diff --git "a/trainer_state.json" "b/trainer_state.json" --- "a/trainer_state.json" +++ "b/trainer_state.json" @@ -10,8657 +10,8657 @@ "log_history": [ { "epoch": 0.004048582995951417, - "grad_norm": 8.904837529217238, + "grad_norm": 7.408691527358415, "learning_rate": 3.2258064516129035e-07, - "loss": 1.6221, + "loss": 1.3995, "step": 1 }, { "epoch": 0.008097165991902834, - "grad_norm": 8.76839644538671, + "grad_norm": 7.3491348766342846, "learning_rate": 6.451612903225807e-07, - "loss": 1.6157, + "loss": 1.3765, "step": 2 }, { "epoch": 0.012145748987854251, - "grad_norm": 8.684413390149196, + "grad_norm": 7.434161959303196, "learning_rate": 9.67741935483871e-07, - "loss": 1.6613, + "loss": 1.4601, "step": 3 }, { "epoch": 0.016194331983805668, - "grad_norm": 8.46500455720241, + "grad_norm": 7.083460170977558, "learning_rate": 1.2903225806451614e-06, - "loss": 1.6262, + "loss": 1.3903, "step": 4 }, { "epoch": 0.020242914979757085, - "grad_norm": 8.063098204633494, + "grad_norm": 6.822349788146192, "learning_rate": 1.6129032258064516e-06, - "loss": 1.6023, + "loss": 1.3758, "step": 5 }, { "epoch": 0.024291497975708502, - "grad_norm": 8.020573314845281, + "grad_norm": 6.670422955710694, "learning_rate": 1.935483870967742e-06, - "loss": 1.5848, + "loss": 1.3761, "step": 6 }, { "epoch": 0.02834008097165992, - "grad_norm": 6.496061404314003, + "grad_norm": 5.1885545657135115, "learning_rate": 2.2580645161290324e-06, - "loss": 1.5517, + "loss": 1.3435, "step": 7 }, { "epoch": 0.032388663967611336, - "grad_norm": 6.023368541171357, + "grad_norm": 4.691011420689284, "learning_rate": 2.580645161290323e-06, - "loss": 1.5369, + "loss": 1.3308, "step": 8 }, { "epoch": 0.03643724696356275, - "grad_norm": 3.5908767868124523, + "grad_norm": 2.532141797228516, "learning_rate": 2.903225806451613e-06, - "loss": 1.4968, + "loss": 1.293, "step": 9 }, { "epoch": 0.04048582995951417, - "grad_norm": 3.218263191011134, + "grad_norm": 2.2850518449463064, "learning_rate": 3.225806451612903e-06, - "loss": 1.4586, + "loss": 1.2673, "step": 10 }, { "epoch": 0.044534412955465584, - "grad_norm": 2.9778464144015135, + "grad_norm": 2.1296814164814113, "learning_rate": 3.548387096774194e-06, - "loss": 1.4714, + "loss": 1.2769, "step": 11 }, { "epoch": 0.048582995951417005, - "grad_norm": 5.844242109428821, + "grad_norm": 6.121477662020106, "learning_rate": 3.870967741935484e-06, - "loss": 1.4783, + "loss": 1.2934, "step": 12 }, { "epoch": 0.05263157894736842, - "grad_norm": 6.117187623859754, + "grad_norm": 6.483932494484684, "learning_rate": 4.193548387096774e-06, - "loss": 1.477, + "loss": 1.2899, "step": 13 }, { "epoch": 0.05668016194331984, - "grad_norm": 5.858192813836667, + "grad_norm": 6.207842169892318, "learning_rate": 4.516129032258065e-06, - "loss": 1.4858, + "loss": 1.295, "step": 14 }, { "epoch": 0.06072874493927125, - "grad_norm": 5.503802778578249, + "grad_norm": 5.528000619898938, "learning_rate": 4.838709677419355e-06, - "loss": 1.4483, + "loss": 1.2745, "step": 15 }, { "epoch": 0.06477732793522267, - "grad_norm": 3.630346651724757, + "grad_norm": 3.0986083970945297, "learning_rate": 5.161290322580646e-06, - "loss": 1.3883, + "loss": 1.2224, "step": 16 }, { "epoch": 0.06882591093117409, - "grad_norm": 3.0063134935298517, + "grad_norm": 2.3779277734838873, "learning_rate": 5.483870967741935e-06, - "loss": 1.3735, + "loss": 1.2011, "step": 17 }, { "epoch": 0.0728744939271255, - "grad_norm": 2.2467968001298835, + "grad_norm": 2.1323308018280933, "learning_rate": 5.806451612903226e-06, - "loss": 1.3653, + "loss": 1.2219, "step": 18 }, { "epoch": 0.07692307692307693, - "grad_norm": 1.909865524009955, + "grad_norm": 2.034526346746421, "learning_rate": 6.129032258064517e-06, - "loss": 1.3734, + "loss": 1.2197, "step": 19 }, { "epoch": 0.08097165991902834, - "grad_norm": 1.6142065786265911, + "grad_norm": 1.7311924396749865, "learning_rate": 6.451612903225806e-06, - "loss": 1.3279, + "loss": 1.202, "step": 20 }, { "epoch": 0.08502024291497975, - "grad_norm": 1.6584760602059594, + "grad_norm": 1.3130620855147759, "learning_rate": 6.774193548387097e-06, - "loss": 1.3102, + "loss": 1.1643, "step": 21 }, { "epoch": 0.08906882591093117, - "grad_norm": 1.537567588796272, + "grad_norm": 1.3284224125882784, "learning_rate": 7.096774193548388e-06, - "loss": 1.3213, + "loss": 1.1698, "step": 22 }, { "epoch": 0.0931174089068826, - "grad_norm": 1.7543442570470116, + "grad_norm": 1.3801512182027635, "learning_rate": 7.4193548387096784e-06, - "loss": 1.287, + "loss": 1.1418, "step": 23 }, { "epoch": 0.09716599190283401, - "grad_norm": 1.3548591693541097, + "grad_norm": 1.1093910271566902, "learning_rate": 7.741935483870968e-06, - "loss": 1.2954, + "loss": 1.1216, "step": 24 }, { "epoch": 0.10121457489878542, - "grad_norm": 1.5409888155548248, + "grad_norm": 1.149560921363145, "learning_rate": 8.064516129032258e-06, - "loss": 1.307, + "loss": 1.1684, "step": 25 }, { "epoch": 0.10526315789473684, - "grad_norm": 1.3996634971334376, + "grad_norm": 1.2023462837967012, "learning_rate": 8.387096774193549e-06, - "loss": 1.2806, + "loss": 1.1461, "step": 26 }, { "epoch": 0.10931174089068826, - "grad_norm": 1.2859003955624342, + "grad_norm": 0.9519022314022481, "learning_rate": 8.70967741935484e-06, - "loss": 1.2696, + "loss": 1.1318, "step": 27 }, { "epoch": 0.11336032388663968, - "grad_norm": 1.2846094486627682, + "grad_norm": 1.0409954585120875, "learning_rate": 9.03225806451613e-06, - "loss": 1.2506, + "loss": 1.0862, "step": 28 }, { "epoch": 0.11740890688259109, - "grad_norm": 1.2532052531781255, + "grad_norm": 0.9829580982924351, "learning_rate": 9.35483870967742e-06, - "loss": 1.2536, + "loss": 1.1041, "step": 29 }, { "epoch": 0.1214574898785425, - "grad_norm": 1.236262077500035, + "grad_norm": 0.8820055206541261, "learning_rate": 9.67741935483871e-06, - "loss": 1.2687, + "loss": 1.1179, "step": 30 }, { "epoch": 0.12550607287449392, - "grad_norm": 1.326397134096534, + "grad_norm": 0.9227104843503833, "learning_rate": 1e-05, - "loss": 1.2606, + "loss": 1.1041, "step": 31 }, { "epoch": 0.12955465587044535, - "grad_norm": 1.172920507059507, + "grad_norm": 0.8889812317323886, "learning_rate": 1.0322580645161291e-05, - "loss": 1.2878, + "loss": 1.1312, "step": 32 }, { "epoch": 0.13360323886639677, - "grad_norm": 1.2895767640915314, + "grad_norm": 0.9316405762417062, "learning_rate": 1.0645161290322582e-05, - "loss": 1.2187, + "loss": 1.0763, "step": 33 }, { "epoch": 0.13765182186234817, - "grad_norm": 1.0395752157086429, + "grad_norm": 0.9622031009167106, "learning_rate": 1.096774193548387e-05, - "loss": 1.2769, + "loss": 1.1172, "step": 34 }, { "epoch": 0.1417004048582996, - "grad_norm": 1.0027760993473553, + "grad_norm": 0.8687090144971021, "learning_rate": 1.1290322580645164e-05, - "loss": 1.2363, + "loss": 1.0972, "step": 35 }, { "epoch": 0.145748987854251, - "grad_norm": 1.0434707822858609, + "grad_norm": 0.792800349281562, "learning_rate": 1.1612903225806453e-05, - "loss": 1.2122, + "loss": 1.0628, "step": 36 }, { "epoch": 0.14979757085020243, - "grad_norm": 1.0612490355101944, + "grad_norm": 0.8962261253552802, "learning_rate": 1.1935483870967743e-05, - "loss": 1.2135, + "loss": 1.0638, "step": 37 }, { "epoch": 0.15384615384615385, - "grad_norm": 0.894488961179703, + "grad_norm": 0.7334136695664331, "learning_rate": 1.2258064516129034e-05, - "loss": 1.2121, + "loss": 1.0714, "step": 38 }, { "epoch": 0.15789473684210525, - "grad_norm": 0.8261642896123903, + "grad_norm": 0.8813635971994432, "learning_rate": 1.2580645161290324e-05, - "loss": 1.2093, + "loss": 1.0706, "step": 39 }, { "epoch": 0.16194331983805668, - "grad_norm": 0.8640506323824492, + "grad_norm": 0.6761467220057631, "learning_rate": 1.2903225806451613e-05, - "loss": 1.1958, + "loss": 1.0467, "step": 40 }, { "epoch": 0.1659919028340081, - "grad_norm": 0.9094551010235326, + "grad_norm": 0.8236215339755097, "learning_rate": 1.3225806451612903e-05, - "loss": 1.2017, + "loss": 1.0574, "step": 41 }, { "epoch": 0.1700404858299595, - "grad_norm": 0.7016432717086865, + "grad_norm": 0.7278055520576935, "learning_rate": 1.3548387096774194e-05, - "loss": 1.1979, + "loss": 1.0595, "step": 42 }, { "epoch": 0.17408906882591094, - "grad_norm": 0.9908918903711591, + "grad_norm": 0.6733915254774951, "learning_rate": 1.3870967741935486e-05, - "loss": 1.2036, + "loss": 1.0519, "step": 43 }, { "epoch": 0.17813765182186234, - "grad_norm": 0.7144171223378187, + "grad_norm": 0.7931404617753777, "learning_rate": 1.4193548387096776e-05, - "loss": 1.1786, + "loss": 1.0493, "step": 44 }, { "epoch": 0.18218623481781376, - "grad_norm": 0.8848464063667321, + "grad_norm": 0.8372800068529676, "learning_rate": 1.4516129032258066e-05, - "loss": 1.2065, + "loss": 1.0743, "step": 45 }, { "epoch": 0.1862348178137652, - "grad_norm": 0.7443902040561414, + "grad_norm": 0.6530056477352766, "learning_rate": 1.4838709677419357e-05, - "loss": 1.2152, + "loss": 1.0596, "step": 46 }, { "epoch": 0.1902834008097166, - "grad_norm": 1.0144974241541531, + "grad_norm": 0.8910362994082557, "learning_rate": 1.5161290322580646e-05, - "loss": 1.1882, + "loss": 1.0481, "step": 47 }, { "epoch": 0.19433198380566802, - "grad_norm": 0.6980594441249316, + "grad_norm": 0.7399353184000506, "learning_rate": 1.5483870967741936e-05, - "loss": 1.2131, + "loss": 1.0593, "step": 48 }, { "epoch": 0.19838056680161945, - "grad_norm": 0.7514112478071557, + "grad_norm": 0.7734092910054149, "learning_rate": 1.5806451612903226e-05, - "loss": 1.1745, + "loss": 1.0389, "step": 49 }, { "epoch": 0.20242914979757085, - "grad_norm": 0.7894931975430158, + "grad_norm": 0.7258580136827824, "learning_rate": 1.6129032258064517e-05, - "loss": 1.2171, + "loss": 1.0629, "step": 50 }, { "epoch": 0.20647773279352227, - "grad_norm": 0.8829818396513377, + "grad_norm": 0.7333801210002687, "learning_rate": 1.6451612903225807e-05, - "loss": 1.1841, + "loss": 1.0357, "step": 51 }, { "epoch": 0.21052631578947367, - "grad_norm": 0.9547125420627887, + "grad_norm": 0.7336090194336299, "learning_rate": 1.6774193548387098e-05, - "loss": 1.1718, + "loss": 1.0328, "step": 52 }, { "epoch": 0.2145748987854251, - "grad_norm": 0.8950277624784226, + "grad_norm": 0.8854217601037879, "learning_rate": 1.7096774193548388e-05, - "loss": 1.2293, + "loss": 1.0756, "step": 53 }, { "epoch": 0.21862348178137653, - "grad_norm": 0.8668053710658594, + "grad_norm": 0.8856981657944366, "learning_rate": 1.741935483870968e-05, - "loss": 1.194, + "loss": 1.0217, "step": 54 }, { "epoch": 0.22267206477732793, - "grad_norm": 1.1251366360979487, + "grad_norm": 0.5594716925738864, "learning_rate": 1.774193548387097e-05, - "loss": 1.1864, + "loss": 1.0353, "step": 55 }, { "epoch": 0.22672064777327935, - "grad_norm": 0.7127427617295803, + "grad_norm": 0.8029608167349068, "learning_rate": 1.806451612903226e-05, - "loss": 1.2039, + "loss": 1.0529, "step": 56 }, { "epoch": 0.23076923076923078, - "grad_norm": 1.132376632568397, + "grad_norm": 1.0496652624874956, "learning_rate": 1.838709677419355e-05, - "loss": 1.1792, + "loss": 1.0295, "step": 57 }, { "epoch": 0.23481781376518218, - "grad_norm": 0.800373905171592, + "grad_norm": 0.7929302144172623, "learning_rate": 1.870967741935484e-05, - "loss": 1.1603, + "loss": 1.0113, "step": 58 }, { "epoch": 0.2388663967611336, - "grad_norm": 0.7915550535594651, + "grad_norm": 0.588983718459387, "learning_rate": 1.903225806451613e-05, - "loss": 1.1814, + "loss": 1.025, "step": 59 }, { "epoch": 0.242914979757085, - "grad_norm": 0.8235810716542652, + "grad_norm": 0.6012540574125316, "learning_rate": 1.935483870967742e-05, - "loss": 1.204, + "loss": 1.0526, "step": 60 }, { "epoch": 0.24696356275303644, - "grad_norm": 0.8932460949529389, + "grad_norm": 0.7670090146478642, "learning_rate": 1.967741935483871e-05, - "loss": 1.1932, + "loss": 1.0625, "step": 61 }, { "epoch": 0.25101214574898784, - "grad_norm": 0.7668928245672696, + "grad_norm": 0.7847960082293395, "learning_rate": 2e-05, - "loss": 1.1574, + "loss": 1.0175, "step": 62 }, { "epoch": 0.2550607287449393, - "grad_norm": 0.9492582200050924, + "grad_norm": 0.880091978985326, "learning_rate": 2.0322580645161292e-05, - "loss": 1.1629, + "loss": 1.004, "step": 63 }, { "epoch": 0.2591093117408907, - "grad_norm": 0.8789880488756853, + "grad_norm": 0.9477419147789407, "learning_rate": 2.0645161290322582e-05, - "loss": 1.1953, + "loss": 1.0492, "step": 64 }, { "epoch": 0.2631578947368421, - "grad_norm": 1.0599260710216811, + "grad_norm": 0.7623714963669758, "learning_rate": 2.0967741935483873e-05, - "loss": 1.1796, + "loss": 1.042, "step": 65 }, { "epoch": 0.26720647773279355, - "grad_norm": 0.9010423858317235, + "grad_norm": 0.8780442851482717, "learning_rate": 2.1290322580645163e-05, - "loss": 1.156, + "loss": 1.0038, "step": 66 }, { "epoch": 0.27125506072874495, - "grad_norm": 1.3495974622230575, + "grad_norm": 0.8745939558507999, "learning_rate": 2.161290322580645e-05, - "loss": 1.1582, + "loss": 1.0189, "step": 67 }, { "epoch": 0.27530364372469635, - "grad_norm": 0.9828827789350647, + "grad_norm": 1.1838145154108364, "learning_rate": 2.193548387096774e-05, - "loss": 1.1708, + "loss": 1.0025, "step": 68 }, { "epoch": 0.2793522267206478, - "grad_norm": 0.9594638211383509, + "grad_norm": 1.3506283546428999, "learning_rate": 2.225806451612903e-05, - "loss": 1.1854, + "loss": 1.0352, "step": 69 }, { "epoch": 0.2834008097165992, - "grad_norm": 0.8575953949631719, + "grad_norm": 0.8194707065762359, "learning_rate": 2.2580645161290328e-05, - "loss": 1.1614, + "loss": 1.0188, "step": 70 }, { "epoch": 0.2874493927125506, - "grad_norm": 1.3735282085289946, + "grad_norm": 0.9261425066415593, "learning_rate": 2.2903225806451615e-05, - "loss": 1.1672, + "loss": 1.0226, "step": 71 }, { "epoch": 0.291497975708502, - "grad_norm": 1.140421190671952, + "grad_norm": 1.3051507880198747, "learning_rate": 2.3225806451612906e-05, - "loss": 1.1679, + "loss": 1.0346, "step": 72 }, { "epoch": 0.29554655870445345, - "grad_norm": 1.0930532793669856, + "grad_norm": 1.131916770908472, "learning_rate": 2.3548387096774196e-05, - "loss": 1.1645, + "loss": 1.0214, "step": 73 }, { "epoch": 0.29959514170040485, - "grad_norm": 1.3780859304310613, + "grad_norm": 0.9960974910826763, "learning_rate": 2.3870967741935486e-05, - "loss": 1.148, + "loss": 0.9968, "step": 74 }, { "epoch": 0.30364372469635625, - "grad_norm": 1.1059664119183346, + "grad_norm": 1.0901842005770983, "learning_rate": 2.4193548387096777e-05, - "loss": 1.1677, + "loss": 1.0079, "step": 75 }, { "epoch": 0.3076923076923077, - "grad_norm": 1.1768360581998731, + "grad_norm": 1.3761905598375581, "learning_rate": 2.4516129032258067e-05, - "loss": 1.1675, + "loss": 1.0092, "step": 76 }, { "epoch": 0.3117408906882591, - "grad_norm": 1.2834666003278978, + "grad_norm": 1.4388714641395732, "learning_rate": 2.4838709677419358e-05, - "loss": 1.1805, + "loss": 1.0158, "step": 77 }, { "epoch": 0.3157894736842105, - "grad_norm": 1.0854102904352574, + "grad_norm": 0.9394682813764781, "learning_rate": 2.5161290322580648e-05, - "loss": 1.1724, + "loss": 1.0433, "step": 78 }, { "epoch": 0.31983805668016196, - "grad_norm": 1.0119835277537526, + "grad_norm": 1.8899292364667286, "learning_rate": 2.548387096774194e-05, - "loss": 1.1725, + "loss": 1.0392, "step": 79 }, { "epoch": 0.32388663967611336, - "grad_norm": 1.0303676297618163, + "grad_norm": 0.9299155584723263, "learning_rate": 2.5806451612903226e-05, - "loss": 1.1525, + "loss": 0.991, "step": 80 }, { "epoch": 0.32793522267206476, - "grad_norm": 1.0139933204807439, + "grad_norm": 2.634205747120425, "learning_rate": 2.6129032258064516e-05, - "loss": 1.1732, + "loss": 1.0351, "step": 81 }, { "epoch": 0.3319838056680162, - "grad_norm": 1.1518569985216913, + "grad_norm": 2.1989793219928786, "learning_rate": 2.6451612903225806e-05, - "loss": 1.1826, + "loss": 1.0266, "step": 82 }, { "epoch": 0.3360323886639676, - "grad_norm": 1.323076742640062, + "grad_norm": 1.6047965967975384, "learning_rate": 2.6774193548387097e-05, - "loss": 1.1609, + "loss": 1.0107, "step": 83 }, { "epoch": 0.340080971659919, - "grad_norm": 0.8592933224524022, + "grad_norm": 1.5281389413627957, "learning_rate": 2.7096774193548387e-05, - "loss": 1.1691, + "loss": 1.0266, "step": 84 }, { "epoch": 0.3441295546558704, - "grad_norm": 1.3706521307838602, + "grad_norm": 1.5546608211338266, "learning_rate": 2.7419354838709678e-05, - "loss": 1.1471, + "loss": 0.9898, "step": 85 }, { "epoch": 0.3481781376518219, - "grad_norm": 0.9096574798325738, + "grad_norm": 1.1767692948781652, "learning_rate": 2.774193548387097e-05, - "loss": 1.1872, + "loss": 1.0055, "step": 86 }, { "epoch": 0.3522267206477733, - "grad_norm": 1.2039461728209269, + "grad_norm": 1.3811317270312848, "learning_rate": 2.8064516129032262e-05, - "loss": 1.1452, + "loss": 1.0161, "step": 87 }, { "epoch": 0.3562753036437247, - "grad_norm": 0.8584182757586538, + "grad_norm": 1.5183961679003033, "learning_rate": 2.8387096774193552e-05, - "loss": 1.161, + "loss": 1.0197, "step": 88 }, { "epoch": 0.3603238866396761, - "grad_norm": 1.1288929208765355, + "grad_norm": 1.4918256975279003, "learning_rate": 2.8709677419354843e-05, - "loss": 1.1661, + "loss": 1.0198, "step": 89 }, { "epoch": 0.3643724696356275, - "grad_norm": 0.9814930749724674, + "grad_norm": 1.5531908680093214, "learning_rate": 2.9032258064516133e-05, - "loss": 1.181, + "loss": 1.0417, "step": 90 }, { "epoch": 0.3684210526315789, - "grad_norm": 1.5294318977844779, + "grad_norm": 1.2212252964329473, "learning_rate": 2.9354838709677423e-05, - "loss": 1.1353, + "loss": 0.9803, "step": 91 }, { "epoch": 0.3724696356275304, - "grad_norm": 1.0681821873312574, + "grad_norm": 1.6463522701727396, "learning_rate": 2.9677419354838714e-05, - "loss": 1.1318, + "loss": 0.9894, "step": 92 }, { "epoch": 0.3765182186234818, - "grad_norm": 1.3755196252523367, + "grad_norm": 1.1244454929230798, "learning_rate": 3.0000000000000004e-05, - "loss": 1.167, + "loss": 0.9996, "step": 93 }, { "epoch": 0.3805668016194332, - "grad_norm": 1.2345014067744862, + "grad_norm": 1.9812785097527656, "learning_rate": 3.032258064516129e-05, - "loss": 1.2123, + "loss": 1.0485, "step": 94 }, { "epoch": 0.38461538461538464, - "grad_norm": 1.2375411488612973, + "grad_norm": 1.3950978278267776, "learning_rate": 3.0645161290322585e-05, - "loss": 1.1752, + "loss": 0.9929, "step": 95 }, { "epoch": 0.38866396761133604, - "grad_norm": 1.0404968242044534, + "grad_norm": 1.5410821853494987, "learning_rate": 3.096774193548387e-05, - "loss": 1.1318, + "loss": 0.9731, "step": 96 }, { "epoch": 0.39271255060728744, - "grad_norm": 1.3528633030568809, + "grad_norm": 1.356532012532559, "learning_rate": 3.1290322580645166e-05, - "loss": 1.1447, + "loss": 0.9909, "step": 97 }, { "epoch": 0.3967611336032389, - "grad_norm": 1.01159996875333, + "grad_norm": 1.2637473151865988, "learning_rate": 3.161290322580645e-05, - "loss": 1.1584, + "loss": 0.9845, "step": 98 }, { "epoch": 0.4008097165991903, - "grad_norm": 1.367534134201412, + "grad_norm": 1.4940747292321284, "learning_rate": 3.193548387096774e-05, - "loss": 1.1571, + "loss": 0.9944, "step": 99 }, { "epoch": 0.4048582995951417, - "grad_norm": 1.23107455221171, + "grad_norm": 1.2501253708557871, "learning_rate": 3.2258064516129034e-05, - "loss": 1.1768, + "loss": 1.0115, "step": 100 }, { "epoch": 0.4089068825910931, - "grad_norm": 1.1402104041323293, + "grad_norm": 1.7045097660547837, "learning_rate": 3.258064516129033e-05, - "loss": 1.1634, + "loss": 1.0015, "step": 101 }, { "epoch": 0.41295546558704455, - "grad_norm": 1.2103265548384168, + "grad_norm": 1.2744501685433578, "learning_rate": 3.2903225806451614e-05, - "loss": 1.1428, + "loss": 0.9875, "step": 102 }, { "epoch": 0.41700404858299595, - "grad_norm": 0.9736091156065979, + "grad_norm": 1.4925481831506031, "learning_rate": 3.322580645161291e-05, - "loss": 1.1694, + "loss": 1.0082, "step": 103 }, { "epoch": 0.42105263157894735, - "grad_norm": 1.1047419078876415, + "grad_norm": 1.396963615988694, "learning_rate": 3.3548387096774195e-05, - "loss": 1.1624, + "loss": 1.0014, "step": 104 }, { "epoch": 0.4251012145748988, - "grad_norm": 1.1821402442576308, + "grad_norm": 1.5129005081053877, "learning_rate": 3.387096774193549e-05, - "loss": 1.1754, + "loss": 1.0268, "step": 105 }, { "epoch": 0.4291497975708502, - "grad_norm": 1.3683972475321688, + "grad_norm": 1.1834624520941643, "learning_rate": 3.4193548387096776e-05, - "loss": 1.183, + "loss": 1.022, "step": 106 }, { "epoch": 0.4331983805668016, - "grad_norm": 1.2619402591658873, + "grad_norm": 1.761982035994383, "learning_rate": 3.451612903225807e-05, - "loss": 1.2015, + "loss": 1.0441, "step": 107 }, { "epoch": 0.43724696356275305, - "grad_norm": 1.1264316724448091, + "grad_norm": 1.2448039944994347, "learning_rate": 3.483870967741936e-05, - "loss": 1.1073, + "loss": 0.9664, "step": 108 }, { "epoch": 0.44129554655870445, - "grad_norm": 1.4152557435106927, + "grad_norm": 1.5309054836406017, "learning_rate": 3.516129032258065e-05, - "loss": 1.1541, + "loss": 1.0026, "step": 109 }, { "epoch": 0.44534412955465585, - "grad_norm": 1.0255480579753737, + "grad_norm": 1.2399733355246783, "learning_rate": 3.548387096774194e-05, - "loss": 1.1637, + "loss": 1.0028, "step": 110 }, { "epoch": 0.4493927125506073, - "grad_norm": 1.1147458587952046, + "grad_norm": 1.9466267703932512, "learning_rate": 3.5806451612903225e-05, - "loss": 1.1676, + "loss": 1.0325, "step": 111 }, { "epoch": 0.4534412955465587, - "grad_norm": 0.899011390498638, + "grad_norm": 1.5613548023009343, "learning_rate": 3.612903225806452e-05, - "loss": 1.1746, + "loss": 1.0089, "step": 112 }, { "epoch": 0.4574898785425101, - "grad_norm": 1.2294638415509205, + "grad_norm": 1.5018008916095047, "learning_rate": 3.6451612903225805e-05, - "loss": 1.112, + "loss": 0.9656, "step": 113 }, { "epoch": 0.46153846153846156, - "grad_norm": 1.2713174390678026, + "grad_norm": 1.6165536220202155, "learning_rate": 3.67741935483871e-05, - "loss": 1.143, + "loss": 0.9961, "step": 114 }, { "epoch": 0.46558704453441296, - "grad_norm": 1.3780909089839308, + "grad_norm": 1.5567794627653193, "learning_rate": 3.7096774193548386e-05, - "loss": 1.1791, + "loss": 0.9999, "step": 115 }, { "epoch": 0.46963562753036436, - "grad_norm": 1.0226900605630365, + "grad_norm": 1.7150618174618357, "learning_rate": 3.741935483870968e-05, - "loss": 1.1535, + "loss": 0.9995, "step": 116 }, { "epoch": 0.47368421052631576, - "grad_norm": 1.4580068874108452, + "grad_norm": 1.3386164725927332, "learning_rate": 3.7741935483870974e-05, - "loss": 1.1804, + "loss": 1.0209, "step": 117 }, { "epoch": 0.4777327935222672, - "grad_norm": 1.087731507360505, + "grad_norm": 1.2762194998606182, "learning_rate": 3.806451612903226e-05, - "loss": 1.1578, + "loss": 1.0057, "step": 118 }, { "epoch": 0.4817813765182186, - "grad_norm": 1.2223803181523674, + "grad_norm": 1.9435747023612966, "learning_rate": 3.8387096774193555e-05, - "loss": 1.1507, + "loss": 0.9924, "step": 119 }, { "epoch": 0.48582995951417, - "grad_norm": 1.8590464461944065, + "grad_norm": 1.6016104885179308, "learning_rate": 3.870967741935484e-05, - "loss": 1.1673, + "loss": 0.9919, "step": 120 }, { "epoch": 0.4898785425101215, - "grad_norm": 1.110347784853984, + "grad_norm": 1.352151722747662, "learning_rate": 3.9032258064516135e-05, - "loss": 1.163, + "loss": 0.9965, "step": 121 }, { "epoch": 0.4939271255060729, - "grad_norm": 1.9025400864263613, + "grad_norm": 1.619478594064343, "learning_rate": 3.935483870967742e-05, - "loss": 1.1363, + "loss": 0.9904, "step": 122 }, { "epoch": 0.4979757085020243, - "grad_norm": 1.507893339266089, + "grad_norm": 1.390383011633699, "learning_rate": 3.9677419354838716e-05, - "loss": 1.1729, + "loss": 1.0084, "step": 123 }, { "epoch": 0.5020242914979757, - "grad_norm": 1.6458070841277188, + "grad_norm": 1.376536369028652, "learning_rate": 4e-05, - "loss": 1.1542, + "loss": 1.0087, "step": 124 }, { "epoch": 0.5060728744939271, - "grad_norm": 1.4988674191406195, + "grad_norm": 1.815011890210962, "learning_rate": 3.9999920040266474e-05, - "loss": 1.147, + "loss": 0.9876, "step": 125 }, { "epoch": 0.5101214574898786, - "grad_norm": 1.6778746443795984, + "grad_norm": 1.0883649984434585, "learning_rate": 3.999968016170525e-05, - "loss": 1.172, + "loss": 1.0051, "step": 126 }, { "epoch": 0.5141700404858299, - "grad_norm": 1.2480801082468136, + "grad_norm": 1.4791182466600161, "learning_rate": 3.9999280366234394e-05, - "loss": 1.178, + "loss": 1.027, "step": 127 }, { "epoch": 0.5182186234817814, - "grad_norm": 1.8780370809746492, + "grad_norm": 1.4905222492068115, "learning_rate": 3.999872065705065e-05, - "loss": 1.1378, + "loss": 0.9819, "step": 128 }, { "epoch": 0.5222672064777328, - "grad_norm": 1.336997339081644, + "grad_norm": 1.5581426027657719, "learning_rate": 3.999800103862945e-05, - "loss": 1.1345, + "loss": 0.9753, "step": 129 }, { "epoch": 0.5263157894736842, - "grad_norm": 1.829707361591979, + "grad_norm": 1.8067021542741448, "learning_rate": 3.9997121516724834e-05, - "loss": 1.1436, + "loss": 0.9807, "step": 130 }, { "epoch": 0.5303643724696356, - "grad_norm": 1.4725525304125124, + "grad_norm": 1.1295480773574968, "learning_rate": 3.999608209836944e-05, - "loss": 1.1356, + "loss": 0.9765, "step": 131 }, { "epoch": 0.5344129554655871, - "grad_norm": 1.2507250146173101, + "grad_norm": 1.755852069493054, "learning_rate": 3.999488279187443e-05, - "loss": 1.1331, + "loss": 0.9817, "step": 132 }, { "epoch": 0.5384615384615384, - "grad_norm": 1.2095218735790005, + "grad_norm": 0.7932487768385548, "learning_rate": 3.9993523606829425e-05, - "loss": 1.1373, + "loss": 0.9834, "step": 133 }, { "epoch": 0.5425101214574899, - "grad_norm": 2.0627960131791268, + "grad_norm": 1.5954130810683043, "learning_rate": 3.9992004554102435e-05, - "loss": 1.1784, + "loss": 1.0121, "step": 134 }, { "epoch": 0.5465587044534413, - "grad_norm": 1.4446956627203702, + "grad_norm": 0.8761479668514884, "learning_rate": 3.999032564583976e-05, - "loss": 1.1461, + "loss": 0.9963, "step": 135 }, { "epoch": 0.5506072874493927, - "grad_norm": 2.0245565279630298, + "grad_norm": 1.9652771408350451, "learning_rate": 3.998848689546591e-05, - "loss": 1.1371, + "loss": 0.989, "step": 136 }, { "epoch": 0.5546558704453441, - "grad_norm": 1.935194669216886, + "grad_norm": 1.746892441760076, "learning_rate": 3.9986488317683496e-05, - "loss": 1.1736, + "loss": 1.0, "step": 137 }, { "epoch": 0.5587044534412956, - "grad_norm": 1.6881730510012092, + "grad_norm": 1.2712736391184383, "learning_rate": 3.9984329928473067e-05, - "loss": 1.1353, + "loss": 0.9773, "step": 138 }, { "epoch": 0.562753036437247, - "grad_norm": 1.544795579176874, + "grad_norm": 1.9511587314465966, "learning_rate": 3.998201174509306e-05, - "loss": 1.1464, + "loss": 1.0051, "step": 139 }, { "epoch": 0.5668016194331984, - "grad_norm": 1.3367163687832162, + "grad_norm": 1.661070586857481, "learning_rate": 3.997953378607961e-05, - "loss": 1.1743, + "loss": 1.0257, "step": 140 }, { "epoch": 0.5708502024291497, - "grad_norm": 1.5396566509349454, + "grad_norm": 1.12878395539416, "learning_rate": 3.9976896071246403e-05, - "loss": 1.1488, + "loss": 0.9966, "step": 141 }, { "epoch": 0.5748987854251012, - "grad_norm": 1.2203504453548248, + "grad_norm": 1.9751028716353851, "learning_rate": 3.997409862168454e-05, - "loss": 1.1413, + "loss": 0.975, "step": 142 }, { "epoch": 0.5789473684210527, - "grad_norm": 1.5467910165169128, + "grad_norm": 1.3647102061374379, "learning_rate": 3.997114145976237e-05, - "loss": 1.1421, + "loss": 0.9997, "step": 143 }, { "epoch": 0.582995951417004, - "grad_norm": 1.3545522792618352, + "grad_norm": 1.8262365660508646, "learning_rate": 3.9968024609125245e-05, - "loss": 1.1643, + "loss": 0.9903, "step": 144 }, { "epoch": 0.5870445344129555, - "grad_norm": 0.9789898430136221, + "grad_norm": 1.4502095147320848, "learning_rate": 3.996474809469545e-05, - "loss": 1.1406, + "loss": 0.9833, "step": 145 }, { "epoch": 0.5910931174089069, - "grad_norm": 1.4219467137808297, + "grad_norm": 1.3616108287315498, "learning_rate": 3.996131194267189e-05, - "loss": 1.143, + "loss": 0.9819, "step": 146 }, { "epoch": 0.5951417004048583, - "grad_norm": 0.8972229447552691, + "grad_norm": 1.52113371267984, "learning_rate": 3.995771618052995e-05, - "loss": 1.1302, + "loss": 0.9777, "step": 147 }, { "epoch": 0.5991902834008097, - "grad_norm": 1.0586185122916958, + "grad_norm": 1.2478868678229602, "learning_rate": 3.9953960837021256e-05, - "loss": 1.1853, + "loss": 1.0092, "step": 148 }, { "epoch": 0.6032388663967612, - "grad_norm": 1.2859152068582889, + "grad_norm": 1.3463650214383267, "learning_rate": 3.995004594217343e-05, - "loss": 1.1346, + "loss": 0.9848, "step": 149 }, { "epoch": 0.6072874493927125, - "grad_norm": 1.1435188619784535, + "grad_norm": 1.3005272853479581, "learning_rate": 3.994597152728987e-05, - "loss": 1.1658, + "loss": 0.9916, "step": 150 }, { "epoch": 0.611336032388664, - "grad_norm": 1.36724328438588, + "grad_norm": 1.461826790403466, "learning_rate": 3.994173762494947e-05, - "loss": 1.1105, + "loss": 0.952, "step": 151 }, { "epoch": 0.6153846153846154, - "grad_norm": 1.089362464336718, + "grad_norm": 1.2195262899084016, "learning_rate": 3.9937344269006424e-05, - "loss": 1.1404, + "loss": 0.9799, "step": 152 }, { "epoch": 0.6194331983805668, - "grad_norm": 1.285464649584717, + "grad_norm": 1.8219416656153171, "learning_rate": 3.993279149458988e-05, - "loss": 1.1793, + "loss": 1.01, "step": 153 }, { "epoch": 0.6234817813765182, - "grad_norm": 1.0640340690535945, + "grad_norm": 1.008599228954731, "learning_rate": 3.9928079338103695e-05, - "loss": 1.1168, + "loss": 0.9732, "step": 154 }, { "epoch": 0.6275303643724697, - "grad_norm": 1.3358153248093656, + "grad_norm": 1.989235085684208, "learning_rate": 3.992320783722616e-05, - "loss": 