diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,8687 @@ +{ + "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": 6.99357402667285, + "learning_rate": 3.2258064516129035e-07, + "loss": 1.0703, + "step": 1 + }, + { + "epoch": 0.008097165991902834, + "grad_norm": 6.721489218361523, + "learning_rate": 6.451612903225807e-07, + "loss": 1.0422, + "step": 2 + }, + { + "epoch": 0.012145748987854251, + "grad_norm": 6.794493242844992, + "learning_rate": 9.67741935483871e-07, + "loss": 1.0455, + "step": 3 + }, + { + "epoch": 0.016194331983805668, + "grad_norm": 6.586741682932585, + "learning_rate": 1.2903225806451614e-06, + "loss": 1.0278, + "step": 4 + }, + { + "epoch": 0.020242914979757085, + "grad_norm": 6.434679439050592, + "learning_rate": 1.6129032258064516e-06, + "loss": 1.0335, + "step": 5 + }, + { + "epoch": 0.024291497975708502, + "grad_norm": 6.084167301323337, + "learning_rate": 1.935483870967742e-06, + "loss": 1.036, + "step": 6 + }, + { + "epoch": 0.02834008097165992, + "grad_norm": 4.992537290338143, + "learning_rate": 2.2580645161290324e-06, + "loss": 1.0065, + "step": 7 + }, + { + "epoch": 0.032388663967611336, + "grad_norm": 4.662475009334241, + "learning_rate": 2.580645161290323e-06, + "loss": 0.9905, + "step": 8 + }, + { + "epoch": 0.03643724696356275, + "grad_norm": 2.930177245931705, + "learning_rate": 2.903225806451613e-06, + "loss": 0.9438, + "step": 9 + }, + { + "epoch": 0.04048582995951417, + "grad_norm": 2.8080554017719606, + "learning_rate": 3.225806451612903e-06, + "loss": 0.9581, + "step": 10 + }, + { + "epoch": 0.044534412955465584, + "grad_norm": 2.6589380653632206, + "learning_rate": 3.548387096774194e-06, + "loss": 0.9406, + "step": 11 + }, + { + "epoch": 0.048582995951417005, + "grad_norm": 3.9389525358274273, + "learning_rate": 3.870967741935484e-06, + "loss": 0.8996, + "step": 12 + }, + { + "epoch": 0.05263157894736842, + "grad_norm": 4.3664166858395985, + "learning_rate": 4.193548387096774e-06, + "loss": 0.914, + "step": 13 + }, + { + "epoch": 0.05668016194331984, + "grad_norm": 4.182982803258953, + "learning_rate": 4.516129032258065e-06, + "loss": 0.8942, + "step": 14 + }, + { + "epoch": 0.06072874493927125, + "grad_norm": 3.979475587381267, + "learning_rate": 4.838709677419355e-06, + "loss": 0.9039, + "step": 15 + }, + { + "epoch": 0.06477732793522267, + "grad_norm": 2.6189107686677726, + "learning_rate": 5.161290322580646e-06, + "loss": 0.8537, + "step": 16 + }, + { + "epoch": 0.06882591093117409, + "grad_norm": 2.510763004027838, + "learning_rate": 5.483870967741935e-06, + "loss": 0.834, + "step": 17 + }, + { + "epoch": 0.0728744939271255, + "grad_norm": 2.0657305326264166, + "learning_rate": 5.806451612903226e-06, + "loss": 0.8115, + "step": 18 + }, + { + "epoch": 0.07692307692307693, + "grad_norm": 1.4949157748103197, + "learning_rate": 6.129032258064517e-06, + "loss": 0.7984, + "step": 19 + }, + { + "epoch": 0.08097165991902834, + "grad_norm": 1.3571548638872974, + "learning_rate": 6.451612903225806e-06, + "loss": 0.7891, + "step": 20 + }, + { + "epoch": 0.08502024291497975, + "grad_norm": 1.2498863238814686, + "learning_rate": 6.774193548387097e-06, + "loss": 0.7791, + "step": 21 + }, + { + "epoch": 0.08906882591093117, + "grad_norm": 1.4408195567440185, + "learning_rate": 7.096774193548388e-06, + "loss": 0.7646, + "step": 22 + }, + { + "epoch": 0.0931174089068826, + "grad_norm": 1.2142250761825433, + "learning_rate": 7.4193548387096784e-06, + "loss": 0.751, + "step": 23 + }, + { + "epoch": 0.09716599190283401, + "grad_norm": 1.0792292987171785, + "learning_rate": 7.741935483870968e-06, + "loss": 0.7295, + "step": 24 + }, + { + "epoch": 0.10121457489878542, + "grad_norm": 1.0857870029191334, + "learning_rate": 8.064516129032258e-06, + "loss": 0.7168, + "step": 25 + }, + { + "epoch": 0.10526315789473684, + "grad_norm": 1.1103467019080724, + "learning_rate": 8.387096774193549e-06, + "loss": 0.7367, + "step": 26 + }, + { + "epoch": 0.10931174089068826, + "grad_norm": 1.079024227449241, + "learning_rate": 8.70967741935484e-06, + "loss": 0.7287, + "step": 27 + }, + { + "epoch": 0.11336032388663968, + "grad_norm": 0.9506764724291974, + "learning_rate": 9.03225806451613e-06, + "loss": 0.7247, + "step": 28 + }, + { + "epoch": 0.11740890688259109, + "grad_norm": 0.907505125192566, + "learning_rate": 9.35483870967742e-06, + "loss": 0.7155, + "step": 29 + }, + { + "epoch": 0.1214574898785425, + "grad_norm": 1.0007251361865117, + "learning_rate": 9.67741935483871e-06, + "loss": 0.7071, + "step": 30 + }, + { + "epoch": 0.12550607287449392, + "grad_norm": 0.9136640473646399, + "learning_rate": 1e-05, + "loss": 0.7014, + "step": 31 + }, + { + "epoch": 0.12955465587044535, + "grad_norm": 1.0570686402447855, + "learning_rate": 1.0322580645161291e-05, + "loss": 0.6751, + "step": 32 + }, + { + "epoch": 0.13360323886639677, + "grad_norm": 0.778945836563289, + "learning_rate": 1.0645161290322582e-05, + "loss": 0.6789, + "step": 33 + }, + { + "epoch": 0.13765182186234817, + "grad_norm": 0.9056628502418778, + "learning_rate": 1.096774193548387e-05, + "loss": 0.6794, + "step": 34 + }, + { + "epoch": 0.1417004048582996, + "grad_norm": 1.0723027799258256, + "learning_rate": 1.1290322580645164e-05, + "loss": 0.6858, + "step": 35 + }, + { + "epoch": 0.145748987854251, + "grad_norm": 0.6850386206301797, + "learning_rate": 1.1612903225806453e-05, + "loss": 0.671, + "step": 36 + }, + { + "epoch": 0.14979757085020243, + "grad_norm": 0.9261025733170712, + "learning_rate": 1.1935483870967743e-05, + "loss": 0.6698, + "step": 37 + }, + { + "epoch": 0.15384615384615385, + "grad_norm": 0.832450104261302, + "learning_rate": 1.2258064516129034e-05, + "loss": 0.6601, + "step": 38 + }, + { + "epoch": 0.15789473684210525, + "grad_norm": 0.8565536506794375, + "learning_rate": 1.2580645161290324e-05, + "loss": 0.6585, + "step": 39 + }, + { + "epoch": 0.16194331983805668, + "grad_norm": 1.004107349289588, + "learning_rate": 1.2903225806451613e-05, + "loss": 0.6517, + "step": 40 + }, + { + "epoch": 0.1659919028340081, + "grad_norm": 0.8074464082138924, + "learning_rate": 1.3225806451612903e-05, + "loss": 0.6557, + "step": 41 + }, + { + "epoch": 0.1700404858299595, + "grad_norm": 0.9051777505468113, + "learning_rate": 1.3548387096774194e-05, + "loss": 0.6485, + "step": 42 + }, + { + "epoch": 0.17408906882591094, + "grad_norm": 0.7294028777431649, + "learning_rate": 1.3870967741935486e-05, + "loss": 0.6398, + "step": 43 + }, + { + "epoch": 0.17813765182186234, + "grad_norm": 0.6846999434635919, + "learning_rate": 1.4193548387096776e-05, + "loss": 0.6333, + "step": 44 + }, + { + "epoch": 0.18218623481781376, + "grad_norm": 0.8914034050029572, + "learning_rate": 1.4516129032258066e-05, + "loss": 0.6366, + "step": 45 + }, + { + "epoch": 0.1862348178137652, + "grad_norm": 0.631513344361098, + "learning_rate": 1.4838709677419357e-05, + "loss": 0.6225, + "step": 46 + }, + { + "epoch": 0.1902834008097166, + "grad_norm": 0.7187248964456495, + "learning_rate": 1.5161290322580646e-05, + "loss": 0.6268, + "step": 47 + }, + { + "epoch": 0.19433198380566802, + "grad_norm": 0.7602312537953997, + "learning_rate": 1.5483870967741936e-05, + "loss": 0.6446, + "step": 48 + }, + { + "epoch": 0.19838056680161945, + "grad_norm": 0.6924594075975052, + "learning_rate": 1.5806451612903226e-05, + "loss": 0.606, + "step": 49 + }, + { + "epoch": 0.20242914979757085, + "grad_norm": 0.8146777460429339, + "learning_rate": 1.6129032258064517e-05, + "loss": 0.6354, + "step": 50 + }, + { + "epoch": 0.20647773279352227, + "grad_norm": 0.6478266665652556, + "learning_rate": 1.6451612903225807e-05, + "loss": 0.6309, + "step": 51 + }, + { + "epoch": 0.21052631578947367, + "grad_norm": 0.7375644916283652, + "learning_rate": 1.6774193548387098e-05, + "loss": 0.6116, + "step": 52 + }, + { + "epoch": 0.2145748987854251, + "grad_norm": 0.8441489116346235, + "learning_rate": 1.7096774193548388e-05, + "loss": 0.611, + "step": 53 + }, + { + "epoch": 0.21862348178137653, + "grad_norm": 0.8104301702702456, + "learning_rate": 1.741935483870968e-05, + "loss": 0.6212, + "step": 54 + }, + { + "epoch": 0.22267206477732793, + "grad_norm": 0.691193093201488, + "learning_rate": 1.774193548387097e-05, + "loss": 0.6078, + "step": 55 + }, + { + "epoch": 0.22672064777327935, + "grad_norm": 0.6934069372879649, + "learning_rate": 1.806451612903226e-05, + "loss": 0.6114, + "step": 56 + }, + { + "epoch": 0.23076923076923078, + "grad_norm": 0.7773544802965093, + "learning_rate": 1.838709677419355e-05, + "loss": 0.6041, + "step": 57 + }, + { + "epoch": 0.23481781376518218, + "grad_norm": 0.6688255765482306, + "learning_rate": 1.870967741935484e-05, + "loss": 0.6, + "step": 58 + }, + { + "epoch": 0.2388663967611336, + "grad_norm": 0.8252169846955861, + "learning_rate": 1.903225806451613e-05, + "loss": 0.6062, + "step": 59 + }, + { + "epoch": 0.242914979757085, + "grad_norm": 0.7277774318553778, + "learning_rate": 1.935483870967742e-05, + "loss": 0.6011, + "step": 60 + }, + { + "epoch": 0.24696356275303644, + "grad_norm": 0.7449614303661842, + "learning_rate": 1.967741935483871e-05, + "loss": 0.5918, + "step": 61 + }, + { + "epoch": 0.25101214574898784, + "grad_norm": 0.6499731576228293, + "learning_rate": 2e-05, + "loss": 0.5736, + "step": 62 + }, + { + "epoch": 0.2550607287449393, + "grad_norm": 0.9075212780145497, + "learning_rate": 2.0322580645161292e-05, + "loss": 0.5696, + "step": 63 + }, + { + "epoch": 0.2591093117408907, + "grad_norm": 1.120506057765324, + "learning_rate": 2.0645161290322582e-05, + "loss": 0.5908, + "step": 64 + }, + { + "epoch": 0.2631578947368421, + "grad_norm": 1.1470935930118864, + "learning_rate": 2.0967741935483873e-05, + "loss": 0.5765, + "step": 65 + }, + { + "epoch": 0.26720647773279355, + "grad_norm": 0.8181649082666178, + "learning_rate": 2.1290322580645163e-05, + "loss": 0.5716, + "step": 66 + }, + { + "epoch": 0.27125506072874495, + "grad_norm": 1.0329192244168772, + "learning_rate": 2.161290322580645e-05, + "loss": 0.5823, + "step": 67 + }, + { + "epoch": 0.27530364372469635, + "grad_norm": 1.3463080645094907, + "learning_rate": 2.193548387096774e-05, + "loss": 0.5695, + "step": 68 + }, + { + "epoch": 0.2793522267206478, + "grad_norm": 0.8882626606523801, + "learning_rate": 2.225806451612903e-05, + "loss": 0.5673, + "step": 69 + }, + { + "epoch": 0.2834008097165992, + "grad_norm": 1.0527974216397888, + "learning_rate": 2.2580645161290328e-05, + "loss": 0.5734, + "step": 70 + }, + { + "epoch": 0.2874493927125506, + "grad_norm": 1.3938997121823604, + "learning_rate": 2.2903225806451615e-05, + "loss": 0.5608, + "step": 71 + }, + { + "epoch": 0.291497975708502, + "grad_norm": 0.8513836884774304, + "learning_rate": 2.3225806451612906e-05, + "loss": 0.5682, + "step": 72 + }, + { + "epoch": 0.29554655870445345, + "grad_norm": 1.4354686983388167, + "learning_rate": 2.3548387096774196e-05, + "loss": 0.557, + "step": 73 + }, + { + "epoch": 0.29959514170040485, + "grad_norm": 0.8835654632712755, + "learning_rate": 2.3870967741935486e-05, + "loss": 0.5476, + "step": 74 + }, + { + "epoch": 0.30364372469635625, + "grad_norm": 0.8968597873878523, + "learning_rate": 2.4193548387096777e-05, + "loss": 0.5424, + "step": 75 + }, + { + "epoch": 0.3076923076923077, + "grad_norm": 1.0172790097850437, + "learning_rate": 2.4516129032258067e-05, + "loss": 0.5294, + "step": 76 + }, + { + "epoch": 0.3117408906882591, + "grad_norm": 1.1390968244180721, + "learning_rate": 2.4838709677419358e-05, + "loss": 0.5337, + "step": 77 + }, + { + "epoch": 0.3157894736842105, + "grad_norm": 0.9443925103621406, + "learning_rate": 2.5161290322580648e-05, + "loss": 0.5513, + "step": 78 + }, + { + "epoch": 0.31983805668016196, + "grad_norm": 1.1887118049919119, + "learning_rate": 2.548387096774194e-05, + "loss": 0.5274, + "step": 79 + }, + { + "epoch": 0.32388663967611336, + "grad_norm": 1.2250499246541346, + "learning_rate": 2.5806451612903226e-05, + "loss": 0.556, + "step": 80 + }, + { + "epoch": 0.32793522267206476, + "grad_norm": 1.337326002408912, + "learning_rate": 2.6129032258064516e-05, + "loss": 0.5213, + "step": 81 + }, + { + "epoch": 0.3319838056680162, + "grad_norm": 0.9131430914855333, + "learning_rate": 2.6451612903225806e-05, + "loss": 0.5249, + "step": 82 + }, + { + "epoch": 0.3360323886639676, + "grad_norm": 1.20797274677986, + "learning_rate": 2.6774193548387097e-05, + "loss": 0.5315, + "step": 83 + }, + { + "epoch": 0.340080971659919, + "grad_norm": 1.135089219614009, + "learning_rate": 2.7096774193548387e-05, + "loss": 0.5354, + "step": 84 + }, + { + "epoch": 0.3441295546558704, + "grad_norm": 1.467026926375247, + "learning_rate": 2.7419354838709678e-05, + "loss": 0.537, + "step": 85 + }, + { + "epoch": 0.3481781376518219, + "grad_norm": 0.8562144096038729, + "learning_rate": 2.774193548387097e-05, + "loss": 0.528, + "step": 86 + }, + { + "epoch": 0.3522267206477733, + "grad_norm": 1.437160201781753, + "learning_rate": 2.8064516129032262e-05, + "loss": 0.5171, + "step": 87 + }, + { + "epoch": 0.3562753036437247, + "grad_norm": 1.365017610780837, + "learning_rate": 2.8387096774193552e-05, + "loss": 0.5084, + "step": 88 + }, + { + "epoch": 0.3603238866396761, + "grad_norm": 1.0424291786859952, + "learning_rate": 2.8709677419354843e-05, + "loss": 0.5207, + "step": 89 + }, + { + "epoch": 0.3643724696356275, + "grad_norm": 2.265244820302624, + "learning_rate": 2.9032258064516133e-05, + "loss": 0.5105, + "step": 90 + }, + { + "epoch": 0.3684210526315789, + "grad_norm": 1.5072918807408553, + "learning_rate": 2.9354838709677423e-05, + "loss": 0.5232, + "step": 91 + }, + { + "epoch": 0.3724696356275304, + "grad_norm": 2.0422388651634726, + "learning_rate": 2.9677419354838714e-05, + "loss": 0.507, + "step": 92 + }, + { + "epoch": 0.3765182186234818, + "grad_norm": 2.0101577192664926, + "learning_rate": 3.0000000000000004e-05, + "loss": 0.5278, + "step": 93 + }, + { + "epoch": 0.3805668016194332, + "grad_norm": 2.0154418776442897, + "learning_rate": 3.032258064516129e-05, + "loss": 0.503, + "step": 94 + }, + { + "epoch": 0.38461538461538464, + "grad_norm": 1.7329099680388307, + "learning_rate": 3.0645161290322585e-05, + "loss": 0.4979, + "step": 95 + }, + { + "epoch": 0.38866396761133604, + "grad_norm": 1.7463584845253686, + "learning_rate": 3.096774193548387e-05, + "loss": 0.4822, + "step": 96 + }, + { + "epoch": 0.39271255060728744, + "grad_norm": 1.8653838475455666, + "learning_rate": 3.1290322580645166e-05, + "loss": 0.4951, + "step": 97 + }, + { + "epoch": 0.3967611336032389, + "grad_norm": 1.6536210899512287, + "learning_rate": 3.161290322580645e-05, + "loss": 0.4757, + "step": 98 + }, + { + "epoch": 0.4008097165991903, + "grad_norm": 1.4194729775505108, + "learning_rate": 3.193548387096774e-05, + "loss": 0.465, + "step": 99 + }, + { + "epoch": 0.4048582995951417, + "grad_norm": 1.54666867280196, + "learning_rate": 3.2258064516129034e-05, + "loss": 0.4692, + "step": 100 + }, + { + "epoch": 0.4089068825910931, + "grad_norm": 1.3393470296310128, + "learning_rate": 3.258064516129033e-05, + "loss": 0.4443, + "step": 101 + }, + { + "epoch": 0.41295546558704455, + "grad_norm": 1.6239835314720361, + "learning_rate": 3.2903225806451614e-05, + "loss": 0.4793, + "step": 102 + }, + { + "epoch": 0.41700404858299595, + "grad_norm": 1.2886539410625921, + "learning_rate": 3.322580645161291e-05, + "loss": 0.4749, + "step": 103 + }, + { + "epoch": 0.42105263157894735, + "grad_norm": 1.7787758585205053, + "learning_rate": 3.3548387096774195e-05, + "loss": 0.4657, + "step": 104 + }, + { + "epoch": 0.4251012145748988, + "grad_norm": 1.5556047892323346, + "learning_rate": 3.387096774193549e-05, + "loss": 0.4726, + "step": 105 + }, + { + "epoch": 0.4291497975708502, + "grad_norm": 1.5244821366888675, + "learning_rate": 3.4193548387096776e-05, + "loss": 0.4491, + "step": 106 + }, + { + "epoch": 0.4331983805668016, + "grad_norm": 1.5280788692491822, + "learning_rate": 3.451612903225807e-05, + "loss": 0.4569, + "step": 107 + }, + { + "epoch": 0.43724696356275305, + "grad_norm": 1.3568552700411742, + "learning_rate": 3.483870967741936e-05, + "loss": 0.4566, + "step": 108 + }, + { + "epoch": 0.44129554655870445, + "grad_norm": 1.388276373123974, + "learning_rate": 3.516129032258065e-05, + "loss": 0.4445, + "step": 109 + }, + { + "epoch": 0.44534412955465585, + "grad_norm": 1.490674994644851, + "learning_rate": 3.548387096774194e-05, + "loss": 0.4777, + "step": 110 + }, + { + "epoch": 0.4493927125506073, + "grad_norm": 1.240786498654692, + "learning_rate": 3.5806451612903225e-05, + "loss": 0.4241, + "step": 111 + }, + { + "epoch": 0.4534412955465587, + "grad_norm": 1.4129066662011576, + "learning_rate": 3.612903225806452e-05, + "loss": 0.4476, + "step": 112 + }, + { + "epoch": 0.4574898785425101, + "grad_norm": 1.143887299625655, + "learning_rate": 3.6451612903225805e-05, + "loss": 0.4336, + "step": 113 + }, + { + "epoch": 0.46153846153846156, + "grad_norm": 1.4968226364515826, + "learning_rate": 3.67741935483871e-05, + "loss": 0.4371, + "step": 114 + }, + { + "epoch": 0.46558704453441296, + "grad_norm": 0.9417910551845765, + "learning_rate": 3.7096774193548386e-05, + "loss": 0.438, + "step": 115 + }, + { + "epoch": 0.46963562753036436, + "grad_norm": 1.6577658178032788, + "learning_rate": 3.741935483870968e-05, + "loss": 0.4354, + "step": 116 + }, + { + "epoch": 0.47368421052631576, + "grad_norm": 1.1834015565870626, + "learning_rate": 3.7741935483870974e-05, + "loss": 0.4272, + "step": 117 + }, + { + "epoch": 0.4777327935222672, + "grad_norm": 1.6838352164216592, + "learning_rate": 3.806451612903226e-05, + "loss": 0.4411, + "step": 118 + }, + { + "epoch": 0.4817813765182186, + "grad_norm": 1.5590837981845118, + "learning_rate": 3.8387096774193555e-05, + "loss": 0.458, + "step": 119 + }, + { + "epoch": 0.48582995951417, + "grad_norm": 1.1196197351300552, + "learning_rate": 3.870967741935484e-05, + "loss": 0.4514, + "step": 120 + }, + { + "epoch": 0.4898785425101215, + "grad_norm": 1.2452353903277258, + "learning_rate": 3.9032258064516135e-05, + "loss": 0.4315, + "step": 121 + }, + { + "epoch": 0.4939271255060729, + "grad_norm": 0.94448571196041, + "learning_rate": 3.935483870967742e-05, + "loss": 0.4123, + "step": 122 + }, + { + "epoch": 0.4979757085020243, + "grad_norm": 1.6742376160491346, + "learning_rate": 3.9677419354838716e-05, + "loss": 0.4361, + "step": 123 + }, + { + "epoch": 0.5020242914979757, + "grad_norm": 1.0426887365389157, + "learning_rate": 4e-05, + "loss": 0.4523, + "step": 124 + }, + { + "epoch": 0.5060728744939271, + "grad_norm": 1.0777195358379013, + "learning_rate": 3.9999920040266474e-05, + "loss": 0.3959, + "step": 125 + }, + { + "epoch": 0.5101214574898786, + "grad_norm": 1.3089918983861526, + "learning_rate": 3.999968016170525e-05, + "loss": 0.4207, + "step": 126 + }, + { + "epoch": 0.5141700404858299, + "grad_norm": 1.1149996059524532, + "learning_rate": 3.9999280366234394e-05, + "loss": 0.4228, + "step": 127 + }, + { + "epoch": 0.5182186234817814, + "grad_norm": 1.2622116733081405, + "learning_rate": 3.999872065705065e-05, + "loss": 0.3942, + "step": 128 + }, + { + "epoch": 0.5222672064777328, + "grad_norm": 1.2960435353086937, + "learning_rate": 3.999800103862945e-05, + "loss": 0.4022, + "step": 129 + }, + { + "epoch": 0.5263157894736842, + "grad_norm": 1.1978461780547458, + "learning_rate": 3.9997121516724834e-05, + "loss": 0.3736, + "step": 130 + }, + { + "epoch": 0.5303643724696356, + "grad_norm": 1.3744436030697438, + "learning_rate": 3.999608209836944e-05, + "loss": 0.4222, + "step": 131 + }, + { + "epoch": 0.5344129554655871, + "grad_norm": 1.1054879096106962, + "learning_rate": 3.999488279187443e-05, + "loss": 0.4046, + "step": 132 + }, + { + "epoch": 0.5384615384615384, + "grad_norm": 0.9850849106996765, + "learning_rate": 3.9993523606829425e-05, + "loss": 0.4068, + "step": 133 + }, + { + "epoch": 0.5425101214574899, + "grad_norm": 1.032845985492106, + "learning_rate": 3.9992004554102435e-05, + "loss": 0.3998, + "step": 134 + }, + { + "epoch": 0.5465587044534413, + "grad_norm": 1.2890590809795395, + "learning_rate": 3.999032564583976e-05, + "loss": 0.3854, + "step": 135 + }, + { + "epoch": 0.5506072874493927, + "grad_norm": 1.0877344020485236, + "learning_rate": 3.998848689546591e-05, + "loss": 0.4123, + "step": 136 + }, + { + "epoch": 0.5546558704453441, + "grad_norm": 1.4482692874762881, + "learning_rate": 3.9986488317683496e-05, + "loss": 0.3966, + "step": 137 + }, + { + "epoch": 0.5587044534412956, + "grad_norm": 1.132180046399313, + "learning_rate": 3.9984329928473067e-05, + "loss": 0.3466, + "step": 138 + }, + { + "epoch": 0.562753036437247, + "grad_norm": 1.5413011425954193, + "learning_rate": 3.998201174509306e-05, + "loss": 0.4012, + "step": 139 + }, + { + "epoch": 0.5668016194331984, + "grad_norm": 1.219252806615975, + "learning_rate": 3.997953378607961e-05, + "loss": 0.3895, + "step": 140 + }, + { + "epoch": 0.5708502024291497, + "grad_norm": 1.1932067796369097, + "learning_rate": 3.9976896071246403e-05, + "loss": 0.3619, + "step": 141 + }, + { + "epoch": 0.5748987854251012, + "grad_norm": 1.261628181575705, + "learning_rate": 3.997409862168454e-05, + "loss": 0.3742, + "step": 142 + }, + { + "epoch": 0.5789473684210527, + "grad_norm": 1.340984306587018, + "learning_rate": 3.997114145976237e-05, + "loss": 0.3607, + "step": 143 + }, + { + "epoch": 0.582995951417004, + "grad_norm": 1.0562003206668498, + "learning_rate": 3.9968024609125245e-05, + "loss": 0.3455, + "step": 144 + }, + { + "epoch": 0.5870445344129555, + "grad_norm": 1.5830820781527875, + "learning_rate": 3.996474809469545e-05, + "loss": 0.3807, + "step": 145 + }, + { + "epoch": 0.5910931174089069, + "grad_norm": 1.0745761110334133, + "learning_rate": 3.996131194267189e-05, + "loss": 0.3785, + "step": 146 + }, + { + "epoch": 0.5951417004048583, + "grad_norm": 1.366934058606418, + "learning_rate": 3.995771618052995e-05, + "loss": 0.3503, + "step": 147 + }, + { + "epoch": 0.5991902834008097, + "grad_norm": 0.9898312784448842, + "learning_rate": 3.9953960837021256e-05, + "loss": 0.3412, + "step": 148 + }, + { + "epoch": 0.6032388663967612, + "grad_norm": 1.5446948029590657, + "learning_rate": 3.995004594217343e-05, + "loss": 0.3474, + "step": 149 + }, + { + "epoch": 0.6072874493927125, + "grad_norm": 1.283507833851253, + "learning_rate": 3.994597152728987e-05, + "loss": 0.365, + "step": 150 + }, + { + "epoch": 0.611336032388664, + "grad_norm": 1.1804639966580295, + "learning_rate": 3.994173762494947e-05, + "loss": 0.3545, + "step": 151 + }, + { + "epoch": 0.6153846153846154, + "grad_norm": 1.126810584136089, + "learning_rate": 3.9937344269006424e-05, + "loss": 0.3223, + "step": 152 + }, + { + "epoch": 0.6194331983805668, + "grad_norm": 1.018029627254414, + "learning_rate": 3.993279149458988e-05, + "loss": 0.3184, + "step": 153 + }, + { + "epoch": 0.6234817813765182, + "grad_norm": 0.9675380041302128, + "learning_rate": 3.9928079338103695e-05, + "loss": 0.3163, + "step": 154 + }, + { + "epoch": 0.6275303643724697, + "grad_norm": 1.0578023056518158, + "learning_rate": 3.992320783722616e-05, + "loss": 0.3294, + "step": 155 + }, + { + "epoch": 