{ "best_metric": null, "best_model_checkpoint": null, "epoch": 5.0, "eval_steps": 500, "global_step": 1235, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.004048582995951417, "grad_norm": 7.882094273500173, "learning_rate": 3.2258064516129035e-07, "loss": 1.1073, "step": 1 }, { "epoch": 0.008097165991902834, "grad_norm": 7.878077210714097, "learning_rate": 6.451612903225807e-07, "loss": 1.1137, "step": 2 }, { "epoch": 0.012145748987854251, "grad_norm": 7.898899456990533, "learning_rate": 9.67741935483871e-07, "loss": 1.1075, "step": 3 }, { "epoch": 0.016194331983805668, "grad_norm": 7.799963117134728, "learning_rate": 1.2903225806451614e-06, "loss": 1.1002, "step": 4 }, { "epoch": 0.020242914979757085, "grad_norm": 7.007443589332562, "learning_rate": 1.6129032258064516e-06, "loss": 1.0961, "step": 5 }, { "epoch": 0.024291497975708502, "grad_norm": 6.888166105481446, "learning_rate": 1.935483870967742e-06, "loss": 1.0914, "step": 6 }, { "epoch": 0.02834008097165992, "grad_norm": 5.131096847796136, "learning_rate": 2.2580645161290324e-06, "loss": 1.0366, "step": 7 }, { "epoch": 0.032388663967611336, "grad_norm": 4.800195899669063, "learning_rate": 2.580645161290323e-06, "loss": 1.037, "step": 8 }, { "epoch": 0.03643724696356275, "grad_norm": 3.0744366328787964, "learning_rate": 2.903225806451613e-06, "loss": 0.9854, "step": 9 }, { "epoch": 0.04048582995951417, "grad_norm": 2.775513252084914, "learning_rate": 3.225806451612903e-06, "loss": 0.9877, "step": 10 }, { "epoch": 0.044534412955465584, "grad_norm": 2.896204756530653, "learning_rate": 3.548387096774194e-06, "loss": 0.9922, "step": 11 }, { "epoch": 0.048582995951417005, "grad_norm": 3.78085631456449, "learning_rate": 3.870967741935484e-06, "loss": 0.9451, "step": 12 }, { "epoch": 0.05263157894736842, "grad_norm": 4.262300846259237, "learning_rate": 4.193548387096774e-06, "loss": 0.962, "step": 13 }, { "epoch": 0.05668016194331984, "grad_norm": 3.854793075424327, "learning_rate": 4.516129032258065e-06, "loss": 0.9378, "step": 14 }, { "epoch": 0.06072874493927125, "grad_norm": 3.0664598825870075, "learning_rate": 4.838709677419355e-06, "loss": 0.9031, "step": 15 }, { "epoch": 0.06477732793522267, "grad_norm": 2.542370772664176, "learning_rate": 5.161290322580646e-06, "loss": 0.8852, "step": 16 }, { "epoch": 0.06882591093117409, "grad_norm": 2.6707348977901963, "learning_rate": 5.483870967741935e-06, "loss": 0.862, "step": 17 }, { "epoch": 0.0728744939271255, "grad_norm": 2.2913856935751657, "learning_rate": 5.806451612903226e-06, "loss": 0.8487, "step": 18 }, { "epoch": 0.07692307692307693, "grad_norm": 1.5550922441417883, "learning_rate": 6.129032258064517e-06, "loss": 0.8341, "step": 19 }, { "epoch": 0.08097165991902834, "grad_norm": 1.404560720829427, "learning_rate": 6.451612903225806e-06, "loss": 0.8194, "step": 20 }, { "epoch": 0.08502024291497975, "grad_norm": 1.3556945097795654, "learning_rate": 6.774193548387097e-06, "loss": 0.8089, "step": 21 }, { "epoch": 0.08906882591093117, "grad_norm": 1.372001371092715, "learning_rate": 7.096774193548388e-06, "loss": 0.8086, "step": 22 }, { "epoch": 0.0931174089068826, "grad_norm": 1.3767242354663116, "learning_rate": 7.4193548387096784e-06, "loss": 0.81, "step": 23 }, { "epoch": 0.09716599190283401, "grad_norm": 1.3456273721436534, "learning_rate": 7.741935483870968e-06, "loss": 0.7901, "step": 24 }, { "epoch": 0.10121457489878542, "grad_norm": 0.9220913185558441, "learning_rate": 8.064516129032258e-06, "loss": 0.7788, "step": 25 }, { "epoch": 0.10526315789473684, "grad_norm": 0.9696510870501455, "learning_rate": 8.387096774193549e-06, "loss": 0.769, "step": 26 }, { "epoch": 0.10931174089068826, "grad_norm": 1.0857823395570407, "learning_rate": 8.70967741935484e-06, "loss": 0.7647, "step": 27 }, { "epoch": 0.11336032388663968, "grad_norm": 0.9673503577045061, "learning_rate": 9.03225806451613e-06, "loss": 0.7624, "step": 28 }, { "epoch": 0.11740890688259109, "grad_norm": 0.8154660435335195, "learning_rate": 9.35483870967742e-06, "loss": 0.7423, "step": 29 }, { "epoch": 0.1214574898785425, "grad_norm": 0.848587359003483, "learning_rate": 9.67741935483871e-06, "loss": 0.7421, "step": 30 }, { "epoch": 0.12550607287449392, "grad_norm": 0.8994208182784287, "learning_rate": 1e-05, "loss": 0.7466, "step": 31 }, { "epoch": 0.12955465587044535, "grad_norm": 0.8143601333782581, "learning_rate": 1.0322580645161291e-05, "loss": 0.7364, "step": 32 }, { "epoch": 0.13360323886639677, "grad_norm": 0.7524351488394027, "learning_rate": 1.0645161290322582e-05, "loss": 0.7367, "step": 33 }, { "epoch": 0.13765182186234817, "grad_norm": 0.7778232988622295, "learning_rate": 1.096774193548387e-05, "loss": 0.7214, "step": 34 }, { "epoch": 0.1417004048582996, "grad_norm": 0.7972893872570641, "learning_rate": 1.1290322580645164e-05, "loss": 0.719, "step": 35 }, { "epoch": 0.145748987854251, "grad_norm": 0.7531504282250008, "learning_rate": 1.1612903225806453e-05, "loss": 0.7076, "step": 36 }, { "epoch": 0.14979757085020243, "grad_norm": 0.9084969209274426, "learning_rate": 1.1935483870967743e-05, "loss": 0.7205, "step": 37 }, { "epoch": 0.15384615384615385, "grad_norm": 0.9794622183239722, "learning_rate": 1.2258064516129034e-05, "loss": 0.7261, "step": 38 }, { "epoch": 0.15789473684210525, "grad_norm": 0.6979746463551584, "learning_rate": 1.2580645161290324e-05, "loss": 0.7162, "step": 39 }, { "epoch": 0.16194331983805668, "grad_norm": 0.7627642395938141, "learning_rate": 1.2903225806451613e-05, "loss": 0.7155, "step": 40 }, { "epoch": 0.1659919028340081, "grad_norm": 0.9052636481311727, "learning_rate": 1.3225806451612903e-05, "loss": 0.7138, "step": 41 }, { "epoch": 0.1700404858299595, "grad_norm": 1.0412771895222608, "learning_rate": 1.3548387096774194e-05, "loss": 0.7079, "step": 42 }, { "epoch": 0.17408906882591094, "grad_norm": 0.7098196842679796, "learning_rate": 1.3870967741935486e-05, "loss": 0.7124, "step": 43 }, { "epoch": 0.17813765182186234, "grad_norm": 0.8180764530402803, "learning_rate": 1.4193548387096776e-05, "loss": 0.6968, "step": 44 }, { "epoch": 0.18218623481781376, "grad_norm": 0.6919764862319594, "learning_rate": 1.4516129032258066e-05, "loss": 0.7066, "step": 45 }, { "epoch": 0.1862348178137652, "grad_norm": 0.6740184407128497, "learning_rate": 1.4838709677419357e-05, "loss": 0.6994, "step": 46 }, { "epoch": 0.1902834008097166, "grad_norm": 0.739948993834859, "learning_rate": 1.5161290322580646e-05, "loss": 0.6969, "step": 47 }, { "epoch": 0.19433198380566802, "grad_norm": 0.7521927154879392, "learning_rate": 1.5483870967741936e-05, "loss": 0.686, "step": 48 }, { "epoch": 0.19838056680161945, "grad_norm": 0.7153091810934201, "learning_rate": 1.5806451612903226e-05, "loss": 0.686, "step": 49 }, { "epoch": 0.20242914979757085, "grad_norm": 0.8086860396280745, "learning_rate": 1.6129032258064517e-05, "loss": 0.7075, "step": 50 }, { "epoch": 0.20647773279352227, "grad_norm": 0.7177074473143203, "learning_rate": 1.6451612903225807e-05, "loss": 0.7129, "step": 51 }, { "epoch": 0.21052631578947367, "grad_norm": 0.7976607365795638, "learning_rate": 1.6774193548387098e-05, "loss": 0.7036, "step": 52 }, { "epoch": 0.2145748987854251, "grad_norm": 0.6895866989937592, "learning_rate": 1.7096774193548388e-05, "loss": 0.6782, "step": 53 }, { "epoch": 0.21862348178137653, "grad_norm": 0.7456168936353316, "learning_rate": 1.741935483870968e-05, "loss": 0.6853, "step": 54 }, { "epoch": 0.22267206477732793, "grad_norm": 0.8771756457732699, "learning_rate": 1.774193548387097e-05, "loss": 0.6874, "step": 55 }, { "epoch": 0.22672064777327935, "grad_norm": 0.7584450300990598, "learning_rate": 1.806451612903226e-05, "loss": 0.691, "step": 56 }, { "epoch": 0.23076923076923078, "grad_norm": 0.8443786867958131, "learning_rate": 1.838709677419355e-05, "loss": 0.6863, "step": 57 }, { "epoch": 0.23481781376518218, "grad_norm": 0.7519909897405547, "learning_rate": 1.870967741935484e-05, "loss": 0.6872, "step": 58 }, { "epoch": 0.2388663967611336, "grad_norm": 0.8711403516648242, "learning_rate": 1.903225806451613e-05, "loss": 0.6792, "step": 59 }, { "epoch": 0.242914979757085, "grad_norm": 0.657205742874964, "learning_rate": 1.935483870967742e-05, "loss": 0.6846, "step": 60 }, { "epoch": 0.24696356275303644, "grad_norm": 0.8090824782920741, "learning_rate": 1.967741935483871e-05, "loss": 0.682, "step": 61 }, { "epoch": 0.25101214574898784, "grad_norm": 1.1260983695663622, "learning_rate": 2e-05, "loss": 0.6836, "step": 62 }, { "epoch": 0.2550607287449393, "grad_norm": 0.7340148194253817, "learning_rate": 2.0322580645161292e-05, "loss": 0.6879, "step": 63 }, { "epoch": 0.2591093117408907, "grad_norm": 1.0565060945375306, "learning_rate": 2.0645161290322582e-05, "loss": 0.6768, "step": 64 }, { "epoch": 0.2631578947368421, "grad_norm": 1.4046960777987767, "learning_rate": 2.0967741935483873e-05, "loss": 0.6808, "step": 65 }, { "epoch": 0.26720647773279355, "grad_norm": 0.8307867984649079, "learning_rate": 2.1290322580645163e-05, "loss": 0.6714, "step": 66 }, { "epoch": 0.27125506072874495, "grad_norm": 1.4484160576508127, "learning_rate": 2.161290322580645e-05, "loss": 0.6839, "step": 67 }, { "epoch": 0.27530364372469635, "grad_norm": 0.9776248773826525, "learning_rate": 2.193548387096774e-05, "loss": 0.6709, "step": 68 }, { "epoch": 0.2793522267206478, "grad_norm": 1.4895212449047839, "learning_rate": 2.225806451612903e-05, "loss": 0.6747, "step": 69 }, { "epoch": 0.2834008097165992, "grad_norm": 1.0305241918881496, "learning_rate": 2.2580645161290328e-05, "loss": 0.6836, "step": 70 }, { "epoch": 0.2874493927125506, "grad_norm": 1.5883626343937964, "learning_rate": 2.2903225806451615e-05, "loss": 0.6947, "step": 71 }, { "epoch": 0.291497975708502, "grad_norm": 0.9737445868537087, "learning_rate": 2.3225806451612906e-05, "loss": 0.6804, "step": 72 }, { "epoch": 0.29554655870445345, "grad_norm": 1.6809954921790682, "learning_rate": 2.3548387096774196e-05, "loss": 0.6762, "step": 73 }, { "epoch": 0.29959514170040485, "grad_norm": 1.5389300291614563, "learning_rate": 2.3870967741935486e-05, "loss": 0.6726, "step": 74 }, { "epoch": 0.30364372469635625, "grad_norm": 1.3911177323441521, "learning_rate": 2.4193548387096777e-05, "loss": 0.6841, "step": 75 }, { "epoch": 0.3076923076923077, "grad_norm": 1.7373878004344372, "learning_rate": 2.4516129032258067e-05, "loss": 0.662, "step": 76 }, { "epoch": 0.3117408906882591, "grad_norm": 1.2712498925004307, "learning_rate": 2.4838709677419358e-05, "loss": 0.6789, "step": 77 }, { "epoch": 0.3157894736842105, "grad_norm": 1.8579590153214287, "learning_rate": 2.5161290322580648e-05, "loss": 0.6738, "step": 78 }, { "epoch": 0.31983805668016196, "grad_norm": 1.519653642723494, "learning_rate": 2.548387096774194e-05, "loss": 0.6764, "step": 79 }, { "epoch": 0.32388663967611336, "grad_norm": 1.332777263294369, "learning_rate": 2.5806451612903226e-05, "loss": 0.6823, "step": 80 }, { "epoch": 0.32793522267206476, "grad_norm": 1.2941649224567768, "learning_rate": 2.6129032258064516e-05, "loss": 0.6857, "step": 81 }, { "epoch": 0.3319838056680162, "grad_norm": 1.1084930403350992, "learning_rate": 2.6451612903225806e-05, "loss": 0.6727, "step": 82 }, { "epoch": 0.3360323886639676, "grad_norm": 1.0605554599182503, "learning_rate": 2.6774193548387097e-05, "loss": 0.6645, "step": 83 }, { "epoch": 0.340080971659919, "grad_norm": 1.0107604869788853, "learning_rate": 2.7096774193548387e-05, "loss": 0.6588, "step": 84 }, { "epoch": 0.3441295546558704, "grad_norm": 1.4089016954492848, "learning_rate": 2.7419354838709678e-05, "loss": 0.669, "step": 85 }, { "epoch": 0.3481781376518219, "grad_norm": 0.9616972711656484, "learning_rate": 2.774193548387097e-05, "loss": 0.6564, "step": 86 }, { "epoch": 0.3522267206477733, "grad_norm": 1.7907087164232462, "learning_rate": 2.8064516129032262e-05, "loss": 0.68, "step": 87 }, { "epoch": 0.3562753036437247, "grad_norm": 1.1784082027314597, "learning_rate": 2.8387096774193552e-05, "loss": 0.6679, "step": 88 }, { "epoch": 0.3603238866396761, "grad_norm": 1.710107632238156, "learning_rate": 2.8709677419354843e-05, "loss": 0.675, "step": 89 }, { "epoch": 0.3643724696356275, "grad_norm": 1.2566250241075074, "learning_rate": 2.9032258064516133e-05, "loss": 0.6687, "step": 90 }, { "epoch": 0.3684210526315789, "grad_norm": 1.7150620241139247, "learning_rate": 2.9354838709677423e-05, "loss": 0.6753, "step": 91 }, { "epoch": 0.3724696356275304, "grad_norm": 1.2917048805499056, "learning_rate": 2.9677419354838714e-05, "loss": 0.6675, "step": 92 }, { "epoch": 0.3765182186234818, "grad_norm": 1.501558051534323, "learning_rate": 3.0000000000000004e-05, "loss": 0.6762, "step": 93 }, { "epoch": 0.3805668016194332, "grad_norm": 1.2957390540726195, "learning_rate": 3.032258064516129e-05, "loss": 0.6849, "step": 94 }, { "epoch": 0.38461538461538464, "grad_norm": 1.1737638818271559, "learning_rate": 3.0645161290322585e-05, "loss": 0.6688, "step": 95 }, { "epoch": 0.38866396761133604, "grad_norm": 1.3437403861610935, "learning_rate": 3.096774193548387e-05, "loss": 0.6714, "step": 96 }, { "epoch": 0.39271255060728744, "grad_norm": 1.1799648489322407, "learning_rate": 3.1290322580645166e-05, "loss": 0.6745, "step": 97 }, { "epoch": 0.3967611336032389, "grad_norm": 0.9776560337041099, "learning_rate": 3.161290322580645e-05, "loss": 0.669, "step": 98 }, { "epoch": 0.4008097165991903, "grad_norm": 1.4179545278771335, "learning_rate": 3.193548387096774e-05, "loss": 0.6688, "step": 99 }, { "epoch": 0.4048582995951417, "grad_norm": 0.8171918645875557, "learning_rate": 3.2258064516129034e-05, "loss": 0.6561, "step": 100 }, { "epoch": 0.4089068825910931, "grad_norm": 1.430422217757808, "learning_rate": 3.258064516129033e-05, "loss": 0.6708, "step": 101 }, { "epoch": 0.41295546558704455, "grad_norm": 0.9643659222291623, "learning_rate": 3.2903225806451614e-05, "loss": 0.6593, "step": 102 }, { "epoch": 0.41700404858299595, "grad_norm": 1.0875643785841147, "learning_rate": 3.322580645161291e-05, "loss": 0.668, "step": 103 }, { "epoch": 0.42105263157894735, "grad_norm": 0.8069912782665241, "learning_rate": 3.3548387096774195e-05, "loss": 0.6672, "step": 104 }, { "epoch": 0.4251012145748988, "grad_norm": 0.9578855120939409, "learning_rate": 3.387096774193549e-05, "loss": 0.6709, "step": 105 }, { "epoch": 0.4291497975708502, "grad_norm": 1.2239941997055563, "learning_rate": 3.4193548387096776e-05, "loss": 0.6662, "step": 106 }, { "epoch": 0.4331983805668016, "grad_norm": 0.7022716356388247, "learning_rate": 3.451612903225807e-05, "loss": 0.6746, "step": 107 }, { "epoch": 0.43724696356275305, "grad_norm": 1.0567196188271675, "learning_rate": 3.483870967741936e-05, "loss": 0.6628, "step": 108 }, { "epoch": 0.44129554655870445, "grad_norm": 1.0331572331084202, "learning_rate": 3.516129032258065e-05, "loss": 0.6789, "step": 109 }, { "epoch": 0.44534412955465585, "grad_norm": 0.9331066931247813, "learning_rate": 3.548387096774194e-05, "loss": 0.6732, "step": 110 }, { "epoch": 0.4493927125506073, "grad_norm": 0.956405097670445, "learning_rate": 3.5806451612903225e-05, "loss": 0.6704, "step": 111 }, { "epoch": 0.4534412955465587, "grad_norm": 1.1185889578606147, "learning_rate": 3.612903225806452e-05, "loss": 0.6687, "step": 112 }, { "epoch": 0.4574898785425101, "grad_norm": 0.8143219355412407, "learning_rate": 3.6451612903225805e-05, "loss": 0.6686, "step": 113 }, { "epoch": 0.46153846153846156, "grad_norm": 1.0186890238498854, "learning_rate": 3.67741935483871e-05, "loss": 0.6723, "step": 114 }, { "epoch": 0.46558704453441296, "grad_norm": 1.0008381887579187, "learning_rate": 3.7096774193548386e-05, "loss": 0.6774, "step": 115 }, { "epoch": 0.46963562753036436, "grad_norm": 1.1627107394241238, "learning_rate": 3.741935483870968e-05, "loss": 0.6676, "step": 116 }, { "epoch": 0.47368421052631576, "grad_norm": 1.065534255716223, "learning_rate": 3.7741935483870974e-05, "loss": 0.6659, "step": 117 }, { "epoch": 0.4777327935222672, "grad_norm": 0.8985518350761134, "learning_rate": 3.806451612903226e-05, "loss": 0.6614, "step": 118 }, { "epoch": 0.4817813765182186, "grad_norm": 1.7624115769296718, "learning_rate": 3.8387096774193555e-05, "loss": 0.6827, "step": 119 }, { "epoch": 0.48582995951417, "grad_norm": 0.9885715591271229, "learning_rate": 3.870967741935484e-05, "loss": 0.6632, "step": 120 }, { "epoch": 0.4898785425101215, "grad_norm": 1.7065215847437567, "learning_rate": 3.9032258064516135e-05, "loss": 0.6668, "step": 121 }, { "epoch": 0.4939271255060729, "grad_norm": 1.0926929554964997, "learning_rate": 3.935483870967742e-05, "loss": 0.6772, "step": 122 }, { "epoch": 0.4979757085020243, "grad_norm": 1.6048883780499372, "learning_rate": 3.9677419354838716e-05, "loss": 0.6788, "step": 123 }, { "epoch": 0.5020242914979757, "grad_norm": 1.0361400678252042, "learning_rate": 4e-05, "loss": 0.6791, "step": 124 }, { "epoch": 0.5060728744939271, "grad_norm": 1.449548453251311, "learning_rate": 3.9999920040266474e-05, "loss": 0.6718, "step": 125 }, { "epoch": 0.5101214574898786, "grad_norm": 1.1480515006389913, "learning_rate": 3.999968016170525e-05, "loss": 0.6621, "step": 126 }, { "epoch": 0.5141700404858299, "grad_norm": 1.2130886503614051, "learning_rate": 3.9999280366234394e-05, "loss": 0.6647, "step": 127 }, { "epoch": 0.5182186234817814, "grad_norm": 1.2767515580681137, "learning_rate": 3.999872065705065e-05, "loss": 0.6802, "step": 128 }, { "epoch": 0.5222672064777328, "grad_norm": 1.0626700782229634, "learning_rate": 3.999800103862945e-05, "loss": 0.6507, "step": 129 }, { "epoch": 0.5263157894736842, "grad_norm": 1.2938582227957467, "learning_rate": 3.9997121516724834e-05, "loss": 0.6668, "step": 130 }, { "epoch": 0.5303643724696356, "grad_norm": 0.9855789507110971, "learning_rate": 3.999608209836944e-05, "loss": 0.6708, "step": 131 }, { "epoch": 0.5344129554655871, "grad_norm": 1.008675636050608, "learning_rate": 3.999488279187443e-05, "loss": 0.6562, "step": 132 }, { "epoch": 0.5384615384615384, "grad_norm": 1.5549269558454728, "learning_rate": 3.9993523606829425e-05, "loss": 0.6468, "step": 133 }, { "epoch": 0.5425101214574899, "grad_norm": 1.116732824053692, "learning_rate": 3.9992004554102435e-05, "loss": 0.6646, "step": 134 }, { "epoch": 0.5465587044534413, "grad_norm": 1.815572049197193, "learning_rate": 3.999032564583976e-05, "loss": 0.6649, "step": 135 }, { "epoch": 0.5506072874493927, "grad_norm": 1.5020494048276833, "learning_rate": 3.998848689546591e-05, "loss": 0.6758, "step": 136 }, { "epoch": 0.5546558704453441, "grad_norm": 1.30276440989443, "learning_rate": 3.9986488317683496e-05, "loss": 0.6684, "step": 137 }, { "epoch": 0.5587044534412956, "grad_norm": 0.8825189610870944, "learning_rate": 3.9984329928473067e-05, "loss": 0.6645, "step": 138 }, { "epoch": 0.562753036437247, "grad_norm": 1.1115673767578254, "learning_rate": 3.998201174509306e-05, "loss": 0.663, "step": 139 }, { "epoch": 0.5668016194331984, "grad_norm": 0.9259897556815295, "learning_rate": 3.997953378607961e-05, "loss": 0.6664, "step": 140 }, { "epoch": 0.5708502024291497, "grad_norm": 0.848235511465222, "learning_rate": 3.9976896071246403e-05, "loss": 0.6502, "step": 141 }, { "epoch": 0.5748987854251012, "grad_norm": 0.9399044748791721, "learning_rate": 3.997409862168454e-05, "loss": 0.6575, "step": 142 }, { "epoch": 0.5789473684210527, "grad_norm": 0.9227194908530267, "learning_rate": 3.997114145976237e-05, "loss": 0.6472, "step": 143 }, { "epoch": 0.582995951417004, "grad_norm": 1.1387147173646979, "learning_rate": 3.9968024609125245e-05, "loss": 0.6613, "step": 144 }, { "epoch": 0.5870445344129555, "grad_norm": 0.9382257165336029, "learning_rate": 3.996474809469545e-05, "loss": 0.653, "step": 145 }, { "epoch": 0.5910931174089069, "grad_norm": 0.8255587725208898, "learning_rate": 3.996131194267189e-05, "loss": 0.6605, "step": 146 }, { "epoch": 0.5951417004048583, "grad_norm": 1.3683472551487146, "learning_rate": 3.995771618052995e-05, "loss": 0.6662, "step": 147 }, { "epoch": 0.5991902834008097, "grad_norm": 0.8417208476981121, "learning_rate": 3.9953960837021256e-05, "loss": 0.6659, "step": 148 }, { "epoch": 0.6032388663967612, "grad_norm": 0.9870595341776506, "learning_rate": 3.995004594217343e-05, "loss": 0.6541, "step": 149 }, { "epoch": 0.6072874493927125, "grad_norm": 1.0518868041847447, "learning_rate": 3.994597152728987e-05, "loss": 0.6576, "step": 150 }, { "epoch": 0.611336032388664, "grad_norm": 0.9286763543378607, "learning_rate": 3.994173762494947e-05, "loss": 0.6666, "step": 151 }, { "epoch": 0.6153846153846154, "grad_norm": 1.027925963732866, "learning_rate": 3.9937344269006424e-05, "loss": 0.6557, "step": 152 }, { "epoch": 0.6194331983805668, "grad_norm": 0.9350761827267431, "learning_rate": 3.993279149458988e-05, "loss": 0.6647, "step": 153 }, { "epoch": 0.6234817813765182, "grad_norm": 0.8487024232065322, "learning_rate": 3.9928079338103695e-05, "loss": 0.6613, "step": 154 }, { "epoch": 0.6275303643724697, "grad_norm": 