diff --git "a/checkpoint-14240/trainer_state.json" "b/checkpoint-14240/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-14240/trainer_state.json" @@ -0,0 +1,99713 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 5.0, + "eval_steps": 500, + "global_step": 14240, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00035112359550561797, + "grad_norm": 2.8951518535614014, + "learning_rate": 5.0000000000000004e-08, + "loss": 1.0953, + "step": 1 + }, + { + "epoch": 0.0007022471910112359, + "grad_norm": 2.683199167251587, + "learning_rate": 1.0000000000000001e-07, + "loss": 1.0614, + "step": 2 + }, + { + "epoch": 0.001053370786516854, + "grad_norm": 2.72992205619812, + "learning_rate": 1.5000000000000002e-07, + "loss": 1.0674, + "step": 3 + }, + { + "epoch": 0.0014044943820224719, + "grad_norm": 2.6386663913726807, + "learning_rate": 2.0000000000000002e-07, + "loss": 1.058, + "step": 4 + }, + { + "epoch": 0.0017556179775280898, + "grad_norm": 2.672447443008423, + "learning_rate": 2.5000000000000004e-07, + "loss": 1.0523, + "step": 5 + }, + { + "epoch": 0.002106741573033708, + "grad_norm": 2.5482404232025146, + "learning_rate": 3.0000000000000004e-07, + "loss": 1.0454, + "step": 6 + }, + { + "epoch": 0.0024578651685393258, + "grad_norm": 2.651028633117676, + "learning_rate": 3.5000000000000004e-07, + "loss": 1.0646, + "step": 7 + }, + { + "epoch": 0.0028089887640449437, + "grad_norm": 2.76011323928833, + "learning_rate": 4.0000000000000003e-07, + "loss": 1.061, + "step": 8 + }, + { + "epoch": 0.0031601123595505617, + "grad_norm": 2.4421677589416504, + "learning_rate": 4.5000000000000003e-07, + "loss": 1.0827, + "step": 9 + }, + { + "epoch": 0.0035112359550561797, + "grad_norm": 2.460603713989258, + "learning_rate": 5.000000000000001e-07, + "loss": 1.0925, + "step": 10 + }, + { + "epoch": 0.0038623595505617976, + "grad_norm": 2.3482112884521484, + "learning_rate": 5.5e-07, + "loss": 1.0679, + "step": 11 + }, + { + "epoch": 0.004213483146067416, + "grad_norm": 2.4199249744415283, + "learning_rate": 6.000000000000001e-07, + "loss": 1.0685, + "step": 12 + }, + { + "epoch": 0.004564606741573034, + "grad_norm": 2.171464443206787, + "learning_rate": 6.5e-07, + "loss": 1.0455, + "step": 13 + }, + { + "epoch": 0.0049157303370786515, + "grad_norm": 2.212657928466797, + "learning_rate": 7.000000000000001e-07, + "loss": 1.0718, + "step": 14 + }, + { + "epoch": 0.0052668539325842695, + "grad_norm": 2.0458426475524902, + "learning_rate": 7.5e-07, + "loss": 1.0213, + "step": 15 + }, + { + "epoch": 0.0056179775280898875, + "grad_norm": 1.8622595071792603, + "learning_rate": 8.000000000000001e-07, + "loss": 1.0266, + "step": 16 + }, + { + "epoch": 0.0059691011235955055, + "grad_norm": 1.7484514713287354, + "learning_rate": 8.500000000000001e-07, + "loss": 1.0439, + "step": 17 + }, + { + "epoch": 0.006320224719101123, + "grad_norm": 1.626994252204895, + "learning_rate": 9.000000000000001e-07, + "loss": 1.0274, + "step": 18 + }, + { + "epoch": 0.006671348314606741, + "grad_norm": 1.462096095085144, + "learning_rate": 9.500000000000001e-07, + "loss": 1.036, + "step": 19 + }, + { + "epoch": 0.007022471910112359, + "grad_norm": 1.4871413707733154, + "learning_rate": 1.0000000000000002e-06, + "loss": 1.033, + "step": 20 + }, + { + "epoch": 0.007373595505617977, + "grad_norm": 1.4984582662582397, + "learning_rate": 1.0500000000000001e-06, + "loss": 1.0279, + "step": 21 + }, + { + "epoch": 0.007724719101123595, + "grad_norm": 1.5213954448699951, + "learning_rate": 1.1e-06, + "loss": 1.0468, + "step": 22 + }, + { + "epoch": 0.008075842696629214, + "grad_norm": 1.3945717811584473, + "learning_rate": 1.1500000000000002e-06, + "loss": 1.0252, + "step": 23 + }, + { + "epoch": 0.008426966292134831, + "grad_norm": 1.2632803916931152, + "learning_rate": 1.2000000000000002e-06, + "loss": 0.9637, + "step": 24 + }, + { + "epoch": 0.00877808988764045, + "grad_norm": 1.2537329196929932, + "learning_rate": 1.25e-06, + "loss": 1.0194, + "step": 25 + }, + { + "epoch": 0.009129213483146067, + "grad_norm": 1.287577748298645, + "learning_rate": 1.3e-06, + "loss": 0.9491, + "step": 26 + }, + { + "epoch": 0.009480337078651686, + "grad_norm": 1.1792410612106323, + "learning_rate": 1.3500000000000002e-06, + "loss": 1.0, + "step": 27 + }, + { + "epoch": 0.009831460674157303, + "grad_norm": 1.1308109760284424, + "learning_rate": 1.4000000000000001e-06, + "loss": 1.002, + "step": 28 + }, + { + "epoch": 0.010182584269662922, + "grad_norm": 1.1292983293533325, + "learning_rate": 1.45e-06, + "loss": 1.0114, + "step": 29 + }, + { + "epoch": 0.010533707865168539, + "grad_norm": 1.0696202516555786, + "learning_rate": 1.5e-06, + "loss": 0.9371, + "step": 30 + }, + { + "epoch": 0.010884831460674158, + "grad_norm": 1.0532927513122559, + "learning_rate": 1.5500000000000002e-06, + "loss": 0.9805, + "step": 31 + }, + { + "epoch": 0.011235955056179775, + "grad_norm": 1.039662480354309, + "learning_rate": 1.6000000000000001e-06, + "loss": 0.9591, + "step": 32 + }, + { + "epoch": 0.011587078651685394, + "grad_norm": 0.9709979891777039, + "learning_rate": 1.6500000000000003e-06, + "loss": 0.9599, + "step": 33 + }, + { + "epoch": 0.011938202247191011, + "grad_norm": 0.9826909303665161, + "learning_rate": 1.7000000000000002e-06, + "loss": 0.9492, + "step": 34 + }, + { + "epoch": 0.01228932584269663, + "grad_norm": 1.01341712474823, + "learning_rate": 1.75e-06, + "loss": 1.0066, + "step": 35 + }, + { + "epoch": 0.012640449438202247, + "grad_norm": 1.0003409385681152, + "learning_rate": 1.8000000000000001e-06, + "loss": 0.9641, + "step": 36 + }, + { + "epoch": 0.012991573033707866, + "grad_norm": 0.9853388667106628, + "learning_rate": 1.85e-06, + "loss": 0.9681, + "step": 37 + }, + { + "epoch": 0.013342696629213483, + "grad_norm": 0.9508251547813416, + "learning_rate": 1.9000000000000002e-06, + "loss": 0.9242, + "step": 38 + }, + { + "epoch": 0.013693820224719102, + "grad_norm": 0.9218304753303528, + "learning_rate": 1.9500000000000004e-06, + "loss": 0.8917, + "step": 39 + }, + { + "epoch": 0.014044943820224719, + "grad_norm": 0.9143016934394836, + "learning_rate": 2.0000000000000003e-06, + "loss": 0.8715, + "step": 40 + }, + { + "epoch": 0.014396067415730338, + "grad_norm": 0.9491222500801086, + "learning_rate": 2.05e-06, + "loss": 0.9157, + "step": 41 + }, + { + "epoch": 0.014747191011235955, + "grad_norm": 0.9453691244125366, + "learning_rate": 2.1000000000000002e-06, + "loss": 0.9537, + "step": 42 + }, + { + "epoch": 0.015098314606741573, + "grad_norm": 0.9372655153274536, + "learning_rate": 2.15e-06, + "loss": 0.9288, + "step": 43 + }, + { + "epoch": 0.01544943820224719, + "grad_norm": 0.9183003306388855, + "learning_rate": 2.2e-06, + "loss": 0.8841, + "step": 44 + }, + { + "epoch": 0.01580056179775281, + "grad_norm": 0.924336314201355, + "learning_rate": 2.25e-06, + "loss": 0.9305, + "step": 45 + }, + { + "epoch": 0.016151685393258428, + "grad_norm": 0.9038196206092834, + "learning_rate": 2.3000000000000004e-06, + "loss": 0.9376, + "step": 46 + }, + { + "epoch": 0.016502808988764044, + "grad_norm": 0.9091538786888123, + "learning_rate": 2.35e-06, + "loss": 0.9243, + "step": 47 + }, + { + "epoch": 0.016853932584269662, + "grad_norm": 0.9212149381637573, + "learning_rate": 2.4000000000000003e-06, + "loss": 0.9515, + "step": 48 + }, + { + "epoch": 0.01720505617977528, + "grad_norm": 0.8974660038948059, + "learning_rate": 2.4500000000000003e-06, + "loss": 0.893, + "step": 49 + }, + { + "epoch": 0.0175561797752809, + "grad_norm": 0.8716650009155273, + "learning_rate": 2.5e-06, + "loss": 0.9485, + "step": 50 + }, + { + "epoch": 0.017907303370786515, + "grad_norm": 0.8729946613311768, + "learning_rate": 2.55e-06, + "loss": 0.912, + "step": 51 + }, + { + "epoch": 0.018258426966292134, + "grad_norm": 0.8486078977584839, + "learning_rate": 2.6e-06, + "loss": 0.9479, + "step": 52 + }, + { + "epoch": 0.018609550561797753, + "grad_norm": 0.8713150024414062, + "learning_rate": 2.6500000000000005e-06, + "loss": 0.9049, + "step": 53 + }, + { + "epoch": 0.018960674157303372, + "grad_norm": 0.855265200138092, + "learning_rate": 2.7000000000000004e-06, + "loss": 0.921, + "step": 54 + }, + { + "epoch": 0.019311797752808987, + "grad_norm": 0.8279822468757629, + "learning_rate": 2.7500000000000004e-06, + "loss": 0.8507, + "step": 55 + }, + { + "epoch": 0.019662921348314606, + "grad_norm": 0.8624402284622192, + "learning_rate": 2.8000000000000003e-06, + "loss": 0.9292, + "step": 56 + }, + { + "epoch": 0.020014044943820225, + "grad_norm": 0.8482896089553833, + "learning_rate": 2.85e-06, + "loss": 0.881, + "step": 57 + }, + { + "epoch": 0.020365168539325844, + "grad_norm": 0.8510951399803162, + "learning_rate": 2.9e-06, + "loss": 0.8702, + "step": 58 + }, + { + "epoch": 0.02071629213483146, + "grad_norm": 0.8771702647209167, + "learning_rate": 2.95e-06, + "loss": 0.9329, + "step": 59 + }, + { + "epoch": 0.021067415730337078, + "grad_norm": 0.8551446199417114, + "learning_rate": 3e-06, + "loss": 0.8908, + "step": 60 + }, + { + "epoch": 0.021418539325842697, + "grad_norm": 0.8884496688842773, + "learning_rate": 3.05e-06, + "loss": 0.8791, + "step": 61 + }, + { + "epoch": 0.021769662921348316, + "grad_norm": 0.8115338087081909, + "learning_rate": 3.1000000000000004e-06, + "loss": 0.8544, + "step": 62 + }, + { + "epoch": 0.02212078651685393, + "grad_norm": 0.8549612760543823, + "learning_rate": 3.1500000000000003e-06, + "loss": 0.8963, + "step": 63 + }, + { + "epoch": 0.02247191011235955, + "grad_norm": 0.8259643316268921, + "learning_rate": 3.2000000000000003e-06, + "loss": 0.8299, + "step": 64 + }, + { + "epoch": 0.02282303370786517, + "grad_norm": 0.8829634785652161, + "learning_rate": 3.2500000000000002e-06, + "loss": 0.8873, + "step": 65 + }, + { + "epoch": 0.023174157303370788, + "grad_norm": 0.8536106944084167, + "learning_rate": 3.3000000000000006e-06, + "loss": 0.9009, + "step": 66 + }, + { + "epoch": 0.023525280898876403, + "grad_norm": 0.8468846082687378, + "learning_rate": 3.3500000000000005e-06, + "loss": 0.8921, + "step": 67 + }, + { + "epoch": 0.023876404494382022, + "grad_norm": 0.8253934383392334, + "learning_rate": 3.4000000000000005e-06, + "loss": 0.848, + "step": 68 + }, + { + "epoch": 0.02422752808988764, + "grad_norm": 0.8778877854347229, + "learning_rate": 3.45e-06, + "loss": 0.8887, + "step": 69 + }, + { + "epoch": 0.02457865168539326, + "grad_norm": 0.8790880441665649, + "learning_rate": 3.5e-06, + "loss": 0.8849, + "step": 70 + }, + { + "epoch": 0.024929775280898875, + "grad_norm": 0.843599259853363, + "learning_rate": 3.5500000000000003e-06, + "loss": 0.8972, + "step": 71 + }, + { + "epoch": 0.025280898876404494, + "grad_norm": 0.8428463339805603, + "learning_rate": 3.6000000000000003e-06, + "loss": 0.9027, + "step": 72 + }, + { + "epoch": 0.025632022471910113, + "grad_norm": 0.8088714480400085, + "learning_rate": 3.65e-06, + "loss": 0.8562, + "step": 73 + }, + { + "epoch": 0.02598314606741573, + "grad_norm": 0.857280969619751, + "learning_rate": 3.7e-06, + "loss": 0.8725, + "step": 74 + }, + { + "epoch": 0.026334269662921347, + "grad_norm": 0.8321940898895264, + "learning_rate": 3.7500000000000005e-06, + "loss": 0.8656, + "step": 75 + }, + { + "epoch": 0.026685393258426966, + "grad_norm": 0.8289700150489807, + "learning_rate": 3.8000000000000005e-06, + "loss": 0.8211, + "step": 76 + }, + { + "epoch": 0.027036516853932584, + "grad_norm": 0.8454659581184387, + "learning_rate": 3.85e-06, + "loss": 0.9003, + "step": 77 + }, + { + "epoch": 0.027387640449438203, + "grad_norm": 0.8450126647949219, + "learning_rate": 3.900000000000001e-06, + "loss": 0.8595, + "step": 78 + }, + { + "epoch": 0.02773876404494382, + "grad_norm": 0.8216248750686646, + "learning_rate": 3.95e-06, + "loss": 0.8503, + "step": 79 + }, + { + "epoch": 0.028089887640449437, + "grad_norm": 0.8196912407875061, + "learning_rate": 4.000000000000001e-06, + "loss": 0.8965, + "step": 80 + }, + { + "epoch": 0.028441011235955056, + "grad_norm": 0.8264586329460144, + "learning_rate": 4.05e-06, + "loss": 0.879, + "step": 81 + }, + { + "epoch": 0.028792134831460675, + "grad_norm": 0.8712279200553894, + "learning_rate": 4.1e-06, + "loss": 0.8825, + "step": 82 + }, + { + "epoch": 0.02914325842696629, + "grad_norm": 0.8609493970870972, + "learning_rate": 4.15e-06, + "loss": 0.8901, + "step": 83 + }, + { + "epoch": 0.02949438202247191, + "grad_norm": 0.8273038864135742, + "learning_rate": 4.2000000000000004e-06, + "loss": 0.8589, + "step": 84 + }, + { + "epoch": 0.029845505617977528, + "grad_norm": 0.8164820075035095, + "learning_rate": 4.25e-06, + "loss": 0.8099, + "step": 85 + }, + { + "epoch": 0.030196629213483147, + "grad_norm": 0.8666368722915649, + "learning_rate": 4.3e-06, + "loss": 0.8535, + "step": 86 + }, + { + "epoch": 0.030547752808988762, + "grad_norm": 0.8173773288726807, + "learning_rate": 4.350000000000001e-06, + "loss": 0.806, + "step": 87 + }, + { + "epoch": 0.03089887640449438, + "grad_norm": 0.8484762907028198, + "learning_rate": 4.4e-06, + "loss": 0.8678, + "step": 88 + }, + { + "epoch": 0.03125, + "grad_norm": 0.8572366833686829, + "learning_rate": 4.450000000000001e-06, + "loss": 0.8939, + "step": 89 + }, + { + "epoch": 0.03160112359550562, + "grad_norm": 0.8265636563301086, + "learning_rate": 4.5e-06, + "loss": 0.8373, + "step": 90 + }, + { + "epoch": 0.03195224719101124, + "grad_norm": 0.8756334185600281, + "learning_rate": 4.5500000000000005e-06, + "loss": 0.8322, + "step": 91 + }, + { + "epoch": 0.032303370786516857, + "grad_norm": 0.8236197829246521, + "learning_rate": 4.600000000000001e-06, + "loss": 0.8425, + "step": 92 + }, + { + "epoch": 0.032654494382022475, + "grad_norm": 0.855594277381897, + "learning_rate": 4.65e-06, + "loss": 0.8779, + "step": 93 + }, + { + "epoch": 0.03300561797752809, + "grad_norm": 0.8004137873649597, + "learning_rate": 4.7e-06, + "loss": 0.8078, + "step": 94 + }, + { + "epoch": 0.033356741573033706, + "grad_norm": 0.7953411340713501, + "learning_rate": 4.75e-06, + "loss": 0.8363, + "step": 95 + }, + { + "epoch": 0.033707865168539325, + "grad_norm": 0.8278580904006958, + "learning_rate": 4.800000000000001e-06, + "loss": 0.8301, + "step": 96 + }, + { + "epoch": 0.034058988764044944, + "grad_norm": 0.8473982810974121, + "learning_rate": 4.85e-06, + "loss": 0.8408, + "step": 97 + }, + { + "epoch": 0.03441011235955056, + "grad_norm": 0.8142527341842651, + "learning_rate": 4.9000000000000005e-06, + "loss": 0.8295, + "step": 98 + }, + { + "epoch": 0.03476123595505618, + "grad_norm": 0.8610883355140686, + "learning_rate": 4.95e-06, + "loss": 0.8405, + "step": 99 + }, + { + "epoch": 0.0351123595505618, + "grad_norm": 0.8903759717941284, + "learning_rate": 5e-06, + "loss": 0.8793, + "step": 100 + }, + { + "epoch": 0.03546348314606742, + "grad_norm": 0.8775374293327332, + "learning_rate": 4.999999957251071e-06, + "loss": 0.8516, + "step": 101 + }, + { + "epoch": 0.03581460674157303, + "grad_norm": 0.8515768647193909, + "learning_rate": 4.999999829004281e-06, + "loss": 0.8045, + "step": 102 + }, + { + "epoch": 0.03616573033707865, + "grad_norm": 1.0148895978927612, + "learning_rate": 4.9999996152596355e-06, + "loss": 0.8247, + "step": 103 + }, + { + "epoch": 0.03651685393258427, + "grad_norm": 0.829903244972229, + "learning_rate": 4.999999316017144e-06, + "loss": 0.8282, + "step": 104 + }, + { + "epoch": 0.03686797752808989, + "grad_norm": 0.8467276096343994, + "learning_rate": 4.999998931276815e-06, + "loss": 0.8325, + "step": 105 + }, + { + "epoch": 0.037219101123595506, + "grad_norm": 0.8355855941772461, + "learning_rate": 4.999998461038661e-06, + "loss": 0.813, + "step": 106 + }, + { + "epoch": 0.037570224719101125, + "grad_norm": 0.8558890223503113, + "learning_rate": 4.999997905302699e-06, + "loss": 0.8592, + "step": 107 + }, + { + "epoch": 0.037921348314606744, + "grad_norm": 0.9446994662284851, + "learning_rate": 4.999997264068948e-06, + "loss": 0.8323, + "step": 108 + }, + { + "epoch": 0.03827247191011236, + "grad_norm": 0.8493000268936157, + "learning_rate": 4.99999653733743e-06, + "loss": 0.8178, + "step": 109 + }, + { + "epoch": 0.038623595505617975, + "grad_norm": 0.8132368922233582, + "learning_rate": 4.99999572510817e-06, + "loss": 0.7977, + "step": 110 + }, + { + "epoch": 0.038974719101123594, + "grad_norm": 0.8313632607460022, + "learning_rate": 4.999994827381196e-06, + "loss": 0.8407, + "step": 111 + }, + { + "epoch": 0.03932584269662921, + "grad_norm": 0.8673725724220276, + "learning_rate": 4.999993844156537e-06, + "loss": 0.862, + "step": 112 + }, + { + "epoch": 0.03967696629213483, + "grad_norm": 0.8794373273849487, + "learning_rate": 4.999992775434229e-06, + "loss": 0.8599, + "step": 113 + }, + { + "epoch": 0.04002808988764045, + "grad_norm": 0.8612070083618164, + "learning_rate": 4.9999916212143065e-06, + "loss": 0.8013, + "step": 114 + }, + { + "epoch": 0.04037921348314607, + "grad_norm": 0.8925568461418152, + "learning_rate": 4.9999903814968095e-06, + "loss": 0.7529, + "step": 115 + }, + { + "epoch": 0.04073033707865169, + "grad_norm": 0.8791881799697876, + "learning_rate": 4.999989056281781e-06, + "loss": 0.8756, + "step": 116 + }, + { + "epoch": 0.04108146067415731, + "grad_norm": 0.8983411192893982, + "learning_rate": 4.999987645569267e-06, + "loss": 0.8329, + "step": 117 + }, + { + "epoch": 0.04143258426966292, + "grad_norm": 0.8358251452445984, + "learning_rate": 4.999986149359314e-06, + "loss": 0.7792, + "step": 118 + }, + { + "epoch": 0.04178370786516854, + "grad_norm": 0.8848925828933716, + "learning_rate": 4.999984567651974e-06, + "loss": 0.8242, + "step": 119 + }, + { + "epoch": 0.042134831460674156, + "grad_norm": 0.9365871548652649, + "learning_rate": 4.999982900447301e-06, + "loss": 0.7875, + "step": 120 + }, + { + "epoch": 0.042485955056179775, + "grad_norm": 0.8330186605453491, + "learning_rate": 4.999981147745352e-06, + "loss": 0.8387, + "step": 121 + }, + { + "epoch": 0.042837078651685394, + "grad_norm": 0.8438044190406799, + "learning_rate": 4.999979309546187e-06, + "loss": 0.8547, + "step": 122 + }, + { + "epoch": 0.04318820224719101, + "grad_norm": 0.8812524676322937, + "learning_rate": 4.999977385849869e-06, + "loss": 0.7891, + "step": 123 + }, + { + "epoch": 0.04353932584269663, + "grad_norm": 0.8498764634132385, + "learning_rate": 4.999975376656464e-06, + "loss": 0.8259, + "step": 124 + }, + { + "epoch": 0.04389044943820225, + "grad_norm": 0.8646953105926514, + "learning_rate": 4.99997328196604e-06, + "loss": 0.8269, + "step": 125 + }, + { + "epoch": 0.04424157303370786, + "grad_norm": 0.8175535798072815, + "learning_rate": 4.9999711017786686e-06, + "loss": 0.7943, + "step": 126 + }, + { + "epoch": 0.04459269662921348, + "grad_norm": 0.8761390447616577, + "learning_rate": 4.999968836094425e-06, + "loss": 0.8668, + "step": 127 + }, + { + "epoch": 0.0449438202247191, + "grad_norm": 0.8925103545188904, + "learning_rate": 4.999966484913387e-06, + "loss": 0.8562, + "step": 128 + }, + { + "epoch": 0.04529494382022472, + "grad_norm": 0.8750078678131104, + "learning_rate": 4.999964048235634e-06, + "loss": 0.8058, + "step": 129 + }, + { + "epoch": 0.04564606741573034, + "grad_norm": 0.8446925282478333, + "learning_rate": 4.999961526061251e-06, + "loss": 0.8197, + "step": 130 + }, + { + "epoch": 0.045997191011235956, + "grad_norm": 0.8947305679321289, + "learning_rate": 4.999958918390321e-06, + "loss": 0.8347, + "step": 131 + }, + { + "epoch": 0.046348314606741575, + "grad_norm": 0.8591403961181641, + "learning_rate": 4.999956225222937e-06, + "loss": 0.8117, + "step": 132 + }, + { + "epoch": 0.046699438202247194, + "grad_norm": 0.8698204755783081, + "learning_rate": 4.999953446559188e-06, + "loss": 0.8378, + "step": 133 + }, + { + "epoch": 0.047050561797752806, + "grad_norm": 0.8451970219612122, + "learning_rate": 4.999950582399171e-06, + "loss": 0.8324, + "step": 134 + }, + { + "epoch": 0.047401685393258425, + "grad_norm": 0.8550758957862854, + "learning_rate": 4.999947632742984e-06, + "loss": 0.8291, + "step": 135 + }, + { + "epoch": 0.047752808988764044, + "grad_norm": 0.8345015645027161, + "learning_rate": 4.999944597590726e-06, + "loss": 0.8315, + "step": 136 + }, + { + "epoch": 0.04810393258426966, + "grad_norm": 0.9103791117668152, + "learning_rate": 4.999941476942502e-06, + "loss": 0.8009, + "step": 137 + }, + { + "epoch": 0.04845505617977528, + "grad_norm": 0.9898645877838135, + "learning_rate": 4.99993827079842e-06, + "loss": 0.8574, + "step": 138 + }, + { + "epoch": 0.0488061797752809, + "grad_norm": 0.8579599261283875, + "learning_rate": 4.9999349791585876e-06, + "loss": 0.7747, + "step": 139 + }, + { + "epoch": 0.04915730337078652, + "grad_norm": 0.8869126439094543, + "learning_rate": 4.999931602023118e-06, + "loss": 0.8472, + "step": 140 + }, + { + "epoch": 0.04950842696629214, + "grad_norm": 0.9130380749702454, + "learning_rate": 4.999928139392127e-06, + "loss": 0.8247, + "step": 141 + }, + { + "epoch": 0.04985955056179775, + "grad_norm": 0.8806836605072021, + "learning_rate": 4.999924591265732e-06, + "loss": 0.8658, + "step": 142 + }, + { + "epoch": 0.05021067415730337, + "grad_norm": 0.9065203666687012, + "learning_rate": 4.999920957644057e-06, + "loss": 0.8097, + "step": 143 + }, + { + "epoch": 0.05056179775280899, + "grad_norm": 0.8982106447219849, + "learning_rate": 4.999917238527223e-06, + "loss": 0.8207, + "step": 144 + }, + { + "epoch": 0.050912921348314606, + "grad_norm": 0.9152344465255737, + "learning_rate": 4.999913433915359e-06, + "loss": 0.845, + "step": 145 + }, + { + "epoch": 0.051264044943820225, + "grad_norm": 0.8733872771263123, + "learning_rate": 4.999909543808595e-06, + "loss": 0.8337, + "step": 146 + }, + { + "epoch": 0.051615168539325844, + "grad_norm": 0.8482681512832642, + "learning_rate": 4.999905568207064e-06, + "loss": 0.8226, + "step": 147 + }, + { + "epoch": 0.05196629213483146, + "grad_norm": 0.8690842986106873, + "learning_rate": 4.999901507110901e-06, + "loss": 0.8677, + "step": 148 + }, + { + "epoch": 0.05231741573033708, + "grad_norm": 0.8329693675041199, + "learning_rate": 4.999897360520246e-06, + "loss": 0.7628, + "step": 149 + }, + { + "epoch": 0.05266853932584269, + "grad_norm": 0.913270890712738, + "learning_rate": 4.999893128435241e-06, + "loss": 0.8374, + "step": 150 + }, + { + "epoch": 0.05301966292134831, + "grad_norm": 0.843010425567627, + "learning_rate": 4.999888810856029e-06, + "loss": 0.8343, + "step": 151 + }, + { + "epoch": 0.05337078651685393, + "grad_norm": 0.8601850867271423, + "learning_rate": 4.99988440778276e-06, + "loss": 0.8254, + "step": 152 + }, + { + "epoch": 0.05372191011235955, + "grad_norm": 0.8458811640739441, + "learning_rate": 4.999879919215583e-06, + "loss": 0.8238, + "step": 153 + }, + { + "epoch": 0.05407303370786517, + "grad_norm": 0.8649472594261169, + "learning_rate": 4.999875345154652e-06, + "loss": 0.8545, + "step": 154 + }, + { + "epoch": 0.05442415730337079, + "grad_norm": 0.8692646622657776, + "learning_rate": 4.999870685600123e-06, + "loss": 0.769, + "step": 155 + }, + { + "epoch": 0.054775280898876406, + "grad_norm": 0.8382206559181213, + "learning_rate": 4.999865940552157e-06, + "loss": 0.8076, + "step": 156 + }, + { + "epoch": 0.055126404494382025, + "grad_norm": 0.8909122943878174, + "learning_rate": 4.999861110010914e-06, + "loss": 0.8118, + "step": 157 + }, + { + "epoch": 0.05547752808988764, + "grad_norm": 0.8760329484939575, + "learning_rate": 4.99985619397656e-06, + "loss": 0.8625, + "step": 158 + }, + { + "epoch": 0.055828651685393256, + "grad_norm": 1.0767908096313477, + "learning_rate": 4.999851192449263e-06, + "loss": 0.8038, + "step": 159 + }, + { + "epoch": 0.056179775280898875, + "grad_norm": 0.8571433424949646, + "learning_rate": 4.999846105429196e-06, + "loss": 0.8123, + "step": 160 + }, + { + "epoch": 0.056530898876404494, + "grad_norm": 0.8204314112663269, + "learning_rate": 4.99984093291653e-06, + "loss": 0.7713, + "step": 161 + }, + { + "epoch": 0.05688202247191011, + "grad_norm": 0.8622918128967285, + "learning_rate": 4.9998356749114434e-06, + "loss": 0.7816, + "step": 162 + }, + { + "epoch": 0.05723314606741573, + "grad_norm": 0.991470456123352, + "learning_rate": 4.999830331414116e-06, + "loss": 0.8171, + "step": 163 + }, + { + "epoch": 0.05758426966292135, + "grad_norm": 0.9167494177818298, + "learning_rate": 4.99982490242473e-06, + "loss": 0.8096, + "step": 164 + }, + { + "epoch": 0.05793539325842697, + "grad_norm": 0.8794699311256409, + "learning_rate": 4.999819387943472e-06, + "loss": 0.8284, + "step": 165 + }, + { + "epoch": 0.05828651685393258, + "grad_norm": 0.9374570250511169, + "learning_rate": 4.99981378797053e-06, + "loss": 0.7887, + "step": 166 + }, + { + "epoch": 0.0586376404494382, + "grad_norm": 0.8803961277008057, + "learning_rate": 4.999808102506095e-06, + "loss": 0.8331, + "step": 167 + }, + { + "epoch": 0.05898876404494382, + "grad_norm": 0.903806209564209, + "learning_rate": 4.999802331550363e-06, + "loss": 0.8002, + "step": 168 + }, + { + "epoch": 0.05933988764044944, + "grad_norm": 0.8599410057067871, + "learning_rate": 4.99979647510353e-06, + "loss": 0.8342, + "step": 169 + }, + { + "epoch": 0.059691011235955056, + "grad_norm": 0.9048354625701904, + "learning_rate": 4.999790533165797e-06, + "loss": 0.8197, + "step": 170 + }, + { + "epoch": 0.060042134831460675, + "grad_norm": 0.8594435453414917, + "learning_rate": 4.999784505737366e-06, + "loss": 0.7994, + "step": 171 + }, + { + "epoch": 0.060393258426966294, + "grad_norm": 0.8999492526054382, + "learning_rate": 4.999778392818444e-06, + "loss": 0.8062, + "step": 172 + }, + { + "epoch": 0.06074438202247191, + "grad_norm": 0.8730378150939941, + "learning_rate": 4.999772194409241e-06, + "loss": 0.8348, + "step": 173 + }, + { + "epoch": 0.061095505617977525, + "grad_norm": 0.8853862881660461, + "learning_rate": 4.999765910509967e-06, + "loss": 0.8235, + "step": 174 + }, + { + "epoch": 0.061446629213483143, + "grad_norm": 0.8512648344039917, + "learning_rate": 4.999759541120839e-06, + "loss": 0.8254, + "step": 175 + }, + { + "epoch": 0.06179775280898876, + "grad_norm": 0.925890326499939, + "learning_rate": 4.999753086242073e-06, + "loss": 0.8498, + "step": 176 + }, + { + "epoch": 0.06214887640449438, + "grad_norm": 0.9359338283538818, + "learning_rate": 4.999746545873891e-06, + "loss": 0.8006, + "step": 177 + }, + { + "epoch": 0.0625, + "grad_norm": 0.8511932492256165, + "learning_rate": 4.999739920016516e-06, + "loss": 0.7843, + "step": 178 + }, + { + "epoch": 0.06285112359550561, + "grad_norm": 0.920088529586792, + "learning_rate": 4.999733208670174e-06, + "loss": 0.8173, + "step": 179 + }, + { + "epoch": 0.06320224719101124, + "grad_norm": 0.8991772532463074, + "learning_rate": 4.9997264118350965e-06, + "loss": 0.8227, + "step": 180 + }, + { + "epoch": 0.06355337078651685, + "grad_norm": 0.9123396873474121, + "learning_rate": 4.9997195295115145e-06, + "loss": 0.7767, + "step": 181 + }, + { + "epoch": 0.06390449438202248, + "grad_norm": 0.888533353805542, + "learning_rate": 4.9997125616996635e-06, + "loss": 0.8093, + "step": 182 + }, + { + "epoch": 0.06425561797752809, + "grad_norm": 0.9099261164665222, + "learning_rate": 4.999705508399782e-06, + "loss": 0.8142, + "step": 183 + }, + { + "epoch": 0.06460674157303371, + "grad_norm": 0.9250622987747192, + "learning_rate": 4.999698369612111e-06, + "loss": 0.7913, + "step": 184 + }, + { + "epoch": 0.06495786516853932, + "grad_norm": 0.8754624128341675, + "learning_rate": 4.999691145336895e-06, + "loss": 0.7755, + "step": 185 + }, + { + "epoch": 0.06530898876404495, + "grad_norm": 0.8488731384277344, + "learning_rate": 4.99968383557438e-06, + "loss": 0.8209, + "step": 186 + }, + { + "epoch": 0.06566011235955056, + "grad_norm": 0.8827488422393799, + "learning_rate": 4.9996764403248175e-06, + "loss": 0.8121, + "step": 187 + }, + { + "epoch": 0.06601123595505617, + "grad_norm": 0.8604486584663391, + "learning_rate": 4.99966895958846e-06, + "loss": 0.8046, + "step": 188 + }, + { + "epoch": 0.0663623595505618, + "grad_norm": 0.855309247970581, + "learning_rate": 4.999661393365562e-06, + "loss": 0.7345, + "step": 189 + }, + { + "epoch": 0.06671348314606741, + "grad_norm": 0.8943816423416138, + "learning_rate": 4.999653741656385e-06, + "loss": 0.7937, + "step": 190 + }, + { + "epoch": 0.06706460674157304, + "grad_norm": 0.9106631278991699, + "learning_rate": 4.999646004461188e-06, + "loss": 0.8157, + "step": 191 + }, + { + "epoch": 0.06741573033707865, + "grad_norm": 0.9037092924118042, + "learning_rate": 4.999638181780237e-06, + "loss": 0.8091, + "step": 192 + }, + { + "epoch": 0.06776685393258428, + "grad_norm": 0.854783833026886, + "learning_rate": 4.9996302736137994e-06, + "loss": 0.7784, + "step": 193 + }, + { + "epoch": 0.06811797752808989, + "grad_norm": 0.8792631030082703, + "learning_rate": 4.999622279962144e-06, + "loss": 0.7737, + "step": 194 + }, + { + "epoch": 0.0684691011235955, + "grad_norm": 0.8999345898628235, + "learning_rate": 4.9996142008255465e-06, + "loss": 0.7858, + "step": 195 + }, + { + "epoch": 0.06882022471910113, + "grad_norm": 0.8954832553863525, + "learning_rate": 4.999606036204282e-06, + "loss": 0.8045, + "step": 196 + }, + { + "epoch": 0.06917134831460674, + "grad_norm": 0.8854857087135315, + "learning_rate": 4.999597786098631e-06, + "loss": 0.7842, + "step": 197 + }, + { + "epoch": 0.06952247191011236, + "grad_norm": 0.8617668747901917, + "learning_rate": 4.999589450508874e-06, + "loss": 0.807, + "step": 198 + }, + { + "epoch": 0.06987359550561797, + "grad_norm": 0.8737844824790955, + "learning_rate": 4.999581029435296e-06, + "loss": 0.8268, + "step": 199 + }, + { + "epoch": 0.0702247191011236, + "grad_norm": 0.8421310782432556, + "learning_rate": 4.999572522878186e-06, + "loss": 0.7701, + "step": 200 + }, + { + "epoch": 0.07057584269662921, + "grad_norm": 0.8911473155021667, + "learning_rate": 4.9995639308378365e-06, + "loss": 0.756, + "step": 201 + }, + { + "epoch": 0.07092696629213484, + "grad_norm": 0.879984974861145, + "learning_rate": 4.999555253314538e-06, + "loss": 0.8023, + "step": 202 + }, + { + "epoch": 0.07127808988764045, + "grad_norm": 0.9726935625076294, + "learning_rate": 4.9995464903085885e-06, + "loss": 0.8017, + "step": 203 + }, + { + "epoch": 0.07162921348314606, + "grad_norm": 0.9257283210754395, + "learning_rate": 4.999537641820288e-06, + "loss": 0.8336, + "step": 204 + }, + { + "epoch": 0.07198033707865169, + "grad_norm": 0.9612401723861694, + "learning_rate": 4.9995287078499385e-06, + "loss": 0.8014, + "step": 205 + }, + { + "epoch": 0.0723314606741573, + "grad_norm": 0.8790804743766785, + "learning_rate": 4.999519688397847e-06, + "loss": 0.7592, + "step": 206 + }, + { + "epoch": 0.07268258426966293, + "grad_norm": 0.9160227179527283, + "learning_rate": 4.999510583464321e-06, + "loss": 0.7905, + "step": 207 + }, + { + "epoch": 0.07303370786516854, + "grad_norm": 0.9160066246986389, + "learning_rate": 4.999501393049671e-06, + "loss": 0.8382, + "step": 208 + }, + { + "epoch": 0.07338483146067416, + "grad_norm": 0.8971685767173767, + "learning_rate": 4.999492117154214e-06, + "loss": 0.7881, + "step": 209 + }, + { + "epoch": 0.07373595505617977, + "grad_norm": 0.9171693921089172, + "learning_rate": 4.9994827557782644e-06, + "loss": 0.8101, + "step": 210 + }, + { + "epoch": 0.07408707865168539, + "grad_norm": 0.8446071743965149, + "learning_rate": 4.999473308922143e-06, + "loss": 0.7749, + "step": 211 + }, + { + "epoch": 0.07443820224719101, + "grad_norm": 0.8867083787918091, + "learning_rate": 4.999463776586174e-06, + "loss": 0.7656, + "step": 212 + }, + { + "epoch": 0.07478932584269662, + "grad_norm": 0.8746408820152283, + "learning_rate": 4.9994541587706815e-06, + "loss": 0.7778, + "step": 213 + }, + { + "epoch": 0.07514044943820225, + "grad_norm": 0.884986162185669, + "learning_rate": 4.999444455475997e-06, + "loss": 0.7975, + "step": 214 + }, + { + "epoch": 0.07549157303370786, + "grad_norm": 0.8919047713279724, + "learning_rate": 4.9994346667024505e-06, + "loss": 0.7552, + "step": 215 + }, + { + "epoch": 0.07584269662921349, + "grad_norm": 0.8840435147285461, + "learning_rate": 4.999424792450377e-06, + "loss": 0.7876, + "step": 216 + }, + { + "epoch": 0.0761938202247191, + "grad_norm": 0.9024021625518799, + "learning_rate": 4.999414832720114e-06, + "loss": 0.7921, + "step": 217 + }, + { + "epoch": 0.07654494382022473, + "grad_norm": 0.8887689113616943, + "learning_rate": 4.999404787512003e-06, + "loss": 0.7843, + "step": 218 + }, + { + "epoch": 0.07689606741573034, + "grad_norm": 0.903285026550293, + "learning_rate": 4.9993946568263866e-06, + "loss": 0.8164, + "step": 219 + }, + { + "epoch": 0.07724719101123595, + "grad_norm": 0.8626529574394226, + "learning_rate": 4.9993844406636124e-06, + "loss": 0.7941, + "step": 220 + }, + { + "epoch": 0.07759831460674158, + "grad_norm": 0.8919528126716614, + "learning_rate": 4.999374139024028e-06, + "loss": 0.8054, + "step": 221 + }, + { + "epoch": 0.07794943820224719, + "grad_norm": 0.9159940481185913, + "learning_rate": 4.999363751907988e-06, + "loss": 0.8071, + "step": 222 + }, + { + "epoch": 0.07830056179775281, + "grad_norm": 0.8792368173599243, + "learning_rate": 4.999353279315846e-06, + "loss": 0.7949, + "step": 223 + }, + { + "epoch": 0.07865168539325842, + "grad_norm": 0.9184228181838989, + "learning_rate": 4.9993427212479604e-06, + "loss": 0.8013, + "step": 224 + }, + { + "epoch": 0.07900280898876405, + "grad_norm": 0.8925946354866028, + "learning_rate": 4.999332077704692e-06, + "loss": 0.813, + "step": 225 + }, + { + "epoch": 0.07935393258426966, + "grad_norm": 0.9273242354393005, + "learning_rate": 4.999321348686406e-06, + "loss": 0.8107, + "step": 226 + }, + { + "epoch": 0.07970505617977527, + "grad_norm": 0.8914328813552856, + "learning_rate": 4.999310534193468e-06, + "loss": 0.8229, + "step": 227 + }, + { + "epoch": 0.0800561797752809, + "grad_norm": 0.9217993021011353, + "learning_rate": 4.999299634226249e-06, + "loss": 0.7722, + "step": 228 + }, + { + "epoch": 0.08040730337078651, + "grad_norm": 0.9615241289138794, + "learning_rate": 4.999288648785121e-06, + "loss": 0.8105, + "step": 229 + }, + { + "epoch": 0.08075842696629214, + "grad_norm": 0.9098490476608276, + "learning_rate": 4.99927757787046e-06, + "loss": 0.814, + "step": 230 + }, + { + "epoch": 0.08110955056179775, + "grad_norm": 0.9211421012878418, + "learning_rate": 4.999266421482645e-06, + "loss": 0.7542, + "step": 231 + }, + { + "epoch": 0.08146067415730338, + "grad_norm": 0.841416597366333, + "learning_rate": 4.999255179622056e-06, + "loss": 0.7961, + "step": 232 + }, + { + "epoch": 0.08181179775280899, + "grad_norm": 0.8942486643791199, + "learning_rate": 4.9992438522890785e-06, + "loss": 0.8095, + "step": 233 + }, + { + "epoch": 0.08216292134831461, + "grad_norm": 0.8535089492797852, + "learning_rate": 4.9992324394841005e-06, + "loss": 0.801, + "step": 234 + }, + { + "epoch": 0.08251404494382023, + "grad_norm": 1.02827787399292, + "learning_rate": 4.999220941207511e-06, + "loss": 0.8057, + "step": 235 + }, + { + "epoch": 0.08286516853932584, + "grad_norm": 0.9189361929893494, + "learning_rate": 4.9992093574597046e-06, + "loss": 0.7682, + "step": 236 + }, + { + "epoch": 0.08321629213483146, + "grad_norm": 0.9160380959510803, + "learning_rate": 4.999197688241076e-06, + "loss": 0.8218, + "step": 237 + }, + { + "epoch": 0.08356741573033707, + "grad_norm": 0.8860989809036255, + "learning_rate": 4.999185933552027e-06, + "loss": 0.7907, + "step": 238 + }, + { + "epoch": 0.0839185393258427, + "grad_norm": 0.8413179516792297, + "learning_rate": 4.999174093392956e-06, + "loss": 0.7711, + "step": 239 + }, + { + "epoch": 0.08426966292134831, + "grad_norm": 0.8843936324119568, + "learning_rate": 4.99916216776427e-06, + "loss": 0.7356, + "step": 240 + }, + { + "epoch": 0.08462078651685394, + "grad_norm": 0.9339604377746582, + "learning_rate": 4.999150156666376e-06, + "loss": 0.8017, + "step": 241 + }, + { + "epoch": 0.08497191011235955, + "grad_norm": 0.8932393789291382, + "learning_rate": 4.999138060099685e-06, + "loss": 0.7773, + "step": 242 + }, + { + "epoch": 0.08532303370786516, + "grad_norm": 0.9117473363876343, + "learning_rate": 4.999125878064611e-06, + "loss": 0.7828, + "step": 243 + }, + { + "epoch": 0.08567415730337079, + "grad_norm": 0.8696773052215576, + "learning_rate": 4.999113610561571e-06, + "loss": 0.7681, + "step": 244 + }, + { + "epoch": 0.0860252808988764, + "grad_norm": 0.9062797427177429, + "learning_rate": 4.999101257590985e-06, + "loss": 0.7766, + "step": 245 + }, + { + "epoch": 0.08637640449438203, + "grad_norm": 0.893018364906311, + "learning_rate": 4.999088819153273e-06, + "loss": 0.735, + "step": 246 + }, + { + "epoch": 0.08672752808988764, + "grad_norm": 0.9077386856079102, + "learning_rate": 4.999076295248863e-06, + "loss": 0.8092, + "step": 247 + }, + { + "epoch": 0.08707865168539326, + "grad_norm": 0.9474624395370483, + "learning_rate": 4.9990636858781805e-06, + "loss": 0.7841, + "step": 248 + }, + { + "epoch": 0.08742977528089887, + "grad_norm": 0.8823204040527344, + "learning_rate": 4.9990509910416595e-06, + "loss": 0.7827, + "step": 249 + }, + { + "epoch": 0.0877808988764045, + "grad_norm": 0.9361265897750854, + "learning_rate": 4.999038210739733e-06, + "loss": 0.8274, + "step": 250 + }, + { + "epoch": 0.08813202247191011, + "grad_norm": 0.9128723740577698, + "learning_rate": 4.999025344972838e-06, + "loss": 0.7681, + "step": 251 + }, + { + "epoch": 0.08848314606741572, + "grad_norm": 0.8730833530426025, + "learning_rate": 4.999012393741415e-06, + "loss": 0.8086, + "step": 252 + }, + { + "epoch": 0.08883426966292135, + "grad_norm": 0.8662939071655273, + "learning_rate": 4.998999357045906e-06, + "loss": 0.7587, + "step": 253 + }, + { + "epoch": 0.08918539325842696, + "grad_norm": 0.8748407959938049, + "learning_rate": 4.998986234886758e-06, + "loss": 0.7937, + "step": 254 + }, + { + "epoch": 0.08953651685393259, + "grad_norm": 0.9770292639732361, + "learning_rate": 4.998973027264419e-06, + "loss": 0.7815, + "step": 255 + }, + { + "epoch": 0.0898876404494382, + "grad_norm": 0.8764484524726868, + "learning_rate": 4.99895973417934e-06, + "loss": 0.766, + "step": 256 + }, + { + "epoch": 0.09023876404494383, + "grad_norm": 0.8940339684486389, + "learning_rate": 4.998946355631978e-06, + "loss": 0.8166, + "step": 257 + }, + { + "epoch": 0.09058988764044944, + "grad_norm": 0.903728723526001, + "learning_rate": 4.998932891622788e-06, + "loss": 0.7801, + "step": 258 + }, + { + "epoch": 0.09094101123595505, + "grad_norm": 0.9147090315818787, + "learning_rate": 4.998919342152232e-06, + "loss": 0.814, + "step": 259 + }, + { + "epoch": 0.09129213483146068, + "grad_norm": 0.9121170043945312, + "learning_rate": 4.9989057072207725e-06, + "loss": 0.7718, + "step": 260 + }, + { + "epoch": 0.09164325842696629, + "grad_norm": 0.8610017895698547, + "learning_rate": 4.998891986828877e-06, + "loss": 0.7539, + "step": 261 + }, + { + "epoch": 0.09199438202247191, + "grad_norm": 0.9325037598609924, + "learning_rate": 4.9988781809770124e-06, + "loss": 0.7699, + "step": 262 + }, + { + "epoch": 0.09234550561797752, + "grad_norm": 0.92354416847229, + "learning_rate": 4.998864289665654e-06, + "loss": 0.8224, + "step": 263 + }, + { + "epoch": 0.09269662921348315, + "grad_norm": 0.9121785759925842, + "learning_rate": 4.998850312895274e-06, + "loss": 0.7881, + "step": 264 + }, + { + "epoch": 0.09304775280898876, + "grad_norm": 0.8718385696411133, + "learning_rate": 4.998836250666352e-06, + "loss": 0.7865, + "step": 265 + }, + { + "epoch": 0.09339887640449439, + "grad_norm": 0.8944951295852661, + "learning_rate": 4.9988221029793685e-06, + "loss": 0.8315, + "step": 266 + }, + { + "epoch": 0.09375, + "grad_norm": 0.9362013339996338, + "learning_rate": 4.998807869834807e-06, + "loss": 0.7572, + "step": 267 + }, + { + "epoch": 0.09410112359550561, + "grad_norm": 0.8949217796325684, + "learning_rate": 4.998793551233155e-06, + "loss": 0.7614, + "step": 268 + }, + { + "epoch": 0.09445224719101124, + "grad_norm": 0.9898912310600281, + "learning_rate": 4.9987791471749015e-06, + "loss": 0.8111, + "step": 269 + }, + { + "epoch": 0.09480337078651685, + "grad_norm": 0.8814307451248169, + "learning_rate": 4.9987646576605395e-06, + "loss": 0.7774, + "step": 270 + }, + { + "epoch": 0.09515449438202248, + "grad_norm": 0.9214469790458679, + "learning_rate": 4.998750082690564e-06, + "loss": 0.7944, + "step": 271 + }, + { + "epoch": 0.09550561797752809, + "grad_norm": 0.8990268707275391, + "learning_rate": 4.998735422265475e-06, + "loss": 0.7661, + "step": 272 + }, + { + "epoch": 0.09585674157303371, + "grad_norm": 0.8504504561424255, + "learning_rate": 4.998720676385772e-06, + "loss": 0.7643, + "step": 273 + }, + { + "epoch": 0.09620786516853932, + "grad_norm": 0.9197403788566589, + "learning_rate": 4.99870584505196e-06, + "loss": 0.7763, + "step": 274 + }, + { + "epoch": 0.09655898876404495, + "grad_norm": 0.9017580151557922, + "learning_rate": 4.998690928264547e-06, + "loss": 0.8137, + "step": 275 + }, + { + "epoch": 0.09691011235955056, + "grad_norm": 0.880264937877655, + "learning_rate": 4.9986759260240416e-06, + "loss": 0.8088, + "step": 276 + }, + { + "epoch": 0.09726123595505617, + "grad_norm": 0.8751420378684998, + "learning_rate": 4.998660838330958e-06, + "loss": 0.7802, + "step": 277 + }, + { + "epoch": 0.0976123595505618, + "grad_norm": 0.8619885444641113, + "learning_rate": 4.998645665185812e-06, + "loss": 0.7584, + "step": 278 + }, + { + "epoch": 0.09796348314606741, + "grad_norm": 0.8772746324539185, + "learning_rate": 4.998630406589122e-06, + "loss": 0.7739, + "step": 279 + }, + { + "epoch": 0.09831460674157304, + "grad_norm": 0.9019191265106201, + "learning_rate": 4.998615062541411e-06, + "loss": 0.806, + "step": 280 + }, + { + "epoch": 0.09866573033707865, + "grad_norm": 0.9600672721862793, + "learning_rate": 4.998599633043202e-06, + "loss": 0.7902, + "step": 281 + }, + { + "epoch": 0.09901685393258428, + "grad_norm": 0.9067400097846985, + "learning_rate": 4.9985841180950245e-06, + "loss": 0.7779, + "step": 282 + }, + { + "epoch": 0.09936797752808989, + "grad_norm": 0.8911014795303345, + "learning_rate": 4.998568517697408e-06, + "loss": 0.7605, + "step": 283 + }, + { + "epoch": 0.0997191011235955, + "grad_norm": 0.8919551968574524, + "learning_rate": 4.998552831850886e-06, + "loss": 0.7416, + "step": 284 + }, + { + "epoch": 0.10007022471910113, + "grad_norm": 0.8880231380462646, + "learning_rate": 4.998537060555995e-06, + "loss": 0.7757, + "step": 285 + }, + { + "epoch": 0.10042134831460674, + "grad_norm": 0.9274663329124451, + "learning_rate": 4.998521203813275e-06, + "loss": 0.7925, + "step": 286 + }, + { + "epoch": 0.10077247191011236, + "grad_norm": 0.9023240804672241, + "learning_rate": 4.998505261623266e-06, + "loss": 0.8081, + "step": 287 + }, + { + "epoch": 0.10112359550561797, + "grad_norm": 0.8825272917747498, + "learning_rate": 4.998489233986518e-06, + "loss": 0.7864, + "step": 288 + }, + { + "epoch": 0.1014747191011236, + "grad_norm": 0.885826826095581, + "learning_rate": 4.9984731209035745e-06, + "loss": 0.7759, + "step": 289 + }, + { + "epoch": 0.10182584269662921, + "grad_norm": 0.8802645802497864, + "learning_rate": 4.998456922374988e-06, + "loss": 0.769, + "step": 290 + }, + { + "epoch": 0.10217696629213484, + "grad_norm": 0.9090613126754761, + "learning_rate": 4.998440638401314e-06, + "loss": 0.7931, + "step": 291 + }, + { + "epoch": 0.10252808988764045, + "grad_norm": 0.9373996257781982, + "learning_rate": 4.998424268983107e-06, + "loss": 0.7443, + "step": 292 + }, + { + "epoch": 0.10287921348314606, + "grad_norm": 0.9003362655639648, + "learning_rate": 4.998407814120928e-06, + "loss": 0.7983, + "step": 293 + }, + { + "epoch": 0.10323033707865169, + "grad_norm": 0.9281851053237915, + "learning_rate": 4.998391273815341e-06, + "loss": 0.8026, + "step": 294 + }, + { + "epoch": 0.1035814606741573, + "grad_norm": 0.8824150562286377, + "learning_rate": 4.998374648066909e-06, + "loss": 0.7435, + "step": 295 + }, + { + "epoch": 0.10393258426966293, + "grad_norm": 0.9006941318511963, + "learning_rate": 4.998357936876202e-06, + "loss": 0.7567, + "step": 296 + }, + { + "epoch": 0.10428370786516854, + "grad_norm": 0.8622567057609558, + "learning_rate": 4.998341140243792e-06, + "loss": 0.7532, + "step": 297 + }, + { + "epoch": 0.10463483146067416, + "grad_norm": 0.8983448147773743, + "learning_rate": 4.9983242581702525e-06, + "loss": 0.7686, + "step": 298 + }, + { + "epoch": 0.10498595505617977, + "grad_norm": 0.8893060684204102, + "learning_rate": 4.998307290656162e-06, + "loss": 0.7697, + "step": 299 + }, + { + "epoch": 0.10533707865168539, + "grad_norm": 0.8906247615814209, + "learning_rate": 4.998290237702098e-06, + "loss": 0.7628, + "step": 300 + }, + { + "epoch": 0.10568820224719101, + "grad_norm": 0.8763146996498108, + "learning_rate": 4.998273099308647e-06, + "loss": 0.7742, + "step": 301 + }, + { + "epoch": 0.10603932584269662, + "grad_norm": 0.8954668045043945, + "learning_rate": 4.998255875476394e-06, + "loss": 0.7666, + "step": 302 + }, + { + "epoch": 0.10639044943820225, + "grad_norm": 0.9178535342216492, + "learning_rate": 4.9982385662059275e-06, + "loss": 0.7679, + "step": 303 + }, + { + "epoch": 0.10674157303370786, + "grad_norm": 0.9487611651420593, + "learning_rate": 4.99822117149784e-06, + "loss": 0.7744, + "step": 304 + }, + { + "epoch": 0.10709269662921349, + "grad_norm": 0.9466339945793152, + "learning_rate": 4.998203691352726e-06, + "loss": 0.7608, + "step": 305 + }, + { + "epoch": 0.1074438202247191, + "grad_norm": 0.9768984913825989, + "learning_rate": 4.998186125771184e-06, + "loss": 0.8028, + "step": 306 + }, + { + "epoch": 0.10779494382022473, + "grad_norm": 0.9045122861862183, + "learning_rate": 4.998168474753814e-06, + "loss": 0.7365, + "step": 307 + }, + { + "epoch": 0.10814606741573034, + "grad_norm": 0.9124319553375244, + "learning_rate": 4.99815073830122e-06, + "loss": 0.7686, + "step": 308 + }, + { + "epoch": 0.10849719101123595, + "grad_norm": 0.9172065854072571, + "learning_rate": 4.998132916414008e-06, + "loss": 0.778, + "step": 309 + }, + { + "epoch": 0.10884831460674158, + "grad_norm": 0.9291918873786926, + "learning_rate": 4.9981150090927876e-06, + "loss": 0.7828, + "step": 310 + }, + { + "epoch": 0.10919943820224719, + "grad_norm": 0.9258780479431152, + "learning_rate": 4.998097016338172e-06, + "loss": 0.7822, + "step": 311 + }, + { + "epoch": 0.10955056179775281, + "grad_norm": 0.9169499278068542, + "learning_rate": 4.998078938150776e-06, + "loss": 0.7735, + "step": 312 + }, + { + "epoch": 0.10990168539325842, + "grad_norm": 0.8983737826347351, + "learning_rate": 4.998060774531218e-06, + "loss": 0.7729, + "step": 313 + }, + { + "epoch": 0.11025280898876405, + "grad_norm": 0.949192225933075, + "learning_rate": 4.99804252548012e-06, + "loss": 0.8153, + "step": 314 + }, + { + "epoch": 0.11060393258426966, + "grad_norm": 0.9175098538398743, + "learning_rate": 4.998024190998104e-06, + "loss": 0.7784, + "step": 315 + }, + { + "epoch": 0.11095505617977527, + "grad_norm": 0.9033719897270203, + "learning_rate": 4.998005771085799e-06, + "loss": 0.748, + "step": 316 + }, + { + "epoch": 0.1113061797752809, + "grad_norm": 0.9692884683609009, + "learning_rate": 4.997987265743834e-06, + "loss": 0.7795, + "step": 317 + }, + { + "epoch": 0.11165730337078651, + "grad_norm": 0.8999868035316467, + "learning_rate": 4.997968674972842e-06, + "loss": 0.7564, + "step": 318 + }, + { + "epoch": 0.11200842696629214, + "grad_norm": 0.9252747893333435, + "learning_rate": 4.997949998773458e-06, + "loss": 0.8054, + "step": 319 + }, + { + "epoch": 0.11235955056179775, + "grad_norm": 0.8910780549049377, + "learning_rate": 4.997931237146323e-06, + "loss": 0.7671, + "step": 320 + }, + { + "epoch": 0.11271067415730338, + "grad_norm": 0.9301937818527222, + "learning_rate": 4.997912390092077e-06, + "loss": 0.7906, + "step": 321 + }, + { + "epoch": 0.11306179775280899, + "grad_norm": 0.9235373735427856, + "learning_rate": 4.997893457611364e-06, + "loss": 0.7626, + "step": 322 + }, + { + "epoch": 0.11341292134831461, + "grad_norm": 0.9015495777130127, + "learning_rate": 4.997874439704833e-06, + "loss": 0.7634, + "step": 323 + }, + { + "epoch": 0.11376404494382023, + "grad_norm": 0.8859757781028748, + "learning_rate": 4.997855336373133e-06, + "loss": 0.7435, + "step": 324 + }, + { + "epoch": 0.11411516853932584, + "grad_norm": 0.9129098057746887, + "learning_rate": 4.9978361476169176e-06, + "loss": 0.7505, + "step": 325 + }, + { + "epoch": 0.11446629213483146, + "grad_norm": 0.9024680256843567, + "learning_rate": 4.997816873436845e-06, + "loss": 0.7402, + "step": 326 + }, + { + "epoch": 0.11481741573033707, + "grad_norm": 0.9123002290725708, + "learning_rate": 4.997797513833571e-06, + "loss": 0.7751, + "step": 327 + }, + { + "epoch": 0.1151685393258427, + "grad_norm": 0.8788150548934937, + "learning_rate": 4.9977780688077606e-06, + "loss": 0.7882, + "step": 328 + }, + { + "epoch": 0.11551966292134831, + "grad_norm": 0.9502024054527283, + "learning_rate": 4.997758538360077e-06, + "loss": 0.7777, + "step": 329 + }, + { + "epoch": 0.11587078651685394, + "grad_norm": 0.9458324909210205, + "learning_rate": 4.9977389224911884e-06, + "loss": 0.7666, + "step": 330 + }, + { + "epoch": 0.11622191011235955, + "grad_norm": 0.9191945195198059, + "learning_rate": 4.997719221201766e-06, + "loss": 0.7678, + "step": 331 + }, + { + "epoch": 0.11657303370786516, + "grad_norm": 0.9368025660514832, + "learning_rate": 4.997699434492485e-06, + "loss": 0.7689, + "step": 332 + }, + { + "epoch": 0.11692415730337079, + "grad_norm": 0.9120011925697327, + "learning_rate": 4.99767956236402e-06, + "loss": 0.7489, + "step": 333 + }, + { + "epoch": 0.1172752808988764, + "grad_norm": 0.9374188780784607, + "learning_rate": 4.997659604817051e-06, + "loss": 0.7358, + "step": 334 + }, + { + "epoch": 0.11762640449438203, + "grad_norm": 0.9129819273948669, + "learning_rate": 4.997639561852261e-06, + "loss": 0.7591, + "step": 335 + }, + { + "epoch": 0.11797752808988764, + "grad_norm": 0.9249037504196167, + "learning_rate": 4.997619433470334e-06, + "loss": 0.782, + "step": 336 + }, + { + "epoch": 0.11832865168539326, + "grad_norm": 0.8872626423835754, + "learning_rate": 4.9975992196719614e-06, + "loss": 0.7477, + "step": 337 + }, + { + "epoch": 0.11867977528089887, + "grad_norm": 0.9343289732933044, + "learning_rate": 4.9975789204578315e-06, + "loss": 0.7702, + "step": 338 + }, + { + "epoch": 0.1190308988764045, + "grad_norm": 0.9379141926765442, + "learning_rate": 4.997558535828641e-06, + "loss": 0.7847, + "step": 339 + }, + { + "epoch": 0.11938202247191011, + "grad_norm": 0.9160316586494446, + "learning_rate": 4.997538065785085e-06, + "loss": 0.7658, + "step": 340 + }, + { + "epoch": 0.11973314606741572, + "grad_norm": 0.9021292328834534, + "learning_rate": 4.997517510327864e-06, + "loss": 0.8114, + "step": 341 + }, + { + "epoch": 0.12008426966292135, + "grad_norm": 0.9155046343803406, + "learning_rate": 4.997496869457681e-06, + "loss": 0.7579, + "step": 342 + }, + { + "epoch": 0.12043539325842696, + "grad_norm": 0.8777235746383667, + "learning_rate": 4.997476143175243e-06, + "loss": 0.7832, + "step": 343 + }, + { + "epoch": 0.12078651685393259, + "grad_norm": 0.8728359341621399, + "learning_rate": 4.997455331481258e-06, + "loss": 0.7647, + "step": 344 + }, + { + "epoch": 0.1211376404494382, + "grad_norm": 0.9153422713279724, + "learning_rate": 4.997434434376437e-06, + "loss": 0.7396, + "step": 345 + }, + { + "epoch": 0.12148876404494383, + "grad_norm": 0.900204598903656, + "learning_rate": 4.9974134518614964e-06, + "loss": 0.7619, + "step": 346 + }, + { + "epoch": 0.12183988764044944, + "grad_norm": 0.9725254774093628, + "learning_rate": 4.997392383937153e-06, + "loss": 0.7588, + "step": 347 + }, + { + "epoch": 0.12219101123595505, + "grad_norm": 0.9041215777397156, + "learning_rate": 4.997371230604126e-06, + "loss": 0.7712, + "step": 348 + }, + { + "epoch": 0.12254213483146068, + "grad_norm": 0.9517892599105835, + "learning_rate": 4.997349991863141e-06, + "loss": 0.7758, + "step": 349 + }, + { + "epoch": 0.12289325842696629, + "grad_norm": 0.9502705335617065, + "learning_rate": 4.997328667714922e-06, + "loss": 0.8003, + "step": 350 + }, + { + "epoch": 0.12324438202247191, + "grad_norm": 0.9965865612030029, + "learning_rate": 4.9973072581602005e-06, + "loss": 0.751, + "step": 351 + }, + { + "epoch": 0.12359550561797752, + "grad_norm": 0.9221112132072449, + "learning_rate": 4.997285763199707e-06, + "loss": 0.7679, + "step": 352 + }, + { + "epoch": 0.12394662921348315, + "grad_norm": 0.940747857093811, + "learning_rate": 4.997264182834179e-06, + "loss": 0.7863, + "step": 353 + }, + { + "epoch": 0.12429775280898876, + "grad_norm": 0.925940215587616, + "learning_rate": 4.997242517064351e-06, + "loss": 0.7711, + "step": 354 + }, + { + "epoch": 0.12464887640449439, + "grad_norm": 0.8775484561920166, + "learning_rate": 4.997220765890967e-06, + "loss": 0.7504, + "step": 355 + }, + { + "epoch": 0.125, + "grad_norm": 0.9700306057929993, + "learning_rate": 4.997198929314769e-06, + "loss": 0.7385, + "step": 356 + }, + { + "epoch": 0.12535112359550563, + "grad_norm": 0.9070924520492554, + "learning_rate": 4.997177007336505e-06, + "loss": 0.7538, + "step": 357 + }, + { + "epoch": 0.12570224719101122, + "grad_norm": 0.8817124962806702, + "learning_rate": 4.997154999956924e-06, + "loss": 0.7549, + "step": 358 + }, + { + "epoch": 0.12605337078651685, + "grad_norm": 0.9093615412712097, + "learning_rate": 4.997132907176779e-06, + "loss": 0.7621, + "step": 359 + }, + { + "epoch": 0.12640449438202248, + "grad_norm": 0.8783479332923889, + "learning_rate": 4.9971107289968256e-06, + "loss": 0.7369, + "step": 360 + }, + { + "epoch": 0.1267556179775281, + "grad_norm": 0.9039416313171387, + "learning_rate": 4.997088465417822e-06, + "loss": 0.7664, + "step": 361 + }, + { + "epoch": 0.1271067415730337, + "grad_norm": 0.9135423302650452, + "learning_rate": 4.997066116440529e-06, + "loss": 0.7887, + "step": 362 + }, + { + "epoch": 0.12745786516853932, + "grad_norm": 0.8956524133682251, + "learning_rate": 4.997043682065712e-06, + "loss": 0.7303, + "step": 363 + }, + { + "epoch": 0.12780898876404495, + "grad_norm": 0.9221348166465759, + "learning_rate": 4.997021162294138e-06, + "loss": 0.7707, + "step": 364 + }, + { + "epoch": 0.12816011235955055, + "grad_norm": 0.920876681804657, + "learning_rate": 4.996998557126577e-06, + "loss": 0.7825, + "step": 365 + }, + { + "epoch": 0.12851123595505617, + "grad_norm": 0.9243044853210449, + "learning_rate": 4.996975866563802e-06, + "loss": 0.7362, + "step": 366 + }, + { + "epoch": 0.1288623595505618, + "grad_norm": 0.89571613073349, + "learning_rate": 4.996953090606589e-06, + "loss": 0.7645, + "step": 367 + }, + { + "epoch": 0.12921348314606743, + "grad_norm": 0.9237456917762756, + "learning_rate": 4.9969302292557165e-06, + "loss": 0.766, + "step": 368 + }, + { + "epoch": 0.12956460674157302, + "grad_norm": 0.9664248824119568, + "learning_rate": 4.996907282511968e-06, + "loss": 0.7801, + "step": 369 + }, + { + "epoch": 0.12991573033707865, + "grad_norm": 0.9590004682540894, + "learning_rate": 4.9968842503761265e-06, + "loss": 0.7544, + "step": 370 + }, + { + "epoch": 0.13026685393258428, + "grad_norm": 0.9026381969451904, + "learning_rate": 4.99686113284898e-06, + "loss": 0.7756, + "step": 371 + }, + { + "epoch": 0.1306179775280899, + "grad_norm": 0.9528935551643372, + "learning_rate": 4.9968379299313195e-06, + "loss": 0.7673, + "step": 372 + }, + { + "epoch": 0.1309691011235955, + "grad_norm": 0.9700410962104797, + "learning_rate": 4.996814641623937e-06, + "loss": 0.7746, + "step": 373 + }, + { + "epoch": 0.13132022471910113, + "grad_norm": 0.931694746017456, + "learning_rate": 4.996791267927632e-06, + "loss": 0.8078, + "step": 374 + }, + { + "epoch": 0.13167134831460675, + "grad_norm": 0.9812837839126587, + "learning_rate": 4.996767808843203e-06, + "loss": 0.7662, + "step": 375 + }, + { + "epoch": 0.13202247191011235, + "grad_norm": 0.9266939163208008, + "learning_rate": 4.996744264371449e-06, + "loss": 0.7468, + "step": 376 + }, + { + "epoch": 0.13237359550561797, + "grad_norm": 0.8732329607009888, + "learning_rate": 4.996720634513179e-06, + "loss": 0.7774, + "step": 377 + }, + { + "epoch": 0.1327247191011236, + "grad_norm": 0.9104029536247253, + "learning_rate": 4.9966969192692e-06, + "loss": 0.7423, + "step": 378 + }, + { + "epoch": 0.13307584269662923, + "grad_norm": 0.8799782991409302, + "learning_rate": 4.996673118640323e-06, + "loss": 0.7086, + "step": 379 + }, + { + "epoch": 0.13342696629213482, + "grad_norm": 0.9318313598632812, + "learning_rate": 4.9966492326273605e-06, + "loss": 0.7492, + "step": 380 + }, + { + "epoch": 0.13377808988764045, + "grad_norm": 0.9577075839042664, + "learning_rate": 4.996625261231131e-06, + "loss": 0.7935, + "step": 381 + }, + { + "epoch": 0.13412921348314608, + "grad_norm": 0.9912057518959045, + "learning_rate": 4.996601204452455e-06, + "loss": 0.7515, + "step": 382 + }, + { + "epoch": 0.13448033707865167, + "grad_norm": 0.9166970252990723, + "learning_rate": 4.996577062292154e-06, + "loss": 0.7364, + "step": 383 + }, + { + "epoch": 0.1348314606741573, + "grad_norm": 0.9245953559875488, + "learning_rate": 4.996552834751053e-06, + "loss": 0.7895, + "step": 384 + }, + { + "epoch": 0.13518258426966293, + "grad_norm": 0.9169378876686096, + "learning_rate": 4.996528521829982e-06, + "loss": 0.7632, + "step": 385 + }, + { + "epoch": 0.13553370786516855, + "grad_norm": 0.8847085237503052, + "learning_rate": 4.996504123529772e-06, + "loss": 0.7607, + "step": 386 + }, + { + "epoch": 0.13588483146067415, + "grad_norm": 0.8744500875473022, + "learning_rate": 4.996479639851256e-06, + "loss": 0.7146, + "step": 387 + }, + { + "epoch": 0.13623595505617977, + "grad_norm": 0.9195184707641602, + "learning_rate": 4.996455070795274e-06, + "loss": 0.771, + "step": 388 + }, + { + "epoch": 0.1365870786516854, + "grad_norm": 0.8896746039390564, + "learning_rate": 4.996430416362664e-06, + "loss": 0.7239, + "step": 389 + }, + { + "epoch": 0.136938202247191, + "grad_norm": 0.9265782237052917, + "learning_rate": 4.99640567655427e-06, + "loss": 0.7607, + "step": 390 + }, + { + "epoch": 0.13728932584269662, + "grad_norm": 0.9531838893890381, + "learning_rate": 4.996380851370939e-06, + "loss": 0.7487, + "step": 391 + }, + { + "epoch": 0.13764044943820225, + "grad_norm": 0.9503312706947327, + "learning_rate": 4.996355940813518e-06, + "loss": 0.7889, + "step": 392 + }, + { + "epoch": 0.13799157303370788, + "grad_norm": 0.8666608929634094, + "learning_rate": 4.99633094488286e-06, + "loss": 0.7539, + "step": 393 + }, + { + "epoch": 0.13834269662921347, + "grad_norm": 0.9313355684280396, + "learning_rate": 4.996305863579821e-06, + "loss": 0.7674, + "step": 394 + }, + { + "epoch": 0.1386938202247191, + "grad_norm": 0.9345743656158447, + "learning_rate": 4.9962806969052565e-06, + "loss": 0.8057, + "step": 395 + }, + { + "epoch": 0.13904494382022473, + "grad_norm": 0.922758936882019, + "learning_rate": 4.996255444860029e-06, + "loss": 0.7404, + "step": 396 + }, + { + "epoch": 0.13939606741573032, + "grad_norm": 0.8888351321220398, + "learning_rate": 4.996230107445001e-06, + "loss": 0.753, + "step": 397 + }, + { + "epoch": 0.13974719101123595, + "grad_norm": 0.9118607640266418, + "learning_rate": 4.996204684661039e-06, + "loss": 0.7562, + "step": 398 + }, + { + "epoch": 0.14009831460674158, + "grad_norm": 0.9427488446235657, + "learning_rate": 4.996179176509013e-06, + "loss": 0.7085, + "step": 399 + }, + { + "epoch": 0.1404494382022472, + "grad_norm": 0.9284588098526001, + "learning_rate": 4.996153582989795e-06, + "loss": 0.7674, + "step": 400 + }, + { + "epoch": 0.1408005617977528, + "grad_norm": 0.8856985569000244, + "learning_rate": 4.996127904104261e-06, + "loss": 0.7357, + "step": 401 + }, + { + "epoch": 0.14115168539325842, + "grad_norm": 0.9342722296714783, + "learning_rate": 4.996102139853289e-06, + "loss": 0.7387, + "step": 402 + }, + { + "epoch": 0.14150280898876405, + "grad_norm": 0.9672777056694031, + "learning_rate": 4.996076290237759e-06, + "loss": 0.7681, + "step": 403 + }, + { + "epoch": 0.14185393258426968, + "grad_norm": 0.929104745388031, + "learning_rate": 4.996050355258556e-06, + "loss": 0.7266, + "step": 404 + }, + { + "epoch": 0.14220505617977527, + "grad_norm": 0.9371368288993835, + "learning_rate": 4.996024334916567e-06, + "loss": 0.7841, + "step": 405 + }, + { + "epoch": 0.1425561797752809, + "grad_norm": 0.8848541975021362, + "learning_rate": 4.995998229212681e-06, + "loss": 0.7659, + "step": 406 + }, + { + "epoch": 0.14290730337078653, + "grad_norm": 0.9632852673530579, + "learning_rate": 4.995972038147792e-06, + "loss": 0.7871, + "step": 407 + }, + { + "epoch": 0.14325842696629212, + "grad_norm": 0.9157865047454834, + "learning_rate": 4.9959457617227946e-06, + "loss": 0.7529, + "step": 408 + }, + { + "epoch": 0.14360955056179775, + "grad_norm": 0.9337014555931091, + "learning_rate": 4.995919399938588e-06, + "loss": 0.7427, + "step": 409 + }, + { + "epoch": 0.14396067415730338, + "grad_norm": 0.9515252113342285, + "learning_rate": 4.995892952796074e-06, + "loss": 0.7577, + "step": 410 + }, + { + "epoch": 0.144311797752809, + "grad_norm": 0.9301478266716003, + "learning_rate": 4.995866420296157e-06, + "loss": 0.7825, + "step": 411 + }, + { + "epoch": 0.1446629213483146, + "grad_norm": 1.4020230770111084, + "learning_rate": 4.995839802439745e-06, + "loss": 0.7434, + "step": 412 + }, + { + "epoch": 0.14501404494382023, + "grad_norm": 0.9150766730308533, + "learning_rate": 4.995813099227745e-06, + "loss": 0.7038, + "step": 413 + }, + { + "epoch": 0.14536516853932585, + "grad_norm": 0.9467854499816895, + "learning_rate": 4.995786310661075e-06, + "loss": 0.7721, + "step": 414 + }, + { + "epoch": 0.14571629213483145, + "grad_norm": 0.8949883580207825, + "learning_rate": 4.995759436740648e-06, + "loss": 0.7174, + "step": 415 + }, + { + "epoch": 0.14606741573033707, + "grad_norm": 0.9632487893104553, + "learning_rate": 4.995732477467383e-06, + "loss": 0.8002, + "step": 416 + }, + { + "epoch": 0.1464185393258427, + "grad_norm": 0.8924590945243835, + "learning_rate": 4.995705432842204e-06, + "loss": 0.7575, + "step": 417 + }, + { + "epoch": 0.14676966292134833, + "grad_norm": 0.8902977705001831, + "learning_rate": 4.995678302866035e-06, + "loss": 0.73, + "step": 418 + }, + { + "epoch": 0.14712078651685392, + "grad_norm": 0.9031338095664978, + "learning_rate": 4.9956510875398035e-06, + "loss": 0.7697, + "step": 419 + }, + { + "epoch": 0.14747191011235955, + "grad_norm": 0.9241436719894409, + "learning_rate": 4.99562378686444e-06, + "loss": 0.7594, + "step": 420 + }, + { + "epoch": 0.14782303370786518, + "grad_norm": 1.0044499635696411, + "learning_rate": 4.995596400840879e-06, + "loss": 0.7938, + "step": 421 + }, + { + "epoch": 0.14817415730337077, + "grad_norm": 0.9393342733383179, + "learning_rate": 4.995568929470055e-06, + "loss": 0.8142, + "step": 422 + }, + { + "epoch": 0.1485252808988764, + "grad_norm": 0.9527561664581299, + "learning_rate": 4.99554137275291e-06, + "loss": 0.7825, + "step": 423 + }, + { + "epoch": 0.14887640449438203, + "grad_norm": 0.9195716381072998, + "learning_rate": 4.995513730690386e-06, + "loss": 0.7577, + "step": 424 + }, + { + "epoch": 0.14922752808988765, + "grad_norm": 0.9349954128265381, + "learning_rate": 4.995486003283428e-06, + "loss": 0.7159, + "step": 425 + }, + { + "epoch": 0.14957865168539325, + "grad_norm": 0.922205924987793, + "learning_rate": 4.9954581905329845e-06, + "loss": 0.8017, + "step": 426 + }, + { + "epoch": 0.14992977528089887, + "grad_norm": 0.955886721611023, + "learning_rate": 4.995430292440005e-06, + "loss": 0.7813, + "step": 427 + }, + { + "epoch": 0.1502808988764045, + "grad_norm": 0.9729349613189697, + "learning_rate": 4.995402309005445e-06, + "loss": 0.7744, + "step": 428 + }, + { + "epoch": 0.1506320224719101, + "grad_norm": 0.8943726420402527, + "learning_rate": 4.995374240230262e-06, + "loss": 0.775, + "step": 429 + }, + { + "epoch": 0.15098314606741572, + "grad_norm": 0.9235265254974365, + "learning_rate": 4.995346086115416e-06, + "loss": 0.759, + "step": 430 + }, + { + "epoch": 0.15133426966292135, + "grad_norm": 0.9518915414810181, + "learning_rate": 4.995317846661868e-06, + "loss": 0.7563, + "step": 431 + }, + { + "epoch": 0.15168539325842698, + "grad_norm": 0.9186070561408997, + "learning_rate": 4.9952895218705845e-06, + "loss": 0.7551, + "step": 432 + }, + { + "epoch": 0.15203651685393257, + "grad_norm": 0.9712167978286743, + "learning_rate": 4.995261111742536e-06, + "loss": 0.7765, + "step": 433 + }, + { + "epoch": 0.1523876404494382, + "grad_norm": 0.9043035507202148, + "learning_rate": 4.995232616278691e-06, + "loss": 0.7224, + "step": 434 + }, + { + "epoch": 0.15273876404494383, + "grad_norm": 1.357176661491394, + "learning_rate": 4.995204035480027e-06, + "loss": 0.7538, + "step": 435 + }, + { + "epoch": 0.15308988764044945, + "grad_norm": 0.9199062585830688, + "learning_rate": 4.995175369347521e-06, + "loss": 0.7396, + "step": 436 + }, + { + "epoch": 0.15344101123595505, + "grad_norm": 0.9276919960975647, + "learning_rate": 4.995146617882151e-06, + "loss": 0.7626, + "step": 437 + }, + { + "epoch": 0.15379213483146068, + "grad_norm": 0.8831456303596497, + "learning_rate": 4.995117781084901e-06, + "loss": 0.7692, + "step": 438 + }, + { + "epoch": 0.1541432584269663, + "grad_norm": 0.9020641446113586, + "learning_rate": 4.99508885895676e-06, + "loss": 0.7578, + "step": 439 + }, + { + "epoch": 0.1544943820224719, + "grad_norm": 0.9140202403068542, + "learning_rate": 4.995059851498714e-06, + "loss": 0.7728, + "step": 440 + }, + { + "epoch": 0.15484550561797752, + "grad_norm": 0.9247411489486694, + "learning_rate": 4.995030758711756e-06, + "loss": 0.7397, + "step": 441 + }, + { + "epoch": 0.15519662921348315, + "grad_norm": 0.8848091959953308, + "learning_rate": 4.995001580596882e-06, + "loss": 0.6819, + "step": 442 + }, + { + "epoch": 0.15554775280898878, + "grad_norm": 1.0068150758743286, + "learning_rate": 4.994972317155088e-06, + "loss": 0.7437, + "step": 443 + }, + { + "epoch": 0.15589887640449437, + "grad_norm": 0.9932318329811096, + "learning_rate": 4.994942968387376e-06, + "loss": 0.7693, + "step": 444 + }, + { + "epoch": 0.15625, + "grad_norm": 0.920421838760376, + "learning_rate": 4.99491353429475e-06, + "loss": 0.7649, + "step": 445 + }, + { + "epoch": 0.15660112359550563, + "grad_norm": 0.9176994562149048, + "learning_rate": 4.9948840148782165e-06, + "loss": 0.7391, + "step": 446 + }, + { + "epoch": 0.15695224719101122, + "grad_norm": 0.9588678479194641, + "learning_rate": 4.994854410138783e-06, + "loss": 0.7615, + "step": 447 + }, + { + "epoch": 0.15730337078651685, + "grad_norm": 0.9149226546287537, + "learning_rate": 4.9948247200774645e-06, + "loss": 0.75, + "step": 448 + }, + { + "epoch": 0.15765449438202248, + "grad_norm": 0.8968827724456787, + "learning_rate": 4.994794944695275e-06, + "loss": 0.7582, + "step": 449 + }, + { + "epoch": 0.1580056179775281, + "grad_norm": 0.8770960569381714, + "learning_rate": 4.994765083993234e-06, + "loss": 0.7121, + "step": 450 + }, + { + "epoch": 0.1583567415730337, + "grad_norm": 0.9019699692726135, + "learning_rate": 4.9947351379723605e-06, + "loss": 0.7131, + "step": 451 + }, + { + "epoch": 0.15870786516853932, + "grad_norm": 0.8884302973747253, + "learning_rate": 4.994705106633682e-06, + "loss": 0.7292, + "step": 452 + }, + { + "epoch": 0.15905898876404495, + "grad_norm": 0.9590429663658142, + "learning_rate": 4.994674989978222e-06, + "loss": 0.7454, + "step": 453 + }, + { + "epoch": 0.15941011235955055, + "grad_norm": 0.8948072195053101, + "learning_rate": 4.9946447880070124e-06, + "loss": 0.7667, + "step": 454 + }, + { + "epoch": 0.15976123595505617, + "grad_norm": 0.9360845685005188, + "learning_rate": 4.994614500721086e-06, + "loss": 0.7633, + "step": 455 + }, + { + "epoch": 0.1601123595505618, + "grad_norm": 0.9061880111694336, + "learning_rate": 4.9945841281214785e-06, + "loss": 0.7026, + "step": 456 + }, + { + "epoch": 0.16046348314606743, + "grad_norm": 0.918328583240509, + "learning_rate": 4.994553670209228e-06, + "loss": 0.7367, + "step": 457 + }, + { + "epoch": 0.16081460674157302, + "grad_norm": 0.9201732277870178, + "learning_rate": 4.994523126985377e-06, + "loss": 0.7326, + "step": 458 + }, + { + "epoch": 0.16116573033707865, + "grad_norm": 0.9703920483589172, + "learning_rate": 4.9944924984509695e-06, + "loss": 0.7638, + "step": 459 + }, + { + "epoch": 0.16151685393258428, + "grad_norm": 0.8904596567153931, + "learning_rate": 4.994461784607053e-06, + "loss": 0.7717, + "step": 460 + }, + { + "epoch": 0.1618679775280899, + "grad_norm": 0.9822039604187012, + "learning_rate": 4.994430985454678e-06, + "loss": 0.7674, + "step": 461 + }, + { + "epoch": 0.1622191011235955, + "grad_norm": 0.9268808364868164, + "learning_rate": 4.9944001009948986e-06, + "loss": 0.7725, + "step": 462 + }, + { + "epoch": 0.16257022471910113, + "grad_norm": 0.8941001296043396, + "learning_rate": 4.994369131228769e-06, + "loss": 0.7361, + "step": 463 + }, + { + "epoch": 0.16292134831460675, + "grad_norm": 0.8854597210884094, + "learning_rate": 4.99433807615735e-06, + "loss": 0.7426, + "step": 464 + }, + { + "epoch": 0.16327247191011235, + "grad_norm": 1.0173115730285645, + "learning_rate": 4.994306935781704e-06, + "loss": 0.7724, + "step": 465 + }, + { + "epoch": 0.16362359550561797, + "grad_norm": 0.9098212122917175, + "learning_rate": 4.994275710102894e-06, + "loss": 0.6989, + "step": 466 + }, + { + "epoch": 0.1639747191011236, + "grad_norm": 0.9020840525627136, + "learning_rate": 4.99424439912199e-06, + "loss": 0.7312, + "step": 467 + }, + { + "epoch": 0.16432584269662923, + "grad_norm": 0.9145157337188721, + "learning_rate": 4.994213002840062e-06, + "loss": 0.7334, + "step": 468 + }, + { + "epoch": 0.16467696629213482, + "grad_norm": 0.9347027540206909, + "learning_rate": 4.994181521258183e-06, + "loss": 0.7508, + "step": 469 + }, + { + "epoch": 0.16502808988764045, + "grad_norm": 0.9202942252159119, + "learning_rate": 4.994149954377431e-06, + "loss": 0.7351, + "step": 470 + }, + { + "epoch": 0.16537921348314608, + "grad_norm": 0.9013196229934692, + "learning_rate": 4.994118302198884e-06, + "loss": 0.7618, + "step": 471 + }, + { + "epoch": 0.16573033707865167, + "grad_norm": 0.9281770586967468, + "learning_rate": 4.9940865647236255e-06, + "loss": 0.7661, + "step": 472 + }, + { + "epoch": 0.1660814606741573, + "grad_norm": 0.9201061129570007, + "learning_rate": 4.99405474195274e-06, + "loss": 0.7203, + "step": 473 + }, + { + "epoch": 0.16643258426966293, + "grad_norm": 0.9211746454238892, + "learning_rate": 4.994022833887318e-06, + "loss": 0.724, + "step": 474 + }, + { + "epoch": 0.16678370786516855, + "grad_norm": 0.9289520978927612, + "learning_rate": 4.993990840528448e-06, + "loss": 0.7907, + "step": 475 + }, + { + "epoch": 0.16713483146067415, + "grad_norm": 0.9343749284744263, + "learning_rate": 4.993958761877227e-06, + "loss": 0.7707, + "step": 476 + }, + { + "epoch": 0.16748595505617977, + "grad_norm": 0.9140953421592712, + "learning_rate": 4.9939265979347495e-06, + "loss": 0.7558, + "step": 477 + }, + { + "epoch": 0.1678370786516854, + "grad_norm": 0.8889614939689636, + "learning_rate": 4.993894348702117e-06, + "loss": 0.7615, + "step": 478 + }, + { + "epoch": 0.168188202247191, + "grad_norm": 0.8894888162612915, + "learning_rate": 4.993862014180431e-06, + "loss": 0.7536, + "step": 479 + }, + { + "epoch": 0.16853932584269662, + "grad_norm": 0.8904688954353333, + "learning_rate": 4.993829594370798e-06, + "loss": 0.7471, + "step": 480 + }, + { + "epoch": 0.16889044943820225, + "grad_norm": 0.9064297080039978, + "learning_rate": 4.993797089274327e-06, + "loss": 0.7524, + "step": 481 + }, + { + "epoch": 0.16924157303370788, + "grad_norm": 0.8890877962112427, + "learning_rate": 4.99376449889213e-06, + "loss": 0.7727, + "step": 482 + }, + { + "epoch": 0.16959269662921347, + "grad_norm": 0.9089619517326355, + "learning_rate": 4.993731823225321e-06, + "loss": 0.7661, + "step": 483 + }, + { + "epoch": 0.1699438202247191, + "grad_norm": 0.9284274578094482, + "learning_rate": 4.993699062275017e-06, + "loss": 0.7337, + "step": 484 + }, + { + "epoch": 0.17029494382022473, + "grad_norm": 0.9275715947151184, + "learning_rate": 4.99366621604234e-06, + "loss": 0.7417, + "step": 485 + }, + { + "epoch": 0.17064606741573032, + "grad_norm": 0.9078076481819153, + "learning_rate": 4.9936332845284105e-06, + "loss": 0.7366, + "step": 486 + }, + { + "epoch": 0.17099719101123595, + "grad_norm": 0.9106430411338806, + "learning_rate": 4.993600267734359e-06, + "loss": 0.7738, + "step": 487 + }, + { + "epoch": 0.17134831460674158, + "grad_norm": 1.0632219314575195, + "learning_rate": 4.99356716566131e-06, + "loss": 0.724, + "step": 488 + }, + { + "epoch": 0.1716994382022472, + "grad_norm": 0.9347037076950073, + "learning_rate": 4.9935339783103985e-06, + "loss": 0.7273, + "step": 489 + }, + { + "epoch": 0.1720505617977528, + "grad_norm": 0.9145859479904175, + "learning_rate": 4.9935007056827586e-06, + "loss": 0.7603, + "step": 490 + }, + { + "epoch": 0.17240168539325842, + "grad_norm": 0.924089252948761, + "learning_rate": 4.993467347779529e-06, + "loss": 0.7289, + "step": 491 + }, + { + "epoch": 0.17275280898876405, + "grad_norm": 0.9254995584487915, + "learning_rate": 4.993433904601849e-06, + "loss": 0.7614, + "step": 492 + }, + { + "epoch": 0.17310393258426968, + "grad_norm": 0.9493343234062195, + "learning_rate": 4.993400376150863e-06, + "loss": 0.7889, + "step": 493 + }, + { + "epoch": 0.17345505617977527, + "grad_norm": 0.8984596729278564, + "learning_rate": 4.9933667624277184e-06, + "loss": 0.7304, + "step": 494 + }, + { + "epoch": 0.1738061797752809, + "grad_norm": 0.9199086427688599, + "learning_rate": 4.9933330634335634e-06, + "loss": 0.7386, + "step": 495 + }, + { + "epoch": 0.17415730337078653, + "grad_norm": 0.9047686457633972, + "learning_rate": 4.993299279169552e-06, + "loss": 0.7469, + "step": 496 + }, + { + "epoch": 0.17450842696629212, + "grad_norm": 0.9360693097114563, + "learning_rate": 4.993265409636839e-06, + "loss": 0.7595, + "step": 497 + }, + { + "epoch": 0.17485955056179775, + "grad_norm": 0.9289848804473877, + "learning_rate": 4.993231454836581e-06, + "loss": 0.7701, + "step": 498 + }, + { + "epoch": 0.17521067415730338, + "grad_norm": 0.8970044255256653, + "learning_rate": 4.993197414769942e-06, + "loss": 0.7191, + "step": 499 + }, + { + "epoch": 0.175561797752809, + "grad_norm": 0.92835932970047, + "learning_rate": 4.993163289438085e-06, + "loss": 0.7395, + "step": 500 + }, + { + "epoch": 0.1759129213483146, + "grad_norm": 0.9345782995223999, + "learning_rate": 4.993129078842176e-06, + "loss": 0.8146, + "step": 501 + }, + { + "epoch": 0.17626404494382023, + "grad_norm": 0.9650789499282837, + "learning_rate": 4.993094782983386e-06, + "loss": 0.7634, + "step": 502 + }, + { + "epoch": 0.17661516853932585, + "grad_norm": 0.9201986193656921, + "learning_rate": 4.993060401862888e-06, + "loss": 0.7476, + "step": 503 + }, + { + "epoch": 0.17696629213483145, + "grad_norm": 0.964146077632904, + "learning_rate": 4.993025935481858e-06, + "loss": 0.7822, + "step": 504 + }, + { + "epoch": 0.17731741573033707, + "grad_norm": 0.8705427646636963, + "learning_rate": 4.992991383841475e-06, + "loss": 0.6773, + "step": 505 + }, + { + "epoch": 0.1776685393258427, + "grad_norm": 0.9599820971488953, + "learning_rate": 4.992956746942919e-06, + "loss": 0.7516, + "step": 506 + }, + { + "epoch": 0.17801966292134833, + "grad_norm": 0.9984298944473267, + "learning_rate": 4.992922024787374e-06, + "loss": 0.7714, + "step": 507 + }, + { + "epoch": 0.17837078651685392, + "grad_norm": 0.9363481402397156, + "learning_rate": 4.992887217376032e-06, + "loss": 0.784, + "step": 508 + }, + { + "epoch": 0.17872191011235955, + "grad_norm": 0.8900370597839355, + "learning_rate": 4.992852324710078e-06, + "loss": 0.7749, + "step": 509 + }, + { + "epoch": 0.17907303370786518, + "grad_norm": 0.9664030075073242, + "learning_rate": 4.9928173467907085e-06, + "loss": 0.7635, + "step": 510 + }, + { + "epoch": 0.17942415730337077, + "grad_norm": 0.9362422227859497, + "learning_rate": 4.9927822836191185e-06, + "loss": 0.7497, + "step": 511 + }, + { + "epoch": 0.1797752808988764, + "grad_norm": 0.8915205001831055, + "learning_rate": 4.992747135196508e-06, + "loss": 0.7257, + "step": 512 + }, + { + "epoch": 0.18012640449438203, + "grad_norm": 0.9136397838592529, + "learning_rate": 4.992711901524079e-06, + "loss": 0.733, + "step": 513 + }, + { + "epoch": 0.18047752808988765, + "grad_norm": 0.9412914514541626, + "learning_rate": 4.992676582603035e-06, + "loss": 0.7757, + "step": 514 + }, + { + "epoch": 0.18082865168539325, + "grad_norm": 0.8885975480079651, + "learning_rate": 4.992641178434586e-06, + "loss": 0.7219, + "step": 515 + }, + { + "epoch": 0.18117977528089887, + "grad_norm": 0.950423538684845, + "learning_rate": 4.992605689019941e-06, + "loss": 0.7596, + "step": 516 + }, + { + "epoch": 0.1815308988764045, + "grad_norm": 0.9234071969985962, + "learning_rate": 4.992570114360314e-06, + "loss": 0.7624, + "step": 517 + }, + { + "epoch": 0.1818820224719101, + "grad_norm": 0.9259780049324036, + "learning_rate": 4.9925344544569225e-06, + "loss": 0.7755, + "step": 518 + }, + { + "epoch": 0.18223314606741572, + "grad_norm": 0.9503250122070312, + "learning_rate": 4.992498709310986e-06, + "loss": 0.7051, + "step": 519 + }, + { + "epoch": 0.18258426966292135, + "grad_norm": 0.9408488273620605, + "learning_rate": 4.992462878923725e-06, + "loss": 0.7281, + "step": 520 + }, + { + "epoch": 0.18293539325842698, + "grad_norm": 0.9426605105400085, + "learning_rate": 4.992426963296369e-06, + "loss": 0.7666, + "step": 521 + }, + { + "epoch": 0.18328651685393257, + "grad_norm": 0.9386591911315918, + "learning_rate": 4.992390962430142e-06, + "loss": 0.7352, + "step": 522 + }, + { + "epoch": 0.1836376404494382, + "grad_norm": 0.9154810309410095, + "learning_rate": 4.992354876326276e-06, + "loss": 0.7597, + "step": 523 + }, + { + "epoch": 0.18398876404494383, + "grad_norm": 0.936019241809845, + "learning_rate": 4.992318704986007e-06, + "loss": 0.7497, + "step": 524 + }, + { + "epoch": 0.18433988764044945, + "grad_norm": 0.9355201721191406, + "learning_rate": 4.99228244841057e-06, + "loss": 0.7868, + "step": 525 + }, + { + "epoch": 0.18469101123595505, + "grad_norm": 0.9210653305053711, + "learning_rate": 4.9922461066012075e-06, + "loss": 0.6998, + "step": 526 + }, + { + "epoch": 0.18504213483146068, + "grad_norm": 0.9298537969589233, + "learning_rate": 4.99220967955916e-06, + "loss": 0.7253, + "step": 527 + }, + { + "epoch": 0.1853932584269663, + "grad_norm": 0.9049412608146667, + "learning_rate": 4.9921731672856746e-06, + "loss": 0.7621, + "step": 528 + }, + { + "epoch": 0.1857443820224719, + "grad_norm": 0.9317775964736938, + "learning_rate": 4.992136569781999e-06, + "loss": 0.7393, + "step": 529 + }, + { + "epoch": 0.18609550561797752, + "grad_norm": 0.923850417137146, + "learning_rate": 4.9920998870493856e-06, + "loss": 0.7384, + "step": 530 + }, + { + "epoch": 0.18644662921348315, + "grad_norm": 0.9435550570487976, + "learning_rate": 4.992063119089088e-06, + "loss": 0.7811, + "step": 531 + }, + { + "epoch": 0.18679775280898878, + "grad_norm": 0.9249152541160583, + "learning_rate": 4.992026265902364e-06, + "loss": 0.7451, + "step": 532 + }, + { + "epoch": 0.18714887640449437, + "grad_norm": 0.9371122717857361, + "learning_rate": 4.991989327490475e-06, + "loss": 0.7507, + "step": 533 + }, + { + "epoch": 0.1875, + "grad_norm": 0.8902071714401245, + "learning_rate": 4.991952303854683e-06, + "loss": 0.7445, + "step": 534 + }, + { + "epoch": 0.18785112359550563, + "grad_norm": 0.8958204984664917, + "learning_rate": 4.991915194996254e-06, + "loss": 0.7029, + "step": 535 + }, + { + "epoch": 0.18820224719101122, + "grad_norm": 0.9265782237052917, + "learning_rate": 4.991878000916459e-06, + "loss": 0.7421, + "step": 536 + }, + { + "epoch": 0.18855337078651685, + "grad_norm": 0.8842232823371887, + "learning_rate": 4.991840721616568e-06, + "loss": 0.7678, + "step": 537 + }, + { + "epoch": 0.18890449438202248, + "grad_norm": 0.9105944037437439, + "learning_rate": 4.9918033570978565e-06, + "loss": 0.7257, + "step": 538 + }, + { + "epoch": 0.1892556179775281, + "grad_norm": 0.875582754611969, + "learning_rate": 4.991765907361603e-06, + "loss": 0.739, + "step": 539 + }, + { + "epoch": 0.1896067415730337, + "grad_norm": 0.90901780128479, + "learning_rate": 4.991728372409087e-06, + "loss": 0.7467, + "step": 540 + }, + { + "epoch": 0.18995786516853932, + "grad_norm": 0.9725927710533142, + "learning_rate": 4.991690752241594e-06, + "loss": 0.7779, + "step": 541 + }, + { + "epoch": 0.19030898876404495, + "grad_norm": 0.9324540495872498, + "learning_rate": 4.991653046860408e-06, + "loss": 0.7337, + "step": 542 + }, + { + "epoch": 0.19066011235955055, + "grad_norm": 0.929425060749054, + "learning_rate": 4.9916152562668205e-06, + "loss": 0.7432, + "step": 543 + }, + { + "epoch": 0.19101123595505617, + "grad_norm": 0.9117162227630615, + "learning_rate": 4.991577380462124e-06, + "loss": 0.7556, + "step": 544 + }, + { + "epoch": 0.1913623595505618, + "grad_norm": 0.9356319904327393, + "learning_rate": 4.9915394194476115e-06, + "loss": 0.783, + "step": 545 + }, + { + "epoch": 0.19171348314606743, + "grad_norm": 0.9007960557937622, + "learning_rate": 4.991501373224584e-06, + "loss": 0.7485, + "step": 546 + }, + { + "epoch": 0.19206460674157302, + "grad_norm": 0.9083287715911865, + "learning_rate": 4.991463241794342e-06, + "loss": 0.7497, + "step": 547 + }, + { + "epoch": 0.19241573033707865, + "grad_norm": 0.9226256608963013, + "learning_rate": 4.991425025158188e-06, + "loss": 0.7687, + "step": 548 + }, + { + "epoch": 0.19276685393258428, + "grad_norm": 0.8791014552116394, + "learning_rate": 4.991386723317432e-06, + "loss": 0.7542, + "step": 549 + }, + { + "epoch": 0.1931179775280899, + "grad_norm": 0.8970210552215576, + "learning_rate": 4.991348336273381e-06, + "loss": 0.7169, + "step": 550 + }, + { + "epoch": 0.1934691011235955, + "grad_norm": 0.9191828966140747, + "learning_rate": 4.9913098640273475e-06, + "loss": 0.7261, + "step": 551 + }, + { + "epoch": 0.19382022471910113, + "grad_norm": 0.9275765419006348, + "learning_rate": 4.991271306580649e-06, + "loss": 0.7469, + "step": 552 + }, + { + "epoch": 0.19417134831460675, + "grad_norm": 0.9179903864860535, + "learning_rate": 4.9912326639346045e-06, + "loss": 0.7385, + "step": 553 + }, + { + "epoch": 0.19452247191011235, + "grad_norm": 0.9225823283195496, + "learning_rate": 4.991193936090535e-06, + "loss": 0.6986, + "step": 554 + }, + { + "epoch": 0.19487359550561797, + "grad_norm": 0.9241051077842712, + "learning_rate": 4.991155123049764e-06, + "loss": 0.783, + "step": 555 + }, + { + "epoch": 0.1952247191011236, + "grad_norm": 0.9368206858634949, + "learning_rate": 4.991116224813619e-06, + "loss": 0.7712, + "step": 556 + }, + { + "epoch": 0.19557584269662923, + "grad_norm": 0.9148329496383667, + "learning_rate": 4.991077241383431e-06, + "loss": 0.8179, + "step": 557 + }, + { + "epoch": 0.19592696629213482, + "grad_norm": 0.8920343518257141, + "learning_rate": 4.991038172760533e-06, + "loss": 0.7271, + "step": 558 + }, + { + "epoch": 0.19627808988764045, + "grad_norm": 0.9275470972061157, + "learning_rate": 4.990999018946262e-06, + "loss": 0.7704, + "step": 559 + }, + { + "epoch": 0.19662921348314608, + "grad_norm": 0.9231950640678406, + "learning_rate": 4.990959779941955e-06, + "loss": 0.6871, + "step": 560 + }, + { + "epoch": 0.19698033707865167, + "grad_norm": 0.9179824590682983, + "learning_rate": 4.990920455748955e-06, + "loss": 0.7573, + "step": 561 + }, + { + "epoch": 0.1973314606741573, + "grad_norm": 0.9076899290084839, + "learning_rate": 4.9908810463686076e-06, + "loss": 0.7463, + "step": 562 + }, + { + "epoch": 0.19768258426966293, + "grad_norm": 0.9685149788856506, + "learning_rate": 4.99084155180226e-06, + "loss": 0.7207, + "step": 563 + }, + { + "epoch": 0.19803370786516855, + "grad_norm": 1.3696738481521606, + "learning_rate": 4.990801972051262e-06, + "loss": 0.768, + "step": 564 + }, + { + "epoch": 0.19838483146067415, + "grad_norm": 0.9174782633781433, + "learning_rate": 4.990762307116969e-06, + "loss": 0.7711, + "step": 565 + }, + { + "epoch": 0.19873595505617977, + "grad_norm": 0.9449765086174011, + "learning_rate": 4.990722557000736e-06, + "loss": 0.7267, + "step": 566 + }, + { + "epoch": 0.1990870786516854, + "grad_norm": 0.9291553497314453, + "learning_rate": 4.990682721703922e-06, + "loss": 0.7017, + "step": 567 + }, + { + "epoch": 0.199438202247191, + "grad_norm": 0.920346736907959, + "learning_rate": 4.9906428012278915e-06, + "loss": 0.7486, + "step": 568 + }, + { + "epoch": 0.19978932584269662, + "grad_norm": 0.9206670522689819, + "learning_rate": 4.990602795574007e-06, + "loss": 0.7833, + "step": 569 + }, + { + "epoch": 0.20014044943820225, + "grad_norm": 0.9318204522132874, + "learning_rate": 4.99056270474364e-06, + "loss": 0.7467, + "step": 570 + }, + { + "epoch": 0.20049157303370788, + "grad_norm": 0.9259383082389832, + "learning_rate": 4.990522528738159e-06, + "loss": 0.7196, + "step": 571 + }, + { + "epoch": 0.20084269662921347, + "grad_norm": 0.9336618185043335, + "learning_rate": 4.9904822675589385e-06, + "loss": 0.7498, + "step": 572 + }, + { + "epoch": 0.2011938202247191, + "grad_norm": 0.9120561480522156, + "learning_rate": 4.990441921207356e-06, + "loss": 0.7621, + "step": 573 + }, + { + "epoch": 0.20154494382022473, + "grad_norm": 0.9047495126724243, + "learning_rate": 4.99040148968479e-06, + "loss": 0.7103, + "step": 574 + }, + { + "epoch": 0.20189606741573032, + "grad_norm": 0.9809283018112183, + "learning_rate": 4.990360972992625e-06, + "loss": 0.7341, + "step": 575 + }, + { + "epoch": 0.20224719101123595, + "grad_norm": 1.0163681507110596, + "learning_rate": 4.990320371132245e-06, + "loss": 0.7219, + "step": 576 + }, + { + "epoch": 0.20259831460674158, + "grad_norm": 0.8804516792297363, + "learning_rate": 4.99027968410504e-06, + "loss": 0.718, + "step": 577 + }, + { + "epoch": 0.2029494382022472, + "grad_norm": 0.9172428846359253, + "learning_rate": 4.9902389119124005e-06, + "loss": 0.7361, + "step": 578 + }, + { + "epoch": 0.2033005617977528, + "grad_norm": 0.9390194416046143, + "learning_rate": 4.990198054555721e-06, + "loss": 0.7195, + "step": 579 + }, + { + "epoch": 0.20365168539325842, + "grad_norm": 0.9496874213218689, + "learning_rate": 4.990157112036399e-06, + "loss": 0.7305, + "step": 580 + }, + { + "epoch": 0.20400280898876405, + "grad_norm": 0.9417531490325928, + "learning_rate": 4.990116084355835e-06, + "loss": 0.7647, + "step": 581 + }, + { + "epoch": 0.20435393258426968, + "grad_norm": 0.8786060810089111, + "learning_rate": 4.99007497151543e-06, + "loss": 0.7654, + "step": 582 + }, + { + "epoch": 0.20470505617977527, + "grad_norm": 0.8952715992927551, + "learning_rate": 4.990033773516594e-06, + "loss": 0.7149, + "step": 583 + }, + { + "epoch": 0.2050561797752809, + "grad_norm": 0.9071550369262695, + "learning_rate": 4.989992490360734e-06, + "loss": 0.7368, + "step": 584 + }, + { + "epoch": 0.20540730337078653, + "grad_norm": 0.8937875628471375, + "learning_rate": 4.989951122049259e-06, + "loss": 0.7469, + "step": 585 + }, + { + "epoch": 0.20575842696629212, + "grad_norm": 0.9116646647453308, + "learning_rate": 4.989909668583588e-06, + "loss": 0.7348, + "step": 586 + }, + { + "epoch": 0.20610955056179775, + "grad_norm": 0.9020063877105713, + "learning_rate": 4.989868129965137e-06, + "loss": 0.6871, + "step": 587 + }, + { + "epoch": 0.20646067415730338, + "grad_norm": 0.8991970419883728, + "learning_rate": 4.989826506195326e-06, + "loss": 0.7136, + "step": 588 + }, + { + "epoch": 0.206811797752809, + "grad_norm": 0.9156064987182617, + "learning_rate": 4.9897847972755795e-06, + "loss": 0.7082, + "step": 589 + }, + { + "epoch": 0.2071629213483146, + "grad_norm": 0.9545648097991943, + "learning_rate": 4.989743003207323e-06, + "loss": 0.7628, + "step": 590 + }, + { + "epoch": 0.20751404494382023, + "grad_norm": 0.9077872633934021, + "learning_rate": 4.989701123991987e-06, + "loss": 0.7595, + "step": 591 + }, + { + "epoch": 0.20786516853932585, + "grad_norm": 0.9414958953857422, + "learning_rate": 4.989659159631003e-06, + "loss": 0.7604, + "step": 592 + }, + { + "epoch": 0.20821629213483145, + "grad_norm": 0.9313225150108337, + "learning_rate": 4.989617110125805e-06, + "loss": 0.7175, + "step": 593 + }, + { + "epoch": 0.20856741573033707, + "grad_norm": 0.9247497320175171, + "learning_rate": 4.989574975477833e-06, + "loss": 0.7497, + "step": 594 + }, + { + "epoch": 0.2089185393258427, + "grad_norm": 0.8851250410079956, + "learning_rate": 4.989532755688527e-06, + "loss": 0.665, + "step": 595 + }, + { + "epoch": 0.20926966292134833, + "grad_norm": 0.9276209473609924, + "learning_rate": 4.989490450759331e-06, + "loss": 0.6919, + "step": 596 + }, + { + "epoch": 0.20962078651685392, + "grad_norm": 0.9292905330657959, + "learning_rate": 4.9894480606916925e-06, + "loss": 0.7545, + "step": 597 + }, + { + "epoch": 0.20997191011235955, + "grad_norm": 0.9364456534385681, + "learning_rate": 4.98940558548706e-06, + "loss": 0.8026, + "step": 598 + }, + { + "epoch": 0.21032303370786518, + "grad_norm": 0.9156911969184875, + "learning_rate": 4.9893630251468874e-06, + "loss": 0.7279, + "step": 599 + }, + { + "epoch": 0.21067415730337077, + "grad_norm": 0.9316357374191284, + "learning_rate": 4.989320379672629e-06, + "loss": 0.7664, + "step": 600 + }, + { + "epoch": 0.2110252808988764, + "grad_norm": 0.945326566696167, + "learning_rate": 4.989277649065744e-06, + "loss": 0.7456, + "step": 601 + }, + { + "epoch": 0.21137640449438203, + "grad_norm": 0.9577007293701172, + "learning_rate": 4.989234833327693e-06, + "loss": 0.7097, + "step": 602 + }, + { + "epoch": 0.21172752808988765, + "grad_norm": 0.898679792881012, + "learning_rate": 4.989191932459941e-06, + "loss": 0.7085, + "step": 603 + }, + { + "epoch": 0.21207865168539325, + "grad_norm": 0.9334394931793213, + "learning_rate": 4.989148946463955e-06, + "loss": 0.7441, + "step": 604 + }, + { + "epoch": 0.21242977528089887, + "grad_norm": 0.9490911364555359, + "learning_rate": 4.989105875341206e-06, + "loss": 0.7571, + "step": 605 + }, + { + "epoch": 0.2127808988764045, + "grad_norm": 0.9073877930641174, + "learning_rate": 4.989062719093164e-06, + "loss": 0.748, + "step": 606 + }, + { + "epoch": 0.2131320224719101, + "grad_norm": 0.9007843136787415, + "learning_rate": 4.989019477721309e-06, + "loss": 0.7099, + "step": 607 + }, + { + "epoch": 0.21348314606741572, + "grad_norm": 0.9324343204498291, + "learning_rate": 4.988976151227116e-06, + "loss": 0.7688, + "step": 608 + }, + { + "epoch": 0.21383426966292135, + "grad_norm": 0.9213019013404846, + "learning_rate": 4.988932739612069e-06, + "loss": 0.7302, + "step": 609 + }, + { + "epoch": 0.21418539325842698, + "grad_norm": 0.9618836641311646, + "learning_rate": 4.988889242877653e-06, + "loss": 0.7484, + "step": 610 + }, + { + "epoch": 0.21453651685393257, + "grad_norm": 0.9354860186576843, + "learning_rate": 4.988845661025353e-06, + "loss": 0.7192, + "step": 611 + }, + { + "epoch": 0.2148876404494382, + "grad_norm": 0.9579604268074036, + "learning_rate": 4.988801994056663e-06, + "loss": 0.7727, + "step": 612 + }, + { + "epoch": 0.21523876404494383, + "grad_norm": 0.8972249627113342, + "learning_rate": 4.988758241973073e-06, + "loss": 0.7816, + "step": 613 + }, + { + "epoch": 0.21558988764044945, + "grad_norm": 0.9025394916534424, + "learning_rate": 4.988714404776082e-06, + "loss": 0.6994, + "step": 614 + }, + { + "epoch": 0.21594101123595505, + "grad_norm": 0.9159374833106995, + "learning_rate": 4.9886704824671874e-06, + "loss": 0.7319, + "step": 615 + }, + { + "epoch": 0.21629213483146068, + "grad_norm": 0.9703385829925537, + "learning_rate": 4.988626475047892e-06, + "loss": 0.7739, + "step": 616 + }, + { + "epoch": 0.2166432584269663, + "grad_norm": 0.9332022070884705, + "learning_rate": 4.9885823825197e-06, + "loss": 0.7209, + "step": 617 + }, + { + "epoch": 0.2169943820224719, + "grad_norm": 0.9919439554214478, + "learning_rate": 4.988538204884121e-06, + "loss": 0.7631, + "step": 618 + }, + { + "epoch": 0.21734550561797752, + "grad_norm": 0.9095845818519592, + "learning_rate": 4.988493942142664e-06, + "loss": 0.7311, + "step": 619 + }, + { + "epoch": 0.21769662921348315, + "grad_norm": 0.9250304698944092, + "learning_rate": 4.988449594296845e-06, + "loss": 0.7616, + "step": 620 + }, + { + "epoch": 0.21804775280898878, + "grad_norm": 0.9156744480133057, + "learning_rate": 4.988405161348178e-06, + "loss": 0.7446, + "step": 621 + }, + { + "epoch": 0.21839887640449437, + "grad_norm": 0.9540274143218994, + "learning_rate": 4.988360643298184e-06, + "loss": 0.745, + "step": 622 + }, + { + "epoch": 0.21875, + "grad_norm": 0.9059514999389648, + "learning_rate": 4.988316040148386e-06, + "loss": 0.7169, + "step": 623 + }, + { + "epoch": 0.21910112359550563, + "grad_norm": 0.8970401287078857, + "learning_rate": 4.988271351900308e-06, + "loss": 0.7551, + "step": 624 + }, + { + "epoch": 0.21945224719101122, + "grad_norm": 0.8862085938453674, + "learning_rate": 4.988226578555479e-06, + "loss": 0.7141, + "step": 625 + }, + { + "epoch": 0.21980337078651685, + "grad_norm": 0.9373227953910828, + "learning_rate": 4.98818172011543e-06, + "loss": 0.7637, + "step": 626 + }, + { + "epoch": 0.22015449438202248, + "grad_norm": 0.9211307168006897, + "learning_rate": 4.988136776581697e-06, + "loss": 0.7428, + "step": 627 + }, + { + "epoch": 0.2205056179775281, + "grad_norm": 0.9092309474945068, + "learning_rate": 4.9880917479558135e-06, + "loss": 0.7341, + "step": 628 + }, + { + "epoch": 0.2208567415730337, + "grad_norm": 0.9342793822288513, + "learning_rate": 4.988046634239322e-06, + "loss": 0.7397, + "step": 629 + }, + { + "epoch": 0.22120786516853932, + "grad_norm": 0.9277600646018982, + "learning_rate": 4.9880014354337645e-06, + "loss": 0.718, + "step": 630 + }, + { + "epoch": 0.22155898876404495, + "grad_norm": 1.1010304689407349, + "learning_rate": 4.987956151540687e-06, + "loss": 0.7199, + "step": 631 + }, + { + "epoch": 0.22191011235955055, + "grad_norm": 0.9733719229698181, + "learning_rate": 4.9879107825616375e-06, + "loss": 0.7518, + "step": 632 + }, + { + "epoch": 0.22226123595505617, + "grad_norm": 0.9306866526603699, + "learning_rate": 4.9878653284981694e-06, + "loss": 0.7388, + "step": 633 + }, + { + "epoch": 0.2226123595505618, + "grad_norm": 0.9351629614830017, + "learning_rate": 4.987819789351835e-06, + "loss": 0.7326, + "step": 634 + }, + { + "epoch": 0.22296348314606743, + "grad_norm": 0.9454215168952942, + "learning_rate": 4.987774165124194e-06, + "loss": 0.7929, + "step": 635 + }, + { + "epoch": 0.22331460674157302, + "grad_norm": 0.8752790093421936, + "learning_rate": 4.987728455816804e-06, + "loss": 0.7282, + "step": 636 + }, + { + "epoch": 0.22366573033707865, + "grad_norm": 0.9457923173904419, + "learning_rate": 4.98768266143123e-06, + "loss": 0.7297, + "step": 637 + }, + { + "epoch": 0.22401685393258428, + "grad_norm": 0.9447070360183716, + "learning_rate": 4.987636781969038e-06, + "loss": 0.7502, + "step": 638 + }, + { + "epoch": 0.2243679775280899, + "grad_norm": 1.4757858514785767, + "learning_rate": 4.987590817431797e-06, + "loss": 0.754, + "step": 639 + }, + { + "epoch": 0.2247191011235955, + "grad_norm": 0.9355869889259338, + "learning_rate": 4.987544767821078e-06, + "loss": 0.7468, + "step": 640 + }, + { + "epoch": 0.22507022471910113, + "grad_norm": 0.9021878242492676, + "learning_rate": 4.9874986331384566e-06, + "loss": 0.6914, + "step": 641 + }, + { + "epoch": 0.22542134831460675, + "grad_norm": 0.9627546072006226, + "learning_rate": 4.98745241338551e-06, + "loss": 0.6998, + "step": 642 + }, + { + "epoch": 0.22577247191011235, + "grad_norm": 0.9815264940261841, + "learning_rate": 4.98740610856382e-06, + "loss": 0.7392, + "step": 643 + }, + { + "epoch": 0.22612359550561797, + "grad_norm": 0.9428890943527222, + "learning_rate": 4.98735971867497e-06, + "loss": 0.737, + "step": 644 + }, + { + "epoch": 0.2264747191011236, + "grad_norm": 0.9130154252052307, + "learning_rate": 4.987313243720546e-06, + "loss": 0.7363, + "step": 645 + }, + { + "epoch": 0.22682584269662923, + "grad_norm": 0.9495055079460144, + "learning_rate": 4.987266683702137e-06, + "loss": 0.7104, + "step": 646 + }, + { + "epoch": 0.22717696629213482, + "grad_norm": 0.9111932516098022, + "learning_rate": 4.987220038621336e-06, + "loss": 0.7107, + "step": 647 + }, + { + "epoch": 0.22752808988764045, + "grad_norm": 0.9328299760818481, + "learning_rate": 4.987173308479738e-06, + "loss": 0.7294, + "step": 648 + }, + { + "epoch": 0.22787921348314608, + "grad_norm": 0.9401968717575073, + "learning_rate": 4.987126493278942e-06, + "loss": 0.7089, + "step": 649 + }, + { + "epoch": 0.22823033707865167, + "grad_norm": 0.9041705131530762, + "learning_rate": 4.987079593020547e-06, + "loss": 0.7249, + "step": 650 + }, + { + "epoch": 0.2285814606741573, + "grad_norm": 0.8919897675514221, + "learning_rate": 4.987032607706158e-06, + "loss": 0.6662, + "step": 651 + }, + { + "epoch": 0.22893258426966293, + "grad_norm": 0.9634642004966736, + "learning_rate": 4.986985537337384e-06, + "loss": 0.7547, + "step": 652 + }, + { + "epoch": 0.22928370786516855, + "grad_norm": 0.9561633467674255, + "learning_rate": 4.9869383819158305e-06, + "loss": 0.7448, + "step": 653 + }, + { + "epoch": 0.22963483146067415, + "grad_norm": 0.9113203287124634, + "learning_rate": 4.986891141443113e-06, + "loss": 0.7064, + "step": 654 + }, + { + "epoch": 0.22998595505617977, + "grad_norm": 0.8830083012580872, + "learning_rate": 4.986843815920847e-06, + "loss": 0.6909, + "step": 655 + }, + { + "epoch": 0.2303370786516854, + "grad_norm": 0.9515541791915894, + "learning_rate": 4.98679640535065e-06, + "loss": 0.7262, + "step": 656 + }, + { + "epoch": 0.230688202247191, + "grad_norm": 0.9469254612922668, + "learning_rate": 4.986748909734145e-06, + "loss": 0.7177, + "step": 657 + }, + { + "epoch": 0.23103932584269662, + "grad_norm": 0.9044115543365479, + "learning_rate": 4.986701329072954e-06, + "loss": 0.7371, + "step": 658 + }, + { + "epoch": 0.23139044943820225, + "grad_norm": 0.9210987687110901, + "learning_rate": 4.986653663368706e-06, + "loss": 0.7327, + "step": 659 + }, + { + "epoch": 0.23174157303370788, + "grad_norm": 0.8826014995574951, + "learning_rate": 4.98660591262303e-06, + "loss": 0.7455, + "step": 660 + }, + { + "epoch": 0.23209269662921347, + "grad_norm": 0.8898245096206665, + "learning_rate": 4.986558076837561e-06, + "loss": 0.7242, + "step": 661 + }, + { + "epoch": 0.2324438202247191, + "grad_norm": 0.9125939607620239, + "learning_rate": 4.9865101560139324e-06, + "loss": 0.7587, + "step": 662 + }, + { + "epoch": 0.23279494382022473, + "grad_norm": 0.9783604741096497, + "learning_rate": 4.9864621501537845e-06, + "loss": 0.7356, + "step": 663 + }, + { + "epoch": 0.23314606741573032, + "grad_norm": 0.9296882152557373, + "learning_rate": 4.986414059258758e-06, + "loss": 0.7059, + "step": 664 + }, + { + "epoch": 0.23349719101123595, + "grad_norm": 0.9341956973075867, + "learning_rate": 4.9863658833305e-06, + "loss": 0.7342, + "step": 665 + }, + { + "epoch": 0.23384831460674158, + "grad_norm": 0.9274594783782959, + "learning_rate": 4.986317622370655e-06, + "loss": 0.7254, + "step": 666 + }, + { + "epoch": 0.2341994382022472, + "grad_norm": 0.8959493637084961, + "learning_rate": 4.986269276380875e-06, + "loss": 0.7138, + "step": 667 + }, + { + "epoch": 0.2345505617977528, + "grad_norm": 0.9621686935424805, + "learning_rate": 4.986220845362814e-06, + "loss": 0.7301, + "step": 668 + }, + { + "epoch": 0.23490168539325842, + "grad_norm": 0.9666121602058411, + "learning_rate": 4.9861723293181275e-06, + "loss": 0.754, + "step": 669 + }, + { + "epoch": 0.23525280898876405, + "grad_norm": 0.9423872828483582, + "learning_rate": 4.986123728248474e-06, + "loss": 0.7464, + "step": 670 + }, + { + "epoch": 0.23560393258426968, + "grad_norm": 0.9208926558494568, + "learning_rate": 4.986075042155517e-06, + "loss": 0.7092, + "step": 671 + }, + { + "epoch": 0.23595505617977527, + "grad_norm": 0.880066454410553, + "learning_rate": 4.986026271040921e-06, + "loss": 0.6935, + "step": 672 + }, + { + "epoch": 0.2363061797752809, + "grad_norm": 0.9050423502922058, + "learning_rate": 4.985977414906353e-06, + "loss": 0.742, + "step": 673 + }, + { + "epoch": 0.23665730337078653, + "grad_norm": 0.9089376330375671, + "learning_rate": 4.985928473753486e-06, + "loss": 0.7366, + "step": 674 + }, + { + "epoch": 0.23700842696629212, + "grad_norm": 0.9702712893486023, + "learning_rate": 4.985879447583992e-06, + "loss": 0.7314, + "step": 675 + }, + { + "epoch": 0.23735955056179775, + "grad_norm": 0.9272444248199463, + "learning_rate": 4.985830336399548e-06, + "loss": 0.7156, + "step": 676 + }, + { + "epoch": 0.23771067415730338, + "grad_norm": 0.9623541831970215, + "learning_rate": 4.985781140201833e-06, + "loss": 0.705, + "step": 677 + }, + { + "epoch": 0.238061797752809, + "grad_norm": 0.9228547215461731, + "learning_rate": 4.985731858992531e-06, + "loss": 0.7254, + "step": 678 + }, + { + "epoch": 0.2384129213483146, + "grad_norm": 0.8912785649299622, + "learning_rate": 4.985682492773326e-06, + "loss": 0.7254, + "step": 679 + }, + { + "epoch": 0.23876404494382023, + "grad_norm": 0.9913116693496704, + "learning_rate": 4.985633041545906e-06, + "loss": 0.7172, + "step": 680 + }, + { + "epoch": 0.23911516853932585, + "grad_norm": 0.9851749539375305, + "learning_rate": 4.985583505311965e-06, + "loss": 0.7711, + "step": 681 + }, + { + "epoch": 0.23946629213483145, + "grad_norm": 0.9252983331680298, + "learning_rate": 4.9855338840731934e-06, + "loss": 0.7356, + "step": 682 + }, + { + "epoch": 0.23981741573033707, + "grad_norm": 0.9541698694229126, + "learning_rate": 4.9854841778312905e-06, + "loss": 0.7111, + "step": 683 + }, + { + "epoch": 0.2401685393258427, + "grad_norm": 0.9459705948829651, + "learning_rate": 4.985434386587956e-06, + "loss": 0.7254, + "step": 684 + }, + { + "epoch": 0.24051966292134833, + "grad_norm": 1.0094040632247925, + "learning_rate": 4.985384510344892e-06, + "loss": 0.7638, + "step": 685 + }, + { + "epoch": 0.24087078651685392, + "grad_norm": 0.9937487244606018, + "learning_rate": 4.985334549103804e-06, + "loss": 0.7373, + "step": 686 + }, + { + "epoch": 0.24122191011235955, + "grad_norm": 0.9358577132225037, + "learning_rate": 4.985284502866401e-06, + "loss": 0.697, + "step": 687 + }, + { + "epoch": 0.24157303370786518, + "grad_norm": 0.9751777052879333, + "learning_rate": 4.985234371634395e-06, + "loss": 0.7626, + "step": 688 + }, + { + "epoch": 0.24192415730337077, + "grad_norm": 0.9775291085243225, + "learning_rate": 4.9851841554095e-06, + "loss": 0.7031, + "step": 689 + }, + { + "epoch": 0.2422752808988764, + "grad_norm": 0.9366273880004883, + "learning_rate": 4.9851338541934345e-06, + "loss": 0.7268, + "step": 690 + }, + { + "epoch": 0.24262640449438203, + "grad_norm": 0.8989854454994202, + "learning_rate": 4.985083467987917e-06, + "loss": 0.7102, + "step": 691 + }, + { + "epoch": 0.24297752808988765, + "grad_norm": 0.9326997399330139, + "learning_rate": 4.9850329967946715e-06, + "loss": 0.6963, + "step": 692 + }, + { + "epoch": 0.24332865168539325, + "grad_norm": 0.9004776477813721, + "learning_rate": 4.984982440615423e-06, + "loss": 0.7133, + "step": 693 + }, + { + "epoch": 0.24367977528089887, + "grad_norm": 0.915814220905304, + "learning_rate": 4.984931799451903e-06, + "loss": 0.7197, + "step": 694 + }, + { + "epoch": 0.2440308988764045, + "grad_norm": 0.9562146067619324, + "learning_rate": 4.984881073305841e-06, + "loss": 0.7175, + "step": 695 + }, + { + "epoch": 0.2443820224719101, + "grad_norm": 0.9829107522964478, + "learning_rate": 4.984830262178974e-06, + "loss": 0.7343, + "step": 696 + }, + { + "epoch": 0.24473314606741572, + "grad_norm": 0.9095830917358398, + "learning_rate": 4.984779366073036e-06, + "loss": 0.7474, + "step": 697 + }, + { + "epoch": 0.24508426966292135, + "grad_norm": 0.9279952049255371, + "learning_rate": 4.984728384989772e-06, + "loss": 0.7574, + "step": 698 + }, + { + "epoch": 0.24543539325842698, + "grad_norm": 1.07504141330719, + "learning_rate": 4.984677318930922e-06, + "loss": 0.7309, + "step": 699 + }, + { + "epoch": 0.24578651685393257, + "grad_norm": 0.9406894445419312, + "learning_rate": 4.984626167898235e-06, + "loss": 0.7187, + "step": 700 + }, + { + "epoch": 0.2461376404494382, + "grad_norm": 0.9229064583778381, + "learning_rate": 4.9845749318934575e-06, + "loss": 0.7379, + "step": 701 + }, + { + "epoch": 0.24648876404494383, + "grad_norm": 0.9469361901283264, + "learning_rate": 4.984523610918345e-06, + "loss": 0.7585, + "step": 702 + }, + { + "epoch": 0.24683988764044945, + "grad_norm": 0.9020796418190002, + "learning_rate": 4.984472204974651e-06, + "loss": 0.714, + "step": 703 + }, + { + "epoch": 0.24719101123595505, + "grad_norm": 0.9876411557197571, + "learning_rate": 4.9844207140641336e-06, + "loss": 0.7013, + "step": 704 + }, + { + "epoch": 0.24754213483146068, + "grad_norm": 0.9139506816864014, + "learning_rate": 4.984369138188553e-06, + "loss": 0.7252, + "step": 705 + }, + { + "epoch": 0.2478932584269663, + "grad_norm": 0.9223290681838989, + "learning_rate": 4.9843174773496735e-06, + "loss": 0.7603, + "step": 706 + }, + { + "epoch": 0.2482443820224719, + "grad_norm": 0.9541411995887756, + "learning_rate": 4.984265731549263e-06, + "loss": 0.6997, + "step": 707 + }, + { + "epoch": 0.24859550561797752, + "grad_norm": 0.9302672147750854, + "learning_rate": 4.98421390078909e-06, + "loss": 0.7252, + "step": 708 + }, + { + "epoch": 0.24894662921348315, + "grad_norm": 0.9554900527000427, + "learning_rate": 4.984161985070927e-06, + "loss": 0.7305, + "step": 709 + }, + { + "epoch": 0.24929775280898878, + "grad_norm": 0.9729623794555664, + "learning_rate": 4.98410998439655e-06, + "loss": 0.7318, + "step": 710 + }, + { + "epoch": 0.24964887640449437, + "grad_norm": 0.9298332929611206, + "learning_rate": 4.984057898767738e-06, + "loss": 0.7348, + "step": 711 + }, + { + "epoch": 0.25, + "grad_norm": 0.9653823375701904, + "learning_rate": 4.984005728186271e-06, + "loss": 0.7566, + "step": 712 + }, + { + "epoch": 0.2503511235955056, + "grad_norm": 0.9283461570739746, + "learning_rate": 4.983953472653933e-06, + "loss": 0.6999, + "step": 713 + }, + { + "epoch": 0.25070224719101125, + "grad_norm": 0.9600170254707336, + "learning_rate": 4.983901132172513e-06, + "loss": 0.7378, + "step": 714 + }, + { + "epoch": 0.25105337078651685, + "grad_norm": 0.9261863827705383, + "learning_rate": 4.983848706743799e-06, + "loss": 0.7245, + "step": 715 + }, + { + "epoch": 0.25140449438202245, + "grad_norm": 0.9267969727516174, + "learning_rate": 4.983796196369585e-06, + "loss": 0.6921, + "step": 716 + }, + { + "epoch": 0.2517556179775281, + "grad_norm": 0.9277015328407288, + "learning_rate": 4.983743601051666e-06, + "loss": 0.715, + "step": 717 + }, + { + "epoch": 0.2521067415730337, + "grad_norm": 0.9029396176338196, + "learning_rate": 4.9836909207918415e-06, + "loss": 0.7739, + "step": 718 + }, + { + "epoch": 0.25245786516853935, + "grad_norm": 0.9498785138130188, + "learning_rate": 4.983638155591913e-06, + "loss": 0.7071, + "step": 719 + }, + { + "epoch": 0.25280898876404495, + "grad_norm": 0.934562087059021, + "learning_rate": 4.983585305453685e-06, + "loss": 0.7183, + "step": 720 + }, + { + "epoch": 0.25316011235955055, + "grad_norm": 0.9572330713272095, + "learning_rate": 4.983532370378964e-06, + "loss": 0.7535, + "step": 721 + }, + { + "epoch": 0.2535112359550562, + "grad_norm": 0.9443022608757019, + "learning_rate": 4.983479350369562e-06, + "loss": 0.7127, + "step": 722 + }, + { + "epoch": 0.2538623595505618, + "grad_norm": 0.9258742928504944, + "learning_rate": 4.983426245427291e-06, + "loss": 0.7002, + "step": 723 + }, + { + "epoch": 0.2542134831460674, + "grad_norm": 0.939900279045105, + "learning_rate": 4.983373055553968e-06, + "loss": 0.7234, + "step": 724 + }, + { + "epoch": 0.25456460674157305, + "grad_norm": 0.9580402374267578, + "learning_rate": 4.983319780751411e-06, + "loss": 0.7519, + "step": 725 + }, + { + "epoch": 0.25491573033707865, + "grad_norm": 0.9556792974472046, + "learning_rate": 4.983266421021442e-06, + "loss": 0.7701, + "step": 726 + }, + { + "epoch": 0.25526685393258425, + "grad_norm": 0.9831913709640503, + "learning_rate": 4.983212976365887e-06, + "loss": 0.7189, + "step": 727 + }, + { + "epoch": 0.2556179775280899, + "grad_norm": 0.9363849759101868, + "learning_rate": 4.983159446786573e-06, + "loss": 0.7061, + "step": 728 + }, + { + "epoch": 0.2559691011235955, + "grad_norm": 0.9291892647743225, + "learning_rate": 4.983105832285331e-06, + "loss": 0.6942, + "step": 729 + }, + { + "epoch": 0.2563202247191011, + "grad_norm": 0.9837853908538818, + "learning_rate": 4.983052132863995e-06, + "loss": 0.7251, + "step": 730 + }, + { + "epoch": 0.25667134831460675, + "grad_norm": 0.9630160331726074, + "learning_rate": 4.9829983485244e-06, + "loss": 0.694, + "step": 731 + }, + { + "epoch": 0.25702247191011235, + "grad_norm": 0.972981333732605, + "learning_rate": 4.982944479268386e-06, + "loss": 0.7232, + "step": 732 + }, + { + "epoch": 0.257373595505618, + "grad_norm": 0.9112554788589478, + "learning_rate": 4.9828905250977955e-06, + "loss": 0.7553, + "step": 733 + }, + { + "epoch": 0.2577247191011236, + "grad_norm": 0.9407148361206055, + "learning_rate": 4.982836486014474e-06, + "loss": 0.7931, + "step": 734 + }, + { + "epoch": 0.2580758426966292, + "grad_norm": 0.9683741927146912, + "learning_rate": 4.98278236202027e-06, + "loss": 0.76, + "step": 735 + }, + { + "epoch": 0.25842696629213485, + "grad_norm": 0.9538892507553101, + "learning_rate": 4.9827281531170325e-06, + "loss": 0.7491, + "step": 736 + }, + { + "epoch": 0.25877808988764045, + "grad_norm": 0.9792529344558716, + "learning_rate": 4.982673859306617e-06, + "loss": 0.752, + "step": 737 + }, + { + "epoch": 0.25912921348314605, + "grad_norm": 0.9384421110153198, + "learning_rate": 4.98261948059088e-06, + "loss": 0.7587, + "step": 738 + }, + { + "epoch": 0.2594803370786517, + "grad_norm": 0.9048983454704285, + "learning_rate": 4.982565016971682e-06, + "loss": 0.7429, + "step": 739 + }, + { + "epoch": 0.2598314606741573, + "grad_norm": 0.893500804901123, + "learning_rate": 4.982510468450884e-06, + "loss": 0.6985, + "step": 740 + }, + { + "epoch": 0.2601825842696629, + "grad_norm": 0.9900831580162048, + "learning_rate": 4.982455835030352e-06, + "loss": 0.7336, + "step": 741 + }, + { + "epoch": 0.26053370786516855, + "grad_norm": 0.9558880925178528, + "learning_rate": 4.982401116711955e-06, + "loss": 0.7561, + "step": 742 + }, + { + "epoch": 0.26088483146067415, + "grad_norm": 0.8868340849876404, + "learning_rate": 4.982346313497564e-06, + "loss": 0.6967, + "step": 743 + }, + { + "epoch": 0.2612359550561798, + "grad_norm": 0.9625622034072876, + "learning_rate": 4.9822914253890535e-06, + "loss": 0.7382, + "step": 744 + }, + { + "epoch": 0.2615870786516854, + "grad_norm": 0.9891008138656616, + "learning_rate": 4.9822364523883e-06, + "loss": 0.7316, + "step": 745 + }, + { + "epoch": 0.261938202247191, + "grad_norm": 0.9161337018013, + "learning_rate": 4.982181394497184e-06, + "loss": 0.7225, + "step": 746 + }, + { + "epoch": 0.26228932584269665, + "grad_norm": 0.9532581567764282, + "learning_rate": 4.982126251717589e-06, + "loss": 0.7412, + "step": 747 + }, + { + "epoch": 0.26264044943820225, + "grad_norm": 0.90705406665802, + "learning_rate": 4.982071024051399e-06, + "loss": 0.6938, + "step": 748 + }, + { + "epoch": 0.26299157303370785, + "grad_norm": 0.9263889789581299, + "learning_rate": 4.982015711500505e-06, + "loss": 0.7575, + "step": 749 + }, + { + "epoch": 0.2633426966292135, + "grad_norm": 0.9247738718986511, + "learning_rate": 4.981960314066797e-06, + "loss": 0.7184, + "step": 750 + }, + { + "epoch": 0.2636938202247191, + "grad_norm": 0.9485296607017517, + "learning_rate": 4.98190483175217e-06, + "loss": 0.7297, + "step": 751 + }, + { + "epoch": 0.2640449438202247, + "grad_norm": 0.9109437465667725, + "learning_rate": 4.981849264558523e-06, + "loss": 0.7422, + "step": 752 + }, + { + "epoch": 0.26439606741573035, + "grad_norm": 0.9430893063545227, + "learning_rate": 4.981793612487753e-06, + "loss": 0.7634, + "step": 753 + }, + { + "epoch": 0.26474719101123595, + "grad_norm": 0.9232211709022522, + "learning_rate": 4.981737875541765e-06, + "loss": 0.7587, + "step": 754 + }, + { + "epoch": 0.26509831460674155, + "grad_norm": 0.8920214176177979, + "learning_rate": 4.9816820537224675e-06, + "loss": 0.7222, + "step": 755 + }, + { + "epoch": 0.2654494382022472, + "grad_norm": 0.9339509010314941, + "learning_rate": 4.981626147031765e-06, + "loss": 0.7253, + "step": 756 + }, + { + "epoch": 0.2658005617977528, + "grad_norm": 0.890581488609314, + "learning_rate": 4.981570155471573e-06, + "loss": 0.7166, + "step": 757 + }, + { + "epoch": 0.26615168539325845, + "grad_norm": 0.9297827482223511, + "learning_rate": 4.981514079043806e-06, + "loss": 0.713, + "step": 758 + }, + { + "epoch": 0.26650280898876405, + "grad_norm": 0.9213199019432068, + "learning_rate": 4.98145791775038e-06, + "loss": 0.7312, + "step": 759 + }, + { + "epoch": 0.26685393258426965, + "grad_norm": 0.9185107946395874, + "learning_rate": 4.981401671593216e-06, + "loss": 0.7215, + "step": 760 + }, + { + "epoch": 0.2672050561797753, + "grad_norm": 0.8874849081039429, + "learning_rate": 4.981345340574239e-06, + "loss": 0.7082, + "step": 761 + }, + { + "epoch": 0.2675561797752809, + "grad_norm": 0.9179455637931824, + "learning_rate": 4.981288924695374e-06, + "loss": 0.7243, + "step": 762 + }, + { + "epoch": 0.2679073033707865, + "grad_norm": 0.8956063985824585, + "learning_rate": 4.981232423958552e-06, + "loss": 0.704, + "step": 763 + }, + { + "epoch": 0.26825842696629215, + "grad_norm": 0.9451838731765747, + "learning_rate": 4.981175838365704e-06, + "loss": 0.7518, + "step": 764 + }, + { + "epoch": 0.26860955056179775, + "grad_norm": 0.9265857338905334, + "learning_rate": 4.981119167918765e-06, + "loss": 0.7143, + "step": 765 + }, + { + "epoch": 0.26896067415730335, + "grad_norm": 0.9493303298950195, + "learning_rate": 4.981062412619674e-06, + "loss": 0.7172, + "step": 766 + }, + { + "epoch": 0.269311797752809, + "grad_norm": 0.9223463535308838, + "learning_rate": 4.981005572470372e-06, + "loss": 0.7383, + "step": 767 + }, + { + "epoch": 0.2696629213483146, + "grad_norm": 0.9502999782562256, + "learning_rate": 4.980948647472802e-06, + "loss": 0.7137, + "step": 768 + }, + { + "epoch": 0.2700140449438202, + "grad_norm": 0.9198105335235596, + "learning_rate": 4.980891637628911e-06, + "loss": 0.706, + "step": 769 + }, + { + "epoch": 0.27036516853932585, + "grad_norm": 0.9435210227966309, + "learning_rate": 4.980834542940649e-06, + "loss": 0.6921, + "step": 770 + }, + { + "epoch": 0.27071629213483145, + "grad_norm": 0.9600005745887756, + "learning_rate": 4.980777363409969e-06, + "loss": 0.72, + "step": 771 + }, + { + "epoch": 0.2710674157303371, + "grad_norm": 0.9591335654258728, + "learning_rate": 4.980720099038825e-06, + "loss": 0.7386, + "step": 772 + }, + { + "epoch": 0.2714185393258427, + "grad_norm": 0.9055672883987427, + "learning_rate": 4.980662749829177e-06, + "loss": 0.7568, + "step": 773 + }, + { + "epoch": 0.2717696629213483, + "grad_norm": 0.9077644348144531, + "learning_rate": 4.9806053157829855e-06, + "loss": 0.7122, + "step": 774 + }, + { + "epoch": 0.27212078651685395, + "grad_norm": 0.9248702526092529, + "learning_rate": 4.980547796902216e-06, + "loss": 0.6866, + "step": 775 + }, + { + "epoch": 0.27247191011235955, + "grad_norm": 0.9374347925186157, + "learning_rate": 4.980490193188834e-06, + "loss": 0.7283, + "step": 776 + }, + { + "epoch": 0.27282303370786515, + "grad_norm": 0.9280509352684021, + "learning_rate": 4.980432504644809e-06, + "loss": 0.6896, + "step": 777 + }, + { + "epoch": 0.2731741573033708, + "grad_norm": 0.9158201813697815, + "learning_rate": 4.980374731272115e-06, + "loss": 0.7222, + "step": 778 + }, + { + "epoch": 0.2735252808988764, + "grad_norm": 0.918114960193634, + "learning_rate": 4.980316873072729e-06, + "loss": 0.7528, + "step": 779 + }, + { + "epoch": 0.273876404494382, + "grad_norm": 0.891121506690979, + "learning_rate": 4.980258930048627e-06, + "loss": 0.7265, + "step": 780 + }, + { + "epoch": 0.27422752808988765, + "grad_norm": 0.9570785760879517, + "learning_rate": 4.980200902201792e-06, + "loss": 0.6957, + "step": 781 + }, + { + "epoch": 0.27457865168539325, + "grad_norm": 0.8888586759567261, + "learning_rate": 4.980142789534209e-06, + "loss": 0.7258, + "step": 782 + }, + { + "epoch": 0.2749297752808989, + "grad_norm": 0.9602726101875305, + "learning_rate": 4.980084592047866e-06, + "loss": 0.741, + "step": 783 + }, + { + "epoch": 0.2752808988764045, + "grad_norm": 0.9402101635932922, + "learning_rate": 4.9800263097447505e-06, + "loss": 0.7789, + "step": 784 + }, + { + "epoch": 0.2756320224719101, + "grad_norm": 0.916023850440979, + "learning_rate": 4.9799679426268575e-06, + "loss": 0.6746, + "step": 785 + }, + { + "epoch": 0.27598314606741575, + "grad_norm": 1.0306038856506348, + "learning_rate": 4.979909490696184e-06, + "loss": 0.7396, + "step": 786 + }, + { + "epoch": 0.27633426966292135, + "grad_norm": 0.9112997651100159, + "learning_rate": 4.979850953954727e-06, + "loss": 0.7225, + "step": 787 + }, + { + "epoch": 0.27668539325842695, + "grad_norm": 0.9124306440353394, + "learning_rate": 4.97979233240449e-06, + "loss": 0.7244, + "step": 788 + }, + { + "epoch": 0.2770365168539326, + "grad_norm": 0.8682541847229004, + "learning_rate": 4.979733626047477e-06, + "loss": 0.6983, + "step": 789 + }, + { + "epoch": 0.2773876404494382, + "grad_norm": 0.9099353551864624, + "learning_rate": 4.979674834885696e-06, + "loss": 0.723, + "step": 790 + }, + { + "epoch": 0.2777387640449438, + "grad_norm": 0.9057777523994446, + "learning_rate": 4.979615958921158e-06, + "loss": 0.7063, + "step": 791 + }, + { + "epoch": 0.27808988764044945, + "grad_norm": 0.9777807593345642, + "learning_rate": 4.979556998155876e-06, + "loss": 0.7047, + "step": 792 + }, + { + "epoch": 0.27844101123595505, + "grad_norm": 0.952314019203186, + "learning_rate": 4.979497952591865e-06, + "loss": 0.7589, + "step": 793 + }, + { + "epoch": 0.27879213483146065, + "grad_norm": 0.8948432207107544, + "learning_rate": 4.979438822231147e-06, + "loss": 0.6893, + "step": 794 + }, + { + "epoch": 0.2791432584269663, + "grad_norm": 0.9998204708099365, + "learning_rate": 4.979379607075743e-06, + "loss": 0.7176, + "step": 795 + }, + { + "epoch": 0.2794943820224719, + "grad_norm": 0.9392699003219604, + "learning_rate": 4.979320307127678e-06, + "loss": 0.7607, + "step": 796 + }, + { + "epoch": 0.27984550561797755, + "grad_norm": 0.916340708732605, + "learning_rate": 4.97926092238898e-06, + "loss": 0.6797, + "step": 797 + }, + { + "epoch": 0.28019662921348315, + "grad_norm": 0.9007791876792908, + "learning_rate": 4.979201452861679e-06, + "loss": 0.7015, + "step": 798 + }, + { + "epoch": 0.28054775280898875, + "grad_norm": 0.9042453765869141, + "learning_rate": 4.979141898547811e-06, + "loss": 0.7113, + "step": 799 + }, + { + "epoch": 0.2808988764044944, + "grad_norm": 0.9455578327178955, + "learning_rate": 4.9790822594494116e-06, + "loss": 0.7031, + "step": 800 + }, + { + "epoch": 0.28125, + "grad_norm": 0.9787691235542297, + "learning_rate": 4.979022535568519e-06, + "loss": 0.7328, + "step": 801 + }, + { + "epoch": 0.2816011235955056, + "grad_norm": 0.9504345059394836, + "learning_rate": 4.978962726907179e-06, + "loss": 0.7301, + "step": 802 + }, + { + "epoch": 0.28195224719101125, + "grad_norm": 0.9304360747337341, + "learning_rate": 4.978902833467434e-06, + "loss": 0.7195, + "step": 803 + }, + { + "epoch": 0.28230337078651685, + "grad_norm": 0.9509701132774353, + "learning_rate": 4.978842855251333e-06, + "loss": 0.7532, + "step": 804 + }, + { + "epoch": 0.28265449438202245, + "grad_norm": 0.9462277889251709, + "learning_rate": 4.9787827922609276e-06, + "loss": 0.708, + "step": 805 + }, + { + "epoch": 0.2830056179775281, + "grad_norm": 0.9095754027366638, + "learning_rate": 4.978722644498272e-06, + "loss": 0.7484, + "step": 806 + }, + { + "epoch": 0.2833567415730337, + "grad_norm": 0.8992451429367065, + "learning_rate": 4.978662411965424e-06, + "loss": 0.7033, + "step": 807 + }, + { + "epoch": 0.28370786516853935, + "grad_norm": 0.913910984992981, + "learning_rate": 4.978602094664441e-06, + "loss": 0.7333, + "step": 808 + }, + { + "epoch": 0.28405898876404495, + "grad_norm": 0.9041484594345093, + "learning_rate": 4.978541692597388e-06, + "loss": 0.7196, + "step": 809 + }, + { + "epoch": 0.28441011235955055, + "grad_norm": 0.9570571184158325, + "learning_rate": 4.97848120576633e-06, + "loss": 0.7199, + "step": 810 + }, + { + "epoch": 0.2847612359550562, + "grad_norm": 0.9184895753860474, + "learning_rate": 4.978420634173336e-06, + "loss": 0.6982, + "step": 811 + }, + { + "epoch": 0.2851123595505618, + "grad_norm": 0.9278954267501831, + "learning_rate": 4.978359977820476e-06, + "loss": 0.7475, + "step": 812 + }, + { + "epoch": 0.2854634831460674, + "grad_norm": 0.9236921668052673, + "learning_rate": 4.978299236709826e-06, + "loss": 0.7343, + "step": 813 + }, + { + "epoch": 0.28581460674157305, + "grad_norm": 0.9183304309844971, + "learning_rate": 4.978238410843464e-06, + "loss": 0.7111, + "step": 814 + }, + { + "epoch": 0.28616573033707865, + "grad_norm": 0.895050585269928, + "learning_rate": 4.978177500223468e-06, + "loss": 0.6746, + "step": 815 + }, + { + "epoch": 0.28651685393258425, + "grad_norm": 0.9193557500839233, + "learning_rate": 4.9781165048519216e-06, + "loss": 0.7025, + "step": 816 + }, + { + "epoch": 0.2868679775280899, + "grad_norm": 0.9046199917793274, + "learning_rate": 4.978055424730912e-06, + "loss": 0.701, + "step": 817 + }, + { + "epoch": 0.2872191011235955, + "grad_norm": 0.9200829267501831, + "learning_rate": 4.977994259862527e-06, + "loss": 0.755, + "step": 818 + }, + { + "epoch": 0.2875702247191011, + "grad_norm": 0.929964542388916, + "learning_rate": 4.977933010248859e-06, + "loss": 0.7274, + "step": 819 + }, + { + "epoch": 0.28792134831460675, + "grad_norm": 0.9332923293113708, + "learning_rate": 4.977871675892003e-06, + "loss": 0.7503, + "step": 820 + }, + { + "epoch": 0.28827247191011235, + "grad_norm": 0.8993156552314758, + "learning_rate": 4.977810256794055e-06, + "loss": 0.6954, + "step": 821 + }, + { + "epoch": 0.288623595505618, + "grad_norm": 0.8985475301742554, + "learning_rate": 4.977748752957116e-06, + "loss": 0.7287, + "step": 822 + }, + { + "epoch": 0.2889747191011236, + "grad_norm": 0.9083667397499084, + "learning_rate": 4.977687164383291e-06, + "loss": 0.7352, + "step": 823 + }, + { + "epoch": 0.2893258426966292, + "grad_norm": 0.9311912059783936, + "learning_rate": 4.977625491074686e-06, + "loss": 0.7199, + "step": 824 + }, + { + "epoch": 0.28967696629213485, + "grad_norm": 0.8959439396858215, + "learning_rate": 4.977563733033408e-06, + "loss": 0.6992, + "step": 825 + }, + { + "epoch": 0.29002808988764045, + "grad_norm": 1.018567681312561, + "learning_rate": 4.9775018902615715e-06, + "loss": 0.7278, + "step": 826 + }, + { + "epoch": 0.29037921348314605, + "grad_norm": 0.9925201535224915, + "learning_rate": 4.977439962761289e-06, + "loss": 0.7423, + "step": 827 + }, + { + "epoch": 0.2907303370786517, + "grad_norm": 1.0001310110092163, + "learning_rate": 4.977377950534681e-06, + "loss": 0.7395, + "step": 828 + }, + { + "epoch": 0.2910814606741573, + "grad_norm": 0.9196740984916687, + "learning_rate": 4.977315853583866e-06, + "loss": 0.7588, + "step": 829 + }, + { + "epoch": 0.2914325842696629, + "grad_norm": 0.9672794342041016, + "learning_rate": 4.977253671910969e-06, + "loss": 0.7378, + "step": 830 + }, + { + "epoch": 0.29178370786516855, + "grad_norm": 1.0301952362060547, + "learning_rate": 4.977191405518116e-06, + "loss": 0.7757, + "step": 831 + }, + { + "epoch": 0.29213483146067415, + "grad_norm": 0.9453630447387695, + "learning_rate": 4.977129054407437e-06, + "loss": 0.7335, + "step": 832 + }, + { + "epoch": 0.2924859550561798, + "grad_norm": 0.9876764416694641, + "learning_rate": 4.977066618581065e-06, + "loss": 0.7437, + "step": 833 + }, + { + "epoch": 0.2928370786516854, + "grad_norm": 0.9211217164993286, + "learning_rate": 4.977004098041133e-06, + "loss": 0.7333, + "step": 834 + }, + { + "epoch": 0.293188202247191, + "grad_norm": 0.9148357510566711, + "learning_rate": 4.976941492789781e-06, + "loss": 0.677, + "step": 835 + }, + { + "epoch": 0.29353932584269665, + "grad_norm": 0.9743471741676331, + "learning_rate": 4.976878802829149e-06, + "loss": 0.7586, + "step": 836 + }, + { + "epoch": 0.29389044943820225, + "grad_norm": 0.9516198039054871, + "learning_rate": 4.976816028161382e-06, + "loss": 0.7364, + "step": 837 + }, + { + "epoch": 0.29424157303370785, + "grad_norm": 0.8972299098968506, + "learning_rate": 4.9767531687886265e-06, + "loss": 0.6912, + "step": 838 + }, + { + "epoch": 0.2945926966292135, + "grad_norm": 0.8950793147087097, + "learning_rate": 4.976690224713031e-06, + "loss": 0.6969, + "step": 839 + }, + { + "epoch": 0.2949438202247191, + "grad_norm": 0.9539402723312378, + "learning_rate": 4.976627195936749e-06, + "loss": 0.7348, + "step": 840 + }, + { + "epoch": 0.2952949438202247, + "grad_norm": 1.32026207447052, + "learning_rate": 4.976564082461938e-06, + "loss": 0.7192, + "step": 841 + }, + { + "epoch": 0.29564606741573035, + "grad_norm": 0.9319935441017151, + "learning_rate": 4.976500884290753e-06, + "loss": 0.6957, + "step": 842 + }, + { + "epoch": 0.29599719101123595, + "grad_norm": 0.9559399485588074, + "learning_rate": 4.976437601425358e-06, + "loss": 0.7394, + "step": 843 + }, + { + "epoch": 0.29634831460674155, + "grad_norm": 0.9115216135978699, + "learning_rate": 4.976374233867915e-06, + "loss": 0.6955, + "step": 844 + }, + { + "epoch": 0.2966994382022472, + "grad_norm": 0.8987361788749695, + "learning_rate": 4.976310781620592e-06, + "loss": 0.7057, + "step": 845 + }, + { + "epoch": 0.2970505617977528, + "grad_norm": 0.9885380268096924, + "learning_rate": 4.976247244685561e-06, + "loss": 0.7542, + "step": 846 + }, + { + "epoch": 0.29740168539325845, + "grad_norm": 0.9176011681556702, + "learning_rate": 4.9761836230649914e-06, + "loss": 0.6836, + "step": 847 + }, + { + "epoch": 0.29775280898876405, + "grad_norm": 0.9235329031944275, + "learning_rate": 4.976119916761063e-06, + "loss": 0.7541, + "step": 848 + }, + { + "epoch": 0.29810393258426965, + "grad_norm": 0.9174650311470032, + "learning_rate": 4.9760561257759504e-06, + "loss": 0.7149, + "step": 849 + }, + { + "epoch": 0.2984550561797753, + "grad_norm": 0.9441576600074768, + "learning_rate": 4.975992250111837e-06, + "loss": 0.7222, + "step": 850 + }, + { + "epoch": 0.2988061797752809, + "grad_norm": 0.9702540636062622, + "learning_rate": 4.975928289770909e-06, + "loss": 0.7193, + "step": 851 + }, + { + "epoch": 0.2991573033707865, + "grad_norm": 0.9554623961448669, + "learning_rate": 4.9758642447553505e-06, + "loss": 0.7537, + "step": 852 + }, + { + "epoch": 0.29950842696629215, + "grad_norm": 0.9210932850837708, + "learning_rate": 4.975800115067355e-06, + "loss": 0.7159, + "step": 853 + }, + { + "epoch": 0.29985955056179775, + "grad_norm": 0.902009904384613, + "learning_rate": 4.975735900709113e-06, + "loss": 0.6952, + "step": 854 + }, + { + "epoch": 0.30021067415730335, + "grad_norm": 0.9207450151443481, + "learning_rate": 4.975671601682822e-06, + "loss": 0.6705, + "step": 855 + }, + { + "epoch": 0.300561797752809, + "grad_norm": 0.9092718362808228, + "learning_rate": 4.975607217990681e-06, + "loss": 0.723, + "step": 856 + }, + { + "epoch": 0.3009129213483146, + "grad_norm": 0.955410897731781, + "learning_rate": 4.975542749634891e-06, + "loss": 0.728, + "step": 857 + }, + { + "epoch": 0.3012640449438202, + "grad_norm": 0.9404045343399048, + "learning_rate": 4.975478196617658e-06, + "loss": 0.7381, + "step": 858 + }, + { + "epoch": 0.30161516853932585, + "grad_norm": 0.9133435487747192, + "learning_rate": 4.9754135589411875e-06, + "loss": 0.7306, + "step": 859 + }, + { + "epoch": 0.30196629213483145, + "grad_norm": 0.9013397693634033, + "learning_rate": 4.975348836607693e-06, + "loss": 0.7189, + "step": 860 + }, + { + "epoch": 0.3023174157303371, + "grad_norm": 0.9037486910820007, + "learning_rate": 4.975284029619386e-06, + "loss": 0.7152, + "step": 861 + }, + { + "epoch": 0.3026685393258427, + "grad_norm": 0.9141895771026611, + "learning_rate": 4.975219137978484e-06, + "loss": 0.71, + "step": 862 + }, + { + "epoch": 0.3030196629213483, + "grad_norm": 0.942010223865509, + "learning_rate": 4.975154161687205e-06, + "loss": 0.7255, + "step": 863 + }, + { + "epoch": 0.30337078651685395, + "grad_norm": 0.9464311599731445, + "learning_rate": 4.9750891007477716e-06, + "loss": 0.7012, + "step": 864 + }, + { + "epoch": 0.30372191011235955, + "grad_norm": 0.9062106609344482, + "learning_rate": 4.975023955162409e-06, + "loss": 0.6909, + "step": 865 + }, + { + "epoch": 0.30407303370786515, + "grad_norm": 0.921981155872345, + "learning_rate": 4.974958724933344e-06, + "loss": 0.7312, + "step": 866 + }, + { + "epoch": 0.3044241573033708, + "grad_norm": 0.902531087398529, + "learning_rate": 4.97489341006281e-06, + "loss": 0.674, + "step": 867 + }, + { + "epoch": 0.3047752808988764, + "grad_norm": 0.9487978219985962, + "learning_rate": 4.974828010553039e-06, + "loss": 0.7054, + "step": 868 + }, + { + "epoch": 0.305126404494382, + "grad_norm": 0.9971505999565125, + "learning_rate": 4.9747625264062675e-06, + "loss": 0.7676, + "step": 869 + }, + { + "epoch": 0.30547752808988765, + "grad_norm": 0.9108356237411499, + "learning_rate": 4.974696957624735e-06, + "loss": 0.7199, + "step": 870 + }, + { + "epoch": 0.30582865168539325, + "grad_norm": 0.9270719885826111, + "learning_rate": 4.974631304210684e-06, + "loss": 0.748, + "step": 871 + }, + { + "epoch": 0.3061797752808989, + "grad_norm": 0.9216623306274414, + "learning_rate": 4.97456556616636e-06, + "loss": 0.7333, + "step": 872 + }, + { + "epoch": 0.3065308988764045, + "grad_norm": 0.980196475982666, + "learning_rate": 4.9744997434940125e-06, + "loss": 0.7175, + "step": 873 + }, + { + "epoch": 0.3068820224719101, + "grad_norm": 0.8955034613609314, + "learning_rate": 4.97443383619589e-06, + "loss": 0.7133, + "step": 874 + }, + { + "epoch": 0.30723314606741575, + "grad_norm": 0.9280632734298706, + "learning_rate": 4.974367844274248e-06, + "loss": 0.6953, + "step": 875 + }, + { + "epoch": 0.30758426966292135, + "grad_norm": 0.9301517605781555, + "learning_rate": 4.974301767731343e-06, + "loss": 0.7057, + "step": 876 + }, + { + "epoch": 0.30793539325842695, + "grad_norm": 0.9015063047409058, + "learning_rate": 4.974235606569434e-06, + "loss": 0.699, + "step": 877 + }, + { + "epoch": 0.3082865168539326, + "grad_norm": 0.9064421057701111, + "learning_rate": 4.974169360790786e-06, + "loss": 0.6937, + "step": 878 + }, + { + "epoch": 0.3086376404494382, + "grad_norm": 0.9030662775039673, + "learning_rate": 4.974103030397662e-06, + "loss": 0.7455, + "step": 879 + }, + { + "epoch": 0.3089887640449438, + "grad_norm": 0.9049381613731384, + "learning_rate": 4.974036615392332e-06, + "loss": 0.7203, + "step": 880 + }, + { + "epoch": 0.30933988764044945, + "grad_norm": 0.9038983583450317, + "learning_rate": 4.973970115777067e-06, + "loss": 0.7122, + "step": 881 + }, + { + "epoch": 0.30969101123595505, + "grad_norm": 0.96694415807724, + "learning_rate": 4.973903531554141e-06, + "loss": 0.7292, + "step": 882 + }, + { + "epoch": 0.31004213483146065, + "grad_norm": 0.9090932011604309, + "learning_rate": 4.973836862725831e-06, + "loss": 0.7463, + "step": 883 + }, + { + "epoch": 0.3103932584269663, + "grad_norm": 0.9024604558944702, + "learning_rate": 4.973770109294416e-06, + "loss": 0.7488, + "step": 884 + }, + { + "epoch": 0.3107443820224719, + "grad_norm": 0.9796269536018372, + "learning_rate": 4.973703271262181e-06, + "loss": 0.7398, + "step": 885 + }, + { + "epoch": 0.31109550561797755, + "grad_norm": 0.9187832474708557, + "learning_rate": 4.973636348631412e-06, + "loss": 0.7162, + "step": 886 + }, + { + "epoch": 0.31144662921348315, + "grad_norm": 0.8963474631309509, + "learning_rate": 4.973569341404395e-06, + "loss": 0.6812, + "step": 887 + }, + { + "epoch": 0.31179775280898875, + "grad_norm": 0.934047281742096, + "learning_rate": 4.973502249583424e-06, + "loss": 0.7176, + "step": 888 + }, + { + "epoch": 0.3121488764044944, + "grad_norm": 0.9109435677528381, + "learning_rate": 4.9734350731707925e-06, + "loss": 0.684, + "step": 889 + }, + { + "epoch": 0.3125, + "grad_norm": 0.9629603028297424, + "learning_rate": 4.973367812168798e-06, + "loss": 0.756, + "step": 890 + }, + { + "epoch": 0.3128511235955056, + "grad_norm": 0.8999819159507751, + "learning_rate": 4.973300466579741e-06, + "loss": 0.7211, + "step": 891 + }, + { + "epoch": 0.31320224719101125, + "grad_norm": 0.9093006253242493, + "learning_rate": 4.9732330364059245e-06, + "loss": 0.6893, + "step": 892 + }, + { + "epoch": 0.31355337078651685, + "grad_norm": 0.925269603729248, + "learning_rate": 4.973165521649655e-06, + "loss": 0.6631, + "step": 893 + }, + { + "epoch": 0.31390449438202245, + "grad_norm": 0.9515416622161865, + "learning_rate": 4.973097922313239e-06, + "loss": 0.7578, + "step": 894 + }, + { + "epoch": 0.3142556179775281, + "grad_norm": 0.9192783236503601, + "learning_rate": 4.973030238398992e-06, + "loss": 0.6884, + "step": 895 + }, + { + "epoch": 0.3146067415730337, + "grad_norm": 0.905805230140686, + "learning_rate": 4.972962469909227e-06, + "loss": 0.7037, + "step": 896 + }, + { + "epoch": 0.31495786516853935, + "grad_norm": 0.9037303328514099, + "learning_rate": 4.972894616846261e-06, + "loss": 0.6989, + "step": 897 + }, + { + "epoch": 0.31530898876404495, + "grad_norm": 0.8707857728004456, + "learning_rate": 4.9728266792124155e-06, + "loss": 0.6865, + "step": 898 + }, + { + "epoch": 0.31566011235955055, + "grad_norm": 0.9623267650604248, + "learning_rate": 4.972758657010014e-06, + "loss": 0.7028, + "step": 899 + }, + { + "epoch": 0.3160112359550562, + "grad_norm": 0.9062806367874146, + "learning_rate": 4.972690550241383e-06, + "loss": 0.6803, + "step": 900 + }, + { + "epoch": 0.3163623595505618, + "grad_norm": 0.9363368153572083, + "learning_rate": 4.97262235890885e-06, + "loss": 0.7359, + "step": 901 + }, + { + "epoch": 0.3167134831460674, + "grad_norm": 0.9362224340438843, + "learning_rate": 4.972554083014748e-06, + "loss": 0.7432, + "step": 902 + }, + { + "epoch": 0.31706460674157305, + "grad_norm": 0.9579148292541504, + "learning_rate": 4.972485722561412e-06, + "loss": 0.7348, + "step": 903 + }, + { + "epoch": 0.31741573033707865, + "grad_norm": 0.9559216499328613, + "learning_rate": 4.97241727755118e-06, + "loss": 0.7095, + "step": 904 + }, + { + "epoch": 0.31776685393258425, + "grad_norm": 0.9043166637420654, + "learning_rate": 4.972348747986394e-06, + "loss": 0.7289, + "step": 905 + }, + { + "epoch": 0.3181179775280899, + "grad_norm": 0.925491988658905, + "learning_rate": 4.972280133869396e-06, + "loss": 0.6921, + "step": 906 + }, + { + "epoch": 0.3184691011235955, + "grad_norm": 0.9442777037620544, + "learning_rate": 4.972211435202533e-06, + "loss": 0.7104, + "step": 907 + }, + { + "epoch": 0.3188202247191011, + "grad_norm": 0.943748950958252, + "learning_rate": 4.972142651988153e-06, + "loss": 0.7688, + "step": 908 + }, + { + "epoch": 0.31917134831460675, + "grad_norm": 0.8944805860519409, + "learning_rate": 4.972073784228612e-06, + "loss": 0.7321, + "step": 909 + }, + { + "epoch": 0.31952247191011235, + "grad_norm": 0.9088547825813293, + "learning_rate": 4.972004831926261e-06, + "loss": 0.6973, + "step": 910 + }, + { + "epoch": 0.319873595505618, + "grad_norm": 0.9270226359367371, + "learning_rate": 4.971935795083461e-06, + "loss": 0.6956, + "step": 911 + }, + { + "epoch": 0.3202247191011236, + "grad_norm": 0.910492479801178, + "learning_rate": 4.971866673702573e-06, + "loss": 0.681, + "step": 912 + }, + { + "epoch": 0.3205758426966292, + "grad_norm": 0.9591485857963562, + "learning_rate": 4.971797467785958e-06, + "loss": 0.7763, + "step": 913 + }, + { + "epoch": 0.32092696629213485, + "grad_norm": 0.9132441878318787, + "learning_rate": 4.971728177335986e-06, + "loss": 0.7294, + "step": 914 + }, + { + "epoch": 0.32127808988764045, + "grad_norm": 0.9387941956520081, + "learning_rate": 4.971658802355025e-06, + "loss": 0.755, + "step": 915 + }, + { + "epoch": 0.32162921348314605, + "grad_norm": 0.930505096912384, + "learning_rate": 4.971589342845449e-06, + "loss": 0.7337, + "step": 916 + }, + { + "epoch": 0.3219803370786517, + "grad_norm": 0.8818848133087158, + "learning_rate": 4.971519798809631e-06, + "loss": 0.7114, + "step": 917 + }, + { + "epoch": 0.3223314606741573, + "grad_norm": 0.9763118028640747, + "learning_rate": 4.971450170249952e-06, + "loss": 0.73, + "step": 918 + }, + { + "epoch": 0.3226825842696629, + "grad_norm": 0.865503191947937, + "learning_rate": 4.971380457168791e-06, + "loss": 0.6733, + "step": 919 + }, + { + "epoch": 0.32303370786516855, + "grad_norm": 0.9227094650268555, + "learning_rate": 4.9713106595685336e-06, + "loss": 0.6975, + "step": 920 + }, + { + "epoch": 0.32338483146067415, + "grad_norm": 0.9164808988571167, + "learning_rate": 4.971240777451566e-06, + "loss": 0.7266, + "step": 921 + }, + { + "epoch": 0.3237359550561798, + "grad_norm": 0.9115430116653442, + "learning_rate": 4.971170810820279e-06, + "loss": 0.6775, + "step": 922 + }, + { + "epoch": 0.3240870786516854, + "grad_norm": 0.9133177399635315, + "learning_rate": 4.971100759677064e-06, + "loss": 0.7163, + "step": 923 + }, + { + "epoch": 0.324438202247191, + "grad_norm": 0.9289098381996155, + "learning_rate": 4.971030624024319e-06, + "loss": 0.7448, + "step": 924 + }, + { + "epoch": 0.32478932584269665, + "grad_norm": 0.872006356716156, + "learning_rate": 4.97096040386444e-06, + "loss": 0.6492, + "step": 925 + }, + { + "epoch": 0.32514044943820225, + "grad_norm": 0.9258365035057068, + "learning_rate": 4.97089009919983e-06, + "loss": 0.6931, + "step": 926 + }, + { + "epoch": 0.32549157303370785, + "grad_norm": 0.9044914841651917, + "learning_rate": 4.970819710032893e-06, + "loss": 0.7002, + "step": 927 + }, + { + "epoch": 0.3258426966292135, + "grad_norm": 0.9293612837791443, + "learning_rate": 4.970749236366037e-06, + "loss": 0.7471, + "step": 928 + }, + { + "epoch": 0.3261938202247191, + "grad_norm": 0.8999515771865845, + "learning_rate": 4.9706786782016706e-06, + "loss": 0.6992, + "step": 929 + }, + { + "epoch": 0.3265449438202247, + "grad_norm": 0.9285963177680969, + "learning_rate": 4.970608035542207e-06, + "loss": 0.7229, + "step": 930 + }, + { + "epoch": 0.32689606741573035, + "grad_norm": 0.9883245229721069, + "learning_rate": 4.970537308390063e-06, + "loss": 0.7343, + "step": 931 + }, + { + "epoch": 0.32724719101123595, + "grad_norm": 0.9650030732154846, + "learning_rate": 4.970466496747658e-06, + "loss": 0.7206, + "step": 932 + }, + { + "epoch": 0.32759831460674155, + "grad_norm": 0.9341951012611389, + "learning_rate": 4.9703956006174125e-06, + "loss": 0.7029, + "step": 933 + }, + { + "epoch": 0.3279494382022472, + "grad_norm": 0.9105697274208069, + "learning_rate": 4.9703246200017504e-06, + "loss": 0.7127, + "step": 934 + }, + { + "epoch": 0.3283005617977528, + "grad_norm": 0.8891416192054749, + "learning_rate": 4.970253554903102e-06, + "loss": 0.6952, + "step": 935 + }, + { + "epoch": 0.32865168539325845, + "grad_norm": 0.9152025580406189, + "learning_rate": 4.970182405323894e-06, + "loss": 0.7126, + "step": 936 + }, + { + "epoch": 0.32900280898876405, + "grad_norm": 0.9279428720474243, + "learning_rate": 4.970111171266563e-06, + "loss": 0.7066, + "step": 937 + }, + { + "epoch": 0.32935393258426965, + "grad_norm": 0.9324163198471069, + "learning_rate": 4.970039852733542e-06, + "loss": 0.6796, + "step": 938 + }, + { + "epoch": 0.3297050561797753, + "grad_norm": 0.9290011525154114, + "learning_rate": 4.9699684497272736e-06, + "loss": 0.7192, + "step": 939 + }, + { + "epoch": 0.3300561797752809, + "grad_norm": 0.8959920406341553, + "learning_rate": 4.969896962250196e-06, + "loss": 0.7132, + "step": 940 + }, + { + "epoch": 0.3304073033707865, + "grad_norm": 0.8849387168884277, + "learning_rate": 4.969825390304757e-06, + "loss": 0.6618, + "step": 941 + }, + { + "epoch": 0.33075842696629215, + "grad_norm": 0.8804863691329956, + "learning_rate": 4.969753733893403e-06, + "loss": 0.7021, + "step": 942 + }, + { + "epoch": 0.33110955056179775, + "grad_norm": 0.9022026062011719, + "learning_rate": 4.969681993018584e-06, + "loss": 0.6842, + "step": 943 + }, + { + "epoch": 0.33146067415730335, + "grad_norm": 0.9250811338424683, + "learning_rate": 4.969610167682754e-06, + "loss": 0.6977, + "step": 944 + }, + { + "epoch": 0.331811797752809, + "grad_norm": 0.9264232516288757, + "learning_rate": 4.969538257888371e-06, + "loss": 0.7135, + "step": 945 + }, + { + "epoch": 0.3321629213483146, + "grad_norm": 0.9282289147377014, + "learning_rate": 4.969466263637892e-06, + "loss": 0.7304, + "step": 946 + }, + { + "epoch": 0.3325140449438202, + "grad_norm": 0.8798340559005737, + "learning_rate": 4.9693941849337804e-06, + "loss": 0.6996, + "step": 947 + }, + { + "epoch": 0.33286516853932585, + "grad_norm": 0.9778987169265747, + "learning_rate": 4.9693220217785e-06, + "loss": 0.7132, + "step": 948 + }, + { + "epoch": 0.33321629213483145, + "grad_norm": 0.9029282331466675, + "learning_rate": 4.969249774174519e-06, + "loss": 0.6951, + "step": 949 + }, + { + "epoch": 0.3335674157303371, + "grad_norm": 0.9430015087127686, + "learning_rate": 4.96917744212431e-06, + "loss": 0.6943, + "step": 950 + }, + { + "epoch": 0.3339185393258427, + "grad_norm": 0.9794940948486328, + "learning_rate": 4.9691050256303445e-06, + "loss": 0.7054, + "step": 951 + }, + { + "epoch": 0.3342696629213483, + "grad_norm": 0.9288370013237, + "learning_rate": 4.969032524695101e-06, + "loss": 0.7122, + "step": 952 + }, + { + "epoch": 0.33462078651685395, + "grad_norm": 0.9557552933692932, + "learning_rate": 4.968959939321057e-06, + "loss": 0.7598, + "step": 953 + }, + { + "epoch": 0.33497191011235955, + "grad_norm": 0.9375066161155701, + "learning_rate": 4.968887269510696e-06, + "loss": 0.6924, + "step": 954 + }, + { + "epoch": 0.33532303370786515, + "grad_norm": 0.9338163733482361, + "learning_rate": 4.968814515266503e-06, + "loss": 0.7267, + "step": 955 + }, + { + "epoch": 0.3356741573033708, + "grad_norm": 0.9154375791549683, + "learning_rate": 4.968741676590966e-06, + "loss": 0.7321, + "step": 956 + }, + { + "epoch": 0.3360252808988764, + "grad_norm": 0.9233065843582153, + "learning_rate": 4.968668753486576e-06, + "loss": 0.7245, + "step": 957 + }, + { + "epoch": 0.336376404494382, + "grad_norm": 0.9419113993644714, + "learning_rate": 4.968595745955828e-06, + "loss": 0.7007, + "step": 958 + }, + { + "epoch": 0.33672752808988765, + "grad_norm": 0.961121678352356, + "learning_rate": 4.968522654001217e-06, + "loss": 0.7308, + "step": 959 + }, + { + "epoch": 0.33707865168539325, + "grad_norm": 0.934877336025238, + "learning_rate": 4.9684494776252445e-06, + "loss": 0.7388, + "step": 960 + }, + { + "epoch": 0.3374297752808989, + "grad_norm": 0.912398636341095, + "learning_rate": 4.9683762168304115e-06, + "loss": 0.7088, + "step": 961 + }, + { + "epoch": 0.3377808988764045, + "grad_norm": 0.948833167552948, + "learning_rate": 4.9683028716192246e-06, + "loss": 0.7278, + "step": 962 + }, + { + "epoch": 0.3381320224719101, + "grad_norm": 0.9257057905197144, + "learning_rate": 4.968229441994191e-06, + "loss": 0.7338, + "step": 963 + }, + { + "epoch": 0.33848314606741575, + "grad_norm": 0.9129422903060913, + "learning_rate": 4.968155927957823e-06, + "loss": 0.7058, + "step": 964 + }, + { + "epoch": 0.33883426966292135, + "grad_norm": 0.9381750822067261, + "learning_rate": 4.968082329512634e-06, + "loss": 0.7326, + "step": 965 + }, + { + "epoch": 0.33918539325842695, + "grad_norm": 0.9108783602714539, + "learning_rate": 4.968008646661141e-06, + "loss": 0.66, + "step": 966 + }, + { + "epoch": 0.3395365168539326, + "grad_norm": 0.9329748153686523, + "learning_rate": 4.9679348794058655e-06, + "loss": 0.7117, + "step": 967 + }, + { + "epoch": 0.3398876404494382, + "grad_norm": 0.9643568992614746, + "learning_rate": 4.967861027749327e-06, + "loss": 0.7555, + "step": 968 + }, + { + "epoch": 0.3402387640449438, + "grad_norm": 0.9551818370819092, + "learning_rate": 4.967787091694055e-06, + "loss": 0.7474, + "step": 969 + }, + { + "epoch": 0.34058988764044945, + "grad_norm": 0.9381354451179504, + "learning_rate": 4.967713071242576e-06, + "loss": 0.7155, + "step": 970 + }, + { + "epoch": 0.34094101123595505, + "grad_norm": 0.9513474702835083, + "learning_rate": 4.967638966397421e-06, + "loss": 0.765, + "step": 971 + }, + { + "epoch": 0.34129213483146065, + "grad_norm": 0.9263571500778198, + "learning_rate": 4.967564777161124e-06, + "loss": 0.7112, + "step": 972 + }, + { + "epoch": 0.3416432584269663, + "grad_norm": 0.9278532862663269, + "learning_rate": 4.967490503536225e-06, + "loss": 0.7183, + "step": 973 + }, + { + "epoch": 0.3419943820224719, + "grad_norm": 0.9487691521644592, + "learning_rate": 4.967416145525261e-06, + "loss": 0.7312, + "step": 974 + }, + { + "epoch": 0.34234550561797755, + "grad_norm": 0.8866797089576721, + "learning_rate": 4.967341703130777e-06, + "loss": 0.704, + "step": 975 + }, + { + "epoch": 0.34269662921348315, + "grad_norm": 0.9202519655227661, + "learning_rate": 4.967267176355318e-06, + "loss": 0.7434, + "step": 976 + }, + { + "epoch": 0.34304775280898875, + "grad_norm": 0.9155771732330322, + "learning_rate": 4.967192565201432e-06, + "loss": 0.7166, + "step": 977 + }, + { + "epoch": 0.3433988764044944, + "grad_norm": 0.9582788944244385, + "learning_rate": 4.967117869671672e-06, + "loss": 0.7536, + "step": 978 + }, + { + "epoch": 0.34375, + "grad_norm": 0.9524344801902771, + "learning_rate": 4.967043089768592e-06, + "loss": 0.7739, + "step": 979 + }, + { + "epoch": 0.3441011235955056, + "grad_norm": 0.9230315685272217, + "learning_rate": 4.96696822549475e-06, + "loss": 0.7181, + "step": 980 + }, + { + "epoch": 0.34445224719101125, + "grad_norm": 0.9080458283424377, + "learning_rate": 4.966893276852704e-06, + "loss": 0.7107, + "step": 981 + }, + { + "epoch": 0.34480337078651685, + "grad_norm": 0.9102587699890137, + "learning_rate": 4.966818243845021e-06, + "loss": 0.7051, + "step": 982 + }, + { + "epoch": 0.34515449438202245, + "grad_norm": 0.8925254940986633, + "learning_rate": 4.966743126474264e-06, + "loss": 0.7106, + "step": 983 + }, + { + "epoch": 0.3455056179775281, + "grad_norm": 0.9102403521537781, + "learning_rate": 4.966667924743002e-06, + "loss": 0.6849, + "step": 984 + }, + { + "epoch": 0.3458567415730337, + "grad_norm": 0.9164739847183228, + "learning_rate": 4.966592638653808e-06, + "loss": 0.7027, + "step": 985 + }, + { + "epoch": 0.34620786516853935, + "grad_norm": 0.89588463306427, + "learning_rate": 4.966517268209258e-06, + "loss": 0.7086, + "step": 986 + }, + { + "epoch": 0.34655898876404495, + "grad_norm": 0.8812805414199829, + "learning_rate": 4.966441813411928e-06, + "loss": 0.7073, + "step": 987 + }, + { + "epoch": 0.34691011235955055, + "grad_norm": 0.8935489654541016, + "learning_rate": 4.966366274264397e-06, + "loss": 0.6749, + "step": 988 + }, + { + "epoch": 0.3472612359550562, + "grad_norm": 0.935444712638855, + "learning_rate": 4.966290650769251e-06, + "loss": 0.7283, + "step": 989 + }, + { + "epoch": 0.3476123595505618, + "grad_norm": 0.9046710729598999, + "learning_rate": 4.966214942929074e-06, + "loss": 0.7138, + "step": 990 + }, + { + "epoch": 0.3479634831460674, + "grad_norm": 0.9328753352165222, + "learning_rate": 4.966139150746458e-06, + "loss": 0.7022, + "step": 991 + }, + { + "epoch": 0.34831460674157305, + "grad_norm": 0.9144490361213684, + "learning_rate": 4.966063274223992e-06, + "loss": 0.738, + "step": 992 + }, + { + "epoch": 0.34866573033707865, + "grad_norm": 0.9606062173843384, + "learning_rate": 4.965987313364272e-06, + "loss": 0.6937, + "step": 993 + }, + { + "epoch": 0.34901685393258425, + "grad_norm": 0.94340980052948, + "learning_rate": 4.965911268169897e-06, + "loss": 0.6811, + "step": 994 + }, + { + "epoch": 0.3493679775280899, + "grad_norm": 0.8755514025688171, + "learning_rate": 4.9658351386434665e-06, + "loss": 0.6707, + "step": 995 + }, + { + "epoch": 0.3497191011235955, + "grad_norm": 0.8640532493591309, + "learning_rate": 4.965758924787584e-06, + "loss": 0.6378, + "step": 996 + }, + { + "epoch": 0.3500702247191011, + "grad_norm": 0.9261877536773682, + "learning_rate": 4.965682626604856e-06, + "loss": 0.6953, + "step": 997 + }, + { + "epoch": 0.35042134831460675, + "grad_norm": 0.8998399376869202, + "learning_rate": 4.965606244097894e-06, + "loss": 0.7154, + "step": 998 + }, + { + "epoch": 0.35077247191011235, + "grad_norm": 0.9440179467201233, + "learning_rate": 4.9655297772693066e-06, + "loss": 0.6992, + "step": 999 + }, + { + "epoch": 0.351123595505618, + "grad_norm": 0.9105573892593384, + "learning_rate": 4.965453226121711e-06, + "loss": 0.7033, + "step": 1000 + }, + { + "epoch": 0.3514747191011236, + "grad_norm": 0.9109785556793213, + "learning_rate": 4.965376590657724e-06, + "loss": 0.7057, + "step": 1001 + }, + { + "epoch": 0.3518258426966292, + "grad_norm": 0.9462383985519409, + "learning_rate": 4.9652998708799684e-06, + "loss": 0.6979, + "step": 1002 + }, + { + "epoch": 0.35217696629213485, + "grad_norm": 0.9546105861663818, + "learning_rate": 4.965223066791066e-06, + "loss": 0.7227, + "step": 1003 + }, + { + "epoch": 0.35252808988764045, + "grad_norm": 1.0831300020217896, + "learning_rate": 4.965146178393645e-06, + "loss": 0.6849, + "step": 1004 + }, + { + "epoch": 0.35287921348314605, + "grad_norm": 0.9323068261146545, + "learning_rate": 4.965069205690334e-06, + "loss": 0.739, + "step": 1005 + }, + { + "epoch": 0.3532303370786517, + "grad_norm": 0.9407396912574768, + "learning_rate": 4.964992148683765e-06, + "loss": 0.7386, + "step": 1006 + }, + { + "epoch": 0.3535814606741573, + "grad_norm": 0.9513671398162842, + "learning_rate": 4.964915007376574e-06, + "loss": 0.666, + "step": 1007 + }, + { + "epoch": 0.3539325842696629, + "grad_norm": 0.9194923043251038, + "learning_rate": 4.964837781771399e-06, + "loss": 0.7601, + "step": 1008 + }, + { + "epoch": 0.35428370786516855, + "grad_norm": 0.9639756083488464, + "learning_rate": 4.964760471870882e-06, + "loss": 0.7124, + "step": 1009 + }, + { + "epoch": 0.35463483146067415, + "grad_norm": 0.9628319144248962, + "learning_rate": 4.964683077677664e-06, + "loss": 0.7003, + "step": 1010 + }, + { + "epoch": 0.3549859550561798, + "grad_norm": 0.9003061652183533, + "learning_rate": 4.964605599194395e-06, + "loss": 0.7199, + "step": 1011 + }, + { + "epoch": 0.3553370786516854, + "grad_norm": 0.9365134835243225, + "learning_rate": 4.964528036423723e-06, + "loss": 0.6651, + "step": 1012 + }, + { + "epoch": 0.355688202247191, + "grad_norm": 0.9309297204017639, + "learning_rate": 4.964450389368303e-06, + "loss": 0.7354, + "step": 1013 + }, + { + "epoch": 0.35603932584269665, + "grad_norm": 0.9624804258346558, + "learning_rate": 4.964372658030786e-06, + "loss": 0.6955, + "step": 1014 + }, + { + "epoch": 0.35639044943820225, + "grad_norm": 0.9497263431549072, + "learning_rate": 4.964294842413835e-06, + "loss": 0.7374, + "step": 1015 + }, + { + "epoch": 0.35674157303370785, + "grad_norm": 0.9235740900039673, + "learning_rate": 4.964216942520108e-06, + "loss": 0.6906, + "step": 1016 + }, + { + "epoch": 0.3570926966292135, + "grad_norm": 0.9985111951828003, + "learning_rate": 4.9641389583522705e-06, + "loss": 0.7131, + "step": 1017 + }, + { + "epoch": 0.3574438202247191, + "grad_norm": 0.9399976134300232, + "learning_rate": 4.964060889912988e-06, + "loss": 0.6935, + "step": 1018 + }, + { + "epoch": 0.3577949438202247, + "grad_norm": 0.9676353931427002, + "learning_rate": 4.9639827372049335e-06, + "loss": 0.7513, + "step": 1019 + }, + { + "epoch": 0.35814606741573035, + "grad_norm": 0.9015031456947327, + "learning_rate": 4.9639045002307765e-06, + "loss": 0.687, + "step": 1020 + }, + { + "epoch": 0.35849719101123595, + "grad_norm": 0.9401846528053284, + "learning_rate": 4.963826178993195e-06, + "loss": 0.6963, + "step": 1021 + }, + { + "epoch": 0.35884831460674155, + "grad_norm": 1.00020170211792, + "learning_rate": 4.9637477734948654e-06, + "loss": 0.6747, + "step": 1022 + }, + { + "epoch": 0.3591994382022472, + "grad_norm": 0.8997200727462769, + "learning_rate": 4.963669283738471e-06, + "loss": 0.7253, + "step": 1023 + }, + { + "epoch": 0.3595505617977528, + "grad_norm": 0.8761312365531921, + "learning_rate": 4.963590709726695e-06, + "loss": 0.6832, + "step": 1024 + }, + { + "epoch": 0.35990168539325845, + "grad_norm": 0.9415677189826965, + "learning_rate": 4.963512051462226e-06, + "loss": 0.6998, + "step": 1025 + }, + { + "epoch": 0.36025280898876405, + "grad_norm": 0.9139187932014465, + "learning_rate": 4.963433308947751e-06, + "loss": 0.7292, + "step": 1026 + }, + { + "epoch": 0.36060393258426965, + "grad_norm": 0.9207375049591064, + "learning_rate": 4.963354482185966e-06, + "loss": 0.6779, + "step": 1027 + }, + { + "epoch": 0.3609550561797753, + "grad_norm": 0.9517902135848999, + "learning_rate": 4.963275571179565e-06, + "loss": 0.7525, + "step": 1028 + }, + { + "epoch": 0.3613061797752809, + "grad_norm": 0.9088241457939148, + "learning_rate": 4.963196575931248e-06, + "loss": 0.7196, + "step": 1029 + }, + { + "epoch": 0.3616573033707865, + "grad_norm": 0.8898507356643677, + "learning_rate": 4.963117496443715e-06, + "loss": 0.7205, + "step": 1030 + }, + { + "epoch": 0.36200842696629215, + "grad_norm": 0.9045488834381104, + "learning_rate": 4.963038332719672e-06, + "loss": 0.6937, + "step": 1031 + }, + { + "epoch": 0.36235955056179775, + "grad_norm": 0.9505866169929504, + "learning_rate": 4.962959084761826e-06, + "loss": 0.7006, + "step": 1032 + }, + { + "epoch": 0.36271067415730335, + "grad_norm": 0.8908640742301941, + "learning_rate": 4.962879752572885e-06, + "loss": 0.7191, + "step": 1033 + }, + { + "epoch": 0.363061797752809, + "grad_norm": 0.925959050655365, + "learning_rate": 4.962800336155567e-06, + "loss": 0.6963, + "step": 1034 + }, + { + "epoch": 0.3634129213483146, + "grad_norm": 0.9082480072975159, + "learning_rate": 4.962720835512583e-06, + "loss": 0.7084, + "step": 1035 + }, + { + "epoch": 0.3637640449438202, + "grad_norm": 0.9313839077949524, + "learning_rate": 4.962641250646654e-06, + "loss": 0.7425, + "step": 1036 + }, + { + "epoch": 0.36411516853932585, + "grad_norm": 0.9484413266181946, + "learning_rate": 4.962561581560501e-06, + "loss": 0.6814, + "step": 1037 + }, + { + "epoch": 0.36446629213483145, + "grad_norm": 0.9550908207893372, + "learning_rate": 4.96248182825685e-06, + "loss": 0.6747, + "step": 1038 + }, + { + "epoch": 0.3648174157303371, + "grad_norm": 0.9393624663352966, + "learning_rate": 4.962401990738428e-06, + "loss": 0.7172, + "step": 1039 + }, + { + "epoch": 0.3651685393258427, + "grad_norm": 0.961174726486206, + "learning_rate": 4.962322069007964e-06, + "loss": 0.6967, + "step": 1040 + }, + { + "epoch": 0.3655196629213483, + "grad_norm": 0.9074354767799377, + "learning_rate": 4.962242063068193e-06, + "loss": 0.7019, + "step": 1041 + }, + { + "epoch": 0.36587078651685395, + "grad_norm": 0.9572650194168091, + "learning_rate": 4.962161972921851e-06, + "loss": 0.717, + "step": 1042 + }, + { + "epoch": 0.36622191011235955, + "grad_norm": 0.9262009263038635, + "learning_rate": 4.962081798571675e-06, + "loss": 0.7332, + "step": 1043 + }, + { + "epoch": 0.36657303370786515, + "grad_norm": 0.8846391439437866, + "learning_rate": 4.9620015400204094e-06, + "loss": 0.7359, + "step": 1044 + }, + { + "epoch": 0.3669241573033708, + "grad_norm": 0.9351332187652588, + "learning_rate": 4.961921197270797e-06, + "loss": 0.7809, + "step": 1045 + }, + { + "epoch": 0.3672752808988764, + "grad_norm": 0.9081857204437256, + "learning_rate": 4.961840770325587e-06, + "loss": 0.7169, + "step": 1046 + }, + { + "epoch": 0.367626404494382, + "grad_norm": 0.9414200782775879, + "learning_rate": 4.961760259187529e-06, + "loss": 0.7154, + "step": 1047 + }, + { + "epoch": 0.36797752808988765, + "grad_norm": 0.9009916186332703, + "learning_rate": 4.961679663859377e-06, + "loss": 0.7013, + "step": 1048 + }, + { + "epoch": 0.36832865168539325, + "grad_norm": 0.9196215271949768, + "learning_rate": 4.961598984343886e-06, + "loss": 0.706, + "step": 1049 + }, + { + "epoch": 0.3686797752808989, + "grad_norm": 0.947223424911499, + "learning_rate": 4.961518220643817e-06, + "loss": 0.712, + "step": 1050 + }, + { + "epoch": 0.3690308988764045, + "grad_norm": 0.8798839449882507, + "learning_rate": 4.961437372761932e-06, + "loss": 0.6821, + "step": 1051 + }, + { + "epoch": 0.3693820224719101, + "grad_norm": 0.9458513259887695, + "learning_rate": 4.961356440700994e-06, + "loss": 0.7503, + "step": 1052 + }, + { + "epoch": 0.36973314606741575, + "grad_norm": 0.9466217160224915, + "learning_rate": 4.961275424463772e-06, + "loss": 0.7086, + "step": 1053 + }, + { + "epoch": 0.37008426966292135, + "grad_norm": 0.9832736849784851, + "learning_rate": 4.961194324053037e-06, + "loss": 0.7332, + "step": 1054 + }, + { + "epoch": 0.37043539325842695, + "grad_norm": 0.9869272708892822, + "learning_rate": 4.961113139471562e-06, + "loss": 0.7595, + "step": 1055 + }, + { + "epoch": 0.3707865168539326, + "grad_norm": 0.9590395092964172, + "learning_rate": 4.961031870722123e-06, + "loss": 0.7471, + "step": 1056 + }, + { + "epoch": 0.3711376404494382, + "grad_norm": 0.9194419980049133, + "learning_rate": 4.9609505178075015e-06, + "loss": 0.7147, + "step": 1057 + }, + { + "epoch": 0.3714887640449438, + "grad_norm": 0.8812661170959473, + "learning_rate": 4.960869080730477e-06, + "loss": 0.6723, + "step": 1058 + }, + { + "epoch": 0.37183988764044945, + "grad_norm": 0.9303286671638489, + "learning_rate": 4.960787559493836e-06, + "loss": 0.7253, + "step": 1059 + }, + { + "epoch": 0.37219101123595505, + "grad_norm": 0.9536207318305969, + "learning_rate": 4.960705954100366e-06, + "loss": 0.7325, + "step": 1060 + }, + { + "epoch": 0.37254213483146065, + "grad_norm": 0.9165772795677185, + "learning_rate": 4.960624264552858e-06, + "loss": 0.6949, + "step": 1061 + }, + { + "epoch": 0.3728932584269663, + "grad_norm": 0.9144523739814758, + "learning_rate": 4.9605424908541065e-06, + "loss": 0.7355, + "step": 1062 + }, + { + "epoch": 0.3732443820224719, + "grad_norm": 0.8873782753944397, + "learning_rate": 4.960460633006906e-06, + "loss": 0.6819, + "step": 1063 + }, + { + "epoch": 0.37359550561797755, + "grad_norm": 0.9444409012794495, + "learning_rate": 4.960378691014058e-06, + "loss": 0.6851, + "step": 1064 + }, + { + "epoch": 0.37394662921348315, + "grad_norm": 0.959851861000061, + "learning_rate": 4.960296664878364e-06, + "loss": 0.7127, + "step": 1065 + }, + { + "epoch": 0.37429775280898875, + "grad_norm": 0.9353641867637634, + "learning_rate": 4.9602145546026305e-06, + "loss": 0.7208, + "step": 1066 + }, + { + "epoch": 0.3746488764044944, + "grad_norm": 0.9105174541473389, + "learning_rate": 4.960132360189663e-06, + "loss": 0.6717, + "step": 1067 + }, + { + "epoch": 0.375, + "grad_norm": 0.9150325655937195, + "learning_rate": 4.960050081642274e-06, + "loss": 0.7233, + "step": 1068 + }, + { + "epoch": 0.3753511235955056, + "grad_norm": 0.9387450814247131, + "learning_rate": 4.959967718963279e-06, + "loss": 0.7104, + "step": 1069 + }, + { + "epoch": 0.37570224719101125, + "grad_norm": 0.9268593788146973, + "learning_rate": 4.959885272155491e-06, + "loss": 0.7209, + "step": 1070 + }, + { + "epoch": 0.37605337078651685, + "grad_norm": 0.9124265909194946, + "learning_rate": 4.959802741221733e-06, + "loss": 0.6982, + "step": 1071 + }, + { + "epoch": 0.37640449438202245, + "grad_norm": 0.9302802681922913, + "learning_rate": 4.9597201261648255e-06, + "loss": 0.7105, + "step": 1072 + }, + { + "epoch": 0.3767556179775281, + "grad_norm": 0.9560679793357849, + "learning_rate": 4.959637426987595e-06, + "loss": 0.7115, + "step": 1073 + }, + { + "epoch": 0.3771067415730337, + "grad_norm": 0.9246500730514526, + "learning_rate": 4.9595546436928685e-06, + "loss": 0.7319, + "step": 1074 + }, + { + "epoch": 0.37745786516853935, + "grad_norm": 0.9188322424888611, + "learning_rate": 4.959471776283479e-06, + "loss": 0.6829, + "step": 1075 + }, + { + "epoch": 0.37780898876404495, + "grad_norm": 0.9240451455116272, + "learning_rate": 4.959388824762258e-06, + "loss": 0.7183, + "step": 1076 + }, + { + "epoch": 0.37816011235955055, + "grad_norm": 0.8934212923049927, + "learning_rate": 4.959305789132044e-06, + "loss": 0.6982, + "step": 1077 + }, + { + "epoch": 0.3785112359550562, + "grad_norm": 0.9386299848556519, + "learning_rate": 4.959222669395677e-06, + "loss": 0.6942, + "step": 1078 + }, + { + "epoch": 0.3788623595505618, + "grad_norm": 0.9435991644859314, + "learning_rate": 4.959139465555999e-06, + "loss": 0.7382, + "step": 1079 + }, + { + "epoch": 0.3792134831460674, + "grad_norm": 0.9004935026168823, + "learning_rate": 4.959056177615855e-06, + "loss": 0.6944, + "step": 1080 + }, + { + "epoch": 0.37956460674157305, + "grad_norm": 0.9269630312919617, + "learning_rate": 4.958972805578095e-06, + "loss": 0.7037, + "step": 1081 + }, + { + "epoch": 0.37991573033707865, + "grad_norm": 0.9682616591453552, + "learning_rate": 4.9588893494455684e-06, + "loss": 0.6874, + "step": 1082 + }, + { + "epoch": 0.38026685393258425, + "grad_norm": 0.8845170140266418, + "learning_rate": 4.958805809221131e-06, + "loss": 0.6945, + "step": 1083 + }, + { + "epoch": 0.3806179775280899, + "grad_norm": 0.940234363079071, + "learning_rate": 4.958722184907638e-06, + "loss": 0.6813, + "step": 1084 + }, + { + "epoch": 0.3809691011235955, + "grad_norm": 0.9013071656227112, + "learning_rate": 4.95863847650795e-06, + "loss": 0.694, + "step": 1085 + }, + { + "epoch": 0.3813202247191011, + "grad_norm": 0.8996568322181702, + "learning_rate": 4.958554684024931e-06, + "loss": 0.6558, + "step": 1086 + }, + { + "epoch": 0.38167134831460675, + "grad_norm": 0.9193875193595886, + "learning_rate": 4.9584708074614455e-06, + "loss": 0.6996, + "step": 1087 + }, + { + "epoch": 0.38202247191011235, + "grad_norm": 0.9180407524108887, + "learning_rate": 4.958386846820361e-06, + "loss": 0.7147, + "step": 1088 + }, + { + "epoch": 0.382373595505618, + "grad_norm": 1.0345251560211182, + "learning_rate": 4.958302802104551e-06, + "loss": 0.7393, + "step": 1089 + }, + { + "epoch": 0.3827247191011236, + "grad_norm": 0.9022161364555359, + "learning_rate": 4.958218673316889e-06, + "loss": 0.6949, + "step": 1090 + }, + { + "epoch": 0.3830758426966292, + "grad_norm": 0.9014359712600708, + "learning_rate": 4.958134460460251e-06, + "loss": 0.6947, + "step": 1091 + }, + { + "epoch": 0.38342696629213485, + "grad_norm": 0.9596666693687439, + "learning_rate": 4.958050163537519e-06, + "loss": 0.7763, + "step": 1092 + }, + { + "epoch": 0.38377808988764045, + "grad_norm": 0.8888493180274963, + "learning_rate": 4.957965782551574e-06, + "loss": 0.7093, + "step": 1093 + }, + { + "epoch": 0.38412921348314605, + "grad_norm": 0.9068475365638733, + "learning_rate": 4.957881317505303e-06, + "loss": 0.6893, + "step": 1094 + }, + { + "epoch": 0.3844803370786517, + "grad_norm": 0.8918416500091553, + "learning_rate": 4.957796768401594e-06, + "loss": 0.6926, + "step": 1095 + }, + { + "epoch": 0.3848314606741573, + "grad_norm": 0.8772871494293213, + "learning_rate": 4.957712135243339e-06, + "loss": 0.6714, + "step": 1096 + }, + { + "epoch": 0.3851825842696629, + "grad_norm": 0.9466787576675415, + "learning_rate": 4.9576274180334315e-06, + "loss": 0.693, + "step": 1097 + }, + { + "epoch": 0.38553370786516855, + "grad_norm": 0.8940508961677551, + "learning_rate": 4.957542616774768e-06, + "loss": 0.6727, + "step": 1098 + }, + { + "epoch": 0.38588483146067415, + "grad_norm": 0.92746901512146, + "learning_rate": 4.957457731470253e-06, + "loss": 0.6832, + "step": 1099 + }, + { + "epoch": 0.3862359550561798, + "grad_norm": 0.9406936168670654, + "learning_rate": 4.9573727621227845e-06, + "loss": 0.7121, + "step": 1100 + }, + { + "epoch": 0.3865870786516854, + "grad_norm": 0.9280776381492615, + "learning_rate": 4.957287708735271e-06, + "loss": 0.7149, + "step": 1101 + }, + { + "epoch": 0.386938202247191, + "grad_norm": 0.9443319439888, + "learning_rate": 4.9572025713106195e-06, + "loss": 0.7205, + "step": 1102 + }, + { + "epoch": 0.38728932584269665, + "grad_norm": 0.9101063013076782, + "learning_rate": 4.957117349851745e-06, + "loss": 0.703, + "step": 1103 + }, + { + "epoch": 0.38764044943820225, + "grad_norm": 0.8856722116470337, + "learning_rate": 4.957032044361557e-06, + "loss": 0.7079, + "step": 1104 + }, + { + "epoch": 0.38799157303370785, + "grad_norm": 0.928821325302124, + "learning_rate": 4.956946654842977e-06, + "loss": 0.6935, + "step": 1105 + }, + { + "epoch": 0.3883426966292135, + "grad_norm": 0.9314661026000977, + "learning_rate": 4.956861181298924e-06, + "loss": 0.7062, + "step": 1106 + }, + { + "epoch": 0.3886938202247191, + "grad_norm": 0.9111589193344116, + "learning_rate": 4.956775623732321e-06, + "loss": 0.7343, + "step": 1107 + }, + { + "epoch": 0.3890449438202247, + "grad_norm": 0.8977031707763672, + "learning_rate": 4.9566899821460935e-06, + "loss": 0.6702, + "step": 1108 + }, + { + "epoch": 0.38939606741573035, + "grad_norm": 0.9246605634689331, + "learning_rate": 4.956604256543171e-06, + "loss": 0.7062, + "step": 1109 + }, + { + "epoch": 0.38974719101123595, + "grad_norm": 0.9252159595489502, + "learning_rate": 4.956518446926485e-06, + "loss": 0.6981, + "step": 1110 + }, + { + "epoch": 0.39009831460674155, + "grad_norm": 0.9128794074058533, + "learning_rate": 4.956432553298971e-06, + "loss": 0.6802, + "step": 1111 + }, + { + "epoch": 0.3904494382022472, + "grad_norm": 0.9178376197814941, + "learning_rate": 4.956346575663564e-06, + "loss": 0.6753, + "step": 1112 + }, + { + "epoch": 0.3908005617977528, + "grad_norm": 0.8804501891136169, + "learning_rate": 4.956260514023207e-06, + "loss": 0.7378, + "step": 1113 + }, + { + "epoch": 0.39115168539325845, + "grad_norm": 0.9001629948616028, + "learning_rate": 4.956174368380842e-06, + "loss": 0.7061, + "step": 1114 + }, + { + "epoch": 0.39150280898876405, + "grad_norm": 0.8731476664543152, + "learning_rate": 4.956088138739416e-06, + "loss": 0.6749, + "step": 1115 + }, + { + "epoch": 0.39185393258426965, + "grad_norm": 0.9055764079093933, + "learning_rate": 4.956001825101877e-06, + "loss": 0.7172, + "step": 1116 + }, + { + "epoch": 0.3922050561797753, + "grad_norm": 0.9398018717765808, + "learning_rate": 4.955915427471178e-06, + "loss": 0.711, + "step": 1117 + }, + { + "epoch": 0.3925561797752809, + "grad_norm": 0.9070323705673218, + "learning_rate": 4.9558289458502715e-06, + "loss": 0.7467, + "step": 1118 + }, + { + "epoch": 0.3929073033707865, + "grad_norm": 0.916070818901062, + "learning_rate": 4.9557423802421166e-06, + "loss": 0.6929, + "step": 1119 + }, + { + "epoch": 0.39325842696629215, + "grad_norm": 0.9212610721588135, + "learning_rate": 4.955655730649674e-06, + "loss": 0.7098, + "step": 1120 + }, + { + "epoch": 0.39360955056179775, + "grad_norm": 0.8813148736953735, + "learning_rate": 4.955568997075908e-06, + "loss": 0.7241, + "step": 1121 + }, + { + "epoch": 0.39396067415730335, + "grad_norm": 0.960132360458374, + "learning_rate": 4.955482179523782e-06, + "loss": 0.7531, + "step": 1122 + }, + { + "epoch": 0.394311797752809, + "grad_norm": 0.8893491625785828, + "learning_rate": 4.955395277996268e-06, + "loss": 0.708, + "step": 1123 + }, + { + "epoch": 0.3946629213483146, + "grad_norm": 0.9316296577453613, + "learning_rate": 4.955308292496336e-06, + "loss": 0.7129, + "step": 1124 + }, + { + "epoch": 0.3950140449438202, + "grad_norm": 0.8866797685623169, + "learning_rate": 4.955221223026961e-06, + "loss": 0.6986, + "step": 1125 + }, + { + "epoch": 0.39536516853932585, + "grad_norm": 0.9073642492294312, + "learning_rate": 4.955134069591121e-06, + "loss": 0.7192, + "step": 1126 + }, + { + "epoch": 0.39571629213483145, + "grad_norm": 0.8996924757957458, + "learning_rate": 4.9550468321917975e-06, + "loss": 0.6902, + "step": 1127 + }, + { + "epoch": 0.3960674157303371, + "grad_norm": 0.9362848401069641, + "learning_rate": 4.9549595108319726e-06, + "loss": 0.7331, + "step": 1128 + }, + { + "epoch": 0.3964185393258427, + "grad_norm": 0.9147123098373413, + "learning_rate": 4.954872105514633e-06, + "loss": 0.731, + "step": 1129 + }, + { + "epoch": 0.3967696629213483, + "grad_norm": 0.9537502527236938, + "learning_rate": 4.954784616242769e-06, + "loss": 0.7347, + "step": 1130 + }, + { + "epoch": 0.39712078651685395, + "grad_norm": 0.9200554490089417, + "learning_rate": 4.9546970430193705e-06, + "loss": 0.7205, + "step": 1131 + }, + { + "epoch": 0.39747191011235955, + "grad_norm": 0.929223358631134, + "learning_rate": 4.954609385847434e-06, + "loss": 0.7108, + "step": 1132 + }, + { + "epoch": 0.39782303370786515, + "grad_norm": 0.9166401624679565, + "learning_rate": 4.954521644729958e-06, + "loss": 0.732, + "step": 1133 + }, + { + "epoch": 0.3981741573033708, + "grad_norm": 0.9084526300430298, + "learning_rate": 4.954433819669941e-06, + "loss": 0.7078, + "step": 1134 + }, + { + "epoch": 0.3985252808988764, + "grad_norm": 0.9210569858551025, + "learning_rate": 4.954345910670388e-06, + "loss": 0.6905, + "step": 1135 + }, + { + "epoch": 0.398876404494382, + "grad_norm": 0.890738308429718, + "learning_rate": 4.9542579177343054e-06, + "loss": 0.6975, + "step": 1136 + }, + { + "epoch": 0.39922752808988765, + "grad_norm": 0.8724343776702881, + "learning_rate": 4.954169840864701e-06, + "loss": 0.6323, + "step": 1137 + }, + { + "epoch": 0.39957865168539325, + "grad_norm": 0.9185892343521118, + "learning_rate": 4.95408168006459e-06, + "loss": 0.7106, + "step": 1138 + }, + { + "epoch": 0.3999297752808989, + "grad_norm": 0.9605926871299744, + "learning_rate": 4.953993435336983e-06, + "loss": 0.7208, + "step": 1139 + }, + { + "epoch": 0.4002808988764045, + "grad_norm": 0.9934019446372986, + "learning_rate": 4.953905106684902e-06, + "loss": 0.7394, + "step": 1140 + }, + { + "epoch": 0.4006320224719101, + "grad_norm": 0.8674790263175964, + "learning_rate": 4.953816694111367e-06, + "loss": 0.6588, + "step": 1141 + }, + { + "epoch": 0.40098314606741575, + "grad_norm": 0.9420093894004822, + "learning_rate": 4.9537281976194e-06, + "loss": 0.6934, + "step": 1142 + }, + { + "epoch": 0.40133426966292135, + "grad_norm": 0.8882573246955872, + "learning_rate": 4.953639617212028e-06, + "loss": 0.7244, + "step": 1143 + }, + { + "epoch": 0.40168539325842695, + "grad_norm": 1.0398857593536377, + "learning_rate": 4.95355095289228e-06, + "loss": 0.7402, + "step": 1144 + }, + { + "epoch": 0.4020365168539326, + "grad_norm": 0.9207865595817566, + "learning_rate": 4.95346220466319e-06, + "loss": 0.6942, + "step": 1145 + }, + { + "epoch": 0.4023876404494382, + "grad_norm": 0.9174979329109192, + "learning_rate": 4.9533733725277924e-06, + "loss": 0.7241, + "step": 1146 + }, + { + "epoch": 0.4027387640449438, + "grad_norm": 0.9054567217826843, + "learning_rate": 4.953284456489124e-06, + "loss": 0.6666, + "step": 1147 + }, + { + "epoch": 0.40308988764044945, + "grad_norm": 0.9376466870307922, + "learning_rate": 4.953195456550227e-06, + "loss": 0.718, + "step": 1148 + }, + { + "epoch": 0.40344101123595505, + "grad_norm": 0.9617171883583069, + "learning_rate": 4.953106372714144e-06, + "loss": 0.7226, + "step": 1149 + }, + { + "epoch": 0.40379213483146065, + "grad_norm": 0.8789179921150208, + "learning_rate": 4.953017204983923e-06, + "loss": 0.6981, + "step": 1150 + }, + { + "epoch": 0.4041432584269663, + "grad_norm": 0.8860397934913635, + "learning_rate": 4.952927953362612e-06, + "loss": 0.6894, + "step": 1151 + }, + { + "epoch": 0.4044943820224719, + "grad_norm": 0.9390692114830017, + "learning_rate": 4.952838617853265e-06, + "loss": 0.6788, + "step": 1152 + }, + { + "epoch": 0.40484550561797755, + "grad_norm": 0.9032955765724182, + "learning_rate": 4.952749198458935e-06, + "loss": 0.7417, + "step": 1153 + }, + { + "epoch": 0.40519662921348315, + "grad_norm": 0.9163577556610107, + "learning_rate": 4.952659695182682e-06, + "loss": 0.7195, + "step": 1154 + }, + { + "epoch": 0.40554775280898875, + "grad_norm": 0.8781930804252625, + "learning_rate": 4.952570108027566e-06, + "loss": 0.6691, + "step": 1155 + }, + { + "epoch": 0.4058988764044944, + "grad_norm": 0.887157678604126, + "learning_rate": 4.9524804369966505e-06, + "loss": 0.6993, + "step": 1156 + }, + { + "epoch": 0.40625, + "grad_norm": 0.9891164302825928, + "learning_rate": 4.9523906820930035e-06, + "loss": 0.7357, + "step": 1157 + }, + { + "epoch": 0.4066011235955056, + "grad_norm": 0.9419492483139038, + "learning_rate": 4.952300843319693e-06, + "loss": 0.674, + "step": 1158 + }, + { + "epoch": 0.40695224719101125, + "grad_norm": 0.9248577356338501, + "learning_rate": 4.952210920679793e-06, + "loss": 0.724, + "step": 1159 + }, + { + "epoch": 0.40730337078651685, + "grad_norm": 0.9277257919311523, + "learning_rate": 4.9521209141763775e-06, + "loss": 0.7179, + "step": 1160 + }, + { + "epoch": 0.40765449438202245, + "grad_norm": 0.9654591083526611, + "learning_rate": 4.952030823812525e-06, + "loss": 0.6796, + "step": 1161 + }, + { + "epoch": 0.4080056179775281, + "grad_norm": 0.9730082154273987, + "learning_rate": 4.951940649591316e-06, + "loss": 0.715, + "step": 1162 + }, + { + "epoch": 0.4083567415730337, + "grad_norm": 0.9230988621711731, + "learning_rate": 4.9518503915158356e-06, + "loss": 0.7218, + "step": 1163 + }, + { + "epoch": 0.40870786516853935, + "grad_norm": 0.9274692535400391, + "learning_rate": 4.951760049589169e-06, + "loss": 0.6773, + "step": 1164 + }, + { + "epoch": 0.40905898876404495, + "grad_norm": 0.894001841545105, + "learning_rate": 4.951669623814408e-06, + "loss": 0.7118, + "step": 1165 + }, + { + "epoch": 0.40941011235955055, + "grad_norm": 0.924102783203125, + "learning_rate": 4.951579114194643e-06, + "loss": 0.6648, + "step": 1166 + }, + { + "epoch": 0.4097612359550562, + "grad_norm": 0.9237919449806213, + "learning_rate": 4.95148852073297e-06, + "loss": 0.697, + "step": 1167 + }, + { + "epoch": 0.4101123595505618, + "grad_norm": 0.9063787460327148, + "learning_rate": 4.951397843432489e-06, + "loss": 0.699, + "step": 1168 + }, + { + "epoch": 0.4104634831460674, + "grad_norm": 0.8865030407905579, + "learning_rate": 4.951307082296297e-06, + "loss": 0.6768, + "step": 1169 + }, + { + "epoch": 0.41081460674157305, + "grad_norm": 0.9309532046318054, + "learning_rate": 4.9512162373275015e-06, + "loss": 0.7168, + "step": 1170 + }, + { + "epoch": 0.41116573033707865, + "grad_norm": 0.9091811776161194, + "learning_rate": 4.951125308529208e-06, + "loss": 0.6962, + "step": 1171 + }, + { + "epoch": 0.41151685393258425, + "grad_norm": 0.8983654975891113, + "learning_rate": 4.951034295904527e-06, + "loss": 0.7193, + "step": 1172 + }, + { + "epoch": 0.4118679775280899, + "grad_norm": 0.8983482718467712, + "learning_rate": 4.95094319945657e-06, + "loss": 0.7234, + "step": 1173 + }, + { + "epoch": 0.4122191011235955, + "grad_norm": 0.9178565740585327, + "learning_rate": 4.950852019188454e-06, + "loss": 0.7133, + "step": 1174 + }, + { + "epoch": 0.4125702247191011, + "grad_norm": 0.9206750988960266, + "learning_rate": 4.950760755103294e-06, + "loss": 0.7417, + "step": 1175 + }, + { + "epoch": 0.41292134831460675, + "grad_norm": 0.9141495227813721, + "learning_rate": 4.950669407204215e-06, + "loss": 0.7238, + "step": 1176 + }, + { + "epoch": 0.41327247191011235, + "grad_norm": 0.9251378178596497, + "learning_rate": 4.9505779754943385e-06, + "loss": 0.6944, + "step": 1177 + }, + { + "epoch": 0.413623595505618, + "grad_norm": 0.9240232706069946, + "learning_rate": 4.950486459976793e-06, + "loss": 0.7241, + "step": 1178 + }, + { + "epoch": 0.4139747191011236, + "grad_norm": 0.9050787687301636, + "learning_rate": 4.950394860654707e-06, + "loss": 0.7044, + "step": 1179 + }, + { + "epoch": 0.4143258426966292, + "grad_norm": 0.9517340064048767, + "learning_rate": 4.950303177531213e-06, + "loss": 0.7363, + "step": 1180 + }, + { + "epoch": 0.41467696629213485, + "grad_norm": 0.9461115598678589, + "learning_rate": 4.950211410609448e-06, + "loss": 0.6786, + "step": 1181 + }, + { + "epoch": 0.41502808988764045, + "grad_norm": 0.9148724675178528, + "learning_rate": 4.950119559892549e-06, + "loss": 0.6873, + "step": 1182 + }, + { + "epoch": 0.41537921348314605, + "grad_norm": 0.9085513353347778, + "learning_rate": 4.950027625383657e-06, + "loss": 0.7209, + "step": 1183 + }, + { + "epoch": 0.4157303370786517, + "grad_norm": 0.9327560067176819, + "learning_rate": 4.949935607085919e-06, + "loss": 0.7086, + "step": 1184 + }, + { + "epoch": 0.4160814606741573, + "grad_norm": 0.936128556728363, + "learning_rate": 4.949843505002478e-06, + "loss": 0.6894, + "step": 1185 + }, + { + "epoch": 0.4164325842696629, + "grad_norm": 0.931978702545166, + "learning_rate": 4.9497513191364855e-06, + "loss": 0.7209, + "step": 1186 + }, + { + "epoch": 0.41678370786516855, + "grad_norm": 0.8966789245605469, + "learning_rate": 4.949659049491095e-06, + "loss": 0.7211, + "step": 1187 + }, + { + "epoch": 0.41713483146067415, + "grad_norm": 0.9003615975379944, + "learning_rate": 4.9495666960694596e-06, + "loss": 0.6835, + "step": 1188 + }, + { + "epoch": 0.4174859550561798, + "grad_norm": 0.933268666267395, + "learning_rate": 4.9494742588747405e-06, + "loss": 0.7335, + "step": 1189 + }, + { + "epoch": 0.4178370786516854, + "grad_norm": 0.8876200318336487, + "learning_rate": 4.949381737910098e-06, + "loss": 0.7132, + "step": 1190 + }, + { + "epoch": 0.418188202247191, + "grad_norm": 0.9531140923500061, + "learning_rate": 4.949289133178695e-06, + "loss": 0.746, + "step": 1191 + }, + { + "epoch": 0.41853932584269665, + "grad_norm": 0.9262682795524597, + "learning_rate": 4.949196444683701e-06, + "loss": 0.7155, + "step": 1192 + }, + { + "epoch": 0.41889044943820225, + "grad_norm": 0.9318521022796631, + "learning_rate": 4.949103672428282e-06, + "loss": 0.7234, + "step": 1193 + }, + { + "epoch": 0.41924157303370785, + "grad_norm": 0.922127366065979, + "learning_rate": 4.949010816415616e-06, + "loss": 0.7116, + "step": 1194 + }, + { + "epoch": 0.4195926966292135, + "grad_norm": 0.9445079565048218, + "learning_rate": 4.948917876648874e-06, + "loss": 0.6993, + "step": 1195 + }, + { + "epoch": 0.4199438202247191, + "grad_norm": 0.9283545613288879, + "learning_rate": 4.948824853131237e-06, + "loss": 0.7078, + "step": 1196 + }, + { + "epoch": 0.4202949438202247, + "grad_norm": 0.8923372626304626, + "learning_rate": 4.9487317458658845e-06, + "loss": 0.6818, + "step": 1197 + }, + { + "epoch": 0.42064606741573035, + "grad_norm": 0.907924234867096, + "learning_rate": 4.948638554856002e-06, + "loss": 0.6799, + "step": 1198 + }, + { + "epoch": 0.42099719101123595, + "grad_norm": 0.8753466606140137, + "learning_rate": 4.948545280104777e-06, + "loss": 0.6789, + "step": 1199 + }, + { + "epoch": 0.42134831460674155, + "grad_norm": 0.965340256690979, + "learning_rate": 4.948451921615398e-06, + "loss": 0.6911, + "step": 1200 + }, + { + "epoch": 0.4216994382022472, + "grad_norm": 0.9519873261451721, + "learning_rate": 4.948358479391059e-06, + "loss": 0.6899, + "step": 1201 + }, + { + "epoch": 0.4220505617977528, + "grad_norm": 0.907811164855957, + "learning_rate": 4.948264953434955e-06, + "loss": 0.7401, + "step": 1202 + }, + { + "epoch": 0.42240168539325845, + "grad_norm": 0.9398930072784424, + "learning_rate": 4.948171343750284e-06, + "loss": 0.7036, + "step": 1203 + }, + { + "epoch": 0.42275280898876405, + "grad_norm": 0.9245709776878357, + "learning_rate": 4.948077650340248e-06, + "loss": 0.7042, + "step": 1204 + }, + { + "epoch": 0.42310393258426965, + "grad_norm": 0.8742227554321289, + "learning_rate": 4.947983873208052e-06, + "loss": 0.6941, + "step": 1205 + }, + { + "epoch": 0.4234550561797753, + "grad_norm": 0.9183447957038879, + "learning_rate": 4.947890012356902e-06, + "loss": 0.7432, + "step": 1206 + }, + { + "epoch": 0.4238061797752809, + "grad_norm": 0.9594230055809021, + "learning_rate": 4.947796067790009e-06, + "loss": 0.7486, + "step": 1207 + }, + { + "epoch": 0.4241573033707865, + "grad_norm": 0.9002628326416016, + "learning_rate": 4.947702039510583e-06, + "loss": 0.6652, + "step": 1208 + }, + { + "epoch": 0.42450842696629215, + "grad_norm": 0.9353854060173035, + "learning_rate": 4.947607927521844e-06, + "loss": 0.7501, + "step": 1209 + }, + { + "epoch": 0.42485955056179775, + "grad_norm": 0.8836169242858887, + "learning_rate": 4.947513731827007e-06, + "loss": 0.6904, + "step": 1210 + }, + { + "epoch": 0.42521067415730335, + "grad_norm": 0.9366084337234497, + "learning_rate": 4.947419452429295e-06, + "loss": 0.7013, + "step": 1211 + }, + { + "epoch": 0.425561797752809, + "grad_norm": 0.9252647161483765, + "learning_rate": 4.947325089331932e-06, + "loss": 0.7343, + "step": 1212 + }, + { + "epoch": 0.4259129213483146, + "grad_norm": 0.9203290343284607, + "learning_rate": 4.947230642538145e-06, + "loss": 0.7129, + "step": 1213 + }, + { + "epoch": 0.4262640449438202, + "grad_norm": 0.8908094763755798, + "learning_rate": 4.947136112051164e-06, + "loss": 0.6785, + "step": 1214 + }, + { + "epoch": 0.42661516853932585, + "grad_norm": 0.9249038696289062, + "learning_rate": 4.947041497874223e-06, + "loss": 0.6664, + "step": 1215 + }, + { + "epoch": 0.42696629213483145, + "grad_norm": 0.9546428918838501, + "learning_rate": 4.946946800010556e-06, + "loss": 0.7276, + "step": 1216 + }, + { + "epoch": 0.4273174157303371, + "grad_norm": 0.9032455682754517, + "learning_rate": 4.946852018463403e-06, + "loss": 0.6757, + "step": 1217 + }, + { + "epoch": 0.4276685393258427, + "grad_norm": 0.9056017994880676, + "learning_rate": 4.946757153236003e-06, + "loss": 0.6878, + "step": 1218 + }, + { + "epoch": 0.4280196629213483, + "grad_norm": 0.9265672564506531, + "learning_rate": 4.946662204331603e-06, + "loss": 0.7043, + "step": 1219 + }, + { + "epoch": 0.42837078651685395, + "grad_norm": 0.8721897006034851, + "learning_rate": 4.94656717175345e-06, + "loss": 0.627, + "step": 1220 + }, + { + "epoch": 0.42872191011235955, + "grad_norm": 0.9220138192176819, + "learning_rate": 4.9464720555047935e-06, + "loss": 0.7233, + "step": 1221 + }, + { + "epoch": 0.42907303370786515, + "grad_norm": 0.9333030581474304, + "learning_rate": 4.9463768555888855e-06, + "loss": 0.7065, + "step": 1222 + }, + { + "epoch": 0.4294241573033708, + "grad_norm": 0.9611704349517822, + "learning_rate": 4.946281572008982e-06, + "loss": 0.7044, + "step": 1223 + }, + { + "epoch": 0.4297752808988764, + "grad_norm": 0.9443504810333252, + "learning_rate": 4.946186204768343e-06, + "loss": 0.7322, + "step": 1224 + }, + { + "epoch": 0.430126404494382, + "grad_norm": 0.9180960059165955, + "learning_rate": 4.946090753870229e-06, + "loss": 0.7435, + "step": 1225 + }, + { + "epoch": 0.43047752808988765, + "grad_norm": 0.9297470450401306, + "learning_rate": 4.945995219317903e-06, + "loss": 0.7211, + "step": 1226 + }, + { + "epoch": 0.43082865168539325, + "grad_norm": 0.9243416786193848, + "learning_rate": 4.945899601114634e-06, + "loss": 0.7255, + "step": 1227 + }, + { + "epoch": 0.4311797752808989, + "grad_norm": 0.9293515086174011, + "learning_rate": 4.945803899263692e-06, + "loss": 0.7466, + "step": 1228 + }, + { + "epoch": 0.4315308988764045, + "grad_norm": 0.8804196715354919, + "learning_rate": 4.94570811376835e-06, + "loss": 0.6495, + "step": 1229 + }, + { + "epoch": 0.4318820224719101, + "grad_norm": 0.8709548115730286, + "learning_rate": 4.945612244631884e-06, + "loss": 0.6573, + "step": 1230 + }, + { + "epoch": 0.43223314606741575, + "grad_norm": 0.9267548322677612, + "learning_rate": 4.94551629185757e-06, + "loss": 0.6949, + "step": 1231 + }, + { + "epoch": 0.43258426966292135, + "grad_norm": 0.9049427509307861, + "learning_rate": 4.945420255448693e-06, + "loss": 0.6724, + "step": 1232 + }, + { + "epoch": 0.43293539325842695, + "grad_norm": 0.9540070295333862, + "learning_rate": 4.945324135408535e-06, + "loss": 0.7162, + "step": 1233 + }, + { + "epoch": 0.4332865168539326, + "grad_norm": 0.9069761633872986, + "learning_rate": 4.945227931740384e-06, + "loss": 0.6851, + "step": 1234 + }, + { + "epoch": 0.4336376404494382, + "grad_norm": 0.9255860447883606, + "learning_rate": 4.94513164444753e-06, + "loss": 0.7108, + "step": 1235 + }, + { + "epoch": 0.4339887640449438, + "grad_norm": 0.9611523151397705, + "learning_rate": 4.945035273533267e-06, + "loss": 0.7156, + "step": 1236 + }, + { + "epoch": 0.43433988764044945, + "grad_norm": 0.9171714782714844, + "learning_rate": 4.9449388190008896e-06, + "loss": 0.6744, + "step": 1237 + }, + { + "epoch": 0.43469101123595505, + "grad_norm": 0.8625194430351257, + "learning_rate": 4.944842280853696e-06, + "loss": 0.6792, + "step": 1238 + }, + { + "epoch": 0.43504213483146065, + "grad_norm": 0.945565938949585, + "learning_rate": 4.94474565909499e-06, + "loss": 0.7015, + "step": 1239 + }, + { + "epoch": 0.4353932584269663, + "grad_norm": 0.8865808248519897, + "learning_rate": 4.944648953728073e-06, + "loss": 0.6664, + "step": 1240 + }, + { + "epoch": 0.4357443820224719, + "grad_norm": 0.8900501728057861, + "learning_rate": 4.9445521647562545e-06, + "loss": 0.6831, + "step": 1241 + }, + { + "epoch": 0.43609550561797755, + "grad_norm": 0.9169394373893738, + "learning_rate": 4.944455292182844e-06, + "loss": 0.6377, + "step": 1242 + }, + { + "epoch": 0.43644662921348315, + "grad_norm": 0.9147849082946777, + "learning_rate": 4.944358336011154e-06, + "loss": 0.735, + "step": 1243 + }, + { + "epoch": 0.43679775280898875, + "grad_norm": 0.9505800604820251, + "learning_rate": 4.9442612962445e-06, + "loss": 0.7309, + "step": 1244 + }, + { + "epoch": 0.4371488764044944, + "grad_norm": 0.8858315348625183, + "learning_rate": 4.9441641728862015e-06, + "loss": 0.6904, + "step": 1245 + }, + { + "epoch": 0.4375, + "grad_norm": 0.9335656762123108, + "learning_rate": 4.94406696593958e-06, + "loss": 0.7219, + "step": 1246 + }, + { + "epoch": 0.4378511235955056, + "grad_norm": 0.8928138613700867, + "learning_rate": 4.94396967540796e-06, + "loss": 0.6521, + "step": 1247 + }, + { + "epoch": 0.43820224719101125, + "grad_norm": 0.9080971479415894, + "learning_rate": 4.943872301294668e-06, + "loss": 0.6925, + "step": 1248 + }, + { + "epoch": 0.43855337078651685, + "grad_norm": 0.9383649826049805, + "learning_rate": 4.943774843603034e-06, + "loss": 0.708, + "step": 1249 + }, + { + "epoch": 0.43890449438202245, + "grad_norm": 0.9101510047912598, + "learning_rate": 4.943677302336393e-06, + "loss": 0.6901, + "step": 1250 + }, + { + "epoch": 0.4392556179775281, + "grad_norm": 0.9402358531951904, + "learning_rate": 4.943579677498078e-06, + "loss": 0.7169, + "step": 1251 + }, + { + "epoch": 0.4396067415730337, + "grad_norm": 0.8998861908912659, + "learning_rate": 4.943481969091429e-06, + "loss": 0.6962, + "step": 1252 + }, + { + "epoch": 0.43995786516853935, + "grad_norm": 0.9605402946472168, + "learning_rate": 4.943384177119788e-06, + "loss": 0.6783, + "step": 1253 + }, + { + "epoch": 0.44030898876404495, + "grad_norm": 0.9242511987686157, + "learning_rate": 4.9432863015865e-06, + "loss": 0.71, + "step": 1254 + }, + { + "epoch": 0.44066011235955055, + "grad_norm": 0.9048755168914795, + "learning_rate": 4.943188342494909e-06, + "loss": 0.7143, + "step": 1255 + }, + { + "epoch": 0.4410112359550562, + "grad_norm": 0.883234977722168, + "learning_rate": 4.9430902998483685e-06, + "loss": 0.6922, + "step": 1256 + }, + { + "epoch": 0.4413623595505618, + "grad_norm": 0.9546087980270386, + "learning_rate": 4.942992173650231e-06, + "loss": 0.725, + "step": 1257 + }, + { + "epoch": 0.4417134831460674, + "grad_norm": 0.9286230206489563, + "learning_rate": 4.942893963903851e-06, + "loss": 0.7263, + "step": 1258 + }, + { + "epoch": 0.44206460674157305, + "grad_norm": 0.901672899723053, + "learning_rate": 4.942795670612588e-06, + "loss": 0.6826, + "step": 1259 + }, + { + "epoch": 0.44241573033707865, + "grad_norm": 0.9127205610275269, + "learning_rate": 4.9426972937798035e-06, + "loss": 0.6727, + "step": 1260 + }, + { + "epoch": 0.44276685393258425, + "grad_norm": 0.9450622797012329, + "learning_rate": 4.942598833408861e-06, + "loss": 0.726, + "step": 1261 + }, + { + "epoch": 0.4431179775280899, + "grad_norm": 0.9145720601081848, + "learning_rate": 4.94250028950313e-06, + "loss": 0.702, + "step": 1262 + }, + { + "epoch": 0.4434691011235955, + "grad_norm": 0.9019625782966614, + "learning_rate": 4.9424016620659786e-06, + "loss": 0.6682, + "step": 1263 + }, + { + "epoch": 0.4438202247191011, + "grad_norm": 0.9257623553276062, + "learning_rate": 4.9423029511007795e-06, + "loss": 0.6981, + "step": 1264 + }, + { + "epoch": 0.44417134831460675, + "grad_norm": 0.9080988764762878, + "learning_rate": 4.94220415661091e-06, + "loss": 0.708, + "step": 1265 + }, + { + "epoch": 0.44452247191011235, + "grad_norm": 0.9560781121253967, + "learning_rate": 4.942105278599748e-06, + "loss": 0.7021, + "step": 1266 + }, + { + "epoch": 0.444873595505618, + "grad_norm": 0.9422893524169922, + "learning_rate": 4.942006317070676e-06, + "loss": 0.6861, + "step": 1267 + }, + { + "epoch": 0.4452247191011236, + "grad_norm": 0.8909929990768433, + "learning_rate": 4.9419072720270764e-06, + "loss": 0.7143, + "step": 1268 + }, + { + "epoch": 0.4455758426966292, + "grad_norm": 0.9080603718757629, + "learning_rate": 4.941808143472339e-06, + "loss": 0.7073, + "step": 1269 + }, + { + "epoch": 0.44592696629213485, + "grad_norm": 0.8691872358322144, + "learning_rate": 4.941708931409851e-06, + "loss": 0.6425, + "step": 1270 + }, + { + "epoch": 0.44627808988764045, + "grad_norm": 0.9223096966743469, + "learning_rate": 4.9416096358430084e-06, + "loss": 0.6876, + "step": 1271 + }, + { + "epoch": 0.44662921348314605, + "grad_norm": 0.9030603766441345, + "learning_rate": 4.9415102567752045e-06, + "loss": 0.6893, + "step": 1272 + }, + { + "epoch": 0.4469803370786517, + "grad_norm": 0.9421838521957397, + "learning_rate": 4.94141079420984e-06, + "loss": 0.7267, + "step": 1273 + }, + { + "epoch": 0.4473314606741573, + "grad_norm": 0.8724151253700256, + "learning_rate": 4.941311248150314e-06, + "loss": 0.6587, + "step": 1274 + }, + { + "epoch": 0.4476825842696629, + "grad_norm": 0.8822656869888306, + "learning_rate": 4.9412116186000346e-06, + "loss": 0.6969, + "step": 1275 + }, + { + "epoch": 0.44803370786516855, + "grad_norm": 0.8846755623817444, + "learning_rate": 4.941111905562405e-06, + "loss": 0.6848, + "step": 1276 + }, + { + "epoch": 0.44838483146067415, + "grad_norm": 0.8999640941619873, + "learning_rate": 4.941012109040838e-06, + "loss": 0.6465, + "step": 1277 + }, + { + "epoch": 0.4487359550561798, + "grad_norm": 0.9194796681404114, + "learning_rate": 4.940912229038746e-06, + "loss": 0.6995, + "step": 1278 + }, + { + "epoch": 0.4490870786516854, + "grad_norm": 0.9296586513519287, + "learning_rate": 4.940812265559543e-06, + "loss": 0.6399, + "step": 1279 + }, + { + "epoch": 0.449438202247191, + "grad_norm": 0.9012690782546997, + "learning_rate": 4.94071221860665e-06, + "loss": 0.7109, + "step": 1280 + }, + { + "epoch": 0.44978932584269665, + "grad_norm": 0.9507624506950378, + "learning_rate": 4.940612088183488e-06, + "loss": 0.7498, + "step": 1281 + }, + { + "epoch": 0.45014044943820225, + "grad_norm": 0.8964364528656006, + "learning_rate": 4.940511874293481e-06, + "loss": 0.674, + "step": 1282 + }, + { + "epoch": 0.45049157303370785, + "grad_norm": 0.9396047592163086, + "learning_rate": 4.940411576940056e-06, + "loss": 0.699, + "step": 1283 + }, + { + "epoch": 0.4508426966292135, + "grad_norm": 0.9038118720054626, + "learning_rate": 4.940311196126643e-06, + "loss": 0.6598, + "step": 1284 + }, + { + "epoch": 0.4511938202247191, + "grad_norm": 0.9001778960227966, + "learning_rate": 4.940210731856675e-06, + "loss": 0.7129, + "step": 1285 + }, + { + "epoch": 0.4515449438202247, + "grad_norm": 0.9313659071922302, + "learning_rate": 4.94011018413359e-06, + "loss": 0.7138, + "step": 1286 + }, + { + "epoch": 0.45189606741573035, + "grad_norm": 0.9225174784660339, + "learning_rate": 4.9400095529608226e-06, + "loss": 0.7192, + "step": 1287 + }, + { + "epoch": 0.45224719101123595, + "grad_norm": 0.9100403189659119, + "learning_rate": 4.939908838341818e-06, + "loss": 0.6593, + "step": 1288 + }, + { + "epoch": 0.45259831460674155, + "grad_norm": 0.8900116086006165, + "learning_rate": 4.939808040280017e-06, + "loss": 0.6799, + "step": 1289 + }, + { + "epoch": 0.4529494382022472, + "grad_norm": 0.885908305644989, + "learning_rate": 4.939707158778872e-06, + "loss": 0.6889, + "step": 1290 + }, + { + "epoch": 0.4533005617977528, + "grad_norm": 0.9219464659690857, + "learning_rate": 4.939606193841827e-06, + "loss": 0.7198, + "step": 1291 + }, + { + "epoch": 0.45365168539325845, + "grad_norm": 0.9270339012145996, + "learning_rate": 4.939505145472339e-06, + "loss": 0.7121, + "step": 1292 + }, + { + "epoch": 0.45400280898876405, + "grad_norm": 0.9297556281089783, + "learning_rate": 4.939404013673862e-06, + "loss": 0.7119, + "step": 1293 + }, + { + "epoch": 0.45435393258426965, + "grad_norm": 0.9266655445098877, + "learning_rate": 4.9393027984498555e-06, + "loss": 0.6994, + "step": 1294 + }, + { + "epoch": 0.4547050561797753, + "grad_norm": 0.9152787923812866, + "learning_rate": 4.939201499803781e-06, + "loss": 0.6618, + "step": 1295 + }, + { + "epoch": 0.4550561797752809, + "grad_norm": 0.9240756034851074, + "learning_rate": 4.939100117739102e-06, + "loss": 0.7359, + "step": 1296 + }, + { + "epoch": 0.4554073033707865, + "grad_norm": 0.9413486123085022, + "learning_rate": 4.9389986522592855e-06, + "loss": 0.7266, + "step": 1297 + }, + { + "epoch": 0.45575842696629215, + "grad_norm": 0.8976860046386719, + "learning_rate": 4.938897103367803e-06, + "loss": 0.7063, + "step": 1298 + }, + { + "epoch": 0.45610955056179775, + "grad_norm": 0.9674167633056641, + "learning_rate": 4.938795471068126e-06, + "loss": 0.7191, + "step": 1299 + }, + { + "epoch": 0.45646067415730335, + "grad_norm": 0.9469913840293884, + "learning_rate": 4.938693755363731e-06, + "loss": 0.7263, + "step": 1300 + }, + { + "epoch": 0.456811797752809, + "grad_norm": 0.9106466174125671, + "learning_rate": 4.938591956258097e-06, + "loss": 0.7324, + "step": 1301 + }, + { + "epoch": 0.4571629213483146, + "grad_norm": 0.9906820058822632, + "learning_rate": 4.938490073754705e-06, + "loss": 0.7273, + "step": 1302 + }, + { + "epoch": 0.4575140449438202, + "grad_norm": 0.9152560234069824, + "learning_rate": 4.938388107857038e-06, + "loss": 0.6648, + "step": 1303 + }, + { + "epoch": 0.45786516853932585, + "grad_norm": 0.9459682703018188, + "learning_rate": 4.938286058568585e-06, + "loss": 0.6869, + "step": 1304 + }, + { + "epoch": 0.45821629213483145, + "grad_norm": 0.9387467503547668, + "learning_rate": 4.938183925892835e-06, + "loss": 0.7028, + "step": 1305 + }, + { + "epoch": 0.4585674157303371, + "grad_norm": 0.9118907451629639, + "learning_rate": 4.938081709833281e-06, + "loss": 0.6875, + "step": 1306 + }, + { + "epoch": 0.4589185393258427, + "grad_norm": 0.9032610654830933, + "learning_rate": 4.937979410393419e-06, + "loss": 0.7157, + "step": 1307 + }, + { + "epoch": 0.4592696629213483, + "grad_norm": 0.9495104551315308, + "learning_rate": 4.937877027576748e-06, + "loss": 0.6914, + "step": 1308 + }, + { + "epoch": 0.45962078651685395, + "grad_norm": 0.8904275894165039, + "learning_rate": 4.937774561386768e-06, + "loss": 0.7351, + "step": 1309 + }, + { + "epoch": 0.45997191011235955, + "grad_norm": 0.8993387222290039, + "learning_rate": 4.937672011826984e-06, + "loss": 0.7004, + "step": 1310 + }, + { + "epoch": 0.46032303370786515, + "grad_norm": 0.9243264198303223, + "learning_rate": 4.937569378900903e-06, + "loss": 0.7059, + "step": 1311 + }, + { + "epoch": 0.4606741573033708, + "grad_norm": 0.8777183294296265, + "learning_rate": 4.937466662612036e-06, + "loss": 0.6795, + "step": 1312 + }, + { + "epoch": 0.4610252808988764, + "grad_norm": 0.9348961114883423, + "learning_rate": 4.937363862963895e-06, + "loss": 0.6726, + "step": 1313 + }, + { + "epoch": 0.461376404494382, + "grad_norm": 0.9088345766067505, + "learning_rate": 4.937260979959994e-06, + "loss": 0.7114, + "step": 1314 + }, + { + "epoch": 0.46172752808988765, + "grad_norm": 0.9143536686897278, + "learning_rate": 4.937158013603854e-06, + "loss": 0.6819, + "step": 1315 + }, + { + "epoch": 0.46207865168539325, + "grad_norm": 0.9480457901954651, + "learning_rate": 4.937054963898996e-06, + "loss": 0.7335, + "step": 1316 + }, + { + "epoch": 0.4624297752808989, + "grad_norm": 0.9207438230514526, + "learning_rate": 4.9369518308489425e-06, + "loss": 0.7091, + "step": 1317 + }, + { + "epoch": 0.4627808988764045, + "grad_norm": 0.9547027349472046, + "learning_rate": 4.936848614457223e-06, + "loss": 0.7388, + "step": 1318 + }, + { + "epoch": 0.4631320224719101, + "grad_norm": 0.9442094564437866, + "learning_rate": 4.9367453147273645e-06, + "loss": 0.709, + "step": 1319 + }, + { + "epoch": 0.46348314606741575, + "grad_norm": 0.9056798815727234, + "learning_rate": 4.9366419316629024e-06, + "loss": 0.6919, + "step": 1320 + }, + { + "epoch": 0.46383426966292135, + "grad_norm": 0.8800448179244995, + "learning_rate": 4.936538465267371e-06, + "loss": 0.7141, + "step": 1321 + }, + { + "epoch": 0.46418539325842695, + "grad_norm": 0.9331648349761963, + "learning_rate": 4.93643491554431e-06, + "loss": 0.6871, + "step": 1322 + }, + { + "epoch": 0.4645365168539326, + "grad_norm": 0.9699183702468872, + "learning_rate": 4.9363312824972586e-06, + "loss": 0.729, + "step": 1323 + }, + { + "epoch": 0.4648876404494382, + "grad_norm": 0.9243112206459045, + "learning_rate": 4.9362275661297634e-06, + "loss": 0.6751, + "step": 1324 + }, + { + "epoch": 0.4652387640449438, + "grad_norm": 0.9310767650604248, + "learning_rate": 4.936123766445368e-06, + "loss": 0.7263, + "step": 1325 + }, + { + "epoch": 0.46558988764044945, + "grad_norm": 0.9163933992385864, + "learning_rate": 4.936019883447627e-06, + "loss": 0.715, + "step": 1326 + }, + { + "epoch": 0.46594101123595505, + "grad_norm": 0.9764875173568726, + "learning_rate": 4.935915917140089e-06, + "loss": 0.7139, + "step": 1327 + }, + { + "epoch": 0.46629213483146065, + "grad_norm": 0.9456989169120789, + "learning_rate": 4.935811867526311e-06, + "loss": 0.6667, + "step": 1328 + }, + { + "epoch": 0.4666432584269663, + "grad_norm": 0.9617751836776733, + "learning_rate": 4.935707734609852e-06, + "loss": 0.691, + "step": 1329 + }, + { + "epoch": 0.4669943820224719, + "grad_norm": 0.8945748209953308, + "learning_rate": 4.935603518394273e-06, + "loss": 0.711, + "step": 1330 + }, + { + "epoch": 0.46734550561797755, + "grad_norm": 0.9066562652587891, + "learning_rate": 4.935499218883137e-06, + "loss": 0.685, + "step": 1331 + }, + { + "epoch": 0.46769662921348315, + "grad_norm": 0.8701803684234619, + "learning_rate": 4.935394836080012e-06, + "loss": 0.6735, + "step": 1332 + }, + { + "epoch": 0.46804775280898875, + "grad_norm": 0.9355669021606445, + "learning_rate": 4.935290369988468e-06, + "loss": 0.7335, + "step": 1333 + }, + { + "epoch": 0.4683988764044944, + "grad_norm": 0.882905900478363, + "learning_rate": 4.9351858206120774e-06, + "loss": 0.7196, + "step": 1334 + }, + { + "epoch": 0.46875, + "grad_norm": 0.9127156734466553, + "learning_rate": 4.935081187954415e-06, + "loss": 0.714, + "step": 1335 + }, + { + "epoch": 0.4691011235955056, + "grad_norm": 1.284045934677124, + "learning_rate": 4.9349764720190595e-06, + "loss": 0.7167, + "step": 1336 + }, + { + "epoch": 0.46945224719101125, + "grad_norm": 0.919476330280304, + "learning_rate": 4.934871672809594e-06, + "loss": 0.6804, + "step": 1337 + }, + { + "epoch": 0.46980337078651685, + "grad_norm": 0.9066704511642456, + "learning_rate": 4.934766790329599e-06, + "loss": 0.706, + "step": 1338 + }, + { + "epoch": 0.47015449438202245, + "grad_norm": 0.9020872116088867, + "learning_rate": 4.934661824582664e-06, + "loss": 0.7277, + "step": 1339 + }, + { + "epoch": 0.4705056179775281, + "grad_norm": 0.9096556901931763, + "learning_rate": 4.934556775572377e-06, + "loss": 0.7184, + "step": 1340 + }, + { + "epoch": 0.4708567415730337, + "grad_norm": 0.8852599859237671, + "learning_rate": 4.934451643302332e-06, + "loss": 0.6952, + "step": 1341 + }, + { + "epoch": 0.47120786516853935, + "grad_norm": 0.9317033290863037, + "learning_rate": 4.934346427776125e-06, + "loss": 0.7087, + "step": 1342 + }, + { + "epoch": 0.47155898876404495, + "grad_norm": 0.9349064230918884, + "learning_rate": 4.934241128997353e-06, + "loss": 0.7295, + "step": 1343 + }, + { + "epoch": 0.47191011235955055, + "grad_norm": 0.8978156447410583, + "learning_rate": 4.934135746969617e-06, + "loss": 0.7129, + "step": 1344 + }, + { + "epoch": 0.4722612359550562, + "grad_norm": 0.8900869488716125, + "learning_rate": 4.934030281696521e-06, + "loss": 0.6737, + "step": 1345 + }, + { + "epoch": 0.4726123595505618, + "grad_norm": 0.9193541407585144, + "learning_rate": 4.933924733181673e-06, + "loss": 0.7189, + "step": 1346 + }, + { + "epoch": 0.4729634831460674, + "grad_norm": 0.8839302659034729, + "learning_rate": 4.933819101428681e-06, + "loss": 0.6883, + "step": 1347 + }, + { + "epoch": 0.47331460674157305, + "grad_norm": 0.8768748641014099, + "learning_rate": 4.933713386441159e-06, + "loss": 0.6825, + "step": 1348 + }, + { + "epoch": 0.47366573033707865, + "grad_norm": 0.9012611508369446, + "learning_rate": 4.933607588222721e-06, + "loss": 0.7477, + "step": 1349 + }, + { + "epoch": 0.47401685393258425, + "grad_norm": 0.9348760843276978, + "learning_rate": 4.933501706776986e-06, + "loss": 0.7071, + "step": 1350 + }, + { + "epoch": 0.4743679775280899, + "grad_norm": 0.9126866459846497, + "learning_rate": 4.933395742107576e-06, + "loss": 0.7259, + "step": 1351 + }, + { + "epoch": 0.4747191011235955, + "grad_norm": 1.1260316371917725, + "learning_rate": 4.9332896942181135e-06, + "loss": 0.6647, + "step": 1352 + }, + { + "epoch": 0.4750702247191011, + "grad_norm": 0.9544892907142639, + "learning_rate": 4.933183563112226e-06, + "loss": 0.7228, + "step": 1353 + }, + { + "epoch": 0.47542134831460675, + "grad_norm": 0.9179894924163818, + "learning_rate": 4.933077348793542e-06, + "loss": 0.7115, + "step": 1354 + }, + { + "epoch": 0.47577247191011235, + "grad_norm": 0.8785374164581299, + "learning_rate": 4.932971051265695e-06, + "loss": 0.6651, + "step": 1355 + }, + { + "epoch": 0.476123595505618, + "grad_norm": 0.9006591439247131, + "learning_rate": 4.93286467053232e-06, + "loss": 0.7166, + "step": 1356 + }, + { + "epoch": 0.4764747191011236, + "grad_norm": 0.9105557799339294, + "learning_rate": 4.932758206597055e-06, + "loss": 0.6876, + "step": 1357 + }, + { + "epoch": 0.4768258426966292, + "grad_norm": 0.8968350887298584, + "learning_rate": 4.932651659463541e-06, + "loss": 0.6873, + "step": 1358 + }, + { + "epoch": 0.47717696629213485, + "grad_norm": 0.9052225947380066, + "learning_rate": 4.932545029135422e-06, + "loss": 0.7134, + "step": 1359 + }, + { + "epoch": 0.47752808988764045, + "grad_norm": 0.946305513381958, + "learning_rate": 4.932438315616345e-06, + "loss": 0.7061, + "step": 1360 + }, + { + "epoch": 0.47787921348314605, + "grad_norm": 0.9128836393356323, + "learning_rate": 4.9323315189099586e-06, + "loss": 0.7101, + "step": 1361 + }, + { + "epoch": 0.4782303370786517, + "grad_norm": 0.9199392199516296, + "learning_rate": 4.9322246390199155e-06, + "loss": 0.7073, + "step": 1362 + }, + { + "epoch": 0.4785814606741573, + "grad_norm": 1.077110767364502, + "learning_rate": 4.932117675949871e-06, + "loss": 0.6828, + "step": 1363 + }, + { + "epoch": 0.4789325842696629, + "grad_norm": 0.9046890735626221, + "learning_rate": 4.932010629703484e-06, + "loss": 0.6991, + "step": 1364 + }, + { + "epoch": 0.47928370786516855, + "grad_norm": 0.8711754083633423, + "learning_rate": 4.931903500284414e-06, + "loss": 0.6788, + "step": 1365 + }, + { + "epoch": 0.47963483146067415, + "grad_norm": 0.8938144445419312, + "learning_rate": 4.931796287696325e-06, + "loss": 0.7219, + "step": 1366 + }, + { + "epoch": 0.4799859550561798, + "grad_norm": 0.9003678560256958, + "learning_rate": 4.931688991942884e-06, + "loss": 0.6908, + "step": 1367 + }, + { + "epoch": 0.4803370786516854, + "grad_norm": 0.9107855558395386, + "learning_rate": 4.93158161302776e-06, + "loss": 0.6821, + "step": 1368 + }, + { + "epoch": 0.480688202247191, + "grad_norm": 0.9014522433280945, + "learning_rate": 4.931474150954626e-06, + "loss": 0.7071, + "step": 1369 + }, + { + "epoch": 0.48103932584269665, + "grad_norm": 0.9273253083229065, + "learning_rate": 4.931366605727157e-06, + "loss": 0.7027, + "step": 1370 + }, + { + "epoch": 0.48139044943820225, + "grad_norm": 0.9555942416191101, + "learning_rate": 4.93125897734903e-06, + "loss": 0.6952, + "step": 1371 + }, + { + "epoch": 0.48174157303370785, + "grad_norm": 0.9457711577415466, + "learning_rate": 4.931151265823927e-06, + "loss": 0.6893, + "step": 1372 + }, + { + "epoch": 0.4820926966292135, + "grad_norm": 0.9268205761909485, + "learning_rate": 4.931043471155531e-06, + "loss": 0.7169, + "step": 1373 + }, + { + "epoch": 0.4824438202247191, + "grad_norm": 0.9136787056922913, + "learning_rate": 4.930935593347529e-06, + "loss": 0.7326, + "step": 1374 + }, + { + "epoch": 0.4827949438202247, + "grad_norm": 0.925116777420044, + "learning_rate": 4.930827632403608e-06, + "loss": 0.7037, + "step": 1375 + }, + { + "epoch": 0.48314606741573035, + "grad_norm": 0.9161933064460754, + "learning_rate": 4.930719588327464e-06, + "loss": 0.7222, + "step": 1376 + }, + { + "epoch": 0.48349719101123595, + "grad_norm": 0.9238081574440002, + "learning_rate": 4.93061146112279e-06, + "loss": 0.7186, + "step": 1377 + }, + { + "epoch": 0.48384831460674155, + "grad_norm": 0.9430643320083618, + "learning_rate": 4.930503250793283e-06, + "loss": 0.7622, + "step": 1378 + }, + { + "epoch": 0.4841994382022472, + "grad_norm": 0.9295469522476196, + "learning_rate": 4.930394957342645e-06, + "loss": 0.7354, + "step": 1379 + }, + { + "epoch": 0.4845505617977528, + "grad_norm": 0.9064199924468994, + "learning_rate": 4.9302865807745795e-06, + "loss": 0.6487, + "step": 1380 + }, + { + "epoch": 0.48490168539325845, + "grad_norm": 0.9455371499061584, + "learning_rate": 4.930178121092792e-06, + "loss": 0.7468, + "step": 1381 + }, + { + "epoch": 0.48525280898876405, + "grad_norm": 0.9263501763343811, + "learning_rate": 4.930069578300992e-06, + "loss": 0.7099, + "step": 1382 + }, + { + "epoch": 0.48560393258426965, + "grad_norm": 0.8904842734336853, + "learning_rate": 4.929960952402892e-06, + "loss": 0.6519, + "step": 1383 + }, + { + "epoch": 0.4859550561797753, + "grad_norm": 0.8946258425712585, + "learning_rate": 4.929852243402208e-06, + "loss": 0.7011, + "step": 1384 + }, + { + "epoch": 0.4863061797752809, + "grad_norm": 0.9584969282150269, + "learning_rate": 4.929743451302656e-06, + "loss": 0.724, + "step": 1385 + }, + { + "epoch": 0.4866573033707865, + "grad_norm": 0.9212526082992554, + "learning_rate": 4.929634576107956e-06, + "loss": 0.6821, + "step": 1386 + }, + { + "epoch": 0.48700842696629215, + "grad_norm": 0.9416231513023376, + "learning_rate": 4.929525617821834e-06, + "loss": 0.7457, + "step": 1387 + }, + { + "epoch": 0.48735955056179775, + "grad_norm": 0.9377028346061707, + "learning_rate": 4.929416576448014e-06, + "loss": 0.6871, + "step": 1388 + }, + { + "epoch": 0.48771067415730335, + "grad_norm": 0.9065041542053223, + "learning_rate": 4.929307451990226e-06, + "loss": 0.7261, + "step": 1389 + }, + { + "epoch": 0.488061797752809, + "grad_norm": 0.844350278377533, + "learning_rate": 4.929198244452202e-06, + "loss": 0.7099, + "step": 1390 + }, + { + "epoch": 0.4884129213483146, + "grad_norm": 0.9054086804389954, + "learning_rate": 4.9290889538376775e-06, + "loss": 0.669, + "step": 1391 + }, + { + "epoch": 0.4887640449438202, + "grad_norm": 0.8729206919670105, + "learning_rate": 4.928979580150388e-06, + "loss": 0.6816, + "step": 1392 + }, + { + "epoch": 0.48911516853932585, + "grad_norm": 0.9076043367385864, + "learning_rate": 4.928870123394077e-06, + "loss": 0.7044, + "step": 1393 + }, + { + "epoch": 0.48946629213483145, + "grad_norm": 0.9312844276428223, + "learning_rate": 4.928760583572485e-06, + "loss": 0.7107, + "step": 1394 + }, + { + "epoch": 0.4898174157303371, + "grad_norm": 0.9900015592575073, + "learning_rate": 4.928650960689359e-06, + "loss": 0.703, + "step": 1395 + }, + { + "epoch": 0.4901685393258427, + "grad_norm": 0.9289678931236267, + "learning_rate": 4.92854125474845e-06, + "loss": 0.714, + "step": 1396 + }, + { + "epoch": 0.4905196629213483, + "grad_norm": 0.9454885125160217, + "learning_rate": 4.928431465753507e-06, + "loss": 0.7312, + "step": 1397 + }, + { + "epoch": 0.49087078651685395, + "grad_norm": 0.9117372035980225, + "learning_rate": 4.928321593708286e-06, + "loss": 0.6984, + "step": 1398 + }, + { + "epoch": 0.49122191011235955, + "grad_norm": 0.9010490775108337, + "learning_rate": 4.928211638616545e-06, + "loss": 0.672, + "step": 1399 + }, + { + "epoch": 0.49157303370786515, + "grad_norm": 0.9090813994407654, + "learning_rate": 4.928101600482043e-06, + "loss": 0.7132, + "step": 1400 + }, + { + "epoch": 0.4919241573033708, + "grad_norm": 0.905497133731842, + "learning_rate": 4.927991479308545e-06, + "loss": 0.677, + "step": 1401 + }, + { + "epoch": 0.4922752808988764, + "grad_norm": 0.902889609336853, + "learning_rate": 4.927881275099815e-06, + "loss": 0.6601, + "step": 1402 + }, + { + "epoch": 0.492626404494382, + "grad_norm": 0.9104518294334412, + "learning_rate": 4.927770987859624e-06, + "loss": 0.7192, + "step": 1403 + }, + { + "epoch": 0.49297752808988765, + "grad_norm": 0.8784206509590149, + "learning_rate": 4.9276606175917415e-06, + "loss": 0.6556, + "step": 1404 + }, + { + "epoch": 0.49332865168539325, + "grad_norm": 0.9406437277793884, + "learning_rate": 4.927550164299944e-06, + "loss": 0.679, + "step": 1405 + }, + { + "epoch": 0.4936797752808989, + "grad_norm": 0.8956720232963562, + "learning_rate": 4.9274396279880075e-06, + "loss": 0.662, + "step": 1406 + }, + { + "epoch": 0.4940308988764045, + "grad_norm": 0.9246423840522766, + "learning_rate": 4.927329008659714e-06, + "loss": 0.7155, + "step": 1407 + }, + { + "epoch": 0.4943820224719101, + "grad_norm": 0.9392451047897339, + "learning_rate": 4.927218306318845e-06, + "loss": 0.6819, + "step": 1408 + }, + { + "epoch": 0.49473314606741575, + "grad_norm": 0.9309847950935364, + "learning_rate": 4.927107520969187e-06, + "loss": 0.7195, + "step": 1409 + }, + { + "epoch": 0.49508426966292135, + "grad_norm": 0.890052080154419, + "learning_rate": 4.926996652614529e-06, + "loss": 0.6703, + "step": 1410 + }, + { + "epoch": 0.49543539325842695, + "grad_norm": 0.9011236429214478, + "learning_rate": 4.926885701258661e-06, + "loss": 0.6822, + "step": 1411 + }, + { + "epoch": 0.4957865168539326, + "grad_norm": 0.915877103805542, + "learning_rate": 4.92677466690538e-06, + "loss": 0.7156, + "step": 1412 + }, + { + "epoch": 0.4961376404494382, + "grad_norm": 0.8800505995750427, + "learning_rate": 4.9266635495584814e-06, + "loss": 0.6817, + "step": 1413 + }, + { + "epoch": 0.4964887640449438, + "grad_norm": 0.9167606234550476, + "learning_rate": 4.926552349221766e-06, + "loss": 0.6982, + "step": 1414 + }, + { + "epoch": 0.49683988764044945, + "grad_norm": 0.9250441193580627, + "learning_rate": 4.926441065899037e-06, + "loss": 0.6559, + "step": 1415 + }, + { + "epoch": 0.49719101123595505, + "grad_norm": 0.910899817943573, + "learning_rate": 4.9263296995941004e-06, + "loss": 0.6654, + "step": 1416 + }, + { + "epoch": 0.49754213483146065, + "grad_norm": 0.8955602049827576, + "learning_rate": 4.926218250310764e-06, + "loss": 0.7022, + "step": 1417 + }, + { + "epoch": 0.4978932584269663, + "grad_norm": 0.9195224642753601, + "learning_rate": 4.926106718052839e-06, + "loss": 0.7058, + "step": 1418 + }, + { + "epoch": 0.4982443820224719, + "grad_norm": 0.9715495705604553, + "learning_rate": 4.92599510282414e-06, + "loss": 0.6917, + "step": 1419 + }, + { + "epoch": 0.49859550561797755, + "grad_norm": 0.8591254353523254, + "learning_rate": 4.925883404628485e-06, + "loss": 0.6415, + "step": 1420 + }, + { + "epoch": 0.49894662921348315, + "grad_norm": 0.903109610080719, + "learning_rate": 4.925771623469694e-06, + "loss": 0.702, + "step": 1421 + }, + { + "epoch": 0.49929775280898875, + "grad_norm": 0.9081088900566101, + "learning_rate": 4.9256597593515885e-06, + "loss": 0.7098, + "step": 1422 + }, + { + "epoch": 0.4996488764044944, + "grad_norm": 0.9099250435829163, + "learning_rate": 4.925547812277994e-06, + "loss": 0.6752, + "step": 1423 + }, + { + "epoch": 0.5, + "grad_norm": 0.9163993000984192, + "learning_rate": 4.925435782252742e-06, + "loss": 0.7098, + "step": 1424 + }, + { + "epoch": 0.5003511235955056, + "grad_norm": 0.9282369017601013, + "learning_rate": 4.92532366927966e-06, + "loss": 0.7003, + "step": 1425 + }, + { + "epoch": 0.5007022471910112, + "grad_norm": 0.9375564455986023, + "learning_rate": 4.925211473362585e-06, + "loss": 0.6981, + "step": 1426 + }, + { + "epoch": 0.5010533707865169, + "grad_norm": 0.9100632071495056, + "learning_rate": 4.925099194505352e-06, + "loss": 0.6794, + "step": 1427 + }, + { + "epoch": 0.5014044943820225, + "grad_norm": 0.8592130541801453, + "learning_rate": 4.924986832711802e-06, + "loss": 0.6547, + "step": 1428 + }, + { + "epoch": 0.5017556179775281, + "grad_norm": 0.9379070401191711, + "learning_rate": 4.924874387985777e-06, + "loss": 0.6829, + "step": 1429 + }, + { + "epoch": 0.5021067415730337, + "grad_norm": 0.9097589254379272, + "learning_rate": 4.924761860331125e-06, + "loss": 0.682, + "step": 1430 + }, + { + "epoch": 0.5024578651685393, + "grad_norm": 0.925594687461853, + "learning_rate": 4.92464924975169e-06, + "loss": 0.7312, + "step": 1431 + }, + { + "epoch": 0.5028089887640449, + "grad_norm": 0.8913887739181519, + "learning_rate": 4.924536556251326e-06, + "loss": 0.6613, + "step": 1432 + }, + { + "epoch": 0.5031601123595506, + "grad_norm": 0.8830445408821106, + "learning_rate": 4.924423779833887e-06, + "loss": 0.689, + "step": 1433 + }, + { + "epoch": 0.5035112359550562, + "grad_norm": 0.9521420001983643, + "learning_rate": 4.924310920503229e-06, + "loss": 0.678, + "step": 1434 + }, + { + "epoch": 0.5038623595505618, + "grad_norm": 0.918493390083313, + "learning_rate": 4.924197978263212e-06, + "loss": 0.7188, + "step": 1435 + }, + { + "epoch": 0.5042134831460674, + "grad_norm": 0.9049700498580933, + "learning_rate": 4.9240849531176986e-06, + "loss": 0.6846, + "step": 1436 + }, + { + "epoch": 0.504564606741573, + "grad_norm": 0.8993126153945923, + "learning_rate": 4.9239718450705545e-06, + "loss": 0.7298, + "step": 1437 + }, + { + "epoch": 0.5049157303370787, + "grad_norm": 0.9134989380836487, + "learning_rate": 4.923858654125647e-06, + "loss": 0.6945, + "step": 1438 + }, + { + "epoch": 0.5052668539325843, + "grad_norm": 0.9160113334655762, + "learning_rate": 4.9237453802868475e-06, + "loss": 0.7074, + "step": 1439 + }, + { + "epoch": 0.5056179775280899, + "grad_norm": 0.9044978022575378, + "learning_rate": 4.923632023558031e-06, + "loss": 0.6814, + "step": 1440 + }, + { + "epoch": 0.5059691011235955, + "grad_norm": 0.9281947612762451, + "learning_rate": 4.923518583943072e-06, + "loss": 0.6806, + "step": 1441 + }, + { + "epoch": 0.5063202247191011, + "grad_norm": 0.8679924607276917, + "learning_rate": 4.923405061445852e-06, + "loss": 0.6654, + "step": 1442 + }, + { + "epoch": 0.5066713483146067, + "grad_norm": 0.905013918876648, + "learning_rate": 4.923291456070252e-06, + "loss": 0.6938, + "step": 1443 + }, + { + "epoch": 0.5070224719101124, + "grad_norm": 0.8952536582946777, + "learning_rate": 4.923177767820158e-06, + "loss": 0.6802, + "step": 1444 + }, + { + "epoch": 0.507373595505618, + "grad_norm": 0.9175031185150146, + "learning_rate": 4.923063996699458e-06, + "loss": 0.6984, + "step": 1445 + }, + { + "epoch": 0.5077247191011236, + "grad_norm": 0.8895412087440491, + "learning_rate": 4.922950142712044e-06, + "loss": 0.7255, + "step": 1446 + }, + { + "epoch": 0.5080758426966292, + "grad_norm": 0.874991238117218, + "learning_rate": 4.922836205861807e-06, + "loss": 0.6709, + "step": 1447 + }, + { + "epoch": 0.5084269662921348, + "grad_norm": 0.8750325441360474, + "learning_rate": 4.922722186152646e-06, + "loss": 0.6414, + "step": 1448 + }, + { + "epoch": 0.5087780898876404, + "grad_norm": 0.8989003300666809, + "learning_rate": 4.922608083588459e-06, + "loss": 0.6928, + "step": 1449 + }, + { + "epoch": 0.5091292134831461, + "grad_norm": 0.9451534152030945, + "learning_rate": 4.922493898173148e-06, + "loss": 0.7143, + "step": 1450 + }, + { + "epoch": 0.5094803370786517, + "grad_norm": 0.9093509912490845, + "learning_rate": 4.922379629910619e-06, + "loss": 0.7027, + "step": 1451 + }, + { + "epoch": 0.5098314606741573, + "grad_norm": 0.932784914970398, + "learning_rate": 4.9222652788047795e-06, + "loss": 0.7024, + "step": 1452 + }, + { + "epoch": 0.5101825842696629, + "grad_norm": 0.9492952227592468, + "learning_rate": 4.92215084485954e-06, + "loss": 0.6856, + "step": 1453 + }, + { + "epoch": 0.5105337078651685, + "grad_norm": 0.8805875182151794, + "learning_rate": 4.922036328078815e-06, + "loss": 0.7114, + "step": 1454 + }, + { + "epoch": 0.5108848314606742, + "grad_norm": 0.8962169289588928, + "learning_rate": 4.92192172846652e-06, + "loss": 0.6962, + "step": 1455 + }, + { + "epoch": 0.5112359550561798, + "grad_norm": 0.923744797706604, + "learning_rate": 4.921807046026574e-06, + "loss": 0.7325, + "step": 1456 + }, + { + "epoch": 0.5115870786516854, + "grad_norm": 0.9285497069358826, + "learning_rate": 4.9216922807628995e-06, + "loss": 0.7208, + "step": 1457 + }, + { + "epoch": 0.511938202247191, + "grad_norm": 0.879386842250824, + "learning_rate": 4.921577432679421e-06, + "loss": 0.6404, + "step": 1458 + }, + { + "epoch": 0.5122893258426966, + "grad_norm": 0.9378132820129395, + "learning_rate": 4.921462501780067e-06, + "loss": 0.7286, + "step": 1459 + }, + { + "epoch": 0.5126404494382022, + "grad_norm": 0.9210785031318665, + "learning_rate": 4.921347488068767e-06, + "loss": 0.7344, + "step": 1460 + }, + { + "epoch": 0.5129915730337079, + "grad_norm": 0.9228889346122742, + "learning_rate": 4.921232391549456e-06, + "loss": 0.7108, + "step": 1461 + }, + { + "epoch": 0.5133426966292135, + "grad_norm": 0.8554604649543762, + "learning_rate": 4.9211172122260675e-06, + "loss": 0.6272, + "step": 1462 + }, + { + "epoch": 0.5136938202247191, + "grad_norm": 0.8758804202079773, + "learning_rate": 4.921001950102543e-06, + "loss": 0.6463, + "step": 1463 + }, + { + "epoch": 0.5140449438202247, + "grad_norm": 0.9541190266609192, + "learning_rate": 4.920886605182823e-06, + "loss": 0.7274, + "step": 1464 + }, + { + "epoch": 0.5143960674157303, + "grad_norm": 0.9359861016273499, + "learning_rate": 4.920771177470853e-06, + "loss": 0.6961, + "step": 1465 + }, + { + "epoch": 0.514747191011236, + "grad_norm": 0.8696016073226929, + "learning_rate": 4.92065566697058e-06, + "loss": 0.7263, + "step": 1466 + }, + { + "epoch": 0.5150983146067416, + "grad_norm": 0.8890820741653442, + "learning_rate": 4.920540073685956e-06, + "loss": 0.7063, + "step": 1467 + }, + { + "epoch": 0.5154494382022472, + "grad_norm": 0.90772545337677, + "learning_rate": 4.920424397620931e-06, + "loss": 0.6858, + "step": 1468 + }, + { + "epoch": 0.5158005617977528, + "grad_norm": 0.9410818219184875, + "learning_rate": 4.920308638779463e-06, + "loss": 0.6655, + "step": 1469 + }, + { + "epoch": 0.5161516853932584, + "grad_norm": 0.8993090987205505, + "learning_rate": 4.920192797165511e-06, + "loss": 0.7215, + "step": 1470 + }, + { + "epoch": 0.516502808988764, + "grad_norm": 0.8889127969741821, + "learning_rate": 4.920076872783036e-06, + "loss": 0.7214, + "step": 1471 + }, + { + "epoch": 0.5168539325842697, + "grad_norm": 0.9573481678962708, + "learning_rate": 4.919960865636004e-06, + "loss": 0.6754, + "step": 1472 + }, + { + "epoch": 0.5172050561797753, + "grad_norm": 0.9702395796775818, + "learning_rate": 4.91984477572838e-06, + "loss": 0.6986, + "step": 1473 + }, + { + "epoch": 0.5175561797752809, + "grad_norm": 0.9364273548126221, + "learning_rate": 4.919728603064136e-06, + "loss": 0.7098, + "step": 1474 + }, + { + "epoch": 0.5179073033707865, + "grad_norm": 1.0006465911865234, + "learning_rate": 4.919612347647244e-06, + "loss": 0.7143, + "step": 1475 + }, + { + "epoch": 0.5182584269662921, + "grad_norm": 0.9335517287254333, + "learning_rate": 4.91949600948168e-06, + "loss": 0.6837, + "step": 1476 + }, + { + "epoch": 0.5186095505617978, + "grad_norm": 0.9131624698638916, + "learning_rate": 4.919379588571424e-06, + "loss": 0.7254, + "step": 1477 + }, + { + "epoch": 0.5189606741573034, + "grad_norm": 0.9354121685028076, + "learning_rate": 4.919263084920455e-06, + "loss": 0.707, + "step": 1478 + }, + { + "epoch": 0.519311797752809, + "grad_norm": 0.8909806609153748, + "learning_rate": 4.919146498532758e-06, + "loss": 0.6655, + "step": 1479 + }, + { + "epoch": 0.5196629213483146, + "grad_norm": 0.9220510721206665, + "learning_rate": 4.919029829412323e-06, + "loss": 0.6859, + "step": 1480 + }, + { + "epoch": 0.5200140449438202, + "grad_norm": 0.9528585076332092, + "learning_rate": 4.918913077563137e-06, + "loss": 0.694, + "step": 1481 + }, + { + "epoch": 0.5203651685393258, + "grad_norm": 0.9287347793579102, + "learning_rate": 4.918796242989193e-06, + "loss": 0.6775, + "step": 1482 + }, + { + "epoch": 0.5207162921348315, + "grad_norm": 0.9118848443031311, + "learning_rate": 4.9186793256944885e-06, + "loss": 0.6869, + "step": 1483 + }, + { + "epoch": 0.5210674157303371, + "grad_norm": 0.8987846374511719, + "learning_rate": 4.91856232568302e-06, + "loss": 0.6883, + "step": 1484 + }, + { + "epoch": 0.5214185393258427, + "grad_norm": 0.8864469528198242, + "learning_rate": 4.918445242958789e-06, + "loss": 0.7144, + "step": 1485 + }, + { + "epoch": 0.5217696629213483, + "grad_norm": 0.8598546385765076, + "learning_rate": 4.918328077525802e-06, + "loss": 0.6752, + "step": 1486 + }, + { + "epoch": 0.5221207865168539, + "grad_norm": 0.8941078782081604, + "learning_rate": 4.918210829388061e-06, + "loss": 0.6733, + "step": 1487 + }, + { + "epoch": 0.5224719101123596, + "grad_norm": 0.914069414138794, + "learning_rate": 4.918093498549582e-06, + "loss": 0.7032, + "step": 1488 + }, + { + "epoch": 0.5228230337078652, + "grad_norm": 0.9472395181655884, + "learning_rate": 4.917976085014373e-06, + "loss": 0.6666, + "step": 1489 + }, + { + "epoch": 0.5231741573033708, + "grad_norm": 0.9576021432876587, + "learning_rate": 4.9178585887864515e-06, + "loss": 0.7056, + "step": 1490 + }, + { + "epoch": 0.5235252808988764, + "grad_norm": 0.9019284248352051, + "learning_rate": 4.9177410098698354e-06, + "loss": 0.6841, + "step": 1491 + }, + { + "epoch": 0.523876404494382, + "grad_norm": 0.8970656394958496, + "learning_rate": 4.917623348268546e-06, + "loss": 0.7169, + "step": 1492 + }, + { + "epoch": 0.5242275280898876, + "grad_norm": 0.8948590159416199, + "learning_rate": 4.917505603986605e-06, + "loss": 0.6666, + "step": 1493 + }, + { + "epoch": 0.5245786516853933, + "grad_norm": 0.9220554828643799, + "learning_rate": 4.9173877770280425e-06, + "loss": 0.7249, + "step": 1494 + }, + { + "epoch": 0.5249297752808989, + "grad_norm": 0.9420086741447449, + "learning_rate": 4.917269867396887e-06, + "loss": 0.7205, + "step": 1495 + }, + { + "epoch": 0.5252808988764045, + "grad_norm": 0.9278038144111633, + "learning_rate": 4.917151875097169e-06, + "loss": 0.7032, + "step": 1496 + }, + { + "epoch": 0.5256320224719101, + "grad_norm": 0.8682548999786377, + "learning_rate": 4.917033800132927e-06, + "loss": 0.7, + "step": 1497 + }, + { + "epoch": 0.5259831460674157, + "grad_norm": 0.8908578753471375, + "learning_rate": 4.916915642508196e-06, + "loss": 0.6891, + "step": 1498 + }, + { + "epoch": 0.5263342696629213, + "grad_norm": 0.8928547501564026, + "learning_rate": 4.916797402227019e-06, + "loss": 0.7247, + "step": 1499 + }, + { + "epoch": 0.526685393258427, + "grad_norm": 0.9188054203987122, + "learning_rate": 4.916679079293438e-06, + "loss": 0.7001, + "step": 1500 + }, + { + "epoch": 0.5270365168539326, + "grad_norm": 0.8950523138046265, + "learning_rate": 4.916560673711502e-06, + "loss": 0.6473, + "step": 1501 + }, + { + "epoch": 0.5273876404494382, + "grad_norm": 0.9325989484786987, + "learning_rate": 4.916442185485258e-06, + "loss": 0.7096, + "step": 1502 + }, + { + "epoch": 0.5277387640449438, + "grad_norm": 0.9268350005149841, + "learning_rate": 4.9163236146187586e-06, + "loss": 0.7089, + "step": 1503 + }, + { + "epoch": 0.5280898876404494, + "grad_norm": 1.009692907333374, + "learning_rate": 4.91620496111606e-06, + "loss": 0.6721, + "step": 1504 + }, + { + "epoch": 0.5284410112359551, + "grad_norm": 0.8952979445457458, + "learning_rate": 4.916086224981219e-06, + "loss": 0.7074, + "step": 1505 + }, + { + "epoch": 0.5287921348314607, + "grad_norm": 0.896232008934021, + "learning_rate": 4.915967406218296e-06, + "loss": 0.6765, + "step": 1506 + }, + { + "epoch": 0.5291432584269663, + "grad_norm": 0.9074482917785645, + "learning_rate": 4.915848504831356e-06, + "loss": 0.7296, + "step": 1507 + }, + { + "epoch": 0.5294943820224719, + "grad_norm": 0.8722943067550659, + "learning_rate": 4.9157295208244635e-06, + "loss": 0.636, + "step": 1508 + }, + { + "epoch": 0.5298455056179775, + "grad_norm": 0.917375922203064, + "learning_rate": 4.915610454201689e-06, + "loss": 0.6844, + "step": 1509 + }, + { + "epoch": 0.5301966292134831, + "grad_norm": 0.8942883610725403, + "learning_rate": 4.9154913049671045e-06, + "loss": 0.6589, + "step": 1510 + }, + { + "epoch": 0.5305477528089888, + "grad_norm": 0.894601583480835, + "learning_rate": 4.915372073124784e-06, + "loss": 0.6919, + "step": 1511 + }, + { + "epoch": 0.5308988764044944, + "grad_norm": 1.0230928659439087, + "learning_rate": 4.9152527586788056e-06, + "loss": 0.6949, + "step": 1512 + }, + { + "epoch": 0.53125, + "grad_norm": 1.0016642808914185, + "learning_rate": 4.915133361633249e-06, + "loss": 0.6792, + "step": 1513 + }, + { + "epoch": 0.5316011235955056, + "grad_norm": 0.8965452909469604, + "learning_rate": 4.9150138819921985e-06, + "loss": 0.6993, + "step": 1514 + }, + { + "epoch": 0.5319522471910112, + "grad_norm": 0.9045464396476746, + "learning_rate": 4.91489431975974e-06, + "loss": 0.6905, + "step": 1515 + }, + { + "epoch": 0.5323033707865169, + "grad_norm": 0.9282147288322449, + "learning_rate": 4.914774674939962e-06, + "loss": 0.7121, + "step": 1516 + }, + { + "epoch": 0.5326544943820225, + "grad_norm": 0.9581758379936218, + "learning_rate": 4.914654947536956e-06, + "loss": 0.7156, + "step": 1517 + }, + { + "epoch": 0.5330056179775281, + "grad_norm": 0.8986783623695374, + "learning_rate": 4.9145351375548184e-06, + "loss": 0.6827, + "step": 1518 + }, + { + "epoch": 0.5333567415730337, + "grad_norm": 1.0093053579330444, + "learning_rate": 4.9144152449976444e-06, + "loss": 0.6642, + "step": 1519 + }, + { + "epoch": 0.5337078651685393, + "grad_norm": 1.0274914503097534, + "learning_rate": 4.914295269869535e-06, + "loss": 0.6667, + "step": 1520 + }, + { + "epoch": 0.5340589887640449, + "grad_norm": 0.9218011498451233, + "learning_rate": 4.914175212174594e-06, + "loss": 0.6902, + "step": 1521 + }, + { + "epoch": 0.5344101123595506, + "grad_norm": 0.902390718460083, + "learning_rate": 4.914055071916925e-06, + "loss": 0.7009, + "step": 1522 + }, + { + "epoch": 0.5347612359550562, + "grad_norm": 0.9239642024040222, + "learning_rate": 4.9139348491006396e-06, + "loss": 0.7038, + "step": 1523 + }, + { + "epoch": 0.5351123595505618, + "grad_norm": 0.8971140384674072, + "learning_rate": 4.913814543729847e-06, + "loss": 0.7187, + "step": 1524 + }, + { + "epoch": 0.5354634831460674, + "grad_norm": 0.9525965452194214, + "learning_rate": 4.913694155808663e-06, + "loss": 0.6694, + "step": 1525 + }, + { + "epoch": 0.535814606741573, + "grad_norm": 0.915665864944458, + "learning_rate": 4.913573685341205e-06, + "loss": 0.7181, + "step": 1526 + }, + { + "epoch": 0.5361657303370787, + "grad_norm": 0.913693368434906, + "learning_rate": 4.913453132331592e-06, + "loss": 0.7122, + "step": 1527 + }, + { + "epoch": 0.5365168539325843, + "grad_norm": 0.91572505235672, + "learning_rate": 4.913332496783947e-06, + "loss": 0.7143, + "step": 1528 + }, + { + "epoch": 0.5368679775280899, + "grad_norm": 0.947823166847229, + "learning_rate": 4.913211778702395e-06, + "loss": 0.7239, + "step": 1529 + }, + { + "epoch": 0.5372191011235955, + "grad_norm": 0.9384491443634033, + "learning_rate": 4.913090978091066e-06, + "loss": 0.7147, + "step": 1530 + }, + { + "epoch": 0.5375702247191011, + "grad_norm": 0.882896363735199, + "learning_rate": 4.912970094954091e-06, + "loss": 0.6548, + "step": 1531 + }, + { + "epoch": 0.5379213483146067, + "grad_norm": 0.8977818489074707, + "learning_rate": 4.912849129295601e-06, + "loss": 0.6898, + "step": 1532 + }, + { + "epoch": 0.5382724719101124, + "grad_norm": 0.8661186099052429, + "learning_rate": 4.912728081119738e-06, + "loss": 0.6605, + "step": 1533 + }, + { + "epoch": 0.538623595505618, + "grad_norm": 0.8856679797172546, + "learning_rate": 4.9126069504306385e-06, + "loss": 0.6891, + "step": 1534 + }, + { + "epoch": 0.5389747191011236, + "grad_norm": 0.9184430241584778, + "learning_rate": 4.9124857372324455e-06, + "loss": 0.7221, + "step": 1535 + }, + { + "epoch": 0.5393258426966292, + "grad_norm": 0.9198493957519531, + "learning_rate": 4.912364441529306e-06, + "loss": 0.7011, + "step": 1536 + }, + { + "epoch": 0.5396769662921348, + "grad_norm": 0.9443540573120117, + "learning_rate": 4.912243063325365e-06, + "loss": 0.6523, + "step": 1537 + }, + { + "epoch": 0.5400280898876404, + "grad_norm": 0.9004515409469604, + "learning_rate": 4.912121602624777e-06, + "loss": 0.6674, + "step": 1538 + }, + { + "epoch": 0.5403792134831461, + "grad_norm": 0.9147352576255798, + "learning_rate": 4.912000059431693e-06, + "loss": 0.684, + "step": 1539 + }, + { + "epoch": 0.5407303370786517, + "grad_norm": 0.9233749508857727, + "learning_rate": 4.911878433750271e-06, + "loss": 0.702, + "step": 1540 + }, + { + "epoch": 0.5410814606741573, + "grad_norm": 0.8945138454437256, + "learning_rate": 4.911756725584672e-06, + "loss": 0.6694, + "step": 1541 + }, + { + "epoch": 0.5414325842696629, + "grad_norm": 0.9332498908042908, + "learning_rate": 4.911634934939055e-06, + "loss": 0.703, + "step": 1542 + }, + { + "epoch": 0.5417837078651685, + "grad_norm": 0.8984339237213135, + "learning_rate": 4.911513061817588e-06, + "loss": 0.6607, + "step": 1543 + }, + { + "epoch": 0.5421348314606742, + "grad_norm": 0.9096390604972839, + "learning_rate": 4.9113911062244364e-06, + "loss": 0.6786, + "step": 1544 + }, + { + "epoch": 0.5424859550561798, + "grad_norm": 0.8916593790054321, + "learning_rate": 4.911269068163775e-06, + "loss": 0.6644, + "step": 1545 + }, + { + "epoch": 0.5428370786516854, + "grad_norm": 0.9222617149353027, + "learning_rate": 4.911146947639772e-06, + "loss": 0.6983, + "step": 1546 + }, + { + "epoch": 0.543188202247191, + "grad_norm": 0.8953000903129578, + "learning_rate": 4.911024744656608e-06, + "loss": 0.7026, + "step": 1547 + }, + { + "epoch": 0.5435393258426966, + "grad_norm": 0.9042890071868896, + "learning_rate": 4.910902459218461e-06, + "loss": 0.7196, + "step": 1548 + }, + { + "epoch": 0.5438904494382022, + "grad_norm": 0.9504362940788269, + "learning_rate": 4.910780091329513e-06, + "loss": 0.6707, + "step": 1549 + }, + { + "epoch": 0.5442415730337079, + "grad_norm": 0.9307934641838074, + "learning_rate": 4.910657640993948e-06, + "loss": 0.683, + "step": 1550 + }, + { + "epoch": 0.5445926966292135, + "grad_norm": 0.9294170141220093, + "learning_rate": 4.910535108215955e-06, + "loss": 0.6969, + "step": 1551 + }, + { + "epoch": 0.5449438202247191, + "grad_norm": 0.9138461351394653, + "learning_rate": 4.910412492999725e-06, + "loss": 0.7347, + "step": 1552 + }, + { + "epoch": 0.5452949438202247, + "grad_norm": 0.8926948308944702, + "learning_rate": 4.910289795349449e-06, + "loss": 0.7217, + "step": 1553 + }, + { + "epoch": 0.5456460674157303, + "grad_norm": 0.8452489376068115, + "learning_rate": 4.910167015269325e-06, + "loss": 0.6478, + "step": 1554 + }, + { + "epoch": 0.545997191011236, + "grad_norm": 0.8928716778755188, + "learning_rate": 4.9100441527635516e-06, + "loss": 0.6246, + "step": 1555 + }, + { + "epoch": 0.5463483146067416, + "grad_norm": 0.8877882361412048, + "learning_rate": 4.90992120783633e-06, + "loss": 0.6771, + "step": 1556 + }, + { + "epoch": 0.5466994382022472, + "grad_norm": 0.8971658945083618, + "learning_rate": 4.909798180491865e-06, + "loss": 0.6664, + "step": 1557 + }, + { + "epoch": 0.5470505617977528, + "grad_norm": 0.9224132895469666, + "learning_rate": 4.909675070734365e-06, + "loss": 0.6958, + "step": 1558 + }, + { + "epoch": 0.5474016853932584, + "grad_norm": 0.9234408140182495, + "learning_rate": 4.9095518785680395e-06, + "loss": 0.6944, + "step": 1559 + }, + { + "epoch": 0.547752808988764, + "grad_norm": 0.9604536294937134, + "learning_rate": 4.909428603997101e-06, + "loss": 0.6597, + "step": 1560 + }, + { + "epoch": 0.5481039325842697, + "grad_norm": 0.9268457889556885, + "learning_rate": 4.909305247025767e-06, + "loss": 0.6942, + "step": 1561 + }, + { + "epoch": 0.5484550561797753, + "grad_norm": 0.9222556352615356, + "learning_rate": 4.909181807658256e-06, + "loss": 0.6747, + "step": 1562 + }, + { + "epoch": 0.5488061797752809, + "grad_norm": 0.8885440826416016, + "learning_rate": 4.909058285898787e-06, + "loss": 0.6555, + "step": 1563 + }, + { + "epoch": 0.5491573033707865, + "grad_norm": 0.9378191232681274, + "learning_rate": 4.908934681751586e-06, + "loss": 0.7272, + "step": 1564 + }, + { + "epoch": 0.5495084269662921, + "grad_norm": 0.9337999820709229, + "learning_rate": 4.908810995220882e-06, + "loss": 0.6721, + "step": 1565 + }, + { + "epoch": 0.5498595505617978, + "grad_norm": 0.8756289482116699, + "learning_rate": 4.908687226310902e-06, + "loss": 0.6733, + "step": 1566 + }, + { + "epoch": 0.5502106741573034, + "grad_norm": 0.8794893026351929, + "learning_rate": 4.90856337502588e-06, + "loss": 0.6225, + "step": 1567 + }, + { + "epoch": 0.550561797752809, + "grad_norm": 0.8720929026603699, + "learning_rate": 4.9084394413700515e-06, + "loss": 0.7074, + "step": 1568 + }, + { + "epoch": 0.5509129213483146, + "grad_norm": 0.8621073365211487, + "learning_rate": 4.908315425347655e-06, + "loss": 0.6439, + "step": 1569 + }, + { + "epoch": 0.5512640449438202, + "grad_norm": 0.937415361404419, + "learning_rate": 4.908191326962932e-06, + "loss": 0.7121, + "step": 1570 + }, + { + "epoch": 0.5516151685393258, + "grad_norm": 0.908561646938324, + "learning_rate": 4.908067146220127e-06, + "loss": 0.699, + "step": 1571 + }, + { + "epoch": 0.5519662921348315, + "grad_norm": 0.8949597477912903, + "learning_rate": 4.907942883123485e-06, + "loss": 0.7148, + "step": 1572 + }, + { + "epoch": 0.5523174157303371, + "grad_norm": 0.9320929646492004, + "learning_rate": 4.907818537677258e-06, + "loss": 0.6894, + "step": 1573 + }, + { + "epoch": 0.5526685393258427, + "grad_norm": 0.8810732960700989, + "learning_rate": 4.907694109885697e-06, + "loss": 0.6814, + "step": 1574 + }, + { + "epoch": 0.5530196629213483, + "grad_norm": 0.9485206007957458, + "learning_rate": 4.907569599753058e-06, + "loss": 0.7125, + "step": 1575 + }, + { + "epoch": 0.5533707865168539, + "grad_norm": 0.8823009729385376, + "learning_rate": 4.907445007283598e-06, + "loss": 0.7239, + "step": 1576 + }, + { + "epoch": 0.5537219101123596, + "grad_norm": 0.9091922640800476, + "learning_rate": 4.907320332481579e-06, + "loss": 0.6949, + "step": 1577 + }, + { + "epoch": 0.5540730337078652, + "grad_norm": 0.9078152775764465, + "learning_rate": 4.907195575351265e-06, + "loss": 0.701, + "step": 1578 + }, + { + "epoch": 0.5544241573033708, + "grad_norm": 0.9011525511741638, + "learning_rate": 4.907070735896923e-06, + "loss": 0.7039, + "step": 1579 + }, + { + "epoch": 0.5547752808988764, + "grad_norm": 0.8794327974319458, + "learning_rate": 4.90694581412282e-06, + "loss": 0.6543, + "step": 1580 + }, + { + "epoch": 0.555126404494382, + "grad_norm": 0.921090841293335, + "learning_rate": 4.90682081003323e-06, + "loss": 0.7031, + "step": 1581 + }, + { + "epoch": 0.5554775280898876, + "grad_norm": 0.8931931853294373, + "learning_rate": 4.906695723632429e-06, + "loss": 0.6938, + "step": 1582 + }, + { + "epoch": 0.5558286516853933, + "grad_norm": 0.9168527722358704, + "learning_rate": 4.9065705549246935e-06, + "loss": 0.7472, + "step": 1583 + }, + { + "epoch": 0.5561797752808989, + "grad_norm": 0.9369787573814392, + "learning_rate": 4.906445303914303e-06, + "loss": 0.6945, + "step": 1584 + }, + { + "epoch": 0.5565308988764045, + "grad_norm": 0.9574870467185974, + "learning_rate": 4.906319970605544e-06, + "loss": 0.7117, + "step": 1585 + }, + { + "epoch": 0.5568820224719101, + "grad_norm": 0.9056572318077087, + "learning_rate": 4.9061945550027e-06, + "loss": 0.7117, + "step": 1586 + }, + { + "epoch": 0.5572331460674157, + "grad_norm": 0.9152799844741821, + "learning_rate": 4.906069057110062e-06, + "loss": 0.7262, + "step": 1587 + }, + { + "epoch": 0.5575842696629213, + "grad_norm": 0.9061770439147949, + "learning_rate": 4.9059434769319205e-06, + "loss": 0.6746, + "step": 1588 + }, + { + "epoch": 0.557935393258427, + "grad_norm": 0.8756861090660095, + "learning_rate": 4.905817814472572e-06, + "loss": 0.6547, + "step": 1589 + }, + { + "epoch": 0.5582865168539326, + "grad_norm": 0.8801287412643433, + "learning_rate": 4.905692069736312e-06, + "loss": 0.7026, + "step": 1590 + }, + { + "epoch": 0.5586376404494382, + "grad_norm": 0.871894359588623, + "learning_rate": 4.9055662427274415e-06, + "loss": 0.6967, + "step": 1591 + }, + { + "epoch": 0.5589887640449438, + "grad_norm": 0.8951716423034668, + "learning_rate": 4.905440333450264e-06, + "loss": 0.6769, + "step": 1592 + }, + { + "epoch": 0.5593398876404494, + "grad_norm": 0.8958552479743958, + "learning_rate": 4.905314341909086e-06, + "loss": 0.6462, + "step": 1593 + }, + { + "epoch": 0.5596910112359551, + "grad_norm": 0.8950044512748718, + "learning_rate": 4.9051882681082155e-06, + "loss": 0.6943, + "step": 1594 + }, + { + "epoch": 0.5600421348314607, + "grad_norm": 0.8931154608726501, + "learning_rate": 4.905062112051965e-06, + "loss": 0.6712, + "step": 1595 + }, + { + "epoch": 0.5603932584269663, + "grad_norm": 0.9284601211547852, + "learning_rate": 4.904935873744647e-06, + "loss": 0.7057, + "step": 1596 + }, + { + "epoch": 0.5607443820224719, + "grad_norm": 0.8948965072631836, + "learning_rate": 4.904809553190582e-06, + "loss": 0.6923, + "step": 1597 + }, + { + "epoch": 0.5610955056179775, + "grad_norm": 0.9073783755302429, + "learning_rate": 4.904683150394086e-06, + "loss": 0.7183, + "step": 1598 + }, + { + "epoch": 0.5614466292134831, + "grad_norm": 0.9169145226478577, + "learning_rate": 4.904556665359485e-06, + "loss": 0.71, + "step": 1599 + }, + { + "epoch": 0.5617977528089888, + "grad_norm": 1.0053675174713135, + "learning_rate": 4.904430098091104e-06, + "loss": 0.7078, + "step": 1600 + }, + { + "epoch": 0.5621488764044944, + "grad_norm": 0.893157422542572, + "learning_rate": 4.904303448593271e-06, + "loss": 0.6814, + "step": 1601 + }, + { + "epoch": 0.5625, + "grad_norm": 0.8943068981170654, + "learning_rate": 4.904176716870317e-06, + "loss": 0.6915, + "step": 1602 + }, + { + "epoch": 0.5628511235955056, + "grad_norm": 0.8920650482177734, + "learning_rate": 4.904049902926576e-06, + "loss": 0.6953, + "step": 1603 + }, + { + "epoch": 0.5632022471910112, + "grad_norm": 0.8903055787086487, + "learning_rate": 4.9039230067663865e-06, + "loss": 0.6708, + "step": 1604 + }, + { + "epoch": 0.5635533707865169, + "grad_norm": 0.9348129034042358, + "learning_rate": 4.903796028394086e-06, + "loss": 0.6983, + "step": 1605 + }, + { + "epoch": 0.5639044943820225, + "grad_norm": 0.946652889251709, + "learning_rate": 4.903668967814021e-06, + "loss": 0.6997, + "step": 1606 + }, + { + "epoch": 0.5642556179775281, + "grad_norm": 0.9114198684692383, + "learning_rate": 4.9035418250305314e-06, + "loss": 0.6741, + "step": 1607 + }, + { + "epoch": 0.5646067415730337, + "grad_norm": 0.8928046226501465, + "learning_rate": 4.903414600047969e-06, + "loss": 0.6866, + "step": 1608 + }, + { + "epoch": 0.5649578651685393, + "grad_norm": 0.9179702997207642, + "learning_rate": 4.903287292870684e-06, + "loss": 0.696, + "step": 1609 + }, + { + "epoch": 0.5653089887640449, + "grad_norm": 0.9323627948760986, + "learning_rate": 4.903159903503031e-06, + "loss": 0.6939, + "step": 1610 + }, + { + "epoch": 0.5656601123595506, + "grad_norm": 0.8928480744361877, + "learning_rate": 4.903032431949365e-06, + "loss": 0.6819, + "step": 1611 + }, + { + "epoch": 0.5660112359550562, + "grad_norm": 0.8628221154212952, + "learning_rate": 4.9029048782140454e-06, + "loss": 0.687, + "step": 1612 + }, + { + "epoch": 0.5663623595505618, + "grad_norm": 0.8917481303215027, + "learning_rate": 4.9027772423014354e-06, + "loss": 0.7087, + "step": 1613 + }, + { + "epoch": 0.5667134831460674, + "grad_norm": 0.8859778642654419, + "learning_rate": 4.9026495242159e-06, + "loss": 0.6619, + "step": 1614 + }, + { + "epoch": 0.567064606741573, + "grad_norm": 0.9722530841827393, + "learning_rate": 4.902521723961807e-06, + "loss": 0.7133, + "step": 1615 + }, + { + "epoch": 0.5674157303370787, + "grad_norm": 0.8985425233840942, + "learning_rate": 4.9023938415435265e-06, + "loss": 0.6599, + "step": 1616 + }, + { + "epoch": 0.5677668539325843, + "grad_norm": 0.9105679392814636, + "learning_rate": 4.902265876965434e-06, + "loss": 0.6985, + "step": 1617 + }, + { + "epoch": 0.5681179775280899, + "grad_norm": 0.8673763871192932, + "learning_rate": 4.902137830231903e-06, + "loss": 0.692, + "step": 1618 + }, + { + "epoch": 0.5684691011235955, + "grad_norm": 0.8669525384902954, + "learning_rate": 4.902009701347313e-06, + "loss": 0.6515, + "step": 1619 + }, + { + "epoch": 0.5688202247191011, + "grad_norm": 0.9088237881660461, + "learning_rate": 4.9018814903160475e-06, + "loss": 0.6847, + "step": 1620 + }, + { + "epoch": 0.5691713483146067, + "grad_norm": 0.9107149243354797, + "learning_rate": 4.90175319714249e-06, + "loss": 0.6769, + "step": 1621 + }, + { + "epoch": 0.5695224719101124, + "grad_norm": 0.8820178508758545, + "learning_rate": 4.901624821831028e-06, + "loss": 0.7169, + "step": 1622 + }, + { + "epoch": 0.569873595505618, + "grad_norm": 0.8633939623832703, + "learning_rate": 4.901496364386053e-06, + "loss": 0.6395, + "step": 1623 + }, + { + "epoch": 0.5702247191011236, + "grad_norm": 0.9044683575630188, + "learning_rate": 4.901367824811958e-06, + "loss": 0.7011, + "step": 1624 + }, + { + "epoch": 0.5705758426966292, + "grad_norm": 0.9092311263084412, + "learning_rate": 4.901239203113137e-06, + "loss": 0.6643, + "step": 1625 + }, + { + "epoch": 0.5709269662921348, + "grad_norm": 0.9932267665863037, + "learning_rate": 4.90111049929399e-06, + "loss": 0.7203, + "step": 1626 + }, + { + "epoch": 0.5712780898876404, + "grad_norm": 0.8947046995162964, + "learning_rate": 4.9009817133589186e-06, + "loss": 0.683, + "step": 1627 + }, + { + "epoch": 0.5716292134831461, + "grad_norm": 0.9035671353340149, + "learning_rate": 4.900852845312328e-06, + "loss": 0.6761, + "step": 1628 + }, + { + "epoch": 0.5719803370786517, + "grad_norm": 0.905048668384552, + "learning_rate": 4.900723895158624e-06, + "loss": 0.7018, + "step": 1629 + }, + { + "epoch": 0.5723314606741573, + "grad_norm": 0.9121686220169067, + "learning_rate": 4.9005948629022175e-06, + "loss": 0.7506, + "step": 1630 + }, + { + "epoch": 0.5726825842696629, + "grad_norm": 0.9041679501533508, + "learning_rate": 4.90046574854752e-06, + "loss": 0.7058, + "step": 1631 + }, + { + "epoch": 0.5730337078651685, + "grad_norm": 0.8760393857955933, + "learning_rate": 4.900336552098949e-06, + "loss": 0.7067, + "step": 1632 + }, + { + "epoch": 0.5733848314606742, + "grad_norm": 0.9228975772857666, + "learning_rate": 4.900207273560921e-06, + "loss": 0.6959, + "step": 1633 + }, + { + "epoch": 0.5737359550561798, + "grad_norm": 0.8803319334983826, + "learning_rate": 4.900077912937858e-06, + "loss": 0.634, + "step": 1634 + }, + { + "epoch": 0.5740870786516854, + "grad_norm": 0.8787641525268555, + "learning_rate": 4.899948470234185e-06, + "loss": 0.7098, + "step": 1635 + }, + { + "epoch": 0.574438202247191, + "grad_norm": 0.8598030209541321, + "learning_rate": 4.899818945454328e-06, + "loss": 0.7062, + "step": 1636 + }, + { + "epoch": 0.5747893258426966, + "grad_norm": 0.920392632484436, + "learning_rate": 4.899689338602716e-06, + "loss": 0.6514, + "step": 1637 + }, + { + "epoch": 0.5751404494382022, + "grad_norm": 0.9178962111473083, + "learning_rate": 4.899559649683783e-06, + "loss": 0.6626, + "step": 1638 + }, + { + "epoch": 0.5754915730337079, + "grad_norm": 0.8671183586120605, + "learning_rate": 4.899429878701963e-06, + "loss": 0.704, + "step": 1639 + }, + { + "epoch": 0.5758426966292135, + "grad_norm": 0.8965294361114502, + "learning_rate": 4.899300025661694e-06, + "loss": 0.7058, + "step": 1640 + }, + { + "epoch": 0.5761938202247191, + "grad_norm": 0.8801050186157227, + "learning_rate": 4.8991700905674165e-06, + "loss": 0.695, + "step": 1641 + }, + { + "epoch": 0.5765449438202247, + "grad_norm": 0.9138525724411011, + "learning_rate": 4.899040073423577e-06, + "loss": 0.7061, + "step": 1642 + }, + { + "epoch": 0.5768960674157303, + "grad_norm": 0.8814200162887573, + "learning_rate": 4.898909974234618e-06, + "loss": 0.6659, + "step": 1643 + }, + { + "epoch": 0.577247191011236, + "grad_norm": 0.927841305732727, + "learning_rate": 4.898779793004991e-06, + "loss": 0.6864, + "step": 1644 + }, + { + "epoch": 0.5775983146067416, + "grad_norm": 0.9304884076118469, + "learning_rate": 4.898649529739148e-06, + "loss": 0.66, + "step": 1645 + }, + { + "epoch": 0.5779494382022472, + "grad_norm": 0.8991161584854126, + "learning_rate": 4.898519184441544e-06, + "loss": 0.6901, + "step": 1646 + }, + { + "epoch": 0.5783005617977528, + "grad_norm": 0.8874373435974121, + "learning_rate": 4.898388757116636e-06, + "loss": 0.6866, + "step": 1647 + }, + { + "epoch": 0.5786516853932584, + "grad_norm": 0.8752896785736084, + "learning_rate": 4.898258247768885e-06, + "loss": 0.6966, + "step": 1648 + }, + { + "epoch": 0.579002808988764, + "grad_norm": 0.9005473852157593, + "learning_rate": 4.898127656402753e-06, + "loss": 0.7192, + "step": 1649 + }, + { + "epoch": 0.5793539325842697, + "grad_norm": 0.8968993425369263, + "learning_rate": 4.897996983022709e-06, + "loss": 0.7156, + "step": 1650 + }, + { + "epoch": 0.5797050561797753, + "grad_norm": 0.881846010684967, + "learning_rate": 4.897866227633219e-06, + "loss": 0.6848, + "step": 1651 + }, + { + "epoch": 0.5800561797752809, + "grad_norm": 0.9290207028388977, + "learning_rate": 4.897735390238756e-06, + "loss": 0.7263, + "step": 1652 + }, + { + "epoch": 0.5804073033707865, + "grad_norm": 0.8702242374420166, + "learning_rate": 4.897604470843795e-06, + "loss": 0.6796, + "step": 1653 + }, + { + "epoch": 0.5807584269662921, + "grad_norm": 0.9286425709724426, + "learning_rate": 4.897473469452812e-06, + "loss": 0.7203, + "step": 1654 + }, + { + "epoch": 0.5811095505617978, + "grad_norm": 0.9703604578971863, + "learning_rate": 4.897342386070289e-06, + "loss": 0.6574, + "step": 1655 + }, + { + "epoch": 0.5814606741573034, + "grad_norm": 0.9450727701187134, + "learning_rate": 4.897211220700707e-06, + "loss": 0.7004, + "step": 1656 + }, + { + "epoch": 0.581811797752809, + "grad_norm": 0.8609772324562073, + "learning_rate": 4.897079973348553e-06, + "loss": 0.658, + "step": 1657 + }, + { + "epoch": 0.5821629213483146, + "grad_norm": 0.8902559876441956, + "learning_rate": 4.896948644018315e-06, + "loss": 0.6995, + "step": 1658 + }, + { + "epoch": 0.5825140449438202, + "grad_norm": 0.8805756568908691, + "learning_rate": 4.896817232714484e-06, + "loss": 0.6919, + "step": 1659 + }, + { + "epoch": 0.5828651685393258, + "grad_norm": 0.8647357225418091, + "learning_rate": 4.896685739441555e-06, + "loss": 0.7053, + "step": 1660 + }, + { + "epoch": 0.5832162921348315, + "grad_norm": 0.8912749290466309, + "learning_rate": 4.896554164204025e-06, + "loss": 0.6848, + "step": 1661 + }, + { + "epoch": 0.5835674157303371, + "grad_norm": 0.8912277221679688, + "learning_rate": 4.896422507006392e-06, + "loss": 0.687, + "step": 1662 + }, + { + "epoch": 0.5839185393258427, + "grad_norm": 0.8887133598327637, + "learning_rate": 4.896290767853161e-06, + "loss": 0.6703, + "step": 1663 + }, + { + "epoch": 0.5842696629213483, + "grad_norm": 0.9464722871780396, + "learning_rate": 4.896158946748837e-06, + "loss": 0.6834, + "step": 1664 + }, + { + "epoch": 0.5846207865168539, + "grad_norm": 0.8864126801490784, + "learning_rate": 4.8960270436979265e-06, + "loss": 0.6983, + "step": 1665 + }, + { + "epoch": 0.5849719101123596, + "grad_norm": 0.9276545643806458, + "learning_rate": 4.895895058704941e-06, + "loss": 0.6989, + "step": 1666 + }, + { + "epoch": 0.5853230337078652, + "grad_norm": 0.9041825532913208, + "learning_rate": 4.8957629917743955e-06, + "loss": 0.7188, + "step": 1667 + }, + { + "epoch": 0.5856741573033708, + "grad_norm": 0.9835352301597595, + "learning_rate": 4.895630842910805e-06, + "loss": 0.6981, + "step": 1668 + }, + { + "epoch": 0.5860252808988764, + "grad_norm": 0.8841162323951721, + "learning_rate": 4.89549861211869e-06, + "loss": 0.6686, + "step": 1669 + }, + { + "epoch": 0.586376404494382, + "grad_norm": 0.9026679992675781, + "learning_rate": 4.895366299402572e-06, + "loss": 0.6782, + "step": 1670 + }, + { + "epoch": 0.5867275280898876, + "grad_norm": 0.8805919885635376, + "learning_rate": 4.895233904766978e-06, + "loss": 0.7267, + "step": 1671 + }, + { + "epoch": 0.5870786516853933, + "grad_norm": 0.8851740956306458, + "learning_rate": 4.895101428216432e-06, + "loss": 0.6841, + "step": 1672 + }, + { + "epoch": 0.5874297752808989, + "grad_norm": 0.911161482334137, + "learning_rate": 4.894968869755467e-06, + "loss": 0.6938, + "step": 1673 + }, + { + "epoch": 0.5877808988764045, + "grad_norm": 0.900994598865509, + "learning_rate": 4.894836229388617e-06, + "loss": 0.6809, + "step": 1674 + }, + { + "epoch": 0.5881320224719101, + "grad_norm": 0.9121220111846924, + "learning_rate": 4.894703507120417e-06, + "loss": 0.6941, + "step": 1675 + }, + { + "epoch": 0.5884831460674157, + "grad_norm": 0.8991403579711914, + "learning_rate": 4.894570702955406e-06, + "loss": 0.6925, + "step": 1676 + }, + { + "epoch": 0.5888342696629213, + "grad_norm": 0.8850685358047485, + "learning_rate": 4.894437816898125e-06, + "loss": 0.6904, + "step": 1677 + }, + { + "epoch": 0.589185393258427, + "grad_norm": 0.8999640345573425, + "learning_rate": 4.894304848953122e-06, + "loss": 0.6959, + "step": 1678 + }, + { + "epoch": 0.5895365168539326, + "grad_norm": 0.9228804707527161, + "learning_rate": 4.89417179912494e-06, + "loss": 0.6663, + "step": 1679 + }, + { + "epoch": 0.5898876404494382, + "grad_norm": 0.8978350162506104, + "learning_rate": 4.894038667418132e-06, + "loss": 0.6948, + "step": 1680 + }, + { + "epoch": 0.5902387640449438, + "grad_norm": 0.9258869290351868, + "learning_rate": 4.89390545383725e-06, + "loss": 0.7268, + "step": 1681 + }, + { + "epoch": 0.5905898876404494, + "grad_norm": 0.9010106325149536, + "learning_rate": 4.893772158386851e-06, + "loss": 0.6819, + "step": 1682 + }, + { + "epoch": 0.5909410112359551, + "grad_norm": 0.9146981835365295, + "learning_rate": 4.8936387810714904e-06, + "loss": 0.6942, + "step": 1683 + }, + { + "epoch": 0.5912921348314607, + "grad_norm": 0.959784746170044, + "learning_rate": 4.893505321895733e-06, + "loss": 0.7312, + "step": 1684 + }, + { + "epoch": 0.5916432584269663, + "grad_norm": 0.9104639887809753, + "learning_rate": 4.8933717808641415e-06, + "loss": 0.6842, + "step": 1685 + }, + { + "epoch": 0.5919943820224719, + "grad_norm": 0.8836262822151184, + "learning_rate": 4.893238157981283e-06, + "loss": 0.6613, + "step": 1686 + }, + { + "epoch": 0.5923455056179775, + "grad_norm": 0.903002142906189, + "learning_rate": 4.893104453251728e-06, + "loss": 0.6558, + "step": 1687 + }, + { + "epoch": 0.5926966292134831, + "grad_norm": 0.9008224010467529, + "learning_rate": 4.892970666680048e-06, + "loss": 0.7283, + "step": 1688 + }, + { + "epoch": 0.5930477528089888, + "grad_norm": 0.9016857743263245, + "learning_rate": 4.892836798270819e-06, + "loss": 0.704, + "step": 1689 + }, + { + "epoch": 0.5933988764044944, + "grad_norm": 0.9839923977851868, + "learning_rate": 4.8927028480286195e-06, + "loss": 0.6804, + "step": 1690 + }, + { + "epoch": 0.59375, + "grad_norm": 0.9112361669540405, + "learning_rate": 4.8925688159580294e-06, + "loss": 0.6885, + "step": 1691 + }, + { + "epoch": 0.5941011235955056, + "grad_norm": 0.9427493214607239, + "learning_rate": 4.892434702063634e-06, + "loss": 0.6816, + "step": 1692 + }, + { + "epoch": 0.5944522471910112, + "grad_norm": 0.9189691543579102, + "learning_rate": 4.892300506350018e-06, + "loss": 0.7258, + "step": 1693 + }, + { + "epoch": 0.5948033707865169, + "grad_norm": 0.9497784972190857, + "learning_rate": 4.892166228821772e-06, + "loss": 0.6805, + "step": 1694 + }, + { + "epoch": 0.5951544943820225, + "grad_norm": 0.867682933807373, + "learning_rate": 4.892031869483489e-06, + "loss": 0.6952, + "step": 1695 + }, + { + "epoch": 0.5955056179775281, + "grad_norm": 0.9247297644615173, + "learning_rate": 4.891897428339763e-06, + "loss": 0.6892, + "step": 1696 + }, + { + "epoch": 0.5958567415730337, + "grad_norm": 0.9473358392715454, + "learning_rate": 4.89176290539519e-06, + "loss": 0.704, + "step": 1697 + }, + { + "epoch": 0.5962078651685393, + "grad_norm": 0.9231151938438416, + "learning_rate": 4.891628300654373e-06, + "loss": 0.725, + "step": 1698 + }, + { + "epoch": 0.5965589887640449, + "grad_norm": 0.9092977046966553, + "learning_rate": 4.891493614121915e-06, + "loss": 0.7209, + "step": 1699 + }, + { + "epoch": 0.5969101123595506, + "grad_norm": 0.9357640147209167, + "learning_rate": 4.891358845802422e-06, + "loss": 0.6577, + "step": 1700 + }, + { + "epoch": 0.5972612359550562, + "grad_norm": 0.9201985001564026, + "learning_rate": 4.891223995700503e-06, + "loss": 0.7037, + "step": 1701 + }, + { + "epoch": 0.5976123595505618, + "grad_norm": 0.8873964548110962, + "learning_rate": 4.8910890638207685e-06, + "loss": 0.6655, + "step": 1702 + }, + { + "epoch": 0.5979634831460674, + "grad_norm": 0.8884997367858887, + "learning_rate": 4.890954050167835e-06, + "loss": 0.6927, + "step": 1703 + }, + { + "epoch": 0.598314606741573, + "grad_norm": 0.9006896018981934, + "learning_rate": 4.890818954746318e-06, + "loss": 0.6879, + "step": 1704 + }, + { + "epoch": 0.5986657303370787, + "grad_norm": 0.8921645283699036, + "learning_rate": 4.8906837775608395e-06, + "loss": 0.7091, + "step": 1705 + }, + { + "epoch": 0.5990168539325843, + "grad_norm": 0.8784016370773315, + "learning_rate": 4.890548518616022e-06, + "loss": 0.6611, + "step": 1706 + }, + { + "epoch": 0.5993679775280899, + "grad_norm": 0.8995918035507202, + "learning_rate": 4.89041317791649e-06, + "loss": 0.699, + "step": 1707 + }, + { + "epoch": 0.5997191011235955, + "grad_norm": 0.9034433364868164, + "learning_rate": 4.890277755466872e-06, + "loss": 0.6861, + "step": 1708 + }, + { + "epoch": 0.6000702247191011, + "grad_norm": 0.896384596824646, + "learning_rate": 4.890142251271801e-06, + "loss": 0.6658, + "step": 1709 + }, + { + "epoch": 0.6004213483146067, + "grad_norm": 0.8753630518913269, + "learning_rate": 4.890006665335911e-06, + "loss": 0.6886, + "step": 1710 + }, + { + "epoch": 0.6007724719101124, + "grad_norm": 0.9117717146873474, + "learning_rate": 4.889870997663837e-06, + "loss": 0.6885, + "step": 1711 + }, + { + "epoch": 0.601123595505618, + "grad_norm": 0.9002839922904968, + "learning_rate": 4.889735248260221e-06, + "loss": 0.6707, + "step": 1712 + }, + { + "epoch": 0.6014747191011236, + "grad_norm": 0.849831223487854, + "learning_rate": 4.889599417129703e-06, + "loss": 0.6458, + "step": 1713 + }, + { + "epoch": 0.6018258426966292, + "grad_norm": 0.9035419225692749, + "learning_rate": 4.889463504276931e-06, + "loss": 0.6465, + "step": 1714 + }, + { + "epoch": 0.6021769662921348, + "grad_norm": 0.8928185701370239, + "learning_rate": 4.889327509706552e-06, + "loss": 0.6596, + "step": 1715 + }, + { + "epoch": 0.6025280898876404, + "grad_norm": 0.9307925701141357, + "learning_rate": 4.889191433423216e-06, + "loss": 0.6855, + "step": 1716 + }, + { + "epoch": 0.6028792134831461, + "grad_norm": 0.8939083814620972, + "learning_rate": 4.889055275431578e-06, + "loss": 0.7251, + "step": 1717 + }, + { + "epoch": 0.6032303370786517, + "grad_norm": 0.8944793343544006, + "learning_rate": 4.888919035736293e-06, + "loss": 0.6952, + "step": 1718 + }, + { + "epoch": 0.6035814606741573, + "grad_norm": 0.92487633228302, + "learning_rate": 4.888782714342022e-06, + "loss": 0.685, + "step": 1719 + }, + { + "epoch": 0.6039325842696629, + "grad_norm": 0.9212489724159241, + "learning_rate": 4.888646311253426e-06, + "loss": 0.6487, + "step": 1720 + }, + { + "epoch": 0.6042837078651685, + "grad_norm": 0.9022998213768005, + "learning_rate": 4.888509826475171e-06, + "loss": 0.6699, + "step": 1721 + }, + { + "epoch": 0.6046348314606742, + "grad_norm": 0.868482232093811, + "learning_rate": 4.8883732600119225e-06, + "loss": 0.6771, + "step": 1722 + }, + { + "epoch": 0.6049859550561798, + "grad_norm": 0.8934330344200134, + "learning_rate": 4.888236611868353e-06, + "loss": 0.7027, + "step": 1723 + }, + { + "epoch": 0.6053370786516854, + "grad_norm": 0.9520792961120605, + "learning_rate": 4.888099882049135e-06, + "loss": 0.701, + "step": 1724 + }, + { + "epoch": 0.605688202247191, + "grad_norm": 0.921665370464325, + "learning_rate": 4.887963070558944e-06, + "loss": 0.7058, + "step": 1725 + }, + { + "epoch": 0.6060393258426966, + "grad_norm": 0.8855586647987366, + "learning_rate": 4.88782617740246e-06, + "loss": 0.6789, + "step": 1726 + }, + { + "epoch": 0.6063904494382022, + "grad_norm": 0.9030165672302246, + "learning_rate": 4.887689202584364e-06, + "loss": 0.6883, + "step": 1727 + }, + { + "epoch": 0.6067415730337079, + "grad_norm": 0.8876542448997498, + "learning_rate": 4.88755214610934e-06, + "loss": 0.713, + "step": 1728 + }, + { + "epoch": 0.6070926966292135, + "grad_norm": 0.9762006998062134, + "learning_rate": 4.887415007982076e-06, + "loss": 0.7003, + "step": 1729 + }, + { + "epoch": 0.6074438202247191, + "grad_norm": 0.9035171270370483, + "learning_rate": 4.887277788207261e-06, + "loss": 0.7089, + "step": 1730 + }, + { + "epoch": 0.6077949438202247, + "grad_norm": 0.8891435265541077, + "learning_rate": 4.887140486789589e-06, + "loss": 0.6846, + "step": 1731 + }, + { + "epoch": 0.6081460674157303, + "grad_norm": 0.899974524974823, + "learning_rate": 4.887003103733755e-06, + "loss": 0.6662, + "step": 1732 + }, + { + "epoch": 0.608497191011236, + "grad_norm": 0.958362340927124, + "learning_rate": 4.886865639044458e-06, + "loss": 0.6732, + "step": 1733 + }, + { + "epoch": 0.6088483146067416, + "grad_norm": 0.9114941954612732, + "learning_rate": 4.8867280927263985e-06, + "loss": 0.6872, + "step": 1734 + }, + { + "epoch": 0.6091994382022472, + "grad_norm": 0.9042050242424011, + "learning_rate": 4.88659046478428e-06, + "loss": 0.6654, + "step": 1735 + }, + { + "epoch": 0.6095505617977528, + "grad_norm": 0.8774390816688538, + "learning_rate": 4.88645275522281e-06, + "loss": 0.666, + "step": 1736 + }, + { + "epoch": 0.6099016853932584, + "grad_norm": 0.8930482268333435, + "learning_rate": 4.8863149640466975e-06, + "loss": 0.6796, + "step": 1737 + }, + { + "epoch": 0.610252808988764, + "grad_norm": 0.9068736433982849, + "learning_rate": 4.886177091260656e-06, + "loss": 0.6599, + "step": 1738 + }, + { + "epoch": 0.6106039325842697, + "grad_norm": 0.8848494291305542, + "learning_rate": 4.8860391368694e-06, + "loss": 0.7131, + "step": 1739 + }, + { + "epoch": 0.6109550561797753, + "grad_norm": 0.9204060435295105, + "learning_rate": 4.885901100877647e-06, + "loss": 0.7236, + "step": 1740 + }, + { + "epoch": 0.6113061797752809, + "grad_norm": 0.912101149559021, + "learning_rate": 4.885762983290117e-06, + "loss": 0.7401, + "step": 1741 + }, + { + "epoch": 0.6116573033707865, + "grad_norm": 0.8925957083702087, + "learning_rate": 4.885624784111536e-06, + "loss": 0.6837, + "step": 1742 + }, + { + "epoch": 0.6120084269662921, + "grad_norm": 0.9438378214836121, + "learning_rate": 4.885486503346628e-06, + "loss": 0.6398, + "step": 1743 + }, + { + "epoch": 0.6123595505617978, + "grad_norm": 0.885036826133728, + "learning_rate": 4.8853481410001225e-06, + "loss": 0.6341, + "step": 1744 + }, + { + "epoch": 0.6127106741573034, + "grad_norm": 0.8819115161895752, + "learning_rate": 4.885209697076752e-06, + "loss": 0.673, + "step": 1745 + }, + { + "epoch": 0.613061797752809, + "grad_norm": 0.9269865155220032, + "learning_rate": 4.8850711715812515e-06, + "loss": 0.7188, + "step": 1746 + }, + { + "epoch": 0.6134129213483146, + "grad_norm": 0.9204302430152893, + "learning_rate": 4.8849325645183575e-06, + "loss": 0.7151, + "step": 1747 + }, + { + "epoch": 0.6137640449438202, + "grad_norm": 0.9573183655738831, + "learning_rate": 4.88479387589281e-06, + "loss": 0.7048, + "step": 1748 + }, + { + "epoch": 0.6141151685393258, + "grad_norm": 1.0055536031723022, + "learning_rate": 4.884655105709354e-06, + "loss": 0.6865, + "step": 1749 + }, + { + "epoch": 0.6144662921348315, + "grad_norm": 0.9006555080413818, + "learning_rate": 4.8845162539727335e-06, + "loss": 0.6931, + "step": 1750 + }, + { + "epoch": 0.6148174157303371, + "grad_norm": 0.8895266056060791, + "learning_rate": 4.884377320687697e-06, + "loss": 0.6788, + "step": 1751 + }, + { + "epoch": 0.6151685393258427, + "grad_norm": 0.8907827138900757, + "learning_rate": 4.884238305858997e-06, + "loss": 0.6559, + "step": 1752 + }, + { + "epoch": 0.6155196629213483, + "grad_norm": 0.8671404719352722, + "learning_rate": 4.884099209491386e-06, + "loss": 0.692, + "step": 1753 + }, + { + "epoch": 0.6158707865168539, + "grad_norm": 0.8965876698493958, + "learning_rate": 4.883960031589624e-06, + "loss": 0.6662, + "step": 1754 + }, + { + "epoch": 0.6162219101123596, + "grad_norm": 0.8783154487609863, + "learning_rate": 4.883820772158467e-06, + "loss": 0.6802, + "step": 1755 + }, + { + "epoch": 0.6165730337078652, + "grad_norm": 0.9085580706596375, + "learning_rate": 4.883681431202682e-06, + "loss": 0.7184, + "step": 1756 + }, + { + "epoch": 0.6169241573033708, + "grad_norm": 0.9406449794769287, + "learning_rate": 4.883542008727029e-06, + "loss": 0.6998, + "step": 1757 + }, + { + "epoch": 0.6172752808988764, + "grad_norm": 0.9234215021133423, + "learning_rate": 4.88340250473628e-06, + "loss": 0.7069, + "step": 1758 + }, + { + "epoch": 0.617626404494382, + "grad_norm": 0.8770806789398193, + "learning_rate": 4.883262919235205e-06, + "loss": 0.738, + "step": 1759 + }, + { + "epoch": 0.6179775280898876, + "grad_norm": 0.9186333417892456, + "learning_rate": 4.8831232522285786e-06, + "loss": 0.7019, + "step": 1760 + }, + { + "epoch": 0.6183286516853933, + "grad_norm": 0.9067515730857849, + "learning_rate": 4.8829835037211745e-06, + "loss": 0.6724, + "step": 1761 + }, + { + "epoch": 0.6186797752808989, + "grad_norm": 0.962090790271759, + "learning_rate": 4.8828436737177744e-06, + "loss": 0.6521, + "step": 1762 + }, + { + "epoch": 0.6190308988764045, + "grad_norm": 0.8877134323120117, + "learning_rate": 4.88270376222316e-06, + "loss": 0.7029, + "step": 1763 + }, + { + "epoch": 0.6193820224719101, + "grad_norm": 0.9146241545677185, + "learning_rate": 4.882563769242116e-06, + "loss": 0.6999, + "step": 1764 + }, + { + "epoch": 0.6197331460674157, + "grad_norm": 0.8948286175727844, + "learning_rate": 4.882423694779429e-06, + "loss": 0.6677, + "step": 1765 + }, + { + "epoch": 0.6200842696629213, + "grad_norm": 0.958892285823822, + "learning_rate": 4.88228353883989e-06, + "loss": 0.7261, + "step": 1766 + }, + { + "epoch": 0.620435393258427, + "grad_norm": 0.906204104423523, + "learning_rate": 4.882143301428293e-06, + "loss": 0.671, + "step": 1767 + }, + { + "epoch": 0.6207865168539326, + "grad_norm": 0.9508899450302124, + "learning_rate": 4.882002982549435e-06, + "loss": 0.6718, + "step": 1768 + }, + { + "epoch": 0.6211376404494382, + "grad_norm": 0.8806086778640747, + "learning_rate": 4.8818625822081116e-06, + "loss": 0.6835, + "step": 1769 + }, + { + "epoch": 0.6214887640449438, + "grad_norm": 0.9053204655647278, + "learning_rate": 4.8817221004091266e-06, + "loss": 0.7018, + "step": 1770 + }, + { + "epoch": 0.6218398876404494, + "grad_norm": 0.8768627047538757, + "learning_rate": 4.881581537157284e-06, + "loss": 0.6819, + "step": 1771 + }, + { + "epoch": 0.6221910112359551, + "grad_norm": 0.8701360821723938, + "learning_rate": 4.881440892457391e-06, + "loss": 0.6766, + "step": 1772 + }, + { + "epoch": 0.6225421348314607, + "grad_norm": 0.9027068614959717, + "learning_rate": 4.881300166314257e-06, + "loss": 0.6819, + "step": 1773 + }, + { + "epoch": 0.6228932584269663, + "grad_norm": 0.9427441954612732, + "learning_rate": 4.881159358732695e-06, + "loss": 0.6904, + "step": 1774 + }, + { + "epoch": 0.6232443820224719, + "grad_norm": 0.9292178750038147, + "learning_rate": 4.88101846971752e-06, + "loss": 0.7045, + "step": 1775 + }, + { + "epoch": 0.6235955056179775, + "grad_norm": 0.864090085029602, + "learning_rate": 4.880877499273551e-06, + "loss": 0.6496, + "step": 1776 + }, + { + "epoch": 0.6239466292134831, + "grad_norm": 0.9168904423713684, + "learning_rate": 4.880736447405609e-06, + "loss": 0.6805, + "step": 1777 + }, + { + "epoch": 0.6242977528089888, + "grad_norm": 0.9072659015655518, + "learning_rate": 4.880595314118518e-06, + "loss": 0.6699, + "step": 1778 + }, + { + "epoch": 0.6246488764044944, + "grad_norm": 0.9015219807624817, + "learning_rate": 4.880454099417104e-06, + "loss": 0.6781, + "step": 1779 + }, + { + "epoch": 0.625, + "grad_norm": 0.9149530529975891, + "learning_rate": 4.880312803306197e-06, + "loss": 0.7069, + "step": 1780 + }, + { + "epoch": 0.6253511235955056, + "grad_norm": 0.9412938952445984, + "learning_rate": 4.880171425790629e-06, + "loss": 0.6851, + "step": 1781 + }, + { + "epoch": 0.6257022471910112, + "grad_norm": 0.9359396696090698, + "learning_rate": 4.880029966875235e-06, + "loss": 0.6738, + "step": 1782 + }, + { + "epoch": 0.6260533707865169, + "grad_norm": 