0.631578947368421, + "grad_norm": 0.9130057269236581, + "learning_rate": 3.991817703090965e-05, + "loss": 0.3196, + "step": 156 + }, + { + "epoch": 0.6356275303643725, + "grad_norm": 0.8924266073982151, + "learning_rate": 3.991298695938038e-05, + "loss": 0.3274, + "step": 157 + }, + { + "epoch": 0.6396761133603239, + "grad_norm": 0.8262338681269364, + "learning_rate": 3.9907637664138e-05, + "loss": 0.3533, + "step": 158 + }, + { + "epoch": 0.6437246963562753, + "grad_norm": 1.047840419893635, + "learning_rate": 3.990212918795534e-05, + "loss": 0.3029, + "step": 159 + }, + { + "epoch": 0.6477732793522267, + "grad_norm": 0.8513277996700815, + "learning_rate": 3.9896461574878036e-05, + "loss": 0.3001, + "step": 160 + }, + { + "epoch": 0.6518218623481782, + "grad_norm": 0.9221396089420381, + "learning_rate": 3.989063487022416e-05, + "loss": 0.3023, + "step": 161 + }, + { + "epoch": 0.6558704453441295, + "grad_norm": 0.9018209279762236, + "learning_rate": 3.9884649120583903e-05, + "loss": 0.3123, + "step": 162 + }, + { + "epoch": 0.659919028340081, + "grad_norm": 0.7940468902572237, + "learning_rate": 3.987850437381915e-05, + "loss": 0.2891, + "step": 163 + }, + { + "epoch": 0.6639676113360324, + "grad_norm": 0.9946332671672933, + "learning_rate": 3.987220067906313e-05, + "loss": 0.3332, + "step": 164 + }, + { + "epoch": 0.6680161943319838, + "grad_norm": 1.1520986266098894, + "learning_rate": 3.986573808672002e-05, + "loss": 0.3062, + "step": 165 + }, + { + "epoch": 0.6720647773279352, + "grad_norm": 0.8805269550986113, + "learning_rate": 3.9859116648464534e-05, + "loss": 0.2762, + "step": 166 + }, + { + "epoch": 0.6761133603238867, + "grad_norm": 1.0457277030894179, + "learning_rate": 3.985233641724152e-05, + "loss": 0.3029, + "step": 167 + }, + { + "epoch": 0.680161943319838, + "grad_norm": 0.8925960668734015, + "learning_rate": 3.984539744726554e-05, + "loss": 0.293, + "step": 168 + }, + { + "epoch": 0.6842105263157895, + "grad_norm": 0.9174138455608537, + "learning_rate": 3.983829979402038e-05, + "loss": 0.3287, + "step": 169 + }, + { + "epoch": 0.6882591093117408, + "grad_norm": 0.8639801705715331, + "learning_rate": 3.983104351425871e-05, + "loss": 0.2949, + "step": 170 + }, + { + "epoch": 0.6923076923076923, + "grad_norm": 1.1943765479310793, + "learning_rate": 3.982362866600154e-05, + "loss": 0.2804, + "step": 171 + }, + { + "epoch": 0.6963562753036437, + "grad_norm": 0.8998213404807535, + "learning_rate": 3.9816055308537806e-05, + "loss": 0.2874, + "step": 172 + }, + { + "epoch": 0.7004048582995951, + "grad_norm": 0.9018947480841283, + "learning_rate": 3.980832350242388e-05, + "loss": 0.2709, + "step": 173 + }, + { + "epoch": 0.7044534412955465, + "grad_norm": 0.9452545705543409, + "learning_rate": 3.980043330948306e-05, + "loss": 0.2662, + "step": 174 + }, + { + "epoch": 0.708502024291498, + "grad_norm": 0.894692120597729, + "learning_rate": 3.9792384792805124e-05, + "loss": 0.2713, + "step": 175 + }, + { + "epoch": 0.7125506072874493, + "grad_norm": 0.6735835784264312, + "learning_rate": 3.978417801674579e-05, + "loss": 0.2726, + "step": 176 + }, + { + "epoch": 0.7165991902834008, + "grad_norm": 0.8134747105771756, + "learning_rate": 3.9775813046926244e-05, + "loss": 0.2779, + "step": 177 + }, + { + "epoch": 0.7206477732793523, + "grad_norm": 0.8406873420963594, + "learning_rate": 3.976728995023254e-05, + "loss": 0.2619, + "step": 178 + }, + { + "epoch": 0.7246963562753036, + "grad_norm": 0.7900539470852405, + "learning_rate": 3.975860879481514e-05, + "loss": 0.2546, + "step": 179 + }, + { + "epoch": 0.728744939271255, + "grad_norm": 0.7626627378868537, + "learning_rate": 3.974976965008833e-05, + "loss": 0.2712, + "step": 180 + }, + { + "epoch": 0.7327935222672065, + "grad_norm": 0.7964952294394693, + "learning_rate": 3.9740772586729674e-05, + "loss": 0.2518, + "step": 181 + }, + { + "epoch": 0.7368421052631579, + "grad_norm": 0.9357412055647546, + "learning_rate": 3.973161767667946e-05, + "loss": 0.2647, + "step": 182 + }, + { + "epoch": 0.7408906882591093, + "grad_norm": 0.660123322338945, + "learning_rate": 3.972230499314009e-05, + "loss": 0.2547, + "step": 183 + }, + { + "epoch": 0.7449392712550608, + "grad_norm": 0.8563686265027166, + "learning_rate": 3.9712834610575545e-05, + "loss": 0.2526, + "step": 184 + }, + { + "epoch": 0.7489878542510121, + "grad_norm": 0.8380408494724138, + "learning_rate": 3.9703206604710746e-05, + "loss": 0.2671, + "step": 185 + }, + { + "epoch": 0.7530364372469636, + "grad_norm": 0.6095317311728123, + "learning_rate": 3.969342105253098e-05, + "loss": 0.2449, + "step": 186 + }, + { + "epoch": 0.757085020242915, + "grad_norm": 0.7784860451431362, + "learning_rate": 3.9683478032281246e-05, + "loss": 0.2304, + "step": 187 + }, + { + "epoch": 0.7611336032388664, + "grad_norm": 0.8255884733951765, + "learning_rate": 3.967337762346568e-05, + "loss": 0.2351, + "step": 188 + }, + { + "epoch": 0.7651821862348178, + "grad_norm": 0.7119100341143461, + "learning_rate": 3.9663119906846885e-05, + "loss": 0.2587, + "step": 189 + }, + { + "epoch": 0.7692307692307693, + "grad_norm": 0.7181531055849719, + "learning_rate": 3.965270496444528e-05, + "loss": 0.2061, + "step": 190 + }, + { + "epoch": 0.7732793522267206, + "grad_norm": 0.6950950232145006, + "learning_rate": 3.9642132879538465e-05, + "loss": 0.2598, + "step": 191 + }, + { + "epoch": 0.7773279352226721, + "grad_norm": 0.6741038213707056, + "learning_rate": 3.9631403736660566e-05, + "loss": 0.2488, + "step": 192 + }, + { + "epoch": 0.7813765182186235, + "grad_norm": 0.7950816780920087, + "learning_rate": 3.962051762160151e-05, + "loss": 0.2191, + "step": 193 + }, + { + "epoch": 0.7854251012145749, + "grad_norm": 0.6456291310025344, + "learning_rate": 3.960947462140638e-05, + "loss": 0.2078, + "step": 194 + }, + { + "epoch": 0.7894736842105263, + "grad_norm": 0.6491906422503163, + "learning_rate": 3.959827482437472e-05, + "loss": 0.239, + "step": 195 + }, + { + "epoch": 0.7935222672064778, + "grad_norm": 0.7248899469001897, + "learning_rate": 3.958691832005981e-05, + "loss": 0.2362, + "step": 196 + }, + { + "epoch": 0.7975708502024291, + "grad_norm": 0.7761575526543828, + "learning_rate": 3.9575405199267944e-05, + "loss": 0.2082, + "step": 197 + }, + { + "epoch": 0.8016194331983806, + "grad_norm": 0.7296959036621308, + "learning_rate": 3.956373555405774e-05, + "loss": 0.2013, + "step": 198 + }, + { + "epoch": 0.805668016194332, + "grad_norm": 0.7584639466430377, + "learning_rate": 3.955190947773937e-05, + "loss": 0.2296, + "step": 199 + }, + { + "epoch": 0.8097165991902834, + "grad_norm": 0.6816241815891586, + "learning_rate": 3.9539927064873824e-05, + "loss": 0.2193, + "step": 200 + }, + { + "epoch": 0.8137651821862348, + "grad_norm": 0.7200167298264689, + "learning_rate": 3.952778841127214e-05, + "loss": 0.2093, + "step": 201 + }, + { + "epoch": 0.8178137651821862, + "grad_norm": 0.7475271823765884, + "learning_rate": 3.95154936139947e-05, + "loss": 0.213, + "step": 202 + }, + { + "epoch": 0.8218623481781376, + "grad_norm": 0.6484840066825885, + "learning_rate": 3.950304277135034e-05, + "loss": 0.2235, + "step": 203 + }, + { + "epoch": 0.8259109311740891, + "grad_norm": 0.6658175256790314, + "learning_rate": 3.94904359828957e-05, + "loss": 0.1974, + "step": 204 + }, + { + "epoch": 0.8299595141700404, + "grad_norm": 0.6734496201735305, + "learning_rate": 3.94776733494343e-05, + "loss": 0.189, + "step": 205 + }, + { + "epoch": 0.8340080971659919, + "grad_norm": 0.6954778538631582, + "learning_rate": 3.9464754973015826e-05, + "loss": 0.2126, + "step": 206 + }, + { + "epoch": 0.8380566801619433, + "grad_norm": 0.719819050053502, + "learning_rate": 3.9451680956935275e-05, + "loss": 0.2108, + "step": 207 + }, + { + "epoch": 0.8421052631578947, + "grad_norm": 0.6388627609564798, + "learning_rate": 3.943845140573213e-05, + "loss": 0.1962, + "step": 208 + }, + { + "epoch": 0.8461538461538461, + "grad_norm": 0.6115626972679965, + "learning_rate": 3.942506642518952e-05, + "loss": 0.2035, + "step": 209 + }, + { + "epoch": 0.8502024291497976, + "grad_norm": 0.5804738320342732, + "learning_rate": 3.9411526122333396e-05, + "loss": 0.1715, + "step": 210 + }, + { + "epoch": 0.854251012145749, + "grad_norm": 0.5845797144238533, + "learning_rate": 3.939783060543167e-05, + "loss": 0.175, + "step": 211 + }, + { + "epoch": 0.8582995951417004, + "grad_norm": 0.598841517493047, + "learning_rate": 3.938397998399333e-05, + "loss": 0.179, + "step": 212 + }, + { + "epoch": 0.8623481781376519, + "grad_norm": 0.5722263635422322, + "learning_rate": 3.936997436876756e-05, + "loss": 0.1563, + "step": 213 + }, + { + "epoch": 0.8663967611336032, + "grad_norm": 0.6033826956217871, + "learning_rate": 3.935581387174289e-05, + "loss": 0.1819, + "step": 214 + }, + { + "epoch": 0.8704453441295547, + "grad_norm": 0.566741018641793, + "learning_rate": 3.9341498606146295e-05, + "loss": 0.1644, + "step": 215 + }, + { + "epoch": 0.8744939271255061, + "grad_norm": 0.5470950453271776, + "learning_rate": 3.932702868644224e-05, + "loss": 0.1597, + "step": 216 + }, + { + "epoch": 0.8785425101214575, + "grad_norm": 0.6227376918812997, + "learning_rate": 3.931240422833181e-05, + "loss": 0.1787, + "step": 217 + }, + { + "epoch": 0.8825910931174089, + "grad_norm": 0.6023485142403455, + "learning_rate": 3.9297625348751806e-05, + "loss": 0.155, + "step": 218 + }, + { + "epoch": 0.8866396761133604, + "grad_norm": 0.6017983047195229, + "learning_rate": 3.9282692165873744e-05, + "loss": 0.1619, + "step": 219 + }, + { + "epoch": 0.8906882591093117, + "grad_norm": 0.5688980065831615, + "learning_rate": 3.9267604799102945e-05, + "loss": 0.1681, + "step": 220 + }, + { + "epoch": 0.8947368421052632, + "grad_norm": 0.5270725531575751, + "learning_rate": 3.9252363369077615e-05, + "loss": 0.1495, + "step": 221 + }, + { + "epoch": 0.8987854251012146, + "grad_norm": 0.7207423539242678, + "learning_rate": 3.9236967997667796e-05, + "loss": 0.1884, + "step": 222 + }, + { + "epoch": 0.902834008097166, + "grad_norm": 0.6616331472549493, + "learning_rate": 3.9221418807974496e-05, + "loss": 0.1449, + "step": 223 + }, + { + "epoch": 0.9068825910931174, + "grad_norm": 0.5204053432842333, + "learning_rate": 3.92057159243286e-05, + "loss": 0.145, + "step": 224 + }, + { + "epoch": 0.9109311740890689, + "grad_norm": 0.8268261511960855, + "learning_rate": 3.9189859472289956e-05, + "loss": 0.1519, + "step": 225 + }, + { + "epoch": 0.9149797570850202, + "grad_norm": 0.6153982984916607, + "learning_rate": 3.9173849578646324e-05, + "loss": 0.1398, + "step": 226 + }, + { + "epoch": 0.9190283400809717, + "grad_norm": 0.6149662103774857, + "learning_rate": 3.9157686371412395e-05, + "loss": 0.1477, + "step": 227 + }, + { + "epoch": 0.9230769230769231, + "grad_norm": 0.6690191140979355, + "learning_rate": 3.914136997982874e-05, + "loss": 0.1318, + "step": 228 + }, + { + "epoch": 0.9271255060728745, + "grad_norm": 0.5729910382568165, + "learning_rate": 3.912490053436079e-05, + "loss": 0.1275, + "step": 229 + }, + { + "epoch": 0.9311740890688259, + "grad_norm": 0.5689208532003758, + "learning_rate": 3.9108278166697805e-05, + "loss": 0.1412, + "step": 230 + }, + { + "epoch": 0.9352226720647774, + "grad_norm": 0.6072170300834397, + "learning_rate": 3.909150300975178e-05, + "loss": 0.1389, + "step": 231 + }, + { + "epoch": 0.9392712550607287, + "grad_norm": 0.5913037127311208, + "learning_rate": 3.907457519765642e-05, + "loss": 0.15, + "step": 232 + }, + { + "epoch": 0.9433198380566802, + "grad_norm": 0.6033199309084496, + "learning_rate": 3.905749486576607e-05, + "loss": 0.1396, + "step": 233 + }, + { + "epoch": 0.9473684210526315, + "grad_norm": 0.4821009986173103, + "learning_rate": 3.90402621506546e-05, + "loss": 0.1218, + "step": 234 + }, + { + "epoch": 0.951417004048583, + "grad_norm": 0.5294417531421356, + "learning_rate": 3.902287719011435e-05, + "loss": 0.1219, + "step": 235 + }, + { + "epoch": 0.9554655870445344, + "grad_norm": 0.5514573401848548, + "learning_rate": 3.900534012315499e-05, + "loss": 0.1186, + "step": 236 + }, + { + "epoch": 0.9595141700404858, + "grad_norm": 0.46266604525161753, + "learning_rate": 3.8987651090002454e-05, + "loss": 0.1112, + "step": 237 + }, + { + "epoch": 0.9635627530364372, + "grad_norm": 0.5532512437422721, + "learning_rate": 3.896981023209777e-05, + "loss": 0.1324, + "step": 238 + }, + { + "epoch": 0.9676113360323887, + "grad_norm": 0.4803502732631606, + "learning_rate": 3.8951817692095966e-05, + "loss": 0.114, + "step": 239 + }, + { + "epoch": 0.97165991902834, + "grad_norm": 0.5371927057669014, + "learning_rate": 3.893367361386491e-05, + "loss": 0.1195, + "step": 240 + }, + { + "epoch": 0.9757085020242915, + "grad_norm": 0.4745810970444339, + "learning_rate": 3.891537814248417e-05, + "loss": 0.109, + "step": 241 + }, + { + "epoch": 0.979757085020243, + "grad_norm": 0.5484521676580363, + "learning_rate": 3.889693142424384e-05, + "loss": 0.1104, + "step": 242 + }, + { + "epoch": 0.9838056680161943, + "grad_norm": 0.5365519470837069, + "learning_rate": 3.8878333606643405e-05, + "loss": 0.098, + "step": 243 + }, + { + "epoch": 0.9878542510121457, + "grad_norm": 0.6122472720141737, + "learning_rate": 3.885958483839049e-05, + "loss": 0.1211, + "step": 244 + }, + { + "epoch": 0.9919028340080972, + "grad_norm": 0.4824947958444269, + "learning_rate": 3.8840685269399786e-05, + "loss": 0.1039, + "step": 245 + }, + { + "epoch": 0.9959514170040485, + "grad_norm": 0.5068802532648521, + "learning_rate": 3.882163505079171e-05, + "loss": 0.114, + "step": 246 + }, + { + "epoch": 1.0, + "grad_norm": 0.5270597008946245, + "learning_rate": 3.880243433489132e-05, + "loss": 0.1047, + "step": 247 + }, + { + "epoch": 1.0040485829959513, + "grad_norm": 0.476218960998343, + "learning_rate": 3.878308327522702e-05, + "loss": 0.0775, + "step": 248 + }, + { + "epoch": 1.008097165991903, + "grad_norm": 0.4635670333762413, + "learning_rate": 3.876358202652936e-05, + "loss": 0.0847, + "step": 249 + }, + { + "epoch": 1.0121457489878543, + "grad_norm": 0.44421574050124624, + "learning_rate": 3.8743930744729835e-05, + "loss": 0.0826, + "step": 250 + }, + { + "epoch": 1.0161943319838056, + "grad_norm": 0.43502200910593314, + "learning_rate": 3.8724129586959535e-05, + "loss": 0.0785, + "step": 251 + }, + { + "epoch": 1.0202429149797572, + "grad_norm": 0.44677475187312965, + "learning_rate": 3.870417871154801e-05, + "loss": 0.0735, + "step": 252 + }, + { + "epoch": 1.0242914979757085, + "grad_norm": 0.46020475626887547, + "learning_rate": 3.868407827802192e-05, + "loss": 0.0722, + "step": 253 + }, + { + "epoch": 1.0283400809716599, + "grad_norm": 0.4529055739452721, + "learning_rate": 3.8663828447103804e-05, + "loss": 0.08, + "step": 254 + }, + { + "epoch": 1.0323886639676114, + "grad_norm": 0.41938269491109, + "learning_rate": 3.8643429380710765e-05, + "loss": 0.0741, + "step": 255 + }, + { + "epoch": 1.0364372469635628, + "grad_norm": 0.4098558159092878, + "learning_rate": 3.862288124195319e-05, + "loss": 0.0535, + "step": 256 + }, + { + "epoch": 1.040485829959514, + "grad_norm": 0.4795598466629109, + "learning_rate": 3.8602184195133464e-05, + "loss": 0.0884, + "step": 257 + }, + { + "epoch": 1.0445344129554657, + "grad_norm": 0.43998401796671394, + "learning_rate": 3.8581338405744604e-05, + "loss": 0.0767, + "step": 258 + }, + { + "epoch": 1.048582995951417, + "grad_norm": 0.44009925916500453, + "learning_rate": 3.8560344040468996e-05, + "loss": 0.078, + "step": 259 + }, + { + "epoch": 1.0526315789473684, + "grad_norm": 0.41199630036532925, + "learning_rate": 3.853920126717703e-05, + "loss": 0.0784, + "step": 260 + }, + { + "epoch": 1.05668016194332, + "grad_norm": 0.4199930359956907, + "learning_rate": 3.851791025492574e-05, + "loss": 0.077, + "step": 261 + }, + { + "epoch": 1.0607287449392713, + "grad_norm": 0.4191267106437191, + "learning_rate": 3.8496471173957515e-05, + "loss": 0.0725, + "step": 262 + }, + { + "epoch": 1.0647773279352226, + "grad_norm": 0.399198325703814, + "learning_rate": 3.847488419569866e-05, + "loss": 0.078, + "step": 263 + }, + { + "epoch": 1.0688259109311742, + "grad_norm": 0.37498634807325326, + "learning_rate": 3.8453149492758077e-05, + "loss": 0.0686, + "step": 264 + }, + { + "epoch": 1.0728744939271255, + "grad_norm": 0.3875910761243164, + "learning_rate": 3.8431267238925886e-05, + "loss": 0.062, + "step": 265 + }, + { + "epoch": 1.0769230769230769, + "grad_norm": 0.40568956484987145, + "learning_rate": 3.840923760917199e-05, + "loss": 0.0748, + "step": 266 + }, + { + "epoch": 1.0809716599190284, + "grad_norm": 0.3821698926132872, + "learning_rate": 3.838706077964473e-05, + "loss": 0.0632, + "step": 267 + }, + { + "epoch": 1.0850202429149798, + "grad_norm": 0.35435649248031503, + "learning_rate": 3.836473692766944e-05, + "loss": 0.0592, + "step": 268 + }, + { + "epoch": 1.0890688259109311, + "grad_norm": 0.39288871149101, + "learning_rate": 3.834226623174705e-05, + "loss": 0.0699, + "step": 269 + }, + { + "epoch": 1.0931174089068827, + "grad_norm": 0.38452216963507807, + "learning_rate": 3.831964887155264e-05, + "loss": 0.0633, + "step": 270 + }, + { + "epoch": 1.097165991902834, + "grad_norm": 0.3796914452907749, + "learning_rate": 3.829688502793403e-05, + "loss": 0.0552, + "step": 271 + }, + { + "epoch": 1.1012145748987854, + "grad_norm": 0.36457420240250593, + "learning_rate": 3.827397488291029e-05, + "loss": 0.0587, + "step": 272 + }, + { + "epoch": 1.1052631578947367, + "grad_norm": 0.40678542850084876, + "learning_rate": 3.8250918619670343e-05, + "loss": 0.0656, + "step": 273 + }, + { + "epoch": 1.1093117408906883, + "grad_norm": 0.36946764180685904, + "learning_rate": 3.822771642257145e-05, + "loss": 0.0611, + "step": 274 + }, + { + "epoch": 1.1133603238866396, + "grad_norm": 0.35617918906704, + "learning_rate": 3.820436847713776e-05, + "loss": 0.0611, + "step": 275 + }, + { + "epoch": 1.117408906882591, + "grad_norm": 0.3658218076634524, + "learning_rate": 3.8180874970058826e-05, + "loss": 0.0638, + "step": 276 + }, + { + "epoch": 1.1214574898785425, + "grad_norm": 0.37508770683277715, + "learning_rate": 3.8157236089188103e-05, + "loss": 0.0661, + "step": 277 + }, + { + "epoch": 1.125506072874494, + "grad_norm": 0.3755639953440946, + "learning_rate": 3.813345202354145e-05, + "loss": 0.0571, + "step": 278 + }, + { + "epoch": 1.1295546558704452, + "grad_norm": 0.3415377687904576, + "learning_rate": 3.810952296329563e-05, + "loss": 0.0497, + "step": 279 + }, + { + "epoch": 1.1336032388663968, + "grad_norm": 0.3965176921528856, + "learning_rate": 3.808544909978676e-05, + "loss": 0.0651, + "step": 280 + }, + { + "epoch": 1.1376518218623481, + "grad_norm": 0.3518936076388079, + "learning_rate": 3.806123062550882e-05, + "loss": 0.0577, + "step": 281 + }, + { + "epoch": 1.1417004048582995, + "grad_norm": 0.35740728777272746, + "learning_rate": 3.803686773411208e-05, + "loss": 0.0526, + "step": 282 + }, + { + "epoch": 1.145748987854251, + "grad_norm": 0.3585843868617206, + "learning_rate": 3.801236062040158e-05, + "loss": 0.0564, + "step": 283 + }, + { + "epoch": 1.1497975708502024, + "grad_norm": 0.375936912732562, + "learning_rate": 3.798770948033553e-05, + "loss": 0.0621, + "step": 284 + }, + { + "epoch": 1.1538461538461537, + "grad_norm": 0.3257825386249052, + "learning_rate": 3.796291451102381e-05, + "loss": 0.0425, + "step": 285 + }, + { + "epoch": 1.1578947368421053, + "grad_norm": 0.38640724599702897, + "learning_rate": 3.793797591072631e-05, + "loss": 0.0456, + "step": 286 + }, + { + "epoch": 1.1619433198380567, + "grad_norm": 0.3493073358148721, + "learning_rate": 3.791289387885144e-05, + "loss": 0.0423, + "step": 287 + }, + { + "epoch": 1.165991902834008, + "grad_norm": 0.37409289567699033, + "learning_rate": 3.788766861595443e-05, + "loss": 0.0616, + "step": 288 + }, + { + "epoch": 1.1700404858299596, + "grad_norm": 0.32194846534391564, + "learning_rate": 3.7862300323735835e-05, + "loss": 0.0376, + "step": 289 + }, + { + "epoch": 1.174089068825911, + "grad_norm": 0.3603590580176283, + "learning_rate": 3.783678920503983e-05, + "loss": 0.049, + "step": 290 + }, + { + "epoch": 1.1781376518218623, + "grad_norm": 0.32410599639398524, + "learning_rate": 3.781113546385265e-05, + "loss": 0.0517, + "step": 291 + }, + { + "epoch": 1.1821862348178138, + "grad_norm": 0.3321888144430527, + "learning_rate": 3.7785339305300916e-05, + "loss": 0.0433, + "step": 292 + }, + { + "epoch": 1.1862348178137652, + "grad_norm": 0.34189424140938707, + "learning_rate": 3.7759400935650025e-05, + "loss": 0.0511, + "step": 293 + }, + { + "epoch": 1.1902834008097165, + "grad_norm": 0.33128721280553103, + "learning_rate": 3.77333205623025e-05, + "loss": 0.0355, + "step": 294 + }, + { + "epoch": 1.194331983805668, + "grad_norm": 0.3131005761233874, + "learning_rate": 3.7707098393796295e-05, + "loss": 0.042, + "step": 295 + }, + { + "epoch": 1.1983805668016194, + "grad_norm": 0.36274083435512566, + "learning_rate": 3.7680734639803185e-05, + "loss": 0.0446, + "step": 296 + }, + { + "epoch": 1.2024291497975708, + "grad_norm": 0.3287616534910013, + "learning_rate": 3.765422951112704e-05, + "loss": 0.0489, + "step": 297 + }, + { + "epoch": 1.2064777327935223, + "grad_norm": 0.3631601214427436, + "learning_rate": 3.762758321970216e-05, + "loss": 0.04, + "step": 298 + }, + { + "epoch": 1.2105263157894737, + "grad_norm": 0.35899481384020737, + "learning_rate": 3.7600795978591584e-05, + "loss": 0.0444, + "step": 299 + }, + { + "epoch": 1.214574898785425, + "grad_norm": 0.3377398161011398, + "learning_rate": 3.757386800198538e-05, + "loss": 0.0493, + "step": 300 + }, + { + "epoch": 1.2186234817813766, + "grad_norm": 0.36231356881486304, + "learning_rate": 3.7546799505198925e-05, + "loss": 0.0433, + "step": 301 + }, + { + "epoch": 1.222672064777328, + "grad_norm": 0.3212818701835057, + "learning_rate": 3.7519590704671197e-05, + "loss": 0.0431, + "step": 302 + }, + { + "epoch": 1.2267206477732793, + "grad_norm": 0.3893654863484072, + "learning_rate": 3.749224181796305e-05, + "loss": 0.0412, + "step": 303 + }, + { + "epoch": 1.2307692307692308, + "grad_norm": 0.37320766904716185, + "learning_rate": 3.746475306375545e-05, + "loss": 0.0485, + "step": 304 + }, + { + "epoch": 1.2348178137651822, + "grad_norm": 0.3394365128517062, + "learning_rate": 3.743712466184774e-05, + "loss": 0.0343, + "step": 305 + }, + { + "epoch": 1.2388663967611335, + "grad_norm": 0.34112522715428906, + "learning_rate": 3.7409356833155885e-05, + "loss": 0.0455, + "step": 306 + }, + { + "epoch": 1.242914979757085, + "grad_norm": 0.34248044495328833, + "learning_rate": 3.73814497997107e-05, + "loss": 0.0363, + "step": 307 + }, + { + "epoch": 1.2469635627530364, + "grad_norm": 0.33070199802138994, + "learning_rate": 3.7353403784656095e-05, + "loss": 0.0392, + "step": 308 + }, + { + "epoch": 1.2510121457489878, + "grad_norm": 0.35879601061019856, + "learning_rate": 3.732521901224724e-05, + "loss": 0.037, + "step": 309 + }, + { + "epoch": 1.2550607287449393, + "grad_norm": 0.34462052849904334, + "learning_rate": 3.7296895707848844e-05, + "loss": 0.0396, + "step": 310 + }, + { + "epoch": 1.2591093117408907, + "grad_norm": 