4.57655681835427, "learning_rate": 3.992320783722616e-05, "loss": 0.6578, "step": 155 }, { "epoch": 0.631578947368421, "grad_norm": 1.4129797393417678, "learning_rate": 3.991817703090965e-05, "loss": 0.6782, "step": 156 }, { "epoch": 0.6356275303643725, "grad_norm": 0.6427904797915323, "learning_rate": 3.991298695938038e-05, "loss": 0.6542, "step": 157 }, { "epoch": 0.6396761133603239, "grad_norm": 1.402756836678152, "learning_rate": 3.9907637664138e-05, "loss": 0.6715, "step": 158 }, { "epoch": 0.6437246963562753, "grad_norm": 0.760835008202141, "learning_rate": 3.990212918795534e-05, "loss": 0.6578, "step": 159 }, { "epoch": 0.6477732793522267, "grad_norm": 1.3222955009872337, "learning_rate": 3.9896461574878036e-05, "loss": 0.6426, "step": 160 }, { "epoch": 0.6518218623481782, "grad_norm": 0.8430778997083266, "learning_rate": 3.989063487022416e-05, "loss": 0.6512, "step": 161 }, { "epoch": 0.6558704453441295, "grad_norm": 1.0947846877202903, "learning_rate": 3.9884649120583903e-05, "loss": 0.6614, "step": 162 }, { "epoch": 0.659919028340081, "grad_norm": 1.013177392173862, "learning_rate": 3.987850437381915e-05, "loss": 0.6652, "step": 163 }, { "epoch": 0.6639676113360324, "grad_norm": 0.7907276381250335, "learning_rate": 3.987220067906313e-05, "loss": 0.6601, "step": 164 }, { "epoch": 0.6680161943319838, "grad_norm": 1.1849228985023774, "learning_rate": 3.986573808672002e-05, "loss": 0.6624, "step": 165 }, { "epoch": 0.6720647773279352, "grad_norm": 0.6594625770701594, "learning_rate": 3.9859116648464534e-05, "loss": 0.6525, "step": 166 }, { "epoch": 0.6761133603238867, "grad_norm": 1.1646357339686204, "learning_rate": 3.985233641724152e-05, "loss": 0.6565, "step": 167 }, { "epoch": 0.680161943319838, "grad_norm": 0.8034788207799688, "learning_rate": 3.984539744726554e-05, "loss": 0.6541, "step": 168 }, { "epoch": 0.6842105263157895, "grad_norm": 0.8626390856968242, "learning_rate": 3.983829979402038e-05, "loss": 0.6422, "step": 169 }, { "epoch": 0.6882591093117408, "grad_norm": 0.9493380647646935, "learning_rate": 3.983104351425871e-05, "loss": 0.6596, "step": 170 }, { "epoch": 0.6923076923076923, "grad_norm": 0.7151410491672231, "learning_rate": 3.982362866600154e-05, "loss": 0.641, "step": 171 }, { "epoch": 0.6963562753036437, "grad_norm": 0.8750103560706539, "learning_rate": 3.9816055308537806e-05, "loss": 0.6566, "step": 172 }, { "epoch": 0.7004048582995951, "grad_norm": 0.7032885558972359, "learning_rate": 3.980832350242388e-05, "loss": 0.6495, "step": 173 }, { "epoch": 0.7044534412955465, "grad_norm": 0.699256477976818, "learning_rate": 3.980043330948306e-05, "loss": 0.6475, "step": 174 }, { "epoch": 0.708502024291498, "grad_norm": 0.8401949218700564, "learning_rate": 3.9792384792805124e-05, "loss": 0.6428, "step": 175 }, { "epoch": 0.7125506072874493, "grad_norm": 0.6026088490473384, "learning_rate": 3.978417801674579e-05, "loss": 0.6467, "step": 176 }, { "epoch": 0.7165991902834008, "grad_norm": 0.676856734999651, "learning_rate": 3.9775813046926244e-05, "loss": 0.6526, "step": 177 }, { "epoch": 0.7206477732793523, "grad_norm": 0.8506936180969309, "learning_rate": 3.976728995023254e-05, "loss": 0.6649, "step": 178 }, { "epoch": 0.7246963562753036, "grad_norm": 0.6079747452052633, "learning_rate": 3.975860879481514e-05, "loss": 0.6524, "step": 179 }, { "epoch": 0.728744939271255, "grad_norm": 0.8175372210811437, "learning_rate": 3.974976965008833e-05, "loss": 0.6519, "step": 180 }, { "epoch": 0.7327935222672065, "grad_norm": 0.7729415168260595, "learning_rate": 3.9740772586729674e-05, "loss": 0.662, "step": 181 }, { "epoch": 0.7368421052631579, "grad_norm": 0.6707234918622768, "learning_rate": 3.973161767667946e-05, "loss": 0.6383, "step": 182 }, { "epoch": 0.7408906882591093, "grad_norm": 0.5677191163933328, "learning_rate": 3.972230499314009e-05, "loss": 0.6537, "step": 183 }, { "epoch": 0.7449392712550608, "grad_norm": 0.7633744626097358, "learning_rate": 3.9712834610575545e-05, "loss": 0.6413, "step": 184 }, { "epoch": 0.7489878542510121, "grad_norm": 0.7417549740044391, "learning_rate": 3.9703206604710746e-05, "loss": 0.6539, "step": 185 }, { "epoch": 0.7530364372469636, "grad_norm": 0.7062579755509876, "learning_rate": 3.969342105253098e-05, "loss": 0.6492, "step": 186 }, { "epoch": 0.757085020242915, "grad_norm": 0.6650351693034551, "learning_rate": 3.9683478032281246e-05, "loss": 0.6583, "step": 187 }, { "epoch": 0.7611336032388664, "grad_norm": 0.6993263941012752, "learning_rate": 3.967337762346568e-05, "loss": 0.6496, "step": 188 }, { "epoch": 0.7651821862348178, "grad_norm": 0.7189768352414295, "learning_rate": 3.9663119906846885e-05, "loss": 0.6533, "step": 189 }, { "epoch": 0.7692307692307693, "grad_norm": 0.6691245679064446, "learning_rate": 3.965270496444528e-05, "loss": 0.6628, "step": 190 }, { "epoch": 0.7732793522267206, "grad_norm": 0.6424998107516554, "learning_rate": 3.9642132879538465e-05, "loss": 0.6653, "step": 191 }, { "epoch": 0.7773279352226721, "grad_norm": 0.6550922270175151, "learning_rate": 3.9631403736660566e-05, "loss": 0.6625, "step": 192 }, { "epoch": 0.7813765182186235, "grad_norm": 0.6023432515929331, "learning_rate": 3.962051762160151e-05, "loss": 0.6385, "step": 193 }, { "epoch": 0.7854251012145749, "grad_norm": 0.5584887570771214, "learning_rate": 3.960947462140638e-05, "loss": 0.65, "step": 194 }, { "epoch": 0.7894736842105263, "grad_norm": 0.6210203484605729, "learning_rate": 3.959827482437472e-05, "loss": 0.6496, "step": 195 }, { "epoch": 0.7935222672064778, "grad_norm": 0.5842279151116365, "learning_rate": 3.958691832005981e-05, "loss": 0.6349, "step": 196 }, { "epoch": 0.7975708502024291, "grad_norm": 0.5557938578846497, "learning_rate": 3.9575405199267944e-05, "loss": 0.6514, "step": 197 }, { "epoch": 0.8016194331983806, "grad_norm": 0.608956989257208, "learning_rate": 3.956373555405774e-05, "loss": 0.6522, "step": 198 }, { "epoch": 0.805668016194332, "grad_norm": 0.6111932093340493, "learning_rate": 3.955190947773937e-05, "loss": 0.6375, "step": 199 }, { "epoch": 0.8097165991902834, "grad_norm": 0.6255130108975574, "learning_rate": 3.9539927064873824e-05, "loss": 0.6564, "step": 200 }, { "epoch": 0.8137651821862348, "grad_norm": 0.6652174489338144, "learning_rate": 3.952778841127214e-05, "loss": 0.648, "step": 201 }, { "epoch": 0.8178137651821862, "grad_norm": 0.5910354625097416, "learning_rate": 3.95154936139947e-05, "loss": 0.6383, "step": 202 }, { "epoch": 0.8218623481781376, "grad_norm": 0.5634238948681023, "learning_rate": 3.950304277135034e-05, "loss": 0.6386, "step": 203 }, { "epoch": 0.8259109311740891, "grad_norm": 0.6653837438713891, "learning_rate": 3.94904359828957e-05, "loss": 0.6482, "step": 204 }, { "epoch": 0.8299595141700404, "grad_norm": 0.5529477720981693, "learning_rate": 3.94776733494343e-05, "loss": 0.6572, "step": 205 }, { "epoch": 0.8340080971659919, "grad_norm": 0.5107489343815749, "learning_rate": 3.9464754973015826e-05, "loss": 0.6432, "step": 206 }, { "epoch": 0.8380566801619433, "grad_norm": 0.5901551449713127, "learning_rate": 3.9451680956935275e-05, "loss": 0.6414, "step": 207 }, { "epoch": 0.8421052631578947, "grad_norm": 0.581665561424497, "learning_rate": 3.943845140573213e-05, "loss": 0.6385, "step": 208 }, { "epoch": 0.8461538461538461, "grad_norm": 0.5435604045599538, "learning_rate": 3.942506642518952e-05, "loss": 0.6431, "step": 209 }, { "epoch": 0.8502024291497976, "grad_norm": 0.6875483760717378, "learning_rate": 3.9411526122333396e-05, "loss": 0.6426, "step": 210 }, { "epoch": 0.854251012145749, "grad_norm": 0.702396901278839, "learning_rate": 3.939783060543167e-05, "loss": 0.657, "step": 211 }, { "epoch": 0.8582995951417004, "grad_norm": 0.5246915929973963, "learning_rate": 3.938397998399333e-05, "loss": 0.6433, "step": 212 }, { "epoch": 0.8623481781376519, "grad_norm": 0.5950347881280286, "learning_rate": 3.936997436876756e-05, "loss": 0.6454, "step": 213 }, { "epoch": 0.8663967611336032, "grad_norm": 0.5152054468992084, "learning_rate": 3.935581387174289e-05, "loss": 0.6509, "step": 214 }, { "epoch": 0.8704453441295547, "grad_norm": 0.5589375950134066, "learning_rate": 3.9341498606146295e-05, "loss": 0.6464, "step": 215 }, { "epoch": 0.8744939271255061, "grad_norm": 0.6458302316100499, "learning_rate": 3.932702868644224e-05, "loss": 0.6461, "step": 216 }, { "epoch": 0.8785425101214575, "grad_norm": 0.6118259828326609, "learning_rate": 3.931240422833181e-05, "loss": 0.6527, "step": 217 }, { "epoch": 0.8825910931174089, "grad_norm": 0.5824116177593026, "learning_rate": 3.9297625348751806e-05, "loss": 0.6436, "step": 218 }, { "epoch": 0.8866396761133604, "grad_norm": 0.6173254963524052, "learning_rate": 3.9282692165873744e-05, "loss": 0.6413, "step": 219 }, { "epoch": 0.8906882591093117, "grad_norm": 0.5324063504105212, "learning_rate": 3.9267604799102945e-05, "loss": 0.6465, "step": 220 }, { "epoch": 0.8947368421052632, "grad_norm": 0.6349358433711989, "learning_rate": 3.9252363369077615e-05, "loss": 0.6386, "step": 221 }, { "epoch": 0.8987854251012146, "grad_norm": 0.6137727261025042, "learning_rate": 3.9236967997667796e-05, "loss": 0.6348, "step": 222 }, { "epoch": 0.902834008097166, "grad_norm": 0.7560763650953864, "learning_rate": 3.9221418807974496e-05, "loss": 0.6375, "step": 223 }, { "epoch": 0.9068825910931174, "grad_norm": 0.5903599511766886, "learning_rate": 3.92057159243286e-05, "loss": 0.6402, "step": 224 }, { "epoch": 0.9109311740890689, "grad_norm": 0.5466544733529144, "learning_rate": 3.9189859472289956e-05, "loss": 0.6508, "step": 225 }, { "epoch": 0.9149797570850202, "grad_norm": 0.614249501614761, "learning_rate": 3.9173849578646324e-05, "loss": 0.6227, "step": 226 }, { "epoch": 0.9190283400809717, "grad_norm": 0.6527517310731538, "learning_rate": 3.9157686371412395e-05, "loss": 0.6556, "step": 227 }, { "epoch": 0.9230769230769231, "grad_norm": 0.7080859201173951, "learning_rate": 3.914136997982874e-05, "loss": 0.6439, "step": 228 }, { "epoch": 0.9271255060728745, "grad_norm": 0.516305978209034, "learning_rate": 3.912490053436079e-05, "loss": 0.6328, "step": 229 }, { "epoch": 0.9311740890688259, "grad_norm": 0.5837315960684512, "learning_rate": 3.9108278166697805e-05, "loss": 0.6464, "step": 230 }, { "epoch": 0.9352226720647774, "grad_norm": 0.6610507644090599, "learning_rate": 3.909150300975178e-05, "loss": 0.6386, "step": 231 }, { "epoch": 0.9392712550607287, "grad_norm": 0.6716804712773802, "learning_rate": 3.907457519765642e-05, "loss": 0.6342, "step": 232 }, { "epoch": 0.9433198380566802, "grad_norm": 0.5894878513071112, "learning_rate": 3.905749486576607e-05, "loss": 0.6511, "step": 233 }, { "epoch": 0.9473684210526315, "grad_norm": 0.6999203847015966, "learning_rate": 3.90402621506546e-05, "loss": 0.6439, "step": 234 }, { "epoch": 0.951417004048583, "grad_norm": 0.7310722222703663, "learning_rate": 3.902287719011435e-05, "loss": 0.642, "step": 235 }, { "epoch": 0.9554655870445344, "grad_norm": 0.5981042830619328, "learning_rate": 3.900534012315499e-05, "loss": 0.6277, "step": 236 }, { "epoch": 0.9595141700404858, "grad_norm": 0.5070372925575286, "learning_rate": 3.8987651090002454e-05, "loss": 0.6311, "step": 237 }, { "epoch": 0.9635627530364372, "grad_norm": 0.5889083927119861, "learning_rate": 3.896981023209777e-05, "loss": 0.6384, "step": 238 }, { "epoch": 0.9676113360323887, "grad_norm": 0.5942375334677052, "learning_rate": 3.8951817692095966e-05, "loss": 0.6419, "step": 239 }, { "epoch": 0.97165991902834, "grad_norm": 0.6074204926847859, "learning_rate": 3.893367361386491e-05, "loss": 0.6375, "step": 240 }, { "epoch": 0.9757085020242915, "grad_norm": 0.6506991664906648, "learning_rate": 3.891537814248417e-05, "loss": 0.6517, "step": 241 }, { "epoch": 0.979757085020243, "grad_norm": 0.578031198906845, "learning_rate": 3.889693142424384e-05, "loss": 0.644, "step": 242 }, { "epoch": 0.9838056680161943, "grad_norm": 0.5331384351326597, "learning_rate": 3.8878333606643405e-05, "loss": 0.6316, "step": 243 }, { "epoch": 0.9878542510121457, "grad_norm": 0.5181587636021754, "learning_rate": 3.885958483839049e-05, "loss": 0.6398, "step": 244 }, { "epoch": 0.9919028340080972, "grad_norm": 0.48806094386589993, "learning_rate": 3.8840685269399786e-05, "loss": 0.64, "step": 245 }, { "epoch": 0.9959514170040485, "grad_norm": 0.5308181020496036, "learning_rate": 3.882163505079171e-05, "loss": 0.6458, "step": 246 }, { "epoch": 1.0, "grad_norm": 0.48064824092109026, "learning_rate": 3.880243433489132e-05, "loss": 0.6499, "step": 247 }, { "epoch": 1.0040485829959513, "grad_norm": 0.6023759495481228, "learning_rate": 3.878308327522702e-05, "loss": 0.574, "step": 248 }, { "epoch": 1.008097165991903, "grad_norm": 0.5306383882844021, "learning_rate": 3.876358202652936e-05, "loss": 0.5661, "step": 249 }, { "epoch": 1.0121457489878543, "grad_norm": 0.7759353332188046, "learning_rate": 3.8743930744729835e-05, "loss": 0.5625, "step": 250 }, { "epoch": 1.0161943319838056, "grad_norm": 0.7208424976424159, "learning_rate": 3.8724129586959535e-05, "loss": 0.5621, "step": 251 }, { "epoch": 1.0202429149797572, "grad_norm": 0.5989240073685215, "learning_rate": 3.870417871154801e-05, "loss": 0.5606, "step": 252 }, { "epoch": 1.0242914979757085, "grad_norm": 0.7236456119640045, "learning_rate": 3.868407827802192e-05, "loss": 0.5543, "step": 253 }, { "epoch": 1.0283400809716599, "grad_norm": 0.6081199067322328, "learning_rate": 3.8663828447103804e-05, "loss": 0.5765, "step": 254 }, { "epoch": 1.0323886639676114, "grad_norm": 0.5996570791433284, "learning_rate": 3.8643429380710765e-05, "loss": 0.5709, "step": 255 }, { "epoch": 1.0364372469635628, "grad_norm": 0.707004820728192, "learning_rate": 3.862288124195319e-05, "loss": 0.5742, "step": 256 }, { "epoch": 1.040485829959514, "grad_norm": 0.6239959480438237, "learning_rate": 3.8602184195133464e-05, "loss": 0.5652, "step": 257 }, { "epoch": 1.0445344129554657, "grad_norm": 0.7879943330980024, "learning_rate": 3.8581338405744604e-05, "loss": 0.5579, "step": 258 }, { "epoch": 1.048582995951417, "grad_norm": 0.7181855976599116, "learning_rate": 3.8560344040468996e-05, "loss": 0.5781, "step": 259 }, { "epoch": 1.0526315789473684, "grad_norm": 0.7085273512651136, "learning_rate": 3.853920126717703e-05, "loss": 0.5597, "step": 260 }, { "epoch": 1.05668016194332, "grad_norm": 0.7367051401964193, "learning_rate": 3.851791025492574e-05, "loss": 0.5703, "step": 261 }, { "epoch": 1.0607287449392713, "grad_norm": 0.585306708092073, "learning_rate": 3.8496471173957515e-05, "loss": 0.5606, "step": 262 }, { "epoch": 1.0647773279352226, "grad_norm": 0.6440658671647084, "learning_rate": 3.847488419569866e-05, "loss": 0.5523, "step": 263 }, { "epoch": 1.0688259109311742, "grad_norm": 0.6029920016221995, "learning_rate": 3.8453149492758077e-05, "loss": 0.5689, "step": 264 }, { "epoch": 1.0728744939271255, "grad_norm": 0.555800889226909, "learning_rate": 3.8431267238925886e-05, "loss": 0.5558, "step": 265 }, { "epoch": 1.0769230769230769, "grad_norm": 0.5881973011990694, "learning_rate": 3.840923760917199e-05, "loss": 0.5483, "step": 266 }, { "epoch": 1.0809716599190284, "grad_norm": 0.5884180237430069, "learning_rate": 3.838706077964473e-05, "loss": 0.554, "step": 267 }, { "epoch": 1.0850202429149798, "grad_norm": 0.5235035770279134, "learning_rate": 3.836473692766944e-05, "loss": 0.5674, "step": 268 }, { "epoch": 1.0890688259109311, "grad_norm": 0.6215833164355709, "learning_rate": 3.834226623174705e-05, "loss": 0.5502, "step": 269 }, { "epoch": 1.0931174089068827, "grad_norm": 0.695528410433733, "learning_rate": 3.831964887155264e-05, "loss": 0.5648, "step": 270 }, { "epoch": 1.097165991902834, "grad_norm": 0.55427037387917, "learning_rate": 3.829688502793403e-05, "loss": 0.5594, "step": 271 }, { "epoch": 1.1012145748987854, "grad_norm": 1.0320055913133694, "learning_rate": 3.827397488291029e-05, "loss": 0.5739, "step": 272 }, { "epoch": 1.1052631578947367, "grad_norm": 0.9019667772956959, "learning_rate": 3.8250918619670343e-05, "loss": 0.5588, "step": 273 }, { "epoch": 1.1093117408906883, "grad_norm": 0.5793535963019937, "learning_rate": 3.822771642257145e-05, "loss": 0.5544, "step": 274 }, { "epoch": 1.1133603238866396, "grad_norm": 0.8733034487292399, "learning_rate": 3.820436847713776e-05, "loss": 0.5577, "step": 275 }, { "epoch": 1.117408906882591, "grad_norm": 0.8837602039967556, "learning_rate": 3.8180874970058826e-05, "loss": 0.5629, "step": 276 }, { "epoch": 1.1214574898785425, "grad_norm": 0.6133900284967956, "learning_rate": 3.8157236089188103e-05, "loss": 0.5464, "step": 277 }, { "epoch": 1.125506072874494, "grad_norm": 0.8194837377889654, "learning_rate": 3.813345202354145e-05, "loss": 0.5655, "step": 278 }, { "epoch": 1.1295546558704452, "grad_norm": 0.7561671257282725, "learning_rate": 3.810952296329563e-05, "loss": 0.5616, "step": 279 }, { "epoch": 1.1336032388663968, "grad_norm": 0.682467542944483, "learning_rate": 3.808544909978676e-05, "loss": 0.5605, "step": 280 }, { "epoch": 1.1376518218623481, "grad_norm": 0.6250926095950333, "learning_rate": 3.806123062550882e-05, "loss": 0.5769, "step": 281 }, { "epoch": 1.1417004048582995, "grad_norm": 0.8182217336990009, "learning_rate": 3.803686773411208e-05, "loss": 0.5548, "step": 282 }, { "epoch": 1.145748987854251, "grad_norm": 0.7005255958809157, "learning_rate": 3.801236062040158e-05, "loss": 0.5518, "step": 283 }, { "epoch": 1.1497975708502024, "grad_norm": 0.8056073343870679, "learning_rate": 3.798770948033553e-05, "loss": 0.5657, "step": 284 }, { "epoch": 1.1538461538461537, "grad_norm": 0.8144052894968139, "learning_rate": 3.796291451102381e-05, "loss": 0.5621, "step": 285 }, { "epoch": 1.1578947368421053, "grad_norm": 0.6202762088247553, "learning_rate": 3.793797591072631e-05, "loss": 0.5514, "step": 286 }, { "epoch": 1.1619433198380567, "grad_norm": 0.6336741562842518, "learning_rate": 3.791289387885144e-05, "loss": 0.5852, "step": 287 }, { "epoch": 1.165991902834008, "grad_norm": 0.7373233197561097, "learning_rate": 3.788766861595443e-05, "loss": 0.5702, "step": 288 }, { "epoch": 1.1700404858299596, "grad_norm": 0.5666277060654938, "learning_rate": 3.7862300323735835e-05, "loss": 0.559, "step": 289 }, { "epoch": 1.174089068825911, "grad_norm": 0.7145229793311932, "learning_rate": 3.783678920503983e-05, "loss": 0.5616, "step": 290 }, { "epoch": 1.1781376518218623, "grad_norm": 0.7480628977935009, "learning_rate": 3.781113546385265e-05, "loss": 0.5483, "step": 291 }, { "epoch": 1.1821862348178138, "grad_norm": 0.5163896365449409, "learning_rate": 3.7785339305300916e-05, "loss": 0.5577, "step": 292 }, { "epoch": 1.1862348178137652, "grad_norm": 1.021734439278501, "learning_rate": 3.7759400935650025e-05, "loss": 0.5523, "step": 293 }, { "epoch": 1.1902834008097165, "grad_norm": 0.7683993601887955, "learning_rate": 3.77333205623025e-05, "loss": 0.5617, "step": 294 }, { "epoch": 1.194331983805668, "grad_norm": 0.653301831331644, "learning_rate": 3.7707098393796295e-05, "loss": 0.5714, "step": 295 }, { "epoch": 1.1983805668016194, "grad_norm": 0.8754793767151443, "learning_rate": 3.7680734639803185e-05, "loss": 0.5579, "step": 296 }, { "epoch": 1.2024291497975708, "grad_norm": 0.601118009726606, "learning_rate": 3.765422951112704e-05, "loss": 0.5683, "step": 297 }, { "epoch": 1.2064777327935223, "grad_norm": 0.5647711840161407, "learning_rate": 3.762758321970216e-05, "loss": 0.5593, "step": 298 }, { "epoch": 1.2105263157894737, "grad_norm": 0.6799786552552459, "learning_rate": 3.7600795978591584e-05, "loss": 0.5488, "step": 299 }, { "epoch": 1.214574898785425, "grad_norm": 0.5700607519091326, "learning_rate": 3.757386800198538e-05, "loss": 0.5626, "step": 300 }, { "epoch": 1.2186234817813766, "grad_norm": 0.6892281286586782, "learning_rate": 3.7546799505198925e-05, "loss": 0.5497, "step": 301 }, { "epoch": 1.222672064777328, "grad_norm": 0.6854176474321403, "learning_rate": 3.7519590704671197e-05, "loss": 0.5626, "step": 302 }, { "epoch": 1.2267206477732793, "grad_norm": 0.6019672069260693, "learning_rate": 3.749224181796305e-05, "loss": 0.5694, "step": 303 }, { "epoch": 1.2307692307692308, "grad_norm": 0.5301637203371643, "learning_rate": 3.746475306375545e-05, "loss": 0.5766, "step": 304 }, { "epoch": 1.2348178137651822, "grad_norm": 0.6417031188590442, "learning_rate": 3.743712466184774e-05, "loss": 0.5656, "step": 305 }, { "epoch": 1.2388663967611335, "grad_norm": 0.5814498046714109, "learning_rate": 3.7409356833155885e-05, "loss": 0.5616, "step": 306 }, { "epoch": 1.242914979757085, "grad_norm": 0.574006835527926, "learning_rate": 3.73814497997107e-05, "loss": 0.5738, "step": 307 }, { "epoch": 1.2469635627530364, "grad_norm": 0.6577759764561382, "learning_rate": 3.7353403784656095e-05, "loss": 0.5674, "step": 308 }, { "epoch": 1.2510121457489878, "grad_norm": 0.6381152861495564, "learning_rate": 3.732521901224724e-05, "loss": 0.542, "step": 309 }, { "epoch": 1.2550607287449393, "grad_norm": 