1.1309, + "loss": 0.9673, "step": 155 }, { "epoch": 0.631578947368421, - "grad_norm": 1.0939294230759464, + "grad_norm": 1.5808809507164603, "learning_rate": 3.991817703090965e-05, - "loss": 1.133, + "loss": 0.9742, "step": 156 }, { "epoch": 0.6356275303643725, - "grad_norm": 1.1309838733670428, + "grad_norm": 1.8594219651600195, "learning_rate": 3.991298695938038e-05, - "loss": 1.1915, + "loss": 1.0233, "step": 157 }, { "epoch": 0.6396761133603239, - "grad_norm": 1.0070901283908733, + "grad_norm": 1.861658202238838, "learning_rate": 3.9907637664138e-05, - "loss": 1.1431, + "loss": 0.9954, "step": 158 }, { "epoch": 0.6437246963562753, - "grad_norm": 1.1787481801068096, + "grad_norm": 1.8062508120275291, "learning_rate": 3.990212918795534e-05, - "loss": 1.1611, + "loss": 1.0003, "step": 159 }, { "epoch": 0.6477732793522267, - "grad_norm": 1.0283842450107852, + "grad_norm": 1.6887905186160936, "learning_rate": 3.9896461574878036e-05, - "loss": 1.1204, + "loss": 0.9643, "step": 160 }, { "epoch": 0.6518218623481782, - "grad_norm": 1.0929382078152947, + "grad_norm": 1.8394458632653032, "learning_rate": 3.989063487022416e-05, - "loss": 1.1448, + "loss": 0.9854, "step": 161 }, { "epoch": 0.6558704453441295, - "grad_norm": 1.1049256212161167, + "grad_norm": 1.3605284984788892, "learning_rate": 3.9884649120583903e-05, - "loss": 1.1304, + "loss": 0.9763, "step": 162 }, { "epoch": 0.659919028340081, - "grad_norm": 0.946443393515843, + "grad_norm": 2.0225632683613917, "learning_rate": 3.987850437381915e-05, - "loss": 1.1483, + "loss": 0.9887, "step": 163 }, { "epoch": 0.6639676113360324, - "grad_norm": 1.0661317162669037, + "grad_norm": 1.6458200712581244, "learning_rate": 3.987220067906313e-05, - "loss": 1.1273, + "loss": 0.9829, "step": 164 }, { "epoch": 0.6680161943319838, - "grad_norm": 0.9483358788950069, + "grad_norm": 1.6559880062120282, "learning_rate": 3.986573808672002e-05, - "loss": 1.1341, + "loss": 0.9847, "step": 165 }, { "epoch": 0.6720647773279352, - "grad_norm": 1.1025100034599538, + "grad_norm": 1.5067482270500034, "learning_rate": 3.9859116648464534e-05, - "loss": 1.1151, + "loss": 0.9748, "step": 166 }, { "epoch": 0.6761133603238867, - "grad_norm": 0.8685191669757499, + "grad_norm": 1.0248242871210778, "learning_rate": 3.985233641724152e-05, - "loss": 1.1461, + "loss": 0.98, "step": 167 }, { "epoch": 0.680161943319838, - "grad_norm": 1.1974721509898754, + "grad_norm": 1.641728738728529, "learning_rate": 3.984539744726554e-05, - "loss": 1.1349, + "loss": 0.9819, "step": 168 }, { "epoch": 0.6842105263157895, - "grad_norm": 1.1272997953127608, + "grad_norm": 0.8202173297718903, "learning_rate": 3.983829979402038e-05, - "loss": 1.1273, + "loss": 0.9787, "step": 169 }, { "epoch": 0.6882591093117408, - "grad_norm": 1.3493810383823597, + "grad_norm": 1.5115724904410357, "learning_rate": 3.983104351425871e-05, - "loss": 1.1652, + "loss": 1.0081, "step": 170 }, { "epoch": 0.6923076923076923, - "grad_norm": 1.1463158300854484, + "grad_norm": 1.000630385009676, "learning_rate": 3.982362866600154e-05, - "loss": 1.1287, + "loss": 0.9733, "step": 171 }, { "epoch": 0.6963562753036437, - "grad_norm": 1.3335387609220315, + "grad_norm": 1.780768953432777, "learning_rate": 3.9816055308537806e-05, - "loss": 1.1429, + "loss": 0.9928, "step": 172 }, { "epoch": 0.7004048582995951, - "grad_norm": 1.2159218279166035, + "grad_norm": 1.6849155529587272, "learning_rate": 3.980832350242388e-05, - "loss": 1.1539, + "loss": 0.9844, "step": 173 }, { "epoch": 0.7044534412955465, - "grad_norm": 1.150258968906525, + "grad_norm": 1.1043620695190473, "learning_rate": 3.980043330948306e-05, - "loss": 1.1512, + "loss": 0.9987, "step": 174 }, { "epoch": 0.708502024291498, - "grad_norm": 0.9127213067376059, + "grad_norm": 2.040842360416658, "learning_rate": 3.9792384792805124e-05, - "loss": 1.1671, + "loss": 1.0021, "step": 175 }, { "epoch": 0.7125506072874493, - "grad_norm": 1.2531936912483919, + "grad_norm": 1.6825940555803278, "learning_rate": 3.978417801674579e-05, - "loss": 1.1451, + "loss": 1.0038, "step": 176 }, { "epoch": 0.7165991902834008, - "grad_norm": 1.4157303952726707, + "grad_norm": 1.5467276709186584, "learning_rate": 3.9775813046926244e-05, - "loss": 1.1828, + "loss": 1.0048, "step": 177 }, { "epoch": 0.7206477732793523, - "grad_norm": 0.7524302199736917, + "grad_norm": 1.4950164268678294, "learning_rate": 3.976728995023254e-05, - "loss": 1.1217, + "loss": 0.9582, "step": 178 }, { "epoch": 0.7246963562753036, - "grad_norm": 1.2749741343722472, + "grad_norm": 1.3810682345831136, "learning_rate": 3.975860879481514e-05, - "loss": 1.1594, + "loss": 1.0126, "step": 179 }, { "epoch": 0.728744939271255, - "grad_norm": 0.8165246262857367, + "grad_norm": 1.4959289747768403, "learning_rate": 3.974976965008833e-05, - "loss": 1.1393, + "loss": 0.9845, "step": 180 }, { "epoch": 0.7327935222672065, - "grad_norm": 0.8747025326331545, + "grad_norm": 1.1791925745188367, "learning_rate": 3.9740772586729674e-05, - "loss": 1.138, + "loss": 1.0047, "step": 181 }, { "epoch": 0.7368421052631579, - "grad_norm": 0.8674161783364317, + "grad_norm": 1.02067650709497, "learning_rate": 3.973161767667946e-05, - "loss": 1.1308, + "loss": 0.9876, "step": 182 }, { "epoch": 0.7408906882591093, - "grad_norm": 1.0757670994245205, + "grad_norm": 1.6901006722168235, "learning_rate": 3.972230499314009e-05, - "loss": 1.1596, + "loss": 0.9818, "step": 183 }, { "epoch": 0.7449392712550608, - "grad_norm": 1.0955523796943372, + "grad_norm": 1.0179333164537712, "learning_rate": 3.9712834610575545e-05, - "loss": 1.1365, + "loss": 0.9687, "step": 184 }, { "epoch": 0.7489878542510121, - "grad_norm": 0.8945651250456106, + "grad_norm": 1.353506727271105, "learning_rate": 3.9703206604710746e-05, - "loss": 1.1507, + "loss": 0.9869, "step": 185 }, { "epoch": 0.7530364372469636, - "grad_norm": 0.9126366875497873, + "grad_norm": 1.2892391817876527, "learning_rate": 3.969342105253098e-05, - "loss": 1.1448, + "loss": 0.9777, "step": 186 }, { "epoch": 0.757085020242915, - "grad_norm": 0.9349527286168777, + "grad_norm": 1.2169102369579685, "learning_rate": 3.9683478032281246e-05, - "loss": 1.1515, + "loss": 0.9919, "step": 187 }, { "epoch": 0.7611336032388664, - "grad_norm": 1.2506347912377482, + "grad_norm": 1.4034080560850042, "learning_rate": 3.967337762346568e-05, - "loss": 1.1932, + "loss": 1.0362, "step": 188 }, { "epoch": 0.7651821862348178, - "grad_norm": 0.9495562726614051, + "grad_norm": 1.5433961850429194, "learning_rate": 3.9663119906846885e-05, - "loss": 1.1624, + "loss": 1.0039, "step": 189 }, { "epoch": 0.7692307692307693, - "grad_norm": 0.7143256566820063, + "grad_norm": 1.041054671932874, "learning_rate": 3.965270496444528e-05, - "loss": 1.1005, + "loss": 0.9492, "step": 190 }, { "epoch": 0.7732793522267206, - "grad_norm": 1.2524894139283158, + "grad_norm": 1.2540755578803082, "learning_rate": 3.9642132879538465e-05, - "loss": 1.1502, + "loss": 0.9953, "step": 191 }, { "epoch": 0.7773279352226721, - "grad_norm": 0.8267983066227882, + "grad_norm": 0.9784774260714353, "learning_rate": 3.9631403736660566e-05, - "loss": 1.1628, + "loss": 0.987, "step": 192 }, { "epoch": 0.7813765182186235, - "grad_norm": 0.8656824823928665, + "grad_norm": 2.0947649072280017, "learning_rate": 3.962051762160151e-05, - "loss": 1.1544, + "loss": 0.9989, "step": 193 }, { "epoch": 0.7854251012145749, - "grad_norm": 0.8070929180272656, + "grad_norm": 0.9345973675263064, "learning_rate": 3.960947462140638e-05, - "loss": 1.1256, + "loss": 0.9732, "step": 194 }, { "epoch": 0.7894736842105263, - "grad_norm": 1.0916648577471206, + "grad_norm": 2.0318581133465594, "learning_rate": 3.959827482437472e-05, - "loss": 1.1427, + "loss": 1.0023, "step": 195 }, { "epoch": 0.7935222672064778, - "grad_norm": 1.1994818226689143, + "grad_norm": 1.410692904936324, "learning_rate": 3.958691832005981e-05, - "loss": 1.1377, + "loss": 0.9795, "step": 196 }, { "epoch": 0.7975708502024291, - "grad_norm": 0.6959185800605986, + "grad_norm": 1.7044044735789916, "learning_rate": 3.9575405199267944e-05, - "loss": 1.1427, + "loss": 0.9821, "step": 197 }, { "epoch": 0.8016194331983806, - "grad_norm": 1.3676264278669668, + "grad_norm": 1.450611503874692, "learning_rate": 3.956373555405774e-05, - "loss": 1.1336, + "loss": 0.9801, "step": 198 }, { "epoch": 0.805668016194332, - "grad_norm": 0.8016160643751276, + "grad_norm": 1.8950214773730447, "learning_rate": 3.955190947773937e-05, - "loss": 1.1496, + "loss": 0.9928, "step": 199 }, { "epoch": 0.8097165991902834, - "grad_norm": 1.1079823552692, + "grad_norm": 1.5105651223860057, "learning_rate": 3.9539927064873824e-05, - "loss": 1.1436, + "loss": 0.9738, "step": 200 }, { "epoch": 0.8137651821862348, - "grad_norm": 1.092660520356088, + "grad_norm": 1.5221180553169806, "learning_rate": 3.952778841127214e-05, - "loss": 1.1303, + "loss": 0.9787, "step": 201 }, { "epoch": 0.8178137651821862, - "grad_norm": 0.968582937075508, + "grad_norm": 1.5134492256034628, "learning_rate": 3.95154936139947e-05, - "loss": 1.1325, + "loss": 0.9946, "step": 202 }, { "epoch": 0.8218623481781376, - "grad_norm": 0.7777358325513107, + "grad_norm": 1.2645601201712975, "learning_rate": 3.950304277135034e-05, - "loss": 1.1401, + "loss": 0.9793, "step": 203 }, { "epoch": 0.8259109311740891, - "grad_norm": 0.7598447294423123, + "grad_norm": 1.7036977343986468, "learning_rate": 3.94904359828957e-05, - "loss": 1.1119, + "loss": 0.9511, "step": 204 }, { "epoch": 0.8299595141700404, - "grad_norm": 0.7875095153633885, + "grad_norm": 1.0217303953741086, "learning_rate": 3.94776733494343e-05, - "loss": 1.1362, + "loss": 0.9736, "step": 205 }, { "epoch": 0.8340080971659919, - "grad_norm": 0.7937883327951576, + "grad_norm": 1.7727305511203737, "learning_rate": 3.9464754973015826e-05, - "loss": 1.1258, + "loss": 0.9773, "step": 206 }, { "epoch": 0.8380566801619433, - "grad_norm": 0.6172123863529246, + "grad_norm": 1.4442085071886064, "learning_rate": 3.9451680956935275e-05, - "loss": 1.133, + "loss": 0.9769, "step": 207 }, { "epoch": 0.8421052631578947, - "grad_norm": 0.7005778046474148, + "grad_norm": 1.6296331059126898, "learning_rate": 3.943845140573213e-05, - "loss": 1.1399, + "loss": 1.0125, "step": 208 }, { "epoch": 0.8461538461538461, - "grad_norm": 0.7696830455058138, + "grad_norm": 1.8384479618389868, "learning_rate": 3.942506642518952e-05, - "loss": 1.1557, + "loss": 0.9993, "step": 209 }, { "epoch": 0.8502024291497976, - "grad_norm": 0.7167256012191972, + "grad_norm": 1.3924212762021668, "learning_rate": 3.9411526122333396e-05, - "loss": 1.1232, + "loss": 0.9653, "step": 210 }, { "epoch": 0.854251012145749, - "grad_norm": 0.6246607199209433, + "grad_norm": 1.4523366113611085, "learning_rate": 3.939783060543167e-05, - "loss": 1.1662, + "loss": 0.9992, "step": 211 }, { "epoch": 0.8582995951417004, - "grad_norm": 0.6154553266613308, + "grad_norm": 1.2780750765465883, "learning_rate": 3.938397998399333e-05, - "loss": 1.1399, + "loss": 0.9829, "step": 212 }, { "epoch": 0.8623481781376519, - "grad_norm": 0.74486199683668, + "grad_norm": 1.1965324739869692, "learning_rate": 3.936997436876756e-05, - "loss": 1.1353, + "loss": 0.982, "step": 213 }, { "epoch": 0.8663967611336032, - "grad_norm": 0.7481405590080246, + "grad_norm": 1.3956819264913638, "learning_rate": 3.935581387174289e-05, - "loss": 1.1456, + "loss": 0.9823, "step": 214 }, { "epoch": 0.8704453441295547, - "grad_norm": 0.7425693953975653, + "grad_norm": 1.0383139824775591, "learning_rate": 3.9341498606146295e-05, - "loss": 1.151, + "loss": 0.9826, "step": 215 }, { "epoch": 0.8744939271255061, - "grad_norm": 0.795498859635868, + "grad_norm": 1.3255790598258013, "learning_rate": 3.932702868644224e-05, - "loss": 1.1512, + "loss": 0.9807, "step": 216 }, { "epoch": 0.8785425101214575, - "grad_norm": 1.05064863580846, + "grad_norm": 1.1983038806560533, "learning_rate": 3.931240422833181e-05, - "loss": 1.1422, + "loss": 0.9782, "step": 217 }, { "epoch": 0.8825910931174089, - "grad_norm": 0.9000685756332608, + "grad_norm": 1.1811201555804478, "learning_rate": 3.9297625348751806e-05, - "loss": 1.1133, + "loss": 0.9669, "step": 218 }, { "epoch": 0.8866396761133604, - "grad_norm": 0.9882692942824614, + "grad_norm": 1.5326643830815496, "learning_rate": 3.9282692165873744e-05, - "loss": 1.1149, + "loss": 0.9633, "step": 219 }, { "epoch": 0.8906882591093117, - "grad_norm": 0.6522293657278715, + "grad_norm": 1.1962623731171746, "learning_rate": 3.9267604799102945e-05, - "loss": 1.1157, + "loss": 0.9645, "step": 220 }, { "epoch": 0.8947368421052632, - "grad_norm": 0.7801197279819774, + "grad_norm": 1.2496037820674863, "learning_rate": 3.9252363369077615e-05, - "loss": 1.1096, + "loss": 0.9566, "step": 221 }, { "epoch": 0.8987854251012146, - "grad_norm": 0.6401647221667957, + "grad_norm": 1.1672907031546669, "learning_rate": 3.9236967997667796e-05, - "loss": 1.1135, + "loss": 0.9754, "step": 222 }, { "epoch": 0.902834008097166, - "grad_norm": 0.8734900035751407, + "grad_norm": 1.017991909191838, "learning_rate": 3.9221418807974496e-05, - "loss": 1.1436, + "loss": 0.9843, "step": 223 }, { "epoch": 0.9068825910931174, - "grad_norm": 0.8066896135527183, + "grad_norm": 1.2044816845331532, "learning_rate": 3.92057159243286e-05, - "loss": 1.1295, + "loss": 0.967, "step": 224 }, { "epoch": 0.9109311740890689, - "grad_norm": 0.6893534751686538, + "grad_norm": 0.9169043936602551, "learning_rate": 3.9189859472289956e-05, - "loss": 1.1429, + "loss": 0.9951, "step": 225 }, { "epoch": 0.9149797570850202, - "grad_norm": 0.8118072981229562, + "grad_norm": 1.2459976667493324, "learning_rate": 3.9173849578646324e-05, - "loss": 1.1514, + "loss": 0.9852, "step": 226 }, { "epoch": 0.9190283400809717, - "grad_norm": 0.7784843315160537, + "grad_norm": 1.405946358141886, "learning_rate": 3.9157686371412395e-05, - "loss": 1.1283, + "loss": 0.9669, "step": 227 }, { "epoch": 0.9230769230769231, - "grad_norm": 0.6568448208273742, + "grad_norm": 1.233490806627217, "learning_rate": 3.914136997982874e-05, - "loss": 1.1187, + "loss": 0.9684, "step": 228 }, { "epoch": 0.9271255060728745, - "grad_norm": 0.9590563647130507, + "grad_norm": 1.0910199469974624, "learning_rate": 3.912490053436079e-05, - "loss": 1.1104, + "loss": 0.9388, "step": 229 }, { "epoch": 0.9311740890688259, - "grad_norm": 0.7749527157754575, + "grad_norm": 1.1893671296816164, "learning_rate": 3.9108278166697805e-05, - "loss": 1.1409, + "loss": 0.9997, "step": 230 }, { "epoch": 0.9352226720647774, - "grad_norm": 0.8446574309330264, + "grad_norm": 1.0283211790708235, "learning_rate": 3.909150300975178e-05, - "loss": 1.1246, + "loss": 0.9561, "step": 231 }, { "epoch": 0.9392712550607287, - "grad_norm": 0.905253457686092, + "grad_norm": 1.8152537502251187, "learning_rate": 3.907457519765642e-05, - "loss": 1.1224, + "loss": 0.9619, "step": 232 }, { "epoch": 0.9433198380566802, - "grad_norm": 0.8821080409850728, + "grad_norm": 1.0352986591870572, "learning_rate": 3.905749486576607e-05, - "loss": 1.1657, + "loss": 0.9981, "step": 233 }, { "epoch": 0.9473684210526315, - "grad_norm": 0.6687637190324838, + "grad_norm": 1.6920950087676914, "learning_rate": 3.90402621506546e-05, - "loss": 1.1204, + "loss": 0.9581, "step": 234 }, { "epoch": 0.951417004048583, - "grad_norm": 0.9137578491309333, + "grad_norm": 1.4388605745919798, "learning_rate": 3.902287719011435e-05, - "loss": 1.1244, + "loss": 0.9678, "step": 235 }, { "epoch": 0.9554655870445344, - "grad_norm": 0.6829496238937356, + "grad_norm": 1.2473271808731832, "learning_rate": 3.900534012315499e-05, - "loss": 1.1342, + "loss": 0.9551, "step": 236 }, { "epoch": 0.9595141700404858, - "grad_norm": 0.7268661516885349, + "grad_norm": 1.3242045158186027, "learning_rate": 3.8987651090002454e-05, - "loss": 1.1398, + "loss": 0.9601, "step": 237 }, { "epoch": 0.9635627530364372, - "grad_norm": 0.8368121290596675, + "grad_norm": 1.2145741795040474, "learning_rate": 3.896981023209777e-05, - "loss": 1.128, + "loss": 0.9663, "step": 238 }, { "epoch": 0.9676113360323887, - "grad_norm": 0.6793854864064489, + "grad_norm": 1.0242281125160628, "learning_rate": 3.8951817692095966e-05, - "loss": 1.1252, + "loss": 0.9647, "step": 239 }, { "epoch": 0.97165991902834, - "grad_norm": 0.6857692719127351, + "grad_norm": 1.1165952712293816, "learning_rate": 3.893367361386491e-05, - "loss": 1.1391, + "loss": 0.9762, "step": 240 }, { "epoch": 0.9757085020242915, - "grad_norm": 0.7515509810988164, + "grad_norm": 1.05833573212546, "learning_rate": 3.891537814248417e-05, - "loss": 1.1513, + "loss": 0.9745, "step": 241 }, { "epoch": 0.979757085020243, - "grad_norm": 0.9776088604787949, + "grad_norm": 1.2908018295422334, "learning_rate": 3.889693142424384e-05, - "loss": 1.141, + "loss": 0.959, "step": 242 }, { "epoch": 0.9838056680161943, - "grad_norm": 1.0573862187985852, + "grad_norm": 0.9150743236498503, "learning_rate": 3.8878333606643405e-05, - "loss": 1.1134, + "loss": 0.9614, "step": 243 }, { "epoch": 0.9878542510121457, - "grad_norm": 0.8674244150090901, + "grad_norm": 1.6096094568088561, "learning_rate": 3.885958483839049e-05, - "loss": 1.1522, + "loss": 0.9692, "step": 244 }, { "epoch": 0.9919028340080972, - "grad_norm": 1.0591569467276056, + "grad_norm": 1.0957042727081412, "learning_rate": 3.8840685269399786e-05, - "loss": 1.1249, + "loss": 0.9873, "step": 245 }, { "epoch": 0.9959514170040485, - "grad_norm": 1.080510181418393, + "grad_norm": 1.401301419117137, "learning_rate": 3.882163505079171e-05, - "loss": 1.189, + "loss": 1.0092, "step": 246 }, { "epoch": 1.0, - "grad_norm": 0.7644299900225966, + "grad_norm": 1.3784863120541866, "learning_rate": 3.880243433489132e-05, - "loss": 1.1347, + "loss": 0.9878, "step": 247 }, { "epoch": 1.0040485829959513, - "grad_norm": 0.8443179895923126, + "grad_norm": 1.2601508284519107, "learning_rate": 3.878308327522702e-05, - "loss": 1.0883, + "loss": 0.9476, "step": 248 }, { "epoch": 1.008097165991903, - "grad_norm": 0.6597041439735067, + "grad_norm": 1.4643843545438244, "learning_rate": 3.876358202652936e-05, - "loss": 1.0842, + "loss": 0.964, "step": 249 }, { "epoch": 1.0121457489878543, - "grad_norm": 0.699180271173301, + "grad_norm": 0.9818092586841592, "learning_rate": 3.8743930744729835e-05, - "loss": 1.0441, + "loss": 0.9152, "step": 250 }, { "epoch": 1.0161943319838056, - "grad_norm": 0.7417502180312221, + "grad_norm": 1.4903744965000287, "learning_rate": 3.8724129586959535e-05, - "loss": 1.0638, + "loss": 0.9495, "step": 251 }, { "epoch": 1.0202429149797572, - "grad_norm": 0.7962851323121076, + "grad_norm": 0.9527091185365648, "learning_rate": 3.870417871154801e-05, - "loss": 1.0597, + "loss": 0.9394, "step": 252 }, { "epoch": 1.0242914979757085, - "grad_norm": 0.8108763566624371, + "grad_norm": 1.8116636778925408, "learning_rate": 3.868407827802192e-05, - "loss": 1.054, + "loss": 0.9211, "step": 253 }, { "epoch": 1.0283400809716599, - "grad_norm": 0.7882580588950429, + "grad_norm": 1.3343758305460551, "learning_rate": 3.8663828447103804e-05, - "loss": 1.0362, + "loss": 0.9222, "step": 254 }, { "epoch": 1.0323886639676114, - "grad_norm": 0.7414904444678652, + "grad_norm": 1.7023105786578305, "learning_rate": 3.8643429380710765e-05, - "loss": 1.0462, + "loss": 0.9191, "step": 255 }, { "epoch": 1.0364372469635628, - "grad_norm": 0.897759352718327, + "grad_norm": 1.7105675281046444, "learning_rate": 3.862288124195319e-05, - "loss": 1.0735, + "loss": 0.9487, "step": 256 }, { "epoch": 1.040485829959514, - "grad_norm": 0.6485867521144105, + "grad_norm": 1.0565073430701226, "learning_rate": 3.8602184195133464e-05, - "loss": 1.0643, + "loss": 0.9602, "step": 257 }, { "epoch": 1.0445344129554657, - "grad_norm": 0.6594623644575873, + "grad_norm": 1.2081619128021956, "learning_rate": 3.8581338405744604e-05, - "loss": 1.0669, + "loss": 0.9525, "step": 258 }, { "epoch": 1.048582995951417, - "grad_norm": 0.6469684402695344, + "grad_norm": 1.2298707883770363, "learning_rate": 3.8560344040468996e-05, - "loss": 1.0911, + "loss": 0.9543, "step": 259 }, { "epoch": 1.0526315789473684, - "grad_norm": 0.660713961312183, + "grad_norm": 0.8991075354475959, "learning_rate": 3.853920126717703e-05, - "loss": 1.0613, + "loss": 0.9486, "step": 260 }, { "epoch": 1.05668016194332, - "grad_norm": 0.6655432009694497, + "grad_norm": 1.3219290384174023, "learning_rate": 3.851791025492574e-05, - "loss": 1.0689, + "loss": 0.9382, "step": 261 }, { "epoch": 1.0607287449392713, - "grad_norm": 0.744393087697742, + "grad_norm": 1.0306484047639277, "learning_rate": 3.8496471173957515e-05, - "loss": 1.0629, + "loss": 0.9327, "step": 262 }, { "epoch": 1.0647773279352226, - "grad_norm": 0.7282625430194793, + "grad_norm": 1.2863879835589798, "learning_rate": 3.847488419569866e-05, - "loss": 1.0511, + "loss": 0.9324, "step": 263 }, { "epoch": 1.0688259109311742, - "grad_norm": 0.67119991257505, + "grad_norm": 1.0948486048442023, "learning_rate": 3.8453149492758077e-05, - "loss": 1.0476, + "loss": 0.9488, "step": 264 }, { "epoch": 1.0728744939271255, - "grad_norm": 0.739320899021367, + "grad_norm": 1.180583426670717, "learning_rate": 3.8431267238925886e-05, - "loss": 1.0532, + "loss": 0.9332, "step": 265 }, { "epoch": 1.0769230769230769, - "grad_norm": 0.62555452334972, + "grad_norm": 1.0152940509068298, "learning_rate": 3.840923760917199e-05, - "loss": 1.0569, + "loss": 0.9547, "step": 266 }, { "epoch": 1.0809716599190284, - "grad_norm": 0.7054804997807504, + "grad_norm": 1.0531010017040114, "learning_rate": 3.838706077964473e-05, - "loss": 1.0701, + "loss": 0.9486, "step": 267 }, { "epoch": 1.0850202429149798, - "grad_norm": 0.6343998149873769, + "grad_norm": 1.189151717194549, "learning_rate": 3.836473692766944e-05, - "loss": 1.0861, + "loss": 0.9732, "step": 268 }, { "epoch": 1.0890688259109311, - "grad_norm": 0.6635573609010429, + "grad_norm": 1.2152858661974875, "learning_rate": 3.834226623174705e-05, - "loss": 1.0724, + "loss": 0.9563, "step": 269 }, { "epoch": 1.0931174089068827, - "grad_norm": 0.7251144803401505, + "grad_norm": 0.8463805812231769, "learning_rate": 3.831964887155264e-05, - "loss": 1.0316, + "loss": 0.9293, "step": 270 }, { "epoch": 1.097165991902834, - "grad_norm": 0.6708661840501796, + "grad_norm": 1.1392917843312798, "learning_rate": 3.829688502793403e-05, - "loss": 1.1123, + "loss": 0.983, "step": 271 }, { "epoch": 1.1012145748987854, - "grad_norm": 0.791935497493942, + "grad_norm": 1.3052827892538923, "learning_rate": 3.827397488291029e-05, - "loss": 1.0596, + "loss": 0.9367, "step": 272 }, { "epoch": 1.1052631578947367, - "grad_norm": 0.6908715788395017, + "grad_norm": 1.002779557202567, "learning_rate": 3.8250918619670343e-05, - "loss": 1.0575, + "loss": 0.9358, "step": 273 }, { "epoch": 1.1093117408906883, - "grad_norm": 0.6950652982222986, + "grad_norm": 0.9332003075952789, "learning_rate": 3.822771642257145e-05, - "loss": 1.0584, + "loss": 0.9298, "step": 274 }, { "epoch": 1.1133603238866396, - "grad_norm": 0.6419502460001266, + "grad_norm": 1.3123983925657976, "learning_rate": 3.820436847713776e-05, - "loss": 1.0889, + "loss": 0.9589, "step": 275 }, { "epoch": 1.117408906882591, - "grad_norm": 0.7195383974693976, + "grad_norm": 1.0936151678736576, "learning_rate": 3.8180874970058826e-05, - "loss": 1.0659, + "loss": 0.9507, "step": 276 }, { "epoch": 1.1214574898785425, - "grad_norm": 0.6917159522853528, + "grad_norm": 0.9950986107412927, "learning_rate": 3.8157236089188103e-05, - "loss": 1.0821, + "loss": 0.9545, "step": 277 }, { "epoch": 1.125506072874494, - "grad_norm": 0.6281905057843152, + "grad_norm": 0.8784068825826009, "learning_rate": 3.813345202354145e-05, - "loss": 1.0722, + "loss": 0.937, "step": 278 }, { "epoch": 1.1295546558704452, - "grad_norm": 0.6213722520004633, + "grad_norm": 1.059992480611983, "learning_rate": 3.810952296329563e-05, - "loss": 1.08, + "loss": 0.9488, "step": 279 }, { "epoch": 1.1336032388663968, - "grad_norm": 0.6652084629645737, + "grad_norm": 1.411763705523762, "learning_rate": 3.808544909978676e-05, - "loss": 1.0482, + "loss": 0.9257, "step": 280 }, { "epoch": 1.1376518218623481, - "grad_norm": 0.6907135880660236, + "grad_norm": 1.0750907918081896, "learning_rate": 3.806123062550882e-05, - "loss": 1.0965, + "loss": 0.9817, "step": 281 }, { "epoch": 1.1417004048582995, - "grad_norm": 0.632257673258633, + "grad_norm": 0.9406912641442262, "learning_rate": 3.803686773411208e-05, - "loss": 1.0425, + "loss": 0.9194, "step": 282 }, { "epoch": 1.145748987854251, - "grad_norm": 0.6345007129443234, + "grad_norm": 1.2392855664726456, "learning_rate": 3.801236062040158e-05, - "loss": 1.1106, + "loss": 0.9908, "step": 283 }, { "epoch": 1.1497975708502024, - "grad_norm": 0.6803102157205853, + "grad_norm": 0.7360814897964765, "learning_rate": 3.798770948033553e-05, - "loss": 1.0497, + "loss": 0.9363, "step": 284 }, { "epoch": 1.1538461538461537, - "grad_norm": 0.7228342175795381, + "grad_norm": 0.9357343722746014, "learning_rate": 3.796291451102381e-05, - "loss": 1.0879, + "loss": 0.9604, "step": 285 }, { "epoch": 1.1578947368421053, - "grad_norm": 0.8475822634328514, + "grad_norm": 1.388440685377715, "learning_rate": 3.793797591072631e-05, - "loss": 1.0825, + "loss": 0.9769, "step": 286 }, { "epoch": 1.1619433198380567, - "grad_norm": 0.6779182216974409, + "grad_norm": 0.782683044064809, "learning_rate": 3.791289387885144e-05, - "loss": 1.0922, + "loss": 0.9639, "step": 287 }, { "epoch": 1.165991902834008, - "grad_norm": 0.7802482315865281, + "grad_norm": 1.236178577147841, "learning_rate": 3.788766861595443e-05, - "loss": 1.0654, + "loss": 0.9471, "step": 288 }, { "epoch": 1.1700404858299596, - "grad_norm": 0.8066866394352527, + "grad_norm": 1.1861811723216849, "learning_rate": 3.7862300323735835e-05, - "loss": 1.0629, + "loss": 0.9288, "step": 289 }, { "epoch": 1.174089068825911, - "grad_norm": 0.6219653285226472, + "grad_norm": 0.8087251179700615, "learning_rate": 3.783678920503983e-05, - "loss": 1.0508, + "loss": 0.9344, "step": 290 }, { "epoch": 1.1781376518218623, - "grad_norm": 0.6427672745581562, + "grad_norm": 0.8546077035923103, "learning_rate": 3.781113546385265e-05, - "loss": 1.0779, + "loss": 0.96, "step": 291 }, { "epoch": 1.1821862348178138, - "grad_norm": 0.5757616601142316, + "grad_norm": 1.7085128036006627, "learning_rate": 3.7785339305300916e-05, - "loss": 1.105, + "loss": 0.9581, "step": 292 }, { "epoch": 1.1862348178137652, - "grad_norm": 0.6183724001842665, + "grad_norm": 0.8991138875496323, "learning_rate": 3.7759400935650025e-05, - "loss": 1.0656, + "loss": 0.9434, "step": 293 }, { "epoch": 1.1902834008097165, - "grad_norm": 0.6950553990282842, + "grad_norm": 1.9154461242553054, "learning_rate": 3.77333205623025e-05, - "loss": 1.0776, + "loss": 0.9504, "step": 294 }, { "epoch": 1.194331983805668, - "grad_norm": 0.5003460495047228, + "grad_norm": 1.3135211376093052, "learning_rate": 3.7707098393796295e-05, - "loss": 1.0293, + "loss": 0.9185, "step": 295 }, { "epoch": 1.1983805668016194, - "grad_norm": 0.6786339107276377, + "grad_norm": 1.9824877751687107, "learning_rate": 3.7680734639803185e-05, - "loss": 1.0843, + "loss": 0.9669, "step": 296 }, { "epoch": 1.2024291497975708, - "grad_norm": 0.6553558355973711, + "grad_norm": 1.728506157287037, "learning_rate": 3.765422951112704e-05, - "loss": 1.0489, + "loss": 0.931, "step": 297 }, { "epoch": 1.2064777327935223, - "grad_norm": 0.676158767314287, + "grad_norm": 1.6011976486884048, "learning_rate": 3.762758321970216e-05, - "loss": 1.0671, + "loss": 0.9556, "step": 298 }, { "epoch": 1.2105263157894737, - "grad_norm": 0.5832436337903422, + "grad_norm": 1.587064483755446, "learning_rate": 3.7600795978591584e-05, - "loss": 1.0744, + "loss": 0.9606, "step": 299 }, { "epoch": 1.214574898785425, - "grad_norm": 0.646506411939147, + "grad_norm": 1.5577072418437223, "learning_rate": 3.757386800198538e-05, - "loss": 1.0765, + "loss": 0.9593, "step": 300 }, { "epoch": 1.2186234817813766, - "grad_norm": 0.6176278645942233, + "grad_norm": 1.4656822332211439, "learning_rate": 3.7546799505198925e-05, - "loss": 1.1081, + "loss": 0.9839, "step": 301 }, { "epoch": 1.222672064777328, - "grad_norm": 0.6684849554015623, + "grad_norm": 1.3382926754527191, "learning_rate": 3.7519590704671197e-05, - "loss": 1.0809, + "loss": 0.9424, "step": 302 }, { "epoch": 1.2267206477732793, - "grad_norm": 0.6930884876871414, + "grad_norm": 1.0133765195713287, "learning_rate": 3.749224181796305e-05, - "loss": 1.0969, + "loss": 0.9625, "step": 303 }, { "epoch": 1.2307692307692308, - "grad_norm": 0.5475506735787983, + "grad_norm": 1.364272427344207, "learning_rate": 3.746475306375545e-05, - "loss": 1.0844, + "loss": 0.9459, "step": 304 }, { "epoch": 1.2348178137651822, - "grad_norm": 0.6199271349054885, + "grad_norm": 0.8038099872870424, "learning_rate": 3.743712466184774e-05, - "loss": 1.0826, + "loss": 0.9619, "step": 305 }, { "epoch": 1.2388663967611335, - "grad_norm": 0.7036475713773428, + "grad_norm": 1.005465463619015, "learning_rate": 3.7409356833155885e-05, - "loss": 1.0781, + "loss": 0.9651, "step": 306 }, { "epoch": 1.242914979757085, - "grad_norm": 0.5727663627728128, + "grad_norm": 0.8811382093585958, "learning_rate": 3.73814497997107e-05, - "loss": 1.0467, + "loss": 0.9224, "step": 307 }, { "epoch": 1.2469635627530364, - "grad_norm": 0.6567730382642047, + "grad_norm": 0.9325421665023609, "learning_rate": 3.7353403784656095e-05, - "loss": 1.1097, + "loss": 0.9634, "step": 308 }, { "epoch": 1.2510121457489878, - "grad_norm": 0.7597407870848478, + "grad_norm": 1.0112648460330802, "learning_rate": 3.732521901224724e-05, - "loss": 1.0766, + "loss": 0.9525, "step": 309 }, { "epoch": 1.2550607287449393, - "grad_norm": 0.6812198151302539, + "grad_norm": 1.1100784856709203, "learning_rate": 3.7296895707848844e-05, - "loss": 1.0979, + "loss": 0.9699, "step": 