0.2979366238214731, + "learning_rate": 3.7268434097933275e-05, + "loss": 0.0332, + "step": 311 + }, + { + "epoch": 1.263157894736842, + "grad_norm": 0.3025840975196208, + "learning_rate": 3.723983441007882e-05, + "loss": 0.0296, + "step": 312 + }, + { + "epoch": 1.2672064777327936, + "grad_norm": 0.3115505886714221, + "learning_rate": 3.721109687296781e-05, + "loss": 0.0371, + "step": 313 + }, + { + "epoch": 1.271255060728745, + "grad_norm": 0.2954674987963979, + "learning_rate": 3.718222171638484e-05, + "loss": 0.0349, + "step": 314 + }, + { + "epoch": 1.2753036437246963, + "grad_norm": 0.3352876029136561, + "learning_rate": 3.715320917121488e-05, + "loss": 0.0424, + "step": 315 + }, + { + "epoch": 1.2793522267206479, + "grad_norm": 0.31994502937436675, + "learning_rate": 3.712405946944149e-05, + "loss": 0.036, + "step": 316 + }, + { + "epoch": 1.2834008097165992, + "grad_norm": 0.2879553628662006, + "learning_rate": 3.709477284414488e-05, + "loss": 0.0289, + "step": 317 + }, + { + "epoch": 1.2874493927125505, + "grad_norm": 0.29244705126857684, + "learning_rate": 3.7065349529500136e-05, + "loss": 0.0309, + "step": 318 + }, + { + "epoch": 1.291497975708502, + "grad_norm": 0.2821652791258732, + "learning_rate": 3.7035789760775305e-05, + "loss": 0.0327, + "step": 319 + }, + { + "epoch": 1.2955465587044535, + "grad_norm": 0.2968203601690023, + "learning_rate": 3.70060937743295e-05, + "loss": 0.0314, + "step": 320 + }, + { + "epoch": 1.2995951417004048, + "grad_norm": 0.2827338561793179, + "learning_rate": 3.6976261807611045e-05, + "loss": 0.032, + "step": 321 + }, + { + "epoch": 1.3036437246963564, + "grad_norm": 0.30669762398726863, + "learning_rate": 3.694629409915555e-05, + "loss": 0.0322, + "step": 322 + }, + { + "epoch": 1.3076923076923077, + "grad_norm": 0.2871714184713645, + "learning_rate": 3.691619088858401e-05, + "loss": 0.0287, + "step": 323 + }, + { + "epoch": 1.311740890688259, + "grad_norm": 0.2755236498187533, + "learning_rate": 3.68859524166009e-05, + "loss": 0.0262, + "step": 324 + }, + { + "epoch": 1.3157894736842106, + "grad_norm": 0.28078638000717704, + "learning_rate": 3.6855578924992226e-05, + "loss": 0.0257, + "step": 325 + }, + { + "epoch": 1.319838056680162, + "grad_norm": 0.29812775030304123, + "learning_rate": 3.6825070656623626e-05, + "loss": 0.0319, + "step": 326 + }, + { + "epoch": 1.3238866396761133, + "grad_norm": 0.2866723345492833, + "learning_rate": 3.67944278554384e-05, + "loss": 0.0326, + "step": 327 + }, + { + "epoch": 1.3279352226720649, + "grad_norm": 0.3132654084900415, + "learning_rate": 3.676365076645557e-05, + "loss": 0.0369, + "step": 328 + }, + { + "epoch": 1.3319838056680162, + "grad_norm": 0.26884619293297063, + "learning_rate": 3.673273963576791e-05, + "loss": 0.0252, + "step": 329 + }, + { + "epoch": 1.3360323886639676, + "grad_norm": 0.2753083218897676, + "learning_rate": 3.670169471054001e-05, + "loss": 0.0268, + "step": 330 + }, + { + "epoch": 1.3400809716599191, + "grad_norm": 0.27052626367187127, + "learning_rate": 3.6670516239006254e-05, + "loss": 0.0251, + "step": 331 + }, + { + "epoch": 1.3441295546558705, + "grad_norm": 0.28334753087970044, + "learning_rate": 3.663920447046888e-05, + "loss": 0.0326, + "step": 332 + }, + { + "epoch": 1.3481781376518218, + "grad_norm": 0.2750576456086292, + "learning_rate": 3.6607759655295954e-05, + "loss": 0.0261, + "step": 333 + }, + { + "epoch": 1.3522267206477734, + "grad_norm": 0.2942038153279117, + "learning_rate": 3.657618204491937e-05, + "loss": 0.0272, + "step": 334 + }, + { + "epoch": 1.3562753036437247, + "grad_norm": 0.27780596345001934, + "learning_rate": 3.6544471891832875e-05, + "loss": 0.0224, + "step": 335 + }, + { + "epoch": 1.360323886639676, + "grad_norm": 0.2642821709476628, + "learning_rate": 3.6512629449589983e-05, + "loss": 0.0259, + "step": 336 + }, + { + "epoch": 1.3643724696356276, + "grad_norm": 0.292285262060371, + "learning_rate": 3.6480654972802045e-05, + "loss": 0.0289, + "step": 337 + }, + { + "epoch": 1.368421052631579, + "grad_norm": 0.2650278047594277, + "learning_rate": 3.644854871713611e-05, + "loss": 0.0296, + "step": 338 + }, + { + "epoch": 1.3724696356275303, + "grad_norm": 0.26142617037370397, + "learning_rate": 3.6416310939312924e-05, + "loss": 0.0243, + "step": 339 + }, + { + "epoch": 1.376518218623482, + "grad_norm": 0.26966845309887866, + "learning_rate": 3.638394189710493e-05, + "loss": 0.0249, + "step": 340 + }, + { + "epoch": 1.3805668016194332, + "grad_norm": 0.2900675892421605, + "learning_rate": 3.635144184933412e-05, + "loss": 0.0286, + "step": 341 + }, + { + "epoch": 1.3846153846153846, + "grad_norm": 0.25085119494902886, + "learning_rate": 3.6318811055869994e-05, + "loss": 0.0207, + "step": 342 + }, + { + "epoch": 1.3886639676113361, + "grad_norm": 0.2645928208911707, + "learning_rate": 3.628604977762752e-05, + "loss": 0.0195, + "step": 343 + }, + { + "epoch": 1.3927125506072875, + "grad_norm": 0.26754812371054776, + "learning_rate": 3.6253158276565004e-05, + "loss": 0.0222, + "step": 344 + }, + { + "epoch": 1.3967611336032388, + "grad_norm": 0.27720667432756685, + "learning_rate": 3.6220136815682016e-05, + "loss": 0.0265, + "step": 345 + }, + { + "epoch": 1.4008097165991904, + "grad_norm": 0.23798167132959455, + "learning_rate": 3.618698565901727e-05, + "loss": 0.0178, + "step": 346 + }, + { + "epoch": 1.4048582995951417, + "grad_norm": 0.2830257359922065, + "learning_rate": 3.615370507164654e-05, + "loss": 0.0237, + "step": 347 + }, + { + "epoch": 1.408906882591093, + "grad_norm": 0.25621842540185924, + "learning_rate": 3.61202953196805e-05, + "loss": 0.0201, + "step": 348 + }, + { + "epoch": 1.4129554655870447, + "grad_norm": 0.2711281277335374, + "learning_rate": 3.6086756670262654e-05, + "loss": 0.0229, + "step": 349 + }, + { + "epoch": 1.417004048582996, + "grad_norm": 0.25266298219349365, + "learning_rate": 3.6053089391567135e-05, + "loss": 0.0215, + "step": 350 + }, + { + "epoch": 1.4210526315789473, + "grad_norm": 0.2574121201673984, + "learning_rate": 3.601929375279662e-05, + "loss": 0.0174, + "step": 351 + }, + { + "epoch": 1.425101214574899, + "grad_norm": 0.2796184117859232, + "learning_rate": 3.598537002418012e-05, + "loss": 0.025, + "step": 352 + }, + { + "epoch": 1.4291497975708503, + "grad_norm": 0.2505841171437992, + "learning_rate": 3.5951318476970885e-05, + "loss": 0.0209, + "step": 353 + }, + { + "epoch": 1.4331983805668016, + "grad_norm": 0.22748273003631045, + "learning_rate": 3.591713938344416e-05, + "loss": 0.0238, + "step": 354 + }, + { + "epoch": 1.4372469635627532, + "grad_norm": 0.23510828215132012, + "learning_rate": 3.588283301689507e-05, + "loss": 0.0195, + "step": 355 + }, + { + "epoch": 1.4412955465587045, + "grad_norm": 0.2606231148096484, + "learning_rate": 3.5848399651636424e-05, + "loss": 0.0211, + "step": 356 + }, + { + "epoch": 1.4453441295546559, + "grad_norm": 0.23800922615123796, + "learning_rate": 3.5813839562996473e-05, + "loss": 0.0204, + "step": 357 + }, + { + "epoch": 1.4493927125506074, + "grad_norm": 0.2389297731494306, + "learning_rate": 3.5779153027316775e-05, + "loss": 0.0161, + "step": 358 + }, + { + "epoch": 1.4534412955465588, + "grad_norm": 0.24012223875436042, + "learning_rate": 3.574434032194994e-05, + "loss": 0.0168, + "step": 359 + }, + { + "epoch": 1.45748987854251, + "grad_norm": 0.2488193031322965, + "learning_rate": 3.5709401725257436e-05, + "loss": 0.0168, + "step": 360 + }, + { + "epoch": 1.4615384615384617, + "grad_norm": 0.2634274360264893, + "learning_rate": 3.5674337516607346e-05, + "loss": 0.0193, + "step": 361 + }, + { + "epoch": 1.465587044534413, + "grad_norm": 0.2398329744406485, + "learning_rate": 3.563914797637216e-05, + "loss": 0.0165, + "step": 362 + }, + { + "epoch": 1.4696356275303644, + "grad_norm": 0.2383737155753903, + "learning_rate": 3.560383338592649e-05, + "loss": 0.0209, + "step": 363 + }, + { + "epoch": 1.4736842105263157, + "grad_norm": 0.22544825293129397, + "learning_rate": 3.556839402764487e-05, + "loss": 0.0144, + "step": 364 + }, + { + "epoch": 1.4777327935222673, + "grad_norm": 0.22037351831246482, + "learning_rate": 3.553283018489946e-05, + "loss": 0.0138, + "step": 365 + }, + { + "epoch": 1.4817813765182186, + "grad_norm": 0.2609223753015208, + "learning_rate": 3.54971421420578e-05, + "loss": 0.0168, + "step": 366 + }, + { + "epoch": 1.48582995951417, + "grad_norm": 0.2579392070893139, + "learning_rate": 3.5461330184480524e-05, + "loss": 0.0178, + "step": 367 + }, + { + "epoch": 1.4898785425101215, + "grad_norm": 0.21975087895390036, + "learning_rate": 3.54253945985191e-05, + "loss": 0.0132, + "step": 368 + }, + { + "epoch": 1.4939271255060729, + "grad_norm": 0.25149882927665185, + "learning_rate": 3.5389335671513524e-05, + "loss": 0.0186, + "step": 369 + }, + { + "epoch": 1.4979757085020242, + "grad_norm": 0.23893243920419416, + "learning_rate": 3.5353153691789986e-05, + "loss": 0.0136, + "step": 370 + }, + { + "epoch": 1.5020242914979756, + "grad_norm": 0.21728100179289447, + "learning_rate": 3.531684894865866e-05, + "loss": 0.0135, + "step": 371 + }, + { + "epoch": 1.5060728744939271, + "grad_norm": 0.2272515476360262, + "learning_rate": 3.528042173241128e-05, + "loss": 0.014, + "step": 372 + }, + { + "epoch": 1.5101214574898787, + "grad_norm": 0.23856759437513642, + "learning_rate": 3.524387233431892e-05, + "loss": 0.0168, + "step": 373 + }, + { + "epoch": 1.5141700404858298, + "grad_norm": 0.24108728303112265, + "learning_rate": 3.520720104662958e-05, + "loss": 0.0146, + "step": 374 + }, + { + "epoch": 1.5182186234817814, + "grad_norm": 0.2360252771794431, + "learning_rate": 3.5170408162565904e-05, + "loss": 0.0152, + "step": 375 + }, + { + "epoch": 1.522267206477733, + "grad_norm": 0.21862906017134978, + "learning_rate": 3.513349397632281e-05, + "loss": 0.0119, + "step": 376 + }, + { + "epoch": 1.526315789473684, + "grad_norm": 0.23296856790093445, + "learning_rate": 3.5096458783065145e-05, + "loss": 0.0149, + "step": 377 + }, + { + "epoch": 1.5303643724696356, + "grad_norm": 0.24958904998345008, + "learning_rate": 3.505930287892533e-05, + "loss": 0.0174, + "step": 378 + }, + { + "epoch": 1.5344129554655872, + "grad_norm": 0.23021075395925128, + "learning_rate": 3.502202656100099e-05, + "loss": 0.0136, + "step": 379 + }, + { + "epoch": 1.5384615384615383, + "grad_norm": 0.21087505282531865, + "learning_rate": 3.498463012735256e-05, + "loss": 0.0127, + "step": 380 + }, + { + "epoch": 1.54251012145749, + "grad_norm": 0.23486444358257885, + "learning_rate": 3.494711387700094e-05, + "loss": 0.017, + "step": 381 + }, + { + "epoch": 1.5465587044534415, + "grad_norm": 0.23385231162028328, + "learning_rate": 3.490947810992505e-05, + "loss": 0.0167, + "step": 382 + }, + { + "epoch": 1.5506072874493926, + "grad_norm": 0.22045683347429876, + "learning_rate": 3.48717231270595e-05, + "loss": 0.0131, + "step": 383 + }, + { + "epoch": 1.5546558704453441, + "grad_norm": 0.19213309583562121, + "learning_rate": 3.483384923029211e-05, + "loss": 0.01, + "step": 384 + }, + { + "epoch": 1.5587044534412957, + "grad_norm": 0.2239470254881885, + "learning_rate": 3.479585672246156e-05, + "loss": 0.0123, + "step": 385 + }, + { + "epoch": 1.5627530364372468, + "grad_norm": 0.21937787956203322, + "learning_rate": 3.4757745907354924e-05, + "loss": 0.0146, + "step": 386 + }, + { + "epoch": 1.5668016194331984, + "grad_norm": 0.22424861473709587, + "learning_rate": 3.471951708970527e-05, + "loss": 0.0114, + "step": 387 + }, + { + "epoch": 1.5708502024291497, + "grad_norm": 0.21984312886203508, + "learning_rate": 3.468117057518921e-05, + "loss": 0.012, + "step": 388 + }, + { + "epoch": 1.574898785425101, + "grad_norm": 0.21151557136338123, + "learning_rate": 3.4642706670424426e-05, + "loss": 0.0117, + "step": 389 + }, + { + "epoch": 1.5789473684210527, + "grad_norm": 0.21989793113938913, + "learning_rate": 3.460412568296731e-05, + "loss": 0.0124, + "step": 390 + }, + { + "epoch": 1.582995951417004, + "grad_norm": 0.21425411643603523, + "learning_rate": 3.456542792131039e-05, + "loss": 0.0112, + "step": 391 + }, + { + "epoch": 1.5870445344129553, + "grad_norm": 0.23059305790542023, + "learning_rate": 3.4526613694879936e-05, + "loss": 0.0165, + "step": 392 + }, + { + "epoch": 1.591093117408907, + "grad_norm": 0.23283955509000268, + "learning_rate": 3.448768331403348e-05, + "loss": 0.0159, + "step": 393 + }, + { + "epoch": 1.5951417004048583, + "grad_norm": 0.2048739382820979, + "learning_rate": 3.44486370900573e-05, + "loss": 0.0109, + "step": 394 + }, + { + "epoch": 1.5991902834008096, + "grad_norm": 0.2027210242648886, + "learning_rate": 3.440947533516396e-05, + "loss": 0.0101, + "step": 395 + }, + { + "epoch": 1.6032388663967612, + "grad_norm": 0.23377394322529446, + "learning_rate": 3.4370198362489816e-05, + "loss": 0.0143, + "step": 396 + }, + { + "epoch": 1.6072874493927125, + "grad_norm": 0.20728256535277817, + "learning_rate": 3.43308064860925e-05, + "loss": 0.0116, + "step": 397 + }, + { + "epoch": 1.6113360323886639, + "grad_norm": 0.20103014343447764, + "learning_rate": 3.4291300020948393e-05, + "loss": 0.0102, + "step": 398 + }, + { + "epoch": 1.6153846153846154, + "grad_norm": 0.2042812966406501, + "learning_rate": 3.425167928295015e-05, + "loss": 0.0102, + "step": 399 + }, + { + "epoch": 1.6194331983805668, + "grad_norm": 0.19454509018674146, + "learning_rate": 3.421194458890411e-05, + "loss": 0.0088, + "step": 400 + }, + { + "epoch": 1.623481781376518, + "grad_norm": 0.23744928760381068, + "learning_rate": 3.4172096256527865e-05, + "loss": 0.0122, + "step": 401 + }, + { + "epoch": 1.6275303643724697, + "grad_norm": 0.18395664974209025, + "learning_rate": 3.4132134604447595e-05, + "loss": 0.0081, + "step": 402 + }, + { + "epoch": 1.631578947368421, + "grad_norm": 0.22786615468878046, + "learning_rate": 3.409205995219561e-05, + "loss": 0.0113, + "step": 403 + }, + { + "epoch": 1.6356275303643724, + "grad_norm": 0.1961012737013468, + "learning_rate": 3.4051872620207765e-05, + "loss": 0.0089, + "step": 404 + }, + { + "epoch": 1.639676113360324, + "grad_norm": 0.20806692398679313, + "learning_rate": 3.40115729298209e-05, + "loss": 0.0098, + "step": 405 + }, + { + "epoch": 1.6437246963562753, + "grad_norm": 0.20177081563029137, + "learning_rate": 3.397116120327025e-05, + "loss": 0.0105, + "step": 406 + }, + { + "epoch": 1.6477732793522266, + "grad_norm": 0.18675565858119783, + "learning_rate": 3.3930637763686916e-05, + "loss": 0.0082, + "step": 407 + }, + { + "epoch": 1.6518218623481782, + "grad_norm": 0.2077170128592385, + "learning_rate": 3.389000293509524e-05, + "loss": 0.0128, + "step": 408 + }, + { + "epoch": 1.6558704453441295, + "grad_norm": 0.19428138066803613, + "learning_rate": 3.3849257042410225e-05, + "loss": 0.0091, + "step": 409 + }, + { + "epoch": 1.6599190283400809, + "grad_norm": 0.19310157535259947, + "learning_rate": 3.380840041143494e-05, + "loss": 0.0086, + "step": 410 + }, + { + "epoch": 1.6639676113360324, + "grad_norm": 0.21255482323418964, + "learning_rate": 3.376743336885792e-05, + "loss": 0.0114, + "step": 411 + }, + { + "epoch": 1.6680161943319838, + "grad_norm": 0.17519900096670427, + "learning_rate": 3.372635624225055e-05, + "loss": 0.0085, + "step": 412 + }, + { + "epoch": 1.6720647773279351, + "grad_norm": 0.19192433653073265, + "learning_rate": 3.3685169360064445e-05, + "loss": 0.0084, + "step": 413 + }, + { + "epoch": 1.6761133603238867, + "grad_norm": 0.18722909673591975, + "learning_rate": 3.3643873051628804e-05, + "loss": 0.0074, + "step": 414 + }, + { + "epoch": 1.680161943319838, + "grad_norm": 0.18682583918997364, + "learning_rate": 3.3602467647147814e-05, + "loss": 0.0085, + "step": 415 + }, + { + "epoch": 1.6842105263157894, + "grad_norm": 0.18811639952134346, + "learning_rate": 3.356095347769799e-05, + "loss": 0.0073, + "step": 416 + }, + { + "epoch": 1.688259109311741, + "grad_norm": 0.22019592926084067, + "learning_rate": 3.351933087522552e-05, + "loss": 0.0116, + "step": 417 + }, + { + "epoch": 1.6923076923076923, + "grad_norm": 0.21336318820607345, + "learning_rate": 3.3477600172543626e-05, + "loss": 0.0104, + "step": 418 + }, + { + "epoch": 1.6963562753036436, + "grad_norm": 0.18513431470109426, + "learning_rate": 3.3435761703329894e-05, + "loss": 0.0087, + "step": 419 + }, + { + "epoch": 1.7004048582995952, + "grad_norm": 0.20231812486300338, + "learning_rate": 3.3393815802123606e-05, + "loss": 0.0097, + "step": 420 + }, + { + "epoch": 1.7044534412955465, + "grad_norm": 0.18735032312381955, + "learning_rate": 3.335176280432308e-05, + "loss": 0.0083, + "step": 421 + }, + { + "epoch": 1.708502024291498, + "grad_norm": 0.20035995730428888, + "learning_rate": 3.330960304618295e-05, + "loss": 0.0087, + "step": 422 + }, + { + "epoch": 1.7125506072874495, + "grad_norm": 0.20887310871409348, + "learning_rate": 3.326733686481153e-05, + "loss": 0.0114, + "step": 423 + }, + { + "epoch": 1.7165991902834008, + "grad_norm": 0.17532941275648625, + "learning_rate": 3.3224964598168084e-05, + "loss": 0.008, + "step": 424 + }, + { + "epoch": 1.7206477732793521, + "grad_norm": 0.20366252097538504, + "learning_rate": 3.318248658506012e-05, + "loss": 0.0097, + "step": 425 + }, + { + "epoch": 1.7246963562753037, + "grad_norm": 0.20495229813410956, + "learning_rate": 3.313990316514069e-05, + "loss": 0.0082, + "step": 426 + }, + { + "epoch": 1.728744939271255, + "grad_norm": 0.16510367494652767, + "learning_rate": 3.309721467890571e-05, + "loss": 0.0069, + "step": 427 + }, + { + "epoch": 1.7327935222672064, + "grad_norm": 0.17495983074924704, + "learning_rate": 3.305442146769115e-05, + "loss": 0.0069, + "step": 428 + }, + { + "epoch": 1.736842105263158, + "grad_norm": 0.16766197011218736, + "learning_rate": 3.30115238736704e-05, + "loss": 0.0061, + "step": 429 + }, + { + "epoch": 1.7408906882591093, + "grad_norm": 0.18099102846736834, + "learning_rate": 3.296852223985148e-05, + "loss": 0.0068, + "step": 430 + }, + { + "epoch": 1.7449392712550607, + "grad_norm": 0.1761521870249612, + "learning_rate": 3.2925416910074315e-05, + "loss": 0.0061, + "step": 431 + }, + { + "epoch": 1.7489878542510122, + "grad_norm": 0.183877468615866, + "learning_rate": 3.288220822900796e-05, + "loss": 0.0074, + "step": 432 + }, + { + "epoch": 1.7530364372469636, + "grad_norm": 0.1671280039846581, + "learning_rate": 3.283889654214788e-05, + "loss": 0.0068, + "step": 433 + }, + { + "epoch": 1.757085020242915, + "grad_norm": 0.18591287418614744, + "learning_rate": 3.279548219581317e-05, + "loss": 0.0078, + "step": 434 + }, + { + "epoch": 1.7611336032388665, + "grad_norm": 0.18926218924058139, + "learning_rate": 3.275196553714379e-05, + "loss": 0.0085, + "step": 435 + }, + { + "epoch": 1.7651821862348178, + "grad_norm": 0.17457758904084017, + "learning_rate": 3.270834691409778e-05, + "loss": 0.0067, + "step": 436 + }, + { + "epoch": 1.7692307692307692, + "grad_norm": 0.17036351750037523, + "learning_rate": 3.266462667544849e-05, + "loss": 0.0069, + "step": 437 + }, + { + "epoch": 1.7732793522267207, + "grad_norm": 0.2052513588752089, + "learning_rate": 3.262080517078178e-05, + "loss": 0.0093, + "step": 438 + }, + { + "epoch": 1.777327935222672, + "grad_norm": 0.16708803256899096, + "learning_rate": 3.257688275049323e-05, + "loss": 0.0061, + "step": 439 + }, + { + "epoch": 1.7813765182186234, + "grad_norm": 0.1674351970135278, + "learning_rate": 3.253285976578535e-05, + "loss": 0.0071, + "step": 440 + }, + { + "epoch": 1.785425101214575, + "grad_norm": 0.19470676356007338, + "learning_rate": 3.2488736568664756e-05, + "loss": 0.0083, + "step": 441 + }, + { + "epoch": 1.7894736842105263, + "grad_norm": 0.17537355035418706, + "learning_rate": 3.244451351193935e-05, + "loss": 0.0068, + "step": 442 + }, + { + "epoch": 1.7935222672064777, + "grad_norm": 0.19180258899899966, + "learning_rate": 3.240019094921551e-05, + "loss": 0.0077, + "step": 443 + }, + { + "epoch": 1.7975708502024292, + "grad_norm": 0.16786857643373065, + "learning_rate": 3.235576923489527e-05, + "loss": 0.0057, + "step": 444 + }, + { + "epoch": 1.8016194331983806, + "grad_norm": 0.16227524043797809, + "learning_rate": 3.2311248724173476e-05, + "loss": 0.0062, + "step": 445 + }, + { + "epoch": 1.805668016194332, + "grad_norm": 0.1920571153970536, + "learning_rate": 3.226662977303494e-05, + "loss": 0.0063, + "step": 446 + }, + { + "epoch": 1.8097165991902835, + "grad_norm": 0.20433928382049835, + "learning_rate": 3.2221912738251626e-05, + "loss": 0.0087, + "step": 447 + }, + { + "epoch": 1.8137651821862348, + "grad_norm": 0.18275947474228008, + "learning_rate": 3.2177097977379736e-05, + "loss": 0.0069, + "step": 448 + }, + { + "epoch": 1.8178137651821862, + "grad_norm": 0.1755947221516228, + "learning_rate": 3.21321858487569e-05, + "loss": 0.0065, + "step": 449 + }, + { + "epoch": 1.8218623481781377, + "grad_norm": 0.1757584849378906, + "learning_rate": 3.208717671149931e-05, + "loss": 0.0059, + "step": 450 + }, + { + "epoch": 1.825910931174089, + "grad_norm": 0.16460057996590402, + "learning_rate": 3.204207092549882e-05, + "loss": 0.006, + "step": 451 + }, + { + "epoch": 1.8299595141700404, + "grad_norm": 0.1686220360043713, + "learning_rate": 3.19968688514201e-05, + "loss": 0.0066, + "step": 452 + }, + { + "epoch": 1.834008097165992, + "grad_norm": 0.1751296590381264, + "learning_rate": 3.195157085069774e-05, + "loss": 0.0079, + "step": 453 + }, + { + "epoch": 1.8380566801619433, + "grad_norm": 0.1588832640642069, + "learning_rate": 3.190617728553332e-05, + "loss": 0.0067, + "step": 454 + }, + { + "epoch": 1.8421052631578947, + "grad_norm": 0.16025606618446808, + "learning_rate": 3.1860688518892606e-05, + "loss": 0.0057, + "step": 455 + }, + { + "epoch": 1.8461538461538463, + "grad_norm": 0.1738741171954513, + "learning_rate": 3.181510491450254e-05, + "loss": 0.0073, + "step": 456 + }, + { + "epoch": 1.8502024291497976, + "grad_norm": 0.17079966119578927, + "learning_rate": 3.176942683684842e-05, + "loss": 0.0063, + "step": 457 + }, + { + "epoch": 1.854251012145749, + "grad_norm": 0.17422587329773445, + "learning_rate": 3.1723654651170934e-05, + "loss": 0.0058, + "step": 458 + }, + { + "epoch": 1.8582995951417005, + "grad_norm": 0.1749801826339792, + "learning_rate": 3.167778872346327e-05, + "loss": 0.0059, + "step": 459 + }, + { + "epoch": 1.8623481781376519, + "grad_norm": 0.16396065626347045, + "learning_rate": 3.163182942046815e-05, + "loss": 0.0062, + "step": 460 + }, + { + "epoch": 1.8663967611336032, + "grad_norm": 0.180714136397594, + "learning_rate": 3.158577710967494e-05, + "loss": 0.0059, + "step": 461 + }, + { + "epoch": 1.8704453441295548, + "grad_norm": 0.17188579857956154, + "learning_rate": 3.153963215931669e-05, + "loss": 0.0061, + "step": 462 + }, + { + "epoch": 1.874493927125506, + "grad_norm": 0.159011835514648, + "learning_rate": 3.14933949383672e-05, + "loss": 0.0054, + "step": 463 + }, + { + "epoch": 1.8785425101214575, + "grad_norm": 0.16113321304393674, + "learning_rate": 3.144706581653805e-05, + "loss": 0.0055, + "step": 464 + }, + { + "epoch": 1.882591093117409, + "grad_norm": 0.1613032085132324, + "learning_rate": 3.140064516427566e-05, + "loss": 