0.49867792745156364, "learning_rate": 3.7296895707848844e-05, "loss": 0.565, "step": 310 }, { "epoch": 1.2591093117408907, "grad_norm": 0.5507806485829998, "learning_rate": 3.7268434097933275e-05, "loss": 0.5639, "step": 311 }, { "epoch": 1.263157894736842, "grad_norm": 0.48240095518433956, "learning_rate": 3.723983441007882e-05, "loss": 0.5676, "step": 312 }, { "epoch": 1.2672064777327936, "grad_norm": 0.513522472783623, "learning_rate": 3.721109687296781e-05, "loss": 0.5463, "step": 313 }, { "epoch": 1.271255060728745, "grad_norm": 0.5949046825135923, "learning_rate": 3.718222171638484e-05, "loss": 0.5535, "step": 314 }, { "epoch": 1.2753036437246963, "grad_norm": 0.47301715479841877, "learning_rate": 3.715320917121488e-05, "loss": 0.566, "step": 315 }, { "epoch": 1.2793522267206479, "grad_norm": 0.5581779800078297, "learning_rate": 3.712405946944149e-05, "loss": 0.5668, "step": 316 }, { "epoch": 1.2834008097165992, "grad_norm": 0.507516497449386, "learning_rate": 3.709477284414488e-05, "loss": 0.5612, "step": 317 }, { "epoch": 1.2874493927125505, "grad_norm": 0.5494992258582981, "learning_rate": 3.7065349529500136e-05, "loss": 0.5561, "step": 318 }, { "epoch": 1.291497975708502, "grad_norm": 0.559223109032633, "learning_rate": 3.7035789760775305e-05, "loss": 0.5695, "step": 319 }, { "epoch": 1.2955465587044535, "grad_norm": 0.44172927295642006, "learning_rate": 3.70060937743295e-05, "loss": 0.5658, "step": 320 }, { "epoch": 1.2995951417004048, "grad_norm": 0.5144774828180002, "learning_rate": 3.6976261807611045e-05, "loss": 0.552, "step": 321 }, { "epoch": 1.3036437246963564, "grad_norm": 0.5238800407992859, "learning_rate": 3.694629409915555e-05, "loss": 0.5693, "step": 322 }, { "epoch": 1.3076923076923077, "grad_norm": 0.4841087017035178, "learning_rate": 3.691619088858401e-05, "loss": 0.579, "step": 323 }, { "epoch": 1.311740890688259, "grad_norm": 0.5622477503037598, "learning_rate": 3.68859524166009e-05, "loss": 0.5499, "step": 324 }, { "epoch": 1.3157894736842106, "grad_norm": 0.4626960260108458, "learning_rate": 3.6855578924992226e-05, "loss": 0.5681, "step": 325 }, { "epoch": 1.319838056680162, "grad_norm": 0.5597661709785657, "learning_rate": 3.6825070656623626e-05, "loss": 0.5557, "step": 326 }, { "epoch": 1.3238866396761133, "grad_norm": 0.5009348642750153, "learning_rate": 3.67944278554384e-05, "loss": 0.5616, "step": 327 }, { "epoch": 1.3279352226720649, "grad_norm": 0.5682553516233895, "learning_rate": 3.676365076645557e-05, "loss": 0.5577, "step": 328 }, { "epoch": 1.3319838056680162, "grad_norm": 0.8027953172852961, "learning_rate": 3.673273963576791e-05, "loss": 0.549, "step": 329 }, { "epoch": 1.3360323886639676, "grad_norm": 0.5521208472875332, "learning_rate": 3.670169471054001e-05, "loss": 0.5534, "step": 330 }, { "epoch": 1.3400809716599191, "grad_norm": 0.7331901298360598, "learning_rate": 3.6670516239006254e-05, "loss": 0.5519, "step": 331 }, { "epoch": 1.3441295546558705, "grad_norm": 0.7328225442812393, "learning_rate": 3.663920447046888e-05, "loss": 0.5638, "step": 332 }, { "epoch": 1.3481781376518218, "grad_norm": 0.5083130932146668, "learning_rate": 3.6607759655295954e-05, "loss": 0.567, "step": 333 }, { "epoch": 1.3522267206477734, "grad_norm": 0.7061162286527239, "learning_rate": 3.657618204491937e-05, "loss": 0.5678, "step": 334 }, { "epoch": 1.3562753036437247, "grad_norm": 0.7358201840621564, "learning_rate": 3.6544471891832875e-05, "loss": 0.5665, "step": 335 }, { "epoch": 1.360323886639676, "grad_norm": 0.49611706355768265, "learning_rate": 3.6512629449589983e-05, "loss": 0.5692, "step": 336 }, { "epoch": 1.3643724696356276, "grad_norm": 0.7039498384427378, "learning_rate": 3.6480654972802045e-05, "loss": 0.5821, "step": 337 }, { "epoch": 1.368421052631579, "grad_norm": 0.75494774485807, "learning_rate": 3.644854871713611e-05, "loss": 0.5673, "step": 338 }, { "epoch": 1.3724696356275303, "grad_norm": 0.5315797247298614, "learning_rate": 3.6416310939312924e-05, "loss": 0.5621, "step": 339 }, { "epoch": 1.376518218623482, "grad_norm": 0.779562656129718, "learning_rate": 3.638394189710493e-05, "loss": 0.5576, "step": 340 }, { "epoch": 1.3805668016194332, "grad_norm": 0.6997350795352502, "learning_rate": 3.635144184933412e-05, "loss": 0.5566, "step": 341 }, { "epoch": 1.3846153846153846, "grad_norm": 0.5149332241634124, "learning_rate": 3.6318811055869994e-05, "loss": 0.5548, "step": 342 }, { "epoch": 1.3886639676113361, "grad_norm": 0.731665845166358, "learning_rate": 3.628604977762752e-05, "loss": 0.5643, "step": 343 }, { "epoch": 1.3927125506072875, "grad_norm": 0.5016999814159485, "learning_rate": 3.6253158276565004e-05, "loss": 0.5621, "step": 344 }, { "epoch": 1.3967611336032388, "grad_norm": 0.5880192866624583, "learning_rate": 3.6220136815682016e-05, "loss": 0.5673, "step": 345 }, { "epoch": 1.4008097165991904, "grad_norm": 0.5669158089715279, "learning_rate": 3.618698565901727e-05, "loss": 0.5706, "step": 346 }, { "epoch": 1.4048582995951417, "grad_norm": 0.5109715645547244, "learning_rate": 3.615370507164654e-05, "loss": 0.5673, "step": 347 }, { "epoch": 1.408906882591093, "grad_norm": 0.4766761456922858, "learning_rate": 3.61202953196805e-05, "loss": 0.5703, "step": 348 }, { "epoch": 1.4129554655870447, "grad_norm": 0.4966652316797784, "learning_rate": 3.6086756670262654e-05, "loss": 0.5597, "step": 349 }, { "epoch": 1.417004048582996, "grad_norm": 0.4948824658195191, "learning_rate": 3.6053089391567135e-05, "loss": 0.5596, "step": 350 }, { "epoch": 1.4210526315789473, "grad_norm": 0.5033482472753584, "learning_rate": 3.601929375279662e-05, "loss": 0.5658, "step": 351 }, { "epoch": 1.425101214574899, "grad_norm": 0.5949201109984797, "learning_rate": 3.598537002418012e-05, "loss": 0.5748, "step": 352 }, { "epoch": 1.4291497975708503, "grad_norm": 0.5032526285714101, "learning_rate": 3.5951318476970885e-05, "loss": 0.5597, "step": 353 }, { "epoch": 1.4331983805668016, "grad_norm": 0.556409431840639, "learning_rate": 3.591713938344416e-05, "loss": 0.555, "step": 354 }, { "epoch": 1.4372469635627532, "grad_norm": 0.47899653868997444, "learning_rate": 3.588283301689507e-05, "loss": 0.5552, "step": 355 }, { "epoch": 1.4412955465587045, "grad_norm": 0.5254845362011457, "learning_rate": 3.5848399651636424e-05, "loss": 0.5789, "step": 356 }, { "epoch": 1.4453441295546559, "grad_norm": 0.5053963093017578, "learning_rate": 3.5813839562996473e-05, "loss": 0.5686, "step": 357 }, { "epoch": 1.4493927125506074, "grad_norm": 0.5344741925306095, "learning_rate": 3.5779153027316775e-05, "loss": 0.5612, "step": 358 }, { "epoch": 1.4534412955465588, "grad_norm": 0.4523818444931352, "learning_rate": 3.574434032194994e-05, "loss": 0.5629, "step": 359 }, { "epoch": 1.45748987854251, "grad_norm": 0.49123294811802276, "learning_rate": 3.5709401725257436e-05, "loss": 0.5656, "step": 360 }, { "epoch": 1.4615384615384617, "grad_norm": 0.4971951385810586, "learning_rate": 3.5674337516607346e-05, "loss": 0.5743, "step": 361 }, { "epoch": 1.465587044534413, "grad_norm": 0.5593603068595943, "learning_rate": 3.563914797637216e-05, "loss": 0.5555, "step": 362 }, { "epoch": 1.4696356275303644, "grad_norm": 0.5177513883233846, "learning_rate": 3.560383338592649e-05, "loss": 0.5628, "step": 363 }, { "epoch": 1.4736842105263157, "grad_norm": 0.5589656946779594, "learning_rate": 3.556839402764487e-05, "loss": 0.562, "step": 364 }, { "epoch": 1.4777327935222673, "grad_norm": 0.479854297702755, "learning_rate": 3.553283018489946e-05, "loss": 0.5655, "step": 365 }, { "epoch": 1.4817813765182186, "grad_norm": 0.5089893818957172, "learning_rate": 3.54971421420578e-05, "loss": 0.569, "step": 366 }, { "epoch": 1.48582995951417, "grad_norm": 0.5283699755922399, "learning_rate": 3.5461330184480524e-05, "loss": 0.5664, "step": 367 }, { "epoch": 1.4898785425101215, "grad_norm": 0.5230301682585914, "learning_rate": 3.54253945985191e-05, "loss": 0.5597, "step": 368 }, { "epoch": 1.4939271255060729, "grad_norm": 0.5577571529081262, "learning_rate": 3.5389335671513524e-05, "loss": 0.5555, "step": 369 }, { "epoch": 1.4979757085020242, "grad_norm": 0.526479820090276, "learning_rate": 3.5353153691789986e-05, "loss": 0.5673, "step": 370 }, { "epoch": 1.5020242914979756, "grad_norm": 0.5371872487829664, "learning_rate": 3.531684894865866e-05, "loss": 0.5659, "step": 371 }, { "epoch": 1.5060728744939271, "grad_norm": 0.5820383783789912, "learning_rate": 3.528042173241128e-05, "loss": 0.5506, "step": 372 }, { "epoch": 1.5101214574898787, "grad_norm": 0.5004722263495871, "learning_rate": 3.524387233431892e-05, "loss": 0.5622, "step": 373 }, { "epoch": 1.5141700404858298, "grad_norm": 0.4829474833583402, "learning_rate": 3.520720104662958e-05, "loss": 0.5619, "step": 374 }, { "epoch": 1.5182186234817814, "grad_norm": 0.5712906332459863, "learning_rate": 3.5170408162565904e-05, "loss": 0.5513, "step": 375 }, { "epoch": 1.522267206477733, "grad_norm": 0.5932469940060278, "learning_rate": 3.513349397632281e-05, "loss": 0.557, "step": 376 }, { "epoch": 1.526315789473684, "grad_norm": 0.5015596667209188, "learning_rate": 3.5096458783065145e-05, "loss": 0.5652, "step": 377 }, { "epoch": 1.5303643724696356, "grad_norm": 0.6786814246597243, "learning_rate": 3.505930287892533e-05, "loss": 0.5631, "step": 378 }, { "epoch": 1.5344129554655872, "grad_norm": 0.49707467720077725, "learning_rate": 3.502202656100099e-05, "loss": 0.5625, "step": 379 }, { "epoch": 1.5384615384615383, "grad_norm": 0.6154318620842238, "learning_rate": 3.498463012735256e-05, "loss": 0.5657, "step": 380 }, { "epoch": 1.54251012145749, "grad_norm": 0.4124538555203125, "learning_rate": 3.494711387700094e-05, "loss": 0.5643, "step": 381 }, { "epoch": 1.5465587044534415, "grad_norm": 0.6538396876335862, "learning_rate": 3.490947810992505e-05, "loss": 0.5564, "step": 382 }, { "epoch": 1.5506072874493926, "grad_norm": 0.5226009832793632, "learning_rate": 3.48717231270595e-05, "loss": 0.5611, "step": 383 }, { "epoch": 1.5546558704453441, "grad_norm": 0.6551216780566383, "learning_rate": 3.483384923029211e-05, "loss": 0.5767, "step": 384 }, { "epoch": 1.5587044534412957, "grad_norm": 0.6978543595298701, "learning_rate": 3.479585672246156e-05, "loss": 0.5797, "step": 385 }, { "epoch": 1.5627530364372468, "grad_norm": 0.5498489231737607, "learning_rate": 3.4757745907354924e-05, "loss": 0.5653, "step": 386 }, { "epoch": 1.5668016194331984, "grad_norm": 0.6096267947481211, "learning_rate": 3.471951708970527e-05, "loss": 0.5634, "step": 387 }, { "epoch": 1.5708502024291497, "grad_norm": 0.5708831414946063, "learning_rate": 3.468117057518921e-05, "loss": 0.5744, "step": 388 }, { "epoch": 1.574898785425101, "grad_norm": 0.5031942023638978, "learning_rate": 3.4642706670424426e-05, "loss": 0.5719, "step": 389 }, { "epoch": 1.5789473684210527, "grad_norm": 0.5309441335528693, "learning_rate": 3.460412568296731e-05, "loss": 0.5579, "step": 390 }, { "epoch": 1.582995951417004, "grad_norm": 0.5439714600381402, "learning_rate": 3.456542792131039e-05, "loss": 0.5498, "step": 391 }, { "epoch": 1.5870445344129553, "grad_norm": 0.5775959041507577, "learning_rate": 3.4526613694879936e-05, "loss": 0.5688, "step": 392 }, { "epoch": 1.591093117408907, "grad_norm": 0.4811045399244056, "learning_rate": 3.448768331403348e-05, "loss": 0.5564, "step": 393 }, { "epoch": 1.5951417004048583, "grad_norm": 0.5693238682610406, "learning_rate": 3.44486370900573e-05, "loss": 0.5491, "step": 394 }, { "epoch": 1.5991902834008096, "grad_norm": 0.5022900831244657, "learning_rate": 3.440947533516396e-05, "loss": 0.556, "step": 395 }, { "epoch": 1.6032388663967612, "grad_norm": 0.5721025240905128, "learning_rate": 3.4370198362489816e-05, "loss": 0.5636, "step": 396 }, { "epoch": 1.6072874493927125, "grad_norm": 0.47898967599098935, "learning_rate": 3.43308064860925e-05, "loss": 0.5623, "step": 397 }, { "epoch": 1.6113360323886639, "grad_norm": 0.4980170615639608, "learning_rate": 3.4291300020948393e-05, "loss": 0.5436, "step": 398 }, { "epoch": 1.6153846153846154, "grad_norm": 0.48102808575456085, "learning_rate": 3.425167928295015e-05, "loss": 0.5627, "step": 399 }, { "epoch": 1.6194331983805668, "grad_norm": 0.4159063004880129, "learning_rate": 3.421194458890411e-05, "loss": 0.5617, "step": 400 }, { "epoch": 1.623481781376518, "grad_norm": 0.5221267542480305, "learning_rate": 3.4172096256527865e-05, "loss": 0.571, "step": 401 }, { "epoch": 1.6275303643724697, "grad_norm": 0.4440498146949328, "learning_rate": 3.4132134604447595e-05, "loss": 0.5639, "step": 402 }, { "epoch": 1.631578947368421, "grad_norm": 0.4414941193933932, "learning_rate": 3.409205995219561e-05, "loss": 0.5447, "step": 403 }, { "epoch": 1.6356275303643724, "grad_norm": 0.5055801014114144, "learning_rate": 3.4051872620207765e-05, "loss": 0.5695, "step": 404 }, { "epoch": 1.639676113360324, "grad_norm": 0.5087713698010695, "learning_rate": 3.40115729298209e-05, "loss": 0.5751, "step": 405 }, { "epoch": 1.6437246963562753, "grad_norm": 0.4458136831507761, "learning_rate": 3.397116120327025e-05, "loss": 0.5655, "step": 406 }, { "epoch": 1.6477732793522266, "grad_norm": 0.5071087238903562, "learning_rate": 3.3930637763686916e-05, "loss": 0.5733, "step": 407 }, { "epoch": 1.6518218623481782, "grad_norm": 0.41510583286259417, "learning_rate": 3.389000293509524e-05, "loss": 0.559, "step": 408 }, { "epoch": 1.6558704453441295, "grad_norm": 0.48660291336585365, "learning_rate": 3.3849257042410225e-05, "loss": 0.5558, "step": 409 }, { "epoch": 1.6599190283400809, "grad_norm": 0.4713284761406323, "learning_rate": 3.380840041143494e-05, "loss": 0.5633, "step": 410 }, { "epoch": 1.6639676113360324, "grad_norm": 0.4647874465887159, "learning_rate": 3.376743336885792e-05, "loss": 0.5606, "step": 411 }, { "epoch": 1.6680161943319838, "grad_norm": 0.6015545906429657, "learning_rate": 3.372635624225055e-05, "loss": 0.5755, "step": 412 }, { "epoch": 1.6720647773279351, "grad_norm": 0.4420585622508115, "learning_rate": 3.3685169360064445e-05, "loss": 0.5608, "step": 413 }, { "epoch": 1.6761133603238867, "grad_norm": 0.5050701306700736, "learning_rate": 3.3643873051628804e-05, "loss": 0.5626, "step": 414 }, { "epoch": 1.680161943319838, "grad_norm": 0.5958344465073608, "learning_rate": 3.3602467647147814e-05, "loss": 0.5667, "step": 415 }, { "epoch": 1.6842105263157894, "grad_norm": 0.4647858354240761, "learning_rate": 3.356095347769799e-05, "loss": 0.5419, "step": 416 }, { "epoch": 1.688259109311741, "grad_norm": 0.5973937599293945, "learning_rate": 3.351933087522552e-05, "loss": 0.5428, "step": 417 }, { "epoch": 1.6923076923076923, "grad_norm": 0.478884991491656, "learning_rate": 3.3477600172543626e-05, "loss": 0.5773, "step": 418 }, { "epoch": 1.6963562753036436, "grad_norm": 0.6159391345555267, "learning_rate": 3.3435761703329894e-05, "loss": 0.5705, "step": 419 }, { "epoch": 1.7004048582995952, "grad_norm": 0.4649459648706748, "learning_rate": 3.3393815802123606e-05, "loss": 0.5534, "step": 420 }, { "epoch": 1.7044534412955465, "grad_norm": 0.5778701958947005, "learning_rate": 3.335176280432308e-05, "loss": 0.5616, "step": 421 }, { "epoch": 1.708502024291498, "grad_norm": 0.42431271845243973, "learning_rate": 3.330960304618295e-05, "loss": 0.5685, "step": 422 }, { "epoch": 1.7125506072874495, "grad_norm": 0.4900770660367405, "learning_rate": 3.326733686481153e-05, "loss": 0.5529, "step": 423 }, { "epoch": 1.7165991902834008, "grad_norm": 0.45909619909251786, "learning_rate": 3.3224964598168084e-05, "loss": 0.561, "step": 424 }, { "epoch": 1.7206477732793521, "grad_norm": 0.4732112150961493, "learning_rate": 3.318248658506012e-05, "loss": 0.5618, "step": 425 }, { "epoch": 1.7246963562753037, "grad_norm": 0.40222671331373167, "learning_rate": 3.313990316514069e-05, "loss": 0.5623, "step": 426 }, { "epoch": 1.728744939271255, "grad_norm": 0.47940861911331073, "learning_rate": 3.309721467890571e-05, "loss": 0.5462, "step": 427 }, { "epoch": 1.7327935222672064, "grad_norm": 0.47265250726015584, "learning_rate": 3.305442146769115e-05, "loss": 0.5743, "step": 428 }, { "epoch": 1.736842105263158, "grad_norm": 0.4536873766205685, "learning_rate": 3.30115238736704e-05, "loss": 0.5574, "step": 429 }, { "epoch": 1.7408906882591093, "grad_norm": 0.5539872676565223, "learning_rate": 3.296852223985148e-05, "loss": 0.562, "step": 430 }, { "epoch": 1.7449392712550607, "grad_norm": 0.4722733743822937, "learning_rate": 3.2925416910074315e-05, "loss": 0.5738, "step": 431 }, { "epoch": 1.7489878542510122, "grad_norm": 0.5877080492313781, "learning_rate": 3.288220822900796e-05, "loss": 0.5608, "step": 432 }, { "epoch": 1.7530364372469636, "grad_norm": 0.44761065953406504, "learning_rate": 3.283889654214788e-05, "loss": 0.5641, "step": 433 }, { "epoch": 1.757085020242915, "grad_norm": 0.7085237836935455, "learning_rate": 3.279548219581317e-05, "loss": 0.5659, "step": 434 }, { "epoch": 1.7611336032388665, "grad_norm": 0.487594296253294, "learning_rate": 3.275196553714379e-05, "loss": 0.5574, "step": 435 }, { "epoch": 1.7651821862348178, "grad_norm": 0.6547787041065113, "learning_rate": 3.270834691409778e-05, "loss": 0.5711, "step": 436 }, { "epoch": 1.7692307692307692, "grad_norm": 0.5991982708286507, "learning_rate": 3.266462667544849e-05, "loss": 0.5652, "step": 437 }, { "epoch": 1.7732793522267207, "grad_norm": 0.6133901203643713, "learning_rate": 3.262080517078178e-05, "loss": 0.5755, "step": 438 }, { "epoch": 1.777327935222672, "grad_norm": 0.489886496734961, "learning_rate": 3.257688275049323e-05, "loss": 0.568, "step": 439 }, { "epoch": 1.7813765182186234, "grad_norm": 0.690425026070907, "learning_rate": 3.253285976578535e-05, "loss": 0.5705, "step": 440 }, { "epoch": 1.785425101214575, "grad_norm": 0.48432536404571663, "learning_rate": 3.2488736568664756e-05, "loss": 0.5729, "step": 441 }, { "epoch": 1.7894736842105263, "grad_norm": 0.6219389122732883, "learning_rate": 3.244451351193935e-05, "loss": 0.5701, "step": 442 }, { "epoch": 1.7935222672064777, "grad_norm": 0.5773875494846213, "learning_rate": 3.240019094921551e-05, "loss": 0.5684, "step": 443 }, { "epoch": 1.7975708502024292, "grad_norm": 0.6101713056875694, "learning_rate": 3.235576923489527e-05, "loss": 0.5476, "step": 444 }, { "epoch": 1.8016194331983806, "grad_norm": 0.6409445757911334, "learning_rate": 3.2311248724173476e-05, "loss": 0.5565, "step": 445 }, { "epoch": 1.805668016194332, "grad_norm": 0.6122625489166813, "learning_rate": 3.226662977303494e-05, "loss": 0.5633, "step": 446 }, { "epoch": 1.8097165991902835, "grad_norm": 0.635373104334155, "learning_rate": 3.2221912738251626e-05, "loss": 0.5728, "step": 447 }, { "epoch": 1.8137651821862348, "grad_norm": 0.573149767421344, "learning_rate": 3.2177097977379736e-05, "loss": 0.5659, "step": 448 }, { "epoch": 1.8178137651821862, "grad_norm": 0.7320850523442965, "learning_rate": 3.21321858487569e-05, "loss": 0.5678, "step": 449 }, { "epoch": 1.8218623481781377, "grad_norm": 0.601704040698926, "learning_rate": 3.208717671149931e-05, "loss": 0.5611, "step": 450 }, { "epoch": 1.825910931174089, "grad_norm": 0.7009937918001687, "learning_rate": 3.204207092549882e-05, "loss": 0.5636, "step": 451 }, { "epoch": 1.8299595141700404, "grad_norm": 0.544297863746043, "learning_rate": 3.19968688514201e-05, "loss": 0.5699, "step": 452 }, { "epoch": 1.834008097165992, "grad_norm": 0.6497685335364936, "learning_rate": 3.195157085069774e-05, "loss": 0.5635, "step": 453 }, { "epoch": 1.8380566801619433, "grad_norm": 0.5116191941811591, "learning_rate": 3.190617728553332e-05, "loss": 0.5837, "step": 454 }, { "epoch": 1.8421052631578947, "grad_norm": 0.5442097302306264, "learning_rate": 3.1860688518892606e-05, "loss": 0.5672, "step": 455 }, { "epoch": 1.8461538461538463, "grad_norm": 0.582577794040474, "learning_rate": 3.181510491450254e-05, "loss": 0.5655, "step": 456 }, { "epoch": 1.8502024291497976, "grad_norm": 0.5023044677396404, "learning_rate": 3.176942683684842e-05, "loss": 0.5588, "step": 457 }, { "epoch": 1.854251012145749, "grad_norm": 0.48984877474287236, "learning_rate": 3.1723654651170934e-05, "loss": 0.5629, "step": 458 }, { "epoch": 1.8582995951417005, "grad_norm": 0.4886694515314368, "learning_rate": 3.167778872346327e-05, "loss": 0.5494, "step": 459 }, { "epoch": 1.8623481781376519, "grad_norm": 0.4846138935004901, "learning_rate": 3.163182942046815e-05, "loss": 0.5602, "step": 460 }, { "epoch": 1.8663967611336032, "grad_norm": 0.5421868554974124, "learning_rate": 3.158577710967494e-05, "loss": 0.5638, "step": 461 }, { "epoch": 1.8704453441295548, "grad_norm": 0.4666246319702372, "learning_rate": 3.153963215931669e-05, "loss": 0.5463, "step": 462 }, { "epoch": 1.874493927125506, "grad_norm": 0.5345218861401371, "learning_rate": 3.14933949383672e-05, "loss": 0.5577, "step": 463 }, { "epoch": 1.8785425101214575, "grad_norm": 0.4837909515920889, "learning_rate": 3.144706581653805e-05, "loss": 0.5566, "step": 464 }, { "epoch": 1.882591093117409, "grad_norm": 