0869345664978027, + "learning_rate": 1.7874610399976934e-05, + "loss": 0.9168, + "step": 9210 + }, + { + "epoch": 0.6328844303971417, + "grad_norm": 1.106177806854248, + "learning_rate": 1.7868862993816442e-05, + "loss": 0.9535, + "step": 9211 + }, + { + "epoch": 0.6329531400302323, + "grad_norm": 1.0771653652191162, + "learning_rate": 1.786311599790591e-05, + "loss": 0.9643, + "step": 9212 + }, + { + "epoch": 0.6330218496633228, + "grad_norm": 1.0602526664733887, + "learning_rate": 1.7857369412575964e-05, + "loss": 0.9227, + "step": 9213 + }, + { + "epoch": 0.6330905592964133, + "grad_norm": 1.1169806718826294, + "learning_rate": 1.7851623238157198e-05, + "loss": 1.0104, + "step": 9214 + }, + { + "epoch": 0.6331592689295039, + "grad_norm": 1.1009067296981812, + "learning_rate": 1.7845877474980198e-05, + "loss": 0.9457, + "step": 9215 + }, + { + "epoch": 0.6332279785625945, + "grad_norm": 1.01157546043396, + "learning_rate": 1.78401321233755e-05, + "loss": 0.9241, + "step": 9216 + }, + { + "epoch": 0.6332966881956851, + "grad_norm": 1.1066205501556396, + "learning_rate": 1.7834387183673627e-05, + "loss": 0.9364, + "step": 9217 + }, + { + "epoch": 0.6333653978287755, + "grad_norm": 1.0912973880767822, + "learning_rate": 1.7828642656205097e-05, + "loss": 0.9334, + "step": 9218 + }, + { + "epoch": 0.6334341074618661, + "grad_norm": 1.049113392829895, + "learning_rate": 1.7822898541300382e-05, + "loss": 0.9456, + "step": 9219 + }, + { + "epoch": 0.6335028170949567, + "grad_norm": 1.0973479747772217, + "learning_rate": 1.7817154839289945e-05, + "loss": 0.9413, + "step": 9220 + }, + { + "epoch": 0.6335715267280473, + "grad_norm": 1.1183563470840454, + "learning_rate": 1.7811411550504202e-05, + "loss": 0.9673, + "step": 9221 + }, + { + "epoch": 0.6336402363611379, + "grad_norm": 1.023095726966858, + "learning_rate": 1.7805668675273575e-05, + "loss": 0.9298, + "step": 9222 + }, + { + "epoch": 0.6337089459942284, + "grad_norm": 1.069937825202942, + "learning_rate": 1.7799926213928437e-05, + "loss": 0.9315, + "step": 9223 + }, + { + "epoch": 0.6337776556273189, + "grad_norm": 1.070133090019226, + "learning_rate": 1.779418416679915e-05, + "loss": 0.964, + "step": 9224 + }, + { + "epoch": 0.6338463652604095, + "grad_norm": 1.0762412548065186, + "learning_rate": 1.7788442534216072e-05, + "loss": 0.9394, + "step": 9225 + }, + { + "epoch": 0.6339150748935001, + "grad_norm": 1.0615770816802979, + "learning_rate": 1.778270131650948e-05, + "loss": 0.9377, + "step": 9226 + }, + { + "epoch": 0.6339837845265907, + "grad_norm": 1.058005452156067, + "learning_rate": 1.7776960514009682e-05, + "loss": 0.9356, + "step": 9227 + }, + { + "epoch": 0.6340524941596812, + "grad_norm": 1.0919277667999268, + "learning_rate": 1.7771220127046944e-05, + "loss": 0.971, + "step": 9228 + }, + { + "epoch": 0.6341212037927717, + "grad_norm": 1.1246205568313599, + "learning_rate": 1.7765480155951505e-05, + "loss": 0.9691, + "step": 9229 + }, + { + "epoch": 0.6341899134258623, + "grad_norm": 1.0460125207901, + "learning_rate": 1.775974060105359e-05, + "loss": 0.9732, + "step": 9230 + }, + { + "epoch": 0.6342586230589529, + "grad_norm": 1.0747910737991333, + "learning_rate": 1.7754001462683375e-05, + "loss": 0.9525, + "step": 9231 + }, + { + "epoch": 0.6343273326920434, + "grad_norm": 1.0171281099319458, + "learning_rate": 1.7748262741171037e-05, + "loss": 0.9582, + "step": 9232 + }, + { + "epoch": 0.634396042325134, + "grad_norm": 1.036426305770874, + "learning_rate": 1.7742524436846726e-05, + "loss": 0.9256, + "step": 9233 + }, + { + "epoch": 0.6344647519582245, + "grad_norm": 1.0973256826400757, + "learning_rate": 1.7736786550040557e-05, + "loss": 1.0015, + "step": 9234 + }, + { + "epoch": 0.6345334615913151, + "grad_norm": 0.9808809757232666, + "learning_rate": 1.773104908108263e-05, + "loss": 0.9491, + "step": 9235 + }, + { + "epoch": 0.6346021712244057, + "grad_norm": 1.0440443754196167, + "learning_rate": 1.772531203030302e-05, + "loss": 0.9164, + "step": 9236 + }, + { + "epoch": 0.6346708808574962, + "grad_norm": 1.0787426233291626, + "learning_rate": 1.7719575398031767e-05, + "loss": 0.9561, + "step": 9237 + }, + { + "epoch": 0.6347395904905868, + "grad_norm": 1.0832940340042114, + "learning_rate": 1.77138391845989e-05, + "loss": 0.9293, + "step": 9238 + }, + { + "epoch": 0.6348083001236774, + "grad_norm": 1.091361403465271, + "learning_rate": 1.7708103390334428e-05, + "loss": 0.9836, + "step": 9239 + }, + { + "epoch": 0.6348770097567679, + "grad_norm": 1.134049654006958, + "learning_rate": 1.7702368015568328e-05, + "loss": 0.9965, + "step": 9240 + }, + { + "epoch": 0.6349457193898584, + "grad_norm": 1.0090056657791138, + "learning_rate": 1.7696633060630537e-05, + "loss": 0.9552, + "step": 9241 + }, + { + "epoch": 0.635014429022949, + "grad_norm": 1.0910189151763916, + "learning_rate": 1.769089852585099e-05, + "loss": 0.9596, + "step": 9242 + }, + { + "epoch": 0.6350831386560396, + "grad_norm": 1.045535922050476, + "learning_rate": 1.76851644115596e-05, + "loss": 0.9371, + "step": 9243 + }, + { + "epoch": 0.6351518482891302, + "grad_norm": 1.040502667427063, + "learning_rate": 1.7679430718086243e-05, + "loss": 0.9223, + "step": 9244 + }, + { + "epoch": 0.6352205579222207, + "grad_norm": 0.9996009469032288, + "learning_rate": 1.7673697445760767e-05, + "loss": 0.965, + "step": 9245 + }, + { + "epoch": 0.6352892675553112, + "grad_norm": 1.0425422191619873, + "learning_rate": 1.7667964594913017e-05, + "loss": 0.9605, + "step": 9246 + }, + { + "epoch": 0.6353579771884018, + "grad_norm": 1.066724181175232, + "learning_rate": 1.7662232165872794e-05, + "loss": 0.9458, + "step": 9247 + }, + { + "epoch": 0.6354266868214924, + "grad_norm": 1.0272765159606934, + "learning_rate": 1.7656500158969876e-05, + "loss": 0.9527, + "step": 9248 + }, + { + "epoch": 0.635495396454583, + "grad_norm": 1.0337785482406616, + "learning_rate": 1.765076857453402e-05, + "loss": 0.9588, + "step": 9249 + }, + { + "epoch": 0.6355641060876734, + "grad_norm": 1.018971562385559, + "learning_rate": 1.7645037412894974e-05, + "loss": 0.9741, + "step": 9250 + }, + { + "epoch": 0.635632815720764, + "grad_norm": 1.1086952686309814, + "learning_rate": 1.763930667438245e-05, + "loss": 0.9766, + "step": 9251 + }, + { + "epoch": 0.6357015253538546, + "grad_norm": 1.1000022888183594, + "learning_rate": 1.763357635932611e-05, + "loss": 0.9764, + "step": 9252 + }, + { + "epoch": 0.6357702349869452, + "grad_norm": 1.1838270425796509, + "learning_rate": 1.7627846468055638e-05, + "loss": 0.9999, + "step": 9253 + }, + { + "epoch": 0.6358389446200358, + "grad_norm": 1.0546129941940308, + "learning_rate": 1.7622117000900655e-05, + "loss": 0.9659, + "step": 9254 + }, + { + "epoch": 0.6359076542531263, + "grad_norm": 0.9950518608093262, + "learning_rate": 1.7616387958190793e-05, + "loss": 0.9211, + "step": 9255 + }, + { + "epoch": 0.6359763638862168, + "grad_norm": 1.0961092710494995, + "learning_rate": 1.761065934025563e-05, + "loss": 0.9941, + "step": 9256 + }, + { + "epoch": 0.6360450735193074, + "grad_norm": 1.049829363822937, + "learning_rate": 1.7604931147424722e-05, + "loss": 0.956, + "step": 9257 + }, + { + "epoch": 0.636113783152398, + "grad_norm": 1.1050364971160889, + "learning_rate": 1.7599203380027624e-05, + "loss": 0.9502, + "step": 9258 + }, + { + "epoch": 0.6361824927854886, + "grad_norm": 1.0146108865737915, + "learning_rate": 1.7593476038393836e-05, + "loss": 0.9511, + "step": 9259 + }, + { + "epoch": 0.6362512024185791, + "grad_norm": 1.0671719312667847, + "learning_rate": 1.7587749122852854e-05, + "loss": 0.9627, + "step": 9260 + }, + { + "epoch": 0.6363199120516696, + "grad_norm": 1.0815879106521606, + "learning_rate": 1.7582022633734156e-05, + "loss": 0.9838, + "step": 9261 + }, + { + "epoch": 0.6363886216847602, + "grad_norm": 1.1234703063964844, + "learning_rate": 1.7576296571367168e-05, + "loss": 0.991, + "step": 9262 + }, + { + "epoch": 0.6364573313178508, + "grad_norm": 1.0232243537902832, + "learning_rate": 1.7570570936081306e-05, + "loss": 0.9598, + "step": 9263 + }, + { + "epoch": 0.6365260409509413, + "grad_norm": 1.0483990907669067, + "learning_rate": 1.756484572820597e-05, + "loss": 0.9337, + "step": 9264 + }, + { + "epoch": 0.6365947505840319, + "grad_norm": 1.0247613191604614, + "learning_rate": 1.7559120948070524e-05, + "loss": 0.9541, + "step": 9265 + }, + { + "epoch": 0.6366634602171224, + "grad_norm": 0.9718225598335266, + "learning_rate": 1.7553396596004328e-05, + "loss": 0.9161, + "step": 9266 + }, + { + "epoch": 0.636732169850213, + "grad_norm": 1.0977792739868164, + "learning_rate": 1.7547672672336672e-05, + "loss": 0.9308, + "step": 9267 + }, + { + "epoch": 0.6368008794833036, + "grad_norm": 1.0288573503494263, + "learning_rate": 1.7541949177396864e-05, + "loss": 0.8998, + "step": 9268 + }, + { + "epoch": 0.6368695891163941, + "grad_norm": 1.0482990741729736, + "learning_rate": 1.7536226111514176e-05, + "loss": 0.9475, + "step": 9269 + }, + { + "epoch": 0.6369382987494847, + "grad_norm": 1.0405807495117188, + "learning_rate": 1.7530503475017845e-05, + "loss": 0.9509, + "step": 9270 + }, + { + "epoch": 0.6370070083825753, + "grad_norm": 1.1005021333694458, + "learning_rate": 1.7524781268237103e-05, + "loss": 0.9523, + "step": 9271 + }, + { + "epoch": 0.6370757180156658, + "grad_norm": 0.9818710088729858, + "learning_rate": 1.7519059491501128e-05, + "loss": 0.8977, + "step": 9272 + }, + { + "epoch": 0.6371444276487563, + "grad_norm": 1.0364394187927246, + "learning_rate": 1.75133381451391e-05, + "loss": 0.9467, + "step": 9273 + }, + { + "epoch": 0.6372131372818469, + "grad_norm": 1.0697542428970337, + "learning_rate": 1.750761722948016e-05, + "loss": 0.9308, + "step": 9274 + }, + { + "epoch": 0.6372818469149375, + "grad_norm": 1.0673881769180298, + "learning_rate": 1.7501896744853437e-05, + "loss": 0.966, + "step": 9275 + }, + { + "epoch": 0.6373505565480281, + "grad_norm": 1.0634392499923706, + "learning_rate": 1.749617669158803e-05, + "loss": 0.9102, + "step": 9276 + }, + { + "epoch": 0.6374192661811185, + "grad_norm": 1.0727767944335938, + "learning_rate": 1.7490457070012996e-05, + "loss": 0.9211, + "step": 9277 + }, + { + "epoch": 0.6374879758142091, + "grad_norm": 1.1075037717819214, + "learning_rate": 1.7484737880457386e-05, + "loss": 1.0175, + "step": 9278 + }, + { + "epoch": 0.6375566854472997, + "grad_norm": 1.0734046697616577, + "learning_rate": 1.747901912325022e-05, + "loss": 0.9599, + "step": 9279 + }, + { + "epoch": 0.6376253950803903, + "grad_norm": 1.0393682718276978, + "learning_rate": 1.7473300798720506e-05, + "loss": 0.9736, + "step": 9280 + }, + { + "epoch": 0.6376941047134809, + "grad_norm": 1.1159290075302124, + "learning_rate": 1.746758290719721e-05, + "loss": 0.9495, + "step": 9281 + }, + { + "epoch": 0.6377628143465713, + "grad_norm": 1.0305448770523071, + "learning_rate": 1.7461865449009272e-05, + "loss": 0.9387, + "step": 9282 + }, + { + "epoch": 0.6378315239796619, + "grad_norm": 1.0666402578353882, + "learning_rate": 1.7456148424485617e-05, + "loss": 0.9441, + "step": 9283 + }, + { + "epoch": 0.6379002336127525, + "grad_norm": 1.0460336208343506, + "learning_rate": 1.7450431833955138e-05, + "loss": 0.9203, + "step": 9284 + }, + { + "epoch": 0.6379689432458431, + "grad_norm": 1.0633962154388428, + "learning_rate": 1.744471567774672e-05, + "loss": 0.9438, + "step": 9285 + }, + { + "epoch": 0.6380376528789337, + "grad_norm": 1.0282819271087646, + "learning_rate": 1.7438999956189206e-05, + "loss": 0.9384, + "step": 9286 + }, + { + "epoch": 0.6381063625120242, + "grad_norm": 1.016845703125, + "learning_rate": 1.743328466961141e-05, + "loss": 0.9473, + "step": 9287 + }, + { + "epoch": 0.6381750721451147, + "grad_norm": 1.08345627784729, + "learning_rate": 1.742756981834213e-05, + "loss": 0.9649, + "step": 9288 + }, + { + "epoch": 0.6382437817782053, + "grad_norm": 1.0542383193969727, + "learning_rate": 1.742185540271014e-05, + "loss": 0.9856, + "step": 9289 + }, + { + "epoch": 0.6383124914112959, + "grad_norm": 1.055372953414917, + "learning_rate": 1.741614142304419e-05, + "loss": 0.9794, + "step": 9290 + }, + { + "epoch": 0.6383812010443864, + "grad_norm": 0.972689151763916, + "learning_rate": 1.741042787967301e-05, + "loss": 0.9296, + "step": 9291 + }, + { + "epoch": 0.638449910677477, + "grad_norm": 1.06538987159729, + "learning_rate": 1.7404714772925274e-05, + "loss": 0.9529, + "step": 9292 + }, + { + "epoch": 0.6385186203105675, + "grad_norm": 0.9882972836494446, + "learning_rate": 1.7399002103129676e-05, + "loss": 0.9034, + "step": 9293 + }, + { + "epoch": 0.6385873299436581, + "grad_norm": 1.0637542009353638, + "learning_rate": 1.7393289870614844e-05, + "loss": 0.9228, + "step": 9294 + }, + { + "epoch": 0.6386560395767487, + "grad_norm": 1.0147038698196411, + "learning_rate": 1.7387578075709403e-05, + "loss": 0.9359, + "step": 9295 + }, + { + "epoch": 0.6387247492098392, + "grad_norm": 1.0489134788513184, + "learning_rate": 1.7381866718741975e-05, + "loss": 0.9352, + "step": 9296 + }, + { + "epoch": 0.6387934588429298, + "grad_norm": 1.0673773288726807, + "learning_rate": 1.737615580004109e-05, + "loss": 0.9749, + "step": 9297 + }, + { + "epoch": 0.6388621684760203, + "grad_norm": 1.061004400253296, + "learning_rate": 1.7370445319935317e-05, + "loss": 0.9624, + "step": 9298 + }, + { + "epoch": 0.6389308781091109, + "grad_norm": 1.0884246826171875, + "learning_rate": 1.7364735278753172e-05, + "loss": 0.9322, + "step": 9299 + }, + { + "epoch": 0.6389995877422014, + "grad_norm": 1.116464376449585, + "learning_rate": 1.735902567682315e-05, + "loss": 0.9747, + "step": 9300 + }, + { + "epoch": 0.639068297375292, + "grad_norm": 1.0477598905563354, + "learning_rate": 1.7353316514473723e-05, + "loss": 0.9595, + "step": 9301 + }, + { + "epoch": 0.6391370070083826, + "grad_norm": 1.0215545892715454, + "learning_rate": 1.7347607792033345e-05, + "loss": 0.932, + "step": 9302 + }, + { + "epoch": 0.6392057166414732, + "grad_norm": 1.073856234550476, + "learning_rate": 1.7341899509830416e-05, + "loss": 0.9609, + "step": 9303 + }, + { + "epoch": 0.6392744262745637, + "grad_norm": 1.0806636810302734, + "learning_rate": 1.7336191668193342e-05, + "loss": 0.9492, + "step": 9304 + }, + { + "epoch": 0.6393431359076542, + "grad_norm": 1.0029590129852295, + "learning_rate": 1.7330484267450483e-05, + "loss": 0.9117, + "step": 9305 + }, + { + "epoch": 0.6394118455407448, + "grad_norm": 1.0425065755844116, + "learning_rate": 1.732477730793019e-05, + "loss": 0.9072, + "step": 9306 + }, + { + "epoch": 0.6394805551738354, + "grad_norm": 1.0407477617263794, + "learning_rate": 1.731907078996079e-05, + "loss": 0.9184, + "step": 9307 + }, + { + "epoch": 0.639549264806926, + "grad_norm": 1.0793689489364624, + "learning_rate": 1.7313364713870546e-05, + "loss": 0.9406, + "step": 9308 + }, + { + "epoch": 0.6396179744400164, + "grad_norm": 1.0328149795532227, + "learning_rate": 1.730765907998775e-05, + "loss": 0.9378, + "step": 9309 + }, + { + "epoch": 0.639686684073107, + "grad_norm": 1.0230951309204102, + "learning_rate": 1.7301953888640635e-05, + "loss": 0.9213, + "step": 9310 + }, + { + "epoch": 0.6397553937061976, + "grad_norm": 1.0693126916885376, + "learning_rate": 1.729624914015742e-05, + "loss": 0.9584, + "step": 9311 + }, + { + "epoch": 0.6398241033392882, + "grad_norm": 1.0657416582107544, + "learning_rate": 1.7290544834866305e-05, + "loss": 0.9501, + "step": 9312 + }, + { + "epoch": 0.6398928129723788, + "grad_norm": 1.1087887287139893, + "learning_rate": 1.7284840973095435e-05, + "loss": 0.9895, + "step": 9313 + }, + { + "epoch": 0.6399615226054692, + "grad_norm": 1.0702455043792725, + "learning_rate": 1.7279137555172956e-05, + "loss": 0.9482, + "step": 9314 + }, + { + "epoch": 0.6400302322385598, + "grad_norm": 1.024269461631775, + "learning_rate": 1.7273434581426996e-05, + "loss": 0.961, + "step": 9315 + }, + { + "epoch": 0.6400989418716504, + "grad_norm": 1.096817135810852, + "learning_rate": 1.7267732052185624e-05, + "loss": 0.9465, + "step": 9316 + }, + { + "epoch": 0.640167651504741, + "grad_norm": 1.1585501432418823, + "learning_rate": 1.726202996777692e-05, + "loss": 1.0093, + "step": 9317 + }, + { + "epoch": 0.6402363611378316, + "grad_norm": 1.055987000465393, + "learning_rate": 1.725632832852891e-05, + "loss": 0.9558, + "step": 9318 + }, + { + "epoch": 0.6403050707709221, + "grad_norm": 1.031204342842102, + "learning_rate": 1.725062713476961e-05, + "loss": 0.9464, + "step": 9319 + }, + { + "epoch": 0.6403737804040126, + "grad_norm": 1.0340118408203125, + "learning_rate": 1.7244926386827004e-05, + "loss": 0.9239, + "step": 9320 + }, + { + "epoch": 0.6404424900371032, + "grad_norm": 1.0385890007019043, + "learning_rate": 1.7239226085029054e-05, + "loss": 0.9046, + "step": 9321 + }, + { + "epoch": 0.6405111996701938, + "grad_norm": 1.024200677871704, + "learning_rate": 1.7233526229703705e-05, + "loss": 0.9146, + "step": 9322 + }, + { + "epoch": 0.6405799093032843, + "grad_norm": 1.0791324377059937, + "learning_rate": 1.7227826821178848e-05, + "loss": 0.9641, + "step": 9323 + }, + { + "epoch": 0.6406486189363749, + "grad_norm": 1.0535975694656372, + "learning_rate": 1.722212785978237e-05, + "loss": 0.9361, + "step": 9324 + }, + { + "epoch": 0.6407173285694654, + "grad_norm": 1.0508419275283813, + "learning_rate": 1.721642934584214e-05, + "loss": 0.9534, + "step": 9325 + }, + { + "epoch": 0.640786038202556, + "grad_norm": 1.090990424156189, + "learning_rate": 1.721073127968599e-05, + "loss": 1.0417, + "step": 9326 + }, + { + "epoch": 0.6408547478356466, + "grad_norm": 1.098088026046753, + "learning_rate": 1.7205033661641713e-05, + "loss": 0.9645, + "step": 9327 + }, + { + "epoch": 0.6409234574687371, + "grad_norm": 1.0333209037780762, + "learning_rate": 1.7199336492037097e-05, + "loss": 0.9317, + "step": 9328 + }, + { + "epoch": 0.6409921671018277, + "grad_norm": 0.9690247178077698, + "learning_rate": 1.7193639771199898e-05, + "loss": 0.9307, + "step": 9329 + }, + { + "epoch": 0.6410608767349182, + "grad_norm": 1.0371533632278442, + "learning_rate": 1.7187943499457842e-05, + "loss": 0.913, + "step": 9330 + }, + { + "epoch": 0.6411295863680088, + "grad_norm": 1.1449401378631592, + "learning_rate": 1.718224767713863e-05, + "loss": 0.9761, + "step": 9331 + }, + { + "epoch": 0.6411982960010993, + "grad_norm": 1.030651569366455, + "learning_rate": 1.7176552304569954e-05, + "loss": 0.9825, + "step": 9332 + }, + { + "epoch": 0.6412670056341899, + "grad_norm": 1.0922852754592896, + "learning_rate": 1.7170857382079443e-05, + "loss": 0.9669, + "step": 9333 + }, + { + "epoch": 0.6413357152672805, + "grad_norm": 1.0506037473678589, + "learning_rate": 1.716516290999473e-05, + "loss": 0.9996, + "step": 9334 + }, + { + "epoch": 0.6414044249003711, + "grad_norm": 1.0702197551727295, + "learning_rate": 1.7159468888643424e-05, + "loss": 0.9764, + "step": 9335 + }, + { + "epoch": 0.6414731345334616, + "grad_norm": 1.1256073713302612, + "learning_rate": 1.7153775318353088e-05, + "loss": 0.9892, + "step": 9336 + }, + { + "epoch": 0.6415418441665521, + "grad_norm": 1.1426829099655151, + "learning_rate": 1.714808219945129e-05, + "loss": 0.9403, + "step": 9337 + }, + { + "epoch": 0.6416105537996427, + "grad_norm": 1.0746374130249023, + "learning_rate": 1.7142389532265516e-05, + "loss": 0.9089, + "step": 9338 + }, + { + "epoch": 0.6416792634327333, + "grad_norm": 1.105007290840149, + "learning_rate": 1.7136697317123284e-05, + "loss": 0.9874, + "step": 9339 + }, + { + "epoch": 0.6417479730658239, + "grad_norm": 1.039623498916626, + "learning_rate": 1.713100555435207e-05, + "loss": 0.9219, + "step": 9340 + }, + { + "epoch": 0.6418166826989143, + "grad_norm": 1.0141379833221436, + "learning_rate": 1.71253142442793e-05, + "loss": 0.9506, + "step": 9341 + }, + { + "epoch": 0.6418853923320049, + "grad_norm": 1.1499899625778198, + "learning_rate": 1.7119623387232405e-05, + "loss": 0.9704, + "step": 9342 + }, + { + "epoch": 0.6419541019650955, + "grad_norm": 1.0246450901031494, + "learning_rate": 1.711393298353877e-05, + "loss": 0.926, + "step": 9343 + }, + { + "epoch": 0.6420228115981861, + "grad_norm": 1.0380580425262451, + "learning_rate": 1.710824303352576e-05, + "loss": 0.9583, + "step": 9344 + }, + { + "epoch": 0.6420915212312767, + "grad_norm": 1.0996205806732178, + "learning_rate": 1.7102553537520718e-05, + "loss": 0.9591, + "step": 9345 + }, + { + "epoch": 0.6421602308643671, + "grad_norm": 1.01565682888031, + "learning_rate": 1.7096864495850953e-05, + "loss": 0.9453, + "step": 9346 + }, + { + "epoch": 0.6422289404974577, + "grad_norm": 0.9977409839630127, + "learning_rate": 1.7091175908843765e-05, + "loss": 0.9107, + "step": 9347 + }, + { + "epoch": 0.6422976501305483, + "grad_norm": 1.086776614189148, + "learning_rate": 1.7085487776826397e-05, + "loss": 0.9381, + "step": 9348 + }, + { + "epoch": 0.6423663597636389, + "grad_norm": 1.0729197263717651, + "learning_rate": 1.707980010012609e-05, + "loss": 0.9233, + "step": 9349 + }, + { + "epoch": 0.6424350693967295, + "grad_norm": 1.0610092878341675, + "learning_rate": 1.7074112879070053e-05, + "loss": 0.9642, + "step": 9350 + }, + { + "epoch": 0.64250377902982, + "grad_norm": 1.0771852731704712, + "learning_rate": 1.7068426113985476e-05, + "loss": 0.9289, + "step": 9351 + }, + { + "epoch": 0.6425724886629105, + "grad_norm": 1.043266773223877, + "learning_rate": 1.7062739805199506e-05, + "loss": 0.9192, + "step": 9352 + }, + { + "epoch": 0.6426411982960011, + "grad_norm": 1.016170620918274, + "learning_rate": 1.705705395303927e-05, + "loss": 0.9369, + "step": 9353 + }, + { + "epoch": 0.6427099079290917, + "grad_norm": 1.0085843801498413, + "learning_rate": 1.7051368557831887e-05, + "loss": 0.9739, + "step": 9354 + }, + { + "epoch": 0.6427786175621822, + "grad_norm": 1.028591513633728, + "learning_rate": 1.7045683619904415e-05, + "loss": 0.9268, + "step": 9355 + }, + { + "epoch": 0.6428473271952728, + "grad_norm": 1.0834933519363403, + "learning_rate": 1.7039999139583915e-05, + "loss": 0.9289, + "step": 9356 + }, + { + "epoch": 0.6429160368283633, + "grad_norm": 1.0640244483947754, + "learning_rate": 1.7034315117197415e-05, + "loss": 0.9384, + "step": 9357 + }, + { + "epoch": 0.6429847464614539, + "grad_norm": 1.0563634634017944, + "learning_rate": 1.7028631553071914e-05, + "loss": 0.9272, + "step": 9358 + }, + { + "epoch": 0.6430534560945445, + "grad_norm": 1.0585639476776123, + "learning_rate": 1.702294844753437e-05, + "loss": 0.9376, + "step": 9359 + }, + { + "epoch": 0.643122165727635, + "grad_norm": 1.0610586404800415, + "learning_rate": 1.701726580091174e-05, + "loss": 0.9404, + "step": 9360 + }, + { + "epoch": 0.6431908753607256, + "grad_norm": 1.056135892868042, + "learning_rate": 1.701158361353094e-05, + "loss": 0.9807, + "step": 9361 + }, + { + "epoch": 0.6432595849938161, + "grad_norm": 1.0935980081558228, + "learning_rate": 1.700590188571887e-05, + "loss": 0.9688, + "step": 9362 + }, + { + "epoch": 0.6433282946269067, + "grad_norm": 1.0543628931045532, + "learning_rate": 1.7000220617802388e-05, + "loss": 0.9618, + "step": 9363 + }, + { + "epoch": 0.6433970042599972, + "grad_norm": 1.0794028043746948, + "learning_rate": 1.6994539810108333e-05, + "loss": 0.9488, + "step": 9364 + }, + { + "epoch": 0.6434657138930878, + "grad_norm": 1.0685639381408691, + "learning_rate": 1.6988859462963527e-05, + "loss": 0.917, + "step": 9365 + }, + { + "epoch": 0.6435344235261784, + "grad_norm": 1.0659822225570679, + "learning_rate": 1.6983179576694748e-05, + "loss": 0.9436, + "step": 9366 + }, + { + "epoch": 0.643603133159269, + "grad_norm": 0.9930995106697083, + "learning_rate": 1.6977500151628755e-05, + "loss": 0.9252, + "step": 9367 + }, + { + "epoch": 0.6436718427923594, + "grad_norm": 1.0296815633773804, + "learning_rate": 1.69718211880923e-05, + "loss": 0.9261, + "step": 9368 + }, + { + "epoch": 0.64374055242545, + "grad_norm": 1.080217957496643, + "learning_rate": 1.6966142686412066e-05, + "loss": 0.9747, + "step": 9369 + }, + { + "epoch": 0.6438092620585406, + "grad_norm": 1.0714720487594604, + "learning_rate": 1.6960464646914742e-05, + "loss": 0.9412, + "step": 9370 + }, + { + "epoch": 0.6438779716916312, + "grad_norm": 1.1042429208755493, + "learning_rate": 1.6954787069926988e-05, + "loss": 0.9636, + "step": 9371 + }, + { + "epoch": 0.6439466813247218, + "grad_norm": 1.0541629791259766, + "learning_rate": 1.694910995577543e-05, + "loss": 0.9431, + "step": 9372 + }, + { + "epoch": 0.6440153909578122, + "grad_norm": 1.0632387399673462, + "learning_rate": 1.6943433304786678e-05, + "loss": 0.9807, + "step": 9373 + }, + { + "epoch": 0.6440841005909028, + "grad_norm": 1.0605145692825317, + "learning_rate": 1.6937757117287278e-05, + "loss": 0.9764, + "step": 9374 + }, + { + "epoch": 0.6441528102239934, + "grad_norm": 1.104828119277954, + "learning_rate": 1.6932081393603805e-05, + "loss": 0.9796, + "step": 9375 + }, + { + "epoch": 0.644221519857084, + "grad_norm": 1.068032145500183, + "learning_rate": 1.6926406134062765e-05, + "loss": 0.9725, + "step": 9376 + }, + { + "epoch": 0.6442902294901746, + "grad_norm": 1.0633193254470825, + "learning_rate": 1.692073133899066e-05, + "loss": 0.9548, + "step": 9377 + }, + { + "epoch": 0.644358939123265, + "grad_norm": 1.1506671905517578, + "learning_rate": 1.6915057008713962e-05, + "loss": 0.9772, + "step": 9378 + }, + { + "epoch": 0.6444276487563556, + "grad_norm": 1.0467450618743896, + "learning_rate": 1.69093831435591e-05, + "loss": 0.9593, + "step": 9379 + }, + { + "epoch": 0.6444963583894462, + "grad_norm": 0.977731466293335, + "learning_rate": 1.6903709743852486e-05, + "loss": 0.9142, + "step": 9380 + }, + { + "epoch": 0.6445650680225368, + "grad_norm": 1.0185165405273438, + "learning_rate": 1.6898036809920515e-05, + "loss": 0.9148, + "step": 9381 + }, + { + "epoch": 0.6446337776556273, + "grad_norm": 1.0128260850906372, + "learning_rate": 1.689236434208955e-05, + "loss": 0.9058, + "step": 9382 + }, + { + "epoch": 0.6447024872887179, + "grad_norm": 1.049363136291504, + "learning_rate": 1.688669234068593e-05, + "loss": 0.9127, + "step": 9383 + }, + { + "epoch": 0.6447711969218084, + "grad_norm": 1.0789111852645874, + "learning_rate": 1.6881020806035934e-05, + "loss": 0.9348, + "step": 9384 + }, + { + "epoch": 0.644839906554899, + "grad_norm": 1.047218918800354, + "learning_rate": 1.687534973846587e-05, + "loss": 0.9237, + "step": 9385 + }, + { + "epoch": 0.6449086161879896, + "grad_norm": 1.0662130117416382, + "learning_rate": 1.686967913830198e-05, + "loss": 0.9628, + "step": 9386 + }, + { + "epoch": 0.6449773258210801, + "grad_norm": 1.1112006902694702, + "learning_rate": 1.6864009005870495e-05, + "loss": 0.9733, + "step": 9387 + }, + { + "epoch": 0.6450460354541707, + "grad_norm": 1.134528398513794, + "learning_rate": 1.6858339341497614e-05, + "loss": 0.9521, + "step": 9388 + }, + { + "epoch": 0.6451147450872612, + "grad_norm": 1.0899807214736938, + "learning_rate": 1.6852670145509503e-05, + "loss": 0.9423, + "step": 9389 + }, + { + "epoch": 0.6451834547203518, + "grad_norm": 1.0887806415557861, + "learning_rate": 1.6847001418232305e-05, + "loss": 0.9282, + "step": 9390 + }, + { + "epoch": 0.6452521643534423, + "grad_norm": 1.049600601196289, + "learning_rate": 1.6841333159992148e-05, + "loss": 0.9643, + "step": 9391 + }, + { + "epoch": 0.6453208739865329, + "grad_norm": 1.000217080116272, + "learning_rate": 1.6835665371115115e-05, + "loss": 0.9341, + "step": 9392 + }, + { + "epoch": 0.6453895836196235, + "grad_norm": 1.0427350997924805, + "learning_rate": 1.682999805192729e-05, + "loss": 0.9276, + "step": 9393 + }, + { + "epoch": 0.645458293252714, + "grad_norm": 1.0145314931869507, + "learning_rate": 1.6824331202754686e-05, + "loss": 0.9289, + "step": 9394 + }, + { + "epoch": 0.6455270028858046, + "grad_norm": 1.0370968580245972, + "learning_rate": 1.681866482392332e-05, + "loss": 0.9054, + "step": 9395 + }, + { + "epoch": 0.6455957125188951, + "grad_norm": 1.0411571264266968, + "learning_rate": 1.6812998915759182e-05, + "loss": 0.9325, + "step": 9396 + }, + { + "epoch": 0.6456644221519857, + "grad_norm": 1.0378344058990479, + "learning_rate": 1.6807333478588223e-05, + "loss": 0.9542, + "step": 9397 + }, + { + "epoch": 0.6457331317850763, + "grad_norm": 1.0069037675857544, + "learning_rate": 1.680166851273638e-05, + "loss": 0.9047, + "step": 9398 + }, + { + "epoch": 0.6458018414181669, + "grad_norm": 1.07395601272583, + "learning_rate": 1.679600401852954e-05, + "loss": 0.9564, + "step": 9399 + }, + { + "epoch": 0.6458705510512573, + "grad_norm": 1.0908012390136719, + "learning_rate": 1.679033999629359e-05, + "loss": 0.9851, + "step": 9400 + }, + { + "epoch": 0.6459392606843479, + "grad_norm": 0.9989377856254578, + "learning_rate": 1.6784676446354376e-05, + "loss": 0.9341, + "step": 9401 + }, + { + "epoch": 0.6460079703174385, + "grad_norm": 1.007010579109192, + "learning_rate": 1.6779013369037712e-05, + "loss": 0.9264, + "step": 9402 + }, + { + "epoch": 0.6460766799505291, + "grad_norm": 1.0871707201004028, + "learning_rate": 1.6773350764669406e-05, + "loss": 0.9894, + "step": 9403 + }, + { + "epoch": 0.6461453895836197, + "grad_norm": 1.0511913299560547, + "learning_rate": 1.6767688633575206e-05, + "loss": 0.9477, + "step": 9404 + }, + { + "epoch": 0.6462140992167101, + "grad_norm": 1.1157947778701782, + "learning_rate": 1.6762026976080857e-05, + "loss": 0.9111, + "step": 9405 + }, + { + "epoch": 0.6462828088498007, + "grad_norm": 1.0951743125915527, + "learning_rate": 1.6756365792512078e-05, + "loss": 0.9601, + "step": 9406 + }, + { + "epoch": 0.6463515184828913, + "grad_norm": 1.0387495756149292, + "learning_rate": 1.6750705083194545e-05, + "loss": 0.9064, + "step": 9407 + }, + { + "epoch": 0.6464202281159819, + "grad_norm": 1.1148618459701538, + "learning_rate": 1.6745044848453926e-05, + "loss": 0.9357, + "step": 9408 + }, + { + "epoch": 0.6464889377490725, + "grad_norm": 1.0463578701019287, + "learning_rate": 1.6739385088615838e-05, + "loss": 0.9462, + "step": 9409 + }, + { + "epoch": 0.6465576473821629, + "grad_norm": 1.0441473722457886, + "learning_rate": 1.673372580400589e-05, + "loss": 0.9583, + "step": 9410 + }, + { + "epoch": 0.6466263570152535, + "grad_norm": 1.0768635272979736, + "learning_rate": 1.672806699494966e-05, + "loss": 0.9155, + "step": 9411 + }, + { + "epoch": 0.6466950666483441, + "grad_norm": 0.9916418194770813, + "learning_rate": 1.6722408661772686e-05, + "loss": 0.8982, + "step": 9412 + }, + { + "epoch": 0.6467637762814347, + "grad_norm": 1.1077418327331543, + "learning_rate": 1.6716750804800493e-05, + "loss": 0.9368, + "step": 9413 + }, + { + "epoch": 0.6468324859145252, + "grad_norm": 1.0688971281051636, + "learning_rate": 1.6711093424358588e-05, + "loss": 0.949, + "step": 9414 + }, + { + "epoch": 0.6469011955476158, + "grad_norm": 1.0094419717788696, + "learning_rate": 1.6705436520772413e-05, + "loss": 0.9462, + "step": 9415 + }, + { + "epoch": 0.6469699051807063, + "grad_norm": 1.0473567247390747, + "learning_rate": 1.6699780094367416e-05, + "loss": 0.9698, + "step": 9416 + }, + { + "epoch": 0.6470386148137969, + "grad_norm": 1.0122911930084229, + "learning_rate": 1.6694124145469014e-05, + "loss": 0.936, + "step": 9417 + }, + { + "epoch": 0.6471073244468875, + "grad_norm": 1.0834100246429443, + "learning_rate": 1.668846867440258e-05, + "loss": 0.9709, + "step": 9418 + }, + { + "epoch": 0.647176034079978, + "grad_norm": 1.1822707653045654, + "learning_rate": 1.668281368149349e-05, + "loss": 0.961, + "step": 9419 + }, + { + "epoch": 0.6472447437130686, + "grad_norm": 1.1505343914031982, + "learning_rate": 1.6677159167067045e-05, + "loss": 0.9563, + "step": 9420 + }, + { + "epoch": 0.6473134533461591, + "grad_norm": 1.006380558013916, + "learning_rate": 1.667150513144856e-05, + "loss": 0.9198, + "step": 9421 + }, + { + "epoch": 0.6473821629792497, + "grad_norm": 1.0914827585220337, + "learning_rate": 1.6665851574963313e-05, + "loss": 0.9598, + "step": 9422 + }, + { + "epoch": 0.6474508726123402, + "grad_norm": 1.0458711385726929, + "learning_rate": 1.6660198497936536e-05, + "loss": 0.9479, + "step": 9423 + }, + { + "epoch": 0.6475195822454308, + "grad_norm": 1.140194296836853, + "learning_rate": 1.6654545900693458e-05, + "loss": 0.9791, + "step": 9424 + }, + { + "epoch": 0.6475882918785214, + "grad_norm": 1.0177401304244995, + "learning_rate": 1.664889378355927e-05, + "loss": 0.9528, + "step": 9425 + }, + { + "epoch": 0.6476570015116119, + "grad_norm": 1.116042971611023, + "learning_rate": 1.6643242146859123e-05, + "loss": 0.9939, + "step": 9426 + }, + { + "epoch": 0.6477257111447025, + "grad_norm": 1.0266642570495605, + "learning_rate": 1.663759099091816e-05, + "loss": 0.9356, + "step": 9427 + }, + { + "epoch": 0.647794420777793, + "grad_norm": 1.0096572637557983, + "learning_rate": 1.6631940316061486e-05, + "loss": 0.99, + "step": 9428 + }, + { + "epoch": 0.6478631304108836, + "grad_norm": 1.0239652395248413, + "learning_rate": 1.6626290122614195e-05, + "loss": 0.9614, + "step": 9429 + }, + { + "epoch": 0.6479318400439742, + "grad_norm": 1.0547447204589844, + "learning_rate": 1.662064041090132e-05, + "loss": 0.961, + "step": 9430 + }, + { + "epoch": 0.6480005496770648, + "grad_norm": 1.2216269969940186, + "learning_rate": 1.661499118124789e-05, + "loss": 0.9651, + "step": 9431 + }, + { + "epoch": 0.6480692593101552, + "grad_norm": 1.1049797534942627, + "learning_rate": 1.6609342433978908e-05, + "loss": 0.9643, + "step": 9432 + }, + { + "epoch": 0.6481379689432458, + "grad_norm": 1.1100835800170898, + "learning_rate": 1.6603694169419344e-05, + "loss": 0.9649, + "step": 9433 + }, + { + "epoch": 0.6482066785763364, + "grad_norm": 1.093865990638733, + "learning_rate": 1.6598046387894136e-05, + "loss": 0.9845, + "step": 9434 + }, + { + "epoch": 0.648275388209427, + "grad_norm": 1.0292706489562988, + "learning_rate": 1.6592399089728183e-05, + "loss": 0.9624, + "step": 9435 + }, + { + "epoch": 0.6483440978425176, + "grad_norm": 1.0896087884902954, + "learning_rate": 1.65867522752464e-05, + "loss": 0.9313, + "step": 9436 + }, + { + "epoch": 0.648412807475608, + "grad_norm": 1.0399786233901978, + "learning_rate": 1.6581105944773617e-05, + "loss": 0.947, + "step": 9437 + }, + { + "epoch": 0.6484815171086986, + "grad_norm": 1.019562005996704, + "learning_rate": 1.6575460098634677e-05, + "loss": 0.9523, + "step": 9438 + }, + { + "epoch": 0.6485502267417892, + "grad_norm": 1.0322123765945435, + "learning_rate": 1.656981473715439e-05, + "loss": 0.9649, + "step": 9439 + }, + { + "epoch": 0.6486189363748798, + "grad_norm": 1.0079293251037598, + "learning_rate": 1.6564169860657515e-05, + "loss": 0.9621, + "step": 9440 + }, + { + "epoch": 0.6486876460079704, + "grad_norm": 1.0190883874893188, + "learning_rate": 1.65585254694688e-05, + "loss": 0.9262, + "step": 9441 + }, + { + "epoch": 0.6487563556410608, + "grad_norm": 1.0506446361541748, + "learning_rate": 1.655288156391297e-05, + "loss": 0.8947, + "step": 9442 + }, + { + "epoch": 0.6488250652741514, + "grad_norm": 0.999724268913269, + "learning_rate": 1.6547238144314713e-05, + "loss": 0.9185, + "step": 9443 + }, + { + "epoch": 0.648893774907242, + "grad_norm": 1.046817660331726, + "learning_rate": 1.65415952109987e-05, + "loss": 0.967, + "step": 9444 + }, + { + "epoch": 0.6489624845403326, + "grad_norm": 1.084911823272705, + "learning_rate": 1.653595276428955e-05, + "loss": 0.9296, + "step": 9445 + }, + { + "epoch": 0.6490311941734231, + "grad_norm": 1.102226734161377, + "learning_rate": 1.6530310804511874e-05, + "loss": 0.9647, + "step": 9446 + }, + { + "epoch": 0.6490999038065137, + "grad_norm": 1.0546311140060425, + "learning_rate": 1.6524669331990257e-05, + "loss": 0.958, + "step": 9447 + }, + { + "epoch": 0.6491686134396042, + "grad_norm": 1.0634257793426514, + "learning_rate": 1.651902834704924e-05, + "loss": 0.9948, + "step": 9448 + }, + { + "epoch": 0.6492373230726948, + "grad_norm": 1.147200584411621, + "learning_rate": 1.6513387850013366e-05, + "loss": 0.9356, + "step": 9449 + }, + { + "epoch": 0.6493060327057854, + "grad_norm": 1.0307393074035645, + "learning_rate": 1.6507747841207105e-05, + "loss": 0.9545, + "step": 9450 + }, + { + "epoch": 0.6493747423388759, + "grad_norm": 0.9869968891143799, + "learning_rate": 1.650210832095493e-05, + "loss": 0.935, + "step": 9451 + }, + { + "epoch": 0.6494434519719665, + "grad_norm": 1.1133650541305542, + "learning_rate": 1.6496469289581278e-05, + "loss": 0.9635, + "step": 9452 + }, + { + "epoch": 0.649512161605057, + "grad_norm": 1.0730571746826172, + "learning_rate": 1.649083074741057e-05, + "loss": 0.9379, + "step": 9453 + }, + { + "epoch": 0.6495808712381476, + "grad_norm": 1.0851109027862549, + "learning_rate": 1.648519269476719e-05, + "loss": 0.89, + "step": 9454 + }, + { + "epoch": 0.6496495808712381, + "grad_norm": 1.0668662786483765, + "learning_rate": 1.6479555131975466e-05, + "loss": 0.9532, + "step": 9455 + }, + { + "epoch": 0.6497182905043287, + "grad_norm": 1.0579426288604736, + "learning_rate": 1.6473918059359745e-05, + "loss": 0.9414, + "step": 9456 + }, + { + "epoch": 0.6497870001374193, + "grad_norm": 1.0571266412734985, + "learning_rate": 1.646828147724432e-05, + "loss": 0.9553, + "step": 9457 + }, + { + "epoch": 0.6498557097705098, + "grad_norm": 1.0725433826446533, + "learning_rate": 1.6462645385953463e-05, + "loss": 0.9655, + "step": 9458 + }, + { + "epoch": 0.6499244194036004, + "grad_norm": 1.0331743955612183, + "learning_rate": 1.6457009785811408e-05, + "loss": 0.9248, + "step": 9459 + }, + { + "epoch": 0.6499931290366909, + "grad_norm": 1.022273302078247, + "learning_rate": 1.6451374677142375e-05, + "loss": 0.9264, + "step": 9460 + }, + { + "epoch": 0.6500618386697815, + "grad_norm": 1.0854086875915527, + "learning_rate": 1.6445740060270538e-05, + "loss": 0.9464, + "step": 9461 + }, + { + "epoch": 0.6501305483028721, + "grad_norm": 1.0669682025909424, + "learning_rate": 1.6440105935520063e-05, + "loss": 0.9244, + "step": 9462 + }, + { + "epoch": 0.6501992579359627, + "grad_norm": 1.0462726354599, + "learning_rate": 1.643447230321507e-05, + "loss": 0.9789, + "step": 9463 + }, + { + "epoch": 0.6502679675690531, + "grad_norm": 1.0476559400558472, + "learning_rate": 1.6428839163679665e-05, + "loss": 0.9739, + "step": 9464 + }, + { + "epoch": 0.6503366772021437, + "grad_norm": 1.083757996559143, + "learning_rate": 1.6423206517237927e-05, + "loss": 0.9762, + "step": 9465 + }, + { + "epoch": 0.6504053868352343, + "grad_norm": 1.0654547214508057, + "learning_rate": 1.6417574364213877e-05, + "loss": 0.9668, + "step": 9466 + }, + { + "epoch": 0.6504740964683249, + "grad_norm": 1.0377668142318726, + "learning_rate": 1.641194270493154e-05, + "loss": 0.9565, + "step": 9467 + }, + { + "epoch": 0.6505428061014155, + "grad_norm": 1.1074761152267456, + "learning_rate": 1.6406311539714906e-05, + "loss": 0.9615, + "step": 9468 + }, + { + "epoch": 0.6506115157345059, + "grad_norm": 1.0738368034362793, + "learning_rate": 1.640068086888793e-05, + "loss": 0.9356, + "step": 9469 + }, + { + "epoch": 0.6506802253675965, + "grad_norm": 1.0983469486236572, + "learning_rate": 1.6395050692774544e-05, + "loss": 0.925, + "step": 9470 + }, + { + "epoch": 0.6507489350006871, + "grad_norm": 1.1034202575683594, + "learning_rate": 1.6389421011698647e-05, + "loss": 0.9868, + "step": 9471 + }, + { + "epoch": 0.6508176446337777, + "grad_norm": 1.010327935218811, + "learning_rate": 1.63837918259841e-05, + "loss": 0.919, + "step": 9472 + }, + { + "epoch": 0.6508863542668683, + "grad_norm": 1.130110263824463, + "learning_rate": 1.6378163135954755e-05, + "loss": 0.9332, + "step": 9473 + }, + { + "epoch": 0.6509550638999587, + "grad_norm": 1.0835227966308594, + "learning_rate": 1.637253494193443e-05, + "loss": 0.9434, + "step": 9474 + }, + { + "epoch": 0.6510237735330493, + "grad_norm": 1.0580713748931885, + "learning_rate": 1.6366907244246917e-05, + "loss": 0.9263, + "step": 9475 + }, + { + "epoch": 0.6510924831661399, + "grad_norm": 1.049073576927185, + "learning_rate": 1.636128004321596e-05, + "loss": 0.9556, + "step": 9476 + }, + { + "epoch": 0.6511611927992305, + "grad_norm": 1.05020272731781, + "learning_rate": 1.6355653339165292e-05, + "loss": 0.9391, + "step": 9477 + }, + { + "epoch": 0.651229902432321, + "grad_norm": 1.1297780275344849, + "learning_rate": 1.6350027132418617e-05, + "loss": 0.9645, + "step": 9478 + }, + { + "epoch": 0.6512986120654116, + "grad_norm": 1.0592851638793945, + "learning_rate": 1.6344401423299608e-05, + "loss": 0.9517, + "step": 9479 + }, + { + "epoch": 0.6513673216985021, + "grad_norm": 1.059904932975769, + "learning_rate": 1.633877621213192e-05, + "loss": 0.966, + "step": 9480 + }, + { + "epoch": 0.6514360313315927, + "grad_norm": 1.0304231643676758, + "learning_rate": 1.633315149923914e-05, + "loss": 0.9408, + "step": 9481 + }, + { + "epoch": 0.6515047409646832, + "grad_norm": 1.0701556205749512, + "learning_rate": 1.6327527284944878e-05, + "loss": 0.9475, + "step": 9482 + }, + { + "epoch": 0.6515734505977738, + "grad_norm": 1.0721352100372314, + "learning_rate": 1.6321903569572677e-05, + "loss": 0.9313, + "step": 9483 + }, + { + "epoch": 0.6516421602308644, + "grad_norm": 1.0504326820373535, + "learning_rate": 1.6316280353446073e-05, + "loss": 0.9285, + "step": 9484 + }, + { + "epoch": 0.6517108698639549, + "grad_norm": 1.0779072046279907, + "learning_rate": 1.6310657636888574e-05, + "loss": 0.9556, + "step": 9485 + }, + { + "epoch": 0.6517795794970455, + "grad_norm": 1.1182833909988403, + "learning_rate": 1.6305035420223636e-05, + "loss": 0.9631, + "step": 9486 + }, + { + "epoch": 0.651848289130136, + "grad_norm": 1.07615327835083, + "learning_rate": 1.6299413703774706e-05, + "loss": 0.9214, + "step": 9487 + }, + { + "epoch": 0.6519169987632266, + "grad_norm": 1.047699213027954, + "learning_rate": 1.6293792487865202e-05, + "loss": 0.9538, + "step": 9488 + }, + { + "epoch": 0.6519857083963172, + "grad_norm": 1.0425164699554443, + "learning_rate": 1.6288171772818507e-05, + "loss": 0.9222, + "step": 9489 + }, + { + "epoch": 0.6520544180294077, + "grad_norm": 1.017171025276184, + "learning_rate": 1.6282551558957985e-05, + "loss": 0.938, + "step": 9490 + }, + { + "epoch": 0.6521231276624982, + "grad_norm": 1.0753272771835327, + "learning_rate": 1.6276931846606952e-05, + "loss": 0.9444, + "step": 9491 + }, + { + "epoch": 0.6521918372955888, + "grad_norm": 1.0751473903656006, + "learning_rate": 1.627131263608871e-05, + "loss": 0.9674, + "step": 9492 + }, + { + "epoch": 0.6522605469286794, + "grad_norm": 1.0657403469085693, + "learning_rate": 1.6265693927726534e-05, + "loss": 0.8895, + "step": 9493 + }, + { + "epoch": 0.65232925656177, + "grad_norm": 1.0988469123840332, + "learning_rate": 1.6260075721843655e-05, + "loss": 0.977, + "step": 9494 + }, + { + "epoch": 0.6523979661948606, + "grad_norm": 1.1157580614089966, + "learning_rate": 1.6254458018763298e-05, + "loss": 0.9628, + "step": 9495 + }, + { + "epoch": 0.652466675827951, + "grad_norm": 1.067623257637024, + "learning_rate": 1.6248840818808636e-05, + "loss": 0.9765, + "step": 9496 + }, + { + "epoch": 0.6525353854610416, + "grad_norm": 1.0667779445648193, + "learning_rate": 1.624322412230282e-05, + "loss": 0.9524, + "step": 9497 + }, + { + "epoch": 0.6526040950941322, + "grad_norm": 1.0511884689331055, + "learning_rate": 1.623760792956898e-05, + "loss": 0.9803, + "step": 9498 + }, + { + "epoch": 0.6526728047272228, + "grad_norm": 1.085105299949646, + "learning_rate": 1.623199224093022e-05, + "loss": 0.9721, + "step": 9499 + }, + { + "epoch": 0.6527415143603134, + "grad_norm": 1.0149073600769043, + "learning_rate": 1.6226377056709606e-05, + "loss": 0.9313, + "step": 9500 + }, + { + "epoch": 0.6528102239934038, + "grad_norm": 1.0686280727386475, + "learning_rate": 1.622076237723016e-05, + "loss": 0.9567, + "step": 9501 + }, + { + "epoch": 0.6528789336264944, + "grad_norm": 1.0417331457138062, + "learning_rate": 1.6215148202814902e-05, + "loss": 0.9583, + "step": 9502 + }, + { + "epoch": 0.652947643259585, + "grad_norm": 1.036383032798767, + "learning_rate": 1.620953453378681e-05, + "loss": 0.8864, + "step": 9503 + }, + { + "epoch": 0.6530163528926756, + "grad_norm": 1.0748662948608398, + "learning_rate": 1.620392137046884e-05, + "loss": 0.9181, + "step": 9504 + }, + { + "epoch": 0.6530850625257661, + "grad_norm": 1.024985432624817, + "learning_rate": 1.619830871318392e-05, + "loss": 0.9488, + "step": 9505 + }, + { + "epoch": 0.6531537721588567, + "grad_norm": 1.0288057327270508, + "learning_rate": 1.6192696562254918e-05, + "loss": 0.9437, + "step": 9506 + }, + { + "epoch": 0.6532224817919472, + "grad_norm": 1.0662815570831299, + "learning_rate": 1.618708491800472e-05, + "loss": 0.9355, + "step": 9507 + }, + { + "epoch": 0.6532911914250378, + "grad_norm": 1.1136234998703003, + "learning_rate": 1.618147378075615e-05, + "loss": 0.9482, + "step": 9508 + }, + { + "epoch": 0.6533599010581284, + "grad_norm": 1.0013643503189087, + "learning_rate": 1.6175863150832015e-05, + "loss": 0.9438, + "step": 9509 + }, + { + "epoch": 0.6534286106912189, + "grad_norm": 1.0461889505386353, + "learning_rate": 1.6170253028555105e-05, + "loss": 0.9096, + "step": 9510 + }, + { + "epoch": 0.6534973203243095, + "grad_norm": 1.002821922302246, + "learning_rate": 1.616464341424814e-05, + "loss": 0.9249, + "step": 9511 + }, + { + "epoch": 0.6535660299574, + "grad_norm": 1.087815761566162, + "learning_rate": 1.6159034308233855e-05, + "loss": 0.9454, + "step": 9512 + }, + { + "epoch": 0.6536347395904906, + "grad_norm": 1.061657428741455, + "learning_rate": 1.6153425710834934e-05, + "loss": 0.9509, + "step": 9513 + }, + { + "epoch": 0.6537034492235811, + "grad_norm": 1.0127061605453491, + "learning_rate": 1.614781762237404e-05, + "loss": 0.9571, + "step": 9514 + }, + { + "epoch": 0.6537721588566717, + "grad_norm": 1.0464764833450317, + "learning_rate": 1.614221004317381e-05, + "loss": 0.95, + "step": 9515 + }, + { + "epoch": 0.6538408684897623, + "grad_norm": 1.097933292388916, + "learning_rate": 1.6136602973556826e-05, + "loss": 0.9368, + "step": 9516 + }, + { + "epoch": 0.6539095781228528, + "grad_norm": 1.090559720993042, + "learning_rate": 1.6130996413845666e-05, + "loss": 0.9453, + "step": 9517 + }, + { + "epoch": 0.6539782877559434, + "grad_norm": 1.1192824840545654, + "learning_rate": 1.612539036436288e-05, + "loss": 0.9784, + "step": 9518 + }, + { + "epoch": 0.6540469973890339, + "grad_norm": 1.124162197113037, + "learning_rate": 1.6119784825430972e-05, + "loss": 0.9744, + "step": 9519 + }, + { + "epoch": 0.6541157070221245, + "grad_norm": 1.0520377159118652, + "learning_rate": 1.611417979737243e-05, + "loss": 0.9056, + "step": 9520 + }, + { + "epoch": 0.6541844166552151, + "grad_norm": 1.098837971687317, + "learning_rate": 1.6108575280509714e-05, + "loss": 0.9353, + "step": 9521 + }, + { + "epoch": 0.6542531262883057, + "grad_norm": 1.0186420679092407, + "learning_rate": 1.6102971275165228e-05, + "loss": 0.9182, + "step": 9522 + }, + { + "epoch": 0.6543218359213961, + "grad_norm": 1.0846196413040161, + "learning_rate": 1.6097367781661386e-05, + "loss": 0.9542, + "step": 9523 + }, + { + "epoch": 0.6543905455544867, + "grad_norm": 1.0439462661743164, + "learning_rate": 1.6091764800320543e-05, + "loss": 0.9677, + "step": 9524 + }, + { + "epoch": 0.6544592551875773, + "grad_norm": 1.143890619277954, + "learning_rate": 1.608616233146504e-05, + "loss": 1.0034, + "step": 9525 + }, + { + "epoch": 0.6545279648206679, + "grad_norm": 1.0723732709884644, + "learning_rate": 1.6080560375417193e-05, + "loss": 0.9538, + "step": 9526 + }, + { + "epoch": 0.6545966744537585, + "grad_norm": 1.0590659379959106, + "learning_rate": 1.607495893249926e-05, + "loss": 0.9592, + "step": 9527 + }, + { + "epoch": 0.6546653840868489, + "grad_norm": 1.0296647548675537, + "learning_rate": 1.6069358003033497e-05, + "loss": 0.9373, + "step": 9528 + }, + { + "epoch": 0.6547340937199395, + "grad_norm": 1.1775163412094116, + "learning_rate": 1.6063757587342125e-05, + "loss": 0.9945, + "step": 9529 + }, + { + "epoch": 0.6548028033530301, + "grad_norm": 1.0197101831436157, + "learning_rate": 1.6058157685747334e-05, + "loss": 0.9432, + "step": 9530 + }, + { + "epoch": 0.6548715129861207, + "grad_norm": 1.1117708683013916, + "learning_rate": 1.6052558298571275e-05, + "loss": 0.9332, + "step": 9531 + }, + { + "epoch": 0.6549402226192113, + "grad_norm": 1.0198911428451538, + "learning_rate": 1.6046959426136088e-05, + "loss": 0.9521, + "step": 9532 + }, + { + "epoch": 0.6550089322523017, + "grad_norm": 1.0535697937011719, + "learning_rate": 1.6041361068763854e-05, + "loss": 0.9486, + "step": 9533 + }, + { + "epoch": 0.6550776418853923, + "grad_norm": 1.0842589139938354, + "learning_rate": 1.603576322677666e-05, + "loss": 0.9716, + "step": 9534 + }, + { + "epoch": 0.6551463515184829, + "grad_norm": 1.0778952836990356, + "learning_rate": 1.603016590049654e-05, + "loss": 0.9662, + "step": 9535 + }, + { + "epoch": 0.6552150611515735, + "grad_norm": 1.074806571006775, + "learning_rate": 1.6024569090245524e-05, + "loss": 0.9723, + "step": 9536 + }, + { + "epoch": 0.655283770784664, + "grad_norm": 1.0980268716812134, + "learning_rate": 1.6018972796345554e-05, + "loss": 0.9665, + "step": 9537 + }, + { + "epoch": 0.6553524804177546, + "grad_norm": 1.0740636587142944, + "learning_rate": 1.601337701911861e-05, + "loss": 0.9308, + "step": 9538 + }, + { + "epoch": 0.6554211900508451, + "grad_norm": 1.049880862236023, + "learning_rate": 1.6007781758886607e-05, + "loss": 0.9665, + "step": 9539 + }, + { + "epoch": 0.6554898996839357, + "grad_norm": 1.0160764455795288, + "learning_rate": 1.6002187015971438e-05, + "loss": 0.9421, + "step": 9540 + }, + { + "epoch": 0.6555586093170263, + "grad_norm": 1.0005269050598145, + "learning_rate": 1.5996592790694963e-05, + "loss": 0.9273, + "step": 9541 + }, + { + "epoch": 0.6556273189501168, + "grad_norm": 1.0342754125595093, + "learning_rate": 1.5990999083379017e-05, + "loss": 0.9401, + "step": 9542 + }, + { + "epoch": 0.6556960285832074, + "grad_norm": 1.0094529390335083, + "learning_rate": 1.59854058943454e-05, + "loss": 0.9173, + "step": 9543 + }, + { + "epoch": 0.6557647382162979, + "grad_norm": 1.0912024974822998, + "learning_rate": 1.5979813223915878e-05, + "loss": 0.9778, + "step": 9544 + }, + { + "epoch": 0.6558334478493885, + "grad_norm": 1.1204403638839722, + "learning_rate": 1.5974221072412207e-05, + "loss": 0.955, + "step": 9545 + }, + { + "epoch": 0.655902157482479, + "grad_norm": 1.025503396987915, + "learning_rate": 1.59686294401561e-05, + "loss": 0.9508, + "step": 9546 + }, + { + "epoch": 0.6559708671155696, + "grad_norm": 1.027134895324707, + "learning_rate": 1.5963038327469233e-05, + "loss": 0.9647, + "step": 9547 + }, + { + "epoch": 0.6560395767486602, + "grad_norm": 1.1638413667678833, + "learning_rate": 1.5957447734673254e-05, + "loss": 0.9155, + "step": 9548 + }, + { + "epoch": 0.6561082863817507, + "grad_norm": 0.9980229735374451, + "learning_rate": 1.5951857662089793e-05, + "loss": 0.9096, + "step": 9549 + }, + { + "epoch": 0.6561769960148413, + "grad_norm": 1.0842288732528687, + "learning_rate": 1.5946268110040445e-05, + "loss": 0.9542, + "step": 9550 + }, + { + "epoch": 0.6562457056479318, + "grad_norm": 1.0455060005187988, + "learning_rate": 1.594067907884678e-05, + "loss": 0.9093, + "step": 9551 + }, + { + "epoch": 0.6563144152810224, + "grad_norm": 1.031103491783142, + "learning_rate": 1.5935090568830317e-05, + "loss": 0.9733, + "step": 9552 + }, + { + "epoch": 0.656383124914113, + "grad_norm": 1.0886881351470947, + "learning_rate": 1.5929502580312565e-05, + "loss": 0.9699, + "step": 9553 + }, + { + "epoch": 0.6564518345472036, + "grad_norm": 1.0455683469772339, + "learning_rate": 1.5923915113615005e-05, + "loss": 0.9505, + "step": 9554 + }, + { + "epoch": 0.656520544180294, + "grad_norm": 1.05660879611969, + "learning_rate": 1.5918328169059065e-05, + "loss": 0.942, + "step": 9555 + }, + { + "epoch": 0.6565892538133846, + "grad_norm": 1.0513314008712769, + "learning_rate": 1.5912741746966182e-05, + "loss": 0.947, + "step": 9556 + }, + { + "epoch": 0.6566579634464752, + "grad_norm": 1.0694783926010132, + "learning_rate": 1.5907155847657718e-05, + "loss": 0.9696, + "step": 9557 + }, + { + "epoch": 0.6567266730795658, + "grad_norm": 1.0133699178695679, + "learning_rate": 1.590157047145503e-05, + "loss": 0.9147, + "step": 9558 + }, + { + "epoch": 0.6567953827126564, + "grad_norm": 1.073147177696228, + "learning_rate": 1.5895985618679445e-05, + "loss": 0.8891, + "step": 9559 + }, + { + "epoch": 0.6568640923457468, + "grad_norm": 1.0763036012649536, + "learning_rate": 1.589040128965226e-05, + "loss": 0.9746, + "step": 9560 + }, + { + "epoch": 0.6569328019788374, + "grad_norm": 1.0709739923477173, + "learning_rate": 1.588481748469474e-05, + "loss": 0.9799, + "step": 9561 + }, + { + "epoch": 0.657001511611928, + "grad_norm": 1.0443042516708374, + "learning_rate": 1.5879234204128108e-05, + "loss": 0.9745, + "step": 9562 + }, + { + "epoch": 0.6570702212450186, + "grad_norm": 1.092483401298523, + "learning_rate": 1.5873651448273562e-05, + "loss": 0.9651, + "step": 9563 + }, + { + "epoch": 0.6571389308781092, + "grad_norm": 1.0934749841690063, + "learning_rate": 1.586806921745229e-05, + "loss": 0.9843, + "step": 9564 + }, + { + "epoch": 0.6572076405111996, + "grad_norm": 1.0416386127471924, + "learning_rate": 1.5862487511985433e-05, + "loss": 0.9372, + "step": 9565 + }, + { + "epoch": 0.6572763501442902, + "grad_norm": 1.08963942527771, + "learning_rate": 1.5856906332194098e-05, + "loss": 0.9619, + "step": 9566 + }, + { + "epoch": 0.6573450597773808, + "grad_norm": 1.0406506061553955, + "learning_rate": 1.5851325678399364e-05, + "loss": 0.9463, + "step": 9567 + }, + { + "epoch": 0.6574137694104714, + "grad_norm": 1.020248532295227, + "learning_rate": 1.584574555092228e-05, + "loss": 0.9814, + "step": 9568 + }, + { + "epoch": 0.6574824790435619, + "grad_norm": 1.0855865478515625, + "learning_rate": 1.584016595008388e-05, + "loss": 0.952, + "step": 9569 + }, + { + "epoch": 0.6575511886766525, + "grad_norm": 1.1485133171081543, + "learning_rate": 1.583458687620514e-05, + "loss": 0.9889, + "step": 9570 + }, + { + "epoch": 0.657619898309743, + "grad_norm": 1.0725189447402954, + "learning_rate": 1.582900832960704e-05, + "loss": 0.9672, + "step": 9571 + }, + { + "epoch": 0.6576886079428336, + "grad_norm": 1.0752733945846558, + "learning_rate": 1.5823430310610492e-05, + "loss": 0.9505, + "step": 9572 + }, + { + "epoch": 0.6577573175759241, + "grad_norm": 1.0368574857711792, + "learning_rate": 1.58178528195364e-05, + "loss": 0.9369, + "step": 9573 + }, + { + "epoch": 0.6578260272090147, + "grad_norm": 1.0505764484405518, + "learning_rate": 1.581227585670564e-05, + "loss": 0.9393, + "step": 9574 + }, + { + "epoch": 0.6578947368421053, + "grad_norm": 0.9970071911811829, + "learning_rate": 1.5806699422439044e-05, + "loss": 0.9209, + "step": 9575 + }, + { + "epoch": 0.6579634464751958, + "grad_norm": 1.0773032903671265, + "learning_rate": 1.5801123517057432e-05, + "loss": 0.9287, + "step": 9576 + }, + { + "epoch": 0.6580321561082864, + "grad_norm": 1.0500065088272095, + "learning_rate": 1.5795548140881575e-05, + "loss": 0.9674, + "step": 9577 + }, + { + "epoch": 0.6581008657413769, + "grad_norm": 1.0363905429840088, + "learning_rate": 1.5789973294232218e-05, + "loss": 0.9212, + "step": 9578 + }, + { + "epoch": 0.6581695753744675, + "grad_norm": 1.0401099920272827, + "learning_rate": 1.578439897743008e-05, + "loss": 0.9486, + "step": 9579 + }, + { + "epoch": 0.6582382850075581, + "grad_norm": 1.0228471755981445, + "learning_rate": 1.5778825190795852e-05, + "loss": 0.9016, + "step": 9580 + }, + { + "epoch": 0.6583069946406486, + "grad_norm": 1.021952748298645, + "learning_rate": 1.5773251934650184e-05, + "loss": 0.9335, + "step": 9581 + }, + { + "epoch": 0.6583757042737391, + "grad_norm": 1.085869550704956, + "learning_rate": 1.5767679209313717e-05, + "loss": 0.9848, + "step": 9582 + }, + { + "epoch": 0.6584444139068297, + "grad_norm": 1.0310308933258057, + "learning_rate": 1.576210701510703e-05, + "loss": 0.9459, + "step": 9583 + }, + { + "epoch": 0.6585131235399203, + "grad_norm": 1.0713156461715698, + "learning_rate": 1.5756535352350686e-05, + "loss": 0.9612, + "step": 9584 + }, + { + "epoch": 0.6585818331730109, + "grad_norm": 1.1065833568572998, + "learning_rate": 1.5750964221365234e-05, + "loss": 0.9391, + "step": 9585 + }, + { + "epoch": 0.6586505428061015, + "grad_norm": 1.0982797145843506, + "learning_rate": 1.574539362247117e-05, + "loss": 0.9685, + "step": 9586 + }, + { + "epoch": 0.6587192524391919, + "grad_norm": 1.08128821849823, + "learning_rate": 1.5739823555988975e-05, + "loss": 0.9755, + "step": 9587 + }, + { + "epoch": 0.6587879620722825, + "grad_norm": 1.0402908325195312, + "learning_rate": 1.5734254022239077e-05, + "loss": 0.9385, + "step": 9588 + }, + { + "epoch": 0.6588566717053731, + "grad_norm": 1.092411994934082, + "learning_rate": 1.57286850215419e-05, + "loss": 0.9623, + "step": 9589 + }, + { + "epoch": 0.6589253813384637, + "grad_norm": 1.011046051979065, + "learning_rate": 1.5723116554217814e-05, + "loss": 0.9702, + "step": 9590 + }, + { + "epoch": 0.6589940909715543, + "grad_norm": 0.9737231135368347, + "learning_rate": 1.571754862058718e-05, + "loss": 0.8588, + "step": 9591 + }, + { + "epoch": 0.6590628006046447, + "grad_norm": 1.0066982507705688, + "learning_rate": 1.5711981220970323e-05, + "loss": 0.9448, + "step": 9592 + }, + { + "epoch": 0.6591315102377353, + "grad_norm": 1.0316354036331177, + "learning_rate": 1.570641435568751e-05, + "loss": 0.9288, + "step": 9593 + }, + { + "epoch": 0.6592002198708259, + "grad_norm": 1.0277682542800903, + "learning_rate": 1.5700848025059014e-05, + "loss": 0.9331, + "step": 9594 + }, + { + "epoch": 0.6592689295039165, + "grad_norm": 1.075909972190857, + "learning_rate": 1.5695282229405063e-05, + "loss": 0.9714, + "step": 9595 + }, + { + "epoch": 0.659337639137007, + "grad_norm": 1.1715363264083862, + "learning_rate": 1.5689716969045848e-05, + "loss": 0.9641, + "step": 9596 + }, + { + "epoch": 0.6594063487700975, + "grad_norm": 1.044851303100586, + "learning_rate": 1.5684152244301554e-05, + "loss": 0.9315, + "step": 9597 + }, + { + "epoch": 0.6594750584031881, + "grad_norm": 1.136749505996704, + "learning_rate": 1.567858805549229e-05, + "loss": 0.9432, + "step": 9598 + }, + { + "epoch": 0.6595437680362787, + "grad_norm": 1.018057942390442, + "learning_rate": 1.567302440293817e-05, + "loss": 0.9661, + "step": 9599 + }, + { + "epoch": 0.6596124776693693, + "grad_norm": 1.0713025331497192, + "learning_rate": 1.5667461286959274e-05, + "loss": 0.9599, + "step": 9600 + }, + { + "epoch": 0.6596811873024598, + "grad_norm": 1.0841219425201416, + "learning_rate": 1.5661898707875633e-05, + "loss": 0.928, + "step": 9601 + }, + { + "epoch": 0.6597498969355504, + "grad_norm": 1.0301212072372437, + "learning_rate": 1.5656336666007274e-05, + "loss": 0.9218, + "step": 9602 + }, + { + "epoch": 0.6598186065686409, + "grad_norm": 1.038662075996399, + "learning_rate": 1.565077516167417e-05, + "loss": 0.9293, + "step": 9603 + }, + { + "epoch": 0.6598873162017315, + "grad_norm": 1.0660136938095093, + "learning_rate": 1.564521419519626e-05, + "loss": 0.9555, + "step": 9604 + }, + { + "epoch": 0.659956025834822, + "grad_norm": 1.126620888710022, + "learning_rate": 1.5639653766893476e-05, + "loss": 0.9423, + "step": 9605 + }, + { + "epoch": 0.6600247354679126, + "grad_norm": 1.0374277830123901, + "learning_rate": 1.5634093877085708e-05, + "loss": 0.9459, + "step": 9606 + }, + { + "epoch": 0.6600934451010032, + "grad_norm": 1.0901525020599365, + "learning_rate": 1.5628534526092815e-05, + "loss": 0.941, + "step": 9607 + }, + { + "epoch": 0.6601621547340937, + "grad_norm": 1.0629656314849854, + "learning_rate": 1.5622975714234604e-05, + "loss": 0.9352, + "step": 9608 + }, + { + "epoch": 0.6602308643671843, + "grad_norm": 1.0332965850830078, + "learning_rate": 1.561741744183089e-05, + "loss": 0.9193, + "step": 9609 + }, + { + "epoch": 0.6602995740002748, + "grad_norm": 1.0487346649169922, + "learning_rate": 1.5611859709201422e-05, + "loss": 0.9594, + "step": 9610 + }, + { + "epoch": 0.6603682836333654, + "grad_norm": 1.0804316997528076, + "learning_rate": 1.560630251666595e-05, + "loss": 0.9282, + "step": 9611 + }, + { + "epoch": 0.660436993266456, + "grad_norm": 1.0912959575653076, + "learning_rate": 1.5600745864544165e-05, + "loss": 0.9492, + "step": 9612 + }, + { + "epoch": 0.6605057028995465, + "grad_norm": 1.1291351318359375, + "learning_rate": 1.5595189753155733e-05, + "loss": 0.9936, + "step": 9613 + }, + { + "epoch": 0.660574412532637, + "grad_norm": 1.072161078453064, + "learning_rate": 1.5589634182820305e-05, + "loss": 0.959, + "step": 9614 + }, + { + "epoch": 0.6606431221657276, + "grad_norm": 1.0082405805587769, + "learning_rate": 1.5584079153857485e-05, + "loss": 0.9272, + "step": 9615 + }, + { + "epoch": 0.6607118317988182, + "grad_norm": 1.0947083234786987, + "learning_rate": 1.5578524666586848e-05, + "loss": 0.9807, + "step": 9616 + }, + { + "epoch": 0.6607805414319088, + "grad_norm": 1.083504319190979, + "learning_rate": 1.5572970721327948e-05, + "loss": 0.9413, + "step": 9617 + }, + { + "epoch": 0.6608492510649994, + "grad_norm": 1.1052539348602295, + "learning_rate": 1.5567417318400287e-05, + "loss": 0.9608, + "step": 9618 + }, + { + "epoch": 0.6609179606980898, + "grad_norm": 1.151816487312317, + "learning_rate": 1.5561864458123356e-05, + "loss": 0.9533, + "step": 9619 + }, + { + "epoch": 0.6609866703311804, + "grad_norm": 1.0955969095230103, + "learning_rate": 1.555631214081661e-05, + "loss": 0.9588, + "step": 9620 + }, + { + "epoch": 0.661055379964271, + "grad_norm": 1.1055485010147095, + "learning_rate": 1.5550760366799467e-05, + "loss": 0.9613, + "step": 9621 + }, + { + "epoch": 0.6611240895973616, + "grad_norm": 1.0137484073638916, + "learning_rate": 1.554520913639133e-05, + "loss": 0.937, + "step": 9622 + }, + { + "epoch": 0.6611927992304522, + "grad_norm": 1.0213980674743652, + "learning_rate": 1.5539658449911533e-05, + "loss": 0.8989, + "step": 9623 + }, + { + "epoch": 0.6612615088635426, + "grad_norm": 1.0334148406982422, + "learning_rate": 1.553410830767942e-05, + "loss": 0.9966, + "step": 9624 + }, + { + "epoch": 0.6613302184966332, + "grad_norm": 1.039332628250122, + "learning_rate": 1.5528558710014286e-05, + "loss": 0.9274, + "step": 9625 + }, + { + "epoch": 0.6613989281297238, + "grad_norm": 1.1012451648712158, + "learning_rate": 1.5523009657235393e-05, + "loss": 0.9653, + "step": 9626 + }, + { + "epoch": 0.6614676377628144, + "grad_norm": 1.0183933973312378, + "learning_rate": 1.5517461149661986e-05, + "loss": 0.925, + "step": 9627 + }, + { + "epoch": 0.661536347395905, + "grad_norm": 1.0469343662261963, + "learning_rate": 1.5511913187613246e-05, + "loss": 0.9513, + "step": 9628 + }, + { + "epoch": 0.6616050570289954, + "grad_norm": 1.044559359550476, + "learning_rate": 1.5506365771408354e-05, + "loss": 0.9559, + "step": 9629 + }, + { + "epoch": 0.661673766662086, + "grad_norm": 1.0771582126617432, + "learning_rate": 1.5500818901366453e-05, + "loss": 0.9679, + "step": 9630 + }, + { + "epoch": 0.6617424762951766, + "grad_norm": 1.069027304649353, + "learning_rate": 1.5495272577806646e-05, + "loss": 0.9695, + "step": 9631 + }, + { + "epoch": 0.6618111859282672, + "grad_norm": 1.1259610652923584, + "learning_rate": 1.5489726801048016e-05, + "loss": 0.9309, + "step": 9632 + }, + { + "epoch": 0.6618798955613577, + "grad_norm": 1.0531896352767944, + "learning_rate": 1.5484181571409613e-05, + "loss": 0.8994, + "step": 9633 + }, + { + "epoch": 0.6619486051944483, + "grad_norm": 1.0184522867202759, + "learning_rate": 1.547863688921043e-05, + "loss": 0.9605, + "step": 9634 + }, + { + "epoch": 0.6620173148275388, + "grad_norm": 1.0551947355270386, + "learning_rate": 1.5473092754769464e-05, + "loss": 0.9544, + "step": 9635 + }, + { + "epoch": 0.6620860244606294, + "grad_norm": 1.0927377939224243, + "learning_rate": 1.5467549168405665e-05, + "loss": 0.9787, + "step": 9636 + }, + { + "epoch": 0.6621547340937199, + "grad_norm": 1.0685898065567017, + "learning_rate": 1.5462006130437946e-05, + "loss": 0.9226, + "step": 9637 + }, + { + "epoch": 0.6622234437268105, + "grad_norm": 1.0728185176849365, + "learning_rate": 1.5456463641185204e-05, + "loss": 0.923, + "step": 9638 + }, + { + "epoch": 0.6622921533599011, + "grad_norm": 1.0227898359298706, + "learning_rate": 1.5450921700966292e-05, + "loss": 0.9562, + "step": 9639 + }, + { + "epoch": 0.6623608629929916, + "grad_norm": 1.0486568212509155, + "learning_rate": 1.544538031010002e-05, + "loss": 0.9132, + "step": 9640 + }, + { + "epoch": 0.6624295726260822, + "grad_norm": 0.9853944182395935, + "learning_rate": 1.5439839468905202e-05, + "loss": 0.9482, + "step": 9641 + }, + { + "epoch": 0.6624982822591727, + "grad_norm": 1.1699930429458618, + "learning_rate": 1.5434299177700583e-05, + "loss": 0.9703, + "step": 9642 + }, + { + "epoch": 0.6625669918922633, + "grad_norm": 1.040206789970398, + "learning_rate": 1.5428759436804914e-05, + "loss": 0.9262, + "step": 9643 + }, + { + "epoch": 0.6626357015253539, + "grad_norm": 1.0240930318832397, + "learning_rate": 1.5423220246536867e-05, + "loss": 0.967, + "step": 9644 + }, + { + "epoch": 0.6627044111584444, + "grad_norm": 1.0190777778625488, + "learning_rate": 1.5417681607215122e-05, + "loss": 0.9387, + "step": 9645 + }, + { + "epoch": 0.6627731207915349, + "grad_norm": 1.0619068145751953, + "learning_rate": 1.5412143519158313e-05, + "loss": 0.9568, + "step": 9646 + }, + { + "epoch": 0.6628418304246255, + "grad_norm": 1.0130702257156372, + "learning_rate": 1.540660598268504e-05, + "loss": 0.934, + "step": 9647 + }, + { + "epoch": 0.6629105400577161, + "grad_norm": 1.0430219173431396, + "learning_rate": 1.5401068998113884e-05, + "loss": 0.9398, + "step": 9648 + }, + { + "epoch": 0.6629792496908067, + "grad_norm": 1.0275392532348633, + "learning_rate": 1.539553256576336e-05, + "loss": 0.9343, + "step": 9649 + }, + { + "epoch": 0.6630479593238973, + "grad_norm": 1.0858607292175293, + "learning_rate": 1.5389996685952003e-05, + "loss": 0.9465, + "step": 9650 + }, + { + "epoch": 0.6631166689569877, + "grad_norm": 1.0317832231521606, + "learning_rate": 1.5384461358998274e-05, + "loss": 0.9136, + "step": 9651 + }, + { + "epoch": 0.6631853785900783, + "grad_norm": 1.0237042903900146, + "learning_rate": 1.5378926585220617e-05, + "loss": 0.9372, + "step": 9652 + }, + { + "epoch": 0.6632540882231689, + "grad_norm": 1.0752166509628296, + "learning_rate": 1.5373392364937456e-05, + "loss": 0.948, + "step": 9653 + }, + { + "epoch": 0.6633227978562595, + "grad_norm": 1.050750494003296, + "learning_rate": 1.5367858698467154e-05, + "loss": 0.9304, + "step": 9654 + }, + { + "epoch": 0.66339150748935, + "grad_norm": 1.1580113172531128, + "learning_rate": 1.536232558612807e-05, + "loss": 0.9828, + "step": 9655 + }, + { + "epoch": 0.6634602171224405, + "grad_norm": 1.0796608924865723, + "learning_rate": 1.5356793028238516e-05, + "loss": 0.9373, + "step": 9656 + }, + { + "epoch": 0.6635289267555311, + "grad_norm": 1.0386464595794678, + "learning_rate": 1.535126102511678e-05, + "loss": 0.9288, + "step": 9657 + }, + { + "epoch": 0.6635976363886217, + "grad_norm": 1.0319926738739014, + "learning_rate": 1.5345729577081125e-05, + "loss": 0.9361, + "step": 9658 + }, + { + "epoch": 0.6636663460217123, + "grad_norm": 1.0476855039596558, + "learning_rate": 1.5340198684449747e-05, + "loss": 0.9653, + "step": 9659 + }, + { + "epoch": 0.6637350556548028, + "grad_norm": 1.017019271850586, + "learning_rate": 1.5334668347540854e-05, + "loss": 0.9221, + "step": 9660 + }, + { + "epoch": 0.6638037652878933, + "grad_norm": 0.9963772892951965, + "learning_rate": 1.53291385666726e-05, + "loss": 0.9504, + "step": 9661 + }, + { + "epoch": 0.6638724749209839, + "grad_norm": 1.0286186933517456, + "learning_rate": 1.5323609342163104e-05, + "loss": 0.9221, + "step": 9662 + }, + { + "epoch": 0.6639411845540745, + "grad_norm": 1.1584316492080688, + "learning_rate": 1.531808067433047e-05, + "loss": 0.9719, + "step": 9663 + }, + { + "epoch": 0.664009894187165, + "grad_norm": 1.0835368633270264, + "learning_rate": 1.5312552563492742e-05, + "loss": 0.9116, + "step": 9664 + }, + { + "epoch": 0.6640786038202556, + "grad_norm": 1.0550106763839722, + "learning_rate": 1.530702500996796e-05, + "loss": 0.9825, + "step": 9665 + }, + { + "epoch": 0.6641473134533462, + "grad_norm": 1.1050738096237183, + "learning_rate": 1.5301498014074117e-05, + "loss": 0.9418, + "step": 9666 + }, + { + "epoch": 0.6642160230864367, + "grad_norm": 1.0355093479156494, + "learning_rate": 1.5295971576129183e-05, + "loss": 0.9238, + "step": 9667 + }, + { + "epoch": 0.6642847327195273, + "grad_norm": 1.0281798839569092, + "learning_rate": 1.5290445696451094e-05, + "loss": 0.9478, + "step": 9668 + }, + { + "epoch": 0.6643534423526178, + "grad_norm": 1.0303374528884888, + "learning_rate": 1.5284920375357736e-05, + "loss": 0.9202, + "step": 9669 + }, + { + "epoch": 0.6644221519857084, + "grad_norm": 1.1000007390975952, + "learning_rate": 1.5279395613166986e-05, + "loss": 0.9468, + "step": 9670 + }, + { + "epoch": 0.664490861618799, + "grad_norm": 1.0537149906158447, + "learning_rate": 1.5273871410196677e-05, + "loss": 0.9714, + "step": 9671 + }, + { + "epoch": 0.6645595712518895, + "grad_norm": 1.106668472290039, + "learning_rate": 1.526834776676462e-05, + "loss": 0.9586, + "step": 9672 + }, + { + "epoch": 0.66462828088498, + "grad_norm": 1.071576476097107, + "learning_rate": 1.5262824683188587e-05, + "loss": 0.9415, + "step": 9673 + }, + { + "epoch": 0.6646969905180706, + "grad_norm": 1.0744764804840088, + "learning_rate": 1.5257302159786307e-05, + "loss": 0.9508, + "step": 9674 + }, + { + "epoch": 0.6647657001511612, + "grad_norm": 1.0760442018508911, + "learning_rate": 1.5251780196875492e-05, + "loss": 0.9508, + "step": 9675 + }, + { + "epoch": 0.6648344097842518, + "grad_norm": 1.0606130361557007, + "learning_rate": 1.5246258794773815e-05, + "loss": 0.9069, + "step": 9676 + }, + { + "epoch": 0.6649031194173423, + "grad_norm": 1.0503331422805786, + "learning_rate": 1.5240737953798929e-05, + "loss": 0.9022, + "step": 9677 + }, + { + "epoch": 0.6649718290504328, + "grad_norm": 1.0731998682022095, + "learning_rate": 1.5235217674268442e-05, + "loss": 0.9493, + "step": 9678 + }, + { + "epoch": 0.6650405386835234, + "grad_norm": 1.1261701583862305, + "learning_rate": 1.5229697956499919e-05, + "loss": 0.9879, + "step": 9679 + }, + { + "epoch": 0.665109248316614, + "grad_norm": 1.0267302989959717, + "learning_rate": 1.5224178800810917e-05, + "loss": 0.9353, + "step": 9680 + }, + { + "epoch": 0.6651779579497046, + "grad_norm": 1.1211589574813843, + "learning_rate": 1.5218660207518948e-05, + "loss": 0.9623, + "step": 9681 + }, + { + "epoch": 0.6652466675827952, + "grad_norm": 1.0461859703063965, + "learning_rate": 1.5213142176941491e-05, + "loss": 0.9188, + "step": 9682 + }, + { + "epoch": 0.6653153772158856, + "grad_norm": 1.0233980417251587, + "learning_rate": 1.5207624709396006e-05, + "loss": 0.9245, + "step": 9683 + }, + { + "epoch": 0.6653840868489762, + "grad_norm": 1.0751748085021973, + "learning_rate": 1.5202107805199895e-05, + "loss": 0.9357, + "step": 9684 + }, + { + "epoch": 0.6654527964820668, + "grad_norm": 1.1123335361480713, + "learning_rate": 1.5196591464670552e-05, + "loss": 0.972, + "step": 9685 + }, + { + "epoch": 0.6655215061151574, + "grad_norm": 1.1014925241470337, + "learning_rate": 1.5191075688125317e-05, + "loss": 0.9444, + "step": 9686 + }, + { + "epoch": 0.665590215748248, + "grad_norm": 1.0306106805801392, + "learning_rate": 1.5185560475881519e-05, + "loss": 0.9407, + "step": 9687 + }, + { + "epoch": 0.6656589253813384, + "grad_norm": 1.2023634910583496, + "learning_rate": 1.5180045828256443e-05, + "loss": 1.0016, + "step": 9688 + }, + { + "epoch": 0.665727635014429, + "grad_norm": 1.0494089126586914, + "learning_rate": 1.5174531745567353e-05, + "loss": 0.9559, + "step": 9689 + }, + { + "epoch": 0.6657963446475196, + "grad_norm": 1.0697550773620605, + "learning_rate": 1.5169018228131449e-05, + "loss": 0.9244, + "step": 9690 + }, + { + "epoch": 0.6658650542806102, + "grad_norm": 1.0758553743362427, + "learning_rate": 1.5163505276265932e-05, + "loss": 0.9677, + "step": 9691 + }, + { + "epoch": 0.6659337639137007, + "grad_norm": 1.0519806146621704, + "learning_rate": 1.5157992890287958e-05, + "loss": 0.948, + "step": 9692 + }, + { + "epoch": 0.6660024735467912, + "grad_norm": 1.0663433074951172, + "learning_rate": 1.5152481070514657e-05, + "loss": 0.9181, + "step": 9693 + }, + { + "epoch": 0.6660711831798818, + "grad_norm": 1.0139350891113281, + "learning_rate": 1.5146969817263123e-05, + "loss": 0.9334, + "step": 9694 + }, + { + "epoch": 0.6661398928129724, + "grad_norm": 1.0737441778182983, + "learning_rate": 1.5141459130850397e-05, + "loss": 0.9577, + "step": 9695 + }, + { + "epoch": 0.666208602446063, + "grad_norm": 1.057545781135559, + "learning_rate": 1.513594901159352e-05, + "loss": 0.9475, + "step": 9696 + }, + { + "epoch": 0.6662773120791535, + "grad_norm": 1.0614303350448608, + "learning_rate": 1.5130439459809479e-05, + "loss": 0.9633, + "step": 9697 + }, + { + "epoch": 0.6663460217122441, + "grad_norm": 1.022010326385498, + "learning_rate": 1.5124930475815236e-05, + "loss": 0.9114, + "step": 9698 + }, + { + "epoch": 0.6664147313453346, + "grad_norm": 1.08286714553833, + "learning_rate": 1.5119422059927735e-05, + "loss": 0.962, + "step": 9699 + }, + { + "epoch": 0.6664834409784252, + "grad_norm": 1.0290014743804932, + "learning_rate": 1.5113914212463845e-05, + "loss": 0.945, + "step": 9700 + }, + { + "epoch": 0.6665521506115157, + "grad_norm": 1.036109447479248, + "learning_rate": 1.5108406933740443e-05, + "loss": 0.8996, + "step": 9701 + }, + { + "epoch": 0.6666208602446063, + "grad_norm": 1.0310558080673218, + "learning_rate": 1.5102900224074363e-05, + "loss": 0.9086, + "step": 9702 + }, + { + "epoch": 0.6666895698776969, + "grad_norm": 1.1514776945114136, + "learning_rate": 1.5097394083782393e-05, + "loss": 0.9816, + "step": 9703 + }, + { + "epoch": 0.6667582795107874, + "grad_norm": 1.0108811855316162, + "learning_rate": 1.5091888513181315e-05, + "loss": 0.9136, + "step": 9704 + }, + { + "epoch": 0.666826989143878, + "grad_norm": 1.061644434928894, + "learning_rate": 1.508638351258784e-05, + "loss": 0.9461, + "step": 9705 + }, + { + "epoch": 0.6668956987769685, + "grad_norm": 1.1432838439941406, + "learning_rate": 1.5080879082318678e-05, + "loss": 0.9783, + "step": 9706 + }, + { + "epoch": 0.6669644084100591, + "grad_norm": 1.0164135694503784, + "learning_rate": 1.5075375222690496e-05, + "loss": 0.9039, + "step": 9707 + }, + { + "epoch": 0.6670331180431497, + "grad_norm": 1.0601919889450073, + "learning_rate": 1.5069871934019924e-05, + "loss": 0.9611, + "step": 9708 + }, + { + "epoch": 0.6671018276762402, + "grad_norm": 1.0965865850448608, + "learning_rate": 1.506436921662357e-05, + "loss": 0.9746, + "step": 9709 + }, + { + "epoch": 0.6671705373093307, + "grad_norm": 0.9926850199699402, + "learning_rate": 1.5058867070817994e-05, + "loss": 0.8766, + "step": 9710 + }, + { + "epoch": 0.6672392469424213, + "grad_norm": 1.113710880279541, + "learning_rate": 1.5053365496919734e-05, + "loss": 0.9578, + "step": 9711 + }, + { + "epoch": 0.6673079565755119, + "grad_norm": 1.0828931331634521, + "learning_rate": 1.5047864495245289e-05, + "loss": 0.9797, + "step": 9712 + }, + { + "epoch": 0.6673766662086025, + "grad_norm": 1.03443443775177, + "learning_rate": 1.5042364066111132e-05, + "loss": 0.9117, + "step": 9713 + }, + { + "epoch": 0.6674453758416931, + "grad_norm": 1.0234042406082153, + "learning_rate": 1.5036864209833707e-05, + "loss": 0.946, + "step": 9714 + }, + { + "epoch": 0.6675140854747835, + "grad_norm": 1.0375559329986572, + "learning_rate": 1.5031364926729403e-05, + "loss": 0.955, + "step": 9715 + }, + { + "epoch": 0.6675827951078741, + "grad_norm": 1.0545481443405151, + "learning_rate": 1.5025866217114592e-05, + "loss": 0.9279, + "step": 9716 + }, + { + "epoch": 0.6676515047409647, + "grad_norm": 1.0607072114944458, + "learning_rate": 1.5020368081305619e-05, + "loss": 0.9571, + "step": 9717 + }, + { + "epoch": 0.6677202143740553, + "grad_norm": 1.1276445388793945, + "learning_rate": 1.501487051961879e-05, + "loss": 0.9697, + "step": 9718 + }, + { + "epoch": 0.6677889240071458, + "grad_norm": 1.1047979593276978, + "learning_rate": 1.5009373532370372e-05, + "loss": 0.9582, + "step": 9719 + }, + { + "epoch": 0.6678576336402363, + "grad_norm": 1.0688501596450806, + "learning_rate": 1.5003877119876596e-05, + "loss": 0.9621, + "step": 9720 + }, + { + "epoch": 0.6679263432733269, + "grad_norm": 1.0540462732315063, + "learning_rate": 1.4998381282453683e-05, + "loss": 0.9224, + "step": 9721 + }, + { + "epoch": 0.6679950529064175, + "grad_norm": 1.0647104978561401, + "learning_rate": 1.4992886020417787e-05, + "loss": 0.9316, + "step": 9722 + }, + { + "epoch": 0.668063762539508, + "grad_norm": 1.1127614974975586, + "learning_rate": 1.498739133408506e-05, + "loss": 0.9631, + "step": 9723 + }, + { + "epoch": 0.6681324721725986, + "grad_norm": 1.068098545074463, + "learning_rate": 1.4981897223771613e-05, + "loss": 0.95, + "step": 9724 + }, + { + "epoch": 0.6682011818056891, + "grad_norm": 1.1003265380859375, + "learning_rate": 1.49764036897935e-05, + "loss": 0.9454, + "step": 9725 + }, + { + "epoch": 0.6682698914387797, + "grad_norm": 1.0414202213287354, + "learning_rate": 1.4970910732466772e-05, + "loss": 0.9388, + "step": 9726 + }, + { + "epoch": 0.6683386010718703, + "grad_norm": 1.072139024734497, + "learning_rate": 1.4965418352107436e-05, + "loss": 0.9382, + "step": 9727 + }, + { + "epoch": 0.6684073107049608, + "grad_norm": 1.098517656326294, + "learning_rate": 1.495992654903146e-05, + "loss": 0.9807, + "step": 9728 + }, + { + "epoch": 0.6684760203380514, + "grad_norm": 1.0758949518203735, + "learning_rate": 1.49544353235548e-05, + "loss": 0.9691, + "step": 9729 + }, + { + "epoch": 0.668544729971142, + "grad_norm": 1.0555839538574219, + "learning_rate": 1.4948944675993343e-05, + "loss": 0.9529, + "step": 9730 + }, + { + "epoch": 0.6686134396042325, + "grad_norm": 1.0223873853683472, + "learning_rate": 1.4943454606662965e-05, + "loss": 0.9628, + "step": 9731 + }, + { + "epoch": 0.668682149237323, + "grad_norm": 1.031630516052246, + "learning_rate": 1.493796511587952e-05, + "loss": 0.9193, + "step": 9732 + }, + { + "epoch": 0.6687508588704136, + "grad_norm": 1.0268555879592896, + "learning_rate": 1.49324762039588e-05, + "loss": 0.9731, + "step": 9733 + }, + { + "epoch": 0.6688195685035042, + "grad_norm": 1.0659376382827759, + "learning_rate": 1.4926987871216597e-05, + "loss": 0.9246, + "step": 9734 + }, + { + "epoch": 0.6688882781365948, + "grad_norm": 1.0285446643829346, + "learning_rate": 1.4921500117968631e-05, + "loss": 0.9108, + "step": 9735 + }, + { + "epoch": 0.6689569877696853, + "grad_norm": 1.081947684288025, + "learning_rate": 1.4916012944530616e-05, + "loss": 0.9384, + "step": 9736 + }, + { + "epoch": 0.6690256974027758, + "grad_norm": 1.0076204538345337, + "learning_rate": 1.4910526351218229e-05, + "loss": 0.9112, + "step": 9737 + }, + { + "epoch": 0.6690944070358664, + "grad_norm": 1.0614306926727295, + "learning_rate": 1.4905040338347109e-05, + "loss": 0.9474, + "step": 9738 + }, + { + "epoch": 0.669163116668957, + "grad_norm": 1.0514225959777832, + "learning_rate": 1.4899554906232862e-05, + "loss": 0.9793, + "step": 9739 + }, + { + "epoch": 0.6692318263020476, + "grad_norm": 1.0862559080123901, + "learning_rate": 1.4894070055191073e-05, + "loss": 0.9789, + "step": 9740 + }, + { + "epoch": 0.669300535935138, + "grad_norm": 1.0502941608428955, + "learning_rate": 1.4888585785537262e-05, + "loss": 0.906, + "step": 9741 + }, + { + "epoch": 0.6693692455682286, + "grad_norm": 1.047721028327942, + "learning_rate": 1.4883102097586949e-05, + "loss": 0.965, + "step": 9742 + }, + { + "epoch": 0.6694379552013192, + "grad_norm": 1.0150614976882935, + "learning_rate": 1.4877618991655604e-05, + "loss": 0.9489, + "step": 9743 + }, + { + "epoch": 0.6695066648344098, + "grad_norm": 1.0632251501083374, + "learning_rate": 1.4872136468058661e-05, + "loss": 0.9462, + "step": 9744 + }, + { + "epoch": 0.6695753744675004, + "grad_norm": 1.1118444204330444, + "learning_rate": 1.4866654527111546e-05, + "loss": 0.9645, + "step": 9745 + }, + { + "epoch": 0.669644084100591, + "grad_norm": 1.0381581783294678, + "learning_rate": 1.4861173169129605e-05, + "loss": 0.9667, + "step": 9746 + }, + { + "epoch": 0.6697127937336814, + "grad_norm": 1.1118931770324707, + "learning_rate": 1.4855692394428189e-05, + "loss": 0.94, + "step": 9747 + }, + { + "epoch": 0.669781503366772, + "grad_norm": 1.0544782876968384, + "learning_rate": 1.4850212203322605e-05, + "loss": 0.9205, + "step": 9748 + }, + { + "epoch": 0.6698502129998626, + "grad_norm": 1.101656436920166, + "learning_rate": 1.4844732596128125e-05, + "loss": 0.9757, + "step": 9749 + }, + { + "epoch": 0.6699189226329532, + "grad_norm": 0.9848392009735107, + "learning_rate": 1.4839253573159994e-05, + "loss": 0.9628, + "step": 9750 + }, + { + "epoch": 0.6699876322660437, + "grad_norm": 1.1020584106445312, + "learning_rate": 1.4833775134733403e-05, + "loss": 0.9356, + "step": 9751 + }, + { + "epoch": 0.6700563418991342, + "grad_norm": 1.0192079544067383, + "learning_rate": 1.4828297281163527e-05, + "loss": 0.9174, + "step": 9752 + }, + { + "epoch": 0.6701250515322248, + "grad_norm": 1.129977822303772, + "learning_rate": 1.482282001276551e-05, + "loss": 0.9837, + "step": 9753 + }, + { + "epoch": 0.6701937611653154, + "grad_norm": 1.088987946510315, + "learning_rate": 1.4817343329854452e-05, + "loss": 0.913, + "step": 9754 + }, + { + "epoch": 0.670262470798406, + "grad_norm": 1.0077518224716187, + "learning_rate": 1.4811867232745427e-05, + "loss": 0.9212, + "step": 9755 + }, + { + "epoch": 0.6703311804314965, + "grad_norm": 1.0991222858428955, + "learning_rate": 1.4806391721753466e-05, + "loss": 0.9288, + "step": 9756 + }, + { + "epoch": 0.670399890064587, + "grad_norm": 1.026908278465271, + "learning_rate": 1.480091679719357e-05, + "loss": 0.9741, + "step": 9757 + }, + { + "epoch": 0.6704685996976776, + "grad_norm": 1.1274383068084717, + "learning_rate": 1.4795442459380715e-05, + "loss": 0.9351, + "step": 9758 + }, + { + "epoch": 0.6705373093307682, + "grad_norm": 1.0517991781234741, + "learning_rate": 1.478996870862983e-05, + "loss": 0.9495, + "step": 9759 + }, + { + "epoch": 0.6706060189638587, + "grad_norm": 1.0135105848312378, + "learning_rate": 1.478449554525583e-05, + "loss": 0.9238, + "step": 9760 + }, + { + "epoch": 0.6706747285969493, + "grad_norm": 1.034998893737793, + "learning_rate": 1.4779022969573563e-05, + "loss": 0.9244, + "step": 9761 + }, + { + "epoch": 0.6707434382300399, + "grad_norm": 1.0033791065216064, + "learning_rate": 1.477355098189787e-05, + "loss": 0.9162, + "step": 9762 + }, + { + "epoch": 0.6708121478631304, + "grad_norm": 1.0587234497070312, + "learning_rate": 1.476807958254356e-05, + "loss": 0.9697, + "step": 9763 + }, + { + "epoch": 0.670880857496221, + "grad_norm": 1.0534720420837402, + "learning_rate": 1.4762608771825387e-05, + "loss": 0.938, + "step": 9764 + }, + { + "epoch": 0.6709495671293115, + "grad_norm": 1.1150599718093872, + "learning_rate": 1.4757138550058102e-05, + "loss": 0.9481, + "step": 9765 + }, + { + "epoch": 0.6710182767624021, + "grad_norm": 1.188169002532959, + "learning_rate": 1.475166891755638e-05, + "loss": 0.9277, + "step": 9766 + }, + { + "epoch": 0.6710869863954927, + "grad_norm": 1.0716763734817505, + "learning_rate": 1.4746199874634903e-05, + "loss": 0.9727, + "step": 9767 + }, + { + "epoch": 0.6711556960285832, + "grad_norm": 1.0985987186431885, + "learning_rate": 1.4740731421608287e-05, + "loss": 0.9776, + "step": 9768 + }, + { + "epoch": 0.6712244056616737, + "grad_norm": 1.0843641757965088, + "learning_rate": 1.4735263558791137e-05, + "loss": 0.9495, + "step": 9769 + }, + { + "epoch": 0.6712931152947643, + "grad_norm": 1.0139052867889404, + "learning_rate": 1.4729796286498027e-05, + "loss": 0.9415, + "step": 9770 + }, + { + "epoch": 0.6713618249278549, + "grad_norm": 1.0808860063552856, + "learning_rate": 1.4724329605043466e-05, + "loss": 0.9677, + "step": 9771 + }, + { + "epoch": 0.6714305345609455, + "grad_norm": 1.0221257209777832, + "learning_rate": 1.4718863514741954e-05, + "loss": 0.9203, + "step": 9772 + }, + { + "epoch": 0.671499244194036, + "grad_norm": 1.0560588836669922, + "learning_rate": 1.4713398015907958e-05, + "loss": 0.9341, + "step": 9773 + }, + { + "epoch": 0.6715679538271265, + "grad_norm": 1.0808379650115967, + "learning_rate": 1.4707933108855904e-05, + "loss": 0.9235, + "step": 9774 + }, + { + "epoch": 0.6716366634602171, + "grad_norm": 1.0513911247253418, + "learning_rate": 1.4702468793900188e-05, + "loss": 0.9325, + "step": 9775 + }, + { + "epoch": 0.6717053730933077, + "grad_norm": 1.1072925329208374, + "learning_rate": 1.4697005071355157e-05, + "loss": 0.9211, + "step": 9776 + }, + { + "epoch": 0.6717740827263983, + "grad_norm": 1.0880600214004517, + "learning_rate": 1.4691541941535141e-05, + "loss": 0.945, + "step": 9777 + }, + { + "epoch": 0.6718427923594888, + "grad_norm": 1.030088186264038, + "learning_rate": 1.4686079404754439e-05, + "loss": 0.8972, + "step": 9778 + }, + { + "epoch": 0.6719115019925793, + "grad_norm": 1.137538194656372, + "learning_rate": 1.468061746132729e-05, + "loss": 0.9829, + "step": 9779 + }, + { + "epoch": 0.6719802116256699, + "grad_norm": 1.0150407552719116, + "learning_rate": 1.4675156111567933e-05, + "loss": 0.95, + "step": 9780 + }, + { + "epoch": 0.6720489212587605, + "grad_norm": 1.060448169708252, + "learning_rate": 1.4669695355790552e-05, + "loss": 0.9343, + "step": 9781 + }, + { + "epoch": 0.6721176308918511, + "grad_norm": 1.0745545625686646, + "learning_rate": 1.4664235194309292e-05, + "loss": 0.9617, + "step": 9782 + }, + { + "epoch": 0.6721863405249416, + "grad_norm": 0.9798993468284607, + "learning_rate": 1.4658775627438281e-05, + "loss": 0.929, + "step": 9783 + }, + { + "epoch": 0.6722550501580321, + "grad_norm": 1.1060466766357422, + "learning_rate": 1.4653316655491601e-05, + "loss": 0.964, + "step": 9784 + }, + { + "epoch": 0.6723237597911227, + "grad_norm": 1.0837386846542358, + "learning_rate": 1.4647858278783313e-05, + "loss": 0.9532, + "step": 9785 + }, + { + "epoch": 0.6723924694242133, + "grad_norm": 1.0909960269927979, + "learning_rate": 1.4642400497627417e-05, + "loss": 0.927, + "step": 9786 + }, + { + "epoch": 0.6724611790573038, + "grad_norm": 1.1404807567596436, + "learning_rate": 1.4636943312337909e-05, + "loss": 0.9754, + "step": 9787 + }, + { + "epoch": 0.6725298886903944, + "grad_norm": 1.023071527481079, + "learning_rate": 1.4631486723228732e-05, + "loss": 0.9081, + "step": 9788 + }, + { + "epoch": 0.6725985983234849, + "grad_norm": 0.9954321980476379, + "learning_rate": 1.4626030730613804e-05, + "loss": 0.9177, + "step": 9789 + }, + { + "epoch": 0.6726673079565755, + "grad_norm": 1.0626475811004639, + "learning_rate": 1.4620575334807005e-05, + "loss": 0.9024, + "step": 9790 + }, + { + "epoch": 0.6727360175896661, + "grad_norm": 1.0688321590423584, + "learning_rate": 1.4615120536122173e-05, + "loss": 0.968, + "step": 9791 + }, + { + "epoch": 0.6728047272227566, + "grad_norm": 1.061169147491455, + "learning_rate": 1.460966633487313e-05, + "loss": 0.9745, + "step": 9792 + }, + { + "epoch": 0.6728734368558472, + "grad_norm": 1.020169734954834, + "learning_rate": 1.460421273137364e-05, + "loss": 0.9329, + "step": 9793 + }, + { + "epoch": 0.6729421464889378, + "grad_norm": 1.048397183418274, + "learning_rate": 1.4598759725937456e-05, + "loss": 0.9378, + "step": 9794 + }, + { + "epoch": 0.6730108561220283, + "grad_norm": 1.0596565008163452, + "learning_rate": 1.4593307318878286e-05, + "loss": 0.9518, + "step": 9795 + }, + { + "epoch": 0.6730795657551188, + "grad_norm": 1.0735567808151245, + "learning_rate": 1.4587855510509796e-05, + "loss": 0.9751, + "step": 9796 + }, + { + "epoch": 0.6731482753882094, + "grad_norm": 1.0404331684112549, + "learning_rate": 1.4582404301145631e-05, + "loss": 0.9348, + "step": 9797 + }, + { + "epoch": 0.6732169850213, + "grad_norm": 1.0055757761001587, + "learning_rate": 1.4576953691099404e-05, + "loss": 0.9109, + "step": 9798 + }, + { + "epoch": 0.6732856946543906, + "grad_norm": 1.0146472454071045, + "learning_rate": 1.4571503680684664e-05, + "loss": 0.9232, + "step": 9799 + }, + { + "epoch": 0.6733544042874811, + "grad_norm": 1.0385620594024658, + "learning_rate": 1.456605427021496e-05, + "loss": 0.9299, + "step": 9800 + }, + { + "epoch": 0.6734231139205716, + "grad_norm": 1.067636251449585, + "learning_rate": 1.4560605460003801e-05, + "loss": 0.9586, + "step": 9801 + }, + { + "epoch": 0.6734918235536622, + "grad_norm": 1.0774990320205688, + "learning_rate": 1.4555157250364636e-05, + "loss": 0.9141, + "step": 9802 + }, + { + "epoch": 0.6735605331867528, + "grad_norm": 1.0440915822982788, + "learning_rate": 1.4549709641610907e-05, + "loss": 0.9128, + "step": 9803 + }, + { + "epoch": 0.6736292428198434, + "grad_norm": 1.099121332168579, + "learning_rate": 1.4544262634056011e-05, + "loss": 0.9677, + "step": 9804 + }, + { + "epoch": 0.6736979524529338, + "grad_norm": 1.066196322441101, + "learning_rate": 1.4538816228013308e-05, + "loss": 0.9234, + "step": 9805 + }, + { + "epoch": 0.6737666620860244, + "grad_norm": 1.076225757598877, + "learning_rate": 1.4533370423796139e-05, + "loss": 0.971, + "step": 9806 + }, + { + "epoch": 0.673835371719115, + "grad_norm": 1.0097370147705078, + "learning_rate": 1.4527925221717776e-05, + "loss": 0.9275, + "step": 9807 + }, + { + "epoch": 0.6739040813522056, + "grad_norm": 1.1028448343276978, + "learning_rate": 1.4522480622091494e-05, + "loss": 0.948, + "step": 9808 + }, + { + "epoch": 0.6739727909852962, + "grad_norm": 1.06082022190094, + "learning_rate": 1.451703662523051e-05, + "loss": 0.968, + "step": 9809 + }, + { + "epoch": 0.6740415006183867, + "grad_norm": 1.0585206747055054, + "learning_rate": 1.4511593231448017e-05, + "loss": 0.9537, + "step": 9810 + }, + { + "epoch": 0.6741102102514772, + "grad_norm": 1.081167221069336, + "learning_rate": 