0.0059, + "step": 465 + }, + { + "epoch": 1.8866396761133604, + "grad_norm": 0.16124030027626188, + "learning_rate": 3.135413335275833e-05, + "loss": 0.006, + "step": 466 + }, + { + "epoch": 1.8906882591093117, + "grad_norm": 0.16261101426208163, + "learning_rate": 3.130753075389327e-05, + "loss": 0.006, + "step": 467 + }, + { + "epoch": 1.8947368421052633, + "grad_norm": 0.1627597268737651, + "learning_rate": 3.1260837740313625e-05, + "loss": 0.0051, + "step": 468 + }, + { + "epoch": 1.8987854251012146, + "grad_norm": 0.15852239124410028, + "learning_rate": 3.121405468537547e-05, + "loss": 0.0054, + "step": 469 + }, + { + "epoch": 1.902834008097166, + "grad_norm": 0.18505121987526874, + "learning_rate": 3.116718196315488e-05, + "loss": 0.0058, + "step": 470 + }, + { + "epoch": 1.9068825910931175, + "grad_norm": 0.15083580907857927, + "learning_rate": 3.1120219948444894e-05, + "loss": 0.0056, + "step": 471 + }, + { + "epoch": 1.9109311740890689, + "grad_norm": 0.15387067013875938, + "learning_rate": 3.107316901675252e-05, + "loss": 0.0056, + "step": 472 + }, + { + "epoch": 1.9149797570850202, + "grad_norm": 0.1646552814726087, + "learning_rate": 3.1026029544295745e-05, + "loss": 0.0058, + "step": 473 + }, + { + "epoch": 1.9190283400809718, + "grad_norm": 0.15502858477327788, + "learning_rate": 3.097880190800057e-05, + "loss": 0.0046, + "step": 474 + }, + { + "epoch": 1.9230769230769231, + "grad_norm": 0.14962112617535997, + "learning_rate": 3.093148648549788e-05, + "loss": 0.0051, + "step": 475 + }, + { + "epoch": 1.9271255060728745, + "grad_norm": 0.14904974166718218, + "learning_rate": 3.088408365512055e-05, + "loss": 0.0043, + "step": 476 + }, + { + "epoch": 1.931174089068826, + "grad_norm": 0.15679710325414856, + "learning_rate": 3.083659379590034e-05, + "loss": 0.0051, + "step": 477 + }, + { + "epoch": 1.9352226720647774, + "grad_norm": 0.16412212698040274, + "learning_rate": 3.0789017287564913e-05, + "loss": 0.0052, + "step": 478 + }, + { + "epoch": 1.9392712550607287, + "grad_norm": 0.14249022804879338, + "learning_rate": 3.074135451053475e-05, + "loss": 0.0042, + "step": 479 + }, + { + "epoch": 1.9433198380566803, + "grad_norm": 0.1388698608740847, + "learning_rate": 3.069360584592015e-05, + "loss": 0.0039, + "step": 480 + }, + { + "epoch": 1.9473684210526314, + "grad_norm": 0.14866721664139596, + "learning_rate": 3.064577167551816e-05, + "loss": 0.0051, + "step": 481 + }, + { + "epoch": 1.951417004048583, + "grad_norm": 0.15689468269886842, + "learning_rate": 3.059785238180953e-05, + "loss": 0.0052, + "step": 482 + }, + { + "epoch": 1.9554655870445345, + "grad_norm": 0.14331753340582484, + "learning_rate": 3.054984834795566e-05, + "loss": 0.0043, + "step": 483 + }, + { + "epoch": 1.9595141700404857, + "grad_norm": 0.1504499008958325, + "learning_rate": 3.0501759957795537e-05, + "loss": 0.005, + "step": 484 + }, + { + "epoch": 1.9635627530364372, + "grad_norm": 0.15376482112962506, + "learning_rate": 3.0453587595842627e-05, + "loss": 0.0047, + "step": 485 + }, + { + "epoch": 1.9676113360323888, + "grad_norm": 0.1426300256290253, + "learning_rate": 3.0405331647281856e-05, + "loss": 0.0042, + "step": 486 + }, + { + "epoch": 1.97165991902834, + "grad_norm": 0.15466127241797967, + "learning_rate": 3.0356992497966508e-05, + "loss": 0.0047, + "step": 487 + }, + { + "epoch": 1.9757085020242915, + "grad_norm": 0.14955265653438327, + "learning_rate": 3.0308570534415137e-05, + "loss": 0.0048, + "step": 488 + }, + { + "epoch": 1.979757085020243, + "grad_norm": 0.14309770793959614, + "learning_rate": 3.0260066143808467e-05, + "loss": 0.0043, + "step": 489 + }, + { + "epoch": 1.9838056680161942, + "grad_norm": 0.14777225166108104, + "learning_rate": 3.021147971398632e-05, + "loss": 0.0046, + "step": 490 + }, + { + "epoch": 1.9878542510121457, + "grad_norm": 0.14425740944157603, + "learning_rate": 3.0162811633444478e-05, + "loss": 0.0044, + "step": 491 + }, + { + "epoch": 1.9919028340080973, + "grad_norm": 0.13077675773946432, + "learning_rate": 3.0114062291331635e-05, + "loss": 0.0037, + "step": 492 + }, + { + "epoch": 1.9959514170040484, + "grad_norm": 0.16021435879966406, + "learning_rate": 3.0065232077446217e-05, + "loss": 0.0044, + "step": 493 + }, + { + "epoch": 2.0, + "grad_norm": 0.1369932203350117, + "learning_rate": 3.001632138223332e-05, + "loss": 0.0044, + "step": 494 + }, + { + "epoch": 2.0040485829959516, + "grad_norm": 0.12960454604195717, + "learning_rate": 2.9967330596781565e-05, + "loss": 0.0031, + "step": 495 + }, + { + "epoch": 2.0080971659919027, + "grad_norm": 0.1282421861341016, + "learning_rate": 2.9918260112819952e-05, + "loss": 0.0031, + "step": 496 + }, + { + "epoch": 2.0121457489878543, + "grad_norm": 0.1414365087089226, + "learning_rate": 2.9869110322714778e-05, + "loss": 0.0036, + "step": 497 + }, + { + "epoch": 2.016194331983806, + "grad_norm": 0.13498756464941367, + "learning_rate": 2.9819881619466447e-05, + "loss": 0.0038, + "step": 498 + }, + { + "epoch": 2.020242914979757, + "grad_norm": 0.1358114775488299, + "learning_rate": 2.9770574396706364e-05, + "loss": 0.0036, + "step": 499 + }, + { + "epoch": 2.0242914979757085, + "grad_norm": 0.1360998959392684, + "learning_rate": 2.972118904869376e-05, + "loss": 0.0031, + "step": 500 + }, + { + "epoch": 2.02834008097166, + "grad_norm": 0.13773919589001601, + "learning_rate": 2.9671725970312563e-05, + "loss": 0.0033, + "step": 501 + }, + { + "epoch": 2.032388663967611, + "grad_norm": 0.14321503399591512, + "learning_rate": 2.9622185557068238e-05, + "loss": 0.0034, + "step": 502 + }, + { + "epoch": 2.0364372469635628, + "grad_norm": 0.12552548599760066, + "learning_rate": 2.95725682050846e-05, + "loss": 0.0031, + "step": 503 + }, + { + "epoch": 2.0404858299595143, + "grad_norm": 0.17769553379309408, + "learning_rate": 2.9522874311100684e-05, + "loss": 0.0031, + "step": 504 + }, + { + "epoch": 2.0445344129554655, + "grad_norm": 0.14885694813096034, + "learning_rate": 2.9473104272467526e-05, + "loss": 0.0043, + "step": 505 + }, + { + "epoch": 2.048582995951417, + "grad_norm": 0.12978282107370198, + "learning_rate": 2.942325848714505e-05, + "loss": 0.0035, + "step": 506 + }, + { + "epoch": 2.0526315789473686, + "grad_norm": 0.12930118300443738, + "learning_rate": 2.9373337353698804e-05, + "loss": 0.0033, + "step": 507 + }, + { + "epoch": 2.0566801619433197, + "grad_norm": 0.13357777503065282, + "learning_rate": 2.932334127129686e-05, + "loss": 0.0033, + "step": 508 + }, + { + "epoch": 2.0607287449392713, + "grad_norm": 0.12528218633086216, + "learning_rate": 2.9273270639706547e-05, + "loss": 0.0033, + "step": 509 + }, + { + "epoch": 2.064777327935223, + "grad_norm": 0.11640162204608752, + "learning_rate": 2.922312585929131e-05, + "loss": 0.0028, + "step": 510 + }, + { + "epoch": 2.068825910931174, + "grad_norm": 0.12109694733640984, + "learning_rate": 2.9172907331007478e-05, + "loss": 0.0028, + "step": 511 + }, + { + "epoch": 2.0728744939271255, + "grad_norm": 0.13248140491668753, + "learning_rate": 2.9122615456401053e-05, + "loss": 0.0035, + "step": 512 + }, + { + "epoch": 2.076923076923077, + "grad_norm": 0.13815333477108416, + "learning_rate": 2.9072250637604535e-05, + "loss": 0.0036, + "step": 513 + }, + { + "epoch": 2.080971659919028, + "grad_norm": 0.13596788685484124, + "learning_rate": 2.9021813277333667e-05, + "loss": 0.0037, + "step": 514 + }, + { + "epoch": 2.08502024291498, + "grad_norm": 0.1140848675110523, + "learning_rate": 2.8971303778884244e-05, + "loss": 0.0028, + "step": 515 + }, + { + "epoch": 2.0890688259109313, + "grad_norm": 0.11223418360553557, + "learning_rate": 2.8920722546128863e-05, + "loss": 0.0027, + "step": 516 + }, + { + "epoch": 2.0931174089068825, + "grad_norm": 0.14073288159408137, + "learning_rate": 2.887006998351372e-05, + "loss": 0.0035, + "step": 517 + }, + { + "epoch": 2.097165991902834, + "grad_norm": 0.11376603437612794, + "learning_rate": 2.881934649605535e-05, + "loss": 0.0028, + "step": 518 + }, + { + "epoch": 2.1012145748987856, + "grad_norm": 0.1294080335234807, + "learning_rate": 2.8768552489337408e-05, + "loss": 0.0031, + "step": 519 + }, + { + "epoch": 2.1052631578947367, + "grad_norm": 0.1310330294434518, + "learning_rate": 2.8717688369507425e-05, + "loss": 0.003, + "step": 520 + }, + { + "epoch": 2.1093117408906883, + "grad_norm": 0.13424185246363862, + "learning_rate": 2.8666754543273535e-05, + "loss": 0.0035, + "step": 521 + }, + { + "epoch": 2.11336032388664, + "grad_norm": 0.11244964013825477, + "learning_rate": 2.8615751417901273e-05, + "loss": 0.0024, + "step": 522 + }, + { + "epoch": 2.117408906882591, + "grad_norm": 0.13167000771924192, + "learning_rate": 2.856467940121025e-05, + "loss": 0.0036, + "step": 523 + }, + { + "epoch": 2.1214574898785425, + "grad_norm": 0.12539685828003772, + "learning_rate": 2.8513538901570964e-05, + "loss": 0.0027, + "step": 524 + }, + { + "epoch": 2.125506072874494, + "grad_norm": 0.12000511861127591, + "learning_rate": 2.846233032790149e-05, + "loss": 0.0031, + "step": 525 + }, + { + "epoch": 2.1295546558704452, + "grad_norm": 0.13264490546222618, + "learning_rate": 2.84110540896642e-05, + "loss": 0.0034, + "step": 526 + }, + { + "epoch": 2.133603238866397, + "grad_norm": 0.12739819034069674, + "learning_rate": 2.8359710596862555e-05, + "loss": 0.0032, + "step": 527 + }, + { + "epoch": 2.1376518218623484, + "grad_norm": 0.11361637247040562, + "learning_rate": 2.8308300260037734e-05, + "loss": 0.0027, + "step": 528 + }, + { + "epoch": 2.1417004048582995, + "grad_norm": 0.11503452349016562, + "learning_rate": 2.8256823490265424e-05, + "loss": 0.0026, + "step": 529 + }, + { + "epoch": 2.145748987854251, + "grad_norm": 0.13659482785350346, + "learning_rate": 2.8205280699152515e-05, + "loss": 0.0032, + "step": 530 + }, + { + "epoch": 2.1497975708502026, + "grad_norm": 0.11635286536306273, + "learning_rate": 2.8153672298833777e-05, + "loss": 0.0027, + "step": 531 + }, + { + "epoch": 2.1538461538461537, + "grad_norm": 0.10375233808351494, + "learning_rate": 2.8101998701968615e-05, + "loss": 0.0022, + "step": 532 + }, + { + "epoch": 2.1578947368421053, + "grad_norm": 0.13310261257445796, + "learning_rate": 2.805026032173773e-05, + "loss": 0.0028, + "step": 533 + }, + { + "epoch": 2.161943319838057, + "grad_norm": 0.11694482750588328, + "learning_rate": 2.7998457571839832e-05, + "loss": 0.0027, + "step": 534 + }, + { + "epoch": 2.165991902834008, + "grad_norm": 0.11163379013259818, + "learning_rate": 2.7946590866488322e-05, + "loss": 0.0025, + "step": 535 + }, + { + "epoch": 2.1700404858299596, + "grad_norm": 0.11564125886510764, + "learning_rate": 2.7894660620408e-05, + "loss": 0.0027, + "step": 536 + }, + { + "epoch": 2.174089068825911, + "grad_norm": 0.11653545397831337, + "learning_rate": 2.784266724883173e-05, + "loss": 0.0026, + "step": 537 + }, + { + "epoch": 2.1781376518218623, + "grad_norm": 0.11952925530893235, + "learning_rate": 2.7790611167497128e-05, + "loss": 0.0028, + "step": 538 + }, + { + "epoch": 2.182186234817814, + "grad_norm": 0.11792969218267083, + "learning_rate": 2.7738492792643224e-05, + "loss": 0.0026, + "step": 539 + }, + { + "epoch": 2.1862348178137654, + "grad_norm": 0.12954972452342084, + "learning_rate": 2.768631254100716e-05, + "loss": 0.003, + "step": 540 + }, + { + "epoch": 2.1902834008097165, + "grad_norm": 0.1126559855149955, + "learning_rate": 2.7634070829820836e-05, + "loss": 0.0025, + "step": 541 + }, + { + "epoch": 2.194331983805668, + "grad_norm": 0.11375855517330709, + "learning_rate": 2.7581768076807592e-05, + "loss": 0.0024, + "step": 542 + }, + { + "epoch": 2.1983805668016196, + "grad_norm": 0.12737188032560495, + "learning_rate": 2.7529404700178832e-05, + "loss": 0.003, + "step": 543 + }, + { + "epoch": 2.2024291497975708, + "grad_norm": 0.09762169093858009, + "learning_rate": 2.7476981118630735e-05, + "loss": 0.002, + "step": 544 + }, + { + "epoch": 2.2064777327935223, + "grad_norm": 0.11412534266973201, + "learning_rate": 2.742449775134085e-05, + "loss": 0.0024, + "step": 545 + }, + { + "epoch": 2.2105263157894735, + "grad_norm": 0.1313341462638357, + "learning_rate": 2.7371955017964788e-05, + "loss": 0.0029, + "step": 546 + }, + { + "epoch": 2.214574898785425, + "grad_norm": 0.1327174753520099, + "learning_rate": 2.7319353338632845e-05, + "loss": 0.0033, + "step": 547 + }, + { + "epoch": 2.2186234817813766, + "grad_norm": 0.1128304720196744, + "learning_rate": 2.7266693133946645e-05, + "loss": 0.0027, + "step": 548 + }, + { + "epoch": 2.2226720647773277, + "grad_norm": 0.12484643298744519, + "learning_rate": 2.7213974824975785e-05, + "loss": 0.0029, + "step": 549 + }, + { + "epoch": 2.2267206477732793, + "grad_norm": 0.1232876985376994, + "learning_rate": 2.7161198833254464e-05, + "loss": 0.0027, + "step": 550 + }, + { + "epoch": 2.230769230769231, + "grad_norm": 0.11500648676797273, + "learning_rate": 2.7108365580778095e-05, + "loss": 0.0027, + "step": 551 + }, + { + "epoch": 2.234817813765182, + "grad_norm": 0.12389545799230041, + "learning_rate": 2.705547548999996e-05, + "loss": 0.0032, + "step": 552 + }, + { + "epoch": 2.2388663967611335, + "grad_norm": 0.11223527600687436, + "learning_rate": 2.7002528983827817e-05, + "loss": 0.0027, + "step": 553 + }, + { + "epoch": 2.242914979757085, + "grad_norm": 0.10581484711559534, + "learning_rate": 2.6949526485620517e-05, + "loss": 0.0024, + "step": 554 + }, + { + "epoch": 2.246963562753036, + "grad_norm": 0.11204799914584336, + "learning_rate": 2.6896468419184623e-05, + "loss": 0.0026, + "step": 555 + }, + { + "epoch": 2.251012145748988, + "grad_norm": 0.10078328522548928, + "learning_rate": 2.6843355208771024e-05, + "loss": 0.0022, + "step": 556 + }, + { + "epoch": 2.2550607287449393, + "grad_norm": 0.1012190481139576, + "learning_rate": 2.6790187279071536e-05, + "loss": 0.0022, + "step": 557 + }, + { + "epoch": 2.2591093117408905, + "grad_norm": 0.117268060246357, + "learning_rate": 2.67369650552155e-05, + "loss": 0.0026, + "step": 558 + }, + { + "epoch": 2.263157894736842, + "grad_norm": 0.09858422061859334, + "learning_rate": 2.6683688962766407e-05, + "loss": 0.0017, + "step": 559 + }, + { + "epoch": 2.2672064777327936, + "grad_norm": 0.10470194148436021, + "learning_rate": 2.6630359427718468e-05, + "loss": 0.0022, + "step": 560 + }, + { + "epoch": 2.2712550607287447, + "grad_norm": 0.1066716935848619, + "learning_rate": 2.6576976876493225e-05, + "loss": 0.0023, + "step": 561 + }, + { + "epoch": 2.2753036437246963, + "grad_norm": 0.11177732061630462, + "learning_rate": 2.652354173593614e-05, + "loss": 0.0021, + "step": 562 + }, + { + "epoch": 2.279352226720648, + "grad_norm": 0.110887193036861, + "learning_rate": 2.647005443331317e-05, + "loss": 0.0025, + "step": 563 + }, + { + "epoch": 2.283400809716599, + "grad_norm": 0.10502886712187892, + "learning_rate": 2.6416515396307354e-05, + "loss": 0.002, + "step": 564 + }, + { + "epoch": 2.2874493927125505, + "grad_norm": 0.0980989609801273, + "learning_rate": 2.636292505301542e-05, + "loss": 0.0017, + "step": 565 + }, + { + "epoch": 2.291497975708502, + "grad_norm": 0.1218890133481451, + "learning_rate": 2.630928383194431e-05, + "loss": 0.0022, + "step": 566 + }, + { + "epoch": 2.2955465587044532, + "grad_norm": 0.10689107351467333, + "learning_rate": 2.625559216200781e-05, + "loss": 0.0022, + "step": 567 + }, + { + "epoch": 2.299595141700405, + "grad_norm": 0.09490725178499129, + "learning_rate": 2.620185047252308e-05, + "loss": 0.0017, + "step": 568 + }, + { + "epoch": 2.3036437246963564, + "grad_norm": 0.09911919477463059, + "learning_rate": 2.614805919320723e-05, + "loss": 0.0018, + "step": 569 + }, + { + "epoch": 2.3076923076923075, + "grad_norm": 0.12373795521236076, + "learning_rate": 2.6094218754173895e-05, + "loss": 0.003, + "step": 570 + }, + { + "epoch": 2.311740890688259, + "grad_norm": 0.11883575926446971, + "learning_rate": 2.60403295859298e-05, + "loss": 0.002, + "step": 571 + }, + { + "epoch": 2.3157894736842106, + "grad_norm": 0.09873326485853942, + "learning_rate": 2.5986392119371297e-05, + "loss": 0.0018, + "step": 572 + }, + { + "epoch": 2.3198380566801617, + "grad_norm": 0.10821117851040493, + "learning_rate": 2.5932406785780928e-05, + "loss": 0.0023, + "step": 573 + }, + { + "epoch": 2.3238866396761133, + "grad_norm": 0.09971032960804566, + "learning_rate": 2.5878374016823987e-05, + "loss": 0.0018, + "step": 574 + }, + { + "epoch": 2.327935222672065, + "grad_norm": 0.09757560652872671, + "learning_rate": 2.5824294244545045e-05, + "loss": 0.0016, + "step": 575 + }, + { + "epoch": 2.331983805668016, + "grad_norm": 0.09413549309630384, + "learning_rate": 2.577016790136453e-05, + "loss": 0.0017, + "step": 576 + }, + { + "epoch": 2.3360323886639676, + "grad_norm": 0.09543395047076698, + "learning_rate": 2.5715995420075225e-05, + "loss": 0.0017, + "step": 577 + }, + { + "epoch": 2.340080971659919, + "grad_norm": 0.10502005849398484, + "learning_rate": 2.5661777233838864e-05, + "loss": 0.0018, + "step": 578 + }, + { + "epoch": 2.3441295546558703, + "grad_norm": 0.09868735651183642, + "learning_rate": 2.5607513776182616e-05, + "loss": 0.0019, + "step": 579 + }, + { + "epoch": 2.348178137651822, + "grad_norm": 0.10330035787018109, + "learning_rate": 2.5553205480995633e-05, + "loss": 0.002, + "step": 580 + }, + { + "epoch": 2.3522267206477734, + "grad_norm": 0.10044371051655197, + "learning_rate": 2.5498852782525604e-05, + "loss": 0.0017, + "step": 581 + }, + { + "epoch": 2.3562753036437245, + "grad_norm": 0.11758328872517873, + "learning_rate": 2.5444456115375252e-05, + "loss": 0.0024, + "step": 582 + }, + { + "epoch": 2.360323886639676, + "grad_norm": 0.09762212742927469, + "learning_rate": 2.539001591449889e-05, + "loss": 0.0015, + "step": 583 + }, + { + "epoch": 2.3643724696356276, + "grad_norm": 0.0952177066991939, + "learning_rate": 2.53355326151989e-05, + "loss": 0.002, + "step": 584 + }, + { + "epoch": 2.3684210526315788, + "grad_norm": 0.09210280802116268, + "learning_rate": 2.5281006653122288e-05, + "loss": 0.0016, + "step": 585 + }, + { + "epoch": 2.3724696356275303, + "grad_norm": 0.10977175822349436, + "learning_rate": 2.5226438464257217e-05, + "loss": 0.0022, + "step": 586 + }, + { + "epoch": 2.376518218623482, + "grad_norm": 0.09906252679231259, + "learning_rate": 2.5171828484929447e-05, + "loss": 0.0016, + "step": 587 + }, + { + "epoch": 2.380566801619433, + "grad_norm": 0.08534000282943922, + "learning_rate": 2.5117177151798934e-05, + "loss": 0.0014, + "step": 588 + }, + { + "epoch": 2.3846153846153846, + "grad_norm": 0.09633962793803941, + "learning_rate": 2.506248490185627e-05, + "loss": 0.0021, + "step": 589 + }, + { + "epoch": 2.388663967611336, + "grad_norm": 0.09048816254774779, + "learning_rate": 2.500775217241923e-05, + "loss": 0.0015, + "step": 590 + }, + { + "epoch": 2.3927125506072873, + "grad_norm": 0.09471003500327373, + "learning_rate": 2.4952979401129266e-05, + "loss": 0.002, + "step": 591 + }, + { + "epoch": 2.396761133603239, + "grad_norm": 0.10233991052759076, + "learning_rate": 2.4898167025948e-05, + "loss": 0.0019, + "step": 592 + }, + { + "epoch": 2.4008097165991904, + "grad_norm": 0.0871154065396495, + "learning_rate": 2.4843315485153703e-05, + "loss": 0.0017, + "step": 593 + }, + { + "epoch": 2.4048582995951415, + "grad_norm": 0.08262044184682142, + "learning_rate": 2.4788425217337856e-05, + "loss": 0.0013, + "step": 594 + }, + { + "epoch": 2.408906882591093, + "grad_norm": 0.09068031291325453, + "learning_rate": 2.4733496661401572e-05, + "loss": 0.0015, + "step": 595 + }, + { + "epoch": 2.4129554655870447, + "grad_norm": 0.09289510714254343, + "learning_rate": 2.4678530256552114e-05, + "loss": 0.0017, + "step": 596 + }, + { + "epoch": 2.417004048582996, + "grad_norm": 0.09982830895747369, + "learning_rate": 2.46235264422994e-05, + "loss": 0.0015, + "step": 597 + }, + { + "epoch": 2.4210526315789473, + "grad_norm": 0.08585545227447924, + "learning_rate": 2.4568485658452452e-05, + "loss": 0.0014, + "step": 598 + }, + { + "epoch": 2.425101214574899, + "grad_norm": 0.09557332801750142, + "learning_rate": 2.451340834511592e-05, + "loss": 0.0012, + "step": 599 + }, + { + "epoch": 2.42914979757085, + "grad_norm": 0.08470610904102943, + "learning_rate": 2.4458294942686528e-05, + "loss": 0.0016, + "step": 600 + }, + { + "epoch": 2.4331983805668016, + "grad_norm": 0.07410991770645954, + "learning_rate": 2.440314589184958e-05, + "loss": 0.0011, + "step": 601 + }, + { + "epoch": 2.437246963562753, + "grad_norm": 0.0954540144540923, + "learning_rate": 2.434796163357541e-05, + "loss": 0.0017, + "step": 602 + }, + { + "epoch": 2.4412955465587043, + "grad_norm": 0.1015647653282561, + "learning_rate": 2.429274260911588e-05, + "loss": 0.0017, + "step": 603 + }, + { + "epoch": 2.445344129554656, + "grad_norm": 0.10437968208312502, + "learning_rate": 2.423748926000084e-05, + "loss": 0.0018, + "step": 604 + }, + { + "epoch": 2.4493927125506074, + "grad_norm": 0.08306529747636363, + "learning_rate": 2.418220202803459e-05, + "loss": 0.0012, + "step": 605 + }, + { + "epoch": 2.4534412955465585, + "grad_norm": 0.09346846264726467, + "learning_rate": 2.412688135529237e-05, + "loss": 0.0016, + "step": 606 + }, + { + "epoch": 2.45748987854251, + "grad_norm": 0.07420341621034773, + "learning_rate": 2.4071527684116807e-05, + "loss": 0.0011, + "step": 607 + }, + { + "epoch": 2.4615384615384617, + "grad_norm": 0.09329708595297592, + "learning_rate": 2.4016141457114372e-05, + "loss": 0.0013, + "step": 608 + }, + { + "epoch": 2.465587044534413, + "grad_norm": 0.09311983602977335, + "learning_rate": 2.396072311715187e-05, + "loss": 0.0016, + "step": 609 + }, + { + "epoch": 2.4696356275303644, + "grad_norm": 0.08490856708849338, + "learning_rate": 2.3905273107352863e-05, + "loss": 0.0016, + "step": 610 + }, + { + "epoch": 2.473684210526316, + "grad_norm": 0.08501231586680766, + "learning_rate": 2.3849791871094157e-05, + "loss": 0.0014, + "step": 611 + }, + { + "epoch": 2.477732793522267, + "grad_norm": 0.0929500786282343, + "learning_rate": 2.379427985200223e-05, + "loss": 0.0013, + "step": 612 + }, + { + "epoch": 2.4817813765182186, + "grad_norm": 0.06949590924785025, + "learning_rate": 2.373873749394972e-05, + "loss": 0.0012, + "step": 613 + }, + { + "epoch": 2.48582995951417, + "grad_norm": 0.08301763070107807, + "learning_rate": 2.3683165241051836e-05, + "loss": 0.0017, + "step": 614 + }, + { + "epoch": 2.4898785425101213, + "grad_norm": 0.08649545615062469, + "learning_rate": 2.3627563537662823e-05, + "loss": 0.0014, + "step": 615 + }, + { + "epoch": 2.493927125506073, + "grad_norm": 0.09552486138759125, + "learning_rate": 2.3571932828372433e-05, + "loss": 0.0017, + "step": 616 + }, + { + "epoch": 2.4979757085020244, + "grad_norm": 0.0854621365431451, + "learning_rate": 2.351627355800233e-05, + "loss": 0.0013, + "step": 617 + }, + { + "epoch": 2.5020242914979756, + "grad_norm": 0.07506056222974641, + "learning_rate": 2.346058617160256e-05, + "loss": 0.0012, + "step": 618 + }, + { + "epoch": 2.506072874493927, + "grad_norm": 0.08204636733503627, + "learning_rate": 2.3404871114447976e-05, + "loss": 0.0014, + "step": 619 + }, + { + "epoch": 