0.9232779741287231, + "learning_rate": 4.879888426564853e-06, + "loss": 0.6688, + "step": 1783 + }, + { + "epoch": 0.6264044943820225, + "grad_norm": 0.9085204005241394, + "learning_rate": 4.879746804864323e-06, + "loss": 0.6738, + "step": 1784 + }, + { + "epoch": 0.6267556179775281, + "grad_norm": 0.916195809841156, + "learning_rate": 4.879605101778488e-06, + "loss": 0.6691, + "step": 1785 + }, + { + "epoch": 0.6271067415730337, + "grad_norm": 0.891173243522644, + "learning_rate": 4.8794633173121954e-06, + "loss": 0.6941, + "step": 1786 + }, + { + "epoch": 0.6274578651685393, + "grad_norm": 0.8935718536376953, + "learning_rate": 4.879321451470294e-06, + "loss": 0.65, + "step": 1787 + }, + { + "epoch": 0.6278089887640449, + "grad_norm": 0.9326649308204651, + "learning_rate": 4.879179504257634e-06, + "loss": 0.7116, + "step": 1788 + }, + { + "epoch": 0.6281601123595506, + "grad_norm": 0.9544984102249146, + "learning_rate": 4.879037475679071e-06, + "loss": 0.6531, + "step": 1789 + }, + { + "epoch": 0.6285112359550562, + "grad_norm": 0.8751811981201172, + "learning_rate": 4.878895365739463e-06, + "loss": 0.7009, + "step": 1790 + }, + { + "epoch": 0.6288623595505618, + "grad_norm": 0.860333263874054, + "learning_rate": 4.878753174443669e-06, + "loss": 0.7003, + "step": 1791 + }, + { + "epoch": 0.6292134831460674, + "grad_norm": 0.9002732038497925, + "learning_rate": 4.878610901796551e-06, + "loss": 0.6768, + "step": 1792 + }, + { + "epoch": 0.629564606741573, + "grad_norm": 0.8901023864746094, + "learning_rate": 4.8784685478029765e-06, + "loss": 0.6572, + "step": 1793 + }, + { + "epoch": 0.6299157303370787, + "grad_norm": 0.8898550271987915, + "learning_rate": 4.878326112467813e-06, + "loss": 0.7141, + "step": 1794 + }, + { + "epoch": 0.6302668539325843, + "grad_norm": 0.9085826277732849, + "learning_rate": 4.878183595795931e-06, + "loss": 0.7033, + "step": 1795 + }, + { + "epoch": 0.6306179775280899, + "grad_norm": 0.9423730969429016, + "learning_rate": 4.878040997792205e-06, + "loss": 0.6763, + "step": 1796 + }, + { + "epoch": 0.6309691011235955, + "grad_norm": 0.9046456217765808, + "learning_rate": 4.877898318461512e-06, + "loss": 0.6438, + "step": 1797 + }, + { + "epoch": 0.6313202247191011, + "grad_norm": 0.9388726353645325, + "learning_rate": 4.877755557808731e-06, + "loss": 0.6921, + "step": 1798 + }, + { + "epoch": 0.6316713483146067, + "grad_norm": 0.8939900398254395, + "learning_rate": 4.877612715838746e-06, + "loss": 0.6729, + "step": 1799 + }, + { + "epoch": 0.6320224719101124, + "grad_norm": 0.919826328754425, + "learning_rate": 4.877469792556439e-06, + "loss": 0.7064, + "step": 1800 + }, + { + "epoch": 0.632373595505618, + "grad_norm": 0.8846567273139954, + "learning_rate": 4.8773267879667e-06, + "loss": 0.6463, + "step": 1801 + }, + { + "epoch": 0.6327247191011236, + "grad_norm": 0.9120215177536011, + "learning_rate": 4.8771837020744195e-06, + "loss": 0.7055, + "step": 1802 + }, + { + "epoch": 0.6330758426966292, + "grad_norm": 0.9156368970870972, + "learning_rate": 4.87704053488449e-06, + "loss": 0.7379, + "step": 1803 + }, + { + "epoch": 0.6334269662921348, + "grad_norm": 0.9091489315032959, + "learning_rate": 4.876897286401808e-06, + "loss": 0.6618, + "step": 1804 + }, + { + "epoch": 0.6337780898876404, + "grad_norm": 0.8682206273078918, + "learning_rate": 4.876753956631274e-06, + "loss": 0.6773, + "step": 1805 + }, + { + "epoch": 0.6341292134831461, + "grad_norm": 0.9142380952835083, + "learning_rate": 4.876610545577788e-06, + "loss": 0.7102, + "step": 1806 + }, + { + "epoch": 0.6344803370786517, + "grad_norm": 0.938529908657074, + "learning_rate": 4.876467053246255e-06, + "loss": 0.6967, + "step": 1807 + }, + { + "epoch": 0.6348314606741573, + "grad_norm": 0.9012477397918701, + "learning_rate": 4.876323479641582e-06, + "loss": 0.6548, + "step": 1808 + }, + { + "epoch": 0.6351825842696629, + "grad_norm": 0.8724120855331421, + "learning_rate": 4.8761798247686795e-06, + "loss": 0.6205, + "step": 1809 + }, + { + "epoch": 0.6355337078651685, + "grad_norm": 0.9118915796279907, + "learning_rate": 4.876036088632461e-06, + "loss": 0.6931, + "step": 1810 + }, + { + "epoch": 0.6358848314606742, + "grad_norm": 0.8679929375648499, + "learning_rate": 4.875892271237841e-06, + "loss": 0.6808, + "step": 1811 + }, + { + "epoch": 0.6362359550561798, + "grad_norm": 0.8951993584632874, + "learning_rate": 4.875748372589738e-06, + "loss": 0.6665, + "step": 1812 + }, + { + "epoch": 0.6365870786516854, + "grad_norm": 0.9559075832366943, + "learning_rate": 4.875604392693074e-06, + "loss": 0.6634, + "step": 1813 + }, + { + "epoch": 0.636938202247191, + "grad_norm": 0.8831387758255005, + "learning_rate": 4.875460331552772e-06, + "loss": 0.7054, + "step": 1814 + }, + { + "epoch": 0.6372893258426966, + "grad_norm": 0.8540201783180237, + "learning_rate": 4.875316189173761e-06, + "loss": 0.6676, + "step": 1815 + }, + { + "epoch": 0.6376404494382022, + "grad_norm": 0.8935166001319885, + "learning_rate": 4.875171965560967e-06, + "loss": 0.6789, + "step": 1816 + }, + { + "epoch": 0.6379915730337079, + "grad_norm": 0.9121273159980774, + "learning_rate": 4.875027660719326e-06, + "loss": 0.6645, + "step": 1817 + }, + { + "epoch": 0.6383426966292135, + "grad_norm": 0.9133923053741455, + "learning_rate": 4.87488327465377e-06, + "loss": 0.7053, + "step": 1818 + }, + { + "epoch": 0.6386938202247191, + "grad_norm": 0.9285811185836792, + "learning_rate": 4.874738807369239e-06, + "loss": 0.7172, + "step": 1819 + }, + { + "epoch": 0.6390449438202247, + "grad_norm": 0.9026995301246643, + "learning_rate": 4.874594258870673e-06, + "loss": 0.6848, + "step": 1820 + }, + { + "epoch": 0.6393960674157303, + "grad_norm": 0.9294858574867249, + "learning_rate": 4.874449629163015e-06, + "loss": 0.7073, + "step": 1821 + }, + { + "epoch": 0.639747191011236, + "grad_norm": 0.9029496908187866, + "learning_rate": 4.874304918251212e-06, + "loss": 0.7049, + "step": 1822 + }, + { + "epoch": 0.6400983146067416, + "grad_norm": 0.8920079469680786, + "learning_rate": 4.874160126140211e-06, + "loss": 0.6649, + "step": 1823 + }, + { + "epoch": 0.6404494382022472, + "grad_norm": 0.8667442798614502, + "learning_rate": 4.874015252834966e-06, + "loss": 0.6327, + "step": 1824 + }, + { + "epoch": 0.6408005617977528, + "grad_norm": 0.8910001516342163, + "learning_rate": 4.8738702983404315e-06, + "loss": 0.6964, + "step": 1825 + }, + { + "epoch": 0.6411516853932584, + "grad_norm": 0.8773640394210815, + "learning_rate": 4.873725262661564e-06, + "loss": 0.6763, + "step": 1826 + }, + { + "epoch": 0.641502808988764, + "grad_norm": 0.9394330978393555, + "learning_rate": 4.873580145803323e-06, + "loss": 0.7796, + "step": 1827 + }, + { + "epoch": 0.6418539325842697, + "grad_norm": 0.8596976399421692, + "learning_rate": 4.8734349477706725e-06, + "loss": 0.6229, + "step": 1828 + }, + { + "epoch": 0.6422050561797753, + "grad_norm": 0.9220896363258362, + "learning_rate": 4.873289668568578e-06, + "loss": 0.6801, + "step": 1829 + }, + { + "epoch": 0.6425561797752809, + "grad_norm": 0.8582680821418762, + "learning_rate": 4.873144308202008e-06, + "loss": 0.6475, + "step": 1830 + }, + { + "epoch": 0.6429073033707865, + "grad_norm": 0.916015088558197, + "learning_rate": 4.872998866675933e-06, + "loss": 0.6296, + "step": 1831 + }, + { + "epoch": 0.6432584269662921, + "grad_norm": 0.8732319474220276, + "learning_rate": 4.872853343995327e-06, + "loss": 0.6294, + "step": 1832 + }, + { + "epoch": 0.6436095505617978, + "grad_norm": 0.8985139727592468, + "learning_rate": 4.872707740165168e-06, + "loss": 0.6647, + "step": 1833 + }, + { + "epoch": 0.6439606741573034, + "grad_norm": 0.9100089073181152, + "learning_rate": 4.872562055190433e-06, + "loss": 0.66, + "step": 1834 + }, + { + "epoch": 0.644311797752809, + "grad_norm": 0.8807663917541504, + "learning_rate": 4.872416289076107e-06, + "loss": 0.714, + "step": 1835 + }, + { + "epoch": 0.6446629213483146, + "grad_norm": 0.924670934677124, + "learning_rate": 4.872270441827174e-06, + "loss": 0.6851, + "step": 1836 + }, + { + "epoch": 0.6450140449438202, + "grad_norm": 0.9182237982749939, + "learning_rate": 4.872124513448622e-06, + "loss": 0.6728, + "step": 1837 + }, + { + "epoch": 0.6453651685393258, + "grad_norm": 0.8587349057197571, + "learning_rate": 4.871978503945442e-06, + "loss": 0.6632, + "step": 1838 + }, + { + "epoch": 0.6457162921348315, + "grad_norm": 0.8839679956436157, + "learning_rate": 4.871832413322626e-06, + "loss": 0.6531, + "step": 1839 + }, + { + "epoch": 0.6460674157303371, + "grad_norm": 0.9729541540145874, + "learning_rate": 4.871686241585171e-06, + "loss": 0.6855, + "step": 1840 + }, + { + "epoch": 0.6464185393258427, + "grad_norm": 0.895285964012146, + "learning_rate": 4.871539988738076e-06, + "loss": 0.6738, + "step": 1841 + }, + { + "epoch": 0.6467696629213483, + "grad_norm": 0.8678845167160034, + "learning_rate": 4.871393654786344e-06, + "loss": 0.6494, + "step": 1842 + }, + { + "epoch": 0.6471207865168539, + "grad_norm": 0.8896480798721313, + "learning_rate": 4.871247239734977e-06, + "loss": 0.6823, + "step": 1843 + }, + { + "epoch": 0.6474719101123596, + "grad_norm": 0.8339593410491943, + "learning_rate": 4.871100743588983e-06, + "loss": 0.6193, + "step": 1844 + }, + { + "epoch": 0.6478230337078652, + "grad_norm": 0.8970949053764343, + "learning_rate": 4.870954166353374e-06, + "loss": 0.6944, + "step": 1845 + }, + { + "epoch": 0.6481741573033708, + "grad_norm": 0.8810001015663147, + "learning_rate": 4.870807508033161e-06, + "loss": 0.7222, + "step": 1846 + }, + { + "epoch": 0.6485252808988764, + "grad_norm": 0.8821990489959717, + "learning_rate": 4.870660768633359e-06, + "loss": 0.6881, + "step": 1847 + }, + { + "epoch": 0.648876404494382, + "grad_norm": 0.9234027862548828, + "learning_rate": 4.8705139481589885e-06, + "loss": 0.7074, + "step": 1848 + }, + { + "epoch": 0.6492275280898876, + "grad_norm": 0.9003174901008606, + "learning_rate": 4.870367046615069e-06, + "loss": 0.6754, + "step": 1849 + }, + { + "epoch": 0.6495786516853933, + "grad_norm": 0.9838910698890686, + "learning_rate": 4.870220064006625e-06, + "loss": 0.716, + "step": 1850 + }, + { + "epoch": 0.6499297752808989, + "grad_norm": 0.9417749047279358, + "learning_rate": 4.870073000338683e-06, + "loss": 0.6826, + "step": 1851 + }, + { + "epoch": 0.6502808988764045, + "grad_norm": 0.9156447052955627, + "learning_rate": 4.869925855616272e-06, + "loss": 0.6472, + "step": 1852 + }, + { + "epoch": 0.6506320224719101, + "grad_norm": 0.8980277180671692, + "learning_rate": 4.869778629844425e-06, + "loss": 0.6536, + "step": 1853 + }, + { + "epoch": 0.6509831460674157, + "grad_norm": 0.9398206472396851, + "learning_rate": 4.8696313230281776e-06, + "loss": 0.6618, + "step": 1854 + }, + { + "epoch": 0.6513342696629213, + "grad_norm": 0.9095361232757568, + "learning_rate": 4.869483935172566e-06, + "loss": 0.6758, + "step": 1855 + }, + { + "epoch": 0.651685393258427, + "grad_norm": 0.9065535664558411, + "learning_rate": 4.86933646628263e-06, + "loss": 0.6785, + "step": 1856 + }, + { + "epoch": 0.6520365168539326, + "grad_norm": 0.8876499533653259, + "learning_rate": 4.869188916363417e-06, + "loss": 0.642, + "step": 1857 + }, + { + "epoch": 0.6523876404494382, + "grad_norm": 0.8867465853691101, + "learning_rate": 4.869041285419968e-06, + "loss": 0.6926, + "step": 1858 + }, + { + "epoch": 0.6527387640449438, + "grad_norm": 0.9016279578208923, + "learning_rate": 4.868893573457335e-06, + "loss": 0.7007, + "step": 1859 + }, + { + "epoch": 0.6530898876404494, + "grad_norm": 0.9355380535125732, + "learning_rate": 4.868745780480569e-06, + "loss": 0.7025, + "step": 1860 + }, + { + "epoch": 0.6534410112359551, + "grad_norm": 0.8738915324211121, + "learning_rate": 4.868597906494725e-06, + "loss": 0.6633, + "step": 1861 + }, + { + "epoch": 0.6537921348314607, + "grad_norm": 0.9167248010635376, + "learning_rate": 4.868449951504859e-06, + "loss": 0.6669, + "step": 1862 + }, + { + "epoch": 0.6541432584269663, + "grad_norm": 0.9220420122146606, + "learning_rate": 4.868301915516031e-06, + "loss": 0.6419, + "step": 1863 + }, + { + "epoch": 0.6544943820224719, + "grad_norm": 0.9482148885726929, + "learning_rate": 4.868153798533304e-06, + "loss": 0.6934, + "step": 1864 + }, + { + "epoch": 0.6548455056179775, + "grad_norm": 0.928871214389801, + "learning_rate": 4.8680056005617436e-06, + "loss": 0.6506, + "step": 1865 + }, + { + "epoch": 0.6551966292134831, + "grad_norm": 0.9240955114364624, + "learning_rate": 4.867857321606418e-06, + "loss": 0.6745, + "step": 1866 + }, + { + "epoch": 0.6555477528089888, + "grad_norm": 0.8898888230323792, + "learning_rate": 4.867708961672399e-06, + "loss": 0.6925, + "step": 1867 + }, + { + "epoch": 0.6558988764044944, + "grad_norm": 0.9390647411346436, + "learning_rate": 4.8675605207647595e-06, + "loss": 0.7212, + "step": 1868 + }, + { + "epoch": 0.65625, + "grad_norm": 0.9547630548477173, + "learning_rate": 4.8674119988885756e-06, + "loss": 0.7168, + "step": 1869 + }, + { + "epoch": 0.6566011235955056, + "grad_norm": 0.8770663738250732, + "learning_rate": 4.867263396048927e-06, + "loss": 0.6373, + "step": 1870 + }, + { + "epoch": 0.6569522471910112, + "grad_norm": 0.899808406829834, + "learning_rate": 4.867114712250897e-06, + "loss": 0.708, + "step": 1871 + }, + { + "epoch": 0.6573033707865169, + "grad_norm": 0.9099377989768982, + "learning_rate": 4.866965947499568e-06, + "loss": 0.6888, + "step": 1872 + }, + { + "epoch": 0.6576544943820225, + "grad_norm": 0.896460235118866, + "learning_rate": 4.866817101800029e-06, + "loss": 0.6995, + "step": 1873 + }, + { + "epoch": 0.6580056179775281, + "grad_norm": 0.8833721280097961, + "learning_rate": 4.866668175157372e-06, + "loss": 0.659, + "step": 1874 + }, + { + "epoch": 0.6583567415730337, + "grad_norm": 0.830697774887085, + "learning_rate": 4.866519167576687e-06, + "loss": 0.6501, + "step": 1875 + }, + { + "epoch": 0.6587078651685393, + "grad_norm": 0.8897443413734436, + "learning_rate": 4.866370079063073e-06, + "loss": 0.6769, + "step": 1876 + }, + { + "epoch": 0.6590589887640449, + "grad_norm": 0.8772610425949097, + "learning_rate": 4.866220909621625e-06, + "loss": 0.6474, + "step": 1877 + }, + { + "epoch": 0.6594101123595506, + "grad_norm": 0.9042930006980896, + "learning_rate": 4.8660716592574486e-06, + "loss": 0.6827, + "step": 1878 + }, + { + "epoch": 0.6597612359550562, + "grad_norm": 0.9102811217308044, + "learning_rate": 4.865922327975646e-06, + "loss": 0.7126, + "step": 1879 + }, + { + "epoch": 0.6601123595505618, + "grad_norm": 0.9031997323036194, + "learning_rate": 4.865772915781324e-06, + "loss": 0.6963, + "step": 1880 + }, + { + "epoch": 0.6604634831460674, + "grad_norm": 0.89446622133255, + "learning_rate": 4.865623422679593e-06, + "loss": 0.6718, + "step": 1881 + }, + { + "epoch": 0.660814606741573, + "grad_norm": 0.879215657711029, + "learning_rate": 4.865473848675565e-06, + "loss": 0.6711, + "step": 1882 + }, + { + "epoch": 0.6611657303370787, + "grad_norm": 0.916566789150238, + "learning_rate": 4.865324193774355e-06, + "loss": 0.66, + "step": 1883 + }, + { + "epoch": 0.6615168539325843, + "grad_norm": 0.8872631192207336, + "learning_rate": 4.865174457981082e-06, + "loss": 0.6994, + "step": 1884 + }, + { + "epoch": 0.6618679775280899, + "grad_norm": 0.9287483096122742, + "learning_rate": 4.865024641300866e-06, + "loss": 0.7071, + "step": 1885 + }, + { + "epoch": 0.6622191011235955, + "grad_norm": 0.8984839916229248, + "learning_rate": 4.864874743738832e-06, + "loss": 0.6512, + "step": 1886 + }, + { + "epoch": 0.6625702247191011, + "grad_norm": 0.8624014258384705, + "learning_rate": 4.864724765300105e-06, + "loss": 0.6649, + "step": 1887 + }, + { + "epoch": 0.6629213483146067, + "grad_norm": 0.8853806257247925, + "learning_rate": 4.864574705989814e-06, + "loss": 0.6562, + "step": 1888 + }, + { + "epoch": 0.6632724719101124, + "grad_norm": 0.9202688336372375, + "learning_rate": 4.864424565813092e-06, + "loss": 0.7005, + "step": 1889 + }, + { + "epoch": 0.663623595505618, + "grad_norm": 0.9426562786102295, + "learning_rate": 4.8642743447750735e-06, + "loss": 0.6802, + "step": 1890 + }, + { + "epoch": 0.6639747191011236, + "grad_norm": 0.9027799367904663, + "learning_rate": 4.864124042880895e-06, + "loss": 0.6718, + "step": 1891 + }, + { + "epoch": 0.6643258426966292, + "grad_norm": 0.9138725399971008, + "learning_rate": 4.863973660135697e-06, + "loss": 0.7225, + "step": 1892 + }, + { + "epoch": 0.6646769662921348, + "grad_norm": 0.9003073573112488, + "learning_rate": 4.863823196544623e-06, + "loss": 0.6603, + "step": 1893 + }, + { + "epoch": 0.6650280898876404, + "grad_norm": 0.9345358610153198, + "learning_rate": 4.863672652112819e-06, + "loss": 0.6648, + "step": 1894 + }, + { + "epoch": 0.6653792134831461, + "grad_norm": 0.9087908864021301, + "learning_rate": 4.863522026845432e-06, + "loss": 0.701, + "step": 1895 + }, + { + "epoch": 0.6657303370786517, + "grad_norm": 0.889224112033844, + "learning_rate": 4.863371320747615e-06, + "loss": 0.7157, + "step": 1896 + }, + { + "epoch": 0.6660814606741573, + "grad_norm": 0.8854907751083374, + "learning_rate": 4.863220533824521e-06, + "loss": 0.6965, + "step": 1897 + }, + { + "epoch": 0.6664325842696629, + "grad_norm": 0.9223704934120178, + "learning_rate": 4.863069666081307e-06, + "loss": 0.6962, + "step": 1898 + }, + { + "epoch": 0.6667837078651685, + "grad_norm": 0.9045940637588501, + "learning_rate": 4.8629187175231314e-06, + "loss": 0.6394, + "step": 1899 + }, + { + "epoch": 0.6671348314606742, + "grad_norm": 0.8898254036903381, + "learning_rate": 4.862767688155159e-06, + "loss": 0.6781, + "step": 1900 + }, + { + "epoch": 0.6674859550561798, + "grad_norm": 0.882418692111969, + "learning_rate": 4.862616577982553e-06, + "loss": 0.6654, + "step": 1901 + }, + { + "epoch": 0.6678370786516854, + "grad_norm": 0.8898456692695618, + "learning_rate": 4.862465387010482e-06, + "loss": 0.6733, + "step": 1902 + }, + { + "epoch": 0.668188202247191, + "grad_norm": 0.9240460395812988, + "learning_rate": 4.862314115244116e-06, + "loss": 0.6661, + "step": 1903 + }, + { + "epoch": 0.6685393258426966, + "grad_norm": 0.8646409511566162, + "learning_rate": 4.862162762688629e-06, + "loss": 0.6751, + "step": 1904 + }, + { + "epoch": 0.6688904494382022, + "grad_norm": 0.9161674976348877, + "learning_rate": 4.862011329349196e-06, + "loss": 0.6762, + "step": 1905 + }, + { + "epoch": 0.6692415730337079, + "grad_norm": 0.8989123702049255, + "learning_rate": 4.861859815230998e-06, + "loss": 0.6948, + "step": 1906 + }, + { + "epoch": 0.6695926966292135, + "grad_norm": 0.9007300138473511, + "learning_rate": 4.861708220339215e-06, + "loss": 0.6954, + "step": 1907 + }, + { + "epoch": 0.6699438202247191, + "grad_norm": 0.9009419679641724, + "learning_rate": 4.861556544679032e-06, + "loss": 0.6968, + "step": 1908 + }, + { + "epoch": 0.6702949438202247, + "grad_norm": 0.8684524297714233, + "learning_rate": 4.861404788255636e-06, + "loss": 0.6548, + "step": 1909 + }, + { + "epoch": 0.6706460674157303, + "grad_norm": 0.8866965174674988, + "learning_rate": 4.861252951074216e-06, + "loss": 0.6867, + "step": 1910 + }, + { + "epoch": 0.670997191011236, + "grad_norm": 0.8514602184295654, + "learning_rate": 4.861101033139967e-06, + "loss": 0.6677, + "step": 1911 + }, + { + "epoch": 0.6713483146067416, + "grad_norm": 0.9115993976593018, + "learning_rate": 4.8609490344580815e-06, + "loss": 0.7163, + "step": 1912 + }, + { + "epoch": 0.6716994382022472, + "grad_norm": 0.9031485319137573, + "learning_rate": 4.86079695503376e-06, + "loss": 0.6557, + "step": 1913 + }, + { + "epoch": 0.6720505617977528, + "grad_norm": 0.9129601716995239, + "learning_rate": 4.8606447948722036e-06, + "loss": 0.6857, + "step": 1914 + }, + { + "epoch": 0.6724016853932584, + "grad_norm": 0.8930435180664062, + "learning_rate": 4.860492553978614e-06, + "loss": 0.6888, + "step": 1915 + }, + { + "epoch": 0.672752808988764, + "grad_norm": 0.8881438374519348, + "learning_rate": 4.860340232358199e-06, + "loss": 0.6979, + "step": 1916 + }, + { + "epoch": 0.6731039325842697, + "grad_norm": 0.8837933540344238, + "learning_rate": 4.860187830016169e-06, + "loss": 0.6729, + "step": 1917 + }, + { + "epoch": 0.6734550561797753, + "grad_norm": 0.8822264671325684, + "learning_rate": 4.860035346957734e-06, + "loss": 0.6883, + "step": 1918 + }, + { + "epoch": 0.6738061797752809, + "grad_norm": 0.8592543005943298, + "learning_rate": 4.8598827831881105e-06, + "loss": 0.6579, + "step": 1919 + }, + { + "epoch": 0.6741573033707865, + "grad_norm": 0.8520956039428711, + "learning_rate": 4.859730138712514e-06, + "loss": 0.6547, + "step": 1920 + }, + { + "epoch": 0.6745084269662921, + "grad_norm": 0.8868786096572876, + "learning_rate": 4.859577413536167e-06, + "loss": 0.6834, + "step": 1921 + }, + { + "epoch": 0.6748595505617978, + "grad_norm": 0.894353449344635, + "learning_rate": 4.859424607664291e-06, + "loss": 0.6622, + "step": 1922 + }, + { + "epoch": 0.6752106741573034, + "grad_norm": 0.8527143597602844, + "learning_rate": 4.859271721102112e-06, + "loss": 0.6557, + "step": 1923 + }, + { + "epoch": 0.675561797752809, + "grad_norm": 0.840531587600708, + "learning_rate": 4.85911875385486e-06, + "loss": 0.6392, + "step": 1924 + }, + { + "epoch": 0.6759129213483146, + "grad_norm": 0.8790932297706604, + "learning_rate": 4.858965705927765e-06, + "loss": 0.7031, + "step": 1925 + }, + { + "epoch": 0.6762640449438202, + "grad_norm": 0.8507084250450134, + "learning_rate": 4.85881257732606e-06, + "loss": 0.6459, + "step": 1926 + }, + { + "epoch": 0.6766151685393258, + "grad_norm": 0.894951581954956, + "learning_rate": 4.858659368054986e-06, + "loss": 0.6885, + "step": 1927 + }, + { + "epoch": 0.6769662921348315, + "grad_norm": 0.9011364579200745, + "learning_rate": 4.858506078119778e-06, + "loss": 0.7135, + "step": 1928 + }, + { + "epoch": 0.6773174157303371, + "grad_norm": 0.8751960396766663, + "learning_rate": 4.858352707525681e-06, + "loss": 0.6964, + "step": 1929 + }, + { + "epoch": 0.6776685393258427, + "grad_norm": 0.8583717942237854, + "learning_rate": 4.85819925627794e-06, + "loss": 0.6495, + "step": 1930 + }, + { + "epoch": 0.6780196629213483, + "grad_norm": 0.9073567390441895, + "learning_rate": 4.858045724381801e-06, + "loss": 0.6693, + "step": 1931 + }, + { + "epoch": 0.6783707865168539, + "grad_norm": 0.875085711479187, + "learning_rate": 4.857892111842517e-06, + "loss": 0.6552, + "step": 1932 + }, + { + "epoch": 0.6787219101123596, + "grad_norm": 0.890121340751648, + "learning_rate": 4.8577384186653405e-06, + "loss": 0.6816, + "step": 1933 + }, + { + "epoch": 0.6790730337078652, + "grad_norm": 0.9079762697219849, + "learning_rate": 4.8575846448555275e-06, + "loss": 0.7012, + "step": 1934 + }, + { + "epoch": 0.6794241573033708, + "grad_norm": 0.9067320823669434, + "learning_rate": 4.857430790418337e-06, + "loss": 0.6707, + "step": 1935 + }, + { + "epoch": 0.6797752808988764, + "grad_norm": 0.863487184047699, + "learning_rate": 4.857276855359031e-06, + "loss": 0.6532, + "step": 1936 + }, + { + "epoch": 0.680126404494382, + "grad_norm": 0.887187659740448, + "learning_rate": 4.857122839682873e-06, + "loss": 0.68, + "step": 1937 + }, + { + "epoch": 0.6804775280898876, + "grad_norm": 0.9105770587921143, + "learning_rate": 4.856968743395132e-06, + "loss": 0.7263, + "step": 1938 + }, + { + "epoch": 0.6808286516853933, + "grad_norm": 0.859266459941864, + "learning_rate": 4.856814566501075e-06, + "loss": 0.6612, + "step": 1939 + }, + { + "epoch": 0.6811797752808989, + "grad_norm": 0.8715479969978333, + "learning_rate": 4.856660309005979e-06, + "loss": 0.7078, + "step": 1940 + }, + { + "epoch": 0.6815308988764045, + "grad_norm": 0.8754439949989319, + "learning_rate": 4.856505970915115e-06, + "loss": 0.7088, + "step": 1941 + }, + { + "epoch": 0.6818820224719101, + "grad_norm": 0.8877387642860413, + "learning_rate": 4.856351552233764e-06, + "loss": 0.6944, + "step": 1942 + }, + { + "epoch": 0.6822331460674157, + "grad_norm": 0.9069793224334717, + "learning_rate": 4.856197052967207e-06, + "loss": 0.6567, + "step": 1943 + }, + { + "epoch": 0.6825842696629213, + "grad_norm": 0.9112107157707214, + "learning_rate": 4.856042473120726e-06, + "loss": 0.6617, + "step": 1944 + }, + { + "epoch": 0.682935393258427, + "grad_norm": 0.9331689476966858, + "learning_rate": 4.855887812699609e-06, + "loss": 0.7051, + "step": 1945 + }, + { + "epoch": 0.6832865168539326, + "grad_norm": 0.909070611000061, + "learning_rate": 4.855733071709145e-06, + "loss": 0.6827, + "step": 1946 + }, + { + "epoch": 0.6836376404494382, + "grad_norm": 0.9068970084190369, + "learning_rate": 4.855578250154624e-06, + "loss": 0.658, + "step": 1947 + }, + { + "epoch": 0.6839887640449438, + "grad_norm": 0.9052385687828064, + "learning_rate": 4.855423348041345e-06, + "loss": 0.6991, + "step": 1948 + }, + { + "epoch": 0.6843398876404494, + "grad_norm": 0.8857588768005371, + "learning_rate": 4.855268365374601e-06, + "loss": 0.6604, + "step": 1949 + }, + { + "epoch": 0.6846910112359551, + "grad_norm": 0.8484088778495789, + "learning_rate": 4.855113302159697e-06, + "loss": 0.6591, + "step": 1950 + }, + { + "epoch": 0.6850421348314607, + "grad_norm": 0.9859545826911926, + "learning_rate": 4.854958158401931e-06, + "loss": 0.6652, + "step": 1951 + }, + { + "epoch": 0.6853932584269663, + "grad_norm": 0.8704702854156494, + "learning_rate": 4.854802934106611e-06, + "loss": 0.6792, + "step": 1952 + }, + { + "epoch": 0.6857443820224719, + "grad_norm": 0.8603319525718689, + "learning_rate": 4.854647629279047e-06, + "loss": 0.6768, + "step": 1953 + }, + { + "epoch": 0.6860955056179775, + "grad_norm": 0.9137224555015564, + "learning_rate": 4.854492243924549e-06, + "loss": 0.6865, + "step": 1954 + }, + { + "epoch": 0.6864466292134831, + "grad_norm": 0.8924287557601929, + "learning_rate": 4.8543367780484304e-06, + "loss": 0.685, + "step": 1955 + }, + { + "epoch": 0.6867977528089888, + "grad_norm": 0.8736245632171631, + "learning_rate": 4.854181231656009e-06, + "loss": 0.6747, + "step": 1956 + }, + { + "epoch": 0.6871488764044944, + "grad_norm": 0.8849700093269348, + "learning_rate": 4.854025604752604e-06, + "loss": 0.679, + "step": 1957 + }, + { + "epoch": 0.6875, + "grad_norm": 0.9219119548797607, + "learning_rate": 4.853869897343537e-06, + "loss": 0.7069, + "step": 1958 + }, + { + "epoch": 0.6878511235955056, + "grad_norm": 0.8864405751228333, + "learning_rate": 4.853714109434134e-06, + "loss": 0.648, + "step": 1959 + }, + { + "epoch": 0.6882022471910112, + "grad_norm": 0.9157871603965759, + "learning_rate": 4.853558241029723e-06, + "loss": 0.6838, + "step": 1960 + }, + { + "epoch": 0.6885533707865169, + "grad_norm": 0.8845330476760864, + "learning_rate": 4.853402292135634e-06, + "loss": 0.67, + "step": 1961 + }, + { + "epoch": 0.6889044943820225, + "grad_norm": 0.8602401614189148, + "learning_rate": 4.8532462627572e-06, + "loss": 0.6661, + "step": 1962 + }, + { + "epoch": 0.6892556179775281, + "grad_norm": 0.9007517695426941, + "learning_rate": 4.853090152899758e-06, + "loss": 0.6717, + "step": 1963 + }, + { + "epoch": 0.6896067415730337, + "grad_norm": 0.8941755890846252, + "learning_rate": 4.852933962568647e-06, + "loss": 0.6824, + "step": 1964 + }, + { + "epoch": 0.6899578651685393, + "grad_norm": 0.9077978134155273, + "learning_rate": 4.852777691769207e-06, + "loss": 0.695, + "step": 1965 + }, + { + "epoch": 0.6903089887640449, + "grad_norm": 0.9053329229354858, + "learning_rate": 4.852621340506782e-06, + "loss": 0.6781, + "step": 1966 + }, + { + "epoch": 0.6906601123595506, + "grad_norm": 0.885201096534729, + "learning_rate": 4.852464908786722e-06, + "loss": 0.6307, + "step": 1967 + }, + { + "epoch": 0.6910112359550562, + "grad_norm": 0.9176447987556458, + "learning_rate": 4.8523083966143755e-06, + "loss": 0.6797, + "step": 1968 + }, + { + "epoch": 0.6913623595505618, + "grad_norm": 0.8907782435417175, + "learning_rate": 4.852151803995094e-06, + "loss": 0.661, + "step": 1969 + }, + { + "epoch": 0.6917134831460674, + "grad_norm": 0.9267505407333374, + "learning_rate": 4.8519951309342335e-06, + "loss": 0.6711, + "step": 1970 + }, + { + "epoch": 0.692064606741573, + "grad_norm": 0.9260626435279846, + "learning_rate": 4.8518383774371525e-06, + "loss": 0.6667, + "step": 1971 + }, + { + "epoch": 0.6924157303370787, + "grad_norm": 0.947289228439331, + "learning_rate": 4.851681543509211e-06, + "loss": 0.6802, + "step": 1972 + }, + { + "epoch": 0.6927668539325843, + "grad_norm": 0.9016532897949219, + "learning_rate": 4.851524629155773e-06, + "loss": 0.6788, + "step": 1973 + }, + { + "epoch": 0.6931179775280899, + "grad_norm": 0.926960825920105, + "learning_rate": 4.851367634382206e-06, + "loss": 0.6868, + "step": 1974 + }, + { + "epoch": 0.6934691011235955, + "grad_norm": 0.886313259601593, + "learning_rate": 4.851210559193877e-06, + "loss": 0.664, + "step": 1975 + }, + { + "epoch": 0.6938202247191011, + "grad_norm": 0.8867488503456116, + "learning_rate": 4.851053403596158e-06, + "loss": 0.6903, + "step": 1976 + }, + { + "epoch": 0.6941713483146067, + "grad_norm": 0.8759608864784241, + "learning_rate": 4.850896167594426e-06, + "loss": 0.6542, + "step": 1977 + }, + { + "epoch": 0.6945224719101124, + "grad_norm": 0.8738287091255188, + "learning_rate": 4.8507388511940565e-06, + "loss": 0.6844, + "step": 1978 + }, + { + "epoch": 0.694873595505618, + "grad_norm": 0.875167191028595, + "learning_rate": 4.85058145440043e-06, + "loss": 0.6515, + "step": 1979 + }, + { + "epoch": 0.6952247191011236, + "grad_norm": 0.8686906099319458, + "learning_rate": 4.850423977218929e-06, + "loss": 0.6558, + "step": 1980 + }, + { + "epoch": 0.6955758426966292, + "grad_norm": 0.8697443604469299, + "learning_rate": 4.850266419654939e-06, + "loss": 0.6431, + "step": 1981 + }, + { + "epoch": 0.6959269662921348, + "grad_norm": 0.9007163047790527, + "learning_rate": 4.850108781713848e-06, + "loss": 0.6996, + "step": 1982 + }, + { + "epoch": 0.6962780898876404, + "grad_norm": 0.9064307808876038, + "learning_rate": 4.849951063401049e-06, + "loss": 0.6625, + "step": 1983 + }, + { + "epoch": 0.6966292134831461, + "grad_norm": 0.9589636921882629, + "learning_rate": 4.849793264721934e-06, + "loss": 0.7276, + "step": 1984 + }, + { + "epoch": 0.6969803370786517, + "grad_norm": 0.9150680899620056, + "learning_rate": 4.8496353856819e-06, + "loss": 0.6808, + "step": 1985 + }, + { + "epoch": 0.6973314606741573, + "grad_norm": 0.9379653334617615, + "learning_rate": 4.849477426286346e-06, + "loss": 0.7148, + "step": 1986 + }, + { + "epoch": 0.6976825842696629, + "grad_norm": 0.8566982746124268, + "learning_rate": 4.849319386540675e-06, + "loss": 0.676, + "step": 1987 + }, + { + "epoch": 0.6980337078651685, + "grad_norm": 0.9016757607460022, + "learning_rate": 4.84916126645029e-06, + "loss": 0.6672, + "step": 1988 + }, + { + "epoch": 0.6983848314606742, + "grad_norm": 0.9286409020423889, + "learning_rate": 4.849003066020602e-06, + "loss": 0.7028, + "step": 1989 + }, + { + "epoch": 0.6987359550561798, + "grad_norm": 0.8700134754180908, + "learning_rate": 4.848844785257019e-06, + "loss": 0.6919, + "step": 1990 + }, + { + "epoch": 0.6990870786516854, + "grad_norm": 0.8518657088279724, + "learning_rate": 4.848686424164953e-06, + "loss": 0.663, + "step": 1991 + }, + { + "epoch": 0.699438202247191, + "grad_norm": 0.8883944749832153, + "learning_rate": 4.848527982749822e-06, + "loss": 0.6249, + "step": 1992 + }, + { + "epoch": 0.6997893258426966, + "grad_norm": 0.9609950184822083, + "learning_rate": 4.848369461017043e-06, + "loss": 0.7114, + "step": 1993 + }, + { + "epoch": 0.7001404494382022, + "grad_norm": 0.8920437097549438, + "learning_rate": 4.848210858972039e-06, + "loss": 0.7093, + "step": 1994 + }, + { + "epoch": 0.7004915730337079, + "grad_norm": 0.9256541132926941, + "learning_rate": 4.848052176620232e-06, + "loss": 0.6713, + "step": 1995 + }, + { + "epoch": 0.7008426966292135, + "grad_norm": 0.8984493613243103, + "learning_rate": 4.847893413967049e-06, + "loss": 0.6921, + "step": 1996 + }, + { + "epoch": 0.7011938202247191, + "grad_norm": 0.889815628528595, + "learning_rate": 4.847734571017923e-06, + "loss": 0.6539, + "step": 1997 + }, + { + "epoch": 0.7015449438202247, + "grad_norm": 0.856476366519928, + "learning_rate": 4.847575647778281e-06, + "loss": 0.6538, + "step": 1998 + }, + { + "epoch": 0.7018960674157303, + "grad_norm": 0.9042733311653137, + "learning_rate": 4.847416644253562e-06, + "loss": 0.672, + "step": 1999 + }, + { + "epoch": 0.702247191011236, + "grad_norm": 0.847193717956543, + "learning_rate": 4.847257560449202e-06, + "loss": 0.6296, + "step": 2000 + }, + { + "epoch": 0.7025983146067416, + "grad_norm": 0.922281801700592, + "learning_rate": 4.847098396370643e-06, + "loss": 0.6306, + "step": 2001 + }, + { + "epoch": 0.7029494382022472, + "grad_norm": 0.9229936599731445, + "learning_rate": 4.846939152023326e-06, + "loss": 0.7127, + "step": 2002 + }, + { + "epoch": 0.7033005617977528, + "grad_norm": 0.8789034485816956, + "learning_rate": 4.8467798274126995e-06, + "loss": 0.6477, + "step": 2003 + }, + { + "epoch": 0.7036516853932584, + "grad_norm": 0.87799471616745, + "learning_rate": 4.84662042254421e-06, + "loss": 0.646, + "step": 2004 + }, + { + "epoch": 0.704002808988764, + "grad_norm": 0.8978516459465027, + "learning_rate": 4.846460937423311e-06, + "loss": 0.6327, + "step": 2005 + }, + { + "epoch": 0.7043539325842697, + "grad_norm": 0.9151999354362488, + "learning_rate": 4.846301372055455e-06, + "loss": 0.6726, + "step": 2006 + }, + { + "epoch": 0.7047050561797753, + "grad_norm": 0.9256817698478699, + "learning_rate": 4.846141726446101e-06, + "loss": 0.6842, + "step": 2007 + }, + { + "epoch": 0.7050561797752809, + "grad_norm": 0.89638352394104, + "learning_rate": 4.845982000600706e-06, + "loss": 0.6792, + "step": 2008 + }, + { + "epoch": 0.7054073033707865, + "grad_norm": 0.8957139849662781, + "learning_rate": 4.845822194524735e-06, + "loss": 0.7032, + "step": 2009 + }, + { + "epoch": 0.7057584269662921, + "grad_norm": 0.9059911370277405, + "learning_rate": 4.845662308223652e-06, + "loss": 0.7272, + "step": 2010 + }, + { + "epoch": 0.7061095505617978, + "grad_norm": 0.8813321590423584, + "learning_rate": 4.845502341702926e-06, + "loss": 0.6918, + "step": 2011 + }, + { + "epoch": 0.7064606741573034, + "grad_norm": 0.9412849545478821, + "learning_rate": 4.845342294968026e-06, + "loss": 0.7301, + "step": 2012 + }, + { + "epoch": 0.706811797752809, + "grad_norm": 0.9104598164558411, + "learning_rate": 4.845182168024427e-06, + "loss": 0.6655, + "step": 2013 + }, + { + "epoch": 0.7071629213483146, + "grad_norm": 0.8984184861183167, + "learning_rate": 4.845021960877605e-06, + "loss": 0.6816, + "step": 2014 + }, + { + "epoch": 0.7075140449438202, + "grad_norm": 0.9031736850738525, + "learning_rate": 4.844861673533038e-06, + "loss": 0.6941, + "step": 2015 + }, + { + "epoch": 0.7078651685393258, + "grad_norm": 0.9205414652824402, + "learning_rate": 4.8447013059962085e-06, + "loss": 0.6976, + "step": 2016 + }, + { + "epoch": 0.7082162921348315, + "grad_norm": 0.9202782511711121, + "learning_rate": 4.8445408582726e-06, + "loss": 0.6859, + "step": 2017 + }, + { + "epoch": 0.7085674157303371, + "grad_norm": 0.9281151294708252, + "learning_rate": 4.844380330367701e-06, + "loss": 0.6827, + "step": 2018 + }, + { + "epoch": 0.7089185393258427, + "grad_norm": 0.8505613207817078, + "learning_rate": 4.844219722287001e-06, + "loss": 0.6198, + "step": 2019 + }, + { + "epoch": 0.7092696629213483, + "grad_norm": 0.8852872252464294, + "learning_rate": 4.844059034035993e-06, + "loss": 0.7026, + "step": 2020 + }, + { + "epoch": 0.7096207865168539, + "grad_norm": 0.8684709072113037, + "learning_rate": 4.843898265620171e-06, + "loss": 0.6602, + "step": 2021 + }, + { + "epoch": 0.7099719101123596, + "grad_norm": 0.9983372688293457, + "learning_rate": 4.843737417045035e-06, + "loss": 0.6646, + "step": 2022 + }, + { + "epoch": 0.7103230337078652, + "grad_norm": 0.8986789584159851, + "learning_rate": 4.843576488316084e-06, + "loss": 0.7246, + "step": 2023 + }, + { + "epoch": 0.7106741573033708, + "grad_norm": 0.898418664932251, + "learning_rate": 4.843415479438822e-06, + "loss": 0.7011, + "step": 2024 + }, + { + "epoch": 0.7110252808988764, + "grad_norm": 0.9410746097564697, + "learning_rate": 4.843254390418757e-06, + "loss": 0.6753, + "step": 2025 + }, + { + "epoch": 0.711376404494382, + "grad_norm": 0.9196873307228088, + "learning_rate": 4.843093221261396e-06, + "loss": 0.7028, + "step": 2026 + }, + { + "epoch": 0.7117275280898876, + "grad_norm": 0.8739240169525146, + "learning_rate": 4.842931971972252e-06, + "loss": 0.6477, + "step": 2027 + }, + { + "epoch": 0.7120786516853933, + "grad_norm": 0.863125205039978, + "learning_rate": 4.842770642556841e-06, + "loss": 0.6855, + "step": 2028 + }, + { + "epoch": 0.7124297752808989, + "grad_norm": 0.8746199607849121, + "learning_rate": 4.842609233020677e-06, + "loss": 0.6368, + "step": 2029 + }, + { + "epoch": 0.7127808988764045, + "grad_norm": 0.9002125859260559, + "learning_rate": 4.842447743369283e-06, + "loss": 0.666, + "step": 2030 + }, + { + "epoch": 0.7131320224719101, + "grad_norm": 0.8976905345916748, + "learning_rate": 4.84228617360818e-06, + "loss": 0.6807, + "step": 2031 + }, + { + "epoch": 0.7134831460674157, + "grad_norm": 0.9237125515937805, + "learning_rate": 4.8421245237428935e-06, + "loss": 0.6749, + "step": 2032 + }, + { + "epoch": 0.7138342696629213, + "grad_norm": 0.8990357518196106, + "learning_rate": 4.841962793778954e-06, + "loss": 0.7136, + "step": 2033 + }, + { + "epoch": 0.714185393258427, + "grad_norm": 0.9369171857833862, + "learning_rate": 4.84180098372189e-06, + "loss": 0.7157, + "step": 2034 + }, + { + "epoch": 0.7145365168539326, + "grad_norm": 0.9113399386405945, + "learning_rate": 4.841639093577236e-06, + "loss": 0.6733, + "step": 2035 + }, + { + "epoch": 0.7148876404494382, + "grad_norm": 0.9038696885108948, + "learning_rate": 4.84147712335053e-06, + "loss": 0.7065, + "step": 2036 + }, + { + "epoch": 0.7152387640449438, + "grad_norm": 0.8818309903144836, + "learning_rate": 4.841315073047309e-06, + "loss": 0.6633, + "step": 2037 + }, + { + "epoch": 0.7155898876404494, + "grad_norm": 0.8951486349105835, + "learning_rate": 4.841152942673116e-06, + "loss": 0.6965, + "step": 2038 + }, + { + "epoch": 0.7159410112359551, + "grad_norm": 0.8958115577697754, + "learning_rate": 4.840990732233496e-06, + "loss": 0.6536, + "step": 2039 + }, + { + "epoch": 0.7162921348314607, + "grad_norm": 0.8991239070892334, + "learning_rate": 4.840828441733996e-06, + "loss": 0.7002, + "step": 2040 + }, + { + "epoch": 0.7166432584269663, + "grad_norm": 0.8847957849502563, + "learning_rate": 4.840666071180167e-06, + "loss": 0.7011, + "step": 2041 + }, + { + "epoch": 0.7169943820224719, + "grad_norm": 0.8697375655174255, + "learning_rate": 4.840503620577561e-06, + "loss": 0.6869, + "step": 2042 + }, + { + "epoch": 0.7173455056179775, + "grad_norm": 0.8789458274841309, + "learning_rate": 4.840341089931734e-06, + "loss": 0.6624, + "step": 2043 + }, + { + "epoch": 0.7176966292134831, + "grad_norm": 0.868335485458374, + "learning_rate": 4.8401784792482436e-06, + "loss": 0.6624, + "step": 2044 + }, + { + "epoch": 0.7180477528089888, + "grad_norm": 0.8949484825134277, + "learning_rate": 4.840015788532653e-06, + "loss": 0.6621, + "step": 2045 + }, + { + "epoch": 0.7183988764044944, + "grad_norm": 0.8744524717330933, + "learning_rate": 4.839853017790523e-06, + "loss": 0.6055, + "step": 2046 + }, + { + "epoch": 0.71875, + "grad_norm": 0.8852532505989075, + "learning_rate": 4.8396901670274236e-06, + "loss": 0.7299, + "step": 2047 + }, + { + "epoch": 0.7191011235955056, + "grad_norm": 0.8787648677825928, + "learning_rate": 4.839527236248922e-06, + "loss": 0.6687, + "step": 2048 + }, + { + "epoch": 0.7194522471910112, + "grad_norm": 0.8882533311843872, + "learning_rate": 4.839364225460591e-06, + "loss": 0.6488, + "step": 2049 + }, + { + "epoch": 0.7198033707865169, + "grad_norm": 0.8676881194114685, + "learning_rate": 4.839201134668006e-06, + "loss": 0.7084, + "step": 2050 + }, + { + "epoch": 0.7201544943820225, + "grad_norm": 0.9025247097015381, + "learning_rate": 4.839037963876743e-06, + "loss": 0.6805, + "step": 2051 + }, + { + "epoch": 0.7205056179775281, + "grad_norm": 0.8956056833267212, + "learning_rate": 4.838874713092383e-06, + "loss": 0.6905, + "step": 2052 + }, + { + "epoch": 0.7208567415730337, + "grad_norm": 0.8418490290641785, + "learning_rate": 4.8387113823205104e-06, + "loss": 0.6224, + "step": 2053 + }, + { + "epoch": 0.7212078651685393, + "grad_norm": 0.897898256778717, + "learning_rate": 4.838547971566708e-06, + "loss": 0.6505, + "step": 2054 + }, + { + "epoch": 0.7215589887640449, + "grad_norm": 0.9430592060089111, + "learning_rate": 4.838384480836568e-06, + "loss": 0.7031, + "step": 2055 + }, + { + "epoch": 0.7219101123595506, + "grad_norm": 0.9099498987197876, + "learning_rate": 4.838220910135678e-06, + "loss": 0.7372, + "step": 2056 + }, + { + "epoch": 0.7222612359550562, + "grad_norm": 0.8936906456947327, + "learning_rate": 4.838057259469635e-06, + "loss": 0.6782, + "step": 2057 + }, + { + "epoch": 0.7226123595505618, + "grad_norm": 0.8931390643119812, + "learning_rate": 4.8378935288440346e-06, + "loss": 0.6552, + "step": 2058 + }, + { + "epoch": 0.7229634831460674, + "grad_norm": 0.893640398979187, + "learning_rate": 4.837729718264475e-06, + "loss": 0.6971, + "step": 2059 + }, + { + "epoch": 0.723314606741573, + "grad_norm": 0.9098536968231201, + "learning_rate": 4.83756582773656e-06, + "loss": 0.6816, + "step": 2060 + }, + { + "epoch": 0.7236657303370787, + "grad_norm": 0.8403578400611877, + "learning_rate": 4.837401857265894e-06, + "loss": 0.6233, + "step": 2061 + }, + { + "epoch": 0.7240168539325843, + "grad_norm": 0.894605815410614, + "learning_rate": 4.8372378068580835e-06, + "loss": 0.6747, + "step": 2062 + }, + { + "epoch": 0.7243679775280899, + "grad_norm": 0.8893696665763855, + "learning_rate": 4.837073676518742e-06, + "loss": 0.6769, + "step": 2063 + }, + { + "epoch": 0.7247191011235955, + "grad_norm": 0.9210422039031982, + "learning_rate": 4.836909466253479e-06, + "loss": 0.7177, + "step": 2064 + }, + { + "epoch": 0.7250702247191011, + "grad_norm": 0.9131370782852173, + "learning_rate": 4.836745176067913e-06, + "loss": 0.6899, + "step": 2065 + }, + { + "epoch": 0.7254213483146067, + "grad_norm": 0.8986336588859558, + "learning_rate": 4.836580805967661e-06, + "loss": 0.68, + "step": 2066 + }, + { + "epoch": 0.7257724719101124, + "grad_norm": 0.8927890658378601, + "learning_rate": 4.836416355958345e-06, + "loss": 0.6971, + "step": 2067 + }, + { + "epoch": 0.726123595505618, + "grad_norm": 0.9166001677513123, + "learning_rate": 4.836251826045589e-06, + "loss": 0.7008, + "step": 2068 + }, + { + "epoch": 0.7264747191011236, + "grad_norm": 0.8931081891059875, + "learning_rate": 4.836087216235021e-06, + "loss": 0.6816, + "step": 2069 + }, + { + "epoch": 0.7268258426966292, + "grad_norm": 0.8896619081497192, + "learning_rate": 4.835922526532267e-06, + "loss": 0.7011, + "step": 2070 + }, + { + "epoch": 0.7271769662921348, + "grad_norm": 0.8900657296180725, + "learning_rate": 4.835757756942963e-06, + "loss": 0.6754, + "step": 2071 + }, + { + "epoch": 0.7275280898876404, + "grad_norm": 0.860923707485199, + "learning_rate": 4.8355929074727415e-06, + "loss": 0.6407, + "step": 2072 + }, + { + "epoch": 0.7278792134831461, + "grad_norm": 0.8783923983573914, + "learning_rate": 4.835427978127241e-06, + "loss": 0.7052, + "step": 2073 + }, + { + "epoch": 0.7282303370786517, + "grad_norm": 0.9070163369178772, + "learning_rate": 4.835262968912103e-06, + "loss": 0.7003, + "step": 2074 + }, + { + "epoch": 0.7285814606741573, + "grad_norm": 1.025547742843628, + "learning_rate": 4.83509787983297e-06, + "loss": 0.6829, + "step": 2075 + }, + { + "epoch": 0.7289325842696629, + "grad_norm": 0.9013466835021973, + "learning_rate": 4.834932710895486e-06, + "loss": 0.6475, + "step": 2076 + }, + { + "epoch": 0.7292837078651685, + "grad_norm": 0.894381582736969, + "learning_rate": 4.834767462105303e-06, + "loss": 0.6414, + "step": 2077 + }, + { + "epoch": 0.7296348314606742, + "grad_norm": 0.8694182634353638, + "learning_rate": 4.83460213346807e-06, + "loss": 0.6442, + "step": 2078 + }, + { + "epoch": 0.7299859550561798, + "grad_norm": 0.8485740423202515, + "learning_rate": 4.834436724989442e-06, + "loss": 0.6105, + "step": 2079 + }, + { + "epoch": 0.7303370786516854, + "grad_norm": 0.8988609910011292, + "learning_rate": 4.834271236675076e-06, + "loss": 0.6615, + "step": 2080 + }, + { + "epoch": 0.730688202247191, + "grad_norm": 0.9217643141746521, + "learning_rate": 4.8341056685306295e-06, + "loss": 0.7194, + "step": 2081 + }, + { + "epoch": 0.7310393258426966, + "grad_norm": 0.8651581406593323, + "learning_rate": 4.833940020561768e-06, + "loss": 0.6831, + "step": 2082 + }, + { + "epoch": 0.7313904494382022, + "grad_norm": 0.9073237776756287, + "learning_rate": 4.833774292774154e-06, + "loss": 0.7034, + "step": 2083 + }, + { + "epoch": 0.7317415730337079, + "grad_norm": 0.9331315159797668, + "learning_rate": 4.833608485173458e-06, + "loss": 0.7328, + "step": 2084 + }, + { + "epoch": 0.7320926966292135, + "grad_norm": 0.9014981985092163, + "learning_rate": 4.833442597765346e-06, + "loss": 0.6776, + "step": 2085 + }, + { + "epoch": 0.7324438202247191, + "grad_norm": 0.8719879388809204, + "learning_rate": 4.833276630555496e-06, + "loss": 0.6672, + "step": 2086 + }, + { + "epoch": 0.7327949438202247, + "grad_norm": 0.8750829100608826, + "learning_rate": 4.8331105835495816e-06, + "loss": 0.6344, + "step": 2087 + }, + { + "epoch": 0.7331460674157303, + "grad_norm": 0.8622833490371704, + "learning_rate": 4.832944456753281e-06, + "loss": 0.6453, + "step": 2088 + }, + { + "epoch": 0.733497191011236, + "grad_norm": 0.8901452422142029, + "learning_rate": 4.8327782501722766e-06, + "loss": 0.6809, + "step": 2089 + }, + { + "epoch": 0.7338483146067416, + "grad_norm": 0.9039089679718018, + "learning_rate": 4.8326119638122525e-06, + "loss": 0.7185, + "step": 2090 + }, + { + "epoch": 0.7341994382022472, + "grad_norm": 0.9443939328193665, + "learning_rate": 4.832445597678895e-06, + "loss": 0.6647, + "step": 2091 + }, + { + "epoch": 0.7345505617977528, + "grad_norm": 0.9074206352233887, + "learning_rate": 4.8322791517778945e-06, + "loss": 0.6696, + "step": 2092 + }, + { + "epoch": 0.7349016853932584, + "grad_norm": 0.9349908232688904, + "learning_rate": 4.832112626114942e-06, + "loss": 0.6735, + "step": 2093 + }, + { + "epoch": 0.735252808988764, + "grad_norm": 0.8699564933776855, + "learning_rate": 4.831946020695733e-06, + "loss": 0.7143, + "step": 2094 + }, + { + "epoch": 0.7356039325842697, + "grad_norm": 0.8945705890655518, + "learning_rate": 4.831779335525967e-06, + "loss": 0.6748, + "step": 2095 + }, + { + "epoch": 0.7359550561797753, + "grad_norm": 0.8274276256561279, + "learning_rate": 4.831612570611343e-06, + "loss": 0.647, + "step": 2096 + }, + { + "epoch": 0.7363061797752809, + "grad_norm": 0.8812181949615479, + "learning_rate": 4.831445725957563e-06, + "loss": 0.6644, + "step": 2097 + }, + { + "epoch": 0.7366573033707865, + "grad_norm": 0.8417778015136719, + "learning_rate": 4.831278801570335e-06, + "loss": 0.613, + "step": 2098 + }, + { + "epoch": 0.7370084269662921, + "grad_norm": 0.862417459487915, + "learning_rate": 4.831111797455366e-06, + "loss": 0.6592, + "step": 2099 + }, + { + "epoch": 0.7373595505617978, + "grad_norm": 0.8994525074958801, + "learning_rate": 4.830944713618369e-06, + "loss": 0.7038, + "step": 2100 + }, + { + "epoch": 0.7377106741573034, + "grad_norm": 0.8670651316642761, + "learning_rate": 4.830777550065058e-06, + "loss": 0.6344, + "step": 2101 + }, + { + "epoch": 0.738061797752809, + "grad_norm": 0.9172879457473755, + "learning_rate": 4.830610306801148e-06, + "loss": 0.657, + "step": 2102 + }, + { + "epoch": 0.7384129213483146, + "grad_norm": 0.8903508186340332, + "learning_rate": 4.830442983832361e-06, + "loss": 0.6419, + "step": 2103 + }, + { + "epoch": 0.7387640449438202, + "grad_norm": 0.913711428642273, + "learning_rate": 4.830275581164417e-06, + "loss": 0.6551, + "step": 2104 + }, + { + "epoch": 0.7391151685393258, + "grad_norm": 0.8964325189590454, + "learning_rate": 4.830108098803043e-06, + "loss": 0.6839, + "step": 2105 + }, + { + "epoch": 0.7394662921348315, + "grad_norm": 0.8451694846153259, + "learning_rate": 4.829940536753965e-06, + "loss": 0.6846, + "step": 2106 + }, + { + "epoch": 0.7398174157303371, + "grad_norm": 0.8537585735321045, + "learning_rate": 4.829772895022916e-06, + "loss": 0.6632, + "step": 2107 + }, + { + "epoch": 0.7401685393258427, + "grad_norm": 0.882422924041748, + "learning_rate": 4.829605173615627e-06, + "loss": 0.6827, + "step": 2108 + }, + { + "epoch": 0.7405196629213483, + "grad_norm": 0.9608381986618042, + "learning_rate": 4.829437372537834e-06, + "loss": 0.6698, + "step": 2109 + }, + { + "epoch": 0.7408707865168539, + "grad_norm": 0.8602509498596191, + "learning_rate": 4.829269491795277e-06, + "loss": 0.6842, + "step": 2110 + }, + { + "epoch": 0.7412219101123596, + "grad_norm": 0.9101216197013855, + "learning_rate": 4.829101531393695e-06, + "loss": 0.6909, + "step": 2111 + }, + { + "epoch": 0.7415730337078652, + "grad_norm": 0.9232078194618225, + "learning_rate": 4.8289334913388344e-06, + "loss": 0.6544, + "step": 2112 + }, + { + "epoch": 0.7419241573033708, + "grad_norm": 0.9027913212776184, + "learning_rate": 4.828765371636442e-06, + "loss": 0.7055, + "step": 2113 + }, + { + "epoch": 0.7422752808988764, + "grad_norm": 0.8847595453262329, + "learning_rate": 4.828597172292266e-06, + "loss": 0.6585, + "step": 2114 + }, + { + "epoch": 0.742626404494382, + "grad_norm": 0.9057716131210327, + "learning_rate": 4.8284288933120595e-06, + "loss": 0.6815, + "step": 2115 + }, + { + "epoch": 0.7429775280898876, + "grad_norm": 0.833390474319458, + "learning_rate": 4.828260534701577e-06, + "loss": 0.6113, + "step": 2116 + }, + { + "epoch": 0.7433286516853933, + "grad_norm": 0.9086408615112305, + "learning_rate": 4.828092096466578e-06, + "loss": 0.6537, + "step": 2117 + }, + { + "epoch": 0.7436797752808989, + "grad_norm": 0.9064264297485352, + "learning_rate": 4.82792357861282e-06, + "loss": 0.71, + "step": 2118 + }, + { + "epoch": 0.7440308988764045, + "grad_norm": 0.9102289080619812, + "learning_rate": 4.827754981146068e-06, + "loss": 0.6711, + "step": 2119 + }, + { + "epoch": 0.7443820224719101, + "grad_norm": 0.9276719093322754, + "learning_rate": 4.827586304072087e-06, + "loss": 0.6667, + "step": 2120 + }, + { + "epoch": 0.7447331460674157, + "grad_norm": 0.9276329278945923, + "learning_rate": 4.827417547396647e-06, + "loss": 0.7072, + "step": 2121 + }, + { + "epoch": 0.7450842696629213, + "grad_norm": 0.8715528249740601, + "learning_rate": 4.827248711125518e-06, + "loss": 0.6921, + "step": 2122 + }, + { + "epoch": 0.745435393258427, + "grad_norm": 0.8835285902023315, + "learning_rate": 4.827079795264475e-06, + "loss": 0.6557, + "step": 2123 + }, + { + "epoch": 0.7457865168539326, + "grad_norm": 0.9126173853874207, + "learning_rate": 4.826910799819293e-06, + "loss": 0.6965, + "step": 2124 + }, + { + "epoch": 0.7461376404494382, + "grad_norm": 0.8943221569061279, + "learning_rate": 4.8267417247957545e-06, + "loss": 0.6532, + "step": 2125 + }, + { + "epoch": 0.7464887640449438, + "grad_norm": 0.8731197714805603, + "learning_rate": 4.826572570199639e-06, + "loss": 0.6462, + "step": 2126 + }, + { + "epoch": 0.7468398876404494, + "grad_norm": 0.8800104856491089, + "learning_rate": 4.826403336036733e-06, + "loss": 0.6862, + "step": 2127 + }, + { + "epoch": 0.7471910112359551, + "grad_norm": 0.8647268414497375, + "learning_rate": 4.826234022312823e-06, + "loss": 0.6901, + "step": 2128 + }, + { + "epoch": 0.7475421348314607, + "grad_norm": 0.8699238896369934, + "learning_rate": 4.826064629033701e-06, + "loss": 0.6848, + "step": 2129 + }, + { + "epoch": 0.7478932584269663, + "grad_norm": 0.881797730922699, + "learning_rate": 4.825895156205158e-06, + "loss": 0.7022, + "step": 2130 + }, + { + "epoch": 0.7482443820224719, + "grad_norm": 0.8925601840019226, + "learning_rate": 4.825725603832992e-06, + "loss": 0.6668, + "step": 2131 + }, + { + "epoch": 0.7485955056179775, + "grad_norm": 0.874530553817749, + "learning_rate": 4.825555971923e-06, + "loss": 0.6914, + "step": 2132 + }, + { + "epoch": 0.7489466292134831, + "grad_norm": 0.8640453815460205, + "learning_rate": 4.8253862604809836e-06, + "loss": 0.6301, + "step": 2133 + }, + { + "epoch": 0.7492977528089888, + "grad_norm": 0.9012243151664734, + "learning_rate": 4.8252164695127475e-06, + "loss": 0.6876, + "step": 2134 + }, + { + "epoch": 0.7496488764044944, + "grad_norm": 0.8837961554527283, + "learning_rate": 4.825046599024098e-06, + "loss": 0.6898, + "step": 2135 + }, + { + "epoch": 0.75, + "grad_norm": 0.8629834651947021, + "learning_rate": 4.8248766490208434e-06, + "loss": 0.6709, + "step": 2136 + }, + { + "epoch": 0.7503511235955056, + "grad_norm": 0.8934418559074402, + "learning_rate": 4.824706619508797e-06, + "loss": 0.6406, + "step": 2137 + }, + { + "epoch": 0.7507022471910112, + "grad_norm": 0.8818619251251221, + "learning_rate": 4.8245365104937734e-06, + "loss": 0.6751, + "step": 2138 + }, + { + "epoch": 0.7510533707865169, + "grad_norm": 0.871660590171814, + "learning_rate": 4.824366321981591e-06, + "loss": 0.6628, + "step": 2139 + }, + { + "epoch": 0.7514044943820225, + "grad_norm": 0.8862018585205078, + "learning_rate": 4.824196053978069e-06, + "loss": 0.655, + "step": 2140 + }, + { + "epoch": 0.7517556179775281, + "grad_norm": 0.8964344263076782, + "learning_rate": 4.824025706489031e-06, + "loss": 0.6784, + "step": 2141 + }, + { + "epoch": 0.7521067415730337, + "grad_norm": 0.8510351181030273, + "learning_rate": 4.8238552795203025e-06, + "loss": 0.6521, + "step": 2142 + }, + { + "epoch": 0.7524578651685393, + "grad_norm": 0.8770909905433655, + "learning_rate": 4.823684773077712e-06, + "loss": 0.6828, + "step": 2143 + }, + { + "epoch": 0.7528089887640449, + "grad_norm": 0.9084598422050476, + "learning_rate": 4.823514187167091e-06, + "loss": 0.7081, + "step": 2144 + }, + { + "epoch": 0.7531601123595506, + "grad_norm": 0.8897272944450378, + "learning_rate": 4.823343521794273e-06, + "loss": 0.6812, + "step": 2145 + }, + { + "epoch": 0.7535112359550562, + "grad_norm": 0.8769894242286682, + "learning_rate": 4.8231727769650934e-06, + "loss": 0.69, + "step": 2146 + }, + { + "epoch": 0.7538623595505618, + "grad_norm": 0.8911151885986328, + "learning_rate": 4.823001952685394e-06, + "loss": 0.6573, + "step": 2147 + }, + { + "epoch": 0.7542134831460674, + "grad_norm": 0.8679448366165161, + "learning_rate": 4.822831048961015e-06, + "loss": 0.7047, + "step": 2148 + }, + { + "epoch": 0.754564606741573, + "grad_norm": 0.8955689072608948, + "learning_rate": 4.8226600657978024e-06, + "loss": 0.6775, + "step": 2149 + }, + { + "epoch": 0.7549157303370787, + "grad_norm": 0.8726634383201599, + "learning_rate": 4.822489003201602e-06, + "loss": 0.6744, + "step": 2150 + }, + { + "epoch": 0.7552668539325843, + "grad_norm": 0.8858023881912231, + "learning_rate": 4.822317861178266e-06, + "loss": 0.6975, + "step": 2151 + }, + { + "epoch": 0.7556179775280899, + "grad_norm": 0.8668063879013062, + "learning_rate": 4.822146639733647e-06, + "loss": 0.6866, + "step": 2152 + }, + { + "epoch": 0.7559691011235955, + "grad_norm": 0.8951655626296997, + "learning_rate": 4.821975338873599e-06, + "loss": 0.7165, + "step": 2153 + }, + { + "epoch": 0.7563202247191011, + "grad_norm": 0.9189617037773132, + "learning_rate": 4.821803958603982e-06, + "loss": 0.6899, + "step": 2154 + }, + { + "epoch": 0.7566713483146067, + "grad_norm": 0.8782833814620972, + "learning_rate": 4.821632498930656e-06, + "loss": 0.6613, + "step": 2155 + }, + { + "epoch": 0.7570224719101124, + "grad_norm": 0.882354736328125, + "learning_rate": 4.821460959859485e-06, + "loss": 0.7047, + "step": 2156 + }, + { + "epoch": 0.757373595505618, + "grad_norm": 0.8961043357849121, + "learning_rate": 4.821289341396336e-06, + "loss": 0.6699, + "step": 2157 + }, + { + "epoch": 0.7577247191011236, + "grad_norm": 0.9026042819023132, + "learning_rate": 4.8211176435470774e-06, + "loss": 0.6874, + "step": 2158 + }, + { + "epoch": 0.7580758426966292, + "grad_norm": 0.89781653881073, + "learning_rate": 4.820945866317582e-06, + "loss": 0.6764, + "step": 2159 + }, + { + "epoch": 0.7584269662921348, + "grad_norm": 0.9029053449630737, + "learning_rate": 4.820774009713725e-06, + "loss": 0.671, + "step": 2160 + }, + { + "epoch": 0.7587780898876404, + "grad_norm": 0.8790138959884644, + "learning_rate": 4.820602073741381e-06, + "loss": 0.6855, + "step": 2161 + }, + { + "epoch": 0.7591292134831461, + "grad_norm": 0.8737415671348572, + "learning_rate": 4.820430058406433e-06, + "loss": 0.6746, + "step": 2162 + }, + { + "epoch": 0.7594803370786517, + "grad_norm": 0.8937375545501709, + "learning_rate": 4.8202579637147615e-06, + "loss": 0.6742, + "step": 2163 + }, + { + "epoch": 0.7598314606741573, + "grad_norm": 0.8693381547927856, + "learning_rate": 4.820085789672254e-06, + "loss": 0.643, + "step": 2164 + }, + { + "epoch": 0.7601825842696629, + "grad_norm": 0.8404312133789062, + "learning_rate": 4.819913536284798e-06, + "loss": 0.6247, + "step": 2165 + }, + { + "epoch": 0.7605337078651685, + "grad_norm": 0.9133322238922119, + "learning_rate": 4.819741203558284e-06, + "loss": 0.6999, + "step": 2166 + }, + { + "epoch": 0.7608848314606742, + "grad_norm": 0.878828763961792, + "learning_rate": 4.8195687914986055e-06, + "loss": 0.642, + "step": 2167 + }, + { + "epoch": 0.7612359550561798, + "grad_norm": 0.8653201460838318, + "learning_rate": 4.81939630011166e-06, + "loss": 0.6668, + "step": 2168 + }, + { + "epoch": 0.7615870786516854, + "grad_norm": 0.8635656833648682, + "learning_rate": 4.819223729403345e-06, + "loss": 0.6749, + "step": 2169 + }, + { + "epoch": 0.761938202247191, + "grad_norm": 0.9288333654403687, + "learning_rate": 4.819051079379563e-06, + "loss": 0.6687, + "step": 2170 + }, + { + "epoch": 0.7622893258426966, + "grad_norm": 0.8941252827644348, + "learning_rate": 4.818878350046219e-06, + "loss": 0.6795, + "step": 2171 + }, + { + "epoch": 0.7626404494382022, + "grad_norm": 0.8719800114631653, + "learning_rate": 4.81870554140922e-06, + "loss": 0.6273, + "step": 2172 + }, + { + "epoch": 0.7629915730337079, + "grad_norm": 0.9125568866729736, + "learning_rate": 4.818532653474475e-06, + "loss": 0.6992, + "step": 2173 + }, + { + "epoch": 0.7633426966292135, + "grad_norm": 0.9224469661712646, + "learning_rate": 4.818359686247898e-06, + "loss": 0.6984, + "step": 2174 + }, + { + "epoch": 0.7636938202247191, + "grad_norm": 0.9007505178451538, + "learning_rate": 4.818186639735403e-06, + "loss": 0.6994, + "step": 2175 + }, + { + "epoch": 0.7640449438202247, + "grad_norm": 0.9869738817214966, + "learning_rate": 4.818013513942908e-06, + "loss": 0.7248, + "step": 2176 + }, + { + "epoch": 0.7643960674157303, + "grad_norm": 0.9010109305381775, + "learning_rate": 4.817840308876335e-06, + "loss": 0.6987, + "step": 2177 + }, + { + "epoch": 0.764747191011236, + "grad_norm": 0.8850472569465637, + "learning_rate": 4.8176670245416076e-06, + "loss": 0.6561, + "step": 2178 + }, + { + "epoch": 0.7650983146067416, + "grad_norm": 0.9040166139602661, + "learning_rate": 4.81749366094465e-06, + "loss": 0.6789, + "step": 2179 + }, + { + "epoch": 0.7654494382022472, + "grad_norm": 0.879569411277771, + "learning_rate": 4.817320218091393e-06, + "loss": 0.6728, + "step": 2180 + }, + { + "epoch": 0.7658005617977528, + "grad_norm": 0.8991327285766602, + "learning_rate": 4.817146695987768e-06, + "loss": 0.6919, + "step": 2181 + }, + { + "epoch": 0.7661516853932584, + "grad_norm": 0.8922879695892334, + "learning_rate": 4.816973094639708e-06, + "loss": 0.6696, + "step": 2182 + }, + { + "epoch": 0.766502808988764, + "grad_norm": 0.9196280837059021, + "learning_rate": 4.8167994140531504e-06, + "loss": 0.6928, + "step": 2183 + }, + { + "epoch": 0.7668539325842697, + "grad_norm": 0.8855655193328857, + "learning_rate": 4.816625654234036e-06, + "loss": 0.67, + "step": 2184 + }, + { + "epoch": 0.7672050561797753, + "grad_norm": 0.9374614953994751, + "learning_rate": 4.816451815188306e-06, + "loss": 0.6514, + "step": 2185 + }, + { + "epoch": 0.7675561797752809, + "grad_norm": 0.8807061910629272, + "learning_rate": 4.816277896921907e-06, + "loss": 0.666, + "step": 2186 + }, + { + "epoch": 0.7679073033707865, + "grad_norm": 0.8804175853729248, + "learning_rate": 4.816103899440785e-06, + "loss": 0.6687, + "step": 2187 + }, + { + "epoch": 0.7682584269662921, + "grad_norm": 0.8722162842750549, + "learning_rate": 4.815929822750893e-06, + "loss": 0.6924, + "step": 2188 + }, + { + "epoch": 0.7686095505617978, + "grad_norm": 0.8696702718734741, + "learning_rate": 4.815755666858182e-06, + "loss": 0.6883, + "step": 2189 + }, + { + "epoch": 0.7689606741573034, + "grad_norm": 0.8709911704063416, + "learning_rate": 4.815581431768609e-06, + "loss": 0.6533, + "step": 2190 + }, + { + "epoch": 0.769311797752809, + "grad_norm": 0.9215819835662842, + "learning_rate": 4.815407117488132e-06, + "loss": 0.6519, + "step": 2191 + }, + { + "epoch": 0.7696629213483146, + "grad_norm": 0.9619512557983398, + "learning_rate": 4.815232724022713e-06, + "loss": 0.6523, + "step": 2192 + }, + { + "epoch": 0.7700140449438202, + "grad_norm": 0.8896926045417786, + "learning_rate": 4.815058251378316e-06, + "loss": 0.6772, + "step": 2193 + }, + { + "epoch": 0.7703651685393258, + "grad_norm": 0.9150664806365967, + "learning_rate": 4.814883699560908e-06, + "loss": 0.7005, + "step": 2194 + }, + { + "epoch": 0.7707162921348315, + "grad_norm": 0.8808259963989258, + "learning_rate": 4.8147090685764585e-06, + "loss": 0.683, + "step": 2195 + }, + { + "epoch": 0.7710674157303371, + "grad_norm": 0.8727423548698425, + "learning_rate": 4.8145343584309404e-06, + "loss": 0.6499, + "step": 2196 + }, + { + "epoch": 0.7714185393258427, + "grad_norm": 0.8704618811607361, + "learning_rate": 4.814359569130326e-06, + "loss": 0.6678, + "step": 2197 + }, + { + "epoch": 0.7717696629213483, + "grad_norm": 0.8853801488876343, + "learning_rate": 4.814184700680596e-06, + "loss": 0.6788, + "step": 2198 + }, + { + "epoch": 0.7721207865168539, + "grad_norm": 0.8947190046310425, + "learning_rate": 4.81400975308773e-06, + "loss": 0.6919, + "step": 2199 + }, + { + "epoch": 0.7724719101123596, + "grad_norm": 0.8671098351478577, + "learning_rate": 4.813834726357709e-06, + "loss": 0.6344, + "step": 2200 + }, + { + "epoch": 0.7728230337078652, + "grad_norm": 0.8669817447662354, + "learning_rate": 4.813659620496521e-06, + "loss": 0.6655, + "step": 2201 + }, + { + "epoch": 0.7731741573033708, + "grad_norm": 0.8825884461402893, + "learning_rate": 4.813484435510154e-06, + "loss": 0.687, + "step": 2202 + }, + { + "epoch": 0.7735252808988764, + "grad_norm": 0.9186983704566956, + "learning_rate": 4.8133091714045985e-06, + "loss": 0.6807, + "step": 2203 + }, + { + "epoch": 0.773876404494382, + "grad_norm": 0.8848013877868652, + "learning_rate": 4.81313382818585e-06, + "loss": 0.6974, + "step": 2204 + }, + { + "epoch": 0.7742275280898876, + "grad_norm": 0.8753053545951843, + "learning_rate": 4.812958405859903e-06, + "loss": 0.7025, + "step": 2205 + }, + { + "epoch": 0.7745786516853933, + "grad_norm": 0.8476365208625793, + "learning_rate": 4.812782904432758e-06, + "loss": 0.6466, + "step": 2206 + }, + { + "epoch": 0.7749297752808989, + "grad_norm": 0.8795397281646729, + "learning_rate": 4.812607323910416e-06, + "loss": 0.6669, + "step": 2207 + }, + { + "epoch": 0.7752808988764045, + "grad_norm": 0.8774227499961853, + "learning_rate": 4.812431664298884e-06, + "loss": 0.6488, + "step": 2208 + }, + { + "epoch": 0.7756320224719101, + "grad_norm": 0.888466477394104, + "learning_rate": 4.812255925604167e-06, + "loss": 0.6367, + "step": 2209 + }, + { + "epoch": 0.7759831460674157, + "grad_norm": 0.8345751166343689, + "learning_rate": 4.812080107832275e-06, + "loss": 0.6204, + "step": 2210 + }, + { + "epoch": 0.7763342696629213, + "grad_norm": 0.8765496611595154, + "learning_rate": 4.811904210989223e-06, + "loss": 0.697, + "step": 2211 + }, + { + "epoch": 0.776685393258427, + "grad_norm": 0.893101692199707, + "learning_rate": 4.8117282350810255e-06, + "loss": 0.6673, + "step": 2212 + }, + { + "epoch": 0.7770365168539326, + "grad_norm": 0.909160852432251, + "learning_rate": 4.8115521801137e-06, + "loss": 0.6973, + "step": 2213 + }, + { + "epoch": 0.7773876404494382, + "grad_norm": 0.8834896087646484, + "learning_rate": 4.811376046093268e-06, + "loss": 0.6878, + "step": 2214 + }, + { + "epoch": 0.7777387640449438, + "grad_norm": 0.8608538508415222, + "learning_rate": 4.8111998330257535e-06, + "loss": 0.6688, + "step": 2215 + }, + { + "epoch": 0.7780898876404494, + "grad_norm": 0.9701086282730103, + "learning_rate": 4.811023540917181e-06, + "loss": 0.6721, + "step": 2216 + }, + { + "epoch": 0.7784410112359551, + "grad_norm": 0.8717615008354187, + "learning_rate": 4.8108471697735826e-06, + "loss": 0.6834, + "step": 2217 + }, + { + "epoch": 0.7787921348314607, + "grad_norm": 0.9188772439956665, + "learning_rate": 4.810670719600987e-06, + "loss": 0.7135, + "step": 2218 + }, + { + "epoch": 0.7791432584269663, + "grad_norm": 0.8877284526824951, + "learning_rate": 4.810494190405432e-06, + "loss": 0.7184, + "step": 2219 + }, + { + "epoch": 0.7794943820224719, + "grad_norm": 0.8861910104751587, + "learning_rate": 4.8103175821929515e-06, + "loss": 0.6545, + "step": 2220 + }, + { + "epoch": 0.7798455056179775, + "grad_norm": 0.9048117995262146, + "learning_rate": 4.810140894969587e-06, + "loss": 0.6995, + "step": 2221 + }, + { + "epoch": 0.7801966292134831, + "grad_norm": 0.8994266390800476, + "learning_rate": 4.809964128741381e-06, + "loss": 0.6837, + "step": 2222 + }, + { + "epoch": 0.7805477528089888, + "grad_norm": 0.9110033512115479, + "learning_rate": 4.8097872835143775e-06, + "loss": 0.6575, + "step": 2223 + }, + { + "epoch": 0.7808988764044944, + "grad_norm": 0.8434987664222717, + "learning_rate": 4.8096103592946265e-06, + "loss": 0.6535, + "step": 2224 + }, + { + "epoch": 0.78125, + "grad_norm": 0.91326504945755, + "learning_rate": 4.8094333560881765e-06, + "loss": 0.6714, + "step": 2225 + }, + { + "epoch": 0.7816011235955056, + "grad_norm": 0.8931390643119812, + "learning_rate": 4.809256273901083e-06, + "loss": 0.667, + "step": 2226 + }, + { + "epoch": 0.7819522471910112, + "grad_norm": 0.8923385143280029, + "learning_rate": 4.8090791127393995e-06, + "loss": 0.6792, + "step": 2227 + }, + { + "epoch": 0.7823033707865169, + "grad_norm": 0.8582352995872498, + "learning_rate": 4.808901872609188e-06, + "loss": 0.674, + "step": 2228 + }, + { + "epoch": 0.7826544943820225, + "grad_norm": 1.004380702972412, + "learning_rate": 4.808724553516507e-06, + "loss": 0.6736, + "step": 2229 + }, + { + "epoch": 0.7830056179775281, + "grad_norm": 0.9153397083282471, + "learning_rate": 4.808547155467422e-06, + "loss": 0.6842, + "step": 2230 + }, + { + "epoch": 0.7833567415730337, + "grad_norm": 0.8752025961875916, + "learning_rate": 4.808369678468001e-06, + "loss": 0.702, + "step": 2231 + }, + { + "epoch": 0.7837078651685393, + "grad_norm": 0.9121509790420532, + "learning_rate": 4.808192122524311e-06, + "loss": 0.6794, + "step": 2232 + }, + { + "epoch": 0.7840589887640449, + "grad_norm": 0.893732488155365, + "learning_rate": 4.808014487642426e-06, + "loss": 0.6783, + "step": 2233 + }, + { + "epoch": 0.7844101123595506, + "grad_norm": 0.8836367726325989, + "learning_rate": 4.8078367738284205e-06, + "loss": 0.7095, + "step": 2234 + }, + { + "epoch": 0.7847612359550562, + "grad_norm": 0.8630282878875732, + "learning_rate": 4.8076589810883725e-06, + "loss": 0.6262, + "step": 2235 + }, + { + "epoch": 0.7851123595505618, + "grad_norm": 0.8813517093658447, + "learning_rate": 4.807481109428362e-06, + "loss": 0.6728, + "step": 2236 + }, + { + "epoch": 0.7854634831460674, + "grad_norm": 0.8723793625831604, + "learning_rate": 4.80730315885447e-06, + "loss": 0.6554, + "step": 2237 + }, + { + "epoch": 0.785814606741573, + "grad_norm": 0.8989778757095337, + "learning_rate": 4.807125129372787e-06, + "loss": 0.6587, + "step": 2238 + }, + { + "epoch": 0.7861657303370787, + "grad_norm": 0.8957408666610718, + "learning_rate": 4.8069470209893976e-06, + "loss": 0.6604, + "step": 2239 + }, + { + "epoch": 0.7865168539325843, + "grad_norm": 0.9145181775093079, + "learning_rate": 4.806768833710395e-06, + "loss": 0.6794, + "step": 2240 + }, + { + "epoch": 0.7868679775280899, + "grad_norm": 0.8940600752830505, + "learning_rate": 4.806590567541871e-06, + "loss": 0.686, + "step": 2241 + }, + { + "epoch": 0.7872191011235955, + "grad_norm": 0.8759947419166565, + "learning_rate": 4.806412222489925e-06, + "loss": 0.7002, + "step": 2242 + }, + { + "epoch": 0.7875702247191011, + "grad_norm": 0.8928653001785278, + "learning_rate": 4.8062337985606535e-06, + "loss": 0.6343, + "step": 2243 + }, + { + "epoch": 0.7879213483146067, + "grad_norm": 0.9009339809417725, + "learning_rate": 4.806055295760159e-06, + "loss": 0.7096, + "step": 2244 + }, + { + "epoch": 0.7882724719101124, + "grad_norm": 0.8576456308364868, + "learning_rate": 4.8058767140945475e-06, + "loss": 0.6449, + "step": 2245 + }, + { + "epoch": 0.788623595505618, + "grad_norm": 0.8837461471557617, + "learning_rate": 4.805698053569926e-06, + "loss": 0.6565, + "step": 2246 + }, + { + "epoch": 0.7889747191011236, + "grad_norm": 1.0404770374298096, + "learning_rate": 4.805519314192403e-06, + "loss": 0.6726, + "step": 2247 + }, + { + "epoch": 0.7893258426966292, + "grad_norm": 0.9084009528160095, + "learning_rate": 4.805340495968093e-06, + "loss": 0.6518, + "step": 2248 + }, + { + "epoch": 0.7896769662921348, + "grad_norm": 0.9170441031455994, + "learning_rate": 4.805161598903111e-06, + "loss": 0.6733, + "step": 2249 + }, + { + "epoch": 0.7900280898876404, + "grad_norm": 0.9394940733909607, + "learning_rate": 4.804982623003574e-06, + "loss": 0.6898, + "step": 2250 + }, + { + "epoch": 0.7903792134831461, + "grad_norm": 0.9983392953872681, + "learning_rate": 4.8048035682756046e-06, + "loss": 0.6739, + "step": 2251 + }, + { + "epoch": 0.7907303370786517, + "grad_norm": 0.9523327946662903, + "learning_rate": 4.8046244347253245e-06, + "loss": 0.6999, + "step": 2252 + }, + { + "epoch": 0.7910814606741573, + "grad_norm": 0.8619253635406494, + "learning_rate": 4.804445222358862e-06, + "loss": 0.6699, + "step": 2253 + }, + { + "epoch": 0.7914325842696629, + "grad_norm": 0.9346986413002014, + "learning_rate": 4.804265931182343e-06, + "loss": 0.6971, + "step": 2254 + }, + { + "epoch": 0.7917837078651685, + "grad_norm": 0.912071943283081, + "learning_rate": 4.804086561201903e-06, + "loss": 0.6913, + "step": 2255 + }, + { + "epoch": 0.7921348314606742, + "grad_norm": 0.9198614358901978, + "learning_rate": 4.8039071124236734e-06, + "loss": 0.6787, + "step": 2256 + }, + { + "epoch": 0.7924859550561798, + "grad_norm": 0.8504485487937927, + "learning_rate": 4.803727584853792e-06, + "loss": 0.6574, + "step": 2257 + }, + { + "epoch": 0.7928370786516854, + "grad_norm": 0.860133707523346, + "learning_rate": 4.8035479784983984e-06, + "loss": 0.7023, + "step": 2258 + }, + { + "epoch": 0.793188202247191, + "grad_norm": 0.8657998442649841, + "learning_rate": 4.803368293363635e-06, + "loss": 0.667, + "step": 2259 + }, + { + "epoch": 0.7935393258426966, + "grad_norm": 0.8903993368148804, + "learning_rate": 4.803188529455647e-06, + "loss": 0.6504, + "step": 2260 + }, + { + "epoch": 0.7938904494382022, + "grad_norm": 0.9583306908607483, + "learning_rate": 4.803008686780582e-06, + "loss": 0.6624, + "step": 2261 + }, + { + "epoch": 0.7942415730337079, + "grad_norm": 0.9079137444496155, + "learning_rate": 4.802828765344592e-06, + "loss": 0.6845, + "step": 2262 + }, + { + "epoch": 0.7945926966292135, + "grad_norm": 0.8480777740478516, + "learning_rate": 4.802648765153828e-06, + "loss": 0.6416, + "step": 2263 + }, + { + "epoch": 0.7949438202247191, + "grad_norm": 0.9052068591117859, + "learning_rate": 4.802468686214446e-06, + "loss": 0.676, + "step": 2264 + }, + { + "epoch": 0.7952949438202247, + "grad_norm": 0.8713710904121399, + "learning_rate": 4.802288528532606e-06, + "loss": 0.6823, + "step": 2265 + }, + { + "epoch": 0.7956460674157303, + "grad_norm": 0.8888510465621948, + "learning_rate": 4.802108292114468e-06, + "loss": 0.6932, + "step": 2266 + }, + { + "epoch": 0.795997191011236, + "grad_norm": 0.9380448460578918, + "learning_rate": 4.8019279769661965e-06, + "loss": 0.7137, + "step": 2267 + }, + { + "epoch": 0.7963483146067416, + "grad_norm": 0.8584917783737183, + "learning_rate": 4.801747583093958e-06, + "loss": 0.66, + "step": 2268 + }, + { + "epoch": 0.7966994382022472, + "grad_norm": 0.8634847402572632, + "learning_rate": 4.8015671105039205e-06, + "loss": 0.6553, + "step": 2269 + }, + { + "epoch": 0.7970505617977528, + "grad_norm": 1.1672545671463013, + "learning_rate": 4.801386559202259e-06, + "loss": 0.7002, + "step": 2270 + }, + { + "epoch": 0.7974016853932584, + "grad_norm": 0.8866782188415527, + "learning_rate": 4.801205929195146e-06, + "loss": 0.6791, + "step": 2271 + }, + { + "epoch": 0.797752808988764, + "grad_norm": 0.92203289270401, + "learning_rate": 4.801025220488759e-06, + "loss": 0.6586, + "step": 2272 + }, + { + "epoch": 0.7981039325842697, + "grad_norm": 0.9017732739448547, + "learning_rate": 4.8008444330892785e-06, + "loss": 0.6559, + "step": 2273 + }, + { + "epoch": 0.7984550561797753, + "grad_norm": 0.9230211973190308, + "learning_rate": 4.800663567002888e-06, + "loss": 0.6968, + "step": 2274 + }, + { + "epoch": 0.7988061797752809, + "grad_norm": 0.9016463160514832, + "learning_rate": 4.8004826222357715e-06, + "loss": 0.6624, + "step": 2275 + }, + { + "epoch": 0.7991573033707865, + "grad_norm": 0.851717472076416, + "learning_rate": 4.8003015987941185e-06, + "loss": 0.6454, + "step": 2276 + }, + { + "epoch": 0.7995084269662921, + "grad_norm": 0.8657079935073853, + "learning_rate": 4.800120496684119e-06, + "loss": 0.6537, + "step": 2277 + }, + { + "epoch": 0.7998595505617978, + "grad_norm": 0.8756804466247559, + "learning_rate": 4.7999393159119665e-06, + "loss": 0.6777, + "step": 2278 + }, + { + "epoch": 0.8002106741573034, + "grad_norm": 0.8858351111412048, + "learning_rate": 4.799758056483858e-06, + "loss": 0.673, + "step": 2279 + }, + { + "epoch": 0.800561797752809, + "grad_norm": 1.0514605045318604, + "learning_rate": 4.799576718405991e-06, + "loss": 0.6417, + "step": 2280 + }, + { + "epoch": 0.8009129213483146, + "grad_norm": 0.8649700284004211, + "learning_rate": 4.799395301684569e-06, + "loss": 0.6753, + "step": 2281 + }, + { + "epoch": 0.8012640449438202, + "grad_norm": 0.893130362033844, + "learning_rate": 4.799213806325794e-06, + "loss": 0.7144, + "step": 2282 + }, + { + "epoch": 0.8016151685393258, + "grad_norm": 0.8683095574378967, + "learning_rate": 4.799032232335875e-06, + "loss": 0.6565, + "step": 2283 + }, + { + "epoch": 0.8019662921348315, + "grad_norm": 0.8900540471076965, + "learning_rate": 4.798850579721022e-06, + "loss": 0.7065, + "step": 2284 + }, + { + "epoch": 0.8023174157303371, + "grad_norm": 0.9185386300086975, + "learning_rate": 4.7986688484874446e-06, + "loss": 0.6371, + "step": 2285 + }, + { + "epoch": 0.8026685393258427, + "grad_norm": 0.8507232666015625, + "learning_rate": 4.79848703864136e-06, + "loss": 0.66, + "step": 2286 + }, + { + "epoch": 0.8030196629213483, + "grad_norm": 0.8661529421806335, + "learning_rate": 4.798305150188986e-06, + "loss": 0.6643, + "step": 2287 + }, + { + "epoch": 0.8033707865168539, + "grad_norm": 0.8859586119651794, + "learning_rate": 4.798123183136542e-06, + "loss": 0.652, + "step": 2288 + }, + { + "epoch": 0.8037219101123596, + "grad_norm": 0.9179227948188782, + "learning_rate": 4.797941137490253e-06, + "loss": 0.7033, + "step": 2289 + }, + { + "epoch": 0.8040730337078652, + "grad_norm": 0.8715042471885681, + "learning_rate": 4.797759013256343e-06, + "loss": 0.6482, + "step": 2290 + }, + { + "epoch": 0.8044241573033708, + "grad_norm": 0.8953105211257935, + "learning_rate": 4.797576810441041e-06, + "loss": 0.6986, + "step": 2291 + }, + { + "epoch": 0.8047752808988764, + "grad_norm": 0.9580156803131104, + "learning_rate": 4.797394529050577e-06, + "loss": 0.6512, + "step": 2292 + }, + { + "epoch": 0.805126404494382, + "grad_norm": 0.9238266348838806, + "learning_rate": 4.7972121690911865e-06, + "loss": 0.6796, + "step": 2293 + }, + { + "epoch": 0.8054775280898876, + "grad_norm": 0.8915815949440002, + "learning_rate": 4.797029730569106e-06, + "loss": 0.707, + "step": 2294 + }, + { + "epoch": 0.8058286516853933, + "grad_norm": 0.8978826999664307, + "learning_rate": 4.796847213490574e-06, + "loss": 0.7162, + "step": 2295 + }, + { + "epoch": 0.8061797752808989, + "grad_norm": 0.8888537287712097, + "learning_rate": 4.796664617861832e-06, + "loss": 0.6832, + "step": 2296 + }, + { + "epoch": 0.8065308988764045, + "grad_norm": 0.8834527730941772, + "learning_rate": 4.796481943689127e-06, + "loss": 0.6863, + "step": 2297 + }, + { + "epoch": 0.8068820224719101, + "grad_norm": 0.8643321394920349, + "learning_rate": 4.796299190978704e-06, + "loss": 0.6739, + "step": 2298 + }, + { + "epoch": 0.8072331460674157, + "grad_norm": 0.8717115521430969, + "learning_rate": 4.796116359736813e-06, + "loss": 0.6502, + "step": 2299 + }, + { + "epoch": 0.8075842696629213, + "grad_norm": 0.8730166554450989, + "learning_rate": 4.7959334499697085e-06, + "loss": 0.6604, + "step": 2300 + }, + { + "epoch": 0.807935393258427, + "grad_norm": 0.8617185950279236, + "learning_rate": 4.795750461683643e-06, + "loss": 0.667, + "step": 2301 + }, + { + "epoch": 0.8082865168539326, + "grad_norm": 0.826866626739502, + "learning_rate": 4.795567394884878e-06, + "loss": 0.6273, + "step": 2302 + }, + { + "epoch": 0.8086376404494382, + "grad_norm": 0.9160782098770142, + "learning_rate": 4.795384249579672e-06, + "loss": 0.6841, + "step": 2303 + }, + { + "epoch": 0.8089887640449438, + "grad_norm": 0.8521285653114319, + "learning_rate": 4.795201025774287e-06, + "loss": 0.6311, + "step": 2304 + }, + { + "epoch": 0.8093398876404494, + "grad_norm": 0.9138549566268921, + "learning_rate": 4.795017723474993e-06, + "loss": 0.6848, + "step": 2305 + }, + { + "epoch": 0.8096910112359551, + "grad_norm": 0.8857465386390686, + "learning_rate": 4.794834342688056e-06, + "loss": 0.6457, + "step": 2306 + }, + { + "epoch": 0.8100421348314607, + "grad_norm": 0.934290885925293, + "learning_rate": 4.7946508834197495e-06, + "loss": 0.6853, + "step": 2307 + }, + { + "epoch": 0.8103932584269663, + "grad_norm": 0.8928509950637817, + "learning_rate": 4.794467345676345e-06, + "loss": 0.6884, + "step": 2308 + }, + { + "epoch": 0.8107443820224719, + "grad_norm": 0.9139081835746765, + "learning_rate": 4.79428372946412e-06, + "loss": 0.7047, + "step": 2309 + }, + { + "epoch": 0.8110955056179775, + "grad_norm": 0.8674109578132629, + "learning_rate": 4.794100034789356e-06, + "loss": 0.6397, + "step": 2310 + }, + { + "epoch": 0.8114466292134831, + "grad_norm": 0.9009836912155151, + "learning_rate": 4.793916261658333e-06, + "loss": 0.6954, + "step": 2311 + }, + { + "epoch": 0.8117977528089888, + "grad_norm": 0.8968256115913391, + "learning_rate": 4.7937324100773375e-06, + "loss": 0.6983, + "step": 2312 + }, + { + "epoch": 0.8121488764044944, + "grad_norm": 0.9040331244468689, + "learning_rate": 4.793548480052656e-06, + "loss": 0.6932, + "step": 2313 + }, + { + "epoch": 0.8125, + "grad_norm": 0.9160118699073792, + "learning_rate": 4.7933644715905794e-06, + "loss": 0.702, + "step": 2314 + }, + { + "epoch": 0.8128511235955056, + "grad_norm": 0.8957852721214294, + "learning_rate": 4.7931803846974005e-06, + "loss": 0.6737, + "step": 2315 + }, + { + "epoch": 0.8132022471910112, + "grad_norm": 0.9111645817756653, + "learning_rate": 4.7929962193794145e-06, + "loss": 0.6716, + "step": 2316 + }, + { + "epoch": 0.8135533707865169, + "grad_norm": 0.9393349885940552, + "learning_rate": 4.792811975642919e-06, + "loss": 0.7166, + "step": 2317 + }, + { + "epoch": 0.8139044943820225, + "grad_norm": 0.8632532358169556, + "learning_rate": 4.792627653494217e-06, + "loss": 0.6545, + "step": 2318 + }, + { + "epoch": 0.8142556179775281, + "grad_norm": 0.8554111123085022, + "learning_rate": 4.792443252939612e-06, + "loss": 0.6421, + "step": 2319 + }, + { + "epoch": 0.8146067415730337, + "grad_norm": 0.8907461166381836, + "learning_rate": 4.792258773985408e-06, + "loss": 0.6874, + "step": 2320 + }, + { + "epoch": 0.8149578651685393, + "grad_norm": 0.8739638924598694, + "learning_rate": 4.792074216637916e-06, + "loss": 0.6662, + "step": 2321 + }, + { + "epoch": 0.8153089887640449, + "grad_norm": 0.8694685101509094, + "learning_rate": 4.791889580903446e-06, + "loss": 0.6536, + "step": 2322 + }, + { + "epoch": 0.8156601123595506, + "grad_norm": 0.9057086706161499, + "learning_rate": 4.791704866788314e-06, + "loss": 0.6655, + "step": 2323 + }, + { + "epoch": 0.8160112359550562, + "grad_norm": 0.9522676467895508, + "learning_rate": 4.791520074298837e-06, + "loss": 0.7235, + "step": 2324 + }, + { + "epoch": 0.8163623595505618, + "grad_norm": 0.8711917996406555, + "learning_rate": 4.791335203441335e-06, + "loss": 0.7015, + "step": 2325 + }, + { + "epoch": 0.8167134831460674, + "grad_norm": 0.8643499612808228, + "learning_rate": 4.791150254222129e-06, + "loss": 0.6229, + "step": 2326 + }, + { + "epoch": 0.817064606741573, + "grad_norm": 0.8863927721977234, + "learning_rate": 4.790965226647545e-06, + "loss": 0.7017, + "step": 2327 + }, + { + "epoch": 0.8174157303370787, + "grad_norm": 0.8969694375991821, + "learning_rate": 4.790780120723911e-06, + "loss": 0.6926, + "step": 2328 + }, + { + "epoch": 0.8177668539325843, + "grad_norm": 0.923423171043396, + "learning_rate": 4.790594936457556e-06, + "loss": 0.707, + "step": 2329 + }, + { + "epoch": 0.8181179775280899, + "grad_norm": 0.8579330444335938, + "learning_rate": 4.790409673854815e-06, + "loss": 0.6224, + "step": 2330 + }, + { + "epoch": 0.8184691011235955, + "grad_norm": 0.8940487504005432, + "learning_rate": 4.790224332922023e-06, + "loss": 0.6859, + "step": 2331 + }, + { + "epoch": 0.8188202247191011, + "grad_norm": 0.8670837879180908, + "learning_rate": 4.790038913665519e-06, + "loss": 0.6669, + "step": 2332 + }, + { + "epoch": 0.8191713483146067, + "grad_norm": 0.8588036298751831, + "learning_rate": 4.789853416091643e-06, + "loss": 0.6951, + "step": 2333 + }, + { + "epoch": 0.8195224719101124, + "grad_norm": 0.8730223774909973, + "learning_rate": 4.78966784020674e-06, + "loss": 0.6751, + "step": 2334 + }, + { + "epoch": 0.819873595505618, + "grad_norm": 0.8985008001327515, + "learning_rate": 4.789482186017157e-06, + "loss": 0.6927, + "step": 2335 + }, + { + "epoch": 0.8202247191011236, + "grad_norm": 0.8898709416389465, + "learning_rate": 4.78929645352924e-06, + "loss": 0.6307, + "step": 2336 + }, + { + "epoch": 0.8205758426966292, + "grad_norm": 0.8730369806289673, + "learning_rate": 4.789110642749346e-06, + "loss": 0.6548, + "step": 2337 + }, + { + "epoch": 0.8209269662921348, + "grad_norm": 0.8780831694602966, + "learning_rate": 4.788924753683826e-06, + "loss": 0.6451, + "step": 2338 + }, + { + "epoch": 0.8212780898876404, + "grad_norm": 0.8665900230407715, + "learning_rate": 4.788738786339037e-06, + "loss": 0.6656, + "step": 2339 + }, + { + "epoch": 0.8216292134831461, + "grad_norm": 0.9199172854423523, + "learning_rate": 4.788552740721341e-06, + "loss": 0.6629, + "step": 2340 + }, + { + "epoch": 0.8219803370786517, + "grad_norm": 0.8883637189865112, + "learning_rate": 4.7883666168370996e-06, + "loss": 0.6789, + "step": 2341 + }, + { + "epoch": 0.8223314606741573, + "grad_norm": 0.8635849356651306, + "learning_rate": 4.788180414692678e-06, + "loss": 0.6823, + "step": 2342 + }, + { + "epoch": 0.8226825842696629, + "grad_norm": 0.9055063128471375, + "learning_rate": 4.787994134294444e-06, + "loss": 0.6667, + "step": 2343 + }, + { + "epoch": 0.8230337078651685, + "grad_norm": 0.8952557444572449, + "learning_rate": 4.787807775648768e-06, + "loss": 0.6694, + "step": 2344 + }, + { + "epoch": 0.8233848314606742, + "grad_norm": 0.9171862006187439, + "learning_rate": 4.787621338762025e-06, + "loss": 0.6938, + "step": 2345 + }, + { + "epoch": 0.8237359550561798, + "grad_norm": 0.9036487936973572, + "learning_rate": 4.787434823640589e-06, + "loss": 0.6565, + "step": 2346 + }, + { + "epoch": 0.8240870786516854, + "grad_norm": 0.8997727036476135, + "learning_rate": 4.787248230290839e-06, + "loss": 0.6586, + "step": 2347 + }, + { + "epoch": 0.824438202247191, + "grad_norm": 0.9203174114227295, + "learning_rate": 4.787061558719158e-06, + "loss": 0.6684, + "step": 2348 + }, + { + "epoch": 0.8247893258426966, + "grad_norm": 0.8566135168075562, + "learning_rate": 4.786874808931928e-06, + "loss": 0.6366, + "step": 2349 + }, + { + "epoch": 0.8251404494382022, + "grad_norm": 0.8947523832321167, + "learning_rate": 4.786687980935536e-06, + "loss": 0.6954, + "step": 2350 + }, + { + "epoch": 0.8254915730337079, + "grad_norm": 0.8714900016784668, + "learning_rate": 4.7865010747363725e-06, + "loss": 0.6678, + "step": 2351 + }, + { + "epoch": 0.8258426966292135, + "grad_norm": 0.8742727041244507, + "learning_rate": 4.786314090340829e-06, + "loss": 0.6356, + "step": 2352 + }, + { + "epoch": 0.8261938202247191, + "grad_norm": 0.8638887405395508, + "learning_rate": 4.7861270277553e-06, + "loss": 0.6643, + "step": 2353 + }, + { + "epoch": 0.8265449438202247, + "grad_norm": 0.8582507967948914, + "learning_rate": 4.785939886986182e-06, + "loss": 0.6943, + "step": 2354 + }, + { + "epoch": 0.8268960674157303, + "grad_norm": 0.8985755443572998, + "learning_rate": 4.785752668039877e-06, + "loss": 0.6899, + "step": 2355 + }, + { + "epoch": 0.827247191011236, + "grad_norm": 0.889373242855072, + "learning_rate": 4.7855653709227856e-06, + "loss": 0.6687, + "step": 2356 + }, + { + "epoch": 0.8275983146067416, + "grad_norm": 0.8748716711997986, + "learning_rate": 4.785377995641315e-06, + "loss": 0.6813, + "step": 2357 + }, + { + "epoch": 0.8279494382022472, + "grad_norm": 0.912598729133606, + "learning_rate": 4.785190542201873e-06, + "loss": 0.7037, + "step": 2358 + }, + { + "epoch": 0.8283005617977528, + "grad_norm": 0.8579078912734985, + "learning_rate": 4.785003010610869e-06, + "loss": 0.651, + "step": 2359 + }, + { + "epoch": 0.8286516853932584, + "grad_norm": 0.8841166496276855, + "learning_rate": 4.7848154008747185e-06, + "loss": 0.6527, + "step": 2360 + }, + { + "epoch": 0.829002808988764, + "grad_norm": 0.9118697047233582, + "learning_rate": 4.784627712999836e-06, + "loss": 0.7196, + "step": 2361 + }, + { + "epoch": 0.8293539325842697, + "grad_norm": 0.8655892610549927, + "learning_rate": 4.78443994699264e-06, + "loss": 0.6813, + "step": 2362 + }, + { + "epoch": 0.8297050561797753, + "grad_norm": 0.8933113217353821, + "learning_rate": 4.784252102859553e-06, + "loss": 0.6276, + "step": 2363 + }, + { + "epoch": 0.8300561797752809, + "grad_norm": 0.8977401852607727, + "learning_rate": 4.784064180606999e-06, + "loss": 0.7107, + "step": 2364 + }, + { + "epoch": 0.8304073033707865, + "grad_norm": 0.8428375720977783, + "learning_rate": 4.783876180241404e-06, + "loss": 0.6187, + "step": 2365 + }, + { + "epoch": 0.8307584269662921, + "grad_norm": 0.8943214416503906, + "learning_rate": 4.783688101769199e-06, + "loss": 0.71, + "step": 2366 + }, + { + "epoch": 0.8311095505617978, + "grad_norm": 0.8977285027503967, + "learning_rate": 4.783499945196814e-06, + "loss": 0.6895, + "step": 2367 + }, + { + "epoch": 0.8314606741573034, + "grad_norm": 0.8746622800827026, + "learning_rate": 4.783311710530685e-06, + "loss": 0.656, + "step": 2368 + }, + { + "epoch": 0.831811797752809, + "grad_norm": 0.9107025265693665, + "learning_rate": 4.783123397777249e-06, + "loss": 0.6869, + "step": 2369 + }, + { + "epoch": 0.8321629213483146, + "grad_norm": 0.8555697202682495, + "learning_rate": 4.782935006942948e-06, + "loss": 0.6607, + "step": 2370 + }, + { + "epoch": 0.8325140449438202, + "grad_norm": 0.9654397368431091, + "learning_rate": 4.782746538034222e-06, + "loss": 0.7033, + "step": 2371 + }, + { + "epoch": 0.8328651685393258, + "grad_norm": 0.8786224722862244, + "learning_rate": 4.782557991057519e-06, + "loss": 0.6855, + "step": 2372 + }, + { + "epoch": 0.8332162921348315, + "grad_norm": 0.8756742477416992, + "learning_rate": 4.782369366019285e-06, + "loss": 0.6665, + "step": 2373 + }, + { + "epoch": 0.8335674157303371, + "grad_norm": 0.8499640226364136, + "learning_rate": 4.782180662925971e-06, + "loss": 0.6311, + "step": 2374 + }, + { + "epoch": 0.8339185393258427, + "grad_norm": 0.8776542544364929, + "learning_rate": 4.781991881784033e-06, + "loss": 0.6894, + "step": 2375 + }, + { + "epoch": 0.8342696629213483, + "grad_norm": 0.8762704730033875, + "learning_rate": 4.7818030225999245e-06, + "loss": 0.6637, + "step": 2376 + }, + { + "epoch": 0.8346207865168539, + "grad_norm": 0.8762661218643188, + "learning_rate": 4.7816140853801054e-06, + "loss": 0.6672, + "step": 2377 + }, + { + "epoch": 0.8349719101123596, + "grad_norm": 0.8843420743942261, + "learning_rate": 4.781425070131036e-06, + "loss": 0.6614, + "step": 2378 + }, + { + "epoch": 0.8353230337078652, + "grad_norm": 0.8788619041442871, + "learning_rate": 4.781235976859183e-06, + "loss": 0.6853, + "step": 2379 + }, + { + "epoch": 0.8356741573033708, + "grad_norm": 0.9134128093719482, + "learning_rate": 4.78104680557101e-06, + "loss": 0.6904, + "step": 2380 + }, + { + "epoch": 0.8360252808988764, + "grad_norm": 0.8329752683639526, + "learning_rate": 4.7808575562729894e-06, + "loss": 0.6318, + "step": 2381 + }, + { + "epoch": 0.836376404494382, + "grad_norm": 0.8506373763084412, + "learning_rate": 4.780668228971592e-06, + "loss": 0.6757, + "step": 2382 + }, + { + "epoch": 0.8367275280898876, + "grad_norm": 0.8934277892112732, + "learning_rate": 4.780478823673292e-06, + "loss": 0.6961, + "step": 2383 + }, + { + "epoch": 0.8370786516853933, + "grad_norm": 0.90306556224823, + "learning_rate": 4.780289340384569e-06, + "loss": 0.729, + "step": 2384 + }, + { + "epoch": 0.8374297752808989, + "grad_norm": 0.8692042231559753, + "learning_rate": 4.780099779111901e-06, + "loss": 0.6695, + "step": 2385 + }, + { + "epoch": 0.8377808988764045, + "grad_norm": 0.8926328420639038, + "learning_rate": 4.779910139861771e-06, + "loss": 0.6895, + "step": 2386 + }, + { + "epoch": 0.8381320224719101, + "grad_norm": 0.8823839426040649, + "learning_rate": 4.779720422640666e-06, + "loss": 0.6587, + "step": 2387 + }, + { + "epoch": 0.8384831460674157, + "grad_norm": 0.8437162637710571, + "learning_rate": 4.7795306274550735e-06, + "loss": 0.6577, + "step": 2388 + }, + { + "epoch": 0.8388342696629213, + "grad_norm": 0.8754181265830994, + "learning_rate": 4.779340754311483e-06, + "loss": 0.6751, + "step": 2389 + }, + { + "epoch": 0.839185393258427, + "grad_norm": 0.8797729015350342, + "learning_rate": 4.77915080321639e-06, + "loss": 0.675, + "step": 2390 + }, + { + "epoch": 0.8395365168539326, + "grad_norm": 0.8657658100128174, + "learning_rate": 4.778960774176291e-06, + "loss": 0.6504, + "step": 2391 + }, + { + "epoch": 0.8398876404494382, + "grad_norm": 0.8967252969741821, + "learning_rate": 4.778770667197682e-06, + "loss": 0.6623, + "step": 2392 + }, + { + "epoch": 0.8402387640449438, + "grad_norm": 0.881457507610321, + "learning_rate": 4.778580482287067e-06, + "loss": 0.6514, + "step": 2393 + }, + { + "epoch": 0.8405898876404494, + "grad_norm": 0.8704044222831726, + "learning_rate": 4.778390219450949e-06, + "loss": 0.6695, + "step": 2394 + }, + { + "epoch": 0.8409410112359551, + "grad_norm": 0.9062277674674988, + "learning_rate": 4.778199878695835e-06, + "loss": 0.6394, + "step": 2395 + }, + { + "epoch": 0.8412921348314607, + "grad_norm": 0.873767614364624, + "learning_rate": 4.778009460028235e-06, + "loss": 0.6059, + "step": 2396 + }, + { + "epoch": 0.8416432584269663, + "grad_norm": 0.9467976689338684, + "learning_rate": 4.77781896345466e-06, + "loss": 0.6834, + "step": 2397 + }, + { + "epoch": 0.8419943820224719, + "grad_norm": 0.8977733254432678, + "learning_rate": 4.777628388981626e-06, + "loss": 0.703, + "step": 2398 + }, + { + "epoch": 0.8423455056179775, + "grad_norm": 0.8968080282211304, + "learning_rate": 4.7774377366156505e-06, + "loss": 0.678, + "step": 2399 + }, + { + "epoch": 0.8426966292134831, + "grad_norm": 0.8892068862915039, + "learning_rate": 4.777247006363253e-06, + "loss": 0.7203, + "step": 2400 + }, + { + "epoch": 0.8430477528089888, + "grad_norm": 0.8844308853149414, + "learning_rate": 4.777056198230956e-06, + "loss": 0.6694, + "step": 2401 + }, + { + "epoch": 0.8433988764044944, + "grad_norm": 0.8951604962348938, + "learning_rate": 4.7768653122252865e-06, + "loss": 0.6701, + "step": 2402 + }, + { + "epoch": 0.84375, + "grad_norm": 0.8704971075057983, + "learning_rate": 4.7766743483527715e-06, + "loss": 0.6616, + "step": 2403 + }, + { + "epoch": 0.8441011235955056, + "grad_norm": 0.908189058303833, + "learning_rate": 4.776483306619941e-06, + "loss": 0.6668, + "step": 2404 + }, + { + "epoch": 0.8444522471910112, + "grad_norm": 0.8829559087753296, + "learning_rate": 4.77629218703333e-06, + "loss": 0.6891, + "step": 2405 + }, + { + "epoch": 0.8448033707865169, + "grad_norm": 0.912239134311676, + "learning_rate": 4.776100989599475e-06, + "loss": 0.7016, + "step": 2406 + }, + { + "epoch": 0.8451544943820225, + "grad_norm": 0.8914303779602051, + "learning_rate": 4.775909714324912e-06, + "loss": 0.6593, + "step": 2407 + }, + { + "epoch": 0.8455056179775281, + "grad_norm": 0.9197940826416016, + "learning_rate": 4.775718361216186e-06, + "loss": 0.677, + "step": 2408 + }, + { + "epoch": 0.8458567415730337, + "grad_norm": 0.9217466115951538, + "learning_rate": 4.775526930279839e-06, + "loss": 0.7026, + "step": 2409 + }, + { + "epoch": 0.8462078651685393, + "grad_norm": 0.8749791383743286, + "learning_rate": 4.775335421522418e-06, + "loss": 0.6502, + "step": 2410 + }, + { + "epoch": 0.8465589887640449, + "grad_norm": 0.8961182236671448, + "learning_rate": 4.7751438349504716e-06, + "loss": 0.651, + "step": 2411 + }, + { + "epoch": 0.8469101123595506, + "grad_norm": 0.8679896593093872, + "learning_rate": 4.7749521705705546e-06, + "loss": 0.6445, + "step": 2412 + }, + { + "epoch": 0.8472612359550562, + "grad_norm": 0.8874735236167908, + "learning_rate": 4.774760428389219e-06, + "loss": 0.685, + "step": 2413 + }, + { + "epoch": 0.8476123595505618, + "grad_norm": 0.8967884182929993, + "learning_rate": 4.774568608413024e-06, + "loss": 0.6812, + "step": 2414 + }, + { + "epoch": 0.8479634831460674, + "grad_norm": 0.8984326720237732, + "learning_rate": 4.774376710648528e-06, + "loss": 0.6847, + "step": 2415 + }, + { + "epoch": 0.848314606741573, + "grad_norm": 0.861842155456543, + "learning_rate": 4.7741847351022955e-06, + "loss": 0.6643, + "step": 2416 + }, + { + "epoch": 0.8486657303370787, + "grad_norm": 0.8478919267654419, + "learning_rate": 4.773992681780891e-06, + "loss": 0.6309, + "step": 2417 + }, + { + "epoch": 0.8490168539325843, + "grad_norm": 0.8753276467323303, + "learning_rate": 4.773800550690882e-06, + "loss": 0.6906, + "step": 2418 + }, + { + "epoch": 0.8493679775280899, + "grad_norm": 0.9123186469078064, + "learning_rate": 4.7736083418388404e-06, + "loss": 0.6698, + "step": 2419 + }, + { + "epoch": 0.8497191011235955, + "grad_norm": 0.8448005318641663, + "learning_rate": 4.773416055231339e-06, + "loss": 0.6671, + "step": 2420 + }, + { + "epoch": 0.8500702247191011, + "grad_norm": 0.886045515537262, + "learning_rate": 4.773223690874954e-06, + "loss": 0.6925, + "step": 2421 + }, + { + "epoch": 0.8504213483146067, + "grad_norm": 0.8761569857597351, + "learning_rate": 4.773031248776264e-06, + "loss": 0.6988, + "step": 2422 + }, + { + "epoch": 0.8507724719101124, + "grad_norm": 0.8618139028549194, + "learning_rate": 4.772838728941851e-06, + "loss": 0.6345, + "step": 2423 + }, + { + "epoch": 0.851123595505618, + "grad_norm": 0.8982943296432495, + "learning_rate": 4.772646131378297e-06, + "loss": 0.6775, + "step": 2424 + }, + { + "epoch": 0.8514747191011236, + "grad_norm": 0.9190080761909485, + "learning_rate": 4.772453456092191e-06, + "loss": 0.6683, + "step": 2425 + }, + { + "epoch": 0.8518258426966292, + "grad_norm": 0.8613485097885132, + "learning_rate": 4.772260703090121e-06, + "loss": 0.6378, + "step": 2426 + }, + { + "epoch": 0.8521769662921348, + "grad_norm": 0.8885965943336487, + "learning_rate": 4.7720678723786796e-06, + "loss": 0.6971, + "step": 2427 + }, + { + "epoch": 0.8525280898876404, + "grad_norm": 0.8576337099075317, + "learning_rate": 4.771874963964461e-06, + "loss": 0.6417, + "step": 2428 + }, + { + "epoch": 0.8528792134831461, + "grad_norm": 0.8758781552314758, + "learning_rate": 4.771681977854062e-06, + "loss": 0.668, + "step": 2429 + }, + { + "epoch": 0.8532303370786517, + "grad_norm": 0.881989061832428, + "learning_rate": 4.771488914054085e-06, + "loss": 0.655, + "step": 2430 + }, + { + "epoch": 0.8535814606741573, + "grad_norm": 0.8963753581047058, + "learning_rate": 4.77129577257113e-06, + "loss": 0.647, + "step": 2431 + }, + { + "epoch": 0.8539325842696629, + "grad_norm": 0.8885678648948669, + "learning_rate": 4.771102553411803e-06, + "loss": 0.6832, + "step": 2432 + }, + { + "epoch": 0.8542837078651685, + "grad_norm": 0.8854667544364929, + "learning_rate": 4.770909256582712e-06, + "loss": 0.6387, + "step": 2433 + }, + { + "epoch": 0.8546348314606742, + "grad_norm": 0.9203517436981201, + "learning_rate": 4.770715882090468e-06, + "loss": 0.6606, + "step": 2434 + }, + { + "epoch": 0.8549859550561798, + "grad_norm": 0.9271900653839111, + "learning_rate": 4.770522429941683e-06, + "loss": 0.6787, + "step": 2435 + }, + { + "epoch": 0.8553370786516854, + "grad_norm": 0.8828155994415283, + "learning_rate": 4.770328900142975e-06, + "loss": 0.6742, + "step": 2436 + }, + { + "epoch": 0.855688202247191, + "grad_norm": 0.8759010434150696, + "learning_rate": 4.770135292700961e-06, + "loss": 0.6742, + "step": 2437 + }, + { + "epoch": 0.8560393258426966, + "grad_norm": 0.9363099932670593, + "learning_rate": 4.769941607622262e-06, + "loss": 0.6997, + "step": 2438 + }, + { + "epoch": 0.8563904494382022, + "grad_norm": 0.89815753698349, + "learning_rate": 4.769747844913503e-06, + "loss": 0.6643, + "step": 2439 + }, + { + "epoch": 0.8567415730337079, + "grad_norm": 0.8725033402442932, + "learning_rate": 4.76955400458131e-06, + "loss": 0.6433, + "step": 2440 + }, + { + "epoch": 0.8570926966292135, + "grad_norm": 0.8719140291213989, + "learning_rate": 4.769360086632312e-06, + "loss": 0.6215, + "step": 2441 + }, + { + "epoch": 0.8574438202247191, + "grad_norm": 0.8736863732337952, + "learning_rate": 4.76916609107314e-06, + "loss": 0.6854, + "step": 2442 + }, + { + "epoch": 0.8577949438202247, + "grad_norm": 0.8882306814193726, + "learning_rate": 4.768972017910432e-06, + "loss": 0.678, + "step": 2443 + }, + { + "epoch": 0.8581460674157303, + "grad_norm": 0.8788784146308899, + "learning_rate": 4.76877786715082e-06, + "loss": 0.6413, + "step": 2444 + }, + { + "epoch": 0.858497191011236, + "grad_norm": 0.8777339458465576, + "learning_rate": 4.768583638800947e-06, + "loss": 0.703, + "step": 2445 + }, + { + "epoch": 0.8588483146067416, + "grad_norm": 0.8905528783798218, + "learning_rate": 4.7683893328674556e-06, + "loss": 0.6653, + "step": 2446 + }, + { + "epoch": 0.8591994382022472, + "grad_norm": 0.8809261322021484, + "learning_rate": 4.76819494935699e-06, + "loss": 0.6357, + "step": 2447 + }, + { + "epoch": 0.8595505617977528, + "grad_norm": 0.9078806638717651, + "learning_rate": 4.768000488276198e-06, + "loss": 0.6924, + "step": 2448 + }, + { + "epoch": 0.8599016853932584, + "grad_norm": 0.8684207201004028, + "learning_rate": 4.767805949631729e-06, + "loss": 0.6753, + "step": 2449 + }, + { + "epoch": 0.860252808988764, + "grad_norm": 0.9299638867378235, + "learning_rate": 4.767611333430238e-06, + "loss": 0.6986, + "step": 2450 + }, + { + "epoch": 0.8606039325842697, + "grad_norm": 0.9025764465332031, + "learning_rate": 4.7674166396783796e-06, + "loss": 0.6832, + "step": 2451 + }, + { + "epoch": 0.8609550561797753, + "grad_norm": 0.8947761058807373, + "learning_rate": 4.767221868382812e-06, + "loss": 0.6927, + "step": 2452 + }, + { + "epoch": 0.8613061797752809, + "grad_norm": 0.9044678807258606, + "learning_rate": 4.767027019550198e-06, + "loss": 0.6803, + "step": 2453 + }, + { + "epoch": 0.8616573033707865, + "grad_norm": 0.8838576078414917, + "learning_rate": 4.766832093187199e-06, + "loss": 0.7069, + "step": 2454 + }, + { + "epoch": 0.8620084269662921, + "grad_norm": 0.8987936973571777, + "learning_rate": 4.766637089300481e-06, + "loss": 0.6986, + "step": 2455 + }, + { + "epoch": 0.8623595505617978, + "grad_norm": 0.9236810207366943, + "learning_rate": 4.766442007896716e-06, + "loss": 0.6912, + "step": 2456 + }, + { + "epoch": 0.8627106741573034, + "grad_norm": 0.9139038920402527, + "learning_rate": 4.766246848982573e-06, + "loss": 0.6647, + "step": 2457 + }, + { + "epoch": 0.863061797752809, + "grad_norm": 0.8569420576095581, + "learning_rate": 4.766051612564726e-06, + "loss": 0.6092, + "step": 2458 + }, + { + "epoch": 0.8634129213483146, + "grad_norm": 0.895006537437439, + "learning_rate": 4.765856298649855e-06, + "loss": 0.7071, + "step": 2459 + }, + { + "epoch": 0.8637640449438202, + "grad_norm": 0.8703930377960205, + "learning_rate": 4.7656609072446355e-06, + "loss": 0.6839, + "step": 2460 + }, + { + "epoch": 0.8641151685393258, + "grad_norm": 0.8899871706962585, + "learning_rate": 4.765465438355753e-06, + "loss": 0.6543, + "step": 2461 + }, + { + "epoch": 0.8644662921348315, + "grad_norm": 0.929097592830658, + "learning_rate": 4.765269891989891e-06, + "loss": 0.6848, + "step": 2462 + }, + { + "epoch": 0.8648174157303371, + "grad_norm": 0.8808541297912598, + "learning_rate": 4.765074268153736e-06, + "loss": 0.6625, + "step": 2463 + }, + { + "epoch": 0.8651685393258427, + "grad_norm": 0.9321450591087341, + "learning_rate": 4.764878566853981e-06, + "loss": 0.6483, + "step": 2464 + }, + { + "epoch": 0.8655196629213483, + "grad_norm": 0.8869593143463135, + "learning_rate": 4.764682788097317e-06, + "loss": 0.6546, + "step": 2465 + }, + { + "epoch": 0.8658707865168539, + "grad_norm": 0.9124717116355896, + "learning_rate": 4.764486931890438e-06, + "loss": 0.6637, + "step": 2466 + }, + { + "epoch": 0.8662219101123596, + "grad_norm": 0.9037688374519348, + "learning_rate": 4.764290998240044e-06, + "loss": 0.6656, + "step": 2467 + }, + { + "epoch": 0.8665730337078652, + "grad_norm": 0.9029697179794312, + "learning_rate": 4.7640949871528365e-06, + "loss": 0.6837, + "step": 2468 + }, + { + "epoch": 0.8669241573033708, + "grad_norm": 0.9010921716690063, + "learning_rate": 4.763898898635517e-06, + "loss": 0.6527, + "step": 2469 + }, + { + "epoch": 0.8672752808988764, + "grad_norm": 0.8654772043228149, + "learning_rate": 4.763702732694792e-06, + "loss": 0.6983, + "step": 2470 + }, + { + "epoch": 0.867626404494382, + "grad_norm": 0.8883764147758484, + "learning_rate": 4.763506489337372e-06, + "loss": 0.7035, + "step": 2471 + }, + { + "epoch": 0.8679775280898876, + "grad_norm": 0.902722179889679, + "learning_rate": 4.763310168569966e-06, + "loss": 0.6928, + "step": 2472 + }, + { + "epoch": 0.8683286516853933, + "grad_norm": 0.900360643863678, + "learning_rate": 4.76311377039929e-06, + "loss": 0.653, + "step": 2473 + }, + { + "epoch": 0.8686797752808989, + "grad_norm": 0.8905735015869141, + "learning_rate": 4.762917294832058e-06, + "loss": 0.6644, + "step": 2474 + }, + { + "epoch": 0.8690308988764045, + "grad_norm": 0.8971735835075378, + "learning_rate": 4.762720741874992e-06, + "loss": 0.6751, + "step": 2475 + }, + { + "epoch": 0.8693820224719101, + "grad_norm": 0.9062357544898987, + "learning_rate": 4.762524111534813e-06, + "loss": 0.6207, + "step": 2476 + }, + { + "epoch": 0.8697331460674157, + "grad_norm": 0.8765476942062378, + "learning_rate": 4.762327403818245e-06, + "loss": 0.6972, + "step": 2477 + }, + { + "epoch": 0.8700842696629213, + "grad_norm": 0.8708880543708801, + "learning_rate": 4.762130618732016e-06, + "loss": 0.6704, + "step": 2478 + }, + { + "epoch": 0.870435393258427, + "grad_norm": 0.8837103247642517, + "learning_rate": 4.761933756282856e-06, + "loss": 0.6314, + "step": 2479 + }, + { + "epoch": 0.8707865168539326, + "grad_norm": 0.8385997414588928, + "learning_rate": 4.7617368164774956e-06, + "loss": 0.6511, + "step": 2480 + }, + { + "epoch": 0.8711376404494382, + "grad_norm": 0.8760436773300171, + "learning_rate": 4.761539799322672e-06, + "loss": 0.6927, + "step": 2481 + }, + { + "epoch": 0.8714887640449438, + "grad_norm": 0.8821952939033508, + "learning_rate": 4.761342704825124e-06, + "loss": 0.6682, + "step": 2482 + }, + { + "epoch": 0.8718398876404494, + "grad_norm": 0.8891538381576538, + "learning_rate": 4.761145532991589e-06, + "loss": 0.6796, + "step": 2483 + }, + { + "epoch": 0.8721910112359551, + "grad_norm": 0.9022207260131836, + "learning_rate": 4.760948283828812e-06, + "loss": 0.7124, + "step": 2484 + }, + { + "epoch": 0.8725421348314607, + "grad_norm": 0.8623314499855042, + "learning_rate": 4.760750957343538e-06, + "loss": 0.6677, + "step": 2485 + }, + { + "epoch": 0.8728932584269663, + "grad_norm": 0.8784610033035278, + "learning_rate": 4.760553553542517e-06, + "loss": 0.6731, + "step": 2486 + }, + { + "epoch": 0.8732443820224719, + "grad_norm": 0.8761183619499207, + "learning_rate": 4.7603560724324985e-06, + "loss": 0.6246, + "step": 2487 + }, + { + "epoch": 0.8735955056179775, + "grad_norm": 0.8424711227416992, + "learning_rate": 4.7601585140202354e-06, + "loss": 0.6453, + "step": 2488 + }, + { + "epoch": 0.8739466292134831, + "grad_norm": 0.8858649730682373, + "learning_rate": 4.7599608783124865e-06, + "loss": 0.6869, + "step": 2489 + }, + { + "epoch": 0.8742977528089888, + "grad_norm": 0.8897294402122498, + "learning_rate": 4.759763165316008e-06, + "loss": 0.673, + "step": 2490 + }, + { + "epoch": 0.8746488764044944, + "grad_norm": 0.8692971467971802, + "learning_rate": 4.759565375037564e-06, + "loss": 0.6691, + "step": 2491 + }, + { + "epoch": 0.875, + "grad_norm": 0.9153478741645813, + "learning_rate": 4.759367507483918e-06, + "loss": 0.7012, + "step": 2492 + }, + { + "epoch": 0.8753511235955056, + "grad_norm": 0.9016615748405457, + "learning_rate": 4.759169562661836e-06, + "loss": 0.6758, + "step": 2493 + }, + { + "epoch": 0.8757022471910112, + "grad_norm": 0.8769146203994751, + "learning_rate": 4.758971540578088e-06, + "loss": 0.6989, + "step": 2494 + }, + { + "epoch": 0.8760533707865169, + "grad_norm": 0.8640443086624146, + "learning_rate": 4.758773441239447e-06, + "loss": 0.6355, + "step": 2495 + }, + { + "epoch": 0.8764044943820225, + "grad_norm": 0.9241236448287964, + "learning_rate": 4.758575264652687e-06, + "loss": 0.642, + "step": 2496 + }, + { + "epoch": 0.8767556179775281, + "grad_norm": 0.8472598791122437, + "learning_rate": 4.758377010824585e-06, + "loss": 0.6229, + "step": 2497 + }, + { + "epoch": 0.8771067415730337, + "grad_norm": 0.8446848392486572, + "learning_rate": 4.758178679761922e-06, + "loss": 0.6234, + "step": 2498 + }, + { + "epoch": 0.8774578651685393, + "grad_norm": 0.8826449513435364, + "learning_rate": 4.7579802714714815e-06, + "loss": 0.6799, + "step": 2499 + }, + { + "epoch": 0.8778089887640449, + "grad_norm": 0.8934155106544495, + "learning_rate": 4.757781785960047e-06, + "loss": 0.655, + "step": 2500 + }, + { + "epoch": 0.8781601123595506, + "grad_norm": 0.8733129501342773, + "learning_rate": 4.757583223234408e-06, + "loss": 0.6754, + "step": 2501 + }, + { + "epoch": 0.8785112359550562, + "grad_norm": 0.8651498556137085, + "learning_rate": 4.757384583301353e-06, + "loss": 0.6839, + "step": 2502 + }, + { + "epoch": 0.8788623595505618, + "grad_norm": 0.9135331511497498, + "learning_rate": 4.757185866167679e-06, + "loss": 0.6593, + "step": 2503 + }, + { + "epoch": 0.8792134831460674, + "grad_norm": 0.86785888671875, + "learning_rate": 4.756987071840179e-06, + "loss": 0.6568, + "step": 2504 + }, + { + "epoch": 0.879564606741573, + "grad_norm": 0.9271560311317444, + "learning_rate": 4.756788200325652e-06, + "loss": 0.6364, + "step": 2505 + }, + { + "epoch": 0.8799157303370787, + "grad_norm": 0.911849319934845, + "learning_rate": 4.7565892516309e-06, + "loss": 0.675, + "step": 2506 + }, + { + "epoch": 0.8802668539325843, + "grad_norm": 0.8646685481071472, + "learning_rate": 4.756390225762727e-06, + "loss": 0.6056, + "step": 2507 + }, + { + "epoch": 0.8806179775280899, + "grad_norm": 0.8820457458496094, + "learning_rate": 4.756191122727939e-06, + "loss": 0.6632, + "step": 2508 + }, + { + "epoch": 0.8809691011235955, + "grad_norm": 0.8895183801651001, + "learning_rate": 4.755991942533344e-06, + "loss": 0.6819, + "step": 2509 + }, + { + "epoch": 0.8813202247191011, + "grad_norm": 0.8657563924789429, + "learning_rate": 4.755792685185758e-06, + "loss": 0.649, + "step": 2510 + }, + { + "epoch": 0.8816713483146067, + "grad_norm": 0.9095283150672913, + "learning_rate": 4.755593350691989e-06, + "loss": 0.718, + "step": 2511 + }, + { + "epoch": 0.8820224719101124, + "grad_norm": 0.8777075409889221, + "learning_rate": 4.75539393905886e-06, + "loss": 0.6915, + "step": 2512 + }, + { + "epoch": 0.882373595505618, + "grad_norm": 0.8759570121765137, + "learning_rate": 4.755194450293187e-06, + "loss": 0.6749, + "step": 2513 + }, + { + "epoch": 0.8827247191011236, + "grad_norm": 0.8764015436172485, + "learning_rate": 4.754994884401794e-06, + "loss": 0.6652, + "step": 2514 + }, + { + "epoch": 0.8830758426966292, + "grad_norm": 0.9175446033477783, + "learning_rate": 4.754795241391505e-06, + "loss": 0.6626, + "step": 2515 + }, + { + "epoch": 0.8834269662921348, + "grad_norm": 0.871394693851471, + "learning_rate": 4.754595521269149e-06, + "loss": 0.694, + "step": 2516 + }, + { + "epoch": 0.8837780898876404, + "grad_norm": 0.9084363579750061, + "learning_rate": 4.754395724041556e-06, + "loss": 0.6725, + "step": 2517 + }, + { + "epoch": 0.8841292134831461, + "grad_norm": 0.8901112079620361, + "learning_rate": 4.754195849715557e-06, + "loss": 0.6614, + "step": 2518 + }, + { + "epoch": 0.8844803370786517, + "grad_norm": 0.8882237672805786, + "learning_rate": 4.7539958982979894e-06, + "loss": 0.6605, + "step": 2519 + }, + { + "epoch": 0.8848314606741573, + "grad_norm": 0.899763286113739, + "learning_rate": 4.753795869795691e-06, + "loss": 0.6423, + "step": 2520 + }, + { + "epoch": 0.8851825842696629, + "grad_norm": 0.9296014904975891, + "learning_rate": 4.753595764215503e-06, + "loss": 0.7166, + "step": 2521 + }, + { + "epoch": 0.8855337078651685, + "grad_norm": 0.883310854434967, + "learning_rate": 4.753395581564267e-06, + "loss": 0.6404, + "step": 2522 + }, + { + "epoch": 0.8858848314606742, + "grad_norm": 0.8889510035514832, + "learning_rate": 4.753195321848831e-06, + "loss": 0.649, + "step": 2523 + }, + { + "epoch": 0.8862359550561798, + "grad_norm": 0.9097694754600525, + "learning_rate": 4.752994985076044e-06, + "loss": 0.6604, + "step": 2524 + }, + { + "epoch": 0.8865870786516854, + "grad_norm": 0.8716392517089844, + "learning_rate": 4.752794571252755e-06, + "loss": 0.6989, + "step": 2525 + }, + { + "epoch": 0.886938202247191, + "grad_norm": 0.8778386116027832, + "learning_rate": 4.75259408038582e-06, + "loss": 0.638, + "step": 2526 + }, + { + "epoch": 0.8872893258426966, + "grad_norm": 0.8789075016975403, + "learning_rate": 4.752393512482095e-06, + "loss": 0.6713, + "step": 2527 + }, + { + "epoch": 0.8876404494382022, + "grad_norm": 0.8854068517684937, + "learning_rate": 4.752192867548439e-06, + "loss": 0.6534, + "step": 2528 + }, + { + "epoch": 0.8879915730337079, + "grad_norm": 0.8878395557403564, + "learning_rate": 4.7519921455917145e-06, + "loss": 0.684, + "step": 2529 + }, + { + "epoch": 0.8883426966292135, + "grad_norm": 0.8584818840026855, + "learning_rate": 4.751791346618786e-06, + "loss": 0.6259, + "step": 2530 + }, + { + "epoch": 0.8886938202247191, + "grad_norm": 0.8890157341957092, + "learning_rate": 4.75159047063652e-06, + "loss": 0.6607, + "step": 2531 + }, + { + "epoch": 0.8890449438202247, + "grad_norm": 0.851600706577301, + "learning_rate": 4.751389517651787e-06, + "loss": 0.66, + "step": 2532 + }, + { + "epoch": 0.8893960674157303, + "grad_norm": 0.8763266801834106, + "learning_rate": 4.751188487671457e-06, + "loss": 0.6427, + "step": 2533 + }, + { + "epoch": 0.889747191011236, + "grad_norm": 0.9055187702178955, + "learning_rate": 4.75098738070241e-06, + "loss": 0.6804, + "step": 2534 + }, + { + "epoch": 0.8900983146067416, + "grad_norm": 0.8860245943069458, + "learning_rate": 4.750786196751519e-06, + "loss": 0.6534, + "step": 2535 + }, + { + "epoch": 0.8904494382022472, + "grad_norm": 0.8951229453086853, + "learning_rate": 4.7505849358256665e-06, + "loss": 0.6637, + "step": 2536 + }, + { + "epoch": 0.8908005617977528, + "grad_norm": 0.9056205153465271, + "learning_rate": 4.7503835979317345e-06, + "loss": 0.6742, + "step": 2537 + }, + { + "epoch": 0.8911516853932584, + "grad_norm": 0.8903935551643372, + "learning_rate": 4.75018218307661e-06, + "loss": 0.6889, + "step": 2538 + }, + { + "epoch": 0.891502808988764, + "grad_norm": 0.9006560444831848, + "learning_rate": 4.74998069126718e-06, + "loss": 0.7042, + "step": 2539 + }, + { + "epoch": 0.8918539325842697, + "grad_norm": 0.8635199069976807, + "learning_rate": 4.749779122510337e-06, + "loss": 0.6692, + "step": 2540 + }, + { + "epoch": 0.8922050561797753, + "grad_norm": 0.8788262009620667, + "learning_rate": 4.749577476812971e-06, + "loss": 0.6652, + "step": 2541 + }, + { + "epoch": 0.8925561797752809, + "grad_norm": 0.8723106384277344, + "learning_rate": 4.749375754181981e-06, + "loss": 0.6373, + "step": 2542 + }, + { + "epoch": 0.8929073033707865, + "grad_norm": 0.8863840103149414, + "learning_rate": 4.749173954624265e-06, + "loss": 0.6701, + "step": 2543 + }, + { + "epoch": 0.8932584269662921, + "grad_norm": 0.9265256524085999, + "learning_rate": 4.748972078146725e-06, + "loss": 0.6591, + "step": 2544 + }, + { + "epoch": 0.8936095505617978, + "grad_norm": 0.9035506844520569, + "learning_rate": 4.748770124756265e-06, + "loss": 0.6585, + "step": 2545 + }, + { + "epoch": 0.8939606741573034, + "grad_norm": 0.8865588307380676, + "learning_rate": 4.748568094459789e-06, + "loss": 0.6903, + "step": 2546 + }, + { + "epoch": 0.894311797752809, + "grad_norm": 0.8840475678443909, + "learning_rate": 4.74836598726421e-06, + "loss": 0.6704, + "step": 2547 + }, + { + "epoch": 0.8946629213483146, + "grad_norm": 0.8937922120094299, + "learning_rate": 4.748163803176437e-06, + "loss": 0.6649, + "step": 2548 + }, + { + "epoch": 0.8950140449438202, + "grad_norm": 0.897407054901123, + "learning_rate": 4.747961542203387e-06, + "loss": 0.6705, + "step": 2549 + }, + { + "epoch": 0.8953651685393258, + "grad_norm": 0.8580518364906311, + "learning_rate": 4.747759204351975e-06, + "loss": 0.6217, + "step": 2550 + }, + { + "epoch": 0.8957162921348315, + "grad_norm": 0.8982558250427246, + "learning_rate": 4.747556789629121e-06, + "loss": 0.6873, + "step": 2551 + }, + { + "epoch": 0.8960674157303371, + "grad_norm": 0.922828197479248, + "learning_rate": 4.747354298041749e-06, + "loss": 0.6833, + "step": 2552 + }, + { + "epoch": 0.8964185393258427, + "grad_norm": 0.9047195315361023, + "learning_rate": 4.747151729596783e-06, + "loss": 0.6731, + "step": 2553 + }, + { + "epoch": 0.8967696629213483, + "grad_norm": 0.9132267236709595, + "learning_rate": 4.7469490843011515e-06, + "loss": 0.6752, + "step": 2554 + }, + { + "epoch": 0.8971207865168539, + "grad_norm": 0.8471392393112183, + "learning_rate": 4.7467463621617825e-06, + "loss": 0.6552, + "step": 2555 + }, + { + "epoch": 0.8974719101123596, + "grad_norm": 0.8814915418624878, + "learning_rate": 4.746543563185612e-06, + "loss": 0.676, + "step": 2556 + }, + { + "epoch": 0.8978230337078652, + "grad_norm": 0.8573102355003357, + "learning_rate": 4.746340687379573e-06, + "loss": 0.6397, + "step": 2557 + }, + { + "epoch": 0.8981741573033708, + "grad_norm": 0.8620328903198242, + "learning_rate": 4.746137734750604e-06, + "loss": 0.625, + "step": 2558 + }, + { + "epoch": 0.8985252808988764, + "grad_norm": 0.8743171691894531, + "learning_rate": 4.745934705305648e-06, + "loss": 0.6729, + "step": 2559 + }, + { + "epoch": 0.898876404494382, + "grad_norm": 0.838996410369873, + "learning_rate": 4.745731599051646e-06, + "loss": 0.6507, + "step": 2560 + }, + { + "epoch": 0.8992275280898876, + "grad_norm": 0.9150429964065552, + "learning_rate": 4.745528415995546e-06, + "loss": 0.6869, + "step": 2561 + }, + { + "epoch": 0.8995786516853933, + "grad_norm": 0.8738797903060913, + "learning_rate": 4.745325156144295e-06, + "loss": 0.6843, + "step": 2562 + }, + { + "epoch": 0.8999297752808989, + "grad_norm": 0.871146559715271, + "learning_rate": 4.745121819504845e-06, + "loss": 0.6773, + "step": 2563 + }, + { + "epoch": 0.9002808988764045, + "grad_norm": 0.8615120649337769, + "learning_rate": 4.74491840608415e-06, + "loss": 0.6383, + "step": 2564 + }, + { + "epoch": 0.9006320224719101, + "grad_norm": 0.8572788834571838, + "learning_rate": 4.744714915889167e-06, + "loss": 0.6488, + "step": 2565 + }, + { + "epoch": 0.9009831460674157, + "grad_norm": 0.8589732050895691, + "learning_rate": 4.744511348926855e-06, + "loss": 0.6464, + "step": 2566 + }, + { + "epoch": 0.9013342696629213, + "grad_norm": 0.8711311221122742, + "learning_rate": 4.744307705204174e-06, + "loss": 0.6809, + "step": 2567 + }, + { + "epoch": 0.901685393258427, + "grad_norm": 0.9025472402572632, + "learning_rate": 4.744103984728091e-06, + "loss": 0.6973, + "step": 2568 + }, + { + "epoch": 0.9020365168539326, + "grad_norm": 0.8648887872695923, + "learning_rate": 4.743900187505572e-06, + "loss": 0.6666, + "step": 2569 + }, + { + "epoch": 0.9023876404494382, + "grad_norm": 0.8922040462493896, + "learning_rate": 4.7436963135435864e-06, + "loss": 0.6875, + "step": 2570 + }, + { + "epoch": 0.9027387640449438, + "grad_norm": 0.8693889379501343, + "learning_rate": 4.743492362849107e-06, + "loss": 0.6926, + "step": 2571 + }, + { + "epoch": 0.9030898876404494, + "grad_norm": 0.8809205889701843, + "learning_rate": 4.743288335429109e-06, + "loss": 0.6841, + "step": 2572 + }, + { + "epoch": 0.9034410112359551, + "grad_norm": 0.8538480401039124, + "learning_rate": 4.743084231290569e-06, + "loss": 0.6416, + "step": 2573 + }, + { + "epoch": 0.9037921348314607, + "grad_norm": 0.8628817200660706, + "learning_rate": 4.742880050440468e-06, + "loss": 0.649, + "step": 2574 + }, + { + "epoch": 0.9041432584269663, + "grad_norm": 0.9055478572845459, + "learning_rate": 4.742675792885788e-06, + "loss": 0.6819, + "step": 2575 + }, + { + "epoch": 0.9044943820224719, + "grad_norm": 0.8601617217063904, + "learning_rate": 4.742471458633516e-06, + "loss": 0.6675, + "step": 2576 + }, + { + "epoch": 0.9048455056179775, + "grad_norm": 0.8635382056236267, + "learning_rate": 4.742267047690638e-06, + "loss": 0.6537, + "step": 2577 + }, + { + "epoch": 0.9051966292134831, + "grad_norm": 0.892162561416626, + "learning_rate": 4.742062560064146e-06, + "loss": 0.6627, + "step": 2578 + }, + { + "epoch": 0.9055477528089888, + "grad_norm": 0.9016808867454529, + "learning_rate": 4.741857995761033e-06, + "loss": 0.666, + "step": 2579 + }, + { + "epoch": 0.9058988764044944, + "grad_norm": 0.8793184161186218, + "learning_rate": 4.741653354788295e-06, + "loss": 0.638, + "step": 2580 + }, + { + "epoch": 0.90625, + "grad_norm": 0.8721537590026855, + "learning_rate": 4.741448637152931e-06, + "loss": 0.6816, + "step": 2581 + }, + { + "epoch": 0.9066011235955056, + "grad_norm": 0.9141138792037964, + "learning_rate": 4.741243842861941e-06, + "loss": 0.6959, + "step": 2582 + }, + { + "epoch": 0.9069522471910112, + "grad_norm": 0.8552782535552979, + "learning_rate": 4.741038971922329e-06, + "loss": 0.6868, + "step": 2583 + }, + { + "epoch": 0.9073033707865169, + "grad_norm": 0.8622416257858276, + "learning_rate": 4.740834024341103e-06, + "loss": 0.6861, + "step": 2584 + }, + { + "epoch": 0.9076544943820225, + "grad_norm": 0.8589463829994202, + "learning_rate": 4.74062900012527e-06, + "loss": 0.6525, + "step": 2585 + }, + { + "epoch": 0.9080056179775281, + "grad_norm": 0.9055203795433044, + "learning_rate": 4.740423899281843e-06, + "loss": 0.6724, + "step": 2586 + }, + { + "epoch": 0.9083567415730337, + "grad_norm": 0.9066650867462158, + "learning_rate": 4.740218721817836e-06, + "loss": 0.6929, + "step": 2587 + }, + { + "epoch": 0.9087078651685393, + "grad_norm": 0.8654912114143372, + "learning_rate": 4.7400134677402655e-06, + "loss": 0.6704, + "step": 2588 + }, + { + "epoch": 0.9090589887640449, + "grad_norm": 0.9301373958587646, + "learning_rate": 4.739808137056151e-06, + "loss": 0.6797, + "step": 2589 + }, + { + "epoch": 0.9094101123595506, + "grad_norm": 0.8802602291107178, + "learning_rate": 4.739602729772515e-06, + "loss": 0.6657, + "step": 2590 + }, + { + "epoch": 0.9097612359550562, + "grad_norm": 0.873859167098999, + "learning_rate": 4.739397245896382e-06, + "loss": 0.6364, + "step": 2591 + }, + { + "epoch": 0.9101123595505618, + "grad_norm": 0.8515645861625671, + "learning_rate": 4.739191685434779e-06, + "loss": 0.6472, + "step": 2592 + }, + { + "epoch": 0.9104634831460674, + "grad_norm": 0.8645114898681641, + "learning_rate": 4.738986048394737e-06, + "loss": 0.6792, + "step": 2593 + }, + { + "epoch": 0.910814606741573, + "grad_norm": 0.862397313117981, + "learning_rate": 4.738780334783288e-06, + "loss": 0.6522, + "step": 2594 + }, + { + "epoch": 0.9111657303370787, + "grad_norm": 0.8489444255828857, + "learning_rate": 4.738574544607467e-06, + "loss": 0.6757, + "step": 2595 + }, + { + "epoch": 0.9115168539325843, + "grad_norm": 0.8827484250068665, + "learning_rate": 4.738368677874313e-06, + "loss": 0.6842, + "step": 2596 + }, + { + "epoch": 0.9118679775280899, + "grad_norm": 0.8966019749641418, + "learning_rate": 4.738162734590865e-06, + "loss": 0.6658, + "step": 2597 + }, + { + "epoch": 0.9122191011235955, + "grad_norm": 0.8434582948684692, + "learning_rate": 4.7379567147641665e-06, + "loss": 0.6195, + "step": 2598 + }, + { + "epoch": 0.9125702247191011, + "grad_norm": 0.9089177250862122, + "learning_rate": 4.737750618401264e-06, + "loss": 0.7168, + "step": 2599 + }, + { + "epoch": 0.9129213483146067, + "grad_norm": 0.9108870625495911, + "learning_rate": 4.737544445509204e-06, + "loss": 0.6995, + "step": 2600 + }, + { + "epoch": 0.9132724719101124, + "grad_norm": 0.8936570882797241, + "learning_rate": 4.73733819609504e-06, + "loss": 0.6726, + "step": 2601 + }, + { + "epoch": 0.913623595505618, + "grad_norm": 0.8761941194534302, + "learning_rate": 4.737131870165824e-06, + "loss": 0.6642, + "step": 2602 + }, + { + "epoch": 0.9139747191011236, + "grad_norm": 0.870694100856781, + "learning_rate": 4.736925467728612e-06, + "loss": 0.6945, + "step": 2603 + }, + { + "epoch": 0.9143258426966292, + "grad_norm": 0.8931704759597778, + "learning_rate": 4.736718988790464e-06, + "loss": 0.6777, + "step": 2604 + }, + { + "epoch": 0.9146769662921348, + "grad_norm": 0.8875551223754883, + "learning_rate": 4.73651243335844e-06, + "loss": 0.6664, + "step": 2605 + }, + { + "epoch": 0.9150280898876404, + "grad_norm": 0.8898186087608337, + "learning_rate": 4.736305801439603e-06, + "loss": 0.6652, + "step": 2606 + }, + { + "epoch": 0.9153792134831461, + "grad_norm": 0.8893354535102844, + "learning_rate": 4.736099093041023e-06, + "loss": 0.6587, + "step": 2607 + }, + { + "epoch": 0.9157303370786517, + "grad_norm": 0.9278960824012756, + "learning_rate": 4.735892308169768e-06, + "loss": 0.6674, + "step": 2608 + }, + { + "epoch": 0.9160814606741573, + "grad_norm": 0.9112112522125244, + "learning_rate": 4.735685446832908e-06, + "loss": 0.6435, + "step": 2609 + }, + { + "epoch": 0.9164325842696629, + "grad_norm": 0.902930736541748, + "learning_rate": 4.735478509037519e-06, + "loss": 0.6402, + "step": 2610 + }, + { + "epoch": 0.9167837078651685, + "grad_norm": 0.8808385729789734, + "learning_rate": 4.7352714947906785e-06, + "loss": 0.707, + "step": 2611 + }, + { + "epoch": 0.9171348314606742, + "grad_norm": 0.8619163036346436, + "learning_rate": 4.735064404099465e-06, + "loss": 0.6728, + "step": 2612 + }, + { + "epoch": 0.9174859550561798, + "grad_norm": 0.8795300126075745, + "learning_rate": 4.734857236970962e-06, + "loss": 0.6543, + "step": 2613 + }, + { + "epoch": 0.9178370786516854, + "grad_norm": 0.9281381964683533, + "learning_rate": 4.734649993412254e-06, + "loss": 0.6594, + "step": 2614 + }, + { + "epoch": 0.918188202247191, + "grad_norm": 0.8637987971305847, + "learning_rate": 4.734442673430428e-06, + "loss": 0.6664, + "step": 2615 + }, + { + "epoch": 0.9185393258426966, + "grad_norm": 0.8663271069526672, + "learning_rate": 4.734235277032575e-06, + "loss": 0.6658, + "step": 2616 + }, + { + "epoch": 0.9188904494382022, + "grad_norm": 0.9028787016868591, + "learning_rate": 4.734027804225787e-06, + "loss": 0.6473, + "step": 2617 + }, + { + "epoch": 0.9192415730337079, + "grad_norm": 0.8826063275337219, + "learning_rate": 4.733820255017161e-06, + "loss": 0.626, + "step": 2618 + }, + { + "epoch": 0.9195926966292135, + "grad_norm": 0.8579033613204956, + "learning_rate": 4.7336126294137915e-06, + "loss": 0.6527, + "step": 2619 + }, + { + "epoch": 0.9199438202247191, + "grad_norm": 0.8992764353752136, + "learning_rate": 4.733404927422783e-06, + "loss": 0.6841, + "step": 2620 + }, + { + "epoch": 0.9202949438202247, + "grad_norm": 0.9202234148979187, + "learning_rate": 4.733197149051237e-06, + "loss": 0.6871, + "step": 2621 + }, + { + "epoch": 0.9206460674157303, + "grad_norm": 0.8881893754005432, + "learning_rate": 4.73298929430626e-06, + "loss": 0.6851, + "step": 2622 + }, + { + "epoch": 0.920997191011236, + "grad_norm": 0.9517401456832886, + "learning_rate": 4.732781363194958e-06, + "loss": 0.6748, + "step": 2623 + }, + { + "epoch": 0.9213483146067416, + "grad_norm": 0.9222083687782288, + "learning_rate": 4.7325733557244455e-06, + "loss": 0.6939, + "step": 2624 + }, + { + "epoch": 0.9216994382022472, + "grad_norm": 0.8639568090438843, + "learning_rate": 4.732365271901834e-06, + "loss": 0.664, + "step": 2625 + }, + { + "epoch": 0.9220505617977528, + "grad_norm": 0.9032568335533142, + "learning_rate": 4.732157111734241e-06, + "loss": 0.6779, + "step": 2626 + }, + { + "epoch": 0.9224016853932584, + "grad_norm": 0.9278046488761902, + "learning_rate": 4.731948875228784e-06, + "loss": 0.682, + "step": 2627 + }, + { + "epoch": 0.922752808988764, + "grad_norm": 0.9298899173736572, + "learning_rate": 4.731740562392587e-06, + "loss": 0.687, + "step": 2628 + }, + { + "epoch": 0.9231039325842697, + "grad_norm": 0.8798907995223999, + "learning_rate": 4.7315321732327715e-06, + "loss": 0.6722, + "step": 2629 + }, + { + "epoch": 0.9234550561797753, + "grad_norm": 0.8544273376464844, + "learning_rate": 4.731323707756465e-06, + "loss": 0.6801, + "step": 2630 + }, + { + "epoch": 0.9238061797752809, + "grad_norm": 0.8940860033035278, + "learning_rate": 4.731115165970797e-06, + "loss": 0.6548, + "step": 2631 + }, + { + "epoch": 0.9241573033707865, + "grad_norm": 0.9600421190261841, + "learning_rate": 4.730906547882901e-06, + "loss": 0.6935, + "step": 2632 + }, + { + "epoch": 0.9245084269662921, + "grad_norm": 0.9062276482582092, + "learning_rate": 4.730697853499908e-06, + "loss": 0.7043, + "step": 2633 + }, + { + "epoch": 0.9248595505617978, + "grad_norm": 0.8218638300895691, + "learning_rate": 4.730489082828959e-06, + "loss": 0.5993, + "step": 2634 + }, + { + "epoch": 0.9252106741573034, + "grad_norm": 0.8894577622413635, + "learning_rate": 4.7302802358771905e-06, + "loss": 0.6411, + "step": 2635 + }, + { + "epoch": 0.925561797752809, + "grad_norm": 0.881628155708313, + "learning_rate": 4.730071312651747e-06, + "loss": 0.6686, + "step": 2636 + }, + { + "epoch": 0.9259129213483146, + "grad_norm": 0.925878643989563, + "learning_rate": 4.7298623131597735e-06, + "loss": 0.6674, + "step": 2637 + }, + { + "epoch": 0.9262640449438202, + "grad_norm": 0.8464183211326599, + "learning_rate": 4.729653237408416e-06, + "loss": 0.6902, + "step": 2638 + }, + { + "epoch": 0.9266151685393258, + "grad_norm": 0.873918890953064, + "learning_rate": 4.729444085404826e-06, + "loss": 0.6737, + "step": 2639 + }, + { + "epoch": 0.9269662921348315, + "grad_norm": 0.8666852116584778, + "learning_rate": 4.7292348571561566e-06, + "loss": 0.6601, + "step": 2640 + }, + { + "epoch": 0.9273174157303371, + "grad_norm": 1.0832674503326416, + "learning_rate": 4.729025552669563e-06, + "loss": 0.654, + "step": 2641 + }, + { + "epoch": 0.9276685393258427, + "grad_norm": 0.8796826004981995, + "learning_rate": 4.728816171952202e-06, + "loss": 0.6365, + "step": 2642 + }, + { + "epoch": 0.9280196629213483, + "grad_norm": 0.8894056081771851, + "learning_rate": 4.728606715011235e-06, + "loss": 0.6955, + "step": 2643 + }, + { + "epoch": 0.9283707865168539, + "grad_norm": 0.9130917191505432, + "learning_rate": 4.728397181853826e-06, + "loss": 0.6807, + "step": 2644 + }, + { + "epoch": 0.9287219101123596, + "grad_norm": 0.8895679116249084, + "learning_rate": 4.72818757248714e-06, + "loss": 0.71, + "step": 2645 + }, + { + "epoch": 0.9290730337078652, + "grad_norm": 0.8796029090881348, + "learning_rate": 4.727977886918345e-06, + "loss": 0.666, + "step": 2646 + }, + { + "epoch": 0.9294241573033708, + "grad_norm": 0.8870053887367249, + "learning_rate": 4.727768125154614e-06, + "loss": 0.6724, + "step": 2647 + }, + { + "epoch": 0.9297752808988764, + "grad_norm": 0.8785473108291626, + "learning_rate": 4.727558287203119e-06, + "loss": 0.6735, + "step": 2648 + }, + { + "epoch": 0.930126404494382, + "grad_norm": 0.8722987174987793, + "learning_rate": 4.727348373071037e-06, + "loss": 0.6816, + "step": 2649 + }, + { + "epoch": 0.9304775280898876, + "grad_norm": 0.8995219469070435, + "learning_rate": 4.7271383827655455e-06, + "loss": 0.6657, + "step": 2650 + }, + { + "epoch": 0.9308286516853933, + "grad_norm": 0.8835882544517517, + "learning_rate": 4.726928316293828e-06, + "loss": 0.6973, + "step": 2651 + }, + { + "epoch": 0.9311797752808989, + "grad_norm": 0.8654088973999023, + "learning_rate": 4.726718173663068e-06, + "loss": 0.6083, + "step": 2652 + }, + { + "epoch": 0.9315308988764045, + "grad_norm": 0.832149088382721, + "learning_rate": 4.726507954880451e-06, + "loss": 0.5915, + "step": 2653 + }, + { + "epoch": 0.9318820224719101, + "grad_norm": 0.8820677995681763, + "learning_rate": 4.726297659953169e-06, + "loss": 0.641, + "step": 2654 + }, + { + "epoch": 0.9322331460674157, + "grad_norm": 0.9029566049575806, + "learning_rate": 4.7260872888884105e-06, + "loss": 0.6845, + "step": 2655 + }, + { + "epoch": 0.9325842696629213, + "grad_norm": 0.8395662307739258, + "learning_rate": 4.725876841693372e-06, + "loss": 0.6566, + "step": 2656 + }, + { + "epoch": 0.932935393258427, + "grad_norm": 0.894307017326355, + "learning_rate": 4.7256663183752495e-06, + "loss": 0.6985, + "step": 2657 + }, + { + "epoch": 0.9332865168539326, + "grad_norm": 0.9100427031517029, + "learning_rate": 4.725455718941245e-06, + "loss": 0.6687, + "step": 2658 + }, + { + "epoch": 0.9336376404494382, + "grad_norm": 0.874563455581665, + "learning_rate": 4.725245043398558e-06, + "loss": 0.6692, + "step": 2659 + }, + { + "epoch": 0.9339887640449438, + "grad_norm": 0.8920844793319702, + "learning_rate": 4.725034291754394e-06, + "loss": 0.6334, + "step": 2660 + }, + { + "epoch": 0.9343398876404494, + "grad_norm": 0.8333306908607483, + "learning_rate": 4.724823464015963e-06, + "loss": 0.6322, + "step": 2661 + }, + { + "epoch": 0.9346910112359551, + "grad_norm": 0.8971813917160034, + "learning_rate": 4.724612560190473e-06, + "loss": 0.6823, + "step": 2662 + }, + { + "epoch": 0.9350421348314607, + "grad_norm": 0.8798514008522034, + "learning_rate": 4.724401580285136e-06, + "loss": 0.6742, + "step": 2663 + }, + { + "epoch": 0.9353932584269663, + "grad_norm": 0.827732503414154, + "learning_rate": 4.724190524307169e-06, + "loss": 0.6574, + "step": 2664 + }, + { + "epoch": 0.9357443820224719, + "grad_norm": 0.8697452545166016, + "learning_rate": 4.7239793922637896e-06, + "loss": 0.637, + "step": 2665 + }, + { + "epoch": 0.9360955056179775, + "grad_norm": 0.887448251247406, + "learning_rate": 4.723768184162217e-06, + "loss": 0.6762, + "step": 2666 + }, + { + "epoch": 0.9364466292134831, + "grad_norm": 0.9078186750411987, + "learning_rate": 4.723556900009677e-06, + "loss": 0.6832, + "step": 2667 + }, + { + "epoch": 0.9367977528089888, + "grad_norm": 0.8617350459098816, + "learning_rate": 4.723345539813393e-06, + "loss": 0.6591, + "step": 2668 + }, + { + "epoch": 0.9371488764044944, + "grad_norm": 0.8676525354385376, + "learning_rate": 4.723134103580594e-06, + "loss": 0.6941, + "step": 2669 + }, + { + "epoch": 0.9375, + "grad_norm": 0.8551449179649353, + "learning_rate": 4.722922591318511e-06, + "loss": 0.6142, + "step": 2670 + }, + { + "epoch": 0.9378511235955056, + "grad_norm": 0.9027611017227173, + "learning_rate": 4.722711003034378e-06, + "loss": 0.6893, + "step": 2671 + }, + { + "epoch": 0.9382022471910112, + "grad_norm": 0.8602604269981384, + "learning_rate": 4.7224993387354305e-06, + "loss": 0.644, + "step": 2672 + }, + { + "epoch": 0.9385533707865169, + "grad_norm": 0.9275681376457214, + "learning_rate": 4.722287598428907e-06, + "loss": 0.676, + "step": 2673 + }, + { + "epoch": 0.9389044943820225, + "grad_norm": 0.880574107170105, + "learning_rate": 4.7220757821220495e-06, + "loss": 0.6068, + "step": 2674 + }, + { + "epoch": 0.9392556179775281, + "grad_norm": 0.8658181428909302, + "learning_rate": 4.721863889822102e-06, + "loss": 0.6714, + "step": 2675 + }, + { + "epoch": 0.9396067415730337, + "grad_norm": 0.890250563621521, + "learning_rate": 4.7216519215363095e-06, + "loss": 0.6624, + "step": 2676 + }, + { + "epoch": 0.9399578651685393, + "grad_norm": 0.870218813419342, + "learning_rate": 4.721439877271924e-06, + "loss": 0.6692, + "step": 2677 + }, + { + "epoch": 0.9403089887640449, + "grad_norm": 0.8716661334037781, + "learning_rate": 4.721227757036195e-06, + "loss": 0.6513, + "step": 2678 + }, + { + "epoch": 0.9406601123595506, + "grad_norm": 0.8979821801185608, + "learning_rate": 4.721015560836377e-06, + "loss": 0.6658, + "step": 2679 + }, + { + "epoch": 0.9410112359550562, + "grad_norm": 1.3192576169967651, + "learning_rate": 4.7208032886797285e-06, + "loss": 0.6726, + "step": 2680 + }, + { + "epoch": 0.9413623595505618, + "grad_norm": 0.9058758020401001, + "learning_rate": 4.720590940573507e-06, + "loss": 0.6764, + "step": 2681 + }, + { + "epoch": 0.9417134831460674, + "grad_norm": 0.8641902208328247, + "learning_rate": 4.720378516524976e-06, + "loss": 0.7075, + "step": 2682 + }, + { + "epoch": 0.942064606741573, + "grad_norm": 0.8593077659606934, + "learning_rate": 4.7201660165413994e-06, + "loss": 0.6575, + "step": 2683 + }, + { + "epoch": 0.9424157303370787, + "grad_norm": 0.8111205697059631, + "learning_rate": 4.719953440630045e-06, + "loss": 0.5865, + "step": 2684 + }, + { + "epoch": 0.9427668539325843, + "grad_norm": 0.8845921754837036, + "learning_rate": 4.719740788798182e-06, + "loss": 0.6787, + "step": 2685 + }, + { + "epoch": 0.9431179775280899, + "grad_norm": 0.893483579158783, + "learning_rate": 4.719528061053084e-06, + "loss": 0.657, + "step": 2686 + }, + { + "epoch": 0.9434691011235955, + "grad_norm": 0.882802426815033, + "learning_rate": 4.719315257402026e-06, + "loss": 0.6678, + "step": 2687 + }, + { + "epoch": 0.9438202247191011, + "grad_norm": 0.8567876815795898, + "learning_rate": 4.7191023778522844e-06, + "loss": 0.6167, + "step": 2688 + }, + { + "epoch": 0.9441713483146067, + "grad_norm": 0.8992242217063904, + "learning_rate": 4.718889422411141e-06, + "loss": 0.6691, + "step": 2689 + }, + { + "epoch": 0.9445224719101124, + "grad_norm": 0.8735215663909912, + "learning_rate": 4.718676391085878e-06, + "loss": 0.63, + "step": 2690 + }, + { + "epoch": 0.944873595505618, + "grad_norm": 0.8774729371070862, + "learning_rate": 4.71846328388378e-06, + "loss": 0.6702, + "step": 2691 + }, + { + "epoch": 0.9452247191011236, + "grad_norm": 0.9470273852348328, + "learning_rate": 4.718250100812138e-06, + "loss": 0.6997, + "step": 2692 + }, + { + "epoch": 0.9455758426966292, + "grad_norm": 0.9050356149673462, + "learning_rate": 4.71803684187824e-06, + "loss": 0.6451, + "step": 2693 + }, + { + "epoch": 0.9459269662921348, + "grad_norm": 0.9217947125434875, + "learning_rate": 4.7178235070893796e-06, + "loss": 0.6556, + "step": 2694 + }, + { + "epoch": 0.9462780898876404, + "grad_norm": 0.909965991973877, + "learning_rate": 4.717610096452852e-06, + "loss": 0.6772, + "step": 2695 + }, + { + "epoch": 0.9466292134831461, + "grad_norm": 0.8807794451713562, + "learning_rate": 4.717396609975959e-06, + "loss": 0.6805, + "step": 2696 + }, + { + "epoch": 0.9469803370786517, + "grad_norm": 0.9008699059486389, + "learning_rate": 4.717183047665998e-06, + "loss": 0.6718, + "step": 2697 + }, + { + "epoch": 0.9473314606741573, + "grad_norm": 0.9289125204086304, + "learning_rate": 4.716969409530274e-06, + "loss": 0.6705, + "step": 2698 + }, + { + "epoch": 0.9476825842696629, + "grad_norm": 0.9074335098266602, + "learning_rate": 4.716755695576094e-06, + "loss": 0.6897, + "step": 2699 + }, + { + "epoch": 0.9480337078651685, + "grad_norm": 0.9094159603118896, + "learning_rate": 4.716541905810766e-06, + "loss": 0.6625, + "step": 2700 + }, + { + "epoch": 0.9483848314606742, + "grad_norm": 0.8951701521873474, + "learning_rate": 4.716328040241601e-06, + "loss": 0.6234, + "step": 2701 + }, + { + "epoch": 0.9487359550561798, + "grad_norm": 0.8777403831481934, + "learning_rate": 4.716114098875914e-06, + "loss": 0.6418, + "step": 2702 + }, + { + "epoch": 0.9490870786516854, + "grad_norm": 0.9449620842933655, + "learning_rate": 4.715900081721021e-06, + "loss": 0.6814, + "step": 2703 + }, + { + "epoch": 0.949438202247191, + "grad_norm": 0.8588343262672424, + "learning_rate": 4.7156859887842416e-06, + "loss": 0.6042, + "step": 2704 + }, + { + "epoch": 0.9497893258426966, + "grad_norm": 0.8675066232681274, + "learning_rate": 4.715471820072896e-06, + "loss": 0.6591, + "step": 2705 + }, + { + "epoch": 0.9501404494382022, + "grad_norm": 0.8843618631362915, + "learning_rate": 4.715257575594312e-06, + "loss": 0.6594, + "step": 2706 + }, + { + "epoch": 0.9504915730337079, + "grad_norm": 0.8740993142127991, + "learning_rate": 4.715043255355814e-06, + "loss": 0.6578, + "step": 2707 + }, + { + "epoch": 0.9508426966292135, + "grad_norm": 0.8732353448867798, + "learning_rate": 4.7148288593647315e-06, + "loss": 0.6399, + "step": 2708 + }, + { + "epoch": 0.9511938202247191, + "grad_norm": 0.8841140866279602, + "learning_rate": 4.714614387628398e-06, + "loss": 0.6879, + "step": 2709 + }, + { + "epoch": 0.9515449438202247, + "grad_norm": 0.9016351103782654, + "learning_rate": 4.714399840154147e-06, + "loss": 0.6618, + "step": 2710 + }, + { + "epoch": 0.9518960674157303, + "grad_norm": 0.8737030625343323, + "learning_rate": 4.714185216949317e-06, + "loss": 0.6616, + "step": 2711 + }, + { + "epoch": 0.952247191011236, + "grad_norm": 0.8543287515640259, + "learning_rate": 4.713970518021246e-06, + "loss": 0.627, + "step": 2712 + }, + { + "epoch": 0.9525983146067416, + "grad_norm": 0.8959774374961853, + "learning_rate": 4.713755743377278e-06, + "loss": 0.6936, + "step": 2713 + }, + { + "epoch": 0.9529494382022472, + "grad_norm": 0.9039850831031799, + "learning_rate": 4.713540893024759e-06, + "loss": 0.6465, + "step": 2714 + }, + { + "epoch": 0.9533005617977528, + "grad_norm": 0.8787552118301392, + "learning_rate": 4.713325966971035e-06, + "loss": 0.6127, + "step": 2715 + }, + { + "epoch": 0.9536516853932584, + "grad_norm": 0.8473005294799805, + "learning_rate": 4.713110965223457e-06, + "loss": 0.6648, + "step": 2716 + }, + { + "epoch": 0.954002808988764, + "grad_norm": 0.937426745891571, + "learning_rate": 4.712895887789377e-06, + "loss": 0.6594, + "step": 2717 + }, + { + "epoch": 0.9543539325842697, + "grad_norm": 0.8719070553779602, + "learning_rate": 4.712680734676153e-06, + "loss": 0.6652, + "step": 2718 + }, + { + "epoch": 0.9547050561797753, + "grad_norm": 0.8872880935668945, + "learning_rate": 4.7124655058911395e-06, + "loss": 0.6987, + "step": 2719 + }, + { + "epoch": 0.9550561797752809, + "grad_norm": 0.8662359714508057, + "learning_rate": 4.7122502014417e-06, + "loss": 0.6723, + "step": 2720 + }, + { + "epoch": 0.9554073033707865, + "grad_norm": 0.8790738582611084, + "learning_rate": 4.712034821335196e-06, + "loss": 0.6421, + "step": 2721 + }, + { + "epoch": 0.9557584269662921, + "grad_norm": 0.8946492671966553, + "learning_rate": 4.711819365578995e-06, + "loss": 0.6439, + "step": 2722 + }, + { + "epoch": 0.9561095505617978, + "grad_norm": 0.9047174453735352, + "learning_rate": 4.711603834180464e-06, + "loss": 0.6859, + "step": 2723 + }, + { + "epoch": 0.9564606741573034, + "grad_norm": 0.9128047823905945, + "learning_rate": 4.711388227146974e-06, + "loss": 0.671, + "step": 2724 + }, + { + "epoch": 0.956811797752809, + "grad_norm": 0.8675251007080078, + "learning_rate": 4.711172544485898e-06, + "loss": 0.672, + "step": 2725 + }, + { + "epoch": 0.9571629213483146, + "grad_norm": 0.8741472363471985, + "learning_rate": 4.710956786204614e-06, + "loss": 0.6614, + "step": 2726 + }, + { + "epoch": 0.9575140449438202, + "grad_norm": 0.851613461971283, + "learning_rate": 4.710740952310501e-06, + "loss": 0.645, + "step": 2727 + }, + { + "epoch": 0.9578651685393258, + "grad_norm": 0.9370160698890686, + "learning_rate": 4.7105250428109375e-06, + "loss": 0.6743, + "step": 2728 + }, + { + "epoch": 0.9582162921348315, + "grad_norm": 0.9051042199134827, + "learning_rate": 4.71030905771331e-06, + "loss": 0.6166, + "step": 2729 + }, + { + "epoch": 0.9585674157303371, + "grad_norm": 0.8878385424613953, + "learning_rate": 4.7100929970250035e-06, + "loss": 0.6447, + "step": 2730 + }, + { + "epoch": 0.9589185393258427, + "grad_norm": 0.857430100440979, + "learning_rate": 4.709876860753409e-06, + "loss": 0.6557, + "step": 2731 + }, + { + "epoch": 0.9592696629213483, + "grad_norm": 0.8934675455093384, + "learning_rate": 4.709660648905916e-06, + "loss": 0.675, + "step": 2732 + }, + { + "epoch": 0.9596207865168539, + "grad_norm": 0.8855241537094116, + "learning_rate": 4.709444361489918e-06, + "loss": 0.6705, + "step": 2733 + }, + { + "epoch": 0.9599719101123596, + "grad_norm": 0.8473418354988098, + "learning_rate": 4.7092279985128165e-06, + "loss": 0.6398, + "step": 2734 + }, + { + "epoch": 0.9603230337078652, + "grad_norm": 0.9071454405784607, + "learning_rate": 4.709011559982006e-06, + "loss": 0.6824, + "step": 2735 + }, + { + "epoch": 0.9606741573033708, + "grad_norm": 0.8947198390960693, + "learning_rate": 4.708795045904891e-06, + "loss": 0.6794, + "step": 2736 + }, + { + "epoch": 0.9610252808988764, + "grad_norm": 0.8824878931045532, + "learning_rate": 4.708578456288875e-06, + "loss": 0.6453, + "step": 2737 + }, + { + "epoch": 0.961376404494382, + "grad_norm": 0.8897850513458252, + "learning_rate": 4.708361791141367e-06, + "loss": 0.6708, + "step": 2738 + }, + { + "epoch": 0.9617275280898876, + "grad_norm": 0.8753004670143127, + "learning_rate": 4.708145050469774e-06, + "loss": 0.6574, + "step": 2739 + }, + { + "epoch": 0.9620786516853933, + "grad_norm": 0.8702830672264099, + "learning_rate": 4.7079282342815105e-06, + "loss": 0.6363, + "step": 2740 + }, + { + "epoch": 0.9624297752808989, + "grad_norm": 0.8557823896408081, + "learning_rate": 4.707711342583991e-06, + "loss": 0.6607, + "step": 2741 + }, + { + "epoch": 0.9627808988764045, + "grad_norm": 0.9189360737800598, + "learning_rate": 4.7074943753846324e-06, + "loss": 0.6732, + "step": 2742 + }, + { + "epoch": 0.9631320224719101, + "grad_norm": 0.8712623715400696, + "learning_rate": 4.707277332690855e-06, + "loss": 0.6962, + "step": 2743 + }, + { + "epoch": 0.9634831460674157, + "grad_norm": 0.8758993744850159, + "learning_rate": 4.707060214510082e-06, + "loss": 0.688, + "step": 2744 + }, + { + "epoch": 0.9638342696629213, + "grad_norm": 0.9074702262878418, + "learning_rate": 4.706843020849738e-06, + "loss": 0.6668, + "step": 2745 + }, + { + "epoch": 0.964185393258427, + "grad_norm": 0.8798408508300781, + "learning_rate": 4.706625751717251e-06, + "loss": 0.6728, + "step": 2746 + }, + { + "epoch": 0.9645365168539326, + "grad_norm": 0.8691076636314392, + "learning_rate": 4.706408407120053e-06, + "loss": 0.6217, + "step": 2747 + }, + { + "epoch": 0.9648876404494382, + "grad_norm": 0.8737530708312988, + "learning_rate": 4.7061909870655734e-06, + "loss": 0.6618, + "step": 2748 + }, + { + "epoch": 0.9652387640449438, + "grad_norm": 0.893994152545929, + "learning_rate": 4.70597349156125e-06, + "loss": 0.6899, + "step": 2749 + }, + { + "epoch": 0.9655898876404494, + "grad_norm": 0.8533192276954651, + "learning_rate": 4.705755920614522e-06, + "loss": 0.601, + "step": 2750 + }, + { + "epoch": 0.9659410112359551, + "grad_norm": 0.8612004518508911, + "learning_rate": 4.705538274232829e-06, + "loss": 0.6407, + "step": 2751 + }, + { + "epoch": 0.9662921348314607, + "grad_norm": 0.8780084848403931, + "learning_rate": 4.7053205524236136e-06, + "loss": 0.6689, + "step": 2752 + }, + { + "epoch": 0.9666432584269663, + "grad_norm": 0.8959993124008179, + "learning_rate": 4.705102755194323e-06, + "loss": 0.6588, + "step": 2753 + }, + { + "epoch": 0.9669943820224719, + "grad_norm": 0.9221241474151611, + "learning_rate": 4.704884882552405e-06, + "loss": 0.6454, + "step": 2754 + }, + { + "epoch": 0.9673455056179775, + "grad_norm": 0.8571012020111084, + "learning_rate": 4.70466693450531e-06, + "loss": 0.6101, + "step": 2755 + }, + { + "epoch": 0.9676966292134831, + "grad_norm": 0.9270880222320557, + "learning_rate": 4.704448911060493e-06, + "loss": 0.6561, + "step": 2756 + }, + { + "epoch": 0.9680477528089888, + "grad_norm": 0.8872186541557312, + "learning_rate": 4.70423081222541e-06, + "loss": 0.6704, + "step": 2757 + }, + { + "epoch": 0.9683988764044944, + "grad_norm": 0.8734565377235413, + "learning_rate": 4.704012638007519e-06, + "loss": 0.6626, + "step": 2758 + }, + { + "epoch": 0.96875, + "grad_norm": 0.9009286761283875, + "learning_rate": 4.703794388414281e-06, + "loss": 0.6727, + "step": 2759 + }, + { + "epoch": 0.9691011235955056, + "grad_norm": 0.8499953746795654, + "learning_rate": 4.703576063453162e-06, + "loss": 0.6574, + "step": 2760 + }, + { + "epoch": 0.9694522471910112, + "grad_norm": 0.8824490308761597, + "learning_rate": 4.703357663131627e-06, + "loss": 0.6618, + "step": 2761 + }, + { + "epoch": 0.9698033707865169, + "grad_norm": 0.8579809665679932, + "learning_rate": 4.703139187457145e-06, + "loss": 0.6677, + "step": 2762 + }, + { + "epoch": 0.9701544943820225, + "grad_norm": 0.8739168643951416, + "learning_rate": 4.702920636437188e-06, + "loss": 0.6648, + "step": 2763 + }, + { + "epoch": 0.9705056179775281, + "grad_norm": 0.8557081818580627, + "learning_rate": 4.702702010079231e-06, + "loss": 0.6588, + "step": 2764 + }, + { + "epoch": 0.9708567415730337, + "grad_norm": 0.8667842149734497, + "learning_rate": 4.702483308390749e-06, + "loss": 0.6432, + "step": 2765 + }, + { + "epoch": 0.9712078651685393, + "grad_norm": 0.8367511034011841, + "learning_rate": 4.7022645313792235e-06, + "loss": 0.6428, + "step": 2766 + }, + { + "epoch": 0.9715589887640449, + "grad_norm": 0.8607786297798157, + "learning_rate": 4.702045679052136e-06, + "loss": 0.628, + "step": 2767 + }, + { + "epoch": 0.9719101123595506, + "grad_norm": 0.8698323369026184, + "learning_rate": 4.701826751416969e-06, + "loss": 0.6241, + "step": 2768 + }, + { + "epoch": 0.9722612359550562, + "grad_norm": 0.8620527386665344, + "learning_rate": 4.701607748481213e-06, + "loss": 0.6698, + "step": 2769 + }, + { + "epoch": 0.9726123595505618, + "grad_norm": 0.8832960724830627, + "learning_rate": 4.701388670252355e-06, + "loss": 0.6332, + "step": 2770 + }, + { + "epoch": 0.9729634831460674, + "grad_norm": 0.8846637606620789, + "learning_rate": 4.701169516737888e-06, + "loss": 0.6968, + "step": 2771 + }, + { + "epoch": 0.973314606741573, + "grad_norm": 0.8862726092338562, + "learning_rate": 4.7009502879453076e-06, + "loss": 0.7282, + "step": 2772 + }, + { + "epoch": 0.9736657303370787, + "grad_norm": 0.8758227229118347, + "learning_rate": 4.70073098388211e-06, + "loss": 0.658, + "step": 2773 + }, + { + "epoch": 0.9740168539325843, + "grad_norm": 0.8702247738838196, + "learning_rate": 4.700511604555796e-06, + "loss": 0.65, + "step": 2774 + }, + { + "epoch": 0.9743679775280899, + "grad_norm": 0.8507943153381348, + "learning_rate": 4.700292149973869e-06, + "loss": 0.6625, + "step": 2775 + }, + { + "epoch": 0.9747191011235955, + "grad_norm": 0.8701155185699463, + "learning_rate": 4.700072620143833e-06, + "loss": 0.6742, + "step": 2776 + }, + { + "epoch": 0.9750702247191011, + "grad_norm": 0.8407930135726929, + "learning_rate": 4.6998530150731955e-06, + "loss": 0.642, + "step": 2777 + }, + { + "epoch": 0.9754213483146067, + "grad_norm": 0.858070969581604, + "learning_rate": 4.6996333347694675e-06, + "loss": 0.6752, + "step": 2778 + }, + { + "epoch": 0.9757724719101124, + "grad_norm": 0.8673107624053955, + "learning_rate": 4.6994135792401615e-06, + "loss": 0.6826, + "step": 2779 + }, + { + "epoch": 0.976123595505618, + "grad_norm": 0.8591476082801819, + "learning_rate": 4.699193748492794e-06, + "loss": 0.6972, + "step": 2780 + }, + { + "epoch": 0.9764747191011236, + "grad_norm": 0.8738571405410767, + "learning_rate": 4.698973842534882e-06, + "loss": 0.6559, + "step": 2781 + }, + { + "epoch": 0.9768258426966292, + "grad_norm": 0.8602177500724792, + "learning_rate": 4.698753861373946e-06, + "loss": 0.6433, + "step": 2782 + }, + { + "epoch": 0.9771769662921348, + "grad_norm": 0.8604179620742798, + "learning_rate": 4.6985338050175096e-06, + "loss": 0.6499, + "step": 2783 + }, + { + "epoch": 0.9775280898876404, + "grad_norm": 0.8430880904197693, + "learning_rate": 4.698313673473098e-06, + "loss": 0.6008, + "step": 2784 + }, + { + "epoch": 0.9778792134831461, + "grad_norm": 0.8904216885566711, + "learning_rate": 4.69809346674824e-06, + "loss": 0.6869, + "step": 2785 + }, + { + "epoch": 0.9782303370786517, + "grad_norm": 0.8424625396728516, + "learning_rate": 4.6978731848504665e-06, + "loss": 0.6519, + "step": 2786 + }, + { + "epoch": 0.9785814606741573, + "grad_norm": 0.8511930704116821, + "learning_rate": 4.697652827787311e-06, + "loss": 0.6511, + "step": 2787 + }, + { + "epoch": 0.9789325842696629, + "grad_norm": 0.8731825351715088, + "learning_rate": 4.697432395566309e-06, + "loss": 0.6756, + "step": 2788 + }, + { + "epoch": 0.9792837078651685, + "grad_norm": 0.8926572203636169, + "learning_rate": 4.697211888194999e-06, + "loss": 0.6457, + "step": 2789 + }, + { + "epoch": 0.9796348314606742, + "grad_norm": 0.8796569108963013, + "learning_rate": 4.696991305680923e-06, + "loss": 0.6435, + "step": 2790 + }, + { + "epoch": 0.9799859550561798, + "grad_norm": 0.89739990234375, + "learning_rate": 4.6967706480316256e-06, + "loss": 0.666, + "step": 2791 + }, + { + "epoch": 0.9803370786516854, + "grad_norm": 0.8666273951530457, + "learning_rate": 4.69654991525465e-06, + "loss": 0.6658, + "step": 2792 + }, + { + "epoch": 0.980688202247191, + "grad_norm": 0.8648870587348938, + "learning_rate": 4.696329107357548e-06, + "loss": 0.6541, + "step": 2793 + }, + { + "epoch": 0.9810393258426966, + "grad_norm": 0.8609452247619629, + "learning_rate": 4.69610822434787e-06, + "loss": 0.6646, + "step": 2794 + }, + { + "epoch": 0.9813904494382022, + "grad_norm": 0.8994244933128357, + "learning_rate": 4.695887266233169e-06, + "loss": 0.6778, + "step": 2795 + }, + { + "epoch": 0.9817415730337079, + "grad_norm": 0.8689225316047668, + "learning_rate": 4.6956662330210035e-06, + "loss": 0.6581, + "step": 2796 + }, + { + "epoch": 0.9820926966292135, + "grad_norm": 0.8597514629364014, + "learning_rate": 4.695445124718932e-06, + "loss": 0.6322, + "step": 2797 + }, + { + "epoch": 0.9824438202247191, + "grad_norm": 0.8898642063140869, + "learning_rate": 4.695223941334514e-06, + "loss": 0.6866, + "step": 2798 + }, + { + "epoch": 0.9827949438202247, + "grad_norm": 0.9043296575546265, + "learning_rate": 4.695002682875317e-06, + "loss": 0.686, + "step": 2799 + }, + { + "epoch": 0.9831460674157303, + "grad_norm": 0.8787916898727417, + "learning_rate": 4.694781349348907e-06, + "loss": 0.6819, + "step": 2800 + }, + { + "epoch": 0.983497191011236, + "grad_norm": 0.86040198802948, + "learning_rate": 4.694559940762853e-06, + "loss": 0.6591, + "step": 2801 + }, + { + "epoch": 0.9838483146067416, + "grad_norm": 0.8670945167541504, + "learning_rate": 4.694338457124726e-06, + "loss": 0.6828, + "step": 2802 + }, + { + "epoch": 0.9841994382022472, + "grad_norm": 0.8818464875221252, + "learning_rate": 4.694116898442102e-06, + "loss": 0.6556, + "step": 2803 + }, + { + "epoch": 0.9845505617977528, + "grad_norm": 0.850309431552887, + "learning_rate": 4.693895264722557e-06, + "loss": 0.6803, + "step": 2804 + }, + { + "epoch": 0.9849016853932584, + "grad_norm": 0.8555431365966797, + "learning_rate": 4.6936735559736715e-06, + "loss": 0.6484, + "step": 2805 + }, + { + "epoch": 0.985252808988764, + "grad_norm": 0.8949291110038757, + "learning_rate": 4.693451772203028e-06, + "loss": 0.6889, + "step": 2806 + }, + { + "epoch": 0.9856039325842697, + "grad_norm": 0.8468831181526184, + "learning_rate": 4.6932299134182104e-06, + "loss": 0.6398, + "step": 2807 + }, + { + "epoch": 0.9859550561797753, + "grad_norm": 0.8642711043357849, + "learning_rate": 4.693007979626807e-06, + "loss": 0.6789, + "step": 2808 + }, + { + "epoch": 0.9863061797752809, + "grad_norm": 0.8652757406234741, + "learning_rate": 4.692785970836407e-06, + "loss": 0.6862, + "step": 2809 + }, + { + "epoch": 0.9866573033707865, + "grad_norm": 0.8652501106262207, + "learning_rate": 4.692563887054603e-06, + "loss": 0.647, + "step": 2810 + }, + { + "epoch": 0.9870084269662921, + "grad_norm": 0.8379038572311401, + "learning_rate": 4.69234172828899e-06, + "loss": 0.616, + "step": 2811 + }, + { + "epoch": 0.9873595505617978, + "grad_norm": 0.8986461758613586, + "learning_rate": 4.692119494547167e-06, + "loss": 0.6746, + "step": 2812 + }, + { + "epoch": 0.9877106741573034, + "grad_norm": 0.8548451662063599, + "learning_rate": 4.691897185836733e-06, + "loss": 0.6607, + "step": 2813 + }, + { + "epoch": 0.988061797752809, + "grad_norm": 0.8519273400306702, + "learning_rate": 4.69167480216529e-06, + "loss": 0.6341, + "step": 2814 + }, + { + "epoch": 0.9884129213483146, + "grad_norm": 0.8934301137924194, + "learning_rate": 4.691452343540445e-06, + "loss": 0.6601, + "step": 2815 + }, + { + "epoch": 0.9887640449438202, + "grad_norm": 0.8649933338165283, + "learning_rate": 4.6912298099698046e-06, + "loss": 0.6406, + "step": 2816 + }, + { + "epoch": 0.9891151685393258, + "grad_norm": 0.8976009488105774, + "learning_rate": 4.69100720146098e-06, + "loss": 0.6391, + "step": 2817 + }, + { + "epoch": 0.9894662921348315, + "grad_norm": 0.8619791865348816, + "learning_rate": 4.690784518021584e-06, + "loss": 0.6932, + "step": 2818 + }, + { + "epoch": 0.9898174157303371, + "grad_norm": 0.8594460487365723, + "learning_rate": 4.690561759659232e-06, + "loss": 0.6666, + "step": 2819 + }, + { + "epoch": 0.9901685393258427, + "grad_norm": 0.8755744099617004, + "learning_rate": 4.690338926381542e-06, + "loss": 0.6458, + "step": 2820 + }, + { + "epoch": 0.9905196629213483, + "grad_norm": 0.8844261765480042, + "learning_rate": 4.690116018196135e-06, + "loss": 0.6344, + "step": 2821 + }, + { + "epoch": 0.9908707865168539, + "grad_norm": 0.871732234954834, + "learning_rate": 4.689893035110635e-06, + "loss": 0.6316, + "step": 2822 + }, + { + "epoch": 0.9912219101123596, + "grad_norm": 0.8930003046989441, + "learning_rate": 4.689669977132667e-06, + "loss": 0.6887, + "step": 2823 + }, + { + "epoch": 0.9915730337078652, + "grad_norm": 0.9592486619949341, + "learning_rate": 4.689446844269859e-06, + "loss": 0.653, + "step": 2824 + }, + { + "epoch": 0.9919241573033708, + "grad_norm": 0.8865694999694824, + "learning_rate": 4.6892236365298425e-06, + "loss": 0.648, + "step": 2825 + }, + { + "epoch": 0.9922752808988764, + "grad_norm": 0.8786531686782837, + "learning_rate": 4.689000353920251e-06, + "loss": 0.647, + "step": 2826 + }, + { + "epoch": 0.992626404494382, + "grad_norm": 0.8980904817581177, + "learning_rate": 4.6887769964487215e-06, + "loss": 0.6556, + "step": 2827 + }, + { + "epoch": 0.9929775280898876, + "grad_norm": 0.8920478224754333, + "learning_rate": 4.688553564122891e-06, + "loss": 0.6802, + "step": 2828 + }, + { + "epoch": 0.9933286516853933, + "grad_norm": 0.8804338574409485, + "learning_rate": 4.688330056950401e-06, + "loss": 0.6695, + "step": 2829 + }, + { + "epoch": 0.9936797752808989, + "grad_norm": 0.8846019506454468, + "learning_rate": 4.688106474938896e-06, + "loss": 0.6484, + "step": 2830 + }, + { + "epoch": 0.9940308988764045, + "grad_norm": 0.8795493841171265, + "learning_rate": 4.687882818096022e-06, + "loss": 0.6465, + "step": 2831 + }, + { + "epoch": 0.9943820224719101, + "grad_norm": 0.858067512512207, + "learning_rate": 4.687659086429427e-06, + "loss": 0.6471, + "step": 2832 + }, + { + "epoch": 0.9947331460674157, + "grad_norm": 0.8714587092399597, + "learning_rate": 4.687435279946765e-06, + "loss": 0.6446, + "step": 2833 + }, + { + "epoch": 0.9950842696629213, + "grad_norm": 0.8674041628837585, + "learning_rate": 4.687211398655686e-06, + "loss": 0.6419, + "step": 2834 + }, + { + "epoch": 0.995435393258427, + "grad_norm": 0.8507161736488342, + "learning_rate": 4.6869874425638505e-06, + "loss": 0.6202, + "step": 2835 + }, + { + "epoch": 0.9957865168539326, + "grad_norm": 0.8718125224113464, + "learning_rate": 4.686763411678915e-06, + "loss": 0.6807, + "step": 2836 + }, + { + "epoch": 0.9961376404494382, + "grad_norm": 0.8641071319580078, + "learning_rate": 4.686539306008543e-06, + "loss": 0.6492, + "step": 2837 + }, + { + "epoch": 0.9964887640449438, + "grad_norm": 0.8599060773849487, + "learning_rate": 4.686315125560397e-06, + "loss": 0.6779, + "step": 2838 + }, + { + "epoch": 0.9968398876404494, + "grad_norm": 0.8444222211837769, + "learning_rate": 4.686090870342145e-06, + "loss": 0.6841, + "step": 2839 + }, + { + "epoch": 0.9971910112359551, + "grad_norm": 0.86264568567276, + "learning_rate": 4.685866540361456e-06, + "loss": 0.6627, + "step": 2840 + }, + { + "epoch": 0.9975421348314607, + "grad_norm": 0.9181270599365234, + "learning_rate": 4.685642135626001e-06, + "loss": 0.6537, + "step": 2841 + }, + { + "epoch": 0.9978932584269663, + "grad_norm": 0.8701741695404053, + "learning_rate": 4.685417656143456e-06, + "loss": 0.6545, + "step": 2842 + }, + { + "epoch": 0.9982443820224719, + "grad_norm": 0.8788411021232605, + "learning_rate": 4.685193101921497e-06, + "loss": 0.6431, + "step": 2843 + }, + { + "epoch": 0.9985955056179775, + "grad_norm": 0.8748607635498047, + "learning_rate": 4.684968472967804e-06, + "loss": 0.6698, + "step": 2844 + }, + { + "epoch": 0.9989466292134831, + "grad_norm": 0.9054082632064819, + "learning_rate": 4.684743769290059e-06, + "loss": 0.6801, + "step": 2845 + }, + { + "epoch": 0.9992977528089888, + "grad_norm": 0.8799982666969299, + "learning_rate": 4.684518990895945e-06, + "loss": 0.6764, + "step": 2846 + }, + { + "epoch": 0.9996488764044944, + "grad_norm": 0.9407274127006531, + "learning_rate": 4.684294137793153e-06, + "loss": 0.7098, + "step": 2847 + }, + { + "epoch": 1.0, + "grad_norm": 0.91474848985672, + "learning_rate": 4.684069209989369e-06, + "loss": 0.6617, + "step": 2848 + }, + { + "epoch": 1.0003511235955056, + "grad_norm": 0.845862090587616, + "learning_rate": 4.683844207492289e-06, + "loss": 0.6516, + "step": 2849 + }, + { + "epoch": 1.0007022471910112, + "grad_norm": 0.8708998560905457, + "learning_rate": 4.683619130309604e-06, + "loss": 0.6668, + "step": 2850 + }, + { + "epoch": 1.0010533707865168, + "grad_norm": 0.8879108428955078, + "learning_rate": 4.683393978449014e-06, + "loss": 0.6155, + "step": 2851 + }, + { + "epoch": 1.0014044943820224, + "grad_norm": 0.8889144062995911, + "learning_rate": 4.683168751918218e-06, + "loss": 0.6625, + "step": 2852 + }, + { + "epoch": 1.001755617977528, + "grad_norm": 0.8704389929771423, + "learning_rate": 4.682943450724919e-06, + "loss": 0.6769, + "step": 2853 + }, + { + "epoch": 1.0021067415730338, + "grad_norm": 0.8852386474609375, + "learning_rate": 4.682718074876822e-06, + "loss": 0.6532, + "step": 2854 + }, + { + "epoch": 1.0024578651685394, + "grad_norm": 0.9099629521369934, + "learning_rate": 4.682492624381635e-06, + "loss": 0.6217, + "step": 2855 + }, + { + "epoch": 1.002808988764045, + "grad_norm": 0.9146444201469421, + "learning_rate": 4.682267099247066e-06, + "loss": 0.6483, + "step": 2856 + }, + { + "epoch": 1.0031601123595506, + "grad_norm": 0.9506009221076965, + "learning_rate": 4.682041499480832e-06, + "loss": 0.6958, + "step": 2857 + }, + { + "epoch": 1.0035112359550562, + "grad_norm": 0.9254932999610901, + "learning_rate": 4.681815825090645e-06, + "loss": 0.6637, + "step": 2858 + }, + { + "epoch": 1.0038623595505618, + "grad_norm": 0.9010646939277649, + "learning_rate": 4.681590076084224e-06, + "loss": 0.6594, + "step": 2859 + }, + { + "epoch": 1.0042134831460674, + "grad_norm": 0.8690000772476196, + "learning_rate": 4.6813642524692886e-06, + "loss": 0.6238, + "step": 2860 + }, + { + "epoch": 1.004564606741573, + "grad_norm": 0.8916565179824829, + "learning_rate": 4.681138354253563e-06, + "loss": 0.6279, + "step": 2861 + }, + { + "epoch": 1.0049157303370786, + "grad_norm": 0.9105271697044373, + "learning_rate": 4.680912381444771e-06, + "loss": 0.6636, + "step": 2862 + }, + { + "epoch": 1.0052668539325842, + "grad_norm": 0.9027879238128662, + "learning_rate": 4.680686334050643e-06, + "loss": 0.6818, + "step": 2863 + }, + { + "epoch": 1.0056179775280898, + "grad_norm": 0.8352002501487732, + "learning_rate": 4.6804602120789075e-06, + "loss": 0.6071, + "step": 2864 + }, + { + "epoch": 1.0059691011235956, + "grad_norm": 0.9015585780143738, + "learning_rate": 4.680234015537299e-06, + "loss": 0.6091, + "step": 2865 + }, + { + "epoch": 1.0063202247191012, + "grad_norm": 0.8762530088424683, + "learning_rate": 4.680007744433553e-06, + "loss": 0.6307, + "step": 2866 + }, + { + "epoch": 1.0066713483146068, + "grad_norm": 0.9097511768341064, + "learning_rate": 4.679781398775408e-06, + "loss": 0.6567, + "step": 2867 + }, + { + "epoch": 1.0070224719101124, + "grad_norm": 0.9116111993789673, + "learning_rate": 4.679554978570604e-06, + "loss": 0.6421, + "step": 2868 + }, + { + "epoch": 1.007373595505618, + "grad_norm": 0.8672003746032715, + "learning_rate": 4.679328483826885e-06, + "loss": 0.6465, + "step": 2869 + }, + { + "epoch": 1.0077247191011236, + "grad_norm": 0.9330756068229675, + "learning_rate": 4.679101914551997e-06, + "loss": 0.6187, + "step": 2870 + }, + { + "epoch": 1.0080758426966292, + "grad_norm": 0.8700326085090637, + "learning_rate": 4.678875270753687e-06, + "loss": 0.6089, + "step": 2871 + }, + { + "epoch": 1.0084269662921348, + "grad_norm": 0.9126592874526978, + "learning_rate": 4.678648552439708e-06, + "loss": 0.6386, + "step": 2872 + }, + { + "epoch": 1.0087780898876404, + "grad_norm": 0.9096201658248901, + "learning_rate": 4.678421759617813e-06, + "loss": 0.6059, + "step": 2873 + }, + { + "epoch": 1.009129213483146, + "grad_norm": 0.8730313777923584, + "learning_rate": 4.678194892295758e-06, + "loss": 0.6228, + "step": 2874 + }, + { + "epoch": 1.0094803370786516, + "grad_norm": 0.9104415774345398, + "learning_rate": 4.6779679504813e-06, + "loss": 0.6251, + "step": 2875 + }, + { + "epoch": 1.0098314606741574, + "grad_norm": 0.8940228819847107, + "learning_rate": 4.677740934182204e-06, + "loss": 0.6705, + "step": 2876 + }, + { + "epoch": 1.010182584269663, + "grad_norm": 0.8782382607460022, + "learning_rate": 4.67751384340623e-06, + "loss": 0.6203, + "step": 2877 + }, + { + "epoch": 1.0105337078651686, + "grad_norm": 0.8854807615280151, + "learning_rate": 4.6772866781611474e-06, + "loss": 0.6414, + "step": 2878 + }, + { + "epoch": 1.0108848314606742, + "grad_norm": 0.8533816933631897, + "learning_rate": 4.677059438454722e-06, + "loss": 0.6019, + "step": 2879 + }, + { + "epoch": 1.0112359550561798, + "grad_norm": 0.8798636198043823, + "learning_rate": 4.6768321242947276e-06, + "loss": 0.643, + "step": 2880 + }, + { + "epoch": 1.0115870786516854, + "grad_norm": 0.8717870712280273, + "learning_rate": 4.676604735688937e-06, + "loss": 0.6125, + "step": 2881 + }, + { + "epoch": 1.011938202247191, + "grad_norm": 0.8863524198532104, + "learning_rate": 4.676377272645127e-06, + "loss": 0.6475, + "step": 2882 + }, + { + "epoch": 1.0122893258426966, + "grad_norm": 0.9243369698524475, + "learning_rate": 4.6761497351710775e-06, + "loss": 0.654, + "step": 2883 + }, + { + "epoch": 1.0126404494382022, + "grad_norm": 0.9152374267578125, + "learning_rate": 4.675922123274568e-06, + "loss": 0.685, + "step": 2884 + }, + { + "epoch": 1.0129915730337078, + "grad_norm": 0.8751019835472107, + "learning_rate": 4.675694436963384e-06, + "loss": 0.6177, + "step": 2885 + }, + { + "epoch": 1.0133426966292134, + "grad_norm": 0.9058424830436707, + "learning_rate": 4.675466676245313e-06, + "loss": 0.6462, + "step": 2886 + }, + { + "epoch": 1.0136938202247192, + "grad_norm": 0.8993322849273682, + "learning_rate": 4.675238841128142e-06, + "loss": 0.6424, + "step": 2887 + }, + { + "epoch": 1.0140449438202248, + "grad_norm": 0.9268403053283691, + "learning_rate": 4.675010931619665e-06, + "loss": 0.6514, + "step": 2888 + }, + { + "epoch": 1.0143960674157304, + "grad_norm": 0.9016578197479248, + "learning_rate": 4.674782947727674e-06, + "loss": 0.6512, + "step": 2889 + }, + { + "epoch": 1.014747191011236, + "grad_norm": 0.8681126236915588, + "learning_rate": 4.674554889459969e-06, + "loss": 0.6703, + "step": 2890 + }, + { + "epoch": 1.0150983146067416, + "grad_norm": 0.8943293690681458, + "learning_rate": 4.6743267568243455e-06, + "loss": 0.6363, + "step": 2891 + }, + { + "epoch": 1.0154494382022472, + "grad_norm": 0.9124616384506226, + "learning_rate": 4.674098549828609e-06, + "loss": 0.6177, + "step": 2892 + }, + { + "epoch": 1.0158005617977528, + "grad_norm": 0.8788502216339111, + "learning_rate": 4.673870268480563e-06, + "loss": 0.6459, + "step": 2893 + }, + { + "epoch": 1.0161516853932584, + "grad_norm": 0.8843396306037903, + "learning_rate": 4.6736419127880126e-06, + "loss": 0.6513, + "step": 2894 + }, + { + "epoch": 1.016502808988764, + "grad_norm": 0.8527891039848328, + "learning_rate": 4.673413482758769e-06, + "loss": 0.6349, + "step": 2895 + }, + { + "epoch": 1.0168539325842696, + "grad_norm": 0.8894069194793701, + "learning_rate": 4.673184978400645e-06, + "loss": 0.627, + "step": 2896 + }, + { + "epoch": 1.0172050561797752, + "grad_norm": 0.913414478302002, + "learning_rate": 4.6729563997214534e-06, + "loss": 0.6433, + "step": 2897 + }, + { + "epoch": 1.0175561797752808, + "grad_norm": 0.9203279614448547, + "learning_rate": 4.672727746729013e-06, + "loss": 0.642, + "step": 2898 + }, + { + "epoch": 1.0179073033707866, + "grad_norm": 0.9143649935722351, + "learning_rate": 4.672499019431143e-06, + "loss": 0.6529, + "step": 2899 + }, + { + "epoch": 1.0182584269662922, + "grad_norm": 0.9485595226287842, + "learning_rate": 4.672270217835665e-06, + "loss": 0.6509, + "step": 2900 + }, + { + "epoch": 1.0186095505617978, + "grad_norm": 0.9203194379806519, + "learning_rate": 4.672041341950404e-06, + "loss": 0.69, + "step": 2901 + }, + { + "epoch": 1.0189606741573034, + "grad_norm": 0.9102271199226379, + "learning_rate": 4.671812391783188e-06, + "loss": 0.6387, + "step": 2902 + }, + { + "epoch": 1.019311797752809, + "grad_norm": 0.8916968107223511, + "learning_rate": 4.671583367341848e-06, + "loss": 0.6411, + "step": 2903 + }, + { + "epoch": 1.0196629213483146, + "grad_norm": 0.9091634750366211, + "learning_rate": 4.671354268634214e-06, + "loss": 0.6601, + "step": 2904 + }, + { + "epoch": 1.0200140449438202, + "grad_norm": 0.8721140027046204, + "learning_rate": 4.671125095668122e-06, + "loss": 0.6299, + "step": 2905 + }, + { + "epoch": 1.0203651685393258, + "grad_norm": 0.8712757229804993, + "learning_rate": 4.67089584845141e-06, + "loss": 0.6649, + "step": 2906 + }, + { + "epoch": 1.0207162921348314, + "grad_norm": 0.9056061506271362, + "learning_rate": 4.670666526991917e-06, + "loss": 0.5834, + "step": 2907 + }, + { + "epoch": 1.021067415730337, + "grad_norm": 0.9042145013809204, + "learning_rate": 4.670437131297487e-06, + "loss": 0.6487, + "step": 2908 + }, + { + "epoch": 1.0214185393258426, + "grad_norm": 0.914989173412323, + "learning_rate": 4.6702076613759635e-06, + "loss": 0.6595, + "step": 2909 + }, + { + "epoch": 1.0217696629213484, + "grad_norm": 0.850035548210144, + "learning_rate": 4.669978117235195e-06, + "loss": 0.6253, + "step": 2910 + }, + { + "epoch": 1.022120786516854, + "grad_norm": 0.9134849309921265, + "learning_rate": 4.6697484988830334e-06, + "loss": 0.6486, + "step": 2911 + }, + { + "epoch": 1.0224719101123596, + "grad_norm": 0.8706740736961365, + "learning_rate": 4.669518806327328e-06, + "loss": 0.6734, + "step": 2912 + }, + { + "epoch": 1.0228230337078652, + "grad_norm": 0.8819549083709717, + "learning_rate": 4.669289039575937e-06, + "loss": 0.6034, + "step": 2913 + }, + { + "epoch": 1.0231741573033708, + "grad_norm": 0.8862223625183105, + "learning_rate": 4.6690591986367175e-06, + "loss": 0.637, + "step": 2914 + }, + { + "epoch": 1.0235252808988764, + "grad_norm": 0.895828127861023, + "learning_rate": 4.668829283517528e-06, + "loss": 0.6243, + "step": 2915 + }, + { + "epoch": 1.023876404494382, + "grad_norm": 0.9082605838775635, + "learning_rate": 4.668599294226234e-06, + "loss": 0.6333, + "step": 2916 + }, + { + "epoch": 1.0242275280898876, + "grad_norm": 0.9978623390197754, + "learning_rate": 4.6683692307707e-06, + "loss": 0.6772, + "step": 2917 + }, + { + "epoch": 1.0245786516853932, + "grad_norm": 0.8527755737304688, + "learning_rate": 4.668139093158793e-06, + "loss": 0.6118, + "step": 2918 + }, + { + "epoch": 1.0249297752808988, + "grad_norm": 0.8660287857055664, + "learning_rate": 4.667908881398384e-06, + "loss": 0.6458, + "step": 2919 + }, + { + "epoch": 1.0252808988764044, + "grad_norm": 0.8769094347953796, + "learning_rate": 4.667678595497348e-06, + "loss": 0.6458, + "step": 2920 + }, + { + "epoch": 1.0256320224719102, + "grad_norm": 0.8812835812568665, + "learning_rate": 4.6674482354635576e-06, + "loss": 0.6761, + "step": 2921 + }, + { + "epoch": 1.0259831460674158, + "grad_norm": 0.9102697372436523, + "learning_rate": 4.6672178013048925e-06, + "loss": 0.62, + "step": 2922 + }, + { + "epoch": 1.0263342696629214, + "grad_norm": 0.8357862234115601, + "learning_rate": 4.6669872930292336e-06, + "loss": 0.5875, + "step": 2923 + }, + { + "epoch": 1.026685393258427, + "grad_norm": 0.8696123361587524, + "learning_rate": 4.666756710644463e-06, + "loss": 0.6224, + "step": 2924 + }, + { + "epoch": 1.0270365168539326, + "grad_norm": 0.9066466689109802, + "learning_rate": 4.666526054158467e-06, + "loss": 0.6789, + "step": 2925 + }, + { + "epoch": 1.0273876404494382, + "grad_norm": 0.8718449473381042, + "learning_rate": 4.666295323579134e-06, + "loss": 0.633, + "step": 2926 + }, + { + "epoch": 1.0277387640449438, + "grad_norm": 0.9176349639892578, + "learning_rate": 4.666064518914355e-06, + "loss": 0.6343, + "step": 2927 + }, + { + "epoch": 1.0280898876404494, + "grad_norm": 0.8917323350906372, + "learning_rate": 4.6658336401720226e-06, + "loss": 0.6122, + "step": 2928 + }, + { + "epoch": 1.028441011235955, + "grad_norm": 0.884037971496582, + "learning_rate": 4.665602687360034e-06, + "loss": 0.6387, + "step": 2929 + }, + { + "epoch": 1.0287921348314606, + "grad_norm": 0.901550829410553, + "learning_rate": 4.665371660486285e-06, + "loss": 0.6496, + "step": 2930 + }, + { + "epoch": 1.0291432584269662, + "grad_norm": 0.8866931200027466, + "learning_rate": 4.66514055955868e-06, + "loss": 0.6266, + "step": 2931 + }, + { + "epoch": 1.029494382022472, + "grad_norm": 0.8960362672805786, + "learning_rate": 4.664909384585119e-06, + "loss": 0.6186, + "step": 2932 + }, + { + "epoch": 1.0298455056179776, + "grad_norm": 0.8991187810897827, + "learning_rate": 4.66467813557351e-06, + "loss": 0.6298, + "step": 2933 + }, + { + "epoch": 1.0301966292134832, + "grad_norm": 0.869250476360321, + "learning_rate": 4.664446812531761e-06, + "loss": 0.5864, + "step": 2934 + }, + { + "epoch": 1.0305477528089888, + "grad_norm": 0.8768630027770996, + "learning_rate": 4.664215415467784e-06, + "loss": 0.6434, + "step": 2935 + }, + { + "epoch": 1.0308988764044944, + "grad_norm": 0.9055302143096924, + "learning_rate": 4.663983944389491e-06, + "loss": 0.6467, + "step": 2936 + }, + { + "epoch": 1.03125, + "grad_norm": 0.9194045066833496, + "learning_rate": 4.663752399304799e-06, + "loss": 0.6136, + "step": 2937 + }, + { + "epoch": 1.0316011235955056, + "grad_norm": 0.8887422680854797, + "learning_rate": 4.663520780221626e-06, + "loss": 0.6614, + "step": 2938 + }, + { + "epoch": 1.0319522471910112, + "grad_norm": 0.9037759304046631, + "learning_rate": 4.663289087147894e-06, + "loss": 0.6892, + "step": 2939 + }, + { + "epoch": 1.0323033707865168, + "grad_norm": 0.8857134580612183, + "learning_rate": 4.6630573200915275e-06, + "loss": 0.6227, + "step": 2940 + }, + { + "epoch": 1.0326544943820224, + "grad_norm": 0.8977053165435791, + "learning_rate": 4.66282547906045e-06, + "loss": 0.6445, + "step": 2941 + }, + { + "epoch": 1.033005617977528, + "grad_norm": 0.9049233794212341, + "learning_rate": 4.662593564062593e-06, + "loss": 0.6139, + "step": 2942 + }, + { + "epoch": 1.0333567415730338, + "grad_norm": 0.894765317440033, + "learning_rate": 4.6623615751058856e-06, + "loss": 0.6186, + "step": 2943 + }, + { + "epoch": 1.0337078651685394, + "grad_norm": 0.9289073944091797, + "learning_rate": 4.6621295121982634e-06, + "loss": 0.6496, + "step": 2944 + }, + { + "epoch": 1.034058988764045, + "grad_norm": 0.8969010710716248, + "learning_rate": 4.661897375347662e-06, + "loss": 0.6634, + "step": 2945 + }, + { + "epoch": 1.0344101123595506, + "grad_norm": 0.9372804164886475, + "learning_rate": 4.66166516456202e-06, + "loss": 0.6445, + "step": 2946 + }, + { + "epoch": 1.0347612359550562, + "grad_norm": 0.9191506505012512, + "learning_rate": 4.66143287984928e-06, + "loss": 0.6105, + "step": 2947 + }, + { + "epoch": 1.0351123595505618, + "grad_norm": 0.9287760853767395, + "learning_rate": 4.661200521217384e-06, + "loss": 0.6174, + "step": 2948 + }, + { + "epoch": 1.0354634831460674, + "grad_norm": 0.8911864757537842, + "learning_rate": 4.66096808867428e-06, + "loss": 0.6857, + "step": 2949 + }, + { + "epoch": 1.035814606741573, + "grad_norm": 0.8997723460197449, + "learning_rate": 4.660735582227917e-06, + "loss": 0.6199, + "step": 2950 + }, + { + "epoch": 1.0361657303370786, + "grad_norm": 0.9067447185516357, + "learning_rate": 4.660503001886245e-06, + "loss": 0.684, + "step": 2951 + }, + { + "epoch": 1.0365168539325842, + "grad_norm": 0.9156118631362915, + "learning_rate": 4.66027034765722e-06, + "loss": 0.6705, + "step": 2952 + }, + { + "epoch": 1.0368679775280898, + "grad_norm": 0.8432906866073608, + "learning_rate": 4.660037619548797e-06, + "loss": 0.6146, + "step": 2953 + }, + { + "epoch": 1.0372191011235956, + "grad_norm": 0.8450725674629211, + "learning_rate": 4.659804817568937e-06, + "loss": 0.6166, + "step": 2954 + }, + { + "epoch": 1.0375702247191012, + "grad_norm": 0.854310929775238, + "learning_rate": 4.659571941725599e-06, + "loss": 0.6527, + "step": 2955 + }, + { + "epoch": 1.0379213483146068, + "grad_norm": 0.8777969479560852, + "learning_rate": 4.659338992026749e-06, + "loss": 0.6587, + "step": 2956 + }, + { + "epoch": 1.0382724719101124, + "grad_norm": 0.8768917918205261, + "learning_rate": 4.659105968480353e-06, + "loss": 0.6239, + "step": 2957 + }, + { + "epoch": 1.038623595505618, + "grad_norm": 0.8562551736831665, + "learning_rate": 4.658872871094381e-06, + "loss": 0.6188, + "step": 2958 + }, + { + "epoch": 1.0389747191011236, + "grad_norm": 0.8880004286766052, + "learning_rate": 4.658639699876804e-06, + "loss": 0.6379, + "step": 2959 + }, + { + "epoch": 1.0393258426966292, + "grad_norm": 0.8730093240737915, + "learning_rate": 4.658406454835596e-06, + "loss": 0.6575, + "step": 2960 + }, + { + "epoch": 1.0396769662921348, + "grad_norm": 0.8812423944473267, + "learning_rate": 4.658173135978735e-06, + "loss": 0.6451, + "step": 2961 + }, + { + "epoch": 1.0400280898876404, + "grad_norm": 0.8783746957778931, + "learning_rate": 4.657939743314199e-06, + "loss": 0.6775, + "step": 2962 + }, + { + "epoch": 1.040379213483146, + "grad_norm": 0.8458737134933472, + "learning_rate": 4.657706276849971e-06, + "loss": 0.644, + "step": 2963 + }, + { + "epoch": 1.0407303370786516, + "grad_norm": 0.8714883923530579, + "learning_rate": 4.6574727365940344e-06, + "loss": 0.6193, + "step": 2964 + }, + { + "epoch": 1.0410814606741574, + "grad_norm": 0.8596876859664917, + "learning_rate": 4.657239122554376e-06, + "loss": 0.6392, + "step": 2965 + }, + { + "epoch": 1.041432584269663, + "grad_norm": 0.831123948097229, + "learning_rate": 4.657005434738986e-06, + "loss": 0.6212, + "step": 2966 + }, + { + "epoch": 1.0417837078651686, + "grad_norm": 0.8762812614440918, + "learning_rate": 4.656771673155856e-06, + "loss": 0.6422, + "step": 2967 + }, + { + "epoch": 1.0421348314606742, + "grad_norm": 0.8940248489379883, + "learning_rate": 4.65653783781298e-06, + "loss": 0.5996, + "step": 2968 + }, + { + "epoch": 1.0424859550561798, + "grad_norm": 0.9237715601921082, + "learning_rate": 4.6563039287183556e-06, + "loss": 0.6476, + "step": 2969 + }, + { + "epoch": 1.0428370786516854, + "grad_norm": 0.8804140686988831, + "learning_rate": 4.656069945879982e-06, + "loss": 0.6311, + "step": 2970 + }, + { + "epoch": 1.043188202247191, + "grad_norm": 0.8878976702690125, + "learning_rate": 4.655835889305861e-06, + "loss": 0.637, + "step": 2971 + }, + { + "epoch": 1.0435393258426966, + "grad_norm": 0.9103527665138245, + "learning_rate": 4.655601759003997e-06, + "loss": 0.611, + "step": 2972 + }, + { + "epoch": 1.0438904494382022, + "grad_norm": 0.8697876334190369, + "learning_rate": 4.6553675549823975e-06, + "loss": 0.6484, + "step": 2973 + }, + { + "epoch": 1.0442415730337078, + "grad_norm": 0.8702844381332397, + "learning_rate": 4.655133277249073e-06, + "loss": 0.6206, + "step": 2974 + }, + { + "epoch": 1.0445926966292134, + "grad_norm": 0.8851619362831116, + "learning_rate": 4.654898925812034e-06, + "loss": 0.6368, + "step": 2975 + }, + { + "epoch": 1.0449438202247192, + "grad_norm": 0.9111335873603821, + "learning_rate": 4.654664500679295e-06, + "loss": 0.6178, + "step": 2976 + }, + { + "epoch": 1.0452949438202248, + "grad_norm": 0.8853289484977722, + "learning_rate": 4.654430001858874e-06, + "loss": 0.6461, + "step": 2977 + }, + { + "epoch": 1.0456460674157304, + "grad_norm": 0.8976671099662781, + "learning_rate": 4.654195429358792e-06, + "loss": 0.6459, + "step": 2978 + }, + { + "epoch": 1.045997191011236, + "grad_norm": 0.8783121705055237, + "learning_rate": 4.6539607831870685e-06, + "loss": 0.695, + "step": 2979 + }, + { + "epoch": 1.0463483146067416, + "grad_norm": 0.8905385732650757, + "learning_rate": 4.653726063351729e-06, + "loss": 0.6458, + "step": 2980 + }, + { + "epoch": 1.0466994382022472, + "grad_norm": 0.89134681224823, + "learning_rate": 4.6534912698608015e-06, + "loss": 0.6405, + "step": 2981 + }, + { + "epoch": 1.0470505617977528, + "grad_norm": 0.8397007584571838, + "learning_rate": 4.653256402722316e-06, + "loss": 0.6003, + "step": 2982 + }, + { + "epoch": 1.0474016853932584, + "grad_norm": 0.8812448978424072, + "learning_rate": 4.653021461944304e-06, + "loss": 0.6249, + "step": 2983 + }, + { + "epoch": 1.047752808988764, + "grad_norm": 0.8805206418037415, + "learning_rate": 4.652786447534799e-06, + "loss": 0.6545, + "step": 2984 + }, + { + "epoch": 1.0481039325842696, + "grad_norm": 0.8674848675727844, + "learning_rate": 4.652551359501841e-06, + "loss": 0.615, + "step": 2985 + }, + { + "epoch": 1.0484550561797752, + "grad_norm": 0.8602030277252197, + "learning_rate": 4.652316197853468e-06, + "loss": 0.6166, + "step": 2986 + }, + { + "epoch": 1.0488061797752808, + "grad_norm": 0.866513192653656, + "learning_rate": 4.6520809625977224e-06, + "loss": 0.6312, + "step": 2987 + }, + { + "epoch": 1.0491573033707866, + "grad_norm": 0.8837951421737671, + "learning_rate": 4.65184565374265e-06, + "loss": 0.627, + "step": 2988 + }, + { + "epoch": 1.0495084269662922, + "grad_norm": 0.8847421407699585, + "learning_rate": 4.651610271296298e-06, + "loss": 0.617, + "step": 2989 + }, + { + "epoch": 1.0498595505617978, + "grad_norm": 0.8955681324005127, + "learning_rate": 4.651374815266715e-06, + "loss": 0.6153, + "step": 2990 + }, + { + "epoch": 1.0502106741573034, + "grad_norm": 0.8633450865745544, + "learning_rate": 4.651139285661955e-06, + "loss": 0.6217, + "step": 2991 + }, + { + "epoch": 1.050561797752809, + "grad_norm": 0.886147677898407, + "learning_rate": 4.650903682490072e-06, + "loss": 0.5844, + "step": 2992 + }, + { + "epoch": 1.0509129213483146, + "grad_norm": 0.9161800146102905, + "learning_rate": 4.650668005759124e-06, + "loss": 0.6577, + "step": 2993 + }, + { + "epoch": 1.0512640449438202, + "grad_norm": 0.9195558428764343, + "learning_rate": 4.6504322554771694e-06, + "loss": 0.5948, + "step": 2994 + }, + { + "epoch": 1.0516151685393258, + "grad_norm": 0.8970798850059509, + "learning_rate": 4.650196431652272e-06, + "loss": 0.6012, + "step": 2995 + }, + { + "epoch": 1.0519662921348314, + "grad_norm": 0.9370929002761841, + "learning_rate": 4.649960534292497e-06, + "loss": 0.6284, + "step": 2996 + }, + { + "epoch": 1.052317415730337, + "grad_norm": 0.8850135207176208, + "learning_rate": 4.649724563405912e-06, + "loss": 0.6382, + "step": 2997 + }, + { + "epoch": 1.0526685393258426, + "grad_norm": 0.8644798398017883, + "learning_rate": 4.649488519000586e-06, + "loss": 0.6199, + "step": 2998 + }, + { + "epoch": 1.0530196629213484, + "grad_norm": 0.8610007166862488, + "learning_rate": 4.649252401084592e-06, + "loss": 0.6561, + "step": 2999 + }, + { + "epoch": 1.053370786516854, + "grad_norm": 0.9027438163757324, + "learning_rate": 4.649016209666006e-06, + "loss": 0.641, + "step": 3000 + }, + { + "epoch": 1.0537219101123596, + "grad_norm": 0.8504754304885864, + "learning_rate": 4.6487799447529025e-06, + "loss": 0.6128, + "step": 3001 + }, + { + "epoch": 1.0540730337078652, + "grad_norm": 0.8692281246185303, + "learning_rate": 4.648543606353365e-06, + "loss": 0.6375, + "step": 3002 + }, + { + "epoch": 1.0544241573033708, + "grad_norm": 0.9047029614448547, + "learning_rate": 4.648307194475474e-06, + "loss": 0.6364, + "step": 3003 + }, + { + "epoch": 1.0547752808988764, + "grad_norm": 0.9719060659408569, + "learning_rate": 4.648070709127316e-06, + "loss": 0.6765, + "step": 3004 + }, + { + "epoch": 1.055126404494382, + "grad_norm": 0.889021635055542, + "learning_rate": 4.647834150316978e-06, + "loss": 0.6384, + "step": 3005 + }, + { + "epoch": 1.0554775280898876, + "grad_norm": 0.8828878998756409, + "learning_rate": 4.647597518052549e-06, + "loss": 0.628, + "step": 3006 + }, + { + "epoch": 1.0558286516853932, + "grad_norm": 0.8976204991340637, + "learning_rate": 4.647360812342123e-06, + "loss": 0.6497, + "step": 3007 + }, + { + "epoch": 1.0561797752808988, + "grad_norm": 0.9564626216888428, + "learning_rate": 4.647124033193795e-06, + "loss": 0.6396, + "step": 3008 + }, + { + "epoch": 1.0565308988764044, + "grad_norm": 0.8927937150001526, + "learning_rate": 4.6468871806156625e-06, + "loss": 0.6332, + "step": 3009 + }, + { + "epoch": 1.0568820224719102, + "grad_norm": 0.9105674624443054, + "learning_rate": 4.646650254615825e-06, + "loss": 0.6592, + "step": 3010 + }, + { + "epoch": 1.0572331460674158, + "grad_norm": 0.869835376739502, + "learning_rate": 4.646413255202386e-06, + "loss": 0.6322, + "step": 3011 + }, + { + "epoch": 1.0575842696629214, + "grad_norm": 0.9017779231071472, + "learning_rate": 4.646176182383451e-06, + "loss": 0.585, + "step": 3012 + }, + { + "epoch": 1.057935393258427, + "grad_norm": 0.8828255534172058, + "learning_rate": 4.645939036167127e-06, + "loss": 0.6305, + "step": 3013 + }, + { + "epoch": 1.0582865168539326, + "grad_norm": 0.8809278011322021, + "learning_rate": 4.645701816561523e-06, + "loss": 0.6758, + "step": 3014 + }, + { + "epoch": 1.0586376404494382, + "grad_norm": 0.8752210736274719, + "learning_rate": 4.645464523574753e-06, + "loss": 0.6448, + "step": 3015 + }, + { + "epoch": 1.0589887640449438, + "grad_norm": 0.8653308749198914, + "learning_rate": 4.645227157214933e-06, + "loss": 0.6405, + "step": 3016 + }, + { + "epoch": 1.0593398876404494, + "grad_norm": 1.4249169826507568, + "learning_rate": 4.64498971749018e-06, + "loss": 0.6015, + "step": 3017 + }, + { + "epoch": 1.059691011235955, + "grad_norm": 0.8942536115646362, + "learning_rate": 4.6447522044086144e-06, + "loss": 0.6818, + "step": 3018 + }, + { + "epoch": 1.0600421348314606, + "grad_norm": 0.8870991468429565, + "learning_rate": 4.644514617978359e-06, + "loss": 0.6296, + "step": 3019 + }, + { + "epoch": 1.0603932584269662, + "grad_norm": 0.9000036716461182, + "learning_rate": 4.644276958207538e-06, + "loss": 0.6122, + "step": 3020 + }, + { + "epoch": 1.060744382022472, + "grad_norm": 0.9074209332466125, + "learning_rate": 4.644039225104279e-06, + "loss": 0.658, + "step": 3021 + }, + { + "epoch": 1.0610955056179776, + "grad_norm": 0.9330301880836487, + "learning_rate": 4.643801418676714e-06, + "loss": 0.6222, + "step": 3022 + }, + { + "epoch": 1.0614466292134832, + "grad_norm": 0.892718493938446, + "learning_rate": 4.643563538932976e-06, + "loss": 0.6382, + "step": 3023 + }, + { + "epoch": 1.0617977528089888, + "grad_norm": 0.8774312734603882, + "learning_rate": 4.643325585881198e-06, + "loss": 0.61, + "step": 3024 + }, + { + "epoch": 1.0621488764044944, + "grad_norm": 0.9106227159500122, + "learning_rate": 4.643087559529519e-06, + "loss": 0.6362, + "step": 3025 + }, + { + "epoch": 1.0625, + "grad_norm": 0.8989148139953613, + "learning_rate": 4.642849459886079e-06, + "loss": 0.6264, + "step": 3026 + }, + { + "epoch": 1.0628511235955056, + "grad_norm": 0.9128755927085876, + "learning_rate": 4.6426112869590224e-06, + "loss": 0.6537, + "step": 3027 + }, + { + "epoch": 1.0632022471910112, + "grad_norm": 0.9222116470336914, + "learning_rate": 4.642373040756493e-06, + "loss": 0.6387, + "step": 3028 + }, + { + "epoch": 1.0635533707865168, + "grad_norm": 0.8915761113166809, + "learning_rate": 4.6421347212866385e-06, + "loss": 0.6207, + "step": 3029 + }, + { + "epoch": 1.0639044943820224, + "grad_norm": 0.9350106120109558, + "learning_rate": 4.64189632855761e-06, + "loss": 0.6882, + "step": 3030 + }, + { + "epoch": 1.064255617977528, + "grad_norm": 0.9001282453536987, + "learning_rate": 4.641657862577559e-06, + "loss": 0.6126, + "step": 3031 + }, + { + "epoch": 1.0646067415730338, + "grad_norm": 0.8905743360519409, + "learning_rate": 4.6414193233546426e-06, + "loss": 0.6418, + "step": 3032 + }, + { + "epoch": 1.0649578651685394, + "grad_norm": 0.9091342687606812, + "learning_rate": 4.6411807108970175e-06, + "loss": 0.632, + "step": 3033 + }, + { + "epoch": 1.065308988764045, + "grad_norm": 0.879478931427002, + "learning_rate": 4.640942025212844e-06, + "loss": 0.6223, + "step": 3034 + }, + { + "epoch": 1.0656601123595506, + "grad_norm": 0.8716598153114319, + "learning_rate": 4.640703266310286e-06, + "loss": 0.6576, + "step": 3035 + }, + { + "epoch": 1.0660112359550562, + "grad_norm": 0.867374837398529, + "learning_rate": 4.640464434197507e-06, + "loss": 0.618, + "step": 3036 + }, + { + "epoch": 1.0663623595505618, + "grad_norm": 0.899344801902771, + "learning_rate": 4.640225528882677e-06, + "loss": 0.6314, + "step": 3037 + }, + { + "epoch": 1.0667134831460674, + "grad_norm": 0.8644077181816101, + "learning_rate": 4.639986550373965e-06, + "loss": 0.6707, + "step": 3038 + }, + { + "epoch": 1.067064606741573, + "grad_norm": 0.8869829177856445, + "learning_rate": 4.639747498679544e-06, + "loss": 0.6249, + "step": 3039 + }, + { + "epoch": 1.0674157303370786, + "grad_norm": 0.9011994004249573, + "learning_rate": 4.63950837380759e-06, + "loss": 0.6363, + "step": 3040 + }, + { + "epoch": 1.0677668539325842, + "grad_norm": 0.870620846748352, + "learning_rate": 4.63926917576628e-06, + "loss": 0.6256, + "step": 3041 + }, + { + "epoch": 1.0681179775280898, + "grad_norm": 0.8807886838912964, + "learning_rate": 4.639029904563795e-06, + "loss": 0.6254, + "step": 3042 + }, + { + "epoch": 1.0684691011235956, + "grad_norm": 0.8803632259368896, + "learning_rate": 4.638790560208318e-06, + "loss": 0.6605, + "step": 3043 + }, + { + "epoch": 1.0688202247191012, + "grad_norm": 0.8775250315666199, + "learning_rate": 4.638551142708033e-06, + "loss": 0.6239, + "step": 3044 + }, + { + "epoch": 1.0691713483146068, + "grad_norm": 0.8915308117866516, + "learning_rate": 4.63831165207113e-06, + "loss": 0.6403, + "step": 3045 + }, + { + "epoch": 1.0695224719101124, + "grad_norm": 0.8814626336097717, + "learning_rate": 4.638072088305798e-06, + "loss": 0.6237, + "step": 3046 + }, + { + "epoch": 1.069873595505618, + "grad_norm": 0.9150183200836182, + "learning_rate": 4.63783245142023e-06, + "loss": 0.652, + "step": 3047 + }, + { + "epoch": 1.0702247191011236, + "grad_norm": 0.876789927482605, + "learning_rate": 4.637592741422621e-06, + "loss": 0.6419, + "step": 3048 + }, + { + "epoch": 1.0705758426966292, + "grad_norm": 0.8699705600738525, + "learning_rate": 4.63735295832117e-06, + "loss": 0.6399, + "step": 3049 + }, + { + "epoch": 1.0709269662921348, + "grad_norm": 0.8871046900749207, + "learning_rate": 4.637113102124077e-06, + "loss": 0.6427, + "step": 3050 + }, + { + "epoch": 1.0712780898876404, + "grad_norm": 0.877758800983429, + "learning_rate": 4.6368731728395444e-06, + "loss": 0.6534, + "step": 3051 + }, + { + "epoch": 1.071629213483146, + "grad_norm": 0.8570012450218201, + "learning_rate": 4.636633170475777e-06, + "loss": 0.6073, + "step": 3052 + }, + { + "epoch": 1.0719803370786516, + "grad_norm": 0.8933810591697693, + "learning_rate": 4.636393095040985e-06, + "loss": 0.608, + "step": 3053 + }, + { + "epoch": 1.0723314606741572, + "grad_norm": 0.8845889568328857, + "learning_rate": 4.636152946543377e-06, + "loss": 0.6466, + "step": 3054 + }, + { + "epoch": 1.072682584269663, + "grad_norm": 0.9182982444763184, + "learning_rate": 4.635912724991165e-06, + "loss": 0.6424, + "step": 3055 + }, + { + "epoch": 1.0730337078651686, + "grad_norm": 0.906626284122467, + "learning_rate": 4.635672430392566e-06, + "loss": 0.6481, + "step": 3056 + }, + { + "epoch": 1.0733848314606742, + "grad_norm": 0.9121250510215759, + "learning_rate": 4.635432062755798e-06, + "loss": 0.6403, + "step": 3057 + }, + { + "epoch": 1.0737359550561798, + "grad_norm": 0.8549202680587769, + "learning_rate": 4.63519162208908e-06, + "loss": 0.5976, + "step": 3058 + }, + { + "epoch": 1.0740870786516854, + "grad_norm": 0.9251717329025269, + "learning_rate": 4.634951108400636e-06, + "loss": 0.6198, + "step": 3059 + }, + { + "epoch": 1.074438202247191, + "grad_norm": 0.8714932799339294, + "learning_rate": 4.634710521698692e-06, + "loss": 0.6565, + "step": 3060 + }, + { + "epoch": 1.0747893258426966, + "grad_norm": 0.8903185725212097, + "learning_rate": 4.634469861991473e-06, + "loss": 0.6119, + "step": 3061 + }, + { + "epoch": 1.0751404494382022, + "grad_norm": 0.8958722352981567, + "learning_rate": 4.6342291292872125e-06, + "loss": 0.6417, + "step": 3062 + }, + { + "epoch": 1.0754915730337078, + "grad_norm": 0.8739234209060669, + "learning_rate": 4.633988323594142e-06, + "loss": 0.6374, + "step": 3063 + }, + { + "epoch": 1.0758426966292134, + "grad_norm": 0.8859426975250244, + "learning_rate": 4.633747444920497e-06, + "loss": 0.6547, + "step": 3064 + }, + { + "epoch": 1.0761938202247192, + "grad_norm": 0.9002133011817932, + "learning_rate": 4.633506493274514e-06, + "loss": 0.6693, + "step": 3065 + }, + { + "epoch": 1.0765449438202248, + "grad_norm": 0.8716120719909668, + "learning_rate": 4.633265468664436e-06, + "loss": 0.5971, + "step": 3066 + }, + { + "epoch": 1.0768960674157304, + "grad_norm": 0.9050167798995972, + "learning_rate": 4.633024371098503e-06, + "loss": 0.6686, + "step": 3067 + }, + { + "epoch": 1.077247191011236, + "grad_norm": 0.8593562841415405, + "learning_rate": 4.6327832005849635e-06, + "loss": 0.6134, + "step": 3068 + }, + { + "epoch": 1.0775983146067416, + "grad_norm": 0.8681020736694336, + "learning_rate": 4.632541957132063e-06, + "loss": 0.6232, + "step": 3069 + }, + { + "epoch": 1.0779494382022472, + "grad_norm": 0.8903972506523132, + "learning_rate": 4.632300640748052e-06, + "loss": 0.6333, + "step": 3070 + }, + { + "epoch": 1.0783005617977528, + "grad_norm": 0.8613920211791992, + "learning_rate": 4.632059251441184e-06, + "loss": 0.6129, + "step": 3071 + }, + { + "epoch": 1.0786516853932584, + "grad_norm": 0.9118038415908813, + "learning_rate": 4.631817789219714e-06, + "loss": 0.6453, + "step": 3072 + }, + { + "epoch": 1.079002808988764, + "grad_norm": 0.9000958800315857, + "learning_rate": 4.6315762540919e-06, + "loss": 0.6578, + "step": 3073 + }, + { + "epoch": 1.0793539325842696, + "grad_norm": 0.8761112093925476, + "learning_rate": 4.631334646066001e-06, + "loss": 0.6345, + "step": 3074 + }, + { + "epoch": 1.0797050561797752, + "grad_norm": 0.9070879817008972, + "learning_rate": 4.6310929651502825e-06, + "loss": 0.6475, + "step": 3075 + }, + { + "epoch": 1.0800561797752808, + "grad_norm": 0.869162380695343, + "learning_rate": 4.630851211353007e-06, + "loss": 0.6341, + "step": 3076 + }, + { + "epoch": 1.0804073033707866, + "grad_norm": 0.9210894703865051, + "learning_rate": 4.630609384682444e-06, + "loss": 0.6839, + "step": 3077 + }, + { + "epoch": 1.0807584269662922, + "grad_norm": 0.9321752786636353, + "learning_rate": 4.630367485146863e-06, + "loss": 0.6581, + "step": 3078 + }, + { + "epoch": 1.0811095505617978, + "grad_norm": 0.8789924383163452, + "learning_rate": 4.630125512754537e-06, + "loss": 0.6129, + "step": 3079 + }, + { + "epoch": 1.0814606741573034, + "grad_norm": 0.8936173915863037, + "learning_rate": 4.629883467513741e-06, + "loss": 0.6021, + "step": 3080 + }, + { + "epoch": 1.081811797752809, + "grad_norm": 0.8786038160324097, + "learning_rate": 4.629641349432754e-06, + "loss": 0.633, + "step": 3081 + }, + { + "epoch": 1.0821629213483146, + "grad_norm": 0.8804035782814026, + "learning_rate": 4.629399158519855e-06, + "loss": 0.6404, + "step": 3082 + }, + { + "epoch": 1.0825140449438202, + "grad_norm": 0.9330071210861206, + "learning_rate": 4.629156894783326e-06, + "loss": 0.6629, + "step": 3083 + }, + { + "epoch": 1.0828651685393258, + "grad_norm": 0.8635822534561157, + "learning_rate": 4.628914558231455e-06, + "loss": 0.6409, + "step": 3084 + }, + { + "epoch": 1.0832162921348314, + "grad_norm": 0.9098319411277771, + "learning_rate": 4.628672148872526e-06, + "loss": 0.6533, + "step": 3085 + }, + { + "epoch": 1.083567415730337, + "grad_norm": 0.8650401830673218, + "learning_rate": 4.628429666714831e-06, + "loss": 0.6031, + "step": 3086 + }, + { + "epoch": 1.0839185393258428, + "grad_norm": 0.8966626524925232, + "learning_rate": 4.6281871117666634e-06, + "loss": 0.6209, + "step": 3087 + }, + { + "epoch": 1.0842696629213484, + "grad_norm": 0.9028195142745972, + "learning_rate": 4.627944484036318e-06, + "loss": 0.6343, + "step": 3088 + }, + { + "epoch": 1.084620786516854, + "grad_norm": 0.9410665035247803, + "learning_rate": 4.627701783532092e-06, + "loss": 0.6796, + "step": 3089 + }, + { + "epoch": 1.0849719101123596, + "grad_norm": 0.9015296697616577, + "learning_rate": 4.627459010262285e-06, + "loss": 0.6649, + "step": 3090 + }, + { + "epoch": 1.0853230337078652, + "grad_norm": 0.8640577793121338, + "learning_rate": 4.627216164235201e-06, + "loss": 0.6224, + "step": 3091 + }, + { + "epoch": 1.0856741573033708, + "grad_norm": 0.8947779536247253, + "learning_rate": 4.626973245459144e-06, + "loss": 0.6254, + "step": 3092 + }, + { + "epoch": 1.0860252808988764, + "grad_norm": 0.8579085469245911, + "learning_rate": 4.626730253942422e-06, + "loss": 0.6158, + "step": 3093 + }, + { + "epoch": 1.086376404494382, + "grad_norm": 0.8828961849212646, + "learning_rate": 4.626487189693345e-06, + "loss": 0.6429, + "step": 3094 + }, + { + "epoch": 1.0867275280898876, + "grad_norm": 0.8998451232910156, + "learning_rate": 4.626244052720227e-06, + "loss": 0.6176, + "step": 3095 + }, + { + "epoch": 1.0870786516853932, + "grad_norm": 0.8975029587745667, + "learning_rate": 4.626000843031381e-06, + "loss": 0.6106, + "step": 3096 + }, + { + "epoch": 1.0874297752808988, + "grad_norm": 0.8873660564422607, + "learning_rate": 4.625757560635126e-06, + "loss": 0.6279, + "step": 3097 + }, + { + "epoch": 1.0877808988764044, + "grad_norm": 0.8927737474441528, + "learning_rate": 4.625514205539781e-06, + "loss": 0.6208, + "step": 3098 + }, + { + "epoch": 1.0881320224719102, + "grad_norm": 0.9284412264823914, + "learning_rate": 4.6252707777536685e-06, + "loss": 0.6701, + "step": 3099 + }, + { + "epoch": 1.0884831460674158, + "grad_norm": 0.8798356652259827, + "learning_rate": 4.625027277285114e-06, + "loss": 0.5872, + "step": 3100 + }, + { + "epoch": 1.0888342696629214, + "grad_norm": 0.9197069406509399, + "learning_rate": 4.624783704142446e-06, + "loss": 0.6478, + "step": 3101 + }, + { + "epoch": 1.089185393258427, + "grad_norm": 0.8777607083320618, + "learning_rate": 4.624540058333993e-06, + "loss": 0.6248, + "step": 3102 + }, + { + "epoch": 1.0895365168539326, + "grad_norm": 0.8981837630271912, + "learning_rate": 4.624296339868089e-06, + "loss": 0.6398, + "step": 3103 + }, + { + "epoch": 1.0898876404494382, + "grad_norm": 1.1026835441589355, + "learning_rate": 4.624052548753066e-06, + "loss": 0.701, + "step": 3104 + }, + { + "epoch": 1.0902387640449438, + "grad_norm": 0.8540518879890442, + "learning_rate": 4.6238086849972655e-06, + "loss": 0.6133, + "step": 3105 + }, + { + "epoch": 1.0905898876404494, + "grad_norm": 0.8801087141036987, + "learning_rate": 4.623564748609024e-06, + "loss": 0.6557, + "step": 3106 + }, + { + "epoch": 1.090941011235955, + "grad_norm": 0.8963208794593811, + "learning_rate": 4.623320739596685e-06, + "loss": 0.6368, + "step": 3107 + }, + { + "epoch": 1.0912921348314606, + "grad_norm": 0.8750952482223511, + "learning_rate": 4.623076657968594e-06, + "loss": 0.6296, + "step": 3108 + }, + { + "epoch": 1.0916432584269662, + "grad_norm": 0.9236363768577576, + "learning_rate": 4.622832503733099e-06, + "loss": 0.6424, + "step": 3109 + }, + { + "epoch": 1.091994382022472, + "grad_norm": 0.8751869797706604, + "learning_rate": 4.6225882768985485e-06, + "loss": 0.6033, + "step": 3110 + }, + { + "epoch": 1.0923455056179776, + "grad_norm": 0.9356908202171326, + "learning_rate": 4.6223439774732944e-06, + "loss": 0.6372, + "step": 3111 + }, + { + "epoch": 1.0926966292134832, + "grad_norm": 0.881905734539032, + "learning_rate": 4.622099605465693e-06, + "loss": 0.6139, + "step": 3112 + }, + { + "epoch": 1.0930477528089888, + "grad_norm": 0.8708788156509399, + "learning_rate": 4.6218551608841005e-06, + "loss": 0.6441, + "step": 3113 + }, + { + "epoch": 1.0933988764044944, + "grad_norm": 0.8892695903778076, + "learning_rate": 4.621610643736878e-06, + "loss": 0.6329, + "step": 3114 + }, + { + "epoch": 1.09375, + "grad_norm": 0.8857613801956177, + "learning_rate": 4.6213660540323865e-06, + "loss": 0.6215, + "step": 3115 + }, + { + "epoch": 1.0941011235955056, + "grad_norm": 0.8977625370025635, + "learning_rate": 4.621121391778991e-06, + "loss": 0.5709, + "step": 3116 + }, + { + "epoch": 1.0944522471910112, + "grad_norm": 0.9355443120002747, + "learning_rate": 4.62087665698506e-06, + "loss": 0.6367, + "step": 3117 + }, + { + "epoch": 1.0948033707865168, + "grad_norm": 0.8514768481254578, + "learning_rate": 4.620631849658961e-06, + "loss": 0.6059, + "step": 3118 + }, + { + "epoch": 1.0951544943820224, + "grad_norm": 0.897706925868988, + "learning_rate": 4.6203869698090676e-06, + "loss": 0.6327, + "step": 3119 + }, + { + "epoch": 1.095505617977528, + "grad_norm": 0.8812986612319946, + "learning_rate": 4.620142017443755e-06, + "loss": 0.625, + "step": 3120 + }, + { + "epoch": 1.0958567415730338, + "grad_norm": 0.878730833530426, + "learning_rate": 4.619896992571399e-06, + "loss": 0.6629, + "step": 3121 + }, + { + "epoch": 1.0962078651685394, + "grad_norm": 0.8637169003486633, + "learning_rate": 4.61965189520038e-06, + "loss": 0.6148, + "step": 3122 + }, + { + "epoch": 1.096558988764045, + "grad_norm": 0.8571529388427734, + "learning_rate": 4.61940672533908e-06, + "loss": 0.5992, + "step": 3123 + }, + { + "epoch": 1.0969101123595506, + "grad_norm": 0.8880246877670288, + "learning_rate": 4.619161482995883e-06, + "loss": 0.6257, + "step": 3124 + }, + { + "epoch": 1.0972612359550562, + "grad_norm": 0.8742407560348511, + "learning_rate": 4.618916168179178e-06, + "loss": 0.6463, + "step": 3125 + }, + { + "epoch": 1.0976123595505618, + "grad_norm": 0.9125159978866577, + "learning_rate": 4.6186707808973524e-06, + "loss": 0.6534, + "step": 3126 + }, + { + "epoch": 1.0979634831460674, + "grad_norm": 0.9120495319366455, + "learning_rate": 4.6184253211588e-06, + "loss": 0.6858, + "step": 3127 + }, + { + "epoch": 1.098314606741573, + "grad_norm": 0.8634641170501709, + "learning_rate": 4.618179788971913e-06, + "loss": 0.6343, + "step": 3128 + }, + { + "epoch": 1.0986657303370786, + "grad_norm": 0.9329657554626465, + "learning_rate": 4.6179341843450895e-06, + "loss": 0.6626, + "step": 3129 + }, + { + "epoch": 1.0990168539325842, + "grad_norm": 0.9169970750808716, + "learning_rate": 4.617688507286731e-06, + "loss": 0.6737, + "step": 3130 + }, + { + "epoch": 1.0993679775280898, + "grad_norm": 0.9113175272941589, + "learning_rate": 4.617442757805236e-06, + "loss": 0.6285, + "step": 3131 + }, + { + "epoch": 1.0997191011235956, + "grad_norm": 0.8834072351455688, + "learning_rate": 4.617196935909012e-06, + "loss": 0.6344, + "step": 3132 + }, + { + "epoch": 1.1000702247191012, + "grad_norm": 0.8792498707771301, + "learning_rate": 4.6169510416064635e-06, + "loss": 0.635, + "step": 3133 + }, + { + "epoch": 1.1004213483146068, + "grad_norm": 0.864789605140686, + "learning_rate": 4.6167050749060014e-06, + "loss": 0.6498, + "step": 3134 + }, + { + "epoch": 1.1007724719101124, + "grad_norm": 0.8957346677780151, + "learning_rate": 4.616459035816037e-06, + "loss": 0.6484, + "step": 3135 + }, + { + "epoch": 1.101123595505618, + "grad_norm": 0.9335662126541138, + "learning_rate": 4.6162129243449845e-06, + "loss": 0.651, + "step": 3136 + }, + { + "epoch": 1.1014747191011236, + "grad_norm": 0.8907780051231384, + "learning_rate": 4.615966740501261e-06, + "loss": 0.6279, + "step": 3137 + }, + { + "epoch": 1.1018258426966292, + "grad_norm": 0.8416582942008972, + "learning_rate": 4.615720484293286e-06, + "loss": 0.6183, + "step": 3138 + }, + { + "epoch": 1.1021769662921348, + "grad_norm": 0.8414410352706909, + "learning_rate": 4.6154741557294805e-06, + "loss": 0.6091, + "step": 3139 + }, + { + "epoch": 1.1025280898876404, + "grad_norm": 0.8924736380577087, + "learning_rate": 4.6152277548182685e-06, + "loss": 0.6466, + "step": 3140 + }, + { + "epoch": 1.102879213483146, + "grad_norm": 0.952515721321106, + "learning_rate": 4.614981281568078e-06, + "loss": 0.5875, + "step": 3141 + }, + { + "epoch": 1.1032303370786516, + "grad_norm": 1.0150952339172363, + "learning_rate": 4.6147347359873376e-06, + "loss": 0.6295, + "step": 3142 + }, + { + "epoch": 1.1035814606741572, + "grad_norm": 0.8959284424781799, + "learning_rate": 4.614488118084478e-06, + "loss": 0.6483, + "step": 3143 + }, + { + "epoch": 1.103932584269663, + "grad_norm": 0.8429542183876038, + "learning_rate": 4.614241427867935e-06, + "loss": 0.6017, + "step": 3144 + }, + { + "epoch": 1.1042837078651686, + "grad_norm": 0.8937910795211792, + "learning_rate": 4.613994665346144e-06, + "loss": 0.6685, + "step": 3145 + }, + { + "epoch": 1.1046348314606742, + "grad_norm": 0.9087120890617371, + "learning_rate": 4.6137478305275444e-06, + "loss": 0.656, + "step": 3146 + }, + { + "epoch": 1.1049859550561798, + "grad_norm": 0.9055848717689514, + "learning_rate": 4.613500923420577e-06, + "loss": 0.641, + "step": 3147 + }, + { + "epoch": 1.1053370786516854, + "grad_norm": 0.8999933004379272, + "learning_rate": 4.613253944033688e-06, + "loss": 0.6282, + "step": 3148 + }, + { + "epoch": 1.105688202247191, + "grad_norm": 0.901351809501648, + "learning_rate": 4.6130068923753226e-06, + "loss": 0.6409, + "step": 3149 + }, + { + "epoch": 1.1060393258426966, + "grad_norm": 0.8660926818847656, + "learning_rate": 4.612759768453928e-06, + "loss": 0.6238, + "step": 3150 + }, + { + "epoch": 1.1063904494382022, + "grad_norm": 0.8839932084083557, + "learning_rate": 4.612512572277959e-06, + "loss": 0.6394, + "step": 3151 + }, + { + "epoch": 1.1067415730337078, + "grad_norm": 0.882322371006012, + "learning_rate": 4.6122653038558665e-06, + "loss": 0.6587, + "step": 3152 + }, + { + "epoch": 1.1070926966292134, + "grad_norm": 0.8952814936637878, + "learning_rate": 4.6120179631961086e-06, + "loss": 0.647, + "step": 3153 + }, + { + "epoch": 1.1074438202247192, + "grad_norm": 0.8898069262504578, + "learning_rate": 4.611770550307143e-06, + "loss": 0.6238, + "step": 3154 + }, + { + "epoch": 1.1077949438202248, + "grad_norm": 0.8928719758987427, + "learning_rate": 4.6115230651974315e-06, + "loss": 0.6393, + "step": 3155 + }, + { + "epoch": 1.1081460674157304, + "grad_norm": 0.908116340637207, + "learning_rate": 4.611275507875439e-06, + "loss": 0.6431, + "step": 3156 + }, + { + "epoch": 1.108497191011236, + "grad_norm": 0.9306859970092773, + "learning_rate": 4.611027878349631e-06, + "loss": 0.6692, + "step": 3157 + }, + { + "epoch": 1.1088483146067416, + "grad_norm": 0.8866667151451111, + "learning_rate": 4.610780176628475e-06, + "loss": 0.621, + "step": 3158 + }, + { + "epoch": 1.1091994382022472, + "grad_norm": 0.8539839386940002, + "learning_rate": 4.610532402720444e-06, + "loss": 0.601, + "step": 3159 + }, + { + "epoch": 1.1095505617977528, + "grad_norm": 0.8730782270431519, + "learning_rate": 4.61028455663401e-06, + "loss": 0.6, + "step": 3160 + }, + { + "epoch": 1.1099016853932584, + "grad_norm": 0.8877293467521667, + "learning_rate": 4.6100366383776505e-06, + "loss": 0.629, + "step": 3161 + }, + { + "epoch": 1.110252808988764, + "grad_norm": 0.8557886481285095, + "learning_rate": 4.609788647959843e-06, + "loss": 0.6303, + "step": 3162 + }, + { + "epoch": 1.1106039325842696, + "grad_norm": 0.8975784778594971, + "learning_rate": 4.609540585389069e-06, + "loss": 0.645, + "step": 3163 + }, + { + "epoch": 1.1109550561797752, + "grad_norm": 0.8701896667480469, + "learning_rate": 4.6092924506738134e-06, + "loss": 0.6208, + "step": 3164 + }, + { + "epoch": 1.1113061797752808, + "grad_norm": 0.8492256999015808, + "learning_rate": 4.60904424382256e-06, + "loss": 0.6441, + "step": 3165 + }, + { + "epoch": 1.1116573033707866, + "grad_norm": 0.8746402263641357, + "learning_rate": 4.608795964843798e-06, + "loss": 0.6269, + "step": 3166 + }, + { + "epoch": 1.1120084269662922, + "grad_norm": 0.9015893936157227, + "learning_rate": 4.60854761374602e-06, + "loss": 0.6396, + "step": 3167 + }, + { + "epoch": 1.1123595505617978, + "grad_norm": 0.9147549867630005, + "learning_rate": 4.608299190537717e-06, + "loss": 0.6055, + "step": 3168 + }, + { + "epoch": 1.1127106741573034, + "grad_norm": 0.8999795317649841, + "learning_rate": 4.6080506952273855e-06, + "loss": 0.6343, + "step": 3169 + }, + { + "epoch": 1.113061797752809, + "grad_norm": 0.8840816617012024, + "learning_rate": 4.607802127823524e-06, + "loss": 0.6173, + "step": 3170 + }, + { + "epoch": 1.1134129213483146, + "grad_norm": 0.9025754332542419, + "learning_rate": 4.6075534883346335e-06, + "loss": 0.6426, + "step": 3171 + }, + { + "epoch": 1.1137640449438202, + "grad_norm": 0.9157222509384155, + "learning_rate": 4.607304776769217e-06, + "loss": 0.6639, + "step": 3172 + }, + { + "epoch": 1.1141151685393258, + "grad_norm": 0.8551729321479797, + "learning_rate": 4.607055993135781e-06, + "loss": 0.62, + "step": 3173 + }, + { + "epoch": 1.1144662921348314, + "grad_norm": 0.9303814172744751, + "learning_rate": 4.606807137442833e-06, + "loss": 0.6504, + "step": 3174 + }, + { + "epoch": 1.114817415730337, + "grad_norm": 0.8886232972145081, + "learning_rate": 4.606558209698884e-06, + "loss": 0.6301, + "step": 3175 + }, + { + "epoch": 1.1151685393258428, + "grad_norm": 0.8753377795219421, + "learning_rate": 4.606309209912446e-06, + "loss": 0.6206, + "step": 3176 + }, + { + "epoch": 1.1155196629213484, + "grad_norm": 0.9052675366401672, + "learning_rate": 4.606060138092036e-06, + "loss": 0.6563, + "step": 3177 + }, + { + "epoch": 1.115870786516854, + "grad_norm": 0.8677799105644226, + "learning_rate": 4.6058109942461705e-06, + "loss": 0.6203, + "step": 3178 + }, + { + "epoch": 1.1162219101123596, + "grad_norm": 0.8847718834877014, + "learning_rate": 4.605561778383372e-06, + "loss": 0.6473, + "step": 3179 + }, + { + "epoch": 1.1165730337078652, + "grad_norm": 0.8971977829933167, + "learning_rate": 4.605312490512162e-06, + "loss": 0.6363, + "step": 3180 + }, + { + "epoch": 1.1169241573033708, + "grad_norm": 0.8800894021987915, + "learning_rate": 4.605063130641067e-06, + "loss": 0.6071, + "step": 3181 + }, + { + "epoch": 1.1172752808988764, + "grad_norm": 0.884075403213501, + "learning_rate": 4.604813698778613e-06, + "loss": 0.6281, + "step": 3182 + }, + { + "epoch": 1.117626404494382, + "grad_norm": 0.8670194745063782, + "learning_rate": 4.604564194933332e-06, + "loss": 0.644, + "step": 3183 + }, + { + "epoch": 1.1179775280898876, + "grad_norm": 0.8559613227844238, + "learning_rate": 4.604314619113756e-06, + "loss": 0.6348, + "step": 3184 + }, + { + "epoch": 1.1183286516853932, + "grad_norm": 0.8893458247184753, + "learning_rate": 4.604064971328422e-06, + "loss": 0.646, + "step": 3185 + }, + { + "epoch": 1.1186797752808988, + "grad_norm": 0.9015454053878784, + "learning_rate": 4.603815251585867e-06, + "loss": 0.6798, + "step": 3186 + }, + { + "epoch": 1.1190308988764044, + "grad_norm": 0.881079912185669, + "learning_rate": 4.603565459894629e-06, + "loss": 0.6282, + "step": 3187 + }, + { + "epoch": 1.1193820224719102, + "grad_norm": 0.8383495211601257, + "learning_rate": 4.6033155962632535e-06, + "loss": 0.613, + "step": 3188 + }, + { + "epoch": 1.1197331460674158, + "grad_norm": 0.845001757144928, + "learning_rate": 4.603065660700283e-06, + "loss": 0.6104, + "step": 3189 + }, + { + "epoch": 1.1200842696629214, + "grad_norm": 0.8651373982429504, + "learning_rate": 4.602815653214267e-06, + "loss": 0.6286, + "step": 3190 + }, + { + "epoch": 1.120435393258427, + "grad_norm": 0.8765587210655212, + "learning_rate": 4.6025655738137566e-06, + "loss": 0.6206, + "step": 3191 + }, + { + "epoch": 1.1207865168539326, + "grad_norm": 0.8903856873512268, + "learning_rate": 4.602315422507301e-06, + "loss": 0.6162, + "step": 3192 + }, + { + "epoch": 1.1211376404494382, + "grad_norm": 0.8526650667190552, + "learning_rate": 4.602065199303458e-06, + "loss": 0.6418, + "step": 3193 + }, + { + "epoch": 1.1214887640449438, + "grad_norm": 0.8819793462753296, + "learning_rate": 4.601814904210783e-06, + "loss": 0.6513, + "step": 3194 + }, + { + "epoch": 1.1218398876404494, + "grad_norm": 0.9131810069084167, + "learning_rate": 4.601564537237837e-06, + "loss": 0.6234, + "step": 3195 + }, + { + "epoch": 1.122191011235955, + "grad_norm": 0.8469668030738831, + "learning_rate": 4.601314098393183e-06, + "loss": 0.5831, + "step": 3196 + }, + { + "epoch": 1.1225421348314606, + "grad_norm": 0.8926812410354614, + "learning_rate": 4.601063587685384e-06, + "loss": 0.6649, + "step": 3197 + }, + { + "epoch": 1.1228932584269662, + "grad_norm": 0.8658381700515747, + "learning_rate": 4.600813005123009e-06, + "loss": 0.6095, + "step": 3198 + }, + { + "epoch": 1.123244382022472, + "grad_norm": 0.8796584606170654, + "learning_rate": 4.600562350714627e-06, + "loss": 0.6189, + "step": 3199 + }, + { + "epoch": 1.1235955056179776, + "grad_norm": 0.8917720913887024, + "learning_rate": 4.6003116244688095e-06, + "loss": 0.6294, + "step": 3200 + }, + { + "epoch": 1.1239466292134832, + "grad_norm": 0.8931936621665955, + "learning_rate": 4.600060826394131e-06, + "loss": 0.6428, + "step": 3201 + }, + { + "epoch": 1.1242977528089888, + "grad_norm": 0.8980125784873962, + "learning_rate": 4.599809956499171e-06, + "loss": 0.6325, + "step": 3202 + }, + { + "epoch": 1.1246488764044944, + "grad_norm": 0.9097473621368408, + "learning_rate": 4.599559014792507e-06, + "loss": 0.6634, + "step": 3203 + }, + { + "epoch": 1.125, + "grad_norm": 0.920569121837616, + "learning_rate": 4.59930800128272e-06, + "loss": 0.6751, + "step": 3204 + }, + { + "epoch": 1.1253511235955056, + "grad_norm": 0.8496858477592468, + "learning_rate": 4.599056915978396e-06, + "loss": 0.6123, + "step": 3205 + }, + { + "epoch": 1.1257022471910112, + "grad_norm": 0.9153531193733215, + "learning_rate": 4.598805758888122e-06, + "loss": 0.6713, + "step": 3206 + }, + { + "epoch": 1.1260533707865168, + "grad_norm": 0.8894977569580078, + "learning_rate": 4.5985545300204875e-06, + "loss": 0.6154, + "step": 3207 + }, + { + "epoch": 1.1264044943820224, + "grad_norm": 0.9272782206535339, + "learning_rate": 4.598303229384083e-06, + "loss": 0.6311, + "step": 3208 + }, + { + "epoch": 1.126755617977528, + "grad_norm": 0.8554537296295166, + "learning_rate": 4.598051856987504e-06, + "loss": 0.6121, + "step": 3209 + }, + { + "epoch": 1.1271067415730336, + "grad_norm": 0.8680627346038818, + "learning_rate": 4.5978004128393475e-06, + "loss": 0.6256, + "step": 3210 + }, + { + "epoch": 1.1274578651685394, + "grad_norm": 0.8919551372528076, + "learning_rate": 4.597548896948212e-06, + "loss": 0.6632, + "step": 3211 + }, + { + "epoch": 1.127808988764045, + "grad_norm": 0.90400230884552, + "learning_rate": 4.597297309322698e-06, + "loss": 0.6558, + "step": 3212 + }, + { + "epoch": 1.1281601123595506, + "grad_norm": 0.8687441945075989, + "learning_rate": 4.597045649971411e-06, + "loss": 0.6301, + "step": 3213 + }, + { + "epoch": 1.1285112359550562, + "grad_norm": 0.8631216883659363, + "learning_rate": 4.596793918902957e-06, + "loss": 0.6264, + "step": 3214 + }, + { + "epoch": 1.1288623595505618, + "grad_norm": 0.8835992813110352, + "learning_rate": 4.596542116125947e-06, + "loss": 0.6416, + "step": 3215 + }, + { + "epoch": 1.1292134831460674, + "grad_norm": 0.8880035281181335, + "learning_rate": 4.596290241648989e-06, + "loss": 0.6072, + "step": 3216 + }, + { + "epoch": 1.129564606741573, + "grad_norm": 0.9440239667892456, + "learning_rate": 4.596038295480699e-06, + "loss": 0.6522, + "step": 3217 + }, + { + "epoch": 1.1299157303370786, + "grad_norm": 0.8469729423522949, + "learning_rate": 4.595786277629693e-06, + "loss": 0.6141, + "step": 3218 + }, + { + "epoch": 1.1302668539325842, + "grad_norm": 0.8801910281181335, + "learning_rate": 4.595534188104589e-06, + "loss": 0.648, + "step": 3219 + }, + { + "epoch": 1.13061797752809, + "grad_norm": 0.926419198513031, + "learning_rate": 4.59528202691401e-06, + "loss": 0.646, + "step": 3220 + }, + { + "epoch": 1.1309691011235956, + "grad_norm": 0.8675405383110046, + "learning_rate": 4.595029794066578e-06, + "loss": 0.626, + "step": 3221 + }, + { + "epoch": 1.1313202247191012, + "grad_norm": 0.8670094013214111, + "learning_rate": 4.594777489570919e-06, + "loss": 0.618, + "step": 3222 + }, + { + "epoch": 1.1316713483146068, + "grad_norm": 0.8746402263641357, + "learning_rate": 4.5945251134356635e-06, + "loss": 0.6172, + "step": 3223 + }, + { + "epoch": 1.1320224719101124, + "grad_norm": 0.9028223156929016, + "learning_rate": 4.594272665669442e-06, + "loss": 0.6329, + "step": 3224 + }, + { + "epoch": 1.132373595505618, + "grad_norm": 0.9109153151512146, + "learning_rate": 4.594020146280887e-06, + "loss": 0.646, + "step": 3225 + }, + { + "epoch": 1.1327247191011236, + "grad_norm": 0.8849334716796875, + "learning_rate": 4.593767555278634e-06, + "loss": 0.6491, + "step": 3226 + }, + { + "epoch": 1.1330758426966292, + "grad_norm": 0.9008032083511353, + "learning_rate": 4.593514892671323e-06, + "loss": 0.642, + "step": 3227 + }, + { + "epoch": 1.1334269662921348, + "grad_norm": 0.8774647116661072, + "learning_rate": 4.593262158467594e-06, + "loss": 0.6362, + "step": 3228 + }, + { + "epoch": 1.1337780898876404, + "grad_norm": 0.8808940649032593, + "learning_rate": 4.59300935267609e-06, + "loss": 0.6469, + "step": 3229 + }, + { + "epoch": 1.134129213483146, + "grad_norm": 0.9236598610877991, + "learning_rate": 4.592756475305458e-06, + "loss": 0.6227, + "step": 3230 + }, + { + "epoch": 1.1344803370786516, + "grad_norm": 0.8788805603981018, + "learning_rate": 4.592503526364344e-06, + "loss": 0.6168, + "step": 3231 + }, + { + "epoch": 1.1348314606741572, + "grad_norm": 1.109196424484253, + "learning_rate": 4.5922505058614016e-06, + "loss": 0.6154, + "step": 3232 + }, + { + "epoch": 1.135182584269663, + "grad_norm": 0.8943460583686829, + "learning_rate": 4.591997413805281e-06, + "loss": 0.6442, + "step": 3233 + }, + { + "epoch": 1.1355337078651686, + "grad_norm": 0.8823835849761963, + "learning_rate": 4.59174425020464e-06, + "loss": 0.6075, + "step": 3234 + }, + { + "epoch": 1.1358848314606742, + "grad_norm": 0.9072199463844299, + "learning_rate": 4.591491015068135e-06, + "loss": 0.6295, + "step": 3235 + }, + { + "epoch": 1.1362359550561798, + "grad_norm": 0.8751364946365356, + "learning_rate": 4.591237708404427e-06, + "loss": 0.6398, + "step": 3236 + }, + { + "epoch": 1.1365870786516854, + "grad_norm": 0.8724543452262878, + "learning_rate": 4.590984330222179e-06, + "loss": 0.5811, + "step": 3237 + }, + { + "epoch": 1.136938202247191, + "grad_norm": 0.9179860949516296, + "learning_rate": 4.590730880530056e-06, + "loss": 0.6846, + "step": 3238 + }, + { + "epoch": 1.1372893258426966, + "grad_norm": 0.8932046890258789, + "learning_rate": 4.590477359336726e-06, + "loss": 0.6447, + "step": 3239 + }, + { + "epoch": 1.1376404494382022, + "grad_norm": 0.9021968245506287, + "learning_rate": 4.590223766650859e-06, + "loss": 0.6503, + "step": 3240 + }, + { + "epoch": 1.1379915730337078, + "grad_norm": 0.8907482028007507, + "learning_rate": 4.5899701024811285e-06, + "loss": 0.6339, + "step": 3241 + }, + { + "epoch": 1.1383426966292134, + "grad_norm": 0.9072391390800476, + "learning_rate": 4.589716366836208e-06, + "loss": 0.65, + "step": 3242 + }, + { + "epoch": 1.1386938202247192, + "grad_norm": 0.8935176134109497, + "learning_rate": 4.5894625597247765e-06, + "loss": 0.6105, + "step": 3243 + }, + { + "epoch": 1.1390449438202248, + "grad_norm": 0.9080550670623779, + "learning_rate": 4.589208681155513e-06, + "loss": 0.6641, + "step": 3244 + }, + { + "epoch": 1.1393960674157304, + "grad_norm": 0.8920962810516357, + "learning_rate": 4.5889547311371006e-06, + "loss": 0.6482, + "step": 3245 + }, + { + "epoch": 1.139747191011236, + "grad_norm": 0.868358850479126, + "learning_rate": 4.588700709678224e-06, + "loss": 0.615, + "step": 3246 + }, + { + "epoch": 1.1400983146067416, + "grad_norm": 0.9104098677635193, + "learning_rate": 4.588446616787571e-06, + "loss": 0.6472, + "step": 3247 + }, + { + "epoch": 1.1404494382022472, + "grad_norm": 0.9489714503288269, + "learning_rate": 4.58819245247383e-06, + "loss": 0.6441, + "step": 3248 + }, + { + "epoch": 1.1408005617977528, + "grad_norm": 0.8975799679756165, + "learning_rate": 4.587938216745694e-06, + "loss": 0.6469, + "step": 3249 + }, + { + "epoch": 1.1411516853932584, + "grad_norm": 0.9056989550590515, + "learning_rate": 4.587683909611858e-06, + "loss": 0.6643, + "step": 3250 + }, + { + "epoch": 1.141502808988764, + "grad_norm": 0.890923023223877, + "learning_rate": 4.587429531081019e-06, + "loss": 0.6421, + "step": 3251 + }, + { + "epoch": 1.1418539325842696, + "grad_norm": 0.8975485563278198, + "learning_rate": 4.587175081161875e-06, + "loss": 0.6422, + "step": 3252 + }, + { + "epoch": 1.1422050561797752, + "grad_norm": 0.9034876823425293, + "learning_rate": 4.586920559863131e-06, + "loss": 0.6406, + "step": 3253 + }, + { + "epoch": 1.1425561797752808, + "grad_norm": 0.8907468318939209, + "learning_rate": 4.586665967193489e-06, + "loss": 0.6191, + "step": 3254 + }, + { + "epoch": 1.1429073033707866, + "grad_norm": 0.8739985227584839, + "learning_rate": 4.586411303161656e-06, + "loss": 0.6278, + "step": 3255 + }, + { + "epoch": 1.1432584269662922, + "grad_norm": 0.866232693195343, + "learning_rate": 4.586156567776343e-06, + "loss": 0.6497, + "step": 3256 + }, + { + "epoch": 1.1436095505617978, + "grad_norm": 0.8944258093833923, + "learning_rate": 4.585901761046259e-06, + "loss": 0.637, + "step": 3257 + }, + { + "epoch": 1.1439606741573034, + "grad_norm": 0.8632298707962036, + "learning_rate": 4.585646882980121e-06, + "loss": 0.6197, + "step": 3258 + }, + { + "epoch": 1.144311797752809, + "grad_norm": 0.8586609363555908, + "learning_rate": 4.585391933586644e-06, + "loss": 0.6163, + "step": 3259 + }, + { + "epoch": 1.1446629213483146, + "grad_norm": 0.8893579244613647, + "learning_rate": 4.585136912874547e-06, + "loss": 0.6173, + "step": 3260 + }, + { + "epoch": 1.1450140449438202, + "grad_norm": 0.8902631998062134, + "learning_rate": 4.584881820852551e-06, + "loss": 0.644, + "step": 3261 + }, + { + "epoch": 1.1453651685393258, + "grad_norm": 0.8647865653038025, + "learning_rate": 4.584626657529382e-06, + "loss": 0.6156, + "step": 3262 + }, + { + "epoch": 1.1457162921348314, + "grad_norm": 0.8853922486305237, + "learning_rate": 4.584371422913765e-06, + "loss": 0.6659, + "step": 3263 + }, + { + "epoch": 1.146067415730337, + "grad_norm": 0.8677497506141663, + "learning_rate": 4.584116117014429e-06, + "loss": 0.6403, + "step": 3264 + }, + { + "epoch": 1.1464185393258428, + "grad_norm": 0.8832502961158752, + "learning_rate": 4.583860739840105e-06, + "loss": 0.6243, + "step": 3265 + }, + { + "epoch": 1.1467696629213484, + "grad_norm": 0.952691912651062, + "learning_rate": 4.583605291399527e-06, + "loss": 0.6237, + "step": 3266 + }, + { + "epoch": 1.147120786516854, + "grad_norm": 0.8712973594665527, + "learning_rate": 4.5833497717014306e-06, + "loss": 0.6303, + "step": 3267 + }, + { + "epoch": 1.1474719101123596, + "grad_norm": 0.9440521597862244, + "learning_rate": 4.583094180754556e-06, + "loss": 0.6276, + "step": 3268 + }, + { + "epoch": 1.1478230337078652, + "grad_norm": 0.862451434135437, + "learning_rate": 4.582838518567641e-06, + "loss": 0.6363, + "step": 3269 + }, + { + "epoch": 1.1481741573033708, + "grad_norm": 0.8932682275772095, + "learning_rate": 4.582582785149432e-06, + "loss": 0.6508, + "step": 3270 + }, + { + "epoch": 1.1485252808988764, + "grad_norm": 0.9098778963088989, + "learning_rate": 4.582326980508675e-06, + "loss": 0.6377, + "step": 3271 + }, + { + "epoch": 1.148876404494382, + "grad_norm": 0.9052352905273438, + "learning_rate": 4.582071104654115e-06, + "loss": 0.6199, + "step": 3272 + }, + { + "epoch": 1.1492275280898876, + "grad_norm": 0.8726509213447571, + "learning_rate": 4.581815157594507e-06, + "loss": 0.6352, + "step": 3273 + }, + { + "epoch": 1.1495786516853932, + "grad_norm": 0.8598045110702515, + "learning_rate": 4.5815591393386e-06, + "loss": 0.6432, + "step": 3274 + }, + { + "epoch": 1.1499297752808988, + "grad_norm": 0.9145620465278625, + "learning_rate": 4.581303049895153e-06, + "loss": 0.6256, + "step": 3275 + }, + { + "epoch": 1.1502808988764044, + "grad_norm": 0.9343267679214478, + "learning_rate": 4.5810468892729225e-06, + "loss": 0.6535, + "step": 3276 + }, + { + "epoch": 1.15063202247191, + "grad_norm": 0.891581654548645, + "learning_rate": 4.580790657480668e-06, + "loss": 0.7085, + "step": 3277 + }, + { + "epoch": 1.1509831460674158, + "grad_norm": 0.8569180369377136, + "learning_rate": 4.5805343545271555e-06, + "loss": 0.6165, + "step": 3278 + }, + { + "epoch": 1.1513342696629214, + "grad_norm": 0.856910765171051, + "learning_rate": 4.5802779804211475e-06, + "loss": 0.6347, + "step": 3279 + }, + { + "epoch": 1.151685393258427, + "grad_norm": 0.8790578842163086, + "learning_rate": 4.5800215351714124e-06, + "loss": 0.6518, + "step": 3280 + }, + { + "epoch": 1.1520365168539326, + "grad_norm": 0.8809064626693726, + "learning_rate": 4.5797650187867205e-06, + "loss": 0.654, + "step": 3281 + }, + { + "epoch": 1.1523876404494382, + "grad_norm": 0.8870155215263367, + "learning_rate": 4.579508431275845e-06, + "loss": 0.651, + "step": 3282 + }, + { + "epoch": 1.1527387640449438, + "grad_norm": 0.8868922591209412, + "learning_rate": 4.57925177264756e-06, + "loss": 0.6077, + "step": 3283 + }, + { + "epoch": 1.1530898876404494, + "grad_norm": 0.8815183639526367, + "learning_rate": 4.578995042910645e-06, + "loss": 0.6332, + "step": 3284 + }, + { + "epoch": 1.153441011235955, + "grad_norm": 0.8861362934112549, + "learning_rate": 4.578738242073878e-06, + "loss": 0.64, + "step": 3285 + }, + { + "epoch": 1.1537921348314606, + "grad_norm": 0.8777415156364441, + "learning_rate": 4.5784813701460415e-06, + "loss": 0.6357, + "step": 3286 + }, + { + "epoch": 1.1541432584269664, + "grad_norm": 0.924778163433075, + "learning_rate": 4.578224427135921e-06, + "loss": 0.6345, + "step": 3287 + }, + { + "epoch": 1.154494382022472, + "grad_norm": 0.8765180110931396, + "learning_rate": 4.577967413052303e-06, + "loss": 0.6345, + "step": 3288 + }, + { + "epoch": 1.1548455056179776, + "grad_norm": 0.8863521814346313, + "learning_rate": 4.577710327903978e-06, + "loss": 0.6391, + "step": 3289 + }, + { + "epoch": 1.1551966292134832, + "grad_norm": 0.8893427848815918, + "learning_rate": 4.5774531716997384e-06, + "loss": 0.6325, + "step": 3290 + }, + { + "epoch": 1.1555477528089888, + "grad_norm": 0.8832981586456299, + "learning_rate": 4.577195944448377e-06, + "loss": 0.625, + "step": 3291 + }, + { + "epoch": 1.1558988764044944, + "grad_norm": 0.8558886647224426, + "learning_rate": 4.576938646158693e-06, + "loss": 0.5583, + "step": 3292 + }, + { + "epoch": 1.15625, + "grad_norm": 0.9013493657112122, + "learning_rate": 4.576681276839483e-06, + "loss": 0.641, + "step": 3293 + }, + { + "epoch": 1.1566011235955056, + "grad_norm": 0.8543421626091003, + "learning_rate": 4.576423836499552e-06, + "loss": 0.6452, + "step": 3294 + }, + { + "epoch": 1.1569522471910112, + "grad_norm": 0.8888267874717712, + "learning_rate": 4.5761663251477006e-06, + "loss": 0.6525, + "step": 3295 + }, + { + "epoch": 1.1573033707865168, + "grad_norm": 0.8754894733428955, + "learning_rate": 4.575908742792739e-06, + "loss": 0.631, + "step": 3296 + }, + { + "epoch": 1.1576544943820224, + "grad_norm": 0.8829214572906494, + "learning_rate": 4.575651089443474e-06, + "loss": 0.6478, + "step": 3297 + }, + { + "epoch": 1.158005617977528, + "grad_norm": 0.8613840937614441, + "learning_rate": 4.575393365108719e-06, + "loss": 0.5818, + "step": 3298 + }, + { + "epoch": 1.1583567415730336, + "grad_norm": 0.850951611995697, + "learning_rate": 4.575135569797285e-06, + "loss": 0.6349, + "step": 3299 + }, + { + "epoch": 1.1587078651685394, + "grad_norm": 0.9041875004768372, + "learning_rate": 4.574877703517992e-06, + "loss": 0.6544, + "step": 3300 + }, + { + "epoch": 1.159058988764045, + "grad_norm": 0.8524355888366699, + "learning_rate": 4.5746197662796565e-06, + "loss": 0.6145, + "step": 3301 + }, + { + "epoch": 1.1594101123595506, + "grad_norm": 0.8648415803909302, + "learning_rate": 4.5743617580911e-06, + "loss": 0.6221, + "step": 3302 + }, + { + "epoch": 1.1597612359550562, + "grad_norm": 0.8579859733581543, + "learning_rate": 4.574103678961147e-06, + "loss": 0.6011, + "step": 3303 + }, + { + "epoch": 1.1601123595505618, + "grad_norm": 0.8796312808990479, + "learning_rate": 4.5738455288986225e-06, + "loss": 0.6453, + "step": 3304 + }, + { + "epoch": 1.1604634831460674, + "grad_norm": 0.871971845626831, + "learning_rate": 4.573587307912356e-06, + "loss": 0.6096, + "step": 3305 + }, + { + "epoch": 1.160814606741573, + "grad_norm": 0.8905230760574341, + "learning_rate": 4.573329016011177e-06, + "loss": 0.6272, + "step": 3306 + }, + { + "epoch": 1.1611657303370786, + "grad_norm": 0.8950706720352173, + "learning_rate": 4.57307065320392e-06, + "loss": 0.6285, + "step": 3307 + }, + { + "epoch": 1.1615168539325842, + "grad_norm": 0.8785974979400635, + "learning_rate": 4.572812219499422e-06, + "loss": 0.6175, + "step": 3308 + }, + { + "epoch": 1.16186797752809, + "grad_norm": 0.8762933015823364, + "learning_rate": 4.572553714906519e-06, + "loss": 0.6128, + "step": 3309 + }, + { + "epoch": 1.1622191011235956, + "grad_norm": 0.9125222563743591, + "learning_rate": 4.572295139434052e-06, + "loss": 0.634, + "step": 3310 + }, + { + "epoch": 1.1625702247191012, + "grad_norm": 0.8794988989830017, + "learning_rate": 4.572036493090864e-06, + "loss": 0.6494, + "step": 3311 + }, + { + "epoch": 1.1629213483146068, + "grad_norm": 0.901485800743103, + "learning_rate": 4.571777775885803e-06, + "loss": 0.6836, + "step": 3312 + }, + { + "epoch": 1.1632724719101124, + "grad_norm": 0.9252587556838989, + "learning_rate": 4.5715189878277135e-06, + "loss": 0.6082, + "step": 3313 + }, + { + "epoch": 1.163623595505618, + "grad_norm": 0.8742091655731201, + "learning_rate": 4.571260128925448e-06, + "loss": 0.6542, + "step": 3314 + }, + { + "epoch": 1.1639747191011236, + "grad_norm": 0.8586685657501221, + "learning_rate": 4.571001199187857e-06, + "loss": 0.6165, + "step": 3315 + }, + { + "epoch": 1.1643258426966292, + "grad_norm": 0.8647174835205078, + "learning_rate": 4.570742198623798e-06, + "loss": 0.6262, + "step": 3316 + }, + { + "epoch": 1.1646769662921348, + "grad_norm": 0.8538329601287842, + "learning_rate": 4.570483127242129e-06, + "loss": 0.6523, + "step": 3317 + }, + { + "epoch": 1.1650280898876404, + "grad_norm": 0.8604421019554138, + "learning_rate": 4.570223985051707e-06, + "loss": 0.6367, + "step": 3318 + }, + { + "epoch": 1.165379213483146, + "grad_norm": 0.9220116138458252, + "learning_rate": 4.569964772061397e-06, + "loss": 0.6565, + "step": 3319 + }, + { + "epoch": 1.1657303370786516, + "grad_norm": 0.8763860464096069, + "learning_rate": 4.569705488280063e-06, + "loss": 0.6075, + "step": 3320 + }, + { + "epoch": 1.1660814606741572, + "grad_norm": 0.8693764805793762, + "learning_rate": 4.569446133716572e-06, + "loss": 0.596, + "step": 3321 + }, + { + "epoch": 1.166432584269663, + "grad_norm": 0.8876716494560242, + "learning_rate": 4.569186708379795e-06, + "loss": 0.6617, + "step": 3322 + }, + { + "epoch": 1.1667837078651686, + "grad_norm": 0.9221491813659668, + "learning_rate": 4.568927212278602e-06, + "loss": 0.6395, + "step": 3323 + }, + { + "epoch": 1.1671348314606742, + "grad_norm": 0.9208928346633911, + "learning_rate": 4.56866764542187e-06, + "loss": 0.6224, + "step": 3324 + }, + { + "epoch": 1.1674859550561798, + "grad_norm": 0.8737736344337463, + "learning_rate": 4.5684080078184745e-06, + "loss": 0.6247, + "step": 3325 + }, + { + "epoch": 1.1678370786516854, + "grad_norm": 0.8767399787902832, + "learning_rate": 4.568148299477294e-06, + "loss": 0.6394, + "step": 3326 + }, + { + "epoch": 1.168188202247191, + "grad_norm": 0.8892948031425476, + "learning_rate": 4.567888520407212e-06, + "loss": 0.6176, + "step": 3327 + }, + { + "epoch": 1.1685393258426966, + "grad_norm": 0.858261227607727, + "learning_rate": 4.567628670617113e-06, + "loss": 0.611, + "step": 3328 + }, + { + "epoch": 1.1688904494382022, + "grad_norm": 0.8876168131828308, + "learning_rate": 4.567368750115882e-06, + "loss": 0.6314, + "step": 3329 + }, + { + "epoch": 1.1692415730337078, + "grad_norm": 0.9200357794761658, + "learning_rate": 4.567108758912408e-06, + "loss": 0.6621, + "step": 3330 + }, + { + "epoch": 1.1695926966292134, + "grad_norm": 0.8937010765075684, + "learning_rate": 4.5668486970155845e-06, + "loss": 0.6381, + "step": 3331 + }, + { + "epoch": 1.1699438202247192, + "grad_norm": 0.9102723002433777, + "learning_rate": 4.566588564434304e-06, + "loss": 0.6355, + "step": 3332 + }, + { + "epoch": 1.1702949438202248, + "grad_norm": 0.8944071531295776, + "learning_rate": 4.566328361177463e-06, + "loss": 0.618, + "step": 3333 + }, + { + "epoch": 1.1706460674157304, + "grad_norm": 0.9432087540626526, + "learning_rate": 4.56606808725396e-06, + "loss": 0.6654, + "step": 3334 + }, + { + "epoch": 1.170997191011236, + "grad_norm": 0.8926756978034973, + "learning_rate": 4.5658077426726965e-06, + "loss": 0.6635, + "step": 3335 + }, + { + "epoch": 1.1713483146067416, + "grad_norm": 0.9085322022438049, + "learning_rate": 4.5655473274425756e-06, + "loss": 0.625, + "step": 3336 + }, + { + "epoch": 1.1716994382022472, + "grad_norm": 0.8761731386184692, + "learning_rate": 4.565286841572504e-06, + "loss": 0.6263, + "step": 3337 + }, + { + "epoch": 1.1720505617977528, + "grad_norm": 0.902802586555481, + "learning_rate": 4.5650262850713895e-06, + "loss": 0.5946, + "step": 3338 + }, + { + "epoch": 1.1724016853932584, + "grad_norm": 0.9183904528617859, + "learning_rate": 4.564765657948143e-06, + "loss": 0.6103, + "step": 3339 + }, + { + "epoch": 1.172752808988764, + "grad_norm": 0.8977819085121155, + "learning_rate": 4.564504960211678e-06, + "loss": 0.6408, + "step": 3340 + }, + { + "epoch": 1.1731039325842696, + "grad_norm": 0.9143522381782532, + "learning_rate": 4.56424419187091e-06, + "loss": 0.6379, + "step": 3341 + }, + { + "epoch": 1.1734550561797752, + "grad_norm": 0.9088187217712402, + "learning_rate": 4.563983352934758e-06, + "loss": 0.6473, + "step": 3342 + }, + { + "epoch": 1.1738061797752808, + "grad_norm": 0.8903788328170776, + "learning_rate": 4.563722443412139e-06, + "loss": 0.5842, + "step": 3343 + }, + { + "epoch": 1.1741573033707866, + "grad_norm": 0.8306713700294495, + "learning_rate": 4.563461463311981e-06, + "loss": 0.5916, + "step": 3344 + }, + { + "epoch": 1.1745084269662922, + "grad_norm": 0.8774393200874329, + "learning_rate": 4.5632004126432055e-06, + "loss": 0.628, + "step": 3345 + }, + { + "epoch": 1.1748595505617978, + "grad_norm": 0.8590057492256165, + "learning_rate": 4.5629392914147415e-06, + "loss": 0.6111, + "step": 3346 + }, + { + "epoch": 1.1752106741573034, + "grad_norm": 0.8803392648696899, + "learning_rate": 4.562678099635519e-06, + "loss": 0.6313, + "step": 3347 + }, + { + "epoch": 1.175561797752809, + "grad_norm": 0.8844352960586548, + "learning_rate": 4.562416837314471e-06, + "loss": 0.6392, + "step": 3348 + }, + { + "epoch": 1.1759129213483146, + "grad_norm": 0.9030863642692566, + "learning_rate": 4.5621555044605315e-06, + "loss": 0.6407, + "step": 3349 + }, + { + "epoch": 1.1762640449438202, + "grad_norm": 0.8938463926315308, + "learning_rate": 4.561894101082638e-06, + "loss": 0.6224, + "step": 3350 + }, + { + "epoch": 1.1766151685393258, + "grad_norm": 0.869136393070221, + "learning_rate": 4.5616326271897315e-06, + "loss": 0.6239, + "step": 3351 + }, + { + "epoch": 1.1769662921348314, + "grad_norm": 0.8866480588912964, + "learning_rate": 4.5613710827907535e-06, + "loss": 0.6319, + "step": 3352 + }, + { + "epoch": 1.177317415730337, + "grad_norm": 0.911381185054779, + "learning_rate": 4.561109467894647e-06, + "loss": 0.6779, + "step": 3353 + }, + { + "epoch": 1.1776685393258428, + "grad_norm": 0.8881227374076843, + "learning_rate": 4.560847782510361e-06, + "loss": 0.6174, + "step": 3354 + }, + { + "epoch": 1.1780196629213484, + "grad_norm": 0.8753611445426941, + "learning_rate": 4.560586026646845e-06, + "loss": 0.6337, + "step": 3355 + }, + { + "epoch": 1.178370786516854, + "grad_norm": 0.898257315158844, + "learning_rate": 4.56032420031305e-06, + "loss": 0.6547, + "step": 3356 + }, + { + "epoch": 1.1787219101123596, + "grad_norm": 0.871589720249176, + "learning_rate": 4.56006230351793e-06, + "loss": 0.6093, + "step": 3357 + }, + { + "epoch": 1.1790730337078652, + "grad_norm": 0.8686171174049377, + "learning_rate": 4.559800336270443e-06, + "loss": 0.6396, + "step": 3358 + }, + { + "epoch": 1.1794241573033708, + "grad_norm": 0.8978163599967957, + "learning_rate": 4.559538298579546e-06, + "loss": 0.6505, + "step": 3359 + }, + { + "epoch": 1.1797752808988764, + "grad_norm": 0.9143759608268738, + "learning_rate": 4.559276190454201e-06, + "loss": 0.6452, + "step": 3360 + }, + { + "epoch": 1.180126404494382, + "grad_norm": 0.9073001146316528, + "learning_rate": 4.559014011903373e-06, + "loss": 0.6252, + "step": 3361 + }, + { + "epoch": 1.1804775280898876, + "grad_norm": 0.9082154035568237, + "learning_rate": 4.558751762936029e-06, + "loss": 0.6401, + "step": 3362 + }, + { + "epoch": 1.1808286516853932, + "grad_norm": 0.949808657169342, + "learning_rate": 4.558489443561135e-06, + "loss": 0.6716, + "step": 3363 + }, + { + "epoch": 1.1811797752808988, + "grad_norm": 0.8602761626243591, + "learning_rate": 4.558227053787664e-06, + "loss": 0.6112, + "step": 3364 + }, + { + "epoch": 1.1815308988764044, + "grad_norm": 0.9209701418876648, + "learning_rate": 4.5579645936245886e-06, + "loss": 0.6205, + "step": 3365 + }, + { + "epoch": 1.18188202247191, + "grad_norm": 0.8840071558952332, + "learning_rate": 4.557702063080886e-06, + "loss": 0.6242, + "step": 3366 + }, + { + "epoch": 1.1822331460674158, + "grad_norm": 0.906648576259613, + "learning_rate": 4.557439462165532e-06, + "loss": 0.6455, + "step": 3367 + }, + { + "epoch": 1.1825842696629214, + "grad_norm": 0.8441259264945984, + "learning_rate": 4.55717679088751e-06, + "loss": 0.5916, + "step": 3368 + }, + { + "epoch": 1.182935393258427, + "grad_norm": 0.9262234568595886, + "learning_rate": 4.556914049255803e-06, + "loss": 0.659, + "step": 3369 + }, + { + "epoch": 1.1832865168539326, + "grad_norm": 0.8849537372589111, + "learning_rate": 4.556651237279394e-06, + "loss": 0.6503, + "step": 3370 + }, + { + "epoch": 1.1836376404494382, + "grad_norm": 0.8733585476875305, + "learning_rate": 4.5563883549672735e-06, + "loss": 0.6309, + "step": 3371 + }, + { + "epoch": 1.1839887640449438, + "grad_norm": 0.91927570104599, + "learning_rate": 4.556125402328431e-06, + "loss": 0.6148, + "step": 3372 + }, + { + "epoch": 1.1843398876404494, + "grad_norm": 0.9334448575973511, + "learning_rate": 4.555862379371858e-06, + "loss": 0.6483, + "step": 3373 + }, + { + "epoch": 1.184691011235955, + "grad_norm": 0.8952125906944275, + "learning_rate": 4.555599286106553e-06, + "loss": 0.6218, + "step": 3374 + }, + { + "epoch": 1.1850421348314606, + "grad_norm": 0.8810850381851196, + "learning_rate": 4.55533612254151e-06, + "loss": 0.6391, + "step": 3375 + }, + { + "epoch": 1.1853932584269664, + "grad_norm": 0.9044777154922485, + "learning_rate": 4.5550728886857305e-06, + "loss": 0.6462, + "step": 3376 + }, + { + "epoch": 1.185744382022472, + "grad_norm": 0.9001294374465942, + "learning_rate": 4.554809584548217e-06, + "loss": 0.609, + "step": 3377 + }, + { + "epoch": 1.1860955056179776, + "grad_norm": 0.8865534663200378, + "learning_rate": 4.554546210137973e-06, + "loss": 0.6016, + "step": 3378 + }, + { + "epoch": 1.1864466292134832, + "grad_norm": 0.8913449645042419, + "learning_rate": 4.554282765464009e-06, + "loss": 0.6301, + "step": 3379 + }, + { + "epoch": 1.1867977528089888, + "grad_norm": 0.8913347125053406, + "learning_rate": 4.55401925053533e-06, + "loss": 0.6369, + "step": 3380 + }, + { + "epoch": 1.1871488764044944, + "grad_norm": 0.9004356265068054, + "learning_rate": 4.553755665360952e-06, + "loss": 0.646, + "step": 3381 + }, + { + "epoch": 1.1875, + "grad_norm": 0.85580974817276, + "learning_rate": 4.553492009949886e-06, + "loss": 0.6096, + "step": 3382 + }, + { + "epoch": 1.1878511235955056, + "grad_norm": 0.860244631767273, + "learning_rate": 4.553228284311152e-06, + "loss": 0.6043, + "step": 3383 + }, + { + "epoch": 1.1882022471910112, + "grad_norm": 0.8783630132675171, + "learning_rate": 4.552964488453767e-06, + "loss": 0.6359, + "step": 3384 + }, + { + "epoch": 1.1885533707865168, + "grad_norm": 0.8642095923423767, + "learning_rate": 4.552700622386753e-06, + "loss": 0.6038, + "step": 3385 + }, + { + "epoch": 1.1889044943820224, + "grad_norm": 0.867074728012085, + "learning_rate": 4.5524366861191345e-06, + "loss": 0.6415, + "step": 3386 + }, + { + "epoch": 1.189255617977528, + "grad_norm": 0.8984413743019104, + "learning_rate": 4.552172679659938e-06, + "loss": 0.6519, + "step": 3387 + }, + { + "epoch": 1.1896067415730336, + "grad_norm": 0.869999885559082, + "learning_rate": 4.551908603018191e-06, + "loss": 0.6269, + "step": 3388 + }, + { + "epoch": 1.1899578651685394, + "grad_norm": 0.8916454911231995, + "learning_rate": 4.551644456202926e-06, + "loss": 0.6486, + "step": 3389 + }, + { + "epoch": 1.190308988764045, + "grad_norm": 0.888932466506958, + "learning_rate": 4.551380239223176e-06, + "loss": 0.6084, + "step": 3390 + }, + { + "epoch": 1.1906601123595506, + "grad_norm": 0.9049292802810669, + "learning_rate": 4.5511159520879785e-06, + "loss": 0.6134, + "step": 3391 + }, + { + "epoch": 1.1910112359550562, + "grad_norm": 0.8797271847724915, + "learning_rate": 4.550851594806369e-06, + "loss": 0.6367, + "step": 3392 + }, + { + "epoch": 1.1913623595505618, + "grad_norm": 0.9155986309051514, + "learning_rate": 4.550587167387391e-06, + "loss": 0.6415, + "step": 3393 + }, + { + "epoch": 1.1917134831460674, + "grad_norm": 0.8914098143577576, + "learning_rate": 4.550322669840086e-06, + "loss": 0.6352, + "step": 3394 + }, + { + "epoch": 1.192064606741573, + "grad_norm": 0.8818766474723816, + "learning_rate": 4.550058102173501e-06, + "loss": 0.6219, + "step": 3395 + }, + { + "epoch": 1.1924157303370786, + "grad_norm": 0.8712998628616333, + "learning_rate": 4.5497934643966825e-06, + "loss": 0.5773, + "step": 3396 + }, + { + "epoch": 1.1927668539325842, + "grad_norm": 0.8595727682113647, + "learning_rate": 4.5495287565186806e-06, + "loss": 0.6171, + "step": 3397 + }, + { + "epoch": 1.19311797752809, + "grad_norm": 0.8803826570510864, + "learning_rate": 4.549263978548551e-06, + "loss": 0.5942, + "step": 3398 + }, + { + "epoch": 1.1934691011235956, + "grad_norm": 0.8828043937683105, + "learning_rate": 4.548999130495345e-06, + "loss": 0.651, + "step": 3399 + }, + { + "epoch": 1.1938202247191012, + "grad_norm": 0.8621920347213745, + "learning_rate": 4.548734212368123e-06, + "loss": 0.5715, + "step": 3400 + }, + { + "epoch": 1.1941713483146068, + "grad_norm": 0.9156777858734131, + "learning_rate": 4.548469224175944e-06, + "loss": 0.6669, + "step": 3401 + }, + { + "epoch": 1.1945224719101124, + "grad_norm": 0.8646230101585388, + "learning_rate": 4.548204165927871e-06, + "loss": 0.5863, + "step": 3402 + }, + { + "epoch": 1.194873595505618, + "grad_norm": 0.9154274463653564, + "learning_rate": 4.5479390376329676e-06, + "loss": 0.6236, + "step": 3403 + }, + { + "epoch": 1.1952247191011236, + "grad_norm": 0.8905937671661377, + "learning_rate": 4.547673839300301e-06, + "loss": 0.5949, + "step": 3404 + }, + { + "epoch": 1.1955758426966292, + "grad_norm": 0.9460303783416748, + "learning_rate": 4.547408570938942e-06, + "loss": 0.6367, + "step": 3405 + }, + { + "epoch": 1.1959269662921348, + "grad_norm": 0.8888306617736816, + "learning_rate": 4.547143232557962e-06, + "loss": 0.6533, + "step": 3406 + }, + { + "epoch": 1.1962780898876404, + "grad_norm": 0.8990828394889832, + "learning_rate": 4.546877824166436e-06, + "loss": 0.6418, + "step": 3407 + }, + { + "epoch": 1.196629213483146, + "grad_norm": 0.9120289087295532, + "learning_rate": 4.5466123457734385e-06, + "loss": 0.6308, + "step": 3408 + }, + { + "epoch": 1.1969803370786516, + "grad_norm": 0.8693567514419556, + "learning_rate": 4.546346797388051e-06, + "loss": 0.6784, + "step": 3409 + }, + { + "epoch": 1.1973314606741572, + "grad_norm": 0.8695741295814514, + "learning_rate": 4.546081179019353e-06, + "loss": 0.6344, + "step": 3410 + }, + { + "epoch": 1.197682584269663, + "grad_norm": 0.8742298483848572, + "learning_rate": 4.545815490676431e-06, + "loss": 0.584, + "step": 3411 + }, + { + "epoch": 1.1980337078651686, + "grad_norm": 0.8857693076133728, + "learning_rate": 4.545549732368368e-06, + "loss": 0.6591, + "step": 3412 + }, + { + "epoch": 1.1983848314606742, + "grad_norm": 0.887653648853302, + "learning_rate": 4.545283904104257e-06, + "loss": 0.6695, + "step": 3413 + }, + { + "epoch": 1.1987359550561798, + "grad_norm": 0.9010708332061768, + "learning_rate": 4.545018005893185e-06, + "loss": 0.6184, + "step": 3414 + }, + { + "epoch": 1.1990870786516854, + "grad_norm": 0.9171410799026489, + "learning_rate": 4.544752037744247e-06, + "loss": 0.6684, + "step": 3415 + }, + { + "epoch": 1.199438202247191, + "grad_norm": 0.8993377685546875, + "learning_rate": 4.5444859996665395e-06, + "loss": 0.6397, + "step": 3416 + }, + { + "epoch": 1.1997893258426966, + "grad_norm": 0.8996485471725464, + "learning_rate": 4.54421989166916e-06, + "loss": 0.6625, + "step": 3417 + }, + { + "epoch": 1.2001404494382022, + "grad_norm": 0.8873034715652466, + "learning_rate": 4.54395371376121e-06, + "loss": 0.5966, + "step": 3418 + }, + { + "epoch": 1.2004915730337078, + "grad_norm": 0.9241093397140503, + "learning_rate": 4.543687465951791e-06, + "loss": 0.6443, + "step": 3419 + }, + { + "epoch": 1.2008426966292134, + "grad_norm": 0.9346911907196045, + "learning_rate": 4.5434211482500116e-06, + "loss": 0.6544, + "step": 3420 + }, + { + "epoch": 1.2011938202247192, + "grad_norm": 0.8910781741142273, + "learning_rate": 4.543154760664975e-06, + "loss": 0.6357, + "step": 3421 + }, + { + "epoch": 1.2015449438202248, + "grad_norm": 0.8794376850128174, + "learning_rate": 4.542888303205796e-06, + "loss": 0.642, + "step": 3422 + }, + { + "epoch": 1.2018960674157304, + "grad_norm": 0.8754070401191711, + "learning_rate": 4.542621775881584e-06, + "loss": 0.6145, + "step": 3423 + }, + { + "epoch": 1.202247191011236, + "grad_norm": 0.881120502948761, + "learning_rate": 4.542355178701456e-06, + "loss": 0.6168, + "step": 3424 + }, + { + "epoch": 1.2025983146067416, + "grad_norm": 0.9401150941848755, + "learning_rate": 4.5420885116745275e-06, + "loss": 0.6636, + "step": 3425 + }, + { + "epoch": 1.2029494382022472, + "grad_norm": 0.9032625555992126, + "learning_rate": 4.541821774809919e-06, + "loss": 0.6496, + "step": 3426 + }, + { + "epoch": 1.2033005617977528, + "grad_norm": 0.9384265542030334, + "learning_rate": 4.541554968116754e-06, + "loss": 0.668, + "step": 3427 + }, + { + "epoch": 1.2036516853932584, + "grad_norm": 0.8976498246192932, + "learning_rate": 4.541288091604156e-06, + "loss": 0.6287, + "step": 3428 + }, + { + "epoch": 1.204002808988764, + "grad_norm": 0.9575441479682922, + "learning_rate": 4.5410211452812515e-06, + "loss": 0.657, + "step": 3429 + }, + { + "epoch": 1.2043539325842696, + "grad_norm": 0.8660023212432861, + "learning_rate": 4.540754129157171e-06, + "loss": 0.6206, + "step": 3430 + }, + { + "epoch": 1.2047050561797752, + "grad_norm": 0.8459751605987549, + "learning_rate": 4.540487043241045e-06, + "loss": 0.6318, + "step": 3431 + }, + { + "epoch": 1.2050561797752808, + "grad_norm": 0.8574265241622925, + "learning_rate": 4.540219887542009e-06, + "loss": 0.6434, + "step": 3432 + }, + { + "epoch": 1.2054073033707866, + "grad_norm": 0.8773812055587769, + "learning_rate": 4.539952662069198e-06, + "loss": 0.6217, + "step": 3433 + }, + { + "epoch": 1.2057584269662922, + "grad_norm": 0.8617153763771057, + "learning_rate": 4.539685366831752e-06, + "loss": 0.6403, + "step": 3434 + }, + { + "epoch": 1.2061095505617978, + "grad_norm": 0.8632030487060547, + "learning_rate": 4.539418001838812e-06, + "loss": 0.6054, + "step": 3435 + }, + { + "epoch": 1.2064606741573034, + "grad_norm": 0.8510541319847107, + "learning_rate": 4.539150567099521e-06, + "loss": 0.6207, + "step": 3436 + }, + { + "epoch": 1.206811797752809, + "grad_norm": 0.8945329785346985, + "learning_rate": 4.538883062623025e-06, + "loss": 0.6413, + "step": 3437 + }, + { + "epoch": 1.2071629213483146, + "grad_norm": 0.9274084568023682, + "learning_rate": 4.538615488418474e-06, + "loss": 0.647, + "step": 3438 + }, + { + "epoch": 1.2075140449438202, + "grad_norm": 0.8531306385993958, + "learning_rate": 4.538347844495018e-06, + "loss": 0.6365, + "step": 3439 + }, + { + "epoch": 1.2078651685393258, + "grad_norm": 0.9376130700111389, + "learning_rate": 4.538080130861809e-06, + "loss": 0.6194, + "step": 3440 + }, + { + "epoch": 1.2082162921348314, + "grad_norm": 0.8854097723960876, + "learning_rate": 4.537812347528004e-06, + "loss": 0.6196, + "step": 3441 + }, + { + "epoch": 1.208567415730337, + "grad_norm": 0.8933984637260437, + "learning_rate": 4.537544494502761e-06, + "loss": 0.6661, + "step": 3442 + }, + { + "epoch": 1.2089185393258428, + "grad_norm": 0.8640488386154175, + "learning_rate": 4.537276571795238e-06, + "loss": 0.5997, + "step": 3443 + }, + { + "epoch": 1.2092696629213484, + "grad_norm": 0.8864790201187134, + "learning_rate": 4.537008579414601e-06, + "loss": 0.6444, + "step": 3444 + }, + { + "epoch": 1.209620786516854, + "grad_norm": 0.8434916138648987, + "learning_rate": 4.536740517370013e-06, + "loss": 0.6138, + "step": 3445 + }, + { + "epoch": 1.2099719101123596, + "grad_norm": 0.8927525281906128, + "learning_rate": 4.536472385670643e-06, + "loss": 0.6309, + "step": 3446 + }, + { + "epoch": 1.2103230337078652, + "grad_norm": 0.8696897625923157, + "learning_rate": 4.53620418432566e-06, + "loss": 0.655, + "step": 3447 + }, + { + "epoch": 1.2106741573033708, + "grad_norm": 0.8829634189605713, + "learning_rate": 4.535935913344236e-06, + "loss": 0.623, + "step": 3448 + }, + { + "epoch": 1.2110252808988764, + "grad_norm": 0.8809219598770142, + "learning_rate": 4.535667572735546e-06, + "loss": 0.6806, + "step": 3449 + }, + { + "epoch": 1.211376404494382, + "grad_norm": 0.8975908756256104, + "learning_rate": 4.535399162508767e-06, + "loss": 0.6567, + "step": 3450 + }, + { + "epoch": 1.2117275280898876, + "grad_norm": 0.8763452768325806, + "learning_rate": 4.535130682673078e-06, + "loss": 0.6758, + "step": 3451 + }, + { + "epoch": 1.2120786516853932, + "grad_norm": 0.8537138104438782, + "learning_rate": 4.534862133237661e-06, + "loss": 0.623, + "step": 3452 + }, + { + "epoch": 1.2124297752808988, + "grad_norm": 0.8764957189559937, + "learning_rate": 4.534593514211701e-06, + "loss": 0.661, + "step": 3453 + }, + { + "epoch": 1.2127808988764044, + "grad_norm": 0.8892989158630371, + "learning_rate": 4.534324825604383e-06, + "loss": 0.6597, + "step": 3454 + }, + { + "epoch": 1.21313202247191, + "grad_norm": 0.902550220489502, + "learning_rate": 4.534056067424898e-06, + "loss": 0.6569, + "step": 3455 + }, + { + "epoch": 1.2134831460674158, + "grad_norm": 0.872238039970398, + "learning_rate": 4.5337872396824355e-06, + "loss": 0.6272, + "step": 3456 + }, + { + "epoch": 1.2138342696629214, + "grad_norm": 0.8700047135353088, + "learning_rate": 4.533518342386189e-06, + "loss": 0.6396, + "step": 3457 + }, + { + "epoch": 1.214185393258427, + "grad_norm": 0.8908973336219788, + "learning_rate": 4.5332493755453555e-06, + "loss": 0.6108, + "step": 3458 + }, + { + "epoch": 1.2145365168539326, + "grad_norm": 0.9054476618766785, + "learning_rate": 4.532980339169134e-06, + "loss": 0.6372, + "step": 3459 + }, + { + "epoch": 1.2148876404494382, + "grad_norm": 0.8539584279060364, + "learning_rate": 4.532711233266724e-06, + "loss": 0.6418, + "step": 3460 + }, + { + "epoch": 1.2152387640449438, + "grad_norm": 0.8940576314926147, + "learning_rate": 4.532442057847329e-06, + "loss": 0.6454, + "step": 3461 + }, + { + "epoch": 1.2155898876404494, + "grad_norm": 0.9023147225379944, + "learning_rate": 4.532172812920156e-06, + "loss": 0.6301, + "step": 3462 + }, + { + "epoch": 1.215941011235955, + "grad_norm": 0.9048089385032654, + "learning_rate": 4.53190349849441e-06, + "loss": 0.6469, + "step": 3463 + }, + { + "epoch": 1.2162921348314606, + "grad_norm": 0.8976174592971802, + "learning_rate": 4.531634114579304e-06, + "loss": 0.6518, + "step": 3464 + }, + { + "epoch": 1.2166432584269664, + "grad_norm": 0.8753100633621216, + "learning_rate": 4.53136466118405e-06, + "loss": 0.6173, + "step": 3465 + }, + { + "epoch": 1.216994382022472, + "grad_norm": 0.9097129106521606, + "learning_rate": 4.531095138317862e-06, + "loss": 0.6753, + "step": 3466 + }, + { + "epoch": 1.2173455056179776, + "grad_norm": 0.8648528456687927, + "learning_rate": 4.53082554598996e-06, + "loss": 0.5863, + "step": 3467 + }, + { + "epoch": 1.2176966292134832, + "grad_norm": 0.9543090462684631, + "learning_rate": 4.530555884209561e-06, + "loss": 0.6562, + "step": 3468 + }, + { + "epoch": 1.2180477528089888, + "grad_norm": 0.8979066610336304, + "learning_rate": 4.530286152985888e-06, + "loss": 0.6224, + "step": 3469 + }, + { + "epoch": 1.2183988764044944, + "grad_norm": 0.9094343185424805, + "learning_rate": 4.530016352328167e-06, + "loss": 0.6453, + "step": 3470 + }, + { + "epoch": 1.21875, + "grad_norm": 0.861716628074646, + "learning_rate": 4.5297464822456225e-06, + "loss": 0.5886, + "step": 3471 + }, + { + "epoch": 1.2191011235955056, + "grad_norm": 0.928870439529419, + "learning_rate": 4.5294765427474865e-06, + "loss": 0.6492, + "step": 3472 + }, + { + "epoch": 1.2194522471910112, + "grad_norm": 0.8802708387374878, + "learning_rate": 4.529206533842988e-06, + "loss": 0.6405, + "step": 3473 + }, + { + "epoch": 1.2198033707865168, + "grad_norm": 1.2119126319885254, + "learning_rate": 4.528936455541364e-06, + "loss": 0.6119, + "step": 3474 + }, + { + "epoch": 1.2201544943820224, + "grad_norm": 0.8638689517974854, + "learning_rate": 4.528666307851848e-06, + "loss": 0.6467, + "step": 3475 + }, + { + "epoch": 1.220505617977528, + "grad_norm": 0.8760889172554016, + "learning_rate": 4.528396090783681e-06, + "loss": 0.6113, + "step": 3476 + }, + { + "epoch": 1.2208567415730336, + "grad_norm": 0.8935677409172058, + "learning_rate": 4.5281258043461035e-06, + "loss": 0.6029, + "step": 3477 + }, + { + "epoch": 1.2212078651685394, + "grad_norm": 0.881455659866333, + "learning_rate": 4.527855448548359e-06, + "loss": 0.6389, + "step": 3478 + }, + { + "epoch": 1.221558988764045, + "grad_norm": 0.8549000024795532, + "learning_rate": 4.527585023399693e-06, + "loss": 0.6052, + "step": 3479 + }, + { + "epoch": 1.2219101123595506, + "grad_norm": 0.8719263672828674, + "learning_rate": 4.527314528909354e-06, + "loss": 0.635, + "step": 3480 + }, + { + "epoch": 1.2222612359550562, + "grad_norm": 0.8731090426445007, + "learning_rate": 4.527043965086594e-06, + "loss": 0.643, + "step": 3481 + }, + { + "epoch": 1.2226123595505618, + "grad_norm": 0.8634929060935974, + "learning_rate": 4.5267733319406635e-06, + "loss": 0.6184, + "step": 3482 + }, + { + "epoch": 1.2229634831460674, + "grad_norm": 0.8768059611320496, + "learning_rate": 4.52650262948082e-06, + "loss": 0.636, + "step": 3483 + }, + { + "epoch": 1.223314606741573, + "grad_norm": 0.8799298405647278, + "learning_rate": 4.526231857716321e-06, + "loss": 0.6273, + "step": 3484 + }, + { + "epoch": 1.2236657303370786, + "grad_norm": 0.8915926814079285, + "learning_rate": 4.525961016656426e-06, + "loss": 0.608, + "step": 3485 + }, + { + "epoch": 1.2240168539325842, + "grad_norm": 0.8969928026199341, + "learning_rate": 4.525690106310399e-06, + "loss": 0.6404, + "step": 3486 + }, + { + "epoch": 1.22436797752809, + "grad_norm": 0.9175208210945129, + "learning_rate": 4.525419126687503e-06, + "loss": 0.6429, + "step": 3487 + }, + { + "epoch": 1.2247191011235956, + "grad_norm": 0.8877578377723694, + "learning_rate": 4.5251480777970045e-06, + "loss": 0.6268, + "step": 3488 + }, + { + "epoch": 1.2250702247191012, + "grad_norm": 0.8742526173591614, + "learning_rate": 4.524876959648176e-06, + "loss": 0.6216, + "step": 3489 + }, + { + "epoch": 1.2254213483146068, + "grad_norm": 0.8701285123825073, + "learning_rate": 4.524605772250289e-06, + "loss": 0.6139, + "step": 3490 + }, + { + "epoch": 1.2257724719101124, + "grad_norm": 0.8775697946548462, + "learning_rate": 4.5243345156126166e-06, + "loss": 0.6187, + "step": 3491 + }, + { + "epoch": 1.226123595505618, + "grad_norm": 0.8713624477386475, + "learning_rate": 4.524063189744435e-06, + "loss": 0.6025, + "step": 3492 + }, + { + "epoch": 1.2264747191011236, + "grad_norm": 0.8762874007225037, + "learning_rate": 4.5237917946550244e-06, + "loss": 0.6542, + "step": 3493 + }, + { + "epoch": 1.2268258426966292, + "grad_norm": 0.9075011610984802, + "learning_rate": 4.523520330353668e-06, + "loss": 0.6346, + "step": 3494 + }, + { + "epoch": 1.2271769662921348, + "grad_norm": 0.9128796458244324, + "learning_rate": 4.523248796849647e-06, + "loss": 0.6474, + "step": 3495 + }, + { + "epoch": 1.2275280898876404, + "grad_norm": 0.888933002948761, + "learning_rate": 4.522977194152247e-06, + "loss": 0.5965, + "step": 3496 + }, + { + "epoch": 1.227879213483146, + "grad_norm": 0.8765919804573059, + "learning_rate": 4.522705522270759e-06, + "loss": 0.644, + "step": 3497 + }, + { + "epoch": 1.2282303370786516, + "grad_norm": 0.9075853228569031, + "learning_rate": 4.5224337812144735e-06, + "loss": 0.628, + "step": 3498 + }, + { + "epoch": 1.2285814606741572, + "grad_norm": 0.9263275265693665, + "learning_rate": 4.522161970992683e-06, + "loss": 0.6301, + "step": 3499 + }, + { + "epoch": 1.228932584269663, + "grad_norm": 0.8859231472015381, + "learning_rate": 4.521890091614684e-06, + "loss": 0.6347, + "step": 3500 + }, + { + "epoch": 1.2292837078651686, + "grad_norm": 0.879749596118927, + "learning_rate": 4.521618143089773e-06, + "loss": 0.649, + "step": 3501 + }, + { + "epoch": 1.2296348314606742, + "grad_norm": 0.8523240685462952, + "learning_rate": 4.521346125427252e-06, + "loss": 0.6284, + "step": 3502 + }, + { + "epoch": 1.2299859550561798, + "grad_norm": 0.8656728863716125, + "learning_rate": 4.521074038636422e-06, + "loss": 0.6338, + "step": 3503 + }, + { + "epoch": 1.2303370786516854, + "grad_norm": 0.8819969892501831, + "learning_rate": 4.5208018827265915e-06, + "loss": 0.6169, + "step": 3504 + }, + { + "epoch": 1.230688202247191, + "grad_norm": 0.868812084197998, + "learning_rate": 4.520529657707065e-06, + "loss": 0.6601, + "step": 3505 + }, + { + "epoch": 1.2310393258426966, + "grad_norm": 0.8616583347320557, + "learning_rate": 4.520257363587152e-06, + "loss": 0.5738, + "step": 3506 + }, + { + "epoch": 1.2313904494382022, + "grad_norm": 0.8898042440414429, + "learning_rate": 4.519985000376168e-06, + "loss": 0.671, + "step": 3507 + }, + { + "epoch": 1.2317415730337078, + "grad_norm": 0.8364049196243286, + "learning_rate": 4.519712568083424e-06, + "loss": 0.63, + "step": 3508 + }, + { + "epoch": 1.2320926966292134, + "grad_norm": 0.8617914915084839, + "learning_rate": 4.519440066718239e-06, + "loss": 0.5971, + "step": 3509 + }, + { + "epoch": 1.2324438202247192, + "grad_norm": 0.8895637392997742, + "learning_rate": 4.5191674962899314e-06, + "loss": 0.5993, + "step": 3510 + }, + { + "epoch": 1.2327949438202248, + "grad_norm": 0.910457968711853, + "learning_rate": 4.518894856807824e-06, + "loss": 0.6661, + "step": 3511 + }, + { + "epoch": 1.2331460674157304, + "grad_norm": 0.8428799510002136, + "learning_rate": 4.518622148281239e-06, + "loss": 0.5736, + "step": 3512 + }, + { + "epoch": 1.233497191011236, + "grad_norm": 0.8717475533485413, + "learning_rate": 4.518349370719505e-06, + "loss": 0.6315, + "step": 3513 + }, + { + "epoch": 1.2338483146067416, + "grad_norm": 0.9214996695518494, + "learning_rate": 4.518076524131949e-06, + "loss": 0.617, + "step": 3514 + }, + { + "epoch": 1.2341994382022472, + "grad_norm": 0.88681960105896, + "learning_rate": 4.517803608527903e-06, + "loss": 0.6427, + "step": 3515 + }, + { + "epoch": 1.2345505617977528, + "grad_norm": 0.8749536275863647, + "learning_rate": 4.517530623916701e-06, + "loss": 0.6326, + "step": 3516 + }, + { + "epoch": 1.2349016853932584, + "grad_norm": 0.8672427535057068, + "learning_rate": 4.517257570307676e-06, + "loss": 0.6264, + "step": 3517 + }, + { + "epoch": 1.235252808988764, + "grad_norm": 0.8522306680679321, + "learning_rate": 4.5169844477101695e-06, + "loss": 0.591, + "step": 3518 + }, + { + "epoch": 1.2356039325842696, + "grad_norm": 0.9023152589797974, + "learning_rate": 4.516711256133522e-06, + "loss": 0.641, + "step": 3519 + }, + { + "epoch": 1.2359550561797752, + "grad_norm": 0.8903506398200989, + "learning_rate": 4.516437995587074e-06, + "loss": 0.6356, + "step": 3520 + }, + { + "epoch": 1.2363061797752808, + "grad_norm": 0.8734146356582642, + "learning_rate": 4.516164666080172e-06, + "loss": 0.6365, + "step": 3521 + }, + { + "epoch": 1.2366573033707866, + "grad_norm": 0.8641245365142822, + "learning_rate": 4.5158912676221635e-06, + "loss": 0.6536, + "step": 3522 + }, + { + "epoch": 1.2370084269662922, + "grad_norm": 0.8999871015548706, + "learning_rate": 4.515617800222399e-06, + "loss": 0.6096, + "step": 3523 + }, + { + "epoch": 1.2373595505617978, + "grad_norm": 0.9102576971054077, + "learning_rate": 4.515344263890231e-06, + "loss": 0.6285, + "step": 3524 + }, + { + "epoch": 1.2377106741573034, + "grad_norm": 0.8963742256164551, + "learning_rate": 4.515070658635013e-06, + "loss": 0.6625, + "step": 3525 + }, + { + "epoch": 1.238061797752809, + "grad_norm": 0.9126663208007812, + "learning_rate": 4.514796984466103e-06, + "loss": 0.6398, + "step": 3526 + }, + { + "epoch": 1.2384129213483146, + "grad_norm": 0.8794665336608887, + "learning_rate": 4.514523241392861e-06, + "loss": 0.6545, + "step": 3527 + }, + { + "epoch": 1.2387640449438202, + "grad_norm": 0.9098983407020569, + "learning_rate": 4.514249429424647e-06, + "loss": 0.6394, + "step": 3528 + }, + { + "epoch": 1.2391151685393258, + "grad_norm": 0.8858410120010376, + "learning_rate": 4.513975548570826e-06, + "loss": 0.6219, + "step": 3529 + }, + { + "epoch": 1.2394662921348314, + "grad_norm": 0.9184651970863342, + "learning_rate": 4.5137015988407655e-06, + "loss": 0.647, + "step": 3530 + }, + { + "epoch": 1.239817415730337, + "grad_norm": 0.9131940603256226, + "learning_rate": 4.513427580243833e-06, + "loss": 0.615, + "step": 3531 + }, + { + "epoch": 1.2401685393258428, + "grad_norm": 0.8828357458114624, + "learning_rate": 4.5131534927894e-06, + "loss": 0.6288, + "step": 3532 + }, + { + "epoch": 1.2405196629213484, + "grad_norm": 0.8712863922119141, + "learning_rate": 4.512879336486841e-06, + "loss": 0.6083, + "step": 3533 + }, + { + "epoch": 1.240870786516854, + "grad_norm": 0.8657607436180115, + "learning_rate": 4.5126051113455295e-06, + "loss": 0.6189, + "step": 3534 + }, + { + "epoch": 1.2412219101123596, + "grad_norm": 0.896507203578949, + "learning_rate": 4.512330817374847e-06, + "loss": 0.6512, + "step": 3535 + }, + { + "epoch": 1.2415730337078652, + "grad_norm": 0.8922907710075378, + "learning_rate": 4.512056454584172e-06, + "loss": 0.6269, + "step": 3536 + }, + { + "epoch": 1.2419241573033708, + "grad_norm": 0.9080265760421753, + "learning_rate": 4.511782022982888e-06, + "loss": 0.6319, + "step": 3537 + }, + { + "epoch": 1.2422752808988764, + "grad_norm": 0.8720276355743408, + "learning_rate": 4.511507522580381e-06, + "loss": 0.6038, + "step": 3538 + }, + { + "epoch": 1.242626404494382, + "grad_norm": 0.8793643712997437, + "learning_rate": 4.511232953386037e-06, + "loss": 0.6672, + "step": 3539 + }, + { + "epoch": 1.2429775280898876, + "grad_norm": 0.8374327421188354, + "learning_rate": 4.510958315409247e-06, + "loss": 0.5826, + "step": 3540 + }, + { + "epoch": 1.2433286516853932, + "grad_norm": 0.8608018159866333, + "learning_rate": 4.510683608659404e-06, + "loss": 0.61, + "step": 3541 + }, + { + "epoch": 1.2436797752808988, + "grad_norm": 0.9034033417701721, + "learning_rate": 4.510408833145902e-06, + "loss": 0.6382, + "step": 3542 + }, + { + "epoch": 1.2440308988764044, + "grad_norm": 0.908312201499939, + "learning_rate": 4.510133988878138e-06, + "loss": 0.6437, + "step": 3543 + }, + { + "epoch": 1.24438202247191, + "grad_norm": 0.9202480316162109, + "learning_rate": 4.5098590758655115e-06, + "loss": 0.6462, + "step": 3544 + }, + { + "epoch": 1.2447331460674158, + "grad_norm": 0.8709115386009216, + "learning_rate": 4.509584094117424e-06, + "loss": 0.6146, + "step": 3545 + }, + { + "epoch": 1.2450842696629214, + "grad_norm": 0.8764071464538574, + "learning_rate": 4.509309043643282e-06, + "loss": 0.6002, + "step": 3546 + }, + { + "epoch": 1.245435393258427, + "grad_norm": 0.8512903451919556, + "learning_rate": 4.5090339244524886e-06, + "loss": 0.5995, + "step": 3547 + }, + { + "epoch": 1.2457865168539326, + "grad_norm": 0.846396803855896, + "learning_rate": 4.508758736554454e-06, + "loss": 0.5954, + "step": 3548 + }, + { + "epoch": 1.2461376404494382, + "grad_norm": 0.8854002952575684, + "learning_rate": 4.508483479958591e-06, + "loss": 0.619, + "step": 3549 + }, + { + "epoch": 1.2464887640449438, + "grad_norm": 0.8938324451446533, + "learning_rate": 4.50820815467431e-06, + "loss": 0.6382, + "step": 3550 + }, + { + "epoch": 1.2468398876404494, + "grad_norm": 0.8843767046928406, + "learning_rate": 4.50793276071103e-06, + "loss": 0.6169, + "step": 3551 + }, + { + "epoch": 1.247191011235955, + "grad_norm": 0.8539121747016907, + "learning_rate": 4.507657298078167e-06, + "loss": 0.6312, + "step": 3552 + }, + { + "epoch": 1.2475421348314606, + "grad_norm": 0.8683538436889648, + "learning_rate": 4.5073817667851415e-06, + "loss": 0.6121, + "step": 3553 + }, + { + "epoch": 1.2478932584269664, + "grad_norm": 0.8984917998313904, + "learning_rate": 4.507106166841378e-06, + "loss": 0.6588, + "step": 3554 + }, + { + "epoch": 1.248244382022472, + "grad_norm": 0.9375321865081787, + "learning_rate": 4.506830498256302e-06, + "loss": 0.6943, + "step": 3555 + }, + { + "epoch": 1.2485955056179776, + "grad_norm": 0.8783092498779297, + "learning_rate": 4.5065547610393395e-06, + "loss": 0.6103, + "step": 3556 + }, + { + "epoch": 1.2489466292134832, + "grad_norm": 0.8783857822418213, + "learning_rate": 4.506278955199922e-06, + "loss": 0.6034, + "step": 3557 + }, + { + "epoch": 1.2492977528089888, + "grad_norm": 0.8634523749351501, + "learning_rate": 4.506003080747479e-06, + "loss": 0.5898, + "step": 3558 + }, + { + "epoch": 1.2496488764044944, + "grad_norm": 0.8823550343513489, + "learning_rate": 4.5057271376914495e-06, + "loss": 0.5877, + "step": 3559 + }, + { + "epoch": 1.25, + "grad_norm": 0.8883979320526123, + "learning_rate": 4.505451126041267e-06, + "loss": 0.6384, + "step": 3560 + }, + { + "epoch": 1.2503511235955056, + "grad_norm": 0.8742533922195435, + "learning_rate": 4.5051750458063725e-06, + "loss": 0.6099, + "step": 3561 + }, + { + "epoch": 1.2507022471910112, + "grad_norm": 0.9132082462310791, + "learning_rate": 4.504898896996207e-06, + "loss": 0.6418, + "step": 3562 + }, + { + "epoch": 1.2510533707865168, + "grad_norm": 0.8753030300140381, + "learning_rate": 4.504622679620215e-06, + "loss": 0.6074, + "step": 3563 + }, + { + "epoch": 1.2514044943820224, + "grad_norm": 0.8916899561882019, + "learning_rate": 4.504346393687843e-06, + "loss": 0.6614, + "step": 3564 + }, + { + "epoch": 1.251755617977528, + "grad_norm": 