1.4506150441057181e-05, + "loss": 0.9545, + "step": 9811 + }, + { + "epoch": 0.6741789198845678, + "grad_norm": 1.1179332733154297, + "learning_rate": 1.4500708254371104e-05, + "loss": 0.9953, + "step": 9812 + }, + { + "epoch": 0.6742476295176584, + "grad_norm": 1.0352346897125244, + "learning_rate": 1.4495266671702879e-05, + "loss": 0.9326, + "step": 9813 + }, + { + "epoch": 0.674316339150749, + "grad_norm": 1.0343797206878662, + "learning_rate": 1.4489825693365555e-05, + "loss": 0.956, + "step": 9814 + }, + { + "epoch": 0.6743850487838395, + "grad_norm": 1.0542852878570557, + "learning_rate": 1.448438531967215e-05, + "loss": 0.9574, + "step": 9815 + }, + { + "epoch": 0.67445375841693, + "grad_norm": 1.0685688257217407, + "learning_rate": 1.4478945550935658e-05, + "loss": 0.9507, + "step": 9816 + }, + { + "epoch": 0.6745224680500206, + "grad_norm": 1.0835617780685425, + "learning_rate": 1.4473506387469002e-05, + "loss": 0.9581, + "step": 9817 + }, + { + "epoch": 0.6745911776831112, + "grad_norm": 1.0836971998214722, + "learning_rate": 1.4468067829585108e-05, + "loss": 0.9336, + "step": 9818 + }, + { + "epoch": 0.6746598873162017, + "grad_norm": 1.1054803133010864, + "learning_rate": 1.4462629877596846e-05, + "loss": 0.9825, + "step": 9819 + }, + { + "epoch": 0.6747285969492923, + "grad_norm": 0.9972478151321411, + "learning_rate": 1.4457192531817074e-05, + "loss": 0.921, + "step": 9820 + }, + { + "epoch": 0.6747973065823829, + "grad_norm": 1.0624303817749023, + "learning_rate": 1.4451755792558578e-05, + "loss": 0.9431, + "step": 9821 + }, + { + "epoch": 0.6748660162154734, + "grad_norm": 1.0164567232131958, + "learning_rate": 1.444631966013415e-05, + "loss": 0.9069, + "step": 9822 + }, + { + "epoch": 0.674934725848564, + "grad_norm": 1.1025251150131226, + "learning_rate": 1.4440884134856506e-05, + "loss": 0.943, + "step": 9823 + }, + { + "epoch": 0.6750034354816545, + "grad_norm": 1.1311540603637695, + "learning_rate": 1.443544921703836e-05, + "loss": 0.964, + "step": 9824 + }, + { + "epoch": 0.6750721451147451, + "grad_norm": 1.0150939226150513, + "learning_rate": 1.4430014906992379e-05, + "loss": 0.9377, + "step": 9825 + }, + { + "epoch": 0.6751408547478357, + "grad_norm": 1.0499811172485352, + "learning_rate": 1.4424581205031202e-05, + "loss": 0.9248, + "step": 9826 + }, + { + "epoch": 0.6752095643809262, + "grad_norm": 1.1364659070968628, + "learning_rate": 1.4419148111467415e-05, + "loss": 0.9637, + "step": 9827 + }, + { + "epoch": 0.6752782740140167, + "grad_norm": 1.1012448072433472, + "learning_rate": 1.4413715626613583e-05, + "loss": 0.9181, + "step": 9828 + }, + { + "epoch": 0.6753469836471073, + "grad_norm": 1.1108556985855103, + "learning_rate": 1.4408283750782236e-05, + "loss": 0.951, + "step": 9829 + }, + { + "epoch": 0.6754156932801979, + "grad_norm": 1.0204650163650513, + "learning_rate": 1.4402852484285867e-05, + "loss": 0.967, + "step": 9830 + }, + { + "epoch": 0.6754844029132885, + "grad_norm": 1.056457281112671, + "learning_rate": 1.4397421827436946e-05, + "loss": 0.9328, + "step": 9831 + }, + { + "epoch": 0.675553112546379, + "grad_norm": 1.0379619598388672, + "learning_rate": 1.439199178054787e-05, + "loss": 0.8915, + "step": 9832 + }, + { + "epoch": 0.6756218221794695, + "grad_norm": 1.0543686151504517, + "learning_rate": 1.4386562343931037e-05, + "loss": 1.01, + "step": 9833 + }, + { + "epoch": 0.6756905318125601, + "grad_norm": 1.1120225191116333, + "learning_rate": 1.4381133517898804e-05, + "loss": 0.9581, + "step": 9834 + }, + { + "epoch": 0.6757592414456507, + "grad_norm": 1.082160472869873, + "learning_rate": 1.4375705302763486e-05, + "loss": 0.9332, + "step": 9835 + }, + { + "epoch": 0.6758279510787413, + "grad_norm": 1.0124506950378418, + "learning_rate": 1.437027769883737e-05, + "loss": 0.8969, + "step": 9836 + }, + { + "epoch": 0.6758966607118319, + "grad_norm": 1.070431113243103, + "learning_rate": 1.4364850706432692e-05, + "loss": 0.9496, + "step": 9837 + }, + { + "epoch": 0.6759653703449223, + "grad_norm": 1.1459230184555054, + "learning_rate": 1.4359424325861665e-05, + "loss": 0.9604, + "step": 9838 + }, + { + "epoch": 0.6760340799780129, + "grad_norm": 1.0610145330429077, + "learning_rate": 1.4353998557436465e-05, + "loss": 0.9625, + "step": 9839 + }, + { + "epoch": 0.6761027896111035, + "grad_norm": 1.0172457695007324, + "learning_rate": 1.4348573401469242e-05, + "loss": 0.9423, + "step": 9840 + }, + { + "epoch": 0.6761714992441941, + "grad_norm": 1.0803329944610596, + "learning_rate": 1.4343148858272103e-05, + "loss": 0.9598, + "step": 9841 + }, + { + "epoch": 0.6762402088772846, + "grad_norm": 1.09394371509552, + "learning_rate": 1.4337724928157108e-05, + "loss": 0.9035, + "step": 9842 + }, + { + "epoch": 0.6763089185103751, + "grad_norm": 1.0315351486206055, + "learning_rate": 1.4332301611436294e-05, + "loss": 0.9297, + "step": 9843 + }, + { + "epoch": 0.6763776281434657, + "grad_norm": 1.0892926454544067, + "learning_rate": 1.432687890842167e-05, + "loss": 0.9456, + "step": 9844 + }, + { + "epoch": 0.6764463377765563, + "grad_norm": 1.1215534210205078, + "learning_rate": 1.4321456819425189e-05, + "loss": 0.9684, + "step": 9845 + }, + { + "epoch": 0.6765150474096469, + "grad_norm": 1.0731594562530518, + "learning_rate": 1.4316035344758785e-05, + "loss": 0.8973, + "step": 9846 + }, + { + "epoch": 0.6765837570427374, + "grad_norm": 1.0741324424743652, + "learning_rate": 1.4310614484734363e-05, + "loss": 0.9319, + "step": 9847 + }, + { + "epoch": 0.6766524666758279, + "grad_norm": 1.041144847869873, + "learning_rate": 1.4305194239663763e-05, + "loss": 0.9296, + "step": 9848 + }, + { + "epoch": 0.6767211763089185, + "grad_norm": 1.0528687238693237, + "learning_rate": 1.4299774609858818e-05, + "loss": 0.9445, + "step": 9849 + }, + { + "epoch": 0.6767898859420091, + "grad_norm": 1.059277057647705, + "learning_rate": 1.4294355595631312e-05, + "loss": 0.9105, + "step": 9850 + }, + { + "epoch": 0.6768585955750996, + "grad_norm": 1.0819824934005737, + "learning_rate": 1.4288937197293004e-05, + "loss": 0.9141, + "step": 9851 + }, + { + "epoch": 0.6769273052081902, + "grad_norm": 1.0445505380630493, + "learning_rate": 1.4283519415155616e-05, + "loss": 0.9428, + "step": 9852 + }, + { + "epoch": 0.6769960148412808, + "grad_norm": 1.01947021484375, + "learning_rate": 1.4278102249530814e-05, + "loss": 0.9148, + "step": 9853 + }, + { + "epoch": 0.6770647244743713, + "grad_norm": 1.1448924541473389, + "learning_rate": 1.427268570073025e-05, + "loss": 0.9741, + "step": 9854 + }, + { + "epoch": 0.6771334341074619, + "grad_norm": 1.0706193447113037, + "learning_rate": 1.4267269769065538e-05, + "loss": 0.9719, + "step": 9855 + }, + { + "epoch": 0.6772021437405524, + "grad_norm": 1.0965685844421387, + "learning_rate": 1.4261854454848253e-05, + "loss": 0.9646, + "step": 9856 + }, + { + "epoch": 0.677270853373643, + "grad_norm": 0.9968788027763367, + "learning_rate": 1.4256439758389944e-05, + "loss": 0.9273, + "step": 9857 + }, + { + "epoch": 0.6773395630067336, + "grad_norm": 1.1182136535644531, + "learning_rate": 1.4251025680002092e-05, + "loss": 0.9676, + "step": 9858 + }, + { + "epoch": 0.6774082726398241, + "grad_norm": 1.0362474918365479, + "learning_rate": 1.424561221999618e-05, + "loss": 0.9428, + "step": 9859 + }, + { + "epoch": 0.6774769822729146, + "grad_norm": 1.084301233291626, + "learning_rate": 1.4240199378683642e-05, + "loss": 0.8998, + "step": 9860 + }, + { + "epoch": 0.6775456919060052, + "grad_norm": 1.0430705547332764, + "learning_rate": 1.423478715637587e-05, + "loss": 0.9389, + "step": 9861 + }, + { + "epoch": 0.6776144015390958, + "grad_norm": 1.124014973640442, + "learning_rate": 1.4229375553384239e-05, + "loss": 0.9417, + "step": 9862 + }, + { + "epoch": 0.6776831111721864, + "grad_norm": 1.09843909740448, + "learning_rate": 1.4223964570020059e-05, + "loss": 0.9184, + "step": 9863 + }, + { + "epoch": 0.6777518208052769, + "grad_norm": 1.0656840801239014, + "learning_rate": 1.4218554206594623e-05, + "loss": 0.9601, + "step": 9864 + }, + { + "epoch": 0.6778205304383674, + "grad_norm": 1.0277411937713623, + "learning_rate": 1.4213144463419193e-05, + "loss": 0.9482, + "step": 9865 + }, + { + "epoch": 0.677889240071458, + "grad_norm": 1.077882170677185, + "learning_rate": 1.4207735340804989e-05, + "loss": 0.9497, + "step": 9866 + }, + { + "epoch": 0.6779579497045486, + "grad_norm": 1.034564733505249, + "learning_rate": 1.4202326839063184e-05, + "loss": 0.9453, + "step": 9867 + }, + { + "epoch": 0.6780266593376392, + "grad_norm": 1.0941283702850342, + "learning_rate": 1.4196918958504935e-05, + "loss": 0.9682, + "step": 9868 + }, + { + "epoch": 0.6780953689707298, + "grad_norm": 1.1420485973358154, + "learning_rate": 1.4191511699441357e-05, + "loss": 0.9502, + "step": 9869 + }, + { + "epoch": 0.6781640786038202, + "grad_norm": 1.0730211734771729, + "learning_rate": 1.4186105062183513e-05, + "loss": 0.9221, + "step": 9870 + }, + { + "epoch": 0.6782327882369108, + "grad_norm": 1.0598208904266357, + "learning_rate": 1.418069904704245e-05, + "loss": 0.9845, + "step": 9871 + }, + { + "epoch": 0.6783014978700014, + "grad_norm": 1.0651814937591553, + "learning_rate": 1.4175293654329186e-05, + "loss": 0.9226, + "step": 9872 + }, + { + "epoch": 0.678370207503092, + "grad_norm": 1.0072981119155884, + "learning_rate": 1.4169888884354666e-05, + "loss": 0.9281, + "step": 9873 + }, + { + "epoch": 0.6784389171361825, + "grad_norm": 1.1340142488479614, + "learning_rate": 1.4164484737429838e-05, + "loss": 0.9917, + "step": 9874 + }, + { + "epoch": 0.678507626769273, + "grad_norm": 1.0543406009674072, + "learning_rate": 1.4159081213865593e-05, + "loss": 0.9808, + "step": 9875 + }, + { + "epoch": 0.6785763364023636, + "grad_norm": 1.0826665163040161, + "learning_rate": 1.4153678313972795e-05, + "loss": 0.9678, + "step": 9876 + }, + { + "epoch": 0.6786450460354542, + "grad_norm": 1.0255457162857056, + "learning_rate": 1.414827603806228e-05, + "loss": 0.92, + "step": 9877 + }, + { + "epoch": 0.6787137556685447, + "grad_norm": 1.1026859283447266, + "learning_rate": 1.4142874386444822e-05, + "loss": 0.9276, + "step": 9878 + }, + { + "epoch": 0.6787824653016353, + "grad_norm": 1.05731999874115, + "learning_rate": 1.4137473359431178e-05, + "loss": 0.9486, + "step": 9879 + }, + { + "epoch": 0.6788511749347258, + "grad_norm": 1.0571603775024414, + "learning_rate": 1.4132072957332068e-05, + "loss": 0.9545, + "step": 9880 + }, + { + "epoch": 0.6789198845678164, + "grad_norm": 1.0815380811691284, + "learning_rate": 1.4126673180458175e-05, + "loss": 0.958, + "step": 9881 + }, + { + "epoch": 0.678988594200907, + "grad_norm": 1.062762975692749, + "learning_rate": 1.4121274029120155e-05, + "loss": 0.9756, + "step": 9882 + }, + { + "epoch": 0.6790573038339975, + "grad_norm": 1.1128430366516113, + "learning_rate": 1.4115875503628595e-05, + "loss": 0.9667, + "step": 9883 + }, + { + "epoch": 0.6791260134670881, + "grad_norm": 1.038321614265442, + "learning_rate": 1.4110477604294081e-05, + "loss": 0.9466, + "step": 9884 + }, + { + "epoch": 0.6791947231001787, + "grad_norm": 1.055408000946045, + "learning_rate": 1.410508033142715e-05, + "loss": 0.912, + "step": 9885 + }, + { + "epoch": 0.6792634327332692, + "grad_norm": 1.0614259243011475, + "learning_rate": 1.4099683685338308e-05, + "loss": 0.9545, + "step": 9886 + }, + { + "epoch": 0.6793321423663597, + "grad_norm": 1.0728437900543213, + "learning_rate": 1.4094287666338024e-05, + "loss": 0.9285, + "step": 9887 + }, + { + "epoch": 0.6794008519994503, + "grad_norm": 1.0849145650863647, + "learning_rate": 1.4088892274736715e-05, + "loss": 0.9539, + "step": 9888 + }, + { + "epoch": 0.6794695616325409, + "grad_norm": 1.1882696151733398, + "learning_rate": 1.408349751084478e-05, + "loss": 0.9575, + "step": 9889 + }, + { + "epoch": 0.6795382712656315, + "grad_norm": 1.0608775615692139, + "learning_rate": 1.4078103374972573e-05, + "loss": 0.9566, + "step": 9890 + }, + { + "epoch": 0.679606980898722, + "grad_norm": 1.0937247276306152, + "learning_rate": 1.4072709867430429e-05, + "loss": 1.0274, + "step": 9891 + }, + { + "epoch": 0.6796756905318125, + "grad_norm": 1.1284302473068237, + "learning_rate": 1.4067316988528617e-05, + "loss": 0.9154, + "step": 9892 + }, + { + "epoch": 0.6797444001649031, + "grad_norm": 1.0394936800003052, + "learning_rate": 1.4061924738577403e-05, + "loss": 0.9569, + "step": 9893 + }, + { + "epoch": 0.6798131097979937, + "grad_norm": 1.0633758306503296, + "learning_rate": 1.4056533117886978e-05, + "loss": 0.9489, + "step": 9894 + }, + { + "epoch": 0.6798818194310843, + "grad_norm": 1.1140398979187012, + "learning_rate": 1.405114212676753e-05, + "loss": 0.9369, + "step": 9895 + }, + { + "epoch": 0.6799505290641747, + "grad_norm": 1.0476384162902832, + "learning_rate": 1.4045751765529197e-05, + "loss": 0.9263, + "step": 9896 + }, + { + "epoch": 0.6800192386972653, + "grad_norm": 1.010194182395935, + "learning_rate": 1.4040362034482101e-05, + "loss": 0.928, + "step": 9897 + }, + { + "epoch": 0.6800879483303559, + "grad_norm": 1.0343812704086304, + "learning_rate": 1.403497293393628e-05, + "loss": 0.911, + "step": 9898 + }, + { + "epoch": 0.6801566579634465, + "grad_norm": 1.0544252395629883, + "learning_rate": 1.4029584464201783e-05, + "loss": 0.9314, + "step": 9899 + }, + { + "epoch": 0.6802253675965371, + "grad_norm": 1.027385950088501, + "learning_rate": 1.4024196625588603e-05, + "loss": 0.8972, + "step": 9900 + }, + { + "epoch": 0.6802940772296276, + "grad_norm": 1.0672171115875244, + "learning_rate": 1.4018809418406698e-05, + "loss": 0.9413, + "step": 9901 + }, + { + "epoch": 0.6803627868627181, + "grad_norm": 1.0565165281295776, + "learning_rate": 1.4013422842966e-05, + "loss": 0.9354, + "step": 9902 + }, + { + "epoch": 0.6804314964958087, + "grad_norm": 1.0548560619354248, + "learning_rate": 1.4008036899576382e-05, + "loss": 0.9471, + "step": 9903 + }, + { + "epoch": 0.6805002061288993, + "grad_norm": 1.0265611410140991, + "learning_rate": 1.4002651588547694e-05, + "loss": 0.9361, + "step": 9904 + }, + { + "epoch": 0.6805689157619899, + "grad_norm": 1.0600091218948364, + "learning_rate": 1.3997266910189754e-05, + "loss": 0.9363, + "step": 9905 + }, + { + "epoch": 0.6806376253950804, + "grad_norm": 1.0797358751296997, + "learning_rate": 1.3991882864812345e-05, + "loss": 0.9451, + "step": 9906 + }, + { + "epoch": 0.6807063350281709, + "grad_norm": 1.0592069625854492, + "learning_rate": 1.39864994527252e-05, + "loss": 0.9405, + "step": 9907 + }, + { + "epoch": 0.6807750446612615, + "grad_norm": 1.021193265914917, + "learning_rate": 1.3981116674238037e-05, + "loss": 0.9296, + "step": 9908 + }, + { + "epoch": 0.6808437542943521, + "grad_norm": 1.0793302059173584, + "learning_rate": 1.39757345296605e-05, + "loss": 0.9348, + "step": 9909 + }, + { + "epoch": 0.6809124639274426, + "grad_norm": 1.0437065362930298, + "learning_rate": 1.3970353019302238e-05, + "loss": 0.8988, + "step": 9910 + }, + { + "epoch": 0.6809811735605332, + "grad_norm": 1.0870474576950073, + "learning_rate": 1.3964972143472837e-05, + "loss": 0.9675, + "step": 9911 + }, + { + "epoch": 0.6810498831936237, + "grad_norm": 1.1112221479415894, + "learning_rate": 1.3959591902481867e-05, + "loss": 0.987, + "step": 9912 + }, + { + "epoch": 0.6811185928267143, + "grad_norm": 1.0595511198043823, + "learning_rate": 1.3954212296638847e-05, + "loss": 0.9336, + "step": 9913 + }, + { + "epoch": 0.6811873024598049, + "grad_norm": 1.0678778886795044, + "learning_rate": 1.3948833326253252e-05, + "loss": 0.9562, + "step": 9914 + }, + { + "epoch": 0.6812560120928954, + "grad_norm": 1.1243958473205566, + "learning_rate": 1.3943454991634545e-05, + "loss": 0.9982, + "step": 9915 + }, + { + "epoch": 0.681324721725986, + "grad_norm": 1.075380563735962, + "learning_rate": 1.3938077293092123e-05, + "loss": 0.907, + "step": 9916 + }, + { + "epoch": 0.6813934313590766, + "grad_norm": 1.0228816270828247, + "learning_rate": 1.3932700230935367e-05, + "loss": 0.9092, + "step": 9917 + }, + { + "epoch": 0.6814621409921671, + "grad_norm": 1.0472288131713867, + "learning_rate": 1.392732380547363e-05, + "loss": 0.9108, + "step": 9918 + }, + { + "epoch": 0.6815308506252576, + "grad_norm": 1.1159783601760864, + "learning_rate": 1.3921948017016193e-05, + "loss": 0.964, + "step": 9919 + }, + { + "epoch": 0.6815995602583482, + "grad_norm": 1.0535534620285034, + "learning_rate": 1.3916572865872332e-05, + "loss": 0.9077, + "step": 9920 + }, + { + "epoch": 0.6816682698914388, + "grad_norm": 1.055747628211975, + "learning_rate": 1.3911198352351275e-05, + "loss": 0.9442, + "step": 9921 + }, + { + "epoch": 0.6817369795245294, + "grad_norm": 1.020817756652832, + "learning_rate": 1.3905824476762219e-05, + "loss": 0.9495, + "step": 9922 + }, + { + "epoch": 0.6818056891576199, + "grad_norm": 1.0707547664642334, + "learning_rate": 1.3900451239414323e-05, + "loss": 0.9034, + "step": 9923 + }, + { + "epoch": 0.6818743987907104, + "grad_norm": 1.1148624420166016, + "learning_rate": 1.389507864061669e-05, + "loss": 0.971, + "step": 9924 + }, + { + "epoch": 0.681943108423801, + "grad_norm": 1.1359556913375854, + "learning_rate": 1.3889706680678414e-05, + "loss": 0.9642, + "step": 9925 + }, + { + "epoch": 0.6820118180568916, + "grad_norm": 1.0595487356185913, + "learning_rate": 1.3884335359908535e-05, + "loss": 0.9109, + "step": 9926 + }, + { + "epoch": 0.6820805276899822, + "grad_norm": 1.0359642505645752, + "learning_rate": 1.3878964678616069e-05, + "loss": 0.9358, + "step": 9927 + }, + { + "epoch": 0.6821492373230726, + "grad_norm": 1.0101878643035889, + "learning_rate": 1.3873594637109988e-05, + "loss": 0.9602, + "step": 9928 + }, + { + "epoch": 0.6822179469561632, + "grad_norm": 1.0070090293884277, + "learning_rate": 1.3868225235699217e-05, + "loss": 0.9276, + "step": 9929 + }, + { + "epoch": 0.6822866565892538, + "grad_norm": 1.0377146005630493, + "learning_rate": 1.3862856474692665e-05, + "loss": 0.911, + "step": 9930 + }, + { + "epoch": 0.6823553662223444, + "grad_norm": 1.0850478410720825, + "learning_rate": 1.3857488354399184e-05, + "loss": 0.9431, + "step": 9931 + }, + { + "epoch": 0.682424075855435, + "grad_norm": 1.0409212112426758, + "learning_rate": 1.3852120875127606e-05, + "loss": 0.9369, + "step": 9932 + }, + { + "epoch": 0.6824927854885255, + "grad_norm": 1.0893747806549072, + "learning_rate": 1.3846754037186727e-05, + "loss": 0.9617, + "step": 9933 + }, + { + "epoch": 0.682561495121616, + "grad_norm": 1.0558363199234009, + "learning_rate": 1.3841387840885276e-05, + "loss": 0.9094, + "step": 9934 + }, + { + "epoch": 0.6826302047547066, + "grad_norm": 1.043721318244934, + "learning_rate": 1.383602228653198e-05, + "loss": 0.9471, + "step": 9935 + }, + { + "epoch": 0.6826989143877972, + "grad_norm": 1.0859349966049194, + "learning_rate": 1.3830657374435512e-05, + "loss": 0.9188, + "step": 9936 + }, + { + "epoch": 0.6827676240208878, + "grad_norm": 1.1050572395324707, + "learning_rate": 1.382529310490453e-05, + "loss": 0.9502, + "step": 9937 + }, + { + "epoch": 0.6828363336539783, + "grad_norm": 1.0509904623031616, + "learning_rate": 1.381992947824761e-05, + "loss": 0.924, + "step": 9938 + }, + { + "epoch": 0.6829050432870688, + "grad_norm": 1.1206679344177246, + "learning_rate": 1.3814566494773329e-05, + "loss": 0.9631, + "step": 9939 + }, + { + "epoch": 0.6829737529201594, + "grad_norm": 1.0238085985183716, + "learning_rate": 1.3809204154790229e-05, + "loss": 0.9549, + "step": 9940 + }, + { + "epoch": 0.68304246255325, + "grad_norm": 1.0077166557312012, + "learning_rate": 1.380384245860678e-05, + "loss": 0.9198, + "step": 9941 + }, + { + "epoch": 0.6831111721863405, + "grad_norm": 1.0834643840789795, + "learning_rate": 1.3798481406531447e-05, + "loss": 0.9037, + "step": 9942 + }, + { + "epoch": 0.6831798818194311, + "grad_norm": 1.0233262777328491, + "learning_rate": 1.3793120998872658e-05, + "loss": 0.9175, + "step": 9943 + }, + { + "epoch": 0.6832485914525216, + "grad_norm": 1.1011801958084106, + "learning_rate": 1.3787761235938776e-05, + "loss": 0.9314, + "step": 9944 + }, + { + "epoch": 0.6833173010856122, + "grad_norm": 1.0358115434646606, + "learning_rate": 1.3782402118038154e-05, + "loss": 0.9593, + "step": 9945 + }, + { + "epoch": 0.6833860107187028, + "grad_norm": 1.055061936378479, + "learning_rate": 1.3777043645479099e-05, + "loss": 0.9781, + "step": 9946 + }, + { + "epoch": 0.6834547203517933, + "grad_norm": 1.1104766130447388, + "learning_rate": 1.377168581856988e-05, + "loss": 1.0045, + "step": 9947 + }, + { + "epoch": 0.6835234299848839, + "grad_norm": 1.0500905513763428, + "learning_rate": 1.3766328637618736e-05, + "loss": 0.9321, + "step": 9948 + }, + { + "epoch": 0.6835921396179745, + "grad_norm": 1.0301849842071533, + "learning_rate": 1.376097210293385e-05, + "loss": 0.8832, + "step": 9949 + }, + { + "epoch": 0.683660849251065, + "grad_norm": 1.0780112743377686, + "learning_rate": 1.3755616214823386e-05, + "loss": 0.9377, + "step": 9950 + }, + { + "epoch": 0.6837295588841555, + "grad_norm": 1.1531747579574585, + "learning_rate": 1.3750260973595463e-05, + "loss": 0.988, + "step": 9951 + }, + { + "epoch": 0.6837982685172461, + "grad_norm": 1.065659523010254, + "learning_rate": 1.3744906379558165e-05, + "loss": 0.9241, + "step": 9952 + }, + { + "epoch": 0.6838669781503367, + "grad_norm": 1.0430439710617065, + "learning_rate": 1.3739552433019553e-05, + "loss": 0.9348, + "step": 9953 + }, + { + "epoch": 0.6839356877834273, + "grad_norm": 1.0835860967636108, + "learning_rate": 1.3734199134287614e-05, + "loss": 0.9518, + "step": 9954 + }, + { + "epoch": 0.6840043974165178, + "grad_norm": 1.0384414196014404, + "learning_rate": 1.3728846483670332e-05, + "loss": 0.9325, + "step": 9955 + }, + { + "epoch": 0.6840731070496083, + "grad_norm": 1.0959677696228027, + "learning_rate": 1.3723494481475635e-05, + "loss": 0.9349, + "step": 9956 + }, + { + "epoch": 0.6841418166826989, + "grad_norm": 1.0240474939346313, + "learning_rate": 1.3718143128011426e-05, + "loss": 0.9179, + "step": 9957 + }, + { + "epoch": 0.6842105263157895, + "grad_norm": 1.04267418384552, + "learning_rate": 1.3712792423585575e-05, + "loss": 0.9317, + "step": 9958 + }, + { + "epoch": 0.6842792359488801, + "grad_norm": 1.0366648435592651, + "learning_rate": 1.3707442368505885e-05, + "loss": 0.9462, + "step": 9959 + }, + { + "epoch": 0.6843479455819705, + "grad_norm": 1.1315028667449951, + "learning_rate": 1.370209296308015e-05, + "loss": 1.013, + "step": 9960 + }, + { + "epoch": 0.6844166552150611, + "grad_norm": 0.9911446571350098, + "learning_rate": 1.3696744207616113e-05, + "loss": 0.9359, + "step": 9961 + }, + { + "epoch": 0.6844853648481517, + "grad_norm": 1.0528935194015503, + "learning_rate": 1.3691396102421506e-05, + "loss": 0.9467, + "step": 9962 + }, + { + "epoch": 0.6845540744812423, + "grad_norm": 1.053894281387329, + "learning_rate": 1.368604864780397e-05, + "loss": 0.9261, + "step": 9963 + }, + { + "epoch": 0.6846227841143329, + "grad_norm": 1.065690517425537, + "learning_rate": 1.3680701844071158e-05, + "loss": 0.9077, + "step": 9964 + }, + { + "epoch": 0.6846914937474234, + "grad_norm": 1.0448973178863525, + "learning_rate": 1.3675355691530683e-05, + "loss": 0.9304, + "step": 9965 + }, + { + "epoch": 0.6847602033805139, + "grad_norm": 1.107424259185791, + "learning_rate": 1.3670010190490073e-05, + "loss": 0.9599, + "step": 9966 + }, + { + "epoch": 0.6848289130136045, + "grad_norm": 1.0386308431625366, + "learning_rate": 1.3664665341256875e-05, + "loss": 0.9169, + "step": 9967 + }, + { + "epoch": 0.6848976226466951, + "grad_norm": 1.0199452638626099, + "learning_rate": 1.3659321144138566e-05, + "loss": 0.9274, + "step": 9968 + }, + { + "epoch": 0.6849663322797857, + "grad_norm": 1.1128604412078857, + "learning_rate": 1.3653977599442602e-05, + "loss": 0.9395, + "step": 9969 + }, + { + "epoch": 0.6850350419128762, + "grad_norm": 1.0525315999984741, + "learning_rate": 1.3648634707476381e-05, + "loss": 0.9117, + "step": 9970 + }, + { + "epoch": 0.6851037515459667, + "grad_norm": 1.0305957794189453, + "learning_rate": 1.364329246854729e-05, + "loss": 0.9527, + "step": 9971 + }, + { + "epoch": 0.6851724611790573, + "grad_norm": 1.0484294891357422, + "learning_rate": 1.3637950882962652e-05, + "loss": 0.9849, + "step": 9972 + }, + { + "epoch": 0.6852411708121479, + "grad_norm": 1.0616588592529297, + "learning_rate": 1.3632609951029774e-05, + "loss": 0.9262, + "step": 9973 + }, + { + "epoch": 0.6853098804452384, + "grad_norm": 1.0625346899032593, + "learning_rate": 1.3627269673055925e-05, + "loss": 0.9458, + "step": 9974 + }, + { + "epoch": 0.685378590078329, + "grad_norm": 1.0217620134353638, + "learning_rate": 1.3621930049348308e-05, + "loss": 0.9387, + "step": 9975 + }, + { + "epoch": 0.6854472997114195, + "grad_norm": 1.125374436378479, + "learning_rate": 1.3616591080214115e-05, + "loss": 0.9613, + "step": 9976 + }, + { + "epoch": 0.6855160093445101, + "grad_norm": 1.1505721807479858, + "learning_rate": 1.3611252765960496e-05, + "loss": 0.9277, + "step": 9977 + }, + { + "epoch": 0.6855847189776006, + "grad_norm": 0.9951162338256836, + "learning_rate": 1.3605915106894568e-05, + "loss": 0.937, + "step": 9978 + }, + { + "epoch": 0.6856534286106912, + "grad_norm": 1.1081562042236328, + "learning_rate": 1.36005781033234e-05, + "loss": 0.9527, + "step": 9979 + }, + { + "epoch": 0.6857221382437818, + "grad_norm": 1.0042976140975952, + "learning_rate": 1.3595241755554014e-05, + "loss": 0.9066, + "step": 9980 + }, + { + "epoch": 0.6857908478768724, + "grad_norm": 1.0760669708251953, + "learning_rate": 1.3589906063893418e-05, + "loss": 0.9726, + "step": 9981 + }, + { + "epoch": 0.6858595575099629, + "grad_norm": 1.0220143795013428, + "learning_rate": 1.3584571028648568e-05, + "loss": 0.9287, + "step": 9982 + }, + { + "epoch": 0.6859282671430534, + "grad_norm": 1.0361909866333008, + "learning_rate": 1.3579236650126387e-05, + "loss": 0.9589, + "step": 9983 + }, + { + "epoch": 0.685996976776144, + "grad_norm": 1.0258262157440186, + "learning_rate": 1.3573902928633763e-05, + "loss": 0.9679, + "step": 9984 + }, + { + "epoch": 0.6860656864092346, + "grad_norm": 1.17508065700531, + "learning_rate": 1.3568569864477531e-05, + "loss": 0.9475, + "step": 9985 + }, + { + "epoch": 0.6861343960423252, + "grad_norm": 1.0053454637527466, + "learning_rate": 1.3563237457964506e-05, + "loss": 0.9569, + "step": 9986 + }, + { + "epoch": 0.6862031056754156, + "grad_norm": 1.0415973663330078, + "learning_rate": 1.3557905709401464e-05, + "loss": 0.9153, + "step": 9987 + }, + { + "epoch": 0.6862718153085062, + "grad_norm": 1.101430892944336, + "learning_rate": 1.3552574619095121e-05, + "loss": 0.9391, + "step": 9988 + }, + { + "epoch": 0.6863405249415968, + "grad_norm": 1.0099087953567505, + "learning_rate": 1.3547244187352192e-05, + "loss": 0.9248, + "step": 9989 + }, + { + "epoch": 0.6864092345746874, + "grad_norm": 1.0486844778060913, + "learning_rate": 1.354191441447931e-05, + "loss": 0.9341, + "step": 9990 + }, + { + "epoch": 0.686477944207778, + "grad_norm": 1.121656060218811, + "learning_rate": 1.3536585300783106e-05, + "loss": 0.951, + "step": 9991 + }, + { + "epoch": 0.6865466538408684, + "grad_norm": 1.0474376678466797, + "learning_rate": 1.3531256846570159e-05, + "loss": 0.9734, + "step": 9992 + }, + { + "epoch": 0.686615363473959, + "grad_norm": 1.0935112237930298, + "learning_rate": 1.3525929052147012e-05, + "loss": 0.9449, + "step": 9993 + }, + { + "epoch": 0.6866840731070496, + "grad_norm": 1.053495168685913, + "learning_rate": 1.352060191782018e-05, + "loss": 0.9622, + "step": 9994 + }, + { + "epoch": 0.6867527827401402, + "grad_norm": 1.0363267660140991, + "learning_rate": 1.3515275443896114e-05, + "loss": 0.9225, + "step": 9995 + }, + { + "epoch": 0.6868214923732308, + "grad_norm": 1.0584819316864014, + "learning_rate": 1.3509949630681246e-05, + "loss": 0.9685, + "step": 9996 + }, + { + "epoch": 0.6868902020063213, + "grad_norm": 1.0569312572479248, + "learning_rate": 1.350462447848197e-05, + "loss": 0.911, + "step": 9997 + }, + { + "epoch": 0.6869589116394118, + "grad_norm": 1.0997318029403687, + "learning_rate": 1.3499299987604641e-05, + "loss": 0.9501, + "step": 9998 + }, + { + "epoch": 0.6870276212725024, + "grad_norm": 1.058506965637207, + "learning_rate": 1.3493976158355582e-05, + "loss": 0.9306, + "step": 9999 + }, + { + "epoch": 0.687096330905593, + "grad_norm": 1.0625427961349487, + "learning_rate": 1.3488652991041046e-05, + "loss": 0.9467, + "step": 10000 + }, + { + "epoch": 0.6871650405386835, + "grad_norm": 1.1237294673919678, + "learning_rate": 1.3483330485967283e-05, + "loss": 0.9639, + "step": 10001 + }, + { + "epoch": 0.6872337501717741, + "grad_norm": 1.0989130735397339, + "learning_rate": 1.3478008643440498e-05, + "loss": 0.9682, + "step": 10002 + }, + { + "epoch": 0.6873024598048646, + "grad_norm": 1.0363198518753052, + "learning_rate": 1.347268746376685e-05, + "loss": 0.9371, + "step": 10003 + }, + { + "epoch": 0.6873711694379552, + "grad_norm": 1.06019926071167, + "learning_rate": 1.3467366947252474e-05, + "loss": 0.9384, + "step": 10004 + }, + { + "epoch": 0.6874398790710458, + "grad_norm": 1.1651358604431152, + "learning_rate": 1.3462047094203434e-05, + "loss": 0.9909, + "step": 10005 + }, + { + "epoch": 0.6875085887041363, + "grad_norm": 1.0016683340072632, + "learning_rate": 1.3456727904925792e-05, + "loss": 0.9342, + "step": 10006 + }, + { + "epoch": 0.6875772983372269, + "grad_norm": 1.0629944801330566, + "learning_rate": 1.3451409379725554e-05, + "loss": 0.9436, + "step": 10007 + }, + { + "epoch": 0.6876460079703174, + "grad_norm": 1.0983762741088867, + "learning_rate": 1.3446091518908693e-05, + "loss": 0.9545, + "step": 10008 + }, + { + "epoch": 0.687714717603408, + "grad_norm": 1.0748600959777832, + "learning_rate": 1.3440774322781152e-05, + "loss": 0.9541, + "step": 10009 + }, + { + "epoch": 0.6877834272364985, + "grad_norm": 1.0299426317214966, + "learning_rate": 1.343545779164881e-05, + "loss": 0.9253, + "step": 10010 + }, + { + "epoch": 0.6878521368695891, + "grad_norm": 1.036433458328247, + "learning_rate": 1.3430141925817532e-05, + "loss": 0.9595, + "step": 10011 + }, + { + "epoch": 0.6879208465026797, + "grad_norm": 1.0249226093292236, + "learning_rate": 1.342482672559313e-05, + "loss": 0.942, + "step": 10012 + }, + { + "epoch": 0.6879895561357703, + "grad_norm": 1.034467339515686, + "learning_rate": 1.3419512191281387e-05, + "loss": 0.9576, + "step": 10013 + }, + { + "epoch": 0.6880582657688608, + "grad_norm": 1.0003615617752075, + "learning_rate": 1.3414198323188049e-05, + "loss": 0.9213, + "step": 10014 + }, + { + "epoch": 0.6881269754019513, + "grad_norm": 1.1094913482666016, + "learning_rate": 1.3408885121618826e-05, + "loss": 0.9483, + "step": 10015 + }, + { + "epoch": 0.6881956850350419, + "grad_norm": 1.018928050994873, + "learning_rate": 1.340357258687936e-05, + "loss": 0.9021, + "step": 10016 + }, + { + "epoch": 0.6882643946681325, + "grad_norm": 1.0870201587677002, + "learning_rate": 1.3398260719275295e-05, + "loss": 0.9441, + "step": 10017 + }, + { + "epoch": 0.6883331043012231, + "grad_norm": 1.042027473449707, + "learning_rate": 1.3392949519112218e-05, + "loss": 0.9366, + "step": 10018 + }, + { + "epoch": 0.6884018139343135, + "grad_norm": 1.0616165399551392, + "learning_rate": 1.3387638986695681e-05, + "loss": 0.9189, + "step": 10019 + }, + { + "epoch": 0.6884705235674041, + "grad_norm": 1.0511523485183716, + "learning_rate": 1.33823291223312e-05, + "loss": 0.9579, + "step": 10020 + }, + { + "epoch": 0.6885392332004947, + "grad_norm": 1.0461100339889526, + "learning_rate": 1.337701992632423e-05, + "loss": 0.9668, + "step": 10021 + }, + { + "epoch": 0.6886079428335853, + "grad_norm": 1.0583930015563965, + "learning_rate": 1.3371711398980216e-05, + "loss": 0.9212, + "step": 10022 + }, + { + "epoch": 0.6886766524666759, + "grad_norm": 1.05500328540802, + "learning_rate": 1.3366403540604555e-05, + "loss": 0.9321, + "step": 10023 + }, + { + "epoch": 0.6887453620997663, + "grad_norm": 1.0426751375198364, + "learning_rate": 1.3361096351502609e-05, + "loss": 0.9673, + "step": 10024 + }, + { + "epoch": 0.6888140717328569, + "grad_norm": 1.185544729232788, + "learning_rate": 1.3355789831979698e-05, + "loss": 0.9797, + "step": 10025 + }, + { + "epoch": 0.6888827813659475, + "grad_norm": 1.0338943004608154, + "learning_rate": 1.3350483982341091e-05, + "loss": 0.919, + "step": 10026 + }, + { + "epoch": 0.6889514909990381, + "grad_norm": 1.1590653657913208, + "learning_rate": 1.3345178802892038e-05, + "loss": 0.9682, + "step": 10027 + }, + { + "epoch": 0.6890202006321287, + "grad_norm": 1.1114665269851685, + "learning_rate": 1.3339874293937742e-05, + "loss": 0.9915, + "step": 10028 + }, + { + "epoch": 0.6890889102652192, + "grad_norm": 1.0850629806518555, + "learning_rate": 1.3334570455783368e-05, + "loss": 0.9412, + "step": 10029 + }, + { + "epoch": 0.6891576198983097, + "grad_norm": 1.0212265253067017, + "learning_rate": 1.332926728873405e-05, + "loss": 0.8914, + "step": 10030 + }, + { + "epoch": 0.6892263295314003, + "grad_norm": 1.090462565422058, + "learning_rate": 1.3323964793094861e-05, + "loss": 0.9513, + "step": 10031 + }, + { + "epoch": 0.6892950391644909, + "grad_norm": 1.013574481010437, + "learning_rate": 1.3318662969170859e-05, + "loss": 0.9056, + "step": 10032 + }, + { + "epoch": 0.6893637487975814, + "grad_norm": 1.0272841453552246, + "learning_rate": 1.3313361817267061e-05, + "loss": 0.9319, + "step": 10033 + }, + { + "epoch": 0.689432458430672, + "grad_norm": 1.125007152557373, + "learning_rate": 1.3308061337688426e-05, + "loss": 0.9598, + "step": 10034 + }, + { + "epoch": 0.6895011680637625, + "grad_norm": 1.0179619789123535, + "learning_rate": 1.3302761530739893e-05, + "loss": 0.9528, + "step": 10035 + }, + { + "epoch": 0.6895698776968531, + "grad_norm": 1.0158811807632446, + "learning_rate": 1.3297462396726362e-05, + "loss": 0.9306, + "step": 10036 + }, + { + "epoch": 0.6896385873299437, + "grad_norm": 1.0817699432373047, + "learning_rate": 1.3292163935952679e-05, + "loss": 0.9174, + "step": 10037 + }, + { + "epoch": 0.6897072969630342, + "grad_norm": 1.1888208389282227, + "learning_rate": 1.3286866148723665e-05, + "loss": 0.9308, + "step": 10038 + }, + { + "epoch": 0.6897760065961248, + "grad_norm": 1.0630122423171997, + "learning_rate": 1.32815690353441e-05, + "loss": 0.9895, + "step": 10039 + }, + { + "epoch": 0.6898447162292153, + "grad_norm": 1.0336320400238037, + "learning_rate": 1.327627259611873e-05, + "loss": 0.9581, + "step": 10040 + }, + { + "epoch": 0.6899134258623059, + "grad_norm": 1.0955049991607666, + "learning_rate": 1.3270976831352245e-05, + "loss": 0.9468, + "step": 10041 + }, + { + "epoch": 0.6899821354953964, + "grad_norm": 1.0484153032302856, + "learning_rate": 1.3265681741349306e-05, + "loss": 0.9439, + "step": 10042 + }, + { + "epoch": 0.690050845128487, + "grad_norm": 1.0150309801101685, + "learning_rate": 1.3260387326414541e-05, + "loss": 0.9325, + "step": 10043 + }, + { + "epoch": 0.6901195547615776, + "grad_norm": 1.014923095703125, + "learning_rate": 1.3255093586852541e-05, + "loss": 0.9481, + "step": 10044 + }, + { + "epoch": 0.6901882643946682, + "grad_norm": 1.0268523693084717, + "learning_rate": 1.3249800522967848e-05, + "loss": 0.9233, + "step": 10045 + }, + { + "epoch": 0.6902569740277587, + "grad_norm": 1.083986759185791, + "learning_rate": 1.3244508135064962e-05, + "loss": 0.9645, + "step": 10046 + }, + { + "epoch": 0.6903256836608492, + "grad_norm": 1.0260378122329712, + "learning_rate": 1.3239216423448353e-05, + "loss": 0.9107, + "step": 10047 + }, + { + "epoch": 0.6903943932939398, + "grad_norm": 1.0486265420913696, + "learning_rate": 1.3233925388422452e-05, + "loss": 0.9386, + "step": 10048 + }, + { + "epoch": 0.6904631029270304, + "grad_norm": 0.9997297525405884, + "learning_rate": 1.3228635030291653e-05, + "loss": 0.9224, + "step": 10049 + }, + { + "epoch": 0.690531812560121, + "grad_norm": 1.0356979370117188, + "learning_rate": 1.322334534936031e-05, + "loss": 0.936, + "step": 10050 + }, + { + "epoch": 0.6906005221932114, + "grad_norm": 1.0212668180465698, + "learning_rate": 1.3218056345932719e-05, + "loss": 0.9319, + "step": 10051 + }, + { + "epoch": 0.690669231826302, + "grad_norm": 1.0644168853759766, + "learning_rate": 1.3212768020313162e-05, + "loss": 0.9358, + "step": 10052 + }, + { + "epoch": 0.6907379414593926, + "grad_norm": 1.0710089206695557, + "learning_rate": 1.3207480372805875e-05, + "loss": 0.9476, + "step": 10053 + }, + { + "epoch": 0.6908066510924832, + "grad_norm": 1.0605669021606445, + "learning_rate": 1.3202193403715052e-05, + "loss": 0.9632, + "step": 10054 + }, + { + "epoch": 0.6908753607255738, + "grad_norm": 1.1178377866744995, + "learning_rate": 1.3196907113344858e-05, + "loss": 0.9374, + "step": 10055 + }, + { + "epoch": 0.6909440703586642, + "grad_norm": 1.030096173286438, + "learning_rate": 1.319162150199939e-05, + "loss": 0.9343, + "step": 10056 + }, + { + "epoch": 0.6910127799917548, + "grad_norm": 1.0474200248718262, + "learning_rate": 1.3186336569982737e-05, + "loss": 0.921, + "step": 10057 + }, + { + "epoch": 0.6910814896248454, + "grad_norm": 1.0342066287994385, + "learning_rate": 1.3181052317598953e-05, + "loss": 0.9596, + "step": 10058 + }, + { + "epoch": 0.691150199257936, + "grad_norm": 1.0651354789733887, + "learning_rate": 1.3175768745152007e-05, + "loss": 0.9391, + "step": 10059 + }, + { + "epoch": 0.6912189088910266, + "grad_norm": 1.1025583744049072, + "learning_rate": 1.3170485852945889e-05, + "loss": 0.9793, + "step": 10060 + }, + { + "epoch": 0.6912876185241171, + "grad_norm": 1.0783214569091797, + "learning_rate": 1.3165203641284496e-05, + "loss": 0.9957, + "step": 10061 + }, + { + "epoch": 0.6913563281572076, + "grad_norm": 1.0816882848739624, + "learning_rate": 1.3159922110471717e-05, + "loss": 0.9874, + "step": 10062 + }, + { + "epoch": 0.6914250377902982, + "grad_norm": 1.0008788108825684, + "learning_rate": 1.3154641260811407e-05, + "loss": 0.9096, + "step": 10063 + }, + { + "epoch": 0.6914937474233888, + "grad_norm": 1.0846962928771973, + "learning_rate": 1.3149361092607359e-05, + "loss": 0.9588, + "step": 10064 + }, + { + "epoch": 0.6915624570564793, + "grad_norm": 1.0028098821640015, + "learning_rate": 1.314408160616335e-05, + "loss": 0.9041, + "step": 10065 + }, + { + "epoch": 0.6916311666895699, + "grad_norm": 1.086205244064331, + "learning_rate": 1.313880280178309e-05, + "loss": 0.9202, + "step": 10066 + }, + { + "epoch": 0.6916998763226604, + "grad_norm": 1.0956975221633911, + "learning_rate": 1.313352467977027e-05, + "loss": 0.9655, + "step": 10067 + }, + { + "epoch": 0.691768585955751, + "grad_norm": 1.074902892112732, + "learning_rate": 1.3128247240428543e-05, + "loss": 0.9362, + "step": 10068 + }, + { + "epoch": 0.6918372955888415, + "grad_norm": 1.078230381011963, + "learning_rate": 1.3122970484061514e-05, + "loss": 0.9125, + "step": 10069 + }, + { + "epoch": 0.6919060052219321, + "grad_norm": 1.0621697902679443, + "learning_rate": 1.3117694410972748e-05, + "loss": 0.9515, + "step": 10070 + }, + { + "epoch": 0.6919747148550227, + "grad_norm": 1.106017827987671, + "learning_rate": 1.3112419021465789e-05, + "loss": 0.9576, + "step": 10071 + }, + { + "epoch": 0.6920434244881132, + "grad_norm": 1.0506360530853271, + "learning_rate": 1.3107144315844106e-05, + "loss": 0.9066, + "step": 10072 + }, + { + "epoch": 0.6921121341212038, + "grad_norm": 1.0784257650375366, + "learning_rate": 1.3101870294411158e-05, + "loss": 0.9387, + "step": 10073 + }, + { + "epoch": 0.6921808437542943, + "grad_norm": 1.0013086795806885, + "learning_rate": 1.3096596957470358e-05, + "loss": 0.9111, + "step": 10074 + }, + { + "epoch": 0.6922495533873849, + "grad_norm": 1.0494831800460815, + "learning_rate": 1.3091324305325076e-05, + "loss": 0.931, + "step": 10075 + }, + { + "epoch": 0.6923182630204755, + "grad_norm": 1.0280492305755615, + "learning_rate": 1.3086052338278659e-05, + "loss": 0.9829, + "step": 10076 + }, + { + "epoch": 0.6923869726535661, + "grad_norm": 1.0395652055740356, + "learning_rate": 1.3080781056634373e-05, + "loss": 0.9192, + "step": 10077 + }, + { + "epoch": 0.6924556822866565, + "grad_norm": 1.0374199151992798, + "learning_rate": 1.3075510460695487e-05, + "loss": 0.938, + "step": 10078 + }, + { + "epoch": 0.6925243919197471, + "grad_norm": 1.150742530822754, + "learning_rate": 1.3070240550765212e-05, + "loss": 0.9769, + "step": 10079 + }, + { + "epoch": 0.6925931015528377, + "grad_norm": 1.074981927871704, + "learning_rate": 1.3064971327146735e-05, + "loss": 0.9547, + "step": 10080 + }, + { + "epoch": 0.6926618111859283, + "grad_norm": 1.0372240543365479, + "learning_rate": 1.3059702790143172e-05, + "loss": 0.9646, + "step": 10081 + }, + { + "epoch": 0.6927305208190189, + "grad_norm": 1.017575740814209, + "learning_rate": 1.3054434940057636e-05, + "loss": 0.9757, + "step": 10082 + }, + { + "epoch": 0.6927992304521093, + "grad_norm": 1.0687235593795776, + "learning_rate": 1.3049167777193161e-05, + "loss": 0.9374, + "step": 10083 + }, + { + "epoch": 0.6928679400851999, + "grad_norm": 1.0826841592788696, + "learning_rate": 1.3043901301852781e-05, + "loss": 0.9619, + "step": 10084 + }, + { + "epoch": 0.6929366497182905, + "grad_norm": 1.0481898784637451, + "learning_rate": 1.303863551433947e-05, + "loss": 0.9636, + "step": 10085 + }, + { + "epoch": 0.6930053593513811, + "grad_norm": 1.014359712600708, + "learning_rate": 1.303337041495617e-05, + "loss": 0.9298, + "step": 10086 + }, + { + "epoch": 0.6930740689844717, + "grad_norm": 1.0685925483703613, + "learning_rate": 1.3028106004005766e-05, + "loss": 0.917, + "step": 10087 + }, + { + "epoch": 0.6931427786175621, + "grad_norm": 1.0335195064544678, + "learning_rate": 1.3022842281791126e-05, + "loss": 0.9555, + "step": 10088 + }, + { + "epoch": 0.6932114882506527, + "grad_norm": 1.064267873764038, + "learning_rate": 1.3017579248615064e-05, + "loss": 0.9672, + "step": 10089 + }, + { + "epoch": 0.6932801978837433, + "grad_norm": 1.0418916940689087, + "learning_rate": 1.3012316904780362e-05, + "loss": 0.8972, + "step": 10090 + }, + { + "epoch": 0.6933489075168339, + "grad_norm": 1.0859742164611816, + "learning_rate": 1.300705525058977e-05, + "loss": 0.932, + "step": 10091 + }, + { + "epoch": 0.6934176171499244, + "grad_norm": 1.0216747522354126, + "learning_rate": 1.3001794286345964e-05, + "loss": 0.9478, + "step": 10092 + }, + { + "epoch": 0.693486326783015, + "grad_norm": 1.0484598875045776, + "learning_rate": 1.2996534012351622e-05, + "loss": 0.9361, + "step": 10093 + }, + { + "epoch": 0.6935550364161055, + "grad_norm": 1.0621616840362549, + "learning_rate": 1.2991274428909356e-05, + "loss": 0.9418, + "step": 10094 + }, + { + "epoch": 0.6936237460491961, + "grad_norm": 0.9914979338645935, + "learning_rate": 1.298601553632175e-05, + "loss": 0.9398, + "step": 10095 + }, + { + "epoch": 0.6936924556822867, + "grad_norm": 1.0896406173706055, + "learning_rate": 1.2980757334891357e-05, + "loss": 0.9275, + "step": 10096 + }, + { + "epoch": 0.6937611653153772, + "grad_norm": 1.0017164945602417, + "learning_rate": 1.2975499824920656e-05, + "loss": 0.9215, + "step": 10097 + }, + { + "epoch": 0.6938298749484678, + "grad_norm": 1.0615588426589966, + "learning_rate": 1.2970243006712115e-05, + "loss": 0.9152, + "step": 10098 + }, + { + "epoch": 0.6938985845815583, + "grad_norm": 1.1402199268341064, + "learning_rate": 1.2964986880568161e-05, + "loss": 0.972, + "step": 10099 + }, + { + "epoch": 0.6939672942146489, + "grad_norm": 1.1045976877212524, + "learning_rate": 1.2959731446791174e-05, + "loss": 0.956, + "step": 10100 + }, + { + "epoch": 0.6940360038477394, + "grad_norm": 1.1218618154525757, + "learning_rate": 1.2954476705683505e-05, + "loss": 0.9272, + "step": 10101 + }, + { + "epoch": 0.69410471348083, + "grad_norm": 1.0445016622543335, + "learning_rate": 1.2949222657547438e-05, + "loss": 0.9442, + "step": 10102 + }, + { + "epoch": 0.6941734231139206, + "grad_norm": 1.0394103527069092, + "learning_rate": 1.294396930268524e-05, + "loss": 0.9298, + "step": 10103 + }, + { + "epoch": 0.6942421327470111, + "grad_norm": 1.1307380199432373, + "learning_rate": 1.293871664139914e-05, + "loss": 0.9274, + "step": 10104 + }, + { + "epoch": 0.6943108423801017, + "grad_norm": 1.0330243110656738, + "learning_rate": 1.2933464673991327e-05, + "loss": 0.9248, + "step": 10105 + }, + { + "epoch": 0.6943795520131922, + "grad_norm": 1.0361647605895996, + "learning_rate": 1.2928213400763925e-05, + "loss": 0.9131, + "step": 10106 + }, + { + "epoch": 0.6944482616462828, + "grad_norm": 1.0659209489822388, + "learning_rate": 1.2922962822019052e-05, + "loss": 0.9155, + "step": 10107 + }, + { + "epoch": 0.6945169712793734, + "grad_norm": 1.107013463973999, + "learning_rate": 1.2917712938058756e-05, + "loss": 0.9589, + "step": 10108 + }, + { + "epoch": 0.694585680912464, + "grad_norm": 1.0000529289245605, + "learning_rate": 1.2912463749185073e-05, + "loss": 0.9301, + "step": 10109 + }, + { + "epoch": 0.6946543905455544, + "grad_norm": 1.044594645500183, + "learning_rate": 1.2907215255699972e-05, + "loss": 0.9294, + "step": 10110 + }, + { + "epoch": 0.694723100178645, + "grad_norm": 1.099448800086975, + "learning_rate": 1.2901967457905418e-05, + "loss": 0.9461, + "step": 10111 + }, + { + "epoch": 0.6947918098117356, + "grad_norm": 1.0027967691421509, + "learning_rate": 1.289672035610329e-05, + "loss": 0.9252, + "step": 10112 + }, + { + "epoch": 0.6948605194448262, + "grad_norm": 1.0247440338134766, + "learning_rate": 1.2891473950595456e-05, + "loss": 0.9454, + "step": 10113 + }, + { + "epoch": 0.6949292290779168, + "grad_norm": 1.035027027130127, + "learning_rate": 1.2886228241683749e-05, + "loss": 0.9455, + "step": 10114 + }, + { + "epoch": 0.6949979387110072, + "grad_norm": 1.027968406677246, + "learning_rate": 1.288098322966994e-05, + "loss": 0.9341, + "step": 10115 + }, + { + "epoch": 0.6950666483440978, + "grad_norm": 1.104258418083191, + "learning_rate": 1.287573891485579e-05, + "loss": 0.9523, + "step": 10116 + }, + { + "epoch": 0.6951353579771884, + "grad_norm": 1.0612114667892456, + "learning_rate": 1.2870495297542975e-05, + "loss": 0.9381, + "step": 10117 + }, + { + "epoch": 0.695204067610279, + "grad_norm": 1.0531208515167236, + "learning_rate": 1.2865252378033172e-05, + "loss": 0.9206, + "step": 10118 + }, + { + "epoch": 0.6952727772433696, + "grad_norm": 1.0652137994766235, + "learning_rate": 1.2860010156628e-05, + "loss": 0.9785, + "step": 10119 + }, + { + "epoch": 0.69534148687646, + "grad_norm": 1.0444027185440063, + "learning_rate": 1.2854768633629039e-05, + "loss": 0.9385, + "step": 10120 + }, + { + "epoch": 0.6954101965095506, + "grad_norm": 1.053674340248108, + "learning_rate": 1.2849527809337847e-05, + "loss": 0.8984, + "step": 10121 + }, + { + "epoch": 0.6954789061426412, + "grad_norm": 1.0727972984313965, + "learning_rate": 1.28442876840559e-05, + "loss": 0.9635, + "step": 10122 + }, + { + "epoch": 0.6955476157757318, + "grad_norm": 1.1337391138076782, + "learning_rate": 1.2839048258084667e-05, + "loss": 0.9444, + "step": 10123 + }, + { + "epoch": 0.6956163254088223, + "grad_norm": 1.0361815690994263, + "learning_rate": 1.2833809531725577e-05, + "loss": 0.8925, + "step": 10124 + }, + { + "epoch": 0.6956850350419129, + "grad_norm": 1.088839054107666, + "learning_rate": 1.2828571505280007e-05, + "loss": 0.9331, + "step": 10125 + }, + { + "epoch": 0.6957537446750034, + "grad_norm": 1.0345633029937744, + "learning_rate": 1.2823334179049295e-05, + "loss": 0.957, + "step": 10126 + }, + { + "epoch": 0.695822454308094, + "grad_norm": 1.0812045335769653, + "learning_rate": 1.2818097553334758e-05, + "loss": 0.9514, + "step": 10127 + }, + { + "epoch": 0.6958911639411846, + "grad_norm": 1.0200421810150146, + "learning_rate": 1.2812861628437622e-05, + "loss": 0.9127, + "step": 10128 + }, + { + "epoch": 0.6959598735742751, + "grad_norm": 1.116336703300476, + "learning_rate": 1.2807626404659142e-05, + "loss": 0.9508, + "step": 10129 + }, + { + "epoch": 0.6960285832073657, + "grad_norm": 1.048421859741211, + "learning_rate": 1.2802391882300468e-05, + "loss": 0.9427, + "step": 10130 + }, + { + "epoch": 0.6960972928404562, + "grad_norm": 1.0944476127624512, + "learning_rate": 1.2797158061662757e-05, + "loss": 0.9749, + "step": 10131 + }, + { + "epoch": 0.6961660024735468, + "grad_norm": 1.0182697772979736, + "learning_rate": 1.2791924943047106e-05, + "loss": 0.9067, + "step": 10132 + }, + { + "epoch": 0.6962347121066373, + "grad_norm": 1.0650880336761475, + "learning_rate": 1.2786692526754562e-05, + "loss": 0.9255, + "step": 10133 + }, + { + "epoch": 0.6963034217397279, + "grad_norm": 1.0490162372589111, + "learning_rate": 1.2781460813086149e-05, + "loss": 0.922, + "step": 10134 + }, + { + "epoch": 0.6963721313728185, + "grad_norm": 1.0452065467834473, + "learning_rate": 1.2776229802342848e-05, + "loss": 0.936, + "step": 10135 + }, + { + "epoch": 0.6964408410059091, + "grad_norm": 1.0399588346481323, + "learning_rate": 1.2770999494825591e-05, + "loss": 0.9114, + "step": 10136 + }, + { + "epoch": 0.6965095506389996, + "grad_norm": 1.0572952032089233, + "learning_rate": 1.2765769890835288e-05, + "loss": 0.9199, + "step": 10137 + }, + { + "epoch": 0.6965782602720901, + "grad_norm": 1.0166881084442139, + "learning_rate": 1.2760540990672773e-05, + "loss": 0.9321, + "step": 10138 + }, + { + "epoch": 0.6966469699051807, + "grad_norm": 1.0147428512573242, + "learning_rate": 1.2755312794638871e-05, + "loss": 0.9033, + "step": 10139 + }, + { + "epoch": 0.6967156795382713, + "grad_norm": 1.0475112199783325, + "learning_rate": 1.2750085303034359e-05, + "loss": 0.9202, + "step": 10140 + }, + { + "epoch": 0.6967843891713619, + "grad_norm": 1.059555172920227, + "learning_rate": 1.274485851615997e-05, + "loss": 0.9679, + "step": 10141 + }, + { + "epoch": 0.6968530988044523, + "grad_norm": 1.0597151517868042, + "learning_rate": 1.2739632434316406e-05, + "loss": 0.9527, + "step": 10142 + }, + { + "epoch": 0.6969218084375429, + "grad_norm": 1.0593091249465942, + "learning_rate": 1.2734407057804307e-05, + "loss": 0.9759, + "step": 10143 + }, + { + "epoch": 0.6969905180706335, + "grad_norm": 1.1072897911071777, + "learning_rate": 1.2729182386924287e-05, + "loss": 0.9424, + "step": 10144 + }, + { + "epoch": 0.6970592277037241, + "grad_norm": 1.031618356704712, + "learning_rate": 1.2723958421976926e-05, + "loss": 0.9393, + "step": 10145 + }, + { + "epoch": 0.6971279373368147, + "grad_norm": 1.0658783912658691, + "learning_rate": 1.2718735163262751e-05, + "loss": 0.9271, + "step": 10146 + }, + { + "epoch": 0.6971966469699051, + "grad_norm": 1.0529934167861938, + "learning_rate": 1.2713512611082263e-05, + "loss": 0.9016, + "step": 10147 + }, + { + "epoch": 0.6972653566029957, + "grad_norm": 1.102037787437439, + "learning_rate": 1.2708290765735891e-05, + "loss": 0.9326, + "step": 10148 + }, + { + "epoch": 0.6973340662360863, + "grad_norm": 1.0375200510025024, + "learning_rate": 1.2703069627524061e-05, + "loss": 0.9299, + "step": 10149 + }, + { + "epoch": 0.6974027758691769, + "grad_norm": 1.0694140195846558, + "learning_rate": 1.2697849196747134e-05, + "loss": 0.9396, + "step": 10150 + }, + { + "epoch": 0.6974714855022675, + "grad_norm": 1.0342503786087036, + "learning_rate": 1.2692629473705453e-05, + "loss": 0.9023, + "step": 10151 + }, + { + "epoch": 0.697540195135358, + "grad_norm": 1.043169617652893, + "learning_rate": 1.2687410458699284e-05, + "loss": 0.9474, + "step": 10152 + }, + { + "epoch": 0.6976089047684485, + "grad_norm": 1.0511994361877441, + "learning_rate": 1.2682192152028882e-05, + "loss": 0.9224, + "step": 10153 + }, + { + "epoch": 0.6976776144015391, + "grad_norm": 1.0626704692840576, + "learning_rate": 1.2676974553994467e-05, + "loss": 0.891, + "step": 10154 + }, + { + "epoch": 0.6977463240346297, + "grad_norm": 1.0155867338180542, + "learning_rate": 1.2671757664896183e-05, + "loss": 0.9208, + "step": 10155 + }, + { + "epoch": 0.6978150336677202, + "grad_norm": 1.0558419227600098, + "learning_rate": 1.2666541485034161e-05, + "loss": 0.9269, + "step": 10156 + }, + { + "epoch": 0.6978837433008108, + "grad_norm": 1.070788025856018, + "learning_rate": 1.26613260147085e-05, + "loss": 0.9334, + "step": 10157 + }, + { + "epoch": 0.6979524529339013, + "grad_norm": 1.0638681650161743, + "learning_rate": 1.2656111254219217e-05, + "loss": 0.9002, + "step": 10158 + }, + { + "epoch": 0.6980211625669919, + "grad_norm": 0.9863821864128113, + "learning_rate": 1.265089720386633e-05, + "loss": 0.8909, + "step": 10159 + }, + { + "epoch": 0.6980898722000825, + "grad_norm": 1.1120686531066895, + "learning_rate": 1.2645683863949798e-05, + "loss": 0.9475, + "step": 10160 + }, + { + "epoch": 0.698158581833173, + "grad_norm": 1.0984565019607544, + "learning_rate": 1.2640471234769534e-05, + "loss": 0.9275, + "step": 10161 + }, + { + "epoch": 0.6982272914662636, + "grad_norm": 1.0495084524154663, + "learning_rate": 1.2635259316625443e-05, + "loss": 0.9303, + "step": 10162 + }, + { + "epoch": 0.6982960010993541, + "grad_norm": 1.0694568157196045, + "learning_rate": 1.2630048109817328e-05, + "loss": 0.9913, + "step": 10163 + }, + { + "epoch": 0.6983647107324447, + "grad_norm": 1.035286545753479, + "learning_rate": 1.2624837614645002e-05, + "loss": 0.9215, + "step": 10164 + }, + { + "epoch": 0.6984334203655352, + "grad_norm": 1.0970993041992188, + "learning_rate": 1.2619627831408226e-05, + "loss": 0.9417, + "step": 10165 + }, + { + "epoch": 0.6985021299986258, + "grad_norm": 1.0642520189285278, + "learning_rate": 1.2614418760406713e-05, + "loss": 0.9415, + "step": 10166 + }, + { + "epoch": 0.6985708396317164, + "grad_norm": 1.0494821071624756, + "learning_rate": 1.2609210401940144e-05, + "loss": 0.902, + "step": 10167 + }, + { + "epoch": 0.698639549264807, + "grad_norm": 1.0684747695922852, + "learning_rate": 1.2604002756308136e-05, + "loss": 0.9939, + "step": 10168 + }, + { + "epoch": 0.6987082588978974, + "grad_norm": 1.1379363536834717, + "learning_rate": 1.2598795823810294e-05, + "loss": 0.9787, + "step": 10169 + }, + { + "epoch": 0.698776968530988, + "grad_norm": 1.068223237991333, + "learning_rate": 1.2593589604746165e-05, + "loss": 0.9186, + "step": 10170 + }, + { + "epoch": 0.6988456781640786, + "grad_norm": 1.0687717199325562, + "learning_rate": 1.258838409941526e-05, + "loss": 0.9416, + "step": 10171 + }, + { + "epoch": 0.6989143877971692, + "grad_norm": 1.0204503536224365, + "learning_rate": 1.258317930811706e-05, + "loss": 0.9365, + "step": 10172 + }, + { + "epoch": 0.6989830974302598, + "grad_norm": 1.058455467224121, + "learning_rate": 1.2577975231150974e-05, + "loss": 0.9255, + "step": 10173 + }, + { + "epoch": 0.6990518070633502, + "grad_norm": 1.0614203214645386, + "learning_rate": 1.2572771868816398e-05, + "loss": 0.9085, + "step": 10174 + }, + { + "epoch": 0.6991205166964408, + "grad_norm": 1.0092847347259521, + "learning_rate": 1.2567569221412681e-05, + "loss": 0.9323, + "step": 10175 + }, + { + "epoch": 0.6991892263295314, + "grad_norm": 1.0247079133987427, + "learning_rate": 1.2562367289239136e-05, + "loss": 0.9061, + "step": 10176 + }, + { + "epoch": 0.699257935962622, + "grad_norm": 1.0598825216293335, + "learning_rate": 1.2557166072595005e-05, + "loss": 0.9243, + "step": 10177 + }, + { + "epoch": 0.6993266455957126, + "grad_norm": 1.04507315158844, + "learning_rate": 1.2551965571779537e-05, + "loss": 0.9552, + "step": 10178 + }, + { + "epoch": 0.699395355228803, + "grad_norm": 1.0565040111541748, + "learning_rate": 1.2546765787091886e-05, + "loss": 0.9074, + "step": 10179 + }, + { + "epoch": 0.6994640648618936, + "grad_norm": 1.0161770582199097, + "learning_rate": 1.254156671883121e-05, + "loss": 0.9109, + "step": 10180 + }, + { + "epoch": 0.6995327744949842, + "grad_norm": 1.0073391199111938, + "learning_rate": 1.2536368367296603e-05, + "loss": 0.9666, + "step": 10181 + }, + { + "epoch": 0.6996014841280748, + "grad_norm": 1.1052775382995605, + "learning_rate": 1.2531170732787123e-05, + "loss": 0.9171, + "step": 10182 + }, + { + "epoch": 0.6996701937611653, + "grad_norm": 1.0909000635147095, + "learning_rate": 1.2525973815601797e-05, + "loss": 0.9688, + "step": 10183 + }, + { + "epoch": 0.6997389033942559, + "grad_norm": 1.0750923156738281, + "learning_rate": 1.252077761603958e-05, + "loss": 0.9085, + "step": 10184 + }, + { + "epoch": 0.6998076130273464, + "grad_norm": 1.0351327657699585, + "learning_rate": 1.2515582134399422e-05, + "loss": 0.9451, + "step": 10185 + }, + { + "epoch": 0.699876322660437, + "grad_norm": 1.037603735923767, + "learning_rate": 1.2510387370980211e-05, + "loss": 0.95, + "step": 10186 + }, + { + "epoch": 0.6999450322935276, + "grad_norm": 1.0514975786209106, + "learning_rate": 1.25051933260808e-05, + "loss": 0.9021, + "step": 10187 + }, + { + "epoch": 0.7000137419266181, + "grad_norm": 1.1274750232696533, + "learning_rate": 1.2500000000000006e-05, + "loss": 0.9638, + "step": 10188 + }, + { + "epoch": 0.7000824515597087, + "grad_norm": 1.0644009113311768, + "learning_rate": 1.2494807393036581e-05, + "loss": 0.9103, + "step": 10189 + }, + { + "epoch": 0.7001511611927992, + "grad_norm": 0.9951073527336121, + "learning_rate": 1.2489615505489264e-05, + "loss": 0.9164, + "step": 10190 + }, + { + "epoch": 0.7002198708258898, + "grad_norm": 1.0825108289718628, + "learning_rate": 1.2484424337656736e-05, + "loss": 0.9363, + "step": 10191 + }, + { + "epoch": 0.7002885804589803, + "grad_norm": 1.1257621049880981, + "learning_rate": 1.2479233889837646e-05, + "loss": 0.9488, + "step": 10192 + }, + { + "epoch": 0.7003572900920709, + "grad_norm": 1.0881621837615967, + "learning_rate": 1.2474044162330607e-05, + "loss": 0.9558, + "step": 10193 + }, + { + "epoch": 0.7004259997251615, + "grad_norm": 1.0896193981170654, + "learning_rate": 1.2468855155434156e-05, + "loss": 0.9385, + "step": 10194 + }, + { + "epoch": 0.700494709358252, + "grad_norm": 1.0698318481445312, + "learning_rate": 1.246366686944683e-05, + "loss": 0.9857, + "step": 10195 + }, + { + "epoch": 0.7005634189913426, + "grad_norm": 1.069299340248108, + "learning_rate": 1.2458479304667106e-05, + "loss": 0.946, + "step": 10196 + }, + { + "epoch": 0.7006321286244331, + "grad_norm": 1.046854019165039, + "learning_rate": 1.245329246139342e-05, + "loss": 0.9274, + "step": 10197 + }, + { + "epoch": 0.7007008382575237, + "grad_norm": 1.0533292293548584, + "learning_rate": 1.244810633992418e-05, + "loss": 0.9131, + "step": 10198 + }, + { + "epoch": 0.7007695478906143, + "grad_norm": 1.026625394821167, + "learning_rate": 1.2442920940557715e-05, + "loss": 0.9135, + "step": 10199 + }, + { + "epoch": 0.7008382575237049, + "grad_norm": 1.0148520469665527, + "learning_rate": 1.2437736263592365e-05, + "loss": 0.9124, + "step": 10200 + }, + { + "epoch": 0.7009069671567953, + "grad_norm": 1.1268762350082397, + "learning_rate": 1.2432552309326375e-05, + "loss": 0.9637, + "step": 10201 + }, + { + "epoch": 0.7009756767898859, + "grad_norm": 1.1022484302520752, + "learning_rate": 1.242736907805799e-05, + "loss": 0.9405, + "step": 10202 + }, + { + "epoch": 0.7010443864229765, + "grad_norm": 1.0148968696594238, + "learning_rate": 1.2422186570085406e-05, + "loss": 0.9259, + "step": 10203 + }, + { + "epoch": 0.7011130960560671, + "grad_norm": 1.1362463235855103, + "learning_rate": 1.2417004785706751e-05, + "loss": 1.0026, + "step": 10204 + }, + { + "epoch": 0.7011818056891577, + "grad_norm": 1.1339049339294434, + "learning_rate": 1.2411823725220136e-05, + "loss": 0.9501, + "step": 10205 + }, + { + "epoch": 0.7012505153222481, + "grad_norm": 1.1198676824569702, + "learning_rate": 1.2406643388923628e-05, + "loss": 0.9426, + "step": 10206 + }, + { + "epoch": 0.7013192249553387, + "grad_norm": 1.0682169198989868, + "learning_rate": 1.2401463777115246e-05, + "loss": 0.9292, + "step": 10207 + }, + { + "epoch": 0.7013879345884293, + "grad_norm": 1.0551016330718994, + "learning_rate": 1.2396284890092982e-05, + "loss": 0.9465, + "step": 10208 + }, + { + "epoch": 0.7014566442215199, + "grad_norm": 1.0184324979782104, + "learning_rate": 1.2391106728154755e-05, + "loss": 0.9156, + "step": 10209 + }, + { + "epoch": 0.7015253538546105, + "grad_norm": 1.0149813890457153, + "learning_rate": 1.2385929291598469e-05, + "loss": 0.9312, + "step": 10210 + }, + { + "epoch": 0.7015940634877009, + "grad_norm": 0.9962772130966187, + "learning_rate": 1.2380752580721977e-05, + "loss": 0.9039, + "step": 10211 + }, + { + "epoch": 0.7016627731207915, + "grad_norm": 1.125170111656189, + "learning_rate": 1.2375576595823099e-05, + "loss": 0.9833, + "step": 10212 + }, + { + "epoch": 0.7017314827538821, + "grad_norm": 1.062168836593628, + "learning_rate": 1.2370401337199609e-05, + "loss": 0.9767, + "step": 10213 + }, + { + "epoch": 0.7018001923869727, + "grad_norm": 1.0856049060821533, + "learning_rate": 1.2365226805149219e-05, + "loss": 0.9272, + "step": 10214 + }, + { + "epoch": 0.7018689020200632, + "grad_norm": 1.0862267017364502, + "learning_rate": 1.2360052999969629e-05, + "loss": 0.9569, + "step": 10215 + }, + { + "epoch": 0.7019376116531538, + "grad_norm": 1.0425400733947754, + "learning_rate": 1.2354879921958482e-05, + "loss": 0.9375, + "step": 10216 + }, + { + "epoch": 0.7020063212862443, + "grad_norm": 1.0682432651519775, + "learning_rate": 1.2349707571413382e-05, + "loss": 0.9837, + "step": 10217 + }, + { + "epoch": 0.7020750309193349, + "grad_norm": 1.026438593864441, + "learning_rate": 1.2344535948631902e-05, + "loss": 0.9185, + "step": 10218 + }, + { + "epoch": 0.7021437405524255, + "grad_norm": 1.1409727334976196, + "learning_rate": 1.2339365053911545e-05, + "loss": 0.9692, + "step": 10219 + }, + { + "epoch": 0.702212450185516, + "grad_norm": 1.1222177743911743, + "learning_rate": 1.2334194887549794e-05, + "loss": 0.9521, + "step": 10220 + }, + { + "epoch": 0.7022811598186066, + "grad_norm": 0.9874287247657776, + "learning_rate": 1.2329025449844089e-05, + "loss": 0.9098, + "step": 10221 + }, + { + "epoch": 0.7023498694516971, + "grad_norm": 1.1049540042877197, + "learning_rate": 1.2323856741091835e-05, + "loss": 0.9677, + "step": 10222 + }, + { + "epoch": 0.7024185790847877, + "grad_norm": 1.0338765382766724, + "learning_rate": 1.2318688761590364e-05, + "loss": 0.8872, + "step": 10223 + }, + { + "epoch": 0.7024872887178782, + "grad_norm": 1.1013096570968628, + "learning_rate": 1.2313521511636995e-05, + "loss": 0.9265, + "step": 10224 + }, + { + "epoch": 0.7025559983509688, + "grad_norm": 1.1421321630477905, + "learning_rate": 1.2308354991529009e-05, + "loss": 0.9879, + "step": 10225 + }, + { + "epoch": 0.7026247079840594, + "grad_norm": 1.0420914888381958, + "learning_rate": 1.2303189201563612e-05, + "loss": 0.9124, + "step": 10226 + }, + { + "epoch": 0.7026934176171499, + "grad_norm": 1.0187976360321045, + "learning_rate": 1.2298024142037998e-05, + "loss": 0.927, + "step": 10227 + }, + { + "epoch": 0.7027621272502405, + "grad_norm": 1.0720094442367554, + "learning_rate": 1.229285981324932e-05, + "loss": 0.9023, + "step": 10228 + }, + { + "epoch": 0.702830836883331, + "grad_norm": 1.0483520030975342, + "learning_rate": 1.2287696215494657e-05, + "loss": 0.9439, + "step": 10229 + }, + { + "epoch": 0.7028995465164216, + "grad_norm": 1.045654535293579, + "learning_rate": 1.2282533349071082e-05, + "loss": 0.9208, + "step": 10230 + }, + { + "epoch": 0.7029682561495122, + "grad_norm": 1.0894196033477783, + "learning_rate": 1.227737121427561e-05, + "loss": 0.9282, + "step": 10231 + }, + { + "epoch": 0.7030369657826028, + "grad_norm": 1.0899680852890015, + "learning_rate": 1.2272209811405214e-05, + "loss": 0.919, + "step": 10232 + }, + { + "epoch": 0.7031056754156932, + "grad_norm": 1.066543459892273, + "learning_rate": 1.2267049140756837e-05, + "loss": 0.9243, + "step": 10233 + }, + { + "epoch": 0.7031743850487838, + "grad_norm": 1.0984374284744263, + "learning_rate": 1.2261889202627347e-05, + "loss": 0.9683, + "step": 10234 + }, + { + "epoch": 0.7032430946818744, + "grad_norm": 1.1305752992630005, + "learning_rate": 1.2256729997313607e-05, + "loss": 0.97, + "step": 10235 + }, + { + "epoch": 0.703311804314965, + "grad_norm": 1.0631970167160034, + "learning_rate": 1.2251571525112421e-05, + "loss": 0.9487, + "step": 10236 + }, + { + "epoch": 0.7033805139480556, + "grad_norm": 1.1001654863357544, + "learning_rate": 1.2246413786320549e-05, + "loss": 0.9227, + "step": 10237 + }, + { + "epoch": 0.703449223581146, + "grad_norm": 1.0801620483398438, + "learning_rate": 1.2241256781234717e-05, + "loss": 0.9571, + "step": 10238 + }, + { + "epoch": 0.7035179332142366, + "grad_norm": 1.0520836114883423, + "learning_rate": 1.2236100510151613e-05, + "loss": 0.9296, + "step": 10239 + }, + { + "epoch": 0.7035866428473272, + "grad_norm": 1.0772486925125122, + "learning_rate": 1.2230944973367852e-05, + "loss": 0.9572, + "step": 10240 + }, + { + "epoch": 0.7036553524804178, + "grad_norm": 1.0575194358825684, + "learning_rate": 1.2225790171180046e-05, + "loss": 0.9474, + "step": 10241 + }, + { + "epoch": 0.7037240621135084, + "grad_norm": 1.0242794752120972, + "learning_rate": 1.2220636103884738e-05, + "loss": 0.9158, + "step": 10242 + }, + { + "epoch": 0.7037927717465988, + "grad_norm": 1.0152530670166016, + "learning_rate": 1.2215482771778445e-05, + "loss": 0.9326, + "step": 10243 + }, + { + "epoch": 0.7038614813796894, + "grad_norm": 1.0773507356643677, + "learning_rate": 1.2210330175157641e-05, + "loss": 0.9771, + "step": 10244 + }, + { + "epoch": 0.70393019101278, + "grad_norm": 1.061579942703247, + "learning_rate": 1.2205178314318739e-05, + "loss": 0.9478, + "step": 10245 + }, + { + "epoch": 0.7039989006458706, + "grad_norm": 1.1076260805130005, + "learning_rate": 1.2200027189558122e-05, + "loss": 0.9492, + "step": 10246 + }, + { + "epoch": 0.7040676102789611, + "grad_norm": 1.0710415840148926, + "learning_rate": 1.2194876801172148e-05, + "loss": 0.9466, + "step": 10247 + }, + { + "epoch": 0.7041363199120517, + "grad_norm": 1.060288667678833, + "learning_rate": 1.2189727149457092e-05, + "loss": 0.9339, + "step": 10248 + }, + { + "epoch": 0.7042050295451422, + "grad_norm": 1.052958369255066, + "learning_rate": 1.2184578234709223e-05, + "loss": 0.9165, + "step": 10249 + }, + { + "epoch": 0.7042737391782328, + "grad_norm": 1.0930140018463135, + "learning_rate": 1.2179430057224766e-05, + "loss": 0.9708, + "step": 10250 + }, + { + "epoch": 0.7043424488113234, + "grad_norm": 1.0846339464187622, + "learning_rate": 1.217428261729987e-05, + "loss": 0.9759, + "step": 10251 + }, + { + "epoch": 0.7044111584444139, + "grad_norm": 1.0371477603912354, + "learning_rate": 1.2169135915230676e-05, + "loss": 0.9646, + "step": 10252 + }, + { + "epoch": 0.7044798680775045, + "grad_norm": 1.1222434043884277, + "learning_rate": 1.2163989951313268e-05, + "loss": 0.9673, + "step": 10253 + }, + { + "epoch": 0.704548577710595, + "grad_norm": 1.043664813041687, + "learning_rate": 1.2158844725843698e-05, + "loss": 0.9505, + "step": 10254 + }, + { + "epoch": 0.7046172873436856, + "grad_norm": 1.0157376527786255, + "learning_rate": 1.2153700239117957e-05, + "loss": 0.9466, + "step": 10255 + }, + { + "epoch": 0.7046859969767761, + "grad_norm": 1.0622284412384033, + "learning_rate": 1.2148556491432006e-05, + "loss": 0.9846, + "step": 10256 + }, + { + "epoch": 0.7047547066098667, + "grad_norm": 1.091665506362915, + "learning_rate": 1.2143413483081764e-05, + "loss": 0.9497, + "step": 10257 + }, + { + "epoch": 0.7048234162429573, + "grad_norm": 1.0354337692260742, + "learning_rate": 1.2138271214363106e-05, + "loss": 0.9574, + "step": 10258 + }, + { + "epoch": 0.7048921258760478, + "grad_norm": 1.0857865810394287, + "learning_rate": 1.213312968557187e-05, + "loss": 0.9646, + "step": 10259 + }, + { + "epoch": 0.7049608355091384, + "grad_norm": 1.08686101436615, + "learning_rate": 1.212798889700383e-05, + "loss": 0.9517, + "step": 10260 + }, + { + "epoch": 0.7050295451422289, + "grad_norm": 1.067836046218872, + "learning_rate": 1.2122848848954741e-05, + "loss": 0.9503, + "step": 10261 + }, + { + "epoch": 0.7050982547753195, + "grad_norm": 1.0882861614227295, + "learning_rate": 1.2117709541720306e-05, + "loss": 0.9377, + "step": 10262 + }, + { + "epoch": 0.7051669644084101, + "grad_norm": 1.0516151189804077, + "learning_rate": 1.2112570975596185e-05, + "loss": 0.9418, + "step": 10263 + }, + { + "epoch": 0.7052356740415007, + "grad_norm": 1.106612205505371, + "learning_rate": 1.2107433150878009e-05, + "loss": 0.962, + "step": 10264 + }, + { + "epoch": 0.7053043836745911, + "grad_norm": 1.0218961238861084, + "learning_rate": 1.2102296067861332e-05, + "loss": 0.9434, + "step": 10265 + }, + { + "epoch": 0.7053730933076817, + "grad_norm": 1.0660046339035034, + "learning_rate": 1.20971597268417e-05, + "loss": 0.9826, + "step": 10266 + }, + { + "epoch": 0.7054418029407723, + "grad_norm": 1.042354941368103, + "learning_rate": 1.2092024128114598e-05, + "loss": 0.9828, + "step": 10267 + }, + { + "epoch": 0.7055105125738629, + "grad_norm": 1.0094794034957886, + "learning_rate": 1.2086889271975482e-05, + "loss": 0.906, + "step": 10268 + }, + { + "epoch": 0.7055792222069535, + "grad_norm": 1.0130831003189087, + "learning_rate": 1.208175515871976e-05, + "loss": 0.9044, + "step": 10269 + }, + { + "epoch": 0.7056479318400439, + "grad_norm": 0.9909226298332214, + "learning_rate": 1.2076621788642775e-05, + "loss": 0.8798, + "step": 10270 + }, + { + "epoch": 0.7057166414731345, + "grad_norm": 1.0408366918563843, + "learning_rate": 1.2071489162039862e-05, + "loss": 0.9265, + "step": 10271 + }, + { + "epoch": 0.7057853511062251, + "grad_norm": 1.073441505432129, + "learning_rate": 1.2066357279206302e-05, + "loss": 0.9392, + "step": 10272 + }, + { + "epoch": 0.7058540607393157, + "grad_norm": 1.050159215927124, + "learning_rate": 1.2061226140437316e-05, + "loss": 0.9631, + "step": 10273 + }, + { + "epoch": 0.7059227703724062, + "grad_norm": 1.0360403060913086, + "learning_rate": 1.2056095746028106e-05, + "loss": 0.9235, + "step": 10274 + }, + { + "epoch": 0.7059914800054967, + "grad_norm": 0.9966999292373657, + "learning_rate": 1.2050966096273808e-05, + "loss": 0.8615, + "step": 10275 + }, + { + "epoch": 0.7060601896385873, + "grad_norm": 1.1170179843902588, + "learning_rate": 1.2045837191469536e-05, + "loss": 0.9713, + "step": 10276 + }, + { + "epoch": 0.7061288992716779, + "grad_norm": 1.0522605180740356, + "learning_rate": 1.204070903191035e-05, + "loss": 0.9543, + "step": 10277 + }, + { + "epoch": 0.7061976089047685, + "grad_norm": 0.9935382604598999, + "learning_rate": 1.2035581617891274e-05, + "loss": 0.9185, + "step": 10278 + }, + { + "epoch": 0.706266318537859, + "grad_norm": 1.0578762292861938, + "learning_rate": 1.2030454949707288e-05, + "loss": 0.932, + "step": 10279 + }, + { + "epoch": 0.7063350281709496, + "grad_norm": 1.0849902629852295, + "learning_rate": 1.2025329027653313e-05, + "loss": 0.9214, + "step": 10280 + }, + { + "epoch": 0.7064037378040401, + "grad_norm": 1.0876010656356812, + "learning_rate": 1.2020203852024247e-05, + "loss": 0.9512, + "step": 10281 + }, + { + "epoch": 0.7064724474371307, + "grad_norm": 1.0532152652740479, + "learning_rate": 1.2015079423114941e-05, + "loss": 0.9381, + "step": 10282 + }, + { + "epoch": 0.7065411570702212, + "grad_norm": 1.0908509492874146, + "learning_rate": 1.20099557412202e-05, + "loss": 0.928, + "step": 10283 + }, + { + "epoch": 0.7066098667033118, + "grad_norm": 1.0406219959259033, + "learning_rate": 1.2004832806634792e-05, + "loss": 0.9746, + "step": 10284 + }, + { + "epoch": 0.7066785763364024, + "grad_norm": 1.0415191650390625, + "learning_rate": 1.1999710619653421e-05, + "loss": 0.9344, + "step": 10285 + }, + { + "epoch": 0.7067472859694929, + "grad_norm": 1.0716891288757324, + "learning_rate": 1.199458918057077e-05, + "loss": 0.9552, + "step": 10286 + }, + { + "epoch": 0.7068159956025835, + "grad_norm": 1.0864830017089844, + "learning_rate": 1.1989468489681473e-05, + "loss": 0.9219, + "step": 10287 + }, + { + "epoch": 0.706884705235674, + "grad_norm": 1.0587663650512695, + "learning_rate": 1.1984348547280121e-05, + "loss": 0.972, + "step": 10288 + }, + { + "epoch": 0.7069534148687646, + "grad_norm": 1.0868418216705322, + "learning_rate": 1.1979229353661265e-05, + "loss": 0.9471, + "step": 10289 + }, + { + "epoch": 0.7070221245018552, + "grad_norm": 1.0632882118225098, + "learning_rate": 1.197411090911941e-05, + "loss": 0.9556, + "step": 10290 + }, + { + "epoch": 0.7070908341349457, + "grad_norm": 1.1768345832824707, + "learning_rate": 1.1968993213949004e-05, + "loss": 0.9588, + "step": 10291 + }, + { + "epoch": 0.7071595437680362, + "grad_norm": 0.9704309105873108, + "learning_rate": 1.1963876268444472e-05, + "loss": 0.9298, + "step": 10292 + }, + { + "epoch": 0.7072282534011268, + "grad_norm": 1.1122305393218994, + "learning_rate": 1.195876007290019e-05, + "loss": 0.9733, + "step": 10293 + }, + { + "epoch": 0.7072969630342174, + "grad_norm": 1.0358214378356934, + "learning_rate": 1.19536446276105e-05, + "loss": 0.9723, + "step": 10294 + }, + { + "epoch": 0.707365672667308, + "grad_norm": 1.0803366899490356, + "learning_rate": 1.194852993286967e-05, + "loss": 0.9013, + "step": 10295 + }, + { + "epoch": 0.7074343823003986, + "grad_norm": 1.0578386783599854, + "learning_rate": 1.1943415988971965e-05, + "loss": 0.9635, + "step": 10296 + }, + { + "epoch": 0.707503091933489, + "grad_norm": 1.0431954860687256, + "learning_rate": 1.1938302796211568e-05, + "loss": 0.9447, + "step": 10297 + }, + { + "epoch": 0.7075718015665796, + "grad_norm": 1.144016981124878, + "learning_rate": 1.1933190354882645e-05, + "loss": 0.9623, + "step": 10298 + }, + { + "epoch": 0.7076405111996702, + "grad_norm": 1.0320897102355957, + "learning_rate": 1.1928078665279313e-05, + "loss": 0.9192, + "step": 10299 + }, + { + "epoch": 0.7077092208327608, + "grad_norm": 1.0308774709701538, + "learning_rate": 1.1922967727695658e-05, + "loss": 0.9183, + "step": 10300 + }, + { + "epoch": 0.7077779304658514, + "grad_norm": 1.1008193492889404, + "learning_rate": 1.1917857542425683e-05, + "loss": 0.9533, + "step": 10301 + }, + { + "epoch": 0.7078466400989418, + "grad_norm": 0.9869599342346191, + "learning_rate": 1.1912748109763388e-05, + "loss": 0.9284, + "step": 10302 + }, + { + "epoch": 0.7079153497320324, + "grad_norm": 1.0431915521621704, + "learning_rate": 1.1907639430002716e-05, + "loss": 0.9324, + "step": 10303 + }, + { + "epoch": 0.707984059365123, + "grad_norm": 1.1021430492401123, + "learning_rate": 1.1902531503437564e-05, + "loss": 0.966, + "step": 10304 + }, + { + "epoch": 0.7080527689982136, + "grad_norm": 1.0257787704467773, + "learning_rate": 1.1897424330361797e-05, + "loss": 0.9237, + "step": 10305 + }, + { + "epoch": 0.7081214786313041, + "grad_norm": 1.0217431783676147, + "learning_rate": 1.1892317911069212e-05, + "loss": 0.9395, + "step": 10306 + }, + { + "epoch": 0.7081901882643946, + "grad_norm": 1.0800336599349976, + "learning_rate": 1.1887212245853584e-05, + "loss": 0.9479, + "step": 10307 + }, + { + "epoch": 0.7082588978974852, + "grad_norm": 1.0628023147583008, + "learning_rate": 1.188210733500864e-05, + "loss": 0.9338, + "step": 10308 + }, + { + "epoch": 0.7083276075305758, + "grad_norm": 1.0731933116912842, + "learning_rate": 1.1877003178828064e-05, + "loss": 0.9445, + "step": 10309 + }, + { + "epoch": 0.7083963171636664, + "grad_norm": 1.057875156402588, + "learning_rate": 1.1871899777605502e-05, + "loss": 0.9037, + "step": 10310 + }, + { + "epoch": 0.7084650267967569, + "grad_norm": 1.0935338735580444, + "learning_rate": 1.186679713163453e-05, + "loss": 0.9649, + "step": 10311 + }, + { + "epoch": 0.7085337364298475, + "grad_norm": 1.066871166229248, + "learning_rate": 1.1861695241208715e-05, + "loss": 0.9212, + "step": 10312 + }, + { + "epoch": 0.708602446062938, + "grad_norm": 0.9912291765213013, + "learning_rate": 1.1856594106621558e-05, + "loss": 0.9091, + "step": 10313 + }, + { + "epoch": 0.7086711556960286, + "grad_norm": 1.0272419452667236, + "learning_rate": 1.185149372816653e-05, + "loss": 0.9309, + "step": 10314 + }, + { + "epoch": 0.7087398653291191, + "grad_norm": 1.117290735244751, + "learning_rate": 1.184639410613706e-05, + "loss": 0.9354, + "step": 10315 + }, + { + "epoch": 0.7088085749622097, + "grad_norm": 1.1061519384384155, + "learning_rate": 1.1841295240826508e-05, + "loss": 0.9616, + "step": 10316 + }, + { + "epoch": 0.7088772845953003, + "grad_norm": 1.0369411706924438, + "learning_rate": 1.1836197132528217e-05, + "loss": 0.959, + "step": 10317 + }, + { + "epoch": 0.7089459942283908, + "grad_norm": 1.0718905925750732, + "learning_rate": 1.1831099781535488e-05, + "loss": 0.917, + "step": 10318 + }, + { + "epoch": 0.7090147038614814, + "grad_norm": 1.1756998300552368, + "learning_rate": 1.1826003188141549e-05, + "loss": 0.9667, + "step": 10319 + }, + { + "epoch": 0.7090834134945719, + "grad_norm": 1.0964422225952148, + "learning_rate": 1.1820907352639616e-05, + "loss": 0.9319, + "step": 10320 + }, + { + "epoch": 0.7091521231276625, + "grad_norm": 1.1094491481781006, + "learning_rate": 1.1815812275322854e-05, + "loss": 0.937, + "step": 10321 + }, + { + "epoch": 0.7092208327607531, + "grad_norm": 1.037083625793457, + "learning_rate": 1.1810717956484366e-05, + "loss": 0.9025, + "step": 10322 + }, + { + "epoch": 0.7092895423938436, + "grad_norm": 1.104870319366455, + "learning_rate": 1.1805624396417232e-05, + "loss": 0.913, + "step": 10323 + }, + { + "epoch": 0.7093582520269341, + "grad_norm": 1.013112187385559, + "learning_rate": 1.180053159541448e-05, + "loss": 0.9158, + "step": 10324 + }, + { + "epoch": 0.7094269616600247, + "grad_norm": 1.1393145322799683, + "learning_rate": 1.1795439553769105e-05, + "loss": 0.9309, + "step": 10325 + }, + { + "epoch": 0.7094956712931153, + "grad_norm": 1.0204856395721436, + "learning_rate": 1.1790348271774036e-05, + "loss": 0.9324, + "step": 10326 + }, + { + "epoch": 0.7095643809262059, + "grad_norm": 1.0139158964157104, + "learning_rate": 1.1785257749722173e-05, + "loss": 0.9234, + "step": 10327 + }, + { + "epoch": 0.7096330905592965, + "grad_norm": 1.1070433855056763, + "learning_rate": 1.1780167987906377e-05, + "loss": 0.9415, + "step": 10328 + }, + { + "epoch": 0.7097018001923869, + "grad_norm": 1.0506809949874878, + "learning_rate": 1.1775078986619454e-05, + "loss": 0.9364, + "step": 10329 + }, + { + "epoch": 0.7097705098254775, + "grad_norm": 1.1355235576629639, + "learning_rate": 1.1769990746154183e-05, + "loss": 0.9862, + "step": 10330 + }, + { + "epoch": 0.7098392194585681, + "grad_norm": 1.0425270795822144, + "learning_rate": 1.1764903266803267e-05, + "loss": 0.9332, + "step": 10331 + }, + { + "epoch": 0.7099079290916587, + "grad_norm": 1.0530389547348022, + "learning_rate": 1.1759816548859399e-05, + "loss": 0.9244, + "step": 10332 + }, + { + "epoch": 0.7099766387247493, + "grad_norm": 1.0628525018692017, + "learning_rate": 1.1754730592615209e-05, + "loss": 0.9385, + "step": 10333 + }, + { + "epoch": 0.7100453483578397, + "grad_norm": 1.0539469718933105, + "learning_rate": 1.1749645398363297e-05, + "loss": 0.9199, + "step": 10334 + }, + { + "epoch": 0.7101140579909303, + "grad_norm": 1.0519758462905884, + "learning_rate": 1.1744560966396208e-05, + "loss": 0.9402, + "step": 10335 + }, + { + "epoch": 0.7101827676240209, + "grad_norm": 1.0859721899032593, + "learning_rate": 1.173947729700644e-05, + "loss": 0.9149, + "step": 10336 + }, + { + "epoch": 0.7102514772571115, + "grad_norm": 1.0621627569198608, + "learning_rate": 1.1734394390486455e-05, + "loss": 0.9349, + "step": 10337 + }, + { + "epoch": 0.710320186890202, + "grad_norm": 1.1278173923492432, + "learning_rate": 1.1729312247128677e-05, + "loss": 0.9293, + "step": 10338 + }, + { + "epoch": 0.7103888965232925, + "grad_norm": 1.056837558746338, + "learning_rate": 1.172423086722547e-05, + "loss": 0.9834, + "step": 10339 + }, + { + "epoch": 0.7104576061563831, + "grad_norm": 1.0500952005386353, + "learning_rate": 1.1719150251069174e-05, + "loss": 0.9242, + "step": 10340 + }, + { + "epoch": 0.7105263157894737, + "grad_norm": 1.0708461999893188, + "learning_rate": 1.1714070398952062e-05, + "loss": 0.9673, + "step": 10341 + }, + { + "epoch": 0.7105950254225643, + "grad_norm": 1.066550374031067, + "learning_rate": 1.1708991311166375e-05, + "loss": 0.9687, + "step": 10342 + }, + { + "epoch": 0.7106637350556548, + "grad_norm": 1.0756677389144897, + "learning_rate": 1.1703912988004327e-05, + "loss": 0.9286, + "step": 10343 + }, + { + "epoch": 0.7107324446887454, + "grad_norm": 1.0930988788604736, + "learning_rate": 1.1698835429758046e-05, + "loss": 0.9592, + "step": 10344 + }, + { + "epoch": 0.7108011543218359, + "grad_norm": 1.0939050912857056, + "learning_rate": 1.1693758636719652e-05, + "loss": 0.9398, + "step": 10345 + }, + { + "epoch": 0.7108698639549265, + "grad_norm": 1.0978971719741821, + "learning_rate": 1.168868260918122e-05, + "loss": 0.9402, + "step": 10346 + }, + { + "epoch": 0.710938573588017, + "grad_norm": 1.0256884098052979, + "learning_rate": 1.1683607347434753e-05, + "loss": 0.9499, + "step": 10347 + }, + { + "epoch": 0.7110072832211076, + "grad_norm": 1.1395130157470703, + "learning_rate": 1.1678532851772233e-05, + "loss": 0.9969, + "step": 10348 + }, + { + "epoch": 0.7110759928541982, + "grad_norm": 1.0519005060195923, + "learning_rate": 1.16734591224856e-05, + "loss": 0.9193, + "step": 10349 + }, + { + "epoch": 0.7111447024872887, + "grad_norm": 1.1490398645401, + "learning_rate": 1.1668386159866733e-05, + "loss": 0.965, + "step": 10350 + }, + { + "epoch": 0.7112134121203793, + "grad_norm": 1.0234878063201904, + "learning_rate": 1.1663313964207495e-05, + "loss": 0.9447, + "step": 10351 + }, + { + "epoch": 0.7112821217534698, + "grad_norm": 1.0136213302612305, + "learning_rate": 1.1658242535799662e-05, + "loss": 0.9413, + "step": 10352 + }, + { + "epoch": 0.7113508313865604, + "grad_norm": 1.0617849826812744, + "learning_rate": 1.1653171874935001e-05, + "loss": 0.9555, + "step": 10353 + }, + { + "epoch": 0.711419541019651, + "grad_norm": 1.0690511465072632, + "learning_rate": 1.1648101981905225e-05, + "loss": 0.9398, + "step": 10354 + }, + { + "epoch": 0.7114882506527415, + "grad_norm": 1.0732485055923462, + "learning_rate": 1.1643032857002001e-05, + "loss": 0.9737, + "step": 10355 + }, + { + "epoch": 0.711556960285832, + "grad_norm": 1.0012025833129883, + "learning_rate": 1.163796450051696e-05, + "loss": 0.9301, + "step": 10356 + }, + { + "epoch": 0.7116256699189226, + "grad_norm": 1.0898197889328003, + "learning_rate": 1.1632896912741664e-05, + "loss": 0.9184, + "step": 10357 + }, + { + "epoch": 0.7116943795520132, + "grad_norm": 1.1148701906204224, + "learning_rate": 1.162783009396766e-05, + "loss": 0.9549, + "step": 10358 + }, + { + "epoch": 0.7117630891851038, + "grad_norm": 1.0208441019058228, + "learning_rate": 1.1622764044486436e-05, + "loss": 0.917, + "step": 10359 + }, + { + "epoch": 0.7118317988181944, + "grad_norm": 0.9883199334144592, + "learning_rate": 1.1617698764589443e-05, + "loss": 0.9261, + "step": 10360 + }, + { + "epoch": 0.7119005084512848, + "grad_norm": 0.9960642457008362, + "learning_rate": 1.1612634254568086e-05, + "loss": 0.8625, + "step": 10361 + }, + { + "epoch": 0.7119692180843754, + "grad_norm": 1.0271233320236206, + "learning_rate": 1.160757051471371e-05, + "loss": 0.945, + "step": 10362 + }, + { + "epoch": 0.712037927717466, + "grad_norm": 1.082725167274475, + "learning_rate": 1.1602507545317637e-05, + "loss": 0.9834, + "step": 10363 + }, + { + "epoch": 0.7121066373505566, + "grad_norm": 1.0419739484786987, + "learning_rate": 1.1597445346671137e-05, + "loss": 0.9652, + "step": 10364 + }, + { + "epoch": 0.7121753469836472, + "grad_norm": 1.0847220420837402, + "learning_rate": 1.1592383919065442e-05, + "loss": 0.983, + "step": 10365 + }, + { + "epoch": 0.7122440566167376, + "grad_norm": 1.098192811012268, + "learning_rate": 1.1587323262791716e-05, + "loss": 0.9255, + "step": 10366 + }, + { + "epoch": 0.7123127662498282, + "grad_norm": 1.0677491426467896, + "learning_rate": 1.1582263378141111e-05, + "loss": 0.9486, + "step": 10367 + }, + { + "epoch": 0.7123814758829188, + "grad_norm": 1.0720473527908325, + "learning_rate": 1.1577204265404709e-05, + "loss": 0.9352, + "step": 10368 + }, + { + "epoch": 0.7124501855160094, + "grad_norm": 1.039028525352478, + "learning_rate": 1.1572145924873556e-05, + "loss": 0.9093, + "step": 10369 + }, + { + "epoch": 0.7125188951490999, + "grad_norm": 1.050854206085205, + "learning_rate": 1.1567088356838665e-05, + "loss": 0.9511, + "step": 10370 + }, + { + "epoch": 0.7125876047821904, + "grad_norm": 1.0871373414993286, + "learning_rate": 1.1562031561590996e-05, + "loss": 0.9602, + "step": 10371 + }, + { + "epoch": 0.712656314415281, + "grad_norm": 1.1137315034866333, + "learning_rate": 1.155697553942145e-05, + "loss": 0.9319, + "step": 10372 + }, + { + "epoch": 0.7127250240483716, + "grad_norm": 1.147262454032898, + "learning_rate": 1.1551920290620904e-05, + "loss": 0.9495, + "step": 10373 + }, + { + "epoch": 0.7127937336814621, + "grad_norm": 1.0246400833129883, + "learning_rate": 1.1546865815480184e-05, + "loss": 0.9471, + "step": 10374 + }, + { + "epoch": 0.7128624433145527, + "grad_norm": 1.0321416854858398, + "learning_rate": 1.154181211429007e-05, + "loss": 0.9006, + "step": 10375 + }, + { + "epoch": 0.7129311529476433, + "grad_norm": 1.073285460472107, + "learning_rate": 1.1536759187341312e-05, + "loss": 0.9229, + "step": 10376 + }, + { + "epoch": 0.7129998625807338, + "grad_norm": 1.0786052942276, + "learning_rate": 1.1531707034924577e-05, + "loss": 0.95, + "step": 10377 + }, + { + "epoch": 0.7130685722138244, + "grad_norm": 1.044491171836853, + "learning_rate": 1.1526655657330523e-05, + "loss": 0.9132, + "step": 10378 + }, + { + "epoch": 0.7131372818469149, + "grad_norm": 1.0607937574386597, + "learning_rate": 1.1521605054849758e-05, + "loss": 0.9018, + "step": 10379 + }, + { + "epoch": 0.7132059914800055, + "grad_norm": 1.0083708763122559, + "learning_rate": 1.1516555227772832e-05, + "loss": 0.913, + "step": 10380 + }, + { + "epoch": 0.7132747011130961, + "grad_norm": 1.0574110746383667, + "learning_rate": 1.1511506176390274e-05, + "loss": 0.966, + "step": 10381 + }, + { + "epoch": 0.7133434107461866, + "grad_norm": 1.0928927659988403, + "learning_rate": 1.1506457900992532e-05, + "loss": 0.9461, + "step": 10382 + }, + { + "epoch": 0.7134121203792771, + "grad_norm": 1.0110780000686646, + "learning_rate": 1.1501410401870035e-05, + "loss": 0.9147, + "step": 10383 + }, + { + "epoch": 0.7134808300123677, + "grad_norm": 1.1171927452087402, + "learning_rate": 1.149636367931317e-05, + "loss": 1.0006, + "step": 10384 + }, + { + "epoch": 0.7135495396454583, + "grad_norm": 1.0742472410202026, + "learning_rate": 1.1491317733612272e-05, + "loss": 0.9549, + "step": 10385 + }, + { + "epoch": 0.7136182492785489, + "grad_norm": 1.1105912923812866, + "learning_rate": 1.1486272565057635e-05, + "loss": 0.9318, + "step": 10386 + }, + { + "epoch": 0.7136869589116394, + "grad_norm": 1.0619312524795532, + "learning_rate": 1.1481228173939487e-05, + "loss": 0.9672, + "step": 10387 + }, + { + "epoch": 0.7137556685447299, + "grad_norm": 1.0393964052200317, + "learning_rate": 1.1476184560548042e-05, + "loss": 0.9119, + "step": 10388 + }, + { + "epoch": 0.7138243781778205, + "grad_norm": 1.0672105550765991, + "learning_rate": 1.1471141725173463e-05, + "loss": 0.904, + "step": 10389 + }, + { + "epoch": 0.7138930878109111, + "grad_norm": 1.058805227279663, + "learning_rate": 1.1466099668105837e-05, + "loss": 0.9629, + "step": 10390 + }, + { + "epoch": 0.7139617974440017, + "grad_norm": 1.1669725179672241, + "learning_rate": 1.1461058389635248e-05, + "loss": 0.9626, + "step": 10391 + }, + { + "epoch": 0.7140305070770923, + "grad_norm": 1.1334036588668823, + "learning_rate": 1.1456017890051726e-05, + "loss": 0.9138, + "step": 10392 + }, + { + "epoch": 0.7140992167101827, + "grad_norm": 1.0854016542434692, + "learning_rate": 1.1450978169645227e-05, + "loss": 0.9639, + "step": 10393 + }, + { + "epoch": 0.7141679263432733, + "grad_norm": 1.025499939918518, + "learning_rate": 1.144593922870569e-05, + "loss": 0.9008, + "step": 10394 + }, + { + "epoch": 0.7142366359763639, + "grad_norm": 1.0928813219070435, + "learning_rate": 1.1440901067523005e-05, + "loss": 0.9493, + "step": 10395 + }, + { + "epoch": 0.7143053456094545, + "grad_norm": 1.0572808980941772, + "learning_rate": 1.1435863686387024e-05, + "loss": 0.9335, + "step": 10396 + }, + { + "epoch": 0.714374055242545, + "grad_norm": 1.008244514465332, + "learning_rate": 1.1430827085587529e-05, + "loss": 0.9174, + "step": 10397 + }, + { + "epoch": 0.7144427648756355, + "grad_norm": 1.1213462352752686, + "learning_rate": 1.1425791265414273e-05, + "loss": 0.9584, + "step": 10398 + }, + { + "epoch": 0.7145114745087261, + "grad_norm": 1.0077474117279053, + "learning_rate": 1.1420756226156972e-05, + "loss": 0.9473, + "step": 10399 + }, + { + "epoch": 0.7145801841418167, + "grad_norm": 1.1187238693237305, + "learning_rate": 1.1415721968105283e-05, + "loss": 0.9368, + "step": 10400 + }, + { + "epoch": 0.7146488937749073, + "grad_norm": 1.0930172204971313, + "learning_rate": 1.141068849154883e-05, + "loss": 0.963, + "step": 10401 + }, + { + "epoch": 0.7147176034079978, + "grad_norm": 1.0345680713653564, + "learning_rate": 1.140565579677719e-05, + "loss": 0.9122, + "step": 10402 + }, + { + "epoch": 0.7147863130410883, + "grad_norm": 1.0137773752212524, + "learning_rate": 1.1400623884079873e-05, + "loss": 0.9085, + "step": 10403 + }, + { + "epoch": 0.7148550226741789, + "grad_norm": 1.0440287590026855, + "learning_rate": 1.139559275374637e-05, + "loss": 0.9186, + "step": 10404 + }, + { + "epoch": 0.7149237323072695, + "grad_norm": 1.142944097518921, + "learning_rate": 1.1390562406066125e-05, + "loss": 0.9589, + "step": 10405 + }, + { + "epoch": 0.71499244194036, + "grad_norm": 1.1423041820526123, + "learning_rate": 1.1385532841328525e-05, + "loss": 0.9433, + "step": 10406 + }, + { + "epoch": 0.7150611515734506, + "grad_norm": 1.03865385055542, + "learning_rate": 1.1380504059822927e-05, + "loss": 0.9165, + "step": 10407 + }, + { + "epoch": 0.7151298612065412, + "grad_norm": 1.0963908433914185, + "learning_rate": 1.1375476061838619e-05, + "loss": 0.9552, + "step": 10408 + }, + { + "epoch": 0.7151985708396317, + "grad_norm": 