2.5101214574898787, + "grad_norm": 0.07141038456091979, + "learning_rate": 2.334912883203469e-05, + "loss": 0.0009, + "step": 620 + }, + { + "epoch": 2.51417004048583, + "grad_norm": 0.06908954747367257, + "learning_rate": 2.3293359770076505e-05, + "loss": 0.0008, + "step": 621 + }, + { + "epoch": 2.5182186234817814, + "grad_norm": 0.07328090486858517, + "learning_rate": 2.323756437450136e-05, + "loss": 0.0009, + "step": 622 + }, + { + "epoch": 2.522267206477733, + "grad_norm": 0.08556880784240821, + "learning_rate": 2.3181743091447754e-05, + "loss": 0.001, + "step": 623 + }, + { + "epoch": 2.526315789473684, + "grad_norm": 0.08940655249859024, + "learning_rate": 2.3125896367261164e-05, + "loss": 0.0015, + "step": 624 + }, + { + "epoch": 2.5303643724696356, + "grad_norm": 0.07866478399215428, + "learning_rate": 2.307002464849052e-05, + "loss": 0.0012, + "step": 625 + }, + { + "epoch": 2.534412955465587, + "grad_norm": 0.08458480088520309, + "learning_rate": 2.30141283818846e-05, + "loss": 0.0012, + "step": 626 + }, + { + "epoch": 2.5384615384615383, + "grad_norm": 0.0726085787989922, + "learning_rate": 2.2958208014388462e-05, + "loss": 0.001, + "step": 627 + }, + { + "epoch": 2.54251012145749, + "grad_norm": 0.07403654691156016, + "learning_rate": 2.2902263993139857e-05, + "loss": 0.0009, + "step": 628 + }, + { + "epoch": 2.5465587044534415, + "grad_norm": 0.09743666474725009, + "learning_rate": 2.2846296765465708e-05, + "loss": 0.0018, + "step": 629 + }, + { + "epoch": 2.5506072874493926, + "grad_norm": 0.07514455715866586, + "learning_rate": 2.2790306778878465e-05, + "loss": 0.0011, + "step": 630 + }, + { + "epoch": 2.554655870445344, + "grad_norm": 0.05710987647564663, + "learning_rate": 2.273429448107257e-05, + "loss": 0.0006, + "step": 631 + }, + { + "epoch": 2.5587044534412957, + "grad_norm": 0.07834651971489404, + "learning_rate": 2.2678260319920868e-05, + "loss": 0.0011, + "step": 632 + }, + { + "epoch": 2.562753036437247, + "grad_norm": 0.07761255001756283, + "learning_rate": 2.262220474347102e-05, + "loss": 0.0012, + "step": 633 + }, + { + "epoch": 2.5668016194331984, + "grad_norm": 0.07653959711487222, + "learning_rate": 2.2566128199941918e-05, + "loss": 0.0011, + "step": 634 + }, + { + "epoch": 2.57085020242915, + "grad_norm": 0.0847047805638039, + "learning_rate": 2.2510031137720104e-05, + "loss": 0.0013, + "step": 635 + }, + { + "epoch": 2.574898785425101, + "grad_norm": 0.07007838160269851, + "learning_rate": 2.2453914005356206e-05, + "loss": 0.0009, + "step": 636 + }, + { + "epoch": 2.5789473684210527, + "grad_norm": 0.07430543590561914, + "learning_rate": 2.2397777251561302e-05, + "loss": 0.0012, + "step": 637 + }, + { + "epoch": 2.582995951417004, + "grad_norm": 0.08135607477401885, + "learning_rate": 2.2341621325203393e-05, + "loss": 0.0011, + "step": 638 + }, + { + "epoch": 2.5870445344129553, + "grad_norm": 0.08039343183287319, + "learning_rate": 2.2285446675303764e-05, + "loss": 0.001, + "step": 639 + }, + { + "epoch": 2.591093117408907, + "grad_norm": 0.07097900033248751, + "learning_rate": 2.2229253751033423e-05, + "loss": 0.0011, + "step": 640 + }, + { + "epoch": 2.5951417004048585, + "grad_norm": 0.05951126088803695, + "learning_rate": 2.2173043001709492e-05, + "loss": 0.0007, + "step": 641 + }, + { + "epoch": 2.5991902834008096, + "grad_norm": 0.07432404451496874, + "learning_rate": 2.2116814876791627e-05, + "loss": 0.0011, + "step": 642 + }, + { + "epoch": 2.603238866396761, + "grad_norm": 0.06894117355691685, + "learning_rate": 2.2060569825878413e-05, + "loss": 0.001, + "step": 643 + }, + { + "epoch": 2.6072874493927127, + "grad_norm": 0.061021139705098884, + "learning_rate": 2.2004308298703777e-05, + "loss": 0.0007, + "step": 644 + }, + { + "epoch": 2.611336032388664, + "grad_norm": 0.06576287634929692, + "learning_rate": 2.1948030745133396e-05, + "loss": 0.0008, + "step": 645 + }, + { + "epoch": 2.6153846153846154, + "grad_norm": 0.07694280538824322, + "learning_rate": 2.1891737615161083e-05, + "loss": 0.001, + "step": 646 + }, + { + "epoch": 2.619433198380567, + "grad_norm": 0.0700513243645308, + "learning_rate": 2.1835429358905218e-05, + "loss": 0.0009, + "step": 647 + }, + { + "epoch": 2.623481781376518, + "grad_norm": 0.06100624679122712, + "learning_rate": 2.17791064266051e-05, + "loss": 0.0007, + "step": 648 + }, + { + "epoch": 2.6275303643724697, + "grad_norm": 0.07110726640429209, + "learning_rate": 2.172276926861741e-05, + "loss": 0.0009, + "step": 649 + }, + { + "epoch": 2.6315789473684212, + "grad_norm": 0.06693379338797115, + "learning_rate": 2.166641833541255e-05, + "loss": 0.001, + "step": 650 + }, + { + "epoch": 2.6356275303643724, + "grad_norm": 0.05809911901011359, + "learning_rate": 2.161005407757109e-05, + "loss": 0.0007, + "step": 651 + }, + { + "epoch": 2.639676113360324, + "grad_norm": 0.07068377233425482, + "learning_rate": 2.155367694578013e-05, + "loss": 0.0011, + "step": 652 + }, + { + "epoch": 2.6437246963562755, + "grad_norm": 0.06316453120470827, + "learning_rate": 2.1497287390829712e-05, + "loss": 0.0008, + "step": 653 + }, + { + "epoch": 2.6477732793522266, + "grad_norm": 0.07046243631798743, + "learning_rate": 2.1440885863609208e-05, + "loss": 0.0008, + "step": 654 + }, + { + "epoch": 2.651821862348178, + "grad_norm": 0.07389827304433455, + "learning_rate": 2.1384472815103737e-05, + "loss": 0.001, + "step": 655 + }, + { + "epoch": 2.6558704453441297, + "grad_norm": 0.06947813466828594, + "learning_rate": 2.1328048696390534e-05, + "loss": 0.0009, + "step": 656 + }, + { + "epoch": 2.659919028340081, + "grad_norm": 0.056720514214845925, + "learning_rate": 2.1271613958635342e-05, + "loss": 0.0007, + "step": 657 + }, + { + "epoch": 2.6639676113360324, + "grad_norm": 0.0630323147689184, + "learning_rate": 2.1215169053088825e-05, + "loss": 0.0008, + "step": 658 + }, + { + "epoch": 2.668016194331984, + "grad_norm": 0.06680914058671518, + "learning_rate": 2.1158714431082937e-05, + "loss": 0.001, + "step": 659 + }, + { + "epoch": 2.672064777327935, + "grad_norm": 0.06477428368886747, + "learning_rate": 2.110225054402733e-05, + "loss": 0.0008, + "step": 660 + }, + { + "epoch": 2.6761133603238867, + "grad_norm": 0.05729080220979301, + "learning_rate": 2.1045777843405747e-05, + "loss": 0.0006, + "step": 661 + }, + { + "epoch": 2.6801619433198383, + "grad_norm": 0.059869954929180985, + "learning_rate": 2.09892967807724e-05, + "loss": 0.0006, + "step": 662 + }, + { + "epoch": 2.6842105263157894, + "grad_norm": 0.09371013843040668, + "learning_rate": 2.093280780774835e-05, + "loss": 0.001, + "step": 663 + }, + { + "epoch": 2.688259109311741, + "grad_norm": 0.06673844257275008, + "learning_rate": 2.087631137601793e-05, + "loss": 0.0009, + "step": 664 + }, + { + "epoch": 2.6923076923076925, + "grad_norm": 0.05633371057992201, + "learning_rate": 2.0819807937325104e-05, + "loss": 0.0005, + "step": 665 + }, + { + "epoch": 2.6963562753036436, + "grad_norm": 0.06515076780641861, + "learning_rate": 2.0763297943469847e-05, + "loss": 0.0007, + "step": 666 + }, + { + "epoch": 2.700404858299595, + "grad_norm": 0.0548423445954722, + "learning_rate": 2.070678184630458e-05, + "loss": 0.0008, + "step": 667 + }, + { + "epoch": 2.7044534412955468, + "grad_norm": 0.05720546618000139, + "learning_rate": 2.06502600977305e-05, + "loss": 0.0007, + "step": 668 + }, + { + "epoch": 2.708502024291498, + "grad_norm": 0.05560020309598184, + "learning_rate": 2.0593733149694008e-05, + "loss": 0.0008, + "step": 669 + }, + { + "epoch": 2.7125506072874495, + "grad_norm": 0.05751194996185061, + "learning_rate": 2.0537201454183074e-05, + "loss": 0.0006, + "step": 670 + }, + { + "epoch": 2.716599190283401, + "grad_norm": 0.05387616849692982, + "learning_rate": 2.0480665463223626e-05, + "loss": 0.0005, + "step": 671 + }, + { + "epoch": 2.720647773279352, + "grad_norm": 0.06226423981126559, + "learning_rate": 2.042412562887594e-05, + "loss": 0.001, + "step": 672 + }, + { + "epoch": 2.7246963562753037, + "grad_norm": 0.0603298895783454, + "learning_rate": 2.0367582403231033e-05, + "loss": 0.0008, + "step": 673 + }, + { + "epoch": 2.7287449392712553, + "grad_norm": 0.05574973622922628, + "learning_rate": 2.0311036238407023e-05, + "loss": 0.0007, + "step": 674 + }, + { + "epoch": 2.7327935222672064, + "grad_norm": 0.058383440162777744, + "learning_rate": 2.0254487586545534e-05, + "loss": 0.0008, + "step": 675 + }, + { + "epoch": 2.736842105263158, + "grad_norm": 0.07073527570790639, + "learning_rate": 2.0197936899808088e-05, + "loss": 0.0006, + "step": 676 + }, + { + "epoch": 2.7408906882591095, + "grad_norm": 0.06097870479444593, + "learning_rate": 2.0141384630372466e-05, + "loss": 0.0006, + "step": 677 + }, + { + "epoch": 2.7449392712550607, + "grad_norm": 0.04942530158130887, + "learning_rate": 2.0084831230429106e-05, + "loss": 0.0005, + "step": 678 + }, + { + "epoch": 2.748987854251012, + "grad_norm": 0.05729339002857787, + "learning_rate": 2.0028277152177492e-05, + "loss": 0.0005, + "step": 679 + }, + { + "epoch": 2.753036437246964, + "grad_norm": 0.053546995871736384, + "learning_rate": 1.9971722847822518e-05, + "loss": 0.0006, + "step": 680 + }, + { + "epoch": 2.757085020242915, + "grad_norm": 0.0510395524444078, + "learning_rate": 1.99151687695709e-05, + "loss": 0.0005, + "step": 681 + }, + { + "epoch": 2.7611336032388665, + "grad_norm": 0.04982199470214237, + "learning_rate": 1.9858615369627537e-05, + "loss": 0.0005, + "step": 682 + }, + { + "epoch": 2.765182186234818, + "grad_norm": 0.05271399757198629, + "learning_rate": 1.9802063100191916e-05, + "loss": 0.0007, + "step": 683 + }, + { + "epoch": 2.769230769230769, + "grad_norm": 0.0514100027209595, + "learning_rate": 1.974551241345447e-05, + "loss": 0.0005, + "step": 684 + }, + { + "epoch": 2.7732793522267207, + "grad_norm": 0.04321561839871366, + "learning_rate": 1.9688963761592984e-05, + "loss": 0.0003, + "step": 685 + }, + { + "epoch": 2.7773279352226723, + "grad_norm": 0.05169485117339652, + "learning_rate": 1.963241759676897e-05, + "loss": 0.0004, + "step": 686 + }, + { + "epoch": 2.7813765182186234, + "grad_norm": 0.0534758760126439, + "learning_rate": 1.9575874371124066e-05, + "loss": 0.0005, + "step": 687 + }, + { + "epoch": 2.785425101214575, + "grad_norm": 0.04673840719962971, + "learning_rate": 1.951933453677638e-05, + "loss": 0.0004, + "step": 688 + }, + { + "epoch": 2.7894736842105265, + "grad_norm": 0.05476661255796678, + "learning_rate": 1.9462798545816932e-05, + "loss": 0.0005, + "step": 689 + }, + { + "epoch": 2.7935222672064777, + "grad_norm": 0.05043267774592582, + "learning_rate": 1.9406266850305996e-05, + "loss": 0.0004, + "step": 690 + }, + { + "epoch": 2.7975708502024292, + "grad_norm": 0.05485643972318966, + "learning_rate": 1.9349739902269503e-05, + "loss": 0.0006, + "step": 691 + }, + { + "epoch": 2.801619433198381, + "grad_norm": 0.04448496887263086, + "learning_rate": 1.9293218153695425e-05, + "loss": 0.0004, + "step": 692 + }, + { + "epoch": 2.805668016194332, + "grad_norm": 0.052077658267235004, + "learning_rate": 1.9236702056530153e-05, + "loss": 0.0005, + "step": 693 + }, + { + "epoch": 2.8097165991902835, + "grad_norm": 0.051401591650914825, + "learning_rate": 1.9180192062674907e-05, + "loss": 0.0004, + "step": 694 + }, + { + "epoch": 2.813765182186235, + "grad_norm": 0.039396691214105335, + "learning_rate": 1.9123688623982076e-05, + "loss": 0.0005, + "step": 695 + }, + { + "epoch": 2.817813765182186, + "grad_norm": 0.03938894604509275, + "learning_rate": 1.906719219225165e-05, + "loss": 0.0005, + "step": 696 + }, + { + "epoch": 2.8218623481781377, + "grad_norm": 0.05811062566878592, + "learning_rate": 1.9010703219227603e-05, + "loss": 0.0006, + "step": 697 + }, + { + "epoch": 2.8259109311740893, + "grad_norm": 0.0560014964877496, + "learning_rate": 1.8954222156594256e-05, + "loss": 0.0006, + "step": 698 + }, + { + "epoch": 2.8299595141700404, + "grad_norm": 0.04548212472813346, + "learning_rate": 1.8897749455972673e-05, + "loss": 0.0005, + "step": 699 + }, + { + "epoch": 2.834008097165992, + "grad_norm": 0.05218026614504433, + "learning_rate": 1.884128556891707e-05, + "loss": 0.0007, + "step": 700 + }, + { + "epoch": 2.8380566801619436, + "grad_norm": 0.03687787569742056, + "learning_rate": 1.878483094691119e-05, + "loss": 0.0003, + "step": 701 + }, + { + "epoch": 2.8421052631578947, + "grad_norm": 0.035740013187986586, + "learning_rate": 1.8728386041364664e-05, + "loss": 0.0004, + "step": 702 + }, + { + "epoch": 2.8461538461538463, + "grad_norm": 0.039987972569202536, + "learning_rate": 1.867195130360947e-05, + "loss": 0.0004, + "step": 703 + }, + { + "epoch": 2.850202429149798, + "grad_norm": 0.04660766516993981, + "learning_rate": 1.8615527184896263e-05, + "loss": 0.0004, + "step": 704 + }, + { + "epoch": 2.854251012145749, + "grad_norm": 0.042519975294850375, + "learning_rate": 1.8559114136390795e-05, + "loss": 0.0004, + "step": 705 + }, + { + "epoch": 2.8582995951417005, + "grad_norm": 0.05558406915782309, + "learning_rate": 1.8502712609170298e-05, + "loss": 0.0004, + "step": 706 + }, + { + "epoch": 2.862348178137652, + "grad_norm": 0.045517677064072885, + "learning_rate": 1.8446323054219876e-05, + "loss": 0.0005, + "step": 707 + }, + { + "epoch": 2.866396761133603, + "grad_norm": 0.04728774528837013, + "learning_rate": 1.838994592242891e-05, + "loss": 0.0003, + "step": 708 + }, + { + "epoch": 2.8704453441295548, + "grad_norm": 0.039458844375343825, + "learning_rate": 1.8333581664587453e-05, + "loss": 0.0003, + "step": 709 + }, + { + "epoch": 2.8744939271255063, + "grad_norm": 0.05494191012823419, + "learning_rate": 1.8277230731382593e-05, + "loss": 0.0004, + "step": 710 + }, + { + "epoch": 2.8785425101214575, + "grad_norm": 0.039792831749577574, + "learning_rate": 1.82208935733949e-05, + "loss": 0.0003, + "step": 711 + }, + { + "epoch": 2.882591093117409, + "grad_norm": 0.04195917024301811, + "learning_rate": 1.8164570641094793e-05, + "loss": 0.0004, + "step": 712 + }, + { + "epoch": 2.8866396761133606, + "grad_norm": 0.04655243166218935, + "learning_rate": 1.810826238483892e-05, + "loss": 0.0004, + "step": 713 + }, + { + "epoch": 2.8906882591093117, + "grad_norm": 0.031122522694084546, + "learning_rate": 1.8051969254866608e-05, + "loss": 0.0003, + "step": 714 + }, + { + "epoch": 2.8947368421052633, + "grad_norm": 0.043101880789527786, + "learning_rate": 1.7995691701296226e-05, + "loss": 0.0005, + "step": 715 + }, + { + "epoch": 2.898785425101215, + "grad_norm": 0.04493229210509924, + "learning_rate": 1.7939430174121593e-05, + "loss": 0.0003, + "step": 716 + }, + { + "epoch": 2.902834008097166, + "grad_norm": 0.03463859487966806, + "learning_rate": 1.788318512320838e-05, + "loss": 0.0002, + "step": 717 + }, + { + "epoch": 2.9068825910931175, + "grad_norm": 0.05095643828062113, + "learning_rate": 1.782695699829051e-05, + "loss": 0.0005, + "step": 718 + }, + { + "epoch": 2.910931174089069, + "grad_norm": 0.07246027859718954, + "learning_rate": 1.7770746248966587e-05, + "loss": 0.0004, + "step": 719 + }, + { + "epoch": 2.91497975708502, + "grad_norm": 0.04162235934145595, + "learning_rate": 1.7714553324696243e-05, + "loss": 0.0003, + "step": 720 + }, + { + "epoch": 2.919028340080972, + "grad_norm": 0.033793795364664536, + "learning_rate": 1.7658378674796614e-05, + "loss": 0.0003, + "step": 721 + }, + { + "epoch": 2.9230769230769234, + "grad_norm": 0.034837675707095145, + "learning_rate": 1.76022227484387e-05, + "loss": 0.0003, + "step": 722 + }, + { + "epoch": 2.9271255060728745, + "grad_norm": 0.045156168632629154, + "learning_rate": 1.7546085994643807e-05, + "loss": 0.0003, + "step": 723 + }, + { + "epoch": 2.931174089068826, + "grad_norm": 0.047497999389383296, + "learning_rate": 1.7489968862279902e-05, + "loss": 0.0003, + "step": 724 + }, + { + "epoch": 2.9352226720647776, + "grad_norm": 0.031214224208327673, + "learning_rate": 1.743387180005809e-05, + "loss": 0.0002, + "step": 725 + }, + { + "epoch": 2.9392712550607287, + "grad_norm": 0.0330838713234608, + "learning_rate": 1.737779525652899e-05, + "loss": 0.0003, + "step": 726 + }, + { + "epoch": 2.9433198380566803, + "grad_norm": 0.04127973277616298, + "learning_rate": 1.732173968007914e-05, + "loss": 0.0003, + "step": 727 + }, + { + "epoch": 2.9473684210526314, + "grad_norm": 0.04179063762604572, + "learning_rate": 1.7265705518927437e-05, + "loss": 0.0003, + "step": 728 + }, + { + "epoch": 2.951417004048583, + "grad_norm": 0.03225284000292102, + "learning_rate": 1.7209693221121542e-05, + "loss": 0.0003, + "step": 729 + }, + { + "epoch": 2.9554655870445345, + "grad_norm": 0.06206122761420357, + "learning_rate": 1.7153703234534302e-05, + "loss": 0.0005, + "step": 730 + }, + { + "epoch": 2.9595141700404857, + "grad_norm": 0.028232972447193885, + "learning_rate": 1.709773600686015e-05, + "loss": 0.0002, + "step": 731 + }, + { + "epoch": 2.9635627530364372, + "grad_norm": 0.03886425438941887, + "learning_rate": 1.7041791985611548e-05, + "loss": 0.0003, + "step": 732 + }, + { + "epoch": 2.967611336032389, + "grad_norm": 0.04462192561912251, + "learning_rate": 1.69858716181154e-05, + "loss": 0.0003, + "step": 733 + }, + { + "epoch": 2.97165991902834, + "grad_norm": 0.038759428975961484, + "learning_rate": 1.692997535150948e-05, + "loss": 0.0003, + "step": 734 + }, + { + "epoch": 2.9757085020242915, + "grad_norm": 0.03718952789579252, + "learning_rate": 1.687410363273884e-05, + "loss": 0.0003, + "step": 735 + }, + { + "epoch": 2.979757085020243, + "grad_norm": 0.0371117102106491, + "learning_rate": 1.6818256908552257e-05, + "loss": 0.0002, + "step": 736 + }, + { + "epoch": 2.983805668016194, + "grad_norm": 0.03976876793538474, + "learning_rate": 1.6762435625498646e-05, + "loss": 0.0003, + "step": 737 + }, + { + "epoch": 2.9878542510121457, + "grad_norm": 0.039398714419491804, + "learning_rate": 1.67066402299235e-05, + "loss": 0.0004, + "step": 738 + }, + { + "epoch": 2.9919028340080973, + "grad_norm": 0.03146414217036324, + "learning_rate": 1.6650871167965313e-05, + "loss": 0.0002, + "step": 739 + }, + { + "epoch": 2.9959514170040484, + "grad_norm": 0.029653095877203765, + "learning_rate": 1.6595128885552028e-05, + "loss": 0.0003, + "step": 740 + }, + { + "epoch": 3.0, + "grad_norm": 0.03191615868648229, + "learning_rate": 1.653941382839745e-05, + "loss": 0.0003, + "step": 741 + }, + { + "epoch": 3.0040485829959516, + "grad_norm": 0.029145094354245105, + "learning_rate": 1.6483726441997673e-05, + "loss": 0.0002, + "step": 742 + }, + { + "epoch": 3.0080971659919027, + "grad_norm": 0.022588341951087082, + "learning_rate": 1.642806717162757e-05, + "loss": 0.0001, + "step": 743 + }, + { + "epoch": 3.0121457489878543, + "grad_norm": 0.03154009588762076, + "learning_rate": 1.637243646233718e-05, + "loss": 0.0002, + "step": 744 + }, + { + "epoch": 3.016194331983806, + "grad_norm": 0.03456803086070107, + "learning_rate": 1.6316834758948174e-05, + "loss": 0.0002, + "step": 745 + }, + { + "epoch": 3.020242914979757, + "grad_norm": 0.033017672758467446, + "learning_rate": 1.6261262506050282e-05, + "loss": 0.0002, + "step": 746 + }, + { + "epoch": 3.0242914979757085, + "grad_norm": 0.026253542100348325, + "learning_rate": 1.620572014799777e-05, + "loss": 0.0002, + "step": 747 + }, + { + "epoch": 3.02834008097166, + "grad_norm": 0.03666197505192881, + "learning_rate": 1.6150208128905857e-05, + "loss": 0.0003, + "step": 748 + }, + { + "epoch": 3.032388663967611, + "grad_norm": 0.02772349999488188, + "learning_rate": 1.6094726892647147e-05, + "loss": 0.0001, + "step": 749 + }, + { + "epoch": 3.0364372469635628, + "grad_norm": 0.025312439741653148, + "learning_rate": 1.6039276882848135e-05, + "loss": 0.0002, + "step": 750 + }, + { + "epoch": 3.0404858299595143, + "grad_norm": 0.03693531581905942, + "learning_rate": 1.5983858542885635e-05, + "loss": 0.0002, + "step": 751 + }, + { + "epoch": 3.0445344129554655, + "grad_norm": 0.026844044599295562, + "learning_rate": 1.5928472315883203e-05, + "loss": 0.0003, + "step": 752 + }, + { + "epoch": 3.048582995951417, + "grad_norm": 0.04786462136175038, + "learning_rate": 1.5873118644707633e-05, + "loss": 0.0002, + "step": 753 + }, + { + "epoch": 3.0526315789473686, + "grad_norm": 0.030598730799439998, + "learning_rate": 1.5817797971965413e-05, + "loss": 0.0002, + "step": 754 + }, + { + "epoch": 3.0566801619433197, + "grad_norm": 0.014953540820073533, + "learning_rate": 1.576251073999917e-05, + "loss": 0.0001, + "step": 755 + }, + { + "epoch": 3.0607287449392713, + "grad_norm": 0.026856669218570457, + "learning_rate": 1.5707257390884126e-05, + "loss": 0.0002, + "step": 756 + }, + { + "epoch": 3.064777327935223, + "grad_norm": 0.019064758193315646, + "learning_rate": 1.5652038366424595e-05, + "loss": 0.0002, + "step": 757 + }, + { + "epoch": 3.068825910931174, + "grad_norm": 0.02743259927401312, + "learning_rate": 1.5596854108150424e-05, + "loss": 0.0002, + "step": 758 + }, + { + "epoch": 3.0728744939271255, + "grad_norm": 0.027447031448971933, + "learning_rate": 1.5541705057313476e-05, + "loss": 0.0002, + "step": 759 + }, + { + "epoch": 3.076923076923077, + "grad_norm": 0.023639478317692482, + "learning_rate": 1.5486591654884086e-05, + "loss": 0.0002, + "step": 760 + }, + { + "epoch": 3.080971659919028, + "grad_norm": 0.03190276085616586, + "learning_rate": 1.543151434154755e-05, + "loss": 0.0002, + "step": 761 + }, + { + "epoch": 3.08502024291498, + "grad_norm": 0.030696828597046148, + "learning_rate": 1.5376473557700615e-05, + "loss": 0.0002, + "step": 762 + }, + { + "epoch": 3.0890688259109313, + "grad_norm": 0.029209196533392608, + "learning_rate": 1.5321469743447893e-05, + "loss": 0.0002, + "step": 763 + }, + { + "epoch": 3.0931174089068825, + "grad_norm": 0.033282858619750194, + "learning_rate": 1.5266503338598434e-05, + "loss": 0.0002, + "step": 764 + }, + { + "epoch": 3.097165991902834, + "grad_norm": 0.01132358672010993, + "learning_rate": 1.5211574782662149e-05, + "loss": 0.0001, + "step": 765 + }, + { + "epoch": 3.1012145748987856, + "grad_norm": 0.023486895056680623, + "learning_rate": 1.5156684514846307e-05, + "loss": 0.0001, + "step": 766 + }, + { + "epoch": 3.1052631578947367, + "grad_norm": 0.022923193281233982, + "learning_rate": 1.5101832974052012e-05, + "loss": 0.0001, + "step": 767 + }, + { + "epoch": 3.1093117408906883, + "grad_norm": 0.019645492393430032, + "learning_rate": 1.5047020598870737e-05, + "loss": 0.0002, + "step": 768 + }, + { + "epoch": 3.11336032388664, + "grad_norm": 0.016599997429465527, + "learning_rate": 1.4992247827580778e-05, + "loss": 0.0001, + "step": 769 + }, + { + "epoch": 3.117408906882591, + "grad_norm": 0.024704408717812928, + "learning_rate": 1.493751509814374e-05, + "loss": 0.0002, + "step": 770 + }, + { + "epoch": 3.1214574898785425, + "grad_norm": 0.019401719027300538, + "learning_rate": 1.4882822848201073e-05, + "loss": 0.0002, + "step": 771 + }, + { + "epoch": 3.125506072874494, + "grad_norm": 0.027152293754764165, + "learning_rate": 1.4828171515070553e-05, + "loss": 0.0002, + "step": 772 + }, + { + "epoch": 3.1295546558704452, + "grad_norm": 0.038468604263252625, + "learning_rate": 1.4773561535742793e-05, + "loss": 0.0002, + "step": 