310 }, { "epoch": 1.2591093117408907, - "grad_norm": 0.586093912709632, + "grad_norm": 1.23372105529037, "learning_rate": 3.7268434097933275e-05, - "loss": 1.0485, + "loss": 0.9319, "step": 311 }, { "epoch": 1.263157894736842, - "grad_norm": 0.5184003544911839, + "grad_norm": 1.078112699610884, "learning_rate": 3.723983441007882e-05, - "loss": 1.0743, + "loss": 0.9467, "step": 312 }, { "epoch": 1.2672064777327936, - "grad_norm": 0.5804156921302127, + "grad_norm": 1.162714030610333, "learning_rate": 3.721109687296781e-05, - "loss": 1.0664, + "loss": 0.9374, "step": 313 }, { "epoch": 1.271255060728745, - "grad_norm": 0.5356762907815773, + "grad_norm": 0.8387918124507938, "learning_rate": 3.718222171638484e-05, - "loss": 1.0408, + "loss": 0.9249, "step": 314 }, { "epoch": 1.2753036437246963, - "grad_norm": 0.5498694527734794, + "grad_norm": 1.329110800004607, "learning_rate": 3.715320917121488e-05, - "loss": 1.0826, + "loss": 0.9433, "step": 315 }, { "epoch": 1.2793522267206479, - "grad_norm": 0.6169734726700452, + "grad_norm": 1.2761488696707923, "learning_rate": 3.712405946944149e-05, - "loss": 1.0528, + "loss": 0.9235, "step": 316 }, { "epoch": 1.2834008097165992, - "grad_norm": 0.6105688377218171, + "grad_norm": 0.7580407486584213, "learning_rate": 3.709477284414488e-05, - "loss": 1.0701, + "loss": 0.9423, "step": 317 }, { "epoch": 1.2874493927125505, - "grad_norm": 0.6259072124827054, + "grad_norm": 1.1877649442288238, "learning_rate": 3.7065349529500136e-05, - "loss": 1.065, + "loss": 0.9377, "step": 318 }, { "epoch": 1.291497975708502, - "grad_norm": 0.6064540137580364, + "grad_norm": 1.1419999860485783, "learning_rate": 3.7035789760775305e-05, - "loss": 1.0535, + "loss": 0.9354, "step": 319 }, { "epoch": 1.2955465587044535, - "grad_norm": 0.707864064563686, + "grad_norm": 1.17107363600178, "learning_rate": 3.70060937743295e-05, - "loss": 1.0678, + "loss": 0.9432, "step": 320 }, { "epoch": 1.2995951417004048, - "grad_norm": 0.6265064438665333, + "grad_norm": 0.7741549547856117, "learning_rate": 3.6976261807611045e-05, - "loss": 1.0713, + "loss": 0.9411, "step": 321 }, { "epoch": 1.3036437246963564, - "grad_norm": 0.7374978040306303, + "grad_norm": 0.9172176721412969, "learning_rate": 3.694629409915555e-05, - "loss": 1.0269, + "loss": 0.9149, "step": 322 }, { "epoch": 1.3076923076923077, - "grad_norm": 0.6124093193411688, + "grad_norm": 0.7701041194570237, "learning_rate": 3.691619088858401e-05, - "loss": 1.067, + "loss": 0.9314, "step": 323 }, { "epoch": 1.311740890688259, - "grad_norm": 0.8203433461167243, + "grad_norm": 1.1581063537877019, "learning_rate": 3.68859524166009e-05, - "loss": 1.0669, + "loss": 0.9296, "step": 324 }, { "epoch": 1.3157894736842106, - "grad_norm": 0.7845462014942922, + "grad_norm": 1.3867731582735707, "learning_rate": 3.6855578924992226e-05, - "loss": 1.087, + "loss": 0.9596, "step": 325 }, { "epoch": 1.319838056680162, - "grad_norm": 0.5009517460604084, + "grad_norm": 0.5908107848823839, "learning_rate": 3.6825070656623626e-05, - "loss": 1.0544, + "loss": 0.93, "step": 326 }, { "epoch": 1.3238866396761133, - "grad_norm": 0.7326218182947578, + "grad_norm": 0.9327656425339789, "learning_rate": 3.67944278554384e-05, - "loss": 1.0784, + "loss": 0.9528, "step": 327 }, { "epoch": 1.3279352226720649, - "grad_norm": 0.7679717612781879, + "grad_norm": 1.3145462013126092, "learning_rate": 3.676365076645557e-05, - "loss": 1.0804, + "loss": 0.9432, "step": 328 }, { "epoch": 1.3319838056680162, - "grad_norm": 0.7071008510514112, + "grad_norm": 0.9223040348940436, "learning_rate": 3.673273963576791e-05, - "loss": 1.0568, + "loss": 0.9225, "step": 329 }, { "epoch": 1.3360323886639676, - "grad_norm": 0.7365769774142052, + "grad_norm": 0.8322955244626308, "learning_rate": 3.670169471054001e-05, - "loss": 1.0822, + "loss": 0.9507, "step": 330 }, { "epoch": 1.3400809716599191, - "grad_norm": 0.8767254035362417, + "grad_norm": 0.7105310829699437, "learning_rate": 3.6670516239006254e-05, - "loss": 1.098, + "loss": 0.9631, "step": 331 }, { "epoch": 1.3441295546558705, - "grad_norm": 0.7044300047224916, + "grad_norm": 0.7388148326787155, "learning_rate": 3.663920447046888e-05, - "loss": 1.0907, + "loss": 0.9684, "step": 332 }, { "epoch": 1.3481781376518218, - "grad_norm": 0.7382949507859735, + "grad_norm": 0.7631410956280685, "learning_rate": 3.6607759655295954e-05, - "loss": 1.0719, + "loss": 0.9448, "step": 333 }, { "epoch": 1.3522267206477734, - "grad_norm": 0.7912928865825757, + "grad_norm": 0.8263834331319808, "learning_rate": 3.657618204491937e-05, - "loss": 1.083, + "loss": 0.962, "step": 334 }, { "epoch": 1.3562753036437247, - "grad_norm": 0.7640240832244506, + "grad_norm": 0.6239847649348641, "learning_rate": 3.6544471891832875e-05, - "loss": 1.0475, + "loss": 0.9263, "step": 335 }, { "epoch": 1.360323886639676, - "grad_norm": 0.5958716807646797, + "grad_norm": 0.8358274371983642, "learning_rate": 3.6512629449589983e-05, - "loss": 1.0645, + "loss": 0.93, "step": 336 }, { "epoch": 1.3643724696356276, - "grad_norm": 0.5821407125043725, + "grad_norm": 0.6746377646712112, "learning_rate": 3.6480654972802045e-05, - "loss": 1.0781, + "loss": 0.9343, "step": 337 }, { "epoch": 1.368421052631579, - "grad_norm": 0.6356505776607826, + "grad_norm": 0.901823699815308, "learning_rate": 3.644854871713611e-05, - "loss": 1.072, + "loss": 0.9348, "step": 338 }, { "epoch": 1.3724696356275303, - "grad_norm": 0.6361398057149624, + "grad_norm": 1.0283922943494666, "learning_rate": 3.6416310939312924e-05, - "loss": 1.1057, + "loss": 0.9668, "step": 339 }, { "epoch": 1.376518218623482, - "grad_norm": 0.5543593014020454, + "grad_norm": 1.0920385982787542, "learning_rate": 3.638394189710493e-05, - "loss": 1.0694, + "loss": 0.9343, "step": 340 }, { "epoch": 1.3805668016194332, - "grad_norm": 0.5596047622054691, + "grad_norm": 0.9517003244013462, "learning_rate": 3.635144184933412e-05, - "loss": 1.0608, + "loss": 0.9187, "step": 341 }, { "epoch": 1.3846153846153846, - "grad_norm": 0.5489838269583212, + "grad_norm": 0.9657491699179347, "learning_rate": 3.6318811055869994e-05, - "loss": 1.0837, + "loss": 0.944, "step": 342 }, { "epoch": 1.3886639676113361, - "grad_norm": 0.5609485242624515, + "grad_norm": 0.9217728593515332, "learning_rate": 3.628604977762752e-05, - "loss": 1.0641, + "loss": 0.9362, "step": 343 }, { "epoch": 1.3927125506072875, - "grad_norm": 0.6073751191310837, + "grad_norm": 1.1842660808690335, "learning_rate": 3.6253158276565004e-05, - "loss": 1.0443, + "loss": 0.9155, "step": 344 }, { "epoch": 1.3967611336032388, - "grad_norm": 0.5562291111079254, + "grad_norm": 1.1155156510312787, "learning_rate": 3.6220136815682016e-05, - "loss": 1.0783, + "loss": 0.9427, "step": 345 }, { "epoch": 1.4008097165991904, - "grad_norm": 0.6713971617632366, + "grad_norm": 1.1039206312793404, "learning_rate": 3.618698565901727e-05, - "loss": 1.0709, + "loss": 0.9427, "step": 346 }, { "epoch": 1.4048582995951417, - "grad_norm": 0.5660688480906494, + "grad_norm": 0.778074502255895, "learning_rate": 3.615370507164654e-05, - "loss": 1.0802, + "loss": 0.947, "step": 347 }, { "epoch": 1.408906882591093, - "grad_norm": 0.6352774504564733, + "grad_norm": 1.33404419691114, "learning_rate": 3.61202953196805e-05, - "loss": 1.0749, + "loss": 0.9428, "step": 348 }, { "epoch": 1.4129554655870447, - "grad_norm": 0.7296047306531193, + "grad_norm": 1.089385797809555, "learning_rate": 3.6086756670262654e-05, - "loss": 1.0819, + "loss": 0.9424, "step": 349 }, { "epoch": 1.417004048582996, - "grad_norm": 0.5397917361138082, + "grad_norm": 1.1901314933763787, "learning_rate": 3.6053089391567135e-05, - "loss": 1.0743, + "loss": 0.9445, "step": 350 }, { "epoch": 1.4210526315789473, - "grad_norm": 0.6528800158602562, + "grad_norm": 0.5858874705321329, "learning_rate": 3.601929375279662e-05, - "loss": 1.0706, + "loss": 0.9407, "step": 351 }, { "epoch": 1.425101214574899, - "grad_norm": 0.689535174618146, + "grad_norm": 0.8963373086310701, "learning_rate": 3.598537002418012e-05, - "loss": 1.0366, + "loss": 0.9208, "step": 352 }, { "epoch": 1.4291497975708503, - "grad_norm": 0.5544855819375744, + "grad_norm": 1.0386241831904415, "learning_rate": 3.5951318476970885e-05, - "loss": 1.0739, + "loss": 0.9454, "step": 353 }, { "epoch": 1.4331983805668016, - "grad_norm": 0.7389547601101709, + "grad_norm": 1.1455698409790502, "learning_rate": 3.591713938344416e-05, - "loss": 1.0787, + "loss": 0.9602, "step": 354 }, { "epoch": 1.4372469635627532, - "grad_norm": 0.6065025932583743, + "grad_norm": 0.7627382721354394, "learning_rate": 3.588283301689507e-05, - "loss": 1.0979, + "loss": 0.9572, "step": 355 }, { "epoch": 1.4412955465587045, - "grad_norm": 0.558501688290143, + "grad_norm": 1.0095531912110582, "learning_rate": 3.5848399651636424e-05, - "loss": 1.0723, + "loss": 0.9488, "step": 356 }, { "epoch": 1.4453441295546559, - "grad_norm": 0.5350356301369611, + "grad_norm": 0.7983284114814717, "learning_rate": 3.5813839562996473e-05, - "loss": 1.0776, + "loss": 0.9514, "step": 357 }, { "epoch": 1.4493927125506074, - "grad_norm": 0.6159401445890172, + "grad_norm": 0.6804890698929813, "learning_rate": 3.5779153027316775e-05, - "loss": 1.0905, + "loss": 0.9377, "step": 358 }, { "epoch": 1.4534412955465588, - "grad_norm": 0.5462635510126347, + "grad_norm": 0.8959552452558969, "learning_rate": 3.574434032194994e-05, - "loss": 1.0425, + "loss": 0.9218, "step": 359 }, { "epoch": 1.45748987854251, - "grad_norm": 0.6703558098255935, + "grad_norm": 0.6719461277890947, "learning_rate": 3.5709401725257436e-05, - "loss": 1.0863, + "loss": 0.9526, "step": 360 }, { "epoch": 1.4615384615384617, - "grad_norm": 0.6103529920690267, + "grad_norm": 0.9099064382764196, "learning_rate": 3.5674337516607346e-05, - "loss": 1.0514, + "loss": 0.9219, "step": 361 }, { "epoch": 1.465587044534413, - "grad_norm": 0.6152716126418628, + "grad_norm": 1.1538627684048577, "learning_rate": 3.563914797637216e-05, - "loss": 1.0672, + "loss": 0.9497, "step": 362 }, { "epoch": 1.4696356275303644, - "grad_norm": 0.5798544905775808, + "grad_norm": 0.8870832611585405, "learning_rate": 3.560383338592649e-05, - "loss": 1.0941, + "loss": 0.9493, "step": 363 }, { "epoch": 1.4736842105263157, - "grad_norm": 0.6032576950713833, + "grad_norm": 0.6336151684019032, "learning_rate": 3.556839402764487e-05, - "loss": 1.0749, + "loss": 0.9295, "step": 364 }, { "epoch": 1.4777327935222673, - "grad_norm": 0.5743747001714082, + "grad_norm": 0.6823600747563382, "learning_rate": 3.553283018489946e-05, - "loss": 1.0731, + "loss": 0.9358, "step": 365 }, { "epoch": 1.4817813765182186, - "grad_norm": 0.6257409088175863, + "grad_norm": 0.7695913096030961, "learning_rate": 3.54971421420578e-05, - "loss": 1.0563, + "loss": 0.9448, "step": 366 }, { "epoch": 1.48582995951417, - "grad_norm": 0.5052479368992682, + "grad_norm": 0.6644954150624587, "learning_rate": 3.5461330184480524e-05, - "loss": 1.0792, + "loss": 0.9674, "step": 367 }, { "epoch": 1.4898785425101215, - "grad_norm": 0.6134525619598572, + "grad_norm": 0.840904941548884, "learning_rate": 3.54253945985191e-05, - "loss": 1.0506, + "loss": 0.9258, "step": 368 }, { "epoch": 1.4939271255060729, - "grad_norm": 0.5877067772770359, + "grad_norm": 1.2413545418504297, "learning_rate": 3.5389335671513524e-05, - "loss": 1.0743, + "loss": 0.946, "step": 369 }, { "epoch": 1.4979757085020242, - "grad_norm": 0.6824295247470854, + "grad_norm": 0.6986911580844979, "learning_rate": 3.5353153691789986e-05, - "loss": 1.0385, + "loss": 0.9253, "step": 370 }, { "epoch": 1.5020242914979756, - "grad_norm": 0.6124262692625976, + "grad_norm": 0.7357823303633868, "learning_rate": 3.531684894865866e-05, - "loss": 1.0691, + "loss": 0.9277, "step": 371 }, { "epoch": 1.5060728744939271, - "grad_norm": 0.6628464565489501, + "grad_norm": 0.6764560007033392, "learning_rate": 3.528042173241128e-05, - "loss": 1.074, + "loss": 0.9336, "step": 372 }, { "epoch": 1.5101214574898787, - "grad_norm": 0.6012489806738498, + "grad_norm": 0.7781513939335954, "learning_rate": 3.524387233431892e-05, - "loss": 1.071, + "loss": 0.9466, "step": 373 }, { "epoch": 1.5141700404858298, - "grad_norm": 0.5198582855795919, + "grad_norm": 0.9510003107626882, "learning_rate": 3.520720104662958e-05, - "loss": 1.0839, + "loss": 0.9479, "step": 374 }, { "epoch": 1.5182186234817814, - "grad_norm": 0.5732275242822916, + "grad_norm": 1.1010294445381488, "learning_rate": 3.5170408162565904e-05, - "loss": 1.0559, + "loss": 0.941, "step": 375 }, { "epoch": 1.522267206477733, - "grad_norm": 0.6000426692580564, + "grad_norm": 1.3463235278818668, "learning_rate": 3.513349397632281e-05, - "loss": 1.0599, + "loss": 0.9303, "step": 376 }, { "epoch": 1.526315789473684, - "grad_norm": 0.6503237376405522, + "grad_norm": 0.5831423187095385, "learning_rate": 3.5096458783065145e-05, - "loss": 1.0796, + "loss": 0.9472, "step": 377 }, { "epoch": 1.5303643724696356, - "grad_norm": 0.5351010916366247, + "grad_norm": 0.8413936541195572, "learning_rate": 3.505930287892533e-05, - "loss": 1.0713, + "loss": 0.944, "step": 378 }, { "epoch": 1.5344129554655872, - "grad_norm": 0.688644733150878, + "grad_norm": 1.3670789753447343, "learning_rate": 3.502202656100099e-05, - "loss": 1.0523, + "loss": 0.9263, "step": 379 }, { "epoch": 1.5384615384615383, - "grad_norm": 0.6320024196389538, + "grad_norm": 0.5713795899583805, "learning_rate": 3.498463012735256e-05, - "loss": 1.1055, + "loss": 0.9647, "step": 380 }, { "epoch": 1.54251012145749, - "grad_norm": 0.5238335027077707, + "grad_norm": 0.7620490131958497, "learning_rate": 3.494711387700094e-05, - "loss": 1.0971, + "loss": 0.9647, "step": 381 }, { "epoch": 1.5465587044534415, - "grad_norm": 0.6203198785088283, + "grad_norm": 1.3238509369915523, "learning_rate": 3.490947810992505e-05, - "loss": 1.0606, + "loss": 0.9273, "step": 382 }, { "epoch": 1.5506072874493926, - "grad_norm": 0.636405612493776, + "grad_norm": 0.8320467196720917, "learning_rate": 3.48717231270595e-05, - "loss": 1.0522, + "loss": 0.9234, "step": 383 }, { "epoch": 1.5546558704453441, - "grad_norm": 0.6249969504079518, + "grad_norm": 0.7952411148068089, "learning_rate": 3.483384923029211e-05, - "loss": 1.086, + "loss": 0.9575, "step": 384 }, { "epoch": 1.5587044534412957, - "grad_norm": 0.5212954380475604, + "grad_norm": 0.9322962225029922, "learning_rate": 3.479585672246156e-05, - "loss": 1.0771, + "loss": 0.9379, "step": 385 }, { "epoch": 1.5627530364372468, - "grad_norm": 0.7499672540039222, + "grad_norm": 1.045952056285868, "learning_rate": 3.4757745907354924e-05, - "loss": 1.0671, + "loss": 0.9392, "step": 386 }, { "epoch": 1.5668016194331984, - "grad_norm": 0.6298364477711781, + "grad_norm": 0.9386341347427138, "learning_rate": 3.471951708970527e-05, - "loss": 1.0625, + "loss": 0.926, "step": 387 }, { "epoch": 1.5708502024291497, - "grad_norm": 0.6673288004291112, + "grad_norm": 1.1204763737329761, "learning_rate": 3.468117057518921e-05, - "loss": 1.079, + "loss": 0.9516, "step": 388 }, { "epoch": 1.574898785425101, - "grad_norm": 0.7507497463659425, + "grad_norm": 0.6958082460098163, "learning_rate": 3.4642706670424426e-05, - "loss": 1.0869, + "loss": 0.9551, "step": 389 }, { "epoch": 1.5789473684210527, - "grad_norm": 0.5356827886057539, + "grad_norm": 0.6974704146764579, "learning_rate": 3.460412568296731e-05, - "loss": 1.0531, + "loss": 0.9192, "step": 390 }, { "epoch": 1.582995951417004, - "grad_norm": 0.6480917384598469, + "grad_norm": 0.886441897060612, "learning_rate": 3.456542792131039e-05, - "loss": 1.0678, + "loss": 0.9497, "step": 391 }, { "epoch": 1.5870445344129553, - "grad_norm": 0.6375156247885815, + "grad_norm": 0.8833409001600455, "learning_rate": 3.4526613694879936e-05, - "loss": 1.076, + "loss": 0.9324, "step": 392 }, { "epoch": 1.591093117408907, - "grad_norm": 0.5646777147530785, + "grad_norm": 1.0175107410806334, "learning_rate": 3.448768331403348e-05, - "loss": 1.063, + "loss": 0.9333, "step": 393 }, { "epoch": 1.5951417004048583, - "grad_norm": 0.6619823741442639, + "grad_norm": 0.9213486133955586, "learning_rate": 3.44486370900573e-05, - "loss": 1.085, + "loss": 0.9585, "step": 394 }, { "epoch": 1.5991902834008096, - "grad_norm": 0.5683499890052642, + "grad_norm": 0.9240866650025109, "learning_rate": 3.440947533516396e-05, - "loss": 1.0599, + "loss": 0.9257, "step": 395 }, { "epoch": 1.6032388663967612, - "grad_norm": 0.5269539159128876, + "grad_norm": 0.5260005271426192, "learning_rate": 3.4370198362489816e-05, - "loss": 1.0797, + "loss": 0.9351, "step": 396 }, { "epoch": 1.6072874493927125, - "grad_norm": 0.504156960575437, + "grad_norm": 0.668874308720013, "learning_rate": 3.43308064860925e-05, - "loss": 1.084, + "loss": 0.9699, "step": 397 }, { "epoch": 1.6113360323886639, - "grad_norm": 0.6173467255186729, + "grad_norm": 0.6073375196038894, "learning_rate": 3.4291300020948393e-05, - "loss": 1.0338, + "loss": 0.9084, "step": 398 }, { "epoch": 1.6153846153846154, - "grad_norm": 0.556836749544865, + "grad_norm": 0.7829536322457934, "learning_rate": 3.425167928295015e-05, - "loss": 1.075, + "loss": 0.9521, "step": 399 }, { "epoch": 1.6194331983805668, - "grad_norm": 0.666228935521094, + "grad_norm": 0.7585119579139188, "learning_rate": 3.421194458890411e-05, - "loss": 1.0912, + "loss": 0.9516, "step": 400 }, { "epoch": 1.623481781376518, - "grad_norm": 0.5359739036666953, + "grad_norm": 0.64457603858366, "learning_rate": 3.4172096256527865e-05, - "loss": 1.0678, + "loss": 0.9386, "step": 401 }, { "epoch": 1.6275303643724697, - "grad_norm": 0.5753964224845992, + "grad_norm": 0.6337402871760657, "learning_rate": 3.4132134604447595e-05, - "loss": 1.0775, + "loss": 0.9415, "step": 402 }, { "epoch": 1.631578947368421, - "grad_norm": 0.5946087494968246, + "grad_norm": 0.6816896628258218, "learning_rate": 3.409205995219561e-05, - "loss": 1.0369, + "loss": 0.9293, "step": 403 }, { "epoch": 1.6356275303643724, - "grad_norm": 0.5399246994967527, + "grad_norm": 0.6584104637254395, "learning_rate": 3.4051872620207765e-05, - "loss": 1.0859, + "loss": 0.9642, "step": 404 }, { "epoch": 1.639676113360324, - "grad_norm": 0.5561413467709494, + "grad_norm": 0.562575148325179, "learning_rate": 3.40115729298209e-05, - "loss": 1.0645, + "loss": 0.9387, "step": 405 }, { "epoch": 1.6437246963562753, - "grad_norm": 0.5876854854614066, + "grad_norm": 0.7421271127777155, "learning_rate": 3.397116120327025e-05, - "loss": 1.0574, + "loss": 0.9369, "step": 406 }, { "epoch": 1.6477732793522266, - "grad_norm": 0.6537542453485794, + "grad_norm": 0.8705618349094377, "learning_rate": 3.3930637763686916e-05, - "loss": 1.0944, + "loss": 0.9556, "step": 407 }, { "epoch": 1.6518218623481782, - "grad_norm": 0.7383750205301087, + "grad_norm": 0.6869762702168648, "learning_rate": 3.389000293509524e-05, - "loss": 1.0868, + "loss": 0.9555, "step": 408 }, { "epoch": 1.6558704453441295, - "grad_norm": 0.5344083369076991, + "grad_norm": 0.9712123023636346, "learning_rate": 3.3849257042410225e-05, - "loss": 1.08, + "loss": 0.9589, "step": 409 }, { "epoch": 1.6599190283400809, - "grad_norm": 0.6042704688837963, + "grad_norm": 0.885990836656725, "learning_rate": 3.380840041143494e-05, - "loss": 1.0533, + "loss": 0.9338, "step": 410 }, { "epoch": 1.6639676113360324, - "grad_norm": 0.5257800965475016, + "grad_norm": 0.9489075605452604, "learning_rate": 3.376743336885792e-05, - "loss": 1.0663, + "loss": 0.9387, "step": 411 }, { "epoch": 1.6680161943319838, - "grad_norm": 0.5353299423161705, + "grad_norm": 0.9266554858037437, "learning_rate": 3.372635624225055e-05, - "loss": 1.0895, + "loss": 0.9556, "step": 412 }, { "epoch": 1.6720647773279351, - "grad_norm": 0.5908591689476014, + "grad_norm": 0.7267215946608958, "learning_rate": 3.3685169360064445e-05, - "loss": 1.0823, + "loss": 0.9492, "step": 413 }, { "epoch": 1.6761133603238867, - "grad_norm": 0.6514145455944433, + "grad_norm": 0.6220962089256087, "learning_rate": 3.3643873051628804e-05, - "loss": 1.0665, + "loss": 0.9196, "step": 414 }, { "epoch": 1.680161943319838, - "grad_norm": 0.5436545506311266, + "grad_norm": 0.6159901192331753, "learning_rate": 3.3602467647147814e-05, - "loss": 1.0649, + "loss": 0.9261, "step": 415 }, { "epoch": 1.6842105263157894, - "grad_norm": 0.6705260670891561, + "grad_norm": 0.7643407224796857, "learning_rate": 3.356095347769799e-05, - "loss": 1.0751, + "loss": 0.9568, "step": 416 }, { "epoch": 1.688259109311741, - "grad_norm": 0.5158883577816178, + "grad_norm": 1.1468790152518473, "learning_rate": 3.351933087522552e-05, - "loss": 1.0635, + "loss": 0.9272, "step": 417 }, { "epoch": 1.6923076923076923, - "grad_norm": 0.5679085440157943, + "grad_norm": 0.8308424006037879, "learning_rate": 3.3477600172543626e-05, - "loss": 1.0562, + "loss": 0.9148, "step": 418 }, { "epoch": 1.6963562753036436, - "grad_norm": 0.6288096110094743, + "grad_norm": 0.5913399626137447, "learning_rate": 3.3435761703329894e-05, - "loss": 1.0939, + "loss": 0.9486, "step": 419 }, { "epoch": 1.7004048582995952, - "grad_norm": 0.5268313989934008, + "grad_norm": 0.8960191873119582, "learning_rate": 3.3393815802123606e-05, - "loss": 1.0849, + "loss": 0.9381, "step": 420 }, { "epoch": 1.7044534412955465, - "grad_norm": 0.5838688570297814, + "grad_norm": 0.9290940097339209, "learning_rate": 3.335176280432308e-05, - "loss": 1.0825, + "loss": 0.9598, "step": 421 }, { "epoch": 1.708502024291498, - "grad_norm": 0.6605548297312172, + "grad_norm": 1.2090574088535924, "learning_rate": 3.330960304618295e-05, - "loss": 1.0932, + "loss": 0.9516, "step": 422 }, { "epoch": 1.7125506072874495, - "grad_norm": 0.5475941666179691, + "grad_norm": 0.6539838969370315, "learning_rate": 3.326733686481153e-05, - "loss": 1.062, + "loss": 0.933, "step": 423 }, { "epoch": 1.7165991902834008, - "grad_norm": 0.5754361352745795, + "grad_norm": 0.7596906686781987, "learning_rate": 3.3224964598168084e-05, - "loss": 1.0662, + "loss": 0.9345, "step": 424 }, { "epoch": 1.7206477732793521, - "grad_norm": 0.525067780817146, + "grad_norm": 0.7380621449700827, "learning_rate": 3.318248658506012e-05, - "loss": 1.0659, + "loss": 0.9212, "step": 425 }, { "epoch": 1.7246963562753037, - "grad_norm": 0.5881763936014724, + "grad_norm": 0.8561622732066066, "learning_rate": 3.313990316514069e-05, - "loss": 1.0961, + "loss": 0.958, "step": 426 }, { "epoch": 1.728744939271255, - "grad_norm": 0.6577320260639545, + "grad_norm": 0.7341002572623938, "learning_rate": 3.309721467890571e-05, - "loss": 1.0438, + "loss": 0.9303, "step": 427 }, { "epoch": 1.7327935222672064, - "grad_norm": 0.5201001466090973, + "grad_norm": 0.6267457618730351, "learning_rate": 3.305442146769115e-05, - "loss": 1.0708, + "loss": 0.937, "step": 428 }, { "epoch": 1.736842105263158, - "grad_norm": 0.5587878570619761, + "grad_norm": 0.6101109000002897, "learning_rate": 3.30115238736704e-05, - "loss": 1.0384, + "loss": 0.918, "step": 429 }, { "epoch": 1.7408906882591093, - "grad_norm": 0.6265074902699193, + "grad_norm": 0.6318375153602251, "learning_rate": 3.296852223985148e-05, - "loss": 1.0733, + "loss": 0.9593, "step": 430 }, { "epoch": 1.7449392712550607, - "grad_norm": 0.6806204904417946, + "grad_norm": 0.5948877628792898, "learning_rate": 3.2925416910074315e-05, - "loss": 1.0822, + "loss": 0.9349, "step": 431 }, { "epoch": 1.7489878542510122, - "grad_norm": 0.6013195516369267, + "grad_norm": 0.6344049459582574, "learning_rate": 3.288220822900796e-05, - "loss": 1.0575, + "loss": 0.9327, "step": 432 }, { "epoch": 1.7530364372469636, - "grad_norm": 0.47358032435473346, + "grad_norm": 0.6461977998296194, "learning_rate": 3.283889654214788e-05, - "loss": 1.0516, + "loss": 0.9329, "step": 433 }, { "epoch": 1.757085020242915, - "grad_norm": 0.5225663615340297, + "grad_norm": 0.5775129599536977, "learning_rate": 3.279548219581317e-05, - "loss": 1.1069, + "loss": 0.9773, "step": 434 }, { "epoch": 1.7611336032388665, - "grad_norm": 0.4560484833164418, + "grad_norm": 0.5884483209807626, "learning_rate": 3.275196553714379e-05, - "loss": 1.0918, + "loss": 0.9559, "step": 435 }, { "epoch": 1.7651821862348178, - "grad_norm": 0.5406080134444121, + "grad_norm": 0.5470088018590539, "learning_rate": 3.270834691409778e-05, - "loss": 1.058, + "loss": 0.92, "step": 436 }, { "epoch": 1.7692307692307692, - "grad_norm": 0.5008423563691626, + "grad_norm": 0.7177756722437414, "learning_rate": 3.266462667544849e-05, - "loss": 1.0554, + "loss": 0.9176, "step": 437 }, { "epoch": 1.7732793522267207, - "grad_norm": 0.5539940707238067, + "grad_norm": 0.5605463343143123, "learning_rate": 3.262080517078178e-05, - "loss": 1.0874, + "loss": 0.9522, "step": 438 }, { "epoch": 1.777327935222672, - "grad_norm": 0.5151746359113535, + "grad_norm": 0.6270965417712344, "learning_rate": 3.257688275049323e-05, - "loss": 1.0681, + "loss": 0.93, "step": 439 }, { "epoch": 1.7813765182186234, - "grad_norm": 0.6672411028850963, + "grad_norm": 0.6839546603903204, "learning_rate": 3.253285976578535e-05, - "loss": 1.0947, + "loss": 0.9313, "step": 440 }, { "epoch": 1.785425101214575, - "grad_norm": 0.6190971427764351, + "grad_norm": 0.5615911706781257, "learning_rate": 3.2488736568664756e-05, - "loss": 1.0697, + "loss": 0.9353, "step": 441 }, { "epoch": 1.7894736842105263, - "grad_norm": 0.5239697698164889, + "grad_norm": 0.5708004351003453, "learning_rate": 3.244451351193935e-05, - "loss": 1.0675, + "loss": 0.9416, "step": 442 }, { "epoch": 1.7935222672064777, - "grad_norm": 0.8099341539389999, + "grad_norm": 0.5949958255362322, "learning_rate": 3.240019094921551e-05, - "loss": 1.0587, + "loss": 0.9208, "step": 443 }, { "epoch": 1.7975708502024292, - "grad_norm": 0.5736803802671554, + "grad_norm": 0.5351638390410699, "learning_rate": 3.235576923489527e-05, - "loss": 1.0451, + "loss": 0.931, "step": 444 }, { "epoch": 1.8016194331983806, - "grad_norm": 0.6372527076907771, + "grad_norm": 0.5258087593774992, "learning_rate": 3.2311248724173476e-05, - "loss": 1.0593, + "loss": 0.9193, "step": 445 }, { "epoch": 1.805668016194332, - "grad_norm": 0.6417121925642931, + "grad_norm": 0.5127441857966027, "learning_rate": 3.226662977303494e-05, - "loss": 1.0549, + "loss": 0.9129, "step": 446 }, { "epoch": 1.8097165991902835, - "grad_norm": 0.5301608370647144, + "grad_norm": 0.5054557154257424, "learning_rate": 3.2221912738251626e-05, - "loss": 1.0646, + "loss": 0.9228, "step": 447 }, { "epoch": 1.8137651821862348, - "grad_norm": 0.6289351322051107, + "grad_norm": 0.4934682378395103, "learning_rate": 3.2177097977379736e-05, - "loss": 1.0642, + "loss": 0.9221, "step": 448 }, { "epoch": 1.8178137651821862, - "grad_norm": 0.5588455023728987, + "grad_norm": 0.5493566186654533, "learning_rate": 3.21321858487569e-05, - "loss": 1.0936, + "loss": 0.9556, "step": 449 }, { "epoch": 1.8218623481781377, - "grad_norm": 0.5925904407038142, + "grad_norm": 0.46495341652277544, "learning_rate": 3.208717671149931e-05, - "loss": 1.0931, + "loss": 0.9541, "step": 450 }, { "epoch": 1.825910931174089, - "grad_norm": 0.6166411578480747, + "grad_norm": 0.590102414724878, "learning_rate": 3.204207092549882e-05, - "loss": 1.0708, + "loss": 0.9203, "step": 451 }, { "epoch": 1.8299595141700404, - "grad_norm": 0.5136355479386089, + "grad_norm": 0.5138290936760064, "learning_rate": 3.19968688514201e-05, - "loss": 1.0752, + "loss": 0.938, "step": 452 }, { "epoch": 1.834008097165992, - "grad_norm": 0.5614976380168106, + "grad_norm": 0.6285161499192754, "learning_rate": 3.195157085069774e-05, - "loss": 1.048, + "loss": 0.908, "step": 453 }, { "epoch": 1.8380566801619433, - "grad_norm": 0.41730695745287705, + "grad_norm": 0.5779250156942504, "learning_rate": 3.190617728553332e-05, - "loss": 1.0589, + "loss": 0.9213, "step": 454 }, { "epoch": 1.8421052631578947, - "grad_norm": 0.5641941064648709, + "grad_norm": 0.7054323407227598, "learning_rate": 3.1860688518892606e-05, - "loss": 1.0752, + "loss": 0.9426, "step": 455 }, { "epoch": 1.8461538461538463, - "grad_norm": 0.538004945404975, + "grad_norm": 0.6645465318973225, "learning_rate": 3.181510491450254e-05, - "loss": 1.0878, + "loss": 0.9368, "step": 456 }, { "epoch": 1.8502024291497976, - "grad_norm": 0.5559104567502047, + "grad_norm": 0.6793240838252204, "learning_rate": 3.176942683684842e-05, - "loss": 1.095, + "loss": 0.9738, "step": 457 }, { "epoch": 1.854251012145749, - "grad_norm": 0.6035798621110092, + "grad_norm": 0.8280247119305022, "learning_rate": 3.1723654651170934e-05, - "loss": 1.0928, + "loss": 0.9645, "step": 458 }, { "epoch": 1.8582995951417005, - "grad_norm": 0.6005001380443482, + "grad_norm": 0.7347481086065992, "learning_rate": 3.167778872346327e-05, - "loss": 1.0716, + "loss": 0.9339, "step": 459 }, { "epoch": 1.8623481781376519, - "grad_norm": 0.5203767741115847, + "grad_norm": 0.6083904078540814, "learning_rate": 3.163182942046815e-05, - "loss": 1.0576, + "loss": 0.9331, "step": 460 }, { "epoch": 1.8663967611336032, - "grad_norm": 0.6557534331573859, + "grad_norm": 0.5512558228418888, "learning_rate": 3.158577710967494e-05, - "loss": 1.0699, + "loss": 0.9476, "step": 461 }, { "epoch": 1.8704453441295548, - "grad_norm": 0.5364587844028244, + "grad_norm": 0.5729491664063789, "learning_rate": 3.153963215931669e-05, - "loss": 1.0763, + "loss": 0.9395, "step": 462 }, { "epoch": 1.874493927125506, - "grad_norm": 0.6380534053060034, + "grad_norm": 0.8490224814606387, "learning_rate": 3.14933949383672e-05, - "loss": 1.0593, + "loss": 0.9263, "step": 463 }, { "epoch": 1.8785425101214575, - "grad_norm": 0.5699341788293785, + "grad_norm": 0.7436791845780646, "learning_rate": 3.144706581653805e-05, - "loss": 1.078, + "loss": 0.956, "step": 464 }, { "epoch": 1.882591093117409, - "grad_norm": 0.5787488137522467, + "grad_norm": 0.6783667229275504, "learning_rate": 3.140064516427566e-05, - "loss": 1.09, + "loss": 0.9391, "step": 465 }, { "epoch": 1.8866396761133604, - "grad_norm": 