1.0531195402145386, + "learning_rate": 1.1370448847664863e-05, + "loss": 0.9669, + "step": 10409 + }, + { + "epoch": 0.7152672804727223, + "grad_norm": 1.1227391958236694, + "learning_rate": 1.1365422417590878e-05, + "loss": 0.9226, + "step": 10410 + }, + { + "epoch": 0.7153359901058128, + "grad_norm": 1.132771372795105, + "learning_rate": 1.1360396771905827e-05, + "loss": 0.9878, + "step": 10411 + }, + { + "epoch": 0.7154046997389034, + "grad_norm": 1.0415507555007935, + "learning_rate": 1.135537191089884e-05, + "loss": 0.9441, + "step": 10412 + }, + { + "epoch": 0.715473409371994, + "grad_norm": 1.0770169496536255, + "learning_rate": 1.1350347834858982e-05, + "loss": 0.9448, + "step": 10413 + }, + { + "epoch": 0.7155421190050845, + "grad_norm": 1.1010266542434692, + "learning_rate": 1.1345324544075298e-05, + "loss": 0.9218, + "step": 10414 + }, + { + "epoch": 0.715610828638175, + "grad_norm": 1.0523728132247925, + "learning_rate": 1.1340302038836756e-05, + "loss": 0.898, + "step": 10415 + }, + { + "epoch": 0.7156795382712656, + "grad_norm": 1.0772809982299805, + "learning_rate": 1.1335280319432311e-05, + "loss": 0.9664, + "step": 10416 + }, + { + "epoch": 0.7157482479043562, + "grad_norm": 1.0753710269927979, + "learning_rate": 1.1330259386150865e-05, + "loss": 0.9897, + "step": 10417 + }, + { + "epoch": 0.7158169575374468, + "grad_norm": 1.114174246788025, + "learning_rate": 1.1325239239281254e-05, + "loss": 0.951, + "step": 10418 + }, + { + "epoch": 0.7158856671705373, + "grad_norm": 1.0076642036437988, + "learning_rate": 1.1320219879112293e-05, + "loss": 0.945, + "step": 10419 + }, + { + "epoch": 0.7159543768036278, + "grad_norm": 1.1255062818527222, + "learning_rate": 1.1315201305932741e-05, + "loss": 0.9925, + "step": 10420 + }, + { + "epoch": 0.7160230864367184, + "grad_norm": 1.0621085166931152, + "learning_rate": 1.1310183520031312e-05, + "loss": 0.9521, + "step": 10421 + }, + { + "epoch": 0.716091796069809, + "grad_norm": 1.062387466430664, + "learning_rate": 1.1305166521696686e-05, + "loss": 0.9323, + "step": 10422 + }, + { + "epoch": 0.7161605057028996, + "grad_norm": 1.0299705266952515, + "learning_rate": 1.1300150311217473e-05, + "loss": 0.9065, + "step": 10423 + }, + { + "epoch": 0.7162292153359902, + "grad_norm": 1.0039163827896118, + "learning_rate": 1.1295134888882258e-05, + "loss": 0.9233, + "step": 10424 + }, + { + "epoch": 0.7162979249690806, + "grad_norm": 0.9880144596099854, + "learning_rate": 1.1290120254979577e-05, + "loss": 0.8997, + "step": 10425 + }, + { + "epoch": 0.7163666346021712, + "grad_norm": 0.9739450216293335, + "learning_rate": 1.1285106409797919e-05, + "loss": 0.8724, + "step": 10426 + }, + { + "epoch": 0.7164353442352618, + "grad_norm": 1.0496768951416016, + "learning_rate": 1.1280093353625733e-05, + "loss": 0.9321, + "step": 10427 + }, + { + "epoch": 0.7165040538683524, + "grad_norm": 1.0347297191619873, + "learning_rate": 1.1275081086751408e-05, + "loss": 0.9506, + "step": 10428 + }, + { + "epoch": 0.7165727635014429, + "grad_norm": 1.0323611497879028, + "learning_rate": 1.1270069609463293e-05, + "loss": 0.9587, + "step": 10429 + }, + { + "epoch": 0.7166414731345334, + "grad_norm": 1.067716121673584, + "learning_rate": 1.1265058922049704e-05, + "loss": 0.9291, + "step": 10430 + }, + { + "epoch": 0.716710182767624, + "grad_norm": 1.0806891918182373, + "learning_rate": 1.1260049024798902e-05, + "loss": 0.9277, + "step": 10431 + }, + { + "epoch": 0.7167788924007146, + "grad_norm": 1.1062195301055908, + "learning_rate": 1.125503991799911e-05, + "loss": 0.9388, + "step": 10432 + }, + { + "epoch": 0.7168476020338052, + "grad_norm": 1.0406098365783691, + "learning_rate": 1.1250031601938482e-05, + "loss": 0.9159, + "step": 10433 + }, + { + "epoch": 0.7169163116668957, + "grad_norm": 1.0532225370407104, + "learning_rate": 1.1245024076905153e-05, + "loss": 0.9545, + "step": 10434 + }, + { + "epoch": 0.7169850212999862, + "grad_norm": 1.1056262254714966, + "learning_rate": 1.1240017343187198e-05, + "loss": 0.9765, + "step": 10435 + }, + { + "epoch": 0.7170537309330768, + "grad_norm": 0.9924778342247009, + "learning_rate": 1.1235011401072667e-05, + "loss": 0.9202, + "step": 10436 + }, + { + "epoch": 0.7171224405661674, + "grad_norm": 1.1241153478622437, + "learning_rate": 1.1230006250849531e-05, + "loss": 0.9828, + "step": 10437 + }, + { + "epoch": 0.7171911501992579, + "grad_norm": 1.0768640041351318, + "learning_rate": 1.1225001892805736e-05, + "loss": 0.9351, + "step": 10438 + }, + { + "epoch": 0.7172598598323485, + "grad_norm": 1.0659903287887573, + "learning_rate": 1.1219998327229192e-05, + "loss": 0.9283, + "step": 10439 + }, + { + "epoch": 0.7173285694654391, + "grad_norm": 1.0611958503723145, + "learning_rate": 1.1214995554407735e-05, + "loss": 0.9374, + "step": 10440 + }, + { + "epoch": 0.7173972790985296, + "grad_norm": 1.0813357830047607, + "learning_rate": 1.1209993574629176e-05, + "loss": 0.9573, + "step": 10441 + }, + { + "epoch": 0.7174659887316202, + "grad_norm": 1.1247265338897705, + "learning_rate": 1.120499238818129e-05, + "loss": 0.9479, + "step": 10442 + }, + { + "epoch": 0.7175346983647107, + "grad_norm": 1.0503910779953003, + "learning_rate": 1.1199991995351769e-05, + "loss": 0.9164, + "step": 10443 + }, + { + "epoch": 0.7176034079978013, + "grad_norm": 1.0569065809249878, + "learning_rate": 1.1194992396428297e-05, + "loss": 0.935, + "step": 10444 + }, + { + "epoch": 0.7176721176308919, + "grad_norm": 1.0360028743743896, + "learning_rate": 1.1189993591698494e-05, + "loss": 0.9243, + "step": 10445 + }, + { + "epoch": 0.7177408272639824, + "grad_norm": 1.105680227279663, + "learning_rate": 1.118499558144994e-05, + "loss": 0.9574, + "step": 10446 + }, + { + "epoch": 0.7178095368970729, + "grad_norm": 1.1030470132827759, + "learning_rate": 1.1179998365970174e-05, + "loss": 0.9384, + "step": 10447 + }, + { + "epoch": 0.7178782465301635, + "grad_norm": 1.086781620979309, + "learning_rate": 1.1175001945546665e-05, + "loss": 0.9574, + "step": 10448 + }, + { + "epoch": 0.7179469561632541, + "grad_norm": 1.0652399063110352, + "learning_rate": 1.1170006320466867e-05, + "loss": 0.9498, + "step": 10449 + }, + { + "epoch": 0.7180156657963447, + "grad_norm": 1.0587466955184937, + "learning_rate": 1.1165011491018173e-05, + "loss": 0.9629, + "step": 10450 + }, + { + "epoch": 0.7180843754294353, + "grad_norm": 1.0532668828964233, + "learning_rate": 1.116001745748793e-05, + "loss": 0.9624, + "step": 10451 + }, + { + "epoch": 0.7181530850625257, + "grad_norm": 1.0259840488433838, + "learning_rate": 1.1155024220163456e-05, + "loss": 0.9482, + "step": 10452 + }, + { + "epoch": 0.7182217946956163, + "grad_norm": 1.0133285522460938, + "learning_rate": 1.1150031779331987e-05, + "loss": 0.9139, + "step": 10453 + }, + { + "epoch": 0.7182905043287069, + "grad_norm": 1.020128607749939, + "learning_rate": 1.1145040135280745e-05, + "loss": 0.9356, + "step": 10454 + }, + { + "epoch": 0.7183592139617975, + "grad_norm": 1.0285550355911255, + "learning_rate": 1.1140049288296894e-05, + "loss": 0.9097, + "step": 10455 + }, + { + "epoch": 0.718427923594888, + "grad_norm": 1.0672239065170288, + "learning_rate": 1.1135059238667558e-05, + "loss": 0.9102, + "step": 10456 + }, + { + "epoch": 0.7184966332279785, + "grad_norm": 1.0073384046554565, + "learning_rate": 1.113006998667981e-05, + "loss": 0.9289, + "step": 10457 + }, + { + "epoch": 0.7185653428610691, + "grad_norm": 1.0639455318450928, + "learning_rate": 1.112508153262069e-05, + "loss": 0.9618, + "step": 10458 + }, + { + "epoch": 0.7186340524941597, + "grad_norm": 1.065381407737732, + "learning_rate": 1.112009387677716e-05, + "loss": 0.9247, + "step": 10459 + }, + { + "epoch": 0.7187027621272503, + "grad_norm": 1.0555601119995117, + "learning_rate": 1.1115107019436164e-05, + "loss": 0.8994, + "step": 10460 + }, + { + "epoch": 0.7187714717603408, + "grad_norm": 1.0626225471496582, + "learning_rate": 1.1110120960884604e-05, + "loss": 0.9337, + "step": 10461 + }, + { + "epoch": 0.7188401813934313, + "grad_norm": 1.0634663105010986, + "learning_rate": 1.1105135701409309e-05, + "loss": 0.9494, + "step": 10462 + }, + { + "epoch": 0.7189088910265219, + "grad_norm": 1.0092283487319946, + "learning_rate": 1.1100151241297091e-05, + "loss": 0.904, + "step": 10463 + }, + { + "epoch": 0.7189776006596125, + "grad_norm": 0.985611081123352, + "learning_rate": 1.1095167580834691e-05, + "loss": 0.9204, + "step": 10464 + }, + { + "epoch": 0.719046310292703, + "grad_norm": 1.098615288734436, + "learning_rate": 1.1090184720308821e-05, + "loss": 0.9512, + "step": 10465 + }, + { + "epoch": 0.7191150199257936, + "grad_norm": 1.058921217918396, + "learning_rate": 1.1085202660006142e-05, + "loss": 0.9388, + "step": 10466 + }, + { + "epoch": 0.7191837295588842, + "grad_norm": 1.0878654718399048, + "learning_rate": 1.108022140021327e-05, + "loss": 0.9615, + "step": 10467 + }, + { + "epoch": 0.7192524391919747, + "grad_norm": 0.974331796169281, + "learning_rate": 1.1075240941216783e-05, + "loss": 0.9454, + "step": 10468 + }, + { + "epoch": 0.7193211488250653, + "grad_norm": 1.0636093616485596, + "learning_rate": 1.1070261283303184e-05, + "loss": 0.9913, + "step": 10469 + }, + { + "epoch": 0.7193898584581558, + "grad_norm": 1.092969536781311, + "learning_rate": 1.1065282426758961e-05, + "loss": 0.9577, + "step": 10470 + }, + { + "epoch": 0.7194585680912464, + "grad_norm": 1.112321138381958, + "learning_rate": 1.1060304371870541e-05, + "loss": 0.9611, + "step": 10471 + }, + { + "epoch": 0.719527277724337, + "grad_norm": 1.0782049894332886, + "learning_rate": 1.1055327118924313e-05, + "loss": 0.9517, + "step": 10472 + }, + { + "epoch": 0.7195959873574275, + "grad_norm": 1.1723222732543945, + "learning_rate": 1.1050350668206625e-05, + "loss": 0.9652, + "step": 10473 + }, + { + "epoch": 0.719664696990518, + "grad_norm": 1.0597472190856934, + "learning_rate": 1.1045375020003746e-05, + "loss": 0.9373, + "step": 10474 + }, + { + "epoch": 0.7197334066236086, + "grad_norm": 1.1114006042480469, + "learning_rate": 1.1040400174601934e-05, + "loss": 0.9777, + "step": 10475 + }, + { + "epoch": 0.7198021162566992, + "grad_norm": 1.0244944095611572, + "learning_rate": 1.1035426132287385e-05, + "loss": 0.9264, + "step": 10476 + }, + { + "epoch": 0.7198708258897898, + "grad_norm": 1.032484531402588, + "learning_rate": 1.103045289334626e-05, + "loss": 0.9317, + "step": 10477 + }, + { + "epoch": 0.7199395355228803, + "grad_norm": 1.126259684562683, + "learning_rate": 1.102548045806467e-05, + "loss": 0.9766, + "step": 10478 + }, + { + "epoch": 0.7200082451559708, + "grad_norm": 1.0568069219589233, + "learning_rate": 1.1020508826728659e-05, + "loss": 0.96, + "step": 10479 + }, + { + "epoch": 0.7200769547890614, + "grad_norm": 1.0708178281784058, + "learning_rate": 1.1015537999624253e-05, + "loss": 0.9267, + "step": 10480 + }, + { + "epoch": 0.720145664422152, + "grad_norm": 1.055657148361206, + "learning_rate": 1.1010567977037416e-05, + "loss": 0.9435, + "step": 10481 + }, + { + "epoch": 0.7202143740552426, + "grad_norm": 1.0735838413238525, + "learning_rate": 1.1005598759254077e-05, + "loss": 0.9518, + "step": 10482 + }, + { + "epoch": 0.7202830836883332, + "grad_norm": 1.0683715343475342, + "learning_rate": 1.1000630346560117e-05, + "loss": 0.9176, + "step": 10483 + }, + { + "epoch": 0.7203517933214236, + "grad_norm": 0.9923836588859558, + "learning_rate": 1.0995662739241347e-05, + "loss": 0.9227, + "step": 10484 + }, + { + "epoch": 0.7204205029545142, + "grad_norm": 1.0548421144485474, + "learning_rate": 1.099069593758357e-05, + "loss": 0.9489, + "step": 10485 + }, + { + "epoch": 0.7204892125876048, + "grad_norm": 1.047763705253601, + "learning_rate": 1.0985729941872506e-05, + "loss": 0.9524, + "step": 10486 + }, + { + "epoch": 0.7205579222206954, + "grad_norm": 1.0630903244018555, + "learning_rate": 1.0980764752393853e-05, + "loss": 0.9469, + "step": 10487 + }, + { + "epoch": 0.720626631853786, + "grad_norm": 1.0679905414581299, + "learning_rate": 1.0975800369433265e-05, + "loss": 0.9356, + "step": 10488 + }, + { + "epoch": 0.7206953414868764, + "grad_norm": 1.0127553939819336, + "learning_rate": 1.097083679327632e-05, + "loss": 0.9339, + "step": 10489 + }, + { + "epoch": 0.720764051119967, + "grad_norm": 1.0952990055084229, + "learning_rate": 1.0965874024208581e-05, + "loss": 0.9633, + "step": 10490 + }, + { + "epoch": 0.7208327607530576, + "grad_norm": 1.0864876508712769, + "learning_rate": 1.0960912062515553e-05, + "loss": 0.9278, + "step": 10491 + }, + { + "epoch": 0.7209014703861482, + "grad_norm": 1.0361995697021484, + "learning_rate": 1.0955950908482694e-05, + "loss": 0.9608, + "step": 10492 + }, + { + "epoch": 0.7209701800192387, + "grad_norm": 1.037843108177185, + "learning_rate": 1.0950990562395425e-05, + "loss": 0.9083, + "step": 10493 + }, + { + "epoch": 0.7210388896523292, + "grad_norm": 0.9975468516349792, + "learning_rate": 1.0946031024539091e-05, + "loss": 0.9193, + "step": 10494 + }, + { + "epoch": 0.7211075992854198, + "grad_norm": 0.9984990954399109, + "learning_rate": 1.0941072295199025e-05, + "loss": 0.9302, + "step": 10495 + }, + { + "epoch": 0.7211763089185104, + "grad_norm": 1.0564992427825928, + "learning_rate": 1.09361143746605e-05, + "loss": 0.9947, + "step": 10496 + }, + { + "epoch": 0.721245018551601, + "grad_norm": 1.1362237930297852, + "learning_rate": 1.0931157263208736e-05, + "loss": 0.9391, + "step": 10497 + }, + { + "epoch": 0.7213137281846915, + "grad_norm": 1.0328575372695923, + "learning_rate": 1.0926200961128926e-05, + "loss": 0.8578, + "step": 10498 + }, + { + "epoch": 0.7213824378177821, + "grad_norm": 1.1271284818649292, + "learning_rate": 1.0921245468706187e-05, + "loss": 0.975, + "step": 10499 + }, + { + "epoch": 0.7214511474508726, + "grad_norm": 1.0388541221618652, + "learning_rate": 1.0916290786225611e-05, + "loss": 0.9209, + "step": 10500 + }, + { + "epoch": 0.7215198570839632, + "grad_norm": 1.1020371913909912, + "learning_rate": 1.091133691397224e-05, + "loss": 0.9376, + "step": 10501 + }, + { + "epoch": 0.7215885667170537, + "grad_norm": 1.172690510749817, + "learning_rate": 1.0906383852231064e-05, + "loss": 0.977, + "step": 10502 + }, + { + "epoch": 0.7216572763501443, + "grad_norm": 1.0447735786437988, + "learning_rate": 1.0901431601287041e-05, + "loss": 0.9272, + "step": 10503 + }, + { + "epoch": 0.7217259859832349, + "grad_norm": 1.0573700666427612, + "learning_rate": 1.0896480161425052e-05, + "loss": 0.948, + "step": 10504 + }, + { + "epoch": 0.7217946956163254, + "grad_norm": 1.0940039157867432, + "learning_rate": 1.0891529532929959e-05, + "loss": 0.8646, + "step": 10505 + }, + { + "epoch": 0.721863405249416, + "grad_norm": 1.0804256200790405, + "learning_rate": 1.0886579716086572e-05, + "loss": 0.9323, + "step": 10506 + }, + { + "epoch": 0.7219321148825065, + "grad_norm": 1.09125816822052, + "learning_rate": 1.0881630711179658e-05, + "loss": 0.9046, + "step": 10507 + }, + { + "epoch": 0.7220008245155971, + "grad_norm": 1.150881052017212, + "learning_rate": 1.0876682518493908e-05, + "loss": 0.9445, + "step": 10508 + }, + { + "epoch": 0.7220695341486877, + "grad_norm": 1.0725876092910767, + "learning_rate": 1.0871735138314004e-05, + "loss": 0.9231, + "step": 10509 + }, + { + "epoch": 0.7221382437817782, + "grad_norm": 1.0940682888031006, + "learning_rate": 1.0866788570924572e-05, + "loss": 0.9307, + "step": 10510 + }, + { + "epoch": 0.7222069534148687, + "grad_norm": 1.1284595727920532, + "learning_rate": 1.0861842816610168e-05, + "loss": 0.9662, + "step": 10511 + }, + { + "epoch": 0.7222756630479593, + "grad_norm": 1.0751370191574097, + "learning_rate": 1.0856897875655325e-05, + "loss": 0.9492, + "step": 10512 + }, + { + "epoch": 0.7223443726810499, + "grad_norm": 1.0267751216888428, + "learning_rate": 1.0851953748344524e-05, + "loss": 0.926, + "step": 10513 + }, + { + "epoch": 0.7224130823141405, + "grad_norm": 1.020842432975769, + "learning_rate": 1.0847010434962205e-05, + "loss": 0.94, + "step": 10514 + }, + { + "epoch": 0.7224817919472311, + "grad_norm": 1.0439521074295044, + "learning_rate": 1.0842067935792737e-05, + "loss": 0.9064, + "step": 10515 + }, + { + "epoch": 0.7225505015803215, + "grad_norm": 1.026591181755066, + "learning_rate": 1.0837126251120469e-05, + "loss": 0.9014, + "step": 10516 + }, + { + "epoch": 0.7226192112134121, + "grad_norm": 1.0867438316345215, + "learning_rate": 1.0832185381229692e-05, + "loss": 0.9348, + "step": 10517 + }, + { + "epoch": 0.7226879208465027, + "grad_norm": 1.0612016916275024, + "learning_rate": 1.0827245326404653e-05, + "loss": 0.9287, + "step": 10518 + }, + { + "epoch": 0.7227566304795933, + "grad_norm": 1.025403380393982, + "learning_rate": 1.0822306086929556e-05, + "loss": 0.9516, + "step": 10519 + }, + { + "epoch": 0.7228253401126838, + "grad_norm": 1.0987098217010498, + "learning_rate": 1.0817367663088537e-05, + "loss": 0.9444, + "step": 10520 + }, + { + "epoch": 0.7228940497457743, + "grad_norm": 1.0717564821243286, + "learning_rate": 1.081243005516571e-05, + "loss": 0.9431, + "step": 10521 + }, + { + "epoch": 0.7229627593788649, + "grad_norm": 1.0557794570922852, + "learning_rate": 1.080749326344513e-05, + "loss": 0.9149, + "step": 10522 + }, + { + "epoch": 0.7230314690119555, + "grad_norm": 1.0930571556091309, + "learning_rate": 1.0802557288210807e-05, + "loss": 0.9559, + "step": 10523 + }, + { + "epoch": 0.723100178645046, + "grad_norm": 1.1836634874343872, + "learning_rate": 1.0797622129746718e-05, + "loss": 0.9558, + "step": 10524 + }, + { + "epoch": 0.7231688882781366, + "grad_norm": 1.1008373498916626, + "learning_rate": 1.0792687788336758e-05, + "loss": 0.9553, + "step": 10525 + }, + { + "epoch": 0.7232375979112271, + "grad_norm": 1.0515031814575195, + "learning_rate": 1.0787754264264807e-05, + "loss": 0.9233, + "step": 10526 + }, + { + "epoch": 0.7233063075443177, + "grad_norm": 0.9983488321304321, + "learning_rate": 1.0782821557814688e-05, + "loss": 0.935, + "step": 10527 + }, + { + "epoch": 0.7233750171774083, + "grad_norm": 1.1054335832595825, + "learning_rate": 1.0777889669270177e-05, + "loss": 0.9602, + "step": 10528 + }, + { + "epoch": 0.7234437268104988, + "grad_norm": 1.0797756910324097, + "learning_rate": 1.0772958598915008e-05, + "loss": 0.922, + "step": 10529 + }, + { + "epoch": 0.7235124364435894, + "grad_norm": 1.0118011236190796, + "learning_rate": 1.0768028347032848e-05, + "loss": 0.921, + "step": 10530 + }, + { + "epoch": 0.72358114607668, + "grad_norm": 1.0060865879058838, + "learning_rate": 1.0763098913907343e-05, + "loss": 0.9054, + "step": 10531 + }, + { + "epoch": 0.7236498557097705, + "grad_norm": 1.0509040355682373, + "learning_rate": 1.0758170299822082e-05, + "loss": 0.9634, + "step": 10532 + }, + { + "epoch": 0.723718565342861, + "grad_norm": 1.0726872682571411, + "learning_rate": 1.0753242505060594e-05, + "loss": 0.9377, + "step": 10533 + }, + { + "epoch": 0.7237872749759516, + "grad_norm": 1.0992424488067627, + "learning_rate": 1.0748315529906374e-05, + "loss": 0.9687, + "step": 10534 + }, + { + "epoch": 0.7238559846090422, + "grad_norm": 1.1008166074752808, + "learning_rate": 1.0743389374642882e-05, + "loss": 0.955, + "step": 10535 + }, + { + "epoch": 0.7239246942421328, + "grad_norm": 1.110621690750122, + "learning_rate": 1.07384640395535e-05, + "loss": 0.9349, + "step": 10536 + }, + { + "epoch": 0.7239934038752233, + "grad_norm": 1.0179458856582642, + "learning_rate": 1.0733539524921584e-05, + "loss": 0.9334, + "step": 10537 + }, + { + "epoch": 0.7240621135083138, + "grad_norm": 1.0973821878433228, + "learning_rate": 1.0728615831030439e-05, + "loss": 0.9417, + "step": 10538 + }, + { + "epoch": 0.7241308231414044, + "grad_norm": 1.0815573930740356, + "learning_rate": 1.0723692958163336e-05, + "loss": 0.9658, + "step": 10539 + }, + { + "epoch": 0.724199532774495, + "grad_norm": 1.0889575481414795, + "learning_rate": 1.0718770906603463e-05, + "loss": 0.9233, + "step": 10540 + }, + { + "epoch": 0.7242682424075856, + "grad_norm": 1.0724064111709595, + "learning_rate": 1.0713849676633989e-05, + "loss": 0.9701, + "step": 10541 + }, + { + "epoch": 0.724336952040676, + "grad_norm": 1.0835145711898804, + "learning_rate": 1.0708929268538034e-05, + "loss": 0.9931, + "step": 10542 + }, + { + "epoch": 0.7244056616737666, + "grad_norm": 1.097731351852417, + "learning_rate": 1.0704009682598664e-05, + "loss": 0.9434, + "step": 10543 + }, + { + "epoch": 0.7244743713068572, + "grad_norm": 1.0263519287109375, + "learning_rate": 1.0699090919098908e-05, + "loss": 0.9215, + "step": 10544 + }, + { + "epoch": 0.7245430809399478, + "grad_norm": 1.095165491104126, + "learning_rate": 1.0694172978321726e-05, + "loss": 0.9782, + "step": 10545 + }, + { + "epoch": 0.7246117905730384, + "grad_norm": 1.0786579847335815, + "learning_rate": 1.068925586055005e-05, + "loss": 0.9118, + "step": 10546 + }, + { + "epoch": 0.724680500206129, + "grad_norm": 1.0628483295440674, + "learning_rate": 1.0684339566066756e-05, + "loss": 0.9336, + "step": 10547 + }, + { + "epoch": 0.7247492098392194, + "grad_norm": 1.1294699907302856, + "learning_rate": 1.0679424095154677e-05, + "loss": 0.9403, + "step": 10548 + }, + { + "epoch": 0.72481791947231, + "grad_norm": 1.02962064743042, + "learning_rate": 1.067450944809661e-05, + "loss": 0.9337, + "step": 10549 + }, + { + "epoch": 0.7248866291054006, + "grad_norm": 1.1060630083084106, + "learning_rate": 1.0669595625175269e-05, + "loss": 0.9504, + "step": 10550 + }, + { + "epoch": 0.7249553387384912, + "grad_norm": 1.0140546560287476, + "learning_rate": 1.0664682626673356e-05, + "loss": 0.8766, + "step": 10551 + }, + { + "epoch": 0.7250240483715817, + "grad_norm": 1.0908913612365723, + "learning_rate": 1.0659770452873508e-05, + "loss": 0.9201, + "step": 10552 + }, + { + "epoch": 0.7250927580046722, + "grad_norm": 1.047551155090332, + "learning_rate": 1.0654859104058323e-05, + "loss": 0.9465, + "step": 10553 + }, + { + "epoch": 0.7251614676377628, + "grad_norm": 1.1084462404251099, + "learning_rate": 1.0649948580510358e-05, + "loss": 0.9573, + "step": 10554 + }, + { + "epoch": 0.7252301772708534, + "grad_norm": 1.061165452003479, + "learning_rate": 1.064503888251209e-05, + "loss": 0.9246, + "step": 10555 + }, + { + "epoch": 0.725298886903944, + "grad_norm": 1.004926323890686, + "learning_rate": 1.0640130010345983e-05, + "loss": 0.908, + "step": 10556 + }, + { + "epoch": 0.7253675965370345, + "grad_norm": 1.0437192916870117, + "learning_rate": 1.063522196429445e-05, + "loss": 0.9168, + "step": 10557 + }, + { + "epoch": 0.725436306170125, + "grad_norm": 1.0477526187896729, + "learning_rate": 1.063031474463983e-05, + "loss": 0.966, + "step": 10558 + }, + { + "epoch": 0.7255050158032156, + "grad_norm": 1.0297839641571045, + "learning_rate": 1.0625408351664451e-05, + "loss": 0.9418, + "step": 10559 + }, + { + "epoch": 0.7255737254363062, + "grad_norm": 1.0388331413269043, + "learning_rate": 1.0620502785650555e-05, + "loss": 0.9359, + "step": 10560 + }, + { + "epoch": 0.7256424350693967, + "grad_norm": 1.0470805168151855, + "learning_rate": 1.0615598046880365e-05, + "loss": 0.9311, + "step": 10561 + }, + { + "epoch": 0.7257111447024873, + "grad_norm": 1.0756930112838745, + "learning_rate": 1.0610694135636052e-05, + "loss": 0.982, + "step": 10562 + }, + { + "epoch": 0.7257798543355779, + "grad_norm": 1.079820990562439, + "learning_rate": 1.0605791052199728e-05, + "loss": 0.9796, + "step": 10563 + }, + { + "epoch": 0.7258485639686684, + "grad_norm": 1.0384831428527832, + "learning_rate": 1.0600888796853478e-05, + "loss": 0.9133, + "step": 10564 + }, + { + "epoch": 0.725917273601759, + "grad_norm": 1.0266402959823608, + "learning_rate": 1.0595987369879306e-05, + "loss": 0.9122, + "step": 10565 + }, + { + "epoch": 0.7259859832348495, + "grad_norm": 1.0278812646865845, + "learning_rate": 1.05910867715592e-05, + "loss": 0.9651, + "step": 10566 + }, + { + "epoch": 0.7260546928679401, + "grad_norm": 1.0190808773040771, + "learning_rate": 1.0586187002175087e-05, + "loss": 0.9215, + "step": 10567 + }, + { + "epoch": 0.7261234025010307, + "grad_norm": 1.0157463550567627, + "learning_rate": 1.0581288062008846e-05, + "loss": 0.9407, + "step": 10568 + }, + { + "epoch": 0.7261921121341212, + "grad_norm": 1.0178042650222778, + "learning_rate": 1.0576389951342311e-05, + "loss": 0.9372, + "step": 10569 + }, + { + "epoch": 0.7262608217672117, + "grad_norm": 1.0981800556182861, + "learning_rate": 1.0571492670457278e-05, + "loss": 0.9702, + "step": 10570 + }, + { + "epoch": 0.7263295314003023, + "grad_norm": 1.0311137437820435, + "learning_rate": 1.0566596219635464e-05, + "loss": 0.9207, + "step": 10571 + }, + { + "epoch": 0.7263982410333929, + "grad_norm": 1.0551267862319946, + "learning_rate": 1.056170059915857e-05, + "loss": 0.9468, + "step": 10572 + }, + { + "epoch": 0.7264669506664835, + "grad_norm": 1.0515937805175781, + "learning_rate": 1.0556805809308235e-05, + "loss": 0.9618, + "step": 10573 + }, + { + "epoch": 0.726535660299574, + "grad_norm": 1.0586750507354736, + "learning_rate": 1.0551911850366056e-05, + "loss": 0.9452, + "step": 10574 + }, + { + "epoch": 0.7266043699326645, + "grad_norm": 1.061012625694275, + "learning_rate": 1.0547018722613589e-05, + "loss": 0.9396, + "step": 10575 + }, + { + "epoch": 0.7266730795657551, + "grad_norm": 1.0379128456115723, + "learning_rate": 1.054212642633231e-05, + "loss": 0.8957, + "step": 10576 + }, + { + "epoch": 0.7267417891988457, + "grad_norm": 1.0411511659622192, + "learning_rate": 1.0537234961803685e-05, + "loss": 0.9337, + "step": 10577 + }, + { + "epoch": 0.7268104988319363, + "grad_norm": 1.0393060445785522, + "learning_rate": 1.0532344329309116e-05, + "loss": 0.9519, + "step": 10578 + }, + { + "epoch": 0.7268792084650268, + "grad_norm": 1.072749376296997, + "learning_rate": 1.0527454529129962e-05, + "loss": 0.9877, + "step": 10579 + }, + { + "epoch": 0.7269479180981173, + "grad_norm": 1.0455228090286255, + "learning_rate": 1.0522565561547518e-05, + "loss": 0.9202, + "step": 10580 + }, + { + "epoch": 0.7270166277312079, + "grad_norm": 1.0579915046691895, + "learning_rate": 1.0517677426843056e-05, + "loss": 0.9778, + "step": 10581 + }, + { + "epoch": 0.7270853373642985, + "grad_norm": 1.0857168436050415, + "learning_rate": 1.0512790125297775e-05, + "loss": 0.9348, + "step": 10582 + }, + { + "epoch": 0.7271540469973891, + "grad_norm": 1.0034801959991455, + "learning_rate": 1.0507903657192845e-05, + "loss": 0.8988, + "step": 10583 + }, + { + "epoch": 0.7272227566304796, + "grad_norm": 1.053571105003357, + "learning_rate": 1.0503018022809378e-05, + "loss": 0.9294, + "step": 10584 + }, + { + "epoch": 0.7272914662635701, + "grad_norm": 1.1243093013763428, + "learning_rate": 1.0498133222428455e-05, + "loss": 0.9894, + "step": 10585 + }, + { + "epoch": 0.7273601758966607, + "grad_norm": 1.0732489824295044, + "learning_rate": 1.0493249256331078e-05, + "loss": 1.008, + "step": 10586 + }, + { + "epoch": 0.7274288855297513, + "grad_norm": 0.9992889165878296, + "learning_rate": 1.0488366124798224e-05, + "loss": 0.9584, + "step": 10587 + }, + { + "epoch": 0.7274975951628418, + "grad_norm": 1.0084904432296753, + "learning_rate": 1.0483483828110821e-05, + "loss": 0.9082, + "step": 10588 + }, + { + "epoch": 0.7275663047959324, + "grad_norm": 1.0727334022521973, + "learning_rate": 1.0478602366549742e-05, + "loss": 0.933, + "step": 10589 + }, + { + "epoch": 0.7276350144290229, + "grad_norm": 1.0505907535552979, + "learning_rate": 1.0473721740395823e-05, + "loss": 0.9312, + "step": 10590 + }, + { + "epoch": 0.7277037240621135, + "grad_norm": 1.0640655755996704, + "learning_rate": 1.0468841949929825e-05, + "loss": 0.9396, + "step": 10591 + }, + { + "epoch": 0.7277724336952041, + "grad_norm": 1.0447934865951538, + "learning_rate": 1.046396299543249e-05, + "loss": 0.9833, + "step": 10592 + }, + { + "epoch": 0.7278411433282946, + "grad_norm": 1.0223581790924072, + "learning_rate": 1.0459084877184502e-05, + "loss": 0.9103, + "step": 10593 + }, + { + "epoch": 0.7279098529613852, + "grad_norm": 1.0398589372634888, + "learning_rate": 1.0454207595466494e-05, + "loss": 0.9354, + "step": 10594 + }, + { + "epoch": 0.7279785625944758, + "grad_norm": 1.065869688987732, + "learning_rate": 1.0449331150559064e-05, + "loss": 0.9131, + "step": 10595 + }, + { + "epoch": 0.7280472722275663, + "grad_norm": 1.0658668279647827, + "learning_rate": 1.0444455542742729e-05, + "loss": 0.9501, + "step": 10596 + }, + { + "epoch": 0.7281159818606568, + "grad_norm": 1.028347134590149, + "learning_rate": 1.0439580772297993e-05, + "loss": 0.928, + "step": 10597 + }, + { + "epoch": 0.7281846914937474, + "grad_norm": 1.071157693862915, + "learning_rate": 1.0434706839505298e-05, + "loss": 0.9357, + "step": 10598 + }, + { + "epoch": 0.728253401126838, + "grad_norm": 1.0290151834487915, + "learning_rate": 1.042983374464504e-05, + "loss": 0.9262, + "step": 10599 + }, + { + "epoch": 0.7283221107599286, + "grad_norm": 1.1622487306594849, + "learning_rate": 1.0424961487997568e-05, + "loss": 0.9242, + "step": 10600 + }, + { + "epoch": 0.7283908203930191, + "grad_norm": 1.090722680091858, + "learning_rate": 1.0420090069843167e-05, + "loss": 0.9388, + "step": 10601 + }, + { + "epoch": 0.7284595300261096, + "grad_norm": 1.1016205549240112, + "learning_rate": 1.0415219490462094e-05, + "loss": 0.9213, + "step": 10602 + }, + { + "epoch": 0.7285282396592002, + "grad_norm": 1.046151876449585, + "learning_rate": 1.041034975013456e-05, + "loss": 0.9325, + "step": 10603 + }, + { + "epoch": 0.7285969492922908, + "grad_norm": 1.061715841293335, + "learning_rate": 1.04054808491407e-05, + "loss": 0.9349, + "step": 10604 + }, + { + "epoch": 0.7286656589253814, + "grad_norm": 1.0297428369522095, + "learning_rate": 1.040061278776063e-05, + "loss": 0.887, + "step": 10605 + }, + { + "epoch": 0.7287343685584718, + "grad_norm": 1.0572718381881714, + "learning_rate": 1.0395745566274411e-05, + "loss": 0.9829, + "step": 10606 + }, + { + "epoch": 0.7288030781915624, + "grad_norm": 0.9960128664970398, + "learning_rate": 1.0390879184962034e-05, + "loss": 0.9179, + "step": 10607 + }, + { + "epoch": 0.728871787824653, + "grad_norm": 1.0863250494003296, + "learning_rate": 1.0386013644103476e-05, + "loss": 0.9865, + "step": 10608 + }, + { + "epoch": 0.7289404974577436, + "grad_norm": 1.089569330215454, + "learning_rate": 1.0381148943978639e-05, + "loss": 0.9213, + "step": 10609 + }, + { + "epoch": 0.7290092070908342, + "grad_norm": 1.0603888034820557, + "learning_rate": 1.0376285084867399e-05, + "loss": 0.9053, + "step": 10610 + }, + { + "epoch": 0.7290779167239247, + "grad_norm": 1.1037791967391968, + "learning_rate": 1.0371422067049551e-05, + "loss": 0.9194, + "step": 10611 + }, + { + "epoch": 0.7291466263570152, + "grad_norm": 1.1256232261657715, + "learning_rate": 1.0366559890804872e-05, + "loss": 0.9455, + "step": 10612 + }, + { + "epoch": 0.7292153359901058, + "grad_norm": 1.0740755796432495, + "learning_rate": 1.0361698556413082e-05, + "loss": 0.9187, + "step": 10613 + }, + { + "epoch": 0.7292840456231964, + "grad_norm": 1.0419055223464966, + "learning_rate": 1.0356838064153848e-05, + "loss": 0.9568, + "step": 10614 + }, + { + "epoch": 0.729352755256287, + "grad_norm": 1.064253807067871, + "learning_rate": 1.0351978414306801e-05, + "loss": 0.9516, + "step": 10615 + }, + { + "epoch": 0.7294214648893775, + "grad_norm": 1.0535125732421875, + "learning_rate": 1.0347119607151495e-05, + "loss": 0.9542, + "step": 10616 + }, + { + "epoch": 0.729490174522468, + "grad_norm": 1.0652676820755005, + "learning_rate": 1.0342261642967463e-05, + "loss": 0.9663, + "step": 10617 + }, + { + "epoch": 0.7295588841555586, + "grad_norm": 1.1213279962539673, + "learning_rate": 1.033740452203418e-05, + "loss": 0.9219, + "step": 10618 + }, + { + "epoch": 0.7296275937886492, + "grad_norm": 1.0792313814163208, + "learning_rate": 1.0332548244631077e-05, + "loss": 0.9307, + "step": 10619 + }, + { + "epoch": 0.7296963034217397, + "grad_norm": 1.0404101610183716, + "learning_rate": 1.0327692811037531e-05, + "loss": 0.9467, + "step": 10620 + }, + { + "epoch": 0.7297650130548303, + "grad_norm": 1.082176923751831, + "learning_rate": 1.032283822153288e-05, + "loss": 0.9589, + "step": 10621 + }, + { + "epoch": 0.7298337226879208, + "grad_norm": 1.0560648441314697, + "learning_rate": 1.0317984476396386e-05, + "loss": 0.9334, + "step": 10622 + }, + { + "epoch": 0.7299024323210114, + "grad_norm": 1.037778615951538, + "learning_rate": 1.0313131575907295e-05, + "loss": 0.9463, + "step": 10623 + }, + { + "epoch": 0.729971141954102, + "grad_norm": 0.9987453818321228, + "learning_rate": 1.0308279520344788e-05, + "loss": 0.9016, + "step": 10624 + }, + { + "epoch": 0.7300398515871925, + "grad_norm": 1.0841368436813354, + "learning_rate": 1.0303428309988011e-05, + "loss": 0.9436, + "step": 10625 + }, + { + "epoch": 0.7301085612202831, + "grad_norm": 1.030166745185852, + "learning_rate": 1.0298577945116034e-05, + "loss": 0.926, + "step": 10626 + }, + { + "epoch": 0.7301772708533737, + "grad_norm": 1.159122347831726, + "learning_rate": 1.0293728426007903e-05, + "loss": 0.9629, + "step": 10627 + }, + { + "epoch": 0.7302459804864642, + "grad_norm": 1.1523473262786865, + "learning_rate": 1.0288879752942618e-05, + "loss": 0.9342, + "step": 10628 + }, + { + "epoch": 0.7303146901195547, + "grad_norm": 1.0791006088256836, + "learning_rate": 1.0284031926199103e-05, + "loss": 0.9277, + "step": 10629 + }, + { + "epoch": 0.7303833997526453, + "grad_norm": 1.0852370262145996, + "learning_rate": 1.0279184946056258e-05, + "loss": 0.9502, + "step": 10630 + }, + { + "epoch": 0.7304521093857359, + "grad_norm": 1.0625624656677246, + "learning_rate": 1.0274338812792936e-05, + "loss": 0.9331, + "step": 10631 + }, + { + "epoch": 0.7305208190188265, + "grad_norm": 1.044410228729248, + "learning_rate": 1.0269493526687915e-05, + "loss": 0.9221, + "step": 10632 + }, + { + "epoch": 0.730589528651917, + "grad_norm": 1.1492104530334473, + "learning_rate": 1.0264649088019951e-05, + "loss": 0.9682, + "step": 10633 + }, + { + "epoch": 0.7306582382850075, + "grad_norm": 1.0613466501235962, + "learning_rate": 1.0259805497067738e-05, + "loss": 0.9266, + "step": 10634 + }, + { + "epoch": 0.7307269479180981, + "grad_norm": 1.0255385637283325, + "learning_rate": 1.0254962754109928e-05, + "loss": 0.919, + "step": 10635 + }, + { + "epoch": 0.7307956575511887, + "grad_norm": 1.1093690395355225, + "learning_rate": 1.0250120859425127e-05, + "loss": 0.9067, + "step": 10636 + }, + { + "epoch": 0.7308643671842793, + "grad_norm": 1.0587921142578125, + "learning_rate": 1.0245279813291875e-05, + "loss": 0.9239, + "step": 10637 + }, + { + "epoch": 0.7309330768173697, + "grad_norm": 1.0204333066940308, + "learning_rate": 1.0240439615988676e-05, + "loss": 0.893, + "step": 10638 + }, + { + "epoch": 0.7310017864504603, + "grad_norm": 1.1053403615951538, + "learning_rate": 1.0235600267793988e-05, + "loss": 0.9272, + "step": 10639 + }, + { + "epoch": 0.7310704960835509, + "grad_norm": 1.096094012260437, + "learning_rate": 1.023076176898621e-05, + "loss": 0.9497, + "step": 10640 + }, + { + "epoch": 0.7311392057166415, + "grad_norm": 1.0840705633163452, + "learning_rate": 1.0225924119843716e-05, + "loss": 0.9172, + "step": 10641 + }, + { + "epoch": 0.7312079153497321, + "grad_norm": 1.0765053033828735, + "learning_rate": 1.022108732064479e-05, + "loss": 0.9498, + "step": 10642 + }, + { + "epoch": 0.7312766249828226, + "grad_norm": 1.0854297876358032, + "learning_rate": 1.0216251371667695e-05, + "loss": 0.9309, + "step": 10643 + }, + { + "epoch": 0.7313453346159131, + "grad_norm": 1.0418504476547241, + "learning_rate": 1.0211416273190649e-05, + "loss": 0.9196, + "step": 10644 + }, + { + "epoch": 0.7314140442490037, + "grad_norm": 0.9856492280960083, + "learning_rate": 1.0206582025491806e-05, + "loss": 0.924, + "step": 10645 + }, + { + "epoch": 0.7314827538820943, + "grad_norm": 1.0407143831253052, + "learning_rate": 1.0201748628849287e-05, + "loss": 0.9545, + "step": 10646 + }, + { + "epoch": 0.7315514635151849, + "grad_norm": 1.048038363456726, + "learning_rate": 1.019691608354114e-05, + "loss": 0.9382, + "step": 10647 + }, + { + "epoch": 0.7316201731482754, + "grad_norm": 1.0971375703811646, + "learning_rate": 1.0192084389845382e-05, + "loss": 0.919, + "step": 10648 + }, + { + "epoch": 0.7316888827813659, + "grad_norm": 0.9833446145057678, + "learning_rate": 1.0187253548039983e-05, + "loss": 0.8715, + "step": 10649 + }, + { + "epoch": 0.7317575924144565, + "grad_norm": 1.1155288219451904, + "learning_rate": 1.0182423558402863e-05, + "loss": 0.968, + "step": 10650 + }, + { + "epoch": 0.7318263020475471, + "grad_norm": 1.0669167041778564, + "learning_rate": 1.0177594421211869e-05, + "loss": 0.9302, + "step": 10651 + }, + { + "epoch": 0.7318950116806376, + "grad_norm": 1.0692188739776611, + "learning_rate": 1.0172766136744844e-05, + "loss": 0.9401, + "step": 10652 + }, + { + "epoch": 0.7319637213137282, + "grad_norm": 1.1596182584762573, + "learning_rate": 1.016793870527953e-05, + "loss": 0.8941, + "step": 10653 + }, + { + "epoch": 0.7320324309468187, + "grad_norm": 1.0990722179412842, + "learning_rate": 1.016311212709366e-05, + "loss": 0.9304, + "step": 10654 + }, + { + "epoch": 0.7321011405799093, + "grad_norm": 1.1055644750595093, + "learning_rate": 1.0158286402464904e-05, + "loss": 0.9501, + "step": 10655 + }, + { + "epoch": 0.7321698502129999, + "grad_norm": 1.1190119981765747, + "learning_rate": 1.0153461531670889e-05, + "loss": 0.975, + "step": 10656 + }, + { + "epoch": 0.7322385598460904, + "grad_norm": 1.010693073272705, + "learning_rate": 1.014863751498917e-05, + "loss": 0.9147, + "step": 10657 + }, + { + "epoch": 0.732307269479181, + "grad_norm": 1.0387660264968872, + "learning_rate": 1.0143814352697283e-05, + "loss": 0.9497, + "step": 10658 + }, + { + "epoch": 0.7323759791122716, + "grad_norm": 1.0335235595703125, + "learning_rate": 1.0138992045072695e-05, + "loss": 0.9319, + "step": 10659 + }, + { + "epoch": 0.7324446887453621, + "grad_norm": 1.0142531394958496, + "learning_rate": 1.0134170592392836e-05, + "loss": 0.9202, + "step": 10660 + }, + { + "epoch": 0.7325133983784526, + "grad_norm": 1.0646055936813354, + "learning_rate": 1.0129349994935083e-05, + "loss": 0.9034, + "step": 10661 + }, + { + "epoch": 0.7325821080115432, + "grad_norm": 1.0282220840454102, + "learning_rate": 1.0124530252976752e-05, + "loss": 0.9164, + "step": 10662 + }, + { + "epoch": 0.7326508176446338, + "grad_norm": 1.1439166069030762, + "learning_rate": 1.0119711366795127e-05, + "loss": 0.9507, + "step": 10663 + }, + { + "epoch": 0.7327195272777244, + "grad_norm": 1.016404390335083, + "learning_rate": 1.011489333666743e-05, + "loss": 0.9366, + "step": 10664 + }, + { + "epoch": 0.7327882369108148, + "grad_norm": 1.054412841796875, + "learning_rate": 1.0110076162870847e-05, + "loss": 0.9258, + "step": 10665 + }, + { + "epoch": 0.7328569465439054, + "grad_norm": 1.081539273262024, + "learning_rate": 1.010525984568251e-05, + "loss": 0.9223, + "step": 10666 + }, + { + "epoch": 0.732925656176996, + "grad_norm": 1.0577874183654785, + "learning_rate": 1.0100444385379484e-05, + "loss": 0.9073, + "step": 10667 + }, + { + "epoch": 0.7329943658100866, + "grad_norm": 1.0479739904403687, + "learning_rate": 1.0095629782238808e-05, + "loss": 0.905, + "step": 10668 + }, + { + "epoch": 0.7330630754431772, + "grad_norm": 1.0569604635238647, + "learning_rate": 1.0090816036537463e-05, + "loss": 0.9635, + "step": 10669 + }, + { + "epoch": 0.7331317850762676, + "grad_norm": 1.0918333530426025, + "learning_rate": 1.008600314855238e-05, + "loss": 0.9359, + "step": 10670 + }, + { + "epoch": 0.7332004947093582, + "grad_norm": 1.056321620941162, + "learning_rate": 1.0081191118560453e-05, + "loss": 0.9446, + "step": 10671 + }, + { + "epoch": 0.7332692043424488, + "grad_norm": 1.0892523527145386, + "learning_rate": 1.0076379946838494e-05, + "loss": 0.9277, + "step": 10672 + }, + { + "epoch": 0.7333379139755394, + "grad_norm": 1.0023776292800903, + "learning_rate": 1.0071569633663297e-05, + "loss": 0.9083, + "step": 10673 + }, + { + "epoch": 0.73340662360863, + "grad_norm": 1.1367217302322388, + "learning_rate": 1.0066760179311607e-05, + "loss": 0.9267, + "step": 10674 + }, + { + "epoch": 0.7334753332417205, + "grad_norm": 1.0673900842666626, + "learning_rate": 1.0061951584060089e-05, + "loss": 0.9244, + "step": 10675 + }, + { + "epoch": 0.733544042874811, + "grad_norm": 1.0823545455932617, + "learning_rate": 1.0057143848185388e-05, + "loss": 0.9396, + "step": 10676 + }, + { + "epoch": 0.7336127525079016, + "grad_norm": 1.0860174894332886, + "learning_rate": 1.0052336971964099e-05, + "loss": 0.9251, + "step": 10677 + }, + { + "epoch": 0.7336814621409922, + "grad_norm": 1.05278480052948, + "learning_rate": 1.0047530955672741e-05, + "loss": 0.9034, + "step": 10678 + }, + { + "epoch": 0.7337501717740827, + "grad_norm": 1.093625545501709, + "learning_rate": 1.004272579958781e-05, + "loss": 0.926, + "step": 10679 + }, + { + "epoch": 0.7338188814071733, + "grad_norm": 1.023781418800354, + "learning_rate": 1.0037921503985745e-05, + "loss": 0.9388, + "step": 10680 + }, + { + "epoch": 0.7338875910402638, + "grad_norm": 1.1015321016311646, + "learning_rate": 1.0033118069142933e-05, + "loss": 0.937, + "step": 10681 + }, + { + "epoch": 0.7339563006733544, + "grad_norm": 1.0327248573303223, + "learning_rate": 1.0028315495335719e-05, + "loss": 0.9479, + "step": 10682 + }, + { + "epoch": 0.734025010306445, + "grad_norm": 1.2370318174362183, + "learning_rate": 1.002351378284038e-05, + "loss": 1.0036, + "step": 10683 + }, + { + "epoch": 0.7340937199395355, + "grad_norm": 1.0638352632522583, + "learning_rate": 1.0018712931933161e-05, + "loss": 0.9536, + "step": 10684 + }, + { + "epoch": 0.7341624295726261, + "grad_norm": 1.083115816116333, + "learning_rate": 1.001391294289025e-05, + "loss": 0.9632, + "step": 10685 + }, + { + "epoch": 0.7342311392057166, + "grad_norm": 1.1063560247421265, + "learning_rate": 1.0009113815987791e-05, + "loss": 0.9391, + "step": 10686 + }, + { + "epoch": 0.7342998488388072, + "grad_norm": 1.063589334487915, + "learning_rate": 1.0004315551501883e-05, + "loss": 0.9291, + "step": 10687 + }, + { + "epoch": 0.7343685584718977, + "grad_norm": 0.9882360696792603, + "learning_rate": 9.999518149708551e-06, + "loss": 0.8997, + "step": 10688 + }, + { + "epoch": 0.7344372681049883, + "grad_norm": 1.0714462995529175, + "learning_rate": 9.994721610883789e-06, + "loss": 0.9702, + "step": 10689 + }, + { + "epoch": 0.7345059777380789, + "grad_norm": 1.068459153175354, + "learning_rate": 9.989925935303549e-06, + "loss": 0.9482, + "step": 10690 + }, + { + "epoch": 0.7345746873711695, + "grad_norm": 1.0406988859176636, + "learning_rate": 9.985131123243712e-06, + "loss": 0.9247, + "step": 10691 + }, + { + "epoch": 0.73464339700426, + "grad_norm": 1.0861320495605469, + "learning_rate": 9.980337174980139e-06, + "loss": 0.9653, + "step": 10692 + }, + { + "epoch": 0.7347121066373505, + "grad_norm": 1.1062018871307373, + "learning_rate": 9.975544090788603e-06, + "loss": 0.9389, + "step": 10693 + }, + { + "epoch": 0.7347808162704411, + "grad_norm": 1.0809016227722168, + "learning_rate": 9.97075187094485e-06, + "loss": 0.9579, + "step": 10694 + }, + { + "epoch": 0.7348495259035317, + "grad_norm": 1.111506700515747, + "learning_rate": 9.965960515724582e-06, + "loss": 0.9183, + "step": 10695 + }, + { + "epoch": 0.7349182355366223, + "grad_norm": 1.1260634660720825, + "learning_rate": 9.961170025403435e-06, + "loss": 0.9776, + "step": 10696 + }, + { + "epoch": 0.7349869451697127, + "grad_norm": 1.0517634153366089, + "learning_rate": 9.956380400257017e-06, + "loss": 0.9525, + "step": 10697 + }, + { + "epoch": 0.7350556548028033, + "grad_norm": 1.1024914979934692, + "learning_rate": 9.951591640560857e-06, + "loss": 0.9513, + "step": 10698 + }, + { + "epoch": 0.7351243644358939, + "grad_norm": 1.0697249174118042, + "learning_rate": 9.946803746590458e-06, + "loss": 0.9352, + "step": 10699 + }, + { + "epoch": 0.7351930740689845, + "grad_norm": 1.0372809171676636, + "learning_rate": 9.942016718621256e-06, + "loss": 0.9377, + "step": 10700 + }, + { + "epoch": 0.7352617837020751, + "grad_norm": 1.0803731679916382, + "learning_rate": 9.937230556928648e-06, + "loss": 0.9589, + "step": 10701 + }, + { + "epoch": 0.7353304933351655, + "grad_norm": 1.0787280797958374, + "learning_rate": 9.932445261787992e-06, + "loss": 0.9375, + "step": 10702 + }, + { + "epoch": 0.7353992029682561, + "grad_norm": 1.0028847455978394, + "learning_rate": 9.927660833474564e-06, + "loss": 0.9258, + "step": 10703 + }, + { + "epoch": 0.7354679126013467, + "grad_norm": 1.081936240196228, + "learning_rate": 9.922877272263618e-06, + "loss": 0.9936, + "step": 10704 + }, + { + "epoch": 0.7355366222344373, + "grad_norm": 1.1763992309570312, + "learning_rate": 9.918094578430347e-06, + "loss": 0.955, + "step": 10705 + }, + { + "epoch": 0.7356053318675279, + "grad_norm": 1.0379489660263062, + "learning_rate": 9.913312752249903e-06, + "loss": 0.9482, + "step": 10706 + }, + { + "epoch": 0.7356740415006184, + "grad_norm": 1.120192289352417, + "learning_rate": 9.908531793997385e-06, + "loss": 0.9508, + "step": 10707 + }, + { + "epoch": 0.7357427511337089, + "grad_norm": 1.060225248336792, + "learning_rate": 9.903751703947823e-06, + "loss": 0.9586, + "step": 10708 + }, + { + "epoch": 0.7358114607667995, + "grad_norm": 1.1261171102523804, + "learning_rate": 9.898972482376217e-06, + "loss": 0.9617, + "step": 10709 + }, + { + "epoch": 0.7358801703998901, + "grad_norm": 1.0692431926727295, + "learning_rate": 9.89419412955752e-06, + "loss": 0.9179, + "step": 10710 + }, + { + "epoch": 0.7359488800329806, + "grad_norm": 1.0227808952331543, + "learning_rate": 9.889416645766626e-06, + "loss": 0.9342, + "step": 10711 + }, + { + "epoch": 0.7360175896660712, + "grad_norm": 1.0695381164550781, + "learning_rate": 9.884640031278384e-06, + "loss": 0.959, + "step": 10712 + }, + { + "epoch": 0.7360862992991617, + "grad_norm": 1.1033470630645752, + "learning_rate": 9.879864286367577e-06, + "loss": 0.8967, + "step": 10713 + }, + { + "epoch": 0.7361550089322523, + "grad_norm": 1.0028473138809204, + "learning_rate": 9.875089411308958e-06, + "loss": 0.9259, + "step": 10714 + }, + { + "epoch": 0.7362237185653429, + "grad_norm": 1.081248164176941, + "learning_rate": 9.870315406377223e-06, + "loss": 0.9365, + "step": 10715 + }, + { + "epoch": 0.7362924281984334, + "grad_norm": 0.9966160655021667, + "learning_rate": 9.865542271847019e-06, + "loss": 0.9313, + "step": 10716 + }, + { + "epoch": 0.736361137831524, + "grad_norm": 1.070142149925232, + "learning_rate": 9.860770007992947e-06, + "loss": 0.9721, + "step": 10717 + }, + { + "epoch": 0.7364298474646145, + "grad_norm": 0.9463125467300415, + "learning_rate": 9.85599861508954e-06, + "loss": 0.8933, + "step": 10718 + }, + { + "epoch": 0.7364985570977051, + "grad_norm": 1.074183702468872, + "learning_rate": 9.851228093411296e-06, + "loss": 0.9224, + "step": 10719 + }, + { + "epoch": 0.7365672667307956, + "grad_norm": 0.9819031953811646, + "learning_rate": 9.846458443232664e-06, + "loss": 0.9288, + "step": 10720 + }, + { + "epoch": 0.7366359763638862, + "grad_norm": 1.0416293144226074, + "learning_rate": 9.841689664828047e-06, + "loss": 0.9063, + "step": 10721 + }, + { + "epoch": 0.7367046859969768, + "grad_norm": 1.14048171043396, + "learning_rate": 9.836921758471771e-06, + "loss": 1.0207, + "step": 10722 + }, + { + "epoch": 0.7367733956300674, + "grad_norm": 1.1127756834030151, + "learning_rate": 9.832154724438142e-06, + "loss": 0.9136, + "step": 10723 + }, + { + "epoch": 0.7368421052631579, + "grad_norm": 1.0698039531707764, + "learning_rate": 9.827388563001411e-06, + "loss": 0.9558, + "step": 10724 + }, + { + "epoch": 0.7369108148962484, + "grad_norm": 1.0504387617111206, + "learning_rate": 9.822623274435757e-06, + "loss": 0.9756, + "step": 10725 + }, + { + "epoch": 0.736979524529339, + "grad_norm": 1.0867204666137695, + "learning_rate": 9.817858859015333e-06, + "loss": 0.984, + "step": 10726 + }, + { + "epoch": 0.7370482341624296, + "grad_norm": 1.0738413333892822, + "learning_rate": 9.81309531701424e-06, + "loss": 0.928, + "step": 10727 + }, + { + "epoch": 0.7371169437955202, + "grad_norm": 1.106066107749939, + "learning_rate": 9.808332648706503e-06, + "loss": 0.9068, + "step": 10728 + }, + { + "epoch": 0.7371856534286106, + "grad_norm": 1.1095545291900635, + "learning_rate": 9.803570854366127e-06, + "loss": 0.9272, + "step": 10729 + }, + { + "epoch": 0.7372543630617012, + "grad_norm": 1.0785975456237793, + "learning_rate": 9.798809934267052e-06, + "loss": 0.9452, + "step": 10730 + }, + { + "epoch": 0.7373230726947918, + "grad_norm": 1.0155789852142334, + "learning_rate": 9.794049888683177e-06, + "loss": 0.8611, + "step": 10731 + }, + { + "epoch": 0.7373917823278824, + "grad_norm": 1.0375257730484009, + "learning_rate": 9.789290717888337e-06, + "loss": 0.9278, + "step": 10732 + }, + { + "epoch": 0.737460491960973, + "grad_norm": 1.0080921649932861, + "learning_rate": 9.784532422156337e-06, + "loss": 0.9347, + "step": 10733 + }, + { + "epoch": 0.7375292015940634, + "grad_norm": 1.0526609420776367, + "learning_rate": 9.779775001760903e-06, + "loss": 0.9127, + "step": 10734 + }, + { + "epoch": 0.737597911227154, + "grad_norm": 1.0633960962295532, + "learning_rate": 9.77501845697573e-06, + "loss": 0.9451, + "step": 10735 + }, + { + "epoch": 0.7376666208602446, + "grad_norm": 1.068284273147583, + "learning_rate": 9.770262788074461e-06, + "loss": 0.9623, + "step": 10736 + }, + { + "epoch": 0.7377353304933352, + "grad_norm": 1.0503281354904175, + "learning_rate": 9.765507995330689e-06, + "loss": 0.8865, + "step": 10737 + }, + { + "epoch": 0.7378040401264258, + "grad_norm": 1.0768800973892212, + "learning_rate": 9.760754079017964e-06, + "loss": 0.9637, + "step": 10738 + }, + { + "epoch": 0.7378727497595163, + "grad_norm": 1.0395779609680176, + "learning_rate": 9.756001039409753e-06, + "loss": 0.9489, + "step": 10739 + }, + { + "epoch": 0.7379414593926068, + "grad_norm": 1.0439552068710327, + "learning_rate": 9.75124887677951e-06, + "loss": 0.9344, + "step": 10740 + }, + { + "epoch": 0.7380101690256974, + "grad_norm": 1.132172703742981, + "learning_rate": 9.746497591400619e-06, + "loss": 0.9854, + "step": 10741 + }, + { + "epoch": 0.738078878658788, + "grad_norm": 1.0661505460739136, + "learning_rate": 9.741747183546424e-06, + "loss": 0.9715, + "step": 10742 + }, + { + "epoch": 0.7381475882918785, + "grad_norm": 1.0798612833023071, + "learning_rate": 9.736997653490215e-06, + "loss": 0.955, + "step": 10743 + }, + { + "epoch": 0.7382162979249691, + "grad_norm": 1.0590826272964478, + "learning_rate": 9.73224900150522e-06, + "loss": 0.9693, + "step": 10744 + }, + { + "epoch": 0.7382850075580596, + "grad_norm": 1.0748775005340576, + "learning_rate": 9.727501227864627e-06, + "loss": 0.9267, + "step": 10745 + }, + { + "epoch": 0.7383537171911502, + "grad_norm": 1.0298945903778076, + "learning_rate": 9.722754332841588e-06, + "loss": 0.9109, + "step": 10746 + }, + { + "epoch": 0.7384224268242408, + "grad_norm": 1.028701663017273, + "learning_rate": 9.718008316709167e-06, + "loss": 0.9115, + "step": 10747 + }, + { + "epoch": 0.7384911364573313, + "grad_norm": 1.0856415033340454, + "learning_rate": 9.713263179740418e-06, + "loss": 0.9174, + "step": 10748 + }, + { + "epoch": 0.7385598460904219, + "grad_norm": 1.0910308361053467, + "learning_rate": 9.70851892220831e-06, + "loss": 0.9419, + "step": 10749 + }, + { + "epoch": 0.7386285557235124, + "grad_norm": 1.0301094055175781, + "learning_rate": 9.703775544385785e-06, + "loss": 0.9175, + "step": 10750 + }, + { + "epoch": 0.738697265356603, + "grad_norm": 1.0078212022781372, + "learning_rate": 9.699033046545724e-06, + "loss": 0.9016, + "step": 10751 + }, + { + "epoch": 0.7387659749896935, + "grad_norm": 1.087937593460083, + "learning_rate": 9.694291428960964e-06, + "loss": 0.9445, + "step": 10752 + }, + { + "epoch": 0.7388346846227841, + "grad_norm": 1.0946056842803955, + "learning_rate": 9.689550691904292e-06, + "loss": 0.9936, + "step": 10753 + }, + { + "epoch": 0.7389033942558747, + "grad_norm": 1.1220331192016602, + "learning_rate": 9.684810835648428e-06, + "loss": 0.9596, + "step": 10754 + }, + { + "epoch": 0.7389721038889653, + "grad_norm": 1.1228654384613037, + "learning_rate": 9.680071860466056e-06, + "loss": 0.9437, + "step": 10755 + }, + { + "epoch": 0.7390408135220558, + "grad_norm": 1.0667768716812134, + "learning_rate": 9.67533376662981e-06, + "loss": 0.9291, + "step": 10756 + }, + { + "epoch": 0.7391095231551463, + "grad_norm": 1.0867093801498413, + "learning_rate": 9.67059655441227e-06, + "loss": 0.9527, + "step": 10757 + }, + { + "epoch": 0.7391782327882369, + "grad_norm": 1.0326204299926758, + "learning_rate": 9.66586022408597e-06, + "loss": 0.9404, + "step": 10758 + }, + { + "epoch": 0.7392469424213275, + "grad_norm": 1.0752100944519043, + "learning_rate": 9.661124775923371e-06, + "loss": 0.9461, + "step": 10759 + }, + { + "epoch": 0.7393156520544181, + "grad_norm": 1.002463698387146, + "learning_rate": 9.656390210196913e-06, + "loss": 0.9081, + "step": 10760 + }, + { + "epoch": 0.7393843616875085, + "grad_norm": 1.0690124034881592, + "learning_rate": 9.651656527178974e-06, + "loss": 0.9182, + "step": 10761 + }, + { + "epoch": 0.7394530713205991, + "grad_norm": 1.0665291547775269, + "learning_rate": 9.646923727141873e-06, + "loss": 0.9294, + "step": 10762 + }, + { + "epoch": 0.7395217809536897, + "grad_norm": 1.071245789527893, + "learning_rate": 9.642191810357897e-06, + "loss": 0.9182, + "step": 10763 + }, + { + "epoch": 0.7395904905867803, + "grad_norm": 1.042386531829834, + "learning_rate": 9.637460777099255e-06, + "loss": 0.9609, + "step": 10764 + }, + { + "epoch": 0.7396592002198709, + "grad_norm": 1.0334948301315308, + "learning_rate": 9.632730627638126e-06, + "loss": 0.9101, + "step": 10765 + }, + { + "epoch": 0.7397279098529614, + "grad_norm": 1.019099473953247, + "learning_rate": 9.628001362246636e-06, + "loss": 0.9612, + "step": 10766 + }, + { + "epoch": 0.7397966194860519, + "grad_norm": 1.02719247341156, + "learning_rate": 9.623272981196854e-06, + "loss": 0.9358, + "step": 10767 + }, + { + "epoch": 0.7398653291191425, + "grad_norm": 1.027716875076294, + "learning_rate": 9.61854548476081e-06, + "loss": 0.9401, + "step": 10768 + }, + { + "epoch": 0.7399340387522331, + "grad_norm": 1.111466646194458, + "learning_rate": 9.613818873210457e-06, + "loss": 0.914, + "step": 10769 + }, + { + "epoch": 0.7400027483853236, + "grad_norm": 1.0672813653945923, + "learning_rate": 9.60909314681773e-06, + "loss": 0.9346, + "step": 10770 + }, + { + "epoch": 0.7400714580184142, + "grad_norm": 1.0485233068466187, + "learning_rate": 9.604368305854483e-06, + "loss": 0.9499, + "step": 10771 + }, + { + "epoch": 0.7401401676515047, + "grad_norm": 1.0574589967727661, + "learning_rate": 9.59964435059254e-06, + "loss": 0.9423, + "step": 10772 + }, + { + "epoch": 0.7402088772845953, + "grad_norm": 1.0720980167388916, + "learning_rate": 9.594921281303676e-06, + "loss": 0.9639, + "step": 10773 + }, + { + "epoch": 0.7402775869176859, + "grad_norm": 1.1065869331359863, + "learning_rate": 9.590199098259588e-06, + "loss": 0.9395, + "step": 10774 + }, + { + "epoch": 0.7403462965507764, + "grad_norm": 1.0746580362319946, + "learning_rate": 9.585477801731952e-06, + "loss": 0.9383, + "step": 10775 + }, + { + "epoch": 0.740415006183867, + "grad_norm": 1.0708314180374146, + "learning_rate": 9.58075739199238e-06, + "loss": 0.9132, + "step": 10776 + }, + { + "epoch": 0.7404837158169575, + "grad_norm": 1.0326954126358032, + "learning_rate": 9.576037869312428e-06, + "loss": 0.9463, + "step": 10777 + }, + { + "epoch": 0.7405524254500481, + "grad_norm": 1.017802119255066, + "learning_rate": 9.571319233963627e-06, + "loss": 0.9357, + "step": 10778 + }, + { + "epoch": 0.7406211350831386, + "grad_norm": 1.0174678564071655, + "learning_rate": 9.566601486217414e-06, + "loss": 0.9086, + "step": 10779 + }, + { + "epoch": 0.7406898447162292, + "grad_norm": 1.1010048389434814, + "learning_rate": 9.561884626345205e-06, + "loss": 0.9285, + "step": 10780 + }, + { + "epoch": 0.7407585543493198, + "grad_norm": 1.0916260480880737, + "learning_rate": 9.557168654618363e-06, + "loss": 0.956, + "step": 10781 + }, + { + "epoch": 0.7408272639824104, + "grad_norm": 1.0207035541534424, + "learning_rate": 9.55245357130819e-06, + "loss": 0.9116, + "step": 10782 + }, + { + "epoch": 0.7408959736155009, + "grad_norm": 1.0911533832550049, + "learning_rate": 9.547739376685952e-06, + "loss": 0.9233, + "step": 10783 + }, + { + "epoch": 0.7409646832485914, + "grad_norm": 1.040333867073059, + "learning_rate": 9.54302607102284e-06, + "loss": 0.8906, + "step": 10784 + }, + { + "epoch": 0.741033392881682, + "grad_norm": 1.0138956308364868, + "learning_rate": 9.538313654590011e-06, + "loss": 0.8987, + "step": 10785 + }, + { + "epoch": 0.7411021025147726, + "grad_norm": 1.106356143951416, + "learning_rate": 9.533602127658572e-06, + "loss": 0.9647, + "step": 10786 + }, + { + "epoch": 0.7411708121478632, + "grad_norm": 1.0612725019454956, + "learning_rate": 9.528891490499573e-06, + "loss": 0.9343, + "step": 10787 + }, + { + "epoch": 0.7412395217809536, + "grad_norm": 1.0206851959228516, + "learning_rate": 9.524181743384011e-06, + "loss": 0.9537, + "step": 10788 + }, + { + "epoch": 0.7413082314140442, + "grad_norm": 1.0545891523361206, + "learning_rate": 9.519472886582848e-06, + "loss": 0.9418, + "step": 10789 + }, + { + "epoch": 0.7413769410471348, + "grad_norm": 1.1057813167572021, + "learning_rate": 9.514764920366958e-06, + "loss": 0.9388, + "step": 10790 + }, + { + "epoch": 0.7414456506802254, + "grad_norm": 1.0961343050003052, + "learning_rate": 9.510057845007203e-06, + "loss": 0.9095, + "step": 10791 + }, + { + "epoch": 0.741514360313316, + "grad_norm": 1.0706862211227417, + "learning_rate": 9.505351660774386e-06, + "loss": 0.9406, + "step": 10792 + }, + { + "epoch": 0.7415830699464064, + "grad_norm": 1.1213654279708862, + "learning_rate": 9.500646367939228e-06, + "loss": 0.9249, + "step": 10793 + }, + { + "epoch": 0.741651779579497, + "grad_norm": 1.083260178565979, + "learning_rate": 9.495941966772437e-06, + "loss": 0.9538, + "step": 10794 + }, + { + "epoch": 0.7417204892125876, + "grad_norm": 1.0904791355133057, + "learning_rate": 9.49123845754466e-06, + "loss": 0.9341, + "step": 10795 + }, + { + "epoch": 0.7417891988456782, + "grad_norm": 1.0050517320632935, + "learning_rate": 9.486535840526468e-06, + "loss": 0.9468, + "step": 10796 + }, + { + "epoch": 0.7418579084787688, + "grad_norm": 1.0464086532592773, + "learning_rate": 9.481834115988412e-06, + "loss": 0.943, + "step": 10797 + }, + { + "epoch": 0.7419266181118593, + "grad_norm": 1.1157701015472412, + "learning_rate": 9.477133284200973e-06, + "loss": 0.9346, + "step": 10798 + }, + { + "epoch": 0.7419953277449498, + "grad_norm": 1.0458106994628906, + "learning_rate": 9.472433345434603e-06, + "loss": 0.9245, + "step": 10799 + }, + { + "epoch": 0.7420640373780404, + "grad_norm": 1.0757474899291992, + "learning_rate": 9.467734299959666e-06, + "loss": 1.0025, + "step": 10800 + }, + { + "epoch": 0.742132747011131, + "grad_norm": 0.9822719097137451, + "learning_rate": 9.463036148046506e-06, + "loss": 0.8944, + "step": 10801 + }, + { + "epoch": 0.7422014566442215, + "grad_norm": 1.0020331144332886, + "learning_rate": 9.458338889965401e-06, + "loss": 0.9246, + "step": 10802 + }, + { + "epoch": 0.7422701662773121, + "grad_norm": 1.0701545476913452, + "learning_rate": 9.453642525986583e-06, + "loss": 0.9321, + "step": 10803 + }, + { + "epoch": 0.7423388759104026, + "grad_norm": 1.0485159158706665, + "learning_rate": 9.448947056380241e-06, + "loss": 0.9589, + "step": 10804 + }, + { + "epoch": 0.7424075855434932, + "grad_norm": 1.1139644384384155, + "learning_rate": 9.444252481416486e-06, + "loss": 0.9513, + "step": 10805 + }, + { + "epoch": 0.7424762951765838, + "grad_norm": 1.0603004693984985, + "learning_rate": 9.439558801365397e-06, + "loss": 0.9309, + "step": 10806 + }, + { + "epoch": 0.7425450048096743, + "grad_norm": 1.060904860496521, + "learning_rate": 9.434866016497004e-06, + "loss": 0.9398, + "step": 10807 + }, + { + "epoch": 0.7426137144427649, + "grad_norm": 1.106801152229309, + "learning_rate": 9.430174127081279e-06, + "loss": 0.9329, + "step": 10808 + }, + { + "epoch": 0.7426824240758554, + "grad_norm": 1.0248252153396606, + "learning_rate": 9.425483133388151e-06, + "loss": 0.9228, + "step": 10809 + }, + { + "epoch": 0.742751133708946, + "grad_norm": 1.0715668201446533, + "learning_rate": 9.420793035687476e-06, + "loss": 0.9569, + "step": 10810 + }, + { + "epoch": 0.7428198433420365, + "grad_norm": 1.0593082904815674, + "learning_rate": 9.416103834249073e-06, + "loss": 0.9598, + "step": 10811 + }, + { + "epoch": 0.7428885529751271, + "grad_norm": 1.0291932821273804, + "learning_rate": 9.41141552934272e-06, + "loss": 0.9031, + "step": 10812 + }, + { + "epoch": 0.7429572626082177, + "grad_norm": 1.0900002717971802, + "learning_rate": 9.406728121238124e-06, + "loss": 0.9362, + "step": 10813 + }, + { + "epoch": 0.7430259722413083, + "grad_norm": 1.1592369079589844, + "learning_rate": 9.402041610204962e-06, + "loss": 0.9458, + "step": 10814 + }, + { + "epoch": 0.7430946818743988, + "grad_norm": 1.0809086561203003, + "learning_rate": 9.397355996512825e-06, + "loss": 0.9328, + "step": 10815 + }, + { + "epoch": 0.7431633915074893, + "grad_norm": 1.0654081106185913, + "learning_rate": 9.392671280431284e-06, + "loss": 0.9561, + "step": 10816 + }, + { + "epoch": 0.7432321011405799, + "grad_norm": 1.0303250551223755, + "learning_rate": 9.387987462229859e-06, + "loss": 0.8993, + "step": 10817 + }, + { + "epoch": 0.7433008107736705, + "grad_norm": 0.9961594343185425, + "learning_rate": 9.383304542177985e-06, + "loss": 0.9017, + "step": 10818 + }, + { + "epoch": 0.7433695204067611, + "grad_norm": 1.012691855430603, + "learning_rate": 9.378622520545086e-06, + "loss": 0.9479, + "step": 10819 + }, + { + "epoch": 0.7434382300398515, + "grad_norm": 1.053274393081665, + "learning_rate": 9.373941397600502e-06, + "loss": 0.925, + "step": 10820 + }, + { + "epoch": 0.7435069396729421, + "grad_norm": 1.0288370847702026, + "learning_rate": 9.369261173613541e-06, + "loss": 0.9147, + "step": 10821 + }, + { + "epoch": 0.7435756493060327, + "grad_norm": 1.0626873970031738, + "learning_rate": 9.364581848853456e-06, + "loss": 0.9459, + "step": 10822 + }, + { + "epoch": 0.7436443589391233, + "grad_norm": 0.9996181726455688, + "learning_rate": 9.359903423589442e-06, + "loss": 0.9303, + "step": 10823 + }, + { + "epoch": 0.7437130685722139, + "grad_norm": 1.0008670091629028, + "learning_rate": 9.355225898090658e-06, + "loss": 0.9085, + "step": 10824 + }, + { + "epoch": 0.7437817782053043, + "grad_norm": 1.1610400676727295, + "learning_rate": 9.350549272626177e-06, + "loss": 0.9803, + "step": 10825 + }, + { + "epoch": 0.7438504878383949, + "grad_norm": 1.0294535160064697, + "learning_rate": 9.345873547465059e-06, + "loss": 0.9692, + "step": 10826 + }, + { + "epoch": 0.7439191974714855, + "grad_norm": 1.080539584159851, + "learning_rate": 9.341198722876288e-06, + "loss": 0.9749, + "step": 10827 + }, + { + "epoch": 0.7439879071045761, + "grad_norm": 1.094181776046753, + "learning_rate": 9.336524799128808e-06, + "loss": 0.9485, + "step": 10828 + }, + { + "epoch": 0.7440566167376667, + "grad_norm": 1.1231629848480225, + "learning_rate": 9.331851776491512e-06, + "loss": 0.9638, + "step": 10829 + }, + { + "epoch": 0.7441253263707572, + "grad_norm": 1.0672005414962769, + "learning_rate": 9.327179655233223e-06, + "loss": 0.9346, + "step": 10830 + }, + { + "epoch": 0.7441940360038477, + "grad_norm": 1.058367371559143, + "learning_rate": 9.322508435622732e-06, + "loss": 0.937, + "step": 10831 + }, + { + "epoch": 0.7442627456369383, + "grad_norm": 1.0471217632293701, + "learning_rate": 9.317838117928773e-06, + "loss": 0.9548, + "step": 10832 + }, + { + "epoch": 0.7443314552700289, + "grad_norm": 1.0455840826034546, + "learning_rate": 9.313168702420022e-06, + "loss": 0.912, + "step": 10833 + }, + { + "epoch": 0.7444001649031194, + "grad_norm": 1.1267049312591553, + "learning_rate": 9.30850018936512e-06, + "loss": 0.9588, + "step": 10834 + }, + { + "epoch": 0.74446887453621, + "grad_norm": 1.0533207654953003, + "learning_rate": 9.303832579032629e-06, + "loss": 0.927, + "step": 10835 + }, + { + "epoch": 0.7445375841693005, + "grad_norm": 1.0039846897125244, + "learning_rate": 9.29916587169108e-06, + "loss": 0.9264, + "step": 10836 + }, + { + "epoch": 0.7446062938023911, + "grad_norm": 1.0375231504440308, + "learning_rate": 9.29450006760894e-06, + "loss": 0.9043, + "step": 10837 + }, + { + "epoch": 0.7446750034354817, + "grad_norm": 1.0902869701385498, + "learning_rate": 9.28983516705464e-06, + "loss": 0.9552, + "step": 10838 + }, + { + "epoch": 0.7447437130685722, + "grad_norm": 1.0729756355285645, + "learning_rate": 9.285171170296548e-06, + "loss": 0.9659, + "step": 10839 + }, + { + "epoch": 0.7448124227016628, + "grad_norm": 1.0418963432312012, + "learning_rate": 9.28050807760297e-06, + "loss": 0.9577, + "step": 10840 + }, + { + "epoch": 0.7448811323347533, + "grad_norm": 1.0575143098831177, + "learning_rate": 9.275845889242176e-06, + "loss": 0.9174, + "step": 10841 + }, + { + "epoch": 0.7449498419678439, + "grad_norm": 1.0280042886734009, + "learning_rate": 9.271184605482391e-06, + "loss": 0.9293, + "step": 10842 + }, + { + "epoch": 0.7450185516009344, + "grad_norm": 1.0627816915512085, + "learning_rate": 9.266524226591758e-06, + "loss": 0.9342, + "step": 10843 + }, + { + "epoch": 0.745087261234025, + "grad_norm": 1.084511160850525, + "learning_rate": 9.26186475283839e-06, + "loss": 0.9493, + "step": 10844 + }, + { + "epoch": 0.7451559708671156, + "grad_norm": 1.0325291156768799, + "learning_rate": 9.257206184490357e-06, + "loss": 0.9452, + "step": 10845 + }, + { + "epoch": 0.7452246805002062, + "grad_norm": 1.0610055923461914, + "learning_rate": 9.252548521815643e-06, + "loss": 0.9245, + "step": 10846 + }, + { + "epoch": 0.7452933901332967, + "grad_norm": 1.0613821744918823, + "learning_rate": 9.247891765082215e-06, + "loss": 0.9585, + "step": 10847 + }, + { + "epoch": 0.7453620997663872, + "grad_norm": 1.031156063079834, + "learning_rate": 9.243235914557966e-06, + "loss": 0.932, + "step": 10848 + }, + { + "epoch": 0.7454308093994778, + "grad_norm": 1.0650570392608643, + "learning_rate": 9.238580970510746e-06, + "loss": 0.927, + "step": 10849 + }, + { + "epoch": 0.7454995190325684, + "grad_norm": 1.1101408004760742, + "learning_rate": 9.233926933208364e-06, + "loss": 0.945, + "step": 10850 + }, + { + "epoch": 0.745568228665659, + "grad_norm": 1.079102873802185, + "learning_rate": 9.229273802918542e-06, + "loss": 0.9326, + "step": 10851 + }, + { + "epoch": 0.7456369382987494, + "grad_norm": 1.0674251317977905, + "learning_rate": 9.224621579908982e-06, + "loss": 0.9438, + "step": 10852 + }, + { + "epoch": 0.74570564793184, + "grad_norm": 1.037210464477539, + "learning_rate": 9.219970264447324e-06, + "loss": 0.9433, + "step": 10853 + }, + { + "epoch": 0.7457743575649306, + "grad_norm": 1.0336652994155884, + "learning_rate": 9.215319856801158e-06, + "loss": 0.9376, + "step": 10854 + }, + { + "epoch": 0.7458430671980212, + "grad_norm": 1.0572863817214966, + "learning_rate": 9.210670357238022e-06, + "loss": 0.9369, + "step": 10855 + }, + { + "epoch": 0.7459117768311118, + "grad_norm": 1.0114870071411133, + "learning_rate": 9.206021766025385e-06, + "loss": 0.9066, + "step": 10856 + }, + { + "epoch": 0.7459804864642022, + "grad_norm": 1.0997319221496582, + "learning_rate": 9.201374083430686e-06, + "loss": 0.967, + "step": 10857 + }, + { + "epoch": 0.7460491960972928, + "grad_norm": 0.9741802215576172, + "learning_rate": 9.196727309721304e-06, + "loss": 0.9393, + "step": 10858 + }, + { + "epoch": 0.7461179057303834, + "grad_norm": 1.0421040058135986, + "learning_rate": 9.19208144516456e-06, + "loss": 0.9283, + "step": 10859 + }, + { + "epoch": 0.746186615363474, + "grad_norm": 1.1899583339691162, + "learning_rate": 9.187436490027745e-06, + "loss": 1.0107, + "step": 10860 + }, + { + "epoch": 0.7462553249965646, + "grad_norm": 1.0948444604873657, + "learning_rate": 9.182792444578056e-06, + "loss": 0.9485, + "step": 10861 + }, + { + "epoch": 0.7463240346296551, + "grad_norm": 1.0640710592269897, + "learning_rate": 9.178149309082673e-06, + "loss": 0.9525, + "step": 10862 + }, + { + "epoch": 0.7463927442627456, + "grad_norm": 1.0176061391830444, + "learning_rate": 9.173507083808716e-06, + "loss": 0.924, + "step": 10863 + }, + { + "epoch": 0.7464614538958362, + "grad_norm": 1.0429296493530273, + "learning_rate": 9.168865769023255e-06, + "loss": 0.9003, + "step": 10864 + }, + { + "epoch": 0.7465301635289268, + "grad_norm": 1.0299333333969116, + "learning_rate": 9.164225364993281e-06, + "loss": 0.9539, + "step": 10865 + }, + { + "epoch": 0.7465988731620173, + "grad_norm": 0.9703050851821899, + "learning_rate": 9.159585871985779e-06, + "loss": 0.8926, + "step": 10866 + }, + { + "epoch": 0.7466675827951079, + "grad_norm": 1.0465753078460693, + "learning_rate": 9.154947290267632e-06, + "loss": 0.9285, + "step": 10867 + }, + { + "epoch": 0.7467362924281984, + "grad_norm": 1.048132061958313, + "learning_rate": 9.150309620105706e-06, + "loss": 0.9493, + "step": 10868 + }, + { + "epoch": 0.746805002061289, + "grad_norm": 1.056383728981018, + "learning_rate": 9.145672861766807e-06, + "loss": 0.9044, + "step": 10869 + }, + { + "epoch": 0.7468737116943795, + "grad_norm": 1.062782645225525, + "learning_rate": 9.141037015517687e-06, + "loss": 0.9214, + "step": 10870 + }, + { + "epoch": 0.7469424213274701, + "grad_norm": 1.1148982048034668, + "learning_rate": 9.136402081625028e-06, + "loss": 0.9257, + "step": 10871 + }, + { + "epoch": 0.7470111309605607, + "grad_norm": 1.0848408937454224, + "learning_rate": 9.131768060355489e-06, + "loss": 0.9178, + "step": 10872 + }, + { + "epoch": 0.7470798405936512, + "grad_norm": 1.0501412153244019, + "learning_rate": 9.127134951975655e-06, + "loss": 0.9341, + "step": 10873 + }, + { + "epoch": 0.7471485502267418, + "grad_norm": 1.078199028968811, + "learning_rate": 9.12250275675207e-06, + "loss": 0.9557, + "step": 10874 + }, + { + "epoch": 0.7472172598598323, + "grad_norm": 1.0570032596588135, + "learning_rate": 9.117871474951228e-06, + "loss": 0.9652, + "step": 10875 + }, + { + "epoch": 0.7472859694929229, + "grad_norm": 1.1004071235656738, + "learning_rate": 9.113241106839549e-06, + "loss": 0.9302, + "step": 10876 + }, + { + "epoch": 0.7473546791260135, + "grad_norm": 1.0613913536071777, + "learning_rate": 9.10861165268342e-06, + "loss": 0.9483, + "step": 10877 + }, + { + "epoch": 0.7474233887591041, + "grad_norm": 1.095384955406189, + "learning_rate": 9.103983112749175e-06, + "loss": 0.9166, + "step": 10878 + }, + { + "epoch": 0.7474920983921945, + "grad_norm": 1.0516737699508667, + "learning_rate": 9.099355487303088e-06, + "loss": 0.9651, + "step": 10879 + }, + { + "epoch": 0.7475608080252851, + "grad_norm": 1.0469037294387817, + "learning_rate": 9.094728776611395e-06, + "loss": 0.9379, + "step": 10880 + }, + { + "epoch": 0.7476295176583757, + "grad_norm": 1.0201318264007568, + "learning_rate": 9.090102980940249e-06, + "loss": 0.9464, + "step": 10881 + }, + { + "epoch": 0.7476982272914663, + "grad_norm": 1.031165599822998, + "learning_rate": 9.085478100555775e-06, + "loss": 0.9193, + "step": 10882 + }, + { + "epoch": 0.7477669369245569, + "grad_norm": 1.0522700548171997, + "learning_rate": 9.080854135724045e-06, + "loss": 0.9245, + "step": 10883 + }, + { + "epoch": 0.7478356465576473, + "grad_norm": 1.0572763681411743, + "learning_rate": 9.076231086711067e-06, + "loss": 0.9671, + "step": 10884 + }, + { + "epoch": 0.7479043561907379, + "grad_norm": 1.0733600854873657, + "learning_rate": 9.071608953782817e-06, + "loss": 0.9525, + "step": 10885 + }, + { + "epoch": 0.7479730658238285, + "grad_norm": 1.0045781135559082, + "learning_rate": 9.066987737205182e-06, + "loss": 0.9049, + "step": 10886 + }, + { + "epoch": 0.7480417754569191, + "grad_norm": 1.076159119606018, + "learning_rate": 9.06236743724403e-06, + "loss": 0.9574, + "step": 10887 + }, + { + "epoch": 0.7481104850900097, + "grad_norm": 1.0771615505218506, + "learning_rate": 9.057748054165166e-06, + "loss": 0.9564, + "step": 10888 + }, + { + "epoch": 0.7481791947231001, + "grad_norm": 1.0921021699905396, + "learning_rate": 9.05312958823433e-06, + "loss": 0.9615, + "step": 10889 + }, + { + "epoch": 0.7482479043561907, + "grad_norm": 1.0551886558532715, + "learning_rate": 9.048512039717225e-06, + "loss": 0.933, + "step": 10890 + }, + { + "epoch": 0.7483166139892813, + "grad_norm": 1.0861611366271973, + "learning_rate": 9.043895408879505e-06, + "loss": 0.9261, + "step": 10891 + }, + { + "epoch": 0.7483853236223719, + "grad_norm": 1.0642049312591553, + "learning_rate": 9.039279695986743e-06, + "loss": 0.9414, + "step": 10892 + }, + { + "epoch": 0.7484540332554624, + "grad_norm": 0.9858757853507996, + "learning_rate": 9.034664901304491e-06, + "loss": 0.9244, + "step": 10893 + }, + { + "epoch": 0.748522742888553, + "grad_norm": 1.1048741340637207, + "learning_rate": 9.030051025098233e-06, + "loss": 0.9365, + "step": 10894 + }, + { + "epoch": 0.7485914525216435, + "grad_norm": 1.0088608264923096, + "learning_rate": 9.0254380676334e-06, + "loss": 0.9237, + "step": 10895 + }, + { + "epoch": 0.7486601621547341, + "grad_norm": 1.0899052619934082, + "learning_rate": 9.020826029175384e-06, + "loss": 0.9522, + "step": 10896 + }, + { + "epoch": 0.7487288717878247, + "grad_norm": 1.0305616855621338, + "learning_rate": 9.016214909989493e-06, + "loss": 0.9067, + "step": 10897 + }, + { + "epoch": 0.7487975814209152, + "grad_norm": 1.0614858865737915, + "learning_rate": 9.011604710341014e-06, + "loss": 0.962, + "step": 10898 + }, + { + "epoch": 0.7488662910540058, + "grad_norm": 1.0280407667160034, + "learning_rate": 9.006995430495168e-06, + "loss": 0.9299, + "step": 10899 + }, + { + "epoch": 0.7489350006870963, + "grad_norm": 1.039110541343689, + "learning_rate": 9.002387070717122e-06, + "loss": 0.9365, + "step": 10900 + }, + { + "epoch": 0.7490037103201869, + "grad_norm": 1.0113791227340698, + "learning_rate": 8.997779631272002e-06, + "loss": 0.9173, + "step": 10901 + }, + { + "epoch": 0.7490724199532774, + "grad_norm": 1.0729562044143677, + "learning_rate": 8.993173112424857e-06, + "loss": 0.9854, + "step": 10902 + }, + { + "epoch": 0.749141129586368, + "grad_norm": 1.1184097528457642, + "learning_rate": 8.988567514440702e-06, + "loss": 0.9114, + "step": 10903 + }, + { + "epoch": 0.7492098392194586, + "grad_norm": 1.084522008895874, + "learning_rate": 8.983962837584491e-06, + "loss": 0.9595, + "step": 10904 + }, + { + "epoch": 0.7492785488525491, + "grad_norm": 1.0928810834884644, + "learning_rate": 8.979359082121135e-06, + "loss": 0.8846, + "step": 10905 + }, + { + "epoch": 0.7493472584856397, + "grad_norm": 1.0339455604553223, + "learning_rate": 8.974756248315493e-06, + "loss": 0.947, + "step": 10906 + }, + { + "epoch": 0.7494159681187302, + "grad_norm": 1.058512568473816, + "learning_rate": 8.970154336432344e-06, + "loss": 0.9262, + "step": 10907 + }, + { + "epoch": 0.7494846777518208, + "grad_norm": 1.0491212606430054, + "learning_rate": 8.965553346736441e-06, + "loss": 0.9617, + "step": 10908 + }, + { + "epoch": 0.7495533873849114, + "grad_norm": 1.0232044458389282, + "learning_rate": 8.96095327949248e-06, + "loss": 0.94, + "step": 10909 + }, + { + "epoch": 0.749622097018002, + "grad_norm": 1.064279317855835, + "learning_rate": 8.956354134965104e-06, + "loss": 0.9386, + "step": 10910 + }, + { + "epoch": 0.7496908066510924, + "grad_norm": 1.077505111694336, + "learning_rate": 8.95175591341888e-06, + "loss": 0.94, + "step": 10911 + }, + { + "epoch": 0.749759516284183, + "grad_norm": 1.1122398376464844, + "learning_rate": 8.947158615118357e-06, + "loss": 0.9376, + "step": 10912 + }, + { + "epoch": 0.7498282259172736, + "grad_norm": 1.043391466140747, + "learning_rate": 8.94256224032802e-06, + "loss": 0.9048, + "step": 10913 + }, + { + "epoch": 0.7498969355503642, + "grad_norm": 1.138579249382019, + "learning_rate": 8.937966789312277e-06, + "loss": 0.9668, + "step": 10914 + }, + { + "epoch": 0.7499656451834548, + "grad_norm": 1.06857168674469, + "learning_rate": 8.93337226233551e-06, + "loss": 0.9469, + "step": 10915 + }, + { + "epoch": 0.7500343548165452, + "grad_norm": 1.0961549282073975, + "learning_rate": 8.92877865966205e-06, + "loss": 0.9702, + "step": 10916 + }, + { + "epoch": 0.7501030644496358, + "grad_norm": 1.0747307538986206, + "learning_rate": 8.924185981556146e-06, + "loss": 0.9192, + "step": 10917 + }, + { + "epoch": 0.7501717740827264, + "grad_norm": 0.9753084182739258, + "learning_rate": 8.919594228282019e-06, + "loss": 0.9274, + "step": 10918 + }, + { + "epoch": 0.750240483715817, + "grad_norm": 1.0522807836532593, + "learning_rate": 8.915003400103833e-06, + "loss": 0.9379, + "step": 10919 + }, + { + "epoch": 0.7503091933489076, + "grad_norm": 1.0472303628921509, + "learning_rate": 8.910413497285694e-06, + "loss": 0.907, + "step": 10920 + }, + { + "epoch": 0.750377902981998, + "grad_norm": 1.0883376598358154, + "learning_rate": 8.905824520091663e-06, + "loss": 0.9651, + "step": 10921 + }, + { + "epoch": 0.7504466126150886, + "grad_norm": 1.1036115884780884, + "learning_rate": 8.901236468785728e-06, + "loss": 0.9497, + "step": 10922 + }, + { + "epoch": 0.7505153222481792, + "grad_norm": 1.1053669452667236, + "learning_rate": 8.89664934363184e-06, + "loss": 0.9693, + "step": 10923 + }, + { + "epoch": 0.7505840318812698, + "grad_norm": 1.0823371410369873, + "learning_rate": 8.8920631448939e-06, + "loss": 0.9739, + "step": 10924 + }, + { + "epoch": 0.7506527415143603, + "grad_norm": 1.0491889715194702, + "learning_rate": 8.887477872835746e-06, + "loss": 0.9408, + "step": 10925 + }, + { + "epoch": 0.7507214511474509, + "grad_norm": 1.03838312625885, + "learning_rate": 8.88289352772117e-06, + "loss": 0.9214, + "step": 10926 + }, + { + "epoch": 0.7507901607805414, + "grad_norm": 1.0918807983398438, + "learning_rate": 8.878310109813895e-06, + "loss": 0.9358, + "step": 10927 + }, + { + "epoch": 0.750858870413632, + "grad_norm": 1.0989811420440674, + "learning_rate": 8.873727619377611e-06, + "loss": 0.9524, + "step": 10928 + }, + { + "epoch": 0.7509275800467226, + "grad_norm": 1.0646593570709229, + "learning_rate": 8.869146056675946e-06, + "loss": 0.9206, + "step": 10929 + }, + { + "epoch": 0.7509962896798131, + "grad_norm": 1.0235515832901, + "learning_rate": 8.864565421972467e-06, + "loss": 0.9352, + "step": 10930 + }, + { + "epoch": 0.7510649993129037, + "grad_norm": 1.0996390581130981, + "learning_rate": 8.859985715530713e-06, + "loss": 0.9274, + "step": 10931 + }, + { + "epoch": 0.7511337089459942, + "grad_norm": 1.1007221937179565, + "learning_rate": 8.855406937614134e-06, + "loss": 0.958, + "step": 10932 + }, + { + "epoch": 0.7512024185790848, + "grad_norm": 1.0588819980621338, + "learning_rate": 8.850829088486145e-06, + "loss": 0.9209, + "step": 10933 + }, + { + "epoch": 0.7512711282121753, + "grad_norm": 1.0726850032806396, + "learning_rate": 8.846252168410115e-06, + "loss": 0.9476, + "step": 10934 + }, + { + "epoch": 0.7513398378452659, + "grad_norm": 1.0697299242019653, + "learning_rate": 8.841676177649355e-06, + "loss": 0.9318, + "step": 10935 + }, + { + "epoch": 0.7514085474783565, + "grad_norm": 0.9898485541343689, + "learning_rate": 8.837101116467106e-06, + "loss": 0.905, + "step": 10936 + }, + { + "epoch": 0.751477257111447, + "grad_norm": 1.080683946609497, + "learning_rate": 8.832526985126582e-06, + "loss": 0.9628, + "step": 10937 + }, + { + "epoch": 0.7515459667445376, + "grad_norm": 1.076112151145935, + "learning_rate": 8.827953783890912e-06, + "loss": 0.9111, + "step": 10938 + }, + { + "epoch": 0.7516146763776281, + "grad_norm": 1.0450438261032104, + "learning_rate": 8.823381513023205e-06, + "loss": 0.9538, + "step": 10939 + }, + { + "epoch": 0.7516833860107187, + "grad_norm": 1.1012063026428223, + "learning_rate": 8.818810172786493e-06, + "loss": 0.9255, + "step": 10940 + }, + { + "epoch": 0.7517520956438093, + "grad_norm": 1.1210353374481201, + "learning_rate": 8.814239763443775e-06, + "loss": 0.9094, + "step": 10941 + }, + { + "epoch": 0.7518208052768999, + "grad_norm": 1.1055635213851929, + "learning_rate": 8.809670285257965e-06, + "loss": 1.0035, + "step": 10942 + }, + { + "epoch": 0.7518895149099903, + "grad_norm": 1.0574274063110352, + "learning_rate": 8.805101738491949e-06, + "loss": 0.9331, + "step": 10943 + }, + { + "epoch": 0.7519582245430809, + "grad_norm": 1.0531038045883179, + "learning_rate": 8.80053412340856e-06, + "loss": 0.9469, + "step": 10944 + }, + { + "epoch": 0.7520269341761715, + "grad_norm": 1.094329595565796, + "learning_rate": 8.795967440270563e-06, + "loss": 0.9766, + "step": 10945 + }, + { + "epoch": 0.7520956438092621, + "grad_norm": 1.0915780067443848, + "learning_rate": 8.79140168934069e-06, + "loss": 0.9355, + "step": 10946 + }, + { + "epoch": 0.7521643534423527, + "grad_norm": 1.0988585948944092, + "learning_rate": 8.786836870881584e-06, + "loss": 0.9381, + "step": 10947 + }, + { + "epoch": 0.7522330630754431, + "grad_norm": 1.07710599899292, + "learning_rate": 8.782272985155868e-06, + "loss": 0.8977, + "step": 10948 + }, + { + "epoch": 0.7523017727085337, + "grad_norm": 1.09869384765625, + "learning_rate": 8.777710032426098e-06, + "loss": 0.9433, + "step": 10949 + }, + { + "epoch": 0.7523704823416243, + "grad_norm": 1.0458794832229614, + "learning_rate": 8.77314801295478e-06, + "loss": 0.9576, + "step": 10950 + }, + { + "epoch": 0.7524391919747149, + "grad_norm": 1.078449010848999, + "learning_rate": 8.768586927004363e-06, + "loss": 0.9421, + "step": 10951 + }, + { + "epoch": 0.7525079016078055, + "grad_norm": 1.0410972833633423, + "learning_rate": 8.76402677483725e-06, + "loss": 0.9222, + "step": 10952 + }, + { + "epoch": 0.7525766112408959, + "grad_norm": 1.0941736698150635, + "learning_rate": 8.75946755671577e-06, + "loss": 0.9752, + "step": 10953 + }, + { + "epoch": 0.7526453208739865, + "grad_norm": 1.0276827812194824, + "learning_rate": 8.754909272902218e-06, + "loss": 0.941, + "step": 10954 + }, + { + "epoch": 0.7527140305070771, + "grad_norm": 1.072723627090454, + "learning_rate": 8.750351923658832e-06, + "loss": 0.946, + "step": 10955 + }, + { + "epoch": 0.7527827401401677, + "grad_norm": 1.0179141759872437, + "learning_rate": 8.74579550924779e-06, + "loss": 0.9173, + "step": 10956 + }, + { + "epoch": 0.7528514497732582, + "grad_norm": 1.0212821960449219, + "learning_rate": 8.74124002993123e-06, + "loss": 0.9101, + "step": 10957 + }, + { + "epoch": 0.7529201594063488, + "grad_norm": 1.0771634578704834, + "learning_rate": 8.73668548597121e-06, + "loss": 0.9428, + "step": 10958 + }, + { + "epoch": 0.7529888690394393, + "grad_norm": 1.1253271102905273, + "learning_rate": 8.732131877629765e-06, + "loss": 0.9562, + "step": 10959 + }, + { + "epoch": 0.7530575786725299, + "grad_norm": 1.0725256204605103, + "learning_rate": 8.727579205168847e-06, + "loss": 0.9756, + "step": 10960 + }, + { + "epoch": 0.7531262883056205, + "grad_norm": 1.0345438718795776, + "learning_rate": 8.723027468850375e-06, + "loss": 0.9545, + "step": 10961 + }, + { + "epoch": 0.753194997938711, + "grad_norm": 1.1094098091125488, + "learning_rate": 8.718476668936217e-06, + "loss": 0.8939, + "step": 10962 + }, + { + "epoch": 0.7532637075718016, + "grad_norm": 1.0469424724578857, + "learning_rate": 8.71392680568816e-06, + "loss": 0.9179, + "step": 10963 + }, + { + "epoch": 0.7533324172048921, + "grad_norm": 1.1053873300552368, + "learning_rate": 8.709377879367963e-06, + "loss": 0.9499, + "step": 10964 + }, + { + "epoch": 0.7534011268379827, + "grad_norm": 1.0948655605316162, + "learning_rate": 8.704829890237326e-06, + "loss": 0.9677, + "step": 10965 + }, + { + "epoch": 0.7534698364710732, + "grad_norm": 1.027886152267456, + "learning_rate": 8.700282838557886e-06, + "loss": 0.9125, + "step": 10966 + }, + { + "epoch": 0.7535385461041638, + "grad_norm": 1.0831319093704224, + "learning_rate": 8.695736724591247e-06, + "loss": 0.9005, + "step": 10967 + }, + { + "epoch": 0.7536072557372544, + "grad_norm": 1.0981451272964478, + "learning_rate": 8.691191548598923e-06, + "loss": 0.9404, + "step": 10968 + }, + { + "epoch": 0.7536759653703449, + "grad_norm": 1.1061797142028809, + "learning_rate": 8.686647310842407e-06, + "loss": 0.9403, + "step": 10969 + }, + { + "epoch": 0.7537446750034354, + "grad_norm": 1.040846586227417, + "learning_rate": 8.682104011583123e-06, + "loss": 0.9105, + "step": 10970 + }, + { + "epoch": 0.753813384636526, + "grad_norm": 1.1352466344833374, + "learning_rate": 8.677561651082447e-06, + "loss": 0.9745, + "step": 10971 + }, + { + "epoch": 0.7538820942696166, + "grad_norm": 1.110055923461914, + "learning_rate": 8.673020229601705e-06, + "loss": 0.9477, + "step": 10972 + }, + { + "epoch": 0.7539508039027072, + "grad_norm": 1.0057859420776367, + "learning_rate": 8.668479747402145e-06, + "loss": 0.9022, + "step": 10973 + }, + { + "epoch": 0.7540195135357978, + "grad_norm": 1.100290298461914, + "learning_rate": 8.663940204744987e-06, + "loss": 0.9567, + "step": 10974 + }, + { + "epoch": 0.7540882231688882, + "grad_norm": 1.0389660596847534, + "learning_rate": 8.65940160189139e-06, + "loss": 0.9027, + "step": 10975 + }, + { + "epoch": 0.7541569328019788, + "grad_norm": 1.0447784662246704, + "learning_rate": 8.654863939102453e-06, + "loss": 0.9365, + "step": 10976 + }, + { + "epoch": 0.7542256424350694, + "grad_norm": 1.0479130744934082, + "learning_rate": 8.650327216639237e-06, + "loss": 0.9276, + "step": 10977 + }, + { + "epoch": 0.75429435206816, + "grad_norm": 1.0638858079910278, + "learning_rate": 8.645791434762718e-06, + "loss": 0.9634, + "step": 10978 + }, + { + "epoch": 0.7543630617012506, + "grad_norm": 1.0437277555465698, + "learning_rate": 8.64125659373385e-06, + "loss": 0.9282, + "step": 10979 + }, + { + "epoch": 0.754431771334341, + "grad_norm": 1.0648707151412964, + "learning_rate": 8.63672269381351e-06, + "loss": 0.9087, + "step": 10980 + }, + { + "epoch": 0.7545004809674316, + "grad_norm": 1.0880236625671387, + "learning_rate": 8.632189735262547e-06, + "loss": 0.9832, + "step": 10981 + }, + { + "epoch": 0.7545691906005222, + "grad_norm": 1.030439853668213, + "learning_rate": 8.627657718341722e-06, + "loss": 0.9605, + "step": 10982 + }, + { + "epoch": 0.7546379002336128, + "grad_norm": 1.0544850826263428, + "learning_rate": 8.623126643311762e-06, + "loss": 0.9261, + "step": 10983 + }, + { + "epoch": 0.7547066098667033, + "grad_norm": 1.1628659963607788, + "learning_rate": 8.618596510433352e-06, + "loss": 0.9508, + "step": 10984 + }, + { + "epoch": 0.7547753194997938, + "grad_norm": 1.086205005645752, + "learning_rate": 8.614067319967086e-06, + "loss": 0.9454, + "step": 10985 + }, + { + "epoch": 0.7548440291328844, + "grad_norm": 1.0593351125717163, + "learning_rate": 8.609539072173539e-06, + "loss": 0.9543, + "step": 10986 + }, + { + "epoch": 0.754912738765975, + "grad_norm": 1.038906216621399, + "learning_rate": 8.605011767313223e-06, + "loss": 0.915, + "step": 10987 + }, + { + "epoch": 0.7549814483990656, + "grad_norm": 1.0327271223068237, + "learning_rate": 8.600485405646575e-06, + "loss": 0.9276, + "step": 10988 + }, + { + "epoch": 0.7550501580321561, + "grad_norm": 1.0748878717422485, + "learning_rate": 8.595959987434e-06, + "loss": 0.9364, + "step": 10989 + }, + { + "epoch": 0.7551188676652467, + "grad_norm": 1.0427024364471436, + "learning_rate": 8.591435512935847e-06, + "loss": 0.9329, + "step": 10990 + }, + { + "epoch": 0.7551875772983372, + "grad_norm": 1.0395221710205078, + "learning_rate": 8.586911982412401e-06, + "loss": 0.9235, + "step": 10991 + }, + { + "epoch": 0.7552562869314278, + "grad_norm": 1.0939809083938599, + "learning_rate": 8.582389396123914e-06, + "loss": 0.9214, + "step": 10992 + }, + { + "epoch": 0.7553249965645183, + "grad_norm": 1.0022865533828735, + "learning_rate": 8.577867754330544e-06, + "loss": 0.8837, + "step": 10993 + }, + { + "epoch": 0.7553937061976089, + "grad_norm": 1.0316601991653442, + "learning_rate": 8.573347057292428e-06, + "loss": 0.9301, + "step": 10994 + }, + { + "epoch": 0.7554624158306995, + "grad_norm": 1.0373286008834839, + "learning_rate": 8.568827305269641e-06, + "loss": 0.9347, + "step": 10995 + }, + { + "epoch": 0.75553112546379, + "grad_norm": 1.0735617876052856, + "learning_rate": 8.5643084985222e-06, + "loss": 0.9633, + "step": 10996 + }, + { + "epoch": 0.7555998350968806, + "grad_norm": 1.0778905153274536, + "learning_rate": 8.559790637310081e-06, + "loss": 0.9399, + "step": 10997 + }, + { + "epoch": 0.7556685447299711, + "grad_norm": 1.1075770854949951, + "learning_rate": 8.55527372189317e-06, + "loss": 0.9668, + "step": 10998 + }, + { + "epoch": 0.7557372543630617, + "grad_norm": 1.122143030166626, + "learning_rate": 8.550757752531335e-06, + "loss": 0.9522, + "step": 10999 + }, + { + "epoch": 0.7558059639961523, + "grad_norm": 1.041176199913025, + "learning_rate": 8.54624272948438e-06, + "loss": 0.9648, + "step": 11000 + }, + { + "epoch": 0.7558746736292428, + "grad_norm": 1.02454674243927, + "learning_rate": 8.541728653012046e-06, + "loss": 0.9117, + "step": 11001 + }, + { + "epoch": 0.7559433832623333, + "grad_norm": 1.0545276403427124, + "learning_rate": 8.537215523374038e-06, + "loss": 0.9636, + "step": 11002 + }, + { + "epoch": 0.7560120928954239, + "grad_norm": 1.0425225496292114, + "learning_rate": 8.532703340829973e-06, + "loss": 0.8772, + "step": 11003 + }, + { + "epoch": 0.7560808025285145, + "grad_norm": 1.0803571939468384, + "learning_rate": 8.528192105639443e-06, + "loss": 0.957, + "step": 11004 + }, + { + "epoch": 0.7561495121616051, + "grad_norm": 1.0398104190826416, + "learning_rate": 8.52368181806198e-06, + "loss": 0.9572, + "step": 11005 + }, + { + "epoch": 0.7562182217946957, + "grad_norm": 1.1106301546096802, + "learning_rate": 8.519172478357063e-06, + "loss": 0.9471, + "step": 11006 + }, + { + "epoch": 0.7562869314277861, + "grad_norm": 1.0280600786209106, + "learning_rate": 8.514664086784097e-06, + "loss": 0.912, + "step": 11007 + }, + { + "epoch": 