0.48593639071450956, "learning_rate": 3.140064516427566e-05, "loss": 0.5599, "step": 465 }, { "epoch": 1.8866396761133604, "grad_norm": 0.42267264808096466, "learning_rate": 3.135413335275833e-05, "loss": 0.5535, "step": 466 }, { "epoch": 1.8906882591093117, "grad_norm": 0.5067584990845313, "learning_rate": 3.130753075389327e-05, "loss": 0.5618, "step": 467 }, { "epoch": 1.8947368421052633, "grad_norm": 0.49195557337769535, "learning_rate": 3.1260837740313625e-05, "loss": 0.5404, "step": 468 }, { "epoch": 1.8987854251012146, "grad_norm": 0.48694038561732145, "learning_rate": 3.121405468537547e-05, "loss": 0.5601, "step": 469 }, { "epoch": 1.902834008097166, "grad_norm": 0.5023926690777727, "learning_rate": 3.116718196315488e-05, "loss": 0.5654, "step": 470 }, { "epoch": 1.9068825910931175, "grad_norm": 0.4506754288870844, "learning_rate": 3.1120219948444894e-05, "loss": 0.5557, "step": 471 }, { "epoch": 1.9109311740890689, "grad_norm": 0.479506441371951, "learning_rate": 3.107316901675252e-05, "loss": 0.5643, "step": 472 }, { "epoch": 1.9149797570850202, "grad_norm": 0.4520225779706977, "learning_rate": 3.1026029544295745e-05, "loss": 0.5627, "step": 473 }, { "epoch": 1.9190283400809718, "grad_norm": 0.4298984097022692, "learning_rate": 3.097880190800057e-05, "loss": 0.567, "step": 474 }, { "epoch": 1.9230769230769231, "grad_norm": 0.47303083341565944, "learning_rate": 3.093148648549788e-05, "loss": 0.5604, "step": 475 }, { "epoch": 1.9271255060728745, "grad_norm": 0.46139557340284704, "learning_rate": 3.088408365512055e-05, "loss": 0.5517, "step": 476 }, { "epoch": 1.931174089068826, "grad_norm": 0.5530954355104039, "learning_rate": 3.083659379590034e-05, "loss": 0.5652, "step": 477 }, { "epoch": 1.9352226720647774, "grad_norm": 0.48637983746180763, "learning_rate": 3.0789017287564913e-05, "loss": 0.5461, "step": 478 }, { "epoch": 1.9392712550607287, "grad_norm": 0.478139317564781, "learning_rate": 3.074135451053475e-05, "loss": 0.5587, "step": 479 }, { "epoch": 1.9433198380566803, "grad_norm": 0.4476366535023883, "learning_rate": 3.069360584592015e-05, "loss": 0.5636, "step": 480 }, { "epoch": 1.9473684210526314, "grad_norm": 0.43622427859045926, "learning_rate": 3.064577167551816e-05, "loss": 0.5532, "step": 481 }, { "epoch": 1.951417004048583, "grad_norm": 0.43098676137358183, "learning_rate": 3.059785238180953e-05, "loss": 0.5642, "step": 482 }, { "epoch": 1.9554655870445345, "grad_norm": 0.45359395239194683, "learning_rate": 3.054984834795566e-05, "loss": 0.5606, "step": 483 }, { "epoch": 1.9595141700404857, "grad_norm": 0.44686023384821094, "learning_rate": 3.0501759957795537e-05, "loss": 0.5598, "step": 484 }, { "epoch": 1.9635627530364372, "grad_norm": 0.4152548111826416, "learning_rate": 3.0453587595842627e-05, "loss": 0.5459, "step": 485 }, { "epoch": 1.9676113360323888, "grad_norm": 0.40256653337572185, "learning_rate": 3.0405331647281856e-05, "loss": 0.5691, "step": 486 }, { "epoch": 1.97165991902834, "grad_norm": 0.40797753512469886, "learning_rate": 3.0356992497966508e-05, "loss": 0.5622, "step": 487 }, { "epoch": 1.9757085020242915, "grad_norm": 0.40725740634813734, "learning_rate": 3.0308570534415137e-05, "loss": 0.5579, "step": 488 }, { "epoch": 1.979757085020243, "grad_norm": 0.44480242862612507, "learning_rate": 3.0260066143808467e-05, "loss": 0.5741, "step": 489 }, { "epoch": 1.9838056680161942, "grad_norm": 0.45609349200118826, "learning_rate": 3.021147971398632e-05, "loss": 0.5507, "step": 490 }, { "epoch": 1.9878542510121457, "grad_norm": 0.4884077772039339, "learning_rate": 3.0162811633444478e-05, "loss": 0.5693, "step": 491 }, { "epoch": 1.9919028340080973, "grad_norm": 0.41043782725844413, "learning_rate": 3.0114062291331635e-05, "loss": 0.5596, "step": 492 }, { "epoch": 1.9959514170040484, "grad_norm": 0.5076428529147672, "learning_rate": 3.0065232077446217e-05, "loss": 0.5644, "step": 493 }, { "epoch": 2.0, "grad_norm": 0.405814367197076, "learning_rate": 3.001632138223332e-05, "loss": 0.5512, "step": 494 }, { "epoch": 2.0040485829959516, "grad_norm": 0.6806348932791161, "learning_rate": 2.9967330596781565e-05, "loss": 0.4463, "step": 495 }, { "epoch": 2.0080971659919027, "grad_norm": 0.8438820197838203, "learning_rate": 2.9918260112819952e-05, "loss": 0.449, "step": 496 }, { "epoch": 2.0121457489878543, "grad_norm": 0.5126844831733204, "learning_rate": 2.9869110322714778e-05, "loss": 0.45, "step": 497 }, { "epoch": 2.016194331983806, "grad_norm": 0.5735687765056449, "learning_rate": 2.9819881619466447e-05, "loss": 0.4492, "step": 498 }, { "epoch": 2.020242914979757, "grad_norm": 0.5048663910829285, "learning_rate": 2.9770574396706364e-05, "loss": 0.4534, "step": 499 }, { "epoch": 2.0242914979757085, "grad_norm": 0.5734801031769581, "learning_rate": 2.972118904869376e-05, "loss": 0.4488, "step": 500 }, { "epoch": 2.02834008097166, "grad_norm": 0.4824321061798042, "learning_rate": 2.9671725970312563e-05, "loss": 0.4415, "step": 501 }, { "epoch": 2.032388663967611, "grad_norm": 0.5056192682235657, "learning_rate": 2.9622185557068238e-05, "loss": 0.4457, "step": 502 }, { "epoch": 2.0364372469635628, "grad_norm": 0.5122715694747858, "learning_rate": 2.95725682050846e-05, "loss": 0.442, "step": 503 }, { "epoch": 2.0404858299595143, "grad_norm": 0.5262481856523125, "learning_rate": 2.9522874311100684e-05, "loss": 0.4637, "step": 504 }, { "epoch": 2.0445344129554655, "grad_norm": 0.47815128017833003, "learning_rate": 2.9473104272467526e-05, "loss": 0.4326, "step": 505 }, { "epoch": 2.048582995951417, "grad_norm": 0.482799565113372, "learning_rate": 2.942325848714505e-05, "loss": 0.4403, "step": 506 }, { "epoch": 2.0526315789473686, "grad_norm": 0.4847463994586617, "learning_rate": 2.9373337353698804e-05, "loss": 0.431, "step": 507 }, { "epoch": 2.0566801619433197, "grad_norm": 0.5159461153170337, "learning_rate": 2.932334127129686e-05, "loss": 0.4428, "step": 508 }, { "epoch": 2.0607287449392713, "grad_norm": 0.5146840754407, "learning_rate": 2.9273270639706547e-05, "loss": 0.4537, "step": 509 }, { "epoch": 2.064777327935223, "grad_norm": 0.5175341891657566, "learning_rate": 2.922312585929131e-05, "loss": 0.4378, "step": 510 }, { "epoch": 2.068825910931174, "grad_norm": 0.5056406916033661, "learning_rate": 2.9172907331007478e-05, "loss": 0.4495, "step": 511 }, { "epoch": 2.0728744939271255, "grad_norm": 0.519331087043867, "learning_rate": 2.9122615456401053e-05, "loss": 0.453, "step": 512 }, { "epoch": 2.076923076923077, "grad_norm": 0.5512301936237192, "learning_rate": 2.9072250637604535e-05, "loss": 0.4472, "step": 513 }, { "epoch": 2.080971659919028, "grad_norm": 0.48092347010362035, "learning_rate": 2.9021813277333667e-05, "loss": 0.44, "step": 514 }, { "epoch": 2.08502024291498, "grad_norm": 0.5321469185503414, "learning_rate": 2.8971303778884244e-05, "loss": 0.4472, "step": 515 }, { "epoch": 2.0890688259109313, "grad_norm": 0.48139611603234206, "learning_rate": 2.8920722546128863e-05, "loss": 0.4358, "step": 516 }, { "epoch": 2.0931174089068825, "grad_norm": 0.4973022256254829, "learning_rate": 2.887006998351372e-05, "loss": 0.4625, "step": 517 }, { "epoch": 2.097165991902834, "grad_norm": 0.4393623536944921, "learning_rate": 2.881934649605535e-05, "loss": 0.4401, "step": 518 }, { "epoch": 2.1012145748987856, "grad_norm": 0.5508996981541542, "learning_rate": 2.8768552489337408e-05, "loss": 0.4433, "step": 519 }, { "epoch": 2.1052631578947367, "grad_norm": 0.49573890003282123, "learning_rate": 2.8717688369507425e-05, "loss": 0.4475, "step": 520 }, { "epoch": 2.1093117408906883, "grad_norm": 0.5104443503157965, "learning_rate": 2.8666754543273535e-05, "loss": 0.4294, "step": 521 }, { "epoch": 2.11336032388664, "grad_norm": 0.4415549672452191, "learning_rate": 2.8615751417901273e-05, "loss": 0.4304, "step": 522 }, { "epoch": 2.117408906882591, "grad_norm": 0.46407074347071486, "learning_rate": 2.856467940121025e-05, "loss": 0.461, "step": 523 }, { "epoch": 2.1214574898785425, "grad_norm": 0.5244792759912424, "learning_rate": 2.8513538901570964e-05, "loss": 0.4489, "step": 524 }, { "epoch": 2.125506072874494, "grad_norm": 0.46664775576985995, "learning_rate": 2.846233032790149e-05, "loss": 0.429, "step": 525 }, { "epoch": 2.1295546558704452, "grad_norm": 0.5157238521169747, "learning_rate": 2.84110540896642e-05, "loss": 0.4567, "step": 526 }, { "epoch": 2.133603238866397, "grad_norm": 0.48028728039196605, "learning_rate": 2.8359710596862555e-05, "loss": 0.4417, "step": 527 }, { "epoch": 2.1376518218623484, "grad_norm": 0.529230291820357, "learning_rate": 2.8308300260037734e-05, "loss": 0.4384, "step": 528 }, { "epoch": 2.1417004048582995, "grad_norm": 0.47694595853747157, "learning_rate": 2.8256823490265424e-05, "loss": 0.4457, "step": 529 }, { "epoch": 2.145748987854251, "grad_norm": 0.46353982033509633, "learning_rate": 2.8205280699152515e-05, "loss": 0.4387, "step": 530 }, { "epoch": 2.1497975708502026, "grad_norm": 0.45344245197869326, "learning_rate": 2.8153672298833777e-05, "loss": 0.4406, "step": 531 }, { "epoch": 2.1538461538461537, "grad_norm": 0.6048709542336242, "learning_rate": 2.8101998701968615e-05, "loss": 0.4644, "step": 532 }, { "epoch": 2.1578947368421053, "grad_norm": 0.5491597442333501, "learning_rate": 2.805026032173773e-05, "loss": 0.4457, "step": 533 }, { "epoch": 2.161943319838057, "grad_norm": 0.4905801801299945, "learning_rate": 2.7998457571839832e-05, "loss": 0.4495, "step": 534 }, { "epoch": 2.165991902834008, "grad_norm": 0.5050802893240358, "learning_rate": 2.7946590866488322e-05, "loss": 0.4524, "step": 535 }, { "epoch": 2.1700404858299596, "grad_norm": 0.41171896751046105, "learning_rate": 2.7894660620408e-05, "loss": 0.4461, "step": 536 }, { "epoch": 2.174089068825911, "grad_norm": 0.4827821802209062, "learning_rate": 2.784266724883173e-05, "loss": 0.4598, "step": 537 }, { "epoch": 2.1781376518218623, "grad_norm": 0.4331319838821782, "learning_rate": 2.7790611167497128e-05, "loss": 0.4494, "step": 538 }, { "epoch": 2.182186234817814, "grad_norm": 0.4580873973005053, "learning_rate": 2.7738492792643224e-05, "loss": 0.4457, "step": 539 }, { "epoch": 2.1862348178137654, "grad_norm": 0.43616521038300377, "learning_rate": 2.768631254100716e-05, "loss": 0.4543, "step": 540 }, { "epoch": 2.1902834008097165, "grad_norm": 0.4541652395368612, "learning_rate": 2.7634070829820836e-05, "loss": 0.4418, "step": 541 }, { "epoch": 2.194331983805668, "grad_norm": 0.4658938505479704, "learning_rate": 2.7581768076807592e-05, "loss": 0.4365, "step": 542 }, { "epoch": 2.1983805668016196, "grad_norm": 0.44051416729591036, "learning_rate": 2.7529404700178832e-05, "loss": 0.4341, "step": 543 }, { "epoch": 2.2024291497975708, "grad_norm": 0.4482153648322522, "learning_rate": 2.7476981118630735e-05, "loss": 0.4486, "step": 544 }, { "epoch": 2.2064777327935223, "grad_norm": 0.4542285718977572, "learning_rate": 2.742449775134085e-05, "loss": 0.4447, "step": 545 }, { "epoch": 2.2105263157894735, "grad_norm": 0.45629133975144237, "learning_rate": 2.7371955017964788e-05, "loss": 0.4414, "step": 546 }, { "epoch": 2.214574898785425, "grad_norm": 0.4512585931116791, "learning_rate": 2.7319353338632845e-05, "loss": 0.4479, "step": 547 }, { "epoch": 2.2186234817813766, "grad_norm": 0.45078806416864137, "learning_rate": 2.7266693133946645e-05, "loss": 0.4427, "step": 548 }, { "epoch": 2.2226720647773277, "grad_norm": 0.47898606894007734, "learning_rate": 2.7213974824975785e-05, "loss": 0.4355, "step": 549 }, { "epoch": 2.2267206477732793, "grad_norm": 0.40082850662176406, "learning_rate": 2.7161198833254464e-05, "loss": 0.4538, "step": 550 }, { "epoch": 2.230769230769231, "grad_norm": 0.4880967749798117, "learning_rate": 2.7108365580778095e-05, "loss": 0.4575, "step": 551 }, { "epoch": 2.234817813765182, "grad_norm": 0.45376936402895945, "learning_rate": 2.705547548999996e-05, "loss": 0.4573, "step": 552 }, { "epoch": 2.2388663967611335, "grad_norm": 0.5140958110730779, "learning_rate": 2.7002528983827817e-05, "loss": 0.4421, "step": 553 }, { "epoch": 2.242914979757085, "grad_norm": 0.3975975644908091, "learning_rate": 2.6949526485620517e-05, "loss": 0.4529, "step": 554 }, { "epoch": 2.246963562753036, "grad_norm": 0.4597985915663013, "learning_rate": 2.6896468419184623e-05, "loss": 0.454, "step": 555 }, { "epoch": 2.251012145748988, "grad_norm": 0.43979822723016615, "learning_rate": 2.6843355208771024e-05, "loss": 0.4441, "step": 556 }, { "epoch": 2.2550607287449393, "grad_norm": 0.4480622267374088, "learning_rate": 2.6790187279071536e-05, "loss": 0.4375, "step": 557 }, { "epoch": 2.2591093117408905, "grad_norm": 0.43007277202563066, "learning_rate": 2.67369650552155e-05, "loss": 0.4414, "step": 558 }, { "epoch": 2.263157894736842, "grad_norm": 0.3832242502834972, "learning_rate": 2.6683688962766407e-05, "loss": 0.4416, "step": 559 }, { "epoch": 2.2672064777327936, "grad_norm": 0.4630710001462427, "learning_rate": 2.6630359427718468e-05, "loss": 0.4462, "step": 560 }, { "epoch": 2.2712550607287447, "grad_norm": 0.42069201742837625, "learning_rate": 2.6576976876493225e-05, "loss": 0.4577, "step": 561 }, { "epoch": 2.2753036437246963, "grad_norm": 0.46371354227860484, "learning_rate": 2.652354173593614e-05, "loss": 0.4426, "step": 562 }, { "epoch": 2.279352226720648, "grad_norm": 0.4306448169728055, "learning_rate": 2.647005443331317e-05, "loss": 0.4558, "step": 563 }, { "epoch": 2.283400809716599, "grad_norm": 0.4752007992124324, "learning_rate": 2.6416515396307354e-05, "loss": 0.4361, "step": 564 }, { "epoch": 2.2874493927125505, "grad_norm": 0.43686529643790606, "learning_rate": 2.636292505301542e-05, "loss": 0.4468, "step": 565 }, { "epoch": 2.291497975708502, "grad_norm": 0.48269063357926384, "learning_rate": 2.630928383194431e-05, "loss": 0.4424, "step": 566 }, { "epoch": 2.2955465587044532, "grad_norm": 0.4138727289165304, "learning_rate": 2.625559216200781e-05, "loss": 0.4355, "step": 567 }, { "epoch": 2.299595141700405, "grad_norm": 0.4641126733461425, "learning_rate": 2.620185047252308e-05, "loss": 0.4478, "step": 568 }, { "epoch": 2.3036437246963564, "grad_norm": 0.3907283375787215, "learning_rate": 2.614805919320723e-05, "loss": 0.4517, "step": 569 }, { "epoch": 2.3076923076923075, "grad_norm": 0.41165204068083566, "learning_rate": 2.6094218754173895e-05, "loss": 0.4525, "step": 570 }, { "epoch": 2.311740890688259, "grad_norm": 0.3985629556773705, "learning_rate": 2.60403295859298e-05, "loss": 0.4431, "step": 571 }, { "epoch": 2.3157894736842106, "grad_norm": 0.4375195512267774, "learning_rate": 2.5986392119371297e-05, "loss": 0.4507, "step": 572 }, { "epoch": 2.3198380566801617, "grad_norm": 0.403704808945473, "learning_rate": 2.5932406785780928e-05, "loss": 0.4386, "step": 573 }, { "epoch": 2.3238866396761133, "grad_norm": 0.45371852314779854, "learning_rate": 2.5878374016823987e-05, "loss": 0.442, "step": 574 }, { "epoch": 2.327935222672065, "grad_norm": 0.4240294457041315, "learning_rate": 2.5824294244545045e-05, "loss": 0.4472, "step": 575 }, { "epoch": 2.331983805668016, "grad_norm": 0.4491580757000291, "learning_rate": 2.577016790136453e-05, "loss": 0.4508, "step": 576 }, { "epoch": 2.3360323886639676, "grad_norm": 0.5112077869834296, "learning_rate": 2.5715995420075225e-05, "loss": 0.4464, "step": 577 }, { "epoch": 2.340080971659919, "grad_norm": 0.4888080574949571, "learning_rate": 2.5661777233838864e-05, "loss": 0.4444, "step": 578 }, { "epoch": 2.3441295546558703, "grad_norm": 0.47743460291600337, "learning_rate": 2.5607513776182616e-05, "loss": 0.4499, "step": 579 }, { "epoch": 2.348178137651822, "grad_norm": 0.5329469710129096, "learning_rate": 2.5553205480995633e-05, "loss": 0.4549, "step": 580 }, { "epoch": 2.3522267206477734, "grad_norm": 0.4827744117065264, "learning_rate": 2.5498852782525604e-05, "loss": 0.4454, "step": 581 }, { "epoch": 2.3562753036437245, "grad_norm": 0.4662974777653642, "learning_rate": 2.5444456115375252e-05, "loss": 0.4549, "step": 582 }, { "epoch": 2.360323886639676, "grad_norm": 0.4686343506674191, "learning_rate": 2.539001591449889e-05, "loss": 0.4416, "step": 583 }, { "epoch": 2.3643724696356276, "grad_norm": 0.5458358777510788, "learning_rate": 2.53355326151989e-05, "loss": 0.4319, "step": 584 }, { "epoch": 2.3684210526315788, "grad_norm": 0.5039088919997798, "learning_rate": 2.5281006653122288e-05, "loss": 0.4375, "step": 585 }, { "epoch": 2.3724696356275303, "grad_norm": 0.5747865758569068, "learning_rate": 2.5226438464257217e-05, "loss": 0.4477, "step": 586 }, { "epoch": 2.376518218623482, "grad_norm": 0.48868629620264153, "learning_rate": 2.5171828484929447e-05, "loss": 0.4521, "step": 587 }, { "epoch": 2.380566801619433, "grad_norm": 0.533240876633436, "learning_rate": 2.5117177151798934e-05, "loss": 0.4419, "step": 588 }, { "epoch": 2.3846153846153846, "grad_norm": 0.4896294077391343, "learning_rate": 2.506248490185627e-05, "loss": 0.4442, "step": 589 }, { "epoch": 2.388663967611336, "grad_norm": 0.4993396231548543, "learning_rate": 2.500775217241923e-05, "loss": 0.4643, "step": 590 }, { "epoch": 2.3927125506072873, "grad_norm": 0.5212987427174546, "learning_rate": 2.4952979401129266e-05, "loss": 0.4426, "step": 591 }, { "epoch": 2.396761133603239, "grad_norm": 0.4310687421186266, "learning_rate": 2.4898167025948e-05, "loss": 0.4349, "step": 592 }, { "epoch": 2.4008097165991904, "grad_norm": 0.5408612865787172, "learning_rate": 2.4843315485153703e-05, "loss": 0.4543, "step": 593 }, { "epoch": 2.4048582995951415, "grad_norm": 0.4140906913487582, "learning_rate": 2.4788425217337856e-05, "loss": 0.444, "step": 594 }, { "epoch": 2.408906882591093, "grad_norm": 0.5659992362134582, "learning_rate": 2.4733496661401572e-05, "loss": 0.4588, "step": 595 }, { "epoch": 2.4129554655870447, "grad_norm": 0.4287957665659547, "learning_rate": 2.4678530256552114e-05, "loss": 0.4529, "step": 596 }, { "epoch": 2.417004048582996, "grad_norm": 0.5365639380838007, "learning_rate": 2.46235264422994e-05, "loss": 0.4458, "step": 597 }, { "epoch": 2.4210526315789473, "grad_norm": 0.4783562074174893, "learning_rate": 2.4568485658452452e-05, "loss": 0.4484, "step": 598 }, { "epoch": 2.425101214574899, "grad_norm": 0.47917893288145386, "learning_rate": 2.451340834511592e-05, "loss": 0.448, "step": 599 }, { "epoch": 2.42914979757085, "grad_norm": 0.44334744835598583, "learning_rate": 2.4458294942686528e-05, "loss": 0.4391, "step": 600 }, { "epoch": 2.4331983805668016, "grad_norm": 0.43363475557085035, "learning_rate": 2.440314589184958e-05, "loss": 0.4381, "step": 601 }, { "epoch": 2.437246963562753, "grad_norm": 0.4043758774133227, "learning_rate": 2.434796163357541e-05, "loss": 0.4437, "step": 602 }, { "epoch": 2.4412955465587043, "grad_norm": 0.4233170040105381, "learning_rate": 2.429274260911588e-05, "loss": 0.44, "step": 603 }, { "epoch": 2.445344129554656, "grad_norm": 0.45698496342971584, "learning_rate": 2.423748926000084e-05, "loss": 0.4389, "step": 604 }, { "epoch": 2.4493927125506074, "grad_norm": 0.3962770941236353, "learning_rate": 2.418220202803459e-05, "loss": 0.4456, "step": 605 }, { "epoch": 2.4534412955465585, "grad_norm": 0.4363423750360274, "learning_rate": 2.412688135529237e-05, "loss": 0.4547, "step": 606 }, { "epoch": 2.45748987854251, "grad_norm": 0.4012716283812568, "learning_rate": 2.4071527684116807e-05, "loss": 0.4369, "step": 607 }, { "epoch": 2.4615384615384617, "grad_norm": 0.4302368411842425, "learning_rate": 2.4016141457114372e-05, "loss": 0.4601, "step": 608 }, { "epoch": 2.465587044534413, "grad_norm": 0.5351624780637227, "learning_rate": 2.396072311715187e-05, "loss": 0.4599, "step": 609 }, { "epoch": 2.4696356275303644, "grad_norm": 0.4385520238664781, "learning_rate": 2.3905273107352863e-05, "loss": 0.4467, "step": 610 }, { "epoch": 2.473684210526316, "grad_norm": 0.5182557516384074, "learning_rate": 2.3849791871094157e-05, "loss": 0.4407, "step": 611 }, { "epoch": 2.477732793522267, "grad_norm": 0.47543393751194163, "learning_rate": 2.379427985200223e-05, "loss": 0.4577, "step": 612 }, { "epoch": 2.4817813765182186, "grad_norm": 0.519168570235394, "learning_rate": 2.373873749394972e-05, "loss": 0.4632, "step": 613 }, { "epoch": 2.48582995951417, "grad_norm": 0.4429786663154713, "learning_rate": 2.3683165241051836e-05, "loss": 0.4454, "step": 614 }, { "epoch": 2.4898785425101213, "grad_norm": 0.5322304783391678, "learning_rate": 2.3627563537662823e-05, "loss": 0.4516, "step": 615 }, { "epoch": 2.493927125506073, "grad_norm": 0.4934389460708677, "learning_rate": 2.3571932828372433e-05, "loss": 0.4538, "step": 616 }, { "epoch": 2.4979757085020244, "grad_norm": 0.563474178733795, "learning_rate": 2.351627355800233e-05, "loss": 0.4462, "step": 617 }, { "epoch": 2.5020242914979756, "grad_norm": 0.49260630521390325, "learning_rate": 2.346058617160256e-05, "loss": 0.4469, "step": 618 }, { "epoch": 2.506072874493927, "grad_norm": 0.4696260354557719, "learning_rate": 2.3404871114447976e-05, "loss": 0.4451, "step": 619 }, { "epoch": 