0.5284597120847465, + "grad_norm": 0.624127230523046, "learning_rate": 3.135413335275833e-05, - "loss": 1.0462, + "loss": 0.9079, "step": 466 }, { "epoch": 1.8906882591093117, - "grad_norm": 0.6310093194026928, + "grad_norm": 0.5554169959523534, "learning_rate": 3.130753075389327e-05, - "loss": 1.0667, + "loss": 0.9331, "step": 467 }, { "epoch": 1.8947368421052633, - "grad_norm": 0.535290188793271, + "grad_norm": 0.6814806585973552, "learning_rate": 3.1260837740313625e-05, - "loss": 1.0785, + "loss": 0.9508, "step": 468 }, { "epoch": 1.8987854251012146, - "grad_norm": 0.7277238647743011, + "grad_norm": 0.7044762444416278, "learning_rate": 3.121405468537547e-05, - "loss": 1.0714, + "loss": 0.9275, "step": 469 }, { "epoch": 1.902834008097166, - "grad_norm": 0.522964636149426, + "grad_norm": 0.6593170765367392, "learning_rate": 3.116718196315488e-05, - "loss": 1.0552, + "loss": 0.9391, "step": 470 }, { "epoch": 1.9068825910931175, - "grad_norm": 0.7208127381747786, + "grad_norm": 0.7008280293960216, "learning_rate": 3.1120219948444894e-05, - "loss": 1.0906, + "loss": 0.9551, "step": 471 }, { "epoch": 1.9109311740890689, - "grad_norm": 0.6055514896898919, + "grad_norm": 0.5928148461037925, "learning_rate": 3.107316901675252e-05, - "loss": 1.0729, + "loss": 0.9288, "step": 472 }, { "epoch": 1.9149797570850202, - "grad_norm": 0.643748810382086, + "grad_norm": 0.5532980688428584, "learning_rate": 3.1026029544295745e-05, - "loss": 1.0479, + "loss": 0.9153, "step": 473 }, { "epoch": 1.9190283400809718, - "grad_norm": 0.5567358108434363, + "grad_norm": 0.5800334404403348, "learning_rate": 3.097880190800057e-05, - "loss": 1.0898, + "loss": 0.9466, "step": 474 }, { "epoch": 1.9230769230769231, - "grad_norm": 0.5130281231623158, + "grad_norm": 0.7263034230369951, "learning_rate": 3.093148648549788e-05, - "loss": 1.072, + "loss": 0.9315, "step": 475 }, { "epoch": 1.9271255060728745, - "grad_norm": 0.5542448732460946, + "grad_norm": 0.6458027656560078, "learning_rate": 3.088408365512055e-05, - "loss": 1.0624, + "loss": 0.9367, "step": 476 }, { "epoch": 1.931174089068826, - "grad_norm": 0.5507345379469023, + "grad_norm": 0.8969399166010747, "learning_rate": 3.083659379590034e-05, - "loss": 1.0625, + "loss": 0.9407, "step": 477 }, { "epoch": 1.9352226720647774, - "grad_norm": 0.4829435365197085, + "grad_norm": 0.694186323291845, "learning_rate": 3.0789017287564913e-05, - "loss": 1.068, + "loss": 0.9291, "step": 478 }, { "epoch": 1.9392712550607287, - "grad_norm": 0.5776594164930579, + "grad_norm": 0.5511745792730092, "learning_rate": 3.074135451053475e-05, - "loss": 1.0635, + "loss": 0.929, "step": 479 }, { "epoch": 1.9433198380566803, - "grad_norm": 0.49220985620751295, + "grad_norm": 0.6848241275814233, "learning_rate": 3.069360584592015e-05, - "loss": 1.0798, + "loss": 0.9519, "step": 480 }, { "epoch": 1.9473684210526314, - "grad_norm": 0.5504146032209003, + "grad_norm": 0.6499276833008483, "learning_rate": 3.064577167551816e-05, - "loss": 1.0907, + "loss": 0.9472, "step": 481 }, { "epoch": 1.951417004048583, - "grad_norm": 0.5007069597145724, + "grad_norm": 0.6457026596435437, "learning_rate": 3.059785238180953e-05, - "loss": 1.0894, + "loss": 0.944, "step": 482 }, { "epoch": 1.9554655870445345, - "grad_norm": 0.510501877850237, + "grad_norm": 0.6106311075264185, "learning_rate": 3.054984834795566e-05, - "loss": 1.0732, + "loss": 0.9439, "step": 483 }, { "epoch": 1.9595141700404857, - "grad_norm": 0.4976988228836568, + "grad_norm": 0.5406889485959955, "learning_rate": 3.0501759957795537e-05, - "loss": 1.0457, + "loss": 0.9307, "step": 484 }, { "epoch": 1.9635627530364372, - "grad_norm": 0.4673791861722963, + "grad_norm": 0.7236100067821493, "learning_rate": 3.0453587595842627e-05, - "loss": 1.0509, + "loss": 0.9211, "step": 485 }, { "epoch": 1.9676113360323888, - "grad_norm": 0.5286581334210425, + "grad_norm": 0.5620660238123119, "learning_rate": 3.0405331647281856e-05, - "loss": 1.1013, + "loss": 0.9516, "step": 486 }, { "epoch": 1.97165991902834, - "grad_norm": 0.527112425559158, + "grad_norm": 0.6206464082750514, "learning_rate": 3.0356992497966508e-05, - "loss": 1.0811, + "loss": 0.9386, "step": 487 }, { "epoch": 1.9757085020242915, - "grad_norm": 0.4758062259798258, + "grad_norm": 0.6945782050033186, "learning_rate": 3.0308570534415137e-05, - "loss": 1.0758, + "loss": 0.933, "step": 488 }, { "epoch": 1.979757085020243, - "grad_norm": 0.4911942865191961, + "grad_norm": 0.7133662666598716, "learning_rate": 3.0260066143808467e-05, - "loss": 1.0641, + "loss": 0.9344, "step": 489 }, { "epoch": 1.9838056680161942, - "grad_norm": 0.537887727719594, + "grad_norm": 0.5855684318076528, "learning_rate": 3.021147971398632e-05, - "loss": 1.0728, + "loss": 0.9431, "step": 490 }, { "epoch": 1.9878542510121457, - "grad_norm": 0.5439313495035273, + "grad_norm": 0.5531222615444646, "learning_rate": 3.0162811633444478e-05, - "loss": 1.0692, + "loss": 0.9253, "step": 491 }, { "epoch": 1.9919028340080973, - "grad_norm": 0.4967933309941132, + "grad_norm": 0.6526502373182302, "learning_rate": 3.0114062291331635e-05, - "loss": 1.0636, + "loss": 0.9264, "step": 492 }, { "epoch": 1.9959514170040484, - "grad_norm": 0.5680540076062215, + "grad_norm": 0.4280978015965861, "learning_rate": 3.0065232077446217e-05, - "loss": 1.0433, + "loss": 0.927, "step": 493 }, { "epoch": 2.0, - "grad_norm": 0.5168824344698549, + "grad_norm": 0.5132239754624033, "learning_rate": 3.001632138223332e-05, - "loss": 1.0381, + "loss": 0.9196, "step": 494 }, { "epoch": 2.0040485829959516, - "grad_norm": 0.710685723258479, + "grad_norm": 0.5680768421144835, "learning_rate": 2.9967330596781565e-05, - "loss": 0.9889, + "loss": 0.9068, "step": 495 }, { "epoch": 2.0080971659919027, - "grad_norm": 0.5321857125214415, + "grad_norm": 0.5154792137721976, "learning_rate": 2.9918260112819952e-05, - "loss": 0.9625, + "loss": 0.9041, "step": 496 }, { "epoch": 2.0121457489878543, - "grad_norm": 0.8031269314179498, + "grad_norm": 0.716430529034805, "learning_rate": 2.9869110322714778e-05, - "loss": 0.9662, + "loss": 0.8936, "step": 497 }, { "epoch": 2.016194331983806, - "grad_norm": 0.7376844509826479, + "grad_norm": 0.7729364109665723, "learning_rate": 2.9819881619466447e-05, - "loss": 0.9679, + "loss": 0.8893, "step": 498 }, { "epoch": 2.020242914979757, - "grad_norm": 0.8519110829824833, + "grad_norm": 0.6633071481502456, "learning_rate": 2.9770574396706364e-05, - "loss": 0.9658, + "loss": 0.8889, "step": 499 }, { "epoch": 2.0242914979757085, - "grad_norm": 0.8400559493448463, + "grad_norm": 0.5782986903645834, "learning_rate": 2.972118904869376e-05, - "loss": 0.9663, + "loss": 0.8907, "step": 500 }, { "epoch": 2.02834008097166, - "grad_norm": 0.7475106567939678, + "grad_norm": 0.5488411446049505, "learning_rate": 2.9671725970312563e-05, - "loss": 0.9642, + "loss": 0.8945, "step": 501 }, { "epoch": 2.032388663967611, - "grad_norm": 0.8726773696326442, + "grad_norm": 0.593945351142755, "learning_rate": 2.9622185557068238e-05, - "loss": 0.9727, + "loss": 0.902, "step": 502 }, { "epoch": 2.0364372469635628, - "grad_norm": 0.711955888277968, + "grad_norm": 0.6292897057194601, "learning_rate": 2.95725682050846e-05, - "loss": 0.9549, + "loss": 0.8937, "step": 503 }, { "epoch": 2.0404858299595143, - "grad_norm": 0.8810911112546097, + "grad_norm": 0.6903137137443031, "learning_rate": 2.9522874311100684e-05, - "loss": 0.9671, + "loss": 0.8764, "step": 504 }, { "epoch": 2.0445344129554655, - "grad_norm": 0.5662961488470308, + "grad_norm": 0.6907850686575893, "learning_rate": 2.9473104272467526e-05, - "loss": 0.9594, + "loss": 0.9009, "step": 505 }, { "epoch": 2.048582995951417, - "grad_norm": 0.8185321874743823, + "grad_norm": 0.8356989496224466, "learning_rate": 2.942325848714505e-05, - "loss": 0.9639, + "loss": 0.8903, "step": 506 }, { "epoch": 2.0526315789473686, - "grad_norm": 0.6632449504389368, + "grad_norm": 0.5794503787082406, "learning_rate": 2.9373337353698804e-05, - "loss": 0.9677, + "loss": 0.8871, "step": 507 }, { "epoch": 2.0566801619433197, - "grad_norm": 0.6269896313844657, + "grad_norm": 0.7366744633270342, "learning_rate": 2.932334127129686e-05, - "loss": 0.9581, + "loss": 0.8929, "step": 508 }, { "epoch": 2.0607287449392713, - "grad_norm": 0.7868532317667555, + "grad_norm": 0.6858094471228489, "learning_rate": 2.9273270639706547e-05, - "loss": 0.9519, + "loss": 0.8747, "step": 509 }, { "epoch": 2.064777327935223, - "grad_norm": 0.5294500146198006, + "grad_norm": 0.7775305889874989, "learning_rate": 2.922312585929131e-05, - "loss": 0.974, + "loss": 0.9135, "step": 510 }, { "epoch": 2.068825910931174, - "grad_norm": 0.7172383691164358, + "grad_norm": 0.592745527583372, "learning_rate": 2.9172907331007478e-05, - "loss": 0.9598, + "loss": 0.8983, "step": 511 }, { "epoch": 2.0728744939271255, - "grad_norm": 0.5275936230138877, + "grad_norm": 0.637910219042102, "learning_rate": 2.9122615456401053e-05, - "loss": 0.9568, + "loss": 0.8902, "step": 512 }, { "epoch": 2.076923076923077, - "grad_norm": 0.5369340410169218, + "grad_norm": 0.6742349448180341, "learning_rate": 2.9072250637604535e-05, - "loss": 0.9328, + "loss": 0.8768, "step": 513 }, { "epoch": 2.080971659919028, - "grad_norm": 0.655274467256145, + "grad_norm": 0.8250592958692304, "learning_rate": 2.9021813277333667e-05, - "loss": 0.9875, + "loss": 0.9128, "step": 514 }, { "epoch": 2.08502024291498, - "grad_norm": 0.502891610340589, + "grad_norm": 0.6521757885291003, "learning_rate": 2.8971303778884244e-05, - "loss": 0.9414, + "loss": 0.8663, "step": 515 }, { "epoch": 2.0890688259109313, - "grad_norm": 0.5285815181160415, + "grad_norm": 0.5423960335775939, "learning_rate": 2.8920722546128863e-05, - "loss": 0.9637, + "loss": 0.9009, "step": 516 }, { "epoch": 2.0931174089068825, - "grad_norm": 0.6106013437988467, + "grad_norm": 0.7300638909543451, "learning_rate": 2.887006998351372e-05, - "loss": 0.9712, + "loss": 0.8877, "step": 517 }, { "epoch": 2.097165991902834, - "grad_norm": 0.487333397593873, + "grad_norm": 0.8359344608792905, "learning_rate": 2.881934649605535e-05, - "loss": 0.9585, + "loss": 0.9055, "step": 518 }, { "epoch": 2.1012145748987856, - "grad_norm": 0.5633264273068143, + "grad_norm": 0.5943991610556894, "learning_rate": 2.8768552489337408e-05, - "loss": 0.968, + "loss": 0.8958, "step": 519 }, { "epoch": 2.1052631578947367, - "grad_norm": 0.5622093457125968, + "grad_norm": 0.584968428391503, "learning_rate": 2.8717688369507425e-05, - "loss": 0.9779, + "loss": 0.9295, "step": 520 }, { "epoch": 2.1093117408906883, - "grad_norm": 0.5393069372283978, + "grad_norm": 0.5287040107776219, "learning_rate": 2.8666754543273535e-05, - "loss": 0.9799, + "loss": 0.897, "step": 521 }, { "epoch": 2.11336032388664, - "grad_norm": 0.5483885026017794, + "grad_norm": 0.7023688237011424, "learning_rate": 2.8615751417901273e-05, - "loss": 0.9347, + "loss": 0.884, "step": 522 }, { "epoch": 2.117408906882591, - "grad_norm": 0.5826095046526765, + "grad_norm": 0.6050214313064772, "learning_rate": 2.856467940121025e-05, - "loss": 0.9574, + "loss": 0.8901, "step": 523 }, { "epoch": 2.1214574898785425, - "grad_norm": 0.45505914899320044, + "grad_norm": 0.5330188134722201, "learning_rate": 2.8513538901570964e-05, - "loss": 0.9618, + "loss": 0.8998, "step": 524 }, { "epoch": 2.125506072874494, - "grad_norm": 0.5481406994733198, + "grad_norm": 0.4209461803038103, "learning_rate": 2.846233032790149e-05, - "loss": 0.9253, + "loss": 0.8611, "step": 525 }, { "epoch": 2.1295546558704452, - "grad_norm": 0.5652769433684767, + "grad_norm": 0.5854056866255548, "learning_rate": 2.84110540896642e-05, - "loss": 0.9639, + "loss": 0.8784, "step": 526 }, { "epoch": 2.133603238866397, - "grad_norm": 0.5330442876266908, + "grad_norm": 0.4769790939825551, "learning_rate": 2.8359710596862555e-05, - "loss": 0.964, + "loss": 0.8893, "step": 527 }, { "epoch": 2.1376518218623484, - "grad_norm": 0.6952919600130274, + "grad_norm": 0.6730492114381329, "learning_rate": 2.8308300260037734e-05, - "loss": 0.9636, + "loss": 0.8861, "step": 528 }, { "epoch": 2.1417004048582995, - "grad_norm": 0.44595945285220484, + "grad_norm": 0.4867308512234734, "learning_rate": 2.8256823490265424e-05, - "loss": 0.962, + "loss": 0.8922, "step": 529 }, { "epoch": 2.145748987854251, - "grad_norm": 0.5951123380152432, + "grad_norm": 0.6147861078704552, "learning_rate": 2.8205280699152515e-05, - "loss": 0.9736, + "loss": 0.903, "step": 530 }, { "epoch": 2.1497975708502026, - "grad_norm": 0.5709041000237276, + "grad_norm": 0.44362910907650827, "learning_rate": 2.8153672298833777e-05, - "loss": 0.9764, + "loss": 0.9098, "step": 531 }, { "epoch": 2.1538461538461537, - "grad_norm": 0.5961373862016145, + "grad_norm": 0.4971316804629912, "learning_rate": 2.8101998701968615e-05, - "loss": 0.9743, + "loss": 0.8993, "step": 532 }, { "epoch": 2.1578947368421053, - "grad_norm": 0.5593160147575131, + "grad_norm": 0.542998141649332, "learning_rate": 2.805026032173773e-05, - "loss": 0.9764, + "loss": 0.8881, "step": 533 }, { "epoch": 2.161943319838057, - "grad_norm": 0.5008333600497884, + "grad_norm": 0.425665499381359, "learning_rate": 2.7998457571839832e-05, - "loss": 0.9559, + "loss": 0.887, "step": 534 }, { "epoch": 2.165991902834008, - "grad_norm": 0.4700282218933336, + "grad_norm": 0.44860085480924294, "learning_rate": 2.7946590866488322e-05, - "loss": 0.962, + "loss": 0.887, "step": 535 }, { "epoch": 2.1700404858299596, - "grad_norm": 0.47250328330808655, + "grad_norm": 0.5721862755499443, "learning_rate": 2.7894660620408e-05, - "loss": 0.9602, + "loss": 0.8943, "step": 536 }, { "epoch": 2.174089068825911, - "grad_norm": 0.511671837789726, + "grad_norm": 0.6037083296487049, "learning_rate": 2.784266724883173e-05, - "loss": 0.9468, + "loss": 0.8871, "step": 537 }, { "epoch": 2.1781376518218623, - "grad_norm": 0.4508110010187165, + "grad_norm": 0.5844893004025372, "learning_rate": 2.7790611167497128e-05, - "loss": 0.9314, + "loss": 0.876, "step": 538 }, { "epoch": 2.182186234817814, - "grad_norm": 0.46560685791323836, + "grad_norm": 0.5172509523687889, "learning_rate": 2.7738492792643224e-05, - "loss": 0.9577, + "loss": 0.892, "step": 539 }, { "epoch": 2.1862348178137654, - "grad_norm": 0.5436128763813932, + "grad_norm": 0.5660905777334954, "learning_rate": 2.768631254100716e-05, - "loss": 0.9625, + "loss": 0.8878, "step": 540 }, { "epoch": 2.1902834008097165, - "grad_norm": 0.45389249911131313, + "grad_norm": 0.6093565017649889, "learning_rate": 2.7634070829820836e-05, - "loss": 0.9576, + "loss": 0.8941, "step": 541 }, { "epoch": 2.194331983805668, - "grad_norm": 0.5182389283399222, + "grad_norm": 0.6598840445361779, "learning_rate": 2.7581768076807592e-05, - "loss": 0.9734, + "loss": 0.8976, "step": 542 }, { "epoch": 2.1983805668016196, - "grad_norm": 0.5069489642826847, + "grad_norm": 0.4359430860696874, "learning_rate": 2.7529404700178832e-05, - "loss": 0.9666, + "loss": 0.8762, "step": 543 }, { "epoch": 2.2024291497975708, - "grad_norm": 0.5029321008120494, + "grad_norm": 0.6157997002602271, "learning_rate": 2.7476981118630735e-05, - "loss": 0.9373, + "loss": 0.8745, "step": 544 }, { "epoch": 2.2064777327935223, - "grad_norm": 0.5147663579482141, + "grad_norm": 0.6942940892136789, "learning_rate": 2.742449775134085e-05, - "loss": 0.9639, + "loss": 0.8978, "step": 545 }, { "epoch": 2.2105263157894735, - "grad_norm": 0.5899056132864572, + "grad_norm": 0.6869778851730045, "learning_rate": 2.7371955017964788e-05, - "loss": 0.9522, + "loss": 0.875, "step": 546 }, { "epoch": 2.214574898785425, - "grad_norm": 0.5179537834510582, + "grad_norm": 0.6012157929738973, "learning_rate": 2.7319353338632845e-05, - "loss": 0.9775, + "loss": 0.9153, "step": 547 }, { "epoch": 2.2186234817813766, - "grad_norm": 0.4934206299263931, + "grad_norm": 0.40931000693792574, "learning_rate": 2.7266693133946645e-05, - "loss": 0.9716, + "loss": 0.8998, "step": 548 }, { "epoch": 2.2226720647773277, - "grad_norm": 0.6161333758299129, + "grad_norm": 0.4848690688041366, "learning_rate": 2.7213974824975785e-05, - "loss": 0.9552, + "loss": 0.8876, "step": 549 }, { "epoch": 2.2267206477732793, - "grad_norm": 0.47146559545266487, + "grad_norm": 0.578422479395952, "learning_rate": 2.7161198833254464e-05, - "loss": 0.9581, + "loss": 0.8952, "step": 550 }, { "epoch": 2.230769230769231, - "grad_norm": 0.5500079692504681, + "grad_norm": 0.5491520220857032, "learning_rate": 2.7108365580778095e-05, - "loss": 0.9616, + "loss": 0.9065, "step": 551 }, { "epoch": 2.234817813765182, - "grad_norm": 0.5545805361289696, + "grad_norm": 0.4199496876887628, "learning_rate": 2.705547548999996e-05, - "loss": 0.9794, + "loss": 0.8975, "step": 552 }, { "epoch": 2.2388663967611335, - "grad_norm": 0.4558169892304106, + "grad_norm": 0.5460721535468303, "learning_rate": 2.7002528983827817e-05, - "loss": 0.9701, + "loss": 0.9046, "step": 553 }, { "epoch": 2.242914979757085, - "grad_norm": 0.6173218573568598, + "grad_norm": 0.5188880005179151, "learning_rate": 2.6949526485620517e-05, - "loss": 0.941, + "loss": 0.8725, "step": 554 }, { "epoch": 2.246963562753036, - "grad_norm": 0.5135889100138243, + "grad_norm": 0.5288038528907943, "learning_rate": 2.6896468419184623e-05, - "loss": 0.9767, + "loss": 0.8955, "step": 555 }, { "epoch": 2.251012145748988, - "grad_norm": 0.5019591407572072, + "grad_norm": 0.4551110722509293, "learning_rate": 2.6843355208771024e-05, - "loss": 0.9713, + "loss": 0.8864, "step": 556 }, { "epoch": 2.2550607287449393, - "grad_norm": 0.552354893674315, + "grad_norm": 0.48803557894870625, "learning_rate": 2.6790187279071536e-05, - "loss": 0.9743, + "loss": 0.9032, "step": 557 }, { "epoch": 2.2591093117408905, - "grad_norm": 0.5059885669206644, + "grad_norm": 0.4072041639624624, "learning_rate": 2.67369650552155e-05, - "loss": 0.978, + "loss": 0.89, "step": 558 }, { "epoch": 2.263157894736842, - "grad_norm": 0.47630505792443534, + "grad_norm": 0.4406440235159891, "learning_rate": 2.6683688962766407e-05, - "loss": 0.9684, + "loss": 0.9041, "step": 559 }, { "epoch": 2.2672064777327936, - "grad_norm": 0.5542629265346289, + "grad_norm": 0.4604641390768294, "learning_rate": 2.6630359427718468e-05, - "loss": 0.961, + "loss": 0.9029, "step": 560 }, { "epoch": 2.2712550607287447, - "grad_norm": 0.49238351631550253, + "grad_norm": 0.4390849739491862, "learning_rate": 2.6576976876493225e-05, - "loss": 0.9551, + "loss": 0.881, "step": 561 }, { "epoch": 2.2753036437246963, - "grad_norm": 0.5368071906373748, + "grad_norm": 0.4582639482753826, "learning_rate": 2.652354173593614e-05, - "loss": 0.9879, + "loss": 0.9042, "step": 562 }, { "epoch": 2.279352226720648, - "grad_norm": 0.5217995737335323, + "grad_norm": 0.4268148697964676, "learning_rate": 2.647005443331317e-05, - "loss": 0.9615, + "loss": 0.8885, "step": 563 }, { "epoch": 2.283400809716599, - "grad_norm": 0.46760949672577173, + "grad_norm": 0.41542780047582484, "learning_rate": 2.6416515396307354e-05, - "loss": 0.9751, + "loss": 0.8994, "step": 564 }, { "epoch": 2.2874493927125505, - "grad_norm": 0.4953215975498102, + "grad_norm": 0.4737419438147517, "learning_rate": 2.636292505301542e-05, - "loss": 0.98, + "loss": 0.9002, "step": 565 }, { "epoch": 2.291497975708502, - "grad_norm": 0.49525214010334295, + "grad_norm": 0.4982819906332182, "learning_rate": 2.630928383194431e-05, - "loss": 0.9549, + "loss": 0.8964, "step": 566 }, { "epoch": 2.2955465587044532, - "grad_norm": 0.4345854422532677, + "grad_norm": 0.4658647065134157, "learning_rate": 2.625559216200781e-05, - "loss": 0.9542, + "loss": 0.8863, "step": 567 }, { "epoch": 2.299595141700405, - "grad_norm": 0.44175105294623795, + "grad_norm": 0.5590453372146358, "learning_rate": 2.620185047252308e-05, - "loss": 0.9942, + "loss": 0.9208, "step": 568 }, { "epoch": 2.3036437246963564, - "grad_norm": 0.48169860730979475, + "grad_norm": 0.48439857106739226, "learning_rate": 2.614805919320723e-05, - "loss": 0.9683, + "loss": 0.8881, "step": 569 }, { "epoch": 2.3076923076923075, - "grad_norm": 0.4877146316408746, + "grad_norm": 0.5262042753486375, "learning_rate": 2.6094218754173895e-05, - "loss": 0.9781, + "loss": 0.8973, "step": 570 }, { "epoch": 2.311740890688259, - "grad_norm": 0.5129207120154343, + "grad_norm": 0.4282957331355162, "learning_rate": 2.60403295859298e-05, - "loss": 0.9555, + "loss": 0.882, "step": 571 }, { "epoch": 2.3157894736842106, - "grad_norm": 0.47841047182318724, + "grad_norm": 0.5367741729765978, "learning_rate": 2.5986392119371297e-05, - "loss": 0.9623, + "loss": 0.9048, "step": 572 }, { "epoch": 2.3198380566801617, - "grad_norm": 0.4545332449931287, + "grad_norm": 0.4164338416851721, "learning_rate": 2.5932406785780928e-05, - "loss": 0.9354, + "loss": 0.8609, "step": 573 }, { "epoch": 2.3238866396761133, - "grad_norm": 0.4833439729686478, + "grad_norm": 0.4825288763645802, "learning_rate": 2.5878374016823987e-05, - "loss": 0.9811, + "loss": 0.8919, "step": 574 }, { "epoch": 2.327935222672065, - "grad_norm": 0.4720190004020666, + "grad_norm": 0.46019765834942383, "learning_rate": 2.5824294244545045e-05, - "loss": 0.9589, + "loss": 0.8966, "step": 575 }, { "epoch": 2.331983805668016, - "grad_norm": 0.5230768831110423, + "grad_norm": 0.4544628863751653, "learning_rate": 2.577016790136453e-05, - "loss": 0.9944, + "loss": 0.899, "step": 576 }, { "epoch": 2.3360323886639676, - "grad_norm": 0.5268214376749982, + "grad_norm": 0.4628281081063216, "learning_rate": 2.5715995420075225e-05, - "loss": 0.9758, + "loss": 0.9066, "step": 577 }, { "epoch": 2.340080971659919, - "grad_norm": 0.47272420215378086, + "grad_norm": 0.5398130832165038, "learning_rate": 2.5661777233838864e-05, - "loss": 0.9683, + "loss": 0.8909, "step": 578 }, { "epoch": 2.3441295546558703, - "grad_norm": 0.6008004304970271, + "grad_norm": 0.5009061695015287, "learning_rate": 2.5607513776182616e-05, - "loss": 0.975, + "loss": 0.8855, "step": 579 }, { "epoch": 2.348178137651822, - "grad_norm": 0.4775282978150046, + "grad_norm": 0.547525711721424, "learning_rate": 2.5553205480995633e-05, - "loss": 0.9577, + "loss": 0.8811, "step": 580 }, { "epoch": 2.3522267206477734, - "grad_norm": 0.6183480590145304, + "grad_norm": 0.5101405444799098, "learning_rate": 2.5498852782525604e-05, - "loss": 0.9963, + "loss": 0.9258, "step": 581 }, { "epoch": 2.3562753036437245, - "grad_norm": 0.5029298266977416, + "grad_norm": 0.5574223761698288, "learning_rate": 2.5444456115375252e-05, - "loss": 0.967, + "loss": 0.9071, "step": 582 }, { "epoch": 2.360323886639676, - "grad_norm": 0.5062202235753198, + "grad_norm": 0.5503863099835893, "learning_rate": 2.539001591449889e-05, - "loss": 0.9915, + "loss": 0.9232, "step": 583 }, { "epoch": 2.3643724696356276, - "grad_norm": 0.5045211228148797, + "grad_norm": 0.4897877445228228, "learning_rate": 2.53355326151989e-05, - "loss": 0.9583, + "loss": 0.8745, "step": 584 }, { "epoch": 2.3684210526315788, - "grad_norm": 0.5231672726130052, + "grad_norm": 0.5171963852855628, "learning_rate": 2.5281006653122288e-05, - "loss": 0.9493, + "loss": 0.9011, "step": 585 }, { "epoch": 2.3724696356275303, - "grad_norm": 0.5945746027580209, + "grad_norm": 0.5257208394004133, "learning_rate": 2.5226438464257217e-05, - "loss": 0.9595, + "loss": 0.8776, "step": 586 }, { "epoch": 2.376518218623482, - "grad_norm": 0.4736110599255813, + "grad_norm": 0.45508888919795654, "learning_rate": 2.5171828484929447e-05, - "loss": 0.9485, + "loss": 0.8684, "step": 587 }, { "epoch": 2.380566801619433, - "grad_norm": 0.5267912352825795, + "grad_norm": 0.6382133809157001, "learning_rate": 2.5117177151798934e-05, - "loss": 0.993, + "loss": 0.9206, "step": 588 }, { "epoch": 2.3846153846153846, - "grad_norm": 0.5571714347366336, + "grad_norm": 0.4584692151391565, "learning_rate": 2.506248490185627e-05, - "loss": 0.9448, + "loss": 0.8728, "step": 589 }, { "epoch": 2.388663967611336, - "grad_norm": 0.49129145290226056, + "grad_norm": 0.555449182725766, "learning_rate": 2.500775217241923e-05, - "loss": 0.9731, + "loss": 0.8999, "step": 590 }, { "epoch": 2.3927125506072873, - "grad_norm": 0.4715482106688902, + "grad_norm": 0.5931415494507466, "learning_rate": 2.4952979401129266e-05, - "loss": 0.9819, + "loss": 0.9235, "step": 591 }, { "epoch": 2.396761133603239, - "grad_norm": 0.4808358038351141, + "grad_norm": 0.5156452983684486, "learning_rate": 2.4898167025948e-05, - "loss": 0.9869, + "loss": 0.894, "step": 592 }, { "epoch": 2.4008097165991904, - "grad_norm": 0.4445798099861959, + "grad_norm": 0.5542695674453554, "learning_rate": 2.4843315485153703e-05, - "loss": 0.9888, + "loss": 0.9228, "step": 593 }, { "epoch": 2.4048582995951415, - "grad_norm": 0.4370701166884317, + "grad_norm": 0.5592935887201808, "learning_rate": 2.4788425217337856e-05, - "loss": 0.9972, + "loss": 0.9271, "step": 594 }, { "epoch": 2.408906882591093, - "grad_norm": 0.4382858467984727, + "grad_norm": 0.4592046390358905, "learning_rate": 2.4733496661401572e-05, - "loss": 0.9521, + "loss": 0.8706, "step": 595 }, { "epoch": 2.4129554655870447, - "grad_norm": 0.4817474095341445, + "grad_norm": 0.612104858986405, "learning_rate": 2.4678530256552114e-05, - "loss": 0.9667, + "loss": 0.8973, "step": 596 }, { "epoch": 2.417004048582996, - "grad_norm": 0.45777526776126787, + "grad_norm": 0.4520235421079003, "learning_rate": 2.46235264422994e-05, - "loss": 0.9726, + "loss": 0.9031, "step": 597 }, { "epoch": 2.4210526315789473, - "grad_norm": 0.4684300244928374, + "grad_norm": 0.7155454118630583, "learning_rate": 2.4568485658452452e-05, - "loss": 0.9677, + "loss": 0.8884, "step": 598 }, { "epoch": 2.425101214574899, - "grad_norm": 0.449981239802626, + "grad_norm": 0.49210330017397336, "learning_rate": 2.451340834511592e-05, - "loss": 0.9732, + "loss": 0.8924, "step": 599 }, { "epoch": 2.42914979757085, - "grad_norm": 0.4377709062325774, + "grad_norm": 0.5448568945759027, "learning_rate": 2.4458294942686528e-05, - "loss": 0.9753, + "loss": 0.9045, "step": 600 }, { "epoch": 2.4331983805668016, - "grad_norm": 0.3978729899452302, + "grad_norm": 0.5446832502215982, "learning_rate": 2.440314589184958e-05, - "loss": 0.9544, + "loss": 0.8897, "step": 601 }, { "epoch": 2.437246963562753, - "grad_norm": 0.4532963349975613, + "grad_norm": 0.5151084848255604, "learning_rate": 2.434796163357541e-05, - "loss": 0.9787, + "loss": 0.9034, "step": 602 }, { "epoch": 2.4412955465587043, - "grad_norm": 0.40826086270123063, + "grad_norm": 0.5071646079409017, "learning_rate": 2.429274260911588e-05, - "loss": 0.9343, + "loss": 0.8736, "step": 603 }, { "epoch": 2.445344129554656, - "grad_norm": 0.5027866810198305, + "grad_norm": 0.4153205724487963, "learning_rate": 2.423748926000084e-05, - "loss": 0.9927, + "loss": 0.9193, "step": 604 }, { "epoch": 2.4493927125506074, - "grad_norm": 0.4843394423666105, + "grad_norm": 0.5643934816524688, "learning_rate": 2.418220202803459e-05, - "loss": 0.9744, + "loss": 0.9044, "step": 605 }, { "epoch": 2.4534412955465585, - "grad_norm": 0.5083218433083703, + "grad_norm": 0.4453477343034834, "learning_rate": 2.412688135529237e-05, - "loss": 0.9457, + "loss": 0.8801, "step": 606 }, { "epoch": 2.45748987854251, - "grad_norm": 0.4307537639166097, + "grad_norm": 0.5254713380086502, "learning_rate": 2.4071527684116807e-05, - "loss": 0.9797, + "loss": 0.8946, "step": 607 }, { "epoch": 2.4615384615384617, - "grad_norm": 0.500777458165871, + "grad_norm": 0.5756756572136466, "learning_rate": 2.4016141457114372e-05, - "loss": 0.9625, + "loss": 0.8908, "step": 608 }, { "epoch": 2.465587044534413, - "grad_norm": 0.46727975443101144, + "grad_norm": 0.475744077204086, "learning_rate": 2.396072311715187e-05, - "loss": 0.9848, + "loss": 0.8919, "step": 609 }, { "epoch": 2.4696356275303644, - "grad_norm": 0.47547727914849186, + "grad_norm": 0.47036184839161815, "learning_rate": 2.3905273107352863e-05, - "loss": 0.951, + "loss": 0.8693, "step": 610 }, { "epoch": 2.473684210526316, - "grad_norm": 0.47639632733293796, + "grad_norm": 0.5653766536299526, "learning_rate": 2.3849791871094157e-05, - "loss": 0.9542, + "loss": 0.894, "step": 611 }, { "epoch": 2.477732793522267, - "grad_norm": 0.49905223894688044, + "grad_norm": 0.4596720553163141, "learning_rate": 2.379427985200223e-05, - "loss": 0.9902, + "loss": 0.9294, "step": 612 }, { "epoch": 2.4817813765182186, - "grad_norm": 0.5095549441197701, + "grad_norm": 0.5245633732390688, "learning_rate": 2.373873749394972e-05, - "loss": 0.9639, + "loss": 0.8909, "step": 613 }, { "epoch": 2.48582995951417, - "grad_norm": 0.5649581122884083, + "grad_norm": 0.5588293368639196, "learning_rate": 2.3683165241051836e-05, - "loss": 0.9735, + "loss": 0.9186, "step": 614 }, { "epoch": 2.4898785425101213, - "grad_norm": 0.4955982101625742, + "grad_norm": 0.592290789346531, "learning_rate": 2.3627563537662823e-05, - "loss": 0.9703, + "loss": 0.9088, "step": 615 }, { "epoch": 2.493927125506073, - "grad_norm": 0.5301901960919295, + "grad_norm": 0.49467167011139884, "learning_rate": 2.3571932828372433e-05, - "loss": 0.9718, + "loss": 0.9051, "step": 616 }, { "epoch": 2.4979757085020244, - "grad_norm": 0.5628622862082505, + "grad_norm": 0.4989609049589017, "learning_rate": 2.351627355800233e-05, - "loss": 0.9635, + "loss": 0.8814, "step": 617 }, { "epoch": 2.5020242914979756, - "grad_norm": 0.46052380534066545, + "grad_norm": 0.6524692849978467, "learning_rate": 2.346058617160256e-05, - "loss": 0.95, + "loss": 0.8777, "step": 618 }, { "epoch": 2.506072874493927, - "grad_norm": 0.5842296581050297, + "grad_norm": 0.47958387347659054, "learning_rate": 2.3404871114447976e-05, - "loss": 1.0067, + "loss": 0.9222, "step": 619 }, { "epoch": 2.5101214574898787, - "grad_norm": 0.5206205854467717, + "grad_norm": 0.4361785793350719, "learning_rate": 2.334912883203469e-05, - "loss": 0.9844, + "loss": 0.9198, "step": 620 }, { "epoch": 2.51417004048583, - "grad_norm": 