0.7563556410608767, + "grad_norm": 1.0619004964828491, + "learning_rate": 8.510156643602454e-06, + "loss": 0.9053, + "step": 11008 + }, + { + "epoch": 0.7564243506939673, + "grad_norm": 1.052268624305725, + "learning_rate": 8.505650149071453e-06, + "loss": 0.8929, + "step": 11009 + }, + { + "epoch": 0.7564930603270579, + "grad_norm": 1.1699137687683105, + "learning_rate": 8.501144603450333e-06, + "loss": 0.9083, + "step": 11010 + }, + { + "epoch": 0.7565617699601485, + "grad_norm": 1.0572820901870728, + "learning_rate": 8.496640006998305e-06, + "loss": 0.918, + "step": 11011 + }, + { + "epoch": 0.7566304795932389, + "grad_norm": 0.9705283641815186, + "learning_rate": 8.492136359974515e-06, + "loss": 0.9023, + "step": 11012 + }, + { + "epoch": 0.7566991892263295, + "grad_norm": 1.08558189868927, + "learning_rate": 8.48763366263806e-06, + "loss": 0.9733, + "step": 11013 + }, + { + "epoch": 0.7567678988594201, + "grad_norm": 0.9857194423675537, + "learning_rate": 8.483131915247968e-06, + "loss": 0.9184, + "step": 11014 + }, + { + "epoch": 0.7568366084925107, + "grad_norm": 1.149946928024292, + "learning_rate": 8.478631118063223e-06, + "loss": 0.9366, + "step": 11015 + }, + { + "epoch": 0.7569053181256012, + "grad_norm": 1.1478796005249023, + "learning_rate": 8.474131271342762e-06, + "loss": 0.9965, + "step": 11016 + }, + { + "epoch": 0.7569740277586917, + "grad_norm": 1.0142592191696167, + "learning_rate": 8.469632375345446e-06, + "loss": 0.9208, + "step": 11017 + }, + { + "epoch": 0.7570427373917823, + "grad_norm": 1.075839638710022, + "learning_rate": 8.46513443033011e-06, + "loss": 0.9428, + "step": 11018 + }, + { + "epoch": 0.7571114470248729, + "grad_norm": 1.1420314311981201, + "learning_rate": 8.460637436555502e-06, + "loss": 0.9589, + "step": 11019 + }, + { + "epoch": 0.7571801566579635, + "grad_norm": 1.0960354804992676, + "learning_rate": 8.456141394280338e-06, + "loss": 0.9366, + "step": 11020 + }, + { + "epoch": 0.757248866291054, + "grad_norm": 1.0292688608169556, + "learning_rate": 8.451646303763269e-06, + "loss": 0.9037, + "step": 11021 + }, + { + "epoch": 0.7573175759241446, + "grad_norm": 1.0231105089187622, + "learning_rate": 8.4471521652629e-06, + "loss": 0.9295, + "step": 11022 + }, + { + "epoch": 0.7573862855572351, + "grad_norm": 1.045095682144165, + "learning_rate": 8.442658979037782e-06, + "loss": 0.9348, + "step": 11023 + }, + { + "epoch": 0.7574549951903257, + "grad_norm": 1.0737255811691284, + "learning_rate": 8.43816674534639e-06, + "loss": 0.9296, + "step": 11024 + }, + { + "epoch": 0.7575237048234162, + "grad_norm": 1.0755504369735718, + "learning_rate": 8.433675464447166e-06, + "loss": 0.9163, + "step": 11025 + }, + { + "epoch": 0.7575924144565068, + "grad_norm": 1.1545671224594116, + "learning_rate": 8.42918513659849e-06, + "loss": 0.9842, + "step": 11026 + }, + { + "epoch": 0.7576611240895974, + "grad_norm": 1.0701504945755005, + "learning_rate": 8.424695762058692e-06, + "loss": 0.8988, + "step": 11027 + }, + { + "epoch": 0.7577298337226879, + "grad_norm": 1.1010606288909912, + "learning_rate": 8.420207341086048e-06, + "loss": 0.9401, + "step": 11028 + }, + { + "epoch": 0.7577985433557785, + "grad_norm": 1.0744140148162842, + "learning_rate": 8.415719873938757e-06, + "loss": 0.9464, + "step": 11029 + }, + { + "epoch": 0.757867252988869, + "grad_norm": 1.0206507444381714, + "learning_rate": 8.411233360874993e-06, + "loss": 0.9372, + "step": 11030 + }, + { + "epoch": 0.7579359626219596, + "grad_norm": 1.0539140701293945, + "learning_rate": 8.406747802152865e-06, + "loss": 0.9222, + "step": 11031 + }, + { + "epoch": 0.7580046722550502, + "grad_norm": 1.1149860620498657, + "learning_rate": 8.402263198030407e-06, + "loss": 0.9649, + "step": 11032 + }, + { + "epoch": 0.7580733818881407, + "grad_norm": 1.0515234470367432, + "learning_rate": 8.397779548765642e-06, + "loss": 0.9294, + "step": 11033 + }, + { + "epoch": 0.7581420915212312, + "grad_norm": 1.0620290040969849, + "learning_rate": 8.393296854616483e-06, + "loss": 0.9201, + "step": 11034 + }, + { + "epoch": 0.7582108011543218, + "grad_norm": 1.0313959121704102, + "learning_rate": 8.388815115840832e-06, + "loss": 0.942, + "step": 11035 + }, + { + "epoch": 0.7582795107874124, + "grad_norm": 1.0680822134017944, + "learning_rate": 8.38433433269652e-06, + "loss": 0.9484, + "step": 11036 + }, + { + "epoch": 0.758348220420503, + "grad_norm": 1.0964330434799194, + "learning_rate": 8.37985450544132e-06, + "loss": 0.9121, + "step": 11037 + }, + { + "epoch": 0.7584169300535936, + "grad_norm": 1.0642518997192383, + "learning_rate": 8.375375634332965e-06, + "loss": 0.931, + "step": 11038 + }, + { + "epoch": 0.758485639686684, + "grad_norm": 1.0222986936569214, + "learning_rate": 8.370897719629109e-06, + "loss": 0.9411, + "step": 11039 + }, + { + "epoch": 0.7585543493197746, + "grad_norm": 1.1067192554473877, + "learning_rate": 8.366420761587365e-06, + "loss": 0.984, + "step": 11040 + }, + { + "epoch": 0.7586230589528652, + "grad_norm": 1.0888422727584839, + "learning_rate": 8.361944760465293e-06, + "loss": 0.8986, + "step": 11041 + }, + { + "epoch": 0.7586917685859558, + "grad_norm": 1.0441977977752686, + "learning_rate": 8.35746971652039e-06, + "loss": 0.8907, + "step": 11042 + }, + { + "epoch": 0.7587604782190464, + "grad_norm": 1.0533286333084106, + "learning_rate": 8.352995630010122e-06, + "loss": 0.9016, + "step": 11043 + }, + { + "epoch": 0.7588291878521368, + "grad_norm": 1.0419100522994995, + "learning_rate": 8.348522501191853e-06, + "loss": 0.9099, + "step": 11044 + }, + { + "epoch": 0.7588978974852274, + "grad_norm": 1.1214816570281982, + "learning_rate": 8.344050330322934e-06, + "loss": 0.9335, + "step": 11045 + }, + { + "epoch": 0.758966607118318, + "grad_norm": 1.0849274396896362, + "learning_rate": 8.33957911766064e-06, + "loss": 0.9625, + "step": 11046 + }, + { + "epoch": 0.7590353167514086, + "grad_norm": 1.0654568672180176, + "learning_rate": 8.335108863462204e-06, + "loss": 0.9437, + "step": 11047 + }, + { + "epoch": 0.7591040263844991, + "grad_norm": 1.091776967048645, + "learning_rate": 8.3306395679848e-06, + "loss": 0.9707, + "step": 11048 + }, + { + "epoch": 0.7591727360175896, + "grad_norm": 1.0747005939483643, + "learning_rate": 8.326171231485533e-06, + "loss": 0.9184, + "step": 11049 + }, + { + "epoch": 0.7592414456506802, + "grad_norm": 1.0603246688842773, + "learning_rate": 8.321703854221465e-06, + "loss": 0.916, + "step": 11050 + }, + { + "epoch": 0.7593101552837708, + "grad_norm": 1.1144541501998901, + "learning_rate": 8.317237436449606e-06, + "loss": 0.9821, + "step": 11051 + }, + { + "epoch": 0.7593788649168614, + "grad_norm": 1.122541069984436, + "learning_rate": 8.312771978426906e-06, + "loss": 0.9721, + "step": 11052 + }, + { + "epoch": 0.7594475745499519, + "grad_norm": 1.030753254890442, + "learning_rate": 8.308307480410266e-06, + "loss": 0.9388, + "step": 11053 + }, + { + "epoch": 0.7595162841830425, + "grad_norm": 1.0798627138137817, + "learning_rate": 8.303843942656512e-06, + "loss": 0.9575, + "step": 11054 + }, + { + "epoch": 0.759584993816133, + "grad_norm": 1.0192960500717163, + "learning_rate": 8.299381365422446e-06, + "loss": 0.9341, + "step": 11055 + }, + { + "epoch": 0.7596537034492236, + "grad_norm": 1.0283061265945435, + "learning_rate": 8.294919748964775e-06, + "loss": 0.9267, + "step": 11056 + }, + { + "epoch": 0.7597224130823141, + "grad_norm": 1.0353591442108154, + "learning_rate": 8.290459093540187e-06, + "loss": 0.9395, + "step": 11057 + }, + { + "epoch": 0.7597911227154047, + "grad_norm": 1.0814632177352905, + "learning_rate": 8.285999399405309e-06, + "loss": 0.9444, + "step": 11058 + }, + { + "epoch": 0.7598598323484953, + "grad_norm": 1.0849076509475708, + "learning_rate": 8.281540666816684e-06, + "loss": 0.9649, + "step": 11059 + }, + { + "epoch": 0.7599285419815858, + "grad_norm": 1.019100308418274, + "learning_rate": 8.27708289603083e-06, + "loss": 0.9026, + "step": 11060 + }, + { + "epoch": 0.7599972516146764, + "grad_norm": 1.1244791746139526, + "learning_rate": 8.272626087304203e-06, + "loss": 0.9712, + "step": 11061 + }, + { + "epoch": 0.7600659612477669, + "grad_norm": 1.0842761993408203, + "learning_rate": 8.268170240893195e-06, + "loss": 0.9625, + "step": 11062 + }, + { + "epoch": 0.7601346708808575, + "grad_norm": 1.078717589378357, + "learning_rate": 8.263715357054153e-06, + "loss": 0.9266, + "step": 11063 + }, + { + "epoch": 0.7602033805139481, + "grad_norm": 1.0332452058792114, + "learning_rate": 8.25926143604337e-06, + "loss": 0.9218, + "step": 11064 + }, + { + "epoch": 0.7602720901470386, + "grad_norm": 0.9785361886024475, + "learning_rate": 8.25480847811706e-06, + "loss": 0.9277, + "step": 11065 + }, + { + "epoch": 0.7603407997801291, + "grad_norm": 1.072564959526062, + "learning_rate": 8.25035648353141e-06, + "loss": 0.926, + "step": 11066 + }, + { + "epoch": 0.7604095094132197, + "grad_norm": 1.047780990600586, + "learning_rate": 8.245905452542538e-06, + "loss": 0.9331, + "step": 11067 + }, + { + "epoch": 0.7604782190463103, + "grad_norm": 1.0948518514633179, + "learning_rate": 8.24145538540651e-06, + "loss": 0.9054, + "step": 11068 + }, + { + "epoch": 0.7605469286794009, + "grad_norm": 1.0701090097427368, + "learning_rate": 8.237006282379342e-06, + "loss": 0.939, + "step": 11069 + }, + { + "epoch": 0.7606156383124915, + "grad_norm": 1.0641071796417236, + "learning_rate": 8.232558143716976e-06, + "loss": 0.9552, + "step": 11070 + }, + { + "epoch": 0.7606843479455819, + "grad_norm": 1.0439032316207886, + "learning_rate": 8.228110969675313e-06, + "loss": 0.9499, + "step": 11071 + }, + { + "epoch": 0.7607530575786725, + "grad_norm": 1.151658058166504, + "learning_rate": 8.223664760510202e-06, + "loss": 1.0015, + "step": 11072 + }, + { + "epoch": 0.7608217672117631, + "grad_norm": 1.0540932416915894, + "learning_rate": 8.219219516477428e-06, + "loss": 0.9301, + "step": 11073 + }, + { + "epoch": 0.7608904768448537, + "grad_norm": 1.0697367191314697, + "learning_rate": 8.214775237832731e-06, + "loss": 0.9621, + "step": 11074 + }, + { + "epoch": 0.7609591864779442, + "grad_norm": 0.9985745549201965, + "learning_rate": 8.210331924831771e-06, + "loss": 0.9082, + "step": 11075 + }, + { + "epoch": 0.7610278961110347, + "grad_norm": 1.030806541442871, + "learning_rate": 8.20588957773018e-06, + "loss": 0.9072, + "step": 11076 + }, + { + "epoch": 0.7610966057441253, + "grad_norm": 1.1212027072906494, + "learning_rate": 8.201448196783528e-06, + "loss": 0.9437, + "step": 11077 + }, + { + "epoch": 0.7611653153772159, + "grad_norm": 1.08612060546875, + "learning_rate": 8.197007782247313e-06, + "loss": 0.9223, + "step": 11078 + }, + { + "epoch": 0.7612340250103065, + "grad_norm": 1.0211478471755981, + "learning_rate": 8.192568334376994e-06, + "loss": 0.8998, + "step": 11079 + }, + { + "epoch": 0.761302734643397, + "grad_norm": 1.061509609222412, + "learning_rate": 8.188129853427978e-06, + "loss": 0.9561, + "step": 11080 + }, + { + "epoch": 0.7613714442764876, + "grad_norm": 1.037561058998108, + "learning_rate": 8.183692339655591e-06, + "loss": 0.9193, + "step": 11081 + }, + { + "epoch": 0.7614401539095781, + "grad_norm": 1.0482712984085083, + "learning_rate": 8.179255793315132e-06, + "loss": 0.9066, + "step": 11082 + }, + { + "epoch": 0.7615088635426687, + "grad_norm": 1.0617033243179321, + "learning_rate": 8.174820214661832e-06, + "loss": 0.9502, + "step": 11083 + }, + { + "epoch": 0.7615775731757592, + "grad_norm": 1.1277352571487427, + "learning_rate": 8.170385603950873e-06, + "loss": 0.9468, + "step": 11084 + }, + { + "epoch": 0.7616462828088498, + "grad_norm": 1.0527620315551758, + "learning_rate": 8.165951961437363e-06, + "loss": 0.9513, + "step": 11085 + }, + { + "epoch": 0.7617149924419404, + "grad_norm": 1.0693130493164062, + "learning_rate": 8.161519287376368e-06, + "loss": 0.9344, + "step": 11086 + }, + { + "epoch": 0.7617837020750309, + "grad_norm": 1.0074228048324585, + "learning_rate": 8.157087582022904e-06, + "loss": 0.8855, + "step": 11087 + }, + { + "epoch": 0.7618524117081215, + "grad_norm": 1.0490732192993164, + "learning_rate": 8.152656845631923e-06, + "loss": 0.9341, + "step": 11088 + }, + { + "epoch": 0.761921121341212, + "grad_norm": 1.0335043668746948, + "learning_rate": 8.148227078458326e-06, + "loss": 0.94, + "step": 11089 + }, + { + "epoch": 0.7619898309743026, + "grad_norm": 1.0789945125579834, + "learning_rate": 8.143798280756945e-06, + "loss": 0.9066, + "step": 11090 + }, + { + "epoch": 0.7620585406073932, + "grad_norm": 1.011289358139038, + "learning_rate": 8.139370452782568e-06, + "loss": 0.8916, + "step": 11091 + }, + { + "epoch": 0.7621272502404837, + "grad_norm": 1.0248743295669556, + "learning_rate": 8.134943594789932e-06, + "loss": 0.9291, + "step": 11092 + }, + { + "epoch": 0.7621959598735742, + "grad_norm": 1.0352097749710083, + "learning_rate": 8.130517707033708e-06, + "loss": 0.9008, + "step": 11093 + }, + { + "epoch": 0.7622646695066648, + "grad_norm": 1.0296422243118286, + "learning_rate": 8.126092789768519e-06, + "loss": 0.8967, + "step": 11094 + }, + { + "epoch": 0.7623333791397554, + "grad_norm": 1.0122946500778198, + "learning_rate": 8.121668843248917e-06, + "loss": 0.9171, + "step": 11095 + }, + { + "epoch": 0.762402088772846, + "grad_norm": 1.0184937715530396, + "learning_rate": 8.117245867729415e-06, + "loss": 0.9138, + "step": 11096 + }, + { + "epoch": 0.7624707984059366, + "grad_norm": 1.134658694267273, + "learning_rate": 8.112823863464464e-06, + "loss": 0.9507, + "step": 11097 + }, + { + "epoch": 0.762539508039027, + "grad_norm": 1.0143793821334839, + "learning_rate": 8.108402830708461e-06, + "loss": 0.9264, + "step": 11098 + }, + { + "epoch": 0.7626082176721176, + "grad_norm": 1.0220500230789185, + "learning_rate": 8.10398276971575e-06, + "loss": 0.9662, + "step": 11099 + }, + { + "epoch": 0.7626769273052082, + "grad_norm": 1.0598366260528564, + "learning_rate": 8.099563680740601e-06, + "loss": 0.9458, + "step": 11100 + }, + { + "epoch": 0.7627456369382988, + "grad_norm": 1.1485726833343506, + "learning_rate": 8.095145564037249e-06, + "loss": 0.9722, + "step": 11101 + }, + { + "epoch": 0.7628143465713894, + "grad_norm": 1.0405409336090088, + "learning_rate": 8.090728419859872e-06, + "loss": 0.939, + "step": 11102 + }, + { + "epoch": 0.7628830562044798, + "grad_norm": 1.0473769903182983, + "learning_rate": 8.086312248462572e-06, + "loss": 0.9395, + "step": 11103 + }, + { + "epoch": 0.7629517658375704, + "grad_norm": 1.1031358242034912, + "learning_rate": 8.081897050099424e-06, + "loss": 0.9472, + "step": 11104 + }, + { + "epoch": 0.763020475470661, + "grad_norm": 1.0040346384048462, + "learning_rate": 8.077482825024413e-06, + "loss": 0.9381, + "step": 11105 + }, + { + "epoch": 0.7630891851037516, + "grad_norm": 1.0478094816207886, + "learning_rate": 8.073069573491498e-06, + "loss": 0.9418, + "step": 11106 + }, + { + "epoch": 0.7631578947368421, + "grad_norm": 1.0742228031158447, + "learning_rate": 8.06865729575457e-06, + "loss": 0.962, + "step": 11107 + }, + { + "epoch": 0.7632266043699326, + "grad_norm": 1.0426807403564453, + "learning_rate": 8.064245992067463e-06, + "loss": 0.9309, + "step": 11108 + }, + { + "epoch": 0.7632953140030232, + "grad_norm": 0.9990839958190918, + "learning_rate": 8.059835662683968e-06, + "loss": 0.9118, + "step": 11109 + }, + { + "epoch": 0.7633640236361138, + "grad_norm": 1.1115576028823853, + "learning_rate": 8.05542630785779e-06, + "loss": 0.9654, + "step": 11110 + }, + { + "epoch": 0.7634327332692044, + "grad_norm": 0.9986082315444946, + "learning_rate": 8.051017927842602e-06, + "loss": 0.9032, + "step": 11111 + }, + { + "epoch": 0.7635014429022949, + "grad_norm": 1.0754740238189697, + "learning_rate": 8.04661052289202e-06, + "loss": 0.9402, + "step": 11112 + }, + { + "epoch": 0.7635701525353855, + "grad_norm": 1.0430762767791748, + "learning_rate": 8.042204093259598e-06, + "loss": 0.9135, + "step": 11113 + }, + { + "epoch": 0.763638862168476, + "grad_norm": 0.9944958090782166, + "learning_rate": 8.037798639198844e-06, + "loss": 0.9149, + "step": 11114 + }, + { + "epoch": 0.7637075718015666, + "grad_norm": 1.0875730514526367, + "learning_rate": 8.03339416096318e-06, + "loss": 0.9328, + "step": 11115 + }, + { + "epoch": 0.7637762814346571, + "grad_norm": 1.0756598711013794, + "learning_rate": 8.028990658806007e-06, + "loss": 0.9157, + "step": 11116 + }, + { + "epoch": 0.7638449910677477, + "grad_norm": 1.0015400648117065, + "learning_rate": 8.024588132980654e-06, + "loss": 0.926, + "step": 11117 + }, + { + "epoch": 0.7639137007008383, + "grad_norm": 1.0648759603500366, + "learning_rate": 8.020186583740394e-06, + "loss": 0.9184, + "step": 11118 + }, + { + "epoch": 0.7639824103339288, + "grad_norm": 1.007530689239502, + "learning_rate": 8.01578601133845e-06, + "loss": 0.9028, + "step": 11119 + }, + { + "epoch": 0.7640511199670194, + "grad_norm": 1.0232187509536743, + "learning_rate": 8.011386416027985e-06, + "loss": 0.9468, + "step": 11120 + }, + { + "epoch": 0.7641198296001099, + "grad_norm": 1.1177618503570557, + "learning_rate": 8.006987798062096e-06, + "loss": 0.9186, + "step": 11121 + }, + { + "epoch": 0.7641885392332005, + "grad_norm": 1.060634970664978, + "learning_rate": 8.002590157693834e-06, + "loss": 0.9419, + "step": 11122 + }, + { + "epoch": 0.7642572488662911, + "grad_norm": 1.0957165956497192, + "learning_rate": 7.998193495176201e-06, + "loss": 0.9126, + "step": 11123 + }, + { + "epoch": 0.7643259584993816, + "grad_norm": 0.9977865815162659, + "learning_rate": 7.993797810762136e-06, + "loss": 0.9055, + "step": 11124 + }, + { + "epoch": 0.7643946681324721, + "grad_norm": 1.1255637407302856, + "learning_rate": 7.989403104704504e-06, + "loss": 0.9643, + "step": 11125 + }, + { + "epoch": 0.7644633777655627, + "grad_norm": 1.0251085758209229, + "learning_rate": 7.985009377256147e-06, + "loss": 0.9108, + "step": 11126 + }, + { + "epoch": 0.7645320873986533, + "grad_norm": 1.0263293981552124, + "learning_rate": 7.98061662866982e-06, + "loss": 0.9109, + "step": 11127 + }, + { + "epoch": 0.7646007970317439, + "grad_norm": 1.124448299407959, + "learning_rate": 7.976224859198242e-06, + "loss": 0.9428, + "step": 11128 + }, + { + "epoch": 0.7646695066648345, + "grad_norm": 1.0550551414489746, + "learning_rate": 7.971834069094063e-06, + "loss": 0.94, + "step": 11129 + }, + { + "epoch": 0.7647382162979249, + "grad_norm": 1.0325058698654175, + "learning_rate": 7.967444258609902e-06, + "loss": 0.9221, + "step": 11130 + }, + { + "epoch": 0.7648069259310155, + "grad_norm": 1.120269775390625, + "learning_rate": 7.963055427998276e-06, + "loss": 0.9823, + "step": 11131 + }, + { + "epoch": 0.7648756355641061, + "grad_norm": 1.0590237379074097, + "learning_rate": 7.958667577511683e-06, + "loss": 0.9377, + "step": 11132 + }, + { + "epoch": 0.7649443451971967, + "grad_norm": 1.0397861003875732, + "learning_rate": 7.954280707402554e-06, + "loss": 0.9231, + "step": 11133 + }, + { + "epoch": 0.7650130548302873, + "grad_norm": 1.0047417879104614, + "learning_rate": 7.949894817923263e-06, + "loss": 0.8745, + "step": 11134 + }, + { + "epoch": 0.7650817644633777, + "grad_norm": 1.0335538387298584, + "learning_rate": 7.945509909326135e-06, + "loss": 0.9254, + "step": 11135 + }, + { + "epoch": 0.7651504740964683, + "grad_norm": 1.0571271181106567, + "learning_rate": 7.941125981863416e-06, + "loss": 0.8822, + "step": 11136 + }, + { + "epoch": 0.7652191837295589, + "grad_norm": 1.0635457038879395, + "learning_rate": 7.936743035787314e-06, + "loss": 0.8907, + "step": 11137 + }, + { + "epoch": 0.7652878933626495, + "grad_norm": 1.009445071220398, + "learning_rate": 7.932361071349983e-06, + "loss": 0.9072, + "step": 11138 + }, + { + "epoch": 0.76535660299574, + "grad_norm": 1.0103905200958252, + "learning_rate": 7.927980088803517e-06, + "loss": 0.9442, + "step": 11139 + }, + { + "epoch": 0.7654253126288305, + "grad_norm": 1.048002004623413, + "learning_rate": 7.923600088399951e-06, + "loss": 0.9241, + "step": 11140 + }, + { + "epoch": 0.7654940222619211, + "grad_norm": 0.9952563643455505, + "learning_rate": 7.919221070391253e-06, + "loss": 0.8963, + "step": 11141 + }, + { + "epoch": 0.7655627318950117, + "grad_norm": 1.1094937324523926, + "learning_rate": 7.914843035029351e-06, + "loss": 0.9409, + "step": 11142 + }, + { + "epoch": 0.7656314415281023, + "grad_norm": 1.0189684629440308, + "learning_rate": 7.910465982566112e-06, + "loss": 0.9345, + "step": 11143 + }, + { + "epoch": 0.7657001511611928, + "grad_norm": 1.0391961336135864, + "learning_rate": 7.906089913253348e-06, + "loss": 0.9472, + "step": 11144 + }, + { + "epoch": 0.7657688607942834, + "grad_norm": 1.096741795539856, + "learning_rate": 7.901714827342813e-06, + "loss": 0.924, + "step": 11145 + }, + { + "epoch": 0.7658375704273739, + "grad_norm": 1.0197635889053345, + "learning_rate": 7.89734072508619e-06, + "loss": 0.9114, + "step": 11146 + }, + { + "epoch": 0.7659062800604645, + "grad_norm": 1.0633102655410767, + "learning_rate": 7.89296760673513e-06, + "loss": 0.9422, + "step": 11147 + }, + { + "epoch": 0.765974989693555, + "grad_norm": 1.0365355014801025, + "learning_rate": 7.888595472541213e-06, + "loss": 0.9027, + "step": 11148 + }, + { + "epoch": 0.7660436993266456, + "grad_norm": 1.0368895530700684, + "learning_rate": 7.88422432275597e-06, + "loss": 0.884, + "step": 11149 + }, + { + "epoch": 0.7661124089597362, + "grad_norm": 1.0483636856079102, + "learning_rate": 7.879854157630861e-06, + "loss": 0.9046, + "step": 11150 + }, + { + "epoch": 0.7661811185928267, + "grad_norm": 1.0556209087371826, + "learning_rate": 7.87548497741731e-06, + "loss": 0.9252, + "step": 11151 + }, + { + "epoch": 0.7662498282259173, + "grad_norm": 1.028160810470581, + "learning_rate": 7.871116782366658e-06, + "loss": 0.8945, + "step": 11152 + }, + { + "epoch": 0.7663185378590078, + "grad_norm": 0.9882762432098389, + "learning_rate": 7.866749572730217e-06, + "loss": 0.9005, + "step": 11153 + }, + { + "epoch": 0.7663872474920984, + "grad_norm": 1.0705714225769043, + "learning_rate": 7.862383348759222e-06, + "loss": 0.9402, + "step": 11154 + }, + { + "epoch": 0.766455957125189, + "grad_norm": 1.0580456256866455, + "learning_rate": 7.858018110704873e-06, + "loss": 0.9419, + "step": 11155 + }, + { + "epoch": 0.7665246667582795, + "grad_norm": 1.0837332010269165, + "learning_rate": 7.853653858818283e-06, + "loss": 0.9594, + "step": 11156 + }, + { + "epoch": 0.76659337639137, + "grad_norm": 1.02823007106781, + "learning_rate": 7.84929059335053e-06, + "loss": 0.902, + "step": 11157 + }, + { + "epoch": 0.7666620860244606, + "grad_norm": 1.065269112586975, + "learning_rate": 7.844928314552636e-06, + "loss": 0.9302, + "step": 11158 + }, + { + "epoch": 0.7667307956575512, + "grad_norm": 1.0872458219528198, + "learning_rate": 7.840567022675552e-06, + "loss": 0.9717, + "step": 11159 + }, + { + "epoch": 0.7667995052906418, + "grad_norm": 1.0619072914123535, + "learning_rate": 7.836206717970193e-06, + "loss": 0.9278, + "step": 11160 + }, + { + "epoch": 0.7668682149237324, + "grad_norm": 1.0509952306747437, + "learning_rate": 7.83184740068739e-06, + "loss": 0.9191, + "step": 11161 + }, + { + "epoch": 0.7669369245568228, + "grad_norm": 1.1002849340438843, + "learning_rate": 7.827489071077937e-06, + "loss": 0.9288, + "step": 11162 + }, + { + "epoch": 0.7670056341899134, + "grad_norm": 1.0621320009231567, + "learning_rate": 7.823131729392566e-06, + "loss": 0.9018, + "step": 11163 + }, + { + "epoch": 0.767074343823004, + "grad_norm": 1.0776278972625732, + "learning_rate": 7.818775375881953e-06, + "loss": 0.9746, + "step": 11164 + }, + { + "epoch": 0.7671430534560946, + "grad_norm": 1.0073994398117065, + "learning_rate": 7.814420010796728e-06, + "loss": 0.9094, + "step": 11165 + }, + { + "epoch": 0.7672117630891852, + "grad_norm": 1.1433724164962769, + "learning_rate": 7.81006563438743e-06, + "loss": 0.9578, + "step": 11166 + }, + { + "epoch": 0.7672804727222756, + "grad_norm": 1.0510663986206055, + "learning_rate": 7.805712246904576e-06, + "loss": 0.9478, + "step": 11167 + }, + { + "epoch": 0.7673491823553662, + "grad_norm": 1.0747885704040527, + "learning_rate": 7.801359848598614e-06, + "loss": 0.9343, + "step": 11168 + }, + { + "epoch": 0.7674178919884568, + "grad_norm": 1.0032528638839722, + "learning_rate": 7.79700843971993e-06, + "loss": 0.9172, + "step": 11169 + }, + { + "epoch": 0.7674866016215474, + "grad_norm": 1.0066473484039307, + "learning_rate": 7.792658020518864e-06, + "loss": 0.9264, + "step": 11170 + }, + { + "epoch": 0.7675553112546379, + "grad_norm": 1.0594090223312378, + "learning_rate": 7.788308591245697e-06, + "loss": 0.9323, + "step": 11171 + }, + { + "epoch": 0.7676240208877284, + "grad_norm": 1.1363641023635864, + "learning_rate": 7.783960152150632e-06, + "loss": 0.9942, + "step": 11172 + }, + { + "epoch": 0.767692730520819, + "grad_norm": 1.0388374328613281, + "learning_rate": 7.779612703483855e-06, + "loss": 0.9328, + "step": 11173 + }, + { + "epoch": 0.7677614401539096, + "grad_norm": 1.0224254131317139, + "learning_rate": 7.775266245495446e-06, + "loss": 0.906, + "step": 11174 + }, + { + "epoch": 0.7678301497870001, + "grad_norm": 1.0742204189300537, + "learning_rate": 7.770920778435473e-06, + "loss": 0.9326, + "step": 11175 + }, + { + "epoch": 0.7678988594200907, + "grad_norm": 1.0260727405548096, + "learning_rate": 7.766576302553927e-06, + "loss": 0.9283, + "step": 11176 + }, + { + "epoch": 0.7679675690531813, + "grad_norm": 1.062631368637085, + "learning_rate": 7.762232818100732e-06, + "loss": 0.9072, + "step": 11177 + }, + { + "epoch": 0.7680362786862718, + "grad_norm": 1.0301467180252075, + "learning_rate": 7.75789032532577e-06, + "loss": 0.9156, + "step": 11178 + }, + { + "epoch": 0.7681049883193624, + "grad_norm": 1.0935660600662231, + "learning_rate": 7.753548824478868e-06, + "loss": 0.9801, + "step": 11179 + }, + { + "epoch": 0.7681736979524529, + "grad_norm": 1.1058202981948853, + "learning_rate": 7.749208315809784e-06, + "loss": 0.9226, + "step": 11180 + }, + { + "epoch": 0.7682424075855435, + "grad_norm": 1.0222420692443848, + "learning_rate": 7.744868799568236e-06, + "loss": 0.9185, + "step": 11181 + }, + { + "epoch": 0.7683111172186341, + "grad_norm": 1.1157913208007812, + "learning_rate": 7.740530276003857e-06, + "loss": 0.945, + "step": 11182 + }, + { + "epoch": 0.7683798268517246, + "grad_norm": 1.051162600517273, + "learning_rate": 7.736192745366248e-06, + "loss": 0.9323, + "step": 11183 + }, + { + "epoch": 0.7684485364848151, + "grad_norm": 0.9976167678833008, + "learning_rate": 7.731856207904941e-06, + "loss": 0.887, + "step": 11184 + }, + { + "epoch": 0.7685172461179057, + "grad_norm": 1.0335952043533325, + "learning_rate": 7.727520663869423e-06, + "loss": 0.9399, + "step": 11185 + }, + { + "epoch": 0.7685859557509963, + "grad_norm": 0.9851285219192505, + "learning_rate": 7.723186113509112e-06, + "loss": 0.9274, + "step": 11186 + }, + { + "epoch": 0.7686546653840869, + "grad_norm": 1.0954639911651611, + "learning_rate": 7.718852557073366e-06, + "loss": 0.9186, + "step": 11187 + }, + { + "epoch": 0.7687233750171774, + "grad_norm": 1.0088008642196655, + "learning_rate": 7.714519994811491e-06, + "loss": 0.9369, + "step": 11188 + }, + { + "epoch": 0.7687920846502679, + "grad_norm": 1.0102341175079346, + "learning_rate": 7.710188426972744e-06, + "loss": 0.8981, + "step": 11189 + }, + { + "epoch": 0.7688607942833585, + "grad_norm": 1.0161590576171875, + "learning_rate": 7.705857853806317e-06, + "loss": 0.9253, + "step": 11190 + }, + { + "epoch": 0.7689295039164491, + "grad_norm": 1.0378055572509766, + "learning_rate": 7.701528275561348e-06, + "loss": 0.9285, + "step": 11191 + }, + { + "epoch": 0.7689982135495397, + "grad_norm": 0.9665999412536621, + "learning_rate": 7.697199692486904e-06, + "loss": 0.9091, + "step": 11192 + }, + { + "epoch": 0.7690669231826303, + "grad_norm": 1.0853599309921265, + "learning_rate": 7.692872104832008e-06, + "loss": 0.946, + "step": 11193 + }, + { + "epoch": 0.7691356328157207, + "grad_norm": 1.0774275064468384, + "learning_rate": 7.68854551284563e-06, + "loss": 0.9142, + "step": 11194 + }, + { + "epoch": 0.7692043424488113, + "grad_norm": 1.0411955118179321, + "learning_rate": 7.684219916776678e-06, + "loss": 0.9003, + "step": 11195 + }, + { + "epoch": 0.7692730520819019, + "grad_norm": 1.11282217502594, + "learning_rate": 7.679895316873991e-06, + "loss": 0.942, + "step": 11196 + }, + { + "epoch": 0.7693417617149925, + "grad_norm": 1.0617843866348267, + "learning_rate": 7.675571713386365e-06, + "loss": 0.9504, + "step": 11197 + }, + { + "epoch": 0.769410471348083, + "grad_norm": 0.9736995100975037, + "learning_rate": 7.671249106562545e-06, + "loss": 0.863, + "step": 11198 + }, + { + "epoch": 0.7694791809811735, + "grad_norm": 1.0602563619613647, + "learning_rate": 7.666927496651185e-06, + "loss": 0.9365, + "step": 11199 + }, + { + "epoch": 0.7695478906142641, + "grad_norm": 1.0195090770721436, + "learning_rate": 7.662606883900922e-06, + "loss": 0.904, + "step": 11200 + }, + { + "epoch": 0.7696166002473547, + "grad_norm": 0.9914824366569519, + "learning_rate": 7.65828726856032e-06, + "loss": 0.926, + "step": 11201 + }, + { + "epoch": 0.7696853098804453, + "grad_norm": 1.0702794790267944, + "learning_rate": 7.653968650877871e-06, + "loss": 0.889, + "step": 11202 + }, + { + "epoch": 0.7697540195135358, + "grad_norm": 1.093085527420044, + "learning_rate": 7.649651031102028e-06, + "loss": 0.9388, + "step": 11203 + }, + { + "epoch": 0.7698227291466263, + "grad_norm": 1.0678292512893677, + "learning_rate": 7.645334409481184e-06, + "loss": 0.9334, + "step": 11204 + }, + { + "epoch": 0.7698914387797169, + "grad_norm": 1.0745975971221924, + "learning_rate": 7.641018786263666e-06, + "loss": 0.9589, + "step": 11205 + }, + { + "epoch": 0.7699601484128075, + "grad_norm": 1.0571296215057373, + "learning_rate": 7.636704161697763e-06, + "loss": 0.9579, + "step": 11206 + }, + { + "epoch": 0.770028858045898, + "grad_norm": 1.0034962892532349, + "learning_rate": 7.632390536031674e-06, + "loss": 0.9273, + "step": 11207 + }, + { + "epoch": 0.7700975676789886, + "grad_norm": 1.0498548746109009, + "learning_rate": 7.6280779095135706e-06, + "loss": 0.9071, + "step": 11208 + }, + { + "epoch": 0.7701662773120792, + "grad_norm": 1.08661687374115, + "learning_rate": 7.623766282391548e-06, + "loss": 0.9715, + "step": 11209 + }, + { + "epoch": 0.7702349869451697, + "grad_norm": 1.0083556175231934, + "learning_rate": 7.61945565491366e-06, + "loss": 0.9062, + "step": 11210 + }, + { + "epoch": 0.7703036965782603, + "grad_norm": 1.0207915306091309, + "learning_rate": 7.615146027327896e-06, + "loss": 0.9298, + "step": 11211 + }, + { + "epoch": 0.7703724062113508, + "grad_norm": 1.133719801902771, + "learning_rate": 7.610837399882176e-06, + "loss": 0.9289, + "step": 11212 + }, + { + "epoch": 0.7704411158444414, + "grad_norm": 1.0522226095199585, + "learning_rate": 7.606529772824375e-06, + "loss": 0.9209, + "step": 11213 + }, + { + "epoch": 0.770509825477532, + "grad_norm": 1.0895500183105469, + "learning_rate": 7.602223146402312e-06, + "loss": 0.9197, + "step": 11214 + }, + { + "epoch": 0.7705785351106225, + "grad_norm": 1.0625128746032715, + "learning_rate": 7.597917520863743e-06, + "loss": 0.9108, + "step": 11215 + }, + { + "epoch": 0.770647244743713, + "grad_norm": 1.0935611724853516, + "learning_rate": 7.593612896456379e-06, + "loss": 0.9694, + "step": 11216 + }, + { + "epoch": 0.7707159543768036, + "grad_norm": 1.070924162864685, + "learning_rate": 7.589309273427842e-06, + "loss": 0.953, + "step": 11217 + }, + { + "epoch": 0.7707846640098942, + "grad_norm": 1.0399484634399414, + "learning_rate": 7.5850066520257255e-06, + "loss": 0.9493, + "step": 11218 + }, + { + "epoch": 0.7708533736429848, + "grad_norm": 1.0562150478363037, + "learning_rate": 7.580705032497556e-06, + "loss": 0.931, + "step": 11219 + }, + { + "epoch": 0.7709220832760753, + "grad_norm": 1.1277976036071777, + "learning_rate": 7.576404415090815e-06, + "loss": 0.9394, + "step": 11220 + }, + { + "epoch": 0.7709907929091658, + "grad_norm": 1.0545899868011475, + "learning_rate": 7.572104800052898e-06, + "loss": 0.9744, + "step": 11221 + }, + { + "epoch": 0.7710595025422564, + "grad_norm": 1.054510235786438, + "learning_rate": 7.567806187631171e-06, + "loss": 0.9766, + "step": 11222 + }, + { + "epoch": 0.771128212175347, + "grad_norm": 1.0081079006195068, + "learning_rate": 7.5635085780729195e-06, + "loss": 0.8995, + "step": 11223 + }, + { + "epoch": 0.7711969218084376, + "grad_norm": 1.0206013917922974, + "learning_rate": 7.5592119716253855e-06, + "loss": 0.9268, + "step": 11224 + }, + { + "epoch": 0.7712656314415282, + "grad_norm": 1.130367636680603, + "learning_rate": 7.5549163685357555e-06, + "loss": 0.9462, + "step": 11225 + }, + { + "epoch": 0.7713343410746186, + "grad_norm": 1.030052661895752, + "learning_rate": 7.550621769051147e-06, + "loss": 0.9315, + "step": 11226 + }, + { + "epoch": 0.7714030507077092, + "grad_norm": 1.1022742986679077, + "learning_rate": 7.546328173418637e-06, + "loss": 0.9762, + "step": 11227 + }, + { + "epoch": 0.7714717603407998, + "grad_norm": 1.0271949768066406, + "learning_rate": 7.542035581885218e-06, + "loss": 0.9418, + "step": 11228 + }, + { + "epoch": 0.7715404699738904, + "grad_norm": 1.049164056777954, + "learning_rate": 7.537743994697846e-06, + "loss": 0.8928, + "step": 11229 + }, + { + "epoch": 0.7716091796069809, + "grad_norm": 1.0817490816116333, + "learning_rate": 7.533453412103417e-06, + "loss": 0.9233, + "step": 11230 + }, + { + "epoch": 0.7716778892400714, + "grad_norm": 1.0581201314926147, + "learning_rate": 7.529163834348762e-06, + "loss": 0.9548, + "step": 11231 + }, + { + "epoch": 0.771746598873162, + "grad_norm": 1.0671364068984985, + "learning_rate": 7.524875261680667e-06, + "loss": 0.9395, + "step": 11232 + }, + { + "epoch": 0.7718153085062526, + "grad_norm": 1.061273217201233, + "learning_rate": 7.520587694345835e-06, + "loss": 0.9553, + "step": 11233 + }, + { + "epoch": 0.7718840181393432, + "grad_norm": 1.0492606163024902, + "learning_rate": 7.516301132590936e-06, + "loss": 0.9497, + "step": 11234 + }, + { + "epoch": 0.7719527277724337, + "grad_norm": 1.0131582021713257, + "learning_rate": 7.51201557666257e-06, + "loss": 0.9104, + "step": 11235 + }, + { + "epoch": 0.7720214374055242, + "grad_norm": 1.098305583000183, + "learning_rate": 7.507731026807288e-06, + "loss": 0.9573, + "step": 11236 + }, + { + "epoch": 0.7720901470386148, + "grad_norm": 1.0176239013671875, + "learning_rate": 7.50344748327158e-06, + "loss": 0.9464, + "step": 11237 + }, + { + "epoch": 0.7721588566717054, + "grad_norm": 1.1639012098312378, + "learning_rate": 7.4991649463018646e-06, + "loss": 0.9292, + "step": 11238 + }, + { + "epoch": 0.7722275663047959, + "grad_norm": 1.1272298097610474, + "learning_rate": 7.494883416144516e-06, + "loss": 0.9112, + "step": 11239 + }, + { + "epoch": 0.7722962759378865, + "grad_norm": 1.0695654153823853, + "learning_rate": 7.490602893045853e-06, + "loss": 0.9616, + "step": 11240 + }, + { + "epoch": 0.7723649855709771, + "grad_norm": 1.0419836044311523, + "learning_rate": 7.486323377252131e-06, + "loss": 0.93, + "step": 11241 + }, + { + "epoch": 0.7724336952040676, + "grad_norm": 1.0382312536239624, + "learning_rate": 7.48204486900955e-06, + "loss": 0.9383, + "step": 11242 + }, + { + "epoch": 0.7725024048371582, + "grad_norm": 1.020923376083374, + "learning_rate": 7.4777673685642425e-06, + "loss": 0.9397, + "step": 11243 + }, + { + "epoch": 0.7725711144702487, + "grad_norm": 0.977903425693512, + "learning_rate": 7.473490876162301e-06, + "loss": 0.8948, + "step": 11244 + }, + { + "epoch": 0.7726398241033393, + "grad_norm": 1.070840835571289, + "learning_rate": 7.469215392049733e-06, + "loss": 0.9369, + "step": 11245 + }, + { + "epoch": 0.7727085337364299, + "grad_norm": 1.1425968408584595, + "learning_rate": 7.464940916472516e-06, + "loss": 0.973, + "step": 11246 + }, + { + "epoch": 0.7727772433695204, + "grad_norm": 1.0985238552093506, + "learning_rate": 7.4606674496765646e-06, + "loss": 0.9227, + "step": 11247 + }, + { + "epoch": 0.7728459530026109, + "grad_norm": 1.0838301181793213, + "learning_rate": 7.4563949919077114e-06, + "loss": 0.9387, + "step": 11248 + }, + { + "epoch": 0.7729146626357015, + "grad_norm": 1.1101782321929932, + "learning_rate": 7.452123543411759e-06, + "loss": 0.9478, + "step": 11249 + }, + { + "epoch": 0.7729833722687921, + "grad_norm": 1.08644437789917, + "learning_rate": 7.447853104434438e-06, + "loss": 0.9238, + "step": 11250 + }, + { + "epoch": 0.7730520819018827, + "grad_norm": 1.0263257026672363, + "learning_rate": 7.4435836752214275e-06, + "loss": 0.9051, + "step": 11251 + }, + { + "epoch": 0.7731207915349732, + "grad_norm": 1.030397891998291, + "learning_rate": 7.4393152560183505e-06, + "loss": 0.9498, + "step": 11252 + }, + { + "epoch": 0.7731895011680637, + "grad_norm": 1.0335458517074585, + "learning_rate": 7.435047847070753e-06, + "loss": 0.9516, + "step": 11253 + }, + { + "epoch": 0.7732582108011543, + "grad_norm": 1.0698161125183105, + "learning_rate": 7.430781448624141e-06, + "loss": 0.9278, + "step": 11254 + }, + { + "epoch": 0.7733269204342449, + "grad_norm": 1.06540846824646, + "learning_rate": 7.426516060923961e-06, + "loss": 0.954, + "step": 11255 + }, + { + "epoch": 0.7733956300673355, + "grad_norm": 1.046854853630066, + "learning_rate": 7.422251684215595e-06, + "loss": 0.9475, + "step": 11256 + }, + { + "epoch": 0.773464339700426, + "grad_norm": 1.0790868997573853, + "learning_rate": 7.417988318744382e-06, + "loss": 0.9465, + "step": 11257 + }, + { + "epoch": 0.7735330493335165, + "grad_norm": 1.0594249963760376, + "learning_rate": 7.413725964755572e-06, + "loss": 0.9106, + "step": 11258 + }, + { + "epoch": 0.7736017589666071, + "grad_norm": 1.0902270078659058, + "learning_rate": 7.409464622494383e-06, + "loss": 0.9724, + "step": 11259 + }, + { + "epoch": 0.7736704685996977, + "grad_norm": 1.1418912410736084, + "learning_rate": 7.405204292205969e-06, + "loss": 0.9764, + "step": 11260 + }, + { + "epoch": 0.7737391782327883, + "grad_norm": 1.1072559356689453, + "learning_rate": 7.400944974135427e-06, + "loss": 0.9655, + "step": 11261 + }, + { + "epoch": 0.7738078878658788, + "grad_norm": 1.0889666080474854, + "learning_rate": 7.396686668527794e-06, + "loss": 0.9706, + "step": 11262 + }, + { + "epoch": 0.7738765974989693, + "grad_norm": 1.0908647775650024, + "learning_rate": 7.392429375628038e-06, + "loss": 0.9424, + "step": 11263 + }, + { + "epoch": 0.7739453071320599, + "grad_norm": 1.081585168838501, + "learning_rate": 7.388173095681083e-06, + "loss": 0.9344, + "step": 11264 + }, + { + "epoch": 0.7740140167651505, + "grad_norm": 1.02164888381958, + "learning_rate": 7.383917828931791e-06, + "loss": 0.953, + "step": 11265 + }, + { + "epoch": 0.774082726398241, + "grad_norm": 0.9739809632301331, + "learning_rate": 7.3796635756249735e-06, + "loss": 0.9343, + "step": 11266 + }, + { + "epoch": 0.7741514360313316, + "grad_norm": 1.0394399166107178, + "learning_rate": 7.375410336005359e-06, + "loss": 0.9302, + "step": 11267 + }, + { + "epoch": 0.7742201456644221, + "grad_norm": 1.0551484823226929, + "learning_rate": 7.371158110317641e-06, + "loss": 0.9539, + "step": 11268 + }, + { + "epoch": 0.7742888552975127, + "grad_norm": 1.0887806415557861, + "learning_rate": 7.366906898806458e-06, + "loss": 0.9301, + "step": 11269 + }, + { + "epoch": 0.7743575649306033, + "grad_norm": 1.086157202720642, + "learning_rate": 7.362656701716361e-06, + "loss": 0.9517, + "step": 11270 + }, + { + "epoch": 0.7744262745636938, + "grad_norm": 1.0666874647140503, + "learning_rate": 7.3584075192918685e-06, + "loss": 0.9537, + "step": 11271 + }, + { + "epoch": 0.7744949841967844, + "grad_norm": 0.9848765134811401, + "learning_rate": 7.354159351777443e-06, + "loss": 0.9054, + "step": 11272 + }, + { + "epoch": 0.774563693829875, + "grad_norm": 1.080238938331604, + "learning_rate": 7.349912199417463e-06, + "loss": 0.9665, + "step": 11273 + }, + { + "epoch": 0.7746324034629655, + "grad_norm": 1.0851774215698242, + "learning_rate": 7.345666062456274e-06, + "loss": 0.9407, + "step": 11274 + }, + { + "epoch": 0.774701113096056, + "grad_norm": 1.080657958984375, + "learning_rate": 7.3414209411381496e-06, + "loss": 0.9551, + "step": 11275 + }, + { + "epoch": 0.7747698227291466, + "grad_norm": 0.9977281093597412, + "learning_rate": 7.337176835707313e-06, + "loss": 0.933, + "step": 11276 + }, + { + "epoch": 0.7748385323622372, + "grad_norm": 1.0251100063323975, + "learning_rate": 7.33293374640793e-06, + "loss": 0.9155, + "step": 11277 + }, + { + "epoch": 0.7749072419953278, + "grad_norm": 1.0365337133407593, + "learning_rate": 7.328691673484089e-06, + "loss": 0.8966, + "step": 11278 + }, + { + "epoch": 0.7749759516284183, + "grad_norm": 1.03394615650177, + "learning_rate": 7.324450617179843e-06, + "loss": 0.9358, + "step": 11279 + }, + { + "epoch": 0.7750446612615088, + "grad_norm": 1.113948941230774, + "learning_rate": 7.320210577739173e-06, + "loss": 0.9435, + "step": 11280 + }, + { + "epoch": 0.7751133708945994, + "grad_norm": 1.0470188856124878, + "learning_rate": 7.315971555406009e-06, + "loss": 0.932, + "step": 11281 + }, + { + "epoch": 0.77518208052769, + "grad_norm": 1.1336698532104492, + "learning_rate": 7.311733550424221e-06, + "loss": 0.9483, + "step": 11282 + }, + { + "epoch": 0.7752507901607806, + "grad_norm": 1.0722298622131348, + "learning_rate": 7.307496563037622e-06, + "loss": 0.9435, + "step": 11283 + }, + { + "epoch": 0.775319499793871, + "grad_norm": 1.0785202980041504, + "learning_rate": 7.303260593489952e-06, + "loss": 0.9367, + "step": 11284 + }, + { + "epoch": 0.7753882094269616, + "grad_norm": 1.0170793533325195, + "learning_rate": 7.299025642024909e-06, + "loss": 0.9205, + "step": 11285 + }, + { + "epoch": 0.7754569190600522, + "grad_norm": 1.0570614337921143, + "learning_rate": 7.294791708886129e-06, + "loss": 0.9221, + "step": 11286 + }, + { + "epoch": 0.7755256286931428, + "grad_norm": 1.0006442070007324, + "learning_rate": 7.2905587943171846e-06, + "loss": 0.883, + "step": 11287 + }, + { + "epoch": 0.7755943383262334, + "grad_norm": 1.10288667678833, + "learning_rate": 7.286326898561605e-06, + "loss": 0.9804, + "step": 11288 + }, + { + "epoch": 0.775663047959324, + "grad_norm": 1.0150667428970337, + "learning_rate": 7.2820960218628276e-06, + "loss": 0.913, + "step": 11289 + }, + { + "epoch": 0.7757317575924144, + "grad_norm": 1.070617437362671, + "learning_rate": 7.277866164464265e-06, + "loss": 0.9486, + "step": 11290 + }, + { + "epoch": 0.775800467225505, + "grad_norm": 1.0711106061935425, + "learning_rate": 7.273637326609262e-06, + "loss": 0.955, + "step": 11291 + }, + { + "epoch": 0.7758691768585956, + "grad_norm": 1.0501461029052734, + "learning_rate": 7.269409508541089e-06, + "loss": 0.9381, + "step": 11292 + }, + { + "epoch": 0.7759378864916862, + "grad_norm": 1.074154019355774, + "learning_rate": 7.265182710502977e-06, + "loss": 0.9531, + "step": 11293 + }, + { + "epoch": 0.7760065961247767, + "grad_norm": 1.060126543045044, + "learning_rate": 7.260956932738094e-06, + "loss": 0.953, + "step": 11294 + }, + { + "epoch": 0.7760753057578672, + "grad_norm": 1.0903040170669556, + "learning_rate": 7.256732175489539e-06, + "loss": 0.9163, + "step": 11295 + }, + { + "epoch": 0.7761440153909578, + "grad_norm": 1.0574991703033447, + "learning_rate": 7.25250843900036e-06, + "loss": 0.9652, + "step": 11296 + }, + { + "epoch": 0.7762127250240484, + "grad_norm": 1.0342036485671997, + "learning_rate": 7.248285723513554e-06, + "loss": 0.9501, + "step": 11297 + }, + { + "epoch": 0.776281434657139, + "grad_norm": 1.0712807178497314, + "learning_rate": 7.244064029272049e-06, + "loss": 0.9251, + "step": 11298 + }, + { + "epoch": 0.7763501442902295, + "grad_norm": 1.0438671112060547, + "learning_rate": 7.239843356518711e-06, + "loss": 0.9654, + "step": 11299 + }, + { + "epoch": 0.77641885392332, + "grad_norm": 1.041881799697876, + "learning_rate": 7.2356237054963526e-06, + "loss": 0.9223, + "step": 11300 + }, + { + "epoch": 0.7764875635564106, + "grad_norm": 1.0915493965148926, + "learning_rate": 7.231405076447731e-06, + "loss": 0.9678, + "step": 11301 + }, + { + "epoch": 0.7765562731895012, + "grad_norm": 1.0638701915740967, + "learning_rate": 7.2271874696155424e-06, + "loss": 0.9342, + "step": 11302 + }, + { + "epoch": 0.7766249828225917, + "grad_norm": 1.0481542348861694, + "learning_rate": 7.222970885242428e-06, + "loss": 0.9299, + "step": 11303 + }, + { + "epoch": 0.7766936924556823, + "grad_norm": 1.0577876567840576, + "learning_rate": 7.218755323570953e-06, + "loss": 0.9427, + "step": 11304 + }, + { + "epoch": 0.7767624020887729, + "grad_norm": 1.0669842958450317, + "learning_rate": 7.214540784843643e-06, + "loss": 0.9374, + "step": 11305 + }, + { + "epoch": 0.7768311117218634, + "grad_norm": 1.1035196781158447, + "learning_rate": 7.210327269302955e-06, + "loss": 0.9396, + "step": 11306 + }, + { + "epoch": 0.776899821354954, + "grad_norm": 1.055740237236023, + "learning_rate": 7.206114777191295e-06, + "loss": 0.9577, + "step": 11307 + }, + { + "epoch": 0.7769685309880445, + "grad_norm": 1.0907450914382935, + "learning_rate": 7.201903308751009e-06, + "loss": 0.9603, + "step": 11308 + }, + { + "epoch": 0.7770372406211351, + "grad_norm": 1.017159342765808, + "learning_rate": 7.197692864224365e-06, + "loss": 0.9237, + "step": 11309 + }, + { + "epoch": 0.7771059502542257, + "grad_norm": 1.04702889919281, + "learning_rate": 7.193483443853597e-06, + "loss": 0.9396, + "step": 11310 + }, + { + "epoch": 0.7771746598873162, + "grad_norm": 1.1086963415145874, + "learning_rate": 7.1892750478808666e-06, + "loss": 0.9411, + "step": 11311 + }, + { + "epoch": 0.7772433695204067, + "grad_norm": 1.0508073568344116, + "learning_rate": 7.185067676548288e-06, + "loss": 0.9383, + "step": 11312 + }, + { + "epoch": 0.7773120791534973, + "grad_norm": 1.0999748706817627, + "learning_rate": 7.180861330097907e-06, + "loss": 0.9214, + "step": 11313 + }, + { + "epoch": 0.7773807887865879, + "grad_norm": 1.0076106786727905, + "learning_rate": 7.176656008771701e-06, + "loss": 0.9125, + "step": 11314 + }, + { + "epoch": 0.7774494984196785, + "grad_norm": 1.0496320724487305, + "learning_rate": 7.172451712811609e-06, + "loss": 0.9715, + "step": 11315 + }, + { + "epoch": 0.7775182080527689, + "grad_norm": 1.0475164651870728, + "learning_rate": 7.168248442459507e-06, + "loss": 0.9927, + "step": 11316 + }, + { + "epoch": 0.7775869176858595, + "grad_norm": 1.0658979415893555, + "learning_rate": 7.164046197957194e-06, + "loss": 0.9114, + "step": 11317 + }, + { + "epoch": 0.7776556273189501, + "grad_norm": 1.0691827535629272, + "learning_rate": 7.159844979546435e-06, + "loss": 0.8936, + "step": 11318 + }, + { + "epoch": 0.7777243369520407, + "grad_norm": 1.0584213733673096, + "learning_rate": 7.155644787468907e-06, + "loss": 0.9326, + "step": 11319 + }, + { + "epoch": 0.7777930465851313, + "grad_norm": 1.026157259941101, + "learning_rate": 7.1514456219662546e-06, + "loss": 0.9131, + "step": 11320 + }, + { + "epoch": 0.7778617562182218, + "grad_norm": 1.072222352027893, + "learning_rate": 7.147247483280053e-06, + "loss": 0.9231, + "step": 11321 + }, + { + "epoch": 0.7779304658513123, + "grad_norm": 1.0340559482574463, + "learning_rate": 7.143050371651819e-06, + "loss": 0.8962, + "step": 11322 + }, + { + "epoch": 0.7779991754844029, + "grad_norm": 0.9959412217140198, + "learning_rate": 7.138854287323015e-06, + "loss": 0.921, + "step": 11323 + }, + { + "epoch": 0.7780678851174935, + "grad_norm": 1.024063229560852, + "learning_rate": 7.134659230535026e-06, + "loss": 0.9088, + "step": 11324 + }, + { + "epoch": 0.778136594750584, + "grad_norm": 1.0831291675567627, + "learning_rate": 7.130465201529199e-06, + "loss": 0.9079, + "step": 11325 + }, + { + "epoch": 0.7782053043836746, + "grad_norm": 1.139046311378479, + "learning_rate": 7.1262722005468155e-06, + "loss": 0.936, + "step": 11326 + }, + { + "epoch": 0.7782740140167651, + "grad_norm": 1.0659739971160889, + "learning_rate": 7.122080227829092e-06, + "loss": 0.9537, + "step": 11327 + }, + { + "epoch": 0.7783427236498557, + "grad_norm": 1.0495383739471436, + "learning_rate": 7.117889283617199e-06, + "loss": 0.9333, + "step": 11328 + }, + { + "epoch": 0.7784114332829463, + "grad_norm": 1.0863866806030273, + "learning_rate": 7.113699368152227e-06, + "loss": 0.9449, + "step": 11329 + }, + { + "epoch": 0.7784801429160368, + "grad_norm": 1.0157911777496338, + "learning_rate": 7.109510481675221e-06, + "loss": 0.9213, + "step": 11330 + }, + { + "epoch": 0.7785488525491274, + "grad_norm": 1.099753499031067, + "learning_rate": 7.105322624427172e-06, + "loss": 0.9586, + "step": 11331 + }, + { + "epoch": 0.7786175621822179, + "grad_norm": 1.140572428703308, + "learning_rate": 7.101135796649e-06, + "loss": 0.9614, + "step": 11332 + }, + { + "epoch": 0.7786862718153085, + "grad_norm": 0.9935078024864197, + "learning_rate": 7.096949998581579e-06, + "loss": 0.939, + "step": 11333 + }, + { + "epoch": 0.778754981448399, + "grad_norm": 1.0967822074890137, + "learning_rate": 7.092765230465701e-06, + "loss": 0.9518, + "step": 11334 + }, + { + "epoch": 0.7788236910814896, + "grad_norm": 1.0662418603897095, + "learning_rate": 7.088581492542121e-06, + "loss": 0.9387, + "step": 11335 + }, + { + "epoch": 0.7788924007145802, + "grad_norm": 1.054711937904358, + "learning_rate": 7.084398785051527e-06, + "loss": 0.9706, + "step": 11336 + }, + { + "epoch": 0.7789611103476708, + "grad_norm": 1.0703057050704956, + "learning_rate": 7.080217108234546e-06, + "loss": 0.9586, + "step": 11337 + }, + { + "epoch": 0.7790298199807613, + "grad_norm": 1.0208739042282104, + "learning_rate": 7.076036462331759e-06, + "loss": 0.9412, + "step": 11338 + }, + { + "epoch": 0.7790985296138518, + "grad_norm": 1.0670357942581177, + "learning_rate": 7.071856847583655e-06, + "loss": 0.8962, + "step": 11339 + }, + { + "epoch": 0.7791672392469424, + "grad_norm": 1.0256359577178955, + "learning_rate": 7.067678264230704e-06, + "loss": 0.9585, + "step": 11340 + }, + { + "epoch": 0.779235948880033, + "grad_norm": 1.061535120010376, + "learning_rate": 7.063500712513279e-06, + "loss": 0.8994, + "step": 11341 + }, + { + "epoch": 0.7793046585131236, + "grad_norm": 0.9738025665283203, + "learning_rate": 7.059324192671723e-06, + "loss": 0.918, + "step": 11342 + }, + { + "epoch": 0.779373368146214, + "grad_norm": 1.0384243726730347, + "learning_rate": 7.05514870494631e-06, + "loss": 0.94, + "step": 11343 + }, + { + "epoch": 0.7794420777793046, + "grad_norm": 1.0652520656585693, + "learning_rate": 7.0509742495772555e-06, + "loss": 0.937, + "step": 11344 + }, + { + "epoch": 0.7795107874123952, + "grad_norm": 1.0844571590423584, + "learning_rate": 7.0468008268047e-06, + "loss": 0.9396, + "step": 11345 + }, + { + "epoch": 0.7795794970454858, + "grad_norm": 1.0181151628494263, + "learning_rate": 7.042628436868748e-06, + "loss": 0.9431, + "step": 11346 + }, + { + "epoch": 0.7796482066785764, + "grad_norm": 1.0547021627426147, + "learning_rate": 7.038457080009431e-06, + "loss": 0.9414, + "step": 11347 + }, + { + "epoch": 0.7797169163116668, + "grad_norm": 1.01651132106781, + "learning_rate": 7.0342867564667265e-06, + "loss": 0.9595, + "step": 11348 + }, + { + "epoch": 0.7797856259447574, + "grad_norm": 1.066749095916748, + "learning_rate": 7.030117466480559e-06, + "loss": 0.9285, + "step": 11349 + }, + { + "epoch": 0.779854335577848, + "grad_norm": 1.1528449058532715, + "learning_rate": 7.025949210290769e-06, + "loss": 0.9142, + "step": 11350 + }, + { + "epoch": 0.7799230452109386, + "grad_norm": 1.0213229656219482, + "learning_rate": 7.021781988137161e-06, + "loss": 0.9336, + "step": 11351 + }, + { + "epoch": 0.7799917548440292, + "grad_norm": 1.0789563655853271, + "learning_rate": 7.017615800259475e-06, + "loss": 0.9164, + "step": 11352 + }, + { + "epoch": 0.7800604644771197, + "grad_norm": 1.0933570861816406, + "learning_rate": 7.0134506468973846e-06, + "loss": 0.9487, + "step": 11353 + }, + { + "epoch": 0.7801291741102102, + "grad_norm": 1.1550296545028687, + "learning_rate": 7.009286528290521e-06, + "loss": 0.9361, + "step": 11354 + }, + { + "epoch": 0.7801978837433008, + "grad_norm": 1.110629916191101, + "learning_rate": 7.005123444678427e-06, + "loss": 0.943, + "step": 11355 + }, + { + "epoch": 0.7802665933763914, + "grad_norm": 1.1205428838729858, + "learning_rate": 7.000961396300609e-06, + "loss": 0.9487, + "step": 11356 + }, + { + "epoch": 0.780335303009482, + "grad_norm": 0.9906435608863831, + "learning_rate": 6.996800383396507e-06, + "loss": 0.9128, + "step": 11357 + }, + { + "epoch": 0.7804040126425725, + "grad_norm": 1.016190767288208, + "learning_rate": 6.9926404062055e-06, + "loss": 0.9036, + "step": 11358 + }, + { + "epoch": 0.780472722275663, + "grad_norm": 1.1759507656097412, + "learning_rate": 6.98848146496692e-06, + "loss": 0.9815, + "step": 11359 + }, + { + "epoch": 0.7805414319087536, + "grad_norm": 1.1056677103042603, + "learning_rate": 6.984323559920011e-06, + "loss": 0.9221, + "step": 11360 + }, + { + "epoch": 0.7806101415418442, + "grad_norm": 1.032957673072815, + "learning_rate": 6.980166691303985e-06, + "loss": 0.9195, + "step": 11361 + }, + { + "epoch": 0.7806788511749347, + "grad_norm": 1.0823115110397339, + "learning_rate": 6.976010859357987e-06, + "loss": 0.9065, + "step": 11362 + }, + { + "epoch": 0.7807475608080253, + "grad_norm": 1.0647059679031372, + "learning_rate": 6.9718560643210865e-06, + "loss": 0.9538, + "step": 11363 + }, + { + "epoch": 0.7808162704411158, + "grad_norm": 1.0224820375442505, + "learning_rate": 6.967702306432316e-06, + "loss": 0.9256, + "step": 11364 + }, + { + "epoch": 0.7808849800742064, + "grad_norm": 1.0825207233428955, + "learning_rate": 6.963549585930643e-06, + "loss": 0.9699, + "step": 11365 + }, + { + "epoch": 0.780953689707297, + "grad_norm": 1.0080149173736572, + "learning_rate": 6.959397903054959e-06, + "loss": 0.9036, + "step": 11366 + }, + { + "epoch": 0.7810223993403875, + "grad_norm": 1.1118303537368774, + "learning_rate": 6.955247258044112e-06, + "loss": 0.9452, + "step": 11367 + }, + { + "epoch": 0.7810911089734781, + "grad_norm": 1.0393720865249634, + "learning_rate": 6.951097651136889e-06, + "loss": 0.935, + "step": 11368 + }, + { + "epoch": 0.7811598186065687, + "grad_norm": 1.0832747220993042, + "learning_rate": 6.946949082572021e-06, + "loss": 0.9518, + "step": 11369 + }, + { + "epoch": 0.7812285282396592, + "grad_norm": 1.037435531616211, + "learning_rate": 6.942801552588157e-06, + "loss": 0.9484, + "step": 11370 + }, + { + "epoch": 0.7812972378727497, + "grad_norm": 1.017539620399475, + "learning_rate": 6.9386550614239075e-06, + "loss": 0.9346, + "step": 11371 + }, + { + "epoch": 0.7813659475058403, + "grad_norm": 1.017238974571228, + "learning_rate": 6.934509609317821e-06, + "loss": 0.9251, + "step": 11372 + }, + { + "epoch": 0.7814346571389309, + "grad_norm": 1.0649861097335815, + "learning_rate": 6.930365196508382e-06, + "loss": 0.9615, + "step": 11373 + }, + { + "epoch": 0.7815033667720215, + "grad_norm": 1.0256270170211792, + "learning_rate": 6.926221823234025e-06, + "loss": 0.9406, + "step": 11374 + }, + { + "epoch": 0.781572076405112, + "grad_norm": 1.0797004699707031, + "learning_rate": 6.922079489733097e-06, + "loss": 0.9766, + "step": 11375 + }, + { + "epoch": 0.7816407860382025, + "grad_norm": 1.0640177726745605, + "learning_rate": 6.917938196243912e-06, + "loss": 0.9029, + "step": 11376 + }, + { + "epoch": 0.7817094956712931, + "grad_norm": 1.1672126054763794, + "learning_rate": 6.9137979430047184e-06, + "loss": 0.9787, + "step": 11377 + }, + { + "epoch": 0.7817782053043837, + "grad_norm": 1.0561076402664185, + "learning_rate": 6.909658730253702e-06, + "loss": 0.8983, + "step": 11378 + }, + { + "epoch": 0.7818469149374743, + "grad_norm": 0.9789278507232666, + "learning_rate": 6.905520558228995e-06, + "loss": 0.873, + "step": 11379 + }, + { + "epoch": 0.7819156245705647, + "grad_norm": 1.0250256061553955, + "learning_rate": 6.901383427168651e-06, + "loss": 0.8914, + "step": 11380 + }, + { + "epoch": 0.7819843342036553, + "grad_norm": 1.0644088983535767, + "learning_rate": 6.8972473373106806e-06, + "loss": 0.9276, + "step": 11381 + }, + { + "epoch": 0.7820530438367459, + "grad_norm": 1.0739426612854004, + "learning_rate": 6.893112288893036e-06, + "loss": 0.9336, + "step": 11382 + }, + { + "epoch": 0.7821217534698365, + "grad_norm": 1.0845881700515747, + "learning_rate": 6.8889782821535995e-06, + "loss": 0.9491, + "step": 11383 + }, + { + "epoch": 0.7821904631029271, + "grad_norm": 1.0287699699401855, + "learning_rate": 6.884845317330208e-06, + "loss": 0.9441, + "step": 11384 + }, + { + "epoch": 0.7822591727360176, + "grad_norm": 1.0260072946548462, + "learning_rate": 6.880713394660612e-06, + "loss": 0.9555, + "step": 11385 + }, + { + "epoch": 0.7823278823691081, + "grad_norm": 1.048412799835205, + "learning_rate": 6.876582514382526e-06, + "loss": 0.9437, + "step": 11386 + }, + { + "epoch": 0.7823965920021987, + "grad_norm": 1.0969111919403076, + "learning_rate": 6.8724526767336065e-06, + "loss": 0.9544, + "step": 11387 + }, + { + "epoch": 0.7824653016352893, + "grad_norm": 1.0312974452972412, + "learning_rate": 6.868323881951422e-06, + "loss": 0.9177, + "step": 11388 + }, + { + "epoch": 0.7825340112683798, + "grad_norm": 1.072742223739624, + "learning_rate": 6.8641961302735175e-06, + "loss": 0.9285, + "step": 11389 + }, + { + "epoch": 0.7826027209014704, + "grad_norm": 1.058100938796997, + "learning_rate": 6.860069421937346e-06, + "loss": 0.9417, + "step": 11390 + }, + { + "epoch": 0.7826714305345609, + "grad_norm": 1.0491666793823242, + "learning_rate": 6.8559437571803165e-06, + "loss": 0.8901, + "step": 11391 + }, + { + "epoch": 0.7827401401676515, + "grad_norm": 1.0847704410552979, + "learning_rate": 6.851819136239782e-06, + "loss": 0.9696, + "step": 11392 + }, + { + "epoch": 0.7828088498007421, + "grad_norm": 1.112045168876648, + "learning_rate": 6.847695559353027e-06, + "loss": 0.9103, + "step": 11393 + }, + { + "epoch": 0.7828775594338326, + "grad_norm": 1.0463449954986572, + "learning_rate": 6.843573026757277e-06, + "loss": 0.9262, + "step": 11394 + }, + { + "epoch": 0.7829462690669232, + "grad_norm": 1.0854395627975464, + "learning_rate": 6.839451538689709e-06, + "loss": 0.9585, + "step": 11395 + }, + { + "epoch": 0.7830149787000138, + "grad_norm": 1.031728744506836, + "learning_rate": 6.835331095387412e-06, + "loss": 0.9087, + "step": 11396 + }, + { + "epoch": 0.7830836883331043, + "grad_norm": 1.0257502794265747, + "learning_rate": 6.831211697087445e-06, + "loss": 0.9039, + "step": 11397 + }, + { + "epoch": 0.7831523979661948, + "grad_norm": 1.0397664308547974, + "learning_rate": 6.8270933440267855e-06, + "loss": 0.9074, + "step": 11398 + }, + { + "epoch": 0.7832211075992854, + "grad_norm": 1.117340326309204, + "learning_rate": 6.822976036442369e-06, + "loss": 0.9443, + "step": 11399 + }, + { + "epoch": 0.783289817232376, + "grad_norm": 1.0605794191360474, + "learning_rate": 6.818859774571065e-06, + "loss": 0.9301, + "step": 11400 + }, + { + "epoch": 0.7833585268654666, + "grad_norm": 1.0152332782745361, + "learning_rate": 6.814744558649666e-06, + "loss": 0.9429, + "step": 11401 + }, + { + "epoch": 0.7834272364985571, + "grad_norm": 1.0303245782852173, + "learning_rate": 6.810630388914924e-06, + "loss": 0.9332, + "step": 11402 + }, + { + "epoch": 0.7834959461316476, + "grad_norm": 1.025172472000122, + "learning_rate": 6.806517265603524e-06, + "loss": 0.898, + "step": 11403 + }, + { + "epoch": 0.7835646557647382, + "grad_norm": 1.1009929180145264, + "learning_rate": 6.8024051889520905e-06, + "loss": 0.9317, + "step": 11404 + }, + { + "epoch": 0.7836333653978288, + "grad_norm": 1.0137901306152344, + "learning_rate": 6.798294159197202e-06, + "loss": 0.9116, + "step": 11405 + }, + { + "epoch": 0.7837020750309194, + "grad_norm": 1.0406399965286255, + "learning_rate": 6.794184176575344e-06, + "loss": 0.97, + "step": 11406 + }, + { + "epoch": 0.7837707846640098, + "grad_norm": 1.145795464515686, + "learning_rate": 6.7900752413229685e-06, + "loss": 0.9486, + "step": 11407 + }, + { + "epoch": 0.7838394942971004, + "grad_norm": 1.1460551023483276, + "learning_rate": 6.785967353676462e-06, + "loss": 0.9451, + "step": 11408 + }, + { + "epoch": 0.783908203930191, + "grad_norm": 1.0716302394866943, + "learning_rate": 6.781860513872154e-06, + "loss": 0.9418, + "step": 11409 + }, + { + "epoch": 0.7839769135632816, + "grad_norm": 1.0436677932739258, + "learning_rate": 6.777754722146296e-06, + "loss": 0.9179, + "step": 11410 + }, + { + "epoch": 0.7840456231963722, + "grad_norm": 1.0211008787155151, + "learning_rate": 6.773649978735106e-06, + "loss": 0.921, + "step": 11411 + }, + { + "epoch": 0.7841143328294627, + "grad_norm": 1.044496774673462, + "learning_rate": 6.769546283874712e-06, + "loss": 0.9144, + "step": 11412 + }, + { + "epoch": 0.7841830424625532, + "grad_norm": 1.0637022256851196, + "learning_rate": 6.765443637801205e-06, + "loss": 0.9223, + "step": 11413 + }, + { + "epoch": 0.7842517520956438, + "grad_norm": 1.0653178691864014, + "learning_rate": 6.761342040750607e-06, + "loss": 0.9124, + "step": 11414 + }, + { + "epoch": 0.7843204617287344, + "grad_norm": 1.0701606273651123, + "learning_rate": 6.7572414929588875e-06, + "loss": 0.9669, + "step": 11415 + }, + { + "epoch": 0.784389171361825, + "grad_norm": 1.0728638172149658, + "learning_rate": 6.753141994661935e-06, + "loss": 0.9125, + "step": 11416 + }, + { + "epoch": 0.7844578809949155, + "grad_norm": 1.0245819091796875, + "learning_rate": 6.749043546095596e-06, + "loss": 0.9508, + "step": 11417 + }, + { + "epoch": 0.784526590628006, + "grad_norm": 1.0564842224121094, + "learning_rate": 6.744946147495657e-06, + "loss": 0.9394, + "step": 11418 + }, + { + "epoch": 0.7845953002610966, + "grad_norm": 1.0359463691711426, + "learning_rate": 6.740849799097831e-06, + "loss": 0.93, + "step": 11419 + }, + { + "epoch": 0.7846640098941872, + "grad_norm": 1.0343736410140991, + "learning_rate": 6.736754501137793e-06, + "loss": 0.9341, + "step": 11420 + }, + { + "epoch": 0.7847327195272777, + "grad_norm": 1.1047230958938599, + "learning_rate": 6.7326602538511254e-06, + "loss": 0.9772, + "step": 11421 + }, + { + "epoch": 0.7848014291603683, + "grad_norm": 1.0418800115585327, + "learning_rate": 6.728567057473376e-06, + "loss": 0.9148, + "step": 11422 + }, + { + "epoch": 0.7848701387934588, + "grad_norm": 1.1013453006744385, + "learning_rate": 6.72447491224002e-06, + "loss": 0.9674, + "step": 11423 + }, + { + "epoch": 0.7849388484265494, + "grad_norm": 1.1103286743164062, + "learning_rate": 6.720383818386483e-06, + "loss": 0.9368, + "step": 11424 + }, + { + "epoch": 0.78500755805964, + "grad_norm": 1.0025559663772583, + "learning_rate": 6.716293776148125e-06, + "loss": 0.9146, + "step": 11425 + }, + { + "epoch": 0.7850762676927305, + "grad_norm": 1.0309451818466187, + "learning_rate": 6.712204785760229e-06, + "loss": 0.9388, + "step": 11426 + }, + { + "epoch": 0.7851449773258211, + "grad_norm": 1.024834156036377, + "learning_rate": 6.708116847458043e-06, + "loss": 0.9463, + "step": 11427 + }, + { + "epoch": 0.7852136869589117, + "grad_norm": 1.0325305461883545, + "learning_rate": 6.70402996147674e-06, + "loss": 0.9295, + "step": 11428 + }, + { + "epoch": 0.7852823965920022, + "grad_norm": 1.0375984907150269, + "learning_rate": 6.699944128051439e-06, + "loss": 0.9071, + "step": 11429 + }, + { + "epoch": 0.7853511062250927, + "grad_norm": 0.9940739274024963, + "learning_rate": 6.695859347417199e-06, + "loss": 0.9158, + "step": 11430 + }, + { + "epoch": 0.7854198158581833, + "grad_norm": 0.9828243255615234, + "learning_rate": 6.691775619809007e-06, + "loss": 0.898, + "step": 11431 + }, + { + "epoch": 0.7854885254912739, + "grad_norm": 1.0405348539352417, + "learning_rate": 6.687692945461799e-06, + "loss": 0.9425, + "step": 11432 + }, + { + "epoch": 0.7855572351243645, + "grad_norm": 1.043686866760254, + "learning_rate": 6.683611324610447e-06, + "loss": 0.9336, + "step": 11433 + }, + { + "epoch": 0.785625944757455, + "grad_norm": 1.160144329071045, + "learning_rate": 6.679530757489777e-06, + "loss": 0.9789, + "step": 11434 + }, + { + "epoch": 0.7856946543905455, + "grad_norm": 1.0270087718963623, + "learning_rate": 6.675451244334527e-06, + "loss": 0.9497, + "step": 11435 + }, + { + "epoch": 0.7857633640236361, + "grad_norm": 1.0687663555145264, + "learning_rate": 6.671372785379398e-06, + "loss": 0.8985, + "step": 11436 + }, + { + "epoch": 0.7858320736567267, + "grad_norm": 1.0227136611938477, + "learning_rate": 6.6672953808590094e-06, + "loss": 0.9261, + "step": 11437 + }, + { + "epoch": 0.7859007832898173, + "grad_norm": 1.0541880130767822, + "learning_rate": 6.663219031007942e-06, + "loss": 0.9586, + "step": 11438 + }, + { + "epoch": 0.7859694929229077, + "grad_norm": 1.1114091873168945, + "learning_rate": 6.659143736060702e-06, + "loss": 0.9593, + "step": 11439 + }, + { + "epoch": 0.7860382025559983, + "grad_norm": 1.0239410400390625, + "learning_rate": 6.655069496251748e-06, + "loss": 0.8975, + "step": 11440 + }, + { + "epoch": 0.7861069121890889, + "grad_norm": 1.1059589385986328, + "learning_rate": 6.650996311815452e-06, + "loss": 0.9141, + "step": 11441 + }, + { + "epoch": 0.7861756218221795, + "grad_norm": 1.0577946901321411, + "learning_rate": 6.64692418298615e-06, + "loss": 0.9179, + "step": 11442 + }, + { + "epoch": 0.7862443314552701, + "grad_norm": 1.0289427042007446, + "learning_rate": 6.642853109998109e-06, + "loss": 0.9098, + "step": 11443 + }, + { + "epoch": 0.7863130410883606, + "grad_norm": 1.0496045351028442, + "learning_rate": 6.638783093085535e-06, + "loss": 0.9398, + "step": 11444 + }, + { + "epoch": 0.7863817507214511, + "grad_norm": 1.0249372720718384, + "learning_rate": 6.634714132482575e-06, + "loss": 0.9405, + "step": 11445 + }, + { + "epoch": 0.7864504603545417, + "grad_norm": 1.041040301322937, + "learning_rate": 6.6306462284233234e-06, + "loss": 0.952, + "step": 11446 + }, + { + "epoch": 0.7865191699876323, + "grad_norm": 1.0506705045700073, + "learning_rate": 6.626579381141781e-06, + "loss": 0.9296, + "step": 11447 + }, + { + "epoch": 0.7865878796207229, + "grad_norm": 1.097313642501831, + "learning_rate": 6.622513590871929e-06, + "loss": 0.9418, + "step": 11448 + }, + { + "epoch": 0.7866565892538134, + "grad_norm": 1.0449053049087524, + "learning_rate": 6.618448857847662e-06, + "loss": 0.9545, + "step": 11449 + }, + { + "epoch": 0.7867252988869039, + "grad_norm": 1.0704703330993652, + "learning_rate": 6.6143851823028265e-06, + "loss": 0.9395, + "step": 11450 + }, + { + "epoch": 0.7867940085199945, + "grad_norm": 1.0661752223968506, + "learning_rate": 6.6103225644712054e-06, + "loss": 0.9532, + "step": 11451 + }, + { + "epoch": 0.7868627181530851, + "grad_norm": 1.0136748552322388, + "learning_rate": 6.606261004586509e-06, + "loss": 0.9255, + "step": 11452 + }, + { + "epoch": 0.7869314277861756, + "grad_norm": 1.0685080289840698, + "learning_rate": 6.602200502882402e-06, + "loss": 0.8859, + "step": 11453 + }, + { + "epoch": 0.7870001374192662, + "grad_norm": 1.0589470863342285, + "learning_rate": 6.5981410595924804e-06, + "loss": 0.9152, + "step": 11454 + }, + { + "epoch": 0.7870688470523567, + "grad_norm": 1.0096189975738525, + "learning_rate": 6.594082674950286e-06, + "loss": 0.9179, + "step": 11455 + }, + { + "epoch": 0.7871375566854473, + "grad_norm": 1.1115105152130127, + "learning_rate": 6.5900253491892966e-06, + "loss": 0.973, + "step": 11456 + }, + { + "epoch": 0.7872062663185379, + "grad_norm": 1.106547474861145, + "learning_rate": 6.58596908254292e-06, + "loss": 0.9528, + "step": 11457 + }, + { + "epoch": 0.7872749759516284, + "grad_norm": 1.0236320495605469, + "learning_rate": 6.581913875244519e-06, + "loss": 0.9334, + "step": 11458 + }, + { + "epoch": 0.787343685584719, + "grad_norm": 1.0746588706970215, + "learning_rate": 6.577859727527377e-06, + "loss": 0.9282, + "step": 11459 + }, + { + "epoch": 0.7874123952178096, + "grad_norm": 1.090335488319397, + "learning_rate": 6.573806639624733e-06, + "loss": 0.9736, + "step": 11460 + }, + { + "epoch": 0.7874811048509001, + "grad_norm": 1.055191159248352, + "learning_rate": 6.569754611769763e-06, + "loss": 0.9593, + "step": 11461 + }, + { + "epoch": 0.7875498144839906, + "grad_norm": 1.130560278892517, + "learning_rate": 6.565703644195567e-06, + "loss": 0.9481, + "step": 11462 + }, + { + "epoch": 0.7876185241170812, + "grad_norm": 1.0842150449752808, + "learning_rate": 6.561653737135198e-06, + "loss": 0.9475, + "step": 11463 + }, + { + "epoch": 0.7876872337501718, + "grad_norm": 1.0333510637283325, + "learning_rate": 6.55760489082165e-06, + "loss": 0.9155, + "step": 11464 + }, + { + "epoch": 0.7877559433832624, + "grad_norm": 1.048877477645874, + "learning_rate": 6.553557105487845e-06, + "loss": 0.9491, + "step": 11465 + }, + { + "epoch": 0.7878246530163528, + "grad_norm": 1.0618062019348145, + "learning_rate": 6.549510381366664e-06, + "loss": 0.8989, + "step": 11466 + }, + { + "epoch": 0.7878933626494434, + "grad_norm": 1.0276756286621094, + "learning_rate": 6.545464718690889e-06, + "loss": 0.9472, + "step": 11467 + }, + { + "epoch": 0.787962072282534, + "grad_norm": 1.0742301940917969, + "learning_rate": 6.5414201176932784e-06, + "loss": 0.9362, + "step": 11468 + }, + { + "epoch": 0.7880307819156246, + "grad_norm": 1.0416162014007568, + "learning_rate": 6.5373765786065135e-06, + "loss": 0.9066, + "step": 11469 + }, + { + "epoch": 0.7880994915487152, + "grad_norm": 1.0488742589950562, + "learning_rate": 6.533334101663216e-06, + "loss": 0.9065, + "step": 11470 + }, + { + "epoch": 0.7881682011818056, + "grad_norm": 1.0417462587356567, + "learning_rate": 6.529292687095956e-06, + "loss": 0.9424, + "step": 11471 + }, + { + "epoch": 0.7882369108148962, + "grad_norm": 1.0203732252120972, + "learning_rate": 6.525252335137219e-06, + "loss": 0.9064, + "step": 11472 + }, + { + "epoch": 0.7883056204479868, + "grad_norm": 1.0186842679977417, + "learning_rate": 6.5212130460194475e-06, + "loss": 0.91, + "step": 11473 + }, + { + "epoch": 0.7883743300810774, + "grad_norm": 1.0631422996520996, + "learning_rate": 6.517174819975025e-06, + "loss": 0.9156, + "step": 11474 + }, + { + "epoch": 0.788443039714168, + "grad_norm": 1.0493265390396118, + "learning_rate": 6.5131376572362655e-06, + "loss": 0.8809, + "step": 11475 + }, + { + "epoch": 0.7885117493472585, + "grad_norm": 1.0546350479125977, + "learning_rate": 6.50910155803543e-06, + "loss": 0.9611, + "step": 11476 + }, + { + "epoch": 0.788580458980349, + "grad_norm": 1.0617057085037231, + "learning_rate": 6.505066522604703e-06, + "loss": 0.9106, + "step": 11477 + }, + { + "epoch": 0.7886491686134396, + "grad_norm": 1.0478339195251465, + "learning_rate": 6.50103255117622e-06, + "loss": 0.8918, + "step": 11478 + }, + { + "epoch": 0.7887178782465302, + "grad_norm": 1.0337799787521362, + "learning_rate": 6.4969996439820545e-06, + "loss": 0.9379, + "step": 11479 + }, + { + "epoch": 0.7887865878796207, + "grad_norm": 1.069728136062622, + "learning_rate": 6.492967801254224e-06, + "loss": 0.9395, + "step": 11480 + }, + { + "epoch": 0.7888552975127113, + "grad_norm": 1.119390845298767, + "learning_rate": 6.488937023224667e-06, + "loss": 0.9599, + "step": 11481 + }, + { + "epoch": 0.7889240071458018, + "grad_norm": 1.0740989446640015, + "learning_rate": 6.484907310125274e-06, + "loss": 0.9365, + "step": 11482 + }, + { + "epoch": 0.7889927167788924, + "grad_norm": 1.1008912324905396, + "learning_rate": 6.480878662187884e-06, + "loss": 0.9696, + "step": 11483 + }, + { + "epoch": 0.789061426411983, + "grad_norm": 1.0535311698913574, + "learning_rate": 6.476851079644241e-06, + "loss": 0.8966, + "step": 11484 + }, + { + "epoch": 0.7891301360450735, + "grad_norm": 1.0722283124923706, + "learning_rate": 6.472824562726066e-06, + "loss": 0.9066, + "step": 11485 + }, + { + "epoch": 0.7891988456781641, + "grad_norm": 1.1679106950759888, + "learning_rate": 6.468799111665003e-06, + "loss": 0.9091, + "step": 11486 + }, + { + "epoch": 0.7892675553112546, + "grad_norm": 1.1037468910217285, + "learning_rate": 6.464774726692621e-06, + "loss": 0.9274, + "step": 11487 + }, + { + "epoch": 0.7893362649443452, + "grad_norm": 1.0774158239364624, + "learning_rate": 6.460751408040447e-06, + "loss": 0.9304, + "step": 11488 + }, + { + "epoch": 0.7894049745774357, + "grad_norm": 1.0136849880218506, + "learning_rate": 6.456729155939939e-06, + "loss": 0.8812, + "step": 11489 + }, + { + "epoch": 0.7894736842105263, + "grad_norm": 1.066968560218811, + "learning_rate": 6.452707970622499e-06, + "loss": 0.9341, + "step": 11490 + }, + { + "epoch": 0.7895423938436169, + "grad_norm": 1.0568910837173462, + "learning_rate": 6.4486878523194654e-06, + "loss": 0.9106, + "step": 11491 + }, + { + "epoch": 0.7896111034767075, + "grad_norm": 1.0115140676498413, + "learning_rate": 6.444668801262102e-06, + "loss": 0.9177, + "step": 11492 + }, + { + "epoch": 0.789679813109798, + "grad_norm": 1.1011557579040527, + "learning_rate": 6.440650817681629e-06, + "loss": 0.9374, + "step": 11493 + }, + { + "epoch": 0.7897485227428885, + "grad_norm": 1.0322316884994507, + "learning_rate": 6.436633901809197e-06, + "loss": 0.9397, + "step": 11494 + }, + { + "epoch": 0.7898172323759791, + "grad_norm": 1.0081034898757935, + "learning_rate": 6.432618053875899e-06, + "loss": 0.9277, + "step": 11495 + }, + { + "epoch": 0.7898859420090697, + "grad_norm": 1.0438402891159058, + "learning_rate": 6.428603274112771e-06, + "loss": 0.9501, + "step": 11496 + }, + { + "epoch": 0.7899546516421603, + "grad_norm": 1.08469820022583, + "learning_rate": 6.4245895627507654e-06, + "loss": 0.9218, + "step": 11497 + }, + { + "epoch": 0.7900233612752507, + "grad_norm": 1.0833675861358643, + "learning_rate": 6.420576920020796e-06, + "loss": 0.9704, + "step": 11498 + }, + { + "epoch": 0.7900920709083413, + "grad_norm": 1.0759292840957642, + "learning_rate": 6.416565346153708e-06, + "loss": 0.9047, + "step": 11499 + }, + { + "epoch": 0.7901607805414319, + "grad_norm": 1.0150835514068604, + "learning_rate": 6.412554841380283e-06, + "loss": 0.9299, + "step": 11500 + }, + { + "epoch": 0.7902294901745225, + "grad_norm": 1.1053199768066406, + "learning_rate": 6.408545405931246e-06, + "loss": 0.9364, + "step": 11501 + }, + { + "epoch": 0.7902981998076131, + "grad_norm": 0.9805290102958679, + "learning_rate": 6.404537040037264e-06, + "loss": 0.8988, + "step": 11502 + }, + { + "epoch": 0.7903669094407035, + "grad_norm": 1.0365692377090454, + "learning_rate": 6.400529743928921e-06, + "loss": 0.9678, + "step": 11503 + }, + { + "epoch": 0.7904356190737941, + "grad_norm": 1.0432629585266113, + "learning_rate": 6.396523517836761e-06, + "loss": 0.9454, + "step": 11504 + }, + { + "epoch": 0.7905043287068847, + "grad_norm": 1.0106346607208252, + "learning_rate": 6.392518361991265e-06, + "loss": 0.9054, + "step": 11505 + }, + { + "epoch": 0.7905730383399753, + "grad_norm": 1.0170528888702393, + "learning_rate": 6.3885142766228365e-06, + "loss": 0.9283, + "step": 11506 + }, + { + "epoch": 0.7906417479730659, + "grad_norm": 1.0416487455368042, + "learning_rate": 6.3845112619618395e-06, + "loss": 0.9115, + "step": 11507 + }, + { + "epoch": 0.7907104576061564, + "grad_norm": 1.0502170324325562, + "learning_rate": 6.3805093182385536e-06, + "loss": 0.9266, + "step": 11508 + }, + { + "epoch": 0.7907791672392469, + "grad_norm": 1.0173760652542114, + "learning_rate": 6.376508445683213e-06, + "loss": 0.9087, + "step": 11509 + }, + { + "epoch": 0.7908478768723375, + "grad_norm": 1.0345549583435059, + "learning_rate": 6.372508644525985e-06, + "loss": 0.9371, + "step": 11510 + }, + { + "epoch": 0.7909165865054281, + "grad_norm": 1.0501104593276978, + "learning_rate": 6.368509914996978e-06, + "loss": 0.9257, + "step": 11511 + }, + { + "epoch": 0.7909852961385186, + "grad_norm": 1.0804096460342407, + "learning_rate": 6.364512257326244e-06, + "loss": 0.9427, + "step": 11512 + }, + { + "epoch": 0.7910540057716092, + "grad_norm": 1.0732718706130981, + "learning_rate": 6.360515671743747e-06, + "loss": 0.9433, + "step": 11513 + }, + { + "epoch": 0.7911227154046997, + "grad_norm": 1.0796312093734741, + "learning_rate": 6.35652015847942e-06, + "loss": 0.9537, + "step": 11514 + }, + { + "epoch": 0.7911914250377903, + "grad_norm": 1.0960420370101929, + "learning_rate": 6.352525717763119e-06, + "loss": 0.9316, + "step": 11515 + }, + { + "epoch": 0.7912601346708809, + "grad_norm": 1.0264256000518799, + "learning_rate": 6.348532349824646e-06, + "loss": 0.9238, + "step": 11516 + }, + { + "epoch": 0.7913288443039714, + "grad_norm": 1.0287814140319824, + "learning_rate": 6.3445400548937415e-06, + "loss": 0.9278, + "step": 11517 + }, + { + "epoch": 0.791397553937062, + "grad_norm": 1.0529388189315796, + "learning_rate": 6.340548833200063e-06, + "loss": 0.9188, + "step": 11518 + }, + { + "epoch": 0.7914662635701525, + "grad_norm": 1.0223127603530884, + "learning_rate": 6.336558684973234e-06, + "loss": 0.9372, + "step": 11519 + }, + { + "epoch": 0.7915349732032431, + "grad_norm": 1.0724483728408813, + "learning_rate": 6.332569610442807e-06, + "loss": 0.939, + "step": 11520 + }, + { + "epoch": 0.7916036828363336, + "grad_norm": 1.0429496765136719, + "learning_rate": 6.328581609838266e-06, + "loss": 0.9239, + "step": 11521 + }, + { + "epoch": 0.7916723924694242, + "grad_norm": 1.050032615661621, + "learning_rate": 6.3245946833890455e-06, + "loss": 0.9744, + "step": 11522 + }, + { + "epoch": 0.7917411021025148, + "grad_norm": 1.0253535509109497, + "learning_rate": 6.320608831324501e-06, + "loss": 0.9448, + "step": 11523 + }, + { + "epoch": 0.7918098117356054, + "grad_norm": 1.0146050453186035, + "learning_rate": 6.31662405387394e-06, + "loss": 0.8804, + "step": 11524 + }, + { + "epoch": 0.7918785213686959, + "grad_norm": 1.0273096561431885, + "learning_rate": 6.312640351266608e-06, + "loss": 0.8954, + "step": 11525 + }, + { + "epoch": 0.7919472310017864, + "grad_norm": 1.043968677520752, + "learning_rate": 6.308657723731678e-06, + "loss": 0.9237, + "step": 11526 + }, + { + "epoch": 0.792015940634877, + "grad_norm": 1.028630018234253, + "learning_rate": 6.304676171498283e-06, + "loss": 0.9126, + "step": 11527 + }, + { + "epoch": 0.7920846502679676, + "grad_norm": 1.0769487619400024, + "learning_rate": 6.300695694795461e-06, + "loss": 0.9774, + "step": 11528 + }, + { + "epoch": 0.7921533599010582, + "grad_norm": 1.0537245273590088, + "learning_rate": 6.29671629385222e-06, + "loss": 0.915, + "step": 11529 + }, + { + "epoch": 0.7922220695341486, + "grad_norm": 1.072756052017212, + "learning_rate": 6.29273796889748e-06, + "loss": 0.9382, + "step": 11530 + }, + { + "epoch": 0.7922907791672392, + "grad_norm": 1.0974918603897095, + "learning_rate": 6.288760720160117e-06, + "loss": 0.9636, + "step": 11531 + }, + { + "epoch": 0.7923594888003298, + "grad_norm": 1.08718740940094, + "learning_rate": 6.284784547868949e-06, + "loss": 0.919, + "step": 11532 + }, + { + "epoch": 0.7924281984334204, + "grad_norm": 1.1348490715026855, + "learning_rate": 6.280809452252709e-06, + "loss": 0.9304, + "step": 11533 + }, + { + "epoch": 0.792496908066511, + "grad_norm": 1.076356053352356, + "learning_rate": 6.2768354335400876e-06, + "loss": 0.9068, + "step": 11534 + }, + { + "epoch": 0.7925656176996014, + "grad_norm": 1.0743906497955322, + "learning_rate": 6.272862491959705e-06, + "loss": 0.9191, + "step": 11535 + }, + { + "epoch": 0.792634327332692, + "grad_norm": 1.056504249572754, + "learning_rate": 6.268890627740126e-06, + "loss": 0.9436, + "step": 11536 + }, + { + "epoch": 0.7927030369657826, + "grad_norm": 1.018349289894104, + "learning_rate": 6.264919841109856e-06, + "loss": 0.916, + "step": 11537 + }, + { + "epoch": 0.7927717465988732, + "grad_norm": 1.0646520853042603, + "learning_rate": 6.260950132297316e-06, + "loss": 0.9298, + "step": 11538 + }, + { + "epoch": 0.7928404562319638, + "grad_norm": 1.0136034488677979, + "learning_rate": 6.25698150153089e-06, + "loss": 0.8943, + "step": 11539 + }, + { + "epoch": 0.7929091658650543, + "grad_norm": 1.0792763233184814, + "learning_rate": 6.253013949038889e-06, + "loss": 0.9431, + "step": 11540 + }, + { + "epoch": 0.7929778754981448, + "grad_norm": 1.1137851476669312, + "learning_rate": 6.2490474750495635e-06, + "loss": 0.9453, + "step": 11541 + }, + { + "epoch": 0.7930465851312354, + "grad_norm": 1.054161548614502, + "learning_rate": 6.245082079791115e-06, + "loss": 0.9152, + "step": 11542 + }, + { + "epoch": 0.793115294764326, + "grad_norm": 1.0663604736328125, + "learning_rate": 6.241117763491647e-06, + "loss": 0.9061, + "step": 11543 + }, + { + "epoch": 0.7931840043974165, + "grad_norm": 1.0978482961654663, + "learning_rate": 6.2371545263792384e-06, + "loss": 0.8996, + "step": 11544 + }, + { + "epoch": 0.7932527140305071, + "grad_norm": 1.0110918283462524, + "learning_rate": 6.23319236868189e-06, + "loss": 0.9067, + "step": 11545 + }, + { + "epoch": 0.7933214236635976, + "grad_norm": 1.0410999059677124, + "learning_rate": 6.229231290627541e-06, + "loss": 0.9123, + "step": 11546 + }, + { + "epoch": 0.7933901332966882, + "grad_norm": 1.0415472984313965, + "learning_rate": 6.225271292444079e-06, + "loss": 0.9018, + "step": 11547 + }, + { + "epoch": 0.7934588429297788, + "grad_norm": 1.0353918075561523, + "learning_rate": 6.221312374359306e-06, + "loss": 0.9212, + "step": 11548 + }, + { + "epoch": 0.7935275525628693, + "grad_norm": 1.051829218864441, + "learning_rate": 6.217354536600978e-06, + "loss": 0.9406, + "step": 11549 + }, + { + "epoch": 0.7935962621959599, + "grad_norm": 1.0481865406036377, + "learning_rate": 6.213397779396795e-06, + "loss": 0.9263, + "step": 11550 + }, + { + "epoch": 0.7936649718290504, + "grad_norm": 1.0549087524414062, + "learning_rate": 6.209442102974388e-06, + "loss": 0.9306, + "step": 11551 + }, + { + "epoch": 0.793733681462141, + "grad_norm": 1.0449284315109253, + "learning_rate": 6.205487507561314e-06, + "loss": 0.8868, + "step": 11552 + }, + { + "epoch": 0.7938023910952315, + "grad_norm": 1.0649715662002563, + "learning_rate": 6.2015339933850805e-06, + "loss": 0.9072, + "step": 11553 + }, + { + "epoch": 0.7938711007283221, + "grad_norm": 1.0927804708480835, + "learning_rate": 6.197581560673146e-06, + "loss": 0.9491, + "step": 11554 + }, + { + "epoch": 0.7939398103614127, + "grad_norm": 1.0938606262207031, + "learning_rate": 6.19363020965287e-06, + "loss": 0.9607, + "step": 11555 + }, + { + "epoch": 0.7940085199945033, + "grad_norm": 1.1191046237945557, + "learning_rate": 6.189679940551582e-06, + "loss": 0.9504, + "step": 11556 + }, + { + "epoch": 0.7940772296275938, + "grad_norm": 1.1288071870803833, + "learning_rate": 6.185730753596539e-06, + "loss": 0.9111, + "step": 11557 + }, + { + "epoch": 0.7941459392606843, + "grad_norm": 1.0501677989959717, + "learning_rate": 6.181782649014939e-06, + "loss": 0.9634, + "step": 11558 + }, + { + "epoch": 0.7942146488937749, + "grad_norm": 1.0321974754333496, + "learning_rate": 6.1778356270339035e-06, + "loss": 0.9072, + "step": 11559 + }, + { + "epoch": 0.7942833585268655, + "grad_norm": 1.0684430599212646, + "learning_rate": 6.1738896878805075e-06, + "loss": 0.9364, + "step": 11560 + }, + { + "epoch": 0.7943520681599561, + "grad_norm": 1.0652915239334106, + "learning_rate": 6.169944831781757e-06, + "loss": 0.9228, + "step": 11561 + }, + { + "epoch": 0.7944207777930465, + "grad_norm": 0.9925357103347778, + "learning_rate": 6.166001058964601e-06, + "loss": 0.9237, + "step": 11562 + }, + { + "epoch": 0.7944894874261371, + "grad_norm": 1.0616137981414795, + "learning_rate": 6.162058369655926e-06, + "loss": 0.9412, + "step": 11563 + }, + { + "epoch": 0.7945581970592277, + "grad_norm": 1.0343931913375854, + "learning_rate": 6.1581167640825415e-06, + "loss": 0.9067, + "step": 11564 + }, + { + "epoch": 0.7946269066923183, + "grad_norm": 1.0054335594177246, + "learning_rate": 6.15417624247121e-06, + "loss": 0.9025, + "step": 11565 + }, + { + "epoch": 0.7946956163254089, + "grad_norm": 1.1415972709655762, + "learning_rate": 6.150236805048629e-06, + "loss": 0.9176, + "step": 11566 + }, + { + "epoch": 0.7947643259584993, + "grad_norm": 1.0536071062088013, + "learning_rate": 6.1462984520414324e-06, + "loss": 0.9138, + "step": 11567 + }, + { + "epoch": 0.7948330355915899, + "grad_norm": 1.0072383880615234, + "learning_rate": 6.142361183676196e-06, + "loss": 0.919, + "step": 11568 + }, + { + "epoch": 0.7949017452246805, + "grad_norm": 1.0520766973495483, + "learning_rate": 6.1384250001794156e-06, + "loss": 0.9411, + "step": 11569 + }, + { + "epoch": 0.7949704548577711, + "grad_norm": 1.1488786935806274, + "learning_rate": 6.134489901777543e-06, + "loss": 0.9798, + "step": 11570 + }, + { + "epoch": 0.7950391644908616, + "grad_norm": 1.1086355447769165, + "learning_rate": 6.130555888696965e-06, + "loss": 0.937, + "step": 11571 + }, + { + "epoch": 0.7951078741239522, + "grad_norm": 1.0519154071807861, + "learning_rate": 6.126622961164e-06, + "loss": 0.9565, + "step": 11572 + }, + { + "epoch": 0.7951765837570427, + "grad_norm": 1.0474170446395874, + "learning_rate": 6.122691119404916e-06, + "loss": 0.9677, + "step": 11573 + }, + { + "epoch": 0.7952452933901333, + "grad_norm": 1.047635793685913, + "learning_rate": 6.118760363645895e-06, + "loss": 0.9539, + "step": 11574 + }, + { + "epoch": 0.7953140030232239, + "grad_norm": 1.0831470489501953, + "learning_rate": 6.114830694113077e-06, + "loss": 0.9317, + "step": 11575 + }, + { + "epoch": 0.7953827126563144, + "grad_norm": 1.1103484630584717, + "learning_rate": 6.110902111032543e-06, + "loss": 0.9459, + "step": 11576 + }, + { + "epoch": 0.795451422289405, + "grad_norm": 1.0671085119247437, + "learning_rate": 6.106974614630285e-06, + "loss": 0.9423, + "step": 11577 + }, + { + "epoch": 0.7955201319224955, + "grad_norm": 1.0709739923477173, + "learning_rate": 6.103048205132258e-06, + "loss": 0.9492, + "step": 11578 + }, + { + "epoch": 0.7955888415555861, + "grad_norm": 1.0664708614349365, + "learning_rate": 6.099122882764352e-06, + "loss": 0.9618, + "step": 11579 + }, + { + "epoch": 0.7956575511886766, + "grad_norm": 1.014922857284546, + "learning_rate": 6.095198647752376e-06, + "loss": 0.9072, + "step": 11580 + }, + { + "epoch": 0.7957262608217672, + "grad_norm": 1.0850597620010376, + "learning_rate": 6.091275500322097e-06, + "loss": 0.9302, + "step": 11581 + }, + { + "epoch": 0.7957949704548578, + "grad_norm": 1.0427442789077759, + "learning_rate": 6.087353440699206e-06, + "loss": 0.9232, + "step": 11582 + }, + { + "epoch": 0.7958636800879483, + "grad_norm": 1.056435465812683, + "learning_rate": 6.0834324691093486e-06, + "loss": 0.965, + "step": 11583 + }, + { + "epoch": 0.7959323897210389, + "grad_norm": 1.115054726600647, + "learning_rate": 6.079512585778077e-06, + "loss": 0.9558, + "step": 11584 + }, + { + "epoch": 0.7960010993541294, + "grad_norm": 1.0913327932357788, + "learning_rate": 6.075593790930914e-06, + "loss": 0.9591, + "step": 11585 + }, + { + "epoch": 0.79606980898722, + "grad_norm": 1.0423665046691895, + "learning_rate": 6.071676084793299e-06, + "loss": 0.9126, + "step": 11586 + }, + { + "epoch": 0.7961385186203106, + "grad_norm": 1.0615965127944946, + "learning_rate": 6.067759467590617e-06, + "loss": 0.9334, + "step": 11587 + }, + { + "epoch": 0.7962072282534012, + "grad_norm": 1.028915524482727, + "learning_rate": 6.0638439395482e-06, + "loss": 0.94, + "step": 11588 + }, + { + "epoch": 0.7962759378864916, + "grad_norm": 1.11745285987854, + "learning_rate": 6.0599295008912846e-06, + "loss": 0.9485, + "step": 11589 + }, + { + "epoch": 0.7963446475195822, + "grad_norm": 1.0714564323425293, + "learning_rate": 6.056016151845076e-06, + "loss": 0.932, + "step": 11590 + }, + { + "epoch": 0.7964133571526728, + "grad_norm": 1.053489327430725, + "learning_rate": 6.052103892634712e-06, + "loss": 0.9082, + "step": 11591 + }, + { + "epoch": 0.7964820667857634, + "grad_norm": 1.059496283531189, + "learning_rate": 6.048192723485252e-06, + "loss": 0.9, + "step": 11592 + }, + { + "epoch": 0.796550776418854, + "grad_norm": 1.0601105690002441, + "learning_rate": 6.044282644621721e-06, + "loss": 0.943, + "step": 11593 + }, + { + "epoch": 0.7966194860519444, + "grad_norm": 1.1705859899520874, + "learning_rate": 6.040373656269041e-06, + "loss": 0.9527, + "step": 11594 + }, + { + "epoch": 0.796688195685035, + "grad_norm": 1.0966030359268188, + "learning_rate": 6.036465758652107e-06, + "loss": 0.9877, + "step": 11595 + }, + { + "epoch": 0.7967569053181256, + "grad_norm": 1.0284279584884644, + "learning_rate": 6.032558951995734e-06, + "loss": 0.91, + "step": 11596 + }, + { + "epoch": 0.7968256149512162, + "grad_norm": 1.082731008529663, + "learning_rate": 6.028653236524678e-06, + "loss": 0.9514, + "step": 11597 + }, + { + "epoch": 0.7968943245843068, + "grad_norm": 1.1137397289276123, + "learning_rate": 6.024748612463643e-06, + "loss": 0.9597, + "step": 11598 + }, + { + "epoch": 0.7969630342173972, + "grad_norm": 1.0260486602783203, + "learning_rate": 6.020845080037243e-06, + "loss": 0.9083, + "step": 11599 + }, + { + "epoch": 0.7970317438504878, + "grad_norm": 1.0417282581329346, + "learning_rate": 6.016942639470053e-06, + "loss": 0.9624, + "step": 11600 + }, + { + "epoch": 0.7971004534835784, + "grad_norm": 1.0258989334106445, + "learning_rate": 6.0130412909865845e-06, + "loss": 0.9456, + "step": 11601 + }, + { + "epoch": 0.797169163116669, + "grad_norm": 1.0872849225997925, + "learning_rate": 6.009141034811269e-06, + "loss": 0.9533, + "step": 11602 + }, + { + "epoch": 0.7972378727497595, + "grad_norm": 1.0258392095565796, + "learning_rate": 6.005241871168496e-06, + "loss": 0.9243, + "step": 11603 + }, + { + "epoch": 0.7973065823828501, + "grad_norm": 1.0648971796035767, + "learning_rate": 6.001343800282569e-06, + "loss": 0.9231, + "step": 11604 + }, + { + "epoch": 0.7973752920159406, + "grad_norm": 1.0730621814727783, + "learning_rate": 5.997446822377751e-06, + "loss": 0.93, + "step": 11605 + }, + { + "epoch": 0.7974440016490312, + "grad_norm": 1.0736360549926758, + "learning_rate": 5.993550937678228e-06, + "loss": 0.9439, + "step": 11606 + }, + { + "epoch": 0.7975127112821218, + "grad_norm": 1.0489094257354736, + "learning_rate": 5.989656146408132e-06, + "loss": 0.9375, + "step": 11607 + }, + { + "epoch": 