0.8677763342857361, + "learning_rate": 4.504070039208539e-06, + "loss": 0.6256, + "step": 3565 + }, + { + "epoch": 1.2521067415730336, + "grad_norm": 0.8957414627075195, + "learning_rate": 4.503793616191755e-06, + "loss": 0.6742, + "step": 3566 + }, + { + "epoch": 1.2524578651685394, + "grad_norm": 0.8888939619064331, + "learning_rate": 4.5035171246469435e-06, + "loss": 0.6346, + "step": 3567 + }, + { + "epoch": 1.252808988764045, + "grad_norm": 0.8643205165863037, + "learning_rate": 4.5032405645835605e-06, + "loss": 0.6214, + "step": 3568 + }, + { + "epoch": 1.2531601123595506, + "grad_norm": 0.8865513801574707, + "learning_rate": 4.502963936011065e-06, + "loss": 0.6469, + "step": 3569 + }, + { + "epoch": 1.2535112359550562, + "grad_norm": 0.8715183734893799, + "learning_rate": 4.502687238938917e-06, + "loss": 0.6104, + "step": 3570 + }, + { + "epoch": 1.2538623595505618, + "grad_norm": 0.8890028595924377, + "learning_rate": 4.5024104733765795e-06, + "loss": 0.6414, + "step": 3571 + }, + { + "epoch": 1.2542134831460674, + "grad_norm": 0.8931072354316711, + "learning_rate": 4.502133639333516e-06, + "loss": 0.6321, + "step": 3572 + }, + { + "epoch": 1.254564606741573, + "grad_norm": 0.8449422121047974, + "learning_rate": 4.501856736819196e-06, + "loss": 0.6246, + "step": 3573 + }, + { + "epoch": 1.2549157303370786, + "grad_norm": 0.8641847968101501, + "learning_rate": 4.501579765843089e-06, + "loss": 0.5858, + "step": 3574 + }, + { + "epoch": 1.2552668539325842, + "grad_norm": 0.890620768070221, + "learning_rate": 4.501302726414666e-06, + "loss": 0.6045, + "step": 3575 + }, + { + "epoch": 1.25561797752809, + "grad_norm": 0.8621348738670349, + "learning_rate": 4.501025618543403e-06, + "loss": 0.622, + "step": 3576 + }, + { + "epoch": 1.2559691011235956, + "grad_norm": 0.8901969194412231, + "learning_rate": 4.500748442238776e-06, + "loss": 0.6475, + "step": 3577 + }, + { + "epoch": 1.2563202247191012, + "grad_norm": 0.8945297598838806, + "learning_rate": 4.500471197510264e-06, + "loss": 0.6374, + "step": 3578 + }, + { + "epoch": 1.2566713483146068, + "grad_norm": 0.8774127960205078, + "learning_rate": 4.500193884367349e-06, + "loss": 0.6243, + "step": 3579 + }, + { + "epoch": 1.2570224719101124, + "grad_norm": 0.8826004862785339, + "learning_rate": 4.499916502819514e-06, + "loss": 0.6227, + "step": 3580 + }, + { + "epoch": 1.257373595505618, + "grad_norm": 0.9239538311958313, + "learning_rate": 4.499639052876247e-06, + "loss": 0.6432, + "step": 3581 + }, + { + "epoch": 1.2577247191011236, + "grad_norm": 0.8692966103553772, + "learning_rate": 4.499361534547035e-06, + "loss": 0.6343, + "step": 3582 + }, + { + "epoch": 1.2580758426966292, + "grad_norm": 0.837487518787384, + "learning_rate": 4.499083947841371e-06, + "loss": 0.599, + "step": 3583 + }, + { + "epoch": 1.2584269662921348, + "grad_norm": 0.9392856955528259, + "learning_rate": 4.498806292768745e-06, + "loss": 0.5877, + "step": 3584 + }, + { + "epoch": 1.2587780898876404, + "grad_norm": 0.9112626910209656, + "learning_rate": 4.498528569338654e-06, + "loss": 0.6188, + "step": 3585 + }, + { + "epoch": 1.259129213483146, + "grad_norm": 0.863733172416687, + "learning_rate": 4.498250777560596e-06, + "loss": 0.6268, + "step": 3586 + }, + { + "epoch": 1.2594803370786516, + "grad_norm": 0.8512037992477417, + "learning_rate": 4.497972917444072e-06, + "loss": 0.6109, + "step": 3587 + }, + { + "epoch": 1.2598314606741572, + "grad_norm": 0.8677399158477783, + "learning_rate": 4.497694988998584e-06, + "loss": 0.6133, + "step": 3588 + }, + { + "epoch": 1.2601825842696628, + "grad_norm": 0.8969357013702393, + "learning_rate": 4.497416992233637e-06, + "loss": 0.6317, + "step": 3589 + }, + { + "epoch": 1.2605337078651686, + "grad_norm": 0.8791055679321289, + "learning_rate": 4.497138927158737e-06, + "loss": 0.6711, + "step": 3590 + }, + { + "epoch": 1.2608848314606742, + "grad_norm": 0.8862146735191345, + "learning_rate": 4.496860793783397e-06, + "loss": 0.6027, + "step": 3591 + }, + { + "epoch": 1.2612359550561798, + "grad_norm": 0.8788702487945557, + "learning_rate": 4.496582592117123e-06, + "loss": 0.6441, + "step": 3592 + }, + { + "epoch": 1.2615870786516854, + "grad_norm": 0.879822850227356, + "learning_rate": 4.496304322169436e-06, + "loss": 0.605, + "step": 3593 + }, + { + "epoch": 1.261938202247191, + "grad_norm": 0.870986819267273, + "learning_rate": 4.496025983949848e-06, + "loss": 0.6143, + "step": 3594 + }, + { + "epoch": 1.2622893258426966, + "grad_norm": 0.8678484559059143, + "learning_rate": 4.495747577467879e-06, + "loss": 0.6081, + "step": 3595 + }, + { + "epoch": 1.2626404494382022, + "grad_norm": 0.8320144414901733, + "learning_rate": 4.495469102733052e-06, + "loss": 0.6052, + "step": 3596 + }, + { + "epoch": 1.2629915730337078, + "grad_norm": 0.8532273769378662, + "learning_rate": 4.495190559754888e-06, + "loss": 0.6202, + "step": 3597 + }, + { + "epoch": 1.2633426966292136, + "grad_norm": 0.873294472694397, + "learning_rate": 4.494911948542915e-06, + "loss": 0.6292, + "step": 3598 + }, + { + "epoch": 1.2636938202247192, + "grad_norm": 0.8932523131370544, + "learning_rate": 4.4946332691066605e-06, + "loss": 0.6189, + "step": 3599 + }, + { + "epoch": 1.2640449438202248, + "grad_norm": 0.9090549945831299, + "learning_rate": 4.494354521455655e-06, + "loss": 0.66, + "step": 3600 + }, + { + "epoch": 1.2643960674157304, + "grad_norm": 0.8804965615272522, + "learning_rate": 4.49407570559943e-06, + "loss": 0.6105, + "step": 3601 + }, + { + "epoch": 1.264747191011236, + "grad_norm": 0.8950492143630981, + "learning_rate": 4.493796821547523e-06, + "loss": 0.6281, + "step": 3602 + }, + { + "epoch": 1.2650983146067416, + "grad_norm": 0.9223471283912659, + "learning_rate": 4.493517869309472e-06, + "loss": 0.6455, + "step": 3603 + }, + { + "epoch": 1.2654494382022472, + "grad_norm": 0.9145454168319702, + "learning_rate": 4.493238848894815e-06, + "loss": 0.637, + "step": 3604 + }, + { + "epoch": 1.2658005617977528, + "grad_norm": 0.9027042388916016, + "learning_rate": 4.4929597603130944e-06, + "loss": 0.6292, + "step": 3605 + }, + { + "epoch": 1.2661516853932584, + "grad_norm": 0.9141017198562622, + "learning_rate": 4.492680603573857e-06, + "loss": 0.621, + "step": 3606 + }, + { + "epoch": 1.266502808988764, + "grad_norm": 0.903798520565033, + "learning_rate": 4.492401378686646e-06, + "loss": 0.6279, + "step": 3607 + }, + { + "epoch": 1.2668539325842696, + "grad_norm": 0.9175305962562561, + "learning_rate": 4.492122085661014e-06, + "loss": 0.6319, + "step": 3608 + }, + { + "epoch": 1.2672050561797752, + "grad_norm": 0.9265148043632507, + "learning_rate": 4.491842724506511e-06, + "loss": 0.6561, + "step": 3609 + }, + { + "epoch": 1.2675561797752808, + "grad_norm": 0.8984575867652893, + "learning_rate": 4.491563295232692e-06, + "loss": 0.6515, + "step": 3610 + }, + { + "epoch": 1.2679073033707864, + "grad_norm": 0.9043526649475098, + "learning_rate": 4.491283797849112e-06, + "loss": 0.6505, + "step": 3611 + }, + { + "epoch": 1.2682584269662922, + "grad_norm": 0.8922470211982727, + "learning_rate": 4.4910042323653295e-06, + "loss": 0.6638, + "step": 3612 + }, + { + "epoch": 1.2686095505617978, + "grad_norm": 0.8709413409233093, + "learning_rate": 4.4907245987909066e-06, + "loss": 0.6109, + "step": 3613 + }, + { + "epoch": 1.2689606741573034, + "grad_norm": 0.931884229183197, + "learning_rate": 4.490444897135405e-06, + "loss": 0.6791, + "step": 3614 + }, + { + "epoch": 1.269311797752809, + "grad_norm": 0.892665445804596, + "learning_rate": 4.490165127408393e-06, + "loss": 0.6312, + "step": 3615 + }, + { + "epoch": 1.2696629213483146, + "grad_norm": 0.8642548322677612, + "learning_rate": 4.489885289619435e-06, + "loss": 0.6095, + "step": 3616 + }, + { + "epoch": 1.2700140449438202, + "grad_norm": 0.8878061771392822, + "learning_rate": 4.489605383778103e-06, + "loss": 0.6334, + "step": 3617 + }, + { + "epoch": 1.2703651685393258, + "grad_norm": 0.8912963271141052, + "learning_rate": 4.48932540989397e-06, + "loss": 0.6384, + "step": 3618 + }, + { + "epoch": 1.2707162921348314, + "grad_norm": 0.8930538892745972, + "learning_rate": 4.489045367976611e-06, + "loss": 0.6209, + "step": 3619 + }, + { + "epoch": 1.2710674157303372, + "grad_norm": 0.9626955986022949, + "learning_rate": 4.488765258035601e-06, + "loss": 0.6584, + "step": 3620 + }, + { + "epoch": 1.2714185393258428, + "grad_norm": 0.9230532050132751, + "learning_rate": 4.488485080080522e-06, + "loss": 0.6289, + "step": 3621 + }, + { + "epoch": 1.2717696629213484, + "grad_norm": 0.9188664555549622, + "learning_rate": 4.488204834120954e-06, + "loss": 0.6185, + "step": 3622 + }, + { + "epoch": 1.272120786516854, + "grad_norm": 0.9008336663246155, + "learning_rate": 4.4879245201664825e-06, + "loss": 0.6234, + "step": 3623 + }, + { + "epoch": 1.2724719101123596, + "grad_norm": 0.8780505657196045, + "learning_rate": 4.487644138226693e-06, + "loss": 0.6431, + "step": 3624 + }, + { + "epoch": 1.2728230337078652, + "grad_norm": 0.9006023406982422, + "learning_rate": 4.487363688311176e-06, + "loss": 0.6304, + "step": 3625 + }, + { + "epoch": 1.2731741573033708, + "grad_norm": 0.894845724105835, + "learning_rate": 4.487083170429521e-06, + "loss": 0.6455, + "step": 3626 + }, + { + "epoch": 1.2735252808988764, + "grad_norm": 0.9087368249893188, + "learning_rate": 4.486802584591321e-06, + "loss": 0.6616, + "step": 3627 + }, + { + "epoch": 1.273876404494382, + "grad_norm": 0.8746863007545471, + "learning_rate": 4.486521930806174e-06, + "loss": 0.6158, + "step": 3628 + }, + { + "epoch": 1.2742275280898876, + "grad_norm": 0.8722705245018005, + "learning_rate": 4.486241209083676e-06, + "loss": 0.59, + "step": 3629 + }, + { + "epoch": 1.2745786516853932, + "grad_norm": 0.8320393562316895, + "learning_rate": 4.485960419433428e-06, + "loss": 0.6003, + "step": 3630 + }, + { + "epoch": 1.2749297752808988, + "grad_norm": 0.9099650382995605, + "learning_rate": 4.485679561865034e-06, + "loss": 0.6279, + "step": 3631 + }, + { + "epoch": 1.2752808988764044, + "grad_norm": 0.905540406703949, + "learning_rate": 4.485398636388098e-06, + "loss": 0.6963, + "step": 3632 + }, + { + "epoch": 1.27563202247191, + "grad_norm": 0.9020867347717285, + "learning_rate": 4.485117643012227e-06, + "loss": 0.6448, + "step": 3633 + }, + { + "epoch": 1.2759831460674158, + "grad_norm": 0.8654907941818237, + "learning_rate": 4.484836581747032e-06, + "loss": 0.6488, + "step": 3634 + }, + { + "epoch": 1.2763342696629214, + "grad_norm": 0.878855288028717, + "learning_rate": 4.4845554526021235e-06, + "loss": 0.6241, + "step": 3635 + }, + { + "epoch": 1.276685393258427, + "grad_norm": 0.8361636400222778, + "learning_rate": 4.484274255587117e-06, + "loss": 0.6078, + "step": 3636 + }, + { + "epoch": 1.2770365168539326, + "grad_norm": 0.8940629959106445, + "learning_rate": 4.48399299071163e-06, + "loss": 0.6096, + "step": 3637 + }, + { + "epoch": 1.2773876404494382, + "grad_norm": 0.8670988082885742, + "learning_rate": 4.483711657985279e-06, + "loss": 0.5922, + "step": 3638 + }, + { + "epoch": 1.2777387640449438, + "grad_norm": 0.9459436535835266, + "learning_rate": 4.483430257417687e-06, + "loss": 0.648, + "step": 3639 + }, + { + "epoch": 1.2780898876404494, + "grad_norm": 0.889330267906189, + "learning_rate": 4.483148789018478e-06, + "loss": 0.6526, + "step": 3640 + }, + { + "epoch": 1.278441011235955, + "grad_norm": 0.8704626560211182, + "learning_rate": 4.4828672527972775e-06, + "loss": 0.6545, + "step": 3641 + }, + { + "epoch": 1.2787921348314606, + "grad_norm": 0.8923251628875732, + "learning_rate": 4.482585648763713e-06, + "loss": 0.6321, + "step": 3642 + }, + { + "epoch": 1.2791432584269664, + "grad_norm": 0.8882480263710022, + "learning_rate": 4.482303976927417e-06, + "loss": 0.6371, + "step": 3643 + }, + { + "epoch": 1.279494382022472, + "grad_norm": 0.8774885535240173, + "learning_rate": 4.482022237298021e-06, + "loss": 0.5873, + "step": 3644 + }, + { + "epoch": 1.2798455056179776, + "grad_norm": 0.8971747756004333, + "learning_rate": 4.481740429885159e-06, + "loss": 0.6418, + "step": 3645 + }, + { + "epoch": 1.2801966292134832, + "grad_norm": 0.860084593296051, + "learning_rate": 4.481458554698471e-06, + "loss": 0.6259, + "step": 3646 + }, + { + "epoch": 1.2805477528089888, + "grad_norm": 0.885615885257721, + "learning_rate": 4.481176611747597e-06, + "loss": 0.5883, + "step": 3647 + }, + { + "epoch": 1.2808988764044944, + "grad_norm": 0.9061590433120728, + "learning_rate": 4.480894601042177e-06, + "loss": 0.6295, + "step": 3648 + }, + { + "epoch": 1.28125, + "grad_norm": 0.8891565203666687, + "learning_rate": 4.480612522591857e-06, + "loss": 0.6505, + "step": 3649 + }, + { + "epoch": 1.2816011235955056, + "grad_norm": 0.8980679512023926, + "learning_rate": 4.480330376406284e-06, + "loss": 0.6035, + "step": 3650 + }, + { + "epoch": 1.2819522471910112, + "grad_norm": 0.8772475719451904, + "learning_rate": 4.480048162495106e-06, + "loss": 0.6141, + "step": 3651 + }, + { + "epoch": 1.2823033707865168, + "grad_norm": 0.883439302444458, + "learning_rate": 4.479765880867976e-06, + "loss": 0.6349, + "step": 3652 + }, + { + "epoch": 1.2826544943820224, + "grad_norm": 0.872630774974823, + "learning_rate": 4.479483531534546e-06, + "loss": 0.6381, + "step": 3653 + }, + { + "epoch": 1.283005617977528, + "grad_norm": 0.8689930438995361, + "learning_rate": 4.479201114504474e-06, + "loss": 0.6256, + "step": 3654 + }, + { + "epoch": 1.2833567415730336, + "grad_norm": 0.9189596176147461, + "learning_rate": 4.478918629787417e-06, + "loss": 0.6296, + "step": 3655 + }, + { + "epoch": 1.2837078651685394, + "grad_norm": 0.8905652165412903, + "learning_rate": 4.478636077393037e-06, + "loss": 0.6283, + "step": 3656 + }, + { + "epoch": 1.284058988764045, + "grad_norm": 0.889506995677948, + "learning_rate": 4.478353457330995e-06, + "loss": 0.6433, + "step": 3657 + }, + { + "epoch": 1.2844101123595506, + "grad_norm": 0.881960391998291, + "learning_rate": 4.478070769610959e-06, + "loss": 0.6477, + "step": 3658 + }, + { + "epoch": 1.2847612359550562, + "grad_norm": 0.9423824548721313, + "learning_rate": 4.477788014242594e-06, + "loss": 0.6636, + "step": 3659 + }, + { + "epoch": 1.2851123595505618, + "grad_norm": 0.9165781140327454, + "learning_rate": 4.477505191235572e-06, + "loss": 0.6498, + "step": 3660 + }, + { + "epoch": 1.2854634831460674, + "grad_norm": 0.8826702237129211, + "learning_rate": 4.477222300599566e-06, + "loss": 0.6255, + "step": 3661 + }, + { + "epoch": 1.285814606741573, + "grad_norm": 0.8841883540153503, + "learning_rate": 4.476939342344246e-06, + "loss": 0.6496, + "step": 3662 + }, + { + "epoch": 1.2861657303370786, + "grad_norm": 0.8767368197441101, + "learning_rate": 4.476656316479295e-06, + "loss": 0.6227, + "step": 3663 + }, + { + "epoch": 1.2865168539325842, + "grad_norm": 0.8614138960838318, + "learning_rate": 4.4763732230143885e-06, + "loss": 0.6168, + "step": 3664 + }, + { + "epoch": 1.28686797752809, + "grad_norm": 0.8860598802566528, + "learning_rate": 4.476090061959209e-06, + "loss": 0.6414, + "step": 3665 + }, + { + "epoch": 1.2872191011235956, + "grad_norm": 0.8853031396865845, + "learning_rate": 4.4758068333234405e-06, + "loss": 0.6207, + "step": 3666 + }, + { + "epoch": 1.2875702247191012, + "grad_norm": 0.889927089214325, + "learning_rate": 4.4755235371167685e-06, + "loss": 0.6297, + "step": 3667 + }, + { + "epoch": 1.2879213483146068, + "grad_norm": 0.8700869083404541, + "learning_rate": 4.475240173348883e-06, + "loss": 0.6291, + "step": 3668 + }, + { + "epoch": 1.2882724719101124, + "grad_norm": 0.8681374788284302, + "learning_rate": 4.474956742029473e-06, + "loss": 0.6304, + "step": 3669 + }, + { + "epoch": 1.288623595505618, + "grad_norm": 0.8633899688720703, + "learning_rate": 4.474673243168233e-06, + "loss": 0.5838, + "step": 3670 + }, + { + "epoch": 1.2889747191011236, + "grad_norm": 0.8746448755264282, + "learning_rate": 4.4743896767748566e-06, + "loss": 0.6498, + "step": 3671 + }, + { + "epoch": 1.2893258426966292, + "grad_norm": 0.8993739485740662, + "learning_rate": 4.474106042859044e-06, + "loss": 0.6524, + "step": 3672 + }, + { + "epoch": 1.2896769662921348, + "grad_norm": 0.9252314567565918, + "learning_rate": 4.473822341430493e-06, + "loss": 0.668, + "step": 3673 + }, + { + "epoch": 1.2900280898876404, + "grad_norm": 0.8796794414520264, + "learning_rate": 4.473538572498908e-06, + "loss": 0.6452, + "step": 3674 + }, + { + "epoch": 1.290379213483146, + "grad_norm": 0.8701682686805725, + "learning_rate": 4.473254736073992e-06, + "loss": 0.6003, + "step": 3675 + }, + { + "epoch": 1.2907303370786516, + "grad_norm": 0.9012110233306885, + "learning_rate": 4.472970832165453e-06, + "loss": 0.6412, + "step": 3676 + }, + { + "epoch": 1.2910814606741572, + "grad_norm": 0.8365998864173889, + "learning_rate": 4.4726868607829995e-06, + "loss": 0.559, + "step": 3677 + }, + { + "epoch": 1.2914325842696628, + "grad_norm": 0.8749606609344482, + "learning_rate": 4.472402821936344e-06, + "loss": 0.6241, + "step": 3678 + }, + { + "epoch": 1.2917837078651686, + "grad_norm": 0.8832113146781921, + "learning_rate": 4.472118715635199e-06, + "loss": 0.6543, + "step": 3679 + }, + { + "epoch": 1.2921348314606742, + "grad_norm": 0.8898352384567261, + "learning_rate": 4.471834541889282e-06, + "loss": 0.6716, + "step": 3680 + }, + { + "epoch": 1.2924859550561798, + "grad_norm": 0.8850893974304199, + "learning_rate": 4.47155030070831e-06, + "loss": 0.6269, + "step": 3681 + }, + { + "epoch": 1.2928370786516854, + "grad_norm": 0.8936760425567627, + "learning_rate": 4.471265992102005e-06, + "loss": 0.658, + "step": 3682 + }, + { + "epoch": 1.293188202247191, + "grad_norm": 0.8844362497329712, + "learning_rate": 4.4709816160800915e-06, + "loss": 0.6128, + "step": 3683 + }, + { + "epoch": 1.2935393258426966, + "grad_norm": 0.9699074625968933, + "learning_rate": 4.470697172652292e-06, + "loss": 0.653, + "step": 3684 + }, + { + "epoch": 1.2938904494382022, + "grad_norm": 0.9202075600624084, + "learning_rate": 4.470412661828335e-06, + "loss": 0.6514, + "step": 3685 + }, + { + "epoch": 1.2942415730337078, + "grad_norm": 0.8599914312362671, + "learning_rate": 4.470128083617953e-06, + "loss": 0.6347, + "step": 3686 + }, + { + "epoch": 1.2945926966292136, + "grad_norm": 0.8901776075363159, + "learning_rate": 4.469843438030875e-06, + "loss": 0.6073, + "step": 3687 + }, + { + "epoch": 1.2949438202247192, + "grad_norm": 0.8744489550590515, + "learning_rate": 4.4695587250768365e-06, + "loss": 0.5886, + "step": 3688 + }, + { + "epoch": 1.2952949438202248, + "grad_norm": 0.8581476211547852, + "learning_rate": 4.469273944765576e-06, + "loss": 0.6239, + "step": 3689 + }, + { + "epoch": 1.2956460674157304, + "grad_norm": 0.8890603184700012, + "learning_rate": 4.468989097106832e-06, + "loss": 0.6113, + "step": 3690 + }, + { + "epoch": 1.295997191011236, + "grad_norm": 0.9091203808784485, + "learning_rate": 4.468704182110346e-06, + "loss": 0.6044, + "step": 3691 + }, + { + "epoch": 1.2963483146067416, + "grad_norm": 0.8641930818557739, + "learning_rate": 4.46841919978586e-06, + "loss": 0.6264, + "step": 3692 + }, + { + "epoch": 1.2966994382022472, + "grad_norm": 0.8812434673309326, + "learning_rate": 4.468134150143123e-06, + "loss": 0.6027, + "step": 3693 + }, + { + "epoch": 1.2970505617977528, + "grad_norm": 0.9063153266906738, + "learning_rate": 4.467849033191882e-06, + "loss": 0.6429, + "step": 3694 + }, + { + "epoch": 1.2974016853932584, + "grad_norm": 0.8584569096565247, + "learning_rate": 4.467563848941888e-06, + "loss": 0.6003, + "step": 3695 + }, + { + "epoch": 1.297752808988764, + "grad_norm": 0.8659058809280396, + "learning_rate": 4.4672785974028935e-06, + "loss": 0.5802, + "step": 3696 + }, + { + "epoch": 1.2981039325842696, + "grad_norm": 0.8678346872329712, + "learning_rate": 4.4669932785846545e-06, + "loss": 0.6162, + "step": 3697 + }, + { + "epoch": 1.2984550561797752, + "grad_norm": 0.8793441653251648, + "learning_rate": 4.466707892496929e-06, + "loss": 0.6187, + "step": 3698 + }, + { + "epoch": 1.2988061797752808, + "grad_norm": 0.8773770332336426, + "learning_rate": 4.466422439149477e-06, + "loss": 0.6057, + "step": 3699 + }, + { + "epoch": 1.2991573033707864, + "grad_norm": 0.8449274897575378, + "learning_rate": 4.4661369185520594e-06, + "loss": 0.6015, + "step": 3700 + }, + { + "epoch": 1.2995084269662922, + "grad_norm": 0.9012969136238098, + "learning_rate": 4.465851330714441e-06, + "loss": 0.6815, + "step": 3701 + }, + { + "epoch": 1.2998595505617978, + "grad_norm": 0.8823035359382629, + "learning_rate": 4.465565675646391e-06, + "loss": 0.6406, + "step": 3702 + }, + { + "epoch": 1.3002106741573034, + "grad_norm": 0.8793365955352783, + "learning_rate": 4.4652799533576765e-06, + "loss": 0.6172, + "step": 3703 + }, + { + "epoch": 1.300561797752809, + "grad_norm": 0.9065350890159607, + "learning_rate": 4.464994163858069e-06, + "loss": 0.661, + "step": 3704 + }, + { + "epoch": 1.3009129213483146, + "grad_norm": 0.960344672203064, + "learning_rate": 4.464708307157342e-06, + "loss": 0.6326, + "step": 3705 + }, + { + "epoch": 1.3012640449438202, + "grad_norm": 0.8882971405982971, + "learning_rate": 4.4644223832652725e-06, + "loss": 0.6869, + "step": 3706 + }, + { + "epoch": 1.3016151685393258, + "grad_norm": 0.9107559323310852, + "learning_rate": 4.464136392191638e-06, + "loss": 0.669, + "step": 3707 + }, + { + "epoch": 1.3019662921348314, + "grad_norm": 0.93108069896698, + "learning_rate": 4.46385033394622e-06, + "loss": 0.6978, + "step": 3708 + }, + { + "epoch": 1.3023174157303372, + "grad_norm": 0.8690692782402039, + "learning_rate": 4.4635642085388016e-06, + "loss": 0.6164, + "step": 3709 + }, + { + "epoch": 1.3026685393258428, + "grad_norm": 0.882405698299408, + "learning_rate": 4.463278015979166e-06, + "loss": 0.6205, + "step": 3710 + }, + { + "epoch": 1.3030196629213484, + "grad_norm": 0.8657339215278625, + "learning_rate": 4.462991756277104e-06, + "loss": 0.6307, + "step": 3711 + }, + { + "epoch": 1.303370786516854, + "grad_norm": 0.8758717775344849, + "learning_rate": 4.462705429442403e-06, + "loss": 0.6318, + "step": 3712 + }, + { + "epoch": 1.3037219101123596, + "grad_norm": 0.909834623336792, + "learning_rate": 4.4624190354848564e-06, + "loss": 0.6526, + "step": 3713 + }, + { + "epoch": 1.3040730337078652, + "grad_norm": 0.8672031760215759, + "learning_rate": 4.462132574414257e-06, + "loss": 0.6516, + "step": 3714 + }, + { + "epoch": 1.3044241573033708, + "grad_norm": 0.89857017993927, + "learning_rate": 4.461846046240404e-06, + "loss": 0.6645, + "step": 3715 + }, + { + "epoch": 1.3047752808988764, + "grad_norm": 0.8839460015296936, + "learning_rate": 4.461559450973093e-06, + "loss": 0.5743, + "step": 3716 + }, + { + "epoch": 1.305126404494382, + "grad_norm": 0.8894768953323364, + "learning_rate": 4.461272788622129e-06, + "loss": 0.6357, + "step": 3717 + }, + { + "epoch": 1.3054775280898876, + "grad_norm": 0.9299485683441162, + "learning_rate": 4.4609860591973134e-06, + "loss": 0.6118, + "step": 3718 + }, + { + "epoch": 1.3058286516853932, + "grad_norm": 0.8966097831726074, + "learning_rate": 4.460699262708452e-06, + "loss": 0.6128, + "step": 3719 + }, + { + "epoch": 1.3061797752808988, + "grad_norm": 0.9211109280586243, + "learning_rate": 4.460412399165355e-06, + "loss": 0.6307, + "step": 3720 + }, + { + "epoch": 1.3065308988764044, + "grad_norm": 0.8726346492767334, + "learning_rate": 4.46012546857783e-06, + "loss": 0.588, + "step": 3721 + }, + { + "epoch": 1.30688202247191, + "grad_norm": 0.937791645526886, + "learning_rate": 4.459838470955692e-06, + "loss": 0.6367, + "step": 3722 + }, + { + "epoch": 1.3072331460674158, + "grad_norm": 0.9048507213592529, + "learning_rate": 4.459551406308755e-06, + "loss": 0.6337, + "step": 3723 + }, + { + "epoch": 1.3075842696629214, + "grad_norm": 0.8875606060028076, + "learning_rate": 4.459264274646837e-06, + "loss": 0.6341, + "step": 3724 + }, + { + "epoch": 1.307935393258427, + "grad_norm": 0.8814994096755981, + "learning_rate": 4.458977075979758e-06, + "loss": 0.6406, + "step": 3725 + }, + { + "epoch": 1.3082865168539326, + "grad_norm": 0.8935707211494446, + "learning_rate": 4.458689810317338e-06, + "loss": 0.683, + "step": 3726 + }, + { + "epoch": 1.3086376404494382, + "grad_norm": 0.9421975612640381, + "learning_rate": 4.458402477669403e-06, + "loss": 0.6653, + "step": 3727 + }, + { + "epoch": 1.3089887640449438, + "grad_norm": 0.871554434299469, + "learning_rate": 4.45811507804578e-06, + "loss": 0.6291, + "step": 3728 + }, + { + "epoch": 1.3093398876404494, + "grad_norm": 0.9223468899726868, + "learning_rate": 4.457827611456296e-06, + "loss": 0.6447, + "step": 3729 + }, + { + "epoch": 1.309691011235955, + "grad_norm": 0.9151185154914856, + "learning_rate": 4.457540077910784e-06, + "loss": 0.6106, + "step": 3730 + }, + { + "epoch": 1.3100421348314606, + "grad_norm": 0.8625356554985046, + "learning_rate": 4.4572524774190754e-06, + "loss": 0.6024, + "step": 3731 + }, + { + "epoch": 1.3103932584269664, + "grad_norm": 0.8800901174545288, + "learning_rate": 4.456964809991007e-06, + "loss": 0.6199, + "step": 3732 + }, + { + "epoch": 1.310744382022472, + "grad_norm": 0.8656051754951477, + "learning_rate": 4.4566770756364165e-06, + "loss": 0.6249, + "step": 3733 + }, + { + "epoch": 1.3110955056179776, + "grad_norm": 0.8878893256187439, + "learning_rate": 4.456389274365145e-06, + "loss": 0.6181, + "step": 3734 + }, + { + "epoch": 1.3114466292134832, + "grad_norm": 0.8846556544303894, + "learning_rate": 4.456101406187035e-06, + "loss": 0.6395, + "step": 3735 + }, + { + "epoch": 1.3117977528089888, + "grad_norm": 0.9168640375137329, + "learning_rate": 4.4558134711119295e-06, + "loss": 0.6336, + "step": 3736 + }, + { + "epoch": 1.3121488764044944, + "grad_norm": 0.9098040461540222, + "learning_rate": 4.455525469149677e-06, + "loss": 0.6639, + "step": 3737 + }, + { + "epoch": 1.3125, + "grad_norm": 0.8935442566871643, + "learning_rate": 4.455237400310127e-06, + "loss": 0.6221, + "step": 3738 + }, + { + "epoch": 1.3128511235955056, + "grad_norm": 0.8903330564498901, + "learning_rate": 4.454949264603132e-06, + "loss": 0.6484, + "step": 3739 + }, + { + "epoch": 1.3132022471910112, + "grad_norm": 0.8853354454040527, + "learning_rate": 4.454661062038544e-06, + "loss": 0.5822, + "step": 3740 + }, + { + "epoch": 1.3135533707865168, + "grad_norm": 0.9109371304512024, + "learning_rate": 4.45437279262622e-06, + "loss": 0.6844, + "step": 3741 + }, + { + "epoch": 1.3139044943820224, + "grad_norm": 0.8940867781639099, + "learning_rate": 4.45408445637602e-06, + "loss": 0.6693, + "step": 3742 + }, + { + "epoch": 1.314255617977528, + "grad_norm": 0.9142806529998779, + "learning_rate": 4.453796053297803e-06, + "loss": 0.6416, + "step": 3743 + }, + { + "epoch": 1.3146067415730336, + "grad_norm": 0.9041429162025452, + "learning_rate": 4.453507583401433e-06, + "loss": 0.6545, + "step": 3744 + }, + { + "epoch": 1.3149578651685394, + "grad_norm": 0.9050527811050415, + "learning_rate": 4.453219046696775e-06, + "loss": 0.6182, + "step": 3745 + }, + { + "epoch": 1.315308988764045, + "grad_norm": 0.8687748312950134, + "learning_rate": 4.4529304431936976e-06, + "loss": 0.5813, + "step": 3746 + }, + { + "epoch": 1.3156601123595506, + "grad_norm": 0.8562371730804443, + "learning_rate": 4.45264177290207e-06, + "loss": 0.6052, + "step": 3747 + }, + { + "epoch": 1.3160112359550562, + "grad_norm": 0.8941411375999451, + "learning_rate": 4.452353035831764e-06, + "loss": 0.6132, + "step": 3748 + }, + { + "epoch": 1.3163623595505618, + "grad_norm": 0.8909905552864075, + "learning_rate": 4.452064231992655e-06, + "loss": 0.6323, + "step": 3749 + }, + { + "epoch": 1.3167134831460674, + "grad_norm": 0.8690940737724304, + "learning_rate": 4.451775361394621e-06, + "loss": 0.6171, + "step": 3750 + }, + { + "epoch": 1.317064606741573, + "grad_norm": 0.8960789442062378, + "learning_rate": 4.451486424047538e-06, + "loss": 0.6539, + "step": 3751 + }, + { + "epoch": 1.3174157303370786, + "grad_norm": 0.8996719717979431, + "learning_rate": 4.451197419961291e-06, + "loss": 0.6298, + "step": 3752 + }, + { + "epoch": 1.3177668539325842, + "grad_norm": 0.9086701273918152, + "learning_rate": 4.4509083491457605e-06, + "loss": 0.6719, + "step": 3753 + }, + { + "epoch": 1.31811797752809, + "grad_norm": 0.8379474878311157, + "learning_rate": 4.450619211610835e-06, + "loss": 0.6051, + "step": 3754 + }, + { + "epoch": 1.3184691011235956, + "grad_norm": 0.8519045114517212, + "learning_rate": 4.4503300073664e-06, + "loss": 0.6239, + "step": 3755 + }, + { + "epoch": 1.3188202247191012, + "grad_norm": 0.8749707341194153, + "learning_rate": 4.450040736422349e-06, + "loss": 0.614, + "step": 3756 + }, + { + "epoch": 1.3191713483146068, + "grad_norm": 0.9106235504150391, + "learning_rate": 4.4497513987885735e-06, + "loss": 0.6247, + "step": 3757 + }, + { + "epoch": 1.3195224719101124, + "grad_norm": 0.8531911969184875, + "learning_rate": 4.449461994474968e-06, + "loss": 0.5988, + "step": 3758 + }, + { + "epoch": 1.319873595505618, + "grad_norm": 0.8912167549133301, + "learning_rate": 4.449172523491431e-06, + "loss": 0.6383, + "step": 3759 + }, + { + "epoch": 1.3202247191011236, + "grad_norm": 0.8568180203437805, + "learning_rate": 4.448882985847861e-06, + "loss": 0.6366, + "step": 3760 + }, + { + "epoch": 1.3205758426966292, + "grad_norm": 0.9863873720169067, + "learning_rate": 4.448593381554161e-06, + "loss": 0.6223, + "step": 3761 + }, + { + "epoch": 1.3209269662921348, + "grad_norm": 0.8997707366943359, + "learning_rate": 4.448303710620234e-06, + "loss": 0.6224, + "step": 3762 + }, + { + "epoch": 1.3212780898876404, + "grad_norm": 0.8545982837677002, + "learning_rate": 4.448013973055988e-06, + "loss": 0.597, + "step": 3763 + }, + { + "epoch": 1.321629213483146, + "grad_norm": 0.9015237092971802, + "learning_rate": 4.447724168871331e-06, + "loss": 0.6257, + "step": 3764 + }, + { + "epoch": 1.3219803370786516, + "grad_norm": 0.8829646110534668, + "learning_rate": 4.447434298076174e-06, + "loss": 0.6348, + "step": 3765 + }, + { + "epoch": 1.3223314606741572, + "grad_norm": 0.8937842845916748, + "learning_rate": 4.447144360680431e-06, + "loss": 0.6363, + "step": 3766 + }, + { + "epoch": 1.3226825842696628, + "grad_norm": 0.8497073650360107, + "learning_rate": 4.446854356694017e-06, + "loss": 0.5725, + "step": 3767 + }, + { + "epoch": 1.3230337078651686, + "grad_norm": 0.8760591149330139, + "learning_rate": 4.446564286126849e-06, + "loss": 0.6221, + "step": 3768 + }, + { + "epoch": 1.3233848314606742, + "grad_norm": 0.8975864052772522, + "learning_rate": 4.446274148988849e-06, + "loss": 0.6172, + "step": 3769 + }, + { + "epoch": 1.3237359550561798, + "grad_norm": 0.9180589914321899, + "learning_rate": 4.4459839452899385e-06, + "loss": 0.6418, + "step": 3770 + }, + { + "epoch": 1.3240870786516854, + "grad_norm": 0.8982850313186646, + "learning_rate": 4.4456936750400415e-06, + "loss": 0.6271, + "step": 3771 + }, + { + "epoch": 1.324438202247191, + "grad_norm": 0.8551401495933533, + "learning_rate": 4.445403338249087e-06, + "loss": 0.6145, + "step": 3772 + }, + { + "epoch": 1.3247893258426966, + "grad_norm": 0.9100740551948547, + "learning_rate": 4.445112934927002e-06, + "loss": 0.6249, + "step": 3773 + }, + { + "epoch": 1.3251404494382022, + "grad_norm": 0.8740219473838806, + "learning_rate": 4.444822465083719e-06, + "loss": 0.6523, + "step": 3774 + }, + { + "epoch": 1.3254915730337078, + "grad_norm": 0.9058376550674438, + "learning_rate": 4.444531928729173e-06, + "loss": 0.6331, + "step": 3775 + }, + { + "epoch": 1.3258426966292136, + "grad_norm": 0.9305749535560608, + "learning_rate": 4.444241325873298e-06, + "loss": 0.6519, + "step": 3776 + }, + { + "epoch": 1.3261938202247192, + "grad_norm": 0.8894662261009216, + "learning_rate": 4.443950656526034e-06, + "loss": 0.6155, + "step": 3777 + }, + { + "epoch": 1.3265449438202248, + "grad_norm": 0.9070370197296143, + "learning_rate": 4.4436599206973205e-06, + "loss": 0.6019, + "step": 3778 + }, + { + "epoch": 1.3268960674157304, + "grad_norm": 0.8799253106117249, + "learning_rate": 4.4433691183971006e-06, + "loss": 0.6568, + "step": 3779 + }, + { + "epoch": 1.327247191011236, + "grad_norm": 0.9257767200469971, + "learning_rate": 4.4430782496353205e-06, + "loss": 0.6571, + "step": 3780 + }, + { + "epoch": 1.3275983146067416, + "grad_norm": 0.8728551864624023, + "learning_rate": 4.442787314421927e-06, + "loss": 0.6364, + "step": 3781 + }, + { + "epoch": 1.3279494382022472, + "grad_norm": 0.8830699324607849, + "learning_rate": 4.442496312766869e-06, + "loss": 0.629, + "step": 3782 + }, + { + "epoch": 1.3283005617977528, + "grad_norm": 0.9129642248153687, + "learning_rate": 4.4422052446801e-06, + "loss": 0.6377, + "step": 3783 + }, + { + "epoch": 1.3286516853932584, + "grad_norm": 0.8649726510047913, + "learning_rate": 4.441914110171574e-06, + "loss": 0.6565, + "step": 3784 + }, + { + "epoch": 1.329002808988764, + "grad_norm": 0.8841148018836975, + "learning_rate": 4.441622909251246e-06, + "loss": 0.6232, + "step": 3785 + }, + { + "epoch": 1.3293539325842696, + "grad_norm": 0.9116290211677551, + "learning_rate": 4.4413316419290775e-06, + "loss": 0.6522, + "step": 3786 + }, + { + "epoch": 1.3297050561797752, + "grad_norm": 0.8940332531929016, + "learning_rate": 4.441040308215027e-06, + "loss": 0.6147, + "step": 3787 + }, + { + "epoch": 1.3300561797752808, + "grad_norm": 0.8921476602554321, + "learning_rate": 4.4407489081190594e-06, + "loss": 0.6297, + "step": 3788 + }, + { + "epoch": 1.3304073033707864, + "grad_norm": 0.8484114408493042, + "learning_rate": 4.440457441651139e-06, + "loss": 0.5943, + "step": 3789 + }, + { + "epoch": 1.3307584269662922, + "grad_norm": 0.8836881518363953, + "learning_rate": 4.4401659088212356e-06, + "loss": 0.613, + "step": 3790 + }, + { + "epoch": 1.3311095505617978, + "grad_norm": 0.8993842005729675, + "learning_rate": 4.439874309639318e-06, + "loss": 0.653, + "step": 3791 + }, + { + "epoch": 1.3314606741573034, + "grad_norm": 0.896313488483429, + "learning_rate": 4.439582644115359e-06, + "loss": 0.6625, + "step": 3792 + }, + { + "epoch": 1.331811797752809, + "grad_norm": 0.8841575980186462, + "learning_rate": 4.439290912259333e-06, + "loss": 0.6328, + "step": 3793 + }, + { + "epoch": 1.3321629213483146, + "grad_norm": 0.9332913160324097, + "learning_rate": 4.438999114081218e-06, + "loss": 0.6589, + "step": 3794 + }, + { + "epoch": 1.3325140449438202, + "grad_norm": 0.872435450553894, + "learning_rate": 4.438707249590992e-06, + "loss": 0.6348, + "step": 3795 + }, + { + "epoch": 1.3328651685393258, + "grad_norm": 0.8649969100952148, + "learning_rate": 4.4384153187986365e-06, + "loss": 0.6116, + "step": 3796 + }, + { + "epoch": 1.3332162921348314, + "grad_norm": 0.8806115984916687, + "learning_rate": 4.438123321714137e-06, + "loss": 0.618, + "step": 3797 + }, + { + "epoch": 1.3335674157303372, + "grad_norm": 0.905692994594574, + "learning_rate": 4.437831258347477e-06, + "loss": 0.6558, + "step": 3798 + }, + { + "epoch": 1.3339185393258428, + "grad_norm": 0.906639039516449, + "learning_rate": 4.437539128708647e-06, + "loss": 0.6338, + "step": 3799 + }, + { + "epoch": 1.3342696629213484, + "grad_norm": 0.925403356552124, + "learning_rate": 4.4372469328076375e-06, + "loss": 0.6615, + "step": 3800 + }, + { + "epoch": 1.334620786516854, + "grad_norm": 0.8646488189697266, + "learning_rate": 4.436954670654439e-06, + "loss": 0.6128, + "step": 3801 + }, + { + "epoch": 1.3349719101123596, + "grad_norm": 0.8448339700698853, + "learning_rate": 4.436662342259049e-06, + "loss": 0.6352, + "step": 3802 + }, + { + "epoch": 1.3353230337078652, + "grad_norm": 0.8907808661460876, + "learning_rate": 4.436369947631465e-06, + "loss": 0.6108, + "step": 3803 + }, + { + "epoch": 1.3356741573033708, + "grad_norm": 0.8870221972465515, + "learning_rate": 4.436077486781685e-06, + "loss": 0.6391, + "step": 3804 + }, + { + "epoch": 1.3360252808988764, + "grad_norm": 0.8831346035003662, + "learning_rate": 4.435784959719711e-06, + "loss": 0.6947, + "step": 3805 + }, + { + "epoch": 1.336376404494382, + "grad_norm": 0.9102858304977417, + "learning_rate": 4.4354923664555485e-06, + "loss": 0.6248, + "step": 3806 + }, + { + "epoch": 1.3367275280898876, + "grad_norm": 0.8967962265014648, + "learning_rate": 4.435199706999204e-06, + "loss": 0.6066, + "step": 3807 + }, + { + "epoch": 1.3370786516853932, + "grad_norm": 0.8433191180229187, + "learning_rate": 4.434906981360684e-06, + "loss": 0.62, + "step": 3808 + }, + { + "epoch": 1.3374297752808988, + "grad_norm": 0.8778981566429138, + "learning_rate": 4.434614189550003e-06, + "loss": 0.6213, + "step": 3809 + }, + { + "epoch": 1.3377808988764044, + "grad_norm": 0.9490671753883362, + "learning_rate": 4.4343213315771695e-06, + "loss": 0.6183, + "step": 3810 + }, + { + "epoch": 1.33813202247191, + "grad_norm": 0.8874966502189636, + "learning_rate": 4.434028407452203e-06, + "loss": 0.6473, + "step": 3811 + }, + { + "epoch": 1.3384831460674158, + "grad_norm": 0.8727471828460693, + "learning_rate": 4.433735417185121e-06, + "loss": 0.6234, + "step": 3812 + }, + { + "epoch": 1.3388342696629214, + "grad_norm": 0.9094645977020264, + "learning_rate": 4.433442360785941e-06, + "loss": 0.6511, + "step": 3813 + }, + { + "epoch": 1.339185393258427, + "grad_norm": 0.934386670589447, + "learning_rate": 4.433149238264688e-06, + "loss": 0.6563, + "step": 3814 + }, + { + "epoch": 1.3395365168539326, + "grad_norm": 0.8715890049934387, + "learning_rate": 4.432856049631385e-06, + "loss": 0.6598, + "step": 3815 + }, + { + "epoch": 1.3398876404494382, + "grad_norm": 0.8467699289321899, + "learning_rate": 4.432562794896058e-06, + "loss": 0.6224, + "step": 3816 + }, + { + "epoch": 1.3402387640449438, + "grad_norm": 0.8690305948257446, + "learning_rate": 4.4322694740687375e-06, + "loss": 0.6075, + "step": 3817 + }, + { + "epoch": 1.3405898876404494, + "grad_norm": 0.8769239783287048, + "learning_rate": 4.431976087159455e-06, + "loss": 0.6145, + "step": 3818 + }, + { + "epoch": 1.340941011235955, + "grad_norm": 0.8641560673713684, + "learning_rate": 4.431682634178243e-06, + "loss": 0.629, + "step": 3819 + }, + { + "epoch": 1.3412921348314606, + "grad_norm": 0.8964089155197144, + "learning_rate": 4.431389115135138e-06, + "loss": 0.6456, + "step": 3820 + }, + { + "epoch": 1.3416432584269664, + "grad_norm": 0.8796816468238831, + "learning_rate": 4.4310955300401785e-06, + "loss": 0.6511, + "step": 3821 + }, + { + "epoch": 1.341994382022472, + "grad_norm": 0.8870885372161865, + "learning_rate": 4.430801878903403e-06, + "loss": 0.6458, + "step": 3822 + }, + { + "epoch": 1.3423455056179776, + "grad_norm": 0.8509821891784668, + "learning_rate": 4.430508161734857e-06, + "loss": 0.6291, + "step": 3823 + }, + { + "epoch": 1.3426966292134832, + "grad_norm": 0.8842064738273621, + "learning_rate": 4.430214378544583e-06, + "loss": 0.6223, + "step": 3824 + }, + { + "epoch": 1.3430477528089888, + "grad_norm": 0.882311224937439, + "learning_rate": 4.429920529342629e-06, + "loss": 0.5903, + "step": 3825 + }, + { + "epoch": 1.3433988764044944, + "grad_norm": 0.9211931228637695, + "learning_rate": 4.429626614139044e-06, + "loss": 0.6096, + "step": 3826 + }, + { + "epoch": 1.34375, + "grad_norm": 0.9198620319366455, + "learning_rate": 4.429332632943881e-06, + "loss": 0.6021, + "step": 3827 + }, + { + "epoch": 1.3441011235955056, + "grad_norm": 0.8927902579307556, + "learning_rate": 4.4290385857671924e-06, + "loss": 0.6681, + "step": 3828 + }, + { + "epoch": 1.3444522471910112, + "grad_norm": 0.8720464706420898, + "learning_rate": 4.4287444726190346e-06, + "loss": 0.643, + "step": 3829 + }, + { + "epoch": 1.3448033707865168, + "grad_norm": 0.9204714298248291, + "learning_rate": 4.428450293509467e-06, + "loss": 0.6289, + "step": 3830 + }, + { + "epoch": 1.3451544943820224, + "grad_norm": 0.8725011348724365, + "learning_rate": 4.428156048448549e-06, + "loss": 0.6461, + "step": 3831 + }, + { + "epoch": 1.345505617977528, + "grad_norm": 0.9051564335823059, + "learning_rate": 4.427861737446344e-06, + "loss": 0.6235, + "step": 3832 + }, + { + "epoch": 1.3458567415730336, + "grad_norm": 0.9251704216003418, + "learning_rate": 4.427567360512917e-06, + "loss": 0.648, + "step": 3833 + }, + { + "epoch": 1.3462078651685394, + "grad_norm": 0.8915055990219116, + "learning_rate": 4.427272917658337e-06, + "loss": 0.6474, + "step": 3834 + }, + { + "epoch": 1.346558988764045, + "grad_norm": 0.8907241225242615, + "learning_rate": 4.426978408892672e-06, + "loss": 0.6431, + "step": 3835 + }, + { + "epoch": 1.3469101123595506, + "grad_norm": 0.8840727806091309, + "learning_rate": 4.426683834225993e-06, + "loss": 0.6302, + "step": 3836 + }, + { + "epoch": 1.3472612359550562, + "grad_norm": 0.8734930157661438, + "learning_rate": 4.4263891936683765e-06, + "loss": 0.5856, + "step": 3837 + }, + { + "epoch": 1.3476123595505618, + "grad_norm": 0.9012126326560974, + "learning_rate": 4.426094487229898e-06, + "loss": 0.6484, + "step": 3838 + }, + { + "epoch": 1.3479634831460674, + "grad_norm": 0.8863605856895447, + "learning_rate": 4.425799714920636e-06, + "loss": 0.6754, + "step": 3839 + }, + { + "epoch": 1.348314606741573, + "grad_norm": 0.8811327815055847, + "learning_rate": 4.4255048767506725e-06, + "loss": 0.6565, + "step": 3840 + }, + { + "epoch": 1.3486657303370786, + "grad_norm": 0.9418013691902161, + "learning_rate": 4.425209972730089e-06, + "loss": 0.6495, + "step": 3841 + }, + { + "epoch": 1.3490168539325842, + "grad_norm": 0.9380903244018555, + "learning_rate": 4.4249150028689726e-06, + "loss": 0.6224, + "step": 3842 + }, + { + "epoch": 1.34936797752809, + "grad_norm": 0.8815275430679321, + "learning_rate": 4.4246199671774094e-06, + "loss": 0.6617, + "step": 3843 + }, + { + "epoch": 1.3497191011235956, + "grad_norm": 0.8603273630142212, + "learning_rate": 4.424324865665491e-06, + "loss": 0.6247, + "step": 3844 + }, + { + "epoch": 1.3500702247191012, + "grad_norm": 0.904721736907959, + "learning_rate": 4.424029698343309e-06, + "loss": 0.6487, + "step": 3845 + }, + { + "epoch": 1.3504213483146068, + "grad_norm": 0.873080313205719, + "learning_rate": 4.423734465220957e-06, + "loss": 0.6169, + "step": 3846 + }, + { + "epoch": 1.3507724719101124, + "grad_norm": 0.8521433472633362, + "learning_rate": 4.4234391663085324e-06, + "loss": 0.6537, + "step": 3847 + }, + { + "epoch": 1.351123595505618, + "grad_norm": 0.9350584745407104, + "learning_rate": 4.423143801616134e-06, + "loss": 0.6279, + "step": 3848 + }, + { + "epoch": 1.3514747191011236, + "grad_norm": 0.8658765554428101, + "learning_rate": 4.422848371153865e-06, + "loss": 0.6363, + "step": 3849 + }, + { + "epoch": 1.3518258426966292, + "grad_norm": 0.9064232110977173, + "learning_rate": 4.422552874931825e-06, + "loss": 0.6337, + "step": 3850 + }, + { + "epoch": 1.3521769662921348, + "grad_norm": 0.8607880473136902, + "learning_rate": 4.422257312960123e-06, + "loss": 0.6058, + "step": 3851 + }, + { + "epoch": 1.3525280898876404, + "grad_norm": 0.8825393319129944, + "learning_rate": 4.421961685248866e-06, + "loss": 0.6546, + "step": 3852 + }, + { + "epoch": 1.352879213483146, + "grad_norm": 0.8812253475189209, + "learning_rate": 4.421665991808163e-06, + "loss": 0.607, + "step": 3853 + }, + { + "epoch": 1.3532303370786516, + "grad_norm": 0.8943750262260437, + "learning_rate": 4.421370232648128e-06, + "loss": 0.6298, + "step": 3854 + }, + { + "epoch": 1.3535814606741572, + "grad_norm": 0.8581367135047913, + "learning_rate": 4.4210744077788755e-06, + "loss": 0.6057, + "step": 3855 + }, + { + "epoch": 1.3539325842696628, + "grad_norm": 0.8943094611167908, + "learning_rate": 4.4207785172105216e-06, + "loss": 0.6333, + "step": 3856 + }, + { + "epoch": 1.3542837078651686, + "grad_norm": 0.8830317854881287, + "learning_rate": 4.420482560953186e-06, + "loss": 0.651, + "step": 3857 + }, + { + "epoch": 1.3546348314606742, + "grad_norm": 0.8857917189598083, + "learning_rate": 4.420186539016991e-06, + "loss": 0.6463, + "step": 3858 + }, + { + "epoch": 1.3549859550561798, + "grad_norm": 0.8663979768753052, + "learning_rate": 4.419890451412059e-06, + "loss": 0.5975, + "step": 3859 + }, + { + "epoch": 1.3553370786516854, + "grad_norm": 0.8694441318511963, + "learning_rate": 4.4195942981485165e-06, + "loss": 0.6093, + "step": 3860 + }, + { + "epoch": 1.355688202247191, + "grad_norm": 0.8855713605880737, + "learning_rate": 4.419298079236491e-06, + "loss": 0.625, + "step": 3861 + }, + { + "epoch": 1.3560393258426966, + "grad_norm": 0.9174140095710754, + "learning_rate": 4.419001794686114e-06, + "loss": 0.6448, + "step": 3862 + }, + { + "epoch": 1.3563904494382022, + "grad_norm": 0.8802694082260132, + "learning_rate": 4.418705444507518e-06, + "loss": 0.681, + "step": 3863 + }, + { + "epoch": 1.3567415730337078, + "grad_norm": 0.8532006740570068, + "learning_rate": 4.418409028710838e-06, + "loss": 0.6378, + "step": 3864 + }, + { + "epoch": 1.3570926966292136, + "grad_norm": 0.8874769806861877, + "learning_rate": 4.41811254730621e-06, + "loss": 0.6397, + "step": 3865 + }, + { + "epoch": 1.3574438202247192, + "grad_norm": 0.9477583765983582, + "learning_rate": 4.417816000303775e-06, + "loss": 0.62, + "step": 3866 + }, + { + "epoch": 1.3577949438202248, + "grad_norm": 0.8806896805763245, + "learning_rate": 4.417519387713673e-06, + "loss": 0.6324, + "step": 3867 + }, + { + "epoch": 1.3581460674157304, + "grad_norm": 0.8547712564468384, + "learning_rate": 4.41722270954605e-06, + "loss": 0.6066, + "step": 3868 + }, + { + "epoch": 1.358497191011236, + "grad_norm": 0.895625114440918, + "learning_rate": 4.416925965811049e-06, + "loss": 0.6373, + "step": 3869 + }, + { + "epoch": 1.3588483146067416, + "grad_norm": 0.895209550857544, + "learning_rate": 4.416629156518822e-06, + "loss": 0.6408, + "step": 3870 + }, + { + "epoch": 1.3591994382022472, + "grad_norm": 0.8948414325714111, + "learning_rate": 4.416332281679517e-06, + "loss": 0.6238, + "step": 3871 + }, + { + "epoch": 1.3595505617977528, + "grad_norm": 0.8785004019737244, + "learning_rate": 4.4160353413032885e-06, + "loss": 0.6141, + "step": 3872 + }, + { + "epoch": 1.3599016853932584, + "grad_norm": 0.9018064737319946, + "learning_rate": 4.415738335400291e-06, + "loss": 0.6668, + "step": 3873 + }, + { + "epoch": 1.360252808988764, + "grad_norm": 0.8808990120887756, + "learning_rate": 4.415441263980681e-06, + "loss": 0.634, + "step": 3874 + }, + { + "epoch": 1.3606039325842696, + "grad_norm": 0.8734375834465027, + "learning_rate": 4.415144127054619e-06, + "loss": 0.65, + "step": 3875 + }, + { + "epoch": 1.3609550561797752, + "grad_norm": 0.8722198605537415, + "learning_rate": 4.414846924632267e-06, + "loss": 0.6227, + "step": 3876 + }, + { + "epoch": 1.3613061797752808, + "grad_norm": 0.8828457593917847, + "learning_rate": 4.414549656723789e-06, + "loss": 0.5958, + "step": 3877 + }, + { + "epoch": 1.3616573033707864, + "grad_norm": 0.8679347634315491, + "learning_rate": 4.41425232333935e-06, + "loss": 0.6228, + "step": 3878 + }, + { + "epoch": 1.3620084269662922, + "grad_norm": 0.8740646839141846, + "learning_rate": 4.413954924489121e-06, + "loss": 0.5995, + "step": 3879 + }, + { + "epoch": 1.3623595505617978, + "grad_norm": 0.867852509021759, + "learning_rate": 4.41365746018327e-06, + "loss": 0.6606, + "step": 3880 + }, + { + "epoch": 1.3627106741573034, + "grad_norm": 0.917854905128479, + "learning_rate": 4.413359930431972e-06, + "loss": 0.6485, + "step": 3881 + }, + { + "epoch": 1.363061797752809, + "grad_norm": 0.8932327628135681, + "learning_rate": 4.413062335245402e-06, + "loss": 0.6356, + "step": 3882 + }, + { + "epoch": 1.3634129213483146, + "grad_norm": 0.8385898470878601, + "learning_rate": 4.412764674633736e-06, + "loss": 0.6072, + "step": 3883 + }, + { + "epoch": 1.3637640449438202, + "grad_norm": 0.9037755727767944, + "learning_rate": 4.412466948607156e-06, + "loss": 0.6339, + "step": 3884 + }, + { + "epoch": 1.3641151685393258, + "grad_norm": 0.882864236831665, + "learning_rate": 4.412169157175843e-06, + "loss": 0.6571, + "step": 3885 + }, + { + "epoch": 1.3644662921348314, + "grad_norm": 0.8602117896080017, + "learning_rate": 4.41187130034998e-06, + "loss": 0.6173, + "step": 3886 + }, + { + "epoch": 1.3648174157303372, + "grad_norm": 0.8562754392623901, + "learning_rate": 4.411573378139755e-06, + "loss": 0.5818, + "step": 3887 + }, + { + "epoch": 1.3651685393258428, + "grad_norm": 0.8554957509040833, + "learning_rate": 4.4112753905553575e-06, + "loss": 0.6, + "step": 3888 + }, + { + "epoch": 1.3655196629213484, + "grad_norm": 0.9020095467567444, + "learning_rate": 4.410977337606975e-06, + "loss": 0.6538, + "step": 3889 + }, + { + "epoch": 1.365870786516854, + "grad_norm": 0.9186870455741882, + "learning_rate": 4.410679219304804e-06, + "loss": 0.6393, + "step": 3890 + }, + { + "epoch": 1.3662219101123596, + "grad_norm": 0.8861616849899292, + "learning_rate": 4.410381035659039e-06, + "loss": 0.6201, + "step": 3891 + }, + { + "epoch": 1.3665730337078652, + "grad_norm": 0.8816412687301636, + "learning_rate": 4.4100827866798765e-06, + "loss": 0.6228, + "step": 3892 + }, + { + "epoch": 1.3669241573033708, + "grad_norm": 0.9039531350135803, + "learning_rate": 4.4097844723775185e-06, + "loss": 0.6368, + "step": 3893 + }, + { + "epoch": 1.3672752808988764, + "grad_norm": 0.882317066192627, + "learning_rate": 4.409486092762164e-06, + "loss": 0.6586, + "step": 3894 + }, + { + "epoch": 1.367626404494382, + "grad_norm": 0.8748815655708313, + "learning_rate": 4.409187647844021e-06, + "loss": 0.6285, + "step": 3895 + }, + { + "epoch": 1.3679775280898876, + "grad_norm": 0.8901538848876953, + "learning_rate": 4.408889137633292e-06, + "loss": 0.608, + "step": 3896 + }, + { + "epoch": 1.3683286516853932, + "grad_norm": 0.8996601104736328, + "learning_rate": 4.40859056214019e-06, + "loss": 0.6687, + "step": 3897 + }, + { + "epoch": 1.3686797752808988, + "grad_norm": 0.8397285342216492, + "learning_rate": 4.408291921374923e-06, + "loss": 0.5793, + "step": 3898 + }, + { + "epoch": 1.3690308988764044, + "grad_norm": 0.8991907835006714, + "learning_rate": 4.407993215347705e-06, + "loss": 0.6129, + "step": 3899 + }, + { + "epoch": 1.36938202247191, + "grad_norm": 0.8629971742630005, + "learning_rate": 4.407694444068752e-06, + "loss": 0.589, + "step": 3900 + }, + { + "epoch": 1.3697331460674158, + "grad_norm": 0.9028024077415466, + "learning_rate": 4.407395607548281e-06, + "loss": 0.6262, + "step": 3901 + }, + { + "epoch": 1.3700842696629214, + "grad_norm": 0.9251904487609863, + "learning_rate": 4.407096705796513e-06, + "loss": 0.645, + "step": 3902 + }, + { + "epoch": 1.370435393258427, + "grad_norm": 0.8927795886993408, + "learning_rate": 4.40679773882367e-06, + "loss": 0.6317, + "step": 3903 + }, + { + "epoch": 1.3707865168539326, + "grad_norm": 0.8791918158531189, + "learning_rate": 4.406498706639976e-06, + "loss": 0.6172, + "step": 3904 + }, + { + "epoch": 1.3711376404494382, + "grad_norm": 0.8712174296379089, + "learning_rate": 4.406199609255657e-06, + "loss": 0.6229, + "step": 3905 + }, + { + "epoch": 1.3714887640449438, + "grad_norm": 0.8883311748504639, + "learning_rate": 4.405900446680942e-06, + "loss": 0.632, + "step": 3906 + }, + { + "epoch": 1.3718398876404494, + "grad_norm": 0.8838800191879272, + "learning_rate": 4.4056012189260634e-06, + "loss": 0.6314, + "step": 3907 + }, + { + "epoch": 1.372191011235955, + "grad_norm": 0.8562436699867249, + "learning_rate": 4.405301926001253e-06, + "loss": 0.6476, + "step": 3908 + }, + { + "epoch": 1.3725421348314606, + "grad_norm": 0.8729368448257446, + "learning_rate": 4.405002567916749e-06, + "loss": 0.631, + "step": 3909 + }, + { + "epoch": 1.3728932584269664, + "grad_norm": 0.8762539625167847, + "learning_rate": 4.4047031446827855e-06, + "loss": 0.6258, + "step": 3910 + }, + { + "epoch": 1.373244382022472, + "grad_norm": 0.8716838955879211, + "learning_rate": 4.404403656309605e-06, + "loss": 0.6441, + "step": 3911 + }, + { + "epoch": 1.3735955056179776, + "grad_norm": 0.8772343993186951, + "learning_rate": 4.404104102807448e-06, + "loss": 0.622, + "step": 3912 + }, + { + "epoch": 1.3739466292134832, + "grad_norm": 0.8888761401176453, + "learning_rate": 4.403804484186561e-06, + "loss": 0.6393, + "step": 3913 + }, + { + "epoch": 1.3742977528089888, + "grad_norm": 0.891753613948822, + "learning_rate": 4.40350480045719e-06, + "loss": 0.6305, + "step": 3914 + }, + { + "epoch": 1.3746488764044944, + "grad_norm": 0.8871645927429199, + "learning_rate": 4.403205051629584e-06, + "loss": 0.6426, + "step": 3915 + }, + { + "epoch": 1.375, + "grad_norm": 0.8718512654304504, + "learning_rate": 4.402905237713993e-06, + "loss": 0.6553, + "step": 3916 + }, + { + "epoch": 1.3753511235955056, + "grad_norm": 0.8967269062995911, + "learning_rate": 4.402605358720672e-06, + "loss": 0.654, + "step": 3917 + }, + { + "epoch": 1.3757022471910112, + "grad_norm": 0.8855593204498291, + "learning_rate": 4.402305414659875e-06, + "loss": 0.6592, + "step": 3918 + }, + { + "epoch": 1.3760533707865168, + "grad_norm": 0.859167754650116, + "learning_rate": 4.40200540554186e-06, + "loss": 0.6189, + "step": 3919 + }, + { + "epoch": 1.3764044943820224, + "grad_norm": 0.9102888703346252, + "learning_rate": 4.4017053313768895e-06, + "loss": 0.655, + "step": 3920 + }, + { + "epoch": 1.376755617977528, + "grad_norm": 0.8694773316383362, + "learning_rate": 4.401405192175223e-06, + "loss": 0.6315, + "step": 3921 + }, + { + "epoch": 1.3771067415730336, + "grad_norm": 0.8884313106536865, + "learning_rate": 4.401104987947126e-06, + "loss": 0.6305, + "step": 3922 + }, + { + "epoch": 1.3774578651685394, + "grad_norm": 0.8785154819488525, + "learning_rate": 4.400804718702865e-06, + "loss": 0.6349, + "step": 3923 + }, + { + "epoch": 1.377808988764045, + "grad_norm": 0.8957915306091309, + "learning_rate": 4.40050438445271e-06, + "loss": 0.6323, + "step": 3924 + }, + { + "epoch": 1.3781601123595506, + "grad_norm": 0.9052011966705322, + "learning_rate": 4.40020398520693e-06, + "loss": 0.6171, + "step": 3925 + }, + { + "epoch": 1.3785112359550562, + "grad_norm": 0.8835782408714294, + "learning_rate": 4.3999035209758015e-06, + "loss": 0.6299, + "step": 3926 + }, + { + "epoch": 1.3788623595505618, + "grad_norm": 0.911495566368103, + "learning_rate": 4.399602991769597e-06, + "loss": 0.6131, + "step": 3927 + }, + { + "epoch": 1.3792134831460674, + "grad_norm": 0.8838707208633423, + "learning_rate": 4.399302397598596e-06, + "loss": 0.6246, + "step": 3928 + }, + { + "epoch": 1.379564606741573, + "grad_norm": 0.9067579507827759, + "learning_rate": 4.399001738473079e-06, + "loss": 0.6554, + "step": 3929 + }, + { + "epoch": 1.3799157303370786, + "grad_norm": 0.9124329686164856, + "learning_rate": 4.398701014403326e-06, + "loss": 0.6115, + "step": 3930 + }, + { + "epoch": 1.3802668539325842, + "grad_norm": 0.9139940142631531, + "learning_rate": 4.398400225399624e-06, + "loss": 0.6373, + "step": 3931 + }, + { + "epoch": 1.38061797752809, + "grad_norm": 0.8919884562492371, + "learning_rate": 4.398099371472259e-06, + "loss": 0.6771, + "step": 3932 + }, + { + "epoch": 1.3809691011235956, + "grad_norm": 3.221696615219116, + "learning_rate": 4.397798452631519e-06, + "loss": 0.6222, + "step": 3933 + }, + { + "epoch": 1.3813202247191012, + "grad_norm": 0.8897351026535034, + "learning_rate": 4.397497468887697e-06, + "loss": 0.6688, + "step": 3934 + }, + { + "epoch": 1.3816713483146068, + "grad_norm": 0.8792717456817627, + "learning_rate": 4.3971964202510844e-06, + "loss": 0.6576, + "step": 3935 + }, + { + "epoch": 1.3820224719101124, + "grad_norm": 0.8989633321762085, + "learning_rate": 4.396895306731978e-06, + "loss": 0.6293, + "step": 3936 + }, + { + "epoch": 1.382373595505618, + "grad_norm": 0.8742619156837463, + "learning_rate": 4.396594128340675e-06, + "loss": 0.6013, + "step": 3937 + }, + { + "epoch": 1.3827247191011236, + "grad_norm": 0.8698790669441223, + "learning_rate": 4.396292885087476e-06, + "loss": 0.6172, + "step": 3938 + }, + { + "epoch": 1.3830758426966292, + "grad_norm": 0.8680218458175659, + "learning_rate": 4.395991576982683e-06, + "loss": 0.6318, + "step": 3939 + }, + { + "epoch": 1.3834269662921348, + "grad_norm": 0.9022037386894226, + "learning_rate": 4.3956902040366e-06, + "loss": 0.6423, + "step": 3940 + }, + { + "epoch": 1.3837780898876404, + "grad_norm": 0.8297759890556335, + "learning_rate": 4.3953887662595356e-06, + "loss": 0.5716, + "step": 3941 + }, + { + "epoch": 1.384129213483146, + "grad_norm": 0.9098703265190125, + "learning_rate": 4.395087263661796e-06, + "loss": 0.6599, + "step": 3942 + }, + { + "epoch": 1.3844803370786516, + "grad_norm": 0.8878948092460632, + "learning_rate": 4.394785696253694e-06, + "loss": 0.613, + "step": 3943 + }, + { + "epoch": 1.3848314606741572, + "grad_norm": 0.8692125678062439, + "learning_rate": 4.394484064045542e-06, + "loss": 0.6057, + "step": 3944 + }, + { + "epoch": 1.3851825842696628, + "grad_norm": 0.9190526008605957, + "learning_rate": 4.394182367047657e-06, + "loss": 0.6382, + "step": 3945 + }, + { + "epoch": 1.3855337078651686, + "grad_norm": 0.8912534713745117, + "learning_rate": 4.393880605270355e-06, + "loss": 0.6535, + "step": 3946 + }, + { + "epoch": 1.3858848314606742, + "grad_norm": 0.8843017220497131, + "learning_rate": 4.393578778723958e-06, + "loss": 0.6332, + "step": 3947 + }, + { + "epoch": 1.3862359550561798, + "grad_norm": 0.8994221687316895, + "learning_rate": 4.393276887418786e-06, + "loss": 0.6337, + "step": 3948 + }, + { + "epoch": 1.3865870786516854, + "grad_norm": 0.8987597823143005, + "learning_rate": 4.392974931365166e-06, + "loss": 0.6465, + "step": 3949 + }, + { + "epoch": 1.386938202247191, + "grad_norm": 0.8565405011177063, + "learning_rate": 4.392672910573422e-06, + "loss": 0.63, + "step": 3950 + }, + { + "epoch": 1.3872893258426966, + "grad_norm": 0.8897543549537659, + "learning_rate": 4.392370825053884e-06, + "loss": 0.6248, + "step": 3951 + }, + { + "epoch": 1.3876404494382022, + "grad_norm": 0.8679741621017456, + "learning_rate": 4.392068674816884e-06, + "loss": 0.6087, + "step": 3952 + }, + { + "epoch": 1.3879915730337078, + "grad_norm": 0.8487875461578369, + "learning_rate": 4.3917664598727535e-06, + "loss": 0.6154, + "step": 3953 + }, + { + "epoch": 1.3883426966292136, + "grad_norm": 0.9262088537216187, + "learning_rate": 4.39146418023183e-06, + "loss": 0.6278, + "step": 3954 + }, + { + "epoch": 1.3886938202247192, + "grad_norm": 0.9284934997558594, + "learning_rate": 4.391161835904449e-06, + "loss": 0.6405, + "step": 3955 + }, + { + "epoch": 1.3890449438202248, + "grad_norm": 0.8615259528160095, + "learning_rate": 4.390859426900952e-06, + "loss": 0.6117, + "step": 3956 + }, + { + "epoch": 1.3893960674157304, + "grad_norm": 0.9133617877960205, + "learning_rate": 4.39055695323168e-06, + "loss": 0.6036, + "step": 3957 + }, + { + "epoch": 1.389747191011236, + "grad_norm": 0.9007285237312317, + "learning_rate": 4.390254414906979e-06, + "loss": 0.6211, + "step": 3958 + }, + { + "epoch": 1.3900983146067416, + "grad_norm": 0.9187461733818054, + "learning_rate": 4.389951811937195e-06, + "loss": 0.6685, + "step": 3959 + }, + { + "epoch": 1.3904494382022472, + "grad_norm": 0.8911426663398743, + "learning_rate": 4.3896491443326746e-06, + "loss": 0.6602, + "step": 3960 + }, + { + "epoch": 1.3908005617977528, + "grad_norm": 0.9059688448905945, + "learning_rate": 4.389346412103771e-06, + "loss": 0.5913, + "step": 3961 + }, + { + "epoch": 1.3911516853932584, + "grad_norm": 0.871945858001709, + "learning_rate": 4.389043615260837e-06, + "loss": 0.6377, + "step": 3962 + }, + { + "epoch": 1.391502808988764, + "grad_norm": 0.8781225085258484, + "learning_rate": 4.388740753814228e-06, + "loss": 0.6198, + "step": 3963 + }, + { + "epoch": 1.3918539325842696, + "grad_norm": 0.8738672733306885, + "learning_rate": 4.388437827774301e-06, + "loss": 0.6079, + "step": 3964 + }, + { + "epoch": 1.3922050561797752, + "grad_norm": 0.88068026304245, + "learning_rate": 4.388134837151417e-06, + "loss": 0.628, + "step": 3965 + }, + { + "epoch": 1.3925561797752808, + "grad_norm": 0.884495735168457, + "learning_rate": 4.3878317819559365e-06, + "loss": 0.6055, + "step": 3966 + }, + { + "epoch": 1.3929073033707864, + "grad_norm": 0.8998886346817017, + "learning_rate": 4.387528662198225e-06, + "loss": 0.6411, + "step": 3967 + }, + { + "epoch": 1.3932584269662922, + "grad_norm": 0.8848565816879272, + "learning_rate": 4.387225477888648e-06, + "loss": 0.6399, + "step": 3968 + }, + { + "epoch": 1.3936095505617978, + "grad_norm": 0.9099523425102234, + "learning_rate": 4.386922229037575e-06, + "loss": 0.6303, + "step": 3969 + }, + { + "epoch": 1.3939606741573034, + "grad_norm": 0.8989171385765076, + "learning_rate": 4.386618915655375e-06, + "loss": 0.6432, + "step": 3970 + }, + { + "epoch": 1.394311797752809, + "grad_norm": 0.8913509845733643, + "learning_rate": 4.386315537752425e-06, + "loss": 0.6323, + "step": 3971 + }, + { + "epoch": 1.3946629213483146, + "grad_norm": 0.8825880885124207, + "learning_rate": 4.3860120953390965e-06, + "loss": 0.6687, + "step": 3972 + }, + { + "epoch": 1.3950140449438202, + "grad_norm": 0.8669257760047913, + "learning_rate": 4.3857085884257675e-06, + "loss": 0.6326, + "step": 3973 + }, + { + "epoch": 1.3953651685393258, + "grad_norm": 0.8898574113845825, + "learning_rate": 4.385405017022819e-06, + "loss": 0.6391, + "step": 3974 + }, + { + "epoch": 1.3957162921348314, + "grad_norm": 0.8731878399848938, + "learning_rate": 4.385101381140633e-06, + "loss": 0.613, + "step": 3975 + }, + { + "epoch": 1.3960674157303372, + "grad_norm": 0.8665676712989807, + "learning_rate": 4.384797680789592e-06, + "loss": 0.6171, + "step": 3976 + }, + { + "epoch": 1.3964185393258428, + "grad_norm": 0.9031374454498291, + "learning_rate": 4.384493915980083e-06, + "loss": 0.6174, + "step": 3977 + }, + { + "epoch": 1.3967696629213484, + "grad_norm": 0.8580880761146545, + "learning_rate": 4.3841900867224955e-06, + "loss": 0.6338, + "step": 3978 + }, + { + "epoch": 1.397120786516854, + "grad_norm": 0.9038748741149902, + "learning_rate": 4.383886193027219e-06, + "loss": 0.6258, + "step": 3979 + }, + { + "epoch": 1.3974719101123596, + "grad_norm": 0.9072511196136475, + "learning_rate": 4.383582234904646e-06, + "loss": 0.6096, + "step": 3980 + }, + { + "epoch": 1.3978230337078652, + "grad_norm": 0.8500005602836609, + "learning_rate": 4.383278212365174e-06, + "loss": 0.6366, + "step": 3981 + }, + { + "epoch": 1.3981741573033708, + "grad_norm": 0.8831138610839844, + "learning_rate": 4.382974125419197e-06, + "loss": 0.6157, + "step": 3982 + }, + { + "epoch": 1.3985252808988764, + "grad_norm": 0.9004981517791748, + "learning_rate": 4.382669974077117e-06, + "loss": 0.6431, + "step": 3983 + }, + { + "epoch": 1.398876404494382, + "grad_norm": 0.8857392072677612, + "learning_rate": 4.382365758349335e-06, + "loss": 0.6413, + "step": 3984 + }, + { + "epoch": 1.3992275280898876, + "grad_norm": 0.9009760022163391, + "learning_rate": 4.382061478246255e-06, + "loss": 0.6753, + "step": 3985 + }, + { + "epoch": 1.3995786516853932, + "grad_norm": 0.8752977252006531, + "learning_rate": 4.381757133778283e-06, + "loss": 0.6297, + "step": 3986 + }, + { + "epoch": 1.3999297752808988, + "grad_norm": 0.8759447336196899, + "learning_rate": 4.381452724955826e-06, + "loss": 0.6598, + "step": 3987 + }, + { + "epoch": 1.4002808988764044, + "grad_norm": 0.8644441962242126, + "learning_rate": 4.381148251789297e-06, + "loss": 0.6237, + "step": 3988 + }, + { + "epoch": 1.40063202247191, + "grad_norm": 0.8749090433120728, + "learning_rate": 4.380843714289107e-06, + "loss": 0.6475, + "step": 3989 + }, + { + "epoch": 1.4009831460674158, + "grad_norm": 0.8733727335929871, + "learning_rate": 4.380539112465672e-06, + "loss": 0.6421, + "step": 3990 + }, + { + "epoch": 1.4013342696629214, + "grad_norm": 0.8818216919898987, + "learning_rate": 4.3802344463294075e-06, + "loss": 0.6208, + "step": 3991 + }, + { + "epoch": 1.401685393258427, + "grad_norm": 0.8928992748260498, + "learning_rate": 4.379929715890735e-06, + "loss": 0.6075, + "step": 3992 + }, + { + "epoch": 1.4020365168539326, + "grad_norm": 0.9166174530982971, + "learning_rate": 4.3796249211600745e-06, + "loss": 0.6294, + "step": 3993 + }, + { + "epoch": 1.4023876404494382, + "grad_norm": 0.9182286858558655, + "learning_rate": 4.37932006214785e-06, + "loss": 0.6487, + "step": 3994 + }, + { + "epoch": 1.4027387640449438, + "grad_norm": 0.917757511138916, + "learning_rate": 4.379015138864487e-06, + "loss": 0.6424, + "step": 3995 + }, + { + "epoch": 1.4030898876404494, + "grad_norm": 0.912039041519165, + "learning_rate": 4.378710151320416e-06, + "loss": 0.635, + "step": 3996 + }, + { + "epoch": 1.403441011235955, + "grad_norm": 0.8986327648162842, + "learning_rate": 4.378405099526064e-06, + "loss": 0.6258, + "step": 3997 + }, + { + "epoch": 1.4037921348314606, + "grad_norm": 0.8343982100486755, + "learning_rate": 4.3780999834918645e-06, + "loss": 0.5669, + "step": 3998 + }, + { + "epoch": 1.4041432584269664, + "grad_norm": 0.8634805679321289, + "learning_rate": 4.377794803228254e-06, + "loss": 0.6714, + "step": 3999 + }, + { + "epoch": 1.404494382022472, + "grad_norm": 0.8799702525138855, + "learning_rate": 4.3774895587456664e-06, + "loss": 0.6306, + "step": 4000 + }, + { + "epoch": 1.4048455056179776, + "grad_norm": 0.8644582629203796, + "learning_rate": 4.377184250054544e-06, + "loss": 0.5853, + "step": 4001 + }, + { + "epoch": 1.4051966292134832, + "grad_norm": 0.8828877210617065, + "learning_rate": 4.3768788771653255e-06, + "loss": 0.6351, + "step": 4002 + }, + { + "epoch": 1.4055477528089888, + "grad_norm": 0.8081719875335693, + "learning_rate": 4.3765734400884565e-06, + "loss": 0.5686, + "step": 4003 + }, + { + "epoch": 1.4058988764044944, + "grad_norm": 0.873838484287262, + "learning_rate": 4.376267938834381e-06, + "loss": 0.5937, + "step": 4004 + }, + { + "epoch": 1.40625, + "grad_norm": 0.8522796630859375, + "learning_rate": 4.3759623734135485e-06, + "loss": 0.5996, + "step": 4005 + }, + { + "epoch": 1.4066011235955056, + "grad_norm": 0.8799022436141968, + "learning_rate": 4.3756567438364075e-06, + "loss": 0.61, + "step": 4006 + }, + { + "epoch": 1.4069522471910112, + "grad_norm": 0.9355555176734924, + "learning_rate": 4.37535105011341e-06, + "loss": 0.6227, + "step": 4007 + }, + { + "epoch": 1.4073033707865168, + "grad_norm": 0.91303950548172, + "learning_rate": 4.375045292255012e-06, + "loss": 0.6437, + "step": 4008 + }, + { + "epoch": 1.4076544943820224, + "grad_norm": 0.8792590498924255, + "learning_rate": 4.37473947027167e-06, + "loss": 0.6135, + "step": 4009 + }, + { + "epoch": 1.408005617977528, + "grad_norm": 0.9129072427749634, + "learning_rate": 4.3744335841738415e-06, + "loss": 0.646, + "step": 4010 + }, + { + "epoch": 1.4083567415730336, + "grad_norm": 0.8504462838172913, + "learning_rate": 4.37412763397199e-06, + "loss": 0.5935, + "step": 4011 + }, + { + "epoch": 1.4087078651685394, + "grad_norm": 0.8766793608665466, + "learning_rate": 4.373821619676575e-06, + "loss": 0.652, + "step": 4012 + }, + { + "epoch": 1.409058988764045, + "grad_norm": 0.8757210373878479, + "learning_rate": 4.373515541298066e-06, + "loss": 0.6559, + "step": 4013 + }, + { + "epoch": 1.4094101123595506, + "grad_norm": 0.8983198404312134, + "learning_rate": 4.373209398846927e-06, + "loss": 0.6412, + "step": 4014 + }, + { + "epoch": 1.4097612359550562, + "grad_norm": 0.8636668920516968, + "learning_rate": 4.37290319233363e-06, + "loss": 0.6076, + "step": 4015 + }, + { + "epoch": 1.4101123595505618, + "grad_norm": 0.9178102016448975, + "learning_rate": 4.372596921768647e-06, + "loss": 0.6572, + "step": 4016 + }, + { + "epoch": 1.4104634831460674, + "grad_norm": 0.8619410991668701, + "learning_rate": 4.372290587162451e-06, + "loss": 0.6625, + "step": 4017 + }, + { + "epoch": 1.410814606741573, + "grad_norm": 0.8461251854896545, + "learning_rate": 4.37198418852552e-06, + "loss": 0.623, + "step": 4018 + }, + { + "epoch": 1.4111657303370786, + "grad_norm": 0.867741584777832, + "learning_rate": 4.37167772586833e-06, + "loss": 0.6362, + "step": 4019 + }, + { + "epoch": 1.4115168539325842, + "grad_norm": 0.8633602261543274, + "learning_rate": 4.3713711992013656e-06, + "loss": 0.592, + "step": 4020 + }, + { + "epoch": 1.41186797752809, + "grad_norm": 0.8842202425003052, + "learning_rate": 4.371064608535106e-06, + "loss": 0.632, + "step": 4021 + }, + { + "epoch": 1.4122191011235956, + "grad_norm": 0.8827239274978638, + "learning_rate": 4.370757953880038e-06, + "loss": 0.6164, + "step": 4022 + }, + { + "epoch": 1.4125702247191012, + "grad_norm": 0.8834839463233948, + "learning_rate": 4.370451235246649e-06, + "loss": 0.5874, + "step": 4023 + }, + { + "epoch": 1.4129213483146068, + "grad_norm": 0.9291568398475647, + "learning_rate": 4.370144452645429e-06, + "loss": 0.6186, + "step": 4024 + }, + { + "epoch": 1.4132724719101124, + "grad_norm": 0.8816742897033691, + "learning_rate": 4.369837606086869e-06, + "loss": 0.6692, + "step": 4025 + }, + { + "epoch": 1.413623595505618, + "grad_norm": 0.8751959204673767, + "learning_rate": 4.369530695581461e-06, + "loss": 0.6598, + "step": 4026 + }, + { + "epoch": 1.4139747191011236, + "grad_norm": 0.896353006362915, + "learning_rate": 4.369223721139705e-06, + "loss": 0.6605, + "step": 4027 + }, + { + "epoch": 1.4143258426966292, + "grad_norm": 0.8932797908782959, + "learning_rate": 4.368916682772097e-06, + "loss": 0.6395, + "step": 4028 + }, + { + "epoch": 1.4146769662921348, + "grad_norm": 0.8765178918838501, + "learning_rate": 4.368609580489137e-06, + "loss": 0.6298, + "step": 4029 + }, + { + "epoch": 1.4150280898876404, + "grad_norm": 0.872016429901123, + "learning_rate": 4.368302414301328e-06, + "loss": 0.6291, + "step": 4030 + }, + { + "epoch": 1.415379213483146, + "grad_norm": 0.878186821937561, + "learning_rate": 4.367995184219176e-06, + "loss": 0.6139, + "step": 4031 + }, + { + "epoch": 1.4157303370786516, + "grad_norm": 0.8804723024368286, + "learning_rate": 4.367687890253187e-06, + "loss": 0.6372, + "step": 4032 + }, + { + "epoch": 1.4160814606741572, + "grad_norm": 0.8700343370437622, + "learning_rate": 4.3673805324138705e-06, + "loss": 0.5968, + "step": 4033 + }, + { + "epoch": 1.4164325842696628, + "grad_norm": 0.8827090859413147, + "learning_rate": 4.367073110711738e-06, + "loss": 0.5944, + "step": 4034 + }, + { + "epoch": 1.4167837078651686, + "grad_norm": 0.8662370443344116, + "learning_rate": 4.366765625157302e-06, + "loss": 0.6339, + "step": 4035 + }, + { + "epoch": 1.4171348314606742, + "grad_norm": 0.8922637701034546, + "learning_rate": 4.3664580757610795e-06, + "loss": 0.6869, + "step": 4036 + }, + { + "epoch": 1.4174859550561798, + "grad_norm": 0.9283745288848877, + "learning_rate": 4.366150462533588e-06, + "loss": 0.6271, + "step": 4037 + }, + { + "epoch": 1.4178370786516854, + "grad_norm": 0.8803569674491882, + "learning_rate": 4.365842785485348e-06, + "loss": 0.6337, + "step": 4038 + }, + { + "epoch": 1.418188202247191, + "grad_norm": 0.8726440072059631, + "learning_rate": 4.36553504462688e-06, + "loss": 0.64, + "step": 4039 + }, + { + "epoch": 1.4185393258426966, + "grad_norm": 0.882938027381897, + "learning_rate": 4.3652272399687105e-06, + "loss": 0.6372, + "step": 4040 + }, + { + "epoch": 1.4188904494382022, + "grad_norm": 0.8737468123435974, + "learning_rate": 4.364919371521365e-06, + "loss": 0.6284, + "step": 4041 + }, + { + "epoch": 1.4192415730337078, + "grad_norm": 0.8580338954925537, + "learning_rate": 4.364611439295374e-06, + "loss": 0.5805, + "step": 4042 + }, + { + "epoch": 1.4195926966292136, + "grad_norm": 0.8792635202407837, + "learning_rate": 4.364303443301267e-06, + "loss": 0.5764, + "step": 4043 + }, + { + "epoch": 1.4199438202247192, + "grad_norm": 0.9049398303031921, + "learning_rate": 4.363995383549578e-06, + "loss": 0.6422, + "step": 4044 + }, + { + "epoch": 1.4202949438202248, + "grad_norm": 0.9274637699127197, + "learning_rate": 4.363687260050841e-06, + "loss": 0.6415, + "step": 4045 + }, + { + "epoch": 1.4206460674157304, + "grad_norm": 0.9082589149475098, + "learning_rate": 4.363379072815594e-06, + "loss": 0.6201, + "step": 4046 + }, + { + "epoch": 1.420997191011236, + "grad_norm": 0.866824209690094, + "learning_rate": 4.363070821854378e-06, + "loss": 0.5856, + "step": 4047 + }, + { + "epoch": 1.4213483146067416, + "grad_norm": 0.8817905187606812, + "learning_rate": 4.362762507177735e-06, + "loss": 0.5998, + "step": 4048 + }, + { + "epoch": 1.4216994382022472, + "grad_norm": 0.8592423796653748, + "learning_rate": 4.3624541287962065e-06, + "loss": 0.6326, + "step": 4049 + }, + { + "epoch": 1.4220505617977528, + "grad_norm": 0.8892386555671692, + "learning_rate": 4.3621456867203414e-06, + "loss": 0.5931, + "step": 4050 + }, + { + "epoch": 1.4224016853932584, + "grad_norm": 0.8761540055274963, + "learning_rate": 4.361837180960687e-06, + "loss": 0.6055, + "step": 4051 + }, + { + "epoch": 1.422752808988764, + "grad_norm": 0.8538140654563904, + "learning_rate": 4.361528611527795e-06, + "loss": 0.6098, + "step": 4052 + }, + { + "epoch": 1.4231039325842696, + "grad_norm": 0.9002669453620911, + "learning_rate": 4.3612199784322165e-06, + "loss": 0.6199, + "step": 4053 + }, + { + "epoch": 1.4234550561797752, + "grad_norm": 0.8686503767967224, + "learning_rate": 4.3609112816845074e-06, + "loss": 0.6301, + "step": 4054 + }, + { + "epoch": 1.4238061797752808, + "grad_norm": 0.8846096992492676, + "learning_rate": 4.360602521295226e-06, + "loss": 0.6258, + "step": 4055 + }, + { + "epoch": 1.4241573033707864, + "grad_norm": 0.9569324851036072, + "learning_rate": 4.36029369727493e-06, + "loss": 0.6516, + "step": 4056 + }, + { + "epoch": 1.4245084269662922, + "grad_norm": 0.8662309050559998, + "learning_rate": 4.35998480963418e-06, + "loss": 0.6247, + "step": 4057 + }, + { + "epoch": 1.4248595505617978, + "grad_norm": 0.8644338250160217, + "learning_rate": 4.359675858383543e-06, + "loss": 0.6197, + "step": 4058 + }, + { + "epoch": 1.4252106741573034, + "grad_norm": 0.8778071999549866, + "learning_rate": 4.359366843533583e-06, + "loss": 0.576, + "step": 4059 + }, + { + "epoch": 1.425561797752809, + "grad_norm": 0.8852523565292358, + "learning_rate": 4.359057765094866e-06, + "loss": 0.6389, + "step": 4060 + }, + { + "epoch": 1.4259129213483146, + "grad_norm": 0.9235852956771851, + "learning_rate": 4.358748623077966e-06, + "loss": 0.614, + "step": 4061 + }, + { + "epoch": 1.4262640449438202, + "grad_norm": 0.8874046206474304, + "learning_rate": 4.358439417493453e-06, + "loss": 0.6236, + "step": 4062 + }, + { + "epoch": 1.4266151685393258, + "grad_norm": 0.8729009628295898, + "learning_rate": 4.3581301483519036e-06, + "loss": 0.611, + "step": 4063 + }, + { + "epoch": 1.4269662921348314, + "grad_norm": 0.9033231735229492, + "learning_rate": 4.3578208156638915e-06, + "loss": 0.6216, + "step": 4064 + }, + { + "epoch": 1.4273174157303372, + "grad_norm": 0.8786650896072388, + "learning_rate": 4.3575114194399974e-06, + "loss": 0.6198, + "step": 4065 + }, + { + "epoch": 1.4276685393258428, + "grad_norm": 0.887387752532959, + "learning_rate": 4.357201959690802e-06, + "loss": 0.6346, + "step": 4066 + }, + { + "epoch": 1.4280196629213484, + "grad_norm": 0.8734090924263, + "learning_rate": 4.356892436426891e-06, + "loss": 0.6107, + "step": 4067 + }, + { + "epoch": 1.428370786516854, + "grad_norm": 0.8874668478965759, + "learning_rate": 4.356582849658845e-06, + "loss": 0.6454, + "step": 4068 + }, + { + "epoch": 1.4287219101123596, + "grad_norm": 0.8698024749755859, + "learning_rate": 4.356273199397256e-06, + "loss": 0.6529, + "step": 4069 + }, + { + "epoch": 1.4290730337078652, + "grad_norm": 0.8861244916915894, + "learning_rate": 4.355963485652711e-06, + "loss": 0.594, + "step": 4070 + }, + { + "epoch": 1.4294241573033708, + "grad_norm": 0.8716466426849365, + "learning_rate": 4.355653708435803e-06, + "loss": 0.6367, + "step": 4071 + }, + { + "epoch": 1.4297752808988764, + "grad_norm": 0.8842322826385498, + "learning_rate": 4.355343867757127e-06, + "loss": 0.6313, + "step": 4072 + }, + { + "epoch": 1.430126404494382, + "grad_norm": 0.8881300687789917, + "learning_rate": 4.355033963627277e-06, + "loss": 0.624, + "step": 4073 + }, + { + "epoch": 1.4304775280898876, + "grad_norm": 0.8615336418151855, + "learning_rate": 4.354723996056855e-06, + "loss": 0.6307, + "step": 4074 + }, + { + "epoch": 1.4308286516853932, + "grad_norm": 0.8932889699935913, + "learning_rate": 4.354413965056457e-06, + "loss": 0.6211, + "step": 4075 + }, + { + "epoch": 1.4311797752808988, + "grad_norm": 0.8843944668769836, + "learning_rate": 4.354103870636689e-06, + "loss": 0.6482, + "step": 4076 + }, + { + "epoch": 1.4315308988764044, + "grad_norm": 0.8917078375816345, + "learning_rate": 4.353793712808156e-06, + "loss": 0.6044, + "step": 4077 + }, + { + "epoch": 1.43188202247191, + "grad_norm": 0.8852561712265015, + "learning_rate": 4.353483491581463e-06, + "loss": 0.6213, + "step": 4078 + }, + { + "epoch": 1.4322331460674158, + "grad_norm": 0.8828893899917603, + "learning_rate": 4.3531732069672215e-06, + "loss": 0.6676, + "step": 4079 + }, + { + "epoch": 1.4325842696629214, + "grad_norm": 0.8896101713180542, + "learning_rate": 4.352862858976043e-06, + "loss": 0.6277, + "step": 4080 + }, + { + "epoch": 1.432935393258427, + "grad_norm": 0.8827319145202637, + "learning_rate": 4.352552447618539e-06, + "loss": 0.6316, + "step": 4081 + }, + { + "epoch": 1.4332865168539326, + "grad_norm": 0.9116552472114563, + "learning_rate": 4.352241972905326e-06, + "loss": 0.6307, + "step": 4082 + }, + { + "epoch": 1.4336376404494382, + "grad_norm": 0.880622386932373, + "learning_rate": 4.351931434847023e-06, + "loss": 0.6102, + "step": 4083 + }, + { + "epoch": 1.4339887640449438, + "grad_norm": 0.8774731159210205, + "learning_rate": 4.351620833454251e-06, + "loss": 0.6243, + "step": 4084 + }, + { + "epoch": 1.4343398876404494, + "grad_norm": 0.8572145104408264, + "learning_rate": 4.351310168737629e-06, + "loss": 0.6149, + "step": 4085 + }, + { + "epoch": 1.434691011235955, + "grad_norm": 0.8389350175857544, + "learning_rate": 4.350999440707785e-06, + "loss": 0.597, + "step": 4086 + }, + { + "epoch": 1.4350421348314606, + "grad_norm": 0.8829629421234131, + "learning_rate": 4.350688649375343e-06, + "loss": 0.6257, + "step": 4087 + }, + { + "epoch": 1.4353932584269664, + "grad_norm": 0.901107132434845, + "learning_rate": 4.350377794750933e-06, + "loss": 0.6468, + "step": 4088 + }, + { + "epoch": 1.435744382022472, + "grad_norm": 0.8687376379966736, + "learning_rate": 4.350066876845186e-06, + "loss": 0.6274, + "step": 4089 + }, + { + "epoch": 1.4360955056179776, + "grad_norm": 0.8897804617881775, + "learning_rate": 4.349755895668735e-06, + "loss": 0.6661, + "step": 4090 + }, + { + "epoch": 1.4364466292134832, + "grad_norm": 0.8794514536857605, + "learning_rate": 4.349444851232215e-06, + "loss": 0.6543, + "step": 4091 + }, + { + "epoch": 1.4367977528089888, + "grad_norm": 0.8795956969261169, + "learning_rate": 4.349133743546265e-06, + "loss": 0.6164, + "step": 4092 + }, + { + "epoch": 1.4371488764044944, + "grad_norm": 0.8774768114089966, + "learning_rate": 4.348822572621521e-06, + "loss": 0.6115, + "step": 4093 + }, + { + "epoch": 1.4375, + "grad_norm": 0.8642878532409668, + "learning_rate": 4.348511338468629e-06, + "loss": 0.6548, + "step": 4094 + }, + { + "epoch": 1.4378511235955056, + "grad_norm": 0.9389157295227051, + "learning_rate": 4.34820004109823e-06, + "loss": 0.6053, + "step": 4095 + }, + { + "epoch": 1.4382022471910112, + "grad_norm": 0.8678531050682068, + "learning_rate": 4.347888680520971e-06, + "loss": 0.6254, + "step": 4096 + }, + { + "epoch": 1.4385533707865168, + "grad_norm": 0.8993881940841675, + "learning_rate": 4.347577256747501e-06, + "loss": 0.6723, + "step": 4097 + }, + { + "epoch": 1.4389044943820224, + "grad_norm": 0.9129367470741272, + "learning_rate": 4.3472657697884705e-06, + "loss": 0.6197, + "step": 4098 + }, + { + "epoch": 1.439255617977528, + "grad_norm": 0.8977501392364502, + "learning_rate": 4.34695421965453e-06, + "loss": 0.6027, + "step": 4099 + }, + { + "epoch": 1.4396067415730336, + "grad_norm": 0.887247622013092, + "learning_rate": 4.346642606356336e-06, + "loss": 0.5911, + "step": 4100 + }, + { + "epoch": 1.4399578651685394, + "grad_norm": 0.8997041583061218, + "learning_rate": 4.346330929904546e-06, + "loss": 0.6262, + "step": 4101 + }, + { + "epoch": 1.440308988764045, + "grad_norm": 0.885622501373291, + "learning_rate": 4.346019190309817e-06, + "loss": 0.6288, + "step": 4102 + }, + { + "epoch": 1.4406601123595506, + "grad_norm": 0.8826760649681091, + "learning_rate": 4.345707387582811e-06, + "loss": 0.6669, + "step": 4103 + }, + { + "epoch": 1.4410112359550562, + "grad_norm": 0.883783221244812, + "learning_rate": 4.345395521734192e-06, + "loss": 0.6235, + "step": 4104 + }, + { + "epoch": 1.4413623595505618, + "grad_norm": 0.8693239092826843, + "learning_rate": 4.345083592774626e-06, + "loss": 0.6066, + "step": 4105 + }, + { + "epoch": 1.4417134831460674, + "grad_norm": 0.8620202541351318, + "learning_rate": 4.34477160071478e-06, + "loss": 0.6027, + "step": 4106 + }, + { + "epoch": 1.442064606741573, + "grad_norm": 0.8705936074256897, + "learning_rate": 4.344459545565324e-06, + "loss": 0.6173, + "step": 4107 + }, + { + "epoch": 1.4424157303370786, + "grad_norm": 0.8645413517951965, + "learning_rate": 4.344147427336929e-06, + "loss": 0.6598, + "step": 4108 + }, + { + "epoch": 1.4427668539325842, + "grad_norm": 0.8727925419807434, + "learning_rate": 4.34383524604027e-06, + "loss": 0.6537, + "step": 4109 + }, + { + "epoch": 1.44311797752809, + "grad_norm": 0.8704939484596252, + "learning_rate": 4.343523001686024e-06, + "loss": 0.6272, + "step": 4110 + }, + { + "epoch": 1.4434691011235956, + "grad_norm": 0.8718854188919067, + "learning_rate": 4.343210694284869e-06, + "loss": 0.6491, + "step": 4111 + }, + { + "epoch": 1.4438202247191012, + "grad_norm": 0.8282216787338257, + "learning_rate": 4.342898323847486e-06, + "loss": 0.559, + "step": 4112 + }, + { + "epoch": 1.4441713483146068, + "grad_norm": 0.8932397961616516, + "learning_rate": 4.3425858903845555e-06, + "loss": 0.6539, + "step": 4113 + }, + { + "epoch": 1.4445224719101124, + "grad_norm": 0.8598067164421082, + "learning_rate": 4.342273393906765e-06, + "loss": 0.585, + "step": 4114 + }, + { + "epoch": 1.444873595505618, + "grad_norm": 0.9065674543380737, + "learning_rate": 4.3419608344248015e-06, + "loss": 0.6237, + "step": 4115 + }, + { + "epoch": 1.4452247191011236, + "grad_norm": 0.8612712621688843, + "learning_rate": 4.341648211949353e-06, + "loss": 0.5795, + "step": 4116 + }, + { + "epoch": 1.4455758426966292, + "grad_norm": 0.9073783755302429, + "learning_rate": 4.341335526491112e-06, + "loss": 0.6542, + "step": 4117 + }, + { + "epoch": 1.4459269662921348, + "grad_norm": 0.8671945333480835, + "learning_rate": 4.341022778060772e-06, + "loss": 0.6384, + "step": 4118 + }, + { + "epoch": 1.4462780898876404, + "grad_norm": 0.8586475849151611, + "learning_rate": 4.340709966669028e-06, + "loss": 0.6043, + "step": 4119 + }, + { + "epoch": 1.446629213483146, + "grad_norm": 0.8946781754493713, + "learning_rate": 4.340397092326578e-06, + "loss": 0.635, + "step": 4120 + }, + { + "epoch": 1.4469803370786516, + "grad_norm": 0.8610424995422363, + "learning_rate": 4.340084155044122e-06, + "loss": 0.6365, + "step": 4121 + }, + { + "epoch": 1.4473314606741572, + "grad_norm": 0.8891358971595764, + "learning_rate": 4.339771154832364e-06, + "loss": 0.6054, + "step": 4122 + }, + { + "epoch": 1.4476825842696628, + "grad_norm": 0.8908564448356628, + "learning_rate": 4.339458091702005e-06, + "loss": 0.6168, + "step": 4123 + }, + { + "epoch": 1.4480337078651686, + "grad_norm": 0.8783697485923767, + "learning_rate": 4.339144965663754e-06, + "loss": 0.6002, + "step": 4124 + }, + { + "epoch": 1.4483848314606742, + "grad_norm": 0.9083595871925354, + "learning_rate": 4.3388317767283195e-06, + "loss": 0.6357, + "step": 4125 + }, + { + "epoch": 1.4487359550561798, + "grad_norm": 0.8990292549133301, + "learning_rate": 4.338518524906411e-06, + "loss": 0.6422, + "step": 4126 + }, + { + "epoch": 1.4490870786516854, + "grad_norm": 0.9985833168029785, + "learning_rate": 4.338205210208742e-06, + "loss": 0.5973, + "step": 4127 + }, + { + "epoch": 1.449438202247191, + "grad_norm": 0.8731505870819092, + "learning_rate": 4.337891832646028e-06, + "loss": 0.6414, + "step": 4128 + }, + { + "epoch": 1.4497893258426966, + "grad_norm": 0.9245936274528503, + "learning_rate": 4.337578392228987e-06, + "loss": 0.6543, + "step": 4129 + }, + { + "epoch": 1.4501404494382022, + "grad_norm": 0.9523738622665405, + "learning_rate": 4.337264888968337e-06, + "loss": 0.6389, + "step": 4130 + }, + { + "epoch": 1.4504915730337078, + "grad_norm": 0.9052737355232239, + "learning_rate": 4.336951322874799e-06, + "loss": 0.6343, + "step": 4131 + }, + { + "epoch": 1.4508426966292136, + "grad_norm": 0.9124898910522461, + "learning_rate": 4.336637693959099e-06, + "loss": 0.6346, + "step": 4132 + }, + { + "epoch": 1.4511938202247192, + "grad_norm": 0.8696838617324829, + "learning_rate": 4.336324002231961e-06, + "loss": 0.6181, + "step": 4133 + }, + { + "epoch": 1.4515449438202248, + "grad_norm": 0.8571642637252808, + "learning_rate": 4.336010247704113e-06, + "loss": 0.6331, + "step": 4134 + }, + { + "epoch": 1.4518960674157304, + "grad_norm": 0.88227379322052, + "learning_rate": 4.335696430386286e-06, + "loss": 0.585, + "step": 4135 + }, + { + "epoch": 1.452247191011236, + "grad_norm": 0.8858445286750793, + "learning_rate": 4.335382550289212e-06, + "loss": 0.6268, + "step": 4136 + }, + { + "epoch": 1.4525983146067416, + "grad_norm": 0.877275288105011, + "learning_rate": 4.335068607423624e-06, + "loss": 0.6317, + "step": 4137 + }, + { + "epoch": 1.4529494382022472, + "grad_norm": 0.8988421559333801, + "learning_rate": 4.334754601800263e-06, + "loss": 0.6305, + "step": 4138 + }, + { + "epoch": 1.4533005617977528, + "grad_norm": 0.9341934323310852, + "learning_rate": 4.334440533429862e-06, + "loss": 0.614, + "step": 4139 + }, + { + "epoch": 1.4536516853932584, + "grad_norm": 0.8875921964645386, + "learning_rate": 4.3341264023231655e-06, + "loss": 0.6094, + "step": 4140 + }, + { + "epoch": 1.454002808988764, + "grad_norm": 0.8801622986793518, + "learning_rate": 4.333812208490915e-06, + "loss": 0.5936, + "step": 4141 + }, + { + "epoch": 1.4543539325842696, + "grad_norm": 0.9054430723190308, + "learning_rate": 4.333497951943856e-06, + "loss": 0.6518, + "step": 4142 + }, + { + "epoch": 1.4547050561797752, + "grad_norm": 0.8847231268882751, + "learning_rate": 4.333183632692737e-06, + "loss": 0.644, + "step": 4143 + }, + { + "epoch": 1.4550561797752808, + "grad_norm": 0.8920429348945618, + "learning_rate": 4.332869250748306e-06, + "loss": 0.6457, + "step": 4144 + }, + { + "epoch": 1.4554073033707864, + "grad_norm": 0.9018110632896423, + "learning_rate": 4.332554806121315e-06, + "loss": 0.6295, + "step": 4145 + }, + { + "epoch": 1.4557584269662922, + "grad_norm": 0.8690390586853027, + "learning_rate": 4.332240298822517e-06, + "loss": 0.6273, + "step": 4146 + }, + { + "epoch": 1.4561095505617978, + "grad_norm": 0.861956000328064, + "learning_rate": 4.33192572886267e-06, + "loss": 0.6252, + "step": 4147 + }, + { + "epoch": 1.4564606741573034, + "grad_norm": 0.8745825886726379, + "learning_rate": 4.33161109625253e-06, + "loss": 0.6328, + "step": 4148 + }, + { + "epoch": 1.456811797752809, + "grad_norm": 0.8459515571594238, + "learning_rate": 4.331296401002858e-06, + "loss": 0.5466, + "step": 4149 + }, + { + "epoch": 1.4571629213483146, + "grad_norm": 0.8832082152366638, + "learning_rate": 4.330981643124416e-06, + "loss": 0.6516, + "step": 4150 + }, + { + "epoch": 1.4575140449438202, + "grad_norm": 0.8869217038154602, + "learning_rate": 4.330666822627968e-06, + "loss": 0.5984, + "step": 4151 + }, + { + "epoch": 1.4578651685393258, + "grad_norm": 0.8645246624946594, + "learning_rate": 4.330351939524282e-06, + "loss": 0.5963, + "step": 4152 + }, + { + "epoch": 1.4582162921348314, + "grad_norm": 0.8830666542053223, + "learning_rate": 4.330036993824126e-06, + "loss": 0.6349, + "step": 4153 + }, + { + "epoch": 1.4585674157303372, + "grad_norm": 0.9177601933479309, + "learning_rate": 4.329721985538271e-06, + "loss": 0.6322, + "step": 4154 + }, + { + "epoch": 1.4589185393258428, + "grad_norm": 0.9256566762924194, + "learning_rate": 4.329406914677489e-06, + "loss": 0.6912, + "step": 4155 + }, + { + "epoch": 1.4592696629213484, + "grad_norm": 0.879478931427002, + "learning_rate": 4.329091781252556e-06, + "loss": 0.6133, + "step": 4156 + }, + { + "epoch": 1.459620786516854, + "grad_norm": 0.8938345909118652, + "learning_rate": 4.32877658527425e-06, + "loss": 0.5806, + "step": 4157 + }, + { + "epoch": 1.4599719101123596, + "grad_norm": 0.8846132159233093, + "learning_rate": 4.328461326753349e-06, + "loss": 0.6136, + "step": 4158 + }, + { + "epoch": 1.4603230337078652, + "grad_norm": 0.8530312180519104, + "learning_rate": 4.328146005700636e-06, + "loss": 0.6212, + "step": 4159 + }, + { + "epoch": 1.4606741573033708, + "grad_norm": 0.8560202121734619, + "learning_rate": 4.327830622126893e-06, + "loss": 0.6011, + "step": 4160 + }, + { + "epoch": 1.4610252808988764, + "grad_norm": 0.8946338295936584, + "learning_rate": 4.327515176042908e-06, + "loss": 0.5978, + "step": 4161 + }, + { + "epoch": 1.461376404494382, + "grad_norm": 0.89649498462677, + "learning_rate": 4.327199667459467e-06, + "loss": 0.6511, + "step": 4162 + }, + { + "epoch": 1.4617275280898876, + "grad_norm": 0.8728743195533752, + "learning_rate": 4.326884096387361e-06, + "loss": 0.614, + "step": 4163 + }, + { + "epoch": 1.4620786516853932, + "grad_norm": 0.8919811844825745, + "learning_rate": 4.326568462837381e-06, + "loss": 0.6341, + "step": 4164 + }, + { + "epoch": 1.4624297752808988, + "grad_norm": 0.8965336680412292, + "learning_rate": 4.326252766820325e-06, + "loss": 0.5954, + "step": 4165 + }, + { + "epoch": 1.4627808988764044, + "grad_norm": 0.919897198677063, + "learning_rate": 4.325937008346985e-06, + "loss": 0.6758, + "step": 4166 + }, + { + "epoch": 1.46313202247191, + "grad_norm": 0.8864952921867371, + "learning_rate": 4.325621187428164e-06, + "loss": 0.6116, + "step": 4167 + }, + { + "epoch": 1.4634831460674158, + "grad_norm": 0.895219624042511, + "learning_rate": 4.325305304074659e-06, + "loss": 0.6196, + "step": 4168 + }, + { + "epoch": 1.4638342696629214, + "grad_norm": 0.8681285381317139, + "learning_rate": 4.3249893582972745e-06, + "loss": 0.6074, + "step": 4169 + }, + { + "epoch": 1.464185393258427, + "grad_norm": 0.8704788088798523, + "learning_rate": 4.324673350106816e-06, + "loss": 0.6145, + "step": 4170 + }, + { + "epoch": 1.4645365168539326, + "grad_norm": 0.8976466655731201, + "learning_rate": 4.324357279514091e-06, + "loss": 0.6499, + "step": 4171 + }, + { + "epoch": 1.4648876404494382, + "grad_norm": 0.8723233938217163, + "learning_rate": 4.324041146529907e-06, + "loss": 0.6227, + "step": 4172 + }, + { + "epoch": 1.4652387640449438, + "grad_norm": 0.9027929902076721, + "learning_rate": 4.323724951165077e-06, + "loss": 0.6279, + "step": 4173 + }, + { + "epoch": 1.4655898876404494, + "grad_norm": 0.859115481376648, + "learning_rate": 4.323408693430414e-06, + "loss": 0.6295, + "step": 4174 + }, + { + "epoch": 1.465941011235955, + "grad_norm": 1.1383740901947021, + "learning_rate": 4.323092373336735e-06, + "loss": 0.6494, + "step": 4175 + }, + { + "epoch": 1.4662921348314606, + "grad_norm": 0.8614624738693237, + "learning_rate": 4.322775990894856e-06, + "loss": 0.5949, + "step": 4176 + }, + { + "epoch": 1.4666432584269664, + "grad_norm": 0.8514102697372437, + "learning_rate": 4.322459546115598e-06, + "loss": 0.5966, + "step": 4177 + }, + { + "epoch": 1.466994382022472, + "grad_norm": 0.899704098701477, + "learning_rate": 4.322143039009783e-06, + "loss": 0.6548, + "step": 4178 + }, + { + "epoch": 1.4673455056179776, + "grad_norm": 0.8559503555297852, + "learning_rate": 4.321826469588236e-06, + "loss": 0.6039, + "step": 4179 + }, + { + "epoch": 1.4676966292134832, + "grad_norm": 0.8766810297966003, + "learning_rate": 4.3215098378617815e-06, + "loss": 0.588, + "step": 4180 + }, + { + "epoch": 1.4680477528089888, + "grad_norm": 0.8811195492744446, + "learning_rate": 4.32119314384125e-06, + "loss": 0.6208, + "step": 4181 + }, + { + "epoch": 1.4683988764044944, + "grad_norm": 0.9778435826301575, + "learning_rate": 4.320876387537472e-06, + "loss": 0.6445, + "step": 4182 + }, + { + "epoch": 1.46875, + "grad_norm": 0.905834436416626, + "learning_rate": 4.320559568961279e-06, + "loss": 0.6574, + "step": 4183 + }, + { + "epoch": 1.4691011235955056, + "grad_norm": 0.8865072131156921, + "learning_rate": 4.320242688123506e-06, + "loss": 0.5757, + "step": 4184 + }, + { + "epoch": 1.4694522471910112, + "grad_norm": 0.8838146328926086, + "learning_rate": 4.319925745034991e-06, + "loss": 0.634, + "step": 4185 + }, + { + "epoch": 1.4698033707865168, + "grad_norm": 0.8725399971008301, + "learning_rate": 4.3196087397065734e-06, + "loss": 0.6075, + "step": 4186 + }, + { + "epoch": 1.4701544943820224, + "grad_norm": 0.8557268977165222, + "learning_rate": 4.3192916721490945e-06, + "loss": 0.6138, + "step": 4187 + }, + { + "epoch": 1.470505617977528, + "grad_norm": 0.8887699246406555, + "learning_rate": 4.318974542373396e-06, + "loss": 0.6389, + "step": 4188 + }, + { + "epoch": 1.4708567415730336, + "grad_norm": 0.8605521321296692, + "learning_rate": 4.318657350390324e-06, + "loss": 0.6261, + "step": 4189 + }, + { + "epoch": 1.4712078651685394, + "grad_norm": 0.8946020007133484, + "learning_rate": 4.318340096210729e-06, + "loss": 0.6294, + "step": 4190 + }, + { + "epoch": 1.471558988764045, + "grad_norm": 0.8895167708396912, + "learning_rate": 4.318022779845457e-06, + "loss": 0.629, + "step": 4191 + }, + { + "epoch": 1.4719101123595506, + "grad_norm": 0.8478034138679504, + "learning_rate": 4.317705401305362e-06, + "loss": 0.6042, + "step": 4192 + }, + { + "epoch": 1.4722612359550562, + "grad_norm": 0.8884312510490417, + "learning_rate": 4.317387960601298e-06, + "loss": 0.6217, + "step": 4193 + }, + { + "epoch": 1.4726123595505618, + "grad_norm": 0.8619526624679565, + "learning_rate": 4.317070457744121e-06, + "loss": 0.5889, + "step": 4194 + }, + { + "epoch": 1.4729634831460674, + "grad_norm": 0.8887055516242981, + "learning_rate": 4.316752892744689e-06, + "loss": 0.6346, + "step": 4195 + }, + { + "epoch": 1.473314606741573, + "grad_norm": 0.8569642901420593, + "learning_rate": 4.316435265613863e-06, + "loss": 0.622, + "step": 4196 + }, + { + "epoch": 1.4736657303370786, + "grad_norm": 0.8919498920440674, + "learning_rate": 4.316117576362505e-06, + "loss": 0.6371, + "step": 4197 + }, + { + "epoch": 1.4740168539325842, + "grad_norm": 0.8512890338897705, + "learning_rate": 4.31579982500148e-06, + "loss": 0.6495, + "step": 4198 + }, + { + "epoch": 1.47436797752809, + "grad_norm": 0.8871182203292847, + "learning_rate": 4.315482011541655e-06, + "loss": 0.6166, + "step": 4199 + }, + { + "epoch": 1.4747191011235956, + "grad_norm": 0.8572426438331604, + "learning_rate": 4.315164135993898e-06, + "loss": 0.6101, + "step": 4200 + }, + { + "epoch": 1.4750702247191012, + "grad_norm": 0.8520934581756592, + "learning_rate": 4.314846198369082e-06, + "loss": 0.6039, + "step": 4201 + }, + { + "epoch": 1.4754213483146068, + "grad_norm": 0.9315975308418274, + "learning_rate": 4.314528198678079e-06, + "loss": 0.6408, + "step": 4202 + }, + { + "epoch": 1.4757724719101124, + "grad_norm": 0.8353304862976074, + "learning_rate": 4.314210136931762e-06, + "loss": 0.5972, + "step": 4203 + }, + { + "epoch": 1.476123595505618, + "grad_norm": 0.8691760897636414, + "learning_rate": 4.313892013141013e-06, + "loss": 0.5916, + "step": 4204 + }, + { + "epoch": 1.4764747191011236, + "grad_norm": 0.8935512900352478, + "learning_rate": 4.313573827316709e-06, + "loss": 0.6402, + "step": 4205 + }, + { + "epoch": 1.4768258426966292, + "grad_norm": 0.8638583421707153, + "learning_rate": 4.313255579469732e-06, + "loss": 0.621, + "step": 4206 + }, + { + "epoch": 1.4771769662921348, + "grad_norm": 0.9053982496261597, + "learning_rate": 4.312937269610965e-06, + "loss": 0.652, + "step": 4207 + }, + { + "epoch": 1.4775280898876404, + "grad_norm": 0.8318531513214111, + "learning_rate": 4.312618897751295e-06, + "loss": 0.5976, + "step": 4208 + }, + { + "epoch": 1.477879213483146, + "grad_norm": 0.8552941083908081, + "learning_rate": 4.3123004639016105e-06, + "loss": 0.5748, + "step": 4209 + }, + { + "epoch": 1.4782303370786516, + "grad_norm": 0.8920894265174866, + "learning_rate": 4.3119819680728e-06, + "loss": 0.6487, + "step": 4210 + }, + { + "epoch": 1.4785814606741572, + "grad_norm": 0.9560785293579102, + "learning_rate": 4.311663410275757e-06, + "loss": 0.6285, + "step": 4211 + }, + { + "epoch": 1.4789325842696628, + "grad_norm": 1.0505443811416626, + "learning_rate": 4.3113447905213765e-06, + "loss": 0.5522, + "step": 4212 + }, + { + "epoch": 1.4792837078651686, + "grad_norm": 0.8848596811294556, + "learning_rate": 4.311026108820553e-06, + "loss": 0.6029, + "step": 4213 + }, + { + "epoch": 1.4796348314606742, + "grad_norm": 0.8782453536987305, + "learning_rate": 4.310707365184187e-06, + "loss": 0.6234, + "step": 4214 + }, + { + "epoch": 1.4799859550561798, + "grad_norm": 0.8838267922401428, + "learning_rate": 4.310388559623179e-06, + "loss": 0.6206, + "step": 4215 + }, + { + "epoch": 1.4803370786516854, + "grad_norm": 0.9158840179443359, + "learning_rate": 4.31006969214843e-06, + "loss": 0.6414, + "step": 4216 + }, + { + "epoch": 1.480688202247191, + "grad_norm": 0.861907422542572, + "learning_rate": 4.309750762770847e-06, + "loss": 0.6319, + "step": 4217 + }, + { + "epoch": 1.4810393258426966, + "grad_norm": 0.8838783502578735, + "learning_rate": 4.3094317715013376e-06, + "loss": 0.5957, + "step": 4218 + }, + { + "epoch": 1.4813904494382022, + "grad_norm": 0.8620510101318359, + "learning_rate": 4.309112718350809e-06, + "loss": 0.6096, + "step": 4219 + }, + { + "epoch": 1.4817415730337078, + "grad_norm": 0.8699179887771606, + "learning_rate": 4.308793603330173e-06, + "loss": 0.6448, + "step": 4220 + }, + { + "epoch": 1.4820926966292136, + "grad_norm": 0.8801244497299194, + "learning_rate": 4.308474426450344e-06, + "loss": 0.6395, + "step": 4221 + }, + { + "epoch": 1.4824438202247192, + "grad_norm": 0.8840081691741943, + "learning_rate": 4.308155187722238e-06, + "loss": 0.6654, + "step": 4222 + }, + { + "epoch": 1.4827949438202248, + "grad_norm": 0.8977017402648926, + "learning_rate": 4.307835887156771e-06, + "loss": 0.6508, + "step": 4223 + }, + { + "epoch": 1.4831460674157304, + "grad_norm": 0.878200113773346, + "learning_rate": 4.307516524764864e-06, + "loss": 0.5663, + "step": 4224 + }, + { + "epoch": 1.483497191011236, + "grad_norm": 0.8168246150016785, + "learning_rate": 4.307197100557438e-06, + "loss": 0.6026, + "step": 4225 + }, + { + "epoch": 1.4838483146067416, + "grad_norm": 0.8974123597145081, + "learning_rate": 4.306877614545419e-06, + "loss": 0.6149, + "step": 4226 + }, + { + "epoch": 1.4841994382022472, + "grad_norm": 0.8998949527740479, + "learning_rate": 4.30655806673973e-06, + "loss": 0.6391, + "step": 4227 + }, + { + "epoch": 1.4845505617977528, + "grad_norm": 0.9134969115257263, + "learning_rate": 4.306238457151302e-06, + "loss": 0.6334, + "step": 4228 + }, + { + "epoch": 1.4849016853932584, + "grad_norm": 0.8966532349586487, + "learning_rate": 4.305918785791064e-06, + "loss": 0.6537, + "step": 4229 + }, + { + "epoch": 1.485252808988764, + "grad_norm": 0.900520384311676, + "learning_rate": 4.305599052669948e-06, + "loss": 0.6194, + "step": 4230 + }, + { + "epoch": 1.4856039325842696, + "grad_norm": 0.8717507719993591, + "learning_rate": 4.3052792577988915e-06, + "loss": 0.6231, + "step": 4231 + }, + { + "epoch": 1.4859550561797752, + "grad_norm": 0.9033085107803345, + "learning_rate": 4.304959401188828e-06, + "loss": 0.6555, + "step": 4232 + }, + { + "epoch": 1.4863061797752808, + "grad_norm": 0.853325605392456, + "learning_rate": 4.304639482850697e-06, + "loss": 0.6104, + "step": 4233 + }, + { + "epoch": 1.4866573033707864, + "grad_norm": 0.8766210675239563, + "learning_rate": 4.304319502795441e-06, + "loss": 0.6367, + "step": 4234 + }, + { + "epoch": 1.4870084269662922, + "grad_norm": 0.8499912619590759, + "learning_rate": 4.303999461034002e-06, + "loss": 0.6074, + "step": 4235 + }, + { + "epoch": 1.4873595505617978, + "grad_norm": 0.8618026375770569, + "learning_rate": 4.303679357577325e-06, + "loss": 0.6203, + "step": 4236 + }, + { + "epoch": 1.4877106741573034, + "grad_norm": 0.9145092964172363, + "learning_rate": 4.303359192436357e-06, + "loss": 0.6205, + "step": 4237 + }, + { + "epoch": 1.488061797752809, + "grad_norm": 0.8490617275238037, + "learning_rate": 4.303038965622048e-06, + "loss": 0.5845, + "step": 4238 + }, + { + "epoch": 1.4884129213483146, + "grad_norm": 0.8706423044204712, + "learning_rate": 4.3027186771453485e-06, + "loss": 0.6171, + "step": 4239 + }, + { + "epoch": 1.4887640449438202, + "grad_norm": 1.030942440032959, + "learning_rate": 4.302398327017214e-06, + "loss": 0.6191, + "step": 4240 + }, + { + "epoch": 1.4891151685393258, + "grad_norm": 0.8936895728111267, + "learning_rate": 4.3020779152485986e-06, + "loss": 0.6483, + "step": 4241 + }, + { + "epoch": 1.4894662921348314, + "grad_norm": 0.8798382878303528, + "learning_rate": 4.30175744185046e-06, + "loss": 0.619, + "step": 4242 + }, + { + "epoch": 1.4898174157303372, + "grad_norm": 0.8868904709815979, + "learning_rate": 4.30143690683376e-06, + "loss": 0.6033, + "step": 4243 + }, + { + "epoch": 1.4901685393258428, + "grad_norm": 0.8829790353775024, + "learning_rate": 4.301116310209459e-06, + "loss": 0.6283, + "step": 4244 + }, + { + "epoch": 1.4905196629213484, + "grad_norm": 0.8758097887039185, + "learning_rate": 4.3007956519885205e-06, + "loss": 0.5976, + "step": 4245 + }, + { + "epoch": 1.490870786516854, + "grad_norm": 0.865586519241333, + "learning_rate": 4.300474932181912e-06, + "loss": 0.6278, + "step": 4246 + }, + { + "epoch": 1.4912219101123596, + "grad_norm": 0.8885650038719177, + "learning_rate": 4.300154150800602e-06, + "loss": 0.5966, + "step": 4247 + }, + { + "epoch": 1.4915730337078652, + "grad_norm": 0.9123331308364868, + "learning_rate": 4.29983330785556e-06, + "loss": 0.6209, + "step": 4248 + }, + { + "epoch": 1.4919241573033708, + "grad_norm": 0.9102221131324768, + "learning_rate": 4.299512403357758e-06, + "loss": 0.6052, + "step": 4249 + }, + { + "epoch": 1.4922752808988764, + "grad_norm": 0.8728795051574707, + "learning_rate": 4.299191437318173e-06, + "loss": 0.5967, + "step": 4250 + }, + { + "epoch": 1.492626404494382, + "grad_norm": 0.8740513920783997, + "learning_rate": 4.29887040974778e-06, + "loss": 0.6309, + "step": 4251 + }, + { + "epoch": 1.4929775280898876, + "grad_norm": 0.8648720979690552, + "learning_rate": 4.298549320657559e-06, + "loss": 0.625, + "step": 4252 + }, + { + "epoch": 1.4933286516853932, + "grad_norm": 0.9204023480415344, + "learning_rate": 4.29822817005849e-06, + "loss": 0.6341, + "step": 4253 + }, + { + "epoch": 1.4936797752808988, + "grad_norm": 0.877215564250946, + "learning_rate": 4.2979069579615566e-06, + "loss": 0.6547, + "step": 4254 + }, + { + "epoch": 1.4940308988764044, + "grad_norm": 0.8751193881034851, + "learning_rate": 4.297585684377742e-06, + "loss": 0.6074, + "step": 4255 + }, + { + "epoch": 1.49438202247191, + "grad_norm": 0.8720304369926453, + "learning_rate": 4.297264349318037e-06, + "loss": 0.6214, + "step": 4256 + }, + { + "epoch": 1.4947331460674158, + "grad_norm": 0.8909294605255127, + "learning_rate": 4.296942952793429e-06, + "loss": 0.6194, + "step": 4257 + }, + { + "epoch": 1.4950842696629214, + "grad_norm": 0.8968172669410706, + "learning_rate": 4.296621494814909e-06, + "loss": 0.6196, + "step": 4258 + }, + { + "epoch": 1.495435393258427, + "grad_norm": 0.8939587473869324, + "learning_rate": 4.296299975393472e-06, + "loss": 0.67, + "step": 4259 + }, + { + "epoch": 1.4957865168539326, + "grad_norm": 0.9261441826820374, + "learning_rate": 4.2959783945401125e-06, + "loss": 0.6346, + "step": 4260 + }, + { + "epoch": 1.4961376404494382, + "grad_norm": 0.8645102381706238, + "learning_rate": 4.295656752265829e-06, + "loss": 0.6358, + "step": 4261 + }, + { + "epoch": 1.4964887640449438, + "grad_norm": 0.8751295804977417, + "learning_rate": 4.2953350485816205e-06, + "loss": 0.6288, + "step": 4262 + }, + { + "epoch": 1.4968398876404494, + "grad_norm": 0.9099578857421875, + "learning_rate": 4.29501328349849e-06, + "loss": 0.6367, + "step": 4263 + }, + { + "epoch": 1.497191011235955, + "grad_norm": 0.8653101325035095, + "learning_rate": 4.294691457027441e-06, + "loss": 0.6113, + "step": 4264 + }, + { + "epoch": 1.4975421348314606, + "grad_norm": 0.8679119348526001, + "learning_rate": 4.29436956917948e-06, + "loss": 0.6003, + "step": 4265 + }, + { + "epoch": 1.4978932584269664, + "grad_norm": 0.8774658441543579, + "learning_rate": 4.294047619965616e-06, + "loss": 0.62, + "step": 4266 + }, + { + "epoch": 1.498244382022472, + "grad_norm": 0.8665967583656311, + "learning_rate": 4.293725609396858e-06, + "loss": 0.6274, + "step": 4267 + }, + { + "epoch": 1.4985955056179776, + "grad_norm": 0.8959212899208069, + "learning_rate": 4.293403537484219e-06, + "loss": 0.5988, + "step": 4268 + }, + { + "epoch": 1.4989466292134832, + "grad_norm": 0.8877567648887634, + "learning_rate": 4.293081404238714e-06, + "loss": 0.6771, + "step": 4269 + }, + { + "epoch": 1.4992977528089888, + "grad_norm": 0.8664157390594482, + "learning_rate": 4.292759209671359e-06, + "loss": 0.6581, + "step": 4270 + }, + { + "epoch": 1.4996488764044944, + "grad_norm": 0.9112155437469482, + "learning_rate": 4.292436953793173e-06, + "loss": 0.6558, + "step": 4271 + }, + { + "epoch": 1.5, + "grad_norm": 0.8777109980583191, + "learning_rate": 4.292114636615176e-06, + "loss": 0.6097, + "step": 4272 + }, + { + "epoch": 1.5003511235955056, + "grad_norm": 0.8593363761901855, + "learning_rate": 4.291792258148394e-06, + "loss": 0.6, + "step": 4273 + }, + { + "epoch": 1.5007022471910112, + "grad_norm": 0.8953386545181274, + "learning_rate": 4.291469818403849e-06, + "loss": 0.6188, + "step": 4274 + }, + { + "epoch": 1.5010533707865168, + "grad_norm": 0.8898375630378723, + "learning_rate": 4.291147317392569e-06, + "loss": 0.6432, + "step": 4275 + }, + { + "epoch": 1.5014044943820224, + "grad_norm": 0.8718478679656982, + "learning_rate": 4.290824755125584e-06, + "loss": 0.5929, + "step": 4276 + }, + { + "epoch": 1.501755617977528, + "grad_norm": 0.8857758641242981, + "learning_rate": 4.290502131613925e-06, + "loss": 0.6434, + "step": 4277 + }, + { + "epoch": 1.5021067415730336, + "grad_norm": 0.8710552453994751, + "learning_rate": 4.290179446868624e-06, + "loss": 0.5877, + "step": 4278 + }, + { + "epoch": 1.5024578651685392, + "grad_norm": 0.8936992883682251, + "learning_rate": 4.289856700900719e-06, + "loss": 0.5977, + "step": 4279 + }, + { + "epoch": 1.5028089887640448, + "grad_norm": 0.8890673518180847, + "learning_rate": 4.289533893721246e-06, + "loss": 0.662, + "step": 4280 + }, + { + "epoch": 1.5031601123595506, + "grad_norm": 0.8563557267189026, + "learning_rate": 4.289211025341246e-06, + "loss": 0.6105, + "step": 4281 + }, + { + "epoch": 1.5035112359550562, + "grad_norm": 0.8855664730072021, + "learning_rate": 4.288888095771759e-06, + "loss": 0.6502, + "step": 4282 + }, + { + "epoch": 1.5038623595505618, + "grad_norm": 0.8829025626182556, + "learning_rate": 4.2885651050238294e-06, + "loss": 0.6375, + "step": 4283 + }, + { + "epoch": 1.5042134831460674, + "grad_norm": 0.8763623237609863, + "learning_rate": 4.288242053108505e-06, + "loss": 0.6173, + "step": 4284 + }, + { + "epoch": 1.504564606741573, + "grad_norm": 0.8628937005996704, + "learning_rate": 4.287918940036831e-06, + "loss": 0.6088, + "step": 4285 + }, + { + "epoch": 1.5049157303370788, + "grad_norm": 0.8726745247840881, + "learning_rate": 4.287595765819861e-06, + "loss": 0.6046, + "step": 4286 + }, + { + "epoch": 1.5052668539325844, + "grad_norm": 0.886001706123352, + "learning_rate": 4.287272530468644e-06, + "loss": 0.6755, + "step": 4287 + }, + { + "epoch": 1.50561797752809, + "grad_norm": 0.9122753143310547, + "learning_rate": 4.286949233994238e-06, + "loss": 0.6319, + "step": 4288 + }, + { + "epoch": 1.5059691011235956, + "grad_norm": 0.8860533833503723, + "learning_rate": 4.286625876407695e-06, + "loss": 0.6237, + "step": 4289 + }, + { + "epoch": 1.5063202247191012, + "grad_norm": 0.8959923386573792, + "learning_rate": 4.286302457720077e-06, + "loss": 0.6397, + "step": 4290 + }, + { + "epoch": 1.5066713483146068, + "grad_norm": 0.872547447681427, + "learning_rate": 4.285978977942443e-06, + "loss": 0.6098, + "step": 4291 + }, + { + "epoch": 1.5070224719101124, + "grad_norm": 0.885536253452301, + "learning_rate": 4.2856554370858565e-06, + "loss": 0.6148, + "step": 4292 + }, + { + "epoch": 1.507373595505618, + "grad_norm": 0.8694351315498352, + "learning_rate": 4.2853318351613825e-06, + "loss": 0.6649, + "step": 4293 + }, + { + "epoch": 1.5077247191011236, + "grad_norm": 0.9179470539093018, + "learning_rate": 4.285008172180087e-06, + "loss": 0.6518, + "step": 4294 + }, + { + "epoch": 1.5080758426966292, + "grad_norm": 0.9301983714103699, + "learning_rate": 4.2846844481530395e-06, + "loss": 0.6733, + "step": 4295 + }, + { + "epoch": 1.5084269662921348, + "grad_norm": 0.874769926071167, + "learning_rate": 4.28436066309131e-06, + "loss": 0.6206, + "step": 4296 + }, + { + "epoch": 1.5087780898876404, + "grad_norm": 0.8874285817146301, + "learning_rate": 4.284036817005974e-06, + "loss": 0.6427, + "step": 4297 + }, + { + "epoch": 1.509129213483146, + "grad_norm": 0.8830395340919495, + "learning_rate": 4.283712909908105e-06, + "loss": 0.6715, + "step": 4298 + }, + { + "epoch": 1.5094803370786516, + "grad_norm": 0.8672149777412415, + "learning_rate": 4.283388941808782e-06, + "loss": 0.6019, + "step": 4299 + }, + { + "epoch": 1.5098314606741572, + "grad_norm": 0.9076434969902039, + "learning_rate": 4.283064912719081e-06, + "loss": 0.6157, + "step": 4300 + }, + { + "epoch": 1.5101825842696628, + "grad_norm": 0.8910434246063232, + "learning_rate": 4.282740822650087e-06, + "loss": 0.614, + "step": 4301 + }, + { + "epoch": 1.5105337078651684, + "grad_norm": 0.9061810374259949, + "learning_rate": 4.282416671612883e-06, + "loss": 0.6202, + "step": 4302 + }, + { + "epoch": 1.5108848314606742, + "grad_norm": 0.877909243106842, + "learning_rate": 4.282092459618553e-06, + "loss": 0.618, + "step": 4303 + }, + { + "epoch": 1.5112359550561798, + "grad_norm": 0.8897304534912109, + "learning_rate": 4.281768186678187e-06, + "loss": 0.646, + "step": 4304 + }, + { + "epoch": 1.5115870786516854, + "grad_norm": 0.9377396106719971, + "learning_rate": 4.281443852802873e-06, + "loss": 0.6323, + "step": 4305 + }, + { + "epoch": 1.511938202247191, + "grad_norm": 0.8810194730758667, + "learning_rate": 4.281119458003704e-06, + "loss": 0.6372, + "step": 4306 + }, + { + "epoch": 1.5122893258426966, + "grad_norm": 0.8763102293014526, + "learning_rate": 4.280795002291773e-06, + "loss": 0.6189, + "step": 4307 + }, + { + "epoch": 1.5126404494382022, + "grad_norm": 0.8697223663330078, + "learning_rate": 4.280470485678177e-06, + "loss": 0.6188, + "step": 4308 + }, + { + "epoch": 1.512991573033708, + "grad_norm": 0.8664148449897766, + "learning_rate": 4.2801459081740144e-06, + "loss": 0.6135, + "step": 4309 + }, + { + "epoch": 1.5133426966292136, + "grad_norm": 0.877975583076477, + "learning_rate": 4.279821269790384e-06, + "loss": 0.607, + "step": 4310 + }, + { + "epoch": 1.5136938202247192, + "grad_norm": 0.8779658079147339, + "learning_rate": 4.2794965705383906e-06, + "loss": 0.5998, + "step": 4311 + }, + { + "epoch": 1.5140449438202248, + "grad_norm": 0.8846172094345093, + "learning_rate": 4.279171810429136e-06, + "loss": 0.6212, + "step": 4312 + }, + { + "epoch": 1.5143960674157304, + "grad_norm": 0.8623836040496826, + "learning_rate": 4.278846989473729e-06, + "loss": 0.6333, + "step": 4313 + }, + { + "epoch": 1.514747191011236, + "grad_norm": 0.9182273149490356, + "learning_rate": 4.2785221076832765e-06, + "loss": 0.6423, + "step": 4314 + }, + { + "epoch": 1.5150983146067416, + "grad_norm": 0.879574716091156, + "learning_rate": 4.278197165068891e-06, + "loss": 0.621, + "step": 4315 + }, + { + "epoch": 1.5154494382022472, + "grad_norm": 0.8660203814506531, + "learning_rate": 4.277872161641682e-06, + "loss": 0.582, + "step": 4316 + }, + { + "epoch": 1.5158005617977528, + "grad_norm": 0.8960110545158386, + "learning_rate": 4.277547097412768e-06, + "loss": 0.6073, + "step": 4317 + }, + { + "epoch": 1.5161516853932584, + "grad_norm": 0.8762567639350891, + "learning_rate": 4.277221972393263e-06, + "loss": 0.5839, + "step": 4318 + }, + { + "epoch": 1.516502808988764, + "grad_norm": 0.8783929944038391, + "learning_rate": 4.276896786594288e-06, + "loss": 0.6342, + "step": 4319 + }, + { + "epoch": 1.5168539325842696, + "grad_norm": 0.8779000043869019, + "learning_rate": 4.276571540026964e-06, + "loss": 0.6274, + "step": 4320 + }, + { + "epoch": 1.5172050561797752, + "grad_norm": 0.8748044967651367, + "learning_rate": 4.276246232702413e-06, + "loss": 0.614, + "step": 4321 + }, + { + "epoch": 1.5175561797752808, + "grad_norm": 0.9069294333457947, + "learning_rate": 4.275920864631761e-06, + "loss": 0.6438, + "step": 4322 + }, + { + "epoch": 1.5179073033707864, + "grad_norm": 0.8882784247398376, + "learning_rate": 4.275595435826134e-06, + "loss": 0.6104, + "step": 4323 + }, + { + "epoch": 1.518258426966292, + "grad_norm": 0.8945493698120117, + "learning_rate": 4.275269946296664e-06, + "loss": 0.5955, + "step": 4324 + }, + { + "epoch": 1.5186095505617978, + "grad_norm": 0.9070472121238708, + "learning_rate": 4.27494439605448e-06, + "loss": 0.6506, + "step": 4325 + }, + { + "epoch": 1.5189606741573034, + "grad_norm": 0.8988576531410217, + "learning_rate": 4.274618785110718e-06, + "loss": 0.619, + "step": 4326 + }, + { + "epoch": 1.519311797752809, + "grad_norm": 0.8801273107528687, + "learning_rate": 4.274293113476511e-06, + "loss": 0.598, + "step": 4327 + }, + { + "epoch": 1.5196629213483146, + "grad_norm": 0.8812402486801147, + "learning_rate": 4.273967381162998e-06, + "loss": 0.6274, + "step": 4328 + }, + { + "epoch": 1.5200140449438202, + "grad_norm": 0.8644037246704102, + "learning_rate": 4.273641588181317e-06, + "loss": 0.5863, + "step": 4329 + }, + { + "epoch": 1.5203651685393258, + "grad_norm": 0.8854338526725769, + "learning_rate": 4.2733157345426135e-06, + "loss": 0.6315, + "step": 4330 + }, + { + "epoch": 1.5207162921348316, + "grad_norm": 0.9194030165672302, + "learning_rate": 4.272989820258029e-06, + "loss": 0.6241, + "step": 4331 + }, + { + "epoch": 1.5210674157303372, + "grad_norm": 0.8962981700897217, + "learning_rate": 4.27266384533871e-06, + "loss": 0.5799, + "step": 4332 + }, + { + "epoch": 1.5214185393258428, + "grad_norm": 0.8707279562950134, + "learning_rate": 4.2723378097958035e-06, + "loss": 0.6517, + "step": 4333 + }, + { + "epoch": 1.5217696629213484, + "grad_norm": 0.8625728487968445, + "learning_rate": 4.272011713640461e-06, + "loss": 0.5913, + "step": 4334 + }, + { + "epoch": 1.522120786516854, + "grad_norm": 0.8925690054893494, + "learning_rate": 4.271685556883833e-06, + "loss": 0.6334, + "step": 4335 + }, + { + "epoch": 1.5224719101123596, + "grad_norm": 0.8702839612960815, + "learning_rate": 4.271359339537077e-06, + "loss": 0.5967, + "step": 4336 + }, + { + "epoch": 1.5228230337078652, + "grad_norm": 0.9461783170700073, + "learning_rate": 4.271033061611346e-06, + "loss": 0.634, + "step": 4337 + }, + { + "epoch": 1.5231741573033708, + "grad_norm": 1.0706638097763062, + "learning_rate": 4.2707067231178e-06, + "loss": 0.6449, + "step": 4338 + }, + { + "epoch": 1.5235252808988764, + "grad_norm": 0.9275455474853516, + "learning_rate": 4.2703803240676e-06, + "loss": 0.6322, + "step": 4339 + }, + { + "epoch": 1.523876404494382, + "grad_norm": 0.8851011991500854, + "learning_rate": 4.270053864471907e-06, + "loss": 0.6354, + "step": 4340 + }, + { + "epoch": 1.5242275280898876, + "grad_norm": 0.8991807699203491, + "learning_rate": 4.269727344341887e-06, + "loss": 0.6617, + "step": 4341 + }, + { + "epoch": 1.5245786516853932, + "grad_norm": 0.8674828410148621, + "learning_rate": 4.269400763688706e-06, + "loss": 0.5909, + "step": 4342 + }, + { + "epoch": 1.5249297752808988, + "grad_norm": 0.8902739882469177, + "learning_rate": 4.2690741225235336e-06, + "loss": 0.6346, + "step": 4343 + }, + { + "epoch": 1.5252808988764044, + "grad_norm": 0.9026355743408203, + "learning_rate": 4.2687474208575395e-06, + "loss": 0.6505, + "step": 4344 + }, + { + "epoch": 1.52563202247191, + "grad_norm": 0.8712281584739685, + "learning_rate": 4.268420658701897e-06, + "loss": 0.6541, + "step": 4345 + }, + { + "epoch": 1.5259831460674156, + "grad_norm": 0.874847412109375, + "learning_rate": 4.268093836067781e-06, + "loss": 0.5961, + "step": 4346 + }, + { + "epoch": 1.5263342696629212, + "grad_norm": 0.8955193758010864, + "learning_rate": 4.267766952966369e-06, + "loss": 0.6221, + "step": 4347 + }, + { + "epoch": 1.526685393258427, + "grad_norm": 0.8489906191825867, + "learning_rate": 4.2674400094088405e-06, + "loss": 0.6062, + "step": 4348 + }, + { + "epoch": 1.5270365168539326, + "grad_norm": 0.9131476283073425, + "learning_rate": 4.267113005406375e-06, + "loss": 0.6277, + "step": 4349 + }, + { + "epoch": 1.5273876404494382, + "grad_norm": 0.9100372195243835, + "learning_rate": 4.266785940970158e-06, + "loss": 0.6168, + "step": 4350 + }, + { + "epoch": 1.5277387640449438, + "grad_norm": 0.8778021335601807, + "learning_rate": 4.266458816111373e-06, + "loss": 0.6156, + "step": 4351 + }, + { + "epoch": 1.5280898876404494, + "grad_norm": 0.8872804045677185, + "learning_rate": 4.266131630841208e-06, + "loss": 0.6325, + "step": 4352 + }, + { + "epoch": 1.5284410112359552, + "grad_norm": 0.8724574446678162, + "learning_rate": 4.265804385170853e-06, + "loss": 0.6053, + "step": 4353 + }, + { + "epoch": 1.5287921348314608, + "grad_norm": 0.875560998916626, + "learning_rate": 4.265477079111499e-06, + "loss": 0.6343, + "step": 4354 + }, + { + "epoch": 1.5291432584269664, + "grad_norm": 0.9058078527450562, + "learning_rate": 4.26514971267434e-06, + "loss": 0.6229, + "step": 4355 + }, + { + "epoch": 1.529494382022472, + "grad_norm": 0.8583801984786987, + "learning_rate": 4.26482228587057e-06, + "loss": 0.6371, + "step": 4356 + }, + { + "epoch": 1.5298455056179776, + "grad_norm": 0.8806894421577454, + "learning_rate": 4.264494798711389e-06, + "loss": 0.6322, + "step": 4357 + }, + { + "epoch": 1.5301966292134832, + "grad_norm": 0.8982591032981873, + "learning_rate": 4.264167251207995e-06, + "loss": 0.6166, + "step": 4358 + }, + { + "epoch": 1.5305477528089888, + "grad_norm": 0.8655708432197571, + "learning_rate": 4.263839643371591e-06, + "loss": 0.6254, + "step": 4359 + }, + { + "epoch": 1.5308988764044944, + "grad_norm": 0.8754801154136658, + "learning_rate": 4.263511975213381e-06, + "loss": 0.6228, + "step": 4360 + }, + { + "epoch": 1.53125, + "grad_norm": 0.8953879475593567, + "learning_rate": 4.263184246744571e-06, + "loss": 0.596, + "step": 4361 + }, + { + "epoch": 1.5316011235955056, + "grad_norm": 0.8852868676185608, + "learning_rate": 4.2628564579763675e-06, + "loss": 0.6318, + "step": 4362 + }, + { + "epoch": 1.5319522471910112, + "grad_norm": 0.8706391453742981, + "learning_rate": 4.262528608919981e-06, + "loss": 0.6429, + "step": 4363 + }, + { + "epoch": 1.5323033707865168, + "grad_norm": 0.8404558897018433, + "learning_rate": 4.262200699586626e-06, + "loss": 0.5882, + "step": 4364 + }, + { + "epoch": 1.5326544943820224, + "grad_norm": 0.8769184350967407, + "learning_rate": 4.261872729987514e-06, + "loss": 0.6284, + "step": 4365 + }, + { + "epoch": 1.533005617977528, + "grad_norm": 0.8632640838623047, + "learning_rate": 4.261544700133863e-06, + "loss": 0.5794, + "step": 4366 + }, + { + "epoch": 1.5333567415730336, + "grad_norm": 0.8795828223228455, + "learning_rate": 4.26121661003689e-06, + "loss": 0.6281, + "step": 4367 + }, + { + "epoch": 1.5337078651685392, + "grad_norm": 0.9072569012641907, + "learning_rate": 4.260888459707816e-06, + "loss": 0.6154, + "step": 4368 + }, + { + "epoch": 1.5340589887640448, + "grad_norm": 0.8855779767036438, + "learning_rate": 4.260560249157864e-06, + "loss": 0.6199, + "step": 4369 + }, + { + "epoch": 1.5344101123595506, + "grad_norm": 0.8932167887687683, + "learning_rate": 4.260231978398257e-06, + "loss": 0.6155, + "step": 4370 + }, + { + "epoch": 1.5347612359550562, + "grad_norm": 0.8883824348449707, + "learning_rate": 4.259903647440223e-06, + "loss": 0.6386, + "step": 4371 + }, + { + "epoch": 1.5351123595505618, + "grad_norm": 0.8885374665260315, + "learning_rate": 4.259575256294991e-06, + "loss": 0.6175, + "step": 4372 + }, + { + "epoch": 1.5354634831460674, + "grad_norm": 0.9104458689689636, + "learning_rate": 4.25924680497379e-06, + "loss": 0.6223, + "step": 4373 + }, + { + "epoch": 1.535814606741573, + "grad_norm": 0.8783615231513977, + "learning_rate": 4.258918293487854e-06, + "loss": 0.645, + "step": 4374 + }, + { + "epoch": 1.5361657303370788, + "grad_norm": 0.8793334364891052, + "learning_rate": 4.2585897218484175e-06, + "loss": 0.6762, + "step": 4375 + }, + { + "epoch": 1.5365168539325844, + "grad_norm": 0.8661759495735168, + "learning_rate": 4.258261090066718e-06, + "loss": 0.5936, + "step": 4376 + }, + { + "epoch": 1.53686797752809, + "grad_norm": 0.9374626874923706, + "learning_rate": 4.257932398153993e-06, + "loss": 0.6332, + "step": 4377 + }, + { + "epoch": 1.5372191011235956, + "grad_norm": 0.8962463736534119, + "learning_rate": 4.2576036461214844e-06, + "loss": 0.6104, + "step": 4378 + }, + { + "epoch": 1.5375702247191012, + "grad_norm": 0.8866143822669983, + "learning_rate": 4.257274833980435e-06, + "loss": 0.6105, + "step": 4379 + }, + { + "epoch": 1.5379213483146068, + "grad_norm": 0.9056102633476257, + "learning_rate": 4.256945961742091e-06, + "loss": 0.6392, + "step": 4380 + }, + { + "epoch": 1.5382724719101124, + "grad_norm": 0.8876556158065796, + "learning_rate": 4.256617029417696e-06, + "loss": 0.5981, + "step": 4381 + }, + { + "epoch": 1.538623595505618, + "grad_norm": 0.8786119222640991, + "learning_rate": 4.256288037018504e-06, + "loss": 0.6489, + "step": 4382 + }, + { + "epoch": 1.5389747191011236, + "grad_norm": 0.9107440114021301, + "learning_rate": 4.255958984555764e-06, + "loss": 0.6233, + "step": 4383 + }, + { + "epoch": 1.5393258426966292, + "grad_norm": 0.8715046048164368, + "learning_rate": 4.255629872040728e-06, + "loss": 0.6196, + "step": 4384 + }, + { + "epoch": 1.5396769662921348, + "grad_norm": 0.8908570408821106, + "learning_rate": 4.255300699484654e-06, + "loss": 0.6353, + "step": 4385 + }, + { + "epoch": 1.5400280898876404, + "grad_norm": 0.9205907583236694, + "learning_rate": 4.254971466898798e-06, + "loss": 0.6375, + "step": 4386 + }, + { + "epoch": 1.540379213483146, + "grad_norm": 0.8485553860664368, + "learning_rate": 4.254642174294419e-06, + "loss": 0.5977, + "step": 4387 + }, + { + "epoch": 1.5407303370786516, + "grad_norm": 0.9237390160560608, + "learning_rate": 4.254312821682779e-06, + "loss": 0.6641, + "step": 4388 + }, + { + "epoch": 1.5410814606741572, + "grad_norm": 0.8933415412902832, + "learning_rate": 4.253983409075142e-06, + "loss": 0.6581, + "step": 4389 + }, + { + "epoch": 1.5414325842696628, + "grad_norm": 0.9153465032577515, + "learning_rate": 4.2536539364827734e-06, + "loss": 0.6331, + "step": 4390 + }, + { + "epoch": 1.5417837078651684, + "grad_norm": 0.8514824509620667, + "learning_rate": 4.25332440391694e-06, + "loss": 0.6189, + "step": 4391 + }, + { + "epoch": 1.5421348314606742, + "grad_norm": 0.8722473382949829, + "learning_rate": 4.252994811388914e-06, + "loss": 0.6107, + "step": 4392 + }, + { + "epoch": 1.5424859550561798, + "grad_norm": 0.8800398707389832, + "learning_rate": 4.252665158909964e-06, + "loss": 0.6291, + "step": 4393 + }, + { + "epoch": 1.5428370786516854, + "grad_norm": 0.8885344862937927, + "learning_rate": 4.252335446491366e-06, + "loss": 0.5903, + "step": 4394 + }, + { + "epoch": 1.543188202247191, + "grad_norm": 0.9298628568649292, + "learning_rate": 4.2520056741443956e-06, + "loss": 0.6482, + "step": 4395 + }, + { + "epoch": 1.5435393258426966, + "grad_norm": 0.9046818614006042, + "learning_rate": 4.251675841880329e-06, + "loss": 0.6256, + "step": 4396 + }, + { + "epoch": 1.5438904494382022, + "grad_norm": 0.902118444442749, + "learning_rate": 4.251345949710449e-06, + "loss": 0.6262, + "step": 4397 + }, + { + "epoch": 1.544241573033708, + "grad_norm": 0.8869757652282715, + "learning_rate": 4.251015997646035e-06, + "loss": 0.6324, + "step": 4398 + }, + { + "epoch": 1.5445926966292136, + "grad_norm": 0.8704875707626343, + "learning_rate": 4.250685985698374e-06, + "loss": 0.6217, + "step": 4399 + }, + { + "epoch": 1.5449438202247192, + "grad_norm": 0.8620396852493286, + "learning_rate": 4.250355913878749e-06, + "loss": 0.653, + "step": 4400 + }, + { + "epoch": 1.5452949438202248, + "grad_norm": 0.9138562679290771, + "learning_rate": 4.250025782198449e-06, + "loss": 0.638, + "step": 4401 + }, + { + "epoch": 1.5456460674157304, + "grad_norm": 0.874911904335022, + "learning_rate": 4.249695590668765e-06, + "loss": 0.6063, + "step": 4402 + }, + { + "epoch": 1.545997191011236, + "grad_norm": 0.867830753326416, + "learning_rate": 4.249365339300991e-06, + "loss": 0.6183, + "step": 4403 + }, + { + "epoch": 1.5463483146067416, + "grad_norm": 0.8954204320907593, + "learning_rate": 4.249035028106418e-06, + "loss": 0.6259, + "step": 4404 + }, + { + "epoch": 1.5466994382022472, + "grad_norm": 0.8731714487075806, + "learning_rate": 4.248704657096343e-06, + "loss": 0.601, + "step": 4405 + }, + { + "epoch": 1.5470505617977528, + "grad_norm": 0.8896400332450867, + "learning_rate": 4.248374226282066e-06, + "loss": 0.6124, + "step": 4406 + }, + { + "epoch": 1.5474016853932584, + "grad_norm": 0.8725085258483887, + "learning_rate": 4.248043735674886e-06, + "loss": 0.6295, + "step": 4407 + }, + { + "epoch": 1.547752808988764, + "grad_norm": 0.9611513018608093, + "learning_rate": 4.2477131852861064e-06, + "loss": 0.6177, + "step": 4408 + }, + { + "epoch": 1.5481039325842696, + "grad_norm": 0.8641888499259949, + "learning_rate": 4.247382575127031e-06, + "loss": 0.6317, + "step": 4409 + }, + { + "epoch": 1.5484550561797752, + "grad_norm": 0.9119146466255188, + "learning_rate": 4.247051905208967e-06, + "loss": 0.6246, + "step": 4410 + }, + { + "epoch": 1.5488061797752808, + "grad_norm": 0.8706852197647095, + "learning_rate": 4.246721175543224e-06, + "loss": 0.6083, + "step": 4411 + }, + { + "epoch": 1.5491573033707864, + "grad_norm": 0.8891316056251526, + "learning_rate": 4.246390386141109e-06, + "loss": 0.6409, + "step": 4412 + }, + { + "epoch": 1.549508426966292, + "grad_norm": 0.8789838552474976, + "learning_rate": 4.246059537013939e-06, + "loss": 0.6488, + "step": 4413 + }, + { + "epoch": 1.5498595505617978, + "grad_norm": 0.8697891235351562, + "learning_rate": 4.245728628173027e-06, + "loss": 0.6155, + "step": 4414 + }, + { + "epoch": 1.5502106741573034, + "grad_norm": 0.8879761695861816, + "learning_rate": 4.2453976596296895e-06, + "loss": 0.6194, + "step": 4415 + }, + { + "epoch": 1.550561797752809, + "grad_norm": 0.8760266304016113, + "learning_rate": 4.245066631395245e-06, + "loss": 0.6037, + "step": 4416 + }, + { + "epoch": 1.5509129213483146, + "grad_norm": 0.8705148696899414, + "learning_rate": 4.244735543481016e-06, + "loss": 0.6262, + "step": 4417 + }, + { + "epoch": 1.5512640449438202, + "grad_norm": 0.8751351833343506, + "learning_rate": 4.244404395898324e-06, + "loss": 0.6264, + "step": 4418 + }, + { + "epoch": 1.5516151685393258, + "grad_norm": 0.8804269433021545, + "learning_rate": 4.244073188658494e-06, + "loss": 0.6613, + "step": 4419 + }, + { + "epoch": 1.5519662921348316, + "grad_norm": 0.8883501291275024, + "learning_rate": 4.243741921772853e-06, + "loss": 0.6459, + "step": 4420 + }, + { + "epoch": 1.5523174157303372, + "grad_norm": 0.8871954679489136, + "learning_rate": 4.243410595252731e-06, + "loss": 0.6259, + "step": 4421 + }, + { + "epoch": 1.5526685393258428, + "grad_norm": 0.8802275657653809, + "learning_rate": 4.243079209109458e-06, + "loss": 0.5989, + "step": 4422 + }, + { + "epoch": 1.5530196629213484, + "grad_norm": 0.8572296500205994, + "learning_rate": 4.242747763354369e-06, + "loss": 0.6087, + "step": 4423 + }, + { + "epoch": 1.553370786516854, + "grad_norm": 0.9028926491737366, + "learning_rate": 4.242416257998797e-06, + "loss": 0.5896, + "step": 4424 + }, + { + "epoch": 1.5537219101123596, + "grad_norm": 0.9015838503837585, + "learning_rate": 4.2420846930540795e-06, + "loss": 0.7101, + "step": 4425 + }, + { + "epoch": 1.5540730337078652, + "grad_norm": 0.8723250031471252, + "learning_rate": 4.241753068531557e-06, + "loss": 0.646, + "step": 4426 + }, + { + "epoch": 1.5544241573033708, + "grad_norm": 0.8892289400100708, + "learning_rate": 4.241421384442569e-06, + "loss": 0.6446, + "step": 4427 + }, + { + "epoch": 1.5547752808988764, + "grad_norm": 0.8631646633148193, + "learning_rate": 4.24108964079846e-06, + "loss": 0.5921, + "step": 4428 + }, + { + "epoch": 1.555126404494382, + "grad_norm": 0.8563498258590698, + "learning_rate": 4.2407578376105765e-06, + "loss": 0.5942, + "step": 4429 + }, + { + "epoch": 1.5554775280898876, + "grad_norm": 0.883753776550293, + "learning_rate": 4.2404259748902635e-06, + "loss": 0.6359, + "step": 4430 + }, + { + "epoch": 1.5558286516853932, + "grad_norm": 0.8877705335617065, + "learning_rate": 4.240094052648872e-06, + "loss": 0.6229, + "step": 4431 + }, + { + "epoch": 1.5561797752808988, + "grad_norm": 0.8870841860771179, + "learning_rate": 4.239762070897752e-06, + "loss": 0.6342, + "step": 4432 + }, + { + "epoch": 1.5565308988764044, + "grad_norm": 0.8865689635276794, + "learning_rate": 4.239430029648258e-06, + "loss": 0.64, + "step": 4433 + }, + { + "epoch": 1.55688202247191, + "grad_norm": 0.8670912981033325, + "learning_rate": 4.239097928911746e-06, + "loss": 0.6176, + "step": 4434 + }, + { + "epoch": 1.5572331460674156, + "grad_norm": 0.8641391396522522, + "learning_rate": 4.238765768699573e-06, + "loss": 0.6329, + "step": 4435 + }, + { + "epoch": 1.5575842696629212, + "grad_norm": 0.8882496356964111, + "learning_rate": 4.238433549023098e-06, + "loss": 0.6506, + "step": 4436 + }, + { + "epoch": 1.557935393258427, + "grad_norm": 0.8790398836135864, + "learning_rate": 4.238101269893684e-06, + "loss": 0.6053, + "step": 4437 + }, + { + "epoch": 1.5582865168539326, + "grad_norm": 0.8824111223220825, + "learning_rate": 4.237768931322694e-06, + "loss": 0.6207, + "step": 4438 + }, + { + "epoch": 1.5586376404494382, + "grad_norm": 0.8859694004058838, + "learning_rate": 4.237436533321492e-06, + "loss": 0.6206, + "step": 4439 + }, + { + "epoch": 1.5589887640449438, + "grad_norm": 0.8975111842155457, + "learning_rate": 4.237104075901449e-06, + "loss": 0.6106, + "step": 4440 + }, + { + "epoch": 1.5593398876404494, + "grad_norm": 0.9005839228630066, + "learning_rate": 4.236771559073932e-06, + "loss": 0.6292, + "step": 4441 + }, + { + "epoch": 1.5596910112359552, + "grad_norm": 0.8936960101127625, + "learning_rate": 4.236438982850315e-06, + "loss": 0.6472, + "step": 4442 + }, + { + "epoch": 1.5600421348314608, + "grad_norm": 0.8928436636924744, + "learning_rate": 4.23610634724197e-06, + "loss": 0.6097, + "step": 4443 + }, + { + "epoch": 1.5603932584269664, + "grad_norm": 0.8569137454032898, + "learning_rate": 4.235773652260274e-06, + "loss": 0.5809, + "step": 4444 + }, + { + "epoch": 1.560744382022472, + "grad_norm": 0.8743844032287598, + "learning_rate": 4.235440897916604e-06, + "loss": 0.6311, + "step": 4445 + }, + { + "epoch": 1.5610955056179776, + "grad_norm": 0.9037197828292847, + "learning_rate": 4.23510808422234e-06, + "loss": 0.6386, + "step": 4446 + }, + { + "epoch": 1.5614466292134832, + "grad_norm": 0.8934659361839294, + "learning_rate": 4.234775211188864e-06, + "loss": 0.5906, + "step": 4447 + }, + { + "epoch": 1.5617977528089888, + "grad_norm": 0.8630779981613159, + "learning_rate": 4.23444227882756e-06, + "loss": 0.6175, + "step": 4448 + }, + { + "epoch": 1.5621488764044944, + "grad_norm": 0.9021822214126587, + "learning_rate": 4.234109287149816e-06, + "loss": 0.6217, + "step": 4449 + }, + { + "epoch": 1.5625, + "grad_norm": 0.9597007036209106, + "learning_rate": 4.233776236167017e-06, + "loss": 0.6448, + "step": 4450 + }, + { + "epoch": 1.5628511235955056, + "grad_norm": 0.8997489809989929, + "learning_rate": 4.233443125890555e-06, + "loss": 0.6307, + "step": 4451 + }, + { + "epoch": 1.5632022471910112, + "grad_norm": 0.8864565491676331, + "learning_rate": 4.233109956331821e-06, + "loss": 0.6215, + "step": 4452 + }, + { + "epoch": 1.5635533707865168, + "grad_norm": 0.8683627843856812, + "learning_rate": 4.2327767275022096e-06, + "loss": 0.6235, + "step": 4453 + }, + { + "epoch": 1.5639044943820224, + "grad_norm": 0.8793233036994934, + "learning_rate": 4.232443439413117e-06, + "loss": 0.6551, + "step": 4454 + }, + { + "epoch": 1.564255617977528, + "grad_norm": 0.8933106064796448, + "learning_rate": 4.232110092075941e-06, + "loss": 0.6277, + "step": 4455 + }, + { + "epoch": 1.5646067415730336, + "grad_norm": 0.8762953281402588, + "learning_rate": 4.231776685502083e-06, + "loss": 0.6052, + "step": 4456 + }, + { + "epoch": 1.5649578651685392, + "grad_norm": 0.8579549193382263, + "learning_rate": 4.231443219702943e-06, + "loss": 0.5982, + "step": 4457 + }, + { + "epoch": 1.5653089887640448, + "grad_norm": 0.8534015417098999, + "learning_rate": 4.231109694689927e-06, + "loss": 0.6071, + "step": 4458 + }, + { + "epoch": 1.5656601123595506, + "grad_norm": 0.8755966424942017, + "learning_rate": 4.23077611047444e-06, + "loss": 0.5972, + "step": 4459 + }, + { + "epoch": 1.5660112359550562, + "grad_norm": 0.8907667398452759, + "learning_rate": 4.230442467067892e-06, + "loss": 0.6195, + "step": 4460 + }, + { + "epoch": 1.5663623595505618, + "grad_norm": 0.8510807156562805, + "learning_rate": 4.230108764481693e-06, + "loss": 0.6359, + "step": 4461 + }, + { + "epoch": 1.5667134831460674, + "grad_norm": 0.8813369274139404, + "learning_rate": 4.2297750027272535e-06, + "loss": 0.6517, + "step": 4462 + }, + { + "epoch": 1.567064606741573, + "grad_norm": 0.8754818439483643, + "learning_rate": 4.22944118181599e-06, + "loss": 0.633, + "step": 4463 + }, + { + "epoch": 1.5674157303370788, + "grad_norm": 0.8787099719047546, + "learning_rate": 4.229107301759317e-06, + "loss": 0.6439, + "step": 4464 + }, + { + "epoch": 1.5677668539325844, + "grad_norm": 0.8702832460403442, + "learning_rate": 4.2287733625686545e-06, + "loss": 0.6221, + "step": 4465 + }, + { + "epoch": 1.56811797752809, + "grad_norm": 0.8664124011993408, + "learning_rate": 4.228439364255422e-06, + "loss": 0.6139, + "step": 4466 + }, + { + "epoch": 1.5684691011235956, + "grad_norm": 0.8977587819099426, + "learning_rate": 4.228105306831043e-06, + "loss": 0.6185, + "step": 4467 + }, + { + "epoch": 1.5688202247191012, + "grad_norm": 0.8700708746910095, + "learning_rate": 4.2277711903069405e-06, + "loss": 0.6421, + "step": 4468 + }, + { + "epoch": 1.5691713483146068, + "grad_norm": 0.8524662256240845, + "learning_rate": 4.227437014694542e-06, + "loss": 0.6432, + "step": 4469 + }, + { + "epoch": 1.5695224719101124, + "grad_norm": 0.8722302317619324, + "learning_rate": 4.227102780005275e-06, + "loss": 0.6245, + "step": 4470 + }, + { + "epoch": 1.569873595505618, + "grad_norm": 0.879640519618988, + "learning_rate": 4.226768486250573e-06, + "loss": 0.6451, + "step": 4471 + }, + { + "epoch": 1.5702247191011236, + "grad_norm": 0.855945348739624, + "learning_rate": 4.226434133441865e-06, + "loss": 0.6301, + "step": 4472 + }, + { + "epoch": 1.5705758426966292, + "grad_norm": 0.8772982954978943, + "learning_rate": 4.226099721590586e-06, + "loss": 0.6495, + "step": 4473 + }, + { + "epoch": 1.5709269662921348, + "grad_norm": 0.9376111626625061, + "learning_rate": 4.225765250708175e-06, + "loss": 0.6047, + "step": 4474 + }, + { + "epoch": 1.5712780898876404, + "grad_norm": 0.8725435733795166, + "learning_rate": 4.225430720806069e-06, + "loss": 0.634, + "step": 4475 + }, + { + "epoch": 1.571629213483146, + "grad_norm": 0.8824449181556702, + "learning_rate": 4.225096131895709e-06, + "loss": 0.6396, + "step": 4476 + }, + { + "epoch": 1.5719803370786516, + "grad_norm": 0.8654618859291077, + "learning_rate": 4.2247614839885365e-06, + "loss": 0.6144, + "step": 4477 + }, + { + "epoch": 1.5723314606741572, + "grad_norm": 0.8552867770195007, + "learning_rate": 4.224426777095998e-06, + "loss": 0.5836, + "step": 4478 + }, + { + "epoch": 1.5726825842696628, + "grad_norm": 0.8514997363090515, + "learning_rate": 4.224092011229539e-06, + "loss": 0.5915, + "step": 4479 + }, + { + "epoch": 1.5730337078651684, + "grad_norm": 0.8661342859268188, + "learning_rate": 4.223757186400609e-06, + "loss": 0.6288, + "step": 4480 + }, + { + "epoch": 1.5733848314606742, + "grad_norm": 0.8661715984344482, + "learning_rate": 4.223422302620657e-06, + "loss": 0.6201, + "step": 4481 + }, + { + "epoch": 1.5737359550561798, + "grad_norm": 0.8975440859794617, + "learning_rate": 4.223087359901138e-06, + "loss": 0.7046, + "step": 4482 + }, + { + "epoch": 1.5740870786516854, + "grad_norm": 0.8667924404144287, + "learning_rate": 4.222752358253506e-06, + "loss": 0.6245, + "step": 4483 + }, + { + "epoch": 1.574438202247191, + "grad_norm": 0.8844180107116699, + "learning_rate": 4.222417297689217e-06, + "loss": 0.6172, + "step": 4484 + }, + { + "epoch": 1.5747893258426966, + "grad_norm": 0.8499074578285217, + "learning_rate": 4.22208217821973e-06, + "loss": 0.5952, + "step": 4485 + }, + { + "epoch": 1.5751404494382022, + "grad_norm": 0.9003639221191406, + "learning_rate": 4.221746999856507e-06, + "loss": 0.6488, + "step": 4486 + }, + { + "epoch": 1.575491573033708, + "grad_norm": 0.8823853731155396, + "learning_rate": 4.22141176261101e-06, + "loss": 0.6171, + "step": 4487 + }, + { + "epoch": 1.5758426966292136, + "grad_norm": 0.8788625597953796, + "learning_rate": 4.221076466494702e-06, + "loss": 0.6026, + "step": 4488 + }, + { + "epoch": 1.5761938202247192, + "grad_norm": 0.8390807509422302, + "learning_rate": 4.220741111519053e-06, + "loss": 0.5835, + "step": 4489 + }, + { + "epoch": 1.5765449438202248, + "grad_norm": 0.8773233294487, + "learning_rate": 4.220405697695531e-06, + "loss": 0.644, + "step": 4490 + }, + { + "epoch": 1.5768960674157304, + "grad_norm": 0.8999008536338806, + "learning_rate": 4.220070225035605e-06, + "loss": 0.6453, + "step": 4491 + }, + { + "epoch": 1.577247191011236, + "grad_norm": 0.8900070190429688, + "learning_rate": 4.21973469355075e-06, + "loss": 0.6586, + "step": 4492 + }, + { + "epoch": 1.5775983146067416, + "grad_norm": 0.8822365403175354, + "learning_rate": 4.21939910325244e-06, + "loss": 0.639, + "step": 4493 + }, + { + "epoch": 1.5779494382022472, + "grad_norm": 0.8647001385688782, + "learning_rate": 4.219063454152152e-06, + "loss": 0.6038, + "step": 4494 + }, + { + "epoch": 1.5783005617977528, + "grad_norm": 0.8594561219215393, + "learning_rate": 4.218727746261364e-06, + "loss": 0.6143, + "step": 4495 + }, + { + "epoch": 1.5786516853932584, + "grad_norm": 0.8780025243759155, + "learning_rate": 4.218391979591559e-06, + "loss": 0.6539, + "step": 4496 + }, + { + "epoch": 1.579002808988764, + "grad_norm": 0.8385159373283386, + "learning_rate": 4.218056154154219e-06, + "loss": 0.6088, + "step": 4497 + }, + { + "epoch": 1.5793539325842696, + "grad_norm": 0.8734780550003052, + "learning_rate": 4.217720269960828e-06, + "loss": 0.6463, + "step": 4498 + }, + { + "epoch": 1.5797050561797752, + "grad_norm": 0.8894116282463074, + "learning_rate": 4.217384327022873e-06, + "loss": 0.623, + "step": 4499 + }, + { + "epoch": 1.5800561797752808, + "grad_norm": 0.9236807227134705, + "learning_rate": 4.217048325351844e-06, + "loss": 0.6192, + "step": 4500 + }, + { + "epoch": 1.5804073033707864, + "grad_norm": 0.8841096758842468, + "learning_rate": 4.216712264959232e-06, + "loss": 0.6471, + "step": 4501 + }, + { + "epoch": 1.580758426966292, + "grad_norm": 0.8719488978385925, + "learning_rate": 4.216376145856529e-06, + "loss": 0.6733, + "step": 4502 + }, + { + "epoch": 1.5811095505617978, + "grad_norm": 0.8742629289627075, + "learning_rate": 4.216039968055231e-06, + "loss": 0.6004, + "step": 4503 + }, + { + "epoch": 1.5814606741573034, + "grad_norm": 0.8602683544158936, + "learning_rate": 4.215703731566835e-06, + "loss": 0.6221, + "step": 4504 + }, + { + "epoch": 1.581811797752809, + "grad_norm": 0.8487911224365234, + "learning_rate": 4.215367436402838e-06, + "loss": 0.6398, + "step": 4505 + }, + { + "epoch": 1.5821629213483146, + "grad_norm": 0.8509633541107178, + "learning_rate": 4.215031082574744e-06, + "loss": 0.593, + "step": 4506 + }, + { + "epoch": 1.5825140449438202, + "grad_norm": 0.9125469923019409, + "learning_rate": 4.2146946700940545e-06, + "loss": 0.6602, + "step": 4507 + }, + { + "epoch": 1.5828651685393258, + "grad_norm": 0.908766508102417, + "learning_rate": 4.2143581989722736e-06, + "loss": 0.6807, + "step": 4508 + }, + { + "epoch": 1.5832162921348316, + "grad_norm": 0.893795371055603, + "learning_rate": 4.21402166922091e-06, + "loss": 0.6677, + "step": 4509 + }, + { + "epoch": 1.5835674157303372, + "grad_norm": 0.8966583609580994, + "learning_rate": 4.2136850808514715e-06, + "loss": 0.6459, + "step": 4510 + }, + { + "epoch": 1.5839185393258428, + "grad_norm": 0.8950080871582031, + "learning_rate": 4.21334843387547e-06, + "loss": 0.6565, + "step": 4511 + }, + { + "epoch": 1.5842696629213484, + "grad_norm": 0.85506671667099, + "learning_rate": 4.213011728304418e-06, + "loss": 0.6113, + "step": 4512 + }, + { + "epoch": 1.584620786516854, + "grad_norm": 0.8737542629241943, + "learning_rate": 4.212674964149831e-06, + "loss": 0.6275, + "step": 4513 + }, + { + "epoch": 1.5849719101123596, + "grad_norm": 0.882484495639801, + "learning_rate": 4.212338141423226e-06, + "loss": 0.6476, + "step": 4514 + }, + { + "epoch": 1.5853230337078652, + "grad_norm": 0.8486616015434265, + "learning_rate": 4.212001260136121e-06, + "loss": 0.5877, + "step": 4515 + }, + { + "epoch": 1.5856741573033708, + "grad_norm": 0.8440214991569519, + "learning_rate": 4.2116643203000386e-06, + "loss": 0.5961, + "step": 4516 + }, + { + "epoch": 1.5860252808988764, + "grad_norm": 0.8951239585876465, + "learning_rate": 4.211327321926501e-06, + "loss": 0.6463, + "step": 4517 + }, + { + "epoch": 1.586376404494382, + "grad_norm": 0.8931921124458313, + "learning_rate": 4.210990265027034e-06, + "loss": 0.6554, + "step": 4518 + }, + { + "epoch": 1.5867275280898876, + "grad_norm": 0.8429591655731201, + "learning_rate": 4.210653149613163e-06, + "loss": 0.5997, + "step": 4519 + }, + { + "epoch": 1.5870786516853932, + "grad_norm": 0.8799921274185181, + "learning_rate": 4.210315975696419e-06, + "loss": 0.6236, + "step": 4520 + }, + { + "epoch": 1.5874297752808988, + "grad_norm": 0.8685789108276367, + "learning_rate": 4.209978743288331e-06, + "loss": 0.6138, + "step": 4521 + }, + { + "epoch": 1.5877808988764044, + "grad_norm": 0.8596302270889282, + "learning_rate": 4.2096414524004335e-06, + "loss": 0.5799, + "step": 4522 + }, + { + "epoch": 1.58813202247191, + "grad_norm": 0.8389763832092285, + "learning_rate": 4.209304103044262e-06, + "loss": 0.6066, + "step": 4523 + }, + { + "epoch": 1.5884831460674156, + "grad_norm": 0.8625580072402954, + "learning_rate": 4.2089666952313525e-06, + "loss": 0.6263, + "step": 4524 + }, + { + "epoch": 1.5888342696629212, + "grad_norm": 0.8957053422927856, + "learning_rate": 4.208629228973245e-06, + "loss": 0.6237, + "step": 4525 + }, + { + "epoch": 1.589185393258427, + "grad_norm": 0.8905929327011108, + "learning_rate": 4.208291704281479e-06, + "loss": 0.6389, + "step": 4526 + }, + { + "epoch": 1.5895365168539326, + "grad_norm": 0.9022085070610046, + "learning_rate": 4.207954121167599e-06, + "loss": 0.62, + "step": 4527 + }, + { + "epoch": 1.5898876404494382, + "grad_norm": 0.8556004762649536, + "learning_rate": 4.20761647964315e-06, + "loss": 0.5949, + "step": 4528 + }, + { + "epoch": 1.5902387640449438, + "grad_norm": 0.8509926795959473, + "learning_rate": 4.207278779719677e-06, + "loss": 0.603, + "step": 4529 + }, + { + "epoch": 1.5905898876404494, + "grad_norm": 0.8614581227302551, + "learning_rate": 4.206941021408732e-06, + "loss": 0.6292, + "step": 4530 + }, + { + "epoch": 1.5909410112359552, + "grad_norm": 0.8607187867164612, + "learning_rate": 4.206603204721865e-06, + "loss": 0.5729, + "step": 4531 + }, + { + "epoch": 1.5912921348314608, + "grad_norm": 0.9069685339927673, + "learning_rate": 4.206265329670628e-06, + "loss": 0.6275, + "step": 4532 + }, + { + "epoch": 1.5916432584269664, + "grad_norm": 0.9023653268814087, + "learning_rate": 4.205927396266577e-06, + "loss": 0.6463, + "step": 4533 + }, + { + "epoch": 1.591994382022472, + "grad_norm": 0.9172576665878296, + "learning_rate": 4.205589404521269e-06, + "loss": 0.6653, + "step": 4534 + }, + { + "epoch": 1.5923455056179776, + "grad_norm": 0.8926611542701721, + "learning_rate": 4.205251354446262e-06, + "loss": 0.6522, + "step": 4535 + }, + { + "epoch": 1.5926966292134832, + "grad_norm": 0.8904423713684082, + "learning_rate": 4.204913246053119e-06, + "loss": 0.6635, + "step": 4536 + }, + { + "epoch": 1.5930477528089888, + "grad_norm": 0.876338005065918, + "learning_rate": 4.204575079353403e-06, + "loss": 0.628, + "step": 4537 + }, + { + "epoch": 1.5933988764044944, + "grad_norm": 0.8750518560409546, + "learning_rate": 4.204236854358676e-06, + "loss": 0.6365, + "step": 4538 + }, + { + "epoch": 1.59375, + "grad_norm": 0.8659214973449707, + "learning_rate": 4.203898571080507e-06, + "loss": 0.6157, + "step": 4539 + }, + { + "epoch": 1.5941011235955056, + "grad_norm": 0.869689404964447, + "learning_rate": 4.203560229530465e-06, + "loss": 0.605, + "step": 4540 + }, + { + "epoch": 1.5944522471910112, + "grad_norm": 0.8688238859176636, + "learning_rate": 4.203221829720121e-06, + "loss": 0.6453, + "step": 4541 + }, + { + "epoch": 1.5948033707865168, + "grad_norm": 0.8444859385490417, + "learning_rate": 4.202883371661048e-06, + "loss": 0.6125, + "step": 4542 + }, + { + "epoch": 1.5951544943820224, + "grad_norm": 0.881079912185669, + "learning_rate": 4.202544855364822e-06, + "loss": 0.6082, + "step": 4543 + }, + { + "epoch": 1.595505617977528, + "grad_norm": 0.8665393590927124, + "learning_rate": 4.202206280843017e-06, + "loss": 0.6304, + "step": 4544 + }, + { + "epoch": 1.5958567415730336, + "grad_norm": 0.8872676491737366, + "learning_rate": 4.201867648107215e-06, + "loss": 0.6206, + "step": 4545 + }, + { + "epoch": 1.5962078651685392, + "grad_norm": 0.8978577852249146, + "learning_rate": 4.201528957168995e-06, + "loss": 0.6471, + "step": 4546 + }, + { + "epoch": 1.5965589887640448, + "grad_norm": 0.9299441576004028, + "learning_rate": 4.201190208039942e-06, + "loss": 0.6454, + "step": 4547 + }, + { + "epoch": 1.5969101123595506, + "grad_norm": 0.8918309807777405, + "learning_rate": 4.200851400731638e-06, + "loss": 0.6372, + "step": 4548 + }, + { + "epoch": 1.5972612359550562, + "grad_norm": 0.8987869024276733, + "learning_rate": 4.200512535255672e-06, + "loss": 0.6298, + "step": 4549 + }, + { + "epoch": 1.5976123595505618, + "grad_norm": 0.8805652260780334, + "learning_rate": 4.200173611623632e-06, + "loss": 0.6621, + "step": 4550 + }, + { + "epoch": 1.5979634831460674, + "grad_norm": 0.8915992975234985, + "learning_rate": 4.19983462984711e-06, + "loss": 0.65, + "step": 4551 + }, + { + "epoch": 1.598314606741573, + "grad_norm": 0.8713775873184204, + "learning_rate": 4.199495589937699e-06, + "loss": 0.6122, + "step": 4552 + }, + { + "epoch": 1.5986657303370788, + "grad_norm": 0.8860073089599609, + "learning_rate": 4.199156491906991e-06, + "loss": 0.653, + "step": 4553 + }, + { + "epoch": 1.5990168539325844, + "grad_norm": 0.8860781788825989, + "learning_rate": 4.198817335766587e-06, + "loss": 0.6147, + "step": 4554 + }, + { + "epoch": 1.59936797752809, + "grad_norm": 0.8467510938644409, + "learning_rate": 4.198478121528084e-06, + "loss": 0.6153, + "step": 4555 + }, + { + "epoch": 1.5997191011235956, + "grad_norm": 0.8688376545906067, + "learning_rate": 4.1981388492030815e-06, + "loss": 0.6415, + "step": 4556 + }, + { + "epoch": 1.6000702247191012, + "grad_norm": 0.8537015318870544, + "learning_rate": 4.197799518803184e-06, + "loss": 0.6312, + "step": 4557 + }, + { + "epoch": 1.6004213483146068, + "grad_norm": 0.883955717086792, + "learning_rate": 4.197460130339995e-06, + "loss": 0.6212, + "step": 4558 + }, + { + "epoch": 1.6007724719101124, + "grad_norm": 0.8804914355278015, + "learning_rate": 4.1971206838251225e-06, + "loss": 0.6224, + "step": 4559 + }, + { + "epoch": 1.601123595505618, + "grad_norm": 0.8551623225212097, + "learning_rate": 4.1967811792701745e-06, + "loss": 0.6037, + "step": 4560 + }, + { + "epoch": 1.6014747191011236, + "grad_norm": 0.9019951820373535, + "learning_rate": 4.196441616686763e-06, + "loss": 0.6206, + "step": 4561 + }, + { + "epoch": 1.6018258426966292, + "grad_norm": 0.8744482398033142, + "learning_rate": 4.196101996086499e-06, + "loss": 0.6206, + "step": 4562 + }, + { + "epoch": 1.6021769662921348, + "grad_norm": 0.8476253747940063, + "learning_rate": 4.195762317480998e-06, + "loss": 0.5789, + "step": 4563 + }, + { + "epoch": 1.6025280898876404, + "grad_norm": 0.8534361720085144, + "learning_rate": 4.195422580881878e-06, + "loss": 0.6131, + "step": 4564 + }, + { + "epoch": 1.602879213483146, + "grad_norm": 0.8521122336387634, + "learning_rate": 4.195082786300756e-06, + "loss": 0.6133, + "step": 4565 + }, + { + "epoch": 1.6032303370786516, + "grad_norm": 0.8880335092544556, + "learning_rate": 4.194742933749253e-06, + "loss": 0.6433, + "step": 4566 + }, + { + "epoch": 1.6035814606741572, + "grad_norm": 0.8391385674476624, + "learning_rate": 4.194403023238992e-06, + "loss": 0.5859, + "step": 4567 + }, + { + "epoch": 1.6039325842696628, + "grad_norm": 0.8838050961494446, + "learning_rate": 4.194063054781597e-06, + "loss": 0.5979, + "step": 4568 + }, + { + "epoch": 1.6042837078651684, + "grad_norm": 0.8805633783340454, + "learning_rate": 4.193723028388696e-06, + "loss": 0.6112, + "step": 4569 + }, + { + "epoch": 1.6046348314606742, + "grad_norm": 0.8977530598640442, + "learning_rate": 4.193382944071917e-06, + "loss": 0.6199, + "step": 4570 + }, + { + "epoch": 1.6049859550561798, + "grad_norm": 0.8989500403404236, + "learning_rate": 4.193042801842889e-06, + "loss": 0.6176, + "step": 4571 + }, + { + "epoch": 1.6053370786516854, + "grad_norm": 0.9075813889503479, + "learning_rate": 4.192702601713247e-06, + "loss": 0.6429, + "step": 4572 + }, + { + "epoch": 1.605688202247191, + "grad_norm": 0.9065417051315308, + "learning_rate": 4.192362343694624e-06, + "loss": 0.6387, + "step": 4573 + }, + { + "epoch": 1.6060393258426966, + "grad_norm": 0.8901644349098206, + "learning_rate": 4.1920220277986574e-06, + "loss": 0.6332, + "step": 4574 + }, + { + "epoch": 1.6063904494382022, + "grad_norm": 0.8711532354354858, + "learning_rate": 4.191681654036985e-06, + "loss": 0.6324, + "step": 4575 + }, + { + "epoch": 1.606741573033708, + "grad_norm": 0.8672968745231628, + "learning_rate": 4.191341222421248e-06, + "loss": 0.6165, + "step": 4576 + }, + { + "epoch": 1.6070926966292136, + "grad_norm": 0.8829415440559387, + "learning_rate": 4.191000732963088e-06, + "loss": 0.6228, + "step": 4577 + }, + { + "epoch": 1.6074438202247192, + "grad_norm": 0.8676837086677551, + "learning_rate": 4.19066018567415e-06, + "loss": 0.6452, + "step": 4578 + }, + { + "epoch": 1.6077949438202248, + "grad_norm": 0.8773974180221558, + "learning_rate": 4.19031958056608e-06, + "loss": 0.674, + "step": 4579 + }, + { + "epoch": 1.6081460674157304, + "grad_norm": 0.8626821041107178, + "learning_rate": 4.189978917650528e-06, + "loss": 0.6104, + "step": 4580 + }, + { + "epoch": 1.608497191011236, + "grad_norm": 0.8596952557563782, + "learning_rate": 4.189638196939143e-06, + "loss": 0.63, + "step": 4581 + }, + { + "epoch": 1.6088483146067416, + "grad_norm": 0.880653977394104, + "learning_rate": 4.189297418443576e-06, + "loss": 0.6064, + "step": 4582 + }, + { + "epoch": 1.6091994382022472, + "grad_norm": 0.8691625595092773, + "learning_rate": 4.188956582175484e-06, + "loss": 0.6144, + "step": 4583 + }, + { + "epoch": 1.6095505617977528, + "grad_norm": 0.8744456171989441, + "learning_rate": 4.1886156881465225e-06, + "loss": 0.5914, + "step": 4584 + }, + { + "epoch": 1.6099016853932584, + "grad_norm": 0.9076862931251526, + "learning_rate": 4.188274736368349e-06, + "loss": 0.5998, + "step": 4585 + }, + { + "epoch": 1.610252808988764, + "grad_norm": 0.8850449323654175, + "learning_rate": 4.187933726852624e-06, + "loss": 0.613, + "step": 4586 + }, + { + "epoch": 1.6106039325842696, + "grad_norm": 0.875081479549408, + "learning_rate": 4.187592659611011e-06, + "loss": 0.5947, + "step": 4587 + }, + { + "epoch": 1.6109550561797752, + "grad_norm": 0.9096956849098206, + "learning_rate": 4.187251534655171e-06, + "loss": 0.6624, + "step": 4588 + }, + { + "epoch": 1.6113061797752808, + "grad_norm": 0.8888775110244751, + "learning_rate": 4.1869103519967745e-06, + "loss": 0.634, + "step": 4589 + }, + { + "epoch": 1.6116573033707864, + "grad_norm": 0.8596838712692261, + "learning_rate": 4.186569111647486e-06, + "loss": 0.6214, + "step": 4590 + }, + { + "epoch": 1.612008426966292, + "grad_norm": 0.8774394392967224, + "learning_rate": 4.186227813618977e-06, + "loss": 0.6341, + "step": 4591 + }, + { + "epoch": 1.6123595505617978, + "grad_norm": 0.902068018913269, + "learning_rate": 4.185886457922921e-06, + "loss": 0.6276, + "step": 4592 + }, + { + "epoch": 1.6127106741573034, + "grad_norm": 0.8684345483779907, + "learning_rate": 4.18554504457099e-06, + "loss": 0.5965, + "step": 4593 + }, + { + "epoch": 1.613061797752809, + "grad_norm": 0.8846623301506042, + "learning_rate": 4.185203573574861e-06, + "loss": 0.6202, + "step": 4594 + }, + { + "epoch": 1.6134129213483146, + "grad_norm": 0.8909761309623718, + "learning_rate": 4.1848620449462115e-06, + "loss": 0.6117, + "step": 4595 + }, + { + "epoch": 1.6137640449438202, + "grad_norm": 0.9091597199440002, + "learning_rate": 4.1845204586967225e-06, + "loss": 0.6311, + "step": 4596 + }, + { + "epoch": 1.6141151685393258, + "grad_norm": 0.8552021384239197, + "learning_rate": 4.184178814838075e-06, + "loss": 0.5772, + "step": 4597 + }, + { + "epoch": 1.6144662921348316, + "grad_norm": 0.8852929472923279, + "learning_rate": 4.1838371133819525e-06, + "loss": 0.6538, + "step": 4598 + }, + { + "epoch": 1.6148174157303372, + "grad_norm": 1.1475235223770142, + "learning_rate": 4.183495354340042e-06, + "loss": 0.6104, + "step": 4599 + }, + { + "epoch": 1.6151685393258428, + "grad_norm": 0.8915920257568359, + "learning_rate": 4.1831535377240306e-06, + "loss": 0.5814, + "step": 4600 + }, + { + "epoch": 1.6155196629213484, + "grad_norm": 0.8609941601753235, + "learning_rate": 4.182811663545609e-06, + "loss": 0.591, + "step": 4601 + }, + { + "epoch": 1.615870786516854, + "grad_norm": 0.8900454640388489, + "learning_rate": 4.182469731816468e-06, + "loss": 0.611, + "step": 4602 + }, + { + "epoch": 1.6162219101123596, + "grad_norm": 0.8634515404701233, + "learning_rate": 4.182127742548303e-06, + "loss": 0.5957, + "step": 4603 + }, + { + "epoch": 1.6165730337078652, + "grad_norm": 0.9392664432525635, + "learning_rate": 4.181785695752808e-06, + "loss": 0.6798, + "step": 4604 + }, + { + "epoch": 1.6169241573033708, + "grad_norm": 0.8867204189300537, + "learning_rate": 4.181443591441682e-06, + "loss": 0.6288, + "step": 4605 + }, + { + "epoch": 1.6172752808988764, + "grad_norm": 0.869739294052124, + "learning_rate": 4.181101429626623e-06, + "loss": 0.641, + "step": 4606 + }, + { + "epoch": 1.617626404494382, + "grad_norm": 0.8330792784690857, + "learning_rate": 4.180759210319334e-06, + "loss": 0.5901, + "step": 4607 + }, + { + "epoch": 1.6179775280898876, + "grad_norm": 0.8733412623405457, + "learning_rate": 4.180416933531518e-06, + "loss": 0.6115, + "step": 4608 + }, + { + "epoch": 1.6183286516853932, + "grad_norm": 0.8878810405731201, + "learning_rate": 4.1800745992748815e-06, + "loss": 0.6408, + "step": 4609 + }, + { + "epoch": 1.6186797752808988, + "grad_norm": 0.8582045435905457, + "learning_rate": 4.179732207561131e-06, + "loss": 0.6173, + "step": 4610 + }, + { + "epoch": 1.6190308988764044, + "grad_norm": 0.8719291687011719, + "learning_rate": 4.179389758401976e-06, + "loss": 0.6223, + "step": 4611 + }, + { + "epoch": 1.61938202247191, + "grad_norm": 0.8811583518981934, + "learning_rate": 4.179047251809129e-06, + "loss": 0.6012, + "step": 4612 + }, + { + "epoch": 1.6197331460674156, + "grad_norm": 0.9122189283370972, + "learning_rate": 4.178704687794302e-06, + "loss": 0.6253, + "step": 4613 + }, + { + "epoch": 1.6200842696629212, + "grad_norm": 0.8523516654968262, + "learning_rate": 4.178362066369211e-06, + "loss": 0.6133, + "step": 4614 + }, + { + "epoch": 1.620435393258427, + "grad_norm": 0.9010724425315857, + "learning_rate": 4.1780193875455745e-06, + "loss": 0.6225, + "step": 4615 + }, + { + "epoch": 1.6207865168539326, + "grad_norm": 0.8976810574531555, + "learning_rate": 4.177676651335111e-06, + "loss": 0.6051, + "step": 4616 + }, + { + "epoch": 1.6211376404494382, + "grad_norm": 0.8751901388168335, + "learning_rate": 4.17733385774954e-06, + "loss": 0.6189, + "step": 4617 + }, + { + "epoch": 1.6214887640449438, + "grad_norm": 0.8850089311599731, + "learning_rate": 4.176991006800588e-06, + "loss": 0.6241, + "step": 4618 + }, + { + "epoch": 1.6218398876404494, + "grad_norm": 0.8844864964485168, + "learning_rate": 4.176648098499978e-06, + "loss": 0.605, + "step": 4619 + }, + { + "epoch": 1.6221910112359552, + "grad_norm": 0.884218156337738, + "learning_rate": 4.176305132859438e-06, + "loss": 0.6216, + "step": 4620 + }, + { + "epoch": 1.6225421348314608, + "grad_norm": 0.8722840547561646, + "learning_rate": 4.175962109890697e-06, + "loss": 0.6316, + "step": 4621 + }, + { + "epoch": 1.6228932584269664, + "grad_norm": 0.8568814992904663, + "learning_rate": 4.175619029605485e-06, + "loss": 0.6037, + "step": 4622 + }, + { + "epoch": 1.623244382022472, + "grad_norm": 0.8812152743339539, + "learning_rate": 4.175275892015537e-06, + "loss": 0.5933, + "step": 4623 + }, + { + "epoch": 1.6235955056179776, + "grad_norm": 0.8655858039855957, + "learning_rate": 4.174932697132587e-06, + "loss": 0.6138, + "step": 4624 + }, + { + "epoch": 1.6239466292134832, + "grad_norm": 0.8626354932785034, + "learning_rate": 4.174589444968371e-06, + "loss": 0.6076, + "step": 4625 + }, + { + "epoch": 1.6242977528089888, + "grad_norm": 0.8693781495094299, + "learning_rate": 4.17424613553463e-06, + "loss": 0.6139, + "step": 4626 + }, + { + "epoch": 1.6246488764044944, + "grad_norm": 0.9022204875946045, + "learning_rate": 4.173902768843104e-06, + "loss": 0.6254, + "step": 4627 + }, + { + "epoch": 1.625, + "grad_norm": 0.8757764101028442, + "learning_rate": 4.173559344905535e-06, + "loss": 0.5988, + "step": 4628 + }, + { + "epoch": 1.6253511235955056, + "grad_norm": 0.8923527002334595, + "learning_rate": 4.173215863733669e-06, + "loss": 0.6333, + "step": 4629 + }, + { + "epoch": 1.6257022471910112, + "grad_norm": 0.8588404059410095, + "learning_rate": 4.172872325339252e-06, + "loss": 0.6117, + "step": 4630 + }, + { + "epoch": 1.6260533707865168, + "grad_norm": 0.8929343819618225, + "learning_rate": 4.1725287297340335e-06, + "loss": 0.6244, + "step": 4631 + }, + { + "epoch": 1.6264044943820224, + "grad_norm": 0.8902310132980347, + "learning_rate": 4.172185076929763e-06, + "loss": 0.6019, + "step": 4632 + }, + { + "epoch": 1.626755617977528, + "grad_norm": 0.8931180238723755, + "learning_rate": 4.171841366938193e-06, + "loss": 0.6306, + "step": 4633 + }, + { + "epoch": 1.6271067415730336, + "grad_norm": 0.8756839632987976, + "learning_rate": 4.1714975997710804e-06, + "loss": 0.6456, + "step": 4634 + }, + { + "epoch": 1.6274578651685392, + "grad_norm": 0.8530634045600891, + "learning_rate": 4.171153775440179e-06, + "loss": 0.5906, + "step": 4635 + }, + { + "epoch": 1.6278089887640448, + "grad_norm": 0.8859505653381348, + "learning_rate": 4.170809893957249e-06, + "loss": 0.6395, + "step": 4636 + }, + { + "epoch": 1.6281601123595506, + "grad_norm": 0.8910026550292969, + "learning_rate": 4.170465955334051e-06, + "loss": 0.6086, + "step": 4637 + }, + { + "epoch": 1.6285112359550562, + "grad_norm": 0.8203482031822205, + "learning_rate": 4.170121959582346e-06, + "loss": 0.5843, + "step": 4638 + }, + { + "epoch": 1.6288623595505618, + "grad_norm": 0.8972336649894714, + "learning_rate": 4.169777906713899e-06, + "loss": 0.6345, + "step": 4639 + }, + { + "epoch": 1.6292134831460674, + "grad_norm": 0.9142724275588989, + "learning_rate": 4.169433796740477e-06, + "loss": 0.6292, + "step": 4640 + }, + { + "epoch": 1.629564606741573, + "grad_norm": 0.889117419719696, + "learning_rate": 4.169089629673848e-06, + "loss": 0.6188, + "step": 4641 + }, + { + "epoch": 1.6299157303370788, + "grad_norm": 0.9001839756965637, + "learning_rate": 4.168745405525781e-06, + "loss": 0.6506, + "step": 4642 + }, + { + "epoch": 1.6302668539325844, + "grad_norm": 0.8708594441413879, + "learning_rate": 4.168401124308049e-06, + "loss": 0.5841, + "step": 4643 + }, + { + "epoch": 1.63061797752809, + "grad_norm": 0.8744320273399353, + "learning_rate": 4.1680567860324275e-06, + "loss": 0.6199, + "step": 4644 + }, + { + "epoch": 1.6309691011235956, + "grad_norm": 0.8642039895057678, + "learning_rate": 4.1677123907106905e-06, + "loss": 0.6318, + "step": 4645 + }, + { + "epoch": 1.6313202247191012, + "grad_norm": 0.875656247138977, + "learning_rate": 4.167367938354616e-06, + "loss": 0.6521, + "step": 4646 + }, + { + "epoch": 1.6316713483146068, + "grad_norm": 0.8626296520233154, + "learning_rate": 4.167023428975986e-06, + "loss": 0.6273, + "step": 4647 + }, + { + "epoch": 1.6320224719101124, + "grad_norm": 0.8608964085578918, + "learning_rate": 4.1666788625865805e-06, + "loss": 0.6305, + "step": 4648 + }, + { + "epoch": 1.632373595505618, + "grad_norm": 0.871910572052002, + "learning_rate": 4.166334239198185e-06, + "loss": 0.6386, + "step": 4649 + }, + { + "epoch": 1.6327247191011236, + "grad_norm": 0.8875142931938171, + "learning_rate": 4.165989558822584e-06, + "loss": 0.6126, + "step": 4650 + }, + { + "epoch": 1.6330758426966292, + "grad_norm": 0.8592172861099243, + "learning_rate": 4.165644821471565e-06, + "loss": 0.6037, + "step": 4651 + }, + { + "epoch": 1.6334269662921348, + "grad_norm": 0.871364951133728, + "learning_rate": 4.165300027156919e-06, + "loss": 0.6198, + "step": 4652 + }, + { + "epoch": 1.6337780898876404, + "grad_norm": 0.8956006169319153, + "learning_rate": 4.164955175890437e-06, + "loss": 0.6009, + "step": 4653 + }, + { + "epoch": 1.634129213483146, + "grad_norm": 0.8735149502754211, + "learning_rate": 4.164610267683913e-06, + "loss": 0.6032, + "step": 4654 + }, + { + "epoch": 1.6344803370786516, + "grad_norm": 0.8705981373786926, + "learning_rate": 4.164265302549142e-06, + "loss": 0.6201, + "step": 4655 + }, + { + "epoch": 1.6348314606741572, + "grad_norm": 0.8664510846138, + "learning_rate": 4.1639202804979215e-06, + "loss": 0.5911, + "step": 4656 + }, + { + "epoch": 1.6351825842696628, + "grad_norm": 0.8341373205184937, + "learning_rate": 4.163575201542052e-06, + "loss": 0.579, + "step": 4657 + }, + { + "epoch": 1.6355337078651684, + "grad_norm": 0.8607082962989807, + "learning_rate": 4.163230065693334e-06, + "loss": 0.5828, + "step": 4658 + }, + { + "epoch": 1.6358848314606742, + "grad_norm": 0.8956010341644287, + "learning_rate": 4.162884872963571e-06, + "loss": 0.6105, + "step": 4659 + }, + { + "epoch": 1.6362359550561798, + "grad_norm": 0.8953211903572083, + "learning_rate": 4.162539623364568e-06, + "loss": 0.6139, + "step": 4660 + }, + { + "epoch": 1.6365870786516854, + "grad_norm": 0.8708750605583191, + "learning_rate": 4.162194316908132e-06, + "loss": 0.6285, + "step": 4661 + }, + { + "epoch": 1.636938202247191, + "grad_norm": 0.8955643773078918, + "learning_rate": 4.161848953606074e-06, + "loss": 0.5956, + "step": 4662 + }, + { + "epoch": 1.6372893258426966, + "grad_norm": 0.8670720458030701, + "learning_rate": 4.161503533470202e-06, + "loss": 0.6217, + "step": 4663 + }, + { + "epoch": 1.6376404494382022, + "grad_norm": 0.8754477500915527, + "learning_rate": 4.1611580565123325e-06, + "loss": 0.6245, + "step": 4664 + }, + { + "epoch": 1.637991573033708, + "grad_norm": 0.8910819888114929, + "learning_rate": 4.160812522744279e-06, + "loss": 0.605, + "step": 4665 + }, + { + "epoch": 1.6383426966292136, + "grad_norm": 0.8849613666534424, + "learning_rate": 4.1604669321778576e-06, + "loss": 0.6272, + "step": 4666 + }, + { + "epoch": 1.6386938202247192, + "grad_norm": 0.8790765404701233, + "learning_rate": 4.160121284824888e-06, + "loss": 0.6461, + "step": 4667 + }, + { + "epoch": 1.6390449438202248, + "grad_norm": 0.8697189092636108, + "learning_rate": 4.1597755806971914e-06, + "loss": 0.6243, + "step": 4668 + }, + { + "epoch": 1.6393960674157304, + "grad_norm": 0.8819729089736938, + "learning_rate": 4.15942981980659e-06, + "loss": 0.6333, + "step": 4669 + }, + { + "epoch": 1.639747191011236, + "grad_norm": 0.8244314193725586, + "learning_rate": 4.15908400216491e-06, + "loss": 0.5749, + "step": 4670 + }, + { + "epoch": 1.6400983146067416, + "grad_norm": 0.9082034826278687, + "learning_rate": 4.158738127783975e-06, + "loss": 0.6297, + "step": 4671 + }, + { + "epoch": 1.6404494382022472, + "grad_norm": 0.8924310803413391, + "learning_rate": 4.1583921966756154e-06, + "loss": 0.6438, + "step": 4672 + }, + { + "epoch": 1.6408005617977528, + "grad_norm": 0.9138229489326477, + "learning_rate": 4.158046208851663e-06, + "loss": 0.6064, + "step": 4673 + }, + { + "epoch": 1.6411516853932584, + "grad_norm": 0.8768520355224609, + "learning_rate": 4.157700164323949e-06, + "loss": 0.5972, + "step": 4674 + }, + { + "epoch": 1.641502808988764, + "grad_norm": 0.8785306215286255, + "learning_rate": 4.157354063104307e-06, + "loss": 0.6111, + "step": 4675 + }, + { + "epoch": 1.6418539325842696, + "grad_norm": 0.8370828032493591, + "learning_rate": 4.157007905204575e-06, + "loss": 0.5849, + "step": 4676 + }, + { + "epoch": 1.6422050561797752, + "grad_norm": 0.887261152267456, + "learning_rate": 4.15666169063659e-06, + "loss": 0.6375, + "step": 4677 + }, + { + "epoch": 1.6425561797752808, + "grad_norm": 0.8963328003883362, + "learning_rate": 4.156315419412194e-06, + "loss": 0.6058, + "step": 4678 + }, + { + "epoch": 1.6429073033707864, + "grad_norm": 0.8939655423164368, + "learning_rate": 4.155969091543226e-06, + "loss": 0.621, + "step": 4679 + }, + { + "epoch": 1.643258426966292, + "grad_norm": 0.9163799285888672, + "learning_rate": 4.155622707041534e-06, + "loss": 0.6343, + "step": 4680 + }, + { + "epoch": 1.6436095505617978, + "grad_norm": 0.8742914199829102, + "learning_rate": 4.155276265918962e-06, + "loss": 0.6141, + "step": 4681 + }, + { + "epoch": 1.6439606741573034, + "grad_norm": 0.8842009902000427, + "learning_rate": 4.154929768187357e-06, + "loss": 0.5868, + "step": 4682 + }, + { + "epoch": 1.644311797752809, + "grad_norm": 0.8529406189918518, + "learning_rate": 4.15458321385857e-06, + "loss": 0.6573, + "step": 4683 + }, + { + "epoch": 1.6446629213483146, + "grad_norm": 0.8780986666679382, + "learning_rate": 4.154236602944454e-06, + "loss": 0.614, + "step": 4684 + }, + { + "epoch": 1.6450140449438202, + "grad_norm": 0.8548287749290466, + "learning_rate": 4.153889935456862e-06, + "loss": 0.622, + "step": 4685 + }, + { + "epoch": 1.6453651685393258, + "grad_norm": 0.8868494629859924, + "learning_rate": 4.153543211407648e-06, + "loss": 0.6204, + "step": 4686 + }, + { + "epoch": 1.6457162921348316, + "grad_norm": 0.8862740993499756, + "learning_rate": 4.153196430808672e-06, + "loss": 0.639, + "step": 4687 + }, + { + "epoch": 1.6460674157303372, + "grad_norm": 0.8767764568328857, + "learning_rate": 4.152849593671793e-06, + "loss": 0.6059, + "step": 4688 + }, + { + "epoch": 1.6464185393258428, + "grad_norm": 0.9215349555015564, + "learning_rate": 4.1525027000088715e-06, + "loss": 0.6434, + "step": 4689 + }, + { + "epoch": 1.6467696629213484, + "grad_norm": 0.868288516998291, + "learning_rate": 4.152155749831772e-06, + "loss": 0.6229, + "step": 4690 + }, + { + "epoch": 1.647120786516854, + "grad_norm": 0.851254940032959, + "learning_rate": 4.15180874315236e-06, + "loss": 0.6259, + "step": 4691 + }, + { + "epoch": 1.6474719101123596, + "grad_norm": 0.8357805609703064, + "learning_rate": 4.151461679982501e-06, + "loss": 0.5411, + "step": 4692 + }, + { + "epoch": 1.6478230337078652, + "grad_norm": 0.8943964838981628, + "learning_rate": 4.1511145603340675e-06, + "loss": 0.6564, + "step": 4693 + }, + { + "epoch": 1.6481741573033708, + "grad_norm": 0.8508518934249878, + "learning_rate": 4.150767384218929e-06, + "loss": 0.6238, + "step": 4694 + }, + { + "epoch": 1.6485252808988764, + "grad_norm": 0.9138011932373047, + "learning_rate": 4.150420151648957e-06, + "loss": 0.6538, + "step": 4695 + }, + { + "epoch": 1.648876404494382, + "grad_norm": 0.8589404821395874, + "learning_rate": 4.150072862636029e-06, + "loss": 0.6307, + "step": 4696 + }, + { + "epoch": 1.6492275280898876, + "grad_norm": 0.8829411864280701, + "learning_rate": 4.149725517192021e-06, + "loss": 0.6113, + "step": 4697 + }, + { + "epoch": 1.6495786516853932, + "grad_norm": 0.9030507206916809, + "learning_rate": 4.149378115328812e-06, + "loss": 0.6494, + "step": 4698 + }, + { + "epoch": 1.6499297752808988, + "grad_norm": 0.8570032715797424, + "learning_rate": 4.149030657058283e-06, + "loss": 0.5724, + "step": 4699 + }, + { + "epoch": 1.6502808988764044, + "grad_norm": 0.886920690536499, + "learning_rate": 4.148683142392317e-06, + "loss": 0.6313, + "step": 4700 + }, + { + "epoch": 1.65063202247191, + "grad_norm": 0.8540245890617371, + "learning_rate": 4.148335571342798e-06, + "loss": 0.624, + "step": 4701 + }, + { + "epoch": 1.6509831460674156, + "grad_norm": 0.8701122999191284, + "learning_rate": 4.1479879439216116e-06, + "loss": 0.6177, + "step": 4702 + }, + { + "epoch": 1.6513342696629212, + "grad_norm": 0.8617565035820007, + "learning_rate": 4.14764026014065e-06, + "loss": 0.5755, + "step": 4703 + }, + { + "epoch": 1.651685393258427, + "grad_norm": 0.8952060341835022, + "learning_rate": 4.1472925200118e-06, + "loss": 0.6087, + "step": 4704 + }, + { + "epoch": 1.6520365168539326, + "grad_norm": 0.9085246920585632, + "learning_rate": 4.146944723546956e-06, + "loss": 0.6407, + "step": 4705 + }, + { + "epoch": 1.6523876404494382, + "grad_norm": 0.890703022480011, + "learning_rate": 4.146596870758012e-06, + "loss": 0.6719, + "step": 4706 + }, + { + "epoch": 1.6527387640449438, + "grad_norm": 0.8996905088424683, + "learning_rate": 4.146248961656863e-06, + "loss": 0.6283, + "step": 4707 + }, + { + "epoch": 1.6530898876404494, + "grad_norm": 0.8666749000549316, + "learning_rate": 4.145900996255409e-06, + "loss": 0.6244, + "step": 4708 + }, + { + "epoch": 1.6534410112359552, + "grad_norm": 0.8974918127059937, + "learning_rate": 4.145552974565549e-06, + "loss": 0.6304, + "step": 4709 + }, + { + "epoch": 1.6537921348314608, + "grad_norm": 0.8758713603019714, + "learning_rate": 4.145204896599186e-06, + "loss": 0.6481, + "step": 4710 + }, + { + "epoch": 1.6541432584269664, + "grad_norm": 0.8611330389976501, + "learning_rate": 4.1448567623682226e-06, + "loss": 0.6096, + "step": 4711 + }, + { + "epoch": 1.654494382022472, + "grad_norm": 0.8383393883705139, + "learning_rate": 4.144508571884565e-06, + "loss": 0.5943, + "step": 4712 + }, + { + "epoch": 1.6548455056179776, + "grad_norm": 0.8979886174201965, + "learning_rate": 4.144160325160122e-06, + "loss": 0.6182, + "step": 4713 + }, + { + "epoch": 1.6551966292134832, + "grad_norm": 0.8730772733688354, + "learning_rate": 4.143812022206804e-06, + "loss": 0.6265, + "step": 4714 + }, + { + "epoch": 1.6555477528089888, + "grad_norm": 0.8903005123138428, + "learning_rate": 4.14346366303652e-06, + "loss": 0.6213, + "step": 4715 + }, + { + "epoch": 1.6558988764044944, + "grad_norm": 0.8978590369224548, + "learning_rate": 4.1431152476611856e-06, + "loss": 0.6048, + "step": 4716 + }, + { + "epoch": 1.65625, + "grad_norm": 0.8998496532440186, + "learning_rate": 4.1427667760927155e-06, + "loss": 0.6435, + "step": 4717 + }, + { + "epoch": 1.6566011235955056, + "grad_norm": 0.882354736328125, + "learning_rate": 4.142418248343028e-06, + "loss": 0.6063, + "step": 4718 + }, + { + "epoch": 1.6569522471910112, + "grad_norm": 0.8468122482299805, + "learning_rate": 4.142069664424041e-06, + "loss": 0.607, + "step": 4719 + }, + { + "epoch": 1.6573033707865168, + "grad_norm": 0.8717797994613647, + "learning_rate": 4.141721024347677e-06, + "loss": 0.6255, + "step": 4720 + }, + { + "epoch": 1.6576544943820224, + "grad_norm": 0.8650950193405151, + "learning_rate": 4.141372328125859e-06, + "loss": 0.5888, + "step": 4721 + }, + { + "epoch": 1.658005617977528, + "grad_norm": 0.8369475603103638, + "learning_rate": 4.141023575770512e-06, + "loss": 0.5846, + "step": 4722 + }, + { + "epoch": 1.6583567415730336, + "grad_norm": 0.8811677694320679, + "learning_rate": 4.140674767293562e-06, + "loss": 0.6263, + "step": 4723 + }, + { + "epoch": 1.6587078651685392, + "grad_norm": 0.8905251026153564, + "learning_rate": 4.14032590270694e-06, + "loss": 0.6204, + "step": 4724 + }, + { + "epoch": 1.6590589887640448, + "grad_norm": 0.8818519711494446, + "learning_rate": 4.139976982022576e-06, + "loss": 0.6142, + "step": 4725 + }, + { + "epoch": 1.6594101123595506, + "grad_norm": 0.8485746383666992, + "learning_rate": 4.139628005252401e-06, + "loss": 0.5798, + "step": 4726 + }, + { + "epoch": 1.6597612359550562, + "grad_norm": 0.8504104614257812, + "learning_rate": 4.139278972408353e-06, + "loss": 0.6418, + "step": 4727 + }, + { + "epoch": 1.6601123595505618, + "grad_norm": 0.8766199946403503, + "learning_rate": 4.138929883502365e-06, + "loss": 0.6014, + "step": 4728 + }, + { + "epoch": 1.6604634831460674, + "grad_norm": 0.8618320226669312, + "learning_rate": 4.1385807385463785e-06, + "loss": 0.5874, + "step": 4729 + }, + { + "epoch": 1.660814606741573, + "grad_norm": 0.8511069416999817, + "learning_rate": 4.138231537552333e-06, + "loss": 0.6173, + "step": 4730 + }, + { + "epoch": 1.6611657303370788, + "grad_norm": 0.859161913394928, + "learning_rate": 4.137882280532171e-06, + "loss": 0.6264, + "step": 4731 + }, + { + "epoch": 1.6615168539325844, + "grad_norm": 0.8392783999443054, + "learning_rate": 4.137532967497837e-06, + "loss": 0.6064, + "step": 4732 + }, + { + "epoch": 1.66186797752809, + "grad_norm": 0.8905898332595825, + "learning_rate": 4.1371835984612755e-06, + "loss": 0.6294, + "step": 4733 + }, + { + "epoch": 1.6622191011235956, + "grad_norm": 0.8814263343811035, + "learning_rate": 4.1368341734344365e-06, + "loss": 0.6053, + "step": 4734 + }, + { + "epoch": 1.6625702247191012, + "grad_norm": 0.8924182653427124, + "learning_rate": 4.13648469242927e-06, + "loss": 0.6454, + "step": 4735 + }, + { + "epoch": 1.6629213483146068, + "grad_norm": 0.8582429885864258, + "learning_rate": 4.136135155457726e-06, + "loss": 0.5989, + "step": 4736 + }, + { + "epoch": 1.6632724719101124, + "grad_norm": 0.9076274037361145, + "learning_rate": 4.135785562531761e-06, + "loss": 0.6454, + "step": 4737 + }, + { + "epoch": 1.663623595505618, + "grad_norm": 0.9228194355964661, + "learning_rate": 4.135435913663328e-06, + "loss": 0.6089, + "step": 4738 + }, + { + "epoch": 1.6639747191011236, + "grad_norm": 0.8946560025215149, + "learning_rate": 4.135086208864388e-06, + "loss": 0.6196, + "step": 4739 + }, + { + "epoch": 1.6643258426966292, + "grad_norm": 0.8732385635375977, + "learning_rate": 4.134736448146898e-06, + "loss": 0.6055, + "step": 4740 + }, + { + "epoch": 1.6646769662921348, + "grad_norm": 0.8972743153572083, + "learning_rate": 4.13438663152282e-06, + "loss": 0.6511, + "step": 4741 + }, + { + "epoch": 1.6650280898876404, + "grad_norm": 0.9056386351585388, + "learning_rate": 4.134036759004119e-06, + "loss": 0.635, + "step": 4742 + }, + { + "epoch": 1.665379213483146, + "grad_norm": 0.8671088814735413, + "learning_rate": 4.133686830602758e-06, + "loss": 0.6208, + "step": 4743 + }, + { + "epoch": 1.6657303370786516, + "grad_norm": 0.8547230958938599, + "learning_rate": 4.133336846330705e-06, + "loss": 0.6195, + "step": 4744 + }, + { + "epoch": 1.6660814606741572, + "grad_norm": 0.8662692308425903, + "learning_rate": 4.132986806199931e-06, + "loss": 0.6028, + "step": 4745 + }, + { + "epoch": 1.6664325842696628, + "grad_norm": 0.9029760956764221, + "learning_rate": 4.132636710222404e-06, + "loss": 0.6218, + "step": 4746 + }, + { + "epoch": 1.6667837078651684, + "grad_norm": 0.8908417224884033, + "learning_rate": 4.1322865584100994e-06, + "loss": 0.6275, + "step": 4747 + }, + { + "epoch": 1.6671348314606742, + "grad_norm": 0.8707289099693298, + "learning_rate": 4.131936350774991e-06, + "loss": 0.6066, + "step": 4748 + }, + { + "epoch": 1.6674859550561798, + "grad_norm": 0.8993298411369324, + "learning_rate": 4.131586087329057e-06, + "loss": 0.5929, + "step": 4749 + }, + { + "epoch": 1.6678370786516854, + "grad_norm": 0.8987124562263489, + "learning_rate": 4.1312357680842735e-06, + "loss": 0.6188, + "step": 4750 + }, + { + "epoch": 1.668188202247191, + "grad_norm": 0.8531199097633362, + "learning_rate": 4.130885393052623e-06, + "loss": 0.6091, + "step": 4751 + }, + { + "epoch": 1.6685393258426966, + "grad_norm": 0.8786628246307373, + "learning_rate": 4.130534962246088e-06, + "loss": 0.6173, + "step": 4752 + }, + { + "epoch": 1.6688904494382022, + "grad_norm": 0.8418239951133728, + "learning_rate": 4.130184475676652e-06, + "loss": 0.6025, + "step": 4753 + }, + { + "epoch": 1.669241573033708, + "grad_norm": 0.8778673410415649, + "learning_rate": 4.129833933356302e-06, + "loss": 0.6347, + "step": 4754 + }, + { + "epoch": 1.6695926966292136, + "grad_norm": 0.8813691735267639, + "learning_rate": 4.129483335297027e-06, + "loss": 0.6317, + "step": 4755 + }, + { + "epoch": 1.6699438202247192, + "grad_norm": 0.8676353096961975, + "learning_rate": 4.129132681510815e-06, + "loss": 0.6313, + "step": 4756 + }, + { + "epoch": 1.6702949438202248, + "grad_norm": 0.848374605178833, + "learning_rate": 4.128781972009661e-06, + "loss": 0.6367, + "step": 4757 + }, + { + "epoch": 1.6706460674157304, + "grad_norm": 0.875219464302063, + "learning_rate": 4.128431206805556e-06, + "loss": 0.6053, + "step": 4758 + }, + { + "epoch": 1.670997191011236, + "grad_norm": 0.8850371241569519, + "learning_rate": 4.1280803859104975e-06, + "loss": 0.6308, + "step": 4759 + }, + { + "epoch": 1.6713483146067416, + "grad_norm": 0.8848199248313904, + "learning_rate": 4.1277295093364835e-06, + "loss": 0.5875, + "step": 4760 + }, + { + "epoch": 1.6716994382022472, + "grad_norm": 0.8953977227210999, + "learning_rate": 4.1273785770955125e-06, + "loss": 0.6082, + "step": 4761 + }, + { + "epoch": 1.6720505617977528, + "grad_norm": 0.8523750305175781, + "learning_rate": 4.127027589199587e-06, + "loss": 0.6132, + "step": 4762 + }, + { + "epoch": 1.6724016853932584, + "grad_norm": 0.8619308471679688, + "learning_rate": 4.126676545660709e-06, + "loss": 0.5715, + "step": 4763 + }, + { + "epoch": 1.672752808988764, + "grad_norm": 0.9020019173622131, + "learning_rate": 4.126325446490887e-06, + "loss": 0.6054, + "step": 4764 + }, + { + "epoch": 1.6731039325842696, + "grad_norm": 0.8890277147293091, + "learning_rate": 4.125974291702126e-06, + "loss": 0.6116, + "step": 4765 + }, + { + "epoch": 1.6734550561797752, + "grad_norm": 0.9404541254043579, + "learning_rate": 4.125623081306435e-06, + "loss": 0.6256, + "step": 4766 + }, + { + "epoch": 1.6738061797752808, + "grad_norm": 0.8866811990737915, + "learning_rate": 4.125271815315826e-06, + "loss": 0.6266, + "step": 4767 + }, + { + "epoch": 1.6741573033707864, + "grad_norm": 0.8981043696403503, + "learning_rate": 4.1249204937423115e-06, + "loss": 0.6314, + "step": 4768 + }, + { + "epoch": 1.674508426966292, + "grad_norm": 0.889731228351593, + "learning_rate": 4.124569116597907e-06, + "loss": 0.5948, + "step": 4769 + }, + { + "epoch": 1.6748595505617978, + "grad_norm": 0.9299228191375732, + "learning_rate": 4.124217683894628e-06, + "loss": 0.6508, + "step": 4770 + }, + { + "epoch": 1.6752106741573034, + "grad_norm": 0.909980058670044, + "learning_rate": 4.1238661956444946e-06, + "loss": 0.6499, + "step": 4771 + }, + { + "epoch": 1.675561797752809, + "grad_norm": 0.8873563408851624, + "learning_rate": 4.123514651859527e-06, + "loss": 0.6122, + "step": 4772 + }, + { + "epoch": 1.6759129213483146, + "grad_norm": 0.8655134439468384, + "learning_rate": 4.123163052551747e-06, + "loss": 0.6395, + "step": 4773 + }, + { + "epoch": 1.6762640449438202, + "grad_norm": 0.8635685443878174, + "learning_rate": 4.122811397733181e-06, + "loss": 0.6194, + "step": 4774 + }, + { + "epoch": 1.6766151685393258, + "grad_norm": 0.8841813802719116, + "learning_rate": 4.122459687415852e-06, + "loss": 0.6524, + "step": 4775 + }, + { + "epoch": 1.6769662921348316, + "grad_norm": 0.8351168632507324, + "learning_rate": 4.122107921611791e-06, + "loss": 0.6083, + "step": 4776 + }, + { + "epoch": 1.6773174157303372, + "grad_norm": 0.8590231537818909, + "learning_rate": 4.121756100333028e-06, + "loss": 0.6243, + "step": 4777 + }, + { + "epoch": 1.6776685393258428, + "grad_norm": 0.8575931191444397, + "learning_rate": 4.121404223591592e-06, + "loss": 0.5959, + "step": 4778 + }, + { + "epoch": 1.6780196629213484, + "grad_norm": 0.8839737772941589, + "learning_rate": 4.12105229139952e-06, + "loss": 0.6432, + "step": 4779 + }, + { + "epoch": 1.678370786516854, + "grad_norm": 0.871910572052002, + "learning_rate": 4.120700303768848e-06, + "loss": 0.6134, + "step": 4780 + }, + { + "epoch": 1.6787219101123596, + "grad_norm": 0.8897039294242859, + "learning_rate": 4.1203482607116105e-06, + "loss": 0.6329, + "step": 4781 + }, + { + "epoch": 1.6790730337078652, + "grad_norm": 0.8837077617645264, + "learning_rate": 4.119996162239851e-06, + "loss": 0.6059, + "step": 4782 + }, + { + "epoch": 1.6794241573033708, + "grad_norm": 0.8870334625244141, + "learning_rate": 4.119644008365607e-06, + "loss": 0.5785, + "step": 4783 + }, + { + "epoch": 1.6797752808988764, + "grad_norm": 0.8958740234375, + "learning_rate": 4.119291799100925e-06, + "loss": 0.6076, + "step": 4784 + }, + { + "epoch": 1.680126404494382, + "grad_norm": 0.8951886892318726, + "learning_rate": 4.118939534457849e-06, + "loss": 0.6134, + "step": 4785 + }, + { + "epoch": 1.6804775280898876, + "grad_norm": 0.8920659422874451, + "learning_rate": 4.118587214448426e-06, + "loss": 0.6176, + "step": 4786 + }, + { + "epoch": 1.6808286516853932, + "grad_norm": 0.9017564058303833, + "learning_rate": 4.118234839084706e-06, + "loss": 0.623, + "step": 4787 + }, + { + "epoch": 1.6811797752808988, + "grad_norm": 0.8780067563056946, + "learning_rate": 4.117882408378739e-06, + "loss": 0.6034, + "step": 4788 + }, + { + "epoch": 1.6815308988764044, + "grad_norm": 0.8677077293395996, + "learning_rate": 4.117529922342578e-06, + "loss": 0.5848, + "step": 4789 + }, + { + "epoch": 1.68188202247191, + "grad_norm": 0.9106276035308838, + "learning_rate": 4.117177380988277e-06, + "loss": 0.634, + "step": 4790 + }, + { + "epoch": 1.6822331460674156, + "grad_norm": 0.9075629711151123, + "learning_rate": 4.116824784327894e-06, + "loss": 0.622, + "step": 4791 + }, + { + "epoch": 1.6825842696629212, + "grad_norm": 0.9017618298530579, + "learning_rate": 4.116472132373488e-06, + "loss": 0.6306, + "step": 4792 + }, + { + "epoch": 1.682935393258427, + "grad_norm": 0.846936821937561, + "learning_rate": 4.116119425137117e-06, + "loss": 0.5728, + "step": 4793 + }, + { + "epoch": 1.6832865168539326, + "grad_norm": 0.8908042311668396, + "learning_rate": 4.115766662630845e-06, + "loss": 0.6305, + "step": 4794 + }, + { + "epoch": 1.6836376404494382, + "grad_norm": 0.8723462224006653, + "learning_rate": 4.115413844866736e-06, + "loss": 0.6326, + "step": 4795 + }, + { + "epoch": 1.6839887640449438, + "grad_norm": 0.9132239818572998, + "learning_rate": 4.115060971856855e-06, + "loss": 0.6706, + "step": 4796 + }, + { + "epoch": 1.6843398876404494, + "grad_norm": 0.8534126281738281, + "learning_rate": 4.1147080436132716e-06, + "loss": 0.597, + "step": 4797 + }, + { + "epoch": 1.6846910112359552, + "grad_norm": 0.8956817388534546, + "learning_rate": 4.1143550601480546e-06, + "loss": 0.6305, + "step": 4798 + }, + { + "epoch": 1.6850421348314608, + "grad_norm": 0.8739310503005981, + "learning_rate": 4.114002021473276e-06, + "loss": 0.6274, + "step": 4799 + }, + { + "epoch": 1.6853932584269664, + "grad_norm": 0.8571768403053284, + "learning_rate": 4.113648927601008e-06, + "loss": 0.6144, + "step": 4800 + }, + { + "epoch": 1.685744382022472, + "grad_norm": 0.9097314476966858, + "learning_rate": 4.113295778543329e-06, + "loss": 0.5915, + "step": 4801 + }, + { + "epoch": 1.6860955056179776, + "grad_norm": 0.8725996613502502, + "learning_rate": 4.112942574312315e-06, + "loss": 0.6315, + "step": 4802 + }, + { + "epoch": 1.6864466292134832, + "grad_norm": 0.8608437776565552, + "learning_rate": 4.112589314920045e-06, + "loss": 0.6396, + "step": 4803 + }, + { + "epoch": 1.6867977528089888, + "grad_norm": 0.8394370675086975, + "learning_rate": 4.1122360003785995e-06, + "loss": 0.5763, + "step": 4804 + }, + { + "epoch": 1.6871488764044944, + "grad_norm": 0.8686445951461792, + "learning_rate": 4.111882630700064e-06, + "loss": 0.6144, + "step": 4805 + }, + { + "epoch": 1.6875, + "grad_norm": 0.8637919425964355, + "learning_rate": 4.11152920589652e-06, + "loss": 0.5964, + "step": 4806 + }, + { + "epoch": 1.6878511235955056, + "grad_norm": 0.8992030620574951, + "learning_rate": 4.111175725980057e-06, + "loss": 0.6218, + "step": 4807 + }, + { + "epoch": 1.6882022471910112, + "grad_norm": 0.8575127124786377, + "learning_rate": 4.110822190962764e-06, + "loss": 0.6043, + "step": 4808 + }, + { + "epoch": 1.6885533707865168, + "grad_norm": 0.8570694327354431, + "learning_rate": 4.110468600856729e-06, + "loss": 0.6018, + "step": 4809 + }, + { + "epoch": 1.6889044943820224, + "grad_norm": 0.8761375546455383, + "learning_rate": 4.110114955674046e-06, + "loss": 0.6141, + "step": 4810 + }, + { + "epoch": 1.689255617977528, + "grad_norm": 0.8680332899093628, + "learning_rate": 4.10976125542681e-06, + "loss": 0.5964, + "step": 4811 + }, + { + "epoch": 1.6896067415730336, + "grad_norm": 0.8575436472892761, + "learning_rate": 4.1094075001271164e-06, + "loss": 0.6062, + "step": 4812 + }, + { + "epoch": 1.6899578651685392, + "grad_norm": 0.9025743007659912, + "learning_rate": 4.109053689787063e-06, + "loss": 0.6434, + "step": 4813 + }, + { + "epoch": 1.6903089887640448, + "grad_norm": 0.9137904047966003, + "learning_rate": 4.108699824418751e-06, + "loss": 0.6291, + "step": 4814 + }, + { + "epoch": 1.6906601123595506, + "grad_norm": 0.8753927946090698, + "learning_rate": 4.108345904034281e-06, + "loss": 0.6151, + "step": 4815 + }, + { + "epoch": 1.6910112359550562, + "grad_norm": 0.8979054689407349, + "learning_rate": 4.107991928645757e-06, + "loss": 0.6381, + "step": 4816 + }, + { + "epoch": 1.6913623595505618, + "grad_norm": 0.8909083008766174, + "learning_rate": 4.107637898265286e-06, + "loss": 0.6121, + "step": 4817 + }, + { + "epoch": 1.6917134831460674, + "grad_norm": 0.8720690011978149, + "learning_rate": 4.107283812904974e-06, + "loss": 0.6342, + "step": 4818 + }, + { + "epoch": 1.692064606741573, + "grad_norm": 0.9081621766090393, + "learning_rate": 4.106929672576932e-06, + "loss": 0.6058, + "step": 4819 + }, + { + "epoch": 1.6924157303370788, + "grad_norm": 0.8846354484558105, + "learning_rate": 4.106575477293269e-06, + "loss": 0.6247, + "step": 4820 + }, + { + "epoch": 1.6927668539325844, + "grad_norm": 0.8892889022827148, + "learning_rate": 4.1062212270661e-06, + "loss": 0.6229, + "step": 4821 + }, + { + "epoch": 1.69311797752809, + "grad_norm": 0.8698278069496155, + "learning_rate": 4.105866921907539e-06, + "loss": 0.6147, + "step": 4822 + }, + { + "epoch": 1.6934691011235956, + "grad_norm": 0.8612379431724548, + "learning_rate": 4.1055125618297036e-06, + "loss": 0.6106, + "step": 4823 + }, + { + "epoch": 1.6938202247191012, + "grad_norm": 0.8882355690002441, + "learning_rate": 4.105158146844713e-06, + "loss": 0.588, + "step": 4824 + }, + { + "epoch": 1.6941713483146068, + "grad_norm": 0.9116877317428589, + "learning_rate": 4.104803676964686e-06, + "loss": 0.663, + "step": 4825 + }, + { + "epoch": 1.6945224719101124, + "grad_norm": 0.8949796557426453, + "learning_rate": 4.1044491522017485e-06, + "loss": 0.6081, + "step": 4826 + }, + { + "epoch": 1.694873595505618, + "grad_norm": 0.8727705478668213, + "learning_rate": 4.104094572568022e-06, + "loss": 0.5666, + "step": 4827 + }, + { + "epoch": 1.6952247191011236, + "grad_norm": 0.8831336498260498, + "learning_rate": 4.1037399380756324e-06, + "loss": 0.627, + "step": 4828 + }, + { + "epoch": 1.6955758426966292, + "grad_norm": 0.8737815618515015, + "learning_rate": 4.103385248736711e-06, + "loss": 0.6177, + "step": 4829 + }, + { + "epoch": 1.6959269662921348, + "grad_norm": 0.8915756344795227, + "learning_rate": 4.103030504563385e-06, + "loss": 0.6209, + "step": 4830 + }, + { + "epoch": 1.6962780898876404, + "grad_norm": 0.9170124530792236, + "learning_rate": 4.102675705567788e-06, + "loss": 0.6134, + "step": 4831 + }, + { + "epoch": 1.696629213483146, + "grad_norm": 0.9044004678726196, + "learning_rate": 4.1023208517620535e-06, + "loss": 0.6432, + "step": 4832 + }, + { + "epoch": 1.6969803370786516, + "grad_norm": 0.9003125429153442, + "learning_rate": 4.101965943158318e-06, + "loss": 0.6029, + "step": 4833 + }, + { + "epoch": 1.6973314606741572, + "grad_norm": 0.9001431465148926, + "learning_rate": 4.101610979768716e-06, + "loss": 0.6043, + "step": 4834 + }, + { + "epoch": 1.6976825842696628, + "grad_norm": 0.8822726607322693, + "learning_rate": 4.10125596160539e-06, + "loss": 0.6257, + "step": 4835 + }, + { + "epoch": 1.6980337078651684, + "grad_norm": 0.8968911170959473, + "learning_rate": 4.10090088868048e-06, + "loss": 0.6129, + "step": 4836 + }, + { + "epoch": 1.6983848314606742, + "grad_norm": 0.9139944314956665, + "learning_rate": 4.10054576100613e-06, + "loss": 0.6493, + "step": 4837 + }, + { + "epoch": 1.6987359550561798, + "grad_norm": 0.8980047106742859, + "learning_rate": 4.1001905785944855e-06, + "loss": 0.615, + "step": 4838 + }, + { + "epoch": 1.6990870786516854, + "grad_norm": 0.9069281816482544, + "learning_rate": 4.099835341457691e-06, + "loss": 0.5864, + "step": 4839 + }, + { + "epoch": 1.699438202247191, + "grad_norm": 1.545245885848999, + "learning_rate": 4.099480049607897e-06, + "loss": 0.6083, + "step": 4840 + }, + { + "epoch": 1.6997893258426966, + "grad_norm": 0.8865572214126587, + "learning_rate": 4.099124703057254e-06, + "loss": 0.6515, + "step": 4841 + }, + { + "epoch": 1.7001404494382022, + "grad_norm": 0.84342360496521, + "learning_rate": 4.0987693018179145e-06, + "loss": 0.6363, + "step": 4842 + }, + { + "epoch": 1.700491573033708, + "grad_norm": 0.8635156750679016, + "learning_rate": 4.098413845902033e-06, + "loss": 0.5762, + "step": 4843 + }, + { + "epoch": 1.7008426966292136, + "grad_norm": 0.8980802297592163, + "learning_rate": 4.098058335321767e-06, + "loss": 0.6078, + "step": 4844 + }, + { + "epoch": 1.7011938202247192, + "grad_norm": 0.888487696647644, + "learning_rate": 4.0977027700892725e-06, + "loss": 0.6108, + "step": 4845 + }, + { + "epoch": 1.7015449438202248, + "grad_norm": 0.8712450265884399, + "learning_rate": 4.09734715021671e-06, + "loss": 0.6579, + "step": 4846 + }, + { + "epoch": 1.7018960674157304, + "grad_norm": 0.8814717531204224, + "learning_rate": 4.096991475716242e-06, + "loss": 0.64, + "step": 4847 + }, + { + "epoch": 1.702247191011236, + "grad_norm": 0.8596243262290955, + "learning_rate": 4.096635746600032e-06, + "loss": 0.594, + "step": 4848 + }, + { + "epoch": 1.7025983146067416, + "grad_norm": 0.8710845708847046, + "learning_rate": 4.096279962880246e-06, + "loss": 0.6383, + "step": 4849 + }, + { + "epoch": 1.7029494382022472, + "grad_norm": 0.8820031881332397, + "learning_rate": 4.095924124569052e-06, + "loss": 0.6421, + "step": 4850 + }, + { + "epoch": 1.7033005617977528, + "grad_norm": 0.9059613347053528, + "learning_rate": 4.095568231678617e-06, + "loss": 0.6502, + "step": 4851 + }, + { + "epoch": 1.7036516853932584, + "grad_norm": 0.9789934158325195, + "learning_rate": 4.095212284221115e-06, + "loss": 0.6336, + "step": 4852 + }, + { + "epoch": 1.704002808988764, + "grad_norm": 0.8713051080703735, + "learning_rate": 4.094856282208717e-06, + "loss": 0.6108, + "step": 4853 + }, + { + "epoch": 1.7043539325842696, + "grad_norm": 0.9158304929733276, + "learning_rate": 4.0945002256536e-06, + "loss": 0.6391, + "step": 4854 + }, + { + "epoch": 1.7047050561797752, + "grad_norm": 0.8795137405395508, + "learning_rate": 4.094144114567939e-06, + "loss": 0.631, + "step": 4855 + }, + { + "epoch": 1.7050561797752808, + "grad_norm": 0.9169765710830688, + "learning_rate": 4.093787948963913e-06, + "loss": 0.6475, + "step": 4856 + }, + { + "epoch": 1.7054073033707864, + "grad_norm": 0.8848928213119507, + "learning_rate": 4.093431728853704e-06, + "loss": 0.6164, + "step": 4857 + }, + { + "epoch": 1.705758426966292, + "grad_norm": 0.8515163660049438, + "learning_rate": 4.0930754542494935e-06, + "loss": 0.6388, + "step": 4858 + }, + { + "epoch": 1.7061095505617978, + "grad_norm": 0.8862313628196716, + "learning_rate": 4.092719125163464e-06, + "loss": 0.6385, + "step": 4859 + }, + { + "epoch": 1.7064606741573034, + "grad_norm": 0.8719597458839417, + "learning_rate": 4.092362741607805e-06, + "loss": 0.5853, + "step": 4860 + }, + { + "epoch": 1.706811797752809, + "grad_norm": 0.8930037617683411, + "learning_rate": 4.092006303594702e-06, + "loss": 0.6277, + "step": 4861 + }, + { + "epoch": 1.7071629213483146, + "grad_norm": 0.8765163421630859, + "learning_rate": 4.091649811136346e-06, + "loss": 0.6454, + "step": 4862 + }, + { + "epoch": 1.7075140449438202, + "grad_norm": 0.8377728462219238, + "learning_rate": 4.091293264244929e-06, + "loss": 0.5957, + "step": 4863 + }, + { + "epoch": 1.7078651685393258, + "grad_norm": 0.8431267738342285, + "learning_rate": 4.090936662932643e-06, + "loss": 0.586, + "step": 4864 + }, + { + "epoch": 1.7082162921348316, + "grad_norm": 0.923705518245697, + "learning_rate": 4.090580007211686e-06, + "loss": 0.6231, + "step": 4865 + }, + { + "epoch": 1.7085674157303372, + "grad_norm": 0.8978968262672424, + "learning_rate": 4.090223297094251e-06, + "loss": 0.6113, + "step": 4866 + }, + { + "epoch": 1.7089185393258428, + "grad_norm": 0.8877233266830444, + "learning_rate": 4.0898665325925425e-06, + "loss": 0.6058, + "step": 4867 + }, + { + "epoch": 1.7092696629213484, + "grad_norm": 0.8544127941131592, + "learning_rate": 4.089509713718759e-06, + "loss": 0.596, + "step": 4868 + }, + { + "epoch": 1.709620786516854, + "grad_norm": 0.885586678981781, + "learning_rate": 4.089152840485102e-06, + "loss": 0.603, + "step": 4869 + }, + { + "epoch": 1.7099719101123596, + "grad_norm": 0.8819291591644287, + "learning_rate": 4.088795912903779e-06, + "loss": 0.6087, + "step": 4870 + }, + { + "epoch": 1.7103230337078652, + "grad_norm": 0.8448066115379333, + "learning_rate": 4.088438930986995e-06, + "loss": 0.6105, + "step": 4871 + }, + { + "epoch": 1.7106741573033708, + "grad_norm": 0.8704783320426941, + "learning_rate": 4.088081894746959e-06, + "loss": 0.664, + "step": 4872 + }, + { + "epoch": 1.7110252808988764, + "grad_norm": 0.9045015573501587, + "learning_rate": 4.08772480419588e-06, + "loss": 0.6106, + "step": 4873 + }, + { + "epoch": 1.711376404494382, + "grad_norm": 0.8924446105957031, + "learning_rate": 4.087367659345973e-06, + "loss": 0.6542, + "step": 4874 + }, + { + "epoch": 1.7117275280898876, + "grad_norm": 0.8980352878570557, + "learning_rate": 4.0870104602094495e-06, + "loss": 0.6459, + "step": 4875 + }, + { + "epoch": 1.7120786516853932, + "grad_norm": 0.8418675661087036, + "learning_rate": 4.086653206798526e-06, + "loss": 0.6147, + "step": 4876 + }, + { + "epoch": 1.7124297752808988, + "grad_norm": 0.8605186343193054, + "learning_rate": 4.086295899125421e-06, + "loss": 0.6553, + "step": 4877 + }, + { + "epoch": 1.7127808988764044, + "grad_norm": 0.8851338624954224, + "learning_rate": 4.085938537202353e-06, + "loss": 0.6092, + "step": 4878 + }, + { + "epoch": 1.71313202247191, + "grad_norm": 0.8552588820457458, + "learning_rate": 4.085581121041545e-06, + "loss": 0.6042, + "step": 4879 + }, + { + "epoch": 1.7134831460674156, + "grad_norm": 0.8800450563430786, + "learning_rate": 4.085223650655219e-06, + "loss": 0.6298, + "step": 4880 + }, + { + "epoch": 1.7138342696629212, + "grad_norm": 0.8970910310745239, + "learning_rate": 4.084866126055601e-06, + "loss": 0.6378, + "step": 4881 + }, + { + "epoch": 1.714185393258427, + "grad_norm": 0.8541218638420105, + "learning_rate": 4.084508547254917e-06, + "loss": 0.5781, + "step": 4882 + }, + { + "epoch": 1.7145365168539326, + "grad_norm": 0.87952721118927, + "learning_rate": 4.084150914265398e-06, + "loss": 0.6068, + "step": 4883 + }, + { + "epoch": 1.7148876404494382, + "grad_norm": 0.9047470092773438, + "learning_rate": 4.083793227099272e-06, + "loss": 0.6484, + "step": 4884 + }, + { + "epoch": 1.7152387640449438, + "grad_norm": 0.8786801099777222, + "learning_rate": 4.0834354857687735e-06, + "loss": 0.6395, + "step": 4885 + }, + { + "epoch": 1.7155898876404494, + "grad_norm": 0.821982741355896, + "learning_rate": 4.0830776902861354e-06, + "loss": 0.5653, + "step": 4886 + }, + { + "epoch": 1.7159410112359552, + "grad_norm": 0.8869472742080688, + "learning_rate": 4.082719840663596e-06, + "loss": 0.6045, + "step": 4887 + }, + { + "epoch": 1.7162921348314608, + "grad_norm": 0.8803991079330444, + "learning_rate": 4.082361936913393e-06, + "loss": 0.5996, + "step": 4888 + }, + { + "epoch": 1.7166432584269664, + "grad_norm": 0.8931458592414856, + "learning_rate": 4.0820039790477635e-06, + "loss": 0.6376, + "step": 4889 + }, + { + "epoch": 1.716994382022472, + "grad_norm": 0.8802606463432312, + "learning_rate": 4.081645967078953e-06, + "loss": 0.6036, + "step": 4890 + }, + { + "epoch": 1.7173455056179776, + "grad_norm": 0.8540336489677429, + "learning_rate": 4.081287901019204e-06, + "loss": 0.6039, + "step": 4891 + }, + { + "epoch": 1.7176966292134832, + "grad_norm": 0.8872647285461426, + "learning_rate": 4.080929780880762e-06, + "loss": 0.6198, + "step": 4892 + }, + { + "epoch": 1.7180477528089888, + "grad_norm": 0.858366072177887, + "learning_rate": 4.080571606675875e-06, + "loss": 0.5973, + "step": 4893 + }, + { + "epoch": 1.7183988764044944, + "grad_norm": 0.8896951675415039, + "learning_rate": 4.08021337841679e-06, + "loss": 0.6307, + "step": 4894 + }, + { + "epoch": 1.71875, + "grad_norm": 0.8642908930778503, + "learning_rate": 4.07985509611576e-06, + "loss": 0.6158, + "step": 4895 + }, + { + "epoch": 1.7191011235955056, + "grad_norm": 0.9126440286636353, + "learning_rate": 4.079496759785039e-06, + "loss": 0.6595, + "step": 4896 + }, + { + "epoch": 1.7194522471910112, + "grad_norm": 0.882386326789856, + "learning_rate": 4.0791383694368795e-06, + "loss": 0.5986, + "step": 4897 + }, + { + "epoch": 1.7198033707865168, + "grad_norm": 0.846931517124176, + "learning_rate": 4.078779925083538e-06, + "loss": 0.6443, + "step": 4898 + }, + { + "epoch": 1.7201544943820224, + "grad_norm": 0.8972097635269165, + "learning_rate": 4.0784214267372755e-06, + "loss": 0.5873, + "step": 4899 + }, + { + "epoch": 1.720505617977528, + "grad_norm": 0.8785168528556824, + "learning_rate": 4.07806287441035e-06, + "loss": 0.6073, + "step": 4900 + }, + { + "epoch": 1.7208567415730336, + "grad_norm": 0.8814970254898071, + "learning_rate": 4.077704268115025e-06, + "loss": 0.6517, + "step": 4901 + }, + { + "epoch": 1.7212078651685392, + "grad_norm": 0.8843238353729248, + "learning_rate": 4.0773456078635635e-06, + "loss": 0.6437, + "step": 4902 + }, + { + "epoch": 1.7215589887640448, + "grad_norm": 0.8787217736244202, + "learning_rate": 4.076986893668233e-06, + "loss": 0.667, + "step": 4903 + }, + { + "epoch": 1.7219101123595506, + "grad_norm": 0.860560953617096, + "learning_rate": 4.076628125541299e-06, + "loss": 0.5985, + "step": 4904 + }, + { + "epoch": 1.7222612359550562, + "grad_norm": 0.8581588268280029, + "learning_rate": 4.076269303495033e-06, + "loss": 0.5936, + "step": 4905 + }, + { + "epoch": 1.7226123595505618, + "grad_norm": 0.8703935742378235, + "learning_rate": 4.0759104275417056e-06, + "loss": 0.6099, + "step": 4906 + }, + { + "epoch": 1.7229634831460674, + "grad_norm": 0.8954237103462219, + "learning_rate": 4.07555149769359e-06, + "loss": 0.5983, + "step": 4907 + }, + { + "epoch": 1.723314606741573, + "grad_norm": 0.8742859363555908, + "learning_rate": 4.075192513962962e-06, + "loss": 0.6193, + "step": 4908 + }, + { + "epoch": 1.7236657303370788, + "grad_norm": 0.865284264087677, + "learning_rate": 4.0748334763620975e-06, + "loss": 0.6157, + "step": 4909 + }, + { + "epoch": 1.7240168539325844, + "grad_norm": 0.8605691194534302, + "learning_rate": 4.074474384903275e-06, + "loss": 0.622, + "step": 4910 + }, + { + "epoch": 1.72436797752809, + "grad_norm": 0.8634774088859558, + "learning_rate": 4.0741152395987765e-06, + "loss": 0.595, + "step": 4911 + }, + { + "epoch": 1.7247191011235956, + "grad_norm": 0.8784124851226807, + "learning_rate": 4.073756040460884e-06, + "loss": 0.6459, + "step": 4912 + }, + { + "epoch": 1.7250702247191012, + "grad_norm": 0.8804491758346558, + "learning_rate": 4.073396787501881e-06, + "loss": 0.6502, + "step": 4913 + }, + { + "epoch": 1.7254213483146068, + "grad_norm": 0.8847582340240479, + "learning_rate": 4.073037480734054e-06, + "loss": 0.6206, + "step": 4914 + }, + { + "epoch": 1.7257724719101124, + "grad_norm": 0.8761957287788391, + "learning_rate": 4.072678120169692e-06, + "loss": 0.6005, + "step": 4915 + }, + { + "epoch": 1.726123595505618, + "grad_norm": 0.8767578601837158, + "learning_rate": 4.0723187058210845e-06, + "loss": 0.6233, + "step": 4916 + }, + { + "epoch": 1.7264747191011236, + "grad_norm": 0.8684590458869934, + "learning_rate": 4.071959237700522e-06, + "loss": 0.6048, + "step": 4917 + }, + { + "epoch": 1.7268258426966292, + "grad_norm": 0.8713444471359253, + "learning_rate": 4.071599715820299e-06, + "loss": 0.6331, + "step": 4918 + }, + { + "epoch": 1.7271769662921348, + "grad_norm": 0.8798062205314636, + "learning_rate": 4.07124014019271e-06, + "loss": 0.5935, + "step": 4919 + }, + { + "epoch": 1.7275280898876404, + "grad_norm": 0.892004668712616, + "learning_rate": 4.070880510830053e-06, + "loss": 0.6378, + "step": 4920 + }, + { + "epoch": 1.727879213483146, + "grad_norm": 0.8701374530792236, + "learning_rate": 4.070520827744627e-06, + "loss": 0.6012, + "step": 4921 + }, + { + "epoch": 1.7282303370786516, + "grad_norm": 0.8508331179618835, + "learning_rate": 4.0701610909487325e-06, + "loss": 0.6149, + "step": 4922 + }, + { + "epoch": 1.7285814606741572, + "grad_norm": 0.8751832842826843, + "learning_rate": 4.0698013004546725e-06, + "loss": 0.6548, + "step": 4923 + }, + { + "epoch": 1.7289325842696628, + "grad_norm": 0.8840449452400208, + "learning_rate": 4.06944145627475e-06, + "loss": 0.6347, + "step": 4924 + }, + { + "epoch": 1.7292837078651684, + "grad_norm": 0.8908770084381104, + "learning_rate": 4.069081558421274e-06, + "loss": 0.6504, + "step": 4925 + }, + { + "epoch": 1.7296348314606742, + "grad_norm": 0.9205425977706909, + "learning_rate": 4.068721606906552e-06, + "loss": 0.673, + "step": 4926 + }, + { + "epoch": 1.7299859550561798, + "grad_norm": 0.8873515725135803, + "learning_rate": 4.068361601742892e-06, + "loss": 0.6288, + "step": 4927 + }, + { + "epoch": 1.7303370786516854, + "grad_norm": 0.8750460743904114, + "learning_rate": 4.068001542942608e-06, + "loss": 0.6161, + "step": 4928 + }, + { + "epoch": 1.730688202247191, + "grad_norm": 0.8630272746086121, + "learning_rate": 4.067641430518013e-06, + "loss": 0.603, + "step": 4929 + }, + { + "epoch": 1.7310393258426966, + "grad_norm": 0.8792911171913147, + "learning_rate": 4.067281264481422e-06, + "loss": 0.6615, + "step": 4930 + }, + { + "epoch": 1.7313904494382022, + "grad_norm": 0.8867348432540894, + "learning_rate": 4.066921044845153e-06, + "loss": 0.623, + "step": 4931 + }, + { + "epoch": 1.731741573033708, + "grad_norm": 0.8691920042037964, + "learning_rate": 4.066560771621525e-06, + "loss": 0.6306, + "step": 4932 + }, + { + "epoch": 1.7320926966292136, + "grad_norm": 0.8616251349449158, + "learning_rate": 4.06620044482286e-06, + "loss": 0.594, + "step": 4933 + }, + { + "epoch": 1.7324438202247192, + "grad_norm": 0.8571128845214844, + "learning_rate": 4.065840064461479e-06, + "loss": 0.5968, + "step": 4934 + }, + { + "epoch": 1.7327949438202248, + "grad_norm": 0.8640409111976624, + "learning_rate": 4.065479630549709e-06, + "loss": 0.6371, + "step": 4935 + }, + { + "epoch": 1.7331460674157304, + "grad_norm": 0.8730684518814087, + "learning_rate": 4.065119143099874e-06, + "loss": 0.5974, + "step": 4936 + }, + { + "epoch": 1.733497191011236, + "grad_norm": 0.9529215097427368, + "learning_rate": 4.064758602124304e-06, + "loss": 0.6094, + "step": 4937 + }, + { + "epoch": 1.7338483146067416, + "grad_norm": 0.8958554863929749, + "learning_rate": 4.0643980076353286e-06, + "loss": 0.6162, + "step": 4938 + }, + { + "epoch": 1.7341994382022472, + "grad_norm": 0.9163009524345398, + "learning_rate": 4.06403735964528e-06, + "loss": 0.6386, + "step": 4939 + }, + { + "epoch": 1.7345505617977528, + "grad_norm": 0.9188485145568848, + "learning_rate": 4.063676658166492e-06, + "loss": 0.6402, + "step": 4940 + }, + { + "epoch": 1.7349016853932584, + "grad_norm": 0.8810327053070068, + "learning_rate": 4.063315903211301e-06, + "loss": 0.5978, + "step": 4941 + }, + { + "epoch": 1.735252808988764, + "grad_norm": 0.8675046563148499, + "learning_rate": 4.062955094792043e-06, + "loss": 0.602, + "step": 4942 + }, + { + "epoch": 1.7356039325842696, + "grad_norm": 0.8730883002281189, + "learning_rate": 4.062594232921059e-06, + "loss": 0.6152, + "step": 4943 + }, + { + "epoch": 1.7359550561797752, + "grad_norm": 0.8923463225364685, + "learning_rate": 4.062233317610689e-06, + "loss": 0.5889, + "step": 4944 + }, + { + "epoch": 1.7363061797752808, + "grad_norm": 0.8387205004692078, + "learning_rate": 4.061872348873277e-06, + "loss": 0.5848, + "step": 4945 + }, + { + "epoch": 1.7366573033707864, + "grad_norm": 0.8802045583724976, + "learning_rate": 4.061511326721166e-06, + "loss": 0.5933, + "step": 4946 + }, + { + "epoch": 1.737008426966292, + "grad_norm": 0.8678444623947144, + "learning_rate": 4.061150251166705e-06, + "loss": 0.6301, + "step": 4947 + }, + { + "epoch": 1.7373595505617978, + "grad_norm": 0.8504912853240967, + "learning_rate": 4.0607891222222406e-06, + "loss": 0.6174, + "step": 4948 + }, + { + "epoch": 1.7377106741573034, + "grad_norm": 0.8655530214309692, + "learning_rate": 4.060427939900123e-06, + "loss": 0.5946, + "step": 4949 + }, + { + "epoch": 1.738061797752809, + "grad_norm": 0.8756310939788818, + "learning_rate": 4.060066704212706e-06, + "loss": 0.6013, + "step": 4950 + }, + { + "epoch": 1.7384129213483146, + "grad_norm": 0.8152898550033569, + "learning_rate": 4.059705415172343e-06, + "loss": 0.5471, + "step": 4951 + }, + { + "epoch": 1.7387640449438202, + "grad_norm": 0.9165331721305847, + "learning_rate": 4.059344072791389e-06, + "loss": 0.6438, + "step": 4952 + }, + { + "epoch": 1.7391151685393258, + "grad_norm": 0.8528552055358887, + "learning_rate": 4.058982677082203e-06, + "loss": 0.5829, + "step": 4953 + }, + { + "epoch": 1.7394662921348316, + "grad_norm": 0.8451977968215942, + "learning_rate": 4.0586212280571424e-06, + "loss": 0.5632, + "step": 4954 + }, + { + "epoch": 1.7398174157303372, + "grad_norm": 0.8505921363830566, + "learning_rate": 4.0582597257285695e-06, + "loss": 0.6139, + "step": 4955 + }, + { + "epoch": 1.7401685393258428, + "grad_norm": 0.9377829432487488, + "learning_rate": 4.0578981701088485e-06, + "loss": 0.6527, + "step": 4956 + }, + { + "epoch": 1.7405196629213484, + "grad_norm": 0.914893388748169, + "learning_rate": 4.0575365612103425e-06, + "loss": 0.6147, + "step": 4957 + }, + { + "epoch": 1.740870786516854, + "grad_norm": 0.8996618986129761, + "learning_rate": 4.057174899045418e-06, + "loss": 0.6074, + "step": 4958 + }, + { + "epoch": 1.7412219101123596, + "grad_norm": 0.8936229348182678, + "learning_rate": 4.056813183626446e-06, + "loss": 0.5993, + "step": 4959 + }, + { + "epoch": 1.7415730337078652, + "grad_norm": 0.899512767791748, + "learning_rate": 4.056451414965795e-06, + "loss": 0.6188, + "step": 4960 + }, + { + "epoch": 1.7419241573033708, + "grad_norm": 0.8578722476959229, + "learning_rate": 4.056089593075837e-06, + "loss": 0.5732, + "step": 4961 + }, + { + "epoch": 1.7422752808988764, + "grad_norm": 0.8505608439445496, + "learning_rate": 4.055727717968946e-06, + "loss": 0.5933, + "step": 4962 + }, + { + "epoch": 1.742626404494382, + "grad_norm": 0.8775436878204346, + "learning_rate": 4.0553657896575e-06, + "loss": 0.604, + "step": 4963 + }, + { + "epoch": 1.7429775280898876, + "grad_norm": 0.8631317019462585, + "learning_rate": 4.055003808153873e-06, + "loss": 0.6071, + "step": 4964 + }, + { + "epoch": 1.7433286516853932, + "grad_norm": 0.8578945398330688, + "learning_rate": 4.054641773470447e-06, + "loss": 0.6116, + "step": 4965 + }, + { + "epoch": 1.7436797752808988, + "grad_norm": 0.8524402976036072, + "learning_rate": 4.054279685619603e-06, + "loss": 0.6225, + "step": 4966 + }, + { + "epoch": 1.7440308988764044, + "grad_norm": 0.8535289168357849, + "learning_rate": 4.053917544613723e-06, + "loss": 0.6195, + "step": 4967 + }, + { + "epoch": 1.74438202247191, + "grad_norm": 0.8799499869346619, + "learning_rate": 4.0535553504651935e-06, + "loss": 0.6221, + "step": 4968 + }, + { + "epoch": 1.7447331460674156, + "grad_norm": 0.8846419453620911, + "learning_rate": 4.053193103186401e-06, + "loss": 0.5955, + "step": 4969 + }, + { + "epoch": 1.7450842696629212, + "grad_norm": 0.8822450637817383, + "learning_rate": 4.052830802789732e-06, + "loss": 0.6114, + "step": 4970 + }, + { + "epoch": 1.745435393258427, + "grad_norm": 0.8850097060203552, + "learning_rate": 4.0524684492875795e-06, + "loss": 0.6098, + "step": 4971 + }, + { + "epoch": 1.7457865168539326, + "grad_norm": 0.9146633744239807, + "learning_rate": 4.0521060426923345e-06, + "loss": 0.6262, + "step": 4972 + }, + { + "epoch": 1.7461376404494382, + "grad_norm": 0.8852578997612, + "learning_rate": 4.05174358301639e-06, + "loss": 0.6145, + "step": 4973 + }, + { + "epoch": 1.7464887640449438, + "grad_norm": 0.9027591943740845, + "learning_rate": 4.051381070272143e-06, + "loss": 0.6252, + "step": 4974 + }, + { + "epoch": 1.7468398876404494, + "grad_norm": 0.8800673484802246, + "learning_rate": 4.051018504471992e-06, + "loss": 0.616, + "step": 4975 + }, + { + "epoch": 1.7471910112359552, + "grad_norm": 0.8417078256607056, + "learning_rate": 4.050655885628335e-06, + "loss": 0.5869, + "step": 4976 + }, + { + "epoch": 1.7475421348314608, + "grad_norm": 0.8745079040527344, + "learning_rate": 4.050293213753573e-06, + "loss": 0.6215, + "step": 4977 + }, + { + "epoch": 1.7478932584269664, + "grad_norm": 0.8856931924819946, + "learning_rate": 4.04993048886011e-06, + "loss": 0.6311, + "step": 4978 + }, + { + "epoch": 1.748244382022472, + "grad_norm": 0.868653416633606, + "learning_rate": 4.04956771096035e-06, + "loss": 0.6088, + "step": 4979 + }, + { + "epoch": 1.7485955056179776, + "grad_norm": 0.9026139974594116, + "learning_rate": 4.049204880066701e-06, + "loss": 0.6032, + "step": 4980 + }, + { + "epoch": 1.7489466292134832, + "grad_norm": 1.1023459434509277, + "learning_rate": 4.0488419961915704e-06, + "loss": 0.6069, + "step": 4981 + }, + { + "epoch": 1.7492977528089888, + "grad_norm": 0.9017621278762817, + "learning_rate": 4.048479059347369e-06, + "loss": 0.652, + "step": 4982 + }, + { + "epoch": 1.7496488764044944, + "grad_norm": 0.8769750595092773, + "learning_rate": 4.048116069546509e-06, + "loss": 0.633, + "step": 4983 + }, + { + "epoch": 1.75, + "grad_norm": 0.8825761675834656, + "learning_rate": 4.047753026801404e-06, + "loss": 0.63, + "step": 4984 + }, + { + "epoch": 1.7503511235955056, + "grad_norm": 0.8574188947677612, + "learning_rate": 4.04738993112447e-06, + "loss": 0.6121, + "step": 4985 + }, + { + "epoch": 1.7507022471910112, + "grad_norm": 0.871944785118103, + "learning_rate": 4.047026782528125e-06, + "loss": 0.6116, + "step": 4986 + }, + { + "epoch": 1.7510533707865168, + "grad_norm": 0.8469028472900391, + "learning_rate": 4.046663581024787e-06, + "loss": 0.5971, + "step": 4987 + }, + { + "epoch": 1.7514044943820224, + "grad_norm": 0.8719568252563477, + "learning_rate": 4.046300326626878e-06, + "loss": 0.5941, + "step": 4988 + }, + { + "epoch": 1.751755617977528, + "grad_norm": 0.879953920841217, + "learning_rate": 4.045937019346822e-06, + "loss": 0.6169, + "step": 4989 + }, + { + "epoch": 1.7521067415730336, + "grad_norm": 0.8611469268798828, + "learning_rate": 4.045573659197042e-06, + "loss": 0.6287, + "step": 4990 + }, + { + "epoch": 1.7524578651685392, + "grad_norm": 0.8662636876106262, + "learning_rate": 4.045210246189966e-06, + "loss": 0.6159, + "step": 4991 + }, + { + "epoch": 1.7528089887640448, + "grad_norm": 0.874043881893158, + "learning_rate": 4.044846780338021e-06, + "loss": 0.6005, + "step": 4992 + }, + { + "epoch": 1.7531601123595506, + "grad_norm": 0.8485775589942932, + "learning_rate": 4.0444832616536385e-06, + "loss": 0.6224, + "step": 4993 + }, + { + "epoch": 1.7535112359550562, + "grad_norm": 0.8543581962585449, + "learning_rate": 4.04411969014925e-06, + "loss": 0.5885, + "step": 4994 + }, + { + "epoch": 1.7538623595505618, + "grad_norm": 0.8683145642280579, + "learning_rate": 4.043756065837289e-06, + "loss": 0.6394, + "step": 4995 + }, + { + "epoch": 1.7542134831460674, + "grad_norm": 0.8860946893692017, + "learning_rate": 4.043392388730192e-06, + "loss": 0.6372, + "step": 4996 + }, + { + "epoch": 1.754564606741573, + "grad_norm": 0.8612002730369568, + "learning_rate": 4.043028658840397e-06, + "loss": 0.6136, + "step": 4997 + }, + { + "epoch": 1.7549157303370788, + "grad_norm": 0.8517924547195435, + "learning_rate": 4.042664876180341e-06, + "loss": 0.5788, + "step": 4998 + }, + { + "epoch": 1.7552668539325844, + "grad_norm": 0.8405445218086243, + "learning_rate": 4.042301040762467e-06, + "loss": 0.5945, + "step": 4999 + }, + { + "epoch": 1.75561797752809, + "grad_norm": 0.8536641597747803, + "learning_rate": 4.041937152599216e-06, + "loss": 0.5888, + "step": 5000 + }, + { + "epoch": 1.7559691011235956, + "grad_norm": 0.9032542705535889, + "learning_rate": 4.041573211703035e-06, + "loss": 0.6133, + "step": 5001 + }, + { + "epoch": 1.7563202247191012, + "grad_norm": 0.8637306094169617, + "learning_rate": 4.041209218086369e-06, + "loss": 0.6245, + "step": 5002 + }, + { + "epoch": 1.7566713483146068, + "grad_norm": 0.8811326026916504, + "learning_rate": 4.0408451717616656e-06, + "loss": 0.6569, + "step": 5003 + }, + { + "epoch": 1.7570224719101124, + "grad_norm": 0.8908126950263977, + "learning_rate": 4.040481072741378e-06, + "loss": 0.6469, + "step": 5004 + }, + { + "epoch": 1.757373595505618, + "grad_norm": 0.886318564414978, + "learning_rate": 4.040116921037954e-06, + "loss": 0.5808, + "step": 5005 + }, + { + "epoch": 1.7577247191011236, + "grad_norm": 0.8739534616470337, + "learning_rate": 4.03975271666385e-06, + "loss": 0.5976, + "step": 5006 + }, + { + "epoch": 1.7580758426966292, + "grad_norm": 0.9307764172554016, + "learning_rate": 4.03938845963152e-06, + "loss": 0.6418, + "step": 5007 + }, + { + "epoch": 1.7584269662921348, + "grad_norm": 0.861684262752533, + "learning_rate": 4.039024149953424e-06, + "loss": 0.6109, + "step": 5008 + }, + { + "epoch": 1.7587780898876404, + "grad_norm": 0.8454620838165283, + "learning_rate": 4.038659787642016e-06, + "loss": 0.5839, + "step": 5009 + }, + { + "epoch": 1.759129213483146, + "grad_norm": 0.8540975451469421, + "learning_rate": 4.038295372709762e-06, + "loss": 0.6302, + "step": 5010 + }, + { + "epoch": 1.7594803370786516, + "grad_norm": 0.8659608364105225, + "learning_rate": 4.0379309051691225e-06, + "loss": 0.6305, + "step": 5011 + }, + { + "epoch": 1.7598314606741572, + "grad_norm": 0.8791552782058716, + "learning_rate": 4.037566385032561e-06, + "loss": 0.6323, + "step": 5012 + }, + { + "epoch": 1.7601825842696628, + "grad_norm": 0.8560447692871094, + "learning_rate": 4.037201812312545e-06, + "loss": 0.5957, + "step": 5013 + }, + { + "epoch": 1.7605337078651684, + "grad_norm": 0.8539266586303711, + "learning_rate": 4.036837187021544e-06, + "loss": 0.6113, + "step": 5014 + }, + { + "epoch": 1.7608848314606742, + "grad_norm": 0.9080570340156555, + "learning_rate": 4.036472509172025e-06, + "loss": 0.6168, + "step": 5015 + }, + { + "epoch": 1.7612359550561798, + "grad_norm": 0.8955601453781128, + "learning_rate": 4.036107778776461e-06, + "loss": 0.6174, + "step": 5016 + }, + { + "epoch": 1.7615870786516854, + "grad_norm": 0.8747341632843018, + "learning_rate": 4.0357429958473256e-06, + "loss": 0.6357, + "step": 5017 + }, + { + "epoch": 1.761938202247191, + "grad_norm": 0.8803605437278748, + "learning_rate": 4.035378160397094e-06, + "loss": 0.5907, + "step": 5018 + }, + { + "epoch": 1.7622893258426966, + "grad_norm": 0.8748698234558105, + "learning_rate": 4.035013272438243e-06, + "loss": 0.6121, + "step": 5019 + }, + { + "epoch": 1.7626404494382022, + "grad_norm": 0.9040994048118591, + "learning_rate": 4.034648331983251e-06, + "loss": 0.6346, + "step": 5020 + }, + { + "epoch": 1.762991573033708, + "grad_norm": 0.8952014446258545, + "learning_rate": 4.0342833390446e-06, + "loss": 0.6016, + "step": 5021 + }, + { + "epoch": 1.7633426966292136, + "grad_norm": 0.8983685374259949, + "learning_rate": 4.033918293634772e-06, + "loss": 0.6257, + "step": 5022 + }, + { + "epoch": 1.7636938202247192, + "grad_norm": 0.8786833882331848, + "learning_rate": 4.03355319576625e-06, + "loss": 0.6118, + "step": 5023 + }, + { + "epoch": 1.7640449438202248, + "grad_norm": 0.8733046054840088, + "learning_rate": 4.033188045451522e-06, + "loss": 0.6078, + "step": 5024 + }, + { + "epoch": 1.7643960674157304, + "grad_norm": 0.8754127025604248, + "learning_rate": 4.032822842703075e-06, + "loss": 0.6507, + "step": 5025 + }, + { + "epoch": 1.764747191011236, + "grad_norm": 0.8841948509216309, + "learning_rate": 4.032457587533397e-06, + "loss": 0.6371, + "step": 5026 + }, + { + "epoch": 1.7650983146067416, + "grad_norm": 0.8662333488464355, + "learning_rate": 4.032092279954982e-06, + "loss": 0.5939, + "step": 5027 + }, + { + "epoch": 1.7654494382022472, + "grad_norm": 0.8241957426071167, + "learning_rate": 4.031726919980321e-06, + "loss": 0.5896, + "step": 5028 + }, + { + "epoch": 1.7658005617977528, + "grad_norm": 0.8761467933654785, + "learning_rate": 4.0313615076219105e-06, + "loss": 0.6228, + "step": 5029 + }, + { + "epoch": 1.7661516853932584, + "grad_norm": 0.8869350552558899, + "learning_rate": 4.030996042892248e-06, + "loss": 0.6509, + "step": 5030 + }, + { + "epoch": 1.766502808988764, + "grad_norm": 0.8962212204933167, + "learning_rate": 4.030630525803829e-06, + "loss": 0.6377, + "step": 5031 + }, + { + "epoch": 1.7668539325842696, + "grad_norm": 0.8528103828430176, + "learning_rate": 4.030264956369158e-06, + "loss": 0.5963, + "step": 5032 + }, + { + "epoch": 1.7672050561797752, + "grad_norm": 0.8822567462921143, + "learning_rate": 4.029899334600733e-06, + "loss": 0.6135, + "step": 5033 + }, + { + "epoch": 1.7675561797752808, + "grad_norm": 0.846155047416687, + "learning_rate": 4.029533660511061e-06, + "loss": 0.6057, + "step": 5034 + }, + { + "epoch": 1.7679073033707864, + "grad_norm": 0.9543619751930237, + "learning_rate": 4.029167934112646e-06, + "loss": 0.6142, + "step": 5035 + }, + { + "epoch": 1.768258426966292, + "grad_norm": 0.9285796880722046, + "learning_rate": 4.028802155417997e-06, + "loss": 0.6105, + "step": 5036 + }, + { + "epoch": 1.7686095505617978, + "grad_norm": 0.857733964920044, + "learning_rate": 4.028436324439622e-06, + "loss": 0.5895, + "step": 5037 + }, + { + "epoch": 1.7689606741573034, + "grad_norm": 0.8516778349876404, + "learning_rate": 4.028070441190032e-06, + "loss": 0.592, + "step": 5038 + }, + { + "epoch": 1.769311797752809, + "grad_norm": 0.8570651412010193, + "learning_rate": 4.027704505681741e-06, + "loss": 0.6106, + "step": 5039 + }, + { + "epoch": 1.7696629213483146, + "grad_norm": 0.8978039622306824, + "learning_rate": 4.027338517927264e-06, + "loss": 0.6333, + "step": 5040 + }, + { + "epoch": 1.7700140449438202, + "grad_norm": 0.8815526366233826, + "learning_rate": 4.026972477939115e-06, + "loss": 0.6127, + "step": 5041 + }, + { + "epoch": 1.7703651685393258, + "grad_norm": 0.8965631723403931, + "learning_rate": 4.026606385729816e-06, + "loss": 0.6252, + "step": 5042 + }, + { + "epoch": 1.7707162921348316, + "grad_norm": 0.8918066620826721, + "learning_rate": 4.026240241311883e-06, + "loss": 0.6325, + "step": 5043 + }, + { + "epoch": 1.7710674157303372, + "grad_norm": 0.8715677857398987, + "learning_rate": 4.025874044697841e-06, + "loss": 0.591, + "step": 5044 + }, + { + "epoch": 1.7714185393258428, + "grad_norm": 0.8741485476493835, + "learning_rate": 4.025507795900212e-06, + "loss": 0.6137, + "step": 5045 + }, + { + "epoch": 1.7717696629213484, + "grad_norm": 0.8711748123168945, + "learning_rate": 4.025141494931522e-06, + "loss": 0.6098, + "step": 5046 + }, + { + "epoch": 1.772120786516854, + "grad_norm": 0.9053213000297546, + "learning_rate": 4.024775141804297e-06, + "loss": 0.6198, + "step": 5047 + }, + { + "epoch": 1.7724719101123596, + "grad_norm": 0.8338163495063782, + "learning_rate": 4.024408736531068e-06, + "loss": 0.5808, + "step": 5048 + }, + { + "epoch": 1.7728230337078652, + "grad_norm": 0.8671820759773254, + "learning_rate": 4.024042279124364e-06, + "loss": 0.5916, + "step": 5049 + }, + { + "epoch": 1.7731741573033708, + "grad_norm": 0.9618306159973145, + "learning_rate": 4.023675769596718e-06, + "loss": 0.6416, + "step": 5050 + }, + { + "epoch": 1.7735252808988764, + "grad_norm": 0.8482176065444946, + "learning_rate": 4.023309207960666e-06, + "loss": 0.5838, + "step": 5051 + }, + { + "epoch": 1.773876404494382, + "grad_norm": 0.891528844833374, + "learning_rate": 4.0229425942287414e-06, + "loss": 0.6196, + "step": 5052 + }, + { + "epoch": 1.7742275280898876, + "grad_norm": 0.8804656267166138, + "learning_rate": 4.0225759284134835e-06, + "loss": 0.5967, + "step": 5053 + }, + { + "epoch": 1.7745786516853932, + "grad_norm": 0.863013744354248, + "learning_rate": 4.022209210527432e-06, + "loss": 0.6211, + "step": 5054 + }, + { + "epoch": 1.7749297752808988, + "grad_norm": 0.851879358291626, + "learning_rate": 4.021842440583127e-06, + "loss": 0.6305, + "step": 5055 + }, + { + "epoch": 1.7752808988764044, + "grad_norm": 0.8675127029418945, + "learning_rate": 4.0214756185931135e-06, + "loss": 0.6076, + "step": 5056 + }, + { + "epoch": 1.77563202247191, + "grad_norm": 0.8953244090080261, + "learning_rate": 4.021108744569936e-06, + "loss": 0.6286, + "step": 5057 + }, + { + "epoch": 1.7759831460674156, + "grad_norm": 0.8707031011581421, + "learning_rate": 4.020741818526141e-06, + "loss": 0.638, + "step": 5058 + }, + { + "epoch": 1.7763342696629212, + "grad_norm": 0.9085872173309326, + "learning_rate": 4.020374840474277e-06, + "loss": 0.6537, + "step": 5059 + }, + { + "epoch": 1.776685393258427, + "grad_norm": 0.8786725401878357, + "learning_rate": 4.020007810426895e-06, + "loss": 0.6362, + "step": 5060 + }, + { + "epoch": 1.7770365168539326, + "grad_norm": 0.889363706111908, + "learning_rate": 4.019640728396545e-06, + "loss": 0.6149, + "step": 5061 + }, + { + "epoch": 1.7773876404494382, + "grad_norm": 0.9281964302062988, + "learning_rate": 4.019273594395784e-06, + "loss": 0.6798, + "step": 5062 + }, + { + "epoch": 1.7777387640449438, + "grad_norm": 0.9001464247703552, + "learning_rate": 4.018906408437166e-06, + "loss": 0.6325, + "step": 5063 + }, + { + "epoch": 1.7780898876404494, + "grad_norm": 0.8923134207725525, + "learning_rate": 4.018539170533248e-06, + "loss": 0.6288, + "step": 5064 + }, + { + "epoch": 1.7784410112359552, + "grad_norm": 0.8892466425895691, + "learning_rate": 4.018171880696589e-06, + "loss": 0.628, + "step": 5065 + }, + { + "epoch": 1.7787921348314608, + "grad_norm": 0.8730875849723816, + "learning_rate": 4.017804538939753e-06, + "loss": 0.6235, + "step": 5066 + }, + { + "epoch": 1.7791432584269664, + "grad_norm": 0.902269721031189, + "learning_rate": 4.0174371452752995e-06, + "loss": 0.6527, + "step": 5067 + }, + { + "epoch": 1.779494382022472, + "grad_norm": 0.8821526169776917, + "learning_rate": 4.017069699715794e-06, + "loss": 0.5963, + "step": 5068 + }, + { + "epoch": 1.7798455056179776, + "grad_norm": 0.8617396354675293, + "learning_rate": 4.016702202273802e-06, + "loss": 0.6127, + "step": 5069 + }, + { + "epoch": 1.7801966292134832, + "grad_norm": 0.9029756784439087, + "learning_rate": 4.016334652961894e-06, + "loss": 0.6271, + "step": 5070 + }, + { + "epoch": 1.7805477528089888, + "grad_norm": 0.8544668555259705, + "learning_rate": 4.015967051792638e-06, + "loss": 0.6131, + "step": 5071 + }, + { + "epoch": 1.7808988764044944, + "grad_norm": 0.8941280841827393, + "learning_rate": 4.015599398778606e-06, + "loss": 0.6029, + "step": 5072 + }, + { + "epoch": 1.78125, + "grad_norm": 0.9132516384124756, + "learning_rate": 4.015231693932372e-06, + "loss": 0.6349, + "step": 5073 + }, + { + "epoch": 1.7816011235955056, + "grad_norm": 0.8514044284820557, + "learning_rate": 4.01486393726651e-06, + "loss": 0.5818, + "step": 5074 + }, + { + "epoch": 1.7819522471910112, + "grad_norm": 0.8844448328018188, + "learning_rate": 4.014496128793598e-06, + "loss": 0.615, + "step": 5075 + }, + { + "epoch": 1.7823033707865168, + "grad_norm": 0.861137330532074, + "learning_rate": 4.014128268526214e-06, + "loss": 0.5902, + "step": 5076 + }, + { + "epoch": 1.7826544943820224, + "grad_norm": 0.86243736743927, + "learning_rate": 4.013760356476939e-06, + "loss": 0.6108, + "step": 5077 + }, + { + "epoch": 1.783005617977528, + "grad_norm": 0.8797655701637268, + "learning_rate": 4.013392392658355e-06, + "loss": 0.6207, + "step": 5078 + }, + { + "epoch": 1.7833567415730336, + "grad_norm": 0.9152445793151855, + "learning_rate": 4.013024377083046e-06, + "loss": 0.6307, + "step": 5079 + }, + { + "epoch": 1.7837078651685392, + "grad_norm": 0.9087762236595154, + "learning_rate": 4.012656309763598e-06, + "loss": 0.6143, + "step": 5080 + }, + { + "epoch": 1.7840589887640448, + "grad_norm": 0.8641596436500549, + "learning_rate": 4.0122881907126e-06, + "loss": 0.6173, + "step": 5081 + }, + { + "epoch": 1.7844101123595506, + "grad_norm": 0.8863744139671326, + "learning_rate": 4.011920019942638e-06, + "loss": 0.6361, + "step": 5082 + }, + { + "epoch": 1.7847612359550562, + "grad_norm": 0.8684104084968567, + "learning_rate": 4.011551797466306e-06, + "loss": 0.6034, + "step": 5083 + }, + { + "epoch": 1.7851123595505618, + "grad_norm": 0.875816822052002, + "learning_rate": 4.011183523296197e-06, + "loss": 0.6085, + "step": 5084 + }, + { + "epoch": 1.7854634831460674, + "grad_norm": 0.8723933100700378, + "learning_rate": 4.010815197444903e-06, + "loss": 0.6009, + "step": 5085 + }, + { + "epoch": 1.785814606741573, + "grad_norm": 0.8689554333686829, + "learning_rate": 4.010446819925023e-06, + "loss": 0.6131, + "step": 5086 + }, + { + "epoch": 1.7861657303370788, + "grad_norm": 0.8560565710067749, + "learning_rate": 4.010078390749154e-06, + "loss": 0.5902, + "step": 5087 + }, + { + "epoch": 1.7865168539325844, + "grad_norm": 0.8722461462020874, + "learning_rate": 4.009709909929897e-06, + "loss": 0.6335, + "step": 5088 + }, + { + "epoch": 1.78686797752809, + "grad_norm": 0.8994833827018738, + "learning_rate": 4.009341377479851e-06, + "loss": 0.6265, + "step": 5089 + }, + { + "epoch": 1.7872191011235956, + "grad_norm": 0.8879294395446777, + "learning_rate": 4.008972793411623e-06, + "loss": 0.6255, + "step": 5090 + }, + { + "epoch": 1.7875702247191012, + "grad_norm": 0.8504826426506042, + "learning_rate": 4.008604157737816e-06, + "loss": 0.6158, + "step": 5091 + }, + { + "epoch": 1.7879213483146068, + "grad_norm": 0.8719592690467834, + "learning_rate": 4.008235470471038e-06, + "loss": 0.6005, + "step": 5092 + }, + { + "epoch": 1.7882724719101124, + "grad_norm": 0.8644405007362366, + "learning_rate": 4.007866731623898e-06, + "loss": 0.5995, + "step": 5093 + }, + { + "epoch": 1.788623595505618, + "grad_norm": 0.8683051466941833, + "learning_rate": 4.007497941209006e-06, + "loss": 0.6372, + "step": 5094 + }, + { + "epoch": 1.7889747191011236, + "grad_norm": 0.8745830655097961, + "learning_rate": 4.007129099238973e-06, + "loss": 0.6063, + "step": 5095 + }, + { + "epoch": 1.7893258426966292, + "grad_norm": 0.8765552639961243, + "learning_rate": 4.006760205726416e-06, + "loss": 0.6434, + "step": 5096 + }, + { + "epoch": 1.7896769662921348, + "grad_norm": 0.8652710318565369, + "learning_rate": 4.00639126068395e-06, + "loss": 0.6183, + "step": 5097 + }, + { + "epoch": 1.7900280898876404, + "grad_norm": 0.863023042678833, + "learning_rate": 4.00602226412419e-06, + "loss": 0.6265, + "step": 5098 + }, + { + "epoch": 1.790379213483146, + "grad_norm": 0.8672717213630676, + "learning_rate": 4.005653216059758e-06, + "loss": 0.6281, + "step": 5099 + }, + { + "epoch": 1.7907303370786516, + "grad_norm": 0.8654349446296692, + "learning_rate": 4.005284116503275e-06, + "loss": 0.6355, + "step": 5100 + }, + { + "epoch": 1.7910814606741572, + "grad_norm": 0.8771569728851318, + "learning_rate": 4.004914965467363e-06, + "loss": 0.6518, + "step": 5101 + }, + { + "epoch": 1.7914325842696628, + "grad_norm": 0.9185158014297485, + "learning_rate": 4.004545762964646e-06, + "loss": 0.6309, + "step": 5102 + }, + { + "epoch": 1.7917837078651684, + "grad_norm": 0.8859367966651917, + "learning_rate": 4.004176509007752e-06, + "loss": 0.6431, + "step": 5103 + }, + { + "epoch": 1.7921348314606742, + "grad_norm": 0.8867841958999634, + "learning_rate": 4.0038072036093086e-06, + "loss": 0.618, + "step": 5104 + }, + { + "epoch": 1.7924859550561798, + "grad_norm": 0.8852524757385254, + "learning_rate": 4.003437846781946e-06, + "loss": 0.6452, + "step": 5105 + }, + { + "epoch": 1.7928370786516854, + "grad_norm": 0.8460947275161743, + "learning_rate": 4.003068438538295e-06, + "loss": 0.6402, + "step": 5106 + }, + { + "epoch": 1.793188202247191, + "grad_norm": 0.8796756267547607, + "learning_rate": 4.00269897889099e-06, + "loss": 0.6215, + "step": 5107 + }, + { + "epoch": 1.7935393258426966, + "grad_norm": 0.8622691035270691, + "learning_rate": 4.002329467852666e-06, + "loss": 0.6182, + "step": 5108 + }, + { + "epoch": 1.7938904494382022, + "grad_norm": 0.8955039978027344, + "learning_rate": 4.001959905435959e-06, + "loss": 0.6284, + "step": 5109 + }, + { + "epoch": 1.794241573033708, + "grad_norm": 0.900213360786438, + "learning_rate": 4.001590291653509e-06, + "loss": 0.6138, + "step": 5110 + }, + { + "epoch": 1.7945926966292136, + "grad_norm": 0.8492575287818909, + "learning_rate": 4.0012206265179555e-06, + "loss": 0.5906, + "step": 5111 + }, + { + "epoch": 1.7949438202247192, + "grad_norm": 0.8939003944396973, + "learning_rate": 4.000850910041943e-06, + "loss": 0.6461, + "step": 5112 + }, + { + "epoch": 1.7952949438202248, + "grad_norm": 0.8977720737457275, + "learning_rate": 4.000481142238112e-06, + "loss": 0.6255, + "step": 5113 + }, + { + "epoch": 1.7956460674157304, + "grad_norm": 0.8857546448707581, + "learning_rate": 4.0001113231191095e-06, + "loss": 0.6419, + "step": 5114 + }, + { + "epoch": 1.795997191011236, + "grad_norm": 0.8970683217048645, + "learning_rate": 3.999741452697584e-06, + "loss": 0.6544, + "step": 5115 + }, + { + "epoch": 1.7963483146067416, + "grad_norm": 0.8892328143119812, + "learning_rate": 3.999371530986184e-06, + "loss": 0.5811, + "step": 5116 + }, + { + "epoch": 1.7966994382022472, + "grad_norm": 0.8691011667251587, + "learning_rate": 3.999001557997561e-06, + "loss": 0.6118, + "step": 5117 + }, + { + "epoch": 1.7970505617977528, + "grad_norm": 0.8691001534461975, + "learning_rate": 3.998631533744367e-06, + "loss": 0.6209, + "step": 5118 + }, + { + "epoch": 1.7974016853932584, + "grad_norm": 0.8801686763763428, + "learning_rate": 3.998261458239257e-06, + "loss": 0.5993, + "step": 5119 + }, + { + "epoch": 1.797752808988764, + "grad_norm": 0.8820109367370605, + "learning_rate": 3.997891331494888e-06, + "loss": 0.6182, + "step": 5120 + }, + { + "epoch": 1.7981039325842696, + "grad_norm": 0.891557514667511, + "learning_rate": 3.997521153523916e-06, + "loss": 0.6214, + "step": 5121 + }, + { + "epoch": 1.7984550561797752, + "grad_norm": 0.8902502059936523, + "learning_rate": 3.997150924339003e-06, + "loss": 0.6082, + "step": 5122 + }, + { + "epoch": 1.7988061797752808, + "grad_norm": 0.8698774576187134, + "learning_rate": 3.99678064395281e-06, + "loss": 0.6319, + "step": 5123 + }, + { + "epoch": 1.7991573033707864, + "grad_norm": 0.8708075284957886, + "learning_rate": 3.996410312377998e-06, + "loss": 0.5903, + "step": 5124 + }, + { + "epoch": 1.799508426966292, + "grad_norm": 0.860001266002655, + "learning_rate": 3.996039929627235e-06, + "loss": 0.6108, + "step": 5125 + }, + { + "epoch": 1.7998595505617978, + "grad_norm": 0.9757430553436279, + "learning_rate": 3.995669495713187e-06, + "loss": 0.6222, + "step": 5126 + }, + { + "epoch": 1.8002106741573034, + "grad_norm": 0.8471954464912415, + "learning_rate": 3.9952990106485216e-06, + "loss": 0.6127, + "step": 5127 + }, + { + "epoch": 1.800561797752809, + "grad_norm": 0.8945921063423157, + "learning_rate": 3.99492847444591e-06, + "loss": 0.6047, + "step": 5128 + }, + { + "epoch": 1.8009129213483146, + "grad_norm": 0.8653395175933838, + "learning_rate": 3.994557887118022e-06, + "loss": 0.6214, + "step": 5129 + }, + { + "epoch": 1.8012640449438202, + "grad_norm": 0.917481005191803, + "learning_rate": 3.994187248677535e-06, + "loss": 0.6226, + "step": 5130 + }, + { + "epoch": 1.8016151685393258, + "grad_norm": 0.862048864364624, + "learning_rate": 3.993816559137122e-06, + "loss": 0.6309, + "step": 5131 + }, + { + "epoch": 1.8019662921348316, + "grad_norm": 0.8727179765701294, + "learning_rate": 3.993445818509461e-06, + "loss": 0.602, + "step": 5132 + }, + { + "epoch": 1.8023174157303372, + "grad_norm": 0.8686352968215942, + "learning_rate": 3.9930750268072315e-06, + "loss": 0.5883, + "step": 5133 + }, + { + "epoch": 1.8026685393258428, + "grad_norm": 0.8954786658287048, + "learning_rate": 3.992704184043113e-06, + "loss": 0.6448, + "step": 5134 + }, + { + "epoch": 1.8030196629213484, + "grad_norm": 0.8747565150260925, + "learning_rate": 3.992333290229789e-06, + "loss": 0.5916, + "step": 5135 + }, + { + "epoch": 1.803370786516854, + "grad_norm": 0.905752956867218, + "learning_rate": 3.9919623453799436e-06, + "loss": 0.6148, + "step": 5136 + }, + { + "epoch": 1.8037219101123596, + "grad_norm": 0.8951407670974731, + "learning_rate": 3.991591349506263e-06, + "loss": 0.6234, + "step": 5137 + }, + { + "epoch": 1.8040730337078652, + "grad_norm": 0.87555992603302, + "learning_rate": 3.991220302621434e-06, + "loss": 0.6072, + "step": 5138 + }, + { + "epoch": 1.8044241573033708, + "grad_norm": 0.8721452355384827, + "learning_rate": 3.990849204738147e-06, + "loss": 0.6289, + "step": 5139 + }, + { + "epoch": 1.8047752808988764, + "grad_norm": 0.8791200518608093, + "learning_rate": 3.990478055869093e-06, + "loss": 0.6394, + "step": 5140 + }, + { + "epoch": 1.805126404494382, + "grad_norm": 0.8407002091407776, + "learning_rate": 3.990106856026965e-06, + "loss": 0.6079, + "step": 5141 + }, + { + "epoch": 1.8054775280898876, + "grad_norm": 0.9156894683837891, + "learning_rate": 3.9897356052244565e-06, + "loss": 0.6131, + "step": 5142 + }, + { + "epoch": 1.8058286516853932, + "grad_norm": 0.8723463416099548, + "learning_rate": 3.9893643034742665e-06, + "loss": 0.6361, + "step": 5143 + }, + { + "epoch": 1.8061797752808988, + "grad_norm": 0.8676952719688416, + "learning_rate": 3.988992950789091e-06, + "loss": 0.6159, + "step": 5144 + }, + { + "epoch": 1.8065308988764044, + "grad_norm": 0.8790785670280457, + "learning_rate": 3.988621547181631e-06, + "loss": 0.6162, + "step": 5145 + }, + { + "epoch": 1.80688202247191, + "grad_norm": 0.8970304727554321, + "learning_rate": 3.988250092664588e-06, + "loss": 0.5972, + "step": 5146 + }, + { + "epoch": 1.8072331460674156, + "grad_norm": 0.8706192970275879, + "learning_rate": 3.987878587250666e-06, + "loss": 0.6323, + "step": 5147 + }, + { + "epoch": 1.8075842696629212, + "grad_norm": 0.913070559501648, + "learning_rate": 3.987507030952568e-06, + "loss": 0.6182, + "step": 5148 + }, + { + "epoch": 1.807935393258427, + "grad_norm": 0.9056144952774048, + "learning_rate": 3.987135423783004e-06, + "loss": 0.6193, + "step": 5149 + }, + { + "epoch": 1.8082865168539326, + "grad_norm": 0.8835223317146301, + "learning_rate": 3.9867637657546795e-06, + "loss": 0.604, + "step": 5150 + }, + { + "epoch": 1.8086376404494382, + "grad_norm": 0.9045801758766174, + "learning_rate": 3.986392056880307e-06, + "loss": 0.6386, + "step": 5151 + }, + { + "epoch": 1.8089887640449438, + "grad_norm": 0.8738263845443726, + "learning_rate": 3.986020297172599e-06, + "loss": 0.6158, + "step": 5152 + }, + { + "epoch": 1.8093398876404494, + "grad_norm": 0.8748819828033447, + "learning_rate": 3.985648486644267e-06, + "loss": 0.6018, + "step": 5153 + }, + { + "epoch": 1.8096910112359552, + "grad_norm": 0.8693463802337646, + "learning_rate": 3.985276625308029e-06, + "loss": 0.6557, + "step": 5154 + }, + { + "epoch": 1.8100421348314608, + "grad_norm": 0.8723728060722351, + "learning_rate": 3.9849047131766015e-06, + "loss": 0.5882, + "step": 5155 + }, + { + "epoch": 1.8103932584269664, + "grad_norm": 0.8742467164993286, + "learning_rate": 3.984532750262703e-06, + "loss": 0.6267, + "step": 5156 + }, + { + "epoch": 1.810744382022472, + "grad_norm": 0.8632952570915222, + "learning_rate": 3.984160736579054e-06, + "loss": 0.6136, + "step": 5157 + }, + { + "epoch": 1.8110955056179776, + "grad_norm": 0.8474553823471069, + "learning_rate": 3.983788672138379e-06, + "loss": 0.6133, + "step": 5158 + }, + { + "epoch": 1.8114466292134832, + "grad_norm": 0.872573733329773, + "learning_rate": 3.983416556953401e-06, + "loss": 0.5971, + "step": 5159 + }, + { + "epoch": 1.8117977528089888, + "grad_norm": 0.877545952796936, + "learning_rate": 3.983044391036846e-06, + "loss": 0.6148, + "step": 5160 + }, + { + "epoch": 1.8121488764044944, + "grad_norm": 0.908349871635437, + "learning_rate": 3.982672174401442e-06, + "loss": 0.657, + "step": 5161 + }, + { + "epoch": 1.8125, + "grad_norm": 0.8775234818458557, + "learning_rate": 3.982299907059918e-06, + "loss": 0.6167, + "step": 5162 + }, + { + "epoch": 1.8128511235955056, + "grad_norm": 0.9253164529800415, + "learning_rate": 3.981927589025006e-06, + "loss": 0.6277, + "step": 5163 + }, + { + "epoch": 1.8132022471910112, + "grad_norm": 0.8476220965385437, + "learning_rate": 3.981555220309439e-06, + "loss": 0.6029, + "step": 5164 + }, + { + "epoch": 1.8135533707865168, + "grad_norm": 0.8830453157424927, + "learning_rate": 3.98118280092595e-06, + "loss": 0.6432, + "step": 5165 + }, + { + "epoch": 1.8139044943820224, + "grad_norm": 0.8966081738471985, + "learning_rate": 3.980810330887277e-06, + "loss": 0.6397, + "step": 5166 + }, + { + "epoch": 1.814255617977528, + "grad_norm": 0.8749344944953918, + "learning_rate": 3.980437810206158e-06, + "loss": 0.6352, + "step": 5167 + }, + { + "epoch": 1.8146067415730336, + "grad_norm": 0.8861076831817627, + "learning_rate": 3.980065238895333e-06, + "loss": 0.6306, + "step": 5168 + }, + { + "epoch": 1.8149578651685392, + "grad_norm": 0.9134885668754578, + "learning_rate": 3.979692616967543e-06, + "loss": 0.6036, + "step": 5169 + }, + { + "epoch": 1.8153089887640448, + "grad_norm": 0.8736245632171631, + "learning_rate": 3.979319944435532e-06, + "loss": 0.6032, + "step": 5170 + }, + { + "epoch": 1.8156601123595506, + "grad_norm": 0.881210207939148, + "learning_rate": 3.978947221312045e-06, + "loss": 0.6334, + "step": 5171 + }, + { + "epoch": 1.8160112359550562, + "grad_norm": 0.8849275708198547, + "learning_rate": 3.978574447609828e-06, + "loss": 0.5735, + "step": 5172 + }, + { + "epoch": 1.8163623595505618, + "grad_norm": 0.8601321578025818, + "learning_rate": 3.978201623341631e-06, + "loss": 0.5696, + "step": 5173 + }, + { + "epoch": 1.8167134831460674, + "grad_norm": 0.875879168510437, + "learning_rate": 3.977828748520203e-06, + "loss": 0.6013, + "step": 5174 + }, + { + "epoch": 1.817064606741573, + "grad_norm": 0.9210758209228516, + "learning_rate": 3.977455823158296e-06, + "loss": 0.6343, + "step": 5175 + }, + { + "epoch": 1.8174157303370788, + "grad_norm": 0.8983147144317627, + "learning_rate": 3.977082847268664e-06, + "loss": 0.598, + "step": 5176 + }, + { + "epoch": 1.8177668539325844, + "grad_norm": 0.8796141147613525, + "learning_rate": 3.976709820864063e-06, + "loss": 0.5995, + "step": 5177 + }, + { + "epoch": 1.81811797752809, + "grad_norm": 0.9013523459434509, + "learning_rate": 3.97633674395725e-06, + "loss": 0.6097, + "step": 5178 + }, + { + "epoch": 1.8184691011235956, + "grad_norm": 0.8761310577392578, + "learning_rate": 3.975963616560983e-06, + "loss": 0.6272, + "step": 5179 + }, + { + "epoch": 1.8188202247191012, + "grad_norm": 0.8740566968917847, + "learning_rate": 3.975590438688024e-06, + "loss": 0.6276, + "step": 5180 + }, + { + "epoch": 1.8191713483146068, + "grad_norm": 0.8199687600135803, + "learning_rate": 3.975217210351133e-06, + "loss": 0.5566, + "step": 5181 + }, + { + "epoch": 1.8195224719101124, + "grad_norm": 0.8775818347930908, + "learning_rate": 3.974843931563078e-06, + "loss": 0.6017, + "step": 5182 + }, + { + "epoch": 1.819873595505618, + "grad_norm": 0.8951563239097595, + "learning_rate": 3.974470602336622e-06, + "loss": 0.6004, + "step": 5183 + }, + { + "epoch": 1.8202247191011236, + "grad_norm": 0.8890185952186584, + "learning_rate": 3.974097222684532e-06, + "loss": 0.6362, + "step": 5184 + }, + { + "epoch": 1.8205758426966292, + "grad_norm": 0.8939943909645081, + "learning_rate": 3.97372379261958e-06, + "loss": 0.6268, + "step": 5185 + }, + { + "epoch": 1.8209269662921348, + "grad_norm": 0.8659777641296387, + "learning_rate": 3.9733503121545345e-06, + "loss": 0.624, + "step": 5186 + }, + { + "epoch": 1.8212780898876404, + "grad_norm": 0.8676909804344177, + "learning_rate": 3.972976781302169e-06, + "loss": 0.6138, + "step": 5187 + }, + { + "epoch": 1.821629213483146, + "grad_norm": 0.8564205169677734, + "learning_rate": 3.972603200075258e-06, + "loss": 0.5945, + "step": 5188 + }, + { + "epoch": 1.8219803370786516, + "grad_norm": 0.8555643558502197, + "learning_rate": 3.972229568486578e-06, + "loss": 0.5785, + "step": 5189 + }, + { + "epoch": 1.8223314606741572, + "grad_norm": 0.9054913520812988, + "learning_rate": 3.971855886548906e-06, + "loss": 0.6445, + "step": 5190 + }, + { + "epoch": 1.8226825842696628, + "grad_norm": 0.9118171334266663, + "learning_rate": 3.9714821542750226e-06, + "loss": 0.6134, + "step": 5191 + }, + { + "epoch": 1.8230337078651684, + "grad_norm": 0.8640024065971375, + "learning_rate": 3.9711083716777085e-06, + "loss": 0.6404, + "step": 5192 + }, + { + "epoch": 1.8233848314606742, + "grad_norm": 0.8471640348434448, + "learning_rate": 3.970734538769747e-06, + "loss": 0.5966, + "step": 5193 + }, + { + "epoch": 1.8237359550561798, + "grad_norm": 0.8837315440177917, + "learning_rate": 3.970360655563923e-06, + "loss": 0.6001, + "step": 5194 + }, + { + "epoch": 1.8240870786516854, + "grad_norm": 0.8849839568138123, + "learning_rate": 3.969986722073023e-06, + "loss": 0.6301, + "step": 5195 + }, + { + "epoch": 1.824438202247191, + "grad_norm": 0.8799655437469482, + "learning_rate": 3.969612738309834e-06, + "loss": 0.5701, + "step": 5196 + }, + { + "epoch": 1.8247893258426966, + "grad_norm": 0.8617206811904907, + "learning_rate": 3.969238704287147e-06, + "loss": 0.6126, + "step": 5197 + }, + { + "epoch": 1.8251404494382022, + "grad_norm": 0.9338657259941101, + "learning_rate": 3.968864620017755e-06, + "loss": 0.6022, + "step": 5198 + }, + { + "epoch": 1.825491573033708, + "grad_norm": 0.8785296678543091, + "learning_rate": 3.968490485514449e-06, + "loss": 0.6073, + "step": 5199 + }, + { + "epoch": 1.8258426966292136, + "grad_norm": 0.8909986019134521, + "learning_rate": 3.9681163007900255e-06, + "loss": 0.6152, + "step": 5200 + }, + { + "epoch": 1.8261938202247192, + "grad_norm": 0.9001063704490662, + "learning_rate": 3.96774206585728e-06, + "loss": 0.6351, + "step": 5201 + }, + { + "epoch": 1.8265449438202248, + "grad_norm": 0.8840953707695007, + "learning_rate": 3.967367780729012e-06, + "loss": 0.6452, + "step": 5202 + }, + { + "epoch": 1.8268960674157304, + "grad_norm": 0.8713771104812622, + "learning_rate": 3.966993445418022e-06, + "loss": 0.6099, + "step": 5203 + }, + { + "epoch": 1.827247191011236, + "grad_norm": 0.8601224422454834, + "learning_rate": 3.9666190599371116e-06, + "loss": 0.5861, + "step": 5204 + }, + { + "epoch": 1.8275983146067416, + "grad_norm": 0.8551958799362183, + "learning_rate": 3.9662446242990834e-06, + "loss": 0.5977, + "step": 5205 + }, + { + "epoch": 1.8279494382022472, + "grad_norm": 0.8728621602058411, + "learning_rate": 3.965870138516744e-06, + "loss": 0.6523, + "step": 5206 + }, + { + "epoch": 1.8283005617977528, + "grad_norm": 0.8756167888641357, + "learning_rate": 3.965495602602901e-06, + "loss": 0.6479, + "step": 5207 + }, + { + "epoch": 1.8286516853932584, + "grad_norm": 0.8521246910095215, + "learning_rate": 3.965121016570361e-06, + "loss": 0.6009, + "step": 5208 + }, + { + "epoch": 1.829002808988764, + "grad_norm": 0.8759570121765137, + "learning_rate": 3.964746380431937e-06, + "loss": 0.6221, + "step": 5209 + }, + { + "epoch": 1.8293539325842696, + "grad_norm": 0.8832354545593262, + "learning_rate": 3.964371694200441e-06, + "loss": 0.6124, + "step": 5210 + }, + { + "epoch": 1.8297050561797752, + "grad_norm": 0.8482329845428467, + "learning_rate": 3.963996957888685e-06, + "loss": 0.5887, + "step": 5211 + }, + { + "epoch": 1.8300561797752808, + "grad_norm": 0.8687066435813904, + "learning_rate": 3.963622171509486e-06, + "loss": 0.6097, + "step": 5212 + }, + { + "epoch": 1.8304073033707864, + "grad_norm": 0.8731704950332642, + "learning_rate": 3.963247335075661e-06, + "loss": 0.5865, + "step": 5213 + }, + { + "epoch": 1.830758426966292, + "grad_norm": 0.8566392660140991, + "learning_rate": 3.9628724486000295e-06, + "loss": 0.6283, + "step": 5214 + }, + { + "epoch": 1.8311095505617978, + "grad_norm": 0.867788553237915, + "learning_rate": 3.962497512095412e-06, + "loss": 0.6278, + "step": 5215 + }, + { + "epoch": 1.8314606741573034, + "grad_norm": 0.913202166557312, + "learning_rate": 3.9621225255746306e-06, + "loss": 0.6382, + "step": 5216 + }, + { + "epoch": 1.831811797752809, + "grad_norm": 0.8661263585090637, + "learning_rate": 3.9617474890505114e-06, + "loss": 0.619, + "step": 5217 + }, + { + "epoch": 1.8321629213483146, + "grad_norm": 0.867225170135498, + "learning_rate": 3.961372402535878e-06, + "loss": 0.6006, + "step": 5218 + }, + { + "epoch": 1.8325140449438202, + "grad_norm": 0.9224227070808411, + "learning_rate": 3.9609972660435586e-06, + "loss": 0.6413, + "step": 5219 + }, + { + "epoch": 1.8328651685393258, + "grad_norm": 0.9165964722633362, + "learning_rate": 3.960622079586384e-06, + "loss": 0.5787, + "step": 5220 + }, + { + "epoch": 1.8332162921348316, + "grad_norm": 0.9111368060112, + "learning_rate": 3.960246843177182e-06, + "loss": 0.637, + "step": 5221 + }, + { + "epoch": 1.8335674157303372, + "grad_norm": 0.892612874507904, + "learning_rate": 3.959871556828789e-06, + "loss": 0.6237, + "step": 5222 + }, + { + "epoch": 1.8339185393258428, + "grad_norm": 0.8713697791099548, + "learning_rate": 3.9594962205540385e-06, + "loss": 0.6418, + "step": 5223 + }, + { + "epoch": 1.8342696629213484, + "grad_norm": 0.8447199463844299, + "learning_rate": 3.959120834365766e-06, + "loss": 0.606, + "step": 5224 + }, + { + "epoch": 1.834620786516854, + "grad_norm": 0.8594667911529541, + "learning_rate": 3.958745398276808e-06, + "loss": 0.6233, + "step": 5225 + }, + { + "epoch": 1.8349719101123596, + "grad_norm": 0.8984555602073669, + "learning_rate": 3.958369912300007e-06, + "loss": 0.6336, + "step": 5226 + }, + { + "epoch": 1.8353230337078652, + "grad_norm": 0.8514502048492432, + "learning_rate": 3.957994376448203e-06, + "loss": 0.5858, + "step": 5227 + }, + { + "epoch": 1.8356741573033708, + "grad_norm": 0.8587605357170105, + "learning_rate": 3.957618790734239e-06, + "loss": 0.6066, + "step": 5228 + }, + { + "epoch": 1.8360252808988764, + "grad_norm": 0.8868040442466736, + "learning_rate": 3.957243155170959e-06, + "loss": 0.6111, + "step": 5229 + }, + { + "epoch": 1.836376404494382, + "grad_norm": 0.8550049662590027, + "learning_rate": 3.956867469771211e-06, + "loss": 0.5917, + "step": 5230 + }, + { + "epoch": 1.8367275280898876, + "grad_norm": 0.885949969291687, + "learning_rate": 3.956491734547841e-06, + "loss": 0.6348, + "step": 5231 + }, + { + "epoch": 1.8370786516853932, + "grad_norm": 0.8566646575927734, + "learning_rate": 3.9561159495137e-06, + "loss": 0.597, + "step": 5232 + }, + { + "epoch": 1.8374297752808988, + "grad_norm": 0.9448535442352295, + "learning_rate": 3.95574011468164e-06, + "loss": 0.6376, + "step": 5233 + }, + { + "epoch": 1.8377808988764044, + "grad_norm": 0.8502497673034668, + "learning_rate": 3.955364230064515e-06, + "loss": 0.6009, + "step": 5234 + }, + { + "epoch": 1.83813202247191, + "grad_norm": 0.8837283849716187, + "learning_rate": 3.9549882956751766e-06, + "loss": 0.6434, + "step": 5235 + }, + { + "epoch": 1.8384831460674156, + "grad_norm": 0.9013311862945557, + "learning_rate": 3.954612311526485e-06, + "loss": 0.5851, + "step": 5236 + }, + { + "epoch": 1.8388342696629212, + "grad_norm": 0.9204567074775696, + "learning_rate": 3.9542362776312974e-06, + "loss": 0.655, + "step": 5237 + }, + { + "epoch": 1.839185393258427, + "grad_norm": 0.8798168301582336, + "learning_rate": 3.953860194002472e-06, + "loss": 0.6303, + "step": 5238 + }, + { + "epoch": 1.8395365168539326, + "grad_norm": 0.9072458744049072, + "learning_rate": 3.953484060652872e-06, + "loss": 0.6276, + "step": 5239 + }, + { + "epoch": 1.8398876404494382, + "grad_norm": 0.8829042911529541, + "learning_rate": 3.9531078775953625e-06, + "loss": 0.631, + "step": 5240 + }, + { + "epoch": 1.8402387640449438, + "grad_norm": 0.8745750784873962, + "learning_rate": 3.952731644842806e-06, + "loss": 0.6143, + "step": 5241 + }, + { + "epoch": 1.8405898876404494, + "grad_norm": 0.9073604345321655, + "learning_rate": 3.95235536240807e-06, + "loss": 0.627, + "step": 5242 + }, + { + "epoch": 1.8409410112359552, + "grad_norm": 0.8719262480735779, + "learning_rate": 3.951979030304024e-06, + "loss": 0.6162, + "step": 5243 + }, + { + "epoch": 1.8412921348314608, + "grad_norm": 0.8346160650253296, + "learning_rate": 3.9516026485435375e-06, + "loss": 0.6049, + "step": 5244 + }, + { + "epoch": 1.8416432584269664, + "grad_norm": 0.8686418533325195, + "learning_rate": 3.9512262171394835e-06, + "loss": 0.6157, + "step": 5245 + }, + { + "epoch": 1.841994382022472, + "grad_norm": 0.8777024745941162, + "learning_rate": 3.9508497361047334e-06, + "loss": 0.6434, + "step": 5246 + }, + { + "epoch": 1.8423455056179776, + "grad_norm": 0.8348974585533142, + "learning_rate": 3.950473205452165e-06, + "loss": 0.5557, + "step": 5247 + }, + { + "epoch": 1.8426966292134832, + "grad_norm": 0.856793224811554, + "learning_rate": 3.950096625194653e-06, + "loss": 0.5629, + "step": 5248 + }, + { + "epoch": 1.8430477528089888, + "grad_norm": 0.8882337212562561, + "learning_rate": 3.949719995345078e-06, + "loss": 0.6041, + "step": 5249 + }, + { + "epoch": 1.8433988764044944, + "grad_norm": 0.9027885794639587, + "learning_rate": 3.9493433159163195e-06, + "loss": 0.6082, + "step": 5250 + }, + { + "epoch": 1.84375, + "grad_norm": 0.8926260471343994, + "learning_rate": 3.94896658692126e-06, + "loss": 0.5864, + "step": 5251 + }, + { + "epoch": 1.8441011235955056, + "grad_norm": 0.8764528632164001, + "learning_rate": 3.948589808372783e-06, + "loss": 0.6275, + "step": 5252 + }, + { + "epoch": 1.8444522471910112, + "grad_norm": 0.8873112797737122, + "learning_rate": 3.948212980283773e-06, + "loss": 0.6121, + "step": 5253 + }, + { + "epoch": 1.8448033707865168, + "grad_norm": 0.8742668628692627, + "learning_rate": 3.947836102667119e-06, + "loss": 0.6105, + "step": 5254 + }, + { + "epoch": 1.8451544943820224, + "grad_norm": 0.8881484270095825, + "learning_rate": 3.947459175535711e-06, + "loss": 0.6315, + "step": 5255 + }, + { + "epoch": 1.845505617977528, + "grad_norm": 0.8762322068214417, + "learning_rate": 3.9470821989024356e-06, + "loss": 0.61, + "step": 5256 + }, + { + "epoch": 1.8458567415730336, + "grad_norm": 0.8735952973365784, + "learning_rate": 3.946705172780189e-06, + "loss": 0.6066, + "step": 5257 + }, + { + "epoch": 1.8462078651685392, + "grad_norm": 0.8771315813064575, + "learning_rate": 3.946328097181862e-06, + "loss": 0.6196, + "step": 5258 + }, + { + "epoch": 1.8465589887640448, + "grad_norm": 0.8824196457862854, + "learning_rate": 3.945950972120351e-06, + "loss": 0.6753, + "step": 5259 + }, + { + "epoch": 1.8469101123595506, + "grad_norm": 0.8754235506057739, + "learning_rate": 3.9455737976085564e-06, + "loss": 0.6097, + "step": 5260 + }, + { + "epoch": 1.8472612359550562, + "grad_norm": 0.8685277700424194, + "learning_rate": 3.945196573659374e-06, + "loss": 0.6107, + "step": 5261 + }, + { + "epoch": 1.8476123595505618, + "grad_norm": 0.8714946508407593, + "learning_rate": 3.944819300285705e-06, + "loss": 0.6071, + "step": 5262 + }, + { + "epoch": 1.8479634831460674, + "grad_norm": 0.8681133985519409, + "learning_rate": 3.9444419775004525e-06, + "loss": 0.6028, + "step": 5263 + }, + { + "epoch": 1.848314606741573, + "grad_norm": 0.8595680594444275, + "learning_rate": 3.9440646053165204e-06, + "loss": 0.5931, + "step": 5264 + }, + { + "epoch": 1.8486657303370788, + "grad_norm": 0.9302463531494141, + "learning_rate": 3.943687183746815e-06, + "loss": 0.6239, + "step": 5265 + }, + { + "epoch": 1.8490168539325844, + "grad_norm": 0.8686720132827759, + "learning_rate": 3.943309712804243e-06, + "loss": 0.6177, + "step": 5266 + }, + { + "epoch": 1.84936797752809, + "grad_norm": 0.8854331970214844, + "learning_rate": 3.9429321925017136e-06, + "loss": 0.6291, + "step": 5267 + }, + { + "epoch": 1.8497191011235956, + "grad_norm": 0.8775175213813782, + "learning_rate": 3.942554622852138e-06, + "loss": 0.6006, + "step": 5268 + }, + { + "epoch": 1.8500702247191012, + "grad_norm": 0.8732326030731201, + "learning_rate": 3.942177003868428e-06, + "loss": 0.6282, + "step": 5269 + }, + { + "epoch": 1.8504213483146068, + "grad_norm": 0.8572641015052795, + "learning_rate": 3.9417993355635e-06, + "loss": 0.613, + "step": 5270 + }, + { + "epoch": 1.8507724719101124, + "grad_norm": 0.8818247318267822, + "learning_rate": 3.941421617950268e-06, + "loss": 0.5826, + "step": 5271 + }, + { + "epoch": 1.851123595505618, + "grad_norm": 0.8704795837402344, + "learning_rate": 3.9410438510416494e-06, + "loss": 0.6254, + "step": 5272 + }, + { + "epoch": 1.8514747191011236, + "grad_norm": 0.8848576545715332, + "learning_rate": 3.940666034850565e-06, + "loss": 0.6199, + "step": 5273 + }, + { + "epoch": 1.8518258426966292, + "grad_norm": 0.8868470191955566, + "learning_rate": 3.940288169389935e-06, + "loss": 0.617, + "step": 5274 + }, + { + "epoch": 1.8521769662921348, + "grad_norm": 0.875606894493103, + "learning_rate": 3.939910254672682e-06, + "loss": 0.5775, + "step": 5275 + }, + { + "epoch": 1.8525280898876404, + "grad_norm": 0.8833727240562439, + "learning_rate": 3.93953229071173e-06, + "loss": 0.607, + "step": 5276 + }, + { + "epoch": 1.852879213483146, + "grad_norm": 0.9035220742225647, + "learning_rate": 3.939154277520006e-06, + "loss": 0.6213, + "step": 5277 + }, + { + "epoch": 1.8532303370786516, + "grad_norm": 0.8979851007461548, + "learning_rate": 3.938776215110437e-06, + "loss": 0.6331, + "step": 5278 + }, + { + "epoch": 1.8535814606741572, + "grad_norm": 0.8894830346107483, + "learning_rate": 3.938398103495952e-06, + "loss": 0.585, + "step": 5279 + }, + { + "epoch": 1.8539325842696628, + "grad_norm": 0.8893154859542847, + "learning_rate": 3.938019942689484e-06, + "loss": 0.5978, + "step": 5280 + }, + { + "epoch": 1.8542837078651684, + "grad_norm": 0.8567217588424683, + "learning_rate": 3.9376417327039636e-06, + "loss": 0.6029, + "step": 5281 + }, + { + "epoch": 1.8546348314606742, + "grad_norm": 0.8965117335319519, + "learning_rate": 3.937263473552326e-06, + "loss": 0.6029, + "step": 5282 + }, + { + "epoch": 1.8549859550561798, + "grad_norm": 0.8815075755119324, + "learning_rate": 3.936885165247508e-06, + "loss": 0.6741, + "step": 5283 + }, + { + "epoch": 1.8553370786516854, + "grad_norm": 0.8753764629364014, + "learning_rate": 3.936506807802446e-06, + "loss": 0.6105, + "step": 5284 + }, + { + "epoch": 1.855688202247191, + "grad_norm": 0.884854257106781, + "learning_rate": 3.9361284012300816e-06, + "loss": 0.6191, + "step": 5285 + }, + { + "epoch": 1.8560393258426966, + "grad_norm": 0.8499030470848083, + "learning_rate": 3.935749945543354e-06, + "loss": 0.6179, + "step": 5286 + }, + { + "epoch": 1.8563904494382022, + "grad_norm": 0.8841039538383484, + "learning_rate": 3.935371440755208e-06, + "loss": 0.6265, + "step": 5287 + }, + { + "epoch": 1.856741573033708, + "grad_norm": 0.855475902557373, + "learning_rate": 3.934992886878586e-06, + "loss": 0.5774, + "step": 5288 + }, + { + "epoch": 1.8570926966292136, + "grad_norm": 0.8800686597824097, + "learning_rate": 3.934614283926435e-06, + "loss": 0.6131, + "step": 5289 + }, + { + "epoch": 1.8574438202247192, + "grad_norm": 0.8433512449264526, + "learning_rate": 3.934235631911704e-06, + "loss": 0.5718, + "step": 5290 + }, + { + "epoch": 1.8577949438202248, + "grad_norm": 0.9216337203979492, + "learning_rate": 3.933856930847341e-06, + "loss": 0.6243, + "step": 5291 + }, + { + "epoch": 1.8581460674157304, + "grad_norm": 0.8634188175201416, + "learning_rate": 3.9334781807462984e-06, + "loss": 0.6016, + "step": 5292 + }, + { + "epoch": 1.858497191011236, + "grad_norm": 0.8891706466674805, + "learning_rate": 3.933099381621528e-06, + "loss": 0.6323, + "step": 5293 + }, + { + "epoch": 1.8588483146067416, + "grad_norm": 0.8881804943084717, + "learning_rate": 3.9327205334859855e-06, + "loss": 0.6182, + "step": 5294 + }, + { + "epoch": 1.8591994382022472, + "grad_norm": 0.8558664917945862, + "learning_rate": 3.932341636352627e-06, + "loss": 0.6333, + "step": 5295 + }, + { + "epoch": 1.8595505617977528, + "grad_norm": 0.8729931116104126, + "learning_rate": 3.93196269023441e-06, + "loss": 0.6282, + "step": 5296 + }, + { + "epoch": 1.8599016853932584, + "grad_norm": 0.9086493849754333, + "learning_rate": 3.931583695144294e-06, + "loss": 0.6378, + "step": 5297 + }, + { + "epoch": 1.860252808988764, + "grad_norm": 0.8665076494216919, + "learning_rate": 3.931204651095241e-06, + "loss": 0.6327, + "step": 5298 + }, + { + "epoch": 1.8606039325842696, + "grad_norm": 0.9046199917793274, + "learning_rate": 3.9308255581002145e-06, + "loss": 0.6651, + "step": 5299 + }, + { + "epoch": 1.8609550561797752, + "grad_norm": 0.89439857006073, + "learning_rate": 3.930446416172177e-06, + "loss": 0.6414, + "step": 5300 + }, + { + "epoch": 1.8613061797752808, + "grad_norm": 0.8834724426269531, + "learning_rate": 3.930067225324096e-06, + "loss": 0.5866, + "step": 5301 + }, + { + "epoch": 1.8616573033707864, + "grad_norm": 0.882892906665802, + "learning_rate": 3.92968798556894e-06, + "loss": 0.6551, + "step": 5302 + }, + { + "epoch": 1.862008426966292, + "grad_norm": 0.8716971278190613, + "learning_rate": 3.929308696919678e-06, + "loss": 0.6237, + "step": 5303 + }, + { + "epoch": 1.8623595505617978, + "grad_norm": 0.8676100373268127, + "learning_rate": 3.928929359389282e-06, + "loss": 0.5929, + "step": 5304 + }, + { + "epoch": 1.8627106741573034, + "grad_norm": 0.922857403755188, + "learning_rate": 3.928549972990725e-06, + "loss": 0.679, + "step": 5305 + }, + { + "epoch": 1.863061797752809, + "grad_norm": 0.8612826466560364, + "learning_rate": 3.9281705377369814e-06, + "loss": 0.6116, + "step": 5306 + }, + { + "epoch": 1.8634129213483146, + "grad_norm": 0.8769382238388062, + "learning_rate": 3.927791053641027e-06, + "loss": 0.6057, + "step": 5307 + }, + { + "epoch": 1.8637640449438202, + "grad_norm": 0.8767523169517517, + "learning_rate": 3.92741152071584e-06, + "loss": 0.6181, + "step": 5308 + }, + { + "epoch": 1.8641151685393258, + "grad_norm": 0.897783637046814, + "learning_rate": 3.927031938974401e-06, + "loss": 0.5993, + "step": 5309 + }, + { + "epoch": 1.8644662921348316, + "grad_norm": 0.8882983922958374, + "learning_rate": 3.926652308429691e-06, + "loss": 0.6075, + "step": 5310 + }, + { + "epoch": 1.8648174157303372, + "grad_norm": 0.9197914600372314, + "learning_rate": 3.926272629094692e-06, + "loss": 0.6058, + "step": 5311 + }, + { + "epoch": 1.8651685393258428, + "grad_norm": 0.8851439356803894, + "learning_rate": 3.9258929009823906e-06, + "loss": 0.6073, + "step": 5312 + }, + { + "epoch": 1.8655196629213484, + "grad_norm": 0.8552735447883606, + "learning_rate": 3.925513124105772e-06, + "loss": 0.6214, + "step": 5313 + }, + { + "epoch": 1.865870786516854, + "grad_norm": 0.8668675422668457, + "learning_rate": 3.925133298477824e-06, + "loss": 0.6129, + "step": 5314 + }, + { + "epoch": 1.8662219101123596, + "grad_norm": 0.8721663951873779, + "learning_rate": 3.924753424111537e-06, + "loss": 0.6205, + "step": 5315 + }, + { + "epoch": 1.8665730337078652, + "grad_norm": 0.8686200976371765, + "learning_rate": 3.9243735010199014e-06, + "loss": 0.6144, + "step": 5316 + }, + { + "epoch": 1.8669241573033708, + "grad_norm": 0.8643429279327393, + "learning_rate": 3.923993529215912e-06, + "loss": 0.618, + "step": 5317 + }, + { + "epoch": 1.8672752808988764, + "grad_norm": 0.8749399185180664, + "learning_rate": 3.923613508712562e-06, + "loss": 0.6436, + "step": 5318 + }, + { + "epoch": 1.867626404494382, + "grad_norm": 0.8470568060874939, + "learning_rate": 3.923233439522847e-06, + "loss": 0.6082, + "step": 5319 + }, + { + "epoch": 1.8679775280898876, + "grad_norm": 0.8877463340759277, + "learning_rate": 3.922853321659767e-06, + "loss": 0.6028, + "step": 5320 + }, + { + "epoch": 1.8683286516853932, + "grad_norm": 0.8590446710586548, + "learning_rate": 3.922473155136322e-06, + "loss": 0.601, + "step": 5321 + }, + { + "epoch": 1.8686797752808988, + "grad_norm": 0.9017147421836853, + "learning_rate": 3.922092939965511e-06, + "loss": 0.6602, + "step": 5322 + }, + { + "epoch": 1.8690308988764044, + "grad_norm": 0.8631618022918701, + "learning_rate": 3.921712676160339e-06, + "loss": 0.6088, + "step": 5323 + }, + { + "epoch": 1.86938202247191, + "grad_norm": 0.8838371634483337, + "learning_rate": 3.921332363733811e-06, + "loss": 0.6288, + "step": 5324 + }, + { + "epoch": 1.8697331460674156, + "grad_norm": 0.8923317193984985, + "learning_rate": 3.9209520026989305e-06, + "loss": 0.6259, + "step": 5325 + }, + { + "epoch": 1.8700842696629212, + "grad_norm": 0.8856344223022461, + "learning_rate": 3.920571593068708e-06, + "loss": 0.6069, + "step": 5326 + }, + { + "epoch": 1.870435393258427, + "grad_norm": 0.8869331479072571, + "learning_rate": 3.920191134856153e-06, + "loss": 0.5994, + "step": 5327 + }, + { + "epoch": 1.8707865168539326, + "grad_norm": 0.8856245875358582, + "learning_rate": 3.9198106280742765e-06, + "loss": 0.6061, + "step": 5328 + }, + { + "epoch": 1.8711376404494382, + "grad_norm": 0.8708471059799194, + "learning_rate": 3.919430072736091e-06, + "loss": 0.597, + "step": 5329 + }, + { + "epoch": 1.8714887640449438, + "grad_norm": 0.9098658561706543, + "learning_rate": 3.919049468854612e-06, + "loss": 0.6303, + "step": 5330 + }, + { + "epoch": 1.8718398876404494, + "grad_norm": 0.8640565276145935, + "learning_rate": 3.918668816442856e-06, + "loss": 0.6217, + "step": 5331 + }, + { + "epoch": 1.8721910112359552, + "grad_norm": 0.8827778697013855, + "learning_rate": 3.918288115513839e-06, + "loss": 0.6286, + "step": 5332 + }, + { + "epoch": 1.8725421348314608, + "grad_norm": 0.867546021938324, + "learning_rate": 3.917907366080583e-06, + "loss": 0.6291, + "step": 5333 + }, + { + "epoch": 1.8728932584269664, + "grad_norm": 0.8924286961555481, + "learning_rate": 3.917526568156109e-06, + "loss": 0.6058, + "step": 5334 + }, + { + "epoch": 1.873244382022472, + "grad_norm": 0.8416984677314758, + "learning_rate": 3.917145721753438e-06, + "loss": 0.5685, + "step": 5335 + }, + { + "epoch": 1.8735955056179776, + "grad_norm": 0.836463212966919, + "learning_rate": 3.916764826885597e-06, + "loss": 0.5749, + "step": 5336 + }, + { + "epoch": 1.8739466292134832, + "grad_norm": 0.888582706451416, + "learning_rate": 3.916383883565612e-06, + "loss": 0.5987, + "step": 5337 + }, + { + "epoch": 1.8742977528089888, + "grad_norm": 0.8800485134124756, + "learning_rate": 3.916002891806508e-06, + "loss": 0.5966, + "step": 5338 + }, + { + "epoch": 1.8746488764044944, + "grad_norm": 0.8612186908721924, + "learning_rate": 3.915621851621318e-06, + "loss": 0.5848, + "step": 5339 + }, + { + "epoch": 1.875, + "grad_norm": 0.8935150504112244, + "learning_rate": 3.915240763023072e-06, + "loss": 0.6036, + "step": 5340 + }, + { + "epoch": 1.8753511235955056, + "grad_norm": 0.8847930431365967, + "learning_rate": 3.914859626024804e-06, + "loss": 0.6171, + "step": 5341 + }, + { + "epoch": 1.8757022471910112, + "grad_norm": 0.8781599998474121, + "learning_rate": 3.914478440639546e-06, + "loss": 0.615, + "step": 5342 + }, + { + "epoch": 1.8760533707865168, + "grad_norm": 0.902090847492218, + "learning_rate": 3.914097206880336e-06, + "loss": 0.6234, + "step": 5343 + }, + { + "epoch": 1.8764044943820224, + "grad_norm": 0.8890019655227661, + "learning_rate": 3.913715924760212e-06, + "loss": 0.6005, + "step": 5344 + }, + { + "epoch": 1.876755617977528, + "grad_norm": 0.853225827217102, + "learning_rate": 3.913334594292213e-06, + "loss": 0.6023, + "step": 5345 + }, + { + "epoch": 1.8771067415730336, + "grad_norm": 0.8685701489448547, + "learning_rate": 3.91295321548938e-06, + "loss": 0.6291, + "step": 5346 + }, + { + "epoch": 1.8774578651685392, + "grad_norm": 0.8687042593955994, + "learning_rate": 3.912571788364756e-06, + "loss": 0.5823, + "step": 5347 + }, + { + "epoch": 