0.48239384846891076, + "grad_norm": 0.5657936096682605, "learning_rate": 2.3293359770076505e-05, - "loss": 0.9822, + "loss": 0.8945, "step": 621 }, { "epoch": 2.5182186234817814, - "grad_norm": 0.5232217148779899, + "grad_norm": 0.4677607325115319, "learning_rate": 2.323756437450136e-05, - "loss": 0.9474, + "loss": 0.8969, "step": 622 }, { "epoch": 2.522267206477733, - "grad_norm": 0.5424834598500746, + "grad_norm": 0.40382312064102766, "learning_rate": 2.3181743091447754e-05, - "loss": 0.9574, + "loss": 0.8946, "step": 623 }, { "epoch": 2.526315789473684, - "grad_norm": 0.5725661411628233, + "grad_norm": 0.6205971958728497, "learning_rate": 2.3125896367261164e-05, - "loss": 0.9733, + "loss": 0.9018, "step": 624 }, { "epoch": 2.5303643724696356, - "grad_norm": 0.4536456712924967, + "grad_norm": 0.8422689988542695, "learning_rate": 2.307002464849052e-05, - "loss": 0.9497, + "loss": 0.8854, "step": 625 }, { "epoch": 2.534412955465587, - "grad_norm": 0.5830372723662819, + "grad_norm": 0.45607740597667956, "learning_rate": 2.30141283818846e-05, - "loss": 0.9454, + "loss": 0.8708, "step": 626 }, { "epoch": 2.5384615384615383, - "grad_norm": 0.4444857185173483, + "grad_norm": 0.8676389144518011, "learning_rate": 2.2958208014388462e-05, - "loss": 0.9845, + "loss": 0.9072, "step": 627 }, { "epoch": 2.54251012145749, - "grad_norm": 0.5268357758246578, + "grad_norm": 0.8756617695091778, "learning_rate": 2.2902263993139857e-05, - "loss": 0.9952, + "loss": 0.9323, "step": 628 }, { "epoch": 2.5465587044534415, - "grad_norm": 0.547874182917779, + "grad_norm": 0.5210293951116022, "learning_rate": 2.2846296765465708e-05, - "loss": 0.9709, + "loss": 0.9035, "step": 629 }, { "epoch": 2.5506072874493926, - "grad_norm": 0.41991377659396784, + "grad_norm": 0.5994601974896789, "learning_rate": 2.2790306778878465e-05, - "loss": 0.9756, + "loss": 0.9083, "step": 630 }, { "epoch": 2.554655870445344, - "grad_norm": 0.5218163714543066, + "grad_norm": 0.6494441551820401, "learning_rate": 2.273429448107257e-05, - "loss": 0.9522, + "loss": 0.8701, "step": 631 }, { "epoch": 2.5587044534412957, - "grad_norm": 0.4732650269269637, + "grad_norm": 0.732636015089332, "learning_rate": 2.2678260319920868e-05, - "loss": 0.9537, + "loss": 0.8881, "step": 632 }, { "epoch": 2.562753036437247, - "grad_norm": 0.4447719709486366, + "grad_norm": 0.4314417072838478, "learning_rate": 2.262220474347102e-05, - "loss": 0.9836, + "loss": 0.9111, "step": 633 }, { "epoch": 2.5668016194331984, - "grad_norm": 0.5680431995344243, + "grad_norm": 0.6640286001623662, "learning_rate": 2.2566128199941918e-05, - "loss": 0.9945, + "loss": 0.9312, "step": 634 }, { "epoch": 2.57085020242915, - "grad_norm": 0.44258324251221853, + "grad_norm": 0.6075461712144616, "learning_rate": 2.2510031137720104e-05, - "loss": 0.9593, + "loss": 0.8879, "step": 635 }, { "epoch": 2.574898785425101, - "grad_norm": 0.5079644239759481, + "grad_norm": 0.41886099146094163, "learning_rate": 2.2453914005356206e-05, - "loss": 0.9849, + "loss": 0.9101, "step": 636 }, { "epoch": 2.5789473684210527, - "grad_norm": 0.5652772967337294, + "grad_norm": 0.6034917743678824, "learning_rate": 2.2397777251561302e-05, - "loss": 0.9574, + "loss": 0.8933, "step": 637 }, { "epoch": 2.582995951417004, - "grad_norm": 0.4514680101421864, + "grad_norm": 0.6206547354153527, "learning_rate": 2.2341621325203393e-05, - "loss": 0.9574, + "loss": 0.8739, "step": 638 }, { "epoch": 2.5870445344129553, - "grad_norm": 0.5705278535840284, + "grad_norm": 0.47600500077647945, "learning_rate": 2.2285446675303764e-05, - "loss": 0.9929, + "loss": 0.9187, "step": 639 }, { "epoch": 2.591093117408907, - "grad_norm": 0.4439078488965908, + "grad_norm": 0.5099111542162451, "learning_rate": 2.2229253751033423e-05, - "loss": 0.9613, + "loss": 0.886, "step": 640 }, { "epoch": 2.5951417004048585, - "grad_norm": 0.45912692495763774, + "grad_norm": 0.4328196623243501, "learning_rate": 2.2173043001709492e-05, - "loss": 0.974, + "loss": 0.8909, "step": 641 }, { "epoch": 2.5991902834008096, - "grad_norm": 0.46394395753962625, + "grad_norm": 0.44921441875935114, "learning_rate": 2.2116814876791627e-05, - "loss": 0.9627, + "loss": 0.8882, "step": 642 }, { "epoch": 2.603238866396761, - "grad_norm": 0.4775071294887102, + "grad_norm": 0.462367346011191, "learning_rate": 2.2060569825878413e-05, - "loss": 0.9589, + "loss": 0.8775, "step": 643 }, { "epoch": 2.6072874493927127, - "grad_norm": 0.5107265535973038, + "grad_norm": 0.5690698719063819, "learning_rate": 2.2004308298703777e-05, - "loss": 0.9749, + "loss": 0.9127, "step": 644 }, { "epoch": 2.611336032388664, - "grad_norm": 0.4894507550139303, + "grad_norm": 0.395020372515611, "learning_rate": 2.1948030745133396e-05, - "loss": 0.9517, + "loss": 0.8852, "step": 645 }, { "epoch": 2.6153846153846154, - "grad_norm": 0.46902016009761877, + "grad_norm": 0.5564567244643494, "learning_rate": 2.1891737615161083e-05, - "loss": 0.9834, + "loss": 0.9047, "step": 646 }, { "epoch": 2.619433198380567, - "grad_norm": 0.43084246525391906, + "grad_norm": 0.6437907126976355, "learning_rate": 2.1835429358905218e-05, - "loss": 0.9567, + "loss": 0.8855, "step": 647 }, { "epoch": 2.623481781376518, - "grad_norm": 0.47611364454415844, + "grad_norm": 0.410596428921453, "learning_rate": 2.17791064266051e-05, - "loss": 0.9797, + "loss": 0.9111, "step": 648 }, { "epoch": 2.6275303643724697, - "grad_norm": 0.4572471854832932, + "grad_norm": 0.7135055788543385, "learning_rate": 2.172276926861741e-05, - "loss": 0.9798, + "loss": 0.8893, "step": 649 }, { "epoch": 2.6315789473684212, - "grad_norm": 0.4318919719076564, + "grad_norm": 0.6017672364078106, "learning_rate": 2.166641833541255e-05, - "loss": 0.9782, + "loss": 0.8987, "step": 650 }, { "epoch": 2.6356275303643724, - "grad_norm": 0.5102540947231776, + "grad_norm": 0.5466540341167984, "learning_rate": 2.161005407757109e-05, - "loss": 0.9441, + "loss": 0.8831, "step": 651 }, { "epoch": 2.639676113360324, - "grad_norm": 0.5179351673120615, + "grad_norm": 0.5038456724438082, "learning_rate": 2.155367694578013e-05, - "loss": 0.9825, + "loss": 0.8987, "step": 652 }, { "epoch": 2.6437246963562755, - "grad_norm": 0.4254238616365703, + "grad_norm": 0.6051079186799884, "learning_rate": 2.1497287390829712e-05, - "loss": 0.9847, + "loss": 0.901, "step": 653 }, { "epoch": 2.6477732793522266, - "grad_norm": 0.502875183534101, + "grad_norm": 0.515931116305921, "learning_rate": 2.1440885863609208e-05, - "loss": 0.9546, + "loss": 0.8903, "step": 654 }, { "epoch": 2.651821862348178, - "grad_norm": 0.49199606797876844, + "grad_norm": 0.5745813970593697, "learning_rate": 2.1384472815103737e-05, - "loss": 0.9539, + "loss": 0.8825, "step": 655 }, { "epoch": 2.6558704453441297, - "grad_norm": 0.45901085433726746, + "grad_norm": 0.7735589853976339, "learning_rate": 2.1328048696390534e-05, - "loss": 0.9827, + "loss": 0.9117, "step": 656 }, { "epoch": 2.659919028340081, - "grad_norm": 0.48906783530124676, + "grad_norm": 0.4062293777822594, "learning_rate": 2.1271613958635342e-05, - "loss": 0.9707, + "loss": 0.8786, "step": 657 }, { "epoch": 2.6639676113360324, - "grad_norm": 0.4893848834151834, + "grad_norm": 0.8276513721218853, "learning_rate": 2.1215169053088825e-05, - "loss": 0.9792, + "loss": 0.88, "step": 658 }, { "epoch": 2.668016194331984, - "grad_norm": 0.4141997898228611, + "grad_norm": 0.540974413849235, "learning_rate": 2.1158714431082937e-05, - "loss": 0.9528, + "loss": 0.8859, "step": 659 }, { "epoch": 2.672064777327935, - "grad_norm": 0.5578434967348158, + "grad_norm": 0.5788509353843393, "learning_rate": 2.110225054402733e-05, - "loss": 0.9998, + "loss": 0.9332, "step": 660 }, { "epoch": 2.6761133603238867, - "grad_norm": 0.5266613127469756, + "grad_norm": 0.6764058301424899, "learning_rate": 2.1045777843405747e-05, - "loss": 1.0031, + "loss": 0.9229, "step": 661 }, { "epoch": 2.6801619433198383, - "grad_norm": 0.5876628248686028, + "grad_norm": 0.5504851368657753, "learning_rate": 2.09892967807724e-05, - "loss": 0.986, + "loss": 0.9093, "step": 662 }, { "epoch": 2.6842105263157894, - "grad_norm": 0.5415695610588738, + "grad_norm": 0.44543750051064135, "learning_rate": 2.093280780774835e-05, - "loss": 0.949, + "loss": 0.8918, "step": 663 }, { "epoch": 2.688259109311741, - "grad_norm": 0.5040092215957862, + "grad_norm": 0.6911358920800158, "learning_rate": 2.087631137601793e-05, - "loss": 0.9928, + "loss": 0.9064, "step": 664 }, { "epoch": 2.6923076923076925, - "grad_norm": 0.5341911456267248, + "grad_norm": 0.4902429967047728, "learning_rate": 2.0819807937325104e-05, - "loss": 0.9823, + "loss": 0.9128, "step": 665 }, { "epoch": 2.6963562753036436, - "grad_norm": 0.472494791659636, + "grad_norm": 0.6176888419169221, "learning_rate": 2.0763297943469847e-05, - "loss": 0.9718, + "loss": 0.8904, "step": 666 }, { "epoch": 2.700404858299595, - "grad_norm": 0.5621209421149959, + "grad_norm": 0.7262953364111677, "learning_rate": 2.070678184630458e-05, - "loss": 0.9745, + "loss": 0.8905, "step": 667 }, { "epoch": 2.7044534412955468, - "grad_norm": 0.45144787715087253, + "grad_norm": 0.43968987914990504, "learning_rate": 2.06502600977305e-05, - "loss": 0.9669, + "loss": 0.9057, "step": 668 }, { "epoch": 2.708502024291498, - "grad_norm": 0.440711059231762, + "grad_norm": 0.563394014293499, "learning_rate": 2.0593733149694008e-05, - "loss": 0.9487, + "loss": 0.8656, "step": 669 }, { "epoch": 2.7125506072874495, - "grad_norm": 0.6234311839295568, + "grad_norm": 0.5707886644945209, "learning_rate": 2.0537201454183074e-05, - "loss": 0.9553, + "loss": 0.8884, "step": 670 }, { "epoch": 2.716599190283401, - "grad_norm": 0.41083729822661574, + "grad_norm": 0.5067863015617026, "learning_rate": 2.0480665463223626e-05, - "loss": 0.9799, + "loss": 0.8882, "step": 671 }, { "epoch": 2.720647773279352, - "grad_norm": 0.5806381796554544, + "grad_norm": 0.5806235312531612, "learning_rate": 2.042412562887594e-05, - "loss": 0.9681, + "loss": 0.9128, "step": 672 }, { "epoch": 2.7246963562753037, - "grad_norm": 0.46219458592226176, + "grad_norm": 0.6860228138463896, "learning_rate": 2.0367582403231033e-05, - "loss": 0.9782, + "loss": 0.8987, "step": 673 }, { "epoch": 2.7287449392712553, - "grad_norm": 0.48631435076436913, + "grad_norm": 0.4425602008931536, "learning_rate": 2.0311036238407023e-05, - "loss": 0.9584, + "loss": 0.8973, "step": 674 }, { "epoch": 2.7327935222672064, - "grad_norm": 0.5636190328202831, + "grad_norm": 0.8594880201773468, "learning_rate": 2.0254487586545534e-05, - "loss": 0.9909, + "loss": 0.8938, "step": 675 }, { "epoch": 2.736842105263158, - "grad_norm": 0.444260466507559, + "grad_norm": 0.554999276014701, "learning_rate": 2.0197936899808088e-05, - "loss": 0.9788, + "loss": 0.8991, "step": 676 }, { "epoch": 2.7408906882591095, - "grad_norm": 0.4838284287877657, + "grad_norm": 0.5332457169516268, "learning_rate": 2.0141384630372466e-05, - "loss": 0.96, + "loss": 0.8955, "step": 677 }, { "epoch": 2.7449392712550607, - "grad_norm": 0.42367715654051297, + "grad_norm": 0.4825189791716621, "learning_rate": 2.0084831230429106e-05, - "loss": 0.9491, + "loss": 0.8839, "step": 678 }, { "epoch": 2.748987854251012, - "grad_norm": 0.4642628894608451, + "grad_norm": 0.5271056354712564, "learning_rate": 2.0028277152177492e-05, - "loss": 0.9664, + "loss": 0.8988, "step": 679 }, { "epoch": 2.753036437246964, - "grad_norm": 0.4514033821586843, + "grad_norm": 0.5020554683050802, "learning_rate": 1.9971722847822518e-05, - "loss": 0.9519, + "loss": 0.8879, "step": 680 }, { "epoch": 2.757085020242915, - "grad_norm": 0.49643839057896827, + "grad_norm": 0.5497499478226384, "learning_rate": 1.99151687695709e-05, - "loss": 0.9489, + "loss": 0.8912, "step": 681 }, { "epoch": 2.7611336032388665, - "grad_norm": 0.41822014049560435, + "grad_norm": 0.45075912220362685, "learning_rate": 1.9858615369627537e-05, - "loss": 0.9844, + "loss": 0.8967, "step": 682 }, { "epoch": 2.765182186234818, - "grad_norm": 0.5305555009796895, + "grad_norm": 0.5111168269264604, "learning_rate": 1.9802063100191916e-05, - "loss": 0.9691, + "loss": 0.9103, "step": 683 }, { "epoch": 2.769230769230769, - "grad_norm": 0.39832118395611943, + "grad_norm": 0.5335171552086255, "learning_rate": 1.974551241345447e-05, - "loss": 0.9865, + "loss": 0.8868, "step": 684 }, { "epoch": 2.7732793522267207, - "grad_norm": 0.43726121597986894, + "grad_norm": 0.41700890203825935, "learning_rate": 1.9688963761592984e-05, - "loss": 0.9648, + "loss": 0.8893, "step": 685 }, { "epoch": 2.7773279352226723, - "grad_norm": 0.4208836168049033, + "grad_norm": 0.4904031264113896, "learning_rate": 1.963241759676897e-05, - "loss": 0.9602, + "loss": 0.8958, "step": 686 }, { "epoch": 2.7813765182186234, - "grad_norm": 0.47563795061459174, + "grad_norm": 0.45481178199408456, "learning_rate": 1.9575874371124066e-05, - "loss": 0.946, + "loss": 0.8854, "step": 687 }, { "epoch": 2.785425101214575, - "grad_norm": 0.43019403015922575, + "grad_norm": 0.447086093770594, "learning_rate": 1.951933453677638e-05, - "loss": 0.9651, + "loss": 0.8942, "step": 688 }, { "epoch": 2.7894736842105265, - "grad_norm": 0.41874904969173404, + "grad_norm": 0.4685885986690239, "learning_rate": 1.9462798545816932e-05, - "loss": 0.9579, + "loss": 0.8743, "step": 689 }, { "epoch": 2.7935222672064777, - "grad_norm": 0.5040271115878224, + "grad_norm": 0.4444656737634961, "learning_rate": 1.9406266850305996e-05, - "loss": 0.9735, + "loss": 0.8997, "step": 690 }, { "epoch": 2.7975708502024292, - "grad_norm": 0.42776733425919794, + "grad_norm": 0.43579618367719836, "learning_rate": 1.9349739902269503e-05, - "loss": 0.9508, + "loss": 0.86, "step": 691 }, { "epoch": 2.801619433198381, - "grad_norm": 0.45256596320582276, + "grad_norm": 0.4983893832106297, "learning_rate": 1.9293218153695425e-05, - "loss": 0.9775, + "loss": 0.9079, "step": 692 }, { "epoch": 2.805668016194332, - "grad_norm": 0.48307589294510717, + "grad_norm": 0.4005263849090156, "learning_rate": 1.9236702056530153e-05, - "loss": 0.9673, + "loss": 0.8801, "step": 693 }, { "epoch": 2.8097165991902835, - "grad_norm": 0.4187545924596713, + "grad_norm": 0.463558891114383, "learning_rate": 1.9180192062674907e-05, - "loss": 0.9784, + "loss": 0.9084, "step": 694 }, { "epoch": 2.813765182186235, - "grad_norm": 0.4859341599351372, + "grad_norm": 0.48231785118719045, "learning_rate": 1.9123688623982076e-05, - "loss": 0.9876, + "loss": 0.902, "step": 695 }, { "epoch": 2.817813765182186, - "grad_norm": 0.41132810546955756, + "grad_norm": 0.4152341257638767, "learning_rate": 1.906719219225165e-05, - "loss": 0.9635, + "loss": 0.8861, "step": 696 }, { "epoch": 2.8218623481781377, - "grad_norm": 0.43295445594236603, + "grad_norm": 0.4585091066674028, "learning_rate": 1.9010703219227603e-05, - "loss": 0.9692, + "loss": 0.8891, "step": 697 }, { "epoch": 2.8259109311740893, - "grad_norm": 0.4445405349296205, + "grad_norm": 0.4924063561354388, "learning_rate": 1.8954222156594256e-05, - "loss": 0.9397, + "loss": 0.8795, "step": 698 }, { "epoch": 2.8299595141700404, - "grad_norm": 0.4163892251808106, + "grad_norm": 0.32239125406973224, "learning_rate": 1.8897749455972673e-05, - "loss": 0.977, + "loss": 0.8951, "step": 699 }, { "epoch": 2.834008097165992, - "grad_norm": 0.4505076519843084, + "grad_norm": 0.45112411637461275, "learning_rate": 1.884128556891707e-05, - "loss": 0.9865, + "loss": 0.9043, "step": 700 }, { "epoch": 2.8380566801619436, - "grad_norm": 0.4176711747574246, + "grad_norm": 0.364786838313807, "learning_rate": 1.878483094691119e-05, - "loss": 0.9741, + "loss": 0.8848, "step": 701 }, { "epoch": 2.8421052631578947, - "grad_norm": 0.4518651329377136, + "grad_norm": 0.4234031777111052, "learning_rate": 1.8728386041364664e-05, - "loss": 0.977, + "loss": 0.8969, "step": 702 }, { "epoch": 2.8461538461538463, - "grad_norm": 0.42579722539260517, + "grad_norm": 0.358278733144445, "learning_rate": 1.867195130360947e-05, - "loss": 0.9348, + "loss": 0.8651, "step": 703 }, { "epoch": 2.850202429149798, - "grad_norm": 0.4713125580015882, + "grad_norm": 0.4045959177007385, "learning_rate": 1.8615527184896263e-05, - "loss": 0.9856, + "loss": 0.9136, "step": 704 }, { "epoch": 2.854251012145749, - "grad_norm": 0.4259690527460967, + "grad_norm": 0.429989643439928, "learning_rate": 1.8559114136390795e-05, - "loss": 0.9458, + "loss": 0.8751, "step": 705 }, { "epoch": 2.8582995951417005, - "grad_norm": 0.5120331852621113, + "grad_norm": 0.39936021860907756, "learning_rate": 1.8502712609170298e-05, - "loss": 0.9713, + "loss": 0.8911, "step": 706 }, { "epoch": 2.862348178137652, - "grad_norm": 0.4415768180226851, + "grad_norm": 0.45641432483924266, "learning_rate": 1.8446323054219876e-05, - "loss": 0.95, + "loss": 0.8734, "step": 707 }, { "epoch": 2.866396761133603, - "grad_norm": 0.5364723716049101, + "grad_norm": 0.46941986524089424, "learning_rate": 1.838994592242891e-05, - "loss": 0.9884, + "loss": 0.9247, "step": 708 }, { "epoch": 2.8704453441295548, - "grad_norm": 0.45061570220167113, + "grad_norm": 0.42535765349763366, "learning_rate": 1.8333581664587453e-05, - "loss": 0.9678, + "loss": 0.8993, "step": 709 }, { "epoch": 2.8744939271255063, - "grad_norm": 0.452096581291582, + "grad_norm": 0.4662877331980058, "learning_rate": 1.8277230731382593e-05, - "loss": 0.974, + "loss": 0.8818, "step": 710 }, { "epoch": 2.8785425101214575, - "grad_norm": 0.47472107481397646, + "grad_norm": 0.42192416681962047, "learning_rate": 1.82208935733949e-05, - "loss": 0.9707, + "loss": 0.891, "step": 711 }, { "epoch": 2.882591093117409, - "grad_norm": 0.45190025754629864, + "grad_norm": 0.4622546672277531, "learning_rate": 1.8164570641094793e-05, - "loss": 0.9551, + "loss": 0.8837, "step": 712 }, { "epoch": 2.8866396761133606, - "grad_norm": 0.3950272106251044, + "grad_norm": 0.4135246086960274, "learning_rate": 1.810826238483892e-05, - "loss": 0.9614, + "loss": 0.8766, "step": 713 }, { "epoch": 2.8906882591093117, - "grad_norm": 0.43501375960046046, + "grad_norm": 0.43383049826862013, "learning_rate": 1.8051969254866608e-05, - "loss": 0.9497, + "loss": 0.8819, "step": 714 }, { "epoch": 2.8947368421052633, - "grad_norm": 0.40924785711703354, + "grad_norm": 0.47914792118832505, "learning_rate": 1.7995691701296226e-05, - "loss": 0.9386, + "loss": 0.868, "step": 715 }, { "epoch": 2.898785425101215, - "grad_norm": 0.4070057839306158, + "grad_norm": 0.5999029834334415, "learning_rate": 1.7939430174121593e-05, - "loss": 0.9642, + "loss": 0.8722, "step": 716 }, { "epoch": 2.902834008097166, - "grad_norm": 0.454010609171541, + "grad_norm": 0.3657206384059301, "learning_rate": 1.788318512320838e-05, - "loss": 0.9748, + "loss": 0.8994, "step": 717 }, { "epoch": 2.9068825910931175, - "grad_norm": 0.41860742725738104, + "grad_norm": 0.6055919447455844, "learning_rate": 1.782695699829051e-05, - "loss": 0.9598, + "loss": 0.9039, "step": 718 }, { "epoch": 2.910931174089069, - "grad_norm": 0.437787555912543, + "grad_norm": 0.39602015045781086, "learning_rate": 1.7770746248966587e-05, - "loss": 0.9383, + "loss": 0.8628, "step": 719 }, { "epoch": 2.91497975708502, - "grad_norm": 0.4416077667556647, + "grad_norm": 0.6194613938564885, "learning_rate": 1.7714553324696243e-05, - "loss": 0.9369, + "loss": 0.8726, "step": 720 }, { "epoch": 2.919028340080972, - "grad_norm": 0.47204038313831376, + "grad_norm": 0.5073979346056203, "learning_rate": 1.7658378674796614e-05, - "loss": 0.9687, + "loss": 0.8625, "step": 721 }, { "epoch": 2.9230769230769234, - "grad_norm": 0.43466464384788434, + "grad_norm": 0.43882654970695184, "learning_rate": 1.76022227484387e-05, - "loss": 0.9618, + "loss": 0.9011, "step": 722 }, { "epoch": 2.9271255060728745, - "grad_norm": 0.42797923825919676, + "grad_norm": 0.51511295551052, "learning_rate": 1.7546085994643807e-05, - "loss": 0.9712, + "loss": 0.8959, "step": 723 }, { "epoch": 2.931174089068826, - "grad_norm": 0.45941240898617086, + "grad_norm": 0.48550593587820484, "learning_rate": 1.7489968862279902e-05, - "loss": 0.9957, + "loss": 0.9217, "step": 724 }, { "epoch": 2.9352226720647776, - "grad_norm": 0.4395990729315826, + "grad_norm": 0.4546302630489191, "learning_rate": 1.743387180005809e-05, - "loss": 0.9784, + "loss": 0.8948, "step": 725 }, { "epoch": 2.9392712550607287, - "grad_norm": 0.40933809006889993, + "grad_norm": 0.5346382832708898, "learning_rate": 1.737779525652899e-05, - "loss": 0.9813, + "loss": 0.9169, "step": 726 }, { "epoch": 2.9433198380566803, - "grad_norm": 0.4251809472812531, + "grad_norm": 0.3688618196088349, "learning_rate": 1.732173968007914e-05, - "loss": 0.982, + "loss": 0.9085, "step": 727 }, { "epoch": 2.9473684210526314, - "grad_norm": 0.4372337265637708, + "grad_norm": 0.44524426922934385, "learning_rate": 1.7265705518927437e-05, - "loss": 0.9892, + "loss": 0.9075, "step": 728 }, { "epoch": 2.951417004048583, - "grad_norm": 0.4033504449949853, + "grad_norm": 0.43262684455387934, "learning_rate": 1.7209693221121542e-05, - "loss": 0.9678, + "loss": 0.8924, "step": 729 }, { "epoch": 2.9554655870445345, - "grad_norm": 0.4457515413823814, + "grad_norm": 0.42951987646945816, "learning_rate": 1.7153703234534302e-05, - "loss": 0.976, + "loss": 0.8931, "step": 730 }, { "epoch": 2.9595141700404857, - "grad_norm": 0.41469779149988295, + "grad_norm": 0.4779713312753201, "learning_rate": 1.709773600686015e-05, - "loss": 0.9922, + "loss": 0.9207, "step": 731 }, { "epoch": 2.9635627530364372, - "grad_norm": 0.41372460499895364, + "grad_norm": 0.40094919651051075, "learning_rate": 1.7041791985611548e-05, - "loss": 0.976, + "loss": 0.9082, "step": 732 }, { "epoch": 2.967611336032389, - "grad_norm": 0.4350958816312875, + "grad_norm": 0.4711212009760646, "learning_rate": 1.69858716181154e-05, - "loss": 0.9603, + "loss": 0.889, "step": 733 }, { "epoch": 2.97165991902834, - "grad_norm": 0.4167684296122814, + "grad_norm": 0.459316491198509, "learning_rate": 1.692997535150948e-05, - "loss": 0.9723, + "loss": 0.9095, "step": 734 }, { "epoch": 2.9757085020242915, - "grad_norm": 0.3756768569494513, + "grad_norm": 0.4086387100542631, "learning_rate": 1.687410363273884e-05, - "loss": 0.9616, + "loss": 0.8893, "step": 735 }, { "epoch": 2.979757085020243, - "grad_norm": 0.4079938112728359, + "grad_norm": 0.5374980868523385, "learning_rate": 1.6818256908552257e-05, - "loss": 0.9641, + "loss": 0.8987, "step": 736 }, { "epoch": 2.983805668016194, - "grad_norm": 0.4659652034336867, + "grad_norm": 0.3777606074513546, "learning_rate": 1.6762435625498646e-05, - "loss": 0.9675, + "loss": 0.8901, "step": 737 }, { "epoch": 2.9878542510121457, - "grad_norm": 0.38683246452690007, + "grad_norm": 0.5104816299258442, "learning_rate": 1.67066402299235e-05, - "loss": 0.9819, + "loss": 0.8972, "step": 738 }, { "epoch": 2.9919028340080973, - "grad_norm": 0.46454673860003715, + "grad_norm": 0.34003424297185436, "learning_rate": 1.6650871167965313e-05, - "loss": 0.989, + "loss": 0.9162, "step": 739 }, { "epoch": 2.9959514170040484, - "grad_norm": 0.391322091636126, + "grad_norm": 0.5392031732910614, "learning_rate": 1.6595128885552028e-05, - "loss": 0.9864, + "loss": 0.9169, "step": 740 }, { "epoch": 3.0, - "grad_norm": 0.40515312647951424, + "grad_norm": 0.45841378260208726, "learning_rate": 1.653941382839745e-05, - "loss": 0.9648, + "loss": 0.898, "step": 741 }, { "epoch": 3.0040485829959516, - "grad_norm": 0.7041433010478754, + "grad_norm": 0.43942287070141384, "learning_rate": 1.6483726441997673e-05, - "loss": 0.8294, + "loss": 0.8362, "step": 742 }, { "epoch": 3.0080971659919027, - "grad_norm": 0.6627729679681239, + "grad_norm": 0.4414458402971482, "learning_rate": 1.642806717162757e-05, - "loss": 0.8328, + "loss": 0.8597, "step": 743 }, { "epoch": 3.0121457489878543, - "grad_norm": 1.7322663163154086, + "grad_norm": 0.3939633804655854, "learning_rate": 1.637243646233718e-05, - "loss": 0.8513, + "loss": 0.8589, "step": 744 }, { "epoch": 3.016194331983806, - "grad_norm": 0.9407849633722015, + "grad_norm": 0.5442165017301382, "learning_rate": 1.6316834758948174e-05, - "loss": 0.8409, + "loss": 0.8657, "step": 745 }, { "epoch": 3.020242914979757, - "grad_norm": 0.7636144617576476, + "grad_norm": 0.481728003073291, "learning_rate": 1.6261262506050282e-05, - "loss": 0.8481, + "loss": 0.8632, "step": 746 }, { "epoch": 3.0242914979757085, - "grad_norm": 0.9051206465184235, + "grad_norm": 0.49390297006575257, "learning_rate": 1.620572014799777e-05, - "loss": 0.8332, + "loss": 0.8553, "step": 747 }, { "epoch": 3.02834008097166, - "grad_norm": 0.6813473655119282, + "grad_norm": 0.489712610320165, "learning_rate": 1.6150208128905857e-05, - "loss": 0.8466, + "loss": 0.8592, "step": 748 }, { "epoch": 3.032388663967611, - "grad_norm": 0.8883311343165888, + "grad_norm": 0.4279551155347087, "learning_rate": 1.6094726892647147e-05, - "loss": 0.8383, + "loss": 0.8473, "step": 749 }, { "epoch": 3.0364372469635628, - "grad_norm": 0.7650343344239662, + "grad_norm": 0.5110290019877501, "learning_rate": 1.6039276882848135e-05, - "loss": 0.8205, + "loss": 0.8333, "step": 750 }, { "epoch": 3.0404858299595143, - "grad_norm": 0.8849420516763339, + "grad_norm": 0.4088346158831528, "learning_rate": 1.5983858542885635e-05, - "loss": 0.821, + "loss": 0.8387, "step": 751 }, { "epoch": 3.0445344129554655, - "grad_norm": 0.6960480102940518, + "grad_norm": 0.48515424976757465, "learning_rate": 1.5928472315883203e-05, - "loss": 0.8287, + "loss": 0.8623, "step": 752 }, { "epoch": 3.048582995951417, - "grad_norm": 0.9557784493320454, + "grad_norm": 0.4469458486534264, "learning_rate": 1.5873118644707633e-05, - "loss": 0.813, + "loss": 0.8263, "step": 753 }, { "epoch": 3.0526315789473686, - "grad_norm": 0.7536529881332438, + "grad_norm": 0.4630707545575395, "learning_rate": 1.5817797971965413e-05, - "loss": 0.8366, + "loss": 0.8736, "step": 754 }, { "epoch": 3.0566801619433197, - "grad_norm": 0.7195953633193024, + "grad_norm": 0.5061394467218068, "learning_rate": 1.576251073999917e-05, - "loss": 0.8229, + "loss": 0.8535, "step": 755 }, { "epoch": 3.0607287449392713, - "grad_norm": 0.8555547267972244, + "grad_norm": 0.39291781585002045, "learning_rate": 1.5707257390884126e-05, - "loss": 0.8388, + "loss": 0.8535, "step": 756 }, { "epoch": 3.064777327935223, - "grad_norm": 0.6565398660406285, + "grad_norm": 0.45009484602824495, "learning_rate": 1.5652038366424595e-05, - "loss": 0.816, + "loss": 0.8595, "step": 757 }, { "epoch": 3.068825910931174, - "grad_norm": 0.7152434440283757, + "grad_norm": 0.35716691295596525, "learning_rate": 1.5596854108150424e-05, - "loss": 0.8473, + "loss": 0.8778, "step": 758 }, { "epoch": 3.0728744939271255, - "grad_norm": 0.5889650387471189, + "grad_norm": 0.3634042426065204, "learning_rate": 1.5541705057313476e-05, - "loss": 0.8321, + "loss": 0.8633, "step": 759 }, { "epoch": 3.076923076923077, - "grad_norm": 0.6337862606562632, + "grad_norm": 0.3729778165664454, "learning_rate": 1.5486591654884086e-05, - "loss": 0.8108, + "loss": 0.8498, "step": 760 }, { "epoch": 3.080971659919028, - "grad_norm": 0.5516994114411624, + "grad_norm": 0.40282008845053724, "learning_rate": 1.543151434154755e-05, - "loss": 0.8287, + "loss": 0.844, "step": 761 }, { "epoch": 3.08502024291498, - "grad_norm": 0.538390011537654, + "grad_norm": 0.3563825529087447, "learning_rate": 1.5376473557700615e-05, - "loss": 0.8071, + "loss": 0.8276, "step": 762 }, { "epoch": 3.0890688259109313, - "grad_norm": 0.5431450325937882, + "grad_norm": 0.42598109207086077, "learning_rate": 1.5321469743447893e-05, - "loss": 0.8509, + "loss": 0.8921, "step": 763 }, { "epoch": 3.0931174089068825, - "grad_norm": 0.5137472459186407, + "grad_norm": 0.37834016724187186, "learning_rate": 1.5266503338598434e-05, - "loss": 0.816, + "loss": 0.8274, "step": 764 }, { "epoch": 3.097165991902834, - "grad_norm": 0.5415346369339564, + "grad_norm": 0.35907540868603405, "learning_rate": 1.5211574782662149e-05, - "loss": 0.817, + "loss": 0.8303, "step": 765 }, { "epoch": 3.1012145748987856, - "grad_norm": 0.5156542697759232, + "grad_norm": 0.39093883258748297, "learning_rate": 1.5156684514846307e-05, - "loss": 0.8446, + "loss": 0.8623, "step": 766 }, { "epoch": 3.1052631578947367, - "grad_norm": 0.47865528225575704, + "grad_norm": 0.3446103954350152, "learning_rate": 1.5101832974052012e-05, - "loss": 0.8223, + "loss": 0.849, "step": 767 }, { "epoch": 3.1093117408906883, - "grad_norm": 0.4391584653188042, + "grad_norm": 0.3184969046506509, "learning_rate": 1.5047020598870737e-05, - "loss": 0.853, + "loss": 0.8664, "step": 768 }, { "epoch": 3.11336032388664, - "grad_norm": 0.4839213357510332, + "grad_norm": 0.34313848600088226, "learning_rate": 1.4992247827580778e-05, - "loss": 0.8012, + "loss": 0.8248, "step": 769 }, { "epoch": 3.117408906882591, - "grad_norm": 0.4687472373158056, + "grad_norm": 0.3589200225176272, "learning_rate": 1.493751509814374e-05, - "loss": 0.8173, + "loss": 0.839, "step": 770 }, { "epoch": 3.1214574898785425, - "grad_norm": 0.4665740856625211, + "grad_norm": 0.31583633549179596, "learning_rate": 1.4882822848201073e-05, - "loss": 0.7952, + "loss": 0.834, "step": 771 }, { "epoch": 3.125506072874494, - "grad_norm": 0.47727380852483714, + "grad_norm": 0.35042133639634715, "learning_rate": 1.4828171515070553e-05, - "loss": 0.8238, + "loss": 0.8617, "step": 772 }, { "epoch": 3.1295546558704452, - "grad_norm": 0.4850289455166604, + "grad_norm": 0.3313361645281366, "learning_rate": 1.4773561535742793e-05, - "loss": 0.8417, + "loss": 0.8602, "step": 773 }, { "epoch": 3.133603238866397, - "grad_norm": 0.49601984964432777, + "grad_norm": 0.3171088510601409, "learning_rate": 1.4718993346877715e-05, - "loss": 0.8177, + "loss": 0.8534, "step": 774 }, { "epoch": 3.1376518218623484, - "grad_norm": 0.4896701471260314, + "grad_norm": 0.40348913030766603, "learning_rate": 1.466446738480111e-05, - "loss": 0.8416, + "loss": 