0.7975814209152123, + "grad_norm": 1.0335297584533691, + "learning_rate": 5.985762448791532e-06, + "loss": 0.9204, + "step": 11608 + }, + { + "epoch": 0.7976501305483029, + "grad_norm": 1.056985855102539, + "learning_rate": 5.98186984505242e-06, + "loss": 0.9568, + "step": 11609 + }, + { + "epoch": 0.7977188401813934, + "grad_norm": 1.0770875215530396, + "learning_rate": 5.977978335414739e-06, + "loss": 0.961, + "step": 11610 + }, + { + "epoch": 0.797787549814484, + "grad_norm": 1.1321660280227661, + "learning_rate": 5.974087920102367e-06, + "loss": 0.8998, + "step": 11611 + }, + { + "epoch": 0.7978562594475745, + "grad_norm": 1.060296893119812, + "learning_rate": 5.970198599339114e-06, + "loss": 0.8929, + "step": 11612 + }, + { + "epoch": 0.7979249690806651, + "grad_norm": 1.0591343641281128, + "learning_rate": 5.9663103733487355e-06, + "loss": 0.9336, + "step": 11613 + }, + { + "epoch": 0.7979936787137557, + "grad_norm": 1.0411432981491089, + "learning_rate": 5.96242324235492e-06, + "loss": 0.912, + "step": 11614 + }, + { + "epoch": 0.7980623883468462, + "grad_norm": 1.1117656230926514, + "learning_rate": 5.95853720658128e-06, + "loss": 0.9351, + "step": 11615 + }, + { + "epoch": 0.7981310979799368, + "grad_norm": 0.9786347150802612, + "learning_rate": 5.954652266251387e-06, + "loss": 0.9205, + "step": 11616 + }, + { + "epoch": 0.7981998076130273, + "grad_norm": 1.0897395610809326, + "learning_rate": 5.950768421588734e-06, + "loss": 0.9414, + "step": 11617 + }, + { + "epoch": 0.7982685172461179, + "grad_norm": 1.0318725109100342, + "learning_rate": 5.946885672816763e-06, + "loss": 0.9101, + "step": 11618 + }, + { + "epoch": 0.7983372268792085, + "grad_norm": 1.14364492893219, + "learning_rate": 5.943004020158846e-06, + "loss": 0.9616, + "step": 11619 + }, + { + "epoch": 0.7984059365122991, + "grad_norm": 1.0156749486923218, + "learning_rate": 5.939123463838281e-06, + "loss": 0.8923, + "step": 11620 + }, + { + "epoch": 0.7984746461453895, + "grad_norm": 1.051455020904541, + "learning_rate": 5.9352440040783195e-06, + "loss": 0.9335, + "step": 11621 + }, + { + "epoch": 0.7985433557784801, + "grad_norm": 1.1013621091842651, + "learning_rate": 5.931365641102149e-06, + "loss": 0.934, + "step": 11622 + }, + { + "epoch": 0.7986120654115707, + "grad_norm": 1.0261526107788086, + "learning_rate": 5.927488375132892e-06, + "loss": 0.9155, + "step": 11623 + }, + { + "epoch": 0.7986807750446613, + "grad_norm": 1.047407627105713, + "learning_rate": 5.923612206393592e-06, + "loss": 0.9312, + "step": 11624 + }, + { + "epoch": 0.7987494846777519, + "grad_norm": 1.0115793943405151, + "learning_rate": 5.919737135107259e-06, + "loss": 0.9082, + "step": 11625 + }, + { + "epoch": 0.7988181943108423, + "grad_norm": 1.0328246355056763, + "learning_rate": 5.915863161496807e-06, + "loss": 0.9309, + "step": 11626 + }, + { + "epoch": 0.7988869039439329, + "grad_norm": 1.0230532884597778, + "learning_rate": 5.91199028578511e-06, + "loss": 0.8849, + "step": 11627 + }, + { + "epoch": 0.7989556135770235, + "grad_norm": 1.1132595539093018, + "learning_rate": 5.9081185081949735e-06, + "loss": 0.9047, + "step": 11628 + }, + { + "epoch": 0.7990243232101141, + "grad_norm": 1.1352921724319458, + "learning_rate": 5.9042478289491474e-06, + "loss": 1.0155, + "step": 11629 + }, + { + "epoch": 0.7990930328432047, + "grad_norm": 1.0787079334259033, + "learning_rate": 5.900378248270291e-06, + "loss": 0.942, + "step": 11630 + }, + { + "epoch": 0.7991617424762951, + "grad_norm": 0.9849157929420471, + "learning_rate": 5.896509766381028e-06, + "loss": 0.9235, + "step": 11631 + }, + { + "epoch": 0.7992304521093857, + "grad_norm": 1.141903042793274, + "learning_rate": 5.892642383503913e-06, + "loss": 0.9823, + "step": 11632 + }, + { + "epoch": 0.7992991617424763, + "grad_norm": 1.0351388454437256, + "learning_rate": 5.888776099861429e-06, + "loss": 0.9323, + "step": 11633 + }, + { + "epoch": 0.7993678713755669, + "grad_norm": 1.0628724098205566, + "learning_rate": 5.884910915676012e-06, + "loss": 0.9351, + "step": 11634 + }, + { + "epoch": 0.7994365810086574, + "grad_norm": 1.0922620296478271, + "learning_rate": 5.8810468311700065e-06, + "loss": 0.9218, + "step": 11635 + }, + { + "epoch": 0.799505290641748, + "grad_norm": 1.0560789108276367, + "learning_rate": 5.87718384656572e-06, + "loss": 0.9404, + "step": 11636 + }, + { + "epoch": 0.7995740002748385, + "grad_norm": 1.072873592376709, + "learning_rate": 5.873321962085384e-06, + "loss": 0.9207, + "step": 11637 + }, + { + "epoch": 0.7996427099079291, + "grad_norm": 1.0630191564559937, + "learning_rate": 5.869461177951177e-06, + "loss": 0.9523, + "step": 11638 + }, + { + "epoch": 0.7997114195410197, + "grad_norm": 1.101401925086975, + "learning_rate": 5.865601494385212e-06, + "loss": 0.9292, + "step": 11639 + }, + { + "epoch": 0.7997801291741102, + "grad_norm": 1.0275269746780396, + "learning_rate": 5.861742911609516e-06, + "loss": 0.954, + "step": 11640 + }, + { + "epoch": 0.7998488388072008, + "grad_norm": 1.0166654586791992, + "learning_rate": 5.857885429846082e-06, + "loss": 0.9224, + "step": 11641 + }, + { + "epoch": 0.7999175484402913, + "grad_norm": 1.1081326007843018, + "learning_rate": 5.8540290493168284e-06, + "loss": 0.9233, + "step": 11642 + }, + { + "epoch": 0.7999862580733819, + "grad_norm": 1.0366677045822144, + "learning_rate": 5.850173770243608e-06, + "loss": 0.9028, + "step": 11643 + }, + { + "epoch": 0.8000549677064724, + "grad_norm": 1.049536108970642, + "learning_rate": 5.846319592848223e-06, + "loss": 0.9386, + "step": 11644 + }, + { + "epoch": 0.800123677339563, + "grad_norm": 1.0672003030776978, + "learning_rate": 5.842466517352385e-06, + "loss": 0.9648, + "step": 11645 + }, + { + "epoch": 0.8001923869726536, + "grad_norm": 1.0394200086593628, + "learning_rate": 5.838614543977766e-06, + "loss": 0.9231, + "step": 11646 + }, + { + "epoch": 0.8002610966057441, + "grad_norm": 1.028274416923523, + "learning_rate": 5.834763672945978e-06, + "loss": 0.9, + "step": 11647 + }, + { + "epoch": 0.8003298062388347, + "grad_norm": 1.0565589666366577, + "learning_rate": 5.830913904478544e-06, + "loss": 0.904, + "step": 11648 + }, + { + "epoch": 0.8003985158719252, + "grad_norm": 1.0370243787765503, + "learning_rate": 5.8270652387969445e-06, + "loss": 0.9618, + "step": 11649 + }, + { + "epoch": 0.8004672255050158, + "grad_norm": 1.0161023139953613, + "learning_rate": 5.823217676122597e-06, + "loss": 0.9535, + "step": 11650 + }, + { + "epoch": 0.8005359351381064, + "grad_norm": 1.036752700805664, + "learning_rate": 5.819371216676839e-06, + "loss": 0.926, + "step": 11651 + }, + { + "epoch": 0.800604644771197, + "grad_norm": 1.0437209606170654, + "learning_rate": 5.815525860680959e-06, + "loss": 0.9338, + "step": 11652 + }, + { + "epoch": 0.8006733544042874, + "grad_norm": 1.0571045875549316, + "learning_rate": 5.811681608356179e-06, + "loss": 0.9051, + "step": 11653 + }, + { + "epoch": 0.800742064037378, + "grad_norm": 1.0673511028289795, + "learning_rate": 5.807838459923662e-06, + "loss": 0.9197, + "step": 11654 + }, + { + "epoch": 0.8008107736704686, + "grad_norm": 1.024465799331665, + "learning_rate": 5.80399641560449e-06, + "loss": 0.8942, + "step": 11655 + }, + { + "epoch": 0.8008794833035592, + "grad_norm": 1.0972195863723755, + "learning_rate": 5.800155475619701e-06, + "loss": 0.9674, + "step": 11656 + }, + { + "epoch": 0.8009481929366498, + "grad_norm": 1.0281952619552612, + "learning_rate": 5.796315640190259e-06, + "loss": 0.9617, + "step": 11657 + }, + { + "epoch": 0.8010169025697402, + "grad_norm": 1.0272712707519531, + "learning_rate": 5.792476909537067e-06, + "loss": 0.904, + "step": 11658 + }, + { + "epoch": 0.8010856122028308, + "grad_norm": 1.0124272108078003, + "learning_rate": 5.788639283880978e-06, + "loss": 0.9272, + "step": 11659 + }, + { + "epoch": 0.8011543218359214, + "grad_norm": 1.0404976606369019, + "learning_rate": 5.784802763442748e-06, + "loss": 0.9572, + "step": 11660 + }, + { + "epoch": 0.801223031469012, + "grad_norm": 1.0271576642990112, + "learning_rate": 5.780967348443101e-06, + "loss": 0.9484, + "step": 11661 + }, + { + "epoch": 0.8012917411021026, + "grad_norm": 1.0614358186721802, + "learning_rate": 5.7771330391026835e-06, + "loss": 0.8853, + "step": 11662 + }, + { + "epoch": 0.801360450735193, + "grad_norm": 1.1545122861862183, + "learning_rate": 5.7732998356420796e-06, + "loss": 0.9654, + "step": 11663 + }, + { + "epoch": 0.8014291603682836, + "grad_norm": 1.0189918279647827, + "learning_rate": 5.76946773828182e-06, + "loss": 0.9169, + "step": 11664 + }, + { + "epoch": 0.8014978700013742, + "grad_norm": 1.0357369184494019, + "learning_rate": 5.7656367472423504e-06, + "loss": 0.9174, + "step": 11665 + }, + { + "epoch": 0.8015665796344648, + "grad_norm": 1.0133332014083862, + "learning_rate": 5.761806862744071e-06, + "loss": 0.9344, + "step": 11666 + }, + { + "epoch": 0.8016352892675553, + "grad_norm": 1.0682774782180786, + "learning_rate": 5.757978085007312e-06, + "loss": 0.9231, + "step": 11667 + }, + { + "epoch": 0.8017039989006459, + "grad_norm": 1.0967419147491455, + "learning_rate": 5.75415041425234e-06, + "loss": 0.9428, + "step": 11668 + }, + { + "epoch": 0.8017727085337364, + "grad_norm": 1.0359203815460205, + "learning_rate": 5.7503238506993686e-06, + "loss": 0.9216, + "step": 11669 + }, + { + "epoch": 0.801841418166827, + "grad_norm": 1.032500982284546, + "learning_rate": 5.7464983945685224e-06, + "loss": 0.889, + "step": 11670 + }, + { + "epoch": 0.8019101277999175, + "grad_norm": 1.0887975692749023, + "learning_rate": 5.742674046079882e-06, + "loss": 0.9347, + "step": 11671 + }, + { + "epoch": 0.8019788374330081, + "grad_norm": 1.0576008558273315, + "learning_rate": 5.738850805453471e-06, + "loss": 0.93, + "step": 11672 + }, + { + "epoch": 0.8020475470660987, + "grad_norm": 1.0849335193634033, + "learning_rate": 5.73502867290922e-06, + "loss": 0.947, + "step": 11673 + }, + { + "epoch": 0.8021162566991892, + "grad_norm": 1.1475919485092163, + "learning_rate": 5.731207648667025e-06, + "loss": 0.9413, + "step": 11674 + }, + { + "epoch": 0.8021849663322798, + "grad_norm": 1.005241870880127, + "learning_rate": 5.72738773294671e-06, + "loss": 0.8856, + "step": 11675 + }, + { + "epoch": 0.8022536759653703, + "grad_norm": 1.1000703573226929, + "learning_rate": 5.723568925968023e-06, + "loss": 0.9191, + "step": 11676 + }, + { + "epoch": 0.8023223855984609, + "grad_norm": 1.1341891288757324, + "learning_rate": 5.719751227950659e-06, + "loss": 0.9685, + "step": 11677 + }, + { + "epoch": 0.8023910952315515, + "grad_norm": 1.0421631336212158, + "learning_rate": 5.7159346391142535e-06, + "loss": 0.9088, + "step": 11678 + }, + { + "epoch": 0.802459804864642, + "grad_norm": 1.072412133216858, + "learning_rate": 5.71211915967837e-06, + "loss": 0.909, + "step": 11679 + }, + { + "epoch": 0.8025285144977325, + "grad_norm": 1.0134072303771973, + "learning_rate": 5.708304789862517e-06, + "loss": 0.9445, + "step": 11680 + }, + { + "epoch": 0.8025972241308231, + "grad_norm": 1.02085542678833, + "learning_rate": 5.704491529886122e-06, + "loss": 0.9204, + "step": 11681 + }, + { + "epoch": 0.8026659337639137, + "grad_norm": 1.0413662195205688, + "learning_rate": 5.7006793799685595e-06, + "loss": 0.9647, + "step": 11682 + }, + { + "epoch": 0.8027346433970043, + "grad_norm": 0.9790916442871094, + "learning_rate": 5.696868340329148e-06, + "loss": 0.8843, + "step": 11683 + }, + { + "epoch": 0.8028033530300949, + "grad_norm": 1.0626850128173828, + "learning_rate": 5.693058411187133e-06, + "loss": 0.9522, + "step": 11684 + }, + { + "epoch": 0.8028720626631853, + "grad_norm": 1.1258840560913086, + "learning_rate": 5.6892495927617e-06, + "loss": 0.9317, + "step": 11685 + }, + { + "epoch": 0.8029407722962759, + "grad_norm": 1.0204970836639404, + "learning_rate": 5.685441885271955e-06, + "loss": 0.9074, + "step": 11686 + }, + { + "epoch": 0.8030094819293665, + "grad_norm": 1.0551564693450928, + "learning_rate": 5.681635288936965e-06, + "loss": 0.9567, + "step": 11687 + }, + { + "epoch": 0.8030781915624571, + "grad_norm": 1.0589076280593872, + "learning_rate": 5.677829803975715e-06, + "loss": 0.9132, + "step": 11688 + }, + { + "epoch": 0.8031469011955477, + "grad_norm": 1.0621774196624756, + "learning_rate": 5.674025430607136e-06, + "loss": 0.909, + "step": 11689 + }, + { + "epoch": 0.8032156108286381, + "grad_norm": 1.0133498907089233, + "learning_rate": 5.670222169050096e-06, + "loss": 0.914, + "step": 11690 + }, + { + "epoch": 0.8032843204617287, + "grad_norm": 1.141825795173645, + "learning_rate": 5.666420019523383e-06, + "loss": 0.9587, + "step": 11691 + }, + { + "epoch": 0.8033530300948193, + "grad_norm": 1.0493803024291992, + "learning_rate": 5.662618982245741e-06, + "loss": 0.9344, + "step": 11692 + }, + { + "epoch": 0.8034217397279099, + "grad_norm": 1.0044583082199097, + "learning_rate": 5.658819057435832e-06, + "loss": 0.9156, + "step": 11693 + }, + { + "epoch": 0.8034904493610004, + "grad_norm": 1.08216392993927, + "learning_rate": 5.655020245312281e-06, + "loss": 0.9544, + "step": 11694 + }, + { + "epoch": 0.8035591589940909, + "grad_norm": 1.0834559202194214, + "learning_rate": 5.651222546093613e-06, + "loss": 0.9422, + "step": 11695 + }, + { + "epoch": 0.8036278686271815, + "grad_norm": 1.0351797342300415, + "learning_rate": 5.64742595999832e-06, + "loss": 0.9334, + "step": 11696 + }, + { + "epoch": 0.8036965782602721, + "grad_norm": 1.0838686227798462, + "learning_rate": 5.643630487244805e-06, + "loss": 0.9806, + "step": 11697 + }, + { + "epoch": 0.8037652878933627, + "grad_norm": 1.037449836730957, + "learning_rate": 5.639836128051426e-06, + "loss": 0.9075, + "step": 11698 + }, + { + "epoch": 0.8038339975264532, + "grad_norm": 1.0273146629333496, + "learning_rate": 5.636042882636469e-06, + "loss": 0.9094, + "step": 11699 + }, + { + "epoch": 0.8039027071595438, + "grad_norm": 1.0581915378570557, + "learning_rate": 5.632250751218165e-06, + "loss": 0.9189, + "step": 11700 + }, + { + "epoch": 0.8039714167926343, + "grad_norm": 1.0848101377487183, + "learning_rate": 5.62845973401466e-06, + "loss": 0.9295, + "step": 11701 + }, + { + "epoch": 0.8040401264257249, + "grad_norm": 1.1011617183685303, + "learning_rate": 5.624669831244056e-06, + "loss": 0.9024, + "step": 11702 + }, + { + "epoch": 0.8041088360588154, + "grad_norm": 1.10554039478302, + "learning_rate": 5.6208810431243815e-06, + "loss": 0.9829, + "step": 11703 + }, + { + "epoch": 0.804177545691906, + "grad_norm": 1.127638578414917, + "learning_rate": 5.617093369873605e-06, + "loss": 0.9677, + "step": 11704 + }, + { + "epoch": 0.8042462553249966, + "grad_norm": 1.1153948307037354, + "learning_rate": 5.613306811709634e-06, + "loss": 0.9294, + "step": 11705 + }, + { + "epoch": 0.8043149649580871, + "grad_norm": 1.0602132081985474, + "learning_rate": 5.6095213688502956e-06, + "loss": 0.9444, + "step": 11706 + }, + { + "epoch": 0.8043836745911777, + "grad_norm": 1.0578782558441162, + "learning_rate": 5.60573704151337e-06, + "loss": 0.9073, + "step": 11707 + }, + { + "epoch": 0.8044523842242682, + "grad_norm": 1.0790313482284546, + "learning_rate": 5.601953829916568e-06, + "loss": 0.9126, + "step": 11708 + }, + { + "epoch": 0.8045210938573588, + "grad_norm": 1.0110929012298584, + "learning_rate": 5.598171734277536e-06, + "loss": 0.9381, + "step": 11709 + }, + { + "epoch": 0.8045898034904494, + "grad_norm": 1.1288154125213623, + "learning_rate": 5.594390754813858e-06, + "loss": 0.9563, + "step": 11710 + }, + { + "epoch": 0.80465851312354, + "grad_norm": 1.1594542264938354, + "learning_rate": 5.5906108917430394e-06, + "loss": 0.9164, + "step": 11711 + }, + { + "epoch": 0.8047272227566304, + "grad_norm": 1.0729150772094727, + "learning_rate": 5.586832145282544e-06, + "loss": 0.9275, + "step": 11712 + }, + { + "epoch": 0.804795932389721, + "grad_norm": 0.9641766548156738, + "learning_rate": 5.5830545156497575e-06, + "loss": 0.9101, + "step": 11713 + }, + { + "epoch": 0.8048646420228116, + "grad_norm": 1.0012494325637817, + "learning_rate": 5.579278003062008e-06, + "loss": 0.9309, + "step": 11714 + }, + { + "epoch": 0.8049333516559022, + "grad_norm": 1.0881421566009521, + "learning_rate": 5.575502607736557e-06, + "loss": 0.9716, + "step": 11715 + }, + { + "epoch": 0.8050020612889928, + "grad_norm": 1.0894229412078857, + "learning_rate": 5.5717283298905934e-06, + "loss": 0.931, + "step": 11716 + }, + { + "epoch": 0.8050707709220832, + "grad_norm": 1.0102348327636719, + "learning_rate": 5.5679551697412514e-06, + "loss": 0.9355, + "step": 11717 + }, + { + "epoch": 0.8051394805551738, + "grad_norm": 1.082130789756775, + "learning_rate": 5.564183127505607e-06, + "loss": 0.9634, + "step": 11718 + }, + { + "epoch": 0.8052081901882644, + "grad_norm": 1.0665228366851807, + "learning_rate": 5.5604122034006505e-06, + "loss": 0.9395, + "step": 11719 + }, + { + "epoch": 0.805276899821355, + "grad_norm": 1.1072492599487305, + "learning_rate": 5.556642397643325e-06, + "loss": 0.9538, + "step": 11720 + }, + { + "epoch": 0.8053456094544456, + "grad_norm": 1.0352468490600586, + "learning_rate": 5.552873710450515e-06, + "loss": 0.924, + "step": 11721 + }, + { + "epoch": 0.805414319087536, + "grad_norm": 1.0429801940917969, + "learning_rate": 5.549106142039018e-06, + "loss": 0.9098, + "step": 11722 + }, + { + "epoch": 0.8054830287206266, + "grad_norm": 1.0650142431259155, + "learning_rate": 5.545339692625581e-06, + "loss": 0.9047, + "step": 11723 + }, + { + "epoch": 0.8055517383537172, + "grad_norm": 1.1271307468414307, + "learning_rate": 5.541574362426893e-06, + "loss": 0.9304, + "step": 11724 + }, + { + "epoch": 0.8056204479868078, + "grad_norm": 1.0517858266830444, + "learning_rate": 5.537810151659564e-06, + "loss": 0.8864, + "step": 11725 + }, + { + "epoch": 0.8056891576198983, + "grad_norm": 1.1146266460418701, + "learning_rate": 5.53404706054016e-06, + "loss": 0.9351, + "step": 11726 + }, + { + "epoch": 0.8057578672529889, + "grad_norm": 1.1061112880706787, + "learning_rate": 5.5302850892851535e-06, + "loss": 0.9745, + "step": 11727 + }, + { + "epoch": 0.8058265768860794, + "grad_norm": 1.0742870569229126, + "learning_rate": 5.526524238110972e-06, + "loss": 0.9113, + "step": 11728 + }, + { + "epoch": 0.80589528651917, + "grad_norm": 1.003791332244873, + "learning_rate": 5.522764507233977e-06, + "loss": 0.9363, + "step": 11729 + }, + { + "epoch": 0.8059639961522606, + "grad_norm": 1.0281296968460083, + "learning_rate": 5.519005896870466e-06, + "loss": 0.9499, + "step": 11730 + }, + { + "epoch": 0.8060327057853511, + "grad_norm": 1.0226004123687744, + "learning_rate": 5.515248407236673e-06, + "loss": 0.8968, + "step": 11731 + }, + { + "epoch": 0.8061014154184417, + "grad_norm": 1.0444284677505493, + "learning_rate": 5.511492038548752e-06, + "loss": 0.9201, + "step": 11732 + }, + { + "epoch": 0.8061701250515322, + "grad_norm": 1.0580583810806274, + "learning_rate": 5.507736791022813e-06, + "loss": 0.9256, + "step": 11733 + }, + { + "epoch": 0.8062388346846228, + "grad_norm": 1.125615119934082, + "learning_rate": 5.50398266487489e-06, + "loss": 0.9877, + "step": 11734 + }, + { + "epoch": 0.8063075443177133, + "grad_norm": 0.988778293132782, + "learning_rate": 5.500229660320958e-06, + "loss": 0.9339, + "step": 11735 + }, + { + "epoch": 0.8063762539508039, + "grad_norm": 1.0384005308151245, + "learning_rate": 5.496477777576928e-06, + "loss": 0.9298, + "step": 11736 + }, + { + "epoch": 0.8064449635838945, + "grad_norm": 1.0354011058807373, + "learning_rate": 5.4927270168586344e-06, + "loss": 0.9458, + "step": 11737 + }, + { + "epoch": 0.806513673216985, + "grad_norm": 1.1005091667175293, + "learning_rate": 5.48897737838186e-06, + "loss": 0.9603, + "step": 11738 + }, + { + "epoch": 0.8065823828500756, + "grad_norm": 1.0614088773727417, + "learning_rate": 5.485228862362324e-06, + "loss": 0.961, + "step": 11739 + }, + { + "epoch": 0.8066510924831661, + "grad_norm": 1.0062161684036255, + "learning_rate": 5.4814814690156705e-06, + "loss": 0.9205, + "step": 11740 + }, + { + "epoch": 0.8067198021162567, + "grad_norm": 1.037297010421753, + "learning_rate": 5.4777351985574934e-06, + "loss": 0.9187, + "step": 11741 + }, + { + "epoch": 0.8067885117493473, + "grad_norm": 1.0842350721359253, + "learning_rate": 5.473990051203298e-06, + "loss": 0.9881, + "step": 11742 + }, + { + "epoch": 0.8068572213824379, + "grad_norm": 1.1445146799087524, + "learning_rate": 5.47024602716856e-06, + "loss": 0.9454, + "step": 11743 + }, + { + "epoch": 0.8069259310155283, + "grad_norm": 1.0431287288665771, + "learning_rate": 5.466503126668651e-06, + "loss": 0.9438, + "step": 11744 + }, + { + "epoch": 0.8069946406486189, + "grad_norm": 1.1095128059387207, + "learning_rate": 5.46276134991891e-06, + "loss": 0.9008, + "step": 11745 + }, + { + "epoch": 0.8070633502817095, + "grad_norm": 1.0710927248001099, + "learning_rate": 5.459020697134601e-06, + "loss": 0.9718, + "step": 11746 + }, + { + "epoch": 0.8071320599148001, + "grad_norm": 1.0191768407821655, + "learning_rate": 5.455281168530907e-06, + "loss": 0.923, + "step": 11747 + }, + { + "epoch": 0.8072007695478907, + "grad_norm": 1.0251054763793945, + "learning_rate": 5.451542764322973e-06, + "loss": 0.9539, + "step": 11748 + }, + { + "epoch": 0.8072694791809811, + "grad_norm": 1.0644619464874268, + "learning_rate": 5.4478054847258655e-06, + "loss": 0.9247, + "step": 11749 + }, + { + "epoch": 0.8073381888140717, + "grad_norm": 1.0785045623779297, + "learning_rate": 5.444069329954585e-06, + "loss": 0.9121, + "step": 11750 + }, + { + "epoch": 0.8074068984471623, + "grad_norm": 1.0514142513275146, + "learning_rate": 5.440334300224079e-06, + "loss": 0.9208, + "step": 11751 + }, + { + "epoch": 0.8074756080802529, + "grad_norm": 1.0696004629135132, + "learning_rate": 5.436600395749209e-06, + "loss": 0.9308, + "step": 11752 + }, + { + "epoch": 0.8075443177133435, + "grad_norm": 1.031059741973877, + "learning_rate": 5.432867616744788e-06, + "loss": 0.9313, + "step": 11753 + }, + { + "epoch": 0.8076130273464339, + "grad_norm": 1.1342127323150635, + "learning_rate": 5.429135963425563e-06, + "loss": 0.9299, + "step": 11754 + }, + { + "epoch": 0.8076817369795245, + "grad_norm": 1.119389533996582, + "learning_rate": 5.425405436006214e-06, + "loss": 0.9276, + "step": 11755 + }, + { + "epoch": 0.8077504466126151, + "grad_norm": 1.1251744031906128, + "learning_rate": 5.421676034701362e-06, + "loss": 0.9545, + "step": 11756 + }, + { + "epoch": 0.8078191562457057, + "grad_norm": 1.0067435503005981, + "learning_rate": 5.417947759725542e-06, + "loss": 0.9506, + "step": 11757 + }, + { + "epoch": 0.8078878658787962, + "grad_norm": 1.071804404258728, + "learning_rate": 5.414220611293247e-06, + "loss": 0.9443, + "step": 11758 + }, + { + "epoch": 0.8079565755118868, + "grad_norm": 1.0896965265274048, + "learning_rate": 5.410494589618897e-06, + "loss": 0.9613, + "step": 11759 + }, + { + "epoch": 0.8080252851449773, + "grad_norm": 1.0067752599716187, + "learning_rate": 5.406769694916852e-06, + "loss": 0.9148, + "step": 11760 + }, + { + "epoch": 0.8080939947780679, + "grad_norm": 1.021650791168213, + "learning_rate": 5.403045927401406e-06, + "loss": 0.9351, + "step": 11761 + }, + { + "epoch": 0.8081627044111585, + "grad_norm": 1.0469319820404053, + "learning_rate": 5.399323287286773e-06, + "loss": 0.9477, + "step": 11762 + }, + { + "epoch": 0.808231414044249, + "grad_norm": 1.0970561504364014, + "learning_rate": 5.395601774787118e-06, + "loss": 0.9523, + "step": 11763 + }, + { + "epoch": 0.8083001236773396, + "grad_norm": 1.1380047798156738, + "learning_rate": 5.39188139011654e-06, + "loss": 0.9532, + "step": 11764 + }, + { + "epoch": 0.8083688333104301, + "grad_norm": 1.0001360177993774, + "learning_rate": 5.388162133489078e-06, + "loss": 0.91, + "step": 11765 + }, + { + "epoch": 0.8084375429435207, + "grad_norm": 1.0724600553512573, + "learning_rate": 5.38444400511868e-06, + "loss": 0.9422, + "step": 11766 + }, + { + "epoch": 0.8085062525766112, + "grad_norm": 1.0454802513122559, + "learning_rate": 5.380727005219261e-06, + "loss": 0.933, + "step": 11767 + }, + { + "epoch": 0.8085749622097018, + "grad_norm": 1.0366263389587402, + "learning_rate": 5.3770111340046615e-06, + "loss": 0.9246, + "step": 11768 + }, + { + "epoch": 0.8086436718427924, + "grad_norm": 1.0018889904022217, + "learning_rate": 5.3732963916886395e-06, + "loss": 0.9039, + "step": 11769 + }, + { + "epoch": 0.8087123814758829, + "grad_norm": 1.0436729192733765, + "learning_rate": 5.3695827784849105e-06, + "loss": 0.9518, + "step": 11770 + }, + { + "epoch": 0.8087810911089734, + "grad_norm": 1.0990538597106934, + "learning_rate": 5.36587029460712e-06, + "loss": 0.9246, + "step": 11771 + }, + { + "epoch": 0.808849800742064, + "grad_norm": 1.089285969734192, + "learning_rate": 5.362158940268833e-06, + "loss": 0.9648, + "step": 11772 + }, + { + "epoch": 0.8089185103751546, + "grad_norm": 1.0700281858444214, + "learning_rate": 5.35844871568357e-06, + "loss": 0.906, + "step": 11773 + }, + { + "epoch": 0.8089872200082452, + "grad_norm": 1.083311915397644, + "learning_rate": 5.3547396210647745e-06, + "loss": 0.9436, + "step": 11774 + }, + { + "epoch": 0.8090559296413358, + "grad_norm": 1.027605652809143, + "learning_rate": 5.3510316566258326e-06, + "loss": 0.9012, + "step": 11775 + }, + { + "epoch": 0.8091246392744262, + "grad_norm": 1.1479980945587158, + "learning_rate": 5.347324822580055e-06, + "loss": 0.9796, + "step": 11776 + }, + { + "epoch": 0.8091933489075168, + "grad_norm": 1.115868091583252, + "learning_rate": 5.343619119140708e-06, + "loss": 0.938, + "step": 11777 + }, + { + "epoch": 0.8092620585406074, + "grad_norm": 1.0676161050796509, + "learning_rate": 5.339914546520958e-06, + "loss": 0.9597, + "step": 11778 + }, + { + "epoch": 0.809330768173698, + "grad_norm": 1.0535883903503418, + "learning_rate": 5.336211104933939e-06, + "loss": 0.9245, + "step": 11779 + }, + { + "epoch": 0.8093994778067886, + "grad_norm": 1.0363796949386597, + "learning_rate": 5.332508794592705e-06, + "loss": 0.9251, + "step": 11780 + }, + { + "epoch": 0.809468187439879, + "grad_norm": 1.0303703546524048, + "learning_rate": 5.328807615710246e-06, + "loss": 0.926, + "step": 11781 + }, + { + "epoch": 0.8095368970729696, + "grad_norm": 1.1172465085983276, + "learning_rate": 5.325107568499502e-06, + "loss": 0.9522, + "step": 11782 + }, + { + "epoch": 0.8096056067060602, + "grad_norm": 1.028916358947754, + "learning_rate": 5.321408653173313e-06, + "loss": 0.8942, + "step": 11783 + }, + { + "epoch": 0.8096743163391508, + "grad_norm": 1.0696191787719727, + "learning_rate": 5.317710869944487e-06, + "loss": 0.9349, + "step": 11784 + }, + { + "epoch": 0.8097430259722413, + "grad_norm": 1.0056804418563843, + "learning_rate": 5.314014219025751e-06, + "loss": 0.9092, + "step": 11785 + }, + { + "epoch": 0.8098117356053318, + "grad_norm": 0.9994490742683411, + "learning_rate": 5.310318700629777e-06, + "loss": 0.9261, + "step": 11786 + }, + { + "epoch": 0.8098804452384224, + "grad_norm": 1.011383295059204, + "learning_rate": 5.306624314969167e-06, + "loss": 0.8917, + "step": 11787 + }, + { + "epoch": 0.809949154871513, + "grad_norm": 1.0488420724868774, + "learning_rate": 5.302931062256447e-06, + "loss": 0.9352, + "step": 11788 + }, + { + "epoch": 0.8100178645046036, + "grad_norm": 1.0116208791732788, + "learning_rate": 5.299238942704094e-06, + "loss": 0.9093, + "step": 11789 + }, + { + "epoch": 0.8100865741376941, + "grad_norm": 1.1069697141647339, + "learning_rate": 5.295547956524519e-06, + "loss": 0.9837, + "step": 11790 + }, + { + "epoch": 0.8101552837707847, + "grad_norm": 1.0043467283248901, + "learning_rate": 5.2918581039300475e-06, + "loss": 0.9175, + "step": 11791 + }, + { + "epoch": 0.8102239934038752, + "grad_norm": 1.1024335622787476, + "learning_rate": 5.288169385132968e-06, + "loss": 0.9327, + "step": 11792 + }, + { + "epoch": 0.8102927030369658, + "grad_norm": 1.0726051330566406, + "learning_rate": 5.284481800345481e-06, + "loss": 0.9321, + "step": 11793 + }, + { + "epoch": 0.8103614126700563, + "grad_norm": 1.0647681951522827, + "learning_rate": 5.280795349779732e-06, + "loss": 0.9173, + "step": 11794 + }, + { + "epoch": 0.8104301223031469, + "grad_norm": 1.0416126251220703, + "learning_rate": 5.277110033647806e-06, + "loss": 0.9698, + "step": 11795 + }, + { + "epoch": 0.8104988319362375, + "grad_norm": 1.1865019798278809, + "learning_rate": 5.273425852161712e-06, + "loss": 0.96, + "step": 11796 + }, + { + "epoch": 0.810567541569328, + "grad_norm": 1.0427240133285522, + "learning_rate": 5.2697428055334096e-06, + "loss": 0.908, + "step": 11797 + }, + { + "epoch": 0.8106362512024186, + "grad_norm": 1.1040053367614746, + "learning_rate": 5.2660608939747644e-06, + "loss": 0.9847, + "step": 11798 + }, + { + "epoch": 0.8107049608355091, + "grad_norm": 1.0280555486679077, + "learning_rate": 5.262380117697605e-06, + "loss": 0.9526, + "step": 11799 + }, + { + "epoch": 0.8107736704685997, + "grad_norm": 1.058249831199646, + "learning_rate": 5.258700476913683e-06, + "loss": 0.9179, + "step": 11800 + }, + { + "epoch": 0.8108423801016903, + "grad_norm": 1.0452220439910889, + "learning_rate": 5.255021971834684e-06, + "loss": 0.9295, + "step": 11801 + }, + { + "epoch": 0.8109110897347808, + "grad_norm": 1.0576424598693848, + "learning_rate": 5.25134460267224e-06, + "loss": 0.9214, + "step": 11802 + }, + { + "epoch": 0.8109797993678713, + "grad_norm": 1.1258591413497925, + "learning_rate": 5.2476683696378926e-06, + "loss": 0.9427, + "step": 11803 + }, + { + "epoch": 0.8110485090009619, + "grad_norm": 1.0657877922058105, + "learning_rate": 5.24399327294314e-06, + "loss": 0.9214, + "step": 11804 + }, + { + "epoch": 0.8111172186340525, + "grad_norm": 1.0888946056365967, + "learning_rate": 5.240319312799408e-06, + "loss": 0.9526, + "step": 11805 + }, + { + "epoch": 0.8111859282671431, + "grad_norm": 1.0677868127822876, + "learning_rate": 5.236646489418059e-06, + "loss": 0.9539, + "step": 11806 + }, + { + "epoch": 0.8112546379002337, + "grad_norm": 1.0346927642822266, + "learning_rate": 5.232974803010393e-06, + "loss": 0.9789, + "step": 11807 + }, + { + "epoch": 0.8113233475333241, + "grad_norm": 1.0726794004440308, + "learning_rate": 5.229304253787626e-06, + "loss": 0.9248, + "step": 11808 + }, + { + "epoch": 0.8113920571664147, + "grad_norm": 1.0564165115356445, + "learning_rate": 5.225634841960933e-06, + "loss": 0.9194, + "step": 11809 + }, + { + "epoch": 0.8114607667995053, + "grad_norm": 1.0626370906829834, + "learning_rate": 5.22196656774141e-06, + "loss": 0.9021, + "step": 11810 + }, + { + "epoch": 0.8115294764325959, + "grad_norm": 1.0926014184951782, + "learning_rate": 5.21829943134009e-06, + "loss": 0.952, + "step": 11811 + }, + { + "epoch": 0.8115981860656865, + "grad_norm": 1.0048149824142456, + "learning_rate": 5.21463343296795e-06, + "loss": 0.8933, + "step": 11812 + }, + { + "epoch": 0.8116668956987769, + "grad_norm": 1.0444202423095703, + "learning_rate": 5.2109685728358755e-06, + "loss": 0.938, + "step": 11813 + }, + { + "epoch": 0.8117356053318675, + "grad_norm": 1.0276453495025635, + "learning_rate": 5.207304851154721e-06, + "loss": 0.9182, + "step": 11814 + }, + { + "epoch": 0.8118043149649581, + "grad_norm": 1.1249465942382812, + "learning_rate": 5.203642268135245e-06, + "loss": 0.9483, + "step": 11815 + }, + { + "epoch": 0.8118730245980487, + "grad_norm": 1.1059436798095703, + "learning_rate": 5.199980823988157e-06, + "loss": 0.9148, + "step": 11816 + }, + { + "epoch": 0.8119417342311392, + "grad_norm": 1.0523172616958618, + "learning_rate": 5.1963205189241085e-06, + "loss": 0.942, + "step": 11817 + }, + { + "epoch": 0.8120104438642297, + "grad_norm": 1.056654691696167, + "learning_rate": 5.192661353153655e-06, + "loss": 0.932, + "step": 11818 + }, + { + "epoch": 0.8120791534973203, + "grad_norm": 1.0391048192977905, + "learning_rate": 5.189003326887321e-06, + "loss": 0.9589, + "step": 11819 + }, + { + "epoch": 0.8121478631304109, + "grad_norm": 1.066387414932251, + "learning_rate": 5.185346440335542e-06, + "loss": 0.9381, + "step": 11820 + }, + { + "epoch": 0.8122165727635015, + "grad_norm": 1.137435793876648, + "learning_rate": 5.181690693708705e-06, + "loss": 0.9458, + "step": 11821 + }, + { + "epoch": 0.812285282396592, + "grad_norm": 1.1278939247131348, + "learning_rate": 5.178036087217122e-06, + "loss": 0.9333, + "step": 11822 + }, + { + "epoch": 0.8123539920296826, + "grad_norm": 1.0784903764724731, + "learning_rate": 5.174382621071031e-06, + "loss": 0.9639, + "step": 11823 + }, + { + "epoch": 0.8124227016627731, + "grad_norm": 1.0153985023498535, + "learning_rate": 5.170730295480619e-06, + "loss": 0.9085, + "step": 11824 + }, + { + "epoch": 0.8124914112958637, + "grad_norm": 1.0806618928909302, + "learning_rate": 5.167079110656003e-06, + "loss": 0.951, + "step": 11825 + }, + { + "epoch": 0.8125601209289542, + "grad_norm": 1.0449268817901611, + "learning_rate": 5.163429066807232e-06, + "loss": 0.9505, + "step": 11826 + }, + { + "epoch": 0.8126288305620448, + "grad_norm": 1.0992900133132935, + "learning_rate": 5.159780164144301e-06, + "loss": 0.9346, + "step": 11827 + }, + { + "epoch": 0.8126975401951354, + "grad_norm": 1.0424680709838867, + "learning_rate": 5.156132402877112e-06, + "loss": 0.9049, + "step": 11828 + }, + { + "epoch": 0.8127662498282259, + "grad_norm": 1.0613329410552979, + "learning_rate": 5.152485783215524e-06, + "loss": 0.9309, + "step": 11829 + }, + { + "epoch": 0.8128349594613165, + "grad_norm": 1.0591589212417603, + "learning_rate": 5.1488403053693315e-06, + "loss": 0.9275, + "step": 11830 + }, + { + "epoch": 0.812903669094407, + "grad_norm": 1.0173945426940918, + "learning_rate": 5.14519596954825e-06, + "loss": 0.8843, + "step": 11831 + }, + { + "epoch": 0.8129723787274976, + "grad_norm": 1.038020372390747, + "learning_rate": 5.141552775961938e-06, + "loss": 0.9305, + "step": 11832 + }, + { + "epoch": 0.8130410883605882, + "grad_norm": 1.0965884923934937, + "learning_rate": 5.137910724819994e-06, + "loss": 0.9225, + "step": 11833 + }, + { + "epoch": 0.8131097979936787, + "grad_norm": 1.0435410737991333, + "learning_rate": 5.134269816331927e-06, + "loss": 0.9251, + "step": 11834 + }, + { + "epoch": 0.8131785076267692, + "grad_norm": 1.0370362997055054, + "learning_rate": 5.130630050707208e-06, + "loss": 0.9504, + "step": 11835 + }, + { + "epoch": 0.8132472172598598, + "grad_norm": 1.0531879663467407, + "learning_rate": 5.1269914281552315e-06, + "loss": 0.922, + "step": 11836 + }, + { + "epoch": 0.8133159268929504, + "grad_norm": 1.0291657447814941, + "learning_rate": 5.123353948885315e-06, + "loss": 0.9267, + "step": 11837 + }, + { + "epoch": 0.813384636526041, + "grad_norm": 1.0033714771270752, + "learning_rate": 5.119717613106725e-06, + "loss": 0.8966, + "step": 11838 + }, + { + "epoch": 0.8134533461591316, + "grad_norm": 1.0076508522033691, + "learning_rate": 5.1160824210286675e-06, + "loss": 0.8891, + "step": 11839 + }, + { + "epoch": 0.813522055792222, + "grad_norm": 1.0634263753890991, + "learning_rate": 5.112448372860257e-06, + "loss": 0.9237, + "step": 11840 + }, + { + "epoch": 0.8135907654253126, + "grad_norm": 1.0747085809707642, + "learning_rate": 5.108815468810563e-06, + "loss": 0.9258, + "step": 11841 + }, + { + "epoch": 0.8136594750584032, + "grad_norm": 1.088746428489685, + "learning_rate": 5.10518370908859e-06, + "loss": 0.9439, + "step": 11842 + }, + { + "epoch": 0.8137281846914938, + "grad_norm": 1.0133808851242065, + "learning_rate": 5.101553093903275e-06, + "loss": 0.8912, + "step": 11843 + }, + { + "epoch": 0.8137968943245844, + "grad_norm": 1.0310932397842407, + "learning_rate": 5.097923623463468e-06, + "loss": 0.9207, + "step": 11844 + }, + { + "epoch": 0.8138656039576748, + "grad_norm": 1.099387764930725, + "learning_rate": 5.094295297977983e-06, + "loss": 0.9301, + "step": 11845 + }, + { + "epoch": 0.8139343135907654, + "grad_norm": 1.0533852577209473, + "learning_rate": 5.090668117655551e-06, + "loss": 0.9283, + "step": 11846 + }, + { + "epoch": 0.814003023223856, + "grad_norm": 1.0363335609436035, + "learning_rate": 5.08704208270484e-06, + "loss": 0.9272, + "step": 11847 + }, + { + "epoch": 0.8140717328569466, + "grad_norm": 1.0332539081573486, + "learning_rate": 5.0834171933344665e-06, + "loss": 0.9171, + "step": 11848 + }, + { + "epoch": 0.8141404424900371, + "grad_norm": 1.0350067615509033, + "learning_rate": 5.079793449752951e-06, + "loss": 0.8981, + "step": 11849 + }, + { + "epoch": 0.8142091521231276, + "grad_norm": 1.0458348989486694, + "learning_rate": 5.076170852168774e-06, + "loss": 0.9063, + "step": 11850 + }, + { + "epoch": 0.8142778617562182, + "grad_norm": 0.9846508502960205, + "learning_rate": 5.072549400790336e-06, + "loss": 0.9306, + "step": 11851 + }, + { + "epoch": 0.8143465713893088, + "grad_norm": 1.1670669317245483, + "learning_rate": 5.068929095825983e-06, + "loss": 0.951, + "step": 11852 + }, + { + "epoch": 0.8144152810223994, + "grad_norm": 1.013431191444397, + "learning_rate": 5.065309937483992e-06, + "loss": 0.8967, + "step": 11853 + }, + { + "epoch": 0.8144839906554899, + "grad_norm": 1.0928983688354492, + "learning_rate": 5.0616919259725586e-06, + "loss": 0.972, + "step": 11854 + }, + { + "epoch": 0.8145527002885805, + "grad_norm": 1.147799015045166, + "learning_rate": 5.058075061499834e-06, + "loss": 0.9277, + "step": 11855 + }, + { + "epoch": 0.814621409921671, + "grad_norm": 1.1020931005477905, + "learning_rate": 5.05445934427389e-06, + "loss": 0.9374, + "step": 11856 + }, + { + "epoch": 0.8146901195547616, + "grad_norm": 1.0248955488204956, + "learning_rate": 5.050844774502739e-06, + "loss": 0.958, + "step": 11857 + }, + { + "epoch": 0.8147588291878521, + "grad_norm": 0.9928145408630371, + "learning_rate": 5.04723135239433e-06, + "loss": 0.9012, + "step": 11858 + }, + { + "epoch": 0.8148275388209427, + "grad_norm": 1.0398377180099487, + "learning_rate": 5.043619078156528e-06, + "loss": 0.9258, + "step": 11859 + }, + { + "epoch": 0.8148962484540333, + "grad_norm": 1.0909353494644165, + "learning_rate": 5.040007951997156e-06, + "loss": 0.9738, + "step": 11860 + }, + { + "epoch": 0.8149649580871238, + "grad_norm": 1.0704090595245361, + "learning_rate": 5.03639797412396e-06, + "loss": 0.9147, + "step": 11861 + }, + { + "epoch": 0.8150336677202143, + "grad_norm": 1.092958688735962, + "learning_rate": 5.03278914474461e-06, + "loss": 0.9675, + "step": 11862 + }, + { + "epoch": 0.8151023773533049, + "grad_norm": 1.0475561618804932, + "learning_rate": 5.0291814640667315e-06, + "loss": 0.9349, + "step": 11863 + }, + { + "epoch": 0.8151710869863955, + "grad_norm": 1.0740195512771606, + "learning_rate": 5.025574932297861e-06, + "loss": 0.9492, + "step": 11864 + }, + { + "epoch": 0.8152397966194861, + "grad_norm": 1.0694596767425537, + "learning_rate": 5.0219695496454865e-06, + "loss": 0.9668, + "step": 11865 + }, + { + "epoch": 0.8153085062525766, + "grad_norm": 1.0615123510360718, + "learning_rate": 5.0183653163170225e-06, + "loss": 0.9258, + "step": 11866 + }, + { + "epoch": 0.8153772158856671, + "grad_norm": 1.0191987752914429, + "learning_rate": 5.014762232519818e-06, + "loss": 0.9361, + "step": 11867 + }, + { + "epoch": 0.8154459255187577, + "grad_norm": 1.0826938152313232, + "learning_rate": 5.011160298461162e-06, + "loss": 0.9422, + "step": 11868 + }, + { + "epoch": 0.8155146351518483, + "grad_norm": 1.0706044435501099, + "learning_rate": 5.007559514348259e-06, + "loss": 0.9168, + "step": 11869 + }, + { + "epoch": 0.8155833447849389, + "grad_norm": 1.146780252456665, + "learning_rate": 5.003959880388271e-06, + "loss": 0.9594, + "step": 11870 + }, + { + "epoch": 0.8156520544180295, + "grad_norm": 1.0358165502548218, + "learning_rate": 5.000361396788275e-06, + "loss": 0.9109, + "step": 11871 + }, + { + "epoch": 0.8157207640511199, + "grad_norm": 1.0497430562973022, + "learning_rate": 4.996764063755294e-06, + "loss": 0.9188, + "step": 11872 + }, + { + "epoch": 0.8157894736842105, + "grad_norm": 1.0311511754989624, + "learning_rate": 4.993167881496288e-06, + "loss": 0.9382, + "step": 11873 + }, + { + "epoch": 0.8158581833173011, + "grad_norm": 1.1617722511291504, + "learning_rate": 4.989572850218127e-06, + "loss": 0.9581, + "step": 11874 + }, + { + "epoch": 0.8159268929503917, + "grad_norm": 1.001004695892334, + "learning_rate": 4.9859789701276385e-06, + "loss": 0.9068, + "step": 11875 + }, + { + "epoch": 0.8159956025834822, + "grad_norm": 0.9895269870758057, + "learning_rate": 4.982386241431577e-06, + "loss": 0.9122, + "step": 11876 + }, + { + "epoch": 0.8160643122165727, + "grad_norm": 0.9781285524368286, + "learning_rate": 4.978794664336631e-06, + "loss": 0.9107, + "step": 11877 + }, + { + "epoch": 0.8161330218496633, + "grad_norm": 1.057621717453003, + "learning_rate": 4.9752042390494244e-06, + "loss": 0.9392, + "step": 11878 + }, + { + "epoch": 0.8162017314827539, + "grad_norm": 1.0212029218673706, + "learning_rate": 4.9716149657765055e-06, + "loss": 0.9078, + "step": 11879 + }, + { + "epoch": 0.8162704411158445, + "grad_norm": 1.0404725074768066, + "learning_rate": 4.968026844724363e-06, + "loss": 0.9273, + "step": 11880 + }, + { + "epoch": 0.816339150748935, + "grad_norm": 1.12155282497406, + "learning_rate": 4.964439876099422e-06, + "loss": 0.958, + "step": 11881 + }, + { + "epoch": 0.8164078603820255, + "grad_norm": 0.9620577096939087, + "learning_rate": 4.960854060108039e-06, + "loss": 0.9212, + "step": 11882 + }, + { + "epoch": 0.8164765700151161, + "grad_norm": 1.0817865133285522, + "learning_rate": 4.95726939695651e-06, + "loss": 0.9781, + "step": 11883 + }, + { + "epoch": 0.8165452796482067, + "grad_norm": 1.0122673511505127, + "learning_rate": 4.953685886851045e-06, + "loss": 0.8949, + "step": 11884 + }, + { + "epoch": 0.8166139892812972, + "grad_norm": 0.9999651908874512, + "learning_rate": 4.950103529997807e-06, + "loss": 0.9402, + "step": 11885 + }, + { + "epoch": 0.8166826989143878, + "grad_norm": 1.1276122331619263, + "learning_rate": 4.946522326602895e-06, + "loss": 0.9322, + "step": 11886 + }, + { + "epoch": 0.8167514085474784, + "grad_norm": 0.9988337755203247, + "learning_rate": 4.942942276872323e-06, + "loss": 0.8966, + "step": 11887 + }, + { + "epoch": 0.8168201181805689, + "grad_norm": 1.0182929039001465, + "learning_rate": 4.93936338101205e-06, + "loss": 0.9273, + "step": 11888 + }, + { + "epoch": 0.8168888278136595, + "grad_norm": 1.0343445539474487, + "learning_rate": 4.935785639227974e-06, + "loss": 0.9308, + "step": 11889 + }, + { + "epoch": 0.81695753744675, + "grad_norm": 1.0577243566513062, + "learning_rate": 4.932209051725914e-06, + "loss": 0.9278, + "step": 11890 + }, + { + "epoch": 0.8170262470798406, + "grad_norm": 1.0506763458251953, + "learning_rate": 4.928633618711628e-06, + "loss": 0.9307, + "step": 11891 + }, + { + "epoch": 0.8170949567129312, + "grad_norm": 1.0392130613327026, + "learning_rate": 4.925059340390814e-06, + "loss": 0.8966, + "step": 11892 + }, + { + "epoch": 0.8171636663460217, + "grad_norm": 1.0430225133895874, + "learning_rate": 4.921486216969093e-06, + "loss": 0.9048, + "step": 11893 + }, + { + "epoch": 0.8172323759791122, + "grad_norm": 1.042061686515808, + "learning_rate": 4.917914248652037e-06, + "loss": 0.9137, + "step": 11894 + }, + { + "epoch": 0.8173010856122028, + "grad_norm": 1.0761884450912476, + "learning_rate": 4.914343435645122e-06, + "loss": 0.9091, + "step": 11895 + }, + { + "epoch": 0.8173697952452934, + "grad_norm": 1.05536949634552, + "learning_rate": 4.910773778153782e-06, + "loss": 0.8977, + "step": 11896 + }, + { + "epoch": 0.817438504878384, + "grad_norm": 1.0347154140472412, + "learning_rate": 4.907205276383378e-06, + "loss": 0.9358, + "step": 11897 + }, + { + "epoch": 0.8175072145114745, + "grad_norm": 1.0592353343963623, + "learning_rate": 4.903637930539201e-06, + "loss": 0.9606, + "step": 11898 + }, + { + "epoch": 0.817575924144565, + "grad_norm": 1.123213291168213, + "learning_rate": 4.900071740826489e-06, + "loss": 0.9297, + "step": 11899 + }, + { + "epoch": 0.8176446337776556, + "grad_norm": 1.038284182548523, + "learning_rate": 4.896506707450388e-06, + "loss": 0.9232, + "step": 11900 + }, + { + "epoch": 0.8177133434107462, + "grad_norm": 1.0113102197647095, + "learning_rate": 4.892942830615999e-06, + "loss": 0.9014, + "step": 11901 + }, + { + "epoch": 0.8177820530438368, + "grad_norm": 1.111604928970337, + "learning_rate": 4.889380110528346e-06, + "loss": 0.9546, + "step": 11902 + }, + { + "epoch": 0.8178507626769274, + "grad_norm": 0.9819891452789307, + "learning_rate": 4.885818547392396e-06, + "loss": 0.8981, + "step": 11903 + }, + { + "epoch": 0.8179194723100178, + "grad_norm": 1.0084627866744995, + "learning_rate": 4.882258141413046e-06, + "loss": 0.9254, + "step": 11904 + }, + { + "epoch": 0.8179881819431084, + "grad_norm": 1.0258758068084717, + "learning_rate": 4.878698892795111e-06, + "loss": 0.9017, + "step": 11905 + }, + { + "epoch": 0.818056891576199, + "grad_norm": 1.0936237573623657, + "learning_rate": 4.8751408017433606e-06, + "loss": 0.9322, + "step": 11906 + }, + { + "epoch": 0.8181256012092896, + "grad_norm": 1.0713608264923096, + "learning_rate": 4.87158386846249e-06, + "loss": 0.8882, + "step": 11907 + }, + { + "epoch": 0.8181943108423801, + "grad_norm": 1.0364099740982056, + "learning_rate": 4.868028093157134e-06, + "loss": 0.9249, + "step": 11908 + }, + { + "epoch": 0.8182630204754706, + "grad_norm": 1.0660088062286377, + "learning_rate": 4.864473476031839e-06, + "loss": 0.9226, + "step": 11909 + }, + { + "epoch": 0.8183317301085612, + "grad_norm": 1.1184165477752686, + "learning_rate": 4.8609200172911155e-06, + "loss": 0.9466, + "step": 11910 + }, + { + "epoch": 0.8184004397416518, + "grad_norm": 1.1603642702102661, + "learning_rate": 4.857367717139374e-06, + "loss": 0.9467, + "step": 11911 + }, + { + "epoch": 0.8184691493747424, + "grad_norm": 1.049157738685608, + "learning_rate": 4.8538165757809895e-06, + "loss": 0.9027, + "step": 11912 + }, + { + "epoch": 0.8185378590078329, + "grad_norm": 1.1322287321090698, + "learning_rate": 4.850266593420252e-06, + "loss": 0.931, + "step": 11913 + }, + { + "epoch": 0.8186065686409234, + "grad_norm": 1.1248031854629517, + "learning_rate": 4.846717770261397e-06, + "loss": 0.9579, + "step": 11914 + }, + { + "epoch": 0.818675278274014, + "grad_norm": 1.0374780893325806, + "learning_rate": 4.843170106508577e-06, + "loss": 0.9259, + "step": 11915 + }, + { + "epoch": 0.8187439879071046, + "grad_norm": 1.0380712747573853, + "learning_rate": 4.839623602365886e-06, + "loss": 0.9392, + "step": 11916 + }, + { + "epoch": 0.8188126975401951, + "grad_norm": 1.1215009689331055, + "learning_rate": 4.836078258037361e-06, + "loss": 0.9672, + "step": 11917 + }, + { + "epoch": 0.8188814071732857, + "grad_norm": 1.0612379312515259, + "learning_rate": 4.832534073726958e-06, + "loss": 0.9426, + "step": 11918 + }, + { + "epoch": 0.8189501168063763, + "grad_norm": 1.040743350982666, + "learning_rate": 4.828991049638579e-06, + "loss": 0.9151, + "step": 11919 + }, + { + "epoch": 0.8190188264394668, + "grad_norm": 1.0512561798095703, + "learning_rate": 4.825449185976039e-06, + "loss": 0.9279, + "step": 11920 + }, + { + "epoch": 0.8190875360725574, + "grad_norm": 1.0923705101013184, + "learning_rate": 4.821908482943108e-06, + "loss": 0.956, + "step": 11921 + }, + { + "epoch": 0.8191562457056479, + "grad_norm": 1.0216012001037598, + "learning_rate": 4.8183689407434754e-06, + "loss": 0.9204, + "step": 11922 + }, + { + "epoch": 0.8192249553387385, + "grad_norm": 1.106872320175171, + "learning_rate": 4.814830559580774e-06, + "loss": 0.9669, + "step": 11923 + }, + { + "epoch": 0.8192936649718291, + "grad_norm": 0.9780815839767456, + "learning_rate": 4.811293339658571e-06, + "loss": 0.8876, + "step": 11924 + }, + { + "epoch": 0.8193623746049196, + "grad_norm": 1.1121786832809448, + "learning_rate": 4.807757281180345e-06, + "loss": 0.9319, + "step": 11925 + }, + { + "epoch": 0.8194310842380101, + "grad_norm": 1.0174258947372437, + "learning_rate": 4.80422238434953e-06, + "loss": 0.8968, + "step": 11926 + }, + { + "epoch": 0.8194997938711007, + "grad_norm": 1.0716726779937744, + "learning_rate": 4.800688649369489e-06, + "loss": 0.9162, + "step": 11927 + }, + { + "epoch": 0.8195685035041913, + "grad_norm": 0.9983399510383606, + "learning_rate": 4.797156076443512e-06, + "loss": 0.9023, + "step": 11928 + }, + { + "epoch": 0.8196372131372819, + "grad_norm": 1.0339381694793701, + "learning_rate": 4.7936246657748346e-06, + "loss": 0.899, + "step": 11929 + }, + { + "epoch": 0.8197059227703724, + "grad_norm": 1.0977448225021362, + "learning_rate": 4.790094417566601e-06, + "loss": 0.9501, + "step": 11930 + }, + { + "epoch": 0.8197746324034629, + "grad_norm": 1.0404492616653442, + "learning_rate": 4.786565332021917e-06, + "loss": 0.9648, + "step": 11931 + }, + { + "epoch": 0.8198433420365535, + "grad_norm": 1.0535337924957275, + "learning_rate": 4.78303740934381e-06, + "loss": 0.967, + "step": 11932 + }, + { + "epoch": 0.8199120516696441, + "grad_norm": 1.0447416305541992, + "learning_rate": 4.779510649735225e-06, + "loss": 0.9245, + "step": 11933 + }, + { + "epoch": 0.8199807613027347, + "grad_norm": 1.018128752708435, + "learning_rate": 4.775985053399066e-06, + "loss": 0.9185, + "step": 11934 + }, + { + "epoch": 0.8200494709358253, + "grad_norm": 1.1540995836257935, + "learning_rate": 4.77246062053816e-06, + "loss": 0.9574, + "step": 11935 + }, + { + "epoch": 0.8201181805689157, + "grad_norm": 1.0896707773208618, + "learning_rate": 4.768937351355254e-06, + "loss": 0.9055, + "step": 11936 + }, + { + "epoch": 0.8201868902020063, + "grad_norm": 1.1001226902008057, + "learning_rate": 4.765415246053051e-06, + "loss": 0.9497, + "step": 11937 + }, + { + "epoch": 0.8202555998350969, + "grad_norm": 1.1269659996032715, + "learning_rate": 4.761894304834166e-06, + "loss": 0.9526, + "step": 11938 + }, + { + "epoch": 0.8203243094681875, + "grad_norm": 1.0373601913452148, + "learning_rate": 4.7583745279011716e-06, + "loss": 0.9442, + "step": 11939 + }, + { + "epoch": 0.820393019101278, + "grad_norm": 1.1042958498001099, + "learning_rate": 4.754855915456541e-06, + "loss": 0.9265, + "step": 11940 + }, + { + "epoch": 0.8204617287343685, + "grad_norm": 1.0517975091934204, + "learning_rate": 4.751338467702707e-06, + "loss": 0.9379, + "step": 11941 + }, + { + "epoch": 0.8205304383674591, + "grad_norm": 1.0549298524856567, + "learning_rate": 4.747822184842022e-06, + "loss": 0.9438, + "step": 11942 + }, + { + "epoch": 0.8205991480005497, + "grad_norm": 1.04336416721344, + "learning_rate": 4.744307067076781e-06, + "loss": 0.9504, + "step": 11943 + }, + { + "epoch": 0.8206678576336403, + "grad_norm": 1.0431169271469116, + "learning_rate": 4.740793114609202e-06, + "loss": 0.9201, + "step": 11944 + }, + { + "epoch": 0.8207365672667308, + "grad_norm": 1.0414525270462036, + "learning_rate": 4.737280327641449e-06, + "loss": 0.9424, + "step": 11945 + }, + { + "epoch": 0.8208052768998213, + "grad_norm": 1.0121171474456787, + "learning_rate": 4.7337687063755965e-06, + "loss": 0.9138, + "step": 11946 + }, + { + "epoch": 0.8208739865329119, + "grad_norm": 1.0782963037490845, + "learning_rate": 4.730258251013675e-06, + "loss": 0.9012, + "step": 11947 + }, + { + "epoch": 0.8209426961660025, + "grad_norm": 1.1272777318954468, + "learning_rate": 4.726748961757638e-06, + "loss": 0.9443, + "step": 11948 + }, + { + "epoch": 0.821011405799093, + "grad_norm": 1.0385048389434814, + "learning_rate": 4.7232408388093705e-06, + "loss": 0.9194, + "step": 11949 + }, + { + "epoch": 0.8210801154321836, + "grad_norm": 0.9900386333465576, + "learning_rate": 4.7197338823706995e-06, + "loss": 0.9071, + "step": 11950 + }, + { + "epoch": 0.8211488250652742, + "grad_norm": 1.076486587524414, + "learning_rate": 4.716228092643369e-06, + "loss": 0.9691, + "step": 11951 + }, + { + "epoch": 0.8212175346983647, + "grad_norm": 1.0454224348068237, + "learning_rate": 4.712723469829067e-06, + "loss": 0.9799, + "step": 11952 + }, + { + "epoch": 0.8212862443314553, + "grad_norm": 1.0385586023330688, + "learning_rate": 4.7092200141294145e-06, + "loss": 0.9474, + "step": 11953 + }, + { + "epoch": 0.8213549539645458, + "grad_norm": 1.0257585048675537, + "learning_rate": 4.705717725745967e-06, + "loss": 0.9002, + "step": 11954 + }, + { + "epoch": 0.8214236635976364, + "grad_norm": 1.0778793096542358, + "learning_rate": 4.702216604880202e-06, + "loss": 0.914, + "step": 11955 + }, + { + "epoch": 0.821492373230727, + "grad_norm": 1.0597443580627441, + "learning_rate": 4.698716651733536e-06, + "loss": 0.9457, + "step": 11956 + }, + { + "epoch": 0.8215610828638175, + "grad_norm": 1.0191400051116943, + "learning_rate": 4.695217866507332e-06, + "loss": 0.9432, + "step": 11957 + }, + { + "epoch": 0.821629792496908, + "grad_norm": 1.0711218118667603, + "learning_rate": 4.691720249402856e-06, + "loss": 0.9423, + "step": 11958 + }, + { + "epoch": 0.8216985021299986, + "grad_norm": 1.005279302597046, + "learning_rate": 4.68822380062133e-06, + "loss": 0.9189, + "step": 11959 + }, + { + "epoch": 0.8217672117630892, + "grad_norm": 1.0733757019042969, + "learning_rate": 4.684728520363913e-06, + "loss": 0.979, + "step": 11960 + }, + { + "epoch": 0.8218359213961798, + "grad_norm": 1.0552107095718384, + "learning_rate": 4.68123440883167e-06, + "loss": 0.9367, + "step": 11961 + }, + { + "epoch": 0.8219046310292702, + "grad_norm": 1.0378472805023193, + "learning_rate": 4.677741466225621e-06, + "loss": 0.9513, + "step": 11962 + }, + { + "epoch": 0.8219733406623608, + "grad_norm": 1.138669729232788, + "learning_rate": 4.674249692746716e-06, + "loss": 0.9448, + "step": 11963 + }, + { + "epoch": 0.8220420502954514, + "grad_norm": 1.05330228805542, + "learning_rate": 4.67075908859583e-06, + "loss": 0.936, + "step": 11964 + }, + { + "epoch": 0.822110759928542, + "grad_norm": 1.0664228200912476, + "learning_rate": 4.667269653973788e-06, + "loss": 0.9143, + "step": 11965 + }, + { + "epoch": 0.8221794695616326, + "grad_norm": 1.007547378540039, + "learning_rate": 4.663781389081318e-06, + "loss": 0.8988, + "step": 11966 + }, + { + "epoch": 0.8222481791947231, + "grad_norm": 1.1166568994522095, + "learning_rate": 4.660294294119102e-06, + "loss": 0.9684, + "step": 11967 + }, + { + "epoch": 0.8223168888278136, + "grad_norm": 1.0548917055130005, + "learning_rate": 4.656808369287758e-06, + "loss": 0.9324, + "step": 11968 + }, + { + "epoch": 0.8223855984609042, + "grad_norm": 1.0765002965927124, + "learning_rate": 4.653323614787821e-06, + "loss": 0.9194, + "step": 11969 + }, + { + "epoch": 0.8224543080939948, + "grad_norm": 1.0432329177856445, + "learning_rate": 4.64984003081978e-06, + "loss": 0.9108, + "step": 11970 + }, + { + "epoch": 0.8225230177270854, + "grad_norm": 1.0909308195114136, + "learning_rate": 4.646357617584027e-06, + "loss": 0.9214, + "step": 11971 + }, + { + "epoch": 0.8225917273601759, + "grad_norm": 0.9721246957778931, + "learning_rate": 4.642876375280908e-06, + "loss": 0.9046, + "step": 11972 + }, + { + "epoch": 0.8226604369932664, + "grad_norm": 1.0956237316131592, + "learning_rate": 4.639396304110699e-06, + "loss": 0.9392, + "step": 11973 + }, + { + "epoch": 0.822729146626357, + "grad_norm": 1.1305361986160278, + "learning_rate": 4.635917404273607e-06, + "loss": 0.9517, + "step": 11974 + }, + { + "epoch": 0.8227978562594476, + "grad_norm": 1.0417087078094482, + "learning_rate": 4.632439675969777e-06, + "loss": 0.8958, + "step": 11975 + }, + { + "epoch": 0.8228665658925381, + "grad_norm": 1.0303623676300049, + "learning_rate": 4.628963119399266e-06, + "loss": 0.9194, + "step": 11976 + }, + { + "epoch": 0.8229352755256287, + "grad_norm": 1.0361160039901733, + "learning_rate": 4.6254877347620855e-06, + "loss": 0.9134, + "step": 11977 + }, + { + "epoch": 0.8230039851587192, + "grad_norm": 1.057433843612671, + "learning_rate": 4.622013522258173e-06, + "loss": 0.934, + "step": 11978 + }, + { + "epoch": 0.8230726947918098, + "grad_norm": 1.071739673614502, + "learning_rate": 4.618540482087405e-06, + "loss": 0.9007, + "step": 11979 + }, + { + "epoch": 0.8231414044249004, + "grad_norm": 1.1177664995193481, + "learning_rate": 4.6150686144495695e-06, + "loss": 0.9372, + "step": 11980 + }, + { + "epoch": 0.8232101140579909, + "grad_norm": 1.0583152770996094, + "learning_rate": 4.611597919544413e-06, + "loss": 0.9348, + "step": 11981 + }, + { + "epoch": 0.8232788236910815, + "grad_norm": 1.1549025774002075, + "learning_rate": 4.60812839757159e-06, + "loss": 0.9332, + "step": 11982 + }, + { + "epoch": 0.8233475333241721, + "grad_norm": 1.032210111618042, + "learning_rate": 4.604660048730705e-06, + "loss": 0.9362, + "step": 11983 + }, + { + "epoch": 0.8234162429572626, + "grad_norm": 1.0282185077667236, + "learning_rate": 4.601192873221294e-06, + "loss": 0.9238, + "step": 11984 + }, + { + "epoch": 0.8234849525903531, + "grad_norm": 1.0669124126434326, + "learning_rate": 4.597726871242827e-06, + "loss": 0.9302, + "step": 11985 + }, + { + "epoch": 0.8235536622234437, + "grad_norm": 1.0341583490371704, + "learning_rate": 4.594262042994685e-06, + "loss": 0.9152, + "step": 11986 + }, + { + "epoch": 0.8236223718565343, + "grad_norm": 1.0457125902175903, + "learning_rate": 4.590798388676204e-06, + "loss": 0.9335, + "step": 11987 + }, + { + "epoch": 0.8236910814896249, + "grad_norm": 1.0592224597930908, + "learning_rate": 4.58733590848665e-06, + "loss": 0.9321, + "step": 11988 + }, + { + "epoch": 0.8237597911227154, + "grad_norm": 1.0309945344924927, + "learning_rate": 4.583874602625213e-06, + "loss": 0.9005, + "step": 11989 + }, + { + "epoch": 0.8238285007558059, + "grad_norm": 1.035056233406067, + "learning_rate": 4.580414471291028e-06, + "loss": 0.905, + "step": 11990 + }, + { + "epoch": 0.8238972103888965, + "grad_norm": 1.0258595943450928, + "learning_rate": 4.576955514683143e-06, + "loss": 0.9431, + "step": 11991 + }, + { + "epoch": 0.8239659200219871, + "grad_norm": 1.0137251615524292, + "learning_rate": 4.573497733000556e-06, + "loss": 0.9434, + "step": 11992 + }, + { + "epoch": 0.8240346296550777, + "grad_norm": 1.0685219764709473, + "learning_rate": 4.570041126442187e-06, + "loss": 0.9227, + "step": 11993 + }, + { + "epoch": 0.8241033392881681, + "grad_norm": 1.1143629550933838, + "learning_rate": 4.566585695206896e-06, + "loss": 0.9689, + "step": 11994 + }, + { + "epoch": 0.8241720489212587, + "grad_norm": 1.0239757299423218, + "learning_rate": 4.56313143949347e-06, + "loss": 0.9047, + "step": 11995 + }, + { + "epoch": 0.8242407585543493, + "grad_norm": 1.111519455909729, + "learning_rate": 4.559678359500641e-06, + "loss": 0.9389, + "step": 11996 + }, + { + "epoch": 0.8243094681874399, + "grad_norm": 1.0728567838668823, + "learning_rate": 4.556226455427043e-06, + "loss": 0.957, + "step": 11997 + }, + { + "epoch": 0.8243781778205305, + "grad_norm": 1.1283117532730103, + "learning_rate": 4.552775727471273e-06, + "loss": 0.9626, + "step": 11998 + }, + { + "epoch": 0.824446887453621, + "grad_norm": 1.0368008613586426, + "learning_rate": 4.549326175831847e-06, + "loss": 0.9078, + "step": 11999 + }, + { + "epoch": 0.8245155970867115, + "grad_norm": 1.0766849517822266, + "learning_rate": 4.545877800707218e-06, + "loss": 0.9311, + "step": 12000 + }, + { + "epoch": 0.8245843067198021, + "grad_norm": 1.0565588474273682, + "learning_rate": 4.542430602295774e-06, + "loss": 0.9282, + "step": 12001 + }, + { + "epoch": 0.8246530163528927, + "grad_norm": 1.1060432195663452, + "learning_rate": 4.538984580795818e-06, + "loss": 0.9693, + "step": 12002 + }, + { + "epoch": 0.8247217259859833, + "grad_norm": 1.0535411834716797, + "learning_rate": 4.535539736405609e-06, + "loss": 0.9456, + "step": 12003 + }, + { + "epoch": 0.8247904356190738, + "grad_norm": 1.0960757732391357, + "learning_rate": 4.5320960693233136e-06, + "loss": 0.927, + "step": 12004 + }, + { + "epoch": 0.8248591452521643, + "grad_norm": 1.0461066961288452, + "learning_rate": 4.528653579747052e-06, + "loss": 0.9647, + "step": 12005 + }, + { + "epoch": 0.8249278548852549, + "grad_norm": 1.157779574394226, + "learning_rate": 4.525212267874876e-06, + "loss": 0.9912, + "step": 12006 + }, + { + "epoch": 0.8249965645183455, + "grad_norm": 1.0784839391708374, + "learning_rate": 4.521772133904745e-06, + "loss": 0.9381, + "step": 12007 + }, + { + "epoch": 0.825065274151436, + "grad_norm": 1.092513918876648, + "learning_rate": 4.518333178034579e-06, + "loss": 0.9395, + "step": 12008 + }, + { + "epoch": 0.8251339837845266, + "grad_norm": 1.2224390506744385, + "learning_rate": 4.514895400462216e-06, + "loss": 0.9798, + "step": 12009 + }, + { + "epoch": 0.8252026934176171, + "grad_norm": 1.1487938165664673, + "learning_rate": 4.5114588013854295e-06, + "loss": 0.968, + "step": 12010 + }, + { + "epoch": 0.8252714030507077, + "grad_norm": 1.0416693687438965, + "learning_rate": 4.508023381001933e-06, + "loss": 0.9591, + "step": 12011 + }, + { + "epoch": 0.8253401126837983, + "grad_norm": 1.026519775390625, + "learning_rate": 4.504589139509349e-06, + "loss": 0.8976, + "step": 12012 + }, + { + "epoch": 0.8254088223168888, + "grad_norm": 1.0577515363693237, + "learning_rate": 4.5011560771052566e-06, + "loss": 0.9521, + "step": 12013 + }, + { + "epoch": 0.8254775319499794, + "grad_norm": 1.0409061908721924, + "learning_rate": 4.497724193987157e-06, + "loss": 0.9517, + "step": 12014 + }, + { + "epoch": 0.82554624158307, + "grad_norm": 1.0058988332748413, + "learning_rate": 4.494293490352483e-06, + "loss": 0.934, + "step": 12015 + }, + { + "epoch": 0.8256149512161605, + "grad_norm": 1.0437754392623901, + "learning_rate": 4.4908639663986065e-06, + "loss": 0.9452, + "step": 12016 + }, + { + "epoch": 0.825683660849251, + "grad_norm": 1.0622509717941284, + "learning_rate": 4.487435622322814e-06, + "loss": 0.9177, + "step": 12017 + }, + { + "epoch": 0.8257523704823416, + "grad_norm": 1.0894349813461304, + "learning_rate": 4.484008458322345e-06, + "loss": 0.9433, + "step": 12018 + }, + { + "epoch": 0.8258210801154322, + "grad_norm": 1.0570003986358643, + "learning_rate": 4.4805824745943585e-06, + "loss": 0.913, + "step": 12019 + }, + { + "epoch": 0.8258897897485228, + "grad_norm": 1.0688834190368652, + "learning_rate": 4.477157671335952e-06, + "loss": 0.9251, + "step": 12020 + }, + { + "epoch": 0.8259584993816133, + "grad_norm": 1.0497019290924072, + "learning_rate": 4.473734048744157e-06, + "loss": 0.9319, + "step": 12021 + }, + { + "epoch": 0.8260272090147038, + "grad_norm": 1.0550751686096191, + "learning_rate": 4.470311607015923e-06, + "loss": 0.9583, + "step": 12022 + }, + { + "epoch": 0.8260959186477944, + "grad_norm": 1.0380605459213257, + "learning_rate": 4.466890346348141e-06, + "loss": 0.9424, + "step": 12023 + }, + { + "epoch": 0.826164628280885, + "grad_norm": 1.133407473564148, + "learning_rate": 4.463470266937638e-06, + "loss": 0.9389, + "step": 12024 + }, + { + "epoch": 0.8262333379139756, + "grad_norm": 1.0402569770812988, + "learning_rate": 4.460051368981177e-06, + "loss": 0.9304, + "step": 12025 + }, + { + "epoch": 0.8263020475470662, + "grad_norm": 1.0881638526916504, + "learning_rate": 4.4566336526754295e-06, + "loss": 0.9055, + "step": 12026 + }, + { + "epoch": 0.8263707571801566, + "grad_norm": 1.0522682666778564, + "learning_rate": 4.4532171182170205e-06, + "loss": 0.8759, + "step": 12027 + }, + { + "epoch": 0.8264394668132472, + "grad_norm": 1.0761957168579102, + "learning_rate": 4.449801765802511e-06, + "loss": 0.9287, + "step": 12028 + }, + { + "epoch": 0.8265081764463378, + "grad_norm": 1.0472270250320435, + "learning_rate": 4.446387595628371e-06, + "loss": 0.9275, + "step": 12029 + }, + { + "epoch": 0.8265768860794284, + "grad_norm": 1.063371181488037, + "learning_rate": 4.442974607891018e-06, + "loss": 0.9579, + "step": 12030 + }, + { + "epoch": 0.8266455957125189, + "grad_norm": 1.0174589157104492, + "learning_rate": 4.4395628027868066e-06, + "loss": 0.934, + "step": 12031 + }, + { + "epoch": 0.8267143053456094, + "grad_norm": 1.0278918743133545, + "learning_rate": 4.4361521805120074e-06, + "loss": 0.9051, + "step": 12032 + }, + { + "epoch": 0.8267830149787, + "grad_norm": 1.0576039552688599, + "learning_rate": 4.4327427412628325e-06, + "loss": 0.9119, + "step": 12033 + }, + { + "epoch": 0.8268517246117906, + "grad_norm": 1.039810061454773, + "learning_rate": 4.429334485235428e-06, + "loss": 0.9408, + "step": 12034 + }, + { + "epoch": 0.8269204342448812, + "grad_norm": 1.0601913928985596, + "learning_rate": 4.425927412625869e-06, + "loss": 0.9471, + "step": 12035 + }, + { + "epoch": 0.8269891438779717, + "grad_norm": 1.0275533199310303, + "learning_rate": 4.422521523630166e-06, + "loss": 0.9235, + "step": 12036 + }, + { + "epoch": 0.8270578535110622, + "grad_norm": 1.0179531574249268, + "learning_rate": 4.419116818444246e-06, + "loss": 0.9097, + "step": 12037 + }, + { + "epoch": 0.8271265631441528, + "grad_norm": 1.0749953985214233, + "learning_rate": 4.415713297263987e-06, + "loss": 0.9541, + "step": 12038 + }, + { + "epoch": 0.8271952727772434, + "grad_norm": 1.0630072355270386, + "learning_rate": 4.4123109602851935e-06, + "loss": 0.9267, + "step": 12039 + }, + { + "epoch": 0.8272639824103339, + "grad_norm": 1.0768604278564453, + "learning_rate": 4.4089098077035955e-06, + "loss": 0.9531, + "step": 12040 + }, + { + "epoch": 0.8273326920434245, + "grad_norm": 1.0274070501327515, + "learning_rate": 4.405509839714869e-06, + "loss": 0.9212, + "step": 12041 + }, + { + "epoch": 0.8274014016765151, + "grad_norm": 1.0060136318206787, + "learning_rate": 4.402111056514596e-06, + "loss": 0.8957, + "step": 12042 + }, + { + "epoch": 0.8274701113096056, + "grad_norm": 1.013948678970337, + "learning_rate": 4.398713458298315e-06, + "loss": 0.9083, + "step": 12043 + }, + { + "epoch": 0.8275388209426962, + "grad_norm": 1.0459071397781372, + "learning_rate": 4.3953170452614875e-06, + "loss": 0.9346, + "step": 12044 + }, + { + "epoch": 0.8276075305757867, + "grad_norm": 1.0612406730651855, + "learning_rate": 4.391921817599509e-06, + "loss": 0.8863, + "step": 12045 + }, + { + "epoch": 0.8276762402088773, + "grad_norm": 1.139643669128418, + "learning_rate": 4.388527775507706e-06, + "loss": 0.9458, + "step": 12046 + }, + { + "epoch": 0.8277449498419679, + "grad_norm": 1.049705982208252, + "learning_rate": 4.38513491918133e-06, + "loss": 0.9546, + "step": 12047 + }, + { + "epoch": 0.8278136594750584, + "grad_norm": 1.1262929439544678, + "learning_rate": 4.38174324881557e-06, + "loss": 0.9845, + "step": 12048 + }, + { + "epoch": 0.8278823691081489, + "grad_norm": 1.0726040601730347, + "learning_rate": 4.37835276460555e-06, + "loss": 0.9224, + "step": 12049 + }, + { + "epoch": 0.8279510787412395, + "grad_norm": 1.0786081552505493, + "learning_rate": 4.3749634667463265e-06, + "loss": 0.9303, + "step": 12050 + }, + { + "epoch": 0.8280197883743301, + "grad_norm": 1.066711187362671, + "learning_rate": 4.371575355432875e-06, + "loss": 0.9632, + "step": 12051 + }, + { + "epoch": 0.8280884980074207, + "grad_norm": 1.0598795413970947, + "learning_rate": 4.368188430860115e-06, + "loss": 0.9736, + "step": 12052 + }, + { + "epoch": 0.8281572076405112, + "grad_norm": 1.093002438545227, + "learning_rate": 4.3648026932229e-06, + "loss": 0.8871, + "step": 12053 + }, + { + "epoch": 0.8282259172736017, + "grad_norm": 1.06345796585083, + "learning_rate": 4.361418142715998e-06, + "loss": 0.9395, + "step": 12054 + }, + { + "epoch": 0.8282946269066923, + "grad_norm": 1.0990221500396729, + "learning_rate": 4.358034779534126e-06, + "loss": 0.9966, + "step": 12055 + }, + { + "epoch": 0.8283633365397829, + "grad_norm": 1.0258439779281616, + "learning_rate": 4.354652603871928e-06, + "loss": 0.911, + "step": 12056 + }, + { + "epoch": 0.8284320461728735, + "grad_norm": 1.0974289178848267, + "learning_rate": 4.351271615923983e-06, + "loss": 0.9569, + "step": 12057 + }, + { + "epoch": 0.828500755805964, + "grad_norm": 1.0985382795333862, + "learning_rate": 4.347891815884786e-06, + "loss": 0.9475, + "step": 12058 + }, + { + "epoch": 0.8285694654390545, + "grad_norm": 1.1332670450210571, + "learning_rate": 4.3445132039487805e-06, + "loss": 0.9287, + "step": 12059 + }, + { + "epoch": 0.8286381750721451, + "grad_norm": 1.1082981824874878, + "learning_rate": 4.341135780310337e-06, + "loss": 0.9351, + "step": 12060 + }, + { + "epoch": 0.8287068847052357, + "grad_norm": 1.0509461164474487, + "learning_rate": 4.337759545163753e-06, + "loss": 0.9389, + "step": 12061 + }, + { + "epoch": 0.8287755943383263, + "grad_norm": 1.018387794494629, + "learning_rate": 4.334384498703273e-06, + "loss": 0.9029, + "step": 12062 + }, + { + "epoch": 0.8288443039714168, + "grad_norm": 1.03978431224823, + "learning_rate": 4.331010641123046e-06, + "loss": 0.9161, + "step": 12063 + }, + { + "epoch": 0.8289130136045073, + "grad_norm": 1.098077416419983, + "learning_rate": 4.327637972617171e-06, + "loss": 0.9643, + "step": 12064 + }, + { + "epoch": 0.8289817232375979, + "grad_norm": 1.1309655904769897, + "learning_rate": 4.3242664933796806e-06, + "loss": 0.9267, + "step": 12065 + }, + { + "epoch": 0.8290504328706885, + "grad_norm": 1.0438659191131592, + "learning_rate": 4.320896203604533e-06, + "loss": 0.9059, + "step": 12066 + }, + { + "epoch": 0.829119142503779, + "grad_norm": 1.0840532779693604, + "learning_rate": 4.317527103485625e-06, + "loss": 0.922, + "step": 12067 + }, + { + "epoch": 0.8291878521368696, + "grad_norm": 1.0298969745635986, + "learning_rate": 4.314159193216763e-06, + "loss": 0.945, + "step": 12068 + }, + { + "epoch": 0.8292565617699601, + "grad_norm": 1.0779303312301636, + "learning_rate": 4.310792472991712e-06, + "loss": 0.944, + "step": 12069 + }, + { + "epoch": 0.8293252714030507, + "grad_norm": 1.0276541709899902, + "learning_rate": 4.3074269430041535e-06, + "loss": 0.9425, + "step": 12070 + }, + { + "epoch": 0.8293939810361413, + "grad_norm": 1.0326775312423706, + "learning_rate": 4.304062603447706e-06, + "loss": 0.9338, + "step": 12071 + }, + { + "epoch": 0.8294626906692318, + "grad_norm": 1.131408452987671, + "learning_rate": 4.300699454515927e-06, + "loss": 0.9327, + "step": 12072 + }, + { + "epoch": 0.8295314003023224, + "grad_norm": 0.9774366617202759, + "learning_rate": 4.297337496402279e-06, + "loss": 0.8943, + "step": 12073 + }, + { + "epoch": 0.829600109935413, + "grad_norm": 1.0542562007904053, + "learning_rate": 4.2939767293001805e-06, + "loss": 0.9347, + "step": 12074 + }, + { + "epoch": 0.8296688195685035, + "grad_norm": 1.036836862564087, + "learning_rate": 4.290617153402984e-06, + "loss": 0.9063, + "step": 12075 + }, + { + "epoch": 0.829737529201594, + "grad_norm": 1.039897084236145, + "learning_rate": 4.2872587689039484e-06, + "loss": 0.9184, + "step": 12076 + }, + { + "epoch": 0.8298062388346846, + "grad_norm": 1.015135407447815, + "learning_rate": 4.283901575996294e-06, + "loss": 0.9293, + "step": 12077 + }, + { + "epoch": 0.8298749484677752, + "grad_norm": 1.0299798250198364, + "learning_rate": 4.280545574873143e-06, + "loss": 0.8936, + "step": 12078 + }, + { + "epoch": 0.8299436581008658, + "grad_norm": 1.094821572303772, + "learning_rate": 4.2771907657275706e-06, + "loss": 0.9602, + "step": 12079 + }, + { + "epoch": 0.8300123677339563, + "grad_norm": 1.0462734699249268, + "learning_rate": 4.27383714875258e-06, + "loss": 0.9337, + "step": 12080 + }, + { + "epoch": 0.8300810773670468, + "grad_norm": 1.0839487314224243, + "learning_rate": 4.2704847241411e-06, + "loss": 0.9298, + "step": 12081 + }, + { + "epoch": 0.8301497870001374, + "grad_norm": 1.1152713298797607, + "learning_rate": 4.267133492086001e-06, + "loss": 0.9361, + "step": 12082 + }, + { + "epoch": 0.830218496633228, + "grad_norm": 1.0404138565063477, + "learning_rate": 4.263783452780065e-06, + "loss": 0.9432, + "step": 12083 + }, + { + "epoch": 0.8302872062663186, + "grad_norm": 1.0535187721252441, + "learning_rate": 4.260434606416022e-06, + "loss": 0.8851, + "step": 12084 + }, + { + "epoch": 0.830355915899409, + "grad_norm": 0.9917646646499634, + "learning_rate": 4.257086953186532e-06, + "loss": 0.8908, + "step": 12085 + }, + { + "epoch": 0.8304246255324996, + "grad_norm": 1.0909714698791504, + "learning_rate": 4.25374049328418e-06, + "loss": 0.9137, + "step": 12086 + }, + { + "epoch": 0.8304933351655902, + "grad_norm": 1.101514220237732, + "learning_rate": 4.2503952269014966e-06, + "loss": 0.9104, + "step": 12087 + }, + { + "epoch": 0.8305620447986808, + "grad_norm": 1.0343421697616577, + "learning_rate": 4.2470511542309164e-06, + "loss": 0.9243, + "step": 12088 + }, + { + "epoch": 0.8306307544317714, + "grad_norm": 1.0496357679367065, + "learning_rate": 4.243708275464828e-06, + "loss": 0.9411, + "step": 12089 + }, + { + "epoch": 0.830699464064862, + "grad_norm": 1.1249151229858398, + "learning_rate": 4.24036659079555e-06, + "loss": 0.9444, + "step": 12090 + }, + { + "epoch": 0.8307681736979524, + "grad_norm": 1.0477831363677979, + "learning_rate": 4.237026100415323e-06, + "loss": 0.9054, + "step": 12091 + }, + { + "epoch": 0.830836883331043, + "grad_norm": 1.0354143381118774, + "learning_rate": 4.233686804516329e-06, + "loss": 0.9231, + "step": 12092 + }, + { + "epoch": 0.8309055929641336, + "grad_norm": 1.0177602767944336, + "learning_rate": 4.230348703290665e-06, + "loss": 0.9128, + "step": 12093 + }, + { + "epoch": 0.8309743025972242, + "grad_norm": 1.1013202667236328, + "learning_rate": 4.227011796930377e-06, + "loss": 0.9506, + "step": 12094 + }, + { + "epoch": 0.8310430122303147, + "grad_norm": 1.0574804544448853, + "learning_rate": 4.223676085627437e-06, + "loss": 0.9465, + "step": 12095 + }, + { + "epoch": 0.8311117218634052, + "grad_norm": 1.0765184164047241, + "learning_rate": 4.220341569573741e-06, + "loss": 0.9541, + "step": 12096 + }, + { + "epoch": 0.8311804314964958, + "grad_norm": 0.9907115697860718, + "learning_rate": 4.217008248961132e-06, + "loss": 0.9051, + "step": 12097 + }, + { + "epoch": 0.8312491411295864, + "grad_norm": 1.1166164875030518, + "learning_rate": 4.213676123981358e-06, + "loss": 0.9416, + "step": 12098 + }, + { + "epoch": 0.831317850762677, + "grad_norm": 1.065187931060791, + "learning_rate": 4.2103451948261294e-06, + "loss": 0.9216, + "step": 12099 + }, + { + "epoch": 0.8313865603957675, + "grad_norm": 1.1608203649520874, + "learning_rate": 4.207015461687061e-06, + "loss": 0.9383, + "step": 12100 + }, + { + "epoch": 0.831455270028858, + "grad_norm": 1.1287217140197754, + "learning_rate": 4.203686924755715e-06, + "loss": 0.9326, + "step": 12101 + }, + { + "epoch": 0.8315239796619486, + "grad_norm": 1.0223361253738403, + "learning_rate": 4.200359584223584e-06, + "loss": 0.9391, + "step": 12102 + }, + { + "epoch": 0.8315926892950392, + "grad_norm": 1.0215238332748413, + "learning_rate": 4.197033440282078e-06, + "loss": 0.9241, + "step": 12103 + }, + { + "epoch": 0.8316613989281297, + "grad_norm": 1.0619462728500366, + "learning_rate": 4.193708493122556e-06, + "loss": 0.9308, + "step": 12104 + }, + { + "epoch": 0.8317301085612203, + "grad_norm": 0.9720618724822998, + "learning_rate": 4.1903847429362955e-06, + "loss": 0.9088, + "step": 12105 + }, + { + "epoch": 0.8317988181943109, + "grad_norm": 1.0965508222579956, + "learning_rate": 4.1870621899145144e-06, + "loss": 0.9351, + "step": 12106 + }, + { + "epoch": 0.8318675278274014, + "grad_norm": 1.0201143026351929, + "learning_rate": 4.183740834248357e-06, + "loss": 0.8921, + "step": 12107 + }, + { + "epoch": 0.831936237460492, + "grad_norm": 1.016323447227478, + "learning_rate": 4.180420676128899e-06, + "loss": 0.9056, + "step": 12108 + }, + { + "epoch": 0.8320049470935825, + "grad_norm": 1.0599288940429688, + "learning_rate": 4.1771017157471396e-06, + "loss": 0.91, + "step": 12109 + }, + { + "epoch": 0.8320736567266731, + "grad_norm": 1.0799387693405151, + "learning_rate": 4.173783953294025e-06, + "loss": 0.9395, + "step": 12110 + }, + { + "epoch": 0.8321423663597637, + "grad_norm": 1.024133324623108, + "learning_rate": 4.170467388960419e-06, + "loss": 0.946, + "step": 12111 + }, + { + "epoch": 0.8322110759928542, + "grad_norm": 1.0044535398483276, + "learning_rate": 4.167152022937124e-06, + "loss": 0.9159, + "step": 12112 + }, + { + "epoch": 0.8322797856259447, + "grad_norm": 1.0090912580490112, + "learning_rate": 4.163837855414876e-06, + "loss": 0.9318, + "step": 12113 + }, + { + "epoch": 0.8323484952590353, + "grad_norm": 1.0415658950805664, + "learning_rate": 4.1605248865843275e-06, + "loss": 0.9363, + "step": 12114 + }, + { + "epoch": 0.8324172048921259, + "grad_norm": 1.0138438940048218, + "learning_rate": 4.157213116636075e-06, + "loss": 0.9119, + "step": 12115 + }, + { + "epoch": 0.8324859145252165, + "grad_norm": 1.0583336353302002, + "learning_rate": 4.153902545760641e-06, + "loss": 0.9485, + "step": 12116 + }, + { + "epoch": 0.8325546241583069, + "grad_norm": 1.0087271928787231, + "learning_rate": 4.150593174148487e-06, + "loss": 0.9207, + "step": 12117 + }, + { + "epoch": 0.8326233337913975, + "grad_norm": 1.0882002115249634, + "learning_rate": 4.147285001989998e-06, + "loss": 0.9591, + "step": 12118 + }, + { + "epoch": 0.8326920434244881, + "grad_norm": 1.070137619972229, + "learning_rate": 4.143978029475484e-06, + "loss": 0.9163, + "step": 12119 + }, + { + "epoch": 0.8327607530575787, + "grad_norm": 1.0163307189941406, + "learning_rate": 4.140672256795197e-06, + "loss": 0.9335, + "step": 12120 + }, + { + "epoch": 0.8328294626906693, + "grad_norm": 1.1111894845962524, + "learning_rate": 4.1373676841393205e-06, + "loss": 0.9572, + "step": 12121 + }, + { + "epoch": 0.8328981723237598, + "grad_norm": 1.0526809692382812, + "learning_rate": 4.1340643116979574e-06, + "loss": 0.9281, + "step": 12122 + }, + { + "epoch": 0.8329668819568503, + "grad_norm": 1.0768104791641235, + "learning_rate": 4.130762139661151e-06, + "loss": 0.9587, + "step": 12123 + }, + { + "epoch": 0.8330355915899409, + "grad_norm": 1.0302846431732178, + "learning_rate": 4.12746116821888e-06, + "loss": 0.9474, + "step": 12124 + }, + { + "epoch": 0.8331043012230315, + "grad_norm": 1.0563215017318726, + "learning_rate": 4.124161397561033e-06, + "loss": 0.9718, + "step": 12125 + }, + { + "epoch": 0.833173010856122, + "grad_norm": 1.0746923685073853, + "learning_rate": 4.1208628278774556e-06, + "loss": 0.9317, + "step": 12126 + }, + { + "epoch": 0.8332417204892126, + "grad_norm": 1.063737392425537, + "learning_rate": 4.117565459357906e-06, + "loss": 0.9312, + "step": 12127 + }, + { + "epoch": 0.8333104301223031, + "grad_norm": 1.0299181938171387, + "learning_rate": 4.114269292192091e-06, + "loss": 0.9551, + "step": 12128 + }, + { + "epoch": 0.8333791397553937, + "grad_norm": 1.060612678527832, + "learning_rate": 4.110974326569622e-06, + "loss": 0.9692, + "step": 12129 + }, + { + "epoch": 0.8334478493884843, + "grad_norm": 1.0758137702941895, + "learning_rate": 4.107680562680063e-06, + "loss": 0.8851, + "step": 12130 + }, + { + "epoch": 0.8335165590215748, + "grad_norm": 1.030537724494934, + "learning_rate": 4.1043880007129024e-06, + "loss": 0.9073, + "step": 12131 + }, + { + "epoch": 0.8335852686546654, + "grad_norm": 1.0567922592163086, + "learning_rate": 4.10109664085756e-06, + "loss": 0.9494, + "step": 12132 + }, + { + "epoch": 0.8336539782877559, + "grad_norm": 1.0864925384521484, + "learning_rate": 4.097806483303393e-06, + "loss": 0.9318, + "step": 12133 + }, + { + "epoch": 0.8337226879208465, + "grad_norm": 1.0311164855957031, + "learning_rate": 4.094517528239663e-06, + "loss": 0.911, + "step": 12134 + }, + { + "epoch": 0.833791397553937, + "grad_norm": 1.0342899560928345, + "learning_rate": 4.091229775855598e-06, + "loss": 0.9195, + "step": 12135 + }, + { + "epoch": 0.8338601071870276, + "grad_norm": 1.0705069303512573, + "learning_rate": 4.087943226340332e-06, + "loss": 0.9268, + "step": 12136 + }, + { + "epoch": 0.8339288168201182, + "grad_norm": 1.0527191162109375, + "learning_rate": 4.084657879882944e-06, + "loss": 0.9186, + "step": 12137 + }, + { + "epoch": 0.8339975264532088, + "grad_norm": 1.050176739692688, + "learning_rate": 4.081373736672439e-06, + "loss": 0.9307, + "step": 12138 + }, + { + "epoch": 0.8340662360862993, + "grad_norm": 1.0696555376052856, + "learning_rate": 4.078090796897743e-06, + "loss": 0.9258, + "step": 12139 + }, + { + "epoch": 0.8341349457193898, + "grad_norm": 1.0391312837600708, + "learning_rate": 4.074809060747728e-06, + "loss": 0.8863, + "step": 12140 + }, + { + "epoch": 0.8342036553524804, + "grad_norm": 1.161088466644287, + "learning_rate": 4.07152852841119e-06, + "loss": 0.931, + "step": 12141 + }, + { + "epoch": 0.834272364985571, + "grad_norm": 1.0392245054244995, + "learning_rate": 4.068249200076851e-06, + "loss": 0.9396, + "step": 12142 + }, + { + "epoch": 0.8343410746186616, + "grad_norm": 1.0933263301849365, + "learning_rate": 4.06497107593338e-06, + "loss": 0.9472, + "step": 12143 + }, + { + "epoch": 0.834409784251752, + "grad_norm": 1.0684095621109009, + "learning_rate": 4.061694156169354e-06, + "loss": 0.919, + "step": 12144 + }, + { + "epoch": 0.8344784938848426, + "grad_norm": 1.12293541431427, + "learning_rate": 4.058418440973294e-06, + "loss": 0.963, + "step": 12145 + }, + { + "epoch": 0.8345472035179332, + "grad_norm": 1.0377813577651978, + "learning_rate": 4.0551439305336615e-06, + "loss": 0.9261, + "step": 12146 + }, + { + "epoch": 0.8346159131510238, + "grad_norm": 1.0568610429763794, + "learning_rate": 4.051870625038818e-06, + "loss": 0.9289, + "step": 12147 + }, + { + "epoch": 0.8346846227841144, + "grad_norm": 1.0130804777145386, + "learning_rate": 4.048598524677094e-06, + "loss": 0.9427, + "step": 12148 + }, + { + "epoch": 0.8347533324172048, + "grad_norm": 1.0425792932510376, + "learning_rate": 4.045327629636714e-06, + "loss": 0.9408, + "step": 12149 + }, + { + "epoch": 0.8348220420502954, + "grad_norm": 1.0257679224014282, + "learning_rate": 4.042057940105859e-06, + "loss": 0.9105, + "step": 12150 + }, + { + "epoch": 0.834890751683386, + "grad_norm": 1.0578770637512207, + "learning_rate": 4.038789456272632e-06, + "loss": 0.9278, + "step": 12151 + }, + { + "epoch": 0.8349594613164766, + "grad_norm": 1.0847866535186768, + "learning_rate": 4.035522178325069e-06, + "loss": 0.9351, + "step": 12152 + }, + { + "epoch": 0.8350281709495672, + "grad_norm": 1.0792934894561768, + "learning_rate": 4.032256106451135e-06, + "loss": 0.9364, + "step": 12153 + }, + { + "epoch": 0.8350968805826577, + "grad_norm": 0.9752399921417236, + "learning_rate": 4.02899124083872e-06, + "loss": 0.8855, + "step": 12154 + }, + { + "epoch": 0.8351655902157482, + "grad_norm": 1.0328949689865112, + "learning_rate": 4.025727581675648e-06, + "loss": 0.9134, + "step": 12155 + }, + { + "epoch": 0.8352342998488388, + "grad_norm": 1.0285309553146362, + "learning_rate": 4.022465129149683e-06, + "loss": 0.9305, + "step": 12156 + }, + { + "epoch": 0.8353030094819294, + "grad_norm": 1.0580674409866333, + "learning_rate": 4.0192038834485075e-06, + "loss": 0.9182, + "step": 12157 + }, + { + "epoch": 0.83537171911502, + "grad_norm": 1.1126210689544678, + "learning_rate": 4.0159438447597455e-06, + "loss": 0.9416, + "step": 12158 + }, + { + "epoch": 0.8354404287481105, + "grad_norm": 1.052819848060608, + "learning_rate": 4.012685013270934e-06, + "loss": 0.9305, + "step": 12159 + }, + { + "epoch": 0.835509138381201, + "grad_norm": 1.1108752489089966, + "learning_rate": 4.009427389169557e-06, + "loss": 0.9174, + "step": 12160 + }, + { + "epoch": 0.8355778480142916, + "grad_norm": 1.066178560256958, + "learning_rate": 4.0061709726430225e-06, + "loss": 0.9593, + "step": 12161 + }, + { + "epoch": 0.8356465576473822, + "grad_norm": 1.0193355083465576, + "learning_rate": 4.00291576387867e-06, + "loss": 0.9066, + "step": 12162 + }, + { + "epoch": 0.8357152672804727, + "grad_norm": 1.1044081449508667, + "learning_rate": 3.999661763063775e-06, + "loss": 0.9776, + "step": 12163 + }, + { + "epoch": 0.8357839769135633, + "grad_norm": 1.0289283990859985, + "learning_rate": 3.996408970385537e-06, + "loss": 0.9151, + "step": 12164 + }, + { + "epoch": 0.8358526865466538, + "grad_norm": 1.055875301361084, + "learning_rate": 3.99315738603108e-06, + "loss": 0.9338, + "step": 12165 + }, + { + "epoch": 0.8359213961797444, + "grad_norm": 1.064091682434082, + "learning_rate": 3.989907010187468e-06, + "loss": 0.9403, + "step": 12166 + }, + { + "epoch": 0.835990105812835, + "grad_norm": 1.0910937786102295, + "learning_rate": 3.986657843041699e-06, + "loss": 0.9332, + "step": 12167 + }, + { + "epoch": 0.8360588154459255, + "grad_norm": 1.006186842918396, + "learning_rate": 3.983409884780695e-06, + "loss": 0.9063, + "step": 12168 + }, + { + "epoch": 0.8361275250790161, + "grad_norm": 1.123480200767517, + "learning_rate": 3.9801631355912995e-06, + "loss": 0.9858, + "step": 12169 + }, + { + "epoch": 0.8361962347121067, + "grad_norm": 1.0228362083435059, + "learning_rate": 3.976917595660304e-06, + "loss": 0.9031, + "step": 12170 + }, + { + "epoch": 0.8362649443451972, + "grad_norm": 1.105780005455017, + "learning_rate": 3.973673265174424e-06, + "loss": 0.9083, + "step": 12171 + }, + { + "epoch": 0.8363336539782877, + "grad_norm": 1.072906732559204, + "learning_rate": 3.970430144320297e-06, + "loss": 0.9213, + "step": 12172 + }, + { + "epoch": 0.8364023636113783, + "grad_norm": 1.0563604831695557, + "learning_rate": 3.967188233284502e-06, + "loss": 0.9323, + "step": 12173 + }, + { + "epoch": 0.8364710732444689, + "grad_norm": 1.103596806526184, + "learning_rate": 3.963947532253548e-06, + "loss": 0.9378, + "step": 12174 + }, + { + "epoch": 0.8365397828775595, + "grad_norm": 1.0556912422180176, + "learning_rate": 3.96070804141386e-06, + "loss": 0.9512, + "step": 12175 + }, + { + "epoch": 0.83660849251065, + "grad_norm": 1.0592446327209473, + "learning_rate": 3.9574697609518095e-06, + "loss": 0.8988, + "step": 12176 + }, + { + "epoch": 0.8366772021437405, + "grad_norm": 1.0048222541809082, + "learning_rate": 3.954232691053697e-06, + "loss": 0.8912, + "step": 12177 + }, + { + "epoch": 0.8367459117768311, + "grad_norm": 0.978837251663208, + "learning_rate": 3.9509968319057425e-06, + "loss": 0.9318, + "step": 12178 + }, + { + "epoch": 0.8368146214099217, + "grad_norm": 1.0888440608978271, + "learning_rate": 3.947762183694112e-06, + "loss": 0.9461, + "step": 12179 + }, + { + "epoch": 0.8368833310430123, + "grad_norm": 1.0201581716537476, + "learning_rate": 3.944528746604881e-06, + "loss": 0.9133, + "step": 12180 + }, + { + "epoch": 0.8369520406761027, + "grad_norm": 1.028889775276184, + "learning_rate": 3.941296520824072e-06, + "loss": 0.9113, + "step": 12181 + }, + { + "epoch": 0.8370207503091933, + "grad_norm": 1.0917303562164307, + "learning_rate": 3.938065506537636e-06, + "loss": 0.9329, + "step": 12182 + }, + { + "epoch": 0.8370894599422839, + "grad_norm": 1.0284708738327026, + "learning_rate": 3.934835703931447e-06, + "loss": 0.9528, + "step": 12183 + }, + { + "epoch": 0.8371581695753745, + "grad_norm": 1.0615824460983276, + "learning_rate": 3.931607113191324e-06, + "loss": 0.9412, + "step": 12184 + }, + { + "epoch": 0.8372268792084651, + "grad_norm": 1.086951494216919, + "learning_rate": 3.92837973450299e-06, + "loss": 0.9507, + "step": 12185 + }, + { + "epoch": 0.8372955888415556, + "grad_norm": 1.051750659942627, + "learning_rate": 3.925153568052123e-06, + "loss": 0.9377, + "step": 12186 + }, + { + "epoch": 0.8373642984746461, + "grad_norm": 1.0441310405731201, + "learning_rate": 3.921928614024323e-06, + "loss": 0.9395, + "step": 12187 + }, + { + "epoch": 0.8374330081077367, + "grad_norm": 1.0565197467803955, + "learning_rate": 3.918704872605114e-06, + "loss": 0.9322, + "step": 12188 + }, + { + "epoch": 0.8375017177408273, + "grad_norm": 1.0811272859573364, + "learning_rate": 3.915482343979967e-06, + "loss": 0.8956, + "step": 12189 + }, + { + "epoch": 0.8375704273739178, + "grad_norm": 1.123972773551941, + "learning_rate": 3.912261028334261e-06, + "loss": 0.917, + "step": 12190 + }, + { + "epoch": 0.8376391370070084, + "grad_norm": 1.0919814109802246, + "learning_rate": 3.9090409258533165e-06, + "loss": 0.8999, + "step": 12191 + }, + { + "epoch": 0.8377078466400989, + "grad_norm": 1.0725699663162231, + "learning_rate": 3.905822036722389e-06, + "loss": 0.9045, + "step": 12192 + }, + { + "epoch": 0.8377765562731895, + "grad_norm": 1.099852204322815, + "learning_rate": 3.902604361126666e-06, + "loss": 0.9173, + "step": 12193 + }, + { + "epoch": 0.8378452659062801, + "grad_norm": 1.0641801357269287, + "learning_rate": 3.8993878992512415e-06, + "loss": 0.9247, + "step": 12194 + }, + { + "epoch": 0.8379139755393706, + "grad_norm": 1.0229814052581787, + "learning_rate": 3.896172651281174e-06, + "loss": 0.9551, + "step": 12195 + }, + { + "epoch": 0.8379826851724612, + "grad_norm": 1.1014561653137207, + "learning_rate": 3.892958617401418e-06, + "loss": 0.9429, + "step": 12196 + }, + { + "epoch": 0.8380513948055517, + "grad_norm": 1.051297664642334, + "learning_rate": 3.889745797796885e-06, + "loss": 0.9465, + "step": 12197 + }, + { + "epoch": 0.8381201044386423, + "grad_norm": 1.089254379272461, + "learning_rate": 3.886534192652402e-06, + "loss": 0.9493, + "step": 12198 + }, + { + "epoch": 0.8381888140717328, + "grad_norm": 1.0974400043487549, + "learning_rate": 3.8833238021527415e-06, + "loss": 0.9605, + "step": 12199 + }, + { + "epoch": 0.8382575237048234, + "grad_norm": 1.0809804201126099, + "learning_rate": 3.880114626482581e-06, + "loss": 0.9178, + "step": 12200 + }, + { + "epoch": 0.838326233337914, + "grad_norm": 1.036906123161316, + "learning_rate": 3.876906665826549e-06, + "loss": 0.9362, + "step": 12201 + }, + { + "epoch": 0.8383949429710046, + "grad_norm": 1.0712566375732422, + "learning_rate": 3.873699920369195e-06, + "loss": 0.9618, + "step": 12202 + }, + { + "epoch": 0.838463652604095, + "grad_norm": 1.1582120656967163, + "learning_rate": 3.8704943902950044e-06, + "loss": 0.9847, + "step": 12203 + }, + { + "epoch": 0.8385323622371856, + "grad_norm": 1.059890627861023, + "learning_rate": 3.867290075788394e-06, + "loss": 0.911, + "step": 12204 + }, + { + "epoch": 0.8386010718702762, + "grad_norm": 1.0406646728515625, + "learning_rate": 3.864086977033693e-06, + "loss": 0.8951, + "step": 12205 + }, + { + "epoch": 0.8386697815033668, + "grad_norm": 1.0459775924682617, + "learning_rate": 3.860885094215183e-06, + "loss": 0.9249, + "step": 12206 + }, + { + "epoch": 0.8387384911364574, + "grad_norm": 1.0221196413040161, + "learning_rate": 