773 + }, + { + "epoch": 3.133603238866397, + "grad_norm": 0.026624436919316416, + "learning_rate": 1.4718993346877715e-05, + "loss": 0.0001, + "step": 774 + }, + { + "epoch": 3.1376518218623484, + "grad_norm": 0.019460828328861493, + "learning_rate": 1.466446738480111e-05, + "loss": 0.0001, + "step": 775 + }, + { + "epoch": 3.1417004048582995, + "grad_norm": 0.016846735666260743, + "learning_rate": 1.460998408550112e-05, + "loss": 0.0001, + "step": 776 + }, + { + "epoch": 3.145748987854251, + "grad_norm": 0.03283037822194332, + "learning_rate": 1.4555543884624751e-05, + "loss": 0.0002, + "step": 777 + }, + { + "epoch": 3.1497975708502026, + "grad_norm": 0.02608131590273961, + "learning_rate": 1.4501147217474402e-05, + "loss": 0.0002, + "step": 778 + }, + { + "epoch": 3.1538461538461537, + "grad_norm": 0.023016721641962137, + "learning_rate": 1.444679451900437e-05, + "loss": 0.0001, + "step": 779 + }, + { + "epoch": 3.1578947368421053, + "grad_norm": 0.01891574010782884, + "learning_rate": 1.4392486223817397e-05, + "loss": 0.0001, + "step": 780 + }, + { + "epoch": 3.161943319838057, + "grad_norm": 0.020209777721135965, + "learning_rate": 1.4338222766161143e-05, + "loss": 0.0001, + "step": 781 + }, + { + "epoch": 3.165991902834008, + "grad_norm": 0.007235903793200024, + "learning_rate": 1.4284004579924777e-05, + "loss": 0.0001, + "step": 782 + }, + { + "epoch": 3.1700404858299596, + "grad_norm": 0.024930561992524393, + "learning_rate": 1.4229832098635479e-05, + "loss": 0.0001, + "step": 783 + }, + { + "epoch": 3.174089068825911, + "grad_norm": 0.015896570182219447, + "learning_rate": 1.4175705755454963e-05, + "loss": 0.0001, + "step": 784 + }, + { + "epoch": 3.1781376518218623, + "grad_norm": 0.023101164411084295, + "learning_rate": 1.412162598317602e-05, + "loss": 0.0001, + "step": 785 + }, + { + "epoch": 3.182186234817814, + "grad_norm": 0.02091024949015636, + "learning_rate": 1.406759321421907e-05, + "loss": 0.0002, + "step": 786 + }, + { + "epoch": 3.1862348178137654, + "grad_norm": 0.021108696194686964, + "learning_rate": 1.401360788062871e-05, + "loss": 0.0001, + "step": 787 + }, + { + "epoch": 3.1902834008097165, + "grad_norm": 0.0249631082383126, + "learning_rate": 1.3959670414070204e-05, + "loss": 0.0001, + "step": 788 + }, + { + "epoch": 3.194331983805668, + "grad_norm": 0.01536335665698139, + "learning_rate": 1.3905781245826108e-05, + "loss": 0.0001, + "step": 789 + }, + { + "epoch": 3.1983805668016196, + "grad_norm": 0.01954161072801277, + "learning_rate": 1.3851940806792778e-05, + "loss": 0.0001, + "step": 790 + }, + { + "epoch": 3.2024291497975708, + "grad_norm": 0.02786149371981257, + "learning_rate": 1.379814952747693e-05, + "loss": 0.0001, + "step": 791 + }, + { + "epoch": 3.2064777327935223, + "grad_norm": 0.015114582083527058, + "learning_rate": 1.3744407837992193e-05, + "loss": 0.0001, + "step": 792 + }, + { + "epoch": 3.2105263157894735, + "grad_norm": 0.009911341432700756, + "learning_rate": 1.3690716168055692e-05, + "loss": 0.0001, + "step": 793 + }, + { + "epoch": 3.214574898785425, + "grad_norm": 0.014632644284844627, + "learning_rate": 1.363707494698459e-05, + "loss": 0.0001, + "step": 794 + }, + { + "epoch": 3.2186234817813766, + "grad_norm": 0.009944633866281062, + "learning_rate": 1.358348460369265e-05, + "loss": 0.0001, + "step": 795 + }, + { + "epoch": 3.2226720647773277, + "grad_norm": 0.020862256477688308, + "learning_rate": 1.352994556668684e-05, + "loss": 0.0001, + "step": 796 + }, + { + "epoch": 3.2267206477732793, + "grad_norm": 0.019336736961453417, + "learning_rate": 1.3476458264063865e-05, + "loss": 0.0001, + "step": 797 + }, + { + "epoch": 3.230769230769231, + "grad_norm": 0.005840017752526392, + "learning_rate": 1.342302312350678e-05, + "loss": 0.0001, + "step": 798 + }, + { + "epoch": 3.234817813765182, + "grad_norm": 0.0074310197093575515, + "learning_rate": 1.3369640572281537e-05, + "loss": 0.0001, + "step": 799 + }, + { + "epoch": 3.2388663967611335, + "grad_norm": 0.029987542745135393, + "learning_rate": 1.3316311037233596e-05, + "loss": 0.0002, + "step": 800 + }, + { + "epoch": 3.242914979757085, + "grad_norm": 0.016152574623853146, + "learning_rate": 1.326303494478451e-05, + "loss": 0.0001, + "step": 801 + }, + { + "epoch": 3.246963562753036, + "grad_norm": 0.007367062459369656, + "learning_rate": 1.3209812720928472e-05, + "loss": 0.0001, + "step": 802 + }, + { + "epoch": 3.251012145748988, + "grad_norm": 0.017808867816428823, + "learning_rate": 1.315664479122898e-05, + "loss": 0.0001, + "step": 803 + }, + { + "epoch": 3.2550607287449393, + "grad_norm": 0.01874141053005802, + "learning_rate": 1.3103531580815378e-05, + "loss": 0.0001, + "step": 804 + }, + { + "epoch": 3.2591093117408905, + "grad_norm": 0.028331207682845697, + "learning_rate": 1.3050473514379488e-05, + "loss": 0.0001, + "step": 805 + }, + { + "epoch": 3.263157894736842, + "grad_norm": 0.011835992035030073, + "learning_rate": 1.2997471016172188e-05, + "loss": 0.0001, + "step": 806 + }, + { + "epoch": 3.2672064777327936, + "grad_norm": 0.017823479644649885, + "learning_rate": 1.2944524510000042e-05, + "loss": 0.0001, + "step": 807 + }, + { + "epoch": 3.2712550607287447, + "grad_norm": 0.015124603702314601, + "learning_rate": 1.2891634419221908e-05, + "loss": 0.0001, + "step": 808 + }, + { + "epoch": 3.2753036437246963, + "grad_norm": 0.016327745531159937, + "learning_rate": 1.2838801166745545e-05, + "loss": 0.0001, + "step": 809 + }, + { + "epoch": 3.279352226720648, + "grad_norm": 0.009730140100698134, + "learning_rate": 1.2786025175024216e-05, + "loss": 0.0001, + "step": 810 + }, + { + "epoch": 3.283400809716599, + "grad_norm": 0.02162051895572787, + "learning_rate": 1.2733306866053357e-05, + "loss": 0.0001, + "step": 811 + }, + { + "epoch": 3.2874493927125505, + "grad_norm": 0.0047492364902748255, + "learning_rate": 1.2680646661367163e-05, + "loss": 0.0001, + "step": 812 + }, + { + "epoch": 3.291497975708502, + "grad_norm": 0.010427659985501521, + "learning_rate": 1.2628044982035219e-05, + "loss": 0.0001, + "step": 813 + }, + { + "epoch": 3.2955465587044532, + "grad_norm": 0.008034673034913889, + "learning_rate": 1.2575502248659155e-05, + "loss": 0.0001, + "step": 814 + }, + { + "epoch": 3.299595141700405, + "grad_norm": 0.008077441497510543, + "learning_rate": 1.2523018881369269e-05, + "loss": 0.0001, + "step": 815 + }, + { + "epoch": 3.3036437246963564, + "grad_norm": 0.008965949290469007, + "learning_rate": 1.2470595299821176e-05, + "loss": 0.0001, + "step": 816 + }, + { + "epoch": 3.3076923076923075, + "grad_norm": 0.014181836014158377, + "learning_rate": 1.2418231923192416e-05, + "loss": 0.0001, + "step": 817 + }, + { + "epoch": 3.311740890688259, + "grad_norm": 0.011646254059374786, + "learning_rate": 1.2365929170179165e-05, + "loss": 0.0001, + "step": 818 + }, + { + "epoch": 3.3157894736842106, + "grad_norm": 0.009953507193745305, + "learning_rate": 1.2313687458992848e-05, + "loss": 0.0001, + "step": 819 + }, + { + "epoch": 3.3198380566801617, + "grad_norm": 0.01899881498027721, + "learning_rate": 1.2261507207356785e-05, + "loss": 0.0001, + "step": 820 + }, + { + "epoch": 3.3238866396761133, + "grad_norm": 0.012478991738353207, + "learning_rate": 1.220938883250288e-05, + "loss": 0.0001, + "step": 821 + }, + { + "epoch": 3.327935222672065, + "grad_norm": 0.009675874869392663, + "learning_rate": 1.215733275116827e-05, + "loss": 0.0001, + "step": 822 + }, + { + "epoch": 3.331983805668016, + "grad_norm": 0.0050667894240743255, + "learning_rate": 1.2105339379592005e-05, + "loss": 0.0001, + "step": 823 + }, + { + "epoch": 3.3360323886639676, + "grad_norm": 0.004237564006443495, + "learning_rate": 1.2053409133511681e-05, + "loss": 0.0001, + "step": 824 + }, + { + "epoch": 3.340080971659919, + "grad_norm": 0.01142206596287372, + "learning_rate": 1.2001542428160173e-05, + "loss": 0.0001, + "step": 825 + }, + { + "epoch": 3.3441295546558703, + "grad_norm": 0.03259612109779952, + "learning_rate": 1.1949739678262278e-05, + "loss": 0.0001, + "step": 826 + }, + { + "epoch": 3.348178137651822, + "grad_norm": 0.01604938306063877, + "learning_rate": 1.189800129803139e-05, + "loss": 0.0001, + "step": 827 + }, + { + "epoch": 3.3522267206477734, + "grad_norm": 0.012629951869883783, + "learning_rate": 1.1846327701166228e-05, + "loss": 0.0001, + "step": 828 + }, + { + "epoch": 3.3562753036437245, + "grad_norm": 0.021766961145668083, + "learning_rate": 1.1794719300847493e-05, + "loss": 0.0001, + "step": 829 + }, + { + "epoch": 3.360323886639676, + "grad_norm": 0.0034807423415103966, + "learning_rate": 1.1743176509734582e-05, + "loss": 0.0001, + "step": 830 + }, + { + "epoch": 3.3643724696356276, + "grad_norm": 0.012445912856943075, + "learning_rate": 1.1691699739962275e-05, + "loss": 0.0001, + "step": 831 + }, + { + "epoch": 3.3684210526315788, + "grad_norm": 0.0032686067243284027, + "learning_rate": 1.1640289403137456e-05, + "loss": 0.0001, + "step": 832 + }, + { + "epoch": 3.3724696356275303, + "grad_norm": 0.013661002631862767, + "learning_rate": 1.1588945910335802e-05, + "loss": 0.0001, + "step": 833 + }, + { + "epoch": 3.376518218623482, + "grad_norm": 0.011573628535845077, + "learning_rate": 1.1537669672098517e-05, + "loss": 0.0001, + "step": 834 + }, + { + "epoch": 3.380566801619433, + "grad_norm": 0.020753318775551453, + "learning_rate": 1.1486461098429036e-05, + "loss": 0.0001, + "step": 835 + }, + { + "epoch": 3.3846153846153846, + "grad_norm": 0.008091337148014606, + "learning_rate": 1.1435320598789747e-05, + "loss": 0.0001, + "step": 836 + }, + { + "epoch": 3.388663967611336, + "grad_norm": 0.006061378421656682, + "learning_rate": 1.1384248582098736e-05, + "loss": 0.0001, + "step": 837 + }, + { + "epoch": 3.3927125506072873, + "grad_norm": 0.009935215069627377, + "learning_rate": 1.1333245456726467e-05, + "loss": 0.0001, + "step": 838 + }, + { + "epoch": 3.396761133603239, + "grad_norm": 0.001994865202853058, + "learning_rate": 1.128231163049258e-05, + "loss": 0.0, + "step": 839 + }, + { + "epoch": 3.4008097165991904, + "grad_norm": 0.014511192324086038, + "learning_rate": 1.123144751066259e-05, + "loss": 0.0001, + "step": 840 + }, + { + "epoch": 3.4048582995951415, + "grad_norm": 0.0053136077468555155, + "learning_rate": 1.1180653503944657e-05, + "loss": 0.0, + "step": 841 + }, + { + "epoch": 3.408906882591093, + "grad_norm": 0.0038231079232079195, + "learning_rate": 1.1129930016486287e-05, + "loss": 0.0, + "step": 842 + }, + { + "epoch": 3.4129554655870447, + "grad_norm": 0.021306415566894818, + "learning_rate": 1.107927745387114e-05, + "loss": 0.0001, + "step": 843 + }, + { + "epoch": 3.417004048582996, + "grad_norm": 0.006568627243653601, + "learning_rate": 1.1028696221115766e-05, + "loss": 0.0, + "step": 844 + }, + { + "epoch": 3.4210526315789473, + "grad_norm": 0.011724337532036387, + "learning_rate": 1.0978186722666341e-05, + "loss": 0.0001, + "step": 845 + }, + { + "epoch": 3.425101214574899, + "grad_norm": 0.0026103708604204266, + "learning_rate": 1.0927749362395475e-05, + "loss": 0.0, + "step": 846 + }, + { + "epoch": 3.42914979757085, + "grad_norm": 0.009654396636165144, + "learning_rate": 1.0877384543598954e-05, + "loss": 0.0001, + "step": 847 + }, + { + "epoch": 3.4331983805668016, + "grad_norm": 0.004572283519089853, + "learning_rate": 1.0827092668992529e-05, + "loss": 0.0, + "step": 848 + }, + { + "epoch": 3.437246963562753, + "grad_norm": 0.0020795434226904344, + "learning_rate": 1.077687414070869e-05, + "loss": 0.0, + "step": 849 + }, + { + "epoch": 3.4412955465587043, + "grad_norm": 0.008174951323464262, + "learning_rate": 1.0726729360293451e-05, + "loss": 0.0001, + "step": 850 + }, + { + "epoch": 3.445344129554656, + "grad_norm": 0.0013501628718344085, + "learning_rate": 1.0676658728703147e-05, + "loss": 0.0, + "step": 851 + }, + { + "epoch": 3.4493927125506074, + "grad_norm": 0.0073486530660730074, + "learning_rate": 1.0626662646301198e-05, + "loss": 0.0001, + "step": 852 + }, + { + "epoch": 3.4534412955465585, + "grad_norm": 0.004987191177217941, + "learning_rate": 1.0576741512854958e-05, + "loss": 0.0, + "step": 853 + }, + { + "epoch": 3.45748987854251, + "grad_norm": 0.002618845132450121, + "learning_rate": 1.052689572753247e-05, + "loss": 0.0, + "step": 854 + }, + { + "epoch": 3.4615384615384617, + "grad_norm": 0.007243804748935872, + "learning_rate": 1.0477125688899328e-05, + "loss": 0.0, + "step": 855 + }, + { + "epoch": 3.465587044534413, + "grad_norm": 0.010131241966742614, + "learning_rate": 1.0427431794915406e-05, + "loss": 0.0001, + "step": 856 + }, + { + "epoch": 3.4696356275303644, + "grad_norm": 0.0012455865366024698, + "learning_rate": 1.0377814442931769e-05, + "loss": 0.0, + "step": 857 + }, + { + "epoch": 3.473684210526316, + "grad_norm": 0.007092997476844224, + "learning_rate": 1.0328274029687439e-05, + "loss": 0.0, + "step": 858 + }, + { + "epoch": 3.477732793522267, + "grad_norm": 0.0036468733299997954, + "learning_rate": 1.0278810951306253e-05, + "loss": 0.0, + "step": 859 + }, + { + "epoch": 3.4817813765182186, + "grad_norm": 0.0026558468340004333, + "learning_rate": 1.0229425603293647e-05, + "loss": 0.0, + "step": 860 + }, + { + "epoch": 3.48582995951417, + "grad_norm": 0.00642635620778787, + "learning_rate": 1.0180118380533558e-05, + "loss": 0.0001, + "step": 861 + }, + { + "epoch": 3.4898785425101213, + "grad_norm": 0.008382035850832038, + "learning_rate": 1.0130889677285225e-05, + "loss": 0.0, + "step": 862 + }, + { + "epoch": 3.493927125506073, + "grad_norm": 0.003341298529512435, + "learning_rate": 1.008173988718005e-05, + "loss": 0.0, + "step": 863 + }, + { + "epoch": 3.4979757085020244, + "grad_norm": 0.0013780543215555522, + "learning_rate": 1.0032669403218439e-05, + "loss": 0.0, + "step": 864 + }, + { + "epoch": 3.5020242914979756, + "grad_norm": 0.006697475959674273, + "learning_rate": 9.983678617766677e-06, + "loss": 0.0001, + "step": 865 + }, + { + "epoch": 3.506072874493927, + "grad_norm": 0.006696102639476047, + "learning_rate": 9.934767922553788e-06, + "loss": 0.0, + "step": 866 + }, + { + "epoch": 3.5101214574898787, + "grad_norm": 0.001083521355886679, + "learning_rate": 9.885937708668371e-06, + "loss": 0.0, + "step": 867 + }, + { + "epoch": 3.51417004048583, + "grad_norm": 0.0015141904969095373, + "learning_rate": 9.837188366555524e-06, + "loss": 0.0, + "step": 868 + }, + { + "epoch": 3.5182186234817814, + "grad_norm": 0.0030752457460708487, + "learning_rate": 9.788520286013691e-06, + "loss": 0.0001, + "step": 869 + }, + { + "epoch": 3.522267206477733, + "grad_norm": 0.0010808705364827224, + "learning_rate": 9.739933856191537e-06, + "loss": 0.0, + "step": 870 + }, + { + "epoch": 3.526315789473684, + "grad_norm": 0.001642534891966009, + "learning_rate": 9.691429465584866e-06, + "loss": 0.0, + "step": 871 + }, + { + "epoch": 3.5303643724696356, + "grad_norm": 0.0012822316761886235, + "learning_rate": 9.643007502033492e-06, + "loss": 0.0, + "step": 872 + }, + { + "epoch": 3.534412955465587, + "grad_norm": 0.0010196788887256603, + "learning_rate": 9.594668352718152e-06, + "loss": 0.0, + "step": 873 + }, + { + "epoch": 3.5384615384615383, + "grad_norm": 0.017331659949776177, + "learning_rate": 9.546412404157383e-06, + "loss": 0.0001, + "step": 874 + }, + { + "epoch": 3.54251012145749, + "grad_norm": 0.001363916615927253, + "learning_rate": 9.498240042204471e-06, + "loss": 0.0, + "step": 875 + }, + { + "epoch": 3.5465587044534415, + "grad_norm": 0.0011891311467995005, + "learning_rate": 9.450151652044339e-06, + "loss": 0.0, + "step": 876 + }, + { + "epoch": 3.5506072874493926, + "grad_norm": 0.0011404216401430893, + "learning_rate": 9.402147618190472e-06, + "loss": 0.0, + "step": 877 + }, + { + "epoch": 3.554655870445344, + "grad_norm": 0.0054636431299249465, + "learning_rate": 9.354228324481841e-06, + "loss": 0.0, + "step": 878 + }, + { + "epoch": 3.5587044534412957, + "grad_norm": 0.0024901005916972687, + "learning_rate": 9.306394154079852e-06, + "loss": 0.0, + "step": 879 + }, + { + "epoch": 3.562753036437247, + "grad_norm": 0.004268987527624796, + "learning_rate": 9.258645489465254e-06, + "loss": 0.0, + "step": 880 + }, + { + "epoch": 3.5668016194331984, + "grad_norm": 0.009644897473899782, + "learning_rate": 9.21098271243509e-06, + "loss": 0.0001, + "step": 881 + }, + { + "epoch": 3.57085020242915, + "grad_norm": 0.0012372417988239712, + "learning_rate": 9.163406204099659e-06, + "loss": 0.0, + "step": 882 + }, + { + "epoch": 3.574898785425101, + "grad_norm": 0.004868535459033586, + "learning_rate": 9.115916344879452e-06, + "loss": 0.0001, + "step": 883 + }, + { + "epoch": 3.5789473684210527, + "grad_norm": 0.002867338600755962, + "learning_rate": 9.068513514502126e-06, + "loss": 0.0, + "step": 884 + }, + { + "epoch": 3.582995951417004, + "grad_norm": 0.0020499937520915665, + "learning_rate": 9.021198091999439e-06, + "loss": 0.0, + "step": 885 + }, + { + "epoch": 3.5870445344129553, + "grad_norm": 0.001101249865023786, + "learning_rate": 8.973970455704251e-06, + "loss": 0.0, + "step": 886 + }, + { + "epoch": 3.591093117408907, + "grad_norm": 0.003215815278994355, + "learning_rate": 8.926830983247495e-06, + "loss": 0.0, + "step": 887 + }, + { + "epoch": 3.5951417004048585, + "grad_norm": 0.0066395128179204755, + "learning_rate": 8.879780051555116e-06, + "loss": 0.0001, + "step": 888 + }, + { + "epoch": 3.5991902834008096, + "grad_norm": 0.0013368967583306062, + "learning_rate": 8.83281803684512e-06, + "loss": 0.0, + "step": 889 + }, + { + "epoch": 3.603238866396761, + "grad_norm": 0.0010887306055677925, + "learning_rate": 8.785945314624532e-06, + "loss": 0.0, + "step": 890 + }, + { + "epoch": 3.6072874493927127, + "grad_norm": 0.0008693733272676528, + "learning_rate": 8.739162259686385e-06, + "loss": 0.0, + "step": 891 + }, + { + "epoch": 3.611336032388664, + "grad_norm": 0.0009203827923491038, + "learning_rate": 8.692469246106734e-06, + "loss": 0.0, + "step": 892 + }, + { + "epoch": 3.6153846153846154, + "grad_norm": 0.0009672802381561115, + "learning_rate": 8.645866647241677e-06, + "loss": 0.0, + "step": 893 + }, + { + "epoch": 3.619433198380567, + "grad_norm": 0.001318635032051758, + "learning_rate": 8.599354835724348e-06, + "loss": 0.0, + "step": 894 + }, + { + "epoch": 3.623481781376518, + "grad_norm": 0.0008863490503982807, + "learning_rate": 8.552934183461958e-06, + "loss": 0.0, + "step": 895 + }, + { + "epoch": 3.6275303643724697, + "grad_norm": 0.006400513132677742, + "learning_rate": 8.506605061632805e-06, + "loss": 0.0, + "step": 896 + }, + { + "epoch": 3.6315789473684212, + "grad_norm": 0.009301920805199917, + "learning_rate": 8.460367840683308e-06, + "loss": 0.0, + "step": 897 + }, + { + "epoch": 3.6356275303643724, + "grad_norm": 0.0019750404569612023, + "learning_rate": 8.414222890325065e-06, + "loss": 0.0, + "step": 898 + }, + { + "epoch": 3.639676113360324, + "grad_norm": 0.0009901679333558617, + "learning_rate": 8.368170579531855e-06, + "loss": 0.0, + "step": 899 + }, + { + "epoch": 3.6437246963562755, + "grad_norm": 0.0009818172154136006, + "learning_rate": 8.322211276536733e-06, + "loss": 0.0, + "step": 900 + }, + { + "epoch": 3.6477732793522266, + "grad_norm": 0.001378538619905964, + "learning_rate": 8.27634534882907e-06, + "loss": 0.0, + "step": 901 + }, + { + "epoch": 3.651821862348178, + "grad_norm": 0.0010360945441312448, + "learning_rate": 8.230573163151585e-06, + "loss": 0.0, + "step": 902 + }, + { + "epoch": 3.6558704453441297, + "grad_norm": 0.0012206318120555038, + "learning_rate": 8.184895085497466e-06, + "loss": 0.0, + "step": 903 + }, + { + "epoch": 3.659919028340081, + "grad_norm": 0.0008533647805521348, + "learning_rate": 8.139311481107401e-06, + "loss": 0.0, + "step": 904 + }, + { + "epoch": 3.6639676113360324, + "grad_norm": 0.001058326338535455, + "learning_rate": 8.093822714466686e-06, + "loss": 0.0, + "step": 905 + }, + { + "epoch": 3.668016194331984, + "grad_norm": 0.000852528873310708, + "learning_rate": 8.048429149302271e-06, + "loss": 0.0, + "step": 906 + }, + { + "epoch": 3.672064777327935, + "grad_norm": 0.0008864860747179085, + "learning_rate": 8.003131148579902e-06, + "loss": 0.0, + "step": 907 + }, + { + "epoch": 3.6761133603238867, + "grad_norm": 0.004036195992358091, + "learning_rate": 7.957929074501182e-06, + "loss": 0.0001, + "step": 908 + }, + { + "epoch": 3.6801619433198383, + "grad_norm": 0.0009535075639458327, + "learning_rate": 7.912823288500695e-06, + "loss": 0.0, + "step": 909 + }, + { + "epoch": 3.6842105263157894, + "grad_norm": 0.000856947460132472, + "learning_rate": 7.867814151243102e-06, + "loss": 0.0, + "step": 910 + }, + { + "epoch": 3.688259109311741, + "grad_norm": 0.0008902077107799619, + "learning_rate": 7.822902022620267e-06, + "loss": 0.0, + "step": 911 + }, + { + "epoch": 3.6923076923076925, + "grad_norm": 0.00451501883191735, + "learning_rate": 7.778087261748377e-06, + "loss": 0.0001, + "step": 912 + }, + { + "epoch": 3.6963562753036436, + "grad_norm": 0.0010865527656616838, + "learning_rate": 7.733370226965056e-06, + "loss": 0.0, + "step": 913 + }, + { + "epoch": 3.700404858299595, + "grad_norm": 0.0008280752269654713, + "learning_rate": 7.688751275826527e-06, + "loss": 0.0, + "step": 914 + }, + { + "epoch": 3.7044534412955468, + "grad_norm": 0.0012317924447913348, + "learning_rate": 7.644230765104735e-06, + "loss": 0.0, + "step": 915 + }, + { + "epoch": 3.708502024291498, + "grad_norm": 0.0011232643564153741, + "learning_rate": 7.599809050784502e-06, + "loss": 0.0, + "step": 916 + }, + { + "epoch": 3.7125506072874495, + "grad_norm": 0.0008226908183058406, + "learning_rate": 7.555486488060661e-06, + "loss": 0.0, + "step": 917 + }, + { + "epoch": 3.716599190283401, + "grad_norm": 0.001488882638494634, + "learning_rate": 7.511263431335249e-06, + "loss": 0.0, + "step": 918 + }, + { + "epoch": 3.720647773279352, + "grad_norm": 0.000812113982826468, + "learning_rate": 7.467140234214656e-06, + "loss": 0.0, + "step": 919 + }, + { + "epoch": 3.7246963562753037, + "grad_norm": 0.0008151346826423783, + "learning_rate": 7.423117249506777e-06, + "loss": 0.0, + "step": 920 + }, + { + "epoch": 3.7287449392712553, + "grad_norm": 0.0008443132303742008, + "learning_rate": 7.379194829218231e-06, + "loss": 0.0, + "step": 921 + }, + { + "epoch": 3.7327935222672064, + "grad_norm": 0.0008561049661321085, + "learning_rate": 7.335373324551518e-06, + "loss": 0.0, + "step": 922 + }, + { + "epoch": 3.736842105263158, + "grad_norm": 0.0013913653747298926, + "learning_rate": 7.291653085902224e-06, + "loss": 0.0, + "step": 923 + }, + { + "epoch": 3.7408906882591095, + "grad_norm": 0.009009611425625107, + "learning_rate": 7.248034462856213e-06, + "loss": 0.0001, + "step": 924 + }, + { + "epoch": 3.7449392712550607, + "grad_norm": 0.0007705219542980646, + "learning_rate": 7.20451780418683e-06, + "loss": 0.0, + "step": 925 + }, + { + "epoch": 3.748987854251012, + "grad_norm": 0.0007515404066827166, + "learning_rate": 7.161103457852128e-06, + "loss": 0.0, + "step": 926 + }, + { + "epoch": 3.753036437246964, + "grad_norm": 0.0007816910290023181, + "learning_rate": 7.117791770992046e-06, + "loss": 0.0, + "step": 927 + }, + { + "epoch": 3.757085020242915, + "grad_norm": 