0.8382, "step": 775 }, { "epoch": 3.1417004048582995, - "grad_norm": 0.4458506959080997, + "grad_norm": 0.36035245574027003, "learning_rate": 1.460998408550112e-05, - "loss": 0.8282, + "loss": 0.8527, "step": 776 }, { "epoch": 3.145748987854251, - "grad_norm": 0.4675504160898783, + "grad_norm": 0.3523564418914791, "learning_rate": 1.4555543884624751e-05, - "loss": 0.8429, + "loss": 0.8566, "step": 777 }, { "epoch": 3.1497975708502026, - "grad_norm": 0.48200592545287224, + "grad_norm": 0.3700512309933563, "learning_rate": 1.4501147217474402e-05, - "loss": 0.8305, + "loss": 0.8505, "step": 778 }, { "epoch": 3.1538461538461537, - "grad_norm": 0.49765050254091847, + "grad_norm": 0.35158719104044883, "learning_rate": 1.444679451900437e-05, - "loss": 0.8412, + "loss": 0.8655, "step": 779 }, { "epoch": 3.1578947368421053, - "grad_norm": 0.4882876154279006, + "grad_norm": 0.3871512871330745, "learning_rate": 1.4392486223817397e-05, - "loss": 0.8343, + "loss": 0.8594, "step": 780 }, { "epoch": 3.161943319838057, - "grad_norm": 0.41472569292358474, + "grad_norm": 0.3598870493606884, "learning_rate": 1.4338222766161143e-05, - "loss": 0.8422, + "loss": 0.8548, "step": 781 }, { "epoch": 3.165991902834008, - "grad_norm": 0.4632563384103101, + "grad_norm": 0.43216340027673644, "learning_rate": 1.4284004579924777e-05, - "loss": 0.8347, + "loss": 0.8561, "step": 782 }, { "epoch": 3.1700404858299596, - "grad_norm": 0.47644883193423004, + "grad_norm": 0.412365640155345, "learning_rate": 1.4229832098635479e-05, - "loss": 0.8291, + "loss": 0.8268, "step": 783 }, { "epoch": 3.174089068825911, - "grad_norm": 0.45277727238177007, + "grad_norm": 0.3449630908546609, "learning_rate": 1.4175705755454963e-05, - "loss": 0.8369, + "loss": 0.8381, "step": 784 }, { "epoch": 3.1781376518218623, - "grad_norm": 0.44249539023747614, + "grad_norm": 0.3683690069237281, "learning_rate": 1.412162598317602e-05, - "loss": 0.837, + "loss": 0.8543, "step": 785 }, { "epoch": 3.182186234817814, - "grad_norm": 0.4360145824080792, + "grad_norm": 0.35984222878915206, "learning_rate": 1.406759321421907e-05, - "loss": 0.8179, + "loss": 0.8466, "step": 786 }, { "epoch": 3.1862348178137654, - "grad_norm": 0.43819567768657774, + "grad_norm": 0.40674383209226866, "learning_rate": 1.401360788062871e-05, - "loss": 0.8383, + "loss": 0.8555, "step": 787 }, { "epoch": 3.1902834008097165, - "grad_norm": 0.44538573202139786, + "grad_norm": 0.390278310167857, "learning_rate": 1.3959670414070204e-05, - "loss": 0.8385, + "loss": 0.8618, "step": 788 }, { "epoch": 3.194331983805668, - "grad_norm": 0.4333540804091088, + "grad_norm": 0.3486457494871084, "learning_rate": 1.3905781245826108e-05, - "loss": 0.8401, + "loss": 0.8436, "step": 789 }, { "epoch": 3.1983805668016196, - "grad_norm": 0.4403212985115629, + "grad_norm": 0.3780404656796387, "learning_rate": 1.3851940806792778e-05, - "loss": 0.8233, + "loss": 0.853, "step": 790 }, { "epoch": 3.2024291497975708, - "grad_norm": 0.43851450183447094, + "grad_norm": 0.3618752525495607, "learning_rate": 1.379814952747693e-05, - "loss": 0.8226, + "loss": 0.8492, "step": 791 }, { "epoch": 3.2064777327935223, - "grad_norm": 0.4854051808315509, + "grad_norm": 0.30952243460037066, "learning_rate": 1.3744407837992193e-05, - "loss": 0.838, + "loss": 0.8522, "step": 792 }, { "epoch": 3.2105263157894735, - "grad_norm": 0.4342827172171298, + "grad_norm": 0.31617724913668216, "learning_rate": 1.3690716168055692e-05, - "loss": 0.8338, + "loss": 0.8628, "step": 793 }, { "epoch": 3.214574898785425, - "grad_norm": 0.4589398953358267, + "grad_norm": 0.3318912244697494, "learning_rate": 1.363707494698459e-05, - "loss": 0.829, + "loss": 0.8582, "step": 794 }, { "epoch": 3.2186234817813766, - "grad_norm": 0.5021540585878653, + "grad_norm": 0.3552690997516602, "learning_rate": 1.358348460369265e-05, - "loss": 0.8315, + "loss": 0.8595, "step": 795 }, { "epoch": 3.2226720647773277, - "grad_norm": 0.4462057492880543, + "grad_norm": 0.3507215208981739, "learning_rate": 1.352994556668684e-05, - "loss": 0.8308, + "loss": 0.8401, "step": 796 }, { "epoch": 3.2267206477732793, - "grad_norm": 0.417330630017805, + "grad_norm": 0.3470274806922485, "learning_rate": 1.3476458264063865e-05, - "loss": 0.8222, + "loss": 0.8321, "step": 797 }, { "epoch": 3.230769230769231, - "grad_norm": 0.45930589556156487, + "grad_norm": 0.31669195812339457, "learning_rate": 1.342302312350678e-05, - "loss": 0.8219, + "loss": 0.8381, "step": 798 }, { "epoch": 3.234817813765182, - "grad_norm": 0.4324787594813373, + "grad_norm": 0.32830085900158806, "learning_rate": 1.3369640572281537e-05, - "loss": 0.8143, + "loss": 0.8413, "step": 799 }, { "epoch": 3.2388663967611335, - "grad_norm": 0.41306924137294215, + "grad_norm": 0.3667660636452027, "learning_rate": 1.3316311037233596e-05, - "loss": 0.8081, + "loss": 0.8423, "step": 800 }, { "epoch": 3.242914979757085, - "grad_norm": 0.4342630030140212, + "grad_norm": 0.334893512511343, "learning_rate": 1.326303494478451e-05, - "loss": 0.8637, + "loss": 0.9062, "step": 801 }, { "epoch": 3.246963562753036, - "grad_norm": 0.4250872939060123, + "grad_norm": 0.3409707749924025, "learning_rate": 1.3209812720928472e-05, - "loss": 0.833, + "loss": 0.8533, "step": 802 }, { "epoch": 3.251012145748988, - "grad_norm": 0.42084897270289223, + "grad_norm": 0.3336366418527692, "learning_rate": 1.315664479122898e-05, - "loss": 0.8363, + "loss": 0.8686, "step": 803 }, { "epoch": 3.2550607287449393, - "grad_norm": 0.3978327659066298, + "grad_norm": 0.3506095700526711, "learning_rate": 1.3103531580815378e-05, - "loss": 0.8071, + "loss": 0.8379, "step": 804 }, { "epoch": 3.2591093117408905, - "grad_norm": 0.46694387193091896, + "grad_norm": 0.3091493178362056, "learning_rate": 1.3050473514379488e-05, - "loss": 0.8406, + "loss": 0.8506, "step": 805 }, { "epoch": 3.263157894736842, - "grad_norm": 0.43358767383324903, + "grad_norm": 0.35125585423677, "learning_rate": 1.2997471016172188e-05, - "loss": 0.8558, + "loss": 0.8784, "step": 806 }, { "epoch": 3.2672064777327936, - "grad_norm": 0.44745159579878774, + "grad_norm": 0.30915112036018955, "learning_rate": 1.2944524510000042e-05, - "loss": 0.8197, + "loss": 0.8495, "step": 807 }, { "epoch": 3.2712550607287447, - "grad_norm": 0.43274516280205827, + "grad_norm": 0.28572684724626396, "learning_rate": 1.2891634419221908e-05, - "loss": 0.8285, + "loss": 0.845, "step": 808 }, { "epoch": 3.2753036437246963, - "grad_norm": 0.4381627444773244, + "grad_norm": 0.34933084406521386, "learning_rate": 1.2838801166745545e-05, - "loss": 0.8176, + "loss": 0.8392, "step": 809 }, { "epoch": 3.279352226720648, - "grad_norm": 0.4020707669706982, + "grad_norm": 0.3007285106454552, "learning_rate": 1.2786025175024216e-05, - "loss": 0.8187, + "loss": 0.8343, "step": 810 }, { "epoch": 3.283400809716599, - "grad_norm": 0.4597564240436003, + "grad_norm": 0.40207973129407804, "learning_rate": 1.2733306866053357e-05, - "loss": 0.8579, + "loss": 0.8958, "step": 811 }, { "epoch": 3.2874493927125505, - "grad_norm": 0.4465748757559242, + "grad_norm": 0.3385314156616887, "learning_rate": 1.2680646661367163e-05, - "loss": 0.8532, + "loss": 0.8647, "step": 812 }, { "epoch": 3.291497975708502, - "grad_norm": 0.41887579971600664, + "grad_norm": 0.3649652047229675, "learning_rate": 1.2628044982035219e-05, - "loss": 0.816, + "loss": 0.8392, "step": 813 }, { "epoch": 3.2955465587044532, - "grad_norm": 0.42514579239307676, + "grad_norm": 0.41535310975054235, "learning_rate": 1.2575502248659155e-05, - "loss": 0.8327, + "loss": 0.8606, "step": 814 }, { "epoch": 3.299595141700405, - "grad_norm": 0.45326360951646294, + "grad_norm": 0.3623574542844954, "learning_rate": 1.2523018881369269e-05, - "loss": 0.8196, + "loss": 0.8406, "step": 815 }, { "epoch": 3.3036437246963564, - "grad_norm": 0.46226702594997043, + "grad_norm": 0.35147287448961984, "learning_rate": 1.2470595299821176e-05, - "loss": 0.8437, + "loss": 0.861, "step": 816 }, { "epoch": 3.3076923076923075, - "grad_norm": 0.4318014694413498, + "grad_norm": 0.419539134223765, "learning_rate": 1.2418231923192416e-05, - "loss": 0.8439, + "loss": 0.8579, "step": 817 }, { "epoch": 3.311740890688259, - "grad_norm": 0.43432111496796794, + "grad_norm": 0.3224467150867655, "learning_rate": 1.2365929170179165e-05, - "loss": 0.8297, + "loss": 0.8469, "step": 818 }, { "epoch": 3.3157894736842106, - "grad_norm": 0.4863391655764384, + "grad_norm": 0.35259268964677165, "learning_rate": 1.2313687458992848e-05, - "loss": 0.8156, + "loss": 0.8435, "step": 819 }, { "epoch": 3.3198380566801617, - "grad_norm": 0.47004797674813714, + "grad_norm": 0.3552449377233297, "learning_rate": 1.2261507207356785e-05, - "loss": 0.8115, + "loss": 0.8316, "step": 820 }, { "epoch": 3.3238866396761133, - "grad_norm": 0.42537971981758915, + "grad_norm": 0.3344985621292718, "learning_rate": 1.220938883250288e-05, - "loss": 0.834, + "loss": 0.8678, "step": 821 }, { "epoch": 3.327935222672065, - "grad_norm": 0.4353096414206896, + "grad_norm": 0.3332835371120531, "learning_rate": 1.215733275116827e-05, - "loss": 0.8002, + "loss": 0.8311, "step": 822 }, { "epoch": 3.331983805668016, - "grad_norm": 0.5178407166971657, + "grad_norm": 0.2998027848390441, "learning_rate": 1.2105339379592005e-05, - "loss": 0.8302, + "loss": 0.856, "step": 823 }, { "epoch": 3.3360323886639676, - "grad_norm": 0.4096270004185005, + "grad_norm": 0.2639774834122571, "learning_rate": 1.2053409133511681e-05, - "loss": 0.8122, + "loss": 0.8355, "step": 824 }, { "epoch": 3.340080971659919, - "grad_norm": 0.4846087309138413, + "grad_norm": 0.35914884368029937, "learning_rate": 1.2001542428160173e-05, - "loss": 0.8323, + "loss": 0.8432, "step": 825 }, { "epoch": 3.3441295546558703, - "grad_norm": 0.4327728576290646, + "grad_norm": 0.3435243568275602, "learning_rate": 1.1949739678262278e-05, - "loss": 0.8242, + "loss": 0.8598, "step": 826 }, { "epoch": 3.348178137651822, - "grad_norm": 0.47961805691116915, + "grad_norm": 0.3163586351693701, "learning_rate": 1.189800129803139e-05, - "loss": 0.8337, + "loss": 0.8658, "step": 827 }, { "epoch": 3.3522267206477734, - "grad_norm": 0.4527435564706166, + "grad_norm": 0.36479787811801595, "learning_rate": 1.1846327701166228e-05, - "loss": 0.8351, + "loss": 0.8733, "step": 828 }, { "epoch": 3.3562753036437245, - "grad_norm": 0.410778602362315, + "grad_norm": 0.30556707622803564, "learning_rate": 1.1794719300847493e-05, - "loss": 0.8239, + "loss": 0.8437, "step": 829 }, { "epoch": 3.360323886639676, - "grad_norm": 0.4475391446217267, + "grad_norm": 0.34762187665526967, "learning_rate": 1.1743176509734582e-05, - "loss": 0.8256, + "loss": 0.8495, "step": 830 }, { "epoch": 3.3643724696356276, - "grad_norm": 0.4256238262971765, + "grad_norm": 0.3527809088915443, "learning_rate": 1.1691699739962275e-05, - "loss": 0.8316, + "loss": 0.8528, "step": 831 }, { "epoch": 3.3684210526315788, - "grad_norm": 0.4491498330152042, + "grad_norm": 0.3223076946468397, "learning_rate": 1.1640289403137456e-05, - "loss": 0.8354, + "loss": 0.875, "step": 832 }, { "epoch": 3.3724696356275303, - "grad_norm": 0.45079955963424784, + "grad_norm": 0.4042811409881392, "learning_rate": 1.1588945910335802e-05, - "loss": 0.8436, + "loss": 0.8634, "step": 833 }, { "epoch": 3.376518218623482, - "grad_norm": 0.4230013393650072, + "grad_norm": 0.3088903454633702, "learning_rate": 1.1537669672098517e-05, - "loss": 0.8129, + "loss": 0.8363, "step": 834 }, { "epoch": 3.380566801619433, - "grad_norm": 0.45598732831437755, + "grad_norm": 0.31407502045446456, "learning_rate": 1.1486461098429036e-05, - "loss": 0.8552, + "loss": 0.8735, "step": 835 }, { "epoch": 3.3846153846153846, - "grad_norm": 0.43953344688261153, + "grad_norm": 0.3352601933625247, "learning_rate": 1.1435320598789747e-05, - "loss": 0.8177, + "loss": 0.8428, "step": 836 }, { "epoch": 3.388663967611336, - "grad_norm": 0.42861608910274496, + "grad_norm": 0.2929084090806428, "learning_rate": 1.1384248582098736e-05, - "loss": 0.8357, + "loss": 0.8493, "step": 837 }, { "epoch": 3.3927125506072873, - "grad_norm": 0.43512782651532705, + "grad_norm": 0.33512492055155413, "learning_rate": 1.1333245456726467e-05, - "loss": 0.8346, + "loss": 0.8443, "step": 838 }, { "epoch": 3.396761133603239, - "grad_norm": 0.45250480122189995, + "grad_norm": 0.3140577799115343, "learning_rate": 1.128231163049258e-05, - "loss": 0.8223, + "loss": 0.8542, "step": 839 }, { "epoch": 3.4008097165991904, - "grad_norm": 0.44198620330585003, + "grad_norm": 0.3097916729513693, "learning_rate": 1.123144751066259e-05, - "loss": 0.8387, + "loss": 0.8615, "step": 840 }, { "epoch": 3.4048582995951415, - "grad_norm": 0.420999879666562, + "grad_norm": 0.30901470317796276, "learning_rate": 1.1180653503944657e-05, - "loss": 0.8083, + "loss": 0.8416, "step": 841 }, { "epoch": 3.408906882591093, - "grad_norm": 0.41038780690495225, + "grad_norm": 0.28697801639781634, "learning_rate": 1.1129930016486287e-05, - "loss": 0.842, + "loss": 0.8492, "step": 842 }, { "epoch": 3.4129554655870447, - "grad_norm": 0.47495384933028306, + "grad_norm": 0.291713280799152, "learning_rate": 1.107927745387114e-05, - "loss": 0.8199, + "loss": 0.8306, "step": 843 }, { "epoch": 3.417004048582996, - "grad_norm": 0.4584368492824357, + "grad_norm": 0.3072937654221214, "learning_rate": 1.1028696221115766e-05, - "loss": 0.832, + "loss": 0.8503, "step": 844 }, { "epoch": 3.4210526315789473, - "grad_norm": 0.40927985257014315, + "grad_norm": 0.3013764636913997, "learning_rate": 1.0978186722666341e-05, - "loss": 0.8254, + "loss": 0.8452, "step": 845 }, { "epoch": 3.425101214574899, - "grad_norm": 0.49055927700895824, + "grad_norm": 0.31965529226485273, "learning_rate": 1.0927749362395475e-05, - "loss": 0.833, + "loss": 0.8603, "step": 846 }, { "epoch": 3.42914979757085, - "grad_norm": 0.4221725206211155, + "grad_norm": 0.31620055109003237, "learning_rate": 1.0877384543598954e-05, - "loss": 0.8212, + "loss": 0.8379, "step": 847 }, { "epoch": 3.4331983805668016, - "grad_norm": 0.45431887932319504, + "grad_norm": 0.3387508830498116, "learning_rate": 1.0827092668992529e-05, - "loss": 0.8389, + "loss": 0.8592, "step": 848 }, { "epoch": 3.437246963562753, - "grad_norm": 0.43523539368853814, + "grad_norm": 0.31162689330126997, "learning_rate": 1.077687414070869e-05, - "loss": 0.8196, + "loss": 0.8438, "step": 849 }, { "epoch": 3.4412955465587043, - "grad_norm": 0.42350490328050255, + "grad_norm": 0.37577275474788485, "learning_rate": 1.0726729360293451e-05, - "loss": 0.8461, + "loss": 0.8656, "step": 850 }, { "epoch": 3.445344129554656, - "grad_norm": 0.43682383820257775, + "grad_norm": 0.3132580423156575, "learning_rate": 1.0676658728703147e-05, - "loss": 0.8417, + "loss": 0.8618, "step": 851 }, { "epoch": 3.4493927125506074, - "grad_norm": 0.4326203813033068, + "grad_norm": 0.3234728293300849, "learning_rate": 1.0626662646301198e-05, - "loss": 0.8405, + "loss": 0.858, "step": 852 }, { "epoch": 3.4534412955465585, - "grad_norm": 0.44099363418481946, + "grad_norm": 0.3656969565454102, "learning_rate": 1.0576741512854958e-05, - "loss": 0.8478, + "loss": 0.8521, "step": 853 }, { "epoch": 3.45748987854251, - "grad_norm": 0.40713077968882105, + "grad_norm": 0.35273349345014, "learning_rate": 1.052689572753247e-05, - "loss": 0.8261, + "loss": 0.859, "step": 854 }, { "epoch": 3.4615384615384617, - "grad_norm": 0.4751119113631104, + "grad_norm": 0.3014211882836784, "learning_rate": 1.0477125688899328e-05, - "loss": 0.8407, + "loss": 0.8568, "step": 855 }, { "epoch": 3.465587044534413, - "grad_norm": 0.4530496313464184, + "grad_norm": 0.3638075714632998, "learning_rate": 1.0427431794915406e-05, - "loss": 0.8547, + "loss": 0.8789, "step": 856 }, { "epoch": 3.4696356275303644, - "grad_norm": 0.43225658794406036, + "grad_norm": 0.3581610042313006, "learning_rate": 1.0377814442931769e-05, - "loss": 0.8364, + "loss": 0.8587, "step": 857 }, { "epoch": 3.473684210526316, - "grad_norm": 0.4185492336072585, + "grad_norm": 0.37131069249911125, "learning_rate": 1.0328274029687439e-05, - "loss": 0.8121, + "loss": 0.8403, "step": 858 }, { "epoch": 3.477732793522267, - "grad_norm": 0.43713801461460616, + "grad_norm": 0.3323026878593802, "learning_rate": 1.0278810951306253e-05, - "loss": 0.8512, + "loss": 0.8705, "step": 859 }, { "epoch": 3.4817813765182186, - "grad_norm": 0.4075887794177538, + "grad_norm": 0.36235596783623203, "learning_rate": 1.0229425603293647e-05, - "loss": 0.8327, + "loss": 0.8507, "step": 860 }, { "epoch": 3.48582995951417, - "grad_norm": 0.4468608488295686, + "grad_norm": 0.3017715975891122, "learning_rate": 1.0180118380533558e-05, - "loss": 0.8308, + "loss": 0.8523, "step": 861 }, { "epoch": 3.4898785425101213, - "grad_norm": 0.41933958997293136, + "grad_norm": 0.3730631188405016, "learning_rate": 1.0130889677285225e-05, - "loss": 0.8197, + "loss": 0.8286, "step": 862 }, { "epoch": 3.493927125506073, - "grad_norm": 0.4124381108429523, + "grad_norm": 0.3101457070164216, "learning_rate": 1.008173988718005e-05, - "loss": 0.8235, + "loss": 0.8286, "step": 863 }, { "epoch": 3.4979757085020244, - "grad_norm": 0.4278707007278682, + "grad_norm": 0.290535257793208, "learning_rate": 1.0032669403218439e-05, - "loss": 0.821, + "loss": 0.8377, "step": 864 }, { "epoch": 3.5020242914979756, - "grad_norm": 0.40046274292173345, + "grad_norm": 0.3302818689664793, "learning_rate": 9.983678617766677e-06, - "loss": 0.8515, + "loss": 0.8575, "step": 865 }, { "epoch": 3.506072874493927, - "grad_norm": 0.45634869143314205, + "grad_norm": 0.2980800496528561, "learning_rate": 9.934767922553788e-06, - "loss": 0.8322, + "loss": 0.8517, "step": 866 }, { "epoch": 3.5101214574898787, - "grad_norm": 0.408891727868476, + "grad_norm": 0.31940073048125056, "learning_rate": 9.885937708668371e-06, - "loss": 0.8249, + "loss": 0.8378, "step": 867 }, { "epoch": 3.51417004048583, - "grad_norm": 0.470334622874235, + "grad_norm": 0.3128099697241155, "learning_rate": 9.837188366555524e-06, - "loss": 0.8435, + "loss": 0.8591, "step": 868 }, { "epoch": 3.5182186234817814, - "grad_norm": 0.4444884288518513, + "grad_norm": 0.2905935195146965, "learning_rate": 9.788520286013691e-06, - "loss": 0.843, + "loss": 0.8612, "step": 869 }, { "epoch": 3.522267206477733, - "grad_norm": 0.39361986469596405, + "grad_norm": 0.29025417592529773, "learning_rate": 9.739933856191537e-06, - "loss": 0.8364, + "loss": 0.867, "step": 870 }, { "epoch": 3.526315789473684, - "grad_norm": 0.4668750111857091, + "grad_norm": 0.31122517558393653, "learning_rate": 9.691429465584866e-06, - "loss": 0.8386, + "loss": 0.8588, "step": 871 }, { "epoch": 3.5303643724696356, - "grad_norm": 0.4411881379962621, + "grad_norm": 0.31090146699153187, "learning_rate": 9.643007502033492e-06, - "loss": 0.8245, + "loss": 0.839, "step": 872 }, { "epoch": 3.534412955465587, - "grad_norm": 0.4025689620602145, + "grad_norm": 0.3001923733194765, "learning_rate": 9.594668352718152e-06, - "loss": 0.836, + "loss": 0.8656, "step": 873 }, { "epoch": 3.5384615384615383, - "grad_norm": 0.42293967578672204, + "grad_norm": 0.3416700647723706, "learning_rate": 9.546412404157383e-06, - "loss": 0.8294, + "loss": 0.8518, "step": 874 }, { "epoch": 3.54251012145749, - "grad_norm": 0.40073693780342, + "grad_norm": 0.27583896493647125, "learning_rate": 9.498240042204471e-06, - "loss": 0.8245, + "loss": 0.8229, "step": 875 }, { "epoch": 3.5465587044534415, - "grad_norm": 0.4055656614750957, + "grad_norm": 0.30400734119931905, "learning_rate": 9.450151652044339e-06, - "loss": 0.8075, + "loss": 0.8333, "step": 876 }, { "epoch": 3.5506072874493926, - "grad_norm": 0.4612110960730307, + "grad_norm": 0.2936434863496949, "learning_rate": 9.402147618190472e-06, - "loss": 0.8331, + "loss": 0.8589, "step": 877 }, { "epoch": 3.554655870445344, - "grad_norm": 0.4396595385207427, + "grad_norm": 0.2916226199558911, "learning_rate": 9.354228324481841e-06, - "loss": 0.8431, + "loss": 0.8535, "step": 878 }, { "epoch": 3.5587044534412957, - "grad_norm": 0.4468966070531797, + "grad_norm": 0.2961618885107933, "learning_rate": 9.306394154079852e-06, - "loss": 0.8206, + "loss": 0.8579, "step": 879 }, { "epoch": 3.562753036437247, - "grad_norm": 0.41708418917449397, + "grad_norm": 0.3129215525200299, "learning_rate": 9.258645489465254e-06, - "loss": 0.8481, + "loss": 0.8563, "step": 880 }, { "epoch": 3.5668016194331984, - "grad_norm": 0.4489068004581832, + "grad_norm": 0.2909791205015997, "learning_rate": 9.21098271243509e-06, - "loss": 0.8356, + "loss": 0.8619, "step": 881 }, { "epoch": 3.57085020242915, - "grad_norm": 0.4185368318765376, + "grad_norm": 0.30129198768709153, "learning_rate": 9.163406204099659e-06, - "loss": 0.842, + "loss": 0.8389, "step": 882 }, { "epoch": 3.574898785425101, - "grad_norm": 0.42531556909176854, + "grad_norm": 0.28489098554695813, "learning_rate": 9.115916344879452e-06, - "loss": 0.8393, + "loss": 0.8527, "step": 883 }, { "epoch": 3.5789473684210527, - "grad_norm": 0.4277799056967363, + "grad_norm": 0.3489754304597703, "learning_rate": 9.068513514502126e-06, - "loss": 0.8342, + "loss": 0.857, "step": 884 }, { "epoch": 3.582995951417004, - "grad_norm": 0.4474914999265442, + "grad_norm": 0.2754719644023403, "learning_rate": 9.021198091999439e-06, - "loss": 0.8509, + "loss": 0.8786, "step": 885 }, { "epoch": 3.5870445344129553, - "grad_norm": 0.40711233097767346, + "grad_norm": 0.31235803110175736, "learning_rate": 8.973970455704251e-06, - "loss": 0.8296, + "loss": 0.8505, "step": 886 }, { "epoch": 3.591093117408907, - "grad_norm": 0.4330771510131619, + "grad_norm": 0.28607611915341347, "learning_rate": 8.926830983247495e-06, - "loss": 0.8216, + "loss": 0.8447, "step": 887 }, { "epoch": 3.5951417004048585, - "grad_norm": 0.4491990031690705, + "grad_norm": 0.28526753259495546, "learning_rate": 8.879780051555116e-06, - "loss": 0.8431, + "loss": 0.8486, "step": 888 }, { "epoch": 3.5991902834008096, - "grad_norm": 0.40783902615520434, + "grad_norm": 0.314309509457857, "learning_rate": 8.83281803684512e-06, - "loss": 0.8081, + "loss": 0.8329, "step": 889 }, { "epoch": 3.603238866396761, - "grad_norm": 0.4053102126779133, + "grad_norm": 0.2787768440976813, "learning_rate": 8.785945314624532e-06, - "loss": 0.8101, + "loss": 0.8228, "step": 890 }, { "epoch": 3.6072874493927127, - "grad_norm": 0.4459612084747026, + "grad_norm": 0.32022773715538727, "learning_rate": 8.739162259686385e-06, - "loss": 0.8377, + "loss": 0.8654, "step": 891 }, { "epoch": 3.611336032388664, - "grad_norm": 0.4318304527856228, + "grad_norm": 0.358365972222895, "learning_rate": 8.692469246106734e-06, - "loss": 0.8264, + "loss": 0.8435, "step": 892 }, { "epoch": 3.6153846153846154, - "grad_norm": 0.39855837867112526, + "grad_norm": 0.27903017142124126, "learning_rate": 8.645866647241677e-06, - "loss": 0.8156, + "loss": 0.8535, "step": 893 }, { "epoch": 3.619433198380567, - "grad_norm": 0.44635535292704404, + "grad_norm": 0.3106034348704156, "learning_rate": 8.599354835724348e-06, - "loss": 0.8427, + "loss": 0.8682, "step": 894 }, { "epoch": 3.623481781376518, - "grad_norm": 0.4131381982813608, + "grad_norm": 0.2883630971341835, "learning_rate": 8.552934183461958e-06, - "loss": 0.8236, + "loss": 0.8535, "step": 895 }, { "epoch": 3.6275303643724697, - "grad_norm": 0.4367584125869411, + "grad_norm": 0.348567850996918, "learning_rate": 8.506605061632805e-06, - "loss": 0.8187, + "loss": 0.837, "step": 896 }, { "epoch": 3.6315789473684212, - "grad_norm": 0.45893239738653946, + "grad_norm": 0.3100986355429715, "learning_rate": 8.460367840683308e-06, - "loss": 0.8455, + "loss": 0.8755, "step": 897 }, { "epoch": 3.6356275303643724, - "grad_norm": 0.42496548783038923, + "grad_norm": 0.3308149431549723, "learning_rate": 8.414222890325065e-06, - "loss": 0.8389, + "loss": 0.8613, "step": 898 }, { "epoch": 3.639676113360324, - "grad_norm": 0.46888939042548133, + "grad_norm": 0.3304734491760831, "learning_rate": 8.368170579531855e-06, - "loss": 0.8323, + "loss": 0.8337, "step": 899 }, { "epoch": 3.6437246963562755, - "grad_norm": 0.39560018308212147, + "grad_norm": 0.3188102532288472, "learning_rate": 8.322211276536733e-06, - "loss": 0.8204, + "loss": 0.834, "step": 900 }, { "epoch": 3.6477732793522266, - "grad_norm": 0.4623326194817346, + "grad_norm": 0.3417018683092118, "learning_rate": 8.27634534882907e-06, - "loss": 0.8362, + "loss": 0.8685, "step": 901 }, { "epoch": 3.651821862348178, - "grad_norm": 0.41952391684052126, + "grad_norm": 0.2882480394406787, "learning_rate": 8.230573163151585e-06, - "loss": 0.8369, + "loss": 0.8577, "step": 902 }, { "epoch": 3.6558704453441297, - "grad_norm": 0.39848658852074026, + "grad_norm": 0.3413878964016963, "learning_rate": 8.184895085497466e-06, - "loss": 0.8175, + "loss": 0.8399, "step": 903 }, { "epoch": 3.659919028340081, - "grad_norm": 0.43538604110004697, + "grad_norm": 0.3104336893706697, "learning_rate": 8.139311481107401e-06, - "loss": 0.8361, + "loss": 0.8509, "step": 904 }, { "epoch": 3.6639676113360324, - "grad_norm": 0.4027865392059006, + "grad_norm": 0.2969720366996896, "learning_rate": 8.093822714466686e-06, - "loss": 0.802, + "loss": 0.8441, "step": 905 }, { "epoch": 3.668016194331984, - "grad_norm": 0.420442880748298, + "grad_norm": 0.32350585809520777, "learning_rate": 8.048429149302271e-06, - "loss": 0.8225, + "loss": 0.8472, "step": 906 }, { "epoch": 3.672064777327935, - "grad_norm": 0.4003636255886726, + "grad_norm": 0.3182079999155654, "learning_rate": 8.003131148579902e-06, - "loss": 0.8241, + "loss": 0.8551, "step": 907 }, { "epoch": 3.6761133603238867, - "grad_norm": 0.4071114557049831, + "grad_norm": 0.3263059042136318, "learning_rate": 7.957929074501182e-06, - "loss": 0.8478, + "loss": 0.8679, "step": 908 }, { "epoch": 3.6801619433198383, - "grad_norm": 0.4082247074323287, + "grad_norm": 0.30989510623201194, "learning_rate": 7.912823288500695e-06, - "loss": 0.8518, + "loss": 0.8864, "step": 909 }, { "epoch": 3.6842105263157894, - "grad_norm": 0.4241113899822108, + "grad_norm": 0.29617619025781605, "learning_rate": 7.867814151243102e-06, - "loss": 0.8095, + "loss": 0.8317, "step": 910 }, { "epoch": 3.688259109311741, - "grad_norm": 0.4343038790846477, + "grad_norm": 0.3219264844431389, "learning_rate": 7.822902022620267e-06, - "loss": 0.8211, + "loss": 0.8363, "step": 911 }, { "epoch": 3.6923076923076925, - "grad_norm": 0.38597112927101307, + "grad_norm": 0.2895942803343257, "learning_rate": 7.778087261748377e-06, - "loss": 0.818, + "loss": 0.838, "step": 912 }, { "epoch": 3.6963562753036436, - "grad_norm": 0.4181283117617946, + "grad_norm": 0.27967429596295845, "learning_rate": 7.733370226965056e-06, - "loss": 0.842, + "loss": 0.8631, "step": 913 }, { "epoch": 3.700404858299595, - "grad_norm": 0.4279496606611242, + "grad_norm": 0.32099982004659633, "learning_rate": 7.688751275826527e-06, - "loss": 0.8103, + "loss": 0.8324, "step": 914 }, { "epoch": 3.7044534412955468, - "grad_norm": 0.4155381755307851, + "grad_norm": 0.28082068908893965, "learning_rate": 7.644230765104735e-06, - "loss": 0.8167, + "loss": 0.8299, "step": 915 }, { "epoch": 3.708502024291498, - "grad_norm": 0.38543229356707964, + "grad_norm": 0.26880345954398266, "learning_rate": 7.599809050784502e-06, - "loss": 0.8139, + "loss": 0.8458, "step": 916 }, { "epoch": 3.7125506072874495, - "grad_norm": 0.4108455198298163, + "grad_norm": 0.2813338660859997, "learning_rate": 7.555486488060661e-06, - "loss": 0.8328, + "loss": 0.8629, "step": 917 }, { "epoch": 3.716599190283401, - "grad_norm": 0.4315913370345731, + "grad_norm": 0.2588497717746283, "learning_rate": 7.511263431335249e-06, - "loss": 0.819, + "loss": 0.8538, "step": 918 }, { "epoch": 3.720647773279352, - "grad_norm": 0.42066563968917625, + "grad_norm": 0.27394024266233746, "learning_rate": 7.467140234214656e-06, - "loss": 0.8375, + "loss": 0.8698, "step": 919 }, { "epoch": 3.7246963562753037, - "grad_norm": 0.4241686276653224, + "grad_norm": 0.2859835686635573, "learning_rate": 7.423117249506777e-06, - "loss": 0.8437, + "loss": 0.8759, "step": 920 }, { "epoch": 3.7287449392712553, - "grad_norm": 0.3969857306771706, + "grad_norm": 0.28865058358792894, "learning_rate": 7.379194829218231e-06, - "loss": 0.8518, + "loss": 0.8587, "step": 921 }, { "epoch": 3.7327935222672064, - "grad_norm": 0.4106917924317347, + "grad_norm": 0.2524415575525996, "learning_rate": 7.335373324551518e-06, - "loss": 0.8324, + "loss": 0.8569, "step": 922 }, { "epoch": 3.736842105263158, - "grad_norm": 0.38834531354803276, + "grad_norm": 0.2764284770139992, "learning_rate": 7.291653085902224e-06, - "loss": 0.8152, + "loss": 0.8474, "step": 923 }, { "epoch": 3.7408906882591095, - "grad_norm": 0.4373644275195999, + "grad_norm": 0.27126105081675134, "learning_rate": 7.248034462856213e-06, - "loss": 0.841, + "loss": 0.8622, "step": 924 }, { "epoch": 3.7449392712550607, - "grad_norm": 0.4104339650273692, + "grad_norm": 0.281056452503062, "learning_rate": 7.20451780418683e-06, - "loss": 0.8213, + "loss": 0.8337, "step": 925 }, { "epoch": 3.748987854251012, - "grad_norm": 0.37604725603069405, + "grad_norm": 0.2980912995204157, "learning_rate": 7.161103457852128e-06, - "loss": 0.8294, + "loss": 0.862, "step": 926 }, { "epoch": 3.753036437246964, - "grad_norm": 0.3906317044625437, + "grad_norm": 0.29934024012427124, "learning_rate": 7.117791770992046e-06, - "loss": 0.8331, + "loss": 0.8661, "step": 927 }, { "epoch": 3.757085020242915, - "grad_norm": 0.3956296805378481, + "grad_norm": 0.28151545056800004, "learning_rate": 7.074583089925688e-06, - "loss": 0.8295, + "loss": 0.8596, "step": 928 }, { "epoch": 3.7611336032388665, - "grad_norm": 0.4197863873256336, + "grad_norm": 0.26345873189212865, "learning_rate": 7.0314777601485155e-06, - "loss": 0.8486, + "loss": 0.8576, "step": 929 }, { "epoch": 3.765182186234818, - "grad_norm": 0.4607096476516237, + "grad_norm": 