3.8576844275170635e-06, + "loss": 0.9485, + "step": 12207 + }, + { + "epoch": 0.8388072007695478, + "grad_norm": 1.0518344640731812, + "learning_rate": 3.854484977123468e-06, + "loss": 0.9209, + "step": 12208 + }, + { + "epoch": 0.8388759104026384, + "grad_norm": 1.072669506072998, + "learning_rate": 3.851286743218466e-06, + "loss": 0.9433, + "step": 12209 + }, + { + "epoch": 0.838944620035729, + "grad_norm": 1.1286430358886719, + "learning_rate": 3.848089725986034e-06, + "loss": 0.9286, + "step": 12210 + }, + { + "epoch": 0.8390133296688196, + "grad_norm": 1.0627402067184448, + "learning_rate": 3.844893925610104e-06, + "loss": 0.9202, + "step": 12211 + }, + { + "epoch": 0.8390820393019102, + "grad_norm": 1.0510495901107788, + "learning_rate": 3.841699342274527e-06, + "loss": 0.9168, + "step": 12212 + }, + { + "epoch": 0.8391507489350006, + "grad_norm": 1.0822217464447021, + "learning_rate": 3.838505976163087e-06, + "loss": 0.9102, + "step": 12213 + }, + { + "epoch": 0.8392194585680912, + "grad_norm": 1.036496639251709, + "learning_rate": 3.8353138274595e-06, + "loss": 0.9309, + "step": 12214 + }, + { + "epoch": 0.8392881682011818, + "grad_norm": 1.0132932662963867, + "learning_rate": 3.832122896347398e-06, + "loss": 0.9218, + "step": 12215 + }, + { + "epoch": 0.8393568778342724, + "grad_norm": 1.0400850772857666, + "learning_rate": 3.8289331830103606e-06, + "loss": 0.9088, + "step": 12216 + }, + { + "epoch": 0.839425587467363, + "grad_norm": 1.0348241329193115, + "learning_rate": 3.825744687631894e-06, + "loss": 0.935, + "step": 12217 + }, + { + "epoch": 0.8394942971004535, + "grad_norm": 1.0311181545257568, + "learning_rate": 3.822557410395416e-06, + "loss": 0.9178, + "step": 12218 + }, + { + "epoch": 0.839563006733544, + "grad_norm": 1.0606298446655273, + "learning_rate": 3.819371351484302e-06, + "loss": 0.9692, + "step": 12219 + }, + { + "epoch": 0.8396317163666346, + "grad_norm": 1.0343350172042847, + "learning_rate": 3.816186511081843e-06, + "loss": 0.9371, + "step": 12220 + }, + { + "epoch": 0.8397004259997252, + "grad_norm": 1.0905903577804565, + "learning_rate": 3.8130028893712532e-06, + "loss": 0.9149, + "step": 12221 + }, + { + "epoch": 0.8397691356328157, + "grad_norm": 1.0850155353546143, + "learning_rate": 3.809820486535687e-06, + "loss": 0.9477, + "step": 12222 + }, + { + "epoch": 0.8398378452659063, + "grad_norm": 1.0777400732040405, + "learning_rate": 3.8066393027582276e-06, + "loss": 0.9156, + "step": 12223 + }, + { + "epoch": 0.8399065548989968, + "grad_norm": 1.0600709915161133, + "learning_rate": 3.8034593382218877e-06, + "loss": 0.9534, + "step": 12224 + }, + { + "epoch": 0.8399752645320874, + "grad_norm": 1.1155648231506348, + "learning_rate": 3.8002805931096146e-06, + "loss": 1.0248, + "step": 12225 + }, + { + "epoch": 0.840043974165178, + "grad_norm": 1.0631171464920044, + "learning_rate": 3.7971030676042655e-06, + "loss": 0.961, + "step": 12226 + }, + { + "epoch": 0.8401126837982685, + "grad_norm": 1.05125093460083, + "learning_rate": 3.79392676188865e-06, + "loss": 0.9414, + "step": 12227 + }, + { + "epoch": 0.8401813934313591, + "grad_norm": 1.083605170249939, + "learning_rate": 3.790751676145496e-06, + "loss": 0.9238, + "step": 12228 + }, + { + "epoch": 0.8402501030644496, + "grad_norm": 1.0624525547027588, + "learning_rate": 3.787577810557469e-06, + "loss": 0.9082, + "step": 12229 + }, + { + "epoch": 0.8403188126975402, + "grad_norm": 1.080212116241455, + "learning_rate": 3.784405165307159e-06, + "loss": 0.9489, + "step": 12230 + }, + { + "epoch": 0.8403875223306307, + "grad_norm": 1.074500560760498, + "learning_rate": 3.781233740577081e-06, + "loss": 0.9352, + "step": 12231 + }, + { + "epoch": 0.8404562319637213, + "grad_norm": 1.1197142601013184, + "learning_rate": 3.7780635365496864e-06, + "loss": 0.9129, + "step": 12232 + }, + { + "epoch": 0.8405249415968119, + "grad_norm": 1.0508078336715698, + "learning_rate": 3.7748945534073573e-06, + "loss": 0.9076, + "step": 12233 + }, + { + "epoch": 0.8405936512299025, + "grad_norm": 1.0557643175125122, + "learning_rate": 3.771726791332403e-06, + "loss": 0.9228, + "step": 12234 + }, + { + "epoch": 0.840662360862993, + "grad_norm": 1.0599395036697388, + "learning_rate": 3.7685602505070693e-06, + "loss": 0.9382, + "step": 12235 + }, + { + "epoch": 0.8407310704960835, + "grad_norm": 1.0825812816619873, + "learning_rate": 3.7653949311135112e-06, + "loss": 0.9466, + "step": 12236 + }, + { + "epoch": 0.8407997801291741, + "grad_norm": 0.9982886910438538, + "learning_rate": 3.762230833333838e-06, + "loss": 0.9269, + "step": 12237 + }, + { + "epoch": 0.8408684897622647, + "grad_norm": 1.054235577583313, + "learning_rate": 3.7590679573500735e-06, + "loss": 0.9036, + "step": 12238 + }, + { + "epoch": 0.8409371993953553, + "grad_norm": 1.0357664823532104, + "learning_rate": 3.755906303344184e-06, + "loss": 0.9018, + "step": 12239 + }, + { + "epoch": 0.8410059090284457, + "grad_norm": 1.0296399593353271, + "learning_rate": 3.752745871498048e-06, + "loss": 0.9125, + "step": 12240 + }, + { + "epoch": 0.8410746186615363, + "grad_norm": 1.0312901735305786, + "learning_rate": 3.7495866619934844e-06, + "loss": 0.9317, + "step": 12241 + }, + { + "epoch": 0.8411433282946269, + "grad_norm": 1.0435298681259155, + "learning_rate": 3.746428675012251e-06, + "loss": 0.9365, + "step": 12242 + }, + { + "epoch": 0.8412120379277175, + "grad_norm": 1.1165523529052734, + "learning_rate": 3.74327191073601e-06, + "loss": 0.932, + "step": 12243 + }, + { + "epoch": 0.8412807475608081, + "grad_norm": 0.9764993786811829, + "learning_rate": 3.7401163693463746e-06, + "loss": 0.8752, + "step": 12244 + }, + { + "epoch": 0.8413494571938985, + "grad_norm": 1.0190304517745972, + "learning_rate": 3.7369620510248863e-06, + "loss": 0.9559, + "step": 12245 + }, + { + "epoch": 0.8414181668269891, + "grad_norm": 1.0533236265182495, + "learning_rate": 3.733808955953003e-06, + "loss": 0.9087, + "step": 12246 + }, + { + "epoch": 0.8414868764600797, + "grad_norm": 1.0907206535339355, + "learning_rate": 3.7306570843121226e-06, + "loss": 0.9569, + "step": 12247 + }, + { + "epoch": 0.8415555860931703, + "grad_norm": 1.118791937828064, + "learning_rate": 3.7275064362835705e-06, + "loss": 0.9734, + "step": 12248 + }, + { + "epoch": 0.8416242957262609, + "grad_norm": 1.0765595436096191, + "learning_rate": 3.7243570120486042e-06, + "loss": 0.9292, + "step": 12249 + }, + { + "epoch": 0.8416930053593514, + "grad_norm": 1.0427714586257935, + "learning_rate": 3.7212088117884127e-06, + "loss": 0.9473, + "step": 12250 + }, + { + "epoch": 0.8417617149924419, + "grad_norm": 1.0489193201065063, + "learning_rate": 3.7180618356840974e-06, + "loss": 0.9437, + "step": 12251 + }, + { + "epoch": 0.8418304246255325, + "grad_norm": 1.0585293769836426, + "learning_rate": 3.7149160839167084e-06, + "loss": 0.9408, + "step": 12252 + }, + { + "epoch": 0.8418991342586231, + "grad_norm": 1.0317072868347168, + "learning_rate": 3.711771556667218e-06, + "loss": 0.9214, + "step": 12253 + }, + { + "epoch": 0.8419678438917136, + "grad_norm": 1.0380688905715942, + "learning_rate": 3.7086282541165307e-06, + "loss": 0.9234, + "step": 12254 + }, + { + "epoch": 0.8420365535248042, + "grad_norm": 1.1011000871658325, + "learning_rate": 3.7054861764454828e-06, + "loss": 0.9333, + "step": 12255 + }, + { + "epoch": 0.8421052631578947, + "grad_norm": 1.056457281112671, + "learning_rate": 3.702345323834827e-06, + "loss": 0.9435, + "step": 12256 + }, + { + "epoch": 0.8421739727909853, + "grad_norm": 1.0854911804199219, + "learning_rate": 3.699205696465258e-06, + "loss": 0.9694, + "step": 12257 + }, + { + "epoch": 0.8422426824240759, + "grad_norm": 1.0809786319732666, + "learning_rate": 3.6960672945173967e-06, + "loss": 0.9613, + "step": 12258 + }, + { + "epoch": 0.8423113920571664, + "grad_norm": 1.040189504623413, + "learning_rate": 3.6929301181717944e-06, + "loss": 0.9251, + "step": 12259 + }, + { + "epoch": 0.842380101690257, + "grad_norm": 1.1371155977249146, + "learning_rate": 3.689794167608937e-06, + "loss": 0.9239, + "step": 12260 + }, + { + "epoch": 0.8424488113233475, + "grad_norm": 1.0222303867340088, + "learning_rate": 3.6866594430092243e-06, + "loss": 0.9178, + "step": 12261 + }, + { + "epoch": 0.8425175209564381, + "grad_norm": 1.0347579717636108, + "learning_rate": 3.6835259445529957e-06, + "loss": 0.8947, + "step": 12262 + }, + { + "epoch": 0.8425862305895286, + "grad_norm": 1.0329539775848389, + "learning_rate": 3.680393672420521e-06, + "loss": 0.9211, + "step": 12263 + }, + { + "epoch": 0.8426549402226192, + "grad_norm": 1.0784316062927246, + "learning_rate": 3.677262626792008e-06, + "loss": 0.9365, + "step": 12264 + }, + { + "epoch": 0.8427236498557098, + "grad_norm": 1.0675005912780762, + "learning_rate": 3.6741328078475664e-06, + "loss": 0.9284, + "step": 12265 + }, + { + "epoch": 0.8427923594888004, + "grad_norm": 1.0418853759765625, + "learning_rate": 3.671004215767268e-06, + "loss": 0.9128, + "step": 12266 + }, + { + "epoch": 0.8428610691218908, + "grad_norm": 1.0363177061080933, + "learning_rate": 3.667876850731086e-06, + "loss": 0.9049, + "step": 12267 + }, + { + "epoch": 0.8429297787549814, + "grad_norm": 1.0266954898834229, + "learning_rate": 3.6647507129189435e-06, + "loss": 0.9465, + "step": 12268 + }, + { + "epoch": 0.842998488388072, + "grad_norm": 1.0354152917861938, + "learning_rate": 3.661625802510682e-06, + "loss": 0.8833, + "step": 12269 + }, + { + "epoch": 0.8430671980211626, + "grad_norm": 1.0724148750305176, + "learning_rate": 3.658502119686083e-06, + "loss": 0.9492, + "step": 12270 + }, + { + "epoch": 0.8431359076542532, + "grad_norm": 1.0334283113479614, + "learning_rate": 3.655379664624839e-06, + "loss": 0.9265, + "step": 12271 + }, + { + "epoch": 0.8432046172873436, + "grad_norm": 1.096497654914856, + "learning_rate": 3.6522584375065893e-06, + "loss": 0.9169, + "step": 12272 + }, + { + "epoch": 0.8432733269204342, + "grad_norm": 1.0817033052444458, + "learning_rate": 3.649138438510896e-06, + "loss": 0.925, + "step": 12273 + }, + { + "epoch": 0.8433420365535248, + "grad_norm": 1.104803204536438, + "learning_rate": 3.6460196678172464e-06, + "loss": 0.963, + "step": 12274 + }, + { + "epoch": 0.8434107461866154, + "grad_norm": 1.0361344814300537, + "learning_rate": 3.642902125605069e-06, + "loss": 0.8809, + "step": 12275 + }, + { + "epoch": 0.843479455819706, + "grad_norm": 1.034854769706726, + "learning_rate": 3.6397858120537142e-06, + "loss": 0.9187, + "step": 12276 + }, + { + "epoch": 0.8435481654527964, + "grad_norm": 1.028459072113037, + "learning_rate": 3.6366707273424504e-06, + "loss": 0.9306, + "step": 12277 + }, + { + "epoch": 0.843616875085887, + "grad_norm": 1.0792453289031982, + "learning_rate": 3.6335568716504952e-06, + "loss": 0.9595, + "step": 12278 + }, + { + "epoch": 0.8436855847189776, + "grad_norm": 1.0722148418426514, + "learning_rate": 3.630444245156986e-06, + "loss": 0.9475, + "step": 12279 + }, + { + "epoch": 0.8437542943520682, + "grad_norm": 1.0396969318389893, + "learning_rate": 3.62733284804099e-06, + "loss": 0.9117, + "step": 12280 + }, + { + "epoch": 0.8438230039851587, + "grad_norm": 1.0327156782150269, + "learning_rate": 3.6242226804815065e-06, + "loss": 0.8944, + "step": 12281 + }, + { + "epoch": 0.8438917136182493, + "grad_norm": 1.100889801979065, + "learning_rate": 3.621113742657456e-06, + "loss": 0.9391, + "step": 12282 + }, + { + "epoch": 0.8439604232513398, + "grad_norm": 1.0637400150299072, + "learning_rate": 3.618006034747695e-06, + "loss": 0.95, + "step": 12283 + }, + { + "epoch": 0.8440291328844304, + "grad_norm": 1.0216749906539917, + "learning_rate": 3.614899556931009e-06, + "loss": 0.9342, + "step": 12284 + }, + { + "epoch": 0.844097842517521, + "grad_norm": 1.0499944686889648, + "learning_rate": 3.6117943093861156e-06, + "loss": 0.9059, + "step": 12285 + }, + { + "epoch": 0.8441665521506115, + "grad_norm": 1.050349473953247, + "learning_rate": 3.6086902922916553e-06, + "loss": 0.9385, + "step": 12286 + }, + { + "epoch": 0.8442352617837021, + "grad_norm": 1.1334173679351807, + "learning_rate": 3.605587505826197e-06, + "loss": 0.9664, + "step": 12287 + }, + { + "epoch": 0.8443039714167926, + "grad_norm": 0.988175630569458, + "learning_rate": 3.6024859501682494e-06, + "loss": 0.9286, + "step": 12288 + }, + { + "epoch": 0.8443726810498832, + "grad_norm": 1.0834704637527466, + "learning_rate": 3.599385625496232e-06, + "loss": 0.9256, + "step": 12289 + }, + { + "epoch": 0.8444413906829737, + "grad_norm": 1.0541126728057861, + "learning_rate": 3.596286531988513e-06, + "loss": 0.913, + "step": 12290 + }, + { + "epoch": 0.8445101003160643, + "grad_norm": 1.0316873788833618, + "learning_rate": 3.593188669823383e-06, + "loss": 0.9342, + "step": 12291 + }, + { + "epoch": 0.8445788099491549, + "grad_norm": 1.0693053007125854, + "learning_rate": 3.590092039179052e-06, + "loss": 0.9497, + "step": 12292 + }, + { + "epoch": 0.8446475195822454, + "grad_norm": 1.0637959241867065, + "learning_rate": 3.5869966402336695e-06, + "loss": 0.9397, + "step": 12293 + }, + { + "epoch": 0.844716229215336, + "grad_norm": 1.033275842666626, + "learning_rate": 3.5839024731653155e-06, + "loss": 0.9221, + "step": 12294 + }, + { + "epoch": 0.8447849388484265, + "grad_norm": 1.1195135116577148, + "learning_rate": 3.580809538151994e-06, + "loss": 0.9654, + "step": 12295 + }, + { + "epoch": 0.8448536484815171, + "grad_norm": 1.011627197265625, + "learning_rate": 3.577717835371647e-06, + "loss": 0.9124, + "step": 12296 + }, + { + "epoch": 0.8449223581146077, + "grad_norm": 1.0907173156738281, + "learning_rate": 3.5746273650021228e-06, + "loss": 0.9744, + "step": 12297 + }, + { + "epoch": 0.8449910677476983, + "grad_norm": 1.132449746131897, + "learning_rate": 3.5715381272212246e-06, + "loss": 0.9833, + "step": 12298 + }, + { + "epoch": 0.8450597773807887, + "grad_norm": 1.0719349384307861, + "learning_rate": 3.568450122206671e-06, + "loss": 0.9063, + "step": 12299 + }, + { + "epoch": 0.8451284870138793, + "grad_norm": 1.0271426439285278, + "learning_rate": 3.5653633501361165e-06, + "loss": 0.9268, + "step": 12300 + }, + { + "epoch": 0.8451971966469699, + "grad_norm": 1.0765966176986694, + "learning_rate": 3.5622778111871425e-06, + "loss": 0.9327, + "step": 12301 + }, + { + "epoch": 0.8452659062800605, + "grad_norm": 1.00552499294281, + "learning_rate": 3.559193505537248e-06, + "loss": 0.9239, + "step": 12302 + }, + { + "epoch": 0.8453346159131511, + "grad_norm": 1.1019383668899536, + "learning_rate": 3.5561104333638796e-06, + "loss": 0.9139, + "step": 12303 + }, + { + "epoch": 0.8454033255462415, + "grad_norm": 1.0156601667404175, + "learning_rate": 3.553028594844404e-06, + "loss": 0.9093, + "step": 12304 + }, + { + "epoch": 0.8454720351793321, + "grad_norm": 1.0635254383087158, + "learning_rate": 3.549947990156113e-06, + "loss": 0.933, + "step": 12305 + }, + { + "epoch": 0.8455407448124227, + "grad_norm": 1.0128878355026245, + "learning_rate": 3.5468686194762456e-06, + "loss": 0.8943, + "step": 12306 + }, + { + "epoch": 0.8456094544455133, + "grad_norm": 1.0036367177963257, + "learning_rate": 3.543790482981937e-06, + "loss": 0.9151, + "step": 12307 + }, + { + "epoch": 0.8456781640786039, + "grad_norm": 1.0809662342071533, + "learning_rate": 3.54071358085028e-06, + "loss": 0.9313, + "step": 12308 + }, + { + "epoch": 0.8457468737116943, + "grad_norm": 1.0605627298355103, + "learning_rate": 3.5376379132582875e-06, + "loss": 0.9084, + "step": 12309 + }, + { + "epoch": 0.8458155833447849, + "grad_norm": 1.038178563117981, + "learning_rate": 3.534563480382902e-06, + "loss": 0.9014, + "step": 12310 + }, + { + "epoch": 0.8458842929778755, + "grad_norm": 1.0709961652755737, + "learning_rate": 3.5314902824009897e-06, + "loss": 0.9499, + "step": 12311 + }, + { + "epoch": 0.8459530026109661, + "grad_norm": 1.0925170183181763, + "learning_rate": 3.5284183194893488e-06, + "loss": 0.9305, + "step": 12312 + }, + { + "epoch": 0.8460217122440566, + "grad_norm": 1.0947346687316895, + "learning_rate": 3.5253475918247152e-06, + "loss": 0.9246, + "step": 12313 + }, + { + "epoch": 0.8460904218771472, + "grad_norm": 1.0003663301467896, + "learning_rate": 3.522278099583737e-06, + "loss": 0.9224, + "step": 12314 + }, + { + "epoch": 0.8461591315102377, + "grad_norm": 1.1072996854782104, + "learning_rate": 3.5192098429430027e-06, + "loss": 0.9623, + "step": 12315 + }, + { + "epoch": 0.8462278411433283, + "grad_norm": 1.0086065530776978, + "learning_rate": 3.5161428220790356e-06, + "loss": 0.9147, + "step": 12316 + }, + { + "epoch": 0.8462965507764189, + "grad_norm": 1.0046271085739136, + "learning_rate": 3.5130770371682664e-06, + "loss": 0.955, + "step": 12317 + }, + { + "epoch": 0.8463652604095094, + "grad_norm": 1.1330538988113403, + "learning_rate": 3.510012488387074e-06, + "loss": 0.9353, + "step": 12318 + }, + { + "epoch": 0.8464339700426, + "grad_norm": 1.0137300491333008, + "learning_rate": 3.5069491759117586e-06, + "loss": 0.9099, + "step": 12319 + }, + { + "epoch": 0.8465026796756905, + "grad_norm": 1.022147297859192, + "learning_rate": 3.5038870999185542e-06, + "loss": 0.9304, + "step": 12320 + }, + { + "epoch": 0.8465713893087811, + "grad_norm": 1.0271625518798828, + "learning_rate": 3.500826260583623e-06, + "loss": 0.8823, + "step": 12321 + }, + { + "epoch": 0.8466400989418716, + "grad_norm": 1.1099499464035034, + "learning_rate": 3.4977666580830403e-06, + "loss": 0.978, + "step": 12322 + }, + { + "epoch": 0.8467088085749622, + "grad_norm": 1.0431894063949585, + "learning_rate": 3.4947082925928356e-06, + "loss": 0.9084, + "step": 12323 + }, + { + "epoch": 0.8467775182080528, + "grad_norm": 1.2028820514678955, + "learning_rate": 3.4916511642889478e-06, + "loss": 0.9867, + "step": 12324 + }, + { + "epoch": 0.8468462278411433, + "grad_norm": 0.9613275527954102, + "learning_rate": 3.488595273347253e-06, + "loss": 0.8758, + "step": 12325 + }, + { + "epoch": 0.8469149374742339, + "grad_norm": 1.0698260068893433, + "learning_rate": 3.4855406199435587e-06, + "loss": 0.9569, + "step": 12326 + }, + { + "epoch": 0.8469836471073244, + "grad_norm": 0.985489010810852, + "learning_rate": 3.482487204253598e-06, + "loss": 0.9114, + "step": 12327 + }, + { + "epoch": 0.847052356740415, + "grad_norm": 1.050599455833435, + "learning_rate": 3.4794350264530236e-06, + "loss": 0.9441, + "step": 12328 + }, + { + "epoch": 0.8471210663735056, + "grad_norm": 1.0188881158828735, + "learning_rate": 3.4763840867174297e-06, + "loss": 0.8971, + "step": 12329 + }, + { + "epoch": 0.8471897760065962, + "grad_norm": 1.0854439735412598, + "learning_rate": 3.473334385222335e-06, + "loss": 0.9581, + "step": 12330 + }, + { + "epoch": 0.8472584856396866, + "grad_norm": 1.0193722248077393, + "learning_rate": 3.4702859221431905e-06, + "loss": 0.9075, + "step": 12331 + }, + { + "epoch": 0.8473271952727772, + "grad_norm": 1.1106929779052734, + "learning_rate": 3.467238697655373e-06, + "loss": 0.9681, + "step": 12332 + }, + { + "epoch": 0.8473959049058678, + "grad_norm": 1.0620572566986084, + "learning_rate": 3.4641927119341775e-06, + "loss": 0.9508, + "step": 12333 + }, + { + "epoch": 0.8474646145389584, + "grad_norm": 1.0489130020141602, + "learning_rate": 3.4611479651548457e-06, + "loss": 0.882, + "step": 12334 + }, + { + "epoch": 0.847533324172049, + "grad_norm": 1.0686298608779907, + "learning_rate": 3.458104457492545e-06, + "loss": 0.9612, + "step": 12335 + }, + { + "epoch": 0.8476020338051394, + "grad_norm": 1.0311698913574219, + "learning_rate": 3.4550621891223515e-06, + "loss": 0.9301, + "step": 12336 + }, + { + "epoch": 0.84767074343823, + "grad_norm": 1.0765355825424194, + "learning_rate": 3.452021160219296e-06, + "loss": 0.8969, + "step": 12337 + }, + { + "epoch": 0.8477394530713206, + "grad_norm": 1.105230450630188, + "learning_rate": 3.448981370958329e-06, + "loss": 0.9787, + "step": 12338 + }, + { + "epoch": 0.8478081627044112, + "grad_norm": 1.148500919342041, + "learning_rate": 3.44594282151432e-06, + "loss": 0.9594, + "step": 12339 + }, + { + "epoch": 0.8478768723375018, + "grad_norm": 1.0324678421020508, + "learning_rate": 3.442905512062075e-06, + "loss": 0.9045, + "step": 12340 + }, + { + "epoch": 0.8479455819705923, + "grad_norm": 1.1574265956878662, + "learning_rate": 3.4398694427763334e-06, + "loss": 0.981, + "step": 12341 + }, + { + "epoch": 0.8480142916036828, + "grad_norm": 1.100232720375061, + "learning_rate": 3.436834613831763e-06, + "loss": 0.9543, + "step": 12342 + }, + { + "epoch": 0.8480830012367734, + "grad_norm": 1.0083119869232178, + "learning_rate": 3.433801025402944e-06, + "loss": 0.9233, + "step": 12343 + }, + { + "epoch": 0.848151710869864, + "grad_norm": 1.0961493253707886, + "learning_rate": 3.4307686776644004e-06, + "loss": 0.9442, + "step": 12344 + }, + { + "epoch": 0.8482204205029545, + "grad_norm": 1.0186933279037476, + "learning_rate": 3.4277375707905845e-06, + "loss": 0.909, + "step": 12345 + }, + { + "epoch": 0.8482891301360451, + "grad_norm": 1.0403319597244263, + "learning_rate": 3.4247077049558733e-06, + "loss": 0.9481, + "step": 12346 + }, + { + "epoch": 0.8483578397691356, + "grad_norm": 1.0920510292053223, + "learning_rate": 3.4216790803345777e-06, + "loss": 0.9391, + "step": 12347 + }, + { + "epoch": 0.8484265494022262, + "grad_norm": 1.063542366027832, + "learning_rate": 3.418651697100922e-06, + "loss": 0.9521, + "step": 12348 + }, + { + "epoch": 0.8484952590353168, + "grad_norm": 1.073070764541626, + "learning_rate": 3.4156255554290777e-06, + "loss": 0.9236, + "step": 12349 + }, + { + "epoch": 0.8485639686684073, + "grad_norm": 1.0102601051330566, + "learning_rate": 3.4126006554931312e-06, + "loss": 0.9221, + "step": 12350 + }, + { + "epoch": 0.8486326783014979, + "grad_norm": 1.0642870664596558, + "learning_rate": 3.4095769974671067e-06, + "loss": 0.9214, + "step": 12351 + }, + { + "epoch": 0.8487013879345884, + "grad_norm": 1.097960352897644, + "learning_rate": 3.4065545815249593e-06, + "loss": 0.972, + "step": 12352 + }, + { + "epoch": 0.848770097567679, + "grad_norm": 0.998414933681488, + "learning_rate": 3.4035334078405564e-06, + "loss": 0.9067, + "step": 12353 + }, + { + "epoch": 0.8488388072007695, + "grad_norm": 1.0659208297729492, + "learning_rate": 3.400513476587708e-06, + "loss": 0.9124, + "step": 12354 + }, + { + "epoch": 0.8489075168338601, + "grad_norm": 1.0828396081924438, + "learning_rate": 3.3974947879401476e-06, + "loss": 0.9538, + "step": 12355 + }, + { + "epoch": 0.8489762264669507, + "grad_norm": 1.0296262502670288, + "learning_rate": 3.394477342071542e-06, + "loss": 0.928, + "step": 12356 + }, + { + "epoch": 0.8490449361000413, + "grad_norm": 1.0949013233184814, + "learning_rate": 3.391461139155486e-06, + "loss": 0.9178, + "step": 12357 + }, + { + "epoch": 0.8491136457331317, + "grad_norm": 1.0867561101913452, + "learning_rate": 3.38844617936549e-06, + "loss": 0.9116, + "step": 12358 + }, + { + "epoch": 0.8491823553662223, + "grad_norm": 1.1055972576141357, + "learning_rate": 3.385432462875007e-06, + "loss": 0.928, + "step": 12359 + }, + { + "epoch": 0.8492510649993129, + "grad_norm": 1.0674328804016113, + "learning_rate": 3.382419989857422e-06, + "loss": 0.9335, + "step": 12360 + }, + { + "epoch": 0.8493197746324035, + "grad_norm": 1.0215785503387451, + "learning_rate": 3.3794087604860303e-06, + "loss": 0.943, + "step": 12361 + }, + { + "epoch": 0.8493884842654941, + "grad_norm": 1.0475304126739502, + "learning_rate": 3.3763987749340723e-06, + "loss": 0.955, + "step": 12362 + }, + { + "epoch": 0.8494571938985845, + "grad_norm": 1.0472979545593262, + "learning_rate": 3.3733900333747044e-06, + "loss": 0.9386, + "step": 12363 + }, + { + "epoch": 0.8495259035316751, + "grad_norm": 1.0362099409103394, + "learning_rate": 3.3703825359810202e-06, + "loss": 0.919, + "step": 12364 + }, + { + "epoch": 0.8495946131647657, + "grad_norm": 1.0591078996658325, + "learning_rate": 3.3673762829260407e-06, + "loss": 0.9221, + "step": 12365 + }, + { + "epoch": 0.8496633227978563, + "grad_norm": 1.049264669418335, + "learning_rate": 3.3643712743827133e-06, + "loss": 0.9244, + "step": 12366 + }, + { + "epoch": 0.8497320324309469, + "grad_norm": 1.1135157346725464, + "learning_rate": 3.361367510523922e-06, + "loss": 0.952, + "step": 12367 + }, + { + "epoch": 0.8498007420640373, + "grad_norm": 1.0415452718734741, + "learning_rate": 3.358364991522456e-06, + "loss": 0.9033, + "step": 12368 + }, + { + "epoch": 0.8498694516971279, + "grad_norm": 1.1484053134918213, + "learning_rate": 3.3553637175510577e-06, + "loss": 0.9322, + "step": 12369 + }, + { + "epoch": 0.8499381613302185, + "grad_norm": 1.033755898475647, + "learning_rate": 3.352363688782387e-06, + "loss": 0.9121, + "step": 12370 + }, + { + "epoch": 0.8500068709633091, + "grad_norm": 1.0500038862228394, + "learning_rate": 3.3493649053890326e-06, + "loss": 0.9341, + "step": 12371 + }, + { + "epoch": 0.8500755805963996, + "grad_norm": 1.0140495300292969, + "learning_rate": 3.346367367543521e-06, + "loss": 0.9289, + "step": 12372 + }, + { + "epoch": 0.8501442902294902, + "grad_norm": 1.0194281339645386, + "learning_rate": 3.3433710754182863e-06, + "loss": 0.9179, + "step": 12373 + }, + { + "epoch": 0.8502129998625807, + "grad_norm": 1.0062729120254517, + "learning_rate": 3.34037602918571e-06, + "loss": 0.9168, + "step": 12374 + }, + { + "epoch": 0.8502817094956713, + "grad_norm": 1.0193568468093872, + "learning_rate": 3.3373822290180933e-06, + "loss": 0.9579, + "step": 12375 + }, + { + "epoch": 0.8503504191287619, + "grad_norm": 1.117473840713501, + "learning_rate": 3.334389675087668e-06, + "loss": 0.9385, + "step": 12376 + }, + { + "epoch": 0.8504191287618524, + "grad_norm": 0.9885859489440918, + "learning_rate": 3.3313983675665986e-06, + "loss": 0.9122, + "step": 12377 + }, + { + "epoch": 0.850487838394943, + "grad_norm": 1.0656583309173584, + "learning_rate": 3.328408306626965e-06, + "loss": 0.9498, + "step": 12378 + }, + { + "epoch": 0.8505565480280335, + "grad_norm": 1.068516492843628, + "learning_rate": 3.32541949244079e-06, + "loss": 0.9197, + "step": 12379 + }, + { + "epoch": 0.8506252576611241, + "grad_norm": 1.0214465856552124, + "learning_rate": 3.322431925180014e-06, + "loss": 0.8975, + "step": 12380 + }, + { + "epoch": 0.8506939672942146, + "grad_norm": 1.0617576837539673, + "learning_rate": 3.319445605016511e-06, + "loss": 0.9663, + "step": 12381 + }, + { + "epoch": 0.8507626769273052, + "grad_norm": 1.0336326360702515, + "learning_rate": 3.316460532122087e-06, + "loss": 0.9378, + "step": 12382 + }, + { + "epoch": 0.8508313865603958, + "grad_norm": 1.0167787075042725, + "learning_rate": 3.313476706668464e-06, + "loss": 0.8865, + "step": 12383 + }, + { + "epoch": 0.8509000961934863, + "grad_norm": 1.0448201894760132, + "learning_rate": 3.31049412882731e-06, + "loss": 0.8991, + "step": 12384 + }, + { + "epoch": 0.8509688058265769, + "grad_norm": 1.05338716506958, + "learning_rate": 3.3075127987701958e-06, + "loss": 0.934, + "step": 12385 + }, + { + "epoch": 0.8510375154596674, + "grad_norm": 1.0663853883743286, + "learning_rate": 3.3045327166686457e-06, + "loss": 0.9611, + "step": 12386 + }, + { + "epoch": 0.851106225092758, + "grad_norm": 1.0753593444824219, + "learning_rate": 3.3015538826941e-06, + "loss": 0.9603, + "step": 12387 + }, + { + "epoch": 0.8511749347258486, + "grad_norm": 1.1009992361068726, + "learning_rate": 3.298576297017933e-06, + "loss": 0.9856, + "step": 12388 + }, + { + "epoch": 0.8512436443589392, + "grad_norm": 1.0413891077041626, + "learning_rate": 3.295599959811435e-06, + "loss": 0.9094, + "step": 12389 + }, + { + "epoch": 0.8513123539920296, + "grad_norm": 1.0967854261398315, + "learning_rate": 3.2926248712458362e-06, + "loss": 0.923, + "step": 12390 + }, + { + "epoch": 0.8513810636251202, + "grad_norm": 1.0926544666290283, + "learning_rate": 3.2896510314922942e-06, + "loss": 0.9002, + "step": 12391 + }, + { + "epoch": 0.8514497732582108, + "grad_norm": 1.068357229232788, + "learning_rate": 3.286678440721891e-06, + "loss": 0.8813, + "step": 12392 + }, + { + "epoch": 0.8515184828913014, + "grad_norm": 1.0758506059646606, + "learning_rate": 3.283707099105643e-06, + "loss": 0.9233, + "step": 12393 + }, + { + "epoch": 0.851587192524392, + "grad_norm": 1.0416373014450073, + "learning_rate": 3.280737006814477e-06, + "loss": 0.8784, + "step": 12394 + }, + { + "epoch": 0.8516559021574824, + "grad_norm": 1.0537190437316895, + "learning_rate": 3.2777681640192708e-06, + "loss": 0.9294, + "step": 12395 + }, + { + "epoch": 0.851724611790573, + "grad_norm": 1.0463240146636963, + "learning_rate": 3.2748005708908154e-06, + "loss": 0.9623, + "step": 12396 + }, + { + "epoch": 0.8517933214236636, + "grad_norm": 1.0585631132125854, + "learning_rate": 3.2718342275998377e-06, + "loss": 0.9643, + "step": 12397 + }, + { + "epoch": 0.8518620310567542, + "grad_norm": 1.1270244121551514, + "learning_rate": 3.268869134316996e-06, + "loss": 0.9501, + "step": 12398 + }, + { + "epoch": 0.8519307406898448, + "grad_norm": 1.0691008567810059, + "learning_rate": 3.2659052912128536e-06, + "loss": 0.9081, + "step": 12399 + }, + { + "epoch": 0.8519994503229352, + "grad_norm": 1.0352188348770142, + "learning_rate": 3.26294269845793e-06, + "loss": 0.9423, + "step": 12400 + }, + { + "epoch": 0.8520681599560258, + "grad_norm": 1.0226788520812988, + "learning_rate": 3.2599813562226577e-06, + "loss": 0.9255, + "step": 12401 + }, + { + "epoch": 0.8521368695891164, + "grad_norm": 1.0987892150878906, + "learning_rate": 3.257021264677404e-06, + "loss": 0.9591, + "step": 12402 + }, + { + "epoch": 0.852205579222207, + "grad_norm": 1.0334628820419312, + "learning_rate": 3.2540624239924626e-06, + "loss": 0.9345, + "step": 12403 + }, + { + "epoch": 0.8522742888552975, + "grad_norm": 1.0195302963256836, + "learning_rate": 3.251104834338048e-06, + "loss": 0.9059, + "step": 12404 + }, + { + "epoch": 0.8523429984883881, + "grad_norm": 1.0285470485687256, + "learning_rate": 3.2481484958843093e-06, + "loss": 0.918, + "step": 12405 + }, + { + "epoch": 0.8524117081214786, + "grad_norm": 1.030425786972046, + "learning_rate": 3.245193408801331e-06, + "loss": 0.9102, + "step": 12406 + }, + { + "epoch": 0.8524804177545692, + "grad_norm": 1.10539710521698, + "learning_rate": 3.2422395732591065e-06, + "loss": 0.9313, + "step": 12407 + }, + { + "epoch": 0.8525491273876598, + "grad_norm": 1.0358641147613525, + "learning_rate": 3.239286989427573e-06, + "loss": 0.9071, + "step": 12408 + }, + { + "epoch": 0.8526178370207503, + "grad_norm": 1.064664363861084, + "learning_rate": 3.236335657476594e-06, + "loss": 0.9013, + "step": 12409 + }, + { + "epoch": 0.8526865466538409, + "grad_norm": 1.0459851026535034, + "learning_rate": 3.233385577575951e-06, + "loss": 0.9352, + "step": 12410 + }, + { + "epoch": 0.8527552562869314, + "grad_norm": 1.0819168090820312, + "learning_rate": 3.230436749895363e-06, + "loss": 0.8984, + "step": 12411 + }, + { + "epoch": 0.852823965920022, + "grad_norm": 1.1116936206817627, + "learning_rate": 3.2274891746044757e-06, + "loss": 0.9358, + "step": 12412 + }, + { + "epoch": 0.8528926755531125, + "grad_norm": 1.0728542804718018, + "learning_rate": 3.2245428518728665e-06, + "loss": 0.917, + "step": 12413 + }, + { + "epoch": 0.8529613851862031, + "grad_norm": 1.1414625644683838, + "learning_rate": 3.2215977818700253e-06, + "loss": 0.97, + "step": 12414 + }, + { + "epoch": 0.8530300948192937, + "grad_norm": 1.064154028892517, + "learning_rate": 3.2186539647653837e-06, + "loss": 0.9344, + "step": 12415 + }, + { + "epoch": 0.8530988044523842, + "grad_norm": 1.006618857383728, + "learning_rate": 3.215711400728297e-06, + "loss": 0.9034, + "step": 12416 + }, + { + "epoch": 0.8531675140854748, + "grad_norm": 1.0414010286331177, + "learning_rate": 3.2127700899280523e-06, + "loss": 0.9689, + "step": 12417 + }, + { + "epoch": 0.8532362237185653, + "grad_norm": 1.111560344696045, + "learning_rate": 3.209830032533867e-06, + "loss": 0.9224, + "step": 12418 + }, + { + "epoch": 0.8533049333516559, + "grad_norm": 1.0432631969451904, + "learning_rate": 3.2068912287148646e-06, + "loss": 0.9927, + "step": 12419 + }, + { + "epoch": 0.8533736429847465, + "grad_norm": 1.0252516269683838, + "learning_rate": 3.2039536786401257e-06, + "loss": 0.9203, + "step": 12420 + }, + { + "epoch": 0.8534423526178371, + "grad_norm": 1.0749858617782593, + "learning_rate": 3.201017382478638e-06, + "loss": 0.9106, + "step": 12421 + }, + { + "epoch": 0.8535110622509275, + "grad_norm": 1.0088677406311035, + "learning_rate": 3.198082340399333e-06, + "loss": 0.9169, + "step": 12422 + }, + { + "epoch": 0.8535797718840181, + "grad_norm": 1.0308411121368408, + "learning_rate": 3.195148552571059e-06, + "loss": 0.9263, + "step": 12423 + }, + { + "epoch": 0.8536484815171087, + "grad_norm": 1.0378782749176025, + "learning_rate": 3.1922160191625917e-06, + "loss": 0.9436, + "step": 12424 + }, + { + "epoch": 0.8537171911501993, + "grad_norm": 1.0742977857589722, + "learning_rate": 3.1892847403426378e-06, + "loss": 0.9367, + "step": 12425 + }, + { + "epoch": 0.8537859007832899, + "grad_norm": 1.0263398885726929, + "learning_rate": 3.1863547162798353e-06, + "loss": 0.9135, + "step": 12426 + }, + { + "epoch": 0.8538546104163803, + "grad_norm": 0.9921544194221497, + "learning_rate": 3.183425947142743e-06, + "loss": 0.9178, + "step": 12427 + }, + { + "epoch": 0.8539233200494709, + "grad_norm": 1.0076355934143066, + "learning_rate": 3.1804984330998627e-06, + "loss": 0.8929, + "step": 12428 + }, + { + "epoch": 0.8539920296825615, + "grad_norm": 1.0498324632644653, + "learning_rate": 3.1775721743195956e-06, + "loss": 0.9001, + "step": 12429 + }, + { + "epoch": 0.8540607393156521, + "grad_norm": 1.1073228120803833, + "learning_rate": 3.1746471709702964e-06, + "loss": 0.9318, + "step": 12430 + }, + { + "epoch": 0.8541294489487427, + "grad_norm": 1.0791555643081665, + "learning_rate": 3.171723423220241e-06, + "loss": 0.9198, + "step": 12431 + }, + { + "epoch": 0.8541981585818331, + "grad_norm": 1.0533887147903442, + "learning_rate": 3.1688009312376255e-06, + "loss": 0.9119, + "step": 12432 + }, + { + "epoch": 0.8542668682149237, + "grad_norm": 1.0684369802474976, + "learning_rate": 3.165879695190585e-06, + "loss": 0.9151, + "step": 12433 + }, + { + "epoch": 0.8543355778480143, + "grad_norm": 1.096433162689209, + "learning_rate": 3.1629597152471656e-06, + "loss": 0.9258, + "step": 12434 + }, + { + "epoch": 0.8544042874811049, + "grad_norm": 0.9714148044586182, + "learning_rate": 3.160040991575361e-06, + "loss": 0.8907, + "step": 12435 + }, + { + "epoch": 0.8544729971141954, + "grad_norm": 1.0394467115402222, + "learning_rate": 3.1571235243430835e-06, + "loss": 0.8962, + "step": 12436 + }, + { + "epoch": 0.854541706747286, + "grad_norm": 1.028713583946228, + "learning_rate": 3.1542073137181687e-06, + "loss": 0.9158, + "step": 12437 + }, + { + "epoch": 0.8546104163803765, + "grad_norm": 1.0966624021530151, + "learning_rate": 3.151292359868388e-06, + "loss": 0.9624, + "step": 12438 + }, + { + "epoch": 0.8546791260134671, + "grad_norm": 1.0287033319473267, + "learning_rate": 3.1483786629614405e-06, + "loss": 0.9506, + "step": 12439 + }, + { + "epoch": 0.8547478356465577, + "grad_norm": 1.0199686288833618, + "learning_rate": 3.145466223164939e-06, + "loss": 0.8851, + "step": 12440 + }, + { + "epoch": 0.8548165452796482, + "grad_norm": 1.1116775274276733, + "learning_rate": 3.1425550406464417e-06, + "loss": 0.9493, + "step": 12441 + }, + { + "epoch": 0.8548852549127388, + "grad_norm": 1.1445835828781128, + "learning_rate": 3.139645115573425e-06, + "loss": 0.9541, + "step": 12442 + }, + { + "epoch": 0.8549539645458293, + "grad_norm": 1.0601056814193726, + "learning_rate": 3.136736448113295e-06, + "loss": 0.9245, + "step": 12443 + }, + { + "epoch": 0.8550226741789199, + "grad_norm": 1.0296072959899902, + "learning_rate": 3.1338290384333946e-06, + "loss": 0.9064, + "step": 12444 + }, + { + "epoch": 0.8550913838120104, + "grad_norm": 1.0869626998901367, + "learning_rate": 3.1309228867009682e-06, + "loss": 0.8831, + "step": 12445 + }, + { + "epoch": 0.855160093445101, + "grad_norm": 1.0973420143127441, + "learning_rate": 3.1280179930832154e-06, + "loss": 0.9374, + "step": 12446 + }, + { + "epoch": 0.8552288030781916, + "grad_norm": 1.019338607788086, + "learning_rate": 3.1251143577472523e-06, + "loss": 0.9217, + "step": 12447 + }, + { + "epoch": 0.8552975127112821, + "grad_norm": 1.0525734424591064, + "learning_rate": 3.1222119808601207e-06, + "loss": 0.9379, + "step": 12448 + }, + { + "epoch": 0.8553662223443727, + "grad_norm": 1.0251481533050537, + "learning_rate": 3.1193108625888007e-06, + "loss": 0.906, + "step": 12449 + }, + { + "epoch": 0.8554349319774632, + "grad_norm": 1.021919846534729, + "learning_rate": 3.116411003100181e-06, + "loss": 0.9177, + "step": 12450 + }, + { + "epoch": 0.8555036416105538, + "grad_norm": 1.0407745838165283, + "learning_rate": 3.113512402561092e-06, + "loss": 0.9291, + "step": 12451 + }, + { + "epoch": 0.8555723512436444, + "grad_norm": 1.1090978384017944, + "learning_rate": 3.110615061138289e-06, + "loss": 0.9464, + "step": 12452 + }, + { + "epoch": 0.855641060876735, + "grad_norm": 1.014161229133606, + "learning_rate": 3.107718978998461e-06, + "loss": 0.8763, + "step": 12453 + }, + { + "epoch": 0.8557097705098254, + "grad_norm": 1.0602693557739258, + "learning_rate": 3.104824156308206e-06, + "loss": 0.9122, + "step": 12454 + }, + { + "epoch": 0.855778480142916, + "grad_norm": 0.9977532625198364, + "learning_rate": 3.101930593234073e-06, + "loss": 0.9044, + "step": 12455 + }, + { + "epoch": 0.8558471897760066, + "grad_norm": 1.053736686706543, + "learning_rate": 3.0990382899425158e-06, + "loss": 0.8998, + "step": 12456 + }, + { + "epoch": 0.8559158994090972, + "grad_norm": 1.0920228958129883, + "learning_rate": 3.0961472465999314e-06, + "loss": 0.9378, + "step": 12457 + }, + { + "epoch": 0.8559846090421878, + "grad_norm": 1.0449419021606445, + "learning_rate": 3.0932574633726395e-06, + "loss": 0.9427, + "step": 12458 + }, + { + "epoch": 0.8560533186752782, + "grad_norm": 1.0214542150497437, + "learning_rate": 3.0903689404268937e-06, + "loss": 0.9275, + "step": 12459 + }, + { + "epoch": 0.8561220283083688, + "grad_norm": 1.020971417427063, + "learning_rate": 3.0874816779288606e-06, + "loss": 0.9222, + "step": 12460 + }, + { + "epoch": 0.8561907379414594, + "grad_norm": 1.0384490489959717, + "learning_rate": 3.0845956760446438e-06, + "loss": 0.9411, + "step": 12461 + }, + { + "epoch": 0.85625944757455, + "grad_norm": 1.0637458562850952, + "learning_rate": 3.0817109349402766e-06, + "loss": 0.965, + "step": 12462 + }, + { + "epoch": 0.8563281572076406, + "grad_norm": 1.0444003343582153, + "learning_rate": 3.078827454781713e-06, + "loss": 0.952, + "step": 12463 + }, + { + "epoch": 0.856396866840731, + "grad_norm": 1.102221131324768, + "learning_rate": 3.075945235734848e-06, + "loss": 0.9726, + "step": 12464 + }, + { + "epoch": 0.8564655764738216, + "grad_norm": 1.0652331113815308, + "learning_rate": 3.073064277965479e-06, + "loss": 0.8754, + "step": 12465 + }, + { + "epoch": 0.8565342861069122, + "grad_norm": 1.0645511150360107, + "learning_rate": 3.0701845816393516e-06, + "loss": 0.9557, + "step": 12466 + }, + { + "epoch": 0.8566029957400028, + "grad_norm": 1.0397380590438843, + "learning_rate": 3.067306146922136e-06, + "loss": 0.9345, + "step": 12467 + }, + { + "epoch": 0.8566717053730933, + "grad_norm": 1.0005074739456177, + "learning_rate": 3.064428973979422e-06, + "loss": 0.938, + "step": 12468 + }, + { + "epoch": 0.8567404150061839, + "grad_norm": 1.0726656913757324, + "learning_rate": 3.0615530629767406e-06, + "loss": 0.9249, + "step": 12469 + }, + { + "epoch": 0.8568091246392744, + "grad_norm": 1.0726382732391357, + "learning_rate": 3.05867841407953e-06, + "loss": 0.93, + "step": 12470 + }, + { + "epoch": 0.856877834272365, + "grad_norm": 1.1153554916381836, + "learning_rate": 3.0558050274531734e-06, + "loss": 0.9105, + "step": 12471 + }, + { + "epoch": 0.8569465439054555, + "grad_norm": 1.0676429271697998, + "learning_rate": 3.05293290326297e-06, + "loss": 0.933, + "step": 12472 + }, + { + "epoch": 0.8570152535385461, + "grad_norm": 1.0443394184112549, + "learning_rate": 3.050062041674159e-06, + "loss": 0.899, + "step": 12473 + }, + { + "epoch": 0.8570839631716367, + "grad_norm": 1.0928031206130981, + "learning_rate": 3.0471924428518984e-06, + "loss": 0.9122, + "step": 12474 + }, + { + "epoch": 0.8571526728047272, + "grad_norm": 1.1017603874206543, + "learning_rate": 3.044324106961266e-06, + "loss": 0.9401, + "step": 12475 + }, + { + "epoch": 0.8572213824378178, + "grad_norm": 1.0890322923660278, + "learning_rate": 3.0414570341672803e-06, + "loss": 0.962, + "step": 12476 + }, + { + "epoch": 0.8572900920709083, + "grad_norm": 1.0616599321365356, + "learning_rate": 3.038591224634882e-06, + "loss": 0.9399, + "step": 12477 + }, + { + "epoch": 0.8573588017039989, + "grad_norm": 1.0718921422958374, + "learning_rate": 3.0357266785289467e-06, + "loss": 0.9563, + "step": 12478 + }, + { + "epoch": 0.8574275113370895, + "grad_norm": 1.09958016872406, + "learning_rate": 3.0328633960142573e-06, + "loss": 0.9545, + "step": 12479 + }, + { + "epoch": 0.85749622097018, + "grad_norm": 1.1745871305465698, + "learning_rate": 3.030001377255545e-06, + "loss": 0.9334, + "step": 12480 + }, + { + "epoch": 0.8575649306032705, + "grad_norm": 1.0835914611816406, + "learning_rate": 3.027140622417457e-06, + "loss": 0.9489, + "step": 12481 + }, + { + "epoch": 0.8576336402363611, + "grad_norm": 1.0603687763214111, + "learning_rate": 3.024281131664569e-06, + "loss": 0.9143, + "step": 12482 + }, + { + "epoch": 0.8577023498694517, + "grad_norm": 1.0256012678146362, + "learning_rate": 3.021422905161389e-06, + "loss": 0.8925, + "step": 12483 + }, + { + "epoch": 0.8577710595025423, + "grad_norm": 1.0399523973464966, + "learning_rate": 3.018565943072352e-06, + "loss": 0.9137, + "step": 12484 + }, + { + "epoch": 0.8578397691356329, + "grad_norm": 1.1051543951034546, + "learning_rate": 3.015710245561809e-06, + "loss": 0.8849, + "step": 12485 + }, + { + "epoch": 0.8579084787687233, + "grad_norm": 1.0705151557922363, + "learning_rate": 3.012855812794052e-06, + "loss": 0.9297, + "step": 12486 + }, + { + "epoch": 0.8579771884018139, + "grad_norm": 1.1427122354507446, + "learning_rate": 3.0100026449332933e-06, + "loss": 0.9262, + "step": 12487 + }, + { + "epoch": 0.8580458980349045, + "grad_norm": 1.021746277809143, + "learning_rate": 3.0071507421436746e-06, + "loss": 0.8866, + "step": 12488 + }, + { + "epoch": 0.8581146076679951, + "grad_norm": 0.9893001914024353, + "learning_rate": 3.0043001045892695e-06, + "loss": 0.9024, + "step": 12489 + }, + { + "epoch": 0.8581833173010857, + "grad_norm": 1.0987331867218018, + "learning_rate": 3.0014507324340607e-06, + "loss": 0.9516, + "step": 12490 + }, + { + "epoch": 0.8582520269341761, + "grad_norm": 1.0792369842529297, + "learning_rate": 2.9986026258419813e-06, + "loss": 0.9524, + "step": 12491 + }, + { + "epoch": 0.8583207365672667, + "grad_norm": 1.0668972730636597, + "learning_rate": 2.9957557849768754e-06, + "loss": 0.9308, + "step": 12492 + }, + { + "epoch": 0.8583894462003573, + "grad_norm": 1.0237042903900146, + "learning_rate": 2.9929102100025226e-06, + "loss": 0.9207, + "step": 12493 + }, + { + "epoch": 0.8584581558334479, + "grad_norm": 1.0875829458236694, + "learning_rate": 2.9900659010826283e-06, + "loss": 0.9174, + "step": 12494 + }, + { + "epoch": 0.8585268654665384, + "grad_norm": 0.9976902008056641, + "learning_rate": 2.9872228583808252e-06, + "loss": 0.932, + "step": 12495 + }, + { + "epoch": 0.8585955750996289, + "grad_norm": 1.1306830644607544, + "learning_rate": 2.9843810820606666e-06, + "loss": 0.9548, + "step": 12496 + }, + { + "epoch": 0.8586642847327195, + "grad_norm": 1.0161389112472534, + "learning_rate": 2.9815405722856402e-06, + "loss": 0.9528, + "step": 12497 + }, + { + "epoch": 0.8587329943658101, + "grad_norm": 1.0810953378677368, + "learning_rate": 2.978701329219158e-06, + "loss": 0.9518, + "step": 12498 + }, + { + "epoch": 0.8588017039989007, + "grad_norm": 1.0911314487457275, + "learning_rate": 2.9758633530245607e-06, + "loss": 0.9398, + "step": 12499 + }, + { + "epoch": 0.8588704136319912, + "grad_norm": 1.107388973236084, + "learning_rate": 2.9730266438651234e-06, + "loss": 0.9478, + "step": 12500 + }, + { + "epoch": 0.8589391232650818, + "grad_norm": 1.0737265348434448, + "learning_rate": 2.970191201904027e-06, + "loss": 0.917, + "step": 12501 + }, + { + "epoch": 0.8590078328981723, + "grad_norm": 1.008888602256775, + "learning_rate": 2.9673570273044014e-06, + "loss": 0.8963, + "step": 12502 + }, + { + "epoch": 0.8590765425312629, + "grad_norm": 1.0683518648147583, + "learning_rate": 2.9645241202292897e-06, + "loss": 0.9476, + "step": 12503 + }, + { + "epoch": 0.8591452521643534, + "grad_norm": 1.0710750818252563, + "learning_rate": 2.9616924808416685e-06, + "loss": 0.9598, + "step": 12504 + }, + { + "epoch": 0.859213961797444, + "grad_norm": 1.0175212621688843, + "learning_rate": 2.958862109304447e-06, + "loss": 0.9179, + "step": 12505 + }, + { + "epoch": 0.8592826714305346, + "grad_norm": 1.0312576293945312, + "learning_rate": 2.9560330057804425e-06, + "loss": 0.9281, + "step": 12506 + }, + { + "epoch": 0.8593513810636251, + "grad_norm": 1.0097825527191162, + "learning_rate": 2.9532051704324186e-06, + "loss": 0.9044, + "step": 12507 + }, + { + "epoch": 0.8594200906967157, + "grad_norm": 1.053127408027649, + "learning_rate": 2.9503786034230597e-06, + "loss": 0.9156, + "step": 12508 + }, + { + "epoch": 0.8594888003298062, + "grad_norm": 1.0358312129974365, + "learning_rate": 2.947553304914974e-06, + "loss": 0.9169, + "step": 12509 + }, + { + "epoch": 0.8595575099628968, + "grad_norm": 1.0573961734771729, + "learning_rate": 2.944729275070707e-06, + "loss": 0.889, + "step": 12510 + }, + { + "epoch": 0.8596262195959874, + "grad_norm": 0.9851250052452087, + "learning_rate": 2.941906514052711e-06, + "loss": 0.9167, + "step": 12511 + }, + { + "epoch": 0.8596949292290779, + "grad_norm": 1.0720226764678955, + "learning_rate": 2.9390850220233858e-06, + "loss": 0.8752, + "step": 12512 + }, + { + "epoch": 0.8597636388621684, + "grad_norm": 1.0977139472961426, + "learning_rate": 2.9362647991450443e-06, + "loss": 0.9227, + "step": 12513 + }, + { + "epoch": 0.859832348495259, + "grad_norm": 1.0839402675628662, + "learning_rate": 2.933445845579938e-06, + "loss": 0.9187, + "step": 12514 + }, + { + "epoch": 0.8599010581283496, + "grad_norm": 1.1230579614639282, + "learning_rate": 2.9306281614902427e-06, + "loss": 0.9312, + "step": 12515 + }, + { + "epoch": 0.8599697677614402, + "grad_norm": 1.0702879428863525, + "learning_rate": 2.927811747038048e-06, + "loss": 0.9598, + "step": 12516 + }, + { + "epoch": 0.8600384773945308, + "grad_norm": 1.1174166202545166, + "learning_rate": 2.9249966023853853e-06, + "loss": 0.9508, + "step": 12517 + }, + { + "epoch": 0.8601071870276212, + "grad_norm": 1.1639851331710815, + "learning_rate": 2.922182727694206e-06, + "loss": 0.9629, + "step": 12518 + }, + { + "epoch": 0.8601758966607118, + "grad_norm": 1.0832562446594238, + "learning_rate": 2.919370123126397e-06, + "loss": 0.9105, + "step": 12519 + }, + { + "epoch": 0.8602446062938024, + "grad_norm": 0.9917428493499756, + "learning_rate": 2.9165587888437623e-06, + "loss": 0.8803, + "step": 12520 + }, + { + "epoch": 0.860313315926893, + "grad_norm": 1.0718384981155396, + "learning_rate": 2.9137487250080335e-06, + "loss": 0.9388, + "step": 12521 + }, + { + "epoch": 0.8603820255599836, + "grad_norm": 1.0817455053329468, + "learning_rate": 2.9109399317808734e-06, + "loss": 0.9521, + "step": 12522 + }, + { + "epoch": 0.860450735193074, + "grad_norm": 1.0382680892944336, + "learning_rate": 2.908132409323869e-06, + "loss": 0.9506, + "step": 12523 + }, + { + "epoch": 0.8605194448261646, + "grad_norm": 1.0884350538253784, + "learning_rate": 2.9053261577985436e-06, + "loss": 0.968, + "step": 12524 + }, + { + "epoch": 0.8605881544592552, + "grad_norm": 1.0268301963806152, + "learning_rate": 2.902521177366324e-06, + "loss": 0.8933, + "step": 12525 + }, + { + "epoch": 0.8606568640923458, + "grad_norm": 1.0578782558441162, + "learning_rate": 2.8997174681885925e-06, + "loss": 0.9195, + "step": 12526 + }, + { + "epoch": 0.8607255737254363, + "grad_norm": 1.0370794534683228, + "learning_rate": 2.896915030426642e-06, + "loss": 0.945, + "step": 12527 + }, + { + "epoch": 0.8607942833585268, + "grad_norm": 1.033591389656067, + "learning_rate": 2.8941138642416876e-06, + "loss": 0.8924, + "step": 12528 + }, + { + "epoch": 0.8608629929916174, + "grad_norm": 1.0548501014709473, + "learning_rate": 2.8913139697948844e-06, + "loss": 0.9232, + "step": 12529 + }, + { + "epoch": 0.860931702624708, + "grad_norm": 1.0553154945373535, + "learning_rate": 2.8885153472473144e-06, + "loss": 0.9363, + "step": 12530 + }, + { + "epoch": 0.8610004122577986, + "grad_norm": 1.0652521848678589, + "learning_rate": 2.885717996759968e-06, + "loss": 0.9523, + "step": 12531 + }, + { + "epoch": 0.8610691218908891, + "grad_norm": 1.060697317123413, + "learning_rate": 2.8829219184937805e-06, + "loss": 0.9536, + "step": 12532 + }, + { + "epoch": 0.8611378315239797, + "grad_norm": 1.0484447479248047, + "learning_rate": 2.8801271126096124e-06, + "loss": 0.9022, + "step": 12533 + }, + { + "epoch": 0.8612065411570702, + "grad_norm": 1.075678825378418, + "learning_rate": 2.8773335792682404e-06, + "loss": 0.8873, + "step": 12534 + }, + { + "epoch": 0.8612752507901608, + "grad_norm": 1.0912504196166992, + "learning_rate": 2.874541318630386e-06, + "loss": 0.9218, + "step": 12535 + }, + { + "epoch": 0.8613439604232513, + "grad_norm": 1.068310022354126, + "learning_rate": 2.8717503308566732e-06, + "loss": 0.9371, + "step": 12536 + }, + { + "epoch": 0.8614126700563419, + "grad_norm": 1.0857332944869995, + "learning_rate": 2.8689606161076742e-06, + "loss": 0.9727, + "step": 12537 + }, + { + "epoch": 0.8614813796894325, + "grad_norm": 1.038854956626892, + "learning_rate": 2.866172174543874e-06, + "loss": 0.9169, + "step": 12538 + }, + { + "epoch": 0.861550089322523, + "grad_norm": 1.0570541620254517, + "learning_rate": 2.8633850063256943e-06, + "loss": 0.9565, + "step": 12539 + }, + { + "epoch": 0.8616187989556136, + "grad_norm": 1.0485256910324097, + "learning_rate": 2.8605991116134823e-06, + "loss": 0.8898, + "step": 12540 + }, + { + "epoch": 0.8616875085887041, + "grad_norm": 1.0271718502044678, + "learning_rate": 2.857814490567498e-06, + "loss": 0.9115, + "step": 12541 + }, + { + "epoch": 0.8617562182217947, + "grad_norm": 1.0815033912658691, + "learning_rate": 2.855031143347947e-06, + "loss": 0.9388, + "step": 12542 + }, + { + "epoch": 0.8618249278548853, + "grad_norm": 1.1187984943389893, + "learning_rate": 2.852249070114951e-06, + "loss": 0.9605, + "step": 12543 + }, + { + "epoch": 0.8618936374879758, + "grad_norm": 1.060380220413208, + "learning_rate": 2.849468271028563e-06, + "loss": 0.9275, + "step": 12544 + }, + { + "epoch": 0.8619623471210663, + "grad_norm": 1.0568920373916626, + "learning_rate": 2.846688746248763e-06, + "loss": 0.923, + "step": 12545 + }, + { + "epoch": 0.8620310567541569, + "grad_norm": 1.0588809251785278, + "learning_rate": 2.843910495935448e-06, + "loss": 0.972, + "step": 12546 + }, + { + "epoch": 0.8620997663872475, + "grad_norm": 1.0441548824310303, + "learning_rate": 2.841133520248451e-06, + "loss": 0.9141, + "step": 12547 + }, + { + "epoch": 0.8621684760203381, + "grad_norm": 1.0836669206619263, + "learning_rate": 2.8383578193475315e-06, + "loss": 0.9473, + "step": 12548 + }, + { + "epoch": 0.8622371856534287, + "grad_norm": 1.0538586378097534, + "learning_rate": 2.8355833933923798e-06, + "loss": 0.9342, + "step": 12549 + }, + { + "epoch": 0.8623058952865191, + "grad_norm": 1.0033296346664429, + "learning_rate": 2.832810242542597e-06, + "loss": 0.9131, + "step": 12550 + }, + { + "epoch": 0.8623746049196097, + "grad_norm": 1.0087082386016846, + "learning_rate": 2.8300383669577296e-06, + "loss": 0.8867, + "step": 12551 + }, + { + "epoch": 0.8624433145527003, + "grad_norm": 1.0281691551208496, + "learning_rate": 2.8272677667972314e-06, + "loss": 0.9113, + "step": 12552 + }, + { + "epoch": 0.8625120241857909, + "grad_norm": 1.1367111206054688, + "learning_rate": 2.8244984422204995e-06, + "loss": 0.9957, + "step": 12553 + }, + { + "epoch": 0.8625807338188815, + "grad_norm": 1.1269606351852417, + "learning_rate": 2.821730393386851e-06, + "loss": 0.9791, + "step": 12554 + }, + { + "epoch": 0.8626494434519719, + "grad_norm": 1.0790257453918457, + "learning_rate": 2.8189636204555307e-06, + "loss": 0.9243, + "step": 12555 + }, + { + "epoch": 0.8627181530850625, + "grad_norm": 1.0233752727508545, + "learning_rate": 2.8161981235857143e-06, + "loss": 0.8997, + "step": 12556 + }, + { + "epoch": 0.8627868627181531, + "grad_norm": 1.1004540920257568, + "learning_rate": 2.813433902936488e-06, + "loss": 0.9417, + "step": 12557 + }, + { + "epoch": 0.8628555723512437, + "grad_norm": 1.0249254703521729, + "learning_rate": 2.8106709586668804e-06, + "loss": 0.9189, + "step": 12558 + }, + { + "epoch": 0.8629242819843342, + "grad_norm": 1.0315186977386475, + "learning_rate": 2.807909290935845e-06, + "loss": 0.9131, + "step": 12559 + }, + { + "epoch": 0.8629929916174247, + "grad_norm": 1.0453499555587769, + "learning_rate": 2.8051488999022545e-06, + "loss": 0.9278, + "step": 12560 + }, + { + "epoch": 0.8630617012505153, + "grad_norm": 1.1122685670852661, + "learning_rate": 2.802389785724921e-06, + "loss": 0.9495, + "step": 12561 + }, + { + "epoch": 0.8631304108836059, + "grad_norm": 1.1044646501541138, + "learning_rate": 2.7996319485625642e-06, + "loss": 0.9287, + "step": 12562 + }, + { + "epoch": 0.8631991205166964, + "grad_norm": 1.112377405166626, + "learning_rate": 2.796875388573844e-06, + "loss": 0.94, + "step": 12563 + }, + { + "epoch": 0.863267830149787, + "grad_norm": 1.0954900979995728, + "learning_rate": 2.7941201059173444e-06, + "loss": 0.916, + "step": 12564 + }, + { + "epoch": 0.8633365397828776, + "grad_norm": 1.050822138786316, + "learning_rate": 2.7913661007515773e-06, + "loss": 0.9638, + "step": 12565 + }, + { + "epoch": 0.8634052494159681, + "grad_norm": 1.081571340560913, + "learning_rate": 2.78861337323498e-06, + "loss": 0.9247, + "step": 12566 + }, + { + "epoch": 0.8634739590490587, + "grad_norm": 1.0439823865890503, + "learning_rate": 2.785861923525909e-06, + "loss": 0.9125, + "step": 12567 + }, + { + "epoch": 0.8635426686821492, + "grad_norm": 1.0489943027496338, + "learning_rate": 2.7831117517826576e-06, + "loss": 0.9154, + "step": 12568 + }, + { + "epoch": 0.8636113783152398, + "grad_norm": 1.072309136390686, + "learning_rate": 2.780362858163438e-06, + "loss": 0.9537, + "step": 12569 + }, + { + "epoch": 0.8636800879483304, + "grad_norm": 1.118922472000122, + "learning_rate": 2.7776152428263953e-06, + "loss": 0.9208, + "step": 12570 + }, + { + "epoch": 0.8637487975814209, + "grad_norm": 1.1350466012954712, + "learning_rate": 2.7748689059296064e-06, + "loss": 0.9331, + "step": 12571 + }, + { + "epoch": 0.8638175072145114, + "grad_norm": 1.1012707948684692, + "learning_rate": 2.77212384763105e-06, + "loss": 0.9139, + "step": 12572 + }, + { + "epoch": 0.863886216847602, + "grad_norm": 1.0364418029785156, + "learning_rate": 2.7693800680886617e-06, + "loss": 0.9255, + "step": 12573 + }, + { + "epoch": 0.8639549264806926, + "grad_norm": 1.0506123304367065, + "learning_rate": 2.7666375674602755e-06, + "loss": 0.9364, + "step": 12574 + }, + { + "epoch": 0.8640236361137832, + "grad_norm": 1.0587756633758545, + "learning_rate": 2.763896345903677e-06, + "loss": 0.9379, + "step": 12575 + }, + { + "epoch": 0.8640923457468737, + "grad_norm": 1.0893889665603638, + "learning_rate": 2.7611564035765676e-06, + "loss": 0.9475, + "step": 12576 + }, + { + "epoch": 0.8641610553799642, + "grad_norm": 1.1306151151657104, + "learning_rate": 2.7584177406365653e-06, + "loss": 0.9805, + "step": 12577 + }, + { + "epoch": 0.8642297650130548, + "grad_norm": 1.0618807077407837, + "learning_rate": 2.7556803572412303e-06, + "loss": 0.9064, + "step": 12578 + }, + { + "epoch": 0.8642984746461454, + "grad_norm": 1.0355290174484253, + "learning_rate": 2.75294425354804e-06, + "loss": 0.892, + "step": 12579 + }, + { + "epoch": 0.864367184279236, + "grad_norm": 1.039946436882019, + "learning_rate": 2.750209429714401e-06, + "loss": 0.9326, + "step": 12580 + }, + { + "epoch": 0.8644358939123266, + "grad_norm": 1.0923621654510498, + "learning_rate": 2.7474758858976546e-06, + "loss": 0.9293, + "step": 12581 + }, + { + "epoch": 0.864504603545417, + "grad_norm": 1.0525271892547607, + "learning_rate": 2.7447436222550444e-06, + "loss": 0.9074, + "step": 12582 + }, + { + "epoch": 0.8645733131785076, + "grad_norm": 1.0574243068695068, + "learning_rate": 2.7420126389437667e-06, + "loss": 0.9484, + "step": 12583 + }, + { + "epoch": 0.8646420228115982, + "grad_norm": 1.0783203840255737, + "learning_rate": 2.7392829361209293e-06, + "loss": 0.923, + "step": 12584 + }, + { + "epoch": 0.8647107324446888, + "grad_norm": 1.1129475831985474, + "learning_rate": 2.7365545139435704e-06, + "loss": 0.9087, + "step": 12585 + }, + { + "epoch": 0.8647794420777793, + "grad_norm": 0.9738619923591614, + "learning_rate": 2.733827372568662e-06, + "loss": 0.8993, + "step": 12586 + }, + { + "epoch": 0.8648481517108698, + "grad_norm": 1.0878139734268188, + "learning_rate": 2.731101512153084e-06, + "loss": 0.924, + "step": 12587 + }, + { + "epoch": 0.8649168613439604, + "grad_norm": 1.0834236145019531, + "learning_rate": 2.7283769328536544e-06, + "loss": 0.9173, + "step": 12588 + }, + { + "epoch": 0.864985570977051, + "grad_norm": 1.074680209159851, + "learning_rate": 2.7256536348271244e-06, + "loss": 0.9589, + "step": 12589 + }, + { + "epoch": 0.8650542806101416, + "grad_norm": 1.0062614679336548, + "learning_rate": 2.7229316182301566e-06, + "loss": 0.9007, + "step": 12590 + }, + { + "epoch": 0.8651229902432321, + "grad_norm": 1.0578126907348633, + "learning_rate": 2.720210883219357e-06, + "loss": 0.9256, + "step": 12591 + }, + { + "epoch": 0.8651916998763226, + "grad_norm": 1.132086157798767, + "learning_rate": 2.7174914299512328e-06, + "loss": 0.9464, + "step": 12592 + }, + { + "epoch": 0.8652604095094132, + "grad_norm": 1.0777111053466797, + "learning_rate": 2.714773258582243e-06, + "loss": 0.921, + "step": 12593 + }, + { + "epoch": 0.8653291191425038, + "grad_norm": 1.0626912117004395, + "learning_rate": 2.712056369268756e-06, + "loss": 0.9351, + "step": 12594 + }, + { + "epoch": 0.8653978287755943, + "grad_norm": 1.0115163326263428, + "learning_rate": 2.709340762167084e-06, + "loss": 0.9161, + "step": 12595 + }, + { + "epoch": 0.8654665384086849, + "grad_norm": 1.0965784788131714, + "learning_rate": 2.7066264374334434e-06, + "loss": 0.9631, + "step": 12596 + }, + { + "epoch": 0.8655352480417755, + "grad_norm": 1.0261839628219604, + "learning_rate": 2.7039133952239865e-06, + "loss": 0.8976, + "step": 12597 + }, + { + "epoch": 0.865603957674866, + "grad_norm": 1.010032296180725, + "learning_rate": 2.7012016356948056e-06, + "loss": 0.9386, + "step": 12598 + }, + { + "epoch": 0.8656726673079566, + "grad_norm": 1.06666898727417, + "learning_rate": 2.6984911590018926e-06, + "loss": 0.9287, + "step": 12599 + }, + { + "epoch": 0.8657413769410471, + "grad_norm": 1.0036234855651855, + "learning_rate": 2.695781965301186e-06, + "loss": 0.9267, + "step": 12600 + }, + { + "epoch": 0.8658100865741377, + "grad_norm": 1.0620659589767456, + "learning_rate": 2.69307405474854e-06, + "loss": 0.9167, + "step": 12601 + }, + { + "epoch": 0.8658787962072283, + "grad_norm": 1.0507009029388428, + "learning_rate": 2.6903674274997514e-06, + "loss": 0.8857, + "step": 12602 + }, + { + "epoch": 0.8659475058403188, + "grad_norm": 1.016994595527649, + "learning_rate": 2.6876620837105153e-06, + "loss": 0.9225, + "step": 12603 + }, + { + "epoch": 0.8660162154734093, + "grad_norm": 1.0535600185394287, + "learning_rate": 2.6849580235364743e-06, + "loss": 0.9347, + "step": 12604 + }, + { + "epoch": 0.8660849251064999, + "grad_norm": 1.0323255062103271, + "learning_rate": 2.682255247133192e-06, + "loss": 0.9039, + "step": 12605 + }, + { + "epoch": 0.8661536347395905, + "grad_norm": 1.103468418121338, + "learning_rate": 2.679553754656158e-06, + "loss": 1.0289, + "step": 12606 + }, + { + "epoch": 0.8662223443726811, + "grad_norm": 1.06203293800354, + "learning_rate": 2.676853546260791e-06, + "loss": 0.9373, + "step": 12607 + }, + { + "epoch": 0.8662910540057716, + "grad_norm": 1.0488885641098022, + "learning_rate": 2.6741546221024233e-06, + "loss": 0.9552, + "step": 12608 + }, + { + "epoch": 0.8663597636388621, + "grad_norm": 1.0035456418991089, + "learning_rate": 2.671456982336329e-06, + "loss": 0.8781, + "step": 12609 + }, + { + "epoch": 0.8664284732719527, + "grad_norm": 1.1242297887802124, + "learning_rate": 2.6687606271177e-06, + "loss": 0.9832, + "step": 12610 + }, + { + "epoch": 0.8664971829050433, + "grad_norm": 1.1403809785842896, + "learning_rate": 2.6660655566016545e-06, + "loss": 0.9164, + "step": 12611 + }, + { + "epoch": 0.8665658925381339, + "grad_norm": 1.0613558292388916, + "learning_rate": 2.663371770943246e-06, + "loss": 0.8908, + "step": 12612 + }, + { + "epoch": 0.8666346021712245, + "grad_norm": 1.0873061418533325, + "learning_rate": 2.6606792702974337e-06, + "loss": 0.9245, + "step": 12613 + }, + { + "epoch": 0.8667033118043149, + "grad_norm": 1.063893437385559, + "learning_rate": 2.6579880548191244e-06, + "loss": 0.9601, + "step": 12614 + }, + { + "epoch": 0.8667720214374055, + "grad_norm": 1.014217734336853, + "learning_rate": 2.655298124663136e-06, + "loss": 0.9123, + "step": 12615 + }, + { + "epoch": 0.8668407310704961, + "grad_norm": 1.0395067930221558, + "learning_rate": 2.6526094799842248e-06, + "loss": 0.8925, + "step": 12616 + }, + { + "epoch": 0.8669094407035867, + "grad_norm": 1.0393218994140625, + "learning_rate": 2.6499221209370704e-06, + "loss": 0.9576, + "step": 12617 + }, + { + "epoch": 0.8669781503366772, + "grad_norm": 1.1044334173202515, + "learning_rate": 2.6472360476762603e-06, + "loss": 0.947, + "step": 12618 + }, + { + "epoch": 0.8670468599697677, + "grad_norm": 1.1028624773025513, + "learning_rate": 2.6445512603563343e-06, + "loss": 0.9665, + "step": 12619 + }, + { + "epoch": 0.8671155696028583, + "grad_norm": 1.060243844985962, + "learning_rate": 2.641867759131747e-06, + "loss": 0.9087, + "step": 12620 + }, + { + "epoch": 0.8671842792359489, + "grad_norm": 1.0309101343154907, + "learning_rate": 2.639185544156869e-06, + "loss": 0.9207, + "step": 12621 + }, + { + "epoch": 0.8672529888690395, + "grad_norm": 1.0253651142120361, + "learning_rate": 2.6365046155860134e-06, + "loss": 0.9129, + "step": 12622 + }, + { + "epoch": 0.86732169850213, + "grad_norm": 1.031005620956421, + "learning_rate": 2.6338249735734175e-06, + "loss": 0.9373, + "step": 12623 + }, + { + "epoch": 0.8673904081352205, + "grad_norm": 1.024843692779541, + "learning_rate": 2.631146618273228e-06, + "loss": 0.9367, + "step": 12624 + }, + { + "epoch": 0.8674591177683111, + "grad_norm": 1.0720263719558716, + "learning_rate": 2.628469549839535e-06, + "loss": 0.9226, + "step": 12625 + }, + { + "epoch": 0.8675278274014017, + "grad_norm": 1.07747220993042, + "learning_rate": 2.625793768426346e-06, + "loss": 0.9313, + "step": 12626 + }, + { + "epoch": 0.8675965370344922, + "grad_norm": 1.1411433219909668, + "learning_rate": 2.623119274187608e-06, + "loss": 0.9238, + "step": 12627 + }, + { + "epoch": 0.8676652466675828, + "grad_norm": 1.0502551794052124, + "learning_rate": 2.620446067277166e-06, + "loss": 0.9179, + "step": 12628 + }, + { + "epoch": 0.8677339563006734, + "grad_norm": 1.1042635440826416, + "learning_rate": 2.6177741478488173e-06, + "loss": 0.9173, + "step": 12629 + }, + { + "epoch": 0.8678026659337639, + "grad_norm": 1.05532705783844, + "learning_rate": 2.615103516056275e-06, + "loss": 0.9512, + "step": 12630 + }, + { + "epoch": 0.8678713755668545, + "grad_norm": 1.0455158948898315, + "learning_rate": 2.6124341720531768e-06, + "loss": 0.9512, + "step": 12631 + }, + { + "epoch": 0.867940085199945, + "grad_norm": 1.0563639402389526, + "learning_rate": 2.6097661159930983e-06, + "loss": 0.9465, + "step": 12632 + }, + { + "epoch": 0.8680087948330356, + "grad_norm": 1.0449485778808594, + "learning_rate": 2.6070993480295155e-06, + "loss": 0.9186, + "step": 12633 + }, + { + "epoch": 0.8680775044661262, + "grad_norm": 1.09683096408844, + "learning_rate": 2.604433868315853e-06, + "loss": 0.9346, + "step": 12634 + }, + { + "epoch": 0.8681462140992167, + "grad_norm": 0.9762722849845886, + "learning_rate": 2.6017696770054556e-06, + "loss": 0.8808, + "step": 12635 + }, + { + "epoch": 0.8682149237323072, + "grad_norm": 1.1477023363113403, + "learning_rate": 2.5991067742515947e-06, + "loss": 0.953, + "step": 12636 + }, + { + "epoch": 0.8682836333653978, + "grad_norm": 1.1019645929336548, + "learning_rate": 2.596445160207464e-06, + "loss": 0.9401, + "step": 12637 + }, + { + "epoch": 0.8683523429984884, + "grad_norm": 1.055335283279419, + "learning_rate": 2.5937848350261778e-06, + "loss": 0.9416, + "step": 12638 + }, + { + "epoch": 0.868421052631579, + "grad_norm": 1.0526190996170044, + "learning_rate": 2.591125798860791e-06, + "loss": 0.9417, + "step": 12639 + }, + { + "epoch": 0.8684897622646695, + "grad_norm": 1.0229415893554688, + "learning_rate": 2.5884680518642706e-06, + "loss": 0.8967, + "step": 12640 + }, + { + "epoch": 0.86855847189776, + "grad_norm": 1.0933613777160645, + "learning_rate": 2.5858115941895217e-06, + "loss": 0.9381, + "step": 12641 + }, + { + "epoch": 0.8686271815308506, + "grad_norm": 1.03510582447052, + "learning_rate": 2.5831564259893666e-06, + "loss": 0.9378, + "step": 12642 + }, + { + "epoch": 0.8686958911639412, + "grad_norm": 1.0346990823745728, + "learning_rate": 2.5805025474165523e-06, + "loss": 0.9461, + "step": 12643 + }, + { + "epoch": 0.8687646007970318, + "grad_norm": 1.057050108909607, + "learning_rate": 2.577849958623754e-06, + "loss": 0.9625, + "step": 12644 + }, + { + "epoch": 0.8688333104301224, + "grad_norm": 1.0937824249267578, + "learning_rate": 2.575198659763581e-06, + "loss": 0.9292, + "step": 12645 + }, + { + "epoch": 0.8689020200632128, + "grad_norm": 1.0162678956985474, + "learning_rate": 2.5725486509885514e-06, + "loss": 0.9169, + "step": 12646 + }, + { + "epoch": 0.8689707296963034, + "grad_norm": 1.0521987676620483, + "learning_rate": 2.569899932451128e-06, + "loss": 0.8839, + "step": 12647 + }, + { + "epoch": 0.869039439329394, + "grad_norm": 1.1450284719467163, + "learning_rate": 2.56725250430368e-06, + "loss": 0.9363, + "step": 12648 + }, + { + "epoch": 0.8691081489624846, + "grad_norm": 1.0703169107437134, + "learning_rate": 2.564606366698519e-06, + "loss": 0.9019, + "step": 12649 + }, + { + "epoch": 0.8691768585955751, + "grad_norm": 1.0377124547958374, + "learning_rate": 2.561961519787873e-06, + "loss": 0.9349, + "step": 12650 + }, + { + "epoch": 0.8692455682286656, + "grad_norm": 1.1422597169876099, + "learning_rate": 2.5593179637239004e-06, + "loss": 0.9279, + "step": 12651 + }, + { + "epoch": 0.8693142778617562, + "grad_norm": 1.0865590572357178, + "learning_rate": 2.5566756986586864e-06, + "loss": 0.9251, + "step": 12652 + }, + { + "epoch": 0.8693829874948468, + "grad_norm": 1.029325246810913, + "learning_rate": 2.554034724744231e-06, + "loss": 0.9265, + "step": 12653 + }, + { + "epoch": 0.8694516971279374, + "grad_norm": 1.059040904045105, + "learning_rate": 2.5513950421324707e-06, + "loss": 0.887, + "step": 12654 + }, + { + "epoch": 0.8695204067610279, + "grad_norm": 1.0424997806549072, + "learning_rate": 2.5487566509752675e-06, + "loss": 0.8984, + "step": 12655 + }, + { + "epoch": 0.8695891163941185, + "grad_norm": 1.0453461408615112, + "learning_rate": 2.546119551424403e-06, + "loss": 0.9139, + "step": 12656 + }, + { + "epoch": 0.869657826027209, + "grad_norm": 1.0649652481079102, + "learning_rate": 2.5434837436315917e-06, + "loss": 0.92, + "step": 12657 + }, + { + "epoch": 0.8697265356602996, + "grad_norm": 1.062168002128601, + "learning_rate": 2.5408492277484708e-06, + "loss": 0.9367, + "step": 12658 + }, + { + "epoch": 0.8697952452933901, + "grad_norm": 1.0684524774551392, + "learning_rate": 2.5382160039265967e-06, + "loss": 0.9273, + "step": 12659 + }, + { + "epoch": 0.8698639549264807, + "grad_norm": 1.0045872926712036, + "learning_rate": 2.5355840723174596e-06, + "loss": 0.9032, + "step": 12660 + }, + { + "epoch": 0.8699326645595713, + "grad_norm": 1.0542972087860107, + "learning_rate": 2.532953433072474e-06, + "loss": 0.888, + "step": 12661 + }, + { + "epoch": 0.8700013741926618, + "grad_norm": 1.0356546640396118, + "learning_rate": 2.5303240863429777e-06, + "loss": 0.9285, + "step": 12662 + }, + { + "epoch": 0.8700700838257523, + "grad_norm": 1.0876593589782715, + "learning_rate": 2.5276960322802412e-06, + "loss": 0.9421, + "step": 12663 + }, + { + "epoch": 0.8701387934588429, + "grad_norm": 1.0239068269729614, + "learning_rate": 2.5250692710354464e-06, + "loss": 0.9268, + "step": 12664 + }, + { + "epoch": 0.8702075030919335, + "grad_norm": 1.0539628267288208, + "learning_rate": 2.5224438027597107e-06, + "loss": 0.9587, + "step": 12665 + }, + { + "epoch": 0.8702762127250241, + "grad_norm": 1.103110432624817, + "learning_rate": 2.5198196276040782e-06, + "loss": 0.9061, + "step": 12666 + }, + { + "epoch": 0.8703449223581146, + "grad_norm": 1.0837026834487915, + "learning_rate": 2.5171967457195216e-06, + "loss": 0.9698, + "step": 12667 + }, + { + "epoch": 0.8704136319912051, + "grad_norm": 1.036129355430603, + "learning_rate": 2.5145751572569208e-06, + "loss": 0.9208, + "step": 12668 + }, + { + "epoch": 0.8704823416242957, + "grad_norm": 0.9962072372436523, + "learning_rate": 2.511954862367108e-06, + "loss": 0.9485, + "step": 12669 + }, + { + "epoch": 0.8705510512573863, + "grad_norm": 1.0518182516098022, + "learning_rate": 2.509335861200815e-06, + "loss": 0.9193, + "step": 12670 + }, + { + "epoch": 0.8706197608904769, + "grad_norm": 1.0732614994049072, + "learning_rate": 2.5067181539087158e-06, + "loss": 0.9345, + "step": 12671 + }, + { + "epoch": 0.8706884705235675, + "grad_norm": 1.087629795074463, + "learning_rate": 2.504101740641407e-06, + "loss": 0.9528, + "step": 12672 + }, + { + "epoch": 0.8707571801566579, + "grad_norm": 1.0478935241699219, + "learning_rate": 2.501486621549412e-06, + "loss": 0.9082, + "step": 12673 + }, + { + "epoch": 0.8708258897897485, + "grad_norm": 1.0237692594528198, + "learning_rate": 2.498872796783172e-06, + "loss": 0.9008, + "step": 12674 + }, + { + "epoch": 0.8708945994228391, + "grad_norm": 1.0841684341430664, + "learning_rate": 2.4962602664930585e-06, + "loss": 0.9499, + "step": 12675 + }, + { + "epoch": 0.8709633090559297, + "grad_norm": 1.0125998258590698, + "learning_rate": 2.49364903082937e-06, + "loss": 0.8986, + "step": 12676 + }, + { + "epoch": 0.8710320186890202, + "grad_norm": 0.9937782883644104, + "learning_rate": 2.491039089942329e-06, + "loss": 0.9114, + "step": 12677 + }, + { + "epoch": 0.8711007283221107, + "grad_norm": 1.021815538406372, + "learning_rate": 2.48843044398209e-06, + "loss": 0.8969, + "step": 12678 + }, + { + "epoch": 0.8711694379552013, + "grad_norm": 1.0504026412963867, + "learning_rate": 2.4858230930987185e-06, + "loss": 0.9131, + "step": 12679 + }, + { + "epoch": 0.8712381475882919, + "grad_norm": 1.0610685348510742, + "learning_rate": 2.483217037442215e-06, + "loss": 0.9522, + "step": 12680 + }, + { + "epoch": 0.8713068572213825, + "grad_norm": 1.1100711822509766, + "learning_rate": 2.4806122771625083e-06, + "loss": 0.9151, + "step": 12681 + }, + { + "epoch": 0.871375566854473, + "grad_norm": 1.0942654609680176, + "learning_rate": 2.4780088124094435e-06, + "loss": 0.9723, + "step": 12682 + }, + { + "epoch": 0.8714442764875635, + "grad_norm": 1.0822467803955078, + "learning_rate": 2.475406643332806e-06, + "loss": 0.9429, + "step": 12683 + }, + { + "epoch": 0.8715129861206541, + "grad_norm": 1.0338889360427856, + "learning_rate": 2.4728057700822844e-06, + "loss": 0.9275, + "step": 12684 + }, + { + "epoch": 0.8715816957537447, + "grad_norm": 1.0936126708984375, + "learning_rate": 2.4702061928075143e-06, + "loss": 0.9186, + "step": 12685 + }, + { + "epoch": 0.8716504053868352, + "grad_norm": 1.1021146774291992, + "learning_rate": 2.4676079116580403e-06, + "loss": 0.8849, + "step": 12686 + }, + { + "epoch": 0.8717191150199258, + "grad_norm": 1.0713605880737305, + "learning_rate": 2.465010926783348e-06, + "loss": 0.9, + "step": 12687 + }, + { + "epoch": 0.8717878246530164, + "grad_norm": 1.0095144510269165, + "learning_rate": 2.4624152383328407e-06, + "loss": 0.8974, + "step": 12688 + }, + { + "epoch": 0.8718565342861069, + "grad_norm": 1.042314887046814, + "learning_rate": 2.4598208464558377e-06, + "loss": 0.927, + "step": 12689 + }, + { + "epoch": 0.8719252439191975, + "grad_norm": 1.0769085884094238, + "learning_rate": 2.4572277513015966e-06, + "loss": 0.9032, + "step": 12690 + }, + { + "epoch": 0.871993953552288, + "grad_norm": 1.0254181623458862, + "learning_rate": 2.4546359530193048e-06, + "loss": 0.9022, + "step": 12691 + }, + { + "epoch": 0.8720626631853786, + "grad_norm": 1.0248929262161255, + "learning_rate": 2.452045451758053e-06, + "loss": 0.9172, + "step": 12692 + }, + { + "epoch": 0.8721313728184692, + "grad_norm": 1.096656084060669, + "learning_rate": 2.4494562476668786e-06, + "loss": 0.9089, + "step": 12693 + }, + { + "epoch": 0.8722000824515597, + "grad_norm": 1.0435140132904053, + "learning_rate": 2.4468683408947422e-06, + "loss": 0.9165, + "step": 12694 + }, + { + "epoch": 0.8722687920846502, + "grad_norm": 1.1709959506988525, + "learning_rate": 2.444281731590514e-06, + "loss": 0.9639, + "step": 12695 + }, + { + "epoch": 0.8723375017177408, + "grad_norm": 1.0675201416015625, + "learning_rate": 2.441696419903006e-06, + "loss": 0.962, + "step": 12696 + }, + { + "epoch": 0.8724062113508314, + "grad_norm": 1.0952551364898682, + "learning_rate": 2.439112405980945e-06, + "loss": 0.9057, + "step": 12697 + }, + { + "epoch": 0.872474920983922, + "grad_norm": 1.0381991863250732, + "learning_rate": 2.4365296899729994e-06, + "loss": 0.8787, + "step": 12698 + }, + { + "epoch": 0.8725436306170125, + "grad_norm": 1.0566483736038208, + "learning_rate": 2.4339482720277364e-06, + "loss": 0.9206, + "step": 12699 + }, + { + "epoch": 0.872612340250103, + "grad_norm": 1.000535011291504, + "learning_rate": 2.43136815229367e-06, + "loss": 0.9229, + "step": 12700 + }, + { + "epoch": 0.8726810498831936, + "grad_norm": 1.1141598224639893, + "learning_rate": 2.4287893309192344e-06, + "loss": 0.8663, + "step": 12701 + }, + { + "epoch": 0.8727497595162842, + "grad_norm": 1.0834587812423706, + "learning_rate": 2.4262118080527835e-06, + "loss": 0.9603, + "step": 12702 + }, + { + "epoch": 0.8728184691493748, + "grad_norm": 1.069095492362976, + "learning_rate": 2.4236355838426093e-06, + "loss": 0.9314, + "step": 12703 + }, + { + "epoch": 0.8728871787824654, + "grad_norm": 1.1027113199234009, + "learning_rate": 2.4210606584369104e-06, + "loss": 0.9344, + "step": 12704 + }, + { + "epoch": 0.8729558884155558, + "grad_norm": 1.08356773853302, + "learning_rate": 2.4184870319838265e-06, + "loss": 0.9212, + "step": 12705 + }, + { + "epoch": 0.8730245980486464, + "grad_norm": 1.0956780910491943, + "learning_rate": 2.4159147046314114e-06, + "loss": 0.9702, + "step": 12706 + }, + { + "epoch": 0.873093307681737, + "grad_norm": 1.031552791595459, + "learning_rate": 2.4133436765276544e-06, + "loss": 0.9388, + "step": 12707 + }, + { + "epoch": 0.8731620173148276, + "grad_norm": 1.0441479682922363, + "learning_rate": 2.410773947820469e-06, + "loss": 0.895, + "step": 12708 + }, + { + "epoch": 0.8732307269479181, + "grad_norm": 1.0709019899368286, + "learning_rate": 2.40820551865768e-06, + "loss": 0.9148, + "step": 12709 + }, + { + "epoch": 0.8732994365810086, + "grad_norm": 1.0234451293945312, + "learning_rate": 2.405638389187054e-06, + "loss": 0.9246, + "step": 12710 + }, + { + "epoch": 0.8733681462140992, + "grad_norm": 1.0668247938156128, + "learning_rate": 2.4030725595562715e-06, + "loss": 0.923, + "step": 12711 + }, + { + "epoch": 0.8734368558471898, + "grad_norm": 1.0797364711761475, + "learning_rate": 2.4005080299129486e-06, + "loss": 0.8984, + "step": 12712 + }, + { + "epoch": 0.8735055654802804, + "grad_norm": 1.0280781984329224, + "learning_rate": 2.397944800404625e-06, + "loss": 0.9237, + "step": 12713 + }, + { + "epoch": 0.8735742751133709, + "grad_norm": 1.0791922807693481, + "learning_rate": 2.39538287117875e-06, + "loss": 0.9087, + "step": 12714 + }, + { + "epoch": 0.8736429847464614, + "grad_norm": 1.1044278144836426, + "learning_rate": 2.392822242382714e-06, + "loss": 0.9579, + "step": 12715 + }, + { + "epoch": 0.873711694379552, + "grad_norm": 1.1120644807815552, + "learning_rate": 2.390262914163838e-06, + "loss": 0.8934, + "step": 12716 + }, + { + "epoch": 0.8737804040126426, + "grad_norm": 1.0763956308364868, + "learning_rate": 2.3877048866693425e-06, + "loss": 0.9303, + "step": 12717 + }, + { + "epoch": 0.8738491136457331, + "grad_norm": 1.0068498849868774, + "learning_rate": 2.3851481600463992e-06, + "loss": 0.9088, + "step": 12718 + }, + { + "epoch": 0.8739178232788237, + "grad_norm": 1.0246543884277344, + "learning_rate": 2.382592734442096e-06, + "loss": 0.9051, + "step": 12719 + }, + { + "epoch": 0.8739865329119143, + "grad_norm": 1.1381518840789795, + "learning_rate": 2.3800386100034404e-06, + "loss": 0.9347, + "step": 12720 + }, + { + "epoch": 0.8740552425450048, + "grad_norm": 1.0137302875518799, + "learning_rate": 2.3774857868773703e-06, + "loss": 0.9127, + "step": 12721 + }, + { + "epoch": 0.8741239521780954, + "grad_norm": 1.0692247152328491, + "learning_rate": 2.3749342652107486e-06, + "loss": 0.9077, + "step": 12722 + }, + { + "epoch": 0.8741926618111859, + "grad_norm": 1.0483098030090332, + "learning_rate": 2.3723840451503636e-06, + "loss": 0.9276, + "step": 12723 + }, + { + "epoch": 0.8742613714442765, + "grad_norm": 0.9664784073829651, + "learning_rate": 2.369835126842934e-06, + "loss": 0.8816, + "step": 12724 + }, + { + "epoch": 0.8743300810773671, + "grad_norm": 1.072679042816162, + "learning_rate": 2.3672875104350844e-06, + "loss": 0.926, + "step": 12725 + }, + { + "epoch": 0.8743987907104576, + "grad_norm": 1.1191953420639038, + "learning_rate": 2.364741196073386e-06, + "loss": 0.9293, + "step": 12726 + }, + { + "epoch": 0.8744675003435481, + "grad_norm": 1.1042975187301636, + "learning_rate": 2.362196183904325e-06, + "loss": 0.9518, + "step": 12727 + }, + { + "epoch": 0.8745362099766387, + "grad_norm": 1.0849822759628296, + "learning_rate": 2.3596524740743142e-06, + "loss": 0.9206, + "step": 12728 + }, + { + "epoch": 0.8746049196097293, + "grad_norm": 1.079372525215149, + "learning_rate": 2.3571100667296953e-06, + "loss": 0.8993, + "step": 12729 + }, + { + "epoch": 0.8746736292428199, + "grad_norm": 1.0345954895019531, + "learning_rate": 2.354568962016723e-06, + "loss": 0.9328, + "step": 12730 + }, + { + "epoch": 0.8747423388759104, + "grad_norm": 1.0740821361541748, + "learning_rate": 2.3520291600815926e-06, + "loss": 0.9466, + "step": 12731 + }, + { + "epoch": 0.8748110485090009, + "grad_norm": 1.023367166519165, + "learning_rate": 2.3494906610704164e-06, + "loss": 0.8947, + "step": 12732 + }, + { + "epoch": 0.8748797581420915, + "grad_norm": 1.0825622081756592, + "learning_rate": 2.3469534651292286e-06, + "loss": 0.9576, + "step": 12733 + }, + { + "epoch": 0.8749484677751821, + "grad_norm": 1.1232730150222778, + "learning_rate": 2.3444175724040035e-06, + "loss": 0.9288, + "step": 12734 + }, + { + "epoch": 0.8750171774082727, + "grad_norm": 1.050775170326233, + "learning_rate": 2.3418829830406163e-06, + "loss": 0.9361, + "step": 12735 + }, + { + "epoch": 0.8750858870413633, + "grad_norm": 1.0687154531478882, + "learning_rate": 2.339349697184884e-06, + "loss": 0.9276, + "step": 12736 + }, + { + "epoch": 0.8751545966744537, + "grad_norm": 1.060617446899414, + "learning_rate": 2.3368177149825475e-06, + "loss": 0.9418, + "step": 12737 + }, + { + "epoch": 0.8752233063075443, + "grad_norm": 1.0397801399230957, + "learning_rate": 2.3342870365792746e-06, + "loss": 0.9324, + "step": 12738 + }, + { + "epoch": 0.8752920159406349, + "grad_norm": 0.9978806376457214, + "learning_rate": 2.3317576621206427e-06, + "loss": 0.9232, + "step": 12739 + }, + { + "epoch": 0.8753607255737255, + "grad_norm": 1.0493706464767456, + "learning_rate": 2.3292295917521745e-06, + "loss": 0.9069, + "step": 12740 + }, + { + "epoch": 0.875429435206816, + "grad_norm": 1.1487212181091309, + "learning_rate": 2.3267028256193036e-06, + "loss": 0.9644, + "step": 12741 + }, + { + "epoch": 0.8754981448399065, + "grad_norm": 1.1012303829193115, + "learning_rate": 2.3241773638673912e-06, + "loss": 0.9446, + "step": 12742 + }, + { + "epoch": 0.8755668544729971, + "grad_norm": 1.107668161392212, + "learning_rate": 2.32165320664173e-06, + "loss": 0.9517, + "step": 12743 + }, + { + "epoch": 0.8756355641060877, + "grad_norm": 1.0407634973526, + "learning_rate": 2.3191303540875368e-06, + "loss": 0.9412, + "step": 12744 + }, + { + "epoch": 0.8757042737391783, + "grad_norm": 1.0326801538467407, + "learning_rate": 2.31660880634994e-06, + "loss": 0.9004, + "step": 12745 + }, + { + "epoch": 0.8757729833722688, + "grad_norm": 1.083022952079773, + "learning_rate": 2.3140885635740102e-06, + "loss": 0.9237, + "step": 12746 + }, + { + "epoch": 0.8758416930053593, + "grad_norm": 1.0830974578857422, + "learning_rate": 2.3115696259047306e-06, + "loss": 0.9252, + "step": 12747 + }, + { + "epoch": 0.8759104026384499, + "grad_norm": 1.0637024641036987, + "learning_rate": 2.309051993487016e-06, + "loss": 0.9531, + "step": 12748 + }, + { + "epoch": 0.8759791122715405, + "grad_norm": 1.1098878383636475, + "learning_rate": 2.3065356664657117e-06, + "loss": 0.9811, + "step": 12749 + }, + { + "epoch": 0.876047821904631, + "grad_norm": 1.0851894617080688, + "learning_rate": 2.3040206449855663e-06, + "loss": 0.951, + "step": 12750 + }, + { + "epoch": 0.8761165315377216, + "grad_norm": 1.054858684539795, + "learning_rate": 2.3015069291912773e-06, + "loss": 0.9361, + "step": 12751 + }, + { + "epoch": 0.8761852411708122, + "grad_norm": 1.0095535516738892, + "learning_rate": 2.2989945192274542e-06, + "loss": 0.913, + "step": 12752 + }, + { + "epoch": 0.8762539508039027, + "grad_norm": 0.9959141612052917, + "learning_rate": 2.2964834152386317e-06, + "loss": 0.9108, + "step": 12753 + }, + { + "epoch": 0.8763226604369933, + "grad_norm": 1.0187008380889893, + "learning_rate": 2.2939736173692825e-06, + "loss": 0.8999, + "step": 12754 + }, + { + "epoch": 0.8763913700700838, + "grad_norm": 1.0615417957305908, + "learning_rate": 2.2914651257637836e-06, + "loss": 0.9182, + "step": 12755 + }, + { + "epoch": 0.8764600797031744, + "grad_norm": 1.0251717567443848, + "learning_rate": 2.288957940566447e-06, + "loss": 0.9232, + "step": 12756 + }, + { + "epoch": 0.876528789336265, + "grad_norm": 1.1313318014144897, + "learning_rate": 2.2864520619215122e-06, + "loss": 0.9724, + "step": 12757 + }, + { + "epoch": 0.8765974989693555, + "grad_norm": 1.052783727645874, + "learning_rate": 2.2839474899731427e-06, + "loss": 0.9308, + "step": 12758 + }, + { + "epoch": 0.876666208602446, + "grad_norm": 1.1280330419540405, + "learning_rate": 2.281444224865428e-06, + "loss": 0.9074, + "step": 12759 + }, + { + "epoch": 0.8767349182355366, + "grad_norm": 1.0214180946350098, + "learning_rate": 2.2789422667423703e-06, + "loss": 0.9028, + "step": 12760 + }, + { + "epoch": 0.8768036278686272, + "grad_norm": 1.0110541582107544, + "learning_rate": 2.2764416157479096e-06, + "loss": 0.8961, + "step": 12761 + }, + { + "epoch": 0.8768723375017178, + "grad_norm": 1.043859839439392, + "learning_rate": 2.273942272025911e-06, + "loss": 0.9092, + "step": 12762 + }, + { + "epoch": 0.8769410471348082, + "grad_norm": 1.0536383390426636, + "learning_rate": 2.271444235720152e-06, + "loss": 0.9018, + "step": 12763 + }, + { + "epoch": 0.8770097567678988, + "grad_norm": 1.0406543016433716, + "learning_rate": 2.2689475069743506e-06, + "loss": 0.9322, + "step": 12764 + }, + { + "epoch": 0.8770784664009894, + "grad_norm": 1.1323765516281128, + "learning_rate": 2.2664520859321415e-06, + "loss": 0.9682, + "step": 12765 + }, + { + "epoch": 0.87714717603408, + "grad_norm": 1.0040720701217651, + "learning_rate": 2.263957972737077e-06, + "loss": 0.9294, + "step": 12766 + }, + { + "epoch": 0.8772158856671706, + "grad_norm": 1.0749176740646362, + "learning_rate": 2.2614651675326506e-06, + "loss": 0.9749, + "step": 12767 + }, + { + "epoch": 0.8772845953002611, + "grad_norm": 1.062064290046692, + "learning_rate": 2.2589736704622664e-06, + "loss": 0.9302, + "step": 12768 + }, + { + "epoch": 0.8773533049333516, + "grad_norm": 1.1221493482589722, + "learning_rate": 2.256483481669261e-06, + "loss": 0.9401, + "step": 12769 + }, + { + "epoch": 0.8774220145664422, + "grad_norm": 1.1109428405761719, + "learning_rate": 2.253994601296899e-06, + "loss": 0.9619, + "step": 12770 + }, + { + "epoch": 0.8774907241995328, + "grad_norm": 1.041804552078247, + "learning_rate": 2.2515070294883534e-06, + "loss": 0.9319, + "step": 12771 + }, + { + "epoch": 0.8775594338326234, + "grad_norm": 1.027209758758545, + "learning_rate": 2.2490207663867393e-06, + "loss": 0.9458, + "step": 12772 + }, + { + "epoch": 0.8776281434657139, + "grad_norm": 1.0646858215332031, + "learning_rate": 2.2465358121350864e-06, + "loss": 0.946, + "step": 12773 + }, + { + "epoch": 0.8776968530988044, + "grad_norm": 1.0480201244354248, + "learning_rate": 2.2440521668763536e-06, + "loss": 0.9091, + "step": 12774 + }, + { + "epoch": 0.877765562731895, + "grad_norm": 1.0583372116088867, + "learning_rate": 2.2415698307534315e-06, + "loss": 0.9191, + "step": 12775 + }, + { + "epoch": 0.8778342723649856, + "grad_norm": 1.1192171573638916, + "learning_rate": 2.2390888039091175e-06, + "loss": 0.9605, + "step": 12776 + }, + { + "epoch": 0.8779029819980761, + "grad_norm": 1.118536114692688, + "learning_rate": 2.236609086486144e-06, + "loss": 0.9369, + "step": 12777 + }, + { + "epoch": 0.8779716916311667, + "grad_norm": 1.0936287641525269, + "learning_rate": 2.2341306786271695e-06, + "loss": 0.9228, + "step": 12778 + }, + { + "epoch": 0.8780404012642572, + "grad_norm": 1.045833945274353, + "learning_rate": 2.2316535804747767e-06, + "loss": 0.9501, + "step": 12779 + }, + { + "epoch": 0.8781091108973478, + "grad_norm": 1.1372578144073486, + "learning_rate": 2.2291777921714773e-06, + "loss": 0.934, + "step": 12780 + }, + { + "epoch": 0.8781778205304384, + "grad_norm": 1.09443199634552, + "learning_rate": 2.2267033138596875e-06, + "loss": 0.9266, + "step": 12781 + }, + { + "epoch": 0.8782465301635289, + "grad_norm": 0.997621476650238, + "learning_rate": 2.2242301456817745e-06, + "loss": 0.9083, + "step": 12782 + }, + { + "epoch": 0.8783152397966195, + "grad_norm": 1.089827299118042, + "learning_rate": 2.22175828778001e-06, + "loss": 0.9475, + "step": 12783 + }, + { + "epoch": 0.8783839494297101, + "grad_norm": 1.0838377475738525, + "learning_rate": 2.219287740296605e-06, + "loss": 0.923, + "step": 12784 + }, + { + "epoch": 0.8784526590628006, + "grad_norm": 1.0900520086288452, + "learning_rate": 2.216818503373691e-06, + "loss": 0.91, + "step": 12785 + }, + { + "epoch": 0.8785213686958911, + "grad_norm": 1.1479871273040771, + "learning_rate": 2.2143505771533097e-06, + "loss": 0.9764, + "step": 12786 + }, + { + "epoch": 0.8785900783289817, + "grad_norm": 1.0553878545761108, + "learning_rate": 2.2118839617774527e-06, + "loss": 0.943, + "step": 12787 + }, + { + "epoch": 0.8786587879620723, + "grad_norm": 1.0499787330627441, + "learning_rate": 2.209418657388013e-06, + "loss": 0.9271, + "step": 12788 + }, + { + "epoch": 0.8787274975951629, + "grad_norm": 1.0286264419555664, + "learning_rate": 2.20695466412682e-06, + "loss": 0.9037, + "step": 12789 + }, + { + "epoch": 0.8787962072282534, + "grad_norm": 1.1262741088867188, + "learning_rate": 2.2044919821356343e-06, + "loss": 0.93, + "step": 12790 + }, + { + "epoch": 0.8788649168613439, + "grad_norm": 1.0516570806503296, + "learning_rate": 2.202030611556119e-06, + "loss": 0.9462, + "step": 12791 + }, + { + "epoch": 0.8789336264944345, + "grad_norm": 1.0343338251113892, + "learning_rate": 2.199570552529884e-06, + "loss": 0.9055, + "step": 12792 + }, + { + "epoch": 0.8790023361275251, + "grad_norm": 1.0640631914138794, + "learning_rate": 2.1971118051984514e-06, + "loss": 0.9368, + "step": 12793 + }, + { + "epoch": 0.8790710457606157, + "grad_norm": 1.0425117015838623, + "learning_rate": 2.1946543697032723e-06, + "loss": 0.9023, + "step": 12794 + }, + { + "epoch": 0.8791397553937061, + "grad_norm": 1.0727957487106323, + "learning_rate": 2.1921982461857282e-06, + "loss": 0.9159, + "step": 12795 + }, + { + "epoch": 0.8792084650267967, + "grad_norm": 1.0420951843261719, + "learning_rate": 2.189743434787106e-06, + "loss": 0.943, + "step": 12796 + }, + { + "epoch": 0.8792771746598873, + "grad_norm": 1.0387425422668457, + "learning_rate": 2.187289935648637e-06, + "loss": 0.9204, + "step": 12797 + }, + { + "epoch": 0.8793458842929779, + "grad_norm": 1.1205580234527588, + "learning_rate": 2.1848377489114697e-06, + "loss": 0.9381, + "step": 12798 + }, + { + "epoch": 0.8794145939260685, + "grad_norm": 1.0701578855514526, + "learning_rate": 2.1823868747166742e-06, + "loss": 0.9049, + "step": 12799 + }, + { + "epoch": 0.879483303559159, + "grad_norm": 1.0177853107452393, + "learning_rate": 2.179937313205255e-06, + "loss": 0.9176, + "step": 12800 + }, + { + "epoch": 0.8795520131922495, + "grad_norm": 1.0620558261871338, + "learning_rate": 2.1774890645181208e-06, + "loss": 0.9211, + "step": 12801 + }, + { + "epoch": 0.8796207228253401, + "grad_norm": 1.0973299741744995, + "learning_rate": 2.1750421287961257e-06, + "loss": 0.9371, + "step": 12802 + }, + { + "epoch": 0.8796894324584307, + "grad_norm": 1.121442198753357, + "learning_rate": 2.1725965061800415e-06, + "loss": 0.9421, + "step": 12803 + }, + { + "epoch": 0.8797581420915213, + "grad_norm": 1.0496548414230347, + "learning_rate": 2.17015219681056e-06, + "loss": 0.9172, + "step": 12804 + }, + { + "epoch": 0.8798268517246118, + "grad_norm": 