0.0008259214920973247, + "learning_rate": 7.074583089925688e-06, + "loss": 0.0, + "step": 928 + }, + { + "epoch": 3.7611336032388665, + "grad_norm": 0.0008534343395371283, + "learning_rate": 7.0314777601485155e-06, + "loss": 0.0, + "step": 929 + }, + { + "epoch": 3.765182186234818, + "grad_norm": 0.000808659909733043, + "learning_rate": 6.988476126329604e-06, + "loss": 0.0, + "step": 930 + }, + { + "epoch": 3.769230769230769, + "grad_norm": 0.0008537090128356473, + "learning_rate": 6.945578532308857e-06, + "loss": 0.0, + "step": 931 + }, + { + "epoch": 3.7732793522267207, + "grad_norm": 0.0007783067431469113, + "learning_rate": 6.902785321094301e-06, + "loss": 0.0, + "step": 932 + }, + { + "epoch": 3.7773279352226723, + "grad_norm": 0.001085440978460737, + "learning_rate": 6.8600968348593115e-06, + "loss": 0.0, + "step": 933 + }, + { + "epoch": 3.7813765182186234, + "grad_norm": 0.0008305328321424452, + "learning_rate": 6.817513414939892e-06, + "loss": 0.0, + "step": 934 + }, + { + "epoch": 3.785425101214575, + "grad_norm": 0.0007659362898653108, + "learning_rate": 6.775035401831924e-06, + "loss": 0.0, + "step": 935 + }, + { + "epoch": 3.7894736842105265, + "grad_norm": 0.0008297919088068707, + "learning_rate": 6.732663135188473e-06, + "loss": 0.0, + "step": 936 + }, + { + "epoch": 3.7935222672064777, + "grad_norm": 0.0007235774192640838, + "learning_rate": 6.690396953817055e-06, + "loss": 0.0, + "step": 937 + }, + { + "epoch": 3.7975708502024292, + "grad_norm": 0.000760275021810467, + "learning_rate": 6.648237195676927e-06, + "loss": 0.0, + "step": 938 + }, + { + "epoch": 3.801619433198381, + "grad_norm": 0.0008034625263543623, + "learning_rate": 6.606184197876395e-06, + "loss": 0.0, + "step": 939 + }, + { + "epoch": 3.805668016194332, + "grad_norm": 0.0008273944922946577, + "learning_rate": 6.564238296670109e-06, + "loss": 0.0, + "step": 940 + }, + { + "epoch": 3.8097165991902835, + "grad_norm": 0.0007868063544668463, + "learning_rate": 6.52239982745638e-06, + "loss": 0.0, + "step": 941 + }, + { + "epoch": 3.813765182186235, + "grad_norm": 0.0007913813127808242, + "learning_rate": 6.480669124774484e-06, + "loss": 0.0, + "step": 942 + }, + { + "epoch": 3.817813765182186, + "grad_norm": 0.0008569049736271083, + "learning_rate": 6.439046522302013e-06, + "loss": 0.0, + "step": 943 + }, + { + "epoch": 3.8218623481781377, + "grad_norm": 0.0008951390984412765, + "learning_rate": 6.397532352852191e-06, + "loss": 0.0, + "step": 944 + }, + { + "epoch": 3.8259109311740893, + "grad_norm": 0.0009856015800993687, + "learning_rate": 6.356126948371202e-06, + "loss": 0.0, + "step": 945 + }, + { + "epoch": 3.8299595141700404, + "grad_norm": 0.000753371900346853, + "learning_rate": 6.314830639935559e-06, + "loss": 0.0, + "step": 946 + }, + { + "epoch": 3.834008097165992, + "grad_norm": 0.0037502945525558586, + "learning_rate": 6.273643757749448e-06, + "loss": 0.0001, + "step": 947 + }, + { + "epoch": 3.8380566801619436, + "grad_norm": 0.000766018590175195, + "learning_rate": 6.2325666311420856e-06, + "loss": 0.0, + "step": 948 + }, + { + "epoch": 3.8421052631578947, + "grad_norm": 0.0007504357686114615, + "learning_rate": 6.1915995885650675e-06, + "loss": 0.0, + "step": 949 + }, + { + "epoch": 3.8461538461538463, + "grad_norm": 0.0007792107858899475, + "learning_rate": 6.150742957589784e-06, + "loss": 0.0, + "step": 950 + }, + { + "epoch": 3.850202429149798, + "grad_norm": 0.0007433550276693925, + "learning_rate": 6.109997064904774e-06, + "loss": 0.0, + "step": 951 + }, + { + "epoch": 3.854251012145749, + "grad_norm": 0.0007871412061319773, + "learning_rate": 6.069362236313092e-06, + "loss": 0.0, + "step": 952 + }, + { + "epoch": 3.8582995951417005, + "grad_norm": 0.000788066348061703, + "learning_rate": 6.0288387967297565e-06, + "loss": 0.0, + "step": 953 + }, + { + "epoch": 3.862348178137652, + "grad_norm": 0.0007835696743170621, + "learning_rate": 5.988427070179108e-06, + "loss": 0.0, + "step": 954 + }, + { + "epoch": 3.866396761133603, + "grad_norm": 0.0006949863669500585, + "learning_rate": 5.948127379792234e-06, + "loss": 0.0, + "step": 955 + }, + { + "epoch": 3.8704453441295548, + "grad_norm": 0.0006886133810053194, + "learning_rate": 5.90794004780439e-06, + "loss": 0.0, + "step": 956 + }, + { + "epoch": 3.8744939271255063, + "grad_norm": 0.0007618578669333108, + "learning_rate": 5.867865395552405e-06, + "loss": 0.0, + "step": 957 + }, + { + "epoch": 3.8785425101214575, + "grad_norm": 0.0007341390169891019, + "learning_rate": 5.8279037434721364e-06, + "loss": 0.0, + "step": 958 + }, + { + "epoch": 3.882591093117409, + "grad_norm": 0.0006696888235975566, + "learning_rate": 5.788055411095892e-06, + "loss": 0.0, + "step": 959 + }, + { + "epoch": 3.8866396761133606, + "grad_norm": 0.000721497603461402, + "learning_rate": 5.7483207170498625e-06, + "loss": 0.0, + "step": 960 + }, + { + "epoch": 3.8906882591093117, + "grad_norm": 0.0006982395953761789, + "learning_rate": 5.708699979051613e-06, + "loss": 0.0, + "step": 961 + }, + { + "epoch": 3.8947368421052633, + "grad_norm": 0.0007145036666536766, + "learning_rate": 5.66919351390751e-06, + "loss": 0.0, + "step": 962 + }, + { + "epoch": 3.898785425101215, + "grad_norm": 0.0007129212167931738, + "learning_rate": 5.629801637510188e-06, + "loss": 0.0, + "step": 963 + }, + { + "epoch": 3.902834008097166, + "grad_norm": 0.0007058143255764351, + "learning_rate": 5.590524664836043e-06, + "loss": 0.0, + "step": 964 + }, + { + "epoch": 3.9068825910931175, + "grad_norm": 0.0007485026610011524, + "learning_rate": 5.551362909942708e-06, + "loss": 0.0, + "step": 965 + }, + { + "epoch": 3.910931174089069, + "grad_norm": 0.0007347135369699002, + "learning_rate": 5.51231668596653e-06, + "loss": 0.0, + "step": 966 + }, + { + "epoch": 3.91497975708502, + "grad_norm": 0.0007727676944674335, + "learning_rate": 5.47338630512007e-06, + "loss": 0.0, + "step": 967 + }, + { + "epoch": 3.919028340080972, + "grad_norm": 0.0006842556689512628, + "learning_rate": 5.434572078689617e-06, + "loss": 0.0, + "step": 968 + }, + { + "epoch": 3.9230769230769234, + "grad_norm": 0.0006960376443855555, + "learning_rate": 5.395874317032694e-06, + "loss": 0.0, + "step": 969 + }, + { + "epoch": 3.9271255060728745, + "grad_norm": 0.0008798746995732086, + "learning_rate": 5.357293329575573e-06, + "loss": 0.0, + "step": 970 + }, + { + "epoch": 3.931174089068826, + "grad_norm": 0.0007134626454036502, + "learning_rate": 5.318829424810797e-06, + "loss": 0.0, + "step": 971 + }, + { + "epoch": 3.9352226720647776, + "grad_norm": 0.0006699755549023481, + "learning_rate": 5.280482910294728e-06, + "loss": 0.0, + "step": 972 + }, + { + "epoch": 3.9392712550607287, + "grad_norm": 0.0007002588722787359, + "learning_rate": 5.242254092645078e-06, + "loss": 0.0, + "step": 973 + }, + { + "epoch": 3.9433198380566803, + "grad_norm": 0.0007220887250971208, + "learning_rate": 5.2041432775384445e-06, + "loss": 0.0, + "step": 974 + }, + { + "epoch": 3.9473684210526314, + "grad_norm": 0.0007247561742555688, + "learning_rate": 5.166150769707892e-06, + "loss": 0.0, + "step": 975 + }, + { + "epoch": 3.951417004048583, + "grad_norm": 0.0007107362524983746, + "learning_rate": 5.128276872940503e-06, + "loss": 0.0, + "step": 976 + }, + { + "epoch": 3.9554655870445345, + "grad_norm": 0.0006837904762835526, + "learning_rate": 5.090521890074953e-06, + "loss": 0.0, + "step": 977 + }, + { + "epoch": 3.9595141700404857, + "grad_norm": 0.0006920991849820811, + "learning_rate": 5.0528861229990675e-06, + "loss": 0.0, + "step": 978 + }, + { + "epoch": 3.9635627530364372, + "grad_norm": 0.0007685612151591167, + "learning_rate": 5.015369872647442e-06, + "loss": 0.0, + "step": 979 + }, + { + "epoch": 3.967611336032389, + "grad_norm": 0.0007581197422333859, + "learning_rate": 4.9779734389990195e-06, + "loss": 0.0, + "step": 980 + }, + { + "epoch": 3.97165991902834, + "grad_norm": 0.0006946867196604539, + "learning_rate": 4.940697121074675e-06, + "loss": 0.0, + "step": 981 + }, + { + "epoch": 3.9757085020242915, + "grad_norm": 0.0007442460035601494, + "learning_rate": 4.903541216934864e-06, + "loss": 0.0, + "step": 982 + }, + { + "epoch": 3.979757085020243, + "grad_norm": 0.0007123862371644612, + "learning_rate": 4.8665060236772e-06, + "loss": 0.0, + "step": 983 + }, + { + "epoch": 3.983805668016194, + "grad_norm": 0.0007179241869437032, + "learning_rate": 4.829591837434102e-06, + "loss": 0.0, + "step": 984 + }, + { + "epoch": 3.9878542510121457, + "grad_norm": 0.0007075227416155394, + "learning_rate": 4.792798953370424e-06, + "loss": 0.0, + "step": 985 + }, + { + "epoch": 3.9919028340080973, + "grad_norm": 0.0006821446182065088, + "learning_rate": 4.756127665681083e-06, + "loss": 0.0, + "step": 986 + }, + { + "epoch": 3.9959514170040484, + "grad_norm": 0.0006991737232274961, + "learning_rate": 4.719578267588725e-06, + "loss": 0.0, + "step": 987 + }, + { + "epoch": 4.0, + "grad_norm": 0.000699035839646747, + "learning_rate": 4.6831510513413525e-06, + "loss": 0.0, + "step": 988 + }, + { + "epoch": 4.004048582995951, + "grad_norm": 0.0006645980974583952, + "learning_rate": 4.6468463082100176e-06, + "loss": 0.0, + "step": 989 + }, + { + "epoch": 4.008097165991903, + "grad_norm": 0.0005988104054450058, + "learning_rate": 4.610664328486485e-06, + "loss": 0.0, + "step": 990 + }, + { + "epoch": 4.012145748987854, + "grad_norm": 0.0006984122853467349, + "learning_rate": 4.5746054014809e-06, + "loss": 0.0, + "step": 991 + }, + { + "epoch": 4.016194331983805, + "grad_norm": 0.0006238808606561296, + "learning_rate": 4.5386698155194766e-06, + "loss": 0.0, + "step": 992 + }, + { + "epoch": 4.020242914979757, + "grad_norm": 0.0006121659529477136, + "learning_rate": 4.502857857942204e-06, + "loss": 0.0, + "step": 993 + }, + { + "epoch": 4.0242914979757085, + "grad_norm": 0.0006547735805803717, + "learning_rate": 4.467169815100549e-06, + "loss": 0.0, + "step": 994 + }, + { + "epoch": 4.02834008097166, + "grad_norm": 0.0019544200657247005, + "learning_rate": 4.431605972355137e-06, + "loss": 0.0, + "step": 995 + }, + { + "epoch": 4.032388663967612, + "grad_norm": 0.0006729393408615201, + "learning_rate": 4.396166614073516e-06, + "loss": 0.0, + "step": 996 + }, + { + "epoch": 4.036437246963563, + "grad_norm": 0.0006822037195586887, + "learning_rate": 4.360852023627846e-06, + "loss": 0.0, + "step": 997 + }, + { + "epoch": 4.040485829959514, + "grad_norm": 0.0006237623517966427, + "learning_rate": 4.325662483392656e-06, + "loss": 0.0, + "step": 998 + }, + { + "epoch": 4.044534412955466, + "grad_norm": 0.000623285625438411, + "learning_rate": 4.29059827474257e-06, + "loss": 0.0, + "step": 999 + }, + { + "epoch": 4.048582995951417, + "grad_norm": 0.0006667898194258289, + "learning_rate": 4.255659678050063e-06, + "loss": 0.0, + "step": 1000 + }, + { + "epoch": 4.052631578947368, + "grad_norm": 0.0006892110357439175, + "learning_rate": 4.22084697268323e-06, + "loss": 0.0, + "step": 1001 + }, + { + "epoch": 4.05668016194332, + "grad_norm": 0.0006477299773839236, + "learning_rate": 4.186160437003535e-06, + "loss": 0.0, + "step": 1002 + }, + { + "epoch": 4.060728744939271, + "grad_norm": 0.0006300012995730281, + "learning_rate": 4.151600348363582e-06, + "loss": 0.0, + "step": 1003 + }, + { + "epoch": 4.064777327935222, + "grad_norm": 0.0007089555276071764, + "learning_rate": 4.11716698310493e-06, + "loss": 0.0, + "step": 1004 + }, + { + "epoch": 4.068825910931174, + "grad_norm": 0.0006548572272331035, + "learning_rate": 4.082860616555848e-06, + "loss": 0.0, + "step": 1005 + }, + { + "epoch": 4.0728744939271255, + "grad_norm": 0.0006361351970509797, + "learning_rate": 4.048681523029125e-06, + "loss": 0.0, + "step": 1006 + }, + { + "epoch": 4.076923076923077, + "grad_norm": 0.0006082906202124486, + "learning_rate": 4.014629975819882e-06, + "loss": 0.0, + "step": 1007 + }, + { + "epoch": 4.080971659919029, + "grad_norm": 0.0006072956516736832, + "learning_rate": 3.9807062472033855e-06, + "loss": 0.0, + "step": 1008 + }, + { + "epoch": 4.08502024291498, + "grad_norm": 0.0006530140540498992, + "learning_rate": 3.94691060843287e-06, + "loss": 0.0, + "step": 1009 + }, + { + "epoch": 4.089068825910931, + "grad_norm": 0.0006302773062699, + "learning_rate": 3.913243329737353e-06, + "loss": 0.0, + "step": 1010 + }, + { + "epoch": 4.093117408906883, + "grad_norm": 0.0006718191090980719, + "learning_rate": 3.879704680319503e-06, + "loss": 0.0, + "step": 1011 + }, + { + "epoch": 4.097165991902834, + "grad_norm": 0.0006282638561708036, + "learning_rate": 3.846294928353471e-06, + "loss": 0.0, + "step": 1012 + }, + { + "epoch": 4.101214574898785, + "grad_norm": 0.0006439465450341752, + "learning_rate": 3.8130143409827345e-06, + "loss": 0.0, + "step": 1013 + }, + { + "epoch": 4.105263157894737, + "grad_norm": 0.0006404373374257043, + "learning_rate": 3.7798631843179887e-06, + "loss": 0.0, + "step": 1014 + }, + { + "epoch": 4.109311740890688, + "grad_norm": 0.0006546133476447945, + "learning_rate": 3.7468417234349973e-06, + "loss": 0.0, + "step": 1015 + }, + { + "epoch": 4.113360323886639, + "grad_norm": 0.0006232586084406421, + "learning_rate": 3.7139502223724823e-06, + "loss": 0.0, + "step": 1016 + }, + { + "epoch": 4.117408906882591, + "grad_norm": 0.0006245988339023158, + "learning_rate": 3.6811889441300095e-06, + "loss": 0.0, + "step": 1017 + }, + { + "epoch": 4.1214574898785425, + "grad_norm": 0.0006682811338005268, + "learning_rate": 3.648558150665886e-06, + "loss": 0.0, + "step": 1018 + }, + { + "epoch": 4.125506072874494, + "grad_norm": 0.0006893366212533165, + "learning_rate": 3.616058102895075e-06, + "loss": 0.0, + "step": 1019 + }, + { + "epoch": 4.129554655870446, + "grad_norm": 0.000645246761591758, + "learning_rate": 3.583689060687079e-06, + "loss": 0.0, + "step": 1020 + }, + { + "epoch": 4.133603238866397, + "grad_norm": 0.0006069936928434769, + "learning_rate": 3.5514512828639023e-06, + "loss": 0.0, + "step": 1021 + }, + { + "epoch": 4.137651821862348, + "grad_norm": 0.0006301465200438003, + "learning_rate": 3.51934502719796e-06, + "loss": 0.0, + "step": 1022 + }, + { + "epoch": 4.1417004048583, + "grad_norm": 0.0006130183322025312, + "learning_rate": 3.4873705504100185e-06, + "loss": 0.0, + "step": 1023 + }, + { + "epoch": 4.145748987854251, + "grad_norm": 0.0006412996532427162, + "learning_rate": 3.4555281081671345e-06, + "loss": 0.0, + "step": 1024 + }, + { + "epoch": 4.149797570850202, + "grad_norm": 0.0006390892269983162, + "learning_rate": 3.4238179550806326e-06, + "loss": 0.0, + "step": 1025 + }, + { + "epoch": 4.153846153846154, + "grad_norm": 0.0006345071262174445, + "learning_rate": 3.3922403447040497e-06, + "loss": 0.0, + "step": 1026 + }, + { + "epoch": 4.157894736842105, + "grad_norm": 0.0006827872032884239, + "learning_rate": 3.3607955295311247e-06, + "loss": 0.0, + "step": 1027 + }, + { + "epoch": 4.161943319838056, + "grad_norm": 0.0005937947152910665, + "learning_rate": 3.32948376099375e-06, + "loss": 0.0, + "step": 1028 + }, + { + "epoch": 4.165991902834008, + "grad_norm": 0.0006495588121820903, + "learning_rate": 3.2983052894599976e-06, + "loss": 0.0, + "step": 1029 + }, + { + "epoch": 4.17004048582996, + "grad_norm": 0.0006142286957712919, + "learning_rate": 3.2672603642320945e-06, + "loss": 0.0, + "step": 1030 + }, + { + "epoch": 4.174089068825911, + "grad_norm": 0.000632316185028267, + "learning_rate": 3.236349233544438e-06, + "loss": 0.0, + "step": 1031 + }, + { + "epoch": 4.178137651821863, + "grad_norm": 0.000644373386149358, + "learning_rate": 3.205572144561604e-06, + "loss": 0.0, + "step": 1032 + }, + { + "epoch": 4.182186234817814, + "grad_norm": 0.000609371504018645, + "learning_rate": 3.174929343376374e-06, + "loss": 0.0, + "step": 1033 + }, + { + "epoch": 4.186234817813765, + "grad_norm": 0.0006570144019449713, + "learning_rate": 3.1444210750077777e-06, + "loss": 0.0, + "step": 1034 + }, + { + "epoch": 4.190283400809717, + "grad_norm": 0.0006389359104168866, + "learning_rate": 3.1140475833991046e-06, + "loss": 0.0, + "step": 1035 + }, + { + "epoch": 4.194331983805668, + "grad_norm": 0.0006127093550475348, + "learning_rate": 3.0838091114159895e-06, + "loss": 0.0, + "step": 1036 + }, + { + "epoch": 4.198380566801619, + "grad_norm": 0.0006592426923235761, + "learning_rate": 3.0537059008444524e-06, + "loss": 0.0, + "step": 1037 + }, + { + "epoch": 4.202429149797571, + "grad_norm": 0.0006641067438386582, + "learning_rate": 3.0237381923889563e-06, + "loss": 0.0, + "step": 1038 + }, + { + "epoch": 4.206477732793522, + "grad_norm": 0.0006679491641360588, + "learning_rate": 2.9939062256704997e-06, + "loss": 0.0, + "step": 1039 + }, + { + "epoch": 4.2105263157894735, + "grad_norm": 0.0005948834679104125, + "learning_rate": 2.964210239224696e-06, + "loss": 0.0, + "step": 1040 + }, + { + "epoch": 4.2145748987854255, + "grad_norm": 0.0006502116494043688, + "learning_rate": 2.934650470499869e-06, + "loss": 0.0, + "step": 1041 + }, + { + "epoch": 4.218623481781377, + "grad_norm": 0.0006231981421501528, + "learning_rate": 2.905227155855128e-06, + "loss": 0.0, + "step": 1042 + }, + { + "epoch": 4.222672064777328, + "grad_norm": 0.000630827489192215, + "learning_rate": 2.875940530558519e-06, + "loss": 0.0, + "step": 1043 + }, + { + "epoch": 4.22672064777328, + "grad_norm": 0.0006542634630131352, + "learning_rate": 2.8467908287851154e-06, + "loss": 0.0, + "step": 1044 + }, + { + "epoch": 4.230769230769231, + "grad_norm": 0.0006202556685675078, + "learning_rate": 2.8177782836151603e-06, + "loss": 0.0, + "step": 1045 + }, + { + "epoch": 4.234817813765182, + "grad_norm": 0.0006141753902589126, + "learning_rate": 2.7889031270321874e-06, + "loss": 0.0, + "step": 1046 + }, + { + "epoch": 4.238866396761134, + "grad_norm": 0.0006279571620514202, + "learning_rate": 2.760165589921184e-06, + "loss": 0.0, + "step": 1047 + }, + { + "epoch": 4.242914979757085, + "grad_norm": 0.000635162956292089, + "learning_rate": 2.7315659020667305e-06, + "loss": 0.0, + "step": 1048 + }, + { + "epoch": 4.246963562753036, + "grad_norm": 0.0006280731986294182, + "learning_rate": 2.703104292151164e-06, + "loss": 0.0, + "step": 1049 + }, + { + "epoch": 4.251012145748988, + "grad_norm": 0.0006202034501147215, + "learning_rate": 2.6747809877527607e-06, + "loss": 0.0, + "step": 1050 + }, + { + "epoch": 4.255060728744939, + "grad_norm": 0.0006398985555763374, + "learning_rate": 2.6465962153439082e-06, + "loss": 0.0, + "step": 1051 + }, + { + "epoch": 4.2591093117408905, + "grad_norm": 0.0006279871174110503, + "learning_rate": 2.618550200289303e-06, + "loss": 0.0, + "step": 1052 + }, + { + "epoch": 4.2631578947368425, + "grad_norm": 0.0006134653696475744, + "learning_rate": 2.59064316684412e-06, + "loss": 0.0, + "step": 1053 + }, + { + "epoch": 4.267206477732794, + "grad_norm": 0.0006228470213125444, + "learning_rate": 2.562875338152264e-06, + "loss": 0.0, + "step": 1054 + }, + { + "epoch": 4.271255060728745, + "grad_norm": 0.0006022858844248829, + "learning_rate": 2.5352469362445552e-06, + "loss": 0.0, + "step": 1055 + }, + { + "epoch": 4.275303643724697, + "grad_norm": 0.0006005159839475773, + "learning_rate": 2.5077581820369523e-06, + "loss": 0.0, + "step": 1056 + }, + { + "epoch": 4.279352226720648, + "grad_norm": 0.0005856010550918477, + "learning_rate": 2.480409295328803e-06, + "loss": 0.0, + "step": 1057 + }, + { + "epoch": 4.283400809716599, + "grad_norm": 0.0006191844234872906, + "learning_rate": 2.45320049480108e-06, + "loss": 0.0, + "step": 1058 + }, + { + "epoch": 4.287449392712551, + "grad_norm": 0.0006172213413323523, + "learning_rate": 2.4261319980146293e-06, + "loss": 0.0, + "step": 1059 + }, + { + "epoch": 4.291497975708502, + "grad_norm": 0.0006392043563073977, + "learning_rate": 2.399204021408421e-06, + "loss": 0.0, + "step": 1060 + }, + { + "epoch": 4.295546558704453, + "grad_norm": 0.0006171787761430552, + "learning_rate": 2.3724167802978395e-06, + "loss": 0.0, + "step": 1061 + }, + { + "epoch": 4.299595141700405, + "grad_norm": 0.0006301277143553499, + "learning_rate": 2.3457704888729625e-06, + "loss": 0.0, + "step": 1062 + }, + { + "epoch": 4.303643724696356, + "grad_norm": 0.0006290917833991652, + "learning_rate": 2.3192653601968162e-06, + "loss": 0.0, + "step": 1063 + }, + { + "epoch": 4.3076923076923075, + "grad_norm": 0.0005908618108318585, + "learning_rate": 2.2929016062037057e-06, + "loss": 0.0, + "step": 1064 + }, + { + "epoch": 4.3117408906882595, + "grad_norm": 0.0006189261982893687, + "learning_rate": 2.2666794376975055e-06, + "loss": 0.0, + "step": 1065 + }, + { + "epoch": 4.315789473684211, + "grad_norm": 0.0006150082711403873, + "learning_rate": 2.24059906434998e-06, + "loss": 0.0, + "step": 1066 + }, + { + "epoch": 4.319838056680162, + "grad_norm": 0.0006012064710944514, + "learning_rate": 2.214660694699091e-06, + "loss": 0.0, + "step": 1067 + }, + { + "epoch": 4.323886639676114, + "grad_norm": 0.0005889455463704733, + "learning_rate": 2.188864536147355e-06, + "loss": 0.0, + "step": 1068 + }, + { + "epoch": 4.327935222672065, + "grad_norm": 0.0006646167262174587, + "learning_rate": 2.1632107949601757e-06, + "loss": 0.0, + "step": 1069 + }, + { + "epoch": 4.331983805668016, + "grad_norm": 0.0005987264469956702, + "learning_rate": 2.137699676264171e-06, + "loss": 0.0, + "step": 1070 + }, + { + "epoch": 4.336032388663968, + "grad_norm": 0.0006272817922194721, + "learning_rate": 2.1123313840455738e-06, + "loss": 0.0, + "step": 1071 + }, + { + "epoch": 4.340080971659919, + "grad_norm": 0.0006078410613434757, + "learning_rate": 2.08710612114857e-06, + "loss": 0.0, + "step": 1072 + }, + { + "epoch": 4.34412955465587, + "grad_norm": 0.0006317462101818889, + "learning_rate": 2.0620240892736953e-06, + "loss": 0.0, + "step": 1073 + }, + { + "epoch": 4.348178137651822, + "grad_norm": 0.0006557863133483099, + "learning_rate": 2.0370854889762e-06, + "loss": 0.0, + "step": 1074 + }, + { + "epoch": 4.352226720647773, + "grad_norm": 0.0005666517662646615, + "learning_rate": 2.0122905196644726e-06, + "loss": 0.0, + "step": 1075 + }, + { + "epoch": 4.3562753036437245, + "grad_norm": 0.0006045909903809599, + "learning_rate": 1.987639379598427e-06, + "loss": 0.0, + "step": 1076 + }, + { + "epoch": 4.3603238866396765, + "grad_norm": 0.0005864654747635421, + "learning_rate": 1.963132265887919e-06, + "loss": 0.0, + "step": 1077 + }, + { + "epoch": 4.364372469635628, + "grad_norm": 0.0006299852816838784, + "learning_rate": 1.9387693744911784e-06, + "loss": 0.0, + "step": 1078 + }, + { + "epoch": 4.368421052631579, + "grad_norm": 0.0005951839326003402, + "learning_rate": 1.9145509002132368e-06, + "loss": 0.0, + "step": 1079 + }, + { + "epoch": 4.372469635627531, + "grad_norm": 0.0006120767548555046, + "learning_rate": 1.8904770367043723e-06, + "loss": 0.0, + "step": 1080 + }, + { + "epoch": 4.376518218623482, + "grad_norm": 0.0005856719300990638, + "learning_rate": 1.8665479764585504e-06, + "loss": 0.0, + "step": 1081 + }, + { + "epoch": 4.380566801619433, + "grad_norm": 0.0006013667719379405, + "learning_rate": 1.8427639108118977e-06, + "loss": 0.0, + "step": 1082 + }, + { + "epoch": 4.384615384615385, + "grad_norm": 0.0006189134243361386, + "learning_rate": 