2.5101214574898787, "grad_norm": 0.469986846596035, "learning_rate": 2.334912883203469e-05, "loss": 0.4393, "step": 620 }, { "epoch": 2.51417004048583, "grad_norm": 0.44662276511349774, "learning_rate": 2.3293359770076505e-05, "loss": 0.4541, "step": 621 }, { "epoch": 2.5182186234817814, "grad_norm": 0.44083282191045414, "learning_rate": 2.323756437450136e-05, "loss": 0.4613, "step": 622 }, { "epoch": 2.522267206477733, "grad_norm": 0.4107592925132484, "learning_rate": 2.3181743091447754e-05, "loss": 0.457, "step": 623 }, { "epoch": 2.526315789473684, "grad_norm": 0.472738396263994, "learning_rate": 2.3125896367261164e-05, "loss": 0.4251, "step": 624 }, { "epoch": 2.5303643724696356, "grad_norm": 0.4650329737283501, "learning_rate": 2.307002464849052e-05, "loss": 0.4269, "step": 625 }, { "epoch": 2.534412955465587, "grad_norm": 0.43988750607160687, "learning_rate": 2.30141283818846e-05, "loss": 0.4588, "step": 626 }, { "epoch": 2.5384615384615383, "grad_norm": 0.5126169772969815, "learning_rate": 2.2958208014388462e-05, "loss": 0.4521, "step": 627 }, { "epoch": 2.54251012145749, "grad_norm": 0.4197507347632169, "learning_rate": 2.2902263993139857e-05, "loss": 0.4458, "step": 628 }, { "epoch": 2.5465587044534415, "grad_norm": 0.5174953239021007, "learning_rate": 2.2846296765465708e-05, "loss": 0.4575, "step": 629 }, { "epoch": 2.5506072874493926, "grad_norm": 0.4474802058190812, "learning_rate": 2.2790306778878465e-05, "loss": 0.442, "step": 630 }, { "epoch": 2.554655870445344, "grad_norm": 0.45981105752070256, "learning_rate": 2.273429448107257e-05, "loss": 0.4292, "step": 631 }, { "epoch": 2.5587044534412957, "grad_norm": 0.475549979863097, "learning_rate": 2.2678260319920868e-05, "loss": 0.4467, "step": 632 }, { "epoch": 2.562753036437247, "grad_norm": 0.4363021777514046, "learning_rate": 2.262220474347102e-05, "loss": 0.434, "step": 633 }, { "epoch": 2.5668016194331984, "grad_norm": 0.5047237070562561, "learning_rate": 2.2566128199941918e-05, "loss": 0.4439, "step": 634 }, { "epoch": 2.57085020242915, "grad_norm": 0.4739491176531391, "learning_rate": 2.2510031137720104e-05, "loss": 0.4599, "step": 635 }, { "epoch": 2.574898785425101, "grad_norm": 0.46401041560257483, "learning_rate": 2.2453914005356206e-05, "loss": 0.442, "step": 636 }, { "epoch": 2.5789473684210527, "grad_norm": 0.46466694341935033, "learning_rate": 2.2397777251561302e-05, "loss": 0.4551, "step": 637 }, { "epoch": 2.582995951417004, "grad_norm": 0.4179372712451944, "learning_rate": 2.2341621325203393e-05, "loss": 0.4567, "step": 638 }, { "epoch": 2.5870445344129553, "grad_norm": 0.41816651009880895, "learning_rate": 2.2285446675303764e-05, "loss": 0.4552, "step": 639 }, { "epoch": 2.591093117408907, "grad_norm": 0.4246914355542648, "learning_rate": 2.2229253751033423e-05, "loss": 0.4528, "step": 640 }, { "epoch": 2.5951417004048585, "grad_norm": 0.41775239896444477, "learning_rate": 2.2173043001709492e-05, "loss": 0.4561, "step": 641 }, { "epoch": 2.5991902834008096, "grad_norm": 0.40406421758629885, "learning_rate": 2.2116814876791627e-05, "loss": 0.4388, "step": 642 }, { "epoch": 2.603238866396761, "grad_norm": 0.4198717489558601, "learning_rate": 2.2060569825878413e-05, "loss": 0.4316, "step": 643 }, { "epoch": 2.6072874493927127, "grad_norm": 0.43594009096428304, "learning_rate": 2.2004308298703777e-05, "loss": 0.4458, "step": 644 }, { "epoch": 2.611336032388664, "grad_norm": 0.4415654218934666, "learning_rate": 2.1948030745133396e-05, "loss": 0.4534, "step": 645 }, { "epoch": 2.6153846153846154, "grad_norm": 0.42896498702255564, "learning_rate": 2.1891737615161083e-05, "loss": 0.4448, "step": 646 }, { "epoch": 2.619433198380567, "grad_norm": 0.3789639081402083, "learning_rate": 2.1835429358905218e-05, "loss": 0.4521, "step": 647 }, { "epoch": 2.623481781376518, "grad_norm": 0.4234226260125319, "learning_rate": 2.17791064266051e-05, "loss": 0.4528, "step": 648 }, { "epoch": 2.6275303643724697, "grad_norm": 0.37704094361323365, "learning_rate": 2.172276926861741e-05, "loss": 0.4425, "step": 649 }, { "epoch": 2.6315789473684212, "grad_norm": 0.47953528632071013, "learning_rate": 2.166641833541255e-05, "loss": 0.4515, "step": 650 }, { "epoch": 2.6356275303643724, "grad_norm": 0.3926432409349685, "learning_rate": 2.161005407757109e-05, "loss": 0.4425, "step": 651 }, { "epoch": 2.639676113360324, "grad_norm": 0.4240906338211075, "learning_rate": 2.155367694578013e-05, "loss": 0.463, "step": 652 }, { "epoch": 2.6437246963562755, "grad_norm": 0.42567386623679176, "learning_rate": 2.1497287390829712e-05, "loss": 0.4494, "step": 653 }, { "epoch": 2.6477732793522266, "grad_norm": 0.43625479820490826, "learning_rate": 2.1440885863609208e-05, "loss": 0.4513, "step": 654 }, { "epoch": 2.651821862348178, "grad_norm": 0.42105683158318047, "learning_rate": 2.1384472815103737e-05, "loss": 0.4646, "step": 655 }, { "epoch": 2.6558704453441297, "grad_norm": 0.46130636339387315, "learning_rate": 2.1328048696390534e-05, "loss": 0.4651, "step": 656 }, { "epoch": 2.659919028340081, "grad_norm": 0.4538848688497604, "learning_rate": 2.1271613958635342e-05, "loss": 0.4571, "step": 657 }, { "epoch": 2.6639676113360324, "grad_norm": 0.502833891222962, "learning_rate": 2.1215169053088825e-05, "loss": 0.4311, "step": 658 }, { "epoch": 2.668016194331984, "grad_norm": 0.479155148332144, "learning_rate": 2.1158714431082937e-05, "loss": 0.4355, "step": 659 }, { "epoch": 2.672064777327935, "grad_norm": 0.5238594759690257, "learning_rate": 2.110225054402733e-05, "loss": 0.4284, "step": 660 }, { "epoch": 2.6761133603238867, "grad_norm": 0.40974842997616134, "learning_rate": 2.1045777843405747e-05, "loss": 0.459, "step": 661 }, { "epoch": 2.6801619433198383, "grad_norm": 0.5496170525155696, "learning_rate": 2.09892967807724e-05, "loss": 0.4398, "step": 662 }, { "epoch": 2.6842105263157894, "grad_norm": 0.5655722632802467, "learning_rate": 2.093280780774835e-05, "loss": 0.4553, "step": 663 }, { "epoch": 2.688259109311741, "grad_norm": 0.49312054688204776, "learning_rate": 2.087631137601793e-05, "loss": 0.4362, "step": 664 }, { "epoch": 2.6923076923076925, "grad_norm": 0.47222560234140437, "learning_rate": 2.0819807937325104e-05, "loss": 0.4295, "step": 665 }, { "epoch": 2.6963562753036436, "grad_norm": 0.4780310932450505, "learning_rate": 2.0763297943469847e-05, "loss": 0.4363, "step": 666 }, { "epoch": 2.700404858299595, "grad_norm": 0.4132861510104456, "learning_rate": 2.070678184630458e-05, "loss": 0.4509, "step": 667 }, { "epoch": 2.7044534412955468, "grad_norm": 0.46769050271659707, "learning_rate": 2.06502600977305e-05, "loss": 0.459, "step": 668 }, { "epoch": 2.708502024291498, "grad_norm": 0.3619020211060213, "learning_rate": 2.0593733149694008e-05, "loss": 0.456, "step": 669 }, { "epoch": 2.7125506072874495, "grad_norm": 0.43724735746001797, "learning_rate": 2.0537201454183074e-05, "loss": 0.4463, "step": 670 }, { "epoch": 2.716599190283401, "grad_norm": 0.3737099051849437, "learning_rate": 2.0480665463223626e-05, "loss": 0.4576, "step": 671 }, { "epoch": 2.720647773279352, "grad_norm": 0.3690787815524404, "learning_rate": 2.042412562887594e-05, "loss": 0.4501, "step": 672 }, { "epoch": 2.7246963562753037, "grad_norm": 0.4330883573462796, "learning_rate": 2.0367582403231033e-05, "loss": 0.4367, "step": 673 }, { "epoch": 2.7287449392712553, "grad_norm": 0.38485739142619874, "learning_rate": 2.0311036238407023e-05, "loss": 0.4492, "step": 674 }, { "epoch": 2.7327935222672064, "grad_norm": 0.4118268272267897, "learning_rate": 2.0254487586545534e-05, "loss": 0.4456, "step": 675 }, { "epoch": 2.736842105263158, "grad_norm": 0.41137642623918963, "learning_rate": 2.0197936899808088e-05, "loss": 0.442, "step": 676 }, { "epoch": 2.7408906882591095, "grad_norm": 0.3567029471358589, "learning_rate": 2.0141384630372466e-05, "loss": 0.4538, "step": 677 }, { "epoch": 2.7449392712550607, "grad_norm": 0.4242311261460955, "learning_rate": 2.0084831230429106e-05, "loss": 0.4547, "step": 678 }, { "epoch": 2.748987854251012, "grad_norm": 0.3538359869434649, "learning_rate": 2.0028277152177492e-05, "loss": 0.4571, "step": 679 }, { "epoch": 2.753036437246964, "grad_norm": 0.49648706620010036, "learning_rate": 1.9971722847822518e-05, "loss": 0.4406, "step": 680 }, { "epoch": 2.757085020242915, "grad_norm": 0.3888014803774814, "learning_rate": 1.99151687695709e-05, "loss": 0.4407, "step": 681 }, { "epoch": 2.7611336032388665, "grad_norm": 0.5473776064805135, "learning_rate": 1.9858615369627537e-05, "loss": 0.4375, "step": 682 }, { "epoch": 2.765182186234818, "grad_norm": 0.44031251938020705, "learning_rate": 1.9802063100191916e-05, "loss": 0.4617, "step": 683 }, { "epoch": 2.769230769230769, "grad_norm": 0.45937994785321534, "learning_rate": 1.974551241345447e-05, "loss": 0.4462, "step": 684 }, { "epoch": 2.7732793522267207, "grad_norm": 0.4672204219958135, "learning_rate": 1.9688963761592984e-05, "loss": 0.439, "step": 685 }, { "epoch": 2.7773279352226723, "grad_norm": 0.4016397211911956, "learning_rate": 1.963241759676897e-05, "loss": 0.4503, "step": 686 }, { "epoch": 2.7813765182186234, "grad_norm": 0.4867800719238056, "learning_rate": 1.9575874371124066e-05, "loss": 0.4574, "step": 687 }, { "epoch": 2.785425101214575, "grad_norm": 0.4176190328116623, "learning_rate": 1.951933453677638e-05, "loss": 0.437, "step": 688 }, { "epoch": 2.7894736842105265, "grad_norm": 0.4090224139275023, "learning_rate": 1.9462798545816932e-05, "loss": 0.4551, "step": 689 }, { "epoch": 2.7935222672064777, "grad_norm": 0.43376577021980006, "learning_rate": 1.9406266850305996e-05, "loss": 0.4496, "step": 690 }, { "epoch": 2.7975708502024292, "grad_norm": 0.3884694979244205, "learning_rate": 1.9349739902269503e-05, "loss": 0.4514, "step": 691 }, { "epoch": 2.801619433198381, "grad_norm": 0.4089931983540309, "learning_rate": 1.9293218153695425e-05, "loss": 0.4371, "step": 692 }, { "epoch": 2.805668016194332, "grad_norm": 0.3904058990665688, "learning_rate": 1.9236702056530153e-05, "loss": 0.444, "step": 693 }, { "epoch": 2.8097165991902835, "grad_norm": 0.38154338490192063, "learning_rate": 1.9180192062674907e-05, "loss": 0.4494, "step": 694 }, { "epoch": 2.813765182186235, "grad_norm": 0.41532969301486067, "learning_rate": 1.9123688623982076e-05, "loss": 0.465, "step": 695 }, { "epoch": 2.817813765182186, "grad_norm": 0.4274897481154345, "learning_rate": 1.906719219225165e-05, "loss": 0.4601, "step": 696 }, { "epoch": 2.8218623481781377, "grad_norm": 0.39408538002911725, "learning_rate": 1.9010703219227603e-05, "loss": 0.4434, "step": 697 }, { "epoch": 2.8259109311740893, "grad_norm": 0.384942780945218, "learning_rate": 1.8954222156594256e-05, "loss": 0.442, "step": 698 }, { "epoch": 2.8299595141700404, "grad_norm": 0.41382117337013763, "learning_rate": 1.8897749455972673e-05, "loss": 0.4441, "step": 699 }, { "epoch": 2.834008097165992, "grad_norm": 0.39601830963262585, "learning_rate": 1.884128556891707e-05, "loss": 0.4463, "step": 700 }, { "epoch": 2.8380566801619436, "grad_norm": 0.4002563310500298, "learning_rate": 1.878483094691119e-05, "loss": 0.4404, "step": 701 }, { "epoch": 2.8421052631578947, "grad_norm": 0.4320187828404716, "learning_rate": 1.8728386041364664e-05, "loss": 0.443, "step": 702 }, { "epoch": 2.8461538461538463, "grad_norm": 0.40141997381401695, "learning_rate": 1.867195130360947e-05, "loss": 0.4533, "step": 703 }, { "epoch": 2.850202429149798, "grad_norm": 0.4004598023196895, "learning_rate": 1.8615527184896263e-05, "loss": 0.454, "step": 704 }, { "epoch": 2.854251012145749, "grad_norm": 0.45111873684295206, "learning_rate": 1.8559114136390795e-05, "loss": 0.4378, "step": 705 }, { "epoch": 2.8582995951417005, "grad_norm": 0.41879458232552197, "learning_rate": 1.8502712609170298e-05, "loss": 0.4371, "step": 706 }, { "epoch": 2.862348178137652, "grad_norm": 0.44148131566016446, "learning_rate": 1.8446323054219876e-05, "loss": 0.4591, "step": 707 }, { "epoch": 2.866396761133603, "grad_norm": 0.38737594135960085, "learning_rate": 1.838994592242891e-05, "loss": 0.4558, "step": 708 }, { "epoch": 2.8704453441295548, "grad_norm": 0.4006254296092388, "learning_rate": 1.8333581664587453e-05, "loss": 0.4331, "step": 709 }, { "epoch": 2.8744939271255063, "grad_norm": 0.39067015955520445, "learning_rate": 1.8277230731382593e-05, "loss": 0.4547, "step": 710 }, { "epoch": 2.8785425101214575, "grad_norm": 0.4387289908857173, "learning_rate": 1.82208935733949e-05, "loss": 0.4416, "step": 711 }, { "epoch": 2.882591093117409, "grad_norm": 0.416884393164527, "learning_rate": 1.8164570641094793e-05, "loss": 0.431, "step": 712 }, { "epoch": 2.8866396761133606, "grad_norm": 0.4433216644975534, "learning_rate": 1.810826238483892e-05, "loss": 0.4388, "step": 713 }, { "epoch": 2.8906882591093117, "grad_norm": 0.44965997944319547, "learning_rate": 1.8051969254866608e-05, "loss": 0.4463, "step": 714 }, { "epoch": 2.8947368421052633, "grad_norm": 0.39829525322460924, "learning_rate": 1.7995691701296226e-05, "loss": 0.427, "step": 715 }, { "epoch": 2.898785425101215, "grad_norm": 0.4889949773995396, "learning_rate": 1.7939430174121593e-05, "loss": 0.4435, "step": 716 }, { "epoch": 2.902834008097166, "grad_norm": 0.39931818589301943, "learning_rate": 1.788318512320838e-05, "loss": 0.452, "step": 717 }, { "epoch": 2.9068825910931175, "grad_norm": 0.3737653502670879, "learning_rate": 1.782695699829051e-05, "loss": 0.4538, "step": 718 }, { "epoch": 2.910931174089069, "grad_norm": 0.3894764769705788, "learning_rate": 1.7770746248966587e-05, "loss": 0.4314, "step": 719 }, { "epoch": 2.91497975708502, "grad_norm": 0.3886459053702682, "learning_rate": 1.7714553324696243e-05, "loss": 0.4532, "step": 720 }, { "epoch": 2.919028340080972, "grad_norm": 0.40147784045976564, "learning_rate": 1.7658378674796614e-05, "loss": 0.4394, "step": 721 }, { "epoch": 2.9230769230769234, "grad_norm": 0.3930098619103071, "learning_rate": 1.76022227484387e-05, "loss": 0.4504, "step": 722 }, { "epoch": 2.9271255060728745, "grad_norm": 0.3878224937078558, "learning_rate": 1.7546085994643807e-05, "loss": 0.4433, "step": 723 }, { "epoch": 2.931174089068826, "grad_norm": 0.40557101319447625, "learning_rate": 1.7489968862279902e-05, "loss": 0.4525, "step": 724 }, { "epoch": 2.9352226720647776, "grad_norm": 0.36588433004592774, "learning_rate": 1.743387180005809e-05, "loss": 0.4318, "step": 725 }, { "epoch": 2.9392712550607287, "grad_norm": 0.3879440278880997, "learning_rate": 1.737779525652899e-05, "loss": 0.4539, "step": 726 }, { "epoch": 2.9433198380566803, "grad_norm": 0.39162397264524784, "learning_rate": 1.732173968007914e-05, "loss": 0.4475, "step": 727 }, { "epoch": 2.9473684210526314, "grad_norm": 0.3763362227929375, "learning_rate": 1.7265705518927437e-05, "loss": 0.4442, "step": 728 }, { "epoch": 2.951417004048583, "grad_norm": 0.3784574964288552, "learning_rate": 1.7209693221121542e-05, "loss": 0.4416, "step": 729 }, { "epoch": 2.9554655870445345, "grad_norm": 0.38423394648392417, "learning_rate": 1.7153703234534302e-05, "loss": 0.448, "step": 730 }, { "epoch": 2.9595141700404857, "grad_norm": 0.3802420731350129, "learning_rate": 1.709773600686015e-05, "loss": 0.4545, "step": 731 }, { "epoch": 2.9635627530364372, "grad_norm": 0.39908336942995043, "learning_rate": 1.7041791985611548e-05, "loss": 0.4489, "step": 732 }, { "epoch": 2.967611336032389, "grad_norm": 0.39354364208018217, "learning_rate": 1.69858716181154e-05, "loss": 0.4472, "step": 733 }, { "epoch": 2.97165991902834, "grad_norm": 0.35108775516690294, "learning_rate": 1.692997535150948e-05, "loss": 0.4302, "step": 734 }, { "epoch": 2.9757085020242915, "grad_norm": 0.4277539599368108, "learning_rate": 1.687410363273884e-05, "loss": 0.434, "step": 735 }, { "epoch": 2.979757085020243, "grad_norm": 0.4335502795261412, "learning_rate": 1.6818256908552257e-05, "loss": 0.4543, "step": 736 }, { "epoch": 2.983805668016194, "grad_norm": 0.4134326553912304, "learning_rate": 1.6762435625498646e-05, "loss": 0.436, "step": 737 }, { "epoch": 2.9878542510121457, "grad_norm": 0.4437380481991515, "learning_rate": 1.67066402299235e-05, "loss": 0.4474, "step": 738 }, { "epoch": 2.9919028340080973, "grad_norm": 0.347834261862446, "learning_rate": 1.6650871167965313e-05, "loss": 0.4647, "step": 739 }, { "epoch": 2.9959514170040484, "grad_norm": 0.40828994659032664, "learning_rate": 1.6595128885552028e-05, "loss": 0.4439, "step": 740 }, { "epoch": 3.0, "grad_norm": 0.43806990027079695, "learning_rate": 1.653941382839745e-05, "loss": 0.4465, "step": 741 }, { "epoch": 3.0040485829959516, "grad_norm": 0.5698741091189348, "learning_rate": 1.6483726441997673e-05, "loss": 0.3438, "step": 742 }, { "epoch": 3.0080971659919027, "grad_norm": 0.45423254399848945, "learning_rate": 1.642806717162757e-05, "loss": 0.3176, "step": 743 }, { "epoch": 3.0121457489878543, "grad_norm": 0.8382401584159618, "learning_rate": 1.637243646233718e-05, "loss": 0.3237, "step": 744 }, { "epoch": 3.016194331983806, "grad_norm": 0.7903950993980752, "learning_rate": 1.6316834758948174e-05, "loss": 0.3281, "step": 745 }, { "epoch": 3.020242914979757, "grad_norm": 0.685262981018357, "learning_rate": 1.6261262506050282e-05, "loss": 0.3592, "step": 746 }, { "epoch": 3.0242914979757085, "grad_norm": 0.5088911765600208, "learning_rate": 1.620572014799777e-05, "loss": 0.3402, "step": 747 }, { "epoch": 3.02834008097166, "grad_norm": 0.46804165360168254, "learning_rate": 1.6150208128905857e-05, "loss": 0.3502, "step": 748 }, { "epoch": 3.032388663967611, "grad_norm": 0.5528363848650792, "learning_rate": 1.6094726892647147e-05, "loss": 0.3208, "step": 749 }, { "epoch": 3.0364372469635628, "grad_norm": 0.45562733830115976, "learning_rate": 1.6039276882848135e-05, "loss": 0.3102, "step": 750 }, { "epoch": 3.0404858299595143, "grad_norm": 0.440222114514405, "learning_rate": 1.5983858542885635e-05, "loss": 0.3275, "step": 751 }, { "epoch": 3.0445344129554655, "grad_norm": 0.4459159031977577, "learning_rate": 1.5928472315883203e-05, "loss": 0.3224, "step": 752 }, { "epoch": 3.048582995951417, "grad_norm": 0.4741075702061753, "learning_rate": 1.5873118644707633e-05, "loss": 0.3275, "step": 753 }, { "epoch": 3.0526315789473686, "grad_norm": 0.43434922364515366, "learning_rate": 1.5817797971965413e-05, "loss": 0.333, "step": 754 }, { "epoch": 3.0566801619433197, "grad_norm": 0.4455003451872164, "learning_rate": 1.576251073999917e-05, "loss": 0.334, "step": 755 }, { "epoch": 3.0607287449392713, "grad_norm": 0.4965776185847766, "learning_rate": 1.5707257390884126e-05, "loss": 0.3311, "step": 756 }, { "epoch": 3.064777327935223, "grad_norm": 0.39175859412009933, "learning_rate": 1.5652038366424595e-05, "loss": 0.327, "step": 757 }, { "epoch": 3.068825910931174, "grad_norm": 0.4868940850674905, "learning_rate": 1.5596854108150424e-05, "loss": 0.3151, "step": 758 }, { "epoch": 3.0728744939271255, "grad_norm": 0.42844817026934834, "learning_rate": 1.5541705057313476e-05, "loss": 0.3233, "step": 759 }, { "epoch": 3.076923076923077, "grad_norm": 0.4165618461054378, "learning_rate": 1.5486591654884086e-05, "loss": 0.3247, "step": 760 }, { "epoch": 3.080971659919028, "grad_norm": 0.4439192883982158, "learning_rate": 1.543151434154755e-05, "loss": 0.3401, "step": 761 }, { "epoch": 3.08502024291498, "grad_norm": 0.4046695107300932, "learning_rate": 1.5376473557700615e-05, "loss": 0.3395, "step": 762 }, { "epoch": 3.0890688259109313, "grad_norm": 0.45032062694599945, "learning_rate": 1.5321469743447893e-05, "loss": 0.3279, "step": 763 }, { "epoch": 3.0931174089068825, "grad_norm": 0.47839367517337117, "learning_rate": 1.5266503338598434e-05, "loss": 0.3383, "step": 764 }, { "epoch": 3.097165991902834, "grad_norm": 0.3897026524962515, "learning_rate": 1.5211574782662149e-05, "loss": 0.3252, "step": 765 }, { "epoch": 3.1012145748987856, "grad_norm": 0.4337950817336917, "learning_rate": 1.5156684514846307e-05, "loss": 0.3418, "step": 766 }, { "epoch": 3.1052631578947367, "grad_norm": 0.3738655599798417, "learning_rate": 1.5101832974052012e-05, "loss": 0.3305, "step": 767 }, { "epoch": 3.1093117408906883, "grad_norm": 0.4155412616784737, "learning_rate": 1.5047020598870737e-05, "loss": 0.3377, "step": 768 }, { "epoch": 3.11336032388664, "grad_norm": 0.40570876866972605, "learning_rate": 1.4992247827580778e-05, "loss": 0.3387, "step": 769 }, { "epoch": 3.117408906882591, "grad_norm": 0.38777876691146235, "learning_rate": 1.493751509814374e-05, "loss": 0.332, "step": 770 }, { "epoch": 3.1214574898785425, "grad_norm": 0.4099190653795303, "learning_rate": 1.4882822848201073e-05, "loss": 0.353, "step": 771 }, { "epoch": 3.125506072874494, "grad_norm": 0.37036010122422924, "learning_rate": 1.4828171515070553e-05, "loss": 0.3328, "step": 772 }, { "epoch": 3.1295546558704452, "grad_norm": 0.4038461226246466, "learning_rate": 1.4773561535742793e-05, "loss": 0.3304, "step": 773 }, { "epoch": 3.133603238866397, "grad_norm": 0.37331033939962033, "learning_rate": 1.4718993346877715e-05, "loss": 