1.105245590209961, + "learning_rate": 2.167709200828308e-06, + "loss": 0.9518, + "step": 12805 + }, + { + "epoch": 0.8798955613577023, + "grad_norm": 1.0841984748840332, + "learning_rate": 2.16526751837382e-06, + "loss": 0.9552, + "step": 12806 + }, + { + "epoch": 0.8799642709907929, + "grad_norm": 1.0469673871994019, + "learning_rate": 2.162827149587568e-06, + "loss": 0.9184, + "step": 12807 + }, + { + "epoch": 0.8800329806238835, + "grad_norm": 1.027161955833435, + "learning_rate": 2.160388094609944e-06, + "loss": 0.8914, + "step": 12808 + }, + { + "epoch": 0.880101690256974, + "grad_norm": 1.070107340812683, + "learning_rate": 2.1579503535812727e-06, + "loss": 0.9423, + "step": 12809 + }, + { + "epoch": 0.8801703998900646, + "grad_norm": 1.0584977865219116, + "learning_rate": 2.155513926641786e-06, + "loss": 0.9056, + "step": 12810 + }, + { + "epoch": 0.8802391095231551, + "grad_norm": 1.0360705852508545, + "learning_rate": 2.1530788139316557e-06, + "loss": 0.9144, + "step": 12811 + }, + { + "epoch": 0.8803078191562457, + "grad_norm": 0.9355236291885376, + "learning_rate": 2.150645015590971e-06, + "loss": 0.8928, + "step": 12812 + }, + { + "epoch": 0.8803765287893363, + "grad_norm": 1.0733075141906738, + "learning_rate": 2.148212531759744e-06, + "loss": 0.9273, + "step": 12813 + }, + { + "epoch": 0.8804452384224268, + "grad_norm": 1.041011929512024, + "learning_rate": 2.1457813625779172e-06, + "loss": 0.9292, + "step": 12814 + }, + { + "epoch": 0.8805139480555174, + "grad_norm": 1.0186043977737427, + "learning_rate": 2.1433515081853596e-06, + "loss": 0.8989, + "step": 12815 + }, + { + "epoch": 0.880582657688608, + "grad_norm": 1.0424853563308716, + "learning_rate": 2.1409229687218453e-06, + "loss": 0.9142, + "step": 12816 + }, + { + "epoch": 0.8806513673216985, + "grad_norm": 1.0195789337158203, + "learning_rate": 2.138495744327096e-06, + "loss": 0.919, + "step": 12817 + }, + { + "epoch": 0.880720076954789, + "grad_norm": 1.0262521505355835, + "learning_rate": 2.136069835140747e-06, + "loss": 0.9229, + "step": 12818 + }, + { + "epoch": 0.8807887865878796, + "grad_norm": 1.1039438247680664, + "learning_rate": 2.1336452413023623e-06, + "loss": 0.9145, + "step": 12819 + }, + { + "epoch": 0.8808574962209702, + "grad_norm": 1.0413477420806885, + "learning_rate": 2.131221962951424e-06, + "loss": 0.9025, + "step": 12820 + }, + { + "epoch": 0.8809262058540608, + "grad_norm": 1.033376932144165, + "learning_rate": 2.1288000002273405e-06, + "loss": 0.8986, + "step": 12821 + }, + { + "epoch": 0.8809949154871513, + "grad_norm": 1.075157880783081, + "learning_rate": 2.1263793532694476e-06, + "loss": 0.907, + "step": 12822 + }, + { + "epoch": 0.8810636251202418, + "grad_norm": 1.124121069908142, + "learning_rate": 2.123960022217006e-06, + "loss": 0.9234, + "step": 12823 + }, + { + "epoch": 0.8811323347533324, + "grad_norm": 1.1020537614822388, + "learning_rate": 2.121542007209193e-06, + "loss": 0.9123, + "step": 12824 + }, + { + "epoch": 0.881201044386423, + "grad_norm": 1.0371689796447754, + "learning_rate": 2.1191253083851197e-06, + "loss": 0.9424, + "step": 12825 + }, + { + "epoch": 0.8812697540195136, + "grad_norm": 1.0182501077651978, + "learning_rate": 2.1167099258838195e-06, + "loss": 0.9199, + "step": 12826 + }, + { + "epoch": 0.881338463652604, + "grad_norm": 1.0562466382980347, + "learning_rate": 2.1142958598442396e-06, + "loss": 0.9225, + "step": 12827 + }, + { + "epoch": 0.8814071732856946, + "grad_norm": 1.0048576593399048, + "learning_rate": 2.1118831104052663e-06, + "loss": 0.9578, + "step": 12828 + }, + { + "epoch": 0.8814758829187852, + "grad_norm": 1.018126368522644, + "learning_rate": 2.1094716777057016e-06, + "loss": 0.9304, + "step": 12829 + }, + { + "epoch": 0.8815445925518758, + "grad_norm": 1.0205572843551636, + "learning_rate": 2.1070615618842744e-06, + "loss": 0.8974, + "step": 12830 + }, + { + "epoch": 0.8816133021849664, + "grad_norm": 1.0211154222488403, + "learning_rate": 2.1046527630796425e-06, + "loss": 0.9177, + "step": 12831 + }, + { + "epoch": 0.8816820118180569, + "grad_norm": 1.0401222705841064, + "learning_rate": 2.1022452814303705e-06, + "loss": 0.9435, + "step": 12832 + }, + { + "epoch": 0.8817507214511474, + "grad_norm": 1.102702021598816, + "learning_rate": 2.099839117074967e-06, + "loss": 0.9551, + "step": 12833 + }, + { + "epoch": 0.881819431084238, + "grad_norm": 1.0375770330429077, + "learning_rate": 2.09743427015186e-06, + "loss": 0.9263, + "step": 12834 + }, + { + "epoch": 0.8818881407173286, + "grad_norm": 1.01217520236969, + "learning_rate": 2.095030740799392e-06, + "loss": 0.9319, + "step": 12835 + }, + { + "epoch": 0.8819568503504192, + "grad_norm": 1.0052419900894165, + "learning_rate": 2.0926285291558436e-06, + "loss": 0.9026, + "step": 12836 + }, + { + "epoch": 0.8820255599835097, + "grad_norm": 1.0513297319412231, + "learning_rate": 2.0902276353594015e-06, + "loss": 0.9236, + "step": 12837 + }, + { + "epoch": 0.8820942696166002, + "grad_norm": 1.0818997621536255, + "learning_rate": 2.0878280595481967e-06, + "loss": 0.935, + "step": 12838 + }, + { + "epoch": 0.8821629792496908, + "grad_norm": 1.0860904455184937, + "learning_rate": 2.0854298018602725e-06, + "loss": 0.9725, + "step": 12839 + }, + { + "epoch": 0.8822316888827814, + "grad_norm": 1.08347749710083, + "learning_rate": 2.083032862433601e-06, + "loss": 0.8739, + "step": 12840 + }, + { + "epoch": 0.8823003985158719, + "grad_norm": 1.0572636127471924, + "learning_rate": 2.0806372414060802e-06, + "loss": 0.9227, + "step": 12841 + }, + { + "epoch": 0.8823691081489625, + "grad_norm": 1.1098676919937134, + "learning_rate": 2.07824293891552e-06, + "loss": 0.9293, + "step": 12842 + }, + { + "epoch": 0.882437817782053, + "grad_norm": 1.04008150100708, + "learning_rate": 2.075849955099668e-06, + "loss": 0.9505, + "step": 12843 + }, + { + "epoch": 0.8825065274151436, + "grad_norm": 1.0083013772964478, + "learning_rate": 2.073458290096189e-06, + "loss": 0.9335, + "step": 12844 + }, + { + "epoch": 0.8825752370482342, + "grad_norm": 1.077082872390747, + "learning_rate": 2.071067944042676e-06, + "loss": 0.9166, + "step": 12845 + }, + { + "epoch": 0.8826439466813247, + "grad_norm": 1.0628694295883179, + "learning_rate": 2.068678917076647e-06, + "loss": 0.9104, + "step": 12846 + }, + { + "epoch": 0.8827126563144153, + "grad_norm": 1.0238902568817139, + "learning_rate": 2.066291209335533e-06, + "loss": 0.9022, + "step": 12847 + }, + { + "epoch": 0.8827813659475059, + "grad_norm": 1.065369963645935, + "learning_rate": 2.063904820956705e-06, + "loss": 0.9083, + "step": 12848 + }, + { + "epoch": 0.8828500755805964, + "grad_norm": 1.0705130100250244, + "learning_rate": 2.0615197520774452e-06, + "loss": 0.9468, + "step": 12849 + }, + { + "epoch": 0.8829187852136869, + "grad_norm": 1.1241978406906128, + "learning_rate": 2.059136002834969e-06, + "loss": 0.9337, + "step": 12850 + }, + { + "epoch": 0.8829874948467775, + "grad_norm": 1.0798101425170898, + "learning_rate": 2.0567535733664154e-06, + "loss": 0.941, + "step": 12851 + }, + { + "epoch": 0.8830562044798681, + "grad_norm": 1.023074984550476, + "learning_rate": 2.0543724638088347e-06, + "loss": 0.9268, + "step": 12852 + }, + { + "epoch": 0.8831249141129587, + "grad_norm": 1.0612672567367554, + "learning_rate": 2.0519926742992164e-06, + "loss": 0.9725, + "step": 12853 + }, + { + "epoch": 0.8831936237460492, + "grad_norm": 1.0450263023376465, + "learning_rate": 2.0496142049744675e-06, + "loss": 0.9426, + "step": 12854 + }, + { + "epoch": 0.8832623333791397, + "grad_norm": 1.0874965190887451, + "learning_rate": 2.0472370559714184e-06, + "loss": 0.9343, + "step": 12855 + }, + { + "epoch": 0.8833310430122303, + "grad_norm": 1.040371298789978, + "learning_rate": 2.0448612274268307e-06, + "loss": 0.9353, + "step": 12856 + }, + { + "epoch": 0.8833997526453209, + "grad_norm": 1.0538419485092163, + "learning_rate": 2.042486719477377e-06, + "loss": 0.9388, + "step": 12857 + }, + { + "epoch": 0.8834684622784115, + "grad_norm": 1.0159032344818115, + "learning_rate": 2.040113532259666e-06, + "loss": 0.9229, + "step": 12858 + }, + { + "epoch": 0.8835371719115019, + "grad_norm": 1.033892273902893, + "learning_rate": 2.0377416659102227e-06, + "loss": 0.9063, + "step": 12859 + }, + { + "epoch": 0.8836058815445925, + "grad_norm": 1.0623453855514526, + "learning_rate": 2.035371120565499e-06, + "loss": 0.9533, + "step": 12860 + }, + { + "epoch": 0.8836745911776831, + "grad_norm": 1.0976921319961548, + "learning_rate": 2.0330018963618774e-06, + "loss": 0.9515, + "step": 12861 + }, + { + "epoch": 0.8837433008107737, + "grad_norm": 1.0644407272338867, + "learning_rate": 2.030633993435649e-06, + "loss": 0.9185, + "step": 12862 + }, + { + "epoch": 0.8838120104438643, + "grad_norm": 1.0615198612213135, + "learning_rate": 2.0282674119230404e-06, + "loss": 0.9419, + "step": 12863 + }, + { + "epoch": 0.8838807200769548, + "grad_norm": 1.0729039907455444, + "learning_rate": 2.025902151960202e-06, + "loss": 0.911, + "step": 12864 + }, + { + "epoch": 0.8839494297100453, + "grad_norm": 1.0091465711593628, + "learning_rate": 2.0235382136832043e-06, + "loss": 0.9119, + "step": 12865 + }, + { + "epoch": 0.8840181393431359, + "grad_norm": 1.0794117450714111, + "learning_rate": 2.0211755972280476e-06, + "loss": 0.9323, + "step": 12866 + }, + { + "epoch": 0.8840868489762265, + "grad_norm": 1.0466866493225098, + "learning_rate": 2.0188143027306456e-06, + "loss": 0.9374, + "step": 12867 + }, + { + "epoch": 0.884155558609317, + "grad_norm": 1.0860905647277832, + "learning_rate": 2.016454330326842e-06, + "loss": 0.9163, + "step": 12868 + }, + { + "epoch": 0.8842242682424076, + "grad_norm": 1.094903588294983, + "learning_rate": 2.0140956801524087e-06, + "loss": 0.9223, + "step": 12869 + }, + { + "epoch": 0.8842929778754981, + "grad_norm": 1.0679643154144287, + "learning_rate": 2.0117383523430343e-06, + "loss": 0.9243, + "step": 12870 + }, + { + "epoch": 0.8843616875085887, + "grad_norm": 1.059394359588623, + "learning_rate": 2.0093823470343415e-06, + "loss": 0.9449, + "step": 12871 + }, + { + "epoch": 0.8844303971416793, + "grad_norm": 1.1003644466400146, + "learning_rate": 2.0070276643618575e-06, + "loss": 0.9153, + "step": 12872 + }, + { + "epoch": 0.8844991067747698, + "grad_norm": 1.089556097984314, + "learning_rate": 2.0046743044610544e-06, + "loss": 0.9162, + "step": 12873 + }, + { + "epoch": 0.8845678164078604, + "grad_norm": 1.0192128419876099, + "learning_rate": 2.002322267467316e-06, + "loss": 0.8774, + "step": 12874 + }, + { + "epoch": 0.8846365260409509, + "grad_norm": 1.075034260749817, + "learning_rate": 1.9999715535159557e-06, + "loss": 0.96, + "step": 12875 + }, + { + "epoch": 0.8847052356740415, + "grad_norm": 1.055349349975586, + "learning_rate": 1.9976221627422104e-06, + "loss": 0.9568, + "step": 12876 + }, + { + "epoch": 0.884773945307132, + "grad_norm": 1.02603280544281, + "learning_rate": 1.995274095281238e-06, + "loss": 0.9577, + "step": 12877 + }, + { + "epoch": 0.8848426549402226, + "grad_norm": 1.0826480388641357, + "learning_rate": 1.9929273512681175e-06, + "loss": 0.9296, + "step": 12878 + }, + { + "epoch": 0.8849113645733132, + "grad_norm": 0.9767555594444275, + "learning_rate": 1.9905819308378566e-06, + "loss": 0.907, + "step": 12879 + }, + { + "epoch": 0.8849800742064038, + "grad_norm": 1.0304615497589111, + "learning_rate": 1.988237834125395e-06, + "loss": 0.9041, + "step": 12880 + }, + { + "epoch": 0.8850487838394943, + "grad_norm": 1.062150478363037, + "learning_rate": 1.985895061265572e-06, + "loss": 0.9281, + "step": 12881 + }, + { + "epoch": 0.8851174934725848, + "grad_norm": 0.9992304444313049, + "learning_rate": 1.983553612393177e-06, + "loss": 0.9192, + "step": 12882 + }, + { + "epoch": 0.8851862031056754, + "grad_norm": 1.0129094123840332, + "learning_rate": 1.9812134876429104e-06, + "loss": 0.9253, + "step": 12883 + }, + { + "epoch": 0.885254912738766, + "grad_norm": 1.113974690437317, + "learning_rate": 1.978874687149393e-06, + "loss": 0.9476, + "step": 12884 + }, + { + "epoch": 0.8853236223718566, + "grad_norm": 1.0960193872451782, + "learning_rate": 1.9765372110471777e-06, + "loss": 0.968, + "step": 12885 + }, + { + "epoch": 0.885392332004947, + "grad_norm": 1.0232598781585693, + "learning_rate": 1.97420105947074e-06, + "loss": 0.9278, + "step": 12886 + }, + { + "epoch": 0.8854610416380376, + "grad_norm": 1.0858030319213867, + "learning_rate": 1.9718662325544784e-06, + "loss": 0.9534, + "step": 12887 + }, + { + "epoch": 0.8855297512711282, + "grad_norm": 1.1168406009674072, + "learning_rate": 1.9695327304327075e-06, + "loss": 0.9482, + "step": 12888 + }, + { + "epoch": 0.8855984609042188, + "grad_norm": 1.0400131940841675, + "learning_rate": 1.9672005532396758e-06, + "loss": 0.9269, + "step": 12889 + }, + { + "epoch": 0.8856671705373094, + "grad_norm": 1.014309287071228, + "learning_rate": 1.9648697011095537e-06, + "loss": 0.9016, + "step": 12890 + }, + { + "epoch": 0.8857358801703998, + "grad_norm": 1.0493816137313843, + "learning_rate": 1.9625401741764332e-06, + "loss": 0.9195, + "step": 12891 + }, + { + "epoch": 0.8858045898034904, + "grad_norm": 1.0382286310195923, + "learning_rate": 1.960211972574333e-06, + "loss": 0.907, + "step": 12892 + }, + { + "epoch": 0.885873299436581, + "grad_norm": 1.0758271217346191, + "learning_rate": 1.9578850964371843e-06, + "loss": 0.9213, + "step": 12893 + }, + { + "epoch": 0.8859420090696716, + "grad_norm": 1.0313291549682617, + "learning_rate": 1.955559545898858e-06, + "loss": 0.9453, + "step": 12894 + }, + { + "epoch": 0.8860107187027622, + "grad_norm": 1.0028270483016968, + "learning_rate": 1.953235321093139e-06, + "loss": 0.9184, + "step": 12895 + }, + { + "epoch": 0.8860794283358527, + "grad_norm": 1.0815415382385254, + "learning_rate": 1.950912422153739e-06, + "loss": 0.9314, + "step": 12896 + }, + { + "epoch": 0.8861481379689432, + "grad_norm": 1.095890998840332, + "learning_rate": 1.948590849214299e-06, + "loss": 0.9554, + "step": 12897 + }, + { + "epoch": 0.8862168476020338, + "grad_norm": 1.0694665908813477, + "learning_rate": 1.9462706024083675e-06, + "loss": 0.93, + "step": 12898 + }, + { + "epoch": 0.8862855572351244, + "grad_norm": 0.9554431438446045, + "learning_rate": 1.9439516818694326e-06, + "loss": 0.8701, + "step": 12899 + }, + { + "epoch": 0.886354266868215, + "grad_norm": 1.053733229637146, + "learning_rate": 1.9416340877309008e-06, + "loss": 0.9077, + "step": 12900 + }, + { + "epoch": 0.8864229765013055, + "grad_norm": 1.055965542793274, + "learning_rate": 1.9393178201260966e-06, + "loss": 0.8967, + "step": 12901 + }, + { + "epoch": 0.886491686134396, + "grad_norm": 1.0262326002120972, + "learning_rate": 1.937002879188285e-06, + "loss": 0.9164, + "step": 12902 + }, + { + "epoch": 0.8865603957674866, + "grad_norm": 1.0776751041412354, + "learning_rate": 1.9346892650506295e-06, + "loss": 0.9145, + "step": 12903 + }, + { + "epoch": 0.8866291054005772, + "grad_norm": 1.098089575767517, + "learning_rate": 1.932376977846237e-06, + "loss": 0.9064, + "step": 12904 + }, + { + "epoch": 0.8866978150336677, + "grad_norm": 1.0546073913574219, + "learning_rate": 1.9300660177081346e-06, + "loss": 0.9416, + "step": 12905 + }, + { + "epoch": 0.8867665246667583, + "grad_norm": 1.0302714109420776, + "learning_rate": 1.9277563847692666e-06, + "loss": 0.9046, + "step": 12906 + }, + { + "epoch": 0.8868352342998488, + "grad_norm": 1.0087051391601562, + "learning_rate": 1.925448079162509e-06, + "loss": 0.8871, + "step": 12907 + }, + { + "epoch": 0.8869039439329394, + "grad_norm": 1.10176682472229, + "learning_rate": 1.923141101020648e-06, + "loss": 0.9472, + "step": 12908 + }, + { + "epoch": 0.88697265356603, + "grad_norm": 1.0255216360092163, + "learning_rate": 1.9208354504764102e-06, + "loss": 0.9328, + "step": 12909 + }, + { + "epoch": 0.8870413631991205, + "grad_norm": 1.143277645111084, + "learning_rate": 1.9185311276624372e-06, + "loss": 0.9519, + "step": 12910 + }, + { + "epoch": 0.8871100728322111, + "grad_norm": 1.1239997148513794, + "learning_rate": 1.916228132711295e-06, + "loss": 0.9555, + "step": 12911 + }, + { + "epoch": 0.8871787824653017, + "grad_norm": 1.03679358959198, + "learning_rate": 1.9139264657554778e-06, + "loss": 0.9131, + "step": 12912 + }, + { + "epoch": 0.8872474920983922, + "grad_norm": 1.0743961334228516, + "learning_rate": 1.9116261269273907e-06, + "loss": 0.904, + "step": 12913 + }, + { + "epoch": 0.8873162017314827, + "grad_norm": 1.08387291431427, + "learning_rate": 1.9093271163593717e-06, + "loss": 0.9418, + "step": 12914 + }, + { + "epoch": 0.8873849113645733, + "grad_norm": 1.05419921875, + "learning_rate": 1.9070294341836882e-06, + "loss": 0.9523, + "step": 12915 + }, + { + "epoch": 0.8874536209976639, + "grad_norm": 1.1249432563781738, + "learning_rate": 1.9047330805325198e-06, + "loss": 0.9381, + "step": 12916 + }, + { + "epoch": 0.8875223306307545, + "grad_norm": 1.030863881111145, + "learning_rate": 1.9024380555379778e-06, + "loss": 0.8922, + "step": 12917 + }, + { + "epoch": 0.8875910402638449, + "grad_norm": 1.071640968322754, + "learning_rate": 1.90014435933209e-06, + "loss": 0.9439, + "step": 12918 + }, + { + "epoch": 0.8876597498969355, + "grad_norm": 1.0976345539093018, + "learning_rate": 1.8978519920468091e-06, + "loss": 0.9591, + "step": 12919 + }, + { + "epoch": 0.8877284595300261, + "grad_norm": 1.0717896223068237, + "learning_rate": 1.8955609538140184e-06, + "loss": 0.9433, + "step": 12920 + }, + { + "epoch": 0.8877971691631167, + "grad_norm": 1.08059561252594, + "learning_rate": 1.8932712447655181e-06, + "loss": 0.9214, + "step": 12921 + }, + { + "epoch": 0.8878658787962073, + "grad_norm": 1.0525034666061401, + "learning_rate": 1.8909828650330391e-06, + "loss": 0.9332, + "step": 12922 + }, + { + "epoch": 0.8879345884292977, + "grad_norm": 1.118007779121399, + "learning_rate": 1.888695814748223e-06, + "loss": 0.9519, + "step": 12923 + }, + { + "epoch": 0.8880032980623883, + "grad_norm": 1.0419671535491943, + "learning_rate": 1.8864100940426426e-06, + "loss": 0.9407, + "step": 12924 + }, + { + "epoch": 0.8880720076954789, + "grad_norm": 1.0801239013671875, + "learning_rate": 1.8841257030477954e-06, + "loss": 0.9122, + "step": 12925 + }, + { + "epoch": 0.8881407173285695, + "grad_norm": 1.0511034727096558, + "learning_rate": 1.881842641895104e-06, + "loss": 0.9584, + "step": 12926 + }, + { + "epoch": 0.88820942696166, + "grad_norm": 1.1063826084136963, + "learning_rate": 1.8795609107159134e-06, + "loss": 0.9457, + "step": 12927 + }, + { + "epoch": 0.8882781365947506, + "grad_norm": 1.0298837423324585, + "learning_rate": 1.8772805096414825e-06, + "loss": 0.9191, + "step": 12928 + }, + { + "epoch": 0.8883468462278411, + "grad_norm": 1.0567148923873901, + "learning_rate": 1.8750014388030035e-06, + "loss": 0.9289, + "step": 12929 + }, + { + "epoch": 0.8884155558609317, + "grad_norm": 1.066214919090271, + "learning_rate": 1.8727236983315965e-06, + "loss": 0.9617, + "step": 12930 + }, + { + "epoch": 0.8884842654940223, + "grad_norm": 1.0662788152694702, + "learning_rate": 1.8704472883582902e-06, + "loss": 0.9338, + "step": 12931 + }, + { + "epoch": 0.8885529751271128, + "grad_norm": 1.107776403427124, + "learning_rate": 1.8681722090140492e-06, + "loss": 0.961, + "step": 12932 + }, + { + "epoch": 0.8886216847602034, + "grad_norm": 1.1383857727050781, + "learning_rate": 1.8658984604297603e-06, + "loss": 0.9688, + "step": 12933 + }, + { + "epoch": 0.8886903943932939, + "grad_norm": 1.119112491607666, + "learning_rate": 1.8636260427362218e-06, + "loss": 0.9698, + "step": 12934 + }, + { + "epoch": 0.8887591040263845, + "grad_norm": 0.9956680536270142, + "learning_rate": 1.8613549560641707e-06, + "loss": 0.9343, + "step": 12935 + }, + { + "epoch": 0.888827813659475, + "grad_norm": 1.1200908422470093, + "learning_rate": 1.859085200544261e-06, + "loss": 0.974, + "step": 12936 + }, + { + "epoch": 0.8888965232925656, + "grad_norm": 1.0576437711715698, + "learning_rate": 1.8568167763070687e-06, + "loss": 0.9149, + "step": 12937 + }, + { + "epoch": 0.8889652329256562, + "grad_norm": 1.0303457975387573, + "learning_rate": 1.8545496834831032e-06, + "loss": 0.9361, + "step": 12938 + }, + { + "epoch": 0.8890339425587467, + "grad_norm": 1.085960865020752, + "learning_rate": 1.8522839222027743e-06, + "loss": 0.9556, + "step": 12939 + }, + { + "epoch": 0.8891026521918373, + "grad_norm": 1.1163914203643799, + "learning_rate": 1.8500194925964387e-06, + "loss": 0.9905, + "step": 12940 + }, + { + "epoch": 0.8891713618249278, + "grad_norm": 1.0381531715393066, + "learning_rate": 1.847756394794367e-06, + "loss": 0.9188, + "step": 12941 + }, + { + "epoch": 0.8892400714580184, + "grad_norm": 1.032021403312683, + "learning_rate": 1.8454946289267527e-06, + "loss": 0.9269, + "step": 12942 + }, + { + "epoch": 0.889308781091109, + "grad_norm": 1.0884228944778442, + "learning_rate": 1.8432341951237192e-06, + "loss": 0.9346, + "step": 12943 + }, + { + "epoch": 0.8893774907241996, + "grad_norm": 1.0538352727890015, + "learning_rate": 1.8409750935152986e-06, + "loss": 0.911, + "step": 12944 + }, + { + "epoch": 0.88944620035729, + "grad_norm": 1.0733946561813354, + "learning_rate": 1.8387173242314592e-06, + "loss": 0.9425, + "step": 12945 + }, + { + "epoch": 0.8895149099903806, + "grad_norm": 1.0252853631973267, + "learning_rate": 1.8364608874020916e-06, + "loss": 0.8861, + "step": 12946 + }, + { + "epoch": 0.8895836196234712, + "grad_norm": 1.1105430126190186, + "learning_rate": 1.8342057831570059e-06, + "loss": 0.9569, + "step": 12947 + }, + { + "epoch": 0.8896523292565618, + "grad_norm": 1.1019396781921387, + "learning_rate": 1.8319520116259369e-06, + "loss": 0.9652, + "step": 12948 + }, + { + "epoch": 0.8897210388896524, + "grad_norm": 1.0934770107269287, + "learning_rate": 1.8296995729385425e-06, + "loss": 0.9063, + "step": 12949 + }, + { + "epoch": 0.8897897485227428, + "grad_norm": 1.0915441513061523, + "learning_rate": 1.827448467224402e-06, + "loss": 0.9289, + "step": 12950 + }, + { + "epoch": 0.8898584581558334, + "grad_norm": 1.0927276611328125, + "learning_rate": 1.8251986946130228e-06, + "loss": 0.9641, + "step": 12951 + }, + { + "epoch": 0.889927167788924, + "grad_norm": 1.0902122259140015, + "learning_rate": 1.8229502552338351e-06, + "loss": 0.9098, + "step": 12952 + }, + { + "epoch": 0.8899958774220146, + "grad_norm": 1.0817852020263672, + "learning_rate": 1.8207031492161824e-06, + "loss": 0.9761, + "step": 12953 + }, + { + "epoch": 0.8900645870551052, + "grad_norm": 0.9801344275474548, + "learning_rate": 1.8184573766893504e-06, + "loss": 0.9021, + "step": 12954 + }, + { + "epoch": 0.8901332966881956, + "grad_norm": 1.0709258317947388, + "learning_rate": 1.816212937782527e-06, + "loss": 0.9491, + "step": 12955 + }, + { + "epoch": 0.8902020063212862, + "grad_norm": 1.0926517248153687, + "learning_rate": 1.8139698326248373e-06, + "loss": 0.914, + "step": 12956 + }, + { + "epoch": 0.8902707159543768, + "grad_norm": 1.0387698411941528, + "learning_rate": 1.8117280613453247e-06, + "loss": 0.9189, + "step": 12957 + }, + { + "epoch": 0.8903394255874674, + "grad_norm": 1.0483827590942383, + "learning_rate": 1.809487624072964e-06, + "loss": 0.9324, + "step": 12958 + }, + { + "epoch": 0.890408135220558, + "grad_norm": 1.0998449325561523, + "learning_rate": 1.8072485209366358e-06, + "loss": 0.9132, + "step": 12959 + }, + { + "epoch": 0.8904768448536485, + "grad_norm": 1.0927729606628418, + "learning_rate": 1.805010752065156e-06, + "loss": 0.9184, + "step": 12960 + }, + { + "epoch": 0.890545554486739, + "grad_norm": 1.0278884172439575, + "learning_rate": 1.8027743175872664e-06, + "loss": 0.9468, + "step": 12961 + }, + { + "epoch": 0.8906142641198296, + "grad_norm": 1.1095607280731201, + "learning_rate": 1.8005392176316276e-06, + "loss": 0.9181, + "step": 12962 + }, + { + "epoch": 0.8906829737529202, + "grad_norm": 1.0512120723724365, + "learning_rate": 1.798305452326826e-06, + "loss": 0.925, + "step": 12963 + }, + { + "epoch": 0.8907516833860107, + "grad_norm": 1.1090726852416992, + "learning_rate": 1.796073021801359e-06, + "loss": 0.9324, + "step": 12964 + }, + { + "epoch": 0.8908203930191013, + "grad_norm": 1.0534465312957764, + "learning_rate": 1.7938419261836648e-06, + "loss": 0.8645, + "step": 12965 + }, + { + "epoch": 0.8908891026521918, + "grad_norm": 1.0552492141723633, + "learning_rate": 1.791612165602094e-06, + "loss": 0.9187, + "step": 12966 + }, + { + "epoch": 0.8909578122852824, + "grad_norm": 1.1158394813537598, + "learning_rate": 1.7893837401849245e-06, + "loss": 0.9653, + "step": 12967 + }, + { + "epoch": 0.891026521918373, + "grad_norm": 1.024705171585083, + "learning_rate": 1.787156650060362e-06, + "loss": 0.938, + "step": 12968 + }, + { + "epoch": 0.8910952315514635, + "grad_norm": 1.0974828004837036, + "learning_rate": 1.784930895356518e-06, + "loss": 0.9442, + "step": 12969 + }, + { + "epoch": 0.8911639411845541, + "grad_norm": 1.081609845161438, + "learning_rate": 1.7827064762014427e-06, + "loss": 0.9336, + "step": 12970 + }, + { + "epoch": 0.8912326508176446, + "grad_norm": 1.021323323249817, + "learning_rate": 1.780483392723109e-06, + "loss": 0.9268, + "step": 12971 + }, + { + "epoch": 0.8913013604507352, + "grad_norm": 1.039209246635437, + "learning_rate": 1.7782616450494088e-06, + "loss": 0.9292, + "step": 12972 + }, + { + "epoch": 0.8913700700838257, + "grad_norm": 1.100865364074707, + "learning_rate": 1.7760412333081622e-06, + "loss": 0.952, + "step": 12973 + }, + { + "epoch": 0.8914387797169163, + "grad_norm": 1.0828189849853516, + "learning_rate": 1.7738221576270975e-06, + "loss": 0.9465, + "step": 12974 + }, + { + "epoch": 0.8915074893500069, + "grad_norm": 1.041393518447876, + "learning_rate": 1.771604418133882e-06, + "loss": 0.9281, + "step": 12975 + }, + { + "epoch": 0.8915761989830975, + "grad_norm": 1.0690085887908936, + "learning_rate": 1.7693880149561055e-06, + "loss": 0.9485, + "step": 12976 + }, + { + "epoch": 0.891644908616188, + "grad_norm": 1.0070768594741821, + "learning_rate": 1.7671729482212684e-06, + "loss": 0.9178, + "step": 12977 + }, + { + "epoch": 0.8917136182492785, + "grad_norm": 1.0720112323760986, + "learning_rate": 1.7649592180568053e-06, + "loss": 0.9711, + "step": 12978 + }, + { + "epoch": 0.8917823278823691, + "grad_norm": 1.0535391569137573, + "learning_rate": 1.7627468245900752e-06, + "loss": 0.9552, + "step": 12979 + }, + { + "epoch": 0.8918510375154597, + "grad_norm": 1.0482792854309082, + "learning_rate": 1.7605357679483486e-06, + "loss": 0.8956, + "step": 12980 + }, + { + "epoch": 0.8919197471485503, + "grad_norm": 1.0133949518203735, + "learning_rate": 1.7583260482588266e-06, + "loss": 0.9106, + "step": 12981 + }, + { + "epoch": 0.8919884567816407, + "grad_norm": 1.0999187231063843, + "learning_rate": 1.7561176656486405e-06, + "loss": 0.9323, + "step": 12982 + }, + { + "epoch": 0.8920571664147313, + "grad_norm": 1.046897053718567, + "learning_rate": 1.7539106202448336e-06, + "loss": 0.9708, + "step": 12983 + }, + { + "epoch": 0.8921258760478219, + "grad_norm": 1.0457667112350464, + "learning_rate": 1.7517049121743705e-06, + "loss": 0.9192, + "step": 12984 + }, + { + "epoch": 0.8921945856809125, + "grad_norm": 1.063186764717102, + "learning_rate": 1.7495005415641503e-06, + "loss": 0.9385, + "step": 12985 + }, + { + "epoch": 0.8922632953140031, + "grad_norm": 1.0732126235961914, + "learning_rate": 1.7472975085409847e-06, + "loss": 0.9688, + "step": 12986 + }, + { + "epoch": 0.8923320049470936, + "grad_norm": 1.0568833351135254, + "learning_rate": 1.7450958132316175e-06, + "loss": 0.8924, + "step": 12987 + }, + { + "epoch": 0.8924007145801841, + "grad_norm": 1.0923070907592773, + "learning_rate": 1.742895455762708e-06, + "loss": 0.9306, + "step": 12988 + }, + { + "epoch": 0.8924694242132747, + "grad_norm": 1.0700801610946655, + "learning_rate": 1.7406964362608469e-06, + "loss": 0.8861, + "step": 12989 + }, + { + "epoch": 0.8925381338463653, + "grad_norm": 1.0431126356124878, + "learning_rate": 1.7384987548525328e-06, + "loss": 0.921, + "step": 12990 + }, + { + "epoch": 0.8926068434794558, + "grad_norm": 1.0076522827148438, + "learning_rate": 1.7363024116642034e-06, + "loss": 0.9092, + "step": 12991 + }, + { + "epoch": 0.8926755531125464, + "grad_norm": 1.015762448310852, + "learning_rate": 1.7341074068222107e-06, + "loss": 0.9302, + "step": 12992 + }, + { + "epoch": 0.8927442627456369, + "grad_norm": 1.0283981561660767, + "learning_rate": 1.7319137404528342e-06, + "loss": 0.8981, + "step": 12993 + }, + { + "epoch": 0.8928129723787275, + "grad_norm": 1.1037921905517578, + "learning_rate": 1.7297214126822752e-06, + "loss": 0.9129, + "step": 12994 + }, + { + "epoch": 0.8928816820118181, + "grad_norm": 1.078298807144165, + "learning_rate": 1.727530423636653e-06, + "loss": 0.9325, + "step": 12995 + }, + { + "epoch": 0.8929503916449086, + "grad_norm": 1.0398716926574707, + "learning_rate": 1.7253407734420136e-06, + "loss": 0.9671, + "step": 12996 + }, + { + "epoch": 0.8930191012779992, + "grad_norm": 1.1059746742248535, + "learning_rate": 1.7231524622243284e-06, + "loss": 0.923, + "step": 12997 + }, + { + "epoch": 0.8930878109110897, + "grad_norm": 1.0619158744812012, + "learning_rate": 1.7209654901094941e-06, + "loss": 0.9047, + "step": 12998 + }, + { + "epoch": 0.8931565205441803, + "grad_norm": 1.0573663711547852, + "learning_rate": 1.7187798572233154e-06, + "loss": 0.9253, + "step": 12999 + }, + { + "epoch": 0.8932252301772708, + "grad_norm": 1.0645639896392822, + "learning_rate": 1.7165955636915392e-06, + "loss": 0.9255, + "step": 13000 + }, + { + "epoch": 0.8932939398103614, + "grad_norm": 1.0888164043426514, + "learning_rate": 1.7144126096398262e-06, + "loss": 0.9685, + "step": 13001 + }, + { + "epoch": 0.893362649443452, + "grad_norm": 1.0780802965164185, + "learning_rate": 1.7122309951937533e-06, + "loss": 0.9324, + "step": 13002 + }, + { + "epoch": 0.8934313590765426, + "grad_norm": 1.0631284713745117, + "learning_rate": 1.7100507204788312e-06, + "loss": 0.9387, + "step": 13003 + }, + { + "epoch": 0.893500068709633, + "grad_norm": 1.0330122709274292, + "learning_rate": 1.7078717856204961e-06, + "loss": 0.8881, + "step": 13004 + }, + { + "epoch": 0.8935687783427236, + "grad_norm": 1.02329421043396, + "learning_rate": 1.7056941907440894e-06, + "loss": 0.8834, + "step": 13005 + }, + { + "epoch": 0.8936374879758142, + "grad_norm": 1.0646148920059204, + "learning_rate": 1.7035179359748937e-06, + "loss": 0.9383, + "step": 13006 + }, + { + "epoch": 0.8937061976089048, + "grad_norm": 1.0398564338684082, + "learning_rate": 1.7013430214381066e-06, + "loss": 0.9253, + "step": 13007 + }, + { + "epoch": 0.8937749072419954, + "grad_norm": 1.0522466897964478, + "learning_rate": 1.6991694472588499e-06, + "loss": 0.9474, + "step": 13008 + }, + { + "epoch": 0.8938436168750858, + "grad_norm": 1.0406442880630493, + "learning_rate": 1.6969972135621709e-06, + "loss": 0.9479, + "step": 13009 + }, + { + "epoch": 0.8939123265081764, + "grad_norm": 1.022267460823059, + "learning_rate": 1.6948263204730281e-06, + "loss": 0.9284, + "step": 13010 + }, + { + "epoch": 0.893981036141267, + "grad_norm": 0.994902491569519, + "learning_rate": 1.6926567681163186e-06, + "loss": 0.8709, + "step": 13011 + }, + { + "epoch": 0.8940497457743576, + "grad_norm": 1.0944427251815796, + "learning_rate": 1.6904885566168566e-06, + "loss": 0.9548, + "step": 13012 + }, + { + "epoch": 0.8941184554074482, + "grad_norm": 1.0448687076568604, + "learning_rate": 1.6883216860993727e-06, + "loss": 0.918, + "step": 13013 + }, + { + "epoch": 0.8941871650405386, + "grad_norm": 1.0163553953170776, + "learning_rate": 1.6861561566885315e-06, + "loss": 0.9016, + "step": 13014 + }, + { + "epoch": 0.8942558746736292, + "grad_norm": 1.1321029663085938, + "learning_rate": 1.6839919685089077e-06, + "loss": 0.9449, + "step": 13015 + }, + { + "epoch": 0.8943245843067198, + "grad_norm": 1.0325450897216797, + "learning_rate": 1.6818291216850108e-06, + "loss": 0.934, + "step": 13016 + }, + { + "epoch": 0.8943932939398104, + "grad_norm": 1.0540755987167358, + "learning_rate": 1.6796676163412656e-06, + "loss": 0.9036, + "step": 13017 + }, + { + "epoch": 0.894462003572901, + "grad_norm": 1.1279776096343994, + "learning_rate": 1.677507452602023e-06, + "loss": 0.9408, + "step": 13018 + }, + { + "epoch": 0.8945307132059915, + "grad_norm": 1.0441670417785645, + "learning_rate": 1.6753486305915611e-06, + "loss": 0.9169, + "step": 13019 + }, + { + "epoch": 0.894599422839082, + "grad_norm": 1.0644164085388184, + "learning_rate": 1.673191150434067e-06, + "loss": 0.9384, + "step": 13020 + }, + { + "epoch": 0.8946681324721726, + "grad_norm": 1.059712529182434, + "learning_rate": 1.6710350122536634e-06, + "loss": 0.929, + "step": 13021 + }, + { + "epoch": 0.8947368421052632, + "grad_norm": 0.9940868616104126, + "learning_rate": 1.66888021617439e-06, + "loss": 0.9182, + "step": 13022 + }, + { + "epoch": 0.8948055517383537, + "grad_norm": 1.0342358350753784, + "learning_rate": 1.6667267623202142e-06, + "loss": 0.8994, + "step": 13023 + }, + { + "epoch": 0.8948742613714443, + "grad_norm": 1.0909676551818848, + "learning_rate": 1.664574650815018e-06, + "loss": 0.9561, + "step": 13024 + }, + { + "epoch": 0.8949429710045348, + "grad_norm": 1.0450118780136108, + "learning_rate": 1.6624238817826183e-06, + "loss": 0.9173, + "step": 13025 + }, + { + "epoch": 0.8950116806376254, + "grad_norm": 1.0423089265823364, + "learning_rate": 1.6602744553467386e-06, + "loss": 0.8936, + "step": 13026 + }, + { + "epoch": 0.895080390270716, + "grad_norm": 1.0698617696762085, + "learning_rate": 1.6581263716310386e-06, + "loss": 0.9198, + "step": 13027 + }, + { + "epoch": 0.8951490999038065, + "grad_norm": 1.0356107950210571, + "learning_rate": 1.6559796307590969e-06, + "loss": 0.9048, + "step": 13028 + }, + { + "epoch": 0.8952178095368971, + "grad_norm": 1.0537760257720947, + "learning_rate": 1.6538342328544176e-06, + "loss": 0.9505, + "step": 13029 + }, + { + "epoch": 0.8952865191699876, + "grad_norm": 1.0229166746139526, + "learning_rate": 1.651690178040416e-06, + "loss": 0.9185, + "step": 13030 + }, + { + "epoch": 0.8953552288030782, + "grad_norm": 1.132956862449646, + "learning_rate": 1.6495474664404403e-06, + "loss": 0.9296, + "step": 13031 + }, + { + "epoch": 0.8954239384361687, + "grad_norm": 1.0755305290222168, + "learning_rate": 1.6474060981777617e-06, + "loss": 0.97, + "step": 13032 + }, + { + "epoch": 0.8954926480692593, + "grad_norm": 1.017390489578247, + "learning_rate": 1.6452660733755731e-06, + "loss": 0.9236, + "step": 13033 + }, + { + "epoch": 0.8955613577023499, + "grad_norm": 1.0660332441329956, + "learning_rate": 1.64312739215699e-06, + "loss": 0.9322, + "step": 13034 + }, + { + "epoch": 0.8956300673354405, + "grad_norm": 1.0739063024520874, + "learning_rate": 1.6409900546450418e-06, + "loss": 0.9262, + "step": 13035 + }, + { + "epoch": 0.895698776968531, + "grad_norm": 1.074304223060608, + "learning_rate": 1.638854060962694e-06, + "loss": 0.9373, + "step": 13036 + }, + { + "epoch": 0.8957674866016215, + "grad_norm": 1.0940072536468506, + "learning_rate": 1.6367194112328288e-06, + "loss": 0.9368, + "step": 13037 + }, + { + "epoch": 0.8958361962347121, + "grad_norm": 1.062418818473816, + "learning_rate": 1.6345861055782508e-06, + "loss": 0.9266, + "step": 13038 + }, + { + "epoch": 0.8959049058678027, + "grad_norm": 1.0754810571670532, + "learning_rate": 1.6324541441216896e-06, + "loss": 0.9562, + "step": 13039 + }, + { + "epoch": 0.8959736155008933, + "grad_norm": 1.0587488412857056, + "learning_rate": 1.6303235269857887e-06, + "loss": 0.9385, + "step": 13040 + }, + { + "epoch": 0.8960423251339837, + "grad_norm": 1.0015391111373901, + "learning_rate": 1.6281942542931279e-06, + "loss": 0.9038, + "step": 13041 + }, + { + "epoch": 0.8961110347670743, + "grad_norm": 1.1216907501220703, + "learning_rate": 1.6260663261662008e-06, + "loss": 0.9206, + "step": 13042 + }, + { + "epoch": 0.8961797444001649, + "grad_norm": 1.0519887208938599, + "learning_rate": 1.6239397427274262e-06, + "loss": 0.9378, + "step": 13043 + }, + { + "epoch": 0.8962484540332555, + "grad_norm": 1.0937221050262451, + "learning_rate": 1.621814504099145e-06, + "loss": 0.9254, + "step": 13044 + }, + { + "epoch": 0.8963171636663461, + "grad_norm": 1.0896426439285278, + "learning_rate": 1.6196906104036264e-06, + "loss": 0.9417, + "step": 13045 + }, + { + "epoch": 0.8963858732994365, + "grad_norm": 1.0070890188217163, + "learning_rate": 1.6175680617630446e-06, + "loss": 0.8854, + "step": 13046 + }, + { + "epoch": 0.8964545829325271, + "grad_norm": 1.0897057056427002, + "learning_rate": 1.6154468582995214e-06, + "loss": 0.9631, + "step": 13047 + }, + { + "epoch": 0.8965232925656177, + "grad_norm": 1.116299033164978, + "learning_rate": 1.613327000135076e-06, + "loss": 0.9512, + "step": 13048 + }, + { + "epoch": 0.8965920021987083, + "grad_norm": 1.1446698904037476, + "learning_rate": 1.6112084873916688e-06, + "loss": 0.9661, + "step": 13049 + }, + { + "epoch": 0.8966607118317989, + "grad_norm": 1.0469019412994385, + "learning_rate": 1.6090913201911807e-06, + "loss": 0.909, + "step": 13050 + }, + { + "epoch": 0.8967294214648894, + "grad_norm": 1.07023286819458, + "learning_rate": 1.6069754986554054e-06, + "loss": 0.9596, + "step": 13051 + }, + { + "epoch": 0.8967981310979799, + "grad_norm": 1.0817533731460571, + "learning_rate": 1.6048610229060624e-06, + "loss": 0.8793, + "step": 13052 + }, + { + "epoch": 0.8968668407310705, + "grad_norm": 1.0897029638290405, + "learning_rate": 1.6027478930648015e-06, + "loss": 0.9285, + "step": 13053 + }, + { + "epoch": 0.8969355503641611, + "grad_norm": 1.0621159076690674, + "learning_rate": 1.6006361092531896e-06, + "loss": 0.9475, + "step": 13054 + }, + { + "epoch": 0.8970042599972516, + "grad_norm": 1.019881010055542, + "learning_rate": 1.5985256715927154e-06, + "loss": 0.8864, + "step": 13055 + }, + { + "epoch": 0.8970729696303422, + "grad_norm": 1.0774048566818237, + "learning_rate": 1.5964165802047904e-06, + "loss": 0.9271, + "step": 13056 + }, + { + "epoch": 0.8971416792634327, + "grad_norm": 1.0337494611740112, + "learning_rate": 1.5943088352107476e-06, + "loss": 0.9584, + "step": 13057 + }, + { + "epoch": 0.8972103888965233, + "grad_norm": 1.0593594312667847, + "learning_rate": 1.5922024367318461e-06, + "loss": 0.9064, + "step": 13058 + }, + { + "epoch": 0.8972790985296138, + "grad_norm": 1.0615147352218628, + "learning_rate": 1.5900973848892693e-06, + "loss": 0.9334, + "step": 13059 + }, + { + "epoch": 0.8973478081627044, + "grad_norm": 1.0546109676361084, + "learning_rate": 1.5879936798041173e-06, + "loss": 0.9218, + "step": 13060 + }, + { + "epoch": 0.897416517795795, + "grad_norm": 1.0399318933486938, + "learning_rate": 1.5858913215974076e-06, + "loss": 0.9236, + "step": 13061 + }, + { + "epoch": 0.8974852274288855, + "grad_norm": 1.077266812324524, + "learning_rate": 1.5837903103900987e-06, + "loss": 0.9486, + "step": 13062 + }, + { + "epoch": 0.8975539370619761, + "grad_norm": 1.0323177576065063, + "learning_rate": 1.5816906463030523e-06, + "loss": 0.9282, + "step": 13063 + }, + { + "epoch": 0.8976226466950666, + "grad_norm": 1.0342739820480347, + "learning_rate": 1.5795923294570635e-06, + "loss": 0.9424, + "step": 13064 + }, + { + "epoch": 0.8976913563281572, + "grad_norm": 1.096860647201538, + "learning_rate": 1.5774953599728554e-06, + "loss": 0.926, + "step": 13065 + }, + { + "epoch": 0.8977600659612478, + "grad_norm": 1.0314956903457642, + "learning_rate": 1.5753997379710506e-06, + "loss": 0.9291, + "step": 13066 + }, + { + "epoch": 0.8978287755943384, + "grad_norm": 1.1085566282272339, + "learning_rate": 1.5733054635722166e-06, + "loss": 0.9548, + "step": 13067 + }, + { + "epoch": 0.8978974852274288, + "grad_norm": 1.0436372756958008, + "learning_rate": 1.571212536896835e-06, + "loss": 0.921, + "step": 13068 + }, + { + "epoch": 0.8979661948605194, + "grad_norm": 1.0498597621917725, + "learning_rate": 1.5691209580653176e-06, + "loss": 0.9369, + "step": 13069 + }, + { + "epoch": 0.89803490449361, + "grad_norm": 1.092474102973938, + "learning_rate": 1.5670307271979768e-06, + "loss": 0.909, + "step": 13070 + }, + { + "epoch": 0.8981036141267006, + "grad_norm": 1.0397502183914185, + "learning_rate": 1.5649418444150743e-06, + "loss": 0.9196, + "step": 13071 + }, + { + "epoch": 0.8981723237597912, + "grad_norm": 1.039810299873352, + "learning_rate": 1.5628543098367782e-06, + "loss": 0.9186, + "step": 13072 + }, + { + "epoch": 0.8982410333928816, + "grad_norm": 1.1619365215301514, + "learning_rate": 1.560768123583184e-06, + "loss": 0.9505, + "step": 13073 + }, + { + "epoch": 0.8983097430259722, + "grad_norm": 1.157943844795227, + "learning_rate": 1.558683285774304e-06, + "loss": 0.9209, + "step": 13074 + }, + { + "epoch": 0.8983784526590628, + "grad_norm": 0.9965066909790039, + "learning_rate": 1.5565997965300865e-06, + "loss": 0.9117, + "step": 13075 + }, + { + "epoch": 0.8984471622921534, + "grad_norm": 1.051535725593567, + "learning_rate": 1.554517655970386e-06, + "loss": 0.9, + "step": 13076 + }, + { + "epoch": 0.898515871925244, + "grad_norm": 1.09810209274292, + "learning_rate": 1.55243686421499e-06, + "loss": 0.9305, + "step": 13077 + }, + { + "epoch": 0.8985845815583344, + "grad_norm": 1.0804853439331055, + "learning_rate": 1.5503574213836024e-06, + "loss": 0.9449, + "step": 13078 + }, + { + "epoch": 0.898653291191425, + "grad_norm": 1.142663836479187, + "learning_rate": 1.5482793275958529e-06, + "loss": 0.9729, + "step": 13079 + }, + { + "epoch": 0.8987220008245156, + "grad_norm": 1.028725028038025, + "learning_rate": 1.5462025829713011e-06, + "loss": 0.8929, + "step": 13080 + }, + { + "epoch": 0.8987907104576062, + "grad_norm": 1.0892670154571533, + "learning_rate": 1.54412718762941e-06, + "loss": 0.91, + "step": 13081 + }, + { + "epoch": 0.8988594200906967, + "grad_norm": 1.0849767923355103, + "learning_rate": 1.5420531416895784e-06, + "loss": 0.9326, + "step": 13082 + }, + { + "epoch": 0.8989281297237873, + "grad_norm": 1.0799999237060547, + "learning_rate": 1.5399804452711276e-06, + "loss": 0.898, + "step": 13083 + }, + { + "epoch": 0.8989968393568778, + "grad_norm": 1.0385297536849976, + "learning_rate": 1.5379090984932981e-06, + "loss": 0.9321, + "step": 13084 + }, + { + "epoch": 0.8990655489899684, + "grad_norm": 1.0446807146072388, + "learning_rate": 1.535839101475256e-06, + "loss": 0.9133, + "step": 13085 + }, + { + "epoch": 0.899134258623059, + "grad_norm": 1.053049921989441, + "learning_rate": 1.5337704543360781e-06, + "loss": 0.9324, + "step": 13086 + }, + { + "epoch": 0.8992029682561495, + "grad_norm": 0.9806519746780396, + "learning_rate": 1.5317031571947803e-06, + "loss": 0.9007, + "step": 13087 + }, + { + "epoch": 0.8992716778892401, + "grad_norm": 1.052438735961914, + "learning_rate": 1.529637210170287e-06, + "loss": 0.8841, + "step": 13088 + }, + { + "epoch": 0.8993403875223306, + "grad_norm": 1.0835044384002686, + "learning_rate": 1.5275726133814588e-06, + "loss": 0.9607, + "step": 13089 + }, + { + "epoch": 0.8994090971554212, + "grad_norm": 1.0984442234039307, + "learning_rate": 1.525509366947067e-06, + "loss": 0.9204, + "step": 13090 + }, + { + "epoch": 0.8994778067885117, + "grad_norm": 1.0672355890274048, + "learning_rate": 1.5234474709858059e-06, + "loss": 0.9008, + "step": 13091 + }, + { + "epoch": 0.8995465164216023, + "grad_norm": 1.0846867561340332, + "learning_rate": 1.5213869256162945e-06, + "loss": 0.9526, + "step": 13092 + }, + { + "epoch": 0.8996152260546929, + "grad_norm": 1.0873157978057861, + "learning_rate": 1.519327730957082e-06, + "loss": 0.9293, + "step": 13093 + }, + { + "epoch": 0.8996839356877834, + "grad_norm": 1.1039434671401978, + "learning_rate": 1.5172698871266295e-06, + "loss": 0.9271, + "step": 13094 + }, + { + "epoch": 0.899752645320874, + "grad_norm": 1.109310269355774, + "learning_rate": 1.5152133942433172e-06, + "loss": 0.9833, + "step": 13095 + }, + { + "epoch": 0.8998213549539645, + "grad_norm": 1.042345643043518, + "learning_rate": 1.5131582524254616e-06, + "loss": 0.8983, + "step": 13096 + }, + { + "epoch": 0.8998900645870551, + "grad_norm": 1.0867055654525757, + "learning_rate": 1.5111044617912933e-06, + "loss": 0.9293, + "step": 13097 + }, + { + "epoch": 0.8999587742201457, + "grad_norm": 1.0368602275848389, + "learning_rate": 1.5090520224589594e-06, + "loss": 0.9089, + "step": 13098 + }, + { + "epoch": 0.9000274838532363, + "grad_norm": 1.0691685676574707, + "learning_rate": 1.5070009345465403e-06, + "loss": 0.887, + "step": 13099 + }, + { + "epoch": 0.9000961934863267, + "grad_norm": 1.1054681539535522, + "learning_rate": 1.5049511981720305e-06, + "loss": 0.9322, + "step": 13100 + }, + { + "epoch": 0.9001649031194173, + "grad_norm": 1.0786595344543457, + "learning_rate": 1.5029028134533578e-06, + "loss": 0.9652, + "step": 13101 + }, + { + "epoch": 0.9002336127525079, + "grad_norm": 1.0149089097976685, + "learning_rate": 1.5008557805083561e-06, + "loss": 0.892, + "step": 13102 + }, + { + "epoch": 0.9003023223855985, + "grad_norm": 1.0883009433746338, + "learning_rate": 1.498810099454795e-06, + "loss": 0.9316, + "step": 13103 + }, + { + "epoch": 0.9003710320186891, + "grad_norm": 1.0742677450180054, + "learning_rate": 1.496765770410355e-06, + "loss": 0.92, + "step": 13104 + }, + { + "epoch": 0.9004397416517795, + "grad_norm": 1.061564326286316, + "learning_rate": 1.4947227934926533e-06, + "loss": 0.9656, + "step": 13105 + }, + { + "epoch": 0.9005084512848701, + "grad_norm": 1.080665111541748, + "learning_rate": 1.4926811688192183e-06, + "loss": 0.9371, + "step": 13106 + }, + { + "epoch": 0.9005771609179607, + "grad_norm": 1.0942721366882324, + "learning_rate": 1.4906408965075003e-06, + "loss": 0.9574, + "step": 13107 + }, + { + "epoch": 0.9006458705510513, + "grad_norm": 1.0893821716308594, + "learning_rate": 1.4886019766748777e-06, + "loss": 0.9593, + "step": 13108 + }, + { + "epoch": 0.9007145801841419, + "grad_norm": 1.1202878952026367, + "learning_rate": 1.4865644094386483e-06, + "loss": 0.9351, + "step": 13109 + }, + { + "epoch": 0.9007832898172323, + "grad_norm": 1.0668233633041382, + "learning_rate": 1.4845281949160294e-06, + "loss": 0.9186, + "step": 13110 + }, + { + "epoch": 0.9008519994503229, + "grad_norm": 1.0346821546554565, + "learning_rate": 1.4824933332241692e-06, + "loss": 0.939, + "step": 13111 + }, + { + "epoch": 0.9009207090834135, + "grad_norm": 1.044079065322876, + "learning_rate": 1.4804598244801264e-06, + "loss": 0.9082, + "step": 13112 + }, + { + "epoch": 0.9009894187165041, + "grad_norm": 1.08379328250885, + "learning_rate": 1.4784276688008858e-06, + "loss": 0.9415, + "step": 13113 + }, + { + "epoch": 0.9010581283495946, + "grad_norm": 1.0461888313293457, + "learning_rate": 1.4763968663033617e-06, + "loss": 0.9421, + "step": 13114 + }, + { + "epoch": 0.9011268379826852, + "grad_norm": 1.1304214000701904, + "learning_rate": 1.4743674171043831e-06, + "loss": 0.9071, + "step": 13115 + }, + { + "epoch": 0.9011955476157757, + "grad_norm": 1.0400066375732422, + "learning_rate": 1.4723393213207065e-06, + "loss": 0.9348, + "step": 13116 + }, + { + "epoch": 0.9012642572488663, + "grad_norm": 1.0441882610321045, + "learning_rate": 1.4703125790689999e-06, + "loss": 0.9192, + "step": 13117 + }, + { + "epoch": 0.9013329668819569, + "grad_norm": 1.1204404830932617, + "learning_rate": 1.4682871904658612e-06, + "loss": 0.9741, + "step": 13118 + }, + { + "epoch": 0.9014016765150474, + "grad_norm": 1.1090376377105713, + "learning_rate": 1.4662631556278173e-06, + "loss": 0.9339, + "step": 13119 + }, + { + "epoch": 0.901470386148138, + "grad_norm": 1.0625122785568237, + "learning_rate": 1.4642404746713024e-06, + "loss": 0.9314, + "step": 13120 + }, + { + "epoch": 0.9015390957812285, + "grad_norm": 1.1104151010513306, + "learning_rate": 1.4622191477126845e-06, + "loss": 0.9134, + "step": 13121 + }, + { + "epoch": 0.9016078054143191, + "grad_norm": 1.0565872192382812, + "learning_rate": 1.4601991748682458e-06, + "loss": 0.9335, + "step": 13122 + }, + { + "epoch": 0.9016765150474096, + "grad_norm": 1.0222047567367554, + "learning_rate": 1.4581805562541934e-06, + "loss": 0.9217, + "step": 13123 + }, + { + "epoch": 0.9017452246805002, + "grad_norm": 1.0900803804397583, + "learning_rate": 1.4561632919866592e-06, + "loss": 0.9352, + "step": 13124 + }, + { + "epoch": 0.9018139343135908, + "grad_norm": 1.058766484260559, + "learning_rate": 1.4541473821816976e-06, + "loss": 0.9388, + "step": 13125 + }, + { + "epoch": 0.9018826439466813, + "grad_norm": 1.0659523010253906, + "learning_rate": 1.4521328269552826e-06, + "loss": 0.9131, + "step": 13126 + }, + { + "epoch": 0.9019513535797719, + "grad_norm": 1.0544633865356445, + "learning_rate": 1.4501196264233019e-06, + "loss": 0.9172, + "step": 13127 + }, + { + "epoch": 0.9020200632128624, + "grad_norm": 1.0370843410491943, + "learning_rate": 1.4481077807015826e-06, + "loss": 0.9189, + "step": 13128 + }, + { + "epoch": 0.902088772845953, + "grad_norm": 1.0081560611724854, + "learning_rate": 1.4460972899058622e-06, + "loss": 0.9399, + "step": 13129 + }, + { + "epoch": 0.9021574824790436, + "grad_norm": 1.0318177938461304, + "learning_rate": 1.4440881541518043e-06, + "loss": 0.907, + "step": 13130 + }, + { + "epoch": 0.9022261921121342, + "grad_norm": 1.0098381042480469, + "learning_rate": 1.4420803735549938e-06, + "loss": 0.9646, + "step": 13131 + }, + { + "epoch": 0.9022949017452246, + "grad_norm": 1.0629576444625854, + "learning_rate": 1.4400739482309329e-06, + "loss": 0.9154, + "step": 13132 + }, + { + "epoch": 0.9023636113783152, + "grad_norm": 1.0264767408370972, + "learning_rate": 1.4380688782950512e-06, + "loss": 0.9315, + "step": 13133 + }, + { + "epoch": 0.9024323210114058, + "grad_norm": 1.0589641332626343, + "learning_rate": 1.4360651638627014e-06, + "loss": 0.9663, + "step": 13134 + }, + { + "epoch": 0.9025010306444964, + "grad_norm": 1.0408155918121338, + "learning_rate": 1.434062805049155e-06, + "loss": 0.918, + "step": 13135 + }, + { + "epoch": 0.902569740277587, + "grad_norm": 1.0446544885635376, + "learning_rate": 1.4320618019696114e-06, + "loss": 0.9, + "step": 13136 + }, + { + "epoch": 0.9026384499106774, + "grad_norm": 1.0927588939666748, + "learning_rate": 1.4300621547391758e-06, + "loss": 0.9476, + "step": 13137 + }, + { + "epoch": 0.902707159543768, + "grad_norm": 1.0284312963485718, + "learning_rate": 1.428063863472895e-06, + "loss": 0.9471, + "step": 13138 + }, + { + "epoch": 0.9027758691768586, + "grad_norm": 1.0374577045440674, + "learning_rate": 1.426066928285727e-06, + "loss": 0.9313, + "step": 13139 + }, + { + "epoch": 0.9028445788099492, + "grad_norm": 1.1011390686035156, + "learning_rate": 1.424071349292555e-06, + "loss": 0.9545, + "step": 13140 + }, + { + "epoch": 0.9029132884430398, + "grad_norm": 1.0860228538513184, + "learning_rate": 1.4220771266081872e-06, + "loss": 0.9466, + "step": 13141 + }, + { + "epoch": 0.9029819980761302, + "grad_norm": 1.0040956735610962, + "learning_rate": 1.420084260347343e-06, + "loss": 0.9058, + "step": 13142 + }, + { + "epoch": 0.9030507077092208, + "grad_norm": 1.055548071861267, + "learning_rate": 1.418092750624675e-06, + "loss": 0.9818, + "step": 13143 + }, + { + "epoch": 0.9031194173423114, + "grad_norm": 0.9999873042106628, + "learning_rate": 1.41610259755475e-06, + "loss": 0.9039, + "step": 13144 + }, + { + "epoch": 0.903188126975402, + "grad_norm": 1.0427792072296143, + "learning_rate": 1.4141138012520649e-06, + "loss": 0.8807, + "step": 13145 + }, + { + "epoch": 0.9032568366084925, + "grad_norm": 0.9750209450721741, + "learning_rate": 1.4121263618310315e-06, + "loss": 0.8939, + "step": 13146 + }, + { + "epoch": 0.9033255462415831, + "grad_norm": 1.054241418838501, + "learning_rate": 1.410140279405986e-06, + "loss": 0.9271, + "step": 13147 + }, + { + "epoch": 0.9033942558746736, + "grad_norm": 1.0622906684875488, + "learning_rate": 1.408155554091184e-06, + "loss": 0.977, + "step": 13148 + }, + { + "epoch": 0.9034629655077642, + "grad_norm": 0.952070951461792, + "learning_rate": 1.4061721860008092e-06, + "loss": 0.8703, + "step": 13149 + }, + { + "epoch": 0.9035316751408548, + "grad_norm": 1.106837272644043, + "learning_rate": 1.4041901752489623e-06, + "loss": 0.9911, + "step": 13150 + }, + { + "epoch": 0.9036003847739453, + "grad_norm": 0.9951278567314148, + "learning_rate": 1.4022095219496684e-06, + "loss": 0.9508, + "step": 13151 + }, + { + "epoch": 0.9036690944070359, + "grad_norm": 1.075540542602539, + "learning_rate": 1.4002302262168753e-06, + "loss": 0.9144, + "step": 13152 + }, + { + "epoch": 0.9037378040401264, + "grad_norm": 1.1081860065460205, + "learning_rate": 1.398252288164445e-06, + "loss": 0.9143, + "step": 13153 + }, + { + "epoch": 0.903806513673217, + "grad_norm": 1.049278974533081, + "learning_rate": 1.3962757079061696e-06, + "loss": 0.9442, + "step": 13154 + }, + { + "epoch": 0.9038752233063075, + "grad_norm": 1.015810489654541, + "learning_rate": 1.3943004855557607e-06, + "loss": 0.9321, + "step": 13155 + }, + { + "epoch": 0.9039439329393981, + "grad_norm": 1.006861686706543, + "learning_rate": 1.39232662122685e-06, + "loss": 0.9088, + "step": 13156 + }, + { + "epoch": 0.9040126425724887, + "grad_norm": 1.0637297630310059, + "learning_rate": 1.390354115032999e-06, + "loss": 0.9236, + "step": 13157 + }, + { + "epoch": 0.9040813522055792, + "grad_norm": 1.1253873109817505, + "learning_rate": 1.3883829670876758e-06, + "loss": 0.9387, + "step": 13158 + }, + { + "epoch": 0.9041500618386697, + "grad_norm": 1.0762877464294434, + "learning_rate": 1.3864131775042866e-06, + "loss": 0.8924, + "step": 13159 + }, + { + "epoch": 0.9042187714717603, + "grad_norm": 1.0776289701461792, + "learning_rate": 1.3844447463961462e-06, + "loss": 0.9203, + "step": 13160 + }, + { + "epoch": 0.9042874811048509, + "grad_norm": 1.064631700515747, + "learning_rate": 1.3824776738765032e-06, + "loss": 0.9221, + "step": 13161 + }, + { + "epoch": 0.9043561907379415, + "grad_norm": 1.040283203125, + "learning_rate": 1.3805119600585197e-06, + "loss": 0.9577, + "step": 13162 + }, + { + "epoch": 0.9044249003710321, + "grad_norm": 1.1169917583465576, + "learning_rate": 1.3785476050552803e-06, + "loss": 0.9599, + "step": 13163 + }, + { + "epoch": 0.9044936100041225, + "grad_norm": 1.0626599788665771, + "learning_rate": 1.3765846089797945e-06, + "loss": 0.9661, + "step": 13164 + }, + { + "epoch": 0.9045623196372131, + "grad_norm": 1.0909199714660645, + "learning_rate": 1.3746229719449944e-06, + "loss": 0.9376, + "step": 13165 + }, + { + "epoch": 0.9046310292703037, + "grad_norm": 1.056915283203125, + "learning_rate": 1.3726626940637255e-06, + "loss": 0.9172, + "step": 13166 + }, + { + "epoch": 0.9046997389033943, + "grad_norm": 1.0416620969772339, + "learning_rate": 1.3707037754487673e-06, + "loss": 0.9096, + "step": 13167 + }, + { + "epoch": 0.9047684485364849, + "grad_norm": 0.9849027395248413, + "learning_rate": 1.3687462162128156e-06, + "loss": 0.8911, + "step": 13168 + }, + { + "epoch": 0.9048371581695753, + "grad_norm": 1.0457537174224854, + "learning_rate": 1.3667900164684833e-06, + "loss": 0.9414, + "step": 13169 + }, + { + "epoch": 0.9049058678026659, + "grad_norm": 1.0697710514068604, + "learning_rate": 1.364835176328308e-06, + "loss": 0.9009, + "step": 13170 + }, + { + "epoch": 0.9049745774357565, + "grad_norm": 1.0894614458084106, + "learning_rate": 1.362881695904758e-06, + "loss": 0.9549, + "step": 13171 + }, + { + "epoch": 0.9050432870688471, + "grad_norm": 1.0244276523590088, + "learning_rate": 1.360929575310213e-06, + "loss": 0.9259, + "step": 13172 + }, + { + "epoch": 0.9051119967019376, + "grad_norm": 1.0716882944107056, + "learning_rate": 1.3589788146569721e-06, + "loss": 0.9545, + "step": 13173 + }, + { + "epoch": 0.9051807063350281, + "grad_norm": 1.11696195602417, + "learning_rate": 1.3570294140572648e-06, + "loss": 0.9173, + "step": 13174 + }, + { + "epoch": 0.9052494159681187, + "grad_norm": 1.0985066890716553, + "learning_rate": 1.3550813736232403e-06, + "loss": 0.9494, + "step": 13175 + }, + { + "epoch": 0.9053181256012093, + "grad_norm": 1.102666974067688, + "learning_rate": 1.3531346934669703e-06, + "loss": 0.9445, + "step": 13176 + }, + { + "epoch": 0.9053868352342999, + "grad_norm": 1.0470762252807617, + "learning_rate": 1.3511893737004426e-06, + "loss": 0.9318, + "step": 13177 + }, + { + "epoch": 0.9054555448673904, + "grad_norm": 1.0273215770721436, + "learning_rate": 1.3492454144355682e-06, + "loss": 0.9463, + "step": 13178 + }, + { + "epoch": 0.905524254500481, + "grad_norm": 1.0703212022781372, + "learning_rate": 1.347302815784185e-06, + "loss": 0.9389, + "step": 13179 + }, + { + "epoch": 0.9055929641335715, + "grad_norm": 1.111424207687378, + "learning_rate": 1.3453615778580481e-06, + "loss": 0.9735, + "step": 13180 + }, + { + "epoch": 0.9056616737666621, + "grad_norm": 1.0446009635925293, + "learning_rate": 1.3434217007688377e-06, + "loss": 0.9643, + "step": 13181 + }, + { + "epoch": 0.9057303833997526, + "grad_norm": 1.0276378393173218, + "learning_rate": 1.3414831846281562e-06, + "loss": 0.9482, + "step": 13182 + }, + { + "epoch": 0.9057990930328432, + "grad_norm": 1.0343124866485596, + "learning_rate": 1.339546029547517e-06, + "loss": 0.898, + "step": 13183 + }, + { + "epoch": 0.9058678026659338, + "grad_norm": 1.0618209838867188, + "learning_rate": 1.3376102356383703e-06, + "loss": 0.9345, + "step": 13184 + }, + { + "epoch": 0.9059365122990243, + "grad_norm": 1.100093960762024, + "learning_rate": 1.3356758030120763e-06, + "loss": 0.9845, + "step": 13185 + }, + { + "epoch": 0.9060052219321149, + "grad_norm": 1.1428935527801514, + "learning_rate": 1.3337427317799273e-06, + "loss": 0.948, + "step": 13186 + }, + { + "epoch": 0.9060739315652054, + "grad_norm": 1.0604543685913086, + "learning_rate": 1.331811022053131e-06, + "loss": 0.9028, + "step": 13187 + }, + { + "epoch": 0.906142641198296, + "grad_norm": 1.0516622066497803, + "learning_rate": 1.32988067394281e-06, + "loss": 0.9097, + "step": 13188 + }, + { + "epoch": 0.9062113508313866, + "grad_norm": 1.1150602102279663, + "learning_rate": 1.3279516875600252e-06, + "loss": 0.9379, + "step": 13189 + }, + { + "epoch": 0.9062800604644771, + "grad_norm": 0.9862825870513916, + "learning_rate": 1.3260240630157462e-06, + "loss": 0.8938, + "step": 13190 + }, + { + "epoch": 0.9063487700975676, + "grad_norm": 1.0251259803771973, + "learning_rate": 1.3240978004208649e-06, + "loss": 0.8954, + "step": 13191 + }, + { + "epoch": 0.9064174797306582, + "grad_norm": 1.0443291664123535, + "learning_rate": 1.3221728998862066e-06, + "loss": 0.9306, + "step": 13192 + }, + { + "epoch": 0.9064861893637488, + "grad_norm": 1.0726091861724854, + "learning_rate": 1.3202493615224991e-06, + "loss": 0.9059, + "step": 13193 + }, + { + "epoch": 0.9065548989968394, + "grad_norm": 1.0897114276885986, + "learning_rate": 1.31832718544041e-06, + "loss": 0.9126, + "step": 13194 + }, + { + "epoch": 0.90662360862993, + "grad_norm": 1.0913474559783936, + "learning_rate": 1.3164063717505143e-06, + "loss": 0.9109, + "step": 13195 + }, + { + "epoch": 0.9066923182630204, + "grad_norm": 1.0816965103149414, + "learning_rate": 1.3144869205633238e-06, + "loss": 0.9532, + "step": 13196 + }, + { + "epoch": 0.906761027896111, + "grad_norm": 1.0452321767807007, + "learning_rate": 1.312568831989258e-06, + "loss": 0.923, + "step": 13197 + }, + { + "epoch": 0.9068297375292016, + "grad_norm": 1.031075119972229, + "learning_rate": 1.3106521061386628e-06, + "loss": 0.9402, + "step": 13198 + }, + { + "epoch": 0.9068984471622922, + "grad_norm": 1.0255765914916992, + "learning_rate": 1.3087367431218078e-06, + "loss": 0.9572, + "step": 13199 + }, + { + "epoch": 0.9069671567953828, + "grad_norm": 1.0763349533081055, + "learning_rate": 1.3068227430488856e-06, + "loss": 0.92, + "step": 13200 + }, + { + "epoch": 0.9070358664284732, + "grad_norm": 1.0762039422988892, + "learning_rate": 1.3049101060299996e-06, + "loss": 0.9397, + "step": 13201 + }, + { + "epoch": 0.9071045760615638, + "grad_norm": 1.0545554161071777, + "learning_rate": 1.3029988321751923e-06, + "loss": 0.8897, + "step": 13202 + }, + { + "epoch": 0.9071732856946544, + "grad_norm": 1.0975980758666992, + "learning_rate": 1.3010889215944122e-06, + "loss": 0.9304, + "step": 13203 + }, + { + "epoch": 0.907241995327745, + "grad_norm": 1.032854676246643, + "learning_rate": 1.299180374397535e-06, + "loss": 0.9251, + "step": 13204 + }, + { + "epoch": 0.9073107049608355, + "grad_norm": 1.0431095361709595, + "learning_rate": 1.297273190694359e-06, + "loss": 0.9355, + "step": 13205 + }, + { + "epoch": 0.907379414593926, + "grad_norm": 1.08451509475708, + "learning_rate": 1.2953673705946052e-06, + "loss": 0.9216, + "step": 13206 + }, + { + "epoch": 0.9074481242270166, + "grad_norm": 1.0575355291366577, + "learning_rate": 1.293462914207913e-06, + "loss": 0.9234, + "step": 13207 + }, + { + "epoch": 0.9075168338601072, + "grad_norm": 1.0837690830230713, + "learning_rate": 1.2915598216438478e-06, + "loss": 0.9492, + "step": 13208 + }, + { + "epoch": 0.9075855434931978, + "grad_norm": 1.0116931200027466, + "learning_rate": 1.2896580930118862e-06, + "loss": 0.9371, + "step": 13209 + }, + { + "epoch": 0.9076542531262883, + "grad_norm": 1.0981192588806152, + "learning_rate": 1.2877577284214376e-06, + "loss": 0.9419, + "step": 13210 + }, + { + "epoch": 0.9077229627593789, + "grad_norm": 1.0959489345550537, + "learning_rate": 1.2858587279818284e-06, + "loss": 0.9113, + "step": 13211 + }, + { + "epoch": 0.9077916723924694, + "grad_norm": 1.0270147323608398, + "learning_rate": 1.2839610918023104e-06, + "loss": 0.9169, + "step": 13212 + }, + { + "epoch": 0.90786038202556, + "grad_norm": 1.0924394130706787, + "learning_rate": 1.2820648199920487e-06, + "loss": 0.9584, + "step": 13213 + }, + { + "epoch": 0.9079290916586505, + "grad_norm": 1.0809760093688965, + "learning_rate": 1.2801699126601368e-06, + "loss": 0.9596, + "step": 13214 + }, + { + "epoch": 0.9079978012917411, + "grad_norm": 1.0401346683502197, + "learning_rate": 1.2782763699155876e-06, + "loss": 0.9358, + "step": 13215 + }, + { + "epoch": 0.9080665109248317, + "grad_norm": 1.045815348625183, + "learning_rate": 1.276384191867333e-06, + "loss": 0.9413, + "step": 13216 + }, + { + "epoch": 0.9081352205579222, + "grad_norm": 1.01192045211792, + "learning_rate": 1.2744933786242308e-06, + "loss": 0.9295, + "step": 13217 + }, + { + "epoch": 0.9082039301910128, + "grad_norm": 1.1019680500030518, + "learning_rate": 1.2726039302950633e-06, + "loss": 0.9791, + "step": 13218 + }, + { + "epoch": 0.9082726398241033, + "grad_norm": 1.0561392307281494, + "learning_rate": 1.2707158469885182e-06, + "loss": 0.9452, + "step": 13219 + }, + { + "epoch": 0.9083413494571939, + "grad_norm": 1.069106936454773, + "learning_rate": 1.2688291288132232e-06, + "loss": 0.9382, + "step": 13220 + }, + { + "epoch": 0.9084100590902845, + "grad_norm": 1.0023616552352905, + "learning_rate": 1.2669437758777186e-06, + "loss": 0.9079, + "step": 13221 + }, + { + "epoch": 0.908478768723375, + "grad_norm": 1.037609577178955, + "learning_rate": 1.265059788290468e-06, + "loss": 0.932, + "step": 13222 + }, + { + "epoch": 0.9085474783564655, + "grad_norm": 1.1371198892593384, + "learning_rate": 1.2631771661598597e-06, + "loss": 0.949, + "step": 13223 + }, + { + "epoch": 0.9086161879895561, + "grad_norm": 1.0016696453094482, + "learning_rate": 1.26129590959419e-06, + "loss": 0.9394, + "step": 13224 + }, + { + "epoch": 0.9086848976226467, + "grad_norm": 1.0934104919433594, + "learning_rate": 1.259416018701695e-06, + "loss": 0.9426, + "step": 13225 + }, + { + "epoch": 0.9087536072557373, + "grad_norm": 1.0150748491287231, + "learning_rate": 1.2575374935905187e-06, + "loss": 0.8643, + "step": 13226 + }, + { + "epoch": 0.9088223168888279, + "grad_norm": 1.0816636085510254, + "learning_rate": 1.2556603343687357e-06, + "loss": 0.9858, + "step": 13227 + }, + { + "epoch": 0.9088910265219183, + "grad_norm": 1.029200792312622, + "learning_rate": 1.25378454114434e-06, + "loss": 0.9736, + "step": 13228 + }, + { + "epoch": 0.9089597361550089, + "grad_norm": 1.0921365022659302, + "learning_rate": 1.2519101140252343e-06, + "loss": 0.9198, + "step": 13229 + }, + { + "epoch": 0.9090284457880995, + "grad_norm": 1.033756971359253, + "learning_rate": 1.2500370531192628e-06, + "loss": 0.8992, + "step": 13230 + }, + { + "epoch": 0.9090971554211901, + "grad_norm": 1.1093958616256714, + "learning_rate": 1.2481653585341785e-06, + "loss": 0.9372, + "step": 13231 + }, + { + "epoch": 0.9091658650542807, + "grad_norm": 1.0208101272583008, + "learning_rate": 1.246295030377656e-06, + "loss": 0.9376, + "step": 13232 + }, + { + "epoch": 0.9092345746873711, + "grad_norm": 1.006331205368042, + "learning_rate": 1.244426068757304e-06, + "loss": 0.9038, + "step": 13233 + }, + { + "epoch": 0.9093032843204617, + "grad_norm": 1.0432593822479248, + "learning_rate": 1.2425584737806279e-06, + "loss": 0.9046, + "step": 13234 + }, + { + "epoch": 0.9093719939535523, + "grad_norm": 1.050887107849121, + "learning_rate": 1.2406922455550808e-06, + "loss": 0.9369, + "step": 13235 + }, + { + "epoch": 0.9094407035866429, + "grad_norm": 1.0607914924621582, + "learning_rate": 1.2388273841880211e-06, + "loss": 0.9398, + "step": 13236 + }, + { + "epoch": 0.9095094132197334, + "grad_norm": 1.0825823545455933, + "learning_rate": 1.2369638897867381e-06, + "loss": 0.939, + "step": 13237 + }, + { + "epoch": 0.9095781228528239, + "grad_norm": 1.0919196605682373, + "learning_rate": 1.2351017624584266e-06, + "loss": 0.9449, + "step": 13238 + }, + { + "epoch": 0.9096468324859145, + "grad_norm": 1.0229957103729248, + "learning_rate": 1.2332410023102285e-06, + "loss": 0.9108, + "step": 13239 + }, + { + "epoch": 0.9097155421190051, + "grad_norm": 1.0504666566848755, + "learning_rate": 1.231381609449178e-06, + "loss": 0.9432, + "step": 13240 + }, + { + "epoch": 0.9097842517520957, + "grad_norm": 1.081551194190979, + "learning_rate": 1.2295235839822505e-06, + "loss": 0.9429, + "step": 13241 + }, + { + "epoch": 0.9098529613851862, + "grad_norm": 1.0237784385681152, + "learning_rate": 1.2276669260163382e-06, + "loss": 0.8714, + "step": 13242 + }, + { + "epoch": 0.9099216710182768, + "grad_norm": 1.103492259979248, + "learning_rate": 1.2258116356582583e-06, + "loss": 0.9465, + "step": 13243 + }, + { + "epoch": 0.9099903806513673, + "grad_norm": 1.0956878662109375, + "learning_rate": 1.223957713014734e-06, + "loss": 0.9036, + "step": 13244 + }, + { + "epoch": 0.9100590902844579, + "grad_norm": 1.0795232057571411, + "learning_rate": 1.2221051581924243e-06, + "loss": 0.9264, + "step": 13245 + }, + { + "epoch": 0.9101277999175484, + "grad_norm": 1.0654866695404053, + "learning_rate": 1.2202539712979078e-06, + "loss": 0.8731, + "step": 13246 + }, + { + "epoch": 0.910196509550639, + "grad_norm": 1.0905359983444214, + "learning_rate": 1.2184041524376828e-06, + "loss": 0.9214, + "step": 13247 + }, + { + "epoch": 0.9102652191837296, + "grad_norm": 1.0471090078353882, + "learning_rate": 1.2165557017181667e-06, + "loss": 0.939, + "step": 13248 + }, + { + "epoch": 0.9103339288168201, + "grad_norm": 1.0555658340454102, + "learning_rate": 1.214708619245697e-06, + "loss": 0.9114, + "step": 13249 + }, + { + "epoch": 0.9104026384499107, + "grad_norm": 1.1151423454284668, + "learning_rate": 1.2128629051265383e-06, + "loss": 0.9501, + "step": 13250 + }, + { + "epoch": 0.9104713480830012, + "grad_norm": 1.1382288932800293, + "learning_rate": 1.2110185594668755e-06, + "loss": 0.9269, + "step": 13251 + }, + { + "epoch": 0.9105400577160918, + "grad_norm": 1.057716727256775, + "learning_rate": 1.2091755823728069e-06, + "loss": 0.8897, + "step": 13252 + }, + { + "epoch": 0.9106087673491824, + "grad_norm": 1.04112708568573, + "learning_rate": 1.2073339739503641e-06, + "loss": 0.9235, + "step": 13253 + }, + { + "epoch": 0.9106774769822729, + "grad_norm": 1.0523428916931152, + "learning_rate": 1.2054937343054878e-06, + "loss": 0.9499, + "step": 13254 + }, + { + "epoch": 0.9107461866153634, + "grad_norm": 1.017913579940796, + "learning_rate": 1.2036548635440487e-06, + "loss": 0.9525, + "step": 13255 + }, + { + "epoch": 0.910814896248454, + "grad_norm": 1.0411854982376099, + "learning_rate": 1.201817361771837e-06, + "loss": 0.8988, + "step": 13256 + }, + { + "epoch": 0.9108836058815446, + "grad_norm": 1.1226369142532349, + "learning_rate": 1.19998122909456e-06, + "loss": 0.9407, + "step": 13257 + }, + { + "epoch": 0.9109523155146352, + "grad_norm": 1.0222104787826538, + "learning_rate": 1.1981464656178554e-06, + "loss": 0.8916, + "step": 13258 + }, + { + "epoch": 0.9110210251477258, + "grad_norm": 1.1259523630142212, + "learning_rate": 1.1963130714472691e-06, + "loss": 0.9614, + "step": 13259 + }, + { + "epoch": 0.9110897347808162, + "grad_norm": 1.0698715448379517, + "learning_rate": 1.1944810466882782e-06, + "loss": 0.9127, + "step": 13260 + }, + { + "epoch": 0.9111584444139068, + "grad_norm": 1.0639251470565796, + "learning_rate": 1.1926503914462817e-06, + "loss": 0.9532, + "step": 13261 + }, + { + "epoch": 0.9112271540469974, + "grad_norm": 0.9995754957199097, + "learning_rate": 1.1908211058265867e-06, + "loss": 0.8726, + "step": 13262 + }, + { + "epoch": 0.911295863680088, + "grad_norm": 1.0837315320968628, + "learning_rate": 1.188993189934437e-06, + "loss": 0.9223, + "step": 13263 + }, + { + "epoch": 0.9113645733131785, + "grad_norm": 1.0286701917648315, + "learning_rate": 1.187166643874993e-06, + "loss": 0.9163, + "step": 13264 + }, + { + "epoch": 0.911433282946269, + "grad_norm": 0.992902934551239, + "learning_rate": 1.1853414677533315e-06, + "loss": 0.8961, + "step": 13265 + }, + { + "epoch": 0.9115019925793596, + "grad_norm": 1.0259356498718262, + "learning_rate": 1.183517661674452e-06, + "loss": 0.898, + "step": 13266 + }, + { + "epoch": 0.9115707022124502, + "grad_norm": 1.0356911420822144, + "learning_rate": 1.1816952257432818e-06, + "loss": 0.9107, + "step": 13267 + }, + { + "epoch": 0.9116394118455408, + "grad_norm": 1.134995460510254, + "learning_rate": 1.1798741600646619e-06, + "loss": 0.976, + "step": 13268 + }, + { + "epoch": 0.9117081214786313, + "grad_norm": 1.0972273349761963, + "learning_rate": 1.1780544647433612e-06, + "loss": 0.9162, + "step": 13269 + }, + { + "epoch": 0.9117768311117218, + "grad_norm": 1.0702658891677856, + "learning_rate": 1.1762361398840572e-06, + "loss": 0.9772, + "step": 13270 + }, + { + "epoch": 0.9118455407448124, + "grad_norm": 1.0515131950378418, + "learning_rate": 1.1744191855913634e-06, + "loss": 0.8756, + "step": 13271 + }, + { + "epoch": 0.911914250377903, + "grad_norm": 1.0724103450775146, + "learning_rate": 1.1726036019698072e-06, + "loss": 0.9482, + "step": 13272 + }, + { + "epoch": 0.9119829600109935, + "grad_norm": 1.0015432834625244, + "learning_rate": 1.1707893891238358e-06, + "loss": 0.893, + "step": 13273 + }, + { + "epoch": 0.9120516696440841, + "grad_norm": 1.118153691291809, + "learning_rate": 1.1689765471578267e-06, + "loss": 0.9497, + "step": 13274 + }, + { + "epoch": 0.9121203792771747, + "grad_norm": 1.0299720764160156, + "learning_rate": 1.1671650761760604e-06, + "loss": 0.9119, + "step": 13275 + }, + { + "epoch": 0.9121890889102652, + "grad_norm": 1.1027858257293701, + "learning_rate": 1.1653549762827593e-06, + "loss": 0.9448, + "step": 13276 + }, + { + "epoch": 0.9122577985433558, + "grad_norm": 1.0253269672393799, + "learning_rate": 1.1635462475820509e-06, + "loss": 0.9053, + "step": 13277 + }, + { + "epoch": 0.9123265081764463, + "grad_norm": 1.0687395334243774, + "learning_rate": 1.1617388901779935e-06, + "loss": 0.9237, + "step": 13278 + }, + { + "epoch": 0.9123952178095369, + "grad_norm": 1.1262568235397339, + "learning_rate": 1.1599329041745655e-06, + "loss": 0.9395, + "step": 13279 + }, + { + "epoch": 0.9124639274426275, + "grad_norm": 1.144048810005188, + "learning_rate": 1.1581282896756585e-06, + "loss": 0.9509, + "step": 13280 + }, + { + "epoch": 0.912532637075718, + "grad_norm": 1.1207120418548584, + "learning_rate": 1.156325046785095e-06, + "loss": 0.9562, + "step": 13281 + }, + { + "epoch": 0.9126013467088085, + "grad_norm": 1.069180965423584, + "learning_rate": 1.1545231756066116e-06, + "loss": 0.9318, + "step": 13282 + }, + { + "epoch": 0.9126700563418991, + "grad_norm": 1.0765842199325562, + "learning_rate": 1.1527226762438753e-06, + "loss": 0.9517, + "step": 13283 + }, + { + "epoch": 0.9127387659749897, + "grad_norm": 1.0824204683303833, + "learning_rate": 1.1509235488004588e-06, + "loss": 0.9248, + "step": 13284 + }, + { + "epoch": 0.9128074756080803, + "grad_norm": 1.03066086769104, + "learning_rate": 1.1491257933798682e-06, + "loss": 0.9137, + "step": 13285 + }, + { + "epoch": 0.9128761852411708, + "grad_norm": 1.0487545728683472, + "learning_rate": 1.1473294100855319e-06, + "loss": 0.9467, + "step": 13286 + }, + { + "epoch": 0.9129448948742613, + "grad_norm": 1.1340062618255615, + "learning_rate": 1.1455343990207896e-06, + "loss": 0.972, + "step": 13287 + }, + { + "epoch": 0.9130136045073519, + "grad_norm": 1.018186092376709, + "learning_rate": 1.1437407602889056e-06, + "loss": 0.9167, + "step": 13288 + }, + { + "epoch": 0.9130823141404425, + "grad_norm": 1.1134487390518188, + "learning_rate": 1.1419484939930753e-06, + "loss": 0.9481, + "step": 13289 + }, + { + "epoch": 0.9131510237735331, + "grad_norm": 1.0667506456375122, + "learning_rate": 1.1401576002363968e-06, + "loss": 0.9392, + "step": 13290 + }, + { + "epoch": 0.9132197334066237, + "grad_norm": 1.090200662612915, + "learning_rate": 1.1383680791219015e-06, + "loss": 0.9476, + "step": 13291 + }, + { + "epoch": 0.9132884430397141, + "grad_norm": 1.1008021831512451, + "learning_rate": 1.1365799307525431e-06, + "loss": 0.9344, + "step": 13292 + }, + { + "epoch": 0.9133571526728047, + "grad_norm": 1.0756031274795532, + "learning_rate": 1.1347931552311925e-06, + "loss": 0.9487, + "step": 13293 + }, + { + "epoch": 0.9134258623058953, + "grad_norm": 1.0843884944915771, + "learning_rate": 1.133007752660642e-06, + "loss": 0.9527, + "step": 13294 + }, + { + "epoch": 0.9134945719389859, + "grad_norm": 1.0243231058120728, + "learning_rate": 1.1312237231435984e-06, + "loss": 0.8968, + "step": 13295 + }, + { + "epoch": 0.9135632815720764, + "grad_norm": 1.0774672031402588, + "learning_rate": 1.129441066782702e-06, + "loss": 0.9563, + "step": 13296 + }, + { + "epoch": 0.9136319912051669, + "grad_norm": 1.036152958869934, + "learning_rate": 1.127659783680507e-06, + "loss": 0.943, + "step": 13297 + }, + { + "epoch": 0.9137007008382575, + "grad_norm": 1.0121266841888428, + "learning_rate": 1.1258798739394866e-06, + "loss": 0.9114, + "step": 13298 + }, + { + "epoch": 0.9137694104713481, + "grad_norm": 1.0375484228134155, + "learning_rate": 1.1241013376620452e-06, + "loss": 0.9433, + "step": 13299 + }, + { + "epoch": 0.9138381201044387, + "grad_norm": 1.0259768962860107, + "learning_rate": 1.1223241749504926e-06, + "loss": 0.9191, + "step": 13300 + }, + { + "epoch": 0.9139068297375292, + "grad_norm": 1.0513620376586914, + "learning_rate": 1.120548385907072e-06, + "loss": 0.9634, + "step": 13301 + }, + { + "epoch": 0.9139755393706198, + "grad_norm": 1.0329508781433105, + "learning_rate": 1.1187739706339406e-06, + "loss": 0.9048, + "step": 13302 + }, + { + "epoch": 0.9140442490037103, + "grad_norm": 1.0487453937530518, + "learning_rate": 1.1170009292331835e-06, + "loss": 0.934, + "step": 13303 + }, + { + "epoch": 0.9141129586368009, + "grad_norm": 1.051682949066162, + "learning_rate": 1.1152292618068021e-06, + "loss": 0.9564, + "step": 13304 + }, + { + "epoch": 0.9141816682698914, + "grad_norm": 1.0588258504867554, + "learning_rate": 1.1134589684567153e-06, + "loss": 0.952, + "step": 13305 + }, + { + "epoch": 0.914250377902982, + "grad_norm": 1.030260443687439, + "learning_rate": 1.1116900492847721e-06, + "loss": 0.9513, + "step": 13306 + }, + { + "epoch": 0.9143190875360726, + "grad_norm": 1.0976778268814087, + "learning_rate": 1.1099225043927324e-06, + "loss": 0.9525, + "step": 13307 + }, + { + "epoch": 0.9143877971691631, + "grad_norm": 1.0087741613388062, + "learning_rate": 1.1081563338822877e-06, + "loss": 0.8593, + "step": 13308 + }, + { + "epoch": 0.9144565068022537, + "grad_norm": 1.0908117294311523, + "learning_rate": 1.1063915378550393e-06, + "loss": 0.9263, + "step": 13309 + }, + { + "epoch": 0.9145252164353442, + "grad_norm": 1.0494683980941772, + "learning_rate": 1.1046281164125177e-06, + "loss": 0.9277, + "step": 13310 + }, + { + "epoch": 0.9145939260684348, + "grad_norm": 1.0468997955322266, + "learning_rate": 1.1028660696561694e-06, + "loss": 0.9403, + "step": 13311 + }, + { + "epoch": 0.9146626357015254, + "grad_norm": 1.0814788341522217, + "learning_rate": 1.1011053976873659e-06, + "loss": 0.9525, + "step": 13312 + }, + { + "epoch": 0.9147313453346159, + "grad_norm": 1.0408122539520264, + "learning_rate": 1.0993461006073958e-06, + "loss": 0.947, + "step": 13313 + }, + { + "epoch": 0.9148000549677064, + "grad_norm": 1.0523922443389893, + "learning_rate": 1.0975881785174753e-06, + "loss": 0.9497, + "step": 13314 + }, + { + "epoch": 0.914868764600797, + "grad_norm": 1.0354009866714478, + "learning_rate": 1.095831631518729e-06, + "loss": 0.9306, + "step": 13315 + }, + { + "epoch": 0.9149374742338876, + "grad_norm": 1.0409446954727173, + "learning_rate": 1.094076459712215e-06, + "loss": 0.9182, + "step": 13316 + }, + { + "epoch": 0.9150061838669782, + "grad_norm": 1.0935618877410889, + "learning_rate": 1.0923226631989052e-06, + "loss": 0.9412, + "step": 13317 + }, + { + "epoch": 0.9150748935000688, + "grad_norm": 1.0451828241348267, + "learning_rate": 1.0905702420796942e-06, + "loss": 0.9308, + "step": 13318 + }, + { + "epoch": 0.9151436031331592, + "grad_norm": 1.0804808139801025, + "learning_rate": 1.0888191964554011e-06, + "loss": 0.9278, + "step": 13319 + }, + { + "epoch": 0.9152123127662498, + "grad_norm": 1.1279479265213013, + "learning_rate": 1.0870695264267622e-06, + "loss": 0.9754, + "step": 13320 + }, + { + "epoch": 0.9152810223993404, + "grad_norm": 1.0778236389160156, + "learning_rate": 1.08532123209443e-06, + "loss": 0.9384, + "step": 13321 + }, + { + "epoch": 0.915349732032431, + "grad_norm": 1.0392324924468994, + "learning_rate": 1.0835743135589854e-06, + "loss": 0.8931, + "step": 13322 + }, + { + "epoch": 0.9154184416655216, + "grad_norm": 1.0466008186340332, + "learning_rate": 1.0818287709209257e-06, + "loss": 0.9084, + "step": 13323 + }, + { + "epoch": 0.915487151298612, + "grad_norm": 1.0489375591278076, + "learning_rate": 1.0800846042806761e-06, + "loss": 0.9401, + "step": 13324 + }, + { + "epoch": 0.9155558609317026, + "grad_norm": 1.098530650138855, + "learning_rate": 1.078341813738576e-06, + "loss": 0.9308, + "step": 13325 + }, + { + "epoch": 0.9156245705647932, + "grad_norm": 1.0616509914398193, + "learning_rate": 1.0766003993948836e-06, + "loss": 0.9068, + "step": 13326 + }, + { + "epoch": 0.9156932801978838, + "grad_norm": 1.01930832862854, + "learning_rate": 1.0748603613497833e-06, + "loss": 0.9235, + "step": 13327 + }, + { + "epoch": 0.9157619898309743, + "grad_norm": 1.0244179964065552, + "learning_rate": 1.0731216997033776e-06, + "loss": 0.9375, + "step": 13328 + }, + { + "epoch": 0.9158306994640648, + "grad_norm": 1.077180027961731, + "learning_rate": 1.0713844145556928e-06, + "loss": 0.9652, + "step": 13329 + }, + { + "epoch": 0.9158994090971554, + "grad_norm": 1.0740606784820557, + "learning_rate": 1.0696485060066764e-06, + "loss": 0.9016, + "step": 13330 + }, + { + "epoch": 0.915968118730246, + "grad_norm": 1.0109751224517822, + "learning_rate": 1.0679139741561845e-06, + "loss": 0.8818, + "step": 13331 + }, + { + "epoch": 0.9160368283633366, + "grad_norm": 1.0530308485031128, + "learning_rate": 1.066180819104015e-06, + "loss": 0.9204, + "step": 13332 + }, + { + "epoch": 0.9161055379964271, + "grad_norm": 1.0287288427352905, + "learning_rate": 1.0644490409498637e-06, + "loss": 0.915, + "step": 13333 + }, + { + "epoch": 0.9161742476295177, + "grad_norm": 1.0415596961975098, + "learning_rate": 1.062718639793367e-06, + "loss": 0.9559, + "step": 13334 + }, + { + "epoch": 0.9162429572626082, + "grad_norm": 1.006760597229004, + "learning_rate": 1.060989615734076e-06, + "loss": 0.9064, + "step": 13335 + }, + { + "epoch": 0.9163116668956988, + "grad_norm": 1.057961106300354, + "learning_rate": 1.05926196887145e-06, + "loss": 0.9023, + "step": 13336 + }, + { + "epoch": 0.9163803765287893, + "grad_norm": 1.0507705211639404, + "learning_rate": 1.0575356993048874e-06, + "loss": 0.9483, + "step": 13337 + }, + { + "epoch": 0.9164490861618799, + "grad_norm": 1.102627158164978, + "learning_rate": 1.0558108071336947e-06, + "loss": 0.9636, + "step": 13338 + }, + { + "epoch": 0.9165177957949705, + "grad_norm": 1.0697261095046997, + "learning_rate": 1.0540872924571094e-06, + "loss": 0.938, + "step": 13339 + }, + { + "epoch": 0.916586505428061, + "grad_norm": 1.080649971961975, + "learning_rate": 1.0523651553742852e-06, + "loss": 0.8941, + "step": 13340 + }, + { + "epoch": 0.9166552150611516, + "grad_norm": 1.0475658178329468, + "learning_rate": 1.0506443959842876e-06, + "loss": 0.9322, + "step": 13341 + }, + { + "epoch": 0.9167239246942421, + "grad_norm": 1.0425633192062378, + "learning_rate": 1.048925014386115e-06, + "loss": 0.9069, + "step": 13342 + }, + { + "epoch": 0.9167926343273327, + "grad_norm": 1.0546197891235352, + "learning_rate": 1.047207010678683e-06, + "loss": 0.916, + "step": 13343 + }, + { + "epoch": 0.9168613439604233, + "grad_norm": 1.0934337377548218, + "learning_rate": 1.0454903849608261e-06, + "loss": 0.9804, + "step": 13344 + }, + { + "epoch": 0.9169300535935138, + "grad_norm": 1.0640398263931274, + "learning_rate": 1.0437751373313042e-06, + "loss": 0.9238, + "step": 13345 + }, + { + "epoch": 0.9169987632266043, + "grad_norm": 1.0146859884262085, + "learning_rate": 1.0420612678887914e-06, + "loss": 0.9394, + "step": 13346 + }, + { + "epoch": 0.9170674728596949, + "grad_norm": 1.0594887733459473, + "learning_rate": 1.0403487767318836e-06, + "loss": 0.9089, + "step": 13347 + }, + { + "epoch": 0.9171361824927855, + "grad_norm": 1.0505613088607788, + "learning_rate": 1.0386376639591022e-06, + "loss": 0.9295, + "step": 13348 + }, + { + "epoch": 0.9172048921258761, + "grad_norm": 1.1024792194366455, + "learning_rate": 1.036927929668885e-06, + "loss": 0.9621, + "step": 13349 + }, + { + "epoch": 0.9172736017589667, + "grad_norm": 1.019732117652893, + "learning_rate": 1.0352195739595977e-06, + "loss": 0.9177, + "step": 13350 + }, + { + "epoch": 0.9173423113920571, + "grad_norm": 1.03745436668396, + "learning_rate": 1.0335125969295118e-06, + "loss": 0.8863, + "step": 13351 + }, + { + "epoch": 0.9174110210251477, + "grad_norm": 0.9836845993995667, + "learning_rate": 1.0318069986768348e-06, + "loss": 0.871, + "step": 13352 + }, + { + "epoch": 0.9174797306582383, + "grad_norm": 1.0455207824707031, + "learning_rate": 1.0301027792996852e-06, + "loss": 0.9416, + "step": 13353 + }, + { + "epoch": 0.9175484402913289, + "grad_norm": 1.0727331638336182, + "learning_rate": 1.0283999388961102e-06, + "loss": 0.9311, + "step": 13354 + }, + { + "epoch": 0.9176171499244195, + "grad_norm": 1.0378307104110718, + "learning_rate": 1.0266984775640699e-06, + "loss": 0.9186, + "step": 13355 + }, + { + "epoch": 0.9176858595575099, + "grad_norm": 1.0824220180511475, + "learning_rate": 1.024998395401447e-06, + "loss": 0.9162, + "step": 13356 + }, + { + "epoch": 0.9177545691906005, + "grad_norm": 1.1192888021469116, + "learning_rate": 1.0232996925060494e-06, + "loss": 0.9236, + "step": 13357 + }, + { + "epoch": 0.9178232788236911, + "grad_norm": 1.1603753566741943, + "learning_rate": 1.0216023689756022e-06, + "loss": 0.958, + "step": 13358 + }, + { + "epoch": 0.9178919884567817, + "grad_norm": 1.0504628419876099, + "learning_rate": 1.0199064249077461e-06, + "loss": 0.9143, + "step": 13359 + }, + { + "epoch": 0.9179606980898722, + "grad_norm": 1.049589991569519, + "learning_rate": 1.018211860400059e-06, + "loss": 0.9249, + "step": 13360 + }, + { + "epoch": 0.9180294077229627, + "grad_norm": 1.03108549118042, + "learning_rate": 1.0165186755500156e-06, + "loss": 0.9443, + "step": 13361 + }, + { + "epoch": 0.9180981173560533, + "grad_norm": 1.0624710321426392, + "learning_rate": 1.0148268704550296e-06, + "loss": 0.9174, + "step": 13362 + }, + { + "epoch": 0.9181668269891439, + "grad_norm": 1.0966057777404785, + "learning_rate": 1.0131364452124287e-06, + "loss": 0.9462, + "step": 13363 + }, + { + "epoch": 0.9182355366222344, + "grad_norm": 1.1255208253860474, + "learning_rate": 1.011447399919463e-06, + "loss": 0.9674, + "step": 13364 + }, + { + "epoch": 0.918304246255325, + "grad_norm": 1.0522114038467407, + "learning_rate": 1.0097597346733045e-06, + "loss": 0.9285, + "step": 13365 + }, + { + "epoch": 0.9183729558884156, + "grad_norm": 1.0643481016159058, + "learning_rate": 1.008073449571037e-06, + "loss": 0.9279, + "step": 13366 + }, + { + "epoch": 0.9184416655215061, + "grad_norm": 1.054445743560791, + "learning_rate": 1.0063885447096772e-06, + "loss": 0.9495, + "step": 13367 + }, + { + "epoch": 0.9185103751545967, + "grad_norm": 1.033866047859192, + "learning_rate": 1.0047050201861557e-06, + "loss": 0.9259, + "step": 13368 + }, + { + "epoch": 0.9185790847876872, + "grad_norm": 0.9596525430679321, + "learning_rate": 1.0030228760973204e-06, + "loss": 0.8582, + "step": 13369 + }, + { + "epoch": 0.9186477944207778, + "grad_norm": 1.0656741857528687, + "learning_rate": 1.0013421125399519e-06, + "loss": 0.9098, + "step": 13370 + }, + { + "epoch": 0.9187165040538684, + "grad_norm": 1.0948814153671265, + "learning_rate": 9.99662729610737e-07, + "loss": 0.911, + "step": 13371 + }, + { + "epoch": 0.9187852136869589, + "grad_norm": 1.131231665611267, + "learning_rate": 9.979847274062897e-07, + "loss": 0.9566, + "step": 13372 + }, + { + "epoch": 0.9188539233200494, + "grad_norm": 1.0932725667953491, + "learning_rate": 9.963081060231472e-07, + "loss": 0.9183, + "step": 13373 + }, + { + "epoch": 0.91892263295314, + "grad_norm": 0.9897722005844116, + "learning_rate": 9.946328655577624e-07, + "loss": 0.8736, + "step": 13374 + }, + { + "epoch": 0.9189913425862306, + "grad_norm": 1.0779529809951782, + "learning_rate": 9.929590061065114e-07, + "loss": 0.9158, + "step": 13375 + }, + { + "epoch": 0.9190600522193212, + "grad_norm": 1.0445200204849243, + "learning_rate": 9.912865277656973e-07, + "loss": 0.954, + "step": 13376 + }, + { + "epoch": 0.9191287618524117, + "grad_norm": 1.0290261507034302, + "learning_rate": 9.896154306315241e-07, + "loss": 0.9378, + "step": 13377 + }, + { + "epoch": 0.9191974714855022, + "grad_norm": 1.043291449546814, + "learning_rate": 9.879457148001369e-07, + "loss": 0.9239, + "step": 13378 + }, + { + "epoch": 0.9192661811185928, + "grad_norm": 1.0611213445663452, + "learning_rate": 9.86277380367595e-07, + "loss": 0.911, + "step": 13379 + }, + { + "epoch": 0.9193348907516834, + "grad_norm": 1.0962588787078857, + "learning_rate": 9.846104274298689e-07, + "loss": 0.9519, + "step": 13380 + }, + { + "epoch": 0.919403600384774, + "grad_norm": 1.1331348419189453, + "learning_rate": 9.829448560828625e-07, + "loss": 0.9139, + "step": 13381 + }, + { + "epoch": 0.9194723100178646, + "grad_norm": 1.0487440824508667, + "learning_rate": 9.812806664223988e-07, + "loss": 0.9426, + "step": 13382 + }, + { + "epoch": 0.919541019650955, + "grad_norm": 0.9993253946304321, + "learning_rate": 9.796178585442073e-07, + "loss": 0.9119, + "step": 13383 + }, + { + "epoch": 0.9196097292840456, + "grad_norm": 1.1200515031814575, + "learning_rate": 9.779564325439583e-07, + "loss": 0.9144, + "step": 13384 + }, + { + "epoch": 0.9196784389171362, + "grad_norm": 1.0458194017410278, + "learning_rate": 9.762963885172255e-07, + "loss": 0.8937, + "step": 13385 + }, + { + "epoch": 0.9197471485502268, + "grad_norm": 1.0735198259353638, + "learning_rate": 9.746377265595186e-07, + "loss": 0.9369, + "step": 13386 + }, + { + "epoch": 0.9198158581833173, + "grad_norm": 1.0511552095413208, + "learning_rate": 9.729804467662502e-07, + "loss": 0.9331, + "step": 13387 + }, + { + "epoch": 0.9198845678164078, + "grad_norm": 1.1136349439620972, + "learning_rate": 9.71324549232766e-07, + "loss": 0.9677, + "step": 13388 + }, + { + "epoch": 0.9199532774494984, + "grad_norm": 1.0854700803756714, + "learning_rate": 9.696700340543318e-07, + "loss": 0.9247, + "step": 13389 + }, + { + "epoch": 0.920021987082589, + "grad_norm": 1.1365166902542114, + "learning_rate": 9.68016901326127e-07, + "loss": 0.9295, + "step": 13390 + }, + { + "epoch": 0.9200906967156796, + "grad_norm": 1.045328974723816, + "learning_rate": 9.663651511432586e-07, + "loss": 0.9512, + "step": 13391 + }, + { + "epoch": 0.9201594063487701, + "grad_norm": 1.0379445552825928, + "learning_rate": 9.647147836007482e-07, + "loss": 0.919, + "step": 13392 + }, + { + "epoch": 0.9202281159818606, + "grad_norm": 1.054007887840271, + "learning_rate": 9.630657987935388e-07, + "loss": 0.9318, + "step": 13393 + }, + { + "epoch": 0.9202968256149512, + "grad_norm": 1.1003557443618774, + "learning_rate": 9.614181968164993e-07, + "loss": 0.9293, + "step": 13394 + }, + { + "epoch": 0.9203655352480418, + "grad_norm": 1.1252599954605103, + "learning_rate": 9.59771977764412e-07, + "loss": 0.9204, + "step": 13395 + }, + { + "epoch": 0.9204342448811323, + "grad_norm": 1.0705373287200928, + "learning_rate": 9.581271417319903e-07, + "loss": 0.9534, + "step": 13396 + }, + { + "epoch": 0.9205029545142229, + "grad_norm": 1.0662646293640137, + "learning_rate": 9.5648368881385e-07, + "loss": 0.9288, + "step": 13397 + }, + { + "epoch": 0.9205716641473135, + "grad_norm": 1.0654219388961792, + "learning_rate": 9.548416191045461e-07, + "loss": 0.9421, + "step": 13398 + }, + { + "epoch": 0.920640373780404, + "grad_norm": 1.087607741355896, + "learning_rate": 9.532009326985392e-07, + "loss": 0.9284, + "step": 13399 + }, + { + "epoch": 0.9207090834134946, + "grad_norm": 1.0946024656295776, + "learning_rate": 9.51561629690223e-07, + "loss": 0.933, + "step": 13400 + }, + { + "epoch": 0.9207777930465851, + "grad_norm": 1.0126949548721313, + "learning_rate": 9.499237101739055e-07, + "loss": 0.9366, + "step": 13401 + }, + { + "epoch": 0.9208465026796757, + "grad_norm": 1.1093968152999878, + "learning_rate": 9.482871742438087e-07, + "loss": 0.9758, + "step": 