0.2517554440346497, "learning_rate": 6.988476126329604e-06, - "loss": 0.8363, + "loss": 0.8631, "step": 930 }, { "epoch": 3.769230769230769, - "grad_norm": 0.40731121911657564, + "grad_norm": 0.32171317185764925, "learning_rate": 6.945578532308857e-06, - "loss": 0.8332, + "loss": 0.8458, "step": 931 }, { "epoch": 3.7732793522267207, - "grad_norm": 0.3865433785728649, + "grad_norm": 0.28322559373756956, "learning_rate": 6.902785321094301e-06, - "loss": 0.8525, + "loss": 0.87, "step": 932 }, { "epoch": 3.7773279352226723, - "grad_norm": 0.40197804335661635, + "grad_norm": 0.26226742561744026, "learning_rate": 6.8600968348593115e-06, - "loss": 0.8259, + "loss": 0.8479, "step": 933 }, { "epoch": 3.7813765182186234, - "grad_norm": 0.4028652192667042, + "grad_norm": 0.25530982055438156, "learning_rate": 6.817513414939892e-06, - "loss": 0.8246, + "loss": 0.8384, "step": 934 }, { "epoch": 3.785425101214575, - "grad_norm": 0.4438250223633903, + "grad_norm": 0.31747674130531467, "learning_rate": 6.775035401831924e-06, - "loss": 0.8304, + "loss": 0.8591, "step": 935 }, { "epoch": 3.7894736842105265, - "grad_norm": 0.39570956672750107, + "grad_norm": 0.28602400393431743, "learning_rate": 6.732663135188473e-06, - "loss": 0.8262, + "loss": 0.84, "step": 936 }, { "epoch": 3.7935222672064777, - "grad_norm": 0.40390300573974414, + "grad_norm": 0.2789687716811859, "learning_rate": 6.690396953817055e-06, - "loss": 0.8307, + "loss": 0.8624, "step": 937 }, { "epoch": 3.7975708502024292, - "grad_norm": 0.41466074159053146, + "grad_norm": 0.3010946353307862, "learning_rate": 6.648237195676927e-06, - "loss": 0.829, + "loss": 0.8536, "step": 938 }, { "epoch": 3.801619433198381, - "grad_norm": 0.4273556084330557, + "grad_norm": 0.2758110333914285, "learning_rate": 6.606184197876395e-06, - "loss": 0.8263, + "loss": 0.8447, "step": 939 }, { "epoch": 3.805668016194332, - "grad_norm": 0.40053229373957766, + "grad_norm": 0.2494796793122086, "learning_rate": 6.564238296670109e-06, - "loss": 0.8149, + "loss": 0.8307, "step": 940 }, { "epoch": 3.8097165991902835, - "grad_norm": 0.4075848067887678, + "grad_norm": 0.2835644157930435, "learning_rate": 6.52239982745638e-06, - "loss": 0.8118, + "loss": 0.8426, "step": 941 }, { "epoch": 3.813765182186235, - "grad_norm": 0.41758212354637736, + "grad_norm": 0.2590850422104621, "learning_rate": 6.480669124774484e-06, - "loss": 0.839, + "loss": 0.8691, "step": 942 }, { "epoch": 3.817813765182186, - "grad_norm": 0.39644735620526267, + "grad_norm": 0.28287067852881187, "learning_rate": 6.439046522302013e-06, - "loss": 0.8224, + "loss": 0.8312, "step": 943 }, { "epoch": 3.8218623481781377, - "grad_norm": 0.4248340652533807, + "grad_norm": 0.24014799486470478, "learning_rate": 6.397532352852191e-06, - "loss": 0.8425, + "loss": 0.85, "step": 944 }, { "epoch": 3.8259109311740893, - "grad_norm": 0.3927043811729782, + "grad_norm": 0.27345252985044155, "learning_rate": 6.356126948371202e-06, - "loss": 0.8596, + "loss": 0.8812, "step": 945 }, { "epoch": 3.8299595141700404, - "grad_norm": 0.4077747766599207, + "grad_norm": 0.26669812977186974, "learning_rate": 6.314830639935559e-06, - "loss": 0.8291, + "loss": 0.8478, "step": 946 }, { "epoch": 3.834008097165992, - "grad_norm": 0.40737582009004353, + "grad_norm": 0.2614627547925244, "learning_rate": 6.273643757749448e-06, - "loss": 0.8549, + "loss": 0.8793, "step": 947 }, { "epoch": 3.8380566801619436, - "grad_norm": 0.4079336014228566, + "grad_norm": 0.2719729622322591, "learning_rate": 6.2325666311420856e-06, - "loss": 0.8258, + "loss": 0.8621, "step": 948 }, { "epoch": 3.8421052631578947, - "grad_norm": 0.4289378841625121, + "grad_norm": 0.27577854797983054, "learning_rate": 6.1915995885650675e-06, - "loss": 0.847, + "loss": 0.8671, "step": 949 }, { "epoch": 3.8461538461538463, - "grad_norm": 0.44679024160777675, + "grad_norm": 0.2849047751315424, "learning_rate": 6.150742957589784e-06, - "loss": 0.8517, + "loss": 0.8725, "step": 950 }, { "epoch": 3.850202429149798, - "grad_norm": 0.3790027259438454, + "grad_norm": 0.2523328615103755, "learning_rate": 6.109997064904774e-06, - "loss": 0.8302, + "loss": 0.8527, "step": 951 }, { "epoch": 3.854251012145749, - "grad_norm": 0.4116671013279036, + "grad_norm": 0.26985286048660395, "learning_rate": 6.069362236313092e-06, - "loss": 0.8131, + "loss": 0.8342, "step": 952 }, { "epoch": 3.8582995951417005, - "grad_norm": 0.41480080631371047, + "grad_norm": 0.27345811086442656, "learning_rate": 6.0288387967297565e-06, - "loss": 0.8255, + "loss": 0.8602, "step": 953 }, { "epoch": 3.862348178137652, - "grad_norm": 0.44548472286307045, + "grad_norm": 0.2840152732370331, "learning_rate": 5.988427070179108e-06, - "loss": 0.8362, + "loss": 0.8751, "step": 954 }, { "epoch": 3.866396761133603, - "grad_norm": 0.40609230277966935, + "grad_norm": 0.2748096281885799, "learning_rate": 5.948127379792234e-06, - "loss": 0.8324, + "loss": 0.8579, "step": 955 }, { "epoch": 3.8704453441295548, - "grad_norm": 0.40875657662750675, + "grad_norm": 0.24655078385572227, "learning_rate": 5.90794004780439e-06, - "loss": 0.8281, + "loss": 0.8516, "step": 956 }, { "epoch": 3.8744939271255063, - "grad_norm": 0.40728929425101806, + "grad_norm": 0.27764811315228016, "learning_rate": 5.867865395552405e-06, - "loss": 0.8361, + "loss": 0.8536, "step": 957 }, { "epoch": 3.8785425101214575, - "grad_norm": 0.4127850832817179, + "grad_norm": 0.25955621224472, "learning_rate": 5.8279037434721364e-06, - "loss": 0.807, + "loss": 0.8376, "step": 958 }, { "epoch": 3.882591093117409, - "grad_norm": 0.4043798435965883, + "grad_norm": 0.24384272595413295, "learning_rate": 5.788055411095892e-06, - "loss": 0.8197, + "loss": 0.853, "step": 959 }, { "epoch": 3.8866396761133606, - "grad_norm": 0.4326684094029027, + "grad_norm": 0.2598529273334053, "learning_rate": 5.7483207170498625e-06, - "loss": 0.8271, + "loss": 0.8566, "step": 960 }, { "epoch": 3.8906882591093117, - "grad_norm": 0.414933730310786, + "grad_norm": 0.2732261444116901, "learning_rate": 5.708699979051613e-06, - "loss": 0.8659, + "loss": 0.8843, "step": 961 }, { "epoch": 3.8947368421052633, - "grad_norm": 0.3957759396010362, + "grad_norm": 0.26652880951522734, "learning_rate": 5.66919351390751e-06, - "loss": 0.8267, + "loss": 0.8402, "step": 962 }, { "epoch": 3.898785425101215, - "grad_norm": 0.38955610090027687, + "grad_norm": 0.2513966862099206, "learning_rate": 5.629801637510188e-06, - "loss": 0.8222, + "loss": 0.8592, "step": 963 }, { "epoch": 3.902834008097166, - "grad_norm": 0.45594620136340147, + "grad_norm": 0.23360886429205252, "learning_rate": 5.590524664836043e-06, - "loss": 0.816, + "loss": 0.8337, "step": 964 }, { "epoch": 3.9068825910931175, - "grad_norm": 0.4037984655637796, + "grad_norm": 0.23062943628961874, "learning_rate": 5.551362909942708e-06, - "loss": 0.8283, + "loss": 0.8533, "step": 965 }, { "epoch": 3.910931174089069, - "grad_norm": 0.3870414934137407, + "grad_norm": 0.24240054504428182, "learning_rate": 5.51231668596653e-06, - "loss": 0.8353, + "loss": 0.8704, "step": 966 }, { "epoch": 3.91497975708502, - "grad_norm": 0.38110817598086527, + "grad_norm": 0.2407632887089734, "learning_rate": 5.47338630512007e-06, - "loss": 0.8416, + "loss": 0.8701, "step": 967 }, { "epoch": 3.919028340080972, - "grad_norm": 0.39918928330373743, + "grad_norm": 0.26111651993815166, "learning_rate": 5.434572078689617e-06, - "loss": 0.8521, + "loss": 0.8818, "step": 968 }, { "epoch": 3.9230769230769234, - "grad_norm": 0.3898370651104759, + "grad_norm": 0.2357336347196473, "learning_rate": 5.395874317032694e-06, - "loss": 0.8175, + "loss": 0.8502, "step": 969 }, { "epoch": 3.9271255060728745, - "grad_norm": 0.3947520323314868, + "grad_norm": 0.24419263453623283, "learning_rate": 5.357293329575573e-06, - "loss": 0.8399, + "loss": 0.8711, "step": 970 }, { "epoch": 3.931174089068826, - "grad_norm": 0.4221485812901026, + "grad_norm": 0.2678700169092211, "learning_rate": 5.318829424810797e-06, - "loss": 0.8202, + "loss": 0.8538, "step": 971 }, { "epoch": 3.9352226720647776, - "grad_norm": 0.3663620678546695, + "grad_norm": 0.23167745678748597, "learning_rate": 5.280482910294728e-06, - "loss": 0.818, + "loss": 0.8424, "step": 972 }, { "epoch": 3.9392712550607287, - "grad_norm": 0.39788695074479397, + "grad_norm": 0.24557710357899323, "learning_rate": 5.242254092645078e-06, - "loss": 0.8508, + "loss": 0.8673, "step": 973 }, { "epoch": 3.9433198380566803, - "grad_norm": 0.39552588825139373, + "grad_norm": 0.3216542264447858, "learning_rate": 5.2041432775384445e-06, - "loss": 0.8126, + "loss": 0.8325, "step": 974 }, { "epoch": 3.9473684210526314, - "grad_norm": 0.40314942997637754, + "grad_norm": 0.297145468065552, "learning_rate": 5.166150769707892e-06, - "loss": 0.8482, + "loss": 0.8676, "step": 975 }, { "epoch": 3.951417004048583, - "grad_norm": 0.405456128359912, + "grad_norm": 0.2588154900237031, "learning_rate": 5.128276872940503e-06, - "loss": 0.8196, + "loss": 0.8479, "step": 976 }, { "epoch": 3.9554655870445345, - "grad_norm": 0.42766479074539077, + "grad_norm": 0.2999374861376341, "learning_rate": 5.090521890074953e-06, - "loss": 0.8313, + "loss": 0.847, "step": 977 }, { "epoch": 3.9595141700404857, - "grad_norm": 0.45227121235570317, + "grad_norm": 0.24644665426513312, "learning_rate": 5.0528861229990675e-06, - "loss": 0.8283, + "loss": 0.8519, "step": 978 }, { "epoch": 3.9635627530364372, - "grad_norm": 0.42547087149990587, + "grad_norm": 0.261509767744411, "learning_rate": 5.015369872647442e-06, - "loss": 0.8532, + "loss": 0.8571, "step": 979 }, { "epoch": 3.967611336032389, - "grad_norm": 0.40627370404267577, + "grad_norm": 0.26966962225676594, "learning_rate": 4.9779734389990195e-06, - "loss": 0.8229, + "loss": 0.8569, "step": 980 }, { "epoch": 3.97165991902834, - "grad_norm": 0.42766474158845674, + "grad_norm": 0.2629090892730354, "learning_rate": 4.940697121074675e-06, - "loss": 0.8284, + "loss": 0.8549, "step": 981 }, { "epoch": 3.9757085020242915, - "grad_norm": 0.43288113711246917, + "grad_norm": 0.23499051308231428, "learning_rate": 4.903541216934864e-06, - "loss": 0.826, + "loss": 0.8502, "step": 982 }, { "epoch": 3.979757085020243, - "grad_norm": 0.3972928933302155, + "grad_norm": 0.3124685736528782, "learning_rate": 4.8665060236772e-06, - "loss": 0.842, + "loss": 0.8825, "step": 983 }, { "epoch": 3.983805668016194, - "grad_norm": 0.3823029131116864, + "grad_norm": 0.2571739938386458, "learning_rate": 4.829591837434102e-06, - "loss": 0.8303, + "loss": 0.8399, "step": 984 }, { "epoch": 3.9878542510121457, - "grad_norm": 0.36915759290268996, + "grad_norm": 0.26863368158909007, "learning_rate": 4.792798953370424e-06, - "loss": 0.8229, + "loss": 0.8439, "step": 985 }, { "epoch": 3.9919028340080973, - "grad_norm": 0.4092086386292386, + "grad_norm": 0.26247924487163865, "learning_rate": 4.756127665681083e-06, - "loss": 0.8456, + "loss": 0.8755, "step": 986 }, { "epoch": 3.9959514170040484, - "grad_norm": 0.4175920659237699, + "grad_norm": 0.26899587122464425, "learning_rate": 4.719578267588725e-06, - "loss": 0.8363, + "loss": 0.867, "step": 987 }, { "epoch": 4.0, - "grad_norm": 0.4376486314386626, + "grad_norm": 0.23766171216966855, "learning_rate": 4.6831510513413525e-06, - "loss": 0.8186, + "loss": 0.8489, "step": 988 }, { "epoch": 4.004048582995951, - "grad_norm": 0.8991706727211234, + "grad_norm": 0.31498304144941597, "learning_rate": 4.6468463082100176e-06, - "loss": 0.73, + "loss": 0.8396, "step": 989 }, { "epoch": 4.008097165991903, - "grad_norm": 0.6994517159495758, + "grad_norm": 0.30688614672172926, "learning_rate": 4.610664328486485e-06, - "loss": 0.7157, + "loss": 0.8179, "step": 990 }, { "epoch": 4.012145748987854, - "grad_norm": 0.5276101737121127, + "grad_norm": 0.2814978591565573, "learning_rate": 4.5746054014809e-06, - "loss": 0.688, + "loss": 0.8172, "step": 991 }, { "epoch": 4.016194331983805, - "grad_norm": 0.8096899910738223, + "grad_norm": 0.29677583976782773, "learning_rate": 4.5386698155194766e-06, - "loss": 0.6947, + "loss": 0.8054, "step": 992 }, { "epoch": 4.020242914979757, - "grad_norm": 1.1157336088235812, + "grad_norm": 0.28009772420074747, "learning_rate": 4.502857857942204e-06, - "loss": 0.7259, + "loss": 0.8433, "step": 993 }, { "epoch": 4.0242914979757085, - "grad_norm": 0.868733299595644, + "grad_norm": 0.30906148532521294, "learning_rate": 4.467169815100549e-06, - "loss": 0.6963, + "loss": 0.8395, "step": 994 }, { "epoch": 4.02834008097166, - "grad_norm": 0.7469865028544674, + "grad_norm": 0.32608722110202676, "learning_rate": 4.431605972355137e-06, - "loss": 0.6927, + "loss": 0.8021, "step": 995 }, { "epoch": 4.032388663967612, - "grad_norm": 0.7641890904450588, + "grad_norm": 0.287473448927692, "learning_rate": 4.396166614073516e-06, - "loss": 0.6986, + "loss": 0.8137, "step": 996 }, { "epoch": 4.036437246963563, - "grad_norm": 0.6995280883533584, + "grad_norm": 0.2769438745261557, "learning_rate": 4.360852023627846e-06, - "loss": 0.6979, + "loss": 0.8021, "step": 997 }, { "epoch": 4.040485829959514, - "grad_norm": 0.6619221596155305, + "grad_norm": 0.32891010334259113, "learning_rate": 4.325662483392656e-06, - "loss": 0.711, + "loss": 0.8299, "step": 998 }, { "epoch": 4.044534412955466, - "grad_norm": 0.6810319357568299, + "grad_norm": 0.31482223835839696, "learning_rate": 4.29059827474257e-06, - "loss": 0.6972, + "loss": 0.8378, "step": 999 }, { "epoch": 4.048582995951417, - "grad_norm": 0.6090559467129231, + "grad_norm": 0.2806197330131218, "learning_rate": 4.255659678050063e-06, - "loss": 0.7163, + "loss": 0.8337, "step": 1000 }, { "epoch": 4.052631578947368, - "grad_norm": 0.6646232401534313, + "grad_norm": 0.32252935228773555, "learning_rate": 4.22084697268323e-06, - "loss": 0.6853, + "loss": 0.8138, "step": 1001 }, { "epoch": 4.05668016194332, - "grad_norm": 0.6095948202671398, + "grad_norm": 0.3323849352034766, "learning_rate": 4.186160437003535e-06, - "loss": 0.6787, + "loss": 0.7984, "step": 1002 }, { "epoch": 4.060728744939271, - "grad_norm": 0.5746990062245557, + "grad_norm": 0.29188197418017997, "learning_rate": 4.151600348363582e-06, - "loss": 0.6894, + "loss": 0.8125, "step": 1003 }, { "epoch": 4.064777327935222, - "grad_norm": 0.5975047838140005, + "grad_norm": 0.27434804007083996, "learning_rate": 4.11716698310493e-06, - "loss": 0.7087, + "loss": 0.8269, "step": 1004 }, { "epoch": 4.068825910931174, - "grad_norm": 0.6564402789910927, + "grad_norm": 0.3039775020897398, "learning_rate": 4.082860616555848e-06, - "loss": 0.6976, + "loss": 0.8278, "step": 1005 }, { "epoch": 4.0728744939271255, - "grad_norm": 0.6075010246120259, + "grad_norm": 0.31753649187707356, "learning_rate": 4.048681523029125e-06, - "loss": 0.6963, + "loss": 0.8162, "step": 1006 }, { "epoch": 4.076923076923077, - "grad_norm": 0.5347399836930657, + "grad_norm": 0.2631123495419154, "learning_rate": 4.014629975819882e-06, - "loss": 0.6907, + "loss": 0.8327, "step": 1007 }, { "epoch": 4.080971659919029, - "grad_norm": 0.5236807337305364, + "grad_norm": 0.2854412098370565, "learning_rate": 3.9807062472033855e-06, - "loss": 0.7007, + "loss": 0.84, "step": 1008 }, { "epoch": 4.08502024291498, - "grad_norm": 0.5704770263885252, + "grad_norm": 0.27476640353310644, "learning_rate": 3.94691060843287e-06, - "loss": 0.6699, + "loss": 0.8015, "step": 1009 }, { "epoch": 4.089068825910931, - "grad_norm": 0.5722568785392083, + "grad_norm": 0.28503886815924834, "learning_rate": 3.913243329737353e-06, - "loss": 0.7139, + "loss": 0.8365, "step": 1010 }, { "epoch": 4.093117408906883, - "grad_norm": 0.5321674988340522, + "grad_norm": 0.28206945453935045, "learning_rate": 3.879704680319503e-06, - "loss": 0.6796, + "loss": 0.8104, "step": 1011 }, { "epoch": 4.097165991902834, - "grad_norm": 0.48880277238657854, + "grad_norm": 0.25680141950222496, "learning_rate": 3.846294928353471e-06, - "loss": 0.7065, + "loss": 0.8354, "step": 1012 }, { "epoch": 4.101214574898785, - "grad_norm": 0.5105600007888825, + "grad_norm": 0.2515172555381191, "learning_rate": 3.8130143409827345e-06, - "loss": 0.7022, + "loss": 0.8398, "step": 1013 }, { "epoch": 4.105263157894737, - "grad_norm": 0.5377918440837455, + "grad_norm": 0.288307704583416, "learning_rate": 3.7798631843179887e-06, - "loss": 0.693, + "loss": 0.8106, "step": 1014 }, { "epoch": 4.109311740890688, - "grad_norm": 0.5172264132863608, + "grad_norm": 0.2783219823182209, "learning_rate": 3.7468417234349973e-06, - "loss": 0.6891, + "loss": 0.8324, "step": 1015 }, { "epoch": 4.113360323886639, - "grad_norm": 0.5228022901882626, + "grad_norm": 0.2658342951854686, "learning_rate": 3.7139502223724823e-06, - "loss": 0.7076, + "loss": 0.8455, "step": 1016 }, { "epoch": 4.117408906882591, - "grad_norm": 0.48830946622853655, + "grad_norm": 0.22684111874875146, "learning_rate": 3.6811889441300095e-06, - "loss": 0.6812, + "loss": 0.8236, "step": 1017 }, { "epoch": 4.1214574898785425, - "grad_norm": 0.49026026046639776, + "grad_norm": 0.2991036568173905, "learning_rate": 3.648558150665886e-06, - "loss": 0.6982, + "loss": 0.8464, "step": 1018 }, { "epoch": 4.125506072874494, - "grad_norm": 0.48965823108821604, + "grad_norm": 0.28366321758358376, "learning_rate": 3.616058102895075e-06, - "loss": 0.7, + "loss": 0.8269, "step": 1019 }, { "epoch": 4.129554655870446, - "grad_norm": 0.497231957551131, + "grad_norm": 0.24366615892613094, "learning_rate": 3.583689060687079e-06, - "loss": 0.6889, + "loss": 0.8138, "step": 1020 }, { "epoch": 4.133603238866397, - "grad_norm": 0.46061138397050233, + "grad_norm": 0.23543985600356815, "learning_rate": 3.5514512828639023e-06, - "loss": 0.6957, + "loss": 0.818, "step": 1021 }, { "epoch": 4.137651821862348, - "grad_norm": 0.46196650981139553, + "grad_norm": 0.23203396649412503, "learning_rate": 3.51934502719796e-06, - "loss": 0.6897, + "loss": 0.8177, "step": 1022 }, { "epoch": 4.1417004048583, - "grad_norm": 0.4964712995151658, + "grad_norm": 0.22919607525114957, "learning_rate": 3.4873705504100185e-06, - "loss": 0.7227, + "loss": 0.8389, "step": 1023 }, { "epoch": 4.145748987854251, - "grad_norm": 0.4570471592575044, + "grad_norm": 0.25153042313916124, "learning_rate": 3.4555281081671345e-06, - "loss": 0.6967, + "loss": 0.831, "step": 1024 }, { "epoch": 4.149797570850202, - "grad_norm": 0.458116591117372, + "grad_norm": 0.2380715278372012, "learning_rate": 3.4238179550806326e-06, - "loss": 0.6825, + "loss": 0.8186, "step": 1025 }, { "epoch": 4.153846153846154, - "grad_norm": 0.4674836289298726, + "grad_norm": 0.24724409022302438, "learning_rate": 3.3922403447040497e-06, - "loss": 0.6975, + "loss": 0.8441, "step": 1026 }, { "epoch": 4.157894736842105, - "grad_norm": 0.45287130331669795, + "grad_norm": 0.27258603869229664, "learning_rate": 3.3607955295311247e-06, - "loss": 0.6943, + "loss": 0.832, "step": 1027 }, { "epoch": 4.161943319838056, - "grad_norm": 0.4410875306842379, + "grad_norm": 0.23944620373831046, "learning_rate": 3.32948376099375e-06, - "loss": 0.6884, + "loss": 0.7984, "step": 1028 }, { "epoch": 4.165991902834008, - "grad_norm": 0.4662979147386088, + "grad_norm": 0.2211540346317158, "learning_rate": 3.2983052894599976e-06, - "loss": 0.6945, + "loss": 0.814, "step": 1029 }, { "epoch": 4.17004048582996, - "grad_norm": 0.4518975107930034, + "grad_norm": 0.2582977425436942, "learning_rate": 3.2672603642320945e-06, - "loss": 0.6931, + "loss": 0.8376, "step": 1030 }, { "epoch": 4.174089068825911, - "grad_norm": 0.4586073176920542, + "grad_norm": 0.2315433343007862, "learning_rate": 3.236349233544438e-06, - "loss": 0.7061, + "loss": 0.8172, "step": 1031 }, { "epoch": 4.178137651821863, - "grad_norm": 0.4452788645653796, + "grad_norm": 0.2165106910187865, "learning_rate": 3.205572144561604e-06, - "loss": 0.6817, + "loss": 0.7922, "step": 1032 }, { "epoch": 4.182186234817814, - "grad_norm": 0.444106102816619, + "grad_norm": 0.22797893083681345, "learning_rate": 3.174929343376374e-06, - "loss": 0.6842, + "loss": 0.7988, "step": 1033 }, { "epoch": 4.186234817813765, - "grad_norm": 0.4498288520021532, + "grad_norm": 0.2596913125704419, "learning_rate": 3.1444210750077777e-06, - "loss": 0.7052, + "loss": 0.8273, "step": 1034 }, { "epoch": 4.190283400809717, - "grad_norm": 0.4496904435133432, + "grad_norm": 0.24248126870445666, "learning_rate": 3.1140475833991046e-06, - "loss": 0.6954, + "loss": 0.8105, "step": 1035 }, { "epoch": 4.194331983805668, - "grad_norm": 0.4441023144388826, + "grad_norm": 0.22117606171942478, "learning_rate": 3.0838091114159895e-06, - "loss": 0.6926, + "loss": 0.81, "step": 1036 }, { "epoch": 4.198380566801619, - "grad_norm": 0.43331758691908157, + "grad_norm": 0.23788485236410326, "learning_rate": 3.0537059008444524e-06, - "loss": 0.6905, + "loss": 0.8137, "step": 1037 }, { "epoch": 4.202429149797571, - "grad_norm": 0.4575685644075248, + "grad_norm": 0.23807088717667976, "learning_rate": 3.0237381923889563e-06, - "loss": 0.6996, + "loss": 0.8271, "step": 1038 }, { "epoch": 4.206477732793522, - "grad_norm": 0.4409080486795876, + "grad_norm": 0.2530601069337019, "learning_rate": 2.9939062256704997e-06, - "loss": 0.6795, + "loss": 0.8242, "step": 1039 }, { "epoch": 4.2105263157894735, - "grad_norm": 0.44284805384940784, + "grad_norm": 0.22694844248142654, "learning_rate": 2.964210239224696e-06, - "loss": 0.6916, + "loss": 0.8348, "step": 1040 }, { "epoch": 4.2145748987854255, - "grad_norm": 0.4408982007413591, + "grad_norm": 0.22323275769981352, "learning_rate": 2.934650470499869e-06, - "loss": 0.6793, + "loss": 0.8033, "step": 1041 }, { "epoch": 4.218623481781377, - "grad_norm": 0.4659692396415994, + "grad_norm": 0.22268413790914812, "learning_rate": 2.905227155855128e-06, - "loss": 0.7172, + "loss": 0.8419, "step": 1042 }, { "epoch": 4.222672064777328, - "grad_norm": 0.44465713838002946, + "grad_norm": 0.22637545853323437, "learning_rate": 2.875940530558519e-06, - "loss": 0.6875, + "loss": 0.8253, "step": 1043 }, { "epoch": 4.22672064777328, - "grad_norm": 0.45310294698855014, + "grad_norm": 0.25735733213672174, "learning_rate": 2.8467908287851154e-06, - "loss": 0.6986, + "loss": 0.8134, "step": 1044 }, { "epoch": 4.230769230769231, - "grad_norm": 0.44749099616708765, + "grad_norm": 0.2539694650782424, "learning_rate": 2.8177782836151603e-06, - "loss": 0.7055, + "loss": 0.8275, "step": 1045 }, { "epoch": 4.234817813765182, - "grad_norm": 0.4390538115035109, + "grad_norm": 0.21536680626726754, "learning_rate": 2.7889031270321874e-06, - "loss": 0.7002, + "loss": 0.8125, "step": 1046 }, { "epoch": 4.238866396761134, - "grad_norm": 0.45675273083561657, + "grad_norm": 0.2167423018667985, "learning_rate": 2.760165589921184e-06, - "loss": 0.6987, + "loss": 0.8182, "step": 1047 }, { "epoch": 4.242914979757085, - "grad_norm": 0.45231486165822604, + "grad_norm": 0.22129183806897926, "learning_rate": 2.7315659020667305e-06, - "loss": 0.716, + "loss": 0.8372, "step": 1048 }, { "epoch": 4.246963562753036, - "grad_norm": 0.4579921216990953, + "grad_norm": 0.21187056279738653, "learning_rate": 2.703104292151164e-06, - "loss": 0.6984, + "loss": 0.8185, "step": 1049 }, { "epoch": 4.251012145748988, - "grad_norm": 0.4463681145112567, + "grad_norm": 0.21753583974969926, "learning_rate": 2.6747809877527607e-06, - "loss": 0.6872, + "loss": 0.8161, "step": 1050 }, { "epoch": 4.255060728744939, - "grad_norm": 0.44371929827432116, + "grad_norm": 0.2243870690077022, "learning_rate": 2.6465962153439082e-06, - "loss": 0.698, + "loss": 0.831, "step": 1051 }, { "epoch": 4.2591093117408905, - "grad_norm": 0.4460437799072536, + "grad_norm": 0.22027449099629887, "learning_rate": 2.618550200289303e-06, - "loss": 0.6917, + "loss": 0.8265, "step": 1052 }, { "epoch": 4.2631578947368425, - "grad_norm": 0.4440680674244864, + "grad_norm": 0.22005143758613824, "learning_rate": 2.59064316684412e-06, - "loss": 0.6808, + "loss": 0.8124, "step": 1053 }, { "epoch": 4.267206477732794, - "grad_norm": 0.44360866626575524, + "grad_norm": 0.21036161834844672, "learning_rate": 2.562875338152264e-06, - "loss": 0.6775, + "loss": 0.7939, "step": 1054 }, { "epoch": 4.271255060728745, - "grad_norm": 0.44309501121825956, + "grad_norm": 0.23242011482480293, "learning_rate": 2.5352469362445552e-06, - "loss": 0.6728, + "loss": 0.8048, "step": 1055 }, { "epoch": 4.275303643724697, - "grad_norm": 0.45205767632882893, + "grad_norm": 0.23710861685519863, "learning_rate": 2.5077581820369523e-06, - "loss": 0.7119, + "loss": 0.8286, "step": 1056 }, { "epoch": 4.279352226720648, - "grad_norm": 0.46011391544247177, + "grad_norm": 0.2346261436521955, "learning_rate": 2.480409295328803e-06, - "loss": 0.7101, + "loss": 0.8273, "step": 1057 }, { "epoch": 4.283400809716599, - "grad_norm": 0.4358779468272973, + "grad_norm": 0.23327776541413672, "learning_rate": 2.45320049480108e-06, - "loss": 0.6828, + "loss": 0.8153, "step": 1058 }, { "epoch": 4.287449392712551, - "grad_norm": 0.44589424677993544, + "grad_norm": 0.2202336333949928, "learning_rate": 2.4261319980146293e-06, - "loss": 0.7281, + "loss": 0.8331, "step": 1059 }, { "epoch": 4.291497975708502, - "grad_norm": 0.42281316947637065, + "grad_norm": 0.22231002011620007, "learning_rate": 2.399204021408421e-06, - "loss": 0.6882, + "loss": 0.8155, "step": 1060 }, { "epoch": 4.295546558704453, - "grad_norm": 0.44017663551409986, + "grad_norm": 0.24398929493554108, "learning_rate": 2.3724167802978395e-06, - "loss": 0.6991, + "loss": 0.8313, "step": 1061 }, { "epoch": 4.299595141700405, - "grad_norm": 0.43352777833251915, + "grad_norm": 0.21114782345902308, "learning_rate": 2.3457704888729625e-06, - "loss": 0.6872, + "loss": 0.8045, "step": 1062 }, { "epoch": 4.303643724696356, - "grad_norm": 0.44649814468631754, + "grad_norm": 0.20554255987540446, "learning_rate": 2.3192653601968162e-06, - "loss": 0.7013, + "loss": 0.8376, "step": 1063 }, { "epoch": 4.3076923076923075, - "grad_norm": 0.44439925078204817, + "grad_norm": 0.20554097414755199, "learning_rate": 2.2929016062037057e-06, - "loss": 0.6935, + "loss": 0.818, "step": 1064 }, { "epoch": 4.3117408906882595, - "grad_norm": 0.43782563931912105, + "grad_norm": 0.2045680450929699, "learning_rate": 2.2666794376975055e-06, - "loss": 0.6797, + "loss": 0.8139, "step": 1065 }, { "epoch": 4.315789473684211, - "grad_norm": 0.45241406261310524, + "grad_norm": 0.22558650322684756, "learning_rate": 2.24059906434998e-06, - "loss": 0.6819, + "loss": 0.8259, "step": 1066 }, { "epoch": 4.319838056680162, - "grad_norm": 0.4531216599688645, + "grad_norm": 0.21375628189965765, "learning_rate": 2.214660694699091e-06, - "loss": 0.7065, + "loss": 0.8439, "step": 1067 }, { "epoch": 4.323886639676114, - "grad_norm": 0.4522792603643063, + "grad_norm": 0.2251257610645498, "learning_rate": 2.188864536147355e-06, - "loss": 0.6809, + "loss": 0.8094, "step": 1068 }, { "epoch": 4.327935222672065, - "grad_norm": 0.43133090407109403, + "grad_norm": 0.2325180915604801, "learning_rate": 2.1632107949601757e-06, - "loss": 0.6872, + "loss": 0.8155, "step": 1069 }, { "epoch": 4.331983805668016, - "grad_norm": 0.44388930298401136, + "grad_norm": 0.21569350297936452, "learning_rate": 2.137699676264171e-06, - "loss": 0.6952, + "loss": 0.8156, "step": 1070 }, { "epoch": 4.336032388663968, - "grad_norm": 0.4395922897814348, + "grad_norm": 0.20588721941187735, "learning_rate": 2.1123313840455738e-06, - "loss": 0.6967, + "loss": 0.8265, "step": 1071 }, { "epoch": 4.340080971659919, - "grad_norm": 0.43989136323243966, + "grad_norm": 0.22548834357094483, "learning_rate": 2.08710612114857e-06, - "loss": 0.7063, + "loss": 0.8252, "step": 1072 }, { "epoch": 4.34412955465587, - "grad_norm": 0.4275919646292657, + "grad_norm": 0.21807920255180194, "learning_rate": 2.0620240892736953e-06, - "loss": 0.6907, + "loss": 0.8132, "step": 1073 }, { "epoch": 4.348178137651822, - "grad_norm": 0.4362318802997258, + "grad_norm": 0.21697117226048962, "learning_rate": 2.0370854889762e-06, - "loss": 0.6933, + "loss": 0.8206, "step": 1074 }, { "epoch": 4.352226720647773, - "grad_norm": 0.44072692723570683, + "grad_norm": 0.2501122411479991, "learning_rate": 2.0122905196644726e-06, - "loss": 0.6886, + "loss": 0.8279, "step": 1075 }, { "epoch": 4.3562753036437245, - "grad_norm": 0.4424228000971391, + "grad_norm": 0.21562793732892602, "learning_rate": 1.987639379598427e-06, - "loss": 0.7005, + "loss": 0.822, "step": 1076 }, { "epoch": 4.3603238866396765, - "grad_norm": 0.4292580169732448, + "grad_norm": 0.20776931119187292, "learning_rate": 1.963132265887919e-06, - "loss": 0.6825, + "loss": 0.8034, "step": 1077 }, { "epoch": 4.364372469635628, - "grad_norm": 0.4398431455674593, + "grad_norm": 0.22006442560800757, "learning_rate": 1.9387693744911784e-06, - "loss": 0.7025, + "loss": 0.8232, "step": 1078 }, { "epoch": 4.368421052631579, - "grad_norm": 0.42821196282830076, + "grad_norm": 0.21571586725960543, "learning_rate": 1.9145509002132368e-06, - "loss": 0.6926, + "loss": 0.8082, "step": 1079 }, { "epoch": 4.372469635627531, - "grad_norm": 0.42936298850345583, + "grad_norm": 0.21739931200629528, "learning_rate": 1.8904770367043723e-06, - "loss": 0.6866, + "loss": 0.816, "step": 1080 }, { "epoch": 4.376518218623482, - "grad_norm": 0.4398368750432568, + "grad_norm": 0.2199861851255427, "learning_rate": 1.8665479764585504e-06, - "loss": 0.7063, + "loss": 0.8202, "step": 1081 }, { "epoch": 4.380566801619433, - "grad_norm": 0.43344880408549546, + "grad_norm": 0.20731806034467812, "learning_rate": 1.8427639108118977e-06, - "loss": 0.6972, + "loss": 0.8401, "step": 1082 }, { "epoch": 4.384615384615385, - "grad_norm": 0.42982375339674633, + "grad_norm": 0.21999031655629087, "learning_rate": 1.8191250299411755e-06, - "loss": 0.679, + "loss": 0.8067, "step": 1083 }, { "epoch": 4.388663967611336, - "grad_norm": 0.4462699775611063, + "grad_norm": 0.20730095535085977, "learning_rate": 1.7956315228622446e-06, - "loss": 