1.8191250299411755e-06, + "loss": 0.0, + "step": 1083 + }, + { + "epoch": 4.388663967611336, + "grad_norm": 0.0005911945500261623, + "learning_rate": 1.7956315228622446e-06, + "loss": 0.0, + "step": 1084 + }, + { + "epoch": 4.392712550607287, + "grad_norm": 0.0005938860994803431, + "learning_rate": 1.7722835774285553e-06, + "loss": 0.0, + "step": 1085 + }, + { + "epoch": 4.396761133603239, + "grad_norm": 0.000607399388844308, + "learning_rate": 1.7490813803296624e-06, + "loss": 0.0, + "step": 1086 + }, + { + "epoch": 4.40080971659919, + "grad_norm": 0.0006311200674338588, + "learning_rate": 1.726025117089718e-06, + "loss": 0.0, + "step": 1087 + }, + { + "epoch": 4.4048582995951415, + "grad_norm": 0.0005956781368214504, + "learning_rate": 1.70311497206598e-06, + "loss": 0.0, + "step": 1088 + }, + { + "epoch": 4.4089068825910935, + "grad_norm": 0.0006162219250532885, + "learning_rate": 1.6803511284473638e-06, + "loss": 0.0, + "step": 1089 + }, + { + "epoch": 4.412955465587045, + "grad_norm": 0.0005756696851283539, + "learning_rate": 1.6577337682529538e-06, + "loss": 0.0, + "step": 1090 + }, + { + "epoch": 4.417004048582996, + "grad_norm": 0.0006075561305466914, + "learning_rate": 1.6352630723305617e-06, + "loss": 0.0, + "step": 1091 + }, + { + "epoch": 4.421052631578947, + "grad_norm": 0.0005839016962119484, + "learning_rate": 1.612939220355272e-06, + "loss": 0.0, + "step": 1092 + }, + { + "epoch": 4.425101214574899, + "grad_norm": 0.000614770368811097, + "learning_rate": 1.5907623908280112e-06, + "loss": 0.0, + "step": 1093 + }, + { + "epoch": 4.42914979757085, + "grad_norm": 0.0006099896123172694, + "learning_rate": 1.568732761074121e-06, + "loss": 0.0, + "step": 1094 + }, + { + "epoch": 4.433198380566802, + "grad_norm": 0.0005974353775716716, + "learning_rate": 1.5468505072419237e-06, + "loss": 0.0, + "step": 1095 + }, + { + "epoch": 4.437246963562753, + "grad_norm": 0.0005727931754063761, + "learning_rate": 1.5251158043013448e-06, + "loss": 0.0, + "step": 1096 + }, + { + "epoch": 4.441295546558704, + "grad_norm": 0.0005783921155749679, + "learning_rate": 1.5035288260424885e-06, + "loss": 0.0, + "step": 1097 + }, + { + "epoch": 4.445344129554655, + "grad_norm": 0.000592762939731971, + "learning_rate": 1.4820897450742623e-06, + "loss": 0.0, + "step": 1098 + }, + { + "epoch": 4.449392712550607, + "grad_norm": 0.0006182867636556517, + "learning_rate": 1.4607987328229767e-06, + "loss": 0.0, + "step": 1099 + }, + { + "epoch": 4.4534412955465585, + "grad_norm": 0.0006055329093803481, + "learning_rate": 1.4396559595310056e-06, + "loss": 0.0, + "step": 1100 + }, + { + "epoch": 4.4574898785425106, + "grad_norm": 0.0006228707575003964, + "learning_rate": 1.4186615942553973e-06, + "loss": 0.0, + "step": 1101 + }, + { + "epoch": 4.461538461538462, + "grad_norm": 0.0005906650294750883, + "learning_rate": 1.3978158048665424e-06, + "loss": 0.0, + "step": 1102 + }, + { + "epoch": 4.465587044534413, + "grad_norm": 0.0006077366477251441, + "learning_rate": 1.3771187580468115e-06, + "loss": 0.0, + "step": 1103 + }, + { + "epoch": 4.469635627530364, + "grad_norm": 0.0005821437247073783, + "learning_rate": 1.3565706192892392e-06, + "loss": 0.0, + "step": 1104 + }, + { + "epoch": 4.473684210526316, + "grad_norm": 0.0005849661593587785, + "learning_rate": 1.3361715528961993e-06, + "loss": 0.0, + "step": 1105 + }, + { + "epoch": 4.477732793522267, + "grad_norm": 0.0006082329187188347, + "learning_rate": 1.3159217219780817e-06, + "loss": 0.0, + "step": 1106 + }, + { + "epoch": 4.481781376518219, + "grad_norm": 0.0006326501201399662, + "learning_rate": 1.2958212884519949e-06, + "loss": 0.0, + "step": 1107 + }, + { + "epoch": 4.48582995951417, + "grad_norm": 0.0006312225046514419, + "learning_rate": 1.2758704130404675e-06, + "loss": 0.0, + "step": 1108 + }, + { + "epoch": 4.489878542510121, + "grad_norm": 0.0006080643875469274, + "learning_rate": 1.2560692552701738e-06, + "loss": 0.0, + "step": 1109 + }, + { + "epoch": 4.493927125506072, + "grad_norm": 0.0005618298831161689, + "learning_rate": 1.2364179734706382e-06, + "loss": 0.0, + "step": 1110 + }, + { + "epoch": 4.497975708502024, + "grad_norm": 0.0005848125825922261, + "learning_rate": 1.2169167247729852e-06, + "loss": 0.0, + "step": 1111 + }, + { + "epoch": 4.502024291497976, + "grad_norm": 0.0005874871238603982, + "learning_rate": 1.1975656651086864e-06, + "loss": 0.0, + "step": 1112 + }, + { + "epoch": 4.506072874493928, + "grad_norm": 0.0005898998616950863, + "learning_rate": 1.1783649492082926e-06, + "loss": 0.0, + "step": 1113 + }, + { + "epoch": 4.510121457489879, + "grad_norm": 0.0005761096397421368, + "learning_rate": 1.1593147306002183e-06, + "loss": 0.0, + "step": 1114 + }, + { + "epoch": 4.51417004048583, + "grad_norm": 0.0005585632410331389, + "learning_rate": 1.140415161609505e-06, + "loss": 0.0, + "step": 1115 + }, + { + "epoch": 4.518218623481781, + "grad_norm": 0.0006332722578664566, + "learning_rate": 1.121666393356604e-06, + "loss": 0.0, + "step": 1116 + }, + { + "epoch": 4.522267206477733, + "grad_norm": 0.0006360931047732007, + "learning_rate": 1.1030685757561632e-06, + "loss": 0.0, + "step": 1117 + }, + { + "epoch": 4.526315789473684, + "grad_norm": 0.0005960407009573104, + "learning_rate": 1.0846218575158373e-06, + "loss": 0.0, + "step": 1118 + }, + { + "epoch": 4.530364372469636, + "grad_norm": 0.000616985649674873, + "learning_rate": 1.066326386135097e-06, + "loss": 0.0, + "step": 1119 + }, + { + "epoch": 4.534412955465587, + "grad_norm": 0.0005733758065486644, + "learning_rate": 1.0481823079040398e-06, + "loss": 0.0, + "step": 1120 + }, + { + "epoch": 4.538461538461538, + "grad_norm": 0.0005843872497682361, + "learning_rate": 1.0301897679022344e-06, + "loss": 0.0, + "step": 1121 + }, + { + "epoch": 4.5425101214574894, + "grad_norm": 0.0005942648664981755, + "learning_rate": 1.0123489099975491e-06, + "loss": 0.0, + "step": 1122 + }, + { + "epoch": 4.5465587044534415, + "grad_norm": 0.0006189948701613911, + "learning_rate": 9.9465987684501e-07, + "loss": 0.0, + "step": 1123 + }, + { + "epoch": 4.550607287449393, + "grad_norm": 0.0006159338148178573, + "learning_rate": 9.771228098856534e-07, + "loss": 0.0, + "step": 1124 + }, + { + "epoch": 4.554655870445345, + "grad_norm": 0.0006150284142169607, + "learning_rate": 9.597378493454013e-07, + "loss": 0.0, + "step": 1125 + }, + { + "epoch": 4.558704453441296, + "grad_norm": 0.0006322783284252932, + "learning_rate": 9.425051342339353e-07, + "loss": 0.0, + "step": 1126 + }, + { + "epoch": 4.562753036437247, + "grad_norm": 0.0006075536408291218, + "learning_rate": 9.254248023435864e-07, + "loss": 0.0, + "step": 1127 + }, + { + "epoch": 4.566801619433198, + "grad_norm": 0.0005979691584407369, + "learning_rate": 9.084969902482288e-07, + "loss": 0.0, + "step": 1128 + }, + { + "epoch": 4.57085020242915, + "grad_norm": 0.0005815235007639147, + "learning_rate": 8.917218333021993e-07, + "loss": 0.0, + "step": 1129 + }, + { + "epoch": 4.574898785425101, + "grad_norm": 0.0006197269601551945, + "learning_rate": 8.750994656392109e-07, + "loss": 0.0, + "step": 1130 + }, + { + "epoch": 4.578947368421053, + "grad_norm": 0.0005955932183545657, + "learning_rate": 8.58630020171265e-07, + "loss": 0.0, + "step": 1131 + }, + { + "epoch": 4.582995951417004, + "grad_norm": 0.0005899135519537605, + "learning_rate": 8.423136285876099e-07, + "loss": 0.0, + "step": 1132 + }, + { + "epoch": 4.587044534412955, + "grad_norm": 0.0005727273199498031, + "learning_rate": 8.26150421353682e-07, + "loss": 0.0, + "step": 1133 + }, + { + "epoch": 4.5910931174089065, + "grad_norm": 0.0005886829288119401, + "learning_rate": 8.101405277100549e-07, + "loss": 0.0, + "step": 1134 + }, + { + "epoch": 4.5951417004048585, + "grad_norm": 0.0006217255001614463, + "learning_rate": 7.942840756714077e-07, + "loss": 0.0, + "step": 1135 + }, + { + "epoch": 4.59919028340081, + "grad_norm": 0.0006302181720958444, + "learning_rate": 7.785811920255093e-07, + "loss": 0.0, + "step": 1136 + }, + { + "epoch": 4.603238866396762, + "grad_norm": 0.000579563916067256, + "learning_rate": 7.630320023322024e-07, + "loss": 0.0, + "step": 1137 + }, + { + "epoch": 4.607287449392713, + "grad_norm": 0.0006285945790970845, + "learning_rate": 7.476366309223903e-07, + "loss": 0.0, + "step": 1138 + }, + { + "epoch": 4.611336032388664, + "grad_norm": 0.0006043173705420381, + "learning_rate": 7.323952008970537e-07, + "loss": 0.0, + "step": 1139 + }, + { + "epoch": 4.615384615384615, + "grad_norm": 0.0005996311414384656, + "learning_rate": 7.173078341262618e-07, + "loss": 0.0, + "step": 1140 + }, + { + "epoch": 4.619433198380567, + "grad_norm": 0.0005841288654131844, + "learning_rate": 7.023746512481989e-07, + "loss": 0.0, + "step": 1141 + }, + { + "epoch": 4.623481781376518, + "grad_norm": 0.0005780766782994903, + "learning_rate": 6.875957716681902e-07, + "loss": 0.0, + "step": 1142 + }, + { + "epoch": 4.62753036437247, + "grad_norm": 0.000671209925161776, + "learning_rate": 6.729713135577687e-07, + "loss": 0.0, + "step": 1143 + }, + { + "epoch": 4.631578947368421, + "grad_norm": 0.0006175901407099344, + "learning_rate": 6.585013938537143e-07, + "loss": 0.0, + "step": 1144 + }, + { + "epoch": 4.635627530364372, + "grad_norm": 0.0006340564082690305, + "learning_rate": 6.441861282571115e-07, + "loss": 0.0, + "step": 1145 + }, + { + "epoch": 4.6396761133603235, + "grad_norm": 0.0005983927980569957, + "learning_rate": 6.300256312324471e-07, + "loss": 0.0, + "step": 1146 + }, + { + "epoch": 4.6437246963562755, + "grad_norm": 0.0005767823980468125, + "learning_rate": 6.160200160066776e-07, + "loss": 0.0, + "step": 1147 + }, + { + "epoch": 4.647773279352227, + "grad_norm": 0.0006117041995134146, + "learning_rate": 6.021693945683327e-07, + "loss": 0.0, + "step": 1148 + }, + { + "epoch": 4.651821862348179, + "grad_norm": 0.0006101860587520911, + "learning_rate": 5.884738776666044e-07, + "loss": 0.0, + "step": 1149 + }, + { + "epoch": 4.65587044534413, + "grad_norm": 0.0005919798709634722, + "learning_rate": 5.749335748104878e-07, + "loss": 0.0, + "step": 1150 + }, + { + "epoch": 4.659919028340081, + "grad_norm": 0.0006024294807045528, + "learning_rate": 5.615485942678778e-07, + "loss": 0.0, + "step": 1151 + }, + { + "epoch": 4.663967611336032, + "grad_norm": 0.0006132624925588825, + "learning_rate": 5.483190430647289e-07, + "loss": 0.0, + "step": 1152 + }, + { + "epoch": 4.668016194331984, + "grad_norm": 0.00057678565386801, + "learning_rate": 5.352450269841747e-07, + "loss": 0.0, + "step": 1153 + }, + { + "epoch": 4.672064777327935, + "grad_norm": 0.0006186394410979935, + "learning_rate": 5.22326650565701e-07, + "loss": 0.0, + "step": 1154 + }, + { + "epoch": 4.676113360323887, + "grad_norm": 0.0005902944150653909, + "learning_rate": 5.095640171043048e-07, + "loss": 0.0, + "step": 1155 + }, + { + "epoch": 4.680161943319838, + "grad_norm": 0.0005993021772950403, + "learning_rate": 4.969572286496594e-07, + "loss": 0.0, + "step": 1156 + }, + { + "epoch": 4.684210526315789, + "grad_norm": 0.0006215287705734527, + "learning_rate": 4.845063860053079e-07, + "loss": 0.0, + "step": 1157 + }, + { + "epoch": 4.6882591093117405, + "grad_norm": 0.000600038209860177, + "learning_rate": 4.7221158872785997e-07, + "loss": 0.0, + "step": 1158 + }, + { + "epoch": 4.6923076923076925, + "grad_norm": 0.0006165075496675, + "learning_rate": 4.600729351261857e-07, + "loss": 0.0, + "step": 1159 + }, + { + "epoch": 4.696356275303644, + "grad_norm": 0.0006250910889058045, + "learning_rate": 4.4809052226063355e-07, + "loss": 0.0, + "step": 1160 + }, + { + "epoch": 4.700404858299595, + "grad_norm": 0.0006379515505008557, + "learning_rate": 4.3626444594226046e-07, + "loss": 0.0, + "step": 1161 + }, + { + "epoch": 4.704453441295547, + "grad_norm": 0.000599614425864646, + "learning_rate": 4.245948007320588e-07, + "loss": 0.0, + "step": 1162 + }, + { + "epoch": 4.708502024291498, + "grad_norm": 0.0005868832934917119, + "learning_rate": 4.130816799401971e-07, + "loss": 0.0, + "step": 1163 + }, + { + "epoch": 4.712550607287449, + "grad_norm": 0.0005856351968412419, + "learning_rate": 4.017251756252827e-07, + "loss": 0.0, + "step": 1164 + }, + { + "epoch": 4.716599190283401, + "grad_norm": 0.000634635581775992, + "learning_rate": 3.9052537859362473e-07, + "loss": 0.0, + "step": 1165 + }, + { + "epoch": 4.720647773279352, + "grad_norm": 0.0005886926715250295, + "learning_rate": 3.7948237839849687e-07, + "loss": 0.0, + "step": 1166 + }, + { + "epoch": 4.724696356275303, + "grad_norm": 0.0005879250156254251, + "learning_rate": 3.6859626333943797e-07, + "loss": 0.0, + "step": 1167 + }, + { + "epoch": 4.728744939271255, + "grad_norm": 0.0006000565555566223, + "learning_rate": 3.578671204615347e-07, + "loss": 0.0, + "step": 1168 + }, + { + "epoch": 4.732793522267206, + "grad_norm": 0.0005788665169839752, + "learning_rate": 3.472950355547267e-07, + "loss": 0.0, + "step": 1169 + }, + { + "epoch": 4.7368421052631575, + "grad_norm": 0.0006026444925370789, + "learning_rate": 3.3688009315312244e-07, + "loss": 0.0, + "step": 1170 + }, + { + "epoch": 4.7408906882591095, + "grad_norm": 0.000585892703775998, + "learning_rate": 3.266223765343224e-07, + "loss": 0.0, + "step": 1171 + }, + { + "epoch": 4.744939271255061, + "grad_norm": 0.000610959030279406, + "learning_rate": 3.16521967718757e-07, + "loss": 0.0, + "step": 1172 + }, + { + "epoch": 4.748987854251012, + "grad_norm": 0.0005874854776296095, + "learning_rate": 3.0657894746902726e-07, + "loss": 0.0, + "step": 1173 + }, + { + "epoch": 4.753036437246964, + "grad_norm": 0.0006369142381279653, + "learning_rate": 2.9679339528925654e-07, + "loss": 0.0, + "step": 1174 + }, + { + "epoch": 4.757085020242915, + "grad_norm": 0.0005766739855207077, + "learning_rate": 2.871653894244597e-07, + "loss": 0.0, + "step": 1175 + }, + { + "epoch": 4.761133603238866, + "grad_norm": 0.0006250058517508081, + "learning_rate": 2.7769500685991266e-07, + "loss": 0.0, + "step": 1176 + }, + { + "epoch": 4.765182186234818, + "grad_norm": 0.000577568399040822, + "learning_rate": 2.683823233205485e-07, + "loss": 0.0, + "step": 1177 + }, + { + "epoch": 4.769230769230769, + "grad_norm": 0.0006179568753314737, + "learning_rate": 2.592274132703287e-07, + "loss": 0.0, + "step": 1178 + }, + { + "epoch": 4.77327935222672, + "grad_norm": 0.0005593073853022722, + "learning_rate": 2.502303499116754e-07, + "loss": 0.0, + "step": 1179 + }, + { + "epoch": 4.777327935222672, + "grad_norm": 0.000656209756728621, + "learning_rate": 2.413912051848666e-07, + "loss": 0.0, + "step": 1180 + }, + { + "epoch": 4.781376518218623, + "grad_norm": 0.0005853927989326852, + "learning_rate": 2.3271004976746613e-07, + "loss": 0.0, + "step": 1181 + }, + { + "epoch": 4.7854251012145745, + "grad_norm": 0.0005899689800598323, + "learning_rate": 2.2418695307376393e-07, + "loss": 0.0, + "step": 1182 + }, + { + "epoch": 4.7894736842105265, + "grad_norm": 0.000574766587009232, + "learning_rate": 2.1582198325420968e-07, + "loss": 0.0, + "step": 1183 + }, + { + "epoch": 4.793522267206478, + "grad_norm": 0.0005809585932743954, + "learning_rate": 2.0761520719488447e-07, + "loss": 0.0, + "step": 1184 + }, + { + "epoch": 4.797570850202429, + "grad_norm": 0.0006279649226488172, + "learning_rate": 1.9956669051694798e-07, + "loss": 0.0, + "step": 1185 + }, + { + "epoch": 4.801619433198381, + "grad_norm": 0.0006148227004877684, + "learning_rate": 1.916764975761254e-07, + "loss": 0.0, + "step": 1186 + }, + { + "epoch": 4.805668016194332, + "grad_norm": 0.0005902762177897672, + "learning_rate": 1.839446914621923e-07, + "loss": 0.0, + "step": 1187 + }, + { + "epoch": 4.809716599190283, + "grad_norm": 0.0006059964615423422, + "learning_rate": 1.7637133399845962e-07, + "loss": 0.0, + "step": 1188 + }, + { + "epoch": 4.813765182186235, + "grad_norm": 0.0006011711941751824, + "learning_rate": 1.6895648574129397e-07, + "loss": 0.0, + "step": 1189 + }, + { + "epoch": 4.817813765182186, + "grad_norm": 0.0005974274069011393, + "learning_rate": 1.6170020597962465e-07, + "loss": 0.0, + "step": 1190 + }, + { + "epoch": 4.821862348178137, + "grad_norm": 0.000611866487630383, + "learning_rate": 1.5460255273447078e-07, + "loss": 0.0, + "step": 1191 + }, + { + "epoch": 4.825910931174089, + "grad_norm": 0.0005790504792108821, + "learning_rate": 1.4766358275847715e-07, + "loss": 0.0, + "step": 1192 + }, + { + "epoch": 4.82995951417004, + "grad_norm": 0.0005854871902112542, + "learning_rate": 1.4088335153546573e-07, + "loss": 0.0, + "step": 1193 + }, + { + "epoch": 4.834008097165992, + "grad_norm": 0.0006173319981470624, + "learning_rate": 1.3426191327998496e-07, + "loss": 0.0, + "step": 1194 + }, + { + "epoch": 4.838056680161944, + "grad_norm": 0.0006309881260558489, + "learning_rate": 1.277993209368744e-07, + "loss": 0.0, + "step": 1195 + }, + { + "epoch": 4.842105263157895, + "grad_norm": 0.0006066080256705906, + "learning_rate": 1.214956261808542e-07, + "loss": 0.0, + "step": 1196 + }, + { + "epoch": 4.846153846153846, + "grad_norm": 0.0006056684096511928, + "learning_rate": 1.1535087941609846e-07, + "loss": 0.0, + "step": 1197 + }, + { + "epoch": 4.850202429149798, + "grad_norm": 0.0005707393769768048, + "learning_rate": 1.0936512977583802e-07, + "loss": 0.0, + "step": 1198 + }, + { + "epoch": 4.854251012145749, + "grad_norm": 0.0006332097223366787, + "learning_rate": 1.0353842512196732e-07, + "loss": 0.0, + "step": 1199 + }, + { + "epoch": 4.8582995951417, + "grad_norm": 0.0005929670500990227, + "learning_rate": 9.787081204466253e-08, + "loss": 0.0, + "step": 1200 + }, + { + "epoch": 4.862348178137652, + "grad_norm": 0.0005865261587131307, + "learning_rate": 9.236233586200627e-08, + "loss": 0.0, + "step": 1201 + }, + { + "epoch": 4.866396761133603, + "grad_norm": 0.0006503166819092894, + "learning_rate": 8.70130406196279e-08, + "loss": 0.0, + "step": 1202 + }, + { + "epoch": 4.870445344129554, + "grad_norm": 0.0006039153632752846, + "learning_rate": 8.182296909035047e-08, + "loss": 0.0, + "step": 1203 + }, + { + "epoch": 4.874493927125506, + "grad_norm": 0.0005940682264946841, + "learning_rate": 7.679216277384439e-08, + "loss": 0.0, + "step": 1204 + }, + { + "epoch": 4.8785425101214575, + "grad_norm": 0.0005736519140830359, + "learning_rate": 7.19206618963053e-08, + "loss": 0.0, + "step": 1205 + }, + { + "epoch": 4.882591093117409, + "grad_norm": 0.0006096474707139318, + "learning_rate": 6.720850541012347e-08, + "loss": 0.0, + "step": 1206 + }, + { + "epoch": 4.886639676113361, + "grad_norm": 0.000607317723404939, + "learning_rate": 6.265573099357714e-08, + "loss": 0.0, + "step": 1207 + }, + { + "epoch": 4.890688259109312, + "grad_norm": 0.0006283153240108185, + "learning_rate": 5.826237505053067e-08, + "loss": 0.0, + "step": 1208 + }, + { + "epoch": 4.894736842105263, + "grad_norm": 0.0006090051852382628, + "learning_rate": 5.4028472710137e-08, + "loss": 0.0, + "step": 1209 + }, + { + "epoch": 4.898785425101215, + "grad_norm": 0.0005796069271065241, + "learning_rate": 4.9954057826571146e-08, + "loss": 0.0, + "step": 1210 + }, + { + "epoch": 4.902834008097166, + "grad_norm": 0.0006049346595066373, + "learning_rate": 4.6039162978743776e-08, + "loss": 0.0, + "step": 1211 + }, + { + "epoch": 4.906882591093117, + "grad_norm": 0.0005885286263151249, + "learning_rate": 4.228381947004812e-08, + "loss": 0.0, + "step": 1212 + }, + { + "epoch": 4.910931174089069, + "grad_norm": 0.0006095230906253506, + "learning_rate": 3.868805732811343e-08, + "loss": 0.0, + "step": 1213 + }, + { + "epoch": 4.91497975708502, + "grad_norm": 0.0006222604598722839, + "learning_rate": 3.525190530455636e-08, + "loss": 0.0, + "step": 1214 + }, + { + "epoch": 4.919028340080971, + "grad_norm": 0.0005888978945670181, + "learning_rate": 3.197539087475887e-08, + "loss": 0.0, + "step": 1215 + }, + { + "epoch": 4.923076923076923, + "grad_norm": 0.0005881085527095157, + "learning_rate": 2.8858540237641786e-08, + "loss": 0.0, + "step": 1216 + }, + { + "epoch": 4.9271255060728745, + "grad_norm": 0.0006437538709122607, + "learning_rate": 2.5901378315458247e-08, + "loss": 0.0, + "step": 1217 + }, + { + "epoch": 4.931174089068826, + "grad_norm": 0.0005580188976378538, + "learning_rate": 2.3103928753598347e-08, + "loss": 0.0, + "step": 1218 + }, + { + "epoch": 4.935222672064778, + "grad_norm": 0.0005916396872469307, + "learning_rate": 2.0466213920393718e-08, + "loss": 0.0, + "step": 1219 + }, + { + "epoch": 4.939271255060729, + "grad_norm": 0.0006066744913368516, + "learning_rate": 1.7988254906942115e-08, + "loss": 0.0, + "step": 1220 + }, + { + "epoch": 4.94331983805668, + "grad_norm": 0.0005616694296764241, + "learning_rate": 1.5670071526936447e-08, + "loss": 0.0, + "step": 1221 + }, + { + "epoch": 4.947368421052632, + "grad_norm": 0.0005882498598412197, + "learning_rate": 1.3511682316509344e-08, + "loss": 0.0, + "step": 1222 + }, + { + "epoch": 4.951417004048583, + "grad_norm": 0.0006248755037612498, + "learning_rate": 1.1513104534086605e-08, + "loss": 0.0, + "step": 1223 + }, + { + "epoch": 4.955465587044534, + "grad_norm": 0.0005918173550893698, + "learning_rate": 9.67435416023843e-09, + "loss": 0.0, + "step": 1224 + }, + { + "epoch": 4.959514170040486, + "grad_norm": 0.0006299515194063453, + "learning_rate": 7.995445897566178e-09, + "loss": 0.0, + "step": 1225 + }, + { + "epoch": 4.963562753036437, + "grad_norm": 0.0006280803079000345, + "learning_rate": 6.476393170578021e-09, + "loss": 0.0, + "step": 1226 + }, + { + "epoch": 4.967611336032388, + "grad_norm": 0.0006105893795200474, + "learning_rate": 5.11720812557348e-09, + "loss": 0.0, + "step": 1227 + }, + { + "epoch": 4.97165991902834, + "grad_norm": 0.0006533767922637341, + "learning_rate": 3.917901630563492e-09, + "loss": 0.0, + "step": 1228 + }, + { + "epoch": 4.9757085020242915, + "grad_norm": 0.0006060747700378738, + "learning_rate": 2.878483275170485e-09, + "loss": 0.0, + "step": 1229 + }, + { + "epoch": 4.979757085020243, + "grad_norm": 0.0005842313857652465, + "learning_rate": 1.998961370555108e-09, + "loss": 0.0, + "step": 1230 + }, + { + "epoch": 4.983805668016195, + "grad_norm": 0.0005905493537970519, + "learning_rate": 1.2793429493518361e-09, + "loss": 0.0, + "step": 1231 + }, + { + "epoch": 4.987854251012146, + "grad_norm": 0.0005671009351108718, + "learning_rate": 7.196337656112384e-10, + "loss": 0.0, + "step": 1232 + }, + { + "epoch": 4.991902834008097, + "grad_norm": 0.0006093483031431026, + "learning_rate": 3.198382947511292e-10, + "loss": 0.0, + "step": 1233 + }, + { + "epoch": 4.995951417004049, + "grad_norm": 0.0005589094450204423, + "learning_rate": 7.995973352770137e-11, + "loss": 0.0, + "step": 1234 + }, + { + "epoch": 5.0, + "grad_norm": 0.0005912690051679989, + "learning_rate": 0.0, + "loss": 0.0, + "step": 1235 + }, + { + "epoch": 5.0, + "step": 1235, + "total_flos": 2.3839246561983857e+18, + "train_loss": 0.09312284464515681, + "train_runtime": 49075.0282, + "train_samples_per_second": 3.22, + "train_steps_per_second": 0.025 + } + ], + "logging_steps": 1.0, + "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": 2.3839246561983857e+18, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}