0.3379, "step": 774 }, { "epoch": 3.1376518218623484, "grad_norm": 0.3908060363890622, "learning_rate": 1.466446738480111e-05, "loss": 0.3371, "step": 775 }, { "epoch": 3.1417004048582995, "grad_norm": 0.3698607111628352, "learning_rate": 1.460998408550112e-05, "loss": 0.3211, "step": 776 }, { "epoch": 3.145748987854251, "grad_norm": 0.38598231325437105, "learning_rate": 1.4555543884624751e-05, "loss": 0.3355, "step": 777 }, { "epoch": 3.1497975708502026, "grad_norm": 0.43717773020463174, "learning_rate": 1.4501147217474402e-05, "loss": 0.3351, "step": 778 }, { "epoch": 3.1538461538461537, "grad_norm": 0.39755099309365954, "learning_rate": 1.444679451900437e-05, "loss": 0.3287, "step": 779 }, { "epoch": 3.1578947368421053, "grad_norm": 0.40049271293381555, "learning_rate": 1.4392486223817397e-05, "loss": 0.33, "step": 780 }, { "epoch": 3.161943319838057, "grad_norm": 0.4214792058594269, "learning_rate": 1.4338222766161143e-05, "loss": 0.3268, "step": 781 }, { "epoch": 3.165991902834008, "grad_norm": 0.39542360603146515, "learning_rate": 1.4284004579924777e-05, "loss": 0.3422, "step": 782 }, { "epoch": 3.1700404858299596, "grad_norm": 0.43623844831044883, "learning_rate": 1.4229832098635479e-05, "loss": 0.3215, "step": 783 }, { "epoch": 3.174089068825911, "grad_norm": 0.43699617849566724, "learning_rate": 1.4175705755454963e-05, "loss": 0.3346, "step": 784 }, { "epoch": 3.1781376518218623, "grad_norm": 0.4239201514123324, "learning_rate": 1.412162598317602e-05, "loss": 0.3269, "step": 785 }, { "epoch": 3.182186234817814, "grad_norm": 0.4485826098557037, "learning_rate": 1.406759321421907e-05, "loss": 0.3436, "step": 786 }, { "epoch": 3.1862348178137654, "grad_norm": 0.43981378730955056, "learning_rate": 1.401360788062871e-05, "loss": 0.3182, "step": 787 }, { "epoch": 3.1902834008097165, "grad_norm": 0.3847531543228635, "learning_rate": 1.3959670414070204e-05, "loss": 0.3339, "step": 788 }, { "epoch": 3.194331983805668, "grad_norm": 0.4250350755464123, "learning_rate": 1.3905781245826108e-05, "loss": 0.3332, "step": 789 }, { "epoch": 3.1983805668016196, "grad_norm": 0.3942247594936936, "learning_rate": 1.3851940806792778e-05, "loss": 0.3258, "step": 790 }, { "epoch": 3.2024291497975708, "grad_norm": 0.3994253089730525, "learning_rate": 1.379814952747693e-05, "loss": 0.3443, "step": 791 }, { "epoch": 3.2064777327935223, "grad_norm": 0.4143136182569084, "learning_rate": 1.3744407837992193e-05, "loss": 0.328, "step": 792 }, { "epoch": 3.2105263157894735, "grad_norm": 0.39297634078600147, "learning_rate": 1.3690716168055692e-05, "loss": 0.3497, "step": 793 }, { "epoch": 3.214574898785425, "grad_norm": 0.39743634656243804, "learning_rate": 1.363707494698459e-05, "loss": 0.3337, "step": 794 }, { "epoch": 3.2186234817813766, "grad_norm": 0.3884031933913506, "learning_rate": 1.358348460369265e-05, "loss": 0.3277, "step": 795 }, { "epoch": 3.2226720647773277, "grad_norm": 0.3793515564171868, "learning_rate": 1.352994556668684e-05, "loss": 0.3268, "step": 796 }, { "epoch": 3.2267206477732793, "grad_norm": 0.39876633181339716, "learning_rate": 1.3476458264063865e-05, "loss": 0.3286, "step": 797 }, { "epoch": 3.230769230769231, "grad_norm": 0.3998353198055001, "learning_rate": 1.342302312350678e-05, "loss": 0.317, "step": 798 }, { "epoch": 3.234817813765182, "grad_norm": 0.4138611145838701, "learning_rate": 1.3369640572281537e-05, "loss": 0.3295, "step": 799 }, { "epoch": 3.2388663967611335, "grad_norm": 0.39535919889560234, "learning_rate": 1.3316311037233596e-05, "loss": 0.3208, "step": 800 }, { "epoch": 3.242914979757085, "grad_norm": 0.38929977419000966, "learning_rate": 1.326303494478451e-05, "loss": 0.3404, "step": 801 }, { "epoch": 3.246963562753036, "grad_norm": 0.43664028528596976, "learning_rate": 1.3209812720928472e-05, "loss": 0.3244, "step": 802 }, { "epoch": 3.251012145748988, "grad_norm": 0.38217164823270805, "learning_rate": 1.315664479122898e-05, "loss": 0.316, "step": 803 }, { "epoch": 3.2550607287449393, "grad_norm": 0.39938774709841945, "learning_rate": 1.3103531580815378e-05, "loss": 0.3516, "step": 804 }, { "epoch": 3.2591093117408905, "grad_norm": 0.4259082389864567, "learning_rate": 1.3050473514379488e-05, "loss": 0.3362, "step": 805 }, { "epoch": 3.263157894736842, "grad_norm": 0.3906897709455996, "learning_rate": 1.2997471016172188e-05, "loss": 0.3205, "step": 806 }, { "epoch": 3.2672064777327936, "grad_norm": 0.4002545389374646, "learning_rate": 1.2944524510000042e-05, "loss": 0.3501, "step": 807 }, { "epoch": 3.2712550607287447, "grad_norm": 0.3779085966552503, "learning_rate": 1.2891634419221908e-05, "loss": 0.3265, "step": 808 }, { "epoch": 3.2753036437246963, "grad_norm": 0.40907825985997454, "learning_rate": 1.2838801166745545e-05, "loss": 0.3015, "step": 809 }, { "epoch": 3.279352226720648, "grad_norm": 0.38749467329659043, "learning_rate": 1.2786025175024216e-05, "loss": 0.3409, "step": 810 }, { "epoch": 3.283400809716599, "grad_norm": 0.4210739084525279, "learning_rate": 1.2733306866053357e-05, "loss": 0.3269, "step": 811 }, { "epoch": 3.2874493927125505, "grad_norm": 0.40402960347344175, "learning_rate": 1.2680646661367163e-05, "loss": 0.3335, "step": 812 }, { "epoch": 3.291497975708502, "grad_norm": 0.39128581228955844, "learning_rate": 1.2628044982035219e-05, "loss": 0.3337, "step": 813 }, { "epoch": 3.2955465587044532, "grad_norm": 0.41986330442080344, "learning_rate": 1.2575502248659155e-05, "loss": 0.322, "step": 814 }, { "epoch": 3.299595141700405, "grad_norm": 0.3755490492893918, "learning_rate": 1.2523018881369269e-05, "loss": 0.3393, "step": 815 }, { "epoch": 3.3036437246963564, "grad_norm": 0.3749536971277909, "learning_rate": 1.2470595299821176e-05, "loss": 0.3351, "step": 816 }, { "epoch": 3.3076923076923075, "grad_norm": 0.3829678878921148, "learning_rate": 1.2418231923192416e-05, "loss": 0.329, "step": 817 }, { "epoch": 3.311740890688259, "grad_norm": 0.41676068816563644, "learning_rate": 1.2365929170179165e-05, "loss": 0.327, "step": 818 }, { "epoch": 3.3157894736842106, "grad_norm": 0.36794357365760555, "learning_rate": 1.2313687458992848e-05, "loss": 0.3211, "step": 819 }, { "epoch": 3.3198380566801617, "grad_norm": 0.40323242915679497, "learning_rate": 1.2261507207356785e-05, "loss": 0.3401, "step": 820 }, { "epoch": 3.3238866396761133, "grad_norm": 0.4289388210284499, "learning_rate": 1.220938883250288e-05, "loss": 0.3246, "step": 821 }, { "epoch": 3.327935222672065, "grad_norm": 0.4145720155062669, "learning_rate": 1.215733275116827e-05, "loss": 0.3511, "step": 822 }, { "epoch": 3.331983805668016, "grad_norm": 0.390544135120874, "learning_rate": 1.2105339379592005e-05, "loss": 0.3281, "step": 823 }, { "epoch": 3.3360323886639676, "grad_norm": 0.45649857145643435, "learning_rate": 1.2053409133511681e-05, "loss": 0.3232, "step": 824 }, { "epoch": 3.340080971659919, "grad_norm": 0.39881626901443407, "learning_rate": 1.2001542428160173e-05, "loss": 0.337, "step": 825 }, { "epoch": 3.3441295546558703, "grad_norm": 0.4468652736330309, "learning_rate": 1.1949739678262278e-05, "loss": 0.327, "step": 826 }, { "epoch": 3.348178137651822, "grad_norm": 0.4618437848917049, "learning_rate": 1.189800129803139e-05, "loss": 0.33, "step": 827 }, { "epoch": 3.3522267206477734, "grad_norm": 0.3991376318013047, "learning_rate": 1.1846327701166228e-05, "loss": 0.3203, "step": 828 }, { "epoch": 3.3562753036437245, "grad_norm": 0.36785197819366516, "learning_rate": 1.1794719300847493e-05, "loss": 0.3486, "step": 829 }, { "epoch": 3.360323886639676, "grad_norm": 0.4012573196150077, "learning_rate": 1.1743176509734582e-05, "loss": 0.3317, "step": 830 }, { "epoch": 3.3643724696356276, "grad_norm": 0.4026818322205964, "learning_rate": 1.1691699739962275e-05, "loss": 0.3395, "step": 831 }, { "epoch": 3.3684210526315788, "grad_norm": 0.36579426242524704, "learning_rate": 1.1640289403137456e-05, "loss": 0.3335, "step": 832 }, { "epoch": 3.3724696356275303, "grad_norm": 0.35493073698396943, "learning_rate": 1.1588945910335802e-05, "loss": 0.3284, "step": 833 }, { "epoch": 3.376518218623482, "grad_norm": 0.3698399046419741, "learning_rate": 1.1537669672098517e-05, "loss": 0.3244, "step": 834 }, { "epoch": 3.380566801619433, "grad_norm": 0.36208608749233373, "learning_rate": 1.1486461098429036e-05, "loss": 0.3357, "step": 835 }, { "epoch": 3.3846153846153846, "grad_norm": 0.4026406498136208, "learning_rate": 1.1435320598789747e-05, "loss": 0.3081, "step": 836 }, { "epoch": 3.388663967611336, "grad_norm": 0.3826464206438068, "learning_rate": 1.1384248582098736e-05, "loss": 0.3467, "step": 837 }, { "epoch": 3.3927125506072873, "grad_norm": 0.38208270677044975, "learning_rate": 1.1333245456726467e-05, "loss": 0.3244, "step": 838 }, { "epoch": 3.396761133603239, "grad_norm": 0.366704047260151, "learning_rate": 1.128231163049258e-05, "loss": 0.3306, "step": 839 }, { "epoch": 3.4008097165991904, "grad_norm": 0.3915445125240143, "learning_rate": 1.123144751066259e-05, "loss": 0.3348, "step": 840 }, { "epoch": 3.4048582995951415, "grad_norm": 0.37224091081633864, "learning_rate": 1.1180653503944657e-05, "loss": 0.3235, "step": 841 }, { "epoch": 3.408906882591093, "grad_norm": 0.382186181464385, "learning_rate": 1.1129930016486287e-05, "loss": 0.3296, "step": 842 }, { "epoch": 3.4129554655870447, "grad_norm": 0.38551105908180006, "learning_rate": 1.107927745387114e-05, "loss": 0.3199, "step": 843 }, { "epoch": 3.417004048582996, "grad_norm": 0.3854724813424922, "learning_rate": 1.1028696221115766e-05, "loss": 0.3557, "step": 844 }, { "epoch": 3.4210526315789473, "grad_norm": 0.38617584188951803, "learning_rate": 1.0978186722666341e-05, "loss": 0.3498, "step": 845 }, { "epoch": 3.425101214574899, "grad_norm": 0.38897511483489783, "learning_rate": 1.0927749362395475e-05, "loss": 0.3363, "step": 846 }, { "epoch": 3.42914979757085, "grad_norm": 0.36504537438772894, "learning_rate": 1.0877384543598954e-05, "loss": 0.3193, "step": 847 }, { "epoch": 3.4331983805668016, "grad_norm": 0.38403010961492734, "learning_rate": 1.0827092668992529e-05, "loss": 0.3295, "step": 848 }, { "epoch": 3.437246963562753, "grad_norm": 0.3774975886118004, "learning_rate": 1.077687414070869e-05, "loss": 0.3427, "step": 849 }, { "epoch": 3.4412955465587043, "grad_norm": 0.3587369124543757, "learning_rate": 1.0726729360293451e-05, "loss": 0.3283, "step": 850 }, { "epoch": 3.445344129554656, "grad_norm": 0.45122448582548713, "learning_rate": 1.0676658728703147e-05, "loss": 0.3554, "step": 851 }, { "epoch": 3.4493927125506074, "grad_norm": 0.3676279646080525, "learning_rate": 1.0626662646301198e-05, "loss": 0.3418, "step": 852 }, { "epoch": 3.4534412955465585, "grad_norm": 0.3589952105616687, "learning_rate": 1.0576741512854958e-05, "loss": 0.3389, "step": 853 }, { "epoch": 3.45748987854251, "grad_norm": 0.3816434056475414, "learning_rate": 1.052689572753247e-05, "loss": 0.3395, "step": 854 }, { "epoch": 3.4615384615384617, "grad_norm": 0.36878992054873505, "learning_rate": 1.0477125688899328e-05, "loss": 0.3159, "step": 855 }, { "epoch": 3.465587044534413, "grad_norm": 0.36406630666742473, "learning_rate": 1.0427431794915406e-05, "loss": 0.3368, "step": 856 }, { "epoch": 3.4696356275303644, "grad_norm": 0.3665315838171894, "learning_rate": 1.0377814442931769e-05, "loss": 0.3293, "step": 857 }, { "epoch": 3.473684210526316, "grad_norm": 0.39593545956978626, "learning_rate": 1.0328274029687439e-05, "loss": 0.333, "step": 858 }, { "epoch": 3.477732793522267, "grad_norm": 0.3585323784908102, "learning_rate": 1.0278810951306253e-05, "loss": 0.3231, "step": 859 }, { "epoch": 3.4817813765182186, "grad_norm": 0.3556124969355949, "learning_rate": 1.0229425603293647e-05, "loss": 0.327, "step": 860 }, { "epoch": 3.48582995951417, "grad_norm": 0.3835886493343863, "learning_rate": 1.0180118380533558e-05, "loss": 0.3049, "step": 861 }, { "epoch": 3.4898785425101213, "grad_norm": 0.3779410994808152, "learning_rate": 1.0130889677285225e-05, "loss": 0.3485, "step": 862 }, { "epoch": 3.493927125506073, "grad_norm": 0.370583389087676, "learning_rate": 1.008173988718005e-05, "loss": 0.3327, "step": 863 }, { "epoch": 3.4979757085020244, "grad_norm": 0.37571952179578816, "learning_rate": 1.0032669403218439e-05, "loss": 0.3468, "step": 864 }, { "epoch": 3.5020242914979756, "grad_norm": 0.37384485081444035, "learning_rate": 9.983678617766677e-06, "loss": 0.3243, "step": 865 }, { "epoch": 3.506072874493927, "grad_norm": 0.368018111853861, "learning_rate": 9.934767922553788e-06, "loss": 0.3395, "step": 866 }, { "epoch": 3.5101214574898787, "grad_norm": 0.3580300906404369, "learning_rate": 9.885937708668371e-06, "loss": 0.3374, "step": 867 }, { "epoch": 3.51417004048583, "grad_norm": 0.3545937824688672, "learning_rate": 9.837188366555524e-06, "loss": 0.3345, "step": 868 }, { "epoch": 3.5182186234817814, "grad_norm": 0.3818040530918038, "learning_rate": 9.788520286013691e-06, "loss": 0.3465, "step": 869 }, { "epoch": 3.522267206477733, "grad_norm": 0.3928780911896249, "learning_rate": 9.739933856191537e-06, "loss": 0.3467, "step": 870 }, { "epoch": 3.526315789473684, "grad_norm": 0.37019496375672634, "learning_rate": 9.691429465584866e-06, "loss": 0.3043, "step": 871 }, { "epoch": 3.5303643724696356, "grad_norm": 0.37278032613038714, "learning_rate": 9.643007502033492e-06, "loss": 0.335, "step": 872 }, { "epoch": 3.534412955465587, "grad_norm": 0.3838652468275984, "learning_rate": 9.594668352718152e-06, "loss": 0.3374, "step": 873 }, { "epoch": 3.5384615384615383, "grad_norm": 0.3597804880581786, "learning_rate": 9.546412404157383e-06, "loss": 0.3396, "step": 874 }, { "epoch": 3.54251012145749, "grad_norm": 0.39768972368993466, "learning_rate": 9.498240042204471e-06, "loss": 0.3294, "step": 875 }, { "epoch": 3.5465587044534415, "grad_norm": 0.3970221299237698, "learning_rate": 9.450151652044339e-06, "loss": 0.3393, "step": 876 }, { "epoch": 3.5506072874493926, "grad_norm": 0.36714628540972694, "learning_rate": 9.402147618190472e-06, "loss": 0.3221, "step": 877 }, { "epoch": 3.554655870445344, "grad_norm": 0.4034071034399899, "learning_rate": 9.354228324481841e-06, "loss": 0.3119, "step": 878 }, { "epoch": 3.5587044534412957, "grad_norm": 0.3736071927172039, "learning_rate": 9.306394154079852e-06, "loss": 0.3313, "step": 879 }, { "epoch": 3.562753036437247, "grad_norm": 0.36658557543627557, "learning_rate": 9.258645489465254e-06, "loss": 0.3259, "step": 880 }, { "epoch": 3.5668016194331984, "grad_norm": 0.37337754516340826, "learning_rate": 9.21098271243509e-06, "loss": 0.333, "step": 881 }, { "epoch": 3.57085020242915, "grad_norm": 0.37813122151349904, "learning_rate": 9.163406204099659e-06, "loss": 0.3244, "step": 882 }, { "epoch": 3.574898785425101, "grad_norm": 0.38982033377007486, "learning_rate": 9.115916344879452e-06, "loss": 0.3287, "step": 883 }, { "epoch": 3.5789473684210527, "grad_norm": 0.3582741506655269, "learning_rate": 9.068513514502126e-06, "loss": 0.3369, "step": 884 }, { "epoch": 3.582995951417004, "grad_norm": 0.39738066872562033, "learning_rate": 9.021198091999439e-06, "loss": 0.3433, "step": 885 }, { "epoch": 3.5870445344129553, "grad_norm": 0.3444122469774166, "learning_rate": 8.973970455704251e-06, "loss": 0.3302, "step": 886 }, { "epoch": 3.591093117408907, "grad_norm": 0.3568581640680521, "learning_rate": 8.926830983247495e-06, "loss": 0.3249, "step": 887 }, { "epoch": 3.5951417004048585, "grad_norm": 0.38335989196331466, "learning_rate": 8.879780051555116e-06, "loss": 0.3308, "step": 888 }, { "epoch": 3.5991902834008096, "grad_norm": 0.3821181611463222, "learning_rate": 8.83281803684512e-06, "loss": 0.3202, "step": 889 }, { "epoch": 3.603238866396761, "grad_norm": 0.3528869252842929, "learning_rate": 8.785945314624532e-06, "loss": 0.33, "step": 890 }, { "epoch": 3.6072874493927127, "grad_norm": 0.37933370247517045, "learning_rate": 8.739162259686385e-06, "loss": 0.324, "step": 891 }, { "epoch": 3.611336032388664, "grad_norm": 0.3649339208839775, "learning_rate": 8.692469246106734e-06, "loss": 0.3203, "step": 892 }, { "epoch": 3.6153846153846154, "grad_norm": 0.37724963222704677, "learning_rate": 8.645866647241677e-06, "loss": 0.3421, "step": 893 }, { "epoch": 3.619433198380567, "grad_norm": 0.3622846624423585, "learning_rate": 8.599354835724348e-06, "loss": 0.3296, "step": 894 }, { "epoch": 3.623481781376518, "grad_norm": 0.3519389750762941, "learning_rate": 8.552934183461958e-06, "loss": 0.3469, "step": 895 }, { "epoch": 3.6275303643724697, "grad_norm": 0.3662360835589883, "learning_rate": 8.506605061632805e-06, "loss": 0.3413, "step": 896 }, { "epoch": 3.6315789473684212, "grad_norm": 0.35250785055830414, "learning_rate": 8.460367840683308e-06, "loss": 0.3266, "step": 897 }, { "epoch": 3.6356275303643724, "grad_norm": 0.364045465515907, "learning_rate": 8.414222890325065e-06, "loss": 0.3228, "step": 898 }, { "epoch": 3.639676113360324, "grad_norm": 0.3578821925680158, "learning_rate": 8.368170579531855e-06, "loss": 0.3218, "step": 899 }, { "epoch": 3.6437246963562755, "grad_norm": 0.37198206749974894, "learning_rate": 8.322211276536733e-06, "loss": 0.3321, "step": 900 }, { "epoch": 3.6477732793522266, "grad_norm": 0.38342686679147814, "learning_rate": 8.27634534882907e-06, "loss": 0.324, "step": 901 }, { "epoch": 3.651821862348178, "grad_norm": 0.37734784068997634, "learning_rate": 8.230573163151585e-06, "loss": 0.3313, "step": 902 }, { "epoch": 3.6558704453441297, "grad_norm": 0.39303013906808487, "learning_rate": 8.184895085497466e-06, "loss": 0.3201, "step": 903 }, { "epoch": 3.659919028340081, "grad_norm": 0.37265718186023156, "learning_rate": 8.139311481107401e-06, "loss": 0.3335, "step": 904 }, { "epoch": 3.6639676113360324, "grad_norm": 0.3621444858315891, "learning_rate": 8.093822714466686e-06, "loss": 0.331, "step": 905 }, { "epoch": 3.668016194331984, "grad_norm": 0.3689822119466648, "learning_rate": 8.048429149302271e-06, "loss": 0.3484, "step": 906 }, { "epoch": 3.672064777327935, "grad_norm": 0.3721272283438131, "learning_rate": 8.003131148579902e-06, "loss": 0.342, "step": 907 }, { "epoch": 3.6761133603238867, "grad_norm": 0.3516227519492819, "learning_rate": 7.957929074501182e-06, "loss": 0.3432, "step": 908 }, { "epoch": 3.6801619433198383, "grad_norm": 0.3678888141472692, "learning_rate": 7.912823288500695e-06, "loss": 0.3093, "step": 909 }, { "epoch": 3.6842105263157894, "grad_norm": 0.36102637725004894, "learning_rate": 7.867814151243102e-06, "loss": 0.3215, "step": 910 }, { "epoch": 3.688259109311741, "grad_norm": 0.3806878257256919, "learning_rate": 7.822902022620267e-06, "loss": 0.3276, "step": 911 }, { "epoch": 3.6923076923076925, "grad_norm": 0.36025574314873193, "learning_rate": 7.778087261748377e-06, "loss": 0.324, "step": 912 }, { "epoch": 3.6963562753036436, "grad_norm": 0.3657031305682369, "learning_rate": 7.733370226965056e-06, "loss": 0.3347, "step": 913 }, { "epoch": 3.700404858299595, "grad_norm": 0.3523173596977649, "learning_rate": 7.688751275826527e-06, "loss": 0.3238, "step": 914 }, { "epoch": 3.7044534412955468, "grad_norm": 0.3579810989895611, "learning_rate": 7.644230765104735e-06, "loss": 0.3265, "step": 915 }, { "epoch": 3.708502024291498, "grad_norm": 0.3690363238116635, "learning_rate": 7.599809050784502e-06, "loss": 0.3263, "step": 916 }, { "epoch": 3.7125506072874495, "grad_norm": 0.369586545631982, "learning_rate": 7.555486488060661e-06, "loss": 0.3255, "step": 917 }, { "epoch": 3.716599190283401, "grad_norm": 0.34848939516791555, "learning_rate": 7.511263431335249e-06, "loss": 0.3276, "step": 918 }, { "epoch": 3.720647773279352, "grad_norm": 0.3993450411706624, "learning_rate": 7.467140234214656e-06, "loss": 0.3233, "step": 919 }, { "epoch": 3.7246963562753037, "grad_norm": 0.384345306443403, "learning_rate": 7.423117249506777e-06, "loss": 0.3269, "step": 920 }, { "epoch": 3.7287449392712553, "grad_norm": 0.3452847467151486, "learning_rate": 7.379194829218231e-06, "loss": 0.322, "step": 921 }, { "epoch": 3.7327935222672064, "grad_norm": 0.36886556462720266, "learning_rate": 7.335373324551518e-06, "loss": 0.3282, "step": 922 }, { "epoch": 3.736842105263158, "grad_norm": 0.36271020247302865, "learning_rate": 7.291653085902224e-06, "loss": 0.3183, "step": 923 }, { "epoch": 3.7408906882591095, "grad_norm": 0.35528784654784096, "learning_rate": 7.248034462856213e-06, "loss": 0.3387, "step": 924 }, { "epoch": 3.7449392712550607, "grad_norm": 0.3592303738339367, "learning_rate": 7.20451780418683e-06, "loss": 0.3242, "step": 925 }, { "epoch": 3.748987854251012, "grad_norm": 0.35684823592402537, "learning_rate": 7.161103457852128e-06, "loss": 0.3316, "step": 926 }, { "epoch": 3.753036437246964, "grad_norm": 0.370330169322077, "learning_rate": 7.117791770992046e-06, "loss": 0.3172, "step": 927 }, { "epoch": 3.757085020242915, "grad_norm": 0.357640409422367, "learning_rate": 7.074583089925688e-06, "loss": 0.3152, "step": 928 }, { "epoch": 3.7611336032388665, "grad_norm": 0.35095953870731783, "learning_rate": 