0.7144, + "loss": 0.8289, "step": 1084 }, { "epoch": 4.392712550607287, - "grad_norm": 0.4306594217285721, + "grad_norm": 0.22197956618806458, "learning_rate": 1.7722835774285553e-06, - "loss": 0.6979, + "loss": 0.8262, "step": 1085 }, { "epoch": 4.396761133603239, - "grad_norm": 0.43165462616496375, + "grad_norm": 0.2245472496291921, "learning_rate": 1.7490813803296624e-06, - "loss": 0.6895, + "loss": 0.8146, "step": 1086 }, { "epoch": 4.40080971659919, - "grad_norm": 0.4392872827848514, + "grad_norm": 0.23153476357099875, "learning_rate": 1.726025117089718e-06, - "loss": 0.6937, + "loss": 0.8276, "step": 1087 }, { "epoch": 4.4048582995951415, - "grad_norm": 0.4281726575936688, + "grad_norm": 0.21410677768121758, "learning_rate": 1.70311497206598e-06, - "loss": 0.6905, + "loss": 0.818, "step": 1088 }, { "epoch": 4.4089068825910935, - "grad_norm": 0.43323406711636286, + "grad_norm": 0.23858134454838378, "learning_rate": 1.6803511284473638e-06, - "loss": 0.6856, + "loss": 0.797, "step": 1089 }, { "epoch": 4.412955465587045, - "grad_norm": 0.42178358504130914, + "grad_norm": 0.21132578039218852, "learning_rate": 1.6577337682529538e-06, - "loss": 0.6938, + "loss": 0.8028, "step": 1090 }, { "epoch": 4.417004048582996, - "grad_norm": 0.4280855370525906, + "grad_norm": 0.22962540909183005, "learning_rate": 1.6352630723305617e-06, - "loss": 0.6952, + "loss": 0.8262, "step": 1091 }, { "epoch": 4.421052631578947, - "grad_norm": 0.4397070861611521, + "grad_norm": 0.21509185966726724, "learning_rate": 1.612939220355272e-06, - "loss": 0.6904, + "loss": 0.8239, "step": 1092 }, { "epoch": 4.425101214574899, - "grad_norm": 0.43935896158651344, + "grad_norm": 0.22402947376315385, "learning_rate": 1.5907623908280112e-06, - "loss": 0.7075, + "loss": 0.828, "step": 1093 }, { "epoch": 4.42914979757085, - "grad_norm": 0.43597776114043907, + "grad_norm": 0.2140174761929923, "learning_rate": 1.568732761074121e-06, - "loss": 0.6895, + "loss": 0.825, "step": 1094 }, { "epoch": 4.433198380566802, - "grad_norm": 0.43921315860437854, + "grad_norm": 0.21548022753328144, "learning_rate": 1.5468505072419237e-06, - "loss": 0.7053, + "loss": 0.8202, "step": 1095 }, { "epoch": 4.437246963562753, - "grad_norm": 0.4338178390165574, + "grad_norm": 0.2025415815760108, "learning_rate": 1.5251158043013448e-06, - "loss": 0.7156, + "loss": 0.8328, "step": 1096 }, { "epoch": 4.441295546558704, - "grad_norm": 0.43128027778377975, + "grad_norm": 0.2139864191869674, "learning_rate": 1.5035288260424885e-06, - "loss": 0.6956, + "loss": 0.8128, "step": 1097 }, { "epoch": 4.445344129554655, - "grad_norm": 0.43117786033940825, + "grad_norm": 0.2453791001851767, "learning_rate": 1.4820897450742623e-06, - "loss": 0.675, + "loss": 0.8008, "step": 1098 }, { "epoch": 4.449392712550607, - "grad_norm": 0.43772231927154037, + "grad_norm": 0.22595075686359892, "learning_rate": 1.4607987328229767e-06, - "loss": 0.6862, + "loss": 0.8108, "step": 1099 }, { "epoch": 4.4534412955465585, - "grad_norm": 0.4408428556274507, + "grad_norm": 0.21517515721635427, "learning_rate": 1.4396559595310056e-06, - "loss": 0.6985, + "loss": 0.8191, "step": 1100 }, { "epoch": 4.4574898785425106, - "grad_norm": 0.44629798614381044, + "grad_norm": 0.21426000984281948, "learning_rate": 1.4186615942553973e-06, - "loss": 0.7056, + "loss": 0.8375, "step": 1101 }, { "epoch": 4.461538461538462, - "grad_norm": 0.4248748275515192, + "grad_norm": 0.21229248156711378, "learning_rate": 1.3978158048665424e-06, - "loss": 0.6893, + "loss": 0.8154, "step": 1102 }, { "epoch": 4.465587044534413, - "grad_norm": 0.4222203442774602, + "grad_norm": 0.21630768870306924, "learning_rate": 1.3771187580468115e-06, - "loss": 0.6988, + "loss": 0.8195, "step": 1103 }, { "epoch": 4.469635627530364, - "grad_norm": 0.4193103529507362, + "grad_norm": 0.2107626339868905, "learning_rate": 1.3565706192892392e-06, - "loss": 0.6847, + "loss": 0.7925, "step": 1104 }, { "epoch": 4.473684210526316, - "grad_norm": 0.4330854379145975, + "grad_norm": 0.21526719136948813, "learning_rate": 1.3361715528961993e-06, - "loss": 0.7074, + "loss": 0.8505, "step": 1105 }, { "epoch": 4.477732793522267, - "grad_norm": 0.4335083260025143, + "grad_norm": 0.2234048006830192, "learning_rate": 1.3159217219780817e-06, - "loss": 0.6901, + "loss": 0.8246, "step": 1106 }, { "epoch": 4.481781376518219, - "grad_norm": 0.43093110361276094, + "grad_norm": 0.2168048129644934, "learning_rate": 1.2958212884519949e-06, - "loss": 0.6928, + "loss": 0.8284, "step": 1107 }, { "epoch": 4.48582995951417, - "grad_norm": 0.4425202475933826, + "grad_norm": 0.20318076361698195, "learning_rate": 1.2758704130404675e-06, - "loss": 0.6858, + "loss": 0.8299, "step": 1108 }, { "epoch": 4.489878542510121, - "grad_norm": 0.44324354093300555, + "grad_norm": 0.205261640455371, "learning_rate": 1.2560692552701738e-06, - "loss": 0.7104, + "loss": 0.8282, "step": 1109 }, { "epoch": 4.493927125506072, - "grad_norm": 0.43633388368696235, + "grad_norm": 0.20218815342477314, "learning_rate": 1.2364179734706382e-06, - "loss": 0.6963, + "loss": 0.8329, "step": 1110 }, { "epoch": 4.497975708502024, - "grad_norm": 0.431056145978849, + "grad_norm": 0.2133462317172052, "learning_rate": 1.2169167247729852e-06, - "loss": 0.7048, + "loss": 0.8361, "step": 1111 }, { "epoch": 4.502024291497976, - "grad_norm": 0.42768029613873443, + "grad_norm": 0.2074101069411439, "learning_rate": 1.1975656651086864e-06, - "loss": 0.6874, + "loss": 0.833, "step": 1112 }, { "epoch": 4.506072874493928, - "grad_norm": 0.4173657208085509, + "grad_norm": 0.2278815899836169, "learning_rate": 1.1783649492082926e-06, - "loss": 0.6822, + "loss": 0.8181, "step": 1113 }, { "epoch": 4.510121457489879, - "grad_norm": 0.42873437968746114, + "grad_norm": 0.20222761336041423, "learning_rate": 1.1593147306002183e-06, - "loss": 0.689, + "loss": 0.8117, "step": 1114 }, { "epoch": 4.51417004048583, - "grad_norm": 0.4445780645978198, + "grad_norm": 0.24068529275024628, "learning_rate": 1.140415161609505e-06, - "loss": 0.7013, + "loss": 0.8194, "step": 1115 }, { "epoch": 4.518218623481781, - "grad_norm": 0.4332219886748684, + "grad_norm": 0.22025040045963223, "learning_rate": 1.121666393356604e-06, - "loss": 0.7019, + "loss": 0.8272, "step": 1116 }, { "epoch": 4.522267206477733, - "grad_norm": 0.42736534854914804, + "grad_norm": 0.2127685679349505, "learning_rate": 1.1030685757561632e-06, - "loss": 0.687, + "loss": 0.8114, "step": 1117 }, { "epoch": 4.526315789473684, - "grad_norm": 0.421043845685808, + "grad_norm": 0.20070504978662512, "learning_rate": 1.0846218575158373e-06, - "loss": 0.6937, + "loss": 0.8291, "step": 1118 }, { "epoch": 4.530364372469636, - "grad_norm": 0.43740121243418434, + "grad_norm": 0.20905285795905265, "learning_rate": 1.066326386135097e-06, - "loss": 0.7072, + "loss": 0.8296, "step": 1119 }, { "epoch": 4.534412955465587, - "grad_norm": 0.4345396842761784, + "grad_norm": 0.20246749587570545, "learning_rate": 1.0481823079040398e-06, - "loss": 0.6852, + "loss": 0.8272, "step": 1120 }, { "epoch": 4.538461538461538, - "grad_norm": 0.43272209661766875, + "grad_norm": 0.20363691549345542, "learning_rate": 1.0301897679022344e-06, - "loss": 0.6798, + "loss": 0.8195, "step": 1121 }, { "epoch": 4.5425101214574894, - "grad_norm": 0.4490362461530953, + "grad_norm": 0.19733337415300373, "learning_rate": 1.0123489099975491e-06, - "loss": 0.7119, + "loss": 0.8297, "step": 1122 }, { "epoch": 4.5465587044534415, - "grad_norm": 0.43799608537818274, + "grad_norm": 0.20917923837339455, "learning_rate": 9.9465987684501e-07, - "loss": 0.6848, + "loss": 0.8067, "step": 1123 }, { "epoch": 4.550607287449393, - "grad_norm": 0.42594178829964324, + "grad_norm": 0.21542430224340822, "learning_rate": 9.771228098856534e-07, - "loss": 0.698, + "loss": 0.8233, "step": 1124 }, { "epoch": 4.554655870445345, - "grad_norm": 0.4383667153976398, + "grad_norm": 0.2026896422320593, "learning_rate": 9.597378493454013e-07, - "loss": 0.7092, + "loss": 0.8475, "step": 1125 }, { "epoch": 4.558704453441296, - "grad_norm": 0.43752461105771145, + "grad_norm": 0.1984601602854306, "learning_rate": 9.425051342339353e-07, - "loss": 0.7021, + "loss": 0.8249, "step": 1126 }, { "epoch": 4.562753036437247, - "grad_norm": 0.4318158096764484, + "grad_norm": 0.20708561314879084, "learning_rate": 9.254248023435864e-07, - "loss": 0.6925, + "loss": 0.8445, "step": 1127 }, { "epoch": 4.566801619433198, - "grad_norm": 0.42973262017197206, + "grad_norm": 0.21446289044667288, "learning_rate": 9.084969902482288e-07, - "loss": 0.6793, + "loss": 0.8123, "step": 1128 }, { "epoch": 4.57085020242915, - "grad_norm": 0.4223628570772846, + "grad_norm": 0.2210376680426339, "learning_rate": 8.917218333021993e-07, - "loss": 0.6938, + "loss": 0.8173, "step": 1129 }, { "epoch": 4.574898785425101, - "grad_norm": 0.4414364446669952, + "grad_norm": 0.20461204934241708, "learning_rate": 8.750994656392109e-07, - "loss": 0.6935, + "loss": 0.8173, "step": 1130 }, { "epoch": 4.578947368421053, - "grad_norm": 0.4477306992265067, + "grad_norm": 0.20887885153787342, "learning_rate": 8.58630020171265e-07, - "loss": 0.7065, + "loss": 0.8335, "step": 1131 }, { "epoch": 4.582995951417004, - "grad_norm": 0.4410482312945271, + "grad_norm": 0.2068788884872065, "learning_rate": 8.423136285876099e-07, - "loss": 0.6997, + "loss": 0.8259, "step": 1132 }, { "epoch": 4.587044534412955, - "grad_norm": 0.42563971966033776, + "grad_norm": 0.2037797315054344, "learning_rate": 8.26150421353682e-07, - "loss": 0.6958, + "loss": 0.8043, "step": 1133 }, { "epoch": 4.5910931174089065, - "grad_norm": 0.4291236476386049, + "grad_norm": 0.202698192674187, "learning_rate": 8.101405277100549e-07, - "loss": 0.6907, + "loss": 0.8213, "step": 1134 }, { "epoch": 4.5951417004048585, - "grad_norm": 0.4488993793622574, + "grad_norm": 0.20887260727353932, "learning_rate": 7.942840756714077e-07, - "loss": 0.7188, + "loss": 0.8377, "step": 1135 }, { "epoch": 4.59919028340081, - "grad_norm": 0.4389951096838271, + "grad_norm": 0.1931600715227766, "learning_rate": 7.785811920255093e-07, - "loss": 0.6934, + "loss": 0.8465, "step": 1136 }, { "epoch": 4.603238866396762, - "grad_norm": 0.4212600442563057, + "grad_norm": 0.19902683867741028, "learning_rate": 7.630320023322024e-07, - "loss": 0.6924, + "loss": 0.8311, "step": 1137 }, { "epoch": 4.607287449392713, - "grad_norm": 0.4366604493266147, + "grad_norm": 0.20735000266101647, "learning_rate": 7.476366309223903e-07, - "loss": 0.6984, + "loss": 0.824, "step": 1138 }, { "epoch": 4.611336032388664, - "grad_norm": 0.4183443145347396, + "grad_norm": 0.22194861299679935, "learning_rate": 7.323952008970537e-07, - "loss": 0.6925, + "loss": 0.8128, "step": 1139 }, { "epoch": 4.615384615384615, - "grad_norm": 0.4349084896884154, + "grad_norm": 0.2118065092394025, "learning_rate": 7.173078341262618e-07, - "loss": 0.6833, + "loss": 0.8186, "step": 1140 }, { "epoch": 4.619433198380567, - "grad_norm": 0.4402833606053142, + "grad_norm": 0.20427023040856718, "learning_rate": 7.023746512481989e-07, - "loss": 0.7049, + "loss": 0.8351, "step": 1141 }, { "epoch": 4.623481781376518, - "grad_norm": 0.4142396550315283, + "grad_norm": 0.20505708330437128, "learning_rate": 6.875957716681902e-07, - "loss": 0.6981, + "loss": 0.8168, "step": 1142 }, { "epoch": 4.62753036437247, - "grad_norm": 0.4439537076839186, + "grad_norm": 0.20271221525804145, "learning_rate": 6.729713135577687e-07, - "loss": 0.6902, + "loss": 0.816, "step": 1143 }, { "epoch": 4.631578947368421, - "grad_norm": 0.42037400720974544, + "grad_norm": 0.19354113624182362, "learning_rate": 6.585013938537143e-07, - "loss": 0.7017, + "loss": 0.827, "step": 1144 }, { "epoch": 4.635627530364372, - "grad_norm": 0.42188380668125586, + "grad_norm": 0.20121255519994796, "learning_rate": 6.441861282571115e-07, - "loss": 0.695, + "loss": 0.8189, "step": 1145 }, { "epoch": 4.6396761133603235, - "grad_norm": 0.42891260064186326, + "grad_norm": 0.1966882909824288, "learning_rate": 6.300256312324471e-07, - "loss": 0.7033, + "loss": 0.8366, "step": 1146 }, { "epoch": 4.6437246963562755, - "grad_norm": 0.42310387650162046, + "grad_norm": 0.20762807575947087, "learning_rate": 6.160200160066776e-07, - "loss": 0.6962, + "loss": 0.8171, "step": 1147 }, { "epoch": 4.647773279352227, - "grad_norm": 0.43603610746178734, + "grad_norm": 0.20076946011114039, "learning_rate": 6.021693945683327e-07, - "loss": 0.704, + "loss": 0.8237, "step": 1148 }, { "epoch": 4.651821862348179, - "grad_norm": 0.4364794827628862, + "grad_norm": 0.2021105282668505, "learning_rate": 5.884738776666044e-07, - "loss": 0.6825, + "loss": 0.8238, "step": 1149 }, { "epoch": 4.65587044534413, - "grad_norm": 0.4188191360984303, + "grad_norm": 0.20351321290750227, "learning_rate": 5.749335748104878e-07, - "loss": 0.7101, + "loss": 0.8448, "step": 1150 }, { "epoch": 4.659919028340081, - "grad_norm": 0.43763614157521374, + "grad_norm": 0.20625773833370545, "learning_rate": 5.615485942678778e-07, - "loss": 0.711, + "loss": 0.8486, "step": 1151 }, { "epoch": 4.663967611336032, - "grad_norm": 0.4172948492635435, + "grad_norm": 0.187885878341998, "learning_rate": 5.483190430647289e-07, - "loss": 0.6864, + "loss": 0.8101, "step": 1152 }, { "epoch": 4.668016194331984, - "grad_norm": 0.43260761996178254, + "grad_norm": 0.19730327516960433, "learning_rate": 5.352450269841747e-07, - "loss": 0.697, + "loss": 0.8319, "step": 1153 }, { "epoch": 4.672064777327935, - "grad_norm": 0.4380307392532172, + "grad_norm": 0.19808398238319463, "learning_rate": 5.22326650565701e-07, - "loss": 0.6818, + "loss": 0.8255, "step": 1154 }, { "epoch": 4.676113360323887, - "grad_norm": 0.42684872322102374, + "grad_norm": 0.21413566112760696, "learning_rate": 5.095640171043048e-07, - "loss": 0.7044, + "loss": 0.8166, "step": 1155 }, { "epoch": 4.680161943319838, - "grad_norm": 0.4205242441676326, + "grad_norm": 0.19066261089283612, "learning_rate": 4.969572286496594e-07, - "loss": 0.6976, + "loss": 0.8243, "step": 1156 }, { "epoch": 4.684210526315789, - "grad_norm": 0.42963727259945966, + "grad_norm": 0.20409010053350415, "learning_rate": 4.845063860053079e-07, - "loss": 0.7083, + "loss": 0.828, "step": 1157 }, { "epoch": 4.6882591093117405, - "grad_norm": 0.426733863606493, + "grad_norm": 0.18984365469720085, "learning_rate": 4.7221158872785997e-07, - "loss": 0.6798, + "loss": 0.8278, "step": 1158 }, { "epoch": 4.6923076923076925, - "grad_norm": 0.4300328945818575, + "grad_norm": 0.20304592598978408, "learning_rate": 4.600729351261857e-07, - "loss": 0.6935, + "loss": 0.8331, "step": 1159 }, { "epoch": 4.696356275303644, - "grad_norm": 0.42492137710233713, + "grad_norm": 0.19499891239194733, "learning_rate": 4.4809052226063355e-07, - "loss": 0.6943, + "loss": 0.8149, "step": 1160 }, { "epoch": 4.700404858299595, - "grad_norm": 0.4226336609003332, + "grad_norm": 0.19295319784130172, "learning_rate": 4.3626444594226046e-07, - "loss": 0.6918, + "loss": 0.8268, "step": 1161 }, { "epoch": 4.704453441295547, - "grad_norm": 0.42574521223643347, + "grad_norm": 0.19950694297195293, "learning_rate": 4.245948007320588e-07, - "loss": 0.6747, + "loss": 0.7967, "step": 1162 }, { "epoch": 4.708502024291498, - "grad_norm": 0.4185019724471728, + "grad_norm": 0.19928763523977552, "learning_rate": 4.130816799401971e-07, - "loss": 0.6877, + "loss": 0.8122, "step": 1163 }, { "epoch": 4.712550607287449, - "grad_norm": 0.4164344179266747, + "grad_norm": 0.20274118567905397, "learning_rate": 4.017251756252827e-07, - "loss": 0.6835, + "loss": 0.8008, "step": 1164 }, { "epoch": 4.716599190283401, - "grad_norm": 0.42837646770732285, + "grad_norm": 0.20478768948691245, "learning_rate": 3.9052537859362473e-07, - "loss": 0.7009, + "loss": 0.8366, "step": 1165 }, { "epoch": 4.720647773279352, - "grad_norm": 0.42708598544358534, + "grad_norm": 0.20560320558862047, "learning_rate": 3.7948237839849687e-07, - "loss": 0.685, + "loss": 0.8275, "step": 1166 }, { "epoch": 4.724696356275303, - "grad_norm": 0.4362016509654501, + "grad_norm": 0.20428608479651042, "learning_rate": 3.6859626333943797e-07, - "loss": 0.7038, + "loss": 0.8438, "step": 1167 }, { "epoch": 4.728744939271255, - "grad_norm": 0.4263933761029795, + "grad_norm": 0.20459370094007687, "learning_rate": 3.578671204615347e-07, - "loss": 0.6914, + "loss": 0.8135, "step": 1168 }, { "epoch": 4.732793522267206, - "grad_norm": 0.4344192405503701, + "grad_norm": 0.20059576651116576, "learning_rate": 3.472950355547267e-07, - "loss": 0.7132, + "loss": 0.8359, "step": 1169 }, { "epoch": 4.7368421052631575, - "grad_norm": 0.43623880192614234, + "grad_norm": 0.19584025780228306, "learning_rate": 3.3688009315312244e-07, - "loss": 0.6931, + "loss": 0.8262, "step": 1170 }, { "epoch": 4.7408906882591095, - "grad_norm": 0.422317142070649, + "grad_norm": 0.1923572555084818, "learning_rate": 3.266223765343224e-07, - "loss": 0.6975, + "loss": 0.8151, "step": 1171 }, { "epoch": 4.744939271255061, - "grad_norm": 0.4373059582208326, + "grad_norm": 0.19988680141611362, "learning_rate": 3.16521967718757e-07, - "loss": 0.7068, + "loss": 0.8395, "step": 1172 }, { "epoch": 4.748987854251012, - "grad_norm": 0.4239548896551374, + "grad_norm": 0.20522951666988712, "learning_rate": 3.0657894746902726e-07, - "loss": 0.7105, + "loss": 0.8271, "step": 1173 }, { "epoch": 4.753036437246964, - "grad_norm": 0.42761328593230546, + "grad_norm": 0.2006351851628511, "learning_rate": 2.9679339528925654e-07, - "loss": 0.6813, + "loss": 0.816, "step": 1174 }, { "epoch": 4.757085020242915, - "grad_norm": 0.4250506801701118, + "grad_norm": 0.1974974031678836, "learning_rate": 2.871653894244597e-07, - "loss": 0.6828, + "loss": 0.7973, "step": 1175 }, { "epoch": 4.761133603238866, - "grad_norm": 0.42268757665637513, + "grad_norm": 0.1989467070784958, "learning_rate": 2.7769500685991266e-07, - "loss": 0.675, + "loss": 0.816, "step": 1176 }, { "epoch": 4.765182186234818, - "grad_norm": 0.41686873407330055, + "grad_norm": 0.20109615635696884, "learning_rate": 2.683823233205485e-07, - "loss": 0.6939, + "loss": 0.8338, "step": 1177 }, { "epoch": 4.769230769230769, - "grad_norm": 0.422514442255524, + "grad_norm": 0.18965870938445314, "learning_rate": 2.592274132703287e-07, - "loss": 0.705, + "loss": 0.832, "step": 1178 }, { "epoch": 4.77327935222672, - "grad_norm": 0.4264998678209214, + "grad_norm": 0.20528394300408795, "learning_rate": 2.502303499116754e-07, - "loss": 0.6894, + "loss": 0.8075, "step": 1179 }, { "epoch": 4.777327935222672, - "grad_norm": 0.4312301979353998, + "grad_norm": 0.19061803193962018, "learning_rate": 2.413912051848666e-07, - "loss": 0.692, + "loss": 0.8204, "step": 1180 }, { "epoch": 4.781376518218623, - "grad_norm": 0.4264686352741417, + "grad_norm": 0.20252028267854927, "learning_rate": 2.3271004976746613e-07, - "loss": 0.7016, + "loss": 0.838, "step": 1181 }, { "epoch": 4.7854251012145745, - "grad_norm": 0.4245554971592264, + "grad_norm": 0.19165288122748222, "learning_rate": 2.2418695307376393e-07, - "loss": 0.6961, + "loss": 0.8322, "step": 1182 }, { "epoch": 4.7894736842105265, - "grad_norm": 0.4152853686814947, + "grad_norm": 0.19266699350752967, "learning_rate": 2.1582198325420968e-07, - "loss": 0.6787, + "loss": 0.8196, "step": 1183 }, { "epoch": 4.793522267206478, - "grad_norm": 0.43315071492637874, + "grad_norm": 0.195589076548436, "learning_rate": 2.0761520719488447e-07, - "loss": 0.6975, + "loss": 0.8337, "step": 1184 }, { "epoch": 4.797570850202429, - "grad_norm": 0.4310833297803439, + "grad_norm": 0.20551180046051526, "learning_rate": 1.9956669051694798e-07, - "loss": 0.7033, + "loss": 0.8391, "step": 1185 }, { "epoch": 4.801619433198381, - "grad_norm": 0.4386762803153357, + "grad_norm": 0.1874768081678635, "learning_rate": 1.916764975761254e-07, - "loss": 0.7207, + "loss": 0.834, "step": 1186 }, { "epoch": 4.805668016194332, - "grad_norm": 0.4200630311528745, + "grad_norm": 0.18422206268542277, "learning_rate": 1.839446914621923e-07, - "loss": 0.6897, + "loss": 0.8078, "step": 1187 }, { "epoch": 4.809716599190283, - "grad_norm": 0.4352782537808006, + "grad_norm": 0.21449196831549638, "learning_rate": 1.7637133399845962e-07, - "loss": 0.7103, + "loss": 0.8399, "step": 1188 }, { "epoch": 4.813765182186235, - "grad_norm": 0.42444277148345494, + "grad_norm": 0.19906238568429183, "learning_rate": 1.6895648574129397e-07, - "loss": 0.6953, + "loss": 0.837, "step": 1189 }, { "epoch": 4.817813765182186, - "grad_norm": 0.42728429306948845, + "grad_norm": 0.19474508929129228, "learning_rate": 1.6170020597962465e-07, - "loss": 0.7012, + "loss": 0.821, "step": 1190 }, { "epoch": 4.821862348178137, - "grad_norm": 0.42624856451485, + "grad_norm": 0.2173169435195749, "learning_rate": 1.5460255273447078e-07, - "loss": 0.685, + "loss": 0.8106, "step": 1191 }, { "epoch": 4.825910931174089, - "grad_norm": 0.4151624752050446, + "grad_norm": 0.2022900649409234, "learning_rate": 1.4766358275847715e-07, - "loss": 0.6923, + "loss": 0.8117, "step": 1192 }, { "epoch": 4.82995951417004, - "grad_norm": 0.4199466488639303, + "grad_norm": 0.18854724388053937, "learning_rate": 1.4088335153546573e-07, - "loss": 0.6896, + "loss": 0.8127, "step": 1193 }, { "epoch": 4.834008097165992, - "grad_norm": 0.4298296207388573, + "grad_norm": 0.19975231366600021, "learning_rate": 1.3426191327998496e-07, - "loss": 0.71, + "loss": 0.8368, "step": 1194 }, { "epoch": 4.838056680161944, - "grad_norm": 0.43832950296961604, + "grad_norm": 0.20885260857986004, "learning_rate": 1.277993209368744e-07, - "loss": 0.6836, + "loss": 0.8409, "step": 1195 }, { "epoch": 4.842105263157895, - "grad_norm": 0.424314195177959, + "grad_norm": 0.19420986099478404, "learning_rate": 1.214956261808542e-07, - "loss": 0.6877, + "loss": 0.8251, "step": 1196 }, { "epoch": 4.846153846153846, - "grad_norm": 0.41965184901784974, + "grad_norm": 0.2010517901873058, "learning_rate": 1.1535087941609846e-07, - "loss": 0.7037, + "loss": 0.8194, "step": 1197 }, { "epoch": 4.850202429149798, - "grad_norm": 0.4312866621160564, + "grad_norm": 0.1952816224047853, "learning_rate": 1.0936512977583802e-07, - "loss": 0.6935, + "loss": 0.8315, "step": 1198 }, { "epoch": 4.854251012145749, - "grad_norm": 0.4207805566345931, + "grad_norm": 0.18499144404754378, "learning_rate": 1.0353842512196732e-07, - "loss": 0.6876, + "loss": 0.8204, "step": 1199 }, { "epoch": 4.8582995951417, - "grad_norm": 0.4182399649062434, + "grad_norm": 0.2209825915963369, "learning_rate": 9.787081204466253e-08, - "loss": 0.6911, + "loss": 0.8237, "step": 1200 }, { "epoch": 4.862348178137652, - "grad_norm": 0.4202432261309142, + "grad_norm": 0.2074677551653603, "learning_rate": 9.236233586200627e-08, - "loss": 0.6825, + "loss": 0.7929, "step": 1201 }, { "epoch": 4.866396761133603, - "grad_norm": 0.4306252419668704, + "grad_norm": 0.19228810649556657, "learning_rate": 8.70130406196279e-08, - "loss": 0.7246, + "loss": 0.848, "step": 1202 }, { "epoch": 4.870445344129554, - "grad_norm": 0.4397711114745392, + "grad_norm": 0.1810627924626934, "learning_rate": 8.182296909035047e-08, - "loss": 0.6732, + "loss": 0.8052, "step": 1203 }, { "epoch": 4.874493927125506, - "grad_norm": 0.42682991096786144, + "grad_norm": 0.20269739864447628, "learning_rate": 7.679216277384439e-08, - "loss": 0.698, + "loss": 0.8457, "step": 1204 }, { "epoch": 4.8785425101214575, - "grad_norm": 0.4230685111365464, + "grad_norm": 0.19559405223355772, "learning_rate": 7.19206618963053e-08, - "loss": 0.7091, + "loss": 0.8337, "step": 1205 }, { "epoch": 4.882591093117409, - "grad_norm": 0.4382328920820113, + "grad_norm": 0.19900441194929736, "learning_rate": 6.720850541012347e-08, - "loss": 0.7102, + "loss": 0.8405, "step": 1206 }, { "epoch": 4.886639676113361, - "grad_norm": 0.40897250977637256, + "grad_norm": 0.18745032504863457, "learning_rate": 6.265573099357714e-08, - "loss": 0.6952, + "loss": 0.808, "step": 1207 }, { "epoch": 4.890688259109312, - "grad_norm": 0.4240250286260564, + "grad_norm": 0.19081137792308533, "learning_rate": 5.826237505053067e-08, - "loss": 0.6773, + "loss": 0.8086, "step": 1208 }, { "epoch": 4.894736842105263, - "grad_norm": 0.42229258553831894, + "grad_norm": 0.20735976151454993, "learning_rate": 5.4028472710137e-08, - "loss": 0.6944, + "loss": 0.8131, "step": 1209 }, { "epoch": 4.898785425101215, - "grad_norm": 0.4106051019106051, + "grad_norm": 0.18897745992164203, "learning_rate": 4.9954057826571146e-08, - "loss": 0.6815, + "loss": 0.7929, "step": 1210 }, { "epoch": 4.902834008097166, - "grad_norm": 0.43869941999700635, + "grad_norm": 0.19043769423874837, "learning_rate": 4.6039162978743776e-08, - "loss": 0.6919, + "loss": 0.8233, "step": 1211 }, { "epoch": 4.906882591093117, - "grad_norm": 0.4201648262153282, + "grad_norm": 0.18953743537333403, "learning_rate": 4.228381947004812e-08, - "loss": 0.6788, + "loss": 0.7954, "step": 1212 }, { "epoch": 4.910931174089069, - "grad_norm": 0.4161642339264833, + "grad_norm": 0.19083121733031624, "learning_rate": 3.868805732811343e-08, - "loss": 0.6877, + "loss": 0.8043, "step": 1213 }, { "epoch": 4.91497975708502, - "grad_norm": 0.4159157506555154, + "grad_norm": 0.19619934316452378, "learning_rate": 3.525190530455636e-08, - "loss": 0.6949, + "loss": 0.8072, "step": 1214 }, { "epoch": 4.919028340080971, - "grad_norm": 0.43238237257048084, + "grad_norm": 0.1997178476079843, "learning_rate": 3.197539087475887e-08, - "loss": 0.694, + "loss": 0.823, "step": 1215 }, { "epoch": 4.923076923076923, - "grad_norm": 0.42185454831749164, + "grad_norm": 0.19360484924206856, "learning_rate": 2.8858540237641786e-08, - "loss": 0.7159, + "loss": 0.8446, "step": 1216 }, { "epoch": 4.9271255060728745, - "grad_norm": 0.4286218170844706, + "grad_norm": 0.1987883343917353, "learning_rate": 2.5901378315458247e-08, - "loss": 0.6948, + "loss": 0.8114, "step": 1217 }, { "epoch": 4.931174089068826, - "grad_norm": 0.42884555513677947, + "grad_norm": 0.20529421376932722, "learning_rate": 2.3103928753598347e-08, - "loss": 0.7043, + "loss": 0.8481, "step": 1218 }, { "epoch": 4.935222672064778, - "grad_norm": 0.42566407724848904, + "grad_norm": 0.1880416355708006, "learning_rate": 2.0466213920393718e-08, - "loss": 0.7048, + "loss": 0.8409, "step": 1219 }, { "epoch": 4.939271255060729, - "grad_norm": 0.4407802517744325, + "grad_norm": 0.20284780112112683, "learning_rate": 1.7988254906942115e-08, - "loss": 0.7148, + "loss": 0.85, "step": 1220 }, { "epoch": 4.94331983805668, - "grad_norm": 0.4188062160199182, + "grad_norm": 0.18635891405507554, "learning_rate": 1.5670071526936447e-08, - "loss": 0.6901, + "loss": 0.8226, "step": 1221 }, { "epoch": 4.947368421052632, - "grad_norm": 0.43068830637816163, + "grad_norm": 0.20339004917489378, "learning_rate": 1.3511682316509344e-08, - "loss": 0.702, + "loss": 0.8216, "step": 1222 }, { "epoch": 4.951417004048583, - "grad_norm": 0.4215224010869817, + "grad_norm": 0.2013435142671592, "learning_rate": 1.1513104534086605e-08, - "loss": 0.6945, + "loss": 0.8141, "step": 1223 }, { "epoch": 4.955465587044534, - "grad_norm": 0.42355939005732013, + "grad_norm": 0.19422793916154213, "learning_rate": 9.67435416023843e-09, - "loss": 0.7043, + "loss": 0.8195, "step": 1224 }, { "epoch": 4.959514170040486, - "grad_norm": 0.4135380310912789, + "grad_norm": 0.19600963980191094, "learning_rate": 7.995445897566178e-09, - "loss": 0.68, + "loss": 0.8289, "step": 1225 }, { "epoch": 4.963562753036437, - "grad_norm": 0.4309118882792459, + "grad_norm": 0.1991568663478795, "learning_rate": 6.476393170578021e-09, - "loss": 0.7076, + "loss": 0.8326, "step": 1226 }, { "epoch": 4.967611336032388, - "grad_norm": 0.4207905684836021, + "grad_norm": 0.2128080613084259, "learning_rate": 5.11720812557348e-09, - "loss": 0.6955, + "loss": 0.8334, "step": 1227 }, { "epoch": 4.97165991902834, - "grad_norm": 0.44379868297658454, + "grad_norm": 0.20199168555613695, "learning_rate": 3.917901630563492e-09, - "loss": 0.6977, + "loss": 0.8252, "step": 1228 }, { "epoch": 4.9757085020242915, - "grad_norm": 0.4212978693324689, + "grad_norm": 0.19918537701090572, "learning_rate": 2.878483275170485e-09, - "loss": 0.7053, + "loss": 0.8339, "step": 1229 }, { "epoch": 4.979757085020243, - "grad_norm": 0.430528091111413, + "grad_norm": 0.19252508303003496, "learning_rate": 1.998961370555108e-09, - "loss": 0.6852, + "loss": 0.8176, "step": 1230 }, { "epoch": 4.983805668016195, - "grad_norm": 0.43270649080488033, + "grad_norm": 0.20265094197418562, "learning_rate": 1.2793429493518361e-09, - "loss": 0.6855, + "loss": 0.8243, "step": 1231 }, { "epoch": 4.987854251012146, - "grad_norm": 0.4197138321920042, + "grad_norm": 0.1945016243060065, "learning_rate": 7.196337656112384e-10, - "loss": 0.6983, + "loss": 0.8143, "step": 1232 }, { "epoch": 4.991902834008097, - "grad_norm": 0.4224363743110873, + "grad_norm": 0.1978959228313507, "learning_rate": 3.198382947511292e-10, - "loss": 0.6902, + "loss": 0.8281, "step": 1233 }, { "epoch": 4.995951417004049, - "grad_norm": 0.4282742571822906, + "grad_norm": 0.2011234281158961, "learning_rate": 7.995973352770137e-11, - "loss": 0.7031, + "loss": 0.829, "step": 1234 }, { "epoch": 5.0, - "grad_norm": 0.4237821336803865, + "grad_norm": 0.18580850601849525, "learning_rate": 0.0, - "loss": 0.6793, + "loss": 0.822, "step": 1235 }, { "epoch": 5.0, "step": 1235, - "total_flos": 1.6787206024137277e+18, - "train_loss": 0.0, - "train_runtime": 3.5426, - "train_samples_per_second": 44600.403, - "train_steps_per_second": 348.617 + "total_flos": 6.239074703412036e+18, + "train_loss": 0.9088270787768036, + "train_runtime": 22292.6502, + "train_samples_per_second": 7.088, + "train_steps_per_second": 0.055 } ], "logging_steps": 1, @@ -8680,7 +8680,7 @@ "attributes": {} } }, - "total_flos": 1.6787206024137277e+18, + "total_flos": 6.239074703412036e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null