13402 + }, + { + "epoch": 0.9209152123127663, + "grad_norm": 1.1513224840164185, + "learning_rate": 9.46652021994085e-07, + "loss": 0.9663, + "step": 13403 + }, + { + "epoch": 0.9209839219458568, + "grad_norm": 1.1423710584640503, + "learning_rate": 9.45018253518809e-07, + "loss": 0.9708, + "step": 13404 + }, + { + "epoch": 0.9210526315789473, + "grad_norm": 1.0701472759246826, + "learning_rate": 9.43385868911964e-07, + "loss": 0.9371, + "step": 13405 + }, + { + "epoch": 0.9211213412120379, + "grad_norm": 1.094874382019043, + "learning_rate": 9.417548682674609e-07, + "loss": 0.9657, + "step": 13406 + }, + { + "epoch": 0.9211900508451285, + "grad_norm": 1.0614808797836304, + "learning_rate": 9.401252516791304e-07, + "loss": 0.9322, + "step": 13407 + }, + { + "epoch": 0.9212587604782191, + "grad_norm": 1.071982741355896, + "learning_rate": 9.384970192407249e-07, + "loss": 0.9131, + "step": 13408 + }, + { + "epoch": 0.9213274701113096, + "grad_norm": 1.0739896297454834, + "learning_rate": 9.368701710459144e-07, + "loss": 0.9416, + "step": 13409 + }, + { + "epoch": 0.9213961797444001, + "grad_norm": 1.0991158485412598, + "learning_rate": 9.352447071882903e-07, + "loss": 0.9414, + "step": 13410 + }, + { + "epoch": 0.9214648893774907, + "grad_norm": 1.0890687704086304, + "learning_rate": 9.336206277613668e-07, + "loss": 0.9884, + "step": 13411 + }, + { + "epoch": 0.9215335990105813, + "grad_norm": 1.0438323020935059, + "learning_rate": 9.319979328585749e-07, + "loss": 0.9029, + "step": 13412 + }, + { + "epoch": 0.9216023086436719, + "grad_norm": 1.0896506309509277, + "learning_rate": 9.303766225732646e-07, + "loss": 0.9657, + "step": 13413 + }, + { + "epoch": 0.9216710182767625, + "grad_norm": 1.0592777729034424, + "learning_rate": 9.287566969987116e-07, + "loss": 0.9607, + "step": 13414 + }, + { + "epoch": 0.9217397279098529, + "grad_norm": 1.0541777610778809, + "learning_rate": 9.271381562281079e-07, + "loss": 0.9092, + "step": 13415 + }, + { + "epoch": 0.9218084375429435, + "grad_norm": 1.0321067571640015, + "learning_rate": 9.255210003545733e-07, + "loss": 0.941, + "step": 13416 + }, + { + "epoch": 0.9218771471760341, + "grad_norm": 1.0869933366775513, + "learning_rate": 9.239052294711309e-07, + "loss": 0.9147, + "step": 13417 + }, + { + "epoch": 0.9219458568091247, + "grad_norm": 1.036386251449585, + "learning_rate": 9.222908436707423e-07, + "loss": 0.9139, + "step": 13418 + }, + { + "epoch": 0.9220145664422152, + "grad_norm": 1.1048283576965332, + "learning_rate": 9.206778430462803e-07, + "loss": 0.9202, + "step": 13419 + }, + { + "epoch": 0.9220832760753057, + "grad_norm": 1.001107096672058, + "learning_rate": 9.190662276905404e-07, + "loss": 0.8993, + "step": 13420 + }, + { + "epoch": 0.9221519857083963, + "grad_norm": 1.0860927104949951, + "learning_rate": 9.1745599769624e-07, + "loss": 0.9343, + "step": 13421 + }, + { + "epoch": 0.9222206953414869, + "grad_norm": 1.0805212259292603, + "learning_rate": 9.158471531560075e-07, + "loss": 0.9586, + "step": 13422 + }, + { + "epoch": 0.9222894049745775, + "grad_norm": 1.0450032949447632, + "learning_rate": 9.142396941624054e-07, + "loss": 0.935, + "step": 13423 + }, + { + "epoch": 0.922358114607668, + "grad_norm": 1.0334240198135376, + "learning_rate": 9.126336208079067e-07, + "loss": 0.9161, + "step": 13424 + }, + { + "epoch": 0.9224268242407585, + "grad_norm": 1.0802313089370728, + "learning_rate": 9.110289331849098e-07, + "loss": 0.9366, + "step": 13425 + }, + { + "epoch": 0.9224955338738491, + "grad_norm": 1.0363805294036865, + "learning_rate": 9.094256313857325e-07, + "loss": 0.9018, + "step": 13426 + }, + { + "epoch": 0.9225642435069397, + "grad_norm": 1.0482131242752075, + "learning_rate": 9.078237155026093e-07, + "loss": 0.9589, + "step": 13427 + }, + { + "epoch": 0.9226329531400302, + "grad_norm": 1.0416425466537476, + "learning_rate": 9.062231856276971e-07, + "loss": 0.9064, + "step": 13428 + }, + { + "epoch": 0.9227016627731208, + "grad_norm": 1.081917405128479, + "learning_rate": 9.046240418530749e-07, + "loss": 0.9329, + "step": 13429 + }, + { + "epoch": 0.9227703724062114, + "grad_norm": 0.9757556319236755, + "learning_rate": 9.030262842707388e-07, + "loss": 0.9248, + "step": 13430 + }, + { + "epoch": 0.9228390820393019, + "grad_norm": 1.0413049459457397, + "learning_rate": 9.014299129726067e-07, + "loss": 0.9447, + "step": 13431 + }, + { + "epoch": 0.9229077916723925, + "grad_norm": 1.1385949850082397, + "learning_rate": 8.998349280505219e-07, + "loss": 0.9383, + "step": 13432 + }, + { + "epoch": 0.922976501305483, + "grad_norm": 1.075589656829834, + "learning_rate": 8.98241329596236e-07, + "loss": 0.9593, + "step": 13433 + }, + { + "epoch": 0.9230452109385736, + "grad_norm": 1.0589936971664429, + "learning_rate": 8.966491177014286e-07, + "loss": 0.9008, + "step": 13434 + }, + { + "epoch": 0.9231139205716642, + "grad_norm": 1.0093755722045898, + "learning_rate": 8.950582924577039e-07, + "loss": 0.9161, + "step": 13435 + }, + { + "epoch": 0.9231826302047547, + "grad_norm": 1.1391654014587402, + "learning_rate": 8.934688539565778e-07, + "loss": 0.9051, + "step": 13436 + }, + { + "epoch": 0.9232513398378452, + "grad_norm": 1.1476147174835205, + "learning_rate": 8.918808022894909e-07, + "loss": 1.0032, + "step": 13437 + }, + { + "epoch": 0.9233200494709358, + "grad_norm": 1.0406272411346436, + "learning_rate": 8.902941375478008e-07, + "loss": 0.9075, + "step": 13438 + }, + { + "epoch": 0.9233887591040264, + "grad_norm": 1.1373332738876343, + "learning_rate": 8.887088598227872e-07, + "loss": 0.9628, + "step": 13439 + }, + { + "epoch": 0.923457468737117, + "grad_norm": 1.0172535181045532, + "learning_rate": 8.871249692056522e-07, + "loss": 0.9009, + "step": 13440 + }, + { + "epoch": 0.9235261783702075, + "grad_norm": 1.090942144393921, + "learning_rate": 8.855424657875144e-07, + "loss": 0.92, + "step": 13441 + }, + { + "epoch": 0.923594888003298, + "grad_norm": 1.1240710020065308, + "learning_rate": 8.839613496594207e-07, + "loss": 0.9714, + "step": 13442 + }, + { + "epoch": 0.9236635976363886, + "grad_norm": 1.0406395196914673, + "learning_rate": 8.823816209123204e-07, + "loss": 0.9702, + "step": 13443 + }, + { + "epoch": 0.9237323072694792, + "grad_norm": 1.06511390209198, + "learning_rate": 8.808032796371019e-07, + "loss": 0.9283, + "step": 13444 + }, + { + "epoch": 0.9238010169025698, + "grad_norm": 1.0668619871139526, + "learning_rate": 8.79226325924562e-07, + "loss": 0.9203, + "step": 13445 + }, + { + "epoch": 0.9238697265356604, + "grad_norm": 1.0394210815429688, + "learning_rate": 8.776507598654283e-07, + "loss": 0.9205, + "step": 13446 + }, + { + "epoch": 0.9239384361687508, + "grad_norm": 1.0238473415374756, + "learning_rate": 8.760765815503391e-07, + "loss": 0.9211, + "step": 13447 + }, + { + "epoch": 0.9240071458018414, + "grad_norm": 1.030302882194519, + "learning_rate": 8.7450379106985e-07, + "loss": 0.9086, + "step": 13448 + }, + { + "epoch": 0.924075855434932, + "grad_norm": 1.1019237041473389, + "learning_rate": 8.729323885144497e-07, + "loss": 0.9283, + "step": 13449 + }, + { + "epoch": 0.9241445650680226, + "grad_norm": 1.0710238218307495, + "learning_rate": 8.713623739745408e-07, + "loss": 0.864, + "step": 13450 + }, + { + "epoch": 0.9242132747011131, + "grad_norm": 1.084230661392212, + "learning_rate": 8.697937475404399e-07, + "loss": 0.9235, + "step": 13451 + }, + { + "epoch": 0.9242819843342036, + "grad_norm": 1.0371005535125732, + "learning_rate": 8.682265093023916e-07, + "loss": 0.924, + "step": 13452 + }, + { + "epoch": 0.9243506939672942, + "grad_norm": 1.0389925241470337, + "learning_rate": 8.6666065935056e-07, + "loss": 0.9265, + "step": 13453 + }, + { + "epoch": 0.9244194036003848, + "grad_norm": 1.0292857885360718, + "learning_rate": 8.650961977750255e-07, + "loss": 0.9258, + "step": 13454 + }, + { + "epoch": 0.9244881132334754, + "grad_norm": 1.037095308303833, + "learning_rate": 8.635331246657885e-07, + "loss": 0.9247, + "step": 13455 + }, + { + "epoch": 0.9245568228665659, + "grad_norm": 1.0799860954284668, + "learning_rate": 8.619714401127743e-07, + "loss": 0.925, + "step": 13456 + }, + { + "epoch": 0.9246255324996564, + "grad_norm": 1.0340871810913086, + "learning_rate": 8.604111442058304e-07, + "loss": 0.9138, + "step": 13457 + }, + { + "epoch": 0.924694242132747, + "grad_norm": 1.091131567955017, + "learning_rate": 8.588522370347101e-07, + "loss": 0.959, + "step": 13458 + }, + { + "epoch": 0.9247629517658376, + "grad_norm": 1.059396505355835, + "learning_rate": 8.572947186891001e-07, + "loss": 0.9131, + "step": 13459 + }, + { + "epoch": 0.9248316613989281, + "grad_norm": 1.0604584217071533, + "learning_rate": 8.557385892586062e-07, + "loss": 0.9096, + "step": 13460 + }, + { + "epoch": 0.9249003710320187, + "grad_norm": 1.0464292764663696, + "learning_rate": 8.541838488327486e-07, + "loss": 0.8936, + "step": 13461 + }, + { + "epoch": 0.9249690806651093, + "grad_norm": 1.0408570766448975, + "learning_rate": 8.526304975009724e-07, + "loss": 0.9061, + "step": 13462 + }, + { + "epoch": 0.9250377902981998, + "grad_norm": 1.0559344291687012, + "learning_rate": 8.510785353526396e-07, + "loss": 0.9151, + "step": 13463 + }, + { + "epoch": 0.9251064999312903, + "grad_norm": 1.0852017402648926, + "learning_rate": 8.495279624770341e-07, + "loss": 0.9529, + "step": 13464 + }, + { + "epoch": 0.9251752095643809, + "grad_norm": 1.002292513847351, + "learning_rate": 8.479787789633598e-07, + "loss": 0.8877, + "step": 13465 + }, + { + "epoch": 0.9252439191974715, + "grad_norm": 0.9807438850402832, + "learning_rate": 8.464309849007424e-07, + "loss": 0.8799, + "step": 13466 + }, + { + "epoch": 0.9253126288305621, + "grad_norm": 1.0342241525650024, + "learning_rate": 8.448845803782218e-07, + "loss": 0.8782, + "step": 13467 + }, + { + "epoch": 0.9253813384636526, + "grad_norm": 1.1408894062042236, + "learning_rate": 8.433395654847631e-07, + "loss": 0.926, + "step": 13468 + }, + { + "epoch": 0.9254500480967431, + "grad_norm": 1.0765719413757324, + "learning_rate": 8.417959403092479e-07, + "loss": 0.9381, + "step": 13469 + }, + { + "epoch": 0.9255187577298337, + "grad_norm": 1.0463052988052368, + "learning_rate": 8.402537049404857e-07, + "loss": 0.8979, + "step": 13470 + }, + { + "epoch": 0.9255874673629243, + "grad_norm": 1.0355851650238037, + "learning_rate": 8.387128594671944e-07, + "loss": 0.9358, + "step": 13471 + }, + { + "epoch": 0.9256561769960149, + "grad_norm": 1.078111171722412, + "learning_rate": 8.371734039780227e-07, + "loss": 0.9214, + "step": 13472 + }, + { + "epoch": 0.9257248866291053, + "grad_norm": 1.0683692693710327, + "learning_rate": 8.356353385615301e-07, + "loss": 0.9548, + "step": 13473 + }, + { + "epoch": 0.9257935962621959, + "grad_norm": 1.080081820487976, + "learning_rate": 8.340986633062015e-07, + "loss": 0.8921, + "step": 13474 + }, + { + "epoch": 0.9258623058952865, + "grad_norm": 1.0537678003311157, + "learning_rate": 8.325633783004466e-07, + "loss": 0.91, + "step": 13475 + }, + { + "epoch": 0.9259310155283771, + "grad_norm": 1.0713887214660645, + "learning_rate": 8.31029483632581e-07, + "loss": 0.9416, + "step": 13476 + }, + { + "epoch": 0.9259997251614677, + "grad_norm": 1.0546942949295044, + "learning_rate": 8.294969793908564e-07, + "loss": 0.9294, + "step": 13477 + }, + { + "epoch": 0.9260684347945582, + "grad_norm": 1.0583291053771973, + "learning_rate": 8.279658656634298e-07, + "loss": 0.8992, + "step": 13478 + }, + { + "epoch": 0.9261371444276487, + "grad_norm": 1.0635831356048584, + "learning_rate": 8.264361425383893e-07, + "loss": 0.9336, + "step": 13479 + }, + { + "epoch": 0.9262058540607393, + "grad_norm": 1.0958783626556396, + "learning_rate": 8.249078101037366e-07, + "loss": 0.9432, + "step": 13480 + }, + { + "epoch": 0.9262745636938299, + "grad_norm": 1.077742576599121, + "learning_rate": 8.233808684473959e-07, + "loss": 0.9467, + "step": 13481 + }, + { + "epoch": 0.9263432733269205, + "grad_norm": 1.0837818384170532, + "learning_rate": 8.218553176572164e-07, + "loss": 0.9263, + "step": 13482 + }, + { + "epoch": 0.926411982960011, + "grad_norm": 1.0755103826522827, + "learning_rate": 8.203311578209583e-07, + "loss": 0.9362, + "step": 13483 + }, + { + "epoch": 0.9264806925931015, + "grad_norm": 1.034294605255127, + "learning_rate": 8.188083890263043e-07, + "loss": 0.9223, + "step": 13484 + }, + { + "epoch": 0.9265494022261921, + "grad_norm": 1.0391767024993896, + "learning_rate": 8.172870113608594e-07, + "loss": 0.9043, + "step": 13485 + }, + { + "epoch": 0.9266181118592827, + "grad_norm": 1.1160047054290771, + "learning_rate": 8.157670249121452e-07, + "loss": 0.9453, + "step": 13486 + }, + { + "epoch": 0.9266868214923732, + "grad_norm": 1.0401147603988647, + "learning_rate": 8.142484297676112e-07, + "loss": 0.9359, + "step": 13487 + }, + { + "epoch": 0.9267555311254638, + "grad_norm": 1.040026307106018, + "learning_rate": 8.12731226014618e-07, + "loss": 0.9281, + "step": 13488 + }, + { + "epoch": 0.9268242407585543, + "grad_norm": 1.0562951564788818, + "learning_rate": 8.112154137404487e-07, + "loss": 0.9261, + "step": 13489 + }, + { + "epoch": 0.9268929503916449, + "grad_norm": 1.0603728294372559, + "learning_rate": 8.097009930323085e-07, + "loss": 0.9301, + "step": 13490 + }, + { + "epoch": 0.9269616600247355, + "grad_norm": 1.089414119720459, + "learning_rate": 8.081879639773193e-07, + "loss": 0.9644, + "step": 13491 + }, + { + "epoch": 0.927030369657826, + "grad_norm": 1.1151241064071655, + "learning_rate": 8.066763266625282e-07, + "loss": 0.9488, + "step": 13492 + }, + { + "epoch": 0.9270990792909166, + "grad_norm": 1.0361946821212769, + "learning_rate": 8.051660811748962e-07, + "loss": 0.9168, + "step": 13493 + }, + { + "epoch": 0.9271677889240072, + "grad_norm": 1.0663511753082275, + "learning_rate": 8.036572276013066e-07, + "loss": 0.9495, + "step": 13494 + }, + { + "epoch": 0.9272364985570977, + "grad_norm": 0.9998550415039062, + "learning_rate": 8.021497660285621e-07, + "loss": 0.9411, + "step": 13495 + }, + { + "epoch": 0.9273052081901882, + "grad_norm": 1.0298115015029907, + "learning_rate": 8.006436965433877e-07, + "loss": 0.9131, + "step": 13496 + }, + { + "epoch": 0.9273739178232788, + "grad_norm": 1.0749861001968384, + "learning_rate": 7.99139019232431e-07, + "loss": 0.9408, + "step": 13497 + }, + { + "epoch": 0.9274426274563694, + "grad_norm": 1.087315559387207, + "learning_rate": 7.976357341822449e-07, + "loss": 0.9648, + "step": 13498 + }, + { + "epoch": 0.92751133708946, + "grad_norm": 1.1806893348693848, + "learning_rate": 7.96133841479324e-07, + "loss": 0.9662, + "step": 13499 + }, + { + "epoch": 0.9275800467225505, + "grad_norm": 1.0586483478546143, + "learning_rate": 7.946333412100604e-07, + "loss": 0.9184, + "step": 13500 + }, + { + "epoch": 0.927648756355641, + "grad_norm": 1.0261198282241821, + "learning_rate": 7.931342334607794e-07, + "loss": 0.9081, + "step": 13501 + }, + { + "epoch": 0.9277174659887316, + "grad_norm": 1.017918348312378, + "learning_rate": 7.916365183177288e-07, + "loss": 0.9329, + "step": 13502 + }, + { + "epoch": 0.9277861756218222, + "grad_norm": 1.071897029876709, + "learning_rate": 7.901401958670729e-07, + "loss": 0.9818, + "step": 13503 + }, + { + "epoch": 0.9278548852549128, + "grad_norm": 1.0414652824401855, + "learning_rate": 7.886452661948845e-07, + "loss": 0.9231, + "step": 13504 + }, + { + "epoch": 0.9279235948880032, + "grad_norm": 1.0751066207885742, + "learning_rate": 7.871517293871727e-07, + "loss": 0.9395, + "step": 13505 + }, + { + "epoch": 0.9279923045210938, + "grad_norm": 1.0638474225997925, + "learning_rate": 7.856595855298604e-07, + "loss": 0.9474, + "step": 13506 + }, + { + "epoch": 0.9280610141541844, + "grad_norm": 1.0161195993423462, + "learning_rate": 7.841688347087844e-07, + "loss": 0.8848, + "step": 13507 + }, + { + "epoch": 0.928129723787275, + "grad_norm": 1.1475509405136108, + "learning_rate": 7.826794770097151e-07, + "loss": 0.9152, + "step": 13508 + }, + { + "epoch": 0.9281984334203656, + "grad_norm": 1.0539036989212036, + "learning_rate": 7.811915125183256e-07, + "loss": 0.9091, + "step": 13509 + }, + { + "epoch": 0.9282671430534561, + "grad_norm": 1.0262460708618164, + "learning_rate": 7.797049413202223e-07, + "loss": 0.9039, + "step": 13510 + }, + { + "epoch": 0.9283358526865466, + "grad_norm": 1.0032883882522583, + "learning_rate": 7.782197635009259e-07, + "loss": 0.8804, + "step": 13511 + }, + { + "epoch": 0.9284045623196372, + "grad_norm": 1.04150390625, + "learning_rate": 7.767359791458761e-07, + "loss": 0.9146, + "step": 13512 + }, + { + "epoch": 0.9284732719527278, + "grad_norm": 1.0564980506896973, + "learning_rate": 7.752535883404382e-07, + "loss": 0.892, + "step": 13513 + }, + { + "epoch": 0.9285419815858184, + "grad_norm": 1.010793924331665, + "learning_rate": 7.737725911698912e-07, + "loss": 0.9042, + "step": 13514 + }, + { + "epoch": 0.9286106912189089, + "grad_norm": 1.0456937551498413, + "learning_rate": 7.722929877194334e-07, + "loss": 0.9403, + "step": 13515 + }, + { + "epoch": 0.9286794008519994, + "grad_norm": 1.0928415060043335, + "learning_rate": 7.708147780741859e-07, + "loss": 0.9782, + "step": 13516 + }, + { + "epoch": 0.92874811048509, + "grad_norm": 1.136845588684082, + "learning_rate": 7.693379623191943e-07, + "loss": 0.906, + "step": 13517 + }, + { + "epoch": 0.9288168201181806, + "grad_norm": 1.0209888219833374, + "learning_rate": 7.678625405394157e-07, + "loss": 0.8974, + "step": 13518 + }, + { + "epoch": 0.9288855297512711, + "grad_norm": 1.0436577796936035, + "learning_rate": 7.663885128197267e-07, + "loss": 0.9243, + "step": 13519 + }, + { + "epoch": 0.9289542393843617, + "grad_norm": 1.0968174934387207, + "learning_rate": 7.649158792449346e-07, + "loss": 0.9243, + "step": 13520 + }, + { + "epoch": 0.9290229490174522, + "grad_norm": 1.035887598991394, + "learning_rate": 7.634446398997519e-07, + "loss": 0.95, + "step": 13521 + }, + { + "epoch": 0.9290916586505428, + "grad_norm": 1.04221510887146, + "learning_rate": 7.619747948688221e-07, + "loss": 0.899, + "step": 13522 + }, + { + "epoch": 0.9291603682836334, + "grad_norm": 1.0382364988327026, + "learning_rate": 7.605063442367056e-07, + "loss": 0.9166, + "step": 13523 + }, + { + "epoch": 0.9292290779167239, + "grad_norm": 1.0457863807678223, + "learning_rate": 7.590392880878788e-07, + "loss": 0.9335, + "step": 13524 + }, + { + "epoch": 0.9292977875498145, + "grad_norm": 1.1109224557876587, + "learning_rate": 7.575736265067385e-07, + "loss": 0.9135, + "step": 13525 + }, + { + "epoch": 0.9293664971829051, + "grad_norm": 1.0216238498687744, + "learning_rate": 7.561093595776086e-07, + "loss": 0.9031, + "step": 13526 + }, + { + "epoch": 0.9294352068159956, + "grad_norm": 1.075106143951416, + "learning_rate": 7.54646487384722e-07, + "loss": 0.9289, + "step": 13527 + }, + { + "epoch": 0.9295039164490861, + "grad_norm": 1.0563124418258667, + "learning_rate": 7.531850100122445e-07, + "loss": 0.9275, + "step": 13528 + }, + { + "epoch": 0.9295726260821767, + "grad_norm": 1.0548735857009888, + "learning_rate": 7.517249275442478e-07, + "loss": 0.9466, + "step": 13529 + }, + { + "epoch": 0.9296413357152673, + "grad_norm": 1.0865415334701538, + "learning_rate": 7.502662400647286e-07, + "loss": 0.9398, + "step": 13530 + }, + { + "epoch": 0.9297100453483579, + "grad_norm": 1.00994873046875, + "learning_rate": 7.48808947657606e-07, + "loss": 0.9288, + "step": 13531 + }, + { + "epoch": 0.9297787549814484, + "grad_norm": 1.0312985181808472, + "learning_rate": 7.473530504067211e-07, + "loss": 0.919, + "step": 13532 + }, + { + "epoch": 0.9298474646145389, + "grad_norm": 1.0846143960952759, + "learning_rate": 7.458985483958291e-07, + "loss": 0.9568, + "step": 13533 + }, + { + "epoch": 0.9299161742476295, + "grad_norm": 1.159743309020996, + "learning_rate": 7.444454417085994e-07, + "loss": 0.9513, + "step": 13534 + }, + { + "epoch": 0.9299848838807201, + "grad_norm": 1.0803881883621216, + "learning_rate": 7.429937304286371e-07, + "loss": 0.9143, + "step": 13535 + }, + { + "epoch": 0.9300535935138107, + "grad_norm": 1.0589045286178589, + "learning_rate": 7.415434146394563e-07, + "loss": 0.9071, + "step": 13536 + }, + { + "epoch": 0.9301223031469011, + "grad_norm": 0.9969520568847656, + "learning_rate": 7.400944944244898e-07, + "loss": 0.9176, + "step": 13537 + }, + { + "epoch": 0.9301910127799917, + "grad_norm": 1.0318896770477295, + "learning_rate": 7.386469698670934e-07, + "loss": 0.9235, + "step": 13538 + }, + { + "epoch": 0.9302597224130823, + "grad_norm": 1.08161461353302, + "learning_rate": 7.372008410505476e-07, + "loss": 0.9133, + "step": 13539 + }, + { + "epoch": 0.9303284320461729, + "grad_norm": 1.018318772315979, + "learning_rate": 7.357561080580416e-07, + "loss": 0.9469, + "step": 13540 + }, + { + "epoch": 0.9303971416792635, + "grad_norm": 1.08855402469635, + "learning_rate": 7.343127709726921e-07, + "loss": 0.949, + "step": 13541 + }, + { + "epoch": 0.930465851312354, + "grad_norm": 1.0358222723007202, + "learning_rate": 7.328708298775328e-07, + "loss": 0.9101, + "step": 13542 + }, + { + "epoch": 0.9305345609454445, + "grad_norm": 1.073020100593567, + "learning_rate": 7.314302848555193e-07, + "loss": 0.9479, + "step": 13543 + }, + { + "epoch": 0.9306032705785351, + "grad_norm": 1.0366981029510498, + "learning_rate": 7.299911359895273e-07, + "loss": 0.9343, + "step": 13544 + }, + { + "epoch": 0.9306719802116257, + "grad_norm": 1.0400125980377197, + "learning_rate": 7.285533833623431e-07, + "loss": 0.9569, + "step": 13545 + }, + { + "epoch": 0.9307406898447163, + "grad_norm": 1.1061885356903076, + "learning_rate": 7.27117027056684e-07, + "loss": 0.9028, + "step": 13546 + }, + { + "epoch": 0.9308093994778068, + "grad_norm": 1.0192348957061768, + "learning_rate": 7.256820671551839e-07, + "loss": 0.9408, + "step": 13547 + }, + { + "epoch": 0.9308781091108973, + "grad_norm": 1.0822604894638062, + "learning_rate": 7.242485037403906e-07, + "loss": 0.9065, + "step": 13548 + }, + { + "epoch": 0.9309468187439879, + "grad_norm": 1.0820719003677368, + "learning_rate": 7.228163368947827e-07, + "loss": 0.9103, + "step": 13549 + }, + { + "epoch": 0.9310155283770785, + "grad_norm": 1.0731737613677979, + "learning_rate": 7.213855667007469e-07, + "loss": 0.9087, + "step": 13550 + }, + { + "epoch": 0.931084238010169, + "grad_norm": 0.975029706954956, + "learning_rate": 7.199561932405952e-07, + "loss": 0.9008, + "step": 13551 + }, + { + "epoch": 0.9311529476432596, + "grad_norm": 1.0893062353134155, + "learning_rate": 7.18528216596559e-07, + "loss": 0.9422, + "step": 13552 + }, + { + "epoch": 0.9312216572763501, + "grad_norm": 1.070655107498169, + "learning_rate": 7.171016368507893e-07, + "loss": 0.9409, + "step": 13553 + }, + { + "epoch": 0.9312903669094407, + "grad_norm": 1.0564311742782593, + "learning_rate": 7.156764540853594e-07, + "loss": 0.9342, + "step": 13554 + }, + { + "epoch": 0.9313590765425313, + "grad_norm": 1.0267950296401978, + "learning_rate": 7.142526683822537e-07, + "loss": 0.9083, + "step": 13555 + }, + { + "epoch": 0.9314277861756218, + "grad_norm": 1.163763165473938, + "learning_rate": 7.128302798233816e-07, + "loss": 0.9533, + "step": 13556 + }, + { + "epoch": 0.9314964958087124, + "grad_norm": 1.0434000492095947, + "learning_rate": 7.114092884905777e-07, + "loss": 0.9216, + "step": 13557 + }, + { + "epoch": 0.931565205441803, + "grad_norm": 1.072986125946045, + "learning_rate": 7.099896944655876e-07, + "loss": 0.8984, + "step": 13558 + }, + { + "epoch": 0.9316339150748935, + "grad_norm": 1.1017206907272339, + "learning_rate": 7.085714978300823e-07, + "loss": 0.9427, + "step": 13559 + }, + { + "epoch": 0.931702624707984, + "grad_norm": 1.0407555103302002, + "learning_rate": 7.071546986656436e-07, + "loss": 0.8841, + "step": 13560 + }, + { + "epoch": 0.9317713343410746, + "grad_norm": 1.0195177793502808, + "learning_rate": 7.057392970537841e-07, + "loss": 0.945, + "step": 13561 + }, + { + "epoch": 0.9318400439741652, + "grad_norm": 1.0180692672729492, + "learning_rate": 7.043252930759303e-07, + "loss": 0.9322, + "step": 13562 + }, + { + "epoch": 0.9319087536072558, + "grad_norm": 1.0354070663452148, + "learning_rate": 7.029126868134256e-07, + "loss": 0.9207, + "step": 13563 + }, + { + "epoch": 0.9319774632403462, + "grad_norm": 1.0881975889205933, + "learning_rate": 7.015014783475466e-07, + "loss": 0.9711, + "step": 13564 + }, + { + "epoch": 0.9320461728734368, + "grad_norm": 1.0314695835113525, + "learning_rate": 7.000916677594644e-07, + "loss": 0.9457, + "step": 13565 + }, + { + "epoch": 0.9321148825065274, + "grad_norm": 1.0615447759628296, + "learning_rate": 6.986832551302946e-07, + "loss": 0.943, + "step": 13566 + }, + { + "epoch": 0.932183592139618, + "grad_norm": 1.0486446619033813, + "learning_rate": 6.972762405410615e-07, + "loss": 0.9212, + "step": 13567 + }, + { + "epoch": 0.9322523017727086, + "grad_norm": 1.1067222356796265, + "learning_rate": 6.958706240727086e-07, + "loss": 0.9351, + "step": 13568 + }, + { + "epoch": 0.932321011405799, + "grad_norm": 1.025692105293274, + "learning_rate": 6.944664058060962e-07, + "loss": 0.8835, + "step": 13569 + }, + { + "epoch": 0.9323897210388896, + "grad_norm": 1.0570175647735596, + "learning_rate": 6.930635858220125e-07, + "loss": 0.9142, + "step": 13570 + }, + { + "epoch": 0.9324584306719802, + "grad_norm": 1.1181020736694336, + "learning_rate": 6.91662164201165e-07, + "loss": 0.9465, + "step": 13571 + }, + { + "epoch": 0.9325271403050708, + "grad_norm": 1.020859956741333, + "learning_rate": 6.902621410241672e-07, + "loss": 0.9178, + "step": 13572 + }, + { + "epoch": 0.9325958499381614, + "grad_norm": 1.0484806299209595, + "learning_rate": 6.888635163715657e-07, + "loss": 0.9614, + "step": 13573 + }, + { + "epoch": 0.9326645595712519, + "grad_norm": 1.0704858303070068, + "learning_rate": 6.874662903238239e-07, + "loss": 0.8975, + "step": 13574 + }, + { + "epoch": 0.9327332692043424, + "grad_norm": 1.0643565654754639, + "learning_rate": 6.860704629613218e-07, + "loss": 0.9476, + "step": 13575 + }, + { + "epoch": 0.932801978837433, + "grad_norm": 1.085831880569458, + "learning_rate": 6.846760343643593e-07, + "loss": 0.9423, + "step": 13576 + }, + { + "epoch": 0.9328706884705236, + "grad_norm": 1.0563087463378906, + "learning_rate": 6.832830046131611e-07, + "loss": 0.9367, + "step": 13577 + }, + { + "epoch": 0.9329393981036141, + "grad_norm": 1.0891960859298706, + "learning_rate": 6.81891373787863e-07, + "loss": 0.9399, + "step": 13578 + }, + { + "epoch": 0.9330081077367047, + "grad_norm": 1.1470333337783813, + "learning_rate": 6.805011419685287e-07, + "loss": 0.9762, + "step": 13579 + }, + { + "epoch": 0.9330768173697952, + "grad_norm": 1.0523097515106201, + "learning_rate": 6.791123092351331e-07, + "loss": 0.9308, + "step": 13580 + }, + { + "epoch": 0.9331455270028858, + "grad_norm": 1.0826939344406128, + "learning_rate": 6.777248756675792e-07, + "loss": 0.9104, + "step": 13581 + }, + { + "epoch": 0.9332142366359764, + "grad_norm": 1.0513566732406616, + "learning_rate": 6.763388413456806e-07, + "loss": 0.9146, + "step": 13582 + }, + { + "epoch": 0.9332829462690669, + "grad_norm": 1.0907342433929443, + "learning_rate": 6.749542063491793e-07, + "loss": 0.9102, + "step": 13583 + }, + { + "epoch": 0.9333516559021575, + "grad_norm": 1.1251717805862427, + "learning_rate": 6.735709707577337e-07, + "loss": 0.9575, + "step": 13584 + }, + { + "epoch": 0.933420365535248, + "grad_norm": 1.128388524055481, + "learning_rate": 6.721891346509134e-07, + "loss": 0.9639, + "step": 13585 + }, + { + "epoch": 0.9334890751683386, + "grad_norm": 1.0609873533248901, + "learning_rate": 6.70808698108219e-07, + "loss": 0.9444, + "step": 13586 + }, + { + "epoch": 0.9335577848014291, + "grad_norm": 1.061338186264038, + "learning_rate": 6.694296612090645e-07, + "loss": 0.9463, + "step": 13587 + }, + { + "epoch": 0.9336264944345197, + "grad_norm": 1.0592765808105469, + "learning_rate": 6.680520240327892e-07, + "loss": 0.971, + "step": 13588 + }, + { + "epoch": 0.9336952040676103, + "grad_norm": 1.087607979774475, + "learning_rate": 6.666757866586465e-07, + "loss": 0.9285, + "step": 13589 + }, + { + "epoch": 0.9337639137007009, + "grad_norm": 1.014660120010376, + "learning_rate": 6.653009491658063e-07, + "loss": 0.9342, + "step": 13590 + }, + { + "epoch": 0.9338326233337914, + "grad_norm": 1.1245197057724, + "learning_rate": 6.639275116333638e-07, + "loss": 0.9819, + "step": 13591 + }, + { + "epoch": 0.9339013329668819, + "grad_norm": 1.1122597455978394, + "learning_rate": 6.625554741403333e-07, + "loss": 0.9329, + "step": 13592 + }, + { + "epoch": 0.9339700425999725, + "grad_norm": 1.0608807802200317, + "learning_rate": 6.61184836765652e-07, + "loss": 0.9258, + "step": 13593 + }, + { + "epoch": 0.9340387522330631, + "grad_norm": 1.0277231931686401, + "learning_rate": 6.598155995881622e-07, + "loss": 0.9662, + "step": 13594 + }, + { + "epoch": 0.9341074618661537, + "grad_norm": 1.0841158628463745, + "learning_rate": 6.584477626866426e-07, + "loss": 0.9209, + "step": 13595 + }, + { + "epoch": 0.9341761714992441, + "grad_norm": 1.068216323852539, + "learning_rate": 6.570813261397773e-07, + "loss": 0.9446, + "step": 13596 + }, + { + "epoch": 0.9342448811323347, + "grad_norm": 1.0880993604660034, + "learning_rate": 6.557162900261843e-07, + "loss": 0.9369, + "step": 13597 + }, + { + "epoch": 0.9343135907654253, + "grad_norm": 1.0496209859848022, + "learning_rate": 6.543526544243866e-07, + "loss": 0.9421, + "step": 13598 + }, + { + "epoch": 0.9343823003985159, + "grad_norm": 0.9482786655426025, + "learning_rate": 6.529904194128384e-07, + "loss": 0.8957, + "step": 13599 + }, + { + "epoch": 0.9344510100316065, + "grad_norm": 1.1314462423324585, + "learning_rate": 6.516295850699072e-07, + "loss": 0.9273, + "step": 13600 + }, + { + "epoch": 0.9345197196646969, + "grad_norm": 1.0101898908615112, + "learning_rate": 6.50270151473878e-07, + "loss": 0.9164, + "step": 13601 + }, + { + "epoch": 0.9345884292977875, + "grad_norm": 1.0345410108566284, + "learning_rate": 6.489121187029601e-07, + "loss": 0.904, + "step": 13602 + }, + { + "epoch": 0.9346571389308781, + "grad_norm": 1.0778532028198242, + "learning_rate": 6.475554868352801e-07, + "loss": 0.936, + "step": 13603 + }, + { + "epoch": 0.9347258485639687, + "grad_norm": 1.1154931783676147, + "learning_rate": 6.462002559488839e-07, + "loss": 0.9452, + "step": 13604 + }, + { + "epoch": 0.9347945581970593, + "grad_norm": 1.0481160879135132, + "learning_rate": 6.448464261217424e-07, + "loss": 0.9321, + "step": 13605 + }, + { + "epoch": 0.9348632678301498, + "grad_norm": 1.0839166641235352, + "learning_rate": 6.434939974317323e-07, + "loss": 0.8959, + "step": 13606 + }, + { + "epoch": 0.9349319774632403, + "grad_norm": 1.1155574321746826, + "learning_rate": 6.421429699566606e-07, + "loss": 0.9532, + "step": 13607 + }, + { + "epoch": 0.9350006870963309, + "grad_norm": 1.0492513179779053, + "learning_rate": 6.407933437742514e-07, + "loss": 0.9512, + "step": 13608 + }, + { + "epoch": 0.9350693967294215, + "grad_norm": 1.0534882545471191, + "learning_rate": 6.394451189621509e-07, + "loss": 0.9339, + "step": 13609 + }, + { + "epoch": 0.935138106362512, + "grad_norm": 1.074899673461914, + "learning_rate": 6.380982955979192e-07, + "loss": 0.9285, + "step": 13610 + }, + { + "epoch": 0.9352068159956026, + "grad_norm": 1.0046107769012451, + "learning_rate": 6.36752873759039e-07, + "loss": 0.8604, + "step": 13611 + }, + { + "epoch": 0.9352755256286931, + "grad_norm": 1.0264453887939453, + "learning_rate": 6.354088535229091e-07, + "loss": 0.8898, + "step": 13612 + }, + { + "epoch": 0.9353442352617837, + "grad_norm": 1.061328411102295, + "learning_rate": 6.340662349668514e-07, + "loss": 0.9663, + "step": 13613 + }, + { + "epoch": 0.9354129448948743, + "grad_norm": 1.106950044631958, + "learning_rate": 6.327250181681066e-07, + "loss": 0.9426, + "step": 13614 + }, + { + "epoch": 0.9354816545279648, + "grad_norm": 1.0664840936660767, + "learning_rate": 6.313852032038381e-07, + "loss": 0.9332, + "step": 13615 + }, + { + "epoch": 0.9355503641610554, + "grad_norm": 1.0474299192428589, + "learning_rate": 6.300467901511176e-07, + "loss": 0.9296, + "step": 13616 + }, + { + "epoch": 0.935619073794146, + "grad_norm": 1.037965178489685, + "learning_rate": 6.287097790869473e-07, + "loss": 0.96, + "step": 13617 + }, + { + "epoch": 0.9356877834272365, + "grad_norm": 1.030075192451477, + "learning_rate": 6.273741700882407e-07, + "loss": 0.9262, + "step": 13618 + }, + { + "epoch": 0.935756493060327, + "grad_norm": 1.0674711465835571, + "learning_rate": 6.260399632318392e-07, + "loss": 0.9662, + "step": 13619 + }, + { + "epoch": 0.9358252026934176, + "grad_norm": 1.00765860080719, + "learning_rate": 6.247071585944981e-07, + "loss": 0.9348, + "step": 13620 + }, + { + "epoch": 0.9358939123265082, + "grad_norm": 1.0846506357192993, + "learning_rate": 6.233757562528892e-07, + "loss": 0.939, + "step": 13621 + }, + { + "epoch": 0.9359626219595988, + "grad_norm": 1.0294060707092285, + "learning_rate": 6.220457562836096e-07, + "loss": 0.9034, + "step": 13622 + }, + { + "epoch": 0.9360313315926893, + "grad_norm": 1.0160534381866455, + "learning_rate": 6.207171587631732e-07, + "loss": 0.8852, + "step": 13623 + }, + { + "epoch": 0.9361000412257798, + "grad_norm": 1.0568008422851562, + "learning_rate": 6.193899637680162e-07, + "loss": 0.9315, + "step": 13624 + }, + { + "epoch": 0.9361687508588704, + "grad_norm": 1.1022918224334717, + "learning_rate": 6.180641713744883e-07, + "loss": 0.9119, + "step": 13625 + }, + { + "epoch": 0.936237460491961, + "grad_norm": 1.0380195379257202, + "learning_rate": 6.167397816588622e-07, + "loss": 0.8916, + "step": 13626 + }, + { + "epoch": 0.9363061701250516, + "grad_norm": 1.0832383632659912, + "learning_rate": 6.154167946973294e-07, + "loss": 0.8825, + "step": 13627 + }, + { + "epoch": 0.936374879758142, + "grad_norm": 1.025261640548706, + "learning_rate": 6.140952105659986e-07, + "loss": 0.9058, + "step": 13628 + }, + { + "epoch": 0.9364435893912326, + "grad_norm": 1.0530073642730713, + "learning_rate": 6.127750293409007e-07, + "loss": 0.9361, + "step": 13629 + }, + { + "epoch": 0.9365122990243232, + "grad_norm": 1.0890897512435913, + "learning_rate": 6.114562510979916e-07, + "loss": 0.9041, + "step": 13630 + }, + { + "epoch": 0.9365810086574138, + "grad_norm": 1.0667521953582764, + "learning_rate": 6.101388759131271e-07, + "loss": 0.927, + "step": 13631 + }, + { + "epoch": 0.9366497182905044, + "grad_norm": 1.0427640676498413, + "learning_rate": 6.088229038621051e-07, + "loss": 0.9487, + "step": 13632 + }, + { + "epoch": 0.9367184279235949, + "grad_norm": 1.0855588912963867, + "learning_rate": 6.07508335020629e-07, + "loss": 0.9622, + "step": 13633 + }, + { + "epoch": 0.9367871375566854, + "grad_norm": 1.005983591079712, + "learning_rate": 6.061951694643242e-07, + "loss": 0.9082, + "step": 13634 + }, + { + "epoch": 0.936855847189776, + "grad_norm": 1.0079463720321655, + "learning_rate": 6.048834072687415e-07, + "loss": 0.9312, + "step": 13635 + }, + { + "epoch": 0.9369245568228666, + "grad_norm": 1.0830857753753662, + "learning_rate": 6.035730485093372e-07, + "loss": 0.9768, + "step": 13636 + }, + { + "epoch": 0.9369932664559572, + "grad_norm": 1.0120187997817993, + "learning_rate": 6.022640932615009e-07, + "loss": 0.8843, + "step": 13637 + }, + { + "epoch": 0.9370619760890477, + "grad_norm": 1.0510632991790771, + "learning_rate": 6.009565416005364e-07, + "loss": 0.8904, + "step": 13638 + }, + { + "epoch": 0.9371306857221382, + "grad_norm": 1.0677907466888428, + "learning_rate": 5.996503936016668e-07, + "loss": 0.9363, + "step": 13639 + }, + { + "epoch": 0.9371993953552288, + "grad_norm": 1.0732225179672241, + "learning_rate": 5.983456493400291e-07, + "loss": 0.9567, + "step": 13640 + }, + { + "epoch": 0.9372681049883194, + "grad_norm": 1.057494878768921, + "learning_rate": 5.970423088906885e-07, + "loss": 0.9268, + "step": 13641 + }, + { + "epoch": 0.9373368146214099, + "grad_norm": 1.0301473140716553, + "learning_rate": 5.957403723286264e-07, + "loss": 0.909, + "step": 13642 + }, + { + "epoch": 0.9374055242545005, + "grad_norm": 1.0627557039260864, + "learning_rate": 5.944398397287387e-07, + "loss": 0.9579, + "step": 13643 + }, + { + "epoch": 0.937474233887591, + "grad_norm": 1.11601984500885, + "learning_rate": 5.931407111658488e-07, + "loss": 0.9695, + "step": 13644 + }, + { + "epoch": 0.9375429435206816, + "grad_norm": 1.090370774269104, + "learning_rate": 5.918429867146913e-07, + "loss": 0.9451, + "step": 13645 + }, + { + "epoch": 0.9376116531537722, + "grad_norm": 1.0607229471206665, + "learning_rate": 5.905466664499231e-07, + "loss": 0.9456, + "step": 13646 + }, + { + "epoch": 0.9376803627868627, + "grad_norm": 1.027464747428894, + "learning_rate": 5.892517504461209e-07, + "loss": 0.9325, + "step": 13647 + }, + { + "epoch": 0.9377490724199533, + "grad_norm": 1.0394841432571411, + "learning_rate": 5.879582387777804e-07, + "loss": 0.9069, + "step": 13648 + }, + { + "epoch": 0.9378177820530439, + "grad_norm": 1.040945053100586, + "learning_rate": 5.8666613151932e-07, + "loss": 0.8944, + "step": 13649 + }, + { + "epoch": 0.9378864916861344, + "grad_norm": 1.0253249406814575, + "learning_rate": 5.853754287450719e-07, + "loss": 0.9064, + "step": 13650 + }, + { + "epoch": 0.9379552013192249, + "grad_norm": 1.0511436462402344, + "learning_rate": 5.840861305292905e-07, + "loss": 0.8967, + "step": 13651 + }, + { + "epoch": 0.9380239109523155, + "grad_norm": 0.9916970133781433, + "learning_rate": 5.827982369461415e-07, + "loss": 0.8898, + "step": 13652 + }, + { + "epoch": 0.9380926205854061, + "grad_norm": 1.0550469160079956, + "learning_rate": 5.815117480697268e-07, + "loss": 0.9165, + "step": 13653 + }, + { + "epoch": 0.9381613302184967, + "grad_norm": 1.0805635452270508, + "learning_rate": 5.802266639740483e-07, + "loss": 0.9494, + "step": 13654 + }, + { + "epoch": 0.9382300398515871, + "grad_norm": 1.069831132888794, + "learning_rate": 5.789429847330441e-07, + "loss": 0.9273, + "step": 13655 + }, + { + "epoch": 0.9382987494846777, + "grad_norm": 1.0466960668563843, + "learning_rate": 5.776607104205578e-07, + "loss": 0.9388, + "step": 13656 + }, + { + "epoch": 0.9383674591177683, + "grad_norm": 1.0484658479690552, + "learning_rate": 5.76379841110361e-07, + "loss": 0.9238, + "step": 13657 + }, + { + "epoch": 0.9384361687508589, + "grad_norm": 1.073557734489441, + "learning_rate": 5.751003768761392e-07, + "loss": 0.9049, + "step": 13658 + }, + { + "epoch": 0.9385048783839495, + "grad_norm": 1.1152877807617188, + "learning_rate": 5.738223177914975e-07, + "loss": 0.9236, + "step": 13659 + }, + { + "epoch": 0.9385735880170399, + "grad_norm": 1.0566132068634033, + "learning_rate": 5.725456639299687e-07, + "loss": 0.8796, + "step": 13660 + }, + { + "epoch": 0.9386422976501305, + "grad_norm": 1.0310187339782715, + "learning_rate": 5.712704153649939e-07, + "loss": 0.9115, + "step": 13661 + }, + { + "epoch": 0.9387110072832211, + "grad_norm": 1.0859973430633545, + "learning_rate": 5.699965721699341e-07, + "loss": 0.958, + "step": 13662 + }, + { + "epoch": 0.9387797169163117, + "grad_norm": 1.0516644716262817, + "learning_rate": 5.68724134418075e-07, + "loss": 0.9238, + "step": 13663 + }, + { + "epoch": 0.9388484265494023, + "grad_norm": 1.0816463232040405, + "learning_rate": 5.674531021826246e-07, + "loss": 0.9183, + "step": 13664 + }, + { + "epoch": 0.9389171361824928, + "grad_norm": 1.075994849205017, + "learning_rate": 5.661834755366969e-07, + "loss": 0.9357, + "step": 13665 + }, + { + "epoch": 0.9389858458155833, + "grad_norm": 1.0415641069412231, + "learning_rate": 5.649152545533332e-07, + "loss": 0.8977, + "step": 13666 + }, + { + "epoch": 0.9390545554486739, + "grad_norm": 1.0315192937850952, + "learning_rate": 5.636484393055003e-07, + "loss": 0.9091, + "step": 13667 + }, + { + "epoch": 0.9391232650817645, + "grad_norm": 0.9858797788619995, + "learning_rate": 5.623830298660704e-07, + "loss": 0.9002, + "step": 13668 + }, + { + "epoch": 0.939191974714855, + "grad_norm": 1.05354905128479, + "learning_rate": 5.611190263078464e-07, + "loss": 0.9499, + "step": 13669 + }, + { + "epoch": 0.9392606843479456, + "grad_norm": 1.0678253173828125, + "learning_rate": 5.598564287035423e-07, + "loss": 0.9515, + "step": 13670 + }, + { + "epoch": 0.9393293939810361, + "grad_norm": 1.0612982511520386, + "learning_rate": 5.585952371257974e-07, + "loss": 0.9345, + "step": 13671 + }, + { + "epoch": 0.9393981036141267, + "grad_norm": 1.0535157918930054, + "learning_rate": 5.573354516471646e-07, + "loss": 0.8832, + "step": 13672 + }, + { + "epoch": 0.9394668132472173, + "grad_norm": 1.0160921812057495, + "learning_rate": 5.560770723401193e-07, + "loss": 0.9253, + "step": 13673 + }, + { + "epoch": 0.9395355228803078, + "grad_norm": 1.032763957977295, + "learning_rate": 5.548200992770536e-07, + "loss": 0.9425, + "step": 13674 + }, + { + "epoch": 0.9396042325133984, + "grad_norm": 1.0428355932235718, + "learning_rate": 5.535645325302846e-07, + "loss": 0.8893, + "step": 13675 + }, + { + "epoch": 0.9396729421464889, + "grad_norm": 1.0542640686035156, + "learning_rate": 5.523103721720463e-07, + "loss": 0.8983, + "step": 13676 + }, + { + "epoch": 0.9397416517795795, + "grad_norm": 1.076854944229126, + "learning_rate": 5.510576182744809e-07, + "loss": 0.9256, + "step": 13677 + }, + { + "epoch": 0.93981036141267, + "grad_norm": 1.0581812858581543, + "learning_rate": 5.49806270909664e-07, + "loss": 0.9559, + "step": 13678 + }, + { + "epoch": 0.9398790710457606, + "grad_norm": 1.097591757774353, + "learning_rate": 5.485563301495827e-07, + "loss": 0.8882, + "step": 13679 + }, + { + "epoch": 0.9399477806788512, + "grad_norm": 1.0718008279800415, + "learning_rate": 5.473077960661488e-07, + "loss": 0.9266, + "step": 13680 + }, + { + "epoch": 0.9400164903119418, + "grad_norm": 1.0613476037979126, + "learning_rate": 5.460606687311909e-07, + "loss": 0.92, + "step": 13681 + }, + { + "epoch": 0.9400851999450323, + "grad_norm": 1.1173455715179443, + "learning_rate": 5.448149482164489e-07, + "loss": 0.9748, + "step": 13682 + }, + { + "epoch": 0.9401539095781228, + "grad_norm": 1.0315524339675903, + "learning_rate": 5.435706345935931e-07, + "loss": 0.9571, + "step": 13683 + }, + { + "epoch": 0.9402226192112134, + "grad_norm": 1.0415469408035278, + "learning_rate": 5.423277279342054e-07, + "loss": 0.9297, + "step": 13684 + }, + { + "epoch": 0.940291328844304, + "grad_norm": 1.0279654264450073, + "learning_rate": 5.410862283097895e-07, + "loss": 0.919, + "step": 13685 + }, + { + "epoch": 0.9403600384773946, + "grad_norm": 1.0952914953231812, + "learning_rate": 5.398461357917745e-07, + "loss": 0.9008, + "step": 13686 + }, + { + "epoch": 0.940428748110485, + "grad_norm": 1.0595383644104004, + "learning_rate": 5.386074504514921e-07, + "loss": 0.9326, + "step": 13687 + }, + { + "epoch": 0.9404974577435756, + "grad_norm": 1.0681649446487427, + "learning_rate": 5.373701723602103e-07, + "loss": 0.9325, + "step": 13688 + }, + { + "epoch": 0.9405661673766662, + "grad_norm": 1.052687168121338, + "learning_rate": 5.361343015891085e-07, + "loss": 0.9335, + "step": 13689 + }, + { + "epoch": 0.9406348770097568, + "grad_norm": 1.0250308513641357, + "learning_rate": 5.348998382092824e-07, + "loss": 0.9613, + "step": 13690 + }, + { + "epoch": 0.9407035866428474, + "grad_norm": 1.0510525703430176, + "learning_rate": 5.336667822917529e-07, + "loss": 0.902, + "step": 13691 + }, + { + "epoch": 0.9407722962759378, + "grad_norm": 1.0519269704818726, + "learning_rate": 5.324351339074524e-07, + "loss": 0.9132, + "step": 13692 + }, + { + "epoch": 0.9408410059090284, + "grad_norm": 0.9954001307487488, + "learning_rate": 5.312048931272406e-07, + "loss": 0.8831, + "step": 13693 + }, + { + "epoch": 0.940909715542119, + "grad_norm": 1.0357645750045776, + "learning_rate": 5.299760600218917e-07, + "loss": 0.9239, + "step": 13694 + }, + { + "epoch": 0.9409784251752096, + "grad_norm": 1.1141507625579834, + "learning_rate": 5.287486346621018e-07, + "loss": 0.9228, + "step": 13695 + }, + { + "epoch": 0.9410471348083002, + "grad_norm": 1.1020655632019043, + "learning_rate": 5.275226171184811e-07, + "loss": 0.9571, + "step": 13696 + }, + { + "epoch": 0.9411158444413907, + "grad_norm": 1.0430338382720947, + "learning_rate": 5.26298007461562e-07, + "loss": 0.914, + "step": 13697 + }, + { + "epoch": 0.9411845540744812, + "grad_norm": 1.080077886581421, + "learning_rate": 5.250748057617966e-07, + "loss": 0.9626, + "step": 13698 + }, + { + "epoch": 0.9412532637075718, + "grad_norm": 0.9891381859779358, + "learning_rate": 5.238530120895535e-07, + "loss": 0.8938, + "step": 13699 + }, + { + "epoch": 0.9413219733406624, + "grad_norm": 1.1355849504470825, + "learning_rate": 5.226326265151238e-07, + "loss": 0.9533, + "step": 13700 + }, + { + "epoch": 0.941390682973753, + "grad_norm": 1.075866937637329, + "learning_rate": 5.214136491087124e-07, + "loss": 0.9231, + "step": 13701 + }, + { + "epoch": 0.9414593926068435, + "grad_norm": 1.045554280281067, + "learning_rate": 5.201960799404521e-07, + "loss": 0.9569, + "step": 13702 + }, + { + "epoch": 0.941528102239934, + "grad_norm": 1.123032808303833, + "learning_rate": 5.18979919080384e-07, + "loss": 0.9333, + "step": 13703 + }, + { + "epoch": 0.9415968118730246, + "grad_norm": 1.0213775634765625, + "learning_rate": 5.177651665984718e-07, + "loss": 0.9254, + "step": 13704 + }, + { + "epoch": 0.9416655215061152, + "grad_norm": 1.0406724214553833, + "learning_rate": 5.165518225646038e-07, + "loss": 0.8856, + "step": 13705 + }, + { + "epoch": 0.9417342311392057, + "grad_norm": 1.0097354650497437, + "learning_rate": 5.153398870485798e-07, + "loss": 0.9016, + "step": 13706 + }, + { + "epoch": 0.9418029407722963, + "grad_norm": 1.0801856517791748, + "learning_rate": 5.141293601201247e-07, + "loss": 0.9319, + "step": 13707 + }, + { + "epoch": 0.9418716504053868, + "grad_norm": 1.1018495559692383, + "learning_rate": 5.129202418488771e-07, + "loss": 0.936, + "step": 13708 + }, + { + "epoch": 0.9419403600384774, + "grad_norm": 1.0883488655090332, + "learning_rate": 5.117125323043981e-07, + "loss": 0.9207, + "step": 13709 + }, + { + "epoch": 0.942009069671568, + "grad_norm": 1.0766500234603882, + "learning_rate": 5.105062315561627e-07, + "loss": 0.9122, + "step": 13710 + }, + { + "epoch": 0.9420777793046585, + "grad_norm": 1.152709722518921, + "learning_rate": 5.093013396735763e-07, + "loss": 0.9171, + "step": 13711 + }, + { + "epoch": 0.9421464889377491, + "grad_norm": 1.1106221675872803, + "learning_rate": 5.080978567259476e-07, + "loss": 0.9739, + "step": 13712 + }, + { + "epoch": 0.9422151985708397, + "grad_norm": 1.1121748685836792, + "learning_rate": 5.06895782782521e-07, + "loss": 0.9693, + "step": 13713 + }, + { + "epoch": 0.9422839082039302, + "grad_norm": 1.0451579093933105, + "learning_rate": 5.056951179124414e-07, + "loss": 0.9508, + "step": 13714 + }, + { + "epoch": 0.9423526178370207, + "grad_norm": 1.0145835876464844, + "learning_rate": 5.04495862184784e-07, + "loss": 0.9012, + "step": 13715 + }, + { + "epoch": 0.9424213274701113, + "grad_norm": 1.0811235904693604, + "learning_rate": 5.032980156685463e-07, + "loss": 0.9167, + "step": 13716 + }, + { + "epoch": 0.9424900371032019, + "grad_norm": 1.1316488981246948, + "learning_rate": 5.021015784326399e-07, + "loss": 0.9308, + "step": 13717 + }, + { + "epoch": 0.9425587467362925, + "grad_norm": 1.0752346515655518, + "learning_rate": 5.009065505458904e-07, + "loss": 0.9462, + "step": 13718 + }, + { + "epoch": 0.9426274563693829, + "grad_norm": 1.009950876235962, + "learning_rate": 4.997129320770483e-07, + "loss": 0.9342, + "step": 13719 + }, + { + "epoch": 0.9426961660024735, + "grad_norm": 1.0688276290893555, + "learning_rate": 4.985207230947836e-07, + "loss": 0.9613, + "step": 13720 + }, + { + "epoch": 0.9427648756355641, + "grad_norm": 1.0561463832855225, + "learning_rate": 4.973299236676831e-07, + "loss": 0.9178, + "step": 13721 + }, + { + "epoch": 0.9428335852686547, + "grad_norm": 1.0008054971694946, + "learning_rate": 4.961405338642533e-07, + "loss": 0.8812, + "step": 13722 + }, + { + "epoch": 0.9429022949017453, + "grad_norm": 1.075090765953064, + "learning_rate": 4.949525537529143e-07, + "loss": 0.9482, + "step": 13723 + }, + { + "epoch": 0.9429710045348357, + "grad_norm": 1.087072491645813, + "learning_rate": 4.937659834020142e-07, + "loss": 0.9397, + "step": 13724 + }, + { + "epoch": 0.9430397141679263, + "grad_norm": 1.0504732131958008, + "learning_rate": 4.925808228798151e-07, + "loss": 0.9197, + "step": 13725 + }, + { + "epoch": 0.9431084238010169, + "grad_norm": 1.043493390083313, + "learning_rate": 4.913970722544986e-07, + "loss": 0.9196, + "step": 13726 + }, + { + "epoch": 0.9431771334341075, + "grad_norm": 1.1784930229187012, + "learning_rate": 4.902147315941657e-07, + "loss": 0.9905, + "step": 13727 + }, + { + "epoch": 0.943245843067198, + "grad_norm": 1.0419013500213623, + "learning_rate": 4.890338009668316e-07, + "loss": 0.9288, + "step": 13728 + }, + { + "epoch": 0.9433145527002886, + "grad_norm": 1.0821878910064697, + "learning_rate": 4.878542804404418e-07, + "loss": 0.9093, + "step": 13729 + }, + { + "epoch": 0.9433832623333791, + "grad_norm": 1.0448570251464844, + "learning_rate": 4.866761700828448e-07, + "loss": 0.9141, + "step": 13730 + }, + { + "epoch": 0.9434519719664697, + "grad_norm": 1.0556389093399048, + "learning_rate": 4.854994699618254e-07, + "loss": 0.9014, + "step": 13731 + }, + { + "epoch": 0.9435206815995603, + "grad_norm": 1.039147973060608, + "learning_rate": 4.843241801450738e-07, + "loss": 0.9223, + "step": 13732 + }, + { + "epoch": 0.9435893912326508, + "grad_norm": 1.0713698863983154, + "learning_rate": 4.831503007002026e-07, + "loss": 0.9288, + "step": 13733 + }, + { + "epoch": 0.9436581008657414, + "grad_norm": 1.0389362573623657, + "learning_rate": 4.819778316947465e-07, + "loss": 0.9364, + "step": 13734 + }, + { + "epoch": 0.9437268104988319, + "grad_norm": 1.0876562595367432, + "learning_rate": 4.808067731961574e-07, + "loss": 0.9339, + "step": 13735 + }, + { + "epoch": 0.9437955201319225, + "grad_norm": 1.1031770706176758, + "learning_rate": 4.796371252718035e-07, + "loss": 0.9733, + "step": 13736 + }, + { + "epoch": 0.943864229765013, + "grad_norm": 1.1045877933502197, + "learning_rate": 4.784688879889726e-07, + "loss": 0.9145, + "step": 13737 + }, + { + "epoch": 0.9439329393981036, + "grad_norm": 1.0862094163894653, + "learning_rate": 4.773020614148805e-07, + "loss": 0.904, + "step": 13738 + }, + { + "epoch": 0.9440016490311942, + "grad_norm": 1.0525258779525757, + "learning_rate": 4.761366456166427e-07, + "loss": 0.9282, + "step": 13739 + }, + { + "epoch": 0.9440703586642847, + "grad_norm": 1.0321564674377441, + "learning_rate": 4.749726406613142e-07, + "loss": 0.8847, + "step": 13740 + }, + { + "epoch": 0.9441390682973753, + "grad_norm": 1.163675308227539, + "learning_rate": 4.738100466158524e-07, + "loss": 0.901, + "step": 13741 + }, + { + "epoch": 0.9442077779304658, + "grad_norm": 1.1248772144317627, + "learning_rate": 4.7264886354714823e-07, + "loss": 0.9777, + "step": 13742 + }, + { + "epoch": 0.9442764875635564, + "grad_norm": 1.1089543104171753, + "learning_rate": 4.714890915219983e-07, + "loss": 0.9489, + "step": 13743 + }, + { + "epoch": 0.944345197196647, + "grad_norm": 1.0761433839797974, + "learning_rate": 4.7033073060712416e-07, + "loss": 0.9146, + "step": 13744 + }, + { + "epoch": 0.9444139068297376, + "grad_norm": 1.0153815746307373, + "learning_rate": 4.691737808691671e-07, + "loss": 0.8798, + "step": 13745 + }, + { + "epoch": 0.944482616462828, + "grad_norm": 1.0740917921066284, + "learning_rate": 4.680182423746848e-07, + "loss": 0.9442, + "step": 13746 + }, + { + "epoch": 0.9445513260959186, + "grad_norm": 1.1119858026504517, + "learning_rate": 4.668641151901576e-07, + "loss": 0.9259, + "step": 13747 + }, + { + "epoch": 0.9446200357290092, + "grad_norm": 1.0890263319015503, + "learning_rate": 4.657113993819767e-07, + "loss": 0.9516, + "step": 13748 + }, + { + "epoch": 0.9446887453620998, + "grad_norm": 1.043960690498352, + "learning_rate": 4.645600950164586e-07, + "loss": 0.9076, + "step": 13749 + }, + { + "epoch": 0.9447574549951904, + "grad_norm": 1.1643061637878418, + "learning_rate": 4.634102021598391e-07, + "loss": 0.9363, + "step": 13750 + }, + { + "epoch": 0.9448261646282808, + "grad_norm": 1.0869498252868652, + "learning_rate": 4.622617208782709e-07, + "loss": 0.9408, + "step": 13751 + }, + { + "epoch": 0.9448948742613714, + "grad_norm": 1.0487897396087646, + "learning_rate": 4.6111465123782617e-07, + "loss": 0.9367, + "step": 13752 + }, + { + "epoch": 0.944963583894462, + "grad_norm": 1.112092137336731, + "learning_rate": 4.5996899330449094e-07, + "loss": 0.9316, + "step": 13753 + }, + { + "epoch": 0.9450322935275526, + "grad_norm": 1.0461105108261108, + "learning_rate": 4.5882474714417643e-07, + "loss": 0.9142, + "step": 13754 + }, + { + "epoch": 0.9451010031606432, + "grad_norm": 1.1499552726745605, + "learning_rate": 4.5768191282271056e-07, + "loss": 0.9977, + "step": 13755 + }, + { + "epoch": 0.9451697127937336, + "grad_norm": 1.0999163389205933, + "learning_rate": 4.565404904058379e-07, + "loss": 0.8956, + "step": 13756 + }, + { + "epoch": 0.9452384224268242, + "grad_norm": 1.06465744972229, + "learning_rate": 4.55400479959231e-07, + "loss": 0.909, + "step": 13757 + }, + { + "epoch": 0.9453071320599148, + "grad_norm": 1.004543423652649, + "learning_rate": 4.5426188154846504e-07, + "loss": 0.9454, + "step": 13758 + }, + { + "epoch": 0.9453758416930054, + "grad_norm": 1.0128220319747925, + "learning_rate": 4.531246952390489e-07, + "loss": 0.9401, + "step": 13759 + }, + { + "epoch": 0.945444551326096, + "grad_norm": 1.1002238988876343, + "learning_rate": 4.519889210964023e-07, + "loss": 0.9239, + "step": 13760 + }, + { + "epoch": 0.9455132609591865, + "grad_norm": 1.0359342098236084, + "learning_rate": 4.508545591858648e-07, + "loss": 0.9365, + "step": 13761 + }, + { + "epoch": 0.945581970592277, + "grad_norm": 1.0536116361618042, + "learning_rate": 4.497216095726953e-07, + "loss": 0.9113, + "step": 13762 + }, + { + "epoch": 0.9456506802253676, + "grad_norm": 1.0561832189559937, + "learning_rate": 4.4859007232207207e-07, + "loss": 0.9124, + "step": 13763 + }, + { + "epoch": 0.9457193898584582, + "grad_norm": 1.0483458042144775, + "learning_rate": 4.474599474990931e-07, + "loss": 0.9416, + "step": 13764 + }, + { + "epoch": 0.9457880994915487, + "grad_norm": 1.0789780616760254, + "learning_rate": 4.463312351687732e-07, + "loss": 0.9499, + "step": 13765 + }, + { + "epoch": 0.9458568091246393, + "grad_norm": 1.0242915153503418, + "learning_rate": 4.4520393539604357e-07, + "loss": 0.9418, + "step": 13766 + }, + { + "epoch": 0.9459255187577298, + "grad_norm": 1.1934844255447388, + "learning_rate": 4.4407804824576074e-07, + "loss": 0.972, + "step": 13767 + }, + { + "epoch": 0.9459942283908204, + "grad_norm": 1.0243760347366333, + "learning_rate": 4.4295357378269785e-07, + "loss": 0.9398, + "step": 13768 + }, + { + "epoch": 0.946062938023911, + "grad_norm": 1.116477370262146, + "learning_rate": 4.4183051207154213e-07, + "loss": 0.9259, + "step": 13769 + }, + { + "epoch": 0.9461316476570015, + "grad_norm": 1.135294795036316, + "learning_rate": 4.4070886317690017e-07, + "loss": 0.942, + "step": 13770 + }, + { + "epoch": 0.9462003572900921, + "grad_norm": 1.1182985305786133, + "learning_rate": 4.3958862716330375e-07, + "loss": 0.923, + "step": 13771 + }, + { + "epoch": 0.9462690669231826, + "grad_norm": 0.9930106997489929, + "learning_rate": 4.3846980409519847e-07, + "loss": 0.8896, + "step": 13772 + }, + { + "epoch": 0.9463377765562732, + "grad_norm": 1.107454538345337, + "learning_rate": 4.3735239403695227e-07, + "loss": 0.9526, + "step": 13773 + }, + { + "epoch": 0.9464064861893637, + "grad_norm": 1.0822296142578125, + "learning_rate": 4.3623639705284157e-07, + "loss": 0.9499, + "step": 13774 + }, + { + "epoch": 0.9464751958224543, + "grad_norm": 1.0029778480529785, + "learning_rate": 4.351218132070761e-07, + "loss": 0.9173, + "step": 13775 + }, + { + "epoch": 0.9465439054555449, + "grad_norm": 1.0827585458755493, + "learning_rate": 4.3400864256377407e-07, + "loss": 0.9785, + "step": 13776 + }, + { + "epoch": 0.9466126150886355, + "grad_norm": 1.0650335550308228, + "learning_rate": 4.328968851869758e-07, + "loss": 0.9181, + "step": 13777 + }, + { + "epoch": 0.946681324721726, + "grad_norm": 1.0367194414138794, + "learning_rate": 4.3178654114064133e-07, + "loss": 0.8994, + "step": 13778 + }, + { + "epoch": 0.9467500343548165, + "grad_norm": 1.0029802322387695, + "learning_rate": 4.306776104886473e-07, + "loss": 0.9281, + "step": 13779 + }, + { + "epoch": 0.9468187439879071, + "grad_norm": 1.0884454250335693, + "learning_rate": 4.2957009329478993e-07, + "loss": 0.9189, + "step": 13780 + }, + { + "epoch": 0.9468874536209977, + "grad_norm": 1.1025524139404297, + "learning_rate": 4.284639896227821e-07, + "loss": 0.9286, + "step": 13781 + }, + { + "epoch": 0.9469561632540883, + "grad_norm": 1.051811695098877, + "learning_rate": 4.273592995362591e-07, + "loss": 0.9218, + "step": 13782 + }, + { + "epoch": 0.9470248728871787, + "grad_norm": 1.033125877380371, + "learning_rate": 4.2625602309877276e-07, + "loss": 0.92, + "step": 13783 + }, + { + "epoch": 0.9470935825202693, + "grad_norm": 1.0456209182739258, + "learning_rate": 4.251541603737974e-07, + "loss": 0.9409, + "step": 13784 + }, + { + "epoch": 0.9471622921533599, + "grad_norm": 0.9870741367340088, + "learning_rate": 4.240537114247156e-07, + "loss": 0.9089, + "step": 13785 + }, + { + "epoch": 0.9472310017864505, + "grad_norm": 1.11373770236969, + "learning_rate": 4.229546763148379e-07, + "loss": 1.0047, + "step": 13786 + }, + { + "epoch": 0.9472997114195411, + "grad_norm": 1.0250385999679565, + "learning_rate": 4.2185705510739415e-07, + "loss": 0.925, + "step": 13787 + }, + { + "epoch": 0.9473684210526315, + "grad_norm": 1.0071606636047363, + "learning_rate": 4.2076084786553116e-07, + "loss": 0.9132, + "step": 13788 + }, + { + "epoch": 0.9474371306857221, + "grad_norm": 1.0528191328048706, + "learning_rate": 4.19666054652304e-07, + "loss": 0.9407, + "step": 13789 + }, + { + "epoch": 0.9475058403188127, + "grad_norm": 1.0912330150604248, + "learning_rate": 4.1857267553070403e-07, + "loss": 0.9013, + "step": 13790 + }, + { + "epoch": 0.9475745499519033, + "grad_norm": 1.1144219636917114, + "learning_rate": 4.1748071056362804e-07, + "loss": 0.9292, + "step": 13791 + }, + { + "epoch": 0.9476432595849938, + "grad_norm": 1.0792875289916992, + "learning_rate": 4.1639015981390086e-07, + "loss": 0.9338, + "step": 13792 + }, + { + "epoch": 0.9477119692180844, + "grad_norm": 1.0534058809280396, + "learning_rate": 4.1530102334425836e-07, + "loss": 0.9328, + "step": 13793 + }, + { + "epoch": 0.9477806788511749, + "grad_norm": 1.0582114458084106, + "learning_rate": 4.14213301217356e-07, + "loss": 0.9218, + "step": 13794 + }, + { + "epoch": 0.9478493884842655, + "grad_norm": 1.0722938776016235, + "learning_rate": 4.1312699349577155e-07, + "loss": 0.9304, + "step": 13795 + }, + { + "epoch": 0.9479180981173561, + "grad_norm": 1.1251505613327026, + "learning_rate": 4.120421002419994e-07, + "loss": 0.9143, + "step": 13796 + }, + { + "epoch": 0.9479868077504466, + "grad_norm": 1.0145269632339478, + "learning_rate": 4.1095862151845356e-07, + "loss": 0.8932, + "step": 13797 + }, + { + "epoch": 0.9480555173835372, + "grad_norm": 1.076309323310852, + "learning_rate": 4.0987655738746746e-07, + "loss": 0.9285, + "step": 13798 + }, + { + "epoch": 0.9481242270166277, + "grad_norm": 1.102840781211853, + "learning_rate": 4.0879590791128853e-07, + "loss": 0.9159, + "step": 13799 + }, + { + "epoch": 0.9481929366497183, + "grad_norm": 1.0561063289642334, + "learning_rate": 4.077166731520865e-07, + "loss": 0.9354, + "step": 13800 + }, + { + "epoch": 0.9482616462828088, + "grad_norm": 1.037518858909607, + "learning_rate": 4.066388531719478e-07, + "loss": 0.9329, + "step": 13801 + }, + { + "epoch": 0.9483303559158994, + "grad_norm": 1.0192112922668457, + "learning_rate": 4.0556244803288126e-07, + "loss": 0.9508, + "step": 13802 + }, + { + "epoch": 0.94839906554899, + "grad_norm": 1.0446038246154785, + "learning_rate": 4.0448745779681494e-07, + "loss": 0.9389, + "step": 13803 + }, + { + "epoch": 0.9484677751820805, + "grad_norm": 1.0574455261230469, + "learning_rate": 4.034138825255857e-07, + "loss": 0.8907, + "step": 13804 + }, + { + "epoch": 0.948536484815171, + "grad_norm": 1.027052640914917, + "learning_rate": 4.023417222809578e-07, + "loss": 0.9214, + "step": 13805 + }, + { + "epoch": 0.9486051944482616, + "grad_norm": 0.9951720237731934, + "learning_rate": 4.0127097712461257e-07, + "loss": 0.9154, + "step": 13806 + }, + { + "epoch": 0.9486739040813522, + "grad_norm": 1.0630730390548706, + "learning_rate": 4.0020164711815076e-07, + "loss": 0.9574, + "step": 13807 + }, + { + "epoch": 0.9487426137144428, + "grad_norm": 1.0507186651229858, + "learning_rate": 3.9913373232308705e-07, + "loss": 0.9692, + "step": 13808 + }, + { + "epoch": 0.9488113233475334, + "grad_norm": 1.0353939533233643, + "learning_rate": 3.9806723280086386e-07, + "loss": 0.8933, + "step": 13809 + }, + { + "epoch": 0.9488800329806238, + "grad_norm": 1.0432510375976562, + "learning_rate": 3.970021486128267e-07, + "loss": 0.9238, + "step": 13810 + }, + { + "epoch": 0.9489487426137144, + "grad_norm": 1.0457963943481445, + "learning_rate": 3.95938479820257e-07, + "loss": 0.9245, + "step": 13811 + }, + { + "epoch": 0.949017452246805, + "grad_norm": 1.0532584190368652, + "learning_rate": 3.9487622648434474e-07, + "loss": 0.9353, + "step": 13812 + }, + { + "epoch": 0.9490861618798956, + "grad_norm": 1.0938020944595337, + "learning_rate": 3.9381538866619937e-07, + "loss": 0.9399, + "step": 13813 + }, + { + "epoch": 0.9491548715129862, + "grad_norm": 1.0815690755844116, + "learning_rate": 3.9275596642685543e-07, + "loss": 0.9608, + "step": 13814 + }, + { + "epoch": 0.9492235811460766, + "grad_norm": 1.0948246717453003, + "learning_rate": 3.91697959827253e-07, + "loss": 0.9406, + "step": 13815 + }, + { + "epoch": 0.9492922907791672, + "grad_norm": 1.0304300785064697, + "learning_rate": 3.9064136892826287e-07, + "loss": 0.9733, + "step": 13816 + }, + { + "epoch": 0.9493610004122578, + "grad_norm": 1.0521223545074463, + "learning_rate": 3.895861937906697e-07, + "loss": 0.9309, + "step": 13817 + }, + { + "epoch": 0.9494297100453484, + "grad_norm": 1.159808874130249, + "learning_rate": 3.8853243447517785e-07, + "loss": 0.9045, + "step": 13818 + }, + { + "epoch": 0.949498419678439, + "grad_norm": 1.0827933549880981, + "learning_rate": 3.8748009104241087e-07, + "loss": 0.9173, + "step": 13819 + }, + { + "epoch": 0.9495671293115294, + "grad_norm": 1.0548981428146362, + "learning_rate": 3.864291635529038e-07, + "loss": 0.9237, + "step": 13820 + }, + { + "epoch": 0.94963583894462, + "grad_norm": 0.9733108282089233, + "learning_rate": 3.8537965206711935e-07, + "loss": 0.8791, + "step": 13821 + }, + { + "epoch": 0.9497045485777106, + "grad_norm": 0.9898475408554077, + "learning_rate": 3.8433155664543697e-07, + "loss": 0.9176, + "step": 13822 + }, + { + "epoch": 0.9497732582108012, + "grad_norm": 1.1292821168899536, + "learning_rate": 3.832848773481501e-07, + "loss": 0.9387, + "step": 13823 + }, + { + "epoch": 0.9498419678438917, + "grad_norm": 1.1099138259887695, + "learning_rate": 3.822396142354745e-07, + "loss": 0.9288, + "step": 13824 + }, + { + "epoch": 0.9499106774769823, + "grad_norm": 1.0526282787322998, + "learning_rate": 3.811957673675426e-07, + "loss": 0.9553, + "step": 13825 + }, + { + "epoch": 0.9499793871100728, + "grad_norm": 1.046433925628662, + "learning_rate": 3.8015333680440914e-07, + "loss": 0.906, + "step": 13826 + }, + { + "epoch": 0.9500480967431634, + "grad_norm": 0.9820589423179626, + "learning_rate": 3.7911232260604277e-07, + "loss": 0.9228, + "step": 13827 + }, + { + "epoch": 0.950116806376254, + "grad_norm": 1.0738213062286377, + "learning_rate": 3.780727248323318e-07, + "loss": 0.985, + "step": 13828 + }, + { + "epoch": 0.9501855160093445, + "grad_norm": 1.1119279861450195, + "learning_rate": 3.7703454354308385e-07, + "loss": 0.9083, + "step": 13829 + }, + { + "epoch": 0.9502542256424351, + "grad_norm": 0.998784601688385, + "learning_rate": 3.7599777879802623e-07, + "loss": 0.9137, + "step": 13830 + }, + { + "epoch": 0.9503229352755256, + "grad_norm": 1.0759962797164917, + "learning_rate": 3.749624306568028e-07, + "loss": 0.9222, + "step": 13831 + }, + { + "epoch": 0.9503916449086162, + "grad_norm": 1.055018663406372, + "learning_rate": 3.739284991789743e-07, + "loss": 0.9254, + "step": 13832 + }, + { + "epoch": 0.9504603545417067, + "grad_norm": 1.0715546607971191, + "learning_rate": 3.7289598442402374e-07, + "loss": 0.9371, + "step": 13833 + }, + { + "epoch": 0.9505290641747973, + "grad_norm": 1.072340726852417, + "learning_rate": 3.718648864513535e-07, + "loss": 0.9485, + "step": 13834 + }, + { + "epoch": 0.9505977738078879, + "grad_norm": 1.073588252067566, + "learning_rate": 3.7083520532027737e-07, + "loss": 0.9198, + "step": 13835 + }, + { + "epoch": 0.9506664834409784, + "grad_norm": 1.0398211479187012, + "learning_rate": 3.69806941090034e-07, + "loss": 0.9203, + "step": 13836 + }, + { + "epoch": 0.950735193074069, + "grad_norm": 1.045859932899475, + "learning_rate": 3.6878009381978164e-07, + "loss": 0.9186, + "step": 13837 + }, + { + "epoch": 0.9508039027071595, + "grad_norm": 1.059051752090454, + "learning_rate": 3.677546635685897e-07, + "loss": 0.9163, + "step": 13838 + }, + { + "epoch": 0.9508726123402501, + "grad_norm": 1.0935602188110352, + "learning_rate": 3.667306503954554e-07, + "loss": 0.9116, + "step": 13839 + }, + { + "epoch": 0.9509413219733407, + "grad_norm": 1.0253630876541138, + "learning_rate": 3.657080543592845e-07, + "loss": 0.9071, + "step": 13840 + }, + { + "epoch": 0.9510100316064313, + "grad_norm": 1.0403231382369995, + "learning_rate": 3.6468687551891034e-07, + "loss": 0.9553, + "step": 13841 + }, + { + "epoch": 0.9510787412395217, + "grad_norm": 1.0199532508850098, + "learning_rate": 3.636671139330805e-07, + "loss": 0.8929, + "step": 13842 + }, + { + "epoch": 0.9511474508726123, + "grad_norm": 1.04966402053833, + "learning_rate": 3.626487696604564e-07, + "loss": 0.8813, + "step": 13843 + }, + { + "epoch": 0.9512161605057029, + "grad_norm": 1.1065456867218018, + "learning_rate": 3.616318427596327e-07, + "loss": 0.9834, + "step": 13844 + }, + { + "epoch": 0.9512848701387935, + "grad_norm": 1.0417481660842896, + "learning_rate": 3.606163332891016e-07, + "loss": 0.9049, + "step": 13845 + }, + { + "epoch": 0.9513535797718841, + "grad_norm": 1.035931944847107, + "learning_rate": 3.5960224130728857e-07, + "loss": 0.9119, + "step": 13846 + }, + { + "epoch": 0.9514222894049745, + "grad_norm": 1.0508596897125244, + "learning_rate": 3.585895668725359e-07, + "loss": 0.9102, + "step": 13847 + }, + { + "epoch": 0.9514909990380651, + "grad_norm": 1.070745587348938, + "learning_rate": 3.575783100430996e-07, + "loss": 0.9004, + "step": 13848 + }, + { + "epoch": 0.9515597086711557, + "grad_norm": 1.0909985303878784, + "learning_rate": 3.5656847087716104e-07, + "loss": 0.9274, + "step": 13849 + }, + { + "epoch": 0.9516284183042463, + "grad_norm": 1.0721975564956665, + "learning_rate": 3.5556004943280986e-07, + "loss": 0.9383, + "step": 13850 + }, + { + "epoch": 0.9516971279373369, + "grad_norm": 1.0468841791152954, + "learning_rate": 3.5455304576806347e-07, + "loss": 0.9307, + "step": 13851 + }, + { + "epoch": 0.9517658375704273, + "grad_norm": 1.0777194499969482, + "learning_rate": 3.5354745994085337e-07, + "loss": 0.9263, + "step": 13852 + }, + { + "epoch": 0.9518345472035179, + "grad_norm": 1.094817876815796, + "learning_rate": 3.5254329200903047e-07, + "loss": 0.8972, + "step": 13853 + }, + { + "epoch": 0.9519032568366085, + "grad_norm": 1.1307185888290405, + "learning_rate": 3.5154054203036247e-07, + "loss": 0.9309, + "step": 13854 + }, + { + "epoch": 0.9519719664696991, + "grad_norm": 1.0559775829315186, + "learning_rate": 3.505392100625393e-07, + "loss": 0.932, + "step": 13855 + }, + { + "epoch": 0.9520406761027896, + "grad_norm": 1.0468181371688843, + "learning_rate": 3.4953929616316494e-07, + "loss": 0.9383, + "step": 13856 + }, + { + "epoch": 0.9521093857358802, + "grad_norm": 1.0748292207717896, + "learning_rate": 3.4854080038976556e-07, + "loss": 0.9067, + "step": 13857 + }, + { + "epoch": 0.9521780953689707, + "grad_norm": 0.9997394680976868, + "learning_rate": 3.475437227997813e-07, + "loss": 0.8785, + "step": 13858 + }, + { + "epoch": 0.9522468050020613, + "grad_norm": 1.0085054636001587, + "learning_rate": 3.4654806345057743e-07, + "loss": 0.9053, + "step": 13859 + }, + { + "epoch": 0.9523155146351518, + "grad_norm": 1.0349355936050415, + "learning_rate": 3.4555382239943037e-07, + "loss": 0.8968, + "step": 13860 + }, + { + "epoch": 0.9523842242682424, + "grad_norm": 1.017142653465271, + "learning_rate": 3.445609997035387e-07, + "loss": 0.8882, + "step": 13861 + }, + { + "epoch": 0.952452933901333, + "grad_norm": 1.041959285736084, + "learning_rate": 3.43569595420018e-07, + "loss": 0.9101, + "step": 13862 + }, + { + "epoch": 0.9525216435344235, + "grad_norm": 1.042968988418579, + "learning_rate": 3.4257960960590585e-07, + "loss": 0.9348, + "step": 13863 + }, + { + "epoch": 0.9525903531675141, + "grad_norm": 1.0649430751800537, + "learning_rate": 3.4159104231815674e-07, + "loss": 0.9429, + "step": 13864 + }, + { + "epoch": 0.9526590628006046, + "grad_norm": 1.047082781791687, + "learning_rate": 3.406038936136391e-07, + "loss": 0.916, + "step": 13865 + }, + { + "epoch": 0.9527277724336952, + "grad_norm": 1.0431965589523315, + "learning_rate": 3.3961816354914356e-07, + "loss": 0.9154, + "step": 13866 + }, + { + "epoch": 0.9527964820667858, + "grad_norm": 1.0933469533920288, + "learning_rate": 3.3863385218137755e-07, + "loss": 0.8897, + "step": 13867 + }, + { + "epoch": 0.9528651916998763, + "grad_norm": 0.9993520975112915, + "learning_rate": 3.37650959566968e-07, + "loss": 0.8773, + "step": 13868 + }, + { + "epoch": 0.9529339013329668, + "grad_norm": 1.0709141492843628, + "learning_rate": 3.3666948576246417e-07, + "loss": 0.9424, + "step": 13869 + }, + { + "epoch": 0.9530026109660574, + "grad_norm": 1.115227460861206, + "learning_rate": 3.3568943082432633e-07, + "loss": 0.9434, + "step": 13870 + }, + { + "epoch": 0.953071320599148, + "grad_norm": 1.0690513849258423, + "learning_rate": 3.3471079480893455e-07, + "loss": 0.9202, + "step": 13871 + }, + { + "epoch": 0.9531400302322386, + "grad_norm": 1.058638572692871, + "learning_rate": 3.3373357777259364e-07, + "loss": 0.9619, + "step": 13872 + }, + { + "epoch": 0.9532087398653292, + "grad_norm": 1.0573188066482544, + "learning_rate": 3.3275777977151976e-07, + "loss": 0.9162, + "step": 13873 + }, + { + "epoch": 0.9532774494984196, + "grad_norm": 1.045134425163269, + "learning_rate": 3.317834008618487e-07, + "loss": 0.9489, + "step": 13874 + }, + { + "epoch": 0.9533461591315102, + "grad_norm": 1.0712639093399048, + "learning_rate": 3.30810441099641e-07, + "loss": 0.9175, + "step": 13875 + }, + { + "epoch": 0.9534148687646008, + "grad_norm": 1.0374528169631958, + "learning_rate": 3.298389005408659e-07, + "loss": 0.9307, + "step": 13876 + }, + { + "epoch": 0.9534835783976914, + "grad_norm": 1.057019591331482, + "learning_rate": 3.288687792414147e-07, + "loss": 0.946, + "step": 13877 + }, + { + "epoch": 0.953552288030782, + "grad_norm": 1.0628130435943604, + "learning_rate": 3.279000772571039e-07, + "loss": 0.881, + "step": 13878 + }, + { + "epoch": 0.9536209976638724, + "grad_norm": 1.0703892707824707, + "learning_rate": 3.2693279464365287e-07, + "loss": 0.8914, + "step": 13879 + }, + { + "epoch": 0.953689707296963, + "grad_norm": 1.0472664833068848, + "learning_rate": 3.259669314567199e-07, + "loss": 0.8998, + "step": 13880 + }, + { + "epoch": 0.9537584169300536, + "grad_norm": 1.037590742111206, + "learning_rate": 3.2500248775186036e-07, + "loss": 0.91, + "step": 13881 + }, + { + "epoch": 0.9538271265631442, + "grad_norm": 1.071506381034851, + "learning_rate": 3.2403946358456337e-07, + "loss": 0.9516, + "step": 13882 + }, + { + "epoch": 0.9538958361962347, + "grad_norm": 1.0286390781402588, + "learning_rate": 3.230778590102318e-07, + "loss": 0.9385, + "step": 13883 + }, + { + "epoch": 0.9539645458293252, + "grad_norm": 1.0790674686431885, + "learning_rate": 3.221176740841825e-07, + "loss": 0.9393, + "step": 13884 + }, + { + "epoch": 0.9540332554624158, + "grad_norm": 1.0797141790390015, + "learning_rate": 3.2115890886165735e-07, + "loss": 0.9495, + "step": 13885 + }, + { + "epoch": 0.9541019650955064, + "grad_norm": 1.047607421875, + "learning_rate": 3.2020156339781503e-07, + "loss": 0.9518, + "step": 13886 + }, + { + "epoch": 0.954170674728597, + "grad_norm": 1.1209027767181396, + "learning_rate": 3.1924563774772543e-07, + "loss": 0.9449, + "step": 13887 + }, + { + "epoch": 0.9542393843616875, + "grad_norm": 1.083793044090271, + "learning_rate": 3.1829113196638614e-07, + "loss": 0.9706, + "step": 13888 + }, + { + "epoch": 0.9543080939947781, + "grad_norm": 1.0457229614257812, + "learning_rate": 3.1733804610870886e-07, + "loss": 0.9396, + "step": 13889 + }, + { + "epoch": 0.9543768036278686, + "grad_norm": 1.0834258794784546, + "learning_rate": 3.1638638022952474e-07, + "loss": 0.9241, + "step": 13890 + }, + { + "epoch": 0.9544455132609592, + "grad_norm": 1.0262199640274048, + "learning_rate": 3.1543613438358155e-07, + "loss": 0.905, + "step": 13891 + }, + { + "epoch": 0.9545142228940497, + "grad_norm": 1.0828512907028198, + "learning_rate": 3.144873086255468e-07, + "loss": 0.9117, + "step": 13892 + }, + { + "epoch": 0.9545829325271403, + "grad_norm": 1.1193939447402954, + "learning_rate": 3.1353990301000447e-07, + "loss": 0.9542, + "step": 13893 + }, + { + "epoch": 0.9546516421602309, + "grad_norm": 1.06670343875885, + "learning_rate": 3.125939175914611e-07, + "loss": 0.9574, + "step": 13894 + }, + { + "epoch": 0.9547203517933214, + "grad_norm": 1.071449637413025, + "learning_rate": 3.116493524243369e-07, + "loss": 0.9055, + "step": 13895 + }, + { + "epoch": 0.954789061426412, + "grad_norm": 1.0665626525878906, + "learning_rate": 3.107062075629691e-07, + "loss": 0.9254, + "step": 13896 + }, + { + "epoch": 0.9548577710595025, + "grad_norm": 1.0433917045593262, + "learning_rate": 3.0976448306162255e-07, + "loss": 0.9392, + "step": 13897 + }, + { + "epoch": 0.9549264806925931, + "grad_norm": 1.0973602533340454, + "learning_rate": 3.088241789744706e-07, + "loss": 0.9505, + "step": 13898 + }, + { + "epoch": 0.9549951903256837, + "grad_norm": 1.0194140672683716, + "learning_rate": 3.0788529535560605e-07, + "loss": 0.9102, + "step": 13899 + }, + { + "epoch": 0.9550638999587742, + "grad_norm": 1.0507646799087524, + "learning_rate": 3.0694783225904687e-07, + "loss": 0.924, + "step": 13900 + }, + { + "epoch": 0.9551326095918647, + "grad_norm": 1.128298044204712, + "learning_rate": 3.060117897387221e-07, + "loss": 0.9609, + "step": 13901 + }, + { + "epoch": 0.9552013192249553, + "grad_norm": 1.0549463033676147, + "learning_rate": 3.0507716784848595e-07, + "loss": 0.9634, + "step": 13902 + }, + { + "epoch": 0.9552700288580459, + "grad_norm": 1.0817883014678955, + "learning_rate": 3.041439666420981e-07, + "loss": 0.9414, + "step": 13903 + }, + { + "epoch": 0.9553387384911365, + "grad_norm": 1.1047072410583496, + "learning_rate": 3.0321218617325176e-07, + "loss": 0.9132, + "step": 13904 + }, + { + "epoch": 0.9554074481242271, + "grad_norm": 1.1194989681243896, + "learning_rate": 3.0228182649555126e-07, + "loss": 0.9602, + "step": 13905 + }, + { + "epoch": 0.9554761577573175, + "grad_norm": 0.9883902668952942, + "learning_rate": 3.013528876625149e-07, + "loss": 0.9249, + "step": 13906 + }, + { + "epoch": 0.9555448673904081, + "grad_norm": 1.0682682991027832, + "learning_rate": 3.0042536972758884e-07, + "loss": 0.937, + "step": 13907 + }, + { + "epoch": 0.9556135770234987, + "grad_norm": 1.0529297590255737, + "learning_rate": 2.994992727441331e-07, + "loss": 0.9346, + "step": 13908 + }, + { + "epoch": 0.9556822866565893, + "grad_norm": 1.04657781124115, + "learning_rate": 2.9857459676542185e-07, + "loss": 0.9225, + "step": 13909 + }, + { + "epoch": 0.9557509962896799, + "grad_norm": 1.0849392414093018, + "learning_rate": 2.9765134184465137e-07, + "loss": 0.9503, + "step": 13910 + }, + { + "epoch": 0.9558197059227703, + "grad_norm": 1.0928218364715576, + "learning_rate": 2.9672950803494027e-07, + "loss": 0.8874, + "step": 13911 + }, + { + "epoch": 0.9558884155558609, + "grad_norm": 1.0382105112075806, + "learning_rate": 2.9580909538931556e-07, + "loss": 0.9409, + "step": 13912 + }, + { + "epoch": 0.9559571251889515, + "grad_norm": 1.0071210861206055, + "learning_rate": 2.9489010396073216e-07, + "loss": 0.8973, + "step": 13913 + }, + { + "epoch": 0.9560258348220421, + "grad_norm": 1.0590420961380005, + "learning_rate": 2.9397253380205604e-07, + "loss": 0.9217, + "step": 13914 + }, + { + "epoch": 0.9560945444551326, + "grad_norm": 1.0115976333618164, + "learning_rate": 2.9305638496607836e-07, + "loss": 0.9296, + "step": 13915 + }, + { + "epoch": 0.9561632540882231, + "grad_norm": 1.0438722372055054, + "learning_rate": 2.9214165750550137e-07, + "loss": 0.9224, + "step": 13916 + }, + { + "epoch": 0.9562319637213137, + "grad_norm": 1.0326769351959229, + "learning_rate": 2.9122835147294693e-07, + "loss": 0.8993, + "step": 13917 + }, + { + "epoch": 0.9563006733544043, + "grad_norm": 0.9892563819885254, + "learning_rate": 2.9031646692096184e-07, + "loss": 0.9012, + "step": 13918 + }, + { + "epoch": 0.9563693829874949, + "grad_norm": 1.06629478931427, + "learning_rate": 2.894060039020041e-07, + "loss": 0.9737, + "step": 13919 + }, + { + "epoch": 0.9564380926205854, + "grad_norm": 0.9878289699554443, + "learning_rate": 2.884969624684541e-07, + "loss": 0.9191, + "step": 13920 + }, + { + "epoch": 0.956506802253676, + "grad_norm": 1.071443796157837, + "learning_rate": 2.875893426726034e-07, + "loss": 0.9346, + "step": 13921 + }, + { + "epoch": 0.9565755118867665, + "grad_norm": 1.015599012374878, + "learning_rate": 2.8668314456667123e-07, + "loss": 0.8975, + "step": 13922 + }, + { + "epoch": 0.9566442215198571, + "grad_norm": 1.106429100036621, + "learning_rate": 2.857783682027881e-07, + "loss": 0.9622, + "step": 13923 + }, + { + "epoch": 0.9567129311529476, + "grad_norm": 1.1126797199249268, + "learning_rate": 2.848750136330097e-07, + "loss": 0.9374, + "step": 13924 + }, + { + "epoch": 0.9567816407860382, + "grad_norm": 1.0609164237976074, + "learning_rate": 2.839730809092972e-07, + "loss": 0.9236, + "step": 13925 + }, + { + "epoch": 0.9568503504191288, + "grad_norm": 1.0506212711334229, + "learning_rate": 2.8307257008354515e-07, + "loss": 0.9414, + "step": 13926 + }, + { + "epoch": 0.9569190600522193, + "grad_norm": 1.0146586894989014, + "learning_rate": 2.8217348120755935e-07, + "loss": 0.9194, + "step": 13927 + }, + { + "epoch": 0.9569877696853099, + "grad_norm": 1.1007003784179688, + "learning_rate": 2.812758143330596e-07, + "loss": 0.9556, + "step": 13928 + }, + { + "epoch": 0.9570564793184004, + "grad_norm": 1.0984687805175781, + "learning_rate": 2.803795695116934e-07, + "loss": 0.9204, + "step": 13929 + }, + { + "epoch": 0.957125188951491, + "grad_norm": 1.084271788597107, + "learning_rate": 2.7948474679501403e-07, + "loss": 0.9394, + "step": 13930 + }, + { + "epoch": 0.9571938985845816, + "grad_norm": 1.11652410030365, + "learning_rate": 2.78591346234508e-07, + "loss": 0.9365, + "step": 13931 + }, + { + "epoch": 0.9572626082176722, + "grad_norm": 1.0958291292190552, + "learning_rate": 2.776993678815676e-07, + "loss": 0.9444, + "step": 13932 + }, + { + "epoch": 0.9573313178507626, + "grad_norm": 1.0388078689575195, + "learning_rate": 2.768088117875045e-07, + "loss": 0.9374, + "step": 13933 + }, + { + "epoch": 0.9574000274838532, + "grad_norm": 1.060337781906128, + "learning_rate": 2.7591967800356113e-07, + "loss": 0.8905, + "step": 13934 + }, + { + "epoch": 0.9574687371169438, + "grad_norm": 1.0973176956176758, + "learning_rate": 2.750319665808798e-07, + "loss": 0.9908, + "step": 13935 + }, + { + "epoch": 0.9575374467500344, + "grad_norm": 1.1191706657409668, + "learning_rate": 2.7414567757053645e-07, + "loss": 0.9631, + "step": 13936 + }, + { + "epoch": 0.957606156383125, + "grad_norm": 1.1209707260131836, + "learning_rate": 2.732608110235152e-07, + "loss": 0.9289, + "step": 13937 + }, + { + "epoch": 0.9576748660162154, + "grad_norm": 1.0096038579940796, + "learning_rate": 2.723773669907226e-07, + "loss": 0.9094, + "step": 13938 + }, + { + "epoch": 0.957743575649306, + "grad_norm": 1.0930761098861694, + "learning_rate": 2.714953455229818e-07, + "loss": 0.956, + "step": 13939 + }, + { + "epoch": 0.9578122852823966, + "grad_norm": 1.0750983953475952, + "learning_rate": 2.706147466710385e-07, + "loss": 0.9351, + "step": 13940 + }, + { + "epoch": 0.9578809949154872, + "grad_norm": 1.0491994619369507, + "learning_rate": 2.697355704855492e-07, + "loss": 0.9314, + "step": 13941 + }, + { + "epoch": 0.9579497045485778, + "grad_norm": 1.0830976963043213, + "learning_rate": 2.6885781701709303e-07, + "loss": 0.9507, + "step": 13942 + }, + { + "epoch": 0.9580184141816682, + "grad_norm": 1.0436726808547974, + "learning_rate": 2.679814863161684e-07, + "loss": 0.9225, + "step": 13943 + }, + { + "epoch": 0.9580871238147588, + "grad_norm": 1.0507835149765015, + "learning_rate": 2.671065784331878e-07, + "loss": 0.8998, + "step": 13944 + }, + { + "epoch": 0.9581558334478494, + "grad_norm": 1.0454647541046143, + "learning_rate": 2.662330934184859e-07, + "loss": 0.9085, + "step": 13945 + }, + { + "epoch": 0.95822454308094, + "grad_norm": 1.0966461896896362, + "learning_rate": 2.6536103132231427e-07, + "loss": 0.9651, + "step": 13946 + }, + { + "epoch": 0.9582932527140305, + "grad_norm": 1.086499571800232, + "learning_rate": 2.644903921948383e-07, + "loss": 0.95, + "step": 13947 + }, + { + "epoch": 0.9583619623471211, + "grad_norm": 1.0477561950683594, + "learning_rate": 2.6362117608614844e-07, + "loss": 0.9706, + "step": 13948 + }, + { + "epoch": 0.9584306719802116, + "grad_norm": 1.0589673519134521, + "learning_rate": 2.6275338304625195e-07, + "loss": 0.9315, + "step": 13949 + }, + { + "epoch": 0.9584993816133022, + "grad_norm": 1.1477296352386475, + "learning_rate": 2.6188701312507e-07, + "loss": 0.946, + "step": 13950 + }, + { + "epoch": 0.9585680912463928, + "grad_norm": 1.045416235923767, + "learning_rate": 2.610220663724433e-07, + "loss": 0.9055, + "step": 13951 + }, + { + "epoch": 0.9586368008794833, + "grad_norm": 1.0205210447311401, + "learning_rate": 2.601585428381348e-07, + "loss": 0.941, + "step": 13952 + }, + { + "epoch": 0.9587055105125739, + "grad_norm": 1.004856824874878, + "learning_rate": 2.592964425718186e-07, + "loss": 0.8902, + "step": 13953 + }, + { + "epoch": 0.9587742201456644, + "grad_norm": 1.045806884765625, + "learning_rate": 2.584357656230912e-07, + "loss": 0.8859, + "step": 13954 + }, + { + "epoch": 0.958842929778755, + "grad_norm": 1.0077078342437744, + "learning_rate": 2.575765120414714e-07, + "loss": 0.9252, + "step": 13955 + }, + { + "epoch": 0.9589116394118455, + "grad_norm": 1.0186705589294434, + "learning_rate": 2.567186818763889e-07, + "loss": 0.9286, + "step": 13956 + }, + { + "epoch": 0.9589803490449361, + "grad_norm": 1.0485761165618896, + "learning_rate": 2.558622751771905e-07, + "loss": 0.912, + "step": 13957 + }, + { + "epoch": 0.9590490586780267, + "grad_norm": 1.0243827104568481, + "learning_rate": 2.5500729199315066e-07, + "loss": 0.9114, + "step": 13958 + }, + { + "epoch": 0.9591177683111172, + "grad_norm": 1.0470324754714966, + "learning_rate": 2.541537323734522e-07, + "loss": 0.9314, + "step": 13959 + }, + { + "epoch": 0.9591864779442077, + "grad_norm": 1.121978521347046, + "learning_rate": 2.5330159636720307e-07, + "loss": 0.9443, + "step": 13960 + }, + { + "epoch": 0.9592551875772983, + "grad_norm": 1.0279465913772583, + "learning_rate": 2.524508840234224e-07, + "loss": 0.9231, + "step": 13961 + }, + { + "epoch": 0.9593238972103889, + "grad_norm": 1.0355910062789917, + "learning_rate": 2.5160159539105443e-07, + "loss": 0.9265, + "step": 13962 + }, + { + "epoch": 0.9593926068434795, + "grad_norm": 1.0219284296035767, + "learning_rate": 2.507537305189572e-07, + "loss": 0.9308, + "step": 13963 + }, + { + "epoch": 0.9594613164765701, + "grad_norm": 1.0405269861221313, + "learning_rate": 2.499072894559057e-07, + "loss": 0.9343, + "step": 13964 + }, + { + "epoch": 0.9595300261096605, + "grad_norm": 1.062942624092102, + "learning_rate": 2.49062272250597e-07, + "loss": 0.923, + "step": 13965 + }, + { + "epoch": 0.9595987357427511, + "grad_norm": 1.0092065334320068, + "learning_rate": 2.4821867895164496e-07, + "loss": 0.8917, + "step": 13966 + }, + { + "epoch": 0.9596674453758417, + "grad_norm": 1.0682332515716553, + "learning_rate": 2.473765096075803e-07, + "loss": 0.9325, + "step": 13967 + }, + { + "epoch": 0.9597361550089323, + "grad_norm": 1.051476240158081, + "learning_rate": 2.465357642668531e-07, + "loss": 0.9356, + "step": 13968 + }, + { + "epoch": 0.9598048646420229, + "grad_norm": 1.0677436590194702, + "learning_rate": 2.456964429778275e-07, + "loss": 0.9173, + "step": 13969 + }, + { + "epoch": 0.9598735742751133, + "grad_norm": 1.1696335077285767, + "learning_rate": 2.448585457887953e-07, + "loss": 0.9574, + "step": 13970 + }, + { + "epoch": 0.9599422839082039, + "grad_norm": 1.1037851572036743, + "learning_rate": 2.4402207274795697e-07, + "loss": 0.9067, + "step": 13971 + }, + { + "epoch": 0.9600109935412945, + "grad_norm": 1.0365538597106934, + "learning_rate": 2.4318702390343226e-07, + "loss": 0.8927, + "step": 13972 + }, + { + "epoch": 0.9600797031743851, + "grad_norm": 1.0767970085144043, + "learning_rate": 2.423533993032634e-07, + "loss": 0.925, + "step": 13973 + }, + { + "epoch": 0.9601484128074756, + "grad_norm": 1.101503849029541, + "learning_rate": 2.4152119899540915e-07, + "loss": 0.9575, + "step": 13974 + }, + { + "epoch": 0.9602171224405661, + "grad_norm": 1.0956326723098755, + "learning_rate": 2.4069042302774236e-07, + "loss": 0.947, + "step": 13975 + }, + { + "epoch": 0.9602858320736567, + "grad_norm": 1.0681174993515015, + "learning_rate": 2.398610714480609e-07, + "loss": 0.9762, + "step": 13976 + }, + { + "epoch": 0.9603545417067473, + "grad_norm": 1.0121204853057861, + "learning_rate": 2.390331443040739e-07, + "loss": 0.9212, + "step": 13977 + }, + { + "epoch": 0.9604232513398379, + "grad_norm": 1.0415809154510498, + "learning_rate": 2.3820664164341256e-07, + "loss": 0.9077, + "step": 13978 + }, + { + "epoch": 0.9604919609729284, + "grad_norm": 1.0300366878509521, + "learning_rate": 2.3738156351362506e-07, + "loss": 0.9215, + "step": 13979 + }, + { + "epoch": 0.960560670606019, + "grad_norm": 1.0652025938034058, + "learning_rate": 2.3655790996217896e-07, + "loss": 0.8987, + "step": 13980 + }, + { + "epoch": 0.9606293802391095, + "grad_norm": 1.0462002754211426, + "learning_rate": 2.3573568103645582e-07, + "loss": 0.9207, + "step": 13981 + }, + { + "epoch": 0.9606980898722001, + "grad_norm": 1.128944993019104, + "learning_rate": 2.3491487678375946e-07, + "loss": 0.9504, + "step": 13982 + }, + { + "epoch": 0.9607667995052906, + "grad_norm": 1.0603771209716797, + "learning_rate": 2.3409549725131042e-07, + "loss": 0.91, + "step": 13983 + }, + { + "epoch": 0.9608355091383812, + "grad_norm": 1.0658942461013794, + "learning_rate": 2.33277542486246e-07, + "loss": 0.9809, + "step": 13984 + }, + { + "epoch": 0.9609042187714718, + "grad_norm": 1.0671719312667847, + "learning_rate": 2.3246101253562303e-07, + "loss": 0.9274, + "step": 13985 + }, + { + "epoch": 0.9609729284045623, + "grad_norm": 1.0822876691818237, + "learning_rate": 2.316459074464178e-07, + "loss": 0.9369, + "step": 13986 + }, + { + "epoch": 0.9610416380376529, + "grad_norm": 1.0309628248214722, + "learning_rate": 2.3083222726552335e-07, + "loss": 0.8993, + "step": 13987 + }, + { + "epoch": 0.9611103476707434, + "grad_norm": 1.0063483715057373, + "learning_rate": 2.3001997203974667e-07, + "loss": 0.909, + "step": 13988 + }, + { + "epoch": 0.961179057303834, + "grad_norm": 1.0353749990463257, + "learning_rate": 2.292091418158171e-07, + "loss": 0.8991, + "step": 13989 + }, + { + "epoch": 0.9612477669369246, + "grad_norm": 1.072662115097046, + "learning_rate": 2.283997366403806e-07, + "loss": 0.9667, + "step": 13990 + }, + { + "epoch": 0.9613164765700151, + "grad_norm": 1.0392874479293823, + "learning_rate": 2.2759175656000554e-07, + "loss": 0.9109, + "step": 13991 + }, + { + "epoch": 0.9613851862031056, + "grad_norm": 1.0304604768753052, + "learning_rate": 2.2678520162117134e-07, + "loss": 0.9596, + "step": 13992 + }, + { + "epoch": 0.9614538958361962, + "grad_norm": 1.089744210243225, + "learning_rate": 2.259800718702798e-07, + "loss": 0.933, + "step": 13993 + }, + { + "epoch": 0.9615226054692868, + "grad_norm": 1.0507222414016724, + "learning_rate": 2.2517636735364945e-07, + "loss": 0.9294, + "step": 13994 + }, + { + "epoch": 0.9615913151023774, + "grad_norm": 1.041906476020813, + "learning_rate": 2.2437408811751547e-07, + "loss": 0.878, + "step": 13995 + }, + { + "epoch": 0.961660024735468, + "grad_norm": 1.0609941482543945, + "learning_rate": 2.2357323420803544e-07, + "loss": 0.9213, + "step": 13996 + }, + { + "epoch": 0.9617287343685584, + "grad_norm": 1.0197694301605225, + "learning_rate": 2.2277380567127805e-07, + "loss": 0.9343, + "step": 13997 + }, + { + "epoch": 0.961797444001649, + "grad_norm": 1.0915803909301758, + "learning_rate": 2.219758025532398e-07, + "loss": 0.9292, + "step": 13998 + }, + { + "epoch": 0.9618661536347396, + "grad_norm": 1.0855954885482788, + "learning_rate": 2.211792248998229e-07, + "loss": 0.9656, + "step": 13999 + }, + { + "epoch": 0.9619348632678302, + "grad_norm": 1.0029369592666626, + "learning_rate": 2.2038407275685458e-07, + "loss": 0.9081, + "step": 14000 + } + ], + "logging_steps": 1, + "max_steps": 14554, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.857042430184286e+19, + "train_batch_size": 256, + "trial_name": null, + "trial_params": null +}