7.0314777601485155e-06, "loss": 0.3368, "step": 929 }, { "epoch": 3.765182186234818, "grad_norm": 0.3565434640861754, "learning_rate": 6.988476126329604e-06, "loss": 0.3278, "step": 930 }, { "epoch": 3.769230769230769, "grad_norm": 0.3585951109444752, "learning_rate": 6.945578532308857e-06, "loss": 0.3075, "step": 931 }, { "epoch": 3.7732793522267207, "grad_norm": 0.34726065517182514, "learning_rate": 6.902785321094301e-06, "loss": 0.3289, "step": 932 }, { "epoch": 3.7773279352226723, "grad_norm": 0.3632160759143023, "learning_rate": 6.8600968348593115e-06, "loss": 0.3245, "step": 933 }, { "epoch": 3.7813765182186234, "grad_norm": 0.36073741634999335, "learning_rate": 6.817513414939892e-06, "loss": 0.3262, "step": 934 }, { "epoch": 3.785425101214575, "grad_norm": 0.34816017403379523, "learning_rate": 6.775035401831924e-06, "loss": 0.3424, "step": 935 }, { "epoch": 3.7894736842105265, "grad_norm": 0.3422252162429211, "learning_rate": 6.732663135188473e-06, "loss": 0.3233, "step": 936 }, { "epoch": 3.7935222672064777, "grad_norm": 0.35842494369044103, "learning_rate": 6.690396953817055e-06, "loss": 0.354, "step": 937 }, { "epoch": 3.7975708502024292, "grad_norm": 0.35989661326663697, "learning_rate": 6.648237195676927e-06, "loss": 0.3169, "step": 938 }, { "epoch": 3.801619433198381, "grad_norm": 0.374510367257931, "learning_rate": 6.606184197876395e-06, "loss": 0.3289, "step": 939 }, { "epoch": 3.805668016194332, "grad_norm": 0.3674082886219193, "learning_rate": 6.564238296670109e-06, "loss": 0.3099, "step": 940 }, { "epoch": 3.8097165991902835, "grad_norm": 0.36213626752183503, "learning_rate": 6.52239982745638e-06, "loss": 0.3378, "step": 941 }, { "epoch": 3.813765182186235, "grad_norm": 0.3480121138429209, "learning_rate": 6.480669124774484e-06, "loss": 0.3307, "step": 942 }, { "epoch": 3.817813765182186, "grad_norm": 0.36541409166869554, "learning_rate": 6.439046522302013e-06, "loss": 0.3279, "step": 943 }, { "epoch": 3.8218623481781377, "grad_norm": 0.34707350367613377, "learning_rate": 6.397532352852191e-06, "loss": 0.3419, "step": 944 }, { "epoch": 3.8259109311740893, "grad_norm": 0.3509820200580403, "learning_rate": 6.356126948371202e-06, "loss": 0.3241, "step": 945 }, { "epoch": 3.8299595141700404, "grad_norm": 0.3565826846277528, "learning_rate": 6.314830639935559e-06, "loss": 0.3439, "step": 946 }, { "epoch": 3.834008097165992, "grad_norm": 0.34729689206930947, "learning_rate": 6.273643757749448e-06, "loss": 0.3197, "step": 947 }, { "epoch": 3.8380566801619436, "grad_norm": 0.36672282725598293, "learning_rate": 6.2325666311420856e-06, "loss": 0.3162, "step": 948 }, { "epoch": 3.8421052631578947, "grad_norm": 0.36097057365447427, "learning_rate": 6.1915995885650675e-06, "loss": 0.3376, "step": 949 }, { "epoch": 3.8461538461538463, "grad_norm": 0.36139514069092477, "learning_rate": 6.150742957589784e-06, "loss": 0.3193, "step": 950 }, { "epoch": 3.850202429149798, "grad_norm": 0.33796193205404557, "learning_rate": 6.109997064904774e-06, "loss": 0.3227, "step": 951 }, { "epoch": 3.854251012145749, "grad_norm": 0.3426863585728826, "learning_rate": 6.069362236313092e-06, "loss": 0.3356, "step": 952 }, { "epoch": 3.8582995951417005, "grad_norm": 0.37310649126758905, "learning_rate": 6.0288387967297565e-06, "loss": 0.3336, "step": 953 }, { "epoch": 3.862348178137652, "grad_norm": 0.3642653753975914, "learning_rate": 5.988427070179108e-06, "loss": 0.3277, "step": 954 }, { "epoch": 3.866396761133603, "grad_norm": 0.34603792137237943, "learning_rate": 5.948127379792234e-06, "loss": 0.336, "step": 955 }, { "epoch": 3.8704453441295548, "grad_norm": 0.3508434221498635, "learning_rate": 5.90794004780439e-06, "loss": 0.3276, "step": 956 }, { "epoch": 3.8744939271255063, "grad_norm": 0.3690059973344492, "learning_rate": 5.867865395552405e-06, "loss": 0.3402, "step": 957 }, { "epoch": 3.8785425101214575, "grad_norm": 0.32667250756138355, "learning_rate": 5.8279037434721364e-06, "loss": 0.343, "step": 958 }, { "epoch": 3.882591093117409, "grad_norm": 0.3536451469728077, "learning_rate": 5.788055411095892e-06, "loss": 0.3293, "step": 959 }, { "epoch": 3.8866396761133606, "grad_norm": 0.3695763205529021, "learning_rate": 5.7483207170498625e-06, "loss": 0.3253, "step": 960 }, { "epoch": 3.8906882591093117, "grad_norm": 0.38609738959851536, "learning_rate": 5.708699979051613e-06, "loss": 0.325, "step": 961 }, { "epoch": 3.8947368421052633, "grad_norm": 0.3502312200347007, "learning_rate": 5.66919351390751e-06, "loss": 0.3121, "step": 962 }, { "epoch": 3.898785425101215, "grad_norm": 0.362046714720532, "learning_rate": 5.629801637510188e-06, "loss": 0.3106, "step": 963 }, { "epoch": 3.902834008097166, "grad_norm": 0.3556457679964099, "learning_rate": 5.590524664836043e-06, "loss": 0.3352, "step": 964 }, { "epoch": 3.9068825910931175, "grad_norm": 0.37410165187776623, "learning_rate": 5.551362909942708e-06, "loss": 0.3434, "step": 965 }, { "epoch": 3.910931174089069, "grad_norm": 0.3517428998577359, "learning_rate": 5.51231668596653e-06, "loss": 0.3329, "step": 966 }, { "epoch": 3.91497975708502, "grad_norm": 0.3623811362140439, "learning_rate": 5.47338630512007e-06, "loss": 0.3289, "step": 967 }, { "epoch": 3.919028340080972, "grad_norm": 0.36834755761894866, "learning_rate": 5.434572078689617e-06, "loss": 0.3353, "step": 968 }, { "epoch": 3.9230769230769234, "grad_norm": 0.34828129689110204, "learning_rate": 5.395874317032694e-06, "loss": 0.3444, "step": 969 }, { "epoch": 3.9271255060728745, "grad_norm": 0.3736764155276103, "learning_rate": 5.357293329575573e-06, "loss": 0.321, "step": 970 }, { "epoch": 3.931174089068826, "grad_norm": 0.3374640264513952, "learning_rate": 5.318829424810797e-06, "loss": 0.3328, "step": 971 }, { "epoch": 3.9352226720647776, "grad_norm": 0.35152365192548984, "learning_rate": 5.280482910294728e-06, "loss": 0.312, "step": 972 }, { "epoch": 3.9392712550607287, "grad_norm": 0.3537095330295948, "learning_rate": 5.242254092645078e-06, "loss": 0.3142, "step": 973 }, { "epoch": 3.9433198380566803, "grad_norm": 0.36750000697199675, "learning_rate": 5.2041432775384445e-06, "loss": 0.3308, "step": 974 }, { "epoch": 3.9473684210526314, "grad_norm": 0.35696406489621707, "learning_rate": 5.166150769707892e-06, "loss": 0.3226, "step": 975 }, { "epoch": 3.951417004048583, "grad_norm": 0.3216471118477728, "learning_rate": 5.128276872940503e-06, "loss": 0.3395, "step": 976 }, { "epoch": 3.9554655870445345, "grad_norm": 0.3556170140333896, "learning_rate": 5.090521890074953e-06, "loss": 0.3296, "step": 977 }, { "epoch": 3.9595141700404857, "grad_norm": 0.3560314044640428, "learning_rate": 5.0528861229990675e-06, "loss": 0.3199, "step": 978 }, { "epoch": 3.9635627530364372, "grad_norm": 0.43289341181709495, "learning_rate": 5.015369872647442e-06, "loss": 0.3078, "step": 979 }, { "epoch": 3.967611336032389, "grad_norm": 0.36337165769519325, "learning_rate": 4.9779734389990195e-06, "loss": 0.3053, "step": 980 }, { "epoch": 3.97165991902834, "grad_norm": 0.3489179732461242, "learning_rate": 4.940697121074675e-06, "loss": 0.3394, "step": 981 }, { "epoch": 3.9757085020242915, "grad_norm": 0.35661753598544227, "learning_rate": 4.903541216934864e-06, "loss": 0.3379, "step": 982 }, { "epoch": 3.979757085020243, "grad_norm": 0.3281293162337046, "learning_rate": 4.8665060236772e-06, "loss": 0.3435, "step": 983 }, { "epoch": 3.983805668016194, "grad_norm": 0.3412052663787814, "learning_rate": 4.829591837434102e-06, "loss": 0.3295, "step": 984 }, { "epoch": 3.9878542510121457, "grad_norm": 0.38179519726900335, "learning_rate": 4.792798953370424e-06, "loss": 0.3135, "step": 985 }, { "epoch": 3.9919028340080973, "grad_norm": 0.3553661623673057, "learning_rate": 4.756127665681083e-06, "loss": 0.3276, "step": 986 }, { "epoch": 3.9959514170040484, "grad_norm": 0.34685536802881395, "learning_rate": 4.719578267588725e-06, "loss": 0.3364, "step": 987 }, { "epoch": 4.0, "grad_norm": 0.32780203735759794, "learning_rate": 4.6831510513413525e-06, "loss": 0.3221, "step": 988 }, { "epoch": 4.004048582995951, "grad_norm": 0.6300562963279791, "learning_rate": 4.6468463082100176e-06, "loss": 0.2651, "step": 989 }, { "epoch": 4.008097165991903, "grad_norm": 0.48806101757020226, "learning_rate": 4.610664328486485e-06, "loss": 0.262, "step": 990 }, { "epoch": 4.012145748987854, "grad_norm": 0.3716727548694136, "learning_rate": 4.5746054014809e-06, "loss": 0.2416, "step": 991 }, { "epoch": 4.016194331983805, "grad_norm": 0.4556959768059351, "learning_rate": 4.5386698155194766e-06, "loss": 0.2539, "step": 992 }, { "epoch": 4.020242914979757, "grad_norm": 0.5756429864347833, "learning_rate": 4.502857857942204e-06, "loss": 0.2431, "step": 993 }, { "epoch": 4.0242914979757085, "grad_norm": 0.6154833282867375, "learning_rate": 4.467169815100549e-06, "loss": 0.2615, "step": 994 }, { "epoch": 4.02834008097166, "grad_norm": 0.5065454020339065, "learning_rate": 4.431605972355137e-06, "loss": 0.2575, "step": 995 }, { "epoch": 4.032388663967612, "grad_norm": 0.5004514788572547, "learning_rate": 4.396166614073516e-06, "loss": 0.2585, "step": 996 }, { "epoch": 4.036437246963563, "grad_norm": 0.5405684721963787, "learning_rate": 4.360852023627846e-06, "loss": 0.2465, "step": 997 }, { "epoch": 4.040485829959514, "grad_norm": 0.5572575915124711, "learning_rate": 4.325662483392656e-06, "loss": 0.2526, "step": 998 }, { "epoch": 4.044534412955466, "grad_norm": 0.4940054111168175, "learning_rate": 4.29059827474257e-06, "loss": 0.2546, "step": 999 }, { "epoch": 4.048582995951417, "grad_norm": 0.3972956971622546, "learning_rate": 4.255659678050063e-06, "loss": 0.2505, "step": 1000 }, { "epoch": 4.052631578947368, "grad_norm": 0.41009464702230847, "learning_rate": 4.22084697268323e-06, "loss": 0.2561, "step": 1001 }, { "epoch": 4.05668016194332, "grad_norm": 0.43331115586558827, "learning_rate": 4.186160437003535e-06, "loss": 0.2432, "step": 1002 }, { "epoch": 4.060728744939271, "grad_norm": 0.5029906330455525, "learning_rate": 4.151600348363582e-06, "loss": 0.2548, "step": 1003 }, { "epoch": 4.064777327935222, "grad_norm": 0.4825820257806159, "learning_rate": 4.11716698310493e-06, "loss": 0.2367, "step": 1004 }, { "epoch": 4.068825910931174, "grad_norm": 0.37905287963437323, "learning_rate": 4.082860616555848e-06, "loss": 0.2403, "step": 1005 }, { "epoch": 4.0728744939271255, "grad_norm": 0.37039627593124136, "learning_rate": 4.048681523029125e-06, "loss": 0.2455, "step": 1006 }, { "epoch": 4.076923076923077, "grad_norm": 0.3872390411198614, "learning_rate": 4.014629975819882e-06, "loss": 0.25, "step": 1007 }, { "epoch": 4.080971659919029, "grad_norm": 0.41653246101123304, "learning_rate": 3.9807062472033855e-06, "loss": 0.243, "step": 1008 }, { "epoch": 4.08502024291498, "grad_norm": 0.41289162042895705, "learning_rate": 3.94691060843287e-06, "loss": 0.2503, "step": 1009 }, { "epoch": 4.089068825910931, "grad_norm": 0.3854956732546739, "learning_rate": 3.913243329737353e-06, "loss": 0.2419, "step": 1010 }, { "epoch": 4.093117408906883, "grad_norm": 0.36987198149599887, "learning_rate": 3.879704680319503e-06, "loss": 0.2582, "step": 1011 }, { "epoch": 4.097165991902834, "grad_norm": 0.39229038916717696, "learning_rate": 3.846294928353471e-06, "loss": 0.2456, "step": 1012 }, { "epoch": 4.101214574898785, "grad_norm": 0.396953846976373, "learning_rate": 3.8130143409827345e-06, "loss": 0.231, "step": 1013 }, { "epoch": 4.105263157894737, "grad_norm": 0.3719937650967542, "learning_rate": 3.7798631843179887e-06, "loss": 0.2291, "step": 1014 }, { "epoch": 4.109311740890688, "grad_norm": 0.3784954523760989, "learning_rate": 3.7468417234349973e-06, "loss": 0.2315, "step": 1015 }, { "epoch": 4.113360323886639, "grad_norm": 0.35865311980375464, "learning_rate": 3.7139502223724823e-06, "loss": 0.2426, "step": 1016 }, { "epoch": 4.117408906882591, "grad_norm": 0.34257431544049205, "learning_rate": 3.6811889441300095e-06, "loss": 0.2432, "step": 1017 }, { "epoch": 4.1214574898785425, "grad_norm": 0.3460305753996829, "learning_rate": 3.648558150665886e-06, "loss": 0.2468, "step": 1018 }, { "epoch": 4.125506072874494, "grad_norm": 0.33161947887136534, "learning_rate": 3.616058102895075e-06, "loss": 0.2485, "step": 1019 }, { "epoch": 4.129554655870446, "grad_norm": 0.3632362520453887, "learning_rate": 3.583689060687079e-06, "loss": 0.2437, "step": 1020 }, { "epoch": 4.133603238866397, "grad_norm": 0.33892225911100465, "learning_rate": 3.5514512828639023e-06, "loss": 0.2541, "step": 1021 }, { "epoch": 4.137651821862348, "grad_norm": 0.3541744806184804, "learning_rate": 3.51934502719796e-06, "loss": 0.2341, "step": 1022 }, { "epoch": 4.1417004048583, "grad_norm": 0.3323822733653191, "learning_rate": 3.4873705504100185e-06, "loss": 0.2588, "step": 1023 }, { "epoch": 4.145748987854251, "grad_norm": 0.3438678951309742, "learning_rate": 3.4555281081671345e-06, "loss": 0.2582, "step": 1024 }, { "epoch": 4.149797570850202, "grad_norm": 0.3543382725437608, "learning_rate": 3.4238179550806326e-06, "loss": 0.2577, "step": 1025 }, { "epoch": 4.153846153846154, "grad_norm": 0.3405581732696806, "learning_rate": 3.3922403447040497e-06, "loss": 0.2473, "step": 1026 }, { "epoch": 4.157894736842105, "grad_norm": 0.3485036046346753, "learning_rate": 3.3607955295311247e-06, "loss": 0.2474, "step": 1027 }, { "epoch": 4.161943319838056, "grad_norm": 0.35448992603859286, "learning_rate": 3.32948376099375e-06, "loss": 0.2611, "step": 1028 }, { "epoch": 4.165991902834008, "grad_norm": 0.3366572837843547, "learning_rate": 3.2983052894599976e-06, "loss": 0.2442, "step": 1029 }, { "epoch": 4.17004048582996, "grad_norm": 0.333115655390729, "learning_rate": 3.2672603642320945e-06, "loss": 0.2474, "step": 1030 }, { "epoch": 4.174089068825911, "grad_norm": 0.32337289313329926, "learning_rate": 3.236349233544438e-06, "loss": 0.2562, "step": 1031 }, { "epoch": 4.178137651821863, "grad_norm": 0.33283356842860057, "learning_rate": 3.205572144561604e-06, "loss": 0.2682, "step": 1032 }, { "epoch": 4.182186234817814, "grad_norm": 0.35129487932322595, "learning_rate": 3.174929343376374e-06, "loss": 0.2313, "step": 1033 }, { "epoch": 4.186234817813765, "grad_norm": 0.33681491072843334, "learning_rate": 3.1444210750077777e-06, "loss": 0.2651, "step": 1034 }, { "epoch": 4.190283400809717, "grad_norm": 0.33442713092236387, "learning_rate": 3.1140475833991046e-06, "loss": 0.2458, "step": 1035 }, { "epoch": 4.194331983805668, "grad_norm": 0.32913364851832627, "learning_rate": 3.0838091114159895e-06, "loss": 0.2499, "step": 1036 }, { "epoch": 4.198380566801619, "grad_norm": 0.31452630280970956, "learning_rate": 3.0537059008444524e-06, "loss": 0.2698, "step": 1037 }, { "epoch": 4.202429149797571, "grad_norm": 0.33867639479857303, "learning_rate": 3.0237381923889563e-06, "loss": 0.2578, "step": 1038 }, { "epoch": 4.206477732793522, "grad_norm": 0.3561312566121435, "learning_rate": 2.9939062256704997e-06, "loss": 0.2534, "step": 1039 }, { "epoch": 4.2105263157894735, "grad_norm": 0.3482876494948953, "learning_rate": 2.964210239224696e-06, "loss": 0.243, "step": 1040 }, { "epoch": 4.2145748987854255, "grad_norm": 0.33733742287367247, "learning_rate": 2.934650470499869e-06, "loss": 0.2546, "step": 1041 }, { "epoch": 4.218623481781377, "grad_norm": 0.3263639661401672, "learning_rate": 2.905227155855128e-06, "loss": 0.2341, "step": 1042 }, { "epoch": 4.222672064777328, "grad_norm": 0.31434530580498915, "learning_rate": 2.875940530558519e-06, "loss": 0.2599, "step": 1043 }, { "epoch": 4.22672064777328, "grad_norm": 0.34165654262036843, "learning_rate": 2.8467908287851154e-06, "loss": 0.2598, "step": 1044 }, { "epoch": 4.230769230769231, "grad_norm": 0.3335698906710206, "learning_rate": 2.8177782836151603e-06, "loss": 0.2297, "step": 1045 }, { "epoch": 4.234817813765182, "grad_norm": 0.3271101074797205, "learning_rate": 2.7889031270321874e-06, "loss": 0.264, "step": 1046 }, { "epoch": 4.238866396761134, "grad_norm": 0.33521491339437887, "learning_rate": 2.760165589921184e-06, "loss": 0.2584, "step": 1047 }, { "epoch": 4.242914979757085, "grad_norm": 0.33397706906447266, "learning_rate": 2.7315659020667305e-06, "loss": 0.2492, "step": 1048 }, { "epoch": 4.246963562753036, "grad_norm": 0.339361077404235, "learning_rate": 2.703104292151164e-06, "loss": 0.2344, "step": 1049 }, { "epoch": 4.251012145748988, "grad_norm": 0.32122126864395634, "learning_rate": 2.6747809877527607e-06, "loss": 0.2512, "step": 1050 }, { "epoch": 4.255060728744939, "grad_norm": 0.36091953889253237, "learning_rate": 2.6465962153439082e-06, "loss": 0.2374, "step": 1051 }, { "epoch": 4.2591093117408905, "grad_norm": 0.3171242870654275, "learning_rate": 2.618550200289303e-06, "loss": 0.2429, "step": 1052 }, { "epoch": 4.2631578947368425, "grad_norm": 0.35036977329960095, "learning_rate": 2.59064316684412e-06, "loss": 0.2435, "step": 1053 }, { "epoch": 4.267206477732794, "grad_norm": 0.3392202485040565, "learning_rate": 2.562875338152264e-06, "loss": 0.2388, "step": 1054 }, { "epoch": 4.271255060728745, "grad_norm": 0.32913713470499123, "learning_rate": 2.5352469362445552e-06, "loss": 0.2386, "step": 1055 }, { "epoch": 4.275303643724697, "grad_norm": 0.345849956615011, "learning_rate": 2.5077581820369523e-06, "loss": 0.2411, "step": 1056 }, { "epoch": 4.279352226720648, "grad_norm": 0.33005988957966176, "learning_rate": 2.480409295328803e-06, "loss": 0.2573, "step": 1057 }, { "epoch": 4.283400809716599, "grad_norm": 0.336340149916154, "learning_rate": 2.45320049480108e-06, "loss": 0.2364, "step": 1058 }, { "epoch": 4.287449392712551, "grad_norm": 0.3436822316670492, "learning_rate": 2.4261319980146293e-06, "loss": 0.2439, "step": 1059 }, { "epoch": 4.291497975708502, "grad_norm": 0.336864464180618, "learning_rate": 2.399204021408421e-06, "loss": 0.2412, "step": 1060 }, { "epoch": 4.295546558704453, "grad_norm": 0.3272045972620206, "learning_rate": 2.3724167802978395e-06, "loss": 0.2398, "step": 1061 }, { "epoch": 4.299595141700405, "grad_norm": 0.3399313783081761, "learning_rate": 2.3457704888729625e-06, "loss": 0.2253, "step": 1062 }, { "epoch": 4.303643724696356, "grad_norm": 0.3420872089088808, "learning_rate": 2.3192653601968162e-06, "loss": 0.2391, "step": 1063 }, { "epoch": 4.3076923076923075, "grad_norm": 0.3317177419911281, "learning_rate": 2.2929016062037057e-06, "loss": 0.2606, "step": 1064 }, { "epoch": 4.3117408906882595, "grad_norm": 0.3473648780167928, "learning_rate": 2.2666794376975055e-06, "loss": 0.243, "step": 1065 }, { "epoch": 4.315789473684211, "grad_norm": 0.3358394180194201, "learning_rate": 2.24059906434998e-06, "loss": 0.2557, "step": 1066 }, { "epoch": 4.319838056680162, "grad_norm": 0.33721174135143495, "learning_rate": 2.214660694699091e-06, "loss": 0.2237, "step": 1067 }, { "epoch": 4.323886639676114, "grad_norm": 0.33454033923815824, "learning_rate": 2.188864536147355e-06, "loss": 0.2366, "step": 1068 }, { "epoch": 4.327935222672065, "grad_norm": 0.32849217501528905, "learning_rate": 2.1632107949601757e-06, "loss": 0.2496, "step": 1069 }, { "epoch": 4.331983805668016, "grad_norm": 0.3377559057835799, "learning_rate": 2.137699676264171e-06, "loss": 0.2387, "step": 1070 }, { "epoch": 4.336032388663968, "grad_norm": 0.3450718685077953, "learning_rate": 2.1123313840455738e-06, "loss": 0.2243, "step": 1071 }, { "epoch": 4.340080971659919, "grad_norm": 0.3180624820016626, "learning_rate": 2.08710612114857e-06, "loss": 0.2485, "step": 1072 }, { "epoch": 4.34412955465587, "grad_norm": 0.3240160625886283, "learning_rate": 2.0620240892736953e-06, "loss": 0.2531, "step": 1073 }, { "epoch": 4.348178137651822, "grad_norm": 0.33931909842070046, "learning_rate": 2.0370854889762e-06, "loss": 0.2266, "step": 1074 }, { "epoch": 4.352226720647773, "grad_norm": 0.3457626067224815, "learning_rate": 2.0122905196644726e-06, "loss": 0.2317, "step": 1075 }, { "epoch": 4.3562753036437245, "grad_norm": 0.3290334740705453, "learning_rate": 1.987639379598427e-06, "loss": 0.2407, "step": 1076 }, { "epoch": 4.3603238866396765, "grad_norm": 0.3300699247967724, "learning_rate": 1.963132265887919e-06, "loss": 0.2421, "step": 1077 }, { "epoch": 4.364372469635628, "grad_norm": 0.3438486482426197, "learning_rate": 1.9387693744911784e-06, "loss": 0.2259, "step": 1078 }, { "epoch": 4.368421052631579, "grad_norm": 0.3400712819947121, "learning_rate": 1.9145509002132368e-06, "loss": 0.2468, "step": 1079 }, { "epoch": 4.372469635627531, "grad_norm": 0.34464421297117703, "learning_rate": 1.8904770367043723e-06, "loss": 0.2294, "step": 1080 }, { "epoch": 4.376518218623482, "grad_norm": 0.3325373811464626, "learning_rate": 1.8665479764585504e-06, "loss": 0.2476, "step": 1081 }, { "epoch": 4.380566801619433, "grad_norm": 0.3280756433571564, "learning_rate": 1.8427639108118977e-06, "loss": 0.2513, "step": 1082 }, { "epoch": 4.384615384615385, "grad_norm": 0.3219392147897554, "learning_rate": 1.8191250299411755e-06, "loss": 0.228, "step": 1083 }, { "epoch": 4.388663967611336, "grad_norm": 