1.8778089887640448, + "grad_norm": 0.8955281972885132, + "learning_rate": 3.9121903129313864e-06, + "loss": 0.6188, + "step": 5348 + }, + { + "epoch": 1.8781601123595506, + "grad_norm": 0.911578893661499, + "learning_rate": 3.911808789202316e-06, + "loss": 0.5788, + "step": 5349 + }, + { + "epoch": 1.8785112359550562, + "grad_norm": 0.8692489862442017, + "learning_rate": 3.911427217190593e-06, + "loss": 0.6135, + "step": 5350 + }, + { + "epoch": 1.8788623595505618, + "grad_norm": 0.8962387442588806, + "learning_rate": 3.911045596909267e-06, + "loss": 0.6012, + "step": 5351 + }, + { + "epoch": 1.8792134831460674, + "grad_norm": 0.8675644993782043, + "learning_rate": 3.910663928371389e-06, + "loss": 0.6014, + "step": 5352 + }, + { + "epoch": 1.879564606741573, + "grad_norm": 0.8898077011108398, + "learning_rate": 3.910282211590012e-06, + "loss": 0.6502, + "step": 5353 + }, + { + "epoch": 1.8799157303370788, + "grad_norm": 0.876844584941864, + "learning_rate": 3.9099004465781895e-06, + "loss": 0.6123, + "step": 5354 + }, + { + "epoch": 1.8802668539325844, + "grad_norm": 0.8768188953399658, + "learning_rate": 3.909518633348979e-06, + "loss": 0.611, + "step": 5355 + }, + { + "epoch": 1.88061797752809, + "grad_norm": 0.9023895859718323, + "learning_rate": 3.909136771915437e-06, + "loss": 0.6135, + "step": 5356 + }, + { + "epoch": 1.8809691011235956, + "grad_norm": 0.9012351632118225, + "learning_rate": 3.9087548622906235e-06, + "loss": 0.6493, + "step": 5357 + }, + { + "epoch": 1.8813202247191012, + "grad_norm": 0.8937700986862183, + "learning_rate": 3.908372904487599e-06, + "loss": 0.6015, + "step": 5358 + }, + { + "epoch": 1.8816713483146068, + "grad_norm": 0.8769406676292419, + "learning_rate": 3.9079908985194255e-06, + "loss": 0.6112, + "step": 5359 + }, + { + "epoch": 1.8820224719101124, + "grad_norm": 0.887407124042511, + "learning_rate": 3.90760884439917e-06, + "loss": 0.5949, + "step": 5360 + }, + { + "epoch": 1.882373595505618, + "grad_norm": 0.8920381665229797, + "learning_rate": 3.907226742139895e-06, + "loss": 0.6141, + "step": 5361 + }, + { + "epoch": 1.8827247191011236, + "grad_norm": 0.9120955467224121, + "learning_rate": 3.906844591754671e-06, + "loss": 0.6315, + "step": 5362 + }, + { + "epoch": 1.8830758426966292, + "grad_norm": 0.8675406575202942, + "learning_rate": 3.906462393256565e-06, + "loss": 0.6008, + "step": 5363 + }, + { + "epoch": 1.8834269662921348, + "grad_norm": 0.8848645091056824, + "learning_rate": 3.906080146658648e-06, + "loss": 0.6364, + "step": 5364 + }, + { + "epoch": 1.8837780898876404, + "grad_norm": 0.873410165309906, + "learning_rate": 3.905697851973994e-06, + "loss": 0.6179, + "step": 5365 + }, + { + "epoch": 1.884129213483146, + "grad_norm": 0.8644540905952454, + "learning_rate": 3.905315509215676e-06, + "loss": 0.5992, + "step": 5366 + }, + { + "epoch": 1.8844803370786516, + "grad_norm": 0.8811582326889038, + "learning_rate": 3.9049331183967694e-06, + "loss": 0.6385, + "step": 5367 + }, + { + "epoch": 1.8848314606741572, + "grad_norm": 0.8806861042976379, + "learning_rate": 3.904550679530354e-06, + "loss": 0.5863, + "step": 5368 + }, + { + "epoch": 1.8851825842696628, + "grad_norm": 0.8641405701637268, + "learning_rate": 3.904168192629506e-06, + "loss": 0.6297, + "step": 5369 + }, + { + "epoch": 1.8855337078651684, + "grad_norm": 0.8805460929870605, + "learning_rate": 3.903785657707308e-06, + "loss": 0.6499, + "step": 5370 + }, + { + "epoch": 1.8858848314606742, + "grad_norm": 0.8589033484458923, + "learning_rate": 3.9034030747768416e-06, + "loss": 0.5999, + "step": 5371 + }, + { + "epoch": 1.8862359550561798, + "grad_norm": 0.8469181060791016, + "learning_rate": 3.903020443851191e-06, + "loss": 0.6029, + "step": 5372 + }, + { + "epoch": 1.8865870786516854, + "grad_norm": 0.8949747681617737, + "learning_rate": 3.902637764943441e-06, + "loss": 0.633, + "step": 5373 + }, + { + "epoch": 1.886938202247191, + "grad_norm": 0.8772246241569519, + "learning_rate": 3.902255038066681e-06, + "loss": 0.6144, + "step": 5374 + }, + { + "epoch": 1.8872893258426966, + "grad_norm": 0.8535354733467102, + "learning_rate": 3.901872263233998e-06, + "loss": 0.6059, + "step": 5375 + }, + { + "epoch": 1.8876404494382022, + "grad_norm": 0.8883733153343201, + "learning_rate": 3.901489440458483e-06, + "loss": 0.5846, + "step": 5376 + }, + { + "epoch": 1.887991573033708, + "grad_norm": 0.8794171810150146, + "learning_rate": 3.901106569753229e-06, + "loss": 0.5991, + "step": 5377 + }, + { + "epoch": 1.8883426966292136, + "grad_norm": 0.8670855164527893, + "learning_rate": 3.900723651131328e-06, + "loss": 0.6293, + "step": 5378 + }, + { + "epoch": 1.8886938202247192, + "grad_norm": 0.8933359384536743, + "learning_rate": 3.900340684605878e-06, + "loss": 0.6334, + "step": 5379 + }, + { + "epoch": 1.8890449438202248, + "grad_norm": 0.8721469640731812, + "learning_rate": 3.899957670189974e-06, + "loss": 0.6133, + "step": 5380 + }, + { + "epoch": 1.8893960674157304, + "grad_norm": 0.8683612942695618, + "learning_rate": 3.899574607896716e-06, + "loss": 0.6315, + "step": 5381 + }, + { + "epoch": 1.889747191011236, + "grad_norm": 0.8970484137535095, + "learning_rate": 3.8991914977392055e-06, + "loss": 0.6686, + "step": 5382 + }, + { + "epoch": 1.8900983146067416, + "grad_norm": 0.8801787495613098, + "learning_rate": 3.898808339730541e-06, + "loss": 0.6471, + "step": 5383 + }, + { + "epoch": 1.8904494382022472, + "grad_norm": 0.907516360282898, + "learning_rate": 3.89842513388383e-06, + "loss": 0.654, + "step": 5384 + }, + { + "epoch": 1.8908005617977528, + "grad_norm": 0.8796942830085754, + "learning_rate": 3.898041880212174e-06, + "loss": 0.5926, + "step": 5385 + }, + { + "epoch": 1.8911516853932584, + "grad_norm": 0.873957633972168, + "learning_rate": 3.897658578728683e-06, + "loss": 0.6269, + "step": 5386 + }, + { + "epoch": 1.891502808988764, + "grad_norm": 0.8868711590766907, + "learning_rate": 3.897275229446466e-06, + "loss": 0.6081, + "step": 5387 + }, + { + "epoch": 1.8918539325842696, + "grad_norm": 0.8819994926452637, + "learning_rate": 3.89689183237863e-06, + "loss": 0.6204, + "step": 5388 + }, + { + "epoch": 1.8922050561797752, + "grad_norm": 0.8778470158576965, + "learning_rate": 3.896508387538289e-06, + "loss": 0.6256, + "step": 5389 + }, + { + "epoch": 1.8925561797752808, + "grad_norm": 0.8505924344062805, + "learning_rate": 3.896124894938556e-06, + "loss": 0.6361, + "step": 5390 + }, + { + "epoch": 1.8929073033707864, + "grad_norm": 0.8763235211372375, + "learning_rate": 3.8957413545925475e-06, + "loss": 0.6421, + "step": 5391 + }, + { + "epoch": 1.893258426966292, + "grad_norm": 0.8803117871284485, + "learning_rate": 3.895357766513378e-06, + "loss": 0.6148, + "step": 5392 + }, + { + "epoch": 1.8936095505617978, + "grad_norm": 0.8600309491157532, + "learning_rate": 3.894974130714167e-06, + "loss": 0.6275, + "step": 5393 + }, + { + "epoch": 1.8939606741573034, + "grad_norm": 0.8821873664855957, + "learning_rate": 3.894590447208034e-06, + "loss": 0.6356, + "step": 5394 + }, + { + "epoch": 1.894311797752809, + "grad_norm": 0.8747897148132324, + "learning_rate": 3.894206716008102e-06, + "loss": 0.6109, + "step": 5395 + }, + { + "epoch": 1.8946629213483146, + "grad_norm": 0.8427042365074158, + "learning_rate": 3.893822937127492e-06, + "loss": 0.5866, + "step": 5396 + }, + { + "epoch": 1.8950140449438202, + "grad_norm": 0.8987433910369873, + "learning_rate": 3.893439110579332e-06, + "loss": 0.6339, + "step": 5397 + }, + { + "epoch": 1.8953651685393258, + "grad_norm": 0.8769526481628418, + "learning_rate": 3.893055236376746e-06, + "loss": 0.6208, + "step": 5398 + }, + { + "epoch": 1.8957162921348316, + "grad_norm": 0.8786547183990479, + "learning_rate": 3.892671314532863e-06, + "loss": 0.655, + "step": 5399 + }, + { + "epoch": 1.8960674157303372, + "grad_norm": 0.8721256256103516, + "learning_rate": 3.892287345060814e-06, + "loss": 0.6026, + "step": 5400 + }, + { + "epoch": 1.8964185393258428, + "grad_norm": 0.9021056294441223, + "learning_rate": 3.891903327973727e-06, + "loss": 0.6335, + "step": 5401 + }, + { + "epoch": 1.8967696629213484, + "grad_norm": 0.8656460642814636, + "learning_rate": 3.891519263284738e-06, + "loss": 0.5922, + "step": 5402 + }, + { + "epoch": 1.897120786516854, + "grad_norm": 0.8590431213378906, + "learning_rate": 3.891135151006981e-06, + "loss": 0.6004, + "step": 5403 + }, + { + "epoch": 1.8974719101123596, + "grad_norm": 0.8588739037513733, + "learning_rate": 3.890750991153592e-06, + "loss": 0.5901, + "step": 5404 + }, + { + "epoch": 1.8978230337078652, + "grad_norm": 0.8661170601844788, + "learning_rate": 3.89036678373771e-06, + "loss": 0.6204, + "step": 5405 + }, + { + "epoch": 1.8981741573033708, + "grad_norm": 0.8917558193206787, + "learning_rate": 3.889982528772474e-06, + "loss": 0.6184, + "step": 5406 + }, + { + "epoch": 1.8985252808988764, + "grad_norm": 0.8691694736480713, + "learning_rate": 3.8895982262710236e-06, + "loss": 0.6294, + "step": 5407 + }, + { + "epoch": 1.898876404494382, + "grad_norm": 0.9101895093917847, + "learning_rate": 3.8892138762465034e-06, + "loss": 0.6445, + "step": 5408 + }, + { + "epoch": 1.8992275280898876, + "grad_norm": 0.8621460795402527, + "learning_rate": 3.888829478712058e-06, + "loss": 0.5856, + "step": 5409 + }, + { + "epoch": 1.8995786516853932, + "grad_norm": 0.8767841458320618, + "learning_rate": 3.8884450336808325e-06, + "loss": 0.6094, + "step": 5410 + }, + { + "epoch": 1.8999297752808988, + "grad_norm": 0.8723804354667664, + "learning_rate": 3.888060541165974e-06, + "loss": 0.6424, + "step": 5411 + }, + { + "epoch": 1.9002808988764044, + "grad_norm": 0.914440393447876, + "learning_rate": 3.887676001180635e-06, + "loss": 0.6079, + "step": 5412 + }, + { + "epoch": 1.90063202247191, + "grad_norm": 0.8787926435470581, + "learning_rate": 3.887291413737962e-06, + "loss": 0.5855, + "step": 5413 + }, + { + "epoch": 1.9009831460674156, + "grad_norm": 0.8624381422996521, + "learning_rate": 3.886906778851111e-06, + "loss": 0.5986, + "step": 5414 + }, + { + "epoch": 1.9013342696629212, + "grad_norm": 0.8594074845314026, + "learning_rate": 3.886522096533235e-06, + "loss": 0.5894, + "step": 5415 + }, + { + "epoch": 1.901685393258427, + "grad_norm": 0.8813564777374268, + "learning_rate": 3.8861373667974885e-06, + "loss": 0.6339, + "step": 5416 + }, + { + "epoch": 1.9020365168539326, + "grad_norm": 0.900949239730835, + "learning_rate": 3.885752589657032e-06, + "loss": 0.6144, + "step": 5417 + }, + { + "epoch": 1.9023876404494382, + "grad_norm": 0.9057722091674805, + "learning_rate": 3.885367765125021e-06, + "loss": 0.5978, + "step": 5418 + }, + { + "epoch": 1.9027387640449438, + "grad_norm": 0.8934159874916077, + "learning_rate": 3.884982893214619e-06, + "loss": 0.642, + "step": 5419 + }, + { + "epoch": 1.9030898876404494, + "grad_norm": 0.888180673122406, + "learning_rate": 3.884597973938987e-06, + "loss": 0.6387, + "step": 5420 + }, + { + "epoch": 1.9034410112359552, + "grad_norm": 0.8636809587478638, + "learning_rate": 3.884213007311289e-06, + "loss": 0.6016, + "step": 5421 + }, + { + "epoch": 1.9037921348314608, + "grad_norm": 0.8603489995002747, + "learning_rate": 3.883827993344691e-06, + "loss": 0.5827, + "step": 5422 + }, + { + "epoch": 1.9041432584269664, + "grad_norm": 0.8876716494560242, + "learning_rate": 3.88344293205236e-06, + "loss": 0.5968, + "step": 5423 + }, + { + "epoch": 1.904494382022472, + "grad_norm": 0.9033610820770264, + "learning_rate": 3.883057823447464e-06, + "loss": 0.6301, + "step": 5424 + }, + { + "epoch": 1.9048455056179776, + "grad_norm": 0.9055176377296448, + "learning_rate": 3.882672667543174e-06, + "loss": 0.6085, + "step": 5425 + }, + { + "epoch": 1.9051966292134832, + "grad_norm": 0.8801239132881165, + "learning_rate": 3.882287464352662e-06, + "loss": 0.5865, + "step": 5426 + }, + { + "epoch": 1.9055477528089888, + "grad_norm": 0.8756425380706787, + "learning_rate": 3.8819022138891e-06, + "loss": 0.6097, + "step": 5427 + }, + { + "epoch": 1.9058988764044944, + "grad_norm": 0.8778197169303894, + "learning_rate": 3.881516916165668e-06, + "loss": 0.6026, + "step": 5428 + }, + { + "epoch": 1.90625, + "grad_norm": 0.8809471726417542, + "learning_rate": 3.881131571195537e-06, + "loss": 0.5967, + "step": 5429 + }, + { + "epoch": 1.9066011235955056, + "grad_norm": 0.854507565498352, + "learning_rate": 3.880746178991889e-06, + "loss": 0.5973, + "step": 5430 + }, + { + "epoch": 1.9069522471910112, + "grad_norm": 0.8528857827186584, + "learning_rate": 3.8803607395679035e-06, + "loss": 0.5809, + "step": 5431 + }, + { + "epoch": 1.9073033707865168, + "grad_norm": 0.8797081112861633, + "learning_rate": 3.879975252936761e-06, + "loss": 0.6207, + "step": 5432 + }, + { + "epoch": 1.9076544943820224, + "grad_norm": 0.862160325050354, + "learning_rate": 3.879589719111646e-06, + "loss": 0.5845, + "step": 5433 + }, + { + "epoch": 1.908005617977528, + "grad_norm": 0.8742191791534424, + "learning_rate": 3.879204138105744e-06, + "loss": 0.6332, + "step": 5434 + }, + { + "epoch": 1.9083567415730336, + "grad_norm": 0.9287744760513306, + "learning_rate": 3.87881850993224e-06, + "loss": 0.6571, + "step": 5435 + }, + { + "epoch": 1.9087078651685392, + "grad_norm": 0.8513810634613037, + "learning_rate": 3.878432834604324e-06, + "loss": 0.5712, + "step": 5436 + }, + { + "epoch": 1.9090589887640448, + "grad_norm": 0.9044253826141357, + "learning_rate": 3.878047112135183e-06, + "loss": 0.6297, + "step": 5437 + }, + { + "epoch": 1.9094101123595506, + "grad_norm": 0.8363608717918396, + "learning_rate": 3.877661342538011e-06, + "loss": 0.6032, + "step": 5438 + }, + { + "epoch": 1.9097612359550562, + "grad_norm": 0.9041305184364319, + "learning_rate": 3.877275525826e-06, + "loss": 0.6192, + "step": 5439 + }, + { + "epoch": 1.9101123595505618, + "grad_norm": 0.8639614582061768, + "learning_rate": 3.876889662012343e-06, + "loss": 0.6213, + "step": 5440 + }, + { + "epoch": 1.9104634831460674, + "grad_norm": 0.8700418472290039, + "learning_rate": 3.876503751110239e-06, + "loss": 0.6209, + "step": 5441 + }, + { + "epoch": 1.910814606741573, + "grad_norm": 0.899722158908844, + "learning_rate": 3.876117793132884e-06, + "loss": 0.6461, + "step": 5442 + }, + { + "epoch": 1.9111657303370788, + "grad_norm": 0.8731403350830078, + "learning_rate": 3.875731788093478e-06, + "loss": 0.6135, + "step": 5443 + }, + { + "epoch": 1.9115168539325844, + "grad_norm": 0.8795063495635986, + "learning_rate": 3.8753457360052235e-06, + "loss": 0.6442, + "step": 5444 + }, + { + "epoch": 1.91186797752809, + "grad_norm": 0.8663721680641174, + "learning_rate": 3.874959636881321e-06, + "loss": 0.6225, + "step": 5445 + }, + { + "epoch": 1.9122191011235956, + "grad_norm": 0.863272488117218, + "learning_rate": 3.874573490734975e-06, + "loss": 0.6031, + "step": 5446 + }, + { + "epoch": 1.9125702247191012, + "grad_norm": 0.8670037388801575, + "learning_rate": 3.874187297579392e-06, + "loss": 0.6105, + "step": 5447 + }, + { + "epoch": 1.9129213483146068, + "grad_norm": 0.898658812046051, + "learning_rate": 3.87380105742778e-06, + "loss": 0.6176, + "step": 5448 + }, + { + "epoch": 1.9132724719101124, + "grad_norm": 0.8771390914916992, + "learning_rate": 3.873414770293347e-06, + "loss": 0.5724, + "step": 5449 + }, + { + "epoch": 1.913623595505618, + "grad_norm": 0.8916599750518799, + "learning_rate": 3.873028436189304e-06, + "loss": 0.6308, + "step": 5450 + }, + { + "epoch": 1.9139747191011236, + "grad_norm": 0.8692615628242493, + "learning_rate": 3.872642055128863e-06, + "loss": 0.5857, + "step": 5451 + }, + { + "epoch": 1.9143258426966292, + "grad_norm": 0.9091446399688721, + "learning_rate": 3.872255627125238e-06, + "loss": 0.6443, + "step": 5452 + }, + { + "epoch": 1.9146769662921348, + "grad_norm": 0.8689329624176025, + "learning_rate": 3.871869152191646e-06, + "loss": 0.6319, + "step": 5453 + }, + { + "epoch": 1.9150280898876404, + "grad_norm": 0.8973026275634766, + "learning_rate": 3.871482630341303e-06, + "loss": 0.6078, + "step": 5454 + }, + { + "epoch": 1.915379213483146, + "grad_norm": 0.8688374161720276, + "learning_rate": 3.871096061587427e-06, + "loss": 0.5782, + "step": 5455 + }, + { + "epoch": 1.9157303370786516, + "grad_norm": 0.9019613862037659, + "learning_rate": 3.8707094459432384e-06, + "loss": 0.6484, + "step": 5456 + }, + { + "epoch": 1.9160814606741572, + "grad_norm": 0.8789473176002502, + "learning_rate": 3.870322783421961e-06, + "loss": 0.5992, + "step": 5457 + }, + { + "epoch": 1.9164325842696628, + "grad_norm": 0.8779850602149963, + "learning_rate": 3.869936074036816e-06, + "loss": 0.5993, + "step": 5458 + }, + { + "epoch": 1.9167837078651684, + "grad_norm": 0.8998771905899048, + "learning_rate": 3.86954931780103e-06, + "loss": 0.6341, + "step": 5459 + }, + { + "epoch": 1.9171348314606742, + "grad_norm": 0.8945295214653015, + "learning_rate": 3.869162514727829e-06, + "loss": 0.6197, + "step": 5460 + }, + { + "epoch": 1.9174859550561798, + "grad_norm": 0.8839888572692871, + "learning_rate": 3.868775664830442e-06, + "loss": 0.6411, + "step": 5461 + }, + { + "epoch": 1.9178370786516854, + "grad_norm": 0.8995748162269592, + "learning_rate": 3.868388768122098e-06, + "loss": 0.6471, + "step": 5462 + }, + { + "epoch": 1.918188202247191, + "grad_norm": 0.8744586706161499, + "learning_rate": 3.868001824616029e-06, + "loss": 0.6023, + "step": 5463 + }, + { + "epoch": 1.9185393258426966, + "grad_norm": 0.8731626272201538, + "learning_rate": 3.867614834325469e-06, + "loss": 0.6145, + "step": 5464 + }, + { + "epoch": 1.9188904494382022, + "grad_norm": 0.8779453635215759, + "learning_rate": 3.867227797263652e-06, + "loss": 0.5688, + "step": 5465 + }, + { + "epoch": 1.919241573033708, + "grad_norm": 0.894486129283905, + "learning_rate": 3.866840713443814e-06, + "loss": 0.5936, + "step": 5466 + }, + { + "epoch": 1.9195926966292136, + "grad_norm": 0.9034679532051086, + "learning_rate": 3.866453582879193e-06, + "loss": 0.6005, + "step": 5467 + }, + { + "epoch": 1.9199438202247192, + "grad_norm": 0.8829547166824341, + "learning_rate": 3.866066405583028e-06, + "loss": 0.6141, + "step": 5468 + }, + { + "epoch": 1.9202949438202248, + "grad_norm": 0.8914617896080017, + "learning_rate": 3.865679181568561e-06, + "loss": 0.6344, + "step": 5469 + }, + { + "epoch": 1.9206460674157304, + "grad_norm": 0.8478212952613831, + "learning_rate": 3.865291910849036e-06, + "loss": 0.5906, + "step": 5470 + }, + { + "epoch": 1.920997191011236, + "grad_norm": 0.9483685493469238, + "learning_rate": 3.864904593437696e-06, + "loss": 0.6285, + "step": 5471 + }, + { + "epoch": 1.9213483146067416, + "grad_norm": 0.8849635124206543, + "learning_rate": 3.8645172293477855e-06, + "loss": 0.6054, + "step": 5472 + }, + { + "epoch": 1.9216994382022472, + "grad_norm": 0.8696728348731995, + "learning_rate": 3.864129818592554e-06, + "loss": 0.63, + "step": 5473 + }, + { + "epoch": 1.9220505617977528, + "grad_norm": 0.8640512824058533, + "learning_rate": 3.86374236118525e-06, + "loss": 0.5815, + "step": 5474 + }, + { + "epoch": 1.9224016853932584, + "grad_norm": 0.9038639068603516, + "learning_rate": 3.8633548571391245e-06, + "loss": 0.6459, + "step": 5475 + }, + { + "epoch": 1.922752808988764, + "grad_norm": 0.8907269835472107, + "learning_rate": 3.862967306467429e-06, + "loss": 0.6009, + "step": 5476 + }, + { + "epoch": 1.9231039325842696, + "grad_norm": 0.9073399305343628, + "learning_rate": 3.862579709183418e-06, + "loss": 0.6359, + "step": 5477 + }, + { + "epoch": 1.9234550561797752, + "grad_norm": 0.8706175684928894, + "learning_rate": 3.862192065300347e-06, + "loss": 0.6203, + "step": 5478 + }, + { + "epoch": 1.9238061797752808, + "grad_norm": 0.874502956867218, + "learning_rate": 3.861804374831474e-06, + "loss": 0.5929, + "step": 5479 + }, + { + "epoch": 1.9241573033707864, + "grad_norm": 0.846248984336853, + "learning_rate": 3.861416637790056e-06, + "loss": 0.5648, + "step": 5480 + }, + { + "epoch": 1.924508426966292, + "grad_norm": 0.8775991797447205, + "learning_rate": 3.861028854189353e-06, + "loss": 0.6009, + "step": 5481 + }, + { + "epoch": 1.9248595505617978, + "grad_norm": 0.9060062766075134, + "learning_rate": 3.860641024042628e-06, + "loss": 0.6432, + "step": 5482 + }, + { + "epoch": 1.9252106741573034, + "grad_norm": 0.8684884309768677, + "learning_rate": 3.8602531473631456e-06, + "loss": 0.5969, + "step": 5483 + }, + { + "epoch": 1.925561797752809, + "grad_norm": 0.9077668190002441, + "learning_rate": 3.859865224164169e-06, + "loss": 0.6311, + "step": 5484 + }, + { + "epoch": 1.9259129213483146, + "grad_norm": 0.8688258528709412, + "learning_rate": 3.8594772544589656e-06, + "loss": 0.6016, + "step": 5485 + }, + { + "epoch": 1.9262640449438202, + "grad_norm": 0.8717395663261414, + "learning_rate": 3.8590892382608025e-06, + "loss": 0.6134, + "step": 5486 + }, + { + "epoch": 1.9266151685393258, + "grad_norm": 0.8912414312362671, + "learning_rate": 3.858701175582951e-06, + "loss": 0.6481, + "step": 5487 + }, + { + "epoch": 1.9269662921348316, + "grad_norm": 0.858781099319458, + "learning_rate": 3.858313066438682e-06, + "loss": 0.6464, + "step": 5488 + }, + { + "epoch": 1.9273174157303372, + "grad_norm": 0.858666718006134, + "learning_rate": 3.857924910841269e-06, + "loss": 0.6094, + "step": 5489 + }, + { + "epoch": 1.9276685393258428, + "grad_norm": 0.8739560842514038, + "learning_rate": 3.857536708803985e-06, + "loss": 0.6047, + "step": 5490 + }, + { + "epoch": 1.9280196629213484, + "grad_norm": 0.8531132936477661, + "learning_rate": 3.857148460340107e-06, + "loss": 0.586, + "step": 5491 + }, + { + "epoch": 1.928370786516854, + "grad_norm": 0.8856545686721802, + "learning_rate": 3.856760165462914e-06, + "loss": 0.5874, + "step": 5492 + }, + { + "epoch": 1.9287219101123596, + "grad_norm": 0.9217509627342224, + "learning_rate": 3.856371824185684e-06, + "loss": 0.5857, + "step": 5493 + }, + { + "epoch": 1.9290730337078652, + "grad_norm": 0.845321774482727, + "learning_rate": 3.855983436521698e-06, + "loss": 0.6045, + "step": 5494 + }, + { + "epoch": 1.9294241573033708, + "grad_norm": 0.84823077917099, + "learning_rate": 3.855595002484239e-06, + "loss": 0.621, + "step": 5495 + }, + { + "epoch": 1.9297752808988764, + "grad_norm": 1.008187174797058, + "learning_rate": 3.855206522086591e-06, + "loss": 0.6289, + "step": 5496 + }, + { + "epoch": 1.930126404494382, + "grad_norm": 0.8827589154243469, + "learning_rate": 3.8548179953420405e-06, + "loss": 0.5957, + "step": 5497 + }, + { + "epoch": 1.9304775280898876, + "grad_norm": 0.8767725825309753, + "learning_rate": 3.854429422263872e-06, + "loss": 0.5905, + "step": 5498 + }, + { + "epoch": 1.9308286516853932, + "grad_norm": 0.8842640519142151, + "learning_rate": 3.854040802865377e-06, + "loss": 0.6496, + "step": 5499 + }, + { + "epoch": 1.9311797752808988, + "grad_norm": 0.8716596364974976, + "learning_rate": 3.853652137159846e-06, + "loss": 0.6259, + "step": 5500 + }, + { + "epoch": 1.9315308988764044, + "grad_norm": 0.8906645774841309, + "learning_rate": 3.85326342516057e-06, + "loss": 0.5954, + "step": 5501 + }, + { + "epoch": 1.93188202247191, + "grad_norm": 0.8781253099441528, + "learning_rate": 3.8528746668808425e-06, + "loss": 0.6137, + "step": 5502 + }, + { + "epoch": 1.9322331460674156, + "grad_norm": 0.9453568458557129, + "learning_rate": 3.8524858623339585e-06, + "loss": 0.6333, + "step": 5503 + }, + { + "epoch": 1.9325842696629212, + "grad_norm": 0.8589761257171631, + "learning_rate": 3.852097011533217e-06, + "loss": 0.6573, + "step": 5504 + }, + { + "epoch": 1.932935393258427, + "grad_norm": 1.0276621580123901, + "learning_rate": 3.851708114491913e-06, + "loss": 0.594, + "step": 5505 + }, + { + "epoch": 1.9332865168539326, + "grad_norm": 0.8756514191627502, + "learning_rate": 3.851319171223349e-06, + "loss": 0.6206, + "step": 5506 + }, + { + "epoch": 1.9336376404494382, + "grad_norm": 0.8564157485961914, + "learning_rate": 3.850930181740826e-06, + "loss": 0.6062, + "step": 5507 + }, + { + "epoch": 1.9339887640449438, + "grad_norm": 0.8514534831047058, + "learning_rate": 3.8505411460576455e-06, + "loss": 0.5625, + "step": 5508 + }, + { + "epoch": 1.9343398876404494, + "grad_norm": 0.9018797874450684, + "learning_rate": 3.850152064187115e-06, + "loss": 0.6357, + "step": 5509 + }, + { + "epoch": 1.9346910112359552, + "grad_norm": 0.8431733250617981, + "learning_rate": 3.849762936142538e-06, + "loss": 0.6041, + "step": 5510 + }, + { + "epoch": 1.9350421348314608, + "grad_norm": 0.8911080360412598, + "learning_rate": 3.849373761937225e-06, + "loss": 0.6077, + "step": 5511 + }, + { + "epoch": 1.9353932584269664, + "grad_norm": 0.8887476921081543, + "learning_rate": 3.848984541584483e-06, + "loss": 0.6041, + "step": 5512 + }, + { + "epoch": 1.935744382022472, + "grad_norm": 0.8878927826881409, + "learning_rate": 3.848595275097624e-06, + "loss": 0.6169, + "step": 5513 + }, + { + "epoch": 1.9360955056179776, + "grad_norm": 0.9070707559585571, + "learning_rate": 3.848205962489961e-06, + "loss": 0.5982, + "step": 5514 + }, + { + "epoch": 1.9364466292134832, + "grad_norm": 1.2515376806259155, + "learning_rate": 3.847816603774807e-06, + "loss": 0.6475, + "step": 5515 + }, + { + "epoch": 1.9367977528089888, + "grad_norm": 0.8596694469451904, + "learning_rate": 3.84742719896548e-06, + "loss": 0.6293, + "step": 5516 + }, + { + "epoch": 1.9371488764044944, + "grad_norm": 0.9003183841705322, + "learning_rate": 3.847037748075294e-06, + "loss": 0.6235, + "step": 5517 + }, + { + "epoch": 1.9375, + "grad_norm": 0.8805832266807556, + "learning_rate": 3.846648251117571e-06, + "loss": 0.5892, + "step": 5518 + }, + { + "epoch": 1.9378511235955056, + "grad_norm": 0.8691025972366333, + "learning_rate": 3.846258708105629e-06, + "loss": 0.5987, + "step": 5519 + }, + { + "epoch": 1.9382022471910112, + "grad_norm": 0.8848376870155334, + "learning_rate": 3.8458691190527915e-06, + "loss": 0.5966, + "step": 5520 + }, + { + "epoch": 1.9385533707865168, + "grad_norm": 0.8585199117660522, + "learning_rate": 3.845479483972383e-06, + "loss": 0.5966, + "step": 5521 + }, + { + "epoch": 1.9389044943820224, + "grad_norm": 0.8973793983459473, + "learning_rate": 3.845089802877727e-06, + "loss": 0.6159, + "step": 5522 + }, + { + "epoch": 1.939255617977528, + "grad_norm": 0.8836932182312012, + "learning_rate": 3.84470007578215e-06, + "loss": 0.6267, + "step": 5523 + }, + { + "epoch": 1.9396067415730336, + "grad_norm": 0.861569344997406, + "learning_rate": 3.844310302698982e-06, + "loss": 0.6287, + "step": 5524 + }, + { + "epoch": 1.9399578651685392, + "grad_norm": 0.9185940027236938, + "learning_rate": 3.843920483641551e-06, + "loss": 0.6334, + "step": 5525 + }, + { + "epoch": 1.9403089887640448, + "grad_norm": 0.8938425779342651, + "learning_rate": 3.8435306186231905e-06, + "loss": 0.609, + "step": 5526 + }, + { + "epoch": 1.9406601123595506, + "grad_norm": 0.8964993357658386, + "learning_rate": 3.8431407076572316e-06, + "loss": 0.5996, + "step": 5527 + }, + { + "epoch": 1.9410112359550562, + "grad_norm": 0.8518646359443665, + "learning_rate": 3.8427507507570104e-06, + "loss": 0.5862, + "step": 5528 + }, + { + "epoch": 1.9413623595505618, + "grad_norm": 0.8727825880050659, + "learning_rate": 3.842360747935863e-06, + "loss": 0.627, + "step": 5529 + }, + { + "epoch": 1.9417134831460674, + "grad_norm": 0.8543184399604797, + "learning_rate": 3.841970699207125e-06, + "loss": 0.5737, + "step": 5530 + }, + { + "epoch": 1.942064606741573, + "grad_norm": 0.8790775537490845, + "learning_rate": 3.841580604584139e-06, + "loss": 0.5938, + "step": 5531 + }, + { + "epoch": 1.9424157303370788, + "grad_norm": 0.8977394104003906, + "learning_rate": 3.8411904640802435e-06, + "loss": 0.6028, + "step": 5532 + }, + { + "epoch": 1.9427668539325844, + "grad_norm": 0.8888459205627441, + "learning_rate": 3.8408002777087825e-06, + "loss": 0.5876, + "step": 5533 + }, + { + "epoch": 1.94311797752809, + "grad_norm": 0.8542887568473816, + "learning_rate": 3.840410045483098e-06, + "loss": 0.5762, + "step": 5534 + }, + { + "epoch": 1.9434691011235956, + "grad_norm": 0.9694353342056274, + "learning_rate": 3.8400197674165385e-06, + "loss": 0.6231, + "step": 5535 + }, + { + "epoch": 1.9438202247191012, + "grad_norm": 0.9098255634307861, + "learning_rate": 3.839629443522449e-06, + "loss": 0.6231, + "step": 5536 + }, + { + "epoch": 1.9441713483146068, + "grad_norm": 0.9094727635383606, + "learning_rate": 3.8392390738141784e-06, + "loss": 0.6039, + "step": 5537 + }, + { + "epoch": 1.9445224719101124, + "grad_norm": 0.8750001788139343, + "learning_rate": 3.838848658305079e-06, + "loss": 0.6298, + "step": 5538 + }, + { + "epoch": 1.944873595505618, + "grad_norm": 0.892129123210907, + "learning_rate": 3.8384581970085e-06, + "loss": 0.6056, + "step": 5539 + }, + { + "epoch": 1.9452247191011236, + "grad_norm": 0.8831581473350525, + "learning_rate": 3.8380676899377965e-06, + "loss": 0.6372, + "step": 5540 + }, + { + "epoch": 1.9455758426966292, + "grad_norm": 0.8667299151420593, + "learning_rate": 3.837677137106322e-06, + "loss": 0.621, + "step": 5541 + }, + { + "epoch": 1.9459269662921348, + "grad_norm": 0.8704750537872314, + "learning_rate": 3.837286538527435e-06, + "loss": 0.6052, + "step": 5542 + }, + { + "epoch": 1.9462780898876404, + "grad_norm": 0.9060012102127075, + "learning_rate": 3.836895894214492e-06, + "loss": 0.5917, + "step": 5543 + }, + { + "epoch": 1.946629213483146, + "grad_norm": 0.8708364963531494, + "learning_rate": 3.836505204180854e-06, + "loss": 0.6443, + "step": 5544 + }, + { + "epoch": 1.9469803370786516, + "grad_norm": 0.881945788860321, + "learning_rate": 3.836114468439882e-06, + "loss": 0.6284, + "step": 5545 + }, + { + "epoch": 1.9473314606741572, + "grad_norm": 0.8595880270004272, + "learning_rate": 3.835723687004938e-06, + "loss": 0.6034, + "step": 5546 + }, + { + "epoch": 1.9476825842696628, + "grad_norm": 0.8725127577781677, + "learning_rate": 3.8353328598893865e-06, + "loss": 0.6082, + "step": 5547 + }, + { + "epoch": 1.9480337078651684, + "grad_norm": 0.8611936569213867, + "learning_rate": 3.834941987106594e-06, + "loss": 0.5814, + "step": 5548 + }, + { + "epoch": 1.9483848314606742, + "grad_norm": 0.8612635731697083, + "learning_rate": 3.834551068669928e-06, + "loss": 0.5771, + "step": 5549 + }, + { + "epoch": 1.9487359550561798, + "grad_norm": 0.8790583610534668, + "learning_rate": 3.834160104592758e-06, + "loss": 0.5968, + "step": 5550 + }, + { + "epoch": 1.9490870786516854, + "grad_norm": 0.9288917183876038, + "learning_rate": 3.833769094888454e-06, + "loss": 0.6283, + "step": 5551 + }, + { + "epoch": 1.949438202247191, + "grad_norm": 0.8635450005531311, + "learning_rate": 3.833378039570388e-06, + "loss": 0.6024, + "step": 5552 + }, + { + "epoch": 1.9497893258426966, + "grad_norm": 0.9010553359985352, + "learning_rate": 3.8329869386519335e-06, + "loss": 0.5933, + "step": 5553 + }, + { + "epoch": 1.9501404494382022, + "grad_norm": 0.8576677441596985, + "learning_rate": 3.832595792146468e-06, + "loss": 0.6018, + "step": 5554 + }, + { + "epoch": 1.950491573033708, + "grad_norm": 0.9220290780067444, + "learning_rate": 3.832204600067365e-06, + "loss": 0.6292, + "step": 5555 + }, + { + "epoch": 1.9508426966292136, + "grad_norm": 0.8435717821121216, + "learning_rate": 3.831813362428005e-06, + "loss": 0.581, + "step": 5556 + }, + { + "epoch": 1.9511938202247192, + "grad_norm": 0.9167920351028442, + "learning_rate": 3.831422079241768e-06, + "loss": 0.621, + "step": 5557 + }, + { + "epoch": 1.9515449438202248, + "grad_norm": 0.8408527374267578, + "learning_rate": 3.831030750522034e-06, + "loss": 0.5921, + "step": 5558 + }, + { + "epoch": 1.9518960674157304, + "grad_norm": 0.8575167655944824, + "learning_rate": 3.830639376282189e-06, + "loss": 0.5961, + "step": 5559 + }, + { + "epoch": 1.952247191011236, + "grad_norm": 0.8758642673492432, + "learning_rate": 3.8302479565356154e-06, + "loss": 0.657, + "step": 5560 + }, + { + "epoch": 1.9525983146067416, + "grad_norm": 0.8745916485786438, + "learning_rate": 3.8298564912957e-06, + "loss": 0.6317, + "step": 5561 + }, + { + "epoch": 1.9529494382022472, + "grad_norm": 0.867461621761322, + "learning_rate": 3.829464980575831e-06, + "loss": 0.6211, + "step": 5562 + }, + { + "epoch": 1.9533005617977528, + "grad_norm": 0.8979873657226562, + "learning_rate": 3.829073424389396e-06, + "loss": 0.6253, + "step": 5563 + }, + { + "epoch": 1.9536516853932584, + "grad_norm": 0.8740954399108887, + "learning_rate": 3.828681822749789e-06, + "loss": 0.6189, + "step": 5564 + }, + { + "epoch": 1.954002808988764, + "grad_norm": 0.8818367123603821, + "learning_rate": 3.828290175670399e-06, + "loss": 0.6066, + "step": 5565 + }, + { + "epoch": 1.9543539325842696, + "grad_norm": 0.9065709114074707, + "learning_rate": 3.8278984831646225e-06, + "loss": 0.638, + "step": 5566 + }, + { + "epoch": 1.9547050561797752, + "grad_norm": 0.8510881066322327, + "learning_rate": 3.827506745245854e-06, + "loss": 0.587, + "step": 5567 + }, + { + "epoch": 1.9550561797752808, + "grad_norm": 0.8661137819290161, + "learning_rate": 3.827114961927491e-06, + "loss": 0.6402, + "step": 5568 + }, + { + "epoch": 1.9554073033707864, + "grad_norm": 0.8897769451141357, + "learning_rate": 3.8267231332229315e-06, + "loss": 0.6144, + "step": 5569 + }, + { + "epoch": 1.955758426966292, + "grad_norm": 0.9311615824699402, + "learning_rate": 3.826331259145576e-06, + "loss": 0.5961, + "step": 5570 + }, + { + "epoch": 1.9561095505617978, + "grad_norm": 0.8808618187904358, + "learning_rate": 3.825939339708827e-06, + "loss": 0.6238, + "step": 5571 + }, + { + "epoch": 1.9564606741573034, + "grad_norm": 0.8663437962532043, + "learning_rate": 3.825547374926087e-06, + "loss": 0.5774, + "step": 5572 + }, + { + "epoch": 1.956811797752809, + "grad_norm": 0.8695099353790283, + "learning_rate": 3.825155364810761e-06, + "loss": 0.6166, + "step": 5573 + }, + { + "epoch": 1.9571629213483146, + "grad_norm": 0.890457272529602, + "learning_rate": 3.824763309376256e-06, + "loss": 0.6004, + "step": 5574 + }, + { + "epoch": 1.9575140449438202, + "grad_norm": 0.8851537108421326, + "learning_rate": 3.8243712086359785e-06, + "loss": 0.5441, + "step": 5575 + }, + { + "epoch": 1.9578651685393258, + "grad_norm": 0.8680998086929321, + "learning_rate": 3.823979062603339e-06, + "loss": 0.634, + "step": 5576 + }, + { + "epoch": 1.9582162921348316, + "grad_norm": 0.8861505389213562, + "learning_rate": 3.823586871291749e-06, + "loss": 0.6083, + "step": 5577 + }, + { + "epoch": 1.9585674157303372, + "grad_norm": 0.853298544883728, + "learning_rate": 3.823194634714621e-06, + "loss": 0.6065, + "step": 5578 + }, + { + "epoch": 1.9589185393258428, + "grad_norm": 0.896482527256012, + "learning_rate": 3.822802352885367e-06, + "loss": 0.6085, + "step": 5579 + }, + { + "epoch": 1.9592696629213484, + "grad_norm": 0.8784202337265015, + "learning_rate": 3.8224100258174066e-06, + "loss": 0.6141, + "step": 5580 + }, + { + "epoch": 1.959620786516854, + "grad_norm": 0.8632230162620544, + "learning_rate": 3.8220176535241546e-06, + "loss": 0.6062, + "step": 5581 + }, + { + "epoch": 1.9599719101123596, + "grad_norm": 0.8777154684066772, + "learning_rate": 3.821625236019029e-06, + "loss": 0.5994, + "step": 5582 + }, + { + "epoch": 1.9603230337078652, + "grad_norm": 0.8801113963127136, + "learning_rate": 3.821232773315452e-06, + "loss": 0.6037, + "step": 5583 + }, + { + "epoch": 1.9606741573033708, + "grad_norm": 0.8997433185577393, + "learning_rate": 3.820840265426844e-06, + "loss": 0.641, + "step": 5584 + }, + { + "epoch": 1.9610252808988764, + "grad_norm": 0.9067327976226807, + "learning_rate": 3.82044771236663e-06, + "loss": 0.6045, + "step": 5585 + }, + { + "epoch": 1.961376404494382, + "grad_norm": 0.8806515336036682, + "learning_rate": 3.820055114148234e-06, + "loss": 0.6139, + "step": 5586 + }, + { + "epoch": 1.9617275280898876, + "grad_norm": 0.8809478282928467, + "learning_rate": 3.819662470785082e-06, + "loss": 0.6166, + "step": 5587 + }, + { + "epoch": 1.9620786516853932, + "grad_norm": 0.8271637558937073, + "learning_rate": 3.819269782290603e-06, + "loss": 0.5549, + "step": 5588 + }, + { + "epoch": 1.9624297752808988, + "grad_norm": 0.866726815700531, + "learning_rate": 3.8188770486782275e-06, + "loss": 0.5978, + "step": 5589 + }, + { + "epoch": 1.9627808988764044, + "grad_norm": 0.8871544003486633, + "learning_rate": 3.818484269961385e-06, + "loss": 0.6149, + "step": 5590 + }, + { + "epoch": 1.96313202247191, + "grad_norm": 0.9260693788528442, + "learning_rate": 3.818091446153508e-06, + "loss": 0.6066, + "step": 5591 + }, + { + "epoch": 1.9634831460674156, + "grad_norm": 0.867003858089447, + "learning_rate": 3.817698577268032e-06, + "loss": 0.6245, + "step": 5592 + }, + { + "epoch": 1.9638342696629212, + "grad_norm": 0.8828530311584473, + "learning_rate": 3.817305663318393e-06, + "loss": 0.6173, + "step": 5593 + }, + { + "epoch": 1.964185393258427, + "grad_norm": 0.8841493725776672, + "learning_rate": 3.8169127043180255e-06, + "loss": 0.5633, + "step": 5594 + }, + { + "epoch": 1.9645365168539326, + "grad_norm": 0.881924569606781, + "learning_rate": 3.816519700280372e-06, + "loss": 0.5844, + "step": 5595 + }, + { + "epoch": 1.9648876404494382, + "grad_norm": 0.8986069560050964, + "learning_rate": 3.816126651218871e-06, + "loss": 0.6792, + "step": 5596 + }, + { + "epoch": 1.9652387640449438, + "grad_norm": 0.8712230920791626, + "learning_rate": 3.815733557146964e-06, + "loss": 0.6117, + "step": 5597 + }, + { + "epoch": 1.9655898876404494, + "grad_norm": 0.9183915853500366, + "learning_rate": 3.815340418078096e-06, + "loss": 0.6554, + "step": 5598 + }, + { + "epoch": 1.9659410112359552, + "grad_norm": 0.8475895524024963, + "learning_rate": 3.814947234025711e-06, + "loss": 0.5581, + "step": 5599 + }, + { + "epoch": 1.9662921348314608, + "grad_norm": 0.881952702999115, + "learning_rate": 3.8145540050032558e-06, + "loss": 0.6173, + "step": 5600 + }, + { + "epoch": 1.9666432584269664, + "grad_norm": 0.8803291916847229, + "learning_rate": 3.814160731024179e-06, + "loss": 0.6322, + "step": 5601 + }, + { + "epoch": 1.966994382022472, + "grad_norm": 0.8611246347427368, + "learning_rate": 3.8137674121019296e-06, + "loss": 0.6011, + "step": 5602 + }, + { + "epoch": 1.9673455056179776, + "grad_norm": 0.8588524460792542, + "learning_rate": 3.8133740482499582e-06, + "loss": 0.614, + "step": 5603 + }, + { + "epoch": 1.9676966292134832, + "grad_norm": 0.872290313243866, + "learning_rate": 3.8129806394817186e-06, + "loss": 0.6002, + "step": 5604 + }, + { + "epoch": 1.9680477528089888, + "grad_norm": 0.892127513885498, + "learning_rate": 3.812587185810665e-06, + "loss": 0.616, + "step": 5605 + }, + { + "epoch": 1.9683988764044944, + "grad_norm": 0.8674480319023132, + "learning_rate": 3.8121936872502523e-06, + "loss": 0.6111, + "step": 5606 + }, + { + "epoch": 1.96875, + "grad_norm": 0.8589068651199341, + "learning_rate": 3.8118001438139384e-06, + "loss": 0.5768, + "step": 5607 + }, + { + "epoch": 1.9691011235955056, + "grad_norm": 0.8877671957015991, + "learning_rate": 3.8114065555151826e-06, + "loss": 0.6497, + "step": 5608 + }, + { + "epoch": 1.9694522471910112, + "grad_norm": 0.8697142004966736, + "learning_rate": 3.8110129223674443e-06, + "loss": 0.6118, + "step": 5609 + }, + { + "epoch": 1.9698033707865168, + "grad_norm": 0.8770889043807983, + "learning_rate": 3.810619244384186e-06, + "loss": 0.6308, + "step": 5610 + }, + { + "epoch": 1.9701544943820224, + "grad_norm": 0.8840969204902649, + "learning_rate": 3.8102255215788713e-06, + "loss": 0.5886, + "step": 5611 + }, + { + "epoch": 1.970505617977528, + "grad_norm": 0.8892825841903687, + "learning_rate": 3.8098317539649656e-06, + "loss": 0.6546, + "step": 5612 + }, + { + "epoch": 1.9708567415730336, + "grad_norm": 0.8668626546859741, + "learning_rate": 3.8094379415559333e-06, + "loss": 0.6218, + "step": 5613 + }, + { + "epoch": 1.9712078651685392, + "grad_norm": 0.8517277836799622, + "learning_rate": 3.8090440843652454e-06, + "loss": 0.6075, + "step": 5614 + }, + { + "epoch": 1.9715589887640448, + "grad_norm": 0.9507593512535095, + "learning_rate": 3.8086501824063687e-06, + "loss": 0.6418, + "step": 5615 + }, + { + "epoch": 1.9719101123595506, + "grad_norm": 0.8373234868049622, + "learning_rate": 3.808256235692776e-06, + "loss": 0.6001, + "step": 5616 + }, + { + "epoch": 1.9722612359550562, + "grad_norm": 0.8520222306251526, + "learning_rate": 3.80786224423794e-06, + "loss": 0.5982, + "step": 5617 + }, + { + "epoch": 1.9726123595505618, + "grad_norm": 0.873327374458313, + "learning_rate": 3.807468208055334e-06, + "loss": 0.6114, + "step": 5618 + }, + { + "epoch": 1.9729634831460674, + "grad_norm": 0.8597673773765564, + "learning_rate": 3.807074127158434e-06, + "loss": 0.5953, + "step": 5619 + }, + { + "epoch": 1.973314606741573, + "grad_norm": 0.8622745275497437, + "learning_rate": 3.8066800015607173e-06, + "loss": 0.6354, + "step": 5620 + }, + { + "epoch": 1.9736657303370788, + "grad_norm": 0.868476390838623, + "learning_rate": 3.8062858312756625e-06, + "loss": 0.6406, + "step": 5621 + }, + { + "epoch": 1.9740168539325844, + "grad_norm": 0.8725282549858093, + "learning_rate": 3.8058916163167504e-06, + "loss": 0.6252, + "step": 5622 + }, + { + "epoch": 1.97436797752809, + "grad_norm": 0.8920749425888062, + "learning_rate": 3.805497356697462e-06, + "loss": 0.6377, + "step": 5623 + }, + { + "epoch": 1.9747191011235956, + "grad_norm": 0.8895111083984375, + "learning_rate": 3.805103052431282e-06, + "loss": 0.5911, + "step": 5624 + }, + { + "epoch": 1.9750702247191012, + "grad_norm": 0.868254542350769, + "learning_rate": 3.804708703531694e-06, + "loss": 0.6213, + "step": 5625 + }, + { + "epoch": 1.9754213483146068, + "grad_norm": 0.8568325638771057, + "learning_rate": 3.8043143100121847e-06, + "loss": 0.6014, + "step": 5626 + }, + { + "epoch": 1.9757724719101124, + "grad_norm": 0.8779681921005249, + "learning_rate": 3.8039198718862424e-06, + "loss": 0.6314, + "step": 5627 + }, + { + "epoch": 1.976123595505618, + "grad_norm": 0.8798127174377441, + "learning_rate": 3.803525389167356e-06, + "loss": 0.6109, + "step": 5628 + }, + { + "epoch": 1.9764747191011236, + "grad_norm": 0.8850237131118774, + "learning_rate": 3.803130861869017e-06, + "loss": 0.6351, + "step": 5629 + }, + { + "epoch": 1.9768258426966292, + "grad_norm": 0.8631384372711182, + "learning_rate": 3.8027362900047177e-06, + "loss": 0.63, + "step": 5630 + }, + { + "epoch": 1.9771769662921348, + "grad_norm": 0.8860955834388733, + "learning_rate": 3.8023416735879513e-06, + "loss": 0.6042, + "step": 5631 + }, + { + "epoch": 1.9775280898876404, + "grad_norm": 0.8704652190208435, + "learning_rate": 3.8019470126322146e-06, + "loss": 0.6089, + "step": 5632 + }, + { + "epoch": 1.977879213483146, + "grad_norm": 0.8756489157676697, + "learning_rate": 3.801552307151004e-06, + "loss": 0.6121, + "step": 5633 + }, + { + "epoch": 1.9782303370786516, + "grad_norm": 0.8789172768592834, + "learning_rate": 3.8011575571578186e-06, + "loss": 0.5878, + "step": 5634 + }, + { + "epoch": 1.9785814606741572, + "grad_norm": 0.8992910385131836, + "learning_rate": 3.800762762666158e-06, + "loss": 0.6156, + "step": 5635 + }, + { + "epoch": 1.9789325842696628, + "grad_norm": 0.8658064603805542, + "learning_rate": 3.800367923689524e-06, + "loss": 0.6162, + "step": 5636 + }, + { + "epoch": 1.9792837078651684, + "grad_norm": 0.861645519733429, + "learning_rate": 3.7999730402414193e-06, + "loss": 0.572, + "step": 5637 + }, + { + "epoch": 1.9796348314606742, + "grad_norm": 0.8935385346412659, + "learning_rate": 3.79957811233535e-06, + "loss": 0.6163, + "step": 5638 + }, + { + "epoch": 1.9799859550561798, + "grad_norm": 0.8381475210189819, + "learning_rate": 3.799183139984821e-06, + "loss": 0.5775, + "step": 5639 + }, + { + "epoch": 1.9803370786516854, + "grad_norm": 0.8763298988342285, + "learning_rate": 3.7987881232033397e-06, + "loss": 0.6381, + "step": 5640 + }, + { + "epoch": 1.980688202247191, + "grad_norm": 0.8951460123062134, + "learning_rate": 3.7983930620044164e-06, + "loss": 0.6457, + "step": 5641 + }, + { + "epoch": 1.9810393258426966, + "grad_norm": 0.8614299893379211, + "learning_rate": 3.797997956401561e-06, + "loss": 0.5989, + "step": 5642 + }, + { + "epoch": 1.9813904494382022, + "grad_norm": 0.8521272540092468, + "learning_rate": 3.797602806408287e-06, + "loss": 0.6016, + "step": 5643 + }, + { + "epoch": 1.981741573033708, + "grad_norm": 0.8658344745635986, + "learning_rate": 3.7972076120381075e-06, + "loss": 0.6222, + "step": 5644 + }, + { + "epoch": 1.9820926966292136, + "grad_norm": 0.8566646575927734, + "learning_rate": 3.796812373304537e-06, + "loss": 0.6105, + "step": 5645 + }, + { + "epoch": 1.9824438202247192, + "grad_norm": 0.8909837007522583, + "learning_rate": 3.7964170902210927e-06, + "loss": 0.6047, + "step": 5646 + }, + { + "epoch": 1.9827949438202248, + "grad_norm": 0.8529945611953735, + "learning_rate": 3.796021762801294e-06, + "loss": 0.6196, + "step": 5647 + }, + { + "epoch": 1.9831460674157304, + "grad_norm": 0.878425657749176, + "learning_rate": 3.79562639105866e-06, + "loss": 0.5868, + "step": 5648 + }, + { + "epoch": 1.983497191011236, + "grad_norm": 0.8768588900566101, + "learning_rate": 3.795230975006712e-06, + "loss": 0.5811, + "step": 5649 + }, + { + "epoch": 1.9838483146067416, + "grad_norm": 0.8540440201759338, + "learning_rate": 3.794835514658972e-06, + "loss": 0.6095, + "step": 5650 + }, + { + "epoch": 1.9841994382022472, + "grad_norm": 0.8583154082298279, + "learning_rate": 3.794440010028967e-06, + "loss": 0.5874, + "step": 5651 + }, + { + "epoch": 1.9845505617977528, + "grad_norm": 0.8400346040725708, + "learning_rate": 3.7940444611302206e-06, + "loss": 0.5943, + "step": 5652 + }, + { + "epoch": 1.9849016853932584, + "grad_norm": 0.8862133026123047, + "learning_rate": 3.7936488679762617e-06, + "loss": 0.5951, + "step": 5653 + }, + { + "epoch": 1.985252808988764, + "grad_norm": 0.9019565582275391, + "learning_rate": 3.7932532305806175e-06, + "loss": 0.6084, + "step": 5654 + }, + { + "epoch": 1.9856039325842696, + "grad_norm": 0.8929518461227417, + "learning_rate": 3.7928575489568197e-06, + "loss": 0.589, + "step": 5655 + }, + { + "epoch": 1.9859550561797752, + "grad_norm": 0.8899116516113281, + "learning_rate": 3.7924618231184e-06, + "loss": 0.6176, + "step": 5656 + }, + { + "epoch": 1.9863061797752808, + "grad_norm": 0.8458961844444275, + "learning_rate": 3.7920660530788923e-06, + "loss": 0.5669, + "step": 5657 + }, + { + "epoch": 1.9866573033707864, + "grad_norm": 0.855290412902832, + "learning_rate": 3.7916702388518313e-06, + "loss": 0.5848, + "step": 5658 + }, + { + "epoch": 1.987008426966292, + "grad_norm": 0.8671900629997253, + "learning_rate": 3.7912743804507522e-06, + "loss": 0.5838, + "step": 5659 + }, + { + "epoch": 1.9873595505617978, + "grad_norm": 0.8991999626159668, + "learning_rate": 3.7908784778891954e-06, + "loss": 0.6014, + "step": 5660 + }, + { + "epoch": 1.9877106741573034, + "grad_norm": 0.8997818827629089, + "learning_rate": 3.7904825311807e-06, + "loss": 0.6297, + "step": 5661 + }, + { + "epoch": 1.988061797752809, + "grad_norm": 0.8679925203323364, + "learning_rate": 3.7900865403388054e-06, + "loss": 0.6217, + "step": 5662 + }, + { + "epoch": 1.9884129213483146, + "grad_norm": 0.8854179382324219, + "learning_rate": 3.789690505377055e-06, + "loss": 0.592, + "step": 5663 + }, + { + "epoch": 1.9887640449438202, + "grad_norm": 0.8907507658004761, + "learning_rate": 3.7892944263089927e-06, + "loss": 0.6142, + "step": 5664 + }, + { + "epoch": 1.9891151685393258, + "grad_norm": 1.1212937831878662, + "learning_rate": 3.788898303148164e-06, + "loss": 0.6249, + "step": 5665 + }, + { + "epoch": 1.9894662921348316, + "grad_norm": 0.8652292490005493, + "learning_rate": 3.788502135908117e-06, + "loss": 0.6164, + "step": 5666 + }, + { + "epoch": 1.9898174157303372, + "grad_norm": 0.8923054337501526, + "learning_rate": 3.7881059246023993e-06, + "loss": 0.6409, + "step": 5667 + }, + { + "epoch": 1.9901685393258428, + "grad_norm": 0.8936302065849304, + "learning_rate": 3.7877096692445613e-06, + "loss": 0.6208, + "step": 5668 + }, + { + "epoch": 1.9905196629213484, + "grad_norm": 0.8705040216445923, + "learning_rate": 3.7873133698481547e-06, + "loss": 0.6326, + "step": 5669 + }, + { + "epoch": 1.990870786516854, + "grad_norm": 0.8782395124435425, + "learning_rate": 3.7869170264267325e-06, + "loss": 0.6189, + "step": 5670 + }, + { + "epoch": 1.9912219101123596, + "grad_norm": 0.8925029635429382, + "learning_rate": 3.786520638993849e-06, + "loss": 0.6151, + "step": 5671 + }, + { + "epoch": 1.9915730337078652, + "grad_norm": 0.8680460453033447, + "learning_rate": 3.78612420756306e-06, + "loss": 0.6144, + "step": 5672 + }, + { + "epoch": 1.9919241573033708, + "grad_norm": 0.859771728515625, + "learning_rate": 3.785727732147925e-06, + "loss": 0.623, + "step": 5673 + }, + { + "epoch": 1.9922752808988764, + "grad_norm": 0.8385575413703918, + "learning_rate": 3.7853312127620003e-06, + "loss": 0.5991, + "step": 5674 + }, + { + "epoch": 1.992626404494382, + "grad_norm": 0.8689984083175659, + "learning_rate": 3.7849346494188487e-06, + "loss": 0.5658, + "step": 5675 + }, + { + "epoch": 1.9929775280898876, + "grad_norm": 0.8608708381652832, + "learning_rate": 3.784538042132031e-06, + "loss": 0.6215, + "step": 5676 + }, + { + "epoch": 1.9933286516853932, + "grad_norm": 0.884060800075531, + "learning_rate": 3.7841413909151127e-06, + "loss": 0.6248, + "step": 5677 + }, + { + "epoch": 1.9936797752808988, + "grad_norm": 0.86598801612854, + "learning_rate": 3.783744695781657e-06, + "loss": 0.5958, + "step": 5678 + }, + { + "epoch": 1.9940308988764044, + "grad_norm": 0.8934966325759888, + "learning_rate": 3.7833479567452308e-06, + "loss": 0.6226, + "step": 5679 + }, + { + "epoch": 1.99438202247191, + "grad_norm": 0.8695813417434692, + "learning_rate": 3.782951173819403e-06, + "loss": 0.59, + "step": 5680 + }, + { + "epoch": 1.9947331460674156, + "grad_norm": 0.9348827004432678, + "learning_rate": 3.782554347017743e-06, + "loss": 0.6402, + "step": 5681 + }, + { + "epoch": 1.9950842696629212, + "grad_norm": 0.8460621237754822, + "learning_rate": 3.7821574763538216e-06, + "loss": 0.5891, + "step": 5682 + }, + { + "epoch": 1.995435393258427, + "grad_norm": 0.8422776460647583, + "learning_rate": 3.7817605618412125e-06, + "loss": 0.6031, + "step": 5683 + }, + { + "epoch": 1.9957865168539326, + "grad_norm": 0.87843257188797, + "learning_rate": 3.7813636034934875e-06, + "loss": 0.6175, + "step": 5684 + }, + { + "epoch": 1.9961376404494382, + "grad_norm": 0.9025244116783142, + "learning_rate": 3.7809666013242245e-06, + "loss": 0.6066, + "step": 5685 + }, + { + "epoch": 1.9964887640449438, + "grad_norm": 0.8787137866020203, + "learning_rate": 3.7805695553470002e-06, + "loss": 0.62, + "step": 5686 + }, + { + "epoch": 1.9968398876404494, + "grad_norm": 0.8768424391746521, + "learning_rate": 3.7801724655753922e-06, + "loss": 0.6046, + "step": 5687 + }, + { + "epoch": 1.9971910112359552, + "grad_norm": 0.8371465802192688, + "learning_rate": 3.7797753320229822e-06, + "loss": 0.582, + "step": 5688 + }, + { + "epoch": 1.9975421348314608, + "grad_norm": 0.8751241564750671, + "learning_rate": 3.7793781547033503e-06, + "loss": 0.597, + "step": 5689 + }, + { + "epoch": 1.9978932584269664, + "grad_norm": 0.8778484463691711, + "learning_rate": 3.778980933630081e-06, + "loss": 0.6299, + "step": 5690 + }, + { + "epoch": 1.998244382022472, + "grad_norm": 0.8325814008712769, + "learning_rate": 3.7785836688167576e-06, + "loss": 0.5842, + "step": 5691 + }, + { + "epoch": 1.9985955056179776, + "grad_norm": 0.8934423923492432, + "learning_rate": 3.778186360276967e-06, + "loss": 0.6192, + "step": 5692 + }, + { + "epoch": 1.9989466292134832, + "grad_norm": 0.8526570796966553, + "learning_rate": 3.7777890080242963e-06, + "loss": 0.5843, + "step": 5693 + }, + { + "epoch": 1.9992977528089888, + "grad_norm": 0.896981418132782, + "learning_rate": 3.7773916120723354e-06, + "loss": 0.651, + "step": 5694 + }, + { + "epoch": 1.9996488764044944, + "grad_norm": 0.8761165738105774, + "learning_rate": 3.776994172434675e-06, + "loss": 0.649, + "step": 5695 + }, + { + "epoch": 2.0, + "grad_norm": 0.8966251611709595, + "learning_rate": 3.776596689124905e-06, + "loss": 0.5836, + "step": 5696 + }, + { + "epoch": 2.0003511235955056, + "grad_norm": 0.8626259565353394, + "learning_rate": 3.776199162156622e-06, + "loss": 0.5788, + "step": 5697 + }, + { + "epoch": 2.000702247191011, + "grad_norm": 0.8231014013290405, + "learning_rate": 3.775801591543419e-06, + "loss": 0.5541, + "step": 5698 + }, + { + "epoch": 2.001053370786517, + "grad_norm": 0.8400091528892517, + "learning_rate": 3.7754039772988936e-06, + "loss": 0.5942, + "step": 5699 + }, + { + "epoch": 2.0014044943820224, + "grad_norm": 0.8540766835212708, + "learning_rate": 3.775006319436644e-06, + "loss": 0.5816, + "step": 5700 + }, + { + "epoch": 2.001755617977528, + "grad_norm": 0.8629932403564453, + "learning_rate": 3.7746086179702683e-06, + "loss": 0.5508, + "step": 5701 + }, + { + "epoch": 2.0021067415730336, + "grad_norm": 0.8767760992050171, + "learning_rate": 3.77421087291337e-06, + "loss": 0.6141, + "step": 5702 + }, + { + "epoch": 2.002457865168539, + "grad_norm": 0.9155377745628357, + "learning_rate": 3.773813084279549e-06, + "loss": 0.5655, + "step": 5703 + }, + { + "epoch": 2.002808988764045, + "grad_norm": 0.901252806186676, + "learning_rate": 3.773415252082411e-06, + "loss": 0.5813, + "step": 5704 + }, + { + "epoch": 2.0031601123595504, + "grad_norm": 0.8925053477287292, + "learning_rate": 3.7730173763355606e-06, + "loss": 0.5966, + "step": 5705 + }, + { + "epoch": 2.003511235955056, + "grad_norm": 0.8810111284255981, + "learning_rate": 3.7726194570526053e-06, + "loss": 0.5947, + "step": 5706 + }, + { + "epoch": 2.0038623595505616, + "grad_norm": 0.9047281742095947, + "learning_rate": 3.7722214942471535e-06, + "loss": 0.606, + "step": 5707 + }, + { + "epoch": 2.0042134831460676, + "grad_norm": 0.866055428981781, + "learning_rate": 3.7718234879328164e-06, + "loss": 0.574, + "step": 5708 + }, + { + "epoch": 2.004564606741573, + "grad_norm": 0.8605414032936096, + "learning_rate": 3.771425438123203e-06, + "loss": 0.58, + "step": 5709 + }, + { + "epoch": 2.004915730337079, + "grad_norm": 0.88720703125, + "learning_rate": 3.771027344831928e-06, + "loss": 0.588, + "step": 5710 + }, + { + "epoch": 2.0052668539325844, + "grad_norm": 0.8635183572769165, + "learning_rate": 3.770629208072606e-06, + "loss": 0.5687, + "step": 5711 + }, + { + "epoch": 2.00561797752809, + "grad_norm": 0.8983670473098755, + "learning_rate": 3.7702310278588515e-06, + "loss": 0.592, + "step": 5712 + }, + { + "epoch": 2.0059691011235956, + "grad_norm": 0.902247428894043, + "learning_rate": 3.769832804204283e-06, + "loss": 0.5914, + "step": 5713 + }, + { + "epoch": 2.006320224719101, + "grad_norm": 0.877448320388794, + "learning_rate": 3.769434537122519e-06, + "loss": 0.5878, + "step": 5714 + }, + { + "epoch": 2.006671348314607, + "grad_norm": 0.8857719302177429, + "learning_rate": 3.769036226627182e-06, + "loss": 0.5602, + "step": 5715 + }, + { + "epoch": 2.0070224719101124, + "grad_norm": 0.8802056312561035, + "learning_rate": 3.7686378727318902e-06, + "loss": 0.5835, + "step": 5716 + }, + { + "epoch": 2.007373595505618, + "grad_norm": 0.8634882569313049, + "learning_rate": 3.7682394754502687e-06, + "loss": 0.5729, + "step": 5717 + }, + { + "epoch": 2.0077247191011236, + "grad_norm": 0.8762107491493225, + "learning_rate": 3.7678410347959426e-06, + "loss": 0.557, + "step": 5718 + }, + { + "epoch": 2.008075842696629, + "grad_norm": 0.8637248873710632, + "learning_rate": 3.7674425507825387e-06, + "loss": 0.5741, + "step": 5719 + }, + { + "epoch": 2.008426966292135, + "grad_norm": 0.884354829788208, + "learning_rate": 3.767044023423683e-06, + "loss": 0.5826, + "step": 5720 + }, + { + "epoch": 2.0087780898876404, + "grad_norm": 0.8706732392311096, + "learning_rate": 3.7666454527330064e-06, + "loss": 0.5674, + "step": 5721 + }, + { + "epoch": 2.009129213483146, + "grad_norm": 0.8681031465530396, + "learning_rate": 3.766246838724139e-06, + "loss": 0.5622, + "step": 5722 + }, + { + "epoch": 2.0094803370786516, + "grad_norm": 0.8685696721076965, + "learning_rate": 3.7658481814107138e-06, + "loss": 0.5824, + "step": 5723 + }, + { + "epoch": 2.009831460674157, + "grad_norm": 0.8874221444129944, + "learning_rate": 3.7654494808063643e-06, + "loss": 0.6057, + "step": 5724 + }, + { + "epoch": 2.010182584269663, + "grad_norm": 0.9122766852378845, + "learning_rate": 3.7650507369247247e-06, + "loss": 0.5942, + "step": 5725 + }, + { + "epoch": 2.0105337078651684, + "grad_norm": 0.8845071196556091, + "learning_rate": 3.764651949779432e-06, + "loss": 0.5834, + "step": 5726 + }, + { + "epoch": 2.010884831460674, + "grad_norm": 0.8612940311431885, + "learning_rate": 3.764253119384126e-06, + "loss": 0.5781, + "step": 5727 + }, + { + "epoch": 2.0112359550561796, + "grad_norm": 0.8638239502906799, + "learning_rate": 3.763854245752445e-06, + "loss": 0.579, + "step": 5728 + }, + { + "epoch": 2.011587078651685, + "grad_norm": 0.8957563638687134, + "learning_rate": 3.763455328898029e-06, + "loss": 0.6285, + "step": 5729 + }, + { + "epoch": 2.011938202247191, + "grad_norm": 0.8769227266311646, + "learning_rate": 3.7630563688345233e-06, + "loss": 0.5862, + "step": 5730 + }, + { + "epoch": 2.012289325842697, + "grad_norm": 0.8635375499725342, + "learning_rate": 3.76265736557557e-06, + "loss": 0.6063, + "step": 5731 + }, + { + "epoch": 2.0126404494382024, + "grad_norm": 0.8824431896209717, + "learning_rate": 3.762258319134815e-06, + "loss": 0.6195, + "step": 5732 + }, + { + "epoch": 2.012991573033708, + "grad_norm": 0.869091808795929, + "learning_rate": 3.761859229525906e-06, + "loss": 0.5701, + "step": 5733 + }, + { + "epoch": 2.0133426966292136, + "grad_norm": 0.8852167129516602, + "learning_rate": 3.7614600967624914e-06, + "loss": 0.6121, + "step": 5734 + }, + { + "epoch": 2.013693820224719, + "grad_norm": 0.8769776225090027, + "learning_rate": 3.7610609208582205e-06, + "loss": 0.5714, + "step": 5735 + }, + { + "epoch": 2.014044943820225, + "grad_norm": 0.860116183757782, + "learning_rate": 3.760661701826745e-06, + "loss": 0.5848, + "step": 5736 + }, + { + "epoch": 2.0143960674157304, + "grad_norm": 0.8777115345001221, + "learning_rate": 3.7602624396817184e-06, + "loss": 0.6069, + "step": 5737 + }, + { + "epoch": 2.014747191011236, + "grad_norm": 0.8745265603065491, + "learning_rate": 3.7598631344367947e-06, + "loss": 0.595, + "step": 5738 + }, + { + "epoch": 2.0150983146067416, + "grad_norm": 0.911849856376648, + "learning_rate": 3.75946378610563e-06, + "loss": 0.5913, + "step": 5739 + }, + { + "epoch": 2.015449438202247, + "grad_norm": 0.8881710767745972, + "learning_rate": 3.759064394701882e-06, + "loss": 0.5995, + "step": 5740 + }, + { + "epoch": 2.015800561797753, + "grad_norm": 0.9085574746131897, + "learning_rate": 3.7586649602392085e-06, + "loss": 0.604, + "step": 5741 + }, + { + "epoch": 2.0161516853932584, + "grad_norm": 0.894646942615509, + "learning_rate": 3.7582654827312702e-06, + "loss": 0.6295, + "step": 5742 + }, + { + "epoch": 2.016502808988764, + "grad_norm": 0.8959636092185974, + "learning_rate": 3.7578659621917298e-06, + "loss": 0.5823, + "step": 5743 + }, + { + "epoch": 2.0168539325842696, + "grad_norm": 0.84996497631073, + "learning_rate": 3.7574663986342493e-06, + "loss": 0.5906, + "step": 5744 + }, + { + "epoch": 2.017205056179775, + "grad_norm": 0.8966703414916992, + "learning_rate": 3.757066792072494e-06, + "loss": 0.6209, + "step": 5745 + }, + { + "epoch": 2.017556179775281, + "grad_norm": 0.8766224980354309, + "learning_rate": 3.7566671425201307e-06, + "loss": 0.5937, + "step": 5746 + }, + { + "epoch": 2.0179073033707864, + "grad_norm": 0.8875607848167419, + "learning_rate": 3.756267449990826e-06, + "loss": 0.5911, + "step": 5747 + }, + { + "epoch": 2.018258426966292, + "grad_norm": 0.8710358142852783, + "learning_rate": 3.7558677144982492e-06, + "loss": 0.5879, + "step": 5748 + }, + { + "epoch": 2.0186095505617976, + "grad_norm": 0.8919982314109802, + "learning_rate": 3.755467936056072e-06, + "loss": 0.5837, + "step": 5749 + }, + { + "epoch": 2.018960674157303, + "grad_norm": 0.8751034140586853, + "learning_rate": 3.7550681146779656e-06, + "loss": 0.5897, + "step": 5750 + }, + { + "epoch": 2.019311797752809, + "grad_norm": 0.9065902233123779, + "learning_rate": 3.7546682503776027e-06, + "loss": 0.5722, + "step": 5751 + }, + { + "epoch": 2.019662921348315, + "grad_norm": 0.9091061949729919, + "learning_rate": 3.7542683431686598e-06, + "loss": 0.6093, + "step": 5752 + }, + { + "epoch": 2.0200140449438204, + "grad_norm": 0.8719105124473572, + "learning_rate": 3.7538683930648135e-06, + "loss": 0.5762, + "step": 5753 + }, + { + "epoch": 2.020365168539326, + "grad_norm": 0.9036155939102173, + "learning_rate": 3.7534684000797404e-06, + "loss": 0.603, + "step": 5754 + }, + { + "epoch": 2.0207162921348316, + "grad_norm": 0.9013174772262573, + "learning_rate": 3.753068364227121e-06, + "loss": 0.5949, + "step": 5755 + }, + { + "epoch": 2.021067415730337, + "grad_norm": 0.8782882690429688, + "learning_rate": 3.7526682855206353e-06, + "loss": 0.6021, + "step": 5756 + }, + { + "epoch": 2.021418539325843, + "grad_norm": 0.876643180847168, + "learning_rate": 3.7522681639739666e-06, + "loss": 0.602, + "step": 5757 + }, + { + "epoch": 2.0217696629213484, + "grad_norm": 0.9358587861061096, + "learning_rate": 3.7518679996007984e-06, + "loss": 0.6086, + "step": 5758 + }, + { + "epoch": 2.022120786516854, + "grad_norm": 0.9080302715301514, + "learning_rate": 3.751467792414816e-06, + "loss": 0.5863, + "step": 5759 + }, + { + "epoch": 2.0224719101123596, + "grad_norm": 0.9168391227722168, + "learning_rate": 3.751067542429705e-06, + "loss": 0.615, + "step": 5760 + }, + { + "epoch": 2.022823033707865, + "grad_norm": 0.8669367432594299, + "learning_rate": 3.750667249659155e-06, + "loss": 0.5801, + "step": 5761 + }, + { + "epoch": 2.023174157303371, + "grad_norm": 0.8877193331718445, + "learning_rate": 3.7502669141168558e-06, + "loss": 0.6102, + "step": 5762 + }, + { + "epoch": 2.0235252808988764, + "grad_norm": 0.8779431581497192, + "learning_rate": 3.7498665358164975e-06, + "loss": 0.5669, + "step": 5763 + }, + { + "epoch": 2.023876404494382, + "grad_norm": 0.8888627290725708, + "learning_rate": 3.749466114771773e-06, + "loss": 0.5751, + "step": 5764 + }, + { + "epoch": 2.0242275280898876, + "grad_norm": 0.9005616903305054, + "learning_rate": 3.7490656509963775e-06, + "loss": 0.5913, + "step": 5765 + }, + { + "epoch": 2.024578651685393, + "grad_norm": 0.9011760950088501, + "learning_rate": 3.748665144504005e-06, + "loss": 0.61, + "step": 5766 + }, + { + "epoch": 2.024929775280899, + "grad_norm": 0.877354085445404, + "learning_rate": 3.7482645953083528e-06, + "loss": 0.5892, + "step": 5767 + }, + { + "epoch": 2.0252808988764044, + "grad_norm": 0.9336481690406799, + "learning_rate": 3.74786400342312e-06, + "loss": 0.5569, + "step": 5768 + }, + { + "epoch": 2.02563202247191, + "grad_norm": 0.9085440635681152, + "learning_rate": 3.747463368862006e-06, + "loss": 0.5844, + "step": 5769 + }, + { + "epoch": 2.0259831460674156, + "grad_norm": 0.9236464500427246, + "learning_rate": 3.747062691638712e-06, + "loss": 0.6144, + "step": 5770 + }, + { + "epoch": 2.026334269662921, + "grad_norm": 0.9236299991607666, + "learning_rate": 3.7466619717669416e-06, + "loss": 0.6021, + "step": 5771 + }, + { + "epoch": 2.026685393258427, + "grad_norm": 0.8903847932815552, + "learning_rate": 3.7462612092603985e-06, + "loss": 0.574, + "step": 5772 + }, + { + "epoch": 2.0270365168539324, + "grad_norm": 0.9148250222206116, + "learning_rate": 3.745860404132788e-06, + "loss": 0.5764, + "step": 5773 + }, + { + "epoch": 2.0273876404494384, + "grad_norm": 0.9070294499397278, + "learning_rate": 3.745459556397818e-06, + "loss": 0.5652, + "step": 5774 + }, + { + "epoch": 2.027738764044944, + "grad_norm": 0.8650291562080383, + "learning_rate": 3.745058666069197e-06, + "loss": 0.5799, + "step": 5775 + }, + { + "epoch": 2.0280898876404496, + "grad_norm": 0.8774101138114929, + "learning_rate": 3.744657733160635e-06, + "loss": 0.5811, + "step": 5776 + }, + { + "epoch": 2.028441011235955, + "grad_norm": 0.8710494041442871, + "learning_rate": 3.744256757685843e-06, + "loss": 0.5532, + "step": 5777 + }, + { + "epoch": 2.028792134831461, + "grad_norm": 0.8983147740364075, + "learning_rate": 3.743855739658536e-06, + "loss": 0.5701, + "step": 5778 + }, + { + "epoch": 2.0291432584269664, + "grad_norm": 0.8525742292404175, + "learning_rate": 3.743454679092426e-06, + "loss": 0.5487, + "step": 5779 + }, + { + "epoch": 2.029494382022472, + "grad_norm": 0.8792864680290222, + "learning_rate": 3.74305357600123e-06, + "loss": 0.6004, + "step": 5780 + }, + { + "epoch": 2.0298455056179776, + "grad_norm": 0.8998715877532959, + "learning_rate": 3.7426524303986656e-06, + "loss": 0.5724, + "step": 5781 + }, + { + "epoch": 2.030196629213483, + "grad_norm": 0.8908255696296692, + "learning_rate": 3.7422512422984513e-06, + "loss": 0.5714, + "step": 5782 + }, + { + "epoch": 2.030547752808989, + "grad_norm": 0.9007064700126648, + "learning_rate": 3.741850011714308e-06, + "loss": 0.5826, + "step": 5783 + }, + { + "epoch": 2.0308988764044944, + "grad_norm": 0.8781507015228271, + "learning_rate": 3.7414487386599567e-06, + "loss": 0.5567, + "step": 5784 + }, + { + "epoch": 2.03125, + "grad_norm": 0.8928130865097046, + "learning_rate": 3.741047423149121e-06, + "loss": 0.5773, + "step": 5785 + }, + { + "epoch": 2.0316011235955056, + "grad_norm": 0.8740100264549255, + "learning_rate": 3.7406460651955252e-06, + "loss": 0.5651, + "step": 5786 + }, + { + "epoch": 2.031952247191011, + "grad_norm": 0.8849120140075684, + "learning_rate": 3.7402446648128954e-06, + "loss": 0.5676, + "step": 5787 + }, + { + "epoch": 2.032303370786517, + "grad_norm": 0.8706261515617371, + "learning_rate": 3.7398432220149595e-06, + "loss": 0.589, + "step": 5788 + }, + { + "epoch": 2.0326544943820224, + "grad_norm": 0.8797366619110107, + "learning_rate": 3.7394417368154466e-06, + "loss": 0.5761, + "step": 5789 + }, + { + "epoch": 2.033005617977528, + "grad_norm": 0.8794291615486145, + "learning_rate": 3.7390402092280875e-06, + "loss": 0.6359, + "step": 5790 + }, + { + "epoch": 2.0333567415730336, + "grad_norm": 0.9492156505584717, + "learning_rate": 3.7386386392666123e-06, + "loss": 0.599, + "step": 5791 + }, + { + "epoch": 2.033707865168539, + "grad_norm": 0.9190279841423035, + "learning_rate": 3.738237026944756e-06, + "loss": 0.6092, + "step": 5792 + }, + { + "epoch": 2.034058988764045, + "grad_norm": 0.8763152360916138, + "learning_rate": 3.7378353722762526e-06, + "loss": 0.5772, + "step": 5793 + }, + { + "epoch": 2.0344101123595504, + "grad_norm": 0.8799238801002502, + "learning_rate": 3.7374336752748396e-06, + "loss": 0.6137, + "step": 5794 + }, + { + "epoch": 2.034761235955056, + "grad_norm": 0.838129997253418, + "learning_rate": 3.7370319359542532e-06, + "loss": 0.5618, + "step": 5795 + }, + { + "epoch": 2.0351123595505616, + "grad_norm": 0.8667547702789307, + "learning_rate": 3.7366301543282336e-06, + "loss": 0.5798, + "step": 5796 + }, + { + "epoch": 2.0354634831460676, + "grad_norm": 0.8789803981781006, + "learning_rate": 3.73622833041052e-06, + "loss": 0.59, + "step": 5797 + }, + { + "epoch": 2.035814606741573, + "grad_norm": 0.9414893388748169, + "learning_rate": 3.7358264642148556e-06, + "loss": 0.5972, + "step": 5798 + }, + { + "epoch": 2.036165730337079, + "grad_norm": 0.8904270529747009, + "learning_rate": 3.735424555754984e-06, + "loss": 0.5949, + "step": 5799 + }, + { + "epoch": 2.0365168539325844, + "grad_norm": 0.8712731003761292, + "learning_rate": 3.7350226050446496e-06, + "loss": 0.5917, + "step": 5800 + }, + { + "epoch": 2.03686797752809, + "grad_norm": 0.8815358877182007, + "learning_rate": 3.7346206120975994e-06, + "loss": 0.599, + "step": 5801 + }, + { + "epoch": 2.0372191011235956, + "grad_norm": 0.8902667760848999, + "learning_rate": 3.73421857692758e-06, + "loss": 0.5867, + "step": 5802 + }, + { + "epoch": 2.037570224719101, + "grad_norm": 0.8855932950973511, + "learning_rate": 3.7338164995483424e-06, + "loss": 0.6142, + "step": 5803 + }, + { + "epoch": 2.037921348314607, + "grad_norm": 0.8844321370124817, + "learning_rate": 3.7334143799736354e-06, + "loss": 0.5723, + "step": 5804 + }, + { + "epoch": 2.0382724719101124, + "grad_norm": 0.8979598879814148, + "learning_rate": 3.7330122182172124e-06, + "loss": 0.5662, + "step": 5805 + }, + { + "epoch": 2.038623595505618, + "grad_norm": 0.8970213532447815, + "learning_rate": 3.7326100142928267e-06, + "loss": 0.5982, + "step": 5806 + }, + { + "epoch": 2.0389747191011236, + "grad_norm": 0.8381531238555908, + "learning_rate": 3.732207768214233e-06, + "loss": 0.5475, + "step": 5807 + }, + { + "epoch": 2.039325842696629, + "grad_norm": 0.8699519038200378, + "learning_rate": 3.7318054799951884e-06, + "loss": 0.5742, + "step": 5808 + }, + { + "epoch": 2.039676966292135, + "grad_norm": 0.846829891204834, + "learning_rate": 3.73140314964945e-06, + "loss": 0.5557, + "step": 5809 + }, + { + "epoch": 2.0400280898876404, + "grad_norm": 0.8873547315597534, + "learning_rate": 3.731000777190778e-06, + "loss": 0.6004, + "step": 5810 + }, + { + "epoch": 2.040379213483146, + "grad_norm": 0.9075663089752197, + "learning_rate": 3.7305983626329323e-06, + "loss": 0.6042, + "step": 5811 + }, + { + "epoch": 2.0407303370786516, + "grad_norm": 0.9091333746910095, + "learning_rate": 3.7301959059896764e-06, + "loss": 0.6123, + "step": 5812 + }, + { + "epoch": 2.041081460674157, + "grad_norm": 0.8782358765602112, + "learning_rate": 3.7297934072747726e-06, + "loss": 0.5734, + "step": 5813 + }, + { + "epoch": 2.041432584269663, + "grad_norm": 0.8886682987213135, + "learning_rate": 3.729390866501987e-06, + "loss": 0.5757, + "step": 5814 + }, + { + "epoch": 2.0417837078651684, + "grad_norm": 0.8818804621696472, + "learning_rate": 3.7289882836850855e-06, + "loss": 0.6084, + "step": 5815 + }, + { + "epoch": 2.042134831460674, + "grad_norm": 0.8960871696472168, + "learning_rate": 3.7285856588378367e-06, + "loss": 0.5863, + "step": 5816 + }, + { + "epoch": 2.0424859550561796, + "grad_norm": 0.8980026245117188, + "learning_rate": 3.728182991974009e-06, + "loss": 0.5574, + "step": 5817 + }, + { + "epoch": 2.042837078651685, + "grad_norm": 0.8893744349479675, + "learning_rate": 3.7277802831073753e-06, + "loss": 0.5954, + "step": 5818 + }, + { + "epoch": 2.043188202247191, + "grad_norm": 0.9305961728096008, + "learning_rate": 3.7273775322517046e-06, + "loss": 0.5581, + "step": 5819 + }, + { + "epoch": 2.043539325842697, + "grad_norm": 0.8795851469039917, + "learning_rate": 3.726974739420774e-06, + "loss": 0.6124, + "step": 5820 + }, + { + "epoch": 2.0438904494382024, + "grad_norm": 0.8998198509216309, + "learning_rate": 3.7265719046283567e-06, + "loss": 0.5869, + "step": 5821 + }, + { + "epoch": 2.044241573033708, + "grad_norm": 0.8911243081092834, + "learning_rate": 3.72616902788823e-06, + "loss": 0.5875, + "step": 5822 + }, + { + "epoch": 2.0445926966292136, + "grad_norm": 0.8849146366119385, + "learning_rate": 3.7257661092141716e-06, + "loss": 0.601, + "step": 5823 + }, + { + "epoch": 2.044943820224719, + "grad_norm": 0.8870599269866943, + "learning_rate": 3.7253631486199617e-06, + "loss": 0.6019, + "step": 5824 + }, + { + "epoch": 2.045294943820225, + "grad_norm": 0.8740954399108887, + "learning_rate": 3.7249601461193804e-06, + "loss": 0.6131, + "step": 5825 + }, + { + "epoch": 2.0456460674157304, + "grad_norm": 0.8803035616874695, + "learning_rate": 3.72455710172621e-06, + "loss": 0.5683, + "step": 5826 + }, + { + "epoch": 2.045997191011236, + "grad_norm": 0.9090285301208496, + "learning_rate": 3.7241540154542354e-06, + "loss": 0.5809, + "step": 5827 + }, + { + "epoch": 2.0463483146067416, + "grad_norm": 0.9098823666572571, + "learning_rate": 3.72375088731724e-06, + "loss": 0.6458, + "step": 5828 + }, + { + "epoch": 2.046699438202247, + "grad_norm": 0.92947918176651, + "learning_rate": 3.7233477173290123e-06, + "loss": 0.5983, + "step": 5829 + }, + { + "epoch": 2.047050561797753, + "grad_norm": 0.8461861610412598, + "learning_rate": 3.7229445055033387e-06, + "loss": 0.6021, + "step": 5830 + }, + { + "epoch": 2.0474016853932584, + "grad_norm": 0.8962950110435486, + "learning_rate": 3.72254125185401e-06, + "loss": 0.572, + "step": 5831 + }, + { + "epoch": 2.047752808988764, + "grad_norm": 0.8702701926231384, + "learning_rate": 3.722137956394817e-06, + "loss": 0.5842, + "step": 5832 + }, + { + "epoch": 2.0481039325842696, + "grad_norm": 0.8926149010658264, + "learning_rate": 3.721734619139551e-06, + "loss": 0.5906, + "step": 5833 + }, + { + "epoch": 2.048455056179775, + "grad_norm": 0.8730502724647522, + "learning_rate": 3.721331240102007e-06, + "loss": 0.5725, + "step": 5834 + }, + { + "epoch": 2.048806179775281, + "grad_norm": 0.9188753366470337, + "learning_rate": 3.720927819295979e-06, + "loss": 0.5947, + "step": 5835 + }, + { + "epoch": 2.0491573033707864, + "grad_norm": 0.8904201984405518, + "learning_rate": 3.7205243567352645e-06, + "loss": 0.6209, + "step": 5836 + }, + { + "epoch": 2.049508426966292, + "grad_norm": 0.8634858131408691, + "learning_rate": 3.720120852433662e-06, + "loss": 0.5741, + "step": 5837 + }, + { + "epoch": 2.0498595505617976, + "grad_norm": 0.86948162317276, + "learning_rate": 3.7197173064049694e-06, + "loss": 0.573, + "step": 5838 + }, + { + "epoch": 2.050210674157303, + "grad_norm": 0.8552312850952148, + "learning_rate": 3.7193137186629897e-06, + "loss": 0.5445, + "step": 5839 + }, + { + "epoch": 2.050561797752809, + "grad_norm": 0.8711963891983032, + "learning_rate": 3.7189100892215234e-06, + "loss": 0.5784, + "step": 5840 + }, + { + "epoch": 2.050912921348315, + "grad_norm": 0.8837445378303528, + "learning_rate": 3.718506418094376e-06, + "loss": 0.5876, + "step": 5841 + }, + { + "epoch": 2.0512640449438204, + "grad_norm": 0.8832021951675415, + "learning_rate": 3.7181027052953515e-06, + "loss": 0.6054, + "step": 5842 + }, + { + "epoch": 2.051615168539326, + "grad_norm": 0.882051944732666, + "learning_rate": 3.7176989508382566e-06, + "loss": 0.5882, + "step": 5843 + }, + { + "epoch": 2.0519662921348316, + "grad_norm": 0.9083414077758789, + "learning_rate": 3.7172951547368996e-06, + "loss": 0.5859, + "step": 5844 + }, + { + "epoch": 2.052317415730337, + "grad_norm": 0.8930639624595642, + "learning_rate": 3.7168913170050903e-06, + "loss": 0.5932, + "step": 5845 + }, + { + "epoch": 2.052668539325843, + "grad_norm": 0.9015393257141113, + "learning_rate": 3.7164874376566385e-06, + "loss": 0.5843, + "step": 5846 + }, + { + "epoch": 2.0530196629213484, + "grad_norm": 0.8611611127853394, + "learning_rate": 3.7160835167053576e-06, + "loss": 0.5666, + "step": 5847 + }, + { + "epoch": 2.053370786516854, + "grad_norm": 0.8690086603164673, + "learning_rate": 3.7156795541650613e-06, + "loss": 0.5773, + "step": 5848 + }, + { + "epoch": 2.0537219101123596, + "grad_norm": 0.9017196297645569, + "learning_rate": 3.715275550049564e-06, + "loss": 0.6076, + "step": 5849 + }, + { + "epoch": 2.054073033707865, + "grad_norm": 0.9030886888504028, + "learning_rate": 3.7148715043726837e-06, + "loss": 0.6035, + "step": 5850 + }, + { + "epoch": 2.054424157303371, + "grad_norm": 0.8992416858673096, + "learning_rate": 3.7144674171482363e-06, + "loss": 0.5965, + "step": 5851 + }, + { + "epoch": 2.0547752808988764, + "grad_norm": 0.897692859172821, + "learning_rate": 3.7140632883900434e-06, + "loss": 0.6229, + "step": 5852 + }, + { + "epoch": 2.055126404494382, + "grad_norm": 0.901557981967926, + "learning_rate": 3.713659118111925e-06, + "loss": 0.5958, + "step": 5853 + }, + { + "epoch": 2.0554775280898876, + "grad_norm": 0.9022448062896729, + "learning_rate": 3.7132549063277033e-06, + "loss": 0.6035, + "step": 5854 + }, + { + "epoch": 2.055828651685393, + "grad_norm": 0.95181804895401, + "learning_rate": 3.7128506530512015e-06, + "loss": 0.6047, + "step": 5855 + }, + { + "epoch": 2.056179775280899, + "grad_norm": 0.8748995661735535, + "learning_rate": 3.7124463582962456e-06, + "loss": 0.6007, + "step": 5856 + }, + { + "epoch": 2.0565308988764044, + "grad_norm": 0.8931772112846375, + "learning_rate": 3.712042022076662e-06, + "loss": 0.575, + "step": 5857 + }, + { + "epoch": 2.05688202247191, + "grad_norm": 0.903578519821167, + "learning_rate": 3.7116376444062776e-06, + "loss": 0.5984, + "step": 5858 + }, + { + "epoch": 2.0572331460674156, + "grad_norm": 0.8780085444450378, + "learning_rate": 3.7112332252989224e-06, + "loss": 0.5793, + "step": 5859 + }, + { + "epoch": 2.057584269662921, + "grad_norm": 0.8822435736656189, + "learning_rate": 3.710828764768428e-06, + "loss": 0.5894, + "step": 5860 + }, + { + "epoch": 2.057935393258427, + "grad_norm": 0.8836818933486938, + "learning_rate": 3.7104242628286263e-06, + "loss": 0.5892, + "step": 5861 + }, + { + "epoch": 2.0582865168539324, + "grad_norm": 0.8829315900802612, + "learning_rate": 3.7100197194933502e-06, + "loss": 0.5736, + "step": 5862 + }, + { + "epoch": 2.0586376404494384, + "grad_norm": 0.911048173904419, + "learning_rate": 3.709615134776435e-06, + "loss": 0.5824, + "step": 5863 + }, + { + "epoch": 2.058988764044944, + "grad_norm": 0.8984778523445129, + "learning_rate": 3.7092105086917176e-06, + "loss": 0.5951, + "step": 5864 + }, + { + "epoch": 2.0593398876404496, + "grad_norm": 0.8903387188911438, + "learning_rate": 3.708805841253035e-06, + "loss": 0.5858, + "step": 5865 + }, + { + "epoch": 2.059691011235955, + "grad_norm": 0.876234769821167, + "learning_rate": 3.7084011324742273e-06, + "loss": 0.5936, + "step": 5866 + }, + { + "epoch": 2.060042134831461, + "grad_norm": 0.8848009705543518, + "learning_rate": 3.707996382369136e-06, + "loss": 0.5994, + "step": 5867 + }, + { + "epoch": 2.0603932584269664, + "grad_norm": 0.8751669526100159, + "learning_rate": 3.7075915909516e-06, + "loss": 0.6312, + "step": 5868 + }, + { + "epoch": 2.060744382022472, + "grad_norm": 0.8718172311782837, + "learning_rate": 3.7071867582354666e-06, + "loss": 0.5668, + "step": 5869 + }, + { + "epoch": 2.0610955056179776, + "grad_norm": 0.85500568151474, + "learning_rate": 3.7067818842345783e-06, + "loss": 0.5622, + "step": 5870 + }, + { + "epoch": 2.061446629213483, + "grad_norm": 0.8555139899253845, + "learning_rate": 3.706376968962782e-06, + "loss": 0.5673, + "step": 5871 + }, + { + "epoch": 2.061797752808989, + "grad_norm": 0.8747328519821167, + "learning_rate": 3.705972012433926e-06, + "loss": 0.5771, + "step": 5872 + }, + { + "epoch": 2.0621488764044944, + "grad_norm": 0.9076268076896667, + "learning_rate": 3.7055670146618594e-06, + "loss": 0.6017, + "step": 5873 + }, + { + "epoch": 2.0625, + "grad_norm": 0.8561304807662964, + "learning_rate": 3.7051619756604324e-06, + "loss": 0.5445, + "step": 5874 + }, + { + "epoch": 2.0628511235955056, + "grad_norm": 0.8934998512268066, + "learning_rate": 3.7047568954434972e-06, + "loss": 0.5957, + "step": 5875 + }, + { + "epoch": 2.063202247191011, + "grad_norm": 0.8787360787391663, + "learning_rate": 3.704351774024907e-06, + "loss": 0.5764, + "step": 5876 + }, + { + "epoch": 2.063553370786517, + "grad_norm": 0.8640685081481934, + "learning_rate": 3.7039466114185164e-06, + "loss": 0.563, + "step": 5877 + }, + { + "epoch": 2.0639044943820224, + "grad_norm": 0.8665021657943726, + "learning_rate": 3.703541407638182e-06, + "loss": 0.596, + "step": 5878 + }, + { + "epoch": 2.064255617977528, + "grad_norm": 0.8836948275566101, + "learning_rate": 3.703136162697762e-06, + "loss": 0.575, + "step": 5879 + }, + { + "epoch": 2.0646067415730336, + "grad_norm": 0.911559522151947, + "learning_rate": 3.7027308766111143e-06, + "loss": 0.6267, + "step": 5880 + }, + { + "epoch": 2.064957865168539, + "grad_norm": 0.8681786060333252, + "learning_rate": 3.7023255493920995e-06, + "loss": 0.5873, + "step": 5881 + }, + { + "epoch": 2.065308988764045, + "grad_norm": 0.8732732534408569, + "learning_rate": 3.70192018105458e-06, + "loss": 0.5614, + "step": 5882 + }, + { + "epoch": 2.0656601123595504, + "grad_norm": 0.8946567177772522, + "learning_rate": 3.7015147716124187e-06, + "loss": 0.5732, + "step": 5883 + }, + { + "epoch": 2.066011235955056, + "grad_norm": 0.8860378861427307, + "learning_rate": 3.7011093210794808e-06, + "loss": 0.5646, + "step": 5884 + }, + { + "epoch": 2.0663623595505616, + "grad_norm": 0.8822278380393982, + "learning_rate": 3.7007038294696317e-06, + "loss": 0.5621, + "step": 5885 + }, + { + "epoch": 2.0667134831460676, + "grad_norm": 0.8843095302581787, + "learning_rate": 3.700298296796739e-06, + "loss": 0.5941, + "step": 5886 + }, + { + "epoch": 2.067064606741573, + "grad_norm": 0.9095739126205444, + "learning_rate": 3.6998927230746718e-06, + "loss": 0.6071, + "step": 5887 + }, + { + "epoch": 2.067415730337079, + "grad_norm": 0.8828830718994141, + "learning_rate": 3.6994871083172995e-06, + "loss": 0.6115, + "step": 5888 + }, + { + "epoch": 2.0677668539325844, + "grad_norm": 0.8572625517845154, + "learning_rate": 3.6990814525384956e-06, + "loss": 0.5735, + "step": 5889 + }, + { + "epoch": 2.06811797752809, + "grad_norm": 0.8938194513320923, + "learning_rate": 3.6986757557521315e-06, + "loss": 0.5856, + "step": 5890 + }, + { + "epoch": 2.0684691011235956, + "grad_norm": 0.927527129650116, + "learning_rate": 3.6982700179720827e-06, + "loss": 0.5602, + "step": 5891 + }, + { + "epoch": 2.068820224719101, + "grad_norm": 0.9165863990783691, + "learning_rate": 3.6978642392122243e-06, + "loss": 0.5852, + "step": 5892 + }, + { + "epoch": 2.069171348314607, + "grad_norm": 0.9599493741989136, + "learning_rate": 3.697458419486434e-06, + "loss": 0.6183, + "step": 5893 + }, + { + "epoch": 2.0695224719101124, + "grad_norm": 0.8913846611976624, + "learning_rate": 3.6970525588085906e-06, + "loss": 0.6221, + "step": 5894 + }, + { + "epoch": 2.069873595505618, + "grad_norm": 0.9317246079444885, + "learning_rate": 3.6966466571925734e-06, + "loss": 0.5825, + "step": 5895 + }, + { + "epoch": 2.0702247191011236, + "grad_norm": 0.91375732421875, + "learning_rate": 3.696240714652265e-06, + "loss": 0.6106, + "step": 5896 + }, + { + "epoch": 2.070575842696629, + "grad_norm": 0.9135724306106567, + "learning_rate": 3.695834731201548e-06, + "loss": 0.5746, + "step": 5897 + }, + { + "epoch": 2.070926966292135, + "grad_norm": 0.9261338114738464, + "learning_rate": 3.695428706854306e-06, + "loss": 0.6093, + "step": 5898 + }, + { + "epoch": 2.0712780898876404, + "grad_norm": 0.8809404969215393, + "learning_rate": 3.6950226416244255e-06, + "loss": 0.5909, + "step": 5899 + }, + { + "epoch": 2.071629213483146, + "grad_norm": 0.8801942467689514, + "learning_rate": 3.694616535525793e-06, + "loss": 0.5791, + "step": 5900 + }, + { + "epoch": 2.0719803370786516, + "grad_norm": 0.9020616412162781, + "learning_rate": 3.6942103885722975e-06, + "loss": 0.5752, + "step": 5901 + }, + { + "epoch": 2.072331460674157, + "grad_norm": 0.9009699821472168, + "learning_rate": 3.6938042007778283e-06, + "loss": 0.59, + "step": 5902 + }, + { + "epoch": 2.072682584269663, + "grad_norm": 0.9019879102706909, + "learning_rate": 3.6933979721562775e-06, + "loss": 0.5919, + "step": 5903 + }, + { + "epoch": 2.0730337078651684, + "grad_norm": 0.8951608538627625, + "learning_rate": 3.692991702721537e-06, + "loss": 0.5898, + "step": 5904 + }, + { + "epoch": 2.073384831460674, + "grad_norm": 0.8855031728744507, + "learning_rate": 3.692585392487501e-06, + "loss": 0.5832, + "step": 5905 + }, + { + "epoch": 2.0737359550561796, + "grad_norm": 0.844199001789093, + "learning_rate": 3.6921790414680654e-06, + "loss": 0.5508, + "step": 5906 + }, + { + "epoch": 2.074087078651685, + "grad_norm": 0.9061780571937561, + "learning_rate": 3.6917726496771268e-06, + "loss": 0.5484, + "step": 5907 + }, + { + "epoch": 2.074438202247191, + "grad_norm": 0.8656120896339417, + "learning_rate": 3.6913662171285827e-06, + "loss": 0.571, + "step": 5908 + }, + { + "epoch": 2.074789325842697, + "grad_norm": 0.8697794675827026, + "learning_rate": 3.6909597438363337e-06, + "loss": 0.5896, + "step": 5909 + }, + { + "epoch": 2.0751404494382024, + "grad_norm": 0.8954538702964783, + "learning_rate": 3.690553229814281e-06, + "loss": 0.5824, + "step": 5910 + }, + { + "epoch": 2.075491573033708, + "grad_norm": 0.8624008297920227, + "learning_rate": 3.6901466750763258e-06, + "loss": 0.542, + "step": 5911 + }, + { + "epoch": 2.0758426966292136, + "grad_norm": 0.9189245700836182, + "learning_rate": 3.6897400796363735e-06, + "loss": 0.5947, + "step": 5912 + }, + { + "epoch": 2.076193820224719, + "grad_norm": 0.9016966223716736, + "learning_rate": 3.6893334435083285e-06, + "loss": 0.6207, + "step": 5913 + }, + { + "epoch": 2.076544943820225, + "grad_norm": 0.9189615249633789, + "learning_rate": 3.6889267667060975e-06, + "loss": 0.5741, + "step": 5914 + }, + { + "epoch": 2.0768960674157304, + "grad_norm": 0.9112511277198792, + "learning_rate": 3.688520049243588e-06, + "loss": 0.615, + "step": 5915 + }, + { + "epoch": 2.077247191011236, + "grad_norm": 0.8809523582458496, + "learning_rate": 3.6881132911347108e-06, + "loss": 0.5591, + "step": 5916 + }, + { + "epoch": 2.0775983146067416, + "grad_norm": 0.8637116551399231, + "learning_rate": 3.687706492393375e-06, + "loss": 0.5605, + "step": 5917 + }, + { + "epoch": 2.077949438202247, + "grad_norm": 0.9029573798179626, + "learning_rate": 3.687299653033493e-06, + "loss": 0.5905, + "step": 5918 + }, + { + "epoch": 2.078300561797753, + "grad_norm": 0.9370858669281006, + "learning_rate": 3.6868927730689796e-06, + "loss": 0.5789, + "step": 5919 + }, + { + "epoch": 2.0786516853932584, + "grad_norm": 0.884752631187439, + "learning_rate": 3.686485852513748e-06, + "loss": 0.5693, + "step": 5920 + }, + { + "epoch": 2.079002808988764, + "grad_norm": 0.8713205456733704, + "learning_rate": 3.686078891381717e-06, + "loss": 0.5818, + "step": 5921 + }, + { + "epoch": 2.0793539325842696, + "grad_norm": 0.934033215045929, + "learning_rate": 3.685671889686802e-06, + "loss": 0.5794, + "step": 5922 + }, + { + "epoch": 2.079705056179775, + "grad_norm": 0.8763996362686157, + "learning_rate": 3.685264847442923e-06, + "loss": 0.5755, + "step": 5923 + }, + { + "epoch": 2.080056179775281, + "grad_norm": 0.8777490258216858, + "learning_rate": 3.684857764664001e-06, + "loss": 0.5521, + "step": 5924 + }, + { + "epoch": 2.0804073033707864, + "grad_norm": 0.8671329021453857, + "learning_rate": 3.684450641363957e-06, + "loss": 0.5734, + "step": 5925 + }, + { + "epoch": 2.080758426966292, + "grad_norm": 0.911651611328125, + "learning_rate": 3.684043477556715e-06, + "loss": 0.5897, + "step": 5926 + }, + { + "epoch": 2.0811095505617976, + "grad_norm": 0.9036709070205688, + "learning_rate": 3.683636273256198e-06, + "loss": 0.6008, + "step": 5927 + }, + { + "epoch": 2.081460674157303, + "grad_norm": 0.917131781578064, + "learning_rate": 3.683229028476334e-06, + "loss": 0.6003, + "step": 5928 + }, + { + "epoch": 2.081811797752809, + "grad_norm": 0.8952845931053162, + "learning_rate": 3.68282174323105e-06, + "loss": 0.5554, + "step": 5929 + }, + { + "epoch": 2.082162921348315, + "grad_norm": 0.8961856961250305, + "learning_rate": 3.6824144175342747e-06, + "loss": 0.5648, + "step": 5930 + }, + { + "epoch": 2.0825140449438204, + "grad_norm": 0.895000696182251, + "learning_rate": 3.682007051399938e-06, + "loss": 0.6074, + "step": 5931 + }, + { + "epoch": 2.082865168539326, + "grad_norm": 0.9262698292732239, + "learning_rate": 3.681599644841971e-06, + "loss": 0.5706, + "step": 5932 + }, + { + "epoch": 2.0832162921348316, + "grad_norm": 0.8890082240104675, + "learning_rate": 3.6811921978743082e-06, + "loss": 0.6214, + "step": 5933 + }, + { + "epoch": 2.083567415730337, + "grad_norm": 0.8488702178001404, + "learning_rate": 3.680784710510882e-06, + "loss": 0.5605, + "step": 5934 + }, + { + "epoch": 2.083918539325843, + "grad_norm": 0.9110900163650513, + "learning_rate": 3.6803771827656297e-06, + "loss": 0.6025, + "step": 5935 + }, + { + "epoch": 2.0842696629213484, + "grad_norm": 0.8841571807861328, + "learning_rate": 3.6799696146524877e-06, + "loss": 0.5465, + "step": 5936 + }, + { + "epoch": 2.084620786516854, + "grad_norm": 0.9424610137939453, + "learning_rate": 3.6795620061853943e-06, + "loss": 0.6325, + "step": 5937 + }, + { + "epoch": 2.0849719101123596, + "grad_norm": 0.8845745921134949, + "learning_rate": 3.67915435737829e-06, + "loss": 0.5901, + "step": 5938 + }, + { + "epoch": 2.085323033707865, + "grad_norm": 0.8947811126708984, + "learning_rate": 3.6787466682451156e-06, + "loss": 0.5885, + "step": 5939 + }, + { + "epoch": 2.085674157303371, + "grad_norm": 0.8605453372001648, + "learning_rate": 3.678338938799814e-06, + "loss": 0.5408, + "step": 5940 + }, + { + "epoch": 2.0860252808988764, + "grad_norm": 0.9256695508956909, + "learning_rate": 3.6779311690563292e-06, + "loss": 0.6102, + "step": 5941 + }, + { + "epoch": 2.086376404494382, + "grad_norm": 0.8742806911468506, + "learning_rate": 3.6775233590286065e-06, + "loss": 0.5772, + "step": 5942 + }, + { + "epoch": 2.0867275280898876, + "grad_norm": 0.8935145735740662, + "learning_rate": 3.6771155087305916e-06, + "loss": 0.5893, + "step": 5943 + }, + { + "epoch": 2.087078651685393, + "grad_norm": 0.9108278751373291, + "learning_rate": 3.676707618176234e-06, + "loss": 0.6283, + "step": 5944 + }, + { + "epoch": 2.087429775280899, + "grad_norm": 0.9008842706680298, + "learning_rate": 3.676299687379483e-06, + "loss": 0.6101, + "step": 5945 + }, + { + "epoch": 2.0877808988764044, + "grad_norm": 0.8601182103157043, + "learning_rate": 3.675891716354289e-06, + "loss": 0.5357, + "step": 5946 + }, + { + "epoch": 2.08813202247191, + "grad_norm": 0.880409300327301, + "learning_rate": 3.6754837051146043e-06, + "loss": 0.5724, + "step": 5947 + }, + { + "epoch": 2.0884831460674156, + "grad_norm": 0.859566330909729, + "learning_rate": 3.675075653674383e-06, + "loss": 0.577, + "step": 5948 + }, + { + "epoch": 2.088834269662921, + "grad_norm": 0.8837049007415771, + "learning_rate": 3.6746675620475804e-06, + "loss": 0.5829, + "step": 5949 + }, + { + "epoch": 2.089185393258427, + "grad_norm": 0.8644298911094666, + "learning_rate": 3.6742594302481516e-06, + "loss": 0.5647, + "step": 5950 + }, + { + "epoch": 2.0895365168539324, + "grad_norm": 0.9024447798728943, + "learning_rate": 3.6738512582900555e-06, + "loss": 0.5594, + "step": 5951 + }, + { + "epoch": 2.0898876404494384, + "grad_norm": 0.9284654855728149, + "learning_rate": 3.6734430461872505e-06, + "loss": 0.5899, + "step": 5952 + }, + { + "epoch": 2.090238764044944, + "grad_norm": 0.9058789610862732, + "learning_rate": 3.6730347939536982e-06, + "loss": 0.5769, + "step": 5953 + }, + { + "epoch": 2.0905898876404496, + "grad_norm": 0.8683139681816101, + "learning_rate": 3.672626501603359e-06, + "loss": 0.5499, + "step": 5954 + }, + { + "epoch": 2.090941011235955, + "grad_norm": 0.89882892370224, + "learning_rate": 3.6722181691501975e-06, + "loss": 0.5811, + "step": 5955 + }, + { + "epoch": 2.091292134831461, + "grad_norm": 0.8749769330024719, + "learning_rate": 3.6718097966081774e-06, + "loss": 0.5754, + "step": 5956 + }, + { + "epoch": 2.0916432584269664, + "grad_norm": 0.8947672247886658, + "learning_rate": 3.671401383991265e-06, + "loss": 0.5967, + "step": 5957 + }, + { + "epoch": 2.091994382022472, + "grad_norm": 0.9050623178482056, + "learning_rate": 3.6709929313134276e-06, + "loss": 0.5691, + "step": 5958 + }, + { + "epoch": 2.0923455056179776, + "grad_norm": 0.9081699848175049, + "learning_rate": 3.6705844385886334e-06, + "loss": 0.6139, + "step": 5959 + }, + { + "epoch": 2.092696629213483, + "grad_norm": 0.8798913359642029, + "learning_rate": 3.670175905830854e-06, + "loss": 0.5839, + "step": 5960 + }, + { + "epoch": 2.093047752808989, + "grad_norm": 0.8618450164794922, + "learning_rate": 3.6697673330540593e-06, + "loss": 0.569, + "step": 5961 + }, + { + "epoch": 2.0933988764044944, + "grad_norm": 0.8755471706390381, + "learning_rate": 3.669358720272223e-06, + "loss": 0.5551, + "step": 5962 + }, + { + "epoch": 2.09375, + "grad_norm": 0.9232277274131775, + "learning_rate": 3.6689500674993193e-06, + "loss": 0.6198, + "step": 5963 + }, + { + "epoch": 2.0941011235955056, + "grad_norm": 0.9025054574012756, + "learning_rate": 3.6685413747493235e-06, + "loss": 0.6024, + "step": 5964 + }, + { + "epoch": 2.094452247191011, + "grad_norm": 0.8963713049888611, + "learning_rate": 3.6681326420362125e-06, + "loss": 0.6056, + "step": 5965 + }, + { + "epoch": 2.094803370786517, + "grad_norm": 0.9302486777305603, + "learning_rate": 3.6677238693739648e-06, + "loss": 0.6118, + "step": 5966 + }, + { + "epoch": 2.0951544943820224, + "grad_norm": 0.8892012238502502, + "learning_rate": 3.6673150567765602e-06, + "loss": 0.6069, + "step": 5967 + }, + { + "epoch": 2.095505617977528, + "grad_norm": 0.8898606896400452, + "learning_rate": 3.6669062042579796e-06, + "loss": 0.5815, + "step": 5968 + }, + { + "epoch": 2.0958567415730336, + "grad_norm": 0.9015076160430908, + "learning_rate": 3.666497311832205e-06, + "loss": 0.5842, + "step": 5969 + }, + { + "epoch": 2.096207865168539, + "grad_norm": 0.9140727519989014, + "learning_rate": 3.666088379513221e-06, + "loss": 0.5993, + "step": 5970 + }, + { + "epoch": 2.096558988764045, + "grad_norm": 0.8970695734024048, + "learning_rate": 3.6656794073150114e-06, + "loss": 0.5748, + "step": 5971 + }, + { + "epoch": 2.0969101123595504, + "grad_norm": 0.8759880065917969, + "learning_rate": 3.665270395251564e-06, + "loss": 0.5805, + "step": 5972 + }, + { + "epoch": 2.097261235955056, + "grad_norm": 0.8815058469772339, + "learning_rate": 3.664861343336866e-06, + "loss": 0.577, + "step": 5973 + }, + { + "epoch": 2.0976123595505616, + "grad_norm": 0.8842485547065735, + "learning_rate": 3.6644522515849075e-06, + "loss": 0.5969, + "step": 5974 + }, + { + "epoch": 2.0979634831460676, + "grad_norm": 0.8830922245979309, + "learning_rate": 3.6640431200096775e-06, + "loss": 0.5992, + "step": 5975 + }, + { + "epoch": 2.098314606741573, + "grad_norm": 0.9191491603851318, + "learning_rate": 3.66363394862517e-06, + "loss": 0.6109, + "step": 5976 + }, + { + "epoch": 2.098665730337079, + "grad_norm": 0.9068108201026917, + "learning_rate": 3.663224737445376e-06, + "loss": 0.6103, + "step": 5977 + }, + { + "epoch": 2.0990168539325844, + "grad_norm": 0.8892364501953125, + "learning_rate": 3.6628154864842923e-06, + "loss": 0.6036, + "step": 5978 + }, + { + "epoch": 2.09936797752809, + "grad_norm": 0.8922228217124939, + "learning_rate": 3.662406195755914e-06, + "loss": 0.5713, + "step": 5979 + }, + { + "epoch": 2.0997191011235956, + "grad_norm": 0.895887553691864, + "learning_rate": 3.661996865274238e-06, + "loss": 0.5926, + "step": 5980 + }, + { + "epoch": 2.100070224719101, + "grad_norm": 0.9129197597503662, + "learning_rate": 3.661587495053264e-06, + "loss": 0.599, + "step": 5981 + }, + { + "epoch": 2.100421348314607, + "grad_norm": 0.9314238429069519, + "learning_rate": 3.661178085106991e-06, + "loss": 0.6058, + "step": 5982 + }, + { + "epoch": 2.1007724719101124, + "grad_norm": 0.8824983239173889, + "learning_rate": 3.6607686354494217e-06, + "loss": 0.5841, + "step": 5983 + }, + { + "epoch": 2.101123595505618, + "grad_norm": 0.9065082669258118, + "learning_rate": 3.660359146094558e-06, + "loss": 0.5937, + "step": 5984 + }, + { + "epoch": 2.1014747191011236, + "grad_norm": 0.8882701992988586, + "learning_rate": 3.659949617056404e-06, + "loss": 0.5831, + "step": 5985 + }, + { + "epoch": 2.101825842696629, + "grad_norm": 0.8943924903869629, + "learning_rate": 3.659540048348967e-06, + "loss": 0.5805, + "step": 5986 + }, + { + "epoch": 2.102176966292135, + "grad_norm": 0.9324249625205994, + "learning_rate": 3.6591304399862515e-06, + "loss": 0.6056, + "step": 5987 + }, + { + "epoch": 2.1025280898876404, + "grad_norm": 0.9193480610847473, + "learning_rate": 3.658720791982267e-06, + "loss": 0.5757, + "step": 5988 + }, + { + "epoch": 2.102879213483146, + "grad_norm": 0.8888943195343018, + "learning_rate": 3.6583111043510234e-06, + "loss": 0.5411, + "step": 5989 + }, + { + "epoch": 2.1032303370786516, + "grad_norm": 0.8851947784423828, + "learning_rate": 3.657901377106531e-06, + "loss": 0.5839, + "step": 5990 + }, + { + "epoch": 2.103581460674157, + "grad_norm": 0.8993967771530151, + "learning_rate": 3.657491610262802e-06, + "loss": 0.5831, + "step": 5991 + }, + { + "epoch": 2.103932584269663, + "grad_norm": 0.9143550992012024, + "learning_rate": 3.657081803833851e-06, + "loss": 0.5838, + "step": 5992 + }, + { + "epoch": 2.1042837078651684, + "grad_norm": 0.913809597492218, + "learning_rate": 3.6566719578336916e-06, + "loss": 0.5919, + "step": 5993 + }, + { + "epoch": 2.104634831460674, + "grad_norm": 0.9145388603210449, + "learning_rate": 3.656262072276342e-06, + "loss": 0.5679, + "step": 5994 + }, + { + "epoch": 2.1049859550561796, + "grad_norm": 0.8764489889144897, + "learning_rate": 3.6558521471758184e-06, + "loss": 0.5483, + "step": 5995 + }, + { + "epoch": 2.105337078651685, + "grad_norm": 0.8881774544715881, + "learning_rate": 3.6554421825461407e-06, + "loss": 0.5877, + "step": 5996 + }, + { + "epoch": 2.105688202247191, + "grad_norm": 0.9042853116989136, + "learning_rate": 3.6550321784013286e-06, + "loss": 0.5799, + "step": 5997 + }, + { + "epoch": 2.106039325842697, + "grad_norm": 0.890293300151825, + "learning_rate": 3.6546221347554045e-06, + "loss": 0.6022, + "step": 5998 + }, + { + "epoch": 2.1063904494382024, + "grad_norm": 0.8841552138328552, + "learning_rate": 3.6542120516223916e-06, + "loss": 0.5725, + "step": 5999 + }, + { + "epoch": 2.106741573033708, + "grad_norm": 0.9131177663803101, + "learning_rate": 3.653801929016314e-06, + "loss": 0.5893, + "step": 6000 + }, + { + "epoch": 2.1070926966292136, + "grad_norm": 0.8825763463973999, + "learning_rate": 3.6533917669511985e-06, + "loss": 0.5733, + "step": 6001 + }, + { + "epoch": 2.107443820224719, + "grad_norm": 0.8770654797554016, + "learning_rate": 3.652981565441071e-06, + "loss": 0.591, + "step": 6002 + }, + { + "epoch": 2.107794943820225, + "grad_norm": 0.9025745987892151, + "learning_rate": 3.6525713244999605e-06, + "loss": 0.5808, + "step": 6003 + }, + { + "epoch": 2.1081460674157304, + "grad_norm": 0.8804476857185364, + "learning_rate": 3.6521610441418976e-06, + "loss": 0.6004, + "step": 6004 + }, + { + "epoch": 2.108497191011236, + "grad_norm": 0.8800620436668396, + "learning_rate": 3.651750724380912e-06, + "loss": 0.6049, + "step": 6005 + }, + { + "epoch": 2.1088483146067416, + "grad_norm": 0.9342840313911438, + "learning_rate": 3.6513403652310374e-06, + "loss": 0.6098, + "step": 6006 + }, + { + "epoch": 2.109199438202247, + "grad_norm": 0.8898195624351501, + "learning_rate": 3.650929966706308e-06, + "loss": 0.6188, + "step": 6007 + }, + { + "epoch": 2.109550561797753, + "grad_norm": 0.8977441191673279, + "learning_rate": 3.6505195288207586e-06, + "loss": 0.5897, + "step": 6008 + }, + { + "epoch": 2.1099016853932584, + "grad_norm": 0.8968367576599121, + "learning_rate": 3.6501090515884252e-06, + "loss": 0.6155, + "step": 6009 + }, + { + "epoch": 2.110252808988764, + "grad_norm": 0.8875924348831177, + "learning_rate": 3.6496985350233466e-06, + "loss": 0.607, + "step": 6010 + }, + { + "epoch": 2.1106039325842696, + "grad_norm": 0.8825415372848511, + "learning_rate": 3.649287979139562e-06, + "loss": 0.5759, + "step": 6011 + }, + { + "epoch": 2.110955056179775, + "grad_norm": 0.8900230526924133, + "learning_rate": 3.6488773839511123e-06, + "loss": 0.5645, + "step": 6012 + }, + { + "epoch": 2.111306179775281, + "grad_norm": 0.89983731508255, + "learning_rate": 3.6484667494720386e-06, + "loss": 0.618, + "step": 6013 + }, + { + "epoch": 2.1116573033707864, + "grad_norm": 0.9116685390472412, + "learning_rate": 3.648056075716385e-06, + "loss": 0.6051, + "step": 6014 + }, + { + "epoch": 2.112008426966292, + "grad_norm": 0.8847039937973022, + "learning_rate": 3.647645362698196e-06, + "loss": 0.5815, + "step": 6015 + }, + { + "epoch": 2.1123595505617976, + "grad_norm": 0.8583051562309265, + "learning_rate": 3.6472346104315184e-06, + "loss": 0.5405, + "step": 6016 + }, + { + "epoch": 2.112710674157303, + "grad_norm": 0.9137873649597168, + "learning_rate": 3.6468238189303985e-06, + "loss": 0.5623, + "step": 6017 + }, + { + "epoch": 2.113061797752809, + "grad_norm": 0.8863442540168762, + "learning_rate": 3.6464129882088843e-06, + "loss": 0.5811, + "step": 6018 + }, + { + "epoch": 2.113412921348315, + "grad_norm": 0.9018019437789917, + "learning_rate": 3.646002118281028e-06, + "loss": 0.6201, + "step": 6019 + }, + { + "epoch": 2.1137640449438204, + "grad_norm": 0.9345055222511292, + "learning_rate": 3.645591209160879e-06, + "loss": 0.595, + "step": 6020 + }, + { + "epoch": 2.114115168539326, + "grad_norm": 0.9012865424156189, + "learning_rate": 3.645180260862492e-06, + "loss": 0.6253, + "step": 6021 + }, + { + "epoch": 2.1144662921348316, + "grad_norm": 0.8870772123336792, + "learning_rate": 3.6447692733999195e-06, + "loss": 0.6149, + "step": 6022 + }, + { + "epoch": 2.114817415730337, + "grad_norm": 0.9296122789382935, + "learning_rate": 3.6443582467872173e-06, + "loss": 0.5996, + "step": 6023 + }, + { + "epoch": 2.115168539325843, + "grad_norm": 0.9270819425582886, + "learning_rate": 3.6439471810384425e-06, + "loss": 0.5929, + "step": 6024 + }, + { + "epoch": 2.1155196629213484, + "grad_norm": 0.9137008190155029, + "learning_rate": 3.643536076167653e-06, + "loss": 0.6059, + "step": 6025 + }, + { + "epoch": 2.115870786516854, + "grad_norm": 0.8802319169044495, + "learning_rate": 3.643124932188908e-06, + "loss": 0.6016, + "step": 6026 + }, + { + "epoch": 2.1162219101123596, + "grad_norm": 0.9111731648445129, + "learning_rate": 3.6427137491162685e-06, + "loss": 0.6027, + "step": 6027 + }, + { + "epoch": 2.116573033707865, + "grad_norm": 0.876023530960083, + "learning_rate": 3.6423025269637968e-06, + "loss": 0.5893, + "step": 6028 + }, + { + "epoch": 2.116924157303371, + "grad_norm": 0.8902382254600525, + "learning_rate": 3.6418912657455565e-06, + "loss": 0.5905, + "step": 6029 + }, + { + "epoch": 2.1172752808988764, + "grad_norm": 0.900912880897522, + "learning_rate": 3.641479965475612e-06, + "loss": 0.5943, + "step": 6030 + }, + { + "epoch": 2.117626404494382, + "grad_norm": 0.876758873462677, + "learning_rate": 3.641068626168029e-06, + "loss": 0.5537, + "step": 6031 + }, + { + "epoch": 2.1179775280898876, + "grad_norm": 0.8998464941978455, + "learning_rate": 3.640657247836875e-06, + "loss": 0.5655, + "step": 6032 + }, + { + "epoch": 2.118328651685393, + "grad_norm": 0.9266892671585083, + "learning_rate": 3.64024583049622e-06, + "loss": 0.6026, + "step": 6033 + }, + { + "epoch": 2.118679775280899, + "grad_norm": 0.8956958055496216, + "learning_rate": 3.639834374160133e-06, + "loss": 0.6097, + "step": 6034 + }, + { + "epoch": 2.1190308988764044, + "grad_norm": 0.8688603043556213, + "learning_rate": 3.639422878842686e-06, + "loss": 0.582, + "step": 6035 + }, + { + "epoch": 2.11938202247191, + "grad_norm": 0.9054641127586365, + "learning_rate": 3.639011344557951e-06, + "loss": 0.5807, + "step": 6036 + }, + { + "epoch": 2.1197331460674156, + "grad_norm": 0.8836588263511658, + "learning_rate": 3.6385997713200034e-06, + "loss": 0.5822, + "step": 6037 + }, + { + "epoch": 2.120084269662921, + "grad_norm": 0.9030302166938782, + "learning_rate": 3.638188159142918e-06, + "loss": 0.6117, + "step": 6038 + }, + { + "epoch": 2.120435393258427, + "grad_norm": 0.8707258701324463, + "learning_rate": 3.6377765080407707e-06, + "loss": 0.5827, + "step": 6039 + }, + { + "epoch": 2.1207865168539324, + "grad_norm": 0.8488860726356506, + "learning_rate": 3.637364818027641e-06, + "loss": 0.5439, + "step": 6040 + }, + { + "epoch": 2.1211376404494384, + "grad_norm": 0.8651903867721558, + "learning_rate": 3.6369530891176076e-06, + "loss": 0.5826, + "step": 6041 + }, + { + "epoch": 2.121488764044944, + "grad_norm": 0.9176006317138672, + "learning_rate": 3.6365413213247513e-06, + "loss": 0.6048, + "step": 6042 + }, + { + "epoch": 2.1218398876404496, + "grad_norm": 0.9069273471832275, + "learning_rate": 3.636129514663155e-06, + "loss": 0.6083, + "step": 6043 + }, + { + "epoch": 2.122191011235955, + "grad_norm": 0.8784558773040771, + "learning_rate": 3.6357176691469016e-06, + "loss": 0.5653, + "step": 6044 + }, + { + "epoch": 2.122542134831461, + "grad_norm": 0.8880614042282104, + "learning_rate": 3.635305784790075e-06, + "loss": 0.596, + "step": 6045 + }, + { + "epoch": 2.1228932584269664, + "grad_norm": 0.8955920934677124, + "learning_rate": 3.634893861606762e-06, + "loss": 0.5787, + "step": 6046 + }, + { + "epoch": 2.123244382022472, + "grad_norm": 0.8876059055328369, + "learning_rate": 3.63448189961105e-06, + "loss": 0.5574, + "step": 6047 + }, + { + "epoch": 2.1235955056179776, + "grad_norm": 0.928648829460144, + "learning_rate": 3.634069898817029e-06, + "loss": 0.6265, + "step": 6048 + }, + { + "epoch": 2.123946629213483, + "grad_norm": 0.9039307236671448, + "learning_rate": 3.633657859238787e-06, + "loss": 0.6118, + "step": 6049 + }, + { + "epoch": 2.124297752808989, + "grad_norm": 0.9031195044517517, + "learning_rate": 3.6332457808904165e-06, + "loss": 0.5996, + "step": 6050 + }, + { + "epoch": 2.1246488764044944, + "grad_norm": 0.8785572052001953, + "learning_rate": 3.6328336637860097e-06, + "loss": 0.5409, + "step": 6051 + }, + { + "epoch": 2.125, + "grad_norm": 0.9082341194152832, + "learning_rate": 3.6324215079396606e-06, + "loss": 0.5867, + "step": 6052 + }, + { + "epoch": 2.1253511235955056, + "grad_norm": 0.9263855218887329, + "learning_rate": 3.632009313365466e-06, + "loss": 0.5874, + "step": 6053 + }, + { + "epoch": 2.125702247191011, + "grad_norm": 0.8875517249107361, + "learning_rate": 3.6315970800775214e-06, + "loss": 0.5949, + "step": 6054 + }, + { + "epoch": 2.126053370786517, + "grad_norm": 0.9085497856140137, + "learning_rate": 3.631184808089925e-06, + "loss": 0.58, + "step": 6055 + }, + { + "epoch": 2.1264044943820224, + "grad_norm": 0.8873892426490784, + "learning_rate": 3.630772497416776e-06, + "loss": 0.611, + "step": 6056 + }, + { + "epoch": 2.126755617977528, + "grad_norm": 0.8809614181518555, + "learning_rate": 3.6303601480721747e-06, + "loss": 0.5774, + "step": 6057 + }, + { + "epoch": 2.1271067415730336, + "grad_norm": 0.865064263343811, + "learning_rate": 3.629947760070224e-06, + "loss": 0.5662, + "step": 6058 + }, + { + "epoch": 2.127457865168539, + "grad_norm": 0.8960385918617249, + "learning_rate": 3.6295353334250266e-06, + "loss": 0.586, + "step": 6059 + }, + { + "epoch": 2.127808988764045, + "grad_norm": 0.863010823726654, + "learning_rate": 3.6291228681506872e-06, + "loss": 0.5551, + "step": 6060 + }, + { + "epoch": 2.1281601123595504, + "grad_norm": 0.8789451718330383, + "learning_rate": 3.6287103642613126e-06, + "loss": 0.5981, + "step": 6061 + }, + { + "epoch": 2.128511235955056, + "grad_norm": 0.9608534574508667, + "learning_rate": 3.6282978217710084e-06, + "loss": 0.6194, + "step": 6062 + }, + { + "epoch": 2.1288623595505616, + "grad_norm": 0.8890073299407959, + "learning_rate": 3.6278852406938854e-06, + "loss": 0.5935, + "step": 6063 + }, + { + "epoch": 2.1292134831460676, + "grad_norm": 0.8967961668968201, + "learning_rate": 3.6274726210440513e-06, + "loss": 0.5621, + "step": 6064 + }, + { + "epoch": 2.129564606741573, + "grad_norm": 0.9044331312179565, + "learning_rate": 3.627059962835619e-06, + "loss": 0.6219, + "step": 6065 + }, + { + "epoch": 2.129915730337079, + "grad_norm": 0.8840842843055725, + "learning_rate": 3.6266472660827e-06, + "loss": 0.5699, + "step": 6066 + }, + { + "epoch": 2.1302668539325844, + "grad_norm": 0.8718449473381042, + "learning_rate": 3.6262345307994083e-06, + "loss": 0.5743, + "step": 6067 + }, + { + "epoch": 2.13061797752809, + "grad_norm": 0.9051425457000732, + "learning_rate": 3.625821756999861e-06, + "loss": 0.566, + "step": 6068 + }, + { + "epoch": 2.1309691011235956, + "grad_norm": 0.8948372006416321, + "learning_rate": 3.6254089446981715e-06, + "loss": 0.5677, + "step": 6069 + }, + { + "epoch": 2.131320224719101, + "grad_norm": 0.8956426382064819, + "learning_rate": 3.6249960939084595e-06, + "loss": 0.595, + "step": 6070 + }, + { + "epoch": 2.131671348314607, + "grad_norm": 0.8849937915802002, + "learning_rate": 3.6245832046448437e-06, + "loss": 0.5789, + "step": 6071 + }, + { + "epoch": 2.1320224719101124, + "grad_norm": 0.8585726618766785, + "learning_rate": 3.624170276921445e-06, + "loss": 0.5718, + "step": 6072 + }, + { + "epoch": 2.132373595505618, + "grad_norm": 0.8997080326080322, + "learning_rate": 3.6237573107523837e-06, + "loss": 0.6301, + "step": 6073 + }, + { + "epoch": 2.1327247191011236, + "grad_norm": 0.9126721620559692, + "learning_rate": 3.6233443061517846e-06, + "loss": 0.6037, + "step": 6074 + }, + { + "epoch": 2.133075842696629, + "grad_norm": 0.9080379605293274, + "learning_rate": 3.622931263133772e-06, + "loss": 0.5676, + "step": 6075 + }, + { + "epoch": 2.133426966292135, + "grad_norm": 0.8928210735321045, + "learning_rate": 3.6225181817124706e-06, + "loss": 0.5599, + "step": 6076 + }, + { + "epoch": 2.1337780898876404, + "grad_norm": 0.9351580142974854, + "learning_rate": 3.622105061902008e-06, + "loss": 0.6021, + "step": 6077 + }, + { + "epoch": 2.134129213483146, + "grad_norm": 0.8837612867355347, + "learning_rate": 3.6216919037165127e-06, + "loss": 0.5782, + "step": 6078 + }, + { + "epoch": 2.1344803370786516, + "grad_norm": 0.8808587193489075, + "learning_rate": 3.6212787071701138e-06, + "loss": 0.5693, + "step": 6079 + }, + { + "epoch": 2.134831460674157, + "grad_norm": 0.8964069485664368, + "learning_rate": 3.6208654722769434e-06, + "loss": 0.5574, + "step": 6080 + }, + { + "epoch": 2.135182584269663, + "grad_norm": 0.8928869366645813, + "learning_rate": 3.6204521990511322e-06, + "loss": 0.6095, + "step": 6081 + }, + { + "epoch": 2.1355337078651684, + "grad_norm": 0.9963878989219666, + "learning_rate": 3.6200388875068147e-06, + "loss": 0.5555, + "step": 6082 + }, + { + "epoch": 2.135884831460674, + "grad_norm": 0.9124773144721985, + "learning_rate": 3.6196255376581258e-06, + "loss": 0.5812, + "step": 6083 + }, + { + "epoch": 2.1362359550561796, + "grad_norm": 0.9189741015434265, + "learning_rate": 3.619212149519201e-06, + "loss": 0.5871, + "step": 6084 + }, + { + "epoch": 2.1365870786516856, + "grad_norm": 0.8763818144798279, + "learning_rate": 3.6187987231041792e-06, + "loss": 0.5439, + "step": 6085 + }, + { + "epoch": 2.136938202247191, + "grad_norm": 0.8844108581542969, + "learning_rate": 3.618385258427198e-06, + "loss": 0.5618, + "step": 6086 + }, + { + "epoch": 2.137289325842697, + "grad_norm": 0.8809760212898254, + "learning_rate": 3.6179717555023986e-06, + "loss": 0.5626, + "step": 6087 + }, + { + "epoch": 2.1376404494382024, + "grad_norm": 0.9006713628768921, + "learning_rate": 3.6175582143439214e-06, + "loss": 0.6011, + "step": 6088 + }, + { + "epoch": 2.137991573033708, + "grad_norm": 0.8796180486679077, + "learning_rate": 3.6171446349659093e-06, + "loss": 0.5996, + "step": 6089 + }, + { + "epoch": 2.1383426966292136, + "grad_norm": 0.8964955806732178, + "learning_rate": 3.616731017382507e-06, + "loss": 0.5712, + "step": 6090 + }, + { + "epoch": 2.138693820224719, + "grad_norm": 0.9046218991279602, + "learning_rate": 3.6163173616078594e-06, + "loss": 0.6266, + "step": 6091 + }, + { + "epoch": 2.139044943820225, + "grad_norm": 0.8758260011672974, + "learning_rate": 3.6159036676561132e-06, + "loss": 0.6097, + "step": 6092 + }, + { + "epoch": 2.1393960674157304, + "grad_norm": 0.903355062007904, + "learning_rate": 3.6154899355414162e-06, + "loss": 0.58, + "step": 6093 + }, + { + "epoch": 2.139747191011236, + "grad_norm": 0.8973618149757385, + "learning_rate": 3.615076165277919e-06, + "loss": 0.5696, + "step": 6094 + }, + { + "epoch": 2.1400983146067416, + "grad_norm": 0.9142894148826599, + "learning_rate": 3.6146623568797705e-06, + "loss": 0.6041, + "step": 6095 + }, + { + "epoch": 2.140449438202247, + "grad_norm": 0.9032093286514282, + "learning_rate": 3.614248510361123e-06, + "loss": 0.6128, + "step": 6096 + }, + { + "epoch": 2.140800561797753, + "grad_norm": 0.8991050720214844, + "learning_rate": 3.6138346257361302e-06, + "loss": 0.5813, + "step": 6097 + }, + { + "epoch": 2.1411516853932584, + "grad_norm": 0.8855283260345459, + "learning_rate": 3.6134207030189462e-06, + "loss": 0.5637, + "step": 6098 + }, + { + "epoch": 2.141502808988764, + "grad_norm": 0.9061201214790344, + "learning_rate": 3.613006742223727e-06, + "loss": 0.5847, + "step": 6099 + }, + { + "epoch": 2.1418539325842696, + "grad_norm": 0.910794198513031, + "learning_rate": 3.612592743364629e-06, + "loss": 0.5999, + "step": 6100 + }, + { + "epoch": 2.142205056179775, + "grad_norm": 0.9515483379364014, + "learning_rate": 3.6121787064558117e-06, + "loss": 0.6242, + "step": 6101 + }, + { + "epoch": 2.142556179775281, + "grad_norm": 0.9159862399101257, + "learning_rate": 3.6117646315114347e-06, + "loss": 0.5692, + "step": 6102 + }, + { + "epoch": 2.1429073033707864, + "grad_norm": 0.8884764909744263, + "learning_rate": 3.6113505185456583e-06, + "loss": 0.558, + "step": 6103 + }, + { + "epoch": 2.143258426966292, + "grad_norm": 0.9172934889793396, + "learning_rate": 3.610936367572645e-06, + "loss": 0.6061, + "step": 6104 + }, + { + "epoch": 2.1436095505617976, + "grad_norm": 0.9251722693443298, + "learning_rate": 3.6105221786065588e-06, + "loss": 0.6039, + "step": 6105 + }, + { + "epoch": 2.143960674157303, + "grad_norm": 0.89174884557724, + "learning_rate": 3.610107951661565e-06, + "loss": 0.5838, + "step": 6106 + }, + { + "epoch": 2.144311797752809, + "grad_norm": 0.9081915020942688, + "learning_rate": 3.6096936867518282e-06, + "loss": 0.5718, + "step": 6107 + }, + { + "epoch": 2.1446629213483144, + "grad_norm": 0.8662707805633545, + "learning_rate": 3.6092793838915176e-06, + "loss": 0.5298, + "step": 6108 + }, + { + "epoch": 2.1450140449438204, + "grad_norm": 0.9019919633865356, + "learning_rate": 3.608865043094801e-06, + "loss": 0.5721, + "step": 6109 + }, + { + "epoch": 2.145365168539326, + "grad_norm": 0.8812065124511719, + "learning_rate": 3.608450664375848e-06, + "loss": 0.5574, + "step": 6110 + }, + { + "epoch": 2.1457162921348316, + "grad_norm": 0.8954291939735413, + "learning_rate": 3.6080362477488317e-06, + "loss": 0.5944, + "step": 6111 + }, + { + "epoch": 2.146067415730337, + "grad_norm": 0.8922247886657715, + "learning_rate": 3.607621793227924e-06, + "loss": 0.5472, + "step": 6112 + }, + { + "epoch": 2.146418539325843, + "grad_norm": 0.9212684631347656, + "learning_rate": 3.6072073008272985e-06, + "loss": 0.5921, + "step": 6113 + }, + { + "epoch": 2.1467696629213484, + "grad_norm": 0.885578989982605, + "learning_rate": 3.6067927705611307e-06, + "loss": 0.6381, + "step": 6114 + }, + { + "epoch": 2.147120786516854, + "grad_norm": 0.8882766962051392, + "learning_rate": 3.6063782024435973e-06, + "loss": 0.5891, + "step": 6115 + }, + { + "epoch": 2.1474719101123596, + "grad_norm": 0.8550765514373779, + "learning_rate": 3.605963596488876e-06, + "loss": 0.5548, + "step": 6116 + }, + { + "epoch": 2.147823033707865, + "grad_norm": 0.9154667258262634, + "learning_rate": 3.6055489527111457e-06, + "loss": 0.6034, + "step": 6117 + }, + { + "epoch": 2.148174157303371, + "grad_norm": 0.8833652138710022, + "learning_rate": 3.6051342711245883e-06, + "loss": 0.5415, + "step": 6118 + }, + { + "epoch": 2.1485252808988764, + "grad_norm": 0.9128112196922302, + "learning_rate": 3.6047195517433843e-06, + "loss": 0.5828, + "step": 6119 + }, + { + "epoch": 2.148876404494382, + "grad_norm": 0.9060637950897217, + "learning_rate": 3.6043047945817166e-06, + "loss": 0.569, + "step": 6120 + }, + { + "epoch": 2.1492275280898876, + "grad_norm": 0.9523305296897888, + "learning_rate": 3.6038899996537694e-06, + "loss": 0.5683, + "step": 6121 + }, + { + "epoch": 2.149578651685393, + "grad_norm": 0.9009748101234436, + "learning_rate": 3.6034751669737294e-06, + "loss": 0.6163, + "step": 6122 + }, + { + "epoch": 2.149929775280899, + "grad_norm": 0.8681451678276062, + "learning_rate": 3.603060296555783e-06, + "loss": 0.5632, + "step": 6123 + }, + { + "epoch": 2.1502808988764044, + "grad_norm": 0.862285852432251, + "learning_rate": 3.602645388414118e-06, + "loss": 0.5835, + "step": 6124 + }, + { + "epoch": 2.15063202247191, + "grad_norm": 0.9381813406944275, + "learning_rate": 3.6022304425629246e-06, + "loss": 0.6179, + "step": 6125 + }, + { + "epoch": 2.1509831460674156, + "grad_norm": 0.8743694424629211, + "learning_rate": 3.6018154590163933e-06, + "loss": 0.6282, + "step": 6126 + }, + { + "epoch": 2.151334269662921, + "grad_norm": 0.8931199908256531, + "learning_rate": 3.6014004377887157e-06, + "loss": 0.6114, + "step": 6127 + }, + { + "epoch": 2.151685393258427, + "grad_norm": 0.911408007144928, + "learning_rate": 3.600985378894086e-06, + "loss": 0.5958, + "step": 6128 + }, + { + "epoch": 2.1520365168539324, + "grad_norm": 0.8921440839767456, + "learning_rate": 3.600570282346698e-06, + "loss": 0.5209, + "step": 6129 + }, + { + "epoch": 2.1523876404494384, + "grad_norm": 0.907931923866272, + "learning_rate": 3.600155148160749e-06, + "loss": 0.6049, + "step": 6130 + }, + { + "epoch": 2.152738764044944, + "grad_norm": 0.9091392755508423, + "learning_rate": 3.5997399763504347e-06, + "loss": 0.5786, + "step": 6131 + }, + { + "epoch": 2.1530898876404496, + "grad_norm": 0.9466739296913147, + "learning_rate": 3.5993247669299548e-06, + "loss": 0.6107, + "step": 6132 + }, + { + "epoch": 2.153441011235955, + "grad_norm": 0.8793186545372009, + "learning_rate": 3.5989095199135084e-06, + "loss": 0.5474, + "step": 6133 + }, + { + "epoch": 2.153792134831461, + "grad_norm": 0.8959071636199951, + "learning_rate": 3.5984942353152965e-06, + "loss": 0.5787, + "step": 6134 + }, + { + "epoch": 2.1541432584269664, + "grad_norm": 0.9494472742080688, + "learning_rate": 3.5980789131495216e-06, + "loss": 0.576, + "step": 6135 + }, + { + "epoch": 2.154494382022472, + "grad_norm": 0.9050858616828918, + "learning_rate": 3.5976635534303877e-06, + "loss": 0.5773, + "step": 6136 + }, + { + "epoch": 2.1548455056179776, + "grad_norm": 0.8835017681121826, + "learning_rate": 3.5972481561720996e-06, + "loss": 0.6013, + "step": 6137 + }, + { + "epoch": 2.155196629213483, + "grad_norm": 0.9269439578056335, + "learning_rate": 3.596832721388863e-06, + "loss": 0.5785, + "step": 6138 + }, + { + "epoch": 2.155547752808989, + "grad_norm": 0.9176435470581055, + "learning_rate": 3.596417249094887e-06, + "loss": 0.6115, + "step": 6139 + }, + { + "epoch": 2.1558988764044944, + "grad_norm": 0.9091086387634277, + "learning_rate": 3.596001739304378e-06, + "loss": 0.5874, + "step": 6140 + }, + { + "epoch": 2.15625, + "grad_norm": 0.8805109858512878, + "learning_rate": 3.5955861920315484e-06, + "loss": 0.589, + "step": 6141 + }, + { + "epoch": 2.1566011235955056, + "grad_norm": 0.8851735591888428, + "learning_rate": 3.595170607290608e-06, + "loss": 0.5554, + "step": 6142 + }, + { + "epoch": 2.156952247191011, + "grad_norm": 0.8817130923271179, + "learning_rate": 3.594754985095771e-06, + "loss": 0.5677, + "step": 6143 + }, + { + "epoch": 2.157303370786517, + "grad_norm": 0.8889925479888916, + "learning_rate": 3.594339325461249e-06, + "loss": 0.6157, + "step": 6144 + }, + { + "epoch": 2.1576544943820224, + "grad_norm": 0.9071047306060791, + "learning_rate": 3.593923628401259e-06, + "loss": 0.5944, + "step": 6145 + }, + { + "epoch": 2.158005617977528, + "grad_norm": 0.8781319260597229, + "learning_rate": 3.5935078939300165e-06, + "loss": 0.5748, + "step": 6146 + }, + { + "epoch": 2.1583567415730336, + "grad_norm": 0.8723105788230896, + "learning_rate": 3.59309212206174e-06, + "loss": 0.5553, + "step": 6147 + }, + { + "epoch": 2.158707865168539, + "grad_norm": 0.9141338467597961, + "learning_rate": 3.592676312810649e-06, + "loss": 0.5944, + "step": 6148 + }, + { + "epoch": 2.159058988764045, + "grad_norm": 0.8858033418655396, + "learning_rate": 3.592260466190962e-06, + "loss": 0.5769, + "step": 6149 + }, + { + "epoch": 2.1594101123595504, + "grad_norm": 0.9247735142707825, + "learning_rate": 3.591844582216903e-06, + "loss": 0.5834, + "step": 6150 + }, + { + "epoch": 2.159761235955056, + "grad_norm": 0.9926352500915527, + "learning_rate": 3.5914286609026926e-06, + "loss": 0.6392, + "step": 6151 + }, + { + "epoch": 2.1601123595505616, + "grad_norm": 0.8952130079269409, + "learning_rate": 3.591012702262556e-06, + "loss": 0.595, + "step": 6152 + }, + { + "epoch": 2.1604634831460676, + "grad_norm": 0.8879304528236389, + "learning_rate": 3.5905967063107193e-06, + "loss": 0.5552, + "step": 6153 + }, + { + "epoch": 2.160814606741573, + "grad_norm": 0.8830671310424805, + "learning_rate": 3.5901806730614082e-06, + "loss": 0.5794, + "step": 6154 + }, + { + "epoch": 2.161165730337079, + "grad_norm": 0.9102872610092163, + "learning_rate": 3.58976460252885e-06, + "loss": 0.6047, + "step": 6155 + }, + { + "epoch": 2.1615168539325844, + "grad_norm": 0.9050349593162537, + "learning_rate": 3.5893484947272765e-06, + "loss": 0.6221, + "step": 6156 + }, + { + "epoch": 2.16186797752809, + "grad_norm": 0.8936917185783386, + "learning_rate": 3.588932349670916e-06, + "loss": 0.5804, + "step": 6157 + }, + { + "epoch": 2.1622191011235956, + "grad_norm": 0.8901835083961487, + "learning_rate": 3.588516167374001e-06, + "loss": 0.5841, + "step": 6158 + }, + { + "epoch": 2.162570224719101, + "grad_norm": 0.8509178161621094, + "learning_rate": 3.588099947850764e-06, + "loss": 0.5562, + "step": 6159 + }, + { + "epoch": 2.162921348314607, + "grad_norm": 0.8937607407569885, + "learning_rate": 3.5876836911154402e-06, + "loss": 0.5847, + "step": 6160 + }, + { + "epoch": 2.1632724719101124, + "grad_norm": 0.8953386545181274, + "learning_rate": 3.5872673971822647e-06, + "loss": 0.569, + "step": 6161 + }, + { + "epoch": 2.163623595505618, + "grad_norm": 0.8592950105667114, + "learning_rate": 3.5868510660654753e-06, + "loss": 0.5542, + "step": 6162 + }, + { + "epoch": 2.1639747191011236, + "grad_norm": 0.9137312173843384, + "learning_rate": 3.5864346977793085e-06, + "loss": 0.6069, + "step": 6163 + }, + { + "epoch": 2.164325842696629, + "grad_norm": 0.9160043597221375, + "learning_rate": 3.5860182923380055e-06, + "loss": 0.5832, + "step": 6164 + }, + { + "epoch": 2.164676966292135, + "grad_norm": 0.8983912467956543, + "learning_rate": 3.585601849755806e-06, + "loss": 0.5671, + "step": 6165 + }, + { + "epoch": 2.1650280898876404, + "grad_norm": 0.8892392516136169, + "learning_rate": 3.585185370046952e-06, + "loss": 0.6019, + "step": 6166 + }, + { + "epoch": 2.165379213483146, + "grad_norm": 0.8900330066680908, + "learning_rate": 3.5847688532256876e-06, + "loss": 0.5586, + "step": 6167 + }, + { + "epoch": 2.1657303370786516, + "grad_norm": 0.9106168746948242, + "learning_rate": 3.5843522993062564e-06, + "loss": 0.6131, + "step": 6168 + }, + { + "epoch": 2.166081460674157, + "grad_norm": 0.8671777844429016, + "learning_rate": 3.583935708302904e-06, + "loss": 0.5971, + "step": 6169 + }, + { + "epoch": 2.166432584269663, + "grad_norm": 0.9169990420341492, + "learning_rate": 3.5835190802298782e-06, + "loss": 0.6024, + "step": 6170 + }, + { + "epoch": 2.1667837078651684, + "grad_norm": 0.9220364689826965, + "learning_rate": 3.583102415101427e-06, + "loss": 0.6046, + "step": 6171 + }, + { + "epoch": 2.167134831460674, + "grad_norm": 0.9101770520210266, + "learning_rate": 3.5826857129318e-06, + "loss": 0.5704, + "step": 6172 + }, + { + "epoch": 2.1674859550561796, + "grad_norm": 0.8814703822135925, + "learning_rate": 3.582268973735248e-06, + "loss": 0.5828, + "step": 6173 + }, + { + "epoch": 2.1678370786516856, + "grad_norm": 0.8779802322387695, + "learning_rate": 3.5818521975260232e-06, + "loss": 0.5759, + "step": 6174 + }, + { + "epoch": 2.168188202247191, + "grad_norm": 0.9279386401176453, + "learning_rate": 3.5814353843183795e-06, + "loss": 0.6022, + "step": 6175 + }, + { + "epoch": 2.168539325842697, + "grad_norm": 0.9070867300033569, + "learning_rate": 3.5810185341265706e-06, + "loss": 0.5819, + "step": 6176 + }, + { + "epoch": 2.1688904494382024, + "grad_norm": 0.9013492465019226, + "learning_rate": 3.580601646964853e-06, + "loss": 0.615, + "step": 6177 + }, + { + "epoch": 2.169241573033708, + "grad_norm": 0.9076384902000427, + "learning_rate": 3.5801847228474835e-06, + "loss": 0.5845, + "step": 6178 + }, + { + "epoch": 2.1695926966292136, + "grad_norm": 0.9299958944320679, + "learning_rate": 3.579767761788721e-06, + "loss": 0.565, + "step": 6179 + }, + { + "epoch": 2.169943820224719, + "grad_norm": 0.8838452696800232, + "learning_rate": 3.5793507638028252e-06, + "loss": 0.576, + "step": 6180 + }, + { + "epoch": 2.170294943820225, + "grad_norm": 0.903094470500946, + "learning_rate": 3.5789337289040566e-06, + "loss": 0.5938, + "step": 6181 + }, + { + "epoch": 2.1706460674157304, + "grad_norm": 0.8720085620880127, + "learning_rate": 3.578516657106678e-06, + "loss": 0.5551, + "step": 6182 + }, + { + "epoch": 2.170997191011236, + "grad_norm": 0.8685125708580017, + "learning_rate": 3.5780995484249523e-06, + "loss": 0.587, + "step": 6183 + }, + { + "epoch": 2.1713483146067416, + "grad_norm": 0.8758479356765747, + "learning_rate": 3.577682402873145e-06, + "loss": 0.566, + "step": 6184 + }, + { + "epoch": 2.171699438202247, + "grad_norm": 0.86467444896698, + "learning_rate": 3.5772652204655213e-06, + "loss": 0.5683, + "step": 6185 + }, + { + "epoch": 2.172050561797753, + "grad_norm": 0.9026554226875305, + "learning_rate": 3.5768480012163493e-06, + "loss": 0.5968, + "step": 6186 + }, + { + "epoch": 2.1724016853932584, + "grad_norm": 0.8844003677368164, + "learning_rate": 3.5764307451398966e-06, + "loss": 0.5646, + "step": 6187 + }, + { + "epoch": 2.172752808988764, + "grad_norm": 0.9018101096153259, + "learning_rate": 3.576013452250434e-06, + "loss": 0.6041, + "step": 6188 + }, + { + "epoch": 2.1731039325842696, + "grad_norm": 0.8885400295257568, + "learning_rate": 3.5755961225622315e-06, + "loss": 0.6002, + "step": 6189 + }, + { + "epoch": 2.173455056179775, + "grad_norm": 0.9019367098808289, + "learning_rate": 3.5751787560895624e-06, + "loss": 0.5495, + "step": 6190 + }, + { + "epoch": 2.173806179775281, + "grad_norm": 0.936189591884613, + "learning_rate": 3.5747613528467e-06, + "loss": 0.5908, + "step": 6191 + }, + { + "epoch": 2.1741573033707864, + "grad_norm": 0.9192583560943604, + "learning_rate": 3.5743439128479192e-06, + "loss": 0.6169, + "step": 6192 + }, + { + "epoch": 2.174508426966292, + "grad_norm": 0.8848391771316528, + "learning_rate": 3.5739264361074956e-06, + "loss": 0.5948, + "step": 6193 + }, + { + "epoch": 2.1748595505617976, + "grad_norm": 0.8885852694511414, + "learning_rate": 3.573508922639707e-06, + "loss": 0.5623, + "step": 6194 + }, + { + "epoch": 2.175210674157303, + "grad_norm": 0.8941972255706787, + "learning_rate": 3.5730913724588316e-06, + "loss": 0.579, + "step": 6195 + }, + { + "epoch": 2.175561797752809, + "grad_norm": 0.9446448683738708, + "learning_rate": 3.57267378557915e-06, + "loss": 0.6239, + "step": 6196 + }, + { + "epoch": 2.1759129213483144, + "grad_norm": 0.9064301252365112, + "learning_rate": 3.572256162014942e-06, + "loss": 0.5775, + "step": 6197 + }, + { + "epoch": 2.1762640449438204, + "grad_norm": 0.9025107026100159, + "learning_rate": 3.5718385017804914e-06, + "loss": 0.6134, + "step": 6198 + }, + { + "epoch": 2.176615168539326, + "grad_norm": 0.8902868032455444, + "learning_rate": 3.571420804890081e-06, + "loss": 0.6039, + "step": 6199 + }, + { + "epoch": 2.1769662921348316, + "grad_norm": 0.8915112018585205, + "learning_rate": 3.5710030713579964e-06, + "loss": 0.5742, + "step": 6200 + }, + { + "epoch": 2.177317415730337, + "grad_norm": 0.9017792344093323, + "learning_rate": 3.5705853011985226e-06, + "loss": 0.5951, + "step": 6201 + }, + { + "epoch": 2.177668539325843, + "grad_norm": 0.8672464489936829, + "learning_rate": 3.570167494425948e-06, + "loss": 0.575, + "step": 6202 + }, + { + "epoch": 2.1780196629213484, + "grad_norm": 0.9439303278923035, + "learning_rate": 3.5697496510545605e-06, + "loss": 0.6018, + "step": 6203 + }, + { + "epoch": 2.178370786516854, + "grad_norm": 0.8855968713760376, + "learning_rate": 3.569331771098651e-06, + "loss": 0.5809, + "step": 6204 + }, + { + "epoch": 2.1787219101123596, + "grad_norm": 0.9145450592041016, + "learning_rate": 3.5689138545725086e-06, + "loss": 0.5616, + "step": 6205 + }, + { + "epoch": 2.179073033707865, + "grad_norm": 0.8996074199676514, + "learning_rate": 3.5684959014904288e-06, + "loss": 0.5971, + "step": 6206 + }, + { + "epoch": 2.179424157303371, + "grad_norm": 0.9426374435424805, + "learning_rate": 3.5680779118667025e-06, + "loss": 0.5711, + "step": 6207 + }, + { + "epoch": 2.1797752808988764, + "grad_norm": 0.9067327380180359, + "learning_rate": 3.5676598857156258e-06, + "loss": 0.588, + "step": 6208 + }, + { + "epoch": 2.180126404494382, + "grad_norm": 0.8810411691665649, + "learning_rate": 3.567241823051495e-06, + "loss": 0.5562, + "step": 6209 + }, + { + "epoch": 2.1804775280898876, + "grad_norm": 0.8785190582275391, + "learning_rate": 3.5668237238886062e-06, + "loss": 0.5746, + "step": 6210 + }, + { + "epoch": 2.180828651685393, + "grad_norm": 0.8730765581130981, + "learning_rate": 3.5664055882412598e-06, + "loss": 0.5877, + "step": 6211 + }, + { + "epoch": 2.181179775280899, + "grad_norm": 0.8951026797294617, + "learning_rate": 3.565987416123754e-06, + "loss": 0.629, + "step": 6212 + }, + { + "epoch": 2.1815308988764044, + "grad_norm": 0.8909263610839844, + "learning_rate": 3.565569207550391e-06, + "loss": 0.593, + "step": 6213 + }, + { + "epoch": 2.18188202247191, + "grad_norm": 0.8497523069381714, + "learning_rate": 3.565150962535473e-06, + "loss": 0.5494, + "step": 6214 + }, + { + "epoch": 2.1822331460674156, + "grad_norm": 0.917820155620575, + "learning_rate": 3.5647326810933034e-06, + "loss": 0.5905, + "step": 6215 + }, + { + "epoch": 2.182584269662921, + "grad_norm": 0.8889433145523071, + "learning_rate": 3.564314363238188e-06, + "loss": 0.5727, + "step": 6216 + }, + { + "epoch": 2.182935393258427, + "grad_norm": 0.8928989171981812, + "learning_rate": 3.563896008984431e-06, + "loss": 0.5449, + "step": 6217 + }, + { + "epoch": 2.1832865168539324, + "grad_norm": 0.9180118441581726, + "learning_rate": 3.5634776183463414e-06, + "loss": 0.5765, + "step": 6218 + }, + { + "epoch": 2.1836376404494384, + "grad_norm": 0.9106191992759705, + "learning_rate": 3.5630591913382284e-06, + "loss": 0.5727, + "step": 6219 + }, + { + "epoch": 2.183988764044944, + "grad_norm": 0.9079222679138184, + "learning_rate": 3.562640727974399e-06, + "loss": 0.6213, + "step": 6220 + }, + { + "epoch": 2.1843398876404496, + "grad_norm": 0.8941214680671692, + "learning_rate": 3.5622222282691673e-06, + "loss": 0.5714, + "step": 6221 + }, + { + "epoch": 2.184691011235955, + "grad_norm": 0.8956150412559509, + "learning_rate": 3.5618036922368433e-06, + "loss": 0.5744, + "step": 6222 + }, + { + "epoch": 2.185042134831461, + "grad_norm": 0.8873118162155151, + "learning_rate": 3.561385119891742e-06, + "loss": 0.5972, + "step": 6223 + }, + { + "epoch": 2.1853932584269664, + "grad_norm": 0.9109763503074646, + "learning_rate": 3.5609665112481777e-06, + "loss": 0.5897, + "step": 6224 + }, + { + "epoch": 2.185744382022472, + "grad_norm": 0.9046030044555664, + "learning_rate": 3.5605478663204675e-06, + "loss": 0.6101, + "step": 6225 + }, + { + "epoch": 2.1860955056179776, + "grad_norm": 0.8899408578872681, + "learning_rate": 3.5601291851229264e-06, + "loss": 0.5848, + "step": 6226 + }, + { + "epoch": 2.186446629213483, + "grad_norm": 0.9096876382827759, + "learning_rate": 3.559710467669875e-06, + "loss": 0.5829, + "step": 6227 + }, + { + "epoch": 2.186797752808989, + "grad_norm": 0.8890770077705383, + "learning_rate": 3.5592917139756323e-06, + "loss": 0.5747, + "step": 6228 + }, + { + "epoch": 2.1871488764044944, + "grad_norm": 0.8742604851722717, + "learning_rate": 3.5588729240545197e-06, + "loss": 0.5857, + "step": 6229 + }, + { + "epoch": 2.1875, + "grad_norm": 0.8967490792274475, + "learning_rate": 3.558454097920859e-06, + "loss": 0.6039, + "step": 6230 + }, + { + "epoch": 2.1878511235955056, + "grad_norm": 0.8852705359458923, + "learning_rate": 3.5580352355889743e-06, + "loss": 0.5927, + "step": 6231 + }, + { + "epoch": 2.188202247191011, + "grad_norm": 0.9024866223335266, + "learning_rate": 3.55761633707319e-06, + "loss": 0.6218, + "step": 6232 + }, + { + "epoch": 2.188553370786517, + "grad_norm": 0.892336368560791, + "learning_rate": 3.5571974023878315e-06, + "loss": 0.5624, + "step": 6233 + }, + { + "epoch": 2.1889044943820224, + "grad_norm": 0.9634124040603638, + "learning_rate": 3.5567784315472263e-06, + "loss": 0.5421, + "step": 6234 + }, + { + "epoch": 2.189255617977528, + "grad_norm": 0.8967486619949341, + "learning_rate": 3.556359424565703e-06, + "loss": 0.5791, + "step": 6235 + }, + { + "epoch": 2.1896067415730336, + "grad_norm": 0.9186691045761108, + "learning_rate": 3.5559403814575917e-06, + "loss": 0.6012, + "step": 6236 + }, + { + "epoch": 2.189957865168539, + "grad_norm": 0.9300957918167114, + "learning_rate": 3.5555213022372227e-06, + "loss": 0.5951, + "step": 6237 + }, + { + "epoch": 2.190308988764045, + "grad_norm": 0.885424017906189, + "learning_rate": 3.5551021869189284e-06, + "loss": 0.576, + "step": 6238 + }, + { + "epoch": 2.1906601123595504, + "grad_norm": 0.9140034914016724, + "learning_rate": 3.5546830355170425e-06, + "loss": 0.6073, + "step": 6239 + }, + { + "epoch": 2.191011235955056, + "grad_norm": 0.9048447608947754, + "learning_rate": 3.554263848045899e-06, + "loss": 0.5903, + "step": 6240 + }, + { + "epoch": 2.1913623595505616, + "grad_norm": 0.8971291184425354, + "learning_rate": 3.5538446245198346e-06, + "loss": 0.6148, + "step": 6241 + }, + { + "epoch": 2.1917134831460676, + "grad_norm": 0.8865309357643127, + "learning_rate": 3.5534253649531846e-06, + "loss": 0.5742, + "step": 6242 + }, + { + "epoch": 2.192064606741573, + "grad_norm": 0.9242413640022278, + "learning_rate": 3.5530060693602897e-06, + "loss": 0.5911, + "step": 6243 + }, + { + "epoch": 2.192415730337079, + "grad_norm": 0.9109138250350952, + "learning_rate": 3.552586737755488e-06, + "loss": 0.5888, + "step": 6244 + }, + { + "epoch": 2.1927668539325844, + "grad_norm": 0.9142351150512695, + "learning_rate": 3.5521673701531208e-06, + "loss": 0.5911, + "step": 6245 + }, + { + "epoch": 2.19311797752809, + "grad_norm": 0.8777997493743896, + "learning_rate": 3.5517479665675294e-06, + "loss": 0.5629, + "step": 6246 + }, + { + "epoch": 2.1934691011235956, + "grad_norm": 0.9284225702285767, + "learning_rate": 3.5513285270130575e-06, + "loss": 0.5894, + "step": 6247 + }, + { + "epoch": 2.193820224719101, + "grad_norm": 0.9144864082336426, + "learning_rate": 3.55090905150405e-06, + "loss": 0.5833, + "step": 6248 + }, + { + "epoch": 2.194171348314607, + "grad_norm": 0.8971447348594666, + "learning_rate": 3.5504895400548516e-06, + "loss": 0.5458, + "step": 6249 + }, + { + "epoch": 2.1945224719101124, + "grad_norm": 0.897527277469635, + "learning_rate": 3.5500699926798105e-06, + "loss": 0.5917, + "step": 6250 + }, + { + "epoch": 2.194873595505618, + "grad_norm": 0.9136509895324707, + "learning_rate": 3.5496504093932737e-06, + "loss": 0.5685, + "step": 6251 + }, + { + "epoch": 2.1952247191011236, + "grad_norm": 0.8840371370315552, + "learning_rate": 3.5492307902095915e-06, + "loss": 0.5725, + "step": 6252 + }, + { + "epoch": 2.195575842696629, + "grad_norm": 0.9275133013725281, + "learning_rate": 3.5488111351431143e-06, + "loss": 0.5984, + "step": 6253 + }, + { + "epoch": 2.195926966292135, + "grad_norm": 0.8992714881896973, + "learning_rate": 3.5483914442081936e-06, + "loss": 0.6006, + "step": 6254 + }, + { + "epoch": 2.1962780898876404, + "grad_norm": 0.8837565779685974, + "learning_rate": 3.5479717174191826e-06, + "loss": 0.5559, + "step": 6255 + }, + { + "epoch": 2.196629213483146, + "grad_norm": 0.8913798332214355, + "learning_rate": 3.547551954790435e-06, + "loss": 0.5792, + "step": 6256 + }, + { + "epoch": 2.1969803370786516, + "grad_norm": 0.8821024298667908, + "learning_rate": 3.547132156336308e-06, + "loss": 0.5971, + "step": 6257 + }, + { + "epoch": 2.197331460674157, + "grad_norm": 0.8911494612693787, + "learning_rate": 3.546712322071156e-06, + "loss": 0.6117, + "step": 6258 + }, + { + "epoch": 2.197682584269663, + "grad_norm": 0.8553071022033691, + "learning_rate": 3.5462924520093394e-06, + "loss": 0.5955, + "step": 6259 + }, + { + "epoch": 2.1980337078651684, + "grad_norm": 0.8499958515167236, + "learning_rate": 3.545872546165216e-06, + "loss": 0.5547, + "step": 6260 + }, + { + "epoch": 2.198384831460674, + "grad_norm": 0.9098100066184998, + "learning_rate": 3.5454526045531466e-06, + "loss": 0.6258, + "step": 6261 + }, + { + "epoch": 2.1987359550561796, + "grad_norm": 0.9093878865242004, + "learning_rate": 3.5450326271874923e-06, + "loss": 0.5953, + "step": 6262 + }, + { + "epoch": 2.1990870786516856, + "grad_norm": 0.8861823678016663, + "learning_rate": 3.544612614082617e-06, + "loss": 0.5795, + "step": 6263 + }, + { + "epoch": 2.199438202247191, + "grad_norm": 0.8943526744842529, + "learning_rate": 3.5441925652528837e-06, + "loss": 0.5554, + "step": 6264 + }, + { + "epoch": 2.199789325842697, + "grad_norm": 0.897099494934082, + "learning_rate": 3.5437724807126583e-06, + "loss": 0.6009, + "step": 6265 + }, + { + "epoch": 2.2001404494382024, + "grad_norm": 0.8690956234931946, + "learning_rate": 3.543352360476307e-06, + "loss": 0.5802, + "step": 6266 + }, + { + "epoch": 2.200491573033708, + "grad_norm": 0.9275516271591187, + "learning_rate": 3.542932204558198e-06, + "loss": 0.6329, + "step": 6267 + }, + { + "epoch": 2.2008426966292136, + "grad_norm": 0.9013081789016724, + "learning_rate": 3.5425120129727004e-06, + "loss": 0.5941, + "step": 6268 + }, + { + "epoch": 2.201193820224719, + "grad_norm": 0.9379423260688782, + "learning_rate": 3.542091785734184e-06, + "loss": 0.5739, + "step": 6269 + }, + { + "epoch": 2.201544943820225, + "grad_norm": 0.9083360433578491, + "learning_rate": 3.5416715228570196e-06, + "loss": 0.5658, + "step": 6270 + }, + { + "epoch": 2.2018960674157304, + "grad_norm": 0.8775901794433594, + "learning_rate": 3.5412512243555806e-06, + "loss": 0.598, + "step": 6271 + }, + { + "epoch": 2.202247191011236, + "grad_norm": 0.9183217883110046, + "learning_rate": 3.540830890244241e-06, + "loss": 0.6014, + "step": 6272 + }, + { + "epoch": 2.2025983146067416, + "grad_norm": 0.8978754281997681, + "learning_rate": 3.5404105205373764e-06, + "loss": 0.5744, + "step": 6273 + }, + { + "epoch": 2.202949438202247, + "grad_norm": 0.8864443898200989, + "learning_rate": 3.539990115249361e-06, + "loss": 0.5965, + "step": 6274 + }, + { + "epoch": 2.203300561797753, + "grad_norm": 0.9106681942939758, + "learning_rate": 3.539569674394574e-06, + "loss": 0.6109, + "step": 6275 + }, + { + "epoch": 2.2036516853932584, + "grad_norm": 0.8936038613319397, + "learning_rate": 3.5391491979873937e-06, + "loss": 0.575, + "step": 6276 + }, + { + "epoch": 2.204002808988764, + "grad_norm": 0.8808788657188416, + "learning_rate": 3.5387286860422006e-06, + "loss": 0.5748, + "step": 6277 + }, + { + "epoch": 2.2043539325842696, + "grad_norm": 0.917221188545227, + "learning_rate": 3.538308138573374e-06, + "loss": 0.5923, + "step": 6278 + }, + { + "epoch": 2.204705056179775, + "grad_norm": 0.8884896636009216, + "learning_rate": 3.5378875555952985e-06, + "loss": 0.5703, + "step": 6279 + }, + { + "epoch": 2.205056179775281, + "grad_norm": 0.9145598411560059, + "learning_rate": 3.5374669371223567e-06, + "loss": 0.5852, + "step": 6280 + }, + { + "epoch": 2.2054073033707864, + "grad_norm": 0.9300435185432434, + "learning_rate": 3.5370462831689336e-06, + "loss": 0.5901, + "step": 6281 + }, + { + "epoch": 2.205758426966292, + "grad_norm": 0.9018733501434326, + "learning_rate": 3.5366255937494147e-06, + "loss": 0.6108, + "step": 6282 + }, + { + "epoch": 2.2061095505617976, + "grad_norm": 0.8998925089836121, + "learning_rate": 3.536204868878188e-06, + "loss": 0.5877, + "step": 6283 + }, + { + "epoch": 2.206460674157303, + "grad_norm": 0.8832637071609497, + "learning_rate": 3.535784108569641e-06, + "loss": 0.5792, + "step": 6284 + }, + { + "epoch": 2.206811797752809, + "grad_norm": 0.9105501174926758, + "learning_rate": 3.5353633128381636e-06, + "loss": 0.5932, + "step": 6285 + }, + { + "epoch": 2.2071629213483144, + "grad_norm": 0.9197086095809937, + "learning_rate": 3.534942481698147e-06, + "loss": 0.6299, + "step": 6286 + }, + { + "epoch": 2.2075140449438204, + "grad_norm": 0.9127769470214844, + "learning_rate": 3.534521615163984e-06, + "loss": 0.6066, + "step": 6287 + }, + { + "epoch": 2.207865168539326, + "grad_norm": 0.8922096490859985, + "learning_rate": 3.534100713250066e-06, + "loss": 0.5562, + "step": 6288 + }, + { + "epoch": 2.2082162921348316, + "grad_norm": 0.8825200200080872, + "learning_rate": 3.5336797759707884e-06, + "loss": 0.5383, + "step": 6289 + }, + { + "epoch": 2.208567415730337, + "grad_norm": 0.8893291354179382, + "learning_rate": 3.5332588033405478e-06, + "loss": 0.6128, + "step": 6290 + }, + { + "epoch": 2.208918539325843, + "grad_norm": 0.8805075883865356, + "learning_rate": 3.5328377953737396e-06, + "loss": 0.5795, + "step": 6291 + }, + { + "epoch": 2.2092696629213484, + "grad_norm": 0.897807240486145, + "learning_rate": 3.532416752084763e-06, + "loss": 0.6169, + "step": 6292 + }, + { + "epoch": 2.209620786516854, + "grad_norm": 0.9111034870147705, + "learning_rate": 3.531995673488017e-06, + "loss": 0.6061, + "step": 6293 + }, + { + "epoch": 2.2099719101123596, + "grad_norm": 0.9339807033538818, + "learning_rate": 3.531574559597902e-06, + "loss": 0.5838, + "step": 6294 + }, + { + "epoch": 2.210323033707865, + "grad_norm": 0.9197667837142944, + "learning_rate": 3.5311534104288197e-06, + "loss": 0.5939, + "step": 6295 + }, + { + "epoch": 2.210674157303371, + "grad_norm": 0.8908889889717102, + "learning_rate": 3.5307322259951733e-06, + "loss": 0.5863, + "step": 6296 + }, + { + "epoch": 2.2110252808988764, + "grad_norm": 0.8924925327301025, + "learning_rate": 3.5303110063113666e-06, + "loss": 0.5967, + "step": 6297 + }, + { + "epoch": 2.211376404494382, + "grad_norm": 0.8802773952484131, + "learning_rate": 3.5298897513918053e-06, + "loss": 0.5631, + "step": 6298 + }, + { + "epoch": 2.2117275280898876, + "grad_norm": 0.8912798762321472, + "learning_rate": 3.5294684612508955e-06, + "loss": 0.5937, + "step": 6299 + }, + { + "epoch": 2.212078651685393, + "grad_norm": 0.913508951663971, + "learning_rate": 3.5290471359030455e-06, + "loss": 0.573, + "step": 6300 + }, + { + "epoch": 2.212429775280899, + "grad_norm": 0.8894520998001099, + "learning_rate": 3.528625775362664e-06, + "loss": 0.5674, + "step": 6301 + }, + { + "epoch": 2.2127808988764044, + "grad_norm": 0.8660706877708435, + "learning_rate": 3.5282043796441608e-06, + "loss": 0.5703, + "step": 6302 + }, + { + "epoch": 2.21313202247191, + "grad_norm": 0.9247276782989502, + "learning_rate": 3.527782948761948e-06, + "loss": 0.5961, + "step": 6303 + }, + { + "epoch": 2.2134831460674156, + "grad_norm": 0.8931673765182495, + "learning_rate": 3.5273614827304377e-06, + "loss": 0.5951, + "step": 6304 + }, + { + "epoch": 2.213834269662921, + "grad_norm": 0.8573269844055176, + "learning_rate": 3.5269399815640436e-06, + "loss": 0.5741, + "step": 6305 + }, + { + "epoch": 2.214185393258427, + "grad_norm": 0.8802708387374878, + "learning_rate": 3.5265184452771817e-06, + "loss": 0.573, + "step": 6306 + }, + { + "epoch": 2.2145365168539324, + "grad_norm": 0.8823794722557068, + "learning_rate": 3.5260968738842672e-06, + "loss": 0.6072, + "step": 6307 + }, + { + "epoch": 2.2148876404494384, + "grad_norm": 0.8586814999580383, + "learning_rate": 3.5256752673997175e-06, + "loss": 0.5745, + "step": 6308 + }, + { + "epoch": 2.215238764044944, + "grad_norm": 0.8755496144294739, + "learning_rate": 3.5252536258379512e-06, + "loss": 0.5808, + "step": 6309 + }, + { + "epoch": 2.2155898876404496, + "grad_norm": 0.894632875919342, + "learning_rate": 3.5248319492133882e-06, + "loss": 0.5605, + "step": 6310 + }, + { + "epoch": 2.215941011235955, + "grad_norm": 0.8956295847892761, + "learning_rate": 3.5244102375404492e-06, + "loss": 0.5801, + "step": 6311 + }, + { + "epoch": 2.216292134831461, + "grad_norm": 0.916114866733551, + "learning_rate": 3.523988490833557e-06, + "loss": 0.5773, + "step": 6312 + }, + { + "epoch": 2.2166432584269664, + "grad_norm": 0.852431058883667, + "learning_rate": 3.5235667091071347e-06, + "loss": 0.5577, + "step": 6313 + }, + { + "epoch": 2.216994382022472, + "grad_norm": 0.8810890316963196, + "learning_rate": 3.5231448923756066e-06, + "loss": 0.5845, + "step": 6314 + }, + { + "epoch": 2.2173455056179776, + "grad_norm": 0.8931310772895813, + "learning_rate": 3.5227230406533986e-06, + "loss": 0.5805, + "step": 6315 + }, + { + "epoch": 2.217696629213483, + "grad_norm": 1.0059645175933838, + "learning_rate": 3.5223011539549386e-06, + "loss": 0.6391, + "step": 6316 + }, + { + "epoch": 2.218047752808989, + "grad_norm": 0.8880115151405334, + "learning_rate": 3.521879232294653e-06, + "loss": 0.5484, + "step": 6317 + }, + { + "epoch": 2.2183988764044944, + "grad_norm": 0.8507689237594604, + "learning_rate": 3.5214572756869723e-06, + "loss": 0.5719, + "step": 6318 + }, + { + "epoch": 2.21875, + "grad_norm": 0.9185780882835388, + "learning_rate": 3.521035284146327e-06, + "loss": 0.6043, + "step": 6319 + }, + { + "epoch": 2.2191011235955056, + "grad_norm": 0.9091411828994751, + "learning_rate": 3.5206132576871495e-06, + "loss": 0.619, + "step": 6320 + }, + { + "epoch": 2.219452247191011, + "grad_norm": 0.9296228885650635, + "learning_rate": 3.520191196323871e-06, + "loss": 0.6237, + "step": 6321 + }, + { + "epoch": 2.219803370786517, + "grad_norm": 0.9283947348594666, + "learning_rate": 3.5197691000709266e-06, + "loss": 0.5937, + "step": 6322 + }, + { + "epoch": 2.2201544943820224, + "grad_norm": 0.915845513343811, + "learning_rate": 3.519346968942752e-06, + "loss": 0.5876, + "step": 6323 + }, + { + "epoch": 2.220505617977528, + "grad_norm": 0.9252262711524963, + "learning_rate": 3.5189248029537827e-06, + "loss": 0.6249, + "step": 6324 + }, + { + "epoch": 2.2208567415730336, + "grad_norm": 0.8870640397071838, + "learning_rate": 3.5185026021184577e-06, + "loss": 0.5851, + "step": 6325 + }, + { + "epoch": 2.221207865168539, + "grad_norm": 0.914588212966919, + "learning_rate": 3.518080366451214e-06, + "loss": 0.6134, + "step": 6326 + }, + { + "epoch": 2.221558988764045, + "grad_norm": 0.8884751796722412, + "learning_rate": 3.517658095966494e-06, + "loss": 0.6021, + "step": 6327 + }, + { + "epoch": 2.2219101123595504, + "grad_norm": 0.9763365983963013, + "learning_rate": 3.517235790678738e-06, + "loss": 0.6007, + "step": 6328 + }, + { + "epoch": 2.222261235955056, + "grad_norm": 0.8730164170265198, + "learning_rate": 3.516813450602388e-06, + "loss": 0.5918, + "step": 6329 + }, + { + "epoch": 2.2226123595505616, + "grad_norm": 0.8899509906768799, + "learning_rate": 3.516391075751889e-06, + "loss": 0.5935, + "step": 6330 + }, + { + "epoch": 2.2229634831460676, + "grad_norm": 0.9134078621864319, + "learning_rate": 3.5159686661416836e-06, + "loss": 0.6109, + "step": 6331 + }, + { + "epoch": 2.223314606741573, + "grad_norm": 0.9032573103904724, + "learning_rate": 3.5155462217862205e-06, + "loss": 0.5821, + "step": 6332 + }, + { + "epoch": 2.223665730337079, + "grad_norm": 0.9035776853561401, + "learning_rate": 3.5151237426999452e-06, + "loss": 0.5922, + "step": 6333 + }, + { + "epoch": 2.2240168539325844, + "grad_norm": 0.9185246825218201, + "learning_rate": 3.5147012288973067e-06, + "loss": 0.6102, + "step": 6334 + }, + { + "epoch": 2.22436797752809, + "grad_norm": 0.8820770978927612, + "learning_rate": 3.5142786803927536e-06, + "loss": 0.5642, + "step": 6335 + }, + { + "epoch": 2.2247191011235956, + "grad_norm": 0.9068747162818909, + "learning_rate": 3.5138560972007386e-06, + "loss": 0.5511, + "step": 6336 + }, + { + "epoch": 2.225070224719101, + "grad_norm": 0.9120933413505554, + "learning_rate": 3.5134334793357115e-06, + "loss": 0.5543, + "step": 6337 + }, + { + "epoch": 2.225421348314607, + "grad_norm": 0.9381104707717896, + "learning_rate": 3.5130108268121277e-06, + "loss": 0.6192, + "step": 6338 + }, + { + "epoch": 2.2257724719101124, + "grad_norm": 0.9110600352287292, + "learning_rate": 3.5125881396444395e-06, + "loss": 0.6079, + "step": 6339 + }, + { + "epoch": 2.226123595505618, + "grad_norm": 0.8691093325614929, + "learning_rate": 3.5121654178471043e-06, + "loss": 0.5461, + "step": 6340 + }, + { + "epoch": 2.2264747191011236, + "grad_norm": 0.975554347038269, + "learning_rate": 3.5117426614345777e-06, + "loss": 0.6036, + "step": 6341 + }, + { + "epoch": 2.226825842696629, + "grad_norm": 0.8655807971954346, + "learning_rate": 3.511319870421318e-06, + "loss": 0.5366, + "step": 6342 + }, + { + "epoch": 2.227176966292135, + "grad_norm": 0.8993242383003235, + "learning_rate": 3.5108970448217837e-06, + "loss": 0.5986, + "step": 6343 + }, + { + "epoch": 2.2275280898876404, + "grad_norm": 0.8908608555793762, + "learning_rate": 3.5104741846504354e-06, + "loss": 0.5785, + "step": 6344 + }, + { + "epoch": 2.227879213483146, + "grad_norm": 0.9227836728096008, + "learning_rate": 3.5100512899217348e-06, + "loss": 0.6036, + "step": 6345 + }, + { + "epoch": 2.2282303370786516, + "grad_norm": 0.923088788986206, + "learning_rate": 3.5096283606501446e-06, + "loss": 0.596, + "step": 6346 + }, + { + "epoch": 2.228581460674157, + "grad_norm": 0.8904281854629517, + "learning_rate": 3.5092053968501283e-06, + "loss": 0.5714, + "step": 6347 + }, + { + "epoch": 2.228932584269663, + "grad_norm": 0.8758519887924194, + "learning_rate": 3.508782398536151e-06, + "loss": 0.5511, + "step": 6348 + }, + { + "epoch": 2.2292837078651684, + "grad_norm": 0.903537929058075, + "learning_rate": 3.5083593657226792e-06, + "loss": 0.5929, + "step": 6349 + }, + { + "epoch": 2.229634831460674, + "grad_norm": 0.9074186682701111, + "learning_rate": 3.5079362984241798e-06, + "loss": 0.6221, + "step": 6350 + }, + { + "epoch": 2.2299859550561796, + "grad_norm": 0.899612545967102, + "learning_rate": 3.5075131966551214e-06, + "loss": 0.5757, + "step": 6351 + }, + { + "epoch": 2.2303370786516856, + "grad_norm": 0.8896338939666748, + "learning_rate": 3.507090060429974e-06, + "loss": 0.5706, + "step": 6352 + }, + { + "epoch": 2.230688202247191, + "grad_norm": 0.8950545191764832, + "learning_rate": 3.506666889763208e-06, + "loss": 0.5956, + "step": 6353 + }, + { + "epoch": 2.231039325842697, + "grad_norm": 0.8823170065879822, + "learning_rate": 3.506243684669296e-06, + "loss": 0.5674, + "step": 6354 + }, + { + "epoch": 2.2313904494382024, + "grad_norm": 0.9043381810188293, + "learning_rate": 3.505820445162711e-06, + "loss": 0.5947, + "step": 6355 + }, + { + "epoch": 2.231741573033708, + "grad_norm": 0.8928703665733337, + "learning_rate": 3.505397171257927e-06, + "loss": 0.5605, + "step": 6356 + }, + { + "epoch": 2.2320926966292136, + "grad_norm": 0.8755289316177368, + "learning_rate": 3.504973862969421e-06, + "loss": 0.5634, + "step": 6357 + }, + { + "epoch": 2.232443820224719, + "grad_norm": 0.8837035298347473, + "learning_rate": 3.5045505203116683e-06, + "loss": 0.5655, + "step": 6358 + }, + { + "epoch": 2.232794943820225, + "grad_norm": 0.8875620365142822, + "learning_rate": 3.5041271432991474e-06, + "loss": 0.613, + "step": 6359 + }, + { + "epoch": 2.2331460674157304, + "grad_norm": 0.8647816181182861, + "learning_rate": 3.5037037319463375e-06, + "loss": 0.5351, + "step": 6360 + }, + { + "epoch": 2.233497191011236, + "grad_norm": 0.8917087912559509, + "learning_rate": 3.5032802862677186e-06, + "loss": 0.5822, + "step": 6361 + }, + { + "epoch": 2.2338483146067416, + "grad_norm": 0.9301523566246033, + "learning_rate": 3.5028568062777728e-06, + "loss": 0.6171, + "step": 6362 + }, + { + "epoch": 2.234199438202247, + "grad_norm": 0.8970189094543457, + "learning_rate": 3.502433291990982e-06, + "loss": 0.5721, + "step": 6363 + }, + { + "epoch": 2.234550561797753, + "grad_norm": 0.8678383827209473, + "learning_rate": 3.502009743421831e-06, + "loss": 0.5813, + "step": 6364 + }, + { + "epoch": 2.2349016853932584, + "grad_norm": 0.9084876775741577, + "learning_rate": 3.5015861605848032e-06, + "loss": 0.5982, + "step": 6365 + }, + { + "epoch": 2.235252808988764, + "grad_norm": 0.8969109058380127, + "learning_rate": 3.501162543494387e-06, + "loss": 0.5774, + "step": 6366 + }, + { + "epoch": 2.2356039325842696, + "grad_norm": 1.1410881280899048, + "learning_rate": 3.5007388921650675e-06, + "loss": 0.5559, + "step": 6367 + }, + { + "epoch": 2.235955056179775, + "grad_norm": 0.9038795828819275, + "learning_rate": 3.500315206611335e-06, + "loss": 0.5963, + "step": 6368 + }, + { + "epoch": 2.236306179775281, + "grad_norm": 0.9086941480636597, + "learning_rate": 3.499891486847678e-06, + "loss": 0.5833, + "step": 6369 + }, + { + "epoch": 2.2366573033707864, + "grad_norm": 0.8891550302505493, + "learning_rate": 3.4994677328885883e-06, + "loss": 0.5711, + "step": 6370 + }, + { + "epoch": 2.237008426966292, + "grad_norm": 1.0190508365631104, + "learning_rate": 3.499043944748557e-06, + "loss": 0.6301, + "step": 6371 + }, + { + "epoch": 2.2373595505617976, + "grad_norm": 0.9191010594367981, + "learning_rate": 3.498620122442078e-06, + "loss": 0.6103, + "step": 6372 + }, + { + "epoch": 2.237710674157303, + "grad_norm": 0.8736655116081238, + "learning_rate": 3.4981962659836442e-06, + "loss": 0.5488, + "step": 6373 + }, + { + "epoch": 2.238061797752809, + "grad_norm": 0.8848807215690613, + "learning_rate": 3.4977723753877535e-06, + "loss": 0.5682, + "step": 6374 + }, + { + "epoch": 2.2384129213483144, + "grad_norm": 0.9219880104064941, + "learning_rate": 3.497348450668901e-06, + "loss": 0.6126, + "step": 6375 + }, + { + "epoch": 2.2387640449438204, + "grad_norm": 0.8889185190200806, + "learning_rate": 3.4969244918415848e-06, + "loss": 0.5713, + "step": 6376 + }, + { + "epoch": 2.239115168539326, + "grad_norm": 0.9257709383964539, + "learning_rate": 3.496500498920304e-06, + "loss": 0.618, + "step": 6377 + }, + { + "epoch": 2.2394662921348316, + "grad_norm": 0.9121856093406677, + "learning_rate": 3.496076471919559e-06, + "loss": 0.5819, + "step": 6378 + }, + { + "epoch": 2.239817415730337, + "grad_norm": 0.899768054485321, + "learning_rate": 3.495652410853851e-06, + "loss": 0.5757, + "step": 6379 + }, + { + "epoch": 2.240168539325843, + "grad_norm": 0.9038087129592896, + "learning_rate": 3.4952283157376826e-06, + "loss": 0.5887, + "step": 6380 + }, + { + "epoch": 2.2405196629213484, + "grad_norm": 0.9292892217636108, + "learning_rate": 3.494804186585557e-06, + "loss": 0.5843, + "step": 6381 + }, + { + "epoch": 2.240870786516854, + "grad_norm": 0.8868176937103271, + "learning_rate": 3.4943800234119803e-06, + "loss": 0.5908, + "step": 6382 + }, + { + "epoch": 2.2412219101123596, + "grad_norm": 0.917927086353302, + "learning_rate": 3.4939558262314577e-06, + "loss": 0.5929, + "step": 6383 + }, + { + "epoch": 2.241573033707865, + "grad_norm": 0.9159899353981018, + "learning_rate": 3.4935315950584957e-06, + "loss": 0.5786, + "step": 6384 + }, + { + "epoch": 2.241924157303371, + "grad_norm": 0.867624819278717, + "learning_rate": 3.4931073299076033e-06, + "loss": 0.6085, + "step": 6385 + }, + { + "epoch": 2.2422752808988764, + "grad_norm": 0.8946213126182556, + "learning_rate": 3.49268303079329e-06, + "loss": 0.5876, + "step": 6386 + }, + { + "epoch": 2.242626404494382, + "grad_norm": 0.9105337262153625, + "learning_rate": 3.492258697730067e-06, + "loss": 0.5804, + "step": 6387 + }, + { + "epoch": 2.2429775280898876, + "grad_norm": 0.8662266135215759, + "learning_rate": 3.4918343307324455e-06, + "loss": 0.5644, + "step": 6388 + }, + { + "epoch": 2.243328651685393, + "grad_norm": 0.9749018549919128, + "learning_rate": 3.4914099298149385e-06, + "loss": 0.6218, + "step": 6389 + }, + { + "epoch": 2.243679775280899, + "grad_norm": 0.8880422115325928, + "learning_rate": 3.49098549499206e-06, + "loss": 0.5894, + "step": 6390 + }, + { + "epoch": 2.2440308988764044, + "grad_norm": 0.8963291049003601, + "learning_rate": 3.490561026278326e-06, + "loss": 0.5895, + "step": 6391 + }, + { + "epoch": 2.24438202247191, + "grad_norm": 0.9514265656471252, + "learning_rate": 3.4901365236882523e-06, + "loss": 0.6194, + "step": 6392 + }, + { + "epoch": 2.2447331460674156, + "grad_norm": 0.9134100079536438, + "learning_rate": 3.489711987236357e-06, + "loss": 0.5704, + "step": 6393 + }, + { + "epoch": 2.245084269662921, + "grad_norm": 0.8674129247665405, + "learning_rate": 3.489287416937159e-06, + "loss": 0.5555, + "step": 6394 + }, + { + "epoch": 2.245435393258427, + "grad_norm": 0.9079374670982361, + "learning_rate": 3.488862812805177e-06, + "loss": 0.602, + "step": 6395 + }, + { + "epoch": 2.2457865168539324, + "grad_norm": 0.9119422435760498, + "learning_rate": 3.4884381748549338e-06, + "loss": 0.5787, + "step": 6396 + }, + { + "epoch": 2.2461376404494384, + "grad_norm": 0.9023412466049194, + "learning_rate": 3.4880135031009505e-06, + "loss": 0.5871, + "step": 6397 + }, + { + "epoch": 2.246488764044944, + "grad_norm": 0.9588386416435242, + "learning_rate": 3.4875887975577506e-06, + "loss": 0.6162, + "step": 6398 + }, + { + "epoch": 2.2468398876404496, + "grad_norm": 0.9681424498558044, + "learning_rate": 3.487164058239859e-06, + "loss": 0.638, + "step": 6399 + }, + { + "epoch": 2.247191011235955, + "grad_norm": 0.9108127355575562, + "learning_rate": 3.4867392851618016e-06, + "loss": 0.6089, + "step": 6400 + }, + { + "epoch": 2.247542134831461, + "grad_norm": 0.8878898024559021, + "learning_rate": 3.486314478338105e-06, + "loss": 0.6078, + "step": 6401 + }, + { + "epoch": 2.2478932584269664, + "grad_norm": 0.8697559237480164, + "learning_rate": 3.4858896377832966e-06, + "loss": 0.5717, + "step": 6402 + }, + { + "epoch": 2.248244382022472, + "grad_norm": 0.9023134708404541, + "learning_rate": 3.4854647635119066e-06, + "loss": 0.6072, + "step": 6403 + }, + { + "epoch": 2.2485955056179776, + "grad_norm": 0.9089261293411255, + "learning_rate": 3.485039855538465e-06, + "loss": 0.5761, + "step": 6404 + }, + { + "epoch": 2.248946629213483, + "grad_norm": 0.8970053791999817, + "learning_rate": 3.484614913877503e-06, + "loss": 0.6064, + "step": 6405 + }, + { + "epoch": 2.249297752808989, + "grad_norm": 0.9035254716873169, + "learning_rate": 3.484189938543554e-06, + "loss": 0.5714, + "step": 6406 + }, + { + "epoch": 2.2496488764044944, + "grad_norm": 0.8840333819389343, + "learning_rate": 3.483764929551151e-06, + "loss": 0.5572, + "step": 6407 + }, + { + "epoch": 2.25, + "grad_norm": 0.8977181315422058, + "learning_rate": 3.48333988691483e-06, + "loss": 0.5737, + "step": 6408 + }, + { + "epoch": 2.2503511235955056, + "grad_norm": 0.8934135437011719, + "learning_rate": 3.4829148106491257e-06, + "loss": 0.6128, + "step": 6409 + }, + { + "epoch": 2.250702247191011, + "grad_norm": 0.8992844820022583, + "learning_rate": 3.482489700768576e-06, + "loss": 0.6048, + "step": 6410 + }, + { + "epoch": 2.251053370786517, + "grad_norm": 0.8466454148292542, + "learning_rate": 3.4820645572877186e-06, + "loss": 0.5575, + "step": 6411 + }, + { + "epoch": 2.2514044943820224, + "grad_norm": 0.8654247522354126, + "learning_rate": 3.4816393802210946e-06, + "loss": 0.5646, + "step": 6412 + }, + { + "epoch": 2.251755617977528, + "grad_norm": 0.8887550234794617, + "learning_rate": 3.4812141695832434e-06, + "loss": 0.5987, + "step": 6413 + }, + { + "epoch": 2.2521067415730336, + "grad_norm": 0.9065608382225037, + "learning_rate": 3.4807889253887074e-06, + "loss": 0.5995, + "step": 6414 + }, + { + "epoch": 2.252457865168539, + "grad_norm": 0.8965268135070801, + "learning_rate": 3.480363647652029e-06, + "loss": 0.5742, + "step": 6415 + }, + { + "epoch": 2.252808988764045, + "grad_norm": 0.909958004951477, + "learning_rate": 3.4799383363877535e-06, + "loss": 0.5781, + "step": 6416 + }, + { + "epoch": 2.2531601123595504, + "grad_norm": 0.8881512880325317, + "learning_rate": 3.4795129916104246e-06, + "loss": 0.5312, + "step": 6417 + }, + { + "epoch": 2.253511235955056, + "grad_norm": 0.8759739995002747, + "learning_rate": 3.47908761333459e-06, + "loss": 0.5637, + "step": 6418 + }, + { + "epoch": 2.2538623595505616, + "grad_norm": 0.8761715292930603, + "learning_rate": 3.4786622015747972e-06, + "loss": 0.5604, + "step": 6419 + }, + { + "epoch": 2.254213483146067, + "grad_norm": 0.933495283126831, + "learning_rate": 3.4782367563455945e-06, + "loss": 0.5896, + "step": 6420 + }, + { + "epoch": 2.254564606741573, + "grad_norm": 0.9232785105705261, + "learning_rate": 3.4778112776615316e-06, + "loss": 0.5848, + "step": 6421 + }, + { + "epoch": 2.254915730337079, + "grad_norm": 0.8967583775520325, + "learning_rate": 3.4773857655371594e-06, + "loss": 0.5968, + "step": 6422 + }, + { + "epoch": 2.2552668539325844, + "grad_norm": 0.8670732975006104, + "learning_rate": 3.476960219987031e-06, + "loss": 0.579, + "step": 6423 + }, + { + "epoch": 2.25561797752809, + "grad_norm": 0.9020293951034546, + "learning_rate": 3.4765346410256984e-06, + "loss": 0.5793, + "step": 6424 + }, + { + "epoch": 2.2559691011235956, + "grad_norm": 0.9154753684997559, + "learning_rate": 3.4761090286677167e-06, + "loss": 0.5819, + "step": 6425 + }, + { + "epoch": 2.256320224719101, + "grad_norm": 0.8987799286842346, + "learning_rate": 3.475683382927642e-06, + "loss": 0.5599, + "step": 6426 + }, + { + "epoch": 2.256671348314607, + "grad_norm": 0.881934404373169, + "learning_rate": 3.475257703820031e-06, + "loss": 0.5876, + "step": 6427 + }, + { + "epoch": 2.2570224719101124, + "grad_norm": 0.8703842163085938, + "learning_rate": 3.4748319913594404e-06, + "loss": 0.5735, + "step": 6428 + }, + { + "epoch": 2.257373595505618, + "grad_norm": 0.888954758644104, + "learning_rate": 3.47440624556043e-06, + "loss": 0.5774, + "step": 6429 + }, + { + "epoch": 2.2577247191011236, + "grad_norm": 0.9289371967315674, + "learning_rate": 3.473980466437559e-06, + "loss": 0.5789, + "step": 6430 + }, + { + "epoch": 2.258075842696629, + "grad_norm": 0.8648634552955627, + "learning_rate": 3.4735546540053906e-06, + "loss": 0.5908, + "step": 6431 + }, + { + "epoch": 2.258426966292135, + "grad_norm": 0.9237593412399292, + "learning_rate": 3.473128808278485e-06, + "loss": 0.6075, + "step": 6432 + }, + { + "epoch": 2.2587780898876404, + "grad_norm": 0.9287500977516174, + "learning_rate": 3.472702929271408e-06, + "loss": 0.6062, + "step": 6433 + }, + { + "epoch": 2.259129213483146, + "grad_norm": 0.8851912617683411, + "learning_rate": 3.472277016998723e-06, + "loss": 0.5821, + "step": 6434 + }, + { + "epoch": 2.2594803370786516, + "grad_norm": 0.8714771270751953, + "learning_rate": 3.471851071474996e-06, + "loss": 0.5556, + "step": 6435 + }, + { + "epoch": 2.259831460674157, + "grad_norm": 0.8845500349998474, + "learning_rate": 3.4714250927147936e-06, + "loss": 0.5613, + "step": 6436 + }, + { + "epoch": 2.260182584269663, + "grad_norm": 0.8737462759017944, + "learning_rate": 3.4709990807326843e-06, + "loss": 0.5707, + "step": 6437 + }, + { + "epoch": 2.2605337078651684, + "grad_norm": 0.9052067399024963, + "learning_rate": 3.4705730355432378e-06, + "loss": 0.5882, + "step": 6438 + }, + { + "epoch": 2.260884831460674, + "grad_norm": 0.895591139793396, + "learning_rate": 3.4701469571610235e-06, + "loss": 0.5612, + "step": 6439 + }, + { + "epoch": 2.26123595505618, + "grad_norm": 0.9322772026062012, + "learning_rate": 3.469720845600614e-06, + "loss": 0.6033, + "step": 6440 + }, + { + "epoch": 2.2615870786516856, + "grad_norm": 0.8765947818756104, + "learning_rate": 3.4692947008765802e-06, + "loss": 0.5671, + "step": 6441 + }, + { + "epoch": 2.261938202247191, + "grad_norm": 0.8841614723205566, + "learning_rate": 3.468868523003498e-06, + "loss": 0.5815, + "step": 6442 + }, + { + "epoch": 2.262289325842697, + "grad_norm": 0.8994764685630798, + "learning_rate": 3.468442311995941e-06, + "loss": 0.582, + "step": 6443 + }, + { + "epoch": 2.2626404494382024, + "grad_norm": 0.9189562201499939, + "learning_rate": 3.468016067868486e-06, + "loss": 0.5645, + "step": 6444 + }, + { + "epoch": 2.262991573033708, + "grad_norm": 0.8909952044487, + "learning_rate": 3.4675897906357097e-06, + "loss": 0.5816, + "step": 6445 + }, + { + "epoch": 2.2633426966292136, + "grad_norm": 0.8972160220146179, + "learning_rate": 3.467163480312191e-06, + "loss": 0.6034, + "step": 6446 + }, + { + "epoch": 2.263693820224719, + "grad_norm": 0.8866010308265686, + "learning_rate": 3.4667371369125084e-06, + "loss": 0.5728, + "step": 6447 + }, + { + "epoch": 2.264044943820225, + "grad_norm": 0.8894676566123962, + "learning_rate": 3.466310760451243e-06, + "loss": 0.5612, + "step": 6448 + }, + { + "epoch": 2.2643960674157304, + "grad_norm": 0.9047231078147888, + "learning_rate": 3.4658843509429763e-06, + "loss": 0.5866, + "step": 6449 + }, + { + "epoch": 2.264747191011236, + "grad_norm": 0.9167290329933167, + "learning_rate": 3.4654579084022914e-06, + "loss": 0.6395, + "step": 6450 + }, + { + "epoch": 2.2650983146067416, + "grad_norm": 0.9025495648384094, + "learning_rate": 3.4650314328437722e-06, + "loss": 0.5827, + "step": 6451 + }, + { + "epoch": 2.265449438202247, + "grad_norm": 0.8762118816375732, + "learning_rate": 3.464604924282004e-06, + "loss": 0.5732, + "step": 6452 + }, + { + "epoch": 2.265800561797753, + "grad_norm": 0.9023401737213135, + "learning_rate": 3.464178382731572e-06, + "loss": 0.588, + "step": 6453 + }, + { + "epoch": 2.2661516853932584, + "grad_norm": 0.9782307744026184, + "learning_rate": 3.4637518082070655e-06, + "loss": 0.5561, + "step": 6454 + }, + { + "epoch": 2.266502808988764, + "grad_norm": 0.9176931977272034, + "learning_rate": 3.4633252007230712e-06, + "loss": 0.6039, + "step": 6455 + }, + { + "epoch": 2.2668539325842696, + "grad_norm": 0.9114319682121277, + "learning_rate": 3.4628985602941802e-06, + "loss": 0.5778, + "step": 6456 + }, + { + "epoch": 2.267205056179775, + "grad_norm": 0.8781710863113403, + "learning_rate": 3.462471886934981e-06, + "loss": 0.5727, + "step": 6457 + }, + { + "epoch": 2.267556179775281, + "grad_norm": 0.91740483045578, + "learning_rate": 3.4620451806600687e-06, + "loss": 0.587, + "step": 6458 + }, + { + "epoch": 2.2679073033707864, + "grad_norm": 0.8923473954200745, + "learning_rate": 3.461618441484033e-06, + "loss": 0.5956, + "step": 6459 + }, + { + "epoch": 2.268258426966292, + "grad_norm": 0.8914666771888733, + "learning_rate": 3.4611916694214698e-06, + "loss": 0.6056, + "step": 6460 + }, + { + "epoch": 2.2686095505617976, + "grad_norm": 0.8831369280815125, + "learning_rate": 3.460764864486974e-06, + "loss": 0.5514, + "step": 6461 + }, + { + "epoch": 2.268960674157303, + "grad_norm": 0.8803779482841492, + "learning_rate": 3.460338026695142e-06, + "loss": 0.5793, + "step": 6462 + }, + { + "epoch": 2.269311797752809, + "grad_norm": 0.8947704434394836, + "learning_rate": 3.4599111560605712e-06, + "loss": 0.5576, + "step": 6463 + }, + { + "epoch": 2.2696629213483144, + "grad_norm": 0.8768846988677979, + "learning_rate": 3.45948425259786e-06, + "loss": 0.5646, + "step": 6464 + }, + { + "epoch": 2.27001404494382, + "grad_norm": 0.8953551054000854, + "learning_rate": 3.459057316321609e-06, + "loss": 0.6257, + "step": 6465 + }, + { + "epoch": 2.270365168539326, + "grad_norm": 0.9006490707397461, + "learning_rate": 3.4586303472464183e-06, + "loss": 0.5982, + "step": 6466 + }, + { + "epoch": 2.2707162921348316, + "grad_norm": 0.9017635583877563, + "learning_rate": 3.4582033453868903e-06, + "loss": 0.591, + "step": 6467 + }, + { + "epoch": 2.271067415730337, + "grad_norm": 0.9035273194313049, + "learning_rate": 3.457776310757627e-06, + "loss": 0.606, + "step": 6468 + }, + { + "epoch": 2.271418539325843, + "grad_norm": 0.8860619068145752, + "learning_rate": 3.4573492433732336e-06, + "loss": 0.5719, + "step": 6469 + }, + { + "epoch": 2.2717696629213484, + "grad_norm": 0.9756413698196411, + "learning_rate": 3.4569221432483156e-06, + "loss": 0.6138, + "step": 6470 + }, + { + "epoch": 2.272120786516854, + "grad_norm": 0.9026461243629456, + "learning_rate": 3.4564950103974797e-06, + "loss": 0.5942, + "step": 6471 + }, + { + "epoch": 2.2724719101123596, + "grad_norm": 0.9319086074829102, + "learning_rate": 3.4560678448353325e-06, + "loss": 0.6151, + "step": 6472 + }, + { + "epoch": 2.272823033707865, + "grad_norm": 0.904626727104187, + "learning_rate": 3.455640646576483e-06, + "loss": 0.5969, + "step": 6473 + }, + { + "epoch": 2.273174157303371, + "grad_norm": 0.8865272998809814, + "learning_rate": 3.455213415635541e-06, + "loss": 0.5651, + "step": 6474 + }, + { + "epoch": 2.2735252808988764, + "grad_norm": 0.8737308979034424, + "learning_rate": 3.4547861520271183e-06, + "loss": 0.5899, + "step": 6475 + }, + { + "epoch": 2.273876404494382, + "grad_norm": 0.9398338794708252, + "learning_rate": 3.454358855765825e-06, + "loss": 0.6428, + "step": 6476 + }, + { + "epoch": 2.2742275280898876, + "grad_norm": 0.8904305100440979, + "learning_rate": 3.4539315268662766e-06, + "loss": 0.5835, + "step": 6477 + }, + { + "epoch": 2.274578651685393, + "grad_norm": 0.876357913017273, + "learning_rate": 3.453504165343086e-06, + "loss": 0.5608, + "step": 6478 + }, + { + "epoch": 2.274929775280899, + "grad_norm": 0.933561384677887, + "learning_rate": 3.453076771210868e-06, + "loss": 0.5958, + "step": 6479 + }, + { + "epoch": 2.2752808988764044, + "grad_norm": 0.931567370891571, + "learning_rate": 3.452649344484241e-06, + "loss": 0.6323, + "step": 6480 + }, + { + "epoch": 2.27563202247191, + "grad_norm": 0.8942869305610657, + "learning_rate": 3.4522218851778213e-06, + "loss": 0.562, + "step": 6481 + }, + { + "epoch": 2.2759831460674156, + "grad_norm": 0.888300895690918, + "learning_rate": 3.4517943933062283e-06, + "loss": 0.5885, + "step": 6482 + }, + { + "epoch": 2.276334269662921, + "grad_norm": 0.896180272102356, + "learning_rate": 3.4513668688840807e-06, + "loss": 0.5724, + "step": 6483 + }, + { + "epoch": 2.276685393258427, + "grad_norm": 0.90582674741745, + "learning_rate": 3.4509393119260017e-06, + "loss": 0.5787, + "step": 6484 + }, + { + "epoch": 2.277036516853933, + "grad_norm": 0.9344244003295898, + "learning_rate": 3.4505117224466106e-06, + "loss": 0.6177, + "step": 6485 + }, + { + "epoch": 2.2773876404494384, + "grad_norm": 0.926701009273529, + "learning_rate": 3.4500841004605324e-06, + "loss": 0.595, + "step": 6486 + }, + { + "epoch": 2.277738764044944, + "grad_norm": 0.8885989785194397, + "learning_rate": 3.4496564459823906e-06, + "loss": 0.5569, + "step": 6487 + }, + { + "epoch": 2.2780898876404496, + "grad_norm": 0.8861546516418457, + "learning_rate": 3.4492287590268113e-06, + "loss": 0.5612, + "step": 6488 + }, + { + "epoch": 2.278441011235955, + "grad_norm": 0.9039327502250671, + "learning_rate": 3.4488010396084208e-06, + "loss": 0.6159, + "step": 6489 + }, + { + "epoch": 2.278792134831461, + "grad_norm": 0.9324183464050293, + "learning_rate": 3.4483732877418464e-06, + "loss": 0.5991, + "step": 6490 + }, + { + "epoch": 2.2791432584269664, + "grad_norm": 0.9026148915290833, + "learning_rate": 3.4479455034417176e-06, + "loss": 0.5405, + "step": 6491 + }, + { + "epoch": 2.279494382022472, + "grad_norm": 0.9265632033348083, + "learning_rate": 3.4475176867226635e-06, + "loss": 0.58, + "step": 6492 + }, + { + "epoch": 2.2798455056179776, + "grad_norm": 0.887611448764801, + "learning_rate": 3.447089837599315e-06, + "loss": 0.6065, + "step": 6493 + }, + { + "epoch": 2.280196629213483, + "grad_norm": 0.9507808685302734, + "learning_rate": 3.4466619560863046e-06, + "loss": 0.591, + "step": 6494 + }, + { + "epoch": 2.280547752808989, + "grad_norm": 0.9046680927276611, + "learning_rate": 3.4462340421982656e-06, + "loss": 0.6161, + "step": 6495 + }, + { + "epoch": 2.2808988764044944, + "grad_norm": 0.9269330501556396, + "learning_rate": 3.4458060959498326e-06, + "loss": 0.6139, + "step": 6496 + }, + { + "epoch": 2.28125, + "grad_norm": 0.8850961327552795, + "learning_rate": 3.4453781173556393e-06, + "loss": 0.5721, + "step": 6497 + }, + { + "epoch": 2.2816011235955056, + "grad_norm": 0.9026087522506714, + "learning_rate": 3.4449501064303238e-06, + "loss": 0.5888, + "step": 6498 + }, + { + "epoch": 2.281952247191011, + "grad_norm": 0.8796538710594177, + "learning_rate": 3.4445220631885236e-06, + "loss": 0.614, + "step": 6499 + }, + { + "epoch": 2.282303370786517, + "grad_norm": 0.9063525199890137, + "learning_rate": 3.4440939876448763e-06, + "loss": 0.601, + "step": 6500 + }, + { + "epoch": 2.2826544943820224, + "grad_norm": 0.8959040641784668, + "learning_rate": 3.4436658798140234e-06, + "loss": 0.5793, + "step": 6501 + }, + { + "epoch": 2.283005617977528, + "grad_norm": 0.8836160898208618, + "learning_rate": 3.4432377397106044e-06, + "loss": 0.5757, + "step": 6502 + }, + { + "epoch": 2.2833567415730336, + "grad_norm": 0.9056631922721863, + "learning_rate": 3.4428095673492617e-06, + "loss": 0.5906, + "step": 6503 + }, + { + "epoch": 2.283707865168539, + "grad_norm": 0.8947301506996155, + "learning_rate": 3.442381362744639e-06, + "loss": 0.5824, + "step": 6504 + }, + { + "epoch": 2.284058988764045, + "grad_norm": 0.8961586356163025, + "learning_rate": 3.44195312591138e-06, + "loss": 0.6011, + "step": 6505 + }, + { + "epoch": 2.2844101123595504, + "grad_norm": 0.9021598100662231, + "learning_rate": 3.4415248568641303e-06, + "loss": 0.5889, + "step": 6506 + }, + { + "epoch": 2.284761235955056, + "grad_norm": 0.914538562297821, + "learning_rate": 3.4410965556175356e-06, + "loss": 0.5912, + "step": 6507 + }, + { + "epoch": 2.2851123595505616, + "grad_norm": 0.9124893546104431, + "learning_rate": 3.440668222186245e-06, + "loss": 0.5739, + "step": 6508 + }, + { + "epoch": 2.285463483146067, + "grad_norm": 0.929883599281311, + "learning_rate": 3.440239856584905e-06, + "loss": 0.5812, + "step": 6509 + }, + { + "epoch": 2.285814606741573, + "grad_norm": 0.8998481035232544, + "learning_rate": 3.4398114588281673e-06, + "loss": 0.6068, + "step": 6510 + }, + { + "epoch": 2.286165730337079, + "grad_norm": 0.8813897371292114, + "learning_rate": 3.439383028930681e-06, + "loss": 0.608, + "step": 6511 + }, + { + "epoch": 2.2865168539325844, + "grad_norm": 0.9093124270439148, + "learning_rate": 3.4389545669071e-06, + "loss": 0.5794, + "step": 6512 + }, + { + "epoch": 2.28686797752809, + "grad_norm": 0.8969611525535583, + "learning_rate": 3.438526072772076e-06, + "loss": 0.5814, + "step": 6513 + }, + { + "epoch": 2.2872191011235956, + "grad_norm": 0.8756739497184753, + "learning_rate": 3.4380975465402634e-06, + "loss": 0.5566, + "step": 6514 + }, + { + "epoch": 2.287570224719101, + "grad_norm": 0.9021661281585693, + "learning_rate": 3.437668988226317e-06, + "loss": 0.6086, + "step": 6515 + }, + { + "epoch": 2.287921348314607, + "grad_norm": 0.928519606590271, + "learning_rate": 3.4372403978448944e-06, + "loss": 0.6097, + "step": 6516 + }, + { + "epoch": 2.2882724719101124, + "grad_norm": 0.8854835033416748, + "learning_rate": 3.4368117754106512e-06, + "loss": 0.5785, + "step": 6517 + }, + { + "epoch": 2.288623595505618, + "grad_norm": 0.908923327922821, + "learning_rate": 3.4363831209382474e-06, + "loss": 0.5846, + "step": 6518 + }, + { + "epoch": 2.2889747191011236, + "grad_norm": 0.9027758240699768, + "learning_rate": 3.4359544344423425e-06, + "loss": 0.5869, + "step": 6519 + }, + { + "epoch": 2.289325842696629, + "grad_norm": 0.9100944399833679, + "learning_rate": 3.4355257159375958e-06, + "loss": 0.5878, + "step": 6520 + }, + { + "epoch": 2.289676966292135, + "grad_norm": 0.904100239276886, + "learning_rate": 3.4350969654386714e-06, + "loss": 0.5791, + "step": 6521 + }, + { + "epoch": 2.2900280898876404, + "grad_norm": 0.8889220952987671, + "learning_rate": 3.4346681829602308e-06, + "loss": 0.5761, + "step": 6522 + }, + { + "epoch": 2.290379213483146, + "grad_norm": 0.8781698942184448, + "learning_rate": 3.434239368516937e-06, + "loss": 0.5802, + "step": 6523 + }, + { + "epoch": 2.2907303370786516, + "grad_norm": 0.9097362756729126, + "learning_rate": 3.4338105221234575e-06, + "loss": 0.6372, + "step": 6524 + }, + { + "epoch": 2.291081460674157, + "grad_norm": 0.9048176407814026, + "learning_rate": 3.433381643794457e-06, + "loss": 0.6145, + "step": 6525 + }, + { + "epoch": 2.291432584269663, + "grad_norm": 0.8805797100067139, + "learning_rate": 3.4329527335446023e-06, + "loss": 0.593, + "step": 6526 + }, + { + "epoch": 2.2917837078651684, + "grad_norm": 0.8993074297904968, + "learning_rate": 3.432523791388563e-06, + "loss": 0.5742, + "step": 6527 + }, + { + "epoch": 2.292134831460674, + "grad_norm": 0.911186695098877, + "learning_rate": 3.4320948173410073e-06, + "loss": 0.6035, + "step": 6528 + }, + { + "epoch": 2.29248595505618, + "grad_norm": 0.9233299493789673, + "learning_rate": 3.4316658114166075e-06, + "loss": 0.6035, + "step": 6529 + }, + { + "epoch": 2.2928370786516856, + "grad_norm": 0.8914572596549988, + "learning_rate": 3.4312367736300334e-06, + "loss": 0.5752, + "step": 6530 + }, + { + "epoch": 2.293188202247191, + "grad_norm": 0.8642270565032959, + "learning_rate": 3.4308077039959586e-06, + "loss": 0.5689, + "step": 6531 + }, + { + "epoch": 2.293539325842697, + "grad_norm": 0.9133194088935852, + "learning_rate": 3.430378602529057e-06, + "loss": 0.5985, + "step": 6532 + }, + { + "epoch": 2.2938904494382024, + "grad_norm": 0.9189753532409668, + "learning_rate": 3.4299494692440033e-06, + "loss": 0.6024, + "step": 6533 + }, + { + "epoch": 2.294241573033708, + "grad_norm": 0.8901280760765076, + "learning_rate": 3.429520304155474e-06, + "loss": 0.6039, + "step": 6534 + }, + { + "epoch": 2.2945926966292136, + "grad_norm": 0.8626114726066589, + "learning_rate": 3.4290911072781452e-06, + "loss": 0.5402, + "step": 6535 + }, + { + "epoch": 2.294943820224719, + "grad_norm": 1.2892146110534668, + "learning_rate": 3.428661878626696e-06, + "loss": 0.6026, + "step": 6536 + }, + { + "epoch": 2.295294943820225, + "grad_norm": 0.8889427781105042, + "learning_rate": 3.4282326182158045e-06, + "loss": 0.5696, + "step": 6537 + }, + { + "epoch": 2.2956460674157304, + "grad_norm": 0.9239266514778137, + "learning_rate": 3.4278033260601525e-06, + "loss": 0.6171, + "step": 6538 + }, + { + "epoch": 2.295997191011236, + "grad_norm": 0.9155535101890564, + "learning_rate": 3.42737400217442e-06, + "loss": 0.5606, + "step": 6539 + }, + { + "epoch": 2.2963483146067416, + "grad_norm": 0.8951740264892578, + "learning_rate": 3.4269446465732902e-06, + "loss": 0.5816, + "step": 6540 + }, + { + "epoch": 2.296699438202247, + "grad_norm": 0.9403988718986511, + "learning_rate": 3.4265152592714475e-06, + "loss": 0.5841, + "step": 6541 + }, + { + "epoch": 2.297050561797753, + "grad_norm": 0.8800614476203918, + "learning_rate": 3.426085840283575e-06, + "loss": 0.5675, + "step": 6542 + }, + { + "epoch": 2.2974016853932584, + "grad_norm": 0.9015397429466248, + "learning_rate": 3.4256563896243596e-06, + "loss": 0.6339, + "step": 6543 + }, + { + "epoch": 2.297752808988764, + "grad_norm": 0.8804484009742737, + "learning_rate": 3.4252269073084874e-06, + "loss": 0.5996, + "step": 6544 + }, + { + "epoch": 2.2981039325842696, + "grad_norm": 0.8954015374183655, + "learning_rate": 3.4247973933506466e-06, + "loss": 0.5747, + "step": 6545 + }, + { + "epoch": 2.298455056179775, + "grad_norm": 0.9067341685295105, + "learning_rate": 3.424367847765527e-06, + "loss": 0.5787, + "step": 6546 + }, + { + "epoch": 2.298806179775281, + "grad_norm": 0.8996054530143738, + "learning_rate": 3.423938270567818e-06, + "loss": 0.6203, + "step": 6547 + }, + { + "epoch": 2.2991573033707864, + "grad_norm": 0.8791823983192444, + "learning_rate": 3.42350866177221e-06, + "loss": 0.5706, + "step": 6548 + }, + { + "epoch": 2.299508426966292, + "grad_norm": 0.9062857031822205, + "learning_rate": 3.4230790213933967e-06, + "loss": 0.5999, + "step": 6549 + }, + { + "epoch": 2.2998595505617976, + "grad_norm": 0.9098970293998718, + "learning_rate": 3.4226493494460707e-06, + "loss": 0.588, + "step": 6550 + }, + { + "epoch": 2.300210674157303, + "grad_norm": 0.8824111819267273, + "learning_rate": 3.422219645944926e-06, + "loss": 0.5788, + "step": 6551 + }, + { + "epoch": 2.300561797752809, + "grad_norm": 0.9092296957969666, + "learning_rate": 3.4217899109046588e-06, + "loss": 0.603, + "step": 6552 + }, + { + "epoch": 2.3009129213483144, + "grad_norm": 0.8823329210281372, + "learning_rate": 3.4213601443399654e-06, + "loss": 0.5806, + "step": 6553 + }, + { + "epoch": 2.30126404494382, + "grad_norm": 0.9040799140930176, + "learning_rate": 3.4209303462655435e-06, + "loss": 0.574, + "step": 6554 + }, + { + "epoch": 2.301615168539326, + "grad_norm": 0.912324070930481, + "learning_rate": 3.420500516696092e-06, + "loss": 0.5993, + "step": 6555 + }, + { + "epoch": 2.3019662921348316, + "grad_norm": 0.9137417078018188, + "learning_rate": 3.4200706556463103e-06, + "loss": 0.593, + "step": 6556 + }, + { + "epoch": 2.302317415730337, + "grad_norm": 0.88844233751297, + "learning_rate": 3.419640763130899e-06, + "loss": 0.5963, + "step": 6557 + }, + { + "epoch": 2.302668539325843, + "grad_norm": 0.8981591463088989, + "learning_rate": 3.4192108391645614e-06, + "loss": 0.5762, + "step": 6558 + }, + { + "epoch": 2.3030196629213484, + "grad_norm": 0.9256020188331604, + "learning_rate": 3.4187808837619997e-06, + "loss": 0.5904, + "step": 6559 + }, + { + "epoch": 2.303370786516854, + "grad_norm": 0.8861855268478394, + "learning_rate": 3.4183508969379174e-06, + "loss": 0.5429, + "step": 6560 + }, + { + "epoch": 2.3037219101123596, + "grad_norm": 0.925940752029419, + "learning_rate": 3.4179208787070206e-06, + "loss": 0.5916, + "step": 6561 + }, + { + "epoch": 2.304073033707865, + "grad_norm": 0.8930290341377258, + "learning_rate": 3.4174908290840155e-06, + "loss": 0.6163, + "step": 6562 + }, + { + "epoch": 2.304424157303371, + "grad_norm": 0.8830251097679138, + "learning_rate": 3.4170607480836087e-06, + "loss": 0.5849, + "step": 6563 + }, + { + "epoch": 2.3047752808988764, + "grad_norm": 0.922913670539856, + "learning_rate": 3.4166306357205094e-06, + "loss": 0.6013, + "step": 6564 + }, + { + "epoch": 2.305126404494382, + "grad_norm": 0.8658748865127563, + "learning_rate": 3.416200492009426e-06, + "loss": 0.5609, + "step": 6565 + }, + { + "epoch": 2.3054775280898876, + "grad_norm": 0.8608705401420593, + "learning_rate": 3.4157703169650702e-06, + "loss": 0.5332, + "step": 6566 + }, + { + "epoch": 2.305828651685393, + "grad_norm": 0.8820616602897644, + "learning_rate": 3.4153401106021534e-06, + "loss": 0.5385, + "step": 6567 + }, + { + "epoch": 2.306179775280899, + "grad_norm": 0.8814651370048523, + "learning_rate": 3.414909872935388e-06, + "loss": 0.5815, + "step": 6568 + }, + { + "epoch": 2.3065308988764044, + "grad_norm": 0.9099449515342712, + "learning_rate": 3.414479603979488e-06, + "loss": 0.5916, + "step": 6569 + }, + { + "epoch": 2.30688202247191, + "grad_norm": 0.90244060754776, + "learning_rate": 3.4140493037491677e-06, + "loss": 0.6006, + "step": 6570 + }, + { + "epoch": 2.3072331460674156, + "grad_norm": 0.9014183878898621, + "learning_rate": 3.4136189722591434e-06, + "loss": 0.597, + "step": 6571 + }, + { + "epoch": 2.307584269662921, + "grad_norm": 0.9025565981864929, + "learning_rate": 3.413188609524133e-06, + "loss": 0.5639, + "step": 6572 + }, + { + "epoch": 2.307935393258427, + "grad_norm": 0.9068440198898315, + "learning_rate": 3.4127582155588523e-06, + "loss": 0.5587, + "step": 6573 + }, + { + "epoch": 2.308286516853933, + "grad_norm": 0.8977717161178589, + "learning_rate": 3.412327790378022e-06, + "loss": 0.5686, + "step": 6574 + }, + { + "epoch": 2.3086376404494384, + "grad_norm": 0.9275607466697693, + "learning_rate": 3.411897333996362e-06, + "loss": 0.6155, + "step": 6575 + }, + { + "epoch": 2.308988764044944, + "grad_norm": 0.8982040286064148, + "learning_rate": 3.4114668464285937e-06, + "loss": 0.6006, + "step": 6576 + }, + { + "epoch": 2.3093398876404496, + "grad_norm": 0.9026185870170593, + "learning_rate": 3.4110363276894393e-06, + "loss": 0.5831, + "step": 6577 + }, + { + "epoch": 2.309691011235955, + "grad_norm": 0.9157429337501526, + "learning_rate": 3.4106057777936217e-06, + "loss": 0.6076, + "step": 6578 + }, + { + "epoch": 2.310042134831461, + "grad_norm": 0.8853725790977478, + "learning_rate": 3.410175196755866e-06, + "loss": 0.5981, + "step": 6579 + }, + { + "epoch": 2.3103932584269664, + "grad_norm": 0.9073501229286194, + "learning_rate": 3.409744584590897e-06, + "loss": 0.597, + "step": 6580 + }, + { + "epoch": 2.310744382022472, + "grad_norm": 0.8882585763931274, + "learning_rate": 3.409313941313442e-06, + "loss": 0.5784, + "step": 6581 + }, + { + "epoch": 2.3110955056179776, + "grad_norm": 0.8733483552932739, + "learning_rate": 3.408883266938229e-06, + "loss": 0.6027, + "step": 6582 + }, + { + "epoch": 2.311446629213483, + "grad_norm": 0.920896589756012, + "learning_rate": 3.408452561479985e-06, + "loss": 0.5869, + "step": 6583 + }, + { + "epoch": 2.311797752808989, + "grad_norm": 0.9173773527145386, + "learning_rate": 3.408021824953442e-06, + "loss": 0.5932, + "step": 6584 + }, + { + "epoch": 2.3121488764044944, + "grad_norm": 0.9067617058753967, + "learning_rate": 3.4075910573733283e-06, + "loss": 0.5835, + "step": 6585 + }, + { + "epoch": 2.3125, + "grad_norm": 0.8864839673042297, + "learning_rate": 3.407160258754378e-06, + "loss": 0.5896, + "step": 6586 + }, + { + "epoch": 2.3128511235955056, + "grad_norm": 0.9122422933578491, + "learning_rate": 3.4067294291113224e-06, + "loss": 0.6165, + "step": 6587 + }, + { + "epoch": 2.313202247191011, + "grad_norm": 0.9099538326263428, + "learning_rate": 3.406298568458896e-06, + "loss": 0.5934, + "step": 6588 + }, + { + "epoch": 2.313553370786517, + "grad_norm": 0.8932905197143555, + "learning_rate": 3.405867676811835e-06, + "loss": 0.5664, + "step": 6589 + }, + { + "epoch": 2.3139044943820224, + "grad_norm": 0.8937992453575134, + "learning_rate": 3.4054367541848736e-06, + "loss": 0.589, + "step": 6590 + }, + { + "epoch": 2.314255617977528, + "grad_norm": 0.9297001361846924, + "learning_rate": 3.405005800592751e-06, + "loss": 0.6157, + "step": 6591 + }, + { + "epoch": 2.3146067415730336, + "grad_norm": 0.8797909021377563, + "learning_rate": 3.4045748160502033e-06, + "loss": 0.5775, + "step": 6592 + }, + { + "epoch": 2.314957865168539, + "grad_norm": 0.8575281500816345, + "learning_rate": 3.404143800571972e-06, + "loss": 0.5249, + "step": 6593 + }, + { + "epoch": 2.315308988764045, + "grad_norm": 0.916915774345398, + "learning_rate": 3.4037127541727964e-06, + "loss": 0.6298, + "step": 6594 + }, + { + "epoch": 2.3156601123595504, + "grad_norm": 0.9410571455955505, + "learning_rate": 3.4032816768674178e-06, + "loss": 0.6113, + "step": 6595 + }, + { + "epoch": 2.316011235955056, + "grad_norm": 0.870938777923584, + "learning_rate": 3.402850568670578e-06, + "loss": 0.5691, + "step": 6596 + }, + { + "epoch": 2.3163623595505616, + "grad_norm": 0.8897520899772644, + "learning_rate": 3.4024194295970226e-06, + "loss": 0.5838, + "step": 6597 + }, + { + "epoch": 2.316713483146067, + "grad_norm": 0.9131793975830078, + "learning_rate": 3.401988259661494e-06, + "loss": 0.6284, + "step": 6598 + }, + { + "epoch": 2.317064606741573, + "grad_norm": 0.9168117642402649, + "learning_rate": 3.4015570588787387e-06, + "loss": 0.6004, + "step": 6599 + }, + { + "epoch": 2.317415730337079, + "grad_norm": 0.9237546920776367, + "learning_rate": 3.401125827263504e-06, + "loss": 0.5393, + "step": 6600 + }, + { + "epoch": 2.3177668539325844, + "grad_norm": 0.8847349882125854, + "learning_rate": 3.4006945648305375e-06, + "loss": 0.5808, + "step": 6601 + }, + { + "epoch": 2.31811797752809, + "grad_norm": 0.9153926968574524, + "learning_rate": 3.4002632715945868e-06, + "loss": 0.5722, + "step": 6602 + }, + { + "epoch": 2.3184691011235956, + "grad_norm": 0.8829538226127625, + "learning_rate": 3.399831947570402e-06, + "loss": 0.6002, + "step": 6603 + }, + { + "epoch": 2.318820224719101, + "grad_norm": 0.9332196116447449, + "learning_rate": 3.399400592772736e-06, + "loss": 0.6061, + "step": 6604 + }, + { + "epoch": 2.319171348314607, + "grad_norm": 0.9451414942741394, + "learning_rate": 3.3989692072163383e-06, + "loss": 0.646, + "step": 6605 + }, + { + "epoch": 2.3195224719101124, + "grad_norm": 0.9025115966796875, + "learning_rate": 3.3985377909159635e-06, + "loss": 0.5723, + "step": 6606 + }, + { + "epoch": 2.319873595505618, + "grad_norm": 0.894688606262207, + "learning_rate": 3.398106343886365e-06, + "loss": 0.5725, + "step": 6607 + }, + { + "epoch": 2.3202247191011236, + "grad_norm": 0.913253128528595, + "learning_rate": 3.3976748661422983e-06, + "loss": 0.5832, + "step": 6608 + }, + { + "epoch": 2.320575842696629, + "grad_norm": 0.8924520015716553, + "learning_rate": 3.3972433576985187e-06, + "loss": 0.5985, + "step": 6609 + }, + { + "epoch": 2.320926966292135, + "grad_norm": 0.9239665865898132, + "learning_rate": 3.396811818569785e-06, + "loss": 0.563, + "step": 6610 + }, + { + "epoch": 2.3212780898876404, + "grad_norm": 0.902673065662384, + "learning_rate": 3.3963802487708543e-06, + "loss": 0.5664, + "step": 6611 + }, + { + "epoch": 2.321629213483146, + "grad_norm": 0.919130802154541, + "learning_rate": 3.3959486483164856e-06, + "loss": 0.5474, + "step": 6612 + }, + { + "epoch": 2.3219803370786516, + "grad_norm": 0.8744391798973083, + "learning_rate": 3.3955170172214404e-06, + "loss": 0.5614, + "step": 6613 + }, + { + "epoch": 2.322331460674157, + "grad_norm": 0.9338593482971191, + "learning_rate": 3.3950853555004786e-06, + "loss": 0.5922, + "step": 6614 + }, + { + "epoch": 2.322682584269663, + "grad_norm": 0.9600446820259094, + "learning_rate": 3.394653663168364e-06, + "loss": 0.608, + "step": 6615 + }, + { + "epoch": 2.3230337078651684, + "grad_norm": 0.9451854825019836, + "learning_rate": 3.3942219402398594e-06, + "loss": 0.5794, + "step": 6616 + }, + { + "epoch": 2.323384831460674, + "grad_norm": 0.9392783641815186, + "learning_rate": 3.3937901867297305e-06, + "loss": 0.5933, + "step": 6617 + }, + { + "epoch": 2.32373595505618, + "grad_norm": 0.8949928879737854, + "learning_rate": 3.3933584026527414e-06, + "loss": 0.5713, + "step": 6618 + }, + { + "epoch": 2.3240870786516856, + "grad_norm": 0.8982742428779602, + "learning_rate": 3.3929265880236594e-06, + "loss": 0.5824, + "step": 6619 + }, + { + "epoch": 2.324438202247191, + "grad_norm": 0.8996788859367371, + "learning_rate": 3.3924947428572523e-06, + "loss": 0.5841, + "step": 6620 + }, + { + "epoch": 2.324789325842697, + "grad_norm": 0.8854213356971741, + "learning_rate": 3.3920628671682887e-06, + "loss": 0.5843, + "step": 6621 + }, + { + "epoch": 2.3251404494382024, + "grad_norm": 0.8895551562309265, + "learning_rate": 3.391630960971539e-06, + "loss": 0.5873, + "step": 6622 + }, + { + "epoch": 2.325491573033708, + "grad_norm": 0.9179404973983765, + "learning_rate": 3.391199024281773e-06, + "loss": 0.6155, + "step": 6623 + }, + { + "epoch": 2.3258426966292136, + "grad_norm": 0.9028970003128052, + "learning_rate": 3.390767057113763e-06, + "loss": 0.5655, + "step": 6624 + }, + { + "epoch": 2.326193820224719, + "grad_norm": 0.9288235306739807, + "learning_rate": 3.3903350594822813e-06, + "loss": 0.5759, + "step": 6625 + }, + { + "epoch": 2.326544943820225, + "grad_norm": 0.9575920104980469, + "learning_rate": 3.389903031402103e-06, + "loss": 0.5572, + "step": 6626 + }, + { + "epoch": 2.3268960674157304, + "grad_norm": 0.9123892784118652, + "learning_rate": 3.3894709728880023e-06, + "loss": 0.5924, + "step": 6627 + }, + { + "epoch": 2.327247191011236, + "grad_norm": 0.884859025478363, + "learning_rate": 3.3890388839547554e-06, + "loss": 0.5785, + "step": 6628 + }, + { + "epoch": 2.3275983146067416, + "grad_norm": 0.8762487173080444, + "learning_rate": 3.3886067646171393e-06, + "loss": 0.5919, + "step": 6629 + }, + { + "epoch": 2.327949438202247, + "grad_norm": 0.8824047446250916, + "learning_rate": 3.388174614889933e-06, + "loss": 0.5827, + "step": 6630 + }, + { + "epoch": 2.328300561797753, + "grad_norm": 0.86397784948349, + "learning_rate": 3.387742434787914e-06, + "loss": 0.5473, + "step": 6631 + }, + { + "epoch": 2.3286516853932584, + "grad_norm": 0.8915233016014099, + "learning_rate": 3.3873102243258637e-06, + "loss": 0.5384, + "step": 6632 + }, + { + "epoch": 2.329002808988764, + "grad_norm": 0.9059692621231079, + "learning_rate": 3.3868779835185633e-06, + "loss": 0.6126, + "step": 6633 + }, + { + "epoch": 2.3293539325842696, + "grad_norm": 0.9174104332923889, + "learning_rate": 3.3864457123807948e-06, + "loss": 0.5811, + "step": 6634 + }, + { + "epoch": 2.329705056179775, + "grad_norm": 0.8908886313438416, + "learning_rate": 3.3860134109273414e-06, + "loss": 0.5585, + "step": 6635 + }, + { + "epoch": 2.330056179775281, + "grad_norm": 0.9034232497215271, + "learning_rate": 3.385581079172987e-06, + "loss": 0.5804, + "step": 6636 + }, + { + "epoch": 2.3304073033707864, + "grad_norm": 0.9230918288230896, + "learning_rate": 3.3851487171325183e-06, + "loss": 0.6266, + "step": 6637 + }, + { + "epoch": 2.330758426966292, + "grad_norm": 0.9082732200622559, + "learning_rate": 3.3847163248207206e-06, + "loss": 0.6025, + "step": 6638 + }, + { + "epoch": 2.3311095505617976, + "grad_norm": 0.8988151550292969, + "learning_rate": 3.384283902252381e-06, + "loss": 0.5873, + "step": 6639 + }, + { + "epoch": 2.331460674157303, + "grad_norm": 0.8844032883644104, + "learning_rate": 3.383851449442289e-06, + "loss": 0.5809, + "step": 6640 + }, + { + "epoch": 2.331811797752809, + "grad_norm": 0.9464019536972046, + "learning_rate": 3.383418966405234e-06, + "loss": 0.5949, + "step": 6641 + }, + { + "epoch": 2.3321629213483144, + "grad_norm": 0.9258586168289185, + "learning_rate": 3.3829864531560065e-06, + "loss": 0.5814, + "step": 6642 + }, + { + "epoch": 2.33251404494382, + "grad_norm": 0.9415653944015503, + "learning_rate": 3.3825539097093973e-06, + "loss": 0.6026, + "step": 6643 + }, + { + "epoch": 2.332865168539326, + "grad_norm": 0.9154693484306335, + "learning_rate": 3.3821213360802e-06, + "loss": 0.6054, + "step": 6644 + }, + { + "epoch": 2.3332162921348316, + "grad_norm": 0.9117648601531982, + "learning_rate": 3.3816887322832075e-06, + "loss": 0.6073, + "step": 6645 + }, + { + "epoch": 2.333567415730337, + "grad_norm": 0.919195830821991, + "learning_rate": 3.3812560983332155e-06, + "loss": 0.6106, + "step": 6646 + }, + { + "epoch": 2.333918539325843, + "grad_norm": 0.8973156213760376, + "learning_rate": 3.3808234342450187e-06, + "loss": 0.5649, + "step": 6647 + }, + { + "epoch": 2.3342696629213484, + "grad_norm": 0.9068472385406494, + "learning_rate": 3.380390740033414e-06, + "loss": 0.5683, + "step": 6648 + }, + { + "epoch": 2.334620786516854, + "grad_norm": 0.8847914338111877, + "learning_rate": 3.3799580157131993e-06, + "loss": 0.5515, + "step": 6649 + }, + { + "epoch": 2.3349719101123596, + "grad_norm": 0.9018244743347168, + "learning_rate": 3.3795252612991738e-06, + "loss": 0.5572, + "step": 6650 + }, + { + "epoch": 2.335323033707865, + "grad_norm": 0.9132426381111145, + "learning_rate": 3.3790924768061367e-06, + "loss": 0.5835, + "step": 6651 + }, + { + "epoch": 2.335674157303371, + "grad_norm": 0.9036231637001038, + "learning_rate": 3.378659662248889e-06, + "loss": 0.5835, + "step": 6652 + }, + { + "epoch": 2.3360252808988764, + "grad_norm": 0.8762897253036499, + "learning_rate": 3.3782268176422334e-06, + "loss": 0.5519, + "step": 6653 + }, + { + "epoch": 2.336376404494382, + "grad_norm": 0.909015417098999, + "learning_rate": 3.3777939430009714e-06, + "loss": 0.5654, + "step": 6654 + }, + { + "epoch": 2.3367275280898876, + "grad_norm": 0.8963292837142944, + "learning_rate": 3.377361038339908e-06, + "loss": 0.5684, + "step": 6655 + }, + { + "epoch": 2.337078651685393, + "grad_norm": 0.9168239235877991, + "learning_rate": 3.3769281036738475e-06, + "loss": 0.5965, + "step": 6656 + }, + { + "epoch": 2.337429775280899, + "grad_norm": 0.9261362552642822, + "learning_rate": 3.376495139017597e-06, + "loss": 0.5592, + "step": 6657 + }, + { + "epoch": 2.3377808988764044, + "grad_norm": 0.8753339052200317, + "learning_rate": 3.376062144385962e-06, + "loss": 0.5625, + "step": 6658 + }, + { + "epoch": 2.33813202247191, + "grad_norm": 0.9048936367034912, + "learning_rate": 3.375629119793752e-06, + "loss": 0.5732, + "step": 6659 + }, + { + "epoch": 2.3384831460674156, + "grad_norm": 0.9059818387031555, + "learning_rate": 3.375196065255776e-06, + "loss": 0.5868, + "step": 6660 + }, + { + "epoch": 2.338834269662921, + "grad_norm": 0.9207652807235718, + "learning_rate": 3.374762980786842e-06, + "loss": 0.6209, + "step": 6661 + }, + { + "epoch": 2.339185393258427, + "grad_norm": 0.892612099647522, + "learning_rate": 3.374329866401763e-06, + "loss": 0.5958, + "step": 6662 + }, + { + "epoch": 2.339536516853933, + "grad_norm": 0.8915576934814453, + "learning_rate": 3.3738967221153512e-06, + "loss": 0.5941, + "step": 6663 + }, + { + "epoch": 2.3398876404494384, + "grad_norm": 0.9067428112030029, + "learning_rate": 3.373463547942419e-06, + "loss": 0.6013, + "step": 6664 + }, + { + "epoch": 2.340238764044944, + "grad_norm": 0.8714523315429688, + "learning_rate": 3.3730303438977806e-06, + "loss": 0.5558, + "step": 6665 + }, + { + "epoch": 2.3405898876404496, + "grad_norm": 0.8962456583976746, + "learning_rate": 3.372597109996252e-06, + "loss": 0.606, + "step": 6666 + }, + { + "epoch": 2.340941011235955, + "grad_norm": 0.8832200765609741, + "learning_rate": 3.3721638462526483e-06, + "loss": 0.5644, + "step": 6667 + }, + { + "epoch": 2.341292134831461, + "grad_norm": 0.8841350078582764, + "learning_rate": 3.371730552681788e-06, + "loss": 0.5762, + "step": 6668 + }, + { + "epoch": 2.3416432584269664, + "grad_norm": 0.8814048767089844, + "learning_rate": 3.3712972292984887e-06, + "loss": 0.5677, + "step": 6669 + }, + { + "epoch": 2.341994382022472, + "grad_norm": 0.88837730884552, + "learning_rate": 3.3708638761175695e-06, + "loss": 0.5547, + "step": 6670 + }, + { + "epoch": 2.3423455056179776, + "grad_norm": 0.9263938665390015, + "learning_rate": 3.370430493153851e-06, + "loss": 0.5832, + "step": 6671 + }, + { + "epoch": 2.342696629213483, + "grad_norm": 0.9187746047973633, + "learning_rate": 3.3699970804221548e-06, + "loss": 0.5975, + "step": 6672 + }, + { + "epoch": 2.343047752808989, + "grad_norm": 0.8915598392486572, + "learning_rate": 3.369563637937303e-06, + "loss": 0.587, + "step": 6673 + }, + { + "epoch": 2.3433988764044944, + "grad_norm": 0.8813969492912292, + "learning_rate": 3.3691301657141183e-06, + "loss": 0.5807, + "step": 6674 + }, + { + "epoch": 2.34375, + "grad_norm": 0.8943729996681213, + "learning_rate": 3.3686966637674257e-06, + "loss": 0.579, + "step": 6675 + }, + { + "epoch": 2.3441011235955056, + "grad_norm": 0.9202485084533691, + "learning_rate": 3.3682631321120507e-06, + "loss": 0.6423, + "step": 6676 + }, + { + "epoch": 2.344452247191011, + "grad_norm": 0.8741579651832581, + "learning_rate": 3.36782957076282e-06, + "loss": 0.5625, + "step": 6677 + }, + { + "epoch": 2.344803370786517, + "grad_norm": 0.8953279256820679, + "learning_rate": 3.36739597973456e-06, + "loss": 0.6042, + "step": 6678 + }, + { + "epoch": 2.3451544943820224, + "grad_norm": 0.904447615146637, + "learning_rate": 3.3669623590420997e-06, + "loss": 0.5852, + "step": 6679 + }, + { + "epoch": 2.345505617977528, + "grad_norm": 0.9037969708442688, + "learning_rate": 3.366528708700269e-06, + "loss": 0.5569, + "step": 6680 + }, + { + "epoch": 2.3458567415730336, + "grad_norm": 0.9001902341842651, + "learning_rate": 3.366095028723898e-06, + "loss": 0.576, + "step": 6681 + }, + { + "epoch": 2.346207865168539, + "grad_norm": 0.8963335752487183, + "learning_rate": 3.3656613191278183e-06, + "loss": 0.5627, + "step": 6682 + }, + { + "epoch": 2.346558988764045, + "grad_norm": 0.9518795609474182, + "learning_rate": 3.3652275799268618e-06, + "loss": 0.6192, + "step": 6683 + }, + { + "epoch": 2.3469101123595504, + "grad_norm": 0.9051629900932312, + "learning_rate": 3.3647938111358624e-06, + "loss": 0.5839, + "step": 6684 + }, + { + "epoch": 2.347261235955056, + "grad_norm": 0.9279952049255371, + "learning_rate": 3.3643600127696545e-06, + "loss": 0.6007, + "step": 6685 + }, + { + "epoch": 2.3476123595505616, + "grad_norm": 0.9282490015029907, + "learning_rate": 3.363926184843074e-06, + "loss": 0.6176, + "step": 6686 + }, + { + "epoch": 2.347963483146067, + "grad_norm": 0.883164644241333, + "learning_rate": 3.3634923273709575e-06, + "loss": 0.566, + "step": 6687 + }, + { + "epoch": 2.348314606741573, + "grad_norm": 0.8973921537399292, + "learning_rate": 3.3630584403681426e-06, + "loss": 0.5896, + "step": 6688 + }, + { + "epoch": 2.348665730337079, + "grad_norm": 0.9248161911964417, + "learning_rate": 3.3626245238494672e-06, + "loss": 0.59, + "step": 6689 + }, + { + "epoch": 2.3490168539325844, + "grad_norm": 0.8894219398498535, + "learning_rate": 3.3621905778297715e-06, + "loss": 0.6139, + "step": 6690 + }, + { + "epoch": 2.34936797752809, + "grad_norm": 0.9183155298233032, + "learning_rate": 3.3617566023238953e-06, + "loss": 0.6017, + "step": 6691 + }, + { + "epoch": 2.3497191011235956, + "grad_norm": 0.8851825594902039, + "learning_rate": 3.3613225973466805e-06, + "loss": 0.567, + "step": 6692 + }, + { + "epoch": 2.350070224719101, + "grad_norm": 0.9135637283325195, + "learning_rate": 3.3608885629129707e-06, + "loss": 0.6206, + "step": 6693 + }, + { + "epoch": 2.350421348314607, + "grad_norm": 0.8638511300086975, + "learning_rate": 3.3604544990376083e-06, + "loss": 0.5817, + "step": 6694 + }, + { + "epoch": 2.3507724719101124, + "grad_norm": 0.8822200894355774, + "learning_rate": 3.360020405735438e-06, + "loss": 0.5801, + "step": 6695 + }, + { + "epoch": 2.351123595505618, + "grad_norm": 0.8683942556381226, + "learning_rate": 3.3595862830213066e-06, + "loss": 0.563, + "step": 6696 + }, + { + "epoch": 2.3514747191011236, + "grad_norm": 0.8999106287956238, + "learning_rate": 3.3591521309100596e-06, + "loss": 0.6232, + "step": 6697 + }, + { + "epoch": 2.351825842696629, + "grad_norm": 0.8885555863380432, + "learning_rate": 3.3587179494165445e-06, + "loss": 0.5508, + "step": 6698 + }, + { + "epoch": 2.352176966292135, + "grad_norm": 0.9269642233848572, + "learning_rate": 3.358283738555611e-06, + "loss": 0.588, + "step": 6699 + }, + { + "epoch": 2.3525280898876404, + "grad_norm": 0.8653067350387573, + "learning_rate": 3.3578494983421073e-06, + "loss": 0.5613, + "step": 6700 + }, + { + "epoch": 2.352879213483146, + "grad_norm": 0.9387295842170715, + "learning_rate": 3.3574152287908853e-06, + "loss": 0.6235, + "step": 6701 + }, + { + "epoch": 2.3532303370786516, + "grad_norm": 0.9514093399047852, + "learning_rate": 3.3569809299167962e-06, + "loss": 0.6198, + "step": 6702 + }, + { + "epoch": 2.353581460674157, + "grad_norm": 0.9142879247665405, + "learning_rate": 3.356546601734692e-06, + "loss": 0.5937, + "step": 6703 + }, + { + "epoch": 2.353932584269663, + "grad_norm": 0.8835601806640625, + "learning_rate": 3.3561122442594274e-06, + "loss": 0.5724, + "step": 6704 + }, + { + "epoch": 2.3542837078651684, + "grad_norm": 0.8973185420036316, + "learning_rate": 3.355677857505857e-06, + "loss": 0.5823, + "step": 6705 + }, + { + "epoch": 2.354634831460674, + "grad_norm": 0.9576224088668823, + "learning_rate": 3.355243441488836e-06, + "loss": 0.5835, + "step": 6706 + }, + { + "epoch": 2.35498595505618, + "grad_norm": 0.9009076356887817, + "learning_rate": 3.3548089962232204e-06, + "loss": 0.5697, + "step": 6707 + }, + { + "epoch": 2.3553370786516856, + "grad_norm": 0.8896247148513794, + "learning_rate": 3.3543745217238695e-06, + "loss": 0.5806, + "step": 6708 + }, + { + "epoch": 2.355688202247191, + "grad_norm": 0.9310916662216187, + "learning_rate": 3.3539400180056403e-06, + "loss": 0.6439, + "step": 6709 + }, + { + "epoch": 2.356039325842697, + "grad_norm": 0.8478217124938965, + "learning_rate": 3.3535054850833943e-06, + "loss": 0.537, + "step": 6710 + }, + { + "epoch": 2.3563904494382024, + "grad_norm": 0.9331678152084351, + "learning_rate": 3.3530709229719905e-06, + "loss": 0.5711, + "step": 6711 + }, + { + "epoch": 2.356741573033708, + "grad_norm": 0.8889136910438538, + "learning_rate": 3.3526363316862907e-06, + "loss": 0.6153, + "step": 6712 + }, + { + "epoch": 2.3570926966292136, + "grad_norm": 0.9165581464767456, + "learning_rate": 3.352201711241158e-06, + "loss": 0.5973, + "step": 6713 + }, + { + "epoch": 2.357443820224719, + "grad_norm": 0.9083494544029236, + "learning_rate": 3.3517670616514565e-06, + "loss": 0.6275, + "step": 6714 + }, + { + "epoch": 2.357794943820225, + "grad_norm": 0.9021841287612915, + "learning_rate": 3.35133238293205e-06, + "loss": 0.5791, + "step": 6715 + }, + { + "epoch": 2.3581460674157304, + "grad_norm": 0.9055184721946716, + "learning_rate": 3.3508976750978045e-06, + "loss": 0.6072, + "step": 6716 + }, + { + "epoch": 2.358497191011236, + "grad_norm": 0.8913466334342957, + "learning_rate": 3.3504629381635866e-06, + "loss": 0.5888, + "step": 6717 + }, + { + "epoch": 2.3588483146067416, + "grad_norm": 0.9372550845146179, + "learning_rate": 3.350028172144264e-06, + "loss": 0.5672, + "step": 6718 + }, + { + "epoch": 2.359199438202247, + "grad_norm": 0.8611996173858643, + "learning_rate": 3.3495933770547057e-06, + "loss": 0.5774, + "step": 6719 + }, + { + "epoch": 2.359550561797753, + "grad_norm": 0.890654981136322, + "learning_rate": 3.3491585529097803e-06, + "loss": 0.538, + "step": 6720 + }, + { + "epoch": 2.3599016853932584, + "grad_norm": 0.8999179601669312, + "learning_rate": 3.3487236997243596e-06, + "loss": 0.5681, + "step": 6721 + }, + { + "epoch": 2.360252808988764, + "grad_norm": 0.9105609655380249, + "learning_rate": 3.3482888175133144e-06, + "loss": 0.5907, + "step": 6722 + }, + { + "epoch": 2.3606039325842696, + "grad_norm": 0.9112907648086548, + "learning_rate": 3.347853906291518e-06, + "loss": 0.6323, + "step": 6723 + }, + { + "epoch": 2.360955056179775, + "grad_norm": 0.9228449463844299, + "learning_rate": 3.3474189660738434e-06, + "loss": 0.5894, + "step": 6724 + }, + { + "epoch": 2.361306179775281, + "grad_norm": 0.8868569731712341, + "learning_rate": 3.3469839968751645e-06, + "loss": 0.5947, + "step": 6725 + }, + { + "epoch": 2.3616573033707864, + "grad_norm": 0.9078425765037537, + "learning_rate": 3.3465489987103584e-06, + "loss": 0.5778, + "step": 6726 + }, + { + "epoch": 2.362008426966292, + "grad_norm": 0.9294246435165405, + "learning_rate": 3.3461139715943003e-06, + "loss": 0.6384, + "step": 6727 + }, + { + "epoch": 2.3623595505617976, + "grad_norm": 0.9035244584083557, + "learning_rate": 3.345678915541869e-06, + "loss": 0.5913, + "step": 6728 + }, + { + "epoch": 2.362710674157303, + "grad_norm": 0.8824789524078369, + "learning_rate": 3.345243830567942e-06, + "loss": 0.5658, + "step": 6729 + }, + { + "epoch": 2.363061797752809, + "grad_norm": 0.8710284233093262, + "learning_rate": 3.3448087166873994e-06, + "loss": 0.5955, + "step": 6730 + }, + { + "epoch": 2.3634129213483144, + "grad_norm": 0.9041821360588074, + "learning_rate": 3.3443735739151216e-06, + "loss": 0.5785, + "step": 6731 + }, + { + "epoch": 2.36376404494382, + "grad_norm": 0.9021846652030945, + "learning_rate": 3.3439384022659904e-06, + "loss": 0.5956, + "step": 6732 + }, + { + "epoch": 2.364115168539326, + "grad_norm": 0.9032467007637024, + "learning_rate": 3.3435032017548875e-06, + "loss": 0.5775, + "step": 6733 + }, + { + "epoch": 2.3644662921348316, + "grad_norm": 0.8861805200576782, + "learning_rate": 3.3430679723966974e-06, + "loss": 0.5792, + "step": 6734 + }, + { + "epoch": 2.364817415730337, + "grad_norm": 0.8970776796340942, + "learning_rate": 3.3426327142063036e-06, + "loss": 0.5994, + "step": 6735 + }, + { + "epoch": 2.365168539325843, + "grad_norm": 0.89840167760849, + "learning_rate": 3.342197427198593e-06, + "loss": 0.5745, + "step": 6736 + }, + { + "epoch": 2.3655196629213484, + "grad_norm": 0.8873356580734253, + "learning_rate": 3.3417621113884504e-06, + "loss": 0.5772, + "step": 6737 + }, + { + "epoch": 2.365870786516854, + "grad_norm": 0.8778063058853149, + "learning_rate": 3.341326766790764e-06, + "loss": 0.573, + "step": 6738 + }, + { + "epoch": 2.3662219101123596, + "grad_norm": 0.9052056670188904, + "learning_rate": 3.340891393420422e-06, + "loss": 0.57, + "step": 6739 + }, + { + "epoch": 2.366573033707865, + "grad_norm": 0.9140535593032837, + "learning_rate": 3.3404559912923134e-06, + "loss": 0.6192, + "step": 6740 + }, + { + "epoch": 2.366924157303371, + "grad_norm": 0.9063019752502441, + "learning_rate": 3.3400205604213297e-06, + "loss": 0.5997, + "step": 6741 + }, + { + "epoch": 2.3672752808988764, + "grad_norm": 0.8880258798599243, + "learning_rate": 3.3395851008223618e-06, + "loss": 0.6006, + "step": 6742 + }, + { + "epoch": 2.367626404494382, + "grad_norm": 0.9156434535980225, + "learning_rate": 3.3391496125103013e-06, + "loss": 0.6131, + "step": 6743 + }, + { + "epoch": 2.3679775280898876, + "grad_norm": 0.9320157766342163, + "learning_rate": 3.3387140955000424e-06, + "loss": 0.5763, + "step": 6744 + }, + { + "epoch": 2.368328651685393, + "grad_norm": 0.9099292159080505, + "learning_rate": 3.3382785498064794e-06, + "loss": 0.6003, + "step": 6745 + }, + { + "epoch": 2.368679775280899, + "grad_norm": 0.891942024230957, + "learning_rate": 3.337842975444507e-06, + "loss": 0.5842, + "step": 6746 + }, + { + "epoch": 2.3690308988764044, + "grad_norm": 0.9034740328788757, + "learning_rate": 3.337407372429022e-06, + "loss": 0.6241, + "step": 6747 + }, + { + "epoch": 2.36938202247191, + "grad_norm": 0.9337887167930603, + "learning_rate": 3.336971740774921e-06, + "loss": 0.5922, + "step": 6748 + }, + { + "epoch": 2.3697331460674156, + "grad_norm": 0.8917706608772278, + "learning_rate": 3.3365360804971035e-06, + "loss": 0.5706, + "step": 6749 + }, + { + "epoch": 2.370084269662921, + "grad_norm": 0.9517117738723755, + "learning_rate": 3.336100391610467e-06, + "loss": 0.5886, + "step": 6750 + }, + { + "epoch": 2.370435393258427, + "grad_norm": 0.8895031213760376, + "learning_rate": 3.3356646741299137e-06, + "loss": 0.5429, + "step": 6751 + }, + { + "epoch": 2.370786516853933, + "grad_norm": 0.9127622842788696, + "learning_rate": 3.3352289280703425e-06, + "loss": 0.6067, + "step": 6752 + }, + { + "epoch": 2.3711376404494384, + "grad_norm": 0.8721568584442139, + "learning_rate": 3.3347931534466573e-06, + "loss": 0.5599, + "step": 6753 + }, + { + "epoch": 2.371488764044944, + "grad_norm": 0.9159287810325623, + "learning_rate": 3.3343573502737605e-06, + "loss": 0.5769, + "step": 6754 + }, + { + "epoch": 2.3718398876404496, + "grad_norm": 0.8926753401756287, + "learning_rate": 3.3339215185665557e-06, + "loss": 0.5892, + "step": 6755 + }, + { + "epoch": 2.372191011235955, + "grad_norm": 0.8853592872619629, + "learning_rate": 3.333485658339949e-06, + "loss": 0.5941, + "step": 6756 + }, + { + "epoch": 2.372542134831461, + "grad_norm": 0.8926652073860168, + "learning_rate": 3.333049769608846e-06, + "loss": 0.5775, + "step": 6757 + }, + { + "epoch": 2.3728932584269664, + "grad_norm": 0.8698339462280273, + "learning_rate": 3.3326138523881536e-06, + "loss": 0.6043, + "step": 6758 + }, + { + "epoch": 2.373244382022472, + "grad_norm": 0.9000748991966248, + "learning_rate": 3.33217790669278e-06, + "loss": 0.608, + "step": 6759 + }, + { + "epoch": 2.3735955056179776, + "grad_norm": 0.9015805721282959, + "learning_rate": 3.331741932537635e-06, + "loss": 0.6106, + "step": 6760 + }, + { + "epoch": 2.373946629213483, + "grad_norm": 0.8738047480583191, + "learning_rate": 3.3313059299376265e-06, + "loss": 0.557, + "step": 6761 + }, + { + "epoch": 2.374297752808989, + "grad_norm": 0.8983538746833801, + "learning_rate": 3.330869898907667e-06, + "loss": 0.5829, + "step": 6762 + }, + { + "epoch": 2.3746488764044944, + "grad_norm": 0.9225078225135803, + "learning_rate": 3.330433839462668e-06, + "loss": 0.5986, + "step": 6763 + }, + { + "epoch": 2.375, + "grad_norm": 0.895306408405304, + "learning_rate": 3.329997751617542e-06, + "loss": 0.5734, + "step": 6764 + }, + { + "epoch": 2.3753511235955056, + "grad_norm": 0.90412437915802, + "learning_rate": 3.329561635387203e-06, + "loss": 0.6339, + "step": 6765 + }, + { + "epoch": 2.375702247191011, + "grad_norm": 0.9264748692512512, + "learning_rate": 3.329125490786566e-06, + "loss": 0.6107, + "step": 6766 + }, + { + "epoch": 2.376053370786517, + "grad_norm": 0.8876569867134094, + "learning_rate": 3.328689317830547e-06, + "loss": 0.5634, + "step": 6767 + }, + { + "epoch": 2.3764044943820224, + "grad_norm": 0.9254834651947021, + "learning_rate": 3.3282531165340633e-06, + "loss": 0.5584, + "step": 6768 + }, + { + "epoch": 2.376755617977528, + "grad_norm": 0.9019694924354553, + "learning_rate": 3.3278168869120302e-06, + "loss": 0.5904, + "step": 6769 + }, + { + "epoch": 2.3771067415730336, + "grad_norm": 0.8998555541038513, + "learning_rate": 3.3273806289793684e-06, + "loss": 0.6055, + "step": 6770 + }, + { + "epoch": 2.377457865168539, + "grad_norm": 0.8702539205551147, + "learning_rate": 3.326944342750997e-06, + "loss": 0.5614, + "step": 6771 + }, + { + "epoch": 2.377808988764045, + "grad_norm": 0.8986842036247253, + "learning_rate": 3.326508028241837e-06, + "loss": 0.5521, + "step": 6772 + }, + { + "epoch": 2.3781601123595504, + "grad_norm": 0.901059091091156, + "learning_rate": 3.3260716854668103e-06, + "loss": 0.5523, + "step": 6773 + }, + { + "epoch": 2.378511235955056, + "grad_norm": 0.8907270431518555, + "learning_rate": 3.3256353144408383e-06, + "loss": 0.5744, + "step": 6774 + }, + { + "epoch": 2.3788623595505616, + "grad_norm": 0.9329801201820374, + "learning_rate": 3.325198915178845e-06, + "loss": 0.5719, + "step": 6775 + }, + { + "epoch": 2.379213483146067, + "grad_norm": 0.8890707492828369, + "learning_rate": 3.3247624876957553e-06, + "loss": 0.5666, + "step": 6776 + }, + { + "epoch": 2.379564606741573, + "grad_norm": 0.9123046398162842, + "learning_rate": 3.3243260320064934e-06, + "loss": 0.6123, + "step": 6777 + }, + { + "epoch": 2.379915730337079, + "grad_norm": 0.9072712659835815, + "learning_rate": 3.323889548125987e-06, + "loss": 0.5869, + "step": 6778 + }, + { + "epoch": 2.3802668539325844, + "grad_norm": 0.9533200263977051, + "learning_rate": 3.323453036069163e-06, + "loss": 0.5818, + "step": 6779 + }, + { + "epoch": 2.38061797752809, + "grad_norm": 0.9296919107437134, + "learning_rate": 3.3230164958509496e-06, + "loss": 0.6057, + "step": 6780 + }, + { + "epoch": 2.3809691011235956, + "grad_norm": 0.8382134437561035, + "learning_rate": 3.322579927486277e-06, + "loss": 0.571, + "step": 6781 + }, + { + "epoch": 2.381320224719101, + "grad_norm": 0.9843775033950806, + "learning_rate": 3.322143330990074e-06, + "loss": 0.6143, + "step": 6782 + }, + { + "epoch": 2.381671348314607, + "grad_norm": 0.8919227719306946, + "learning_rate": 3.3217067063772734e-06, + "loss": 0.5906, + "step": 6783 + }, + { + "epoch": 2.3820224719101124, + "grad_norm": 0.8886569738388062, + "learning_rate": 3.321270053662806e-06, + "loss": 0.6209, + "step": 6784 + }, + { + "epoch": 2.382373595505618, + "grad_norm": 0.8830541968345642, + "learning_rate": 3.3208333728616053e-06, + "loss": 0.5714, + "step": 6785 + }, + { + "epoch": 2.3827247191011236, + "grad_norm": 0.9220032095909119, + "learning_rate": 3.3203966639886063e-06, + "loss": 0.6106, + "step": 6786 + }, + { + "epoch": 2.383075842696629, + "grad_norm": 0.8977261185646057, + "learning_rate": 3.3199599270587434e-06, + "loss": 0.6088, + "step": 6787 + }, + { + "epoch": 2.383426966292135, + "grad_norm": 0.9233003258705139, + "learning_rate": 3.3195231620869527e-06, + "loss": 0.5642, + "step": 6788 + }, + { + "epoch": 2.3837780898876404, + "grad_norm": 0.9116259217262268, + "learning_rate": 3.3190863690881703e-06, + "loss": 0.5774, + "step": 6789 + }, + { + "epoch": 2.384129213483146, + "grad_norm": 0.8908228278160095, + "learning_rate": 3.318649548077336e-06, + "loss": 0.5951, + "step": 6790 + }, + { + "epoch": 2.3844803370786516, + "grad_norm": 0.8890783190727234, + "learning_rate": 3.318212699069387e-06, + "loss": 0.5694, + "step": 6791 + }, + { + "epoch": 2.384831460674157, + "grad_norm": 0.8974951505661011, + "learning_rate": 3.317775822079264e-06, + "loss": 0.5726, + "step": 6792 + }, + { + "epoch": 2.385182584269663, + "grad_norm": 0.9875553846359253, + "learning_rate": 3.317338917121908e-06, + "loss": 0.5892, + "step": 6793 + }, + { + "epoch": 2.3855337078651684, + "grad_norm": 0.9124466776847839, + "learning_rate": 3.3169019842122596e-06, + "loss": 0.6282, + "step": 6794 + }, + { + "epoch": 2.385884831460674, + "grad_norm": 0.8811401724815369, + "learning_rate": 3.316465023365263e-06, + "loss": 0.5538, + "step": 6795 + }, + { + "epoch": 2.38623595505618, + "grad_norm": 0.9008293747901917, + "learning_rate": 3.3160280345958614e-06, + "loss": 0.5398, + "step": 6796 + }, + { + "epoch": 2.3865870786516856, + "grad_norm": 0.9087538123130798, + "learning_rate": 3.3155910179189985e-06, + "loss": 0.6113, + "step": 6797 + }, + { + "epoch": 2.386938202247191, + "grad_norm": 0.929850697517395, + "learning_rate": 3.315153973349622e-06, + "loss": 0.5802, + "step": 6798 + }, + { + "epoch": 2.387289325842697, + "grad_norm": 0.8659695982933044, + "learning_rate": 3.3147169009026763e-06, + "loss": 0.5785, + "step": 6799 + }, + { + "epoch": 2.3876404494382024, + "grad_norm": 0.9129486680030823, + "learning_rate": 3.31427980059311e-06, + "loss": 0.6171, + "step": 6800 + }, + { + "epoch": 2.387991573033708, + "grad_norm": 0.9121725559234619, + "learning_rate": 3.3138426724358714e-06, + "loss": 0.5375, + "step": 6801 + }, + { + "epoch": 2.3883426966292136, + "grad_norm": 0.9205370545387268, + "learning_rate": 3.3134055164459095e-06, + "loss": 0.5975, + "step": 6802 + }, + { + "epoch": 2.388693820224719, + "grad_norm": 0.934188723564148, + "learning_rate": 3.312968332638175e-06, + "loss": 0.5874, + "step": 6803 + }, + { + "epoch": 2.389044943820225, + "grad_norm": 0.8987590670585632, + "learning_rate": 3.312531121027619e-06, + "loss": 0.5566, + "step": 6804 + }, + { + "epoch": 2.3893960674157304, + "grad_norm": 0.8946107029914856, + "learning_rate": 3.312093881629194e-06, + "loss": 0.5796, + "step": 6805 + }, + { + "epoch": 2.389747191011236, + "grad_norm": 0.9183394312858582, + "learning_rate": 3.311656614457854e-06, + "loss": 0.5727, + "step": 6806 + }, + { + "epoch": 2.3900983146067416, + "grad_norm": 0.9294660687446594, + "learning_rate": 3.3112193195285516e-06, + "loss": 0.6077, + "step": 6807 + }, + { + "epoch": 2.390449438202247, + "grad_norm": 0.9613252282142639, + "learning_rate": 3.3107819968562428e-06, + "loss": 0.571, + "step": 6808 + }, + { + "epoch": 2.390800561797753, + "grad_norm": 0.9094938635826111, + "learning_rate": 3.310344646455883e-06, + "loss": 0.5562, + "step": 6809 + }, + { + "epoch": 2.3911516853932584, + "grad_norm": 0.8914387226104736, + "learning_rate": 3.3099072683424304e-06, + "loss": 0.5706, + "step": 6810 + }, + { + "epoch": 2.391502808988764, + "grad_norm": 0.9410868287086487, + "learning_rate": 3.3094698625308423e-06, + "loss": 0.6033, + "step": 6811 + }, + { + "epoch": 2.3918539325842696, + "grad_norm": 0.8659963011741638, + "learning_rate": 3.3090324290360775e-06, + "loss": 0.5543, + "step": 6812 + }, + { + "epoch": 2.392205056179775, + "grad_norm": 0.8899027705192566, + "learning_rate": 3.3085949678730953e-06, + "loss": 0.5743, + "step": 6813 + }, + { + "epoch": 2.392556179775281, + "grad_norm": 0.8891961574554443, + "learning_rate": 3.3081574790568583e-06, + "loss": 0.5885, + "step": 6814 + }, + { + "epoch": 2.3929073033707864, + "grad_norm": 0.8514748811721802, + "learning_rate": 3.307719962602326e-06, + "loss": 0.5149, + "step": 6815 + }, + { + "epoch": 2.393258426966292, + "grad_norm": 0.87742018699646, + "learning_rate": 3.307282418524462e-06, + "loss": 0.5873, + "step": 6816 + }, + { + "epoch": 2.3936095505617976, + "grad_norm": 0.9012399315834045, + "learning_rate": 3.3068448468382307e-06, + "loss": 0.5665, + "step": 6817 + }, + { + "epoch": 2.393960674157303, + "grad_norm": 0.9050118923187256, + "learning_rate": 3.306407247558596e-06, + "loss": 0.5569, + "step": 6818 + }, + { + "epoch": 2.394311797752809, + "grad_norm": 0.8786141872406006, + "learning_rate": 3.305969620700524e-06, + "loss": 0.572, + "step": 6819 + }, + { + "epoch": 2.3946629213483144, + "grad_norm": 0.9159024357795715, + "learning_rate": 3.3055319662789797e-06, + "loss": 0.6136, + "step": 6820 + }, + { + "epoch": 2.39501404494382, + "grad_norm": 0.8937184810638428, + "learning_rate": 3.305094284308932e-06, + "loss": 0.6003, + "step": 6821 + }, + { + "epoch": 2.395365168539326, + "grad_norm": 0.8854658603668213, + "learning_rate": 3.3046565748053485e-06, + "loss": 0.5553, + "step": 6822 + }, + { + "epoch": 2.3957162921348316, + "grad_norm": 0.912428617477417, + "learning_rate": 3.3042188377831985e-06, + "loss": 0.5915, + "step": 6823 + }, + { + "epoch": 2.396067415730337, + "grad_norm": 0.8899787068367004, + "learning_rate": 3.3037810732574534e-06, + "loss": 0.5645, + "step": 6824 + }, + { + "epoch": 2.396418539325843, + "grad_norm": 0.9207813739776611, + "learning_rate": 3.3033432812430828e-06, + "loss": 0.614, + "step": 6825 + }, + { + "epoch": 2.3967696629213484, + "grad_norm": 0.9020202159881592, + "learning_rate": 3.302905461755059e-06, + "loss": 0.5775, + "step": 6826 + }, + { + "epoch": 2.397120786516854, + "grad_norm": 0.8828416466712952, + "learning_rate": 3.3024676148083555e-06, + "loss": 0.6004, + "step": 6827 + }, + { + "epoch": 2.3974719101123596, + "grad_norm": 0.9006657600402832, + "learning_rate": 3.302029740417947e-06, + "loss": 0.5861, + "step": 6828 + }, + { + "epoch": 2.397823033707865, + "grad_norm": 0.8664976358413696, + "learning_rate": 3.3015918385988067e-06, + "loss": 0.5481, + "step": 6829 + }, + { + "epoch": 2.398174157303371, + "grad_norm": 0.870573103427887, + "learning_rate": 3.301153909365912e-06, + "loss": 0.5647, + "step": 6830 + }, + { + "epoch": 2.3985252808988764, + "grad_norm": 0.8845220804214478, + "learning_rate": 3.3007159527342394e-06, + "loss": 0.5962, + "step": 6831 + }, + { + "epoch": 2.398876404494382, + "grad_norm": 0.8879493474960327, + "learning_rate": 3.300277968718766e-06, + "loss": 0.573, + "step": 6832 + }, + { + "epoch": 2.3992275280898876, + "grad_norm": 0.9110060930252075, + "learning_rate": 3.299839957334471e-06, + "loss": 0.5849, + "step": 6833 + }, + { + "epoch": 2.399578651685393, + "grad_norm": 0.9112672209739685, + "learning_rate": 3.2994019185963333e-06, + "loss": 0.5955, + "step": 6834 + }, + { + "epoch": 2.399929775280899, + "grad_norm": 0.91420578956604, + "learning_rate": 3.298963852519335e-06, + "loss": 0.586, + "step": 6835 + }, + { + "epoch": 2.4002808988764044, + "grad_norm": 0.8676735758781433, + "learning_rate": 3.2985257591184565e-06, + "loss": 0.5957, + "step": 6836 + }, + { + "epoch": 2.40063202247191, + "grad_norm": 0.9431807994842529, + "learning_rate": 3.2980876384086803e-06, + "loss": 0.5894, + "step": 6837 + }, + { + "epoch": 2.4009831460674156, + "grad_norm": 0.9138767719268799, + "learning_rate": 3.2976494904049895e-06, + "loss": 0.5998, + "step": 6838 + }, + { + "epoch": 2.401334269662921, + "grad_norm": 0.8867050409317017, + "learning_rate": 3.2972113151223688e-06, + "loss": 0.5843, + "step": 6839 + }, + { + "epoch": 2.401685393258427, + "grad_norm": 0.8942441940307617, + "learning_rate": 3.296773112575803e-06, + "loss": 0.5681, + "step": 6840 + }, + { + "epoch": 2.402036516853933, + "grad_norm": 0.9189730882644653, + "learning_rate": 3.296334882780279e-06, + "loss": 0.6028, + "step": 6841 + }, + { + "epoch": 2.4023876404494384, + "grad_norm": 0.9064764976501465, + "learning_rate": 3.295896625750783e-06, + "loss": 0.5705, + "step": 6842 + }, + { + "epoch": 2.402738764044944, + "grad_norm": 0.9189040064811707, + "learning_rate": 3.295458341502304e-06, + "loss": 0.6349, + "step": 6843 + }, + { + "epoch": 2.4030898876404496, + "grad_norm": 0.9180513024330139, + "learning_rate": 3.2950200300498303e-06, + "loss": 0.6263, + "step": 6844 + }, + { + "epoch": 2.403441011235955, + "grad_norm": 0.9278371930122375, + "learning_rate": 3.2945816914083516e-06, + "loss": 0.6262, + "step": 6845 + }, + { + "epoch": 2.403792134831461, + "grad_norm": 0.9270788431167603, + "learning_rate": 3.294143325592859e-06, + "loss": 0.6263, + "step": 6846 + }, + { + "epoch": 2.4041432584269664, + "grad_norm": 0.8792659044265747, + "learning_rate": 3.2937049326183446e-06, + "loss": 0.5904, + "step": 6847 + }, + { + "epoch": 2.404494382022472, + "grad_norm": 0.8935775756835938, + "learning_rate": 3.2932665124998e-06, + "loss": 0.5748, + "step": 6848 + }, + { + "epoch": 2.4048455056179776, + "grad_norm": 0.9019354581832886, + "learning_rate": 3.2928280652522203e-06, + "loss": 0.5758, + "step": 6849 + }, + { + "epoch": 2.405196629213483, + "grad_norm": 0.9184874892234802, + "learning_rate": 3.292389590890599e-06, + "loss": 0.5941, + "step": 6850 + }, + { + "epoch": 2.405547752808989, + "grad_norm": 0.9245009422302246, + "learning_rate": 3.291951089429931e-06, + "loss": 0.5978, + "step": 6851 + }, + { + "epoch": 2.4058988764044944, + "grad_norm": 0.9085261821746826, + "learning_rate": 3.2915125608852142e-06, + "loss": 0.5809, + "step": 6852 + }, + { + "epoch": 2.40625, + "grad_norm": 0.9185128808021545, + "learning_rate": 3.291074005271445e-06, + "loss": 0.6063, + "step": 6853 + }, + { + "epoch": 2.4066011235955056, + "grad_norm": 1.1241658926010132, + "learning_rate": 3.2906354226036218e-06, + "loss": 0.5926, + "step": 6854 + }, + { + "epoch": 2.406952247191011, + "grad_norm": 0.8945859670639038, + "learning_rate": 3.2901968128967438e-06, + "loss": 0.611, + "step": 6855 + }, + { + "epoch": 2.407303370786517, + "grad_norm": 0.9110241532325745, + "learning_rate": 3.289758176165811e-06, + "loss": 0.5686, + "step": 6856 + }, + { + "epoch": 2.4076544943820224, + "grad_norm": 0.9480373859405518, + "learning_rate": 3.289319512425824e-06, + "loss": 0.6113, + "step": 6857 + }, + { + "epoch": 2.408005617977528, + "grad_norm": 0.9290652871131897, + "learning_rate": 3.288880821691785e-06, + "loss": 0.5677, + "step": 6858 + }, + { + "epoch": 2.4083567415730336, + "grad_norm": 0.9119603633880615, + "learning_rate": 3.2884421039786975e-06, + "loss": 0.5987, + "step": 6859 + }, + { + "epoch": 2.408707865168539, + "grad_norm": 0.9131902456283569, + "learning_rate": 3.288003359301564e-06, + "loss": 0.5907, + "step": 6860 + }, + { + "epoch": 2.409058988764045, + "grad_norm": 0.9263566732406616, + "learning_rate": 3.2875645876753902e-06, + "loss": 0.5897, + "step": 6861 + }, + { + "epoch": 2.4094101123595504, + "grad_norm": 0.8975273966789246, + "learning_rate": 3.287125789115182e-06, + "loss": 0.574, + "step": 6862 + }, + { + "epoch": 2.409761235955056, + "grad_norm": 0.9052065014839172, + "learning_rate": 3.2866869636359448e-06, + "loss": 0.6111, + "step": 6863 + }, + { + "epoch": 2.4101123595505616, + "grad_norm": 0.9193122386932373, + "learning_rate": 3.2862481112526866e-06, + "loss": 0.651, + "step": 6864 + }, + { + "epoch": 2.410463483146067, + "grad_norm": 0.8871622085571289, + "learning_rate": 3.285809231980416e-06, + "loss": 0.5958, + "step": 6865 + }, + { + "epoch": 2.410814606741573, + "grad_norm": 0.9120397567749023, + "learning_rate": 3.2853703258341425e-06, + "loss": 0.6263, + "step": 6866 + }, + { + "epoch": 2.411165730337079, + "grad_norm": 0.9002541899681091, + "learning_rate": 3.284931392828875e-06, + "loss": 0.6082, + "step": 6867 + }, + { + "epoch": 2.4115168539325844, + "grad_norm": 0.8848686218261719, + "learning_rate": 3.2844924329796264e-06, + "loss": 0.5742, + "step": 6868 + }, + { + "epoch": 2.41186797752809, + "grad_norm": 0.9187352061271667, + "learning_rate": 3.2840534463014073e-06, + "loss": 0.5953, + "step": 6869 + }, + { + "epoch": 2.4122191011235956, + "grad_norm": 0.8958747386932373, + "learning_rate": 3.283614432809231e-06, + "loss": 0.582, + "step": 6870 + }, + { + "epoch": 2.412570224719101, + "grad_norm": 0.8955165147781372, + "learning_rate": 3.283175392518112e-06, + "loss": 0.5906, + "step": 6871 + }, + { + "epoch": 2.412921348314607, + "grad_norm": 0.9156850576400757, + "learning_rate": 3.282736325443065e-06, + "loss": 0.5925, + "step": 6872 + }, + { + "epoch": 2.4132724719101124, + "grad_norm": 0.8810732364654541, + "learning_rate": 3.2822972315991053e-06, + "loss": 0.5866, + "step": 6873 + }, + { + "epoch": 2.413623595505618, + "grad_norm": 0.9109708070755005, + "learning_rate": 3.2818581110012503e-06, + "loss": 0.5942, + "step": 6874 + }, + { + "epoch": 2.4139747191011236, + "grad_norm": 0.9490804076194763, + "learning_rate": 3.281418963664516e-06, + "loss": 0.6085, + "step": 6875 + }, + { + "epoch": 2.414325842696629, + "grad_norm": 0.9368224143981934, + "learning_rate": 3.280979789603922e-06, + "loss": 0.5748, + "step": 6876 + }, + { + "epoch": 2.414676966292135, + "grad_norm": 0.9140441417694092, + "learning_rate": 3.2805405888344876e-06, + "loss": 0.5758, + "step": 6877 + }, + { + "epoch": 2.4150280898876404, + "grad_norm": 0.9142122268676758, + "learning_rate": 3.280101361371233e-06, + "loss": 0.5971, + "step": 6878 + }, + { + "epoch": 2.415379213483146, + "grad_norm": 0.9765762686729431, + "learning_rate": 3.2796621072291796e-06, + "loss": 0.6124, + "step": 6879 + }, + { + "epoch": 2.4157303370786516, + "grad_norm": 0.9631723165512085, + "learning_rate": 3.2792228264233495e-06, + "loss": 0.5599, + "step": 6880 + }, + { + "epoch": 2.416081460674157, + "grad_norm": 0.911230206489563, + "learning_rate": 3.2787835189687643e-06, + "loss": 0.5963, + "step": 6881 + }, + { + "epoch": 2.416432584269663, + "grad_norm": 0.8858192563056946, + "learning_rate": 3.2783441848804503e-06, + "loss": 0.589, + "step": 6882 + }, + { + "epoch": 2.4167837078651684, + "grad_norm": 0.9075689315795898, + "learning_rate": 3.2779048241734314e-06, + "loss": 0.5827, + "step": 6883 + }, + { + "epoch": 2.417134831460674, + "grad_norm": 0.9115205407142639, + "learning_rate": 3.2774654368627322e-06, + "loss": 0.6046, + "step": 6884 + }, + { + "epoch": 2.41748595505618, + "grad_norm": 0.9490504264831543, + "learning_rate": 3.277026022963381e-06, + "loss": 0.5783, + "step": 6885 + }, + { + "epoch": 2.4178370786516856, + "grad_norm": 0.8726581335067749, + "learning_rate": 3.276586582490404e-06, + "loss": 0.5502, + "step": 6886 + }, + { + "epoch": 2.418188202247191, + "grad_norm": 0.9013791680335999, + "learning_rate": 3.276147115458832e-06, + "loss": 0.575, + "step": 6887 + }, + { + "epoch": 2.418539325842697, + "grad_norm": 0.9145155549049377, + "learning_rate": 3.2757076218836915e-06, + "loss": 0.5922, + "step": 6888 + }, + { + "epoch": 2.4188904494382024, + "grad_norm": 0.8942053318023682, + "learning_rate": 3.275268101780014e-06, + "loss": 0.579, + "step": 6889 + }, + { + "epoch": 2.419241573033708, + "grad_norm": 0.9338133931159973, + "learning_rate": 3.274828555162832e-06, + "loss": 0.595, + "step": 6890 + }, + { + "epoch": 2.4195926966292136, + "grad_norm": 0.8968507647514343, + "learning_rate": 3.274388982047175e-06, + "loss": 0.5801, + "step": 6891 + }, + { + "epoch": 2.419943820224719, + "grad_norm": 0.8962770104408264, + "learning_rate": 3.273949382448078e-06, + "loss": 0.5751, + "step": 6892 + }, + { + "epoch": 2.420294943820225, + "grad_norm": 0.8636572360992432, + "learning_rate": 3.273509756380575e-06, + "loss": 0.5647, + "step": 6893 + }, + { + "epoch": 2.4206460674157304, + "grad_norm": 0.9011679887771606, + "learning_rate": 3.2730701038596996e-06, + "loss": 0.5761, + "step": 6894 + }, + { + "epoch": 2.420997191011236, + "grad_norm": 0.941214382648468, + "learning_rate": 3.272630424900488e-06, + "loss": 0.5787, + "step": 6895 + }, + { + "epoch": 2.4213483146067416, + "grad_norm": 0.904338538646698, + "learning_rate": 3.2721907195179777e-06, + "loss": 0.5571, + "step": 6896 + }, + { + "epoch": 2.421699438202247, + "grad_norm": 0.9070066809654236, + "learning_rate": 3.271750987727205e-06, + "loss": 0.6053, + "step": 6897 + }, + { + "epoch": 2.422050561797753, + "grad_norm": 0.8987091779708862, + "learning_rate": 3.2713112295432093e-06, + "loss": 0.5846, + "step": 6898 + }, + { + "epoch": 2.4224016853932584, + "grad_norm": 0.894283652305603, + "learning_rate": 3.27087144498103e-06, + "loss": 0.5826, + "step": 6899 + }, + { + "epoch": 2.422752808988764, + "grad_norm": 0.8895407319068909, + "learning_rate": 3.2704316340557062e-06, + "loss": 0.5704, + "step": 6900 + }, + { + "epoch": 2.4231039325842696, + "grad_norm": 0.8823187351226807, + "learning_rate": 3.2699917967822797e-06, + "loss": 0.5906, + "step": 6901 + }, + { + "epoch": 2.423455056179775, + "grad_norm": 0.8966806530952454, + "learning_rate": 3.269551933175793e-06, + "loss": 0.5662, + "step": 6902 + }, + { + "epoch": 2.423806179775281, + "grad_norm": 0.9248377084732056, + "learning_rate": 3.2691120432512883e-06, + "loss": 0.5638, + "step": 6903 + }, + { + "epoch": 2.4241573033707864, + "grad_norm": 0.8964454531669617, + "learning_rate": 3.2686721270238105e-06, + "loss": 0.6019, + "step": 6904 + }, + { + "epoch": 2.424508426966292, + "grad_norm": 0.9169251918792725, + "learning_rate": 3.268232184508403e-06, + "loss": 0.6093, + "step": 6905 + }, + { + "epoch": 2.4248595505617976, + "grad_norm": 0.8706364631652832, + "learning_rate": 3.2677922157201124e-06, + "loss": 0.5659, + "step": 6906 + }, + { + "epoch": 2.425210674157303, + "grad_norm": 0.9040411710739136, + "learning_rate": 3.267352220673985e-06, + "loss": 0.5519, + "step": 6907 + }, + { + "epoch": 2.425561797752809, + "grad_norm": 0.9162287712097168, + "learning_rate": 3.2669121993850683e-06, + "loss": 0.5501, + "step": 6908 + }, + { + "epoch": 2.4259129213483144, + "grad_norm": 0.9564346075057983, + "learning_rate": 3.2664721518684106e-06, + "loss": 0.6097, + "step": 6909 + }, + { + "epoch": 2.42626404494382, + "grad_norm": 0.9077656865119934, + "learning_rate": 3.266032078139062e-06, + "loss": 0.6151, + "step": 6910 + }, + { + "epoch": 2.426615168539326, + "grad_norm": 0.9500285387039185, + "learning_rate": 3.265591978212071e-06, + "loss": 0.5853, + "step": 6911 + }, + { + "epoch": 2.4269662921348316, + "grad_norm": 0.903337836265564, + "learning_rate": 3.26515185210249e-06, + "loss": 0.5818, + "step": 6912 + }, + { + "epoch": 2.427317415730337, + "grad_norm": 0.9333521723747253, + "learning_rate": 3.26471169982537e-06, + "loss": 0.5796, + "step": 6913 + }, + { + "epoch": 2.427668539325843, + "grad_norm": 0.9057077765464783, + "learning_rate": 3.2642715213957644e-06, + "loss": 0.5869, + "step": 6914 + }, + { + "epoch": 2.4280196629213484, + "grad_norm": 0.9188728332519531, + "learning_rate": 3.2638313168287265e-06, + "loss": 0.5917, + "step": 6915 + }, + { + "epoch": 2.428370786516854, + "grad_norm": 0.9184112548828125, + "learning_rate": 3.2633910861393115e-06, + "loss": 0.6279, + "step": 6916 + }, + { + "epoch": 2.4287219101123596, + "grad_norm": 0.8528975248336792, + "learning_rate": 3.262950829342574e-06, + "loss": 0.5757, + "step": 6917 + }, + { + "epoch": 2.429073033707865, + "grad_norm": 0.8992438316345215, + "learning_rate": 3.2625105464535718e-06, + "loss": 0.5949, + "step": 6918 + }, + { + "epoch": 2.429424157303371, + "grad_norm": 0.9076921343803406, + "learning_rate": 3.2620702374873607e-06, + "loss": 0.5852, + "step": 6919 + }, + { + "epoch": 2.4297752808988764, + "grad_norm": 0.8996908068656921, + "learning_rate": 3.2616299024590005e-06, + "loss": 0.5794, + "step": 6920 + }, + { + "epoch": 2.430126404494382, + "grad_norm": 0.9273661375045776, + "learning_rate": 3.2611895413835482e-06, + "loss": 0.6095, + "step": 6921 + }, + { + "epoch": 2.4304775280898876, + "grad_norm": 0.8781593441963196, + "learning_rate": 3.2607491542760656e-06, + "loss": 0.5613, + "step": 6922 + }, + { + "epoch": 2.430828651685393, + "grad_norm": 0.8770261406898499, + "learning_rate": 3.260308741151613e-06, + "loss": 0.594, + "step": 6923 + }, + { + "epoch": 2.431179775280899, + "grad_norm": 0.8968514204025269, + "learning_rate": 3.259868302025252e-06, + "loss": 0.5835, + "step": 6924 + }, + { + "epoch": 2.4315308988764044, + "grad_norm": 0.906139075756073, + "learning_rate": 3.259427836912046e-06, + "loss": 0.6075, + "step": 6925 + }, + { + "epoch": 2.43188202247191, + "grad_norm": 0.8941826224327087, + "learning_rate": 3.2589873458270567e-06, + "loss": 0.5609, + "step": 6926 + }, + { + "epoch": 2.4322331460674156, + "grad_norm": 0.923839807510376, + "learning_rate": 3.2585468287853493e-06, + "loss": 0.5788, + "step": 6927 + }, + { + "epoch": 2.432584269662921, + "grad_norm": 0.8710209727287292, + "learning_rate": 3.2581062858019907e-06, + "loss": 0.5906, + "step": 6928 + }, + { + "epoch": 2.432935393258427, + "grad_norm": 0.9340402483940125, + "learning_rate": 3.257665716892045e-06, + "loss": 0.6117, + "step": 6929 + }, + { + "epoch": 2.433286516853933, + "grad_norm": 0.877828061580658, + "learning_rate": 3.257225122070581e-06, + "loss": 0.5706, + "step": 6930 + }, + { + "epoch": 2.4336376404494384, + "grad_norm": 0.9005634784698486, + "learning_rate": 3.2567845013526646e-06, + "loss": 0.6185, + "step": 6931 + }, + { + "epoch": 2.433988764044944, + "grad_norm": 0.9182244539260864, + "learning_rate": 3.2563438547533665e-06, + "loss": 0.598, + "step": 6932 + }, + { + "epoch": 2.4343398876404496, + "grad_norm": 0.9094123840332031, + "learning_rate": 3.2559031822877557e-06, + "loss": 0.6246, + "step": 6933 + }, + { + "epoch": 2.434691011235955, + "grad_norm": 0.8959712386131287, + "learning_rate": 3.255462483970904e-06, + "loss": 0.5671, + "step": 6934 + }, + { + "epoch": 2.435042134831461, + "grad_norm": 0.8880782723426819, + "learning_rate": 3.25502175981788e-06, + "loss": 0.5792, + "step": 6935 + }, + { + "epoch": 2.4353932584269664, + "grad_norm": 0.8947858810424805, + "learning_rate": 3.254581009843759e-06, + "loss": 0.5598, + "step": 6936 + }, + { + "epoch": 2.435744382022472, + "grad_norm": 0.9413600564002991, + "learning_rate": 3.2541402340636125e-06, + "loss": 0.5889, + "step": 6937 + }, + { + "epoch": 2.4360955056179776, + "grad_norm": 0.8757100105285645, + "learning_rate": 3.2536994324925157e-06, + "loss": 0.61, + "step": 6938 + }, + { + "epoch": 2.436446629213483, + "grad_norm": 1.020088791847229, + "learning_rate": 3.2532586051455426e-06, + "loss": 0.5945, + "step": 6939 + }, + { + "epoch": 2.436797752808989, + "grad_norm": 0.9095579385757446, + "learning_rate": 3.252817752037771e-06, + "loss": 0.5238, + "step": 6940 + }, + { + "epoch": 2.4371488764044944, + "grad_norm": 0.8991023898124695, + "learning_rate": 3.2523768731842765e-06, + "loss": 0.5699, + "step": 6941 + }, + { + "epoch": 2.4375, + "grad_norm": 0.8953317999839783, + "learning_rate": 3.2519359686001357e-06, + "loss": 0.6174, + "step": 6942 + }, + { + "epoch": 2.4378511235955056, + "grad_norm": 0.8995128870010376, + "learning_rate": 3.2514950383004283e-06, + "loss": 0.591, + "step": 6943 + }, + { + "epoch": 2.438202247191011, + "grad_norm": 0.8980720043182373, + "learning_rate": 3.2510540823002344e-06, + "loss": 0.6269, + "step": 6944 + }, + { + "epoch": 2.438553370786517, + "grad_norm": 0.8931188583374023, + "learning_rate": 3.250613100614633e-06, + "loss": 0.5485, + "step": 6945 + }, + { + "epoch": 2.4389044943820224, + "grad_norm": 0.9148392677307129, + "learning_rate": 3.250172093258706e-06, + "loss": 0.6075, + "step": 6946 + }, + { + "epoch": 2.439255617977528, + "grad_norm": 0.8999046683311462, + "learning_rate": 3.2497310602475352e-06, + "loss": 0.5717, + "step": 6947 + }, + { + "epoch": 2.4396067415730336, + "grad_norm": 0.9245252013206482, + "learning_rate": 3.2492900015962036e-06, + "loss": 0.6114, + "step": 6948 + }, + { + "epoch": 2.439957865168539, + "grad_norm": 0.8906577229499817, + "learning_rate": 3.2488489173197956e-06, + "loss": 0.5598, + "step": 6949 + }, + { + "epoch": 2.440308988764045, + "grad_norm": 0.9028708934783936, + "learning_rate": 3.248407807433396e-06, + "loss": 0.5917, + "step": 6950 + }, + { + "epoch": 2.4406601123595504, + "grad_norm": 0.9025599956512451, + "learning_rate": 3.2479666719520887e-06, + "loss": 0.5809, + "step": 6951 + }, + { + "epoch": 2.441011235955056, + "grad_norm": 0.9142019152641296, + "learning_rate": 3.2475255108909616e-06, + "loss": 0.6064, + "step": 6952 + }, + { + "epoch": 2.4413623595505616, + "grad_norm": 0.8819493055343628, + "learning_rate": 3.247084324265102e-06, + "loss": 0.5677, + "step": 6953 + }, + { + "epoch": 2.441713483146067, + "grad_norm": 0.8866091370582581, + "learning_rate": 3.246643112089597e-06, + "loss": 0.5563, + "step": 6954 + }, + { + "epoch": 2.442064606741573, + "grad_norm": 0.9083012938499451, + "learning_rate": 3.2462018743795366e-06, + "loss": 0.5826, + "step": 6955 + }, + { + "epoch": 2.442415730337079, + "grad_norm": 0.9067873358726501, + "learning_rate": 3.245760611150011e-06, + "loss": 0.6065, + "step": 6956 + }, + { + "epoch": 2.4427668539325844, + "grad_norm": 0.8973134756088257, + "learning_rate": 3.2453193224161105e-06, + "loss": 0.597, + "step": 6957 + }, + { + "epoch": 2.44311797752809, + "grad_norm": 0.9013165235519409, + "learning_rate": 3.244878008192927e-06, + "loss": 0.5931, + "step": 6958 + }, + { + "epoch": 2.4434691011235956, + "grad_norm": 0.9038127064704895, + "learning_rate": 3.244436668495553e-06, + "loss": 0.5869, + "step": 6959 + }, + { + "epoch": 2.443820224719101, + "grad_norm": 0.9094183444976807, + "learning_rate": 3.2439953033390823e-06, + "loss": 0.5826, + "step": 6960 + }, + { + "epoch": 2.444171348314607, + "grad_norm": 0.8923176527023315, + "learning_rate": 3.2435539127386083e-06, + "loss": 0.5772, + "step": 6961 + }, + { + "epoch": 2.4445224719101124, + "grad_norm": 0.9041379690170288, + "learning_rate": 3.2431124967092282e-06, + "loss": 0.5753, + "step": 6962 + }, + { + "epoch": 2.444873595505618, + "grad_norm": 0.9126103520393372, + "learning_rate": 3.2426710552660353e-06, + "loss": 0.5827, + "step": 6963 + }, + { + "epoch": 2.4452247191011236, + "grad_norm": 0.9693210124969482, + "learning_rate": 3.242229588424128e-06, + "loss": 0.5839, + "step": 6964 + }, + { + "epoch": 2.445575842696629, + "grad_norm": 0.9098565578460693, + "learning_rate": 3.241788096198604e-06, + "loss": 0.6022, + "step": 6965 + }, + { + "epoch": 2.445926966292135, + "grad_norm": 0.9330307245254517, + "learning_rate": 3.2413465786045617e-06, + "loss": 0.5678, + "step": 6966 + }, + { + "epoch": 2.4462780898876404, + "grad_norm": 0.8912758231163025, + "learning_rate": 3.2409050356571005e-06, + "loss": 0.5449, + "step": 6967 + }, + { + "epoch": 2.446629213483146, + "grad_norm": 0.8749823570251465, + "learning_rate": 3.2404634673713213e-06, + "loss": 0.5641, + "step": 6968 + }, + { + "epoch": 2.4469803370786516, + "grad_norm": 0.913131058216095, + "learning_rate": 3.2400218737623258e-06, + "loss": 0.583, + "step": 6969 + }, + { + "epoch": 2.447331460674157, + "grad_norm": 0.9098646640777588, + "learning_rate": 3.239580254845215e-06, + "loss": 0.5834, + "step": 6970 + }, + { + "epoch": 2.447682584269663, + "grad_norm": 0.9374386072158813, + "learning_rate": 3.239138610635092e-06, + "loss": 0.5887, + "step": 6971 + }, + { + "epoch": 2.4480337078651684, + "grad_norm": 0.9426707625389099, + "learning_rate": 3.2386969411470613e-06, + "loss": 0.615, + "step": 6972 + }, + { + "epoch": 2.448384831460674, + "grad_norm": 0.8913367986679077, + "learning_rate": 3.238255246396227e-06, + "loss": 0.5893, + "step": 6973 + }, + { + "epoch": 2.44873595505618, + "grad_norm": 0.9027500152587891, + "learning_rate": 3.237813526397696e-06, + "loss": 0.5649, + "step": 6974 + }, + { + "epoch": 2.4490870786516856, + "grad_norm": 0.9199661612510681, + "learning_rate": 3.237371781166573e-06, + "loss": 0.6005, + "step": 6975 + }, + { + "epoch": 2.449438202247191, + "grad_norm": 0.8691537976264954, + "learning_rate": 3.2369300107179664e-06, + "loss": 0.6295, + "step": 6976 + }, + { + "epoch": 2.449789325842697, + "grad_norm": 0.8786364197731018, + "learning_rate": 3.2364882150669833e-06, + "loss": 0.6122, + "step": 6977 + }, + { + "epoch": 2.4501404494382024, + "grad_norm": 0.9122487306594849, + "learning_rate": 3.236046394228734e-06, + "loss": 0.6362, + "step": 6978 + }, + { + "epoch": 2.450491573033708, + "grad_norm": 0.9142274260520935, + "learning_rate": 3.235604548218328e-06, + "loss": 0.5622, + "step": 6979 + }, + { + "epoch": 2.4508426966292136, + "grad_norm": 0.8734898567199707, + "learning_rate": 3.2351626770508758e-06, + "loss": 0.5738, + "step": 6980 + }, + { + "epoch": 2.451193820224719, + "grad_norm": 0.8727086186408997, + "learning_rate": 3.234720780741489e-06, + "loss": 0.5818, + "step": 6981 + }, + { + "epoch": 2.451544943820225, + "grad_norm": 0.8629016876220703, + "learning_rate": 3.2342788593052797e-06, + "loss": 0.5707, + "step": 6982 + }, + { + "epoch": 2.4518960674157304, + "grad_norm": 0.9243472218513489, + "learning_rate": 3.2338369127573626e-06, + "loss": 0.5845, + "step": 6983 + }, + { + "epoch": 2.452247191011236, + "grad_norm": 0.8935674428939819, + "learning_rate": 3.2333949411128506e-06, + "loss": 0.6005, + "step": 6984 + }, + { + "epoch": 2.4525983146067416, + "grad_norm": 0.889320433139801, + "learning_rate": 3.2329529443868596e-06, + "loss": 0.5889, + "step": 6985 + }, + { + "epoch": 2.452949438202247, + "grad_norm": 0.8866029977798462, + "learning_rate": 3.232510922594505e-06, + "loss": 0.5323, + "step": 6986 + }, + { + "epoch": 2.453300561797753, + "grad_norm": 0.8897583484649658, + "learning_rate": 3.2320688757509034e-06, + "loss": 0.5739, + "step": 6987 + }, + { + "epoch": 2.4536516853932584, + "grad_norm": 0.9142066836357117, + "learning_rate": 3.231626803871173e-06, + "loss": 0.5877, + "step": 6988 + }, + { + "epoch": 2.454002808988764, + "grad_norm": 0.879382848739624, + "learning_rate": 3.2311847069704315e-06, + "loss": 0.5789, + "step": 6989 + }, + { + "epoch": 2.4543539325842696, + "grad_norm": 0.8784737586975098, + "learning_rate": 3.2307425850638e-06, + "loss": 0.5688, + "step": 6990 + }, + { + "epoch": 2.454705056179775, + "grad_norm": 0.8837603330612183, + "learning_rate": 3.230300438166396e-06, + "loss": 0.5503, + "step": 6991 + }, + { + "epoch": 2.455056179775281, + "grad_norm": 0.903950572013855, + "learning_rate": 3.229858266293342e-06, + "loss": 0.5677, + "step": 6992 + }, + { + "epoch": 2.4554073033707864, + "grad_norm": 0.9059702157974243, + "learning_rate": 3.22941606945976e-06, + "loss": 0.5766, + "step": 6993 + }, + { + "epoch": 2.455758426966292, + "grad_norm": 0.9046807289123535, + "learning_rate": 3.228973847680773e-06, + "loss": 0.5625, + "step": 6994 + }, + { + "epoch": 2.4561095505617976, + "grad_norm": 0.8932182192802429, + "learning_rate": 3.228531600971504e-06, + "loss": 0.5716, + "step": 6995 + }, + { + "epoch": 2.456460674157303, + "grad_norm": 0.9098941087722778, + "learning_rate": 3.228089329347077e-06, + "loss": 0.5774, + "step": 6996 + }, + { + "epoch": 2.456811797752809, + "grad_norm": 0.9209422469139099, + "learning_rate": 3.227647032822619e-06, + "loss": 0.6022, + "step": 6997 + }, + { + "epoch": 2.4571629213483144, + "grad_norm": 0.9338439106941223, + "learning_rate": 3.2272047114132543e-06, + "loss": 0.6005, + "step": 6998 + }, + { + "epoch": 2.45751404494382, + "grad_norm": 0.9156144261360168, + "learning_rate": 3.226762365134112e-06, + "loss": 0.567, + "step": 6999 + }, + { + "epoch": 2.457865168539326, + "grad_norm": 0.8707544803619385, + "learning_rate": 3.2263199940003175e-06, + "loss": 0.5624, + "step": 7000 + }, + { + "epoch": 2.4582162921348316, + "grad_norm": 0.917948305606842, + "learning_rate": 3.2258775980270015e-06, + "loss": 0.6112, + "step": 7001 + }, + { + "epoch": 2.458567415730337, + "grad_norm": 0.9017558693885803, + "learning_rate": 3.2254351772292923e-06, + "loss": 0.601, + "step": 7002 + }, + { + "epoch": 2.458918539325843, + "grad_norm": 0.907769501209259, + "learning_rate": 3.2249927316223206e-06, + "loss": 0.5823, + "step": 7003 + }, + { + "epoch": 2.4592696629213484, + "grad_norm": 0.9365431070327759, + "learning_rate": 3.2245502612212186e-06, + "loss": 0.6082, + "step": 7004 + }, + { + "epoch": 2.459620786516854, + "grad_norm": 0.9162115454673767, + "learning_rate": 3.224107766041117e-06, + "loss": 0.5695, + "step": 7005 + }, + { + "epoch": 2.4599719101123596, + "grad_norm": 0.886081337928772, + "learning_rate": 3.2236652460971497e-06, + "loss": 0.6094, + "step": 7006 + }, + { + "epoch": 2.460323033707865, + "grad_norm": 0.9087616205215454, + "learning_rate": 3.22322270140445e-06, + "loss": 0.5916, + "step": 7007 + }, + { + "epoch": 2.460674157303371, + "grad_norm": 0.9102017879486084, + "learning_rate": 3.222780131978153e-06, + "loss": 0.5894, + "step": 7008 + }, + { + "epoch": 2.4610252808988764, + "grad_norm": 0.8942725658416748, + "learning_rate": 3.2223375378333938e-06, + "loss": 0.6147, + "step": 7009 + }, + { + "epoch": 2.461376404494382, + "grad_norm": 0.8880075216293335, + "learning_rate": 3.2218949189853094e-06, + "loss": 0.5788, + "step": 7010 + }, + { + "epoch": 2.4617275280898876, + "grad_norm": 0.8984467387199402, + "learning_rate": 3.221452275449036e-06, + "loss": 0.587, + "step": 7011 + }, + { + "epoch": 2.462078651685393, + "grad_norm": 0.9137831926345825, + "learning_rate": 3.2210096072397124e-06, + "loss": 0.5764, + "step": 7012 + }, + { + "epoch": 2.462429775280899, + "grad_norm": 0.9108327627182007, + "learning_rate": 3.220566914372477e-06, + "loss": 0.6134, + "step": 7013 + }, + { + "epoch": 2.4627808988764044, + "grad_norm": 0.8871783018112183, + "learning_rate": 3.22012419686247e-06, + "loss": 0.583, + "step": 7014 + }, + { + "epoch": 2.46313202247191, + "grad_norm": 0.9132178425788879, + "learning_rate": 3.2196814547248316e-06, + "loss": 0.596, + "step": 7015 + }, + { + "epoch": 2.4634831460674156, + "grad_norm": 0.8959728479385376, + "learning_rate": 3.2192386879747035e-06, + "loss": 0.5779, + "step": 7016 + }, + { + "epoch": 2.463834269662921, + "grad_norm": 0.8945717215538025, + "learning_rate": 3.218795896627227e-06, + "loss": 0.5819, + "step": 7017 + }, + { + "epoch": 2.464185393258427, + "grad_norm": 0.9090186953544617, + "learning_rate": 3.2183530806975465e-06, + "loss": 0.5788, + "step": 7018 + }, + { + "epoch": 2.464536516853933, + "grad_norm": 0.8839727640151978, + "learning_rate": 3.2179102402008057e-06, + "loss": 0.5631, + "step": 7019 + }, + { + "epoch": 2.4648876404494384, + "grad_norm": 0.9012869000434875, + "learning_rate": 3.217467375152148e-06, + "loss": 0.577, + "step": 7020 + }, + { + "epoch": 2.465238764044944, + "grad_norm": 0.9183435440063477, + "learning_rate": 3.217024485566721e-06, + "loss": 0.5979, + "step": 7021 + }, + { + "epoch": 2.4655898876404496, + "grad_norm": 0.876729428768158, + "learning_rate": 3.21658157145967e-06, + "loss": 0.5978, + "step": 7022 + }, + { + "epoch": 2.465941011235955, + "grad_norm": 0.9012320041656494, + "learning_rate": 3.216138632846142e-06, + "loss": 0.5938, + "step": 7023 + }, + { + "epoch": 2.466292134831461, + "grad_norm": 0.9024367928504944, + "learning_rate": 3.2156956697412854e-06, + "loss": 0.5894, + "step": 7024 + }, + { + "epoch": 2.4666432584269664, + "grad_norm": 0.918143630027771, + "learning_rate": 3.2152526821602498e-06, + "loss": 0.593, + "step": 7025 + }, + { + "epoch": 2.466994382022472, + "grad_norm": 0.8777908682823181, + "learning_rate": 3.2148096701181842e-06, + "loss": 0.5499, + "step": 7026 + }, + { + "epoch": 2.4673455056179776, + "grad_norm": 0.9229316711425781, + "learning_rate": 3.21436663363024e-06, + "loss": 0.5971, + "step": 7027 + }, + { + "epoch": 2.467696629213483, + "grad_norm": 0.9105817079544067, + "learning_rate": 3.213923572711568e-06, + "loss": 0.5769, + "step": 7028 + }, + { + "epoch": 2.468047752808989, + "grad_norm": 0.9043117165565491, + "learning_rate": 3.213480487377321e-06, + "loss": 0.6004, + "step": 7029 + }, + { + "epoch": 2.4683988764044944, + "grad_norm": 0.901021420955658, + "learning_rate": 3.2130373776426515e-06, + "loss": 0.5877, + "step": 7030 + }, + { + "epoch": 2.46875, + "grad_norm": 0.932961642742157, + "learning_rate": 3.2125942435227137e-06, + "loss": 0.6153, + "step": 7031 + }, + { + "epoch": 2.4691011235955056, + "grad_norm": 0.93779456615448, + "learning_rate": 3.2121510850326627e-06, + "loss": 0.6086, + "step": 7032 + }, + { + "epoch": 2.469452247191011, + "grad_norm": 0.8923274874687195, + "learning_rate": 3.211707902187654e-06, + "loss": 0.5873, + "step": 7033 + }, + { + "epoch": 2.469803370786517, + "grad_norm": 0.9306513071060181, + "learning_rate": 3.2112646950028443e-06, + "loss": 0.5999, + "step": 7034 + }, + { + "epoch": 2.4701544943820224, + "grad_norm": 0.9095919728279114, + "learning_rate": 3.2108214634933898e-06, + "loss": 0.6012, + "step": 7035 + }, + { + "epoch": 2.470505617977528, + "grad_norm": 0.8742579817771912, + "learning_rate": 3.2103782076744504e-06, + "loss": 0.5957, + "step": 7036 + }, + { + "epoch": 2.4708567415730336, + "grad_norm": 0.9040021896362305, + "learning_rate": 3.209934927561184e-06, + "loss": 0.6177, + "step": 7037 + }, + { + "epoch": 2.471207865168539, + "grad_norm": 0.8758634924888611, + "learning_rate": 3.209491623168751e-06, + "loss": 0.584, + "step": 7038 + }, + { + "epoch": 2.471558988764045, + "grad_norm": 0.9028114080429077, + "learning_rate": 3.2090482945123107e-06, + "loss": 0.5748, + "step": 7039 + }, + { + "epoch": 2.4719101123595504, + "grad_norm": 0.8887829780578613, + "learning_rate": 3.2086049416070263e-06, + "loss": 0.5793, + "step": 7040 + }, + { + "epoch": 2.472261235955056, + "grad_norm": 0.9378764629364014, + "learning_rate": 3.2081615644680585e-06, + "loss": 0.6022, + "step": 7041 + }, + { + "epoch": 2.4726123595505616, + "grad_norm": 0.8793286085128784, + "learning_rate": 3.2077181631105714e-06, + "loss": 0.5574, + "step": 7042 + }, + { + "epoch": 2.472963483146067, + "grad_norm": 0.9339090585708618, + "learning_rate": 3.2072747375497294e-06, + "loss": 0.6092, + "step": 7043 + }, + { + "epoch": 2.473314606741573, + "grad_norm": 0.9024748802185059, + "learning_rate": 3.2068312878006956e-06, + "loss": 0.6177, + "step": 7044 + }, + { + "epoch": 2.473665730337079, + "grad_norm": 0.9213776588439941, + "learning_rate": 3.206387813878637e-06, + "loss": 0.5859, + "step": 7045 + }, + { + "epoch": 2.4740168539325844, + "grad_norm": 0.9179170727729797, + "learning_rate": 3.2059443157987195e-06, + "loss": 0.6103, + "step": 7046 + }, + { + "epoch": 2.47436797752809, + "grad_norm": 0.9424740076065063, + "learning_rate": 3.2055007935761106e-06, + "loss": 0.5685, + "step": 7047 + }, + { + "epoch": 2.4747191011235956, + "grad_norm": 0.9363864064216614, + "learning_rate": 3.2050572472259783e-06, + "loss": 0.6268, + "step": 7048 + }, + { + "epoch": 2.475070224719101, + "grad_norm": 0.8998108506202698, + "learning_rate": 3.2046136767634916e-06, + "loss": 0.5696, + "step": 7049 + }, + { + "epoch": 2.475421348314607, + "grad_norm": 0.8754726648330688, + "learning_rate": 3.20417008220382e-06, + "loss": 0.5439, + "step": 7050 + }, + { + "epoch": 2.4757724719101124, + "grad_norm": 0.9287495613098145, + "learning_rate": 3.2037264635621346e-06, + "loss": 0.5838, + "step": 7051 + }, + { + "epoch": 2.476123595505618, + "grad_norm": 0.9086918830871582, + "learning_rate": 3.203282820853606e-06, + "loss": 0.5619, + "step": 7052 + }, + { + "epoch": 2.4764747191011236, + "grad_norm": 0.8700255155563354, + "learning_rate": 3.2028391540934066e-06, + "loss": 0.5975, + "step": 7053 + }, + { + "epoch": 2.476825842696629, + "grad_norm": 0.9084519743919373, + "learning_rate": 3.202395463296709e-06, + "loss": 0.6084, + "step": 7054 + }, + { + "epoch": 2.477176966292135, + "grad_norm": 0.9101182222366333, + "learning_rate": 3.2019517484786887e-06, + "loss": 0.6024, + "step": 7055 + }, + { + "epoch": 2.4775280898876404, + "grad_norm": 0.9299564361572266, + "learning_rate": 3.2015080096545188e-06, + "loss": 0.6141, + "step": 7056 + }, + { + "epoch": 2.477879213483146, + "grad_norm": 0.9324676990509033, + "learning_rate": 3.2010642468393748e-06, + "loss": 0.605, + "step": 7057 + }, + { + "epoch": 2.4782303370786516, + "grad_norm": 0.9061481356620789, + "learning_rate": 3.2006204600484338e-06, + "loss": 0.5703, + "step": 7058 + }, + { + "epoch": 2.478581460674157, + "grad_norm": 0.858217716217041, + "learning_rate": 3.200176649296873e-06, + "loss": 0.5715, + "step": 7059 + }, + { + "epoch": 2.478932584269663, + "grad_norm": 0.9061951041221619, + "learning_rate": 3.1997328145998694e-06, + "loss": 0.6095, + "step": 7060 + }, + { + "epoch": 2.4792837078651684, + "grad_norm": 0.9107330441474915, + "learning_rate": 3.1992889559726025e-06, + "loss": 0.5824, + "step": 7061 + }, + { + "epoch": 2.479634831460674, + "grad_norm": 0.9080485105514526, + "learning_rate": 3.1988450734302518e-06, + "loss": 0.581, + "step": 7062 + }, + { + "epoch": 2.47998595505618, + "grad_norm": 0.8972571492195129, + "learning_rate": 3.198401166987998e-06, + "loss": 0.5816, + "step": 7063 + }, + { + "epoch": 2.4803370786516856, + "grad_norm": 0.8946796655654907, + "learning_rate": 3.197957236661021e-06, + "loss": 0.5769, + "step": 7064 + }, + { + "epoch": 2.480688202247191, + "grad_norm": 0.9583724737167358, + "learning_rate": 3.1975132824645044e-06, + "loss": 0.5728, + "step": 7065 + }, + { + "epoch": 2.481039325842697, + "grad_norm": 0.9171568155288696, + "learning_rate": 3.1970693044136302e-06, + "loss": 0.5708, + "step": 7066 + }, + { + "epoch": 2.4813904494382024, + "grad_norm": 0.9041579365730286, + "learning_rate": 3.196625302523582e-06, + "loss": 0.5876, + "step": 7067 + }, + { + "epoch": 2.481741573033708, + "grad_norm": 0.9154647588729858, + "learning_rate": 3.1961812768095452e-06, + "loss": 0.5818, + "step": 7068 + }, + { + "epoch": 2.4820926966292136, + "grad_norm": 0.928131103515625, + "learning_rate": 3.195737227286704e-06, + "loss": 0.5572, + "step": 7069 + }, + { + "epoch": 2.482443820224719, + "grad_norm": 0.8931006193161011, + "learning_rate": 3.1952931539702447e-06, + "loss": 0.5989, + "step": 7070 + }, + { + "epoch": 2.482794943820225, + "grad_norm": 0.8964663743972778, + "learning_rate": 3.1948490568753556e-06, + "loss": 0.5801, + "step": 7071 + }, + { + "epoch": 2.4831460674157304, + "grad_norm": 0.915996253490448, + "learning_rate": 3.1944049360172223e-06, + "loss": 0.6025, + "step": 7072 + }, + { + "epoch": 2.483497191011236, + "grad_norm": 0.9158309698104858, + "learning_rate": 3.193960791411035e-06, + "loss": 0.5906, + "step": 7073 + }, + { + "epoch": 2.4838483146067416, + "grad_norm": 0.8999008536338806, + "learning_rate": 3.1935166230719816e-06, + "loss": 0.5695, + "step": 7074 + }, + { + "epoch": 2.484199438202247, + "grad_norm": 0.8934087157249451, + "learning_rate": 3.193072431015254e-06, + "loss": 0.5601, + "step": 7075 + }, + { + "epoch": 2.484550561797753, + "grad_norm": 0.873293936252594, + "learning_rate": 3.192628215256042e-06, + "loss": 0.5642, + "step": 7076 + }, + { + "epoch": 2.4849016853932584, + "grad_norm": 0.9033623337745667, + "learning_rate": 3.192183975809538e-06, + "loss": 0.5885, + "step": 7077 + }, + { + "epoch": 2.485252808988764, + "grad_norm": 0.912635087966919, + "learning_rate": 3.1917397126909344e-06, + "loss": 0.5778, + "step": 7078 + }, + { + "epoch": 2.4856039325842696, + "grad_norm": 0.8821267485618591, + "learning_rate": 3.1912954259154237e-06, + "loss": 0.6107, + "step": 7079 + }, + { + "epoch": 2.485955056179775, + "grad_norm": 0.87236088514328, + "learning_rate": 3.190851115498202e-06, + "loss": 0.5641, + "step": 7080 + }, + { + "epoch": 2.486306179775281, + "grad_norm": 0.9131118655204773, + "learning_rate": 3.1904067814544627e-06, + "loss": 0.5586, + "step": 7081 + }, + { + "epoch": 2.4866573033707864, + "grad_norm": 0.9087355732917786, + "learning_rate": 3.1899624237994027e-06, + "loss": 0.5739, + "step": 7082 + }, + { + "epoch": 2.487008426966292, + "grad_norm": 0.8918299674987793, + "learning_rate": 3.1895180425482177e-06, + "loss": 0.5989, + "step": 7083 + }, + { + "epoch": 2.4873595505617976, + "grad_norm": 0.8997132182121277, + "learning_rate": 3.189073637716106e-06, + "loss": 0.6219, + "step": 7084 + }, + { + "epoch": 2.487710674157303, + "grad_norm": 0.872758150100708, + "learning_rate": 3.1886292093182657e-06, + "loss": 0.5799, + "step": 7085 + }, + { + "epoch": 2.488061797752809, + "grad_norm": 0.9139792919158936, + "learning_rate": 3.1881847573698954e-06, + "loss": 0.6014, + "step": 7086 + }, + { + "epoch": 2.4884129213483144, + "grad_norm": 0.8912699818611145, + "learning_rate": 3.1877402818861954e-06, + "loss": 0.5512, + "step": 7087 + }, + { + "epoch": 2.48876404494382, + "grad_norm": 0.8925781846046448, + "learning_rate": 3.1872957828823658e-06, + "loss": 0.5476, + "step": 7088 + }, + { + "epoch": 2.489115168539326, + "grad_norm": 0.9175998568534851, + "learning_rate": 3.1868512603736094e-06, + "loss": 0.5486, + "step": 7089 + }, + { + "epoch": 2.4894662921348316, + "grad_norm": 0.8994670510292053, + "learning_rate": 3.1864067143751277e-06, + "loss": 0.5758, + "step": 7090 + }, + { + "epoch": 2.489817415730337, + "grad_norm": 0.9013662338256836, + "learning_rate": 3.185962144902123e-06, + "loss": 0.5592, + "step": 7091 + }, + { + "epoch": 2.490168539325843, + "grad_norm": 0.9046351313591003, + "learning_rate": 3.1855175519698e-06, + "loss": 0.6096, + "step": 7092 + }, + { + "epoch": 2.4905196629213484, + "grad_norm": 0.9141769409179688, + "learning_rate": 3.185072935593364e-06, + "loss": 0.5602, + "step": 7093 + }, + { + "epoch": 2.490870786516854, + "grad_norm": 0.9111477136611938, + "learning_rate": 3.184628295788019e-06, + "loss": 0.5924, + "step": 7094 + }, + { + "epoch": 2.4912219101123596, + "grad_norm": 0.8942867517471313, + "learning_rate": 3.1841836325689722e-06, + "loss": 0.5748, + "step": 7095 + }, + { + "epoch": 2.491573033707865, + "grad_norm": 0.9035190939903259, + "learning_rate": 3.1837389459514313e-06, + "loss": 0.597, + "step": 7096 + }, + { + "epoch": 2.491924157303371, + "grad_norm": 0.8901208639144897, + "learning_rate": 3.183294235950603e-06, + "loss": 0.5631, + "step": 7097 + }, + { + "epoch": 2.4922752808988764, + "grad_norm": 0.8867502212524414, + "learning_rate": 3.182849502581697e-06, + "loss": 0.5716, + "step": 7098 + }, + { + "epoch": 2.492626404494382, + "grad_norm": 0.9300554990768433, + "learning_rate": 3.182404745859922e-06, + "loss": 0.6092, + "step": 7099 + }, + { + "epoch": 2.4929775280898876, + "grad_norm": 0.9296733736991882, + "learning_rate": 3.181959965800489e-06, + "loss": 0.5642, + "step": 7100 + }, + { + "epoch": 2.493328651685393, + "grad_norm": 0.923528790473938, + "learning_rate": 3.1815151624186084e-06, + "loss": 0.5871, + "step": 7101 + }, + { + "epoch": 2.493679775280899, + "grad_norm": 0.930453360080719, + "learning_rate": 3.1810703357294927e-06, + "loss": 0.5785, + "step": 7102 + }, + { + "epoch": 2.4940308988764044, + "grad_norm": 0.8982266187667847, + "learning_rate": 3.1806254857483537e-06, + "loss": 0.5731, + "step": 7103 + }, + { + "epoch": 2.49438202247191, + "grad_norm": 0.9218209385871887, + "learning_rate": 3.1801806124904054e-06, + "loss": 0.5823, + "step": 7104 + }, + { + "epoch": 2.4947331460674156, + "grad_norm": 0.9041141867637634, + "learning_rate": 3.179735715970863e-06, + "loss": 0.5606, + "step": 7105 + }, + { + "epoch": 2.495084269662921, + "grad_norm": 0.8705416321754456, + "learning_rate": 3.1792907962049403e-06, + "loss": 0.5493, + "step": 7106 + }, + { + "epoch": 2.495435393258427, + "grad_norm": 0.9235622882843018, + "learning_rate": 3.1788458532078536e-06, + "loss": 0.6079, + "step": 7107 + }, + { + "epoch": 2.495786516853933, + "grad_norm": 0.8958116769790649, + "learning_rate": 3.1784008869948203e-06, + "loss": 0.6085, + "step": 7108 + }, + { + "epoch": 2.4961376404494384, + "grad_norm": 0.9100772142410278, + "learning_rate": 3.1779558975810565e-06, + "loss": 0.6228, + "step": 7109 + }, + { + "epoch": 2.496488764044944, + "grad_norm": 0.8802205324172974, + "learning_rate": 3.1775108849817817e-06, + "loss": 0.557, + "step": 7110 + }, + { + "epoch": 2.4968398876404496, + "grad_norm": 0.9105343818664551, + "learning_rate": 3.177065849212214e-06, + "loss": 0.5874, + "step": 7111 + }, + { + "epoch": 2.497191011235955, + "grad_norm": 0.887204110622406, + "learning_rate": 3.1766207902875732e-06, + "loss": 0.6083, + "step": 7112 + }, + { + "epoch": 2.497542134831461, + "grad_norm": 0.9102429747581482, + "learning_rate": 3.176175708223081e-06, + "loss": 0.5937, + "step": 7113 + }, + { + "epoch": 2.4978932584269664, + "grad_norm": 0.9321662783622742, + "learning_rate": 3.175730603033958e-06, + "loss": 0.5732, + "step": 7114 + }, + { + "epoch": 2.498244382022472, + "grad_norm": 0.9148409366607666, + "learning_rate": 3.1752854747354266e-06, + "loss": 0.5473, + "step": 7115 + }, + { + "epoch": 2.4985955056179776, + "grad_norm": 0.8857855200767517, + "learning_rate": 3.17484032334271e-06, + "loss": 0.579, + "step": 7116 + }, + { + "epoch": 2.498946629213483, + "grad_norm": 0.8620510101318359, + "learning_rate": 3.1743951488710317e-06, + "loss": 0.5698, + "step": 7117 + }, + { + "epoch": 2.499297752808989, + "grad_norm": 0.9007819294929504, + "learning_rate": 3.1739499513356163e-06, + "loss": 0.5786, + "step": 7118 + }, + { + "epoch": 2.4996488764044944, + "grad_norm": 0.8998503088951111, + "learning_rate": 3.173504730751689e-06, + "loss": 0.5755, + "step": 7119 + }, + { + "epoch": 2.5, + "grad_norm": 0.9349919557571411, + "learning_rate": 3.1730594871344765e-06, + "loss": 0.581, + "step": 7120 + }, + { + "epoch": 2.5003511235955056, + "grad_norm": 0.9016222953796387, + "learning_rate": 3.1726142204992055e-06, + "loss": 0.5595, + "step": 7121 + }, + { + "epoch": 2.500702247191011, + "grad_norm": 0.929345428943634, + "learning_rate": 3.1721689308611035e-06, + "loss": 0.5692, + "step": 7122 + }, + { + "epoch": 2.501053370786517, + "grad_norm": 0.9004616141319275, + "learning_rate": 3.1717236182353993e-06, + "loss": 0.5581, + "step": 7123 + }, + { + "epoch": 2.5014044943820224, + "grad_norm": 0.9339146614074707, + "learning_rate": 3.1712782826373226e-06, + "loss": 0.605, + "step": 7124 + }, + { + "epoch": 2.501755617977528, + "grad_norm": 0.9844611883163452, + "learning_rate": 3.1708329240821027e-06, + "loss": 0.5426, + "step": 7125 + }, + { + "epoch": 2.5021067415730336, + "grad_norm": 0.8789356350898743, + "learning_rate": 3.1703875425849716e-06, + "loss": 0.5797, + "step": 7126 + }, + { + "epoch": 2.502457865168539, + "grad_norm": 0.8834618330001831, + "learning_rate": 3.1699421381611596e-06, + "loss": 0.5492, + "step": 7127 + }, + { + "epoch": 2.502808988764045, + "grad_norm": 0.913087010383606, + "learning_rate": 3.1694967108258995e-06, + "loss": 0.578, + "step": 7128 + }, + { + "epoch": 2.5031601123595504, + "grad_norm": 0.8942289352416992, + "learning_rate": 3.1690512605944245e-06, + "loss": 0.5769, + "step": 7129 + }, + { + "epoch": 2.503511235955056, + "grad_norm": 0.8703804612159729, + "learning_rate": 3.1686057874819698e-06, + "loss": 0.5836, + "step": 7130 + }, + { + "epoch": 2.5038623595505616, + "grad_norm": 0.8990670442581177, + "learning_rate": 3.168160291503769e-06, + "loss": 0.6138, + "step": 7131 + }, + { + "epoch": 2.504213483146067, + "grad_norm": 0.8880472183227539, + "learning_rate": 3.1677147726750577e-06, + "loss": 0.541, + "step": 7132 + }, + { + "epoch": 2.5045646067415728, + "grad_norm": 0.9476865530014038, + "learning_rate": 3.1672692310110724e-06, + "loss": 0.6115, + "step": 7133 + }, + { + "epoch": 2.504915730337079, + "grad_norm": 0.9040087461471558, + "learning_rate": 3.1668236665270506e-06, + "loss": 0.5489, + "step": 7134 + }, + { + "epoch": 2.5052668539325844, + "grad_norm": 0.8872163891792297, + "learning_rate": 3.1663780792382302e-06, + "loss": 0.5705, + "step": 7135 + }, + { + "epoch": 2.50561797752809, + "grad_norm": 0.9129857420921326, + "learning_rate": 3.16593246915985e-06, + "loss": 0.5927, + "step": 7136 + }, + { + "epoch": 2.5059691011235956, + "grad_norm": 0.8994055986404419, + "learning_rate": 3.1654868363071482e-06, + "loss": 0.5906, + "step": 7137 + }, + { + "epoch": 2.506320224719101, + "grad_norm": 0.8948174118995667, + "learning_rate": 3.1650411806953667e-06, + "loss": 0.5555, + "step": 7138 + }, + { + "epoch": 2.506671348314607, + "grad_norm": 0.9184172749519348, + "learning_rate": 3.1645955023397468e-06, + "loss": 0.5872, + "step": 7139 + }, + { + "epoch": 2.5070224719101124, + "grad_norm": 0.9121282696723938, + "learning_rate": 3.1641498012555283e-06, + "loss": 0.5947, + "step": 7140 + }, + { + "epoch": 2.507373595505618, + "grad_norm": 0.9155190587043762, + "learning_rate": 3.1637040774579553e-06, + "loss": 0.619, + "step": 7141 + }, + { + "epoch": 2.5077247191011236, + "grad_norm": 0.8686392903327942, + "learning_rate": 3.1632583309622698e-06, + "loss": 0.5813, + "step": 7142 + }, + { + "epoch": 2.508075842696629, + "grad_norm": 0.886168897151947, + "learning_rate": 3.162812561783718e-06, + "loss": 0.5734, + "step": 7143 + }, + { + "epoch": 2.508426966292135, + "grad_norm": 0.8962077498435974, + "learning_rate": 3.162366769937544e-06, + "loss": 0.5783, + "step": 7144 + }, + { + "epoch": 2.5087780898876404, + "grad_norm": 0.8725407719612122, + "learning_rate": 3.1619209554389922e-06, + "loss": 0.5952, + "step": 7145 + }, + { + "epoch": 2.509129213483146, + "grad_norm": 0.8892403244972229, + "learning_rate": 3.161475118303311e-06, + "loss": 0.5699, + "step": 7146 + }, + { + "epoch": 2.5094803370786516, + "grad_norm": 0.8870375752449036, + "learning_rate": 3.1610292585457462e-06, + "loss": 0.5939, + "step": 7147 + }, + { + "epoch": 2.509831460674157, + "grad_norm": 0.9070790410041809, + "learning_rate": 3.160583376181547e-06, + "loss": 0.5964, + "step": 7148 + }, + { + "epoch": 2.510182584269663, + "grad_norm": 0.890164315700531, + "learning_rate": 3.1601374712259614e-06, + "loss": 0.5675, + "step": 7149 + }, + { + "epoch": 2.5105337078651684, + "grad_norm": 0.8721410036087036, + "learning_rate": 3.1596915436942387e-06, + "loss": 0.5584, + "step": 7150 + }, + { + "epoch": 2.5108848314606744, + "grad_norm": 0.9081694483757019, + "learning_rate": 3.1592455936016296e-06, + "loss": 0.5705, + "step": 7151 + }, + { + "epoch": 2.51123595505618, + "grad_norm": 0.9280213713645935, + "learning_rate": 3.158799620963387e-06, + "loss": 0.6025, + "step": 7152 + }, + { + "epoch": 2.5115870786516856, + "grad_norm": 0.9049160480499268, + "learning_rate": 3.15835362579476e-06, + "loss": 0.5733, + "step": 7153 + }, + { + "epoch": 2.511938202247191, + "grad_norm": 0.9163095355033875, + "learning_rate": 3.1579076081110015e-06, + "loss": 0.5651, + "step": 7154 + }, + { + "epoch": 2.512289325842697, + "grad_norm": 0.9008212685585022, + "learning_rate": 3.1574615679273667e-06, + "loss": 0.6186, + "step": 7155 + }, + { + "epoch": 2.5126404494382024, + "grad_norm": 0.8943555355072021, + "learning_rate": 3.157015505259109e-06, + "loss": 0.5569, + "step": 7156 + }, + { + "epoch": 2.512991573033708, + "grad_norm": 0.9180866479873657, + "learning_rate": 3.156569420121483e-06, + "loss": 0.6358, + "step": 7157 + }, + { + "epoch": 2.5133426966292136, + "grad_norm": 0.8881757259368896, + "learning_rate": 3.1561233125297446e-06, + "loss": 0.6164, + "step": 7158 + }, + { + "epoch": 2.513693820224719, + "grad_norm": 0.9002585411071777, + "learning_rate": 3.1556771824991507e-06, + "loss": 0.5935, + "step": 7159 + }, + { + "epoch": 2.514044943820225, + "grad_norm": 0.9107791185379028, + "learning_rate": 3.1552310300449575e-06, + "loss": 0.6153, + "step": 7160 + }, + { + "epoch": 2.5143960674157304, + "grad_norm": 0.9197893142700195, + "learning_rate": 3.1547848551824247e-06, + "loss": 0.6097, + "step": 7161 + }, + { + "epoch": 2.514747191011236, + "grad_norm": 0.8979735970497131, + "learning_rate": 3.1543386579268094e-06, + "loss": 0.5876, + "step": 7162 + }, + { + "epoch": 2.5150983146067416, + "grad_norm": 0.906789243221283, + "learning_rate": 3.153892438293372e-06, + "loss": 0.5812, + "step": 7163 + }, + { + "epoch": 2.515449438202247, + "grad_norm": 0.8778041005134583, + "learning_rate": 3.1534461962973734e-06, + "loss": 0.5426, + "step": 7164 + }, + { + "epoch": 2.515800561797753, + "grad_norm": 0.9567428231239319, + "learning_rate": 3.1529999319540733e-06, + "loss": 0.5878, + "step": 7165 + }, + { + "epoch": 2.5161516853932584, + "grad_norm": 0.9290637373924255, + "learning_rate": 3.152553645278735e-06, + "loss": 0.6169, + "step": 7166 + }, + { + "epoch": 2.516502808988764, + "grad_norm": 0.8982633948326111, + "learning_rate": 3.1521073362866196e-06, + "loss": 0.5776, + "step": 7167 + }, + { + "epoch": 2.5168539325842696, + "grad_norm": 0.9090446829795837, + "learning_rate": 3.151661004992992e-06, + "loss": 0.5698, + "step": 7168 + }, + { + "epoch": 2.517205056179775, + "grad_norm": 0.8945769667625427, + "learning_rate": 3.1512146514131152e-06, + "loss": 0.568, + "step": 7169 + }, + { + "epoch": 2.517556179775281, + "grad_norm": 0.8889007568359375, + "learning_rate": 3.1507682755622547e-06, + "loss": 0.6075, + "step": 7170 + }, + { + "epoch": 2.5179073033707864, + "grad_norm": 0.9171195030212402, + "learning_rate": 3.150321877455676e-06, + "loss": 0.6038, + "step": 7171 + }, + { + "epoch": 2.518258426966292, + "grad_norm": 0.8888158202171326, + "learning_rate": 3.149875457108646e-06, + "loss": 0.5807, + "step": 7172 + }, + { + "epoch": 2.5186095505617976, + "grad_norm": 0.9068456888198853, + "learning_rate": 3.149429014536431e-06, + "loss": 0.5655, + "step": 7173 + }, + { + "epoch": 2.518960674157303, + "grad_norm": 0.9322959780693054, + "learning_rate": 3.1489825497542993e-06, + "loss": 0.5833, + "step": 7174 + }, + { + "epoch": 2.519311797752809, + "grad_norm": 0.9341587424278259, + "learning_rate": 3.14853606277752e-06, + "loss": 0.6281, + "step": 7175 + }, + { + "epoch": 2.5196629213483144, + "grad_norm": 0.9163209199905396, + "learning_rate": 3.1480895536213623e-06, + "loss": 0.6102, + "step": 7176 + }, + { + "epoch": 2.52001404494382, + "grad_norm": 0.9000644087791443, + "learning_rate": 3.1476430223010975e-06, + "loss": 0.6033, + "step": 7177 + }, + { + "epoch": 2.5203651685393256, + "grad_norm": 0.8934353590011597, + "learning_rate": 3.1471964688319943e-06, + "loss": 0.5461, + "step": 7178 + }, + { + "epoch": 2.5207162921348316, + "grad_norm": 0.9028684496879578, + "learning_rate": 3.146749893229326e-06, + "loss": 0.5782, + "step": 7179 + }, + { + "epoch": 2.521067415730337, + "grad_norm": 0.9014409780502319, + "learning_rate": 3.1463032955083642e-06, + "loss": 0.5735, + "step": 7180 + }, + { + "epoch": 2.521418539325843, + "grad_norm": 0.9010222554206848, + "learning_rate": 3.1458566756843834e-06, + "loss": 0.5835, + "step": 7181 + }, + { + "epoch": 2.5217696629213484, + "grad_norm": 0.9944157004356384, + "learning_rate": 3.145410033772657e-06, + "loss": 0.5937, + "step": 7182 + }, + { + "epoch": 2.522120786516854, + "grad_norm": 0.8764009475708008, + "learning_rate": 3.1449633697884597e-06, + "loss": 0.5632, + "step": 7183 + }, + { + "epoch": 2.5224719101123596, + "grad_norm": 0.8874295949935913, + "learning_rate": 3.1445166837470663e-06, + "loss": 0.5844, + "step": 7184 + }, + { + "epoch": 2.522823033707865, + "grad_norm": 0.8621526956558228, + "learning_rate": 3.144069975663755e-06, + "loss": 0.552, + "step": 7185 + }, + { + "epoch": 2.523174157303371, + "grad_norm": 0.925792396068573, + "learning_rate": 3.1436232455538005e-06, + "loss": 0.5589, + "step": 7186 + }, + { + "epoch": 2.5235252808988764, + "grad_norm": 0.9157830476760864, + "learning_rate": 3.143176493432483e-06, + "loss": 0.5926, + "step": 7187 + }, + { + "epoch": 2.523876404494382, + "grad_norm": 0.9245595932006836, + "learning_rate": 3.1427297193150792e-06, + "loss": 0.5887, + "step": 7188 + }, + { + "epoch": 2.5242275280898876, + "grad_norm": 0.8830627799034119, + "learning_rate": 3.1422829232168688e-06, + "loss": 0.5706, + "step": 7189 + }, + { + "epoch": 2.524578651685393, + "grad_norm": 0.9089692234992981, + "learning_rate": 3.1418361051531323e-06, + "loss": 0.5788, + "step": 7190 + }, + { + "epoch": 2.524929775280899, + "grad_norm": 0.8836162090301514, + "learning_rate": 3.1413892651391498e-06, + "loss": 0.5572, + "step": 7191 + }, + { + "epoch": 2.5252808988764044, + "grad_norm": 0.8822372555732727, + "learning_rate": 3.1409424031902034e-06, + "loss": 0.5717, + "step": 7192 + }, + { + "epoch": 2.52563202247191, + "grad_norm": 0.8716445565223694, + "learning_rate": 3.140495519321576e-06, + "loss": 0.5598, + "step": 7193 + }, + { + "epoch": 2.5259831460674156, + "grad_norm": 0.9181852340698242, + "learning_rate": 3.140048613548549e-06, + "loss": 0.6122, + "step": 7194 + }, + { + "epoch": 2.526334269662921, + "grad_norm": 0.8861651420593262, + "learning_rate": 3.139601685886408e-06, + "loss": 0.559, + "step": 7195 + }, + { + "epoch": 2.5266853932584272, + "grad_norm": 0.8966087102890015, + "learning_rate": 3.139154736350436e-06, + "loss": 0.5739, + "step": 7196 + }, + { + "epoch": 2.527036516853933, + "grad_norm": 0.8727557063102722, + "learning_rate": 3.1387077649559194e-06, + "loss": 0.6033, + "step": 7197 + }, + { + "epoch": 2.5273876404494384, + "grad_norm": 0.8684516549110413, + "learning_rate": 3.1382607717181435e-06, + "loss": 0.579, + "step": 7198 + }, + { + "epoch": 2.527738764044944, + "grad_norm": 0.9035211205482483, + "learning_rate": 3.137813756652395e-06, + "loss": 0.5869, + "step": 7199 + }, + { + "epoch": 2.5280898876404496, + "grad_norm": 0.914918065071106, + "learning_rate": 3.1373667197739623e-06, + "loss": 0.6134, + "step": 7200 + }, + { + "epoch": 2.528441011235955, + "grad_norm": 0.9170656800270081, + "learning_rate": 3.136919661098133e-06, + "loss": 0.6316, + "step": 7201 + }, + { + "epoch": 2.528792134831461, + "grad_norm": 0.9072597622871399, + "learning_rate": 3.136472580640197e-06, + "loss": 0.6056, + "step": 7202 + }, + { + "epoch": 2.5291432584269664, + "grad_norm": 0.8989172577857971, + "learning_rate": 3.1360254784154426e-06, + "loss": 0.6005, + "step": 7203 + }, + { + "epoch": 2.529494382022472, + "grad_norm": 0.9164608120918274, + "learning_rate": 3.135578354439161e-06, + "loss": 0.6286, + "step": 7204 + }, + { + "epoch": 2.5298455056179776, + "grad_norm": 0.8916328549385071, + "learning_rate": 3.1351312087266437e-06, + "loss": 0.5856, + "step": 7205 + }, + { + "epoch": 2.530196629213483, + "grad_norm": 0.9085004329681396, + "learning_rate": 3.134684041293183e-06, + "loss": 0.5609, + "step": 7206 + }, + { + "epoch": 2.530547752808989, + "grad_norm": 0.9137087464332581, + "learning_rate": 3.1342368521540705e-06, + "loss": 0.5908, + "step": 7207 + }, + { + "epoch": 2.5308988764044944, + "grad_norm": 0.8761213421821594, + "learning_rate": 3.1337896413246004e-06, + "loss": 0.6106, + "step": 7208 + }, + { + "epoch": 2.53125, + "grad_norm": 0.8962277173995972, + "learning_rate": 3.1333424088200675e-06, + "loss": 0.5708, + "step": 7209 + }, + { + "epoch": 2.5316011235955056, + "grad_norm": 0.8867594003677368, + "learning_rate": 3.1328951546557662e-06, + "loss": 0.5689, + "step": 7210 + }, + { + "epoch": 2.531952247191011, + "grad_norm": 0.8816742300987244, + "learning_rate": 3.1324478788469915e-06, + "loss": 0.5788, + "step": 7211 + }, + { + "epoch": 2.532303370786517, + "grad_norm": 0.914510190486908, + "learning_rate": 3.132000581409041e-06, + "loss": 0.5882, + "step": 7212 + }, + { + "epoch": 2.5326544943820224, + "grad_norm": 0.9041720628738403, + "learning_rate": 3.1315532623572115e-06, + "loss": 0.561, + "step": 7213 + }, + { + "epoch": 2.533005617977528, + "grad_norm": 0.9328110814094543, + "learning_rate": 3.131105921706801e-06, + "loss": 0.6021, + "step": 7214 + }, + { + "epoch": 2.5333567415730336, + "grad_norm": 0.9069661498069763, + "learning_rate": 3.1306585594731078e-06, + "loss": 0.613, + "step": 7215 + }, + { + "epoch": 2.533707865168539, + "grad_norm": 0.9356861114501953, + "learning_rate": 3.1302111756714317e-06, + "loss": 0.5867, + "step": 7216 + }, + { + "epoch": 2.534058988764045, + "grad_norm": 0.9017901420593262, + "learning_rate": 3.1297637703170717e-06, + "loss": 0.5891, + "step": 7217 + }, + { + "epoch": 2.5344101123595504, + "grad_norm": 0.915593147277832, + "learning_rate": 3.1293163434253303e-06, + "loss": 0.5704, + "step": 7218 + }, + { + "epoch": 2.534761235955056, + "grad_norm": 0.888866662979126, + "learning_rate": 3.128868895011509e-06, + "loss": 0.5744, + "step": 7219 + }, + { + "epoch": 2.5351123595505616, + "grad_norm": 0.8909019827842712, + "learning_rate": 3.1284214250909087e-06, + "loss": 0.5632, + "step": 7220 + }, + { + "epoch": 2.535463483146067, + "grad_norm": 0.921874463558197, + "learning_rate": 3.127973933678834e-06, + "loss": 0.6144, + "step": 7221 + }, + { + "epoch": 2.5358146067415728, + "grad_norm": 0.9166398048400879, + "learning_rate": 3.127526420790588e-06, + "loss": 0.5805, + "step": 7222 + }, + { + "epoch": 2.536165730337079, + "grad_norm": 0.9245310425758362, + "learning_rate": 3.1270788864414756e-06, + "loss": 0.5945, + "step": 7223 + }, + { + "epoch": 2.5365168539325844, + "grad_norm": 0.8960409164428711, + "learning_rate": 3.1266313306468018e-06, + "loss": 0.5651, + "step": 7224 + }, + { + "epoch": 2.53686797752809, + "grad_norm": 0.9255216121673584, + "learning_rate": 3.1261837534218724e-06, + "loss": 0.5848, + "step": 7225 + }, + { + "epoch": 2.5372191011235956, + "grad_norm": 0.8838738203048706, + "learning_rate": 3.125736154781995e-06, + "loss": 0.5778, + "step": 7226 + }, + { + "epoch": 2.537570224719101, + "grad_norm": 0.9194349646568298, + "learning_rate": 3.1252885347424766e-06, + "loss": 0.5995, + "step": 7227 + }, + { + "epoch": 2.537921348314607, + "grad_norm": 0.8944032788276672, + "learning_rate": 3.1248408933186247e-06, + "loss": 0.5653, + "step": 7228 + }, + { + "epoch": 2.5382724719101124, + "grad_norm": 0.899829089641571, + "learning_rate": 3.124393230525749e-06, + "loss": 0.5709, + "step": 7229 + }, + { + "epoch": 2.538623595505618, + "grad_norm": 0.8916481733322144, + "learning_rate": 3.1239455463791596e-06, + "loss": 0.6256, + "step": 7230 + }, + { + "epoch": 2.5389747191011236, + "grad_norm": 0.9506766200065613, + "learning_rate": 3.123497840894166e-06, + "loss": 0.6282, + "step": 7231 + }, + { + "epoch": 2.539325842696629, + "grad_norm": 0.8796055912971497, + "learning_rate": 3.1230501140860805e-06, + "loss": 0.5585, + "step": 7232 + }, + { + "epoch": 2.539676966292135, + "grad_norm": 0.8823270797729492, + "learning_rate": 3.122602365970213e-06, + "loss": 0.5741, + "step": 7233 + }, + { + "epoch": 2.5400280898876404, + "grad_norm": 0.9207660555839539, + "learning_rate": 3.1221545965618784e-06, + "loss": 0.5852, + "step": 7234 + }, + { + "epoch": 2.540379213483146, + "grad_norm": 0.8710954189300537, + "learning_rate": 3.1217068058763883e-06, + "loss": 0.5724, + "step": 7235 + }, + { + "epoch": 2.5407303370786516, + "grad_norm": 0.88435298204422, + "learning_rate": 3.121258993929058e-06, + "loss": 0.5945, + "step": 7236 + }, + { + "epoch": 2.541081460674157, + "grad_norm": 0.8989911079406738, + "learning_rate": 3.1208111607352014e-06, + "loss": 0.5511, + "step": 7237 + }, + { + "epoch": 2.541432584269663, + "grad_norm": 0.8742223978042603, + "learning_rate": 3.1203633063101347e-06, + "loss": 0.5888, + "step": 7238 + }, + { + "epoch": 2.5417837078651684, + "grad_norm": 0.9363758563995361, + "learning_rate": 3.1199154306691732e-06, + "loss": 0.5923, + "step": 7239 + }, + { + "epoch": 2.5421348314606744, + "grad_norm": 0.9339737296104431, + "learning_rate": 3.1194675338276353e-06, + "loss": 0.6101, + "step": 7240 + }, + { + "epoch": 2.54248595505618, + "grad_norm": 0.9465301632881165, + "learning_rate": 3.1190196158008373e-06, + "loss": 0.563, + "step": 7241 + }, + { + "epoch": 2.5428370786516856, + "grad_norm": 0.8891629576683044, + "learning_rate": 3.118571676604098e-06, + "loss": 0.5644, + "step": 7242 + }, + { + "epoch": 2.543188202247191, + "grad_norm": 0.899986743927002, + "learning_rate": 3.1181237162527367e-06, + "loss": 0.5974, + "step": 7243 + }, + { + "epoch": 2.543539325842697, + "grad_norm": 0.8943713903427124, + "learning_rate": 3.1176757347620735e-06, + "loss": 0.6108, + "step": 7244 + }, + { + "epoch": 2.5438904494382024, + "grad_norm": 0.8701123595237732, + "learning_rate": 3.117227732147429e-06, + "loss": 0.5821, + "step": 7245 + }, + { + "epoch": 2.544241573033708, + "grad_norm": 0.9077134728431702, + "learning_rate": 3.116779708424123e-06, + "loss": 0.568, + "step": 7246 + }, + { + "epoch": 2.5445926966292136, + "grad_norm": 0.9160280823707581, + "learning_rate": 3.1163316636074796e-06, + "loss": 0.5801, + "step": 7247 + }, + { + "epoch": 2.544943820224719, + "grad_norm": 0.9029747843742371, + "learning_rate": 3.115883597712821e-06, + "loss": 0.5965, + "step": 7248 + }, + { + "epoch": 2.545294943820225, + "grad_norm": 0.9012978076934814, + "learning_rate": 3.1154355107554697e-06, + "loss": 0.5539, + "step": 7249 + }, + { + "epoch": 2.5456460674157304, + "grad_norm": 0.8863149285316467, + "learning_rate": 3.1149874027507508e-06, + "loss": 0.5659, + "step": 7250 + }, + { + "epoch": 2.545997191011236, + "grad_norm": 0.8892086148262024, + "learning_rate": 3.114539273713989e-06, + "loss": 0.5574, + "step": 7251 + }, + { + "epoch": 2.5463483146067416, + "grad_norm": 0.9138109683990479, + "learning_rate": 3.114091123660511e-06, + "loss": 0.5875, + "step": 7252 + }, + { + "epoch": 2.546699438202247, + "grad_norm": 0.8755885362625122, + "learning_rate": 3.1136429526056405e-06, + "loss": 0.5699, + "step": 7253 + }, + { + "epoch": 2.547050561797753, + "grad_norm": 0.8694639205932617, + "learning_rate": 3.113194760564707e-06, + "loss": 0.5547, + "step": 7254 + }, + { + "epoch": 2.5474016853932584, + "grad_norm": 0.9114827513694763, + "learning_rate": 3.1127465475530374e-06, + "loss": 0.5555, + "step": 7255 + }, + { + "epoch": 2.547752808988764, + "grad_norm": 0.9505554437637329, + "learning_rate": 3.1122983135859596e-06, + "loss": 0.5786, + "step": 7256 + }, + { + "epoch": 2.5481039325842696, + "grad_norm": 0.8955578207969666, + "learning_rate": 3.1118500586788034e-06, + "loss": 0.5751, + "step": 7257 + }, + { + "epoch": 2.548455056179775, + "grad_norm": 0.9145152568817139, + "learning_rate": 3.111401782846899e-06, + "loss": 0.6036, + "step": 7258 + }, + { + "epoch": 2.548806179775281, + "grad_norm": 0.9128950238227844, + "learning_rate": 3.1109534861055767e-06, + "loss": 0.5701, + "step": 7259 + }, + { + "epoch": 2.5491573033707864, + "grad_norm": 0.9094752669334412, + "learning_rate": 3.1105051684701687e-06, + "loss": 0.563, + "step": 7260 + }, + { + "epoch": 2.549508426966292, + "grad_norm": 0.9190934896469116, + "learning_rate": 3.110056829956006e-06, + "loss": 0.5845, + "step": 7261 + }, + { + "epoch": 2.5498595505617976, + "grad_norm": 0.8836715221405029, + "learning_rate": 3.1096084705784213e-06, + "loss": 0.5404, + "step": 7262 + }, + { + "epoch": 2.550210674157303, + "grad_norm": 0.9002956748008728, + "learning_rate": 3.1091600903527487e-06, + "loss": 0.5921, + "step": 7263 + }, + { + "epoch": 2.550561797752809, + "grad_norm": 0.9009602069854736, + "learning_rate": 3.1087116892943227e-06, + "loss": 0.5608, + "step": 7264 + }, + { + "epoch": 2.5509129213483144, + "grad_norm": 0.92513108253479, + "learning_rate": 3.1082632674184777e-06, + "loss": 0.5912, + "step": 7265 + }, + { + "epoch": 2.55126404494382, + "grad_norm": 0.9085976481437683, + "learning_rate": 3.107814824740549e-06, + "loss": 0.5666, + "step": 7266 + }, + { + "epoch": 2.5516151685393256, + "grad_norm": 0.8912241458892822, + "learning_rate": 3.1073663612758735e-06, + "loss": 0.5539, + "step": 7267 + }, + { + "epoch": 2.5519662921348316, + "grad_norm": 0.9152499437332153, + "learning_rate": 3.1069178770397883e-06, + "loss": 0.6109, + "step": 7268 + }, + { + "epoch": 2.552317415730337, + "grad_norm": 0.9883180856704712, + "learning_rate": 3.1064693720476313e-06, + "loss": 0.6003, + "step": 7269 + }, + { + "epoch": 2.552668539325843, + "grad_norm": 0.8970620036125183, + "learning_rate": 3.1060208463147405e-06, + "loss": 0.5708, + "step": 7270 + }, + { + "epoch": 2.5530196629213484, + "grad_norm": 0.9245628714561462, + "learning_rate": 3.1055722998564553e-06, + "loss": 0.6338, + "step": 7271 + }, + { + "epoch": 2.553370786516854, + "grad_norm": 0.8905019164085388, + "learning_rate": 3.1051237326881154e-06, + "loss": 0.5723, + "step": 7272 + }, + { + "epoch": 2.5537219101123596, + "grad_norm": 0.9124685525894165, + "learning_rate": 3.1046751448250623e-06, + "loss": 0.5758, + "step": 7273 + }, + { + "epoch": 2.554073033707865, + "grad_norm": 0.9105271100997925, + "learning_rate": 3.1042265362826367e-06, + "loss": 0.5864, + "step": 7274 + }, + { + "epoch": 2.554424157303371, + "grad_norm": 0.9125889539718628, + "learning_rate": 3.1037779070761803e-06, + "loss": 0.5828, + "step": 7275 + }, + { + "epoch": 2.5547752808988764, + "grad_norm": 0.8931271433830261, + "learning_rate": 3.103329257221036e-06, + "loss": 0.6119, + "step": 7276 + }, + { + "epoch": 2.555126404494382, + "grad_norm": 0.9100943803787231, + "learning_rate": 3.102880586732547e-06, + "loss": 0.6027, + "step": 7277 + }, + { + "epoch": 2.5554775280898876, + "grad_norm": 0.8700313568115234, + "learning_rate": 3.102431895626059e-06, + "loss": 0.5528, + "step": 7278 + }, + { + "epoch": 2.555828651685393, + "grad_norm": 0.9101304411888123, + "learning_rate": 3.1019831839169146e-06, + "loss": 0.5488, + "step": 7279 + }, + { + "epoch": 2.556179775280899, + "grad_norm": 0.8850422501564026, + "learning_rate": 3.101534451620461e-06, + "loss": 0.5724, + "step": 7280 + }, + { + "epoch": 2.5565308988764044, + "grad_norm": 0.9123455882072449, + "learning_rate": 3.1010856987520444e-06, + "loss": 0.5803, + "step": 7281 + }, + { + "epoch": 2.55688202247191, + "grad_norm": 0.90606290102005, + "learning_rate": 3.1006369253270102e-06, + "loss": 0.5849, + "step": 7282 + }, + { + "epoch": 2.5572331460674156, + "grad_norm": 0.8990072011947632, + "learning_rate": 3.1001881313607075e-06, + "loss": 0.5557, + "step": 7283 + }, + { + "epoch": 2.557584269662921, + "grad_norm": 0.8796411156654358, + "learning_rate": 3.099739316868484e-06, + "loss": 0.5937, + "step": 7284 + }, + { + "epoch": 2.5579353932584272, + "grad_norm": 0.9014245271682739, + "learning_rate": 3.09929048186569e-06, + "loss": 0.5772, + "step": 7285 + }, + { + "epoch": 2.558286516853933, + "grad_norm": 0.947607696056366, + "learning_rate": 3.0988416263676736e-06, + "loss": 0.5928, + "step": 7286 + }, + { + "epoch": 2.5586376404494384, + "grad_norm": 0.9210224747657776, + "learning_rate": 3.098392750389786e-06, + "loss": 0.563, + "step": 7287 + }, + { + "epoch": 2.558988764044944, + "grad_norm": 0.9200950264930725, + "learning_rate": 3.097943853947378e-06, + "loss": 0.5757, + "step": 7288 + }, + { + "epoch": 2.5593398876404496, + "grad_norm": 0.9156930446624756, + "learning_rate": 3.0974949370558027e-06, + "loss": 0.5756, + "step": 7289 + }, + { + "epoch": 2.559691011235955, + "grad_norm": 0.9226803779602051, + "learning_rate": 3.097045999730412e-06, + "loss": 0.589, + "step": 7290 + }, + { + "epoch": 2.560042134831461, + "grad_norm": 0.9272333383560181, + "learning_rate": 3.0965970419865582e-06, + "loss": 0.5555, + "step": 7291 + }, + { + "epoch": 2.5603932584269664, + "grad_norm": 0.8942030072212219, + "learning_rate": 3.0961480638395963e-06, + "loss": 0.5719, + "step": 7292 + }, + { + "epoch": 2.560744382022472, + "grad_norm": 0.9034997820854187, + "learning_rate": 3.0956990653048808e-06, + "loss": 0.604, + "step": 7293 + }, + { + "epoch": 2.5610955056179776, + "grad_norm": 0.8922563791275024, + "learning_rate": 3.0952500463977668e-06, + "loss": 0.5967, + "step": 7294 + }, + { + "epoch": 2.561446629213483, + "grad_norm": 0.9301415085792542, + "learning_rate": 3.094801007133611e-06, + "loss": 0.5841, + "step": 7295 + }, + { + "epoch": 2.561797752808989, + "grad_norm": 0.8967776298522949, + "learning_rate": 3.0943519475277695e-06, + "loss": 0.5588, + "step": 7296 + }, + { + "epoch": 2.5621488764044944, + "grad_norm": 0.9253349304199219, + "learning_rate": 3.0939028675956e-06, + "loss": 0.5847, + "step": 7297 + }, + { + "epoch": 2.5625, + "grad_norm": 0.9434443116188049, + "learning_rate": 3.093453767352461e-06, + "loss": 0.5997, + "step": 7298 + }, + { + "epoch": 2.5628511235955056, + "grad_norm": 0.9002946615219116, + "learning_rate": 3.0930046468137102e-06, + "loss": 0.5727, + "step": 7299 + }, + { + "epoch": 2.563202247191011, + "grad_norm": 0.9191354513168335, + "learning_rate": 3.0925555059947086e-06, + "loss": 0.6051, + "step": 7300 + }, + { + "epoch": 2.563553370786517, + "grad_norm": 0.9061764478683472, + "learning_rate": 3.092106344910816e-06, + "loss": 0.5874, + "step": 7301 + }, + { + "epoch": 2.5639044943820224, + "grad_norm": 0.9085385203361511, + "learning_rate": 3.0916571635773924e-06, + "loss": 0.5684, + "step": 7302 + }, + { + "epoch": 2.564255617977528, + "grad_norm": 0.8983359336853027, + "learning_rate": 3.0912079620098006e-06, + "loss": 0.5394, + "step": 7303 + }, + { + "epoch": 2.5646067415730336, + "grad_norm": 0.9051839709281921, + "learning_rate": 3.090758740223402e-06, + "loss": 0.5811, + "step": 7304 + }, + { + "epoch": 2.564957865168539, + "grad_norm": 0.9062090516090393, + "learning_rate": 3.0903094982335603e-06, + "loss": 0.5551, + "step": 7305 + }, + { + "epoch": 2.565308988764045, + "grad_norm": 0.9072140455245972, + "learning_rate": 3.0898602360556386e-06, + "loss": 0.6054, + "step": 7306 + }, + { + "epoch": 2.5656601123595504, + "grad_norm": 0.8808127045631409, + "learning_rate": 3.0894109537050014e-06, + "loss": 0.5771, + "step": 7307 + }, + { + "epoch": 2.566011235955056, + "grad_norm": 0.8979044556617737, + "learning_rate": 3.0889616511970143e-06, + "loss": 0.5981, + "step": 7308 + }, + { + "epoch": 2.5663623595505616, + "grad_norm": 0.8869459629058838, + "learning_rate": 3.0885123285470427e-06, + "loss": 0.5883, + "step": 7309 + }, + { + "epoch": 2.566713483146067, + "grad_norm": 0.8959521651268005, + "learning_rate": 3.088062985770453e-06, + "loss": 0.5753, + "step": 7310 + }, + { + "epoch": 2.5670646067415728, + "grad_norm": 0.8889307975769043, + "learning_rate": 3.087613622882612e-06, + "loss": 0.5468, + "step": 7311 + }, + { + "epoch": 2.567415730337079, + "grad_norm": 0.8885915279388428, + "learning_rate": 3.087164239898889e-06, + "loss": 0.5765, + "step": 7312 + }, + { + "epoch": 2.5677668539325844, + "grad_norm": 0.9297278523445129, + "learning_rate": 3.08671483683465e-06, + "loss": 0.5821, + "step": 7313 + }, + { + "epoch": 2.56811797752809, + "grad_norm": 0.8932828903198242, + "learning_rate": 3.0862654137052665e-06, + "loss": 0.5649, + "step": 7314 + }, + { + "epoch": 2.5684691011235956, + "grad_norm": 0.8938810229301453, + "learning_rate": 3.085815970526108e-06, + "loss": 0.5572, + "step": 7315 + }, + { + "epoch": 2.568820224719101, + "grad_norm": 0.9084933996200562, + "learning_rate": 3.085366507312544e-06, + "loss": 0.5801, + "step": 7316 + }, + { + "epoch": 2.569171348314607, + "grad_norm": 0.8684697151184082, + "learning_rate": 3.0849170240799466e-06, + "loss": 0.5584, + "step": 7317 + }, + { + "epoch": 2.5695224719101124, + "grad_norm": 0.8961203694343567, + "learning_rate": 3.084467520843687e-06, + "loss": 0.5425, + "step": 7318 + }, + { + "epoch": 2.569873595505618, + "grad_norm": 0.8650273680686951, + "learning_rate": 3.0840179976191387e-06, + "loss": 0.5545, + "step": 7319 + }, + { + "epoch": 2.5702247191011236, + "grad_norm": 0.9152853488922119, + "learning_rate": 3.0835684544216743e-06, + "loss": 0.6, + "step": 7320 + }, + { + "epoch": 2.570575842696629, + "grad_norm": 0.8937976956367493, + "learning_rate": 3.0831188912666686e-06, + "loss": 0.5845, + "step": 7321 + }, + { + "epoch": 2.570926966292135, + "grad_norm": 0.8949443697929382, + "learning_rate": 3.0826693081694953e-06, + "loss": 0.5905, + "step": 7322 + }, + { + "epoch": 2.5712780898876404, + "grad_norm": 0.8906983733177185, + "learning_rate": 3.0822197051455305e-06, + "loss": 0.5282, + "step": 7323 + }, + { + "epoch": 2.571629213483146, + "grad_norm": 0.9305698275566101, + "learning_rate": 3.0817700822101505e-06, + "loss": 0.5895, + "step": 7324 + }, + { + "epoch": 2.5719803370786516, + "grad_norm": 0.9006452560424805, + "learning_rate": 3.0813204393787306e-06, + "loss": 0.5538, + "step": 7325 + }, + { + "epoch": 2.572331460674157, + "grad_norm": 0.9058154225349426, + "learning_rate": 3.0808707766666495e-06, + "loss": 0.5953, + "step": 7326 + }, + { + "epoch": 2.572682584269663, + "grad_norm": 0.8668138384819031, + "learning_rate": 3.0804210940892847e-06, + "loss": 0.5387, + "step": 7327 + }, + { + "epoch": 2.5730337078651684, + "grad_norm": 0.901249349117279, + "learning_rate": 3.079971391662016e-06, + "loss": 0.5834, + "step": 7328 + }, + { + "epoch": 2.5733848314606744, + "grad_norm": 0.908290445804596, + "learning_rate": 3.079521669400222e-06, + "loss": 0.6047, + "step": 7329 + }, + { + "epoch": 2.57373595505618, + "grad_norm": 0.8919737935066223, + "learning_rate": 3.0790719273192817e-06, + "loss": 0.5863, + "step": 7330 + }, + { + "epoch": 2.5740870786516856, + "grad_norm": 0.9070049524307251, + "learning_rate": 3.0786221654345778e-06, + "loss": 0.5757, + "step": 7331 + }, + { + "epoch": 2.574438202247191, + "grad_norm": 0.8819604516029358, + "learning_rate": 3.07817238376149e-06, + "loss": 0.5442, + "step": 7332 + }, + { + "epoch": 2.574789325842697, + "grad_norm": 0.877232551574707, + "learning_rate": 3.077722582315402e-06, + "loss": 0.5772, + "step": 7333 + }, + { + "epoch": 2.5751404494382024, + "grad_norm": 0.9180396795272827, + "learning_rate": 3.0772727611116965e-06, + "loss": 0.6038, + "step": 7334 + }, + { + "epoch": 2.575491573033708, + "grad_norm": 0.9079856276512146, + "learning_rate": 3.0768229201657564e-06, + "loss": 0.5947, + "step": 7335 + }, + { + "epoch": 2.5758426966292136, + "grad_norm": 0.858569324016571, + "learning_rate": 3.0763730594929656e-06, + "loss": 0.5127, + "step": 7336 + }, + { + "epoch": 2.576193820224719, + "grad_norm": 0.907561182975769, + "learning_rate": 3.0759231791087094e-06, + "loss": 0.6108, + "step": 7337 + }, + { + "epoch": 2.576544943820225, + "grad_norm": 0.9146140813827515, + "learning_rate": 3.0754732790283737e-06, + "loss": 0.5952, + "step": 7338 + }, + { + "epoch": 2.5768960674157304, + "grad_norm": 0.8925809860229492, + "learning_rate": 3.075023359267344e-06, + "loss": 0.5587, + "step": 7339 + }, + { + "epoch": 2.577247191011236, + "grad_norm": 0.8726143836975098, + "learning_rate": 3.0745734198410076e-06, + "loss": 0.569, + "step": 7340 + }, + { + "epoch": 2.5775983146067416, + "grad_norm": 0.92153400182724, + "learning_rate": 3.0741234607647525e-06, + "loss": 0.6033, + "step": 7341 + }, + { + "epoch": 2.577949438202247, + "grad_norm": 0.8806565403938293, + "learning_rate": 3.0736734820539654e-06, + "loss": 0.5639, + "step": 7342 + }, + { + "epoch": 2.578300561797753, + "grad_norm": 0.8692107200622559, + "learning_rate": 3.0732234837240364e-06, + "loss": 0.5709, + "step": 7343 + }, + { + "epoch": 2.5786516853932584, + "grad_norm": 0.9363458752632141, + "learning_rate": 3.0727734657903552e-06, + "loss": 0.5877, + "step": 7344 + }, + { + "epoch": 2.579002808988764, + "grad_norm": 0.8658397197723389, + "learning_rate": 3.0723234282683113e-06, + "loss": 0.5235, + "step": 7345 + }, + { + "epoch": 2.5793539325842696, + "grad_norm": 0.904514729976654, + "learning_rate": 3.0718733711732956e-06, + "loss": 0.5672, + "step": 7346 + }, + { + "epoch": 2.579705056179775, + "grad_norm": 0.9416945576667786, + "learning_rate": 3.0714232945207e-06, + "loss": 0.6487, + "step": 7347 + }, + { + "epoch": 2.580056179775281, + "grad_norm": 0.8864945769309998, + "learning_rate": 3.0709731983259166e-06, + "loss": 0.5965, + "step": 7348 + }, + { + "epoch": 2.5804073033707864, + "grad_norm": 0.8871757984161377, + "learning_rate": 3.0705230826043387e-06, + "loss": 0.5792, + "step": 7349 + }, + { + "epoch": 2.580758426966292, + "grad_norm": 0.9047237634658813, + "learning_rate": 3.07007294737136e-06, + "loss": 0.6144, + "step": 7350 + }, + { + "epoch": 2.5811095505617976, + "grad_norm": 0.941625714302063, + "learning_rate": 3.069622792642373e-06, + "loss": 0.5864, + "step": 7351 + }, + { + "epoch": 2.581460674157303, + "grad_norm": 0.9283971190452576, + "learning_rate": 3.069172618432775e-06, + "loss": 0.5751, + "step": 7352 + }, + { + "epoch": 2.581811797752809, + "grad_norm": 0.8910781741142273, + "learning_rate": 3.068722424757961e-06, + "loss": 0.6138, + "step": 7353 + }, + { + "epoch": 2.5821629213483144, + "grad_norm": 0.9000306725502014, + "learning_rate": 3.0682722116333257e-06, + "loss": 0.5876, + "step": 7354 + }, + { + "epoch": 2.58251404494382, + "grad_norm": 0.8841826915740967, + "learning_rate": 3.0678219790742673e-06, + "loss": 0.6113, + "step": 7355 + }, + { + "epoch": 2.5828651685393256, + "grad_norm": 0.8769837617874146, + "learning_rate": 3.0673717270961832e-06, + "loss": 0.5732, + "step": 7356 + }, + { + "epoch": 2.5832162921348316, + "grad_norm": 0.9571608304977417, + "learning_rate": 3.0669214557144717e-06, + "loss": 0.5863, + "step": 7357 + }, + { + "epoch": 2.583567415730337, + "grad_norm": 0.8969611525535583, + "learning_rate": 3.0664711649445313e-06, + "loss": 0.5894, + "step": 7358 + }, + { + "epoch": 2.583918539325843, + "grad_norm": 0.8940547704696655, + "learning_rate": 3.0660208548017616e-06, + "loss": 0.589, + "step": 7359 + }, + { + "epoch": 2.5842696629213484, + "grad_norm": 0.8886762857437134, + "learning_rate": 3.0655705253015633e-06, + "loss": 0.5364, + "step": 7360 + }, + { + "epoch": 2.584620786516854, + "grad_norm": 0.9108870029449463, + "learning_rate": 3.0651201764593375e-06, + "loss": 0.5746, + "step": 7361 + }, + { + "epoch": 2.5849719101123596, + "grad_norm": 0.8724221587181091, + "learning_rate": 3.0646698082904847e-06, + "loss": 0.5482, + "step": 7362 + }, + { + "epoch": 2.585323033707865, + "grad_norm": 0.9085797071456909, + "learning_rate": 3.0642194208104076e-06, + "loss": 0.5633, + "step": 7363 + }, + { + "epoch": 2.585674157303371, + "grad_norm": 0.907699704170227, + "learning_rate": 3.0637690140345096e-06, + "loss": 0.5999, + "step": 7364 + }, + { + "epoch": 2.5860252808988764, + "grad_norm": 0.8974177837371826, + "learning_rate": 3.0633185879781928e-06, + "loss": 0.5982, + "step": 7365 + }, + { + "epoch": 2.586376404494382, + "grad_norm": 0.9150027632713318, + "learning_rate": 3.062868142656863e-06, + "loss": 0.5903, + "step": 7366 + }, + { + "epoch": 2.5867275280898876, + "grad_norm": 0.9065232872962952, + "learning_rate": 3.0624176780859244e-06, + "loss": 0.5787, + "step": 7367 + }, + { + "epoch": 2.587078651685393, + "grad_norm": 0.8992656469345093, + "learning_rate": 3.061967194280783e-06, + "loss": 0.6296, + "step": 7368 + }, + { + "epoch": 2.587429775280899, + "grad_norm": 0.8498324155807495, + "learning_rate": 3.061516691256843e-06, + "loss": 0.571, + "step": 7369 + }, + { + "epoch": 2.5877808988764044, + "grad_norm": 0.8866354823112488, + "learning_rate": 3.0610661690295134e-06, + "loss": 0.5957, + "step": 7370 + }, + { + "epoch": 2.58813202247191, + "grad_norm": 0.9311127066612244, + "learning_rate": 3.0606156276142014e-06, + "loss": 0.5961, + "step": 7371 + }, + { + "epoch": 2.5884831460674156, + "grad_norm": 0.8798646330833435, + "learning_rate": 3.060165067026314e-06, + "loss": 0.5894, + "step": 7372 + }, + { + "epoch": 2.588834269662921, + "grad_norm": 0.8873475193977356, + "learning_rate": 3.0597144872812605e-06, + "loss": 0.591, + "step": 7373 + }, + { + "epoch": 2.5891853932584272, + "grad_norm": 0.8915675282478333, + "learning_rate": 3.059263888394451e-06, + "loss": 0.5628, + "step": 7374 + }, + { + "epoch": 2.589536516853933, + "grad_norm": 0.9279215931892395, + "learning_rate": 3.058813270381295e-06, + "loss": 0.6062, + "step": 7375 + }, + { + "epoch": 2.5898876404494384, + "grad_norm": 0.9147570729255676, + "learning_rate": 3.058362633257203e-06, + "loss": 0.6004, + "step": 7376 + }, + { + "epoch": 2.590238764044944, + "grad_norm": 0.8872963190078735, + "learning_rate": 3.057911977037587e-06, + "loss": 0.5745, + "step": 7377 + }, + { + "epoch": 2.5905898876404496, + "grad_norm": 0.8807946443557739, + "learning_rate": 3.0574613017378598e-06, + "loss": 0.5886, + "step": 7378 + }, + { + "epoch": 2.590941011235955, + "grad_norm": 0.8931580781936646, + "learning_rate": 3.057010607373432e-06, + "loss": 0.5554, + "step": 7379 + }, + { + "epoch": 2.591292134831461, + "grad_norm": 0.9233186841011047, + "learning_rate": 3.0565598939597173e-06, + "loss": 0.5526, + "step": 7380 + }, + { + "epoch": 2.5916432584269664, + "grad_norm": 0.8855900168418884, + "learning_rate": 3.0561091615121317e-06, + "loss": 0.5805, + "step": 7381 + }, + { + "epoch": 2.591994382022472, + "grad_norm": 0.8893604874610901, + "learning_rate": 3.0556584100460883e-06, + "loss": 0.5898, + "step": 7382 + }, + { + "epoch": 2.5923455056179776, + "grad_norm": 0.8735301494598389, + "learning_rate": 3.0552076395770025e-06, + "loss": 0.567, + "step": 7383 + }, + { + "epoch": 2.592696629213483, + "grad_norm": 0.8719521164894104, + "learning_rate": 3.0547568501202905e-06, + "loss": 0.5431, + "step": 7384 + }, + { + "epoch": 2.593047752808989, + "grad_norm": 0.9043188691139221, + "learning_rate": 3.0543060416913694e-06, + "loss": 0.5531, + "step": 7385 + }, + { + "epoch": 2.5933988764044944, + "grad_norm": 0.9022269248962402, + "learning_rate": 3.053855214305656e-06, + "loss": 0.599, + "step": 7386 + }, + { + "epoch": 2.59375, + "grad_norm": 0.9153850078582764, + "learning_rate": 3.053404367978568e-06, + "loss": 0.5811, + "step": 7387 + }, + { + "epoch": 2.5941011235955056, + "grad_norm": 0.8980445265769958, + "learning_rate": 3.0529535027255243e-06, + "loss": 0.5868, + "step": 7388 + }, + { + "epoch": 2.594452247191011, + "grad_norm": 0.8804596662521362, + "learning_rate": 3.0525026185619436e-06, + "loss": 0.566, + "step": 7389 + }, + { + "epoch": 2.594803370786517, + "grad_norm": 0.9109121561050415, + "learning_rate": 3.0520517155032462e-06, + "loss": 0.5607, + "step": 7390 + }, + { + "epoch": 2.5951544943820224, + "grad_norm": 0.8740862011909485, + "learning_rate": 3.051600793564853e-06, + "loss": 0.5566, + "step": 7391 + }, + { + "epoch": 2.595505617977528, + "grad_norm": 0.873243510723114, + "learning_rate": 3.0511498527621847e-06, + "loss": 0.5611, + "step": 7392 + }, + { + "epoch": 2.5958567415730336, + "grad_norm": 0.8793222308158875, + "learning_rate": 3.0506988931106627e-06, + "loss": 0.5601, + "step": 7393 + }, + { + "epoch": 2.596207865168539, + "grad_norm": 0.8926047086715698, + "learning_rate": 3.0502479146257104e-06, + "loss": 0.5801, + "step": 7394 + }, + { + "epoch": 2.596558988764045, + "grad_norm": 0.9121901988983154, + "learning_rate": 3.0497969173227494e-06, + "loss": 0.5862, + "step": 7395 + }, + { + "epoch": 2.5969101123595504, + "grad_norm": 0.8930234313011169, + "learning_rate": 3.0493459012172044e-06, + "loss": 0.5742, + "step": 7396 + }, + { + "epoch": 2.597261235955056, + "grad_norm": 0.9104388952255249, + "learning_rate": 3.0488948663245e-06, + "loss": 0.5884, + "step": 7397 + }, + { + "epoch": 2.5976123595505616, + "grad_norm": 0.969490110874176, + "learning_rate": 3.0484438126600606e-06, + "loss": 0.6234, + "step": 7398 + }, + { + "epoch": 2.597963483146067, + "grad_norm": 0.9049099087715149, + "learning_rate": 3.0479927402393124e-06, + "loss": 0.5578, + "step": 7399 + }, + { + "epoch": 2.5983146067415728, + "grad_norm": 0.8937108516693115, + "learning_rate": 3.0475416490776812e-06, + "loss": 0.59, + "step": 7400 + }, + { + "epoch": 2.598665730337079, + "grad_norm": 0.9168557524681091, + "learning_rate": 3.0470905391905944e-06, + "loss": 0.6078, + "step": 7401 + }, + { + "epoch": 2.5990168539325844, + "grad_norm": 0.9291811585426331, + "learning_rate": 3.046639410593479e-06, + "loss": 0.6064, + "step": 7402 + }, + { + "epoch": 2.59936797752809, + "grad_norm": 0.9339472055435181, + "learning_rate": 3.0461882633017636e-06, + "loss": 0.5727, + "step": 7403 + }, + { + "epoch": 2.5997191011235956, + "grad_norm": 0.9143481254577637, + "learning_rate": 3.0457370973308774e-06, + "loss": 0.5795, + "step": 7404 + }, + { + "epoch": 2.600070224719101, + "grad_norm": 0.8766186833381653, + "learning_rate": 3.045285912696249e-06, + "loss": 0.629, + "step": 7405 + }, + { + "epoch": 2.600421348314607, + "grad_norm": 0.8886841535568237, + "learning_rate": 3.044834709413309e-06, + "loss": 0.5558, + "step": 7406 + }, + { + "epoch": 2.6007724719101124, + "grad_norm": 0.8979460597038269, + "learning_rate": 3.0443834874974882e-06, + "loss": 0.5805, + "step": 7407 + }, + { + "epoch": 2.601123595505618, + "grad_norm": 0.9234282374382019, + "learning_rate": 3.043932246964218e-06, + "loss": 0.5875, + "step": 7408 + }, + { + "epoch": 2.6014747191011236, + "grad_norm": 0.884569525718689, + "learning_rate": 3.0434809878289305e-06, + "loss": 0.5797, + "step": 7409 + }, + { + "epoch": 2.601825842696629, + "grad_norm": 0.9241189360618591, + "learning_rate": 3.043029710107058e-06, + "loss": 0.5604, + "step": 7410 + }, + { + "epoch": 2.602176966292135, + "grad_norm": 0.8926496505737305, + "learning_rate": 3.042578413814034e-06, + "loss": 0.5787, + "step": 7411 + }, + { + "epoch": 2.6025280898876404, + "grad_norm": 0.8973783254623413, + "learning_rate": 3.042127098965293e-06, + "loss": 0.5651, + "step": 7412 + }, + { + "epoch": 2.602879213483146, + "grad_norm": 0.8764045834541321, + "learning_rate": 3.0416757655762692e-06, + "loss": 0.5869, + "step": 7413 + }, + { + "epoch": 2.6032303370786516, + "grad_norm": 1.0176976919174194, + "learning_rate": 3.041224413662397e-06, + "loss": 0.5893, + "step": 7414 + }, + { + "epoch": 2.603581460674157, + "grad_norm": 0.8821653723716736, + "learning_rate": 3.0407730432391135e-06, + "loss": 0.5969, + "step": 7415 + }, + { + "epoch": 2.603932584269663, + "grad_norm": 0.9040608406066895, + "learning_rate": 3.040321654321855e-06, + "loss": 0.572, + "step": 7416 + }, + { + "epoch": 2.6042837078651684, + "grad_norm": 0.9187495112419128, + "learning_rate": 3.0398702469260583e-06, + "loss": 0.6031, + "step": 7417 + }, + { + "epoch": 2.6046348314606744, + "grad_norm": 0.9266928434371948, + "learning_rate": 3.0394188210671604e-06, + "loss": 0.6021, + "step": 7418 + }, + { + "epoch": 2.60498595505618, + "grad_norm": 0.8934510350227356, + "learning_rate": 3.038967376760601e-06, + "loss": 0.582, + "step": 7419 + }, + { + "epoch": 2.6053370786516856, + "grad_norm": 0.935521125793457, + "learning_rate": 3.038515914021818e-06, + "loss": 0.6006, + "step": 7420 + }, + { + "epoch": 2.605688202247191, + "grad_norm": 0.9077858328819275, + "learning_rate": 3.0380644328662516e-06, + "loss": 0.5731, + "step": 7421 + }, + { + "epoch": 2.606039325842697, + "grad_norm": 0.8944138288497925, + "learning_rate": 3.0376129333093425e-06, + "loss": 0.5921, + "step": 7422 + }, + { + "epoch": 2.6063904494382024, + "grad_norm": 0.8829587697982788, + "learning_rate": 3.037161415366531e-06, + "loss": 0.6057, + "step": 7423 + }, + { + "epoch": 2.606741573033708, + "grad_norm": 0.8940721154212952, + "learning_rate": 3.0367098790532585e-06, + "loss": 0.585, + "step": 7424 + }, + { + "epoch": 2.6070926966292136, + "grad_norm": 0.8949596285820007, + "learning_rate": 3.0362583243849676e-06, + "loss": 0.551, + "step": 7425 + }, + { + "epoch": 2.607443820224719, + "grad_norm": 0.9193463921546936, + "learning_rate": 3.0358067513771006e-06, + "loss": 0.5901, + "step": 7426 + }, + { + "epoch": 2.607794943820225, + "grad_norm": 0.9386188983917236, + "learning_rate": 3.0353551600451013e-06, + "loss": 0.5716, + "step": 7427 + }, + { + "epoch": 2.6081460674157304, + "grad_norm": 0.9175008535385132, + "learning_rate": 3.0349035504044137e-06, + "loss": 0.6007, + "step": 7428 + }, + { + "epoch": 2.608497191011236, + "grad_norm": 0.9020130038261414, + "learning_rate": 3.0344519224704825e-06, + "loss": 0.5729, + "step": 7429 + }, + { + "epoch": 2.6088483146067416, + "grad_norm": 0.9111397862434387, + "learning_rate": 3.0340002762587536e-06, + "loss": 0.5865, + "step": 7430 + }, + { + "epoch": 2.609199438202247, + "grad_norm": 0.864912748336792, + "learning_rate": 3.0335486117846715e-06, + "loss": 0.5749, + "step": 7431 + }, + { + "epoch": 2.609550561797753, + "grad_norm": 0.8830868005752563, + "learning_rate": 3.033096929063683e-06, + "loss": 0.5632, + "step": 7432 + }, + { + "epoch": 2.6099016853932584, + "grad_norm": 0.8986430168151855, + "learning_rate": 3.0326452281112363e-06, + "loss": 0.6117, + "step": 7433 + }, + { + "epoch": 2.610252808988764, + "grad_norm": 0.8825873136520386, + "learning_rate": 3.032193508942778e-06, + "loss": 0.5707, + "step": 7434 + }, + { + "epoch": 2.6106039325842696, + "grad_norm": 0.9106655120849609, + "learning_rate": 3.0317417715737574e-06, + "loss": 0.5671, + "step": 7435 + }, + { + "epoch": 2.610955056179775, + "grad_norm": 0.9067287445068359, + "learning_rate": 3.0312900160196235e-06, + "loss": 0.6073, + "step": 7436 + }, + { + "epoch": 2.611306179775281, + "grad_norm": 0.8884614109992981, + "learning_rate": 3.030838242295825e-06, + "loss": 0.5462, + "step": 7437 + }, + { + "epoch": 2.6116573033707864, + "grad_norm": 0.9295231103897095, + "learning_rate": 3.030386450417814e-06, + "loss": 0.6004, + "step": 7438 + }, + { + "epoch": 2.612008426966292, + "grad_norm": 0.9013570547103882, + "learning_rate": 3.0299346404010386e-06, + "loss": 0.5906, + "step": 7439 + }, + { + "epoch": 2.6123595505617976, + "grad_norm": 0.8821189403533936, + "learning_rate": 3.0294828122609538e-06, + "loss": 0.6155, + "step": 7440 + }, + { + "epoch": 2.612710674157303, + "grad_norm": 0.9484449625015259, + "learning_rate": 3.029030966013009e-06, + "loss": 0.602, + "step": 7441 + }, + { + "epoch": 2.613061797752809, + "grad_norm": 0.9285559058189392, + "learning_rate": 3.028579101672658e-06, + "loss": 0.5912, + "step": 7442 + }, + { + "epoch": 2.6134129213483144, + "grad_norm": 0.9206721186637878, + "learning_rate": 3.028127219255354e-06, + "loss": 0.5821, + "step": 7443 + }, + { + "epoch": 2.61376404494382, + "grad_norm": 0.8934462070465088, + "learning_rate": 3.0276753187765508e-06, + "loss": 0.626, + "step": 7444 + }, + { + "epoch": 2.6141151685393256, + "grad_norm": 0.9067831039428711, + "learning_rate": 3.0272234002517036e-06, + "loss": 0.5669, + "step": 7445 + }, + { + "epoch": 2.6144662921348316, + "grad_norm": 0.9498097896575928, + "learning_rate": 3.026771463696267e-06, + "loss": 0.6104, + "step": 7446 + }, + { + "epoch": 2.614817415730337, + "grad_norm": 0.8946022987365723, + "learning_rate": 3.026319509125697e-06, + "loss": 0.5919, + "step": 7447 + }, + { + "epoch": 2.615168539325843, + "grad_norm": 0.8940441608428955, + "learning_rate": 3.0258675365554513e-06, + "loss": 0.5961, + "step": 7448 + }, + { + "epoch": 2.6155196629213484, + "grad_norm": 0.8911185264587402, + "learning_rate": 3.025415546000985e-06, + "loss": 0.6026, + "step": 7449 + }, + { + "epoch": 2.615870786516854, + "grad_norm": 0.8965465426445007, + "learning_rate": 3.024963537477756e-06, + "loss": 0.586, + "step": 7450 + }, + { + "epoch": 2.6162219101123596, + "grad_norm": 0.8889825344085693, + "learning_rate": 3.0245115110012235e-06, + "loss": 0.5976, + "step": 7451 + }, + { + "epoch": 2.616573033707865, + "grad_norm": 0.9011339545249939, + "learning_rate": 3.0240594665868466e-06, + "loss": 0.601, + "step": 7452 + }, + { + "epoch": 2.616924157303371, + "grad_norm": 0.8892192840576172, + "learning_rate": 3.023607404250084e-06, + "loss": 0.5995, + "step": 7453 + }, + { + "epoch": 2.6172752808988764, + "grad_norm": 0.9045920968055725, + "learning_rate": 3.0231553240063966e-06, + "loss": 0.5652, + "step": 7454 + }, + { + "epoch": 2.617626404494382, + "grad_norm": 0.9050779342651367, + "learning_rate": 3.0227032258712454e-06, + "loss": 0.5624, + "step": 7455 + }, + { + "epoch": 2.6179775280898876, + "grad_norm": 0.8913336992263794, + "learning_rate": 3.0222511098600902e-06, + "loss": 0.5695, + "step": 7456 + }, + { + "epoch": 2.618328651685393, + "grad_norm": 0.9203185439109802, + "learning_rate": 3.021798975988394e-06, + "loss": 0.5728, + "step": 7457 + }, + { + "epoch": 2.618679775280899, + "grad_norm": 0.9134983420372009, + "learning_rate": 3.021346824271619e-06, + "loss": 0.5982, + "step": 7458 + }, + { + "epoch": 2.6190308988764044, + "grad_norm": 0.9019878506660461, + "learning_rate": 3.0208946547252292e-06, + "loss": 0.5869, + "step": 7459 + }, + { + "epoch": 2.61938202247191, + "grad_norm": 0.8988032937049866, + "learning_rate": 3.0204424673646883e-06, + "loss": 0.5479, + "step": 7460 + }, + { + "epoch": 2.6197331460674156, + "grad_norm": 0.8326671719551086, + "learning_rate": 3.01999026220546e-06, + "loss": 0.5403, + "step": 7461 + }, + { + "epoch": 2.620084269662921, + "grad_norm": 0.8832093477249146, + "learning_rate": 3.0195380392630097e-06, + "loss": 0.5432, + "step": 7462 + }, + { + "epoch": 2.6204353932584272, + "grad_norm": 0.9081989526748657, + "learning_rate": 3.0190857985528033e-06, + "loss": 0.5871, + "step": 7463 + }, + { + "epoch": 2.620786516853933, + "grad_norm": 0.8925137519836426, + "learning_rate": 3.018633540090306e-06, + "loss": 0.5907, + "step": 7464 + }, + { + "epoch": 2.6211376404494384, + "grad_norm": 0.8523365259170532, + "learning_rate": 3.0181812638909864e-06, + "loss": 0.5536, + "step": 7465 + }, + { + "epoch": 2.621488764044944, + "grad_norm": 0.9297746419906616, + "learning_rate": 3.01772896997031e-06, + "loss": 0.6013, + "step": 7466 + }, + { + "epoch": 2.6218398876404496, + "grad_norm": 0.8995739221572876, + "learning_rate": 3.017276658343747e-06, + "loss": 0.6173, + "step": 7467 + }, + { + "epoch": 2.622191011235955, + "grad_norm": 0.9009650945663452, + "learning_rate": 3.016824329026764e-06, + "loss": 0.5587, + "step": 7468 + }, + { + "epoch": 2.622542134831461, + "grad_norm": 0.9247835874557495, + "learning_rate": 3.016371982034832e-06, + "loss": 0.5831, + "step": 7469 + }, + { + "epoch": 2.6228932584269664, + "grad_norm": 0.8929818272590637, + "learning_rate": 3.0159196173834193e-06, + "loss": 0.5574, + "step": 7470 + }, + { + "epoch": 2.623244382022472, + "grad_norm": 0.8876343369483948, + "learning_rate": 3.015467235087998e-06, + "loss": 0.573, + "step": 7471 + }, + { + "epoch": 2.6235955056179776, + "grad_norm": 0.8867413997650146, + "learning_rate": 3.0150148351640373e-06, + "loss": 0.5894, + "step": 7472 + }, + { + "epoch": 2.623946629213483, + "grad_norm": 0.9058734774589539, + "learning_rate": 3.0145624176270105e-06, + "loss": 0.619, + "step": 7473 + }, + { + "epoch": 2.624297752808989, + "grad_norm": 0.9514951705932617, + "learning_rate": 3.0141099824923892e-06, + "loss": 0.6, + "step": 7474 + }, + { + "epoch": 2.6246488764044944, + "grad_norm": 0.8839761018753052, + "learning_rate": 3.0136575297756463e-06, + "loss": 0.5692, + "step": 7475 + }, + { + "epoch": 2.625, + "grad_norm": 0.9004509449005127, + "learning_rate": 3.0132050594922558e-06, + "loss": 0.5688, + "step": 7476 + }, + { + "epoch": 2.6253511235955056, + "grad_norm": 0.8842835426330566, + "learning_rate": 3.012752571657691e-06, + "loss": 0.5793, + "step": 7477 + }, + { + "epoch": 2.625702247191011, + "grad_norm": 0.9110257625579834, + "learning_rate": 3.0123000662874275e-06, + "loss": 0.5876, + "step": 7478 + }, + { + "epoch": 2.626053370786517, + "grad_norm": 0.912479817867279, + "learning_rate": 3.0118475433969403e-06, + "loss": 0.5841, + "step": 7479 + }, + { + "epoch": 2.6264044943820224, + "grad_norm": 0.8699826002120972, + "learning_rate": 3.011395003001705e-06, + "loss": 0.5737, + "step": 7480 + }, + { + "epoch": 2.626755617977528, + "grad_norm": 0.9493129253387451, + "learning_rate": 3.0109424451171976e-06, + "loss": 0.6122, + "step": 7481 + }, + { + "epoch": 2.6271067415730336, + "grad_norm": 0.9309719204902649, + "learning_rate": 3.010489869758896e-06, + "loss": 0.6141, + "step": 7482 + }, + { + "epoch": 2.627457865168539, + "grad_norm": 0.8564309477806091, + "learning_rate": 3.0100372769422777e-06, + "loss": 0.5546, + "step": 7483 + }, + { + "epoch": 2.627808988764045, + "grad_norm": 0.8737847208976746, + "learning_rate": 3.009584666682821e-06, + "loss": 0.574, + "step": 7484 + }, + { + "epoch": 2.6281601123595504, + "grad_norm": 0.9153870344161987, + "learning_rate": 3.009132038996005e-06, + "loss": 0.6203, + "step": 7485 + }, + { + "epoch": 2.628511235955056, + "grad_norm": 0.9143859148025513, + "learning_rate": 3.0086793938973084e-06, + "loss": 0.6005, + "step": 7486 + }, + { + "epoch": 2.6288623595505616, + "grad_norm": 0.8971354365348816, + "learning_rate": 3.0082267314022123e-06, + "loss": 0.5677, + "step": 7487 + }, + { + "epoch": 2.629213483146067, + "grad_norm": 0.8730185031890869, + "learning_rate": 3.0077740515261966e-06, + "loss": 0.5566, + "step": 7488 + }, + { + "epoch": 2.6295646067415728, + "grad_norm": 0.9006599187850952, + "learning_rate": 3.007321354284743e-06, + "loss": 0.6019, + "step": 7489 + }, + { + "epoch": 2.629915730337079, + "grad_norm": 0.9241135120391846, + "learning_rate": 3.006868639693333e-06, + "loss": 0.5804, + "step": 7490 + }, + { + "epoch": 2.6302668539325844, + "grad_norm": 0.964394748210907, + "learning_rate": 3.0064159077674495e-06, + "loss": 0.6091, + "step": 7491 + }, + { + "epoch": 2.63061797752809, + "grad_norm": 0.9133546352386475, + "learning_rate": 3.0059631585225758e-06, + "loss": 0.5748, + "step": 7492 + }, + { + "epoch": 2.6309691011235956, + "grad_norm": 0.8920817971229553, + "learning_rate": 3.0055103919741946e-06, + "loss": 0.6038, + "step": 7493 + }, + { + "epoch": 2.631320224719101, + "grad_norm": 0.9057829976081848, + "learning_rate": 3.0050576081377904e-06, + "loss": 0.5728, + "step": 7494 + }, + { + "epoch": 2.631671348314607, + "grad_norm": 0.901849627494812, + "learning_rate": 3.004604807028848e-06, + "loss": 0.5753, + "step": 7495 + }, + { + "epoch": 2.6320224719101124, + "grad_norm": 0.9440762996673584, + "learning_rate": 3.0041519886628535e-06, + "loss": 0.5834, + "step": 7496 + }, + { + "epoch": 2.632373595505618, + "grad_norm": 0.8661344647407532, + "learning_rate": 3.0036991530552917e-06, + "loss": 0.5601, + "step": 7497 + }, + { + "epoch": 2.6327247191011236, + "grad_norm": 0.89628005027771, + "learning_rate": 3.0032463002216504e-06, + "loss": 0.5977, + "step": 7498 + }, + { + "epoch": 2.633075842696629, + "grad_norm": 0.9727967977523804, + "learning_rate": 3.0027934301774165e-06, + "loss": 0.6285, + "step": 7499 + }, + { + "epoch": 2.633426966292135, + "grad_norm": 0.9229931235313416, + "learning_rate": 3.002340542938077e-06, + "loss": 0.5874, + "step": 7500 + }, + { + "epoch": 2.6337780898876404, + "grad_norm": 0.9079124927520752, + "learning_rate": 3.001887638519121e-06, + "loss": 0.5806, + "step": 7501 + }, + { + "epoch": 2.634129213483146, + "grad_norm": 0.9008191823959351, + "learning_rate": 3.001434716936037e-06, + "loss": 0.5978, + "step": 7502 + }, + { + "epoch": 2.6344803370786516, + "grad_norm": 0.9136261940002441, + "learning_rate": 3.000981778204315e-06, + "loss": 0.6277, + "step": 7503 + }, + { + "epoch": 2.634831460674157, + "grad_norm": 0.935242772102356, + "learning_rate": 3.0005288223394455e-06, + "loss": 0.591, + "step": 7504 + }, + { + "epoch": 2.635182584269663, + "grad_norm": 0.8957575559616089, + "learning_rate": 3.0000758493569183e-06, + "loss": 0.5872, + "step": 7505 + }, + { + "epoch": 2.6355337078651684, + "grad_norm": 0.9074363112449646, + "learning_rate": 2.999622859272224e-06, + "loss": 0.6128, + "step": 7506 + }, + { + "epoch": 2.6358848314606744, + "grad_norm": 0.8998899459838867, + "learning_rate": 2.999169852100856e-06, + "loss": 0.5724, + "step": 7507 + }, + { + "epoch": 2.63623595505618, + "grad_norm": 0.8857282996177673, + "learning_rate": 2.998716827858307e-06, + "loss": 0.5469, + "step": 7508 + }, + { + "epoch": 2.6365870786516856, + "grad_norm": 0.8999072313308716, + "learning_rate": 2.9982637865600682e-06, + "loss": 0.5641, + "step": 7509 + }, + { + "epoch": 2.636938202247191, + "grad_norm": 0.891825795173645, + "learning_rate": 2.9978107282216345e-06, + "loss": 0.5479, + "step": 7510 + }, + { + "epoch": 2.637289325842697, + "grad_norm": 0.8776609897613525, + "learning_rate": 2.9973576528585e-06, + "loss": 0.564, + "step": 7511 + }, + { + "epoch": 2.6376404494382024, + "grad_norm": 0.9149337410926819, + "learning_rate": 2.996904560486159e-06, + "loss": 0.6172, + "step": 7512 + }, + { + "epoch": 2.637991573033708, + "grad_norm": 0.9021109342575073, + "learning_rate": 2.996451451120108e-06, + "loss": 0.5995, + "step": 7513 + }, + { + "epoch": 2.6383426966292136, + "grad_norm": 0.9293187856674194, + "learning_rate": 2.9959983247758416e-06, + "loss": 0.5975, + "step": 7514 + }, + { + "epoch": 2.638693820224719, + "grad_norm": 0.9530065655708313, + "learning_rate": 2.9955451814688567e-06, + "loss": 0.6001, + "step": 7515 + }, + { + "epoch": 2.639044943820225, + "grad_norm": 0.9049083590507507, + "learning_rate": 2.995092021214652e-06, + "loss": 0.5778, + "step": 7516 + }, + { + "epoch": 2.6393960674157304, + "grad_norm": 0.9356215596199036, + "learning_rate": 2.994638844028723e-06, + "loss": 0.5871, + "step": 7517 + }, + { + "epoch": 2.639747191011236, + "grad_norm": 0.9148952960968018, + "learning_rate": 2.9941856499265685e-06, + "loss": 0.5876, + "step": 7518 + }, + { + "epoch": 2.6400983146067416, + "grad_norm": 0.873844563961029, + "learning_rate": 2.9937324389236877e-06, + "loss": 0.5528, + "step": 7519 + }, + { + "epoch": 2.640449438202247, + "grad_norm": 0.8968779444694519, + "learning_rate": 2.9932792110355803e-06, + "loss": 0.5508, + "step": 7520 + }, + { + "epoch": 2.640800561797753, + "grad_norm": 0.907414972782135, + "learning_rate": 2.992825966277746e-06, + "loss": 0.5822, + "step": 7521 + }, + { + "epoch": 2.6411516853932584, + "grad_norm": 1.0141228437423706, + "learning_rate": 2.992372704665685e-06, + "loss": 0.5577, + "step": 7522 + }, + { + "epoch": 2.641502808988764, + "grad_norm": 0.9177740812301636, + "learning_rate": 2.9919194262148987e-06, + "loss": 0.6117, + "step": 7523 + }, + { + "epoch": 2.6418539325842696, + "grad_norm": 0.8881279230117798, + "learning_rate": 2.9914661309408895e-06, + "loss": 0.5747, + "step": 7524 + }, + { + "epoch": 2.642205056179775, + "grad_norm": 0.893361508846283, + "learning_rate": 2.9910128188591586e-06, + "loss": 0.5896, + "step": 7525 + }, + { + "epoch": 2.642556179775281, + "grad_norm": 0.888554036617279, + "learning_rate": 2.9905594899852095e-06, + "loss": 0.58, + "step": 7526 + }, + { + "epoch": 2.6429073033707864, + "grad_norm": 0.897283136844635, + "learning_rate": 2.990106144334547e-06, + "loss": 0.5923, + "step": 7527 + }, + { + "epoch": 2.643258426966292, + "grad_norm": 0.9431992173194885, + "learning_rate": 2.989652781922672e-06, + "loss": 0.5868, + "step": 7528 + }, + { + "epoch": 2.6436095505617976, + "grad_norm": 0.8821288347244263, + "learning_rate": 2.989199402765092e-06, + "loss": 0.5565, + "step": 7529 + }, + { + "epoch": 2.643960674157303, + "grad_norm": 0.9408068656921387, + "learning_rate": 2.9887460068773107e-06, + "loss": 0.6062, + "step": 7530 + }, + { + "epoch": 2.644311797752809, + "grad_norm": 0.9126775860786438, + "learning_rate": 2.988292594274834e-06, + "loss": 0.6274, + "step": 7531 + }, + { + "epoch": 2.6446629213483144, + "grad_norm": 0.9338752627372742, + "learning_rate": 2.9878391649731686e-06, + "loss": 0.579, + "step": 7532 + }, + { + "epoch": 2.64501404494382, + "grad_norm": 0.9266043901443481, + "learning_rate": 2.987385718987821e-06, + "loss": 0.6034, + "step": 7533 + }, + { + "epoch": 2.6453651685393256, + "grad_norm": 0.9097104072570801, + "learning_rate": 2.9869322563342988e-06, + "loss": 0.5916, + "step": 7534 + }, + { + "epoch": 2.6457162921348316, + "grad_norm": 0.9424006342887878, + "learning_rate": 2.9864787770281106e-06, + "loss": 0.563, + "step": 7535 + }, + { + "epoch": 2.646067415730337, + "grad_norm": 0.9103488922119141, + "learning_rate": 2.9860252810847647e-06, + "loss": 0.58, + "step": 7536 + }, + { + "epoch": 2.646418539325843, + "grad_norm": 0.9027099609375, + "learning_rate": 2.9855717685197694e-06, + "loss": 0.5772, + "step": 7537 + }, + { + "epoch": 2.6467696629213484, + "grad_norm": 0.8858651518821716, + "learning_rate": 2.985118239348635e-06, + "loss": 0.5949, + "step": 7538 + }, + { + "epoch": 2.647120786516854, + "grad_norm": 0.9033393263816833, + "learning_rate": 2.9846646935868724e-06, + "loss": 0.5806, + "step": 7539 + }, + { + "epoch": 2.6474719101123596, + "grad_norm": 0.9138420224189758, + "learning_rate": 2.9842111312499916e-06, + "loss": 0.5766, + "step": 7540 + }, + { + "epoch": 2.647823033707865, + "grad_norm": 0.899876058101654, + "learning_rate": 2.9837575523535044e-06, + "loss": 0.5633, + "step": 7541 + }, + { + "epoch": 2.648174157303371, + "grad_norm": 0.9311326146125793, + "learning_rate": 2.983303956912924e-06, + "loss": 0.5929, + "step": 7542 + }, + { + "epoch": 2.6485252808988764, + "grad_norm": 0.9021244049072266, + "learning_rate": 2.982850344943761e-06, + "loss": 0.5871, + "step": 7543 + }, + { + "epoch": 2.648876404494382, + "grad_norm": 0.8931776285171509, + "learning_rate": 2.982396716461529e-06, + "loss": 0.588, + "step": 7544 + }, + { + "epoch": 2.6492275280898876, + "grad_norm": 0.9432953596115112, + "learning_rate": 2.9819430714817426e-06, + "loss": 0.5756, + "step": 7545 + }, + { + "epoch": 2.649578651685393, + "grad_norm": 0.9166197776794434, + "learning_rate": 2.9814894100199155e-06, + "loss": 0.588, + "step": 7546 + }, + { + "epoch": 2.649929775280899, + "grad_norm": 0.9054636359214783, + "learning_rate": 2.981035732091563e-06, + "loss": 0.5804, + "step": 7547 + }, + { + "epoch": 2.6502808988764044, + "grad_norm": 0.9029044508934021, + "learning_rate": 2.980582037712199e-06, + "loss": 0.5705, + "step": 7548 + }, + { + "epoch": 2.65063202247191, + "grad_norm": 0.898112416267395, + "learning_rate": 2.9801283268973413e-06, + "loss": 0.5777, + "step": 7549 + }, + { + "epoch": 2.6509831460674156, + "grad_norm": 0.8770557045936584, + "learning_rate": 2.9796745996625054e-06, + "loss": 0.5519, + "step": 7550 + }, + { + "epoch": 2.651334269662921, + "grad_norm": 0.9042981863021851, + "learning_rate": 2.9792208560232088e-06, + "loss": 0.5831, + "step": 7551 + }, + { + "epoch": 2.6516853932584272, + "grad_norm": 0.8893117904663086, + "learning_rate": 2.978767095994969e-06, + "loss": 0.5533, + "step": 7552 + }, + { + "epoch": 2.652036516853933, + "grad_norm": 0.9130943417549133, + "learning_rate": 2.9783133195933043e-06, + "loss": 0.6057, + "step": 7553 + }, + { + "epoch": 2.6523876404494384, + "grad_norm": 0.8703882098197937, + "learning_rate": 2.9778595268337335e-06, + "loss": 0.5738, + "step": 7554 + }, + { + "epoch": 2.652738764044944, + "grad_norm": 0.904779851436615, + "learning_rate": 2.9774057177317763e-06, + "loss": 0.5934, + "step": 7555 + }, + { + "epoch": 2.6530898876404496, + "grad_norm": 0.9051056504249573, + "learning_rate": 2.9769518923029507e-06, + "loss": 0.5665, + "step": 7556 + }, + { + "epoch": 2.653441011235955, + "grad_norm": 0.92510586977005, + "learning_rate": 2.9764980505627794e-06, + "loss": 0.5823, + "step": 7557 + }, + { + "epoch": 2.653792134831461, + "grad_norm": 0.9277645349502563, + "learning_rate": 2.976044192526782e-06, + "loss": 0.5885, + "step": 7558 + }, + { + "epoch": 2.6541432584269664, + "grad_norm": 0.8838064074516296, + "learning_rate": 2.975590318210481e-06, + "loss": 0.567, + "step": 7559 + }, + { + "epoch": 2.654494382022472, + "grad_norm": 0.9391385316848755, + "learning_rate": 2.9751364276293982e-06, + "loss": 0.5878, + "step": 7560 + }, + { + "epoch": 2.6548455056179776, + "grad_norm": 0.918768584728241, + "learning_rate": 2.9746825207990555e-06, + "loss": 0.5997, + "step": 7561 + }, + { + "epoch": 2.655196629213483, + "grad_norm": 0.9046016335487366, + "learning_rate": 2.9742285977349774e-06, + "loss": 0.5927, + "step": 7562 + }, + { + "epoch": 2.655547752808989, + "grad_norm": 0.9006232023239136, + "learning_rate": 2.9737746584526866e-06, + "loss": 0.5638, + "step": 7563 + }, + { + "epoch": 2.6558988764044944, + "grad_norm": 0.8865199089050293, + "learning_rate": 2.973320702967708e-06, + "loss": 0.5905, + "step": 7564 + }, + { + "epoch": 2.65625, + "grad_norm": 0.8879817128181458, + "learning_rate": 2.9728667312955667e-06, + "loss": 0.5899, + "step": 7565 + }, + { + "epoch": 2.6566011235955056, + "grad_norm": 0.9215937256813049, + "learning_rate": 2.9724127434517873e-06, + "loss": 0.6114, + "step": 7566 + }, + { + "epoch": 2.656952247191011, + "grad_norm": 0.9561192989349365, + "learning_rate": 2.971958739451897e-06, + "loss": 0.5923, + "step": 7567 + }, + { + "epoch": 2.657303370786517, + "grad_norm": 0.889039933681488, + "learning_rate": 2.9715047193114215e-06, + "loss": 0.5944, + "step": 7568 + }, + { + "epoch": 2.6576544943820224, + "grad_norm": 0.9298639297485352, + "learning_rate": 2.9710506830458878e-06, + "loss": 0.6079, + "step": 7569 + }, + { + "epoch": 2.658005617977528, + "grad_norm": 0.8852623105049133, + "learning_rate": 2.970596630670824e-06, + "loss": 0.5673, + "step": 7570 + }, + { + "epoch": 2.6583567415730336, + "grad_norm": 0.9156980514526367, + "learning_rate": 2.9701425622017583e-06, + "loss": 0.5838, + "step": 7571 + }, + { + "epoch": 2.658707865168539, + "grad_norm": 0.8899957537651062, + "learning_rate": 2.969688477654219e-06, + "loss": 0.5732, + "step": 7572 + }, + { + "epoch": 2.659058988764045, + "grad_norm": 0.9103763103485107, + "learning_rate": 2.9692343770437352e-06, + "loss": 0.6099, + "step": 7573 + }, + { + "epoch": 2.6594101123595504, + "grad_norm": 0.9026846885681152, + "learning_rate": 2.9687802603858377e-06, + "loss": 0.5743, + "step": 7574 + }, + { + "epoch": 2.659761235955056, + "grad_norm": 0.9048449993133545, + "learning_rate": 2.9683261276960566e-06, + "loss": 0.6085, + "step": 7575 + }, + { + "epoch": 2.6601123595505616, + "grad_norm": 0.9376624822616577, + "learning_rate": 2.967871978989923e-06, + "loss": 0.5917, + "step": 7576 + }, + { + "epoch": 2.660463483146067, + "grad_norm": 0.9245718717575073, + "learning_rate": 2.9674178142829672e-06, + "loss": 0.568, + "step": 7577 + }, + { + "epoch": 2.6608146067415728, + "grad_norm": 0.9038835167884827, + "learning_rate": 2.9669636335907227e-06, + "loss": 0.5445, + "step": 7578 + }, + { + "epoch": 2.661165730337079, + "grad_norm": 0.8837337493896484, + "learning_rate": 2.966509436928721e-06, + "loss": 0.5929, + "step": 7579 + }, + { + "epoch": 2.6615168539325844, + "grad_norm": 0.8826956748962402, + "learning_rate": 2.966055224312497e-06, + "loss": 0.5655, + "step": 7580 + }, + { + "epoch": 2.66186797752809, + "grad_norm": 0.8991236686706543, + "learning_rate": 2.965600995757583e-06, + "loss": 0.5742, + "step": 7581 + }, + { + "epoch": 2.6622191011235956, + "grad_norm": 0.9094868302345276, + "learning_rate": 2.9651467512795116e-06, + "loss": 0.6193, + "step": 7582 + }, + { + "epoch": 2.662570224719101, + "grad_norm": 0.9033479690551758, + "learning_rate": 2.9646924908938208e-06, + "loss": 0.5995, + "step": 7583 + }, + { + "epoch": 2.662921348314607, + "grad_norm": 0.9999510645866394, + "learning_rate": 2.9642382146160444e-06, + "loss": 0.5816, + "step": 7584 + }, + { + "epoch": 2.6632724719101124, + "grad_norm": 0.9389203190803528, + "learning_rate": 2.9637839224617175e-06, + "loss": 0.6262, + "step": 7585 + }, + { + "epoch": 2.663623595505618, + "grad_norm": 0.9023115634918213, + "learning_rate": 2.9633296144463776e-06, + "loss": 0.5752, + "step": 7586 + }, + { + "epoch": 2.6639747191011236, + "grad_norm": 0.9151685833930969, + "learning_rate": 2.962875290585562e-06, + "loss": 0.5493, + "step": 7587 + }, + { + "epoch": 2.664325842696629, + "grad_norm": 0.8865213990211487, + "learning_rate": 2.9624209508948066e-06, + "loss": 0.5608, + "step": 7588 + }, + { + "epoch": 2.664676966292135, + "grad_norm": 0.8870331645011902, + "learning_rate": 2.961966595389651e-06, + "loss": 0.5837, + "step": 7589 + }, + { + "epoch": 2.6650280898876404, + "grad_norm": 0.8965469598770142, + "learning_rate": 2.961512224085633e-06, + "loss": 0.5695, + "step": 7590 + }, + { + "epoch": 2.665379213483146, + "grad_norm": 0.8908924460411072, + "learning_rate": 2.9610578369982917e-06, + "loss": 0.5509, + "step": 7591 + }, + { + "epoch": 2.6657303370786516, + "grad_norm": 0.921562910079956, + "learning_rate": 2.9606034341431672e-06, + "loss": 0.5824, + "step": 7592 + }, + { + "epoch": 2.666081460674157, + "grad_norm": 0.913086473941803, + "learning_rate": 2.9601490155358e-06, + "loss": 0.5846, + "step": 7593 + }, + { + "epoch": 2.666432584269663, + "grad_norm": 0.9352757334709167, + "learning_rate": 2.9596945811917292e-06, + "loss": 0.601, + "step": 7594 + }, + { + "epoch": 2.6667837078651684, + "grad_norm": 0.8898943662643433, + "learning_rate": 2.959240131126497e-06, + "loss": 0.569, + "step": 7595 + }, + { + "epoch": 2.6671348314606744, + "grad_norm": 0.8933197259902954, + "learning_rate": 2.958785665355646e-06, + "loss": 0.5516, + "step": 7596 + }, + { + "epoch": 2.66748595505618, + "grad_norm": 0.9233509302139282, + "learning_rate": 2.9583311838947178e-06, + "loss": 0.65, + "step": 7597 + }, + { + "epoch": 2.6678370786516856, + "grad_norm": 0.8976076245307922, + "learning_rate": 2.9578766867592545e-06, + "loss": 0.6035, + "step": 7598 + }, + { + "epoch": 2.668188202247191, + "grad_norm": 0.899989128112793, + "learning_rate": 2.957422173964801e-06, + "loss": 0.5602, + "step": 7599 + }, + { + "epoch": 2.668539325842697, + "grad_norm": 0.9107141494750977, + "learning_rate": 2.9569676455269003e-06, + "loss": 0.5786, + "step": 7600 + }, + { + "epoch": 2.6688904494382024, + "grad_norm": 0.9022222757339478, + "learning_rate": 2.9565131014610976e-06, + "loss": 0.5862, + "step": 7601 + }, + { + "epoch": 2.669241573033708, + "grad_norm": 0.856817364692688, + "learning_rate": 2.956058541782937e-06, + "loss": 0.5394, + "step": 7602 + }, + { + "epoch": 2.6695926966292136, + "grad_norm": 0.8951742053031921, + "learning_rate": 2.9556039665079643e-06, + "loss": 0.5911, + "step": 7603 + }, + { + "epoch": 2.669943820224719, + "grad_norm": 0.9154193997383118, + "learning_rate": 2.955149375651726e-06, + "loss": 0.5899, + "step": 7604 + }, + { + "epoch": 2.670294943820225, + "grad_norm": 0.9063310623168945, + "learning_rate": 2.9546947692297702e-06, + "loss": 0.6438, + "step": 7605 + }, + { + "epoch": 2.6706460674157304, + "grad_norm": 0.8875969648361206, + "learning_rate": 2.9542401472576404e-06, + "loss": 0.5695, + "step": 7606 + }, + { + "epoch": 2.670997191011236, + "grad_norm": 0.9202022552490234, + "learning_rate": 2.9537855097508865e-06, + "loss": 0.608, + "step": 7607 + }, + { + "epoch": 2.6713483146067416, + "grad_norm": 0.8828110694885254, + "learning_rate": 2.9533308567250567e-06, + "loss": 0.5545, + "step": 7608 + }, + { + "epoch": 2.671699438202247, + "grad_norm": 0.8986119031906128, + "learning_rate": 2.9528761881956995e-06, + "loss": 0.5906, + "step": 7609 + }, + { + "epoch": 2.672050561797753, + "grad_norm": 0.8997572064399719, + "learning_rate": 2.9524215041783644e-06, + "loss": 0.5527, + "step": 7610 + }, + { + "epoch": 2.6724016853932584, + "grad_norm": 0.9033135175704956, + "learning_rate": 2.9519668046886007e-06, + "loss": 0.6355, + "step": 7611 + }, + { + "epoch": 2.672752808988764, + "grad_norm": 0.9109470248222351, + "learning_rate": 2.9515120897419595e-06, + "loss": 0.5878, + "step": 7612 + }, + { + "epoch": 2.6731039325842696, + "grad_norm": 0.9041316509246826, + "learning_rate": 2.9510573593539908e-06, + "loss": 0.5764, + "step": 7613 + }, + { + "epoch": 2.673455056179775, + "grad_norm": 0.9000607132911682, + "learning_rate": 2.950602613540247e-06, + "loss": 0.5793, + "step": 7614 + }, + { + "epoch": 2.673806179775281, + "grad_norm": 0.9038352966308594, + "learning_rate": 2.950147852316279e-06, + "loss": 0.5838, + "step": 7615 + }, + { + "epoch": 2.6741573033707864, + "grad_norm": 0.933503270149231, + "learning_rate": 2.9496930756976395e-06, + "loss": 0.5814, + "step": 7616 + }, + { + "epoch": 2.674508426966292, + "grad_norm": 0.8916289806365967, + "learning_rate": 2.949238283699882e-06, + "loss": 0.581, + "step": 7617 + }, + { + "epoch": 2.6748595505617976, + "grad_norm": 0.9005141854286194, + "learning_rate": 2.9487834763385585e-06, + "loss": 0.5656, + "step": 7618 + }, + { + "epoch": 2.675210674157303, + "grad_norm": 0.9058454036712646, + "learning_rate": 2.948328653629225e-06, + "loss": 0.5754, + "step": 7619 + }, + { + "epoch": 2.675561797752809, + "grad_norm": 0.8913198113441467, + "learning_rate": 2.9478738155874353e-06, + "loss": 0.5516, + "step": 7620 + }, + { + "epoch": 2.6759129213483144, + "grad_norm": 0.8964958786964417, + "learning_rate": 2.9474189622287445e-06, + "loss": 0.55, + "step": 7621 + }, + { + "epoch": 2.67626404494382, + "grad_norm": 0.9086730480194092, + "learning_rate": 2.9469640935687077e-06, + "loss": 0.584, + "step": 7622 + }, + { + "epoch": 2.6766151685393256, + "grad_norm": 0.8964736461639404, + "learning_rate": 2.9465092096228813e-06, + "loss": 0.5862, + "step": 7623 + }, + { + "epoch": 2.6769662921348316, + "grad_norm": 0.902264416217804, + "learning_rate": 2.9460543104068224e-06, + "loss": 0.5977, + "step": 7624 + }, + { + "epoch": 2.677317415730337, + "grad_norm": 0.9111658334732056, + "learning_rate": 2.9455993959360873e-06, + "loss": 0.5971, + "step": 7625 + }, + { + "epoch": 2.677668539325843, + "grad_norm": 0.9154437184333801, + "learning_rate": 2.945144466226234e-06, + "loss": 0.6027, + "step": 7626 + }, + { + "epoch": 2.6780196629213484, + "grad_norm": 0.9174174666404724, + "learning_rate": 2.9446895212928214e-06, + "loss": 0.5975, + "step": 7627 + }, + { + "epoch": 2.678370786516854, + "grad_norm": 0.9467739462852478, + "learning_rate": 2.944234561151407e-06, + "loss": 0.5644, + "step": 7628 + }, + { + "epoch": 2.6787219101123596, + "grad_norm": 0.8814003467559814, + "learning_rate": 2.9437795858175516e-06, + "loss": 0.5889, + "step": 7629 + }, + { + "epoch": 2.679073033707865, + "grad_norm": 0.9380208253860474, + "learning_rate": 2.943324595306814e-06, + "loss": 0.5922, + "step": 7630 + }, + { + "epoch": 2.679424157303371, + "grad_norm": 0.9046818017959595, + "learning_rate": 2.942869589634755e-06, + "loss": 0.597, + "step": 7631 + }, + { + "epoch": 2.6797752808988764, + "grad_norm": 0.9101907014846802, + "learning_rate": 2.9424145688169335e-06, + "loss": 0.5743, + "step": 7632 + }, + { + "epoch": 2.680126404494382, + "grad_norm": 0.923832893371582, + "learning_rate": 2.9419595328689136e-06, + "loss": 0.5754, + "step": 7633 + }, + { + "epoch": 2.6804775280898876, + "grad_norm": 0.8889986276626587, + "learning_rate": 2.9415044818062553e-06, + "loss": 0.5923, + "step": 7634 + }, + { + "epoch": 2.680828651685393, + "grad_norm": 0.8886246681213379, + "learning_rate": 2.941049415644522e-06, + "loss": 0.5561, + "step": 7635 + }, + { + "epoch": 2.681179775280899, + "grad_norm": 0.9242022633552551, + "learning_rate": 2.9405943343992755e-06, + "loss": 0.5698, + "step": 7636 + }, + { + "epoch": 2.6815308988764044, + "grad_norm": 0.907970130443573, + "learning_rate": 2.94013923808608e-06, + "loss": 0.5456, + "step": 7637 + }, + { + "epoch": 2.68188202247191, + "grad_norm": 0.8920653462409973, + "learning_rate": 2.939684126720499e-06, + "loss": 0.5964, + "step": 7638 + }, + { + "epoch": 2.6822331460674156, + "grad_norm": 0.8510951399803162, + "learning_rate": 2.939229000318097e-06, + "loss": 0.5257, + "step": 7639 + }, + { + "epoch": 2.682584269662921, + "grad_norm": 0.8985562324523926, + "learning_rate": 2.9387738588944397e-06, + "loss": 0.5883, + "step": 7640 + }, + { + "epoch": 2.6829353932584272, + "grad_norm": 0.9025958776473999, + "learning_rate": 2.9383187024650916e-06, + "loss": 0.5739, + "step": 7641 + }, + { + "epoch": 2.683286516853933, + "grad_norm": 0.8917282223701477, + "learning_rate": 2.937863531045619e-06, + "loss": 0.5873, + "step": 7642 + }, + { + "epoch": 2.6836376404494384, + "grad_norm": 0.9282995462417603, + "learning_rate": 2.937408344651589e-06, + "loss": 0.5408, + "step": 7643 + }, + { + "epoch": 2.683988764044944, + "grad_norm": 0.9218565225601196, + "learning_rate": 2.9369531432985672e-06, + "loss": 0.5874, + "step": 7644 + }, + { + "epoch": 2.6843398876404496, + "grad_norm": 0.8886592984199524, + "learning_rate": 2.9364979270021214e-06, + "loss": 0.5855, + "step": 7645 + }, + { + "epoch": 2.684691011235955, + "grad_norm": 0.9213438034057617, + "learning_rate": 2.9360426957778205e-06, + "loss": 0.5571, + "step": 7646 + }, + { + "epoch": 2.685042134831461, + "grad_norm": 0.8927668333053589, + "learning_rate": 2.9355874496412323e-06, + "loss": 0.569, + "step": 7647 + }, + { + "epoch": 2.6853932584269664, + "grad_norm": 0.9174467921257019, + "learning_rate": 2.9351321886079265e-06, + "loss": 0.6059, + "step": 7648 + }, + { + "epoch": 2.685744382022472, + "grad_norm": 0.8778411746025085, + "learning_rate": 2.9346769126934715e-06, + "loss": 0.5465, + "step": 7649 + }, + { + "epoch": 2.6860955056179776, + "grad_norm": 0.8958412408828735, + "learning_rate": 2.9342216219134384e-06, + "loss": 0.5955, + "step": 7650 + }, + { + "epoch": 2.686446629213483, + "grad_norm": 0.9176569581031799, + "learning_rate": 2.9337663162833973e-06, + "loss": 0.5949, + "step": 7651 + }, + { + "epoch": 2.686797752808989, + "grad_norm": 0.8973156213760376, + "learning_rate": 2.93331099581892e-06, + "loss": 0.5946, + "step": 7652 + }, + { + "epoch": 2.6871488764044944, + "grad_norm": 0.8997647762298584, + "learning_rate": 2.9328556605355764e-06, + "loss": 0.6413, + "step": 7653 + }, + { + "epoch": 2.6875, + "grad_norm": 0.9028372764587402, + "learning_rate": 2.9324003104489403e-06, + "loss": 0.5901, + "step": 7654 + }, + { + "epoch": 2.6878511235955056, + "grad_norm": 0.9137160778045654, + "learning_rate": 2.931944945574583e-06, + "loss": 0.5803, + "step": 7655 + }, + { + "epoch": 2.688202247191011, + "grad_norm": 0.8920665979385376, + "learning_rate": 2.9314895659280796e-06, + "loss": 0.5838, + "step": 7656 + }, + { + "epoch": 2.688553370786517, + "grad_norm": 0.8954311609268188, + "learning_rate": 2.931034171525001e-06, + "loss": 0.5716, + "step": 7657 + }, + { + "epoch": 2.6889044943820224, + "grad_norm": 0.8939211964607239, + "learning_rate": 2.9305787623809228e-06, + "loss": 0.5819, + "step": 7658 + }, + { + "epoch": 2.689255617977528, + "grad_norm": 0.9514079689979553, + "learning_rate": 2.9301233385114193e-06, + "loss": 0.5843, + "step": 7659 + }, + { + "epoch": 2.6896067415730336, + "grad_norm": 0.8991907835006714, + "learning_rate": 2.929667899932066e-06, + "loss": 0.5756, + "step": 7660 + }, + { + "epoch": 2.689957865168539, + "grad_norm": 0.8960686922073364, + "learning_rate": 2.929212446658438e-06, + "loss": 0.6055, + "step": 7661 + }, + { + "epoch": 2.690308988764045, + "grad_norm": 0.8831461668014526, + "learning_rate": 2.9287569787061115e-06, + "loss": 0.569, + "step": 7662 + }, + { + "epoch": 2.6906601123595504, + "grad_norm": 0.9181011319160461, + "learning_rate": 2.928301496090663e-06, + "loss": 0.6069, + "step": 7663 + }, + { + "epoch": 2.691011235955056, + "grad_norm": 0.9059464931488037, + "learning_rate": 2.92784599882767e-06, + "loss": 0.5978, + "step": 7664 + }, + { + "epoch": 2.6913623595505616, + "grad_norm": 0.910171389579773, + "learning_rate": 2.927390486932711e-06, + "loss": 0.5893, + "step": 7665 + }, + { + "epoch": 2.691713483146067, + "grad_norm": 0.9183464050292969, + "learning_rate": 2.9269349604213617e-06, + "loss": 0.5887, + "step": 7666 + }, + { + "epoch": 2.6920646067415728, + "grad_norm": 0.8932310342788696, + "learning_rate": 2.926479419309203e-06, + "loss": 0.5697, + "step": 7667 + }, + { + "epoch": 2.692415730337079, + "grad_norm": 0.9090089797973633, + "learning_rate": 2.9260238636118126e-06, + "loss": 0.605, + "step": 7668 + }, + { + "epoch": 2.6927668539325844, + "grad_norm": 0.9095858335494995, + "learning_rate": 2.925568293344771e-06, + "loss": 0.5988, + "step": 7669 + }, + { + "epoch": 2.69311797752809, + "grad_norm": 0.9203164577484131, + "learning_rate": 2.9251127085236575e-06, + "loss": 0.6026, + "step": 7670 + }, + { + "epoch": 2.6934691011235956, + "grad_norm": 0.899074912071228, + "learning_rate": 2.924657109164053e-06, + "loss": 0.5251, + "step": 7671 + }, + { + "epoch": 2.693820224719101, + "grad_norm": 0.9205914735794067, + "learning_rate": 2.924201495281539e-06, + "loss": 0.5795, + "step": 7672 + }, + { + "epoch": 2.694171348314607, + "grad_norm": 0.8893402218818665, + "learning_rate": 2.9237458668916968e-06, + "loss": 0.5735, + "step": 7673 + }, + { + "epoch": 2.6945224719101124, + "grad_norm": 0.9128731489181519, + "learning_rate": 2.923290224010109e-06, + "loss": 0.5754, + "step": 7674 + }, + { + "epoch": 2.694873595505618, + "grad_norm": 0.9015269875526428, + "learning_rate": 2.9228345666523573e-06, + "loss": 0.5929, + "step": 7675 + }, + { + "epoch": 2.6952247191011236, + "grad_norm": 0.9074078798294067, + "learning_rate": 2.9223788948340255e-06, + "loss": 0.5997, + "step": 7676 + }, + { + "epoch": 2.695575842696629, + "grad_norm": 0.8818020820617676, + "learning_rate": 2.9219232085706968e-06, + "loss": 0.5543, + "step": 7677 + }, + { + "epoch": 2.695926966292135, + "grad_norm": 0.8831465840339661, + "learning_rate": 2.9214675078779555e-06, + "loss": 0.5668, + "step": 7678 + }, + { + "epoch": 2.6962780898876404, + "grad_norm": 0.925886869430542, + "learning_rate": 2.9210117927713854e-06, + "loss": 0.5569, + "step": 7679 + }, + { + "epoch": 2.696629213483146, + "grad_norm": 0.9230363965034485, + "learning_rate": 2.920556063266573e-06, + "loss": 0.5915, + "step": 7680 + }, + { + "epoch": 2.6969803370786516, + "grad_norm": 0.9074366688728333, + "learning_rate": 2.9201003193791034e-06, + "loss": 0.5748, + "step": 7681 + }, + { + "epoch": 2.697331460674157, + "grad_norm": 0.9124719500541687, + "learning_rate": 2.9196445611245614e-06, + "loss": 0.6254, + "step": 7682 + }, + { + "epoch": 2.697682584269663, + "grad_norm": 0.9043287038803101, + "learning_rate": 2.9191887885185353e-06, + "loss": 0.5983, + "step": 7683 + }, + { + "epoch": 2.6980337078651684, + "grad_norm": 0.9195630550384521, + "learning_rate": 2.918733001576611e-06, + "loss": 0.5735, + "step": 7684 + }, + { + "epoch": 2.6983848314606744, + "grad_norm": 0.9061943292617798, + "learning_rate": 2.9182772003143767e-06, + "loss": 0.5973, + "step": 7685 + }, + { + "epoch": 2.69873595505618, + "grad_norm": 0.9061102271080017, + "learning_rate": 2.9178213847474197e-06, + "loss": 0.5838, + "step": 7686 + }, + { + "epoch": 2.6990870786516856, + "grad_norm": 0.889696478843689, + "learning_rate": 2.917365554891329e-06, + "loss": 0.5695, + "step": 7687 + }, + { + "epoch": 2.699438202247191, + "grad_norm": 0.9346334934234619, + "learning_rate": 2.916909710761693e-06, + "loss": 0.614, + "step": 7688 + }, + { + "epoch": 2.699789325842697, + "grad_norm": 0.9514817595481873, + "learning_rate": 2.916453852374102e-06, + "loss": 0.5991, + "step": 7689 + }, + { + "epoch": 2.7001404494382024, + "grad_norm": 0.9394652843475342, + "learning_rate": 2.9159979797441456e-06, + "loss": 0.6247, + "step": 7690 + }, + { + "epoch": 2.700491573033708, + "grad_norm": 0.8955004811286926, + "learning_rate": 2.9155420928874145e-06, + "loss": 0.5809, + "step": 7691 + }, + { + "epoch": 2.7008426966292136, + "grad_norm": 0.9160822033882141, + "learning_rate": 2.915086191819499e-06, + "loss": 0.5522, + "step": 7692 + }, + { + "epoch": 2.701193820224719, + "grad_norm": 0.8646555542945862, + "learning_rate": 2.914630276555992e-06, + "loss": 0.5602, + "step": 7693 + }, + { + "epoch": 2.701544943820225, + "grad_norm": 0.9061866402626038, + "learning_rate": 2.9141743471124832e-06, + "loss": 0.5816, + "step": 7694 + }, + { + "epoch": 2.7018960674157304, + "grad_norm": 0.8753962516784668, + "learning_rate": 2.9137184035045666e-06, + "loss": 0.5688, + "step": 7695 + }, + { + "epoch": 2.702247191011236, + "grad_norm": 0.9016565680503845, + "learning_rate": 2.913262445747834e-06, + "loss": 0.5708, + "step": 7696 + }, + { + "epoch": 2.7025983146067416, + "grad_norm": 0.8927076458930969, + "learning_rate": 2.9128064738578804e-06, + "loss": 0.5942, + "step": 7697 + }, + { + "epoch": 2.702949438202247, + "grad_norm": 0.8795783519744873, + "learning_rate": 2.9123504878502978e-06, + "loss": 0.5655, + "step": 7698 + }, + { + "epoch": 2.703300561797753, + "grad_norm": 0.9148332476615906, + "learning_rate": 2.9118944877406816e-06, + "loss": 0.5778, + "step": 7699 + }, + { + "epoch": 2.7036516853932584, + "grad_norm": 0.8686633110046387, + "learning_rate": 2.9114384735446264e-06, + "loss": 0.5726, + "step": 7700 + }, + { + "epoch": 2.704002808988764, + "grad_norm": 0.9103491306304932, + "learning_rate": 2.9109824452777285e-06, + "loss": 0.5789, + "step": 7701 + }, + { + "epoch": 2.7043539325842696, + "grad_norm": 0.9307095408439636, + "learning_rate": 2.9105264029555813e-06, + "loss": 0.5817, + "step": 7702 + }, + { + "epoch": 2.704705056179775, + "grad_norm": 0.914969265460968, + "learning_rate": 2.910070346593783e-06, + "loss": 0.5716, + "step": 7703 + }, + { + "epoch": 2.705056179775281, + "grad_norm": 0.911785364151001, + "learning_rate": 2.9096142762079294e-06, + "loss": 0.6222, + "step": 7704 + }, + { + "epoch": 2.7054073033707864, + "grad_norm": 0.8534267544746399, + "learning_rate": 2.9091581918136184e-06, + "loss": 0.528, + "step": 7705 + }, + { + "epoch": 2.705758426966292, + "grad_norm": 0.8922281265258789, + "learning_rate": 2.908702093426447e-06, + "loss": 0.5405, + "step": 7706 + }, + { + "epoch": 2.7061095505617976, + "grad_norm": 0.9002159833908081, + "learning_rate": 2.908245981062014e-06, + "loss": 0.5873, + "step": 7707 + }, + { + "epoch": 2.706460674157303, + "grad_norm": 0.9031289219856262, + "learning_rate": 2.9077898547359185e-06, + "loss": 0.5839, + "step": 7708 + }, + { + "epoch": 2.706811797752809, + "grad_norm": 0.910094678401947, + "learning_rate": 2.907333714463758e-06, + "loss": 0.5653, + "step": 7709 + }, + { + "epoch": 2.7071629213483144, + "grad_norm": 0.892598032951355, + "learning_rate": 2.906877560261133e-06, + "loss": 0.5771, + "step": 7710 + }, + { + "epoch": 2.70751404494382, + "grad_norm": 0.9047536849975586, + "learning_rate": 2.906421392143644e-06, + "loss": 0.5987, + "step": 7711 + }, + { + "epoch": 2.7078651685393256, + "grad_norm": 0.8951240181922913, + "learning_rate": 2.905965210126891e-06, + "loss": 0.6069, + "step": 7712 + }, + { + "epoch": 2.7082162921348316, + "grad_norm": 0.8984569311141968, + "learning_rate": 2.905509014226475e-06, + "loss": 0.5662, + "step": 7713 + }, + { + "epoch": 2.708567415730337, + "grad_norm": 0.9219905734062195, + "learning_rate": 2.9050528044579977e-06, + "loss": 0.617, + "step": 7714 + }, + { + "epoch": 2.708918539325843, + "grad_norm": 0.8862488865852356, + "learning_rate": 2.9045965808370612e-06, + "loss": 0.621, + "step": 7715 + }, + { + "epoch": 2.7092696629213484, + "grad_norm": 0.8901491761207581, + "learning_rate": 2.904140343379268e-06, + "loss": 0.5368, + "step": 7716 + }, + { + "epoch": 2.709620786516854, + "grad_norm": 0.8888766169548035, + "learning_rate": 2.9036840921002207e-06, + "loss": 0.5873, + "step": 7717 + }, + { + "epoch": 2.7099719101123596, + "grad_norm": 0.8900644183158875, + "learning_rate": 2.903227827015523e-06, + "loss": 0.5905, + "step": 7718 + }, + { + "epoch": 2.710323033707865, + "grad_norm": 0.9076018929481506, + "learning_rate": 2.902771548140779e-06, + "loss": 0.5493, + "step": 7719 + }, + { + "epoch": 2.710674157303371, + "grad_norm": 0.9167888164520264, + "learning_rate": 2.902315255491592e-06, + "loss": 0.6093, + "step": 7720 + }, + { + "epoch": 2.7110252808988764, + "grad_norm": 0.8933820128440857, + "learning_rate": 2.901858949083568e-06, + "loss": 0.5635, + "step": 7721 + }, + { + "epoch": 2.711376404494382, + "grad_norm": 0.875125527381897, + "learning_rate": 2.9014026289323115e-06, + "loss": 0.5709, + "step": 7722 + }, + { + "epoch": 2.7117275280898876, + "grad_norm": 0.8973773717880249, + "learning_rate": 2.900946295053429e-06, + "loss": 0.5803, + "step": 7723 + }, + { + "epoch": 2.712078651685393, + "grad_norm": 0.9032551050186157, + "learning_rate": 2.9004899474625262e-06, + "loss": 0.5729, + "step": 7724 + }, + { + "epoch": 2.712429775280899, + "grad_norm": 0.8532813191413879, + "learning_rate": 2.90003358617521e-06, + "loss": 0.5239, + "step": 7725 + }, + { + "epoch": 2.7127808988764044, + "grad_norm": 0.8868680596351624, + "learning_rate": 2.899577211207087e-06, + "loss": 0.5637, + "step": 7726 + }, + { + "epoch": 2.71313202247191, + "grad_norm": 0.9110879898071289, + "learning_rate": 2.8991208225737656e-06, + "loss": 0.585, + "step": 7727 + }, + { + "epoch": 2.7134831460674156, + "grad_norm": 0.943913996219635, + "learning_rate": 2.8986644202908533e-06, + "loss": 0.5729, + "step": 7728 + }, + { + "epoch": 2.713834269662921, + "grad_norm": 0.9030404686927795, + "learning_rate": 2.898208004373959e-06, + "loss": 0.5818, + "step": 7729 + }, + { + "epoch": 2.7141853932584272, + "grad_norm": 0.9079831838607788, + "learning_rate": 2.897751574838692e-06, + "loss": 0.5809, + "step": 7730 + }, + { + "epoch": 2.714536516853933, + "grad_norm": 0.9192582368850708, + "learning_rate": 2.897295131700662e-06, + "loss": 0.6085, + "step": 7731 + }, + { + "epoch": 2.7148876404494384, + "grad_norm": 0.8968272805213928, + "learning_rate": 2.8968386749754777e-06, + "loss": 0.5953, + "step": 7732 + }, + { + "epoch": 2.715238764044944, + "grad_norm": 0.9105080962181091, + "learning_rate": 2.8963822046787505e-06, + "loss": 0.5879, + "step": 7733 + }, + { + "epoch": 2.7155898876404496, + "grad_norm": 0.9092912673950195, + "learning_rate": 2.895925720826091e-06, + "loss": 0.5734, + "step": 7734 + }, + { + "epoch": 2.715941011235955, + "grad_norm": 0.9069133400917053, + "learning_rate": 2.895469223433111e-06, + "loss": 0.5726, + "step": 7735 + }, + { + "epoch": 2.716292134831461, + "grad_norm": 0.9019514918327332, + "learning_rate": 2.8950127125154214e-06, + "loss": 0.571, + "step": 7736 + }, + { + "epoch": 2.7166432584269664, + "grad_norm": 0.873347818851471, + "learning_rate": 2.8945561880886353e-06, + "loss": 0.552, + "step": 7737 + }, + { + "epoch": 2.716994382022472, + "grad_norm": 0.889513373374939, + "learning_rate": 2.8940996501683653e-06, + "loss": 0.607, + "step": 7738 + }, + { + "epoch": 2.7173455056179776, + "grad_norm": 0.8990209102630615, + "learning_rate": 2.893643098770225e-06, + "loss": 0.5652, + "step": 7739 + }, + { + "epoch": 2.717696629213483, + "grad_norm": 0.9095270037651062, + "learning_rate": 2.8931865339098263e-06, + "loss": 0.5806, + "step": 7740 + }, + { + "epoch": 2.718047752808989, + "grad_norm": 0.9184511303901672, + "learning_rate": 2.8927299556027853e-06, + "loss": 0.5613, + "step": 7741 + }, + { + "epoch": 2.7183988764044944, + "grad_norm": 0.9207443594932556, + "learning_rate": 2.892273363864716e-06, + "loss": 0.5862, + "step": 7742 + }, + { + "epoch": 2.71875, + "grad_norm": 0.9110394716262817, + "learning_rate": 2.8918167587112332e-06, + "loss": 0.5981, + "step": 7743 + }, + { + "epoch": 2.7191011235955056, + "grad_norm": 0.9008548855781555, + "learning_rate": 2.891360140157953e-06, + "loss": 0.5965, + "step": 7744 + }, + { + "epoch": 2.719452247191011, + "grad_norm": 0.9168673753738403, + "learning_rate": 2.8909035082204905e-06, + "loss": 0.5767, + "step": 7745 + }, + { + "epoch": 2.719803370786517, + "grad_norm": 0.9254347085952759, + "learning_rate": 2.890446862914462e-06, + "loss": 0.6047, + "step": 7746 + }, + { + "epoch": 2.7201544943820224, + "grad_norm": 0.93023681640625, + "learning_rate": 2.889990204255486e-06, + "loss": 0.5984, + "step": 7747 + }, + { + "epoch": 2.720505617977528, + "grad_norm": 0.9006367921829224, + "learning_rate": 2.889533532259178e-06, + "loss": 0.5672, + "step": 7748 + }, + { + "epoch": 2.7208567415730336, + "grad_norm": 0.8872804045677185, + "learning_rate": 2.889076846941156e-06, + "loss": 0.5475, + "step": 7749 + }, + { + "epoch": 2.721207865168539, + "grad_norm": 0.8753276467323303, + "learning_rate": 2.8886201483170393e-06, + "loss": 0.5918, + "step": 7750 + }, + { + "epoch": 2.721558988764045, + "grad_norm": 0.9104637503623962, + "learning_rate": 2.888163436402446e-06, + "loss": 0.6049, + "step": 7751 + }, + { + "epoch": 2.7219101123595504, + "grad_norm": 0.8875898122787476, + "learning_rate": 2.887706711212996e-06, + "loss": 0.5834, + "step": 7752 + }, + { + "epoch": 2.722261235955056, + "grad_norm": 0.8867804408073425, + "learning_rate": 2.8872499727643076e-06, + "loss": 0.5555, + "step": 7753 + }, + { + "epoch": 2.7226123595505616, + "grad_norm": 0.9041495323181152, + "learning_rate": 2.886793221072002e-06, + "loss": 0.6145, + "step": 7754 + }, + { + "epoch": 2.722963483146067, + "grad_norm": 0.8953424692153931, + "learning_rate": 2.8863364561516984e-06, + "loss": 0.5916, + "step": 7755 + }, + { + "epoch": 2.7233146067415728, + "grad_norm": 0.8621262311935425, + "learning_rate": 2.8858796780190196e-06, + "loss": 0.5382, + "step": 7756 + }, + { + "epoch": 2.723665730337079, + "grad_norm": 0.8919382691383362, + "learning_rate": 2.8854228866895858e-06, + "loss": 0.5497, + "step": 7757 + }, + { + "epoch": 2.7240168539325844, + "grad_norm": 0.9364498853683472, + "learning_rate": 2.884966082179018e-06, + "loss": 0.5766, + "step": 7758 + }, + { + "epoch": 2.72436797752809, + "grad_norm": 0.8953717947006226, + "learning_rate": 2.8845092645029404e-06, + "loss": 0.578, + "step": 7759 + }, + { + "epoch": 2.7247191011235956, + "grad_norm": 0.8985649943351746, + "learning_rate": 2.8840524336769748e-06, + "loss": 0.6112, + "step": 7760 + }, + { + "epoch": 2.725070224719101, + "grad_norm": 0.9003793597221375, + "learning_rate": 2.883595589716745e-06, + "loss": 0.5486, + "step": 7761 + }, + { + "epoch": 2.725421348314607, + "grad_norm": 0.8872453570365906, + "learning_rate": 2.883138732637874e-06, + "loss": 0.5629, + "step": 7762 + }, + { + "epoch": 2.7257724719101124, + "grad_norm": 0.8960056304931641, + "learning_rate": 2.8826818624559862e-06, + "loss": 0.5889, + "step": 7763 + }, + { + "epoch": 2.726123595505618, + "grad_norm": 0.9092264771461487, + "learning_rate": 2.882224979186706e-06, + "loss": 0.5781, + "step": 7764 + }, + { + "epoch": 2.7264747191011236, + "grad_norm": 0.8974511623382568, + "learning_rate": 2.881768082845659e-06, + "loss": 0.5873, + "step": 7765 + }, + { + "epoch": 2.726825842696629, + "grad_norm": 0.9124070405960083, + "learning_rate": 2.88131117344847e-06, + "loss": 0.5815, + "step": 7766 + }, + { + "epoch": 2.727176966292135, + "grad_norm": 0.8972293138504028, + "learning_rate": 2.880854251010765e-06, + "loss": 0.5551, + "step": 7767 + }, + { + "epoch": 2.7275280898876404, + "grad_norm": 0.9020177125930786, + "learning_rate": 2.8803973155481706e-06, + "loss": 0.5884, + "step": 7768 + }, + { + "epoch": 2.727879213483146, + "grad_norm": 0.897101879119873, + "learning_rate": 2.8799403670763144e-06, + "loss": 0.6013, + "step": 7769 + }, + { + "epoch": 2.7282303370786516, + "grad_norm": 0.9154086112976074, + "learning_rate": 2.8794834056108217e-06, + "loss": 0.5538, + "step": 7770 + }, + { + "epoch": 2.728581460674157, + "grad_norm": 0.916377067565918, + "learning_rate": 2.8790264311673218e-06, + "loss": 0.5684, + "step": 7771 + }, + { + "epoch": 2.728932584269663, + "grad_norm": 0.9238152503967285, + "learning_rate": 2.878569443761442e-06, + "loss": 0.6408, + "step": 7772 + }, + { + "epoch": 2.7292837078651684, + "grad_norm": 0.8774369955062866, + "learning_rate": 2.8781124434088115e-06, + "loss": 0.5844, + "step": 7773 + }, + { + "epoch": 2.7296348314606744, + "grad_norm": 0.9383780360221863, + "learning_rate": 2.877655430125058e-06, + "loss": 0.6058, + "step": 7774 + }, + { + "epoch": 2.72998595505618, + "grad_norm": 0.9093977212905884, + "learning_rate": 2.877198403925813e-06, + "loss": 0.5695, + "step": 7775 + }, + { + "epoch": 2.7303370786516856, + "grad_norm": 0.9040725231170654, + "learning_rate": 2.876741364826705e-06, + "loss": 0.5687, + "step": 7776 + }, + { + "epoch": 2.730688202247191, + "grad_norm": 0.9114583730697632, + "learning_rate": 2.876284312843365e-06, + "loss": 0.5849, + "step": 7777 + }, + { + "epoch": 2.731039325842697, + "grad_norm": 0.9144101142883301, + "learning_rate": 2.8758272479914227e-06, + "loss": 0.5897, + "step": 7778 + }, + { + "epoch": 2.7313904494382024, + "grad_norm": 0.8901776671409607, + "learning_rate": 2.8753701702865106e-06, + "loss": 0.5884, + "step": 7779 + }, + { + "epoch": 2.731741573033708, + "grad_norm": 0.8778916597366333, + "learning_rate": 2.8749130797442602e-06, + "loss": 0.559, + "step": 7780 + }, + { + "epoch": 2.7320926966292136, + "grad_norm": 0.8670718669891357, + "learning_rate": 2.874455976380303e-06, + "loss": 0.5749, + "step": 7781 + }, + { + "epoch": 2.732443820224719, + "grad_norm": 0.9012483358383179, + "learning_rate": 2.873998860210272e-06, + "loss": 0.5932, + "step": 7782 + }, + { + "epoch": 2.732794943820225, + "grad_norm": 0.8850282430648804, + "learning_rate": 2.8735417312497997e-06, + "loss": 0.5611, + "step": 7783 + }, + { + "epoch": 2.7331460674157304, + "grad_norm": 0.8992084264755249, + "learning_rate": 2.8730845895145204e-06, + "loss": 0.5939, + "step": 7784 + }, + { + "epoch": 2.733497191011236, + "grad_norm": 0.9050067663192749, + "learning_rate": 2.8726274350200666e-06, + "loss": 0.5564, + "step": 7785 + }, + { + "epoch": 2.7338483146067416, + "grad_norm": 0.9020808339118958, + "learning_rate": 2.872170267782074e-06, + "loss": 0.5749, + "step": 7786 + }, + { + "epoch": 2.734199438202247, + "grad_norm": 0.9171011447906494, + "learning_rate": 2.8717130878161765e-06, + "loss": 0.5691, + "step": 7787 + }, + { + "epoch": 2.734550561797753, + "grad_norm": 0.884613037109375, + "learning_rate": 2.8712558951380098e-06, + "loss": 0.575, + "step": 7788 + }, + { + "epoch": 2.7349016853932584, + "grad_norm": 0.9079249501228333, + "learning_rate": 2.8707986897632088e-06, + "loss": 0.5796, + "step": 7789 + }, + { + "epoch": 2.735252808988764, + "grad_norm": 0.8976254463195801, + "learning_rate": 2.87034147170741e-06, + "loss": 0.5547, + "step": 7790 + }, + { + "epoch": 2.7356039325842696, + "grad_norm": 0.9098071455955505, + "learning_rate": 2.86988424098625e-06, + "loss": 0.5648, + "step": 7791 + }, + { + "epoch": 2.735955056179775, + "grad_norm": 0.8768701553344727, + "learning_rate": 2.8694269976153648e-06, + "loss": 0.588, + "step": 7792 + }, + { + "epoch": 2.736306179775281, + "grad_norm": 0.8875728845596313, + "learning_rate": 2.868969741610393e-06, + "loss": 0.5888, + "step": 7793 + }, + { + "epoch": 2.7366573033707864, + "grad_norm": 0.8963574171066284, + "learning_rate": 2.8685124729869713e-06, + "loss": 0.597, + "step": 7794 + }, + { + "epoch": 2.737008426966292, + "grad_norm": 0.8692088723182678, + "learning_rate": 2.8680551917607392e-06, + "loss": 0.5886, + "step": 7795 + }, + { + "epoch": 2.7373595505617976, + "grad_norm": 0.8887844085693359, + "learning_rate": 2.867597897947334e-06, + "loss": 0.5876, + "step": 7796 + }, + { + "epoch": 2.737710674157303, + "grad_norm": 0.8845783472061157, + "learning_rate": 2.867140591562395e-06, + "loss": 0.5669, + "step": 7797 + }, + { + "epoch": 2.738061797752809, + "grad_norm": 0.886153519153595, + "learning_rate": 2.866683272621562e-06, + "loss": 0.5703, + "step": 7798 + }, + { + "epoch": 2.7384129213483144, + "grad_norm": 0.9220955967903137, + "learning_rate": 2.866225941140475e-06, + "loss": 0.5773, + "step": 7799 + }, + { + "epoch": 2.73876404494382, + "grad_norm": 0.8992103338241577, + "learning_rate": 2.8657685971347744e-06, + "loss": 0.6053, + "step": 7800 + }, + { + "epoch": 2.7391151685393256, + "grad_norm": 0.9232634902000427, + "learning_rate": 2.8653112406201e-06, + "loss": 0.6207, + "step": 7801 + }, + { + "epoch": 2.7394662921348316, + "grad_norm": 0.9084803462028503, + "learning_rate": 2.8648538716120943e-06, + "loss": 0.6069, + "step": 7802 + }, + { + "epoch": 2.739817415730337, + "grad_norm": 0.9057773947715759, + "learning_rate": 2.8643964901263987e-06, + "loss": 0.5915, + "step": 7803 + }, + { + "epoch": 2.740168539325843, + "grad_norm": 0.9210877418518066, + "learning_rate": 2.8639390961786547e-06, + "loss": 0.5752, + "step": 7804 + }, + { + "epoch": 2.7405196629213484, + "grad_norm": 0.8869462013244629, + "learning_rate": 2.8634816897845047e-06, + "loss": 0.5978, + "step": 7805 + }, + { + "epoch": 2.740870786516854, + "grad_norm": 0.9292904734611511, + "learning_rate": 2.8630242709595923e-06, + "loss": 0.6064, + "step": 7806 + }, + { + "epoch": 2.7412219101123596, + "grad_norm": 0.9136069416999817, + "learning_rate": 2.862566839719561e-06, + "loss": 0.5767, + "step": 7807 + }, + { + "epoch": 2.741573033707865, + "grad_norm": 0.9364684820175171, + "learning_rate": 2.8621093960800534e-06, + "loss": 0.6032, + "step": 7808 + }, + { + "epoch": 2.741924157303371, + "grad_norm": 0.9005317687988281, + "learning_rate": 2.8616519400567146e-06, + "loss": 0.596, + "step": 7809 + }, + { + "epoch": 2.7422752808988764, + "grad_norm": 0.9092980623245239, + "learning_rate": 2.8611944716651883e-06, + "loss": 0.5866, + "step": 7810 + }, + { + "epoch": 2.742626404494382, + "grad_norm": 0.8536505103111267, + "learning_rate": 2.86073699092112e-06, + "loss": 0.5017, + "step": 7811 + }, + { + "epoch": 2.7429775280898876, + "grad_norm": 0.9143251180648804, + "learning_rate": 2.860279497840156e-06, + "loss": 0.5758, + "step": 7812 + }, + { + "epoch": 2.743328651685393, + "grad_norm": 0.9093949794769287, + "learning_rate": 2.8598219924379417e-06, + "loss": 0.5816, + "step": 7813 + }, + { + "epoch": 2.743679775280899, + "grad_norm": 0.9022784233093262, + "learning_rate": 2.8593644747301223e-06, + "loss": 0.5956, + "step": 7814 + }, + { + "epoch": 2.7440308988764044, + "grad_norm": 0.9230907559394836, + "learning_rate": 2.858906944732346e-06, + "loss": 0.5667, + "step": 7815 + }, + { + "epoch": 2.74438202247191, + "grad_norm": 0.9062755703926086, + "learning_rate": 2.8584494024602594e-06, + "loss": 0.5825, + "step": 7816 + }, + { + "epoch": 2.7447331460674156, + "grad_norm": 0.9259464144706726, + "learning_rate": 2.857991847929509e-06, + "loss": 0.5898, + "step": 7817 + }, + { + "epoch": 2.745084269662921, + "grad_norm": 0.9404741525650024, + "learning_rate": 2.857534281155745e-06, + "loss": 0.5936, + "step": 7818 + }, + { + "epoch": 2.7454353932584272, + "grad_norm": 0.8923565149307251, + "learning_rate": 2.857076702154614e-06, + "loss": 0.5866, + "step": 7819 + }, + { + "epoch": 2.745786516853933, + "grad_norm": 0.9064624905586243, + "learning_rate": 2.8566191109417655e-06, + "loss": 0.5727, + "step": 7820 + }, + { + "epoch": 2.7461376404494384, + "grad_norm": 0.9221748113632202, + "learning_rate": 2.8561615075328486e-06, + "loss": 0.5508, + "step": 7821 + }, + { + "epoch": 2.746488764044944, + "grad_norm": 0.9328464269638062, + "learning_rate": 2.8557038919435124e-06, + "loss": 0.5881, + "step": 7822 + }, + { + "epoch": 2.7468398876404496, + "grad_norm": 0.8958388566970825, + "learning_rate": 2.855246264189408e-06, + "loss": 0.5779, + "step": 7823 + }, + { + "epoch": 2.747191011235955, + "grad_norm": 0.9882782101631165, + "learning_rate": 2.854788624286185e-06, + "loss": 0.645, + "step": 7824 + }, + { + "epoch": 2.747542134831461, + "grad_norm": 0.9169550538063049, + "learning_rate": 2.854330972249495e-06, + "loss": 0.5809, + "step": 7825 + }, + { + "epoch": 2.7478932584269664, + "grad_norm": 0.8962950110435486, + "learning_rate": 2.853873308094989e-06, + "loss": 0.613, + "step": 7826 + }, + { + "epoch": 2.748244382022472, + "grad_norm": 0.9145513772964478, + "learning_rate": 2.8534156318383175e-06, + "loss": 0.5652, + "step": 7827 + }, + { + "epoch": 2.7485955056179776, + "grad_norm": 0.8974515795707703, + "learning_rate": 2.852957943495135e-06, + "loss": 0.5668, + "step": 7828 + }, + { + "epoch": 2.748946629213483, + "grad_norm": 0.9432291388511658, + "learning_rate": 2.8525002430810928e-06, + "loss": 0.5686, + "step": 7829 + }, + { + "epoch": 2.749297752808989, + "grad_norm": 0.9684997797012329, + "learning_rate": 2.852042530611844e-06, + "loss": 0.5774, + "step": 7830 + }, + { + "epoch": 2.7496488764044944, + "grad_norm": 0.9242194294929504, + "learning_rate": 2.851584806103041e-06, + "loss": 0.5674, + "step": 7831 + }, + { + "epoch": 2.75, + "grad_norm": 0.919677197933197, + "learning_rate": 2.85112706957034e-06, + "loss": 0.5882, + "step": 7832 + }, + { + "epoch": 2.7503511235955056, + "grad_norm": 0.8811525702476501, + "learning_rate": 2.850669321029392e-06, + "loss": 0.5066, + "step": 7833 + }, + { + "epoch": 2.750702247191011, + "grad_norm": 0.9169440269470215, + "learning_rate": 2.8502115604958546e-06, + "loss": 0.5929, + "step": 7834 + }, + { + "epoch": 2.751053370786517, + "grad_norm": 0.9150686264038086, + "learning_rate": 2.849753787985381e-06, + "loss": 0.6077, + "step": 7835 + }, + { + "epoch": 2.7514044943820224, + "grad_norm": 0.9467943906784058, + "learning_rate": 2.8492960035136268e-06, + "loss": 0.6047, + "step": 7836 + }, + { + "epoch": 2.751755617977528, + "grad_norm": 0.8883975148200989, + "learning_rate": 2.8488382070962487e-06, + "loss": 0.5919, + "step": 7837 + }, + { + "epoch": 2.7521067415730336, + "grad_norm": 0.8869690299034119, + "learning_rate": 2.8483803987489024e-06, + "loss": 0.5955, + "step": 7838 + }, + { + "epoch": 2.752457865168539, + "grad_norm": 0.9173586368560791, + "learning_rate": 2.8479225784872445e-06, + "loss": 0.555, + "step": 7839 + }, + { + "epoch": 2.752808988764045, + "grad_norm": 0.8764431476593018, + "learning_rate": 2.8474647463269315e-06, + "loss": 0.5592, + "step": 7840 + }, + { + "epoch": 2.7531601123595504, + "grad_norm": 0.8995755910873413, + "learning_rate": 2.847006902283622e-06, + "loss": 0.6178, + "step": 7841 + }, + { + "epoch": 2.753511235955056, + "grad_norm": 0.8902384042739868, + "learning_rate": 2.846549046372974e-06, + "loss": 0.5295, + "step": 7842 + }, + { + "epoch": 2.7538623595505616, + "grad_norm": 0.9841292500495911, + "learning_rate": 2.8460911786106444e-06, + "loss": 0.6188, + "step": 7843 + }, + { + "epoch": 2.754213483146067, + "grad_norm": 0.9201949834823608, + "learning_rate": 2.8456332990122927e-06, + "loss": 0.5905, + "step": 7844 + }, + { + "epoch": 2.7545646067415728, + "grad_norm": 0.9177679419517517, + "learning_rate": 2.8451754075935784e-06, + "loss": 0.5706, + "step": 7845 + }, + { + "epoch": 2.754915730337079, + "grad_norm": 0.9101627469062805, + "learning_rate": 2.84471750437016e-06, + "loss": 0.6129, + "step": 7846 + }, + { + "epoch": 2.7552668539325844, + "grad_norm": 0.9324325323104858, + "learning_rate": 2.8442595893576985e-06, + "loss": 0.602, + "step": 7847 + }, + { + "epoch": 2.75561797752809, + "grad_norm": 0.9207699298858643, + "learning_rate": 2.8438016625718528e-06, + "loss": 0.5734, + "step": 7848 + }, + { + "epoch": 2.7559691011235956, + "grad_norm": 0.9031668305397034, + "learning_rate": 2.843343724028285e-06, + "loss": 0.6036, + "step": 7849 + }, + { + "epoch": 2.756320224719101, + "grad_norm": 0.9285680651664734, + "learning_rate": 2.842885773742656e-06, + "loss": 0.5928, + "step": 7850 + }, + { + "epoch": 2.756671348314607, + "grad_norm": 0.8884228467941284, + "learning_rate": 2.842427811730626e-06, + "loss": 0.5791, + "step": 7851 + }, + { + "epoch": 2.7570224719101124, + "grad_norm": 0.9234258532524109, + "learning_rate": 2.8419698380078585e-06, + "loss": 0.6151, + "step": 7852 + }, + { + "epoch": 2.757373595505618, + "grad_norm": 0.8608604073524475, + "learning_rate": 2.841511852590015e-06, + "loss": 0.5728, + "step": 7853 + }, + { + "epoch": 2.7577247191011236, + "grad_norm": 0.9165059328079224, + "learning_rate": 2.841053855492758e-06, + "loss": 0.5796, + "step": 7854 + }, + { + "epoch": 2.758075842696629, + "grad_norm": 0.8790934085845947, + "learning_rate": 2.8405958467317517e-06, + "loss": 0.5977, + "step": 7855 + }, + { + "epoch": 2.758426966292135, + "grad_norm": 0.8862566947937012, + "learning_rate": 2.8401378263226585e-06, + "loss": 0.5525, + "step": 7856 + }, + { + "epoch": 2.7587780898876404, + "grad_norm": 0.8492090702056885, + "learning_rate": 2.8396797942811432e-06, + "loss": 0.5627, + "step": 7857 + }, + { + "epoch": 2.759129213483146, + "grad_norm": 0.8960773348808289, + "learning_rate": 2.8392217506228693e-06, + "loss": 0.583, + "step": 7858 + }, + { + "epoch": 2.7594803370786516, + "grad_norm": 0.9028218388557434, + "learning_rate": 2.8387636953635016e-06, + "loss": 0.5656, + "step": 7859 + }, + { + "epoch": 2.759831460674157, + "grad_norm": 0.8809931874275208, + "learning_rate": 2.8383056285187057e-06, + "loss": 0.5538, + "step": 7860 + }, + { + "epoch": 2.760182584269663, + "grad_norm": 0.9050323963165283, + "learning_rate": 2.837847550104146e-06, + "loss": 0.5874, + "step": 7861 + }, + { + "epoch": 2.7605337078651684, + "grad_norm": 0.9089576005935669, + "learning_rate": 2.83738946013549e-06, + "loss": 0.5723, + "step": 7862 + }, + { + "epoch": 2.7608848314606744, + "grad_norm": 0.9302275776863098, + "learning_rate": 2.836931358628403e-06, + "loss": 0.5748, + "step": 7863 + }, + { + "epoch": 2.76123595505618, + "grad_norm": 0.9302178621292114, + "learning_rate": 2.8364732455985517e-06, + "loss": 0.5692, + "step": 7864 + }, + { + "epoch": 2.7615870786516856, + "grad_norm": 0.8986315727233887, + "learning_rate": 2.836015121061603e-06, + "loss": 0.6024, + "step": 7865 + }, + { + "epoch": 2.761938202247191, + "grad_norm": 0.8517990708351135, + "learning_rate": 2.8355569850332254e-06, + "loss": 0.5705, + "step": 7866 + }, + { + "epoch": 2.762289325842697, + "grad_norm": 0.9088213443756104, + "learning_rate": 2.835098837529086e-06, + "loss": 0.5464, + "step": 7867 + }, + { + "epoch": 2.7626404494382024, + "grad_norm": 0.8970953226089478, + "learning_rate": 2.8346406785648527e-06, + "loss": 0.5704, + "step": 7868 + }, + { + "epoch": 2.762991573033708, + "grad_norm": 0.9165033102035522, + "learning_rate": 2.8341825081561954e-06, + "loss": 0.6051, + "step": 7869 + }, + { + "epoch": 2.7633426966292136, + "grad_norm": 0.9462375640869141, + "learning_rate": 2.8337243263187813e-06, + "loss": 0.5585, + "step": 7870 + }, + { + "epoch": 2.763693820224719, + "grad_norm": 0.8804163932800293, + "learning_rate": 2.833266133068281e-06, + "loss": 0.5647, + "step": 7871 + }, + { + "epoch": 2.764044943820225, + "grad_norm": 0.8424877524375916, + "learning_rate": 2.832807928420364e-06, + "loss": 0.5204, + "step": 7872 + }, + { + "epoch": 2.7643960674157304, + "grad_norm": 0.8887908458709717, + "learning_rate": 2.832349712390701e-06, + "loss": 0.5843, + "step": 7873 + }, + { + "epoch": 2.764747191011236, + "grad_norm": 0.9151390194892883, + "learning_rate": 2.8318914849949617e-06, + "loss": 0.6007, + "step": 7874 + }, + { + "epoch": 2.7650983146067416, + "grad_norm": 0.9235409498214722, + "learning_rate": 2.831433246248819e-06, + "loss": 0.603, + "step": 7875 + }, + { + "epoch": 2.765449438202247, + "grad_norm": 0.8901816606521606, + "learning_rate": 2.8309749961679417e-06, + "loss": 0.548, + "step": 7876 + }, + { + "epoch": 2.765800561797753, + "grad_norm": 0.9216820001602173, + "learning_rate": 2.830516734768003e-06, + "loss": 0.5777, + "step": 7877 + }, + { + "epoch": 2.7661516853932584, + "grad_norm": 0.8966969847679138, + "learning_rate": 2.830058462064675e-06, + "loss": 0.6098, + "step": 7878 + }, + { + "epoch": 2.766502808988764, + "grad_norm": 0.8904604911804199, + "learning_rate": 2.82960017807363e-06, + "loss": 0.5938, + "step": 7879 + }, + { + "epoch": 2.7668539325842696, + "grad_norm": 0.8935307264328003, + "learning_rate": 2.8291418828105405e-06, + "loss": 0.5839, + "step": 7880 + }, + { + "epoch": 2.767205056179775, + "grad_norm": 0.8961401581764221, + "learning_rate": 2.82868357629108e-06, + "loss": 0.5887, + "step": 7881 + }, + { + "epoch": 2.767556179775281, + "grad_norm": 0.8781768083572388, + "learning_rate": 2.828225258530924e-06, + "loss": 0.5643, + "step": 7882 + }, + { + "epoch": 2.7679073033707864, + "grad_norm": 0.8693135976791382, + "learning_rate": 2.827766929545744e-06, + "loss": 0.5971, + "step": 7883 + }, + { + "epoch": 2.768258426966292, + "grad_norm": 0.9095543622970581, + "learning_rate": 2.827308589351215e-06, + "loss": 0.5745, + "step": 7884 + }, + { + "epoch": 2.7686095505617976, + "grad_norm": 0.8940972089767456, + "learning_rate": 2.826850237963013e-06, + "loss": 0.5704, + "step": 7885 + }, + { + "epoch": 2.768960674157303, + "grad_norm": 0.8901918530464172, + "learning_rate": 2.826391875396812e-06, + "loss": 0.5743, + "step": 7886 + }, + { + "epoch": 2.769311797752809, + "grad_norm": 0.9287811517715454, + "learning_rate": 2.825933501668289e-06, + "loss": 0.5732, + "step": 7887 + }, + { + "epoch": 2.7696629213483144, + "grad_norm": 0.9934301376342773, + "learning_rate": 2.8254751167931183e-06, + "loss": 0.5896, + "step": 7888 + }, + { + "epoch": 2.77001404494382, + "grad_norm": 0.8969842791557312, + "learning_rate": 2.825016720786977e-06, + "loss": 0.5684, + "step": 7889 + }, + { + "epoch": 2.7703651685393256, + "grad_norm": 0.9056935906410217, + "learning_rate": 2.824558313665543e-06, + "loss": 0.5882, + "step": 7890 + }, + { + "epoch": 2.7707162921348316, + "grad_norm": 0.8731392621994019, + "learning_rate": 2.8240998954444915e-06, + "loss": 0.5516, + "step": 7891 + }, + { + "epoch": 2.771067415730337, + "grad_norm": 0.9082303047180176, + "learning_rate": 2.823641466139501e-06, + "loss": 0.5636, + "step": 7892 + }, + { + "epoch": 2.771418539325843, + "grad_norm": 0.9359666109085083, + "learning_rate": 2.8231830257662502e-06, + "loss": 0.6009, + "step": 7893 + }, + { + "epoch": 2.7717696629213484, + "grad_norm": 0.8763213157653809, + "learning_rate": 2.8227245743404154e-06, + "loss": 0.6063, + "step": 7894 + }, + { + "epoch": 2.772120786516854, + "grad_norm": 0.9016227126121521, + "learning_rate": 2.8222661118776774e-06, + "loss": 0.5737, + "step": 7895 + }, + { + "epoch": 2.7724719101123596, + "grad_norm": 0.8880155086517334, + "learning_rate": 2.821807638393713e-06, + "loss": 0.5667, + "step": 7896 + }, + { + "epoch": 2.772823033707865, + "grad_norm": 0.8975705504417419, + "learning_rate": 2.8213491539042037e-06, + "loss": 0.5829, + "step": 7897 + }, + { + "epoch": 2.773174157303371, + "grad_norm": 0.8855388164520264, + "learning_rate": 2.8208906584248274e-06, + "loss": 0.5691, + "step": 7898 + }, + { + "epoch": 2.7735252808988764, + "grad_norm": 0.9102710485458374, + "learning_rate": 2.8204321519712656e-06, + "loss": 0.5917, + "step": 7899 + }, + { + "epoch": 2.773876404494382, + "grad_norm": 0.9258704781532288, + "learning_rate": 2.819973634559199e-06, + "loss": 0.6091, + "step": 7900 + }, + { + "epoch": 2.7742275280898876, + "grad_norm": 0.8828044533729553, + "learning_rate": 2.8195151062043067e-06, + "loss": 0.555, + "step": 7901 + }, + { + "epoch": 2.774578651685393, + "grad_norm": 0.9081563949584961, + "learning_rate": 2.819056566922272e-06, + "loss": 0.5774, + "step": 7902 + }, + { + "epoch": 2.774929775280899, + "grad_norm": 0.8960114121437073, + "learning_rate": 2.8185980167287757e-06, + "loss": 0.5791, + "step": 7903 + }, + { + "epoch": 2.7752808988764044, + "grad_norm": 0.8993722200393677, + "learning_rate": 2.818139455639499e-06, + "loss": 0.5458, + "step": 7904 + }, + { + "epoch": 2.77563202247191, + "grad_norm": 0.8939298987388611, + "learning_rate": 2.8176808836701262e-06, + "loss": 0.5509, + "step": 7905 + }, + { + "epoch": 2.7759831460674156, + "grad_norm": 0.9166505932807922, + "learning_rate": 2.8172223008363386e-06, + "loss": 0.549, + "step": 7906 + }, + { + "epoch": 2.776334269662921, + "grad_norm": 0.9310207366943359, + "learning_rate": 2.81676370715382e-06, + "loss": 0.5984, + "step": 7907 + }, + { + "epoch": 2.7766853932584272, + "grad_norm": 0.9218758344650269, + "learning_rate": 2.816305102638254e-06, + "loss": 0.5991, + "step": 7908 + }, + { + "epoch": 2.777036516853933, + "grad_norm": 0.889301061630249, + "learning_rate": 2.8158464873053236e-06, + "loss": 0.5721, + "step": 7909 + }, + { + "epoch": 2.7773876404494384, + "grad_norm": 0.8902767896652222, + "learning_rate": 2.8153878611707137e-06, + "loss": 0.5929, + "step": 7910 + }, + { + "epoch": 2.777738764044944, + "grad_norm": 0.8903549909591675, + "learning_rate": 2.8149292242501086e-06, + "loss": 0.5805, + "step": 7911 + }, + { + "epoch": 2.7780898876404496, + "grad_norm": 0.9385935664176941, + "learning_rate": 2.814470576559194e-06, + "loss": 0.6008, + "step": 7912 + }, + { + "epoch": 2.778441011235955, + "grad_norm": 0.8879544138908386, + "learning_rate": 2.8140119181136544e-06, + "loss": 0.5895, + "step": 7913 + }, + { + "epoch": 2.778792134831461, + "grad_norm": 0.9132834076881409, + "learning_rate": 2.8135532489291755e-06, + "loss": 0.577, + "step": 7914 + }, + { + "epoch": 2.7791432584269664, + "grad_norm": 0.8963463306427002, + "learning_rate": 2.8130945690214447e-06, + "loss": 0.569, + "step": 7915 + }, + { + "epoch": 2.779494382022472, + "grad_norm": 0.8845507502555847, + "learning_rate": 2.812635878406147e-06, + "loss": 0.5755, + "step": 7916 + }, + { + "epoch": 2.7798455056179776, + "grad_norm": 0.8660081028938293, + "learning_rate": 2.81217717709897e-06, + "loss": 0.5569, + "step": 7917 + }, + { + "epoch": 2.780196629213483, + "grad_norm": 0.8760690093040466, + "learning_rate": 2.8117184651156006e-06, + "loss": 0.5422, + "step": 7918 + }, + { + "epoch": 2.780547752808989, + "grad_norm": 0.9052375555038452, + "learning_rate": 2.8112597424717266e-06, + "loss": 0.5934, + "step": 7919 + }, + { + "epoch": 2.7808988764044944, + "grad_norm": 0.936708390712738, + "learning_rate": 2.8108010091830358e-06, + "loss": 0.6088, + "step": 7920 + }, + { + "epoch": 2.78125, + "grad_norm": 0.9265839457511902, + "learning_rate": 2.810342265265216e-06, + "loss": 0.5796, + "step": 7921 + }, + { + "epoch": 2.7816011235955056, + "grad_norm": 0.9283183217048645, + "learning_rate": 2.8098835107339567e-06, + "loss": 0.575, + "step": 7922 + }, + { + "epoch": 2.781952247191011, + "grad_norm": 0.939897358417511, + "learning_rate": 2.8094247456049463e-06, + "loss": 0.5877, + "step": 7923 + }, + { + "epoch": 2.782303370786517, + "grad_norm": 0.8957794904708862, + "learning_rate": 2.808965969893874e-06, + "loss": 0.5622, + "step": 7924 + }, + { + "epoch": 2.7826544943820224, + "grad_norm": 0.925071656703949, + "learning_rate": 2.8085071836164305e-06, + "loss": 0.5925, + "step": 7925 + }, + { + "epoch": 2.783005617977528, + "grad_norm": 0.8879451751708984, + "learning_rate": 2.808048386788305e-06, + "loss": 0.5533, + "step": 7926 + }, + { + "epoch": 2.7833567415730336, + "grad_norm": 0.8894330263137817, + "learning_rate": 2.8075895794251885e-06, + "loss": 0.5928, + "step": 7927 + }, + { + "epoch": 2.783707865168539, + "grad_norm": 1.000893235206604, + "learning_rate": 2.807130761542772e-06, + "loss": 0.603, + "step": 7928 + }, + { + "epoch": 2.784058988764045, + "grad_norm": 0.9041726589202881, + "learning_rate": 2.8066719331567455e-06, + "loss": 0.5769, + "step": 7929 + }, + { + "epoch": 2.7844101123595504, + "grad_norm": 0.8769118785858154, + "learning_rate": 2.8062130942828016e-06, + "loss": 0.5833, + "step": 7930 + }, + { + "epoch": 2.784761235955056, + "grad_norm": 0.8974519371986389, + "learning_rate": 2.8057542449366316e-06, + "loss": 0.5822, + "step": 7931 + }, + { + "epoch": 2.7851123595505616, + "grad_norm": 1.0338870286941528, + "learning_rate": 2.8052953851339283e-06, + "loss": 0.585, + "step": 7932 + }, + { + "epoch": 2.785463483146067, + "grad_norm": 0.8823001384735107, + "learning_rate": 2.804836514890385e-06, + "loss": 0.5348, + "step": 7933 + }, + { + "epoch": 2.7858146067415728, + "grad_norm": 0.930145263671875, + "learning_rate": 2.804377634221693e-06, + "loss": 0.6001, + "step": 7934 + }, + { + "epoch": 2.786165730337079, + "grad_norm": 0.9309945702552795, + "learning_rate": 2.803918743143546e-06, + "loss": 0.5706, + "step": 7935 + }, + { + "epoch": 2.7865168539325844, + "grad_norm": 0.939356803894043, + "learning_rate": 2.803459841671638e-06, + "loss": 0.5847, + "step": 7936 + }, + { + "epoch": 2.78686797752809, + "grad_norm": 0.9293814897537231, + "learning_rate": 2.803000929821664e-06, + "loss": 0.5671, + "step": 7937 + }, + { + "epoch": 2.7872191011235956, + "grad_norm": 0.8978012204170227, + "learning_rate": 2.8025420076093164e-06, + "loss": 0.5831, + "step": 7938 + }, + { + "epoch": 2.787570224719101, + "grad_norm": 0.9063260555267334, + "learning_rate": 2.8020830750502913e-06, + "loss": 0.5324, + "step": 7939 + }, + { + "epoch": 2.787921348314607, + "grad_norm": 0.8795073628425598, + "learning_rate": 2.801624132160284e-06, + "loss": 0.582, + "step": 7940 + }, + { + "epoch": 2.7882724719101124, + "grad_norm": 0.8699506521224976, + "learning_rate": 2.8011651789549886e-06, + "loss": 0.5816, + "step": 7941 + }, + { + "epoch": 2.788623595505618, + "grad_norm": 0.9146275520324707, + "learning_rate": 2.8007062154501026e-06, + "loss": 0.5747, + "step": 7942 + }, + { + "epoch": 2.7889747191011236, + "grad_norm": 0.8866508603096008, + "learning_rate": 2.8002472416613214e-06, + "loss": 0.5662, + "step": 7943 + }, + { + "epoch": 2.789325842696629, + "grad_norm": 0.9378337860107422, + "learning_rate": 2.799788257604341e-06, + "loss": 0.6101, + "step": 7944 + }, + { + "epoch": 2.789676966292135, + "grad_norm": 0.8856747150421143, + "learning_rate": 2.7993292632948594e-06, + "loss": 0.5567, + "step": 7945 + }, + { + "epoch": 2.7900280898876404, + "grad_norm": 0.9383493065834045, + "learning_rate": 2.798870258748573e-06, + "loss": 0.5992, + "step": 7946 + }, + { + "epoch": 2.790379213483146, + "grad_norm": 0.9284090399742126, + "learning_rate": 2.798411243981179e-06, + "loss": 0.6065, + "step": 7947 + }, + { + "epoch": 2.7907303370786516, + "grad_norm": 0.9114540815353394, + "learning_rate": 2.7979522190083756e-06, + "loss": 0.5939, + "step": 7948 + }, + { + "epoch": 2.791081460674157, + "grad_norm": 0.8957846760749817, + "learning_rate": 2.7974931838458618e-06, + "loss": 0.5597, + "step": 7949 + }, + { + "epoch": 2.791432584269663, + "grad_norm": 0.9297586679458618, + "learning_rate": 2.7970341385093353e-06, + "loss": 0.6023, + "step": 7950 + }, + { + "epoch": 2.7917837078651684, + "grad_norm": 0.9461788535118103, + "learning_rate": 2.7965750830144957e-06, + "loss": 0.6018, + "step": 7951 + }, + { + "epoch": 2.7921348314606744, + "grad_norm": 0.9076969623565674, + "learning_rate": 2.7961160173770417e-06, + "loss": 0.5714, + "step": 7952 + }, + { + "epoch": 2.79248595505618, + "grad_norm": 0.9167848825454712, + "learning_rate": 2.795656941612674e-06, + "loss": 0.5797, + "step": 7953 + }, + { + "epoch": 2.7928370786516856, + "grad_norm": 0.8906055092811584, + "learning_rate": 2.7951978557370918e-06, + "loss": 0.5594, + "step": 7954 + }, + { + "epoch": 2.793188202247191, + "grad_norm": 0.9148594737052917, + "learning_rate": 2.7947387597659954e-06, + "loss": 0.5823, + "step": 7955 + }, + { + "epoch": 2.793539325842697, + "grad_norm": 0.9462953209877014, + "learning_rate": 2.794279653715085e-06, + "loss": 0.6038, + "step": 7956 + }, + { + "epoch": 2.7938904494382024, + "grad_norm": 0.8844831585884094, + "learning_rate": 2.7938205376000632e-06, + "loss": 0.584, + "step": 7957 + }, + { + "epoch": 2.794241573033708, + "grad_norm": 0.9542579650878906, + "learning_rate": 2.79336141143663e-06, + "loss": 0.5832, + "step": 7958 + }, + { + "epoch": 2.7945926966292136, + "grad_norm": 0.9120001792907715, + "learning_rate": 2.7929022752404873e-06, + "loss": 0.6023, + "step": 7959 + }, + { + "epoch": 2.794943820224719, + "grad_norm": 0.8945369720458984, + "learning_rate": 2.7924431290273378e-06, + "loss": 0.6109, + "step": 7960 + }, + { + "epoch": 2.795294943820225, + "grad_norm": 0.8717198967933655, + "learning_rate": 2.7919839728128835e-06, + "loss": 0.5686, + "step": 7961 + }, + { + "epoch": 2.7956460674157304, + "grad_norm": 0.9457799792289734, + "learning_rate": 2.7915248066128273e-06, + "loss": 0.5963, + "step": 7962 + }, + { + "epoch": 2.795997191011236, + "grad_norm": 0.9040877819061279, + "learning_rate": 2.791065630442872e-06, + "loss": 0.6075, + "step": 7963 + }, + { + "epoch": 2.7963483146067416, + "grad_norm": 0.8823800683021545, + "learning_rate": 2.790606444318721e-06, + "loss": 0.5616, + "step": 7964 + }, + { + "epoch": 2.796699438202247, + "grad_norm": 0.879108190536499, + "learning_rate": 2.790147248256079e-06, + "loss": 0.5594, + "step": 7965 + }, + { + "epoch": 2.797050561797753, + "grad_norm": 0.9170940518379211, + "learning_rate": 2.7896880422706484e-06, + "loss": 0.5853, + "step": 7966 + }, + { + "epoch": 2.7974016853932584, + "grad_norm": 0.9105141758918762, + "learning_rate": 2.7892288263781346e-06, + "loss": 0.561, + "step": 7967 + }, + { + "epoch": 2.797752808988764, + "grad_norm": 0.9200360178947449, + "learning_rate": 2.7887696005942433e-06, + "loss": 0.5923, + "step": 7968 + }, + { + "epoch": 2.7981039325842696, + "grad_norm": 0.8971552848815918, + "learning_rate": 2.7883103649346784e-06, + "loss": 0.5689, + "step": 7969 + }, + { + "epoch": 2.798455056179775, + "grad_norm": 0.9354571104049683, + "learning_rate": 2.787851119415145e-06, + "loss": 0.5614, + "step": 7970 + }, + { + "epoch": 2.798806179775281, + "grad_norm": 0.9274175763130188, + "learning_rate": 2.7873918640513513e-06, + "loss": 0.587, + "step": 7971 + }, + { + "epoch": 2.7991573033707864, + "grad_norm": 0.8968660831451416, + "learning_rate": 2.7869325988590005e-06, + "loss": 0.5501, + "step": 7972 + }, + { + "epoch": 2.799508426966292, + "grad_norm": 0.9124216437339783, + "learning_rate": 2.786473323853801e-06, + "loss": 0.5921, + "step": 7973 + }, + { + "epoch": 2.7998595505617976, + "grad_norm": 0.9251074194908142, + "learning_rate": 2.7860140390514584e-06, + "loss": 0.5889, + "step": 7974 + }, + { + "epoch": 2.800210674157303, + "grad_norm": 0.8539794683456421, + "learning_rate": 2.7855547444676806e-06, + "loss": 0.552, + "step": 7975 + }, + { + "epoch": 2.800561797752809, + "grad_norm": 0.8544564247131348, + "learning_rate": 2.7850954401181753e-06, + "loss": 0.5418, + "step": 7976 + }, + { + "epoch": 2.8009129213483144, + "grad_norm": 0.8686181902885437, + "learning_rate": 2.7846361260186495e-06, + "loss": 0.5418, + "step": 7977 + }, + { + "epoch": 2.80126404494382, + "grad_norm": 0.9212062358856201, + "learning_rate": 2.7841768021848115e-06, + "loss": 0.6022, + "step": 7978 + }, + { + "epoch": 2.8016151685393256, + "grad_norm": 0.9245893359184265, + "learning_rate": 2.78371746863237e-06, + "loss": 0.5958, + "step": 7979 + }, + { + "epoch": 2.8019662921348316, + "grad_norm": 0.8693844079971313, + "learning_rate": 2.783258125377035e-06, + "loss": 0.5482, + "step": 7980 + }, + { + "epoch": 2.802317415730337, + "grad_norm": 0.901725172996521, + "learning_rate": 2.7827987724345127e-06, + "loss": 0.6031, + "step": 7981 + }, + { + "epoch": 2.802668539325843, + "grad_norm": 0.8873985409736633, + "learning_rate": 2.782339409820516e-06, + "loss": 0.5761, + "step": 7982 + }, + { + "epoch": 2.8030196629213484, + "grad_norm": 0.901635468006134, + "learning_rate": 2.7818800375507526e-06, + "loss": 0.5782, + "step": 7983 + }, + { + "epoch": 2.803370786516854, + "grad_norm": 0.8788841366767883, + "learning_rate": 2.7814206556409327e-06, + "loss": 0.5839, + "step": 7984 + }, + { + "epoch": 2.8037219101123596, + "grad_norm": 0.8784237504005432, + "learning_rate": 2.780961264106768e-06, + "loss": 0.5492, + "step": 7985 + }, + { + "epoch": 2.804073033707865, + "grad_norm": 0.9009540677070618, + "learning_rate": 2.780501862963967e-06, + "loss": 0.6146, + "step": 7986 + }, + { + "epoch": 2.804424157303371, + "grad_norm": 0.9254754781723022, + "learning_rate": 2.780042452228244e-06, + "loss": 0.6106, + "step": 7987 + }, + { + "epoch": 2.8047752808988764, + "grad_norm": 0.9444158673286438, + "learning_rate": 2.7795830319153076e-06, + "loss": 0.6148, + "step": 7988 + }, + { + "epoch": 2.805126404494382, + "grad_norm": 0.8996027708053589, + "learning_rate": 2.779123602040871e-06, + "loss": 0.562, + "step": 7989 + }, + { + "epoch": 2.8054775280898876, + "grad_norm": 0.8970121741294861, + "learning_rate": 2.778664162620647e-06, + "loss": 0.5973, + "step": 7990 + }, + { + "epoch": 2.805828651685393, + "grad_norm": 0.8975568413734436, + "learning_rate": 2.778204713670346e-06, + "loss": 0.5532, + "step": 7991 + }, + { + "epoch": 2.806179775280899, + "grad_norm": 0.9298721551895142, + "learning_rate": 2.777745255205682e-06, + "loss": 0.5879, + "step": 7992 + }, + { + "epoch": 2.8065308988764044, + "grad_norm": 0.9015418887138367, + "learning_rate": 2.777285787242368e-06, + "loss": 0.5732, + "step": 7993 + }, + { + "epoch": 2.80688202247191, + "grad_norm": 0.9223244786262512, + "learning_rate": 2.776826309796118e-06, + "loss": 0.5586, + "step": 7994 + }, + { + "epoch": 2.8072331460674156, + "grad_norm": 0.9270037412643433, + "learning_rate": 2.776366822882644e-06, + "loss": 0.5929, + "step": 7995 + }, + { + "epoch": 2.807584269662921, + "grad_norm": 1.329334020614624, + "learning_rate": 2.775907326517662e-06, + "loss": 0.5746, + "step": 7996 + }, + { + "epoch": 2.8079353932584272, + "grad_norm": 0.8749464154243469, + "learning_rate": 2.7754478207168855e-06, + "loss": 0.5369, + "step": 7997 + }, + { + "epoch": 2.808286516853933, + "grad_norm": 0.888923168182373, + "learning_rate": 2.774988305496029e-06, + "loss": 0.5814, + "step": 7998 + }, + { + "epoch": 2.8086376404494384, + "grad_norm": 0.9036821722984314, + "learning_rate": 2.7745287808708075e-06, + "loss": 0.5665, + "step": 7999 + }, + { + "epoch": 2.808988764044944, + "grad_norm": 0.9323573112487793, + "learning_rate": 2.7740692468569376e-06, + "loss": 0.5929, + "step": 8000 + }, + { + "epoch": 2.8093398876404496, + "grad_norm": 0.9229921102523804, + "learning_rate": 2.7736097034701327e-06, + "loss": 0.6094, + "step": 8001 + }, + { + "epoch": 2.809691011235955, + "grad_norm": 0.9492613077163696, + "learning_rate": 2.7731501507261103e-06, + "loss": 0.6013, + "step": 8002 + }, + { + "epoch": 2.810042134831461, + "grad_norm": 0.8981757760047913, + "learning_rate": 2.7726905886405868e-06, + "loss": 0.5676, + "step": 8003 + }, + { + "epoch": 2.8103932584269664, + "grad_norm": 0.8597955107688904, + "learning_rate": 2.7722310172292788e-06, + "loss": 0.5503, + "step": 8004 + }, + { + "epoch": 2.810744382022472, + "grad_norm": 0.9153335690498352, + "learning_rate": 2.771771436507903e-06, + "loss": 0.5851, + "step": 8005 + }, + { + "epoch": 2.8110955056179776, + "grad_norm": 0.9090461134910583, + "learning_rate": 2.771311846492176e-06, + "loss": 0.5918, + "step": 8006 + }, + { + "epoch": 2.811446629213483, + "grad_norm": 0.9025166034698486, + "learning_rate": 2.770852247197816e-06, + "loss": 0.5768, + "step": 8007 + }, + { + "epoch": 2.811797752808989, + "grad_norm": 0.9335866570472717, + "learning_rate": 2.7703926386405424e-06, + "loss": 0.5905, + "step": 8008 + }, + { + "epoch": 2.8121488764044944, + "grad_norm": 0.979668378829956, + "learning_rate": 2.76993302083607e-06, + "loss": 0.5982, + "step": 8009 + }, + { + "epoch": 2.8125, + "grad_norm": 0.8809782266616821, + "learning_rate": 2.7694733938001207e-06, + "loss": 0.5826, + "step": 8010 + }, + { + "epoch": 2.8128511235955056, + "grad_norm": 0.9384341239929199, + "learning_rate": 2.769013757548411e-06, + "loss": 0.6054, + "step": 8011 + }, + { + "epoch": 2.813202247191011, + "grad_norm": 0.9115980267524719, + "learning_rate": 2.7685541120966613e-06, + "loss": 0.5612, + "step": 8012 + }, + { + "epoch": 2.813553370786517, + "grad_norm": 0.9215322732925415, + "learning_rate": 2.7680944574605906e-06, + "loss": 0.6027, + "step": 8013 + }, + { + "epoch": 2.8139044943820224, + "grad_norm": 0.8628883957862854, + "learning_rate": 2.767634793655919e-06, + "loss": 0.5829, + "step": 8014 + }, + { + "epoch": 2.814255617977528, + "grad_norm": 0.954682469367981, + "learning_rate": 2.7671751206983667e-06, + "loss": 0.6071, + "step": 8015 + }, + { + "epoch": 2.8146067415730336, + "grad_norm": 0.899603009223938, + "learning_rate": 2.7667154386036533e-06, + "loss": 0.5667, + "step": 8016 + }, + { + "epoch": 2.814957865168539, + "grad_norm": 0.9117580056190491, + "learning_rate": 2.766255747387501e-06, + "loss": 0.5952, + "step": 8017 + }, + { + "epoch": 2.815308988764045, + "grad_norm": 0.869078516960144, + "learning_rate": 2.7657960470656294e-06, + "loss": 0.5789, + "step": 8018 + }, + { + "epoch": 2.8156601123595504, + "grad_norm": 0.9149150848388672, + "learning_rate": 2.76533633765376e-06, + "loss": 0.5764, + "step": 8019 + }, + { + "epoch": 2.816011235955056, + "grad_norm": 0.8775571584701538, + "learning_rate": 2.764876619167615e-06, + "loss": 0.5595, + "step": 8020 + }, + { + "epoch": 2.8163623595505616, + "grad_norm": 0.9042192697525024, + "learning_rate": 2.7644168916229175e-06, + "loss": 0.6032, + "step": 8021 + }, + { + "epoch": 2.816713483146067, + "grad_norm": 0.8988972306251526, + "learning_rate": 2.7639571550353876e-06, + "loss": 0.5808, + "step": 8022 + }, + { + "epoch": 2.8170646067415728, + "grad_norm": 0.870978057384491, + "learning_rate": 2.7634974094207493e-06, + "loss": 0.5399, + "step": 8023 + }, + { + "epoch": 2.817415730337079, + "grad_norm": 0.8812955021858215, + "learning_rate": 2.7630376547947245e-06, + "loss": 0.5495, + "step": 8024 + }, + { + "epoch": 2.8177668539325844, + "grad_norm": 0.9151604175567627, + "learning_rate": 2.762577891173038e-06, + "loss": 0.5792, + "step": 8025 + }, + { + "epoch": 2.81811797752809, + "grad_norm": 0.8923425078392029, + "learning_rate": 2.7621181185714115e-06, + "loss": 0.5639, + "step": 8026 + }, + { + "epoch": 2.8184691011235956, + "grad_norm": 0.9231808185577393, + "learning_rate": 2.7616583370055694e-06, + "loss": 0.559, + "step": 8027 + }, + { + "epoch": 2.818820224719101, + "grad_norm": 0.8981329798698425, + "learning_rate": 2.7611985464912367e-06, + "loss": 0.5389, + "step": 8028 + }, + { + "epoch": 2.819171348314607, + "grad_norm": 0.904042661190033, + "learning_rate": 2.7607387470441365e-06, + "loss": 0.5661, + "step": 8029 + }, + { + "epoch": 2.8195224719101124, + "grad_norm": 0.9179330468177795, + "learning_rate": 2.7602789386799948e-06, + "loss": 0.5895, + "step": 8030 + }, + { + "epoch": 2.819873595505618, + "grad_norm": 0.9180135726928711, + "learning_rate": 2.759819121414536e-06, + "loss": 0.5792, + "step": 8031 + }, + { + "epoch": 2.8202247191011236, + "grad_norm": 0.8892014622688293, + "learning_rate": 2.7593592952634857e-06, + "loss": 0.5497, + "step": 8032 + }, + { + "epoch": 2.820575842696629, + "grad_norm": 0.9591269493103027, + "learning_rate": 2.7588994602425696e-06, + "loss": 0.6176, + "step": 8033 + }, + { + "epoch": 2.820926966292135, + "grad_norm": 0.8720344305038452, + "learning_rate": 2.7584396163675135e-06, + "loss": 0.5822, + "step": 8034 + }, + { + "epoch": 2.8212780898876404, + "grad_norm": 0.9185641407966614, + "learning_rate": 2.7579797636540433e-06, + "loss": 0.5742, + "step": 8035 + }, + { + "epoch": 2.821629213483146, + "grad_norm": 0.8980414867401123, + "learning_rate": 2.757519902117886e-06, + "loss": 0.5906, + "step": 8036 + }, + { + "epoch": 2.8219803370786516, + "grad_norm": 0.900624692440033, + "learning_rate": 2.7570600317747685e-06, + "loss": 0.5974, + "step": 8037 + }, + { + "epoch": 2.822331460674157, + "grad_norm": 0.8863587379455566, + "learning_rate": 2.756600152640418e-06, + "loss": 0.565, + "step": 8038 + }, + { + "epoch": 2.822682584269663, + "grad_norm": 0.9254420399665833, + "learning_rate": 2.756140264730562e-06, + "loss": 0.5758, + "step": 8039 + }, + { + "epoch": 2.8230337078651684, + "grad_norm": 0.9113165140151978, + "learning_rate": 2.755680368060927e-06, + "loss": 0.5865, + "step": 8040 + }, + { + "epoch": 2.8233848314606744, + "grad_norm": 0.8823907971382141, + "learning_rate": 2.755220462647244e-06, + "loss": 0.5317, + "step": 8041 + }, + { + "epoch": 2.82373595505618, + "grad_norm": 0.8936343789100647, + "learning_rate": 2.7547605485052385e-06, + "loss": 0.585, + "step": 8042 + }, + { + "epoch": 2.8240870786516856, + "grad_norm": 0.9209809899330139, + "learning_rate": 2.7543006256506404e-06, + "loss": 0.5839, + "step": 8043 + }, + { + "epoch": 2.824438202247191, + "grad_norm": 0.8758163452148438, + "learning_rate": 2.7538406940991785e-06, + "loss": 0.5663, + "step": 8044 + }, + { + "epoch": 2.824789325842697, + "grad_norm": 0.9240387678146362, + "learning_rate": 2.7533807538665826e-06, + "loss": 0.6261, + "step": 8045 + }, + { + "epoch": 2.8251404494382024, + "grad_norm": 0.8865236043930054, + "learning_rate": 2.752920804968581e-06, + "loss": 0.5627, + "step": 8046 + }, + { + "epoch": 2.825491573033708, + "grad_norm": 1.0111860036849976, + "learning_rate": 2.752460847420904e-06, + "loss": 0.5886, + "step": 8047 + }, + { + "epoch": 2.8258426966292136, + "grad_norm": 0.9175224900245667, + "learning_rate": 2.7520008812392836e-06, + "loss": 0.5887, + "step": 8048 + }, + { + "epoch": 2.826193820224719, + "grad_norm": 0.8987218141555786, + "learning_rate": 2.7515409064394472e-06, + "loss": 0.5405, + "step": 8049 + }, + { + "epoch": 2.826544943820225, + "grad_norm": 0.8951777219772339, + "learning_rate": 2.7510809230371277e-06, + "loss": 0.5778, + "step": 8050 + }, + { + "epoch": 2.8268960674157304, + "grad_norm": 0.9131626486778259, + "learning_rate": 2.7506209310480553e-06, + "loss": 0.5833, + "step": 8051 + }, + { + "epoch": 2.827247191011236, + "grad_norm": 0.8816405534744263, + "learning_rate": 2.750160930487961e-06, + "loss": 0.5445, + "step": 8052 + }, + { + "epoch": 2.8275983146067416, + "grad_norm": 0.8860347270965576, + "learning_rate": 2.749700921372578e-06, + "loss": 0.5769, + "step": 8053 + }, + { + "epoch": 2.827949438202247, + "grad_norm": 0.876766562461853, + "learning_rate": 2.7492409037176365e-06, + "loss": 0.5823, + "step": 8054 + }, + { + "epoch": 2.828300561797753, + "grad_norm": 0.8878535628318787, + "learning_rate": 2.748780877538869e-06, + "loss": 0.5773, + "step": 8055 + }, + { + "epoch": 2.8286516853932584, + "grad_norm": 0.8702972531318665, + "learning_rate": 2.7483208428520093e-06, + "loss": 0.5477, + "step": 8056 + }, + { + "epoch": 2.829002808988764, + "grad_norm": 0.9484648108482361, + "learning_rate": 2.7478607996727887e-06, + "loss": 0.5917, + "step": 8057 + }, + { + "epoch": 2.8293539325842696, + "grad_norm": 0.9172254800796509, + "learning_rate": 2.7474007480169406e-06, + "loss": 0.5801, + "step": 8058 + }, + { + "epoch": 2.829705056179775, + "grad_norm": 0.904717743396759, + "learning_rate": 2.7469406879001997e-06, + "loss": 0.5681, + "step": 8059 + }, + { + "epoch": 2.830056179775281, + "grad_norm": 0.9146766662597656, + "learning_rate": 2.746480619338298e-06, + "loss": 0.5594, + "step": 8060 + }, + { + "epoch": 2.8304073033707864, + "grad_norm": 0.9276914596557617, + "learning_rate": 2.74602054234697e-06, + "loss": 0.576, + "step": 8061 + }, + { + "epoch": 2.830758426966292, + "grad_norm": 0.9250832200050354, + "learning_rate": 2.74556045694195e-06, + "loss": 0.6086, + "step": 8062 + }, + { + "epoch": 2.8311095505617976, + "grad_norm": 0.89726322889328, + "learning_rate": 2.745100363138972e-06, + "loss": 0.5828, + "step": 8063 + }, + { + "epoch": 2.831460674157303, + "grad_norm": 0.9294756650924683, + "learning_rate": 2.7446402609537713e-06, + "loss": 0.6005, + "step": 8064 + }, + { + "epoch": 2.831811797752809, + "grad_norm": 0.9135230183601379, + "learning_rate": 2.7441801504020837e-06, + "loss": 0.5672, + "step": 8065 + }, + { + "epoch": 2.8321629213483144, + "grad_norm": 0.8606854677200317, + "learning_rate": 2.7437200314996436e-06, + "loss": 0.5456, + "step": 8066 + }, + { + "epoch": 2.83251404494382, + "grad_norm": 0.911483108997345, + "learning_rate": 2.743259904262187e-06, + "loss": 0.5479, + "step": 8067 + }, + { + "epoch": 2.8328651685393256, + "grad_norm": 0.888810932636261, + "learning_rate": 2.74279976870545e-06, + "loss": 0.5971, + "step": 8068 + }, + { + "epoch": 2.8332162921348316, + "grad_norm": 0.9282062649726868, + "learning_rate": 2.7423396248451684e-06, + "loss": 0.5884, + "step": 8069 + }, + { + "epoch": 2.833567415730337, + "grad_norm": 0.9085161685943604, + "learning_rate": 2.741879472697079e-06, + "loss": 0.5813, + "step": 8070 + }, + { + "epoch": 2.833918539325843, + "grad_norm": 0.9183391332626343, + "learning_rate": 2.741419312276919e-06, + "loss": 0.5545, + "step": 8071 + }, + { + "epoch": 2.8342696629213484, + "grad_norm": 0.867479145526886, + "learning_rate": 2.740959143600425e-06, + "loss": 0.5583, + "step": 8072 + }, + { + "epoch": 2.834620786516854, + "grad_norm": 0.9330912232398987, + "learning_rate": 2.7404989666833342e-06, + "loss": 0.5969, + "step": 8073 + }, + { + "epoch": 2.8349719101123596, + "grad_norm": 0.919543981552124, + "learning_rate": 2.740038781541385e-06, + "loss": 0.615, + "step": 8074 + }, + { + "epoch": 2.835323033707865, + "grad_norm": 0.9216518998146057, + "learning_rate": 2.739578588190314e-06, + "loss": 0.5984, + "step": 8075 + }, + { + "epoch": 2.835674157303371, + "grad_norm": 0.8854014277458191, + "learning_rate": 2.739118386645861e-06, + "loss": 0.5677, + "step": 8076 + }, + { + "epoch": 2.8360252808988764, + "grad_norm": 0.9095461964607239, + "learning_rate": 2.7386581769237636e-06, + "loss": 0.5789, + "step": 8077 + }, + { + "epoch": 2.836376404494382, + "grad_norm": 0.864895224571228, + "learning_rate": 2.7381979590397608e-06, + "loss": 0.5593, + "step": 8078 + }, + { + "epoch": 2.8367275280898876, + "grad_norm": 0.9118948578834534, + "learning_rate": 2.737737733009592e-06, + "loss": 0.6008, + "step": 8079 + }, + { + "epoch": 2.837078651685393, + "grad_norm": 0.9061723351478577, + "learning_rate": 2.7372774988489955e-06, + "loss": 0.6262, + "step": 8080 + }, + { + "epoch": 2.837429775280899, + "grad_norm": 0.901373028755188, + "learning_rate": 2.7368172565737116e-06, + "loss": 0.593, + "step": 8081 + }, + { + "epoch": 2.8377808988764044, + "grad_norm": 0.9072074294090271, + "learning_rate": 2.7363570061994804e-06, + "loss": 0.5839, + "step": 8082 + }, + { + "epoch": 2.83813202247191, + "grad_norm": 0.8903719186782837, + "learning_rate": 2.7358967477420416e-06, + "loss": 0.5636, + "step": 8083 + }, + { + "epoch": 2.8384831460674156, + "grad_norm": 0.9042538404464722, + "learning_rate": 2.735436481217137e-06, + "loss": 0.5947, + "step": 8084 + }, + { + "epoch": 2.838834269662921, + "grad_norm": 0.9497808218002319, + "learning_rate": 2.7349762066405055e-06, + "loss": 0.6019, + "step": 8085 + }, + { + "epoch": 2.8391853932584272, + "grad_norm": 0.8912104368209839, + "learning_rate": 2.7345159240278885e-06, + "loss": 0.5625, + "step": 8086 + }, + { + "epoch": 2.839536516853933, + "grad_norm": 0.8984716534614563, + "learning_rate": 2.7340556333950275e-06, + "loss": 0.5867, + "step": 8087 + }, + { + "epoch": 2.8398876404494384, + "grad_norm": 0.9342149496078491, + "learning_rate": 2.7335953347576644e-06, + "loss": 0.5872, + "step": 8088 + }, + { + "epoch": 2.840238764044944, + "grad_norm": 1.068824052810669, + "learning_rate": 2.733135028131541e-06, + "loss": 0.573, + "step": 8089 + }, + { + "epoch": 2.8405898876404496, + "grad_norm": 0.9081486463546753, + "learning_rate": 2.732674713532399e-06, + "loss": 0.5746, + "step": 8090 + }, + { + "epoch": 2.840941011235955, + "grad_norm": 0.9154818058013916, + "learning_rate": 2.7322143909759803e-06, + "loss": 0.5715, + "step": 8091 + }, + { + "epoch": 2.841292134831461, + "grad_norm": 0.9216861128807068, + "learning_rate": 2.7317540604780286e-06, + "loss": 0.6038, + "step": 8092 + }, + { + "epoch": 2.8416432584269664, + "grad_norm": 0.8931046724319458, + "learning_rate": 2.7312937220542862e-06, + "loss": 0.61, + "step": 8093 + }, + { + "epoch": 2.841994382022472, + "grad_norm": 0.8908752799034119, + "learning_rate": 2.730833375720497e-06, + "loss": 0.5484, + "step": 8094 + }, + { + "epoch": 2.8423455056179776, + "grad_norm": 0.8965240120887756, + "learning_rate": 2.7303730214924027e-06, + "loss": 0.5618, + "step": 8095 + }, + { + "epoch": 2.842696629213483, + "grad_norm": 0.915827214717865, + "learning_rate": 2.7299126593857496e-06, + "loss": 0.5553, + "step": 8096 + }, + { + "epoch": 2.843047752808989, + "grad_norm": 0.905910074710846, + "learning_rate": 2.72945228941628e-06, + "loss": 0.6026, + "step": 8097 + }, + { + "epoch": 2.8433988764044944, + "grad_norm": 0.8721076250076294, + "learning_rate": 2.7289919115997375e-06, + "loss": 0.5659, + "step": 8098 + }, + { + "epoch": 2.84375, + "grad_norm": 0.8757249712944031, + "learning_rate": 2.7285315259518686e-06, + "loss": 0.5817, + "step": 8099 + }, + { + "epoch": 2.8441011235955056, + "grad_norm": 0.8586335182189941, + "learning_rate": 2.7280711324884162e-06, + "loss": 0.5412, + "step": 8100 + }, + { + "epoch": 2.844452247191011, + "grad_norm": 0.8704186081886292, + "learning_rate": 2.727610731225127e-06, + "loss": 0.531, + "step": 8101 + }, + { + "epoch": 2.844803370786517, + "grad_norm": 0.8958937525749207, + "learning_rate": 2.727150322177745e-06, + "loss": 0.5938, + "step": 8102 + }, + { + "epoch": 2.8451544943820224, + "grad_norm": 0.9165218472480774, + "learning_rate": 2.7266899053620166e-06, + "loss": 0.5703, + "step": 8103 + }, + { + "epoch": 2.845505617977528, + "grad_norm": 0.864075779914856, + "learning_rate": 2.7262294807936873e-06, + "loss": 0.564, + "step": 8104 + }, + { + "epoch": 2.8458567415730336, + "grad_norm": 0.885975182056427, + "learning_rate": 2.7257690484885036e-06, + "loss": 0.5647, + "step": 8105 + }, + { + "epoch": 2.846207865168539, + "grad_norm": 0.8838197588920593, + "learning_rate": 2.7253086084622118e-06, + "loss": 0.5584, + "step": 8106 + }, + { + "epoch": 2.846558988764045, + "grad_norm": 0.8782060146331787, + "learning_rate": 2.724848160730558e-06, + "loss": 0.5603, + "step": 8107 + }, + { + "epoch": 2.8469101123595504, + "grad_norm": 0.8954731225967407, + "learning_rate": 2.7243877053092893e-06, + "loss": 0.5635, + "step": 8108 + }, + { + "epoch": 2.847261235955056, + "grad_norm": 0.950038492679596, + "learning_rate": 2.723927242214154e-06, + "loss": 0.572, + "step": 8109 + }, + { + "epoch": 2.8476123595505616, + "grad_norm": 0.9107396006584167, + "learning_rate": 2.7234667714608974e-06, + "loss": 0.6064, + "step": 8110 + }, + { + "epoch": 2.847963483146067, + "grad_norm": 0.9317976832389832, + "learning_rate": 2.723006293065269e-06, + "loss": 0.5904, + "step": 8111 + }, + { + "epoch": 2.8483146067415728, + "grad_norm": 0.8979677557945251, + "learning_rate": 2.7225458070430166e-06, + "loss": 0.5849, + "step": 8112 + }, + { + "epoch": 2.848665730337079, + "grad_norm": 0.9154090285301208, + "learning_rate": 2.7220853134098874e-06, + "loss": 0.632, + "step": 8113 + }, + { + "epoch": 2.8490168539325844, + "grad_norm": 0.8495071530342102, + "learning_rate": 2.7216248121816307e-06, + "loss": 0.5321, + "step": 8114 + }, + { + "epoch": 2.84936797752809, + "grad_norm": 0.85858154296875, + "learning_rate": 2.721164303373995e-06, + "loss": 0.5731, + "step": 8115 + }, + { + "epoch": 2.8497191011235956, + "grad_norm": 0.8827555775642395, + "learning_rate": 2.72070378700273e-06, + "loss": 0.5809, + "step": 8116 + }, + { + "epoch": 2.850070224719101, + "grad_norm": 0.8819757103919983, + "learning_rate": 2.7202432630835832e-06, + "loss": 0.5876, + "step": 8117 + }, + { + "epoch": 2.850421348314607, + "grad_norm": 0.8956786394119263, + "learning_rate": 2.719782731632306e-06, + "loss": 0.6048, + "step": 8118 + }, + { + "epoch": 2.8507724719101124, + "grad_norm": 0.909347653388977, + "learning_rate": 2.719322192664648e-06, + "loss": 0.5887, + "step": 8119 + }, + { + "epoch": 2.851123595505618, + "grad_norm": 0.8959633111953735, + "learning_rate": 2.7188616461963587e-06, + "loss": 0.5769, + "step": 8120 + }, + { + "epoch": 2.8514747191011236, + "grad_norm": 0.8970386981964111, + "learning_rate": 2.7184010922431876e-06, + "loss": 0.5684, + "step": 8121 + }, + { + "epoch": 2.851825842696629, + "grad_norm": 0.8788214325904846, + "learning_rate": 2.717940530820886e-06, + "loss": 0.5563, + "step": 8122 + }, + { + "epoch": 2.852176966292135, + "grad_norm": 0.919266402721405, + "learning_rate": 2.7174799619452048e-06, + "loss": 0.57, + "step": 8123 + }, + { + "epoch": 2.8525280898876404, + "grad_norm": 0.9180001616477966, + "learning_rate": 2.717019385631895e-06, + "loss": 0.6066, + "step": 8124 + }, + { + "epoch": 2.852879213483146, + "grad_norm": 0.9025391340255737, + "learning_rate": 2.7165588018967084e-06, + "loss": 0.5858, + "step": 8125 + }, + { + "epoch": 2.8532303370786516, + "grad_norm": 0.9630336165428162, + "learning_rate": 2.7160982107553956e-06, + "loss": 0.6006, + "step": 8126 + }, + { + "epoch": 2.853581460674157, + "grad_norm": 0.8783957958221436, + "learning_rate": 2.715637612223709e-06, + "loss": 0.5783, + "step": 8127 + }, + { + "epoch": 2.853932584269663, + "grad_norm": 0.9016755819320679, + "learning_rate": 2.715177006317401e-06, + "loss": 0.5848, + "step": 8128 + }, + { + "epoch": 2.8542837078651684, + "grad_norm": 0.88164222240448, + "learning_rate": 2.714716393052223e-06, + "loss": 0.5624, + "step": 8129 + }, + { + "epoch": 2.8546348314606744, + "grad_norm": 0.8823972344398499, + "learning_rate": 2.7142557724439283e-06, + "loss": 0.5662, + "step": 8130 + }, + { + "epoch": 2.85498595505618, + "grad_norm": 0.891019344329834, + "learning_rate": 2.7137951445082692e-06, + "loss": 0.5736, + "step": 8131 + }, + { + "epoch": 2.8553370786516856, + "grad_norm": 0.901480495929718, + "learning_rate": 2.7133345092609997e-06, + "loss": 0.5767, + "step": 8132 + }, + { + "epoch": 2.855688202247191, + "grad_norm": 0.9263578653335571, + "learning_rate": 2.712873866717872e-06, + "loss": 0.5915, + "step": 8133 + }, + { + "epoch": 2.856039325842697, + "grad_norm": 0.8922516703605652, + "learning_rate": 2.7124132168946406e-06, + "loss": 0.5846, + "step": 8134 + }, + { + "epoch": 2.8563904494382024, + "grad_norm": 0.874633252620697, + "learning_rate": 2.7119525598070593e-06, + "loss": 0.5981, + "step": 8135 + }, + { + "epoch": 2.856741573033708, + "grad_norm": 0.9012803435325623, + "learning_rate": 2.7114918954708814e-06, + "loss": 0.5743, + "step": 8136 + }, + { + "epoch": 2.8570926966292136, + "grad_norm": 0.9137008786201477, + "learning_rate": 2.7110312239018616e-06, + "loss": 0.5899, + "step": 8137 + }, + { + "epoch": 2.857443820224719, + "grad_norm": 0.9057254791259766, + "learning_rate": 2.7105705451157544e-06, + "loss": 0.56, + "step": 8138 + }, + { + "epoch": 2.857794943820225, + "grad_norm": 0.901858389377594, + "learning_rate": 2.710109859128315e-06, + "loss": 0.5653, + "step": 8139 + }, + { + "epoch": 2.8581460674157304, + "grad_norm": 0.9124141335487366, + "learning_rate": 2.7096491659552975e-06, + "loss": 0.6001, + "step": 8140 + }, + { + "epoch": 2.858497191011236, + "grad_norm": 0.9009039998054504, + "learning_rate": 2.7091884656124584e-06, + "loss": 0.5762, + "step": 8141 + }, + { + "epoch": 2.8588483146067416, + "grad_norm": 0.9098173975944519, + "learning_rate": 2.708727758115553e-06, + "loss": 0.5942, + "step": 8142 + }, + { + "epoch": 2.859199438202247, + "grad_norm": 0.910575270652771, + "learning_rate": 2.7082670434803367e-06, + "loss": 0.5901, + "step": 8143 + }, + { + "epoch": 2.859550561797753, + "grad_norm": 0.9129311442375183, + "learning_rate": 2.707806321722566e-06, + "loss": 0.5392, + "step": 8144 + }, + { + "epoch": 2.8599016853932584, + "grad_norm": 0.9152742028236389, + "learning_rate": 2.707345592857996e-06, + "loss": 0.5954, + "step": 8145 + }, + { + "epoch": 2.860252808988764, + "grad_norm": 0.9020033478736877, + "learning_rate": 2.7068848569023853e-06, + "loss": 0.5663, + "step": 8146 + }, + { + "epoch": 2.8606039325842696, + "grad_norm": 0.899909257888794, + "learning_rate": 2.706424113871489e-06, + "loss": 0.5721, + "step": 8147 + }, + { + "epoch": 2.860955056179775, + "grad_norm": 0.8965358138084412, + "learning_rate": 2.705963363781065e-06, + "loss": 0.5767, + "step": 8148 + }, + { + "epoch": 2.861306179775281, + "grad_norm": 0.9218393564224243, + "learning_rate": 2.70550260664687e-06, + "loss": 0.5593, + "step": 8149 + }, + { + "epoch": 2.8616573033707864, + "grad_norm": 0.9264180660247803, + "learning_rate": 2.7050418424846616e-06, + "loss": 0.5995, + "step": 8150 + }, + { + "epoch": 2.862008426966292, + "grad_norm": 0.9252122044563293, + "learning_rate": 2.704581071310198e-06, + "loss": 0.6015, + "step": 8151 + }, + { + "epoch": 2.8623595505617976, + "grad_norm": 0.9273534417152405, + "learning_rate": 2.7041202931392356e-06, + "loss": 0.6163, + "step": 8152 + }, + { + "epoch": 2.862710674157303, + "grad_norm": 0.9162545800209045, + "learning_rate": 2.7036595079875348e-06, + "loss": 0.6354, + "step": 8153 + }, + { + "epoch": 2.863061797752809, + "grad_norm": 0.8859727382659912, + "learning_rate": 2.7031987158708535e-06, + "loss": 0.5654, + "step": 8154 + }, + { + "epoch": 2.8634129213483144, + "grad_norm": 0.9059848785400391, + "learning_rate": 2.7027379168049493e-06, + "loss": 0.5664, + "step": 8155 + }, + { + "epoch": 2.86376404494382, + "grad_norm": 0.933712899684906, + "learning_rate": 2.702277110805582e-06, + "loss": 0.6122, + "step": 8156 + }, + { + "epoch": 2.8641151685393256, + "grad_norm": 0.892230749130249, + "learning_rate": 2.7018162978885103e-06, + "loss": 0.5777, + "step": 8157 + }, + { + "epoch": 2.8644662921348316, + "grad_norm": 0.898952066898346, + "learning_rate": 2.7013554780694946e-06, + "loss": 0.554, + "step": 8158 + }, + { + "epoch": 2.864817415730337, + "grad_norm": 0.9052218198776245, + "learning_rate": 2.7008946513642932e-06, + "loss": 0.5708, + "step": 8159 + }, + { + "epoch": 2.865168539325843, + "grad_norm": 0.899046003818512, + "learning_rate": 2.7004338177886675e-06, + "loss": 0.5912, + "step": 8160 + }, + { + "epoch": 2.8655196629213484, + "grad_norm": 0.8891444802284241, + "learning_rate": 2.699972977358376e-06, + "loss": 0.5622, + "step": 8161 + }, + { + "epoch": 2.865870786516854, + "grad_norm": 0.8865372538566589, + "learning_rate": 2.69951213008918e-06, + "loss": 0.5602, + "step": 8162 + }, + { + "epoch": 2.8662219101123596, + "grad_norm": 0.885740339756012, + "learning_rate": 2.69905127599684e-06, + "loss": 0.5772, + "step": 8163 + }, + { + "epoch": 2.866573033707865, + "grad_norm": 0.9158214330673218, + "learning_rate": 2.6985904150971165e-06, + "loss": 0.5985, + "step": 8164 + }, + { + "epoch": 2.866924157303371, + "grad_norm": 0.9154390096664429, + "learning_rate": 2.6981295474057705e-06, + "loss": 0.5651, + "step": 8165 + }, + { + "epoch": 2.8672752808988764, + "grad_norm": 0.9186477661132812, + "learning_rate": 2.6976686729385645e-06, + "loss": 0.6049, + "step": 8166 + }, + { + "epoch": 2.867626404494382, + "grad_norm": 0.8657130599021912, + "learning_rate": 2.6972077917112576e-06, + "loss": 0.5844, + "step": 8167 + }, + { + "epoch": 2.8679775280898876, + "grad_norm": 0.8702936768531799, + "learning_rate": 2.696746903739614e-06, + "loss": 0.5328, + "step": 8168 + }, + { + "epoch": 2.868328651685393, + "grad_norm": 0.8942049741744995, + "learning_rate": 2.6962860090393945e-06, + "loss": 0.5509, + "step": 8169 + }, + { + "epoch": 2.868679775280899, + "grad_norm": 0.8806273341178894, + "learning_rate": 2.6958251076263608e-06, + "loss": 0.5402, + "step": 8170 + }, + { + "epoch": 2.8690308988764044, + "grad_norm": 0.9089407324790955, + "learning_rate": 2.6953641995162766e-06, + "loss": 0.5835, + "step": 8171 + }, + { + "epoch": 2.86938202247191, + "grad_norm": 0.9090007543563843, + "learning_rate": 2.6949032847249046e-06, + "loss": 0.5812, + "step": 8172 + }, + { + "epoch": 2.8697331460674156, + "grad_norm": 0.9046833515167236, + "learning_rate": 2.6944423632680063e-06, + "loss": 0.5984, + "step": 8173 + }, + { + "epoch": 2.870084269662921, + "grad_norm": 0.926001787185669, + "learning_rate": 2.6939814351613454e-06, + "loss": 0.561, + "step": 8174 + }, + { + "epoch": 2.8704353932584272, + "grad_norm": 0.9212695360183716, + "learning_rate": 2.6935205004206856e-06, + "loss": 0.5727, + "step": 8175 + }, + { + "epoch": 2.870786516853933, + "grad_norm": 0.907873809337616, + "learning_rate": 2.6930595590617903e-06, + "loss": 0.5926, + "step": 8176 + }, + { + "epoch": 2.8711376404494384, + "grad_norm": 0.860938310623169, + "learning_rate": 2.6925986111004234e-06, + "loss": 0.5679, + "step": 8177 + }, + { + "epoch": 2.871488764044944, + "grad_norm": 0.8981375694274902, + "learning_rate": 2.6921376565523483e-06, + "loss": 0.5884, + "step": 8178 + }, + { + "epoch": 2.8718398876404496, + "grad_norm": 0.9225471019744873, + "learning_rate": 2.69167669543333e-06, + "loss": 0.59, + "step": 8179 + }, + { + "epoch": 2.872191011235955, + "grad_norm": 0.9178627729415894, + "learning_rate": 2.6912157277591327e-06, + "loss": 0.5929, + "step": 8180 + }, + { + "epoch": 2.872542134831461, + "grad_norm": 0.9277577996253967, + "learning_rate": 2.690754753545521e-06, + "loss": 0.5926, + "step": 8181 + }, + { + "epoch": 2.8728932584269664, + "grad_norm": 0.9066944718360901, + "learning_rate": 2.6902937728082606e-06, + "loss": 0.6079, + "step": 8182 + }, + { + "epoch": 2.873244382022472, + "grad_norm": 0.9437388181686401, + "learning_rate": 2.689832785563116e-06, + "loss": 0.6046, + "step": 8183 + }, + { + "epoch": 2.8735955056179776, + "grad_norm": 0.9208738207817078, + "learning_rate": 2.6893717918258523e-06, + "loss": 0.5726, + "step": 8184 + }, + { + "epoch": 2.873946629213483, + "grad_norm": 0.8848270773887634, + "learning_rate": 2.6889107916122355e-06, + "loss": 0.5506, + "step": 8185 + }, + { + "epoch": 2.874297752808989, + "grad_norm": 0.900601327419281, + "learning_rate": 2.6884497849380307e-06, + "loss": 0.5866, + "step": 8186 + }, + { + "epoch": 2.8746488764044944, + "grad_norm": 0.8968028426170349, + "learning_rate": 2.687988771819005e-06, + "loss": 0.5671, + "step": 8187 + }, + { + "epoch": 2.875, + "grad_norm": 0.8851161599159241, + "learning_rate": 2.6875277522709243e-06, + "loss": 0.5733, + "step": 8188 + }, + { + "epoch": 2.8753511235955056, + "grad_norm": 0.898205041885376, + "learning_rate": 2.6870667263095544e-06, + "loss": 0.5686, + "step": 8189 + }, + { + "epoch": 2.875702247191011, + "grad_norm": 0.9170325994491577, + "learning_rate": 2.686605693950663e-06, + "loss": 0.5626, + "step": 8190 + }, + { + "epoch": 2.876053370786517, + "grad_norm": 0.9037489295005798, + "learning_rate": 2.686144655210016e-06, + "loss": 0.6013, + "step": 8191 + }, + { + "epoch": 2.8764044943820224, + "grad_norm": 0.9015204310417175, + "learning_rate": 2.6856836101033817e-06, + "loss": 0.5977, + "step": 8192 + }, + { + "epoch": 2.876755617977528, + "grad_norm": 0.9107714891433716, + "learning_rate": 2.6852225586465267e-06, + "loss": 0.5712, + "step": 8193 + }, + { + "epoch": 2.8771067415730336, + "grad_norm": 0.8946993350982666, + "learning_rate": 2.6847615008552185e-06, + "loss": 0.5868, + "step": 8194 + }, + { + "epoch": 2.877457865168539, + "grad_norm": 0.9009749293327332, + "learning_rate": 2.6843004367452253e-06, + "loss": 0.5928, + "step": 8195 + }, + { + "epoch": 2.877808988764045, + "grad_norm": 0.9526275992393494, + "learning_rate": 2.6838393663323146e-06, + "loss": 0.6014, + "step": 8196 + }, + { + "epoch": 2.8781601123595504, + "grad_norm": 0.9459334015846252, + "learning_rate": 2.6833782896322545e-06, + "loss": 0.6011, + "step": 8197 + }, + { + "epoch": 2.878511235955056, + "grad_norm": 0.88639897108078, + "learning_rate": 2.6829172066608154e-06, + "loss": 0.5852, + "step": 8198 + }, + { + "epoch": 2.8788623595505616, + "grad_norm": 0.8830696940422058, + "learning_rate": 2.682456117433763e-06, + "loss": 0.5299, + "step": 8199 + }, + { + "epoch": 2.879213483146067, + "grad_norm": 0.9058687686920166, + "learning_rate": 2.6819950219668677e-06, + "loss": 0.5837, + "step": 8200 + }, + { + "epoch": 2.8795646067415728, + "grad_norm": 0.8865855932235718, + "learning_rate": 2.681533920275899e-06, + "loss": 0.5805, + "step": 8201 + }, + { + "epoch": 2.879915730337079, + "grad_norm": 0.8935736417770386, + "learning_rate": 2.6810728123766244e-06, + "loss": 0.6156, + "step": 8202 + }, + { + "epoch": 2.8802668539325844, + "grad_norm": 0.8725783824920654, + "learning_rate": 2.6806116982848153e-06, + "loss": 0.589, + "step": 8203 + }, + { + "epoch": 2.88061797752809, + "grad_norm": 0.9060936570167542, + "learning_rate": 2.680150578016241e-06, + "loss": 0.5511, + "step": 8204 + }, + { + "epoch": 2.8809691011235956, + "grad_norm": 0.8891122937202454, + "learning_rate": 2.67968945158667e-06, + "loss": 0.5614, + "step": 8205 + }, + { + "epoch": 2.881320224719101, + "grad_norm": 0.8917089104652405, + "learning_rate": 2.6792283190118746e-06, + "loss": 0.5791, + "step": 8206 + }, + { + "epoch": 2.881671348314607, + "grad_norm": 0.9016678929328918, + "learning_rate": 2.678767180307624e-06, + "loss": 0.6042, + "step": 8207 + }, + { + "epoch": 2.8820224719101124, + "grad_norm": 0.8830958008766174, + "learning_rate": 2.678306035489688e-06, + "loss": 0.5531, + "step": 8208 + }, + { + "epoch": 2.882373595505618, + "grad_norm": 0.899713933467865, + "learning_rate": 2.677844884573839e-06, + "loss": 0.6191, + "step": 8209 + }, + { + "epoch": 2.8827247191011236, + "grad_norm": 0.918228268623352, + "learning_rate": 2.677383727575847e-06, + "loss": 0.5726, + "step": 8210 + }, + { + "epoch": 2.883075842696629, + "grad_norm": 0.9001905918121338, + "learning_rate": 2.676922564511483e-06, + "loss": 0.581, + "step": 8211 + }, + { + "epoch": 2.883426966292135, + "grad_norm": 0.8910868763923645, + "learning_rate": 2.6764613953965185e-06, + "loss": 0.5635, + "step": 8212 + }, + { + "epoch": 2.8837780898876404, + "grad_norm": 0.9370684027671814, + "learning_rate": 2.6760002202467257e-06, + "loss": 0.5989, + "step": 8213 + }, + { + "epoch": 2.884129213483146, + "grad_norm": 0.9193254709243774, + "learning_rate": 2.6755390390778757e-06, + "loss": 0.5739, + "step": 8214 + }, + { + "epoch": 2.8844803370786516, + "grad_norm": 0.9021379947662354, + "learning_rate": 2.6750778519057415e-06, + "loss": 0.5512, + "step": 8215 + }, + { + "epoch": 2.884831460674157, + "grad_norm": 0.8739520907402039, + "learning_rate": 2.6746166587460936e-06, + "loss": 0.5424, + "step": 8216 + }, + { + "epoch": 2.885182584269663, + "grad_norm": 0.9087174534797668, + "learning_rate": 2.674155459614706e-06, + "loss": 0.5748, + "step": 8217 + }, + { + "epoch": 2.8855337078651684, + "grad_norm": 0.8752431273460388, + "learning_rate": 2.6736942545273505e-06, + "loss": 0.5639, + "step": 8218 + }, + { + "epoch": 2.8858848314606744, + "grad_norm": 0.9016522765159607, + "learning_rate": 2.6732330434998e-06, + "loss": 0.5961, + "step": 8219 + }, + { + "epoch": 2.88623595505618, + "grad_norm": 0.9272969365119934, + "learning_rate": 2.6727718265478276e-06, + "loss": 0.5937, + "step": 8220 + }, + { + "epoch": 2.8865870786516856, + "grad_norm": 0.9172776937484741, + "learning_rate": 2.6723106036872066e-06, + "loss": 0.5962, + "step": 8221 + }, + { + "epoch": 2.886938202247191, + "grad_norm": 0.8959907293319702, + "learning_rate": 2.6718493749337104e-06, + "loss": 0.5699, + "step": 8222 + }, + { + "epoch": 2.887289325842697, + "grad_norm": 0.9008168578147888, + "learning_rate": 2.6713881403031134e-06, + "loss": 0.5996, + "step": 8223 + }, + { + "epoch": 2.8876404494382024, + "grad_norm": 0.8963308930397034, + "learning_rate": 2.6709268998111878e-06, + "loss": 0.5738, + "step": 8224 + }, + { + "epoch": 2.887991573033708, + "grad_norm": 0.8792314529418945, + "learning_rate": 2.670465653473709e-06, + "loss": 0.5704, + "step": 8225 + }, + { + "epoch": 2.8883426966292136, + "grad_norm": 0.9152819514274597, + "learning_rate": 2.67000440130645e-06, + "loss": 0.5897, + "step": 8226 + }, + { + "epoch": 2.888693820224719, + "grad_norm": 0.882333517074585, + "learning_rate": 2.669543143325186e-06, + "loss": 0.569, + "step": 8227 + }, + { + "epoch": 2.889044943820225, + "grad_norm": 0.9098389744758606, + "learning_rate": 2.6690818795456925e-06, + "loss": 0.6012, + "step": 8228 + }, + { + "epoch": 2.8893960674157304, + "grad_norm": 0.8887183666229248, + "learning_rate": 2.6686206099837427e-06, + "loss": 0.58, + "step": 8229 + }, + { + "epoch": 2.889747191011236, + "grad_norm": 0.8996925354003906, + "learning_rate": 2.668159334655112e-06, + "loss": 0.6037, + "step": 8230 + }, + { + "epoch": 2.8900983146067416, + "grad_norm": 0.9456595778465271, + "learning_rate": 2.667698053575577e-06, + "loss": 0.6214, + "step": 8231 + }, + { + "epoch": 2.890449438202247, + "grad_norm": 0.9302764534950256, + "learning_rate": 2.667236766760911e-06, + "loss": 0.5892, + "step": 8232 + }, + { + "epoch": 2.890800561797753, + "grad_norm": 0.8984224796295166, + "learning_rate": 2.6667754742268915e-06, + "loss": 0.6065, + "step": 8233 + }, + { + "epoch": 2.8911516853932584, + "grad_norm": 0.8788968324661255, + "learning_rate": 2.666314175989293e-06, + "loss": 0.5796, + "step": 8234 + }, + { + "epoch": 2.891502808988764, + "grad_norm": 0.9025775790214539, + "learning_rate": 2.665852872063893e-06, + "loss": 0.5717, + "step": 8235 + }, + { + "epoch": 2.8918539325842696, + "grad_norm": 0.9194645881652832, + "learning_rate": 2.6653915624664657e-06, + "loss": 0.5954, + "step": 8236 + }, + { + "epoch": 2.892205056179775, + "grad_norm": 0.8912577033042908, + "learning_rate": 2.6649302472127885e-06, + "loss": 0.538, + "step": 8237 + }, + { + "epoch": 2.892556179775281, + "grad_norm": 0.9098647236824036, + "learning_rate": 2.664468926318638e-06, + "loss": 0.5673, + "step": 8238 + }, + { + "epoch": 2.8929073033707864, + "grad_norm": 0.9245277643203735, + "learning_rate": 2.664007599799791e-06, + "loss": 0.6324, + "step": 8239 + }, + { + "epoch": 2.893258426966292, + "grad_norm": 0.8771243691444397, + "learning_rate": 2.6635462676720247e-06, + "loss": 0.5427, + "step": 8240 + }, + { + "epoch": 2.8936095505617976, + "grad_norm": 0.8846092224121094, + "learning_rate": 2.6630849299511158e-06, + "loss": 0.5732, + "step": 8241 + }, + { + "epoch": 2.893960674157303, + "grad_norm": 0.8946841359138489, + "learning_rate": 2.6626235866528417e-06, + "loss": 0.5557, + "step": 8242 + }, + { + "epoch": 2.894311797752809, + "grad_norm": 0.922015368938446, + "learning_rate": 2.66216223779298e-06, + "loss": 0.583, + "step": 8243 + }, + { + "epoch": 2.8946629213483144, + "grad_norm": 0.9224690794944763, + "learning_rate": 2.6617008833873094e-06, + "loss": 0.5683, + "step": 8244 + }, + { + "epoch": 2.89501404494382, + "grad_norm": 0.8936585187911987, + "learning_rate": 2.6612395234516057e-06, + "loss": 0.5629, + "step": 8245 + }, + { + "epoch": 2.8953651685393256, + "grad_norm": 0.9283834099769592, + "learning_rate": 2.660778158001649e-06, + "loss": 0.5986, + "step": 8246 + }, + { + "epoch": 2.8957162921348316, + "grad_norm": 0.930003821849823, + "learning_rate": 2.660316787053217e-06, + "loss": 0.5894, + "step": 8247 + }, + { + "epoch": 2.896067415730337, + "grad_norm": 0.9406369924545288, + "learning_rate": 2.659855410622088e-06, + "loss": 0.6108, + "step": 8248 + }, + { + "epoch": 2.896418539325843, + "grad_norm": 0.8958816528320312, + "learning_rate": 2.6593940287240406e-06, + "loss": 0.5851, + "step": 8249 + }, + { + "epoch": 2.8967696629213484, + "grad_norm": 0.9197438955307007, + "learning_rate": 2.658932641374854e-06, + "loss": 0.5797, + "step": 8250 + }, + { + "epoch": 2.897120786516854, + "grad_norm": 0.9176791310310364, + "learning_rate": 2.658471248590307e-06, + "loss": 0.6005, + "step": 8251 + }, + { + "epoch": 2.8974719101123596, + "grad_norm": 0.8915731310844421, + "learning_rate": 2.6580098503861786e-06, + "loss": 0.602, + "step": 8252 + }, + { + "epoch": 2.897823033707865, + "grad_norm": 0.8958067893981934, + "learning_rate": 2.6575484467782485e-06, + "loss": 0.5784, + "step": 8253 + }, + { + "epoch": 2.898174157303371, + "grad_norm": 0.9221014976501465, + "learning_rate": 2.6570870377822966e-06, + "loss": 0.5833, + "step": 8254 + }, + { + "epoch": 2.8985252808988764, + "grad_norm": 0.8922723531723022, + "learning_rate": 2.6566256234141025e-06, + "loss": 0.5765, + "step": 8255 + }, + { + "epoch": 2.898876404494382, + "grad_norm": 0.9050207138061523, + "learning_rate": 2.6561642036894458e-06, + "loss": 0.5667, + "step": 8256 + }, + { + "epoch": 2.8992275280898876, + "grad_norm": 0.9206213355064392, + "learning_rate": 2.6557027786241074e-06, + "loss": 0.5523, + "step": 8257 + }, + { + "epoch": 2.899578651685393, + "grad_norm": 0.8873724341392517, + "learning_rate": 2.6552413482338667e-06, + "loss": 0.546, + "step": 8258 + }, + { + "epoch": 2.899929775280899, + "grad_norm": 0.8937140107154846, + "learning_rate": 2.6547799125345053e-06, + "loss": 0.6007, + "step": 8259 + }, + { + "epoch": 2.9002808988764044, + "grad_norm": 0.9157314300537109, + "learning_rate": 2.654318471541803e-06, + "loss": 0.5636, + "step": 8260 + }, + { + "epoch": 2.90063202247191, + "grad_norm": 0.9253490567207336, + "learning_rate": 2.6538570252715414e-06, + "loss": 0.5632, + "step": 8261 + }, + { + "epoch": 2.9009831460674156, + "grad_norm": 0.9119126200675964, + "learning_rate": 2.6533955737395006e-06, + "loss": 0.5567, + "step": 8262 + }, + { + "epoch": 2.901334269662921, + "grad_norm": 0.920519232749939, + "learning_rate": 2.652934116961463e-06, + "loss": 0.596, + "step": 8263 + }, + { + "epoch": 2.9016853932584272, + "grad_norm": 0.9028516411781311, + "learning_rate": 2.652472654953209e-06, + "loss": 0.5655, + "step": 8264 + }, + { + "epoch": 2.902036516853933, + "grad_norm": 0.9474261999130249, + "learning_rate": 2.652011187730521e-06, + "loss": 0.5878, + "step": 8265 + }, + { + "epoch": 2.9023876404494384, + "grad_norm": 0.8954325914382935, + "learning_rate": 2.65154971530918e-06, + "loss": 0.5664, + "step": 8266 + }, + { + "epoch": 2.902738764044944, + "grad_norm": 0.8704952597618103, + "learning_rate": 2.651088237704969e-06, + "loss": 0.5687, + "step": 8267 + }, + { + "epoch": 2.9030898876404496, + "grad_norm": 0.9122330546379089, + "learning_rate": 2.650626754933669e-06, + "loss": 0.5768, + "step": 8268 + }, + { + "epoch": 2.903441011235955, + "grad_norm": 0.9100549817085266, + "learning_rate": 2.6501652670110635e-06, + "loss": 0.5707, + "step": 8269 + }, + { + "epoch": 2.903792134831461, + "grad_norm": 0.8868029117584229, + "learning_rate": 2.6497037739529335e-06, + "loss": 0.577, + "step": 8270 + }, + { + "epoch": 2.9041432584269664, + "grad_norm": 0.9084314107894897, + "learning_rate": 2.649242275775063e-06, + "loss": 0.5809, + "step": 8271 + }, + { + "epoch": 2.904494382022472, + "grad_norm": 0.8618656992912292, + "learning_rate": 2.6487807724932346e-06, + "loss": 0.5474, + "step": 8272 + }, + { + "epoch": 2.9048455056179776, + "grad_norm": 0.9551761746406555, + "learning_rate": 2.6483192641232316e-06, + "loss": 0.5672, + "step": 8273 + }, + { + "epoch": 2.905196629213483, + "grad_norm": 0.8906540870666504, + "learning_rate": 2.6478577506808357e-06, + "loss": 0.596, + "step": 8274 + }, + { + "epoch": 2.905547752808989, + "grad_norm": 0.8918073177337646, + "learning_rate": 2.6473962321818314e-06, + "loss": 0.5825, + "step": 8275 + }, + { + "epoch": 2.9058988764044944, + "grad_norm": 0.8955526947975159, + "learning_rate": 2.646934708642002e-06, + "loss": 0.568, + "step": 8276 + }, + { + "epoch": 2.90625, + "grad_norm": 0.906037449836731, + "learning_rate": 2.646473180077132e-06, + "loss": 0.583, + "step": 8277 + }, + { + "epoch": 2.9066011235955056, + "grad_norm": 0.9098412394523621, + "learning_rate": 2.646011646503004e-06, + "loss": 0.5876, + "step": 8278 + }, + { + "epoch": 2.906952247191011, + "grad_norm": 0.8843075633049011, + "learning_rate": 2.6455501079354027e-06, + "loss": 0.5637, + "step": 8279 + }, + { + "epoch": 2.907303370786517, + "grad_norm": 0.8774104118347168, + "learning_rate": 2.6450885643901124e-06, + "loss": 0.5648, + "step": 8280 + }, + { + "epoch": 2.9076544943820224, + "grad_norm": 0.8732701539993286, + "learning_rate": 2.6446270158829174e-06, + "loss": 0.5373, + "step": 8281 + }, + { + "epoch": 2.908005617977528, + "grad_norm": 0.9018761515617371, + "learning_rate": 2.644165462429602e-06, + "loss": 0.5973, + "step": 8282 + }, + { + "epoch": 2.9083567415730336, + "grad_norm": 0.9379571676254272, + "learning_rate": 2.643703904045951e-06, + "loss": 0.5982, + "step": 8283 + }, + { + "epoch": 2.908707865168539, + "grad_norm": 0.9016522765159607, + "learning_rate": 2.64324234074775e-06, + "loss": 0.5524, + "step": 8284 + }, + { + "epoch": 2.909058988764045, + "grad_norm": 0.8787747621536255, + "learning_rate": 2.6427807725507824e-06, + "loss": 0.5426, + "step": 8285 + }, + { + "epoch": 2.9094101123595504, + "grad_norm": 0.9074045419692993, + "learning_rate": 2.6423191994708364e-06, + "loss": 0.5944, + "step": 8286 + }, + { + "epoch": 2.909761235955056, + "grad_norm": 0.9042428135871887, + "learning_rate": 2.6418576215236946e-06, + "loss": 0.5897, + "step": 8287 + }, + { + "epoch": 2.9101123595505616, + "grad_norm": 0.9205817580223083, + "learning_rate": 2.641396038725143e-06, + "loss": 0.592, + "step": 8288 + }, + { + "epoch": 2.910463483146067, + "grad_norm": 0.9140691757202148, + "learning_rate": 2.6409344510909685e-06, + "loss": 0.6096, + "step": 8289 + }, + { + "epoch": 2.9108146067415728, + "grad_norm": 0.9287146329879761, + "learning_rate": 2.640472858636956e-06, + "loss": 0.5624, + "step": 8290 + }, + { + "epoch": 2.911165730337079, + "grad_norm": 0.8922277092933655, + "learning_rate": 2.6400112613788925e-06, + "loss": 0.607, + "step": 8291 + }, + { + "epoch": 2.9115168539325844, + "grad_norm": 0.8896186947822571, + "learning_rate": 2.639549659332563e-06, + "loss": 0.6023, + "step": 8292 + }, + { + "epoch": 2.91186797752809, + "grad_norm": 0.8995725512504578, + "learning_rate": 2.639088052513755e-06, + "loss": 0.5605, + "step": 8293 + }, + { + "epoch": 2.9122191011235956, + "grad_norm": 0.8958765268325806, + "learning_rate": 2.6386264409382552e-06, + "loss": 0.5694, + "step": 8294 + }, + { + "epoch": 2.912570224719101, + "grad_norm": 0.9619843363761902, + "learning_rate": 2.6381648246218487e-06, + "loss": 0.6198, + "step": 8295 + }, + { + "epoch": 2.912921348314607, + "grad_norm": 0.8933351635932922, + "learning_rate": 2.6377032035803235e-06, + "loss": 0.5608, + "step": 8296 + }, + { + "epoch": 2.9132724719101124, + "grad_norm": 0.8789713978767395, + "learning_rate": 2.637241577829467e-06, + "loss": 0.5505, + "step": 8297 + }, + { + "epoch": 2.913623595505618, + "grad_norm": 0.9287780523300171, + "learning_rate": 2.636779947385066e-06, + "loss": 0.5754, + "step": 8298 + }, + { + "epoch": 2.9139747191011236, + "grad_norm": 0.9333387017250061, + "learning_rate": 2.6363183122629076e-06, + "loss": 0.6068, + "step": 8299 + }, + { + "epoch": 2.914325842696629, + "grad_norm": 0.9124438762664795, + "learning_rate": 2.6358566724787794e-06, + "loss": 0.5425, + "step": 8300 + }, + { + "epoch": 2.914676966292135, + "grad_norm": 0.8794487714767456, + "learning_rate": 2.6353950280484697e-06, + "loss": 0.5784, + "step": 8301 + }, + { + "epoch": 2.9150280898876404, + "grad_norm": 0.9106101393699646, + "learning_rate": 2.6349333789877663e-06, + "loss": 0.5955, + "step": 8302 + }, + { + "epoch": 2.915379213483146, + "grad_norm": 0.8929299712181091, + "learning_rate": 2.634471725312456e-06, + "loss": 0.5882, + "step": 8303 + }, + { + "epoch": 2.9157303370786516, + "grad_norm": 0.9001083374023438, + "learning_rate": 2.6340100670383277e-06, + "loss": 0.6032, + "step": 8304 + }, + { + "epoch": 2.916081460674157, + "grad_norm": 0.9037061929702759, + "learning_rate": 2.63354840418117e-06, + "loss": 0.5885, + "step": 8305 + }, + { + "epoch": 2.916432584269663, + "grad_norm": 0.9323333501815796, + "learning_rate": 2.6330867367567714e-06, + "loss": 0.6266, + "step": 8306 + }, + { + "epoch": 2.9167837078651684, + "grad_norm": 0.8839550614356995, + "learning_rate": 2.6326250647809198e-06, + "loss": 0.5862, + "step": 8307 + }, + { + "epoch": 2.9171348314606744, + "grad_norm": 0.9128544330596924, + "learning_rate": 2.632163388269405e-06, + "loss": 0.5901, + "step": 8308 + }, + { + "epoch": 2.91748595505618, + "grad_norm": 0.892253577709198, + "learning_rate": 2.631701707238015e-06, + "loss": 0.585, + "step": 8309 + }, + { + "epoch": 2.9178370786516856, + "grad_norm": 0.8858360648155212, + "learning_rate": 2.63124002170254e-06, + "loss": 0.5826, + "step": 8310 + }, + { + "epoch": 2.918188202247191, + "grad_norm": 0.9461344480514526, + "learning_rate": 2.630778331678768e-06, + "loss": 0.5896, + "step": 8311 + }, + { + "epoch": 2.918539325842697, + "grad_norm": 0.9293336868286133, + "learning_rate": 2.630316637182489e-06, + "loss": 0.5862, + "step": 8312 + }, + { + "epoch": 2.9188904494382024, + "grad_norm": 0.906002938747406, + "learning_rate": 2.6298549382294924e-06, + "loss": 0.589, + "step": 8313 + }, + { + "epoch": 2.919241573033708, + "grad_norm": 0.8614712357521057, + "learning_rate": 2.6293932348355684e-06, + "loss": 0.5378, + "step": 8314 + }, + { + "epoch": 2.9195926966292136, + "grad_norm": 0.908088207244873, + "learning_rate": 2.628931527016506e-06, + "loss": 0.5963, + "step": 8315 + }, + { + "epoch": 2.919943820224719, + "grad_norm": 0.9295352101325989, + "learning_rate": 2.628469814788096e-06, + "loss": 0.5958, + "step": 8316 + }, + { + "epoch": 2.920294943820225, + "grad_norm": 0.8814553618431091, + "learning_rate": 2.628008098166128e-06, + "loss": 0.5802, + "step": 8317 + }, + { + "epoch": 2.9206460674157304, + "grad_norm": 0.8914538621902466, + "learning_rate": 2.6275463771663927e-06, + "loss": 0.5542, + "step": 8318 + }, + { + "epoch": 2.920997191011236, + "grad_norm": 1.0258550643920898, + "learning_rate": 2.6270846518046806e-06, + "loss": 0.5848, + "step": 8319 + }, + { + "epoch": 2.9213483146067416, + "grad_norm": 0.9142085313796997, + "learning_rate": 2.626622922096782e-06, + "loss": 0.5412, + "step": 8320 + }, + { + "epoch": 2.921699438202247, + "grad_norm": 0.9338873624801636, + "learning_rate": 2.6261611880584882e-06, + "loss": 0.5942, + "step": 8321 + }, + { + "epoch": 2.922050561797753, + "grad_norm": 0.92507404088974, + "learning_rate": 2.625699449705589e-06, + "loss": 0.588, + "step": 8322 + }, + { + "epoch": 2.9224016853932584, + "grad_norm": 0.9166693687438965, + "learning_rate": 2.625237707053877e-06, + "loss": 0.6007, + "step": 8323 + }, + { + "epoch": 2.922752808988764, + "grad_norm": 0.9323252439498901, + "learning_rate": 2.6247759601191427e-06, + "loss": 0.6293, + "step": 8324 + }, + { + "epoch": 2.9231039325842696, + "grad_norm": 0.9325878620147705, + "learning_rate": 2.624314208917177e-06, + "loss": 0.5864, + "step": 8325 + }, + { + "epoch": 2.923455056179775, + "grad_norm": 0.893571138381958, + "learning_rate": 2.623852453463771e-06, + "loss": 0.5532, + "step": 8326 + }, + { + "epoch": 2.923806179775281, + "grad_norm": 0.9113224744796753, + "learning_rate": 2.6233906937747177e-06, + "loss": 0.5676, + "step": 8327 + }, + { + "epoch": 2.9241573033707864, + "grad_norm": 0.9172493815422058, + "learning_rate": 2.6229289298658084e-06, + "loss": 0.5934, + "step": 8328 + }, + { + "epoch": 2.924508426966292, + "grad_norm": 0.9047500491142273, + "learning_rate": 2.6224671617528345e-06, + "loss": 0.5904, + "step": 8329 + }, + { + "epoch": 2.9248595505617976, + "grad_norm": 0.9123345613479614, + "learning_rate": 2.622005389451589e-06, + "loss": 0.5612, + "step": 8330 + }, + { + "epoch": 2.925210674157303, + "grad_norm": 0.9070037603378296, + "learning_rate": 2.621543612977863e-06, + "loss": 0.5929, + "step": 8331 + }, + { + "epoch": 2.925561797752809, + "grad_norm": 0.9156293272972107, + "learning_rate": 2.6210818323474495e-06, + "loss": 0.5998, + "step": 8332 + }, + { + "epoch": 2.9259129213483144, + "grad_norm": 0.8773887753486633, + "learning_rate": 2.620620047576141e-06, + "loss": 0.5361, + "step": 8333 + }, + { + "epoch": 2.92626404494382, + "grad_norm": 0.9225067496299744, + "learning_rate": 2.6201582586797303e-06, + "loss": 0.5539, + "step": 8334 + }, + { + "epoch": 2.9266151685393256, + "grad_norm": 0.9035494327545166, + "learning_rate": 2.61969646567401e-06, + "loss": 0.6095, + "step": 8335 + }, + { + "epoch": 2.9269662921348316, + "grad_norm": 0.9184130430221558, + "learning_rate": 2.619234668574773e-06, + "loss": 0.5638, + "step": 8336 + }, + { + "epoch": 2.927317415730337, + "grad_norm": 0.9005270600318909, + "learning_rate": 2.6187728673978126e-06, + "loss": 0.5996, + "step": 8337 + }, + { + "epoch": 2.927668539325843, + "grad_norm": 0.9133041501045227, + "learning_rate": 2.6183110621589213e-06, + "loss": 0.603, + "step": 8338 + }, + { + "epoch": 2.9280196629213484, + "grad_norm": 0.931492030620575, + "learning_rate": 2.617849252873893e-06, + "loss": 0.5953, + "step": 8339 + }, + { + "epoch": 2.928370786516854, + "grad_norm": 0.8507922887802124, + "learning_rate": 2.6173874395585212e-06, + "loss": 0.5558, + "step": 8340 + }, + { + "epoch": 2.9287219101123596, + "grad_norm": 0.8982647061347961, + "learning_rate": 2.6169256222285992e-06, + "loss": 0.5924, + "step": 8341 + }, + { + "epoch": 2.929073033707865, + "grad_norm": 0.9138845801353455, + "learning_rate": 2.6164638008999215e-06, + "loss": 0.5775, + "step": 8342 + }, + { + "epoch": 2.929424157303371, + "grad_norm": 0.8981946706771851, + "learning_rate": 2.616001975588281e-06, + "loss": 0.5656, + "step": 8343 + }, + { + "epoch": 2.9297752808988764, + "grad_norm": 0.8824996948242188, + "learning_rate": 2.615540146309472e-06, + "loss": 0.5731, + "step": 8344 + }, + { + "epoch": 2.930126404494382, + "grad_norm": 0.9175012707710266, + "learning_rate": 2.6150783130792897e-06, + "loss": 0.5697, + "step": 8345 + }, + { + "epoch": 2.9304775280898876, + "grad_norm": 0.9154291152954102, + "learning_rate": 2.614616475913527e-06, + "loss": 0.6239, + "step": 8346 + }, + { + "epoch": 2.930828651685393, + "grad_norm": 0.905644953250885, + "learning_rate": 2.6141546348279784e-06, + "loss": 0.5457, + "step": 8347 + }, + { + "epoch": 2.931179775280899, + "grad_norm": 0.9273164868354797, + "learning_rate": 2.6136927898384394e-06, + "loss": 0.5688, + "step": 8348 + }, + { + "epoch": 2.9315308988764044, + "grad_norm": 0.9161174893379211, + "learning_rate": 2.6132309409607054e-06, + "loss": 0.5759, + "step": 8349 + }, + { + "epoch": 2.93188202247191, + "grad_norm": 0.8911331295967102, + "learning_rate": 2.612769088210569e-06, + "loss": 0.5609, + "step": 8350 + }, + { + "epoch": 2.9322331460674156, + "grad_norm": 0.9204432964324951, + "learning_rate": 2.612307231603826e-06, + "loss": 0.6021, + "step": 8351 + }, + { + "epoch": 2.932584269662921, + "grad_norm": 0.9142928719520569, + "learning_rate": 2.6118453711562726e-06, + "loss": 0.5427, + "step": 8352 + }, + { + "epoch": 2.9329353932584272, + "grad_norm": 0.9489219188690186, + "learning_rate": 2.6113835068837025e-06, + "loss": 0.5968, + "step": 8353 + }, + { + "epoch": 2.933286516853933, + "grad_norm": 0.9173111915588379, + "learning_rate": 2.610921638801912e-06, + "loss": 0.5764, + "step": 8354 + }, + { + "epoch": 2.9336376404494384, + "grad_norm": 0.9424446225166321, + "learning_rate": 2.610459766926696e-06, + "loss": 0.601, + "step": 8355 + }, + { + "epoch": 2.933988764044944, + "grad_norm": 0.9405547380447388, + "learning_rate": 2.609997891273851e-06, + "loss": 0.6099, + "step": 8356 + }, + { + "epoch": 2.9343398876404496, + "grad_norm": 0.8925829529762268, + "learning_rate": 2.6095360118591722e-06, + "loss": 0.5497, + "step": 8357 + }, + { + "epoch": 2.934691011235955, + "grad_norm": 0.9334905743598938, + "learning_rate": 2.6090741286984556e-06, + "loss": 0.5802, + "step": 8358 + }, + { + "epoch": 2.935042134831461, + "grad_norm": 0.8693034052848816, + "learning_rate": 2.608612241807497e-06, + "loss": 0.5528, + "step": 8359 + }, + { + "epoch": 2.9353932584269664, + "grad_norm": 0.8875693082809448, + "learning_rate": 2.6081503512020926e-06, + "loss": 0.5639, + "step": 8360 + }, + { + "epoch": 2.935744382022472, + "grad_norm": 0.8978365063667297, + "learning_rate": 2.6076884568980392e-06, + "loss": 0.6361, + "step": 8361 + }, + { + "epoch": 2.9360955056179776, + "grad_norm": 0.9226500988006592, + "learning_rate": 2.6072265589111324e-06, + "loss": 0.5805, + "step": 8362 + }, + { + "epoch": 2.936446629213483, + "grad_norm": 0.9010093808174133, + "learning_rate": 2.606764657257169e-06, + "loss": 0.589, + "step": 8363 + }, + { + "epoch": 2.936797752808989, + "grad_norm": 0.9042021036148071, + "learning_rate": 2.606302751951946e-06, + "loss": 0.6008, + "step": 8364 + }, + { + "epoch": 2.9371488764044944, + "grad_norm": 0.8876915574073792, + "learning_rate": 2.6058408430112593e-06, + "loss": 0.5459, + "step": 8365 + }, + { + "epoch": 2.9375, + "grad_norm": 0.8911769986152649, + "learning_rate": 2.6053789304509062e-06, + "loss": 0.5847, + "step": 8366 + }, + { + "epoch": 2.9378511235955056, + "grad_norm": 0.8704618811607361, + "learning_rate": 2.6049170142866846e-06, + "loss": 0.5447, + "step": 8367 + }, + { + "epoch": 2.938202247191011, + "grad_norm": 0.8904538154602051, + "learning_rate": 2.6044550945343903e-06, + "loss": 0.5853, + "step": 8368 + }, + { + "epoch": 2.938553370786517, + "grad_norm": 0.8736263513565063, + "learning_rate": 2.6039931712098214e-06, + "loss": 0.5332, + "step": 8369 + }, + { + "epoch": 2.9389044943820224, + "grad_norm": 0.8789374828338623, + "learning_rate": 2.6035312443287745e-06, + "loss": 0.5604, + "step": 8370 + }, + { + "epoch": 2.939255617977528, + "grad_norm": 0.9802852869033813, + "learning_rate": 2.6030693139070483e-06, + "loss": 0.5817, + "step": 8371 + }, + { + "epoch": 2.9396067415730336, + "grad_norm": 0.9311314225196838, + "learning_rate": 2.6026073799604397e-06, + "loss": 0.5933, + "step": 8372 + }, + { + "epoch": 2.939957865168539, + "grad_norm": 0.8919070363044739, + "learning_rate": 2.6021454425047464e-06, + "loss": 0.5854, + "step": 8373 + }, + { + "epoch": 2.940308988764045, + "grad_norm": 0.9426806569099426, + "learning_rate": 2.601683501555767e-06, + "loss": 0.5826, + "step": 8374 + }, + { + "epoch": 2.9406601123595504, + "grad_norm": 0.9031126499176025, + "learning_rate": 2.6012215571292982e-06, + "loss": 0.5632, + "step": 8375 + }, + { + "epoch": 2.941011235955056, + "grad_norm": 0.911989152431488, + "learning_rate": 2.6007596092411385e-06, + "loss": 0.5758, + "step": 8376 + }, + { + "epoch": 2.9413623595505616, + "grad_norm": 0.8897982835769653, + "learning_rate": 2.600297657907087e-06, + "loss": 0.5613, + "step": 8377 + }, + { + "epoch": 2.941713483146067, + "grad_norm": 0.8923067450523376, + "learning_rate": 2.599835703142941e-06, + "loss": 0.5791, + "step": 8378 + }, + { + "epoch": 2.9420646067415728, + "grad_norm": 0.8883963823318481, + "learning_rate": 2.5993737449644996e-06, + "loss": 0.5702, + "step": 8379 + }, + { + "epoch": 2.942415730337079, + "grad_norm": 0.9059048295021057, + "learning_rate": 2.598911783387561e-06, + "loss": 0.5694, + "step": 8380 + }, + { + "epoch": 2.9427668539325844, + "grad_norm": 0.9108198881149292, + "learning_rate": 2.5984498184279245e-06, + "loss": 0.5597, + "step": 8381 + }, + { + "epoch": 2.94311797752809, + "grad_norm": 0.9599053859710693, + "learning_rate": 2.597987850101388e-06, + "loss": 0.5802, + "step": 8382 + }, + { + "epoch": 2.9434691011235956, + "grad_norm": 0.9044753909111023, + "learning_rate": 2.5975258784237504e-06, + "loss": 0.5982, + "step": 8383 + }, + { + "epoch": 2.943820224719101, + "grad_norm": 0.9053015112876892, + "learning_rate": 2.597063903410812e-06, + "loss": 0.5958, + "step": 8384 + }, + { + "epoch": 2.944171348314607, + "grad_norm": 0.9355586767196655, + "learning_rate": 2.596601925078371e-06, + "loss": 0.5989, + "step": 8385 + }, + { + "epoch": 2.9445224719101124, + "grad_norm": 0.922772228717804, + "learning_rate": 2.5961399434422268e-06, + "loss": 0.5747, + "step": 8386 + }, + { + "epoch": 2.944873595505618, + "grad_norm": 0.9055498242378235, + "learning_rate": 2.5956779585181786e-06, + "loss": 0.5814, + "step": 8387 + }, + { + "epoch": 2.9452247191011236, + "grad_norm": 0.8994905948638916, + "learning_rate": 2.5952159703220265e-06, + "loss": 0.5824, + "step": 8388 + }, + { + "epoch": 2.945575842696629, + "grad_norm": 0.8950543999671936, + "learning_rate": 2.59475397886957e-06, + "loss": 0.5701, + "step": 8389 + }, + { + "epoch": 2.945926966292135, + "grad_norm": 0.9080302119255066, + "learning_rate": 2.5942919841766072e-06, + "loss": 0.6001, + "step": 8390 + }, + { + "epoch": 2.9462780898876404, + "grad_norm": 0.8889274001121521, + "learning_rate": 2.59382998625894e-06, + "loss": 0.5789, + "step": 8391 + }, + { + "epoch": 2.946629213483146, + "grad_norm": 0.8817432522773743, + "learning_rate": 2.5933679851323678e-06, + "loss": 0.6187, + "step": 8392 + }, + { + "epoch": 2.9469803370786516, + "grad_norm": 0.9191221594810486, + "learning_rate": 2.59290598081269e-06, + "loss": 0.608, + "step": 8393 + }, + { + "epoch": 2.947331460674157, + "grad_norm": 0.9250878095626831, + "learning_rate": 2.592443973315707e-06, + "loss": 0.6095, + "step": 8394 + }, + { + "epoch": 2.947682584269663, + "grad_norm": 0.9102625250816345, + "learning_rate": 2.5919819626572196e-06, + "loss": 0.5774, + "step": 8395 + }, + { + "epoch": 2.9480337078651684, + "grad_norm": 0.9281505346298218, + "learning_rate": 2.5915199488530274e-06, + "loss": 0.5857, + "step": 8396 + }, + { + "epoch": 2.9483848314606744, + "grad_norm": 0.9040842056274414, + "learning_rate": 2.5910579319189317e-06, + "loss": 0.5689, + "step": 8397 + }, + { + "epoch": 2.94873595505618, + "grad_norm": 0.920558750629425, + "learning_rate": 2.5905959118707324e-06, + "loss": 0.5957, + "step": 8398 + }, + { + "epoch": 2.9490870786516856, + "grad_norm": 0.8976204991340637, + "learning_rate": 2.590133888724231e-06, + "loss": 0.5538, + "step": 8399 + }, + { + "epoch": 2.949438202247191, + "grad_norm": 0.9071747064590454, + "learning_rate": 2.5896718624952273e-06, + "loss": 0.5895, + "step": 8400 + }, + { + "epoch": 2.949789325842697, + "grad_norm": 0.8880211710929871, + "learning_rate": 2.5892098331995225e-06, + "loss": 0.594, + "step": 8401 + }, + { + "epoch": 2.9501404494382024, + "grad_norm": 0.9073636531829834, + "learning_rate": 2.5887478008529178e-06, + "loss": 0.5893, + "step": 8402 + }, + { + "epoch": 2.950491573033708, + "grad_norm": 0.9249823689460754, + "learning_rate": 2.5882857654712144e-06, + "loss": 0.5878, + "step": 8403 + }, + { + "epoch": 2.9508426966292136, + "grad_norm": 0.9058404564857483, + "learning_rate": 2.5878237270702133e-06, + "loss": 0.5781, + "step": 8404 + }, + { + "epoch": 2.951193820224719, + "grad_norm": 0.881600558757782, + "learning_rate": 2.5873616856657156e-06, + "loss": 0.5717, + "step": 8405 + }, + { + "epoch": 2.951544943820225, + "grad_norm": 0.8910747170448303, + "learning_rate": 2.586899641273523e-06, + "loss": 0.5596, + "step": 8406 + }, + { + "epoch": 2.9518960674157304, + "grad_norm": 0.9196321964263916, + "learning_rate": 2.586437593909438e-06, + "loss": 0.5802, + "step": 8407 + }, + { + "epoch": 2.952247191011236, + "grad_norm": 0.8868768811225891, + "learning_rate": 2.58597554358926e-06, + "loss": 0.5875, + "step": 8408 + }, + { + "epoch": 2.9525983146067416, + "grad_norm": 0.9195485711097717, + "learning_rate": 2.5855134903287924e-06, + "loss": 0.5398, + "step": 8409 + }, + { + "epoch": 2.952949438202247, + "grad_norm": 0.901940643787384, + "learning_rate": 2.5850514341438366e-06, + "loss": 0.5605, + "step": 8410 + }, + { + "epoch": 2.953300561797753, + "grad_norm": 0.9271135330200195, + "learning_rate": 2.584589375050195e-06, + "loss": 0.5873, + "step": 8411 + }, + { + "epoch": 2.9536516853932584, + "grad_norm": 0.8865430951118469, + "learning_rate": 2.584127313063669e-06, + "loss": 0.6156, + "step": 8412 + }, + { + "epoch": 2.954002808988764, + "grad_norm": 0.9074857234954834, + "learning_rate": 2.5836652482000608e-06, + "loss": 0.5827, + "step": 8413 + }, + { + "epoch": 2.9543539325842696, + "grad_norm": 0.9082297086715698, + "learning_rate": 2.583203180475172e-06, + "loss": 0.5923, + "step": 8414 + }, + { + "epoch": 2.954705056179775, + "grad_norm": 0.9126758575439453, + "learning_rate": 2.5827411099048067e-06, + "loss": 0.5807, + "step": 8415 + }, + { + "epoch": 2.955056179775281, + "grad_norm": 0.8857371211051941, + "learning_rate": 2.5822790365047656e-06, + "loss": 0.5606, + "step": 8416 + }, + { + "epoch": 2.9554073033707864, + "grad_norm": 0.8784708380699158, + "learning_rate": 2.581816960290852e-06, + "loss": 0.5499, + "step": 8417 + }, + { + "epoch": 2.955758426966292, + "grad_norm": 0.8785218000411987, + "learning_rate": 2.5813548812788688e-06, + "loss": 0.5617, + "step": 8418 + }, + { + "epoch": 2.9561095505617976, + "grad_norm": 0.9093417525291443, + "learning_rate": 2.580892799484617e-06, + "loss": 0.5524, + "step": 8419 + }, + { + "epoch": 2.956460674157303, + "grad_norm": 0.9172379374504089, + "learning_rate": 2.580430714923902e-06, + "loss": 0.5793, + "step": 8420 + }, + { + "epoch": 2.956811797752809, + "grad_norm": 0.8907975554466248, + "learning_rate": 2.5799686276125248e-06, + "loss": 0.5758, + "step": 8421 + }, + { + "epoch": 2.9571629213483144, + "grad_norm": 0.8963255286216736, + "learning_rate": 2.5795065375662893e-06, + "loss": 0.6042, + "step": 8422 + }, + { + "epoch": 2.95751404494382, + "grad_norm": 0.8777085542678833, + "learning_rate": 2.5790444448009976e-06, + "loss": 0.5429, + "step": 8423 + }, + { + "epoch": 2.9578651685393256, + "grad_norm": 0.9219308495521545, + "learning_rate": 2.5785823493324545e-06, + "loss": 0.6047, + "step": 8424 + }, + { + "epoch": 2.9582162921348316, + "grad_norm": 0.8927537798881531, + "learning_rate": 2.5781202511764618e-06, + "loss": 0.593, + "step": 8425 + }, + { + "epoch": 2.958567415730337, + "grad_norm": 0.9076624512672424, + "learning_rate": 2.5776581503488233e-06, + "loss": 0.5738, + "step": 8426 + }, + { + "epoch": 2.958918539325843, + "grad_norm": 0.8665882349014282, + "learning_rate": 2.5771960468653428e-06, + "loss": 0.5405, + "step": 8427 + }, + { + "epoch": 2.9592696629213484, + "grad_norm": 0.9200811982154846, + "learning_rate": 2.576733940741823e-06, + "loss": 0.5778, + "step": 8428 + }, + { + "epoch": 2.959620786516854, + "grad_norm": 0.9071989059448242, + "learning_rate": 2.5762718319940683e-06, + "loss": 0.5573, + "step": 8429 + }, + { + "epoch": 2.9599719101123596, + "grad_norm": 0.9085105657577515, + "learning_rate": 2.5758097206378824e-06, + "loss": 0.5623, + "step": 8430 + }, + { + "epoch": 2.960323033707865, + "grad_norm": 0.9189997315406799, + "learning_rate": 2.5753476066890688e-06, + "loss": 0.5959, + "step": 8431 + }, + { + "epoch": 2.960674157303371, + "grad_norm": 0.8904687762260437, + "learning_rate": 2.5748854901634314e-06, + "loss": 0.5458, + "step": 8432 + }, + { + "epoch": 2.9610252808988764, + "grad_norm": 0.8914241194725037, + "learning_rate": 2.5744233710767747e-06, + "loss": 0.563, + "step": 8433 + }, + { + "epoch": 2.961376404494382, + "grad_norm": 0.897879958152771, + "learning_rate": 2.5739612494449017e-06, + "loss": 0.5489, + "step": 8434 + }, + { + "epoch": 2.9617275280898876, + "grad_norm": 0.9113520383834839, + "learning_rate": 2.5734991252836177e-06, + "loss": 0.588, + "step": 8435 + }, + { + "epoch": 2.962078651685393, + "grad_norm": 0.9117496013641357, + "learning_rate": 2.5730369986087265e-06, + "loss": 0.5864, + "step": 8436 + }, + { + "epoch": 2.962429775280899, + "grad_norm": 0.9213951230049133, + "learning_rate": 2.572574869436033e-06, + "loss": 0.607, + "step": 8437 + }, + { + "epoch": 2.9627808988764044, + "grad_norm": 0.9336210489273071, + "learning_rate": 2.5721127377813403e-06, + "loss": 0.5933, + "step": 8438 + }, + { + "epoch": 2.96313202247191, + "grad_norm": 0.892371654510498, + "learning_rate": 2.5716506036604544e-06, + "loss": 0.5657, + "step": 8439 + }, + { + "epoch": 2.9634831460674156, + "grad_norm": 0.941336452960968, + "learning_rate": 2.5711884670891784e-06, + "loss": 0.5791, + "step": 8440 + }, + { + "epoch": 2.963834269662921, + "grad_norm": 0.9672850966453552, + "learning_rate": 2.570726328083318e-06, + "loss": 0.6004, + "step": 8441 + }, + { + "epoch": 2.9641853932584272, + "grad_norm": 1.0078755617141724, + "learning_rate": 2.5702641866586775e-06, + "loss": 0.5591, + "step": 8442 + }, + { + "epoch": 2.964536516853933, + "grad_norm": 0.8987733125686646, + "learning_rate": 2.569802042831062e-06, + "loss": 0.5668, + "step": 8443 + }, + { + "epoch": 2.9648876404494384, + "grad_norm": 0.9090468883514404, + "learning_rate": 2.569339896616277e-06, + "loss": 0.5477, + "step": 8444 + }, + { + "epoch": 2.965238764044944, + "grad_norm": 0.9281615018844604, + "learning_rate": 2.5688777480301263e-06, + "loss": 0.6123, + "step": 8445 + }, + { + "epoch": 2.9655898876404496, + "grad_norm": 0.9032601118087769, + "learning_rate": 2.5684155970884154e-06, + "loss": 0.6001, + "step": 8446 + }, + { + "epoch": 2.965941011235955, + "grad_norm": 0.923275887966156, + "learning_rate": 2.5679534438069497e-06, + "loss": 0.5842, + "step": 8447 + }, + { + "epoch": 2.966292134831461, + "grad_norm": 0.8700858950614929, + "learning_rate": 2.5674912882015347e-06, + "loss": 0.5988, + "step": 8448 + }, + { + "epoch": 2.9666432584269664, + "grad_norm": 0.8954595923423767, + "learning_rate": 2.5670291302879753e-06, + "loss": 0.5682, + "step": 8449 + }, + { + "epoch": 2.966994382022472, + "grad_norm": 0.9107581377029419, + "learning_rate": 2.566566970082077e-06, + "loss": 0.5654, + "step": 8450 + }, + { + "epoch": 2.9673455056179776, + "grad_norm": 0.9228717684745789, + "learning_rate": 2.566104807599646e-06, + "loss": 0.6192, + "step": 8451 + }, + { + "epoch": 2.967696629213483, + "grad_norm": 0.8883110284805298, + "learning_rate": 2.5656426428564858e-06, + "loss": 0.5736, + "step": 8452 + }, + { + "epoch": 2.968047752808989, + "grad_norm": 0.8976356983184814, + "learning_rate": 2.565180475868404e-06, + "loss": 0.5666, + "step": 8453 + }, + { + "epoch": 2.9683988764044944, + "grad_norm": 0.8793136477470398, + "learning_rate": 2.5647183066512056e-06, + "loss": 0.5724, + "step": 8454 + }, + { + "epoch": 2.96875, + "grad_norm": 0.9236953258514404, + "learning_rate": 2.564256135220697e-06, + "loss": 0.5579, + "step": 8455 + }, + { + "epoch": 2.9691011235955056, + "grad_norm": 0.9253345131874084, + "learning_rate": 2.563793961592683e-06, + "loss": 0.611, + "step": 8456 + }, + { + "epoch": 2.969452247191011, + "grad_norm": 0.9224141240119934, + "learning_rate": 2.56333178578297e-06, + "loss": 0.6227, + "step": 8457 + }, + { + "epoch": 2.969803370786517, + "grad_norm": 0.8846723437309265, + "learning_rate": 2.562869607807364e-06, + "loss": 0.5536, + "step": 8458 + }, + { + "epoch": 2.9701544943820224, + "grad_norm": 0.954719066619873, + "learning_rate": 2.562407427681672e-06, + "loss": 0.5988, + "step": 8459 + }, + { + "epoch": 2.970505617977528, + "grad_norm": 0.8987382054328918, + "learning_rate": 2.5619452454216993e-06, + "loss": 0.562, + "step": 8460 + }, + { + "epoch": 2.9708567415730336, + "grad_norm": 0.9086666107177734, + "learning_rate": 2.561483061043252e-06, + "loss": 0.5637, + "step": 8461 + }, + { + "epoch": 2.971207865168539, + "grad_norm": 0.910751461982727, + "learning_rate": 2.561020874562136e-06, + "loss": 0.5843, + "step": 8462 + }, + { + "epoch": 2.971558988764045, + "grad_norm": 0.90587317943573, + "learning_rate": 2.560558685994159e-06, + "loss": 0.6149, + "step": 8463 + }, + { + "epoch": 2.9719101123595504, + "grad_norm": 0.8958348035812378, + "learning_rate": 2.5600964953551266e-06, + "loss": 0.5974, + "step": 8464 + }, + { + "epoch": 2.972261235955056, + "grad_norm": 0.9219935536384583, + "learning_rate": 2.559634302660845e-06, + "loss": 0.5913, + "step": 8465 + }, + { + "epoch": 2.9726123595505616, + "grad_norm": 0.8740019798278809, + "learning_rate": 2.559172107927122e-06, + "loss": 0.5545, + "step": 8466 + }, + { + "epoch": 2.972963483146067, + "grad_norm": 0.9145710468292236, + "learning_rate": 2.5587099111697626e-06, + "loss": 0.5856, + "step": 8467 + }, + { + "epoch": 2.9733146067415728, + "grad_norm": 0.8794683218002319, + "learning_rate": 2.558247712404576e-06, + "loss": 0.582, + "step": 8468 + }, + { + "epoch": 2.973665730337079, + "grad_norm": 0.9286167621612549, + "learning_rate": 2.5577855116473665e-06, + "loss": 0.5837, + "step": 8469 + }, + { + "epoch": 2.9740168539325844, + "grad_norm": 0.9220527410507202, + "learning_rate": 2.557323308913942e-06, + "loss": 0.5854, + "step": 8470 + }, + { + "epoch": 2.97436797752809, + "grad_norm": 0.8931424617767334, + "learning_rate": 2.5568611042201095e-06, + "loss": 0.5559, + "step": 8471 + }, + { + "epoch": 2.9747191011235956, + "grad_norm": 0.8948701024055481, + "learning_rate": 2.5563988975816756e-06, + "loss": 0.563, + "step": 8472 + }, + { + "epoch": 2.975070224719101, + "grad_norm": 0.92146235704422, + "learning_rate": 2.5559366890144482e-06, + "loss": 0.5907, + "step": 8473 + }, + { + "epoch": 2.975421348314607, + "grad_norm": 0.8765264749526978, + "learning_rate": 2.5554744785342333e-06, + "loss": 0.5517, + "step": 8474 + }, + { + "epoch": 2.9757724719101124, + "grad_norm": 0.8799095153808594, + "learning_rate": 2.5550122661568395e-06, + "loss": 0.529, + "step": 8475 + }, + { + "epoch": 2.976123595505618, + "grad_norm": 0.9077634811401367, + "learning_rate": 2.5545500518980724e-06, + "loss": 0.587, + "step": 8476 + }, + { + "epoch": 2.9764747191011236, + "grad_norm": 0.9179309010505676, + "learning_rate": 2.5540878357737407e-06, + "loss": 0.6005, + "step": 8477 + }, + { + "epoch": 2.976825842696629, + "grad_norm": 0.9103131890296936, + "learning_rate": 2.5536256177996517e-06, + "loss": 0.5739, + "step": 8478 + }, + { + "epoch": 2.977176966292135, + "grad_norm": 0.9008108973503113, + "learning_rate": 2.5531633979916114e-06, + "loss": 0.5644, + "step": 8479 + }, + { + "epoch": 2.9775280898876404, + "grad_norm": 0.9275205135345459, + "learning_rate": 2.5527011763654292e-06, + "loss": 0.6192, + "step": 8480 + }, + { + "epoch": 2.977879213483146, + "grad_norm": 0.8984926342964172, + "learning_rate": 2.5522389529369113e-06, + "loss": 0.6246, + "step": 8481 + }, + { + "epoch": 2.9782303370786516, + "grad_norm": 0.8759105801582336, + "learning_rate": 2.5517767277218663e-06, + "loss": 0.5709, + "step": 8482 + }, + { + "epoch": 2.978581460674157, + "grad_norm": 0.9202468991279602, + "learning_rate": 2.5513145007361014e-06, + "loss": 0.5719, + "step": 8483 + }, + { + "epoch": 2.978932584269663, + "grad_norm": 0.9187071919441223, + "learning_rate": 2.5508522719954244e-06, + "loss": 0.5936, + "step": 8484 + }, + { + "epoch": 2.9792837078651684, + "grad_norm": 0.9312149286270142, + "learning_rate": 2.5503900415156435e-06, + "loss": 0.581, + "step": 8485 + }, + { + "epoch": 2.9796348314606744, + "grad_norm": 0.8901807069778442, + "learning_rate": 2.5499278093125656e-06, + "loss": 0.556, + "step": 8486 + }, + { + "epoch": 2.97998595505618, + "grad_norm": 0.8926215767860413, + "learning_rate": 2.549465575402e-06, + "loss": 0.589, + "step": 8487 + }, + { + "epoch": 2.9803370786516856, + "grad_norm": 0.8843978047370911, + "learning_rate": 2.5490033397997545e-06, + "loss": 0.5611, + "step": 8488 + }, + { + "epoch": 2.980688202247191, + "grad_norm": 0.9020524024963379, + "learning_rate": 2.548541102521636e-06, + "loss": 0.6027, + "step": 8489 + }, + { + "epoch": 2.981039325842697, + "grad_norm": 0.9096185564994812, + "learning_rate": 2.5480788635834535e-06, + "loss": 0.5814, + "step": 8490 + }, + { + "epoch": 2.9813904494382024, + "grad_norm": 0.9102569818496704, + "learning_rate": 2.5476166230010146e-06, + "loss": 0.5652, + "step": 8491 + }, + { + "epoch": 2.981741573033708, + "grad_norm": 0.9763215184211731, + "learning_rate": 2.5471543807901285e-06, + "loss": 0.5475, + "step": 8492 + }, + { + "epoch": 2.9820926966292136, + "grad_norm": 0.8967247009277344, + "learning_rate": 2.5466921369666026e-06, + "loss": 0.5938, + "step": 8493 + }, + { + "epoch": 2.982443820224719, + "grad_norm": 0.9018954038619995, + "learning_rate": 2.5462298915462456e-06, + "loss": 0.5803, + "step": 8494 + }, + { + "epoch": 2.982794943820225, + "grad_norm": 0.9286206364631653, + "learning_rate": 2.545767644544866e-06, + "loss": 0.6389, + "step": 8495 + }, + { + "epoch": 2.9831460674157304, + "grad_norm": 0.8892673254013062, + "learning_rate": 2.545305395978272e-06, + "loss": 0.5943, + "step": 8496 + }, + { + "epoch": 2.983497191011236, + "grad_norm": 0.9189037680625916, + "learning_rate": 2.5448431458622722e-06, + "loss": 0.5788, + "step": 8497 + }, + { + "epoch": 2.9838483146067416, + "grad_norm": 0.8956021666526794, + "learning_rate": 2.5443808942126752e-06, + "loss": 0.6166, + "step": 8498 + }, + { + "epoch": 2.984199438202247, + "grad_norm": 0.9188178777694702, + "learning_rate": 2.5439186410452897e-06, + "loss": 0.6135, + "step": 8499 + }, + { + "epoch": 2.984550561797753, + "grad_norm": 0.9242973923683167, + "learning_rate": 2.543456386375925e-06, + "loss": 0.6029, + "step": 8500 + }, + { + "epoch": 2.9849016853932584, + "grad_norm": 0.8959596753120422, + "learning_rate": 2.542994130220388e-06, + "loss": 0.5787, + "step": 8501 + }, + { + "epoch": 2.985252808988764, + "grad_norm": 0.907460629940033, + "learning_rate": 2.542531872594489e-06, + "loss": 0.5718, + "step": 8502 + }, + { + "epoch": 2.9856039325842696, + "grad_norm": 0.8827933073043823, + "learning_rate": 2.542069613514036e-06, + "loss": 0.5793, + "step": 8503 + }, + { + "epoch": 2.985955056179775, + "grad_norm": 0.9433537125587463, + "learning_rate": 2.541607352994839e-06, + "loss": 0.5857, + "step": 8504 + }, + { + "epoch": 2.986306179775281, + "grad_norm": 0.9445831775665283, + "learning_rate": 2.5411450910527052e-06, + "loss": 0.5895, + "step": 8505 + }, + { + "epoch": 2.9866573033707864, + "grad_norm": 0.9128881692886353, + "learning_rate": 2.5406828277034445e-06, + "loss": 0.5699, + "step": 8506 + }, + { + "epoch": 2.987008426966292, + "grad_norm": 0.8813773989677429, + "learning_rate": 2.540220562962866e-06, + "loss": 0.5767, + "step": 8507 + }, + { + "epoch": 2.9873595505617976, + "grad_norm": 0.9386189579963684, + "learning_rate": 2.5397582968467794e-06, + "loss": 0.5999, + "step": 8508 + }, + { + "epoch": 2.987710674157303, + "grad_norm": 0.908528208732605, + "learning_rate": 2.5392960293709923e-06, + "loss": 0.5789, + "step": 8509 + }, + { + "epoch": 2.988061797752809, + "grad_norm": 0.8872787952423096, + "learning_rate": 2.5388337605513145e-06, + "loss": 0.5852, + "step": 8510 + }, + { + "epoch": 2.9884129213483144, + "grad_norm": 0.8956595659255981, + "learning_rate": 2.538371490403556e-06, + "loss": 0.5793, + "step": 8511 + }, + { + "epoch": 2.98876404494382, + "grad_norm": 0.9266185164451599, + "learning_rate": 2.537909218943525e-06, + "loss": 0.5626, + "step": 8512 + }, + { + "epoch": 2.9891151685393256, + "grad_norm": 0.9429141879081726, + "learning_rate": 2.537446946187032e-06, + "loss": 0.5939, + "step": 8513 + }, + { + "epoch": 2.9894662921348316, + "grad_norm": 0.9124779105186462, + "learning_rate": 2.536984672149885e-06, + "loss": 0.5527, + "step": 8514 + }, + { + "epoch": 2.989817415730337, + "grad_norm": 0.9191445708274841, + "learning_rate": 2.5365223968478932e-06, + "loss": 0.5964, + "step": 8515 + }, + { + "epoch": 2.990168539325843, + "grad_norm": 0.8742399215698242, + "learning_rate": 2.536060120296867e-06, + "loss": 0.5685, + "step": 8516 + }, + { + "epoch": 2.9905196629213484, + "grad_norm": 0.9216606616973877, + "learning_rate": 2.5355978425126154e-06, + "loss": 0.5507, + "step": 8517 + }, + { + "epoch": 2.990870786516854, + "grad_norm": 0.905522882938385, + "learning_rate": 2.535135563510948e-06, + "loss": 0.6131, + "step": 8518 + }, + { + "epoch": 2.9912219101123596, + "grad_norm": 0.8910292387008667, + "learning_rate": 2.5346732833076748e-06, + "loss": 0.567, + "step": 8519 + }, + { + "epoch": 2.991573033707865, + "grad_norm": 0.8611335158348083, + "learning_rate": 2.534211001918605e-06, + "loss": 0.5022, + "step": 8520 + }, + { + "epoch": 2.991924157303371, + "grad_norm": 0.9413619637489319, + "learning_rate": 2.533748719359548e-06, + "loss": 0.5868, + "step": 8521 + }, + { + "epoch": 2.9922752808988764, + "grad_norm": 0.8834371566772461, + "learning_rate": 2.5332864356463136e-06, + "loss": 0.5503, + "step": 8522 + }, + { + "epoch": 2.992626404494382, + "grad_norm": 0.9130658507347107, + "learning_rate": 2.5328241507947116e-06, + "loss": 0.5899, + "step": 8523 + }, + { + "epoch": 2.9929775280898876, + "grad_norm": 0.9253802299499512, + "learning_rate": 2.5323618648205516e-06, + "loss": 0.5994, + "step": 8524 + }, + { + "epoch": 2.993328651685393, + "grad_norm": 0.9231074452400208, + "learning_rate": 2.5318995777396443e-06, + "loss": 0.616, + "step": 8525 + }, + { + "epoch": 2.993679775280899, + "grad_norm": 0.930930495262146, + "learning_rate": 2.5314372895677985e-06, + "loss": 0.5891, + "step": 8526 + }, + { + "epoch": 2.9940308988764044, + "grad_norm": 0.9062270522117615, + "learning_rate": 2.530975000320824e-06, + "loss": 0.6013, + "step": 8527 + }, + { + "epoch": 2.99438202247191, + "grad_norm": 0.9265364408493042, + "learning_rate": 2.530512710014531e-06, + "loss": 0.6052, + "step": 8528 + }, + { + "epoch": 2.9947331460674156, + "grad_norm": 0.8931238651275635, + "learning_rate": 2.530050418664729e-06, + "loss": 0.5823, + "step": 8529 + }, + { + "epoch": 2.995084269662921, + "grad_norm": 0.9220635890960693, + "learning_rate": 2.529588126287229e-06, + "loss": 0.5802, + "step": 8530 + }, + { + "epoch": 2.9954353932584272, + "grad_norm": 0.9293198585510254, + "learning_rate": 2.5291258328978403e-06, + "loss": 0.5697, + "step": 8531 + }, + { + "epoch": 2.995786516853933, + "grad_norm": 0.9374073147773743, + "learning_rate": 2.5286635385123725e-06, + "loss": 0.6225, + "step": 8532 + }, + { + "epoch": 2.9961376404494384, + "grad_norm": 0.9082289934158325, + "learning_rate": 2.528201243146637e-06, + "loss": 0.6163, + "step": 8533 + }, + { + "epoch": 2.996488764044944, + "grad_norm": 0.9165958762168884, + "learning_rate": 2.5277389468164428e-06, + "loss": 0.5803, + "step": 8534 + }, + { + "epoch": 2.9968398876404496, + "grad_norm": 0.8821650743484497, + "learning_rate": 2.5272766495376e-06, + "loss": 0.5607, + "step": 8535 + }, + { + "epoch": 2.997191011235955, + "grad_norm": 0.9156829714775085, + "learning_rate": 2.52681435132592e-06, + "loss": 0.5973, + "step": 8536 + }, + { + "epoch": 2.997542134831461, + "grad_norm": 0.8864458799362183, + "learning_rate": 2.526352052197211e-06, + "loss": 0.5616, + "step": 8537 + }, + { + "epoch": 2.9978932584269664, + "grad_norm": 0.8939210772514343, + "learning_rate": 2.5258897521672855e-06, + "loss": 0.585, + "step": 8538 + }, + { + "epoch": 2.998244382022472, + "grad_norm": 0.915823221206665, + "learning_rate": 2.5254274512519522e-06, + "loss": 0.5739, + "step": 8539 + }, + { + "epoch": 2.9985955056179776, + "grad_norm": 0.8936715722084045, + "learning_rate": 2.5249651494670214e-06, + "loss": 0.5622, + "step": 8540 + }, + { + "epoch": 2.998946629213483, + "grad_norm": 0.8853635191917419, + "learning_rate": 2.524502846828304e-06, + "loss": 0.5633, + "step": 8541 + }, + { + "epoch": 2.999297752808989, + "grad_norm": 0.8832322359085083, + "learning_rate": 2.5240405433516107e-06, + "loss": 0.5759, + "step": 8542 + }, + { + "epoch": 2.9996488764044944, + "grad_norm": 0.9333004951477051, + "learning_rate": 2.523578239052751e-06, + "loss": 0.6088, + "step": 8543 + }, + { + "epoch": 3.0, + "grad_norm": 0.9140715599060059, + "learning_rate": 2.5231159339475352e-06, + "loss": 0.5582, + "step": 8544 + }, + { + "epoch": 3.0003511235955056, + "grad_norm": 0.892126202583313, + "learning_rate": 2.522653628051775e-06, + "loss": 0.5616, + "step": 8545 + }, + { + "epoch": 3.000702247191011, + "grad_norm": 0.9008687734603882, + "learning_rate": 2.52219132138128e-06, + "loss": 0.5369, + "step": 8546 + }, + { + "epoch": 3.001053370786517, + "grad_norm": 0.8750908374786377, + "learning_rate": 2.5217290139518614e-06, + "loss": 0.554, + "step": 8547 + }, + { + "epoch": 3.0014044943820224, + "grad_norm": 0.8911838531494141, + "learning_rate": 2.5212667057793285e-06, + "loss": 0.5522, + "step": 8548 + }, + { + "epoch": 3.001755617977528, + "grad_norm": 0.8570799827575684, + "learning_rate": 2.520804396879492e-06, + "loss": 0.5216, + "step": 8549 + }, + { + "epoch": 3.0021067415730336, + "grad_norm": 0.8868703842163086, + "learning_rate": 2.520342087268163e-06, + "loss": 0.5455, + "step": 8550 + }, + { + "epoch": 3.002457865168539, + "grad_norm": 0.8640176057815552, + "learning_rate": 2.5198797769611525e-06, + "loss": 0.5282, + "step": 8551 + }, + { + "epoch": 3.002808988764045, + "grad_norm": 0.9003017544746399, + "learning_rate": 2.5194174659742705e-06, + "loss": 0.5253, + "step": 8552 + }, + { + "epoch": 3.0031601123595504, + "grad_norm": 0.9024666547775269, + "learning_rate": 2.5189551543233274e-06, + "loss": 0.5235, + "step": 8553 + }, + { + "epoch": 3.003511235955056, + "grad_norm": 0.920034646987915, + "learning_rate": 2.5184928420241343e-06, + "loss": 0.5783, + "step": 8554 + }, + { + "epoch": 3.0038623595505616, + "grad_norm": 0.9136059880256653, + "learning_rate": 2.5180305290925027e-06, + "loss": 0.5321, + "step": 8555 + }, + { + "epoch": 3.0042134831460676, + "grad_norm": 0.9276214838027954, + "learning_rate": 2.5175682155442422e-06, + "loss": 0.5692, + "step": 8556 + }, + { + "epoch": 3.004564606741573, + "grad_norm": 0.9201930165290833, + "learning_rate": 2.517105901395163e-06, + "loss": 0.5768, + "step": 8557 + }, + { + "epoch": 3.004915730337079, + "grad_norm": 0.9409593343734741, + "learning_rate": 2.516643586661077e-06, + "loss": 0.5564, + "step": 8558 + }, + { + "epoch": 3.0052668539325844, + "grad_norm": 0.8970922827720642, + "learning_rate": 2.516181271357795e-06, + "loss": 0.5565, + "step": 8559 + }, + { + "epoch": 3.00561797752809, + "grad_norm": 0.9575510025024414, + "learning_rate": 2.515718955501127e-06, + "loss": 0.5496, + "step": 8560 + }, + { + "epoch": 3.0059691011235956, + "grad_norm": 0.9126796126365662, + "learning_rate": 2.515256639106884e-06, + "loss": 0.5501, + "step": 8561 + }, + { + "epoch": 3.006320224719101, + "grad_norm": 0.8977692723274231, + "learning_rate": 2.514794322190877e-06, + "loss": 0.5686, + "step": 8562 + }, + { + "epoch": 3.006671348314607, + "grad_norm": 0.9079583883285522, + "learning_rate": 2.5143320047689173e-06, + "loss": 0.5532, + "step": 8563 + }, + { + "epoch": 3.0070224719101124, + "grad_norm": 0.9110571146011353, + "learning_rate": 2.513869686856815e-06, + "loss": 0.5393, + "step": 8564 + }, + { + "epoch": 3.007373595505618, + "grad_norm": 0.911446750164032, + "learning_rate": 2.5134073684703815e-06, + "loss": 0.5363, + "step": 8565 + }, + { + "epoch": 3.0077247191011236, + "grad_norm": 0.9001997113227844, + "learning_rate": 2.512945049625427e-06, + "loss": 0.5651, + "step": 8566 + }, + { + "epoch": 3.008075842696629, + "grad_norm": 0.9125750064849854, + "learning_rate": 2.512482730337763e-06, + "loss": 0.5422, + "step": 8567 + }, + { + "epoch": 3.008426966292135, + "grad_norm": 0.9246050715446472, + "learning_rate": 2.5120204106232006e-06, + "loss": 0.5719, + "step": 8568 + }, + { + "epoch": 3.0087780898876404, + "grad_norm": 0.8888805508613586, + "learning_rate": 2.51155809049755e-06, + "loss": 0.5604, + "step": 8569 + }, + { + "epoch": 3.009129213483146, + "grad_norm": 0.936180591583252, + "learning_rate": 2.5110957699766224e-06, + "loss": 0.5565, + "step": 8570 + }, + { + "epoch": 3.0094803370786516, + "grad_norm": 0.958117663860321, + "learning_rate": 2.51063344907623e-06, + "loss": 0.5782, + "step": 8571 + }, + { + "epoch": 3.009831460674157, + "grad_norm": 0.9103242754936218, + "learning_rate": 2.510171127812182e-06, + "loss": 0.5306, + "step": 8572 + }, + { + "epoch": 3.010182584269663, + "grad_norm": 0.9062320590019226, + "learning_rate": 2.50970880620029e-06, + "loss": 0.5789, + "step": 8573 + }, + { + "epoch": 3.0105337078651684, + "grad_norm": 0.9318956136703491, + "learning_rate": 2.5092464842563652e-06, + "loss": 0.5821, + "step": 8574 + }, + { + "epoch": 3.010884831460674, + "grad_norm": 0.9377192854881287, + "learning_rate": 2.508784161996219e-06, + "loss": 0.5604, + "step": 8575 + }, + { + "epoch": 3.0112359550561796, + "grad_norm": 0.9041692018508911, + "learning_rate": 2.5083218394356624e-06, + "loss": 0.542, + "step": 8576 + }, + { + "epoch": 3.011587078651685, + "grad_norm": 0.8867703080177307, + "learning_rate": 2.507859516590505e-06, + "loss": 0.5294, + "step": 8577 + }, + { + "epoch": 3.011938202247191, + "grad_norm": 0.8993325233459473, + "learning_rate": 2.5073971934765596e-06, + "loss": 0.5414, + "step": 8578 + }, + { + "epoch": 3.012289325842697, + "grad_norm": 0.9128599166870117, + "learning_rate": 2.506934870109636e-06, + "loss": 0.5587, + "step": 8579 + }, + { + "epoch": 3.0126404494382024, + "grad_norm": 0.90801602602005, + "learning_rate": 2.5064725465055455e-06, + "loss": 0.5515, + "step": 8580 + }, + { + "epoch": 3.012991573033708, + "grad_norm": 0.905956506729126, + "learning_rate": 2.5060102226801e-06, + "loss": 0.5198, + "step": 8581 + }, + { + "epoch": 3.0133426966292136, + "grad_norm": 0.9390702843666077, + "learning_rate": 2.50554789864911e-06, + "loss": 0.5917, + "step": 8582 + }, + { + "epoch": 3.013693820224719, + "grad_norm": 0.887134313583374, + "learning_rate": 2.5050855744283864e-06, + "loss": 0.5237, + "step": 8583 + }, + { + "epoch": 3.014044943820225, + "grad_norm": 0.8773907423019409, + "learning_rate": 2.5046232500337403e-06, + "loss": 0.5477, + "step": 8584 + }, + { + "epoch": 3.0143960674157304, + "grad_norm": 0.9288877844810486, + "learning_rate": 2.5041609254809835e-06, + "loss": 0.5513, + "step": 8585 + }, + { + "epoch": 3.014747191011236, + "grad_norm": 0.937523603439331, + "learning_rate": 2.5036986007859266e-06, + "loss": 0.5476, + "step": 8586 + }, + { + "epoch": 3.0150983146067416, + "grad_norm": 0.9449216723442078, + "learning_rate": 2.50323627596438e-06, + "loss": 0.5828, + "step": 8587 + }, + { + "epoch": 3.015449438202247, + "grad_norm": 0.9125329256057739, + "learning_rate": 2.502773951032157e-06, + "loss": 0.5335, + "step": 8588 + }, + { + "epoch": 3.015800561797753, + "grad_norm": 0.9391717910766602, + "learning_rate": 2.5023116260050663e-06, + "loss": 0.5897, + "step": 8589 + }, + { + "epoch": 3.0161516853932584, + "grad_norm": 0.9164744019508362, + "learning_rate": 2.5018493008989197e-06, + "loss": 0.5774, + "step": 8590 + }, + { + "epoch": 3.016502808988764, + "grad_norm": 0.9078039526939392, + "learning_rate": 2.501386975729529e-06, + "loss": 0.543, + "step": 8591 + }, + { + "epoch": 3.0168539325842696, + "grad_norm": 1.451960563659668, + "learning_rate": 2.5009246505127043e-06, + "loss": 0.5468, + "step": 8592 + }, + { + "epoch": 3.017205056179775, + "grad_norm": 0.904181957244873, + "learning_rate": 2.500462325264258e-06, + "loss": 0.5473, + "step": 8593 + }, + { + "epoch": 3.017556179775281, + "grad_norm": 0.8868268132209778, + "learning_rate": 2.5e-06, + "loss": 0.5289, + "step": 8594 + }, + { + "epoch": 3.0179073033707864, + "grad_norm": 0.876323401927948, + "learning_rate": 2.4995376747357434e-06, + "loss": 0.5079, + "step": 8595 + }, + { + "epoch": 3.018258426966292, + "grad_norm": 0.8988457322120667, + "learning_rate": 2.499075349487296e-06, + "loss": 0.5664, + "step": 8596 + }, + { + "epoch": 3.0186095505617976, + "grad_norm": 0.9250447154045105, + "learning_rate": 2.4986130242704728e-06, + "loss": 0.5764, + "step": 8597 + }, + { + "epoch": 3.018960674157303, + "grad_norm": 0.8886809349060059, + "learning_rate": 2.498150699101081e-06, + "loss": 0.5538, + "step": 8598 + }, + { + "epoch": 3.019311797752809, + "grad_norm": 0.9123310446739197, + "learning_rate": 2.4976883739949354e-06, + "loss": 0.5417, + "step": 8599 + }, + { + "epoch": 3.019662921348315, + "grad_norm": 0.9530540108680725, + "learning_rate": 2.497226048967844e-06, + "loss": 0.5364, + "step": 8600 + }, + { + "epoch": 3.0200140449438204, + "grad_norm": 0.9198148250579834, + "learning_rate": 2.4967637240356204e-06, + "loss": 0.5688, + "step": 8601 + }, + { + "epoch": 3.020365168539326, + "grad_norm": 0.888922929763794, + "learning_rate": 2.496301399214074e-06, + "loss": 0.5717, + "step": 8602 + }, + { + "epoch": 3.0207162921348316, + "grad_norm": 0.9528188109397888, + "learning_rate": 2.4958390745190173e-06, + "loss": 0.5885, + "step": 8603 + }, + { + "epoch": 3.021067415730337, + "grad_norm": 0.9578085541725159, + "learning_rate": 2.4953767499662597e-06, + "loss": 0.5584, + "step": 8604 + }, + { + "epoch": 3.021418539325843, + "grad_norm": 0.9170271754264832, + "learning_rate": 2.4949144255716144e-06, + "loss": 0.5444, + "step": 8605 + }, + { + "epoch": 3.0217696629213484, + "grad_norm": 0.9525162577629089, + "learning_rate": 2.4944521013508904e-06, + "loss": 0.5786, + "step": 8606 + }, + { + "epoch": 3.022120786516854, + "grad_norm": 0.9316418170928955, + "learning_rate": 2.4939897773199005e-06, + "loss": 0.5521, + "step": 8607 + }, + { + "epoch": 3.0224719101123596, + "grad_norm": 0.9054294228553772, + "learning_rate": 2.4935274534944558e-06, + "loss": 0.5563, + "step": 8608 + }, + { + "epoch": 3.022823033707865, + "grad_norm": 0.901077926158905, + "learning_rate": 2.493065129890365e-06, + "loss": 0.5576, + "step": 8609 + }, + { + "epoch": 3.023174157303371, + "grad_norm": 0.9124284982681274, + "learning_rate": 2.492602806523442e-06, + "loss": 0.5866, + "step": 8610 + }, + { + "epoch": 3.0235252808988764, + "grad_norm": 0.8995033502578735, + "learning_rate": 2.4921404834094958e-06, + "loss": 0.5505, + "step": 8611 + }, + { + "epoch": 3.023876404494382, + "grad_norm": 0.9043973684310913, + "learning_rate": 2.491678160564339e-06, + "loss": 0.5407, + "step": 8612 + }, + { + "epoch": 3.0242275280898876, + "grad_norm": 0.8896393179893494, + "learning_rate": 2.491215838003781e-06, + "loss": 0.5501, + "step": 8613 + }, + { + "epoch": 3.024578651685393, + "grad_norm": 0.9449522495269775, + "learning_rate": 2.490753515743635e-06, + "loss": 0.5677, + "step": 8614 + }, + { + "epoch": 3.024929775280899, + "grad_norm": 0.9230022430419922, + "learning_rate": 2.4902911937997103e-06, + "loss": 0.5576, + "step": 8615 + }, + { + "epoch": 3.0252808988764044, + "grad_norm": 0.8864214420318604, + "learning_rate": 2.489828872187819e-06, + "loss": 0.539, + "step": 8616 + }, + { + "epoch": 3.02563202247191, + "grad_norm": 0.8782857656478882, + "learning_rate": 2.4893665509237706e-06, + "loss": 0.5191, + "step": 8617 + }, + { + "epoch": 3.0259831460674156, + "grad_norm": 0.9038311243057251, + "learning_rate": 2.488904230023378e-06, + "loss": 0.5441, + "step": 8618 + }, + { + "epoch": 3.026334269662921, + "grad_norm": 0.930645227432251, + "learning_rate": 2.4884419095024507e-06, + "loss": 0.5906, + "step": 8619 + }, + { + "epoch": 3.026685393258427, + "grad_norm": 0.8890973925590515, + "learning_rate": 2.4879795893768e-06, + "loss": 0.5485, + "step": 8620 + }, + { + "epoch": 3.0270365168539324, + "grad_norm": 0.9450518488883972, + "learning_rate": 2.487517269662238e-06, + "loss": 0.5357, + "step": 8621 + }, + { + "epoch": 3.0273876404494384, + "grad_norm": 0.8984273076057434, + "learning_rate": 2.4870549503745738e-06, + "loss": 0.5341, + "step": 8622 + }, + { + "epoch": 3.027738764044944, + "grad_norm": 0.9018667340278625, + "learning_rate": 2.4865926315296197e-06, + "loss": 0.5532, + "step": 8623 + }, + { + "epoch": 3.0280898876404496, + "grad_norm": 0.9440216422080994, + "learning_rate": 2.486130313143186e-06, + "loss": 0.5662, + "step": 8624 + }, + { + "epoch": 3.028441011235955, + "grad_norm": 0.9236498475074768, + "learning_rate": 2.485667995231084e-06, + "loss": 0.5782, + "step": 8625 + }, + { + "epoch": 3.028792134831461, + "grad_norm": 0.9626726508140564, + "learning_rate": 2.4852056778091232e-06, + "loss": 0.5772, + "step": 8626 + }, + { + "epoch": 3.0291432584269664, + "grad_norm": 0.8876867294311523, + "learning_rate": 2.484743360893117e-06, + "loss": 0.5473, + "step": 8627 + }, + { + "epoch": 3.029494382022472, + "grad_norm": 0.9241661429405212, + "learning_rate": 2.484281044498874e-06, + "loss": 0.5911, + "step": 8628 + }, + { + "epoch": 3.0298455056179776, + "grad_norm": 0.9357810020446777, + "learning_rate": 2.483818728642206e-06, + "loss": 0.5485, + "step": 8629 + }, + { + "epoch": 3.030196629213483, + "grad_norm": 0.9297934770584106, + "learning_rate": 2.4833564133389233e-06, + "loss": 0.5691, + "step": 8630 + }, + { + "epoch": 3.030547752808989, + "grad_norm": 0.8987939357757568, + "learning_rate": 2.4828940986048374e-06, + "loss": 0.5536, + "step": 8631 + }, + { + "epoch": 3.0308988764044944, + "grad_norm": 0.9145669937133789, + "learning_rate": 2.482431784455759e-06, + "loss": 0.545, + "step": 8632 + }, + { + "epoch": 3.03125, + "grad_norm": 0.8982566595077515, + "learning_rate": 2.4819694709074977e-06, + "loss": 0.5647, + "step": 8633 + }, + { + "epoch": 3.0316011235955056, + "grad_norm": 0.9082058668136597, + "learning_rate": 2.481507157975866e-06, + "loss": 0.5478, + "step": 8634 + }, + { + "epoch": 3.031952247191011, + "grad_norm": 0.9038417935371399, + "learning_rate": 2.481044845676673e-06, + "loss": 0.5621, + "step": 8635 + }, + { + "epoch": 3.032303370786517, + "grad_norm": 0.9302850365638733, + "learning_rate": 2.4805825340257303e-06, + "loss": 0.5619, + "step": 8636 + }, + { + "epoch": 3.0326544943820224, + "grad_norm": 0.914351761341095, + "learning_rate": 2.4801202230388483e-06, + "loss": 0.539, + "step": 8637 + }, + { + "epoch": 3.033005617977528, + "grad_norm": 0.9237471222877502, + "learning_rate": 2.4796579127318377e-06, + "loss": 0.5664, + "step": 8638 + }, + { + "epoch": 3.0333567415730336, + "grad_norm": 0.8927889466285706, + "learning_rate": 2.479195603120509e-06, + "loss": 0.5358, + "step": 8639 + }, + { + "epoch": 3.033707865168539, + "grad_norm": 0.9440634846687317, + "learning_rate": 2.4787332942206728e-06, + "loss": 0.5588, + "step": 8640 + }, + { + "epoch": 3.034058988764045, + "grad_norm": 0.9017863869667053, + "learning_rate": 2.4782709860481395e-06, + "loss": 0.5137, + "step": 8641 + }, + { + "epoch": 3.0344101123595504, + "grad_norm": 0.9513047337532043, + "learning_rate": 2.477808678618721e-06, + "loss": 0.5877, + "step": 8642 + }, + { + "epoch": 3.034761235955056, + "grad_norm": 0.9166730642318726, + "learning_rate": 2.477346371948225e-06, + "loss": 0.5558, + "step": 8643 + }, + { + "epoch": 3.0351123595505616, + "grad_norm": 0.8981868624687195, + "learning_rate": 2.476884066052465e-06, + "loss": 0.5267, + "step": 8644 + }, + { + "epoch": 3.0354634831460676, + "grad_norm": 0.9089592099189758, + "learning_rate": 2.47642176094725e-06, + "loss": 0.5398, + "step": 8645 + }, + { + "epoch": 3.035814606741573, + "grad_norm": 0.8809298872947693, + "learning_rate": 2.47595945664839e-06, + "loss": 0.5426, + "step": 8646 + }, + { + "epoch": 3.036165730337079, + "grad_norm": 0.9058164954185486, + "learning_rate": 2.4754971531716965e-06, + "loss": 0.5364, + "step": 8647 + }, + { + "epoch": 3.0365168539325844, + "grad_norm": 0.9012526273727417, + "learning_rate": 2.4750348505329794e-06, + "loss": 0.543, + "step": 8648 + }, + { + "epoch": 3.03686797752809, + "grad_norm": 0.8812972903251648, + "learning_rate": 2.4745725487480495e-06, + "loss": 0.5615, + "step": 8649 + }, + { + "epoch": 3.0372191011235956, + "grad_norm": 1.2184324264526367, + "learning_rate": 2.4741102478327153e-06, + "loss": 0.5813, + "step": 8650 + }, + { + "epoch": 3.037570224719101, + "grad_norm": 0.9156524538993835, + "learning_rate": 2.4736479478027898e-06, + "loss": 0.5341, + "step": 8651 + }, + { + "epoch": 3.037921348314607, + "grad_norm": 0.9279172420501709, + "learning_rate": 2.4731856486740806e-06, + "loss": 0.5807, + "step": 8652 + }, + { + "epoch": 3.0382724719101124, + "grad_norm": 0.9355500340461731, + "learning_rate": 2.4727233504624006e-06, + "loss": 0.5693, + "step": 8653 + }, + { + "epoch": 3.038623595505618, + "grad_norm": 0.9756408929824829, + "learning_rate": 2.4722610531835576e-06, + "loss": 0.5683, + "step": 8654 + }, + { + "epoch": 3.0389747191011236, + "grad_norm": 0.9109455943107605, + "learning_rate": 2.471798756853364e-06, + "loss": 0.5714, + "step": 8655 + }, + { + "epoch": 3.039325842696629, + "grad_norm": 0.8939483761787415, + "learning_rate": 2.4713364614876275e-06, + "loss": 0.5529, + "step": 8656 + }, + { + "epoch": 3.039676966292135, + "grad_norm": 0.9166051149368286, + "learning_rate": 2.4708741671021606e-06, + "loss": 0.5625, + "step": 8657 + }, + { + "epoch": 3.0400280898876404, + "grad_norm": 0.9080982804298401, + "learning_rate": 2.4704118737127726e-06, + "loss": 0.5817, + "step": 8658 + }, + { + "epoch": 3.040379213483146, + "grad_norm": 0.8846049904823303, + "learning_rate": 2.4699495813352713e-06, + "loss": 0.5353, + "step": 8659 + }, + { + "epoch": 3.0407303370786516, + "grad_norm": 0.9577983617782593, + "learning_rate": 2.4694872899854705e-06, + "loss": 0.5781, + "step": 8660 + }, + { + "epoch": 3.041081460674157, + "grad_norm": 0.9427290558815002, + "learning_rate": 2.469024999679177e-06, + "loss": 0.5736, + "step": 8661 + }, + { + "epoch": 3.041432584269663, + "grad_norm": 0.9106037616729736, + "learning_rate": 2.468562710432203e-06, + "loss": 0.5578, + "step": 8662 + }, + { + "epoch": 3.0417837078651684, + "grad_norm": 0.9185969829559326, + "learning_rate": 2.468100422260356e-06, + "loss": 0.5589, + "step": 8663 + }, + { + "epoch": 3.042134831460674, + "grad_norm": 0.8964781165122986, + "learning_rate": 2.467638135179449e-06, + "loss": 0.5157, + "step": 8664 + }, + { + "epoch": 3.0424859550561796, + "grad_norm": 0.9552000164985657, + "learning_rate": 2.4671758492052884e-06, + "loss": 0.5843, + "step": 8665 + }, + { + "epoch": 3.042837078651685, + "grad_norm": 0.8914016485214233, + "learning_rate": 2.466713564353687e-06, + "loss": 0.5219, + "step": 8666 + }, + { + "epoch": 3.043188202247191, + "grad_norm": 0.9400820732116699, + "learning_rate": 2.466251280640452e-06, + "loss": 0.5409, + "step": 8667 + }, + { + "epoch": 3.043539325842697, + "grad_norm": 0.9568427205085754, + "learning_rate": 2.4657889980813955e-06, + "loss": 0.5697, + "step": 8668 + }, + { + "epoch": 3.0438904494382024, + "grad_norm": 0.8930754661560059, + "learning_rate": 2.465326716692325e-06, + "loss": 0.5663, + "step": 8669 + }, + { + "epoch": 3.044241573033708, + "grad_norm": 0.9707926511764526, + "learning_rate": 2.4648644364890523e-06, + "loss": 0.6201, + "step": 8670 + }, + { + "epoch": 3.0445926966292136, + "grad_norm": 0.9320633411407471, + "learning_rate": 2.464402157487386e-06, + "loss": 0.5801, + "step": 8671 + }, + { + "epoch": 3.044943820224719, + "grad_norm": 0.9549024105072021, + "learning_rate": 2.4639398797031337e-06, + "loss": 0.5415, + "step": 8672 + }, + { + "epoch": 3.045294943820225, + "grad_norm": 0.9142651557922363, + "learning_rate": 2.463477603152108e-06, + "loss": 0.5592, + "step": 8673 + }, + { + "epoch": 3.0456460674157304, + "grad_norm": 0.8838040828704834, + "learning_rate": 2.4630153278501163e-06, + "loss": 0.563, + "step": 8674 + }, + { + "epoch": 3.045997191011236, + "grad_norm": 0.9340301156044006, + "learning_rate": 2.4625530538129693e-06, + "loss": 0.561, + "step": 8675 + }, + { + "epoch": 3.0463483146067416, + "grad_norm": 0.930112361907959, + "learning_rate": 2.462090781056475e-06, + "loss": 0.5333, + "step": 8676 + }, + { + "epoch": 3.046699438202247, + "grad_norm": 0.955159604549408, + "learning_rate": 2.4616285095964444e-06, + "loss": 0.5517, + "step": 8677 + }, + { + "epoch": 3.047050561797753, + "grad_norm": 0.8946186900138855, + "learning_rate": 2.461166239448685e-06, + "loss": 0.5273, + "step": 8678 + }, + { + "epoch": 3.0474016853932584, + "grad_norm": 0.9050875902175903, + "learning_rate": 2.4607039706290086e-06, + "loss": 0.5476, + "step": 8679 + }, + { + "epoch": 3.047752808988764, + "grad_norm": 0.9322788119316101, + "learning_rate": 2.460241703153221e-06, + "loss": 0.5679, + "step": 8680 + }, + { + "epoch": 3.0481039325842696, + "grad_norm": 0.9337145090103149, + "learning_rate": 2.4597794370371343e-06, + "loss": 0.5507, + "step": 8681 + }, + { + "epoch": 3.048455056179775, + "grad_norm": 0.9402443766593933, + "learning_rate": 2.4593171722965555e-06, + "loss": 0.5733, + "step": 8682 + }, + { + "epoch": 3.048806179775281, + "grad_norm": 0.8997933864593506, + "learning_rate": 2.4588549089472956e-06, + "loss": 0.5562, + "step": 8683 + }, + { + "epoch": 3.0491573033707864, + "grad_norm": 0.9218352437019348, + "learning_rate": 2.458392647005163e-06, + "loss": 0.5487, + "step": 8684 + }, + { + "epoch": 3.049508426966292, + "grad_norm": 0.9463024735450745, + "learning_rate": 2.4579303864859645e-06, + "loss": 0.5865, + "step": 8685 + }, + { + "epoch": 3.0498595505617976, + "grad_norm": 0.924631655216217, + "learning_rate": 2.4574681274055122e-06, + "loss": 0.5718, + "step": 8686 + }, + { + "epoch": 3.050210674157303, + "grad_norm": 0.9213266968727112, + "learning_rate": 2.4570058697796125e-06, + "loss": 0.5463, + "step": 8687 + }, + { + "epoch": 3.050561797752809, + "grad_norm": 0.9161856174468994, + "learning_rate": 2.4565436136240763e-06, + "loss": 0.5523, + "step": 8688 + }, + { + "epoch": 3.050912921348315, + "grad_norm": 0.9059017300605774, + "learning_rate": 2.4560813589547103e-06, + "loss": 0.5537, + "step": 8689 + }, + { + "epoch": 3.0512640449438204, + "grad_norm": 0.9157281517982483, + "learning_rate": 2.455619105787325e-06, + "loss": 0.5577, + "step": 8690 + }, + { + "epoch": 3.051615168539326, + "grad_norm": 0.9055077433586121, + "learning_rate": 2.455156854137728e-06, + "loss": 0.5544, + "step": 8691 + }, + { + "epoch": 3.0519662921348316, + "grad_norm": 0.9005976319313049, + "learning_rate": 2.454694604021729e-06, + "loss": 0.5253, + "step": 8692 + }, + { + "epoch": 3.052317415730337, + "grad_norm": 0.9300852417945862, + "learning_rate": 2.4542323554551344e-06, + "loss": 0.5659, + "step": 8693 + }, + { + "epoch": 3.052668539325843, + "grad_norm": 0.9117161631584167, + "learning_rate": 2.453770108453755e-06, + "loss": 0.536, + "step": 8694 + }, + { + "epoch": 3.0530196629213484, + "grad_norm": 0.9248644709587097, + "learning_rate": 2.4533078630333986e-06, + "loss": 0.5803, + "step": 8695 + }, + { + "epoch": 3.053370786516854, + "grad_norm": 0.9057329297065735, + "learning_rate": 2.4528456192098724e-06, + "loss": 0.53, + "step": 8696 + }, + { + "epoch": 3.0537219101123596, + "grad_norm": 0.8901026248931885, + "learning_rate": 2.4523833769989863e-06, + "loss": 0.5419, + "step": 8697 + }, + { + "epoch": 3.054073033707865, + "grad_norm": 0.8877182006835938, + "learning_rate": 2.4519211364165473e-06, + "loss": 0.5491, + "step": 8698 + }, + { + "epoch": 3.054424157303371, + "grad_norm": 0.9011861085891724, + "learning_rate": 2.4514588974783653e-06, + "loss": 0.536, + "step": 8699 + }, + { + "epoch": 3.0547752808988764, + "grad_norm": 0.9572852253913879, + "learning_rate": 2.4509966602002463e-06, + "loss": 0.5746, + "step": 8700 + }, + { + "epoch": 3.055126404494382, + "grad_norm": 0.9288879036903381, + "learning_rate": 2.4505344245980006e-06, + "loss": 0.5768, + "step": 8701 + }, + { + "epoch": 3.0554775280898876, + "grad_norm": 0.9239123463630676, + "learning_rate": 2.4500721906874343e-06, + "loss": 0.5554, + "step": 8702 + }, + { + "epoch": 3.055828651685393, + "grad_norm": 0.9224625825881958, + "learning_rate": 2.4496099584843573e-06, + "loss": 0.5412, + "step": 8703 + }, + { + "epoch": 3.056179775280899, + "grad_norm": 0.9459335207939148, + "learning_rate": 2.449147728004576e-06, + "loss": 0.5922, + "step": 8704 + }, + { + "epoch": 3.0565308988764044, + "grad_norm": 0.9469919800758362, + "learning_rate": 2.4486854992638994e-06, + "loss": 0.5303, + "step": 8705 + }, + { + "epoch": 3.05688202247191, + "grad_norm": 0.9022507667541504, + "learning_rate": 2.4482232722781345e-06, + "loss": 0.5589, + "step": 8706 + }, + { + "epoch": 3.0572331460674156, + "grad_norm": 0.9310958385467529, + "learning_rate": 2.4477610470630895e-06, + "loss": 0.5591, + "step": 8707 + }, + { + "epoch": 3.057584269662921, + "grad_norm": 0.9631215333938599, + "learning_rate": 2.447298823634572e-06, + "loss": 0.5636, + "step": 8708 + }, + { + "epoch": 3.057935393258427, + "grad_norm": 0.9181795716285706, + "learning_rate": 2.446836602008389e-06, + "loss": 0.5579, + "step": 8709 + }, + { + "epoch": 3.0582865168539324, + "grad_norm": 0.9015101790428162, + "learning_rate": 2.44637438220035e-06, + "loss": 0.558, + "step": 8710 + }, + { + "epoch": 3.0586376404494384, + "grad_norm": 0.9486123323440552, + "learning_rate": 2.4459121642262597e-06, + "loss": 0.5765, + "step": 8711 + }, + { + "epoch": 3.058988764044944, + "grad_norm": 0.9191428422927856, + "learning_rate": 2.4454499481019284e-06, + "loss": 0.5793, + "step": 8712 + }, + { + "epoch": 3.0593398876404496, + "grad_norm": 0.9156162738800049, + "learning_rate": 2.444987733843161e-06, + "loss": 0.5466, + "step": 8713 + }, + { + "epoch": 3.059691011235955, + "grad_norm": 0.9304625391960144, + "learning_rate": 2.4445255214657675e-06, + "loss": 0.5901, + "step": 8714 + }, + { + "epoch": 3.060042134831461, + "grad_norm": 0.9316036105155945, + "learning_rate": 2.4440633109855526e-06, + "loss": 0.5839, + "step": 8715 + }, + { + "epoch": 3.0603932584269664, + "grad_norm": 0.9060834050178528, + "learning_rate": 2.4436011024183248e-06, + "loss": 0.5297, + "step": 8716 + }, + { + "epoch": 3.060744382022472, + "grad_norm": 0.9168406128883362, + "learning_rate": 2.4431388957798913e-06, + "loss": 0.5363, + "step": 8717 + }, + { + "epoch": 3.0610955056179776, + "grad_norm": 0.9034215211868286, + "learning_rate": 2.4426766910860587e-06, + "loss": 0.5189, + "step": 8718 + }, + { + "epoch": 3.061446629213483, + "grad_norm": 0.8810475468635559, + "learning_rate": 2.442214488352634e-06, + "loss": 0.5624, + "step": 8719 + }, + { + "epoch": 3.061797752808989, + "grad_norm": 0.9512900710105896, + "learning_rate": 2.4417522875954247e-06, + "loss": 0.5647, + "step": 8720 + }, + { + "epoch": 3.0621488764044944, + "grad_norm": 0.9337357878684998, + "learning_rate": 2.441290088830238e-06, + "loss": 0.5424, + "step": 8721 + }, + { + "epoch": 3.0625, + "grad_norm": 0.9300682544708252, + "learning_rate": 2.440827892072879e-06, + "loss": 0.5706, + "step": 8722 + }, + { + "epoch": 3.0628511235955056, + "grad_norm": 0.9339327812194824, + "learning_rate": 2.4403656973391557e-06, + "loss": 0.5914, + "step": 8723 + }, + { + "epoch": 3.063202247191011, + "grad_norm": 0.899368166923523, + "learning_rate": 2.4399035046448743e-06, + "loss": 0.5407, + "step": 8724 + }, + { + "epoch": 3.063553370786517, + "grad_norm": 0.9143944978713989, + "learning_rate": 2.4394413140058416e-06, + "loss": 0.5662, + "step": 8725 + }, + { + "epoch": 3.0639044943820224, + "grad_norm": 1.0048120021820068, + "learning_rate": 2.4389791254378646e-06, + "loss": 0.5486, + "step": 8726 + }, + { + "epoch": 3.064255617977528, + "grad_norm": 0.9438900947570801, + "learning_rate": 2.438516938956749e-06, + "loss": 0.5575, + "step": 8727 + }, + { + "epoch": 3.0646067415730336, + "grad_norm": 0.897100567817688, + "learning_rate": 2.4380547545783016e-06, + "loss": 0.5498, + "step": 8728 + }, + { + "epoch": 3.064957865168539, + "grad_norm": 0.9282429218292236, + "learning_rate": 2.4375925723183293e-06, + "loss": 0.5657, + "step": 8729 + }, + { + "epoch": 3.065308988764045, + "grad_norm": 0.8952595591545105, + "learning_rate": 2.437130392192636e-06, + "loss": 0.5525, + "step": 8730 + }, + { + "epoch": 3.0656601123595504, + "grad_norm": 0.9370861053466797, + "learning_rate": 2.436668214217031e-06, + "loss": 0.5564, + "step": 8731 + }, + { + "epoch": 3.066011235955056, + "grad_norm": 0.8888670802116394, + "learning_rate": 2.4362060384073176e-06, + "loss": 0.5339, + "step": 8732 + }, + { + "epoch": 3.0663623595505616, + "grad_norm": 0.919877827167511, + "learning_rate": 2.4357438647793043e-06, + "loss": 0.5478, + "step": 8733 + }, + { + "epoch": 3.0667134831460676, + "grad_norm": 0.9033722877502441, + "learning_rate": 2.435281693348795e-06, + "loss": 0.5512, + "step": 8734 + }, + { + "epoch": 3.067064606741573, + "grad_norm": 0.9495663046836853, + "learning_rate": 2.434819524131597e-06, + "loss": 0.5496, + "step": 8735 + }, + { + "epoch": 3.067415730337079, + "grad_norm": 0.9636256098747253, + "learning_rate": 2.4343573571435155e-06, + "loss": 0.5713, + "step": 8736 + }, + { + "epoch": 3.0677668539325844, + "grad_norm": 0.9488418698310852, + "learning_rate": 2.4338951924003552e-06, + "loss": 0.5452, + "step": 8737 + }, + { + "epoch": 3.06811797752809, + "grad_norm": 0.9174200892448425, + "learning_rate": 2.433433029917924e-06, + "loss": 0.5739, + "step": 8738 + }, + { + "epoch": 3.0684691011235956, + "grad_norm": 0.9007225036621094, + "learning_rate": 2.432970869712025e-06, + "loss": 0.5512, + "step": 8739 + }, + { + "epoch": 3.068820224719101, + "grad_norm": 0.925025999546051, + "learning_rate": 2.432508711798466e-06, + "loss": 0.5294, + "step": 8740 + }, + { + "epoch": 3.069171348314607, + "grad_norm": 0.9268113970756531, + "learning_rate": 2.4320465561930503e-06, + "loss": 0.5599, + "step": 8741 + }, + { + "epoch": 3.0695224719101124, + "grad_norm": 0.9283733367919922, + "learning_rate": 2.4315844029115854e-06, + "loss": 0.5782, + "step": 8742 + }, + { + "epoch": 3.069873595505618, + "grad_norm": 0.8945986032485962, + "learning_rate": 2.431122251969874e-06, + "loss": 0.5212, + "step": 8743 + }, + { + "epoch": 3.0702247191011236, + "grad_norm": 0.9030485153198242, + "learning_rate": 2.4306601033837242e-06, + "loss": 0.545, + "step": 8744 + }, + { + "epoch": 3.070575842696629, + "grad_norm": 0.9275111556053162, + "learning_rate": 2.430197957168938e-06, + "loss": 0.5514, + "step": 8745 + }, + { + "epoch": 3.070926966292135, + "grad_norm": 0.9420635104179382, + "learning_rate": 2.429735813341323e-06, + "loss": 0.5801, + "step": 8746 + }, + { + "epoch": 3.0712780898876404, + "grad_norm": 0.9032498002052307, + "learning_rate": 2.4292736719166833e-06, + "loss": 0.5424, + "step": 8747 + }, + { + "epoch": 3.071629213483146, + "grad_norm": 0.9428184628486633, + "learning_rate": 2.4288115329108224e-06, + "loss": 0.5536, + "step": 8748 + }, + { + "epoch": 3.0719803370786516, + "grad_norm": 0.9107243418693542, + "learning_rate": 2.4283493963395473e-06, + "loss": 0.557, + "step": 8749 + }, + { + "epoch": 3.072331460674157, + "grad_norm": 0.9486559629440308, + "learning_rate": 2.42788726221866e-06, + "loss": 0.5756, + "step": 8750 + }, + { + "epoch": 3.072682584269663, + "grad_norm": 0.9269229173660278, + "learning_rate": 2.4274251305639682e-06, + "loss": 0.5422, + "step": 8751 + }, + { + "epoch": 3.0730337078651684, + "grad_norm": 0.9165647625923157, + "learning_rate": 2.4269630013912735e-06, + "loss": 0.5377, + "step": 8752 + }, + { + "epoch": 3.073384831460674, + "grad_norm": 0.931347668170929, + "learning_rate": 2.4265008747163827e-06, + "loss": 0.5502, + "step": 8753 + }, + { + "epoch": 3.0737359550561796, + "grad_norm": 0.8984931111335754, + "learning_rate": 2.4260387505550983e-06, + "loss": 0.5411, + "step": 8754 + }, + { + "epoch": 3.074087078651685, + "grad_norm": 0.9066592454910278, + "learning_rate": 2.425576628923226e-06, + "loss": 0.5432, + "step": 8755 + }, + { + "epoch": 3.074438202247191, + "grad_norm": 0.9413814544677734, + "learning_rate": 2.4251145098365686e-06, + "loss": 0.5508, + "step": 8756 + }, + { + "epoch": 3.074789325842697, + "grad_norm": 0.9335858225822449, + "learning_rate": 2.424652393310932e-06, + "loss": 0.5338, + "step": 8757 + }, + { + "epoch": 3.0751404494382024, + "grad_norm": 0.9055057764053345, + "learning_rate": 2.424190279362119e-06, + "loss": 0.568, + "step": 8758 + }, + { + "epoch": 3.075491573033708, + "grad_norm": 0.9435656070709229, + "learning_rate": 2.4237281680059325e-06, + "loss": 0.5791, + "step": 8759 + }, + { + "epoch": 3.0758426966292136, + "grad_norm": 0.9286941885948181, + "learning_rate": 2.423266059258178e-06, + "loss": 0.5777, + "step": 8760 + }, + { + "epoch": 3.076193820224719, + "grad_norm": 0.9294689893722534, + "learning_rate": 2.422803953134658e-06, + "loss": 0.584, + "step": 8761 + }, + { + "epoch": 3.076544943820225, + "grad_norm": 0.9134140610694885, + "learning_rate": 2.422341849651178e-06, + "loss": 0.5259, + "step": 8762 + }, + { + "epoch": 3.0768960674157304, + "grad_norm": 0.9329862594604492, + "learning_rate": 2.4218797488235387e-06, + "loss": 0.545, + "step": 8763 + }, + { + "epoch": 3.077247191011236, + "grad_norm": 0.9447376132011414, + "learning_rate": 2.4214176506675468e-06, + "loss": 0.5607, + "step": 8764 + }, + { + "epoch": 3.0775983146067416, + "grad_norm": 0.9274771809577942, + "learning_rate": 2.4209555551990024e-06, + "loss": 0.5437, + "step": 8765 + }, + { + "epoch": 3.077949438202247, + "grad_norm": 0.9024612307548523, + "learning_rate": 2.420493462433712e-06, + "loss": 0.513, + "step": 8766 + }, + { + "epoch": 3.078300561797753, + "grad_norm": 0.9127166867256165, + "learning_rate": 2.4200313723874752e-06, + "loss": 0.5324, + "step": 8767 + }, + { + "epoch": 3.0786516853932584, + "grad_norm": 0.9430844187736511, + "learning_rate": 2.4195692850760987e-06, + "loss": 0.5698, + "step": 8768 + }, + { + "epoch": 3.079002808988764, + "grad_norm": 0.9167546033859253, + "learning_rate": 2.4191072005153824e-06, + "loss": 0.5308, + "step": 8769 + }, + { + "epoch": 3.0793539325842696, + "grad_norm": 0.9421341419219971, + "learning_rate": 2.4186451187211325e-06, + "loss": 0.5562, + "step": 8770 + }, + { + "epoch": 3.079705056179775, + "grad_norm": 0.9473413825035095, + "learning_rate": 2.418183039709149e-06, + "loss": 0.5697, + "step": 8771 + }, + { + "epoch": 3.080056179775281, + "grad_norm": 0.9433597922325134, + "learning_rate": 2.4177209634952353e-06, + "loss": 0.56, + "step": 8772 + }, + { + "epoch": 3.0804073033707864, + "grad_norm": 0.9268821477890015, + "learning_rate": 2.4172588900951945e-06, + "loss": 0.5724, + "step": 8773 + }, + { + "epoch": 3.080758426966292, + "grad_norm": 0.9172853231430054, + "learning_rate": 2.4167968195248283e-06, + "loss": 0.5444, + "step": 8774 + }, + { + "epoch": 3.0811095505617976, + "grad_norm": 0.9481356143951416, + "learning_rate": 2.4163347517999405e-06, + "loss": 0.5567, + "step": 8775 + }, + { + "epoch": 3.081460674157303, + "grad_norm": 0.9161304235458374, + "learning_rate": 2.4158726869363317e-06, + "loss": 0.5443, + "step": 8776 + }, + { + "epoch": 3.081811797752809, + "grad_norm": 0.9579325914382935, + "learning_rate": 2.415410624949806e-06, + "loss": 0.601, + "step": 8777 + }, + { + "epoch": 3.082162921348315, + "grad_norm": 0.9390956163406372, + "learning_rate": 2.4149485658561634e-06, + "loss": 0.5786, + "step": 8778 + }, + { + "epoch": 3.0825140449438204, + "grad_norm": 0.9467632174491882, + "learning_rate": 2.4144865096712084e-06, + "loss": 0.5685, + "step": 8779 + }, + { + "epoch": 3.082865168539326, + "grad_norm": 0.9338750839233398, + "learning_rate": 2.41402445641074e-06, + "loss": 0.5648, + "step": 8780 + }, + { + "epoch": 3.0832162921348316, + "grad_norm": 0.9061920046806335, + "learning_rate": 2.4135624060905633e-06, + "loss": 0.5354, + "step": 8781 + }, + { + "epoch": 3.083567415730337, + "grad_norm": 0.9327059388160706, + "learning_rate": 2.413100358726477e-06, + "loss": 0.5911, + "step": 8782 + }, + { + "epoch": 3.083918539325843, + "grad_norm": 0.9405179619789124, + "learning_rate": 2.4126383143342852e-06, + "loss": 0.5452, + "step": 8783 + }, + { + "epoch": 3.0842696629213484, + "grad_norm": 0.892975389957428, + "learning_rate": 2.4121762729297884e-06, + "loss": 0.5586, + "step": 8784 + }, + { + "epoch": 3.084620786516854, + "grad_norm": 0.9155935645103455, + "learning_rate": 2.4117142345287865e-06, + "loss": 0.51, + "step": 8785 + }, + { + "epoch": 3.0849719101123596, + "grad_norm": 0.9281261563301086, + "learning_rate": 2.4112521991470835e-06, + "loss": 0.5956, + "step": 8786 + }, + { + "epoch": 3.085323033707865, + "grad_norm": 0.9271456003189087, + "learning_rate": 2.4107901668004783e-06, + "loss": 0.5233, + "step": 8787 + }, + { + "epoch": 3.085674157303371, + "grad_norm": 0.9286864995956421, + "learning_rate": 2.410328137504774e-06, + "loss": 0.5642, + "step": 8788 + }, + { + "epoch": 3.0860252808988764, + "grad_norm": 0.9009825587272644, + "learning_rate": 2.4098661112757698e-06, + "loss": 0.5556, + "step": 8789 + }, + { + "epoch": 3.086376404494382, + "grad_norm": 0.910004734992981, + "learning_rate": 2.4094040881292684e-06, + "loss": 0.5485, + "step": 8790 + }, + { + "epoch": 3.0867275280898876, + "grad_norm": 0.9161363244056702, + "learning_rate": 2.4089420680810683e-06, + "loss": 0.559, + "step": 8791 + }, + { + "epoch": 3.087078651685393, + "grad_norm": 0.8899289965629578, + "learning_rate": 2.408480051146973e-06, + "loss": 0.5492, + "step": 8792 + }, + { + "epoch": 3.087429775280899, + "grad_norm": 0.894830584526062, + "learning_rate": 2.4080180373427813e-06, + "loss": 0.545, + "step": 8793 + }, + { + "epoch": 3.0877808988764044, + "grad_norm": 0.9506084322929382, + "learning_rate": 2.4075560266842938e-06, + "loss": 0.5356, + "step": 8794 + }, + { + "epoch": 3.08813202247191, + "grad_norm": 0.908367931842804, + "learning_rate": 2.4070940191873105e-06, + "loss": 0.5319, + "step": 8795 + }, + { + "epoch": 3.0884831460674156, + "grad_norm": 0.9308577179908752, + "learning_rate": 2.406632014867633e-06, + "loss": 0.5877, + "step": 8796 + }, + { + "epoch": 3.088834269662921, + "grad_norm": 0.9469871520996094, + "learning_rate": 2.4061700137410614e-06, + "loss": 0.57, + "step": 8797 + }, + { + "epoch": 3.089185393258427, + "grad_norm": 0.93496173620224, + "learning_rate": 2.405708015823393e-06, + "loss": 0.5576, + "step": 8798 + }, + { + "epoch": 3.0895365168539324, + "grad_norm": 0.9359568953514099, + "learning_rate": 2.405246021130432e-06, + "loss": 0.5374, + "step": 8799 + }, + { + "epoch": 3.0898876404494384, + "grad_norm": 0.9386340975761414, + "learning_rate": 2.404784029677974e-06, + "loss": 0.5526, + "step": 8800 + }, + { + "epoch": 3.090238764044944, + "grad_norm": 0.9635471701622009, + "learning_rate": 2.4043220414818222e-06, + "loss": 0.5743, + "step": 8801 + }, + { + "epoch": 3.0905898876404496, + "grad_norm": 0.8821988105773926, + "learning_rate": 2.403860056557774e-06, + "loss": 0.5589, + "step": 8802 + }, + { + "epoch": 3.090941011235955, + "grad_norm": 0.9300549626350403, + "learning_rate": 2.40339807492163e-06, + "loss": 0.5618, + "step": 8803 + }, + { + "epoch": 3.091292134831461, + "grad_norm": 0.9554167985916138, + "learning_rate": 2.402936096589189e-06, + "loss": 0.5947, + "step": 8804 + }, + { + "epoch": 3.0916432584269664, + "grad_norm": 0.9080701470375061, + "learning_rate": 2.40247412157625e-06, + "loss": 0.5584, + "step": 8805 + }, + { + "epoch": 3.091994382022472, + "grad_norm": 0.899929940700531, + "learning_rate": 2.402012149898613e-06, + "loss": 0.5419, + "step": 8806 + }, + { + "epoch": 3.0923455056179776, + "grad_norm": 0.9399242401123047, + "learning_rate": 2.4015501815720763e-06, + "loss": 0.5878, + "step": 8807 + }, + { + "epoch": 3.092696629213483, + "grad_norm": 0.9202435612678528, + "learning_rate": 2.4010882166124393e-06, + "loss": 0.5772, + "step": 8808 + }, + { + "epoch": 3.093047752808989, + "grad_norm": 0.9238872528076172, + "learning_rate": 2.4006262550355013e-06, + "loss": 0.5915, + "step": 8809 + }, + { + "epoch": 3.0933988764044944, + "grad_norm": 0.931072473526001, + "learning_rate": 2.4001642968570597e-06, + "loss": 0.5711, + "step": 8810 + }, + { + "epoch": 3.09375, + "grad_norm": 0.9183313250541687, + "learning_rate": 2.399702342092914e-06, + "loss": 0.5706, + "step": 8811 + }, + { + "epoch": 3.0941011235955056, + "grad_norm": 0.9476996064186096, + "learning_rate": 2.399240390758862e-06, + "loss": 0.593, + "step": 8812 + }, + { + "epoch": 3.094452247191011, + "grad_norm": 0.9197458624839783, + "learning_rate": 2.3987784428707026e-06, + "loss": 0.5395, + "step": 8813 + }, + { + "epoch": 3.094803370786517, + "grad_norm": 0.940715491771698, + "learning_rate": 2.3983164984442335e-06, + "loss": 0.5813, + "step": 8814 + }, + { + "epoch": 3.0951544943820224, + "grad_norm": 0.9470247626304626, + "learning_rate": 2.3978545574952536e-06, + "loss": 0.5494, + "step": 8815 + }, + { + "epoch": 3.095505617977528, + "grad_norm": 0.9360830187797546, + "learning_rate": 2.3973926200395607e-06, + "loss": 0.5352, + "step": 8816 + }, + { + "epoch": 3.0958567415730336, + "grad_norm": 0.928244411945343, + "learning_rate": 2.3969306860929517e-06, + "loss": 0.5591, + "step": 8817 + }, + { + "epoch": 3.096207865168539, + "grad_norm": 0.9259093403816223, + "learning_rate": 2.396468755671226e-06, + "loss": 0.5466, + "step": 8818 + }, + { + "epoch": 3.096558988764045, + "grad_norm": 0.9337350130081177, + "learning_rate": 2.396006828790179e-06, + "loss": 0.5639, + "step": 8819 + }, + { + "epoch": 3.0969101123595504, + "grad_norm": 0.9089118242263794, + "learning_rate": 2.3955449054656105e-06, + "loss": 0.5365, + "step": 8820 + }, + { + "epoch": 3.097261235955056, + "grad_norm": 0.9315569996833801, + "learning_rate": 2.3950829857133163e-06, + "loss": 0.5629, + "step": 8821 + }, + { + "epoch": 3.0976123595505616, + "grad_norm": 0.924245297908783, + "learning_rate": 2.394621069549094e-06, + "loss": 0.5468, + "step": 8822 + }, + { + "epoch": 3.0979634831460676, + "grad_norm": 0.9293679594993591, + "learning_rate": 2.394159156988742e-06, + "loss": 0.5714, + "step": 8823 + }, + { + "epoch": 3.098314606741573, + "grad_norm": 0.9309301376342773, + "learning_rate": 2.393697248048055e-06, + "loss": 0.5586, + "step": 8824 + }, + { + "epoch": 3.098665730337079, + "grad_norm": 0.8802352547645569, + "learning_rate": 2.393235342742832e-06, + "loss": 0.5275, + "step": 8825 + }, + { + "epoch": 3.0990168539325844, + "grad_norm": 0.9110726118087769, + "learning_rate": 2.392773441088868e-06, + "loss": 0.5723, + "step": 8826 + }, + { + "epoch": 3.09936797752809, + "grad_norm": 0.9279472827911377, + "learning_rate": 2.3923115431019616e-06, + "loss": 0.5548, + "step": 8827 + }, + { + "epoch": 3.0997191011235956, + "grad_norm": 0.9140030741691589, + "learning_rate": 2.391849648797908e-06, + "loss": 0.5754, + "step": 8828 + }, + { + "epoch": 3.100070224719101, + "grad_norm": 0.9206311702728271, + "learning_rate": 2.391387758192504e-06, + "loss": 0.5497, + "step": 8829 + }, + { + "epoch": 3.100421348314607, + "grad_norm": 0.9100523591041565, + "learning_rate": 2.3909258713015448e-06, + "loss": 0.5575, + "step": 8830 + }, + { + "epoch": 3.1007724719101124, + "grad_norm": 0.9200919270515442, + "learning_rate": 2.3904639881408286e-06, + "loss": 0.5747, + "step": 8831 + }, + { + "epoch": 3.101123595505618, + "grad_norm": 0.9382959604263306, + "learning_rate": 2.3900021087261487e-06, + "loss": 0.5618, + "step": 8832 + }, + { + "epoch": 3.1014747191011236, + "grad_norm": 0.9392076730728149, + "learning_rate": 2.3895402330733047e-06, + "loss": 0.536, + "step": 8833 + }, + { + "epoch": 3.101825842696629, + "grad_norm": 0.9338902831077576, + "learning_rate": 2.3890783611980895e-06, + "loss": 0.5477, + "step": 8834 + }, + { + "epoch": 3.102176966292135, + "grad_norm": 0.9275320172309875, + "learning_rate": 2.3886164931162984e-06, + "loss": 0.5747, + "step": 8835 + }, + { + "epoch": 3.1025280898876404, + "grad_norm": 0.8980362415313721, + "learning_rate": 2.388154628843729e-06, + "loss": 0.5182, + "step": 8836 + }, + { + "epoch": 3.102879213483146, + "grad_norm": 0.9216986298561096, + "learning_rate": 2.3876927683961747e-06, + "loss": 0.5968, + "step": 8837 + }, + { + "epoch": 3.1032303370786516, + "grad_norm": 0.936629056930542, + "learning_rate": 2.3872309117894324e-06, + "loss": 0.5564, + "step": 8838 + }, + { + "epoch": 3.103581460674157, + "grad_norm": 0.9568994641304016, + "learning_rate": 2.3867690590392954e-06, + "loss": 0.5811, + "step": 8839 + }, + { + "epoch": 3.103932584269663, + "grad_norm": 0.92243492603302, + "learning_rate": 2.386307210161561e-06, + "loss": 0.5613, + "step": 8840 + }, + { + "epoch": 3.1042837078651684, + "grad_norm": 0.9217270612716675, + "learning_rate": 2.3858453651720216e-06, + "loss": 0.5622, + "step": 8841 + }, + { + "epoch": 3.104634831460674, + "grad_norm": 0.909415066242218, + "learning_rate": 2.385383524086474e-06, + "loss": 0.5488, + "step": 8842 + }, + { + "epoch": 3.1049859550561796, + "grad_norm": 0.8802032470703125, + "learning_rate": 2.3849216869207108e-06, + "loss": 0.5394, + "step": 8843 + }, + { + "epoch": 3.105337078651685, + "grad_norm": 0.9021036624908447, + "learning_rate": 2.3844598536905285e-06, + "loss": 0.5416, + "step": 8844 + }, + { + "epoch": 3.105688202247191, + "grad_norm": 0.9095220565795898, + "learning_rate": 2.383998024411719e-06, + "loss": 0.5731, + "step": 8845 + }, + { + "epoch": 3.106039325842697, + "grad_norm": 0.927535891532898, + "learning_rate": 2.3835361991000793e-06, + "loss": 0.5666, + "step": 8846 + }, + { + "epoch": 3.1063904494382024, + "grad_norm": 0.9082242846488953, + "learning_rate": 2.383074377771402e-06, + "loss": 0.5459, + "step": 8847 + }, + { + "epoch": 3.106741573033708, + "grad_norm": 0.9189441204071045, + "learning_rate": 2.382612560441479e-06, + "loss": 0.5363, + "step": 8848 + }, + { + "epoch": 3.1070926966292136, + "grad_norm": 0.9195154309272766, + "learning_rate": 2.3821507471261084e-06, + "loss": 0.5885, + "step": 8849 + }, + { + "epoch": 3.107443820224719, + "grad_norm": 0.9095465540885925, + "learning_rate": 2.381688937841079e-06, + "loss": 0.5338, + "step": 8850 + }, + { + "epoch": 3.107794943820225, + "grad_norm": 0.9628371596336365, + "learning_rate": 2.3812271326021887e-06, + "loss": 0.5559, + "step": 8851 + }, + { + "epoch": 3.1081460674157304, + "grad_norm": 0.9224883913993835, + "learning_rate": 2.3807653314252273e-06, + "loss": 0.5945, + "step": 8852 + }, + { + "epoch": 3.108497191011236, + "grad_norm": 0.9521452784538269, + "learning_rate": 2.3803035343259906e-06, + "loss": 0.5756, + "step": 8853 + }, + { + "epoch": 3.1088483146067416, + "grad_norm": 0.9127302169799805, + "learning_rate": 2.3798417413202697e-06, + "loss": 0.5364, + "step": 8854 + }, + { + "epoch": 3.109199438202247, + "grad_norm": 0.9071231484413147, + "learning_rate": 2.3793799524238597e-06, + "loss": 0.559, + "step": 8855 + }, + { + "epoch": 3.109550561797753, + "grad_norm": 0.9093103408813477, + "learning_rate": 2.3789181676525505e-06, + "loss": 0.5608, + "step": 8856 + }, + { + "epoch": 3.1099016853932584, + "grad_norm": 0.9037235975265503, + "learning_rate": 2.3784563870221377e-06, + "loss": 0.5431, + "step": 8857 + }, + { + "epoch": 3.110252808988764, + "grad_norm": 0.9057395458221436, + "learning_rate": 2.3779946105484116e-06, + "loss": 0.5661, + "step": 8858 + }, + { + "epoch": 3.1106039325842696, + "grad_norm": 0.9160523414611816, + "learning_rate": 2.377532838247166e-06, + "loss": 0.5318, + "step": 8859 + }, + { + "epoch": 3.110955056179775, + "grad_norm": 0.9095848798751831, + "learning_rate": 2.377071070134193e-06, + "loss": 0.5242, + "step": 8860 + }, + { + "epoch": 3.111306179775281, + "grad_norm": 0.9185169339179993, + "learning_rate": 2.3766093062252827e-06, + "loss": 0.5739, + "step": 8861 + }, + { + "epoch": 3.1116573033707864, + "grad_norm": 0.9330510497093201, + "learning_rate": 2.3761475465362298e-06, + "loss": 0.5556, + "step": 8862 + }, + { + "epoch": 3.112008426966292, + "grad_norm": 0.9140385985374451, + "learning_rate": 2.375685791082824e-06, + "loss": 0.5144, + "step": 8863 + }, + { + "epoch": 3.1123595505617976, + "grad_norm": 0.9058642387390137, + "learning_rate": 2.3752240398808586e-06, + "loss": 0.5616, + "step": 8864 + }, + { + "epoch": 3.112710674157303, + "grad_norm": 0.8959345817565918, + "learning_rate": 2.3747622929461232e-06, + "loss": 0.5435, + "step": 8865 + }, + { + "epoch": 3.113061797752809, + "grad_norm": 0.9209824204444885, + "learning_rate": 2.3743005502944112e-06, + "loss": 0.569, + "step": 8866 + }, + { + "epoch": 3.113412921348315, + "grad_norm": 0.9133533239364624, + "learning_rate": 2.373838811941512e-06, + "loss": 0.5434, + "step": 8867 + }, + { + "epoch": 3.1137640449438204, + "grad_norm": 0.9735870361328125, + "learning_rate": 2.3733770779032185e-06, + "loss": 0.5577, + "step": 8868 + }, + { + "epoch": 3.114115168539326, + "grad_norm": 0.90877366065979, + "learning_rate": 2.3729153481953194e-06, + "loss": 0.5587, + "step": 8869 + }, + { + "epoch": 3.1144662921348316, + "grad_norm": 0.9251682162284851, + "learning_rate": 2.3724536228336077e-06, + "loss": 0.5632, + "step": 8870 + }, + { + "epoch": 3.114817415730337, + "grad_norm": 0.9082393646240234, + "learning_rate": 2.3719919018338722e-06, + "loss": 0.5619, + "step": 8871 + }, + { + "epoch": 3.115168539325843, + "grad_norm": 0.8542476892471313, + "learning_rate": 2.371530185211905e-06, + "loss": 0.5175, + "step": 8872 + }, + { + "epoch": 3.1155196629213484, + "grad_norm": 0.9490677118301392, + "learning_rate": 2.3710684729834955e-06, + "loss": 0.5606, + "step": 8873 + }, + { + "epoch": 3.115870786516854, + "grad_norm": 0.9434581398963928, + "learning_rate": 2.3706067651644325e-06, + "loss": 0.5878, + "step": 8874 + }, + { + "epoch": 3.1162219101123596, + "grad_norm": 0.9130839109420776, + "learning_rate": 2.370145061770509e-06, + "loss": 0.5724, + "step": 8875 + }, + { + "epoch": 3.116573033707865, + "grad_norm": 0.9174863696098328, + "learning_rate": 2.3696833628175117e-06, + "loss": 0.5923, + "step": 8876 + }, + { + "epoch": 3.116924157303371, + "grad_norm": 0.9523452520370483, + "learning_rate": 2.3692216683212334e-06, + "loss": 0.5587, + "step": 8877 + }, + { + "epoch": 3.1172752808988764, + "grad_norm": 0.9290615320205688, + "learning_rate": 2.3687599782974606e-06, + "loss": 0.5525, + "step": 8878 + }, + { + "epoch": 3.117626404494382, + "grad_norm": 0.8855964541435242, + "learning_rate": 2.3682982927619858e-06, + "loss": 0.5157, + "step": 8879 + }, + { + "epoch": 3.1179775280898876, + "grad_norm": 0.8849876523017883, + "learning_rate": 2.367836611730596e-06, + "loss": 0.5131, + "step": 8880 + }, + { + "epoch": 3.118328651685393, + "grad_norm": 0.9100109934806824, + "learning_rate": 2.367374935219081e-06, + "loss": 0.497, + "step": 8881 + }, + { + "epoch": 3.118679775280899, + "grad_norm": 6.5442280769348145, + "learning_rate": 2.3669132632432294e-06, + "loss": 0.5166, + "step": 8882 + }, + { + "epoch": 3.1190308988764044, + "grad_norm": 0.9247830510139465, + "learning_rate": 2.3664515958188305e-06, + "loss": 0.5566, + "step": 8883 + }, + { + "epoch": 3.11938202247191, + "grad_norm": 0.9496949911117554, + "learning_rate": 2.3659899329616736e-06, + "loss": 0.5471, + "step": 8884 + }, + { + "epoch": 3.1197331460674156, + "grad_norm": 0.9136682152748108, + "learning_rate": 2.365528274687545e-06, + "loss": 0.53, + "step": 8885 + }, + { + "epoch": 3.120084269662921, + "grad_norm": 0.9163531064987183, + "learning_rate": 2.3650666210122354e-06, + "loss": 0.5829, + "step": 8886 + }, + { + "epoch": 3.120435393258427, + "grad_norm": 0.938612699508667, + "learning_rate": 2.364604971951531e-06, + "loss": 0.5289, + "step": 8887 + }, + { + "epoch": 3.1207865168539324, + "grad_norm": 0.924331545829773, + "learning_rate": 2.364143327521221e-06, + "loss": 0.5542, + "step": 8888 + }, + { + "epoch": 3.1211376404494384, + "grad_norm": 0.959979236125946, + "learning_rate": 2.3636816877370932e-06, + "loss": 0.5637, + "step": 8889 + }, + { + "epoch": 3.121488764044944, + "grad_norm": 0.9219009876251221, + "learning_rate": 2.363220052614935e-06, + "loss": 0.553, + "step": 8890 + }, + { + "epoch": 3.1218398876404496, + "grad_norm": 0.9053733348846436, + "learning_rate": 2.3627584221705334e-06, + "loss": 0.541, + "step": 8891 + }, + { + "epoch": 3.122191011235955, + "grad_norm": 0.9491739869117737, + "learning_rate": 2.362296796419677e-06, + "loss": 0.5565, + "step": 8892 + }, + { + "epoch": 3.122542134831461, + "grad_norm": 0.9550970196723938, + "learning_rate": 2.361835175378152e-06, + "loss": 0.5625, + "step": 8893 + }, + { + "epoch": 3.1228932584269664, + "grad_norm": 0.9412895441055298, + "learning_rate": 2.361373559061746e-06, + "loss": 0.5671, + "step": 8894 + }, + { + "epoch": 3.123244382022472, + "grad_norm": 0.9184961318969727, + "learning_rate": 2.360911947486245e-06, + "loss": 0.5608, + "step": 8895 + }, + { + "epoch": 3.1235955056179776, + "grad_norm": 0.9310684204101562, + "learning_rate": 2.3604503406674374e-06, + "loss": 0.5652, + "step": 8896 + }, + { + "epoch": 3.123946629213483, + "grad_norm": 0.9393607378005981, + "learning_rate": 2.3599887386211084e-06, + "loss": 0.5423, + "step": 8897 + }, + { + "epoch": 3.124297752808989, + "grad_norm": 0.9042835235595703, + "learning_rate": 2.359527141363045e-06, + "loss": 0.5346, + "step": 8898 + }, + { + "epoch": 3.1246488764044944, + "grad_norm": 0.9197384119033813, + "learning_rate": 2.3590655489090323e-06, + "loss": 0.5724, + "step": 8899 + }, + { + "epoch": 3.125, + "grad_norm": 0.9438782930374146, + "learning_rate": 2.3586039612748576e-06, + "loss": 0.5435, + "step": 8900 + }, + { + "epoch": 3.1253511235955056, + "grad_norm": 0.9054040312767029, + "learning_rate": 2.3581423784763062e-06, + "loss": 0.5546, + "step": 8901 + }, + { + "epoch": 3.125702247191011, + "grad_norm": 0.9769840836524963, + "learning_rate": 2.3576808005291645e-06, + "loss": 0.5947, + "step": 8902 + }, + { + "epoch": 3.126053370786517, + "grad_norm": 0.912523627281189, + "learning_rate": 2.357219227449218e-06, + "loss": 0.5738, + "step": 8903 + }, + { + "epoch": 3.1264044943820224, + "grad_norm": 0.9094953536987305, + "learning_rate": 2.3567576592522505e-06, + "loss": 0.5326, + "step": 8904 + }, + { + "epoch": 3.126755617977528, + "grad_norm": 0.9189884662628174, + "learning_rate": 2.3562960959540498e-06, + "loss": 0.5376, + "step": 8905 + }, + { + "epoch": 3.1271067415730336, + "grad_norm": 0.8994994163513184, + "learning_rate": 2.3558345375703985e-06, + "loss": 0.5758, + "step": 8906 + }, + { + "epoch": 3.127457865168539, + "grad_norm": 0.9371024966239929, + "learning_rate": 2.3553729841170834e-06, + "loss": 0.5711, + "step": 8907 + }, + { + "epoch": 3.127808988764045, + "grad_norm": 0.9175930023193359, + "learning_rate": 2.354911435609888e-06, + "loss": 0.5731, + "step": 8908 + }, + { + "epoch": 3.1281601123595504, + "grad_norm": 0.9270338416099548, + "learning_rate": 2.354449892064598e-06, + "loss": 0.5521, + "step": 8909 + }, + { + "epoch": 3.128511235955056, + "grad_norm": 0.9249902367591858, + "learning_rate": 2.3539883534969972e-06, + "loss": 0.5527, + "step": 8910 + }, + { + "epoch": 3.1288623595505616, + "grad_norm": 0.9235257506370544, + "learning_rate": 2.353526819922869e-06, + "loss": 0.5419, + "step": 8911 + }, + { + "epoch": 3.1292134831460676, + "grad_norm": 0.9523985385894775, + "learning_rate": 2.353065291357999e-06, + "loss": 0.5858, + "step": 8912 + }, + { + "epoch": 3.129564606741573, + "grad_norm": 0.9682888984680176, + "learning_rate": 2.352603767818169e-06, + "loss": 0.5722, + "step": 8913 + }, + { + "epoch": 3.129915730337079, + "grad_norm": 0.9295958876609802, + "learning_rate": 2.3521422493191656e-06, + "loss": 0.5679, + "step": 8914 + }, + { + "epoch": 3.1302668539325844, + "grad_norm": 0.9561300873756409, + "learning_rate": 2.3516807358767693e-06, + "loss": 0.5693, + "step": 8915 + }, + { + "epoch": 3.13061797752809, + "grad_norm": 0.9593166708946228, + "learning_rate": 2.3512192275067663e-06, + "loss": 0.5888, + "step": 8916 + }, + { + "epoch": 3.1309691011235956, + "grad_norm": 0.9141625761985779, + "learning_rate": 2.350757724224937e-06, + "loss": 0.5663, + "step": 8917 + }, + { + "epoch": 3.131320224719101, + "grad_norm": 0.9452791810035706, + "learning_rate": 2.3502962260470673e-06, + "loss": 0.5599, + "step": 8918 + }, + { + "epoch": 3.131671348314607, + "grad_norm": 0.9614371657371521, + "learning_rate": 2.349834732988937e-06, + "loss": 0.5779, + "step": 8919 + }, + { + "epoch": 3.1320224719101124, + "grad_norm": 0.9115386009216309, + "learning_rate": 2.3493732450663317e-06, + "loss": 0.5704, + "step": 8920 + }, + { + "epoch": 3.132373595505618, + "grad_norm": 0.9307755827903748, + "learning_rate": 2.3489117622950315e-06, + "loss": 0.5717, + "step": 8921 + }, + { + "epoch": 3.1327247191011236, + "grad_norm": 0.9371019005775452, + "learning_rate": 2.3484502846908207e-06, + "loss": 0.5298, + "step": 8922 + }, + { + "epoch": 3.133075842696629, + "grad_norm": 0.9290947914123535, + "learning_rate": 2.3479888122694804e-06, + "loss": 0.5636, + "step": 8923 + }, + { + "epoch": 3.133426966292135, + "grad_norm": 0.9441036581993103, + "learning_rate": 2.3475273450467917e-06, + "loss": 0.5647, + "step": 8924 + }, + { + "epoch": 3.1337780898876404, + "grad_norm": 0.9305264949798584, + "learning_rate": 2.3470658830385383e-06, + "loss": 0.5584, + "step": 8925 + }, + { + "epoch": 3.134129213483146, + "grad_norm": 0.9293364882469177, + "learning_rate": 2.3466044262604998e-06, + "loss": 0.5372, + "step": 8926 + }, + { + "epoch": 3.1344803370786516, + "grad_norm": 0.9441460371017456, + "learning_rate": 2.34614297472846e-06, + "loss": 0.5553, + "step": 8927 + }, + { + "epoch": 3.134831460674157, + "grad_norm": 0.9568279981613159, + "learning_rate": 2.3456815284581976e-06, + "loss": 0.5749, + "step": 8928 + }, + { + "epoch": 3.135182584269663, + "grad_norm": 0.982342004776001, + "learning_rate": 2.3452200874654956e-06, + "loss": 0.5494, + "step": 8929 + }, + { + "epoch": 3.1355337078651684, + "grad_norm": 0.9266687631607056, + "learning_rate": 2.3447586517661333e-06, + "loss": 0.5485, + "step": 8930 + }, + { + "epoch": 3.135884831460674, + "grad_norm": 0.9465826153755188, + "learning_rate": 2.3442972213758934e-06, + "loss": 0.5709, + "step": 8931 + }, + { + "epoch": 3.1362359550561796, + "grad_norm": 0.9068037867546082, + "learning_rate": 2.343835796310554e-06, + "loss": 0.5541, + "step": 8932 + }, + { + "epoch": 3.1365870786516856, + "grad_norm": 0.9190397262573242, + "learning_rate": 2.3433743765858983e-06, + "loss": 0.5386, + "step": 8933 + }, + { + "epoch": 3.136938202247191, + "grad_norm": 0.9141831398010254, + "learning_rate": 2.3429129622177034e-06, + "loss": 0.5431, + "step": 8934 + }, + { + "epoch": 3.137289325842697, + "grad_norm": 0.9500133395195007, + "learning_rate": 2.342451553221752e-06, + "loss": 0.5738, + "step": 8935 + }, + { + "epoch": 3.1376404494382024, + "grad_norm": 0.9465636610984802, + "learning_rate": 2.3419901496138227e-06, + "loss": 0.5975, + "step": 8936 + }, + { + "epoch": 3.137991573033708, + "grad_norm": 0.8895576000213623, + "learning_rate": 2.341528751409694e-06, + "loss": 0.5149, + "step": 8937 + }, + { + "epoch": 3.1383426966292136, + "grad_norm": 0.9108017086982727, + "learning_rate": 2.3410673586251475e-06, + "loss": 0.5662, + "step": 8938 + }, + { + "epoch": 3.138693820224719, + "grad_norm": 0.9355370998382568, + "learning_rate": 2.3406059712759598e-06, + "loss": 0.5738, + "step": 8939 + }, + { + "epoch": 3.139044943820225, + "grad_norm": 0.9378455281257629, + "learning_rate": 2.3401445893779128e-06, + "loss": 0.5687, + "step": 8940 + }, + { + "epoch": 3.1393960674157304, + "grad_norm": 1.020998477935791, + "learning_rate": 2.3396832129467833e-06, + "loss": 0.5642, + "step": 8941 + }, + { + "epoch": 3.139747191011236, + "grad_norm": 0.8965001702308655, + "learning_rate": 2.339221841998352e-06, + "loss": 0.5538, + "step": 8942 + }, + { + "epoch": 3.1400983146067416, + "grad_norm": 0.938067615032196, + "learning_rate": 2.3387604765483943e-06, + "loss": 0.5675, + "step": 8943 + }, + { + "epoch": 3.140449438202247, + "grad_norm": 0.8801164031028748, + "learning_rate": 2.338299116612692e-06, + "loss": 0.5307, + "step": 8944 + }, + { + "epoch": 3.140800561797753, + "grad_norm": 0.8848679065704346, + "learning_rate": 2.33783776220702e-06, + "loss": 0.5467, + "step": 8945 + }, + { + "epoch": 3.1411516853932584, + "grad_norm": 0.9651525020599365, + "learning_rate": 2.337376413347159e-06, + "loss": 0.5526, + "step": 8946 + }, + { + "epoch": 3.141502808988764, + "grad_norm": 0.9414547085762024, + "learning_rate": 2.336915070048886e-06, + "loss": 0.589, + "step": 8947 + }, + { + "epoch": 3.1418539325842696, + "grad_norm": 0.9873248338699341, + "learning_rate": 2.336453732327976e-06, + "loss": 0.5682, + "step": 8948 + }, + { + "epoch": 3.142205056179775, + "grad_norm": 0.9017943143844604, + "learning_rate": 2.3359924002002103e-06, + "loss": 0.5271, + "step": 8949 + }, + { + "epoch": 3.142556179775281, + "grad_norm": 0.9191378951072693, + "learning_rate": 2.335531073681363e-06, + "loss": 0.5499, + "step": 8950 + }, + { + "epoch": 3.1429073033707864, + "grad_norm": 1.0262577533721924, + "learning_rate": 2.335069752787213e-06, + "loss": 0.5739, + "step": 8951 + }, + { + "epoch": 3.143258426966292, + "grad_norm": 0.9117028713226318, + "learning_rate": 2.3346084375335355e-06, + "loss": 0.5779, + "step": 8952 + }, + { + "epoch": 3.1436095505617976, + "grad_norm": 0.9521639347076416, + "learning_rate": 2.3341471279361084e-06, + "loss": 0.6043, + "step": 8953 + }, + { + "epoch": 3.143960674157303, + "grad_norm": 0.9468163847923279, + "learning_rate": 2.333685824010707e-06, + "loss": 0.5566, + "step": 8954 + }, + { + "epoch": 3.144311797752809, + "grad_norm": 0.9359827041625977, + "learning_rate": 2.3332245257731097e-06, + "loss": 0.5543, + "step": 8955 + }, + { + "epoch": 3.1446629213483144, + "grad_norm": 0.9230249524116516, + "learning_rate": 2.332763233239089e-06, + "loss": 0.5721, + "step": 8956 + }, + { + "epoch": 3.1450140449438204, + "grad_norm": 0.9995796084403992, + "learning_rate": 2.332301946424424e-06, + "loss": 0.5526, + "step": 8957 + }, + { + "epoch": 3.145365168539326, + "grad_norm": 1.095003366470337, + "learning_rate": 2.3318406653448875e-06, + "loss": 0.576, + "step": 8958 + }, + { + "epoch": 3.1457162921348316, + "grad_norm": 0.9420155882835388, + "learning_rate": 2.331379390016258e-06, + "loss": 0.5849, + "step": 8959 + }, + { + "epoch": 3.146067415730337, + "grad_norm": 0.9077823758125305, + "learning_rate": 2.330918120454309e-06, + "loss": 0.5465, + "step": 8960 + }, + { + "epoch": 3.146418539325843, + "grad_norm": 0.9510911703109741, + "learning_rate": 2.3304568566748144e-06, + "loss": 0.5847, + "step": 8961 + }, + { + "epoch": 3.1467696629213484, + "grad_norm": 0.9385233521461487, + "learning_rate": 2.3299955986935515e-06, + "loss": 0.5697, + "step": 8962 + }, + { + "epoch": 3.147120786516854, + "grad_norm": 0.9283239245414734, + "learning_rate": 2.3295343465262922e-06, + "loss": 0.5716, + "step": 8963 + }, + { + "epoch": 3.1474719101123596, + "grad_norm": 0.9338172674179077, + "learning_rate": 2.3290731001888135e-06, + "loss": 0.5523, + "step": 8964 + }, + { + "epoch": 3.147823033707865, + "grad_norm": 0.9136307239532471, + "learning_rate": 2.3286118596968874e-06, + "loss": 0.5593, + "step": 8965 + }, + { + "epoch": 3.148174157303371, + "grad_norm": 0.9847103953361511, + "learning_rate": 2.3281506250662904e-06, + "loss": 0.5324, + "step": 8966 + }, + { + "epoch": 3.1485252808988764, + "grad_norm": 0.909062385559082, + "learning_rate": 2.3276893963127943e-06, + "loss": 0.5553, + "step": 8967 + }, + { + "epoch": 3.148876404494382, + "grad_norm": 0.9283292293548584, + "learning_rate": 2.3272281734521732e-06, + "loss": 0.5665, + "step": 8968 + }, + { + "epoch": 3.1492275280898876, + "grad_norm": 0.9497380256652832, + "learning_rate": 2.3267669565002012e-06, + "loss": 0.5633, + "step": 8969 + }, + { + "epoch": 3.149578651685393, + "grad_norm": 0.9020211696624756, + "learning_rate": 2.326305745472651e-06, + "loss": 0.5363, + "step": 8970 + }, + { + "epoch": 3.149929775280899, + "grad_norm": 0.9506552815437317, + "learning_rate": 2.325844540385295e-06, + "loss": 0.5479, + "step": 8971 + }, + { + "epoch": 3.1502808988764044, + "grad_norm": 0.9471758008003235, + "learning_rate": 2.325383341253907e-06, + "loss": 0.5998, + "step": 8972 + }, + { + "epoch": 3.15063202247191, + "grad_norm": 0.9534868597984314, + "learning_rate": 2.3249221480942598e-06, + "loss": 0.5713, + "step": 8973 + }, + { + "epoch": 3.1509831460674156, + "grad_norm": 0.9526163935661316, + "learning_rate": 2.324460960922125e-06, + "loss": 0.6194, + "step": 8974 + }, + { + "epoch": 3.151334269662921, + "grad_norm": 0.9661351442337036, + "learning_rate": 2.323999779753275e-06, + "loss": 0.5569, + "step": 8975 + }, + { + "epoch": 3.151685393258427, + "grad_norm": 0.9209139347076416, + "learning_rate": 2.3235386046034823e-06, + "loss": 0.5623, + "step": 8976 + }, + { + "epoch": 3.1520365168539324, + "grad_norm": 0.9163221716880798, + "learning_rate": 2.323077435488518e-06, + "loss": 0.5292, + "step": 8977 + }, + { + "epoch": 3.1523876404494384, + "grad_norm": 0.9472640752792358, + "learning_rate": 2.322616272424154e-06, + "loss": 0.5982, + "step": 8978 + }, + { + "epoch": 3.152738764044944, + "grad_norm": 0.9668310880661011, + "learning_rate": 2.3221551154261616e-06, + "loss": 0.5739, + "step": 8979 + }, + { + "epoch": 3.1530898876404496, + "grad_norm": 0.9095568060874939, + "learning_rate": 2.3216939645103122e-06, + "loss": 0.5286, + "step": 8980 + }, + { + "epoch": 3.153441011235955, + "grad_norm": 0.904996395111084, + "learning_rate": 2.3212328196923766e-06, + "loss": 0.5433, + "step": 8981 + }, + { + "epoch": 3.153792134831461, + "grad_norm": 0.9338881373405457, + "learning_rate": 2.3207716809881254e-06, + "loss": 0.5861, + "step": 8982 + }, + { + "epoch": 3.1541432584269664, + "grad_norm": 1.015492558479309, + "learning_rate": 2.3203105484133302e-06, + "loss": 0.5381, + "step": 8983 + }, + { + "epoch": 3.154494382022472, + "grad_norm": 0.9350138306617737, + "learning_rate": 2.3198494219837596e-06, + "loss": 0.5622, + "step": 8984 + }, + { + "epoch": 3.1548455056179776, + "grad_norm": 0.934749960899353, + "learning_rate": 2.319388301715185e-06, + "loss": 0.5422, + "step": 8985 + }, + { + "epoch": 3.155196629213483, + "grad_norm": 0.9269441366195679, + "learning_rate": 2.318927187623376e-06, + "loss": 0.569, + "step": 8986 + }, + { + "epoch": 3.155547752808989, + "grad_norm": 0.9028807282447815, + "learning_rate": 2.318466079724102e-06, + "loss": 0.5157, + "step": 8987 + }, + { + "epoch": 3.1558988764044944, + "grad_norm": 0.9250398278236389, + "learning_rate": 2.318004978033133e-06, + "loss": 0.5473, + "step": 8988 + }, + { + "epoch": 3.15625, + "grad_norm": 0.917531430721283, + "learning_rate": 2.3175438825662377e-06, + "loss": 0.5803, + "step": 8989 + }, + { + "epoch": 3.1566011235955056, + "grad_norm": 0.9266380071640015, + "learning_rate": 2.317082793339186e-06, + "loss": 0.5465, + "step": 8990 + }, + { + "epoch": 3.156952247191011, + "grad_norm": 0.9941410422325134, + "learning_rate": 2.316621710367745e-06, + "loss": 0.5685, + "step": 8991 + }, + { + "epoch": 3.157303370786517, + "grad_norm": 0.9138315320014954, + "learning_rate": 2.316160633667686e-06, + "loss": 0.5501, + "step": 8992 + }, + { + "epoch": 3.1576544943820224, + "grad_norm": 0.9415439963340759, + "learning_rate": 2.3156995632547747e-06, + "loss": 0.5591, + "step": 8993 + }, + { + "epoch": 3.158005617977528, + "grad_norm": 0.9314430356025696, + "learning_rate": 2.315238499144782e-06, + "loss": 0.529, + "step": 8994 + }, + { + "epoch": 3.1583567415730336, + "grad_norm": 0.9484070539474487, + "learning_rate": 2.3147774413534737e-06, + "loss": 0.5572, + "step": 8995 + }, + { + "epoch": 3.158707865168539, + "grad_norm": 0.9399926066398621, + "learning_rate": 2.3143163898966187e-06, + "loss": 0.5455, + "step": 8996 + }, + { + "epoch": 3.159058988764045, + "grad_norm": 0.9525354504585266, + "learning_rate": 2.313855344789984e-06, + "loss": 0.5754, + "step": 8997 + }, + { + "epoch": 3.1594101123595504, + "grad_norm": 0.9502968192100525, + "learning_rate": 2.3133943060493376e-06, + "loss": 0.5748, + "step": 8998 + }, + { + "epoch": 3.159761235955056, + "grad_norm": 0.8968663215637207, + "learning_rate": 2.312933273690447e-06, + "loss": 0.5423, + "step": 8999 + }, + { + "epoch": 3.1601123595505616, + "grad_norm": 0.8691650629043579, + "learning_rate": 2.3124722477290765e-06, + "loss": 0.4896, + "step": 9000 + }, + { + "epoch": 3.1604634831460676, + "grad_norm": 0.9302195906639099, + "learning_rate": 2.3120112281809962e-06, + "loss": 0.5565, + "step": 9001 + }, + { + "epoch": 3.160814606741573, + "grad_norm": 0.9128553867340088, + "learning_rate": 2.3115502150619697e-06, + "loss": 0.5524, + "step": 9002 + }, + { + "epoch": 3.161165730337079, + "grad_norm": 0.9269843697547913, + "learning_rate": 2.3110892083877658e-06, + "loss": 0.5493, + "step": 9003 + }, + { + "epoch": 3.1615168539325844, + "grad_norm": 0.904180109500885, + "learning_rate": 2.310628208174148e-06, + "loss": 0.5695, + "step": 9004 + }, + { + "epoch": 3.16186797752809, + "grad_norm": 0.8929877281188965, + "learning_rate": 2.310167214436885e-06, + "loss": 0.5569, + "step": 9005 + }, + { + "epoch": 3.1622191011235956, + "grad_norm": 0.938226044178009, + "learning_rate": 2.3097062271917394e-06, + "loss": 0.5589, + "step": 9006 + }, + { + "epoch": 3.162570224719101, + "grad_norm": 0.9409511089324951, + "learning_rate": 2.309245246454479e-06, + "loss": 0.549, + "step": 9007 + }, + { + "epoch": 3.162921348314607, + "grad_norm": 0.9133210182189941, + "learning_rate": 2.3087842722408673e-06, + "loss": 0.5322, + "step": 9008 + }, + { + "epoch": 3.1632724719101124, + "grad_norm": 0.9695359468460083, + "learning_rate": 2.308323304566671e-06, + "loss": 0.6047, + "step": 9009 + }, + { + "epoch": 3.163623595505618, + "grad_norm": 0.9128788113594055, + "learning_rate": 2.307862343447653e-06, + "loss": 0.5704, + "step": 9010 + }, + { + "epoch": 3.1639747191011236, + "grad_norm": 0.9297662973403931, + "learning_rate": 2.3074013888995775e-06, + "loss": 0.5652, + "step": 9011 + }, + { + "epoch": 3.164325842696629, + "grad_norm": 0.9448683857917786, + "learning_rate": 2.306940440938211e-06, + "loss": 0.5764, + "step": 9012 + }, + { + "epoch": 3.164676966292135, + "grad_norm": 0.9009461402893066, + "learning_rate": 2.306479499579315e-06, + "loss": 0.5388, + "step": 9013 + }, + { + "epoch": 3.1650280898876404, + "grad_norm": 0.911253809928894, + "learning_rate": 2.306018564838656e-06, + "loss": 0.5519, + "step": 9014 + }, + { + "epoch": 3.165379213483146, + "grad_norm": 0.9532901048660278, + "learning_rate": 2.3055576367319945e-06, + "loss": 0.5619, + "step": 9015 + }, + { + "epoch": 3.1657303370786516, + "grad_norm": 0.9588057994842529, + "learning_rate": 2.3050967152750966e-06, + "loss": 0.5449, + "step": 9016 + }, + { + "epoch": 3.166081460674157, + "grad_norm": 0.9382750391960144, + "learning_rate": 2.3046358004837233e-06, + "loss": 0.5505, + "step": 9017 + }, + { + "epoch": 3.166432584269663, + "grad_norm": 0.8919453620910645, + "learning_rate": 2.3041748923736397e-06, + "loss": 0.5486, + "step": 9018 + }, + { + "epoch": 3.1667837078651684, + "grad_norm": 0.9534412622451782, + "learning_rate": 2.303713990960606e-06, + "loss": 0.5559, + "step": 9019 + }, + { + "epoch": 3.167134831460674, + "grad_norm": 0.9397498369216919, + "learning_rate": 2.3032530962603865e-06, + "loss": 0.5455, + "step": 9020 + }, + { + "epoch": 3.1674859550561796, + "grad_norm": 0.9306217432022095, + "learning_rate": 2.302792208288742e-06, + "loss": 0.5606, + "step": 9021 + }, + { + "epoch": 3.1678370786516856, + "grad_norm": 0.9721474647521973, + "learning_rate": 2.3023313270614368e-06, + "loss": 0.5637, + "step": 9022 + }, + { + "epoch": 3.168188202247191, + "grad_norm": 0.9403707385063171, + "learning_rate": 2.3018704525942304e-06, + "loss": 0.5479, + "step": 9023 + }, + { + "epoch": 3.168539325842697, + "grad_norm": 0.9606196284294128, + "learning_rate": 2.3014095849028843e-06, + "loss": 0.5188, + "step": 9024 + }, + { + "epoch": 3.1688904494382024, + "grad_norm": 0.937238335609436, + "learning_rate": 2.3009487240031615e-06, + "loss": 0.5644, + "step": 9025 + }, + { + "epoch": 3.169241573033708, + "grad_norm": 0.940178394317627, + "learning_rate": 2.3004878699108206e-06, + "loss": 0.5246, + "step": 9026 + }, + { + "epoch": 3.1695926966292136, + "grad_norm": 0.9183790683746338, + "learning_rate": 2.300027022641625e-06, + "loss": 0.528, + "step": 9027 + }, + { + "epoch": 3.169943820224719, + "grad_norm": 0.9336768388748169, + "learning_rate": 2.299566182211333e-06, + "loss": 0.5484, + "step": 9028 + }, + { + "epoch": 3.170294943820225, + "grad_norm": 0.928656816482544, + "learning_rate": 2.299105348635707e-06, + "loss": 0.543, + "step": 9029 + }, + { + "epoch": 3.1706460674157304, + "grad_norm": 0.9587504267692566, + "learning_rate": 2.2986445219305054e-06, + "loss": 0.5641, + "step": 9030 + }, + { + "epoch": 3.170997191011236, + "grad_norm": 0.9122671484947205, + "learning_rate": 2.29818370211149e-06, + "loss": 0.5691, + "step": 9031 + }, + { + "epoch": 3.1713483146067416, + "grad_norm": 0.9110110998153687, + "learning_rate": 2.297722889194418e-06, + "loss": 0.5676, + "step": 9032 + }, + { + "epoch": 3.171699438202247, + "grad_norm": 0.8846708536148071, + "learning_rate": 2.2972620831950516e-06, + "loss": 0.5148, + "step": 9033 + }, + { + "epoch": 3.172050561797753, + "grad_norm": 0.9163132309913635, + "learning_rate": 2.296801284129147e-06, + "loss": 0.5554, + "step": 9034 + }, + { + "epoch": 3.1724016853932584, + "grad_norm": 0.9429922103881836, + "learning_rate": 2.2963404920124656e-06, + "loss": 0.6062, + "step": 9035 + }, + { + "epoch": 3.172752808988764, + "grad_norm": 0.9419503808021545, + "learning_rate": 2.2958797068607652e-06, + "loss": 0.5543, + "step": 9036 + }, + { + "epoch": 3.1731039325842696, + "grad_norm": 0.9216638803482056, + "learning_rate": 2.295418928689803e-06, + "loss": 0.5538, + "step": 9037 + }, + { + "epoch": 3.173455056179775, + "grad_norm": 0.9066448211669922, + "learning_rate": 2.2949581575153397e-06, + "loss": 0.5433, + "step": 9038 + }, + { + "epoch": 3.173806179775281, + "grad_norm": 0.9317613840103149, + "learning_rate": 2.294497393353131e-06, + "loss": 0.5968, + "step": 9039 + }, + { + "epoch": 3.1741573033707864, + "grad_norm": 0.9725520014762878, + "learning_rate": 2.2940366362189363e-06, + "loss": 0.5653, + "step": 9040 + }, + { + "epoch": 3.174508426966292, + "grad_norm": 0.9258610606193542, + "learning_rate": 2.2935758861285113e-06, + "loss": 0.5542, + "step": 9041 + }, + { + "epoch": 3.1748595505617976, + "grad_norm": 0.935573160648346, + "learning_rate": 2.2931151430976156e-06, + "loss": 0.576, + "step": 9042 + }, + { + "epoch": 3.175210674157303, + "grad_norm": 0.9360717535018921, + "learning_rate": 2.2926544071420038e-06, + "loss": 0.5326, + "step": 9043 + }, + { + "epoch": 3.175561797752809, + "grad_norm": 0.9399312734603882, + "learning_rate": 2.292193678277435e-06, + "loss": 0.5755, + "step": 9044 + }, + { + "epoch": 3.1759129213483144, + "grad_norm": 0.9361773133277893, + "learning_rate": 2.2917329565196637e-06, + "loss": 0.5653, + "step": 9045 + }, + { + "epoch": 3.1762640449438204, + "grad_norm": 0.9085448980331421, + "learning_rate": 2.2912722418844476e-06, + "loss": 0.5434, + "step": 9046 + }, + { + "epoch": 3.176615168539326, + "grad_norm": 0.9225165843963623, + "learning_rate": 2.2908115343875424e-06, + "loss": 0.5791, + "step": 9047 + }, + { + "epoch": 3.1769662921348316, + "grad_norm": 0.9260159134864807, + "learning_rate": 2.2903508340447033e-06, + "loss": 0.5543, + "step": 9048 + }, + { + "epoch": 3.177317415730337, + "grad_norm": 0.9076563119888306, + "learning_rate": 2.289890140871687e-06, + "loss": 0.5219, + "step": 9049 + }, + { + "epoch": 3.177668539325843, + "grad_norm": 0.9252681136131287, + "learning_rate": 2.2894294548842465e-06, + "loss": 0.5613, + "step": 9050 + }, + { + "epoch": 3.1780196629213484, + "grad_norm": 0.9024373888969421, + "learning_rate": 2.28896877609814e-06, + "loss": 0.5257, + "step": 9051 + }, + { + "epoch": 3.178370786516854, + "grad_norm": 0.905224621295929, + "learning_rate": 2.2885081045291194e-06, + "loss": 0.5768, + "step": 9052 + }, + { + "epoch": 3.1787219101123596, + "grad_norm": 0.9322327375411987, + "learning_rate": 2.288047440192942e-06, + "loss": 0.5675, + "step": 9053 + }, + { + "epoch": 3.179073033707865, + "grad_norm": 0.9132179617881775, + "learning_rate": 2.2875867831053602e-06, + "loss": 0.5482, + "step": 9054 + }, + { + "epoch": 3.179424157303371, + "grad_norm": 0.9264069199562073, + "learning_rate": 2.2871261332821284e-06, + "loss": 0.5717, + "step": 9055 + }, + { + "epoch": 3.1797752808988764, + "grad_norm": 0.9629480242729187, + "learning_rate": 2.286665490739001e-06, + "loss": 0.5895, + "step": 9056 + }, + { + "epoch": 3.180126404494382, + "grad_norm": 0.9351604580879211, + "learning_rate": 2.286204855491731e-06, + "loss": 0.5404, + "step": 9057 + }, + { + "epoch": 3.1804775280898876, + "grad_norm": 0.9364950060844421, + "learning_rate": 2.2857442275560725e-06, + "loss": 0.5835, + "step": 9058 + }, + { + "epoch": 3.180828651685393, + "grad_norm": 0.9281259179115295, + "learning_rate": 2.2852836069477773e-06, + "loss": 0.5742, + "step": 9059 + }, + { + "epoch": 3.181179775280899, + "grad_norm": 0.9376089572906494, + "learning_rate": 2.2848229936825997e-06, + "loss": 0.5635, + "step": 9060 + }, + { + "epoch": 3.1815308988764044, + "grad_norm": 1.0491068363189697, + "learning_rate": 2.284362387776292e-06, + "loss": 0.5278, + "step": 9061 + }, + { + "epoch": 3.18188202247191, + "grad_norm": 0.9894011616706848, + "learning_rate": 2.283901789244605e-06, + "loss": 0.5551, + "step": 9062 + }, + { + "epoch": 3.1822331460674156, + "grad_norm": 2.0815200805664062, + "learning_rate": 2.2834411981032925e-06, + "loss": 0.5978, + "step": 9063 + }, + { + "epoch": 3.182584269662921, + "grad_norm": 0.920780599117279, + "learning_rate": 2.2829806143681055e-06, + "loss": 0.5157, + "step": 9064 + }, + { + "epoch": 3.182935393258427, + "grad_norm": 0.9750953316688538, + "learning_rate": 2.282520038054796e-06, + "loss": 0.5616, + "step": 9065 + }, + { + "epoch": 3.1832865168539324, + "grad_norm": 0.9077079892158508, + "learning_rate": 2.282059469179115e-06, + "loss": 0.5336, + "step": 9066 + }, + { + "epoch": 3.1836376404494384, + "grad_norm": 0.9596759080886841, + "learning_rate": 2.2815989077568136e-06, + "loss": 0.5639, + "step": 9067 + }, + { + "epoch": 3.183988764044944, + "grad_norm": 0.9222280979156494, + "learning_rate": 2.2811383538036426e-06, + "loss": 0.5614, + "step": 9068 + }, + { + "epoch": 3.1843398876404496, + "grad_norm": 0.9478800892829895, + "learning_rate": 2.2806778073353523e-06, + "loss": 0.5779, + "step": 9069 + }, + { + "epoch": 3.184691011235955, + "grad_norm": 0.8949623703956604, + "learning_rate": 2.280217268367694e-06, + "loss": 0.5389, + "step": 9070 + }, + { + "epoch": 3.185042134831461, + "grad_norm": 0.9100927710533142, + "learning_rate": 2.2797567369164167e-06, + "loss": 0.5844, + "step": 9071 + }, + { + "epoch": 3.1853932584269664, + "grad_norm": 0.9618512988090515, + "learning_rate": 2.279296212997271e-06, + "loss": 0.5509, + "step": 9072 + }, + { + "epoch": 3.185744382022472, + "grad_norm": 0.9218263626098633, + "learning_rate": 2.2788356966260056e-06, + "loss": 0.5636, + "step": 9073 + }, + { + "epoch": 3.1860955056179776, + "grad_norm": 0.8963178992271423, + "learning_rate": 2.2783751878183697e-06, + "loss": 0.5202, + "step": 9074 + }, + { + "epoch": 3.186446629213483, + "grad_norm": 0.8807015419006348, + "learning_rate": 2.2779146865901134e-06, + "loss": 0.5081, + "step": 9075 + }, + { + "epoch": 3.186797752808989, + "grad_norm": 0.9365087747573853, + "learning_rate": 2.2774541929569842e-06, + "loss": 0.58, + "step": 9076 + }, + { + "epoch": 3.1871488764044944, + "grad_norm": 0.9479805827140808, + "learning_rate": 2.276993706934732e-06, + "loss": 0.5754, + "step": 9077 + }, + { + "epoch": 3.1875, + "grad_norm": 0.9108830094337463, + "learning_rate": 2.276533228539103e-06, + "loss": 0.5865, + "step": 9078 + }, + { + "epoch": 3.1878511235955056, + "grad_norm": 0.938982367515564, + "learning_rate": 2.276072757785848e-06, + "loss": 0.5788, + "step": 9079 + }, + { + "epoch": 3.188202247191011, + "grad_norm": 0.9025770425796509, + "learning_rate": 2.275612294690711e-06, + "loss": 0.5328, + "step": 9080 + }, + { + "epoch": 3.188553370786517, + "grad_norm": 0.9685847759246826, + "learning_rate": 2.2751518392694434e-06, + "loss": 0.5712, + "step": 9081 + }, + { + "epoch": 3.1889044943820224, + "grad_norm": 0.9372316002845764, + "learning_rate": 2.2746913915377886e-06, + "loss": 0.5916, + "step": 9082 + }, + { + "epoch": 3.189255617977528, + "grad_norm": 0.921090304851532, + "learning_rate": 2.2742309515114973e-06, + "loss": 0.5392, + "step": 9083 + }, + { + "epoch": 3.1896067415730336, + "grad_norm": 0.9218523502349854, + "learning_rate": 2.2737705192063127e-06, + "loss": 0.5576, + "step": 9084 + }, + { + "epoch": 3.189957865168539, + "grad_norm": 0.9160637855529785, + "learning_rate": 2.2733100946379843e-06, + "loss": 0.5793, + "step": 9085 + }, + { + "epoch": 3.190308988764045, + "grad_norm": 0.9549495577812195, + "learning_rate": 2.272849677822256e-06, + "loss": 0.577, + "step": 9086 + }, + { + "epoch": 3.1906601123595504, + "grad_norm": 0.9374105930328369, + "learning_rate": 2.272389268774874e-06, + "loss": 0.5664, + "step": 9087 + }, + { + "epoch": 3.191011235955056, + "grad_norm": 0.9102039933204651, + "learning_rate": 2.2719288675115846e-06, + "loss": 0.522, + "step": 9088 + }, + { + "epoch": 3.1913623595505616, + "grad_norm": 0.9222140312194824, + "learning_rate": 2.271468474048132e-06, + "loss": 0.5435, + "step": 9089 + }, + { + "epoch": 3.1917134831460676, + "grad_norm": 0.9319133758544922, + "learning_rate": 2.2710080884002634e-06, + "loss": 0.5269, + "step": 9090 + }, + { + "epoch": 3.192064606741573, + "grad_norm": 0.9387129545211792, + "learning_rate": 2.270547710583721e-06, + "loss": 0.6035, + "step": 9091 + }, + { + "epoch": 3.192415730337079, + "grad_norm": 0.9357312321662903, + "learning_rate": 2.2700873406142517e-06, + "loss": 0.5598, + "step": 9092 + }, + { + "epoch": 3.1927668539325844, + "grad_norm": 0.9478253126144409, + "learning_rate": 2.269626978507597e-06, + "loss": 0.5589, + "step": 9093 + }, + { + "epoch": 3.19311797752809, + "grad_norm": 0.9411143064498901, + "learning_rate": 2.2691666242795042e-06, + "loss": 0.5421, + "step": 9094 + }, + { + "epoch": 3.1934691011235956, + "grad_norm": 0.9287983179092407, + "learning_rate": 2.2687062779457138e-06, + "loss": 0.5696, + "step": 9095 + }, + { + "epoch": 3.193820224719101, + "grad_norm": 0.9401189088821411, + "learning_rate": 2.268245939521972e-06, + "loss": 0.5403, + "step": 9096 + }, + { + "epoch": 3.194171348314607, + "grad_norm": 0.9173681735992432, + "learning_rate": 2.2677856090240196e-06, + "loss": 0.5431, + "step": 9097 + }, + { + "epoch": 3.1945224719101124, + "grad_norm": 0.9410167336463928, + "learning_rate": 2.267325286467602e-06, + "loss": 0.5834, + "step": 9098 + }, + { + "epoch": 3.194873595505618, + "grad_norm": 0.9246461987495422, + "learning_rate": 2.2668649718684603e-06, + "loss": 0.5667, + "step": 9099 + }, + { + "epoch": 3.1952247191011236, + "grad_norm": 0.9043212532997131, + "learning_rate": 2.266404665242336e-06, + "loss": 0.5233, + "step": 9100 + }, + { + "epoch": 3.195575842696629, + "grad_norm": 0.9357842803001404, + "learning_rate": 2.2659443666049734e-06, + "loss": 0.5439, + "step": 9101 + }, + { + "epoch": 3.195926966292135, + "grad_norm": 0.9282718896865845, + "learning_rate": 2.2654840759721123e-06, + "loss": 0.5386, + "step": 9102 + }, + { + "epoch": 3.1962780898876404, + "grad_norm": 0.983466386795044, + "learning_rate": 2.2650237933594957e-06, + "loss": 0.5717, + "step": 9103 + }, + { + "epoch": 3.196629213483146, + "grad_norm": 0.9434092044830322, + "learning_rate": 2.2645635187828637e-06, + "loss": 0.5535, + "step": 9104 + }, + { + "epoch": 3.1969803370786516, + "grad_norm": 0.925925076007843, + "learning_rate": 2.264103252257959e-06, + "loss": 0.5598, + "step": 9105 + }, + { + "epoch": 3.197331460674157, + "grad_norm": 0.9110205769538879, + "learning_rate": 2.2636429938005196e-06, + "loss": 0.5502, + "step": 9106 + }, + { + "epoch": 3.197682584269663, + "grad_norm": 0.9265942573547363, + "learning_rate": 2.263182743426289e-06, + "loss": 0.5715, + "step": 9107 + }, + { + "epoch": 3.1980337078651684, + "grad_norm": 0.9322956204414368, + "learning_rate": 2.262722501151005e-06, + "loss": 0.5452, + "step": 9108 + }, + { + "epoch": 3.198384831460674, + "grad_norm": 0.9277102947235107, + "learning_rate": 2.2622622669904094e-06, + "loss": 0.5212, + "step": 9109 + }, + { + "epoch": 3.1987359550561796, + "grad_norm": 0.9415985941886902, + "learning_rate": 2.261802040960239e-06, + "loss": 0.5435, + "step": 9110 + }, + { + "epoch": 3.1990870786516856, + "grad_norm": 0.9226630330085754, + "learning_rate": 2.2613418230762372e-06, + "loss": 0.5237, + "step": 9111 + }, + { + "epoch": 3.199438202247191, + "grad_norm": 0.9280032515525818, + "learning_rate": 2.26088161335414e-06, + "loss": 0.5792, + "step": 9112 + }, + { + "epoch": 3.199789325842697, + "grad_norm": 0.9120243787765503, + "learning_rate": 2.2604214118096867e-06, + "loss": 0.555, + "step": 9113 + }, + { + "epoch": 3.2001404494382024, + "grad_norm": 0.9228692650794983, + "learning_rate": 2.2599612184586163e-06, + "loss": 0.5546, + "step": 9114 + }, + { + "epoch": 3.200491573033708, + "grad_norm": 0.9252179265022278, + "learning_rate": 2.2595010333166666e-06, + "loss": 0.5445, + "step": 9115 + }, + { + "epoch": 3.2008426966292136, + "grad_norm": 0.9535467624664307, + "learning_rate": 2.2590408563995762e-06, + "loss": 0.5773, + "step": 9116 + }, + { + "epoch": 3.201193820224719, + "grad_norm": 0.9053303003311157, + "learning_rate": 2.2585806877230813e-06, + "loss": 0.5438, + "step": 9117 + }, + { + "epoch": 3.201544943820225, + "grad_norm": 0.9410801529884338, + "learning_rate": 2.2581205273029215e-06, + "loss": 0.538, + "step": 9118 + }, + { + "epoch": 3.2018960674157304, + "grad_norm": 0.930516242980957, + "learning_rate": 2.2576603751548316e-06, + "loss": 0.5722, + "step": 9119 + }, + { + "epoch": 3.202247191011236, + "grad_norm": 0.9345971345901489, + "learning_rate": 2.257200231294551e-06, + "loss": 0.5492, + "step": 9120 + }, + { + "epoch": 3.2025983146067416, + "grad_norm": 0.902501106262207, + "learning_rate": 2.2567400957378133e-06, + "loss": 0.5581, + "step": 9121 + }, + { + "epoch": 3.202949438202247, + "grad_norm": 0.8879126906394958, + "learning_rate": 2.256279968500357e-06, + "loss": 0.5489, + "step": 9122 + }, + { + "epoch": 3.203300561797753, + "grad_norm": 0.9021519422531128, + "learning_rate": 2.2558198495979163e-06, + "loss": 0.5313, + "step": 9123 + }, + { + "epoch": 3.2036516853932584, + "grad_norm": 0.9352854490280151, + "learning_rate": 2.255359739046229e-06, + "loss": 0.5433, + "step": 9124 + }, + { + "epoch": 3.204002808988764, + "grad_norm": 0.9204985499382019, + "learning_rate": 2.2548996368610295e-06, + "loss": 0.5745, + "step": 9125 + }, + { + "epoch": 3.2043539325842696, + "grad_norm": 0.9595812559127808, + "learning_rate": 2.254439543058051e-06, + "loss": 0.5881, + "step": 9126 + }, + { + "epoch": 3.204705056179775, + "grad_norm": 0.9242666959762573, + "learning_rate": 2.2539794576530318e-06, + "loss": 0.5596, + "step": 9127 + }, + { + "epoch": 3.205056179775281, + "grad_norm": 0.9347757697105408, + "learning_rate": 2.2535193806617027e-06, + "loss": 0.5211, + "step": 9128 + }, + { + "epoch": 3.2054073033707864, + "grad_norm": 0.9560678005218506, + "learning_rate": 2.2530593120998015e-06, + "loss": 0.5694, + "step": 9129 + }, + { + "epoch": 3.205758426966292, + "grad_norm": 1.004502773284912, + "learning_rate": 2.252599251983059e-06, + "loss": 0.5463, + "step": 9130 + }, + { + "epoch": 3.2061095505617976, + "grad_norm": 0.9629998207092285, + "learning_rate": 2.252139200327212e-06, + "loss": 0.5567, + "step": 9131 + }, + { + "epoch": 3.206460674157303, + "grad_norm": 0.922271728515625, + "learning_rate": 2.251679157147991e-06, + "loss": 0.5789, + "step": 9132 + }, + { + "epoch": 3.206811797752809, + "grad_norm": 0.900230348110199, + "learning_rate": 2.2512191224611314e-06, + "loss": 0.5231, + "step": 9133 + }, + { + "epoch": 3.2071629213483144, + "grad_norm": 0.930475115776062, + "learning_rate": 2.2507590962823643e-06, + "loss": 0.5528, + "step": 9134 + }, + { + "epoch": 3.2075140449438204, + "grad_norm": 0.9082934856414795, + "learning_rate": 2.250299078627423e-06, + "loss": 0.5319, + "step": 9135 + }, + { + "epoch": 3.207865168539326, + "grad_norm": 0.9630513191223145, + "learning_rate": 2.2498390695120397e-06, + "loss": 0.5982, + "step": 9136 + }, + { + "epoch": 3.2082162921348316, + "grad_norm": 0.9115469455718994, + "learning_rate": 2.2493790689519455e-06, + "loss": 0.5153, + "step": 9137 + }, + { + "epoch": 3.208567415730337, + "grad_norm": 0.928566575050354, + "learning_rate": 2.2489190769628735e-06, + "loss": 0.5614, + "step": 9138 + }, + { + "epoch": 3.208918539325843, + "grad_norm": 0.9137188196182251, + "learning_rate": 2.248459093560553e-06, + "loss": 0.4988, + "step": 9139 + }, + { + "epoch": 3.2092696629213484, + "grad_norm": 0.9302276372909546, + "learning_rate": 2.247999118760718e-06, + "loss": 0.5613, + "step": 9140 + }, + { + "epoch": 3.209620786516854, + "grad_norm": 0.9396297931671143, + "learning_rate": 2.2475391525790967e-06, + "loss": 0.574, + "step": 9141 + }, + { + "epoch": 3.2099719101123596, + "grad_norm": 0.9348545670509338, + "learning_rate": 2.24707919503142e-06, + "loss": 0.5476, + "step": 9142 + }, + { + "epoch": 3.210323033707865, + "grad_norm": 0.9297940135002136, + "learning_rate": 2.2466192461334186e-06, + "loss": 0.5415, + "step": 9143 + }, + { + "epoch": 3.210674157303371, + "grad_norm": 1.0088021755218506, + "learning_rate": 2.2461593059008224e-06, + "loss": 0.585, + "step": 9144 + }, + { + "epoch": 3.2110252808988764, + "grad_norm": 0.9418768882751465, + "learning_rate": 2.24569937434936e-06, + "loss": 0.5669, + "step": 9145 + }, + { + "epoch": 3.211376404494382, + "grad_norm": 0.9185901284217834, + "learning_rate": 2.245239451494762e-06, + "loss": 0.5374, + "step": 9146 + }, + { + "epoch": 3.2117275280898876, + "grad_norm": 0.9211855530738831, + "learning_rate": 2.2447795373527565e-06, + "loss": 0.5924, + "step": 9147 + }, + { + "epoch": 3.212078651685393, + "grad_norm": 0.8803207278251648, + "learning_rate": 2.2443196319390732e-06, + "loss": 0.4945, + "step": 9148 + }, + { + "epoch": 3.212429775280899, + "grad_norm": 0.9109355807304382, + "learning_rate": 2.243859735269439e-06, + "loss": 0.5594, + "step": 9149 + }, + { + "epoch": 3.2127808988764044, + "grad_norm": 0.9102396368980408, + "learning_rate": 2.243399847359583e-06, + "loss": 0.5462, + "step": 9150 + }, + { + "epoch": 3.21313202247191, + "grad_norm": 0.9318421483039856, + "learning_rate": 2.242939968225232e-06, + "loss": 0.5542, + "step": 9151 + }, + { + "epoch": 3.2134831460674156, + "grad_norm": 0.9272435307502747, + "learning_rate": 2.2424800978821148e-06, + "loss": 0.5928, + "step": 9152 + }, + { + "epoch": 3.213834269662921, + "grad_norm": 0.9304667115211487, + "learning_rate": 2.242020236345957e-06, + "loss": 0.5396, + "step": 9153 + }, + { + "epoch": 3.214185393258427, + "grad_norm": 0.9411357641220093, + "learning_rate": 2.2415603836324874e-06, + "loss": 0.6012, + "step": 9154 + }, + { + "epoch": 3.2145365168539324, + "grad_norm": 0.9300580620765686, + "learning_rate": 2.2411005397574317e-06, + "loss": 0.5839, + "step": 9155 + }, + { + "epoch": 3.2148876404494384, + "grad_norm": 0.928591251373291, + "learning_rate": 2.2406407047365142e-06, + "loss": 0.5665, + "step": 9156 + }, + { + "epoch": 3.215238764044944, + "grad_norm": 0.9337812066078186, + "learning_rate": 2.2401808785854646e-06, + "loss": 0.5339, + "step": 9157 + }, + { + "epoch": 3.2155898876404496, + "grad_norm": 0.95014488697052, + "learning_rate": 2.2397210613200052e-06, + "loss": 0.5512, + "step": 9158 + }, + { + "epoch": 3.215941011235955, + "grad_norm": 0.96954745054245, + "learning_rate": 2.239261252955864e-06, + "loss": 0.5352, + "step": 9159 + }, + { + "epoch": 3.216292134831461, + "grad_norm": 0.9551593661308289, + "learning_rate": 2.2388014535087637e-06, + "loss": 0.5836, + "step": 9160 + }, + { + "epoch": 3.2166432584269664, + "grad_norm": 0.9785509705543518, + "learning_rate": 2.2383416629944315e-06, + "loss": 0.5802, + "step": 9161 + }, + { + "epoch": 3.216994382022472, + "grad_norm": 0.9426549673080444, + "learning_rate": 2.23788188142859e-06, + "loss": 0.562, + "step": 9162 + }, + { + "epoch": 3.2173455056179776, + "grad_norm": 0.913628101348877, + "learning_rate": 2.237422108826963e-06, + "loss": 0.5488, + "step": 9163 + }, + { + "epoch": 3.217696629213483, + "grad_norm": 0.9752023220062256, + "learning_rate": 2.2369623452052763e-06, + "loss": 0.5741, + "step": 9164 + }, + { + "epoch": 3.218047752808989, + "grad_norm": 0.9644680023193359, + "learning_rate": 2.2365025905792515e-06, + "loss": 0.558, + "step": 9165 + }, + { + "epoch": 3.2183988764044944, + "grad_norm": 0.930069625377655, + "learning_rate": 2.2360428449646132e-06, + "loss": 0.5762, + "step": 9166 + }, + { + "epoch": 3.21875, + "grad_norm": 0.9585107564926147, + "learning_rate": 2.235583108377083e-06, + "loss": 0.5737, + "step": 9167 + }, + { + "epoch": 3.2191011235955056, + "grad_norm": 0.921271026134491, + "learning_rate": 2.2351233808323853e-06, + "loss": 0.5498, + "step": 9168 + }, + { + "epoch": 3.219452247191011, + "grad_norm": 0.9218440055847168, + "learning_rate": 2.23466366234624e-06, + "loss": 0.5807, + "step": 9169 + }, + { + "epoch": 3.219803370786517, + "grad_norm": 0.9136621952056885, + "learning_rate": 2.234203952934372e-06, + "loss": 0.5531, + "step": 9170 + }, + { + "epoch": 3.2201544943820224, + "grad_norm": 0.908959686756134, + "learning_rate": 2.2337442526124996e-06, + "loss": 0.5771, + "step": 9171 + }, + { + "epoch": 3.220505617977528, + "grad_norm": 0.9029327630996704, + "learning_rate": 2.2332845613963475e-06, + "loss": 0.5666, + "step": 9172 + }, + { + "epoch": 3.2208567415730336, + "grad_norm": 0.9246013760566711, + "learning_rate": 2.2328248793016337e-06, + "loss": 0.5401, + "step": 9173 + }, + { + "epoch": 3.221207865168539, + "grad_norm": 0.9429555535316467, + "learning_rate": 2.232365206344082e-06, + "loss": 0.5738, + "step": 9174 + }, + { + "epoch": 3.221558988764045, + "grad_norm": 0.9304042458534241, + "learning_rate": 2.2319055425394107e-06, + "loss": 0.5956, + "step": 9175 + }, + { + "epoch": 3.2219101123595504, + "grad_norm": 0.903529942035675, + "learning_rate": 2.2314458879033395e-06, + "loss": 0.5371, + "step": 9176 + }, + { + "epoch": 3.222261235955056, + "grad_norm": 0.9709274768829346, + "learning_rate": 2.2309862424515903e-06, + "loss": 0.5798, + "step": 9177 + }, + { + "epoch": 3.2226123595505616, + "grad_norm": 0.9517124891281128, + "learning_rate": 2.2305266061998805e-06, + "loss": 0.5437, + "step": 9178 + }, + { + "epoch": 3.2229634831460676, + "grad_norm": 0.9191957712173462, + "learning_rate": 2.2300669791639307e-06, + "loss": 0.5572, + "step": 9179 + }, + { + "epoch": 3.223314606741573, + "grad_norm": 0.9220535159111023, + "learning_rate": 2.229607361359459e-06, + "loss": 0.5526, + "step": 9180 + }, + { + "epoch": 3.223665730337079, + "grad_norm": 0.9546854496002197, + "learning_rate": 2.2291477528021842e-06, + "loss": 0.6013, + "step": 9181 + }, + { + "epoch": 3.2240168539325844, + "grad_norm": 0.9491085410118103, + "learning_rate": 2.2286881535078243e-06, + "loss": 0.5574, + "step": 9182 + }, + { + "epoch": 3.22436797752809, + "grad_norm": 0.9600939750671387, + "learning_rate": 2.228228563492098e-06, + "loss": 0.5556, + "step": 9183 + }, + { + "epoch": 3.2247191011235956, + "grad_norm": 0.946183443069458, + "learning_rate": 2.2277689827707212e-06, + "loss": 0.5733, + "step": 9184 + }, + { + "epoch": 3.225070224719101, + "grad_norm": 0.934611976146698, + "learning_rate": 2.2273094113594136e-06, + "loss": 0.5603, + "step": 9185 + }, + { + "epoch": 3.225421348314607, + "grad_norm": 0.9424348473548889, + "learning_rate": 2.2268498492738893e-06, + "loss": 0.5822, + "step": 9186 + }, + { + "epoch": 3.2257724719101124, + "grad_norm": 0.9237279295921326, + "learning_rate": 2.2263902965298677e-06, + "loss": 0.5236, + "step": 9187 + }, + { + "epoch": 3.226123595505618, + "grad_norm": 0.9089604616165161, + "learning_rate": 2.225930753143064e-06, + "loss": 0.5395, + "step": 9188 + }, + { + "epoch": 3.2264747191011236, + "grad_norm": 0.9223783016204834, + "learning_rate": 2.225471219129193e-06, + "loss": 0.5484, + "step": 9189 + }, + { + "epoch": 3.226825842696629, + "grad_norm": 0.9446120858192444, + "learning_rate": 2.225011694503972e-06, + "loss": 0.5627, + "step": 9190 + }, + { + "epoch": 3.227176966292135, + "grad_norm": 0.9273456335067749, + "learning_rate": 2.224552179283115e-06, + "loss": 0.5518, + "step": 9191 + }, + { + "epoch": 3.2275280898876404, + "grad_norm": 0.9152151942253113, + "learning_rate": 2.2240926734823387e-06, + "loss": 0.5606, + "step": 9192 + }, + { + "epoch": 3.227879213483146, + "grad_norm": 0.9733384847640991, + "learning_rate": 2.2236331771173557e-06, + "loss": 0.5603, + "step": 9193 + }, + { + "epoch": 3.2282303370786516, + "grad_norm": 0.9348355531692505, + "learning_rate": 2.223173690203883e-06, + "loss": 0.5402, + "step": 9194 + }, + { + "epoch": 3.228581460674157, + "grad_norm": 0.9234939813613892, + "learning_rate": 2.2227142127576322e-06, + "loss": 0.5505, + "step": 9195 + }, + { + "epoch": 3.228932584269663, + "grad_norm": 0.9725620746612549, + "learning_rate": 2.2222547447943183e-06, + "loss": 0.5786, + "step": 9196 + }, + { + "epoch": 3.2292837078651684, + "grad_norm": 0.9163536429405212, + "learning_rate": 2.2217952863296543e-06, + "loss": 0.557, + "step": 9197 + }, + { + "epoch": 3.229634831460674, + "grad_norm": 0.9097366333007812, + "learning_rate": 2.2213358373793544e-06, + "loss": 0.5485, + "step": 9198 + }, + { + "epoch": 3.2299859550561796, + "grad_norm": 0.9418055415153503, + "learning_rate": 2.22087639795913e-06, + "loss": 0.5647, + "step": 9199 + }, + { + "epoch": 3.2303370786516856, + "grad_norm": 0.9442188739776611, + "learning_rate": 2.220416968084693e-06, + "loss": 0.5464, + "step": 9200 + }, + { + "epoch": 3.230688202247191, + "grad_norm": 1.0534002780914307, + "learning_rate": 2.2199575477717573e-06, + "loss": 0.6067, + "step": 9201 + }, + { + "epoch": 3.231039325842697, + "grad_norm": 0.9445778131484985, + "learning_rate": 2.2194981370360332e-06, + "loss": 0.5469, + "step": 9202 + }, + { + "epoch": 3.2313904494382024, + "grad_norm": 0.9010031223297119, + "learning_rate": 2.219038735893234e-06, + "loss": 0.5481, + "step": 9203 + }, + { + "epoch": 3.231741573033708, + "grad_norm": 0.9348849058151245, + "learning_rate": 2.2185793443590677e-06, + "loss": 0.5556, + "step": 9204 + }, + { + "epoch": 3.2320926966292136, + "grad_norm": 0.9249671101570129, + "learning_rate": 2.2181199624492487e-06, + "loss": 0.5491, + "step": 9205 + }, + { + "epoch": 3.232443820224719, + "grad_norm": 0.8897696733474731, + "learning_rate": 2.2176605901794845e-06, + "loss": 0.5325, + "step": 9206 + }, + { + "epoch": 3.232794943820225, + "grad_norm": 0.9049200415611267, + "learning_rate": 2.2172012275654877e-06, + "loss": 0.573, + "step": 9207 + }, + { + "epoch": 3.2331460674157304, + "grad_norm": 0.9447661638259888, + "learning_rate": 2.2167418746229656e-06, + "loss": 0.5644, + "step": 9208 + }, + { + "epoch": 3.233497191011236, + "grad_norm": 0.9540297389030457, + "learning_rate": 2.21628253136763e-06, + "loss": 0.5775, + "step": 9209 + }, + { + "epoch": 3.2338483146067416, + "grad_norm": 0.9026526808738708, + "learning_rate": 2.2158231978151885e-06, + "loss": 0.5652, + "step": 9210 + }, + { + "epoch": 3.234199438202247, + "grad_norm": 0.9415181875228882, + "learning_rate": 2.2153638739813518e-06, + "loss": 0.5601, + "step": 9211 + }, + { + "epoch": 3.234550561797753, + "grad_norm": 0.9067089557647705, + "learning_rate": 2.2149045598818264e-06, + "loss": 0.5699, + "step": 9212 + }, + { + "epoch": 3.2349016853932584, + "grad_norm": 0.9328421354293823, + "learning_rate": 2.2144452555323203e-06, + "loss": 0.5726, + "step": 9213 + }, + { + "epoch": 3.235252808988764, + "grad_norm": 0.9483035802841187, + "learning_rate": 2.213985960948543e-06, + "loss": 0.5896, + "step": 9214 + }, + { + "epoch": 3.2356039325842696, + "grad_norm": 0.910982072353363, + "learning_rate": 2.2135266761462e-06, + "loss": 0.553, + "step": 9215 + }, + { + "epoch": 3.235955056179775, + "grad_norm": 0.9213133454322815, + "learning_rate": 2.2130674011410008e-06, + "loss": 0.5385, + "step": 9216 + }, + { + "epoch": 3.236306179775281, + "grad_norm": 0.9438744783401489, + "learning_rate": 2.212608135948649e-06, + "loss": 0.5531, + "step": 9217 + }, + { + "epoch": 3.2366573033707864, + "grad_norm": 0.9228683710098267, + "learning_rate": 2.2121488805848553e-06, + "loss": 0.5186, + "step": 9218 + }, + { + "epoch": 3.237008426966292, + "grad_norm": 0.9168832302093506, + "learning_rate": 2.211689635065322e-06, + "loss": 0.5445, + "step": 9219 + }, + { + "epoch": 3.2373595505617976, + "grad_norm": 0.9342902898788452, + "learning_rate": 2.2112303994057575e-06, + "loss": 0.5656, + "step": 9220 + }, + { + "epoch": 3.237710674157303, + "grad_norm": 0.9299089312553406, + "learning_rate": 2.2107711736218658e-06, + "loss": 0.5333, + "step": 9221 + }, + { + "epoch": 3.238061797752809, + "grad_norm": 0.9060919880867004, + "learning_rate": 2.2103119577293524e-06, + "loss": 0.5708, + "step": 9222 + }, + { + "epoch": 3.2384129213483144, + "grad_norm": 0.951111912727356, + "learning_rate": 2.2098527517439224e-06, + "loss": 0.5659, + "step": 9223 + }, + { + "epoch": 3.2387640449438204, + "grad_norm": 0.9648430347442627, + "learning_rate": 2.2093935556812796e-06, + "loss": 0.5764, + "step": 9224 + }, + { + "epoch": 3.239115168539326, + "grad_norm": 0.9135909080505371, + "learning_rate": 2.2089343695571293e-06, + "loss": 0.556, + "step": 9225 + }, + { + "epoch": 3.2394662921348316, + "grad_norm": 0.9376388192176819, + "learning_rate": 2.2084751933871735e-06, + "loss": 0.5681, + "step": 9226 + }, + { + "epoch": 3.239817415730337, + "grad_norm": 0.9354874491691589, + "learning_rate": 2.2080160271871177e-06, + "loss": 0.5773, + "step": 9227 + }, + { + "epoch": 3.240168539325843, + "grad_norm": 0.9520635008811951, + "learning_rate": 2.207556870972663e-06, + "loss": 0.5444, + "step": 9228 + }, + { + "epoch": 3.2405196629213484, + "grad_norm": 0.9636841416358948, + "learning_rate": 2.207097724759513e-06, + "loss": 0.5643, + "step": 9229 + }, + { + "epoch": 3.240870786516854, + "grad_norm": 0.981449544429779, + "learning_rate": 2.206638588563371e-06, + "loss": 0.5832, + "step": 9230 + }, + { + "epoch": 3.2412219101123596, + "grad_norm": 0.9497424364089966, + "learning_rate": 2.2061794623999376e-06, + "loss": 0.5339, + "step": 9231 + }, + { + "epoch": 3.241573033707865, + "grad_norm": 0.9614279866218567, + "learning_rate": 2.2057203462849154e-06, + "loss": 0.554, + "step": 9232 + }, + { + "epoch": 3.241924157303371, + "grad_norm": 0.9220987558364868, + "learning_rate": 2.2052612402340054e-06, + "loss": 0.5229, + "step": 9233 + }, + { + "epoch": 3.2422752808988764, + "grad_norm": 0.9503759145736694, + "learning_rate": 2.2048021442629086e-06, + "loss": 0.5663, + "step": 9234 + }, + { + "epoch": 3.242626404494382, + "grad_norm": 0.9253359436988831, + "learning_rate": 2.204343058387327e-06, + "loss": 0.5463, + "step": 9235 + }, + { + "epoch": 3.2429775280898876, + "grad_norm": 0.9331046342849731, + "learning_rate": 2.203883982622958e-06, + "loss": 0.5617, + "step": 9236 + }, + { + "epoch": 3.243328651685393, + "grad_norm": 0.9096701741218567, + "learning_rate": 2.2034249169855047e-06, + "loss": 0.5563, + "step": 9237 + }, + { + "epoch": 3.243679775280899, + "grad_norm": 0.9220385551452637, + "learning_rate": 2.202965861490665e-06, + "loss": 0.5658, + "step": 9238 + }, + { + "epoch": 3.2440308988764044, + "grad_norm": 0.9419352412223816, + "learning_rate": 2.202506816154139e-06, + "loss": 0.5556, + "step": 9239 + }, + { + "epoch": 3.24438202247191, + "grad_norm": 0.9261015057563782, + "learning_rate": 2.202047780991625e-06, + "loss": 0.5407, + "step": 9240 + }, + { + "epoch": 3.2447331460674156, + "grad_norm": 0.9412524104118347, + "learning_rate": 2.2015887560188216e-06, + "loss": 0.5798, + "step": 9241 + }, + { + "epoch": 3.245084269662921, + "grad_norm": 0.9226146340370178, + "learning_rate": 2.2011297412514284e-06, + "loss": 0.5619, + "step": 9242 + }, + { + "epoch": 3.245435393258427, + "grad_norm": 0.954803466796875, + "learning_rate": 2.2006707367051415e-06, + "loss": 0.5813, + "step": 9243 + }, + { + "epoch": 3.2457865168539324, + "grad_norm": 0.9200631976127625, + "learning_rate": 2.20021174239566e-06, + "loss": 0.5508, + "step": 9244 + }, + { + "epoch": 3.2461376404494384, + "grad_norm": 0.9291655421257019, + "learning_rate": 2.199752758338679e-06, + "loss": 0.5738, + "step": 9245 + }, + { + "epoch": 3.246488764044944, + "grad_norm": 0.9119186401367188, + "learning_rate": 2.1992937845498978e-06, + "loss": 0.5672, + "step": 9246 + }, + { + "epoch": 3.2468398876404496, + "grad_norm": 0.9075996279716492, + "learning_rate": 2.198834821045011e-06, + "loss": 0.57, + "step": 9247 + }, + { + "epoch": 3.247191011235955, + "grad_norm": 0.9333869218826294, + "learning_rate": 2.198375867839717e-06, + "loss": 0.569, + "step": 9248 + }, + { + "epoch": 3.247542134831461, + "grad_norm": 0.9820589423179626, + "learning_rate": 2.1979169249497087e-06, + "loss": 0.591, + "step": 9249 + }, + { + "epoch": 3.2478932584269664, + "grad_norm": 0.9149643182754517, + "learning_rate": 2.1974579923906844e-06, + "loss": 0.5492, + "step": 9250 + }, + { + "epoch": 3.248244382022472, + "grad_norm": 0.9473191499710083, + "learning_rate": 2.196999070178338e-06, + "loss": 0.5727, + "step": 9251 + }, + { + "epoch": 3.2485955056179776, + "grad_norm": 0.9668101072311401, + "learning_rate": 2.1965401583283627e-06, + "loss": 0.558, + "step": 9252 + }, + { + "epoch": 3.248946629213483, + "grad_norm": 0.9074317812919617, + "learning_rate": 2.1960812568564556e-06, + "loss": 0.5801, + "step": 9253 + }, + { + "epoch": 3.249297752808989, + "grad_norm": 0.9349655508995056, + "learning_rate": 2.195622365778308e-06, + "loss": 0.5825, + "step": 9254 + }, + { + "epoch": 3.2496488764044944, + "grad_norm": 0.9493439793586731, + "learning_rate": 2.1951634851096164e-06, + "loss": 0.5352, + "step": 9255 + }, + { + "epoch": 3.25, + "grad_norm": 0.9797566533088684, + "learning_rate": 2.1947046148660716e-06, + "loss": 0.5992, + "step": 9256 + }, + { + "epoch": 3.2503511235955056, + "grad_norm": 0.9674732089042664, + "learning_rate": 2.194245755063369e-06, + "loss": 0.5744, + "step": 9257 + }, + { + "epoch": 3.250702247191011, + "grad_norm": 0.9335240125656128, + "learning_rate": 2.1937869057171988e-06, + "loss": 0.5426, + "step": 9258 + }, + { + "epoch": 3.251053370786517, + "grad_norm": 0.9316025972366333, + "learning_rate": 2.1933280668432554e-06, + "loss": 0.5239, + "step": 9259 + }, + { + "epoch": 3.2514044943820224, + "grad_norm": 0.9162511229515076, + "learning_rate": 2.1928692384572286e-06, + "loss": 0.5544, + "step": 9260 + }, + { + "epoch": 3.251755617977528, + "grad_norm": 0.9300464987754822, + "learning_rate": 2.192410420574812e-06, + "loss": 0.5669, + "step": 9261 + }, + { + "epoch": 3.2521067415730336, + "grad_norm": 0.9539361000061035, + "learning_rate": 2.191951613211696e-06, + "loss": 0.5449, + "step": 9262 + }, + { + "epoch": 3.252457865168539, + "grad_norm": 0.9239545464515686, + "learning_rate": 2.19149281638357e-06, + "loss": 0.5656, + "step": 9263 + }, + { + "epoch": 3.252808988764045, + "grad_norm": 0.9301228523254395, + "learning_rate": 2.191034030106127e-06, + "loss": 0.5394, + "step": 9264 + }, + { + "epoch": 3.2531601123595504, + "grad_norm": 0.9327834844589233, + "learning_rate": 2.1905752543950545e-06, + "loss": 0.5937, + "step": 9265 + }, + { + "epoch": 3.253511235955056, + "grad_norm": 0.9334339499473572, + "learning_rate": 2.1901164892660446e-06, + "loss": 0.5371, + "step": 9266 + }, + { + "epoch": 3.2538623595505616, + "grad_norm": 0.9245770573616028, + "learning_rate": 2.1896577347347843e-06, + "loss": 0.5527, + "step": 9267 + }, + { + "epoch": 3.254213483146067, + "grad_norm": 0.9583907127380371, + "learning_rate": 2.1891989908169655e-06, + "loss": 0.5483, + "step": 9268 + }, + { + "epoch": 3.254564606741573, + "grad_norm": 1.0152393579483032, + "learning_rate": 2.188740257528274e-06, + "loss": 0.5419, + "step": 9269 + }, + { + "epoch": 3.254915730337079, + "grad_norm": 0.948883056640625, + "learning_rate": 2.1882815348844002e-06, + "loss": 0.5445, + "step": 9270 + }, + { + "epoch": 3.2552668539325844, + "grad_norm": 0.9664182662963867, + "learning_rate": 2.1878228229010305e-06, + "loss": 0.5529, + "step": 9271 + }, + { + "epoch": 3.25561797752809, + "grad_norm": 0.9603793621063232, + "learning_rate": 2.1873641215938537e-06, + "loss": 0.579, + "step": 9272 + }, + { + "epoch": 3.2559691011235956, + "grad_norm": 0.946603000164032, + "learning_rate": 2.1869054309785558e-06, + "loss": 0.5649, + "step": 9273 + }, + { + "epoch": 3.256320224719101, + "grad_norm": 0.9449673891067505, + "learning_rate": 2.186446751070825e-06, + "loss": 0.5817, + "step": 9274 + }, + { + "epoch": 3.256671348314607, + "grad_norm": 0.9449647068977356, + "learning_rate": 2.1859880818863473e-06, + "loss": 0.595, + "step": 9275 + }, + { + "epoch": 3.2570224719101124, + "grad_norm": 0.8935825824737549, + "learning_rate": 2.185529423440807e-06, + "loss": 0.5583, + "step": 9276 + }, + { + "epoch": 3.257373595505618, + "grad_norm": 0.9158092737197876, + "learning_rate": 2.1850707757498927e-06, + "loss": 0.5463, + "step": 9277 + }, + { + "epoch": 3.2577247191011236, + "grad_norm": 0.9467341899871826, + "learning_rate": 2.184612138829287e-06, + "loss": 0.5655, + "step": 9278 + }, + { + "epoch": 3.258075842696629, + "grad_norm": 0.9283179044723511, + "learning_rate": 2.1841535126946777e-06, + "loss": 0.6165, + "step": 9279 + }, + { + "epoch": 3.258426966292135, + "grad_norm": 0.8861303925514221, + "learning_rate": 2.1836948973617466e-06, + "loss": 0.495, + "step": 9280 + }, + { + "epoch": 3.2587780898876404, + "grad_norm": 0.9263895750045776, + "learning_rate": 2.183236292846181e-06, + "loss": 0.555, + "step": 9281 + }, + { + "epoch": 3.259129213483146, + "grad_norm": 0.9167452454566956, + "learning_rate": 2.1827776991636614e-06, + "loss": 0.5505, + "step": 9282 + }, + { + "epoch": 3.2594803370786516, + "grad_norm": 0.9315223693847656, + "learning_rate": 2.182319116329874e-06, + "loss": 0.5904, + "step": 9283 + }, + { + "epoch": 3.259831460674157, + "grad_norm": 0.9422589540481567, + "learning_rate": 2.1818605443605004e-06, + "loss": 0.5887, + "step": 9284 + }, + { + "epoch": 3.260182584269663, + "grad_norm": 0.9583097100257874, + "learning_rate": 2.181401983271225e-06, + "loss": 0.5617, + "step": 9285 + }, + { + "epoch": 3.2605337078651684, + "grad_norm": 0.9501683712005615, + "learning_rate": 2.180943433077728e-06, + "loss": 0.5492, + "step": 9286 + }, + { + "epoch": 3.260884831460674, + "grad_norm": 0.9554504752159119, + "learning_rate": 2.1804848937956937e-06, + "loss": 0.5734, + "step": 9287 + }, + { + "epoch": 3.26123595505618, + "grad_norm": 0.9478028416633606, + "learning_rate": 2.180026365440803e-06, + "loss": 0.5646, + "step": 9288 + }, + { + "epoch": 3.2615870786516856, + "grad_norm": 0.9106342196464539, + "learning_rate": 2.179567848028735e-06, + "loss": 0.557, + "step": 9289 + }, + { + "epoch": 3.261938202247191, + "grad_norm": 0.9394769072532654, + "learning_rate": 2.1791093415751734e-06, + "loss": 0.5678, + "step": 9290 + }, + { + "epoch": 3.262289325842697, + "grad_norm": 0.9662386775016785, + "learning_rate": 2.178650846095797e-06, + "loss": 0.602, + "step": 9291 + }, + { + "epoch": 3.2626404494382024, + "grad_norm": 0.9246668219566345, + "learning_rate": 2.178192361606288e-06, + "loss": 0.5299, + "step": 9292 + }, + { + "epoch": 3.262991573033708, + "grad_norm": 0.9317671060562134, + "learning_rate": 2.1777338881223234e-06, + "loss": 0.5683, + "step": 9293 + }, + { + "epoch": 3.2633426966292136, + "grad_norm": 0.952355682849884, + "learning_rate": 2.177275425659585e-06, + "loss": 0.5327, + "step": 9294 + }, + { + "epoch": 3.263693820224719, + "grad_norm": 0.8911362290382385, + "learning_rate": 2.17681697423375e-06, + "loss": 0.5395, + "step": 9295 + }, + { + "epoch": 3.264044943820225, + "grad_norm": 0.9451972246170044, + "learning_rate": 2.1763585338604994e-06, + "loss": 0.587, + "step": 9296 + }, + { + "epoch": 3.2643960674157304, + "grad_norm": 0.8921921849250793, + "learning_rate": 2.1759001045555085e-06, + "loss": 0.5451, + "step": 9297 + }, + { + "epoch": 3.264747191011236, + "grad_norm": 0.94584059715271, + "learning_rate": 2.175441686334458e-06, + "loss": 0.5611, + "step": 9298 + }, + { + "epoch": 3.2650983146067416, + "grad_norm": 0.9641897678375244, + "learning_rate": 2.1749832792130232e-06, + "loss": 0.5878, + "step": 9299 + }, + { + "epoch": 3.265449438202247, + "grad_norm": 0.9302607774734497, + "learning_rate": 2.1745248832068826e-06, + "loss": 0.5946, + "step": 9300 + }, + { + "epoch": 3.265800561797753, + "grad_norm": 0.9422127604484558, + "learning_rate": 2.1740664983317124e-06, + "loss": 0.579, + "step": 9301 + }, + { + "epoch": 3.2661516853932584, + "grad_norm": 0.9462207555770874, + "learning_rate": 2.1736081246031883e-06, + "loss": 0.58, + "step": 9302 + }, + { + "epoch": 3.266502808988764, + "grad_norm": 0.9865782260894775, + "learning_rate": 2.1731497620369885e-06, + "loss": 0.6011, + "step": 9303 + }, + { + "epoch": 3.2668539325842696, + "grad_norm": 0.9176053404808044, + "learning_rate": 2.1726914106487857e-06, + "loss": 0.5645, + "step": 9304 + }, + { + "epoch": 3.267205056179775, + "grad_norm": 0.9741256237030029, + "learning_rate": 2.172233070454258e-06, + "loss": 0.582, + "step": 9305 + }, + { + "epoch": 3.267556179775281, + "grad_norm": 0.9165406823158264, + "learning_rate": 2.171774741469077e-06, + "loss": 0.5798, + "step": 9306 + }, + { + "epoch": 3.2679073033707864, + "grad_norm": 0.941644549369812, + "learning_rate": 2.1713164237089202e-06, + "loss": 0.5645, + "step": 9307 + }, + { + "epoch": 3.268258426966292, + "grad_norm": 1.003472089767456, + "learning_rate": 2.1708581171894604e-06, + "loss": 0.593, + "step": 9308 + }, + { + "epoch": 3.2686095505617976, + "grad_norm": 0.9404183030128479, + "learning_rate": 2.1703998219263712e-06, + "loss": 0.5746, + "step": 9309 + }, + { + "epoch": 3.268960674157303, + "grad_norm": 0.9423112273216248, + "learning_rate": 2.169941537935326e-06, + "loss": 0.5563, + "step": 9310 + }, + { + "epoch": 3.269311797752809, + "grad_norm": 0.9333468079566956, + "learning_rate": 2.1694832652319976e-06, + "loss": 0.5417, + "step": 9311 + }, + { + "epoch": 3.2696629213483144, + "grad_norm": 0.9410494565963745, + "learning_rate": 2.1690250038320587e-06, + "loss": 0.585, + "step": 9312 + }, + { + "epoch": 3.27001404494382, + "grad_norm": 0.952239990234375, + "learning_rate": 2.1685667537511817e-06, + "loss": 0.5536, + "step": 9313 + }, + { + "epoch": 3.270365168539326, + "grad_norm": 0.8911036849021912, + "learning_rate": 2.1681085150050387e-06, + "loss": 0.5257, + "step": 9314 + }, + { + "epoch": 3.2707162921348316, + "grad_norm": 0.9210630059242249, + "learning_rate": 2.1676502876093e-06, + "loss": 0.5584, + "step": 9315 + }, + { + "epoch": 3.271067415730337, + "grad_norm": 0.9207329750061035, + "learning_rate": 2.1671920715796364e-06, + "loss": 0.5831, + "step": 9316 + }, + { + "epoch": 3.271418539325843, + "grad_norm": 0.937929630279541, + "learning_rate": 2.1667338669317195e-06, + "loss": 0.5455, + "step": 9317 + }, + { + "epoch": 3.2717696629213484, + "grad_norm": 0.9272195100784302, + "learning_rate": 2.1662756736812195e-06, + "loss": 0.5882, + "step": 9318 + }, + { + "epoch": 3.272120786516854, + "grad_norm": 0.9555923938751221, + "learning_rate": 2.165817491843806e-06, + "loss": 0.5567, + "step": 9319 + }, + { + "epoch": 3.2724719101123596, + "grad_norm": 0.9813551306724548, + "learning_rate": 2.1653593214351477e-06, + "loss": 0.5704, + "step": 9320 + }, + { + "epoch": 3.272823033707865, + "grad_norm": 0.9359728097915649, + "learning_rate": 2.1649011624709143e-06, + "loss": 0.5487, + "step": 9321 + }, + { + "epoch": 3.273174157303371, + "grad_norm": 1.1197147369384766, + "learning_rate": 2.1644430149667754e-06, + "loss": 0.5967, + "step": 9322 + }, + { + "epoch": 3.2735252808988764, + "grad_norm": 0.9386174082756042, + "learning_rate": 2.163984878938397e-06, + "loss": 0.5687, + "step": 9323 + }, + { + "epoch": 3.273876404494382, + "grad_norm": 0.9326093196868896, + "learning_rate": 2.163526754401449e-06, + "loss": 0.5774, + "step": 9324 + }, + { + "epoch": 3.2742275280898876, + "grad_norm": 0.9266312122344971, + "learning_rate": 2.1630686413715977e-06, + "loss": 0.5797, + "step": 9325 + }, + { + "epoch": 3.274578651685393, + "grad_norm": 0.9191955924034119, + "learning_rate": 2.1626105398645107e-06, + "loss": 0.5264, + "step": 9326 + }, + { + "epoch": 3.274929775280899, + "grad_norm": 0.9152282476425171, + "learning_rate": 2.1621524498958547e-06, + "loss": 0.5149, + "step": 9327 + }, + { + "epoch": 3.2752808988764044, + "grad_norm": 0.9302106499671936, + "learning_rate": 2.1616943714812956e-06, + "loss": 0.5573, + "step": 9328 + }, + { + "epoch": 3.27563202247191, + "grad_norm": 0.9237815737724304, + "learning_rate": 2.1612363046364997e-06, + "loss": 0.5694, + "step": 9329 + }, + { + "epoch": 3.2759831460674156, + "grad_norm": 0.9177208542823792, + "learning_rate": 2.1607782493771316e-06, + "loss": 0.5211, + "step": 9330 + }, + { + "epoch": 3.276334269662921, + "grad_norm": 0.938884437084198, + "learning_rate": 2.160320205718858e-06, + "loss": 0.5713, + "step": 9331 + }, + { + "epoch": 3.276685393258427, + "grad_norm": 0.9307376742362976, + "learning_rate": 2.159862173677342e-06, + "loss": 0.5788, + "step": 9332 + }, + { + "epoch": 3.277036516853933, + "grad_norm": 0.8969783782958984, + "learning_rate": 2.159404153268249e-06, + "loss": 0.532, + "step": 9333 + }, + { + "epoch": 3.2773876404494384, + "grad_norm": 0.8999346494674683, + "learning_rate": 2.1589461445072417e-06, + "loss": 0.5586, + "step": 9334 + }, + { + "epoch": 3.277738764044944, + "grad_norm": 0.9217050075531006, + "learning_rate": 2.158488147409986e-06, + "loss": 0.5821, + "step": 9335 + }, + { + "epoch": 3.2780898876404496, + "grad_norm": 0.9110537767410278, + "learning_rate": 2.158030161992142e-06, + "loss": 0.56, + "step": 9336 + }, + { + "epoch": 3.278441011235955, + "grad_norm": 0.9227237105369568, + "learning_rate": 2.1575721882693745e-06, + "loss": 0.5528, + "step": 9337 + }, + { + "epoch": 3.278792134831461, + "grad_norm": 0.8889098763465881, + "learning_rate": 2.1571142262573458e-06, + "loss": 0.5171, + "step": 9338 + }, + { + "epoch": 3.2791432584269664, + "grad_norm": 0.9552954435348511, + "learning_rate": 2.1566562759717157e-06, + "loss": 0.5489, + "step": 9339 + }, + { + "epoch": 3.279494382022472, + "grad_norm": 0.9327992796897888, + "learning_rate": 2.156198337428148e-06, + "loss": 0.5282, + "step": 9340 + }, + { + "epoch": 3.2798455056179776, + "grad_norm": 0.9664953947067261, + "learning_rate": 2.1557404106423023e-06, + "loss": 0.5736, + "step": 9341 + }, + { + "epoch": 3.280196629213483, + "grad_norm": 0.9561286568641663, + "learning_rate": 2.155282495629841e-06, + "loss": 0.6233, + "step": 9342 + }, + { + "epoch": 3.280547752808989, + "grad_norm": 0.966708779335022, + "learning_rate": 2.1548245924064225e-06, + "loss": 0.5857, + "step": 9343 + }, + { + "epoch": 3.2808988764044944, + "grad_norm": 0.9543550610542297, + "learning_rate": 2.154366700987708e-06, + "loss": 0.5695, + "step": 9344 + }, + { + "epoch": 3.28125, + "grad_norm": 0.973105788230896, + "learning_rate": 2.153908821389356e-06, + "loss": 0.5456, + "step": 9345 + }, + { + "epoch": 3.2816011235955056, + "grad_norm": 0.9364118576049805, + "learning_rate": 2.1534509536270272e-06, + "loss": 0.5682, + "step": 9346 + }, + { + "epoch": 3.281952247191011, + "grad_norm": 0.950748085975647, + "learning_rate": 2.152993097716378e-06, + "loss": 0.5603, + "step": 9347 + }, + { + "epoch": 3.282303370786517, + "grad_norm": 0.9152896404266357, + "learning_rate": 2.152535253673069e-06, + "loss": 0.5491, + "step": 9348 + }, + { + "epoch": 3.2826544943820224, + "grad_norm": 0.9358779191970825, + "learning_rate": 2.152077421512756e-06, + "loss": 0.5603, + "step": 9349 + }, + { + "epoch": 3.283005617977528, + "grad_norm": 0.9314160346984863, + "learning_rate": 2.1516196012510984e-06, + "loss": 0.5815, + "step": 9350 + }, + { + "epoch": 3.2833567415730336, + "grad_norm": 0.8896580338478088, + "learning_rate": 2.1511617929037525e-06, + "loss": 0.5549, + "step": 9351 + }, + { + "epoch": 3.283707865168539, + "grad_norm": 0.9332231283187866, + "learning_rate": 2.1507039964863736e-06, + "loss": 0.5519, + "step": 9352 + }, + { + "epoch": 3.284058988764045, + "grad_norm": 0.9459170699119568, + "learning_rate": 2.1502462120146203e-06, + "loss": 0.569, + "step": 9353 + }, + { + "epoch": 3.2844101123595504, + "grad_norm": 0.9523163437843323, + "learning_rate": 2.1497884395041462e-06, + "loss": 0.5571, + "step": 9354 + }, + { + "epoch": 3.284761235955056, + "grad_norm": 0.9250772595405579, + "learning_rate": 2.1493306789706087e-06, + "loss": 0.5372, + "step": 9355 + }, + { + "epoch": 3.2851123595505616, + "grad_norm": 0.9062320590019226, + "learning_rate": 2.148872930429661e-06, + "loss": 0.5126, + "step": 9356 + }, + { + "epoch": 3.285463483146067, + "grad_norm": 0.9260614514350891, + "learning_rate": 2.1484151938969596e-06, + "loss": 0.5531, + "step": 9357 + }, + { + "epoch": 3.285814606741573, + "grad_norm": 0.923161506652832, + "learning_rate": 2.1479574693881565e-06, + "loss": 0.5476, + "step": 9358 + }, + { + "epoch": 3.286165730337079, + "grad_norm": 0.9407051205635071, + "learning_rate": 2.147499756918908e-06, + "loss": 0.5801, + "step": 9359 + }, + { + "epoch": 3.2865168539325844, + "grad_norm": 0.9298778176307678, + "learning_rate": 2.147042056504865e-06, + "loss": 0.5312, + "step": 9360 + }, + { + "epoch": 3.28686797752809, + "grad_norm": 0.9168205261230469, + "learning_rate": 2.146584368161683e-06, + "loss": 0.5633, + "step": 9361 + }, + { + "epoch": 3.2872191011235956, + "grad_norm": 0.8956932425498962, + "learning_rate": 2.1461266919050115e-06, + "loss": 0.5879, + "step": 9362 + }, + { + "epoch": 3.287570224719101, + "grad_norm": 0.9286038875579834, + "learning_rate": 2.145669027750506e-06, + "loss": 0.5542, + "step": 9363 + }, + { + "epoch": 3.287921348314607, + "grad_norm": 0.9422730803489685, + "learning_rate": 2.1452113757138164e-06, + "loss": 0.5564, + "step": 9364 + }, + { + "epoch": 3.2882724719101124, + "grad_norm": 0.9473357796669006, + "learning_rate": 2.1447537358105925e-06, + "loss": 0.5441, + "step": 9365 + }, + { + "epoch": 3.288623595505618, + "grad_norm": 0.9332162737846375, + "learning_rate": 2.1442961080564884e-06, + "loss": 0.5471, + "step": 9366 + }, + { + "epoch": 3.2889747191011236, + "grad_norm": 0.9310396313667297, + "learning_rate": 2.1438384924671522e-06, + "loss": 0.5638, + "step": 9367 + }, + { + "epoch": 3.289325842696629, + "grad_norm": 0.9379040002822876, + "learning_rate": 2.1433808890582358e-06, + "loss": 0.5559, + "step": 9368 + }, + { + "epoch": 3.289676966292135, + "grad_norm": 0.9278322458267212, + "learning_rate": 2.1429232978453863e-06, + "loss": 0.5765, + "step": 9369 + }, + { + "epoch": 3.2900280898876404, + "grad_norm": 1.0483002662658691, + "learning_rate": 2.1424657188442557e-06, + "loss": 0.5395, + "step": 9370 + }, + { + "epoch": 3.290379213483146, + "grad_norm": 0.9193881154060364, + "learning_rate": 2.142008152070491e-06, + "loss": 0.5117, + "step": 9371 + }, + { + "epoch": 3.2907303370786516, + "grad_norm": 0.9345661997795105, + "learning_rate": 2.1415505975397415e-06, + "loss": 0.5562, + "step": 9372 + }, + { + "epoch": 3.291081460674157, + "grad_norm": 0.9620882272720337, + "learning_rate": 2.141093055267654e-06, + "loss": 0.5757, + "step": 9373 + }, + { + "epoch": 3.291432584269663, + "grad_norm": 0.9476869106292725, + "learning_rate": 2.140635525269878e-06, + "loss": 0.5693, + "step": 9374 + }, + { + "epoch": 3.2917837078651684, + "grad_norm": 0.8918929696083069, + "learning_rate": 2.1401780075620587e-06, + "loss": 0.5279, + "step": 9375 + }, + { + "epoch": 3.292134831460674, + "grad_norm": 0.9949395060539246, + "learning_rate": 2.1397205021598445e-06, + "loss": 0.5776, + "step": 9376 + }, + { + "epoch": 3.29248595505618, + "grad_norm": 0.8984315395355225, + "learning_rate": 2.1392630090788808e-06, + "loss": 0.5671, + "step": 9377 + }, + { + "epoch": 3.2928370786516856, + "grad_norm": 0.9293601512908936, + "learning_rate": 2.1388055283348125e-06, + "loss": 0.5697, + "step": 9378 + }, + { + "epoch": 3.293188202247191, + "grad_norm": 0.9109359979629517, + "learning_rate": 2.138348059943287e-06, + "loss": 0.5507, + "step": 9379 + }, + { + "epoch": 3.293539325842697, + "grad_norm": 0.9336657524108887, + "learning_rate": 2.1378906039199475e-06, + "loss": 0.5997, + "step": 9380 + }, + { + "epoch": 3.2938904494382024, + "grad_norm": 0.9552705883979797, + "learning_rate": 2.1374331602804404e-06, + "loss": 0.5531, + "step": 9381 + }, + { + "epoch": 3.294241573033708, + "grad_norm": 0.960089385509491, + "learning_rate": 2.136975729040408e-06, + "loss": 0.5561, + "step": 9382 + }, + { + "epoch": 3.2945926966292136, + "grad_norm": 0.944839596748352, + "learning_rate": 2.136518310215496e-06, + "loss": 0.5475, + "step": 9383 + }, + { + "epoch": 3.294943820224719, + "grad_norm": 0.9402100443840027, + "learning_rate": 2.1360609038213457e-06, + "loss": 0.561, + "step": 9384 + }, + { + "epoch": 3.295294943820225, + "grad_norm": 0.9437503814697266, + "learning_rate": 2.135603509873602e-06, + "loss": 0.5465, + "step": 9385 + }, + { + "epoch": 3.2956460674157304, + "grad_norm": 0.9648163914680481, + "learning_rate": 2.135146128387906e-06, + "loss": 0.5737, + "step": 9386 + }, + { + "epoch": 3.295997191011236, + "grad_norm": 0.9521219730377197, + "learning_rate": 2.1346887593799003e-06, + "loss": 0.5443, + "step": 9387 + }, + { + "epoch": 3.2963483146067416, + "grad_norm": 0.9033349752426147, + "learning_rate": 2.1342314028652272e-06, + "loss": 0.5523, + "step": 9388 + }, + { + "epoch": 3.296699438202247, + "grad_norm": 0.9372509717941284, + "learning_rate": 2.1337740588595258e-06, + "loss": 0.5621, + "step": 9389 + }, + { + "epoch": 3.297050561797753, + "grad_norm": 0.945377767086029, + "learning_rate": 2.1333167273784393e-06, + "loss": 0.5745, + "step": 9390 + }, + { + "epoch": 3.2974016853932584, + "grad_norm": 1.0431073904037476, + "learning_rate": 2.1328594084376057e-06, + "loss": 0.5613, + "step": 9391 + }, + { + "epoch": 3.297752808988764, + "grad_norm": 0.9378055930137634, + "learning_rate": 2.1324021020526676e-06, + "loss": 0.5103, + "step": 9392 + }, + { + "epoch": 3.2981039325842696, + "grad_norm": 0.9704043865203857, + "learning_rate": 2.1319448082392616e-06, + "loss": 0.5662, + "step": 9393 + }, + { + "epoch": 3.298455056179775, + "grad_norm": 0.9491967558860779, + "learning_rate": 2.131487527013029e-06, + "loss": 0.5828, + "step": 9394 + }, + { + "epoch": 3.298806179775281, + "grad_norm": 0.9302085638046265, + "learning_rate": 2.131030258389608e-06, + "loss": 0.5836, + "step": 9395 + }, + { + "epoch": 3.2991573033707864, + "grad_norm": 0.9602435827255249, + "learning_rate": 2.1305730023846356e-06, + "loss": 0.5572, + "step": 9396 + }, + { + "epoch": 3.299508426966292, + "grad_norm": 0.9119904041290283, + "learning_rate": 2.130115759013751e-06, + "loss": 0.5451, + "step": 9397 + }, + { + "epoch": 3.2998595505617976, + "grad_norm": 0.9169926643371582, + "learning_rate": 2.1296585282925906e-06, + "loss": 0.5129, + "step": 9398 + }, + { + "epoch": 3.300210674157303, + "grad_norm": 0.9192485213279724, + "learning_rate": 2.1292013102367916e-06, + "loss": 0.5301, + "step": 9399 + }, + { + "epoch": 3.300561797752809, + "grad_norm": 0.9589872360229492, + "learning_rate": 2.1287441048619906e-06, + "loss": 0.5681, + "step": 9400 + }, + { + "epoch": 3.3009129213483144, + "grad_norm": 0.9310960173606873, + "learning_rate": 2.1282869121838243e-06, + "loss": 0.5287, + "step": 9401 + }, + { + "epoch": 3.30126404494382, + "grad_norm": 0.9305381178855896, + "learning_rate": 2.127829732217927e-06, + "loss": 0.5635, + "step": 9402 + }, + { + "epoch": 3.301615168539326, + "grad_norm": 0.9106603264808655, + "learning_rate": 2.127372564979934e-06, + "loss": 0.535, + "step": 9403 + }, + { + "epoch": 3.3019662921348316, + "grad_norm": 0.9346327781677246, + "learning_rate": 2.1269154104854804e-06, + "loss": 0.5707, + "step": 9404 + }, + { + "epoch": 3.302317415730337, + "grad_norm": 0.9191266894340515, + "learning_rate": 2.1264582687502007e-06, + "loss": 0.5422, + "step": 9405 + }, + { + "epoch": 3.302668539325843, + "grad_norm": 0.9292037487030029, + "learning_rate": 2.1260011397897283e-06, + "loss": 0.5625, + "step": 9406 + }, + { + "epoch": 3.3030196629213484, + "grad_norm": 0.92925626039505, + "learning_rate": 2.1255440236196973e-06, + "loss": 0.5423, + "step": 9407 + }, + { + "epoch": 3.303370786516854, + "grad_norm": 0.9086340665817261, + "learning_rate": 2.12508692025574e-06, + "loss": 0.5434, + "step": 9408 + }, + { + "epoch": 3.3037219101123596, + "grad_norm": 0.9334898591041565, + "learning_rate": 2.1246298297134898e-06, + "loss": 0.5554, + "step": 9409 + }, + { + "epoch": 3.304073033707865, + "grad_norm": 0.9303518533706665, + "learning_rate": 2.1241727520085772e-06, + "loss": 0.5522, + "step": 9410 + }, + { + "epoch": 3.304424157303371, + "grad_norm": 0.9585745930671692, + "learning_rate": 2.1237156871566362e-06, + "loss": 0.6029, + "step": 9411 + }, + { + "epoch": 3.3047752808988764, + "grad_norm": 0.9129349589347839, + "learning_rate": 2.1232586351732954e-06, + "loss": 0.5489, + "step": 9412 + }, + { + "epoch": 3.305126404494382, + "grad_norm": 0.9065384864807129, + "learning_rate": 2.1228015960741875e-06, + "loss": 0.549, + "step": 9413 + }, + { + "epoch": 3.3054775280898876, + "grad_norm": 0.9644752144813538, + "learning_rate": 2.1223445698749425e-06, + "loss": 0.5567, + "step": 9414 + }, + { + "epoch": 3.305828651685393, + "grad_norm": 0.9337021708488464, + "learning_rate": 2.1218875565911893e-06, + "loss": 0.5742, + "step": 9415 + }, + { + "epoch": 3.306179775280899, + "grad_norm": 0.9227051734924316, + "learning_rate": 2.1214305562385592e-06, + "loss": 0.5763, + "step": 9416 + }, + { + "epoch": 3.3065308988764044, + "grad_norm": 0.8803406953811646, + "learning_rate": 2.1209735688326786e-06, + "loss": 0.5011, + "step": 9417 + }, + { + "epoch": 3.30688202247191, + "grad_norm": 0.9227780699729919, + "learning_rate": 2.120516594389179e-06, + "loss": 0.549, + "step": 9418 + }, + { + "epoch": 3.3072331460674156, + "grad_norm": 0.9258955717086792, + "learning_rate": 2.1200596329236864e-06, + "loss": 0.5733, + "step": 9419 + }, + { + "epoch": 3.307584269662921, + "grad_norm": 0.9044305086135864, + "learning_rate": 2.11960268445183e-06, + "loss": 0.5533, + "step": 9420 + }, + { + "epoch": 3.307935393258427, + "grad_norm": 0.9087320566177368, + "learning_rate": 2.119145748989235e-06, + "loss": 0.5577, + "step": 9421 + }, + { + "epoch": 3.308286516853933, + "grad_norm": 0.8886432647705078, + "learning_rate": 2.1186888265515307e-06, + "loss": 0.5601, + "step": 9422 + }, + { + "epoch": 3.3086376404494384, + "grad_norm": 0.9689134955406189, + "learning_rate": 2.118231917154341e-06, + "loss": 0.5803, + "step": 9423 + }, + { + "epoch": 3.308988764044944, + "grad_norm": 0.9433503746986389, + "learning_rate": 2.1177750208132947e-06, + "loss": 0.5923, + "step": 9424 + }, + { + "epoch": 3.3093398876404496, + "grad_norm": 0.9974418878555298, + "learning_rate": 2.1173181375440138e-06, + "loss": 0.5555, + "step": 9425 + }, + { + "epoch": 3.309691011235955, + "grad_norm": 0.971814751625061, + "learning_rate": 2.1168612673621267e-06, + "loss": 0.5742, + "step": 9426 + }, + { + "epoch": 3.310042134831461, + "grad_norm": 0.9225754737854004, + "learning_rate": 2.116404410283256e-06, + "loss": 0.5667, + "step": 9427 + }, + { + "epoch": 3.3103932584269664, + "grad_norm": 0.9115647077560425, + "learning_rate": 2.1159475663230257e-06, + "loss": 0.5242, + "step": 9428 + }, + { + "epoch": 3.310744382022472, + "grad_norm": 0.9351648688316345, + "learning_rate": 2.1154907354970604e-06, + "loss": 0.5294, + "step": 9429 + }, + { + "epoch": 3.3110955056179776, + "grad_norm": 0.906618595123291, + "learning_rate": 2.1150339178209826e-06, + "loss": 0.5298, + "step": 9430 + }, + { + "epoch": 3.311446629213483, + "grad_norm": 0.9469627141952515, + "learning_rate": 2.114577113310416e-06, + "loss": 0.5645, + "step": 9431 + }, + { + "epoch": 3.311797752808989, + "grad_norm": 0.944211483001709, + "learning_rate": 2.114120321980981e-06, + "loss": 0.5533, + "step": 9432 + }, + { + "epoch": 3.3121488764044944, + "grad_norm": 0.9431511759757996, + "learning_rate": 2.113663543848302e-06, + "loss": 0.5812, + "step": 9433 + }, + { + "epoch": 3.3125, + "grad_norm": 0.9242687821388245, + "learning_rate": 2.1132067789279985e-06, + "loss": 0.5261, + "step": 9434 + }, + { + "epoch": 3.3128511235955056, + "grad_norm": 0.9508368372917175, + "learning_rate": 2.1127500272356933e-06, + "loss": 0.56, + "step": 9435 + }, + { + "epoch": 3.313202247191011, + "grad_norm": 0.9544445276260376, + "learning_rate": 2.1122932887870044e-06, + "loss": 0.5651, + "step": 9436 + }, + { + "epoch": 3.313553370786517, + "grad_norm": 0.924429178237915, + "learning_rate": 2.1118365635975542e-06, + "loss": 0.5366, + "step": 9437 + }, + { + "epoch": 3.3139044943820224, + "grad_norm": 0.9226158857345581, + "learning_rate": 2.1113798516829606e-06, + "loss": 0.5554, + "step": 9438 + }, + { + "epoch": 3.314255617977528, + "grad_norm": 0.9302194714546204, + "learning_rate": 2.1109231530588447e-06, + "loss": 0.5766, + "step": 9439 + }, + { + "epoch": 3.3146067415730336, + "grad_norm": 0.9131645560264587, + "learning_rate": 2.110466467740824e-06, + "loss": 0.5284, + "step": 9440 + }, + { + "epoch": 3.314957865168539, + "grad_norm": 0.9383044838905334, + "learning_rate": 2.110009795744515e-06, + "loss": 0.5747, + "step": 9441 + }, + { + "epoch": 3.315308988764045, + "grad_norm": 0.9283998608589172, + "learning_rate": 2.109553137085539e-06, + "loss": 0.5631, + "step": 9442 + }, + { + "epoch": 3.3156601123595504, + "grad_norm": 0.9435434937477112, + "learning_rate": 2.1090964917795104e-06, + "loss": 0.587, + "step": 9443 + }, + { + "epoch": 3.316011235955056, + "grad_norm": 0.9398729205131531, + "learning_rate": 2.108639859842048e-06, + "loss": 0.5367, + "step": 9444 + }, + { + "epoch": 3.3163623595505616, + "grad_norm": 0.9379832744598389, + "learning_rate": 2.108183241288767e-06, + "loss": 0.5555, + "step": 9445 + }, + { + "epoch": 3.316713483146067, + "grad_norm": 0.9372707605361938, + "learning_rate": 2.107726636135285e-06, + "loss": 0.5222, + "step": 9446 + }, + { + "epoch": 3.317064606741573, + "grad_norm": 0.9698647260665894, + "learning_rate": 2.107270044397215e-06, + "loss": 0.5573, + "step": 9447 + }, + { + "epoch": 3.317415730337079, + "grad_norm": 0.9479122161865234, + "learning_rate": 2.106813466090174e-06, + "loss": 0.5684, + "step": 9448 + }, + { + "epoch": 3.3177668539325844, + "grad_norm": 0.9156413078308105, + "learning_rate": 2.1063569012297756e-06, + "loss": 0.5739, + "step": 9449 + }, + { + "epoch": 3.31811797752809, + "grad_norm": 0.9281979203224182, + "learning_rate": 2.1059003498316355e-06, + "loss": 0.5596, + "step": 9450 + }, + { + "epoch": 3.3184691011235956, + "grad_norm": 0.9554669857025146, + "learning_rate": 2.1054438119113655e-06, + "loss": 0.5631, + "step": 9451 + }, + { + "epoch": 3.318820224719101, + "grad_norm": 0.922210156917572, + "learning_rate": 2.1049872874845794e-06, + "loss": 0.5515, + "step": 9452 + }, + { + "epoch": 3.319171348314607, + "grad_norm": 0.9713139533996582, + "learning_rate": 2.1045307765668905e-06, + "loss": 0.5735, + "step": 9453 + }, + { + "epoch": 3.3195224719101124, + "grad_norm": 0.9374245405197144, + "learning_rate": 2.1040742791739093e-06, + "loss": 0.5419, + "step": 9454 + }, + { + "epoch": 3.319873595505618, + "grad_norm": 0.9070262908935547, + "learning_rate": 2.1036177953212507e-06, + "loss": 0.5232, + "step": 9455 + }, + { + "epoch": 3.3202247191011236, + "grad_norm": 0.9400661587715149, + "learning_rate": 2.1031613250245227e-06, + "loss": 0.5527, + "step": 9456 + }, + { + "epoch": 3.320575842696629, + "grad_norm": 0.9582133889198303, + "learning_rate": 2.1027048682993394e-06, + "loss": 0.5461, + "step": 9457 + }, + { + "epoch": 3.320926966292135, + "grad_norm": 0.9912821054458618, + "learning_rate": 2.102248425161308e-06, + "loss": 0.5553, + "step": 9458 + }, + { + "epoch": 3.3212780898876404, + "grad_norm": 0.9318990707397461, + "learning_rate": 2.1017919956260414e-06, + "loss": 0.5532, + "step": 9459 + }, + { + "epoch": 3.321629213483146, + "grad_norm": 0.9134421348571777, + "learning_rate": 2.1013355797091467e-06, + "loss": 0.5332, + "step": 9460 + }, + { + "epoch": 3.3219803370786516, + "grad_norm": 0.9602994322776794, + "learning_rate": 2.1008791774262353e-06, + "loss": 0.5655, + "step": 9461 + }, + { + "epoch": 3.322331460674157, + "grad_norm": 0.9216490387916565, + "learning_rate": 2.1004227887929134e-06, + "loss": 0.5463, + "step": 9462 + }, + { + "epoch": 3.322682584269663, + "grad_norm": 0.9316028356552124, + "learning_rate": 2.099966413824791e-06, + "loss": 0.5524, + "step": 9463 + }, + { + "epoch": 3.3230337078651684, + "grad_norm": 0.9684726595878601, + "learning_rate": 2.099510052537475e-06, + "loss": 0.5318, + "step": 9464 + }, + { + "epoch": 3.323384831460674, + "grad_norm": 0.9174932837486267, + "learning_rate": 2.0990537049465713e-06, + "loss": 0.5537, + "step": 9465 + }, + { + "epoch": 3.32373595505618, + "grad_norm": 0.9524845480918884, + "learning_rate": 2.0985973710676894e-06, + "loss": 0.5748, + "step": 9466 + }, + { + "epoch": 3.3240870786516856, + "grad_norm": 0.9567570686340332, + "learning_rate": 2.0981410509164324e-06, + "loss": 0.5398, + "step": 9467 + }, + { + "epoch": 3.324438202247191, + "grad_norm": 0.9324995279312134, + "learning_rate": 2.097684744508409e-06, + "loss": 0.5659, + "step": 9468 + }, + { + "epoch": 3.324789325842697, + "grad_norm": 0.9247406125068665, + "learning_rate": 2.0972284518592217e-06, + "loss": 0.5635, + "step": 9469 + }, + { + "epoch": 3.3251404494382024, + "grad_norm": 0.947342574596405, + "learning_rate": 2.096772172984478e-06, + "loss": 0.577, + "step": 9470 + }, + { + "epoch": 3.325491573033708, + "grad_norm": 0.9182720184326172, + "learning_rate": 2.0963159078997797e-06, + "loss": 0.5422, + "step": 9471 + }, + { + "epoch": 3.3258426966292136, + "grad_norm": 0.8973906636238098, + "learning_rate": 2.0958596566207327e-06, + "loss": 0.57, + "step": 9472 + }, + { + "epoch": 3.326193820224719, + "grad_norm": 0.9193776845932007, + "learning_rate": 2.0954034191629396e-06, + "loss": 0.5856, + "step": 9473 + }, + { + "epoch": 3.326544943820225, + "grad_norm": 0.9366991519927979, + "learning_rate": 2.094947195542003e-06, + "loss": 0.5784, + "step": 9474 + }, + { + "epoch": 3.3268960674157304, + "grad_norm": 0.9532985687255859, + "learning_rate": 2.0944909857735256e-06, + "loss": 0.5638, + "step": 9475 + }, + { + "epoch": 3.327247191011236, + "grad_norm": 0.9402126669883728, + "learning_rate": 2.09403478987311e-06, + "loss": 0.586, + "step": 9476 + }, + { + "epoch": 3.3275983146067416, + "grad_norm": 0.9697914123535156, + "learning_rate": 2.0935786078563577e-06, + "loss": 0.5461, + "step": 9477 + }, + { + "epoch": 3.327949438202247, + "grad_norm": 0.9512802362442017, + "learning_rate": 2.093122439738868e-06, + "loss": 0.5562, + "step": 9478 + }, + { + "epoch": 3.328300561797753, + "grad_norm": 0.9340483546257019, + "learning_rate": 2.0926662855362433e-06, + "loss": 0.5597, + "step": 9479 + }, + { + "epoch": 3.3286516853932584, + "grad_norm": 0.9480443000793457, + "learning_rate": 2.0922101452640823e-06, + "loss": 0.5563, + "step": 9480 + }, + { + "epoch": 3.329002808988764, + "grad_norm": 0.9373241662979126, + "learning_rate": 2.0917540189379864e-06, + "loss": 0.5599, + "step": 9481 + }, + { + "epoch": 3.3293539325842696, + "grad_norm": 0.971430778503418, + "learning_rate": 2.0912979065735536e-06, + "loss": 0.5945, + "step": 9482 + }, + { + "epoch": 3.329705056179775, + "grad_norm": 0.9626575112342834, + "learning_rate": 2.0908418081863824e-06, + "loss": 0.5459, + "step": 9483 + }, + { + "epoch": 3.330056179775281, + "grad_norm": 0.9483453035354614, + "learning_rate": 2.090385723792071e-06, + "loss": 0.5648, + "step": 9484 + }, + { + "epoch": 3.3304073033707864, + "grad_norm": 0.9109677076339722, + "learning_rate": 2.089929653406218e-06, + "loss": 0.5229, + "step": 9485 + }, + { + "epoch": 3.330758426966292, + "grad_norm": 0.9673324823379517, + "learning_rate": 2.089473597044419e-06, + "loss": 0.5735, + "step": 9486 + }, + { + "epoch": 3.3311095505617976, + "grad_norm": 0.9559398293495178, + "learning_rate": 2.0890175547222724e-06, + "loss": 0.5615, + "step": 9487 + }, + { + "epoch": 3.331460674157303, + "grad_norm": 0.9583089351654053, + "learning_rate": 2.0885615264553736e-06, + "loss": 0.5558, + "step": 9488 + }, + { + "epoch": 3.331811797752809, + "grad_norm": 0.9480681419372559, + "learning_rate": 2.0881055122593188e-06, + "loss": 0.5528, + "step": 9489 + }, + { + "epoch": 3.3321629213483144, + "grad_norm": 0.9278402328491211, + "learning_rate": 2.0876495121497027e-06, + "loss": 0.5279, + "step": 9490 + }, + { + "epoch": 3.33251404494382, + "grad_norm": 0.9504565596580505, + "learning_rate": 2.0871935261421204e-06, + "loss": 0.5526, + "step": 9491 + }, + { + "epoch": 3.332865168539326, + "grad_norm": 0.9514027833938599, + "learning_rate": 2.0867375542521663e-06, + "loss": 0.5852, + "step": 9492 + }, + { + "epoch": 3.3332162921348316, + "grad_norm": 0.9220589995384216, + "learning_rate": 2.0862815964954342e-06, + "loss": 0.5403, + "step": 9493 + }, + { + "epoch": 3.333567415730337, + "grad_norm": 0.9586886167526245, + "learning_rate": 2.085825652887517e-06, + "loss": 0.58, + "step": 9494 + }, + { + "epoch": 3.333918539325843, + "grad_norm": 0.9174534678459167, + "learning_rate": 2.085369723444009e-06, + "loss": 0.5485, + "step": 9495 + }, + { + "epoch": 3.3342696629213484, + "grad_norm": 0.9672203063964844, + "learning_rate": 2.0849138081805013e-06, + "loss": 0.573, + "step": 9496 + }, + { + "epoch": 3.334620786516854, + "grad_norm": 0.9239518046379089, + "learning_rate": 2.084457907112586e-06, + "loss": 0.5513, + "step": 9497 + }, + { + "epoch": 3.3349719101123596, + "grad_norm": 0.9192790985107422, + "learning_rate": 2.084002020255855e-06, + "loss": 0.5407, + "step": 9498 + }, + { + "epoch": 3.335323033707865, + "grad_norm": 0.9636248350143433, + "learning_rate": 2.083546147625898e-06, + "loss": 0.5903, + "step": 9499 + }, + { + "epoch": 3.335674157303371, + "grad_norm": 0.9345899820327759, + "learning_rate": 2.0830902892383075e-06, + "loss": 0.5584, + "step": 9500 + }, + { + "epoch": 3.3360252808988764, + "grad_norm": 0.9466955661773682, + "learning_rate": 2.0826344451086715e-06, + "loss": 0.574, + "step": 9501 + }, + { + "epoch": 3.336376404494382, + "grad_norm": 0.933061957359314, + "learning_rate": 2.082178615252581e-06, + "loss": 0.5451, + "step": 9502 + }, + { + "epoch": 3.3367275280898876, + "grad_norm": 0.9231393337249756, + "learning_rate": 2.081722799685625e-06, + "loss": 0.531, + "step": 9503 + }, + { + "epoch": 3.337078651685393, + "grad_norm": 0.9501417875289917, + "learning_rate": 2.0812669984233896e-06, + "loss": 0.5902, + "step": 9504 + }, + { + "epoch": 3.337429775280899, + "grad_norm": 0.937694787979126, + "learning_rate": 2.080811211481466e-06, + "loss": 0.5611, + "step": 9505 + }, + { + "epoch": 3.3377808988764044, + "grad_norm": 0.9235745668411255, + "learning_rate": 2.080355438875439e-06, + "loss": 0.5282, + "step": 9506 + }, + { + "epoch": 3.33813202247191, + "grad_norm": 0.9877792000770569, + "learning_rate": 2.079899680620898e-06, + "loss": 0.5813, + "step": 9507 + }, + { + "epoch": 3.3384831460674156, + "grad_norm": 0.9050416946411133, + "learning_rate": 2.0794439367334273e-06, + "loss": 0.5341, + "step": 9508 + }, + { + "epoch": 3.338834269662921, + "grad_norm": 0.932400643825531, + "learning_rate": 2.078988207228615e-06, + "loss": 0.5568, + "step": 9509 + }, + { + "epoch": 3.339185393258427, + "grad_norm": 0.9466111063957214, + "learning_rate": 2.078532492122045e-06, + "loss": 0.5477, + "step": 9510 + }, + { + "epoch": 3.339536516853933, + "grad_norm": 0.9461885690689087, + "learning_rate": 2.0780767914293045e-06, + "loss": 0.5671, + "step": 9511 + }, + { + "epoch": 3.3398876404494384, + "grad_norm": 0.9253801107406616, + "learning_rate": 2.077621105165975e-06, + "loss": 0.5256, + "step": 9512 + }, + { + "epoch": 3.340238764044944, + "grad_norm": 0.9370130300521851, + "learning_rate": 2.0771654333476436e-06, + "loss": 0.5418, + "step": 9513 + }, + { + "epoch": 3.3405898876404496, + "grad_norm": 0.9081478118896484, + "learning_rate": 2.0767097759898926e-06, + "loss": 0.5347, + "step": 9514 + }, + { + "epoch": 3.340941011235955, + "grad_norm": 0.9352136254310608, + "learning_rate": 2.0762541331083036e-06, + "loss": 0.5675, + "step": 9515 + }, + { + "epoch": 3.341292134831461, + "grad_norm": 0.9467495083808899, + "learning_rate": 2.075798504718462e-06, + "loss": 0.5591, + "step": 9516 + }, + { + "epoch": 3.3416432584269664, + "grad_norm": 0.8988256454467773, + "learning_rate": 2.0753428908359476e-06, + "loss": 0.5219, + "step": 9517 + }, + { + "epoch": 3.341994382022472, + "grad_norm": 0.9110228419303894, + "learning_rate": 2.0748872914763438e-06, + "loss": 0.5542, + "step": 9518 + }, + { + "epoch": 3.3423455056179776, + "grad_norm": 0.9184408187866211, + "learning_rate": 2.07443170665523e-06, + "loss": 0.5364, + "step": 9519 + }, + { + "epoch": 3.342696629213483, + "grad_norm": 0.924029529094696, + "learning_rate": 2.0739761363881886e-06, + "loss": 0.5415, + "step": 9520 + }, + { + "epoch": 3.343047752808989, + "grad_norm": 0.9409920573234558, + "learning_rate": 2.0735205806907975e-06, + "loss": 0.5528, + "step": 9521 + }, + { + "epoch": 3.3433988764044944, + "grad_norm": 0.9431738257408142, + "learning_rate": 2.0730650395786387e-06, + "loss": 0.565, + "step": 9522 + }, + { + "epoch": 3.34375, + "grad_norm": 0.9130451679229736, + "learning_rate": 2.0726095130672895e-06, + "loss": 0.5463, + "step": 9523 + }, + { + "epoch": 3.3441011235955056, + "grad_norm": 0.9336493015289307, + "learning_rate": 2.0721540011723303e-06, + "loss": 0.5361, + "step": 9524 + }, + { + "epoch": 3.344452247191011, + "grad_norm": 0.9670137763023376, + "learning_rate": 2.0716985039093366e-06, + "loss": 0.5738, + "step": 9525 + }, + { + "epoch": 3.344803370786517, + "grad_norm": 0.8999162912368774, + "learning_rate": 2.071243021293889e-06, + "loss": 0.5158, + "step": 9526 + }, + { + "epoch": 3.3451544943820224, + "grad_norm": 0.936633825302124, + "learning_rate": 2.0707875533415633e-06, + "loss": 0.5575, + "step": 9527 + }, + { + "epoch": 3.345505617977528, + "grad_norm": 0.9230221509933472, + "learning_rate": 2.070332100067935e-06, + "loss": 0.5753, + "step": 9528 + }, + { + "epoch": 3.3458567415730336, + "grad_norm": 0.9422169327735901, + "learning_rate": 2.069876661488582e-06, + "loss": 0.5792, + "step": 9529 + }, + { + "epoch": 3.346207865168539, + "grad_norm": 0.9151440858840942, + "learning_rate": 2.0694212376190776e-06, + "loss": 0.5354, + "step": 9530 + }, + { + "epoch": 3.346558988764045, + "grad_norm": 0.952183187007904, + "learning_rate": 2.068965828475e-06, + "loss": 0.5537, + "step": 9531 + }, + { + "epoch": 3.3469101123595504, + "grad_norm": 0.9554935693740845, + "learning_rate": 2.0685104340719213e-06, + "loss": 0.5881, + "step": 9532 + }, + { + "epoch": 3.347261235955056, + "grad_norm": 0.9139998555183411, + "learning_rate": 2.0680550544254176e-06, + "loss": 0.5914, + "step": 9533 + }, + { + "epoch": 3.3476123595505616, + "grad_norm": 0.9479494690895081, + "learning_rate": 2.06759968955106e-06, + "loss": 0.5603, + "step": 9534 + }, + { + "epoch": 3.347963483146067, + "grad_norm": 0.9495264291763306, + "learning_rate": 2.067144339464424e-06, + "loss": 0.5461, + "step": 9535 + }, + { + "epoch": 3.348314606741573, + "grad_norm": 0.9160444140434265, + "learning_rate": 2.0666890041810804e-06, + "loss": 0.5676, + "step": 9536 + }, + { + "epoch": 3.348665730337079, + "grad_norm": 0.9321492910385132, + "learning_rate": 2.066233683716603e-06, + "loss": 0.5432, + "step": 9537 + }, + { + "epoch": 3.3490168539325844, + "grad_norm": 0.9376959800720215, + "learning_rate": 2.0657783780865616e-06, + "loss": 0.5341, + "step": 9538 + }, + { + "epoch": 3.34936797752809, + "grad_norm": 0.9719323515892029, + "learning_rate": 2.0653230873065293e-06, + "loss": 0.5531, + "step": 9539 + }, + { + "epoch": 3.3497191011235956, + "grad_norm": 0.9366676807403564, + "learning_rate": 2.064867811392075e-06, + "loss": 0.55, + "step": 9540 + }, + { + "epoch": 3.350070224719101, + "grad_norm": 0.968198299407959, + "learning_rate": 2.0644125503587685e-06, + "loss": 0.5615, + "step": 9541 + }, + { + "epoch": 3.350421348314607, + "grad_norm": 0.9379764795303345, + "learning_rate": 2.0639573042221807e-06, + "loss": 0.5338, + "step": 9542 + }, + { + "epoch": 3.3507724719101124, + "grad_norm": 0.9251357913017273, + "learning_rate": 2.063502072997879e-06, + "loss": 0.5634, + "step": 9543 + }, + { + "epoch": 3.351123595505618, + "grad_norm": 0.9515902400016785, + "learning_rate": 2.0630468567014345e-06, + "loss": 0.5489, + "step": 9544 + }, + { + "epoch": 3.3514747191011236, + "grad_norm": 0.916793942451477, + "learning_rate": 2.0625916553484116e-06, + "loss": 0.545, + "step": 9545 + }, + { + "epoch": 3.351825842696629, + "grad_norm": 0.962908148765564, + "learning_rate": 2.0621364689543815e-06, + "loss": 0.5764, + "step": 9546 + }, + { + "epoch": 3.352176966292135, + "grad_norm": 0.9240784049034119, + "learning_rate": 2.0616812975349084e-06, + "loss": 0.5808, + "step": 9547 + }, + { + "epoch": 3.3525280898876404, + "grad_norm": 0.9395306706428528, + "learning_rate": 2.0612261411055608e-06, + "loss": 0.542, + "step": 9548 + }, + { + "epoch": 3.352879213483146, + "grad_norm": 0.9146497249603271, + "learning_rate": 2.0607709996819024e-06, + "loss": 0.5374, + "step": 9549 + }, + { + "epoch": 3.3532303370786516, + "grad_norm": 0.9776683449745178, + "learning_rate": 2.0603158732795013e-06, + "loss": 0.5939, + "step": 9550 + }, + { + "epoch": 3.353581460674157, + "grad_norm": 0.9309123754501343, + "learning_rate": 2.05986076191392e-06, + "loss": 0.5599, + "step": 9551 + }, + { + "epoch": 3.353932584269663, + "grad_norm": 0.9642617106437683, + "learning_rate": 2.0594056656007253e-06, + "loss": 0.5581, + "step": 9552 + }, + { + "epoch": 3.3542837078651684, + "grad_norm": 0.9486116766929626, + "learning_rate": 2.05895058435548e-06, + "loss": 0.5532, + "step": 9553 + }, + { + "epoch": 3.354634831460674, + "grad_norm": 0.9248676896095276, + "learning_rate": 2.0584955181937456e-06, + "loss": 0.5326, + "step": 9554 + }, + { + "epoch": 3.35498595505618, + "grad_norm": 0.9157546162605286, + "learning_rate": 2.0580404671310876e-06, + "loss": 0.542, + "step": 9555 + }, + { + "epoch": 3.3553370786516856, + "grad_norm": 0.961417019367218, + "learning_rate": 2.057585431183067e-06, + "loss": 0.5209, + "step": 9556 + }, + { + "epoch": 3.355688202247191, + "grad_norm": 0.9612290263175964, + "learning_rate": 2.0571304103652467e-06, + "loss": 0.567, + "step": 9557 + }, + { + "epoch": 3.356039325842697, + "grad_norm": 0.9280309677124023, + "learning_rate": 2.0566754046931862e-06, + "loss": 0.5809, + "step": 9558 + }, + { + "epoch": 3.3563904494382024, + "grad_norm": 0.9515618085861206, + "learning_rate": 2.0562204141824492e-06, + "loss": 0.5754, + "step": 9559 + }, + { + "epoch": 3.356741573033708, + "grad_norm": 0.9460400938987732, + "learning_rate": 2.0557654388485932e-06, + "loss": 0.5592, + "step": 9560 + }, + { + "epoch": 3.3570926966292136, + "grad_norm": 0.9356980919837952, + "learning_rate": 2.055310478707179e-06, + "loss": 0.5678, + "step": 9561 + }, + { + "epoch": 3.357443820224719, + "grad_norm": 0.9338058233261108, + "learning_rate": 2.0548555337737667e-06, + "loss": 0.5614, + "step": 9562 + }, + { + "epoch": 3.357794943820225, + "grad_norm": 0.9196421504020691, + "learning_rate": 2.0544006040639136e-06, + "loss": 0.5389, + "step": 9563 + }, + { + "epoch": 3.3581460674157304, + "grad_norm": 0.9025902152061462, + "learning_rate": 2.053945689593179e-06, + "loss": 0.5607, + "step": 9564 + }, + { + "epoch": 3.358497191011236, + "grad_norm": 0.93089359998703, + "learning_rate": 2.0534907903771196e-06, + "loss": 0.5435, + "step": 9565 + }, + { + "epoch": 3.3588483146067416, + "grad_norm": 0.910881757736206, + "learning_rate": 2.0530359064312936e-06, + "loss": 0.5592, + "step": 9566 + }, + { + "epoch": 3.359199438202247, + "grad_norm": 0.965417742729187, + "learning_rate": 2.0525810377712563e-06, + "loss": 0.5551, + "step": 9567 + }, + { + "epoch": 3.359550561797753, + "grad_norm": 0.9655492901802063, + "learning_rate": 2.0521261844125655e-06, + "loss": 0.5637, + "step": 9568 + }, + { + "epoch": 3.3599016853932584, + "grad_norm": 0.9017603993415833, + "learning_rate": 2.0516713463707753e-06, + "loss": 0.5113, + "step": 9569 + }, + { + "epoch": 3.360252808988764, + "grad_norm": 0.9202604293823242, + "learning_rate": 2.051216523661442e-06, + "loss": 0.5346, + "step": 9570 + }, + { + "epoch": 3.3606039325842696, + "grad_norm": 0.9383999109268188, + "learning_rate": 2.0507617163001194e-06, + "loss": 0.5383, + "step": 9571 + }, + { + "epoch": 3.360955056179775, + "grad_norm": 0.911308765411377, + "learning_rate": 2.0503069243023613e-06, + "loss": 0.5695, + "step": 9572 + }, + { + "epoch": 3.361306179775281, + "grad_norm": 0.9378693103790283, + "learning_rate": 2.049852147683722e-06, + "loss": 0.5636, + "step": 9573 + }, + { + "epoch": 3.3616573033707864, + "grad_norm": 0.9124588370323181, + "learning_rate": 2.0493973864597537e-06, + "loss": 0.5939, + "step": 9574 + }, + { + "epoch": 3.362008426966292, + "grad_norm": 0.9333383440971375, + "learning_rate": 2.0489426406460097e-06, + "loss": 0.5609, + "step": 9575 + }, + { + "epoch": 3.3623595505617976, + "grad_norm": 0.9123498797416687, + "learning_rate": 2.0484879102580413e-06, + "loss": 0.5577, + "step": 9576 + }, + { + "epoch": 3.362710674157303, + "grad_norm": 0.9699660539627075, + "learning_rate": 2.0480331953113998e-06, + "loss": 0.5461, + "step": 9577 + }, + { + "epoch": 3.363061797752809, + "grad_norm": 0.9192150235176086, + "learning_rate": 2.047578495821636e-06, + "loss": 0.5091, + "step": 9578 + }, + { + "epoch": 3.3634129213483144, + "grad_norm": 0.9397869110107422, + "learning_rate": 2.047123811804301e-06, + "loss": 0.5517, + "step": 9579 + }, + { + "epoch": 3.36376404494382, + "grad_norm": 0.9263961911201477, + "learning_rate": 2.0466691432749437e-06, + "loss": 0.5402, + "step": 9580 + }, + { + "epoch": 3.364115168539326, + "grad_norm": 0.9266073703765869, + "learning_rate": 2.046214490249114e-06, + "loss": 0.552, + "step": 9581 + }, + { + "epoch": 3.3644662921348316, + "grad_norm": 0.9194657206535339, + "learning_rate": 2.0457598527423604e-06, + "loss": 0.5199, + "step": 9582 + }, + { + "epoch": 3.364817415730337, + "grad_norm": 0.9734546542167664, + "learning_rate": 2.0453052307702315e-06, + "loss": 0.5584, + "step": 9583 + }, + { + "epoch": 3.365168539325843, + "grad_norm": 1.008378267288208, + "learning_rate": 2.0448506243482734e-06, + "loss": 0.5679, + "step": 9584 + }, + { + "epoch": 3.3655196629213484, + "grad_norm": 0.9432476162910461, + "learning_rate": 2.044396033492036e-06, + "loss": 0.5609, + "step": 9585 + }, + { + "epoch": 3.365870786516854, + "grad_norm": 0.9017815589904785, + "learning_rate": 2.043941458217063e-06, + "loss": 0.5454, + "step": 9586 + }, + { + "epoch": 3.3662219101123596, + "grad_norm": 0.9556635022163391, + "learning_rate": 2.0434868985389032e-06, + "loss": 0.5557, + "step": 9587 + }, + { + "epoch": 3.366573033707865, + "grad_norm": 0.9331648349761963, + "learning_rate": 2.0430323544730997e-06, + "loss": 0.5755, + "step": 9588 + }, + { + "epoch": 3.366924157303371, + "grad_norm": 0.9041140079498291, + "learning_rate": 2.0425778260351997e-06, + "loss": 0.5318, + "step": 9589 + }, + { + "epoch": 3.3672752808988764, + "grad_norm": 0.9196086525917053, + "learning_rate": 2.0421233132407463e-06, + "loss": 0.5555, + "step": 9590 + }, + { + "epoch": 3.367626404494382, + "grad_norm": 0.9029288291931152, + "learning_rate": 2.041668816105283e-06, + "loss": 0.5127, + "step": 9591 + }, + { + "epoch": 3.3679775280898876, + "grad_norm": 0.9077665209770203, + "learning_rate": 2.041214334644355e-06, + "loss": 0.5368, + "step": 9592 + }, + { + "epoch": 3.368328651685393, + "grad_norm": 0.9301674365997314, + "learning_rate": 2.0407598688735033e-06, + "loss": 0.5444, + "step": 9593 + }, + { + "epoch": 3.368679775280899, + "grad_norm": 0.9320032000541687, + "learning_rate": 2.040305418808272e-06, + "loss": 0.5877, + "step": 9594 + }, + { + "epoch": 3.3690308988764044, + "grad_norm": 0.9277612566947937, + "learning_rate": 2.039850984464201e-06, + "loss": 0.5742, + "step": 9595 + }, + { + "epoch": 3.36938202247191, + "grad_norm": 0.9756101965904236, + "learning_rate": 2.0393965658568336e-06, + "loss": 0.5667, + "step": 9596 + }, + { + "epoch": 3.3697331460674156, + "grad_norm": 0.8938387036323547, + "learning_rate": 2.0389421630017083e-06, + "loss": 0.5145, + "step": 9597 + }, + { + "epoch": 3.370084269662921, + "grad_norm": 0.9377100467681885, + "learning_rate": 2.0384877759143677e-06, + "loss": 0.5484, + "step": 9598 + }, + { + "epoch": 3.370435393258427, + "grad_norm": 0.9264451861381531, + "learning_rate": 2.0380334046103493e-06, + "loss": 0.5671, + "step": 9599 + }, + { + "epoch": 3.370786516853933, + "grad_norm": 0.9430090188980103, + "learning_rate": 2.0375790491051938e-06, + "loss": 0.5631, + "step": 9600 + }, + { + "epoch": 3.3711376404494384, + "grad_norm": 0.9397012591362, + "learning_rate": 2.0371247094144385e-06, + "loss": 0.55, + "step": 9601 + }, + { + "epoch": 3.371488764044944, + "grad_norm": 0.9561929702758789, + "learning_rate": 2.036670385553623e-06, + "loss": 0.5659, + "step": 9602 + }, + { + "epoch": 3.3718398876404496, + "grad_norm": 0.9514825940132141, + "learning_rate": 2.036216077538284e-06, + "loss": 0.5729, + "step": 9603 + }, + { + "epoch": 3.372191011235955, + "grad_norm": 0.9548146724700928, + "learning_rate": 2.035761785383957e-06, + "loss": 0.5865, + "step": 9604 + }, + { + "epoch": 3.372542134831461, + "grad_norm": 0.9450569152832031, + "learning_rate": 2.0353075091061805e-06, + "loss": 0.5265, + "step": 9605 + }, + { + "epoch": 3.3728932584269664, + "grad_norm": 0.9797646403312683, + "learning_rate": 2.0348532487204888e-06, + "loss": 0.5685, + "step": 9606 + }, + { + "epoch": 3.373244382022472, + "grad_norm": 0.9625208377838135, + "learning_rate": 2.034399004242419e-06, + "loss": 0.5747, + "step": 9607 + }, + { + "epoch": 3.3735955056179776, + "grad_norm": 1.0091040134429932, + "learning_rate": 2.0339447756875036e-06, + "loss": 0.5699, + "step": 9608 + }, + { + "epoch": 3.373946629213483, + "grad_norm": 0.9598633050918579, + "learning_rate": 2.0334905630712794e-06, + "loss": 0.6017, + "step": 9609 + }, + { + "epoch": 3.374297752808989, + "grad_norm": 0.9466208219528198, + "learning_rate": 2.0330363664092773e-06, + "loss": 0.5711, + "step": 9610 + }, + { + "epoch": 3.3746488764044944, + "grad_norm": 0.9376499056816101, + "learning_rate": 2.032582185717033e-06, + "loss": 0.5504, + "step": 9611 + }, + { + "epoch": 3.375, + "grad_norm": 0.9478402733802795, + "learning_rate": 2.0321280210100775e-06, + "loss": 0.5431, + "step": 9612 + }, + { + "epoch": 3.3753511235955056, + "grad_norm": 0.9561818838119507, + "learning_rate": 2.031673872303944e-06, + "loss": 0.5729, + "step": 9613 + }, + { + "epoch": 3.375702247191011, + "grad_norm": 0.9164089560508728, + "learning_rate": 2.031219739614162e-06, + "loss": 0.5437, + "step": 9614 + }, + { + "epoch": 3.376053370786517, + "grad_norm": 0.9594018459320068, + "learning_rate": 2.030765622956265e-06, + "loss": 0.5585, + "step": 9615 + }, + { + "epoch": 3.3764044943820224, + "grad_norm": 0.9758911728858948, + "learning_rate": 2.0303115223457825e-06, + "loss": 0.5389, + "step": 9616 + }, + { + "epoch": 3.376755617977528, + "grad_norm": 0.9144931435585022, + "learning_rate": 2.029857437798243e-06, + "loss": 0.5451, + "step": 9617 + }, + { + "epoch": 3.3771067415730336, + "grad_norm": 0.9265840649604797, + "learning_rate": 2.029403369329177e-06, + "loss": 0.554, + "step": 9618 + }, + { + "epoch": 3.377457865168539, + "grad_norm": 0.9317816495895386, + "learning_rate": 2.0289493169541127e-06, + "loss": 0.5315, + "step": 9619 + }, + { + "epoch": 3.377808988764045, + "grad_norm": 0.9156888127326965, + "learning_rate": 2.0284952806885797e-06, + "loss": 0.5594, + "step": 9620 + }, + { + "epoch": 3.3781601123595504, + "grad_norm": 0.951923131942749, + "learning_rate": 2.0280412605481032e-06, + "loss": 0.5736, + "step": 9621 + }, + { + "epoch": 3.378511235955056, + "grad_norm": 0.9644631743431091, + "learning_rate": 2.027587256548213e-06, + "loss": 0.5948, + "step": 9622 + }, + { + "epoch": 3.3788623595505616, + "grad_norm": 0.9449539184570312, + "learning_rate": 2.0271332687044337e-06, + "loss": 0.5417, + "step": 9623 + }, + { + "epoch": 3.379213483146067, + "grad_norm": 0.9438310861587524, + "learning_rate": 2.0266792970322925e-06, + "loss": 0.5548, + "step": 9624 + }, + { + "epoch": 3.379564606741573, + "grad_norm": 0.9358301162719727, + "learning_rate": 2.026225341547314e-06, + "loss": 0.5714, + "step": 9625 + }, + { + "epoch": 3.379915730337079, + "grad_norm": 0.9529426097869873, + "learning_rate": 2.0257714022650234e-06, + "loss": 0.5801, + "step": 9626 + }, + { + "epoch": 3.3802668539325844, + "grad_norm": 0.9080844521522522, + "learning_rate": 2.025317479200945e-06, + "loss": 0.5183, + "step": 9627 + }, + { + "epoch": 3.38061797752809, + "grad_norm": 0.9297475218772888, + "learning_rate": 2.0248635723706026e-06, + "loss": 0.5627, + "step": 9628 + }, + { + "epoch": 3.3809691011235956, + "grad_norm": 0.9762905836105347, + "learning_rate": 2.02440968178952e-06, + "loss": 0.5935, + "step": 9629 + }, + { + "epoch": 3.381320224719101, + "grad_norm": 0.9484661221504211, + "learning_rate": 2.0239558074732186e-06, + "loss": 0.5498, + "step": 9630 + }, + { + "epoch": 3.381671348314607, + "grad_norm": 0.9228793382644653, + "learning_rate": 2.0235019494372223e-06, + "loss": 0.5717, + "step": 9631 + }, + { + "epoch": 3.3820224719101124, + "grad_norm": 0.8851661086082458, + "learning_rate": 2.0230481076970497e-06, + "loss": 0.5102, + "step": 9632 + }, + { + "epoch": 3.382373595505618, + "grad_norm": 0.9509567022323608, + "learning_rate": 2.0225942822682254e-06, + "loss": 0.5636, + "step": 9633 + }, + { + "epoch": 3.3827247191011236, + "grad_norm": 0.967288076877594, + "learning_rate": 2.022140473166267e-06, + "loss": 0.5813, + "step": 9634 + }, + { + "epoch": 3.383075842696629, + "grad_norm": 0.8972613215446472, + "learning_rate": 2.0216866804066965e-06, + "loss": 0.5375, + "step": 9635 + }, + { + "epoch": 3.383426966292135, + "grad_norm": 0.9016784429550171, + "learning_rate": 2.021232904005031e-06, + "loss": 0.5658, + "step": 9636 + }, + { + "epoch": 3.3837780898876404, + "grad_norm": 1.076409101486206, + "learning_rate": 2.0207791439767917e-06, + "loss": 0.5863, + "step": 9637 + }, + { + "epoch": 3.384129213483146, + "grad_norm": 0.9085844159126282, + "learning_rate": 2.0203254003374945e-06, + "loss": 0.5008, + "step": 9638 + }, + { + "epoch": 3.3844803370786516, + "grad_norm": 0.9410509467124939, + "learning_rate": 2.0198716731026595e-06, + "loss": 0.5764, + "step": 9639 + }, + { + "epoch": 3.384831460674157, + "grad_norm": 0.9477881789207458, + "learning_rate": 2.019417962287802e-06, + "loss": 0.5535, + "step": 9640 + }, + { + "epoch": 3.385182584269663, + "grad_norm": 0.9573963284492493, + "learning_rate": 2.0189642679084383e-06, + "loss": 0.53, + "step": 9641 + }, + { + "epoch": 3.3855337078651684, + "grad_norm": 0.9202030301094055, + "learning_rate": 2.0185105899800857e-06, + "loss": 0.5405, + "step": 9642 + }, + { + "epoch": 3.385884831460674, + "grad_norm": 0.9354908466339111, + "learning_rate": 2.018056928518258e-06, + "loss": 0.5642, + "step": 9643 + }, + { + "epoch": 3.38623595505618, + "grad_norm": 0.9485152959823608, + "learning_rate": 2.0176032835384717e-06, + "loss": 0.5577, + "step": 9644 + }, + { + "epoch": 3.3865870786516856, + "grad_norm": 0.9381886124610901, + "learning_rate": 2.0171496550562395e-06, + "loss": 0.5668, + "step": 9645 + }, + { + "epoch": 3.386938202247191, + "grad_norm": 0.9463789463043213, + "learning_rate": 2.0166960430870774e-06, + "loss": 0.5712, + "step": 9646 + }, + { + "epoch": 3.387289325842697, + "grad_norm": 0.938942015171051, + "learning_rate": 2.016242447646496e-06, + "loss": 0.554, + "step": 9647 + }, + { + "epoch": 3.3876404494382024, + "grad_norm": 0.9304429888725281, + "learning_rate": 2.0157888687500092e-06, + "loss": 0.5681, + "step": 9648 + }, + { + "epoch": 3.387991573033708, + "grad_norm": 0.9593173265457153, + "learning_rate": 2.0153353064131284e-06, + "loss": 0.5802, + "step": 9649 + }, + { + "epoch": 3.3883426966292136, + "grad_norm": 0.9467602372169495, + "learning_rate": 2.0148817606513657e-06, + "loss": 0.5492, + "step": 9650 + }, + { + "epoch": 3.388693820224719, + "grad_norm": 0.9404729008674622, + "learning_rate": 2.0144282314802314e-06, + "loss": 0.5603, + "step": 9651 + }, + { + "epoch": 3.389044943820225, + "grad_norm": 0.9133000373840332, + "learning_rate": 2.0139747189152366e-06, + "loss": 0.5463, + "step": 9652 + }, + { + "epoch": 3.3893960674157304, + "grad_norm": 0.9134009480476379, + "learning_rate": 2.0135212229718902e-06, + "loss": 0.5576, + "step": 9653 + }, + { + "epoch": 3.389747191011236, + "grad_norm": 0.9958639144897461, + "learning_rate": 2.013067743665702e-06, + "loss": 0.56, + "step": 9654 + }, + { + "epoch": 3.3900983146067416, + "grad_norm": 0.943522572517395, + "learning_rate": 2.01261428101218e-06, + "loss": 0.5865, + "step": 9655 + }, + { + "epoch": 3.390449438202247, + "grad_norm": 0.9264531135559082, + "learning_rate": 2.012160835026832e-06, + "loss": 0.5901, + "step": 9656 + }, + { + "epoch": 3.390800561797753, + "grad_norm": 0.9214319586753845, + "learning_rate": 2.0117074057251667e-06, + "loss": 0.5769, + "step": 9657 + }, + { + "epoch": 3.3911516853932584, + "grad_norm": 0.9377774596214294, + "learning_rate": 2.01125399312269e-06, + "loss": 0.5515, + "step": 9658 + }, + { + "epoch": 3.391502808988764, + "grad_norm": 0.9379155039787292, + "learning_rate": 2.0108005972349087e-06, + "loss": 0.5447, + "step": 9659 + }, + { + "epoch": 3.3918539325842696, + "grad_norm": 0.9265129566192627, + "learning_rate": 2.010347218077328e-06, + "loss": 0.5522, + "step": 9660 + }, + { + "epoch": 3.392205056179775, + "grad_norm": 0.9125092625617981, + "learning_rate": 2.009893855665454e-06, + "loss": 0.538, + "step": 9661 + }, + { + "epoch": 3.392556179775281, + "grad_norm": 0.9417039752006531, + "learning_rate": 2.00944051001479e-06, + "loss": 0.5553, + "step": 9662 + }, + { + "epoch": 3.3929073033707864, + "grad_norm": 0.9107561111450195, + "learning_rate": 2.0089871811408422e-06, + "loss": 0.515, + "step": 9663 + }, + { + "epoch": 3.393258426966292, + "grad_norm": 0.9479217529296875, + "learning_rate": 2.008533869059111e-06, + "loss": 0.5709, + "step": 9664 + }, + { + "epoch": 3.3936095505617976, + "grad_norm": 0.9541273713111877, + "learning_rate": 2.008080573785102e-06, + "loss": 0.5534, + "step": 9665 + }, + { + "epoch": 3.393960674157303, + "grad_norm": 0.9499912858009338, + "learning_rate": 2.007627295334316e-06, + "loss": 0.5498, + "step": 9666 + }, + { + "epoch": 3.394311797752809, + "grad_norm": 0.9009503722190857, + "learning_rate": 2.007174033722255e-06, + "loss": 0.5463, + "step": 9667 + }, + { + "epoch": 3.3946629213483144, + "grad_norm": 0.9736745357513428, + "learning_rate": 2.0067207889644205e-06, + "loss": 0.5591, + "step": 9668 + }, + { + "epoch": 3.39501404494382, + "grad_norm": 0.9323515892028809, + "learning_rate": 2.006267561076313e-06, + "loss": 0.5476, + "step": 9669 + }, + { + "epoch": 3.395365168539326, + "grad_norm": 0.9423682689666748, + "learning_rate": 2.0058143500734328e-06, + "loss": 0.5538, + "step": 9670 + }, + { + "epoch": 3.3957162921348316, + "grad_norm": 0.9407555460929871, + "learning_rate": 2.005361155971278e-06, + "loss": 0.545, + "step": 9671 + }, + { + "epoch": 3.396067415730337, + "grad_norm": 0.9046871066093445, + "learning_rate": 2.0049079787853495e-06, + "loss": 0.54, + "step": 9672 + }, + { + "epoch": 3.396418539325843, + "grad_norm": 0.9218787550926208, + "learning_rate": 2.004454818531143e-06, + "loss": 0.5978, + "step": 9673 + }, + { + "epoch": 3.3967696629213484, + "grad_norm": 0.9345743060112, + "learning_rate": 2.0040016752241593e-06, + "loss": 0.5573, + "step": 9674 + }, + { + "epoch": 3.397120786516854, + "grad_norm": 0.9159797430038452, + "learning_rate": 2.003548548879892e-06, + "loss": 0.5328, + "step": 9675 + }, + { + "epoch": 3.3974719101123596, + "grad_norm": 0.9360120296478271, + "learning_rate": 2.0030954395138415e-06, + "loss": 0.5814, + "step": 9676 + }, + { + "epoch": 3.397823033707865, + "grad_norm": 0.939478874206543, + "learning_rate": 2.0026423471415003e-06, + "loss": 0.5782, + "step": 9677 + }, + { + "epoch": 3.398174157303371, + "grad_norm": 0.9562158584594727, + "learning_rate": 2.0021892717783663e-06, + "loss": 0.564, + "step": 9678 + }, + { + "epoch": 3.3985252808988764, + "grad_norm": 0.9122350215911865, + "learning_rate": 2.001736213439933e-06, + "loss": 0.5722, + "step": 9679 + }, + { + "epoch": 3.398876404494382, + "grad_norm": 0.9435994029045105, + "learning_rate": 2.0012831721416944e-06, + "loss": 0.5351, + "step": 9680 + }, + { + "epoch": 3.3992275280898876, + "grad_norm": 0.9426365494728088, + "learning_rate": 2.0008301478991447e-06, + "loss": 0.5431, + "step": 9681 + }, + { + "epoch": 3.399578651685393, + "grad_norm": 0.9355964064598083, + "learning_rate": 2.0003771407277762e-06, + "loss": 0.5552, + "step": 9682 + }, + { + "epoch": 3.399929775280899, + "grad_norm": 0.9391768574714661, + "learning_rate": 1.9999241506430833e-06, + "loss": 0.564, + "step": 9683 + }, + { + "epoch": 3.4002808988764044, + "grad_norm": 0.9375246167182922, + "learning_rate": 1.9994711776605553e-06, + "loss": 0.5646, + "step": 9684 + }, + { + "epoch": 3.40063202247191, + "grad_norm": 0.9437415599822998, + "learning_rate": 1.9990182217956856e-06, + "loss": 0.5595, + "step": 9685 + }, + { + "epoch": 3.4009831460674156, + "grad_norm": 0.9372910857200623, + "learning_rate": 1.998565283063963e-06, + "loss": 0.5749, + "step": 9686 + }, + { + "epoch": 3.401334269662921, + "grad_norm": 0.9351158142089844, + "learning_rate": 1.99811236148088e-06, + "loss": 0.5526, + "step": 9687 + }, + { + "epoch": 3.401685393258427, + "grad_norm": 0.9315171241760254, + "learning_rate": 1.9976594570619233e-06, + "loss": 0.5408, + "step": 9688 + }, + { + "epoch": 3.402036516853933, + "grad_norm": 0.9412782788276672, + "learning_rate": 1.9972065698225847e-06, + "loss": 0.5498, + "step": 9689 + }, + { + "epoch": 3.4023876404494384, + "grad_norm": 0.9596142768859863, + "learning_rate": 1.9967536997783495e-06, + "loss": 0.5321, + "step": 9690 + }, + { + "epoch": 3.402738764044944, + "grad_norm": 0.9105181097984314, + "learning_rate": 1.9963008469447087e-06, + "loss": 0.5395, + "step": 9691 + }, + { + "epoch": 3.4030898876404496, + "grad_norm": 0.9111261963844299, + "learning_rate": 1.9958480113371477e-06, + "loss": 0.5299, + "step": 9692 + }, + { + "epoch": 3.403441011235955, + "grad_norm": 0.9140063524246216, + "learning_rate": 1.995395192971153e-06, + "loss": 0.5259, + "step": 9693 + }, + { + "epoch": 3.403792134831461, + "grad_norm": 0.8964090347290039, + "learning_rate": 1.994942391862211e-06, + "loss": 0.5177, + "step": 9694 + }, + { + "epoch": 3.4041432584269664, + "grad_norm": 0.9768686890602112, + "learning_rate": 1.9944896080258063e-06, + "loss": 0.5924, + "step": 9695 + }, + { + "epoch": 3.404494382022472, + "grad_norm": 0.955869734287262, + "learning_rate": 1.9940368414774255e-06, + "loss": 0.5562, + "step": 9696 + }, + { + "epoch": 3.4048455056179776, + "grad_norm": 0.918975830078125, + "learning_rate": 1.9935840922325505e-06, + "loss": 0.5493, + "step": 9697 + }, + { + "epoch": 3.405196629213483, + "grad_norm": 0.9174593687057495, + "learning_rate": 1.9931313603066677e-06, + "loss": 0.5215, + "step": 9698 + }, + { + "epoch": 3.405547752808989, + "grad_norm": 0.919669508934021, + "learning_rate": 1.9926786457152574e-06, + "loss": 0.549, + "step": 9699 + }, + { + "epoch": 3.4058988764044944, + "grad_norm": 0.9336464405059814, + "learning_rate": 1.9922259484738042e-06, + "loss": 0.5385, + "step": 9700 + }, + { + "epoch": 3.40625, + "grad_norm": 0.9459969401359558, + "learning_rate": 1.991773268597788e-06, + "loss": 0.5541, + "step": 9701 + }, + { + "epoch": 3.4066011235955056, + "grad_norm": 0.9446254372596741, + "learning_rate": 1.991320606102692e-06, + "loss": 0.5449, + "step": 9702 + }, + { + "epoch": 3.406952247191011, + "grad_norm": 0.9209154844284058, + "learning_rate": 1.9908679610039955e-06, + "loss": 0.5632, + "step": 9703 + }, + { + "epoch": 3.407303370786517, + "grad_norm": 0.917721152305603, + "learning_rate": 1.9904153333171796e-06, + "loss": 0.5536, + "step": 9704 + }, + { + "epoch": 3.4076544943820224, + "grad_norm": 0.9303230047225952, + "learning_rate": 1.9899627230577236e-06, + "loss": 0.5484, + "step": 9705 + }, + { + "epoch": 3.408005617977528, + "grad_norm": 0.8980202674865723, + "learning_rate": 1.989510130241105e-06, + "loss": 0.55, + "step": 9706 + }, + { + "epoch": 3.4083567415730336, + "grad_norm": 0.9986125230789185, + "learning_rate": 1.989057554882804e-06, + "loss": 0.574, + "step": 9707 + }, + { + "epoch": 3.408707865168539, + "grad_norm": 0.958685576915741, + "learning_rate": 1.9886049969982963e-06, + "loss": 0.5606, + "step": 9708 + }, + { + "epoch": 3.409058988764045, + "grad_norm": 0.9456254839897156, + "learning_rate": 1.988152456603061e-06, + "loss": 0.5402, + "step": 9709 + }, + { + "epoch": 3.4094101123595504, + "grad_norm": 0.9134203791618347, + "learning_rate": 1.987699933712573e-06, + "loss": 0.58, + "step": 9710 + }, + { + "epoch": 3.409761235955056, + "grad_norm": 0.9379029870033264, + "learning_rate": 1.9872474283423096e-06, + "loss": 0.5646, + "step": 9711 + }, + { + "epoch": 3.4101123595505616, + "grad_norm": 0.9907280802726746, + "learning_rate": 1.986794940507744e-06, + "loss": 0.5404, + "step": 9712 + }, + { + "epoch": 3.410463483146067, + "grad_norm": 0.9637587070465088, + "learning_rate": 1.9863424702243545e-06, + "loss": 0.5847, + "step": 9713 + }, + { + "epoch": 3.410814606741573, + "grad_norm": 0.940900444984436, + "learning_rate": 1.985890017507611e-06, + "loss": 0.5296, + "step": 9714 + }, + { + "epoch": 3.411165730337079, + "grad_norm": 0.9550067782402039, + "learning_rate": 1.9854375823729904e-06, + "loss": 0.5807, + "step": 9715 + }, + { + "epoch": 3.4115168539325844, + "grad_norm": 0.9646278619766235, + "learning_rate": 1.984985164835964e-06, + "loss": 0.5335, + "step": 9716 + }, + { + "epoch": 3.41186797752809, + "grad_norm": 0.9329323172569275, + "learning_rate": 1.9845327649120034e-06, + "loss": 0.5671, + "step": 9717 + }, + { + "epoch": 3.4122191011235956, + "grad_norm": 0.9579326510429382, + "learning_rate": 1.984080382616582e-06, + "loss": 0.5957, + "step": 9718 + }, + { + "epoch": 3.412570224719101, + "grad_norm": 0.9227491021156311, + "learning_rate": 1.983628017965169e-06, + "loss": 0.5582, + "step": 9719 + }, + { + "epoch": 3.412921348314607, + "grad_norm": 0.9401417970657349, + "learning_rate": 1.983175670973237e-06, + "loss": 0.5281, + "step": 9720 + }, + { + "epoch": 3.4132724719101124, + "grad_norm": 0.9340532422065735, + "learning_rate": 1.9827233416562536e-06, + "loss": 0.5523, + "step": 9721 + }, + { + "epoch": 3.413623595505618, + "grad_norm": 0.9342505931854248, + "learning_rate": 1.9822710300296904e-06, + "loss": 0.5459, + "step": 9722 + }, + { + "epoch": 3.4139747191011236, + "grad_norm": 0.907447338104248, + "learning_rate": 1.981818736109014e-06, + "loss": 0.5346, + "step": 9723 + }, + { + "epoch": 3.414325842696629, + "grad_norm": 0.9114577770233154, + "learning_rate": 1.9813664599096945e-06, + "loss": 0.576, + "step": 9724 + }, + { + "epoch": 3.414676966292135, + "grad_norm": 0.9425318241119385, + "learning_rate": 1.980914201447197e-06, + "loss": 0.5399, + "step": 9725 + }, + { + "epoch": 3.4150280898876404, + "grad_norm": 0.9534419775009155, + "learning_rate": 1.980461960736991e-06, + "loss": 0.5465, + "step": 9726 + }, + { + "epoch": 3.415379213483146, + "grad_norm": 0.9463133811950684, + "learning_rate": 1.9800097377945408e-06, + "loss": 0.5768, + "step": 9727 + }, + { + "epoch": 3.4157303370786516, + "grad_norm": 0.920906662940979, + "learning_rate": 1.979557532635312e-06, + "loss": 0.5649, + "step": 9728 + }, + { + "epoch": 3.416081460674157, + "grad_norm": 0.9497731924057007, + "learning_rate": 1.9791053452747716e-06, + "loss": 0.5777, + "step": 9729 + }, + { + "epoch": 3.416432584269663, + "grad_norm": 0.9365454316139221, + "learning_rate": 1.9786531757283813e-06, + "loss": 0.5363, + "step": 9730 + }, + { + "epoch": 3.4167837078651684, + "grad_norm": 0.9350811839103699, + "learning_rate": 1.978201024011607e-06, + "loss": 0.5573, + "step": 9731 + }, + { + "epoch": 3.417134831460674, + "grad_norm": 0.9867560267448425, + "learning_rate": 1.9777488901399106e-06, + "loss": 0.5575, + "step": 9732 + }, + { + "epoch": 3.41748595505618, + "grad_norm": 0.9544400572776794, + "learning_rate": 1.977296774128756e-06, + "loss": 0.5544, + "step": 9733 + }, + { + "epoch": 3.4178370786516856, + "grad_norm": 0.8971168994903564, + "learning_rate": 1.9768446759936038e-06, + "loss": 0.5474, + "step": 9734 + }, + { + "epoch": 3.418188202247191, + "grad_norm": 0.9146912097930908, + "learning_rate": 1.9763925957499166e-06, + "loss": 0.5625, + "step": 9735 + }, + { + "epoch": 3.418539325842697, + "grad_norm": 0.929466724395752, + "learning_rate": 1.975940533413154e-06, + "loss": 0.551, + "step": 9736 + }, + { + "epoch": 3.4188904494382024, + "grad_norm": 0.9285315871238708, + "learning_rate": 1.975488488998777e-06, + "loss": 0.5859, + "step": 9737 + }, + { + "epoch": 3.419241573033708, + "grad_norm": 0.9782288670539856, + "learning_rate": 1.9750364625222446e-06, + "loss": 0.5585, + "step": 9738 + }, + { + "epoch": 3.4195926966292136, + "grad_norm": 0.9722647070884705, + "learning_rate": 1.974584453999016e-06, + "loss": 0.5665, + "step": 9739 + }, + { + "epoch": 3.419943820224719, + "grad_norm": 0.9328365325927734, + "learning_rate": 1.9741324634445496e-06, + "loss": 0.5838, + "step": 9740 + }, + { + "epoch": 3.420294943820225, + "grad_norm": 0.9420405030250549, + "learning_rate": 1.9736804908743036e-06, + "loss": 0.5495, + "step": 9741 + }, + { + "epoch": 3.4206460674157304, + "grad_norm": 0.9678810834884644, + "learning_rate": 1.973228536303734e-06, + "loss": 0.5833, + "step": 9742 + }, + { + "epoch": 3.420997191011236, + "grad_norm": 0.9679741263389587, + "learning_rate": 1.972776599748297e-06, + "loss": 0.5779, + "step": 9743 + }, + { + "epoch": 3.4213483146067416, + "grad_norm": 0.9151297807693481, + "learning_rate": 1.9723246812234496e-06, + "loss": 0.5078, + "step": 9744 + }, + { + "epoch": 3.421699438202247, + "grad_norm": 0.9637489318847656, + "learning_rate": 1.9718727807446466e-06, + "loss": 0.5385, + "step": 9745 + }, + { + "epoch": 3.422050561797753, + "grad_norm": 0.9603188633918762, + "learning_rate": 1.9714208983273423e-06, + "loss": 0.5339, + "step": 9746 + }, + { + "epoch": 3.4224016853932584, + "grad_norm": 0.9456979036331177, + "learning_rate": 1.9709690339869917e-06, + "loss": 0.5921, + "step": 9747 + }, + { + "epoch": 3.422752808988764, + "grad_norm": 0.9333366751670837, + "learning_rate": 1.9705171877390466e-06, + "loss": 0.5537, + "step": 9748 + }, + { + "epoch": 3.4231039325842696, + "grad_norm": 0.9102929830551147, + "learning_rate": 1.970065359598961e-06, + "loss": 0.5846, + "step": 9749 + }, + { + "epoch": 3.423455056179775, + "grad_norm": 0.9263753294944763, + "learning_rate": 1.969613549582187e-06, + "loss": 0.5739, + "step": 9750 + }, + { + "epoch": 3.423806179775281, + "grad_norm": 0.9241089820861816, + "learning_rate": 1.969161757704175e-06, + "loss": 0.5311, + "step": 9751 + }, + { + "epoch": 3.4241573033707864, + "grad_norm": 0.9483809471130371, + "learning_rate": 1.9687099839803773e-06, + "loss": 0.5648, + "step": 9752 + }, + { + "epoch": 3.424508426966292, + "grad_norm": 0.9442264437675476, + "learning_rate": 1.9682582284262426e-06, + "loss": 0.566, + "step": 9753 + }, + { + "epoch": 3.4248595505617976, + "grad_norm": 0.9246250987052917, + "learning_rate": 1.967806491057223e-06, + "loss": 0.5362, + "step": 9754 + }, + { + "epoch": 3.425210674157303, + "grad_norm": 0.9767395257949829, + "learning_rate": 1.967354771888765e-06, + "loss": 0.5671, + "step": 9755 + }, + { + "epoch": 3.425561797752809, + "grad_norm": 0.9626738429069519, + "learning_rate": 1.966903070936318e-06, + "loss": 0.5633, + "step": 9756 + }, + { + "epoch": 3.4259129213483144, + "grad_norm": 0.9538341164588928, + "learning_rate": 1.96645138821533e-06, + "loss": 0.5506, + "step": 9757 + }, + { + "epoch": 3.42626404494382, + "grad_norm": 0.9540991187095642, + "learning_rate": 1.9659997237412477e-06, + "loss": 0.578, + "step": 9758 + }, + { + "epoch": 3.426615168539326, + "grad_norm": 0.9297328591346741, + "learning_rate": 1.9655480775295183e-06, + "loss": 0.5708, + "step": 9759 + }, + { + "epoch": 3.4269662921348316, + "grad_norm": 0.9184114336967468, + "learning_rate": 1.9650964495955867e-06, + "loss": 0.5693, + "step": 9760 + }, + { + "epoch": 3.427317415730337, + "grad_norm": 0.9494081139564514, + "learning_rate": 1.9646448399548995e-06, + "loss": 0.5536, + "step": 9761 + }, + { + "epoch": 3.427668539325843, + "grad_norm": 0.9364960193634033, + "learning_rate": 1.9641932486229e-06, + "loss": 0.5732, + "step": 9762 + }, + { + "epoch": 3.4280196629213484, + "grad_norm": 0.936487078666687, + "learning_rate": 1.9637416756150336e-06, + "loss": 0.5646, + "step": 9763 + }, + { + "epoch": 3.428370786516854, + "grad_norm": 0.9187251925468445, + "learning_rate": 1.963290120946742e-06, + "loss": 0.5118, + "step": 9764 + }, + { + "epoch": 3.4287219101123596, + "grad_norm": 0.9177348017692566, + "learning_rate": 1.96283858463347e-06, + "loss": 0.5226, + "step": 9765 + }, + { + "epoch": 3.429073033707865, + "grad_norm": 0.9376488327980042, + "learning_rate": 1.962387066690658e-06, + "loss": 0.5658, + "step": 9766 + }, + { + "epoch": 3.429424157303371, + "grad_norm": 0.9241811633110046, + "learning_rate": 1.961935567133749e-06, + "loss": 0.5929, + "step": 9767 + }, + { + "epoch": 3.4297752808988764, + "grad_norm": 0.9410795569419861, + "learning_rate": 1.9614840859781832e-06, + "loss": 0.5632, + "step": 9768 + }, + { + "epoch": 3.430126404494382, + "grad_norm": 0.9223426580429077, + "learning_rate": 1.9610326232394e-06, + "loss": 0.5754, + "step": 9769 + }, + { + "epoch": 3.4304775280898876, + "grad_norm": 0.9355698227882385, + "learning_rate": 1.960581178932841e-06, + "loss": 0.562, + "step": 9770 + }, + { + "epoch": 3.430828651685393, + "grad_norm": 0.9344855546951294, + "learning_rate": 1.9601297530739425e-06, + "loss": 0.5384, + "step": 9771 + }, + { + "epoch": 3.431179775280899, + "grad_norm": 0.9835062026977539, + "learning_rate": 1.9596783456781458e-06, + "loss": 0.5155, + "step": 9772 + }, + { + "epoch": 3.4315308988764044, + "grad_norm": 0.9497717022895813, + "learning_rate": 1.9592269567608865e-06, + "loss": 0.5659, + "step": 9773 + }, + { + "epoch": 3.43188202247191, + "grad_norm": 0.9333541989326477, + "learning_rate": 1.9587755863376036e-06, + "loss": 0.5203, + "step": 9774 + }, + { + "epoch": 3.4322331460674156, + "grad_norm": 0.9968510866165161, + "learning_rate": 1.958324234423731e-06, + "loss": 0.5719, + "step": 9775 + }, + { + "epoch": 3.432584269662921, + "grad_norm": 0.9084416627883911, + "learning_rate": 1.957872901034708e-06, + "loss": 0.5347, + "step": 9776 + }, + { + "epoch": 3.432935393258427, + "grad_norm": 0.9413123726844788, + "learning_rate": 1.957421586185966e-06, + "loss": 0.622, + "step": 9777 + }, + { + "epoch": 3.433286516853933, + "grad_norm": 0.9193368554115295, + "learning_rate": 1.9569702898929427e-06, + "loss": 0.5746, + "step": 9778 + }, + { + "epoch": 3.4336376404494384, + "grad_norm": 0.9216201305389404, + "learning_rate": 1.9565190121710708e-06, + "loss": 0.562, + "step": 9779 + }, + { + "epoch": 3.433988764044944, + "grad_norm": 0.9366143345832825, + "learning_rate": 1.956067753035783e-06, + "loss": 0.5665, + "step": 9780 + }, + { + "epoch": 3.4343398876404496, + "grad_norm": 0.912437379360199, + "learning_rate": 1.955616512502513e-06, + "loss": 0.5686, + "step": 9781 + }, + { + "epoch": 3.434691011235955, + "grad_norm": 0.9410976767539978, + "learning_rate": 1.9551652905866915e-06, + "loss": 0.5733, + "step": 9782 + }, + { + "epoch": 3.435042134831461, + "grad_norm": 0.9354014992713928, + "learning_rate": 1.9547140873037524e-06, + "loss": 0.5508, + "step": 9783 + }, + { + "epoch": 3.4353932584269664, + "grad_norm": 0.937870442867279, + "learning_rate": 1.9542629026691234e-06, + "loss": 0.518, + "step": 9784 + }, + { + "epoch": 3.435744382022472, + "grad_norm": 0.9242886900901794, + "learning_rate": 1.9538117366982373e-06, + "loss": 0.5555, + "step": 9785 + }, + { + "epoch": 3.4360955056179776, + "grad_norm": 0.9530436396598816, + "learning_rate": 1.9533605894065215e-06, + "loss": 0.5998, + "step": 9786 + }, + { + "epoch": 3.436446629213483, + "grad_norm": 0.9338531494140625, + "learning_rate": 1.952909460809407e-06, + "loss": 0.5573, + "step": 9787 + }, + { + "epoch": 3.436797752808989, + "grad_norm": 0.9284071922302246, + "learning_rate": 1.952458350922319e-06, + "loss": 0.5506, + "step": 9788 + }, + { + "epoch": 3.4371488764044944, + "grad_norm": 0.9072175621986389, + "learning_rate": 1.9520072597606884e-06, + "loss": 0.5192, + "step": 9789 + }, + { + "epoch": 3.4375, + "grad_norm": 0.9227522015571594, + "learning_rate": 1.95155618733994e-06, + "loss": 0.5335, + "step": 9790 + }, + { + "epoch": 3.4378511235955056, + "grad_norm": 0.9971368312835693, + "learning_rate": 1.951105133675501e-06, + "loss": 0.5874, + "step": 9791 + }, + { + "epoch": 3.438202247191011, + "grad_norm": 0.9470699429512024, + "learning_rate": 1.950654098782797e-06, + "loss": 0.5665, + "step": 9792 + }, + { + "epoch": 3.438553370786517, + "grad_norm": 0.9197167754173279, + "learning_rate": 1.9502030826772515e-06, + "loss": 0.5692, + "step": 9793 + }, + { + "epoch": 3.4389044943820224, + "grad_norm": 0.9388170838356018, + "learning_rate": 1.9497520853742913e-06, + "loss": 0.6087, + "step": 9794 + }, + { + "epoch": 3.439255617977528, + "grad_norm": 0.9088274240493774, + "learning_rate": 1.9493011068893377e-06, + "loss": 0.547, + "step": 9795 + }, + { + "epoch": 3.4396067415730336, + "grad_norm": 0.9366891384124756, + "learning_rate": 1.948850147237816e-06, + "loss": 0.4966, + "step": 9796 + }, + { + "epoch": 3.439957865168539, + "grad_norm": 0.9438656568527222, + "learning_rate": 1.9483992064351472e-06, + "loss": 0.5322, + "step": 9797 + }, + { + "epoch": 3.440308988764045, + "grad_norm": 0.942043125629425, + "learning_rate": 1.947948284496754e-06, + "loss": 0.5891, + "step": 9798 + }, + { + "epoch": 3.4406601123595504, + "grad_norm": 0.9567787051200867, + "learning_rate": 1.9474973814380568e-06, + "loss": 0.5704, + "step": 9799 + }, + { + "epoch": 3.441011235955056, + "grad_norm": 0.9638857841491699, + "learning_rate": 1.9470464972744765e-06, + "loss": 0.5841, + "step": 9800 + }, + { + "epoch": 3.4413623595505616, + "grad_norm": 0.9434655904769897, + "learning_rate": 1.946595632021432e-06, + "loss": 0.5717, + "step": 9801 + }, + { + "epoch": 3.441713483146067, + "grad_norm": 1.0056416988372803, + "learning_rate": 1.9461447856943447e-06, + "loss": 0.5294, + "step": 9802 + }, + { + "epoch": 3.442064606741573, + "grad_norm": 0.964290201663971, + "learning_rate": 1.9456939583086306e-06, + "loss": 0.5924, + "step": 9803 + }, + { + "epoch": 3.442415730337079, + "grad_norm": 0.9279812574386597, + "learning_rate": 1.94524314987971e-06, + "loss": 0.5683, + "step": 9804 + }, + { + "epoch": 3.4427668539325844, + "grad_norm": 0.9285330176353455, + "learning_rate": 1.9447923604229983e-06, + "loss": 0.5192, + "step": 9805 + }, + { + "epoch": 3.44311797752809, + "grad_norm": 0.9312847256660461, + "learning_rate": 1.9443415899539126e-06, + "loss": 0.542, + "step": 9806 + }, + { + "epoch": 3.4434691011235956, + "grad_norm": 0.9284462332725525, + "learning_rate": 1.9438908384878695e-06, + "loss": 0.5586, + "step": 9807 + }, + { + "epoch": 3.443820224719101, + "grad_norm": 0.8944636583328247, + "learning_rate": 1.943440106040283e-06, + "loss": 0.5345, + "step": 9808 + }, + { + "epoch": 3.444171348314607, + "grad_norm": 0.9210460782051086, + "learning_rate": 1.9429893926265697e-06, + "loss": 0.5245, + "step": 9809 + }, + { + "epoch": 3.4445224719101124, + "grad_norm": 0.9241941571235657, + "learning_rate": 1.9425386982621415e-06, + "loss": 0.5748, + "step": 9810 + }, + { + "epoch": 3.444873595505618, + "grad_norm": 0.9474250674247742, + "learning_rate": 1.9420880229624136e-06, + "loss": 0.5753, + "step": 9811 + }, + { + "epoch": 3.4452247191011236, + "grad_norm": 0.952357828617096, + "learning_rate": 1.941637366742797e-06, + "loss": 0.5353, + "step": 9812 + }, + { + "epoch": 3.445575842696629, + "grad_norm": 0.909981906414032, + "learning_rate": 1.941186729618706e-06, + "loss": 0.5815, + "step": 9813 + }, + { + "epoch": 3.445926966292135, + "grad_norm": 0.9442864060401917, + "learning_rate": 1.9407361116055495e-06, + "loss": 0.5627, + "step": 9814 + }, + { + "epoch": 3.4462780898876404, + "grad_norm": 0.9455252289772034, + "learning_rate": 1.94028551271874e-06, + "loss": 0.5633, + "step": 9815 + }, + { + "epoch": 3.446629213483146, + "grad_norm": 0.9374313950538635, + "learning_rate": 1.939834932973687e-06, + "loss": 0.5617, + "step": 9816 + }, + { + "epoch": 3.4469803370786516, + "grad_norm": 0.9269933104515076, + "learning_rate": 1.9393843723858e-06, + "loss": 0.547, + "step": 9817 + }, + { + "epoch": 3.447331460674157, + "grad_norm": 0.9133486151695251, + "learning_rate": 1.9389338309704875e-06, + "loss": 0.5604, + "step": 9818 + }, + { + "epoch": 3.447682584269663, + "grad_norm": 0.90470951795578, + "learning_rate": 1.9384833087431572e-06, + "loss": 0.5417, + "step": 9819 + }, + { + "epoch": 3.4480337078651684, + "grad_norm": 0.9282109141349792, + "learning_rate": 1.9380328057192184e-06, + "loss": 0.5541, + "step": 9820 + }, + { + "epoch": 3.448384831460674, + "grad_norm": 0.9550314545631409, + "learning_rate": 1.9375823219140765e-06, + "loss": 0.5991, + "step": 9821 + }, + { + "epoch": 3.44873595505618, + "grad_norm": 0.9559625387191772, + "learning_rate": 1.9371318573431376e-06, + "loss": 0.5524, + "step": 9822 + }, + { + "epoch": 3.4490870786516856, + "grad_norm": 0.9419597387313843, + "learning_rate": 1.9366814120218076e-06, + "loss": 0.5819, + "step": 9823 + }, + { + "epoch": 3.449438202247191, + "grad_norm": 0.9182565212249756, + "learning_rate": 1.9362309859654916e-06, + "loss": 0.5617, + "step": 9824 + }, + { + "epoch": 3.449789325842697, + "grad_norm": 0.9307287335395813, + "learning_rate": 1.9357805791895928e-06, + "loss": 0.5578, + "step": 9825 + }, + { + "epoch": 3.4501404494382024, + "grad_norm": 0.9588905572891235, + "learning_rate": 1.9353301917095157e-06, + "loss": 0.5828, + "step": 9826 + }, + { + "epoch": 3.450491573033708, + "grad_norm": 0.9346449971199036, + "learning_rate": 1.934879823540663e-06, + "loss": 0.5823, + "step": 9827 + }, + { + "epoch": 3.4508426966292136, + "grad_norm": 0.8852512836456299, + "learning_rate": 1.934429474698437e-06, + "loss": 0.5004, + "step": 9828 + }, + { + "epoch": 3.451193820224719, + "grad_norm": 0.9483822584152222, + "learning_rate": 1.933979145198238e-06, + "loss": 0.598, + "step": 9829 + }, + { + "epoch": 3.451544943820225, + "grad_norm": 0.9277448058128357, + "learning_rate": 1.933528835055469e-06, + "loss": 0.582, + "step": 9830 + }, + { + "epoch": 3.4518960674157304, + "grad_norm": 0.9303699135780334, + "learning_rate": 1.9330785442855287e-06, + "loss": 0.5645, + "step": 9831 + }, + { + "epoch": 3.452247191011236, + "grad_norm": 0.9354785680770874, + "learning_rate": 1.932628272903817e-06, + "loss": 0.5572, + "step": 9832 + }, + { + "epoch": 3.4525983146067416, + "grad_norm": 0.9311044812202454, + "learning_rate": 1.932178020925733e-06, + "loss": 0.5579, + "step": 9833 + }, + { + "epoch": 3.452949438202247, + "grad_norm": 0.9231809973716736, + "learning_rate": 1.9317277883666747e-06, + "loss": 0.5514, + "step": 9834 + }, + { + "epoch": 3.453300561797753, + "grad_norm": 0.9001242518424988, + "learning_rate": 1.9312775752420398e-06, + "loss": 0.5488, + "step": 9835 + }, + { + "epoch": 3.4536516853932584, + "grad_norm": 0.9324273467063904, + "learning_rate": 1.930827381567225e-06, + "loss": 0.5469, + "step": 9836 + }, + { + "epoch": 3.454002808988764, + "grad_norm": 0.9363886117935181, + "learning_rate": 1.9303772073576272e-06, + "loss": 0.5763, + "step": 9837 + }, + { + "epoch": 3.4543539325842696, + "grad_norm": 0.9326059818267822, + "learning_rate": 1.9299270526286406e-06, + "loss": 0.5428, + "step": 9838 + }, + { + "epoch": 3.454705056179775, + "grad_norm": 0.9044217467308044, + "learning_rate": 1.9294769173956617e-06, + "loss": 0.5261, + "step": 9839 + }, + { + "epoch": 3.455056179775281, + "grad_norm": 0.9041662216186523, + "learning_rate": 1.9290268016740834e-06, + "loss": 0.5326, + "step": 9840 + }, + { + "epoch": 3.4554073033707864, + "grad_norm": 0.9469613432884216, + "learning_rate": 1.928576705479301e-06, + "loss": 0.574, + "step": 9841 + }, + { + "epoch": 3.455758426966292, + "grad_norm": 0.9463566541671753, + "learning_rate": 1.9281266288267053e-06, + "loss": 0.5365, + "step": 9842 + }, + { + "epoch": 3.4561095505617976, + "grad_norm": 0.9747692942619324, + "learning_rate": 1.92767657173169e-06, + "loss": 0.5906, + "step": 9843 + }, + { + "epoch": 3.456460674157303, + "grad_norm": 0.9493552446365356, + "learning_rate": 1.9272265342096464e-06, + "loss": 0.5504, + "step": 9844 + }, + { + "epoch": 3.456811797752809, + "grad_norm": 0.9679231643676758, + "learning_rate": 1.926776516275964e-06, + "loss": 0.5991, + "step": 9845 + }, + { + "epoch": 3.4571629213483144, + "grad_norm": 0.946408212184906, + "learning_rate": 1.9263265179460355e-06, + "loss": 0.5435, + "step": 9846 + }, + { + "epoch": 3.45751404494382, + "grad_norm": 0.9287437200546265, + "learning_rate": 1.9258765392352483e-06, + "loss": 0.5427, + "step": 9847 + }, + { + "epoch": 3.457865168539326, + "grad_norm": 0.945772647857666, + "learning_rate": 1.925426580158993e-06, + "loss": 0.555, + "step": 9848 + }, + { + "epoch": 3.4582162921348316, + "grad_norm": 0.9223485589027405, + "learning_rate": 1.9249766407326557e-06, + "loss": 0.5558, + "step": 9849 + }, + { + "epoch": 3.458567415730337, + "grad_norm": 0.893747866153717, + "learning_rate": 1.924526720971627e-06, + "loss": 0.5467, + "step": 9850 + }, + { + "epoch": 3.458918539325843, + "grad_norm": 0.9027898907661438, + "learning_rate": 1.9240768208912905e-06, + "loss": 0.5654, + "step": 9851 + }, + { + "epoch": 3.4592696629213484, + "grad_norm": 0.9133523106575012, + "learning_rate": 1.923626940507035e-06, + "loss": 0.5296, + "step": 9852 + }, + { + "epoch": 3.459620786516854, + "grad_norm": 0.9095272421836853, + "learning_rate": 1.923177079834244e-06, + "loss": 0.5451, + "step": 9853 + }, + { + "epoch": 3.4599719101123596, + "grad_norm": 0.9308523535728455, + "learning_rate": 1.9227272388883044e-06, + "loss": 0.5336, + "step": 9854 + }, + { + "epoch": 3.460323033707865, + "grad_norm": 0.9367560148239136, + "learning_rate": 1.9222774176845986e-06, + "loss": 0.5535, + "step": 9855 + }, + { + "epoch": 3.460674157303371, + "grad_norm": 0.9739148616790771, + "learning_rate": 1.9218276162385103e-06, + "loss": 0.5611, + "step": 9856 + }, + { + "epoch": 3.4610252808988764, + "grad_norm": 0.9131298661231995, + "learning_rate": 1.921377834565424e-06, + "loss": 0.5135, + "step": 9857 + }, + { + "epoch": 3.461376404494382, + "grad_norm": 0.9352442026138306, + "learning_rate": 1.9209280726807187e-06, + "loss": 0.5381, + "step": 9858 + }, + { + "epoch": 3.4617275280898876, + "grad_norm": 0.9587167501449585, + "learning_rate": 1.92047833059978e-06, + "loss": 0.6049, + "step": 9859 + }, + { + "epoch": 3.462078651685393, + "grad_norm": 0.9460211396217346, + "learning_rate": 1.9200286083379846e-06, + "loss": 0.5376, + "step": 9860 + }, + { + "epoch": 3.462429775280899, + "grad_norm": 0.9632991552352905, + "learning_rate": 1.9195789059107157e-06, + "loss": 0.5709, + "step": 9861 + }, + { + "epoch": 3.4627808988764044, + "grad_norm": 0.9615373015403748, + "learning_rate": 1.9191292233333504e-06, + "loss": 0.587, + "step": 9862 + }, + { + "epoch": 3.46313202247191, + "grad_norm": 0.9179255366325378, + "learning_rate": 1.9186795606212698e-06, + "loss": 0.5658, + "step": 9863 + }, + { + "epoch": 3.4634831460674156, + "grad_norm": 0.9290973544120789, + "learning_rate": 1.91822991778985e-06, + "loss": 0.5365, + "step": 9864 + }, + { + "epoch": 3.463834269662921, + "grad_norm": 0.9433035850524902, + "learning_rate": 1.91778029485447e-06, + "loss": 0.5555, + "step": 9865 + }, + { + "epoch": 3.464185393258427, + "grad_norm": 0.9535524845123291, + "learning_rate": 1.9173306918305047e-06, + "loss": 0.5897, + "step": 9866 + }, + { + "epoch": 3.464536516853933, + "grad_norm": 0.90745609998703, + "learning_rate": 1.9168811087333322e-06, + "loss": 0.5419, + "step": 9867 + }, + { + "epoch": 3.4648876404494384, + "grad_norm": 0.9422712326049805, + "learning_rate": 1.9164315455783265e-06, + "loss": 0.5662, + "step": 9868 + }, + { + "epoch": 3.465238764044944, + "grad_norm": 0.9179876446723938, + "learning_rate": 1.915982002380862e-06, + "loss": 0.5436, + "step": 9869 + }, + { + "epoch": 3.4655898876404496, + "grad_norm": 0.9426119923591614, + "learning_rate": 1.915532479156314e-06, + "loss": 0.5693, + "step": 9870 + }, + { + "epoch": 3.465941011235955, + "grad_norm": 0.899300217628479, + "learning_rate": 1.9150829759200542e-06, + "loss": 0.5181, + "step": 9871 + }, + { + "epoch": 3.466292134831461, + "grad_norm": 0.925981342792511, + "learning_rate": 1.9146334926874572e-06, + "loss": 0.5717, + "step": 9872 + }, + { + "epoch": 3.4666432584269664, + "grad_norm": 0.9280354976654053, + "learning_rate": 1.9141840294738927e-06, + "loss": 0.5675, + "step": 9873 + }, + { + "epoch": 3.466994382022472, + "grad_norm": 0.947169840335846, + "learning_rate": 1.913734586294734e-06, + "loss": 0.5494, + "step": 9874 + }, + { + "epoch": 3.4673455056179776, + "grad_norm": 0.925955593585968, + "learning_rate": 1.9132851631653498e-06, + "loss": 0.5432, + "step": 9875 + }, + { + "epoch": 3.467696629213483, + "grad_norm": 0.9399617910385132, + "learning_rate": 1.9128357601011123e-06, + "loss": 0.5236, + "step": 9876 + }, + { + "epoch": 3.468047752808989, + "grad_norm": 0.9419747591018677, + "learning_rate": 1.912386377117388e-06, + "loss": 0.5704, + "step": 9877 + }, + { + "epoch": 3.4683988764044944, + "grad_norm": 0.9297751188278198, + "learning_rate": 1.911937014229548e-06, + "loss": 0.5253, + "step": 9878 + }, + { + "epoch": 3.46875, + "grad_norm": 0.9418967962265015, + "learning_rate": 1.9114876714529577e-06, + "loss": 0.5659, + "step": 9879 + }, + { + "epoch": 3.4691011235955056, + "grad_norm": 0.9735715985298157, + "learning_rate": 1.9110383488029866e-06, + "loss": 0.591, + "step": 9880 + }, + { + "epoch": 3.469452247191011, + "grad_norm": 0.9207716584205627, + "learning_rate": 1.910589046295e-06, + "loss": 0.4944, + "step": 9881 + }, + { + "epoch": 3.469803370786517, + "grad_norm": 0.9339351654052734, + "learning_rate": 1.910139763944362e-06, + "loss": 0.5849, + "step": 9882 + }, + { + "epoch": 3.4701544943820224, + "grad_norm": 0.9344291687011719, + "learning_rate": 1.9096905017664414e-06, + "loss": 0.5618, + "step": 9883 + }, + { + "epoch": 3.470505617977528, + "grad_norm": 0.951622486114502, + "learning_rate": 1.9092412597765985e-06, + "loss": 0.5311, + "step": 9884 + }, + { + "epoch": 3.4708567415730336, + "grad_norm": 0.9389412999153137, + "learning_rate": 1.908792037990201e-06, + "loss": 0.5612, + "step": 9885 + }, + { + "epoch": 3.471207865168539, + "grad_norm": 0.9599605798721313, + "learning_rate": 1.908342836422608e-06, + "loss": 0.562, + "step": 9886 + }, + { + "epoch": 3.471558988764045, + "grad_norm": 0.9610913991928101, + "learning_rate": 1.9078936550891853e-06, + "loss": 0.5715, + "step": 9887 + }, + { + "epoch": 3.4719101123595504, + "grad_norm": 0.9503083825111389, + "learning_rate": 1.907444494005292e-06, + "loss": 0.5471, + "step": 9888 + }, + { + "epoch": 3.472261235955056, + "grad_norm": 0.9268511533737183, + "learning_rate": 1.9069953531862904e-06, + "loss": 0.5429, + "step": 9889 + }, + { + "epoch": 3.4726123595505616, + "grad_norm": 0.9469237923622131, + "learning_rate": 1.9065462326475392e-06, + "loss": 0.5795, + "step": 9890 + }, + { + "epoch": 3.472963483146067, + "grad_norm": 0.9078359603881836, + "learning_rate": 1.9060971324044007e-06, + "loss": 0.5394, + "step": 9891 + }, + { + "epoch": 3.473314606741573, + "grad_norm": 0.9630365371704102, + "learning_rate": 1.9056480524722304e-06, + "loss": 0.5897, + "step": 9892 + }, + { + "epoch": 3.473665730337079, + "grad_norm": 0.9337735772132874, + "learning_rate": 1.9051989928663895e-06, + "loss": 0.5672, + "step": 9893 + }, + { + "epoch": 3.4740168539325844, + "grad_norm": 0.9678616523742676, + "learning_rate": 1.904749953602234e-06, + "loss": 0.5881, + "step": 9894 + }, + { + "epoch": 3.47436797752809, + "grad_norm": 0.9363431930541992, + "learning_rate": 1.9043009346951196e-06, + "loss": 0.5627, + "step": 9895 + }, + { + "epoch": 3.4747191011235956, + "grad_norm": 0.9697135090827942, + "learning_rate": 1.9038519361604046e-06, + "loss": 0.59, + "step": 9896 + }, + { + "epoch": 3.475070224719101, + "grad_norm": 0.9170849323272705, + "learning_rate": 1.9034029580134424e-06, + "loss": 0.541, + "step": 9897 + }, + { + "epoch": 3.475421348314607, + "grad_norm": 0.9574970602989197, + "learning_rate": 1.9029540002695894e-06, + "loss": 0.5626, + "step": 9898 + }, + { + "epoch": 3.4757724719101124, + "grad_norm": 0.9184151887893677, + "learning_rate": 1.9025050629441977e-06, + "loss": 0.5887, + "step": 9899 + }, + { + "epoch": 3.476123595505618, + "grad_norm": 0.9465662837028503, + "learning_rate": 1.9020561460526225e-06, + "loss": 0.5505, + "step": 9900 + }, + { + "epoch": 3.4764747191011236, + "grad_norm": 0.8988319039344788, + "learning_rate": 1.901607249610215e-06, + "loss": 0.5405, + "step": 9901 + }, + { + "epoch": 3.476825842696629, + "grad_norm": 0.9421148300170898, + "learning_rate": 1.9011583736323274e-06, + "loss": 0.56, + "step": 9902 + }, + { + "epoch": 3.477176966292135, + "grad_norm": 0.9460734724998474, + "learning_rate": 1.9007095181343111e-06, + "loss": 0.5834, + "step": 9903 + }, + { + "epoch": 3.4775280898876404, + "grad_norm": 0.9705817699432373, + "learning_rate": 1.9002606831315163e-06, + "loss": 0.578, + "step": 9904 + }, + { + "epoch": 3.477879213483146, + "grad_norm": 0.9191238284111023, + "learning_rate": 1.8998118686392935e-06, + "loss": 0.5687, + "step": 9905 + }, + { + "epoch": 3.4782303370786516, + "grad_norm": 1.1123462915420532, + "learning_rate": 1.8993630746729904e-06, + "loss": 0.5528, + "step": 9906 + }, + { + "epoch": 3.478581460674157, + "grad_norm": 0.9411060214042664, + "learning_rate": 1.898914301247957e-06, + "loss": 0.5363, + "step": 9907 + }, + { + "epoch": 3.478932584269663, + "grad_norm": 0.9148615598678589, + "learning_rate": 1.8984655483795397e-06, + "loss": 0.5327, + "step": 9908 + }, + { + "epoch": 3.4792837078651684, + "grad_norm": 0.92915278673172, + "learning_rate": 1.8980168160830858e-06, + "loss": 0.5609, + "step": 9909 + }, + { + "epoch": 3.479634831460674, + "grad_norm": 0.9737476706504822, + "learning_rate": 1.897568104373942e-06, + "loss": 0.5485, + "step": 9910 + }, + { + "epoch": 3.47998595505618, + "grad_norm": 0.932253897190094, + "learning_rate": 1.897119413267453e-06, + "loss": 0.5755, + "step": 9911 + }, + { + "epoch": 3.4803370786516856, + "grad_norm": 0.9012002944946289, + "learning_rate": 1.8966707427789647e-06, + "loss": 0.5178, + "step": 9912 + }, + { + "epoch": 3.480688202247191, + "grad_norm": 0.9278724193572998, + "learning_rate": 1.8962220929238203e-06, + "loss": 0.5486, + "step": 9913 + }, + { + "epoch": 3.481039325842697, + "grad_norm": 0.951991856098175, + "learning_rate": 1.895773463717364e-06, + "loss": 0.5823, + "step": 9914 + }, + { + "epoch": 3.4813904494382024, + "grad_norm": 0.9457383751869202, + "learning_rate": 1.8953248551749385e-06, + "loss": 0.5857, + "step": 9915 + }, + { + "epoch": 3.481741573033708, + "grad_norm": 0.9369385838508606, + "learning_rate": 1.8948762673118841e-06, + "loss": 0.5295, + "step": 9916 + }, + { + "epoch": 3.4820926966292136, + "grad_norm": 0.9311235547065735, + "learning_rate": 1.8944277001435451e-06, + "loss": 0.5465, + "step": 9917 + }, + { + "epoch": 3.482443820224719, + "grad_norm": 0.9539017081260681, + "learning_rate": 1.8939791536852603e-06, + "loss": 0.58, + "step": 9918 + }, + { + "epoch": 3.482794943820225, + "grad_norm": 0.9526887536048889, + "learning_rate": 1.8935306279523693e-06, + "loss": 0.5636, + "step": 9919 + }, + { + "epoch": 3.4831460674157304, + "grad_norm": 0.9262377619743347, + "learning_rate": 1.893082122960212e-06, + "loss": 0.5635, + "step": 9920 + }, + { + "epoch": 3.483497191011236, + "grad_norm": 0.9161630868911743, + "learning_rate": 1.8926336387241271e-06, + "loss": 0.542, + "step": 9921 + }, + { + "epoch": 3.4838483146067416, + "grad_norm": 0.9114150404930115, + "learning_rate": 1.8921851752594523e-06, + "loss": 0.5537, + "step": 9922 + }, + { + "epoch": 3.484199438202247, + "grad_norm": 0.9345464110374451, + "learning_rate": 1.8917367325815231e-06, + "loss": 0.5876, + "step": 9923 + }, + { + "epoch": 3.484550561797753, + "grad_norm": 0.9176331162452698, + "learning_rate": 1.8912883107056783e-06, + "loss": 0.5495, + "step": 9924 + }, + { + "epoch": 3.4849016853932584, + "grad_norm": 0.9213263392448425, + "learning_rate": 1.8908399096472515e-06, + "loss": 0.5484, + "step": 9925 + }, + { + "epoch": 3.485252808988764, + "grad_norm": 0.9535937905311584, + "learning_rate": 1.8903915294215797e-06, + "loss": 0.548, + "step": 9926 + }, + { + "epoch": 3.4856039325842696, + "grad_norm": 0.9212021827697754, + "learning_rate": 1.8899431700439947e-06, + "loss": 0.5483, + "step": 9927 + }, + { + "epoch": 3.485955056179775, + "grad_norm": 0.986203670501709, + "learning_rate": 1.8894948315298323e-06, + "loss": 0.5614, + "step": 9928 + }, + { + "epoch": 3.486306179775281, + "grad_norm": 1.0019465684890747, + "learning_rate": 1.889046513894423e-06, + "loss": 0.5188, + "step": 9929 + }, + { + "epoch": 3.4866573033707864, + "grad_norm": 0.938369631767273, + "learning_rate": 1.8885982171531014e-06, + "loss": 0.5604, + "step": 9930 + }, + { + "epoch": 3.487008426966292, + "grad_norm": 0.9510270953178406, + "learning_rate": 1.8881499413211978e-06, + "loss": 0.5638, + "step": 9931 + }, + { + "epoch": 3.4873595505617976, + "grad_norm": 0.9275785088539124, + "learning_rate": 1.8877016864140413e-06, + "loss": 0.548, + "step": 9932 + }, + { + "epoch": 3.487710674157303, + "grad_norm": 0.9248453974723816, + "learning_rate": 1.8872534524469643e-06, + "loss": 0.5451, + "step": 9933 + }, + { + "epoch": 3.488061797752809, + "grad_norm": 0.9249113202095032, + "learning_rate": 1.8868052394352938e-06, + "loss": 0.5528, + "step": 9934 + }, + { + "epoch": 3.4884129213483144, + "grad_norm": 1.0094629526138306, + "learning_rate": 1.8863570473943605e-06, + "loss": 0.577, + "step": 9935 + }, + { + "epoch": 3.48876404494382, + "grad_norm": 0.9469836354255676, + "learning_rate": 1.88590887633949e-06, + "loss": 0.5676, + "step": 9936 + }, + { + "epoch": 3.489115168539326, + "grad_norm": 0.9471365809440613, + "learning_rate": 1.8854607262860113e-06, + "loss": 0.5765, + "step": 9937 + }, + { + "epoch": 3.4894662921348316, + "grad_norm": 0.9582838416099548, + "learning_rate": 1.8850125972492492e-06, + "loss": 0.5577, + "step": 9938 + }, + { + "epoch": 3.489817415730337, + "grad_norm": 0.9376549124717712, + "learning_rate": 1.8845644892445307e-06, + "loss": 0.5608, + "step": 9939 + }, + { + "epoch": 3.490168539325843, + "grad_norm": 0.9560538530349731, + "learning_rate": 1.8841164022871794e-06, + "loss": 0.5665, + "step": 9940 + }, + { + "epoch": 3.4905196629213484, + "grad_norm": 0.9541363716125488, + "learning_rate": 1.883668336392521e-06, + "loss": 0.5678, + "step": 9941 + }, + { + "epoch": 3.490870786516854, + "grad_norm": 0.9529964327812195, + "learning_rate": 1.8832202915758768e-06, + "loss": 0.5582, + "step": 9942 + }, + { + "epoch": 3.4912219101123596, + "grad_norm": 0.9583566188812256, + "learning_rate": 1.8827722678525722e-06, + "loss": 0.5832, + "step": 9943 + }, + { + "epoch": 3.491573033707865, + "grad_norm": 0.915770947933197, + "learning_rate": 1.8823242652379277e-06, + "loss": 0.4983, + "step": 9944 + }, + { + "epoch": 3.491924157303371, + "grad_norm": 0.9760103225708008, + "learning_rate": 1.881876283747264e-06, + "loss": 0.5823, + "step": 9945 + }, + { + "epoch": 3.4922752808988764, + "grad_norm": 0.9490340352058411, + "learning_rate": 1.8814283233959032e-06, + "loss": 0.5517, + "step": 9946 + }, + { + "epoch": 3.492626404494382, + "grad_norm": 0.9550537467002869, + "learning_rate": 1.8809803841991636e-06, + "loss": 0.5618, + "step": 9947 + }, + { + "epoch": 3.4929775280898876, + "grad_norm": 0.9249680638313293, + "learning_rate": 1.8805324661723662e-06, + "loss": 0.5497, + "step": 9948 + }, + { + "epoch": 3.493328651685393, + "grad_norm": 0.9591121077537537, + "learning_rate": 1.880084569330827e-06, + "loss": 0.5641, + "step": 9949 + }, + { + "epoch": 3.493679775280899, + "grad_norm": 0.9434618353843689, + "learning_rate": 1.8796366936898664e-06, + "loss": 0.5817, + "step": 9950 + }, + { + "epoch": 3.4940308988764044, + "grad_norm": 0.8885144591331482, + "learning_rate": 1.8791888392647988e-06, + "loss": 0.5116, + "step": 9951 + }, + { + "epoch": 3.49438202247191, + "grad_norm": 0.9607181549072266, + "learning_rate": 1.878741006070943e-06, + "loss": 0.5801, + "step": 9952 + }, + { + "epoch": 3.4947331460674156, + "grad_norm": 0.9534644484519958, + "learning_rate": 1.878293194123612e-06, + "loss": 0.5877, + "step": 9953 + }, + { + "epoch": 3.495084269662921, + "grad_norm": 0.93828946352005, + "learning_rate": 1.8778454034381224e-06, + "loss": 0.556, + "step": 9954 + }, + { + "epoch": 3.495435393258427, + "grad_norm": 0.9643175005912781, + "learning_rate": 1.8773976340297869e-06, + "loss": 0.5508, + "step": 9955 + }, + { + "epoch": 3.495786516853933, + "grad_norm": 0.9178944826126099, + "learning_rate": 1.8769498859139206e-06, + "loss": 0.5258, + "step": 9956 + }, + { + "epoch": 3.4961376404494384, + "grad_norm": 0.9248856902122498, + "learning_rate": 1.8765021591058351e-06, + "loss": 0.5598, + "step": 9957 + }, + { + "epoch": 3.496488764044944, + "grad_norm": 0.9261285662651062, + "learning_rate": 1.876054453620841e-06, + "loss": 0.5619, + "step": 9958 + }, + { + "epoch": 3.4968398876404496, + "grad_norm": 0.9073474407196045, + "learning_rate": 1.8756067694742518e-06, + "loss": 0.5221, + "step": 9959 + }, + { + "epoch": 3.497191011235955, + "grad_norm": 0.9330862164497375, + "learning_rate": 1.875159106681376e-06, + "loss": 0.5701, + "step": 9960 + }, + { + "epoch": 3.497542134831461, + "grad_norm": 0.9415444731712341, + "learning_rate": 1.874711465257525e-06, + "loss": 0.5351, + "step": 9961 + }, + { + "epoch": 3.4978932584269664, + "grad_norm": 0.9090070128440857, + "learning_rate": 1.8742638452180056e-06, + "loss": 0.543, + "step": 9962 + }, + { + "epoch": 3.498244382022472, + "grad_norm": 0.9320966005325317, + "learning_rate": 1.8738162465781284e-06, + "loss": 0.5363, + "step": 9963 + }, + { + "epoch": 3.4985955056179776, + "grad_norm": 0.9806389212608337, + "learning_rate": 1.8733686693531986e-06, + "loss": 0.6212, + "step": 9964 + }, + { + "epoch": 3.498946629213483, + "grad_norm": 0.908397376537323, + "learning_rate": 1.8729211135585252e-06, + "loss": 0.532, + "step": 9965 + }, + { + "epoch": 3.499297752808989, + "grad_norm": 0.966843843460083, + "learning_rate": 1.8724735792094118e-06, + "loss": 0.5479, + "step": 9966 + }, + { + "epoch": 3.4996488764044944, + "grad_norm": 0.9645453095436096, + "learning_rate": 1.8720260663211665e-06, + "loss": 0.545, + "step": 9967 + }, + { + "epoch": 3.5, + "grad_norm": 0.9507824778556824, + "learning_rate": 1.871578574909092e-06, + "loss": 0.5788, + "step": 9968 + }, + { + "epoch": 3.5003511235955056, + "grad_norm": 0.9252168536186218, + "learning_rate": 1.8711311049884916e-06, + "loss": 0.5605, + "step": 9969 + }, + { + "epoch": 3.500702247191011, + "grad_norm": 0.952176034450531, + "learning_rate": 1.8706836565746705e-06, + "loss": 0.5683, + "step": 9970 + }, + { + "epoch": 3.501053370786517, + "grad_norm": 0.9316127896308899, + "learning_rate": 1.8702362296829285e-06, + "loss": 0.5455, + "step": 9971 + }, + { + "epoch": 3.5014044943820224, + "grad_norm": 0.9359353184700012, + "learning_rate": 1.86978882432857e-06, + "loss": 0.6019, + "step": 9972 + }, + { + "epoch": 3.501755617977528, + "grad_norm": 0.9286577105522156, + "learning_rate": 1.8693414405268928e-06, + "loss": 0.5571, + "step": 9973 + }, + { + "epoch": 3.5021067415730336, + "grad_norm": 0.9131038188934326, + "learning_rate": 1.8688940782932002e-06, + "loss": 0.5656, + "step": 9974 + }, + { + "epoch": 3.502457865168539, + "grad_norm": 0.9286239743232727, + "learning_rate": 1.868446737642789e-06, + "loss": 0.5459, + "step": 9975 + }, + { + "epoch": 3.502808988764045, + "grad_norm": 0.9289954304695129, + "learning_rate": 1.8679994185909598e-06, + "loss": 0.5158, + "step": 9976 + }, + { + "epoch": 3.5031601123595504, + "grad_norm": 0.9234467148780823, + "learning_rate": 1.8675521211530084e-06, + "loss": 0.5729, + "step": 9977 + }, + { + "epoch": 3.503511235955056, + "grad_norm": 0.9389896988868713, + "learning_rate": 1.8671048453442348e-06, + "loss": 0.5197, + "step": 9978 + }, + { + "epoch": 3.5038623595505616, + "grad_norm": 0.9272573590278625, + "learning_rate": 1.8666575911799325e-06, + "loss": 0.5359, + "step": 9979 + }, + { + "epoch": 3.504213483146067, + "grad_norm": 0.9517552256584167, + "learning_rate": 1.8662103586753998e-06, + "loss": 0.5428, + "step": 9980 + }, + { + "epoch": 3.5045646067415728, + "grad_norm": 0.9344884753227234, + "learning_rate": 1.8657631478459306e-06, + "loss": 0.5604, + "step": 9981 + }, + { + "epoch": 3.504915730337079, + "grad_norm": 0.9385592937469482, + "learning_rate": 1.865315958706818e-06, + "loss": 0.5478, + "step": 9982 + }, + { + "epoch": 3.5052668539325844, + "grad_norm": 0.9211937785148621, + "learning_rate": 1.8648687912733573e-06, + "loss": 0.5662, + "step": 9983 + }, + { + "epoch": 3.50561797752809, + "grad_norm": 0.9594284892082214, + "learning_rate": 1.8644216455608393e-06, + "loss": 0.5557, + "step": 9984 + }, + { + "epoch": 3.5059691011235956, + "grad_norm": 0.9408329725265503, + "learning_rate": 1.8639745215845587e-06, + "loss": 0.563, + "step": 9985 + }, + { + "epoch": 3.506320224719101, + "grad_norm": 0.9596962332725525, + "learning_rate": 1.8635274193598038e-06, + "loss": 0.5814, + "step": 9986 + }, + { + "epoch": 3.506671348314607, + "grad_norm": 0.9330762028694153, + "learning_rate": 1.8630803389018676e-06, + "loss": 0.5722, + "step": 9987 + }, + { + "epoch": 3.5070224719101124, + "grad_norm": 0.9546787142753601, + "learning_rate": 1.8626332802260383e-06, + "loss": 0.5582, + "step": 9988 + }, + { + "epoch": 3.507373595505618, + "grad_norm": 0.9448786377906799, + "learning_rate": 1.8621862433476057e-06, + "loss": 0.559, + "step": 9989 + }, + { + "epoch": 3.5077247191011236, + "grad_norm": 0.9394882917404175, + "learning_rate": 1.8617392282818575e-06, + "loss": 0.533, + "step": 9990 + }, + { + "epoch": 3.508075842696629, + "grad_norm": 0.9193337559700012, + "learning_rate": 1.8612922350440815e-06, + "loss": 0.5491, + "step": 9991 + }, + { + "epoch": 3.508426966292135, + "grad_norm": 0.9309750199317932, + "learning_rate": 1.8608452636495644e-06, + "loss": 0.5703, + "step": 9992 + }, + { + "epoch": 3.5087780898876404, + "grad_norm": 0.9274532198905945, + "learning_rate": 1.8603983141135929e-06, + "loss": 0.5576, + "step": 9993 + }, + { + "epoch": 3.509129213483146, + "grad_norm": 0.9404646158218384, + "learning_rate": 1.8599513864514519e-06, + "loss": 0.5439, + "step": 9994 + }, + { + "epoch": 3.5094803370786516, + "grad_norm": 0.9496977925300598, + "learning_rate": 1.8595044806784252e-06, + "loss": 0.5712, + "step": 9995 + }, + { + "epoch": 3.509831460674157, + "grad_norm": 0.9119061827659607, + "learning_rate": 1.8590575968097968e-06, + "loss": 0.5216, + "step": 9996 + }, + { + "epoch": 3.510182584269663, + "grad_norm": 0.9275595545768738, + "learning_rate": 1.8586107348608507e-06, + "loss": 0.5473, + "step": 9997 + }, + { + "epoch": 3.5105337078651684, + "grad_norm": 0.9187131524085999, + "learning_rate": 1.8581638948468684e-06, + "loss": 0.528, + "step": 9998 + }, + { + "epoch": 3.5108848314606744, + "grad_norm": 0.9263707995414734, + "learning_rate": 1.857717076783132e-06, + "loss": 0.549, + "step": 9999 + }, + { + "epoch": 3.51123595505618, + "grad_norm": 0.9314154386520386, + "learning_rate": 1.8572702806849216e-06, + "loss": 0.5803, + "step": 10000 + }, + { + "epoch": 3.5115870786516856, + "grad_norm": 0.9460896253585815, + "learning_rate": 1.8568235065675177e-06, + "loss": 0.5562, + "step": 10001 + }, + { + "epoch": 3.511938202247191, + "grad_norm": 0.9514849781990051, + "learning_rate": 1.8563767544462003e-06, + "loss": 0.5936, + "step": 10002 + }, + { + "epoch": 3.512289325842697, + "grad_norm": 0.9125812649726868, + "learning_rate": 1.8559300243362456e-06, + "loss": 0.5374, + "step": 10003 + }, + { + "epoch": 3.5126404494382024, + "grad_norm": 0.9279007315635681, + "learning_rate": 1.8554833162529343e-06, + "loss": 0.5278, + "step": 10004 + }, + { + "epoch": 3.512991573033708, + "grad_norm": 0.9265901446342468, + "learning_rate": 1.8550366302115409e-06, + "loss": 0.5839, + "step": 10005 + }, + { + "epoch": 3.5133426966292136, + "grad_norm": 0.9556569457054138, + "learning_rate": 1.8545899662273437e-06, + "loss": 0.5498, + "step": 10006 + }, + { + "epoch": 3.513693820224719, + "grad_norm": 0.9076559543609619, + "learning_rate": 1.854143324315617e-06, + "loss": 0.5533, + "step": 10007 + }, + { + "epoch": 3.514044943820225, + "grad_norm": 0.9030308723449707, + "learning_rate": 1.8536967044916362e-06, + "loss": 0.5264, + "step": 10008 + }, + { + "epoch": 3.5143960674157304, + "grad_norm": 0.9511495232582092, + "learning_rate": 1.8532501067706754e-06, + "loss": 0.5476, + "step": 10009 + }, + { + "epoch": 3.514747191011236, + "grad_norm": 0.9589691758155823, + "learning_rate": 1.8528035311680065e-06, + "loss": 0.5825, + "step": 10010 + }, + { + "epoch": 3.5150983146067416, + "grad_norm": 0.9432084560394287, + "learning_rate": 1.8523569776989042e-06, + "loss": 0.5593, + "step": 10011 + }, + { + "epoch": 3.515449438202247, + "grad_norm": 0.9391931891441345, + "learning_rate": 1.8519104463786375e-06, + "loss": 0.5554, + "step": 10012 + }, + { + "epoch": 3.515800561797753, + "grad_norm": 0.9526146054267883, + "learning_rate": 1.8514639372224805e-06, + "loss": 0.5552, + "step": 10013 + }, + { + "epoch": 3.5161516853932584, + "grad_norm": 0.9300878643989563, + "learning_rate": 1.8510174502457007e-06, + "loss": 0.5515, + "step": 10014 + }, + { + "epoch": 3.516502808988764, + "grad_norm": 0.9219534993171692, + "learning_rate": 1.8505709854635697e-06, + "loss": 0.5402, + "step": 10015 + }, + { + "epoch": 3.5168539325842696, + "grad_norm": 0.948150634765625, + "learning_rate": 1.8501245428913545e-06, + "loss": 0.5544, + "step": 10016 + }, + { + "epoch": 3.517205056179775, + "grad_norm": 0.9621832370758057, + "learning_rate": 1.8496781225443244e-06, + "loss": 0.5778, + "step": 10017 + }, + { + "epoch": 3.517556179775281, + "grad_norm": 0.9471445083618164, + "learning_rate": 1.8492317244377455e-06, + "loss": 0.576, + "step": 10018 + }, + { + "epoch": 3.5179073033707864, + "grad_norm": 0.9674230813980103, + "learning_rate": 1.8487853485868856e-06, + "loss": 0.5836, + "step": 10019 + }, + { + "epoch": 3.518258426966292, + "grad_norm": 0.9415154457092285, + "learning_rate": 1.8483389950070097e-06, + "loss": 0.5822, + "step": 10020 + }, + { + "epoch": 3.5186095505617976, + "grad_norm": 0.9395483136177063, + "learning_rate": 1.847892663713381e-06, + "loss": 0.557, + "step": 10021 + }, + { + "epoch": 3.518960674157303, + "grad_norm": 0.9537492394447327, + "learning_rate": 1.8474463547212664e-06, + "loss": 0.5517, + "step": 10022 + }, + { + "epoch": 3.519311797752809, + "grad_norm": 0.9584711194038391, + "learning_rate": 1.847000068045927e-06, + "loss": 0.5701, + "step": 10023 + }, + { + "epoch": 3.5196629213483144, + "grad_norm": 0.9907458424568176, + "learning_rate": 1.8465538037026279e-06, + "loss": 0.5655, + "step": 10024 + }, + { + "epoch": 3.52001404494382, + "grad_norm": 0.906841516494751, + "learning_rate": 1.8461075617066281e-06, + "loss": 0.5438, + "step": 10025 + }, + { + "epoch": 3.5203651685393256, + "grad_norm": 0.9018142819404602, + "learning_rate": 1.8456613420731916e-06, + "loss": 0.5395, + "step": 10026 + }, + { + "epoch": 3.5207162921348316, + "grad_norm": 0.9500654935836792, + "learning_rate": 1.845215144817576e-06, + "loss": 0.5546, + "step": 10027 + }, + { + "epoch": 3.521067415730337, + "grad_norm": 0.965528130531311, + "learning_rate": 1.844768969955043e-06, + "loss": 0.5851, + "step": 10028 + }, + { + "epoch": 3.521418539325843, + "grad_norm": 0.943034827709198, + "learning_rate": 1.8443228175008499e-06, + "loss": 0.5925, + "step": 10029 + }, + { + "epoch": 3.5217696629213484, + "grad_norm": 0.9575608372688293, + "learning_rate": 1.8438766874702562e-06, + "loss": 0.5623, + "step": 10030 + }, + { + "epoch": 3.522120786516854, + "grad_norm": 0.9183236360549927, + "learning_rate": 1.8434305798785185e-06, + "loss": 0.5343, + "step": 10031 + }, + { + "epoch": 3.5224719101123596, + "grad_norm": 0.9550362229347229, + "learning_rate": 1.8429844947408917e-06, + "loss": 0.581, + "step": 10032 + }, + { + "epoch": 3.522823033707865, + "grad_norm": 0.9401598572731018, + "learning_rate": 1.8425384320726341e-06, + "loss": 0.5622, + "step": 10033 + }, + { + "epoch": 3.523174157303371, + "grad_norm": 0.9148021340370178, + "learning_rate": 1.8420923918889987e-06, + "loss": 0.5378, + "step": 10034 + }, + { + "epoch": 3.5235252808988764, + "grad_norm": 0.9355801939964294, + "learning_rate": 1.8416463742052415e-06, + "loss": 0.5427, + "step": 10035 + }, + { + "epoch": 3.523876404494382, + "grad_norm": 0.9740668535232544, + "learning_rate": 1.8412003790366139e-06, + "loss": 0.6183, + "step": 10036 + }, + { + "epoch": 3.5242275280898876, + "grad_norm": 0.9630128145217896, + "learning_rate": 1.8407544063983706e-06, + "loss": 0.5671, + "step": 10037 + }, + { + "epoch": 3.524578651685393, + "grad_norm": 0.956880509853363, + "learning_rate": 1.8403084563057617e-06, + "loss": 0.5619, + "step": 10038 + }, + { + "epoch": 3.524929775280899, + "grad_norm": 0.9559294581413269, + "learning_rate": 1.8398625287740396e-06, + "loss": 0.5675, + "step": 10039 + }, + { + "epoch": 3.5252808988764044, + "grad_norm": 0.9689837098121643, + "learning_rate": 1.8394166238184536e-06, + "loss": 0.5698, + "step": 10040 + }, + { + "epoch": 3.52563202247191, + "grad_norm": 0.9151616096496582, + "learning_rate": 1.8389707414542546e-06, + "loss": 0.5389, + "step": 10041 + }, + { + "epoch": 3.5259831460674156, + "grad_norm": 0.9492166042327881, + "learning_rate": 1.8385248816966895e-06, + "loss": 0.5513, + "step": 10042 + }, + { + "epoch": 3.526334269662921, + "grad_norm": 0.9414293169975281, + "learning_rate": 1.8380790445610084e-06, + "loss": 0.5302, + "step": 10043 + }, + { + "epoch": 3.5266853932584272, + "grad_norm": 0.9141148924827576, + "learning_rate": 1.8376332300624577e-06, + "loss": 0.5556, + "step": 10044 + }, + { + "epoch": 3.527036516853933, + "grad_norm": 0.9246927499771118, + "learning_rate": 1.8371874382162824e-06, + "loss": 0.561, + "step": 10045 + }, + { + "epoch": 3.5273876404494384, + "grad_norm": 0.9312998652458191, + "learning_rate": 1.836741669037731e-06, + "loss": 0.537, + "step": 10046 + }, + { + "epoch": 3.527738764044944, + "grad_norm": 0.932227373123169, + "learning_rate": 1.8362959225420457e-06, + "loss": 0.5651, + "step": 10047 + }, + { + "epoch": 3.5280898876404496, + "grad_norm": 0.9300662875175476, + "learning_rate": 1.8358501987444732e-06, + "loss": 0.5289, + "step": 10048 + }, + { + "epoch": 3.528441011235955, + "grad_norm": 0.9585372805595398, + "learning_rate": 1.8354044976602543e-06, + "loss": 0.5882, + "step": 10049 + }, + { + "epoch": 3.528792134831461, + "grad_norm": 0.9552169442176819, + "learning_rate": 1.834958819304634e-06, + "loss": 0.5543, + "step": 10050 + }, + { + "epoch": 3.5291432584269664, + "grad_norm": 0.9097865223884583, + "learning_rate": 1.8345131636928517e-06, + "loss": 0.516, + "step": 10051 + }, + { + "epoch": 3.529494382022472, + "grad_norm": 0.9523643255233765, + "learning_rate": 1.8340675308401511e-06, + "loss": 0.5548, + "step": 10052 + }, + { + "epoch": 3.5298455056179776, + "grad_norm": 0.9539555311203003, + "learning_rate": 1.83362192076177e-06, + "loss": 0.5567, + "step": 10053 + }, + { + "epoch": 3.530196629213483, + "grad_norm": 0.9540579915046692, + "learning_rate": 1.8331763334729498e-06, + "loss": 0.5863, + "step": 10054 + }, + { + "epoch": 3.530547752808989, + "grad_norm": 0.9111067652702332, + "learning_rate": 1.8327307689889275e-06, + "loss": 0.5582, + "step": 10055 + }, + { + "epoch": 3.5308988764044944, + "grad_norm": 0.9437417984008789, + "learning_rate": 1.8322852273249431e-06, + "loss": 0.5734, + "step": 10056 + }, + { + "epoch": 3.53125, + "grad_norm": 0.9586054086685181, + "learning_rate": 1.8318397084962326e-06, + "loss": 0.5475, + "step": 10057 + }, + { + "epoch": 3.5316011235955056, + "grad_norm": 0.9133056402206421, + "learning_rate": 1.831394212518031e-06, + "loss": 0.5443, + "step": 10058 + }, + { + "epoch": 3.531952247191011, + "grad_norm": 0.9457817673683167, + "learning_rate": 1.8309487394055763e-06, + "loss": 0.5916, + "step": 10059 + }, + { + "epoch": 3.532303370786517, + "grad_norm": 0.9076414704322815, + "learning_rate": 1.8305032891741011e-06, + "loss": 0.5619, + "step": 10060 + }, + { + "epoch": 3.5326544943820224, + "grad_norm": 0.9279239177703857, + "learning_rate": 1.8300578618388417e-06, + "loss": 0.5639, + "step": 10061 + }, + { + "epoch": 3.533005617977528, + "grad_norm": 0.9390333294868469, + "learning_rate": 1.8296124574150292e-06, + "loss": 0.5606, + "step": 10062 + }, + { + "epoch": 3.5333567415730336, + "grad_norm": 0.930838942527771, + "learning_rate": 1.8291670759178977e-06, + "loss": 0.5491, + "step": 10063 + }, + { + "epoch": 3.533707865168539, + "grad_norm": 0.9364491105079651, + "learning_rate": 1.8287217173626776e-06, + "loss": 0.6214, + "step": 10064 + }, + { + "epoch": 3.534058988764045, + "grad_norm": 0.9322389960289001, + "learning_rate": 1.8282763817646009e-06, + "loss": 0.559, + "step": 10065 + }, + { + "epoch": 3.5344101123595504, + "grad_norm": 0.9819014072418213, + "learning_rate": 1.8278310691388967e-06, + "loss": 0.5414, + "step": 10066 + }, + { + "epoch": 3.534761235955056, + "grad_norm": 0.931509792804718, + "learning_rate": 1.827385779500795e-06, + "loss": 0.5485, + "step": 10067 + }, + { + "epoch": 3.5351123595505616, + "grad_norm": 0.9373037815093994, + "learning_rate": 1.8269405128655241e-06, + "loss": 0.557, + "step": 10068 + }, + { + "epoch": 3.535463483146067, + "grad_norm": 0.9173663258552551, + "learning_rate": 1.8264952692483114e-06, + "loss": 0.5456, + "step": 10069 + }, + { + "epoch": 3.5358146067415728, + "grad_norm": 0.9410236477851868, + "learning_rate": 1.8260500486643852e-06, + "loss": 0.5435, + "step": 10070 + }, + { + "epoch": 3.536165730337079, + "grad_norm": 0.9534938335418701, + "learning_rate": 1.8256048511289689e-06, + "loss": 0.5881, + "step": 10071 + }, + { + "epoch": 3.5365168539325844, + "grad_norm": 0.9589373469352722, + "learning_rate": 1.825159676657291e-06, + "loss": 0.5554, + "step": 10072 + }, + { + "epoch": 3.53686797752809, + "grad_norm": 0.9317744970321655, + "learning_rate": 1.8247145252645736e-06, + "loss": 0.5355, + "step": 10073 + }, + { + "epoch": 3.5372191011235956, + "grad_norm": 0.9139094352722168, + "learning_rate": 1.824269396966043e-06, + "loss": 0.5483, + "step": 10074 + }, + { + "epoch": 3.537570224719101, + "grad_norm": 0.9360598921775818, + "learning_rate": 1.8238242917769197e-06, + "loss": 0.5854, + "step": 10075 + }, + { + "epoch": 3.537921348314607, + "grad_norm": 0.9223441481590271, + "learning_rate": 1.8233792097124272e-06, + "loss": 0.5464, + "step": 10076 + }, + { + "epoch": 3.5382724719101124, + "grad_norm": 0.9400396347045898, + "learning_rate": 1.8229341507877867e-06, + "loss": 0.5526, + "step": 10077 + }, + { + "epoch": 3.538623595505618, + "grad_norm": 0.9195512533187866, + "learning_rate": 1.8224891150182192e-06, + "loss": 0.5652, + "step": 10078 + }, + { + "epoch": 3.5389747191011236, + "grad_norm": 0.917013943195343, + "learning_rate": 1.822044102418944e-06, + "loss": 0.5992, + "step": 10079 + }, + { + "epoch": 3.539325842696629, + "grad_norm": 0.9347277879714966, + "learning_rate": 1.8215991130051802e-06, + "loss": 0.5579, + "step": 10080 + }, + { + "epoch": 3.539676966292135, + "grad_norm": 0.9546303749084473, + "learning_rate": 1.8211541467921462e-06, + "loss": 0.5593, + "step": 10081 + }, + { + "epoch": 3.5400280898876404, + "grad_norm": 0.9155527949333191, + "learning_rate": 1.8207092037950603e-06, + "loss": 0.5285, + "step": 10082 + }, + { + "epoch": 3.540379213483146, + "grad_norm": 0.9676579833030701, + "learning_rate": 1.8202642840291377e-06, + "loss": 0.5753, + "step": 10083 + }, + { + "epoch": 3.5407303370786516, + "grad_norm": 0.9281370639801025, + "learning_rate": 1.8198193875095948e-06, + "loss": 0.5327, + "step": 10084 + }, + { + "epoch": 3.541081460674157, + "grad_norm": 0.9698655009269714, + "learning_rate": 1.819374514251647e-06, + "loss": 0.5992, + "step": 10085 + }, + { + "epoch": 3.541432584269663, + "grad_norm": 0.9486923217773438, + "learning_rate": 1.8189296642705083e-06, + "loss": 0.5466, + "step": 10086 + }, + { + "epoch": 3.5417837078651684, + "grad_norm": 0.90290367603302, + "learning_rate": 1.8184848375813924e-06, + "loss": 0.5319, + "step": 10087 + }, + { + "epoch": 3.5421348314606744, + "grad_norm": 0.9615541696548462, + "learning_rate": 1.8180400341995116e-06, + "loss": 0.5705, + "step": 10088 + }, + { + "epoch": 3.54248595505618, + "grad_norm": 0.9536952376365662, + "learning_rate": 1.8175952541400788e-06, + "loss": 0.574, + "step": 10089 + }, + { + "epoch": 3.5428370786516856, + "grad_norm": 0.9287588000297546, + "learning_rate": 1.8171504974183035e-06, + "loss": 0.5565, + "step": 10090 + }, + { + "epoch": 3.543188202247191, + "grad_norm": 0.9106992483139038, + "learning_rate": 1.8167057640493973e-06, + "loss": 0.5123, + "step": 10091 + }, + { + "epoch": 3.543539325842697, + "grad_norm": 0.9503855109214783, + "learning_rate": 1.8162610540485687e-06, + "loss": 0.5265, + "step": 10092 + }, + { + "epoch": 3.5438904494382024, + "grad_norm": 0.9160606861114502, + "learning_rate": 1.815816367431028e-06, + "loss": 0.5487, + "step": 10093 + }, + { + "epoch": 3.544241573033708, + "grad_norm": 0.9364210367202759, + "learning_rate": 1.8153717042119815e-06, + "loss": 0.5699, + "step": 10094 + }, + { + "epoch": 3.5445926966292136, + "grad_norm": 0.9420784711837769, + "learning_rate": 1.814927064406637e-06, + "loss": 0.5668, + "step": 10095 + }, + { + "epoch": 3.544943820224719, + "grad_norm": 0.9176422357559204, + "learning_rate": 1.814482448030201e-06, + "loss": 0.5355, + "step": 10096 + }, + { + "epoch": 3.545294943820225, + "grad_norm": 0.9263283014297485, + "learning_rate": 1.8140378550978777e-06, + "loss": 0.5758, + "step": 10097 + }, + { + "epoch": 3.5456460674157304, + "grad_norm": 0.9169784188270569, + "learning_rate": 1.8135932856248736e-06, + "loss": 0.5339, + "step": 10098 + }, + { + "epoch": 3.545997191011236, + "grad_norm": 0.9318415522575378, + "learning_rate": 1.8131487396263907e-06, + "loss": 0.5368, + "step": 10099 + }, + { + "epoch": 3.5463483146067416, + "grad_norm": 0.9441816210746765, + "learning_rate": 1.8127042171176346e-06, + "loss": 0.5639, + "step": 10100 + }, + { + "epoch": 3.546699438202247, + "grad_norm": 0.959679126739502, + "learning_rate": 1.812259718113805e-06, + "loss": 0.5836, + "step": 10101 + }, + { + "epoch": 3.547050561797753, + "grad_norm": 0.9571237564086914, + "learning_rate": 1.8118152426301055e-06, + "loss": 0.5796, + "step": 10102 + }, + { + "epoch": 3.5474016853932584, + "grad_norm": 0.9989253282546997, + "learning_rate": 1.8113707906817345e-06, + "loss": 0.5681, + "step": 10103 + }, + { + "epoch": 3.547752808988764, + "grad_norm": 0.9517219066619873, + "learning_rate": 1.8109263622838945e-06, + "loss": 0.5973, + "step": 10104 + }, + { + "epoch": 3.5481039325842696, + "grad_norm": 0.9144076108932495, + "learning_rate": 1.8104819574517823e-06, + "loss": 0.542, + "step": 10105 + }, + { + "epoch": 3.548455056179775, + "grad_norm": 0.9458330869674683, + "learning_rate": 1.8100375762005981e-06, + "loss": 0.5571, + "step": 10106 + }, + { + "epoch": 3.548806179775281, + "grad_norm": 0.9466350078582764, + "learning_rate": 1.8095932185455383e-06, + "loss": 0.5474, + "step": 10107 + }, + { + "epoch": 3.5491573033707864, + "grad_norm": 0.9249452352523804, + "learning_rate": 1.8091488845017987e-06, + "loss": 0.534, + "step": 10108 + }, + { + "epoch": 3.549508426966292, + "grad_norm": 0.9419516324996948, + "learning_rate": 1.808704574084577e-06, + "loss": 0.5464, + "step": 10109 + }, + { + "epoch": 3.5498595505617976, + "grad_norm": 0.9452080130577087, + "learning_rate": 1.8082602873090666e-06, + "loss": 0.5684, + "step": 10110 + }, + { + "epoch": 3.550210674157303, + "grad_norm": 0.9183087944984436, + "learning_rate": 1.807816024190463e-06, + "loss": 0.5629, + "step": 10111 + }, + { + "epoch": 3.550561797752809, + "grad_norm": 0.9189366102218628, + "learning_rate": 1.8073717847439581e-06, + "loss": 0.5333, + "step": 10112 + }, + { + "epoch": 3.5509129213483144, + "grad_norm": 0.9536154866218567, + "learning_rate": 1.8069275689847468e-06, + "loss": 0.5806, + "step": 10113 + }, + { + "epoch": 3.55126404494382, + "grad_norm": 0.9491174221038818, + "learning_rate": 1.8064833769280184e-06, + "loss": 0.5821, + "step": 10114 + }, + { + "epoch": 3.5516151685393256, + "grad_norm": 0.9274930357933044, + "learning_rate": 1.806039208588966e-06, + "loss": 0.557, + "step": 10115 + }, + { + "epoch": 3.5519662921348316, + "grad_norm": 0.8925730586051941, + "learning_rate": 1.805595063982778e-06, + "loss": 0.5465, + "step": 10116 + }, + { + "epoch": 3.552317415730337, + "grad_norm": 0.9468173980712891, + "learning_rate": 1.8051509431246455e-06, + "loss": 0.5459, + "step": 10117 + }, + { + "epoch": 3.552668539325843, + "grad_norm": 0.9394489526748657, + "learning_rate": 1.8047068460297551e-06, + "loss": 0.5435, + "step": 10118 + }, + { + "epoch": 3.5530196629213484, + "grad_norm": 0.9250186681747437, + "learning_rate": 1.8042627727132967e-06, + "loss": 0.5513, + "step": 10119 + }, + { + "epoch": 3.553370786516854, + "grad_norm": 0.9266967177391052, + "learning_rate": 1.8038187231904565e-06, + "loss": 0.5239, + "step": 10120 + }, + { + "epoch": 3.5537219101123596, + "grad_norm": 0.9252004623413086, + "learning_rate": 1.8033746974764185e-06, + "loss": 0.5246, + "step": 10121 + }, + { + "epoch": 3.554073033707865, + "grad_norm": 0.9506741166114807, + "learning_rate": 1.8029306955863713e-06, + "loss": 0.5661, + "step": 10122 + }, + { + "epoch": 3.554424157303371, + "grad_norm": 0.960401177406311, + "learning_rate": 1.8024867175354962e-06, + "loss": 0.5679, + "step": 10123 + }, + { + "epoch": 3.5547752808988764, + "grad_norm": 0.9277392029762268, + "learning_rate": 1.8020427633389801e-06, + "loss": 0.5686, + "step": 10124 + }, + { + "epoch": 3.555126404494382, + "grad_norm": 0.9276863932609558, + "learning_rate": 1.8015988330120026e-06, + "loss": 0.5514, + "step": 10125 + }, + { + "epoch": 3.5554775280898876, + "grad_norm": 0.9643167853355408, + "learning_rate": 1.8011549265697493e-06, + "loss": 0.6022, + "step": 10126 + }, + { + "epoch": 3.555828651685393, + "grad_norm": 0.9454492330551147, + "learning_rate": 1.800711044027398e-06, + "loss": 0.5397, + "step": 10127 + }, + { + "epoch": 3.556179775280899, + "grad_norm": 0.936007559299469, + "learning_rate": 1.8002671854001314e-06, + "loss": 0.5301, + "step": 10128 + }, + { + "epoch": 3.5565308988764044, + "grad_norm": 0.9523154497146606, + "learning_rate": 1.7998233507031275e-06, + "loss": 0.5713, + "step": 10129 + }, + { + "epoch": 3.55688202247191, + "grad_norm": 0.9625995755195618, + "learning_rate": 1.7993795399515668e-06, + "loss": 0.5452, + "step": 10130 + }, + { + "epoch": 3.5572331460674156, + "grad_norm": 0.925360918045044, + "learning_rate": 1.7989357531606252e-06, + "loss": 0.512, + "step": 10131 + }, + { + "epoch": 3.557584269662921, + "grad_norm": 0.9557311534881592, + "learning_rate": 1.7984919903454823e-06, + "loss": 0.5684, + "step": 10132 + }, + { + "epoch": 3.5579353932584272, + "grad_norm": 0.9492946863174438, + "learning_rate": 1.7980482515213128e-06, + "loss": 0.5326, + "step": 10133 + }, + { + "epoch": 3.558286516853933, + "grad_norm": 0.9301472902297974, + "learning_rate": 1.7976045367032913e-06, + "loss": 0.5517, + "step": 10134 + }, + { + "epoch": 3.5586376404494384, + "grad_norm": 0.9928217530250549, + "learning_rate": 1.7971608459065947e-06, + "loss": 0.5653, + "step": 10135 + }, + { + "epoch": 3.558988764044944, + "grad_norm": 0.9367715120315552, + "learning_rate": 1.796717179146395e-06, + "loss": 0.5486, + "step": 10136 + }, + { + "epoch": 3.5593398876404496, + "grad_norm": 0.9381900429725647, + "learning_rate": 1.7962735364378667e-06, + "loss": 0.5604, + "step": 10137 + }, + { + "epoch": 3.559691011235955, + "grad_norm": 0.9617657661437988, + "learning_rate": 1.7958299177961803e-06, + "loss": 0.5574, + "step": 10138 + }, + { + "epoch": 3.560042134831461, + "grad_norm": 0.930894672870636, + "learning_rate": 1.795386323236509e-06, + "loss": 0.5717, + "step": 10139 + }, + { + "epoch": 3.5603932584269664, + "grad_norm": 0.9762129783630371, + "learning_rate": 1.7949427527740215e-06, + "loss": 0.5565, + "step": 10140 + }, + { + "epoch": 3.560744382022472, + "grad_norm": 0.9345110058784485, + "learning_rate": 1.7944992064238898e-06, + "loss": 0.5502, + "step": 10141 + }, + { + "epoch": 3.5610955056179776, + "grad_norm": 0.935124933719635, + "learning_rate": 1.7940556842012805e-06, + "loss": 0.5389, + "step": 10142 + }, + { + "epoch": 3.561446629213483, + "grad_norm": 0.9299739599227905, + "learning_rate": 1.7936121861213637e-06, + "loss": 0.5601, + "step": 10143 + }, + { + "epoch": 3.561797752808989, + "grad_norm": 0.9300097227096558, + "learning_rate": 1.7931687121993046e-06, + "loss": 0.5956, + "step": 10144 + }, + { + "epoch": 3.5621488764044944, + "grad_norm": 0.9782840609550476, + "learning_rate": 1.7927252624502717e-06, + "loss": 0.5847, + "step": 10145 + }, + { + "epoch": 3.5625, + "grad_norm": 0.9114406704902649, + "learning_rate": 1.7922818368894296e-06, + "loss": 0.4813, + "step": 10146 + }, + { + "epoch": 3.5628511235955056, + "grad_norm": 0.9847769737243652, + "learning_rate": 1.7918384355319421e-06, + "loss": 0.5904, + "step": 10147 + }, + { + "epoch": 3.563202247191011, + "grad_norm": 0.9187098145484924, + "learning_rate": 1.7913950583929752e-06, + "loss": 0.5445, + "step": 10148 + }, + { + "epoch": 3.563553370786517, + "grad_norm": 0.9256553053855896, + "learning_rate": 1.7909517054876895e-06, + "loss": 0.5825, + "step": 10149 + }, + { + "epoch": 3.5639044943820224, + "grad_norm": 0.9589816927909851, + "learning_rate": 1.7905083768312503e-06, + "loss": 0.5772, + "step": 10150 + }, + { + "epoch": 3.564255617977528, + "grad_norm": 0.9310399889945984, + "learning_rate": 1.7900650724388164e-06, + "loss": 0.5954, + "step": 10151 + }, + { + "epoch": 3.5646067415730336, + "grad_norm": 0.9399897456169128, + "learning_rate": 1.7896217923255504e-06, + "loss": 0.5429, + "step": 10152 + }, + { + "epoch": 3.564957865168539, + "grad_norm": 0.9359461665153503, + "learning_rate": 1.7891785365066106e-06, + "loss": 0.5368, + "step": 10153 + }, + { + "epoch": 3.565308988764045, + "grad_norm": 0.9637037515640259, + "learning_rate": 1.7887353049971567e-06, + "loss": 0.5422, + "step": 10154 + }, + { + "epoch": 3.5656601123595504, + "grad_norm": 0.9640803337097168, + "learning_rate": 1.7882920978123466e-06, + "loss": 0.5771, + "step": 10155 + }, + { + "epoch": 3.566011235955056, + "grad_norm": 0.9235727787017822, + "learning_rate": 1.7878489149673377e-06, + "loss": 0.522, + "step": 10156 + }, + { + "epoch": 3.5663623595505616, + "grad_norm": 0.9424094557762146, + "learning_rate": 1.7874057564772873e-06, + "loss": 0.5701, + "step": 10157 + }, + { + "epoch": 3.566713483146067, + "grad_norm": 0.9353017210960388, + "learning_rate": 1.7869626223573494e-06, + "loss": 0.5593, + "step": 10158 + }, + { + "epoch": 3.5670646067415728, + "grad_norm": 0.9460206627845764, + "learning_rate": 1.7865195126226804e-06, + "loss": 0.5638, + "step": 10159 + }, + { + "epoch": 3.567415730337079, + "grad_norm": 0.9309295415878296, + "learning_rate": 1.7860764272884324e-06, + "loss": 0.5794, + "step": 10160 + }, + { + "epoch": 3.5677668539325844, + "grad_norm": 0.9319258332252502, + "learning_rate": 1.7856333663697608e-06, + "loss": 0.5537, + "step": 10161 + }, + { + "epoch": 3.56811797752809, + "grad_norm": 0.9057424664497375, + "learning_rate": 1.7851903298818162e-06, + "loss": 0.5362, + "step": 10162 + }, + { + "epoch": 3.5684691011235956, + "grad_norm": 0.9235795736312866, + "learning_rate": 1.7847473178397506e-06, + "loss": 0.5668, + "step": 10163 + }, + { + "epoch": 3.568820224719101, + "grad_norm": 0.9426140189170837, + "learning_rate": 1.7843043302587148e-06, + "loss": 0.5569, + "step": 10164 + }, + { + "epoch": 3.569171348314607, + "grad_norm": 0.9649958610534668, + "learning_rate": 1.7838613671538588e-06, + "loss": 0.5952, + "step": 10165 + }, + { + "epoch": 3.5695224719101124, + "grad_norm": 0.922294020652771, + "learning_rate": 1.7834184285403309e-06, + "loss": 0.534, + "step": 10166 + }, + { + "epoch": 3.569873595505618, + "grad_norm": 0.9567601680755615, + "learning_rate": 1.7829755144332794e-06, + "loss": 0.602, + "step": 10167 + }, + { + "epoch": 3.5702247191011236, + "grad_norm": 0.898504912853241, + "learning_rate": 1.7825326248478519e-06, + "loss": 0.5058, + "step": 10168 + }, + { + "epoch": 3.570575842696629, + "grad_norm": 0.9048131704330444, + "learning_rate": 1.7820897597991953e-06, + "loss": 0.5274, + "step": 10169 + }, + { + "epoch": 3.570926966292135, + "grad_norm": 0.9627010226249695, + "learning_rate": 1.7816469193024541e-06, + "loss": 0.5789, + "step": 10170 + }, + { + "epoch": 3.5712780898876404, + "grad_norm": 0.9262627363204956, + "learning_rate": 1.7812041033727734e-06, + "loss": 0.5838, + "step": 10171 + }, + { + "epoch": 3.571629213483146, + "grad_norm": 0.9370171427726746, + "learning_rate": 1.7807613120252973e-06, + "loss": 0.5268, + "step": 10172 + }, + { + "epoch": 3.5719803370786516, + "grad_norm": 0.9327815175056458, + "learning_rate": 1.780318545275169e-06, + "loss": 0.5198, + "step": 10173 + }, + { + "epoch": 3.572331460674157, + "grad_norm": 0.9311509728431702, + "learning_rate": 1.7798758031375305e-06, + "loss": 0.5465, + "step": 10174 + }, + { + "epoch": 3.572682584269663, + "grad_norm": 0.9186543822288513, + "learning_rate": 1.7794330856275232e-06, + "loss": 0.5827, + "step": 10175 + }, + { + "epoch": 3.5730337078651684, + "grad_norm": 0.9325449466705322, + "learning_rate": 1.7789903927602886e-06, + "loss": 0.5652, + "step": 10176 + }, + { + "epoch": 3.5733848314606744, + "grad_norm": 0.9380399584770203, + "learning_rate": 1.7785477245509641e-06, + "loss": 0.5807, + "step": 10177 + }, + { + "epoch": 3.57373595505618, + "grad_norm": 0.9053071737289429, + "learning_rate": 1.7781050810146917e-06, + "loss": 0.5187, + "step": 10178 + }, + { + "epoch": 3.5740870786516856, + "grad_norm": 0.929780125617981, + "learning_rate": 1.777662462166606e-06, + "loss": 0.543, + "step": 10179 + }, + { + "epoch": 3.574438202247191, + "grad_norm": 0.9639439582824707, + "learning_rate": 1.7772198680218477e-06, + "loss": 0.5692, + "step": 10180 + }, + { + "epoch": 3.574789325842697, + "grad_norm": 0.9495415091514587, + "learning_rate": 1.77677729859555e-06, + "loss": 0.5564, + "step": 10181 + }, + { + "epoch": 3.5751404494382024, + "grad_norm": 0.936107337474823, + "learning_rate": 1.7763347539028512e-06, + "loss": 0.5671, + "step": 10182 + }, + { + "epoch": 3.575491573033708, + "grad_norm": 0.948896586894989, + "learning_rate": 1.7758922339588842e-06, + "loss": 0.5913, + "step": 10183 + }, + { + "epoch": 3.5758426966292136, + "grad_norm": 1.4140492677688599, + "learning_rate": 1.7754497387787822e-06, + "loss": 0.5804, + "step": 10184 + }, + { + "epoch": 3.576193820224719, + "grad_norm": 0.936802864074707, + "learning_rate": 1.7750072683776804e-06, + "loss": 0.556, + "step": 10185 + }, + { + "epoch": 3.576544943820225, + "grad_norm": 0.9538032412528992, + "learning_rate": 1.7745648227707085e-06, + "loss": 0.5516, + "step": 10186 + }, + { + "epoch": 3.5768960674157304, + "grad_norm": 1.0551891326904297, + "learning_rate": 1.7741224019729996e-06, + "loss": 0.562, + "step": 10187 + }, + { + "epoch": 3.577247191011236, + "grad_norm": 0.9379180669784546, + "learning_rate": 1.7736800059996827e-06, + "loss": 0.5457, + "step": 10188 + }, + { + "epoch": 3.5775983146067416, + "grad_norm": 0.9388746619224548, + "learning_rate": 1.7732376348658891e-06, + "loss": 0.5426, + "step": 10189 + }, + { + "epoch": 3.577949438202247, + "grad_norm": 0.9390409588813782, + "learning_rate": 1.7727952885867455e-06, + "loss": 0.5691, + "step": 10190 + }, + { + "epoch": 3.578300561797753, + "grad_norm": 0.9268652200698853, + "learning_rate": 1.7723529671773818e-06, + "loss": 0.5542, + "step": 10191 + }, + { + "epoch": 3.5786516853932584, + "grad_norm": 0.9327108263969421, + "learning_rate": 1.7719106706529227e-06, + "loss": 0.5493, + "step": 10192 + }, + { + "epoch": 3.579002808988764, + "grad_norm": 0.9177367687225342, + "learning_rate": 1.771468399028497e-06, + "loss": 0.5461, + "step": 10193 + }, + { + "epoch": 3.5793539325842696, + "grad_norm": 0.8948767185211182, + "learning_rate": 1.7710261523192273e-06, + "loss": 0.5051, + "step": 10194 + }, + { + "epoch": 3.579705056179775, + "grad_norm": 0.9690757989883423, + "learning_rate": 1.7705839305402406e-06, + "loss": 0.5867, + "step": 10195 + }, + { + "epoch": 3.580056179775281, + "grad_norm": 0.9315474629402161, + "learning_rate": 1.770141733706659e-06, + "loss": 0.5396, + "step": 10196 + }, + { + "epoch": 3.5804073033707864, + "grad_norm": 0.9016444683074951, + "learning_rate": 1.769699561833605e-06, + "loss": 0.5247, + "step": 10197 + }, + { + "epoch": 3.580758426966292, + "grad_norm": 0.9302697777748108, + "learning_rate": 1.7692574149362018e-06, + "loss": 0.5318, + "step": 10198 + }, + { + "epoch": 3.5811095505617976, + "grad_norm": 0.929118812084198, + "learning_rate": 1.7688152930295689e-06, + "loss": 0.5843, + "step": 10199 + }, + { + "epoch": 3.581460674157303, + "grad_norm": 0.9106625318527222, + "learning_rate": 1.768373196128828e-06, + "loss": 0.5435, + "step": 10200 + }, + { + "epoch": 3.581811797752809, + "grad_norm": 0.927374541759491, + "learning_rate": 1.7679311242490968e-06, + "loss": 0.549, + "step": 10201 + }, + { + "epoch": 3.5821629213483144, + "grad_norm": 0.9355881214141846, + "learning_rate": 1.767489077405496e-06, + "loss": 0.5504, + "step": 10202 + }, + { + "epoch": 3.58251404494382, + "grad_norm": 0.9542840719223022, + "learning_rate": 1.7670470556131406e-06, + "loss": 0.5774, + "step": 10203 + }, + { + "epoch": 3.5828651685393256, + "grad_norm": 0.8945448398590088, + "learning_rate": 1.7666050588871498e-06, + "loss": 0.542, + "step": 10204 + }, + { + "epoch": 3.5832162921348316, + "grad_norm": 0.9620029330253601, + "learning_rate": 1.7661630872426372e-06, + "loss": 0.5861, + "step": 10205 + }, + { + "epoch": 3.583567415730337, + "grad_norm": 0.9557713270187378, + "learning_rate": 1.7657211406947207e-06, + "loss": 0.5597, + "step": 10206 + }, + { + "epoch": 3.583918539325843, + "grad_norm": 0.9401268362998962, + "learning_rate": 1.765279219258511e-06, + "loss": 0.5586, + "step": 10207 + }, + { + "epoch": 3.5842696629213484, + "grad_norm": 0.9724933505058289, + "learning_rate": 1.764837322949125e-06, + "loss": 0.5693, + "step": 10208 + }, + { + "epoch": 3.584620786516854, + "grad_norm": 0.9400120377540588, + "learning_rate": 1.764395451781673e-06, + "loss": 0.518, + "step": 10209 + }, + { + "epoch": 3.5849719101123596, + "grad_norm": 0.9353851079940796, + "learning_rate": 1.7639536057712664e-06, + "loss": 0.5608, + "step": 10210 + }, + { + "epoch": 3.585323033707865, + "grad_norm": 0.897126317024231, + "learning_rate": 1.7635117849330177e-06, + "loss": 0.5389, + "step": 10211 + }, + { + "epoch": 3.585674157303371, + "grad_norm": 0.9246470332145691, + "learning_rate": 1.7630699892820346e-06, + "loss": 0.5121, + "step": 10212 + }, + { + "epoch": 3.5860252808988764, + "grad_norm": 0.965018630027771, + "learning_rate": 1.762628218833428e-06, + "loss": 0.5651, + "step": 10213 + }, + { + "epoch": 3.586376404494382, + "grad_norm": 0.9380896687507629, + "learning_rate": 1.7621864736023048e-06, + "loss": 0.5562, + "step": 10214 + }, + { + "epoch": 3.5867275280898876, + "grad_norm": 0.9403875470161438, + "learning_rate": 1.7617447536037736e-06, + "loss": 0.6357, + "step": 10215 + }, + { + "epoch": 3.587078651685393, + "grad_norm": 0.9712557792663574, + "learning_rate": 1.761303058852939e-06, + "loss": 0.5842, + "step": 10216 + }, + { + "epoch": 3.587429775280899, + "grad_norm": 0.9643117189407349, + "learning_rate": 1.7608613893649087e-06, + "loss": 0.5476, + "step": 10217 + }, + { + "epoch": 3.5877808988764044, + "grad_norm": 0.8838028907775879, + "learning_rate": 1.7604197451547856e-06, + "loss": 0.4873, + "step": 10218 + }, + { + "epoch": 3.58813202247191, + "grad_norm": 0.9296213388442993, + "learning_rate": 1.7599781262376753e-06, + "loss": 0.5387, + "step": 10219 + }, + { + "epoch": 3.5884831460674156, + "grad_norm": 0.9378889799118042, + "learning_rate": 1.7595365326286795e-06, + "loss": 0.5866, + "step": 10220 + }, + { + "epoch": 3.588834269662921, + "grad_norm": 0.9173360466957092, + "learning_rate": 1.7590949643429e-06, + "loss": 0.5576, + "step": 10221 + }, + { + "epoch": 3.5891853932584272, + "grad_norm": 0.941088855266571, + "learning_rate": 1.7586534213954398e-06, + "loss": 0.568, + "step": 10222 + }, + { + "epoch": 3.589536516853933, + "grad_norm": 0.9185635447502136, + "learning_rate": 1.758211903801397e-06, + "loss": 0.5628, + "step": 10223 + }, + { + "epoch": 3.5898876404494384, + "grad_norm": 0.9696146845817566, + "learning_rate": 1.7577704115758734e-06, + "loss": 0.5817, + "step": 10224 + }, + { + "epoch": 3.590238764044944, + "grad_norm": 0.9238273501396179, + "learning_rate": 1.7573289447339655e-06, + "loss": 0.5767, + "step": 10225 + }, + { + "epoch": 3.5905898876404496, + "grad_norm": 0.9485665559768677, + "learning_rate": 1.7568875032907734e-06, + "loss": 0.5691, + "step": 10226 + }, + { + "epoch": 3.590941011235955, + "grad_norm": 0.9562817811965942, + "learning_rate": 1.7564460872613914e-06, + "loss": 0.5451, + "step": 10227 + }, + { + "epoch": 3.591292134831461, + "grad_norm": 0.9339578151702881, + "learning_rate": 1.7560046966609185e-06, + "loss": 0.5227, + "step": 10228 + }, + { + "epoch": 3.5916432584269664, + "grad_norm": 0.9489087462425232, + "learning_rate": 1.7555633315044467e-06, + "loss": 0.5858, + "step": 10229 + }, + { + "epoch": 3.591994382022472, + "grad_norm": 0.9453594088554382, + "learning_rate": 1.7551219918070733e-06, + "loss": 0.5392, + "step": 10230 + }, + { + "epoch": 3.5923455056179776, + "grad_norm": 0.9985160827636719, + "learning_rate": 1.7546806775838893e-06, + "loss": 0.5448, + "step": 10231 + }, + { + "epoch": 3.592696629213483, + "grad_norm": 0.9158629775047302, + "learning_rate": 1.7542393888499895e-06, + "loss": 0.5179, + "step": 10232 + }, + { + "epoch": 3.593047752808989, + "grad_norm": 0.9466661214828491, + "learning_rate": 1.7537981256204645e-06, + "loss": 0.5564, + "step": 10233 + }, + { + "epoch": 3.5933988764044944, + "grad_norm": 1.010447382926941, + "learning_rate": 1.7533568879104037e-06, + "loss": 0.5779, + "step": 10234 + }, + { + "epoch": 3.59375, + "grad_norm": 0.8933452367782593, + "learning_rate": 1.7529156757348997e-06, + "loss": 0.5094, + "step": 10235 + }, + { + "epoch": 3.5941011235955056, + "grad_norm": 0.92055344581604, + "learning_rate": 1.7524744891090392e-06, + "loss": 0.5438, + "step": 10236 + }, + { + "epoch": 3.594452247191011, + "grad_norm": 0.942259669303894, + "learning_rate": 1.7520333280479123e-06, + "loss": 0.5655, + "step": 10237 + }, + { + "epoch": 3.594803370786517, + "grad_norm": 0.9531282186508179, + "learning_rate": 1.7515921925666053e-06, + "loss": 0.6008, + "step": 10238 + }, + { + "epoch": 3.5951544943820224, + "grad_norm": 0.9624236822128296, + "learning_rate": 1.7511510826802052e-06, + "loss": 0.5789, + "step": 10239 + }, + { + "epoch": 3.595505617977528, + "grad_norm": 0.9674206972122192, + "learning_rate": 1.7507099984037968e-06, + "loss": 0.5475, + "step": 10240 + }, + { + "epoch": 3.5958567415730336, + "grad_norm": 0.9103813767433167, + "learning_rate": 1.7502689397524652e-06, + "loss": 0.5318, + "step": 10241 + }, + { + "epoch": 3.596207865168539, + "grad_norm": 0.9277705550193787, + "learning_rate": 1.7498279067412948e-06, + "loss": 0.521, + "step": 10242 + }, + { + "epoch": 3.596558988764045, + "grad_norm": 0.940588116645813, + "learning_rate": 1.7493868993853677e-06, + "loss": 0.5599, + "step": 10243 + }, + { + "epoch": 3.5969101123595504, + "grad_norm": 0.9352573752403259, + "learning_rate": 1.7489459176997664e-06, + "loss": 0.5691, + "step": 10244 + }, + { + "epoch": 3.597261235955056, + "grad_norm": 0.9394773840904236, + "learning_rate": 1.7485049616995719e-06, + "loss": 0.5525, + "step": 10245 + }, + { + "epoch": 3.5976123595505616, + "grad_norm": 0.9550203084945679, + "learning_rate": 1.7480640313998653e-06, + "loss": 0.5475, + "step": 10246 + }, + { + "epoch": 3.597963483146067, + "grad_norm": 0.9414670467376709, + "learning_rate": 1.7476231268157245e-06, + "loss": 0.5637, + "step": 10247 + }, + { + "epoch": 3.5983146067415728, + "grad_norm": 0.9258294105529785, + "learning_rate": 1.7471822479622297e-06, + "loss": 0.5758, + "step": 10248 + }, + { + "epoch": 3.598665730337079, + "grad_norm": 0.9952525496482849, + "learning_rate": 1.7467413948544576e-06, + "loss": 0.6015, + "step": 10249 + }, + { + "epoch": 3.5990168539325844, + "grad_norm": 0.94133061170578, + "learning_rate": 1.7463005675074851e-06, + "loss": 0.5757, + "step": 10250 + }, + { + "epoch": 3.59936797752809, + "grad_norm": 0.8954108953475952, + "learning_rate": 1.7458597659363879e-06, + "loss": 0.5596, + "step": 10251 + }, + { + "epoch": 3.5997191011235956, + "grad_norm": 0.9126954674720764, + "learning_rate": 1.745418990156242e-06, + "loss": 0.5168, + "step": 10252 + }, + { + "epoch": 3.600070224719101, + "grad_norm": 0.9235880970954895, + "learning_rate": 1.7449782401821206e-06, + "loss": 0.5084, + "step": 10253 + }, + { + "epoch": 3.600421348314607, + "grad_norm": 0.9084330797195435, + "learning_rate": 1.7445375160290972e-06, + "loss": 0.5344, + "step": 10254 + }, + { + "epoch": 3.6007724719101124, + "grad_norm": 0.9165638089179993, + "learning_rate": 1.7440968177122443e-06, + "loss": 0.5665, + "step": 10255 + }, + { + "epoch": 3.601123595505618, + "grad_norm": 0.9004682302474976, + "learning_rate": 1.7436561452466342e-06, + "loss": 0.5685, + "step": 10256 + }, + { + "epoch": 3.6014747191011236, + "grad_norm": 0.9474729299545288, + "learning_rate": 1.7432154986473354e-06, + "loss": 0.5644, + "step": 10257 + }, + { + "epoch": 3.601825842696629, + "grad_norm": 0.908247709274292, + "learning_rate": 1.7427748779294201e-06, + "loss": 0.522, + "step": 10258 + }, + { + "epoch": 3.602176966292135, + "grad_norm": 0.911353349685669, + "learning_rate": 1.7423342831079554e-06, + "loss": 0.5659, + "step": 10259 + }, + { + "epoch": 3.6025280898876404, + "grad_norm": 0.9527578353881836, + "learning_rate": 1.7418937141980103e-06, + "loss": 0.5468, + "step": 10260 + }, + { + "epoch": 3.602879213483146, + "grad_norm": 0.949134886264801, + "learning_rate": 1.7414531712146509e-06, + "loss": 0.5538, + "step": 10261 + }, + { + "epoch": 3.6032303370786516, + "grad_norm": 0.943524956703186, + "learning_rate": 1.7410126541729443e-06, + "loss": 0.5667, + "step": 10262 + }, + { + "epoch": 3.603581460674157, + "grad_norm": 1.4641749858856201, + "learning_rate": 1.7405721630879557e-06, + "loss": 0.5204, + "step": 10263 + }, + { + "epoch": 3.603932584269663, + "grad_norm": 0.9282621145248413, + "learning_rate": 1.7401316979747485e-06, + "loss": 0.5468, + "step": 10264 + }, + { + "epoch": 3.6042837078651684, + "grad_norm": 0.9426337480545044, + "learning_rate": 1.7396912588483877e-06, + "loss": 0.555, + "step": 10265 + }, + { + "epoch": 3.6046348314606744, + "grad_norm": 0.9296274781227112, + "learning_rate": 1.7392508457239344e-06, + "loss": 0.5551, + "step": 10266 + }, + { + "epoch": 3.60498595505618, + "grad_norm": 0.9180656671524048, + "learning_rate": 1.7388104586164522e-06, + "loss": 0.5481, + "step": 10267 + }, + { + "epoch": 3.6053370786516856, + "grad_norm": 0.9247082471847534, + "learning_rate": 1.7383700975410002e-06, + "loss": 0.5656, + "step": 10268 + }, + { + "epoch": 3.605688202247191, + "grad_norm": 0.9214598536491394, + "learning_rate": 1.7379297625126399e-06, + "loss": 0.5605, + "step": 10269 + }, + { + "epoch": 3.606039325842697, + "grad_norm": 0.9402819871902466, + "learning_rate": 1.7374894535464286e-06, + "loss": 0.5514, + "step": 10270 + }, + { + "epoch": 3.6063904494382024, + "grad_norm": 0.9422125816345215, + "learning_rate": 1.7370491706574266e-06, + "loss": 0.5463, + "step": 10271 + }, + { + "epoch": 3.606741573033708, + "grad_norm": 1.0062845945358276, + "learning_rate": 1.7366089138606901e-06, + "loss": 0.5542, + "step": 10272 + }, + { + "epoch": 3.6070926966292136, + "grad_norm": 0.9263195991516113, + "learning_rate": 1.7361686831712743e-06, + "loss": 0.5534, + "step": 10273 + }, + { + "epoch": 3.607443820224719, + "grad_norm": 0.9415993690490723, + "learning_rate": 1.735728478604237e-06, + "loss": 0.5711, + "step": 10274 + }, + { + "epoch": 3.607794943820225, + "grad_norm": 0.9498497843742371, + "learning_rate": 1.7352883001746307e-06, + "loss": 0.5536, + "step": 10275 + }, + { + "epoch": 3.6081460674157304, + "grad_norm": 0.9386737942695618, + "learning_rate": 1.7348481478975112e-06, + "loss": 0.5575, + "step": 10276 + }, + { + "epoch": 3.608497191011236, + "grad_norm": 0.9407949447631836, + "learning_rate": 1.7344080217879294e-06, + "loss": 0.5524, + "step": 10277 + }, + { + "epoch": 3.6088483146067416, + "grad_norm": 0.9271396994590759, + "learning_rate": 1.733967921860939e-06, + "loss": 0.5505, + "step": 10278 + }, + { + "epoch": 3.609199438202247, + "grad_norm": 0.9327860474586487, + "learning_rate": 1.7335278481315893e-06, + "loss": 0.5383, + "step": 10279 + }, + { + "epoch": 3.609550561797753, + "grad_norm": 0.9332837462425232, + "learning_rate": 1.7330878006149321e-06, + "loss": 0.5707, + "step": 10280 + }, + { + "epoch": 3.6099016853932584, + "grad_norm": 0.9654833674430847, + "learning_rate": 1.7326477793260149e-06, + "loss": 0.5498, + "step": 10281 + }, + { + "epoch": 3.610252808988764, + "grad_norm": 0.9592836499214172, + "learning_rate": 1.7322077842798885e-06, + "loss": 0.5739, + "step": 10282 + }, + { + "epoch": 3.6106039325842696, + "grad_norm": 0.9181501269340515, + "learning_rate": 1.7317678154915984e-06, + "loss": 0.5227, + "step": 10283 + }, + { + "epoch": 3.610955056179775, + "grad_norm": 0.9529634118080139, + "learning_rate": 1.7313278729761906e-06, + "loss": 0.584, + "step": 10284 + }, + { + "epoch": 3.611306179775281, + "grad_norm": 0.9145504832267761, + "learning_rate": 1.730887956748713e-06, + "loss": 0.5589, + "step": 10285 + }, + { + "epoch": 3.6116573033707864, + "grad_norm": 0.9248044490814209, + "learning_rate": 1.730448066824208e-06, + "loss": 0.5622, + "step": 10286 + }, + { + "epoch": 3.612008426966292, + "grad_norm": 0.9414767622947693, + "learning_rate": 1.7300082032177215e-06, + "loss": 0.5833, + "step": 10287 + }, + { + "epoch": 3.6123595505617976, + "grad_norm": 0.976969838142395, + "learning_rate": 1.7295683659442946e-06, + "loss": 0.5538, + "step": 10288 + }, + { + "epoch": 3.612710674157303, + "grad_norm": 0.9323339462280273, + "learning_rate": 1.7291285550189714e-06, + "loss": 0.5377, + "step": 10289 + }, + { + "epoch": 3.613061797752809, + "grad_norm": 0.9116663336753845, + "learning_rate": 1.728688770456791e-06, + "loss": 0.5285, + "step": 10290 + }, + { + "epoch": 3.6134129213483144, + "grad_norm": 0.913143515586853, + "learning_rate": 1.7282490122727958e-06, + "loss": 0.525, + "step": 10291 + }, + { + "epoch": 3.61376404494382, + "grad_norm": 0.9342637658119202, + "learning_rate": 1.7278092804820227e-06, + "loss": 0.5668, + "step": 10292 + }, + { + "epoch": 3.6141151685393256, + "grad_norm": 0.9866119027137756, + "learning_rate": 1.7273695750995122e-06, + "loss": 0.5436, + "step": 10293 + }, + { + "epoch": 3.6144662921348316, + "grad_norm": 0.9573028087615967, + "learning_rate": 1.7269298961403008e-06, + "loss": 0.5584, + "step": 10294 + }, + { + "epoch": 3.614817415730337, + "grad_norm": 0.9353931546211243, + "learning_rate": 1.7264902436194259e-06, + "loss": 0.5829, + "step": 10295 + }, + { + "epoch": 3.615168539325843, + "grad_norm": 0.9329125881195068, + "learning_rate": 1.7260506175519228e-06, + "loss": 0.536, + "step": 10296 + }, + { + "epoch": 3.6155196629213484, + "grad_norm": 0.9389246106147766, + "learning_rate": 1.7256110179528258e-06, + "loss": 0.5427, + "step": 10297 + }, + { + "epoch": 3.615870786516854, + "grad_norm": 0.8942320346832275, + "learning_rate": 1.7251714448371699e-06, + "loss": 0.5032, + "step": 10298 + }, + { + "epoch": 3.6162219101123596, + "grad_norm": 0.9399652481079102, + "learning_rate": 1.7247318982199862e-06, + "loss": 0.5507, + "step": 10299 + }, + { + "epoch": 3.616573033707865, + "grad_norm": 0.9191274642944336, + "learning_rate": 1.7242923781163095e-06, + "loss": 0.5389, + "step": 10300 + }, + { + "epoch": 3.616924157303371, + "grad_norm": 0.908748209476471, + "learning_rate": 1.723852884541169e-06, + "loss": 0.5493, + "step": 10301 + }, + { + "epoch": 3.6172752808988764, + "grad_norm": 0.9345622658729553, + "learning_rate": 1.7234134175095961e-06, + "loss": 0.5355, + "step": 10302 + }, + { + "epoch": 3.617626404494382, + "grad_norm": 0.8963354825973511, + "learning_rate": 1.7229739770366194e-06, + "loss": 0.5369, + "step": 10303 + }, + { + "epoch": 3.6179775280898876, + "grad_norm": 0.9182429313659668, + "learning_rate": 1.7225345631372686e-06, + "loss": 0.5491, + "step": 10304 + }, + { + "epoch": 3.618328651685393, + "grad_norm": 0.9528826475143433, + "learning_rate": 1.7220951758265692e-06, + "loss": 0.5725, + "step": 10305 + }, + { + "epoch": 3.618679775280899, + "grad_norm": 0.9367262125015259, + "learning_rate": 1.7216558151195503e-06, + "loss": 0.5707, + "step": 10306 + }, + { + "epoch": 3.6190308988764044, + "grad_norm": 0.9196895956993103, + "learning_rate": 1.7212164810312355e-06, + "loss": 0.5693, + "step": 10307 + }, + { + "epoch": 3.61938202247191, + "grad_norm": 0.9537745714187622, + "learning_rate": 1.7207771735766516e-06, + "loss": 0.5561, + "step": 10308 + }, + { + "epoch": 3.6197331460674156, + "grad_norm": 0.91482013463974, + "learning_rate": 1.7203378927708215e-06, + "loss": 0.5071, + "step": 10309 + }, + { + "epoch": 3.620084269662921, + "grad_norm": 0.9603795409202576, + "learning_rate": 1.7198986386287676e-06, + "loss": 0.581, + "step": 10310 + }, + { + "epoch": 3.6204353932584272, + "grad_norm": 0.929828405380249, + "learning_rate": 1.7194594111655133e-06, + "loss": 0.5872, + "step": 10311 + }, + { + "epoch": 3.620786516853933, + "grad_norm": 0.9503136873245239, + "learning_rate": 1.7190202103960786e-06, + "loss": 0.5714, + "step": 10312 + }, + { + "epoch": 3.6211376404494384, + "grad_norm": 0.9395283460617065, + "learning_rate": 1.718581036335485e-06, + "loss": 0.5625, + "step": 10313 + }, + { + "epoch": 3.621488764044944, + "grad_norm": 0.949589729309082, + "learning_rate": 1.7181418889987505e-06, + "loss": 0.5605, + "step": 10314 + }, + { + "epoch": 3.6218398876404496, + "grad_norm": 0.9049926996231079, + "learning_rate": 1.7177027684008954e-06, + "loss": 0.5265, + "step": 10315 + }, + { + "epoch": 3.622191011235955, + "grad_norm": 0.9200312495231628, + "learning_rate": 1.7172636745569352e-06, + "loss": 0.5578, + "step": 10316 + }, + { + "epoch": 3.622542134831461, + "grad_norm": 0.9323643445968628, + "learning_rate": 1.7168246074818883e-06, + "loss": 0.545, + "step": 10317 + }, + { + "epoch": 3.6228932584269664, + "grad_norm": 0.9339997172355652, + "learning_rate": 1.716385567190769e-06, + "loss": 0.4922, + "step": 10318 + }, + { + "epoch": 3.623244382022472, + "grad_norm": 0.9642248153686523, + "learning_rate": 1.7159465536985935e-06, + "loss": 0.5694, + "step": 10319 + }, + { + "epoch": 3.6235955056179776, + "grad_norm": 0.9170717000961304, + "learning_rate": 1.7155075670203747e-06, + "loss": 0.552, + "step": 10320 + }, + { + "epoch": 3.623946629213483, + "grad_norm": 0.9391553997993469, + "learning_rate": 1.7150686071711257e-06, + "loss": 0.5499, + "step": 10321 + }, + { + "epoch": 3.624297752808989, + "grad_norm": 0.9517053961753845, + "learning_rate": 1.7146296741658591e-06, + "loss": 0.5551, + "step": 10322 + }, + { + "epoch": 3.6246488764044944, + "grad_norm": 0.9791589975357056, + "learning_rate": 1.7141907680195847e-06, + "loss": 0.5806, + "step": 10323 + }, + { + "epoch": 3.625, + "grad_norm": 0.9464588761329651, + "learning_rate": 1.7137518887473143e-06, + "loss": 0.5395, + "step": 10324 + }, + { + "epoch": 3.6253511235955056, + "grad_norm": 0.9430926442146301, + "learning_rate": 1.7133130363640554e-06, + "loss": 0.5965, + "step": 10325 + }, + { + "epoch": 3.625702247191011, + "grad_norm": 0.9554885625839233, + "learning_rate": 1.712874210884819e-06, + "loss": 0.5858, + "step": 10326 + }, + { + "epoch": 3.626053370786517, + "grad_norm": 0.9733748435974121, + "learning_rate": 1.7124354123246104e-06, + "loss": 0.5511, + "step": 10327 + }, + { + "epoch": 3.6264044943820224, + "grad_norm": 0.9096484780311584, + "learning_rate": 1.7119966406984367e-06, + "loss": 0.5436, + "step": 10328 + }, + { + "epoch": 3.626755617977528, + "grad_norm": 0.9413081407546997, + "learning_rate": 1.7115578960213037e-06, + "loss": 0.5286, + "step": 10329 + }, + { + "epoch": 3.6271067415730336, + "grad_norm": 0.9204476475715637, + "learning_rate": 1.7111191783082157e-06, + "loss": 0.5357, + "step": 10330 + }, + { + "epoch": 3.627457865168539, + "grad_norm": 0.9445713758468628, + "learning_rate": 1.7106804875741768e-06, + "loss": 0.5539, + "step": 10331 + }, + { + "epoch": 3.627808988764045, + "grad_norm": 0.960208535194397, + "learning_rate": 1.71024182383419e-06, + "loss": 0.5829, + "step": 10332 + }, + { + "epoch": 3.6281601123595504, + "grad_norm": 0.9301666617393494, + "learning_rate": 1.7098031871032566e-06, + "loss": 0.5431, + "step": 10333 + }, + { + "epoch": 3.628511235955056, + "grad_norm": 0.927456259727478, + "learning_rate": 1.7093645773963784e-06, + "loss": 0.5736, + "step": 10334 + }, + { + "epoch": 3.6288623595505616, + "grad_norm": 0.9536610841751099, + "learning_rate": 1.7089259947285558e-06, + "loss": 0.5823, + "step": 10335 + }, + { + "epoch": 3.629213483146067, + "grad_norm": 0.937883734703064, + "learning_rate": 1.7084874391147866e-06, + "loss": 0.5468, + "step": 10336 + }, + { + "epoch": 3.6295646067415728, + "grad_norm": 0.9266161918640137, + "learning_rate": 1.7080489105700692e-06, + "loss": 0.5701, + "step": 10337 + }, + { + "epoch": 3.629915730337079, + "grad_norm": 0.946275532245636, + "learning_rate": 1.707610409109402e-06, + "loss": 0.5832, + "step": 10338 + }, + { + "epoch": 3.6302668539325844, + "grad_norm": 0.936488687992096, + "learning_rate": 1.7071719347477804e-06, + "loss": 0.5538, + "step": 10339 + }, + { + "epoch": 3.63061797752809, + "grad_norm": 0.923230767250061, + "learning_rate": 1.7067334875002e-06, + "loss": 0.5497, + "step": 10340 + }, + { + "epoch": 3.6309691011235956, + "grad_norm": 0.9313597679138184, + "learning_rate": 1.7062950673816558e-06, + "loss": 0.5455, + "step": 10341 + }, + { + "epoch": 3.631320224719101, + "grad_norm": 0.930945634841919, + "learning_rate": 1.7058566744071414e-06, + "loss": 0.5734, + "step": 10342 + }, + { + "epoch": 3.631671348314607, + "grad_norm": 0.9321323037147522, + "learning_rate": 1.7054183085916493e-06, + "loss": 0.5408, + "step": 10343 + }, + { + "epoch": 3.6320224719101124, + "grad_norm": 0.93021160364151, + "learning_rate": 1.70497996995017e-06, + "loss": 0.5481, + "step": 10344 + }, + { + "epoch": 3.632373595505618, + "grad_norm": 0.9149537086486816, + "learning_rate": 1.7045416584976967e-06, + "loss": 0.5442, + "step": 10345 + }, + { + "epoch": 3.6327247191011236, + "grad_norm": 0.9261180758476257, + "learning_rate": 1.7041033742492175e-06, + "loss": 0.5592, + "step": 10346 + }, + { + "epoch": 3.633075842696629, + "grad_norm": 0.9420327544212341, + "learning_rate": 1.7036651172197217e-06, + "loss": 0.5399, + "step": 10347 + }, + { + "epoch": 3.633426966292135, + "grad_norm": 0.9378608465194702, + "learning_rate": 1.7032268874241976e-06, + "loss": 0.5274, + "step": 10348 + }, + { + "epoch": 3.6337780898876404, + "grad_norm": 0.9549940228462219, + "learning_rate": 1.702788684877632e-06, + "loss": 0.553, + "step": 10349 + }, + { + "epoch": 3.634129213483146, + "grad_norm": 0.9735738635063171, + "learning_rate": 1.7023505095950116e-06, + "loss": 0.5542, + "step": 10350 + }, + { + "epoch": 3.6344803370786516, + "grad_norm": 0.9054490923881531, + "learning_rate": 1.7019123615913205e-06, + "loss": 0.5073, + "step": 10351 + }, + { + "epoch": 3.634831460674157, + "grad_norm": 0.9515958428382874, + "learning_rate": 1.7014742408815445e-06, + "loss": 0.5486, + "step": 10352 + }, + { + "epoch": 3.635182584269663, + "grad_norm": 0.9502704739570618, + "learning_rate": 1.7010361474806652e-06, + "loss": 0.5644, + "step": 10353 + }, + { + "epoch": 3.6355337078651684, + "grad_norm": 0.9294485449790955, + "learning_rate": 1.700598081403667e-06, + "loss": 0.5652, + "step": 10354 + }, + { + "epoch": 3.6358848314606744, + "grad_norm": 0.9273135662078857, + "learning_rate": 1.7001600426655294e-06, + "loss": 0.5729, + "step": 10355 + }, + { + "epoch": 3.63623595505618, + "grad_norm": 0.9739341139793396, + "learning_rate": 1.6997220312812348e-06, + "loss": 0.5659, + "step": 10356 + }, + { + "epoch": 3.6365870786516856, + "grad_norm": 0.9643154740333557, + "learning_rate": 1.699284047265761e-06, + "loss": 0.5565, + "step": 10357 + }, + { + "epoch": 3.636938202247191, + "grad_norm": 0.9796493053436279, + "learning_rate": 1.6988460906340887e-06, + "loss": 0.5792, + "step": 10358 + }, + { + "epoch": 3.637289325842697, + "grad_norm": 0.9390765428543091, + "learning_rate": 1.6984081614011944e-06, + "loss": 0.5946, + "step": 10359 + }, + { + "epoch": 3.6376404494382024, + "grad_norm": 0.9188821315765381, + "learning_rate": 1.697970259582054e-06, + "loss": 0.5422, + "step": 10360 + }, + { + "epoch": 3.637991573033708, + "grad_norm": 0.9443793892860413, + "learning_rate": 1.6975323851916453e-06, + "loss": 0.5642, + "step": 10361 + }, + { + "epoch": 3.6383426966292136, + "grad_norm": 0.9560198187828064, + "learning_rate": 1.6970945382449417e-06, + "loss": 0.5637, + "step": 10362 + }, + { + "epoch": 3.638693820224719, + "grad_norm": 0.9777402281761169, + "learning_rate": 1.696656718756919e-06, + "loss": 0.5692, + "step": 10363 + }, + { + "epoch": 3.639044943820225, + "grad_norm": 0.925990879535675, + "learning_rate": 1.6962189267425472e-06, + "loss": 0.5388, + "step": 10364 + }, + { + "epoch": 3.6393960674157304, + "grad_norm": 0.9584459066390991, + "learning_rate": 1.6957811622168019e-06, + "loss": 0.5839, + "step": 10365 + }, + { + "epoch": 3.639747191011236, + "grad_norm": 0.9241669774055481, + "learning_rate": 1.6953434251946515e-06, + "loss": 0.5782, + "step": 10366 + }, + { + "epoch": 3.6400983146067416, + "grad_norm": 0.966725766658783, + "learning_rate": 1.6949057156910687e-06, + "loss": 0.5739, + "step": 10367 + }, + { + "epoch": 3.640449438202247, + "grad_norm": 0.9189884066581726, + "learning_rate": 1.6944680337210203e-06, + "loss": 0.5616, + "step": 10368 + }, + { + "epoch": 3.640800561797753, + "grad_norm": 0.9217503070831299, + "learning_rate": 1.694030379299477e-06, + "loss": 0.5525, + "step": 10369 + }, + { + "epoch": 3.6411516853932584, + "grad_norm": 0.9097515940666199, + "learning_rate": 1.6935927524414038e-06, + "loss": 0.5463, + "step": 10370 + }, + { + "epoch": 3.641502808988764, + "grad_norm": 0.934689462184906, + "learning_rate": 1.6931551531617695e-06, + "loss": 0.571, + "step": 10371 + }, + { + "epoch": 3.6418539325842696, + "grad_norm": 0.9526573419570923, + "learning_rate": 1.6927175814755387e-06, + "loss": 0.5268, + "step": 10372 + }, + { + "epoch": 3.642205056179775, + "grad_norm": 0.9641138911247253, + "learning_rate": 1.6922800373976748e-06, + "loss": 0.5588, + "step": 10373 + }, + { + "epoch": 3.642556179775281, + "grad_norm": 0.9407922029495239, + "learning_rate": 1.6918425209431434e-06, + "loss": 0.5786, + "step": 10374 + }, + { + "epoch": 3.6429073033707864, + "grad_norm": 0.9690650701522827, + "learning_rate": 1.6914050321269049e-06, + "loss": 0.5222, + "step": 10375 + }, + { + "epoch": 3.643258426966292, + "grad_norm": 0.9235990047454834, + "learning_rate": 1.6909675709639236e-06, + "loss": 0.5214, + "step": 10376 + }, + { + "epoch": 3.6436095505617976, + "grad_norm": 0.9372032284736633, + "learning_rate": 1.690530137469158e-06, + "loss": 0.5363, + "step": 10377 + }, + { + "epoch": 3.643960674157303, + "grad_norm": 0.9052777290344238, + "learning_rate": 1.6900927316575704e-06, + "loss": 0.5156, + "step": 10378 + }, + { + "epoch": 3.644311797752809, + "grad_norm": 0.9340805411338806, + "learning_rate": 1.6896553535441168e-06, + "loss": 0.5916, + "step": 10379 + }, + { + "epoch": 3.6446629213483144, + "grad_norm": 0.9523031711578369, + "learning_rate": 1.6892180031437583e-06, + "loss": 0.5912, + "step": 10380 + }, + { + "epoch": 3.64501404494382, + "grad_norm": 0.9108738899230957, + "learning_rate": 1.6887806804714488e-06, + "loss": 0.545, + "step": 10381 + }, + { + "epoch": 3.6453651685393256, + "grad_norm": 0.9437102675437927, + "learning_rate": 1.688343385542147e-06, + "loss": 0.5594, + "step": 10382 + }, + { + "epoch": 3.6457162921348316, + "grad_norm": 0.9325085282325745, + "learning_rate": 1.6879061183708058e-06, + "loss": 0.5581, + "step": 10383 + }, + { + "epoch": 3.646067415730337, + "grad_norm": 0.9605008363723755, + "learning_rate": 1.6874688789723815e-06, + "loss": 0.5415, + "step": 10384 + }, + { + "epoch": 3.646418539325843, + "grad_norm": 0.9242297410964966, + "learning_rate": 1.6870316673618265e-06, + "loss": 0.5288, + "step": 10385 + }, + { + "epoch": 3.6467696629213484, + "grad_norm": 0.9371147155761719, + "learning_rate": 1.6865944835540916e-06, + "loss": 0.5663, + "step": 10386 + }, + { + "epoch": 3.647120786516854, + "grad_norm": 0.9123649597167969, + "learning_rate": 1.6861573275641302e-06, + "loss": 0.5133, + "step": 10387 + }, + { + "epoch": 3.6474719101123596, + "grad_norm": 0.8918284773826599, + "learning_rate": 1.6857201994068906e-06, + "loss": 0.5423, + "step": 10388 + }, + { + "epoch": 3.647823033707865, + "grad_norm": 0.9116357564926147, + "learning_rate": 1.6852830990973245e-06, + "loss": 0.5094, + "step": 10389 + }, + { + "epoch": 3.648174157303371, + "grad_norm": 0.9737594127655029, + "learning_rate": 1.6848460266503786e-06, + "loss": 0.5524, + "step": 10390 + }, + { + "epoch": 3.6485252808988764, + "grad_norm": 0.9336581826210022, + "learning_rate": 1.6844089820810016e-06, + "loss": 0.5657, + "step": 10391 + }, + { + "epoch": 3.648876404494382, + "grad_norm": 0.9270272850990295, + "learning_rate": 1.6839719654041388e-06, + "loss": 0.5399, + "step": 10392 + }, + { + "epoch": 3.6492275280898876, + "grad_norm": 0.94035404920578, + "learning_rate": 1.6835349766347376e-06, + "loss": 0.5469, + "step": 10393 + }, + { + "epoch": 3.649578651685393, + "grad_norm": 0.94282066822052, + "learning_rate": 1.6830980157877401e-06, + "loss": 0.5708, + "step": 10394 + }, + { + "epoch": 3.649929775280899, + "grad_norm": 0.934305727481842, + "learning_rate": 1.682661082878093e-06, + "loss": 0.5502, + "step": 10395 + }, + { + "epoch": 3.6502808988764044, + "grad_norm": 0.9369739890098572, + "learning_rate": 1.6822241779207364e-06, + "loss": 0.5664, + "step": 10396 + }, + { + "epoch": 3.65063202247191, + "grad_norm": 0.9209384322166443, + "learning_rate": 1.6817873009306137e-06, + "loss": 0.5623, + "step": 10397 + }, + { + "epoch": 3.6509831460674156, + "grad_norm": 0.92999666929245, + "learning_rate": 1.6813504519226654e-06, + "loss": 0.5569, + "step": 10398 + }, + { + "epoch": 3.651334269662921, + "grad_norm": 0.9128495454788208, + "learning_rate": 1.68091363091183e-06, + "loss": 0.5696, + "step": 10399 + }, + { + "epoch": 3.6516853932584272, + "grad_norm": 0.9661795496940613, + "learning_rate": 1.6804768379130488e-06, + "loss": 0.5525, + "step": 10400 + }, + { + "epoch": 3.652036516853933, + "grad_norm": 0.919048011302948, + "learning_rate": 1.6800400729412572e-06, + "loss": 0.5485, + "step": 10401 + }, + { + "epoch": 3.6523876404494384, + "grad_norm": 0.9382701516151428, + "learning_rate": 1.6796033360113945e-06, + "loss": 0.5467, + "step": 10402 + }, + { + "epoch": 3.652738764044944, + "grad_norm": 0.9452882409095764, + "learning_rate": 1.6791666271383947e-06, + "loss": 0.551, + "step": 10403 + }, + { + "epoch": 3.6530898876404496, + "grad_norm": 0.9263635277748108, + "learning_rate": 1.6787299463371948e-06, + "loss": 0.5342, + "step": 10404 + }, + { + "epoch": 3.653441011235955, + "grad_norm": 0.9128186702728271, + "learning_rate": 1.678293293622727e-06, + "loss": 0.557, + "step": 10405 + }, + { + "epoch": 3.653792134831461, + "grad_norm": 0.9254305362701416, + "learning_rate": 1.6778566690099265e-06, + "loss": 0.5469, + "step": 10406 + }, + { + "epoch": 3.6541432584269664, + "grad_norm": 0.9371973872184753, + "learning_rate": 1.677420072513724e-06, + "loss": 0.5393, + "step": 10407 + }, + { + "epoch": 3.654494382022472, + "grad_norm": 0.9559010863304138, + "learning_rate": 1.6769835041490506e-06, + "loss": 0.5487, + "step": 10408 + }, + { + "epoch": 3.6548455056179776, + "grad_norm": 0.9466186761856079, + "learning_rate": 1.6765469639308381e-06, + "loss": 0.5473, + "step": 10409 + }, + { + "epoch": 3.655196629213483, + "grad_norm": 0.9220945239067078, + "learning_rate": 1.6761104518740137e-06, + "loss": 0.5542, + "step": 10410 + }, + { + "epoch": 3.655547752808989, + "grad_norm": 0.9204149842262268, + "learning_rate": 1.675673967993508e-06, + "loss": 0.5488, + "step": 10411 + }, + { + "epoch": 3.6558988764044944, + "grad_norm": 0.9334843754768372, + "learning_rate": 1.6752375123042457e-06, + "loss": 0.5377, + "step": 10412 + }, + { + "epoch": 3.65625, + "grad_norm": 0.9263612627983093, + "learning_rate": 1.674801084821156e-06, + "loss": 0.5566, + "step": 10413 + }, + { + "epoch": 3.6566011235955056, + "grad_norm": 0.9593960046768188, + "learning_rate": 1.6743646855591628e-06, + "loss": 0.568, + "step": 10414 + }, + { + "epoch": 3.656952247191011, + "grad_norm": 0.9461026191711426, + "learning_rate": 1.6739283145331905e-06, + "loss": 0.5438, + "step": 10415 + }, + { + "epoch": 3.657303370786517, + "grad_norm": 0.9346083998680115, + "learning_rate": 1.6734919717581632e-06, + "loss": 0.5331, + "step": 10416 + }, + { + "epoch": 3.6576544943820224, + "grad_norm": 0.9119836091995239, + "learning_rate": 1.673055657249003e-06, + "loss": 0.5383, + "step": 10417 + }, + { + "epoch": 3.658005617977528, + "grad_norm": 0.934757649898529, + "learning_rate": 1.6726193710206318e-06, + "loss": 0.5522, + "step": 10418 + }, + { + "epoch": 3.6583567415730336, + "grad_norm": 0.9244552254676819, + "learning_rate": 1.6721831130879702e-06, + "loss": 0.55, + "step": 10419 + }, + { + "epoch": 3.658707865168539, + "grad_norm": 0.9654392600059509, + "learning_rate": 1.6717468834659377e-06, + "loss": 0.5499, + "step": 10420 + }, + { + "epoch": 3.659058988764045, + "grad_norm": 0.9525445699691772, + "learning_rate": 1.6713106821694536e-06, + "loss": 0.5389, + "step": 10421 + }, + { + "epoch": 3.6594101123595504, + "grad_norm": 0.9291754364967346, + "learning_rate": 1.6708745092134342e-06, + "loss": 0.5472, + "step": 10422 + }, + { + "epoch": 3.659761235955056, + "grad_norm": 0.9558712244033813, + "learning_rate": 1.6704383646127976e-06, + "loss": 0.5676, + "step": 10423 + }, + { + "epoch": 3.6601123595505616, + "grad_norm": 1.0040466785430908, + "learning_rate": 1.6700022483824589e-06, + "loss": 0.5421, + "step": 10424 + }, + { + "epoch": 3.660463483146067, + "grad_norm": 0.9507738947868347, + "learning_rate": 1.6695661605373329e-06, + "loss": 0.5478, + "step": 10425 + }, + { + "epoch": 3.6608146067415728, + "grad_norm": 0.9188863039016724, + "learning_rate": 1.6691301010923339e-06, + "loss": 0.5462, + "step": 10426 + }, + { + "epoch": 3.661165730337079, + "grad_norm": 0.9632121324539185, + "learning_rate": 1.668694070062374e-06, + "loss": 0.5656, + "step": 10427 + }, + { + "epoch": 3.6615168539325844, + "grad_norm": 0.9639596343040466, + "learning_rate": 1.6682580674623659e-06, + "loss": 0.5599, + "step": 10428 + }, + { + "epoch": 3.66186797752809, + "grad_norm": 0.9607666730880737, + "learning_rate": 1.6678220933072198e-06, + "loss": 0.5941, + "step": 10429 + }, + { + "epoch": 3.6622191011235956, + "grad_norm": 0.9032201170921326, + "learning_rate": 1.667386147611847e-06, + "loss": 0.5421, + "step": 10430 + }, + { + "epoch": 3.662570224719101, + "grad_norm": 0.9435725212097168, + "learning_rate": 1.666950230391154e-06, + "loss": 0.5283, + "step": 10431 + }, + { + "epoch": 3.662921348314607, + "grad_norm": 0.887864351272583, + "learning_rate": 1.6665143416600516e-06, + "loss": 0.5075, + "step": 10432 + }, + { + "epoch": 3.6632724719101124, + "grad_norm": 0.874033510684967, + "learning_rate": 1.6660784814334443e-06, + "loss": 0.5175, + "step": 10433 + }, + { + "epoch": 3.663623595505618, + "grad_norm": 0.8868166208267212, + "learning_rate": 1.6656426497262408e-06, + "loss": 0.5001, + "step": 10434 + }, + { + "epoch": 3.6639747191011236, + "grad_norm": 0.9332330822944641, + "learning_rate": 1.6652068465533436e-06, + "loss": 0.5724, + "step": 10435 + }, + { + "epoch": 3.664325842696629, + "grad_norm": 0.9384379982948303, + "learning_rate": 1.6647710719296582e-06, + "loss": 0.5749, + "step": 10436 + }, + { + "epoch": 3.664676966292135, + "grad_norm": 0.9640035033226013, + "learning_rate": 1.664335325870088e-06, + "loss": 0.5526, + "step": 10437 + }, + { + "epoch": 3.6650280898876404, + "grad_norm": 0.9176966547966003, + "learning_rate": 1.6638996083895332e-06, + "loss": 0.5696, + "step": 10438 + }, + { + "epoch": 3.665379213483146, + "grad_norm": 0.9119278192520142, + "learning_rate": 1.6634639195028976e-06, + "loss": 0.5403, + "step": 10439 + }, + { + "epoch": 3.6657303370786516, + "grad_norm": 0.8898779153823853, + "learning_rate": 1.663028259225079e-06, + "loss": 0.5348, + "step": 10440 + }, + { + "epoch": 3.666081460674157, + "grad_norm": 0.9395107626914978, + "learning_rate": 1.662592627570979e-06, + "loss": 0.5568, + "step": 10441 + }, + { + "epoch": 3.666432584269663, + "grad_norm": 0.9604291319847107, + "learning_rate": 1.6621570245554932e-06, + "loss": 0.5563, + "step": 10442 + }, + { + "epoch": 3.6667837078651684, + "grad_norm": 0.9661058783531189, + "learning_rate": 1.6617214501935214e-06, + "loss": 0.5827, + "step": 10443 + }, + { + "epoch": 3.6671348314606744, + "grad_norm": 0.951805830001831, + "learning_rate": 1.6612859044999578e-06, + "loss": 0.5575, + "step": 10444 + }, + { + "epoch": 3.66748595505618, + "grad_norm": 0.9112693071365356, + "learning_rate": 1.6608503874896992e-06, + "loss": 0.5318, + "step": 10445 + }, + { + "epoch": 3.6678370786516856, + "grad_norm": 0.9626321196556091, + "learning_rate": 1.6604148991776387e-06, + "loss": 0.5777, + "step": 10446 + }, + { + "epoch": 3.668188202247191, + "grad_norm": 0.9786010384559631, + "learning_rate": 1.659979439578671e-06, + "loss": 0.5401, + "step": 10447 + }, + { + "epoch": 3.668539325842697, + "grad_norm": 0.9816550016403198, + "learning_rate": 1.6595440087076875e-06, + "loss": 0.5574, + "step": 10448 + }, + { + "epoch": 3.6688904494382024, + "grad_norm": 0.9752625823020935, + "learning_rate": 1.659108606579579e-06, + "loss": 0.5788, + "step": 10449 + }, + { + "epoch": 3.669241573033708, + "grad_norm": 0.9331145882606506, + "learning_rate": 1.6586732332092376e-06, + "loss": 0.5138, + "step": 10450 + }, + { + "epoch": 3.6695926966292136, + "grad_norm": 0.9452927112579346, + "learning_rate": 1.6582378886115502e-06, + "loss": 0.5269, + "step": 10451 + }, + { + "epoch": 3.669943820224719, + "grad_norm": 0.9353452920913696, + "learning_rate": 1.657802572801408e-06, + "loss": 0.5425, + "step": 10452 + }, + { + "epoch": 3.670294943820225, + "grad_norm": 0.8967424035072327, + "learning_rate": 1.6573672857936962e-06, + "loss": 0.5374, + "step": 10453 + }, + { + "epoch": 3.6706460674157304, + "grad_norm": 0.9594841599464417, + "learning_rate": 1.6569320276033034e-06, + "loss": 0.5729, + "step": 10454 + }, + { + "epoch": 3.670997191011236, + "grad_norm": 0.931218683719635, + "learning_rate": 1.6564967982451125e-06, + "loss": 0.5419, + "step": 10455 + }, + { + "epoch": 3.6713483146067416, + "grad_norm": 0.9403542280197144, + "learning_rate": 1.6560615977340106e-06, + "loss": 0.5533, + "step": 10456 + }, + { + "epoch": 3.671699438202247, + "grad_norm": 0.9409517645835876, + "learning_rate": 1.6556264260848786e-06, + "loss": 0.5262, + "step": 10457 + }, + { + "epoch": 3.672050561797753, + "grad_norm": 0.9467124938964844, + "learning_rate": 1.6551912833126014e-06, + "loss": 0.5911, + "step": 10458 + }, + { + "epoch": 3.6724016853932584, + "grad_norm": 0.9378759860992432, + "learning_rate": 1.6547561694320583e-06, + "loss": 0.5766, + "step": 10459 + }, + { + "epoch": 3.672752808988764, + "grad_norm": 0.9560098052024841, + "learning_rate": 1.654321084458132e-06, + "loss": 0.541, + "step": 10460 + }, + { + "epoch": 3.6731039325842696, + "grad_norm": 0.9354994893074036, + "learning_rate": 1.653886028405701e-06, + "loss": 0.5372, + "step": 10461 + }, + { + "epoch": 3.673455056179775, + "grad_norm": 0.9480962753295898, + "learning_rate": 1.6534510012896427e-06, + "loss": 0.5778, + "step": 10462 + }, + { + "epoch": 3.673806179775281, + "grad_norm": 0.9233393669128418, + "learning_rate": 1.6530160031248366e-06, + "loss": 0.5521, + "step": 10463 + }, + { + "epoch": 3.6741573033707864, + "grad_norm": 0.9424656629562378, + "learning_rate": 1.6525810339261577e-06, + "loss": 0.5506, + "step": 10464 + }, + { + "epoch": 3.674508426966292, + "grad_norm": 0.9793898463249207, + "learning_rate": 1.6521460937084832e-06, + "loss": 0.5867, + "step": 10465 + }, + { + "epoch": 3.6748595505617976, + "grad_norm": 0.9263428449630737, + "learning_rate": 1.6517111824866858e-06, + "loss": 0.5356, + "step": 10466 + }, + { + "epoch": 3.675210674157303, + "grad_norm": 0.9442005157470703, + "learning_rate": 1.651276300275641e-06, + "loss": 0.556, + "step": 10467 + }, + { + "epoch": 3.675561797752809, + "grad_norm": 0.929419219493866, + "learning_rate": 1.6508414470902196e-06, + "loss": 0.5419, + "step": 10468 + }, + { + "epoch": 3.6759129213483144, + "grad_norm": 0.9190896153450012, + "learning_rate": 1.6504066229452951e-06, + "loss": 0.5645, + "step": 10469 + }, + { + "epoch": 3.67626404494382, + "grad_norm": 1.0218619108200073, + "learning_rate": 1.6499718278557358e-06, + "loss": 0.5766, + "step": 10470 + }, + { + "epoch": 3.6766151685393256, + "grad_norm": 0.9369935989379883, + "learning_rate": 1.6495370618364136e-06, + "loss": 0.4984, + "step": 10471 + }, + { + "epoch": 3.6769662921348316, + "grad_norm": 0.9164863228797913, + "learning_rate": 1.6491023249021964e-06, + "loss": 0.5414, + "step": 10472 + }, + { + "epoch": 3.677317415730337, + "grad_norm": 0.9822987914085388, + "learning_rate": 1.6486676170679506e-06, + "loss": 0.5661, + "step": 10473 + }, + { + "epoch": 3.677668539325843, + "grad_norm": 0.9262935519218445, + "learning_rate": 1.648232938348545e-06, + "loss": 0.5563, + "step": 10474 + }, + { + "epoch": 3.6780196629213484, + "grad_norm": 0.9273950457572937, + "learning_rate": 1.6477982887588425e-06, + "loss": 0.5385, + "step": 10475 + }, + { + "epoch": 3.678370786516854, + "grad_norm": 0.9658692479133606, + "learning_rate": 1.6473636683137107e-06, + "loss": 0.5631, + "step": 10476 + }, + { + "epoch": 3.6787219101123596, + "grad_norm": 0.9198076725006104, + "learning_rate": 1.6469290770280105e-06, + "loss": 0.5434, + "step": 10477 + }, + { + "epoch": 3.679073033707865, + "grad_norm": 0.9337387681007385, + "learning_rate": 1.646494514916607e-06, + "loss": 0.5737, + "step": 10478 + }, + { + "epoch": 3.679424157303371, + "grad_norm": 0.9351328015327454, + "learning_rate": 1.6460599819943595e-06, + "loss": 0.5342, + "step": 10479 + }, + { + "epoch": 3.6797752808988764, + "grad_norm": 0.9353929758071899, + "learning_rate": 1.6456254782761315e-06, + "loss": 0.5513, + "step": 10480 + }, + { + "epoch": 3.680126404494382, + "grad_norm": 0.9427953958511353, + "learning_rate": 1.6451910037767796e-06, + "loss": 0.5164, + "step": 10481 + }, + { + "epoch": 3.6804775280898876, + "grad_norm": 0.9409722089767456, + "learning_rate": 1.6447565585111652e-06, + "loss": 0.5287, + "step": 10482 + }, + { + "epoch": 3.680828651685393, + "grad_norm": 0.9378646016120911, + "learning_rate": 1.6443221424941433e-06, + "loss": 0.5394, + "step": 10483 + }, + { + "epoch": 3.681179775280899, + "grad_norm": 0.9324907660484314, + "learning_rate": 1.643887755740573e-06, + "loss": 0.5835, + "step": 10484 + }, + { + "epoch": 3.6815308988764044, + "grad_norm": 0.9349174499511719, + "learning_rate": 1.6434533982653089e-06, + "loss": 0.5701, + "step": 10485 + }, + { + "epoch": 3.68188202247191, + "grad_norm": 0.9643785953521729, + "learning_rate": 1.6430190700832046e-06, + "loss": 0.5413, + "step": 10486 + }, + { + "epoch": 3.6822331460674156, + "grad_norm": 0.9300989508628845, + "learning_rate": 1.6425847712091158e-06, + "loss": 0.5806, + "step": 10487 + }, + { + "epoch": 3.682584269662921, + "grad_norm": 0.9426717758178711, + "learning_rate": 1.642150501657893e-06, + "loss": 0.5421, + "step": 10488 + }, + { + "epoch": 3.6829353932584272, + "grad_norm": 0.9608815312385559, + "learning_rate": 1.6417162614443904e-06, + "loss": 0.5782, + "step": 10489 + }, + { + "epoch": 3.683286516853933, + "grad_norm": 0.9229572415351868, + "learning_rate": 1.641282050583456e-06, + "loss": 0.5478, + "step": 10490 + }, + { + "epoch": 3.6836376404494384, + "grad_norm": 0.973231852054596, + "learning_rate": 1.6408478690899415e-06, + "loss": 0.542, + "step": 10491 + }, + { + "epoch": 3.683988764044944, + "grad_norm": 0.9333351254463196, + "learning_rate": 1.6404137169786938e-06, + "loss": 0.5507, + "step": 10492 + }, + { + "epoch": 3.6843398876404496, + "grad_norm": 0.925819993019104, + "learning_rate": 1.6399795942645623e-06, + "loss": 0.5727, + "step": 10493 + }, + { + "epoch": 3.684691011235955, + "grad_norm": 0.9236595034599304, + "learning_rate": 1.6395455009623925e-06, + "loss": 0.5331, + "step": 10494 + }, + { + "epoch": 3.685042134831461, + "grad_norm": 0.9373668432235718, + "learning_rate": 1.63911143708703e-06, + "loss": 0.5883, + "step": 10495 + }, + { + "epoch": 3.6853932584269664, + "grad_norm": 0.9281242489814758, + "learning_rate": 1.6386774026533197e-06, + "loss": 0.5408, + "step": 10496 + }, + { + "epoch": 3.685744382022472, + "grad_norm": 0.9281338453292847, + "learning_rate": 1.6382433976761055e-06, + "loss": 0.5593, + "step": 10497 + }, + { + "epoch": 3.6860955056179776, + "grad_norm": 0.9310676455497742, + "learning_rate": 1.63780942217023e-06, + "loss": 0.5406, + "step": 10498 + }, + { + "epoch": 3.686446629213483, + "grad_norm": 0.9333351254463196, + "learning_rate": 1.6373754761505334e-06, + "loss": 0.5855, + "step": 10499 + }, + { + "epoch": 3.686797752808989, + "grad_norm": 0.909023106098175, + "learning_rate": 1.6369415596318582e-06, + "loss": 0.5893, + "step": 10500 + }, + { + "epoch": 3.6871488764044944, + "grad_norm": 0.9206759333610535, + "learning_rate": 1.6365076726290429e-06, + "loss": 0.5459, + "step": 10501 + }, + { + "epoch": 3.6875, + "grad_norm": 0.9401935935020447, + "learning_rate": 1.6360738151569263e-06, + "loss": 0.5781, + "step": 10502 + }, + { + "epoch": 3.6878511235955056, + "grad_norm": 0.9612536430358887, + "learning_rate": 1.6356399872303457e-06, + "loss": 0.5698, + "step": 10503 + }, + { + "epoch": 3.688202247191011, + "grad_norm": 0.9008697867393494, + "learning_rate": 1.6352061888641386e-06, + "loss": 0.5199, + "step": 10504 + }, + { + "epoch": 3.688553370786517, + "grad_norm": 0.9655016660690308, + "learning_rate": 1.634772420073139e-06, + "loss": 0.5657, + "step": 10505 + }, + { + "epoch": 3.6889044943820224, + "grad_norm": 0.9442702531814575, + "learning_rate": 1.6343386808721825e-06, + "loss": 0.5354, + "step": 10506 + }, + { + "epoch": 3.689255617977528, + "grad_norm": 0.9363181591033936, + "learning_rate": 1.6339049712761023e-06, + "loss": 0.5778, + "step": 10507 + }, + { + "epoch": 3.6896067415730336, + "grad_norm": 0.9297296404838562, + "learning_rate": 1.6334712912997316e-06, + "loss": 0.54, + "step": 10508 + }, + { + "epoch": 3.689957865168539, + "grad_norm": 0.9383668303489685, + "learning_rate": 1.6330376409579003e-06, + "loss": 0.5216, + "step": 10509 + }, + { + "epoch": 3.690308988764045, + "grad_norm": 0.9032675623893738, + "learning_rate": 1.6326040202654409e-06, + "loss": 0.5241, + "step": 10510 + }, + { + "epoch": 3.6906601123595504, + "grad_norm": 0.9352113604545593, + "learning_rate": 1.6321704292371809e-06, + "loss": 0.5416, + "step": 10511 + }, + { + "epoch": 3.691011235955056, + "grad_norm": 0.9462507367134094, + "learning_rate": 1.6317368678879497e-06, + "loss": 0.5499, + "step": 10512 + }, + { + "epoch": 3.6913623595505616, + "grad_norm": 0.9889009594917297, + "learning_rate": 1.6313033362325747e-06, + "loss": 0.5975, + "step": 10513 + }, + { + "epoch": 3.691713483146067, + "grad_norm": 0.9450020790100098, + "learning_rate": 1.6308698342858823e-06, + "loss": 0.5927, + "step": 10514 + }, + { + "epoch": 3.6920646067415728, + "grad_norm": 0.9534556269645691, + "learning_rate": 1.6304363620626979e-06, + "loss": 0.5844, + "step": 10515 + }, + { + "epoch": 3.692415730337079, + "grad_norm": 0.9159740805625916, + "learning_rate": 1.6300029195778454e-06, + "loss": 0.5756, + "step": 10516 + }, + { + "epoch": 3.6927668539325844, + "grad_norm": 0.9577197432518005, + "learning_rate": 1.6295695068461499e-06, + "loss": 0.5665, + "step": 10517 + }, + { + "epoch": 3.69311797752809, + "grad_norm": 0.937190592288971, + "learning_rate": 1.6291361238824305e-06, + "loss": 0.5428, + "step": 10518 + }, + { + "epoch": 3.6934691011235956, + "grad_norm": 0.9153808951377869, + "learning_rate": 1.628702770701512e-06, + "loss": 0.5602, + "step": 10519 + }, + { + "epoch": 3.693820224719101, + "grad_norm": 0.9671076536178589, + "learning_rate": 1.628269447318212e-06, + "loss": 0.5608, + "step": 10520 + }, + { + "epoch": 3.694171348314607, + "grad_norm": 0.9625701904296875, + "learning_rate": 1.6278361537473519e-06, + "loss": 0.5677, + "step": 10521 + }, + { + "epoch": 3.6945224719101124, + "grad_norm": 0.913791298866272, + "learning_rate": 1.627402890003748e-06, + "loss": 0.5304, + "step": 10522 + }, + { + "epoch": 3.694873595505618, + "grad_norm": 0.9547209143638611, + "learning_rate": 1.62696965610222e-06, + "loss": 0.5401, + "step": 10523 + }, + { + "epoch": 3.6952247191011236, + "grad_norm": 0.9616701602935791, + "learning_rate": 1.6265364520575823e-06, + "loss": 0.5584, + "step": 10524 + }, + { + "epoch": 3.695575842696629, + "grad_norm": 0.9187096953392029, + "learning_rate": 1.6261032778846498e-06, + "loss": 0.5572, + "step": 10525 + }, + { + "epoch": 3.695926966292135, + "grad_norm": 0.9231456518173218, + "learning_rate": 1.6256701335982381e-06, + "loss": 0.5618, + "step": 10526 + }, + { + "epoch": 3.6962780898876404, + "grad_norm": 0.923202395439148, + "learning_rate": 1.6252370192131586e-06, + "loss": 0.5323, + "step": 10527 + }, + { + "epoch": 3.696629213483146, + "grad_norm": 0.9483276605606079, + "learning_rate": 1.6248039347442258e-06, + "loss": 0.5798, + "step": 10528 + }, + { + "epoch": 3.6969803370786516, + "grad_norm": 0.9359333515167236, + "learning_rate": 1.624370880206248e-06, + "loss": 0.5842, + "step": 10529 + }, + { + "epoch": 3.697331460674157, + "grad_norm": 0.9602817296981812, + "learning_rate": 1.6239378556140383e-06, + "loss": 0.5964, + "step": 10530 + }, + { + "epoch": 3.697682584269663, + "grad_norm": 0.9108102917671204, + "learning_rate": 1.6235048609824033e-06, + "loss": 0.5459, + "step": 10531 + }, + { + "epoch": 3.6980337078651684, + "grad_norm": 0.9500235319137573, + "learning_rate": 1.6230718963261527e-06, + "loss": 0.5361, + "step": 10532 + }, + { + "epoch": 3.6983848314606744, + "grad_norm": 0.9382598400115967, + "learning_rate": 1.622638961660092e-06, + "loss": 0.5977, + "step": 10533 + }, + { + "epoch": 3.69873595505618, + "grad_norm": 0.9336915016174316, + "learning_rate": 1.6222060569990293e-06, + "loss": 0.5578, + "step": 10534 + }, + { + "epoch": 3.6990870786516856, + "grad_norm": 0.9431601762771606, + "learning_rate": 1.621773182357768e-06, + "loss": 0.5942, + "step": 10535 + }, + { + "epoch": 3.699438202247191, + "grad_norm": 1.0670770406723022, + "learning_rate": 1.6213403377511115e-06, + "loss": 0.5493, + "step": 10536 + }, + { + "epoch": 3.699789325842697, + "grad_norm": 0.9157713651657104, + "learning_rate": 1.6209075231938643e-06, + "loss": 0.5506, + "step": 10537 + }, + { + "epoch": 3.7001404494382024, + "grad_norm": 0.9612889289855957, + "learning_rate": 1.6204747387008268e-06, + "loss": 0.595, + "step": 10538 + }, + { + "epoch": 3.700491573033708, + "grad_norm": 0.9259245991706848, + "learning_rate": 1.6200419842868015e-06, + "loss": 0.5376, + "step": 10539 + }, + { + "epoch": 3.7008426966292136, + "grad_norm": 0.9388218522071838, + "learning_rate": 1.6196092599665863e-06, + "loss": 0.5193, + "step": 10540 + }, + { + "epoch": 3.701193820224719, + "grad_norm": 1.0013699531555176, + "learning_rate": 1.6191765657549824e-06, + "loss": 0.5501, + "step": 10541 + }, + { + "epoch": 3.701544943820225, + "grad_norm": 0.9103558659553528, + "learning_rate": 1.618743901666785e-06, + "loss": 0.5276, + "step": 10542 + }, + { + "epoch": 3.7018960674157304, + "grad_norm": 0.939130961894989, + "learning_rate": 1.6183112677167929e-06, + "loss": 0.5624, + "step": 10543 + }, + { + "epoch": 3.702247191011236, + "grad_norm": 0.9693638682365417, + "learning_rate": 1.6178786639198003e-06, + "loss": 0.5748, + "step": 10544 + }, + { + "epoch": 3.7025983146067416, + "grad_norm": 0.9593151211738586, + "learning_rate": 1.6174460902906031e-06, + "loss": 0.568, + "step": 10545 + }, + { + "epoch": 3.702949438202247, + "grad_norm": 0.9666921496391296, + "learning_rate": 1.6170135468439937e-06, + "loss": 0.5624, + "step": 10546 + }, + { + "epoch": 3.703300561797753, + "grad_norm": 0.9075769782066345, + "learning_rate": 1.6165810335947664e-06, + "loss": 0.5319, + "step": 10547 + }, + { + "epoch": 3.7036516853932584, + "grad_norm": 0.9699574112892151, + "learning_rate": 1.6161485505577117e-06, + "loss": 0.5588, + "step": 10548 + }, + { + "epoch": 3.704002808988764, + "grad_norm": 0.9229788780212402, + "learning_rate": 1.6157160977476193e-06, + "loss": 0.5682, + "step": 10549 + }, + { + "epoch": 3.7043539325842696, + "grad_norm": 0.9924206137657166, + "learning_rate": 1.615283675179281e-06, + "loss": 0.5591, + "step": 10550 + }, + { + "epoch": 3.704705056179775, + "grad_norm": 0.9105188250541687, + "learning_rate": 1.6148512828674825e-06, + "loss": 0.5405, + "step": 10551 + }, + { + "epoch": 3.705056179775281, + "grad_norm": 0.9407828450202942, + "learning_rate": 1.614418920827014e-06, + "loss": 0.5691, + "step": 10552 + }, + { + "epoch": 3.7054073033707864, + "grad_norm": 0.902071475982666, + "learning_rate": 1.6139865890726592e-06, + "loss": 0.5628, + "step": 10553 + }, + { + "epoch": 3.705758426966292, + "grad_norm": 0.986543595790863, + "learning_rate": 1.6135542876192063e-06, + "loss": 0.5537, + "step": 10554 + }, + { + "epoch": 3.7061095505617976, + "grad_norm": 0.9592205286026001, + "learning_rate": 1.613122016481437e-06, + "loss": 0.5332, + "step": 10555 + }, + { + "epoch": 3.706460674157303, + "grad_norm": 0.9814830422401428, + "learning_rate": 1.6126897756741367e-06, + "loss": 0.596, + "step": 10556 + }, + { + "epoch": 3.706811797752809, + "grad_norm": 0.9216854572296143, + "learning_rate": 1.612257565212086e-06, + "loss": 0.5235, + "step": 10557 + }, + { + "epoch": 3.7071629213483144, + "grad_norm": 0.9562729001045227, + "learning_rate": 1.611825385110068e-06, + "loss": 0.4864, + "step": 10558 + }, + { + "epoch": 3.70751404494382, + "grad_norm": 0.9615098834037781, + "learning_rate": 1.6113932353828604e-06, + "loss": 0.573, + "step": 10559 + }, + { + "epoch": 3.7078651685393256, + "grad_norm": 0.9580915570259094, + "learning_rate": 1.6109611160452454e-06, + "loss": 0.5758, + "step": 10560 + }, + { + "epoch": 3.7082162921348316, + "grad_norm": 0.9231317043304443, + "learning_rate": 1.610529027111999e-06, + "loss": 0.5871, + "step": 10561 + }, + { + "epoch": 3.708567415730337, + "grad_norm": 0.9523150324821472, + "learning_rate": 1.610096968597898e-06, + "loss": 0.5876, + "step": 10562 + }, + { + "epoch": 3.708918539325843, + "grad_norm": 0.9217310547828674, + "learning_rate": 1.60966494051772e-06, + "loss": 0.5581, + "step": 10563 + }, + { + "epoch": 3.7092696629213484, + "grad_norm": 0.9322973489761353, + "learning_rate": 1.609232942886238e-06, + "loss": 0.5239, + "step": 10564 + }, + { + "epoch": 3.709620786516854, + "grad_norm": 0.9320168495178223, + "learning_rate": 1.6088009757182282e-06, + "loss": 0.5792, + "step": 10565 + }, + { + "epoch": 3.7099719101123596, + "grad_norm": 0.9518136978149414, + "learning_rate": 1.6083690390284617e-06, + "loss": 0.5722, + "step": 10566 + }, + { + "epoch": 3.710323033707865, + "grad_norm": 0.9317528009414673, + "learning_rate": 1.6079371328317117e-06, + "loss": 0.587, + "step": 10567 + }, + { + "epoch": 3.710674157303371, + "grad_norm": 0.9149349331855774, + "learning_rate": 1.6075052571427477e-06, + "loss": 0.5326, + "step": 10568 + }, + { + "epoch": 3.7110252808988764, + "grad_norm": 0.9086408019065857, + "learning_rate": 1.6070734119763413e-06, + "loss": 0.5485, + "step": 10569 + }, + { + "epoch": 3.711376404494382, + "grad_norm": 0.9480844140052795, + "learning_rate": 1.6066415973472587e-06, + "loss": 0.564, + "step": 10570 + }, + { + "epoch": 3.7117275280898876, + "grad_norm": 0.9564575552940369, + "learning_rate": 1.6062098132702703e-06, + "loss": 0.596, + "step": 10571 + }, + { + "epoch": 3.712078651685393, + "grad_norm": 0.9495142698287964, + "learning_rate": 1.6057780597601401e-06, + "loss": 0.5785, + "step": 10572 + }, + { + "epoch": 3.712429775280899, + "grad_norm": 0.9603184461593628, + "learning_rate": 1.6053463368316367e-06, + "loss": 0.5876, + "step": 10573 + }, + { + "epoch": 3.7127808988764044, + "grad_norm": 0.9414350390434265, + "learning_rate": 1.6049146444995225e-06, + "loss": 0.5772, + "step": 10574 + }, + { + "epoch": 3.71313202247191, + "grad_norm": 0.9227333068847656, + "learning_rate": 1.6044829827785607e-06, + "loss": 0.5873, + "step": 10575 + }, + { + "epoch": 3.7134831460674156, + "grad_norm": 0.9541880488395691, + "learning_rate": 1.6040513516835158e-06, + "loss": 0.5912, + "step": 10576 + }, + { + "epoch": 3.713834269662921, + "grad_norm": 0.9479333162307739, + "learning_rate": 1.6036197512291467e-06, + "loss": 0.5192, + "step": 10577 + }, + { + "epoch": 3.7141853932584272, + "grad_norm": 0.9799323678016663, + "learning_rate": 1.603188181430216e-06, + "loss": 0.5426, + "step": 10578 + }, + { + "epoch": 3.714536516853933, + "grad_norm": 0.9110511541366577, + "learning_rate": 1.6027566423014812e-06, + "loss": 0.5345, + "step": 10579 + }, + { + "epoch": 3.7148876404494384, + "grad_norm": 0.9352244138717651, + "learning_rate": 1.6023251338577025e-06, + "loss": 0.5508, + "step": 10580 + }, + { + "epoch": 3.715238764044944, + "grad_norm": 0.9439874887466431, + "learning_rate": 1.6018936561136355e-06, + "loss": 0.5443, + "step": 10581 + }, + { + "epoch": 3.7155898876404496, + "grad_norm": 0.9365434050559998, + "learning_rate": 1.601462209084037e-06, + "loss": 0.55, + "step": 10582 + }, + { + "epoch": 3.715941011235955, + "grad_norm": 0.9551948308944702, + "learning_rate": 1.601030792783662e-06, + "loss": 0.5482, + "step": 10583 + }, + { + "epoch": 3.716292134831461, + "grad_norm": 0.9546279907226562, + "learning_rate": 1.6005994072272645e-06, + "loss": 0.5501, + "step": 10584 + }, + { + "epoch": 3.7166432584269664, + "grad_norm": 1.0423859357833862, + "learning_rate": 1.6001680524295983e-06, + "loss": 0.5242, + "step": 10585 + }, + { + "epoch": 3.716994382022472, + "grad_norm": 0.9398695230484009, + "learning_rate": 1.5997367284054138e-06, + "loss": 0.5365, + "step": 10586 + }, + { + "epoch": 3.7173455056179776, + "grad_norm": 0.9232392311096191, + "learning_rate": 1.599305435169464e-06, + "loss": 0.5642, + "step": 10587 + }, + { + "epoch": 3.717696629213483, + "grad_norm": 0.9366011619567871, + "learning_rate": 1.5988741727364968e-06, + "loss": 0.5468, + "step": 10588 + }, + { + "epoch": 3.718047752808989, + "grad_norm": 0.9445890784263611, + "learning_rate": 1.5984429411212617e-06, + "loss": 0.5605, + "step": 10589 + }, + { + "epoch": 3.7183988764044944, + "grad_norm": 0.9199753999710083, + "learning_rate": 1.5980117403385068e-06, + "loss": 0.5414, + "step": 10590 + }, + { + "epoch": 3.71875, + "grad_norm": 0.9045284390449524, + "learning_rate": 1.5975805704029785e-06, + "loss": 0.5529, + "step": 10591 + }, + { + "epoch": 3.7191011235955056, + "grad_norm": 0.9331473708152771, + "learning_rate": 1.5971494313294223e-06, + "loss": 0.5643, + "step": 10592 + }, + { + "epoch": 3.719452247191011, + "grad_norm": 0.9333986639976501, + "learning_rate": 1.5967183231325829e-06, + "loss": 0.5907, + "step": 10593 + }, + { + "epoch": 3.719803370786517, + "grad_norm": 0.920032799243927, + "learning_rate": 1.5962872458272042e-06, + "loss": 0.514, + "step": 10594 + }, + { + "epoch": 3.7201544943820224, + "grad_norm": 0.9101995229721069, + "learning_rate": 1.5958561994280287e-06, + "loss": 0.5402, + "step": 10595 + }, + { + "epoch": 3.720505617977528, + "grad_norm": 0.9490725994110107, + "learning_rate": 1.5954251839497964e-06, + "loss": 0.5216, + "step": 10596 + }, + { + "epoch": 3.7208567415730336, + "grad_norm": 0.9377645254135132, + "learning_rate": 1.5949941994072498e-06, + "loss": 0.5779, + "step": 10597 + }, + { + "epoch": 3.721207865168539, + "grad_norm": 0.9533388018608093, + "learning_rate": 1.5945632458151266e-06, + "loss": 0.5267, + "step": 10598 + }, + { + "epoch": 3.721558988764045, + "grad_norm": 0.9614124298095703, + "learning_rate": 1.5941323231881662e-06, + "loss": 0.5993, + "step": 10599 + }, + { + "epoch": 3.7219101123595504, + "grad_norm": 0.9291122555732727, + "learning_rate": 1.5937014315411042e-06, + "loss": 0.5386, + "step": 10600 + }, + { + "epoch": 3.722261235955056, + "grad_norm": 0.9296464920043945, + "learning_rate": 1.5932705708886786e-06, + "loss": 0.5503, + "step": 10601 + }, + { + "epoch": 3.7226123595505616, + "grad_norm": 0.9358545541763306, + "learning_rate": 1.5928397412456231e-06, + "loss": 0.5588, + "step": 10602 + }, + { + "epoch": 3.722963483146067, + "grad_norm": 0.9344218969345093, + "learning_rate": 1.592408942626672e-06, + "loss": 0.5611, + "step": 10603 + }, + { + "epoch": 3.7233146067415728, + "grad_norm": 0.9412795305252075, + "learning_rate": 1.5919781750465596e-06, + "loss": 0.5496, + "step": 10604 + }, + { + "epoch": 3.723665730337079, + "grad_norm": 0.9462301731109619, + "learning_rate": 1.591547438520015e-06, + "loss": 0.5518, + "step": 10605 + }, + { + "epoch": 3.7240168539325844, + "grad_norm": 0.9579633474349976, + "learning_rate": 1.591116733061772e-06, + "loss": 0.5663, + "step": 10606 + }, + { + "epoch": 3.72436797752809, + "grad_norm": 0.9176328182220459, + "learning_rate": 1.5906860586865574e-06, + "loss": 0.5468, + "step": 10607 + }, + { + "epoch": 3.7247191011235956, + "grad_norm": 0.9314478635787964, + "learning_rate": 1.5902554154091031e-06, + "loss": 0.5672, + "step": 10608 + }, + { + "epoch": 3.725070224719101, + "grad_norm": 0.9880291223526001, + "learning_rate": 1.5898248032441338e-06, + "loss": 0.578, + "step": 10609 + }, + { + "epoch": 3.725421348314607, + "grad_norm": 0.9414797425270081, + "learning_rate": 1.5893942222063785e-06, + "loss": 0.5907, + "step": 10610 + }, + { + "epoch": 3.7257724719101124, + "grad_norm": 0.9324955344200134, + "learning_rate": 1.5889636723105618e-06, + "loss": 0.5434, + "step": 10611 + }, + { + "epoch": 3.726123595505618, + "grad_norm": 0.9632052183151245, + "learning_rate": 1.5885331535714065e-06, + "loss": 0.6048, + "step": 10612 + }, + { + "epoch": 3.7264747191011236, + "grad_norm": 0.9265936613082886, + "learning_rate": 1.5881026660036387e-06, + "loss": 0.551, + "step": 10613 + }, + { + "epoch": 3.726825842696629, + "grad_norm": 0.9645761251449585, + "learning_rate": 1.5876722096219781e-06, + "loss": 0.5597, + "step": 10614 + }, + { + "epoch": 3.727176966292135, + "grad_norm": 0.9448643326759338, + "learning_rate": 1.5872417844411487e-06, + "loss": 0.5536, + "step": 10615 + }, + { + "epoch": 3.7275280898876404, + "grad_norm": 0.9402912259101868, + "learning_rate": 1.586811390475868e-06, + "loss": 0.5671, + "step": 10616 + }, + { + "epoch": 3.727879213483146, + "grad_norm": 0.9387921690940857, + "learning_rate": 1.586381027740857e-06, + "loss": 0.515, + "step": 10617 + }, + { + "epoch": 3.7282303370786516, + "grad_norm": 0.9861497282981873, + "learning_rate": 1.5859506962508325e-06, + "loss": 0.5832, + "step": 10618 + }, + { + "epoch": 3.728581460674157, + "grad_norm": 0.9264570474624634, + "learning_rate": 1.5855203960205128e-06, + "loss": 0.5725, + "step": 10619 + }, + { + "epoch": 3.728932584269663, + "grad_norm": 0.9177839756011963, + "learning_rate": 1.585090127064612e-06, + "loss": 0.5161, + "step": 10620 + }, + { + "epoch": 3.7292837078651684, + "grad_norm": 0.9530559182167053, + "learning_rate": 1.584659889397847e-06, + "loss": 0.55, + "step": 10621 + }, + { + "epoch": 3.7296348314606744, + "grad_norm": 0.9386327862739563, + "learning_rate": 1.5842296830349295e-06, + "loss": 0.5544, + "step": 10622 + }, + { + "epoch": 3.72998595505618, + "grad_norm": 0.9604566097259521, + "learning_rate": 1.5837995079905743e-06, + "loss": 0.5611, + "step": 10623 + }, + { + "epoch": 3.7303370786516856, + "grad_norm": 0.977691113948822, + "learning_rate": 1.583369364279492e-06, + "loss": 0.5585, + "step": 10624 + }, + { + "epoch": 3.730688202247191, + "grad_norm": 0.9333210587501526, + "learning_rate": 1.582939251916392e-06, + "loss": 0.5475, + "step": 10625 + }, + { + "epoch": 3.731039325842697, + "grad_norm": 0.9364824891090393, + "learning_rate": 1.5825091709159856e-06, + "loss": 0.5353, + "step": 10626 + }, + { + "epoch": 3.7313904494382024, + "grad_norm": 0.931216835975647, + "learning_rate": 1.5820791212929798e-06, + "loss": 0.5449, + "step": 10627 + }, + { + "epoch": 3.731741573033708, + "grad_norm": 0.9222827553749084, + "learning_rate": 1.5816491030620834e-06, + "loss": 0.5653, + "step": 10628 + }, + { + "epoch": 3.7320926966292136, + "grad_norm": 0.9535141587257385, + "learning_rate": 1.581219116238001e-06, + "loss": 0.5755, + "step": 10629 + }, + { + "epoch": 3.732443820224719, + "grad_norm": 0.97788006067276, + "learning_rate": 1.5807891608354392e-06, + "loss": 0.5594, + "step": 10630 + }, + { + "epoch": 3.732794943820225, + "grad_norm": 0.9392472505569458, + "learning_rate": 1.580359236869101e-06, + "loss": 0.5662, + "step": 10631 + }, + { + "epoch": 3.7331460674157304, + "grad_norm": 0.9539934396743774, + "learning_rate": 1.5799293443536905e-06, + "loss": 0.5935, + "step": 10632 + }, + { + "epoch": 3.733497191011236, + "grad_norm": 0.9311276078224182, + "learning_rate": 1.5794994833039084e-06, + "loss": 0.5375, + "step": 10633 + }, + { + "epoch": 3.7338483146067416, + "grad_norm": 0.9635607004165649, + "learning_rate": 1.5790696537344574e-06, + "loss": 0.5615, + "step": 10634 + }, + { + "epoch": 3.734199438202247, + "grad_norm": 0.9642311930656433, + "learning_rate": 1.5786398556600346e-06, + "loss": 0.5802, + "step": 10635 + }, + { + "epoch": 3.734550561797753, + "grad_norm": 0.9364364743232727, + "learning_rate": 1.5782100890953419e-06, + "loss": 0.5519, + "step": 10636 + }, + { + "epoch": 3.7349016853932584, + "grad_norm": 0.955528736114502, + "learning_rate": 1.5777803540550752e-06, + "loss": 0.5383, + "step": 10637 + }, + { + "epoch": 3.735252808988764, + "grad_norm": 0.9443372488021851, + "learning_rate": 1.5773506505539304e-06, + "loss": 0.5812, + "step": 10638 + }, + { + "epoch": 3.7356039325842696, + "grad_norm": 0.9313518404960632, + "learning_rate": 1.5769209786066047e-06, + "loss": 0.5536, + "step": 10639 + }, + { + "epoch": 3.735955056179775, + "grad_norm": 0.9271915555000305, + "learning_rate": 1.5764913382277904e-06, + "loss": 0.5368, + "step": 10640 + }, + { + "epoch": 3.736306179775281, + "grad_norm": 0.9330967664718628, + "learning_rate": 1.5760617294321833e-06, + "loss": 0.5319, + "step": 10641 + }, + { + "epoch": 3.7366573033707864, + "grad_norm": 0.9497278332710266, + "learning_rate": 1.5756321522344733e-06, + "loss": 0.541, + "step": 10642 + }, + { + "epoch": 3.737008426966292, + "grad_norm": 0.9632142186164856, + "learning_rate": 1.575202606649354e-06, + "loss": 0.5701, + "step": 10643 + }, + { + "epoch": 3.7373595505617976, + "grad_norm": 0.9908214807510376, + "learning_rate": 1.5747730926915128e-06, + "loss": 0.573, + "step": 10644 + }, + { + "epoch": 3.737710674157303, + "grad_norm": 1.1243443489074707, + "learning_rate": 1.5743436103756415e-06, + "loss": 0.5372, + "step": 10645 + }, + { + "epoch": 3.738061797752809, + "grad_norm": 0.9059547185897827, + "learning_rate": 1.5739141597164254e-06, + "loss": 0.5291, + "step": 10646 + }, + { + "epoch": 3.7384129213483144, + "grad_norm": 0.9413697719573975, + "learning_rate": 1.5734847407285535e-06, + "loss": 0.5449, + "step": 10647 + }, + { + "epoch": 3.73876404494382, + "grad_norm": 0.9345862865447998, + "learning_rate": 1.5730553534267095e-06, + "loss": 0.5514, + "step": 10648 + }, + { + "epoch": 3.7391151685393256, + "grad_norm": 0.9088484644889832, + "learning_rate": 1.572625997825581e-06, + "loss": 0.5789, + "step": 10649 + }, + { + "epoch": 3.7394662921348316, + "grad_norm": 0.934416651725769, + "learning_rate": 1.5721966739398492e-06, + "loss": 0.5693, + "step": 10650 + }, + { + "epoch": 3.739817415730337, + "grad_norm": 0.9354353547096252, + "learning_rate": 1.571767381784196e-06, + "loss": 0.5367, + "step": 10651 + }, + { + "epoch": 3.740168539325843, + "grad_norm": 0.930545449256897, + "learning_rate": 1.5713381213733054e-06, + "loss": 0.5605, + "step": 10652 + }, + { + "epoch": 3.7405196629213484, + "grad_norm": 0.9724206924438477, + "learning_rate": 1.5709088927218552e-06, + "loss": 0.5505, + "step": 10653 + }, + { + "epoch": 3.740870786516854, + "grad_norm": 0.9474396109580994, + "learning_rate": 1.570479695844527e-06, + "loss": 0.5607, + "step": 10654 + }, + { + "epoch": 3.7412219101123596, + "grad_norm": 0.9180148243904114, + "learning_rate": 1.5700505307559965e-06, + "loss": 0.5289, + "step": 10655 + }, + { + "epoch": 3.741573033707865, + "grad_norm": 0.9182876944541931, + "learning_rate": 1.5696213974709435e-06, + "loss": 0.5592, + "step": 10656 + }, + { + "epoch": 3.741924157303371, + "grad_norm": 0.9211063981056213, + "learning_rate": 1.5691922960040416e-06, + "loss": 0.5418, + "step": 10657 + }, + { + "epoch": 3.7422752808988764, + "grad_norm": 0.9261342883110046, + "learning_rate": 1.5687632263699674e-06, + "loss": 0.5576, + "step": 10658 + }, + { + "epoch": 3.742626404494382, + "grad_norm": 0.9651371240615845, + "learning_rate": 1.5683341885833931e-06, + "loss": 0.5378, + "step": 10659 + }, + { + "epoch": 3.7429775280898876, + "grad_norm": 0.9661044478416443, + "learning_rate": 1.5679051826589931e-06, + "loss": 0.5689, + "step": 10660 + }, + { + "epoch": 3.743328651685393, + "grad_norm": 0.9243354201316833, + "learning_rate": 1.5674762086114385e-06, + "loss": 0.5433, + "step": 10661 + }, + { + "epoch": 3.743679775280899, + "grad_norm": 0.9344622492790222, + "learning_rate": 1.5670472664553985e-06, + "loss": 0.5628, + "step": 10662 + }, + { + "epoch": 3.7440308988764044, + "grad_norm": 0.9268540143966675, + "learning_rate": 1.5666183562055449e-06, + "loss": 0.5555, + "step": 10663 + }, + { + "epoch": 3.74438202247191, + "grad_norm": 0.9837895631790161, + "learning_rate": 1.5661894778765433e-06, + "loss": 0.6233, + "step": 10664 + }, + { + "epoch": 3.7447331460674156, + "grad_norm": 0.9186933040618896, + "learning_rate": 1.5657606314830637e-06, + "loss": 0.5609, + "step": 10665 + }, + { + "epoch": 3.745084269662921, + "grad_norm": 0.9684637188911438, + "learning_rate": 1.5653318170397703e-06, + "loss": 0.5796, + "step": 10666 + }, + { + "epoch": 3.7454353932584272, + "grad_norm": 1.023880124092102, + "learning_rate": 1.5649030345613298e-06, + "loss": 0.5659, + "step": 10667 + }, + { + "epoch": 3.745786516853933, + "grad_norm": 0.9225965142250061, + "learning_rate": 1.5644742840624044e-06, + "loss": 0.5727, + "step": 10668 + }, + { + "epoch": 3.7461376404494384, + "grad_norm": 0.9389476180076599, + "learning_rate": 1.5640455655576585e-06, + "loss": 0.569, + "step": 10669 + }, + { + "epoch": 3.746488764044944, + "grad_norm": 0.9335291385650635, + "learning_rate": 1.5636168790617534e-06, + "loss": 0.5492, + "step": 10670 + }, + { + "epoch": 3.7468398876404496, + "grad_norm": 0.9351773262023926, + "learning_rate": 1.5631882245893492e-06, + "loss": 0.5542, + "step": 10671 + }, + { + "epoch": 3.747191011235955, + "grad_norm": 0.9217349290847778, + "learning_rate": 1.5627596021551066e-06, + "loss": 0.5438, + "step": 10672 + }, + { + "epoch": 3.747542134831461, + "grad_norm": 0.9657183885574341, + "learning_rate": 1.5623310117736835e-06, + "loss": 0.5799, + "step": 10673 + }, + { + "epoch": 3.7478932584269664, + "grad_norm": 0.9370497465133667, + "learning_rate": 1.5619024534597376e-06, + "loss": 0.5325, + "step": 10674 + }, + { + "epoch": 3.748244382022472, + "grad_norm": 0.9162675142288208, + "learning_rate": 1.561473927227925e-06, + "loss": 0.5427, + "step": 10675 + }, + { + "epoch": 3.7485955056179776, + "grad_norm": 0.8944252133369446, + "learning_rate": 1.561045433092901e-06, + "loss": 0.5085, + "step": 10676 + }, + { + "epoch": 3.748946629213483, + "grad_norm": 0.9014416933059692, + "learning_rate": 1.5606169710693192e-06, + "loss": 0.5515, + "step": 10677 + }, + { + "epoch": 3.749297752808989, + "grad_norm": 0.9568307995796204, + "learning_rate": 1.5601885411718337e-06, + "loss": 0.5805, + "step": 10678 + }, + { + "epoch": 3.7496488764044944, + "grad_norm": 0.9294196367263794, + "learning_rate": 1.5597601434150955e-06, + "loss": 0.5569, + "step": 10679 + }, + { + "epoch": 3.75, + "grad_norm": 0.9252208471298218, + "learning_rate": 1.559331777813756e-06, + "loss": 0.5509, + "step": 10680 + }, + { + "epoch": 3.7503511235955056, + "grad_norm": 0.9790807962417603, + "learning_rate": 1.5589034443824646e-06, + "loss": 0.574, + "step": 10681 + }, + { + "epoch": 3.750702247191011, + "grad_norm": 0.9812945127487183, + "learning_rate": 1.5584751431358708e-06, + "loss": 0.5831, + "step": 10682 + }, + { + "epoch": 3.751053370786517, + "grad_norm": 0.9515967965126038, + "learning_rate": 1.5580468740886201e-06, + "loss": 0.5575, + "step": 10683 + }, + { + "epoch": 3.7514044943820224, + "grad_norm": 0.9402060508728027, + "learning_rate": 1.5576186372553617e-06, + "loss": 0.5961, + "step": 10684 + }, + { + "epoch": 3.751755617977528, + "grad_norm": 0.9270188808441162, + "learning_rate": 1.557190432650738e-06, + "loss": 0.5733, + "step": 10685 + }, + { + "epoch": 3.7521067415730336, + "grad_norm": 0.9564749002456665, + "learning_rate": 1.5567622602893962e-06, + "loss": 0.5663, + "step": 10686 + }, + { + "epoch": 3.752457865168539, + "grad_norm": 0.9946996569633484, + "learning_rate": 1.5563341201859774e-06, + "loss": 0.5847, + "step": 10687 + }, + { + "epoch": 3.752808988764045, + "grad_norm": 0.989551842212677, + "learning_rate": 1.5559060123551239e-06, + "loss": 0.5842, + "step": 10688 + }, + { + "epoch": 3.7531601123595504, + "grad_norm": 0.961204469203949, + "learning_rate": 1.5554779368114779e-06, + "loss": 0.5802, + "step": 10689 + }, + { + "epoch": 3.753511235955056, + "grad_norm": 0.9486815929412842, + "learning_rate": 1.5550498935696766e-06, + "loss": 0.5454, + "step": 10690 + }, + { + "epoch": 3.7538623595505616, + "grad_norm": 0.9312151074409485, + "learning_rate": 1.5546218826443618e-06, + "loss": 0.5424, + "step": 10691 + }, + { + "epoch": 3.754213483146067, + "grad_norm": 0.9303691983222961, + "learning_rate": 1.5541939040501685e-06, + "loss": 0.5514, + "step": 10692 + }, + { + "epoch": 3.7545646067415728, + "grad_norm": 0.9411336779594421, + "learning_rate": 1.5537659578017352e-06, + "loss": 0.5921, + "step": 10693 + }, + { + "epoch": 3.754915730337079, + "grad_norm": 0.9188650846481323, + "learning_rate": 1.5533380439136958e-06, + "loss": 0.5697, + "step": 10694 + }, + { + "epoch": 3.7552668539325844, + "grad_norm": 0.964292585849762, + "learning_rate": 1.552910162400686e-06, + "loss": 0.6009, + "step": 10695 + }, + { + "epoch": 3.75561797752809, + "grad_norm": 0.9295787215232849, + "learning_rate": 1.552482313277337e-06, + "loss": 0.5317, + "step": 10696 + }, + { + "epoch": 3.7559691011235956, + "grad_norm": 0.9303566813468933, + "learning_rate": 1.5520544965582835e-06, + "loss": 0.5708, + "step": 10697 + }, + { + "epoch": 3.756320224719101, + "grad_norm": 0.9286878108978271, + "learning_rate": 1.5516267122581536e-06, + "loss": 0.5536, + "step": 10698 + }, + { + "epoch": 3.756671348314607, + "grad_norm": 0.9566774964332581, + "learning_rate": 1.55119896039158e-06, + "loss": 0.5655, + "step": 10699 + }, + { + "epoch": 3.7570224719101124, + "grad_norm": 0.9507827758789062, + "learning_rate": 1.5507712409731895e-06, + "loss": 0.5579, + "step": 10700 + }, + { + "epoch": 3.757373595505618, + "grad_norm": 0.9594137072563171, + "learning_rate": 1.5503435540176098e-06, + "loss": 0.5479, + "step": 10701 + }, + { + "epoch": 3.7577247191011236, + "grad_norm": 0.8982268571853638, + "learning_rate": 1.5499158995394689e-06, + "loss": 0.5344, + "step": 10702 + }, + { + "epoch": 3.758075842696629, + "grad_norm": 0.9873787760734558, + "learning_rate": 1.54948827755339e-06, + "loss": 0.5775, + "step": 10703 + }, + { + "epoch": 3.758426966292135, + "grad_norm": 0.9249011278152466, + "learning_rate": 1.549060688074e-06, + "loss": 0.5438, + "step": 10704 + }, + { + "epoch": 3.7587780898876404, + "grad_norm": 0.9297093749046326, + "learning_rate": 1.5486331311159195e-06, + "loss": 0.5524, + "step": 10705 + }, + { + "epoch": 3.759129213483146, + "grad_norm": 0.9570794105529785, + "learning_rate": 1.5482056066937728e-06, + "loss": 0.5653, + "step": 10706 + }, + { + "epoch": 3.7594803370786516, + "grad_norm": 0.9445897936820984, + "learning_rate": 1.5477781148221787e-06, + "loss": 0.5367, + "step": 10707 + }, + { + "epoch": 3.759831460674157, + "grad_norm": 0.9461711645126343, + "learning_rate": 1.5473506555157597e-06, + "loss": 0.5423, + "step": 10708 + }, + { + "epoch": 3.760182584269663, + "grad_norm": 0.9353588223457336, + "learning_rate": 1.5469232287891316e-06, + "loss": 0.5434, + "step": 10709 + }, + { + "epoch": 3.7605337078651684, + "grad_norm": 0.9335821270942688, + "learning_rate": 1.546495834656915e-06, + "loss": 0.5598, + "step": 10710 + }, + { + "epoch": 3.7608848314606744, + "grad_norm": 0.9592576622962952, + "learning_rate": 1.5460684731337238e-06, + "loss": 0.5849, + "step": 10711 + }, + { + "epoch": 3.76123595505618, + "grad_norm": 0.9749981164932251, + "learning_rate": 1.5456411442341754e-06, + "loss": 0.5137, + "step": 10712 + }, + { + "epoch": 3.7615870786516856, + "grad_norm": 0.9782581329345703, + "learning_rate": 1.5452138479728832e-06, + "loss": 0.5363, + "step": 10713 + }, + { + "epoch": 3.761938202247191, + "grad_norm": 0.9624357223510742, + "learning_rate": 1.5447865843644595e-06, + "loss": 0.5637, + "step": 10714 + }, + { + "epoch": 3.762289325842697, + "grad_norm": 0.9408215880393982, + "learning_rate": 1.544359353423518e-06, + "loss": 0.5923, + "step": 10715 + }, + { + "epoch": 3.7626404494382024, + "grad_norm": 0.9330781698226929, + "learning_rate": 1.543932155164668e-06, + "loss": 0.6022, + "step": 10716 + }, + { + "epoch": 3.762991573033708, + "grad_norm": 0.9183773994445801, + "learning_rate": 1.5435049896025211e-06, + "loss": 0.5337, + "step": 10717 + }, + { + "epoch": 3.7633426966292136, + "grad_norm": 0.9441034197807312, + "learning_rate": 1.5430778567516842e-06, + "loss": 0.5558, + "step": 10718 + }, + { + "epoch": 3.763693820224719, + "grad_norm": 0.946442723274231, + "learning_rate": 1.5426507566267668e-06, + "loss": 0.5771, + "step": 10719 + }, + { + "epoch": 3.764044943820225, + "grad_norm": 0.941146194934845, + "learning_rate": 1.542223689242373e-06, + "loss": 0.5693, + "step": 10720 + }, + { + "epoch": 3.7643960674157304, + "grad_norm": 0.9326491951942444, + "learning_rate": 1.5417966546131108e-06, + "loss": 0.5303, + "step": 10721 + }, + { + "epoch": 3.764747191011236, + "grad_norm": 0.944232702255249, + "learning_rate": 1.5413696527535821e-06, + "loss": 0.5368, + "step": 10722 + }, + { + "epoch": 3.7650983146067416, + "grad_norm": 0.9303635954856873, + "learning_rate": 1.5409426836783917e-06, + "loss": 0.5512, + "step": 10723 + }, + { + "epoch": 3.765449438202247, + "grad_norm": 0.9647197127342224, + "learning_rate": 1.540515747402141e-06, + "loss": 0.6044, + "step": 10724 + }, + { + "epoch": 3.765800561797753, + "grad_norm": 0.989267885684967, + "learning_rate": 1.5400888439394296e-06, + "loss": 0.5506, + "step": 10725 + }, + { + "epoch": 3.7661516853932584, + "grad_norm": 0.9374409914016724, + "learning_rate": 1.5396619733048593e-06, + "loss": 0.5578, + "step": 10726 + }, + { + "epoch": 3.766502808988764, + "grad_norm": 0.9353379607200623, + "learning_rate": 1.5392351355130266e-06, + "loss": 0.5697, + "step": 10727 + }, + { + "epoch": 3.7668539325842696, + "grad_norm": 0.9379695653915405, + "learning_rate": 1.5388083305785315e-06, + "loss": 0.5398, + "step": 10728 + }, + { + "epoch": 3.767205056179775, + "grad_norm": 0.940359354019165, + "learning_rate": 1.5383815585159678e-06, + "loss": 0.5697, + "step": 10729 + }, + { + "epoch": 3.767556179775281, + "grad_norm": 0.9534140229225159, + "learning_rate": 1.537954819339933e-06, + "loss": 0.5597, + "step": 10730 + }, + { + "epoch": 3.7679073033707864, + "grad_norm": 0.9604806303977966, + "learning_rate": 1.5375281130650188e-06, + "loss": 0.5684, + "step": 10731 + }, + { + "epoch": 3.768258426966292, + "grad_norm": 0.9219235181808472, + "learning_rate": 1.537101439705821e-06, + "loss": 0.5387, + "step": 10732 + }, + { + "epoch": 3.7686095505617976, + "grad_norm": 0.9506957530975342, + "learning_rate": 1.5366747992769287e-06, + "loss": 0.5517, + "step": 10733 + }, + { + "epoch": 3.768960674157303, + "grad_norm": 0.9409708380699158, + "learning_rate": 1.536248191792935e-06, + "loss": 0.547, + "step": 10734 + }, + { + "epoch": 3.769311797752809, + "grad_norm": 0.911901593208313, + "learning_rate": 1.5358216172684276e-06, + "loss": 0.5284, + "step": 10735 + }, + { + "epoch": 3.7696629213483144, + "grad_norm": 0.9323476552963257, + "learning_rate": 1.535395075717997e-06, + "loss": 0.5683, + "step": 10736 + }, + { + "epoch": 3.77001404494382, + "grad_norm": 0.9363270998001099, + "learning_rate": 1.534968567156229e-06, + "loss": 0.5485, + "step": 10737 + }, + { + "epoch": 3.7703651685393256, + "grad_norm": 0.9381291270256042, + "learning_rate": 1.5345420915977094e-06, + "loss": 0.5604, + "step": 10738 + }, + { + "epoch": 3.7707162921348316, + "grad_norm": 0.9586318731307983, + "learning_rate": 1.5341156490570248e-06, + "loss": 0.554, + "step": 10739 + }, + { + "epoch": 3.771067415730337, + "grad_norm": 0.9650681018829346, + "learning_rate": 1.533689239548758e-06, + "loss": 0.588, + "step": 10740 + }, + { + "epoch": 3.771418539325843, + "grad_norm": 0.9583778977394104, + "learning_rate": 1.5332628630874929e-06, + "loss": 0.5394, + "step": 10741 + }, + { + "epoch": 3.7717696629213484, + "grad_norm": 0.9228506684303284, + "learning_rate": 1.5328365196878098e-06, + "loss": 0.5153, + "step": 10742 + }, + { + "epoch": 3.772120786516854, + "grad_norm": 0.9376241564750671, + "learning_rate": 1.532410209364291e-06, + "loss": 0.5549, + "step": 10743 + }, + { + "epoch": 3.7724719101123596, + "grad_norm": 0.9708750247955322, + "learning_rate": 1.5319839321315142e-06, + "loss": 0.5718, + "step": 10744 + }, + { + "epoch": 3.772823033707865, + "grad_norm": 0.9377588033676147, + "learning_rate": 1.5315576880040595e-06, + "loss": 0.5334, + "step": 10745 + }, + { + "epoch": 3.773174157303371, + "grad_norm": 0.9808067083358765, + "learning_rate": 1.5311314769965022e-06, + "loss": 0.5653, + "step": 10746 + }, + { + "epoch": 3.7735252808988764, + "grad_norm": 0.9662362337112427, + "learning_rate": 1.5307052991234204e-06, + "loss": 0.5288, + "step": 10747 + }, + { + "epoch": 3.773876404494382, + "grad_norm": 0.9530960917472839, + "learning_rate": 1.5302791543993872e-06, + "loss": 0.5706, + "step": 10748 + }, + { + "epoch": 3.7742275280898876, + "grad_norm": 0.9249573349952698, + "learning_rate": 1.5298530428389773e-06, + "loss": 0.5643, + "step": 10749 + }, + { + "epoch": 3.774578651685393, + "grad_norm": 0.9607812166213989, + "learning_rate": 1.5294269644567637e-06, + "loss": 0.5785, + "step": 10750 + }, + { + "epoch": 3.774929775280899, + "grad_norm": 0.9255026578903198, + "learning_rate": 1.5290009192673161e-06, + "loss": 0.5707, + "step": 10751 + }, + { + "epoch": 3.7752808988764044, + "grad_norm": 0.9447278380393982, + "learning_rate": 1.5285749072852074e-06, + "loss": 0.5351, + "step": 10752 + }, + { + "epoch": 3.77563202247191, + "grad_norm": 0.9822608232498169, + "learning_rate": 1.5281489285250044e-06, + "loss": 0.5527, + "step": 10753 + }, + { + "epoch": 3.7759831460674156, + "grad_norm": 0.9427886605262756, + "learning_rate": 1.527722983001278e-06, + "loss": 0.5567, + "step": 10754 + }, + { + "epoch": 3.776334269662921, + "grad_norm": 0.9581505656242371, + "learning_rate": 1.5272970707285927e-06, + "loss": 0.5867, + "step": 10755 + }, + { + "epoch": 3.7766853932584272, + "grad_norm": 0.956584095954895, + "learning_rate": 1.526871191721515e-06, + "loss": 0.5581, + "step": 10756 + }, + { + "epoch": 3.777036516853933, + "grad_norm": 0.9370353817939758, + "learning_rate": 1.5264453459946102e-06, + "loss": 0.5563, + "step": 10757 + }, + { + "epoch": 3.7773876404494384, + "grad_norm": 0.9345613718032837, + "learning_rate": 1.5260195335624412e-06, + "loss": 0.5302, + "step": 10758 + }, + { + "epoch": 3.777738764044944, + "grad_norm": 0.9434543251991272, + "learning_rate": 1.525593754439571e-06, + "loss": 0.5023, + "step": 10759 + }, + { + "epoch": 3.7780898876404496, + "grad_norm": 0.9286500811576843, + "learning_rate": 1.5251680086405604e-06, + "loss": 0.5286, + "step": 10760 + }, + { + "epoch": 3.778441011235955, + "grad_norm": 0.9689263105392456, + "learning_rate": 1.5247422961799696e-06, + "loss": 0.5621, + "step": 10761 + }, + { + "epoch": 3.778792134831461, + "grad_norm": 0.9467781186103821, + "learning_rate": 1.5243166170723584e-06, + "loss": 0.5531, + "step": 10762 + }, + { + "epoch": 3.7791432584269664, + "grad_norm": 0.9614080786705017, + "learning_rate": 1.5238909713322835e-06, + "loss": 0.5301, + "step": 10763 + }, + { + "epoch": 3.779494382022472, + "grad_norm": 0.9189320206642151, + "learning_rate": 1.523465358974302e-06, + "loss": 0.519, + "step": 10764 + }, + { + "epoch": 3.7798455056179776, + "grad_norm": 0.935393750667572, + "learning_rate": 1.5230397800129697e-06, + "loss": 0.5874, + "step": 10765 + }, + { + "epoch": 3.780196629213483, + "grad_norm": 0.9262731075286865, + "learning_rate": 1.522614234462841e-06, + "loss": 0.5785, + "step": 10766 + }, + { + "epoch": 3.780547752808989, + "grad_norm": 0.9411794543266296, + "learning_rate": 1.522188722338469e-06, + "loss": 0.5489, + "step": 10767 + }, + { + "epoch": 3.7808988764044944, + "grad_norm": 0.9350724220275879, + "learning_rate": 1.5217632436544062e-06, + "loss": 0.5283, + "step": 10768 + }, + { + "epoch": 3.78125, + "grad_norm": 0.9689305424690247, + "learning_rate": 1.5213377984252036e-06, + "loss": 0.5828, + "step": 10769 + }, + { + "epoch": 3.7816011235955056, + "grad_norm": 0.9296193718910217, + "learning_rate": 1.52091238666541e-06, + "loss": 0.5757, + "step": 10770 + }, + { + "epoch": 3.781952247191011, + "grad_norm": 0.9350335597991943, + "learning_rate": 1.5204870083895756e-06, + "loss": 0.5369, + "step": 10771 + }, + { + "epoch": 3.782303370786517, + "grad_norm": 0.944064736366272, + "learning_rate": 1.5200616636122467e-06, + "loss": 0.5502, + "step": 10772 + }, + { + "epoch": 3.7826544943820224, + "grad_norm": 0.9680630564689636, + "learning_rate": 1.5196363523479713e-06, + "loss": 0.5325, + "step": 10773 + }, + { + "epoch": 3.783005617977528, + "grad_norm": 0.9690945744514465, + "learning_rate": 1.5192110746112926e-06, + "loss": 0.5502, + "step": 10774 + }, + { + "epoch": 3.7833567415730336, + "grad_norm": 0.939819872379303, + "learning_rate": 1.5187858304167572e-06, + "loss": 0.5443, + "step": 10775 + }, + { + "epoch": 3.783707865168539, + "grad_norm": 0.9469059705734253, + "learning_rate": 1.5183606197789064e-06, + "loss": 0.5685, + "step": 10776 + }, + { + "epoch": 3.784058988764045, + "grad_norm": 0.9511504173278809, + "learning_rate": 1.5179354427122816e-06, + "loss": 0.5544, + "step": 10777 + }, + { + "epoch": 3.7844101123595504, + "grad_norm": 0.9215365052223206, + "learning_rate": 1.5175102992314253e-06, + "loss": 0.518, + "step": 10778 + }, + { + "epoch": 3.784761235955056, + "grad_norm": 0.9318764209747314, + "learning_rate": 1.5170851893508747e-06, + "loss": 0.5567, + "step": 10779 + }, + { + "epoch": 3.7851123595505616, + "grad_norm": 0.9270498752593994, + "learning_rate": 1.516660113085171e-06, + "loss": 0.5447, + "step": 10780 + }, + { + "epoch": 3.785463483146067, + "grad_norm": 0.942909300327301, + "learning_rate": 1.5162350704488488e-06, + "loss": 0.5601, + "step": 10781 + }, + { + "epoch": 3.7858146067415728, + "grad_norm": 0.9408800601959229, + "learning_rate": 1.5158100614564463e-06, + "loss": 0.5145, + "step": 10782 + }, + { + "epoch": 3.786165730337079, + "grad_norm": 0.9699331521987915, + "learning_rate": 1.5153850861224967e-06, + "loss": 0.5445, + "step": 10783 + }, + { + "epoch": 3.7865168539325844, + "grad_norm": 0.9015965461730957, + "learning_rate": 1.5149601444615353e-06, + "loss": 0.5341, + "step": 10784 + }, + { + "epoch": 3.78686797752809, + "grad_norm": 0.9311888813972473, + "learning_rate": 1.5145352364880934e-06, + "loss": 0.542, + "step": 10785 + }, + { + "epoch": 3.7872191011235956, + "grad_norm": 0.9446443319320679, + "learning_rate": 1.5141103622167042e-06, + "loss": 0.5652, + "step": 10786 + }, + { + "epoch": 3.787570224719101, + "grad_norm": 0.9273098111152649, + "learning_rate": 1.5136855216618964e-06, + "loss": 0.6077, + "step": 10787 + }, + { + "epoch": 3.787921348314607, + "grad_norm": 0.9330839514732361, + "learning_rate": 1.5132607148381992e-06, + "loss": 0.5263, + "step": 10788 + }, + { + "epoch": 3.7882724719101124, + "grad_norm": 0.9516134858131409, + "learning_rate": 1.5128359417601422e-06, + "loss": 0.5678, + "step": 10789 + }, + { + "epoch": 3.788623595505618, + "grad_norm": 0.9730702042579651, + "learning_rate": 1.51241120244225e-06, + "loss": 0.6107, + "step": 10790 + }, + { + "epoch": 3.7889747191011236, + "grad_norm": 0.9270119667053223, + "learning_rate": 1.5119864968990507e-06, + "loss": 0.5484, + "step": 10791 + }, + { + "epoch": 3.789325842696629, + "grad_norm": 0.9622483849525452, + "learning_rate": 1.511561825145067e-06, + "loss": 0.5462, + "step": 10792 + }, + { + "epoch": 3.789676966292135, + "grad_norm": 0.9596713185310364, + "learning_rate": 1.5111371871948239e-06, + "loss": 0.6066, + "step": 10793 + }, + { + "epoch": 3.7900280898876404, + "grad_norm": 0.9476714730262756, + "learning_rate": 1.5107125830628416e-06, + "loss": 0.534, + "step": 10794 + }, + { + "epoch": 3.790379213483146, + "grad_norm": 0.9400805830955505, + "learning_rate": 1.5102880127636437e-06, + "loss": 0.577, + "step": 10795 + }, + { + "epoch": 3.7907303370786516, + "grad_norm": 0.9370384812355042, + "learning_rate": 1.5098634763117475e-06, + "loss": 0.554, + "step": 10796 + }, + { + "epoch": 3.791081460674157, + "grad_norm": 0.9267989993095398, + "learning_rate": 1.5094389737216747e-06, + "loss": 0.5495, + "step": 10797 + }, + { + "epoch": 3.791432584269663, + "grad_norm": 0.9522055983543396, + "learning_rate": 1.5090145050079397e-06, + "loss": 0.5724, + "step": 10798 + }, + { + "epoch": 3.7917837078651684, + "grad_norm": 0.9279165863990784, + "learning_rate": 1.508590070185062e-06, + "loss": 0.5199, + "step": 10799 + }, + { + "epoch": 3.7921348314606744, + "grad_norm": 0.9527968764305115, + "learning_rate": 1.5081656692675556e-06, + "loss": 0.5487, + "step": 10800 + }, + { + "epoch": 3.79248595505618, + "grad_norm": 0.9761722683906555, + "learning_rate": 1.5077413022699334e-06, + "loss": 0.5977, + "step": 10801 + }, + { + "epoch": 3.7928370786516856, + "grad_norm": 0.9520300030708313, + "learning_rate": 1.5073169692067108e-06, + "loss": 0.5322, + "step": 10802 + }, + { + "epoch": 3.793188202247191, + "grad_norm": 0.9618119597434998, + "learning_rate": 1.5068926700923971e-06, + "loss": 0.5474, + "step": 10803 + }, + { + "epoch": 3.793539325842697, + "grad_norm": 0.9429842233657837, + "learning_rate": 1.5064684049415058e-06, + "loss": 0.5465, + "step": 10804 + }, + { + "epoch": 3.7938904494382024, + "grad_norm": 0.9232264757156372, + "learning_rate": 1.5060441737685434e-06, + "loss": 0.5756, + "step": 10805 + }, + { + "epoch": 3.794241573033708, + "grad_norm": 0.9250338077545166, + "learning_rate": 1.505619976588021e-06, + "loss": 0.5523, + "step": 10806 + }, + { + "epoch": 3.7945926966292136, + "grad_norm": 0.9654404520988464, + "learning_rate": 1.5051958134144428e-06, + "loss": 0.5678, + "step": 10807 + }, + { + "epoch": 3.794943820224719, + "grad_norm": 0.9384139776229858, + "learning_rate": 1.5047716842623184e-06, + "loss": 0.5391, + "step": 10808 + }, + { + "epoch": 3.795294943820225, + "grad_norm": 0.9213578701019287, + "learning_rate": 1.5043475891461493e-06, + "loss": 0.5409, + "step": 10809 + }, + { + "epoch": 3.7956460674157304, + "grad_norm": 0.9485815763473511, + "learning_rate": 1.5039235280804418e-06, + "loss": 0.5405, + "step": 10810 + }, + { + "epoch": 3.795997191011236, + "grad_norm": 0.941283106803894, + "learning_rate": 1.503499501079696e-06, + "loss": 0.5466, + "step": 10811 + }, + { + "epoch": 3.7963483146067416, + "grad_norm": 0.9410977959632874, + "learning_rate": 1.5030755081584159e-06, + "loss": 0.5529, + "step": 10812 + }, + { + "epoch": 3.796699438202247, + "grad_norm": 0.9363217353820801, + "learning_rate": 1.5026515493311002e-06, + "loss": 0.5548, + "step": 10813 + }, + { + "epoch": 3.797050561797753, + "grad_norm": 0.9521547555923462, + "learning_rate": 1.502227624612247e-06, + "loss": 0.562, + "step": 10814 + }, + { + "epoch": 3.7974016853932584, + "grad_norm": 1.0684159994125366, + "learning_rate": 1.5018037340163566e-06, + "loss": 0.5552, + "step": 10815 + }, + { + "epoch": 3.797752808988764, + "grad_norm": 0.9399673342704773, + "learning_rate": 1.5013798775579231e-06, + "loss": 0.5621, + "step": 10816 + }, + { + "epoch": 3.7981039325842696, + "grad_norm": 0.9376049041748047, + "learning_rate": 1.5009560552514441e-06, + "loss": 0.5503, + "step": 10817 + }, + { + "epoch": 3.798455056179775, + "grad_norm": 0.9296598434448242, + "learning_rate": 1.5005322671114123e-06, + "loss": 0.5199, + "step": 10818 + }, + { + "epoch": 3.798806179775281, + "grad_norm": 0.9487420320510864, + "learning_rate": 1.5001085131523227e-06, + "loss": 0.5586, + "step": 10819 + }, + { + "epoch": 3.7991573033707864, + "grad_norm": 0.9297412633895874, + "learning_rate": 1.4996847933886651e-06, + "loss": 0.5587, + "step": 10820 + }, + { + "epoch": 3.799508426966292, + "grad_norm": 0.9367296099662781, + "learning_rate": 1.499261107834933e-06, + "loss": 0.5521, + "step": 10821 + }, + { + "epoch": 3.7998595505617976, + "grad_norm": 0.9718416929244995, + "learning_rate": 1.4988374565056133e-06, + "loss": 0.5855, + "step": 10822 + }, + { + "epoch": 3.800210674157303, + "grad_norm": 0.9464272260665894, + "learning_rate": 1.498413839415197e-06, + "loss": 0.5761, + "step": 10823 + }, + { + "epoch": 3.800561797752809, + "grad_norm": 0.950835108757019, + "learning_rate": 1.4979902565781693e-06, + "loss": 0.569, + "step": 10824 + }, + { + "epoch": 3.8009129213483144, + "grad_norm": 0.9220231771469116, + "learning_rate": 1.4975667080090184e-06, + "loss": 0.5133, + "step": 10825 + }, + { + "epoch": 3.80126404494382, + "grad_norm": 0.9331125617027283, + "learning_rate": 1.4971431937222283e-06, + "loss": 0.559, + "step": 10826 + }, + { + "epoch": 3.8016151685393256, + "grad_norm": 0.913788378238678, + "learning_rate": 1.496719713732282e-06, + "loss": 0.5232, + "step": 10827 + }, + { + "epoch": 3.8019662921348316, + "grad_norm": 0.9765790700912476, + "learning_rate": 1.4962962680536635e-06, + "loss": 0.57, + "step": 10828 + }, + { + "epoch": 3.802317415730337, + "grad_norm": 0.9475187063217163, + "learning_rate": 1.4958728567008534e-06, + "loss": 0.5861, + "step": 10829 + }, + { + "epoch": 3.802668539325843, + "grad_norm": 0.8875717520713806, + "learning_rate": 1.4954494796883328e-06, + "loss": 0.5077, + "step": 10830 + }, + { + "epoch": 3.8030196629213484, + "grad_norm": 0.9128801822662354, + "learning_rate": 1.4950261370305797e-06, + "loss": 0.5679, + "step": 10831 + }, + { + "epoch": 3.803370786516854, + "grad_norm": 0.9409915804862976, + "learning_rate": 1.4946028287420739e-06, + "loss": 0.5814, + "step": 10832 + }, + { + "epoch": 3.8037219101123596, + "grad_norm": 0.9061814546585083, + "learning_rate": 1.4941795548372901e-06, + "loss": 0.5228, + "step": 10833 + }, + { + "epoch": 3.804073033707865, + "grad_norm": 0.9344466924667358, + "learning_rate": 1.493756315330705e-06, + "loss": 0.574, + "step": 10834 + }, + { + "epoch": 3.804424157303371, + "grad_norm": 0.9444854259490967, + "learning_rate": 1.4933331102367927e-06, + "loss": 0.5403, + "step": 10835 + }, + { + "epoch": 3.8047752808988764, + "grad_norm": 0.9301469326019287, + "learning_rate": 1.492909939570027e-06, + "loss": 0.5608, + "step": 10836 + }, + { + "epoch": 3.805126404494382, + "grad_norm": 0.9251372218132019, + "learning_rate": 1.4924868033448792e-06, + "loss": 0.5568, + "step": 10837 + }, + { + "epoch": 3.8054775280898876, + "grad_norm": 0.9703600406646729, + "learning_rate": 1.4920637015758208e-06, + "loss": 0.5683, + "step": 10838 + }, + { + "epoch": 3.805828651685393, + "grad_norm": 0.9172004461288452, + "learning_rate": 1.4916406342773218e-06, + "loss": 0.5609, + "step": 10839 + }, + { + "epoch": 3.806179775280899, + "grad_norm": 0.9733784198760986, + "learning_rate": 1.491217601463849e-06, + "loss": 0.5886, + "step": 10840 + }, + { + "epoch": 3.8065308988764044, + "grad_norm": 0.9580109119415283, + "learning_rate": 1.4907946031498723e-06, + "loss": 0.5478, + "step": 10841 + }, + { + "epoch": 3.80688202247191, + "grad_norm": 0.9043340682983398, + "learning_rate": 1.490371639349856e-06, + "loss": 0.5076, + "step": 10842 + }, + { + "epoch": 3.8072331460674156, + "grad_norm": 1.0355217456817627, + "learning_rate": 1.4899487100782656e-06, + "loss": 0.5202, + "step": 10843 + }, + { + "epoch": 3.807584269662921, + "grad_norm": 0.9685026407241821, + "learning_rate": 1.489525815349565e-06, + "loss": 0.576, + "step": 10844 + }, + { + "epoch": 3.8079353932584272, + "grad_norm": 0.8947959542274475, + "learning_rate": 1.489102955178217e-06, + "loss": 0.5071, + "step": 10845 + }, + { + "epoch": 3.808286516853933, + "grad_norm": 0.9445297122001648, + "learning_rate": 1.4886801295786828e-06, + "loss": 0.5564, + "step": 10846 + }, + { + "epoch": 3.8086376404494384, + "grad_norm": 0.8878141641616821, + "learning_rate": 1.488257338565423e-06, + "loss": 0.5265, + "step": 10847 + }, + { + "epoch": 3.808988764044944, + "grad_norm": 0.9618232250213623, + "learning_rate": 1.487834582152896e-06, + "loss": 0.534, + "step": 10848 + }, + { + "epoch": 3.8093398876404496, + "grad_norm": 0.9541359543800354, + "learning_rate": 1.487411860355561e-06, + "loss": 0.5452, + "step": 10849 + }, + { + "epoch": 3.809691011235955, + "grad_norm": 0.9364867806434631, + "learning_rate": 1.4869891731878733e-06, + "loss": 0.546, + "step": 10850 + }, + { + "epoch": 3.810042134831461, + "grad_norm": 0.9496193528175354, + "learning_rate": 1.486566520664289e-06, + "loss": 0.5845, + "step": 10851 + }, + { + "epoch": 3.8103932584269664, + "grad_norm": 0.9321529269218445, + "learning_rate": 1.4861439027992625e-06, + "loss": 0.5863, + "step": 10852 + }, + { + "epoch": 3.810744382022472, + "grad_norm": 0.928046703338623, + "learning_rate": 1.4857213196072468e-06, + "loss": 0.5568, + "step": 10853 + }, + { + "epoch": 3.8110955056179776, + "grad_norm": 0.9556453824043274, + "learning_rate": 1.4852987711026946e-06, + "loss": 0.5521, + "step": 10854 + }, + { + "epoch": 3.811446629213483, + "grad_norm": 0.9275403618812561, + "learning_rate": 1.4848762573000556e-06, + "loss": 0.5912, + "step": 10855 + }, + { + "epoch": 3.811797752808989, + "grad_norm": 0.9700289368629456, + "learning_rate": 1.4844537782137805e-06, + "loss": 0.5581, + "step": 10856 + }, + { + "epoch": 3.8121488764044944, + "grad_norm": 0.9431378245353699, + "learning_rate": 1.4840313338583162e-06, + "loss": 0.5706, + "step": 10857 + }, + { + "epoch": 3.8125, + "grad_norm": 0.9479694962501526, + "learning_rate": 1.4836089242481119e-06, + "loss": 0.5993, + "step": 10858 + }, + { + "epoch": 3.8128511235955056, + "grad_norm": 0.9443580508232117, + "learning_rate": 1.4831865493976117e-06, + "loss": 0.5622, + "step": 10859 + }, + { + "epoch": 3.813202247191011, + "grad_norm": 0.9167327880859375, + "learning_rate": 1.4827642093212623e-06, + "loss": 0.5655, + "step": 10860 + }, + { + "epoch": 3.813553370786517, + "grad_norm": 0.9365440607070923, + "learning_rate": 1.4823419040335058e-06, + "loss": 0.5736, + "step": 10861 + }, + { + "epoch": 3.8139044943820224, + "grad_norm": 0.9381092190742493, + "learning_rate": 1.4819196335487862e-06, + "loss": 0.5593, + "step": 10862 + }, + { + "epoch": 3.814255617977528, + "grad_norm": 0.9740251302719116, + "learning_rate": 1.481497397881544e-06, + "loss": 0.5976, + "step": 10863 + }, + { + "epoch": 3.8146067415730336, + "grad_norm": 0.9156931042671204, + "learning_rate": 1.4810751970462182e-06, + "loss": 0.5349, + "step": 10864 + }, + { + "epoch": 3.814957865168539, + "grad_norm": 0.9777852892875671, + "learning_rate": 1.4806530310572495e-06, + "loss": 0.5881, + "step": 10865 + }, + { + "epoch": 3.815308988764045, + "grad_norm": 0.9901489615440369, + "learning_rate": 1.4802308999290742e-06, + "loss": 0.5698, + "step": 10866 + }, + { + "epoch": 3.8156601123595504, + "grad_norm": 0.9273315072059631, + "learning_rate": 1.4798088036761304e-06, + "loss": 0.5979, + "step": 10867 + }, + { + "epoch": 3.816011235955056, + "grad_norm": 0.9300526976585388, + "learning_rate": 1.4793867423128516e-06, + "loss": 0.5676, + "step": 10868 + }, + { + "epoch": 3.8163623595505616, + "grad_norm": 0.9585113525390625, + "learning_rate": 1.4789647158536734e-06, + "loss": 0.5881, + "step": 10869 + }, + { + "epoch": 3.816713483146067, + "grad_norm": 0.9701324105262756, + "learning_rate": 1.4785427243130279e-06, + "loss": 0.5602, + "step": 10870 + }, + { + "epoch": 3.8170646067415728, + "grad_norm": 0.9350293278694153, + "learning_rate": 1.4781207677053478e-06, + "loss": 0.5507, + "step": 10871 + }, + { + "epoch": 3.817415730337079, + "grad_norm": 0.9317014217376709, + "learning_rate": 1.477698846045062e-06, + "loss": 0.5795, + "step": 10872 + }, + { + "epoch": 3.8177668539325844, + "grad_norm": 0.9305100440979004, + "learning_rate": 1.4772769593466018e-06, + "loss": 0.5305, + "step": 10873 + }, + { + "epoch": 3.81811797752809, + "grad_norm": 0.9462583661079407, + "learning_rate": 1.4768551076243936e-06, + "loss": 0.5673, + "step": 10874 + }, + { + "epoch": 3.8184691011235956, + "grad_norm": 1.0228981971740723, + "learning_rate": 1.4764332908928661e-06, + "loss": 0.5367, + "step": 10875 + }, + { + "epoch": 3.818820224719101, + "grad_norm": 0.9754185080528259, + "learning_rate": 1.4760115091664442e-06, + "loss": 0.5846, + "step": 10876 + }, + { + "epoch": 3.819171348314607, + "grad_norm": 0.9211755990982056, + "learning_rate": 1.4755897624595514e-06, + "loss": 0.5607, + "step": 10877 + }, + { + "epoch": 3.8195224719101124, + "grad_norm": 0.981249213218689, + "learning_rate": 1.4751680507866135e-06, + "loss": 0.587, + "step": 10878 + }, + { + "epoch": 3.819873595505618, + "grad_norm": 0.9556145071983337, + "learning_rate": 1.4747463741620498e-06, + "loss": 0.5726, + "step": 10879 + }, + { + "epoch": 3.8202247191011236, + "grad_norm": 0.9598215222358704, + "learning_rate": 1.4743247326002841e-06, + "loss": 0.549, + "step": 10880 + }, + { + "epoch": 3.820575842696629, + "grad_norm": 0.9482568502426147, + "learning_rate": 1.4739031261157336e-06, + "loss": 0.5719, + "step": 10881 + }, + { + "epoch": 3.820926966292135, + "grad_norm": 0.9245221614837646, + "learning_rate": 1.4734815547228193e-06, + "loss": 0.5734, + "step": 10882 + }, + { + "epoch": 3.8212780898876404, + "grad_norm": 0.9448093771934509, + "learning_rate": 1.4730600184359562e-06, + "loss": 0.5453, + "step": 10883 + }, + { + "epoch": 3.821629213483146, + "grad_norm": 0.9300689101219177, + "learning_rate": 1.4726385172695629e-06, + "loss": 0.5663, + "step": 10884 + }, + { + "epoch": 3.8219803370786516, + "grad_norm": 0.9453216195106506, + "learning_rate": 1.4722170512380524e-06, + "loss": 0.5485, + "step": 10885 + }, + { + "epoch": 3.822331460674157, + "grad_norm": 0.9246537685394287, + "learning_rate": 1.4717956203558398e-06, + "loss": 0.5502, + "step": 10886 + }, + { + "epoch": 3.822682584269663, + "grad_norm": 0.9323095679283142, + "learning_rate": 1.4713742246373363e-06, + "loss": 0.5753, + "step": 10887 + }, + { + "epoch": 3.8230337078651684, + "grad_norm": 0.9466277956962585, + "learning_rate": 1.4709528640969553e-06, + "loss": 0.5429, + "step": 10888 + }, + { + "epoch": 3.8233848314606744, + "grad_norm": 0.9182153940200806, + "learning_rate": 1.470531538749106e-06, + "loss": 0.5681, + "step": 10889 + }, + { + "epoch": 3.82373595505618, + "grad_norm": 0.952373206615448, + "learning_rate": 1.4701102486081955e-06, + "loss": 0.5383, + "step": 10890 + }, + { + "epoch": 3.8240870786516856, + "grad_norm": 0.9180224537849426, + "learning_rate": 1.4696889936886345e-06, + "loss": 0.5472, + "step": 10891 + }, + { + "epoch": 3.824438202247191, + "grad_norm": 0.9640561938285828, + "learning_rate": 1.4692677740048273e-06, + "loss": 0.5559, + "step": 10892 + }, + { + "epoch": 3.824789325842697, + "grad_norm": 0.9628115892410278, + "learning_rate": 1.4688465895711811e-06, + "loss": 0.5572, + "step": 10893 + }, + { + "epoch": 3.8251404494382024, + "grad_norm": 0.9241471290588379, + "learning_rate": 1.4684254404020983e-06, + "loss": 0.546, + "step": 10894 + }, + { + "epoch": 3.825491573033708, + "grad_norm": 0.9464368224143982, + "learning_rate": 1.4680043265119838e-06, + "loss": 0.5515, + "step": 10895 + }, + { + "epoch": 3.8258426966292136, + "grad_norm": 0.9185828566551208, + "learning_rate": 1.4675832479152371e-06, + "loss": 0.5501, + "step": 10896 + }, + { + "epoch": 3.826193820224719, + "grad_norm": 0.9665042161941528, + "learning_rate": 1.467162204626261e-06, + "loss": 0.5866, + "step": 10897 + }, + { + "epoch": 3.826544943820225, + "grad_norm": 0.9414646625518799, + "learning_rate": 1.4667411966594524e-06, + "loss": 0.558, + "step": 10898 + }, + { + "epoch": 3.8268960674157304, + "grad_norm": 0.9399771094322205, + "learning_rate": 1.4663202240292118e-06, + "loss": 0.5514, + "step": 10899 + }, + { + "epoch": 3.827247191011236, + "grad_norm": 0.9222230911254883, + "learning_rate": 1.4658992867499344e-06, + "loss": 0.5164, + "step": 10900 + }, + { + "epoch": 3.8275983146067416, + "grad_norm": 0.9460821747779846, + "learning_rate": 1.4654783848360172e-06, + "loss": 0.5565, + "step": 10901 + }, + { + "epoch": 3.827949438202247, + "grad_norm": 0.9422255158424377, + "learning_rate": 1.465057518301854e-06, + "loss": 0.5463, + "step": 10902 + }, + { + "epoch": 3.828300561797753, + "grad_norm": 0.9388396739959717, + "learning_rate": 1.4646366871618368e-06, + "loss": 0.5627, + "step": 10903 + }, + { + "epoch": 3.8286516853932584, + "grad_norm": 0.9079676866531372, + "learning_rate": 1.4642158914303603e-06, + "loss": 0.5207, + "step": 10904 + }, + { + "epoch": 3.829002808988764, + "grad_norm": 0.9662813544273376, + "learning_rate": 1.463795131121813e-06, + "loss": 0.5695, + "step": 10905 + }, + { + "epoch": 3.8293539325842696, + "grad_norm": 0.9788883328437805, + "learning_rate": 1.4633744062505862e-06, + "loss": 0.5743, + "step": 10906 + }, + { + "epoch": 3.829705056179775, + "grad_norm": 0.9505196213722229, + "learning_rate": 1.4629537168310668e-06, + "loss": 0.5373, + "step": 10907 + }, + { + "epoch": 3.830056179775281, + "grad_norm": 0.942787230014801, + "learning_rate": 1.4625330628776441e-06, + "loss": 0.5354, + "step": 10908 + }, + { + "epoch": 3.8304073033707864, + "grad_norm": 0.9325472712516785, + "learning_rate": 1.4621124444047013e-06, + "loss": 0.538, + "step": 10909 + }, + { + "epoch": 3.830758426966292, + "grad_norm": 0.9609138369560242, + "learning_rate": 1.4616918614266263e-06, + "loss": 0.5584, + "step": 10910 + }, + { + "epoch": 3.8311095505617976, + "grad_norm": 0.9422142505645752, + "learning_rate": 1.4612713139578e-06, + "loss": 0.5493, + "step": 10911 + }, + { + "epoch": 3.831460674157303, + "grad_norm": 0.9521642327308655, + "learning_rate": 1.4608508020126067e-06, + "loss": 0.5369, + "step": 10912 + }, + { + "epoch": 3.831811797752809, + "grad_norm": 0.9577821493148804, + "learning_rate": 1.460430325605427e-06, + "loss": 0.5893, + "step": 10913 + }, + { + "epoch": 3.8321629213483144, + "grad_norm": 0.9307746887207031, + "learning_rate": 1.4600098847506394e-06, + "loss": 0.5629, + "step": 10914 + }, + { + "epoch": 3.83251404494382, + "grad_norm": 0.8962500095367432, + "learning_rate": 1.459589479462626e-06, + "loss": 0.5415, + "step": 10915 + }, + { + "epoch": 3.8328651685393256, + "grad_norm": 0.9056134819984436, + "learning_rate": 1.4591691097557599e-06, + "loss": 0.553, + "step": 10916 + }, + { + "epoch": 3.8332162921348316, + "grad_norm": 0.9169676303863525, + "learning_rate": 1.4587487756444207e-06, + "loss": 0.5544, + "step": 10917 + }, + { + "epoch": 3.833567415730337, + "grad_norm": 0.9228463768959045, + "learning_rate": 1.4583284771429812e-06, + "loss": 0.5313, + "step": 10918 + }, + { + "epoch": 3.833918539325843, + "grad_norm": 0.9828086495399475, + "learning_rate": 1.4579082142658178e-06, + "loss": 0.5601, + "step": 10919 + }, + { + "epoch": 3.8342696629213484, + "grad_norm": 0.9204429984092712, + "learning_rate": 1.4574879870273007e-06, + "loss": 0.5512, + "step": 10920 + }, + { + "epoch": 3.834620786516854, + "grad_norm": 0.9896245002746582, + "learning_rate": 1.457067795441803e-06, + "loss": 0.5497, + "step": 10921 + }, + { + "epoch": 3.8349719101123596, + "grad_norm": 0.9174785614013672, + "learning_rate": 1.4566476395236934e-06, + "loss": 0.5249, + "step": 10922 + }, + { + "epoch": 3.835323033707865, + "grad_norm": 0.9288141131401062, + "learning_rate": 1.456227519287343e-06, + "loss": 0.5311, + "step": 10923 + }, + { + "epoch": 3.835674157303371, + "grad_norm": 1.0028986930847168, + "learning_rate": 1.455807434747117e-06, + "loss": 0.57, + "step": 10924 + }, + { + "epoch": 3.8360252808988764, + "grad_norm": 0.9294567704200745, + "learning_rate": 1.4553873859173843e-06, + "loss": 0.5967, + "step": 10925 + }, + { + "epoch": 3.836376404494382, + "grad_norm": 0.9574340581893921, + "learning_rate": 1.454967372812509e-06, + "loss": 0.5472, + "step": 10926 + }, + { + "epoch": 3.8367275280898876, + "grad_norm": 0.9573356509208679, + "learning_rate": 1.4545473954468542e-06, + "loss": 0.5682, + "step": 10927 + }, + { + "epoch": 3.837078651685393, + "grad_norm": 0.9482593536376953, + "learning_rate": 1.4541274538347852e-06, + "loss": 0.5548, + "step": 10928 + }, + { + "epoch": 3.837429775280899, + "grad_norm": 0.9522284269332886, + "learning_rate": 1.4537075479906613e-06, + "loss": 0.5479, + "step": 10929 + }, + { + "epoch": 3.8377808988764044, + "grad_norm": 0.963201642036438, + "learning_rate": 1.4532876779288447e-06, + "loss": 0.552, + "step": 10930 + }, + { + "epoch": 3.83813202247191, + "grad_norm": 0.9291874170303345, + "learning_rate": 1.452867843663693e-06, + "loss": 0.5206, + "step": 10931 + }, + { + "epoch": 3.8384831460674156, + "grad_norm": 0.9235369563102722, + "learning_rate": 1.4524480452095657e-06, + "loss": 0.5597, + "step": 10932 + }, + { + "epoch": 3.838834269662921, + "grad_norm": 0.9267776608467102, + "learning_rate": 1.4520282825808182e-06, + "loss": 0.5546, + "step": 10933 + }, + { + "epoch": 3.8391853932584272, + "grad_norm": 0.9524484276771545, + "learning_rate": 1.4516085557918075e-06, + "loss": 0.5806, + "step": 10934 + }, + { + "epoch": 3.839536516853933, + "grad_norm": 0.9253097176551819, + "learning_rate": 1.451188864856886e-06, + "loss": 0.5164, + "step": 10935 + }, + { + "epoch": 3.8398876404494384, + "grad_norm": 0.9512264728546143, + "learning_rate": 1.4507692097904092e-06, + "loss": 0.5477, + "step": 10936 + }, + { + "epoch": 3.840238764044944, + "grad_norm": 0.9273902177810669, + "learning_rate": 1.4503495906067261e-06, + "loss": 0.5348, + "step": 10937 + }, + { + "epoch": 3.8405898876404496, + "grad_norm": 0.9562587738037109, + "learning_rate": 1.4499300073201903e-06, + "loss": 0.5754, + "step": 10938 + }, + { + "epoch": 3.840941011235955, + "grad_norm": 0.9613707065582275, + "learning_rate": 1.4495104599451492e-06, + "loss": 0.5514, + "step": 10939 + }, + { + "epoch": 3.841292134831461, + "grad_norm": 0.9185537099838257, + "learning_rate": 1.449090948495951e-06, + "loss": 0.5512, + "step": 10940 + }, + { + "epoch": 3.8416432584269664, + "grad_norm": 0.9627362489700317, + "learning_rate": 1.4486714729869436e-06, + "loss": 0.6037, + "step": 10941 + }, + { + "epoch": 3.841994382022472, + "grad_norm": 0.9748256206512451, + "learning_rate": 1.4482520334324712e-06, + "loss": 0.5643, + "step": 10942 + }, + { + "epoch": 3.8423455056179776, + "grad_norm": 0.9233769178390503, + "learning_rate": 1.4478326298468807e-06, + "loss": 0.5295, + "step": 10943 + }, + { + "epoch": 3.842696629213483, + "grad_norm": 0.9540885090827942, + "learning_rate": 1.4474132622445125e-06, + "loss": 0.5745, + "step": 10944 + }, + { + "epoch": 3.843047752808989, + "grad_norm": 0.9345555305480957, + "learning_rate": 1.4469939306397113e-06, + "loss": 0.5531, + "step": 10945 + }, + { + "epoch": 3.8433988764044944, + "grad_norm": 0.9224537014961243, + "learning_rate": 1.4465746350468153e-06, + "loss": 0.5217, + "step": 10946 + }, + { + "epoch": 3.84375, + "grad_norm": 0.9641884565353394, + "learning_rate": 1.4461553754801667e-06, + "loss": 0.5871, + "step": 10947 + }, + { + "epoch": 3.8441011235955056, + "grad_norm": 0.9210605621337891, + "learning_rate": 1.4457361519541012e-06, + "loss": 0.5593, + "step": 10948 + }, + { + "epoch": 3.844452247191011, + "grad_norm": 0.9269044995307922, + "learning_rate": 1.4453169644829584e-06, + "loss": 0.5575, + "step": 10949 + }, + { + "epoch": 3.844803370786517, + "grad_norm": 0.9450878500938416, + "learning_rate": 1.4448978130810716e-06, + "loss": 0.5608, + "step": 10950 + }, + { + "epoch": 3.8451544943820224, + "grad_norm": 0.9434930682182312, + "learning_rate": 1.4444786977627777e-06, + "loss": 0.5514, + "step": 10951 + }, + { + "epoch": 3.845505617977528, + "grad_norm": 0.9217976331710815, + "learning_rate": 1.4440596185424093e-06, + "loss": 0.5686, + "step": 10952 + }, + { + "epoch": 3.8458567415730336, + "grad_norm": 0.9435510039329529, + "learning_rate": 1.4436405754342975e-06, + "loss": 0.5704, + "step": 10953 + }, + { + "epoch": 3.846207865168539, + "grad_norm": 0.9393649101257324, + "learning_rate": 1.443221568452775e-06, + "loss": 0.5667, + "step": 10954 + }, + { + "epoch": 3.846558988764045, + "grad_norm": 0.9749777317047119, + "learning_rate": 1.4428025976121695e-06, + "loss": 0.5457, + "step": 10955 + }, + { + "epoch": 3.8469101123595504, + "grad_norm": 0.9680777788162231, + "learning_rate": 1.4423836629268116e-06, + "loss": 0.5589, + "step": 10956 + }, + { + "epoch": 3.847261235955056, + "grad_norm": 1.0064245462417603, + "learning_rate": 1.4419647644110263e-06, + "loss": 0.5735, + "step": 10957 + }, + { + "epoch": 3.8476123595505616, + "grad_norm": 0.9689911603927612, + "learning_rate": 1.4415459020791416e-06, + "loss": 0.5338, + "step": 10958 + }, + { + "epoch": 3.847963483146067, + "grad_norm": 0.9638147950172424, + "learning_rate": 1.4411270759454803e-06, + "loss": 0.5902, + "step": 10959 + }, + { + "epoch": 3.8483146067415728, + "grad_norm": 0.9309440851211548, + "learning_rate": 1.440708286024368e-06, + "loss": 0.5284, + "step": 10960 + }, + { + "epoch": 3.848665730337079, + "grad_norm": 0.9388378262519836, + "learning_rate": 1.4402895323301249e-06, + "loss": 0.5665, + "step": 10961 + }, + { + "epoch": 3.8490168539325844, + "grad_norm": 0.9369130730628967, + "learning_rate": 1.439870814877074e-06, + "loss": 0.5783, + "step": 10962 + }, + { + "epoch": 3.84936797752809, + "grad_norm": 0.966088593006134, + "learning_rate": 1.439452133679533e-06, + "loss": 0.5588, + "step": 10963 + }, + { + "epoch": 3.8497191011235956, + "grad_norm": 0.9556235671043396, + "learning_rate": 1.4390334887518227e-06, + "loss": 0.5554, + "step": 10964 + }, + { + "epoch": 3.850070224719101, + "grad_norm": 0.9495468139648438, + "learning_rate": 1.438614880108259e-06, + "loss": 0.5802, + "step": 10965 + }, + { + "epoch": 3.850421348314607, + "grad_norm": 0.9883668422698975, + "learning_rate": 1.4381963077631573e-06, + "loss": 0.5786, + "step": 10966 + }, + { + "epoch": 3.8507724719101124, + "grad_norm": 0.9593429565429688, + "learning_rate": 1.4377777717308342e-06, + "loss": 0.5448, + "step": 10967 + }, + { + "epoch": 3.851123595505618, + "grad_norm": 0.9682921171188354, + "learning_rate": 1.4373592720256014e-06, + "loss": 0.5933, + "step": 10968 + }, + { + "epoch": 3.8514747191011236, + "grad_norm": 0.9346643090248108, + "learning_rate": 1.436940808661773e-06, + "loss": 0.5825, + "step": 10969 + }, + { + "epoch": 3.851825842696629, + "grad_norm": 0.91338050365448, + "learning_rate": 1.4365223816536584e-06, + "loss": 0.5559, + "step": 10970 + }, + { + "epoch": 3.852176966292135, + "grad_norm": 0.9082552790641785, + "learning_rate": 1.4361039910155694e-06, + "loss": 0.5254, + "step": 10971 + }, + { + "epoch": 3.8525280898876404, + "grad_norm": 0.9094631671905518, + "learning_rate": 1.4356856367618124e-06, + "loss": 0.5757, + "step": 10972 + }, + { + "epoch": 3.852879213483146, + "grad_norm": 0.9365786910057068, + "learning_rate": 1.4352673189066968e-06, + "loss": 0.5312, + "step": 10973 + }, + { + "epoch": 3.8532303370786516, + "grad_norm": 0.9188830852508545, + "learning_rate": 1.434849037464527e-06, + "loss": 0.555, + "step": 10974 + }, + { + "epoch": 3.853581460674157, + "grad_norm": 0.996910035610199, + "learning_rate": 1.4344307924496096e-06, + "loss": 0.5784, + "step": 10975 + }, + { + "epoch": 3.853932584269663, + "grad_norm": 0.9554314017295837, + "learning_rate": 1.4340125838762473e-06, + "loss": 0.6044, + "step": 10976 + }, + { + "epoch": 3.8542837078651684, + "grad_norm": 0.9080484509468079, + "learning_rate": 1.4335944117587413e-06, + "loss": 0.5519, + "step": 10977 + }, + { + "epoch": 3.8546348314606744, + "grad_norm": 1.3059234619140625, + "learning_rate": 1.4331762761113948e-06, + "loss": 0.5424, + "step": 10978 + }, + { + "epoch": 3.85498595505618, + "grad_norm": 0.9410378932952881, + "learning_rate": 1.432758176948506e-06, + "loss": 0.5787, + "step": 10979 + }, + { + "epoch": 3.8553370786516856, + "grad_norm": 1.0553710460662842, + "learning_rate": 1.432340114284375e-06, + "loss": 0.5766, + "step": 10980 + }, + { + "epoch": 3.855688202247191, + "grad_norm": 0.9499240517616272, + "learning_rate": 1.4319220881332979e-06, + "loss": 0.5733, + "step": 10981 + }, + { + "epoch": 3.856039325842697, + "grad_norm": 0.9400267601013184, + "learning_rate": 1.4315040985095725e-06, + "loss": 0.5807, + "step": 10982 + }, + { + "epoch": 3.8563904494382024, + "grad_norm": 0.9472957253456116, + "learning_rate": 1.431086145427491e-06, + "loss": 0.5753, + "step": 10983 + }, + { + "epoch": 3.856741573033708, + "grad_norm": 0.947072446346283, + "learning_rate": 1.43066822890135e-06, + "loss": 0.582, + "step": 10984 + }, + { + "epoch": 3.8570926966292136, + "grad_norm": 0.9560330510139465, + "learning_rate": 1.4302503489454395e-06, + "loss": 0.5798, + "step": 10985 + }, + { + "epoch": 3.857443820224719, + "grad_norm": 0.9631170630455017, + "learning_rate": 1.4298325055740528e-06, + "loss": 0.5625, + "step": 10986 + }, + { + "epoch": 3.857794943820225, + "grad_norm": 0.9455559849739075, + "learning_rate": 1.4294146988014774e-06, + "loss": 0.6031, + "step": 10987 + }, + { + "epoch": 3.8581460674157304, + "grad_norm": 0.9362829923629761, + "learning_rate": 1.4289969286420045e-06, + "loss": 0.558, + "step": 10988 + }, + { + "epoch": 3.858497191011236, + "grad_norm": 0.9333298802375793, + "learning_rate": 1.4285791951099199e-06, + "loss": 0.5521, + "step": 10989 + }, + { + "epoch": 3.8588483146067416, + "grad_norm": 0.9125950932502747, + "learning_rate": 1.428161498219509e-06, + "loss": 0.5625, + "step": 10990 + }, + { + "epoch": 3.859199438202247, + "grad_norm": 0.914951741695404, + "learning_rate": 1.4277438379850589e-06, + "loss": 0.5101, + "step": 10991 + }, + { + "epoch": 3.859550561797753, + "grad_norm": 0.9505528211593628, + "learning_rate": 1.4273262144208507e-06, + "loss": 0.5328, + "step": 10992 + }, + { + "epoch": 3.8599016853932584, + "grad_norm": 0.9727820158004761, + "learning_rate": 1.4269086275411693e-06, + "loss": 0.5356, + "step": 10993 + }, + { + "epoch": 3.860252808988764, + "grad_norm": 0.9923450350761414, + "learning_rate": 1.4264910773602937e-06, + "loss": 0.5386, + "step": 10994 + }, + { + "epoch": 3.8606039325842696, + "grad_norm": 0.9496344923973083, + "learning_rate": 1.4260735638925056e-06, + "loss": 0.5443, + "step": 10995 + }, + { + "epoch": 3.860955056179775, + "grad_norm": 0.9616802930831909, + "learning_rate": 1.4256560871520814e-06, + "loss": 0.5573, + "step": 10996 + }, + { + "epoch": 3.861306179775281, + "grad_norm": 0.9689275622367859, + "learning_rate": 1.4252386471533005e-06, + "loss": 0.5795, + "step": 10997 + }, + { + "epoch": 3.8616573033707864, + "grad_norm": 0.9484035968780518, + "learning_rate": 1.4248212439104376e-06, + "loss": 0.5111, + "step": 10998 + }, + { + "epoch": 3.862008426966292, + "grad_norm": 0.9421113133430481, + "learning_rate": 1.424403877437769e-06, + "loss": 0.5583, + "step": 10999 + }, + { + "epoch": 3.8623595505617976, + "grad_norm": 0.9728732109069824, + "learning_rate": 1.4239865477495662e-06, + "loss": 0.5488, + "step": 11000 + }, + { + "epoch": 3.862710674157303, + "grad_norm": 0.9230992794036865, + "learning_rate": 1.4235692548601038e-06, + "loss": 0.5273, + "step": 11001 + }, + { + "epoch": 3.863061797752809, + "grad_norm": 0.9112728238105774, + "learning_rate": 1.423151998783652e-06, + "loss": 0.5668, + "step": 11002 + }, + { + "epoch": 3.8634129213483144, + "grad_norm": 0.9217725992202759, + "learning_rate": 1.4227347795344793e-06, + "loss": 0.5311, + "step": 11003 + }, + { + "epoch": 3.86376404494382, + "grad_norm": 0.9478429555892944, + "learning_rate": 1.422317597126856e-06, + "loss": 0.5469, + "step": 11004 + }, + { + "epoch": 3.8641151685393256, + "grad_norm": 0.9611356854438782, + "learning_rate": 1.4219004515750479e-06, + "loss": 0.5661, + "step": 11005 + }, + { + "epoch": 3.8644662921348316, + "grad_norm": 0.9451982975006104, + "learning_rate": 1.4214833428933229e-06, + "loss": 0.5538, + "step": 11006 + }, + { + "epoch": 3.864817415730337, + "grad_norm": 0.933803915977478, + "learning_rate": 1.4210662710959438e-06, + "loss": 0.5702, + "step": 11007 + }, + { + "epoch": 3.865168539325843, + "grad_norm": 0.955635666847229, + "learning_rate": 1.4206492361971758e-06, + "loss": 0.5351, + "step": 11008 + }, + { + "epoch": 3.8655196629213484, + "grad_norm": 0.9997610449790955, + "learning_rate": 1.4202322382112793e-06, + "loss": 0.5395, + "step": 11009 + }, + { + "epoch": 3.865870786516854, + "grad_norm": 0.9518892168998718, + "learning_rate": 1.419815277152517e-06, + "loss": 0.5495, + "step": 11010 + }, + { + "epoch": 3.8662219101123596, + "grad_norm": 0.903296709060669, + "learning_rate": 1.4193983530351474e-06, + "loss": 0.5441, + "step": 11011 + }, + { + "epoch": 3.866573033707865, + "grad_norm": 0.939941942691803, + "learning_rate": 1.4189814658734302e-06, + "loss": 0.5478, + "step": 11012 + }, + { + "epoch": 3.866924157303371, + "grad_norm": 0.9873878955841064, + "learning_rate": 1.4185646156816207e-06, + "loss": 0.5445, + "step": 11013 + }, + { + "epoch": 3.8672752808988764, + "grad_norm": 0.9623928666114807, + "learning_rate": 1.418147802473977e-06, + "loss": 0.5855, + "step": 11014 + }, + { + "epoch": 3.867626404494382, + "grad_norm": 0.91180819272995, + "learning_rate": 1.4177310262647527e-06, + "loss": 0.4908, + "step": 11015 + }, + { + "epoch": 3.8679775280898876, + "grad_norm": 0.9501622319221497, + "learning_rate": 1.4173142870682006e-06, + "loss": 0.5449, + "step": 11016 + }, + { + "epoch": 3.868328651685393, + "grad_norm": 0.9056792855262756, + "learning_rate": 1.416897584898574e-06, + "loss": 0.5407, + "step": 11017 + }, + { + "epoch": 3.868679775280899, + "grad_norm": 0.9621976613998413, + "learning_rate": 1.4164809197701222e-06, + "loss": 0.5891, + "step": 11018 + }, + { + "epoch": 3.8690308988764044, + "grad_norm": 0.9467156529426575, + "learning_rate": 1.4160642916970967e-06, + "loss": 0.5577, + "step": 11019 + }, + { + "epoch": 3.86938202247191, + "grad_norm": 0.9179587960243225, + "learning_rate": 1.4156477006937442e-06, + "loss": 0.5534, + "step": 11020 + }, + { + "epoch": 3.8697331460674156, + "grad_norm": 0.9302246570587158, + "learning_rate": 1.415231146774313e-06, + "loss": 0.5528, + "step": 11021 + }, + { + "epoch": 3.870084269662921, + "grad_norm": 0.9292799830436707, + "learning_rate": 1.4148146299530476e-06, + "loss": 0.5464, + "step": 11022 + }, + { + "epoch": 3.8704353932584272, + "grad_norm": 0.9325312376022339, + "learning_rate": 1.4143981502441945e-06, + "loss": 0.539, + "step": 11023 + }, + { + "epoch": 3.870786516853933, + "grad_norm": 0.9469289779663086, + "learning_rate": 1.4139817076619943e-06, + "loss": 0.5645, + "step": 11024 + }, + { + "epoch": 3.8711376404494384, + "grad_norm": 0.959275484085083, + "learning_rate": 1.4135653022206913e-06, + "loss": 0.5237, + "step": 11025 + }, + { + "epoch": 3.871488764044944, + "grad_norm": 0.9085878133773804, + "learning_rate": 1.4131489339345245e-06, + "loss": 0.5552, + "step": 11026 + }, + { + "epoch": 3.8718398876404496, + "grad_norm": 0.903184711933136, + "learning_rate": 1.412732602817735e-06, + "loss": 0.5415, + "step": 11027 + }, + { + "epoch": 3.872191011235955, + "grad_norm": 0.9569523930549622, + "learning_rate": 1.4123163088845604e-06, + "loss": 0.5583, + "step": 11028 + }, + { + "epoch": 3.872542134831461, + "grad_norm": 0.9667792916297913, + "learning_rate": 1.411900052149236e-06, + "loss": 0.5752, + "step": 11029 + }, + { + "epoch": 3.8728932584269664, + "grad_norm": 0.9546414613723755, + "learning_rate": 1.4114838326259999e-06, + "loss": 0.5795, + "step": 11030 + }, + { + "epoch": 3.873244382022472, + "grad_norm": 0.959254264831543, + "learning_rate": 1.4110676503290843e-06, + "loss": 0.5707, + "step": 11031 + }, + { + "epoch": 3.8735955056179776, + "grad_norm": 0.961174488067627, + "learning_rate": 1.4106515052727241e-06, + "loss": 0.5686, + "step": 11032 + }, + { + "epoch": 3.873946629213483, + "grad_norm": 0.9194923639297485, + "learning_rate": 1.4102353974711492e-06, + "loss": 0.5483, + "step": 11033 + }, + { + "epoch": 3.874297752808989, + "grad_norm": 0.9240887761116028, + "learning_rate": 1.4098193269385924e-06, + "loss": 0.5586, + "step": 11034 + }, + { + "epoch": 3.8746488764044944, + "grad_norm": 0.9395901560783386, + "learning_rate": 1.4094032936892805e-06, + "loss": 0.5725, + "step": 11035 + }, + { + "epoch": 3.875, + "grad_norm": 0.9781716465950012, + "learning_rate": 1.4089872977374438e-06, + "loss": 0.5574, + "step": 11036 + }, + { + "epoch": 3.8753511235955056, + "grad_norm": 0.9723407030105591, + "learning_rate": 1.4085713390973072e-06, + "loss": 0.5471, + "step": 11037 + }, + { + "epoch": 3.875702247191011, + "grad_norm": 0.9118796586990356, + "learning_rate": 1.4081554177830975e-06, + "loss": 0.5358, + "step": 11038 + }, + { + "epoch": 3.876053370786517, + "grad_norm": 0.9137970209121704, + "learning_rate": 1.4077395338090383e-06, + "loss": 0.5854, + "step": 11039 + }, + { + "epoch": 3.8764044943820224, + "grad_norm": 0.9937852621078491, + "learning_rate": 1.4073236871893514e-06, + "loss": 0.5428, + "step": 11040 + }, + { + "epoch": 3.876755617977528, + "grad_norm": 0.9185106158256531, + "learning_rate": 1.4069078779382602e-06, + "loss": 0.5331, + "step": 11041 + }, + { + "epoch": 3.8771067415730336, + "grad_norm": 0.9094840288162231, + "learning_rate": 1.4064921060699835e-06, + "loss": 0.5334, + "step": 11042 + }, + { + "epoch": 3.877457865168539, + "grad_norm": 0.9433227181434631, + "learning_rate": 1.4060763715987419e-06, + "loss": 0.5218, + "step": 11043 + }, + { + "epoch": 3.877808988764045, + "grad_norm": 0.9375569820404053, + "learning_rate": 1.4056606745387513e-06, + "loss": 0.5592, + "step": 11044 + }, + { + "epoch": 3.8781601123595504, + "grad_norm": 0.9309371709823608, + "learning_rate": 1.4052450149042303e-06, + "loss": 0.5069, + "step": 11045 + }, + { + "epoch": 3.878511235955056, + "grad_norm": 0.9392685890197754, + "learning_rate": 1.404829392709392e-06, + "loss": 0.5819, + "step": 11046 + }, + { + "epoch": 3.8788623595505616, + "grad_norm": 0.9276441931724548, + "learning_rate": 1.4044138079684522e-06, + "loss": 0.5466, + "step": 11047 + }, + { + "epoch": 3.879213483146067, + "grad_norm": 0.9512078762054443, + "learning_rate": 1.4039982606956227e-06, + "loss": 0.5466, + "step": 11048 + }, + { + "epoch": 3.8795646067415728, + "grad_norm": 0.9594552516937256, + "learning_rate": 1.4035827509051137e-06, + "loss": 0.621, + "step": 11049 + }, + { + "epoch": 3.879915730337079, + "grad_norm": 0.949516236782074, + "learning_rate": 1.4031672786111375e-06, + "loss": 0.5696, + "step": 11050 + }, + { + "epoch": 3.8802668539325844, + "grad_norm": 0.9366505146026611, + "learning_rate": 1.4027518438279009e-06, + "loss": 0.5228, + "step": 11051 + }, + { + "epoch": 3.88061797752809, + "grad_norm": 0.9496411085128784, + "learning_rate": 1.4023364465696131e-06, + "loss": 0.5602, + "step": 11052 + }, + { + "epoch": 3.8809691011235956, + "grad_norm": 0.9488803744316101, + "learning_rate": 1.4019210868504788e-06, + "loss": 0.5325, + "step": 11053 + }, + { + "epoch": 3.881320224719101, + "grad_norm": 0.9927105903625488, + "learning_rate": 1.4015057646847044e-06, + "loss": 0.5692, + "step": 11054 + }, + { + "epoch": 3.881671348314607, + "grad_norm": 0.9643698930740356, + "learning_rate": 1.401090480086493e-06, + "loss": 0.5915, + "step": 11055 + }, + { + "epoch": 3.8820224719101124, + "grad_norm": 0.9595552086830139, + "learning_rate": 1.4006752330700456e-06, + "loss": 0.5726, + "step": 11056 + }, + { + "epoch": 3.882373595505618, + "grad_norm": 0.9714942574501038, + "learning_rate": 1.4002600236495662e-06, + "loss": 0.5719, + "step": 11057 + }, + { + "epoch": 3.8827247191011236, + "grad_norm": 0.9988847970962524, + "learning_rate": 1.3998448518392516e-06, + "loss": 0.5801, + "step": 11058 + }, + { + "epoch": 3.883075842696629, + "grad_norm": 0.9708136320114136, + "learning_rate": 1.3994297176533023e-06, + "loss": 0.5593, + "step": 11059 + }, + { + "epoch": 3.883426966292135, + "grad_norm": 0.934221088886261, + "learning_rate": 1.3990146211059141e-06, + "loss": 0.5406, + "step": 11060 + }, + { + "epoch": 3.8837780898876404, + "grad_norm": 0.9569685459136963, + "learning_rate": 1.3985995622112847e-06, + "loss": 0.5831, + "step": 11061 + }, + { + "epoch": 3.884129213483146, + "grad_norm": 0.9517682790756226, + "learning_rate": 1.3981845409836082e-06, + "loss": 0.5239, + "step": 11062 + }, + { + "epoch": 3.8844803370786516, + "grad_norm": 0.9420996308326721, + "learning_rate": 1.397769557437076e-06, + "loss": 0.5276, + "step": 11063 + }, + { + "epoch": 3.884831460674157, + "grad_norm": 0.9268381595611572, + "learning_rate": 1.397354611585883e-06, + "loss": 0.5511, + "step": 11064 + }, + { + "epoch": 3.885182584269663, + "grad_norm": 0.939054012298584, + "learning_rate": 1.3969397034442178e-06, + "loss": 0.5557, + "step": 11065 + }, + { + "epoch": 3.8855337078651684, + "grad_norm": 0.9306736588478088, + "learning_rate": 1.3965248330262714e-06, + "loss": 0.5659, + "step": 11066 + }, + { + "epoch": 3.8858848314606744, + "grad_norm": 0.9777511954307556, + "learning_rate": 1.3961100003462308e-06, + "loss": 0.5651, + "step": 11067 + }, + { + "epoch": 3.88623595505618, + "grad_norm": 0.9447049498558044, + "learning_rate": 1.3956952054182849e-06, + "loss": 0.5769, + "step": 11068 + }, + { + "epoch": 3.8865870786516856, + "grad_norm": 0.95235276222229, + "learning_rate": 1.3952804482566173e-06, + "loss": 0.5529, + "step": 11069 + }, + { + "epoch": 3.886938202247191, + "grad_norm": 0.9654942750930786, + "learning_rate": 1.3948657288754125e-06, + "loss": 0.5688, + "step": 11070 + }, + { + "epoch": 3.887289325842697, + "grad_norm": 0.9709132313728333, + "learning_rate": 1.394451047288855e-06, + "loss": 0.5232, + "step": 11071 + }, + { + "epoch": 3.8876404494382024, + "grad_norm": 0.9386630654335022, + "learning_rate": 1.3940364035111248e-06, + "loss": 0.5644, + "step": 11072 + }, + { + "epoch": 3.887991573033708, + "grad_norm": 0.9413669109344482, + "learning_rate": 1.3936217975564042e-06, + "loss": 0.5578, + "step": 11073 + }, + { + "epoch": 3.8883426966292136, + "grad_norm": 0.9315128922462463, + "learning_rate": 1.3932072294388701e-06, + "loss": 0.541, + "step": 11074 + }, + { + "epoch": 3.888693820224719, + "grad_norm": 0.9553053379058838, + "learning_rate": 1.3927926991727027e-06, + "loss": 0.5729, + "step": 11075 + }, + { + "epoch": 3.889044943820225, + "grad_norm": 0.9553484320640564, + "learning_rate": 1.392378206772077e-06, + "loss": 0.5521, + "step": 11076 + }, + { + "epoch": 3.8893960674157304, + "grad_norm": 0.9333077669143677, + "learning_rate": 1.3919637522511692e-06, + "loss": 0.5951, + "step": 11077 + }, + { + "epoch": 3.889747191011236, + "grad_norm": 1.002480387687683, + "learning_rate": 1.3915493356241532e-06, + "loss": 0.5646, + "step": 11078 + }, + { + "epoch": 3.8900983146067416, + "grad_norm": 0.9518157243728638, + "learning_rate": 1.3911349569052005e-06, + "loss": 0.5772, + "step": 11079 + }, + { + "epoch": 3.890449438202247, + "grad_norm": 0.9598498344421387, + "learning_rate": 1.3907206161084843e-06, + "loss": 0.5888, + "step": 11080 + }, + { + "epoch": 3.890800561797753, + "grad_norm": 1.0056177377700806, + "learning_rate": 1.3903063132481728e-06, + "loss": 0.5553, + "step": 11081 + }, + { + "epoch": 3.8911516853932584, + "grad_norm": 0.9450264573097229, + "learning_rate": 1.389892048338437e-06, + "loss": 0.5561, + "step": 11082 + }, + { + "epoch": 3.891502808988764, + "grad_norm": 0.921626091003418, + "learning_rate": 1.3894778213934418e-06, + "loss": 0.5883, + "step": 11083 + }, + { + "epoch": 3.8918539325842696, + "grad_norm": 0.9008750319480896, + "learning_rate": 1.389063632427356e-06, + "loss": 0.5508, + "step": 11084 + }, + { + "epoch": 3.892205056179775, + "grad_norm": 0.9517634510993958, + "learning_rate": 1.3886494814543425e-06, + "loss": 0.5743, + "step": 11085 + }, + { + "epoch": 3.892556179775281, + "grad_norm": 0.9277675151824951, + "learning_rate": 1.3882353684885666e-06, + "loss": 0.5573, + "step": 11086 + }, + { + "epoch": 3.8929073033707864, + "grad_norm": 0.9603415727615356, + "learning_rate": 1.3878212935441887e-06, + "loss": 0.5602, + "step": 11087 + }, + { + "epoch": 3.893258426966292, + "grad_norm": 0.9504032135009766, + "learning_rate": 1.3874072566353718e-06, + "loss": 0.5544, + "step": 11088 + }, + { + "epoch": 3.8936095505617976, + "grad_norm": 0.9563642740249634, + "learning_rate": 1.386993257776274e-06, + "loss": 0.5621, + "step": 11089 + }, + { + "epoch": 3.893960674157303, + "grad_norm": 0.9566351771354675, + "learning_rate": 1.3865792969810552e-06, + "loss": 0.5704, + "step": 11090 + }, + { + "epoch": 3.894311797752809, + "grad_norm": 0.9284030199050903, + "learning_rate": 1.3861653742638714e-06, + "loss": 0.5539, + "step": 11091 + }, + { + "epoch": 3.8946629213483144, + "grad_norm": 0.9233971238136292, + "learning_rate": 1.385751489638878e-06, + "loss": 0.543, + "step": 11092 + }, + { + "epoch": 3.89501404494382, + "grad_norm": 0.9435145258903503, + "learning_rate": 1.3853376431202312e-06, + "loss": 0.542, + "step": 11093 + }, + { + "epoch": 3.8953651685393256, + "grad_norm": 0.9272922277450562, + "learning_rate": 1.3849238347220822e-06, + "loss": 0.5191, + "step": 11094 + }, + { + "epoch": 3.8957162921348316, + "grad_norm": 0.9704311490058899, + "learning_rate": 1.3845100644585846e-06, + "loss": 0.5782, + "step": 11095 + }, + { + "epoch": 3.896067415730337, + "grad_norm": 0.936034083366394, + "learning_rate": 1.3840963323438872e-06, + "loss": 0.574, + "step": 11096 + }, + { + "epoch": 3.896418539325843, + "grad_norm": 0.9400407075881958, + "learning_rate": 1.3836826383921416e-06, + "loss": 0.5532, + "step": 11097 + }, + { + "epoch": 3.8967696629213484, + "grad_norm": 0.9423049688339233, + "learning_rate": 1.3832689826174937e-06, + "loss": 0.6011, + "step": 11098 + }, + { + "epoch": 3.897120786516854, + "grad_norm": 0.9250279664993286, + "learning_rate": 1.3828553650340915e-06, + "loss": 0.555, + "step": 11099 + }, + { + "epoch": 3.8974719101123596, + "grad_norm": 1.055710792541504, + "learning_rate": 1.3824417856560793e-06, + "loss": 0.5505, + "step": 11100 + }, + { + "epoch": 3.897823033707865, + "grad_norm": 0.9303153157234192, + "learning_rate": 1.3820282444976025e-06, + "loss": 0.5314, + "step": 11101 + }, + { + "epoch": 3.898174157303371, + "grad_norm": 0.9698507785797119, + "learning_rate": 1.3816147415728031e-06, + "loss": 0.5748, + "step": 11102 + }, + { + "epoch": 3.8985252808988764, + "grad_norm": 1.004014253616333, + "learning_rate": 1.3812012768958216e-06, + "loss": 0.563, + "step": 11103 + }, + { + "epoch": 3.898876404494382, + "grad_norm": 0.9176561832427979, + "learning_rate": 1.3807878504807998e-06, + "loss": 0.5754, + "step": 11104 + }, + { + "epoch": 3.8992275280898876, + "grad_norm": 0.9259368181228638, + "learning_rate": 1.3803744623418753e-06, + "loss": 0.5025, + "step": 11105 + }, + { + "epoch": 3.899578651685393, + "grad_norm": 0.9314144253730774, + "learning_rate": 1.3799611124931867e-06, + "loss": 0.5365, + "step": 11106 + }, + { + "epoch": 3.899929775280899, + "grad_norm": 0.9545547366142273, + "learning_rate": 1.3795478009488686e-06, + "loss": 0.5436, + "step": 11107 + }, + { + "epoch": 3.9002808988764044, + "grad_norm": 0.9530776143074036, + "learning_rate": 1.3791345277230579e-06, + "loss": 0.5201, + "step": 11108 + }, + { + "epoch": 3.90063202247191, + "grad_norm": 0.9544113278388977, + "learning_rate": 1.3787212928298864e-06, + "loss": 0.5452, + "step": 11109 + }, + { + "epoch": 3.9009831460674156, + "grad_norm": 0.9274408221244812, + "learning_rate": 1.3783080962834883e-06, + "loss": 0.5272, + "step": 11110 + }, + { + "epoch": 3.901334269662921, + "grad_norm": 0.9226788282394409, + "learning_rate": 1.3778949380979922e-06, + "loss": 0.5804, + "step": 11111 + }, + { + "epoch": 3.9016853932584272, + "grad_norm": 0.9695422053337097, + "learning_rate": 1.3774818182875298e-06, + "loss": 0.5685, + "step": 11112 + }, + { + "epoch": 3.902036516853933, + "grad_norm": 0.9480987191200256, + "learning_rate": 1.3770687368662284e-06, + "loss": 0.5125, + "step": 11113 + }, + { + "epoch": 3.9023876404494384, + "grad_norm": 0.9380653500556946, + "learning_rate": 1.3766556938482156e-06, + "loss": 0.5298, + "step": 11114 + }, + { + "epoch": 3.902738764044944, + "grad_norm": 0.940731406211853, + "learning_rate": 1.3762426892476172e-06, + "loss": 0.5419, + "step": 11115 + }, + { + "epoch": 3.9030898876404496, + "grad_norm": 0.9188136458396912, + "learning_rate": 1.3758297230785561e-06, + "loss": 0.5279, + "step": 11116 + }, + { + "epoch": 3.903441011235955, + "grad_norm": 0.9138721823692322, + "learning_rate": 1.3754167953551576e-06, + "loss": 0.5264, + "step": 11117 + }, + { + "epoch": 3.903792134831461, + "grad_norm": 0.9503632187843323, + "learning_rate": 1.3750039060915413e-06, + "loss": 0.5584, + "step": 11118 + }, + { + "epoch": 3.9041432584269664, + "grad_norm": 0.949775218963623, + "learning_rate": 1.3745910553018296e-06, + "loss": 0.545, + "step": 11119 + }, + { + "epoch": 3.904494382022472, + "grad_norm": 0.9215615391731262, + "learning_rate": 1.3741782430001398e-06, + "loss": 0.5454, + "step": 11120 + }, + { + "epoch": 3.9048455056179776, + "grad_norm": 0.9466119408607483, + "learning_rate": 1.3737654692005919e-06, + "loss": 0.5685, + "step": 11121 + }, + { + "epoch": 3.905196629213483, + "grad_norm": 0.9724555611610413, + "learning_rate": 1.3733527339173003e-06, + "loss": 0.5499, + "step": 11122 + }, + { + "epoch": 3.905547752808989, + "grad_norm": 0.9373450875282288, + "learning_rate": 1.3729400371643819e-06, + "loss": 0.5735, + "step": 11123 + }, + { + "epoch": 3.9058988764044944, + "grad_norm": 0.9042215347290039, + "learning_rate": 1.3725273789559487e-06, + "loss": 0.5352, + "step": 11124 + }, + { + "epoch": 3.90625, + "grad_norm": 0.9520610570907593, + "learning_rate": 1.3721147593061156e-06, + "loss": 0.5634, + "step": 11125 + }, + { + "epoch": 3.9066011235955056, + "grad_norm": 0.9212858080863953, + "learning_rate": 1.3717021782289913e-06, + "loss": 0.5376, + "step": 11126 + }, + { + "epoch": 3.906952247191011, + "grad_norm": 0.9610831141471863, + "learning_rate": 1.3712896357386884e-06, + "loss": 0.5854, + "step": 11127 + }, + { + "epoch": 3.907303370786517, + "grad_norm": 0.942595899105072, + "learning_rate": 1.3708771318493136e-06, + "loss": 0.5505, + "step": 11128 + }, + { + "epoch": 3.9076544943820224, + "grad_norm": 0.9638190269470215, + "learning_rate": 1.3704646665749744e-06, + "loss": 0.5549, + "step": 11129 + }, + { + "epoch": 3.908005617977528, + "grad_norm": 0.9971073865890503, + "learning_rate": 1.3700522399297771e-06, + "loss": 0.594, + "step": 11130 + }, + { + "epoch": 3.9083567415730336, + "grad_norm": 0.9519518613815308, + "learning_rate": 1.369639851927826e-06, + "loss": 0.5439, + "step": 11131 + }, + { + "epoch": 3.908707865168539, + "grad_norm": 0.9552960395812988, + "learning_rate": 1.3692275025832258e-06, + "loss": 0.5498, + "step": 11132 + }, + { + "epoch": 3.909058988764045, + "grad_norm": 0.9835587739944458, + "learning_rate": 1.368815191910076e-06, + "loss": 0.5974, + "step": 11133 + }, + { + "epoch": 3.9094101123595504, + "grad_norm": 0.9330475330352783, + "learning_rate": 1.3684029199224797e-06, + "loss": 0.5203, + "step": 11134 + }, + { + "epoch": 3.909761235955056, + "grad_norm": 0.9409958720207214, + "learning_rate": 1.3679906866345344e-06, + "loss": 0.5353, + "step": 11135 + }, + { + "epoch": 3.9101123595505616, + "grad_norm": 0.9464348554611206, + "learning_rate": 1.3675784920603396e-06, + "loss": 0.5523, + "step": 11136 + }, + { + "epoch": 3.910463483146067, + "grad_norm": 0.941512942314148, + "learning_rate": 1.3671663362139907e-06, + "loss": 0.5941, + "step": 11137 + }, + { + "epoch": 3.9108146067415728, + "grad_norm": 0.9680454134941101, + "learning_rate": 1.3667542191095845e-06, + "loss": 0.5602, + "step": 11138 + }, + { + "epoch": 3.911165730337079, + "grad_norm": 0.9763267636299133, + "learning_rate": 1.3663421407612134e-06, + "loss": 0.5785, + "step": 11139 + }, + { + "epoch": 3.9115168539325844, + "grad_norm": 0.9459673762321472, + "learning_rate": 1.365930101182972e-06, + "loss": 0.57, + "step": 11140 + }, + { + "epoch": 3.91186797752809, + "grad_norm": 0.9394272565841675, + "learning_rate": 1.3655181003889507e-06, + "loss": 0.5611, + "step": 11141 + }, + { + "epoch": 3.9122191011235956, + "grad_norm": 0.9448548555374146, + "learning_rate": 1.3651061383932385e-06, + "loss": 0.5604, + "step": 11142 + }, + { + "epoch": 3.912570224719101, + "grad_norm": 0.9488401412963867, + "learning_rate": 1.3646942152099262e-06, + "loss": 0.5407, + "step": 11143 + }, + { + "epoch": 3.912921348314607, + "grad_norm": 0.9081423282623291, + "learning_rate": 1.3642823308530995e-06, + "loss": 0.5261, + "step": 11144 + }, + { + "epoch": 3.9132724719101124, + "grad_norm": 0.9627257585525513, + "learning_rate": 1.363870485336846e-06, + "loss": 0.5649, + "step": 11145 + }, + { + "epoch": 3.913623595505618, + "grad_norm": 0.9672406911849976, + "learning_rate": 1.3634586786752485e-06, + "loss": 0.5261, + "step": 11146 + }, + { + "epoch": 3.9139747191011236, + "grad_norm": 0.9318965077400208, + "learning_rate": 1.363046910882393e-06, + "loss": 0.5502, + "step": 11147 + }, + { + "epoch": 3.914325842696629, + "grad_norm": 0.9222128391265869, + "learning_rate": 1.3626351819723594e-06, + "loss": 0.5425, + "step": 11148 + }, + { + "epoch": 3.914676966292135, + "grad_norm": 1.066196084022522, + "learning_rate": 1.3622234919592297e-06, + "loss": 0.5078, + "step": 11149 + }, + { + "epoch": 3.9150280898876404, + "grad_norm": 0.964749276638031, + "learning_rate": 1.3618118408570824e-06, + "loss": 0.5766, + "step": 11150 + }, + { + "epoch": 3.915379213483146, + "grad_norm": 0.9151565432548523, + "learning_rate": 1.3614002286799972e-06, + "loss": 0.519, + "step": 11151 + }, + { + "epoch": 3.9157303370786516, + "grad_norm": 0.9877290725708008, + "learning_rate": 1.3609886554420487e-06, + "loss": 0.5642, + "step": 11152 + }, + { + "epoch": 3.916081460674157, + "grad_norm": 0.9168569445610046, + "learning_rate": 1.3605771211573146e-06, + "loss": 0.5209, + "step": 11153 + }, + { + "epoch": 3.916432584269663, + "grad_norm": 0.9288423657417297, + "learning_rate": 1.360165625839868e-06, + "loss": 0.5407, + "step": 11154 + }, + { + "epoch": 3.9167837078651684, + "grad_norm": 0.9477772116661072, + "learning_rate": 1.3597541695037804e-06, + "loss": 0.5762, + "step": 11155 + }, + { + "epoch": 3.9171348314606744, + "grad_norm": 0.9571468830108643, + "learning_rate": 1.3593427521631258e-06, + "loss": 0.5542, + "step": 11156 + }, + { + "epoch": 3.91748595505618, + "grad_norm": 0.915473222732544, + "learning_rate": 1.3589313738319719e-06, + "loss": 0.5112, + "step": 11157 + }, + { + "epoch": 3.9178370786516856, + "grad_norm": 0.9468619227409363, + "learning_rate": 1.3585200345243892e-06, + "loss": 0.5338, + "step": 11158 + }, + { + "epoch": 3.918188202247191, + "grad_norm": 0.944211483001709, + "learning_rate": 1.3581087342544441e-06, + "loss": 0.5767, + "step": 11159 + }, + { + "epoch": 3.918539325842697, + "grad_norm": 0.9007745981216431, + "learning_rate": 1.3576974730362039e-06, + "loss": 0.5176, + "step": 11160 + }, + { + "epoch": 3.9188904494382024, + "grad_norm": 0.9494607448577881, + "learning_rate": 1.3572862508837315e-06, + "loss": 0.5419, + "step": 11161 + }, + { + "epoch": 3.919241573033708, + "grad_norm": 0.9424705505371094, + "learning_rate": 1.3568750678110927e-06, + "loss": 0.5515, + "step": 11162 + }, + { + "epoch": 3.9195926966292136, + "grad_norm": 0.9772363305091858, + "learning_rate": 1.3564639238323475e-06, + "loss": 0.5434, + "step": 11163 + }, + { + "epoch": 3.919943820224719, + "grad_norm": 0.9162544012069702, + "learning_rate": 1.3560528189615586e-06, + "loss": 0.5538, + "step": 11164 + }, + { + "epoch": 3.920294943820225, + "grad_norm": 0.9484769105911255, + "learning_rate": 1.3556417532127841e-06, + "loss": 0.5628, + "step": 11165 + }, + { + "epoch": 3.9206460674157304, + "grad_norm": 0.9486870169639587, + "learning_rate": 1.3552307266000814e-06, + "loss": 0.5292, + "step": 11166 + }, + { + "epoch": 3.920997191011236, + "grad_norm": 0.9628684520721436, + "learning_rate": 1.3548197391375093e-06, + "loss": 0.541, + "step": 11167 + }, + { + "epoch": 3.9213483146067416, + "grad_norm": 0.9496630430221558, + "learning_rate": 1.3544087908391213e-06, + "loss": 0.5523, + "step": 11168 + }, + { + "epoch": 3.921699438202247, + "grad_norm": 0.9491123557090759, + "learning_rate": 1.3539978817189731e-06, + "loss": 0.5682, + "step": 11169 + }, + { + "epoch": 3.922050561797753, + "grad_norm": 0.9477087259292603, + "learning_rate": 1.353587011791116e-06, + "loss": 0.5477, + "step": 11170 + }, + { + "epoch": 3.9224016853932584, + "grad_norm": 0.9674578905105591, + "learning_rate": 1.3531761810696031e-06, + "loss": 0.5695, + "step": 11171 + }, + { + "epoch": 3.922752808988764, + "grad_norm": 0.9546025395393372, + "learning_rate": 1.3527653895684823e-06, + "loss": 0.5597, + "step": 11172 + }, + { + "epoch": 3.9231039325842696, + "grad_norm": 0.958831250667572, + "learning_rate": 1.3523546373018045e-06, + "loss": 0.5388, + "step": 11173 + }, + { + "epoch": 3.923455056179775, + "grad_norm": 0.9288198351860046, + "learning_rate": 1.3519439242836152e-06, + "loss": 0.5855, + "step": 11174 + }, + { + "epoch": 3.923806179775281, + "grad_norm": 0.9407035112380981, + "learning_rate": 1.3515332505279622e-06, + "loss": 0.5659, + "step": 11175 + }, + { + "epoch": 3.9241573033707864, + "grad_norm": 0.9221037030220032, + "learning_rate": 1.351122616048888e-06, + "loss": 0.5543, + "step": 11176 + }, + { + "epoch": 3.924508426966292, + "grad_norm": 0.902523934841156, + "learning_rate": 1.3507120208604385e-06, + "loss": 0.5046, + "step": 11177 + }, + { + "epoch": 3.9248595505617976, + "grad_norm": 0.9847588539123535, + "learning_rate": 1.3503014649766544e-06, + "loss": 0.5739, + "step": 11178 + }, + { + "epoch": 3.925210674157303, + "grad_norm": 0.931635320186615, + "learning_rate": 1.3498909484115752e-06, + "loss": 0.5492, + "step": 11179 + }, + { + "epoch": 3.925561797752809, + "grad_norm": 0.9266260862350464, + "learning_rate": 1.3494804711792425e-06, + "loss": 0.5493, + "step": 11180 + }, + { + "epoch": 3.9259129213483144, + "grad_norm": 0.9200179576873779, + "learning_rate": 1.3490700332936923e-06, + "loss": 0.5274, + "step": 11181 + }, + { + "epoch": 3.92626404494382, + "grad_norm": 0.9368164539337158, + "learning_rate": 1.3486596347689632e-06, + "loss": 0.5552, + "step": 11182 + }, + { + "epoch": 3.9266151685393256, + "grad_norm": 0.9406830668449402, + "learning_rate": 1.348249275619088e-06, + "loss": 0.5812, + "step": 11183 + }, + { + "epoch": 3.9269662921348316, + "grad_norm": 0.9513319134712219, + "learning_rate": 1.3478389558581035e-06, + "loss": 0.5254, + "step": 11184 + }, + { + "epoch": 3.927317415730337, + "grad_norm": 0.9666449427604675, + "learning_rate": 1.3474286755000393e-06, + "loss": 0.5339, + "step": 11185 + }, + { + "epoch": 3.927668539325843, + "grad_norm": 0.9189738631248474, + "learning_rate": 1.3470184345589292e-06, + "loss": 0.563, + "step": 11186 + }, + { + "epoch": 3.9280196629213484, + "grad_norm": 0.9651278853416443, + "learning_rate": 1.3466082330488014e-06, + "loss": 0.5428, + "step": 11187 + }, + { + "epoch": 3.928370786516854, + "grad_norm": 0.9227038621902466, + "learning_rate": 1.3461980709836857e-06, + "loss": 0.5158, + "step": 11188 + }, + { + "epoch": 3.9287219101123596, + "grad_norm": 0.9196301102638245, + "learning_rate": 1.345787948377608e-06, + "loss": 0.5589, + "step": 11189 + }, + { + "epoch": 3.929073033707865, + "grad_norm": 0.9552080035209656, + "learning_rate": 1.3453778652445953e-06, + "loss": 0.5305, + "step": 11190 + }, + { + "epoch": 3.929424157303371, + "grad_norm": 0.9399698376655579, + "learning_rate": 1.344967821598672e-06, + "loss": 0.5559, + "step": 11191 + }, + { + "epoch": 3.9297752808988764, + "grad_norm": 0.9559733271598816, + "learning_rate": 1.3445578174538597e-06, + "loss": 0.5469, + "step": 11192 + }, + { + "epoch": 3.930126404494382, + "grad_norm": 0.928051233291626, + "learning_rate": 1.3441478528241822e-06, + "loss": 0.5209, + "step": 11193 + }, + { + "epoch": 3.9304775280898876, + "grad_norm": 0.9554131031036377, + "learning_rate": 1.3437379277236583e-06, + "loss": 0.5902, + "step": 11194 + }, + { + "epoch": 3.930828651685393, + "grad_norm": 0.9383732676506042, + "learning_rate": 1.3433280421663086e-06, + "loss": 0.5377, + "step": 11195 + }, + { + "epoch": 3.931179775280899, + "grad_norm": 0.9545925259590149, + "learning_rate": 1.3429181961661491e-06, + "loss": 0.5785, + "step": 11196 + }, + { + "epoch": 3.9315308988764044, + "grad_norm": 0.9433280229568481, + "learning_rate": 1.3425083897371983e-06, + "loss": 0.5732, + "step": 11197 + }, + { + "epoch": 3.93188202247191, + "grad_norm": 0.9631233215332031, + "learning_rate": 1.342098622893469e-06, + "loss": 0.5503, + "step": 11198 + }, + { + "epoch": 3.9322331460674156, + "grad_norm": 0.9394797682762146, + "learning_rate": 1.341688895648977e-06, + "loss": 0.5669, + "step": 11199 + }, + { + "epoch": 3.932584269662921, + "grad_norm": 0.9129445552825928, + "learning_rate": 1.3412792080177326e-06, + "loss": 0.5385, + "step": 11200 + }, + { + "epoch": 3.9329353932584272, + "grad_norm": 0.9481983780860901, + "learning_rate": 1.3408695600137487e-06, + "loss": 0.5515, + "step": 11201 + }, + { + "epoch": 3.933286516853933, + "grad_norm": 0.9398009181022644, + "learning_rate": 1.3404599516510331e-06, + "loss": 0.5392, + "step": 11202 + }, + { + "epoch": 3.9336376404494384, + "grad_norm": 0.9436145424842834, + "learning_rate": 1.3400503829435957e-06, + "loss": 0.5369, + "step": 11203 + }, + { + "epoch": 3.933988764044944, + "grad_norm": 0.9475415349006653, + "learning_rate": 1.3396408539054428e-06, + "loss": 0.5857, + "step": 11204 + }, + { + "epoch": 3.9343398876404496, + "grad_norm": 0.9420165419578552, + "learning_rate": 1.3392313645505787e-06, + "loss": 0.5554, + "step": 11205 + }, + { + "epoch": 3.934691011235955, + "grad_norm": 0.9359337687492371, + "learning_rate": 1.3388219148930096e-06, + "loss": 0.5437, + "step": 11206 + }, + { + "epoch": 3.935042134831461, + "grad_norm": 0.9774880409240723, + "learning_rate": 1.3384125049467364e-06, + "loss": 0.5706, + "step": 11207 + }, + { + "epoch": 3.9353932584269664, + "grad_norm": 0.9382474422454834, + "learning_rate": 1.3380031347257627e-06, + "loss": 0.5737, + "step": 11208 + }, + { + "epoch": 3.935744382022472, + "grad_norm": 0.9294614791870117, + "learning_rate": 1.3375938042440862e-06, + "loss": 0.5862, + "step": 11209 + }, + { + "epoch": 3.9360955056179776, + "grad_norm": 0.9472325444221497, + "learning_rate": 1.337184513515708e-06, + "loss": 0.5819, + "step": 11210 + }, + { + "epoch": 3.936446629213483, + "grad_norm": 0.9588031768798828, + "learning_rate": 1.3367752625546235e-06, + "loss": 0.5765, + "step": 11211 + }, + { + "epoch": 3.936797752808989, + "grad_norm": 0.9210266470909119, + "learning_rate": 1.3363660513748305e-06, + "loss": 0.54, + "step": 11212 + }, + { + "epoch": 3.9371488764044944, + "grad_norm": 0.964421808719635, + "learning_rate": 1.3359568799903227e-06, + "loss": 0.5705, + "step": 11213 + }, + { + "epoch": 3.9375, + "grad_norm": 0.9626476168632507, + "learning_rate": 1.335547748415093e-06, + "loss": 0.5459, + "step": 11214 + }, + { + "epoch": 3.9378511235955056, + "grad_norm": 0.9146947860717773, + "learning_rate": 1.3351386566631344e-06, + "loss": 0.5528, + "step": 11215 + }, + { + "epoch": 3.938202247191011, + "grad_norm": 0.9370722770690918, + "learning_rate": 1.3347296047484361e-06, + "loss": 0.5559, + "step": 11216 + }, + { + "epoch": 3.938553370786517, + "grad_norm": 0.9622222185134888, + "learning_rate": 1.3343205926849892e-06, + "loss": 0.5682, + "step": 11217 + }, + { + "epoch": 3.9389044943820224, + "grad_norm": 0.9179551005363464, + "learning_rate": 1.3339116204867797e-06, + "loss": 0.5675, + "step": 11218 + }, + { + "epoch": 3.939255617977528, + "grad_norm": 0.953270435333252, + "learning_rate": 1.3335026881677956e-06, + "loss": 0.5705, + "step": 11219 + }, + { + "epoch": 3.9396067415730336, + "grad_norm": 0.9308480620384216, + "learning_rate": 1.3330937957420206e-06, + "loss": 0.5927, + "step": 11220 + }, + { + "epoch": 3.939957865168539, + "grad_norm": 0.9272353053092957, + "learning_rate": 1.3326849432234402e-06, + "loss": 0.5352, + "step": 11221 + }, + { + "epoch": 3.940308988764045, + "grad_norm": 0.9346489906311035, + "learning_rate": 1.3322761306260359e-06, + "loss": 0.5501, + "step": 11222 + }, + { + "epoch": 3.9406601123595504, + "grad_norm": 0.9393819570541382, + "learning_rate": 1.3318673579637875e-06, + "loss": 0.5554, + "step": 11223 + }, + { + "epoch": 3.941011235955056, + "grad_norm": 0.9396387338638306, + "learning_rate": 1.3314586252506773e-06, + "loss": 0.5662, + "step": 11224 + }, + { + "epoch": 3.9413623595505616, + "grad_norm": 0.9246081709861755, + "learning_rate": 1.3310499325006809e-06, + "loss": 0.5467, + "step": 11225 + }, + { + "epoch": 3.941713483146067, + "grad_norm": 0.9252482056617737, + "learning_rate": 1.3306412797277774e-06, + "loss": 0.5498, + "step": 11226 + }, + { + "epoch": 3.9420646067415728, + "grad_norm": 0.9111382365226746, + "learning_rate": 1.3302326669459409e-06, + "loss": 0.5909, + "step": 11227 + }, + { + "epoch": 3.942415730337079, + "grad_norm": 0.927321195602417, + "learning_rate": 1.3298240941691467e-06, + "loss": 0.5432, + "step": 11228 + }, + { + "epoch": 3.9427668539325844, + "grad_norm": 0.8950797915458679, + "learning_rate": 1.3294155614113674e-06, + "loss": 0.519, + "step": 11229 + }, + { + "epoch": 3.94311797752809, + "grad_norm": 0.931207537651062, + "learning_rate": 1.3290070686865733e-06, + "loss": 0.5252, + "step": 11230 + }, + { + "epoch": 3.9434691011235956, + "grad_norm": 0.9372845888137817, + "learning_rate": 1.3285986160087364e-06, + "loss": 0.5367, + "step": 11231 + }, + { + "epoch": 3.943820224719101, + "grad_norm": 0.9238448143005371, + "learning_rate": 1.3281902033918234e-06, + "loss": 0.5499, + "step": 11232 + }, + { + "epoch": 3.944171348314607, + "grad_norm": 0.9084984660148621, + "learning_rate": 1.3277818308498036e-06, + "loss": 0.5027, + "step": 11233 + }, + { + "epoch": 3.9445224719101124, + "grad_norm": 0.9488505721092224, + "learning_rate": 1.3273734983966414e-06, + "loss": 0.5455, + "step": 11234 + }, + { + "epoch": 3.944873595505618, + "grad_norm": 0.9351166486740112, + "learning_rate": 1.3269652060463028e-06, + "loss": 0.5275, + "step": 11235 + }, + { + "epoch": 3.9452247191011236, + "grad_norm": 0.9588375687599182, + "learning_rate": 1.3265569538127505e-06, + "loss": 0.5825, + "step": 11236 + }, + { + "epoch": 3.945575842696629, + "grad_norm": 0.9853055477142334, + "learning_rate": 1.326148741709945e-06, + "loss": 0.5584, + "step": 11237 + }, + { + "epoch": 3.945926966292135, + "grad_norm": 0.9758622646331787, + "learning_rate": 1.3257405697518495e-06, + "loss": 0.5374, + "step": 11238 + }, + { + "epoch": 3.9462780898876404, + "grad_norm": 0.9346928000450134, + "learning_rate": 1.3253324379524204e-06, + "loss": 0.556, + "step": 11239 + }, + { + "epoch": 3.946629213483146, + "grad_norm": 0.9200893044471741, + "learning_rate": 1.324924346325618e-06, + "loss": 0.5499, + "step": 11240 + }, + { + "epoch": 3.9469803370786516, + "grad_norm": 0.934998631477356, + "learning_rate": 1.3245162948853961e-06, + "loss": 0.5516, + "step": 11241 + }, + { + "epoch": 3.947331460674157, + "grad_norm": 0.9439476132392883, + "learning_rate": 1.3241082836457123e-06, + "loss": 0.5447, + "step": 11242 + }, + { + "epoch": 3.947682584269663, + "grad_norm": 0.9115414023399353, + "learning_rate": 1.323700312620519e-06, + "loss": 0.53, + "step": 11243 + }, + { + "epoch": 3.9480337078651684, + "grad_norm": 0.949103832244873, + "learning_rate": 1.323292381823767e-06, + "loss": 0.5704, + "step": 11244 + }, + { + "epoch": 3.9483848314606744, + "grad_norm": 0.9499713778495789, + "learning_rate": 1.32288449126941e-06, + "loss": 0.5496, + "step": 11245 + }, + { + "epoch": 3.94873595505618, + "grad_norm": 0.957030713558197, + "learning_rate": 1.3224766409713952e-06, + "loss": 0.5882, + "step": 11246 + }, + { + "epoch": 3.9490870786516856, + "grad_norm": 0.9729870557785034, + "learning_rate": 1.3220688309436724e-06, + "loss": 0.5656, + "step": 11247 + }, + { + "epoch": 3.949438202247191, + "grad_norm": 0.9273895025253296, + "learning_rate": 1.3216610612001868e-06, + "loss": 0.5535, + "step": 11248 + }, + { + "epoch": 3.949789325842697, + "grad_norm": 0.9330081343650818, + "learning_rate": 1.3212533317548853e-06, + "loss": 0.5305, + "step": 11249 + }, + { + "epoch": 3.9501404494382024, + "grad_norm": 0.938386857509613, + "learning_rate": 1.3208456426217104e-06, + "loss": 0.5687, + "step": 11250 + }, + { + "epoch": 3.950491573033708, + "grad_norm": 0.9170256853103638, + "learning_rate": 1.3204379938146068e-06, + "loss": 0.5271, + "step": 11251 + }, + { + "epoch": 3.9508426966292136, + "grad_norm": 0.9770430326461792, + "learning_rate": 1.320030385347513e-06, + "loss": 0.5984, + "step": 11252 + }, + { + "epoch": 3.951193820224719, + "grad_norm": 0.9376773238182068, + "learning_rate": 1.3196228172343714e-06, + "loss": 0.5602, + "step": 11253 + }, + { + "epoch": 3.951544943820225, + "grad_norm": 0.9167219996452332, + "learning_rate": 1.3192152894891195e-06, + "loss": 0.5363, + "step": 11254 + }, + { + "epoch": 3.9518960674157304, + "grad_norm": 0.9476504325866699, + "learning_rate": 1.3188078021256935e-06, + "loss": 0.5477, + "step": 11255 + }, + { + "epoch": 3.952247191011236, + "grad_norm": 0.8964136838912964, + "learning_rate": 1.3184003551580304e-06, + "loss": 0.5138, + "step": 11256 + }, + { + "epoch": 3.9525983146067416, + "grad_norm": 0.9326159358024597, + "learning_rate": 1.317992948600063e-06, + "loss": 0.5664, + "step": 11257 + }, + { + "epoch": 3.952949438202247, + "grad_norm": 0.9347344040870667, + "learning_rate": 1.317585582465727e-06, + "loss": 0.5682, + "step": 11258 + }, + { + "epoch": 3.953300561797753, + "grad_norm": 0.9558525681495667, + "learning_rate": 1.3171782567689507e-06, + "loss": 0.566, + "step": 11259 + }, + { + "epoch": 3.9536516853932584, + "grad_norm": 0.9253415465354919, + "learning_rate": 1.316770971523667e-06, + "loss": 0.559, + "step": 11260 + }, + { + "epoch": 3.954002808988764, + "grad_norm": 0.9655329585075378, + "learning_rate": 1.3163637267438026e-06, + "loss": 0.5693, + "step": 11261 + }, + { + "epoch": 3.9543539325842696, + "grad_norm": 0.9636206030845642, + "learning_rate": 1.3159565224432869e-06, + "loss": 0.5545, + "step": 11262 + }, + { + "epoch": 3.954705056179775, + "grad_norm": 0.9305518269538879, + "learning_rate": 1.3155493586360439e-06, + "loss": 0.5661, + "step": 11263 + }, + { + "epoch": 3.955056179775281, + "grad_norm": 0.9323936104774475, + "learning_rate": 1.3151422353360003e-06, + "loss": 0.5314, + "step": 11264 + }, + { + "epoch": 3.9554073033707864, + "grad_norm": 0.9374061226844788, + "learning_rate": 1.3147351525570774e-06, + "loss": 0.5579, + "step": 11265 + }, + { + "epoch": 3.955758426966292, + "grad_norm": 0.9763423204421997, + "learning_rate": 1.3143281103131989e-06, + "loss": 0.5684, + "step": 11266 + }, + { + "epoch": 3.9561095505617976, + "grad_norm": 0.947421133518219, + "learning_rate": 1.3139211086182847e-06, + "loss": 0.5428, + "step": 11267 + }, + { + "epoch": 3.956460674157303, + "grad_norm": 0.9560215473175049, + "learning_rate": 1.3135141474862524e-06, + "loss": 0.5499, + "step": 11268 + }, + { + "epoch": 3.956811797752809, + "grad_norm": 0.9293621778488159, + "learning_rate": 1.313107226931022e-06, + "loss": 0.5596, + "step": 11269 + }, + { + "epoch": 3.9571629213483144, + "grad_norm": 0.93729168176651, + "learning_rate": 1.312700346966508e-06, + "loss": 0.5413, + "step": 11270 + }, + { + "epoch": 3.95751404494382, + "grad_norm": 0.91648268699646, + "learning_rate": 1.3122935076066268e-06, + "loss": 0.525, + "step": 11271 + }, + { + "epoch": 3.9578651685393256, + "grad_norm": 0.9396895170211792, + "learning_rate": 1.3118867088652903e-06, + "loss": 0.5392, + "step": 11272 + }, + { + "epoch": 3.9582162921348316, + "grad_norm": 1.0600090026855469, + "learning_rate": 1.311479950756413e-06, + "loss": 0.5195, + "step": 11273 + }, + { + "epoch": 3.958567415730337, + "grad_norm": 0.9479255676269531, + "learning_rate": 1.311073233293903e-06, + "loss": 0.587, + "step": 11274 + }, + { + "epoch": 3.958918539325843, + "grad_norm": 0.9675742387771606, + "learning_rate": 1.3106665564916724e-06, + "loss": 0.5559, + "step": 11275 + }, + { + "epoch": 3.9592696629213484, + "grad_norm": 0.9563098549842834, + "learning_rate": 1.3102599203636269e-06, + "loss": 0.5614, + "step": 11276 + }, + { + "epoch": 3.959620786516854, + "grad_norm": 0.9340769648551941, + "learning_rate": 1.3098533249236744e-06, + "loss": 0.5378, + "step": 11277 + }, + { + "epoch": 3.9599719101123596, + "grad_norm": 0.9685168862342834, + "learning_rate": 1.3094467701857193e-06, + "loss": 0.5897, + "step": 11278 + }, + { + "epoch": 3.960323033707865, + "grad_norm": 0.9595354795455933, + "learning_rate": 1.3090402561636667e-06, + "loss": 0.576, + "step": 11279 + }, + { + "epoch": 3.960674157303371, + "grad_norm": 1.019031047821045, + "learning_rate": 1.3086337828714186e-06, + "loss": 0.5452, + "step": 11280 + }, + { + "epoch": 3.9610252808988764, + "grad_norm": 0.9618712067604065, + "learning_rate": 1.3082273503228743e-06, + "loss": 0.5853, + "step": 11281 + }, + { + "epoch": 3.961376404494382, + "grad_norm": 0.9459299445152283, + "learning_rate": 1.3078209585319356e-06, + "loss": 0.5638, + "step": 11282 + }, + { + "epoch": 3.9617275280898876, + "grad_norm": 0.9610244631767273, + "learning_rate": 1.3074146075124993e-06, + "loss": 0.589, + "step": 11283 + }, + { + "epoch": 3.962078651685393, + "grad_norm": 0.9495189189910889, + "learning_rate": 1.307008297278464e-06, + "loss": 0.5891, + "step": 11284 + }, + { + "epoch": 3.962429775280899, + "grad_norm": 0.9873005151748657, + "learning_rate": 1.3066020278437231e-06, + "loss": 0.537, + "step": 11285 + }, + { + "epoch": 3.9627808988764044, + "grad_norm": 0.952072262763977, + "learning_rate": 1.3061957992221725e-06, + "loss": 0.5555, + "step": 11286 + }, + { + "epoch": 3.96313202247191, + "grad_norm": 0.9186425805091858, + "learning_rate": 1.3057896114277027e-06, + "loss": 0.5311, + "step": 11287 + }, + { + "epoch": 3.9634831460674156, + "grad_norm": 0.9444823265075684, + "learning_rate": 1.3053834644742076e-06, + "loss": 0.6081, + "step": 11288 + }, + { + "epoch": 3.963834269662921, + "grad_norm": 0.9447528719902039, + "learning_rate": 1.3049773583755749e-06, + "loss": 0.5444, + "step": 11289 + }, + { + "epoch": 3.9641853932584272, + "grad_norm": 0.8909673094749451, + "learning_rate": 1.3045712931456947e-06, + "loss": 0.5119, + "step": 11290 + }, + { + "epoch": 3.964536516853933, + "grad_norm": 0.9647850394248962, + "learning_rate": 1.3041652687984534e-06, + "loss": 0.5434, + "step": 11291 + }, + { + "epoch": 3.9648876404494384, + "grad_norm": 0.9253914952278137, + "learning_rate": 1.3037592853477355e-06, + "loss": 0.5581, + "step": 11292 + }, + { + "epoch": 3.965238764044944, + "grad_norm": 0.955643355846405, + "learning_rate": 1.3033533428074274e-06, + "loss": 0.5642, + "step": 11293 + }, + { + "epoch": 3.9655898876404496, + "grad_norm": 0.9567423462867737, + "learning_rate": 1.3029474411914103e-06, + "loss": 0.5833, + "step": 11294 + }, + { + "epoch": 3.965941011235955, + "grad_norm": 0.933980405330658, + "learning_rate": 1.3025415805135672e-06, + "loss": 0.543, + "step": 11295 + }, + { + "epoch": 3.966292134831461, + "grad_norm": 0.9744176268577576, + "learning_rate": 1.3021357607877765e-06, + "loss": 0.58, + "step": 11296 + }, + { + "epoch": 3.9666432584269664, + "grad_norm": 0.9513632655143738, + "learning_rate": 1.3017299820279183e-06, + "loss": 0.588, + "step": 11297 + }, + { + "epoch": 3.966994382022472, + "grad_norm": 0.9054436087608337, + "learning_rate": 1.3013242442478692e-06, + "loss": 0.5193, + "step": 11298 + }, + { + "epoch": 3.9673455056179776, + "grad_norm": 0.922542154788971, + "learning_rate": 1.3009185474615054e-06, + "loss": 0.545, + "step": 11299 + }, + { + "epoch": 3.967696629213483, + "grad_norm": 0.9639185667037964, + "learning_rate": 1.3005128916827002e-06, + "loss": 0.5819, + "step": 11300 + }, + { + "epoch": 3.968047752808989, + "grad_norm": 0.9758689999580383, + "learning_rate": 1.3001072769253293e-06, + "loss": 0.5622, + "step": 11301 + }, + { + "epoch": 3.9683988764044944, + "grad_norm": 0.9842034578323364, + "learning_rate": 1.2997017032032616e-06, + "loss": 0.5971, + "step": 11302 + }, + { + "epoch": 3.96875, + "grad_norm": 0.9484326243400574, + "learning_rate": 1.2992961705303691e-06, + "loss": 0.591, + "step": 11303 + }, + { + "epoch": 3.9691011235955056, + "grad_norm": 0.9570214748382568, + "learning_rate": 1.2988906789205205e-06, + "loss": 0.5282, + "step": 11304 + }, + { + "epoch": 3.969452247191011, + "grad_norm": 0.9776564836502075, + "learning_rate": 1.2984852283875819e-06, + "loss": 0.5671, + "step": 11305 + }, + { + "epoch": 3.969803370786517, + "grad_norm": 0.9335845708847046, + "learning_rate": 1.298079818945421e-06, + "loss": 0.5458, + "step": 11306 + }, + { + "epoch": 3.9701544943820224, + "grad_norm": 0.9312028884887695, + "learning_rate": 1.297674450607901e-06, + "loss": 0.5515, + "step": 11307 + }, + { + "epoch": 3.970505617977528, + "grad_norm": 0.930756151676178, + "learning_rate": 1.297269123388887e-06, + "loss": 0.5242, + "step": 11308 + }, + { + "epoch": 3.9708567415730336, + "grad_norm": 0.9269701838493347, + "learning_rate": 1.2968638373022387e-06, + "loss": 0.5664, + "step": 11309 + }, + { + "epoch": 3.971207865168539, + "grad_norm": 0.9576120376586914, + "learning_rate": 1.296458592361819e-06, + "loss": 0.5551, + "step": 11310 + }, + { + "epoch": 3.971558988764045, + "grad_norm": 0.9569449424743652, + "learning_rate": 1.2960533885814842e-06, + "loss": 0.5552, + "step": 11311 + }, + { + "epoch": 3.9719101123595504, + "grad_norm": 0.9599092602729797, + "learning_rate": 1.295648225975094e-06, + "loss": 0.5359, + "step": 11312 + }, + { + "epoch": 3.972261235955056, + "grad_norm": 0.9304606318473816, + "learning_rate": 1.2952431045565034e-06, + "loss": 0.5601, + "step": 11313 + }, + { + "epoch": 3.9726123595505616, + "grad_norm": 0.9371951222419739, + "learning_rate": 1.2948380243395683e-06, + "loss": 0.5606, + "step": 11314 + }, + { + "epoch": 3.972963483146067, + "grad_norm": 0.9579379558563232, + "learning_rate": 1.2944329853381408e-06, + "loss": 0.5747, + "step": 11315 + }, + { + "epoch": 3.9733146067415728, + "grad_norm": 0.923941969871521, + "learning_rate": 1.2940279875660745e-06, + "loss": 0.5523, + "step": 11316 + }, + { + "epoch": 3.973665730337079, + "grad_norm": 0.9482721090316772, + "learning_rate": 1.2936230310372188e-06, + "loss": 0.553, + "step": 11317 + }, + { + "epoch": 3.9740168539325844, + "grad_norm": 0.9275593757629395, + "learning_rate": 1.2932181157654225e-06, + "loss": 0.5445, + "step": 11318 + }, + { + "epoch": 3.97436797752809, + "grad_norm": 0.9414995312690735, + "learning_rate": 1.2928132417645351e-06, + "loss": 0.5653, + "step": 11319 + }, + { + "epoch": 3.9747191011235956, + "grad_norm": 0.9395621418952942, + "learning_rate": 1.2924084090484001e-06, + "loss": 0.5368, + "step": 11320 + }, + { + "epoch": 3.975070224719101, + "grad_norm": 0.917155921459198, + "learning_rate": 1.2920036176308659e-06, + "loss": 0.5107, + "step": 11321 + }, + { + "epoch": 3.975421348314607, + "grad_norm": 0.9555845260620117, + "learning_rate": 1.291598867525773e-06, + "loss": 0.5497, + "step": 11322 + }, + { + "epoch": 3.9757724719101124, + "grad_norm": 0.9474613070487976, + "learning_rate": 1.2911941587469655e-06, + "loss": 0.5907, + "step": 11323 + }, + { + "epoch": 3.976123595505618, + "grad_norm": 0.9366769194602966, + "learning_rate": 1.290789491308283e-06, + "loss": 0.5591, + "step": 11324 + }, + { + "epoch": 3.9764747191011236, + "grad_norm": 0.915431559085846, + "learning_rate": 1.2903848652235657e-06, + "loss": 0.5432, + "step": 11325 + }, + { + "epoch": 3.976825842696629, + "grad_norm": 0.9518815279006958, + "learning_rate": 1.2899802805066502e-06, + "loss": 0.556, + "step": 11326 + }, + { + "epoch": 3.977176966292135, + "grad_norm": 0.9109532833099365, + "learning_rate": 1.2895757371713745e-06, + "loss": 0.525, + "step": 11327 + }, + { + "epoch": 3.9775280898876404, + "grad_norm": 0.9725996255874634, + "learning_rate": 1.2891712352315718e-06, + "loss": 0.5911, + "step": 11328 + }, + { + "epoch": 3.977879213483146, + "grad_norm": 0.9072994589805603, + "learning_rate": 1.288766774701078e-06, + "loss": 0.5388, + "step": 11329 + }, + { + "epoch": 3.9782303370786516, + "grad_norm": 0.9454349875450134, + "learning_rate": 1.2883623555937237e-06, + "loss": 0.5656, + "step": 11330 + }, + { + "epoch": 3.978581460674157, + "grad_norm": 0.9769485592842102, + "learning_rate": 1.2879579779233393e-06, + "loss": 0.598, + "step": 11331 + }, + { + "epoch": 3.978932584269663, + "grad_norm": 0.9163880348205566, + "learning_rate": 1.2875536417037555e-06, + "loss": 0.5333, + "step": 11332 + }, + { + "epoch": 3.9792837078651684, + "grad_norm": 0.9899414777755737, + "learning_rate": 1.2871493469487989e-06, + "loss": 0.5435, + "step": 11333 + }, + { + "epoch": 3.9796348314606744, + "grad_norm": 0.9384893178939819, + "learning_rate": 1.286745093672298e-06, + "loss": 0.5311, + "step": 11334 + }, + { + "epoch": 3.97998595505618, + "grad_norm": 0.9648903012275696, + "learning_rate": 1.2863408818880753e-06, + "loss": 0.5688, + "step": 11335 + }, + { + "epoch": 3.9803370786516856, + "grad_norm": 0.9212486743927002, + "learning_rate": 1.2859367116099572e-06, + "loss": 0.5082, + "step": 11336 + }, + { + "epoch": 3.980688202247191, + "grad_norm": 0.9927660226821899, + "learning_rate": 1.2855325828517635e-06, + "loss": 0.5628, + "step": 11337 + }, + { + "epoch": 3.981039325842697, + "grad_norm": 0.9485447406768799, + "learning_rate": 1.2851284956273174e-06, + "loss": 0.551, + "step": 11338 + }, + { + "epoch": 3.9813904494382024, + "grad_norm": 0.9682270288467407, + "learning_rate": 1.2847244499504358e-06, + "loss": 0.5694, + "step": 11339 + }, + { + "epoch": 3.981741573033708, + "grad_norm": 0.9451234340667725, + "learning_rate": 1.2843204458349395e-06, + "loss": 0.5643, + "step": 11340 + }, + { + "epoch": 3.9820926966292136, + "grad_norm": 0.9740580916404724, + "learning_rate": 1.2839164832946426e-06, + "loss": 0.5423, + "step": 11341 + }, + { + "epoch": 3.982443820224719, + "grad_norm": 0.9878503084182739, + "learning_rate": 1.2835125623433619e-06, + "loss": 0.5439, + "step": 11342 + }, + { + "epoch": 3.982794943820225, + "grad_norm": 0.9304189682006836, + "learning_rate": 1.2831086829949113e-06, + "loss": 0.5558, + "step": 11343 + }, + { + "epoch": 3.9831460674157304, + "grad_norm": 0.9342523813247681, + "learning_rate": 1.282704845263101e-06, + "loss": 0.5418, + "step": 11344 + }, + { + "epoch": 3.983497191011236, + "grad_norm": 0.926432192325592, + "learning_rate": 1.2823010491617444e-06, + "loss": 0.5267, + "step": 11345 + }, + { + "epoch": 3.9838483146067416, + "grad_norm": 0.9485119581222534, + "learning_rate": 1.281897294704649e-06, + "loss": 0.5995, + "step": 11346 + }, + { + "epoch": 3.984199438202247, + "grad_norm": 0.9295848608016968, + "learning_rate": 1.2814935819056248e-06, + "loss": 0.5372, + "step": 11347 + }, + { + "epoch": 3.984550561797753, + "grad_norm": 0.9080570936203003, + "learning_rate": 1.2810899107784764e-06, + "loss": 0.5397, + "step": 11348 + }, + { + "epoch": 3.9849016853932584, + "grad_norm": 0.9275252223014832, + "learning_rate": 1.2806862813370109e-06, + "loss": 0.5224, + "step": 11349 + }, + { + "epoch": 3.985252808988764, + "grad_norm": 0.9995993375778198, + "learning_rate": 1.2802826935950301e-06, + "loss": 0.5694, + "step": 11350 + }, + { + "epoch": 3.9856039325842696, + "grad_norm": 0.9535873532295227, + "learning_rate": 1.2798791475663388e-06, + "loss": 0.5663, + "step": 11351 + }, + { + "epoch": 3.985955056179775, + "grad_norm": 0.8975780606269836, + "learning_rate": 1.2794756432647353e-06, + "loss": 0.5135, + "step": 11352 + }, + { + "epoch": 3.986306179775281, + "grad_norm": 0.9328894019126892, + "learning_rate": 1.2790721807040216e-06, + "loss": 0.5483, + "step": 11353 + }, + { + "epoch": 3.9866573033707864, + "grad_norm": 1.1398292779922485, + "learning_rate": 1.2786687598979945e-06, + "loss": 0.5537, + "step": 11354 + }, + { + "epoch": 3.987008426966292, + "grad_norm": 0.9581092596054077, + "learning_rate": 1.2782653808604494e-06, + "loss": 0.5528, + "step": 11355 + }, + { + "epoch": 3.9873595505617976, + "grad_norm": 0.9479995369911194, + "learning_rate": 1.2778620436051842e-06, + "loss": 0.579, + "step": 11356 + }, + { + "epoch": 3.987710674157303, + "grad_norm": 0.9634050130844116, + "learning_rate": 1.2774587481459903e-06, + "loss": 0.541, + "step": 11357 + }, + { + "epoch": 3.988061797752809, + "grad_norm": 0.9477751851081848, + "learning_rate": 1.2770554944966617e-06, + "loss": 0.5408, + "step": 11358 + }, + { + "epoch": 3.9884129213483144, + "grad_norm": 0.9325323104858398, + "learning_rate": 1.2766522826709883e-06, + "loss": 0.5205, + "step": 11359 + }, + { + "epoch": 3.98876404494382, + "grad_norm": 0.9136067628860474, + "learning_rate": 1.2762491126827602e-06, + "loss": 0.5382, + "step": 11360 + }, + { + "epoch": 3.9891151685393256, + "grad_norm": 0.9446619749069214, + "learning_rate": 1.275845984545765e-06, + "loss": 0.5737, + "step": 11361 + }, + { + "epoch": 3.9894662921348316, + "grad_norm": 0.9424066543579102, + "learning_rate": 1.2754428982737903e-06, + "loss": 0.5487, + "step": 11362 + }, + { + "epoch": 3.989817415730337, + "grad_norm": 0.9592906832695007, + "learning_rate": 1.2750398538806196e-06, + "loss": 0.5747, + "step": 11363 + }, + { + "epoch": 3.990168539325843, + "grad_norm": 0.9276706576347351, + "learning_rate": 1.2746368513800387e-06, + "loss": 0.5433, + "step": 11364 + }, + { + "epoch": 3.9905196629213484, + "grad_norm": 0.9547109007835388, + "learning_rate": 1.274233890785828e-06, + "loss": 0.5575, + "step": 11365 + }, + { + "epoch": 3.990870786516854, + "grad_norm": 0.9405946135520935, + "learning_rate": 1.2738309721117703e-06, + "loss": 0.5638, + "step": 11366 + }, + { + "epoch": 3.9912219101123596, + "grad_norm": 0.9517446756362915, + "learning_rate": 1.2734280953716439e-06, + "loss": 0.5274, + "step": 11367 + }, + { + "epoch": 3.991573033707865, + "grad_norm": 0.9440069198608398, + "learning_rate": 1.2730252605792262e-06, + "loss": 0.5583, + "step": 11368 + }, + { + "epoch": 3.991924157303371, + "grad_norm": 0.9279712438583374, + "learning_rate": 1.2726224677482958e-06, + "loss": 0.5616, + "step": 11369 + }, + { + "epoch": 3.9922752808988764, + "grad_norm": 0.9771681427955627, + "learning_rate": 1.2722197168926254e-06, + "loss": 0.5885, + "step": 11370 + }, + { + "epoch": 3.992626404494382, + "grad_norm": 0.9663728475570679, + "learning_rate": 1.2718170080259911e-06, + "loss": 0.5415, + "step": 11371 + }, + { + "epoch": 3.9929775280898876, + "grad_norm": 0.9439047574996948, + "learning_rate": 1.2714143411621635e-06, + "loss": 0.5521, + "step": 11372 + }, + { + "epoch": 3.993328651685393, + "grad_norm": 0.9855775237083435, + "learning_rate": 1.271011716314915e-06, + "loss": 0.5819, + "step": 11373 + }, + { + "epoch": 3.993679775280899, + "grad_norm": 0.9418023228645325, + "learning_rate": 1.270609133498013e-06, + "loss": 0.5275, + "step": 11374 + }, + { + "epoch": 3.9940308988764044, + "grad_norm": 0.9334932565689087, + "learning_rate": 1.2702065927252278e-06, + "loss": 0.5124, + "step": 11375 + }, + { + "epoch": 3.99438202247191, + "grad_norm": 0.9609782695770264, + "learning_rate": 1.2698040940103233e-06, + "loss": 0.5235, + "step": 11376 + }, + { + "epoch": 3.9947331460674156, + "grad_norm": 0.9695614576339722, + "learning_rate": 1.2694016373670676e-06, + "loss": 0.5701, + "step": 11377 + }, + { + "epoch": 3.995084269662921, + "grad_norm": 0.9282645583152771, + "learning_rate": 1.2689992228092218e-06, + "loss": 0.5597, + "step": 11378 + }, + { + "epoch": 3.9954353932584272, + "grad_norm": 0.919155478477478, + "learning_rate": 1.2685968503505503e-06, + "loss": 0.5259, + "step": 11379 + }, + { + "epoch": 3.995786516853933, + "grad_norm": 0.9760883450508118, + "learning_rate": 1.2681945200048124e-06, + "loss": 0.5605, + "step": 11380 + }, + { + "epoch": 3.9961376404494384, + "grad_norm": 0.9728698134422302, + "learning_rate": 1.267792231785767e-06, + "loss": 0.581, + "step": 11381 + }, + { + "epoch": 3.996488764044944, + "grad_norm": 0.9472363591194153, + "learning_rate": 1.267389985707174e-06, + "loss": 0.5604, + "step": 11382 + }, + { + "epoch": 3.9968398876404496, + "grad_norm": 0.9091648459434509, + "learning_rate": 1.2669877817827876e-06, + "loss": 0.5339, + "step": 11383 + }, + { + "epoch": 3.997191011235955, + "grad_norm": 0.9580212235450745, + "learning_rate": 1.266585620026365e-06, + "loss": 0.5237, + "step": 11384 + }, + { + "epoch": 3.997542134831461, + "grad_norm": 0.9276888966560364, + "learning_rate": 1.2661835004516576e-06, + "loss": 0.5252, + "step": 11385 + }, + { + "epoch": 3.9978932584269664, + "grad_norm": 0.9348134398460388, + "learning_rate": 1.26578142307242e-06, + "loss": 0.5574, + "step": 11386 + }, + { + "epoch": 3.998244382022472, + "grad_norm": 0.9321048855781555, + "learning_rate": 1.2653793879024012e-06, + "loss": 0.5082, + "step": 11387 + }, + { + "epoch": 3.9985955056179776, + "grad_norm": 0.9473230242729187, + "learning_rate": 1.2649773949553502e-06, + "loss": 0.5552, + "step": 11388 + }, + { + "epoch": 3.998946629213483, + "grad_norm": 0.9441924691200256, + "learning_rate": 1.2645754442450164e-06, + "loss": 0.5407, + "step": 11389 + }, + { + "epoch": 3.999297752808989, + "grad_norm": 0.953187108039856, + "learning_rate": 1.2641735357851442e-06, + "loss": 0.5432, + "step": 11390 + }, + { + "epoch": 3.9996488764044944, + "grad_norm": 0.9304288625717163, + "learning_rate": 1.2637716695894806e-06, + "loss": 0.5347, + "step": 11391 + }, + { + "epoch": 4.0, + "grad_norm": 0.9777024984359741, + "learning_rate": 1.2633698456717668e-06, + "loss": 0.5185, + "step": 11392 + }, + { + "epoch": 4.000351123595506, + "grad_norm": 0.918589174747467, + "learning_rate": 1.2629680640457476e-06, + "loss": 0.5194, + "step": 11393 + }, + { + "epoch": 4.000702247191011, + "grad_norm": 0.9451001286506653, + "learning_rate": 1.2625663247251617e-06, + "loss": 0.5497, + "step": 11394 + }, + { + "epoch": 4.001053370786517, + "grad_norm": 0.9481091499328613, + "learning_rate": 1.2621646277237476e-06, + "loss": 0.5526, + "step": 11395 + }, + { + "epoch": 4.001404494382022, + "grad_norm": 0.9260170459747314, + "learning_rate": 1.261762973055245e-06, + "loss": 0.568, + "step": 11396 + }, + { + "epoch": 4.001755617977528, + "grad_norm": 0.9369192123413086, + "learning_rate": 1.2613613607333884e-06, + "loss": 0.5756, + "step": 11397 + }, + { + "epoch": 4.002106741573034, + "grad_norm": 0.9233396053314209, + "learning_rate": 1.260959790771914e-06, + "loss": 0.5627, + "step": 11398 + }, + { + "epoch": 4.002457865168539, + "grad_norm": 0.9301872253417969, + "learning_rate": 1.2605582631845536e-06, + "loss": 0.5425, + "step": 11399 + }, + { + "epoch": 4.002808988764045, + "grad_norm": 0.9441287517547607, + "learning_rate": 1.260156777985041e-06, + "loss": 0.5381, + "step": 11400 + }, + { + "epoch": 4.00316011235955, + "grad_norm": 0.9353060722351074, + "learning_rate": 1.2597553351871048e-06, + "loss": 0.5147, + "step": 11401 + }, + { + "epoch": 4.003511235955056, + "grad_norm": 0.9645448923110962, + "learning_rate": 1.2593539348044756e-06, + "loss": 0.5263, + "step": 11402 + }, + { + "epoch": 4.003862359550562, + "grad_norm": 0.9203650951385498, + "learning_rate": 1.2589525768508804e-06, + "loss": 0.5442, + "step": 11403 + }, + { + "epoch": 4.004213483146067, + "grad_norm": 0.931452751159668, + "learning_rate": 1.2585512613400442e-06, + "loss": 0.5162, + "step": 11404 + }, + { + "epoch": 4.004564606741573, + "grad_norm": 0.9072222113609314, + "learning_rate": 1.258149988285693e-06, + "loss": 0.5129, + "step": 11405 + }, + { + "epoch": 4.004915730337078, + "grad_norm": 0.9858955144882202, + "learning_rate": 1.257748757701549e-06, + "loss": 0.5608, + "step": 11406 + }, + { + "epoch": 4.005266853932584, + "grad_norm": 0.9319295287132263, + "learning_rate": 1.2573475696013355e-06, + "loss": 0.5124, + "step": 11407 + }, + { + "epoch": 4.00561797752809, + "grad_norm": 0.9738900065422058, + "learning_rate": 1.2569464239987706e-06, + "loss": 0.5496, + "step": 11408 + }, + { + "epoch": 4.005969101123595, + "grad_norm": 0.949315071105957, + "learning_rate": 1.256545320907575e-06, + "loss": 0.5104, + "step": 11409 + }, + { + "epoch": 4.006320224719101, + "grad_norm": 0.9521793127059937, + "learning_rate": 1.2561442603414658e-06, + "loss": 0.5282, + "step": 11410 + }, + { + "epoch": 4.006671348314606, + "grad_norm": 0.9387788772583008, + "learning_rate": 1.2557432423141574e-06, + "loss": 0.5431, + "step": 11411 + }, + { + "epoch": 4.007022471910112, + "grad_norm": 0.9636697173118591, + "learning_rate": 1.2553422668393667e-06, + "loss": 0.5888, + "step": 11412 + }, + { + "epoch": 4.007373595505618, + "grad_norm": 0.9593473672866821, + "learning_rate": 1.2549413339308038e-06, + "loss": 0.5455, + "step": 11413 + }, + { + "epoch": 4.007724719101123, + "grad_norm": 0.9786216616630554, + "learning_rate": 1.2545404436021835e-06, + "loss": 0.5527, + "step": 11414 + }, + { + "epoch": 4.00807584269663, + "grad_norm": 0.9433626532554626, + "learning_rate": 1.2541395958672128e-06, + "loss": 0.5657, + "step": 11415 + }, + { + "epoch": 4.008426966292135, + "grad_norm": 0.9675050973892212, + "learning_rate": 1.2537387907396032e-06, + "loss": 0.5397, + "step": 11416 + }, + { + "epoch": 4.008778089887641, + "grad_norm": 0.9749513864517212, + "learning_rate": 1.25333802823306e-06, + "loss": 0.5531, + "step": 11417 + }, + { + "epoch": 4.009129213483146, + "grad_norm": 0.9174317717552185, + "learning_rate": 1.252937308361289e-06, + "loss": 0.5094, + "step": 11418 + }, + { + "epoch": 4.009480337078652, + "grad_norm": 0.9586646556854248, + "learning_rate": 1.2525366311379955e-06, + "loss": 0.5431, + "step": 11419 + }, + { + "epoch": 4.009831460674158, + "grad_norm": 0.9000153541564941, + "learning_rate": 1.2521359965768809e-06, + "loss": 0.4878, + "step": 11420 + }, + { + "epoch": 4.010182584269663, + "grad_norm": 0.9353187680244446, + "learning_rate": 1.2517354046916485e-06, + "loss": 0.558, + "step": 11421 + }, + { + "epoch": 4.010533707865169, + "grad_norm": 0.9299389719963074, + "learning_rate": 1.251334855495996e-06, + "loss": 0.566, + "step": 11422 + }, + { + "epoch": 4.010884831460674, + "grad_norm": 0.9069304466247559, + "learning_rate": 1.2509343490036242e-06, + "loss": 0.5062, + "step": 11423 + }, + { + "epoch": 4.01123595505618, + "grad_norm": 0.9658790826797485, + "learning_rate": 1.2505338852282272e-06, + "loss": 0.5855, + "step": 11424 + }, + { + "epoch": 4.011587078651686, + "grad_norm": 0.970927894115448, + "learning_rate": 1.2501334641835037e-06, + "loss": 0.5365, + "step": 11425 + }, + { + "epoch": 4.011938202247191, + "grad_norm": 0.9241825938224792, + "learning_rate": 1.249733085883145e-06, + "loss": 0.5375, + "step": 11426 + }, + { + "epoch": 4.012289325842697, + "grad_norm": 0.9171820878982544, + "learning_rate": 1.249332750340846e-06, + "loss": 0.5395, + "step": 11427 + }, + { + "epoch": 4.012640449438202, + "grad_norm": 0.9379998445510864, + "learning_rate": 1.2489324575702958e-06, + "loss": 0.5271, + "step": 11428 + }, + { + "epoch": 4.012991573033708, + "grad_norm": 0.9273099899291992, + "learning_rate": 1.2485322075851859e-06, + "loss": 0.5747, + "step": 11429 + }, + { + "epoch": 4.013342696629214, + "grad_norm": 0.9754973649978638, + "learning_rate": 1.2481320003992034e-06, + "loss": 0.56, + "step": 11430 + }, + { + "epoch": 4.013693820224719, + "grad_norm": 0.9438334703445435, + "learning_rate": 1.2477318360260344e-06, + "loss": 0.5319, + "step": 11431 + }, + { + "epoch": 4.014044943820225, + "grad_norm": 0.9363061785697937, + "learning_rate": 1.247331714479366e-06, + "loss": 0.5064, + "step": 11432 + }, + { + "epoch": 4.01439606741573, + "grad_norm": 0.9263989925384521, + "learning_rate": 1.24693163577288e-06, + "loss": 0.5069, + "step": 11433 + }, + { + "epoch": 4.014747191011236, + "grad_norm": 0.9806944131851196, + "learning_rate": 1.2465315999202609e-06, + "loss": 0.5464, + "step": 11434 + }, + { + "epoch": 4.015098314606742, + "grad_norm": 0.932845950126648, + "learning_rate": 1.2461316069351876e-06, + "loss": 0.5419, + "step": 11435 + }, + { + "epoch": 4.015449438202247, + "grad_norm": 0.9551830291748047, + "learning_rate": 1.245731656831341e-06, + "loss": 0.5392, + "step": 11436 + }, + { + "epoch": 4.015800561797753, + "grad_norm": 0.9256681799888611, + "learning_rate": 1.2453317496223977e-06, + "loss": 0.5287, + "step": 11437 + }, + { + "epoch": 4.016151685393258, + "grad_norm": 0.9563596844673157, + "learning_rate": 1.244931885322036e-06, + "loss": 0.5632, + "step": 11438 + }, + { + "epoch": 4.016502808988764, + "grad_norm": 0.9646201729774475, + "learning_rate": 1.2445320639439287e-06, + "loss": 0.5696, + "step": 11439 + }, + { + "epoch": 4.01685393258427, + "grad_norm": 0.9635419249534607, + "learning_rate": 1.2441322855017516e-06, + "loss": 0.5421, + "step": 11440 + }, + { + "epoch": 4.017205056179775, + "grad_norm": 0.9631474614143372, + "learning_rate": 1.2437325500091746e-06, + "loss": 0.5312, + "step": 11441 + }, + { + "epoch": 4.017556179775281, + "grad_norm": 0.9501041173934937, + "learning_rate": 1.2433328574798703e-06, + "loss": 0.6049, + "step": 11442 + }, + { + "epoch": 4.017907303370786, + "grad_norm": 0.9379426836967468, + "learning_rate": 1.2429332079275073e-06, + "loss": 0.5138, + "step": 11443 + }, + { + "epoch": 4.018258426966292, + "grad_norm": 0.9597308039665222, + "learning_rate": 1.2425336013657516e-06, + "loss": 0.5387, + "step": 11444 + }, + { + "epoch": 4.018609550561798, + "grad_norm": 0.946087121963501, + "learning_rate": 1.2421340378082717e-06, + "loss": 0.5673, + "step": 11445 + }, + { + "epoch": 4.018960674157303, + "grad_norm": 0.9495368599891663, + "learning_rate": 1.2417345172687304e-06, + "loss": 0.5319, + "step": 11446 + }, + { + "epoch": 4.019311797752809, + "grad_norm": 0.9712097644805908, + "learning_rate": 1.2413350397607925e-06, + "loss": 0.5774, + "step": 11447 + }, + { + "epoch": 4.019662921348314, + "grad_norm": 0.9512820839881897, + "learning_rate": 1.240935605298119e-06, + "loss": 0.5468, + "step": 11448 + }, + { + "epoch": 4.02001404494382, + "grad_norm": 0.9344589114189148, + "learning_rate": 1.2405362138943709e-06, + "loss": 0.5385, + "step": 11449 + }, + { + "epoch": 4.020365168539326, + "grad_norm": 0.9749465584754944, + "learning_rate": 1.2401368655632053e-06, + "loss": 0.5358, + "step": 11450 + }, + { + "epoch": 4.020716292134831, + "grad_norm": 0.9482048749923706, + "learning_rate": 1.2397375603182824e-06, + "loss": 0.5233, + "step": 11451 + }, + { + "epoch": 4.021067415730337, + "grad_norm": 0.9020488858222961, + "learning_rate": 1.2393382981732552e-06, + "loss": 0.5622, + "step": 11452 + }, + { + "epoch": 4.021418539325842, + "grad_norm": 0.9608054161071777, + "learning_rate": 1.2389390791417805e-06, + "loss": 0.5439, + "step": 11453 + }, + { + "epoch": 4.021769662921348, + "grad_norm": 0.9356751441955566, + "learning_rate": 1.2385399032375092e-06, + "loss": 0.5193, + "step": 11454 + }, + { + "epoch": 4.022120786516854, + "grad_norm": 0.9567332863807678, + "learning_rate": 1.2381407704740946e-06, + "loss": 0.5494, + "step": 11455 + }, + { + "epoch": 4.022471910112359, + "grad_norm": 0.9469530582427979, + "learning_rate": 1.237741680865186e-06, + "loss": 0.5632, + "step": 11456 + }, + { + "epoch": 4.022823033707865, + "grad_norm": 0.9731237292289734, + "learning_rate": 1.2373426344244307e-06, + "loss": 0.554, + "step": 11457 + }, + { + "epoch": 4.02317415730337, + "grad_norm": 0.9379258155822754, + "learning_rate": 1.2369436311654778e-06, + "loss": 0.5626, + "step": 11458 + }, + { + "epoch": 4.023525280898877, + "grad_norm": 0.9364827275276184, + "learning_rate": 1.2365446711019712e-06, + "loss": 0.5386, + "step": 11459 + }, + { + "epoch": 4.023876404494382, + "grad_norm": 0.9513711333274841, + "learning_rate": 1.2361457542475562e-06, + "loss": 0.5278, + "step": 11460 + }, + { + "epoch": 4.024227528089888, + "grad_norm": 0.9745139479637146, + "learning_rate": 1.2357468806158746e-06, + "loss": 0.5371, + "step": 11461 + }, + { + "epoch": 4.024578651685394, + "grad_norm": 0.9504488110542297, + "learning_rate": 1.2353480502205683e-06, + "loss": 0.5284, + "step": 11462 + }, + { + "epoch": 4.024929775280899, + "grad_norm": 0.9485361576080322, + "learning_rate": 1.2349492630752757e-06, + "loss": 0.5358, + "step": 11463 + }, + { + "epoch": 4.025280898876405, + "grad_norm": 0.9206878542900085, + "learning_rate": 1.2345505191936366e-06, + "loss": 0.5199, + "step": 11464 + }, + { + "epoch": 4.02563202247191, + "grad_norm": 0.9252239465713501, + "learning_rate": 1.2341518185892862e-06, + "loss": 0.5097, + "step": 11465 + }, + { + "epoch": 4.025983146067416, + "grad_norm": 0.9485441446304321, + "learning_rate": 1.233753161275861e-06, + "loss": 0.5459, + "step": 11466 + }, + { + "epoch": 4.026334269662922, + "grad_norm": 0.9195905923843384, + "learning_rate": 1.2333545472669936e-06, + "loss": 0.4928, + "step": 11467 + }, + { + "epoch": 4.026685393258427, + "grad_norm": 0.9596986770629883, + "learning_rate": 1.2329559765763175e-06, + "loss": 0.5525, + "step": 11468 + }, + { + "epoch": 4.027036516853933, + "grad_norm": 0.9260126352310181, + "learning_rate": 1.2325574492174628e-06, + "loss": 0.5455, + "step": 11469 + }, + { + "epoch": 4.027387640449438, + "grad_norm": 0.9372918009757996, + "learning_rate": 1.2321589652040578e-06, + "loss": 0.5388, + "step": 11470 + }, + { + "epoch": 4.027738764044944, + "grad_norm": 0.958169162273407, + "learning_rate": 1.2317605245497324e-06, + "loss": 0.535, + "step": 11471 + }, + { + "epoch": 4.02808988764045, + "grad_norm": 0.927825391292572, + "learning_rate": 1.2313621272681106e-06, + "loss": 0.5198, + "step": 11472 + }, + { + "epoch": 4.028441011235955, + "grad_norm": 0.9355791211128235, + "learning_rate": 1.2309637733728198e-06, + "loss": 0.5312, + "step": 11473 + }, + { + "epoch": 4.028792134831461, + "grad_norm": 0.9535204768180847, + "learning_rate": 1.2305654628774808e-06, + "loss": 0.5375, + "step": 11474 + }, + { + "epoch": 4.029143258426966, + "grad_norm": 0.9507874846458435, + "learning_rate": 1.2301671957957176e-06, + "loss": 0.5177, + "step": 11475 + }, + { + "epoch": 4.029494382022472, + "grad_norm": 0.9251250624656677, + "learning_rate": 1.2297689721411487e-06, + "loss": 0.5151, + "step": 11476 + }, + { + "epoch": 4.029845505617978, + "grad_norm": 0.9741066098213196, + "learning_rate": 1.229370791927395e-06, + "loss": 0.5239, + "step": 11477 + }, + { + "epoch": 4.030196629213483, + "grad_norm": 0.9586778879165649, + "learning_rate": 1.228972655168072e-06, + "loss": 0.5523, + "step": 11478 + }, + { + "epoch": 4.030547752808989, + "grad_norm": 0.9805047512054443, + "learning_rate": 1.2285745618767978e-06, + "loss": 0.5436, + "step": 11479 + }, + { + "epoch": 4.030898876404494, + "grad_norm": 0.9483461380004883, + "learning_rate": 1.228176512067185e-06, + "loss": 0.5255, + "step": 11480 + }, + { + "epoch": 4.03125, + "grad_norm": 0.9352431893348694, + "learning_rate": 1.2277785057528469e-06, + "loss": 0.5437, + "step": 11481 + }, + { + "epoch": 4.031601123595506, + "grad_norm": 0.9580222368240356, + "learning_rate": 1.2273805429473956e-06, + "loss": 0.5237, + "step": 11482 + }, + { + "epoch": 4.031952247191011, + "grad_norm": 0.9628734588623047, + "learning_rate": 1.2269826236644402e-06, + "loss": 0.5368, + "step": 11483 + }, + { + "epoch": 4.032303370786517, + "grad_norm": 0.9506659507751465, + "learning_rate": 1.2265847479175905e-06, + "loss": 0.576, + "step": 11484 + }, + { + "epoch": 4.032654494382022, + "grad_norm": 0.9460512399673462, + "learning_rate": 1.2261869157204518e-06, + "loss": 0.5434, + "step": 11485 + }, + { + "epoch": 4.033005617977528, + "grad_norm": 0.9506760239601135, + "learning_rate": 1.2257891270866316e-06, + "loss": 0.5614, + "step": 11486 + }, + { + "epoch": 4.033356741573034, + "grad_norm": 0.9429535865783691, + "learning_rate": 1.225391382029732e-06, + "loss": 0.5237, + "step": 11487 + }, + { + "epoch": 4.033707865168539, + "grad_norm": 0.9538012146949768, + "learning_rate": 1.2249936805633572e-06, + "loss": 0.557, + "step": 11488 + }, + { + "epoch": 4.034058988764045, + "grad_norm": 0.9224615693092346, + "learning_rate": 1.2245960227011066e-06, + "loss": 0.5273, + "step": 11489 + }, + { + "epoch": 4.03441011235955, + "grad_norm": 0.9287775754928589, + "learning_rate": 1.2241984084565817e-06, + "loss": 0.4915, + "step": 11490 + }, + { + "epoch": 4.034761235955056, + "grad_norm": 0.9495412111282349, + "learning_rate": 1.2238008378433782e-06, + "loss": 0.5376, + "step": 11491 + }, + { + "epoch": 4.035112359550562, + "grad_norm": 0.9546013474464417, + "learning_rate": 1.2234033108750953e-06, + "loss": 0.5439, + "step": 11492 + }, + { + "epoch": 4.035463483146067, + "grad_norm": 0.9248926043510437, + "learning_rate": 1.2230058275653268e-06, + "loss": 0.5359, + "step": 11493 + }, + { + "epoch": 4.035814606741573, + "grad_norm": 0.9400137662887573, + "learning_rate": 1.2226083879276652e-06, + "loss": 0.5262, + "step": 11494 + }, + { + "epoch": 4.036165730337078, + "grad_norm": 0.9365279674530029, + "learning_rate": 1.2222109919757043e-06, + "loss": 0.5377, + "step": 11495 + }, + { + "epoch": 4.036516853932584, + "grad_norm": 0.969580352306366, + "learning_rate": 1.2218136397230338e-06, + "loss": 0.5596, + "step": 11496 + }, + { + "epoch": 4.03686797752809, + "grad_norm": 0.9641111493110657, + "learning_rate": 1.2214163311832434e-06, + "loss": 0.5668, + "step": 11497 + }, + { + "epoch": 4.037219101123595, + "grad_norm": 1.0307368040084839, + "learning_rate": 1.2210190663699196e-06, + "loss": 0.5634, + "step": 11498 + }, + { + "epoch": 4.037570224719101, + "grad_norm": 0.9193294048309326, + "learning_rate": 1.2206218452966504e-06, + "loss": 0.5222, + "step": 11499 + }, + { + "epoch": 4.037921348314606, + "grad_norm": 0.9522969722747803, + "learning_rate": 1.220224667977018e-06, + "loss": 0.5491, + "step": 11500 + }, + { + "epoch": 4.038272471910112, + "grad_norm": 0.9578246474266052, + "learning_rate": 1.2198275344246082e-06, + "loss": 0.5312, + "step": 11501 + }, + { + "epoch": 4.038623595505618, + "grad_norm": 0.9320897459983826, + "learning_rate": 1.219430444653e-06, + "loss": 0.5327, + "step": 11502 + }, + { + "epoch": 4.038974719101123, + "grad_norm": 0.9773994088172913, + "learning_rate": 1.219033398675776e-06, + "loss": 0.5565, + "step": 11503 + }, + { + "epoch": 4.03932584269663, + "grad_norm": 0.9341104030609131, + "learning_rate": 1.2186363965065125e-06, + "loss": 0.552, + "step": 11504 + }, + { + "epoch": 4.039676966292135, + "grad_norm": 0.9490652680397034, + "learning_rate": 1.2182394381587886e-06, + "loss": 0.5792, + "step": 11505 + }, + { + "epoch": 4.040028089887641, + "grad_norm": 0.9516528248786926, + "learning_rate": 1.2178425236461794e-06, + "loss": 0.563, + "step": 11506 + }, + { + "epoch": 4.040379213483146, + "grad_norm": 0.9184290766716003, + "learning_rate": 1.2174456529822575e-06, + "loss": 0.5253, + "step": 11507 + }, + { + "epoch": 4.040730337078652, + "grad_norm": 0.9341045618057251, + "learning_rate": 1.2170488261805978e-06, + "loss": 0.5528, + "step": 11508 + }, + { + "epoch": 4.041081460674158, + "grad_norm": 0.9579501152038574, + "learning_rate": 1.2166520432547696e-06, + "loss": 0.5571, + "step": 11509 + }, + { + "epoch": 4.041432584269663, + "grad_norm": 0.9443531036376953, + "learning_rate": 1.2162553042183445e-06, + "loss": 0.5536, + "step": 11510 + }, + { + "epoch": 4.041783707865169, + "grad_norm": 0.934834361076355, + "learning_rate": 1.2158586090848881e-06, + "loss": 0.5647, + "step": 11511 + }, + { + "epoch": 4.042134831460674, + "grad_norm": 0.9636829495429993, + "learning_rate": 1.2154619578679693e-06, + "loss": 0.5332, + "step": 11512 + }, + { + "epoch": 4.04248595505618, + "grad_norm": 0.962615430355072, + "learning_rate": 1.215065350581152e-06, + "loss": 0.5388, + "step": 11513 + }, + { + "epoch": 4.042837078651686, + "grad_norm": 0.9584515690803528, + "learning_rate": 1.2146687872380005e-06, + "loss": 0.5419, + "step": 11514 + }, + { + "epoch": 4.043188202247191, + "grad_norm": 0.9497861266136169, + "learning_rate": 1.214272267852076e-06, + "loss": 0.561, + "step": 11515 + }, + { + "epoch": 4.043539325842697, + "grad_norm": 0.9828190207481384, + "learning_rate": 1.2138757924369405e-06, + "loss": 0.5417, + "step": 11516 + }, + { + "epoch": 4.043890449438202, + "grad_norm": 0.9586576223373413, + "learning_rate": 1.2134793610061515e-06, + "loss": 0.5515, + "step": 11517 + }, + { + "epoch": 4.044241573033708, + "grad_norm": 0.9433528184890747, + "learning_rate": 1.2130829735732683e-06, + "loss": 0.5505, + "step": 11518 + }, + { + "epoch": 4.044592696629214, + "grad_norm": 0.9814038872718811, + "learning_rate": 1.2126866301518462e-06, + "loss": 0.5537, + "step": 11519 + }, + { + "epoch": 4.044943820224719, + "grad_norm": 0.9251599311828613, + "learning_rate": 1.212290330755439e-06, + "loss": 0.5364, + "step": 11520 + }, + { + "epoch": 4.045294943820225, + "grad_norm": 0.9376000761985779, + "learning_rate": 1.2118940753976013e-06, + "loss": 0.5777, + "step": 11521 + }, + { + "epoch": 4.04564606741573, + "grad_norm": 0.9281013011932373, + "learning_rate": 1.2114978640918832e-06, + "loss": 0.5453, + "step": 11522 + }, + { + "epoch": 4.045997191011236, + "grad_norm": 0.962309718132019, + "learning_rate": 1.2111016968518364e-06, + "loss": 0.5499, + "step": 11523 + }, + { + "epoch": 4.046348314606742, + "grad_norm": 1.0313557386398315, + "learning_rate": 1.2107055736910075e-06, + "loss": 0.5631, + "step": 11524 + }, + { + "epoch": 4.046699438202247, + "grad_norm": 0.9589264392852783, + "learning_rate": 1.210309494622946e-06, + "loss": 0.5436, + "step": 11525 + }, + { + "epoch": 4.047050561797753, + "grad_norm": 0.8999641537666321, + "learning_rate": 1.2099134596611953e-06, + "loss": 0.5075, + "step": 11526 + }, + { + "epoch": 4.047401685393258, + "grad_norm": 0.9433310031890869, + "learning_rate": 1.209517468819301e-06, + "loss": 0.5508, + "step": 11527 + }, + { + "epoch": 4.047752808988764, + "grad_norm": 0.8785986304283142, + "learning_rate": 1.2091215221108041e-06, + "loss": 0.4932, + "step": 11528 + }, + { + "epoch": 4.04810393258427, + "grad_norm": 0.9323062896728516, + "learning_rate": 1.2087256195492475e-06, + "loss": 0.5361, + "step": 11529 + }, + { + "epoch": 4.048455056179775, + "grad_norm": 0.9622145295143127, + "learning_rate": 1.208329761148169e-06, + "loss": 0.5465, + "step": 11530 + }, + { + "epoch": 4.048806179775281, + "grad_norm": 0.9528011679649353, + "learning_rate": 1.207933946921108e-06, + "loss": 0.5354, + "step": 11531 + }, + { + "epoch": 4.049157303370786, + "grad_norm": 0.9834375977516174, + "learning_rate": 1.2075381768816004e-06, + "loss": 0.5342, + "step": 11532 + }, + { + "epoch": 4.049508426966292, + "grad_norm": 0.9575339555740356, + "learning_rate": 1.2071424510431802e-06, + "loss": 0.5366, + "step": 11533 + }, + { + "epoch": 4.049859550561798, + "grad_norm": 0.9460598230361938, + "learning_rate": 1.2067467694193833e-06, + "loss": 0.5396, + "step": 11534 + }, + { + "epoch": 4.050210674157303, + "grad_norm": 0.9835683703422546, + "learning_rate": 1.2063511320237387e-06, + "loss": 0.562, + "step": 11535 + }, + { + "epoch": 4.050561797752809, + "grad_norm": 0.9278931021690369, + "learning_rate": 1.2059555388697796e-06, + "loss": 0.5241, + "step": 11536 + }, + { + "epoch": 4.050912921348314, + "grad_norm": 0.9398766160011292, + "learning_rate": 1.205559989971033e-06, + "loss": 0.5237, + "step": 11537 + }, + { + "epoch": 4.05126404494382, + "grad_norm": 0.9329536557197571, + "learning_rate": 1.2051644853410277e-06, + "loss": 0.5255, + "step": 11538 + }, + { + "epoch": 4.051615168539326, + "grad_norm": 0.9887698888778687, + "learning_rate": 1.204769024993288e-06, + "loss": 0.5362, + "step": 11539 + }, + { + "epoch": 4.051966292134831, + "grad_norm": 0.9475218057632446, + "learning_rate": 1.2043736089413406e-06, + "loss": 0.5276, + "step": 11540 + }, + { + "epoch": 4.052317415730337, + "grad_norm": 0.9549334645271301, + "learning_rate": 1.2039782371987059e-06, + "loss": 0.5308, + "step": 11541 + }, + { + "epoch": 4.052668539325842, + "grad_norm": 0.9426090717315674, + "learning_rate": 1.2035829097789073e-06, + "loss": 0.518, + "step": 11542 + }, + { + "epoch": 4.053019662921348, + "grad_norm": 0.9231612682342529, + "learning_rate": 1.203187626695464e-06, + "loss": 0.5159, + "step": 11543 + }, + { + "epoch": 4.053370786516854, + "grad_norm": 0.93110191822052, + "learning_rate": 1.2027923879618933e-06, + "loss": 0.5296, + "step": 11544 + }, + { + "epoch": 4.053721910112359, + "grad_norm": 0.9528570175170898, + "learning_rate": 1.2023971935917137e-06, + "loss": 0.5379, + "step": 11545 + }, + { + "epoch": 4.054073033707865, + "grad_norm": 0.9527499675750732, + "learning_rate": 1.2020020435984387e-06, + "loss": 0.5469, + "step": 11546 + }, + { + "epoch": 4.05442415730337, + "grad_norm": 0.926120400428772, + "learning_rate": 1.2016069379955842e-06, + "loss": 0.5334, + "step": 11547 + }, + { + "epoch": 4.054775280898877, + "grad_norm": 0.9735041260719299, + "learning_rate": 1.2012118767966602e-06, + "loss": 0.5355, + "step": 11548 + }, + { + "epoch": 4.055126404494382, + "grad_norm": 0.9418039321899414, + "learning_rate": 1.2008168600151799e-06, + "loss": 0.5294, + "step": 11549 + }, + { + "epoch": 4.055477528089888, + "grad_norm": 1.0284823179244995, + "learning_rate": 1.20042188766465e-06, + "loss": 0.5535, + "step": 11550 + }, + { + "epoch": 4.055828651685394, + "grad_norm": 0.9630689024925232, + "learning_rate": 1.2000269597585807e-06, + "loss": 0.5822, + "step": 11551 + }, + { + "epoch": 4.056179775280899, + "grad_norm": 0.9347305297851562, + "learning_rate": 1.1996320763104757e-06, + "loss": 0.4988, + "step": 11552 + }, + { + "epoch": 4.056530898876405, + "grad_norm": 0.9590454697608948, + "learning_rate": 1.1992372373338421e-06, + "loss": 0.5223, + "step": 11553 + }, + { + "epoch": 4.05688202247191, + "grad_norm": 0.9930384755134583, + "learning_rate": 1.1988424428421819e-06, + "loss": 0.5287, + "step": 11554 + }, + { + "epoch": 4.057233146067416, + "grad_norm": 0.9485338926315308, + "learning_rate": 1.1984476928489958e-06, + "loss": 0.5396, + "step": 11555 + }, + { + "epoch": 4.057584269662922, + "grad_norm": 0.9684705138206482, + "learning_rate": 1.198052987367786e-06, + "loss": 0.5528, + "step": 11556 + }, + { + "epoch": 4.057935393258427, + "grad_norm": 0.944791316986084, + "learning_rate": 1.1976583264120487e-06, + "loss": 0.5575, + "step": 11557 + }, + { + "epoch": 4.058286516853933, + "grad_norm": 0.9494207501411438, + "learning_rate": 1.1972637099952831e-06, + "loss": 0.6161, + "step": 11558 + }, + { + "epoch": 4.058637640449438, + "grad_norm": 0.9696171283721924, + "learning_rate": 1.196869138130983e-06, + "loss": 0.5511, + "step": 11559 + }, + { + "epoch": 4.058988764044944, + "grad_norm": 0.9740244746208191, + "learning_rate": 1.1964746108326445e-06, + "loss": 0.5745, + "step": 11560 + }, + { + "epoch": 4.05933988764045, + "grad_norm": 0.9560799598693848, + "learning_rate": 1.1960801281137584e-06, + "loss": 0.5321, + "step": 11561 + }, + { + "epoch": 4.059691011235955, + "grad_norm": 0.9440670013427734, + "learning_rate": 1.1956856899878154e-06, + "loss": 0.5405, + "step": 11562 + }, + { + "epoch": 4.060042134831461, + "grad_norm": 0.9540131092071533, + "learning_rate": 1.1952912964683067e-06, + "loss": 0.5708, + "step": 11563 + }, + { + "epoch": 4.060393258426966, + "grad_norm": 0.9601548314094543, + "learning_rate": 1.194896947568718e-06, + "loss": 0.5521, + "step": 11564 + }, + { + "epoch": 4.060744382022472, + "grad_norm": 0.9935348629951477, + "learning_rate": 1.1945026433025383e-06, + "loss": 0.55, + "step": 11565 + }, + { + "epoch": 4.061095505617978, + "grad_norm": 0.9812514781951904, + "learning_rate": 1.1941083836832498e-06, + "loss": 0.5291, + "step": 11566 + }, + { + "epoch": 4.061446629213483, + "grad_norm": 0.918543815612793, + "learning_rate": 1.1937141687243382e-06, + "loss": 0.5452, + "step": 11567 + }, + { + "epoch": 4.061797752808989, + "grad_norm": 0.9658193588256836, + "learning_rate": 1.193319998439284e-06, + "loss": 0.5182, + "step": 11568 + }, + { + "epoch": 4.062148876404494, + "grad_norm": 0.9014549255371094, + "learning_rate": 1.192925872841567e-06, + "loss": 0.4899, + "step": 11569 + }, + { + "epoch": 4.0625, + "grad_norm": 0.93558269739151, + "learning_rate": 1.1925317919446675e-06, + "loss": 0.5009, + "step": 11570 + }, + { + "epoch": 4.062851123595506, + "grad_norm": 0.9773933291435242, + "learning_rate": 1.1921377557620608e-06, + "loss": 0.5479, + "step": 11571 + }, + { + "epoch": 4.063202247191011, + "grad_norm": 0.9528326988220215, + "learning_rate": 1.1917437643072249e-06, + "loss": 0.5754, + "step": 11572 + }, + { + "epoch": 4.063553370786517, + "grad_norm": 0.9714740514755249, + "learning_rate": 1.191349817593632e-06, + "loss": 0.5341, + "step": 11573 + }, + { + "epoch": 4.063904494382022, + "grad_norm": 0.9593794345855713, + "learning_rate": 1.190955915634756e-06, + "loss": 0.535, + "step": 11574 + }, + { + "epoch": 4.064255617977528, + "grad_norm": 0.968397855758667, + "learning_rate": 1.1905620584440669e-06, + "loss": 0.5378, + "step": 11575 + }, + { + "epoch": 4.064606741573034, + "grad_norm": 0.9664372205734253, + "learning_rate": 1.1901682460350355e-06, + "loss": 0.545, + "step": 11576 + }, + { + "epoch": 4.064957865168539, + "grad_norm": 1.2266731262207031, + "learning_rate": 1.1897744784211295e-06, + "loss": 0.546, + "step": 11577 + }, + { + "epoch": 4.065308988764045, + "grad_norm": 1.000572681427002, + "learning_rate": 1.1893807556158143e-06, + "loss": 0.5845, + "step": 11578 + }, + { + "epoch": 4.06566011235955, + "grad_norm": 0.9438109993934631, + "learning_rate": 1.1889870776325566e-06, + "loss": 0.5528, + "step": 11579 + }, + { + "epoch": 4.066011235955056, + "grad_norm": 0.9520018696784973, + "learning_rate": 1.188593444484818e-06, + "loss": 0.5404, + "step": 11580 + }, + { + "epoch": 4.066362359550562, + "grad_norm": 0.9677723050117493, + "learning_rate": 1.1881998561860625e-06, + "loss": 0.5695, + "step": 11581 + }, + { + "epoch": 4.066713483146067, + "grad_norm": 0.9632299542427063, + "learning_rate": 1.1878063127497484e-06, + "loss": 0.5544, + "step": 11582 + }, + { + "epoch": 4.067064606741573, + "grad_norm": 0.9520652890205383, + "learning_rate": 1.1874128141893362e-06, + "loss": 0.5455, + "step": 11583 + }, + { + "epoch": 4.067415730337078, + "grad_norm": 0.9522616863250732, + "learning_rate": 1.187019360518283e-06, + "loss": 0.5515, + "step": 11584 + }, + { + "epoch": 4.067766853932584, + "grad_norm": 0.9497724175453186, + "learning_rate": 1.1866259517500428e-06, + "loss": 0.5549, + "step": 11585 + }, + { + "epoch": 4.06811797752809, + "grad_norm": 0.9503533244132996, + "learning_rate": 1.1862325878980723e-06, + "loss": 0.5261, + "step": 11586 + }, + { + "epoch": 4.068469101123595, + "grad_norm": 0.9508625268936157, + "learning_rate": 1.1858392689758223e-06, + "loss": 0.513, + "step": 11587 + }, + { + "epoch": 4.068820224719101, + "grad_norm": 0.9477367401123047, + "learning_rate": 1.1854459949967453e-06, + "loss": 0.5236, + "step": 11588 + }, + { + "epoch": 4.069171348314606, + "grad_norm": 0.9538856744766235, + "learning_rate": 1.1850527659742897e-06, + "loss": 0.534, + "step": 11589 + }, + { + "epoch": 4.069522471910112, + "grad_norm": 0.9533492922782898, + "learning_rate": 1.1846595819219052e-06, + "loss": 0.5396, + "step": 11590 + }, + { + "epoch": 4.069873595505618, + "grad_norm": 0.9724897742271423, + "learning_rate": 1.1842664428530368e-06, + "loss": 0.5559, + "step": 11591 + }, + { + "epoch": 4.070224719101123, + "grad_norm": 0.9832957983016968, + "learning_rate": 1.1838733487811307e-06, + "loss": 0.5865, + "step": 11592 + }, + { + "epoch": 4.07057584269663, + "grad_norm": 0.9699115753173828, + "learning_rate": 1.183480299719629e-06, + "loss": 0.5443, + "step": 11593 + }, + { + "epoch": 4.070926966292135, + "grad_norm": 0.9802961945533752, + "learning_rate": 1.1830872956819756e-06, + "loss": 0.5499, + "step": 11594 + }, + { + "epoch": 4.071278089887641, + "grad_norm": 0.9894610643386841, + "learning_rate": 1.1826943366816097e-06, + "loss": 0.5226, + "step": 11595 + }, + { + "epoch": 4.071629213483146, + "grad_norm": 0.9413086771965027, + "learning_rate": 1.182301422731969e-06, + "loss": 0.5052, + "step": 11596 + }, + { + "epoch": 4.071980337078652, + "grad_norm": 0.9656396508216858, + "learning_rate": 1.1819085538464932e-06, + "loss": 0.5729, + "step": 11597 + }, + { + "epoch": 4.072331460674158, + "grad_norm": 0.9529977440834045, + "learning_rate": 1.1815157300386162e-06, + "loss": 0.5279, + "step": 11598 + }, + { + "epoch": 4.072682584269663, + "grad_norm": 0.9662032723426819, + "learning_rate": 1.181122951321774e-06, + "loss": 0.5526, + "step": 11599 + }, + { + "epoch": 4.073033707865169, + "grad_norm": 0.9562380313873291, + "learning_rate": 1.180730217709397e-06, + "loss": 0.5585, + "step": 11600 + }, + { + "epoch": 4.073384831460674, + "grad_norm": 0.9533992409706116, + "learning_rate": 1.1803375292149188e-06, + "loss": 0.5384, + "step": 11601 + }, + { + "epoch": 4.07373595505618, + "grad_norm": 0.9326551556587219, + "learning_rate": 1.1799448858517668e-06, + "loss": 0.5199, + "step": 11602 + }, + { + "epoch": 4.074087078651686, + "grad_norm": 0.9431266784667969, + "learning_rate": 1.1795522876333711e-06, + "loss": 0.5294, + "step": 11603 + }, + { + "epoch": 4.074438202247191, + "grad_norm": 0.9417838454246521, + "learning_rate": 1.1791597345731562e-06, + "loss": 0.5224, + "step": 11604 + }, + { + "epoch": 4.074789325842697, + "grad_norm": 0.9470933079719543, + "learning_rate": 1.1787672266845493e-06, + "loss": 0.5174, + "step": 11605 + }, + { + "epoch": 4.075140449438202, + "grad_norm": 0.9433825016021729, + "learning_rate": 1.1783747639809726e-06, + "loss": 0.5231, + "step": 11606 + }, + { + "epoch": 4.075491573033708, + "grad_norm": 0.9530733227729797, + "learning_rate": 1.177982346475847e-06, + "loss": 0.5262, + "step": 11607 + }, + { + "epoch": 4.075842696629214, + "grad_norm": 0.917030394077301, + "learning_rate": 1.1775899741825947e-06, + "loss": 0.5067, + "step": 11608 + }, + { + "epoch": 4.076193820224719, + "grad_norm": 0.9349903464317322, + "learning_rate": 1.177197647114633e-06, + "loss": 0.53, + "step": 11609 + }, + { + "epoch": 4.076544943820225, + "grad_norm": 0.9413900971412659, + "learning_rate": 1.1768053652853806e-06, + "loss": 0.5194, + "step": 11610 + }, + { + "epoch": 4.07689606741573, + "grad_norm": 0.9587412476539612, + "learning_rate": 1.1764131287082513e-06, + "loss": 0.5486, + "step": 11611 + }, + { + "epoch": 4.077247191011236, + "grad_norm": 0.9681575298309326, + "learning_rate": 1.1760209373966618e-06, + "loss": 0.5598, + "step": 11612 + }, + { + "epoch": 4.077598314606742, + "grad_norm": 0.918357789516449, + "learning_rate": 1.175628791364022e-06, + "loss": 0.5097, + "step": 11613 + }, + { + "epoch": 4.077949438202247, + "grad_norm": 0.9537950158119202, + "learning_rate": 1.1752366906237456e-06, + "loss": 0.5456, + "step": 11614 + }, + { + "epoch": 4.078300561797753, + "grad_norm": 0.9683903455734253, + "learning_rate": 1.1748446351892394e-06, + "loss": 0.5278, + "step": 11615 + }, + { + "epoch": 4.078651685393258, + "grad_norm": 0.9353767037391663, + "learning_rate": 1.174452625073914e-06, + "loss": 0.5238, + "step": 11616 + }, + { + "epoch": 4.079002808988764, + "grad_norm": 0.9536352753639221, + "learning_rate": 1.1740606602911734e-06, + "loss": 0.53, + "step": 11617 + }, + { + "epoch": 4.07935393258427, + "grad_norm": 0.9400316476821899, + "learning_rate": 1.1736687408544246e-06, + "loss": 0.5377, + "step": 11618 + }, + { + "epoch": 4.079705056179775, + "grad_norm": 0.9384013414382935, + "learning_rate": 1.1732768667770697e-06, + "loss": 0.4796, + "step": 11619 + }, + { + "epoch": 4.080056179775281, + "grad_norm": 0.9437180161476135, + "learning_rate": 1.1728850380725098e-06, + "loss": 0.5306, + "step": 11620 + }, + { + "epoch": 4.080407303370786, + "grad_norm": 0.9359545707702637, + "learning_rate": 1.172493254754147e-06, + "loss": 0.5132, + "step": 11621 + }, + { + "epoch": 4.080758426966292, + "grad_norm": 0.9716233015060425, + "learning_rate": 1.172101516835378e-06, + "loss": 0.566, + "step": 11622 + }, + { + "epoch": 4.081109550561798, + "grad_norm": 0.9926428198814392, + "learning_rate": 1.1717098243296015e-06, + "loss": 0.5829, + "step": 11623 + }, + { + "epoch": 4.081460674157303, + "grad_norm": 0.9643036723136902, + "learning_rate": 1.171318177250212e-06, + "loss": 0.5398, + "step": 11624 + }, + { + "epoch": 4.081811797752809, + "grad_norm": 0.941554069519043, + "learning_rate": 1.1709265756106045e-06, + "loss": 0.5248, + "step": 11625 + }, + { + "epoch": 4.082162921348314, + "grad_norm": 0.9583423137664795, + "learning_rate": 1.1705350194241697e-06, + "loss": 0.5636, + "step": 11626 + }, + { + "epoch": 4.08251404494382, + "grad_norm": 0.9810033440589905, + "learning_rate": 1.1701435087043006e-06, + "loss": 0.528, + "step": 11627 + }, + { + "epoch": 4.082865168539326, + "grad_norm": 0.9614062309265137, + "learning_rate": 1.1697520434643848e-06, + "loss": 0.5349, + "step": 11628 + }, + { + "epoch": 4.083216292134831, + "grad_norm": 0.9315068125724792, + "learning_rate": 1.1693606237178115e-06, + "loss": 0.5167, + "step": 11629 + }, + { + "epoch": 4.083567415730337, + "grad_norm": 0.9869241118431091, + "learning_rate": 1.1689692494779653e-06, + "loss": 0.5489, + "step": 11630 + }, + { + "epoch": 4.083918539325842, + "grad_norm": 0.9442010521888733, + "learning_rate": 1.1685779207582328e-06, + "loss": 0.5769, + "step": 11631 + }, + { + "epoch": 4.084269662921348, + "grad_norm": 0.9418100118637085, + "learning_rate": 1.1681866375719963e-06, + "loss": 0.5413, + "step": 11632 + }, + { + "epoch": 4.084620786516854, + "grad_norm": 0.9764494895935059, + "learning_rate": 1.167795399932636e-06, + "loss": 0.5423, + "step": 11633 + }, + { + "epoch": 4.084971910112359, + "grad_norm": 0.9401649832725525, + "learning_rate": 1.1674042078535339e-06, + "loss": 0.5345, + "step": 11634 + }, + { + "epoch": 4.085323033707865, + "grad_norm": 0.959255039691925, + "learning_rate": 1.1670130613480669e-06, + "loss": 0.5533, + "step": 11635 + }, + { + "epoch": 4.08567415730337, + "grad_norm": 0.9249891042709351, + "learning_rate": 1.1666219604296134e-06, + "loss": 0.53, + "step": 11636 + }, + { + "epoch": 4.086025280898877, + "grad_norm": 0.9851195812225342, + "learning_rate": 1.1662309051115468e-06, + "loss": 0.5722, + "step": 11637 + }, + { + "epoch": 4.086376404494382, + "grad_norm": 0.9953656792640686, + "learning_rate": 1.1658398954072428e-06, + "loss": 0.542, + "step": 11638 + }, + { + "epoch": 4.086727528089888, + "grad_norm": 0.965554416179657, + "learning_rate": 1.1654489313300718e-06, + "loss": 0.5433, + "step": 11639 + }, + { + "epoch": 4.087078651685394, + "grad_norm": 0.9514992237091064, + "learning_rate": 1.1650580128934065e-06, + "loss": 0.5575, + "step": 11640 + }, + { + "epoch": 4.087429775280899, + "grad_norm": 1.1209847927093506, + "learning_rate": 1.1646671401106139e-06, + "loss": 0.62, + "step": 11641 + }, + { + "epoch": 4.087780898876405, + "grad_norm": 0.9434120059013367, + "learning_rate": 1.1642763129950632e-06, + "loss": 0.5546, + "step": 11642 + }, + { + "epoch": 4.08813202247191, + "grad_norm": 0.9528225660324097, + "learning_rate": 1.1638855315601187e-06, + "loss": 0.5358, + "step": 11643 + }, + { + "epoch": 4.088483146067416, + "grad_norm": 0.9393721222877502, + "learning_rate": 1.1634947958191467e-06, + "loss": 0.5188, + "step": 11644 + }, + { + "epoch": 4.088834269662922, + "grad_norm": 0.9560418128967285, + "learning_rate": 1.1631041057855088e-06, + "loss": 0.5309, + "step": 11645 + }, + { + "epoch": 4.089185393258427, + "grad_norm": 0.9546862840652466, + "learning_rate": 1.1627134614725657e-06, + "loss": 0.5338, + "step": 11646 + }, + { + "epoch": 4.089536516853933, + "grad_norm": 0.9545395374298096, + "learning_rate": 1.1623228628936788e-06, + "loss": 0.5276, + "step": 11647 + }, + { + "epoch": 4.089887640449438, + "grad_norm": 0.9744759202003479, + "learning_rate": 1.1619323100622046e-06, + "loss": 0.5737, + "step": 11648 + }, + { + "epoch": 4.090238764044944, + "grad_norm": 0.8839511871337891, + "learning_rate": 1.1615418029915012e-06, + "loss": 0.4978, + "step": 11649 + }, + { + "epoch": 4.09058988764045, + "grad_norm": 0.9239423871040344, + "learning_rate": 1.161151341694922e-06, + "loss": 0.5209, + "step": 11650 + }, + { + "epoch": 4.090941011235955, + "grad_norm": 0.9801817536354065, + "learning_rate": 1.1607609261858218e-06, + "loss": 0.5238, + "step": 11651 + }, + { + "epoch": 4.091292134831461, + "grad_norm": 0.968521773815155, + "learning_rate": 1.1603705564775513e-06, + "loss": 0.5629, + "step": 11652 + }, + { + "epoch": 4.091643258426966, + "grad_norm": 0.9616255164146423, + "learning_rate": 1.1599802325834621e-06, + "loss": 0.5562, + "step": 11653 + }, + { + "epoch": 4.091994382022472, + "grad_norm": 0.9671490788459778, + "learning_rate": 1.1595899545169017e-06, + "loss": 0.5438, + "step": 11654 + }, + { + "epoch": 4.092345505617978, + "grad_norm": 2.4644229412078857, + "learning_rate": 1.1591997222912185e-06, + "loss": 0.4857, + "step": 11655 + }, + { + "epoch": 4.092696629213483, + "grad_norm": 0.9633327126502991, + "learning_rate": 1.1588095359197564e-06, + "loss": 0.5736, + "step": 11656 + }, + { + "epoch": 4.093047752808989, + "grad_norm": 0.9441542625427246, + "learning_rate": 1.1584193954158618e-06, + "loss": 0.5798, + "step": 11657 + }, + { + "epoch": 4.093398876404494, + "grad_norm": 0.9682712554931641, + "learning_rate": 1.158029300792876e-06, + "loss": 0.5401, + "step": 11658 + }, + { + "epoch": 4.09375, + "grad_norm": 0.961936891078949, + "learning_rate": 1.1576392520641382e-06, + "loss": 0.5494, + "step": 11659 + }, + { + "epoch": 4.094101123595506, + "grad_norm": 0.9664946794509888, + "learning_rate": 1.1572492492429906e-06, + "loss": 0.5622, + "step": 11660 + }, + { + "epoch": 4.094452247191011, + "grad_norm": 0.9751226305961609, + "learning_rate": 1.1568592923427688e-06, + "loss": 0.5272, + "step": 11661 + }, + { + "epoch": 4.094803370786517, + "grad_norm": 0.9314601421356201, + "learning_rate": 1.1564693813768107e-06, + "loss": 0.5248, + "step": 11662 + }, + { + "epoch": 4.095154494382022, + "grad_norm": 0.9534780979156494, + "learning_rate": 1.1560795163584494e-06, + "loss": 0.5345, + "step": 11663 + }, + { + "epoch": 4.095505617977528, + "grad_norm": 0.9965634346008301, + "learning_rate": 1.1556896973010192e-06, + "loss": 0.5282, + "step": 11664 + }, + { + "epoch": 4.095856741573034, + "grad_norm": 1.0273871421813965, + "learning_rate": 1.1552999242178502e-06, + "loss": 0.5584, + "step": 11665 + }, + { + "epoch": 4.096207865168539, + "grad_norm": 0.9666382670402527, + "learning_rate": 1.154910197122274e-06, + "loss": 0.5301, + "step": 11666 + }, + { + "epoch": 4.096558988764045, + "grad_norm": 0.9477785229682922, + "learning_rate": 1.154520516027617e-06, + "loss": 0.5092, + "step": 11667 + }, + { + "epoch": 4.09691011235955, + "grad_norm": 0.9624561071395874, + "learning_rate": 1.1541308809472085e-06, + "loss": 0.5099, + "step": 11668 + }, + { + "epoch": 4.097261235955056, + "grad_norm": 0.986264705657959, + "learning_rate": 1.1537412918943715e-06, + "loss": 0.5638, + "step": 11669 + }, + { + "epoch": 4.097612359550562, + "grad_norm": 0.9493083357810974, + "learning_rate": 1.1533517488824297e-06, + "loss": 0.5311, + "step": 11670 + }, + { + "epoch": 4.097963483146067, + "grad_norm": 0.9619984030723572, + "learning_rate": 1.152962251924707e-06, + "loss": 0.5359, + "step": 11671 + }, + { + "epoch": 4.098314606741573, + "grad_norm": 0.975823163986206, + "learning_rate": 1.1525728010345211e-06, + "loss": 0.5762, + "step": 11672 + }, + { + "epoch": 4.098665730337078, + "grad_norm": 0.9852646589279175, + "learning_rate": 1.1521833962251935e-06, + "loss": 0.5691, + "step": 11673 + }, + { + "epoch": 4.099016853932584, + "grad_norm": 0.9467151165008545, + "learning_rate": 1.1517940375100398e-06, + "loss": 0.5828, + "step": 11674 + }, + { + "epoch": 4.09936797752809, + "grad_norm": 0.9277396202087402, + "learning_rate": 1.151404724902377e-06, + "loss": 0.5031, + "step": 11675 + }, + { + "epoch": 4.099719101123595, + "grad_norm": 0.9912547469139099, + "learning_rate": 1.1510154584155176e-06, + "loss": 0.5565, + "step": 11676 + }, + { + "epoch": 4.100070224719101, + "grad_norm": 0.9045020937919617, + "learning_rate": 1.150626238062776e-06, + "loss": 0.5317, + "step": 11677 + }, + { + "epoch": 4.100421348314606, + "grad_norm": 0.9312042593955994, + "learning_rate": 1.1502370638574617e-06, + "loss": 0.532, + "step": 11678 + }, + { + "epoch": 4.100772471910112, + "grad_norm": 0.9585416316986084, + "learning_rate": 1.1498479358128858e-06, + "loss": 0.5482, + "step": 11679 + }, + { + "epoch": 4.101123595505618, + "grad_norm": 0.9324579834938049, + "learning_rate": 1.1494588539423541e-06, + "loss": 0.5374, + "step": 11680 + }, + { + "epoch": 4.101474719101123, + "grad_norm": 0.9350960850715637, + "learning_rate": 1.1490698182591746e-06, + "loss": 0.5643, + "step": 11681 + }, + { + "epoch": 4.10182584269663, + "grad_norm": 0.9588510990142822, + "learning_rate": 1.1486808287766517e-06, + "loss": 0.5554, + "step": 11682 + }, + { + "epoch": 4.102176966292135, + "grad_norm": 0.9810327887535095, + "learning_rate": 1.1482918855080872e-06, + "loss": 0.5366, + "step": 11683 + }, + { + "epoch": 4.102528089887641, + "grad_norm": 0.9594876170158386, + "learning_rate": 1.147902988466784e-06, + "loss": 0.5355, + "step": 11684 + }, + { + "epoch": 4.102879213483146, + "grad_norm": 0.9359371662139893, + "learning_rate": 1.1475141376660412e-06, + "loss": 0.547, + "step": 11685 + }, + { + "epoch": 4.103230337078652, + "grad_norm": 0.9949642419815063, + "learning_rate": 1.1471253331191581e-06, + "loss": 0.5544, + "step": 11686 + }, + { + "epoch": 4.103581460674158, + "grad_norm": 0.9658296704292297, + "learning_rate": 1.1467365748394303e-06, + "loss": 0.5534, + "step": 11687 + }, + { + "epoch": 4.103932584269663, + "grad_norm": 0.9456151723861694, + "learning_rate": 1.1463478628401545e-06, + "loss": 0.5311, + "step": 11688 + }, + { + "epoch": 4.104283707865169, + "grad_norm": 1.0035172700881958, + "learning_rate": 1.1459591971346224e-06, + "loss": 0.5245, + "step": 11689 + }, + { + "epoch": 4.104634831460674, + "grad_norm": 0.9599520564079285, + "learning_rate": 1.1455705777361282e-06, + "loss": 0.5276, + "step": 11690 + }, + { + "epoch": 4.10498595505618, + "grad_norm": 0.9768447279930115, + "learning_rate": 1.1451820046579602e-06, + "loss": 0.5418, + "step": 11691 + }, + { + "epoch": 4.105337078651686, + "grad_norm": 0.9612562656402588, + "learning_rate": 1.144793477913409e-06, + "loss": 0.4952, + "step": 11692 + }, + { + "epoch": 4.105688202247191, + "grad_norm": 0.9589805006980896, + "learning_rate": 1.1444049975157607e-06, + "loss": 0.5211, + "step": 11693 + }, + { + "epoch": 4.106039325842697, + "grad_norm": 0.9986424446105957, + "learning_rate": 1.1440165634783022e-06, + "loss": 0.5574, + "step": 11694 + }, + { + "epoch": 4.106390449438202, + "grad_norm": 0.9862613677978516, + "learning_rate": 1.1436281758143167e-06, + "loss": 0.5492, + "step": 11695 + }, + { + "epoch": 4.106741573033708, + "grad_norm": 0.9589019417762756, + "learning_rate": 1.1432398345370862e-06, + "loss": 0.5504, + "step": 11696 + }, + { + "epoch": 4.107092696629214, + "grad_norm": 0.9814654588699341, + "learning_rate": 1.1428515396598933e-06, + "loss": 0.5441, + "step": 11697 + }, + { + "epoch": 4.107443820224719, + "grad_norm": 0.9308087229728699, + "learning_rate": 1.1424632911960155e-06, + "loss": 0.5595, + "step": 11698 + }, + { + "epoch": 4.107794943820225, + "grad_norm": 0.9518124461174011, + "learning_rate": 1.1420750891587323e-06, + "loss": 0.5304, + "step": 11699 + }, + { + "epoch": 4.10814606741573, + "grad_norm": 0.9655591249465942, + "learning_rate": 1.141686933561318e-06, + "loss": 0.5351, + "step": 11700 + }, + { + "epoch": 4.108497191011236, + "grad_norm": 0.9363293647766113, + "learning_rate": 1.1412988244170495e-06, + "loss": 0.5474, + "step": 11701 + }, + { + "epoch": 4.108848314606742, + "grad_norm": 0.9598181843757629, + "learning_rate": 1.1409107617391975e-06, + "loss": 0.5198, + "step": 11702 + }, + { + "epoch": 4.109199438202247, + "grad_norm": 0.9383897185325623, + "learning_rate": 1.1405227455410357e-06, + "loss": 0.5241, + "step": 11703 + }, + { + "epoch": 4.109550561797753, + "grad_norm": 0.9497976899147034, + "learning_rate": 1.140134775835831e-06, + "loss": 0.5298, + "step": 11704 + }, + { + "epoch": 4.109901685393258, + "grad_norm": 0.9708424806594849, + "learning_rate": 1.1397468526368548e-06, + "loss": 0.5502, + "step": 11705 + }, + { + "epoch": 4.110252808988764, + "grad_norm": 0.965752899646759, + "learning_rate": 1.1393589759573713e-06, + "loss": 0.5524, + "step": 11706 + }, + { + "epoch": 4.11060393258427, + "grad_norm": 1.0007284879684448, + "learning_rate": 1.1389711458106472e-06, + "loss": 0.5716, + "step": 11707 + }, + { + "epoch": 4.110955056179775, + "grad_norm": 0.9619307518005371, + "learning_rate": 1.1385833622099454e-06, + "loss": 0.5556, + "step": 11708 + }, + { + "epoch": 4.111306179775281, + "grad_norm": 0.9458709955215454, + "learning_rate": 1.138195625168527e-06, + "loss": 0.5481, + "step": 11709 + }, + { + "epoch": 4.111657303370786, + "grad_norm": 0.9829618334770203, + "learning_rate": 1.1378079346996534e-06, + "loss": 0.5808, + "step": 11710 + }, + { + "epoch": 4.112008426966292, + "grad_norm": 0.9513505697250366, + "learning_rate": 1.137420290816582e-06, + "loss": 0.5649, + "step": 11711 + }, + { + "epoch": 4.112359550561798, + "grad_norm": 0.9389906525611877, + "learning_rate": 1.1370326935325715e-06, + "loss": 0.527, + "step": 11712 + }, + { + "epoch": 4.112710674157303, + "grad_norm": 0.9555441737174988, + "learning_rate": 1.1366451428608757e-06, + "loss": 0.5482, + "step": 11713 + }, + { + "epoch": 4.113061797752809, + "grad_norm": 0.9764523506164551, + "learning_rate": 1.1362576388147505e-06, + "loss": 0.5371, + "step": 11714 + }, + { + "epoch": 4.113412921348314, + "grad_norm": 0.9734777808189392, + "learning_rate": 1.135870181407446e-06, + "loss": 0.5454, + "step": 11715 + }, + { + "epoch": 4.11376404494382, + "grad_norm": 0.9550744891166687, + "learning_rate": 1.1354827706522149e-06, + "loss": 0.5399, + "step": 11716 + }, + { + "epoch": 4.114115168539326, + "grad_norm": 0.9720668792724609, + "learning_rate": 1.1350954065623045e-06, + "loss": 0.5423, + "step": 11717 + }, + { + "epoch": 4.114466292134831, + "grad_norm": 0.9526565074920654, + "learning_rate": 1.1347080891509643e-06, + "loss": 0.5079, + "step": 11718 + }, + { + "epoch": 4.114817415730337, + "grad_norm": 0.9800282120704651, + "learning_rate": 1.134320818431438e-06, + "loss": 0.5525, + "step": 11719 + }, + { + "epoch": 4.115168539325842, + "grad_norm": 0.9561012983322144, + "learning_rate": 1.1339335944169722e-06, + "loss": 0.4812, + "step": 11720 + }, + { + "epoch": 4.115519662921348, + "grad_norm": 0.9483031034469604, + "learning_rate": 1.1335464171208082e-06, + "loss": 0.5512, + "step": 11721 + }, + { + "epoch": 4.115870786516854, + "grad_norm": 0.9723355770111084, + "learning_rate": 1.133159286556187e-06, + "loss": 0.5408, + "step": 11722 + }, + { + "epoch": 4.116221910112359, + "grad_norm": 0.9843430519104004, + "learning_rate": 1.1327722027363492e-06, + "loss": 0.5399, + "step": 11723 + }, + { + "epoch": 4.116573033707865, + "grad_norm": 0.94383305311203, + "learning_rate": 1.1323851656745311e-06, + "loss": 0.5263, + "step": 11724 + }, + { + "epoch": 4.11692415730337, + "grad_norm": 0.9560931921005249, + "learning_rate": 1.131998175383971e-06, + "loss": 0.5319, + "step": 11725 + }, + { + "epoch": 4.117275280898877, + "grad_norm": 0.9850316643714905, + "learning_rate": 1.1316112318779019e-06, + "loss": 0.5299, + "step": 11726 + }, + { + "epoch": 4.117626404494382, + "grad_norm": 0.9716293215751648, + "learning_rate": 1.1312243351695588e-06, + "loss": 0.5475, + "step": 11727 + }, + { + "epoch": 4.117977528089888, + "grad_norm": 0.9442225098609924, + "learning_rate": 1.1308374852721718e-06, + "loss": 0.5233, + "step": 11728 + }, + { + "epoch": 4.118328651685394, + "grad_norm": 0.9375926852226257, + "learning_rate": 1.1304506821989706e-06, + "loss": 0.5398, + "step": 11729 + }, + { + "epoch": 4.118679775280899, + "grad_norm": 0.9863271117210388, + "learning_rate": 1.1300639259631846e-06, + "loss": 0.5509, + "step": 11730 + }, + { + "epoch": 4.119030898876405, + "grad_norm": 0.9587833285331726, + "learning_rate": 1.1296772165780395e-06, + "loss": 0.5463, + "step": 11731 + }, + { + "epoch": 4.11938202247191, + "grad_norm": 0.9465469121932983, + "learning_rate": 1.129290554056762e-06, + "loss": 0.5231, + "step": 11732 + }, + { + "epoch": 4.119733146067416, + "grad_norm": 0.9779781699180603, + "learning_rate": 1.1289039384125732e-06, + "loss": 0.544, + "step": 11733 + }, + { + "epoch": 4.120084269662922, + "grad_norm": 0.9516111612319946, + "learning_rate": 1.1285173696586978e-06, + "loss": 0.5347, + "step": 11734 + }, + { + "epoch": 4.120435393258427, + "grad_norm": 0.9992930889129639, + "learning_rate": 1.1281308478083547e-06, + "loss": 0.5384, + "step": 11735 + }, + { + "epoch": 4.120786516853933, + "grad_norm": 0.9375869035720825, + "learning_rate": 1.1277443728747617e-06, + "loss": 0.5324, + "step": 11736 + }, + { + "epoch": 4.121137640449438, + "grad_norm": 0.9716718792915344, + "learning_rate": 1.127357944871138e-06, + "loss": 0.5541, + "step": 11737 + }, + { + "epoch": 4.121488764044944, + "grad_norm": 0.9394237399101257, + "learning_rate": 1.1269715638106968e-06, + "loss": 0.5036, + "step": 11738 + }, + { + "epoch": 4.12183988764045, + "grad_norm": 0.9456638693809509, + "learning_rate": 1.1265852297066542e-06, + "loss": 0.5345, + "step": 11739 + }, + { + "epoch": 4.122191011235955, + "grad_norm": 0.9643946290016174, + "learning_rate": 1.1261989425722206e-06, + "loss": 0.5308, + "step": 11740 + }, + { + "epoch": 4.122542134831461, + "grad_norm": 0.9596660137176514, + "learning_rate": 1.1258127024206086e-06, + "loss": 0.562, + "step": 11741 + }, + { + "epoch": 4.122893258426966, + "grad_norm": 0.9589257836341858, + "learning_rate": 1.1254265092650262e-06, + "loss": 0.5394, + "step": 11742 + }, + { + "epoch": 4.123244382022472, + "grad_norm": 0.959556519985199, + "learning_rate": 1.12504036311868e-06, + "loss": 0.5577, + "step": 11743 + }, + { + "epoch": 4.123595505617978, + "grad_norm": 0.9431952238082886, + "learning_rate": 1.1246542639947777e-06, + "loss": 0.5283, + "step": 11744 + }, + { + "epoch": 4.123946629213483, + "grad_norm": 1.0025086402893066, + "learning_rate": 1.1242682119065217e-06, + "loss": 0.5505, + "step": 11745 + }, + { + "epoch": 4.124297752808989, + "grad_norm": 0.9794161915779114, + "learning_rate": 1.1238822068671167e-06, + "loss": 0.5324, + "step": 11746 + }, + { + "epoch": 4.124648876404494, + "grad_norm": 0.9592738151550293, + "learning_rate": 1.1234962488897616e-06, + "loss": 0.561, + "step": 11747 + }, + { + "epoch": 4.125, + "grad_norm": 0.9594618082046509, + "learning_rate": 1.123110337987658e-06, + "loss": 0.5657, + "step": 11748 + }, + { + "epoch": 4.125351123595506, + "grad_norm": 0.9584984183311462, + "learning_rate": 1.1227244741740025e-06, + "loss": 0.5266, + "step": 11749 + }, + { + "epoch": 4.125702247191011, + "grad_norm": 0.988339364528656, + "learning_rate": 1.1223386574619904e-06, + "loss": 0.5631, + "step": 11750 + }, + { + "epoch": 4.126053370786517, + "grad_norm": 0.9413666129112244, + "learning_rate": 1.1219528878648184e-06, + "loss": 0.557, + "step": 11751 + }, + { + "epoch": 4.126404494382022, + "grad_norm": 0.965796172618866, + "learning_rate": 1.1215671653956775e-06, + "loss": 0.5195, + "step": 11752 + }, + { + "epoch": 4.126755617977528, + "grad_norm": 0.9411710500717163, + "learning_rate": 1.1211814900677609e-06, + "loss": 0.5025, + "step": 11753 + }, + { + "epoch": 4.127106741573034, + "grad_norm": 0.966634213924408, + "learning_rate": 1.1207958618942566e-06, + "loss": 0.566, + "step": 11754 + }, + { + "epoch": 4.127457865168539, + "grad_norm": 0.9760687351226807, + "learning_rate": 1.1204102808883547e-06, + "loss": 0.575, + "step": 11755 + }, + { + "epoch": 4.127808988764045, + "grad_norm": 0.957106351852417, + "learning_rate": 1.1200247470632394e-06, + "loss": 0.5419, + "step": 11756 + }, + { + "epoch": 4.12816011235955, + "grad_norm": 0.9930441379547119, + "learning_rate": 1.119639260432098e-06, + "loss": 0.5609, + "step": 11757 + }, + { + "epoch": 4.128511235955056, + "grad_norm": 0.9890742897987366, + "learning_rate": 1.1192538210081126e-06, + "loss": 0.5571, + "step": 11758 + }, + { + "epoch": 4.128862359550562, + "grad_norm": 0.965813159942627, + "learning_rate": 1.118868428804464e-06, + "loss": 0.5474, + "step": 11759 + }, + { + "epoch": 4.129213483146067, + "grad_norm": 0.9289169311523438, + "learning_rate": 1.1184830838343343e-06, + "loss": 0.4888, + "step": 11760 + }, + { + "epoch": 4.129564606741573, + "grad_norm": 0.9506020545959473, + "learning_rate": 1.1180977861109003e-06, + "loss": 0.5172, + "step": 11761 + }, + { + "epoch": 4.129915730337078, + "grad_norm": 0.9677910208702087, + "learning_rate": 1.11771253564734e-06, + "loss": 0.5471, + "step": 11762 + }, + { + "epoch": 4.130266853932584, + "grad_norm": 0.9305959939956665, + "learning_rate": 1.1173273324568274e-06, + "loss": 0.5194, + "step": 11763 + }, + { + "epoch": 4.13061797752809, + "grad_norm": 0.9668494462966919, + "learning_rate": 1.116942176552538e-06, + "loss": 0.5737, + "step": 11764 + }, + { + "epoch": 4.130969101123595, + "grad_norm": 0.9399253726005554, + "learning_rate": 1.1165570679476412e-06, + "loss": 0.5125, + "step": 11765 + }, + { + "epoch": 4.131320224719101, + "grad_norm": 0.974263072013855, + "learning_rate": 1.1161720066553103e-06, + "loss": 0.5709, + "step": 11766 + }, + { + "epoch": 4.131671348314606, + "grad_norm": 0.9727210402488708, + "learning_rate": 1.1157869926887113e-06, + "loss": 0.5143, + "step": 11767 + }, + { + "epoch": 4.132022471910112, + "grad_norm": 0.9477127194404602, + "learning_rate": 1.1154020260610138e-06, + "loss": 0.5435, + "step": 11768 + }, + { + "epoch": 4.132373595505618, + "grad_norm": 0.9573505520820618, + "learning_rate": 1.1150171067853814e-06, + "loss": 0.5465, + "step": 11769 + }, + { + "epoch": 4.132724719101123, + "grad_norm": 0.9859501123428345, + "learning_rate": 1.1146322348749794e-06, + "loss": 0.5528, + "step": 11770 + }, + { + "epoch": 4.13307584269663, + "grad_norm": 0.9343919157981873, + "learning_rate": 1.1142474103429697e-06, + "loss": 0.4993, + "step": 11771 + }, + { + "epoch": 4.133426966292135, + "grad_norm": 0.9656389355659485, + "learning_rate": 1.1138626332025117e-06, + "loss": 0.5444, + "step": 11772 + }, + { + "epoch": 4.133778089887641, + "grad_norm": 0.9633687734603882, + "learning_rate": 1.1134779034667666e-06, + "loss": 0.5277, + "step": 11773 + }, + { + "epoch": 4.134129213483146, + "grad_norm": 0.9559308290481567, + "learning_rate": 1.1130932211488898e-06, + "loss": 0.5471, + "step": 11774 + }, + { + "epoch": 4.134480337078652, + "grad_norm": 0.9728759527206421, + "learning_rate": 1.1127085862620387e-06, + "loss": 0.5412, + "step": 11775 + }, + { + "epoch": 4.134831460674158, + "grad_norm": 0.9551388621330261, + "learning_rate": 1.1123239988193663e-06, + "loss": 0.525, + "step": 11776 + }, + { + "epoch": 4.135182584269663, + "grad_norm": 0.9645889401435852, + "learning_rate": 1.1119394588340262e-06, + "loss": 0.5615, + "step": 11777 + }, + { + "epoch": 4.135533707865169, + "grad_norm": 0.9577795267105103, + "learning_rate": 1.1115549663191684e-06, + "loss": 0.5282, + "step": 11778 + }, + { + "epoch": 4.135884831460674, + "grad_norm": 0.9600687026977539, + "learning_rate": 1.111170521287943e-06, + "loss": 0.523, + "step": 11779 + }, + { + "epoch": 4.13623595505618, + "grad_norm": 0.9659762382507324, + "learning_rate": 1.110786123753497e-06, + "loss": 0.537, + "step": 11780 + }, + { + "epoch": 4.136587078651686, + "grad_norm": 0.9629111886024475, + "learning_rate": 1.1104017737289775e-06, + "loss": 0.5387, + "step": 11781 + }, + { + "epoch": 4.136938202247191, + "grad_norm": 0.9675184488296509, + "learning_rate": 1.110017471227527e-06, + "loss": 0.5597, + "step": 11782 + }, + { + "epoch": 4.137289325842697, + "grad_norm": 0.9262818098068237, + "learning_rate": 1.1096332162622909e-06, + "loss": 0.4922, + "step": 11783 + }, + { + "epoch": 4.137640449438202, + "grad_norm": 0.9290157556533813, + "learning_rate": 1.1092490088464087e-06, + "loss": 0.5404, + "step": 11784 + }, + { + "epoch": 4.137991573033708, + "grad_norm": 0.9216079711914062, + "learning_rate": 1.1088648489930196e-06, + "loss": 0.5252, + "step": 11785 + }, + { + "epoch": 4.138342696629214, + "grad_norm": 0.9277830123901367, + "learning_rate": 1.108480736715263e-06, + "loss": 0.5299, + "step": 11786 + }, + { + "epoch": 4.138693820224719, + "grad_norm": 0.9472830295562744, + "learning_rate": 1.1080966720262737e-06, + "loss": 0.5347, + "step": 11787 + }, + { + "epoch": 4.139044943820225, + "grad_norm": 0.9587204456329346, + "learning_rate": 1.1077126549391882e-06, + "loss": 0.5538, + "step": 11788 + }, + { + "epoch": 4.13939606741573, + "grad_norm": 0.9609595537185669, + "learning_rate": 1.1073286854671376e-06, + "loss": 0.5456, + "step": 11789 + }, + { + "epoch": 4.139747191011236, + "grad_norm": 0.9409713745117188, + "learning_rate": 1.1069447636232547e-06, + "loss": 0.5266, + "step": 11790 + }, + { + "epoch": 4.140098314606742, + "grad_norm": 0.9901164770126343, + "learning_rate": 1.1065608894206684e-06, + "loss": 0.533, + "step": 11791 + }, + { + "epoch": 4.140449438202247, + "grad_norm": 0.9785668849945068, + "learning_rate": 1.106177062872508e-06, + "loss": 0.5728, + "step": 11792 + }, + { + "epoch": 4.140800561797753, + "grad_norm": 0.9377629160881042, + "learning_rate": 1.1057932839918984e-06, + "loss": 0.5038, + "step": 11793 + }, + { + "epoch": 4.141151685393258, + "grad_norm": 0.9569548964500427, + "learning_rate": 1.1054095527919664e-06, + "loss": 0.5315, + "step": 11794 + }, + { + "epoch": 4.141502808988764, + "grad_norm": 0.9425296187400818, + "learning_rate": 1.1050258692858335e-06, + "loss": 0.5302, + "step": 11795 + }, + { + "epoch": 4.14185393258427, + "grad_norm": 0.9674885272979736, + "learning_rate": 1.1046422334866227e-06, + "loss": 0.5257, + "step": 11796 + }, + { + "epoch": 4.142205056179775, + "grad_norm": 0.9491012096405029, + "learning_rate": 1.104258645407454e-06, + "loss": 0.534, + "step": 11797 + }, + { + "epoch": 4.142556179775281, + "grad_norm": 0.9507694840431213, + "learning_rate": 1.1038751050614441e-06, + "loss": 0.5394, + "step": 11798 + }, + { + "epoch": 4.142907303370786, + "grad_norm": 0.9783778786659241, + "learning_rate": 1.1034916124617117e-06, + "loss": 0.5464, + "step": 11799 + }, + { + "epoch": 4.143258426966292, + "grad_norm": 0.9607770442962646, + "learning_rate": 1.1031081676213706e-06, + "loss": 0.5357, + "step": 11800 + }, + { + "epoch": 4.143609550561798, + "grad_norm": 0.9355868697166443, + "learning_rate": 1.1027247705535355e-06, + "loss": 0.5425, + "step": 11801 + }, + { + "epoch": 4.143960674157303, + "grad_norm": 0.9424923062324524, + "learning_rate": 1.1023414212713168e-06, + "loss": 0.4978, + "step": 11802 + }, + { + "epoch": 4.144311797752809, + "grad_norm": 0.9542232155799866, + "learning_rate": 1.1019581197878262e-06, + "loss": 0.5628, + "step": 11803 + }, + { + "epoch": 4.144662921348314, + "grad_norm": 0.979939877986908, + "learning_rate": 1.101574866116171e-06, + "loss": 0.5589, + "step": 11804 + }, + { + "epoch": 4.14501404494382, + "grad_norm": 0.9352039098739624, + "learning_rate": 1.1011916602694594e-06, + "loss": 0.5254, + "step": 11805 + }, + { + "epoch": 4.145365168539326, + "grad_norm": 0.9402464032173157, + "learning_rate": 1.1008085022607951e-06, + "loss": 0.5067, + "step": 11806 + }, + { + "epoch": 4.145716292134831, + "grad_norm": 0.9749277234077454, + "learning_rate": 1.1004253921032837e-06, + "loss": 0.5496, + "step": 11807 + }, + { + "epoch": 4.146067415730337, + "grad_norm": 0.9648091197013855, + "learning_rate": 1.1000423298100263e-06, + "loss": 0.5539, + "step": 11808 + }, + { + "epoch": 4.146418539325842, + "grad_norm": 0.9620043039321899, + "learning_rate": 1.0996593153941223e-06, + "loss": 0.5021, + "step": 11809 + }, + { + "epoch": 4.146769662921348, + "grad_norm": 0.9479560852050781, + "learning_rate": 1.0992763488686725e-06, + "loss": 0.562, + "step": 11810 + }, + { + "epoch": 4.147120786516854, + "grad_norm": 1.000525951385498, + "learning_rate": 1.098893430246772e-06, + "loss": 0.5318, + "step": 11811 + }, + { + "epoch": 4.147471910112359, + "grad_norm": 0.9448985457420349, + "learning_rate": 1.098510559541518e-06, + "loss": 0.549, + "step": 11812 + }, + { + "epoch": 4.147823033707865, + "grad_norm": 0.9978194236755371, + "learning_rate": 1.0981277367660025e-06, + "loss": 0.5674, + "step": 11813 + }, + { + "epoch": 4.14817415730337, + "grad_norm": 0.9788603782653809, + "learning_rate": 1.09774496193332e-06, + "loss": 0.5295, + "step": 11814 + }, + { + "epoch": 4.148525280898877, + "grad_norm": 0.9489254355430603, + "learning_rate": 1.097362235056559e-06, + "loss": 0.5452, + "step": 11815 + }, + { + "epoch": 4.148876404494382, + "grad_norm": 0.9848926663398743, + "learning_rate": 1.09697955614881e-06, + "loss": 0.5203, + "step": 11816 + }, + { + "epoch": 4.149227528089888, + "grad_norm": 0.9830127954483032, + "learning_rate": 1.096596925223159e-06, + "loss": 0.5519, + "step": 11817 + }, + { + "epoch": 4.149578651685394, + "grad_norm": 0.957989513874054, + "learning_rate": 1.096214342292693e-06, + "loss": 0.5418, + "step": 11818 + }, + { + "epoch": 4.149929775280899, + "grad_norm": 0.9694338440895081, + "learning_rate": 1.0958318073704941e-06, + "loss": 0.5327, + "step": 11819 + }, + { + "epoch": 4.150280898876405, + "grad_norm": 0.9744919538497925, + "learning_rate": 1.095449320469647e-06, + "loss": 0.5147, + "step": 11820 + }, + { + "epoch": 4.15063202247191, + "grad_norm": 0.9788051843643188, + "learning_rate": 1.0950668816032312e-06, + "loss": 0.5247, + "step": 11821 + }, + { + "epoch": 4.150983146067416, + "grad_norm": 0.9562873840332031, + "learning_rate": 1.094684490784325e-06, + "loss": 0.5379, + "step": 11822 + }, + { + "epoch": 4.151334269662922, + "grad_norm": 0.9837181568145752, + "learning_rate": 1.0943021480260075e-06, + "loss": 0.5279, + "step": 11823 + }, + { + "epoch": 4.151685393258427, + "grad_norm": 0.9704144597053528, + "learning_rate": 1.0939198533413527e-06, + "loss": 0.5686, + "step": 11824 + }, + { + "epoch": 4.152036516853933, + "grad_norm": 0.9452782869338989, + "learning_rate": 1.0935376067434366e-06, + "loss": 0.5544, + "step": 11825 + }, + { + "epoch": 4.152387640449438, + "grad_norm": 0.9745513796806335, + "learning_rate": 1.09315540824533e-06, + "loss": 0.5064, + "step": 11826 + }, + { + "epoch": 4.152738764044944, + "grad_norm": 0.9334105253219604, + "learning_rate": 1.0927732578601054e-06, + "loss": 0.5462, + "step": 11827 + }, + { + "epoch": 4.15308988764045, + "grad_norm": 0.9693291783332825, + "learning_rate": 1.0923911556008304e-06, + "loss": 0.5841, + "step": 11828 + }, + { + "epoch": 4.153441011235955, + "grad_norm": 0.9618633389472961, + "learning_rate": 1.0920091014805745e-06, + "loss": 0.5348, + "step": 11829 + }, + { + "epoch": 4.153792134831461, + "grad_norm": 0.9433852434158325, + "learning_rate": 1.0916270955124011e-06, + "loss": 0.5467, + "step": 11830 + }, + { + "epoch": 4.154143258426966, + "grad_norm": 0.9491646885871887, + "learning_rate": 1.0912451377093771e-06, + "loss": 0.489, + "step": 11831 + }, + { + "epoch": 4.154494382022472, + "grad_norm": 0.969402015209198, + "learning_rate": 1.090863228084563e-06, + "loss": 0.522, + "step": 11832 + }, + { + "epoch": 4.154845505617978, + "grad_norm": 0.9507244229316711, + "learning_rate": 1.0904813666510214e-06, + "loss": 0.5578, + "step": 11833 + }, + { + "epoch": 4.155196629213483, + "grad_norm": 0.9809635877609253, + "learning_rate": 1.090099553421811e-06, + "loss": 0.5277, + "step": 11834 + }, + { + "epoch": 4.155547752808989, + "grad_norm": 0.9624118208885193, + "learning_rate": 1.0897177884099886e-06, + "loss": 0.5533, + "step": 11835 + }, + { + "epoch": 4.155898876404494, + "grad_norm": 0.9523085951805115, + "learning_rate": 1.089336071628612e-06, + "loss": 0.5517, + "step": 11836 + }, + { + "epoch": 4.15625, + "grad_norm": 0.9282060861587524, + "learning_rate": 1.0889544030907335e-06, + "loss": 0.5533, + "step": 11837 + }, + { + "epoch": 4.156601123595506, + "grad_norm": 0.958351731300354, + "learning_rate": 1.0885727828094077e-06, + "loss": 0.5247, + "step": 11838 + }, + { + "epoch": 4.156952247191011, + "grad_norm": 0.9492023587226868, + "learning_rate": 1.0881912107976845e-06, + "loss": 0.5421, + "step": 11839 + }, + { + "epoch": 4.157303370786517, + "grad_norm": 0.9584190845489502, + "learning_rate": 1.0878096870686144e-06, + "loss": 0.5498, + "step": 11840 + }, + { + "epoch": 4.157654494382022, + "grad_norm": 0.9672701954841614, + "learning_rate": 1.087428211635244e-06, + "loss": 0.5659, + "step": 11841 + }, + { + "epoch": 4.158005617977528, + "grad_norm": 0.954300582408905, + "learning_rate": 1.0870467845106205e-06, + "loss": 0.5426, + "step": 11842 + }, + { + "epoch": 4.158356741573034, + "grad_norm": 0.9448575377464294, + "learning_rate": 1.0866654057077872e-06, + "loss": 0.5278, + "step": 11843 + }, + { + "epoch": 4.158707865168539, + "grad_norm": 0.9641999006271362, + "learning_rate": 1.0862840752397885e-06, + "loss": 0.551, + "step": 11844 + }, + { + "epoch": 4.159058988764045, + "grad_norm": 0.9773526787757874, + "learning_rate": 1.0859027931196639e-06, + "loss": 0.5235, + "step": 11845 + }, + { + "epoch": 4.15941011235955, + "grad_norm": 0.9449295401573181, + "learning_rate": 1.0855215593604542e-06, + "loss": 0.5506, + "step": 11846 + }, + { + "epoch": 4.159761235955056, + "grad_norm": 0.9594520330429077, + "learning_rate": 1.0851403739751973e-06, + "loss": 0.5424, + "step": 11847 + }, + { + "epoch": 4.160112359550562, + "grad_norm": 0.9703198075294495, + "learning_rate": 1.0847592369769277e-06, + "loss": 0.5688, + "step": 11848 + }, + { + "epoch": 4.160463483146067, + "grad_norm": 0.9589354395866394, + "learning_rate": 1.0843781483786823e-06, + "loss": 0.5377, + "step": 11849 + }, + { + "epoch": 4.160814606741573, + "grad_norm": 1.0417757034301758, + "learning_rate": 1.0839971081934916e-06, + "loss": 0.5275, + "step": 11850 + }, + { + "epoch": 4.161165730337078, + "grad_norm": 0.935356855392456, + "learning_rate": 1.083616116434389e-06, + "loss": 0.5029, + "step": 11851 + }, + { + "epoch": 4.161516853932584, + "grad_norm": 1.0120782852172852, + "learning_rate": 1.0832351731144029e-06, + "loss": 0.5701, + "step": 11852 + }, + { + "epoch": 4.16186797752809, + "grad_norm": 0.9857171177864075, + "learning_rate": 1.082854278246562e-06, + "loss": 0.5338, + "step": 11853 + }, + { + "epoch": 4.162219101123595, + "grad_norm": 0.9710322022438049, + "learning_rate": 1.0824734318438912e-06, + "loss": 0.5562, + "step": 11854 + }, + { + "epoch": 4.162570224719101, + "grad_norm": 0.9563541412353516, + "learning_rate": 1.0820926339194168e-06, + "loss": 0.5558, + "step": 11855 + }, + { + "epoch": 4.162921348314606, + "grad_norm": 0.9721680283546448, + "learning_rate": 1.0817118844861604e-06, + "loss": 0.5303, + "step": 11856 + }, + { + "epoch": 4.163272471910112, + "grad_norm": 0.9324816465377808, + "learning_rate": 1.0813311835571447e-06, + "loss": 0.5258, + "step": 11857 + }, + { + "epoch": 4.163623595505618, + "grad_norm": 0.9583902955055237, + "learning_rate": 1.0809505311453878e-06, + "loss": 0.5384, + "step": 11858 + }, + { + "epoch": 4.163974719101123, + "grad_norm": 0.97024005651474, + "learning_rate": 1.080569927263909e-06, + "loss": 0.5561, + "step": 11859 + }, + { + "epoch": 4.16432584269663, + "grad_norm": 0.9666517376899719, + "learning_rate": 1.0801893719257241e-06, + "loss": 0.5099, + "step": 11860 + }, + { + "epoch": 4.164676966292135, + "grad_norm": 0.9955719709396362, + "learning_rate": 1.079808865143847e-06, + "loss": 0.5718, + "step": 11861 + }, + { + "epoch": 4.165028089887641, + "grad_norm": 0.9732844829559326, + "learning_rate": 1.0794284069312925e-06, + "loss": 0.537, + "step": 11862 + }, + { + "epoch": 4.165379213483146, + "grad_norm": 1.0189034938812256, + "learning_rate": 1.0790479973010697e-06, + "loss": 0.5054, + "step": 11863 + }, + { + "epoch": 4.165730337078652, + "grad_norm": 0.961245059967041, + "learning_rate": 1.0786676362661905e-06, + "loss": 0.5171, + "step": 11864 + }, + { + "epoch": 4.166081460674158, + "grad_norm": 0.9637192487716675, + "learning_rate": 1.078287323839661e-06, + "loss": 0.5571, + "step": 11865 + }, + { + "epoch": 4.166432584269663, + "grad_norm": 0.9310654997825623, + "learning_rate": 1.0779070600344894e-06, + "loss": 0.5181, + "step": 11866 + }, + { + "epoch": 4.166783707865169, + "grad_norm": 0.9596859812736511, + "learning_rate": 1.0775268448636782e-06, + "loss": 0.567, + "step": 11867 + }, + { + "epoch": 4.167134831460674, + "grad_norm": 0.9718461632728577, + "learning_rate": 1.077146678340233e-06, + "loss": 0.5342, + "step": 11868 + }, + { + "epoch": 4.16748595505618, + "grad_norm": 0.93833327293396, + "learning_rate": 1.0767665604771528e-06, + "loss": 0.5415, + "step": 11869 + }, + { + "epoch": 4.167837078651686, + "grad_norm": 0.9644821286201477, + "learning_rate": 1.0763864912874391e-06, + "loss": 0.5389, + "step": 11870 + }, + { + "epoch": 4.168188202247191, + "grad_norm": 0.9888043999671936, + "learning_rate": 1.0760064707840892e-06, + "loss": 0.5379, + "step": 11871 + }, + { + "epoch": 4.168539325842697, + "grad_norm": 0.9585707783699036, + "learning_rate": 1.0756264989800988e-06, + "loss": 0.5159, + "step": 11872 + }, + { + "epoch": 4.168890449438202, + "grad_norm": 0.9536551833152771, + "learning_rate": 1.0752465758884638e-06, + "loss": 0.5509, + "step": 11873 + }, + { + "epoch": 4.169241573033708, + "grad_norm": 0.9685904383659363, + "learning_rate": 1.0748667015221764e-06, + "loss": 0.5455, + "step": 11874 + }, + { + "epoch": 4.169592696629214, + "grad_norm": 0.960155725479126, + "learning_rate": 1.074486875894229e-06, + "loss": 0.5394, + "step": 11875 + }, + { + "epoch": 4.169943820224719, + "grad_norm": 0.9726266264915466, + "learning_rate": 1.0741070990176094e-06, + "loss": 0.546, + "step": 11876 + }, + { + "epoch": 4.170294943820225, + "grad_norm": 0.920966386795044, + "learning_rate": 1.0737273709053081e-06, + "loss": 0.5254, + "step": 11877 + }, + { + "epoch": 4.17064606741573, + "grad_norm": 0.9640771150588989, + "learning_rate": 1.0733476915703092e-06, + "loss": 0.5366, + "step": 11878 + }, + { + "epoch": 4.170997191011236, + "grad_norm": 0.9401295185089111, + "learning_rate": 1.0729680610255995e-06, + "loss": 0.5122, + "step": 11879 + }, + { + "epoch": 4.171348314606742, + "grad_norm": 0.9676121473312378, + "learning_rate": 1.0725884792841598e-06, + "loss": 0.5426, + "step": 11880 + }, + { + "epoch": 4.171699438202247, + "grad_norm": 0.951767086982727, + "learning_rate": 1.0722089463589735e-06, + "loss": 0.5277, + "step": 11881 + }, + { + "epoch": 4.172050561797753, + "grad_norm": 0.9472304582595825, + "learning_rate": 1.0718294622630188e-06, + "loss": 0.5328, + "step": 11882 + }, + { + "epoch": 4.172401685393258, + "grad_norm": 0.9775479435920715, + "learning_rate": 1.0714500270092753e-06, + "loss": 0.5695, + "step": 11883 + }, + { + "epoch": 4.172752808988764, + "grad_norm": 0.9557408094406128, + "learning_rate": 1.0710706406107182e-06, + "loss": 0.5483, + "step": 11884 + }, + { + "epoch": 4.17310393258427, + "grad_norm": 0.9796738624572754, + "learning_rate": 1.0706913030803217e-06, + "loss": 0.5501, + "step": 11885 + }, + { + "epoch": 4.173455056179775, + "grad_norm": 0.9526545405387878, + "learning_rate": 1.0703120144310605e-06, + "loss": 0.5478, + "step": 11886 + }, + { + "epoch": 4.173806179775281, + "grad_norm": 0.9609228372573853, + "learning_rate": 1.0699327746759041e-06, + "loss": 0.5448, + "step": 11887 + }, + { + "epoch": 4.174157303370786, + "grad_norm": 0.9506641626358032, + "learning_rate": 1.069553583827824e-06, + "loss": 0.5289, + "step": 11888 + }, + { + "epoch": 4.174508426966292, + "grad_norm": 0.9561762809753418, + "learning_rate": 1.0691744418997866e-06, + "loss": 0.5071, + "step": 11889 + }, + { + "epoch": 4.174859550561798, + "grad_norm": 0.9527935981750488, + "learning_rate": 1.0687953489047594e-06, + "loss": 0.5398, + "step": 11890 + }, + { + "epoch": 4.175210674157303, + "grad_norm": 0.9944334626197815, + "learning_rate": 1.0684163048557058e-06, + "loss": 0.5657, + "step": 11891 + }, + { + "epoch": 4.175561797752809, + "grad_norm": 1.009658932685852, + "learning_rate": 1.0680373097655904e-06, + "loss": 0.5154, + "step": 11892 + }, + { + "epoch": 4.175912921348314, + "grad_norm": 0.9824627637863159, + "learning_rate": 1.067658363647374e-06, + "loss": 0.5205, + "step": 11893 + }, + { + "epoch": 4.17626404494382, + "grad_norm": 0.9627036452293396, + "learning_rate": 1.0672794665140145e-06, + "loss": 0.542, + "step": 11894 + }, + { + "epoch": 4.176615168539326, + "grad_norm": 0.9377517104148865, + "learning_rate": 1.0669006183784726e-06, + "loss": 0.5202, + "step": 11895 + }, + { + "epoch": 4.176966292134831, + "grad_norm": 0.9559123516082764, + "learning_rate": 1.066521819253702e-06, + "loss": 0.5515, + "step": 11896 + }, + { + "epoch": 4.177317415730337, + "grad_norm": 0.9259557127952576, + "learning_rate": 1.0661430691526594e-06, + "loss": 0.5149, + "step": 11897 + }, + { + "epoch": 4.177668539325842, + "grad_norm": 0.9596243500709534, + "learning_rate": 1.0657643680882961e-06, + "loss": 0.5411, + "step": 11898 + }, + { + "epoch": 4.178019662921348, + "grad_norm": 0.9328064918518066, + "learning_rate": 1.065385716073565e-06, + "loss": 0.5154, + "step": 11899 + }, + { + "epoch": 4.178370786516854, + "grad_norm": 0.9461041688919067, + "learning_rate": 1.065007113121414e-06, + "loss": 0.5165, + "step": 11900 + }, + { + "epoch": 4.178721910112359, + "grad_norm": 0.9710589647293091, + "learning_rate": 1.0646285592447924e-06, + "loss": 0.5453, + "step": 11901 + }, + { + "epoch": 4.179073033707865, + "grad_norm": 0.9403935670852661, + "learning_rate": 1.064250054456646e-06, + "loss": 0.5183, + "step": 11902 + }, + { + "epoch": 4.17942415730337, + "grad_norm": 0.9525834321975708, + "learning_rate": 1.0638715987699186e-06, + "loss": 0.5335, + "step": 11903 + }, + { + "epoch": 4.179775280898877, + "grad_norm": 0.9559136033058167, + "learning_rate": 1.063493192197554e-06, + "loss": 0.5329, + "step": 11904 + }, + { + "epoch": 4.180126404494382, + "grad_norm": 0.978340208530426, + "learning_rate": 1.0631148347524923e-06, + "loss": 0.5431, + "step": 11905 + }, + { + "epoch": 4.180477528089888, + "grad_norm": 0.9341151118278503, + "learning_rate": 1.0627365264476746e-06, + "loss": 0.5077, + "step": 11906 + }, + { + "epoch": 4.180828651685394, + "grad_norm": 0.960879385471344, + "learning_rate": 1.062358267296037e-06, + "loss": 0.548, + "step": 11907 + }, + { + "epoch": 4.181179775280899, + "grad_norm": 0.9418762922286987, + "learning_rate": 1.0619800573105171e-06, + "loss": 0.5668, + "step": 11908 + }, + { + "epoch": 4.181530898876405, + "grad_norm": 0.9359615445137024, + "learning_rate": 1.0616018965040486e-06, + "loss": 0.5009, + "step": 11909 + }, + { + "epoch": 4.18188202247191, + "grad_norm": 0.9334201216697693, + "learning_rate": 1.061223784889564e-06, + "loss": 0.5348, + "step": 11910 + }, + { + "epoch": 4.182233146067416, + "grad_norm": 0.9490492939949036, + "learning_rate": 1.0608457224799953e-06, + "loss": 0.5452, + "step": 11911 + }, + { + "epoch": 4.182584269662922, + "grad_norm": 0.9224563837051392, + "learning_rate": 1.0604677092882706e-06, + "loss": 0.542, + "step": 11912 + }, + { + "epoch": 4.182935393258427, + "grad_norm": 0.9579346776008606, + "learning_rate": 1.0600897453273194e-06, + "loss": 0.5348, + "step": 11913 + }, + { + "epoch": 4.183286516853933, + "grad_norm": 0.9404693841934204, + "learning_rate": 1.0597118306100657e-06, + "loss": 0.5204, + "step": 11914 + }, + { + "epoch": 4.183637640449438, + "grad_norm": 0.979247510433197, + "learning_rate": 1.059333965149436e-06, + "loss": 0.5498, + "step": 11915 + }, + { + "epoch": 4.183988764044944, + "grad_norm": 0.9278571009635925, + "learning_rate": 1.0589561489583516e-06, + "loss": 0.5037, + "step": 11916 + }, + { + "epoch": 4.18433988764045, + "grad_norm": 0.9296547174453735, + "learning_rate": 1.0585783820497332e-06, + "loss": 0.5415, + "step": 11917 + }, + { + "epoch": 4.184691011235955, + "grad_norm": 0.9996165633201599, + "learning_rate": 1.0582006644365012e-06, + "loss": 0.5232, + "step": 11918 + }, + { + "epoch": 4.185042134831461, + "grad_norm": 0.9348304867744446, + "learning_rate": 1.0578229961315722e-06, + "loss": 0.5154, + "step": 11919 + }, + { + "epoch": 4.185393258426966, + "grad_norm": 0.9307035803794861, + "learning_rate": 1.0574453771478631e-06, + "loss": 0.5383, + "step": 11920 + }, + { + "epoch": 4.185744382022472, + "grad_norm": 0.9194188714027405, + "learning_rate": 1.0570678074982873e-06, + "loss": 0.5054, + "step": 11921 + }, + { + "epoch": 4.186095505617978, + "grad_norm": 0.9709240794181824, + "learning_rate": 1.0566902871957584e-06, + "loss": 0.5456, + "step": 11922 + }, + { + "epoch": 4.186446629213483, + "grad_norm": 0.9520580172538757, + "learning_rate": 1.056312816253187e-06, + "loss": 0.5379, + "step": 11923 + }, + { + "epoch": 4.186797752808989, + "grad_norm": 0.9486163258552551, + "learning_rate": 1.0559353946834804e-06, + "loss": 0.5117, + "step": 11924 + }, + { + "epoch": 4.187148876404494, + "grad_norm": 0.9477207660675049, + "learning_rate": 1.0555580224995488e-06, + "loss": 0.5234, + "step": 11925 + }, + { + "epoch": 4.1875, + "grad_norm": 0.9590542912483215, + "learning_rate": 1.0551806997142957e-06, + "loss": 0.5287, + "step": 11926 + }, + { + "epoch": 4.187851123595506, + "grad_norm": 1.1737643480300903, + "learning_rate": 1.0548034263406275e-06, + "loss": 0.5289, + "step": 11927 + }, + { + "epoch": 4.188202247191011, + "grad_norm": 0.9814625978469849, + "learning_rate": 1.0544262023914444e-06, + "loss": 0.5463, + "step": 11928 + }, + { + "epoch": 4.188553370786517, + "grad_norm": 0.9348330497741699, + "learning_rate": 1.0540490278796493e-06, + "loss": 0.5334, + "step": 11929 + }, + { + "epoch": 4.188904494382022, + "grad_norm": 1.0037463903427124, + "learning_rate": 1.0536719028181389e-06, + "loss": 0.5501, + "step": 11930 + }, + { + "epoch": 4.189255617977528, + "grad_norm": 0.9765518307685852, + "learning_rate": 1.053294827219813e-06, + "loss": 0.5347, + "step": 11931 + }, + { + "epoch": 4.189606741573034, + "grad_norm": 0.9742205739021301, + "learning_rate": 1.0529178010975649e-06, + "loss": 0.5404, + "step": 11932 + }, + { + "epoch": 4.189957865168539, + "grad_norm": 1.010047197341919, + "learning_rate": 1.0525408244642907e-06, + "loss": 0.5538, + "step": 11933 + }, + { + "epoch": 4.190308988764045, + "grad_norm": 0.9514452815055847, + "learning_rate": 1.0521638973328815e-06, + "loss": 0.5106, + "step": 11934 + }, + { + "epoch": 4.19066011235955, + "grad_norm": 0.9846712350845337, + "learning_rate": 1.0517870197162274e-06, + "loss": 0.5471, + "step": 11935 + }, + { + "epoch": 4.191011235955056, + "grad_norm": 1.0013126134872437, + "learning_rate": 1.051410191627219e-06, + "loss": 0.5323, + "step": 11936 + }, + { + "epoch": 4.191362359550562, + "grad_norm": 0.971915602684021, + "learning_rate": 1.0510334130787414e-06, + "loss": 0.5666, + "step": 11937 + }, + { + "epoch": 4.191713483146067, + "grad_norm": 0.9734273552894592, + "learning_rate": 1.0506566840836821e-06, + "loss": 0.5486, + "step": 11938 + }, + { + "epoch": 4.192064606741573, + "grad_norm": 0.9631562232971191, + "learning_rate": 1.050280004654923e-06, + "loss": 0.5236, + "step": 11939 + }, + { + "epoch": 4.192415730337078, + "grad_norm": 0.9729178547859192, + "learning_rate": 1.0499033748053484e-06, + "loss": 0.5897, + "step": 11940 + }, + { + "epoch": 4.192766853932584, + "grad_norm": 0.9460442662239075, + "learning_rate": 1.0495267945478362e-06, + "loss": 0.5571, + "step": 11941 + }, + { + "epoch": 4.19311797752809, + "grad_norm": 0.9578481912612915, + "learning_rate": 1.0491502638952674e-06, + "loss": 0.5355, + "step": 11942 + }, + { + "epoch": 4.193469101123595, + "grad_norm": 0.9559242129325867, + "learning_rate": 1.0487737828605175e-06, + "loss": 0.5244, + "step": 11943 + }, + { + "epoch": 4.193820224719101, + "grad_norm": 0.9371334910392761, + "learning_rate": 1.048397351456463e-06, + "loss": 0.5, + "step": 11944 + }, + { + "epoch": 4.194171348314606, + "grad_norm": 0.9979387521743774, + "learning_rate": 1.0480209696959762e-06, + "loss": 0.5672, + "step": 11945 + }, + { + "epoch": 4.194522471910112, + "grad_norm": 0.9755778312683105, + "learning_rate": 1.0476446375919305e-06, + "loss": 0.5356, + "step": 11946 + }, + { + "epoch": 4.194873595505618, + "grad_norm": 0.9688037633895874, + "learning_rate": 1.0472683551571955e-06, + "loss": 0.5529, + "step": 11947 + }, + { + "epoch": 4.195224719101123, + "grad_norm": 0.9627313613891602, + "learning_rate": 1.0468921224046385e-06, + "loss": 0.5147, + "step": 11948 + }, + { + "epoch": 4.19557584269663, + "grad_norm": 0.9652628898620605, + "learning_rate": 1.0465159393471287e-06, + "loss": 0.5535, + "step": 11949 + }, + { + "epoch": 4.195926966292135, + "grad_norm": 0.9193136692047119, + "learning_rate": 1.046139805997529e-06, + "loss": 0.5206, + "step": 11950 + }, + { + "epoch": 4.196278089887641, + "grad_norm": 0.9910781383514404, + "learning_rate": 1.0457637223687047e-06, + "loss": 0.5533, + "step": 11951 + }, + { + "epoch": 4.196629213483146, + "grad_norm": 0.9678446650505066, + "learning_rate": 1.045387688473516e-06, + "loss": 0.5235, + "step": 11952 + }, + { + "epoch": 4.196980337078652, + "grad_norm": 0.9570243954658508, + "learning_rate": 1.0450117043248243e-06, + "loss": 0.5274, + "step": 11953 + }, + { + "epoch": 4.197331460674158, + "grad_norm": 0.928067684173584, + "learning_rate": 1.0446357699354862e-06, + "loss": 0.5063, + "step": 11954 + }, + { + "epoch": 4.197682584269663, + "grad_norm": 0.9675750732421875, + "learning_rate": 1.0442598853183605e-06, + "loss": 0.5158, + "step": 11955 + }, + { + "epoch": 4.198033707865169, + "grad_norm": 0.9753603339195251, + "learning_rate": 1.0438840504863e-06, + "loss": 0.5449, + "step": 11956 + }, + { + "epoch": 4.198384831460674, + "grad_norm": 1.008877158164978, + "learning_rate": 1.04350826545216e-06, + "loss": 0.5101, + "step": 11957 + }, + { + "epoch": 4.19873595505618, + "grad_norm": 0.968816339969635, + "learning_rate": 1.04313253022879e-06, + "loss": 0.5358, + "step": 11958 + }, + { + "epoch": 4.199087078651686, + "grad_norm": 0.977059006690979, + "learning_rate": 1.0427568448290421e-06, + "loss": 0.5667, + "step": 11959 + }, + { + "epoch": 4.199438202247191, + "grad_norm": 0.9959145188331604, + "learning_rate": 1.0423812092657626e-06, + "loss": 0.5434, + "step": 11960 + }, + { + "epoch": 4.199789325842697, + "grad_norm": 0.9762945175170898, + "learning_rate": 1.0420056235517978e-06, + "loss": 0.5267, + "step": 11961 + }, + { + "epoch": 4.200140449438202, + "grad_norm": 0.9812824726104736, + "learning_rate": 1.0416300876999938e-06, + "loss": 0.5521, + "step": 11962 + }, + { + "epoch": 4.200491573033708, + "grad_norm": 0.9620629549026489, + "learning_rate": 1.041254601723192e-06, + "loss": 0.5115, + "step": 11963 + }, + { + "epoch": 4.200842696629214, + "grad_norm": 0.9876679182052612, + "learning_rate": 1.0408791656342356e-06, + "loss": 0.5382, + "step": 11964 + }, + { + "epoch": 4.201193820224719, + "grad_norm": 0.9963091611862183, + "learning_rate": 1.040503779445962e-06, + "loss": 0.5627, + "step": 11965 + }, + { + "epoch": 4.201544943820225, + "grad_norm": 0.9676474928855896, + "learning_rate": 1.0401284431712114e-06, + "loss": 0.5441, + "step": 11966 + }, + { + "epoch": 4.20189606741573, + "grad_norm": 0.9420453310012817, + "learning_rate": 1.0397531568228179e-06, + "loss": 0.5266, + "step": 11967 + }, + { + "epoch": 4.202247191011236, + "grad_norm": 0.9545940160751343, + "learning_rate": 1.0393779204136175e-06, + "loss": 0.5243, + "step": 11968 + }, + { + "epoch": 4.202598314606742, + "grad_norm": 0.9663742184638977, + "learning_rate": 1.0390027339564419e-06, + "loss": 0.5507, + "step": 11969 + }, + { + "epoch": 4.202949438202247, + "grad_norm": 0.95212721824646, + "learning_rate": 1.038627597464123e-06, + "loss": 0.5078, + "step": 11970 + }, + { + "epoch": 4.203300561797753, + "grad_norm": 0.9642164707183838, + "learning_rate": 1.0382525109494892e-06, + "loss": 0.5621, + "step": 11971 + }, + { + "epoch": 4.203651685393258, + "grad_norm": 0.9640729427337646, + "learning_rate": 1.0378774744253694e-06, + "loss": 0.5483, + "step": 11972 + }, + { + "epoch": 4.204002808988764, + "grad_norm": 0.9514890313148499, + "learning_rate": 1.0375024879045889e-06, + "loss": 0.5582, + "step": 11973 + }, + { + "epoch": 4.20435393258427, + "grad_norm": 0.9541265368461609, + "learning_rate": 1.037127551399971e-06, + "loss": 0.5324, + "step": 11974 + }, + { + "epoch": 4.204705056179775, + "grad_norm": 0.94986891746521, + "learning_rate": 1.0367526649243397e-06, + "loss": 0.5576, + "step": 11975 + }, + { + "epoch": 4.205056179775281, + "grad_norm": 0.9843858480453491, + "learning_rate": 1.0363778284905146e-06, + "loss": 0.5553, + "step": 11976 + }, + { + "epoch": 4.205407303370786, + "grad_norm": 0.9874837398529053, + "learning_rate": 1.036003042111316e-06, + "loss": 0.5596, + "step": 11977 + }, + { + "epoch": 4.205758426966292, + "grad_norm": 0.9480567574501038, + "learning_rate": 1.0356283057995597e-06, + "loss": 0.5524, + "step": 11978 + }, + { + "epoch": 4.206109550561798, + "grad_norm": 0.944230854511261, + "learning_rate": 1.0352536195680631e-06, + "loss": 0.551, + "step": 11979 + }, + { + "epoch": 4.206460674157303, + "grad_norm": 0.9615346193313599, + "learning_rate": 1.0348789834296385e-06, + "loss": 0.5206, + "step": 11980 + }, + { + "epoch": 4.206811797752809, + "grad_norm": 0.9354829788208008, + "learning_rate": 1.0345043973971e-06, + "loss": 0.5388, + "step": 11981 + }, + { + "epoch": 4.207162921348314, + "grad_norm": 0.9424106478691101, + "learning_rate": 1.034129861483256e-06, + "loss": 0.5227, + "step": 11982 + }, + { + "epoch": 4.20751404494382, + "grad_norm": 0.9449310302734375, + "learning_rate": 1.0337553757009172e-06, + "loss": 0.5115, + "step": 11983 + }, + { + "epoch": 4.207865168539326, + "grad_norm": 0.9741902351379395, + "learning_rate": 1.033380940062889e-06, + "loss": 0.5649, + "step": 11984 + }, + { + "epoch": 4.208216292134831, + "grad_norm": 1.0037405490875244, + "learning_rate": 1.0330065545819787e-06, + "loss": 0.5456, + "step": 11985 + }, + { + "epoch": 4.208567415730337, + "grad_norm": 0.9915852546691895, + "learning_rate": 1.0326322192709889e-06, + "loss": 0.5525, + "step": 11986 + }, + { + "epoch": 4.208918539325842, + "grad_norm": 0.9537054896354675, + "learning_rate": 1.0322579341427205e-06, + "loss": 0.5474, + "step": 11987 + }, + { + "epoch": 4.209269662921348, + "grad_norm": 0.9720523357391357, + "learning_rate": 1.031883699209976e-06, + "loss": 0.541, + "step": 11988 + }, + { + "epoch": 4.209620786516854, + "grad_norm": 0.9590625166893005, + "learning_rate": 1.0315095144855515e-06, + "loss": 0.5347, + "step": 11989 + }, + { + "epoch": 4.209971910112359, + "grad_norm": 0.9825165867805481, + "learning_rate": 1.031135379982246e-06, + "loss": 0.5867, + "step": 11990 + }, + { + "epoch": 4.210323033707865, + "grad_norm": 0.9439297318458557, + "learning_rate": 1.0307612957128524e-06, + "loss": 0.5481, + "step": 11991 + }, + { + "epoch": 4.21067415730337, + "grad_norm": 0.964459240436554, + "learning_rate": 1.0303872616901664e-06, + "loss": 0.5404, + "step": 11992 + }, + { + "epoch": 4.211025280898877, + "grad_norm": 0.9664011001586914, + "learning_rate": 1.030013277926978e-06, + "loss": 0.5479, + "step": 11993 + }, + { + "epoch": 4.211376404494382, + "grad_norm": 0.9102189540863037, + "learning_rate": 1.0296393444360778e-06, + "loss": 0.5345, + "step": 11994 + }, + { + "epoch": 4.211727528089888, + "grad_norm": 0.9293334484100342, + "learning_rate": 1.0292654612302532e-06, + "loss": 0.5179, + "step": 11995 + }, + { + "epoch": 4.212078651685394, + "grad_norm": 0.9401870369911194, + "learning_rate": 1.0288916283222921e-06, + "loss": 0.5155, + "step": 11996 + }, + { + "epoch": 4.212429775280899, + "grad_norm": 0.936691403388977, + "learning_rate": 1.0285178457249787e-06, + "loss": 0.5226, + "step": 11997 + }, + { + "epoch": 4.212780898876405, + "grad_norm": 0.976426899433136, + "learning_rate": 1.0281441134510948e-06, + "loss": 0.5442, + "step": 11998 + }, + { + "epoch": 4.21313202247191, + "grad_norm": 0.9549325704574585, + "learning_rate": 1.0277704315134234e-06, + "loss": 0.5482, + "step": 11999 + }, + { + "epoch": 4.213483146067416, + "grad_norm": 0.9433548450469971, + "learning_rate": 1.0273967999247428e-06, + "loss": 0.5472, + "step": 12000 + }, + { + "epoch": 4.213834269662922, + "grad_norm": 0.9294977188110352, + "learning_rate": 1.0270232186978323e-06, + "loss": 0.5162, + "step": 12001 + }, + { + "epoch": 4.214185393258427, + "grad_norm": 0.9442880749702454, + "learning_rate": 1.0266496878454665e-06, + "loss": 0.5442, + "step": 12002 + }, + { + "epoch": 4.214536516853933, + "grad_norm": 0.9692766666412354, + "learning_rate": 1.0262762073804214e-06, + "loss": 0.5808, + "step": 12003 + }, + { + "epoch": 4.214887640449438, + "grad_norm": 0.9453266263008118, + "learning_rate": 1.0259027773154681e-06, + "loss": 0.5342, + "step": 12004 + }, + { + "epoch": 4.215238764044944, + "grad_norm": 1.0009287595748901, + "learning_rate": 1.0255293976633792e-06, + "loss": 0.5475, + "step": 12005 + }, + { + "epoch": 4.21558988764045, + "grad_norm": 0.9735552668571472, + "learning_rate": 1.0251560684369225e-06, + "loss": 0.5388, + "step": 12006 + }, + { + "epoch": 4.215941011235955, + "grad_norm": 0.98050457239151, + "learning_rate": 1.0247827896488669e-06, + "loss": 0.5541, + "step": 12007 + }, + { + "epoch": 4.216292134831461, + "grad_norm": 0.9864556193351746, + "learning_rate": 1.0244095613119768e-06, + "loss": 0.5584, + "step": 12008 + }, + { + "epoch": 4.216643258426966, + "grad_norm": 0.9478455781936646, + "learning_rate": 1.0240363834390176e-06, + "loss": 0.5431, + "step": 12009 + }, + { + "epoch": 4.216994382022472, + "grad_norm": 0.9750632047653198, + "learning_rate": 1.0236632560427514e-06, + "loss": 0.5367, + "step": 12010 + }, + { + "epoch": 4.217345505617978, + "grad_norm": 0.9553634524345398, + "learning_rate": 1.0232901791359376e-06, + "loss": 0.5636, + "step": 12011 + }, + { + "epoch": 4.217696629213483, + "grad_norm": 0.9504817128181458, + "learning_rate": 1.0229171527313369e-06, + "loss": 0.5413, + "step": 12012 + }, + { + "epoch": 4.218047752808989, + "grad_norm": 0.9573503732681274, + "learning_rate": 1.0225441768417046e-06, + "loss": 0.5564, + "step": 12013 + }, + { + "epoch": 4.218398876404494, + "grad_norm": 0.9700208306312561, + "learning_rate": 1.0221712514797981e-06, + "loss": 0.5652, + "step": 12014 + }, + { + "epoch": 4.21875, + "grad_norm": 0.9369280338287354, + "learning_rate": 1.0217983766583696e-06, + "loss": 0.5362, + "step": 12015 + }, + { + "epoch": 4.219101123595506, + "grad_norm": 0.9755404591560364, + "learning_rate": 1.0214255523901722e-06, + "loss": 0.5641, + "step": 12016 + }, + { + "epoch": 4.219452247191011, + "grad_norm": 0.9621213674545288, + "learning_rate": 1.0210527786879552e-06, + "loss": 0.5562, + "step": 12017 + }, + { + "epoch": 4.219803370786517, + "grad_norm": 0.9463129043579102, + "learning_rate": 1.0206800555644683e-06, + "loss": 0.569, + "step": 12018 + }, + { + "epoch": 4.220154494382022, + "grad_norm": 0.9618059396743774, + "learning_rate": 1.0203073830324566e-06, + "loss": 0.5127, + "step": 12019 + }, + { + "epoch": 4.220505617977528, + "grad_norm": 0.968964695930481, + "learning_rate": 1.0199347611046672e-06, + "loss": 0.5147, + "step": 12020 + }, + { + "epoch": 4.220856741573034, + "grad_norm": 0.9562776684761047, + "learning_rate": 1.0195621897938415e-06, + "loss": 0.5183, + "step": 12021 + }, + { + "epoch": 4.221207865168539, + "grad_norm": 0.950514554977417, + "learning_rate": 1.0191896691127232e-06, + "loss": 0.5193, + "step": 12022 + }, + { + "epoch": 4.221558988764045, + "grad_norm": 0.9560863971710205, + "learning_rate": 1.018817199074051e-06, + "loss": 0.548, + "step": 12023 + }, + { + "epoch": 4.22191011235955, + "grad_norm": 0.9391663074493408, + "learning_rate": 1.018444779690562e-06, + "loss": 0.512, + "step": 12024 + }, + { + "epoch": 4.222261235955056, + "grad_norm": 0.9772560000419617, + "learning_rate": 1.0180724109749948e-06, + "loss": 0.5215, + "step": 12025 + }, + { + "epoch": 4.222612359550562, + "grad_norm": 0.9842314720153809, + "learning_rate": 1.0177000929400822e-06, + "loss": 0.5526, + "step": 12026 + }, + { + "epoch": 4.222963483146067, + "grad_norm": 0.9525455236434937, + "learning_rate": 1.0173278255985588e-06, + "loss": 0.5472, + "step": 12027 + }, + { + "epoch": 4.223314606741573, + "grad_norm": 0.9769579172134399, + "learning_rate": 1.016955608963154e-06, + "loss": 0.5617, + "step": 12028 + }, + { + "epoch": 4.223665730337078, + "grad_norm": 0.9125334024429321, + "learning_rate": 1.0165834430465996e-06, + "loss": 0.4887, + "step": 12029 + }, + { + "epoch": 4.224016853932584, + "grad_norm": 0.9707804322242737, + "learning_rate": 1.016211327861621e-06, + "loss": 0.5544, + "step": 12030 + }, + { + "epoch": 4.22436797752809, + "grad_norm": 0.9723607301712036, + "learning_rate": 1.015839263420946e-06, + "loss": 0.5514, + "step": 12031 + }, + { + "epoch": 4.224719101123595, + "grad_norm": 0.9371212124824524, + "learning_rate": 1.0154672497372972e-06, + "loss": 0.5196, + "step": 12032 + }, + { + "epoch": 4.225070224719101, + "grad_norm": 0.9417607188224792, + "learning_rate": 1.0150952868233993e-06, + "loss": 0.548, + "step": 12033 + }, + { + "epoch": 4.225421348314606, + "grad_norm": 0.9609116911888123, + "learning_rate": 1.014723374691971e-06, + "loss": 0.5134, + "step": 12034 + }, + { + "epoch": 4.225772471910112, + "grad_norm": 0.9785856604576111, + "learning_rate": 1.0143515133557333e-06, + "loss": 0.5426, + "step": 12035 + }, + { + "epoch": 4.226123595505618, + "grad_norm": 0.9839282035827637, + "learning_rate": 1.0139797028274023e-06, + "loss": 0.5433, + "step": 12036 + }, + { + "epoch": 4.226474719101123, + "grad_norm": 0.9570432305335999, + "learning_rate": 1.0136079431196932e-06, + "loss": 0.5611, + "step": 12037 + }, + { + "epoch": 4.22682584269663, + "grad_norm": 0.9816099405288696, + "learning_rate": 1.0132362342453213e-06, + "loss": 0.5223, + "step": 12038 + }, + { + "epoch": 4.227176966292135, + "grad_norm": 0.9681476354598999, + "learning_rate": 1.012864576216997e-06, + "loss": 0.5312, + "step": 12039 + }, + { + "epoch": 4.227528089887641, + "grad_norm": 0.9821230173110962, + "learning_rate": 1.0124929690474328e-06, + "loss": 0.5511, + "step": 12040 + }, + { + "epoch": 4.227879213483146, + "grad_norm": 0.9763071537017822, + "learning_rate": 1.012121412749335e-06, + "loss": 0.5418, + "step": 12041 + }, + { + "epoch": 4.228230337078652, + "grad_norm": 0.9561235904693604, + "learning_rate": 1.0117499073354126e-06, + "loss": 0.5165, + "step": 12042 + }, + { + "epoch": 4.228581460674158, + "grad_norm": 0.98116135597229, + "learning_rate": 1.011378452818369e-06, + "loss": 0.5954, + "step": 12043 + }, + { + "epoch": 4.228932584269663, + "grad_norm": 0.9785438179969788, + "learning_rate": 1.0110070492109095e-06, + "loss": 0.5486, + "step": 12044 + }, + { + "epoch": 4.229283707865169, + "grad_norm": 0.9784141182899475, + "learning_rate": 1.0106356965257339e-06, + "loss": 0.5187, + "step": 12045 + }, + { + "epoch": 4.229634831460674, + "grad_norm": 0.9479644298553467, + "learning_rate": 1.0102643947755437e-06, + "loss": 0.5068, + "step": 12046 + }, + { + "epoch": 4.22998595505618, + "grad_norm": 0.9594654440879822, + "learning_rate": 1.0098931439730355e-06, + "loss": 0.5362, + "step": 12047 + }, + { + "epoch": 4.230337078651686, + "grad_norm": 0.9383675456047058, + "learning_rate": 1.0095219441309077e-06, + "loss": 0.5048, + "step": 12048 + }, + { + "epoch": 4.230688202247191, + "grad_norm": 0.9732776284217834, + "learning_rate": 1.009150795261854e-06, + "loss": 0.5218, + "step": 12049 + }, + { + "epoch": 4.231039325842697, + "grad_norm": 0.9521974325180054, + "learning_rate": 1.0087796973785663e-06, + "loss": 0.5554, + "step": 12050 + }, + { + "epoch": 4.231390449438202, + "grad_norm": 0.9627708196640015, + "learning_rate": 1.008408650493738e-06, + "loss": 0.5523, + "step": 12051 + }, + { + "epoch": 4.231741573033708, + "grad_norm": 0.9919725656509399, + "learning_rate": 1.0080376546200566e-06, + "loss": 0.5535, + "step": 12052 + }, + { + "epoch": 4.232092696629214, + "grad_norm": 0.9991096258163452, + "learning_rate": 1.0076667097702114e-06, + "loss": 0.5416, + "step": 12053 + }, + { + "epoch": 4.232443820224719, + "grad_norm": 0.9273778796195984, + "learning_rate": 1.0072958159568868e-06, + "loss": 0.489, + "step": 12054 + }, + { + "epoch": 4.232794943820225, + "grad_norm": 0.9903126955032349, + "learning_rate": 1.0069249731927691e-06, + "loss": 0.5716, + "step": 12055 + }, + { + "epoch": 4.23314606741573, + "grad_norm": 0.9638489484786987, + "learning_rate": 1.0065541814905387e-06, + "loss": 0.5565, + "step": 12056 + }, + { + "epoch": 4.233497191011236, + "grad_norm": 0.9637631773948669, + "learning_rate": 1.0061834408628782e-06, + "loss": 0.4986, + "step": 12057 + }, + { + "epoch": 4.233848314606742, + "grad_norm": 0.9610356688499451, + "learning_rate": 1.005812751322465e-06, + "loss": 0.5573, + "step": 12058 + }, + { + "epoch": 4.234199438202247, + "grad_norm": 0.9600244760513306, + "learning_rate": 1.0054421128819777e-06, + "loss": 0.546, + "step": 12059 + }, + { + "epoch": 4.234550561797753, + "grad_norm": 0.9890919327735901, + "learning_rate": 1.0050715255540915e-06, + "loss": 0.5497, + "step": 12060 + }, + { + "epoch": 4.234901685393258, + "grad_norm": 0.9436578154563904, + "learning_rate": 1.0047009893514788e-06, + "loss": 0.5554, + "step": 12061 + }, + { + "epoch": 4.235252808988764, + "grad_norm": 0.9864017963409424, + "learning_rate": 1.0043305042868138e-06, + "loss": 0.5627, + "step": 12062 + }, + { + "epoch": 4.23560393258427, + "grad_norm": 0.9991886615753174, + "learning_rate": 1.0039600703727648e-06, + "loss": 0.5301, + "step": 12063 + }, + { + "epoch": 4.235955056179775, + "grad_norm": 0.9468226432800293, + "learning_rate": 1.003589687622002e-06, + "loss": 0.5301, + "step": 12064 + }, + { + "epoch": 4.236306179775281, + "grad_norm": 0.9595242142677307, + "learning_rate": 1.0032193560471904e-06, + "loss": 0.5488, + "step": 12065 + }, + { + "epoch": 4.236657303370786, + "grad_norm": 0.9664261937141418, + "learning_rate": 1.0028490756609972e-06, + "loss": 0.5421, + "step": 12066 + }, + { + "epoch": 4.237008426966292, + "grad_norm": 0.9636686444282532, + "learning_rate": 1.0024788464760842e-06, + "loss": 0.536, + "step": 12067 + }, + { + "epoch": 4.237359550561798, + "grad_norm": 0.9479982852935791, + "learning_rate": 1.0021086685051124e-06, + "loss": 0.547, + "step": 12068 + }, + { + "epoch": 4.237710674157303, + "grad_norm": 0.955987274646759, + "learning_rate": 1.0017385417607434e-06, + "loss": 0.5631, + "step": 12069 + }, + { + "epoch": 4.238061797752809, + "grad_norm": 0.962559700012207, + "learning_rate": 1.001368466255633e-06, + "loss": 0.5613, + "step": 12070 + }, + { + "epoch": 4.238412921348314, + "grad_norm": 0.932537853717804, + "learning_rate": 1.0009984420024396e-06, + "loss": 0.5171, + "step": 12071 + }, + { + "epoch": 4.23876404494382, + "grad_norm": 0.9734734296798706, + "learning_rate": 1.0006284690138162e-06, + "loss": 0.5368, + "step": 12072 + }, + { + "epoch": 4.239115168539326, + "grad_norm": 0.9820423126220703, + "learning_rate": 1.0002585473024169e-06, + "loss": 0.5545, + "step": 12073 + }, + { + "epoch": 4.239466292134831, + "grad_norm": 0.9697520136833191, + "learning_rate": 9.998886768808907e-07, + "loss": 0.5592, + "step": 12074 + }, + { + "epoch": 4.239817415730337, + "grad_norm": 0.9403982162475586, + "learning_rate": 9.995188577618894e-07, + "loss": 0.5127, + "step": 12075 + }, + { + "epoch": 4.240168539325842, + "grad_norm": 0.9671099781990051, + "learning_rate": 9.991490899580588e-07, + "loss": 0.534, + "step": 12076 + }, + { + "epoch": 4.240519662921348, + "grad_norm": 0.9688689708709717, + "learning_rate": 9.987793734820444e-07, + "loss": 0.5332, + "step": 12077 + }, + { + "epoch": 4.240870786516854, + "grad_norm": 0.956724226474762, + "learning_rate": 9.984097083464919e-07, + "loss": 0.4979, + "step": 12078 + }, + { + "epoch": 4.241221910112359, + "grad_norm": 0.952569305896759, + "learning_rate": 9.980400945640411e-07, + "loss": 0.5551, + "step": 12079 + }, + { + "epoch": 4.241573033707865, + "grad_norm": 0.9824742078781128, + "learning_rate": 9.97670532147335e-07, + "loss": 0.5952, + "step": 12080 + }, + { + "epoch": 4.24192415730337, + "grad_norm": 0.9644066095352173, + "learning_rate": 9.973010211090103e-07, + "loss": 0.5401, + "step": 12081 + }, + { + "epoch": 4.242275280898877, + "grad_norm": 0.9289160370826721, + "learning_rate": 9.969315614617056e-07, + "loss": 0.5169, + "step": 12082 + }, + { + "epoch": 4.242626404494382, + "grad_norm": 0.9435855746269226, + "learning_rate": 9.965621532180552e-07, + "loss": 0.5125, + "step": 12083 + }, + { + "epoch": 4.242977528089888, + "grad_norm": 0.9597464799880981, + "learning_rate": 9.961927963906916e-07, + "loss": 0.4968, + "step": 12084 + }, + { + "epoch": 4.243328651685394, + "grad_norm": 0.9404835104942322, + "learning_rate": 9.95823490992249e-07, + "loss": 0.5361, + "step": 12085 + }, + { + "epoch": 4.243679775280899, + "grad_norm": 0.9394482374191284, + "learning_rate": 9.954542370353545e-07, + "loss": 0.5601, + "step": 12086 + }, + { + "epoch": 4.244030898876405, + "grad_norm": 0.9910191297531128, + "learning_rate": 9.950850345326388e-07, + "loss": 0.5267, + "step": 12087 + }, + { + "epoch": 4.24438202247191, + "grad_norm": 0.9620302319526672, + "learning_rate": 9.94715883496726e-07, + "loss": 0.5453, + "step": 12088 + }, + { + "epoch": 4.244733146067416, + "grad_norm": 0.9376868605613708, + "learning_rate": 9.94346783940243e-07, + "loss": 0.539, + "step": 12089 + }, + { + "epoch": 4.245084269662922, + "grad_norm": 0.991488516330719, + "learning_rate": 9.939777358758115e-07, + "loss": 0.5268, + "step": 12090 + }, + { + "epoch": 4.245435393258427, + "grad_norm": 0.9334138631820679, + "learning_rate": 9.936087393160519e-07, + "loss": 0.5524, + "step": 12091 + }, + { + "epoch": 4.245786516853933, + "grad_norm": 0.9720402359962463, + "learning_rate": 9.932397942735849e-07, + "loss": 0.5612, + "step": 12092 + }, + { + "epoch": 4.246137640449438, + "grad_norm": 0.9742700457572937, + "learning_rate": 9.92870900761027e-07, + "loss": 0.5563, + "step": 12093 + }, + { + "epoch": 4.246488764044944, + "grad_norm": 0.980839192867279, + "learning_rate": 9.925020587909957e-07, + "loss": 0.5323, + "step": 12094 + }, + { + "epoch": 4.24683988764045, + "grad_norm": 0.9563094973564148, + "learning_rate": 9.92133268376103e-07, + "loss": 0.5264, + "step": 12095 + }, + { + "epoch": 4.247191011235955, + "grad_norm": 0.96231609582901, + "learning_rate": 9.917645295289627e-07, + "loss": 0.5623, + "step": 12096 + }, + { + "epoch": 4.247542134831461, + "grad_norm": 0.9706680774688721, + "learning_rate": 9.913958422621843e-07, + "loss": 0.5424, + "step": 12097 + }, + { + "epoch": 4.247893258426966, + "grad_norm": 0.9656089544296265, + "learning_rate": 9.91027206588378e-07, + "loss": 0.5577, + "step": 12098 + }, + { + "epoch": 4.248244382022472, + "grad_norm": 0.91878741979599, + "learning_rate": 9.9065862252015e-07, + "loss": 0.5718, + "step": 12099 + }, + { + "epoch": 4.248595505617978, + "grad_norm": 0.9344439506530762, + "learning_rate": 9.902900900701048e-07, + "loss": 0.5253, + "step": 12100 + }, + { + "epoch": 4.248946629213483, + "grad_norm": 0.9415321350097656, + "learning_rate": 9.899216092508473e-07, + "loss": 0.5361, + "step": 12101 + }, + { + "epoch": 4.249297752808989, + "grad_norm": 0.9527208209037781, + "learning_rate": 9.89553180074978e-07, + "loss": 0.5382, + "step": 12102 + }, + { + "epoch": 4.249648876404494, + "grad_norm": 0.9665458798408508, + "learning_rate": 9.89184802555098e-07, + "loss": 0.5667, + "step": 12103 + }, + { + "epoch": 4.25, + "grad_norm": 0.957638144493103, + "learning_rate": 9.888164767038041e-07, + "loss": 0.5628, + "step": 12104 + }, + { + "epoch": 4.250351123595506, + "grad_norm": 0.9924194812774658, + "learning_rate": 9.884482025336948e-07, + "loss": 0.5646, + "step": 12105 + }, + { + "epoch": 4.250702247191011, + "grad_norm": 0.9569982886314392, + "learning_rate": 9.88079980057362e-07, + "loss": 0.553, + "step": 12106 + }, + { + "epoch": 4.251053370786517, + "grad_norm": 0.984036386013031, + "learning_rate": 9.877118092874017e-07, + "loss": 0.5411, + "step": 12107 + }, + { + "epoch": 4.251404494382022, + "grad_norm": 0.9766518473625183, + "learning_rate": 9.87343690236402e-07, + "loss": 0.5442, + "step": 12108 + }, + { + "epoch": 4.251755617977528, + "grad_norm": 1.0098093748092651, + "learning_rate": 9.869756229169548e-07, + "loss": 0.5216, + "step": 12109 + }, + { + "epoch": 4.252106741573034, + "grad_norm": 0.9779828190803528, + "learning_rate": 9.866076073416456e-07, + "loss": 0.5212, + "step": 12110 + }, + { + "epoch": 4.252457865168539, + "grad_norm": 1.005969762802124, + "learning_rate": 9.862396435230622e-07, + "loss": 0.5626, + "step": 12111 + }, + { + "epoch": 4.252808988764045, + "grad_norm": 0.9667978882789612, + "learning_rate": 9.858717314737875e-07, + "loss": 0.5177, + "step": 12112 + }, + { + "epoch": 4.25316011235955, + "grad_norm": 0.9501930475234985, + "learning_rate": 9.855038712064032e-07, + "loss": 0.4781, + "step": 12113 + }, + { + "epoch": 4.253511235955056, + "grad_norm": 0.9569604992866516, + "learning_rate": 9.851360627334916e-07, + "loss": 0.5409, + "step": 12114 + }, + { + "epoch": 4.253862359550562, + "grad_norm": 0.9463083148002625, + "learning_rate": 9.847683060676291e-07, + "loss": 0.5632, + "step": 12115 + }, + { + "epoch": 4.254213483146067, + "grad_norm": 0.9843915700912476, + "learning_rate": 9.84400601221395e-07, + "loss": 0.5615, + "step": 12116 + }, + { + "epoch": 4.254564606741573, + "grad_norm": 0.9753740429878235, + "learning_rate": 9.840329482073625e-07, + "loss": 0.5411, + "step": 12117 + }, + { + "epoch": 4.254915730337078, + "grad_norm": 0.9815272092819214, + "learning_rate": 9.83665347038107e-07, + "loss": 0.5556, + "step": 12118 + }, + { + "epoch": 4.255266853932584, + "grad_norm": 0.9610508680343628, + "learning_rate": 9.83297797726198e-07, + "loss": 0.5331, + "step": 12119 + }, + { + "epoch": 4.25561797752809, + "grad_norm": 1.019180178642273, + "learning_rate": 9.829303002842072e-07, + "loss": 0.5372, + "step": 12120 + }, + { + "epoch": 4.255969101123595, + "grad_norm": 1.0050299167633057, + "learning_rate": 9.825628547247013e-07, + "loss": 0.4925, + "step": 12121 + }, + { + "epoch": 4.256320224719101, + "grad_norm": 0.9625013470649719, + "learning_rate": 9.82195461060248e-07, + "loss": 0.5492, + "step": 12122 + }, + { + "epoch": 4.256671348314606, + "grad_norm": 0.9532650113105774, + "learning_rate": 9.818281193034113e-07, + "loss": 0.518, + "step": 12123 + }, + { + "epoch": 4.257022471910112, + "grad_norm": 0.9721881151199341, + "learning_rate": 9.81460829466753e-07, + "loss": 0.5509, + "step": 12124 + }, + { + "epoch": 4.257373595505618, + "grad_norm": 0.9754394292831421, + "learning_rate": 9.810935915628356e-07, + "loss": 0.5588, + "step": 12125 + }, + { + "epoch": 4.257724719101123, + "grad_norm": 0.9533361196517944, + "learning_rate": 9.807264056042167e-07, + "loss": 0.541, + "step": 12126 + }, + { + "epoch": 4.258075842696629, + "grad_norm": 0.9407110810279846, + "learning_rate": 9.803592716034556e-07, + "loss": 0.5236, + "step": 12127 + }, + { + "epoch": 4.258426966292135, + "grad_norm": 0.9510982632637024, + "learning_rate": 9.799921895731062e-07, + "loss": 0.5349, + "step": 12128 + }, + { + "epoch": 4.258778089887641, + "grad_norm": 0.9578803777694702, + "learning_rate": 9.796251595257241e-07, + "loss": 0.5133, + "step": 12129 + }, + { + "epoch": 4.259129213483146, + "grad_norm": 0.9565126299858093, + "learning_rate": 9.792581814738598e-07, + "loss": 0.545, + "step": 12130 + }, + { + "epoch": 4.259480337078652, + "grad_norm": 0.9521830081939697, + "learning_rate": 9.78891255430065e-07, + "loss": 0.5152, + "step": 12131 + }, + { + "epoch": 4.259831460674158, + "grad_norm": 0.9798305034637451, + "learning_rate": 9.785243814068867e-07, + "loss": 0.5066, + "step": 12132 + }, + { + "epoch": 4.260182584269663, + "grad_norm": 0.9683191180229187, + "learning_rate": 9.781575594168737e-07, + "loss": 0.5275, + "step": 12133 + }, + { + "epoch": 4.260533707865169, + "grad_norm": 0.9992858171463013, + "learning_rate": 9.777907894725691e-07, + "loss": 0.5311, + "step": 12134 + }, + { + "epoch": 4.260884831460674, + "grad_norm": 0.9697720408439636, + "learning_rate": 9.774240715865177e-07, + "loss": 0.5281, + "step": 12135 + }, + { + "epoch": 4.26123595505618, + "grad_norm": 0.9902812838554382, + "learning_rate": 9.7705740577126e-07, + "loss": 0.5559, + "step": 12136 + }, + { + "epoch": 4.261587078651686, + "grad_norm": 0.9708535671234131, + "learning_rate": 9.76690792039335e-07, + "loss": 0.5521, + "step": 12137 + }, + { + "epoch": 4.261938202247191, + "grad_norm": 0.9953432083129883, + "learning_rate": 9.763242304032824e-07, + "loss": 0.5854, + "step": 12138 + }, + { + "epoch": 4.262289325842697, + "grad_norm": 0.9613777995109558, + "learning_rate": 9.759577208756366e-07, + "loss": 0.5586, + "step": 12139 + }, + { + "epoch": 4.262640449438202, + "grad_norm": 0.9466422200202942, + "learning_rate": 9.755912634689333e-07, + "loss": 0.5286, + "step": 12140 + }, + { + "epoch": 4.262991573033708, + "grad_norm": 0.9622014760971069, + "learning_rate": 9.752248581957033e-07, + "loss": 0.5591, + "step": 12141 + }, + { + "epoch": 4.263342696629214, + "grad_norm": 0.9929337501525879, + "learning_rate": 9.748585050684791e-07, + "loss": 0.5467, + "step": 12142 + }, + { + "epoch": 4.263693820224719, + "grad_norm": 0.9867035150527954, + "learning_rate": 9.744922040997884e-07, + "loss": 0.5274, + "step": 12143 + }, + { + "epoch": 4.264044943820225, + "grad_norm": 0.9612558484077454, + "learning_rate": 9.741259553021599e-07, + "loss": 0.5757, + "step": 12144 + }, + { + "epoch": 4.26439606741573, + "grad_norm": 0.9748103022575378, + "learning_rate": 9.73759758688117e-07, + "loss": 0.5477, + "step": 12145 + }, + { + "epoch": 4.264747191011236, + "grad_norm": 0.9668884873390198, + "learning_rate": 9.733936142701853e-07, + "loss": 0.5207, + "step": 12146 + }, + { + "epoch": 4.265098314606742, + "grad_norm": 0.9362105131149292, + "learning_rate": 9.730275220608844e-07, + "loss": 0.5251, + "step": 12147 + }, + { + "epoch": 4.265449438202247, + "grad_norm": 0.9550386667251587, + "learning_rate": 9.726614820727367e-07, + "loss": 0.5601, + "step": 12148 + }, + { + "epoch": 4.265800561797753, + "grad_norm": 1.0099519491195679, + "learning_rate": 9.722954943182597e-07, + "loss": 0.5383, + "step": 12149 + }, + { + "epoch": 4.266151685393258, + "grad_norm": 0.9740354418754578, + "learning_rate": 9.71929558809968e-07, + "loss": 0.5729, + "step": 12150 + }, + { + "epoch": 4.266502808988764, + "grad_norm": 0.93415766954422, + "learning_rate": 9.715636755603792e-07, + "loss": 0.5534, + "step": 12151 + }, + { + "epoch": 4.26685393258427, + "grad_norm": 0.9678096175193787, + "learning_rate": 9.711978445820039e-07, + "loss": 0.5344, + "step": 12152 + }, + { + "epoch": 4.267205056179775, + "grad_norm": 0.9728831052780151, + "learning_rate": 9.708320658873547e-07, + "loss": 0.5787, + "step": 12153 + }, + { + "epoch": 4.267556179775281, + "grad_norm": 0.9662930965423584, + "learning_rate": 9.704663394889396e-07, + "loss": 0.571, + "step": 12154 + }, + { + "epoch": 4.267907303370786, + "grad_norm": 0.9606423377990723, + "learning_rate": 9.701006653992676e-07, + "loss": 0.5487, + "step": 12155 + }, + { + "epoch": 4.268258426966292, + "grad_norm": 0.9602417945861816, + "learning_rate": 9.697350436308428e-07, + "loss": 0.537, + "step": 12156 + }, + { + "epoch": 4.268609550561798, + "grad_norm": 0.9863374829292297, + "learning_rate": 9.693694741961711e-07, + "loss": 0.5616, + "step": 12157 + }, + { + "epoch": 4.268960674157303, + "grad_norm": 0.9549907445907593, + "learning_rate": 9.690039571077526e-07, + "loss": 0.5271, + "step": 12158 + }, + { + "epoch": 4.269311797752809, + "grad_norm": 0.9635226130485535, + "learning_rate": 9.686384923780895e-07, + "loss": 0.563, + "step": 12159 + }, + { + "epoch": 4.269662921348314, + "grad_norm": 0.9742010831832886, + "learning_rate": 9.682730800196788e-07, + "loss": 0.5511, + "step": 12160 + }, + { + "epoch": 4.27001404494382, + "grad_norm": 0.9541518092155457, + "learning_rate": 9.67907720045019e-07, + "loss": 0.5209, + "step": 12161 + }, + { + "epoch": 4.270365168539326, + "grad_norm": 0.9614741802215576, + "learning_rate": 9.675424124666041e-07, + "loss": 0.5389, + "step": 12162 + }, + { + "epoch": 4.270716292134831, + "grad_norm": 0.9687471985816956, + "learning_rate": 9.671771572969262e-07, + "loss": 0.5413, + "step": 12163 + }, + { + "epoch": 4.271067415730337, + "grad_norm": 0.9562928080558777, + "learning_rate": 9.668119545484792e-07, + "loss": 0.5541, + "step": 12164 + }, + { + "epoch": 4.271418539325842, + "grad_norm": 0.9549000859260559, + "learning_rate": 9.6644680423375e-07, + "loss": 0.539, + "step": 12165 + }, + { + "epoch": 4.271769662921348, + "grad_norm": 0.9614023566246033, + "learning_rate": 9.660817063652293e-07, + "loss": 0.5204, + "step": 12166 + }, + { + "epoch": 4.272120786516854, + "grad_norm": 0.9409894943237305, + "learning_rate": 9.657166609554002e-07, + "loss": 0.4869, + "step": 12167 + }, + { + "epoch": 4.272471910112359, + "grad_norm": 0.9689472913742065, + "learning_rate": 9.653516680167493e-07, + "loss": 0.5459, + "step": 12168 + }, + { + "epoch": 4.272823033707865, + "grad_norm": 0.9697874784469604, + "learning_rate": 9.649867275617579e-07, + "loss": 0.5534, + "step": 12169 + }, + { + "epoch": 4.273174157303371, + "grad_norm": 1.0002591609954834, + "learning_rate": 9.64621839602907e-07, + "loss": 0.5192, + "step": 12170 + }, + { + "epoch": 4.273525280898877, + "grad_norm": 0.952660322189331, + "learning_rate": 9.642570041526746e-07, + "loss": 0.5511, + "step": 12171 + }, + { + "epoch": 4.273876404494382, + "grad_norm": 0.9908432960510254, + "learning_rate": 9.638922212235397e-07, + "loss": 0.5649, + "step": 12172 + }, + { + "epoch": 4.274227528089888, + "grad_norm": 0.9364864230155945, + "learning_rate": 9.635274908279755e-07, + "loss": 0.5376, + "step": 12173 + }, + { + "epoch": 4.274578651685394, + "grad_norm": 0.9374816417694092, + "learning_rate": 9.63162812978457e-07, + "loss": 0.5154, + "step": 12174 + }, + { + "epoch": 4.274929775280899, + "grad_norm": 0.923183798789978, + "learning_rate": 9.627981876874553e-07, + "loss": 0.5221, + "step": 12175 + }, + { + "epoch": 4.275280898876405, + "grad_norm": 0.9728631973266602, + "learning_rate": 9.624336149674394e-07, + "loss": 0.5338, + "step": 12176 + }, + { + "epoch": 4.27563202247191, + "grad_norm": 0.9968119859695435, + "learning_rate": 9.620690948308785e-07, + "loss": 0.5397, + "step": 12177 + }, + { + "epoch": 4.275983146067416, + "grad_norm": 1.006558895111084, + "learning_rate": 9.617046272902384e-07, + "loss": 0.5572, + "step": 12178 + }, + { + "epoch": 4.276334269662922, + "grad_norm": 0.9826293587684631, + "learning_rate": 9.61340212357984e-07, + "loss": 0.5324, + "step": 12179 + }, + { + "epoch": 4.276685393258427, + "grad_norm": 0.9578385353088379, + "learning_rate": 9.609758500465771e-07, + "loss": 0.5293, + "step": 12180 + }, + { + "epoch": 4.277036516853933, + "grad_norm": 0.9563983678817749, + "learning_rate": 9.606115403684802e-07, + "loss": 0.6008, + "step": 12181 + }, + { + "epoch": 4.277387640449438, + "grad_norm": 0.9375402927398682, + "learning_rate": 9.602472833361501e-07, + "loss": 0.5287, + "step": 12182 + }, + { + "epoch": 4.277738764044944, + "grad_norm": 0.9420490860939026, + "learning_rate": 9.598830789620464e-07, + "loss": 0.5405, + "step": 12183 + }, + { + "epoch": 4.27808988764045, + "grad_norm": 0.9644142389297485, + "learning_rate": 9.595189272586225e-07, + "loss": 0.5605, + "step": 12184 + }, + { + "epoch": 4.278441011235955, + "grad_norm": 0.9757639169692993, + "learning_rate": 9.591548282383342e-07, + "loss": 0.5545, + "step": 12185 + }, + { + "epoch": 4.278792134831461, + "grad_norm": 1.0038715600967407, + "learning_rate": 9.587907819136321e-07, + "loss": 0.5415, + "step": 12186 + }, + { + "epoch": 4.279143258426966, + "grad_norm": 1.023634910583496, + "learning_rate": 9.584267882969655e-07, + "loss": 0.5757, + "step": 12187 + }, + { + "epoch": 4.279494382022472, + "grad_norm": 0.9729552865028381, + "learning_rate": 9.580628474007844e-07, + "loss": 0.5292, + "step": 12188 + }, + { + "epoch": 4.279845505617978, + "grad_norm": 1.0109634399414062, + "learning_rate": 9.576989592375338e-07, + "loss": 0.5591, + "step": 12189 + }, + { + "epoch": 4.280196629213483, + "grad_norm": 0.9636884927749634, + "learning_rate": 9.5733512381966e-07, + "loss": 0.5047, + "step": 12190 + }, + { + "epoch": 4.280547752808989, + "grad_norm": 0.9373121857643127, + "learning_rate": 9.569713411596037e-07, + "loss": 0.528, + "step": 12191 + }, + { + "epoch": 4.280898876404494, + "grad_norm": 0.9310202598571777, + "learning_rate": 9.56607611269808e-07, + "loss": 0.5268, + "step": 12192 + }, + { + "epoch": 4.28125, + "grad_norm": 0.9454798102378845, + "learning_rate": 9.562439341627108e-07, + "loss": 0.5623, + "step": 12193 + }, + { + "epoch": 4.281601123595506, + "grad_norm": 0.9748424291610718, + "learning_rate": 9.558803098507504e-07, + "loss": 0.6079, + "step": 12194 + }, + { + "epoch": 4.281952247191011, + "grad_norm": 0.9231148958206177, + "learning_rate": 9.555167383463617e-07, + "loss": 0.5126, + "step": 12195 + }, + { + "epoch": 4.282303370786517, + "grad_norm": 0.9463061690330505, + "learning_rate": 9.551532196619795e-07, + "loss": 0.5593, + "step": 12196 + }, + { + "epoch": 4.282654494382022, + "grad_norm": 0.9534937739372253, + "learning_rate": 9.547897538100345e-07, + "loss": 0.5676, + "step": 12197 + }, + { + "epoch": 4.283005617977528, + "grad_norm": 0.9542872905731201, + "learning_rate": 9.544263408029586e-07, + "loss": 0.5266, + "step": 12198 + }, + { + "epoch": 4.283356741573034, + "grad_norm": 0.9882010221481323, + "learning_rate": 9.54062980653179e-07, + "loss": 0.5623, + "step": 12199 + }, + { + "epoch": 4.283707865168539, + "grad_norm": 1.0666993856430054, + "learning_rate": 9.536996733731219e-07, + "loss": 0.5308, + "step": 12200 + }, + { + "epoch": 4.284058988764045, + "grad_norm": 0.941495418548584, + "learning_rate": 9.533364189752137e-07, + "loss": 0.5294, + "step": 12201 + }, + { + "epoch": 4.28441011235955, + "grad_norm": 0.9547412395477295, + "learning_rate": 9.529732174718756e-07, + "loss": 0.5399, + "step": 12202 + }, + { + "epoch": 4.284761235955056, + "grad_norm": 0.9381307363510132, + "learning_rate": 9.526100688755305e-07, + "loss": 0.5355, + "step": 12203 + }, + { + "epoch": 4.285112359550562, + "grad_norm": 0.962881326675415, + "learning_rate": 9.522469731985959e-07, + "loss": 0.5246, + "step": 12204 + }, + { + "epoch": 4.285463483146067, + "grad_norm": 0.9748377799987793, + "learning_rate": 9.518839304534913e-07, + "loss": 0.5479, + "step": 12205 + }, + { + "epoch": 4.285814606741573, + "grad_norm": 0.9694100022315979, + "learning_rate": 9.515209406526307e-07, + "loss": 0.5807, + "step": 12206 + }, + { + "epoch": 4.286165730337078, + "grad_norm": 0.9442321062088013, + "learning_rate": 9.511580038084297e-07, + "loss": 0.5251, + "step": 12207 + }, + { + "epoch": 4.286516853932584, + "grad_norm": 0.9588233232498169, + "learning_rate": 9.507951199332991e-07, + "loss": 0.5432, + "step": 12208 + }, + { + "epoch": 4.28686797752809, + "grad_norm": 0.9826153516769409, + "learning_rate": 9.504322890396503e-07, + "loss": 0.5473, + "step": 12209 + }, + { + "epoch": 4.287219101123595, + "grad_norm": 0.9605292081832886, + "learning_rate": 9.500695111398903e-07, + "loss": 0.5282, + "step": 12210 + }, + { + "epoch": 4.287570224719101, + "grad_norm": 0.9328616261482239, + "learning_rate": 9.497067862464277e-07, + "loss": 0.5395, + "step": 12211 + }, + { + "epoch": 4.287921348314606, + "grad_norm": 0.9636738300323486, + "learning_rate": 9.493441143716664e-07, + "loss": 0.5456, + "step": 12212 + }, + { + "epoch": 4.288272471910112, + "grad_norm": 0.9425281286239624, + "learning_rate": 9.489814955280085e-07, + "loss": 0.4871, + "step": 12213 + }, + { + "epoch": 4.288623595505618, + "grad_norm": 0.9811469912528992, + "learning_rate": 9.486189297278573e-07, + "loss": 0.542, + "step": 12214 + }, + { + "epoch": 4.288974719101123, + "grad_norm": 0.9899441599845886, + "learning_rate": 9.4825641698361e-07, + "loss": 0.5737, + "step": 12215 + }, + { + "epoch": 4.289325842696629, + "grad_norm": 0.9817934036254883, + "learning_rate": 9.478939573076665e-07, + "loss": 0.5307, + "step": 12216 + }, + { + "epoch": 4.289676966292135, + "grad_norm": 0.9447322487831116, + "learning_rate": 9.475315507124205e-07, + "loss": 0.5555, + "step": 12217 + }, + { + "epoch": 4.290028089887641, + "grad_norm": 0.9618937969207764, + "learning_rate": 9.471691972102679e-07, + "loss": 0.5336, + "step": 12218 + }, + { + "epoch": 4.290379213483146, + "grad_norm": 0.923475980758667, + "learning_rate": 9.468068968135993e-07, + "loss": 0.5462, + "step": 12219 + }, + { + "epoch": 4.290730337078652, + "grad_norm": 0.9615676403045654, + "learning_rate": 9.464446495348065e-07, + "loss": 0.569, + "step": 12220 + }, + { + "epoch": 4.291081460674158, + "grad_norm": 0.9349548816680908, + "learning_rate": 9.460824553862763e-07, + "loss": 0.5507, + "step": 12221 + }, + { + "epoch": 4.291432584269663, + "grad_norm": 0.976909875869751, + "learning_rate": 9.457203143803972e-07, + "loss": 0.5508, + "step": 12222 + }, + { + "epoch": 4.291783707865169, + "grad_norm": 0.9604188799858093, + "learning_rate": 9.453582265295527e-07, + "loss": 0.5692, + "step": 12223 + }, + { + "epoch": 4.292134831460674, + "grad_norm": 0.987231969833374, + "learning_rate": 9.449961918461273e-07, + "loss": 0.5197, + "step": 12224 + }, + { + "epoch": 4.29248595505618, + "grad_norm": 0.9228474497795105, + "learning_rate": 9.446342103425016e-07, + "loss": 0.518, + "step": 12225 + }, + { + "epoch": 4.292837078651686, + "grad_norm": 0.9360945820808411, + "learning_rate": 9.442722820310541e-07, + "loss": 0.5303, + "step": 12226 + }, + { + "epoch": 4.293188202247191, + "grad_norm": 0.9767336845397949, + "learning_rate": 9.439104069241639e-07, + "loss": 0.5362, + "step": 12227 + }, + { + "epoch": 4.293539325842697, + "grad_norm": 0.9399197101593018, + "learning_rate": 9.435485850342055e-07, + "loss": 0.539, + "step": 12228 + }, + { + "epoch": 4.293890449438202, + "grad_norm": 0.9559866786003113, + "learning_rate": 9.431868163735547e-07, + "loss": 0.5249, + "step": 12229 + }, + { + "epoch": 4.294241573033708, + "grad_norm": 0.955140233039856, + "learning_rate": 9.428251009545817e-07, + "loss": 0.5488, + "step": 12230 + }, + { + "epoch": 4.294592696629214, + "grad_norm": 0.9935112595558167, + "learning_rate": 9.424634387896583e-07, + "loss": 0.5633, + "step": 12231 + }, + { + "epoch": 4.294943820224719, + "grad_norm": 0.9858177304267883, + "learning_rate": 9.421018298911519e-07, + "loss": 0.541, + "step": 12232 + }, + { + "epoch": 4.295294943820225, + "grad_norm": 0.9698333144187927, + "learning_rate": 9.417402742714304e-07, + "loss": 0.5334, + "step": 12233 + }, + { + "epoch": 4.29564606741573, + "grad_norm": 0.9432562589645386, + "learning_rate": 9.413787719428583e-07, + "loss": 0.5174, + "step": 12234 + }, + { + "epoch": 4.295997191011236, + "grad_norm": 0.9450672268867493, + "learning_rate": 9.410173229177977e-07, + "loss": 0.5275, + "step": 12235 + }, + { + "epoch": 4.296348314606742, + "grad_norm": 0.9782847166061401, + "learning_rate": 9.406559272086114e-07, + "loss": 0.5337, + "step": 12236 + }, + { + "epoch": 4.296699438202247, + "grad_norm": 0.9498922824859619, + "learning_rate": 9.40294584827657e-07, + "loss": 0.5406, + "step": 12237 + }, + { + "epoch": 4.297050561797753, + "grad_norm": 0.9263525605201721, + "learning_rate": 9.399332957872939e-07, + "loss": 0.4967, + "step": 12238 + }, + { + "epoch": 4.297401685393258, + "grad_norm": 1.0348913669586182, + "learning_rate": 9.395720600998768e-07, + "loss": 0.5699, + "step": 12239 + }, + { + "epoch": 4.297752808988764, + "grad_norm": 0.9994657039642334, + "learning_rate": 9.392108777777603e-07, + "loss": 0.5297, + "step": 12240 + }, + { + "epoch": 4.29810393258427, + "grad_norm": 1.0398528575897217, + "learning_rate": 9.388497488332965e-07, + "loss": 0.5162, + "step": 12241 + }, + { + "epoch": 4.298455056179775, + "grad_norm": 0.9929320216178894, + "learning_rate": 9.384886732788342e-07, + "loss": 0.5385, + "step": 12242 + }, + { + "epoch": 4.298806179775281, + "grad_norm": 0.9826210737228394, + "learning_rate": 9.381276511267243e-07, + "loss": 0.5688, + "step": 12243 + }, + { + "epoch": 4.299157303370786, + "grad_norm": 0.9365386366844177, + "learning_rate": 9.377666823893114e-07, + "loss": 0.5155, + "step": 12244 + }, + { + "epoch": 4.299508426966292, + "grad_norm": 0.9702936410903931, + "learning_rate": 9.374057670789416e-07, + "loss": 0.5122, + "step": 12245 + }, + { + "epoch": 4.299859550561798, + "grad_norm": 0.9952878355979919, + "learning_rate": 9.370449052079569e-07, + "loss": 0.5713, + "step": 12246 + }, + { + "epoch": 4.300210674157303, + "grad_norm": 0.9644503593444824, + "learning_rate": 9.366840967886998e-07, + "loss": 0.541, + "step": 12247 + }, + { + "epoch": 4.300561797752809, + "grad_norm": 0.9784395694732666, + "learning_rate": 9.363233418335088e-07, + "loss": 0.5286, + "step": 12248 + }, + { + "epoch": 4.300912921348314, + "grad_norm": 0.9706071615219116, + "learning_rate": 9.359626403547206e-07, + "loss": 0.5516, + "step": 12249 + }, + { + "epoch": 4.30126404494382, + "grad_norm": 0.9603403806686401, + "learning_rate": 9.356019923646726e-07, + "loss": 0.5551, + "step": 12250 + }, + { + "epoch": 4.301615168539326, + "grad_norm": 0.9644148945808411, + "learning_rate": 9.352413978756968e-07, + "loss": 0.5234, + "step": 12251 + }, + { + "epoch": 4.301966292134831, + "grad_norm": 0.9663314819335938, + "learning_rate": 9.348808569001272e-07, + "loss": 0.5089, + "step": 12252 + }, + { + "epoch": 4.302317415730337, + "grad_norm": 0.9643239974975586, + "learning_rate": 9.345203694502921e-07, + "loss": 0.5575, + "step": 12253 + }, + { + "epoch": 4.302668539325842, + "grad_norm": 0.9852828979492188, + "learning_rate": 9.341599355385217e-07, + "loss": 0.5353, + "step": 12254 + }, + { + "epoch": 4.303019662921348, + "grad_norm": 0.9515071511268616, + "learning_rate": 9.337995551771406e-07, + "loss": 0.5402, + "step": 12255 + }, + { + "epoch": 4.303370786516854, + "grad_norm": 0.9371696710586548, + "learning_rate": 9.334392283784754e-07, + "loss": 0.5191, + "step": 12256 + }, + { + "epoch": 4.303721910112359, + "grad_norm": 0.9305912256240845, + "learning_rate": 9.330789551548483e-07, + "loss": 0.52, + "step": 12257 + }, + { + "epoch": 4.304073033707865, + "grad_norm": 0.8983298540115356, + "learning_rate": 9.327187355185791e-07, + "loss": 0.4749, + "step": 12258 + }, + { + "epoch": 4.304424157303371, + "grad_norm": 0.9386540651321411, + "learning_rate": 9.323585694819887e-07, + "loss": 0.5132, + "step": 12259 + }, + { + "epoch": 4.304775280898877, + "grad_norm": 0.973682701587677, + "learning_rate": 9.319984570573931e-07, + "loss": 0.5363, + "step": 12260 + }, + { + "epoch": 4.305126404494382, + "grad_norm": 0.94440096616745, + "learning_rate": 9.316383982571092e-07, + "loss": 0.5294, + "step": 12261 + }, + { + "epoch": 4.305477528089888, + "grad_norm": 0.9566545486450195, + "learning_rate": 9.312783930934494e-07, + "loss": 0.5188, + "step": 12262 + }, + { + "epoch": 4.305828651685394, + "grad_norm": 0.956429660320282, + "learning_rate": 9.309184415787267e-07, + "loss": 0.5164, + "step": 12263 + }, + { + "epoch": 4.306179775280899, + "grad_norm": 0.9590983986854553, + "learning_rate": 9.30558543725251e-07, + "loss": 0.5619, + "step": 12264 + }, + { + "epoch": 4.306530898876405, + "grad_norm": 0.9481155276298523, + "learning_rate": 9.301986995453289e-07, + "loss": 0.5349, + "step": 12265 + }, + { + "epoch": 4.30688202247191, + "grad_norm": 0.9505434632301331, + "learning_rate": 9.298389090512691e-07, + "loss": 0.5264, + "step": 12266 + }, + { + "epoch": 4.307233146067416, + "grad_norm": 0.9352911114692688, + "learning_rate": 9.29479172255374e-07, + "loss": 0.5499, + "step": 12267 + }, + { + "epoch": 4.307584269662922, + "grad_norm": 0.9663898348808289, + "learning_rate": 9.291194891699482e-07, + "loss": 0.5476, + "step": 12268 + }, + { + "epoch": 4.307935393258427, + "grad_norm": 0.950279951095581, + "learning_rate": 9.287598598072909e-07, + "loss": 0.5322, + "step": 12269 + }, + { + "epoch": 4.308286516853933, + "grad_norm": 0.9531329274177551, + "learning_rate": 9.284002841797024e-07, + "loss": 0.5454, + "step": 12270 + }, + { + "epoch": 4.308637640449438, + "grad_norm": 0.9292583465576172, + "learning_rate": 9.280407622994789e-07, + "loss": 0.5235, + "step": 12271 + }, + { + "epoch": 4.308988764044944, + "grad_norm": 0.9885945916175842, + "learning_rate": 9.276812941789168e-07, + "loss": 0.5431, + "step": 12272 + }, + { + "epoch": 4.30933988764045, + "grad_norm": 0.9649800062179565, + "learning_rate": 9.273218798303079e-07, + "loss": 0.5493, + "step": 12273 + }, + { + "epoch": 4.309691011235955, + "grad_norm": 0.9979891180992126, + "learning_rate": 9.269625192659462e-07, + "loss": 0.5414, + "step": 12274 + }, + { + "epoch": 4.310042134831461, + "grad_norm": 0.9611807465553284, + "learning_rate": 9.266032124981203e-07, + "loss": 0.5319, + "step": 12275 + }, + { + "epoch": 4.310393258426966, + "grad_norm": 0.9705820679664612, + "learning_rate": 9.26243959539117e-07, + "loss": 0.5429, + "step": 12276 + }, + { + "epoch": 4.310744382022472, + "grad_norm": 0.9603354334831238, + "learning_rate": 9.258847604012247e-07, + "loss": 0.5342, + "step": 12277 + }, + { + "epoch": 4.311095505617978, + "grad_norm": 0.9562709331512451, + "learning_rate": 9.255256150967257e-07, + "loss": 0.5391, + "step": 12278 + }, + { + "epoch": 4.311446629213483, + "grad_norm": 0.9405698776245117, + "learning_rate": 9.25166523637904e-07, + "loss": 0.5396, + "step": 12279 + }, + { + "epoch": 4.311797752808989, + "grad_norm": 1.004160761833191, + "learning_rate": 9.24807486037039e-07, + "loss": 0.5389, + "step": 12280 + }, + { + "epoch": 4.312148876404494, + "grad_norm": 0.9706826210021973, + "learning_rate": 9.24448502306411e-07, + "loss": 0.5467, + "step": 12281 + }, + { + "epoch": 4.3125, + "grad_norm": 0.9266903400421143, + "learning_rate": 9.240895724582948e-07, + "loss": 0.5056, + "step": 12282 + }, + { + "epoch": 4.312851123595506, + "grad_norm": 0.9716761708259583, + "learning_rate": 9.237306965049678e-07, + "loss": 0.5261, + "step": 12283 + }, + { + "epoch": 4.313202247191011, + "grad_norm": 0.9746789336204529, + "learning_rate": 9.233718744587011e-07, + "loss": 0.5238, + "step": 12284 + }, + { + "epoch": 4.313553370786517, + "grad_norm": 0.9478601813316345, + "learning_rate": 9.230131063317682e-07, + "loss": 0.5329, + "step": 12285 + }, + { + "epoch": 4.313904494382022, + "grad_norm": 0.9756632447242737, + "learning_rate": 9.226543921364367e-07, + "loss": 0.5305, + "step": 12286 + }, + { + "epoch": 4.314255617977528, + "grad_norm": 0.9375220537185669, + "learning_rate": 9.222957318849759e-07, + "loss": 0.5592, + "step": 12287 + }, + { + "epoch": 4.314606741573034, + "grad_norm": 0.9790070056915283, + "learning_rate": 9.219371255896514e-07, + "loss": 0.5339, + "step": 12288 + }, + { + "epoch": 4.314957865168539, + "grad_norm": 0.9641648530960083, + "learning_rate": 9.215785732627258e-07, + "loss": 0.5467, + "step": 12289 + }, + { + "epoch": 4.315308988764045, + "grad_norm": 0.9815196394920349, + "learning_rate": 9.212200749164629e-07, + "loss": 0.5385, + "step": 12290 + }, + { + "epoch": 4.31566011235955, + "grad_norm": 0.9742016792297363, + "learning_rate": 9.208616305631221e-07, + "loss": 0.5529, + "step": 12291 + }, + { + "epoch": 4.316011235955056, + "grad_norm": 0.9334456920623779, + "learning_rate": 9.205032402149627e-07, + "loss": 0.5185, + "step": 12292 + }, + { + "epoch": 4.316362359550562, + "grad_norm": 0.974540114402771, + "learning_rate": 9.201449038842403e-07, + "loss": 0.4932, + "step": 12293 + }, + { + "epoch": 4.316713483146067, + "grad_norm": 0.9646206498146057, + "learning_rate": 9.19786621583211e-07, + "loss": 0.5064, + "step": 12294 + }, + { + "epoch": 4.317064606741573, + "grad_norm": 0.9950608611106873, + "learning_rate": 9.194283933241261e-07, + "loss": 0.5394, + "step": 12295 + }, + { + "epoch": 4.317415730337078, + "grad_norm": 0.9797853827476501, + "learning_rate": 9.190702191192385e-07, + "loss": 0.5598, + "step": 12296 + }, + { + "epoch": 4.317766853932584, + "grad_norm": 0.9890244007110596, + "learning_rate": 9.18712098980796e-07, + "loss": 0.4933, + "step": 12297 + }, + { + "epoch": 4.31811797752809, + "grad_norm": 0.952889621257782, + "learning_rate": 9.183540329210472e-07, + "loss": 0.5427, + "step": 12298 + }, + { + "epoch": 4.318469101123595, + "grad_norm": 0.9722253680229187, + "learning_rate": 9.179960209522362e-07, + "loss": 0.5215, + "step": 12299 + }, + { + "epoch": 4.318820224719101, + "grad_norm": 0.9532808065414429, + "learning_rate": 9.176380630866086e-07, + "loss": 0.5347, + "step": 12300 + }, + { + "epoch": 4.319171348314606, + "grad_norm": 0.9683524966239929, + "learning_rate": 9.17280159336405e-07, + "loss": 0.5467, + "step": 12301 + }, + { + "epoch": 4.319522471910112, + "grad_norm": 0.978545069694519, + "learning_rate": 9.169223097138646e-07, + "loss": 0.5694, + "step": 12302 + }, + { + "epoch": 4.319873595505618, + "grad_norm": 0.9630188941955566, + "learning_rate": 9.165645142312277e-07, + "loss": 0.5305, + "step": 12303 + }, + { + "epoch": 4.320224719101123, + "grad_norm": 0.9584699869155884, + "learning_rate": 9.162067729007285e-07, + "loss": 0.5124, + "step": 12304 + }, + { + "epoch": 4.320575842696629, + "grad_norm": 0.9606171250343323, + "learning_rate": 9.158490857346034e-07, + "loss": 0.5357, + "step": 12305 + }, + { + "epoch": 4.320926966292135, + "grad_norm": 1.002145767211914, + "learning_rate": 9.154914527450829e-07, + "loss": 0.5591, + "step": 12306 + }, + { + "epoch": 4.321278089887641, + "grad_norm": 0.9570406079292297, + "learning_rate": 9.151338739443996e-07, + "loss": 0.5475, + "step": 12307 + }, + { + "epoch": 4.321629213483146, + "grad_norm": 0.964959979057312, + "learning_rate": 9.147763493447811e-07, + "loss": 0.5515, + "step": 12308 + }, + { + "epoch": 4.321980337078652, + "grad_norm": 0.9553263783454895, + "learning_rate": 9.144188789584554e-07, + "loss": 0.5222, + "step": 12309 + }, + { + "epoch": 4.322331460674158, + "grad_norm": 0.9248050451278687, + "learning_rate": 9.140614627976468e-07, + "loss": 0.5157, + "step": 12310 + }, + { + "epoch": 4.322682584269663, + "grad_norm": 0.9889593720436096, + "learning_rate": 9.137041008745798e-07, + "loss": 0.5617, + "step": 12311 + }, + { + "epoch": 4.323033707865169, + "grad_norm": 0.9919673800468445, + "learning_rate": 9.13346793201475e-07, + "loss": 0.5684, + "step": 12312 + }, + { + "epoch": 4.323384831460674, + "grad_norm": 0.9731014370918274, + "learning_rate": 9.129895397905514e-07, + "loss": 0.5485, + "step": 12313 + }, + { + "epoch": 4.32373595505618, + "grad_norm": 0.9841817617416382, + "learning_rate": 9.126323406540283e-07, + "loss": 0.5494, + "step": 12314 + }, + { + "epoch": 4.324087078651686, + "grad_norm": 0.9503588080406189, + "learning_rate": 9.122751958041201e-07, + "loss": 0.537, + "step": 12315 + }, + { + "epoch": 4.324438202247191, + "grad_norm": 0.9729681015014648, + "learning_rate": 9.119181052530423e-07, + "loss": 0.5627, + "step": 12316 + }, + { + "epoch": 4.324789325842697, + "grad_norm": 0.9524460434913635, + "learning_rate": 9.115610690130056e-07, + "loss": 0.5459, + "step": 12317 + }, + { + "epoch": 4.325140449438202, + "grad_norm": 0.9541613459587097, + "learning_rate": 9.112040870962218e-07, + "loss": 0.5601, + "step": 12318 + }, + { + "epoch": 4.325491573033708, + "grad_norm": 1.0021734237670898, + "learning_rate": 9.10847159514898e-07, + "loss": 0.5817, + "step": 12319 + }, + { + "epoch": 4.325842696629214, + "grad_norm": 0.966221034526825, + "learning_rate": 9.104902862812421e-07, + "loss": 0.5155, + "step": 12320 + }, + { + "epoch": 4.326193820224719, + "grad_norm": 0.9274359941482544, + "learning_rate": 9.101334674074574e-07, + "loss": 0.4951, + "step": 12321 + }, + { + "epoch": 4.326544943820225, + "grad_norm": 0.9869076609611511, + "learning_rate": 9.097767029057489e-07, + "loss": 0.5673, + "step": 12322 + }, + { + "epoch": 4.32689606741573, + "grad_norm": 0.9844452142715454, + "learning_rate": 9.094199927883152e-07, + "loss": 0.5521, + "step": 12323 + }, + { + "epoch": 4.327247191011236, + "grad_norm": 0.9828171730041504, + "learning_rate": 9.090633370673576e-07, + "loss": 0.5133, + "step": 12324 + }, + { + "epoch": 4.327598314606742, + "grad_norm": 0.9943015575408936, + "learning_rate": 9.087067357550724e-07, + "loss": 0.5706, + "step": 12325 + }, + { + "epoch": 4.327949438202247, + "grad_norm": 1.0033771991729736, + "learning_rate": 9.083501888636545e-07, + "loss": 0.5602, + "step": 12326 + }, + { + "epoch": 4.328300561797753, + "grad_norm": 0.9902317523956299, + "learning_rate": 9.079936964052987e-07, + "loss": 0.5165, + "step": 12327 + }, + { + "epoch": 4.328651685393258, + "grad_norm": 0.9429242610931396, + "learning_rate": 9.076372583921955e-07, + "loss": 0.5271, + "step": 12328 + }, + { + "epoch": 4.329002808988764, + "grad_norm": 0.946856677532196, + "learning_rate": 9.072808748365364e-07, + "loss": 0.5301, + "step": 12329 + }, + { + "epoch": 4.32935393258427, + "grad_norm": 0.9599604606628418, + "learning_rate": 9.069245457505074e-07, + "loss": 0.5567, + "step": 12330 + }, + { + "epoch": 4.329705056179775, + "grad_norm": 0.9679630398750305, + "learning_rate": 9.065682711462967e-07, + "loss": 0.539, + "step": 12331 + }, + { + "epoch": 4.330056179775281, + "grad_norm": 0.9669332504272461, + "learning_rate": 9.06212051036087e-07, + "loss": 0.5366, + "step": 12332 + }, + { + "epoch": 4.330407303370786, + "grad_norm": 0.9712163805961609, + "learning_rate": 9.058558854320618e-07, + "loss": 0.5263, + "step": 12333 + }, + { + "epoch": 4.330758426966292, + "grad_norm": 0.9649853706359863, + "learning_rate": 9.054997743464006e-07, + "loss": 0.5162, + "step": 12334 + }, + { + "epoch": 4.331109550561798, + "grad_norm": 0.9846929907798767, + "learning_rate": 9.051437177912834e-07, + "loss": 0.5416, + "step": 12335 + }, + { + "epoch": 4.331460674157303, + "grad_norm": 0.9676690101623535, + "learning_rate": 9.047877157788853e-07, + "loss": 0.5468, + "step": 12336 + }, + { + "epoch": 4.331811797752809, + "grad_norm": 0.9918473362922668, + "learning_rate": 9.044317683213835e-07, + "loss": 0.5719, + "step": 12337 + }, + { + "epoch": 4.332162921348314, + "grad_norm": 0.9597248435020447, + "learning_rate": 9.040758754309495e-07, + "loss": 0.5173, + "step": 12338 + }, + { + "epoch": 4.33251404494382, + "grad_norm": 0.9763705730438232, + "learning_rate": 9.037200371197543e-07, + "loss": 0.5307, + "step": 12339 + }, + { + "epoch": 4.332865168539326, + "grad_norm": 0.9629073143005371, + "learning_rate": 9.033642533999687e-07, + "loss": 0.5251, + "step": 12340 + }, + { + "epoch": 4.333216292134831, + "grad_norm": 0.9844147562980652, + "learning_rate": 9.030085242837586e-07, + "loss": 0.5454, + "step": 12341 + }, + { + "epoch": 4.333567415730337, + "grad_norm": 0.9633737802505493, + "learning_rate": 9.02652849783291e-07, + "loss": 0.5465, + "step": 12342 + }, + { + "epoch": 4.333918539325842, + "grad_norm": 0.9700624346733093, + "learning_rate": 9.022972299107283e-07, + "loss": 0.5427, + "step": 12343 + }, + { + "epoch": 4.334269662921348, + "grad_norm": 0.9793705940246582, + "learning_rate": 9.019416646782341e-07, + "loss": 0.5351, + "step": 12344 + }, + { + "epoch": 4.334620786516854, + "grad_norm": 0.9820535778999329, + "learning_rate": 9.015861540979667e-07, + "loss": 0.5498, + "step": 12345 + }, + { + "epoch": 4.334971910112359, + "grad_norm": 0.969109833240509, + "learning_rate": 9.01230698182086e-07, + "loss": 0.5061, + "step": 12346 + }, + { + "epoch": 4.335323033707865, + "grad_norm": 0.9529459476470947, + "learning_rate": 9.008752969427462e-07, + "loss": 0.5371, + "step": 12347 + }, + { + "epoch": 4.335674157303371, + "grad_norm": 0.9738073945045471, + "learning_rate": 9.005199503921036e-07, + "loss": 0.5697, + "step": 12348 + }, + { + "epoch": 4.336025280898877, + "grad_norm": 0.9644418954849243, + "learning_rate": 9.001646585423093e-07, + "loss": 0.5414, + "step": 12349 + }, + { + "epoch": 4.336376404494382, + "grad_norm": 0.971838653087616, + "learning_rate": 8.998094214055159e-07, + "loss": 0.5526, + "step": 12350 + }, + { + "epoch": 4.336727528089888, + "grad_norm": 0.9439349174499512, + "learning_rate": 8.994542389938704e-07, + "loss": 0.5553, + "step": 12351 + }, + { + "epoch": 4.337078651685394, + "grad_norm": 0.9444662928581238, + "learning_rate": 8.990991113195199e-07, + "loss": 0.5374, + "step": 12352 + }, + { + "epoch": 4.337429775280899, + "grad_norm": 0.9644107818603516, + "learning_rate": 8.987440383946106e-07, + "loss": 0.5382, + "step": 12353 + }, + { + "epoch": 4.337780898876405, + "grad_norm": 0.9498382806777954, + "learning_rate": 8.98389020231284e-07, + "loss": 0.5284, + "step": 12354 + }, + { + "epoch": 4.33813202247191, + "grad_norm": 0.9461143016815186, + "learning_rate": 8.980340568416834e-07, + "loss": 0.5517, + "step": 12355 + }, + { + "epoch": 4.338483146067416, + "grad_norm": 0.942456066608429, + "learning_rate": 8.976791482379462e-07, + "loss": 0.4816, + "step": 12356 + }, + { + "epoch": 4.338834269662922, + "grad_norm": 0.9585902094841003, + "learning_rate": 8.973242944322119e-07, + "loss": 0.573, + "step": 12357 + }, + { + "epoch": 4.339185393258427, + "grad_norm": 1.0187699794769287, + "learning_rate": 8.969694954366145e-07, + "loss": 0.5467, + "step": 12358 + }, + { + "epoch": 4.339536516853933, + "grad_norm": 0.9521846771240234, + "learning_rate": 8.966147512632898e-07, + "loss": 0.5231, + "step": 12359 + }, + { + "epoch": 4.339887640449438, + "grad_norm": 0.9272840023040771, + "learning_rate": 8.962600619243672e-07, + "loss": 0.517, + "step": 12360 + }, + { + "epoch": 4.340238764044944, + "grad_norm": 0.9619066119194031, + "learning_rate": 8.959054274319792e-07, + "loss": 0.5276, + "step": 12361 + }, + { + "epoch": 4.34058988764045, + "grad_norm": 0.9244234561920166, + "learning_rate": 8.955508477982522e-07, + "loss": 0.5178, + "step": 12362 + }, + { + "epoch": 4.340941011235955, + "grad_norm": 0.9564957022666931, + "learning_rate": 8.951963230353139e-07, + "loss": 0.5314, + "step": 12363 + }, + { + "epoch": 4.341292134831461, + "grad_norm": 0.9430667757987976, + "learning_rate": 8.948418531552882e-07, + "loss": 0.5077, + "step": 12364 + }, + { + "epoch": 4.341643258426966, + "grad_norm": 0.9410857558250427, + "learning_rate": 8.944874381702968e-07, + "loss": 0.542, + "step": 12365 + }, + { + "epoch": 4.341994382022472, + "grad_norm": 0.9539079070091248, + "learning_rate": 8.94133078092462e-07, + "loss": 0.5402, + "step": 12366 + }, + { + "epoch": 4.342345505617978, + "grad_norm": 1.0087920427322388, + "learning_rate": 8.937787729339009e-07, + "loss": 0.5727, + "step": 12367 + }, + { + "epoch": 4.342696629213483, + "grad_norm": 0.9860296845436096, + "learning_rate": 8.93424522706732e-07, + "loss": 0.5704, + "step": 12368 + }, + { + "epoch": 4.343047752808989, + "grad_norm": 0.9632112383842468, + "learning_rate": 8.930703274230689e-07, + "loss": 0.5241, + "step": 12369 + }, + { + "epoch": 4.343398876404494, + "grad_norm": 0.9831791520118713, + "learning_rate": 8.927161870950263e-07, + "loss": 0.5557, + "step": 12370 + }, + { + "epoch": 4.34375, + "grad_norm": 0.9489305019378662, + "learning_rate": 8.923621017347139e-07, + "loss": 0.5263, + "step": 12371 + }, + { + "epoch": 4.344101123595506, + "grad_norm": 0.9667307734489441, + "learning_rate": 8.920080713542431e-07, + "loss": 0.5656, + "step": 12372 + }, + { + "epoch": 4.344452247191011, + "grad_norm": 0.9814040660858154, + "learning_rate": 8.916540959657191e-07, + "loss": 0.5834, + "step": 12373 + }, + { + "epoch": 4.344803370786517, + "grad_norm": 0.9604358673095703, + "learning_rate": 8.913001755812497e-07, + "loss": 0.5616, + "step": 12374 + }, + { + "epoch": 4.345154494382022, + "grad_norm": 0.9895382523536682, + "learning_rate": 8.909463102129376e-07, + "loss": 0.5572, + "step": 12375 + }, + { + "epoch": 4.345505617977528, + "grad_norm": 0.9979939460754395, + "learning_rate": 8.905924998728841e-07, + "loss": 0.5528, + "step": 12376 + }, + { + "epoch": 4.345856741573034, + "grad_norm": 0.9355661273002625, + "learning_rate": 8.902387445731908e-07, + "loss": 0.5466, + "step": 12377 + }, + { + "epoch": 4.346207865168539, + "grad_norm": 0.9404891729354858, + "learning_rate": 8.898850443259541e-07, + "loss": 0.5644, + "step": 12378 + }, + { + "epoch": 4.346558988764045, + "grad_norm": 0.9663461446762085, + "learning_rate": 8.895313991432716e-07, + "loss": 0.5545, + "step": 12379 + }, + { + "epoch": 4.34691011235955, + "grad_norm": 0.9721162915229797, + "learning_rate": 8.891778090372366e-07, + "loss": 0.5318, + "step": 12380 + }, + { + "epoch": 4.347261235955056, + "grad_norm": 0.9218356609344482, + "learning_rate": 8.888242740199429e-07, + "loss": 0.5237, + "step": 12381 + }, + { + "epoch": 4.347612359550562, + "grad_norm": 0.9509425163269043, + "learning_rate": 8.884707941034796e-07, + "loss": 0.5218, + "step": 12382 + }, + { + "epoch": 4.347963483146067, + "grad_norm": 0.9409254193305969, + "learning_rate": 8.88117369299937e-07, + "loss": 0.5028, + "step": 12383 + }, + { + "epoch": 4.348314606741573, + "grad_norm": 0.9798882603645325, + "learning_rate": 8.877639996213999e-07, + "loss": 0.5226, + "step": 12384 + }, + { + "epoch": 4.348665730337078, + "grad_norm": 0.9576828479766846, + "learning_rate": 8.874106850799554e-07, + "loss": 0.5418, + "step": 12385 + }, + { + "epoch": 4.349016853932584, + "grad_norm": 0.985493540763855, + "learning_rate": 8.870574256876849e-07, + "loss": 0.5476, + "step": 12386 + }, + { + "epoch": 4.34936797752809, + "grad_norm": 0.9660130739212036, + "learning_rate": 8.867042214566709e-07, + "loss": 0.5191, + "step": 12387 + }, + { + "epoch": 4.349719101123595, + "grad_norm": 0.9378551840782166, + "learning_rate": 8.863510723989921e-07, + "loss": 0.5, + "step": 12388 + }, + { + "epoch": 4.350070224719101, + "grad_norm": 0.9422694444656372, + "learning_rate": 8.85997978526725e-07, + "loss": 0.5196, + "step": 12389 + }, + { + "epoch": 4.350421348314606, + "grad_norm": 0.9335485696792603, + "learning_rate": 8.856449398519465e-07, + "loss": 0.5204, + "step": 12390 + }, + { + "epoch": 4.350772471910112, + "grad_norm": 0.9474460482597351, + "learning_rate": 8.85291956386729e-07, + "loss": 0.5233, + "step": 12391 + }, + { + "epoch": 4.351123595505618, + "grad_norm": 0.9562787413597107, + "learning_rate": 8.849390281431458e-07, + "loss": 0.5494, + "step": 12392 + }, + { + "epoch": 4.351474719101123, + "grad_norm": 0.9785548448562622, + "learning_rate": 8.845861551332649e-07, + "loss": 0.5823, + "step": 12393 + }, + { + "epoch": 4.351825842696629, + "grad_norm": 0.9405418634414673, + "learning_rate": 8.842333373691559e-07, + "loss": 0.5275, + "step": 12394 + }, + { + "epoch": 4.352176966292135, + "grad_norm": 0.9441057443618774, + "learning_rate": 8.838805748628831e-07, + "loss": 0.5634, + "step": 12395 + }, + { + "epoch": 4.352528089887641, + "grad_norm": 0.9374129176139832, + "learning_rate": 8.83527867626513e-07, + "loss": 0.5192, + "step": 12396 + }, + { + "epoch": 4.352879213483146, + "grad_norm": 0.9628786444664001, + "learning_rate": 8.831752156721055e-07, + "loss": 0.4974, + "step": 12397 + }, + { + "epoch": 4.353230337078652, + "grad_norm": 0.9277937412261963, + "learning_rate": 8.828226190117229e-07, + "loss": 0.5039, + "step": 12398 + }, + { + "epoch": 4.353581460674158, + "grad_norm": 1.0032870769500732, + "learning_rate": 8.824700776574222e-07, + "loss": 0.5119, + "step": 12399 + }, + { + "epoch": 4.353932584269663, + "grad_norm": 0.9556220173835754, + "learning_rate": 8.821175916212612e-07, + "loss": 0.53, + "step": 12400 + }, + { + "epoch": 4.354283707865169, + "grad_norm": 0.9566677808761597, + "learning_rate": 8.817651609152944e-07, + "loss": 0.5326, + "step": 12401 + }, + { + "epoch": 4.354634831460674, + "grad_norm": 0.9769572019577026, + "learning_rate": 8.814127855515736e-07, + "loss": 0.5212, + "step": 12402 + }, + { + "epoch": 4.35498595505618, + "grad_norm": 0.9610295295715332, + "learning_rate": 8.810604655421511e-07, + "loss": 0.5662, + "step": 12403 + }, + { + "epoch": 4.355337078651686, + "grad_norm": 0.9495554566383362, + "learning_rate": 8.807082008990747e-07, + "loss": 0.5223, + "step": 12404 + }, + { + "epoch": 4.355688202247191, + "grad_norm": 0.9711163640022278, + "learning_rate": 8.803559916343929e-07, + "loss": 0.5428, + "step": 12405 + }, + { + "epoch": 4.356039325842697, + "grad_norm": 0.9295318126678467, + "learning_rate": 8.800038377601494e-07, + "loss": 0.523, + "step": 12406 + }, + { + "epoch": 4.356390449438202, + "grad_norm": 0.9663895964622498, + "learning_rate": 8.796517392883894e-07, + "loss": 0.5404, + "step": 12407 + }, + { + "epoch": 4.356741573033708, + "grad_norm": 0.950190544128418, + "learning_rate": 8.792996962311531e-07, + "loss": 0.5393, + "step": 12408 + }, + { + "epoch": 4.357092696629214, + "grad_norm": 0.941865861415863, + "learning_rate": 8.789477086004797e-07, + "loss": 0.538, + "step": 12409 + }, + { + "epoch": 4.357443820224719, + "grad_norm": 0.9495593309402466, + "learning_rate": 8.785957764084082e-07, + "loss": 0.5272, + "step": 12410 + }, + { + "epoch": 4.357794943820225, + "grad_norm": 0.9695387482643127, + "learning_rate": 8.782438996669731e-07, + "loss": 0.5021, + "step": 12411 + }, + { + "epoch": 4.35814606741573, + "grad_norm": 0.9602741003036499, + "learning_rate": 8.778920783882094e-07, + "loss": 0.5409, + "step": 12412 + }, + { + "epoch": 4.358497191011236, + "grad_norm": 0.9892792105674744, + "learning_rate": 8.775403125841481e-07, + "loss": 0.5616, + "step": 12413 + }, + { + "epoch": 4.358848314606742, + "grad_norm": 1.003823161125183, + "learning_rate": 8.771886022668202e-07, + "loss": 0.5104, + "step": 12414 + }, + { + "epoch": 4.359199438202247, + "grad_norm": 0.9129050970077515, + "learning_rate": 8.768369474482535e-07, + "loss": 0.4824, + "step": 12415 + }, + { + "epoch": 4.359550561797753, + "grad_norm": 0.9881854057312012, + "learning_rate": 8.764853481404736e-07, + "loss": 0.5338, + "step": 12416 + }, + { + "epoch": 4.359901685393258, + "grad_norm": 0.963223934173584, + "learning_rate": 8.761338043555062e-07, + "loss": 0.524, + "step": 12417 + }, + { + "epoch": 4.360252808988764, + "grad_norm": 0.9451430439949036, + "learning_rate": 8.757823161053724e-07, + "loss": 0.52, + "step": 12418 + }, + { + "epoch": 4.36060393258427, + "grad_norm": 0.9510461091995239, + "learning_rate": 8.754308834020944e-07, + "loss": 0.4999, + "step": 12419 + }, + { + "epoch": 4.360955056179775, + "grad_norm": 0.9316567182540894, + "learning_rate": 8.750795062576892e-07, + "loss": 0.5203, + "step": 12420 + }, + { + "epoch": 4.361306179775281, + "grad_norm": 0.9610862135887146, + "learning_rate": 8.747281846841751e-07, + "loss": 0.5364, + "step": 12421 + }, + { + "epoch": 4.361657303370786, + "grad_norm": 0.9891303181648254, + "learning_rate": 8.743769186935663e-07, + "loss": 0.5487, + "step": 12422 + }, + { + "epoch": 4.362008426966292, + "grad_norm": 0.9863368272781372, + "learning_rate": 8.740257082978751e-07, + "loss": 0.542, + "step": 12423 + }, + { + "epoch": 4.362359550561798, + "grad_norm": 0.9842139482498169, + "learning_rate": 8.73674553509114e-07, + "loss": 0.547, + "step": 12424 + }, + { + "epoch": 4.362710674157303, + "grad_norm": 0.9630603790283203, + "learning_rate": 8.73323454339291e-07, + "loss": 0.5519, + "step": 12425 + }, + { + "epoch": 4.363061797752809, + "grad_norm": 1.0026322603225708, + "learning_rate": 8.729724108004145e-07, + "loss": 0.5577, + "step": 12426 + }, + { + "epoch": 4.363412921348314, + "grad_norm": 0.9556251168251038, + "learning_rate": 8.726214229044883e-07, + "loss": 0.5312, + "step": 12427 + }, + { + "epoch": 4.36376404494382, + "grad_norm": 0.9701088070869446, + "learning_rate": 8.72270490663518e-07, + "loss": 0.5482, + "step": 12428 + }, + { + "epoch": 4.364115168539326, + "grad_norm": 0.9620578289031982, + "learning_rate": 8.719196140895039e-07, + "loss": 0.5198, + "step": 12429 + }, + { + "epoch": 4.364466292134831, + "grad_norm": 0.9632093906402588, + "learning_rate": 8.71568793194445e-07, + "loss": 0.5311, + "step": 12430 + }, + { + "epoch": 4.364817415730337, + "grad_norm": 0.9394406676292419, + "learning_rate": 8.712180279903406e-07, + "loss": 0.5202, + "step": 12431 + }, + { + "epoch": 4.365168539325842, + "grad_norm": 0.9807718396186829, + "learning_rate": 8.708673184891853e-07, + "loss": 0.5485, + "step": 12432 + }, + { + "epoch": 4.365519662921348, + "grad_norm": 0.9206869006156921, + "learning_rate": 8.705166647029742e-07, + "loss": 0.4914, + "step": 12433 + }, + { + "epoch": 4.365870786516854, + "grad_norm": 0.9944050312042236, + "learning_rate": 8.701660666436981e-07, + "loss": 0.5567, + "step": 12434 + }, + { + "epoch": 4.366221910112359, + "grad_norm": 0.9486953616142273, + "learning_rate": 8.698155243233489e-07, + "loss": 0.5317, + "step": 12435 + }, + { + "epoch": 4.366573033707865, + "grad_norm": 0.9784730672836304, + "learning_rate": 8.694650377539127e-07, + "loss": 0.5557, + "step": 12436 + }, + { + "epoch": 4.366924157303371, + "grad_norm": 1.0058505535125732, + "learning_rate": 8.69114606947378e-07, + "loss": 0.5398, + "step": 12437 + }, + { + "epoch": 4.367275280898877, + "grad_norm": 0.9485182166099548, + "learning_rate": 8.68764231915728e-07, + "loss": 0.542, + "step": 12438 + }, + { + "epoch": 4.367626404494382, + "grad_norm": 1.0283043384552002, + "learning_rate": 8.684139126709445e-07, + "loss": 0.5361, + "step": 12439 + }, + { + "epoch": 4.367977528089888, + "grad_norm": 0.985747754573822, + "learning_rate": 8.680636492250099e-07, + "loss": 0.557, + "step": 12440 + }, + { + "epoch": 4.368328651685394, + "grad_norm": 0.939697265625, + "learning_rate": 8.677134415899013e-07, + "loss": 0.5497, + "step": 12441 + }, + { + "epoch": 4.368679775280899, + "grad_norm": 0.955487072467804, + "learning_rate": 8.673632897775969e-07, + "loss": 0.5397, + "step": 12442 + }, + { + "epoch": 4.369030898876405, + "grad_norm": 0.9704470038414001, + "learning_rate": 8.670131938000703e-07, + "loss": 0.525, + "step": 12443 + }, + { + "epoch": 4.36938202247191, + "grad_norm": 0.97579026222229, + "learning_rate": 8.666631536692957e-07, + "loss": 0.554, + "step": 12444 + }, + { + "epoch": 4.369733146067416, + "grad_norm": 0.9477248787879944, + "learning_rate": 8.663131693972429e-07, + "loss": 0.534, + "step": 12445 + }, + { + "epoch": 4.370084269662922, + "grad_norm": 0.9262085556983948, + "learning_rate": 8.659632409958827e-07, + "loss": 0.5403, + "step": 12446 + }, + { + "epoch": 4.370435393258427, + "grad_norm": 1.063076376914978, + "learning_rate": 8.656133684771803e-07, + "loss": 0.5803, + "step": 12447 + }, + { + "epoch": 4.370786516853933, + "grad_norm": 0.9366119503974915, + "learning_rate": 8.65263551853103e-07, + "loss": 0.5438, + "step": 12448 + }, + { + "epoch": 4.371137640449438, + "grad_norm": 0.9156193733215332, + "learning_rate": 8.649137911356125e-07, + "loss": 0.4923, + "step": 12449 + }, + { + "epoch": 4.371488764044944, + "grad_norm": 0.9831406474113464, + "learning_rate": 8.645640863366722e-07, + "loss": 0.5392, + "step": 12450 + }, + { + "epoch": 4.37183988764045, + "grad_norm": 0.9549115896224976, + "learning_rate": 8.642144374682406e-07, + "loss": 0.5302, + "step": 12451 + }, + { + "epoch": 4.372191011235955, + "grad_norm": 0.9903922080993652, + "learning_rate": 8.638648445422748e-07, + "loss": 0.5185, + "step": 12452 + }, + { + "epoch": 4.372542134831461, + "grad_norm": 0.9924707412719727, + "learning_rate": 8.635153075707317e-07, + "loss": 0.5342, + "step": 12453 + }, + { + "epoch": 4.372893258426966, + "grad_norm": 0.9407729506492615, + "learning_rate": 8.631658265655643e-07, + "loss": 0.5386, + "step": 12454 + }, + { + "epoch": 4.373244382022472, + "grad_norm": 0.9895833134651184, + "learning_rate": 8.628164015387256e-07, + "loss": 0.5665, + "step": 12455 + }, + { + "epoch": 4.373595505617978, + "grad_norm": 1.0322265625, + "learning_rate": 8.624670325021642e-07, + "loss": 0.5344, + "step": 12456 + }, + { + "epoch": 4.373946629213483, + "grad_norm": 0.9538781046867371, + "learning_rate": 8.621177194678298e-07, + "loss": 0.5246, + "step": 12457 + }, + { + "epoch": 4.374297752808989, + "grad_norm": 0.9721415638923645, + "learning_rate": 8.617684624476672e-07, + "loss": 0.5512, + "step": 12458 + }, + { + "epoch": 4.374648876404494, + "grad_norm": 0.9502307772636414, + "learning_rate": 8.614192614536218e-07, + "loss": 0.5111, + "step": 12459 + }, + { + "epoch": 4.375, + "grad_norm": 0.9485645890235901, + "learning_rate": 8.610701164976352e-07, + "loss": 0.5065, + "step": 12460 + }, + { + "epoch": 4.375351123595506, + "grad_norm": 0.9798952341079712, + "learning_rate": 8.607210275916486e-07, + "loss": 0.562, + "step": 12461 + }, + { + "epoch": 4.375702247191011, + "grad_norm": 0.9813569784164429, + "learning_rate": 8.603719947475994e-07, + "loss": 0.5294, + "step": 12462 + }, + { + "epoch": 4.376053370786517, + "grad_norm": 0.9763311743736267, + "learning_rate": 8.600230179774255e-07, + "loss": 0.5846, + "step": 12463 + }, + { + "epoch": 4.376404494382022, + "grad_norm": 0.912537157535553, + "learning_rate": 8.596740972930615e-07, + "loss": 0.4853, + "step": 12464 + }, + { + "epoch": 4.376755617977528, + "grad_norm": 0.9646182656288147, + "learning_rate": 8.593252327064386e-07, + "loss": 0.5605, + "step": 12465 + }, + { + "epoch": 4.377106741573034, + "grad_norm": 0.9912760257720947, + "learning_rate": 8.589764242294896e-07, + "loss": 0.5586, + "step": 12466 + }, + { + "epoch": 4.377457865168539, + "grad_norm": 0.9669016599655151, + "learning_rate": 8.586276718741421e-07, + "loss": 0.5223, + "step": 12467 + }, + { + "epoch": 4.377808988764045, + "grad_norm": 0.9757612943649292, + "learning_rate": 8.582789756523241e-07, + "loss": 0.5139, + "step": 12468 + }, + { + "epoch": 4.37816011235955, + "grad_norm": 0.939070999622345, + "learning_rate": 8.579303355759596e-07, + "loss": 0.5005, + "step": 12469 + }, + { + "epoch": 4.378511235955056, + "grad_norm": 0.9377471804618835, + "learning_rate": 8.575817516569734e-07, + "loss": 0.5005, + "step": 12470 + }, + { + "epoch": 4.378862359550562, + "grad_norm": 0.9484912157058716, + "learning_rate": 8.572332239072853e-07, + "loss": 0.5442, + "step": 12471 + }, + { + "epoch": 4.379213483146067, + "grad_norm": 0.9636515974998474, + "learning_rate": 8.568847523388158e-07, + "loss": 0.5476, + "step": 12472 + }, + { + "epoch": 4.379564606741573, + "grad_norm": 0.9795300960540771, + "learning_rate": 8.565363369634808e-07, + "loss": 0.5307, + "step": 12473 + }, + { + "epoch": 4.379915730337078, + "grad_norm": 0.9554656744003296, + "learning_rate": 8.561879777931975e-07, + "loss": 0.5181, + "step": 12474 + }, + { + "epoch": 4.380266853932584, + "grad_norm": 0.9723366498947144, + "learning_rate": 8.558396748398781e-07, + "loss": 0.5315, + "step": 12475 + }, + { + "epoch": 4.38061797752809, + "grad_norm": 0.9662505388259888, + "learning_rate": 8.554914281154356e-07, + "loss": 0.5415, + "step": 12476 + }, + { + "epoch": 4.380969101123595, + "grad_norm": 0.9867193102836609, + "learning_rate": 8.551432376317792e-07, + "loss": 0.5576, + "step": 12477 + }, + { + "epoch": 4.381320224719101, + "grad_norm": 0.9826772212982178, + "learning_rate": 8.547951034008156e-07, + "loss": 0.5482, + "step": 12478 + }, + { + "epoch": 4.381671348314606, + "grad_norm": 0.9744497537612915, + "learning_rate": 8.544470254344526e-07, + "loss": 0.5557, + "step": 12479 + }, + { + "epoch": 4.382022471910112, + "grad_norm": 0.9517459273338318, + "learning_rate": 8.54099003744592e-07, + "loss": 0.5619, + "step": 12480 + }, + { + "epoch": 4.382373595505618, + "grad_norm": 0.9961041808128357, + "learning_rate": 8.537510383431383e-07, + "loss": 0.572, + "step": 12481 + }, + { + "epoch": 4.382724719101123, + "grad_norm": 0.9764973521232605, + "learning_rate": 8.534031292419892e-07, + "loss": 0.5834, + "step": 12482 + }, + { + "epoch": 4.383075842696629, + "grad_norm": 0.9588496088981628, + "learning_rate": 8.530552764530453e-07, + "loss": 0.5453, + "step": 12483 + }, + { + "epoch": 4.383426966292135, + "grad_norm": 0.9828209280967712, + "learning_rate": 8.527074799882004e-07, + "loss": 0.535, + "step": 12484 + }, + { + "epoch": 4.383778089887641, + "grad_norm": 0.9616172313690186, + "learning_rate": 8.523597398593511e-07, + "loss": 0.5528, + "step": 12485 + }, + { + "epoch": 4.384129213483146, + "grad_norm": 0.995172917842865, + "learning_rate": 8.520120560783882e-07, + "loss": 0.5538, + "step": 12486 + }, + { + "epoch": 4.384480337078652, + "grad_norm": 0.9149042963981628, + "learning_rate": 8.516644286572032e-07, + "loss": 0.5403, + "step": 12487 + }, + { + "epoch": 4.384831460674158, + "grad_norm": 0.9662986993789673, + "learning_rate": 8.513168576076835e-07, + "loss": 0.526, + "step": 12488 + }, + { + "epoch": 4.385182584269663, + "grad_norm": 0.9925611019134521, + "learning_rate": 8.509693429417173e-07, + "loss": 0.545, + "step": 12489 + }, + { + "epoch": 4.385533707865169, + "grad_norm": 0.9567667245864868, + "learning_rate": 8.506218846711886e-07, + "loss": 0.5239, + "step": 12490 + }, + { + "epoch": 4.385884831460674, + "grad_norm": 1.0146085023880005, + "learning_rate": 8.502744828079792e-07, + "loss": 0.5567, + "step": 12491 + }, + { + "epoch": 4.38623595505618, + "grad_norm": 0.9255802631378174, + "learning_rate": 8.499271373639717e-07, + "loss": 0.4968, + "step": 12492 + }, + { + "epoch": 4.386587078651686, + "grad_norm": 0.9765677452087402, + "learning_rate": 8.495798483510431e-07, + "loss": 0.5314, + "step": 12493 + }, + { + "epoch": 4.386938202247191, + "grad_norm": 0.9631349444389343, + "learning_rate": 8.492326157810724e-07, + "loss": 0.5646, + "step": 12494 + }, + { + "epoch": 4.387289325842697, + "grad_norm": 0.9844106435775757, + "learning_rate": 8.488854396659327e-07, + "loss": 0.5585, + "step": 12495 + }, + { + "epoch": 4.387640449438202, + "grad_norm": 0.9956976771354675, + "learning_rate": 8.485383200174991e-07, + "loss": 0.4949, + "step": 12496 + }, + { + "epoch": 4.387991573033708, + "grad_norm": 0.9718688726425171, + "learning_rate": 8.481912568476408e-07, + "loss": 0.5606, + "step": 12497 + }, + { + "epoch": 4.388342696629214, + "grad_norm": 0.9656421542167664, + "learning_rate": 8.478442501682287e-07, + "loss": 0.5274, + "step": 12498 + }, + { + "epoch": 4.388693820224719, + "grad_norm": 1.016042947769165, + "learning_rate": 8.474972999911288e-07, + "loss": 0.5794, + "step": 12499 + }, + { + "epoch": 4.389044943820225, + "grad_norm": 0.9634303450584412, + "learning_rate": 8.471504063282082e-07, + "loss": 0.5404, + "step": 12500 + }, + { + "epoch": 4.38939606741573, + "grad_norm": 0.9668565988540649, + "learning_rate": 8.46803569191329e-07, + "loss": 0.5471, + "step": 12501 + }, + { + "epoch": 4.389747191011236, + "grad_norm": 0.9913643598556519, + "learning_rate": 8.464567885923525e-07, + "loss": 0.5598, + "step": 12502 + }, + { + "epoch": 4.390098314606742, + "grad_norm": 0.9420990943908691, + "learning_rate": 8.461100645431397e-07, + "loss": 0.5236, + "step": 12503 + }, + { + "epoch": 4.390449438202247, + "grad_norm": 0.9372364282608032, + "learning_rate": 8.457633970555465e-07, + "loss": 0.5266, + "step": 12504 + }, + { + "epoch": 4.390800561797753, + "grad_norm": 0.9351842999458313, + "learning_rate": 8.454167861414303e-07, + "loss": 0.534, + "step": 12505 + }, + { + "epoch": 4.391151685393258, + "grad_norm": 0.9718179106712341, + "learning_rate": 8.450702318126436e-07, + "loss": 0.5715, + "step": 12506 + }, + { + "epoch": 4.391502808988764, + "grad_norm": 0.9804518818855286, + "learning_rate": 8.447237340810394e-07, + "loss": 0.5656, + "step": 12507 + }, + { + "epoch": 4.39185393258427, + "grad_norm": 0.9758808612823486, + "learning_rate": 8.443772929584665e-07, + "loss": 0.5478, + "step": 12508 + }, + { + "epoch": 4.392205056179775, + "grad_norm": 0.9675325751304626, + "learning_rate": 8.440309084567741e-07, + "loss": 0.5408, + "step": 12509 + }, + { + "epoch": 4.392556179775281, + "grad_norm": 1.0077381134033203, + "learning_rate": 8.436845805878066e-07, + "loss": 0.5722, + "step": 12510 + }, + { + "epoch": 4.392907303370786, + "grad_norm": 0.9374393224716187, + "learning_rate": 8.4333830936341e-07, + "loss": 0.5486, + "step": 12511 + }, + { + "epoch": 4.393258426966292, + "grad_norm": 0.946722149848938, + "learning_rate": 8.429920947954251e-07, + "loss": 0.5321, + "step": 12512 + }, + { + "epoch": 4.393609550561798, + "grad_norm": 1.000594973564148, + "learning_rate": 8.426459368956933e-07, + "loss": 0.5496, + "step": 12513 + }, + { + "epoch": 4.393960674157303, + "grad_norm": 0.9472278356552124, + "learning_rate": 8.422998356760523e-07, + "loss": 0.5386, + "step": 12514 + }, + { + "epoch": 4.394311797752809, + "grad_norm": 0.9738829731941223, + "learning_rate": 8.419537911483375e-07, + "loss": 0.5273, + "step": 12515 + }, + { + "epoch": 4.394662921348314, + "grad_norm": 0.9831490516662598, + "learning_rate": 8.416078033243849e-07, + "loss": 0.5387, + "step": 12516 + }, + { + "epoch": 4.39501404494382, + "grad_norm": 0.9683114886283875, + "learning_rate": 8.412618722160257e-07, + "loss": 0.5667, + "step": 12517 + }, + { + "epoch": 4.395365168539326, + "grad_norm": 0.9630364775657654, + "learning_rate": 8.409159978350917e-07, + "loss": 0.5498, + "step": 12518 + }, + { + "epoch": 4.395716292134831, + "grad_norm": 0.9886455535888672, + "learning_rate": 8.405701801934102e-07, + "loss": 0.5565, + "step": 12519 + }, + { + "epoch": 4.396067415730337, + "grad_norm": 0.9638479948043823, + "learning_rate": 8.402244193028095e-07, + "loss": 0.536, + "step": 12520 + }, + { + "epoch": 4.396418539325842, + "grad_norm": 1.0103294849395752, + "learning_rate": 8.398787151751123e-07, + "loss": 0.5534, + "step": 12521 + }, + { + "epoch": 4.396769662921348, + "grad_norm": 0.971401035785675, + "learning_rate": 8.395330678221434e-07, + "loss": 0.4764, + "step": 12522 + }, + { + "epoch": 4.397120786516854, + "grad_norm": 0.9659348130226135, + "learning_rate": 8.391874772557218e-07, + "loss": 0.5309, + "step": 12523 + }, + { + "epoch": 4.397471910112359, + "grad_norm": 0.9599688053131104, + "learning_rate": 8.388419434876682e-07, + "loss": 0.5465, + "step": 12524 + }, + { + "epoch": 4.397823033707865, + "grad_norm": 0.9864976406097412, + "learning_rate": 8.384964665297976e-07, + "loss": 0.5456, + "step": 12525 + }, + { + "epoch": 4.398174157303371, + "grad_norm": 0.9649502038955688, + "learning_rate": 8.381510463939271e-07, + "loss": 0.5249, + "step": 12526 + }, + { + "epoch": 4.398525280898877, + "grad_norm": 0.9375289082527161, + "learning_rate": 8.378056830918688e-07, + "loss": 0.5201, + "step": 12527 + }, + { + "epoch": 4.398876404494382, + "grad_norm": 0.9570899605751038, + "learning_rate": 8.374603766354331e-07, + "loss": 0.5667, + "step": 12528 + }, + { + "epoch": 4.399227528089888, + "grad_norm": 0.9472808837890625, + "learning_rate": 8.371151270364303e-07, + "loss": 0.512, + "step": 12529 + }, + { + "epoch": 4.399578651685394, + "grad_norm": 0.9564096331596375, + "learning_rate": 8.367699343066666e-07, + "loss": 0.532, + "step": 12530 + }, + { + "epoch": 4.399929775280899, + "grad_norm": 0.9419234991073608, + "learning_rate": 8.364247984579488e-07, + "loss": 0.5367, + "step": 12531 + }, + { + "epoch": 4.400280898876405, + "grad_norm": 0.9547091126441956, + "learning_rate": 8.360797195020784e-07, + "loss": 0.5413, + "step": 12532 + }, + { + "epoch": 4.40063202247191, + "grad_norm": 0.9635790586471558, + "learning_rate": 8.357346974508585e-07, + "loss": 0.5179, + "step": 12533 + }, + { + "epoch": 4.400983146067416, + "grad_norm": 0.9510557651519775, + "learning_rate": 8.35389732316087e-07, + "loss": 0.5307, + "step": 12534 + }, + { + "epoch": 4.401334269662922, + "grad_norm": 0.9837446212768555, + "learning_rate": 8.350448241095632e-07, + "loss": 0.5099, + "step": 12535 + }, + { + "epoch": 4.401685393258427, + "grad_norm": 0.9674659371376038, + "learning_rate": 8.34699972843081e-07, + "loss": 0.5454, + "step": 12536 + }, + { + "epoch": 4.402036516853933, + "grad_norm": 0.9770947098731995, + "learning_rate": 8.343551785284351e-07, + "loss": 0.5622, + "step": 12537 + }, + { + "epoch": 4.402387640449438, + "grad_norm": 1.0124224424362183, + "learning_rate": 8.340104411774164e-07, + "loss": 0.5417, + "step": 12538 + }, + { + "epoch": 4.402738764044944, + "grad_norm": 0.9915936589241028, + "learning_rate": 8.336657608018156e-07, + "loss": 0.5759, + "step": 12539 + }, + { + "epoch": 4.40308988764045, + "grad_norm": 0.9608220458030701, + "learning_rate": 8.333211374134198e-07, + "loss": 0.5541, + "step": 12540 + }, + { + "epoch": 4.403441011235955, + "grad_norm": 1.0303380489349365, + "learning_rate": 8.329765710240143e-07, + "loss": 0.5555, + "step": 12541 + }, + { + "epoch": 4.403792134831461, + "grad_norm": 0.9943733215332031, + "learning_rate": 8.326320616453843e-07, + "loss": 0.5687, + "step": 12542 + }, + { + "epoch": 4.404143258426966, + "grad_norm": 0.985022783279419, + "learning_rate": 8.3228760928931e-07, + "loss": 0.5683, + "step": 12543 + }, + { + "epoch": 4.404494382022472, + "grad_norm": 0.9831377863883972, + "learning_rate": 8.319432139675734e-07, + "loss": 0.5538, + "step": 12544 + }, + { + "epoch": 4.404845505617978, + "grad_norm": 0.9781966209411621, + "learning_rate": 8.315988756919507e-07, + "loss": 0.5568, + "step": 12545 + }, + { + "epoch": 4.405196629213483, + "grad_norm": 0.9336716532707214, + "learning_rate": 8.312545944742196e-07, + "loss": 0.4699, + "step": 12546 + }, + { + "epoch": 4.405547752808989, + "grad_norm": 0.9537749290466309, + "learning_rate": 8.309103703261526e-07, + "loss": 0.5271, + "step": 12547 + }, + { + "epoch": 4.405898876404494, + "grad_norm": 0.9865443706512451, + "learning_rate": 8.305662032595235e-07, + "loss": 0.5573, + "step": 12548 + }, + { + "epoch": 4.40625, + "grad_norm": 0.9995547533035278, + "learning_rate": 8.302220932861007e-07, + "loss": 0.5396, + "step": 12549 + }, + { + "epoch": 4.406601123595506, + "grad_norm": 0.9657871127128601, + "learning_rate": 8.298780404176545e-07, + "loss": 0.538, + "step": 12550 + }, + { + "epoch": 4.406952247191011, + "grad_norm": 0.9776629209518433, + "learning_rate": 8.295340446659492e-07, + "loss": 0.5321, + "step": 12551 + }, + { + "epoch": 4.407303370786517, + "grad_norm": 0.9681153297424316, + "learning_rate": 8.29190106042751e-07, + "loss": 0.5843, + "step": 12552 + }, + { + "epoch": 4.407654494382022, + "grad_norm": 1.0087169408798218, + "learning_rate": 8.288462245598214e-07, + "loss": 0.5396, + "step": 12553 + }, + { + "epoch": 4.408005617977528, + "grad_norm": 0.9754542112350464, + "learning_rate": 8.285024002289202e-07, + "loss": 0.545, + "step": 12554 + }, + { + "epoch": 4.408356741573034, + "grad_norm": 0.9784188270568848, + "learning_rate": 8.281586330618072e-07, + "loss": 0.5199, + "step": 12555 + }, + { + "epoch": 4.408707865168539, + "grad_norm": 0.9797220230102539, + "learning_rate": 8.278149230702376e-07, + "loss": 0.5322, + "step": 12556 + }, + { + "epoch": 4.409058988764045, + "grad_norm": 0.9651806950569153, + "learning_rate": 8.274712702659676e-07, + "loss": 0.5748, + "step": 12557 + }, + { + "epoch": 4.40941011235955, + "grad_norm": 0.9760322570800781, + "learning_rate": 8.271276746607482e-07, + "loss": 0.5381, + "step": 12558 + }, + { + "epoch": 4.409761235955056, + "grad_norm": 0.9374415278434753, + "learning_rate": 8.267841362663314e-07, + "loss": 0.504, + "step": 12559 + }, + { + "epoch": 4.410112359550562, + "grad_norm": 0.9144535064697266, + "learning_rate": 8.264406550944648e-07, + "loss": 0.5308, + "step": 12560 + }, + { + "epoch": 4.410463483146067, + "grad_norm": 0.9503322839736938, + "learning_rate": 8.260972311568966e-07, + "loss": 0.5272, + "step": 12561 + }, + { + "epoch": 4.410814606741573, + "grad_norm": 0.9367659687995911, + "learning_rate": 8.257538644653696e-07, + "loss": 0.509, + "step": 12562 + }, + { + "epoch": 4.411165730337078, + "grad_norm": 0.9585871696472168, + "learning_rate": 8.254105550316288e-07, + "loss": 0.5118, + "step": 12563 + }, + { + "epoch": 4.411516853932584, + "grad_norm": 0.9992086887359619, + "learning_rate": 8.250673028674139e-07, + "loss": 0.5397, + "step": 12564 + }, + { + "epoch": 4.41186797752809, + "grad_norm": 0.9538074731826782, + "learning_rate": 8.247241079844631e-07, + "loss": 0.5286, + "step": 12565 + }, + { + "epoch": 4.412219101123595, + "grad_norm": 0.9609680771827698, + "learning_rate": 8.243809703945152e-07, + "loss": 0.5404, + "step": 12566 + }, + { + "epoch": 4.412570224719101, + "grad_norm": 0.9669106602668762, + "learning_rate": 8.240378901093035e-07, + "loss": 0.5434, + "step": 12567 + }, + { + "epoch": 4.412921348314606, + "grad_norm": 1.0268278121948242, + "learning_rate": 8.236948671405625e-07, + "loss": 0.5784, + "step": 12568 + }, + { + "epoch": 4.413272471910112, + "grad_norm": 0.9862893223762512, + "learning_rate": 8.233519015000219e-07, + "loss": 0.5519, + "step": 12569 + }, + { + "epoch": 4.413623595505618, + "grad_norm": 0.9993818402290344, + "learning_rate": 8.230089931994123e-07, + "loss": 0.5302, + "step": 12570 + }, + { + "epoch": 4.413974719101123, + "grad_norm": 0.9311320781707764, + "learning_rate": 8.226661422504593e-07, + "loss": 0.5231, + "step": 12571 + }, + { + "epoch": 4.414325842696629, + "grad_norm": 0.9299443364143372, + "learning_rate": 8.223233486648899e-07, + "loss": 0.5243, + "step": 12572 + }, + { + "epoch": 4.414676966292135, + "grad_norm": 0.956378698348999, + "learning_rate": 8.219806124544261e-07, + "loss": 0.4992, + "step": 12573 + }, + { + "epoch": 4.415028089887641, + "grad_norm": 0.9526459574699402, + "learning_rate": 8.216379336307887e-07, + "loss": 0.5348, + "step": 12574 + }, + { + "epoch": 4.415379213483146, + "grad_norm": 0.9636205434799194, + "learning_rate": 8.212953122056988e-07, + "loss": 0.5322, + "step": 12575 + }, + { + "epoch": 4.415730337078652, + "grad_norm": 0.9841729998588562, + "learning_rate": 8.209527481908716e-07, + "loss": 0.5612, + "step": 12576 + }, + { + "epoch": 4.416081460674158, + "grad_norm": 0.9542888402938843, + "learning_rate": 8.206102415980247e-07, + "loss": 0.5287, + "step": 12577 + }, + { + "epoch": 4.416432584269663, + "grad_norm": 0.9706204533576965, + "learning_rate": 8.202677924388694e-07, + "loss": 0.5135, + "step": 12578 + }, + { + "epoch": 4.416783707865169, + "grad_norm": 0.9357458353042603, + "learning_rate": 8.199254007251192e-07, + "loss": 0.5261, + "step": 12579 + }, + { + "epoch": 4.417134831460674, + "grad_norm": 0.9412339329719543, + "learning_rate": 8.19583066468482e-07, + "loss": 0.5401, + "step": 12580 + }, + { + "epoch": 4.41748595505618, + "grad_norm": 0.9440373778343201, + "learning_rate": 8.192407896806664e-07, + "loss": 0.5381, + "step": 12581 + }, + { + "epoch": 4.417837078651686, + "grad_norm": 0.9799615144729614, + "learning_rate": 8.188985703733778e-07, + "loss": 0.5203, + "step": 12582 + }, + { + "epoch": 4.418188202247191, + "grad_norm": 0.9847771525382996, + "learning_rate": 8.185564085583189e-07, + "loss": 0.5715, + "step": 12583 + }, + { + "epoch": 4.418539325842697, + "grad_norm": 0.9074743986129761, + "learning_rate": 8.182143042471924e-07, + "loss": 0.4938, + "step": 12584 + }, + { + "epoch": 4.418890449438202, + "grad_norm": 0.9650834798812866, + "learning_rate": 8.17872257451697e-07, + "loss": 0.5409, + "step": 12585 + }, + { + "epoch": 4.419241573033708, + "grad_norm": 1.0166469812393188, + "learning_rate": 8.175302681835317e-07, + "loss": 0.529, + "step": 12586 + }, + { + "epoch": 4.419592696629214, + "grad_norm": 0.9783175587654114, + "learning_rate": 8.171883364543909e-07, + "loss": 0.566, + "step": 12587 + }, + { + "epoch": 4.419943820224719, + "grad_norm": 0.9633625745773315, + "learning_rate": 8.168464622759695e-07, + "loss": 0.5406, + "step": 12588 + }, + { + "epoch": 4.420294943820225, + "grad_norm": 0.9770609140396118, + "learning_rate": 8.16504645659959e-07, + "loss": 0.5466, + "step": 12589 + }, + { + "epoch": 4.42064606741573, + "grad_norm": 0.9533287286758423, + "learning_rate": 8.161628866180482e-07, + "loss": 0.5385, + "step": 12590 + }, + { + "epoch": 4.420997191011236, + "grad_norm": 0.9716050028800964, + "learning_rate": 8.158211851619266e-07, + "loss": 0.5481, + "step": 12591 + }, + { + "epoch": 4.421348314606742, + "grad_norm": 0.9453951120376587, + "learning_rate": 8.154795413032784e-07, + "loss": 0.5101, + "step": 12592 + }, + { + "epoch": 4.421699438202247, + "grad_norm": 0.964753270149231, + "learning_rate": 8.151379550537894e-07, + "loss": 0.5432, + "step": 12593 + }, + { + "epoch": 4.422050561797753, + "grad_norm": 0.9370836615562439, + "learning_rate": 8.147964264251396e-07, + "loss": 0.5231, + "step": 12594 + }, + { + "epoch": 4.422401685393258, + "grad_norm": 0.9723945260047913, + "learning_rate": 8.14454955429011e-07, + "loss": 0.5316, + "step": 12595 + }, + { + "epoch": 4.422752808988764, + "grad_norm": 0.9751042127609253, + "learning_rate": 8.141135420770804e-07, + "loss": 0.5476, + "step": 12596 + }, + { + "epoch": 4.42310393258427, + "grad_norm": 1.012180209159851, + "learning_rate": 8.137721863810233e-07, + "loss": 0.5729, + "step": 12597 + }, + { + "epoch": 4.423455056179775, + "grad_norm": 0.9741883873939514, + "learning_rate": 8.134308883525153e-07, + "loss": 0.5302, + "step": 12598 + }, + { + "epoch": 4.423806179775281, + "grad_norm": 0.9805081486701965, + "learning_rate": 8.130896480032266e-07, + "loss": 0.5288, + "step": 12599 + }, + { + "epoch": 4.424157303370786, + "grad_norm": 0.9474637508392334, + "learning_rate": 8.127484653448295e-07, + "loss": 0.5571, + "step": 12600 + }, + { + "epoch": 4.424508426966292, + "grad_norm": 0.9537769556045532, + "learning_rate": 8.124073403889903e-07, + "loss": 0.5405, + "step": 12601 + }, + { + "epoch": 4.424859550561798, + "grad_norm": 0.9508486986160278, + "learning_rate": 8.120662731473766e-07, + "loss": 0.5388, + "step": 12602 + }, + { + "epoch": 4.425210674157303, + "grad_norm": 0.9450559616088867, + "learning_rate": 8.117252636316522e-07, + "loss": 0.5142, + "step": 12603 + }, + { + "epoch": 4.425561797752809, + "grad_norm": 0.9901648759841919, + "learning_rate": 8.113843118534784e-07, + "loss": 0.5563, + "step": 12604 + }, + { + "epoch": 4.425912921348314, + "grad_norm": 0.9724363684654236, + "learning_rate": 8.110434178245166e-07, + "loss": 0.5436, + "step": 12605 + }, + { + "epoch": 4.42626404494382, + "grad_norm": 1.0161991119384766, + "learning_rate": 8.10702581556424e-07, + "loss": 0.5214, + "step": 12606 + }, + { + "epoch": 4.426615168539326, + "grad_norm": 0.9480384588241577, + "learning_rate": 8.103618030608584e-07, + "loss": 0.5444, + "step": 12607 + }, + { + "epoch": 4.426966292134831, + "grad_norm": 0.9667114615440369, + "learning_rate": 8.100210823494726e-07, + "loss": 0.5368, + "step": 12608 + }, + { + "epoch": 4.427317415730337, + "grad_norm": 0.9850381016731262, + "learning_rate": 8.096804194339203e-07, + "loss": 0.5811, + "step": 12609 + }, + { + "epoch": 4.427668539325842, + "grad_norm": 0.9601279497146606, + "learning_rate": 8.093398143258504e-07, + "loss": 0.5351, + "step": 12610 + }, + { + "epoch": 4.428019662921348, + "grad_norm": 1.1035670042037964, + "learning_rate": 8.089992670369128e-07, + "loss": 0.5409, + "step": 12611 + }, + { + "epoch": 4.428370786516854, + "grad_norm": 0.9532591104507446, + "learning_rate": 8.086587775787527e-07, + "loss": 0.5212, + "step": 12612 + }, + { + "epoch": 4.428721910112359, + "grad_norm": 0.9347667098045349, + "learning_rate": 8.083183459630159e-07, + "loss": 0.5247, + "step": 12613 + }, + { + "epoch": 4.429073033707865, + "grad_norm": 0.9657343626022339, + "learning_rate": 8.079779722013434e-07, + "loss": 0.5449, + "step": 12614 + }, + { + "epoch": 4.429424157303371, + "grad_norm": 1.0073562860488892, + "learning_rate": 8.076376563053767e-07, + "loss": 0.5382, + "step": 12615 + }, + { + "epoch": 4.429775280898877, + "grad_norm": 0.9423131346702576, + "learning_rate": 8.072973982867544e-07, + "loss": 0.4748, + "step": 12616 + }, + { + "epoch": 4.430126404494382, + "grad_norm": 0.9727164506912231, + "learning_rate": 8.069571981571117e-07, + "loss": 0.5736, + "step": 12617 + }, + { + "epoch": 4.430477528089888, + "grad_norm": 0.9790947437286377, + "learning_rate": 8.066170559280848e-07, + "loss": 0.5544, + "step": 12618 + }, + { + "epoch": 4.430828651685394, + "grad_norm": 0.9386008977890015, + "learning_rate": 8.062769716113048e-07, + "loss": 0.5046, + "step": 12619 + }, + { + "epoch": 4.431179775280899, + "grad_norm": 0.9868724346160889, + "learning_rate": 8.059369452184038e-07, + "loss": 0.5385, + "step": 12620 + }, + { + "epoch": 4.431530898876405, + "grad_norm": 0.996635913848877, + "learning_rate": 8.055969767610089e-07, + "loss": 0.5462, + "step": 12621 + }, + { + "epoch": 4.43188202247191, + "grad_norm": 0.9695373773574829, + "learning_rate": 8.052570662507481e-07, + "loss": 0.5642, + "step": 12622 + }, + { + "epoch": 4.432233146067416, + "grad_norm": 0.9807968139648438, + "learning_rate": 8.049172136992447e-07, + "loss": 0.5384, + "step": 12623 + }, + { + "epoch": 4.432584269662922, + "grad_norm": 0.9779884815216064, + "learning_rate": 8.04577419118123e-07, + "loss": 0.5637, + "step": 12624 + }, + { + "epoch": 4.432935393258427, + "grad_norm": 0.9662644267082214, + "learning_rate": 8.04237682519002e-07, + "loss": 0.5066, + "step": 12625 + }, + { + "epoch": 4.433286516853933, + "grad_norm": 0.9872508645057678, + "learning_rate": 8.03898003913502e-07, + "loss": 0.5328, + "step": 12626 + }, + { + "epoch": 4.433637640449438, + "grad_norm": 0.9769915342330933, + "learning_rate": 8.035583833132388e-07, + "loss": 0.5594, + "step": 12627 + }, + { + "epoch": 4.433988764044944, + "grad_norm": 0.9504956007003784, + "learning_rate": 8.032188207298264e-07, + "loss": 0.5277, + "step": 12628 + }, + { + "epoch": 4.43433988764045, + "grad_norm": 0.9735992550849915, + "learning_rate": 8.028793161748791e-07, + "loss": 0.5571, + "step": 12629 + }, + { + "epoch": 4.434691011235955, + "grad_norm": 0.9334622025489807, + "learning_rate": 8.025398696600062e-07, + "loss": 0.5368, + "step": 12630 + }, + { + "epoch": 4.435042134831461, + "grad_norm": 0.941819429397583, + "learning_rate": 8.022004811968179e-07, + "loss": 0.5307, + "step": 12631 + }, + { + "epoch": 4.435393258426966, + "grad_norm": 0.963756799697876, + "learning_rate": 8.018611507969198e-07, + "loss": 0.548, + "step": 12632 + }, + { + "epoch": 4.435744382022472, + "grad_norm": 0.9738016724586487, + "learning_rate": 8.015218784719178e-07, + "loss": 0.5388, + "step": 12633 + }, + { + "epoch": 4.436095505617978, + "grad_norm": 0.9689740538597107, + "learning_rate": 8.011826642334133e-07, + "loss": 0.5472, + "step": 12634 + }, + { + "epoch": 4.436446629213483, + "grad_norm": 0.991801381111145, + "learning_rate": 8.008435080930091e-07, + "loss": 0.5433, + "step": 12635 + }, + { + "epoch": 4.436797752808989, + "grad_norm": 0.9678930640220642, + "learning_rate": 8.005044100623019e-07, + "loss": 0.5326, + "step": 12636 + }, + { + "epoch": 4.437148876404494, + "grad_norm": 0.9817230105400085, + "learning_rate": 8.001653701528902e-07, + "loss": 0.5459, + "step": 12637 + }, + { + "epoch": 4.4375, + "grad_norm": 0.9614975452423096, + "learning_rate": 7.998263883763679e-07, + "loss": 0.5942, + "step": 12638 + }, + { + "epoch": 4.437851123595506, + "grad_norm": 0.9712573885917664, + "learning_rate": 7.994874647443288e-07, + "loss": 0.5577, + "step": 12639 + }, + { + "epoch": 4.438202247191011, + "grad_norm": 0.9822211265563965, + "learning_rate": 7.991485992683634e-07, + "loss": 0.5548, + "step": 12640 + }, + { + "epoch": 4.438553370786517, + "grad_norm": 0.9682799577713013, + "learning_rate": 7.988097919600596e-07, + "loss": 0.5437, + "step": 12641 + }, + { + "epoch": 4.438904494382022, + "grad_norm": 0.9659839272499084, + "learning_rate": 7.984710428310061e-07, + "loss": 0.5233, + "step": 12642 + }, + { + "epoch": 4.439255617977528, + "grad_norm": 0.9687818288803101, + "learning_rate": 7.981323518927858e-07, + "loss": 0.551, + "step": 12643 + }, + { + "epoch": 4.439606741573034, + "grad_norm": 0.9474412202835083, + "learning_rate": 7.977937191569837e-07, + "loss": 0.5387, + "step": 12644 + }, + { + "epoch": 4.439957865168539, + "grad_norm": 0.9847413897514343, + "learning_rate": 7.974551446351791e-07, + "loss": 0.5407, + "step": 12645 + }, + { + "epoch": 4.440308988764045, + "grad_norm": 0.9790306091308594, + "learning_rate": 7.971166283389525e-07, + "loss": 0.5297, + "step": 12646 + }, + { + "epoch": 4.44066011235955, + "grad_norm": 1.0515199899673462, + "learning_rate": 7.967781702798791e-07, + "loss": 0.5465, + "step": 12647 + }, + { + "epoch": 4.441011235955056, + "grad_norm": 0.9726919531822205, + "learning_rate": 7.964397704695356e-07, + "loss": 0.5236, + "step": 12648 + }, + { + "epoch": 4.441362359550562, + "grad_norm": 1.0173555612564087, + "learning_rate": 7.961014289194935e-07, + "loss": 0.5432, + "step": 12649 + }, + { + "epoch": 4.441713483146067, + "grad_norm": 0.9604166746139526, + "learning_rate": 7.957631456413254e-07, + "loss": 0.5313, + "step": 12650 + }, + { + "epoch": 4.442064606741573, + "grad_norm": 0.9685438275337219, + "learning_rate": 7.954249206465983e-07, + "loss": 0.5581, + "step": 12651 + }, + { + "epoch": 4.442415730337078, + "grad_norm": 0.9537839293479919, + "learning_rate": 7.950867539468812e-07, + "loss": 0.5194, + "step": 12652 + }, + { + "epoch": 4.442766853932584, + "grad_norm": 0.9733744859695435, + "learning_rate": 7.947486455537381e-07, + "loss": 0.5543, + "step": 12653 + }, + { + "epoch": 4.44311797752809, + "grad_norm": 0.969420313835144, + "learning_rate": 7.944105954787316e-07, + "loss": 0.5394, + "step": 12654 + }, + { + "epoch": 4.443469101123595, + "grad_norm": 0.9809862375259399, + "learning_rate": 7.940726037334237e-07, + "loss": 0.5457, + "step": 12655 + }, + { + "epoch": 4.443820224719101, + "grad_norm": 0.9773897528648376, + "learning_rate": 7.937346703293725e-07, + "loss": 0.5577, + "step": 12656 + }, + { + "epoch": 4.444171348314606, + "grad_norm": 0.96278977394104, + "learning_rate": 7.93396795278136e-07, + "loss": 0.5455, + "step": 12657 + }, + { + "epoch": 4.444522471910112, + "grad_norm": 0.9850781559944153, + "learning_rate": 7.930589785912682e-07, + "loss": 0.5707, + "step": 12658 + }, + { + "epoch": 4.444873595505618, + "grad_norm": 0.9569475650787354, + "learning_rate": 7.927212202803234e-07, + "loss": 0.5387, + "step": 12659 + }, + { + "epoch": 4.445224719101123, + "grad_norm": 0.94439697265625, + "learning_rate": 7.923835203568508e-07, + "loss": 0.5029, + "step": 12660 + }, + { + "epoch": 4.445575842696629, + "grad_norm": 0.9836904406547546, + "learning_rate": 7.920458788324017e-07, + "loss": 0.5322, + "step": 12661 + }, + { + "epoch": 4.445926966292135, + "grad_norm": 0.9911874532699585, + "learning_rate": 7.917082957185212e-07, + "loss": 0.5477, + "step": 12662 + }, + { + "epoch": 4.446278089887641, + "grad_norm": 0.9709537029266357, + "learning_rate": 7.91370771026756e-07, + "loss": 0.5372, + "step": 12663 + }, + { + "epoch": 4.446629213483146, + "grad_norm": 0.9764455556869507, + "learning_rate": 7.910333047686475e-07, + "loss": 0.5229, + "step": 12664 + }, + { + "epoch": 4.446980337078652, + "grad_norm": 0.9782695174217224, + "learning_rate": 7.906958969557386e-07, + "loss": 0.5376, + "step": 12665 + }, + { + "epoch": 4.447331460674158, + "grad_norm": 0.9481970071792603, + "learning_rate": 7.90358547599567e-07, + "loss": 0.5212, + "step": 12666 + }, + { + "epoch": 4.447682584269663, + "grad_norm": 0.9681141376495361, + "learning_rate": 7.900212567116694e-07, + "loss": 0.543, + "step": 12667 + }, + { + "epoch": 4.448033707865169, + "grad_norm": 0.9878993630409241, + "learning_rate": 7.896840243035825e-07, + "loss": 0.5407, + "step": 12668 + }, + { + "epoch": 4.448384831460674, + "grad_norm": 0.9778482913970947, + "learning_rate": 7.893468503868376e-07, + "loss": 0.5875, + "step": 12669 + }, + { + "epoch": 4.44873595505618, + "grad_norm": 0.9612657427787781, + "learning_rate": 7.890097349729672e-07, + "loss": 0.5398, + "step": 12670 + }, + { + "epoch": 4.449087078651686, + "grad_norm": 0.9734082818031311, + "learning_rate": 7.88672678073499e-07, + "loss": 0.5484, + "step": 12671 + }, + { + "epoch": 4.449438202247191, + "grad_norm": 0.9591443538665771, + "learning_rate": 7.88335679699962e-07, + "loss": 0.517, + "step": 12672 + }, + { + "epoch": 4.449789325842697, + "grad_norm": 0.9493557214736938, + "learning_rate": 7.879987398638788e-07, + "loss": 0.5293, + "step": 12673 + }, + { + "epoch": 4.450140449438202, + "grad_norm": 0.9941859245300293, + "learning_rate": 7.876618585767748e-07, + "loss": 0.5437, + "step": 12674 + }, + { + "epoch": 4.450491573033708, + "grad_norm": 0.9591066837310791, + "learning_rate": 7.873250358501688e-07, + "loss": 0.5355, + "step": 12675 + }, + { + "epoch": 4.450842696629214, + "grad_norm": 0.9417831301689148, + "learning_rate": 7.869882716955824e-07, + "loss": 0.5272, + "step": 12676 + }, + { + "epoch": 4.451193820224719, + "grad_norm": 0.9434447884559631, + "learning_rate": 7.8665156612453e-07, + "loss": 0.5366, + "step": 12677 + }, + { + "epoch": 4.451544943820225, + "grad_norm": 0.989935040473938, + "learning_rate": 7.863149191485289e-07, + "loss": 0.5481, + "step": 12678 + }, + { + "epoch": 4.45189606741573, + "grad_norm": 0.9539551138877869, + "learning_rate": 7.859783307790911e-07, + "loss": 0.5451, + "step": 12679 + }, + { + "epoch": 4.452247191011236, + "grad_norm": 0.9888069033622742, + "learning_rate": 7.856418010277267e-07, + "loss": 0.5516, + "step": 12680 + }, + { + "epoch": 4.452598314606742, + "grad_norm": 0.9452155232429504, + "learning_rate": 7.853053299059468e-07, + "loss": 0.5233, + "step": 12681 + }, + { + "epoch": 4.452949438202247, + "grad_norm": 0.9560648202896118, + "learning_rate": 7.849689174252562e-07, + "loss": 0.5454, + "step": 12682 + }, + { + "epoch": 4.453300561797753, + "grad_norm": 0.9505792856216431, + "learning_rate": 7.846325635971622e-07, + "loss": 0.5681, + "step": 12683 + }, + { + "epoch": 4.453651685393258, + "grad_norm": 0.9678177833557129, + "learning_rate": 7.842962684331656e-07, + "loss": 0.5644, + "step": 12684 + }, + { + "epoch": 4.454002808988764, + "grad_norm": 0.9675583243370056, + "learning_rate": 7.839600319447693e-07, + "loss": 0.5428, + "step": 12685 + }, + { + "epoch": 4.45435393258427, + "grad_norm": 0.9662092328071594, + "learning_rate": 7.83623854143471e-07, + "loss": 0.5245, + "step": 12686 + }, + { + "epoch": 4.454705056179775, + "grad_norm": 0.9554389715194702, + "learning_rate": 7.832877350407686e-07, + "loss": 0.5362, + "step": 12687 + }, + { + "epoch": 4.455056179775281, + "grad_norm": 0.9589457511901855, + "learning_rate": 7.829516746481561e-07, + "loss": 0.5327, + "step": 12688 + }, + { + "epoch": 4.455407303370786, + "grad_norm": 0.978738009929657, + "learning_rate": 7.826156729771275e-07, + "loss": 0.5509, + "step": 12689 + }, + { + "epoch": 4.455758426966292, + "grad_norm": 0.9884186387062073, + "learning_rate": 7.822797300391735e-07, + "loss": 0.5543, + "step": 12690 + }, + { + "epoch": 4.456109550561798, + "grad_norm": 0.9775454998016357, + "learning_rate": 7.81943845845782e-07, + "loss": 0.5773, + "step": 12691 + }, + { + "epoch": 4.456460674157303, + "grad_norm": 0.970535933971405, + "learning_rate": 7.816080204084417e-07, + "loss": 0.5293, + "step": 12692 + }, + { + "epoch": 4.456811797752809, + "grad_norm": 0.954775333404541, + "learning_rate": 7.812722537386358e-07, + "loss": 0.5558, + "step": 12693 + }, + { + "epoch": 4.457162921348314, + "grad_norm": 0.9654307961463928, + "learning_rate": 7.809365458478491e-07, + "loss": 0.555, + "step": 12694 + }, + { + "epoch": 4.45751404494382, + "grad_norm": 0.9787588119506836, + "learning_rate": 7.806008967475603e-07, + "loss": 0.5547, + "step": 12695 + }, + { + "epoch": 4.457865168539326, + "grad_norm": 0.9878897666931152, + "learning_rate": 7.802653064492507e-07, + "loss": 0.5814, + "step": 12696 + }, + { + "epoch": 4.458216292134831, + "grad_norm": 0.9414510130882263, + "learning_rate": 7.79929774964395e-07, + "loss": 0.5267, + "step": 12697 + }, + { + "epoch": 4.458567415730337, + "grad_norm": 0.9772969484329224, + "learning_rate": 7.7959430230447e-07, + "loss": 0.5597, + "step": 12698 + }, + { + "epoch": 4.458918539325842, + "grad_norm": 0.9613645672798157, + "learning_rate": 7.792588884809468e-07, + "loss": 0.5141, + "step": 12699 + }, + { + "epoch": 4.459269662921348, + "grad_norm": 0.9718449115753174, + "learning_rate": 7.78923533505298e-07, + "loss": 0.5596, + "step": 12700 + }, + { + "epoch": 4.459620786516854, + "grad_norm": 0.9787622094154358, + "learning_rate": 7.785882373889908e-07, + "loss": 0.5633, + "step": 12701 + }, + { + "epoch": 4.459971910112359, + "grad_norm": 0.9689585566520691, + "learning_rate": 7.782530001434937e-07, + "loss": 0.5296, + "step": 12702 + }, + { + "epoch": 4.460323033707865, + "grad_norm": 0.9697659611701965, + "learning_rate": 7.779178217802705e-07, + "loss": 0.5446, + "step": 12703 + }, + { + "epoch": 4.460674157303371, + "grad_norm": 1.0013611316680908, + "learning_rate": 7.775827023107835e-07, + "loss": 0.5437, + "step": 12704 + }, + { + "epoch": 4.461025280898877, + "grad_norm": 0.9915167093276978, + "learning_rate": 7.772476417464953e-07, + "loss": 0.5193, + "step": 12705 + }, + { + "epoch": 4.461376404494382, + "grad_norm": 0.9753875136375427, + "learning_rate": 7.769126400988622e-07, + "loss": 0.5802, + "step": 12706 + }, + { + "epoch": 4.461727528089888, + "grad_norm": 0.9680429100990295, + "learning_rate": 7.765776973793435e-07, + "loss": 0.5313, + "step": 12707 + }, + { + "epoch": 4.462078651685394, + "grad_norm": 0.9697384238243103, + "learning_rate": 7.762428135993918e-07, + "loss": 0.5609, + "step": 12708 + }, + { + "epoch": 4.462429775280899, + "grad_norm": 0.9744575619697571, + "learning_rate": 7.759079887704618e-07, + "loss": 0.5435, + "step": 12709 + }, + { + "epoch": 4.462780898876405, + "grad_norm": 0.9520298838615417, + "learning_rate": 7.755732229040025e-07, + "loss": 0.5382, + "step": 12710 + }, + { + "epoch": 4.46313202247191, + "grad_norm": 0.943539023399353, + "learning_rate": 7.752385160114642e-07, + "loss": 0.5336, + "step": 12711 + }, + { + "epoch": 4.463483146067416, + "grad_norm": 0.9633198976516724, + "learning_rate": 7.749038681042917e-07, + "loss": 0.5402, + "step": 12712 + }, + { + "epoch": 4.463834269662922, + "grad_norm": 0.9155687689781189, + "learning_rate": 7.745692791939316e-07, + "loss": 0.5389, + "step": 12713 + }, + { + "epoch": 4.464185393258427, + "grad_norm": 0.9766901135444641, + "learning_rate": 7.742347492918251e-07, + "loss": 0.5683, + "step": 12714 + }, + { + "epoch": 4.464536516853933, + "grad_norm": 0.9585535526275635, + "learning_rate": 7.739002784094141e-07, + "loss": 0.5377, + "step": 12715 + }, + { + "epoch": 4.464887640449438, + "grad_norm": 0.9933169484138489, + "learning_rate": 7.735658665581363e-07, + "loss": 0.5258, + "step": 12716 + }, + { + "epoch": 4.465238764044944, + "grad_norm": 0.9576955437660217, + "learning_rate": 7.732315137494279e-07, + "loss": 0.5633, + "step": 12717 + }, + { + "epoch": 4.46558988764045, + "grad_norm": 0.9723733067512512, + "learning_rate": 7.728972199947249e-07, + "loss": 0.551, + "step": 12718 + }, + { + "epoch": 4.465941011235955, + "grad_norm": 0.971680223941803, + "learning_rate": 7.725629853054584e-07, + "loss": 0.5629, + "step": 12719 + }, + { + "epoch": 4.466292134831461, + "grad_norm": 0.9890063405036926, + "learning_rate": 7.722288096930602e-07, + "loss": 0.5478, + "step": 12720 + }, + { + "epoch": 4.466643258426966, + "grad_norm": 0.9588292837142944, + "learning_rate": 7.718946931689578e-07, + "loss": 0.5509, + "step": 12721 + }, + { + "epoch": 4.466994382022472, + "grad_norm": 0.9532110691070557, + "learning_rate": 7.715606357445785e-07, + "loss": 0.5573, + "step": 12722 + }, + { + "epoch": 4.467345505617978, + "grad_norm": 0.9772461652755737, + "learning_rate": 7.712266374313457e-07, + "loss": 0.5426, + "step": 12723 + }, + { + "epoch": 4.467696629213483, + "grad_norm": 0.978580117225647, + "learning_rate": 7.708926982406837e-07, + "loss": 0.5536, + "step": 12724 + }, + { + "epoch": 4.468047752808989, + "grad_norm": 0.9932138323783875, + "learning_rate": 7.705588181840107e-07, + "loss": 0.5865, + "step": 12725 + }, + { + "epoch": 4.468398876404494, + "grad_norm": 0.9834049940109253, + "learning_rate": 7.702249972727471e-07, + "loss": 0.5459, + "step": 12726 + }, + { + "epoch": 4.46875, + "grad_norm": 0.9727849364280701, + "learning_rate": 7.698912355183075e-07, + "loss": 0.5096, + "step": 12727 + }, + { + "epoch": 4.469101123595506, + "grad_norm": 0.9695791006088257, + "learning_rate": 7.695575329321081e-07, + "loss": 0.5411, + "step": 12728 + }, + { + "epoch": 4.469452247191011, + "grad_norm": 0.9950489401817322, + "learning_rate": 7.692238895255603e-07, + "loss": 0.5766, + "step": 12729 + }, + { + "epoch": 4.469803370786517, + "grad_norm": 0.9656262993812561, + "learning_rate": 7.688903053100735e-07, + "loss": 0.5118, + "step": 12730 + }, + { + "epoch": 4.470154494382022, + "grad_norm": 0.9428182244300842, + "learning_rate": 7.685567802970578e-07, + "loss": 0.507, + "step": 12731 + }, + { + "epoch": 4.470505617977528, + "grad_norm": 0.968177318572998, + "learning_rate": 7.682233144979178e-07, + "loss": 0.5143, + "step": 12732 + }, + { + "epoch": 4.470856741573034, + "grad_norm": 0.9822523593902588, + "learning_rate": 7.678899079240593e-07, + "loss": 0.5418, + "step": 12733 + }, + { + "epoch": 4.471207865168539, + "grad_norm": 0.964699923992157, + "learning_rate": 7.675565605868831e-07, + "loss": 0.5829, + "step": 12734 + }, + { + "epoch": 4.471558988764045, + "grad_norm": 0.9715401530265808, + "learning_rate": 7.672232724977907e-07, + "loss": 0.5478, + "step": 12735 + }, + { + "epoch": 4.47191011235955, + "grad_norm": 0.9691911339759827, + "learning_rate": 7.66890043668179e-07, + "loss": 0.5298, + "step": 12736 + }, + { + "epoch": 4.472261235955056, + "grad_norm": 0.9517095685005188, + "learning_rate": 7.665568741094454e-07, + "loss": 0.5294, + "step": 12737 + }, + { + "epoch": 4.472612359550562, + "grad_norm": 0.9577517509460449, + "learning_rate": 7.662237638329826e-07, + "loss": 0.5085, + "step": 12738 + }, + { + "epoch": 4.472963483146067, + "grad_norm": 0.9734330773353577, + "learning_rate": 7.658907128501844e-07, + "loss": 0.5479, + "step": 12739 + }, + { + "epoch": 4.473314606741573, + "grad_norm": 0.9580860137939453, + "learning_rate": 7.655577211724397e-07, + "loss": 0.524, + "step": 12740 + }, + { + "epoch": 4.473665730337078, + "grad_norm": 0.9184178113937378, + "learning_rate": 7.652247888111361e-07, + "loss": 0.4915, + "step": 12741 + }, + { + "epoch": 4.474016853932584, + "grad_norm": 1.0901345014572144, + "learning_rate": 7.64891915777661e-07, + "loss": 0.562, + "step": 12742 + }, + { + "epoch": 4.47436797752809, + "grad_norm": 0.9529563188552856, + "learning_rate": 7.64559102083397e-07, + "loss": 0.5221, + "step": 12743 + }, + { + "epoch": 4.474719101123595, + "grad_norm": 0.9897323250770569, + "learning_rate": 7.642263477397271e-07, + "loss": 0.5732, + "step": 12744 + }, + { + "epoch": 4.475070224719101, + "grad_norm": 0.9532467722892761, + "learning_rate": 7.638936527580304e-07, + "loss": 0.55, + "step": 12745 + }, + { + "epoch": 4.475421348314606, + "grad_norm": 0.9685435891151428, + "learning_rate": 7.635610171496857e-07, + "loss": 0.5437, + "step": 12746 + }, + { + "epoch": 4.475772471910112, + "grad_norm": 0.9630606770515442, + "learning_rate": 7.632284409260684e-07, + "loss": 0.5493, + "step": 12747 + }, + { + "epoch": 4.476123595505618, + "grad_norm": 0.9693646430969238, + "learning_rate": 7.628959240985514e-07, + "loss": 0.5576, + "step": 12748 + }, + { + "epoch": 4.476474719101123, + "grad_norm": 0.9346907138824463, + "learning_rate": 7.625634666785084e-07, + "loss": 0.5619, + "step": 12749 + }, + { + "epoch": 4.476825842696629, + "grad_norm": 0.9611661434173584, + "learning_rate": 7.622310686773068e-07, + "loss": 0.5441, + "step": 12750 + }, + { + "epoch": 4.477176966292135, + "grad_norm": 0.99566251039505, + "learning_rate": 7.618987301063168e-07, + "loss": 0.5436, + "step": 12751 + }, + { + "epoch": 4.477528089887641, + "grad_norm": 0.9718355536460876, + "learning_rate": 7.615664509769019e-07, + "loss": 0.536, + "step": 12752 + }, + { + "epoch": 4.477879213483146, + "grad_norm": 0.938369631767273, + "learning_rate": 7.612342313004279e-07, + "loss": 0.5497, + "step": 12753 + }, + { + "epoch": 4.478230337078652, + "grad_norm": 0.9847626090049744, + "learning_rate": 7.609020710882541e-07, + "loss": 0.5489, + "step": 12754 + }, + { + "epoch": 4.478581460674158, + "grad_norm": 0.9675350785255432, + "learning_rate": 7.605699703517425e-07, + "loss": 0.5407, + "step": 12755 + }, + { + "epoch": 4.478932584269663, + "grad_norm": 0.9715621471405029, + "learning_rate": 7.60237929102249e-07, + "loss": 0.5616, + "step": 12756 + }, + { + "epoch": 4.479283707865169, + "grad_norm": 0.984789252281189, + "learning_rate": 7.599059473511291e-07, + "loss": 0.5767, + "step": 12757 + }, + { + "epoch": 4.479634831460674, + "grad_norm": 0.9862366318702698, + "learning_rate": 7.595740251097378e-07, + "loss": 0.5276, + "step": 12758 + }, + { + "epoch": 4.47998595505618, + "grad_norm": 0.95209801197052, + "learning_rate": 7.592421623894242e-07, + "loss": 0.5224, + "step": 12759 + }, + { + "epoch": 4.480337078651686, + "grad_norm": 0.9772977232933044, + "learning_rate": 7.589103592015401e-07, + "loss": 0.5433, + "step": 12760 + }, + { + "epoch": 4.480688202247191, + "grad_norm": 0.958781361579895, + "learning_rate": 7.585786155574313e-07, + "loss": 0.5487, + "step": 12761 + }, + { + "epoch": 4.481039325842697, + "grad_norm": 0.9563223719596863, + "learning_rate": 7.58246931468444e-07, + "loss": 0.5202, + "step": 12762 + }, + { + "epoch": 4.481390449438202, + "grad_norm": 0.961854875087738, + "learning_rate": 7.579153069459216e-07, + "loss": 0.5345, + "step": 12763 + }, + { + "epoch": 4.481741573033708, + "grad_norm": 0.9485632181167603, + "learning_rate": 7.575837420012041e-07, + "loss": 0.5253, + "step": 12764 + }, + { + "epoch": 4.482092696629214, + "grad_norm": 0.9798651933670044, + "learning_rate": 7.572522366456322e-07, + "loss": 0.5318, + "step": 12765 + }, + { + "epoch": 4.482443820224719, + "grad_norm": 1.0182582139968872, + "learning_rate": 7.569207908905421e-07, + "loss": 0.5148, + "step": 12766 + }, + { + "epoch": 4.482794943820225, + "grad_norm": 0.9517802596092224, + "learning_rate": 7.565894047472699e-07, + "loss": 0.5266, + "step": 12767 + }, + { + "epoch": 4.48314606741573, + "grad_norm": 0.9579293131828308, + "learning_rate": 7.562580782271473e-07, + "loss": 0.5411, + "step": 12768 + }, + { + "epoch": 4.483497191011236, + "grad_norm": 1.0065041780471802, + "learning_rate": 7.559268113415072e-07, + "loss": 0.5161, + "step": 12769 + }, + { + "epoch": 4.483848314606742, + "grad_norm": 0.9756641387939453, + "learning_rate": 7.555956041016779e-07, + "loss": 0.5652, + "step": 12770 + }, + { + "epoch": 4.484199438202247, + "grad_norm": 0.9906870722770691, + "learning_rate": 7.552644565189854e-07, + "loss": 0.5481, + "step": 12771 + }, + { + "epoch": 4.484550561797753, + "grad_norm": 0.9798886775970459, + "learning_rate": 7.549333686047561e-07, + "loss": 0.5231, + "step": 12772 + }, + { + "epoch": 4.484901685393258, + "grad_norm": 0.9306637048721313, + "learning_rate": 7.546023403703115e-07, + "loss": 0.5067, + "step": 12773 + }, + { + "epoch": 4.485252808988764, + "grad_norm": 0.9746548533439636, + "learning_rate": 7.542713718269742e-07, + "loss": 0.5479, + "step": 12774 + }, + { + "epoch": 4.48560393258427, + "grad_norm": 0.9480656385421753, + "learning_rate": 7.539404629860614e-07, + "loss": 0.519, + "step": 12775 + }, + { + "epoch": 4.485955056179775, + "grad_norm": 0.9456484913825989, + "learning_rate": 7.536096138588914e-07, + "loss": 0.5287, + "step": 12776 + }, + { + "epoch": 4.486306179775281, + "grad_norm": 0.9905698895454407, + "learning_rate": 7.532788244567773e-07, + "loss": 0.5547, + "step": 12777 + }, + { + "epoch": 4.486657303370786, + "grad_norm": 0.9721636176109314, + "learning_rate": 7.529480947910337e-07, + "loss": 0.5486, + "step": 12778 + }, + { + "epoch": 4.487008426966292, + "grad_norm": 0.974402666091919, + "learning_rate": 7.5261742487297e-07, + "loss": 0.5299, + "step": 12779 + }, + { + "epoch": 4.487359550561798, + "grad_norm": 1.0090997219085693, + "learning_rate": 7.522868147138948e-07, + "loss": 0.5409, + "step": 12780 + }, + { + "epoch": 4.487710674157303, + "grad_norm": 0.9708750247955322, + "learning_rate": 7.519562643251155e-07, + "loss": 0.5503, + "step": 12781 + }, + { + "epoch": 4.488061797752809, + "grad_norm": 0.9554027915000916, + "learning_rate": 7.516257737179353e-07, + "loss": 0.5518, + "step": 12782 + }, + { + "epoch": 4.488412921348314, + "grad_norm": 0.9723479747772217, + "learning_rate": 7.512953429036582e-07, + "loss": 0.5682, + "step": 12783 + }, + { + "epoch": 4.48876404494382, + "grad_norm": 0.9361486434936523, + "learning_rate": 7.509649718935835e-07, + "loss": 0.5403, + "step": 12784 + }, + { + "epoch": 4.489115168539326, + "grad_norm": 0.9865178465843201, + "learning_rate": 7.506346606990106e-07, + "loss": 0.569, + "step": 12785 + }, + { + "epoch": 4.489466292134831, + "grad_norm": 0.9396985173225403, + "learning_rate": 7.503044093312348e-07, + "loss": 0.5091, + "step": 12786 + }, + { + "epoch": 4.489817415730337, + "grad_norm": 0.9592762589454651, + "learning_rate": 7.499742178015518e-07, + "loss": 0.5318, + "step": 12787 + }, + { + "epoch": 4.490168539325842, + "grad_norm": 1.096723198890686, + "learning_rate": 7.49644086121252e-07, + "loss": 0.5242, + "step": 12788 + }, + { + "epoch": 4.490519662921348, + "grad_norm": 0.9845041036605835, + "learning_rate": 7.493140143016276e-07, + "loss": 0.5004, + "step": 12789 + }, + { + "epoch": 4.490870786516854, + "grad_norm": 0.9781625866889954, + "learning_rate": 7.489840023539649e-07, + "loss": 0.5364, + "step": 12790 + }, + { + "epoch": 4.491221910112359, + "grad_norm": 0.9686986804008484, + "learning_rate": 7.486540502895518e-07, + "loss": 0.5373, + "step": 12791 + }, + { + "epoch": 4.491573033707865, + "grad_norm": 0.990816056728363, + "learning_rate": 7.483241581196715e-07, + "loss": 0.5483, + "step": 12792 + }, + { + "epoch": 4.491924157303371, + "grad_norm": 0.9606095552444458, + "learning_rate": 7.479943258556052e-07, + "loss": 0.4919, + "step": 12793 + }, + { + "epoch": 4.492275280898877, + "grad_norm": 0.9749131202697754, + "learning_rate": 7.476645535086349e-07, + "loss": 0.524, + "step": 12794 + }, + { + "epoch": 4.492626404494382, + "grad_norm": 0.9368486404418945, + "learning_rate": 7.473348410900363e-07, + "loss": 0.5237, + "step": 12795 + }, + { + "epoch": 4.492977528089888, + "grad_norm": 0.9808181524276733, + "learning_rate": 7.470051886110871e-07, + "loss": 0.5278, + "step": 12796 + }, + { + "epoch": 4.493328651685394, + "grad_norm": 0.9603955149650574, + "learning_rate": 7.466755960830596e-07, + "loss": 0.5292, + "step": 12797 + }, + { + "epoch": 4.493679775280899, + "grad_norm": 0.9527784585952759, + "learning_rate": 7.463460635172273e-07, + "loss": 0.5363, + "step": 12798 + }, + { + "epoch": 4.494030898876405, + "grad_norm": 0.9612696766853333, + "learning_rate": 7.460165909248582e-07, + "loss": 0.5617, + "step": 12799 + }, + { + "epoch": 4.49438202247191, + "grad_norm": 0.9543518424034119, + "learning_rate": 7.456871783172214e-07, + "loss": 0.5271, + "step": 12800 + }, + { + "epoch": 4.494733146067416, + "grad_norm": 0.9577986001968384, + "learning_rate": 7.453578257055816e-07, + "loss": 0.5274, + "step": 12801 + }, + { + "epoch": 4.495084269662922, + "grad_norm": 0.9629268050193787, + "learning_rate": 7.450285331012031e-07, + "loss": 0.5314, + "step": 12802 + }, + { + "epoch": 4.495435393258427, + "grad_norm": 0.966696560382843, + "learning_rate": 7.446993005153463e-07, + "loss": 0.542, + "step": 12803 + }, + { + "epoch": 4.495786516853933, + "grad_norm": 0.9762194156646729, + "learning_rate": 7.443701279592722e-07, + "loss": 0.57, + "step": 12804 + }, + { + "epoch": 4.496137640449438, + "grad_norm": 0.9565728306770325, + "learning_rate": 7.440410154442373e-07, + "loss": 0.5173, + "step": 12805 + }, + { + "epoch": 4.496488764044944, + "grad_norm": 0.9744442701339722, + "learning_rate": 7.437119629814962e-07, + "loss": 0.5417, + "step": 12806 + }, + { + "epoch": 4.49683988764045, + "grad_norm": 0.9667987823486328, + "learning_rate": 7.433829705823042e-07, + "loss": 0.5214, + "step": 12807 + }, + { + "epoch": 4.497191011235955, + "grad_norm": 0.9586924314498901, + "learning_rate": 7.430540382579105e-07, + "loss": 0.5353, + "step": 12808 + }, + { + "epoch": 4.497542134831461, + "grad_norm": 0.9782401323318481, + "learning_rate": 7.427251660195658e-07, + "loss": 0.5534, + "step": 12809 + }, + { + "epoch": 4.497893258426966, + "grad_norm": 0.9581220746040344, + "learning_rate": 7.423963538785162e-07, + "loss": 0.5339, + "step": 12810 + }, + { + "epoch": 4.498244382022472, + "grad_norm": 0.9675527215003967, + "learning_rate": 7.420676018460082e-07, + "loss": 0.5387, + "step": 12811 + }, + { + "epoch": 4.498595505617978, + "grad_norm": 0.9581845998764038, + "learning_rate": 7.417389099332827e-07, + "loss": 0.5323, + "step": 12812 + }, + { + "epoch": 4.498946629213483, + "grad_norm": 0.9545124173164368, + "learning_rate": 7.41410278151583e-07, + "loss": 0.5282, + "step": 12813 + }, + { + "epoch": 4.499297752808989, + "grad_norm": 0.9484679102897644, + "learning_rate": 7.410817065121462e-07, + "loss": 0.5643, + "step": 12814 + }, + { + "epoch": 4.499648876404494, + "grad_norm": 0.9697062969207764, + "learning_rate": 7.407531950262106e-07, + "loss": 0.5446, + "step": 12815 + }, + { + "epoch": 4.5, + "grad_norm": 0.9832372069358826, + "learning_rate": 7.404247437050105e-07, + "loss": 0.5366, + "step": 12816 + }, + { + "epoch": 4.500351123595506, + "grad_norm": 0.9499861001968384, + "learning_rate": 7.400963525597773e-07, + "loss": 0.5524, + "step": 12817 + }, + { + "epoch": 4.500702247191011, + "grad_norm": 0.963616669178009, + "learning_rate": 7.397680216017442e-07, + "loss": 0.5413, + "step": 12818 + }, + { + "epoch": 4.501053370786517, + "grad_norm": 0.9776845574378967, + "learning_rate": 7.394397508421373e-07, + "loss": 0.5266, + "step": 12819 + }, + { + "epoch": 4.501404494382022, + "grad_norm": 0.958728015422821, + "learning_rate": 7.391115402921853e-07, + "loss": 0.5644, + "step": 12820 + }, + { + "epoch": 4.501755617977528, + "grad_norm": 0.9251421689987183, + "learning_rate": 7.387833899631108e-07, + "loss": 0.5261, + "step": 12821 + }, + { + "epoch": 4.502106741573034, + "grad_norm": 0.9811964631080627, + "learning_rate": 7.384552998661385e-07, + "loss": 0.5428, + "step": 12822 + }, + { + "epoch": 4.502457865168539, + "grad_norm": 0.9331524968147278, + "learning_rate": 7.381272700124864e-07, + "loss": 0.5109, + "step": 12823 + }, + { + "epoch": 4.502808988764045, + "grad_norm": 0.9874469041824341, + "learning_rate": 7.377993004133749e-07, + "loss": 0.547, + "step": 12824 + }, + { + "epoch": 4.50316011235955, + "grad_norm": 0.9605987668037415, + "learning_rate": 7.374713910800188e-07, + "loss": 0.5529, + "step": 12825 + }, + { + "epoch": 4.503511235955056, + "grad_norm": 0.9361422061920166, + "learning_rate": 7.371435420236334e-07, + "loss": 0.5168, + "step": 12826 + }, + { + "epoch": 4.503862359550562, + "grad_norm": 1.01318359375, + "learning_rate": 7.368157532554298e-07, + "loss": 0.5318, + "step": 12827 + }, + { + "epoch": 4.504213483146067, + "grad_norm": 0.9789500832557678, + "learning_rate": 7.364880247866193e-07, + "loss": 0.5162, + "step": 12828 + }, + { + "epoch": 4.504564606741573, + "grad_norm": 0.9657090306282043, + "learning_rate": 7.361603566284095e-07, + "loss": 0.5332, + "step": 12829 + }, + { + "epoch": 4.504915730337078, + "grad_norm": 1.004984974861145, + "learning_rate": 7.358327487920052e-07, + "loss": 0.5545, + "step": 12830 + }, + { + "epoch": 4.505266853932584, + "grad_norm": 0.9456415772438049, + "learning_rate": 7.355052012886121e-07, + "loss": 0.5397, + "step": 12831 + }, + { + "epoch": 4.50561797752809, + "grad_norm": 0.984649658203125, + "learning_rate": 7.351777141294302e-07, + "loss": 0.5574, + "step": 12832 + }, + { + "epoch": 4.505969101123595, + "grad_norm": 0.9945653676986694, + "learning_rate": 7.348502873256616e-07, + "loss": 0.5724, + "step": 12833 + }, + { + "epoch": 4.506320224719101, + "grad_norm": 0.9753018617630005, + "learning_rate": 7.345229208885016e-07, + "loss": 0.5399, + "step": 12834 + }, + { + "epoch": 4.506671348314606, + "grad_norm": 0.9505728483200073, + "learning_rate": 7.341956148291476e-07, + "loss": 0.542, + "step": 12835 + }, + { + "epoch": 4.507022471910112, + "grad_norm": 0.9761530756950378, + "learning_rate": 7.33868369158792e-07, + "loss": 0.5576, + "step": 12836 + }, + { + "epoch": 4.507373595505618, + "grad_norm": 0.9845876693725586, + "learning_rate": 7.335411838886278e-07, + "loss": 0.5503, + "step": 12837 + }, + { + "epoch": 4.507724719101123, + "grad_norm": 0.9928021430969238, + "learning_rate": 7.332140590298426e-07, + "loss": 0.5417, + "step": 12838 + }, + { + "epoch": 4.508075842696629, + "grad_norm": 0.9697917699813843, + "learning_rate": 7.328869945936254e-07, + "loss": 0.5205, + "step": 12839 + }, + { + "epoch": 4.508426966292134, + "grad_norm": 0.9790090918540955, + "learning_rate": 7.325599905911601e-07, + "loss": 0.5257, + "step": 12840 + }, + { + "epoch": 4.50877808988764, + "grad_norm": 0.9639084935188293, + "learning_rate": 7.322330470336314e-07, + "loss": 0.5116, + "step": 12841 + }, + { + "epoch": 4.509129213483146, + "grad_norm": 0.9632887840270996, + "learning_rate": 7.319061639322201e-07, + "loss": 0.5151, + "step": 12842 + }, + { + "epoch": 4.509480337078652, + "grad_norm": 1.0030144453048706, + "learning_rate": 7.315793412981038e-07, + "loss": 0.5678, + "step": 12843 + }, + { + "epoch": 4.509831460674158, + "grad_norm": 0.9732213616371155, + "learning_rate": 7.312525791424619e-07, + "loss": 0.5127, + "step": 12844 + }, + { + "epoch": 4.510182584269663, + "grad_norm": 1.015194058418274, + "learning_rate": 7.309258774764672e-07, + "loss": 0.5558, + "step": 12845 + }, + { + "epoch": 4.510533707865169, + "grad_norm": 0.9728255867958069, + "learning_rate": 7.305992363112946e-07, + "loss": 0.5586, + "step": 12846 + }, + { + "epoch": 4.510884831460674, + "grad_norm": 1.0415880680084229, + "learning_rate": 7.302726556581133e-07, + "loss": 0.5134, + "step": 12847 + }, + { + "epoch": 4.51123595505618, + "grad_norm": 0.9654303789138794, + "learning_rate": 7.299461355280934e-07, + "loss": 0.5659, + "step": 12848 + }, + { + "epoch": 4.511587078651686, + "grad_norm": 0.9491206407546997, + "learning_rate": 7.296196759324006e-07, + "loss": 0.5229, + "step": 12849 + }, + { + "epoch": 4.511938202247191, + "grad_norm": 0.9433630704879761, + "learning_rate": 7.292932768822003e-07, + "loss": 0.5584, + "step": 12850 + }, + { + "epoch": 4.512289325842697, + "grad_norm": 0.9856569766998291, + "learning_rate": 7.289669383886541e-07, + "loss": 0.564, + "step": 12851 + }, + { + "epoch": 4.512640449438202, + "grad_norm": 0.9495972394943237, + "learning_rate": 7.286406604629239e-07, + "loss": 0.537, + "step": 12852 + }, + { + "epoch": 4.512991573033708, + "grad_norm": 0.9750402569770813, + "learning_rate": 7.283144431161668e-07, + "loss": 0.5754, + "step": 12853 + }, + { + "epoch": 4.513342696629214, + "grad_norm": 0.9973424673080444, + "learning_rate": 7.2798828635954e-07, + "loss": 0.561, + "step": 12854 + }, + { + "epoch": 4.513693820224719, + "grad_norm": 0.9625391960144043, + "learning_rate": 7.276621902041977e-07, + "loss": 0.5308, + "step": 12855 + }, + { + "epoch": 4.514044943820225, + "grad_norm": 0.9718276858329773, + "learning_rate": 7.273361546612912e-07, + "loss": 0.5664, + "step": 12856 + }, + { + "epoch": 4.51439606741573, + "grad_norm": 0.9541653394699097, + "learning_rate": 7.270101797419721e-07, + "loss": 0.5628, + "step": 12857 + }, + { + "epoch": 4.514747191011236, + "grad_norm": 0.9576981067657471, + "learning_rate": 7.266842654573869e-07, + "loss": 0.5226, + "step": 12858 + }, + { + "epoch": 4.515098314606742, + "grad_norm": 0.9699994325637817, + "learning_rate": 7.263584118186828e-07, + "loss": 0.5593, + "step": 12859 + }, + { + "epoch": 4.515449438202247, + "grad_norm": 0.9610205888748169, + "learning_rate": 7.260326188370029e-07, + "loss": 0.5381, + "step": 12860 + }, + { + "epoch": 4.515800561797753, + "grad_norm": 0.9681107401847839, + "learning_rate": 7.257068865234901e-07, + "loss": 0.5143, + "step": 12861 + }, + { + "epoch": 4.516151685393258, + "grad_norm": 0.946168839931488, + "learning_rate": 7.253812148892827e-07, + "loss": 0.5302, + "step": 12862 + }, + { + "epoch": 4.516502808988764, + "grad_norm": 0.9799460768699646, + "learning_rate": 7.250556039455201e-07, + "loss": 0.5493, + "step": 12863 + }, + { + "epoch": 4.51685393258427, + "grad_norm": 0.9965935349464417, + "learning_rate": 7.247300537033361e-07, + "loss": 0.5572, + "step": 12864 + }, + { + "epoch": 4.517205056179775, + "grad_norm": 0.973211407661438, + "learning_rate": 7.244045641738659e-07, + "loss": 0.5413, + "step": 12865 + }, + { + "epoch": 4.517556179775281, + "grad_norm": 0.969127893447876, + "learning_rate": 7.240791353682394e-07, + "loss": 0.5324, + "step": 12866 + }, + { + "epoch": 4.517907303370786, + "grad_norm": 0.9631320834159851, + "learning_rate": 7.237537672975875e-07, + "loss": 0.5271, + "step": 12867 + }, + { + "epoch": 4.518258426966292, + "grad_norm": 0.9655739068984985, + "learning_rate": 7.234284599730371e-07, + "loss": 0.5217, + "step": 12868 + }, + { + "epoch": 4.518609550561798, + "grad_norm": 0.9825718998908997, + "learning_rate": 7.23103213405712e-07, + "loss": 0.5743, + "step": 12869 + }, + { + "epoch": 4.518960674157303, + "grad_norm": 0.9946292042732239, + "learning_rate": 7.227780276067375e-07, + "loss": 0.5634, + "step": 12870 + }, + { + "epoch": 4.519311797752809, + "grad_norm": 0.9635724425315857, + "learning_rate": 7.224529025872329e-07, + "loss": 0.5452, + "step": 12871 + }, + { + "epoch": 4.519662921348314, + "grad_norm": 0.962285041809082, + "learning_rate": 7.221278383583186e-07, + "loss": 0.5228, + "step": 12872 + }, + { + "epoch": 4.52001404494382, + "grad_norm": 0.9775204658508301, + "learning_rate": 7.218028349311101e-07, + "loss": 0.5651, + "step": 12873 + }, + { + "epoch": 4.520365168539326, + "grad_norm": 0.949164092540741, + "learning_rate": 7.21477892316724e-07, + "loss": 0.5102, + "step": 12874 + }, + { + "epoch": 4.520716292134831, + "grad_norm": 0.9366573095321655, + "learning_rate": 7.211530105262712e-07, + "loss": 0.5674, + "step": 12875 + }, + { + "epoch": 4.521067415730337, + "grad_norm": 0.9784334897994995, + "learning_rate": 7.208281895708641e-07, + "loss": 0.519, + "step": 12876 + }, + { + "epoch": 4.521418539325842, + "grad_norm": 0.9681825637817383, + "learning_rate": 7.205034294616095e-07, + "loss": 0.5361, + "step": 12877 + }, + { + "epoch": 4.521769662921348, + "grad_norm": 0.9749162793159485, + "learning_rate": 7.201787302096158e-07, + "loss": 0.5493, + "step": 12878 + }, + { + "epoch": 4.522120786516854, + "grad_norm": 0.9746007323265076, + "learning_rate": 7.198540918259864e-07, + "loss": 0.5333, + "step": 12879 + }, + { + "epoch": 4.52247191011236, + "grad_norm": 0.989389181137085, + "learning_rate": 7.195295143218231e-07, + "loss": 0.546, + "step": 12880 + }, + { + "epoch": 4.522823033707866, + "grad_norm": 0.9763299822807312, + "learning_rate": 7.192049977082274e-07, + "loss": 0.5174, + "step": 12881 + }, + { + "epoch": 4.523174157303371, + "grad_norm": 0.9521219730377197, + "learning_rate": 7.188805419962966e-07, + "loss": 0.5516, + "step": 12882 + }, + { + "epoch": 4.523525280898877, + "grad_norm": 0.9850859642028809, + "learning_rate": 7.185561471971275e-07, + "loss": 0.5477, + "step": 12883 + }, + { + "epoch": 4.523876404494382, + "grad_norm": 0.9792001247406006, + "learning_rate": 7.18231813321813e-07, + "loss": 0.5428, + "step": 12884 + }, + { + "epoch": 4.524227528089888, + "grad_norm": 0.9560434818267822, + "learning_rate": 7.179075403814468e-07, + "loss": 0.5607, + "step": 12885 + }, + { + "epoch": 4.524578651685394, + "grad_norm": 0.9857417345046997, + "learning_rate": 7.175833283871169e-07, + "loss": 0.5322, + "step": 12886 + }, + { + "epoch": 4.524929775280899, + "grad_norm": 0.9827060103416443, + "learning_rate": 7.172591773499127e-07, + "loss": 0.5246, + "step": 12887 + }, + { + "epoch": 4.525280898876405, + "grad_norm": 0.9806876182556152, + "learning_rate": 7.169350872809183e-07, + "loss": 0.5303, + "step": 12888 + }, + { + "epoch": 4.52563202247191, + "grad_norm": 0.9727808237075806, + "learning_rate": 7.166110581912189e-07, + "loss": 0.5241, + "step": 12889 + }, + { + "epoch": 4.525983146067416, + "grad_norm": 0.9523377418518066, + "learning_rate": 7.162870900918947e-07, + "loss": 0.5136, + "step": 12890 + }, + { + "epoch": 4.526334269662922, + "grad_norm": 0.9653171300888062, + "learning_rate": 7.159631829940262e-07, + "loss": 0.5145, + "step": 12891 + }, + { + "epoch": 4.526685393258427, + "grad_norm": 0.9561030268669128, + "learning_rate": 7.156393369086903e-07, + "loss": 0.5266, + "step": 12892 + }, + { + "epoch": 4.527036516853933, + "grad_norm": 0.947825014591217, + "learning_rate": 7.153155518469612e-07, + "loss": 0.5453, + "step": 12893 + }, + { + "epoch": 4.527387640449438, + "grad_norm": 0.9512736201286316, + "learning_rate": 7.149918278199142e-07, + "loss": 0.4982, + "step": 12894 + }, + { + "epoch": 4.527738764044944, + "grad_norm": 0.979557454586029, + "learning_rate": 7.14668164838618e-07, + "loss": 0.5376, + "step": 12895 + }, + { + "epoch": 4.52808988764045, + "grad_norm": 0.9954123497009277, + "learning_rate": 7.143445629141441e-07, + "loss": 0.539, + "step": 12896 + }, + { + "epoch": 4.528441011235955, + "grad_norm": 0.9939441680908203, + "learning_rate": 7.140210220575571e-07, + "loss": 0.561, + "step": 12897 + }, + { + "epoch": 4.528792134831461, + "grad_norm": 1.0182547569274902, + "learning_rate": 7.136975422799236e-07, + "loss": 0.5305, + "step": 12898 + }, + { + "epoch": 4.529143258426966, + "grad_norm": 0.9829132556915283, + "learning_rate": 7.13374123592305e-07, + "loss": 0.5481, + "step": 12899 + }, + { + "epoch": 4.529494382022472, + "grad_norm": 0.9661722183227539, + "learning_rate": 7.130507660057631e-07, + "loss": 0.5399, + "step": 12900 + }, + { + "epoch": 4.529845505617978, + "grad_norm": 0.9631792306900024, + "learning_rate": 7.127274695313552e-07, + "loss": 0.5449, + "step": 12901 + }, + { + "epoch": 4.530196629213483, + "grad_norm": 0.9433115124702454, + "learning_rate": 7.124042341801393e-07, + "loss": 0.5351, + "step": 12902 + }, + { + "epoch": 4.530547752808989, + "grad_norm": 0.9818670749664307, + "learning_rate": 7.120810599631681e-07, + "loss": 0.5248, + "step": 12903 + }, + { + "epoch": 4.530898876404494, + "grad_norm": 0.9435574412345886, + "learning_rate": 7.117579468914953e-07, + "loss": 0.543, + "step": 12904 + }, + { + "epoch": 4.53125, + "grad_norm": 0.9591655135154724, + "learning_rate": 7.11434894976171e-07, + "loss": 0.5083, + "step": 12905 + }, + { + "epoch": 4.531601123595506, + "grad_norm": 0.9437412619590759, + "learning_rate": 7.111119042282416e-07, + "loss": 0.5215, + "step": 12906 + }, + { + "epoch": 4.531952247191011, + "grad_norm": 0.9453641772270203, + "learning_rate": 7.107889746587552e-07, + "loss": 0.4825, + "step": 12907 + }, + { + "epoch": 4.532303370786517, + "grad_norm": 0.9383487105369568, + "learning_rate": 7.104661062787538e-07, + "loss": 0.5246, + "step": 12908 + }, + { + "epoch": 4.532654494382022, + "grad_norm": 0.9562399983406067, + "learning_rate": 7.101432990992813e-07, + "loss": 0.5364, + "step": 12909 + }, + { + "epoch": 4.533005617977528, + "grad_norm": 0.9556291699409485, + "learning_rate": 7.098205531313756e-07, + "loss": 0.5495, + "step": 12910 + }, + { + "epoch": 4.533356741573034, + "grad_norm": 0.9541529417037964, + "learning_rate": 7.094978683860756e-07, + "loss": 0.556, + "step": 12911 + }, + { + "epoch": 4.533707865168539, + "grad_norm": 0.9288599491119385, + "learning_rate": 7.09175244874416e-07, + "loss": 0.4947, + "step": 12912 + }, + { + "epoch": 4.534058988764045, + "grad_norm": 1.0487364530563354, + "learning_rate": 7.08852682607431e-07, + "loss": 0.5895, + "step": 12913 + }, + { + "epoch": 4.53441011235955, + "grad_norm": 0.9803848266601562, + "learning_rate": 7.085301815961518e-07, + "loss": 0.5443, + "step": 12914 + }, + { + "epoch": 4.534761235955056, + "grad_norm": 0.9971548318862915, + "learning_rate": 7.082077418516065e-07, + "loss": 0.5639, + "step": 12915 + }, + { + "epoch": 4.535112359550562, + "grad_norm": 0.9842275381088257, + "learning_rate": 7.07885363384824e-07, + "loss": 0.5257, + "step": 12916 + }, + { + "epoch": 4.535463483146067, + "grad_norm": 0.9587321281433105, + "learning_rate": 7.075630462068276e-07, + "loss": 0.515, + "step": 12917 + }, + { + "epoch": 4.535814606741573, + "grad_norm": 1.0548192262649536, + "learning_rate": 7.072407903286421e-07, + "loss": 0.5334, + "step": 12918 + }, + { + "epoch": 4.536165730337078, + "grad_norm": 1.0311801433563232, + "learning_rate": 7.069185957612867e-07, + "loss": 0.5493, + "step": 12919 + }, + { + "epoch": 4.536516853932584, + "grad_norm": 0.9618310928344727, + "learning_rate": 7.065964625157818e-07, + "loss": 0.5332, + "step": 12920 + }, + { + "epoch": 4.53686797752809, + "grad_norm": 0.9813491106033325, + "learning_rate": 7.062743906031433e-07, + "loss": 0.5583, + "step": 12921 + }, + { + "epoch": 4.537219101123595, + "grad_norm": 0.9763704538345337, + "learning_rate": 7.059523800343848e-07, + "loss": 0.5174, + "step": 12922 + }, + { + "epoch": 4.537570224719101, + "grad_norm": 0.9723789095878601, + "learning_rate": 7.056304308205205e-07, + "loss": 0.5483, + "step": 12923 + }, + { + "epoch": 4.537921348314606, + "grad_norm": 0.9736176133155823, + "learning_rate": 7.053085429725592e-07, + "loss": 0.5426, + "step": 12924 + }, + { + "epoch": 4.538272471910112, + "grad_norm": 0.9422106146812439, + "learning_rate": 7.049867165015109e-07, + "loss": 0.5607, + "step": 12925 + }, + { + "epoch": 4.538623595505618, + "grad_norm": 0.9798263907432556, + "learning_rate": 7.046649514183801e-07, + "loss": 0.5205, + "step": 12926 + }, + { + "epoch": 4.538974719101123, + "grad_norm": 0.9818385243415833, + "learning_rate": 7.043432477341722e-07, + "loss": 0.5388, + "step": 12927 + }, + { + "epoch": 4.539325842696629, + "grad_norm": 0.9972334504127502, + "learning_rate": 7.040216054598889e-07, + "loss": 0.5569, + "step": 12928 + }, + { + "epoch": 4.539676966292134, + "grad_norm": 0.9745475649833679, + "learning_rate": 7.03700024606529e-07, + "loss": 0.5787, + "step": 12929 + }, + { + "epoch": 4.54002808988764, + "grad_norm": 0.977551281452179, + "learning_rate": 7.033785051850919e-07, + "loss": 0.5308, + "step": 12930 + }, + { + "epoch": 4.540379213483146, + "grad_norm": 0.9628678560256958, + "learning_rate": 7.03057047206572e-07, + "loss": 0.5365, + "step": 12931 + }, + { + "epoch": 4.540730337078652, + "grad_norm": 0.9518944025039673, + "learning_rate": 7.027356506819638e-07, + "loss": 0.5023, + "step": 12932 + }, + { + "epoch": 4.541081460674158, + "grad_norm": 0.9573241472244263, + "learning_rate": 7.024143156222579e-07, + "loss": 0.5125, + "step": 12933 + }, + { + "epoch": 4.541432584269663, + "grad_norm": 0.95763099193573, + "learning_rate": 7.02093042038445e-07, + "loss": 0.5144, + "step": 12934 + }, + { + "epoch": 4.541783707865169, + "grad_norm": 0.9614569544792175, + "learning_rate": 7.017718299415105e-07, + "loss": 0.5387, + "step": 12935 + }, + { + "epoch": 4.542134831460674, + "grad_norm": 0.9468940496444702, + "learning_rate": 7.014506793424417e-07, + "loss": 0.5557, + "step": 12936 + }, + { + "epoch": 4.54248595505618, + "grad_norm": 0.9401723742485046, + "learning_rate": 7.011295902522205e-07, + "loss": 0.4813, + "step": 12937 + }, + { + "epoch": 4.542837078651686, + "grad_norm": 0.9067776203155518, + "learning_rate": 7.008085626818273e-07, + "loss": 0.502, + "step": 12938 + }, + { + "epoch": 4.543188202247191, + "grad_norm": 0.9611430168151855, + "learning_rate": 7.004875966422425e-07, + "loss": 0.5151, + "step": 12939 + }, + { + "epoch": 4.543539325842697, + "grad_norm": 0.9415492415428162, + "learning_rate": 7.001666921444411e-07, + "loss": 0.533, + "step": 12940 + }, + { + "epoch": 4.543890449438202, + "grad_norm": 0.9414841532707214, + "learning_rate": 6.998458491993995e-07, + "loss": 0.5021, + "step": 12941 + }, + { + "epoch": 4.544241573033708, + "grad_norm": 0.9724755883216858, + "learning_rate": 6.995250678180887e-07, + "loss": 0.5601, + "step": 12942 + }, + { + "epoch": 4.544592696629214, + "grad_norm": 0.9692052602767944, + "learning_rate": 6.992043480114807e-07, + "loss": 0.5672, + "step": 12943 + }, + { + "epoch": 4.544943820224719, + "grad_norm": 0.9686769247055054, + "learning_rate": 6.98883689790543e-07, + "loss": 0.5544, + "step": 12944 + }, + { + "epoch": 4.545294943820225, + "grad_norm": 0.9692097902297974, + "learning_rate": 6.98563093166241e-07, + "loss": 0.5141, + "step": 12945 + }, + { + "epoch": 4.54564606741573, + "grad_norm": 0.9449496269226074, + "learning_rate": 6.982425581495405e-07, + "loss": 0.5008, + "step": 12946 + }, + { + "epoch": 4.545997191011236, + "grad_norm": 1.0256092548370361, + "learning_rate": 6.979220847514021e-07, + "loss": 0.5209, + "step": 12947 + }, + { + "epoch": 4.546348314606742, + "grad_norm": 0.9972085952758789, + "learning_rate": 6.97601672982787e-07, + "loss": 0.5627, + "step": 12948 + }, + { + "epoch": 4.546699438202247, + "grad_norm": 0.9840613603591919, + "learning_rate": 6.972813228546516e-07, + "loss": 0.6092, + "step": 12949 + }, + { + "epoch": 4.547050561797753, + "grad_norm": 0.9840967655181885, + "learning_rate": 6.969610343779531e-07, + "loss": 0.5486, + "step": 12950 + }, + { + "epoch": 4.547401685393258, + "grad_norm": 0.9666551351547241, + "learning_rate": 6.966408075636435e-07, + "loss": 0.5761, + "step": 12951 + }, + { + "epoch": 4.547752808988764, + "grad_norm": 0.9881908893585205, + "learning_rate": 6.963206424226762e-07, + "loss": 0.5471, + "step": 12952 + }, + { + "epoch": 4.54810393258427, + "grad_norm": 0.9521369934082031, + "learning_rate": 6.960005389659985e-07, + "loss": 0.5222, + "step": 12953 + }, + { + "epoch": 4.548455056179775, + "grad_norm": 1.0193486213684082, + "learning_rate": 6.956804972045595e-07, + "loss": 0.5713, + "step": 12954 + }, + { + "epoch": 4.548806179775281, + "grad_norm": 0.9673452377319336, + "learning_rate": 6.953605171493033e-07, + "loss": 0.5414, + "step": 12955 + }, + { + "epoch": 4.549157303370786, + "grad_norm": 1.0248960256576538, + "learning_rate": 6.950405988111728e-07, + "loss": 0.5467, + "step": 12956 + }, + { + "epoch": 4.549508426966292, + "grad_norm": 0.9794275164604187, + "learning_rate": 6.947207422011096e-07, + "loss": 0.5119, + "step": 12957 + }, + { + "epoch": 4.549859550561798, + "grad_norm": 0.9706592559814453, + "learning_rate": 6.944009473300517e-07, + "loss": 0.5325, + "step": 12958 + }, + { + "epoch": 4.550210674157303, + "grad_norm": 0.9687896370887756, + "learning_rate": 6.940812142089371e-07, + "loss": 0.5471, + "step": 12959 + }, + { + "epoch": 4.550561797752809, + "grad_norm": 0.9878486394882202, + "learning_rate": 6.937615428486988e-07, + "loss": 0.5389, + "step": 12960 + }, + { + "epoch": 4.550912921348314, + "grad_norm": 0.9682524800300598, + "learning_rate": 6.93441933260271e-07, + "loss": 0.5254, + "step": 12961 + }, + { + "epoch": 4.55126404494382, + "grad_norm": 0.970962405204773, + "learning_rate": 6.931223854545823e-07, + "loss": 0.5089, + "step": 12962 + }, + { + "epoch": 4.551615168539326, + "grad_norm": 0.9644586443901062, + "learning_rate": 6.928028994425625e-07, + "loss": 0.5288, + "step": 12963 + }, + { + "epoch": 4.551966292134831, + "grad_norm": 0.9846317768096924, + "learning_rate": 6.924834752351367e-07, + "loss": 0.5375, + "step": 12964 + }, + { + "epoch": 4.552317415730337, + "grad_norm": 0.978790819644928, + "learning_rate": 6.921641128432299e-07, + "loss": 0.5477, + "step": 12965 + }, + { + "epoch": 4.552668539325842, + "grad_norm": 0.9707977175712585, + "learning_rate": 6.918448122777627e-07, + "loss": 0.5612, + "step": 12966 + }, + { + "epoch": 4.553019662921348, + "grad_norm": 0.9378740191459656, + "learning_rate": 6.915255735496562e-07, + "loss": 0.5034, + "step": 12967 + }, + { + "epoch": 4.553370786516854, + "grad_norm": 0.9675323367118835, + "learning_rate": 6.912063966698279e-07, + "loss": 0.5493, + "step": 12968 + }, + { + "epoch": 4.55372191011236, + "grad_norm": 0.9739416241645813, + "learning_rate": 6.908872816491921e-07, + "loss": 0.5297, + "step": 12969 + }, + { + "epoch": 4.554073033707866, + "grad_norm": 0.9861370325088501, + "learning_rate": 6.905682284986639e-07, + "loss": 0.5432, + "step": 12970 + }, + { + "epoch": 4.554424157303371, + "grad_norm": 0.9556821584701538, + "learning_rate": 6.902492372291533e-07, + "loss": 0.5464, + "step": 12971 + }, + { + "epoch": 4.554775280898877, + "grad_norm": 0.958280086517334, + "learning_rate": 6.899303078515707e-07, + "loss": 0.5274, + "step": 12972 + }, + { + "epoch": 4.555126404494382, + "grad_norm": 0.9821887016296387, + "learning_rate": 6.896114403768222e-07, + "loss": 0.5595, + "step": 12973 + }, + { + "epoch": 4.555477528089888, + "grad_norm": 0.973824679851532, + "learning_rate": 6.892926348158138e-07, + "loss": 0.5446, + "step": 12974 + }, + { + "epoch": 4.555828651685394, + "grad_norm": 0.9591623544692993, + "learning_rate": 6.889738911794471e-07, + "loss": 0.5158, + "step": 12975 + }, + { + "epoch": 4.556179775280899, + "grad_norm": 0.9396053552627563, + "learning_rate": 6.886552094786244e-07, + "loss": 0.5455, + "step": 12976 + }, + { + "epoch": 4.556530898876405, + "grad_norm": 0.9526325464248657, + "learning_rate": 6.88336589724243e-07, + "loss": 0.5476, + "step": 12977 + }, + { + "epoch": 4.55688202247191, + "grad_norm": 1.0063426494598389, + "learning_rate": 6.880180319272006e-07, + "loss": 0.5268, + "step": 12978 + }, + { + "epoch": 4.557233146067416, + "grad_norm": 0.9958229064941406, + "learning_rate": 6.876995360983901e-07, + "loss": 0.538, + "step": 12979 + }, + { + "epoch": 4.557584269662922, + "grad_norm": 0.9853613972663879, + "learning_rate": 6.873811022487053e-07, + "loss": 0.5355, + "step": 12980 + }, + { + "epoch": 4.557935393258427, + "grad_norm": 0.9632471203804016, + "learning_rate": 6.870627303890362e-07, + "loss": 0.5562, + "step": 12981 + }, + { + "epoch": 4.558286516853933, + "grad_norm": 0.9674646854400635, + "learning_rate": 6.867444205302692e-07, + "loss": 0.5496, + "step": 12982 + }, + { + "epoch": 4.558637640449438, + "grad_norm": 1.1178067922592163, + "learning_rate": 6.864261726832921e-07, + "loss": 0.553, + "step": 12983 + }, + { + "epoch": 4.558988764044944, + "grad_norm": 0.9752599000930786, + "learning_rate": 6.861079868589876e-07, + "loss": 0.5573, + "step": 12984 + }, + { + "epoch": 4.55933988764045, + "grad_norm": 1.031005859375, + "learning_rate": 6.857898630682383e-07, + "loss": 0.5458, + "step": 12985 + }, + { + "epoch": 4.559691011235955, + "grad_norm": 0.9799385070800781, + "learning_rate": 6.854718013219225e-07, + "loss": 0.5639, + "step": 12986 + }, + { + "epoch": 4.560042134831461, + "grad_norm": 0.9976696968078613, + "learning_rate": 6.851538016309192e-07, + "loss": 0.5299, + "step": 12987 + }, + { + "epoch": 4.560393258426966, + "grad_norm": 0.9926140308380127, + "learning_rate": 6.84835864006102e-07, + "loss": 0.5318, + "step": 12988 + }, + { + "epoch": 4.560744382022472, + "grad_norm": 0.9464899301528931, + "learning_rate": 6.845179884583461e-07, + "loss": 0.4963, + "step": 12989 + }, + { + "epoch": 4.561095505617978, + "grad_norm": 0.9392485022544861, + "learning_rate": 6.842001749985203e-07, + "loss": 0.5203, + "step": 12990 + }, + { + "epoch": 4.561446629213483, + "grad_norm": 0.9531574845314026, + "learning_rate": 6.838824236374957e-07, + "loss": 0.5307, + "step": 12991 + }, + { + "epoch": 4.561797752808989, + "grad_norm": 0.9608098268508911, + "learning_rate": 6.835647343861373e-07, + "loss": 0.5423, + "step": 12992 + }, + { + "epoch": 4.562148876404494, + "grad_norm": 0.9486250281333923, + "learning_rate": 6.832471072553115e-07, + "loss": 0.5233, + "step": 12993 + }, + { + "epoch": 4.5625, + "grad_norm": 0.9564372301101685, + "learning_rate": 6.829295422558798e-07, + "loss": 0.4971, + "step": 12994 + }, + { + "epoch": 4.562851123595506, + "grad_norm": 0.9400876760482788, + "learning_rate": 6.826120393987023e-07, + "loss": 0.5671, + "step": 12995 + }, + { + "epoch": 4.563202247191011, + "grad_norm": 0.9574089050292969, + "learning_rate": 6.822945986946386e-07, + "loss": 0.5141, + "step": 12996 + }, + { + "epoch": 4.563553370786517, + "grad_norm": 0.9941372275352478, + "learning_rate": 6.819772201545432e-07, + "loss": 0.5738, + "step": 12997 + }, + { + "epoch": 4.563904494382022, + "grad_norm": 0.9718570709228516, + "learning_rate": 6.816599037892724e-07, + "loss": 0.5576, + "step": 12998 + }, + { + "epoch": 4.564255617977528, + "grad_norm": 0.9819703102111816, + "learning_rate": 6.813426496096759e-07, + "loss": 0.5417, + "step": 12999 + }, + { + "epoch": 4.564606741573034, + "grad_norm": 0.9906924962997437, + "learning_rate": 6.81025457626605e-07, + "loss": 0.5687, + "step": 13000 + }, + { + "epoch": 4.564957865168539, + "grad_norm": 0.9585727453231812, + "learning_rate": 6.807083278509067e-07, + "loss": 0.5409, + "step": 13001 + }, + { + "epoch": 4.565308988764045, + "grad_norm": 0.9940949082374573, + "learning_rate": 6.803912602934273e-07, + "loss": 0.559, + "step": 13002 + }, + { + "epoch": 4.56566011235955, + "grad_norm": 0.9398554563522339, + "learning_rate": 6.800742549650088e-07, + "loss": 0.5618, + "step": 13003 + }, + { + "epoch": 4.566011235955056, + "grad_norm": 0.9751331210136414, + "learning_rate": 6.797573118764946e-07, + "loss": 0.5575, + "step": 13004 + }, + { + "epoch": 4.566362359550562, + "grad_norm": 0.9710230231285095, + "learning_rate": 6.794404310387226e-07, + "loss": 0.5275, + "step": 13005 + }, + { + "epoch": 4.566713483146067, + "grad_norm": 0.9685956835746765, + "learning_rate": 6.791236124625292e-07, + "loss": 0.5506, + "step": 13006 + }, + { + "epoch": 4.567064606741573, + "grad_norm": 0.959846556186676, + "learning_rate": 6.788068561587508e-07, + "loss": 0.5463, + "step": 13007 + }, + { + "epoch": 4.567415730337078, + "grad_norm": 0.9681848883628845, + "learning_rate": 6.784901621382187e-07, + "loss": 0.521, + "step": 13008 + }, + { + "epoch": 4.567766853932584, + "grad_norm": 0.9552080035209656, + "learning_rate": 6.781735304117651e-07, + "loss": 0.527, + "step": 13009 + }, + { + "epoch": 4.56811797752809, + "grad_norm": 0.9797877073287964, + "learning_rate": 6.778569609902172e-07, + "loss": 0.5162, + "step": 13010 + }, + { + "epoch": 4.568469101123595, + "grad_norm": 0.9839064478874207, + "learning_rate": 6.775404538844027e-07, + "loss": 0.5435, + "step": 13011 + }, + { + "epoch": 4.568820224719101, + "grad_norm": 0.9923290014266968, + "learning_rate": 6.772240091051443e-07, + "loss": 0.5178, + "step": 13012 + }, + { + "epoch": 4.569171348314606, + "grad_norm": 1.01504647731781, + "learning_rate": 6.76907626663266e-07, + "loss": 0.5505, + "step": 13013 + }, + { + "epoch": 4.569522471910112, + "grad_norm": 0.9788626432418823, + "learning_rate": 6.765913065695858e-07, + "loss": 0.5183, + "step": 13014 + }, + { + "epoch": 4.569873595505618, + "grad_norm": 1.007380723953247, + "learning_rate": 6.762750488349238e-07, + "loss": 0.5237, + "step": 13015 + }, + { + "epoch": 4.570224719101123, + "grad_norm": 0.9748814702033997, + "learning_rate": 6.759588534700933e-07, + "loss": 0.5528, + "step": 13016 + }, + { + "epoch": 4.570575842696629, + "grad_norm": 0.953889787197113, + "learning_rate": 6.756427204859101e-07, + "loss": 0.5355, + "step": 13017 + }, + { + "epoch": 4.570926966292134, + "grad_norm": 0.9703826308250427, + "learning_rate": 6.753266498931848e-07, + "loss": 0.5612, + "step": 13018 + }, + { + "epoch": 4.57127808988764, + "grad_norm": 0.9859398603439331, + "learning_rate": 6.750106417027258e-07, + "loss": 0.5728, + "step": 13019 + }, + { + "epoch": 4.571629213483146, + "grad_norm": 0.9914897084236145, + "learning_rate": 6.746946959253423e-07, + "loss": 0.5482, + "step": 13020 + }, + { + "epoch": 4.571980337078652, + "grad_norm": 0.9615387320518494, + "learning_rate": 6.74378812571837e-07, + "loss": 0.5316, + "step": 13021 + }, + { + "epoch": 4.572331460674158, + "grad_norm": 0.9912812113761902, + "learning_rate": 6.74062991653015e-07, + "loss": 0.5148, + "step": 13022 + }, + { + "epoch": 4.572682584269663, + "grad_norm": 0.9682676196098328, + "learning_rate": 6.737472331796757e-07, + "loss": 0.5548, + "step": 13023 + }, + { + "epoch": 4.573033707865169, + "grad_norm": 0.9843235611915588, + "learning_rate": 6.734315371626188e-07, + "loss": 0.5246, + "step": 13024 + }, + { + "epoch": 4.573384831460674, + "grad_norm": 0.957760751247406, + "learning_rate": 6.731159036126397e-07, + "loss": 0.5292, + "step": 13025 + }, + { + "epoch": 4.57373595505618, + "grad_norm": 0.9456557631492615, + "learning_rate": 6.728003325405342e-07, + "loss": 0.5343, + "step": 13026 + }, + { + "epoch": 4.574087078651686, + "grad_norm": 0.9815381765365601, + "learning_rate": 6.724848239570927e-07, + "loss": 0.533, + "step": 13027 + }, + { + "epoch": 4.574438202247191, + "grad_norm": 0.9304009675979614, + "learning_rate": 6.721693778731076e-07, + "loss": 0.5201, + "step": 13028 + }, + { + "epoch": 4.574789325842697, + "grad_norm": 0.9983648061752319, + "learning_rate": 6.718539942993646e-07, + "loss": 0.5366, + "step": 13029 + }, + { + "epoch": 4.575140449438202, + "grad_norm": 0.9526093006134033, + "learning_rate": 6.715386732466514e-07, + "loss": 0.5177, + "step": 13030 + }, + { + "epoch": 4.575491573033708, + "grad_norm": 1.0040122270584106, + "learning_rate": 6.71223414725751e-07, + "loss": 0.5475, + "step": 13031 + }, + { + "epoch": 4.575842696629214, + "grad_norm": 0.9804515838623047, + "learning_rate": 6.709082187474442e-07, + "loss": 0.5369, + "step": 13032 + }, + { + "epoch": 4.576193820224719, + "grad_norm": 0.9872512817382812, + "learning_rate": 6.705930853225118e-07, + "loss": 0.5308, + "step": 13033 + }, + { + "epoch": 4.576544943820225, + "grad_norm": 0.9797122478485107, + "learning_rate": 6.702780144617297e-07, + "loss": 0.5777, + "step": 13034 + }, + { + "epoch": 4.57689606741573, + "grad_norm": 0.9787207841873169, + "learning_rate": 6.699630061758746e-07, + "loss": 0.5333, + "step": 13035 + }, + { + "epoch": 4.577247191011236, + "grad_norm": 0.9692382216453552, + "learning_rate": 6.69648060475718e-07, + "loss": 0.4976, + "step": 13036 + }, + { + "epoch": 4.577598314606742, + "grad_norm": 0.9552083015441895, + "learning_rate": 6.693331773720322e-07, + "loss": 0.5623, + "step": 13037 + }, + { + "epoch": 4.577949438202247, + "grad_norm": 0.941242516040802, + "learning_rate": 6.690183568755843e-07, + "loss": 0.5255, + "step": 13038 + }, + { + "epoch": 4.578300561797753, + "grad_norm": 0.9922299385070801, + "learning_rate": 6.687035989971427e-07, + "loss": 0.5602, + "step": 13039 + }, + { + "epoch": 4.578651685393258, + "grad_norm": 0.9815553426742554, + "learning_rate": 6.683889037474703e-07, + "loss": 0.5582, + "step": 13040 + }, + { + "epoch": 4.579002808988764, + "grad_norm": 0.9601178169250488, + "learning_rate": 6.680742711373306e-07, + "loss": 0.5137, + "step": 13041 + }, + { + "epoch": 4.57935393258427, + "grad_norm": 0.9814119338989258, + "learning_rate": 6.677597011774825e-07, + "loss": 0.573, + "step": 13042 + }, + { + "epoch": 4.579705056179775, + "grad_norm": 0.991087794303894, + "learning_rate": 6.674451938786855e-07, + "loss": 0.5549, + "step": 13043 + }, + { + "epoch": 4.580056179775281, + "grad_norm": 0.958419919013977, + "learning_rate": 6.671307492516949e-07, + "loss": 0.5613, + "step": 13044 + }, + { + "epoch": 4.580407303370786, + "grad_norm": 0.968169629573822, + "learning_rate": 6.668163673072633e-07, + "loss": 0.4943, + "step": 13045 + }, + { + "epoch": 4.580758426966292, + "grad_norm": 0.9397329092025757, + "learning_rate": 6.665020480561443e-07, + "loss": 0.4886, + "step": 13046 + }, + { + "epoch": 4.581109550561798, + "grad_norm": 0.9662849307060242, + "learning_rate": 6.661877915090853e-07, + "loss": 0.5208, + "step": 13047 + }, + { + "epoch": 4.581460674157303, + "grad_norm": 0.9699057340621948, + "learning_rate": 6.658735976768354e-07, + "loss": 0.5203, + "step": 13048 + }, + { + "epoch": 4.581811797752809, + "grad_norm": 0.9455240368843079, + "learning_rate": 6.655594665701381e-07, + "loss": 0.5107, + "step": 13049 + }, + { + "epoch": 4.582162921348314, + "grad_norm": 0.9591947197914124, + "learning_rate": 6.652453981997383e-07, + "loss": 0.557, + "step": 13050 + }, + { + "epoch": 4.58251404494382, + "grad_norm": 1.0016686916351318, + "learning_rate": 6.64931392576375e-07, + "loss": 0.5664, + "step": 13051 + }, + { + "epoch": 4.582865168539326, + "grad_norm": 0.9602221250534058, + "learning_rate": 6.646174497107885e-07, + "loss": 0.4986, + "step": 13052 + }, + { + "epoch": 4.583216292134831, + "grad_norm": 0.9449904561042786, + "learning_rate": 6.64303569613714e-07, + "loss": 0.5284, + "step": 13053 + }, + { + "epoch": 4.583567415730337, + "grad_norm": 0.984626293182373, + "learning_rate": 6.639897522958875e-07, + "loss": 0.5371, + "step": 13054 + }, + { + "epoch": 4.583918539325842, + "grad_norm": 1.0079944133758545, + "learning_rate": 6.636759977680394e-07, + "loss": 0.5227, + "step": 13055 + }, + { + "epoch": 4.584269662921348, + "grad_norm": 0.9727509617805481, + "learning_rate": 6.633623060409015e-07, + "loss": 0.5689, + "step": 13056 + }, + { + "epoch": 4.584620786516854, + "grad_norm": 0.972427487373352, + "learning_rate": 6.630486771252012e-07, + "loss": 0.5106, + "step": 13057 + }, + { + "epoch": 4.58497191011236, + "grad_norm": 0.9831407070159912, + "learning_rate": 6.627351110316635e-07, + "loss": 0.5363, + "step": 13058 + }, + { + "epoch": 4.585323033707866, + "grad_norm": 0.9838858246803284, + "learning_rate": 6.624216077710136e-07, + "loss": 0.59, + "step": 13059 + }, + { + "epoch": 4.585674157303371, + "grad_norm": 0.9687826633453369, + "learning_rate": 6.621081673539714e-07, + "loss": 0.5222, + "step": 13060 + }, + { + "epoch": 4.586025280898877, + "grad_norm": 0.9716885685920715, + "learning_rate": 6.61794789791258e-07, + "loss": 0.5798, + "step": 13061 + }, + { + "epoch": 4.586376404494382, + "grad_norm": 0.9758996963500977, + "learning_rate": 6.614814750935892e-07, + "loss": 0.5471, + "step": 13062 + }, + { + "epoch": 4.586727528089888, + "grad_norm": 0.9454587697982788, + "learning_rate": 6.611682232716812e-07, + "loss": 0.5595, + "step": 13063 + }, + { + "epoch": 4.587078651685394, + "grad_norm": 0.9692065119743347, + "learning_rate": 6.608550343362458e-07, + "loss": 0.5641, + "step": 13064 + }, + { + "epoch": 4.587429775280899, + "grad_norm": 0.9849277138710022, + "learning_rate": 6.605419082979952e-07, + "loss": 0.5243, + "step": 13065 + }, + { + "epoch": 4.587780898876405, + "grad_norm": 0.9846190810203552, + "learning_rate": 6.602288451676367e-07, + "loss": 0.5244, + "step": 13066 + }, + { + "epoch": 4.58813202247191, + "grad_norm": 0.9637678861618042, + "learning_rate": 6.599158449558779e-07, + "loss": 0.5356, + "step": 13067 + }, + { + "epoch": 4.588483146067416, + "grad_norm": 0.9640319347381592, + "learning_rate": 6.596029076734228e-07, + "loss": 0.5491, + "step": 13068 + }, + { + "epoch": 4.588834269662922, + "grad_norm": 0.9582656025886536, + "learning_rate": 6.592900333309727e-07, + "loss": 0.544, + "step": 13069 + }, + { + "epoch": 4.589185393258427, + "grad_norm": 0.9265760779380798, + "learning_rate": 6.589772219392288e-07, + "loss": 0.53, + "step": 13070 + }, + { + "epoch": 4.589536516853933, + "grad_norm": 0.9436482787132263, + "learning_rate": 6.586644735088879e-07, + "loss": 0.5153, + "step": 13071 + }, + { + "epoch": 4.589887640449438, + "grad_norm": 0.9416664242744446, + "learning_rate": 6.583517880506471e-07, + "loss": 0.5449, + "step": 13072 + }, + { + "epoch": 4.590238764044944, + "grad_norm": 0.9959672689437866, + "learning_rate": 6.580391655751986e-07, + "loss": 0.571, + "step": 13073 + }, + { + "epoch": 4.59058988764045, + "grad_norm": 0.9721177220344543, + "learning_rate": 6.577266060932353e-07, + "loss": 0.5398, + "step": 13074 + }, + { + "epoch": 4.590941011235955, + "grad_norm": 0.965003252029419, + "learning_rate": 6.574141096154446e-07, + "loss": 0.524, + "step": 13075 + }, + { + "epoch": 4.591292134831461, + "grad_norm": 0.9521030783653259, + "learning_rate": 6.571016761525153e-07, + "loss": 0.5434, + "step": 13076 + }, + { + "epoch": 4.591643258426966, + "grad_norm": 0.9751164317131042, + "learning_rate": 6.567893057151311e-07, + "loss": 0.5392, + "step": 13077 + }, + { + "epoch": 4.591994382022472, + "grad_norm": 1.0048350095748901, + "learning_rate": 6.564769983139762e-07, + "loss": 0.5493, + "step": 13078 + }, + { + "epoch": 4.592345505617978, + "grad_norm": 0.9547953605651855, + "learning_rate": 6.561647539597296e-07, + "loss": 0.5479, + "step": 13079 + }, + { + "epoch": 4.592696629213483, + "grad_norm": 0.9828712940216064, + "learning_rate": 6.558525726630713e-07, + "loss": 0.5339, + "step": 13080 + }, + { + "epoch": 4.593047752808989, + "grad_norm": 0.9622471928596497, + "learning_rate": 6.555404544346772e-07, + "loss": 0.537, + "step": 13081 + }, + { + "epoch": 4.593398876404494, + "grad_norm": 0.9692254662513733, + "learning_rate": 6.552283992852204e-07, + "loss": 0.5323, + "step": 13082 + }, + { + "epoch": 4.59375, + "grad_norm": 1.000786304473877, + "learning_rate": 6.549164072253744e-07, + "loss": 0.5806, + "step": 13083 + }, + { + "epoch": 4.594101123595506, + "grad_norm": 0.9831902980804443, + "learning_rate": 6.546044782658076e-07, + "loss": 0.5548, + "step": 13084 + }, + { + "epoch": 4.594452247191011, + "grad_norm": 0.9294179677963257, + "learning_rate": 6.542926124171895e-07, + "loss": 0.5233, + "step": 13085 + }, + { + "epoch": 4.594803370786517, + "grad_norm": 0.9782955050468445, + "learning_rate": 6.539808096901837e-07, + "loss": 0.5343, + "step": 13086 + }, + { + "epoch": 4.595154494382022, + "grad_norm": 0.9722889065742493, + "learning_rate": 6.536690700954553e-07, + "loss": 0.5416, + "step": 13087 + }, + { + "epoch": 4.595505617977528, + "grad_norm": 0.9732030630111694, + "learning_rate": 6.533573936436649e-07, + "loss": 0.5847, + "step": 13088 + }, + { + "epoch": 4.595856741573034, + "grad_norm": 0.9683316349983215, + "learning_rate": 6.530457803454706e-07, + "loss": 0.5525, + "step": 13089 + }, + { + "epoch": 4.596207865168539, + "grad_norm": 0.9720481634140015, + "learning_rate": 6.527342302115308e-07, + "loss": 0.5548, + "step": 13090 + }, + { + "epoch": 4.596558988764045, + "grad_norm": 0.9742597937583923, + "learning_rate": 6.52422743252499e-07, + "loss": 0.5499, + "step": 13091 + }, + { + "epoch": 4.59691011235955, + "grad_norm": 0.9776440262794495, + "learning_rate": 6.521113194790291e-07, + "loss": 0.5469, + "step": 13092 + }, + { + "epoch": 4.597261235955056, + "grad_norm": 0.9986089468002319, + "learning_rate": 6.517999589017701e-07, + "loss": 0.5503, + "step": 13093 + }, + { + "epoch": 4.597612359550562, + "grad_norm": 1.0382941961288452, + "learning_rate": 6.514886615313717e-07, + "loss": 0.5113, + "step": 13094 + }, + { + "epoch": 4.597963483146067, + "grad_norm": 0.9720661044120789, + "learning_rate": 6.511774273784794e-07, + "loss": 0.5396, + "step": 13095 + }, + { + "epoch": 4.598314606741573, + "grad_norm": 0.964362382888794, + "learning_rate": 6.50866256453736e-07, + "loss": 0.4965, + "step": 13096 + }, + { + "epoch": 4.598665730337078, + "grad_norm": 0.9602935910224915, + "learning_rate": 6.505551487677853e-07, + "loss": 0.5179, + "step": 13097 + }, + { + "epoch": 4.599016853932584, + "grad_norm": 0.9549167156219482, + "learning_rate": 6.502441043312652e-07, + "loss": 0.5479, + "step": 13098 + }, + { + "epoch": 4.59936797752809, + "grad_norm": 0.9575080275535583, + "learning_rate": 6.499331231548145e-07, + "loss": 0.5622, + "step": 13099 + }, + { + "epoch": 4.599719101123595, + "grad_norm": 0.988008439540863, + "learning_rate": 6.496222052490669e-07, + "loss": 0.564, + "step": 13100 + }, + { + "epoch": 4.600070224719101, + "grad_norm": 0.9859716296195984, + "learning_rate": 6.493113506246576e-07, + "loss": 0.5582, + "step": 13101 + }, + { + "epoch": 4.600421348314606, + "grad_norm": 1.0077072381973267, + "learning_rate": 6.490005592922164e-07, + "loss": 0.5296, + "step": 13102 + }, + { + "epoch": 4.600772471910112, + "grad_norm": 0.9465804100036621, + "learning_rate": 6.486898312623713e-07, + "loss": 0.5336, + "step": 13103 + }, + { + "epoch": 4.601123595505618, + "grad_norm": 0.9484519362449646, + "learning_rate": 6.483791665457503e-07, + "loss": 0.5413, + "step": 13104 + }, + { + "epoch": 4.601474719101123, + "grad_norm": 0.9623340964317322, + "learning_rate": 6.48068565152977e-07, + "loss": 0.551, + "step": 13105 + }, + { + "epoch": 4.601825842696629, + "grad_norm": 0.9676215052604675, + "learning_rate": 6.477580270946745e-07, + "loss": 0.5646, + "step": 13106 + }, + { + "epoch": 4.602176966292134, + "grad_norm": 0.9622493386268616, + "learning_rate": 6.474475523814622e-07, + "loss": 0.5568, + "step": 13107 + }, + { + "epoch": 4.60252808988764, + "grad_norm": 0.9520705342292786, + "learning_rate": 6.471371410239585e-07, + "loss": 0.5401, + "step": 13108 + }, + { + "epoch": 4.602879213483146, + "grad_norm": 0.9789988398551941, + "learning_rate": 6.468267930327787e-07, + "loss": 0.5541, + "step": 13109 + }, + { + "epoch": 4.603230337078652, + "grad_norm": 1.0097423791885376, + "learning_rate": 6.465165084185373e-07, + "loss": 0.5632, + "step": 13110 + }, + { + "epoch": 4.603581460674158, + "grad_norm": 0.9704419374465942, + "learning_rate": 6.462062871918456e-07, + "loss": 0.5216, + "step": 13111 + }, + { + "epoch": 4.603932584269663, + "grad_norm": 0.9579986333847046, + "learning_rate": 6.458961293633115e-07, + "loss": 0.5458, + "step": 13112 + }, + { + "epoch": 4.604283707865169, + "grad_norm": 0.9729369878768921, + "learning_rate": 6.455860349435442e-07, + "loss": 0.5525, + "step": 13113 + }, + { + "epoch": 4.604634831460674, + "grad_norm": 0.9907523393630981, + "learning_rate": 6.452760039431469e-07, + "loss": 0.5299, + "step": 13114 + }, + { + "epoch": 4.60498595505618, + "grad_norm": 0.9946016073226929, + "learning_rate": 6.449660363727236e-07, + "loss": 0.5428, + "step": 13115 + }, + { + "epoch": 4.605337078651686, + "grad_norm": 0.9927729368209839, + "learning_rate": 6.44656132242874e-07, + "loss": 0.5264, + "step": 13116 + }, + { + "epoch": 4.605688202247191, + "grad_norm": 0.9412595629692078, + "learning_rate": 6.443462915641976e-07, + "loss": 0.525, + "step": 13117 + }, + { + "epoch": 4.606039325842697, + "grad_norm": 0.9675086140632629, + "learning_rate": 6.440365143472895e-07, + "loss": 0.5847, + "step": 13118 + }, + { + "epoch": 4.606390449438202, + "grad_norm": 0.9966561198234558, + "learning_rate": 6.437268006027453e-07, + "loss": 0.5537, + "step": 13119 + }, + { + "epoch": 4.606741573033708, + "grad_norm": 0.9701477289199829, + "learning_rate": 6.434171503411557e-07, + "loss": 0.5346, + "step": 13120 + }, + { + "epoch": 4.607092696629214, + "grad_norm": 0.967951774597168, + "learning_rate": 6.431075635731105e-07, + "loss": 0.5347, + "step": 13121 + }, + { + "epoch": 4.607443820224719, + "grad_norm": 0.9650239944458008, + "learning_rate": 6.427980403091983e-07, + "loss": 0.5389, + "step": 13122 + }, + { + "epoch": 4.607794943820225, + "grad_norm": 0.9983237385749817, + "learning_rate": 6.424885805600031e-07, + "loss": 0.5768, + "step": 13123 + }, + { + "epoch": 4.60814606741573, + "grad_norm": 0.9799590706825256, + "learning_rate": 6.421791843361097e-07, + "loss": 0.5561, + "step": 13124 + }, + { + "epoch": 4.608497191011236, + "grad_norm": 0.9402680397033691, + "learning_rate": 6.418698516480976e-07, + "loss": 0.5125, + "step": 13125 + }, + { + "epoch": 4.608848314606742, + "grad_norm": 0.9475598335266113, + "learning_rate": 6.415605825065474e-07, + "loss": 0.5268, + "step": 13126 + }, + { + "epoch": 4.609199438202247, + "grad_norm": 1.209362506866455, + "learning_rate": 6.412513769220341e-07, + "loss": 0.5495, + "step": 13127 + }, + { + "epoch": 4.609550561797753, + "grad_norm": 0.9566494226455688, + "learning_rate": 6.409422349051342e-07, + "loss": 0.5595, + "step": 13128 + }, + { + "epoch": 4.609901685393258, + "grad_norm": 0.973689615726471, + "learning_rate": 6.406331564664181e-07, + "loss": 0.5329, + "step": 13129 + }, + { + "epoch": 4.610252808988764, + "grad_norm": 0.9601119160652161, + "learning_rate": 6.403241416164577e-07, + "loss": 0.5435, + "step": 13130 + }, + { + "epoch": 4.61060393258427, + "grad_norm": 0.9869250655174255, + "learning_rate": 6.400151903658202e-07, + "loss": 0.5448, + "step": 13131 + }, + { + "epoch": 4.610955056179775, + "grad_norm": 0.961050271987915, + "learning_rate": 6.397063027250711e-07, + "loss": 0.5369, + "step": 13132 + }, + { + "epoch": 4.611306179775281, + "grad_norm": 0.9418216347694397, + "learning_rate": 6.393974787047752e-07, + "loss": 0.4964, + "step": 13133 + }, + { + "epoch": 4.611657303370786, + "grad_norm": 0.9514720439910889, + "learning_rate": 6.390887183154929e-07, + "loss": 0.5324, + "step": 13134 + }, + { + "epoch": 4.612008426966292, + "grad_norm": 0.9489847421646118, + "learning_rate": 6.387800215677845e-07, + "loss": 0.5615, + "step": 13135 + }, + { + "epoch": 4.612359550561798, + "grad_norm": 0.9581836462020874, + "learning_rate": 6.384713884722057e-07, + "loss": 0.5363, + "step": 13136 + }, + { + "epoch": 4.612710674157303, + "grad_norm": 0.9362204670906067, + "learning_rate": 6.381628190393136e-07, + "loss": 0.4981, + "step": 13137 + }, + { + "epoch": 4.613061797752809, + "grad_norm": 0.978590190410614, + "learning_rate": 6.378543132796591e-07, + "loss": 0.5629, + "step": 13138 + }, + { + "epoch": 4.613412921348314, + "grad_norm": 0.9572652578353882, + "learning_rate": 6.375458712037943e-07, + "loss": 0.5293, + "step": 13139 + }, + { + "epoch": 4.61376404494382, + "grad_norm": 0.9404938817024231, + "learning_rate": 6.372374928222663e-07, + "loss": 0.518, + "step": 13140 + }, + { + "epoch": 4.614115168539326, + "grad_norm": 0.9696025848388672, + "learning_rate": 6.369291781456227e-07, + "loss": 0.5591, + "step": 13141 + }, + { + "epoch": 4.614466292134831, + "grad_norm": 0.9733339548110962, + "learning_rate": 6.366209271844062e-07, + "loss": 0.4982, + "step": 13142 + }, + { + "epoch": 4.614817415730337, + "grad_norm": 0.9506546854972839, + "learning_rate": 6.363127399491603e-07, + "loss": 0.4923, + "step": 13143 + }, + { + "epoch": 4.615168539325842, + "grad_norm": 0.9769538044929504, + "learning_rate": 6.360046164504238e-07, + "loss": 0.5256, + "step": 13144 + }, + { + "epoch": 4.615519662921348, + "grad_norm": 0.9405174255371094, + "learning_rate": 6.356965566987339e-07, + "loss": 0.5051, + "step": 13145 + }, + { + "epoch": 4.615870786516854, + "grad_norm": 0.9740371108055115, + "learning_rate": 6.35388560704627e-07, + "loss": 0.5582, + "step": 13146 + }, + { + "epoch": 4.61622191011236, + "grad_norm": 0.9787911772727966, + "learning_rate": 6.350806284786348e-07, + "loss": 0.4972, + "step": 13147 + }, + { + "epoch": 4.616573033707866, + "grad_norm": 0.9610333442687988, + "learning_rate": 6.347727600312903e-07, + "loss": 0.5548, + "step": 13148 + }, + { + "epoch": 4.616924157303371, + "grad_norm": 0.9665414094924927, + "learning_rate": 6.344649553731205e-07, + "loss": 0.571, + "step": 13149 + }, + { + "epoch": 4.617275280898877, + "grad_norm": 0.9479056596755981, + "learning_rate": 6.341572145146537e-07, + "loss": 0.5491, + "step": 13150 + }, + { + "epoch": 4.617626404494382, + "grad_norm": 0.9505181312561035, + "learning_rate": 6.338495374664128e-07, + "loss": 0.5327, + "step": 13151 + }, + { + "epoch": 4.617977528089888, + "grad_norm": 0.9968332648277283, + "learning_rate": 6.335419242389213e-07, + "loss": 0.5131, + "step": 13152 + }, + { + "epoch": 4.618328651685394, + "grad_norm": 0.9414198398590088, + "learning_rate": 6.332343748426984e-07, + "loss": 0.5107, + "step": 13153 + }, + { + "epoch": 4.618679775280899, + "grad_norm": 0.970410943031311, + "learning_rate": 6.329268892882631e-07, + "loss": 0.5252, + "step": 13154 + }, + { + "epoch": 4.619030898876405, + "grad_norm": 0.9687212109565735, + "learning_rate": 6.326194675861297e-07, + "loss": 0.5685, + "step": 13155 + }, + { + "epoch": 4.61938202247191, + "grad_norm": 0.9956013560295105, + "learning_rate": 6.323121097468132e-07, + "loss": 0.526, + "step": 13156 + }, + { + "epoch": 4.619733146067416, + "grad_norm": 0.9655634164810181, + "learning_rate": 6.320048157808248e-07, + "loss": 0.5184, + "step": 13157 + }, + { + "epoch": 4.620084269662922, + "grad_norm": 0.9292744994163513, + "learning_rate": 6.316975856986721e-07, + "loss": 0.5492, + "step": 13158 + }, + { + "epoch": 4.620435393258427, + "grad_norm": 0.9820828437805176, + "learning_rate": 6.313904195108642e-07, + "loss": 0.5397, + "step": 13159 + }, + { + "epoch": 4.620786516853933, + "grad_norm": 0.9594804048538208, + "learning_rate": 6.310833172279041e-07, + "loss": 0.5228, + "step": 13160 + }, + { + "epoch": 4.621137640449438, + "grad_norm": 0.9881440997123718, + "learning_rate": 6.307762788602959e-07, + "loss": 0.5636, + "step": 13161 + }, + { + "epoch": 4.621488764044944, + "grad_norm": 0.9336093664169312, + "learning_rate": 6.30469304418539e-07, + "loss": 0.5247, + "step": 13162 + }, + { + "epoch": 4.62183988764045, + "grad_norm": 0.9616424441337585, + "learning_rate": 6.301623939131324e-07, + "loss": 0.5346, + "step": 13163 + }, + { + "epoch": 4.622191011235955, + "grad_norm": 0.9890590310096741, + "learning_rate": 6.298555473545715e-07, + "loss": 0.5265, + "step": 13164 + }, + { + "epoch": 4.622542134831461, + "grad_norm": 0.9748391509056091, + "learning_rate": 6.295487647533514e-07, + "loss": 0.5349, + "step": 13165 + }, + { + "epoch": 4.622893258426966, + "grad_norm": 0.9524152278900146, + "learning_rate": 6.292420461199619e-07, + "loss": 0.5055, + "step": 13166 + }, + { + "epoch": 4.623244382022472, + "grad_norm": 0.9736615419387817, + "learning_rate": 6.289353914648946e-07, + "loss": 0.5501, + "step": 13167 + }, + { + "epoch": 4.623595505617978, + "grad_norm": 1.00412118434906, + "learning_rate": 6.286288007986349e-07, + "loss": 0.5647, + "step": 13168 + }, + { + "epoch": 4.623946629213483, + "grad_norm": 0.9906489849090576, + "learning_rate": 6.283222741316697e-07, + "loss": 0.5504, + "step": 13169 + }, + { + "epoch": 4.624297752808989, + "grad_norm": 0.9487780928611755, + "learning_rate": 6.280158114744811e-07, + "loss": 0.5054, + "step": 13170 + }, + { + "epoch": 4.624648876404494, + "grad_norm": 0.9441604614257812, + "learning_rate": 6.277094128375493e-07, + "loss": 0.5267, + "step": 13171 + }, + { + "epoch": 4.625, + "grad_norm": 0.9667211174964905, + "learning_rate": 6.274030782313539e-07, + "loss": 0.5273, + "step": 13172 + }, + { + "epoch": 4.625351123595506, + "grad_norm": 0.9877879023551941, + "learning_rate": 6.270968076663703e-07, + "loss": 0.5434, + "step": 13173 + }, + { + "epoch": 4.625702247191011, + "grad_norm": 0.9511591196060181, + "learning_rate": 6.267906011530739e-07, + "loss": 0.5415, + "step": 13174 + }, + { + "epoch": 4.626053370786517, + "grad_norm": 1.0068329572677612, + "learning_rate": 6.26484458701935e-07, + "loss": 0.5721, + "step": 13175 + }, + { + "epoch": 4.626404494382022, + "grad_norm": 0.9765903949737549, + "learning_rate": 6.261783803234256e-07, + "loss": 0.5625, + "step": 13176 + }, + { + "epoch": 4.626755617977528, + "grad_norm": 0.9612672924995422, + "learning_rate": 6.258723660280111e-07, + "loss": 0.5543, + "step": 13177 + }, + { + "epoch": 4.627106741573034, + "grad_norm": 0.988845944404602, + "learning_rate": 6.255664158261587e-07, + "loss": 0.5315, + "step": 13178 + }, + { + "epoch": 4.627457865168539, + "grad_norm": 0.9803781509399414, + "learning_rate": 6.252605297283304e-07, + "loss": 0.5629, + "step": 13179 + }, + { + "epoch": 4.627808988764045, + "grad_norm": 0.9516044855117798, + "learning_rate": 6.249547077449883e-07, + "loss": 0.5267, + "step": 13180 + }, + { + "epoch": 4.62816011235955, + "grad_norm": 0.9843720197677612, + "learning_rate": 6.246489498865898e-07, + "loss": 0.5683, + "step": 13181 + }, + { + "epoch": 4.628511235955056, + "grad_norm": 0.9921948313713074, + "learning_rate": 6.243432561635934e-07, + "loss": 0.5545, + "step": 13182 + }, + { + "epoch": 4.628862359550562, + "grad_norm": 0.972173810005188, + "learning_rate": 6.240376265864529e-07, + "loss": 0.533, + "step": 13183 + }, + { + "epoch": 4.629213483146067, + "grad_norm": 1.0266437530517578, + "learning_rate": 6.237320611656192e-07, + "loss": 0.5743, + "step": 13184 + }, + { + "epoch": 4.629564606741573, + "grad_norm": 0.9811228513717651, + "learning_rate": 6.234265599115441e-07, + "loss": 0.5515, + "step": 13185 + }, + { + "epoch": 4.629915730337078, + "grad_norm": 0.9850162267684937, + "learning_rate": 6.231211228346743e-07, + "loss": 0.562, + "step": 13186 + }, + { + "epoch": 4.630266853932584, + "grad_norm": 0.9367568492889404, + "learning_rate": 6.228157499454567e-07, + "loss": 0.5154, + "step": 13187 + }, + { + "epoch": 4.63061797752809, + "grad_norm": 0.9380578994750977, + "learning_rate": 6.225104412543334e-07, + "loss": 0.5342, + "step": 13188 + }, + { + "epoch": 4.630969101123595, + "grad_norm": 0.9941267371177673, + "learning_rate": 6.222051967717471e-07, + "loss": 0.5656, + "step": 13189 + }, + { + "epoch": 4.631320224719101, + "grad_norm": 0.9372175931930542, + "learning_rate": 6.219000165081357e-07, + "loss": 0.5064, + "step": 13190 + }, + { + "epoch": 4.631671348314606, + "grad_norm": 0.9722461104393005, + "learning_rate": 6.215949004739372e-07, + "loss": 0.561, + "step": 13191 + }, + { + "epoch": 4.632022471910112, + "grad_norm": 0.9538682103157043, + "learning_rate": 6.212898486795849e-07, + "loss": 0.5239, + "step": 13192 + }, + { + "epoch": 4.632373595505618, + "grad_norm": 0.9554574489593506, + "learning_rate": 6.209848611355129e-07, + "loss": 0.5182, + "step": 13193 + }, + { + "epoch": 4.632724719101123, + "grad_norm": 0.9552617073059082, + "learning_rate": 6.206799378521508e-07, + "loss": 0.5031, + "step": 13194 + }, + { + "epoch": 4.633075842696629, + "grad_norm": 0.9672900438308716, + "learning_rate": 6.20375078839926e-07, + "loss": 0.5333, + "step": 13195 + }, + { + "epoch": 4.633426966292134, + "grad_norm": 0.9767590165138245, + "learning_rate": 6.20070284109266e-07, + "loss": 0.5161, + "step": 13196 + }, + { + "epoch": 4.63377808988764, + "grad_norm": 0.976703405380249, + "learning_rate": 6.197655536705923e-07, + "loss": 0.4915, + "step": 13197 + }, + { + "epoch": 4.634129213483146, + "grad_norm": 0.9744858741760254, + "learning_rate": 6.194608875343292e-07, + "loss": 0.5361, + "step": 13198 + }, + { + "epoch": 4.634480337078652, + "grad_norm": 0.9808706641197205, + "learning_rate": 6.191562857108932e-07, + "loss": 0.5621, + "step": 13199 + }, + { + "epoch": 4.634831460674158, + "grad_norm": 0.9652357697486877, + "learning_rate": 6.188517482107037e-07, + "loss": 0.5594, + "step": 13200 + }, + { + "epoch": 4.635182584269663, + "grad_norm": 1.0161337852478027, + "learning_rate": 6.18547275044174e-07, + "loss": 0.5759, + "step": 13201 + }, + { + "epoch": 4.635533707865169, + "grad_norm": 0.975017249584198, + "learning_rate": 6.182428662217182e-07, + "loss": 0.486, + "step": 13202 + }, + { + "epoch": 4.635884831460674, + "grad_norm": 0.961396336555481, + "learning_rate": 6.179385217537453e-07, + "loss": 0.523, + "step": 13203 + }, + { + "epoch": 4.63623595505618, + "grad_norm": 0.9972079992294312, + "learning_rate": 6.176342416506655e-07, + "loss": 0.56, + "step": 13204 + }, + { + "epoch": 4.636587078651686, + "grad_norm": 0.9826855659484863, + "learning_rate": 6.173300259228829e-07, + "loss": 0.5414, + "step": 13205 + }, + { + "epoch": 4.636938202247191, + "grad_norm": 0.9823887348175049, + "learning_rate": 6.170258745808033e-07, + "loss": 0.5408, + "step": 13206 + }, + { + "epoch": 4.637289325842697, + "grad_norm": 0.9633069634437561, + "learning_rate": 6.167217876348272e-07, + "loss": 0.5432, + "step": 13207 + }, + { + "epoch": 4.637640449438202, + "grad_norm": 0.9627605676651001, + "learning_rate": 6.16417765095354e-07, + "loss": 0.5222, + "step": 13208 + }, + { + "epoch": 4.637991573033708, + "grad_norm": 0.9803625345230103, + "learning_rate": 6.161138069727821e-07, + "loss": 0.5505, + "step": 13209 + }, + { + "epoch": 4.638342696629214, + "grad_norm": 0.933505117893219, + "learning_rate": 6.15809913277505e-07, + "loss": 0.4835, + "step": 13210 + }, + { + "epoch": 4.638693820224719, + "grad_norm": 1.0216162204742432, + "learning_rate": 6.155060840199173e-07, + "loss": 0.5749, + "step": 13211 + }, + { + "epoch": 4.639044943820225, + "grad_norm": 0.9491509199142456, + "learning_rate": 6.152023192104084e-07, + "loss": 0.5293, + "step": 13212 + }, + { + "epoch": 4.63939606741573, + "grad_norm": 0.9644910097122192, + "learning_rate": 6.148986188593681e-07, + "loss": 0.5903, + "step": 13213 + }, + { + "epoch": 4.639747191011236, + "grad_norm": 1.428976058959961, + "learning_rate": 6.145949829771811e-07, + "loss": 0.5233, + "step": 13214 + }, + { + "epoch": 4.640098314606742, + "grad_norm": 0.9533076286315918, + "learning_rate": 6.142914115742329e-07, + "loss": 0.4969, + "step": 13215 + }, + { + "epoch": 4.640449438202247, + "grad_norm": 0.9433438181877136, + "learning_rate": 6.139879046609043e-07, + "loss": 0.5092, + "step": 13216 + }, + { + "epoch": 4.640800561797753, + "grad_norm": 0.9918370246887207, + "learning_rate": 6.13684462247576e-07, + "loss": 0.5554, + "step": 13217 + }, + { + "epoch": 4.641151685393258, + "grad_norm": 0.9569703936576843, + "learning_rate": 6.133810843446245e-07, + "loss": 0.5208, + "step": 13218 + }, + { + "epoch": 4.641502808988764, + "grad_norm": 0.9539081454277039, + "learning_rate": 6.130777709624258e-07, + "loss": 0.5783, + "step": 13219 + }, + { + "epoch": 4.64185393258427, + "grad_norm": 0.9646636843681335, + "learning_rate": 6.127745221113532e-07, + "loss": 0.5268, + "step": 13220 + }, + { + "epoch": 4.642205056179775, + "grad_norm": 0.9520733952522278, + "learning_rate": 6.124713378017757e-07, + "loss": 0.5333, + "step": 13221 + }, + { + "epoch": 4.642556179775281, + "grad_norm": 0.983492910861969, + "learning_rate": 6.121682180440644e-07, + "loss": 0.5287, + "step": 13222 + }, + { + "epoch": 4.642907303370786, + "grad_norm": 1.2833534479141235, + "learning_rate": 6.118651628485839e-07, + "loss": 0.5318, + "step": 13223 + }, + { + "epoch": 4.643258426966292, + "grad_norm": 0.9538869261741638, + "learning_rate": 6.115621722256998e-07, + "loss": 0.5436, + "step": 13224 + }, + { + "epoch": 4.643609550561798, + "grad_norm": 0.9710142612457275, + "learning_rate": 6.112592461857728e-07, + "loss": 0.5242, + "step": 13225 + }, + { + "epoch": 4.643960674157303, + "grad_norm": 0.976517379283905, + "learning_rate": 6.10956384739164e-07, + "loss": 0.5854, + "step": 13226 + }, + { + "epoch": 4.644311797752809, + "grad_norm": 0.9857500791549683, + "learning_rate": 6.106535878962294e-07, + "loss": 0.5435, + "step": 13227 + }, + { + "epoch": 4.644662921348314, + "grad_norm": 1.0215729475021362, + "learning_rate": 6.103508556673263e-07, + "loss": 0.5767, + "step": 13228 + }, + { + "epoch": 4.64501404494382, + "grad_norm": 0.9794193506240845, + "learning_rate": 6.100481880628062e-07, + "loss": 0.5328, + "step": 13229 + }, + { + "epoch": 4.645365168539326, + "grad_norm": 0.9614959955215454, + "learning_rate": 6.097455850930214e-07, + "loss": 0.5737, + "step": 13230 + }, + { + "epoch": 4.645716292134831, + "grad_norm": 0.9767628908157349, + "learning_rate": 6.094430467683196e-07, + "loss": 0.5639, + "step": 13231 + }, + { + "epoch": 4.646067415730337, + "grad_norm": 1.0160518884658813, + "learning_rate": 6.091405730990485e-07, + "loss": 0.5544, + "step": 13232 + }, + { + "epoch": 4.646418539325842, + "grad_norm": 0.9630218744277954, + "learning_rate": 6.088381640955518e-07, + "loss": 0.5284, + "step": 13233 + }, + { + "epoch": 4.646769662921348, + "grad_norm": 0.9716809391975403, + "learning_rate": 6.085358197681707e-07, + "loss": 0.5365, + "step": 13234 + }, + { + "epoch": 4.647120786516854, + "grad_norm": 1.0035463571548462, + "learning_rate": 6.082335401272469e-07, + "loss": 0.5507, + "step": 13235 + }, + { + "epoch": 4.64747191011236, + "grad_norm": 0.9761444330215454, + "learning_rate": 6.079313251831162e-07, + "loss": 0.5481, + "step": 13236 + }, + { + "epoch": 4.647823033707866, + "grad_norm": 0.9473008513450623, + "learning_rate": 6.076291749461161e-07, + "loss": 0.5117, + "step": 13237 + }, + { + "epoch": 4.648174157303371, + "grad_norm": 0.9540848731994629, + "learning_rate": 6.073270894265782e-07, + "loss": 0.5124, + "step": 13238 + }, + { + "epoch": 4.648525280898877, + "grad_norm": 0.984452486038208, + "learning_rate": 6.070250686348351e-07, + "loss": 0.5144, + "step": 13239 + }, + { + "epoch": 4.648876404494382, + "grad_norm": 0.9752764105796814, + "learning_rate": 6.067231125812137e-07, + "loss": 0.5397, + "step": 13240 + }, + { + "epoch": 4.649227528089888, + "grad_norm": 0.9598808884620667, + "learning_rate": 6.064212212760426e-07, + "loss": 0.522, + "step": 13241 + }, + { + "epoch": 4.649578651685394, + "grad_norm": 0.9570513367652893, + "learning_rate": 6.061193947296448e-07, + "loss": 0.5271, + "step": 13242 + }, + { + "epoch": 4.649929775280899, + "grad_norm": 0.9825669527053833, + "learning_rate": 6.058176329523438e-07, + "loss": 0.5317, + "step": 13243 + }, + { + "epoch": 4.650280898876405, + "grad_norm": 0.970656156539917, + "learning_rate": 6.05515935954458e-07, + "loss": 0.5269, + "step": 13244 + }, + { + "epoch": 4.65063202247191, + "grad_norm": 0.9916336536407471, + "learning_rate": 6.052143037463068e-07, + "loss": 0.5472, + "step": 13245 + }, + { + "epoch": 4.650983146067416, + "grad_norm": 0.9639266729354858, + "learning_rate": 6.049127363382052e-07, + "loss": 0.5131, + "step": 13246 + }, + { + "epoch": 4.651334269662922, + "grad_norm": 0.9721720814704895, + "learning_rate": 6.046112337404655e-07, + "loss": 0.5397, + "step": 13247 + }, + { + "epoch": 4.651685393258427, + "grad_norm": 0.9869018793106079, + "learning_rate": 6.043097959634004e-07, + "loss": 0.5461, + "step": 13248 + }, + { + "epoch": 4.652036516853933, + "grad_norm": 0.9812648296356201, + "learning_rate": 6.040084230173174e-07, + "loss": 0.53, + "step": 13249 + }, + { + "epoch": 4.652387640449438, + "grad_norm": 0.9586833715438843, + "learning_rate": 6.037071149125248e-07, + "loss": 0.5326, + "step": 13250 + }, + { + "epoch": 4.652738764044944, + "grad_norm": 0.9888482093811035, + "learning_rate": 6.034058716593253e-07, + "loss": 0.5739, + "step": 13251 + }, + { + "epoch": 4.65308988764045, + "grad_norm": 0.9887607097625732, + "learning_rate": 6.031046932680229e-07, + "loss": 0.5567, + "step": 13252 + }, + { + "epoch": 4.653441011235955, + "grad_norm": 0.9715023636817932, + "learning_rate": 6.02803579748916e-07, + "loss": 0.5463, + "step": 13253 + }, + { + "epoch": 4.653792134831461, + "grad_norm": 0.9971110224723816, + "learning_rate": 6.025025311123037e-07, + "loss": 0.5312, + "step": 13254 + }, + { + "epoch": 4.654143258426966, + "grad_norm": 0.9475639462471008, + "learning_rate": 6.022015473684814e-07, + "loss": 0.507, + "step": 13255 + }, + { + "epoch": 4.654494382022472, + "grad_norm": 0.9641985893249512, + "learning_rate": 6.019006285277415e-07, + "loss": 0.5615, + "step": 13256 + }, + { + "epoch": 4.654845505617978, + "grad_norm": 0.9498912692070007, + "learning_rate": 6.015997746003765e-07, + "loss": 0.5236, + "step": 13257 + }, + { + "epoch": 4.655196629213483, + "grad_norm": 0.9444243907928467, + "learning_rate": 6.012989855966741e-07, + "loss": 0.4866, + "step": 13258 + }, + { + "epoch": 4.655547752808989, + "grad_norm": 0.9507916569709778, + "learning_rate": 6.009982615269222e-07, + "loss": 0.5131, + "step": 13259 + }, + { + "epoch": 4.655898876404494, + "grad_norm": 0.9335483312606812, + "learning_rate": 6.006976024014042e-07, + "loss": 0.5146, + "step": 13260 + }, + { + "epoch": 4.65625, + "grad_norm": 0.9477214813232422, + "learning_rate": 6.003970082304036e-07, + "loss": 0.5359, + "step": 13261 + }, + { + "epoch": 4.656601123595506, + "grad_norm": 1.0323386192321777, + "learning_rate": 6.000964790241998e-07, + "loss": 0.5204, + "step": 13262 + }, + { + "epoch": 4.656952247191011, + "grad_norm": 0.9974715709686279, + "learning_rate": 5.997960147930699e-07, + "loss": 0.5817, + "step": 13263 + }, + { + "epoch": 4.657303370786517, + "grad_norm": 0.9937880039215088, + "learning_rate": 5.994956155472911e-07, + "loss": 0.544, + "step": 13264 + }, + { + "epoch": 4.657654494382022, + "grad_norm": 0.9514478445053101, + "learning_rate": 5.991952812971352e-07, + "loss": 0.5086, + "step": 13265 + }, + { + "epoch": 4.658005617977528, + "grad_norm": 0.9837133884429932, + "learning_rate": 5.988950120528747e-07, + "loss": 0.5644, + "step": 13266 + }, + { + "epoch": 4.658356741573034, + "grad_norm": 0.9694938659667969, + "learning_rate": 5.985948078247775e-07, + "loss": 0.5404, + "step": 13267 + }, + { + "epoch": 4.658707865168539, + "grad_norm": 0.978111743927002, + "learning_rate": 5.982946686231114e-07, + "loss": 0.5488, + "step": 13268 + }, + { + "epoch": 4.659058988764045, + "grad_norm": 0.974652111530304, + "learning_rate": 5.979945944581403e-07, + "loss": 0.5615, + "step": 13269 + }, + { + "epoch": 4.65941011235955, + "grad_norm": 0.9811543822288513, + "learning_rate": 5.976945853401258e-07, + "loss": 0.5485, + "step": 13270 + }, + { + "epoch": 4.659761235955056, + "grad_norm": 0.9551305174827576, + "learning_rate": 5.973946412793294e-07, + "loss": 0.5222, + "step": 13271 + }, + { + "epoch": 4.660112359550562, + "grad_norm": 0.9718194007873535, + "learning_rate": 5.970947622860077e-07, + "loss": 0.5119, + "step": 13272 + }, + { + "epoch": 4.660463483146067, + "grad_norm": 0.9753769636154175, + "learning_rate": 5.967949483704172e-07, + "loss": 0.5571, + "step": 13273 + }, + { + "epoch": 4.660814606741573, + "grad_norm": 0.976030170917511, + "learning_rate": 5.9649519954281e-07, + "loss": 0.5548, + "step": 13274 + }, + { + "epoch": 4.661165730337078, + "grad_norm": 0.9774072766304016, + "learning_rate": 5.961955158134391e-07, + "loss": 0.5307, + "step": 13275 + }, + { + "epoch": 4.661516853932584, + "grad_norm": 1.004214882850647, + "learning_rate": 5.958958971925524e-07, + "loss": 0.5536, + "step": 13276 + }, + { + "epoch": 4.66186797752809, + "grad_norm": 0.9494681358337402, + "learning_rate": 5.95596343690396e-07, + "loss": 0.5241, + "step": 13277 + }, + { + "epoch": 4.662219101123595, + "grad_norm": 0.9534063339233398, + "learning_rate": 5.952968553172159e-07, + "loss": 0.5298, + "step": 13278 + }, + { + "epoch": 4.662570224719101, + "grad_norm": 0.95174640417099, + "learning_rate": 5.949974320832525e-07, + "loss": 0.5275, + "step": 13279 + }, + { + "epoch": 4.662921348314606, + "grad_norm": 0.9781541228294373, + "learning_rate": 5.946980739987477e-07, + "loss": 0.57, + "step": 13280 + }, + { + "epoch": 4.663272471910112, + "grad_norm": 1.0605838298797607, + "learning_rate": 5.943987810739374e-07, + "loss": 0.5255, + "step": 13281 + }, + { + "epoch": 4.663623595505618, + "grad_norm": 0.9430391192436218, + "learning_rate": 5.940995533190588e-07, + "loss": 0.5181, + "step": 13282 + }, + { + "epoch": 4.663974719101123, + "grad_norm": 0.941925585269928, + "learning_rate": 5.938003907443449e-07, + "loss": 0.5186, + "step": 13283 + }, + { + "epoch": 4.664325842696629, + "grad_norm": 0.9793627858161926, + "learning_rate": 5.935012933600257e-07, + "loss": 0.5287, + "step": 13284 + }, + { + "epoch": 4.664676966292134, + "grad_norm": 1.0032808780670166, + "learning_rate": 5.932022611763313e-07, + "loss": 0.55, + "step": 13285 + }, + { + "epoch": 4.66502808988764, + "grad_norm": 0.9920636415481567, + "learning_rate": 5.929032942034873e-07, + "loss": 0.5565, + "step": 13286 + }, + { + "epoch": 4.665379213483146, + "grad_norm": 1.000809907913208, + "learning_rate": 5.926043924517197e-07, + "loss": 0.5555, + "step": 13287 + }, + { + "epoch": 4.665730337078652, + "grad_norm": 0.9707759022712708, + "learning_rate": 5.923055559312487e-07, + "loss": 0.5389, + "step": 13288 + }, + { + "epoch": 4.666081460674158, + "grad_norm": 1.0142873525619507, + "learning_rate": 5.920067846522959e-07, + "loss": 0.5793, + "step": 13289 + }, + { + "epoch": 4.666432584269663, + "grad_norm": 0.9498804211616516, + "learning_rate": 5.917080786250779e-07, + "loss": 0.5135, + "step": 13290 + }, + { + "epoch": 4.666783707865169, + "grad_norm": 0.9809249639511108, + "learning_rate": 5.914094378598112e-07, + "loss": 0.5501, + "step": 13291 + }, + { + "epoch": 4.667134831460674, + "grad_norm": 0.9796808958053589, + "learning_rate": 5.91110862366708e-07, + "loss": 0.5383, + "step": 13292 + }, + { + "epoch": 4.66748595505618, + "grad_norm": 0.9828783273696899, + "learning_rate": 5.908123521559803e-07, + "loss": 0.5438, + "step": 13293 + }, + { + "epoch": 4.667837078651686, + "grad_norm": 0.9435595273971558, + "learning_rate": 5.90513907237836e-07, + "loss": 0.5209, + "step": 13294 + }, + { + "epoch": 4.668188202247191, + "grad_norm": 0.9471200108528137, + "learning_rate": 5.902155276224828e-07, + "loss": 0.5296, + "step": 13295 + }, + { + "epoch": 4.668539325842697, + "grad_norm": 1.3014353513717651, + "learning_rate": 5.89917213320124e-07, + "loss": 0.5474, + "step": 13296 + }, + { + "epoch": 4.668890449438202, + "grad_norm": 0.9863311648368835, + "learning_rate": 5.896189643409617e-07, + "loss": 0.5657, + "step": 13297 + }, + { + "epoch": 4.669241573033708, + "grad_norm": 0.9620158672332764, + "learning_rate": 5.893207806951965e-07, + "loss": 0.5311, + "step": 13298 + }, + { + "epoch": 4.669592696629214, + "grad_norm": 0.9721419215202332, + "learning_rate": 5.890226623930251e-07, + "loss": 0.5346, + "step": 13299 + }, + { + "epoch": 4.669943820224719, + "grad_norm": 0.9711320996284485, + "learning_rate": 5.887246094446439e-07, + "loss": 0.5348, + "step": 13300 + }, + { + "epoch": 4.670294943820225, + "grad_norm": 0.9560356736183167, + "learning_rate": 5.884266218602449e-07, + "loss": 0.5002, + "step": 13301 + }, + { + "epoch": 4.67064606741573, + "grad_norm": 0.9576439261436462, + "learning_rate": 5.881286996500204e-07, + "loss": 0.5148, + "step": 13302 + }, + { + "epoch": 4.670997191011236, + "grad_norm": 0.9476045966148376, + "learning_rate": 5.878308428241577e-07, + "loss": 0.5437, + "step": 13303 + }, + { + "epoch": 4.671348314606742, + "grad_norm": 0.9543407559394836, + "learning_rate": 5.875330513928448e-07, + "loss": 0.5635, + "step": 13304 + }, + { + "epoch": 4.671699438202247, + "grad_norm": 0.9455150365829468, + "learning_rate": 5.872353253662641e-07, + "loss": 0.5694, + "step": 13305 + }, + { + "epoch": 4.672050561797753, + "grad_norm": 0.9716838002204895, + "learning_rate": 5.869376647545993e-07, + "loss": 0.5401, + "step": 13306 + }, + { + "epoch": 4.672401685393258, + "grad_norm": 1.028611183166504, + "learning_rate": 5.866400695680286e-07, + "loss": 0.5143, + "step": 13307 + }, + { + "epoch": 4.672752808988764, + "grad_norm": 0.9609227776527405, + "learning_rate": 5.863425398167308e-07, + "loss": 0.522, + "step": 13308 + }, + { + "epoch": 4.67310393258427, + "grad_norm": 0.9794772863388062, + "learning_rate": 5.860450755108807e-07, + "loss": 0.524, + "step": 13309 + }, + { + "epoch": 4.673455056179775, + "grad_norm": 0.9549742341041565, + "learning_rate": 5.857476766606504e-07, + "loss": 0.5327, + "step": 13310 + }, + { + "epoch": 4.673806179775281, + "grad_norm": 0.9930747151374817, + "learning_rate": 5.854503432762124e-07, + "loss": 0.5085, + "step": 13311 + }, + { + "epoch": 4.674157303370786, + "grad_norm": 1.005184292793274, + "learning_rate": 5.851530753677337e-07, + "loss": 0.5206, + "step": 13312 + }, + { + "epoch": 4.674508426966292, + "grad_norm": 0.9432843923568726, + "learning_rate": 5.848558729453819e-07, + "loss": 0.524, + "step": 13313 + }, + { + "epoch": 4.674859550561798, + "grad_norm": 0.9463970065116882, + "learning_rate": 5.845587360193198e-07, + "loss": 0.5584, + "step": 13314 + }, + { + "epoch": 4.675210674157303, + "grad_norm": 0.9811241030693054, + "learning_rate": 5.842616645997104e-07, + "loss": 0.5168, + "step": 13315 + }, + { + "epoch": 4.675561797752809, + "grad_norm": 0.97864830493927, + "learning_rate": 5.839646586967121e-07, + "loss": 0.5364, + "step": 13316 + }, + { + "epoch": 4.675912921348314, + "grad_norm": 0.9524725079536438, + "learning_rate": 5.836677183204834e-07, + "loss": 0.564, + "step": 13317 + }, + { + "epoch": 4.67626404494382, + "grad_norm": 0.9429967403411865, + "learning_rate": 5.833708434811786e-07, + "loss": 0.5098, + "step": 13318 + }, + { + "epoch": 4.676615168539326, + "grad_norm": 0.9452599883079529, + "learning_rate": 5.830740341889513e-07, + "loss": 0.5648, + "step": 13319 + }, + { + "epoch": 4.676966292134831, + "grad_norm": 0.9341443181037903, + "learning_rate": 5.827772904539517e-07, + "loss": 0.4935, + "step": 13320 + }, + { + "epoch": 4.677317415730337, + "grad_norm": 0.9825121164321899, + "learning_rate": 5.824806122863274e-07, + "loss": 0.5554, + "step": 13321 + }, + { + "epoch": 4.677668539325842, + "grad_norm": 0.950174868106842, + "learning_rate": 5.821839996962265e-07, + "loss": 0.5554, + "step": 13322 + }, + { + "epoch": 4.678019662921348, + "grad_norm": 0.9251747727394104, + "learning_rate": 5.818874526937904e-07, + "loss": 0.5145, + "step": 13323 + }, + { + "epoch": 4.678370786516854, + "grad_norm": 0.9388288259506226, + "learning_rate": 5.815909712891632e-07, + "loss": 0.535, + "step": 13324 + }, + { + "epoch": 4.67872191011236, + "grad_norm": 0.9996606111526489, + "learning_rate": 5.812945554924823e-07, + "loss": 0.5386, + "step": 13325 + }, + { + "epoch": 4.679073033707866, + "grad_norm": 0.9753063321113586, + "learning_rate": 5.809982053138862e-07, + "loss": 0.5393, + "step": 13326 + }, + { + "epoch": 4.679424157303371, + "grad_norm": 0.9813492298126221, + "learning_rate": 5.807019207635089e-07, + "loss": 0.5562, + "step": 13327 + }, + { + "epoch": 4.679775280898877, + "grad_norm": 0.9517565369606018, + "learning_rate": 5.804057018514844e-07, + "loss": 0.536, + "step": 13328 + }, + { + "epoch": 4.680126404494382, + "grad_norm": 0.9672154784202576, + "learning_rate": 5.801095485879415e-07, + "loss": 0.5426, + "step": 13329 + }, + { + "epoch": 4.680477528089888, + "grad_norm": 0.9989373087882996, + "learning_rate": 5.798134609830097e-07, + "loss": 0.5799, + "step": 13330 + }, + { + "epoch": 4.680828651685394, + "grad_norm": 0.9921289682388306, + "learning_rate": 5.795174390468139e-07, + "loss": 0.5535, + "step": 13331 + }, + { + "epoch": 4.681179775280899, + "grad_norm": 0.9826899766921997, + "learning_rate": 5.792214827894787e-07, + "loss": 0.5386, + "step": 13332 + }, + { + "epoch": 4.681530898876405, + "grad_norm": 0.9516276717185974, + "learning_rate": 5.789255922211256e-07, + "loss": 0.5625, + "step": 13333 + }, + { + "epoch": 4.68188202247191, + "grad_norm": 0.9850350618362427, + "learning_rate": 5.786297673518724e-07, + "loss": 0.5684, + "step": 13334 + }, + { + "epoch": 4.682233146067416, + "grad_norm": 0.9531389474868774, + "learning_rate": 5.783340081918376e-07, + "loss": 0.5151, + "step": 13335 + }, + { + "epoch": 4.682584269662922, + "grad_norm": 0.9853708148002625, + "learning_rate": 5.780383147511348e-07, + "loss": 0.5631, + "step": 13336 + }, + { + "epoch": 4.682935393258427, + "grad_norm": 0.9790744781494141, + "learning_rate": 5.777426870398778e-07, + "loss": 0.5154, + "step": 13337 + }, + { + "epoch": 4.683286516853933, + "grad_norm": 0.9989447593688965, + "learning_rate": 5.774471250681752e-07, + "loss": 0.5329, + "step": 13338 + }, + { + "epoch": 4.683637640449438, + "grad_norm": 0.9633991718292236, + "learning_rate": 5.771516288461362e-07, + "loss": 0.5162, + "step": 13339 + }, + { + "epoch": 4.683988764044944, + "grad_norm": 0.9799330234527588, + "learning_rate": 5.768561983838658e-07, + "loss": 0.538, + "step": 13340 + }, + { + "epoch": 4.68433988764045, + "grad_norm": 0.9949676990509033, + "learning_rate": 5.765608336914683e-07, + "loss": 0.5224, + "step": 13341 + }, + { + "epoch": 4.684691011235955, + "grad_norm": 0.9907997250556946, + "learning_rate": 5.762655347790435e-07, + "loss": 0.5533, + "step": 13342 + }, + { + "epoch": 4.685042134831461, + "grad_norm": 0.9588221907615662, + "learning_rate": 5.759703016566923e-07, + "loss": 0.5486, + "step": 13343 + }, + { + "epoch": 4.685393258426966, + "grad_norm": 0.9830360412597656, + "learning_rate": 5.756751343345094e-07, + "loss": 0.5183, + "step": 13344 + }, + { + "epoch": 4.685744382022472, + "grad_norm": 1.018154263496399, + "learning_rate": 5.753800328225909e-07, + "loss": 0.5577, + "step": 13345 + }, + { + "epoch": 4.686095505617978, + "grad_norm": 0.9674930572509766, + "learning_rate": 5.750849971310285e-07, + "loss": 0.5198, + "step": 13346 + }, + { + "epoch": 4.686446629213483, + "grad_norm": 0.9555802941322327, + "learning_rate": 5.747900272699114e-07, + "loss": 0.5069, + "step": 13347 + }, + { + "epoch": 4.686797752808989, + "grad_norm": 0.9526321887969971, + "learning_rate": 5.744951232493284e-07, + "loss": 0.5285, + "step": 13348 + }, + { + "epoch": 4.687148876404494, + "grad_norm": 0.9421164989471436, + "learning_rate": 5.74200285079364e-07, + "loss": 0.5314, + "step": 13349 + }, + { + "epoch": 4.6875, + "grad_norm": 0.9699839353561401, + "learning_rate": 5.739055127701027e-07, + "loss": 0.5569, + "step": 13350 + }, + { + "epoch": 4.687851123595506, + "grad_norm": 0.9670280814170837, + "learning_rate": 5.736108063316237e-07, + "loss": 0.5308, + "step": 13351 + }, + { + "epoch": 4.688202247191011, + "grad_norm": 0.9437997937202454, + "learning_rate": 5.733161657740074e-07, + "loss": 0.4986, + "step": 13352 + }, + { + "epoch": 4.688553370786517, + "grad_norm": 0.9443026781082153, + "learning_rate": 5.730215911073292e-07, + "loss": 0.5711, + "step": 13353 + }, + { + "epoch": 4.688904494382022, + "grad_norm": 0.9703240394592285, + "learning_rate": 5.727270823416639e-07, + "loss": 0.5329, + "step": 13354 + }, + { + "epoch": 4.689255617977528, + "grad_norm": 0.9744132161140442, + "learning_rate": 5.724326394870827e-07, + "loss": 0.5379, + "step": 13355 + }, + { + "epoch": 4.689606741573034, + "grad_norm": 0.9696474671363831, + "learning_rate": 5.721382625536567e-07, + "loss": 0.5662, + "step": 13356 + }, + { + "epoch": 4.689957865168539, + "grad_norm": 0.936695396900177, + "learning_rate": 5.718439515514515e-07, + "loss": 0.5453, + "step": 13357 + }, + { + "epoch": 4.690308988764045, + "grad_norm": 0.9276344180107117, + "learning_rate": 5.71549706490534e-07, + "loss": 0.4874, + "step": 13358 + }, + { + "epoch": 4.69066011235955, + "grad_norm": 0.9882335066795349, + "learning_rate": 5.712555273809662e-07, + "loss": 0.5232, + "step": 13359 + }, + { + "epoch": 4.691011235955056, + "grad_norm": 0.9965115785598755, + "learning_rate": 5.709614142328082e-07, + "loss": 0.542, + "step": 13360 + }, + { + "epoch": 4.691362359550562, + "grad_norm": 0.9849119186401367, + "learning_rate": 5.7066736705612e-07, + "loss": 0.5753, + "step": 13361 + }, + { + "epoch": 4.691713483146067, + "grad_norm": 0.9677380919456482, + "learning_rate": 5.703733858609561e-07, + "loss": 0.5394, + "step": 13362 + }, + { + "epoch": 4.692064606741573, + "grad_norm": 0.9820301532745361, + "learning_rate": 5.700794706573717e-07, + "loss": 0.5567, + "step": 13363 + }, + { + "epoch": 4.692415730337078, + "grad_norm": 0.9493071436882019, + "learning_rate": 5.697856214554176e-07, + "loss": 0.5164, + "step": 13364 + }, + { + "epoch": 4.692766853932584, + "grad_norm": 0.9814715385437012, + "learning_rate": 5.69491838265144e-07, + "loss": 0.5571, + "step": 13365 + }, + { + "epoch": 4.69311797752809, + "grad_norm": 0.9772918224334717, + "learning_rate": 5.691981210965972e-07, + "loss": 0.524, + "step": 13366 + }, + { + "epoch": 4.693469101123595, + "grad_norm": 0.9611116647720337, + "learning_rate": 5.689044699598229e-07, + "loss": 0.5404, + "step": 13367 + }, + { + "epoch": 4.693820224719101, + "grad_norm": 0.9493620991706848, + "learning_rate": 5.686108848648625e-07, + "loss": 0.5458, + "step": 13368 + }, + { + "epoch": 4.694171348314606, + "grad_norm": 0.9723794460296631, + "learning_rate": 5.683173658217578e-07, + "loss": 0.5513, + "step": 13369 + }, + { + "epoch": 4.694522471910112, + "grad_norm": 0.9831329584121704, + "learning_rate": 5.680239128405457e-07, + "loss": 0.5628, + "step": 13370 + }, + { + "epoch": 4.694873595505618, + "grad_norm": 1.0300928354263306, + "learning_rate": 5.677305259312632e-07, + "loss": 0.5162, + "step": 13371 + }, + { + "epoch": 4.695224719101123, + "grad_norm": 0.9676100015640259, + "learning_rate": 5.674372051039431e-07, + "loss": 0.5503, + "step": 13372 + }, + { + "epoch": 4.695575842696629, + "grad_norm": 0.9751589298248291, + "learning_rate": 5.671439503686163e-07, + "loss": 0.5632, + "step": 13373 + }, + { + "epoch": 4.695926966292134, + "grad_norm": 0.9878960251808167, + "learning_rate": 5.668507617353134e-07, + "loss": 0.564, + "step": 13374 + }, + { + "epoch": 4.69627808988764, + "grad_norm": 0.9981149435043335, + "learning_rate": 5.66557639214059e-07, + "loss": 0.5588, + "step": 13375 + }, + { + "epoch": 4.696629213483146, + "grad_norm": 0.96822589635849, + "learning_rate": 5.662645828148799e-07, + "loss": 0.5213, + "step": 13376 + }, + { + "epoch": 4.696980337078652, + "grad_norm": 1.0003340244293213, + "learning_rate": 5.659715925477966e-07, + "loss": 0.5559, + "step": 13377 + }, + { + "epoch": 4.697331460674158, + "grad_norm": 0.9476814866065979, + "learning_rate": 5.656786684228304e-07, + "loss": 0.5382, + "step": 13378 + }, + { + "epoch": 4.697682584269663, + "grad_norm": 0.9905784130096436, + "learning_rate": 5.653858104499982e-07, + "loss": 0.5284, + "step": 13379 + }, + { + "epoch": 4.698033707865169, + "grad_norm": 0.9925965666770935, + "learning_rate": 5.650930186393161e-07, + "loss": 0.5198, + "step": 13380 + }, + { + "epoch": 4.698384831460674, + "grad_norm": 0.9903517365455627, + "learning_rate": 5.648002930007965e-07, + "loss": 0.5674, + "step": 13381 + }, + { + "epoch": 4.69873595505618, + "grad_norm": 0.947591245174408, + "learning_rate": 5.645076335444516e-07, + "loss": 0.5073, + "step": 13382 + }, + { + "epoch": 4.699087078651686, + "grad_norm": 0.9836434721946716, + "learning_rate": 5.642150402802895e-07, + "loss": 0.5448, + "step": 13383 + }, + { + "epoch": 4.699438202247191, + "grad_norm": 0.9912024140357971, + "learning_rate": 5.639225132183157e-07, + "loss": 0.5379, + "step": 13384 + }, + { + "epoch": 4.699789325842697, + "grad_norm": 0.9760569334030151, + "learning_rate": 5.636300523685361e-07, + "loss": 0.5086, + "step": 13385 + }, + { + "epoch": 4.700140449438202, + "grad_norm": 0.9525435566902161, + "learning_rate": 5.633376577409507e-07, + "loss": 0.5446, + "step": 13386 + }, + { + "epoch": 4.700491573033708, + "grad_norm": 0.9466583728790283, + "learning_rate": 5.630453293455612e-07, + "loss": 0.5176, + "step": 13387 + }, + { + "epoch": 4.700842696629214, + "grad_norm": 1.0678857564926147, + "learning_rate": 5.627530671923631e-07, + "loss": 0.5397, + "step": 13388 + }, + { + "epoch": 4.701193820224719, + "grad_norm": 0.9643000960350037, + "learning_rate": 5.624608712913531e-07, + "loss": 0.5257, + "step": 13389 + }, + { + "epoch": 4.701544943820225, + "grad_norm": 0.933380663394928, + "learning_rate": 5.621687416525228e-07, + "loss": 0.5227, + "step": 13390 + }, + { + "epoch": 4.70189606741573, + "grad_norm": 0.9746999740600586, + "learning_rate": 5.61876678285864e-07, + "loss": 0.572, + "step": 13391 + }, + { + "epoch": 4.702247191011236, + "grad_norm": 0.9840980172157288, + "learning_rate": 5.615846812013634e-07, + "loss": 0.5381, + "step": 13392 + }, + { + "epoch": 4.702598314606742, + "grad_norm": 0.9680678248405457, + "learning_rate": 5.612927504090088e-07, + "loss": 0.5511, + "step": 13393 + }, + { + "epoch": 4.702949438202247, + "grad_norm": 0.9724068641662598, + "learning_rate": 5.610008859187824e-07, + "loss": 0.5572, + "step": 13394 + }, + { + "epoch": 4.703300561797753, + "grad_norm": 0.9613707065582275, + "learning_rate": 5.607090877406673e-07, + "loss": 0.5391, + "step": 13395 + }, + { + "epoch": 4.703651685393258, + "grad_norm": 0.9496790170669556, + "learning_rate": 5.604173558846418e-07, + "loss": 0.5085, + "step": 13396 + }, + { + "epoch": 4.704002808988764, + "grad_norm": 1.0796078443527222, + "learning_rate": 5.601256903606823e-07, + "loss": 0.5697, + "step": 13397 + }, + { + "epoch": 4.70435393258427, + "grad_norm": 0.9817330837249756, + "learning_rate": 5.598340911787651e-07, + "loss": 0.5382, + "step": 13398 + }, + { + "epoch": 4.704705056179775, + "grad_norm": 0.9692642092704773, + "learning_rate": 5.595425583488609e-07, + "loss": 0.5216, + "step": 13399 + }, + { + "epoch": 4.705056179775281, + "grad_norm": 0.9634451866149902, + "learning_rate": 5.592510918809413e-07, + "loss": 0.537, + "step": 13400 + }, + { + "epoch": 4.705407303370786, + "grad_norm": 0.9451289176940918, + "learning_rate": 5.589596917849732e-07, + "loss": 0.5674, + "step": 13401 + }, + { + "epoch": 4.705758426966292, + "grad_norm": 1.0215274095535278, + "learning_rate": 5.586683580709232e-07, + "loss": 0.5951, + "step": 13402 + }, + { + "epoch": 4.706109550561798, + "grad_norm": 0.9778029918670654, + "learning_rate": 5.583770907487534e-07, + "loss": 0.5242, + "step": 13403 + }, + { + "epoch": 4.706460674157303, + "grad_norm": 1.0001003742218018, + "learning_rate": 5.580858898284263e-07, + "loss": 0.5601, + "step": 13404 + }, + { + "epoch": 4.706811797752809, + "grad_norm": 0.994561493396759, + "learning_rate": 5.577947553198995e-07, + "loss": 0.5443, + "step": 13405 + }, + { + "epoch": 4.707162921348314, + "grad_norm": 0.9760117530822754, + "learning_rate": 5.57503687233131e-07, + "loss": 0.5451, + "step": 13406 + }, + { + "epoch": 4.70751404494382, + "grad_norm": 0.9498127102851868, + "learning_rate": 5.572126855780732e-07, + "loss": 0.5198, + "step": 13407 + }, + { + "epoch": 4.707865168539326, + "grad_norm": 1.033523678779602, + "learning_rate": 5.569217503646798e-07, + "loss": 0.5727, + "step": 13408 + }, + { + "epoch": 4.708216292134831, + "grad_norm": 0.9905663132667542, + "learning_rate": 5.566308816028998e-07, + "loss": 0.5649, + "step": 13409 + }, + { + "epoch": 4.708567415730337, + "grad_norm": 0.9440320730209351, + "learning_rate": 5.563400793026799e-07, + "loss": 0.4908, + "step": 13410 + }, + { + "epoch": 4.708918539325842, + "grad_norm": 0.9811357855796814, + "learning_rate": 5.560493434739669e-07, + "loss": 0.5473, + "step": 13411 + }, + { + "epoch": 4.709269662921348, + "grad_norm": 0.9686270356178284, + "learning_rate": 5.557586741267024e-07, + "loss": 0.5458, + "step": 13412 + }, + { + "epoch": 4.709620786516854, + "grad_norm": 0.9791997075080872, + "learning_rate": 5.55468071270828e-07, + "loss": 0.5549, + "step": 13413 + }, + { + "epoch": 4.70997191011236, + "grad_norm": 0.9968413710594177, + "learning_rate": 5.551775349162808e-07, + "loss": 0.5594, + "step": 13414 + }, + { + "epoch": 4.710323033707866, + "grad_norm": 0.9756453037261963, + "learning_rate": 5.548870650729987e-07, + "loss": 0.5365, + "step": 13415 + }, + { + "epoch": 4.710674157303371, + "grad_norm": 1.015196442604065, + "learning_rate": 5.545966617509136e-07, + "loss": 0.5304, + "step": 13416 + }, + { + "epoch": 4.711025280898877, + "grad_norm": 0.9899147152900696, + "learning_rate": 5.543063249599586e-07, + "loss": 0.5319, + "step": 13417 + }, + { + "epoch": 4.711376404494382, + "grad_norm": 0.9654368162155151, + "learning_rate": 5.540160547100618e-07, + "loss": 0.5725, + "step": 13418 + }, + { + "epoch": 4.711727528089888, + "grad_norm": 0.9794710278511047, + "learning_rate": 5.537258510111512e-07, + "loss": 0.5283, + "step": 13419 + }, + { + "epoch": 4.712078651685394, + "grad_norm": 0.9658960103988647, + "learning_rate": 5.534357138731513e-07, + "loss": 0.5204, + "step": 13420 + }, + { + "epoch": 4.712429775280899, + "grad_norm": 0.990121603012085, + "learning_rate": 5.531456433059837e-07, + "loss": 0.5544, + "step": 13421 + }, + { + "epoch": 4.712780898876405, + "grad_norm": 0.9756287932395935, + "learning_rate": 5.528556393195694e-07, + "loss": 0.5591, + "step": 13422 + }, + { + "epoch": 4.71313202247191, + "grad_norm": 0.9581506252288818, + "learning_rate": 5.525657019238256e-07, + "loss": 0.5757, + "step": 13423 + }, + { + "epoch": 4.713483146067416, + "grad_norm": 0.963983952999115, + "learning_rate": 5.522758311286692e-07, + "loss": 0.5575, + "step": 13424 + }, + { + "epoch": 4.713834269662922, + "grad_norm": 0.9443579912185669, + "learning_rate": 5.519860269440116e-07, + "loss": 0.5156, + "step": 13425 + }, + { + "epoch": 4.714185393258427, + "grad_norm": 1.0175230503082275, + "learning_rate": 5.516962893797662e-07, + "loss": 0.5154, + "step": 13426 + }, + { + "epoch": 4.714536516853933, + "grad_norm": 0.9861249327659607, + "learning_rate": 5.5140661844584e-07, + "loss": 0.5729, + "step": 13427 + }, + { + "epoch": 4.714887640449438, + "grad_norm": 0.9646193385124207, + "learning_rate": 5.511170141521396e-07, + "loss": 0.5177, + "step": 13428 + }, + { + "epoch": 4.715238764044944, + "grad_norm": 1.0314533710479736, + "learning_rate": 5.508274765085702e-07, + "loss": 0.5477, + "step": 13429 + }, + { + "epoch": 4.71558988764045, + "grad_norm": 0.9823387861251831, + "learning_rate": 5.505380055250326e-07, + "loss": 0.5105, + "step": 13430 + }, + { + "epoch": 4.715941011235955, + "grad_norm": 0.9577609300613403, + "learning_rate": 5.502486012114275e-07, + "loss": 0.5312, + "step": 13431 + }, + { + "epoch": 4.716292134831461, + "grad_norm": 0.9489021301269531, + "learning_rate": 5.499592635776513e-07, + "loss": 0.5429, + "step": 13432 + }, + { + "epoch": 4.716643258426966, + "grad_norm": 0.98175048828125, + "learning_rate": 5.496699926336002e-07, + "loss": 0.548, + "step": 13433 + }, + { + "epoch": 4.716994382022472, + "grad_norm": 0.9991878271102905, + "learning_rate": 5.493807883891659e-07, + "loss": 0.5762, + "step": 13434 + }, + { + "epoch": 4.717345505617978, + "grad_norm": 0.9637579917907715, + "learning_rate": 5.490916508542399e-07, + "loss": 0.5268, + "step": 13435 + }, + { + "epoch": 4.717696629213483, + "grad_norm": 0.9710743427276611, + "learning_rate": 5.4880258003871e-07, + "loss": 0.5303, + "step": 13436 + }, + { + "epoch": 4.718047752808989, + "grad_norm": 0.956962525844574, + "learning_rate": 5.485135759524618e-07, + "loss": 0.4944, + "step": 13437 + }, + { + "epoch": 4.718398876404494, + "grad_norm": 0.9429409503936768, + "learning_rate": 5.4822463860538e-07, + "loss": 0.5277, + "step": 13438 + }, + { + "epoch": 4.71875, + "grad_norm": 0.9843346476554871, + "learning_rate": 5.479357680073446e-07, + "loss": 0.5503, + "step": 13439 + }, + { + "epoch": 4.719101123595506, + "grad_norm": 1.0026144981384277, + "learning_rate": 5.476469641682364e-07, + "loss": 0.5553, + "step": 13440 + }, + { + "epoch": 4.719452247191011, + "grad_norm": 0.9426062107086182, + "learning_rate": 5.473582270979305e-07, + "loss": 0.4934, + "step": 13441 + }, + { + "epoch": 4.719803370786517, + "grad_norm": 0.9569369554519653, + "learning_rate": 5.470695568063031e-07, + "loss": 0.5328, + "step": 13442 + }, + { + "epoch": 4.720154494382022, + "grad_norm": 0.9447029232978821, + "learning_rate": 5.467809533032259e-07, + "loss": 0.5424, + "step": 13443 + }, + { + "epoch": 4.720505617977528, + "grad_norm": 0.9738330841064453, + "learning_rate": 5.464924165985677e-07, + "loss": 0.563, + "step": 13444 + }, + { + "epoch": 4.720856741573034, + "grad_norm": 0.9861757755279541, + "learning_rate": 5.462039467021981e-07, + "loss": 0.5583, + "step": 13445 + }, + { + "epoch": 4.721207865168539, + "grad_norm": 0.9591412544250488, + "learning_rate": 5.459155436239808e-07, + "loss": 0.5193, + "step": 13446 + }, + { + "epoch": 4.721558988764045, + "grad_norm": 0.9329413175582886, + "learning_rate": 5.456272073737806e-07, + "loss": 0.5269, + "step": 13447 + }, + { + "epoch": 4.72191011235955, + "grad_norm": 0.9648354649543762, + "learning_rate": 5.453389379614568e-07, + "loss": 0.5669, + "step": 13448 + }, + { + "epoch": 4.722261235955056, + "grad_norm": 0.9762459397315979, + "learning_rate": 5.450507353968692e-07, + "loss": 0.5074, + "step": 13449 + }, + { + "epoch": 4.722612359550562, + "grad_norm": 0.9740614295005798, + "learning_rate": 5.447625996898739e-07, + "loss": 0.5541, + "step": 13450 + }, + { + "epoch": 4.722963483146067, + "grad_norm": 0.9706109166145325, + "learning_rate": 5.444745308503235e-07, + "loss": 0.5495, + "step": 13451 + }, + { + "epoch": 4.723314606741573, + "grad_norm": 1.0187245607376099, + "learning_rate": 5.441865288880719e-07, + "loss": 0.5473, + "step": 13452 + }, + { + "epoch": 4.723665730337078, + "grad_norm": 0.9874309301376343, + "learning_rate": 5.438985938129665e-07, + "loss": 0.5398, + "step": 13453 + }, + { + "epoch": 4.724016853932584, + "grad_norm": 0.9665016531944275, + "learning_rate": 5.436107256348558e-07, + "loss": 0.5592, + "step": 13454 + }, + { + "epoch": 4.72436797752809, + "grad_norm": 0.9569320678710938, + "learning_rate": 5.433229243635838e-07, + "loss": 0.5073, + "step": 13455 + }, + { + "epoch": 4.724719101123595, + "grad_norm": 0.9830293655395508, + "learning_rate": 5.430351900089939e-07, + "loss": 0.5475, + "step": 13456 + }, + { + "epoch": 4.725070224719101, + "grad_norm": 0.9916966557502747, + "learning_rate": 5.427475225809253e-07, + "loss": 0.5464, + "step": 13457 + }, + { + "epoch": 4.725421348314606, + "grad_norm": 0.9377747178077698, + "learning_rate": 5.424599220892174e-07, + "loss": 0.5236, + "step": 13458 + }, + { + "epoch": 4.725772471910112, + "grad_norm": 0.9844422340393066, + "learning_rate": 5.421723885437049e-07, + "loss": 0.5466, + "step": 13459 + }, + { + "epoch": 4.726123595505618, + "grad_norm": 0.9431427121162415, + "learning_rate": 5.41884921954221e-07, + "loss": 0.5314, + "step": 13460 + }, + { + "epoch": 4.726474719101123, + "grad_norm": 0.9818122386932373, + "learning_rate": 5.415975223305975e-07, + "loss": 0.5327, + "step": 13461 + }, + { + "epoch": 4.726825842696629, + "grad_norm": 0.9758610129356384, + "learning_rate": 5.413101896826625e-07, + "loss": 0.5279, + "step": 13462 + }, + { + "epoch": 4.727176966292134, + "grad_norm": 0.9411132335662842, + "learning_rate": 5.410229240202433e-07, + "loss": 0.5384, + "step": 13463 + }, + { + "epoch": 4.72752808988764, + "grad_norm": 0.9751479029655457, + "learning_rate": 5.407357253531634e-07, + "loss": 0.5211, + "step": 13464 + }, + { + "epoch": 4.727879213483146, + "grad_norm": 0.9529706835746765, + "learning_rate": 5.404485936912457e-07, + "loss": 0.5289, + "step": 13465 + }, + { + "epoch": 4.728230337078652, + "grad_norm": 0.9648151993751526, + "learning_rate": 5.401615290443085e-07, + "loss": 0.5557, + "step": 13466 + }, + { + "epoch": 4.728581460674158, + "grad_norm": 0.9799836277961731, + "learning_rate": 5.398745314221707e-07, + "loss": 0.589, + "step": 13467 + }, + { + "epoch": 4.728932584269663, + "grad_norm": 0.9352636933326721, + "learning_rate": 5.395876008346459e-07, + "loss": 0.4797, + "step": 13468 + }, + { + "epoch": 4.729283707865169, + "grad_norm": 0.9741494655609131, + "learning_rate": 5.393007372915485e-07, + "loss": 0.5819, + "step": 13469 + }, + { + "epoch": 4.729634831460674, + "grad_norm": 0.9616385698318481, + "learning_rate": 5.390139408026873e-07, + "loss": 0.572, + "step": 13470 + }, + { + "epoch": 4.72998595505618, + "grad_norm": 0.9389507174491882, + "learning_rate": 5.387272113778718e-07, + "loss": 0.5451, + "step": 13471 + }, + { + "epoch": 4.730337078651686, + "grad_norm": 0.9761593341827393, + "learning_rate": 5.384405490269076e-07, + "loss": 0.5425, + "step": 13472 + }, + { + "epoch": 4.730688202247191, + "grad_norm": 0.9516288638114929, + "learning_rate": 5.381539537595973e-07, + "loss": 0.5315, + "step": 13473 + }, + { + "epoch": 4.731039325842697, + "grad_norm": 0.9814075827598572, + "learning_rate": 5.378674255857439e-07, + "loss": 0.5651, + "step": 13474 + }, + { + "epoch": 4.731390449438202, + "grad_norm": 0.9696162939071655, + "learning_rate": 5.375809645151445e-07, + "loss": 0.5339, + "step": 13475 + }, + { + "epoch": 4.731741573033708, + "grad_norm": 0.9847190380096436, + "learning_rate": 5.372945705575977e-07, + "loss": 0.5239, + "step": 13476 + }, + { + "epoch": 4.732092696629214, + "grad_norm": 0.9755910634994507, + "learning_rate": 5.370082437228963e-07, + "loss": 0.5246, + "step": 13477 + }, + { + "epoch": 4.732443820224719, + "grad_norm": 0.9666023850440979, + "learning_rate": 5.36721984020834e-07, + "loss": 0.5288, + "step": 13478 + }, + { + "epoch": 4.732794943820225, + "grad_norm": 0.9984900951385498, + "learning_rate": 5.364357914611989e-07, + "loss": 0.5567, + "step": 13479 + }, + { + "epoch": 4.73314606741573, + "grad_norm": 0.9592170119285583, + "learning_rate": 5.361496660537807e-07, + "loss": 0.5442, + "step": 13480 + }, + { + "epoch": 4.733497191011236, + "grad_norm": 0.9594096541404724, + "learning_rate": 5.358636078083623e-07, + "loss": 0.5598, + "step": 13481 + }, + { + "epoch": 4.733848314606742, + "grad_norm": 0.974115788936615, + "learning_rate": 5.355776167347288e-07, + "loss": 0.5638, + "step": 13482 + }, + { + "epoch": 4.734199438202247, + "grad_norm": 0.9801085591316223, + "learning_rate": 5.352916928426588e-07, + "loss": 0.5476, + "step": 13483 + }, + { + "epoch": 4.734550561797753, + "grad_norm": 0.9405761361122131, + "learning_rate": 5.350058361419325e-07, + "loss": 0.5051, + "step": 13484 + }, + { + "epoch": 4.734901685393258, + "grad_norm": 0.9626070857048035, + "learning_rate": 5.347200466423252e-07, + "loss": 0.4883, + "step": 13485 + }, + { + "epoch": 4.735252808988764, + "grad_norm": 0.9541569948196411, + "learning_rate": 5.344343243536099e-07, + "loss": 0.534, + "step": 13486 + }, + { + "epoch": 4.73560393258427, + "grad_norm": 1.0490716695785522, + "learning_rate": 5.341486692855594e-07, + "loss": 0.5186, + "step": 13487 + }, + { + "epoch": 4.735955056179775, + "grad_norm": 0.9914073944091797, + "learning_rate": 5.338630814479415e-07, + "loss": 0.5749, + "step": 13488 + }, + { + "epoch": 4.736306179775281, + "grad_norm": 0.9878984093666077, + "learning_rate": 5.335775608505242e-07, + "loss": 0.5581, + "step": 13489 + }, + { + "epoch": 4.736657303370786, + "grad_norm": 0.9819514155387878, + "learning_rate": 5.332921075030714e-07, + "loss": 0.5536, + "step": 13490 + }, + { + "epoch": 4.737008426966292, + "grad_norm": 0.9513787627220154, + "learning_rate": 5.330067214153458e-07, + "loss": 0.5342, + "step": 13491 + }, + { + "epoch": 4.737359550561798, + "grad_norm": 0.9780181050300598, + "learning_rate": 5.32721402597107e-07, + "loss": 0.5396, + "step": 13492 + }, + { + "epoch": 4.737710674157303, + "grad_norm": 0.9472973942756653, + "learning_rate": 5.324361510581131e-07, + "loss": 0.5331, + "step": 13493 + }, + { + "epoch": 4.738061797752809, + "grad_norm": 1.0822620391845703, + "learning_rate": 5.321509668081187e-07, + "loss": 0.5221, + "step": 13494 + }, + { + "epoch": 4.738412921348314, + "grad_norm": 0.9777968525886536, + "learning_rate": 5.318658498568779e-07, + "loss": 0.5447, + "step": 13495 + }, + { + "epoch": 4.73876404494382, + "grad_norm": 0.9943186640739441, + "learning_rate": 5.315808002141401e-07, + "loss": 0.5929, + "step": 13496 + }, + { + "epoch": 4.739115168539326, + "grad_norm": 0.9406372308731079, + "learning_rate": 5.312958178896554e-07, + "loss": 0.5161, + "step": 13497 + }, + { + "epoch": 4.739466292134831, + "grad_norm": 0.9879481196403503, + "learning_rate": 5.31010902893169e-07, + "loss": 0.5537, + "step": 13498 + }, + { + "epoch": 4.739817415730337, + "grad_norm": 0.9818180203437805, + "learning_rate": 5.30726055234424e-07, + "loss": 0.5412, + "step": 13499 + }, + { + "epoch": 4.740168539325842, + "grad_norm": 0.9690172672271729, + "learning_rate": 5.304412749231638e-07, + "loss": 0.5508, + "step": 13500 + }, + { + "epoch": 4.740519662921348, + "grad_norm": 1.0102070569992065, + "learning_rate": 5.301565619691256e-07, + "loss": 0.5532, + "step": 13501 + }, + { + "epoch": 4.740870786516854, + "grad_norm": 0.9727211594581604, + "learning_rate": 5.298719163820485e-07, + "loss": 0.4998, + "step": 13502 + }, + { + "epoch": 4.74122191011236, + "grad_norm": 0.9426267743110657, + "learning_rate": 5.295873381716649e-07, + "loss": 0.5488, + "step": 13503 + }, + { + "epoch": 4.741573033707866, + "grad_norm": 0.9724999666213989, + "learning_rate": 5.293028273477089e-07, + "loss": 0.5664, + "step": 13504 + }, + { + "epoch": 4.741924157303371, + "grad_norm": 0.9428138136863708, + "learning_rate": 5.290183839199092e-07, + "loss": 0.5169, + "step": 13505 + }, + { + "epoch": 4.742275280898877, + "grad_norm": 1.0203648805618286, + "learning_rate": 5.287340078979952e-07, + "loss": 0.548, + "step": 13506 + }, + { + "epoch": 4.742626404494382, + "grad_norm": 0.9662187695503235, + "learning_rate": 5.284496992916902e-07, + "loss": 0.5298, + "step": 13507 + }, + { + "epoch": 4.742977528089888, + "grad_norm": 1.0260173082351685, + "learning_rate": 5.281654581107193e-07, + "loss": 0.5709, + "step": 13508 + }, + { + "epoch": 4.743328651685394, + "grad_norm": 0.9403749108314514, + "learning_rate": 5.278812843648023e-07, + "loss": 0.5119, + "step": 13509 + }, + { + "epoch": 4.743679775280899, + "grad_norm": 0.9818673133850098, + "learning_rate": 5.275971780636574e-07, + "loss": 0.5412, + "step": 13510 + }, + { + "epoch": 4.744030898876405, + "grad_norm": 0.9684340953826904, + "learning_rate": 5.273131392170017e-07, + "loss": 0.5232, + "step": 13511 + }, + { + "epoch": 4.74438202247191, + "grad_norm": 0.9591844081878662, + "learning_rate": 5.270291678345477e-07, + "loss": 0.5114, + "step": 13512 + }, + { + "epoch": 4.744733146067416, + "grad_norm": 0.9592310190200806, + "learning_rate": 5.267452639260088e-07, + "loss": 0.5213, + "step": 13513 + }, + { + "epoch": 4.745084269662922, + "grad_norm": 0.9743878841400146, + "learning_rate": 5.264614275010926e-07, + "loss": 0.5443, + "step": 13514 + }, + { + "epoch": 4.745435393258427, + "grad_norm": 0.9500264525413513, + "learning_rate": 5.261776585695075e-07, + "loss": 0.5349, + "step": 13515 + }, + { + "epoch": 4.745786516853933, + "grad_norm": 0.9677380919456482, + "learning_rate": 5.258939571409566e-07, + "loss": 0.5682, + "step": 13516 + }, + { + "epoch": 4.746137640449438, + "grad_norm": 0.9707212448120117, + "learning_rate": 5.256103232251439e-07, + "loss": 0.5427, + "step": 13517 + }, + { + "epoch": 4.746488764044944, + "grad_norm": 0.9275147318840027, + "learning_rate": 5.25326756831768e-07, + "loss": 0.5492, + "step": 13518 + }, + { + "epoch": 4.74683988764045, + "grad_norm": 0.9941521883010864, + "learning_rate": 5.250432579705278e-07, + "loss": 0.5386, + "step": 13519 + }, + { + "epoch": 4.747191011235955, + "grad_norm": 0.9843141436576843, + "learning_rate": 5.247598266511178e-07, + "loss": 0.5385, + "step": 13520 + }, + { + "epoch": 4.747542134831461, + "grad_norm": 0.9643127918243408, + "learning_rate": 5.244764628832319e-07, + "loss": 0.5306, + "step": 13521 + }, + { + "epoch": 4.747893258426966, + "grad_norm": 0.9537690877914429, + "learning_rate": 5.241931666765607e-07, + "loss": 0.544, + "step": 13522 + }, + { + "epoch": 4.748244382022472, + "grad_norm": 0.9782516956329346, + "learning_rate": 5.239099380407916e-07, + "loss": 0.5666, + "step": 13523 + }, + { + "epoch": 4.748595505617978, + "grad_norm": 0.9707499146461487, + "learning_rate": 5.236267769856124e-07, + "loss": 0.5117, + "step": 13524 + }, + { + "epoch": 4.748946629213483, + "grad_norm": 0.9674602746963501, + "learning_rate": 5.233436835207057e-07, + "loss": 0.5272, + "step": 13525 + }, + { + "epoch": 4.749297752808989, + "grad_norm": 0.9807475209236145, + "learning_rate": 5.23060657655754e-07, + "loss": 0.5554, + "step": 13526 + }, + { + "epoch": 4.749648876404494, + "grad_norm": 1.0064703226089478, + "learning_rate": 5.227776994004355e-07, + "loss": 0.5402, + "step": 13527 + }, + { + "epoch": 4.75, + "grad_norm": 0.9903721809387207, + "learning_rate": 5.224948087644286e-07, + "loss": 0.5412, + "step": 13528 + }, + { + "epoch": 4.750351123595506, + "grad_norm": 0.9948616623878479, + "learning_rate": 5.222119857574062e-07, + "loss": 0.5416, + "step": 13529 + }, + { + "epoch": 4.750702247191011, + "grad_norm": 0.9851198196411133, + "learning_rate": 5.219292303890422e-07, + "loss": 0.5674, + "step": 13530 + }, + { + "epoch": 4.751053370786517, + "grad_norm": 0.9577903747558594, + "learning_rate": 5.216465426690051e-07, + "loss": 0.5836, + "step": 13531 + }, + { + "epoch": 4.751404494382022, + "grad_norm": 1.0492733716964722, + "learning_rate": 5.213639226069642e-07, + "loss": 0.5554, + "step": 13532 + }, + { + "epoch": 4.751755617977528, + "grad_norm": 0.9394395351409912, + "learning_rate": 5.210813702125831e-07, + "loss": 0.4926, + "step": 13533 + }, + { + "epoch": 4.752106741573034, + "grad_norm": 0.9649696946144104, + "learning_rate": 5.207988854955265e-07, + "loss": 0.5476, + "step": 13534 + }, + { + "epoch": 4.752457865168539, + "grad_norm": 0.9841094017028809, + "learning_rate": 5.205164684654545e-07, + "loss": 0.5357, + "step": 13535 + }, + { + "epoch": 4.752808988764045, + "grad_norm": 0.9456711411476135, + "learning_rate": 5.202341191320245e-07, + "loss": 0.5181, + "step": 13536 + }, + { + "epoch": 4.75316011235955, + "grad_norm": 0.9933743476867676, + "learning_rate": 5.199518375048945e-07, + "loss": 0.5662, + "step": 13537 + }, + { + "epoch": 4.753511235955056, + "grad_norm": 0.9763473868370056, + "learning_rate": 5.196696235937165e-07, + "loss": 0.5238, + "step": 13538 + }, + { + "epoch": 4.753862359550562, + "grad_norm": 0.9651517868041992, + "learning_rate": 5.193874774081434e-07, + "loss": 0.5117, + "step": 13539 + }, + { + "epoch": 4.754213483146067, + "grad_norm": 0.9977198839187622, + "learning_rate": 5.191053989578232e-07, + "loss": 0.5252, + "step": 13540 + }, + { + "epoch": 4.754564606741573, + "grad_norm": 1.0402394533157349, + "learning_rate": 5.188233882524038e-07, + "loss": 0.5593, + "step": 13541 + }, + { + "epoch": 4.754915730337078, + "grad_norm": 0.960919201374054, + "learning_rate": 5.185414453015286e-07, + "loss": 0.5312, + "step": 13542 + }, + { + "epoch": 4.755266853932584, + "grad_norm": 0.9776978492736816, + "learning_rate": 5.182595701148413e-07, + "loss": 0.5678, + "step": 13543 + }, + { + "epoch": 4.75561797752809, + "grad_norm": 0.9878711104393005, + "learning_rate": 5.1797776270198e-07, + "loss": 0.5178, + "step": 13544 + }, + { + "epoch": 4.755969101123595, + "grad_norm": 0.9552909731864929, + "learning_rate": 5.176960230725839e-07, + "loss": 0.4978, + "step": 13545 + }, + { + "epoch": 4.756320224719101, + "grad_norm": 0.9502068758010864, + "learning_rate": 5.174143512362867e-07, + "loss": 0.5142, + "step": 13546 + }, + { + "epoch": 4.756671348314606, + "grad_norm": 0.9721452593803406, + "learning_rate": 5.171327472027229e-07, + "loss": 0.5174, + "step": 13547 + }, + { + "epoch": 4.757022471910112, + "grad_norm": 0.969758152961731, + "learning_rate": 5.168512109815224e-07, + "loss": 0.525, + "step": 13548 + }, + { + "epoch": 4.757373595505618, + "grad_norm": 0.9905655384063721, + "learning_rate": 5.16569742582313e-07, + "loss": 0.5341, + "step": 13549 + }, + { + "epoch": 4.757724719101123, + "grad_norm": 0.9660579562187195, + "learning_rate": 5.162883420147216e-07, + "loss": 0.5831, + "step": 13550 + }, + { + "epoch": 4.758075842696629, + "grad_norm": 0.9673190712928772, + "learning_rate": 5.160070092883709e-07, + "loss": 0.5354, + "step": 13551 + }, + { + "epoch": 4.758426966292134, + "grad_norm": 0.9768142104148865, + "learning_rate": 5.157257444128832e-07, + "loss": 0.5609, + "step": 13552 + }, + { + "epoch": 4.75877808988764, + "grad_norm": 0.9798534512519836, + "learning_rate": 5.154445473978767e-07, + "loss": 0.5546, + "step": 13553 + }, + { + "epoch": 4.759129213483146, + "grad_norm": 0.9745402932167053, + "learning_rate": 5.151634182529692e-07, + "loss": 0.5612, + "step": 13554 + }, + { + "epoch": 4.759480337078652, + "grad_norm": 1.0018460750579834, + "learning_rate": 5.148823569877734e-07, + "loss": 0.5259, + "step": 13555 + }, + { + "epoch": 4.759831460674158, + "grad_norm": 0.9371743202209473, + "learning_rate": 5.146013636119029e-07, + "loss": 0.5363, + "step": 13556 + }, + { + "epoch": 4.760182584269663, + "grad_norm": 0.9896636009216309, + "learning_rate": 5.143204381349661e-07, + "loss": 0.5373, + "step": 13557 + }, + { + "epoch": 4.760533707865169, + "grad_norm": 1.0546395778656006, + "learning_rate": 5.140395805665721e-07, + "loss": 0.528, + "step": 13558 + }, + { + "epoch": 4.760884831460674, + "grad_norm": 0.9927265048027039, + "learning_rate": 5.137587909163244e-07, + "loss": 0.5317, + "step": 13559 + }, + { + "epoch": 4.76123595505618, + "grad_norm": 0.9910879135131836, + "learning_rate": 5.134780691938271e-07, + "loss": 0.5512, + "step": 13560 + }, + { + "epoch": 4.761587078651686, + "grad_norm": 0.9774528741836548, + "learning_rate": 5.131974154086797e-07, + "loss": 0.5709, + "step": 13561 + }, + { + "epoch": 4.761938202247191, + "grad_norm": 0.9872034192085266, + "learning_rate": 5.1291682957048e-07, + "loss": 0.5507, + "step": 13562 + }, + { + "epoch": 4.762289325842697, + "grad_norm": 0.9848805665969849, + "learning_rate": 5.12636311688825e-07, + "loss": 0.5488, + "step": 13563 + }, + { + "epoch": 4.762640449438202, + "grad_norm": 0.9619590640068054, + "learning_rate": 5.123558617733068e-07, + "loss": 0.5149, + "step": 13564 + }, + { + "epoch": 4.762991573033708, + "grad_norm": 0.9611052870750427, + "learning_rate": 5.120754798335181e-07, + "loss": 0.5454, + "step": 13565 + }, + { + "epoch": 4.763342696629214, + "grad_norm": 0.9357991814613342, + "learning_rate": 5.117951658790463e-07, + "loss": 0.5108, + "step": 13566 + }, + { + "epoch": 4.763693820224719, + "grad_norm": 0.9579005837440491, + "learning_rate": 5.115149199194788e-07, + "loss": 0.551, + "step": 13567 + }, + { + "epoch": 4.764044943820225, + "grad_norm": 1.0311585664749146, + "learning_rate": 5.11234741964399e-07, + "loss": 0.5166, + "step": 13568 + }, + { + "epoch": 4.76439606741573, + "grad_norm": 0.982831597328186, + "learning_rate": 5.109546320233899e-07, + "loss": 0.5665, + "step": 13569 + }, + { + "epoch": 4.764747191011236, + "grad_norm": 0.9572492241859436, + "learning_rate": 5.106745901060298e-07, + "loss": 0.5361, + "step": 13570 + }, + { + "epoch": 4.765098314606742, + "grad_norm": 0.968166708946228, + "learning_rate": 5.10394616221897e-07, + "loss": 0.5041, + "step": 13571 + }, + { + "epoch": 4.765449438202247, + "grad_norm": 0.9681075215339661, + "learning_rate": 5.101147103805656e-07, + "loss": 0.5065, + "step": 13572 + }, + { + "epoch": 4.765800561797753, + "grad_norm": 1.067628264427185, + "learning_rate": 5.09834872591608e-07, + "loss": 0.5394, + "step": 13573 + }, + { + "epoch": 4.766151685393258, + "grad_norm": 0.9676999449729919, + "learning_rate": 5.09555102864595e-07, + "loss": 0.5132, + "step": 13574 + }, + { + "epoch": 4.766502808988764, + "grad_norm": 0.987395167350769, + "learning_rate": 5.09275401209094e-07, + "loss": 0.5294, + "step": 13575 + }, + { + "epoch": 4.76685393258427, + "grad_norm": 0.9893847107887268, + "learning_rate": 5.089957676346713e-07, + "loss": 0.5391, + "step": 13576 + }, + { + "epoch": 4.767205056179775, + "grad_norm": 0.963917076587677, + "learning_rate": 5.087162021508888e-07, + "loss": 0.5023, + "step": 13577 + }, + { + "epoch": 4.767556179775281, + "grad_norm": 0.9899333715438843, + "learning_rate": 5.084367047673092e-07, + "loss": 0.5232, + "step": 13578 + }, + { + "epoch": 4.767907303370786, + "grad_norm": 0.9921440482139587, + "learning_rate": 5.081572754934891e-07, + "loss": 0.5756, + "step": 13579 + }, + { + "epoch": 4.768258426966292, + "grad_norm": 1.0098665952682495, + "learning_rate": 5.078779143389865e-07, + "loss": 0.5216, + "step": 13580 + }, + { + "epoch": 4.768609550561798, + "grad_norm": 0.9490517973899841, + "learning_rate": 5.075986213133538e-07, + "loss": 0.5206, + "step": 13581 + }, + { + "epoch": 4.768960674157303, + "grad_norm": 0.9905654788017273, + "learning_rate": 5.073193964261441e-07, + "loss": 0.5914, + "step": 13582 + }, + { + "epoch": 4.769311797752809, + "grad_norm": 0.9433330297470093, + "learning_rate": 5.070402396869053e-07, + "loss": 0.4933, + "step": 13583 + }, + { + "epoch": 4.769662921348314, + "grad_norm": 0.9669367671012878, + "learning_rate": 5.067611511051857e-07, + "loss": 0.5555, + "step": 13584 + }, + { + "epoch": 4.77001404494382, + "grad_norm": 0.9667124152183533, + "learning_rate": 5.064821306905288e-07, + "loss": 0.5255, + "step": 13585 + }, + { + "epoch": 4.770365168539326, + "grad_norm": 0.9895864725112915, + "learning_rate": 5.062031784524765e-07, + "loss": 0.5624, + "step": 13586 + }, + { + "epoch": 4.770716292134831, + "grad_norm": 0.9594666957855225, + "learning_rate": 5.059242944005702e-07, + "loss": 0.5083, + "step": 13587 + }, + { + "epoch": 4.771067415730337, + "grad_norm": 0.9694823026657104, + "learning_rate": 5.056454785443457e-07, + "loss": 0.5233, + "step": 13588 + }, + { + "epoch": 4.771418539325842, + "grad_norm": 1.0011597871780396, + "learning_rate": 5.053667308933402e-07, + "loss": 0.5373, + "step": 13589 + }, + { + "epoch": 4.771769662921348, + "grad_norm": 0.9510133266448975, + "learning_rate": 5.050880514570849e-07, + "loss": 0.5355, + "step": 13590 + }, + { + "epoch": 4.772120786516854, + "grad_norm": 0.9874480962753296, + "learning_rate": 5.04809440245112e-07, + "loss": 0.5667, + "step": 13591 + }, + { + "epoch": 4.77247191011236, + "grad_norm": 0.9638926386833191, + "learning_rate": 5.04530897266948e-07, + "loss": 0.544, + "step": 13592 + }, + { + "epoch": 4.772823033707866, + "grad_norm": 0.9565342664718628, + "learning_rate": 5.042524225321208e-07, + "loss": 0.5467, + "step": 13593 + }, + { + "epoch": 4.773174157303371, + "grad_norm": 0.9960715770721436, + "learning_rate": 5.039740160501528e-07, + "loss": 0.5882, + "step": 13594 + }, + { + "epoch": 4.773525280898877, + "grad_norm": 0.9889675974845886, + "learning_rate": 5.036956778305646e-07, + "loss": 0.5506, + "step": 13595 + }, + { + "epoch": 4.773876404494382, + "grad_norm": 0.967841625213623, + "learning_rate": 5.034174078828766e-07, + "loss": 0.5583, + "step": 13596 + }, + { + "epoch": 4.774227528089888, + "grad_norm": 0.944546639919281, + "learning_rate": 5.031392062166043e-07, + "loss": 0.5385, + "step": 13597 + }, + { + "epoch": 4.774578651685394, + "grad_norm": 0.961641788482666, + "learning_rate": 5.028610728412631e-07, + "loss": 0.5138, + "step": 13598 + }, + { + "epoch": 4.774929775280899, + "grad_norm": 0.9665804505348206, + "learning_rate": 5.025830077663632e-07, + "loss": 0.5815, + "step": 13599 + }, + { + "epoch": 4.775280898876405, + "grad_norm": 0.9821751117706299, + "learning_rate": 5.023050110014164e-07, + "loss": 0.5409, + "step": 13600 + }, + { + "epoch": 4.77563202247191, + "grad_norm": 0.9686010479927063, + "learning_rate": 5.020270825559284e-07, + "loss": 0.5245, + "step": 13601 + }, + { + "epoch": 4.775983146067416, + "grad_norm": 0.932789146900177, + "learning_rate": 5.017492224394038e-07, + "loss": 0.5064, + "step": 13602 + }, + { + "epoch": 4.776334269662922, + "grad_norm": 0.9546009302139282, + "learning_rate": 5.014714306613466e-07, + "loss": 0.5167, + "step": 13603 + }, + { + "epoch": 4.776685393258427, + "grad_norm": 0.9906979203224182, + "learning_rate": 5.011937072312559e-07, + "loss": 0.5453, + "step": 13604 + }, + { + "epoch": 4.777036516853933, + "grad_norm": 0.9497023820877075, + "learning_rate": 5.009160521586304e-07, + "loss": 0.5289, + "step": 13605 + }, + { + "epoch": 4.777387640449438, + "grad_norm": 0.9796620607376099, + "learning_rate": 5.006384654529647e-07, + "loss": 0.532, + "step": 13606 + }, + { + "epoch": 4.777738764044944, + "grad_norm": 0.9487783908843994, + "learning_rate": 5.003609471237533e-07, + "loss": 0.5324, + "step": 13607 + }, + { + "epoch": 4.77808988764045, + "grad_norm": 0.966387927532196, + "learning_rate": 5.000834971804864e-07, + "loss": 0.511, + "step": 13608 + }, + { + "epoch": 4.778441011235955, + "grad_norm": 0.9807990789413452, + "learning_rate": 4.998061156326517e-07, + "loss": 0.534, + "step": 13609 + }, + { + "epoch": 4.778792134831461, + "grad_norm": 1.0156803131103516, + "learning_rate": 4.995288024897371e-07, + "loss": 0.542, + "step": 13610 + }, + { + "epoch": 4.779143258426966, + "grad_norm": 0.9957343339920044, + "learning_rate": 4.992515577612248e-07, + "loss": 0.5576, + "step": 13611 + }, + { + "epoch": 4.779494382022472, + "grad_norm": 0.9642595648765564, + "learning_rate": 4.989743814565978e-07, + "loss": 0.5474, + "step": 13612 + }, + { + "epoch": 4.779845505617978, + "grad_norm": 0.9806334972381592, + "learning_rate": 4.986972735853343e-07, + "loss": 0.5832, + "step": 13613 + }, + { + "epoch": 4.780196629213483, + "grad_norm": 0.9701009392738342, + "learning_rate": 4.984202341569119e-07, + "loss": 0.5348, + "step": 13614 + }, + { + "epoch": 4.780547752808989, + "grad_norm": 0.9773545861244202, + "learning_rate": 4.981432631808042e-07, + "loss": 0.5248, + "step": 13615 + }, + { + "epoch": 4.780898876404494, + "grad_norm": 0.9561938643455505, + "learning_rate": 4.978663606664844e-07, + "loss": 0.4893, + "step": 13616 + }, + { + "epoch": 4.78125, + "grad_norm": 0.9784732460975647, + "learning_rate": 4.975895266234221e-07, + "loss": 0.5131, + "step": 13617 + }, + { + "epoch": 4.781601123595506, + "grad_norm": 0.9681968688964844, + "learning_rate": 4.973127610610834e-07, + "loss": 0.5455, + "step": 13618 + }, + { + "epoch": 4.781952247191011, + "grad_norm": 0.9768074154853821, + "learning_rate": 4.970360639889354e-07, + "loss": 0.5576, + "step": 13619 + }, + { + "epoch": 4.782303370786517, + "grad_norm": 0.9537262320518494, + "learning_rate": 4.967594354164396e-07, + "loss": 0.5101, + "step": 13620 + }, + { + "epoch": 4.782654494382022, + "grad_norm": 0.9692298173904419, + "learning_rate": 4.964828753530574e-07, + "loss": 0.5417, + "step": 13621 + }, + { + "epoch": 4.783005617977528, + "grad_norm": 0.9538604617118835, + "learning_rate": 4.962063838082457e-07, + "loss": 0.5233, + "step": 13622 + }, + { + "epoch": 4.783356741573034, + "grad_norm": 0.9698537588119507, + "learning_rate": 4.959299607914619e-07, + "loss": 0.5241, + "step": 13623 + }, + { + "epoch": 4.783707865168539, + "grad_norm": 0.9762541651725769, + "learning_rate": 4.956536063121583e-07, + "loss": 0.5416, + "step": 13624 + }, + { + "epoch": 4.784058988764045, + "grad_norm": 0.9798547625541687, + "learning_rate": 4.953773203797857e-07, + "loss": 0.5541, + "step": 13625 + }, + { + "epoch": 4.78441011235955, + "grad_norm": 0.9744400382041931, + "learning_rate": 4.951011030037939e-07, + "loss": 0.5517, + "step": 13626 + }, + { + "epoch": 4.784761235955056, + "grad_norm": 0.9704358577728271, + "learning_rate": 4.948249541936281e-07, + "loss": 0.5602, + "step": 13627 + }, + { + "epoch": 4.785112359550562, + "grad_norm": 0.9661256670951843, + "learning_rate": 4.945488739587336e-07, + "loss": 0.532, + "step": 13628 + }, + { + "epoch": 4.785463483146067, + "grad_norm": 1.0317909717559814, + "learning_rate": 4.942728623085513e-07, + "loss": 0.6004, + "step": 13629 + }, + { + "epoch": 4.785814606741573, + "grad_norm": 0.9873784780502319, + "learning_rate": 4.93996919252521e-07, + "loss": 0.551, + "step": 13630 + }, + { + "epoch": 4.786165730337078, + "grad_norm": 0.9758477210998535, + "learning_rate": 4.93721044800079e-07, + "loss": 0.5691, + "step": 13631 + }, + { + "epoch": 4.786516853932584, + "grad_norm": 0.952049970626831, + "learning_rate": 4.934452389606611e-07, + "loss": 0.5267, + "step": 13632 + }, + { + "epoch": 4.78686797752809, + "grad_norm": 1.0296107530593872, + "learning_rate": 4.931695017436982e-07, + "loss": 0.5107, + "step": 13633 + }, + { + "epoch": 4.787219101123595, + "grad_norm": 0.9379881024360657, + "learning_rate": 4.928938331586219e-07, + "loss": 0.5038, + "step": 13634 + }, + { + "epoch": 4.787570224719101, + "grad_norm": 0.9746097326278687, + "learning_rate": 4.92618233214859e-07, + "loss": 0.5531, + "step": 13635 + }, + { + "epoch": 4.787921348314606, + "grad_norm": 0.9547182321548462, + "learning_rate": 4.923427019218341e-07, + "loss": 0.4764, + "step": 13636 + }, + { + "epoch": 4.788272471910112, + "grad_norm": 0.9933334589004517, + "learning_rate": 4.920672392889716e-07, + "loss": 0.5615, + "step": 13637 + }, + { + "epoch": 4.788623595505618, + "grad_norm": 0.9437901973724365, + "learning_rate": 4.917918453256904e-07, + "loss": 0.5394, + "step": 13638 + }, + { + "epoch": 4.788974719101123, + "grad_norm": 0.970104455947876, + "learning_rate": 4.915165200414105e-07, + "loss": 0.5676, + "step": 13639 + }, + { + "epoch": 4.789325842696629, + "grad_norm": 0.9474315047264099, + "learning_rate": 4.912412634455462e-07, + "loss": 0.5221, + "step": 13640 + }, + { + "epoch": 4.789676966292134, + "grad_norm": 0.9613620042800903, + "learning_rate": 4.909660755475124e-07, + "loss": 0.5248, + "step": 13641 + }, + { + "epoch": 4.79002808988764, + "grad_norm": 0.9648392200469971, + "learning_rate": 4.906909563567192e-07, + "loss": 0.5335, + "step": 13642 + }, + { + "epoch": 4.790379213483146, + "grad_norm": 0.9812972545623779, + "learning_rate": 4.904159058825761e-07, + "loss": 0.5657, + "step": 13643 + }, + { + "epoch": 4.790730337078652, + "grad_norm": 1.0096533298492432, + "learning_rate": 4.901409241344892e-07, + "loss": 0.5606, + "step": 13644 + }, + { + "epoch": 4.791081460674158, + "grad_norm": 0.9440931677818298, + "learning_rate": 4.898660111218634e-07, + "loss": 0.5442, + "step": 13645 + }, + { + "epoch": 4.791432584269663, + "grad_norm": 0.9513047933578491, + "learning_rate": 4.89591166854099e-07, + "loss": 0.52, + "step": 13646 + }, + { + "epoch": 4.791783707865169, + "grad_norm": 0.960460364818573, + "learning_rate": 4.893163913405971e-07, + "loss": 0.5328, + "step": 13647 + }, + { + "epoch": 4.792134831460674, + "grad_norm": 0.9683283567428589, + "learning_rate": 4.890416845907541e-07, + "loss": 0.5178, + "step": 13648 + }, + { + "epoch": 4.79248595505618, + "grad_norm": 0.9579154253005981, + "learning_rate": 4.887670466139643e-07, + "loss": 0.5496, + "step": 13649 + }, + { + "epoch": 4.792837078651686, + "grad_norm": 0.9972191452980042, + "learning_rate": 4.88492477419621e-07, + "loss": 0.5821, + "step": 13650 + }, + { + "epoch": 4.793188202247191, + "grad_norm": 0.983424723148346, + "learning_rate": 4.882179770171127e-07, + "loss": 0.5444, + "step": 13651 + }, + { + "epoch": 4.793539325842697, + "grad_norm": 0.9934150576591492, + "learning_rate": 4.87943545415829e-07, + "loss": 0.5553, + "step": 13652 + }, + { + "epoch": 4.793890449438202, + "grad_norm": 0.964160680770874, + "learning_rate": 4.876691826251536e-07, + "loss": 0.5348, + "step": 13653 + }, + { + "epoch": 4.794241573033708, + "grad_norm": 0.9472085237503052, + "learning_rate": 4.873948886544711e-07, + "loss": 0.51, + "step": 13654 + }, + { + "epoch": 4.794592696629214, + "grad_norm": 0.9641125798225403, + "learning_rate": 4.8712066351316e-07, + "loss": 0.5343, + "step": 13655 + }, + { + "epoch": 4.794943820224719, + "grad_norm": 1.002886414527893, + "learning_rate": 4.86846507210601e-07, + "loss": 0.5749, + "step": 13656 + }, + { + "epoch": 4.795294943820225, + "grad_norm": 1.0168867111206055, + "learning_rate": 4.865724197561677e-07, + "loss": 0.5605, + "step": 13657 + }, + { + "epoch": 4.79564606741573, + "grad_norm": 0.9880848526954651, + "learning_rate": 4.862984011592356e-07, + "loss": 0.5367, + "step": 13658 + }, + { + "epoch": 4.795997191011236, + "grad_norm": 0.979347288608551, + "learning_rate": 4.860244514291743e-07, + "loss": 0.5395, + "step": 13659 + }, + { + "epoch": 4.796348314606742, + "grad_norm": 0.9847811460494995, + "learning_rate": 4.85750570575354e-07, + "loss": 0.5406, + "step": 13660 + }, + { + "epoch": 4.796699438202247, + "grad_norm": 0.9638721346855164, + "learning_rate": 4.854767586071404e-07, + "loss": 0.5552, + "step": 13661 + }, + { + "epoch": 4.797050561797753, + "grad_norm": 1.0127390623092651, + "learning_rate": 4.852030155338974e-07, + "loss": 0.5386, + "step": 13662 + }, + { + "epoch": 4.797401685393258, + "grad_norm": 0.9849705100059509, + "learning_rate": 4.84929341364988e-07, + "loss": 0.5508, + "step": 13663 + }, + { + "epoch": 4.797752808988764, + "grad_norm": 0.974226176738739, + "learning_rate": 4.846557361097698e-07, + "loss": 0.5237, + "step": 13664 + }, + { + "epoch": 4.79810393258427, + "grad_norm": 0.9848212599754333, + "learning_rate": 4.843821997776016e-07, + "loss": 0.571, + "step": 13665 + }, + { + "epoch": 4.798455056179775, + "grad_norm": 0.9636582136154175, + "learning_rate": 4.841087323778368e-07, + "loss": 0.527, + "step": 13666 + }, + { + "epoch": 4.798806179775281, + "grad_norm": 1.0044348239898682, + "learning_rate": 4.838353339198291e-07, + "loss": 0.5464, + "step": 13667 + }, + { + "epoch": 4.799157303370786, + "grad_norm": 0.958339273929596, + "learning_rate": 4.835620044129269e-07, + "loss": 0.5396, + "step": 13668 + }, + { + "epoch": 4.799508426966292, + "grad_norm": 1.0109882354736328, + "learning_rate": 4.832887438664794e-07, + "loss": 0.5114, + "step": 13669 + }, + { + "epoch": 4.799859550561798, + "grad_norm": 0.9450926184654236, + "learning_rate": 4.830155522898303e-07, + "loss": 0.5441, + "step": 13670 + }, + { + "epoch": 4.800210674157303, + "grad_norm": 0.9655445218086243, + "learning_rate": 4.827424296923242e-07, + "loss": 0.5652, + "step": 13671 + }, + { + "epoch": 4.800561797752809, + "grad_norm": 0.9697320461273193, + "learning_rate": 4.824693760833002e-07, + "loss": 0.5323, + "step": 13672 + }, + { + "epoch": 4.800912921348314, + "grad_norm": 1.003684163093567, + "learning_rate": 4.821963914720976e-07, + "loss": 0.5437, + "step": 13673 + }, + { + "epoch": 4.80126404494382, + "grad_norm": 0.9822651147842407, + "learning_rate": 4.819234758680519e-07, + "loss": 0.5517, + "step": 13674 + }, + { + "epoch": 4.801615168539326, + "grad_norm": 0.9789892435073853, + "learning_rate": 4.816506292804956e-07, + "loss": 0.5449, + "step": 13675 + }, + { + "epoch": 4.801966292134831, + "grad_norm": 0.9706618189811707, + "learning_rate": 4.813778517187614e-07, + "loss": 0.5446, + "step": 13676 + }, + { + "epoch": 4.802317415730337, + "grad_norm": 0.9626661539077759, + "learning_rate": 4.811051431921768e-07, + "loss": 0.5309, + "step": 13677 + }, + { + "epoch": 4.802668539325842, + "grad_norm": 1.0480437278747559, + "learning_rate": 4.808325037100692e-07, + "loss": 0.5721, + "step": 13678 + }, + { + "epoch": 4.803019662921348, + "grad_norm": 0.980779230594635, + "learning_rate": 4.805599332817615e-07, + "loss": 0.5399, + "step": 13679 + }, + { + "epoch": 4.803370786516854, + "grad_norm": 0.9921133518218994, + "learning_rate": 4.802874319165765e-07, + "loss": 0.59, + "step": 13680 + }, + { + "epoch": 4.80372191011236, + "grad_norm": 0.9640685319900513, + "learning_rate": 4.800149996238326e-07, + "loss": 0.5383, + "step": 13681 + }, + { + "epoch": 4.804073033707866, + "grad_norm": 0.981501042842865, + "learning_rate": 4.79742636412848e-07, + "loss": 0.5487, + "step": 13682 + }, + { + "epoch": 4.804424157303371, + "grad_norm": 0.970636248588562, + "learning_rate": 4.794703422929356e-07, + "loss": 0.5475, + "step": 13683 + }, + { + "epoch": 4.804775280898877, + "grad_norm": 0.9795015454292297, + "learning_rate": 4.791981172734092e-07, + "loss": 0.5618, + "step": 13684 + }, + { + "epoch": 4.805126404494382, + "grad_norm": 0.959946870803833, + "learning_rate": 4.789259613635772e-07, + "loss": 0.5157, + "step": 13685 + }, + { + "epoch": 4.805477528089888, + "grad_norm": 0.948172390460968, + "learning_rate": 4.786538745727487e-07, + "loss": 0.5184, + "step": 13686 + }, + { + "epoch": 4.805828651685394, + "grad_norm": 0.9416277408599854, + "learning_rate": 4.783818569102277e-07, + "loss": 0.5319, + "step": 13687 + }, + { + "epoch": 4.806179775280899, + "grad_norm": 0.9604785442352295, + "learning_rate": 4.78109908385317e-07, + "loss": 0.5736, + "step": 13688 + }, + { + "epoch": 4.806530898876405, + "grad_norm": 0.9787455201148987, + "learning_rate": 4.778380290073178e-07, + "loss": 0.5507, + "step": 13689 + }, + { + "epoch": 4.80688202247191, + "grad_norm": 0.9981918931007385, + "learning_rate": 4.775662187855268e-07, + "loss": 0.5319, + "step": 13690 + }, + { + "epoch": 4.807233146067416, + "grad_norm": 0.9544591903686523, + "learning_rate": 4.772944777292412e-07, + "loss": 0.5615, + "step": 13691 + }, + { + "epoch": 4.807584269662922, + "grad_norm": 0.9600389003753662, + "learning_rate": 4.770228058477533e-07, + "loss": 0.5505, + "step": 13692 + }, + { + "epoch": 4.807935393258427, + "grad_norm": 0.9813960790634155, + "learning_rate": 4.767512031503546e-07, + "loss": 0.5674, + "step": 13693 + }, + { + "epoch": 4.808286516853933, + "grad_norm": 0.9972847104072571, + "learning_rate": 4.764796696463331e-07, + "loss": 0.5765, + "step": 13694 + }, + { + "epoch": 4.808637640449438, + "grad_norm": 0.9943374395370483, + "learning_rate": 4.762082053449757e-07, + "loss": 0.5236, + "step": 13695 + }, + { + "epoch": 4.808988764044944, + "grad_norm": 0.9727316498756409, + "learning_rate": 4.7593681025556533e-07, + "loss": 0.5511, + "step": 13696 + }, + { + "epoch": 4.80933988764045, + "grad_norm": 0.9753355383872986, + "learning_rate": 4.756654843873848e-07, + "loss": 0.5609, + "step": 13697 + }, + { + "epoch": 4.809691011235955, + "grad_norm": 0.9831970930099487, + "learning_rate": 4.7539422774971225e-07, + "loss": 0.5451, + "step": 13698 + }, + { + "epoch": 4.810042134831461, + "grad_norm": 0.957246720790863, + "learning_rate": 4.7512304035182424e-07, + "loss": 0.5269, + "step": 13699 + }, + { + "epoch": 4.810393258426966, + "grad_norm": 0.9534024596214294, + "learning_rate": 4.74851922202996e-07, + "loss": 0.5546, + "step": 13700 + }, + { + "epoch": 4.810744382022472, + "grad_norm": 0.94466233253479, + "learning_rate": 4.7458087331249845e-07, + "loss": 0.5436, + "step": 13701 + }, + { + "epoch": 4.811095505617978, + "grad_norm": 0.9569569230079651, + "learning_rate": 4.743098936896026e-07, + "loss": 0.539, + "step": 13702 + }, + { + "epoch": 4.811446629213483, + "grad_norm": 0.928993821144104, + "learning_rate": 4.740389833435741e-07, + "loss": 0.5365, + "step": 13703 + }, + { + "epoch": 4.811797752808989, + "grad_norm": 0.9804869890213013, + "learning_rate": 4.7376814228367956e-07, + "loss": 0.5224, + "step": 13704 + }, + { + "epoch": 4.812148876404494, + "grad_norm": 0.9874171614646912, + "learning_rate": 4.7349737051918004e-07, + "loss": 0.5278, + "step": 13705 + }, + { + "epoch": 4.8125, + "grad_norm": 1.041128396987915, + "learning_rate": 4.73226668059337e-07, + "loss": 0.5741, + "step": 13706 + }, + { + "epoch": 4.812851123595506, + "grad_norm": 0.9641829133033752, + "learning_rate": 4.7295603491340685e-07, + "loss": 0.5467, + "step": 13707 + }, + { + "epoch": 4.813202247191011, + "grad_norm": 0.982672393321991, + "learning_rate": 4.7268547109064646e-07, + "loss": 0.5133, + "step": 13708 + }, + { + "epoch": 4.813553370786517, + "grad_norm": 0.9728686213493347, + "learning_rate": 4.724149766003075e-07, + "loss": 0.5435, + "step": 13709 + }, + { + "epoch": 4.813904494382022, + "grad_norm": 0.9826738238334656, + "learning_rate": 4.721445514516421e-07, + "loss": 0.557, + "step": 13710 + }, + { + "epoch": 4.814255617977528, + "grad_norm": 0.9926527142524719, + "learning_rate": 4.718741956538975e-07, + "loss": 0.5453, + "step": 13711 + }, + { + "epoch": 4.814606741573034, + "grad_norm": 0.9998329281806946, + "learning_rate": 4.7160390921631936e-07, + "loss": 0.5549, + "step": 13712 + }, + { + "epoch": 4.814957865168539, + "grad_norm": 0.9961434006690979, + "learning_rate": 4.713336921481526e-07, + "loss": 0.5811, + "step": 13713 + }, + { + "epoch": 4.815308988764045, + "grad_norm": 1.0026239156723022, + "learning_rate": 4.710635444586367e-07, + "loss": 0.5322, + "step": 13714 + }, + { + "epoch": 4.81566011235955, + "grad_norm": 0.9970614910125732, + "learning_rate": 4.707934661570124e-07, + "loss": 0.5686, + "step": 13715 + }, + { + "epoch": 4.816011235955056, + "grad_norm": 1.0040990114212036, + "learning_rate": 4.7052345725251413e-07, + "loss": 0.5346, + "step": 13716 + }, + { + "epoch": 4.816362359550562, + "grad_norm": 0.9407042860984802, + "learning_rate": 4.7025351775437794e-07, + "loss": 0.5248, + "step": 13717 + }, + { + "epoch": 4.816713483146067, + "grad_norm": 0.9944391846656799, + "learning_rate": 4.6998364767183363e-07, + "loss": 0.5684, + "step": 13718 + }, + { + "epoch": 4.817064606741573, + "grad_norm": 0.9613637328147888, + "learning_rate": 4.6971384701411217e-07, + "loss": 0.5331, + "step": 13719 + }, + { + "epoch": 4.817415730337078, + "grad_norm": 0.9962686896324158, + "learning_rate": 4.694441157904392e-07, + "loss": 0.571, + "step": 13720 + }, + { + "epoch": 4.817766853932584, + "grad_norm": 0.9852620959281921, + "learning_rate": 4.691744540100407e-07, + "loss": 0.5371, + "step": 13721 + }, + { + "epoch": 4.81811797752809, + "grad_norm": 0.9495635032653809, + "learning_rate": 4.689048616821373e-07, + "loss": 0.5224, + "step": 13722 + }, + { + "epoch": 4.818469101123595, + "grad_norm": 0.9661790728569031, + "learning_rate": 4.686353388159501e-07, + "loss": 0.5082, + "step": 13723 + }, + { + "epoch": 4.818820224719101, + "grad_norm": 0.9719508290290833, + "learning_rate": 4.683658854206963e-07, + "loss": 0.5209, + "step": 13724 + }, + { + "epoch": 4.819171348314606, + "grad_norm": 1.0281143188476562, + "learning_rate": 4.680965015055899e-07, + "loss": 0.5106, + "step": 13725 + }, + { + "epoch": 4.819522471910112, + "grad_norm": 0.9964945316314697, + "learning_rate": 4.6782718707984507e-07, + "loss": 0.5431, + "step": 13726 + }, + { + "epoch": 4.819873595505618, + "grad_norm": 0.9703767895698547, + "learning_rate": 4.67557942152671e-07, + "loss": 0.5569, + "step": 13727 + }, + { + "epoch": 4.820224719101123, + "grad_norm": 0.944754958152771, + "learning_rate": 4.672887667332765e-07, + "loss": 0.5199, + "step": 13728 + }, + { + "epoch": 4.820575842696629, + "grad_norm": 0.9511826634407043, + "learning_rate": 4.6701966083086636e-07, + "loss": 0.5424, + "step": 13729 + }, + { + "epoch": 4.820926966292134, + "grad_norm": 0.9462835192680359, + "learning_rate": 4.6675062445464494e-07, + "loss": 0.5155, + "step": 13730 + }, + { + "epoch": 4.82127808988764, + "grad_norm": 0.9739974737167358, + "learning_rate": 4.664816576138112e-07, + "loss": 0.5559, + "step": 13731 + }, + { + "epoch": 4.821629213483146, + "grad_norm": 0.9323555827140808, + "learning_rate": 4.662127603175656e-07, + "loss": 0.5054, + "step": 13732 + }, + { + "epoch": 4.821980337078652, + "grad_norm": 0.9768226742744446, + "learning_rate": 4.659439325751025e-07, + "loss": 0.5627, + "step": 13733 + }, + { + "epoch": 4.822331460674158, + "grad_norm": 0.963914692401886, + "learning_rate": 4.6567517439561726e-07, + "loss": 0.5367, + "step": 13734 + }, + { + "epoch": 4.822682584269663, + "grad_norm": 0.9255300164222717, + "learning_rate": 4.654064857882992e-07, + "loss": 0.5129, + "step": 13735 + }, + { + "epoch": 4.823033707865169, + "grad_norm": 0.9537528157234192, + "learning_rate": 4.651378667623394e-07, + "loss": 0.5296, + "step": 13736 + }, + { + "epoch": 4.823384831460674, + "grad_norm": 0.9569952487945557, + "learning_rate": 4.648693173269228e-07, + "loss": 0.5663, + "step": 13737 + }, + { + "epoch": 4.82373595505618, + "grad_norm": 0.9857463836669922, + "learning_rate": 4.646008374912336e-07, + "loss": 0.5503, + "step": 13738 + }, + { + "epoch": 4.824087078651686, + "grad_norm": 0.9919613599777222, + "learning_rate": 4.6433242726445506e-07, + "loss": 0.5515, + "step": 13739 + }, + { + "epoch": 4.824438202247191, + "grad_norm": 0.9597722887992859, + "learning_rate": 4.640640866557644e-07, + "loss": 0.5359, + "step": 13740 + }, + { + "epoch": 4.824789325842697, + "grad_norm": 0.9877889752388, + "learning_rate": 4.637958156743408e-07, + "loss": 0.5316, + "step": 13741 + }, + { + "epoch": 4.825140449438202, + "grad_norm": 0.9800974130630493, + "learning_rate": 4.635276143293571e-07, + "loss": 0.5321, + "step": 13742 + }, + { + "epoch": 4.825491573033708, + "grad_norm": 0.9605726599693298, + "learning_rate": 4.632594826299869e-07, + "loss": 0.5096, + "step": 13743 + }, + { + "epoch": 4.825842696629214, + "grad_norm": 0.9683108925819397, + "learning_rate": 4.6299142058539914e-07, + "loss": 0.5287, + "step": 13744 + }, + { + "epoch": 4.826193820224719, + "grad_norm": 0.9861086010932922, + "learning_rate": 4.627234282047621e-07, + "loss": 0.525, + "step": 13745 + }, + { + "epoch": 4.826544943820225, + "grad_norm": 0.9712125658988953, + "learning_rate": 4.6245550549723987e-07, + "loss": 0.5354, + "step": 13746 + }, + { + "epoch": 4.82689606741573, + "grad_norm": 0.97761470079422, + "learning_rate": 4.621876524719965e-07, + "loss": 0.5362, + "step": 13747 + }, + { + "epoch": 4.827247191011236, + "grad_norm": 0.9935604333877563, + "learning_rate": 4.61919869138191e-07, + "loss": 0.5638, + "step": 13748 + }, + { + "epoch": 4.827598314606742, + "grad_norm": 0.9650956988334656, + "learning_rate": 4.616521555049827e-07, + "loss": 0.5378, + "step": 13749 + }, + { + "epoch": 4.827949438202247, + "grad_norm": 0.975604236125946, + "learning_rate": 4.613845115815263e-07, + "loss": 0.5467, + "step": 13750 + }, + { + "epoch": 4.828300561797753, + "grad_norm": 0.970330536365509, + "learning_rate": 4.611169373769747e-07, + "loss": 0.5155, + "step": 13751 + }, + { + "epoch": 4.828651685393258, + "grad_norm": 0.9494550228118896, + "learning_rate": 4.608494329004795e-07, + "loss": 0.5544, + "step": 13752 + }, + { + "epoch": 4.829002808988764, + "grad_norm": 1.0000163316726685, + "learning_rate": 4.6058199816118837e-07, + "loss": 0.5444, + "step": 13753 + }, + { + "epoch": 4.82935393258427, + "grad_norm": 0.9562383890151978, + "learning_rate": 4.603146331682484e-07, + "loss": 0.5279, + "step": 13754 + }, + { + "epoch": 4.829705056179775, + "grad_norm": 0.9941381812095642, + "learning_rate": 4.6004733793080183e-07, + "loss": 0.5391, + "step": 13755 + }, + { + "epoch": 4.830056179775281, + "grad_norm": 0.9634647369384766, + "learning_rate": 4.597801124579912e-07, + "loss": 0.5608, + "step": 13756 + }, + { + "epoch": 4.830407303370786, + "grad_norm": 0.9499261379241943, + "learning_rate": 4.595129567589546e-07, + "loss": 0.5826, + "step": 13757 + }, + { + "epoch": 4.830758426966292, + "grad_norm": 0.9652177095413208, + "learning_rate": 4.5924587084282906e-07, + "loss": 0.5584, + "step": 13758 + }, + { + "epoch": 4.831109550561798, + "grad_norm": 0.9631256461143494, + "learning_rate": 4.589788547187479e-07, + "loss": 0.5416, + "step": 13759 + }, + { + "epoch": 4.831460674157303, + "grad_norm": 0.97127366065979, + "learning_rate": 4.587119083958441e-07, + "loss": 0.5167, + "step": 13760 + }, + { + "epoch": 4.831811797752809, + "grad_norm": 0.9379510283470154, + "learning_rate": 4.5844503188324647e-07, + "loss": 0.4999, + "step": 13761 + }, + { + "epoch": 4.832162921348314, + "grad_norm": 0.9713377356529236, + "learning_rate": 4.581782251900807e-07, + "loss": 0.5218, + "step": 13762 + }, + { + "epoch": 4.83251404494382, + "grad_norm": 0.9631909132003784, + "learning_rate": 4.579114883254734e-07, + "loss": 0.5394, + "step": 13763 + }, + { + "epoch": 4.832865168539326, + "grad_norm": 0.9955924153327942, + "learning_rate": 4.576448212985449e-07, + "loss": 0.5457, + "step": 13764 + }, + { + "epoch": 4.833216292134831, + "grad_norm": 0.9950634241104126, + "learning_rate": 4.573782241184166e-07, + "loss": 0.5661, + "step": 13765 + }, + { + "epoch": 4.833567415730337, + "grad_norm": 1.0104793310165405, + "learning_rate": 4.571116967942046e-07, + "loss": 0.5618, + "step": 13766 + }, + { + "epoch": 4.833918539325842, + "grad_norm": 1.0396498441696167, + "learning_rate": 4.5684523933502504e-07, + "loss": 0.5405, + "step": 13767 + }, + { + "epoch": 4.834269662921348, + "grad_norm": 0.9696619510650635, + "learning_rate": 4.5657885174998975e-07, + "loss": 0.5428, + "step": 13768 + }, + { + "epoch": 4.834620786516854, + "grad_norm": 0.9402617812156677, + "learning_rate": 4.563125340482086e-07, + "loss": 0.5105, + "step": 13769 + }, + { + "epoch": 4.83497191011236, + "grad_norm": 0.9600562453269958, + "learning_rate": 4.560462862387907e-07, + "loss": 0.5199, + "step": 13770 + }, + { + "epoch": 4.835323033707866, + "grad_norm": 0.9397798180580139, + "learning_rate": 4.5578010833084026e-07, + "loss": 0.5319, + "step": 13771 + }, + { + "epoch": 4.835674157303371, + "grad_norm": 0.9444599151611328, + "learning_rate": 4.555140003334613e-07, + "loss": 0.532, + "step": 13772 + }, + { + "epoch": 4.836025280898877, + "grad_norm": 0.9760026931762695, + "learning_rate": 4.552479622557532e-07, + "loss": 0.5247, + "step": 13773 + }, + { + "epoch": 4.836376404494382, + "grad_norm": 0.963423490524292, + "learning_rate": 4.5498199410681593e-07, + "loss": 0.5339, + "step": 13774 + }, + { + "epoch": 4.836727528089888, + "grad_norm": 1.0056263208389282, + "learning_rate": 4.5471609589574444e-07, + "loss": 0.5739, + "step": 13775 + }, + { + "epoch": 4.837078651685394, + "grad_norm": 0.9658178091049194, + "learning_rate": 4.5445026763163174e-07, + "loss": 0.532, + "step": 13776 + }, + { + "epoch": 4.837429775280899, + "grad_norm": 0.9574733376502991, + "learning_rate": 4.541845093235703e-07, + "loss": 0.5218, + "step": 13777 + }, + { + "epoch": 4.837780898876405, + "grad_norm": 0.9655747413635254, + "learning_rate": 4.539188209806472e-07, + "loss": 0.5411, + "step": 13778 + }, + { + "epoch": 4.83813202247191, + "grad_norm": 0.9800891876220703, + "learning_rate": 4.5365320261195003e-07, + "loss": 0.5409, + "step": 13779 + }, + { + "epoch": 4.838483146067416, + "grad_norm": 0.9645065069198608, + "learning_rate": 4.5338765422656196e-07, + "loss": 0.5515, + "step": 13780 + }, + { + "epoch": 4.838834269662922, + "grad_norm": 0.9558544158935547, + "learning_rate": 4.531221758335652e-07, + "loss": 0.5375, + "step": 13781 + }, + { + "epoch": 4.839185393258427, + "grad_norm": 1.0052781105041504, + "learning_rate": 4.528567674420384e-07, + "loss": 0.543, + "step": 13782 + }, + { + "epoch": 4.839536516853933, + "grad_norm": 0.981176495552063, + "learning_rate": 4.5259142906105805e-07, + "loss": 0.5357, + "step": 13783 + }, + { + "epoch": 4.839887640449438, + "grad_norm": 0.9787960052490234, + "learning_rate": 4.5232616069969926e-07, + "loss": 0.5238, + "step": 13784 + }, + { + "epoch": 4.840238764044944, + "grad_norm": 0.9605594873428345, + "learning_rate": 4.52060962367033e-07, + "loss": 0.5405, + "step": 13785 + }, + { + "epoch": 4.84058988764045, + "grad_norm": 1.5937727689743042, + "learning_rate": 4.5179583407212987e-07, + "loss": 0.491, + "step": 13786 + }, + { + "epoch": 4.840941011235955, + "grad_norm": 0.9915443658828735, + "learning_rate": 4.5153077582405615e-07, + "loss": 0.5433, + "step": 13787 + }, + { + "epoch": 4.841292134831461, + "grad_norm": 0.9745436906814575, + "learning_rate": 4.512657876318774e-07, + "loss": 0.5124, + "step": 13788 + }, + { + "epoch": 4.841643258426966, + "grad_norm": 0.985578715801239, + "learning_rate": 4.51000869504655e-07, + "loss": 0.5561, + "step": 13789 + }, + { + "epoch": 4.841994382022472, + "grad_norm": 1.0020419359207153, + "learning_rate": 4.507360214514503e-07, + "loss": 0.5469, + "step": 13790 + }, + { + "epoch": 4.842345505617978, + "grad_norm": 1.046934723854065, + "learning_rate": 4.5047124348131995e-07, + "loss": 0.5294, + "step": 13791 + }, + { + "epoch": 4.842696629213483, + "grad_norm": 0.9831732511520386, + "learning_rate": 4.5020653560331863e-07, + "loss": 0.5036, + "step": 13792 + }, + { + "epoch": 4.843047752808989, + "grad_norm": 0.9757646918296814, + "learning_rate": 4.499418978265005e-07, + "loss": 0.5548, + "step": 13793 + }, + { + "epoch": 4.843398876404494, + "grad_norm": 0.9905442595481873, + "learning_rate": 4.4967733015991446e-07, + "loss": 0.5566, + "step": 13794 + }, + { + "epoch": 4.84375, + "grad_norm": 0.980782687664032, + "learning_rate": 4.4941283261260986e-07, + "loss": 0.5556, + "step": 13795 + }, + { + "epoch": 4.844101123595506, + "grad_norm": 0.9820467233657837, + "learning_rate": 4.4914840519363094e-07, + "loss": 0.5305, + "step": 13796 + }, + { + "epoch": 4.844452247191011, + "grad_norm": 0.9715821743011475, + "learning_rate": 4.4888404791202266e-07, + "loss": 0.5454, + "step": 13797 + }, + { + "epoch": 4.844803370786517, + "grad_norm": 0.9560681581497192, + "learning_rate": 4.486197607768239e-07, + "loss": 0.5106, + "step": 13798 + }, + { + "epoch": 4.845154494382022, + "grad_norm": 1.00523042678833, + "learning_rate": 4.4835554379707437e-07, + "loss": 0.5315, + "step": 13799 + }, + { + "epoch": 4.845505617977528, + "grad_norm": 0.9645135402679443, + "learning_rate": 4.480913969818099e-07, + "loss": 0.5495, + "step": 13800 + }, + { + "epoch": 4.845856741573034, + "grad_norm": 1.017846703529358, + "learning_rate": 4.4782732034006295e-07, + "loss": 0.515, + "step": 13801 + }, + { + "epoch": 4.846207865168539, + "grad_norm": 0.9741557836532593, + "learning_rate": 4.475633138808663e-07, + "loss": 0.5149, + "step": 13802 + }, + { + "epoch": 4.846558988764045, + "grad_norm": 0.9518179893493652, + "learning_rate": 4.472993776132475e-07, + "loss": 0.5178, + "step": 13803 + }, + { + "epoch": 4.84691011235955, + "grad_norm": 0.9122090339660645, + "learning_rate": 4.470355115462341e-07, + "loss": 0.5182, + "step": 13804 + }, + { + "epoch": 4.847261235955056, + "grad_norm": 1.0055824518203735, + "learning_rate": 4.4677171568884875e-07, + "loss": 0.5221, + "step": 13805 + }, + { + "epoch": 4.847612359550562, + "grad_norm": 1.0172319412231445, + "learning_rate": 4.465079900501143e-07, + "loss": 0.5578, + "step": 13806 + }, + { + "epoch": 4.847963483146067, + "grad_norm": 0.9732767939567566, + "learning_rate": 4.4624433463904883e-07, + "loss": 0.5503, + "step": 13807 + }, + { + "epoch": 4.848314606741573, + "grad_norm": 0.9702237844467163, + "learning_rate": 4.4598074946467035e-07, + "loss": 0.538, + "step": 13808 + }, + { + "epoch": 4.848665730337078, + "grad_norm": 0.9809333682060242, + "learning_rate": 4.45717234535992e-07, + "loss": 0.5388, + "step": 13809 + }, + { + "epoch": 4.849016853932584, + "grad_norm": 0.9945096373558044, + "learning_rate": 4.45453789862027e-07, + "loss": 0.5509, + "step": 13810 + }, + { + "epoch": 4.84936797752809, + "grad_norm": 0.9329445958137512, + "learning_rate": 4.451904154517836e-07, + "loss": 0.5196, + "step": 13811 + }, + { + "epoch": 4.849719101123595, + "grad_norm": 0.9625335931777954, + "learning_rate": 4.449271113142703e-07, + "loss": 0.5256, + "step": 13812 + }, + { + "epoch": 4.850070224719101, + "grad_norm": 1.0091887712478638, + "learning_rate": 4.4466387745849104e-07, + "loss": 0.5545, + "step": 13813 + }, + { + "epoch": 4.850421348314606, + "grad_norm": 0.9560203552246094, + "learning_rate": 4.4440071389344805e-07, + "loss": 0.5145, + "step": 13814 + }, + { + "epoch": 4.850772471910112, + "grad_norm": 0.969840407371521, + "learning_rate": 4.4413762062814227e-07, + "loss": 0.5308, + "step": 13815 + }, + { + "epoch": 4.851123595505618, + "grad_norm": 0.9624013900756836, + "learning_rate": 4.438745976715697e-07, + "loss": 0.5464, + "step": 13816 + }, + { + "epoch": 4.851474719101123, + "grad_norm": 1.0225343704223633, + "learning_rate": 4.436116450327274e-07, + "loss": 0.5504, + "step": 13817 + }, + { + "epoch": 4.851825842696629, + "grad_norm": 0.9848880767822266, + "learning_rate": 4.433487627206065e-07, + "loss": 0.5575, + "step": 13818 + }, + { + "epoch": 4.852176966292134, + "grad_norm": 0.9533563256263733, + "learning_rate": 4.430859507441984e-07, + "loss": 0.5579, + "step": 13819 + }, + { + "epoch": 4.85252808988764, + "grad_norm": 0.9830746054649353, + "learning_rate": 4.428232091124904e-07, + "loss": 0.5131, + "step": 13820 + }, + { + "epoch": 4.852879213483146, + "grad_norm": 0.9772524833679199, + "learning_rate": 4.4256053783446854e-07, + "loss": 0.5575, + "step": 13821 + }, + { + "epoch": 4.853230337078652, + "grad_norm": 0.9994221329689026, + "learning_rate": 4.422979369191152e-07, + "loss": 0.5637, + "step": 13822 + }, + { + "epoch": 4.853581460674158, + "grad_norm": 0.9763613343238831, + "learning_rate": 4.420354063754123e-07, + "loss": 0.5394, + "step": 13823 + }, + { + "epoch": 4.853932584269663, + "grad_norm": 0.9574545621871948, + "learning_rate": 4.4177294621233767e-07, + "loss": 0.5586, + "step": 13824 + }, + { + "epoch": 4.854283707865169, + "grad_norm": 0.970962405204773, + "learning_rate": 4.415105564388661e-07, + "loss": 0.5281, + "step": 13825 + }, + { + "epoch": 4.854634831460674, + "grad_norm": 0.9636043906211853, + "learning_rate": 4.4124823706397287e-07, + "loss": 0.5066, + "step": 13826 + }, + { + "epoch": 4.85498595505618, + "grad_norm": 0.9442651271820068, + "learning_rate": 4.4098598809662744e-07, + "loss": 0.5006, + "step": 13827 + }, + { + "epoch": 4.855337078651686, + "grad_norm": 0.9591049551963806, + "learning_rate": 4.4072380954579996e-07, + "loss": 0.5671, + "step": 13828 + }, + { + "epoch": 4.855688202247191, + "grad_norm": 0.9569978713989258, + "learning_rate": 4.404617014204554e-07, + "loss": 0.5444, + "step": 13829 + }, + { + "epoch": 4.856039325842697, + "grad_norm": 0.9871239066123962, + "learning_rate": 4.401996637295589e-07, + "loss": 0.5637, + "step": 13830 + }, + { + "epoch": 4.856390449438202, + "grad_norm": 0.984786868095398, + "learning_rate": 4.399376964820706e-07, + "loss": 0.531, + "step": 13831 + }, + { + "epoch": 4.856741573033708, + "grad_norm": 0.9720375537872314, + "learning_rate": 4.3967579968695106e-07, + "loss": 0.525, + "step": 13832 + }, + { + "epoch": 4.857092696629214, + "grad_norm": 0.9723108410835266, + "learning_rate": 4.394139733531555e-07, + "loss": 0.5323, + "step": 13833 + }, + { + "epoch": 4.857443820224719, + "grad_norm": 0.9775556325912476, + "learning_rate": 4.3915221748963926e-07, + "loss": 0.5428, + "step": 13834 + }, + { + "epoch": 4.857794943820225, + "grad_norm": 1.0007296800613403, + "learning_rate": 4.388905321053533e-07, + "loss": 0.5608, + "step": 13835 + }, + { + "epoch": 4.85814606741573, + "grad_norm": 0.959686815738678, + "learning_rate": 4.3862891720924784e-07, + "loss": 0.5696, + "step": 13836 + }, + { + "epoch": 4.858497191011236, + "grad_norm": 0.9533791542053223, + "learning_rate": 4.383673728102694e-07, + "loss": 0.5125, + "step": 13837 + }, + { + "epoch": 4.858848314606742, + "grad_norm": 0.9557693600654602, + "learning_rate": 4.381058989173623e-07, + "loss": 0.5463, + "step": 13838 + }, + { + "epoch": 4.859199438202247, + "grad_norm": 0.9538478255271912, + "learning_rate": 4.378444955394695e-07, + "loss": 0.5379, + "step": 13839 + }, + { + "epoch": 4.859550561797753, + "grad_norm": 0.9438117742538452, + "learning_rate": 4.375831626855298e-07, + "loss": 0.5253, + "step": 13840 + }, + { + "epoch": 4.859901685393258, + "grad_norm": 0.9102749824523926, + "learning_rate": 4.3732190036448157e-07, + "loss": 0.4881, + "step": 13841 + }, + { + "epoch": 4.860252808988764, + "grad_norm": 0.9450820088386536, + "learning_rate": 4.370607085852588e-07, + "loss": 0.5249, + "step": 13842 + }, + { + "epoch": 4.86060393258427, + "grad_norm": 0.9308556914329529, + "learning_rate": 4.3679958735679514e-07, + "loss": 0.5488, + "step": 13843 + }, + { + "epoch": 4.860955056179775, + "grad_norm": 0.9534568190574646, + "learning_rate": 4.3653853668801956e-07, + "loss": 0.5377, + "step": 13844 + }, + { + "epoch": 4.861306179775281, + "grad_norm": 0.9743516445159912, + "learning_rate": 4.3627755658786073e-07, + "loss": 0.5304, + "step": 13845 + }, + { + "epoch": 4.861657303370786, + "grad_norm": 0.9498743414878845, + "learning_rate": 4.36016647065243e-07, + "loss": 0.5194, + "step": 13846 + }, + { + "epoch": 4.862008426966292, + "grad_norm": 0.9450740218162537, + "learning_rate": 4.357558081290902e-07, + "loss": 0.5183, + "step": 13847 + }, + { + "epoch": 4.862359550561798, + "grad_norm": 0.9771822690963745, + "learning_rate": 4.3549503978832203e-07, + "loss": 0.5748, + "step": 13848 + }, + { + "epoch": 4.862710674157303, + "grad_norm": 0.968194842338562, + "learning_rate": 4.352343420518573e-07, + "loss": 0.5236, + "step": 13849 + }, + { + "epoch": 4.863061797752809, + "grad_norm": 0.9347674250602722, + "learning_rate": 4.3497371492861123e-07, + "loss": 0.5099, + "step": 13850 + }, + { + "epoch": 4.863412921348314, + "grad_norm": 0.9520711898803711, + "learning_rate": 4.347131584274966e-07, + "loss": 0.5453, + "step": 13851 + }, + { + "epoch": 4.86376404494382, + "grad_norm": 0.9590968489646912, + "learning_rate": 4.344526725574252e-07, + "loss": 0.5328, + "step": 13852 + }, + { + "epoch": 4.864115168539326, + "grad_norm": 0.9699344635009766, + "learning_rate": 4.3419225732730415e-07, + "loss": 0.521, + "step": 13853 + }, + { + "epoch": 4.864466292134831, + "grad_norm": 0.9906556010246277, + "learning_rate": 4.339319127460409e-07, + "loss": 0.5603, + "step": 13854 + }, + { + "epoch": 4.864817415730337, + "grad_norm": 0.9141637682914734, + "learning_rate": 4.336716388225376e-07, + "loss": 0.5051, + "step": 13855 + }, + { + "epoch": 4.865168539325842, + "grad_norm": 0.9674749970436096, + "learning_rate": 4.334114355656968e-07, + "loss": 0.5334, + "step": 13856 + }, + { + "epoch": 4.865519662921348, + "grad_norm": 0.9594097137451172, + "learning_rate": 4.331513029844156e-07, + "loss": 0.5063, + "step": 13857 + }, + { + "epoch": 4.865870786516854, + "grad_norm": 0.9704673886299133, + "learning_rate": 4.328912410875921e-07, + "loss": 0.5143, + "step": 13858 + }, + { + "epoch": 4.86622191011236, + "grad_norm": 0.9658549427986145, + "learning_rate": 4.326312498841187e-07, + "loss": 0.5475, + "step": 13859 + }, + { + "epoch": 4.866573033707866, + "grad_norm": 0.9522914290428162, + "learning_rate": 4.3237132938288796e-07, + "loss": 0.4978, + "step": 13860 + }, + { + "epoch": 4.866924157303371, + "grad_norm": 0.9331061840057373, + "learning_rate": 4.3211147959278774e-07, + "loss": 0.4851, + "step": 13861 + }, + { + "epoch": 4.867275280898877, + "grad_norm": 0.9851291179656982, + "learning_rate": 4.318517005227063e-07, + "loss": 0.5585, + "step": 13862 + }, + { + "epoch": 4.867626404494382, + "grad_norm": 0.9661496877670288, + "learning_rate": 4.3159199218152677e-07, + "loss": 0.5044, + "step": 13863 + }, + { + "epoch": 4.867977528089888, + "grad_norm": 0.9857163429260254, + "learning_rate": 4.313323545781306e-07, + "loss": 0.5338, + "step": 13864 + }, + { + "epoch": 4.868328651685394, + "grad_norm": 0.969819188117981, + "learning_rate": 4.310727877213983e-07, + "loss": 0.5404, + "step": 13865 + }, + { + "epoch": 4.868679775280899, + "grad_norm": 0.9746801853179932, + "learning_rate": 4.308132916202054e-07, + "loss": 0.5374, + "step": 13866 + }, + { + "epoch": 4.869030898876405, + "grad_norm": 0.9699353575706482, + "learning_rate": 4.305538662834283e-07, + "loss": 0.5067, + "step": 13867 + }, + { + "epoch": 4.86938202247191, + "grad_norm": 0.9842150807380676, + "learning_rate": 4.3029451171993727e-07, + "loss": 0.5314, + "step": 13868 + }, + { + "epoch": 4.869733146067416, + "grad_norm": 1.002665638923645, + "learning_rate": 4.3003522793860356e-07, + "loss": 0.5394, + "step": 13869 + }, + { + "epoch": 4.870084269662922, + "grad_norm": 0.9431537389755249, + "learning_rate": 4.297760149482932e-07, + "loss": 0.5372, + "step": 13870 + }, + { + "epoch": 4.870435393258427, + "grad_norm": 1.0049563646316528, + "learning_rate": 4.295168727578722e-07, + "loss": 0.5501, + "step": 13871 + }, + { + "epoch": 4.870786516853933, + "grad_norm": 0.9506394863128662, + "learning_rate": 4.2925780137620194e-07, + "loss": 0.5188, + "step": 13872 + }, + { + "epoch": 4.871137640449438, + "grad_norm": 0.9875645637512207, + "learning_rate": 4.2899880081214316e-07, + "loss": 0.5364, + "step": 13873 + }, + { + "epoch": 4.871488764044944, + "grad_norm": 0.9581896662712097, + "learning_rate": 4.28739871074553e-07, + "loss": 0.5036, + "step": 13874 + }, + { + "epoch": 4.87183988764045, + "grad_norm": 0.9934858679771423, + "learning_rate": 4.284810121722871e-07, + "loss": 0.5699, + "step": 13875 + }, + { + "epoch": 4.872191011235955, + "grad_norm": 1.0033314228057861, + "learning_rate": 4.282222241141984e-07, + "loss": 0.5577, + "step": 13876 + }, + { + "epoch": 4.872542134831461, + "grad_norm": 1.0099718570709229, + "learning_rate": 4.279635069091359e-07, + "loss": 0.5276, + "step": 13877 + }, + { + "epoch": 4.872893258426966, + "grad_norm": 1.014980435371399, + "learning_rate": 4.27704860565949e-07, + "loss": 0.531, + "step": 13878 + }, + { + "epoch": 4.873244382022472, + "grad_norm": 0.9335871338844299, + "learning_rate": 4.2744628509348184e-07, + "loss": 0.5112, + "step": 13879 + }, + { + "epoch": 4.873595505617978, + "grad_norm": 0.9997422695159912, + "learning_rate": 4.271877805005789e-07, + "loss": 0.5297, + "step": 13880 + }, + { + "epoch": 4.873946629213483, + "grad_norm": 0.9704664945602417, + "learning_rate": 4.269293467960797e-07, + "loss": 0.5308, + "step": 13881 + }, + { + "epoch": 4.874297752808989, + "grad_norm": 0.9520581960678101, + "learning_rate": 4.2667098398882336e-07, + "loss": 0.5368, + "step": 13882 + }, + { + "epoch": 4.874648876404494, + "grad_norm": 0.9754824638366699, + "learning_rate": 4.264126920876446e-07, + "loss": 0.5228, + "step": 13883 + }, + { + "epoch": 4.875, + "grad_norm": 0.9592466950416565, + "learning_rate": 4.261544711013782e-07, + "loss": 0.5359, + "step": 13884 + }, + { + "epoch": 4.875351123595506, + "grad_norm": 0.9922242760658264, + "learning_rate": 4.2589632103885337e-07, + "loss": 0.547, + "step": 13885 + }, + { + "epoch": 4.875702247191011, + "grad_norm": 0.9588804244995117, + "learning_rate": 4.256382419089003e-07, + "loss": 0.5302, + "step": 13886 + }, + { + "epoch": 4.876053370786517, + "grad_norm": 0.990556001663208, + "learning_rate": 4.2538023372034415e-07, + "loss": 0.591, + "step": 13887 + }, + { + "epoch": 4.876404494382022, + "grad_norm": 1.0243722200393677, + "learning_rate": 4.251222964820081e-07, + "loss": 0.5222, + "step": 13888 + }, + { + "epoch": 4.876755617977528, + "grad_norm": 0.9843601584434509, + "learning_rate": 4.248644302027149e-07, + "loss": 0.5486, + "step": 13889 + }, + { + "epoch": 4.877106741573034, + "grad_norm": 0.9945630431175232, + "learning_rate": 4.246066348912817e-07, + "loss": 0.5373, + "step": 13890 + }, + { + "epoch": 4.877457865168539, + "grad_norm": 0.9573256969451904, + "learning_rate": 4.243489105565263e-07, + "loss": 0.5308, + "step": 13891 + }, + { + "epoch": 4.877808988764045, + "grad_norm": 0.9691802859306335, + "learning_rate": 4.2409125720726143e-07, + "loss": 0.5609, + "step": 13892 + }, + { + "epoch": 4.87816011235955, + "grad_norm": 0.9944402575492859, + "learning_rate": 4.2383367485229947e-07, + "loss": 0.5657, + "step": 13893 + }, + { + "epoch": 4.878511235955056, + "grad_norm": 0.9956886172294617, + "learning_rate": 4.23576163500449e-07, + "loss": 0.5443, + "step": 13894 + }, + { + "epoch": 4.878862359550562, + "grad_norm": 0.9603670239448547, + "learning_rate": 4.233187231605174e-07, + "loss": 0.551, + "step": 13895 + }, + { + "epoch": 4.879213483146067, + "grad_norm": 0.9782986640930176, + "learning_rate": 4.230613538413078e-07, + "loss": 0.5786, + "step": 13896 + }, + { + "epoch": 4.879564606741573, + "grad_norm": 0.9787992835044861, + "learning_rate": 4.22804055551623e-07, + "loss": 0.5443, + "step": 13897 + }, + { + "epoch": 4.879915730337078, + "grad_norm": 0.9528709053993225, + "learning_rate": 4.2254682830026176e-07, + "loss": 0.5081, + "step": 13898 + }, + { + "epoch": 4.880266853932584, + "grad_norm": 0.9696153402328491, + "learning_rate": 4.222896720960218e-07, + "loss": 0.5515, + "step": 13899 + }, + { + "epoch": 4.88061797752809, + "grad_norm": 0.962026059627533, + "learning_rate": 4.2203258694769713e-07, + "loss": 0.5561, + "step": 13900 + }, + { + "epoch": 4.880969101123595, + "grad_norm": 0.9440157413482666, + "learning_rate": 4.217755728640793e-07, + "loss": 0.5245, + "step": 13901 + }, + { + "epoch": 4.881320224719101, + "grad_norm": 0.9419738054275513, + "learning_rate": 4.215186298539589e-07, + "loss": 0.5016, + "step": 13902 + }, + { + "epoch": 4.881671348314606, + "grad_norm": 0.9815879464149475, + "learning_rate": 4.2126175792612245e-07, + "loss": 0.4957, + "step": 13903 + }, + { + "epoch": 4.882022471910112, + "grad_norm": 0.9668583273887634, + "learning_rate": 4.210049570893554e-07, + "loss": 0.5368, + "step": 13904 + }, + { + "epoch": 4.882373595505618, + "grad_norm": 0.9664719700813293, + "learning_rate": 4.207482273524394e-07, + "loss": 0.5252, + "step": 13905 + }, + { + "epoch": 4.882724719101123, + "grad_norm": 0.9685222506523132, + "learning_rate": 4.2049156872415536e-07, + "loss": 0.5827, + "step": 13906 + }, + { + "epoch": 4.883075842696629, + "grad_norm": 0.9682361483573914, + "learning_rate": 4.2023498121327973e-07, + "loss": 0.5503, + "step": 13907 + }, + { + "epoch": 4.883426966292134, + "grad_norm": 0.956609845161438, + "learning_rate": 4.1997846482858845e-07, + "loss": 0.5032, + "step": 13908 + }, + { + "epoch": 4.88377808988764, + "grad_norm": 0.9811568856239319, + "learning_rate": 4.197220195788532e-07, + "loss": 0.5381, + "step": 13909 + }, + { + "epoch": 4.884129213483146, + "grad_norm": 0.9890134334564209, + "learning_rate": 4.1946564547284546e-07, + "loss": 0.5297, + "step": 13910 + }, + { + "epoch": 4.884480337078652, + "grad_norm": 0.9649168252944946, + "learning_rate": 4.1920934251933136e-07, + "loss": 0.5079, + "step": 13911 + }, + { + "epoch": 4.884831460674158, + "grad_norm": 1.0006872415542603, + "learning_rate": 4.1895311072707835e-07, + "loss": 0.6004, + "step": 13912 + }, + { + "epoch": 4.885182584269663, + "grad_norm": 0.9802055358886719, + "learning_rate": 4.1869695010484765e-07, + "loss": 0.5571, + "step": 13913 + }, + { + "epoch": 4.885533707865169, + "grad_norm": 1.0068734884262085, + "learning_rate": 4.184408606613999e-07, + "loss": 0.6178, + "step": 13914 + }, + { + "epoch": 4.885884831460674, + "grad_norm": 1.0156891345977783, + "learning_rate": 4.1818484240549436e-07, + "loss": 0.543, + "step": 13915 + }, + { + "epoch": 4.88623595505618, + "grad_norm": 0.9861239194869995, + "learning_rate": 4.179288953458849e-07, + "loss": 0.537, + "step": 13916 + }, + { + "epoch": 4.886587078651686, + "grad_norm": 1.0054622888565063, + "learning_rate": 4.1767301949132617e-07, + "loss": 0.5773, + "step": 13917 + }, + { + "epoch": 4.886938202247191, + "grad_norm": 0.9787366390228271, + "learning_rate": 4.1741721485056756e-07, + "loss": 0.5488, + "step": 13918 + }, + { + "epoch": 4.887289325842697, + "grad_norm": 0.9353874921798706, + "learning_rate": 4.171614814323588e-07, + "loss": 0.53, + "step": 13919 + }, + { + "epoch": 4.887640449438202, + "grad_norm": 0.9873981475830078, + "learning_rate": 4.1690581924544476e-07, + "loss": 0.5514, + "step": 13920 + }, + { + "epoch": 4.887991573033708, + "grad_norm": 0.9596479535102844, + "learning_rate": 4.166502282985693e-07, + "loss": 0.5455, + "step": 13921 + }, + { + "epoch": 4.888342696629214, + "grad_norm": 0.9554552435874939, + "learning_rate": 4.16394708600473e-07, + "loss": 0.5421, + "step": 13922 + }, + { + "epoch": 4.888693820224719, + "grad_norm": 0.9573599696159363, + "learning_rate": 4.1613926015989513e-07, + "loss": 0.5178, + "step": 13923 + }, + { + "epoch": 4.889044943820225, + "grad_norm": 0.9957705140113831, + "learning_rate": 4.1588388298557105e-07, + "loss": 0.5773, + "step": 13924 + }, + { + "epoch": 4.88939606741573, + "grad_norm": 0.9698472023010254, + "learning_rate": 4.1562857708623504e-07, + "loss": 0.5203, + "step": 13925 + }, + { + "epoch": 4.889747191011236, + "grad_norm": 0.946613609790802, + "learning_rate": 4.153733424706183e-07, + "loss": 0.5176, + "step": 13926 + }, + { + "epoch": 4.890098314606742, + "grad_norm": 0.968908429145813, + "learning_rate": 4.1511817914744873e-07, + "loss": 0.5087, + "step": 13927 + }, + { + "epoch": 4.890449438202247, + "grad_norm": 0.9514994621276855, + "learning_rate": 4.1486308712545386e-07, + "loss": 0.5472, + "step": 13928 + }, + { + "epoch": 4.890800561797753, + "grad_norm": 0.9502015709877014, + "learning_rate": 4.146080664133567e-07, + "loss": 0.5381, + "step": 13929 + }, + { + "epoch": 4.891151685393258, + "grad_norm": 0.9836103916168213, + "learning_rate": 4.1435311701987974e-07, + "loss": 0.5652, + "step": 13930 + }, + { + "epoch": 4.891502808988764, + "grad_norm": 0.9829474091529846, + "learning_rate": 4.1409823895374094e-07, + "loss": 0.547, + "step": 13931 + }, + { + "epoch": 4.89185393258427, + "grad_norm": 0.9790839552879333, + "learning_rate": 4.1384343222365785e-07, + "loss": 0.5357, + "step": 13932 + }, + { + "epoch": 4.892205056179775, + "grad_norm": 0.934490442276001, + "learning_rate": 4.1358869683834373e-07, + "loss": 0.5019, + "step": 13933 + }, + { + "epoch": 4.892556179775281, + "grad_norm": 0.9849702715873718, + "learning_rate": 4.1333403280651134e-07, + "loss": 0.5768, + "step": 13934 + }, + { + "epoch": 4.892907303370786, + "grad_norm": 1.0108697414398193, + "learning_rate": 4.130794401368696e-07, + "loss": 0.5303, + "step": 13935 + }, + { + "epoch": 4.893258426966292, + "grad_norm": 0.9637621641159058, + "learning_rate": 4.128249188381245e-07, + "loss": 0.5026, + "step": 13936 + }, + { + "epoch": 4.893609550561798, + "grad_norm": 0.9181874394416809, + "learning_rate": 4.125704689189819e-07, + "loss": 0.4893, + "step": 13937 + }, + { + "epoch": 4.893960674157303, + "grad_norm": 0.9479319453239441, + "learning_rate": 4.123160903881421e-07, + "loss": 0.5207, + "step": 13938 + }, + { + "epoch": 4.894311797752809, + "grad_norm": 0.9634531736373901, + "learning_rate": 4.120617832543064e-07, + "loss": 0.5197, + "step": 13939 + }, + { + "epoch": 4.894662921348314, + "grad_norm": 0.9770277738571167, + "learning_rate": 4.1180754752617044e-07, + "loss": 0.5837, + "step": 13940 + }, + { + "epoch": 4.89501404494382, + "grad_norm": 0.9554648399353027, + "learning_rate": 4.115533832124299e-07, + "loss": 0.5366, + "step": 13941 + }, + { + "epoch": 4.895365168539326, + "grad_norm": 0.9581807851791382, + "learning_rate": 4.1129929032177665e-07, + "loss": 0.5372, + "step": 13942 + }, + { + "epoch": 4.895716292134831, + "grad_norm": 1.005018711090088, + "learning_rate": 4.110452688628996e-07, + "loss": 0.5423, + "step": 13943 + }, + { + "epoch": 4.896067415730337, + "grad_norm": 0.9337010979652405, + "learning_rate": 4.1079131884448756e-07, + "loss": 0.513, + "step": 13944 + }, + { + "epoch": 4.896418539325842, + "grad_norm": 0.9477417469024658, + "learning_rate": 4.105374402752241e-07, + "loss": 0.5489, + "step": 13945 + }, + { + "epoch": 4.896769662921348, + "grad_norm": 0.9697373509407043, + "learning_rate": 4.102836331637927e-07, + "loss": 0.5805, + "step": 13946 + }, + { + "epoch": 4.897120786516854, + "grad_norm": 0.9527125358581543, + "learning_rate": 4.100298975188724e-07, + "loss": 0.5375, + "step": 13947 + }, + { + "epoch": 4.89747191011236, + "grad_norm": 0.9601466655731201, + "learning_rate": 4.0977623334914174e-07, + "loss": 0.5435, + "step": 13948 + }, + { + "epoch": 4.897823033707866, + "grad_norm": 0.9511587023735046, + "learning_rate": 4.095226406632752e-07, + "loss": 0.5275, + "step": 13949 + }, + { + "epoch": 4.898174157303371, + "grad_norm": 0.9676934480667114, + "learning_rate": 4.09269119469945e-07, + "loss": 0.5619, + "step": 13950 + }, + { + "epoch": 4.898525280898877, + "grad_norm": 0.9771845936775208, + "learning_rate": 4.090156697778222e-07, + "loss": 0.5393, + "step": 13951 + }, + { + "epoch": 4.898876404494382, + "grad_norm": 0.9738534092903137, + "learning_rate": 4.0876229159557387e-07, + "loss": 0.5407, + "step": 13952 + }, + { + "epoch": 4.899227528089888, + "grad_norm": 0.9602887034416199, + "learning_rate": 4.08508984931866e-07, + "loss": 0.5139, + "step": 13953 + }, + { + "epoch": 4.899578651685394, + "grad_norm": 0.9958669543266296, + "learning_rate": 4.082557497953607e-07, + "loss": 0.5823, + "step": 13954 + }, + { + "epoch": 4.899929775280899, + "grad_norm": 0.9797009825706482, + "learning_rate": 4.0800258619471954e-07, + "loss": 0.5187, + "step": 13955 + }, + { + "epoch": 4.900280898876405, + "grad_norm": 0.9602744579315186, + "learning_rate": 4.0774949413859984e-07, + "loss": 0.5409, + "step": 13956 + }, + { + "epoch": 4.90063202247191, + "grad_norm": 0.9727007150650024, + "learning_rate": 4.074964736356563e-07, + "loss": 0.5442, + "step": 13957 + }, + { + "epoch": 4.900983146067416, + "grad_norm": 0.9836077690124512, + "learning_rate": 4.072435246945433e-07, + "loss": 0.5709, + "step": 13958 + }, + { + "epoch": 4.901334269662922, + "grad_norm": 0.9927993416786194, + "learning_rate": 4.0699064732391047e-07, + "loss": 0.5959, + "step": 13959 + }, + { + "epoch": 4.901685393258427, + "grad_norm": 0.9547233581542969, + "learning_rate": 4.067378415324069e-07, + "loss": 0.5651, + "step": 13960 + }, + { + "epoch": 4.902036516853933, + "grad_norm": 0.9656669497489929, + "learning_rate": 4.064851073286777e-07, + "loss": 0.5168, + "step": 13961 + }, + { + "epoch": 4.902387640449438, + "grad_norm": 1.0139907598495483, + "learning_rate": 4.062324447213667e-07, + "loss": 0.5376, + "step": 13962 + }, + { + "epoch": 4.902738764044944, + "grad_norm": 0.9639943242073059, + "learning_rate": 4.0597985371911454e-07, + "loss": 0.5269, + "step": 13963 + }, + { + "epoch": 4.90308988764045, + "grad_norm": 0.9475921988487244, + "learning_rate": 4.057273343305587e-07, + "loss": 0.5395, + "step": 13964 + }, + { + "epoch": 4.903441011235955, + "grad_norm": 0.9717211723327637, + "learning_rate": 4.054748865643368e-07, + "loss": 0.5331, + "step": 13965 + }, + { + "epoch": 4.903792134831461, + "grad_norm": 0.9735040664672852, + "learning_rate": 4.0522251042908076e-07, + "loss": 0.5397, + "step": 13966 + }, + { + "epoch": 4.904143258426966, + "grad_norm": 0.9963889122009277, + "learning_rate": 4.049702059334229e-07, + "loss": 0.5916, + "step": 13967 + }, + { + "epoch": 4.904494382022472, + "grad_norm": 0.9712895154953003, + "learning_rate": 4.047179730859907e-07, + "loss": 0.5651, + "step": 13968 + }, + { + "epoch": 4.904845505617978, + "grad_norm": 0.9406970143318176, + "learning_rate": 4.0446581189541147e-07, + "loss": 0.5573, + "step": 13969 + }, + { + "epoch": 4.905196629213483, + "grad_norm": 0.9794761538505554, + "learning_rate": 4.042137223703077e-07, + "loss": 0.5733, + "step": 13970 + }, + { + "epoch": 4.905547752808989, + "grad_norm": 0.9956604242324829, + "learning_rate": 4.039617045193017e-07, + "loss": 0.5513, + "step": 13971 + }, + { + "epoch": 4.905898876404494, + "grad_norm": 0.9707120656967163, + "learning_rate": 4.0370975835101133e-07, + "loss": 0.5319, + "step": 13972 + }, + { + "epoch": 4.90625, + "grad_norm": 0.9721235632896423, + "learning_rate": 4.034578838740541e-07, + "loss": 0.5451, + "step": 13973 + }, + { + "epoch": 4.906601123595506, + "grad_norm": 0.968963623046875, + "learning_rate": 4.0320608109704255e-07, + "loss": 0.5163, + "step": 13974 + }, + { + "epoch": 4.906952247191011, + "grad_norm": 0.9734437465667725, + "learning_rate": 4.029543500285893e-07, + "loss": 0.5329, + "step": 13975 + }, + { + "epoch": 4.907303370786517, + "grad_norm": 1.014728307723999, + "learning_rate": 4.0270269067730294e-07, + "loss": 0.5577, + "step": 13976 + }, + { + "epoch": 4.907654494382022, + "grad_norm": 0.9902660250663757, + "learning_rate": 4.024511030517891e-07, + "loss": 0.5545, + "step": 13977 + }, + { + "epoch": 4.908005617977528, + "grad_norm": 0.9565796256065369, + "learning_rate": 4.021995871606532e-07, + "loss": 0.5493, + "step": 13978 + }, + { + "epoch": 4.908356741573034, + "grad_norm": 0.9387690424919128, + "learning_rate": 4.019481430124958e-07, + "loss": 0.5344, + "step": 13979 + }, + { + "epoch": 4.908707865168539, + "grad_norm": 0.9650492668151855, + "learning_rate": 4.0169677061591724e-07, + "loss": 0.5383, + "step": 13980 + }, + { + "epoch": 4.909058988764045, + "grad_norm": 0.9879761338233948, + "learning_rate": 4.014454699795131e-07, + "loss": 0.5343, + "step": 13981 + }, + { + "epoch": 4.90941011235955, + "grad_norm": 0.9763932228088379, + "learning_rate": 4.0119424111187853e-07, + "loss": 0.5561, + "step": 13982 + }, + { + "epoch": 4.909761235955056, + "grad_norm": 0.9760566353797913, + "learning_rate": 4.0094308402160443e-07, + "loss": 0.5554, + "step": 13983 + }, + { + "epoch": 4.910112359550562, + "grad_norm": 0.9669250845909119, + "learning_rate": 4.0069199871728107e-07, + "loss": 0.5199, + "step": 13984 + }, + { + "epoch": 4.910463483146067, + "grad_norm": 0.9624865651130676, + "learning_rate": 4.0044098520749446e-07, + "loss": 0.5873, + "step": 13985 + }, + { + "epoch": 4.910814606741573, + "grad_norm": 0.9590842127799988, + "learning_rate": 4.001900435008302e-07, + "loss": 0.5323, + "step": 13986 + }, + { + "epoch": 4.911165730337078, + "grad_norm": 0.9761930704116821, + "learning_rate": 3.999391736058689e-07, + "loss": 0.5253, + "step": 13987 + }, + { + "epoch": 4.911516853932584, + "grad_norm": 0.9490305185317993, + "learning_rate": 3.996883755311917e-07, + "loss": 0.5163, + "step": 13988 + }, + { + "epoch": 4.91186797752809, + "grad_norm": 0.9677528142929077, + "learning_rate": 3.994376492853744e-07, + "loss": 0.5799, + "step": 13989 + }, + { + "epoch": 4.912219101123595, + "grad_norm": 0.9654319882392883, + "learning_rate": 3.991869948769916e-07, + "loss": 0.5368, + "step": 13990 + }, + { + "epoch": 4.912570224719101, + "grad_norm": 0.9650664925575256, + "learning_rate": 3.989364123146167e-07, + "loss": 0.5625, + "step": 13991 + }, + { + "epoch": 4.912921348314606, + "grad_norm": 0.9692695140838623, + "learning_rate": 3.9868590160681757e-07, + "loss": 0.5411, + "step": 13992 + }, + { + "epoch": 4.913272471910112, + "grad_norm": 0.9447038769721985, + "learning_rate": 3.984354627621634e-07, + "loss": 0.5441, + "step": 13993 + }, + { + "epoch": 4.913623595505618, + "grad_norm": 0.9881811738014221, + "learning_rate": 3.981850957892175e-07, + "loss": 0.5591, + "step": 13994 + }, + { + "epoch": 4.913974719101123, + "grad_norm": 1.067888617515564, + "learning_rate": 3.9793480069654305e-07, + "loss": 0.5726, + "step": 13995 + }, + { + "epoch": 4.914325842696629, + "grad_norm": 0.9648988842964172, + "learning_rate": 3.9768457749269935e-07, + "loss": 0.5611, + "step": 13996 + }, + { + "epoch": 4.914676966292134, + "grad_norm": 0.9906226396560669, + "learning_rate": 3.974344261862445e-07, + "loss": 0.537, + "step": 13997 + }, + { + "epoch": 4.91502808988764, + "grad_norm": 0.9364917874336243, + "learning_rate": 3.971843467857328e-07, + "loss": 0.5419, + "step": 13998 + }, + { + "epoch": 4.915379213483146, + "grad_norm": 0.9761050343513489, + "learning_rate": 3.9693433929971743e-07, + "loss": 0.5098, + "step": 13999 + }, + { + "epoch": 4.915730337078652, + "grad_norm": 0.9704265594482422, + "learning_rate": 3.966844037367473e-07, + "loss": 0.5506, + "step": 14000 + }, + { + "epoch": 4.916081460674158, + "grad_norm": 0.9494706392288208, + "learning_rate": 3.9643454010537155e-07, + "loss": 0.5158, + "step": 14001 + }, + { + "epoch": 4.916432584269663, + "grad_norm": 0.9753944277763367, + "learning_rate": 3.961847484141346e-07, + "loss": 0.5493, + "step": 14002 + }, + { + "epoch": 4.916783707865169, + "grad_norm": 0.9612893462181091, + "learning_rate": 3.9593502867157796e-07, + "loss": 0.5467, + "step": 14003 + }, + { + "epoch": 4.917134831460674, + "grad_norm": 1.01349937915802, + "learning_rate": 3.9568538088624373e-07, + "loss": 0.5511, + "step": 14004 + }, + { + "epoch": 4.91748595505618, + "grad_norm": 0.9375382661819458, + "learning_rate": 3.954358050666682e-07, + "loss": 0.5514, + "step": 14005 + }, + { + "epoch": 4.917837078651686, + "grad_norm": 0.9505131840705872, + "learning_rate": 3.9518630122138774e-07, + "loss": 0.5177, + "step": 14006 + }, + { + "epoch": 4.918188202247191, + "grad_norm": 0.9985142946243286, + "learning_rate": 3.949368693589342e-07, + "loss": 0.5274, + "step": 14007 + }, + { + "epoch": 4.918539325842697, + "grad_norm": 0.9830992221832275, + "learning_rate": 3.9468750948783884e-07, + "loss": 0.557, + "step": 14008 + }, + { + "epoch": 4.918890449438202, + "grad_norm": 0.9737469553947449, + "learning_rate": 3.944382216166287e-07, + "loss": 0.5107, + "step": 14009 + }, + { + "epoch": 4.919241573033708, + "grad_norm": 0.9706385135650635, + "learning_rate": 3.9418900575383e-07, + "loss": 0.5492, + "step": 14010 + }, + { + "epoch": 4.919592696629214, + "grad_norm": 0.9516773223876953, + "learning_rate": 3.9393986190796513e-07, + "loss": 0.5483, + "step": 14011 + }, + { + "epoch": 4.919943820224719, + "grad_norm": 0.9694939255714417, + "learning_rate": 3.93690790087555e-07, + "loss": 0.5295, + "step": 14012 + }, + { + "epoch": 4.920294943820225, + "grad_norm": 1.006238341331482, + "learning_rate": 3.93441790301117e-07, + "loss": 0.5287, + "step": 14013 + }, + { + "epoch": 4.92064606741573, + "grad_norm": 0.931037187576294, + "learning_rate": 3.931928625571677e-07, + "loss": 0.5317, + "step": 14014 + }, + { + "epoch": 4.920997191011236, + "grad_norm": 0.9606152772903442, + "learning_rate": 3.929440068642199e-07, + "loss": 0.5231, + "step": 14015 + }, + { + "epoch": 4.921348314606742, + "grad_norm": 0.9679231643676758, + "learning_rate": 3.926952232307832e-07, + "loss": 0.5433, + "step": 14016 + }, + { + "epoch": 4.921699438202247, + "grad_norm": 0.9856120944023132, + "learning_rate": 3.9244651166536724e-07, + "loss": 0.5446, + "step": 14017 + }, + { + "epoch": 4.922050561797753, + "grad_norm": 0.9580026865005493, + "learning_rate": 3.921978721764766e-07, + "loss": 0.4999, + "step": 14018 + }, + { + "epoch": 4.922401685393258, + "grad_norm": 0.9761483073234558, + "learning_rate": 3.9194930477261557e-07, + "loss": 0.5578, + "step": 14019 + }, + { + "epoch": 4.922752808988764, + "grad_norm": 0.967355489730835, + "learning_rate": 3.917008094622837e-07, + "loss": 0.5498, + "step": 14020 + }, + { + "epoch": 4.92310393258427, + "grad_norm": 0.9449467062950134, + "learning_rate": 3.91452386253981e-07, + "loss": 0.5373, + "step": 14021 + }, + { + "epoch": 4.923455056179775, + "grad_norm": 0.9690815806388855, + "learning_rate": 3.912040351562016e-07, + "loss": 0.5076, + "step": 14022 + }, + { + "epoch": 4.923806179775281, + "grad_norm": 0.97903972864151, + "learning_rate": 3.909557561774402e-07, + "loss": 0.5223, + "step": 14023 + }, + { + "epoch": 4.924157303370786, + "grad_norm": 0.9880831241607666, + "learning_rate": 3.9070754932618666e-07, + "loss": 0.5346, + "step": 14024 + }, + { + "epoch": 4.924508426966292, + "grad_norm": 0.9765170216560364, + "learning_rate": 3.9045941461093083e-07, + "loss": 0.5386, + "step": 14025 + }, + { + "epoch": 4.924859550561798, + "grad_norm": 0.9714359045028687, + "learning_rate": 3.9021135204015765e-07, + "loss": 0.5382, + "step": 14026 + }, + { + "epoch": 4.925210674157303, + "grad_norm": 0.9536328911781311, + "learning_rate": 3.8996336162235025e-07, + "loss": 0.5253, + "step": 14027 + }, + { + "epoch": 4.925561797752809, + "grad_norm": 0.9869883060455322, + "learning_rate": 3.897154433659908e-07, + "loss": 0.5594, + "step": 14028 + }, + { + "epoch": 4.925912921348314, + "grad_norm": 0.963158130645752, + "learning_rate": 3.8946759727955695e-07, + "loss": 0.5758, + "step": 14029 + }, + { + "epoch": 4.92626404494382, + "grad_norm": 1.0001411437988281, + "learning_rate": 3.892198233715258e-07, + "loss": 0.5293, + "step": 14030 + }, + { + "epoch": 4.926615168539326, + "grad_norm": 0.9750561118125916, + "learning_rate": 3.889721216503697e-07, + "loss": 0.5536, + "step": 14031 + }, + { + "epoch": 4.926966292134831, + "grad_norm": 0.94572514295578, + "learning_rate": 3.8872449212456143e-07, + "loss": 0.5098, + "step": 14032 + }, + { + "epoch": 4.927317415730337, + "grad_norm": 0.9577956795692444, + "learning_rate": 3.8847693480256826e-07, + "loss": 0.5216, + "step": 14033 + }, + { + "epoch": 4.927668539325842, + "grad_norm": 0.9728027582168579, + "learning_rate": 3.8822944969285764e-07, + "loss": 0.5385, + "step": 14034 + }, + { + "epoch": 4.928019662921348, + "grad_norm": 0.9888172745704651, + "learning_rate": 3.879820368038922e-07, + "loss": 0.5327, + "step": 14035 + }, + { + "epoch": 4.928370786516854, + "grad_norm": 0.9695999026298523, + "learning_rate": 3.8773469614413444e-07, + "loss": 0.5564, + "step": 14036 + }, + { + "epoch": 4.92872191011236, + "grad_norm": 0.9851166605949402, + "learning_rate": 3.8748742772204193e-07, + "loss": 0.5612, + "step": 14037 + }, + { + "epoch": 4.929073033707866, + "grad_norm": 0.9451625943183899, + "learning_rate": 3.872402315460724e-07, + "loss": 0.5435, + "step": 14038 + }, + { + "epoch": 4.929424157303371, + "grad_norm": 1.0321097373962402, + "learning_rate": 3.8699310762467885e-07, + "loss": 0.526, + "step": 14039 + }, + { + "epoch": 4.929775280898877, + "grad_norm": 0.9877265691757202, + "learning_rate": 3.8674605596631247e-07, + "loss": 0.5477, + "step": 14040 + }, + { + "epoch": 4.930126404494382, + "grad_norm": 0.9770605564117432, + "learning_rate": 3.86499076579423e-07, + "loss": 0.5297, + "step": 14041 + }, + { + "epoch": 4.930477528089888, + "grad_norm": 0.9721372127532959, + "learning_rate": 3.8625216947245607e-07, + "loss": 0.523, + "step": 14042 + }, + { + "epoch": 4.930828651685394, + "grad_norm": 0.9624738097190857, + "learning_rate": 3.86005334653857e-07, + "loss": 0.5409, + "step": 14043 + }, + { + "epoch": 4.931179775280899, + "grad_norm": 0.9515531063079834, + "learning_rate": 3.8575857213206555e-07, + "loss": 0.5042, + "step": 14044 + }, + { + "epoch": 4.931530898876405, + "grad_norm": 0.9918547868728638, + "learning_rate": 3.855118819155226e-07, + "loss": 0.5137, + "step": 14045 + }, + { + "epoch": 4.93188202247191, + "grad_norm": 0.9708138704299927, + "learning_rate": 3.8526526401266334e-07, + "loss": 0.5495, + "step": 14046 + }, + { + "epoch": 4.932233146067416, + "grad_norm": 0.9579374194145203, + "learning_rate": 3.850187184319229e-07, + "loss": 0.527, + "step": 14047 + }, + { + "epoch": 4.932584269662922, + "grad_norm": 0.9874399304389954, + "learning_rate": 3.847722451817318e-07, + "loss": 0.5357, + "step": 14048 + }, + { + "epoch": 4.932935393258427, + "grad_norm": 0.9475987553596497, + "learning_rate": 3.8452584427052075e-07, + "loss": 0.5044, + "step": 14049 + }, + { + "epoch": 4.933286516853933, + "grad_norm": 1.0095301866531372, + "learning_rate": 3.842795157067147e-07, + "loss": 0.5687, + "step": 14050 + }, + { + "epoch": 4.933637640449438, + "grad_norm": 1.040771245956421, + "learning_rate": 3.840332594987395e-07, + "loss": 0.5582, + "step": 14051 + }, + { + "epoch": 4.933988764044944, + "grad_norm": 0.9565936326980591, + "learning_rate": 3.8378707565501635e-07, + "loss": 0.5597, + "step": 14052 + }, + { + "epoch": 4.93433988764045, + "grad_norm": 0.9888427257537842, + "learning_rate": 3.8354096418396355e-07, + "loss": 0.5572, + "step": 14053 + }, + { + "epoch": 4.934691011235955, + "grad_norm": 1.0276050567626953, + "learning_rate": 3.8329492509399933e-07, + "loss": 0.5606, + "step": 14054 + }, + { + "epoch": 4.935042134831461, + "grad_norm": 0.971622884273529, + "learning_rate": 3.8304895839353676e-07, + "loss": 0.5379, + "step": 14055 + }, + { + "epoch": 4.935393258426966, + "grad_norm": 0.9511338472366333, + "learning_rate": 3.828030640909888e-07, + "loss": 0.5217, + "step": 14056 + }, + { + "epoch": 4.935744382022472, + "grad_norm": 0.9792309999465942, + "learning_rate": 3.8255724219476375e-07, + "loss": 0.5134, + "step": 14057 + }, + { + "epoch": 4.936095505617978, + "grad_norm": 0.9811505675315857, + "learning_rate": 3.823114927132698e-07, + "loss": 0.5646, + "step": 14058 + }, + { + "epoch": 4.936446629213483, + "grad_norm": 0.9493322372436523, + "learning_rate": 3.820658156549101e-07, + "loss": 0.5269, + "step": 14059 + }, + { + "epoch": 4.936797752808989, + "grad_norm": 0.9379968643188477, + "learning_rate": 3.8182021102808776e-07, + "loss": 0.5199, + "step": 14060 + }, + { + "epoch": 4.937148876404494, + "grad_norm": 0.9877141714096069, + "learning_rate": 3.8157467884120115e-07, + "loss": 0.5345, + "step": 14061 + }, + { + "epoch": 4.9375, + "grad_norm": 0.9479098916053772, + "learning_rate": 3.8132921910264797e-07, + "loss": 0.5095, + "step": 14062 + }, + { + "epoch": 4.937851123595506, + "grad_norm": 0.9529081583023071, + "learning_rate": 3.8108383182082236e-07, + "loss": 0.5151, + "step": 14063 + }, + { + "epoch": 4.938202247191011, + "grad_norm": 0.9615855813026428, + "learning_rate": 3.8083851700411677e-07, + "loss": 0.5574, + "step": 14064 + }, + { + "epoch": 4.938553370786517, + "grad_norm": 1.0054198503494263, + "learning_rate": 3.8059327466092087e-07, + "loss": 0.546, + "step": 14065 + }, + { + "epoch": 4.938904494382022, + "grad_norm": 0.971982479095459, + "learning_rate": 3.803481047996205e-07, + "loss": 0.5384, + "step": 14066 + }, + { + "epoch": 4.939255617977528, + "grad_norm": 0.9739723205566406, + "learning_rate": 3.801030074286019e-07, + "loss": 0.5662, + "step": 14067 + }, + { + "epoch": 4.939606741573034, + "grad_norm": 0.994340717792511, + "learning_rate": 3.798579825562457e-07, + "loss": 0.538, + "step": 14068 + }, + { + "epoch": 4.939957865168539, + "grad_norm": 0.9558712244033813, + "learning_rate": 3.7961303019093285e-07, + "loss": 0.5251, + "step": 14069 + }, + { + "epoch": 4.940308988764045, + "grad_norm": 0.9394634366035461, + "learning_rate": 3.793681503410393e-07, + "loss": 0.5223, + "step": 14070 + }, + { + "epoch": 4.94066011235955, + "grad_norm": 0.9466896653175354, + "learning_rate": 3.79123343014941e-07, + "loss": 0.5482, + "step": 14071 + }, + { + "epoch": 4.941011235955056, + "grad_norm": 0.9790592193603516, + "learning_rate": 3.7887860822100916e-07, + "loss": 0.5712, + "step": 14072 + }, + { + "epoch": 4.941362359550562, + "grad_norm": 0.9659919738769531, + "learning_rate": 3.7863394596761414e-07, + "loss": 0.5634, + "step": 14073 + }, + { + "epoch": 4.941713483146067, + "grad_norm": 0.9874322414398193, + "learning_rate": 3.7838935626312246e-07, + "loss": 0.5554, + "step": 14074 + }, + { + "epoch": 4.942064606741573, + "grad_norm": 0.978175163269043, + "learning_rate": 3.7814483911589984e-07, + "loss": 0.554, + "step": 14075 + }, + { + "epoch": 4.942415730337078, + "grad_norm": 0.9861769676208496, + "learning_rate": 3.779003945343074e-07, + "loss": 0.538, + "step": 14076 + }, + { + "epoch": 4.942766853932584, + "grad_norm": 0.9993933439254761, + "learning_rate": 3.776560225267062e-07, + "loss": 0.573, + "step": 14077 + }, + { + "epoch": 4.94311797752809, + "grad_norm": 0.9946343302726746, + "learning_rate": 3.774117231014529e-07, + "loss": 0.56, + "step": 14078 + }, + { + "epoch": 4.943469101123595, + "grad_norm": 0.9816747307777405, + "learning_rate": 3.771674962669017e-07, + "loss": 0.5421, + "step": 14079 + }, + { + "epoch": 4.943820224719101, + "grad_norm": 0.9613109230995178, + "learning_rate": 3.7692334203140614e-07, + "loss": 0.5558, + "step": 14080 + }, + { + "epoch": 4.944171348314606, + "grad_norm": 0.9610997438430786, + "learning_rate": 3.766792604033151e-07, + "loss": 0.536, + "step": 14081 + }, + { + "epoch": 4.944522471910112, + "grad_norm": 0.994009792804718, + "learning_rate": 3.7643525139097694e-07, + "loss": 0.5783, + "step": 14082 + }, + { + "epoch": 4.944873595505618, + "grad_norm": 1.0385664701461792, + "learning_rate": 3.761913150027355e-07, + "loss": 0.5778, + "step": 14083 + }, + { + "epoch": 4.945224719101123, + "grad_norm": 0.9195559024810791, + "learning_rate": 3.759474512469341e-07, + "loss": 0.4907, + "step": 14084 + }, + { + "epoch": 4.945575842696629, + "grad_norm": 0.9642199873924255, + "learning_rate": 3.7570366013191167e-07, + "loss": 0.5439, + "step": 14085 + }, + { + "epoch": 4.945926966292134, + "grad_norm": 0.9908018708229065, + "learning_rate": 3.754599416660071e-07, + "loss": 0.532, + "step": 14086 + }, + { + "epoch": 4.94627808988764, + "grad_norm": 0.9647195935249329, + "learning_rate": 3.7521629585755394e-07, + "loss": 0.5028, + "step": 14087 + }, + { + "epoch": 4.946629213483146, + "grad_norm": 0.9858996272087097, + "learning_rate": 3.749727227148858e-07, + "loss": 0.5632, + "step": 14088 + }, + { + "epoch": 4.946980337078652, + "grad_norm": 0.9874307513237, + "learning_rate": 3.7472922224633196e-07, + "loss": 0.5485, + "step": 14089 + }, + { + "epoch": 4.947331460674158, + "grad_norm": 1.1455787420272827, + "learning_rate": 3.7448579446021984e-07, + "loss": 0.5555, + "step": 14090 + }, + { + "epoch": 4.947682584269663, + "grad_norm": 1.4493175745010376, + "learning_rate": 3.7424243936487495e-07, + "loss": 0.5286, + "step": 14091 + }, + { + "epoch": 4.948033707865169, + "grad_norm": 0.967410147190094, + "learning_rate": 3.739991569686194e-07, + "loss": 0.5546, + "step": 14092 + }, + { + "epoch": 4.948384831460674, + "grad_norm": 0.9756190776824951, + "learning_rate": 3.737559472797736e-07, + "loss": 0.5537, + "step": 14093 + }, + { + "epoch": 4.94873595505618, + "grad_norm": 0.9794403314590454, + "learning_rate": 3.735128103066546e-07, + "loss": 0.5256, + "step": 14094 + }, + { + "epoch": 4.949087078651686, + "grad_norm": 0.9577419757843018, + "learning_rate": 3.732697460575785e-07, + "loss": 0.5611, + "step": 14095 + }, + { + "epoch": 4.949438202247191, + "grad_norm": 0.9628543853759766, + "learning_rate": 3.730267545408564e-07, + "loss": 0.5474, + "step": 14096 + }, + { + "epoch": 4.949789325842697, + "grad_norm": 0.9991858005523682, + "learning_rate": 3.7278383576480005e-07, + "loss": 0.5226, + "step": 14097 + }, + { + "epoch": 4.950140449438202, + "grad_norm": 0.9830366969108582, + "learning_rate": 3.7254098973771526e-07, + "loss": 0.5418, + "step": 14098 + }, + { + "epoch": 4.950491573033708, + "grad_norm": 0.974281370639801, + "learning_rate": 3.72298216467909e-07, + "loss": 0.5187, + "step": 14099 + }, + { + "epoch": 4.950842696629214, + "grad_norm": 0.9728162884712219, + "learning_rate": 3.7205551596368293e-07, + "loss": 0.5391, + "step": 14100 + }, + { + "epoch": 4.951193820224719, + "grad_norm": 0.9699310660362244, + "learning_rate": 3.7181288823333686e-07, + "loss": 0.5879, + "step": 14101 + }, + { + "epoch": 4.951544943820225, + "grad_norm": 0.9540532231330872, + "learning_rate": 3.7157033328516945e-07, + "loss": 0.5554, + "step": 14102 + }, + { + "epoch": 4.95189606741573, + "grad_norm": 0.9443933367729187, + "learning_rate": 3.713278511274748e-07, + "loss": 0.5348, + "step": 14103 + }, + { + "epoch": 4.952247191011236, + "grad_norm": 0.9619066715240479, + "learning_rate": 3.710854417685464e-07, + "loss": 0.5517, + "step": 14104 + }, + { + "epoch": 4.952598314606742, + "grad_norm": 0.9470134973526001, + "learning_rate": 3.708431052166739e-07, + "loss": 0.5086, + "step": 14105 + }, + { + "epoch": 4.952949438202247, + "grad_norm": 0.95307856798172, + "learning_rate": 3.706008414801457e-07, + "loss": 0.5167, + "step": 14106 + }, + { + "epoch": 4.953300561797753, + "grad_norm": 0.9307574033737183, + "learning_rate": 3.7035865056724655e-07, + "loss": 0.5034, + "step": 14107 + }, + { + "epoch": 4.953651685393258, + "grad_norm": 0.9607571363449097, + "learning_rate": 3.701165324862585e-07, + "loss": 0.5041, + "step": 14108 + }, + { + "epoch": 4.954002808988764, + "grad_norm": 0.9639852643013, + "learning_rate": 3.698744872454632e-07, + "loss": 0.5435, + "step": 14109 + }, + { + "epoch": 4.95435393258427, + "grad_norm": 0.9775679111480713, + "learning_rate": 3.696325148531371e-07, + "loss": 0.5119, + "step": 14110 + }, + { + "epoch": 4.954705056179775, + "grad_norm": 0.9896337985992432, + "learning_rate": 3.6939061531755663e-07, + "loss": 0.5661, + "step": 14111 + }, + { + "epoch": 4.955056179775281, + "grad_norm": 0.9673769474029541, + "learning_rate": 3.691487886469933e-07, + "loss": 0.5308, + "step": 14112 + }, + { + "epoch": 4.955407303370786, + "grad_norm": 0.9704262614250183, + "learning_rate": 3.689070348497184e-07, + "loss": 0.533, + "step": 14113 + }, + { + "epoch": 4.955758426966292, + "grad_norm": 0.9752947092056274, + "learning_rate": 3.6866535393399905e-07, + "loss": 0.5392, + "step": 14114 + }, + { + "epoch": 4.956109550561798, + "grad_norm": 0.9462851285934448, + "learning_rate": 3.684237459081011e-07, + "loss": 0.5379, + "step": 14115 + }, + { + "epoch": 4.956460674157303, + "grad_norm": 0.956204354763031, + "learning_rate": 3.681822107802871e-07, + "loss": 0.5458, + "step": 14116 + }, + { + "epoch": 4.956811797752809, + "grad_norm": 0.9795507192611694, + "learning_rate": 3.679407485588168e-07, + "loss": 0.5432, + "step": 14117 + }, + { + "epoch": 4.957162921348314, + "grad_norm": 0.9981139302253723, + "learning_rate": 3.67699359251949e-07, + "loss": 0.5598, + "step": 14118 + }, + { + "epoch": 4.95751404494382, + "grad_norm": 0.9527924060821533, + "learning_rate": 3.674580428679378e-07, + "loss": 0.4862, + "step": 14119 + }, + { + "epoch": 4.957865168539326, + "grad_norm": 0.974860429763794, + "learning_rate": 3.672167994150372e-07, + "loss": 0.5486, + "step": 14120 + }, + { + "epoch": 4.958216292134831, + "grad_norm": 0.9825214147567749, + "learning_rate": 3.669756289014967e-07, + "loss": 0.5355, + "step": 14121 + }, + { + "epoch": 4.958567415730337, + "grad_norm": 0.9659982323646545, + "learning_rate": 3.667345313355647e-07, + "loss": 0.5177, + "step": 14122 + }, + { + "epoch": 4.958918539325842, + "grad_norm": 0.950450599193573, + "learning_rate": 3.6649350672548653e-07, + "loss": 0.5217, + "step": 14123 + }, + { + "epoch": 4.959269662921348, + "grad_norm": 1.0028091669082642, + "learning_rate": 3.6625255507950404e-07, + "loss": 0.5157, + "step": 14124 + }, + { + "epoch": 4.959620786516854, + "grad_norm": 0.9729139804840088, + "learning_rate": 3.660116764058588e-07, + "loss": 0.5459, + "step": 14125 + }, + { + "epoch": 4.95997191011236, + "grad_norm": 0.9752771854400635, + "learning_rate": 3.657708707127877e-07, + "loss": 0.5924, + "step": 14126 + }, + { + "epoch": 4.960323033707866, + "grad_norm": 0.9662013649940491, + "learning_rate": 3.655301380085272e-07, + "loss": 0.5526, + "step": 14127 + }, + { + "epoch": 4.960674157303371, + "grad_norm": 0.9705924987792969, + "learning_rate": 3.652894783013089e-07, + "loss": 0.5503, + "step": 14128 + }, + { + "epoch": 4.961025280898877, + "grad_norm": 0.9524622559547424, + "learning_rate": 3.650488915993641e-07, + "loss": 0.5488, + "step": 14129 + }, + { + "epoch": 4.961376404494382, + "grad_norm": 0.9594883918762207, + "learning_rate": 3.6480837791092044e-07, + "loss": 0.531, + "step": 14130 + }, + { + "epoch": 4.961727528089888, + "grad_norm": 0.9866990447044373, + "learning_rate": 3.6456793724420264e-07, + "loss": 0.5162, + "step": 14131 + }, + { + "epoch": 4.962078651685394, + "grad_norm": 0.9387029409408569, + "learning_rate": 3.643275696074347e-07, + "loss": 0.5114, + "step": 14132 + }, + { + "epoch": 4.962429775280899, + "grad_norm": 0.9744367599487305, + "learning_rate": 3.640872750088356e-07, + "loss": 0.5473, + "step": 14133 + }, + { + "epoch": 4.962780898876405, + "grad_norm": 0.9761469960212708, + "learning_rate": 3.6384705345662463e-07, + "loss": 0.5341, + "step": 14134 + }, + { + "epoch": 4.96313202247191, + "grad_norm": 0.9777840375900269, + "learning_rate": 3.6360690495901574e-07, + "loss": 0.5341, + "step": 14135 + }, + { + "epoch": 4.963483146067416, + "grad_norm": 0.9604504108428955, + "learning_rate": 3.633668295242235e-07, + "loss": 0.5576, + "step": 14136 + }, + { + "epoch": 4.963834269662922, + "grad_norm": 0.9545714259147644, + "learning_rate": 3.6312682716045655e-07, + "loss": 0.551, + "step": 14137 + }, + { + "epoch": 4.964185393258427, + "grad_norm": 0.9212827682495117, + "learning_rate": 3.62886897875924e-07, + "loss": 0.4804, + "step": 14138 + }, + { + "epoch": 4.964536516853933, + "grad_norm": 0.9698283076286316, + "learning_rate": 3.626470416788305e-07, + "loss": 0.5219, + "step": 14139 + }, + { + "epoch": 4.964887640449438, + "grad_norm": 0.9760403633117676, + "learning_rate": 3.6240725857737967e-07, + "loss": 0.5315, + "step": 14140 + }, + { + "epoch": 4.965238764044944, + "grad_norm": 0.9715996980667114, + "learning_rate": 3.6216754857977125e-07, + "loss": 0.5189, + "step": 14141 + }, + { + "epoch": 4.96558988764045, + "grad_norm": 0.970618486404419, + "learning_rate": 3.619279116942029e-07, + "loss": 0.5252, + "step": 14142 + }, + { + "epoch": 4.965941011235955, + "grad_norm": 0.989822268486023, + "learning_rate": 3.6168834792887104e-07, + "loss": 0.5793, + "step": 14143 + }, + { + "epoch": 4.966292134831461, + "grad_norm": 0.9693855047225952, + "learning_rate": 3.614488572919672e-07, + "loss": 0.4856, + "step": 14144 + }, + { + "epoch": 4.966643258426966, + "grad_norm": 0.9809362292289734, + "learning_rate": 3.6120943979168323e-07, + "loss": 0.5359, + "step": 14145 + }, + { + "epoch": 4.966994382022472, + "grad_norm": 0.9906484484672546, + "learning_rate": 3.6097009543620525e-07, + "loss": 0.5598, + "step": 14146 + }, + { + "epoch": 4.967345505617978, + "grad_norm": 1.009926676750183, + "learning_rate": 3.607308242337207e-07, + "loss": 0.5392, + "step": 14147 + }, + { + "epoch": 4.967696629213483, + "grad_norm": 0.9799044132232666, + "learning_rate": 3.6049162619241046e-07, + "loss": 0.6006, + "step": 14148 + }, + { + "epoch": 4.968047752808989, + "grad_norm": 0.9816447496414185, + "learning_rate": 3.6025250132045637e-07, + "loss": 0.5198, + "step": 14149 + }, + { + "epoch": 4.968398876404494, + "grad_norm": 0.9886666536331177, + "learning_rate": 3.600134496260352e-07, + "loss": 0.5404, + "step": 14150 + }, + { + "epoch": 4.96875, + "grad_norm": 0.9774203300476074, + "learning_rate": 3.5977447111732345e-07, + "loss": 0.5401, + "step": 14151 + }, + { + "epoch": 4.969101123595506, + "grad_norm": 0.9962584376335144, + "learning_rate": 3.595355658024935e-07, + "loss": 0.5278, + "step": 14152 + }, + { + "epoch": 4.969452247191011, + "grad_norm": 0.9816545248031616, + "learning_rate": 3.592967336897149e-07, + "loss": 0.5398, + "step": 14153 + }, + { + "epoch": 4.969803370786517, + "grad_norm": 0.9939578175544739, + "learning_rate": 3.590579747871567e-07, + "loss": 0.5544, + "step": 14154 + }, + { + "epoch": 4.970154494382022, + "grad_norm": 0.9931473731994629, + "learning_rate": 3.588192891029832e-07, + "loss": 0.5157, + "step": 14155 + }, + { + "epoch": 4.970505617977528, + "grad_norm": 0.9653018712997437, + "learning_rate": 3.585806766453581e-07, + "loss": 0.4956, + "step": 14156 + }, + { + "epoch": 4.970856741573034, + "grad_norm": 0.9831558465957642, + "learning_rate": 3.5834213742244105e-07, + "loss": 0.5141, + "step": 14157 + }, + { + "epoch": 4.971207865168539, + "grad_norm": 0.9760482907295227, + "learning_rate": 3.5810367144239105e-07, + "loss": 0.5159, + "step": 14158 + }, + { + "epoch": 4.971558988764045, + "grad_norm": 0.9628086686134338, + "learning_rate": 3.5786527871336187e-07, + "loss": 0.4874, + "step": 14159 + }, + { + "epoch": 4.97191011235955, + "grad_norm": 0.9556901454925537, + "learning_rate": 3.5762695924350777e-07, + "loss": 0.5467, + "step": 14160 + }, + { + "epoch": 4.972261235955056, + "grad_norm": 0.979988694190979, + "learning_rate": 3.5738871304097754e-07, + "loss": 0.5495, + "step": 14161 + }, + { + "epoch": 4.972612359550562, + "grad_norm": 0.9588579535484314, + "learning_rate": 3.571505401139208e-07, + "loss": 0.5163, + "step": 14162 + }, + { + "epoch": 4.972963483146067, + "grad_norm": 0.9708030223846436, + "learning_rate": 3.56912440470481e-07, + "loss": 0.5382, + "step": 14163 + }, + { + "epoch": 4.973314606741573, + "grad_norm": 0.9929168224334717, + "learning_rate": 3.566744141188028e-07, + "loss": 0.5314, + "step": 14164 + }, + { + "epoch": 4.973665730337078, + "grad_norm": 1.0170319080352783, + "learning_rate": 3.564364610670254e-07, + "loss": 0.542, + "step": 14165 + }, + { + "epoch": 4.974016853932584, + "grad_norm": 0.982647716999054, + "learning_rate": 3.561985813232863e-07, + "loss": 0.5482, + "step": 14166 + }, + { + "epoch": 4.97436797752809, + "grad_norm": 0.9561066627502441, + "learning_rate": 3.559607748957217e-07, + "loss": 0.5087, + "step": 14167 + }, + { + "epoch": 4.974719101123595, + "grad_norm": 0.9759628176689148, + "learning_rate": 3.5572304179246315e-07, + "loss": 0.5294, + "step": 14168 + }, + { + "epoch": 4.975070224719101, + "grad_norm": 0.9834049940109253, + "learning_rate": 3.5548538202164253e-07, + "loss": 0.525, + "step": 14169 + }, + { + "epoch": 4.975421348314606, + "grad_norm": 0.9477645754814148, + "learning_rate": 3.5524779559138607e-07, + "loss": 0.542, + "step": 14170 + }, + { + "epoch": 4.975772471910112, + "grad_norm": 0.9936908483505249, + "learning_rate": 3.5501028250982036e-07, + "loss": 0.5814, + "step": 14171 + }, + { + "epoch": 4.976123595505618, + "grad_norm": 0.9452993869781494, + "learning_rate": 3.547728427850669e-07, + "loss": 0.5364, + "step": 14172 + }, + { + "epoch": 4.976474719101123, + "grad_norm": 0.967536985874176, + "learning_rate": 3.5453547642524707e-07, + "loss": 0.5306, + "step": 14173 + }, + { + "epoch": 4.976825842696629, + "grad_norm": 0.9337233901023865, + "learning_rate": 3.542981834384776e-07, + "loss": 0.5419, + "step": 14174 + }, + { + "epoch": 4.977176966292134, + "grad_norm": 0.9885879755020142, + "learning_rate": 3.5406096383287456e-07, + "loss": 0.5341, + "step": 14175 + }, + { + "epoch": 4.97752808988764, + "grad_norm": 0.9984539747238159, + "learning_rate": 3.5382381761654983e-07, + "loss": 0.5268, + "step": 14176 + }, + { + "epoch": 4.977879213483146, + "grad_norm": 0.9720786809921265, + "learning_rate": 3.5358674479761437e-07, + "loss": 0.5684, + "step": 14177 + }, + { + "epoch": 4.978230337078652, + "grad_norm": 0.9740042686462402, + "learning_rate": 3.533497453841758e-07, + "loss": 0.5262, + "step": 14178 + }, + { + "epoch": 4.978581460674158, + "grad_norm": 0.9543979167938232, + "learning_rate": 3.531128193843383e-07, + "loss": 0.5283, + "step": 14179 + }, + { + "epoch": 4.978932584269663, + "grad_norm": 0.9532420039176941, + "learning_rate": 3.528759668062057e-07, + "loss": 0.5637, + "step": 14180 + }, + { + "epoch": 4.979283707865169, + "grad_norm": 0.9801530838012695, + "learning_rate": 3.5263918765787744e-07, + "loss": 0.5364, + "step": 14181 + }, + { + "epoch": 4.979634831460674, + "grad_norm": 0.9246933460235596, + "learning_rate": 3.5240248194745167e-07, + "loss": 0.5042, + "step": 14182 + }, + { + "epoch": 4.97998595505618, + "grad_norm": 0.9525147080421448, + "learning_rate": 3.5216584968302295e-07, + "loss": 0.5425, + "step": 14183 + }, + { + "epoch": 4.980337078651686, + "grad_norm": 0.9477614164352417, + "learning_rate": 3.5192929087268483e-07, + "loss": 0.5334, + "step": 14184 + }, + { + "epoch": 4.980688202247191, + "grad_norm": 0.9801072478294373, + "learning_rate": 3.516928055245261e-07, + "loss": 0.5336, + "step": 14185 + }, + { + "epoch": 4.981039325842697, + "grad_norm": 0.9953145980834961, + "learning_rate": 3.5145639364663585e-07, + "loss": 0.5407, + "step": 14186 + }, + { + "epoch": 4.981390449438202, + "grad_norm": 0.9521083831787109, + "learning_rate": 3.512200552470979e-07, + "loss": 0.5552, + "step": 14187 + }, + { + "epoch": 4.981741573033708, + "grad_norm": 0.9848597645759583, + "learning_rate": 3.509837903339955e-07, + "loss": 0.553, + "step": 14188 + }, + { + "epoch": 4.982092696629214, + "grad_norm": 1.103560209274292, + "learning_rate": 3.5074759891540826e-07, + "loss": 0.55, + "step": 14189 + }, + { + "epoch": 4.982443820224719, + "grad_norm": 0.9753139019012451, + "learning_rate": 3.505114809994145e-07, + "loss": 0.5323, + "step": 14190 + }, + { + "epoch": 4.982794943820225, + "grad_norm": 0.9437910318374634, + "learning_rate": 3.5027543659408884e-07, + "loss": 0.5487, + "step": 14191 + }, + { + "epoch": 4.98314606741573, + "grad_norm": 0.9593170285224915, + "learning_rate": 3.500394657075029e-07, + "loss": 0.5337, + "step": 14192 + }, + { + "epoch": 4.983497191011236, + "grad_norm": 1.058624505996704, + "learning_rate": 3.4980356834772825e-07, + "loss": 0.5525, + "step": 14193 + }, + { + "epoch": 4.983848314606742, + "grad_norm": 0.9559711217880249, + "learning_rate": 3.49567744522831e-07, + "loss": 0.5204, + "step": 14194 + }, + { + "epoch": 4.984199438202247, + "grad_norm": 0.9469214677810669, + "learning_rate": 3.493319942408774e-07, + "loss": 0.5335, + "step": 14195 + }, + { + "epoch": 4.984550561797753, + "grad_norm": 1.1281222105026245, + "learning_rate": 3.490963175099285e-07, + "loss": 0.5658, + "step": 14196 + }, + { + "epoch": 4.984901685393258, + "grad_norm": 0.9939022064208984, + "learning_rate": 3.4886071433804567e-07, + "loss": 0.544, + "step": 14197 + }, + { + "epoch": 4.985252808988764, + "grad_norm": 0.9560510516166687, + "learning_rate": 3.48625184733285e-07, + "loss": 0.5589, + "step": 14198 + }, + { + "epoch": 4.98560393258427, + "grad_norm": 0.986882209777832, + "learning_rate": 3.483897287037027e-07, + "loss": 0.5451, + "step": 14199 + }, + { + "epoch": 4.985955056179775, + "grad_norm": 0.9682612419128418, + "learning_rate": 3.4815434625735e-07, + "loss": 0.5435, + "step": 14200 + }, + { + "epoch": 4.986306179775281, + "grad_norm": 0.9591750502586365, + "learning_rate": 3.4791903740227755e-07, + "loss": 0.5538, + "step": 14201 + }, + { + "epoch": 4.986657303370786, + "grad_norm": 0.9896494746208191, + "learning_rate": 3.476838021465323e-07, + "loss": 0.5347, + "step": 14202 + }, + { + "epoch": 4.987008426966292, + "grad_norm": 0.9804666638374329, + "learning_rate": 3.4744864049815945e-07, + "loss": 0.5385, + "step": 14203 + }, + { + "epoch": 4.987359550561798, + "grad_norm": 0.9664561748504639, + "learning_rate": 3.472135524652015e-07, + "loss": 0.5097, + "step": 14204 + }, + { + "epoch": 4.987710674157303, + "grad_norm": 0.9723274111747742, + "learning_rate": 3.4697853805569696e-07, + "loss": 0.5426, + "step": 14205 + }, + { + "epoch": 4.988061797752809, + "grad_norm": 0.9617511630058289, + "learning_rate": 3.4674359727768505e-07, + "loss": 0.5268, + "step": 14206 + }, + { + "epoch": 4.988412921348314, + "grad_norm": 0.9909668564796448, + "learning_rate": 3.465087301391987e-07, + "loss": 0.5808, + "step": 14207 + }, + { + "epoch": 4.98876404494382, + "grad_norm": 0.9646721482276917, + "learning_rate": 3.4627393664827145e-07, + "loss": 0.5564, + "step": 14208 + }, + { + "epoch": 4.989115168539326, + "grad_norm": 0.9716416001319885, + "learning_rate": 3.460392168129323e-07, + "loss": 0.5472, + "step": 14209 + }, + { + "epoch": 4.989466292134831, + "grad_norm": 0.9768399000167847, + "learning_rate": 3.458045706412091e-07, + "loss": 0.5258, + "step": 14210 + }, + { + "epoch": 4.989817415730337, + "grad_norm": 0.9879266023635864, + "learning_rate": 3.455699981411259e-07, + "loss": 0.5531, + "step": 14211 + }, + { + "epoch": 4.990168539325842, + "grad_norm": 1.0299594402313232, + "learning_rate": 3.453354993207056e-07, + "loss": 0.5261, + "step": 14212 + }, + { + "epoch": 4.990519662921348, + "grad_norm": 0.9861154556274414, + "learning_rate": 3.451010741879668e-07, + "loss": 0.5542, + "step": 14213 + }, + { + "epoch": 4.990870786516854, + "grad_norm": 0.9851969480514526, + "learning_rate": 3.4486672275092807e-07, + "loss": 0.5363, + "step": 14214 + }, + { + "epoch": 4.99122191011236, + "grad_norm": 0.945633053779602, + "learning_rate": 3.4463244501760305e-07, + "loss": 0.5116, + "step": 14215 + }, + { + "epoch": 4.991573033707866, + "grad_norm": 1.010069727897644, + "learning_rate": 3.4439824099600366e-07, + "loss": 0.5482, + "step": 14216 + }, + { + "epoch": 4.991924157303371, + "grad_norm": 0.9720730185508728, + "learning_rate": 3.441641106941401e-07, + "loss": 0.5648, + "step": 14217 + }, + { + "epoch": 4.992275280898877, + "grad_norm": 0.9643993973731995, + "learning_rate": 3.4393005412001914e-07, + "loss": 0.5539, + "step": 14218 + }, + { + "epoch": 4.992626404494382, + "grad_norm": 1.0163160562515259, + "learning_rate": 3.436960712816453e-07, + "loss": 0.5567, + "step": 14219 + }, + { + "epoch": 4.992977528089888, + "grad_norm": 0.9318769574165344, + "learning_rate": 3.4346216218702047e-07, + "loss": 0.5031, + "step": 14220 + }, + { + "epoch": 4.993328651685394, + "grad_norm": 0.9934771656990051, + "learning_rate": 3.43228326844145e-07, + "loss": 0.5322, + "step": 14221 + }, + { + "epoch": 4.993679775280899, + "grad_norm": 0.9668329358100891, + "learning_rate": 3.4299456526101423e-07, + "loss": 0.5307, + "step": 14222 + }, + { + "epoch": 4.994030898876405, + "grad_norm": 0.9629455804824829, + "learning_rate": 3.4276087744562423e-07, + "loss": 0.5465, + "step": 14223 + }, + { + "epoch": 4.99438202247191, + "grad_norm": 0.9894790053367615, + "learning_rate": 3.425272634059659e-07, + "loss": 0.5615, + "step": 14224 + }, + { + "epoch": 4.994733146067416, + "grad_norm": 0.9606223106384277, + "learning_rate": 3.4229372315002953e-07, + "loss": 0.5374, + "step": 14225 + }, + { + "epoch": 4.995084269662922, + "grad_norm": 0.9684764742851257, + "learning_rate": 3.4206025668580055e-07, + "loss": 0.5595, + "step": 14226 + }, + { + "epoch": 4.995435393258427, + "grad_norm": 0.9797052145004272, + "learning_rate": 3.418268640212652e-07, + "loss": 0.5616, + "step": 14227 + }, + { + "epoch": 4.995786516853933, + "grad_norm": 0.9583964943885803, + "learning_rate": 3.41593545164404e-07, + "loss": 0.5084, + "step": 14228 + }, + { + "epoch": 4.996137640449438, + "grad_norm": 0.9977115988731384, + "learning_rate": 3.4136030012319603e-07, + "loss": 0.5549, + "step": 14229 + }, + { + "epoch": 4.996488764044944, + "grad_norm": 0.976585328578949, + "learning_rate": 3.411271289056195e-07, + "loss": 0.5774, + "step": 14230 + }, + { + "epoch": 4.99683988764045, + "grad_norm": 0.9665892720222473, + "learning_rate": 3.408940315196468e-07, + "loss": 0.5138, + "step": 14231 + }, + { + "epoch": 4.997191011235955, + "grad_norm": 0.9754554033279419, + "learning_rate": 3.406610079732514e-07, + "loss": 0.5451, + "step": 14232 + }, + { + "epoch": 4.997542134831461, + "grad_norm": 0.9498443603515625, + "learning_rate": 3.404280582744013e-07, + "loss": 0.5067, + "step": 14233 + }, + { + "epoch": 4.997893258426966, + "grad_norm": 0.9703800082206726, + "learning_rate": 3.4019518243106426e-07, + "loss": 0.5363, + "step": 14234 + }, + { + "epoch": 4.998244382022472, + "grad_norm": 0.9451833963394165, + "learning_rate": 3.399623804512031e-07, + "loss": 0.5181, + "step": 14235 + }, + { + "epoch": 4.998595505617978, + "grad_norm": 0.9431206583976746, + "learning_rate": 3.397296523427807e-07, + "loss": 0.5267, + "step": 14236 + }, + { + "epoch": 4.998946629213483, + "grad_norm": 0.9900564551353455, + "learning_rate": 3.39496998113755e-07, + "loss": 0.5629, + "step": 14237 + }, + { + "epoch": 4.999297752808989, + "grad_norm": 0.9682465195655823, + "learning_rate": 3.3926441777208384e-07, + "loss": 0.5322, + "step": 14238 + }, + { + "epoch": 4.999648876404494, + "grad_norm": 1.0112299919128418, + "learning_rate": 3.390319113257201e-07, + "loss": 0.5387, + "step": 14239 + }, + { + "epoch": 5.0, + "grad_norm": 0.9732330441474915, + "learning_rate": 3.3879947878261625e-07, + "loss": 0.5599, + "step": 14240 + } + ], + "logging_steps": 1, + "max_steps": 17088, + "num_input_tokens_seen": 0, + "num_train_epochs": 6, + "save_steps": 2848, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 3.5970762561866957e+19, + "train_batch_size": 4, + "trial_name": null, + "trial_params": null +}