0.3516789427310265, "learning_rate": 1.7956315228622446e-06, "loss": 0.2488, "step": 1084 }, { "epoch": 4.392712550607287, "grad_norm": 0.32269542026718934, "learning_rate": 1.7722835774285553e-06, "loss": 0.2558, "step": 1085 }, { "epoch": 4.396761133603239, "grad_norm": 0.3326409627240115, "learning_rate": 1.7490813803296624e-06, "loss": 0.247, "step": 1086 }, { "epoch": 4.40080971659919, "grad_norm": 0.3519969944105961, "learning_rate": 1.726025117089718e-06, "loss": 0.2115, "step": 1087 }, { "epoch": 4.4048582995951415, "grad_norm": 0.32476481393210843, "learning_rate": 1.70311497206598e-06, "loss": 0.2363, "step": 1088 }, { "epoch": 4.4089068825910935, "grad_norm": 0.34273322349431506, "learning_rate": 1.6803511284473638e-06, "loss": 0.2296, "step": 1089 }, { "epoch": 4.412955465587045, "grad_norm": 0.3289239340543521, "learning_rate": 1.6577337682529538e-06, "loss": 0.2577, "step": 1090 }, { "epoch": 4.417004048582996, "grad_norm": 0.32781843265226085, "learning_rate": 1.6352630723305617e-06, "loss": 0.2462, "step": 1091 }, { "epoch": 4.421052631578947, "grad_norm": 0.3344488416450517, "learning_rate": 1.612939220355272e-06, "loss": 0.239, "step": 1092 }, { "epoch": 4.425101214574899, "grad_norm": 0.32886410297408863, "learning_rate": 1.5907623908280112e-06, "loss": 0.2718, "step": 1093 }, { "epoch": 4.42914979757085, "grad_norm": 0.3183599020920596, "learning_rate": 1.568732761074121e-06, "loss": 0.2535, "step": 1094 }, { "epoch": 4.433198380566802, "grad_norm": 0.32412376447384, "learning_rate": 1.5468505072419237e-06, "loss": 0.2525, "step": 1095 }, { "epoch": 4.437246963562753, "grad_norm": 0.317609024730863, "learning_rate": 1.5251158043013448e-06, "loss": 0.2562, "step": 1096 }, { "epoch": 4.441295546558704, "grad_norm": 0.32176653535401395, "learning_rate": 1.5035288260424885e-06, "loss": 0.2455, "step": 1097 }, { "epoch": 4.445344129554655, "grad_norm": 0.33148160700184226, "learning_rate": 1.4820897450742623e-06, "loss": 0.2361, "step": 1098 }, { "epoch": 4.449392712550607, "grad_norm": 0.332186122339974, "learning_rate": 1.4607987328229767e-06, "loss": 0.2576, "step": 1099 }, { "epoch": 4.4534412955465585, "grad_norm": 0.3421076814972842, "learning_rate": 1.4396559595310056e-06, "loss": 0.2429, "step": 1100 }, { "epoch": 4.4574898785425106, "grad_norm": 0.3277921742404346, "learning_rate": 1.4186615942553973e-06, "loss": 0.2482, "step": 1101 }, { "epoch": 4.461538461538462, "grad_norm": 0.3319744933524229, "learning_rate": 1.3978158048665424e-06, "loss": 0.2394, "step": 1102 }, { "epoch": 4.465587044534413, "grad_norm": 0.32425828857212374, "learning_rate": 1.3771187580468115e-06, "loss": 0.2473, "step": 1103 }, { "epoch": 4.469635627530364, "grad_norm": 0.3274893841683791, "learning_rate": 1.3565706192892392e-06, "loss": 0.2409, "step": 1104 }, { "epoch": 4.473684210526316, "grad_norm": 0.31486157039368184, "learning_rate": 1.3361715528961993e-06, "loss": 0.2537, "step": 1105 }, { "epoch": 4.477732793522267, "grad_norm": 0.3345841255934486, "learning_rate": 1.3159217219780817e-06, "loss": 0.2554, "step": 1106 }, { "epoch": 4.481781376518219, "grad_norm": 0.3151019493760541, "learning_rate": 1.2958212884519949e-06, "loss": 0.2688, "step": 1107 }, { "epoch": 4.48582995951417, "grad_norm": 0.31626459763285886, "learning_rate": 1.2758704130404675e-06, "loss": 0.2524, "step": 1108 }, { "epoch": 4.489878542510121, "grad_norm": 0.32990368068154935, "learning_rate": 1.2560692552701738e-06, "loss": 0.2588, "step": 1109 }, { "epoch": 4.493927125506072, "grad_norm": 0.3314313524062429, "learning_rate": 1.2364179734706382e-06, "loss": 0.2412, "step": 1110 }, { "epoch": 4.497975708502024, "grad_norm": 0.3263419740997294, "learning_rate": 1.2169167247729852e-06, "loss": 0.236, "step": 1111 }, { "epoch": 4.502024291497976, "grad_norm": 0.3296659383523177, "learning_rate": 1.1975656651086864e-06, "loss": 0.2517, "step": 1112 }, { "epoch": 4.506072874493928, "grad_norm": 1.6345356152082535, "learning_rate": 1.1783649492082926e-06, "loss": 0.2513, "step": 1113 }, { "epoch": 4.510121457489879, "grad_norm": 0.33207742235448023, "learning_rate": 1.1593147306002183e-06, "loss": 0.2532, "step": 1114 }, { "epoch": 4.51417004048583, "grad_norm": 0.3356629680438009, "learning_rate": 1.140415161609505e-06, "loss": 0.2356, "step": 1115 }, { "epoch": 4.518218623481781, "grad_norm": 0.3261021247464166, "learning_rate": 1.121666393356604e-06, "loss": 0.2715, "step": 1116 }, { "epoch": 4.522267206477733, "grad_norm": 0.33310142896388856, "learning_rate": 1.1030685757561632e-06, "loss": 0.2542, "step": 1117 }, { "epoch": 4.526315789473684, "grad_norm": 0.33533264305107924, "learning_rate": 1.0846218575158373e-06, "loss": 0.2435, "step": 1118 }, { "epoch": 4.530364372469636, "grad_norm": 0.33142355958614633, "learning_rate": 1.066326386135097e-06, "loss": 0.243, "step": 1119 }, { "epoch": 4.534412955465587, "grad_norm": 0.3169086678710102, "learning_rate": 1.0481823079040398e-06, "loss": 0.2666, "step": 1120 }, { "epoch": 4.538461538461538, "grad_norm": 0.32488400963227443, "learning_rate": 1.0301897679022344e-06, "loss": 0.2535, "step": 1121 }, { "epoch": 4.5425101214574894, "grad_norm": 0.3203099871569196, "learning_rate": 1.0123489099975491e-06, "loss": 0.2624, "step": 1122 }, { "epoch": 4.5465587044534415, "grad_norm": 0.3224180129199278, "learning_rate": 9.9465987684501e-07, "loss": 0.2334, "step": 1123 }, { "epoch": 4.550607287449393, "grad_norm": 0.3407289533318673, "learning_rate": 9.771228098856534e-07, "loss": 0.2335, "step": 1124 }, { "epoch": 4.554655870445345, "grad_norm": 0.32818469474902046, "learning_rate": 9.597378493454013e-07, "loss": 0.2359, "step": 1125 }, { "epoch": 4.558704453441296, "grad_norm": 0.34171915480937254, "learning_rate": 9.425051342339353e-07, "loss": 0.2428, "step": 1126 }, { "epoch": 4.562753036437247, "grad_norm": 0.32690214894503566, "learning_rate": 9.254248023435864e-07, "loss": 0.2487, "step": 1127 }, { "epoch": 4.566801619433198, "grad_norm": 0.31069986271327704, "learning_rate": 9.084969902482288e-07, "loss": 0.2466, "step": 1128 }, { "epoch": 4.57085020242915, "grad_norm": 0.3442609693122383, "learning_rate": 8.917218333021993e-07, "loss": 0.252, "step": 1129 }, { "epoch": 4.574898785425101, "grad_norm": 0.3274311767002057, "learning_rate": 8.750994656392109e-07, "loss": 0.2658, "step": 1130 }, { "epoch": 4.578947368421053, "grad_norm": 0.32066610809478097, "learning_rate": 8.58630020171265e-07, "loss": 0.241, "step": 1131 }, { "epoch": 4.582995951417004, "grad_norm": 0.32193109757501864, "learning_rate": 8.423136285876099e-07, "loss": 0.2617, "step": 1132 }, { "epoch": 4.587044534412955, "grad_norm": 0.3154792648875773, "learning_rate": 8.26150421353682e-07, "loss": 0.2432, "step": 1133 }, { "epoch": 4.5910931174089065, "grad_norm": 0.3290147460371909, "learning_rate": 8.101405277100549e-07, "loss": 0.2505, "step": 1134 }, { "epoch": 4.5951417004048585, "grad_norm": 0.32980752261189183, "learning_rate": 7.942840756714077e-07, "loss": 0.2235, "step": 1135 }, { "epoch": 4.59919028340081, "grad_norm": 0.34088662073494563, "learning_rate": 7.785811920255093e-07, "loss": 0.2246, "step": 1136 }, { "epoch": 4.603238866396762, "grad_norm": 0.33720357600047907, "learning_rate": 7.630320023322024e-07, "loss": 0.2305, "step": 1137 }, { "epoch": 4.607287449392713, "grad_norm": 0.3267811889779248, "learning_rate": 7.476366309223903e-07, "loss": 0.2532, "step": 1138 }, { "epoch": 4.611336032388664, "grad_norm": 0.32885001874051706, "learning_rate": 7.323952008970537e-07, "loss": 0.2451, "step": 1139 }, { "epoch": 4.615384615384615, "grad_norm": 0.32083003422379924, "learning_rate": 7.173078341262618e-07, "loss": 0.2439, "step": 1140 }, { "epoch": 4.619433198380567, "grad_norm": 0.32848648749491505, "learning_rate": 7.023746512481989e-07, "loss": 0.2618, "step": 1141 }, { "epoch": 4.623481781376518, "grad_norm": 0.32085657949853175, "learning_rate": 6.875957716681902e-07, "loss": 0.2322, "step": 1142 }, { "epoch": 4.62753036437247, "grad_norm": 0.34221079614242395, "learning_rate": 6.729713135577687e-07, "loss": 0.2437, "step": 1143 }, { "epoch": 4.631578947368421, "grad_norm": 0.3178467366070836, "learning_rate": 6.585013938537143e-07, "loss": 0.2596, "step": 1144 }, { "epoch": 4.635627530364372, "grad_norm": 0.31608811724887614, "learning_rate": 6.441861282571115e-07, "loss": 0.2503, "step": 1145 }, { "epoch": 4.6396761133603235, "grad_norm": 0.3291228741532958, "learning_rate": 6.300256312324471e-07, "loss": 0.2481, "step": 1146 }, { "epoch": 4.6437246963562755, "grad_norm": 0.3129593214390824, "learning_rate": 6.160200160066776e-07, "loss": 0.2607, "step": 1147 }, { "epoch": 4.647773279352227, "grad_norm": 0.32336938256222525, "learning_rate": 6.021693945683327e-07, "loss": 0.2469, "step": 1148 }, { "epoch": 4.651821862348179, "grad_norm": 0.3279911525870542, "learning_rate": 5.884738776666044e-07, "loss": 0.2376, "step": 1149 }, { "epoch": 4.65587044534413, "grad_norm": 0.32929797080313966, "learning_rate": 5.749335748104878e-07, "loss": 0.2563, "step": 1150 }, { "epoch": 4.659919028340081, "grad_norm": 0.32463851713625014, "learning_rate": 5.615485942678778e-07, "loss": 0.2498, "step": 1151 }, { "epoch": 4.663967611336032, "grad_norm": 0.3370378535403398, "learning_rate": 5.483190430647289e-07, "loss": 0.2523, "step": 1152 }, { "epoch": 4.668016194331984, "grad_norm": 0.3257390168511157, "learning_rate": 5.352450269841747e-07, "loss": 0.2375, "step": 1153 }, { "epoch": 4.672064777327935, "grad_norm": 0.3169987645219248, "learning_rate": 5.22326650565701e-07, "loss": 0.2508, "step": 1154 }, { "epoch": 4.676113360323887, "grad_norm": 0.31674987736269766, "learning_rate": 5.095640171043048e-07, "loss": 0.2384, "step": 1155 }, { "epoch": 4.680161943319838, "grad_norm": 0.32891003213393794, "learning_rate": 4.969572286496594e-07, "loss": 0.2486, "step": 1156 }, { "epoch": 4.684210526315789, "grad_norm": 0.3378544350897599, "learning_rate": 4.845063860053079e-07, "loss": 0.2539, "step": 1157 }, { "epoch": 4.6882591093117405, "grad_norm": 0.3293940176825013, "learning_rate": 4.7221158872785997e-07, "loss": 0.2314, "step": 1158 }, { "epoch": 4.6923076923076925, "grad_norm": 0.3477091630393519, "learning_rate": 4.600729351261857e-07, "loss": 0.2397, "step": 1159 }, { "epoch": 4.696356275303644, "grad_norm": 0.3238788980923074, "learning_rate": 4.4809052226063355e-07, "loss": 0.2523, "step": 1160 }, { "epoch": 4.700404858299595, "grad_norm": 0.34037548258919986, "learning_rate": 4.3626444594226046e-07, "loss": 0.2431, "step": 1161 }, { "epoch": 4.704453441295547, "grad_norm": 0.32136190033702666, "learning_rate": 4.245948007320588e-07, "loss": 0.2454, "step": 1162 }, { "epoch": 4.708502024291498, "grad_norm": 0.31552810100818995, "learning_rate": 4.130816799401971e-07, "loss": 0.2549, "step": 1163 }, { "epoch": 4.712550607287449, "grad_norm": 0.32504935272311736, "learning_rate": 4.017251756252827e-07, "loss": 0.2409, "step": 1164 }, { "epoch": 4.716599190283401, "grad_norm": 0.33821112521231905, "learning_rate": 3.9052537859362473e-07, "loss": 0.2557, "step": 1165 }, { "epoch": 4.720647773279352, "grad_norm": 0.34701196981812377, "learning_rate": 3.7948237839849687e-07, "loss": 0.2342, "step": 1166 }, { "epoch": 4.724696356275303, "grad_norm": 0.3313732056936725, "learning_rate": 3.6859626333943797e-07, "loss": 0.2337, "step": 1167 }, { "epoch": 4.728744939271255, "grad_norm": 0.3267819831561852, "learning_rate": 3.578671204615347e-07, "loss": 0.2442, "step": 1168 }, { "epoch": 4.732793522267206, "grad_norm": 0.31265410009682904, "learning_rate": 3.472950355547267e-07, "loss": 0.251, "step": 1169 }, { "epoch": 4.7368421052631575, "grad_norm": 0.32298868967357436, "learning_rate": 3.3688009315312244e-07, "loss": 0.2509, "step": 1170 }, { "epoch": 4.7408906882591095, "grad_norm": 0.3157351046377557, "learning_rate": 3.266223765343224e-07, "loss": 0.2442, "step": 1171 }, { "epoch": 4.744939271255061, "grad_norm": 0.3354218959807653, "learning_rate": 3.16521967718757e-07, "loss": 0.235, "step": 1172 }, { "epoch": 4.748987854251012, "grad_norm": 0.3357740937965887, "learning_rate": 3.0657894746902726e-07, "loss": 0.2362, "step": 1173 }, { "epoch": 4.753036437246964, "grad_norm": 0.332397810857382, "learning_rate": 2.9679339528925654e-07, "loss": 0.2477, "step": 1174 }, { "epoch": 4.757085020242915, "grad_norm": 0.3334836082793273, "learning_rate": 2.871653894244597e-07, "loss": 0.2246, "step": 1175 }, { "epoch": 4.761133603238866, "grad_norm": 0.3327887175511132, "learning_rate": 2.7769500685991266e-07, "loss": 0.2438, "step": 1176 }, { "epoch": 4.765182186234818, "grad_norm": 0.328837940537458, "learning_rate": 2.683823233205485e-07, "loss": 0.2547, "step": 1177 }, { "epoch": 4.769230769230769, "grad_norm": 0.33317033039728333, "learning_rate": 2.592274132703287e-07, "loss": 0.2588, "step": 1178 }, { "epoch": 4.77327935222672, "grad_norm": 0.3378010626491132, "learning_rate": 2.502303499116754e-07, "loss": 0.2475, "step": 1179 }, { "epoch": 4.777327935222672, "grad_norm": 0.3206248040615858, "learning_rate": 2.413912051848666e-07, "loss": 0.218, "step": 1180 }, { "epoch": 4.781376518218623, "grad_norm": 0.31982925085700725, "learning_rate": 2.3271004976746613e-07, "loss": 0.2378, "step": 1181 }, { "epoch": 4.7854251012145745, "grad_norm": 0.3414340838114226, "learning_rate": 2.2418695307376393e-07, "loss": 0.2385, "step": 1182 }, { "epoch": 4.7894736842105265, "grad_norm": 0.32638525848794503, "learning_rate": 2.1582198325420968e-07, "loss": 0.257, "step": 1183 }, { "epoch": 4.793522267206478, "grad_norm": 0.3262890523775591, "learning_rate": 2.0761520719488447e-07, "loss": 0.2447, "step": 1184 }, { "epoch": 4.797570850202429, "grad_norm": 0.32766434581579074, "learning_rate": 1.9956669051694798e-07, "loss": 0.25, "step": 1185 }, { "epoch": 4.801619433198381, "grad_norm": 0.32719247288247, "learning_rate": 1.916764975761254e-07, "loss": 0.2637, "step": 1186 }, { "epoch": 4.805668016194332, "grad_norm": 0.3197686978078804, "learning_rate": 1.839446914621923e-07, "loss": 0.2599, "step": 1187 }, { "epoch": 4.809716599190283, "grad_norm": 0.33714156532219597, "learning_rate": 1.7637133399845962e-07, "loss": 0.2297, "step": 1188 }, { "epoch": 4.813765182186235, "grad_norm": 0.32252014882855223, "learning_rate": 1.6895648574129397e-07, "loss": 0.243, "step": 1189 }, { "epoch": 4.817813765182186, "grad_norm": 0.3207262012613356, "learning_rate": 1.6170020597962465e-07, "loss": 0.2513, "step": 1190 }, { "epoch": 4.821862348178137, "grad_norm": 0.3290793569820913, "learning_rate": 1.5460255273447078e-07, "loss": 0.2399, "step": 1191 }, { "epoch": 4.825910931174089, "grad_norm": 0.32030611936689024, "learning_rate": 1.4766358275847715e-07, "loss": 0.248, "step": 1192 }, { "epoch": 4.82995951417004, "grad_norm": 0.30332290151621977, "learning_rate": 1.4088335153546573e-07, "loss": 0.2635, "step": 1193 }, { "epoch": 4.834008097165992, "grad_norm": 0.31850884600514695, "learning_rate": 1.3426191327998496e-07, "loss": 0.2707, "step": 1194 }, { "epoch": 4.838056680161944, "grad_norm": 0.3257267567862206, "learning_rate": 1.277993209368744e-07, "loss": 0.2481, "step": 1195 }, { "epoch": 4.842105263157895, "grad_norm": 0.31418424140975715, "learning_rate": 1.214956261808542e-07, "loss": 0.2449, "step": 1196 }, { "epoch": 4.846153846153846, "grad_norm": 0.3134842427688657, "learning_rate": 1.1535087941609846e-07, "loss": 0.2505, "step": 1197 }, { "epoch": 4.850202429149798, "grad_norm": 0.3429666310514976, "learning_rate": 1.0936512977583802e-07, "loss": 0.2486, "step": 1198 }, { "epoch": 4.854251012145749, "grad_norm": 0.3126599759138714, "learning_rate": 1.0353842512196732e-07, "loss": 0.2511, "step": 1199 }, { "epoch": 4.8582995951417, "grad_norm": 0.3088994589682609, "learning_rate": 9.787081204466253e-08, "loss": 0.2553, "step": 1200 }, { "epoch": 4.862348178137652, "grad_norm": 0.3057263760063486, "learning_rate": 9.236233586200627e-08, "loss": 0.2576, "step": 1201 }, { "epoch": 4.866396761133603, "grad_norm": 0.3241247716702819, "learning_rate": 8.70130406196279e-08, "loss": 0.262, "step": 1202 }, { "epoch": 4.870445344129554, "grad_norm": 0.3329567188327396, "learning_rate": 8.182296909035047e-08, "loss": 0.2508, "step": 1203 }, { "epoch": 4.874493927125506, "grad_norm": 0.3216254125543982, "learning_rate": 7.679216277384439e-08, "loss": 0.2458, "step": 1204 }, { "epoch": 4.8785425101214575, "grad_norm": 0.3274848891240613, "learning_rate": 7.19206618963053e-08, "loss": 0.251, "step": 1205 }, { "epoch": 4.882591093117409, "grad_norm": 0.3258503781765499, "learning_rate": 6.720850541012347e-08, "loss": 0.2535, "step": 1206 }, { "epoch": 4.886639676113361, "grad_norm": 0.3371255006591497, "learning_rate": 6.265573099357714e-08, "loss": 0.2282, "step": 1207 }, { "epoch": 4.890688259109312, "grad_norm": 0.32250129782741405, "learning_rate": 5.826237505053067e-08, "loss": 0.236, "step": 1208 }, { "epoch": 4.894736842105263, "grad_norm": 0.3093740182677429, "learning_rate": 5.4028472710137e-08, "loss": 0.252, "step": 1209 }, { "epoch": 4.898785425101215, "grad_norm": 0.3088576718298926, "learning_rate": 4.9954057826571146e-08, "loss": 0.2482, "step": 1210 }, { "epoch": 4.902834008097166, "grad_norm": 0.3104065152314839, "learning_rate": 4.6039162978743776e-08, "loss": 0.2654, "step": 1211 }, { "epoch": 4.906882591093117, "grad_norm": 0.3158654517004, "learning_rate": 4.228381947004812e-08, "loss": 0.2552, "step": 1212 }, { "epoch": 4.910931174089069, "grad_norm": 0.3083334103701194, "learning_rate": 3.868805732811343e-08, "loss": 0.2497, "step": 1213 }, { "epoch": 4.91497975708502, "grad_norm": 0.32406889066469474, "learning_rate": 3.525190530455636e-08, "loss": 0.2483, "step": 1214 }, { "epoch": 4.919028340080971, "grad_norm": 0.33195391774161487, "learning_rate": 3.197539087475887e-08, "loss": 0.2496, "step": 1215 }, { "epoch": 4.923076923076923, "grad_norm": 0.3247401811106408, "learning_rate": 2.8858540237641786e-08, "loss": 0.2498, "step": 1216 }, { "epoch": 4.9271255060728745, "grad_norm": 0.3113510445871997, "learning_rate": 2.5901378315458247e-08, "loss": 0.249, "step": 1217 }, { "epoch": 4.931174089068826, "grad_norm": 0.31757423245563404, "learning_rate": 2.3103928753598347e-08, "loss": 0.2618, "step": 1218 }, { "epoch": 4.935222672064778, "grad_norm": 0.3303864451706649, "learning_rate": 2.0466213920393718e-08, "loss": 0.2458, "step": 1219 }, { "epoch": 4.939271255060729, "grad_norm": 0.3314898486950747, "learning_rate": 1.7988254906942115e-08, "loss": 0.2675, "step": 1220 }, { "epoch": 4.94331983805668, "grad_norm": 0.32068880218700696, "learning_rate": 1.5670071526936447e-08, "loss": 0.2396, "step": 1221 }, { "epoch": 4.947368421052632, "grad_norm": 0.3434001159554171, "learning_rate": 1.3511682316509344e-08, "loss": 0.2389, "step": 1222 }, { "epoch": 4.951417004048583, "grad_norm": 0.3256657844439468, "learning_rate": 1.1513104534086605e-08, "loss": 0.2404, "step": 1223 }, { "epoch": 4.955465587044534, "grad_norm": 0.32248628791891865, "learning_rate": 9.67435416023843e-09, "loss": 0.2573, "step": 1224 }, { "epoch": 4.959514170040486, "grad_norm": 0.3331658809873693, "learning_rate": 7.995445897566178e-09, "loss": 0.2349, "step": 1225 }, { "epoch": 4.963562753036437, "grad_norm": 0.3411861846500052, "learning_rate": 6.476393170578021e-09, "loss": 0.2491, "step": 1226 }, { "epoch": 4.967611336032388, "grad_norm": 0.3413738621401125, "learning_rate": 5.11720812557348e-09, "loss": 0.2301, "step": 1227 }, { "epoch": 4.97165991902834, "grad_norm": 0.32918014193498635, "learning_rate": 3.917901630563492e-09, "loss": 0.2363, "step": 1228 }, { "epoch": 4.9757085020242915, "grad_norm": 0.3243411625848123, "learning_rate": 2.878483275170485e-09, "loss": 0.2331, "step": 1229 }, { "epoch": 4.979757085020243, "grad_norm": 0.3135855003807063, "learning_rate": 1.998961370555108e-09, "loss": 0.2614, "step": 1230 }, { "epoch": 4.983805668016195, "grad_norm": 0.3259484629039563, "learning_rate": 1.2793429493518361e-09, "loss": 0.246, "step": 1231 }, { "epoch": 4.987854251012146, "grad_norm": 0.3099751348950676, "learning_rate": 7.196337656112384e-10, "loss": 0.2606, "step": 1232 }, { "epoch": 4.991902834008097, "grad_norm": 0.333858361270268, "learning_rate": 3.198382947511292e-10, "loss": 0.2294, "step": 1233 }, { "epoch": 4.995951417004049, "grad_norm": 0.31283539376575714, "learning_rate": 7.995973352770137e-11, "loss": 0.2553, "step": 1234 }, { "epoch": 5.0, "grad_norm": 0.32080921924510075, "learning_rate": 0.0, "loss": 0.2465, "step": 1235 }, { "epoch": 5.0, "step": 1235, "total_flos": 1402592641024000.0, "train_loss": 0.45633699483717016, "train_runtime": 20091.7844, "train_samples_per_second": 7.864, "train_steps_per_second": 0.061 } ], "logging_steps": 1, "max_steps": 1235, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1402592641024000.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }