diff --git "a/checkpoint-1680/trainer_state.json" "b/checkpoint-1680/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-1680/trainer_state.json" @@ -0,0 +1,11803 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 40.0, + "eval_steps": 25.0, + "global_step": 1680, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.023880597014925373, + "grad_norm": Infinity, + "learning_rate": 1e-05, + "loss": 76.8913, + "step": 1 + }, + { + "epoch": 0.04776119402985075, + "grad_norm": 395.8459777832031, + "learning_rate": 1e-05, + "loss": 77.3235, + "step": 2 + }, + { + "epoch": 0.07164179104477612, + "grad_norm": 161.73968505859375, + "learning_rate": 9.988095238095239e-06, + "loss": 67.3668, + "step": 3 + }, + { + "epoch": 0.0955223880597015, + "grad_norm": 145.1616973876953, + "learning_rate": 9.976190476190477e-06, + "loss": 62.8542, + "step": 4 + }, + { + "epoch": 0.11940298507462686, + "grad_norm": 52.10374069213867, + "learning_rate": 9.964285714285714e-06, + "loss": 59.8627, + "step": 5 + }, + { + "epoch": 0.14328358208955225, + "grad_norm": 31.430763244628906, + "learning_rate": 9.952380952380954e-06, + "loss": 59.2378, + "step": 6 + }, + { + "epoch": 0.16716417910447762, + "grad_norm": 37.42692947387695, + "learning_rate": 9.940476190476192e-06, + "loss": 58.3114, + "step": 7 + }, + { + "epoch": 0.191044776119403, + "grad_norm": 32.284332275390625, + "learning_rate": 9.92857142857143e-06, + "loss": 57.3408, + "step": 8 + }, + { + "epoch": 0.21492537313432836, + "grad_norm": 28.520849227905273, + "learning_rate": 9.916666666666668e-06, + "loss": 56.9256, + "step": 9 + }, + { + "epoch": 0.23880597014925373, + "grad_norm": 40.480167388916016, + "learning_rate": 9.904761904761906e-06, + "loss": 57.4842, + "step": 10 + }, + { + "epoch": 0.2626865671641791, + "grad_norm": 20.3512020111084, + "learning_rate": 9.892857142857143e-06, + "loss": 56.7597, + "step": 11 + }, + { + "epoch": 0.2865671641791045, + "grad_norm": 22.308382034301758, + "learning_rate": 9.880952380952381e-06, + "loss": 56.594, + "step": 12 + }, + { + "epoch": 0.31044776119402984, + "grad_norm": 14.086284637451172, + "learning_rate": 9.869047619047621e-06, + "loss": 56.6327, + "step": 13 + }, + { + "epoch": 0.33432835820895523, + "grad_norm": NaN, + "learning_rate": 9.857142857142859e-06, + "loss": 63.0791, + "step": 14 + }, + { + "epoch": 0.3582089552238806, + "grad_norm": 17.055604934692383, + "learning_rate": 9.857142857142859e-06, + "loss": 56.1551, + "step": 15 + }, + { + "epoch": 0.382089552238806, + "grad_norm": 18.52287483215332, + "learning_rate": 9.845238095238097e-06, + "loss": 54.9502, + "step": 16 + }, + { + "epoch": 0.4059701492537313, + "grad_norm": 32.66905212402344, + "learning_rate": 9.833333333333333e-06, + "loss": 55.6494, + "step": 17 + }, + { + "epoch": 0.4298507462686567, + "grad_norm": 24.075742721557617, + "learning_rate": 9.821428571428573e-06, + "loss": 55.4766, + "step": 18 + }, + { + "epoch": 0.4537313432835821, + "grad_norm": 31.505783081054688, + "learning_rate": 9.80952380952381e-06, + "loss": 55.1481, + "step": 19 + }, + { + "epoch": 0.47761194029850745, + "grad_norm": 30.53020477294922, + "learning_rate": 9.797619047619048e-06, + "loss": 55.158, + "step": 20 + }, + { + "epoch": 0.5014925373134328, + "grad_norm": 14.44444751739502, + "learning_rate": 9.785714285714286e-06, + "loss": 54.8286, + "step": 21 + }, + { + "epoch": 0.5253731343283582, + "grad_norm": 53.601078033447266, + "learning_rate": 9.773809523809524e-06, + "loss": 55.1466, + "step": 22 + }, + { + "epoch": 0.5492537313432836, + "grad_norm": 37.156028747558594, + "learning_rate": 9.761904761904762e-06, + "loss": 54.629, + "step": 23 + }, + { + "epoch": 0.573134328358209, + "grad_norm": 41.84994125366211, + "learning_rate": 9.75e-06, + "loss": 55.2766, + "step": 24 + }, + { + "epoch": 0.5970149253731343, + "grad_norm": 50.61705017089844, + "learning_rate": 9.73809523809524e-06, + "loss": 54.3497, + "step": 25 + }, + { + "epoch": 0.6208955223880597, + "grad_norm": 16.964982986450195, + "learning_rate": 9.726190476190477e-06, + "loss": 55.1673, + "step": 26 + }, + { + "epoch": 0.6447761194029851, + "grad_norm": 22.71157455444336, + "learning_rate": 9.714285714285715e-06, + "loss": 54.8001, + "step": 27 + }, + { + "epoch": 0.6686567164179105, + "grad_norm": 16.372802734375, + "learning_rate": 9.702380952380953e-06, + "loss": 54.4506, + "step": 28 + }, + { + "epoch": 0.6925373134328359, + "grad_norm": 13.21664047241211, + "learning_rate": 9.690476190476191e-06, + "loss": 53.0488, + "step": 29 + }, + { + "epoch": 0.7164179104477612, + "grad_norm": NaN, + "learning_rate": 9.678571428571429e-06, + "loss": 74.7707, + "step": 30 + }, + { + "epoch": 0.7402985074626866, + "grad_norm": 23.045652389526367, + "learning_rate": 9.678571428571429e-06, + "loss": 54.1909, + "step": 31 + }, + { + "epoch": 0.764179104477612, + "grad_norm": 15.034178733825684, + "learning_rate": 9.666666666666667e-06, + "loss": 53.3253, + "step": 32 + }, + { + "epoch": 0.7880597014925373, + "grad_norm": 14.148232460021973, + "learning_rate": 9.654761904761906e-06, + "loss": 53.693, + "step": 33 + }, + { + "epoch": 0.8119402985074626, + "grad_norm": NaN, + "learning_rate": 9.642857142857144e-06, + "loss": 61.3517, + "step": 34 + }, + { + "epoch": 0.835820895522388, + "grad_norm": 14.757994651794434, + "learning_rate": 9.642857142857144e-06, + "loss": 53.3175, + "step": 35 + }, + { + "epoch": 0.8597014925373134, + "grad_norm": 10.875706672668457, + "learning_rate": 9.630952380952382e-06, + "loss": 54.2592, + "step": 36 + }, + { + "epoch": 0.8835820895522388, + "grad_norm": 9.926539421081543, + "learning_rate": 9.61904761904762e-06, + "loss": 53.8721, + "step": 37 + }, + { + "epoch": 0.9074626865671642, + "grad_norm": 17.697235107421875, + "learning_rate": 9.607142857142858e-06, + "loss": 54.2901, + "step": 38 + }, + { + "epoch": 0.9313432835820895, + "grad_norm": 29.19430160522461, + "learning_rate": 9.595238095238096e-06, + "loss": 53.1261, + "step": 39 + }, + { + "epoch": 0.9552238805970149, + "grad_norm": 12.3985595703125, + "learning_rate": 9.583333333333335e-06, + "loss": 53.9815, + "step": 40 + }, + { + "epoch": 0.9791044776119403, + "grad_norm": 17.109691619873047, + "learning_rate": 9.571428571428573e-06, + "loss": 52.5838, + "step": 41 + }, + { + "epoch": 1.0, + "grad_norm": 14.529239654541016, + "learning_rate": 9.559523809523811e-06, + "loss": 46.1888, + "step": 42 + }, + { + "epoch": 1.0238805970149254, + "grad_norm": 15.683514595031738, + "learning_rate": 9.547619047619049e-06, + "loss": 52.2043, + "step": 43 + }, + { + "epoch": 1.0477611940298508, + "grad_norm": 26.219507217407227, + "learning_rate": 9.535714285714287e-06, + "loss": 53.09, + "step": 44 + }, + { + "epoch": 1.0716417910447762, + "grad_norm": 19.859697341918945, + "learning_rate": 9.523809523809525e-06, + "loss": 52.5858, + "step": 45 + }, + { + "epoch": 1.0955223880597016, + "grad_norm": 11.090332984924316, + "learning_rate": 9.511904761904763e-06, + "loss": 53.7674, + "step": 46 + }, + { + "epoch": 1.1194029850746268, + "grad_norm": 15.586993217468262, + "learning_rate": 9.5e-06, + "loss": 53.4901, + "step": 47 + }, + { + "epoch": 1.1432835820895522, + "grad_norm": 22.734928131103516, + "learning_rate": 9.488095238095238e-06, + "loss": 53.6667, + "step": 48 + }, + { + "epoch": 1.1671641791044776, + "grad_norm": 16.382047653198242, + "learning_rate": 9.476190476190476e-06, + "loss": 53.2914, + "step": 49 + }, + { + "epoch": 1.191044776119403, + "grad_norm": 15.916092872619629, + "learning_rate": 9.464285714285714e-06, + "loss": 51.729, + "step": 50 + }, + { + "epoch": 1.2149253731343284, + "grad_norm": NaN, + "learning_rate": 9.452380952380952e-06, + "loss": 66.3151, + "step": 51 + }, + { + "epoch": 1.2388059701492538, + "grad_norm": 20.121395111083984, + "learning_rate": 9.452380952380952e-06, + "loss": 53.866, + "step": 52 + }, + { + "epoch": 1.2626865671641792, + "grad_norm": NaN, + "learning_rate": 9.440476190476192e-06, + "loss": 67.3538, + "step": 53 + }, + { + "epoch": 1.2865671641791045, + "grad_norm": 13.869222640991211, + "learning_rate": 9.440476190476192e-06, + "loss": 53.5431, + "step": 54 + }, + { + "epoch": 1.31044776119403, + "grad_norm": 25.478107452392578, + "learning_rate": 9.42857142857143e-06, + "loss": 52.7126, + "step": 55 + }, + { + "epoch": 1.3343283582089551, + "grad_norm": 35.76942825317383, + "learning_rate": 9.416666666666667e-06, + "loss": 53.8082, + "step": 56 + }, + { + "epoch": 1.3582089552238805, + "grad_norm": 17.95109748840332, + "learning_rate": 9.404761904761905e-06, + "loss": 53.0275, + "step": 57 + }, + { + "epoch": 1.382089552238806, + "grad_norm": 26.80129623413086, + "learning_rate": 9.392857142857143e-06, + "loss": 53.3295, + "step": 58 + }, + { + "epoch": 1.4059701492537313, + "grad_norm": 25.803054809570312, + "learning_rate": 9.380952380952381e-06, + "loss": 52.3771, + "step": 59 + }, + { + "epoch": 1.4298507462686567, + "grad_norm": 35.39850616455078, + "learning_rate": 9.36904761904762e-06, + "loss": 52.9467, + "step": 60 + }, + { + "epoch": 1.4537313432835821, + "grad_norm": 27.43315887451172, + "learning_rate": 9.357142857142859e-06, + "loss": 53.1359, + "step": 61 + }, + { + "epoch": 1.4776119402985075, + "grad_norm": 33.431400299072266, + "learning_rate": 9.345238095238096e-06, + "loss": 52.2006, + "step": 62 + }, + { + "epoch": 1.5014925373134327, + "grad_norm": 33.08237075805664, + "learning_rate": 9.333333333333334e-06, + "loss": 52.6816, + "step": 63 + }, + { + "epoch": 1.5253731343283583, + "grad_norm": 25.711997985839844, + "learning_rate": 9.321428571428572e-06, + "loss": 52.6113, + "step": 64 + }, + { + "epoch": 1.5492537313432835, + "grad_norm": 32.688297271728516, + "learning_rate": 9.30952380952381e-06, + "loss": 51.9086, + "step": 65 + }, + { + "epoch": 1.573134328358209, + "grad_norm": 31.856857299804688, + "learning_rate": 9.297619047619048e-06, + "loss": 53.1913, + "step": 66 + }, + { + "epoch": 1.5970149253731343, + "grad_norm": 26.231773376464844, + "learning_rate": 9.285714285714288e-06, + "loss": 51.464, + "step": 67 + }, + { + "epoch": 1.6208955223880597, + "grad_norm": 29.39109230041504, + "learning_rate": 9.273809523809525e-06, + "loss": 52.0572, + "step": 68 + }, + { + "epoch": 1.644776119402985, + "grad_norm": 28.86277198791504, + "learning_rate": 9.261904761904763e-06, + "loss": 52.7115, + "step": 69 + }, + { + "epoch": 1.6686567164179105, + "grad_norm": 28.82640266418457, + "learning_rate": 9.250000000000001e-06, + "loss": 53.1756, + "step": 70 + }, + { + "epoch": 1.6925373134328359, + "grad_norm": 31.32577133178711, + "learning_rate": 9.238095238095239e-06, + "loss": 52.369, + "step": 71 + }, + { + "epoch": 1.716417910447761, + "grad_norm": 18.739269256591797, + "learning_rate": 9.226190476190477e-06, + "loss": 52.6631, + "step": 72 + }, + { + "epoch": 1.7402985074626867, + "grad_norm": 22.889320373535156, + "learning_rate": 9.214285714285715e-06, + "loss": 52.4786, + "step": 73 + }, + { + "epoch": 1.7641791044776118, + "grad_norm": 24.175626754760742, + "learning_rate": 9.202380952380953e-06, + "loss": 52.1792, + "step": 74 + }, + { + "epoch": 1.7880597014925375, + "grad_norm": 25.257095336914062, + "learning_rate": 9.19047619047619e-06, + "loss": 51.39, + "step": 75 + }, + { + "epoch": 1.8119402985074626, + "grad_norm": 35.745208740234375, + "learning_rate": 9.178571428571429e-06, + "loss": 53.331, + "step": 76 + }, + { + "epoch": 1.835820895522388, + "grad_norm": 23.815813064575195, + "learning_rate": 9.166666666666666e-06, + "loss": 52.7632, + "step": 77 + }, + { + "epoch": 1.8597014925373134, + "grad_norm": 32.405757904052734, + "learning_rate": 9.154761904761906e-06, + "loss": 52.3165, + "step": 78 + }, + { + "epoch": 1.8835820895522388, + "grad_norm": 38.95046615600586, + "learning_rate": 9.142857142857144e-06, + "loss": 52.0931, + "step": 79 + }, + { + "epoch": 1.9074626865671642, + "grad_norm": 22.412342071533203, + "learning_rate": 9.130952380952382e-06, + "loss": 51.8732, + "step": 80 + }, + { + "epoch": 1.9313432835820894, + "grad_norm": 35.088253021240234, + "learning_rate": 9.11904761904762e-06, + "loss": 52.0182, + "step": 81 + }, + { + "epoch": 1.955223880597015, + "grad_norm": 20.136964797973633, + "learning_rate": 9.107142857142858e-06, + "loss": 52.7956, + "step": 82 + }, + { + "epoch": 1.9791044776119402, + "grad_norm": 20.860034942626953, + "learning_rate": 9.095238095238095e-06, + "loss": 50.339, + "step": 83 + }, + { + "epoch": 2.0, + "grad_norm": 20.38931655883789, + "learning_rate": 9.083333333333333e-06, + "loss": 45.3651, + "step": 84 + }, + { + "epoch": 2.023880597014925, + "grad_norm": 21.519498825073242, + "learning_rate": 9.071428571428573e-06, + "loss": 52.1228, + "step": 85 + }, + { + "epoch": 2.047761194029851, + "grad_norm": 30.47000503540039, + "learning_rate": 9.05952380952381e-06, + "loss": 51.1968, + "step": 86 + }, + { + "epoch": 2.071641791044776, + "grad_norm": 19.728044509887695, + "learning_rate": 9.047619047619049e-06, + "loss": 52.1514, + "step": 87 + }, + { + "epoch": 2.0955223880597016, + "grad_norm": 42.647281646728516, + "learning_rate": 9.035714285714287e-06, + "loss": 52.5597, + "step": 88 + }, + { + "epoch": 2.1194029850746268, + "grad_norm": 55.45186233520508, + "learning_rate": 9.023809523809524e-06, + "loss": 53.7621, + "step": 89 + }, + { + "epoch": 2.1432835820895524, + "grad_norm": 14.879026412963867, + "learning_rate": 9.011904761904762e-06, + "loss": 52.0743, + "step": 90 + }, + { + "epoch": 2.1671641791044776, + "grad_norm": 48.55704116821289, + "learning_rate": 9e-06, + "loss": 52.0543, + "step": 91 + }, + { + "epoch": 2.191044776119403, + "grad_norm": 37.96782302856445, + "learning_rate": 8.98809523809524e-06, + "loss": 52.4684, + "step": 92 + }, + { + "epoch": 2.2149253731343284, + "grad_norm": 27.714475631713867, + "learning_rate": 8.976190476190478e-06, + "loss": 52.725, + "step": 93 + }, + { + "epoch": 2.2388059701492535, + "grad_norm": 32.12433624267578, + "learning_rate": 8.964285714285716e-06, + "loss": 52.7161, + "step": 94 + }, + { + "epoch": 2.262686567164179, + "grad_norm": 23.8153018951416, + "learning_rate": 8.952380952380953e-06, + "loss": 51.3124, + "step": 95 + }, + { + "epoch": 2.2865671641791043, + "grad_norm": 31.269794464111328, + "learning_rate": 8.940476190476191e-06, + "loss": 51.9646, + "step": 96 + }, + { + "epoch": 2.31044776119403, + "grad_norm": 16.611865997314453, + "learning_rate": 8.92857142857143e-06, + "loss": 51.8503, + "step": 97 + }, + { + "epoch": 2.334328358208955, + "grad_norm": 26.69631004333496, + "learning_rate": 8.916666666666667e-06, + "loss": 52.4857, + "step": 98 + }, + { + "epoch": 2.3582089552238807, + "grad_norm": 21.10638999938965, + "learning_rate": 8.904761904761905e-06, + "loss": 52.0022, + "step": 99 + }, + { + "epoch": 2.382089552238806, + "grad_norm": 16.273351669311523, + "learning_rate": 8.892857142857143e-06, + "loss": 50.4367, + "step": 100 + }, + { + "epoch": 2.405970149253731, + "grad_norm": 16.407167434692383, + "learning_rate": 8.88095238095238e-06, + "loss": 51.2079, + "step": 101 + }, + { + "epoch": 2.4298507462686567, + "grad_norm": 16.22024154663086, + "learning_rate": 8.869047619047619e-06, + "loss": 50.4939, + "step": 102 + }, + { + "epoch": 2.4537313432835823, + "grad_norm": 27.11235809326172, + "learning_rate": 8.857142857142858e-06, + "loss": 50.0872, + "step": 103 + }, + { + "epoch": 2.4776119402985075, + "grad_norm": 18.912181854248047, + "learning_rate": 8.845238095238096e-06, + "loss": 51.8135, + "step": 104 + }, + { + "epoch": 2.5014925373134327, + "grad_norm": 29.597028732299805, + "learning_rate": 8.833333333333334e-06, + "loss": 49.4789, + "step": 105 + }, + { + "epoch": 2.5253731343283583, + "grad_norm": 30.51687240600586, + "learning_rate": 8.821428571428572e-06, + "loss": 52.5555, + "step": 106 + }, + { + "epoch": 2.5492537313432835, + "grad_norm": 31.4583797454834, + "learning_rate": 8.80952380952381e-06, + "loss": 51.0073, + "step": 107 + }, + { + "epoch": 2.573134328358209, + "grad_norm": 30.35653305053711, + "learning_rate": 8.797619047619048e-06, + "loss": 50.9501, + "step": 108 + }, + { + "epoch": 2.5970149253731343, + "grad_norm": 24.041545867919922, + "learning_rate": 8.785714285714286e-06, + "loss": 49.5162, + "step": 109 + }, + { + "epoch": 2.62089552238806, + "grad_norm": 23.52166175842285, + "learning_rate": 8.773809523809525e-06, + "loss": 52.9747, + "step": 110 + }, + { + "epoch": 2.644776119402985, + "grad_norm": 28.871065139770508, + "learning_rate": 8.761904761904763e-06, + "loss": 50.2273, + "step": 111 + }, + { + "epoch": 2.6686567164179102, + "grad_norm": 26.484140396118164, + "learning_rate": 8.750000000000001e-06, + "loss": 51.2286, + "step": 112 + }, + { + "epoch": 2.692537313432836, + "grad_norm": 37.570743560791016, + "learning_rate": 8.738095238095239e-06, + "loss": 49.7131, + "step": 113 + }, + { + "epoch": 2.716417910447761, + "grad_norm": 23.827178955078125, + "learning_rate": 8.726190476190477e-06, + "loss": 51.913, + "step": 114 + }, + { + "epoch": 2.7402985074626867, + "grad_norm": 33.89924621582031, + "learning_rate": 8.714285714285715e-06, + "loss": 52.2382, + "step": 115 + }, + { + "epoch": 2.764179104477612, + "grad_norm": 29.397851943969727, + "learning_rate": 8.702380952380952e-06, + "loss": 52.1548, + "step": 116 + }, + { + "epoch": 2.7880597014925375, + "grad_norm": 28.73517417907715, + "learning_rate": 8.690476190476192e-06, + "loss": 51.2892, + "step": 117 + }, + { + "epoch": 2.8119402985074626, + "grad_norm": 32.068138122558594, + "learning_rate": 8.67857142857143e-06, + "loss": 51.542, + "step": 118 + }, + { + "epoch": 2.835820895522388, + "grad_norm": 22.76898956298828, + "learning_rate": 8.666666666666668e-06, + "loss": 50.373, + "step": 119 + }, + { + "epoch": 2.8597014925373134, + "grad_norm": 33.528263092041016, + "learning_rate": 8.654761904761906e-06, + "loss": 51.3075, + "step": 120 + }, + { + "epoch": 2.883582089552239, + "grad_norm": 21.655696868896484, + "learning_rate": 8.642857142857144e-06, + "loss": 51.1397, + "step": 121 + }, + { + "epoch": 2.9074626865671642, + "grad_norm": 25.94880485534668, + "learning_rate": 8.630952380952381e-06, + "loss": 51.4326, + "step": 122 + }, + { + "epoch": 2.9313432835820894, + "grad_norm": 36.14421844482422, + "learning_rate": 8.61904761904762e-06, + "loss": 50.9524, + "step": 123 + }, + { + "epoch": 2.955223880597015, + "grad_norm": 28.361459732055664, + "learning_rate": 8.607142857142859e-06, + "loss": 51.3171, + "step": 124 + }, + { + "epoch": 2.97910447761194, + "grad_norm": 30.784954071044922, + "learning_rate": 8.595238095238097e-06, + "loss": 49.9797, + "step": 125 + }, + { + "epoch": 3.0, + "grad_norm": 22.98565101623535, + "learning_rate": 8.583333333333333e-06, + "loss": 44.3471, + "step": 126 + }, + { + "epoch": 3.023880597014925, + "grad_norm": 25.601985931396484, + "learning_rate": 8.571428571428571e-06, + "loss": 51.6574, + "step": 127 + }, + { + "epoch": 3.047761194029851, + "grad_norm": 27.648792266845703, + "learning_rate": 8.55952380952381e-06, + "loss": 51.3385, + "step": 128 + }, + { + "epoch": 3.071641791044776, + "grad_norm": 18.773529052734375, + "learning_rate": 8.547619047619048e-06, + "loss": 51.1451, + "step": 129 + }, + { + "epoch": 3.0955223880597016, + "grad_norm": 29.439353942871094, + "learning_rate": 8.535714285714286e-06, + "loss": 51.6092, + "step": 130 + }, + { + "epoch": 3.1194029850746268, + "grad_norm": 32.41486740112305, + "learning_rate": 8.523809523809524e-06, + "loss": 50.9068, + "step": 131 + }, + { + "epoch": 3.1432835820895524, + "grad_norm": 23.441896438598633, + "learning_rate": 8.511904761904762e-06, + "loss": 51.7453, + "step": 132 + }, + { + "epoch": 3.1671641791044776, + "grad_norm": 29.218734741210938, + "learning_rate": 8.5e-06, + "loss": 49.9124, + "step": 133 + }, + { + "epoch": 3.191044776119403, + "grad_norm": 20.988981246948242, + "learning_rate": 8.488095238095238e-06, + "loss": 50.9788, + "step": 134 + }, + { + "epoch": 3.2149253731343284, + "grad_norm": 22.57052993774414, + "learning_rate": 8.476190476190477e-06, + "loss": 51.4228, + "step": 135 + }, + { + "epoch": 3.2388059701492535, + "grad_norm": 26.112573623657227, + "learning_rate": 8.464285714285715e-06, + "loss": 50.3332, + "step": 136 + }, + { + "epoch": 3.262686567164179, + "grad_norm": 23.8747615814209, + "learning_rate": 8.452380952380953e-06, + "loss": 51.1763, + "step": 137 + }, + { + "epoch": 3.2865671641791043, + "grad_norm": 24.12811851501465, + "learning_rate": 8.440476190476191e-06, + "loss": 49.8539, + "step": 138 + }, + { + "epoch": 3.31044776119403, + "grad_norm": 27.462984085083008, + "learning_rate": 8.428571428571429e-06, + "loss": 50.7766, + "step": 139 + }, + { + "epoch": 3.334328358208955, + "grad_norm": 31.261472702026367, + "learning_rate": 8.416666666666667e-06, + "loss": 49.7599, + "step": 140 + }, + { + "epoch": 3.3582089552238807, + "grad_norm": 21.049545288085938, + "learning_rate": 8.404761904761905e-06, + "loss": 49.6827, + "step": 141 + }, + { + "epoch": 3.382089552238806, + "grad_norm": 30.103389739990234, + "learning_rate": 8.392857142857144e-06, + "loss": 49.3866, + "step": 142 + }, + { + "epoch": 3.405970149253731, + "grad_norm": 31.348888397216797, + "learning_rate": 8.380952380952382e-06, + "loss": 51.4607, + "step": 143 + }, + { + "epoch": 3.4298507462686567, + "grad_norm": 28.910200119018555, + "learning_rate": 8.36904761904762e-06, + "loss": 51.2337, + "step": 144 + }, + { + "epoch": 3.4537313432835823, + "grad_norm": 21.00281524658203, + "learning_rate": 8.357142857142858e-06, + "loss": 50.9557, + "step": 145 + }, + { + "epoch": 3.4776119402985075, + "grad_norm": 45.842002868652344, + "learning_rate": 8.345238095238096e-06, + "loss": 49.7377, + "step": 146 + }, + { + "epoch": 3.5014925373134327, + "grad_norm": 30.77996253967285, + "learning_rate": 8.333333333333334e-06, + "loss": 51.1234, + "step": 147 + }, + { + "epoch": 3.5253731343283583, + "grad_norm": 31.492767333984375, + "learning_rate": 8.321428571428573e-06, + "loss": 50.5733, + "step": 148 + }, + { + "epoch": 3.5492537313432835, + "grad_norm": 36.57206344604492, + "learning_rate": 8.309523809523811e-06, + "loss": 50.6762, + "step": 149 + }, + { + "epoch": 3.573134328358209, + "grad_norm": 33.86347198486328, + "learning_rate": 8.297619047619049e-06, + "loss": 50.8281, + "step": 150 + }, + { + "epoch": 3.5970149253731343, + "grad_norm": 30.812152862548828, + "learning_rate": 8.285714285714287e-06, + "loss": 50.6509, + "step": 151 + }, + { + "epoch": 3.62089552238806, + "grad_norm": 24.536882400512695, + "learning_rate": 8.273809523809523e-06, + "loss": 50.1112, + "step": 152 + }, + { + "epoch": 3.644776119402985, + "grad_norm": 29.8430233001709, + "learning_rate": 8.261904761904763e-06, + "loss": 50.846, + "step": 153 + }, + { + "epoch": 3.6686567164179102, + "grad_norm": 26.18596076965332, + "learning_rate": 8.25e-06, + "loss": 50.3806, + "step": 154 + }, + { + "epoch": 3.692537313432836, + "grad_norm": 38.75019836425781, + "learning_rate": 8.238095238095239e-06, + "loss": 49.8915, + "step": 155 + }, + { + "epoch": 3.716417910447761, + "grad_norm": 34.30149841308594, + "learning_rate": 8.226190476190476e-06, + "loss": 50.7886, + "step": 156 + }, + { + "epoch": 3.7402985074626867, + "grad_norm": 33.179298400878906, + "learning_rate": 8.214285714285714e-06, + "loss": 50.8175, + "step": 157 + }, + { + "epoch": 3.764179104477612, + "grad_norm": 34.90909957885742, + "learning_rate": 8.202380952380952e-06, + "loss": 50.3521, + "step": 158 + }, + { + "epoch": 3.7880597014925375, + "grad_norm": 33.2717399597168, + "learning_rate": 8.190476190476192e-06, + "loss": 51.2006, + "step": 159 + }, + { + "epoch": 3.8119402985074626, + "grad_norm": 33.082672119140625, + "learning_rate": 8.17857142857143e-06, + "loss": 49.5627, + "step": 160 + }, + { + "epoch": 3.835820895522388, + "grad_norm": 23.65228843688965, + "learning_rate": 8.166666666666668e-06, + "loss": 49.9631, + "step": 161 + }, + { + "epoch": 3.8597014925373134, + "grad_norm": 37.3172492980957, + "learning_rate": 8.154761904761905e-06, + "loss": 50.7175, + "step": 162 + }, + { + "epoch": 3.883582089552239, + "grad_norm": 29.369930267333984, + "learning_rate": 8.142857142857143e-06, + "loss": 51.1435, + "step": 163 + }, + { + "epoch": 3.9074626865671642, + "grad_norm": 28.807470321655273, + "learning_rate": 8.130952380952381e-06, + "loss": 50.3349, + "step": 164 + }, + { + "epoch": 3.9313432835820894, + "grad_norm": 33.90628433227539, + "learning_rate": 8.119047619047619e-06, + "loss": 50.7241, + "step": 165 + }, + { + "epoch": 3.955223880597015, + "grad_norm": 21.72952651977539, + "learning_rate": 8.107142857142859e-06, + "loss": 49.9013, + "step": 166 + }, + { + "epoch": 3.97910447761194, + "grad_norm": 26.831520080566406, + "learning_rate": 8.095238095238097e-06, + "loss": 51.0161, + "step": 167 + }, + { + "epoch": 4.0, + "grad_norm": 24.49069595336914, + "learning_rate": 8.083333333333334e-06, + "loss": 44.6758, + "step": 168 + }, + { + "epoch": 4.023880597014926, + "grad_norm": 36.32711410522461, + "learning_rate": 8.071428571428572e-06, + "loss": 49.8601, + "step": 169 + }, + { + "epoch": 4.04776119402985, + "grad_norm": 29.862812042236328, + "learning_rate": 8.05952380952381e-06, + "loss": 51.712, + "step": 170 + }, + { + "epoch": 4.071641791044776, + "grad_norm": 40.245887756347656, + "learning_rate": 8.047619047619048e-06, + "loss": 50.3353, + "step": 171 + }, + { + "epoch": 4.095522388059702, + "grad_norm": 34.22684097290039, + "learning_rate": 8.035714285714286e-06, + "loss": 50.6474, + "step": 172 + }, + { + "epoch": 4.119402985074627, + "grad_norm": 36.754669189453125, + "learning_rate": 8.023809523809526e-06, + "loss": 50.1623, + "step": 173 + }, + { + "epoch": 4.143283582089552, + "grad_norm": 35.76541519165039, + "learning_rate": 8.011904761904763e-06, + "loss": 50.2426, + "step": 174 + }, + { + "epoch": 4.167164179104478, + "grad_norm": 25.851362228393555, + "learning_rate": 8.000000000000001e-06, + "loss": 49.9525, + "step": 175 + }, + { + "epoch": 4.191044776119403, + "grad_norm": 24.48278045654297, + "learning_rate": 7.98809523809524e-06, + "loss": 49.1466, + "step": 176 + }, + { + "epoch": 4.214925373134328, + "grad_norm": 28.79146385192871, + "learning_rate": 7.976190476190477e-06, + "loss": 49.9365, + "step": 177 + }, + { + "epoch": 4.2388059701492535, + "grad_norm": 29.29482650756836, + "learning_rate": 7.964285714285715e-06, + "loss": 50.7427, + "step": 178 + }, + { + "epoch": 4.262686567164179, + "grad_norm": 23.50571060180664, + "learning_rate": 7.952380952380953e-06, + "loss": 49.7287, + "step": 179 + }, + { + "epoch": 4.286567164179105, + "grad_norm": 27.805828094482422, + "learning_rate": 7.94047619047619e-06, + "loss": 50.4316, + "step": 180 + }, + { + "epoch": 4.3104477611940295, + "grad_norm": 28.323888778686523, + "learning_rate": 7.928571428571429e-06, + "loss": 50.0263, + "step": 181 + }, + { + "epoch": 4.334328358208955, + "grad_norm": 25.43438148498535, + "learning_rate": 7.916666666666667e-06, + "loss": 49.426, + "step": 182 + }, + { + "epoch": 4.358208955223881, + "grad_norm": 22.169496536254883, + "learning_rate": 7.904761904761904e-06, + "loss": 51.1048, + "step": 183 + }, + { + "epoch": 4.382089552238806, + "grad_norm": 33.660545349121094, + "learning_rate": 7.892857142857144e-06, + "loss": 49.7654, + "step": 184 + }, + { + "epoch": 4.405970149253731, + "grad_norm": 24.276273727416992, + "learning_rate": 7.880952380952382e-06, + "loss": 50.4976, + "step": 185 + }, + { + "epoch": 4.429850746268657, + "grad_norm": 41.48741149902344, + "learning_rate": 7.86904761904762e-06, + "loss": 52.0386, + "step": 186 + }, + { + "epoch": 4.453731343283582, + "grad_norm": 25.86789894104004, + "learning_rate": 7.857142857142858e-06, + "loss": 49.129, + "step": 187 + }, + { + "epoch": 4.477611940298507, + "grad_norm": 26.607038497924805, + "learning_rate": 7.845238095238096e-06, + "loss": 49.3561, + "step": 188 + }, + { + "epoch": 4.501492537313433, + "grad_norm": 43.54303741455078, + "learning_rate": 7.833333333333333e-06, + "loss": 50.1143, + "step": 189 + }, + { + "epoch": 4.525373134328358, + "grad_norm": 45.6146354675293, + "learning_rate": 7.821428571428571e-06, + "loss": 49.3217, + "step": 190 + }, + { + "epoch": 4.549253731343284, + "grad_norm": 24.00080680847168, + "learning_rate": 7.809523809523811e-06, + "loss": 50.484, + "step": 191 + }, + { + "epoch": 4.573134328358209, + "grad_norm": 29.736740112304688, + "learning_rate": 7.797619047619049e-06, + "loss": 49.748, + "step": 192 + }, + { + "epoch": 4.597014925373134, + "grad_norm": 33.08702850341797, + "learning_rate": 7.785714285714287e-06, + "loss": 50.2142, + "step": 193 + }, + { + "epoch": 4.62089552238806, + "grad_norm": 19.16411018371582, + "learning_rate": 7.773809523809525e-06, + "loss": 50.1073, + "step": 194 + }, + { + "epoch": 4.6447761194029855, + "grad_norm": 32.145721435546875, + "learning_rate": 7.761904761904762e-06, + "loss": 48.5769, + "step": 195 + }, + { + "epoch": 4.66865671641791, + "grad_norm": 38.768341064453125, + "learning_rate": 7.75e-06, + "loss": 49.681, + "step": 196 + }, + { + "epoch": 4.692537313432836, + "grad_norm": 26.108245849609375, + "learning_rate": 7.738095238095238e-06, + "loss": 49.9193, + "step": 197 + }, + { + "epoch": 4.7164179104477615, + "grad_norm": 28.86294174194336, + "learning_rate": 7.726190476190478e-06, + "loss": 50.4584, + "step": 198 + }, + { + "epoch": 4.740298507462686, + "grad_norm": 31.089380264282227, + "learning_rate": 7.714285714285716e-06, + "loss": 50.7873, + "step": 199 + }, + { + "epoch": 4.764179104477612, + "grad_norm": 22.934032440185547, + "learning_rate": 7.702380952380954e-06, + "loss": 50.611, + "step": 200 + }, + { + "epoch": 4.7880597014925375, + "grad_norm": 27.986371994018555, + "learning_rate": 7.690476190476191e-06, + "loss": 49.275, + "step": 201 + }, + { + "epoch": 4.811940298507462, + "grad_norm": 23.44196319580078, + "learning_rate": 7.67857142857143e-06, + "loss": 50.2035, + "step": 202 + }, + { + "epoch": 4.835820895522388, + "grad_norm": 22.05059242248535, + "learning_rate": 7.666666666666667e-06, + "loss": 48.9595, + "step": 203 + }, + { + "epoch": 4.859701492537313, + "grad_norm": 29.709396362304688, + "learning_rate": 7.654761904761905e-06, + "loss": 50.5343, + "step": 204 + }, + { + "epoch": 4.883582089552239, + "grad_norm": 23.702781677246094, + "learning_rate": 7.642857142857143e-06, + "loss": 50.2627, + "step": 205 + }, + { + "epoch": 4.907462686567165, + "grad_norm": 20.144807815551758, + "learning_rate": 7.630952380952381e-06, + "loss": 51.0125, + "step": 206 + }, + { + "epoch": 4.931343283582089, + "grad_norm": 28.83676528930664, + "learning_rate": 7.61904761904762e-06, + "loss": 50.5985, + "step": 207 + }, + { + "epoch": 4.955223880597015, + "grad_norm": 34.40160369873047, + "learning_rate": 7.6071428571428575e-06, + "loss": 49.6469, + "step": 208 + }, + { + "epoch": 4.979104477611941, + "grad_norm": 26.982925415039062, + "learning_rate": 7.595238095238095e-06, + "loss": 50.1666, + "step": 209 + }, + { + "epoch": 5.0, + "grad_norm": 19.569746017456055, + "learning_rate": 7.583333333333333e-06, + "loss": 43.6715, + "step": 210 + }, + { + "epoch": 5.023880597014926, + "grad_norm": 23.753328323364258, + "learning_rate": 7.571428571428572e-06, + "loss": 49.9273, + "step": 211 + }, + { + "epoch": 5.04776119402985, + "grad_norm": 22.463659286499023, + "learning_rate": 7.55952380952381e-06, + "loss": 48.8499, + "step": 212 + }, + { + "epoch": 5.071641791044776, + "grad_norm": 24.507875442504883, + "learning_rate": 7.547619047619048e-06, + "loss": 49.3275, + "step": 213 + }, + { + "epoch": 5.095522388059702, + "grad_norm": 21.727603912353516, + "learning_rate": 7.5357142857142865e-06, + "loss": 49.1879, + "step": 214 + }, + { + "epoch": 5.119402985074627, + "grad_norm": 26.122251510620117, + "learning_rate": 7.523809523809524e-06, + "loss": 50.1094, + "step": 215 + }, + { + "epoch": 5.143283582089552, + "grad_norm": 24.142263412475586, + "learning_rate": 7.511904761904762e-06, + "loss": 50.2708, + "step": 216 + }, + { + "epoch": 5.167164179104478, + "grad_norm": 22.762237548828125, + "learning_rate": 7.500000000000001e-06, + "loss": 50.441, + "step": 217 + }, + { + "epoch": 5.191044776119403, + "grad_norm": 35.74570846557617, + "learning_rate": 7.488095238095239e-06, + "loss": 48.5121, + "step": 218 + }, + { + "epoch": 5.214925373134328, + "grad_norm": 30.92180824279785, + "learning_rate": 7.476190476190477e-06, + "loss": 49.4257, + "step": 219 + }, + { + "epoch": 5.2388059701492535, + "grad_norm": 26.90997314453125, + "learning_rate": 7.464285714285715e-06, + "loss": 50.9712, + "step": 220 + }, + { + "epoch": 5.262686567164179, + "grad_norm": 35.544700622558594, + "learning_rate": 7.4523809523809534e-06, + "loss": 49.7908, + "step": 221 + }, + { + "epoch": 5.286567164179105, + "grad_norm": 33.78145217895508, + "learning_rate": 7.440476190476191e-06, + "loss": 49.2105, + "step": 222 + }, + { + "epoch": 5.3104477611940295, + "grad_norm": 32.16508102416992, + "learning_rate": 7.428571428571429e-06, + "loss": 49.8545, + "step": 223 + }, + { + "epoch": 5.334328358208955, + "grad_norm": 30.4263973236084, + "learning_rate": 7.416666666666668e-06, + "loss": 50.0994, + "step": 224 + }, + { + "epoch": 5.358208955223881, + "grad_norm": 25.801084518432617, + "learning_rate": 7.404761904761906e-06, + "loss": 49.6227, + "step": 225 + }, + { + "epoch": 5.382089552238806, + "grad_norm": 27.16851234436035, + "learning_rate": 7.392857142857144e-06, + "loss": 50.0005, + "step": 226 + }, + { + "epoch": 5.405970149253731, + "grad_norm": 30.102867126464844, + "learning_rate": 7.380952380952382e-06, + "loss": 50.7114, + "step": 227 + }, + { + "epoch": 5.429850746268657, + "grad_norm": 26.032968521118164, + "learning_rate": 7.36904761904762e-06, + "loss": 48.99, + "step": 228 + }, + { + "epoch": 5.453731343283582, + "grad_norm": 24.074424743652344, + "learning_rate": 7.357142857142858e-06, + "loss": 49.6965, + "step": 229 + }, + { + "epoch": 5.477611940298507, + "grad_norm": 24.5870361328125, + "learning_rate": 7.345238095238096e-06, + "loss": 48.8593, + "step": 230 + }, + { + "epoch": 5.501492537313433, + "grad_norm": 22.831932067871094, + "learning_rate": 7.333333333333333e-06, + "loss": 49.9975, + "step": 231 + }, + { + "epoch": 5.525373134328358, + "grad_norm": 26.88197135925293, + "learning_rate": 7.321428571428572e-06, + "loss": 49.5131, + "step": 232 + }, + { + "epoch": 5.549253731343284, + "grad_norm": 28.986154556274414, + "learning_rate": 7.30952380952381e-06, + "loss": 48.9042, + "step": 233 + }, + { + "epoch": 5.573134328358209, + "grad_norm": 17.220605850219727, + "learning_rate": 7.297619047619048e-06, + "loss": 49.453, + "step": 234 + }, + { + "epoch": 5.597014925373134, + "grad_norm": 25.110107421875, + "learning_rate": 7.285714285714286e-06, + "loss": 49.8196, + "step": 235 + }, + { + "epoch": 5.62089552238806, + "grad_norm": 34.680870056152344, + "learning_rate": 7.273809523809524e-06, + "loss": 49.9709, + "step": 236 + }, + { + "epoch": 5.6447761194029855, + "grad_norm": 24.10121726989746, + "learning_rate": 7.261904761904762e-06, + "loss": 49.4685, + "step": 237 + }, + { + "epoch": 5.66865671641791, + "grad_norm": 28.65550422668457, + "learning_rate": 7.25e-06, + "loss": 50.3439, + "step": 238 + }, + { + "epoch": 5.692537313432836, + "grad_norm": 27.49604606628418, + "learning_rate": 7.238095238095239e-06, + "loss": 50.0419, + "step": 239 + }, + { + "epoch": 5.7164179104477615, + "grad_norm": 22.843509674072266, + "learning_rate": 7.226190476190477e-06, + "loss": 49.0357, + "step": 240 + }, + { + "epoch": 5.740298507462686, + "grad_norm": 36.56801223754883, + "learning_rate": 7.2142857142857145e-06, + "loss": 49.4478, + "step": 241 + }, + { + "epoch": 5.764179104477612, + "grad_norm": 38.233734130859375, + "learning_rate": 7.202380952380953e-06, + "loss": 50.4473, + "step": 242 + }, + { + "epoch": 5.7880597014925375, + "grad_norm": 29.198333740234375, + "learning_rate": 7.190476190476191e-06, + "loss": 49.8598, + "step": 243 + }, + { + "epoch": 5.811940298507462, + "grad_norm": 34.49404525756836, + "learning_rate": 7.178571428571429e-06, + "loss": 49.1441, + "step": 244 + }, + { + "epoch": 5.835820895522388, + "grad_norm": 35.568359375, + "learning_rate": 7.166666666666667e-06, + "loss": 49.9402, + "step": 245 + }, + { + "epoch": 5.859701492537313, + "grad_norm": 31.041446685791016, + "learning_rate": 7.154761904761906e-06, + "loss": 50.1265, + "step": 246 + }, + { + "epoch": 5.883582089552239, + "grad_norm": 48.34186935424805, + "learning_rate": 7.1428571428571436e-06, + "loss": 50.7649, + "step": 247 + }, + { + "epoch": 5.907462686567165, + "grad_norm": 39.171661376953125, + "learning_rate": 7.1309523809523814e-06, + "loss": 48.943, + "step": 248 + }, + { + "epoch": 5.931343283582089, + "grad_norm": 28.724523544311523, + "learning_rate": 7.11904761904762e-06, + "loss": 50.8039, + "step": 249 + }, + { + "epoch": 5.955223880597015, + "grad_norm": 36.57830810546875, + "learning_rate": 7.107142857142858e-06, + "loss": 50.2311, + "step": 250 + }, + { + "epoch": 5.979104477611941, + "grad_norm": 39.91551971435547, + "learning_rate": 7.095238095238096e-06, + "loss": 49.1617, + "step": 251 + }, + { + "epoch": 6.0, + "grad_norm": 17.104145050048828, + "learning_rate": 7.083333333333335e-06, + "loss": 42.8003, + "step": 252 + }, + { + "epoch": 6.023880597014926, + "grad_norm": 33.03441619873047, + "learning_rate": 7.0714285714285726e-06, + "loss": 48.2896, + "step": 253 + }, + { + "epoch": 6.04776119402985, + "grad_norm": 26.487470626831055, + "learning_rate": 7.0595238095238105e-06, + "loss": 49.205, + "step": 254 + }, + { + "epoch": 6.071641791044776, + "grad_norm": 26.752981185913086, + "learning_rate": 7.047619047619048e-06, + "loss": 50.3943, + "step": 255 + }, + { + "epoch": 6.095522388059702, + "grad_norm": 22.44376564025879, + "learning_rate": 7.035714285714287e-06, + "loss": 49.285, + "step": 256 + }, + { + "epoch": 6.119402985074627, + "grad_norm": 31.066368103027344, + "learning_rate": 7.023809523809524e-06, + "loss": 49.3131, + "step": 257 + }, + { + "epoch": 6.143283582089552, + "grad_norm": 28.67262840270996, + "learning_rate": 7.011904761904762e-06, + "loss": 50.6188, + "step": 258 + }, + { + "epoch": 6.167164179104478, + "grad_norm": 24.013134002685547, + "learning_rate": 7e-06, + "loss": 50.4382, + "step": 259 + }, + { + "epoch": 6.191044776119403, + "grad_norm": 26.5673828125, + "learning_rate": 6.988095238095239e-06, + "loss": 49.7058, + "step": 260 + }, + { + "epoch": 6.214925373134328, + "grad_norm": 20.803695678710938, + "learning_rate": 6.9761904761904765e-06, + "loss": 48.9389, + "step": 261 + }, + { + "epoch": 6.2388059701492535, + "grad_norm": 23.450183868408203, + "learning_rate": 6.964285714285714e-06, + "loss": 49.0091, + "step": 262 + }, + { + "epoch": 6.262686567164179, + "grad_norm": 36.94446563720703, + "learning_rate": 6.952380952380952e-06, + "loss": 50.2589, + "step": 263 + }, + { + "epoch": 6.286567164179105, + "grad_norm": 39.548095703125, + "learning_rate": 6.940476190476191e-06, + "loss": 49.3129, + "step": 264 + }, + { + "epoch": 6.3104477611940295, + "grad_norm": 30.536083221435547, + "learning_rate": 6.928571428571429e-06, + "loss": 49.1838, + "step": 265 + }, + { + "epoch": 6.334328358208955, + "grad_norm": 27.97296714782715, + "learning_rate": 6.916666666666667e-06, + "loss": 50.3184, + "step": 266 + }, + { + "epoch": 6.358208955223881, + "grad_norm": 25.69655418395996, + "learning_rate": 6.9047619047619055e-06, + "loss": 49.2226, + "step": 267 + }, + { + "epoch": 6.382089552238806, + "grad_norm": 22.114097595214844, + "learning_rate": 6.892857142857143e-06, + "loss": 49.4455, + "step": 268 + }, + { + "epoch": 6.405970149253731, + "grad_norm": 30.47511100769043, + "learning_rate": 6.880952380952381e-06, + "loss": 49.7409, + "step": 269 + }, + { + "epoch": 6.429850746268657, + "grad_norm": 26.32929039001465, + "learning_rate": 6.86904761904762e-06, + "loss": 50.3336, + "step": 270 + }, + { + "epoch": 6.453731343283582, + "grad_norm": 28.09309196472168, + "learning_rate": 6.857142857142858e-06, + "loss": 49.6044, + "step": 271 + }, + { + "epoch": 6.477611940298507, + "grad_norm": 25.840974807739258, + "learning_rate": 6.845238095238096e-06, + "loss": 49.9185, + "step": 272 + }, + { + "epoch": 6.501492537313433, + "grad_norm": 31.89126205444336, + "learning_rate": 6.833333333333334e-06, + "loss": 48.2732, + "step": 273 + }, + { + "epoch": 6.525373134328358, + "grad_norm": 24.013029098510742, + "learning_rate": 6.8214285714285724e-06, + "loss": 49.9752, + "step": 274 + }, + { + "epoch": 6.549253731343284, + "grad_norm": 25.509836196899414, + "learning_rate": 6.80952380952381e-06, + "loss": 50.5493, + "step": 275 + }, + { + "epoch": 6.573134328358209, + "grad_norm": 35.25442886352539, + "learning_rate": 6.797619047619048e-06, + "loss": 49.2553, + "step": 276 + }, + { + "epoch": 6.597014925373134, + "grad_norm": 29.42585563659668, + "learning_rate": 6.785714285714287e-06, + "loss": 48.776, + "step": 277 + }, + { + "epoch": 6.62089552238806, + "grad_norm": 25.90894889831543, + "learning_rate": 6.773809523809525e-06, + "loss": 49.1964, + "step": 278 + }, + { + "epoch": 6.6447761194029855, + "grad_norm": 25.63600730895996, + "learning_rate": 6.761904761904763e-06, + "loss": 48.4528, + "step": 279 + }, + { + "epoch": 6.66865671641791, + "grad_norm": 29.943740844726562, + "learning_rate": 6.750000000000001e-06, + "loss": 49.9026, + "step": 280 + }, + { + "epoch": 6.692537313432836, + "grad_norm": 33.253910064697266, + "learning_rate": 6.738095238095239e-06, + "loss": 49.2364, + "step": 281 + }, + { + "epoch": 6.7164179104477615, + "grad_norm": 23.465354919433594, + "learning_rate": 6.726190476190477e-06, + "loss": 49.2759, + "step": 282 + }, + { + "epoch": 6.740298507462686, + "grad_norm": 31.023218154907227, + "learning_rate": 6.714285714285714e-06, + "loss": 49.3256, + "step": 283 + }, + { + "epoch": 6.764179104477612, + "grad_norm": 32.376991271972656, + "learning_rate": 6.702380952380952e-06, + "loss": 47.7239, + "step": 284 + }, + { + "epoch": 6.7880597014925375, + "grad_norm": 18.388896942138672, + "learning_rate": 6.690476190476191e-06, + "loss": 49.5751, + "step": 285 + }, + { + "epoch": 6.811940298507462, + "grad_norm": 22.1639404296875, + "learning_rate": 6.678571428571429e-06, + "loss": 48.8153, + "step": 286 + }, + { + "epoch": 6.835820895522388, + "grad_norm": 31.39455223083496, + "learning_rate": 6.666666666666667e-06, + "loss": 49.7872, + "step": 287 + }, + { + "epoch": 6.859701492537313, + "grad_norm": 36.480533599853516, + "learning_rate": 6.654761904761905e-06, + "loss": 48.9679, + "step": 288 + }, + { + "epoch": 6.883582089552239, + "grad_norm": 23.432872772216797, + "learning_rate": 6.642857142857143e-06, + "loss": 48.4627, + "step": 289 + }, + { + "epoch": 6.907462686567165, + "grad_norm": 33.31097412109375, + "learning_rate": 6.630952380952381e-06, + "loss": 49.617, + "step": 290 + }, + { + "epoch": 6.931343283582089, + "grad_norm": 34.07685852050781, + "learning_rate": 6.619047619047619e-06, + "loss": 49.3365, + "step": 291 + }, + { + "epoch": 6.955223880597015, + "grad_norm": 25.68811798095703, + "learning_rate": 6.607142857142858e-06, + "loss": 49.7821, + "step": 292 + }, + { + "epoch": 6.979104477611941, + "grad_norm": 24.179588317871094, + "learning_rate": 6.595238095238096e-06, + "loss": 48.5466, + "step": 293 + }, + { + "epoch": 7.0, + "grad_norm": 26.158781051635742, + "learning_rate": 6.5833333333333335e-06, + "loss": 43.2838, + "step": 294 + }, + { + "epoch": 7.023880597014926, + "grad_norm": 22.84689712524414, + "learning_rate": 6.571428571428572e-06, + "loss": 49.3253, + "step": 295 + }, + { + "epoch": 7.04776119402985, + "grad_norm": NaN, + "learning_rate": 6.55952380952381e-06, + "loss": 75.2786, + "step": 296 + }, + { + "epoch": 7.071641791044776, + "grad_norm": 31.13886070251465, + "learning_rate": 6.55952380952381e-06, + "loss": 49.984, + "step": 297 + }, + { + "epoch": 7.095522388059702, + "grad_norm": 32.37982940673828, + "learning_rate": 6.547619047619048e-06, + "loss": 49.6632, + "step": 298 + }, + { + "epoch": 7.119402985074627, + "grad_norm": 22.977916717529297, + "learning_rate": 6.535714285714286e-06, + "loss": 48.7802, + "step": 299 + }, + { + "epoch": 7.143283582089552, + "grad_norm": NaN, + "learning_rate": 6.523809523809525e-06, + "loss": 60.3381, + "step": 300 + }, + { + "epoch": 7.167164179104478, + "grad_norm": 32.18650817871094, + "learning_rate": 6.523809523809525e-06, + "loss": 49.2689, + "step": 301 + }, + { + "epoch": 7.191044776119403, + "grad_norm": 30.0800724029541, + "learning_rate": 6.5119047619047626e-06, + "loss": 49.3891, + "step": 302 + }, + { + "epoch": 7.214925373134328, + "grad_norm": 32.35110855102539, + "learning_rate": 6.5000000000000004e-06, + "loss": 48.4497, + "step": 303 + }, + { + "epoch": 7.2388059701492535, + "grad_norm": 34.08786392211914, + "learning_rate": 6.488095238095239e-06, + "loss": 49.1321, + "step": 304 + }, + { + "epoch": 7.262686567164179, + "grad_norm": 25.25969696044922, + "learning_rate": 6.476190476190477e-06, + "loss": 49.0524, + "step": 305 + }, + { + "epoch": 7.286567164179105, + "grad_norm": 25.843929290771484, + "learning_rate": 6.464285714285715e-06, + "loss": 49.8077, + "step": 306 + }, + { + "epoch": 7.3104477611940295, + "grad_norm": 34.57284927368164, + "learning_rate": 6.452380952380954e-06, + "loss": 49.5393, + "step": 307 + }, + { + "epoch": 7.334328358208955, + "grad_norm": 33.44814682006836, + "learning_rate": 6.4404761904761916e-06, + "loss": 49.0375, + "step": 308 + }, + { + "epoch": 7.358208955223881, + "grad_norm": 25.127429962158203, + "learning_rate": 6.4285714285714295e-06, + "loss": 48.8145, + "step": 309 + }, + { + "epoch": 7.382089552238806, + "grad_norm": 31.81999969482422, + "learning_rate": 6.416666666666667e-06, + "loss": 49.6432, + "step": 310 + }, + { + "epoch": 7.405970149253731, + "grad_norm": 22.428335189819336, + "learning_rate": 6.404761904761904e-06, + "loss": 47.6232, + "step": 311 + }, + { + "epoch": 7.429850746268657, + "grad_norm": 45.87803268432617, + "learning_rate": 6.392857142857143e-06, + "loss": 48.3479, + "step": 312 + }, + { + "epoch": 7.453731343283582, + "grad_norm": 37.441253662109375, + "learning_rate": 6.380952380952381e-06, + "loss": 48.593, + "step": 313 + }, + { + "epoch": 7.477611940298507, + "grad_norm": 23.15785789489746, + "learning_rate": 6.369047619047619e-06, + "loss": 49.1204, + "step": 314 + }, + { + "epoch": 7.501492537313433, + "grad_norm": 35.8905029296875, + "learning_rate": 6.357142857142858e-06, + "loss": 49.2918, + "step": 315 + }, + { + "epoch": 7.525373134328358, + "grad_norm": 37.41954040527344, + "learning_rate": 6.3452380952380955e-06, + "loss": 47.7495, + "step": 316 + }, + { + "epoch": 7.549253731343284, + "grad_norm": 31.173114776611328, + "learning_rate": 6.333333333333333e-06, + "loss": 49.539, + "step": 317 + }, + { + "epoch": 7.573134328358209, + "grad_norm": 23.941965103149414, + "learning_rate": 6.321428571428571e-06, + "loss": 49.0958, + "step": 318 + }, + { + "epoch": 7.597014925373134, + "grad_norm": 31.949769973754883, + "learning_rate": 6.30952380952381e-06, + "loss": 49.1945, + "step": 319 + }, + { + "epoch": 7.62089552238806, + "grad_norm": 21.299409866333008, + "learning_rate": 6.297619047619048e-06, + "loss": 49.3823, + "step": 320 + }, + { + "epoch": 7.6447761194029855, + "grad_norm": 34.93647766113281, + "learning_rate": 6.285714285714286e-06, + "loss": 48.8867, + "step": 321 + }, + { + "epoch": 7.66865671641791, + "grad_norm": 30.189655303955078, + "learning_rate": 6.2738095238095245e-06, + "loss": 49.8644, + "step": 322 + }, + { + "epoch": 7.692537313432836, + "grad_norm": 19.964523315429688, + "learning_rate": 6.261904761904762e-06, + "loss": 49.6489, + "step": 323 + }, + { + "epoch": 7.7164179104477615, + "grad_norm": 22.253337860107422, + "learning_rate": 6.25e-06, + "loss": 48.0582, + "step": 324 + }, + { + "epoch": 7.740298507462686, + "grad_norm": 26.631391525268555, + "learning_rate": 6.238095238095239e-06, + "loss": 48.5585, + "step": 325 + }, + { + "epoch": 7.764179104477612, + "grad_norm": 26.0469913482666, + "learning_rate": 6.226190476190477e-06, + "loss": 49.4969, + "step": 326 + }, + { + "epoch": 7.7880597014925375, + "grad_norm": 30.000507354736328, + "learning_rate": 6.214285714285715e-06, + "loss": 49.6044, + "step": 327 + }, + { + "epoch": 7.811940298507462, + "grad_norm": 29.44800567626953, + "learning_rate": 6.202380952380953e-06, + "loss": 50.3622, + "step": 328 + }, + { + "epoch": 7.835820895522388, + "grad_norm": 24.83717918395996, + "learning_rate": 6.1904761904761914e-06, + "loss": 50.0974, + "step": 329 + }, + { + "epoch": 7.859701492537313, + "grad_norm": 30.0760555267334, + "learning_rate": 6.178571428571429e-06, + "loss": 48.9307, + "step": 330 + }, + { + "epoch": 7.883582089552239, + "grad_norm": 21.087966918945312, + "learning_rate": 6.166666666666667e-06, + "loss": 49.3432, + "step": 331 + }, + { + "epoch": 7.907462686567165, + "grad_norm": 23.193716049194336, + "learning_rate": 6.154761904761906e-06, + "loss": 48.6664, + "step": 332 + }, + { + "epoch": 7.931343283582089, + "grad_norm": 22.764123916625977, + "learning_rate": 6.142857142857144e-06, + "loss": 49.3497, + "step": 333 + }, + { + "epoch": 7.955223880597015, + "grad_norm": 22.411897659301758, + "learning_rate": 6.130952380952382e-06, + "loss": 49.4106, + "step": 334 + }, + { + "epoch": 7.979104477611941, + "grad_norm": 29.535375595092773, + "learning_rate": 6.11904761904762e-06, + "loss": 49.0695, + "step": 335 + }, + { + "epoch": 8.0, + "grad_norm": 21.094457626342773, + "learning_rate": 6.107142857142858e-06, + "loss": 42.1367, + "step": 336 + }, + { + "epoch": 8.023880597014925, + "grad_norm": 33.74859619140625, + "learning_rate": 6.095238095238096e-06, + "loss": 47.0065, + "step": 337 + }, + { + "epoch": 8.047761194029851, + "grad_norm": 32.539127349853516, + "learning_rate": 6.083333333333333e-06, + "loss": 47.9697, + "step": 338 + }, + { + "epoch": 8.071641791044776, + "grad_norm": 19.168655395507812, + "learning_rate": 6.071428571428571e-06, + "loss": 49.4919, + "step": 339 + }, + { + "epoch": 8.0955223880597, + "grad_norm": 30.041269302368164, + "learning_rate": 6.05952380952381e-06, + "loss": 48.7887, + "step": 340 + }, + { + "epoch": 8.119402985074627, + "grad_norm": 21.070598602294922, + "learning_rate": 6.047619047619048e-06, + "loss": 48.5064, + "step": 341 + }, + { + "epoch": 8.143283582089552, + "grad_norm": 29.560287475585938, + "learning_rate": 6.035714285714286e-06, + "loss": 47.7472, + "step": 342 + }, + { + "epoch": 8.167164179104478, + "grad_norm": 24.256393432617188, + "learning_rate": 6.023809523809524e-06, + "loss": 48.8917, + "step": 343 + }, + { + "epoch": 8.191044776119403, + "grad_norm": 29.970674514770508, + "learning_rate": 6.011904761904762e-06, + "loss": 48.3464, + "step": 344 + }, + { + "epoch": 8.214925373134328, + "grad_norm": 25.274595260620117, + "learning_rate": 6e-06, + "loss": 49.1565, + "step": 345 + }, + { + "epoch": 8.238805970149254, + "grad_norm": 24.000280380249023, + "learning_rate": 5.988095238095238e-06, + "loss": 49.3396, + "step": 346 + }, + { + "epoch": 8.26268656716418, + "grad_norm": 25.110261917114258, + "learning_rate": 5.976190476190477e-06, + "loss": 49.786, + "step": 347 + }, + { + "epoch": 8.286567164179104, + "grad_norm": 26.188514709472656, + "learning_rate": 5.964285714285715e-06, + "loss": 50.1652, + "step": 348 + }, + { + "epoch": 8.31044776119403, + "grad_norm": 18.536714553833008, + "learning_rate": 5.9523809523809525e-06, + "loss": 49.7224, + "step": 349 + }, + { + "epoch": 8.334328358208955, + "grad_norm": 33.79502868652344, + "learning_rate": 5.940476190476191e-06, + "loss": 48.2923, + "step": 350 + }, + { + "epoch": 8.35820895522388, + "grad_norm": 33.03609085083008, + "learning_rate": 5.928571428571429e-06, + "loss": 49.128, + "step": 351 + }, + { + "epoch": 8.382089552238806, + "grad_norm": 23.88555145263672, + "learning_rate": 5.916666666666667e-06, + "loss": 49.6072, + "step": 352 + }, + { + "epoch": 8.405970149253731, + "grad_norm": 29.688135147094727, + "learning_rate": 5.904761904761905e-06, + "loss": 49.0984, + "step": 353 + }, + { + "epoch": 8.429850746268656, + "grad_norm": 23.166162490844727, + "learning_rate": 5.892857142857144e-06, + "loss": 48.8104, + "step": 354 + }, + { + "epoch": 8.453731343283582, + "grad_norm": 27.68876838684082, + "learning_rate": 5.8809523809523816e-06, + "loss": 48.7745, + "step": 355 + }, + { + "epoch": 8.477611940298507, + "grad_norm": 26.520286560058594, + "learning_rate": 5.8690476190476194e-06, + "loss": 47.883, + "step": 356 + }, + { + "epoch": 8.501492537313434, + "grad_norm": 28.830135345458984, + "learning_rate": 5.857142857142858e-06, + "loss": 49.1347, + "step": 357 + }, + { + "epoch": 8.525373134328358, + "grad_norm": 27.387250900268555, + "learning_rate": 5.845238095238096e-06, + "loss": 48.2092, + "step": 358 + }, + { + "epoch": 8.549253731343283, + "grad_norm": 23.53616714477539, + "learning_rate": 5.833333333333334e-06, + "loss": 48.437, + "step": 359 + }, + { + "epoch": 8.57313432835821, + "grad_norm": 25.665664672851562, + "learning_rate": 5.821428571428573e-06, + "loss": 49.3006, + "step": 360 + }, + { + "epoch": 8.597014925373134, + "grad_norm": 24.35331153869629, + "learning_rate": 5.8095238095238106e-06, + "loss": 49.5249, + "step": 361 + }, + { + "epoch": 8.620895522388059, + "grad_norm": 28.612688064575195, + "learning_rate": 5.7976190476190485e-06, + "loss": 50.1344, + "step": 362 + }, + { + "epoch": 8.644776119402986, + "grad_norm": 25.055545806884766, + "learning_rate": 5.785714285714286e-06, + "loss": 48.6014, + "step": 363 + }, + { + "epoch": 8.66865671641791, + "grad_norm": 27.645490646362305, + "learning_rate": 5.773809523809523e-06, + "loss": 48.953, + "step": 364 + }, + { + "epoch": 8.692537313432837, + "grad_norm": 26.791471481323242, + "learning_rate": 5.761904761904762e-06, + "loss": 49.5912, + "step": 365 + }, + { + "epoch": 8.716417910447761, + "grad_norm": 27.57213592529297, + "learning_rate": 5.75e-06, + "loss": 48.9958, + "step": 366 + }, + { + "epoch": 8.740298507462686, + "grad_norm": 20.936344146728516, + "learning_rate": 5.738095238095238e-06, + "loss": 48.3449, + "step": 367 + }, + { + "epoch": 8.764179104477613, + "grad_norm": 31.695810317993164, + "learning_rate": 5.726190476190477e-06, + "loss": 49.1015, + "step": 368 + }, + { + "epoch": 8.788059701492537, + "grad_norm": 31.584064483642578, + "learning_rate": 5.7142857142857145e-06, + "loss": 48.8249, + "step": 369 + }, + { + "epoch": 8.811940298507462, + "grad_norm": 30.70412826538086, + "learning_rate": 5.702380952380952e-06, + "loss": 49.2984, + "step": 370 + }, + { + "epoch": 8.835820895522389, + "grad_norm": 36.31315231323242, + "learning_rate": 5.690476190476191e-06, + "loss": 48.6769, + "step": 371 + }, + { + "epoch": 8.859701492537313, + "grad_norm": 28.98838996887207, + "learning_rate": 5.678571428571429e-06, + "loss": 50.2101, + "step": 372 + }, + { + "epoch": 8.883582089552238, + "grad_norm": 29.07052230834961, + "learning_rate": 5.666666666666667e-06, + "loss": 49.9206, + "step": 373 + }, + { + "epoch": 8.907462686567165, + "grad_norm": 31.653087615966797, + "learning_rate": 5.654761904761905e-06, + "loss": 48.3035, + "step": 374 + }, + { + "epoch": 8.93134328358209, + "grad_norm": 27.019704818725586, + "learning_rate": 5.6428571428571435e-06, + "loss": 48.6833, + "step": 375 + }, + { + "epoch": 8.955223880597014, + "grad_norm": 30.919578552246094, + "learning_rate": 5.630952380952381e-06, + "loss": 47.7973, + "step": 376 + }, + { + "epoch": 8.97910447761194, + "grad_norm": 28.002975463867188, + "learning_rate": 5.619047619047619e-06, + "loss": 49.5539, + "step": 377 + }, + { + "epoch": 9.0, + "grad_norm": 27.587263107299805, + "learning_rate": 5.607142857142858e-06, + "loss": 42.9343, + "step": 378 + }, + { + "epoch": 9.023880597014925, + "grad_norm": 31.024024963378906, + "learning_rate": 5.595238095238096e-06, + "loss": 48.6774, + "step": 379 + }, + { + "epoch": 9.047761194029851, + "grad_norm": 27.262426376342773, + "learning_rate": 5.583333333333334e-06, + "loss": 47.8833, + "step": 380 + }, + { + "epoch": 9.071641791044776, + "grad_norm": 29.223133087158203, + "learning_rate": 5.571428571428572e-06, + "loss": 47.1563, + "step": 381 + }, + { + "epoch": 9.0955223880597, + "grad_norm": 21.004749298095703, + "learning_rate": 5.5595238095238104e-06, + "loss": 49.0407, + "step": 382 + }, + { + "epoch": 9.119402985074627, + "grad_norm": 25.157907485961914, + "learning_rate": 5.547619047619048e-06, + "loss": 48.5233, + "step": 383 + }, + { + "epoch": 9.143283582089552, + "grad_norm": 17.611478805541992, + "learning_rate": 5.535714285714286e-06, + "loss": 47.4846, + "step": 384 + }, + { + "epoch": 9.167164179104478, + "grad_norm": 21.00395965576172, + "learning_rate": 5.523809523809525e-06, + "loss": 49.7533, + "step": 385 + }, + { + "epoch": 9.191044776119403, + "grad_norm": 22.07697296142578, + "learning_rate": 5.511904761904763e-06, + "loss": 48.5003, + "step": 386 + }, + { + "epoch": 9.214925373134328, + "grad_norm": 21.743778228759766, + "learning_rate": 5.500000000000001e-06, + "loss": 48.149, + "step": 387 + }, + { + "epoch": 9.238805970149254, + "grad_norm": 23.499980926513672, + "learning_rate": 5.4880952380952394e-06, + "loss": 48.2213, + "step": 388 + }, + { + "epoch": 9.26268656716418, + "grad_norm": 22.22580337524414, + "learning_rate": 5.476190476190477e-06, + "loss": 48.4671, + "step": 389 + }, + { + "epoch": 9.286567164179104, + "grad_norm": 26.5915470123291, + "learning_rate": 5.464285714285714e-06, + "loss": 49.2343, + "step": 390 + }, + { + "epoch": 9.31044776119403, + "grad_norm": 22.510892868041992, + "learning_rate": 5.452380952380952e-06, + "loss": 48.9363, + "step": 391 + }, + { + "epoch": 9.334328358208955, + "grad_norm": 27.17405128479004, + "learning_rate": 5.44047619047619e-06, + "loss": 49.1814, + "step": 392 + }, + { + "epoch": 9.35820895522388, + "grad_norm": 29.143529891967773, + "learning_rate": 5.428571428571429e-06, + "loss": 48.4786, + "step": 393 + }, + { + "epoch": 9.382089552238806, + "grad_norm": 20.24784278869629, + "learning_rate": 5.416666666666667e-06, + "loss": 49.2987, + "step": 394 + }, + { + "epoch": 9.405970149253731, + "grad_norm": 31.44426155090332, + "learning_rate": 5.404761904761905e-06, + "loss": 49.9466, + "step": 395 + }, + { + "epoch": 9.429850746268656, + "grad_norm": 23.775951385498047, + "learning_rate": 5.392857142857143e-06, + "loss": 49.1681, + "step": 396 + }, + { + "epoch": 9.453731343283582, + "grad_norm": 22.168636322021484, + "learning_rate": 5.380952380952381e-06, + "loss": 48.8523, + "step": 397 + }, + { + "epoch": 9.477611940298507, + "grad_norm": 20.944936752319336, + "learning_rate": 5.369047619047619e-06, + "loss": 48.7369, + "step": 398 + }, + { + "epoch": 9.501492537313434, + "grad_norm": 23.880292892456055, + "learning_rate": 5.357142857142857e-06, + "loss": 48.4703, + "step": 399 + }, + { + "epoch": 9.525373134328358, + "grad_norm": 25.316978454589844, + "learning_rate": 5.345238095238096e-06, + "loss": 48.3752, + "step": 400 + }, + { + "epoch": 9.549253731343283, + "grad_norm": 24.398311614990234, + "learning_rate": 5.333333333333334e-06, + "loss": 47.532, + "step": 401 + }, + { + "epoch": 9.57313432835821, + "grad_norm": 23.157140731811523, + "learning_rate": 5.3214285714285715e-06, + "loss": 49.1824, + "step": 402 + }, + { + "epoch": 9.597014925373134, + "grad_norm": 21.641061782836914, + "learning_rate": 5.30952380952381e-06, + "loss": 49.6601, + "step": 403 + }, + { + "epoch": 9.620895522388059, + "grad_norm": 23.863712310791016, + "learning_rate": 5.297619047619048e-06, + "loss": 49.2146, + "step": 404 + }, + { + "epoch": 9.644776119402986, + "grad_norm": 21.876007080078125, + "learning_rate": 5.285714285714286e-06, + "loss": 48.0027, + "step": 405 + }, + { + "epoch": 9.66865671641791, + "grad_norm": 25.783042907714844, + "learning_rate": 5.273809523809525e-06, + "loss": 48.2702, + "step": 406 + }, + { + "epoch": 9.692537313432837, + "grad_norm": 18.782087326049805, + "learning_rate": 5.261904761904763e-06, + "loss": 48.9365, + "step": 407 + }, + { + "epoch": 9.716417910447761, + "grad_norm": 20.206588745117188, + "learning_rate": 5.2500000000000006e-06, + "loss": 49.4144, + "step": 408 + }, + { + "epoch": 9.740298507462686, + "grad_norm": 20.98710823059082, + "learning_rate": 5.2380952380952384e-06, + "loss": 49.7442, + "step": 409 + }, + { + "epoch": 9.764179104477613, + "grad_norm": 19.24452018737793, + "learning_rate": 5.226190476190477e-06, + "loss": 49.0249, + "step": 410 + }, + { + "epoch": 9.788059701492537, + "grad_norm": 23.18075180053711, + "learning_rate": 5.214285714285715e-06, + "loss": 48.8795, + "step": 411 + }, + { + "epoch": 9.811940298507462, + "grad_norm": 17.233261108398438, + "learning_rate": 5.202380952380953e-06, + "loss": 49.2985, + "step": 412 + }, + { + "epoch": 9.835820895522389, + "grad_norm": 24.74007797241211, + "learning_rate": 5.190476190476192e-06, + "loss": 48.8793, + "step": 413 + }, + { + "epoch": 9.859701492537313, + "grad_norm": 20.26863670349121, + "learning_rate": 5.1785714285714296e-06, + "loss": 49.6989, + "step": 414 + }, + { + "epoch": 9.883582089552238, + "grad_norm": 26.168167114257812, + "learning_rate": 5.1666666666666675e-06, + "loss": 48.7413, + "step": 415 + }, + { + "epoch": 9.907462686567165, + "grad_norm": 29.008501052856445, + "learning_rate": 5.1547619047619045e-06, + "loss": 48.7414, + "step": 416 + }, + { + "epoch": 9.93134328358209, + "grad_norm": 18.459829330444336, + "learning_rate": 5.142857142857142e-06, + "loss": 47.7865, + "step": 417 + }, + { + "epoch": 9.955223880597014, + "grad_norm": 20.898181915283203, + "learning_rate": 5.130952380952381e-06, + "loss": 47.7274, + "step": 418 + }, + { + "epoch": 9.97910447761194, + "grad_norm": 23.5065860748291, + "learning_rate": 5.119047619047619e-06, + "loss": 48.471, + "step": 419 + }, + { + "epoch": 10.0, + "grad_norm": 23.147043228149414, + "learning_rate": 5.107142857142857e-06, + "loss": 42.3971, + "step": 420 + }, + { + "epoch": 10.023880597014925, + "grad_norm": 28.423707962036133, + "learning_rate": 5.095238095238096e-06, + "loss": 49.4977, + "step": 421 + }, + { + "epoch": 10.047761194029851, + "grad_norm": 22.017820358276367, + "learning_rate": 5.0833333333333335e-06, + "loss": 47.0638, + "step": 422 + }, + { + "epoch": 10.071641791044776, + "grad_norm": 18.173845291137695, + "learning_rate": 5.071428571428571e-06, + "loss": 48.338, + "step": 423 + }, + { + "epoch": 10.0955223880597, + "grad_norm": 17.628551483154297, + "learning_rate": 5.05952380952381e-06, + "loss": 48.2847, + "step": 424 + }, + { + "epoch": 10.119402985074627, + "grad_norm": 19.974040985107422, + "learning_rate": 5.047619047619048e-06, + "loss": 49.2284, + "step": 425 + }, + { + "epoch": 10.143283582089552, + "grad_norm": 22.45549774169922, + "learning_rate": 5.035714285714286e-06, + "loss": 49.6345, + "step": 426 + }, + { + "epoch": 10.167164179104478, + "grad_norm": 21.609479904174805, + "learning_rate": 5.023809523809524e-06, + "loss": 48.2098, + "step": 427 + }, + { + "epoch": 10.191044776119403, + "grad_norm": 24.7137451171875, + "learning_rate": 5.0119047619047625e-06, + "loss": 47.9527, + "step": 428 + }, + { + "epoch": 10.214925373134328, + "grad_norm": 22.888975143432617, + "learning_rate": 5e-06, + "loss": 49.781, + "step": 429 + }, + { + "epoch": 10.238805970149254, + "grad_norm": 25.53217124938965, + "learning_rate": 4.988095238095238e-06, + "loss": 48.9902, + "step": 430 + }, + { + "epoch": 10.26268656716418, + "grad_norm": 27.80384063720703, + "learning_rate": 4.976190476190477e-06, + "loss": 48.2545, + "step": 431 + }, + { + "epoch": 10.286567164179104, + "grad_norm": 21.421342849731445, + "learning_rate": 4.964285714285715e-06, + "loss": 49.1483, + "step": 432 + }, + { + "epoch": 10.31044776119403, + "grad_norm": 26.178152084350586, + "learning_rate": 4.952380952380953e-06, + "loss": 49.1129, + "step": 433 + }, + { + "epoch": 10.334328358208955, + "grad_norm": 27.993371963500977, + "learning_rate": 4.940476190476191e-06, + "loss": 48.1783, + "step": 434 + }, + { + "epoch": 10.35820895522388, + "grad_norm": 26.75821876525879, + "learning_rate": 4.928571428571429e-06, + "loss": 48.1773, + "step": 435 + }, + { + "epoch": 10.382089552238806, + "grad_norm": 25.641353607177734, + "learning_rate": 4.9166666666666665e-06, + "loss": 48.9295, + "step": 436 + }, + { + "epoch": 10.405970149253731, + "grad_norm": 23.26271629333496, + "learning_rate": 4.904761904761905e-06, + "loss": 49.5486, + "step": 437 + }, + { + "epoch": 10.429850746268656, + "grad_norm": 23.637466430664062, + "learning_rate": 4.892857142857143e-06, + "loss": 48.1263, + "step": 438 + }, + { + "epoch": 10.453731343283582, + "grad_norm": 29.285432815551758, + "learning_rate": 4.880952380952381e-06, + "loss": 48.2424, + "step": 439 + }, + { + "epoch": 10.477611940298507, + "grad_norm": 29.91914939880371, + "learning_rate": 4.86904761904762e-06, + "loss": 48.3695, + "step": 440 + }, + { + "epoch": 10.501492537313434, + "grad_norm": 25.249099731445312, + "learning_rate": 4.857142857142858e-06, + "loss": 48.3644, + "step": 441 + }, + { + "epoch": 10.525373134328358, + "grad_norm": 22.37591552734375, + "learning_rate": 4.8452380952380955e-06, + "loss": 49.397, + "step": 442 + }, + { + "epoch": 10.549253731343283, + "grad_norm": 22.805437088012695, + "learning_rate": 4.833333333333333e-06, + "loss": 48.6522, + "step": 443 + }, + { + "epoch": 10.57313432835821, + "grad_norm": 21.229095458984375, + "learning_rate": 4.821428571428572e-06, + "loss": 47.6681, + "step": 444 + }, + { + "epoch": 10.597014925373134, + "grad_norm": 23.359468460083008, + "learning_rate": 4.80952380952381e-06, + "loss": 48.602, + "step": 445 + }, + { + "epoch": 10.620895522388059, + "grad_norm": 20.953310012817383, + "learning_rate": 4.797619047619048e-06, + "loss": 49.3366, + "step": 446 + }, + { + "epoch": 10.644776119402986, + "grad_norm": 21.970388412475586, + "learning_rate": 4.785714285714287e-06, + "loss": 46.964, + "step": 447 + }, + { + "epoch": 10.66865671641791, + "grad_norm": 24.282426834106445, + "learning_rate": 4.7738095238095245e-06, + "loss": 48.2676, + "step": 448 + }, + { + "epoch": 10.692537313432837, + "grad_norm": 15.47967529296875, + "learning_rate": 4.761904761904762e-06, + "loss": 48.1993, + "step": 449 + }, + { + "epoch": 10.716417910447761, + "grad_norm": 23.230947494506836, + "learning_rate": 4.75e-06, + "loss": 48.5229, + "step": 450 + }, + { + "epoch": 10.740298507462686, + "grad_norm": 20.514225006103516, + "learning_rate": 4.738095238095238e-06, + "loss": 48.062, + "step": 451 + }, + { + "epoch": 10.764179104477613, + "grad_norm": 19.060667037963867, + "learning_rate": 4.726190476190476e-06, + "loss": 48.3893, + "step": 452 + }, + { + "epoch": 10.788059701492537, + "grad_norm": 29.78558349609375, + "learning_rate": 4.714285714285715e-06, + "loss": 48.9921, + "step": 453 + }, + { + "epoch": 10.811940298507462, + "grad_norm": 23.262001037597656, + "learning_rate": 4.702380952380953e-06, + "loss": 48.5597, + "step": 454 + }, + { + "epoch": 10.835820895522389, + "grad_norm": 25.83403778076172, + "learning_rate": 4.6904761904761905e-06, + "loss": 49.2911, + "step": 455 + }, + { + "epoch": 10.859701492537313, + "grad_norm": 21.846391677856445, + "learning_rate": 4.678571428571429e-06, + "loss": 47.3256, + "step": 456 + }, + { + "epoch": 10.883582089552238, + "grad_norm": 17.09532356262207, + "learning_rate": 4.666666666666667e-06, + "loss": 48.3647, + "step": 457 + }, + { + "epoch": 10.907462686567165, + "grad_norm": 31.050525665283203, + "learning_rate": 4.654761904761905e-06, + "loss": 48.3605, + "step": 458 + }, + { + "epoch": 10.93134328358209, + "grad_norm": 22.532379150390625, + "learning_rate": 4.642857142857144e-06, + "loss": 49.0826, + "step": 459 + }, + { + "epoch": 10.955223880597014, + "grad_norm": 23.585033416748047, + "learning_rate": 4.630952380952382e-06, + "loss": 48.5111, + "step": 460 + }, + { + "epoch": 10.97910447761194, + "grad_norm": NaN, + "learning_rate": 4.6190476190476196e-06, + "loss": 66.9717, + "step": 461 + }, + { + "epoch": 11.0, + "grad_norm": 24.73590087890625, + "learning_rate": 4.6190476190476196e-06, + "loss": 41.9122, + "step": 462 + }, + { + "epoch": 11.023880597014925, + "grad_norm": 27.4709415435791, + "learning_rate": 4.6071428571428574e-06, + "loss": 48.4682, + "step": 463 + }, + { + "epoch": 11.047761194029851, + "grad_norm": 26.158245086669922, + "learning_rate": 4.595238095238095e-06, + "loss": 48.1845, + "step": 464 + }, + { + "epoch": 11.071641791044776, + "grad_norm": 25.14693260192871, + "learning_rate": 4.583333333333333e-06, + "loss": 48.4229, + "step": 465 + }, + { + "epoch": 11.0955223880597, + "grad_norm": 22.229764938354492, + "learning_rate": 4.571428571428572e-06, + "loss": 47.8876, + "step": 466 + }, + { + "epoch": 11.119402985074627, + "grad_norm": 24.202686309814453, + "learning_rate": 4.55952380952381e-06, + "loss": 48.4304, + "step": 467 + }, + { + "epoch": 11.143283582089552, + "grad_norm": 21.449726104736328, + "learning_rate": 4.547619047619048e-06, + "loss": 47.6457, + "step": 468 + }, + { + "epoch": 11.167164179104478, + "grad_norm": 23.769763946533203, + "learning_rate": 4.5357142857142865e-06, + "loss": 49.1031, + "step": 469 + }, + { + "epoch": 11.191044776119403, + "grad_norm": 21.20684814453125, + "learning_rate": 4.523809523809524e-06, + "loss": 47.6488, + "step": 470 + }, + { + "epoch": 11.214925373134328, + "grad_norm": 17.992631912231445, + "learning_rate": 4.511904761904762e-06, + "loss": 47.9435, + "step": 471 + }, + { + "epoch": 11.238805970149254, + "grad_norm": 22.017776489257812, + "learning_rate": 4.5e-06, + "loss": 48.5224, + "step": 472 + }, + { + "epoch": 11.26268656716418, + "grad_norm": 22.98673439025879, + "learning_rate": 4.488095238095239e-06, + "loss": 47.9258, + "step": 473 + }, + { + "epoch": 11.286567164179104, + "grad_norm": 16.146743774414062, + "learning_rate": 4.476190476190477e-06, + "loss": 48.3957, + "step": 474 + }, + { + "epoch": 11.31044776119403, + "grad_norm": 23.30071258544922, + "learning_rate": 4.464285714285715e-06, + "loss": 48.5472, + "step": 475 + }, + { + "epoch": 11.334328358208955, + "grad_norm": 24.949913024902344, + "learning_rate": 4.4523809523809525e-06, + "loss": 48.2387, + "step": 476 + }, + { + "epoch": 11.35820895522388, + "grad_norm": 23.10662841796875, + "learning_rate": 4.44047619047619e-06, + "loss": 49.0681, + "step": 477 + }, + { + "epoch": 11.382089552238806, + "grad_norm": 19.024614334106445, + "learning_rate": 4.428571428571429e-06, + "loss": 49.3255, + "step": 478 + }, + { + "epoch": 11.405970149253731, + "grad_norm": 22.34437370300293, + "learning_rate": 4.416666666666667e-06, + "loss": 47.0069, + "step": 479 + }, + { + "epoch": 11.429850746268656, + "grad_norm": 23.563596725463867, + "learning_rate": 4.404761904761905e-06, + "loss": 46.8188, + "step": 480 + }, + { + "epoch": 11.453731343283582, + "grad_norm": 20.5488338470459, + "learning_rate": 4.392857142857143e-06, + "loss": 47.8277, + "step": 481 + }, + { + "epoch": 11.477611940298507, + "grad_norm": 18.416519165039062, + "learning_rate": 4.3809523809523815e-06, + "loss": 48.2203, + "step": 482 + }, + { + "epoch": 11.501492537313434, + "grad_norm": 28.21132469177246, + "learning_rate": 4.369047619047619e-06, + "loss": 48.0691, + "step": 483 + }, + { + "epoch": 11.525373134328358, + "grad_norm": 21.36182975769043, + "learning_rate": 4.357142857142857e-06, + "loss": 48.273, + "step": 484 + }, + { + "epoch": 11.549253731343283, + "grad_norm": 25.726530075073242, + "learning_rate": 4.345238095238096e-06, + "loss": 48.7529, + "step": 485 + }, + { + "epoch": 11.57313432835821, + "grad_norm": 21.686412811279297, + "learning_rate": 4.333333333333334e-06, + "loss": 48.3005, + "step": 486 + }, + { + "epoch": 11.597014925373134, + "grad_norm": 20.56638526916504, + "learning_rate": 4.321428571428572e-06, + "loss": 50.1248, + "step": 487 + }, + { + "epoch": 11.620895522388059, + "grad_norm": 24.193323135375977, + "learning_rate": 4.30952380952381e-06, + "loss": 48.6031, + "step": 488 + }, + { + "epoch": 11.644776119402986, + "grad_norm": 17.18548583984375, + "learning_rate": 4.297619047619048e-06, + "loss": 49.2039, + "step": 489 + }, + { + "epoch": 11.66865671641791, + "grad_norm": 19.07050895690918, + "learning_rate": 4.2857142857142855e-06, + "loss": 48.0961, + "step": 490 + }, + { + "epoch": 11.692537313432837, + "grad_norm": 19.831188201904297, + "learning_rate": 4.273809523809524e-06, + "loss": 48.5481, + "step": 491 + }, + { + "epoch": 11.716417910447761, + "grad_norm": 23.408592224121094, + "learning_rate": 4.261904761904762e-06, + "loss": 48.583, + "step": 492 + }, + { + "epoch": 11.740298507462686, + "grad_norm": 22.152788162231445, + "learning_rate": 4.25e-06, + "loss": 48.6684, + "step": 493 + }, + { + "epoch": 11.764179104477613, + "grad_norm": NaN, + "learning_rate": 4.238095238095239e-06, + "loss": 54.7097, + "step": 494 + }, + { + "epoch": 11.788059701492537, + "grad_norm": 23.1225528717041, + "learning_rate": 4.238095238095239e-06, + "loss": 48.5439, + "step": 495 + }, + { + "epoch": 11.811940298507462, + "grad_norm": 24.673904418945312, + "learning_rate": 4.226190476190477e-06, + "loss": 48.2645, + "step": 496 + }, + { + "epoch": 11.835820895522389, + "grad_norm": 23.318784713745117, + "learning_rate": 4.2142857142857145e-06, + "loss": 47.9159, + "step": 497 + }, + { + "epoch": 11.859701492537313, + "grad_norm": 24.62889289855957, + "learning_rate": 4.202380952380952e-06, + "loss": 48.1392, + "step": 498 + }, + { + "epoch": 11.883582089552238, + "grad_norm": 17.315168380737305, + "learning_rate": 4.190476190476191e-06, + "loss": 49.399, + "step": 499 + }, + { + "epoch": 11.907462686567165, + "grad_norm": 24.458532333374023, + "learning_rate": 4.178571428571429e-06, + "loss": 49.2189, + "step": 500 + }, + { + "epoch": 11.93134328358209, + "grad_norm": 28.294036865234375, + "learning_rate": 4.166666666666667e-06, + "loss": 48.4759, + "step": 501 + }, + { + "epoch": 11.955223880597014, + "grad_norm": 22.393577575683594, + "learning_rate": 4.154761904761906e-06, + "loss": 48.9718, + "step": 502 + }, + { + "epoch": 11.97910447761194, + "grad_norm": 20.199522018432617, + "learning_rate": 4.1428571428571435e-06, + "loss": 47.5364, + "step": 503 + }, + { + "epoch": 12.0, + "grad_norm": 22.080204010009766, + "learning_rate": 4.130952380952381e-06, + "loss": 42.5308, + "step": 504 + }, + { + "epoch": 12.023880597014925, + "grad_norm": 28.897024154663086, + "learning_rate": 4.119047619047619e-06, + "loss": 48.9022, + "step": 505 + }, + { + "epoch": 12.047761194029851, + "grad_norm": 28.31342887878418, + "learning_rate": 4.107142857142857e-06, + "loss": 47.6489, + "step": 506 + }, + { + "epoch": 12.071641791044776, + "grad_norm": 22.62079620361328, + "learning_rate": 4.095238095238096e-06, + "loss": 48.1606, + "step": 507 + }, + { + "epoch": 12.0955223880597, + "grad_norm": 33.49858474731445, + "learning_rate": 4.083333333333334e-06, + "loss": 47.8462, + "step": 508 + }, + { + "epoch": 12.119402985074627, + "grad_norm": 22.20858383178711, + "learning_rate": 4.071428571428572e-06, + "loss": 47.2505, + "step": 509 + }, + { + "epoch": 12.143283582089552, + "grad_norm": 25.425495147705078, + "learning_rate": 4.0595238095238095e-06, + "loss": 48.6289, + "step": 510 + }, + { + "epoch": 12.167164179104478, + "grad_norm": 29.32784652709961, + "learning_rate": 4.047619047619048e-06, + "loss": 47.7772, + "step": 511 + }, + { + "epoch": 12.191044776119403, + "grad_norm": 20.661781311035156, + "learning_rate": 4.035714285714286e-06, + "loss": 47.1414, + "step": 512 + }, + { + "epoch": 12.214925373134328, + "grad_norm": 31.4210205078125, + "learning_rate": 4.023809523809524e-06, + "loss": 47.4312, + "step": 513 + }, + { + "epoch": 12.238805970149254, + "grad_norm": 32.390071868896484, + "learning_rate": 4.011904761904763e-06, + "loss": 49.9899, + "step": 514 + }, + { + "epoch": 12.26268656716418, + "grad_norm": 17.431835174560547, + "learning_rate": 4.000000000000001e-06, + "loss": 48.8975, + "step": 515 + }, + { + "epoch": 12.286567164179104, + "grad_norm": 29.32766342163086, + "learning_rate": 3.9880952380952386e-06, + "loss": 48.8764, + "step": 516 + }, + { + "epoch": 12.31044776119403, + "grad_norm": 29.523069381713867, + "learning_rate": 3.9761904761904764e-06, + "loss": 48.2602, + "step": 517 + }, + { + "epoch": 12.334328358208955, + "grad_norm": 23.866840362548828, + "learning_rate": 3.964285714285714e-06, + "loss": 47.4016, + "step": 518 + }, + { + "epoch": 12.35820895522388, + "grad_norm": 27.464962005615234, + "learning_rate": 3.952380952380952e-06, + "loss": 48.2559, + "step": 519 + }, + { + "epoch": 12.382089552238806, + "grad_norm": 19.796552658081055, + "learning_rate": 3.940476190476191e-06, + "loss": 48.7665, + "step": 520 + }, + { + "epoch": 12.405970149253731, + "grad_norm": 18.637983322143555, + "learning_rate": 3.928571428571429e-06, + "loss": 48.1456, + "step": 521 + }, + { + "epoch": 12.429850746268656, + "grad_norm": 22.065799713134766, + "learning_rate": 3.916666666666667e-06, + "loss": 48.7803, + "step": 522 + }, + { + "epoch": 12.453731343283582, + "grad_norm": 22.648218154907227, + "learning_rate": 3.9047619047619055e-06, + "loss": 47.3376, + "step": 523 + }, + { + "epoch": 12.477611940298507, + "grad_norm": 17.55946922302246, + "learning_rate": 3.892857142857143e-06, + "loss": 47.6002, + "step": 524 + }, + { + "epoch": 12.501492537313434, + "grad_norm": 19.173139572143555, + "learning_rate": 3.880952380952381e-06, + "loss": 48.8976, + "step": 525 + }, + { + "epoch": 12.525373134328358, + "grad_norm": 24.052696228027344, + "learning_rate": 3.869047619047619e-06, + "loss": 48.1851, + "step": 526 + }, + { + "epoch": 12.549253731343283, + "grad_norm": 19.28683090209961, + "learning_rate": 3.857142857142858e-06, + "loss": 48.0342, + "step": 527 + }, + { + "epoch": 12.57313432835821, + "grad_norm": 21.528470993041992, + "learning_rate": 3.845238095238096e-06, + "loss": 49.3597, + "step": 528 + }, + { + "epoch": 12.597014925373134, + "grad_norm": 22.880159378051758, + "learning_rate": 3.833333333333334e-06, + "loss": 47.9594, + "step": 529 + }, + { + "epoch": 12.620895522388059, + "grad_norm": 19.00438117980957, + "learning_rate": 3.8214285714285715e-06, + "loss": 47.2837, + "step": 530 + }, + { + "epoch": 12.644776119402986, + "grad_norm": 22.21845054626465, + "learning_rate": 3.80952380952381e-06, + "loss": 47.1453, + "step": 531 + }, + { + "epoch": 12.66865671641791, + "grad_norm": 18.551712036132812, + "learning_rate": 3.7976190476190477e-06, + "loss": 47.9594, + "step": 532 + }, + { + "epoch": 12.692537313432837, + "grad_norm": 17.805360794067383, + "learning_rate": 3.785714285714286e-06, + "loss": 49.1036, + "step": 533 + }, + { + "epoch": 12.716417910447761, + "grad_norm": 14.508918762207031, + "learning_rate": 3.773809523809524e-06, + "loss": 48.1203, + "step": 534 + }, + { + "epoch": 12.740298507462686, + "grad_norm": 19.395994186401367, + "learning_rate": 3.761904761904762e-06, + "loss": 47.7891, + "step": 535 + }, + { + "epoch": 12.764179104477613, + "grad_norm": 27.492908477783203, + "learning_rate": 3.7500000000000005e-06, + "loss": 48.9027, + "step": 536 + }, + { + "epoch": 12.788059701492537, + "grad_norm": 21.751968383789062, + "learning_rate": 3.7380952380952384e-06, + "loss": 48.0929, + "step": 537 + }, + { + "epoch": 12.811940298507462, + "grad_norm": 24.78274917602539, + "learning_rate": 3.7261904761904767e-06, + "loss": 48.1678, + "step": 538 + }, + { + "epoch": 12.835820895522389, + "grad_norm": 26.319196701049805, + "learning_rate": 3.7142857142857146e-06, + "loss": 49.1874, + "step": 539 + }, + { + "epoch": 12.859701492537313, + "grad_norm": 20.670148849487305, + "learning_rate": 3.702380952380953e-06, + "loss": 48.8441, + "step": 540 + }, + { + "epoch": 12.883582089552238, + "grad_norm": 23.578706741333008, + "learning_rate": 3.690476190476191e-06, + "loss": 47.1627, + "step": 541 + }, + { + "epoch": 12.907462686567165, + "grad_norm": 23.807973861694336, + "learning_rate": 3.678571428571429e-06, + "loss": 47.493, + "step": 542 + }, + { + "epoch": 12.93134328358209, + "grad_norm": 20.977373123168945, + "learning_rate": 3.6666666666666666e-06, + "loss": 49.3489, + "step": 543 + }, + { + "epoch": 12.955223880597014, + "grad_norm": 21.219995498657227, + "learning_rate": 3.654761904761905e-06, + "loss": 49.8562, + "step": 544 + }, + { + "epoch": 12.97910447761194, + "grad_norm": 17.777210235595703, + "learning_rate": 3.642857142857143e-06, + "loss": 48.4018, + "step": 545 + }, + { + "epoch": 13.0, + "grad_norm": 17.52475929260254, + "learning_rate": 3.630952380952381e-06, + "loss": 42.3621, + "step": 546 + }, + { + "epoch": 13.023880597014925, + "grad_norm": 23.431884765625, + "learning_rate": 3.6190476190476194e-06, + "loss": 49.0982, + "step": 547 + }, + { + "epoch": 13.047761194029851, + "grad_norm": 25.512338638305664, + "learning_rate": 3.6071428571428573e-06, + "loss": 47.5758, + "step": 548 + }, + { + "epoch": 13.071641791044776, + "grad_norm": 25.41205406188965, + "learning_rate": 3.5952380952380956e-06, + "loss": 49.0519, + "step": 549 + }, + { + "epoch": 13.0955223880597, + "grad_norm": 20.511945724487305, + "learning_rate": 3.5833333333333335e-06, + "loss": 48.9739, + "step": 550 + }, + { + "epoch": 13.119402985074627, + "grad_norm": 18.88302993774414, + "learning_rate": 3.5714285714285718e-06, + "loss": 47.0551, + "step": 551 + }, + { + "epoch": 13.143283582089552, + "grad_norm": 17.176782608032227, + "learning_rate": 3.55952380952381e-06, + "loss": 48.0771, + "step": 552 + }, + { + "epoch": 13.167164179104478, + "grad_norm": 19.72154426574707, + "learning_rate": 3.547619047619048e-06, + "loss": 49.5084, + "step": 553 + }, + { + "epoch": 13.191044776119403, + "grad_norm": 24.780994415283203, + "learning_rate": 3.5357142857142863e-06, + "loss": 46.5557, + "step": 554 + }, + { + "epoch": 13.214925373134328, + "grad_norm": 20.380996704101562, + "learning_rate": 3.523809523809524e-06, + "loss": 48.841, + "step": 555 + }, + { + "epoch": 13.238805970149254, + "grad_norm": 26.90860939025879, + "learning_rate": 3.511904761904762e-06, + "loss": 47.6185, + "step": 556 + }, + { + "epoch": 13.26268656716418, + "grad_norm": 21.40388298034668, + "learning_rate": 3.5e-06, + "loss": 47.787, + "step": 557 + }, + { + "epoch": 13.286567164179104, + "grad_norm": 24.708845138549805, + "learning_rate": 3.4880952380952383e-06, + "loss": 47.1974, + "step": 558 + }, + { + "epoch": 13.31044776119403, + "grad_norm": 25.317148208618164, + "learning_rate": 3.476190476190476e-06, + "loss": 49.2282, + "step": 559 + }, + { + "epoch": 13.334328358208955, + "grad_norm": 22.903011322021484, + "learning_rate": 3.4642857142857145e-06, + "loss": 47.0762, + "step": 560 + }, + { + "epoch": 13.35820895522388, + "grad_norm": 23.626604080200195, + "learning_rate": 3.4523809523809528e-06, + "loss": 47.3622, + "step": 561 + }, + { + "epoch": 13.382089552238806, + "grad_norm": 16.69061279296875, + "learning_rate": 3.4404761904761907e-06, + "loss": 48.5621, + "step": 562 + }, + { + "epoch": 13.405970149253731, + "grad_norm": 20.52508544921875, + "learning_rate": 3.428571428571429e-06, + "loss": 47.6565, + "step": 563 + }, + { + "epoch": 13.429850746268656, + "grad_norm": 25.125743865966797, + "learning_rate": 3.416666666666667e-06, + "loss": 48.1353, + "step": 564 + }, + { + "epoch": 13.453731343283582, + "grad_norm": 20.697166442871094, + "learning_rate": 3.404761904761905e-06, + "loss": 47.9368, + "step": 565 + }, + { + "epoch": 13.477611940298507, + "grad_norm": 22.396892547607422, + "learning_rate": 3.3928571428571435e-06, + "loss": 48.2956, + "step": 566 + }, + { + "epoch": 13.501492537313434, + "grad_norm": 24.770437240600586, + "learning_rate": 3.3809523809523814e-06, + "loss": 48.4467, + "step": 567 + }, + { + "epoch": 13.525373134328358, + "grad_norm": 19.44706153869629, + "learning_rate": 3.3690476190476197e-06, + "loss": 48.3155, + "step": 568 + }, + { + "epoch": 13.549253731343283, + "grad_norm": 27.680660247802734, + "learning_rate": 3.357142857142857e-06, + "loss": 47.9039, + "step": 569 + }, + { + "epoch": 13.57313432835821, + "grad_norm": 21.89419174194336, + "learning_rate": 3.3452380952380954e-06, + "loss": 47.9416, + "step": 570 + }, + { + "epoch": 13.597014925373134, + "grad_norm": 19.10918426513672, + "learning_rate": 3.3333333333333333e-06, + "loss": 48.45, + "step": 571 + }, + { + "epoch": 13.620895522388059, + "grad_norm": 29.83106231689453, + "learning_rate": 3.3214285714285716e-06, + "loss": 48.9583, + "step": 572 + }, + { + "epoch": 13.644776119402986, + "grad_norm": 28.05882453918457, + "learning_rate": 3.3095238095238095e-06, + "loss": 49.108, + "step": 573 + }, + { + "epoch": 13.66865671641791, + "grad_norm": 17.379384994506836, + "learning_rate": 3.297619047619048e-06, + "loss": 48.4707, + "step": 574 + }, + { + "epoch": 13.692537313432837, + "grad_norm": 19.15117645263672, + "learning_rate": 3.285714285714286e-06, + "loss": 48.476, + "step": 575 + }, + { + "epoch": 13.716417910447761, + "grad_norm": 23.892152786254883, + "learning_rate": 3.273809523809524e-06, + "loss": 48.0321, + "step": 576 + }, + { + "epoch": 13.740298507462686, + "grad_norm": 18.658008575439453, + "learning_rate": 3.2619047619047623e-06, + "loss": 47.2192, + "step": 577 + }, + { + "epoch": 13.764179104477613, + "grad_norm": 16.940099716186523, + "learning_rate": 3.2500000000000002e-06, + "loss": 49.2263, + "step": 578 + }, + { + "epoch": 13.788059701492537, + "grad_norm": 25.7972412109375, + "learning_rate": 3.2380952380952385e-06, + "loss": 47.5039, + "step": 579 + }, + { + "epoch": 13.811940298507462, + "grad_norm": 28.928129196166992, + "learning_rate": 3.226190476190477e-06, + "loss": 47.9264, + "step": 580 + }, + { + "epoch": 13.835820895522389, + "grad_norm": 23.67597007751465, + "learning_rate": 3.2142857142857147e-06, + "loss": 49.1464, + "step": 581 + }, + { + "epoch": 13.859701492537313, + "grad_norm": 18.345443725585938, + "learning_rate": 3.202380952380952e-06, + "loss": 47.888, + "step": 582 + }, + { + "epoch": 13.883582089552238, + "grad_norm": 19.80716896057129, + "learning_rate": 3.1904761904761905e-06, + "loss": 47.4324, + "step": 583 + }, + { + "epoch": 13.907462686567165, + "grad_norm": 20.488346099853516, + "learning_rate": 3.178571428571429e-06, + "loss": 48.3033, + "step": 584 + }, + { + "epoch": 13.93134328358209, + "grad_norm": 22.3657283782959, + "learning_rate": 3.1666666666666667e-06, + "loss": 48.1474, + "step": 585 + }, + { + "epoch": 13.955223880597014, + "grad_norm": 17.457408905029297, + "learning_rate": 3.154761904761905e-06, + "loss": 47.2418, + "step": 586 + }, + { + "epoch": 13.97910447761194, + "grad_norm": NaN, + "learning_rate": 3.142857142857143e-06, + "loss": 54.1812, + "step": 587 + }, + { + "epoch": 14.0, + "grad_norm": 17.137672424316406, + "learning_rate": 3.142857142857143e-06, + "loss": 42.3703, + "step": 588 + }, + { + "epoch": 14.023880597014925, + "grad_norm": 20.55642318725586, + "learning_rate": 3.130952380952381e-06, + "loss": 49.4628, + "step": 589 + }, + { + "epoch": 14.047761194029851, + "grad_norm": 19.925596237182617, + "learning_rate": 3.1190476190476195e-06, + "loss": 47.5266, + "step": 590 + }, + { + "epoch": 14.071641791044776, + "grad_norm": 12.49276065826416, + "learning_rate": 3.1071428571428574e-06, + "loss": 47.8654, + "step": 591 + }, + { + "epoch": 14.0955223880597, + "grad_norm": 17.266550064086914, + "learning_rate": 3.0952380952380957e-06, + "loss": 48.4362, + "step": 592 + }, + { + "epoch": 14.119402985074627, + "grad_norm": 18.234397888183594, + "learning_rate": 3.0833333333333336e-06, + "loss": 48.9532, + "step": 593 + }, + { + "epoch": 14.143283582089552, + "grad_norm": 19.880165100097656, + "learning_rate": 3.071428571428572e-06, + "loss": 48.0088, + "step": 594 + }, + { + "epoch": 14.167164179104478, + "grad_norm": 23.04216766357422, + "learning_rate": 3.05952380952381e-06, + "loss": 48.0934, + "step": 595 + }, + { + "epoch": 14.191044776119403, + "grad_norm": 19.199676513671875, + "learning_rate": 3.047619047619048e-06, + "loss": 48.3845, + "step": 596 + }, + { + "epoch": 14.214925373134328, + "grad_norm": 20.758337020874023, + "learning_rate": 3.0357142857142856e-06, + "loss": 47.4652, + "step": 597 + }, + { + "epoch": 14.238805970149254, + "grad_norm": 17.532787322998047, + "learning_rate": 3.023809523809524e-06, + "loss": 48.0212, + "step": 598 + }, + { + "epoch": 14.26268656716418, + "grad_norm": 16.547094345092773, + "learning_rate": 3.011904761904762e-06, + "loss": 48.6113, + "step": 599 + }, + { + "epoch": 14.286567164179104, + "grad_norm": 16.324464797973633, + "learning_rate": 3e-06, + "loss": 47.9735, + "step": 600 + }, + { + "epoch": 14.31044776119403, + "grad_norm": 16.54167938232422, + "learning_rate": 2.9880952380952384e-06, + "loss": 47.4436, + "step": 601 + }, + { + "epoch": 14.334328358208955, + "grad_norm": 23.455759048461914, + "learning_rate": 2.9761904761904763e-06, + "loss": 47.6631, + "step": 602 + }, + { + "epoch": 14.35820895522388, + "grad_norm": 19.159008026123047, + "learning_rate": 2.9642857142857146e-06, + "loss": 48.0291, + "step": 603 + }, + { + "epoch": 14.382089552238806, + "grad_norm": 18.66881561279297, + "learning_rate": 2.9523809523809525e-06, + "loss": 46.4582, + "step": 604 + }, + { + "epoch": 14.405970149253731, + "grad_norm": 19.129064559936523, + "learning_rate": 2.9404761904761908e-06, + "loss": 49.4455, + "step": 605 + }, + { + "epoch": 14.429850746268656, + "grad_norm": NaN, + "learning_rate": 2.928571428571429e-06, + "loss": 78.6564, + "step": 606 + }, + { + "epoch": 14.453731343283582, + "grad_norm": 18.47364044189453, + "learning_rate": 2.928571428571429e-06, + "loss": 48.1748, + "step": 607 + }, + { + "epoch": 14.477611940298507, + "grad_norm": 17.920883178710938, + "learning_rate": 2.916666666666667e-06, + "loss": 47.6447, + "step": 608 + }, + { + "epoch": 14.501492537313434, + "grad_norm": 18.263038635253906, + "learning_rate": 2.9047619047619053e-06, + "loss": 48.7324, + "step": 609 + }, + { + "epoch": 14.525373134328358, + "grad_norm": 24.323266983032227, + "learning_rate": 2.892857142857143e-06, + "loss": 48.6135, + "step": 610 + }, + { + "epoch": 14.549253731343283, + "grad_norm": 21.56492042541504, + "learning_rate": 2.880952380952381e-06, + "loss": 47.0007, + "step": 611 + }, + { + "epoch": 14.57313432835821, + "grad_norm": 17.741748809814453, + "learning_rate": 2.869047619047619e-06, + "loss": 46.3136, + "step": 612 + }, + { + "epoch": 14.597014925373134, + "grad_norm": 17.218914031982422, + "learning_rate": 2.8571428571428573e-06, + "loss": 47.6417, + "step": 613 + }, + { + "epoch": 14.620895522388059, + "grad_norm": 22.856996536254883, + "learning_rate": 2.8452380952380956e-06, + "loss": 47.6898, + "step": 614 + }, + { + "epoch": 14.644776119402986, + "grad_norm": NaN, + "learning_rate": 2.8333333333333335e-06, + "loss": 53.529, + "step": 615 + }, + { + "epoch": 14.66865671641791, + "grad_norm": 23.29751968383789, + "learning_rate": 2.8333333333333335e-06, + "loss": 48.164, + "step": 616 + }, + { + "epoch": 14.692537313432837, + "grad_norm": 15.633321762084961, + "learning_rate": 2.8214285714285718e-06, + "loss": 46.9866, + "step": 617 + }, + { + "epoch": 14.716417910447761, + "grad_norm": 21.713376998901367, + "learning_rate": 2.8095238095238096e-06, + "loss": 48.2856, + "step": 618 + }, + { + "epoch": 14.740298507462686, + "grad_norm": 17.07369613647461, + "learning_rate": 2.797619047619048e-06, + "loss": 46.4404, + "step": 619 + }, + { + "epoch": 14.764179104477613, + "grad_norm": 14.855449676513672, + "learning_rate": 2.785714285714286e-06, + "loss": 48.2668, + "step": 620 + }, + { + "epoch": 14.788059701492537, + "grad_norm": 16.479616165161133, + "learning_rate": 2.773809523809524e-06, + "loss": 49.2661, + "step": 621 + }, + { + "epoch": 14.811940298507462, + "grad_norm": 14.471490859985352, + "learning_rate": 2.7619047619047625e-06, + "loss": 47.0484, + "step": 622 + }, + { + "epoch": 14.835820895522389, + "grad_norm": 19.018714904785156, + "learning_rate": 2.7500000000000004e-06, + "loss": 49.2253, + "step": 623 + }, + { + "epoch": 14.859701492537313, + "grad_norm": 16.21799087524414, + "learning_rate": 2.7380952380952387e-06, + "loss": 49.0738, + "step": 624 + }, + { + "epoch": 14.883582089552238, + "grad_norm": 20.86383628845215, + "learning_rate": 2.726190476190476e-06, + "loss": 48.4231, + "step": 625 + }, + { + "epoch": 14.907462686567165, + "grad_norm": 20.60930633544922, + "learning_rate": 2.7142857142857144e-06, + "loss": 47.7464, + "step": 626 + }, + { + "epoch": 14.93134328358209, + "grad_norm": 20.909135818481445, + "learning_rate": 2.7023809523809523e-06, + "loss": 48.519, + "step": 627 + }, + { + "epoch": 14.955223880597014, + "grad_norm": 18.555694580078125, + "learning_rate": 2.6904761904761906e-06, + "loss": 48.1625, + "step": 628 + }, + { + "epoch": 14.97910447761194, + "grad_norm": 18.154813766479492, + "learning_rate": 2.6785714285714285e-06, + "loss": 48.9444, + "step": 629 + }, + { + "epoch": 15.0, + "grad_norm": 19.308523178100586, + "learning_rate": 2.666666666666667e-06, + "loss": 42.1936, + "step": 630 + }, + { + "epoch": 15.023880597014925, + "grad_norm": 22.725357055664062, + "learning_rate": 2.654761904761905e-06, + "loss": 49.5597, + "step": 631 + }, + { + "epoch": 15.047761194029851, + "grad_norm": 18.862451553344727, + "learning_rate": 2.642857142857143e-06, + "loss": 46.914, + "step": 632 + }, + { + "epoch": 15.071641791044776, + "grad_norm": 19.017065048217773, + "learning_rate": 2.6309523809523813e-06, + "loss": 47.7233, + "step": 633 + }, + { + "epoch": 15.0955223880597, + "grad_norm": 19.03627586364746, + "learning_rate": 2.6190476190476192e-06, + "loss": 48.0406, + "step": 634 + }, + { + "epoch": 15.119402985074627, + "grad_norm": 18.53116798400879, + "learning_rate": 2.6071428571428575e-06, + "loss": 47.3259, + "step": 635 + }, + { + "epoch": 15.143283582089552, + "grad_norm": 19.265275955200195, + "learning_rate": 2.595238095238096e-06, + "loss": 47.2465, + "step": 636 + }, + { + "epoch": 15.167164179104478, + "grad_norm": 19.497289657592773, + "learning_rate": 2.5833333333333337e-06, + "loss": 48.5984, + "step": 637 + }, + { + "epoch": 15.191044776119403, + "grad_norm": 20.183780670166016, + "learning_rate": 2.571428571428571e-06, + "loss": 46.6221, + "step": 638 + }, + { + "epoch": 15.214925373134328, + "grad_norm": 22.911672592163086, + "learning_rate": 2.5595238095238095e-06, + "loss": 48.0178, + "step": 639 + }, + { + "epoch": 15.238805970149254, + "grad_norm": 20.678709030151367, + "learning_rate": 2.547619047619048e-06, + "loss": 47.0322, + "step": 640 + }, + { + "epoch": 15.26268656716418, + "grad_norm": 18.579042434692383, + "learning_rate": 2.5357142857142857e-06, + "loss": 48.1428, + "step": 641 + }, + { + "epoch": 15.286567164179104, + "grad_norm": 23.61576271057129, + "learning_rate": 2.523809523809524e-06, + "loss": 48.444, + "step": 642 + }, + { + "epoch": 15.31044776119403, + "grad_norm": 19.602746963500977, + "learning_rate": 2.511904761904762e-06, + "loss": 48.2582, + "step": 643 + }, + { + "epoch": 15.334328358208955, + "grad_norm": 12.509607315063477, + "learning_rate": 2.5e-06, + "loss": 48.8531, + "step": 644 + }, + { + "epoch": 15.35820895522388, + "grad_norm": 18.749767303466797, + "learning_rate": 2.4880952380952385e-06, + "loss": 47.6453, + "step": 645 + }, + { + "epoch": 15.382089552238806, + "grad_norm": 20.612041473388672, + "learning_rate": 2.4761904761904764e-06, + "loss": 48.7038, + "step": 646 + }, + { + "epoch": 15.405970149253731, + "grad_norm": 18.65719985961914, + "learning_rate": 2.4642857142857147e-06, + "loss": 47.7954, + "step": 647 + }, + { + "epoch": 15.429850746268656, + "grad_norm": 22.636686325073242, + "learning_rate": 2.4523809523809526e-06, + "loss": 48.1164, + "step": 648 + }, + { + "epoch": 15.453731343283582, + "grad_norm": 20.93446922302246, + "learning_rate": 2.4404761904761905e-06, + "loss": 48.5955, + "step": 649 + }, + { + "epoch": 15.477611940298507, + "grad_norm": 20.77125358581543, + "learning_rate": 2.428571428571429e-06, + "loss": 48.4369, + "step": 650 + }, + { + "epoch": 15.501492537313434, + "grad_norm": 17.003498077392578, + "learning_rate": 2.4166666666666667e-06, + "loss": 49.0355, + "step": 651 + }, + { + "epoch": 15.525373134328358, + "grad_norm": 20.743436813354492, + "learning_rate": 2.404761904761905e-06, + "loss": 47.8368, + "step": 652 + }, + { + "epoch": 15.549253731343283, + "grad_norm": NaN, + "learning_rate": 2.3928571428571433e-06, + "loss": 41.6371, + "step": 653 + }, + { + "epoch": 15.57313432835821, + "grad_norm": 21.716781616210938, + "learning_rate": 2.3928571428571433e-06, + "loss": 48.5806, + "step": 654 + }, + { + "epoch": 15.597014925373134, + "grad_norm": 18.8812198638916, + "learning_rate": 2.380952380952381e-06, + "loss": 49.0707, + "step": 655 + }, + { + "epoch": 15.620895522388059, + "grad_norm": 22.305049896240234, + "learning_rate": 2.369047619047619e-06, + "loss": 47.7556, + "step": 656 + }, + { + "epoch": 15.644776119402986, + "grad_norm": 20.51401710510254, + "learning_rate": 2.3571428571428574e-06, + "loss": 48.1588, + "step": 657 + }, + { + "epoch": 15.66865671641791, + "grad_norm": 17.691770553588867, + "learning_rate": 2.3452380952380953e-06, + "loss": 47.5187, + "step": 658 + }, + { + "epoch": 15.692537313432837, + "grad_norm": 22.343585968017578, + "learning_rate": 2.3333333333333336e-06, + "loss": 47.5725, + "step": 659 + }, + { + "epoch": 15.716417910447761, + "grad_norm": 21.656587600708008, + "learning_rate": 2.321428571428572e-06, + "loss": 47.6903, + "step": 660 + }, + { + "epoch": 15.740298507462686, + "grad_norm": 20.632055282592773, + "learning_rate": 2.3095238095238098e-06, + "loss": 47.4526, + "step": 661 + }, + { + "epoch": 15.764179104477613, + "grad_norm": 22.324811935424805, + "learning_rate": 2.2976190476190477e-06, + "loss": 47.3316, + "step": 662 + }, + { + "epoch": 15.788059701492537, + "grad_norm": 19.320737838745117, + "learning_rate": 2.285714285714286e-06, + "loss": 48.2315, + "step": 663 + }, + { + "epoch": 15.811940298507462, + "grad_norm": 18.58050537109375, + "learning_rate": 2.273809523809524e-06, + "loss": 47.9147, + "step": 664 + }, + { + "epoch": 15.835820895522389, + "grad_norm": 20.37384796142578, + "learning_rate": 2.261904761904762e-06, + "loss": 47.5874, + "step": 665 + }, + { + "epoch": 15.859701492537313, + "grad_norm": 20.893856048583984, + "learning_rate": 2.25e-06, + "loss": 49.225, + "step": 666 + }, + { + "epoch": 15.883582089552238, + "grad_norm": 18.4589786529541, + "learning_rate": 2.2380952380952384e-06, + "loss": 47.3042, + "step": 667 + }, + { + "epoch": 15.907462686567165, + "grad_norm": 20.845996856689453, + "learning_rate": 2.2261904761904763e-06, + "loss": 47.3255, + "step": 668 + }, + { + "epoch": 15.93134328358209, + "grad_norm": 20.149137496948242, + "learning_rate": 2.2142857142857146e-06, + "loss": 48.6543, + "step": 669 + }, + { + "epoch": 15.955223880597014, + "grad_norm": 14.768882751464844, + "learning_rate": 2.2023809523809525e-06, + "loss": 46.8274, + "step": 670 + }, + { + "epoch": 15.97910447761194, + "grad_norm": 26.926074981689453, + "learning_rate": 2.1904761904761908e-06, + "loss": 48.2035, + "step": 671 + }, + { + "epoch": 16.0, + "grad_norm": 22.840618133544922, + "learning_rate": 2.1785714285714286e-06, + "loss": 42.8242, + "step": 672 + }, + { + "epoch": 16.023880597014927, + "grad_norm": 16.183008193969727, + "learning_rate": 2.166666666666667e-06, + "loss": 47.8309, + "step": 673 + }, + { + "epoch": 16.04776119402985, + "grad_norm": 20.603744506835938, + "learning_rate": 2.154761904761905e-06, + "loss": 48.5197, + "step": 674 + }, + { + "epoch": 16.071641791044776, + "grad_norm": 26.492107391357422, + "learning_rate": 2.1428571428571427e-06, + "loss": 47.2312, + "step": 675 + }, + { + "epoch": 16.095522388059702, + "grad_norm": 19.786901473999023, + "learning_rate": 2.130952380952381e-06, + "loss": 49.6201, + "step": 676 + }, + { + "epoch": 16.119402985074625, + "grad_norm": 18.150909423828125, + "learning_rate": 2.1190476190476194e-06, + "loss": 48.7407, + "step": 677 + }, + { + "epoch": 16.143283582089552, + "grad_norm": 18.797983169555664, + "learning_rate": 2.1071428571428572e-06, + "loss": 47.0801, + "step": 678 + }, + { + "epoch": 16.16716417910448, + "grad_norm": 16.088953018188477, + "learning_rate": 2.0952380952380955e-06, + "loss": 47.6509, + "step": 679 + }, + { + "epoch": 16.1910447761194, + "grad_norm": 20.359085083007812, + "learning_rate": 2.0833333333333334e-06, + "loss": 48.9226, + "step": 680 + }, + { + "epoch": 16.214925373134328, + "grad_norm": 21.99265480041504, + "learning_rate": 2.0714285714285717e-06, + "loss": 47.3775, + "step": 681 + }, + { + "epoch": 16.238805970149254, + "grad_norm": 18.616743087768555, + "learning_rate": 2.0595238095238096e-06, + "loss": 45.8448, + "step": 682 + }, + { + "epoch": 16.262686567164177, + "grad_norm": 19.6337947845459, + "learning_rate": 2.047619047619048e-06, + "loss": 48.2077, + "step": 683 + }, + { + "epoch": 16.286567164179104, + "grad_norm": 23.881439208984375, + "learning_rate": 2.035714285714286e-06, + "loss": 48.6796, + "step": 684 + }, + { + "epoch": 16.31044776119403, + "grad_norm": 19.665023803710938, + "learning_rate": 2.023809523809524e-06, + "loss": 48.7275, + "step": 685 + }, + { + "epoch": 16.334328358208957, + "grad_norm": 18.438793182373047, + "learning_rate": 2.011904761904762e-06, + "loss": 49.585, + "step": 686 + }, + { + "epoch": 16.35820895522388, + "grad_norm": 17.073816299438477, + "learning_rate": 2.0000000000000003e-06, + "loss": 47.4548, + "step": 687 + }, + { + "epoch": 16.382089552238806, + "grad_norm": 20.504276275634766, + "learning_rate": 1.9880952380952382e-06, + "loss": 47.3555, + "step": 688 + }, + { + "epoch": 16.405970149253733, + "grad_norm": 21.564546585083008, + "learning_rate": 1.976190476190476e-06, + "loss": 47.6304, + "step": 689 + }, + { + "epoch": 16.429850746268656, + "grad_norm": 16.773197174072266, + "learning_rate": 1.9642857142857144e-06, + "loss": 49.078, + "step": 690 + }, + { + "epoch": 16.453731343283582, + "grad_norm": 22.77934455871582, + "learning_rate": 1.9523809523809527e-06, + "loss": 47.8289, + "step": 691 + }, + { + "epoch": 16.47761194029851, + "grad_norm": 17.375993728637695, + "learning_rate": 1.9404761904761906e-06, + "loss": 48.4812, + "step": 692 + }, + { + "epoch": 16.501492537313432, + "grad_norm": 21.407329559326172, + "learning_rate": 1.928571428571429e-06, + "loss": 48.2934, + "step": 693 + }, + { + "epoch": 16.52537313432836, + "grad_norm": 15.673316955566406, + "learning_rate": 1.916666666666667e-06, + "loss": 46.7304, + "step": 694 + }, + { + "epoch": 16.549253731343285, + "grad_norm": 24.577089309692383, + "learning_rate": 1.904761904761905e-06, + "loss": 47.9352, + "step": 695 + }, + { + "epoch": 16.573134328358208, + "grad_norm": 24.46076774597168, + "learning_rate": 1.892857142857143e-06, + "loss": 48.8173, + "step": 696 + }, + { + "epoch": 16.597014925373134, + "grad_norm": 14.248388290405273, + "learning_rate": 1.880952380952381e-06, + "loss": 48.5858, + "step": 697 + }, + { + "epoch": 16.62089552238806, + "grad_norm": 16.925329208374023, + "learning_rate": 1.8690476190476192e-06, + "loss": 47.8278, + "step": 698 + }, + { + "epoch": 16.644776119402984, + "grad_norm": 25.52614402770996, + "learning_rate": 1.8571428571428573e-06, + "loss": 48.1248, + "step": 699 + }, + { + "epoch": 16.66865671641791, + "grad_norm": 21.011341094970703, + "learning_rate": 1.8452380952380954e-06, + "loss": 47.8154, + "step": 700 + }, + { + "epoch": 16.692537313432837, + "grad_norm": 14.694896697998047, + "learning_rate": 1.8333333333333333e-06, + "loss": 47.9668, + "step": 701 + }, + { + "epoch": 16.71641791044776, + "grad_norm": 22.32903480529785, + "learning_rate": 1.8214285714285716e-06, + "loss": 48.6784, + "step": 702 + }, + { + "epoch": 16.740298507462686, + "grad_norm": 17.19482421875, + "learning_rate": 1.8095238095238097e-06, + "loss": 46.9973, + "step": 703 + }, + { + "epoch": 16.764179104477613, + "grad_norm": 14.590733528137207, + "learning_rate": 1.7976190476190478e-06, + "loss": 47.2393, + "step": 704 + }, + { + "epoch": 16.788059701492536, + "grad_norm": 17.131982803344727, + "learning_rate": 1.7857142857142859e-06, + "loss": 47.9412, + "step": 705 + }, + { + "epoch": 16.811940298507462, + "grad_norm": 18.513992309570312, + "learning_rate": 1.773809523809524e-06, + "loss": 48.8777, + "step": 706 + }, + { + "epoch": 16.83582089552239, + "grad_norm": 17.625539779663086, + "learning_rate": 1.761904761904762e-06, + "loss": 48.3885, + "step": 707 + }, + { + "epoch": 16.85970149253731, + "grad_norm": 16.540056228637695, + "learning_rate": 1.75e-06, + "loss": 47.8561, + "step": 708 + }, + { + "epoch": 16.883582089552238, + "grad_norm": 20.070533752441406, + "learning_rate": 1.738095238095238e-06, + "loss": 46.6418, + "step": 709 + }, + { + "epoch": 16.907462686567165, + "grad_norm": 18.742460250854492, + "learning_rate": 1.7261904761904764e-06, + "loss": 46.7471, + "step": 710 + }, + { + "epoch": 16.93134328358209, + "grad_norm": 17.491954803466797, + "learning_rate": 1.7142857142857145e-06, + "loss": 47.5558, + "step": 711 + }, + { + "epoch": 16.955223880597014, + "grad_norm": 17.457130432128906, + "learning_rate": 1.7023809523809526e-06, + "loss": 47.4441, + "step": 712 + }, + { + "epoch": 16.97910447761194, + "grad_norm": 21.053844451904297, + "learning_rate": 1.6904761904761907e-06, + "loss": 48.1931, + "step": 713 + }, + { + "epoch": 17.0, + "grad_norm": 16.943801879882812, + "learning_rate": 1.6785714285714286e-06, + "loss": 41.9934, + "step": 714 + }, + { + "epoch": 17.023880597014927, + "grad_norm": 21.56785011291504, + "learning_rate": 1.6666666666666667e-06, + "loss": 47.1652, + "step": 715 + }, + { + "epoch": 17.04776119402985, + "grad_norm": 21.193382263183594, + "learning_rate": 1.6547619047619048e-06, + "loss": 47.6751, + "step": 716 + }, + { + "epoch": 17.071641791044776, + "grad_norm": 16.245115280151367, + "learning_rate": 1.642857142857143e-06, + "loss": 47.4133, + "step": 717 + }, + { + "epoch": 17.095522388059702, + "grad_norm": 18.834646224975586, + "learning_rate": 1.6309523809523812e-06, + "loss": 48.145, + "step": 718 + }, + { + "epoch": 17.119402985074625, + "grad_norm": 15.769698143005371, + "learning_rate": 1.6190476190476193e-06, + "loss": 48.1181, + "step": 719 + }, + { + "epoch": 17.143283582089552, + "grad_norm": 13.460511207580566, + "learning_rate": 1.6071428571428574e-06, + "loss": 49.1229, + "step": 720 + }, + { + "epoch": 17.16716417910448, + "grad_norm": 18.58087158203125, + "learning_rate": 1.5952380952380953e-06, + "loss": 47.5095, + "step": 721 + }, + { + "epoch": 17.1910447761194, + "grad_norm": 18.607332229614258, + "learning_rate": 1.5833333333333333e-06, + "loss": 49.1334, + "step": 722 + }, + { + "epoch": 17.214925373134328, + "grad_norm": 15.046488761901855, + "learning_rate": 1.5714285714285714e-06, + "loss": 47.6151, + "step": 723 + }, + { + "epoch": 17.238805970149254, + "grad_norm": 17.442358016967773, + "learning_rate": 1.5595238095238098e-06, + "loss": 47.3771, + "step": 724 + }, + { + "epoch": 17.262686567164177, + "grad_norm": 11.690101623535156, + "learning_rate": 1.5476190476190479e-06, + "loss": 48.1095, + "step": 725 + }, + { + "epoch": 17.286567164179104, + "grad_norm": 17.945192337036133, + "learning_rate": 1.535714285714286e-06, + "loss": 47.8941, + "step": 726 + }, + { + "epoch": 17.31044776119403, + "grad_norm": 13.878116607666016, + "learning_rate": 1.523809523809524e-06, + "loss": 47.6422, + "step": 727 + }, + { + "epoch": 17.334328358208957, + "grad_norm": 15.942928314208984, + "learning_rate": 1.511904761904762e-06, + "loss": 46.9964, + "step": 728 + }, + { + "epoch": 17.35820895522388, + "grad_norm": 13.57482624053955, + "learning_rate": 1.5e-06, + "loss": 47.1832, + "step": 729 + }, + { + "epoch": 17.382089552238806, + "grad_norm": 13.781617164611816, + "learning_rate": 1.4880952380952381e-06, + "loss": 48.9621, + "step": 730 + }, + { + "epoch": 17.405970149253733, + "grad_norm": 14.26857852935791, + "learning_rate": 1.4761904761904762e-06, + "loss": 48.6631, + "step": 731 + }, + { + "epoch": 17.429850746268656, + "grad_norm": 16.23444938659668, + "learning_rate": 1.4642857142857145e-06, + "loss": 46.84, + "step": 732 + }, + { + "epoch": 17.453731343283582, + "grad_norm": 17.442630767822266, + "learning_rate": 1.4523809523809526e-06, + "loss": 48.2996, + "step": 733 + }, + { + "epoch": 17.47761194029851, + "grad_norm": 14.329082489013672, + "learning_rate": 1.4404761904761905e-06, + "loss": 47.396, + "step": 734 + }, + { + "epoch": 17.501492537313432, + "grad_norm": 14.772257804870605, + "learning_rate": 1.4285714285714286e-06, + "loss": 48.5733, + "step": 735 + }, + { + "epoch": 17.52537313432836, + "grad_norm": 14.331324577331543, + "learning_rate": 1.4166666666666667e-06, + "loss": 48.2969, + "step": 736 + }, + { + "epoch": 17.549253731343285, + "grad_norm": 17.498600006103516, + "learning_rate": 1.4047619047619048e-06, + "loss": 48.0221, + "step": 737 + }, + { + "epoch": 17.573134328358208, + "grad_norm": 16.155025482177734, + "learning_rate": 1.392857142857143e-06, + "loss": 47.9848, + "step": 738 + }, + { + "epoch": 17.597014925373134, + "grad_norm": 15.552813529968262, + "learning_rate": 1.3809523809523812e-06, + "loss": 48.4413, + "step": 739 + }, + { + "epoch": 17.62089552238806, + "grad_norm": 15.887310981750488, + "learning_rate": 1.3690476190476193e-06, + "loss": 47.6463, + "step": 740 + }, + { + "epoch": 17.644776119402984, + "grad_norm": 17.783411026000977, + "learning_rate": 1.3571428571428572e-06, + "loss": 47.8009, + "step": 741 + }, + { + "epoch": 17.66865671641791, + "grad_norm": 17.108932495117188, + "learning_rate": 1.3452380952380953e-06, + "loss": 47.9888, + "step": 742 + }, + { + "epoch": 17.692537313432837, + "grad_norm": 19.79203224182129, + "learning_rate": 1.3333333333333334e-06, + "loss": 48.5732, + "step": 743 + }, + { + "epoch": 17.71641791044776, + "grad_norm": 17.06324005126953, + "learning_rate": 1.3214285714285715e-06, + "loss": 48.4815, + "step": 744 + }, + { + "epoch": 17.740298507462686, + "grad_norm": 17.399097442626953, + "learning_rate": 1.3095238095238096e-06, + "loss": 47.5591, + "step": 745 + }, + { + "epoch": 17.764179104477613, + "grad_norm": 15.836935997009277, + "learning_rate": 1.297619047619048e-06, + "loss": 47.994, + "step": 746 + }, + { + "epoch": 17.788059701492536, + "grad_norm": 18.20856475830078, + "learning_rate": 1.2857142857142856e-06, + "loss": 47.9979, + "step": 747 + }, + { + "epoch": 17.811940298507462, + "grad_norm": 19.10239601135254, + "learning_rate": 1.273809523809524e-06, + "loss": 48.1196, + "step": 748 + }, + { + "epoch": 17.83582089552239, + "grad_norm": 17.21087646484375, + "learning_rate": 1.261904761904762e-06, + "loss": 47.8816, + "step": 749 + }, + { + "epoch": 17.85970149253731, + "grad_norm": 14.792268753051758, + "learning_rate": 1.25e-06, + "loss": 47.8182, + "step": 750 + }, + { + "epoch": 17.883582089552238, + "grad_norm": 13.695488929748535, + "learning_rate": 1.2380952380952382e-06, + "loss": 47.4298, + "step": 751 + }, + { + "epoch": 17.907462686567165, + "grad_norm": 15.197646141052246, + "learning_rate": 1.2261904761904763e-06, + "loss": 47.7132, + "step": 752 + }, + { + "epoch": 17.93134328358209, + "grad_norm": 19.13431739807129, + "learning_rate": 1.2142857142857144e-06, + "loss": 48.431, + "step": 753 + }, + { + "epoch": 17.955223880597014, + "grad_norm": 15.690411567687988, + "learning_rate": 1.2023809523809525e-06, + "loss": 47.4529, + "step": 754 + }, + { + "epoch": 17.97910447761194, + "grad_norm": 14.75414752960205, + "learning_rate": 1.1904761904761906e-06, + "loss": 47.9668, + "step": 755 + }, + { + "epoch": 18.0, + "grad_norm": 11.497115135192871, + "learning_rate": 1.1785714285714287e-06, + "loss": 41.8653, + "step": 756 + }, + { + "epoch": 18.023880597014927, + "grad_norm": 16.20159339904785, + "learning_rate": 1.1666666666666668e-06, + "loss": 47.2871, + "step": 757 + }, + { + "epoch": 18.04776119402985, + "grad_norm": 15.400497436523438, + "learning_rate": 1.1547619047619049e-06, + "loss": 46.3673, + "step": 758 + }, + { + "epoch": 18.071641791044776, + "grad_norm": 12.16773509979248, + "learning_rate": 1.142857142857143e-06, + "loss": 47.7463, + "step": 759 + }, + { + "epoch": 18.095522388059702, + "grad_norm": 19.978351593017578, + "learning_rate": 1.130952380952381e-06, + "loss": 47.4632, + "step": 760 + }, + { + "epoch": 18.119402985074625, + "grad_norm": 14.090561866760254, + "learning_rate": 1.1190476190476192e-06, + "loss": 48.9356, + "step": 761 + }, + { + "epoch": 18.143283582089552, + "grad_norm": 13.143173217773438, + "learning_rate": 1.1071428571428573e-06, + "loss": 48.1129, + "step": 762 + }, + { + "epoch": 18.16716417910448, + "grad_norm": 15.609000205993652, + "learning_rate": 1.0952380952380954e-06, + "loss": 48.8554, + "step": 763 + }, + { + "epoch": 18.1910447761194, + "grad_norm": 14.012611389160156, + "learning_rate": 1.0833333333333335e-06, + "loss": 47.6785, + "step": 764 + }, + { + "epoch": 18.214925373134328, + "grad_norm": 13.417494773864746, + "learning_rate": 1.0714285714285714e-06, + "loss": 48.2733, + "step": 765 + }, + { + "epoch": 18.238805970149254, + "grad_norm": 15.562864303588867, + "learning_rate": 1.0595238095238097e-06, + "loss": 48.0488, + "step": 766 + }, + { + "epoch": 18.262686567164177, + "grad_norm": 17.083723068237305, + "learning_rate": 1.0476190476190478e-06, + "loss": 49.3136, + "step": 767 + }, + { + "epoch": 18.286567164179104, + "grad_norm": 16.564395904541016, + "learning_rate": 1.0357142857142859e-06, + "loss": 48.1835, + "step": 768 + }, + { + "epoch": 18.31044776119403, + "grad_norm": 14.694929122924805, + "learning_rate": 1.023809523809524e-06, + "loss": 48.8632, + "step": 769 + }, + { + "epoch": 18.334328358208957, + "grad_norm": 13.4928560256958, + "learning_rate": 1.011904761904762e-06, + "loss": 47.3333, + "step": 770 + }, + { + "epoch": 18.35820895522388, + "grad_norm": 12.980204582214355, + "learning_rate": 1.0000000000000002e-06, + "loss": 47.7961, + "step": 771 + }, + { + "epoch": 18.382089552238806, + "grad_norm": 14.19666862487793, + "learning_rate": 9.88095238095238e-07, + "loss": 47.7806, + "step": 772 + }, + { + "epoch": 18.405970149253733, + "grad_norm": 13.2017183303833, + "learning_rate": 9.761904761904764e-07, + "loss": 46.1119, + "step": 773 + }, + { + "epoch": 18.429850746268656, + "grad_norm": 15.064650535583496, + "learning_rate": 9.642857142857145e-07, + "loss": 47.8105, + "step": 774 + }, + { + "epoch": 18.453731343283582, + "grad_norm": 14.713834762573242, + "learning_rate": 9.523809523809525e-07, + "loss": 47.5723, + "step": 775 + }, + { + "epoch": 18.47761194029851, + "grad_norm": 13.394201278686523, + "learning_rate": 9.404761904761906e-07, + "loss": 48.1632, + "step": 776 + }, + { + "epoch": 18.501492537313432, + "grad_norm": 10.742532730102539, + "learning_rate": 9.285714285714287e-07, + "loss": 48.6118, + "step": 777 + }, + { + "epoch": 18.52537313432836, + "grad_norm": 12.402650833129883, + "learning_rate": 9.166666666666666e-07, + "loss": 48.6597, + "step": 778 + }, + { + "epoch": 18.549253731343285, + "grad_norm": 15.73616886138916, + "learning_rate": 9.047619047619048e-07, + "loss": 47.9931, + "step": 779 + }, + { + "epoch": 18.573134328358208, + "grad_norm": 14.188780784606934, + "learning_rate": 8.928571428571429e-07, + "loss": 47.0081, + "step": 780 + }, + { + "epoch": 18.597014925373134, + "grad_norm": 12.516701698303223, + "learning_rate": 8.80952380952381e-07, + "loss": 49.0287, + "step": 781 + }, + { + "epoch": 18.62089552238806, + "grad_norm": 15.069429397583008, + "learning_rate": 8.69047619047619e-07, + "loss": 47.8721, + "step": 782 + }, + { + "epoch": 18.644776119402984, + "grad_norm": 13.091047286987305, + "learning_rate": 8.571428571428572e-07, + "loss": 48.1678, + "step": 783 + }, + { + "epoch": 18.66865671641791, + "grad_norm": 15.017065048217773, + "learning_rate": 8.452380952380953e-07, + "loss": 47.1277, + "step": 784 + }, + { + "epoch": 18.692537313432837, + "grad_norm": 12.091531753540039, + "learning_rate": 8.333333333333333e-07, + "loss": 47.4962, + "step": 785 + }, + { + "epoch": 18.71641791044776, + "grad_norm": 15.20182991027832, + "learning_rate": 8.214285714285715e-07, + "loss": 48.6224, + "step": 786 + }, + { + "epoch": 18.740298507462686, + "grad_norm": 11.17827320098877, + "learning_rate": 8.095238095238096e-07, + "loss": 47.9759, + "step": 787 + }, + { + "epoch": 18.764179104477613, + "grad_norm": 14.884525299072266, + "learning_rate": 7.976190476190476e-07, + "loss": 47.9749, + "step": 788 + }, + { + "epoch": 18.788059701492536, + "grad_norm": 14.360984802246094, + "learning_rate": 7.857142857142857e-07, + "loss": 48.9952, + "step": 789 + }, + { + "epoch": 18.811940298507462, + "grad_norm": 11.265621185302734, + "learning_rate": 7.738095238095239e-07, + "loss": 47.4274, + "step": 790 + }, + { + "epoch": 18.83582089552239, + "grad_norm": 18.072290420532227, + "learning_rate": 7.61904761904762e-07, + "loss": 47.8815, + "step": 791 + }, + { + "epoch": 18.85970149253731, + "grad_norm": 15.310029029846191, + "learning_rate": 7.5e-07, + "loss": 47.3907, + "step": 792 + }, + { + "epoch": 18.883582089552238, + "grad_norm": 14.032752990722656, + "learning_rate": 7.380952380952381e-07, + "loss": 48.0883, + "step": 793 + }, + { + "epoch": 18.907462686567165, + "grad_norm": 12.853668212890625, + "learning_rate": 7.261904761904763e-07, + "loss": 47.1452, + "step": 794 + }, + { + "epoch": 18.93134328358209, + "grad_norm": 15.6067476272583, + "learning_rate": 7.142857142857143e-07, + "loss": 46.6303, + "step": 795 + }, + { + "epoch": 18.955223880597014, + "grad_norm": 12.828201293945312, + "learning_rate": 7.023809523809524e-07, + "loss": 47.9885, + "step": 796 + }, + { + "epoch": 18.97910447761194, + "grad_norm": 13.336589813232422, + "learning_rate": 6.904761904761906e-07, + "loss": 48.2315, + "step": 797 + }, + { + "epoch": 19.0, + "grad_norm": 13.629434585571289, + "learning_rate": 6.785714285714286e-07, + "loss": 41.9374, + "step": 798 + }, + { + "epoch": 19.023880597014927, + "grad_norm": 13.237930297851562, + "learning_rate": 6.666666666666667e-07, + "loss": 46.6802, + "step": 799 + }, + { + "epoch": 19.04776119402985, + "grad_norm": 13.715863227844238, + "learning_rate": 6.547619047619048e-07, + "loss": 49.0494, + "step": 800 + }, + { + "epoch": 19.071641791044776, + "grad_norm": 13.439970016479492, + "learning_rate": 6.428571428571428e-07, + "loss": 46.3647, + "step": 801 + }, + { + "epoch": 19.095522388059702, + "grad_norm": 15.468942642211914, + "learning_rate": 6.30952380952381e-07, + "loss": 48.4725, + "step": 802 + }, + { + "epoch": 19.119402985074625, + "grad_norm": 14.160257339477539, + "learning_rate": 6.190476190476191e-07, + "loss": 47.4033, + "step": 803 + }, + { + "epoch": 19.143283582089552, + "grad_norm": 13.667155265808105, + "learning_rate": 6.071428571428572e-07, + "loss": 48.4729, + "step": 804 + }, + { + "epoch": 19.16716417910448, + "grad_norm": 12.428313255310059, + "learning_rate": 5.952380952380953e-07, + "loss": 48.8939, + "step": 805 + }, + { + "epoch": 19.1910447761194, + "grad_norm": 12.985882759094238, + "learning_rate": 5.833333333333334e-07, + "loss": 47.0663, + "step": 806 + }, + { + "epoch": 19.214925373134328, + "grad_norm": 12.827404975891113, + "learning_rate": 5.714285714285715e-07, + "loss": 47.5614, + "step": 807 + }, + { + "epoch": 19.238805970149254, + "grad_norm": 11.078653335571289, + "learning_rate": 5.595238095238096e-07, + "loss": 48.564, + "step": 808 + }, + { + "epoch": 19.262686567164177, + "grad_norm": 13.346016883850098, + "learning_rate": 5.476190476190477e-07, + "loss": 48.0823, + "step": 809 + }, + { + "epoch": 19.286567164179104, + "grad_norm": 14.523963928222656, + "learning_rate": 5.357142857142857e-07, + "loss": 48.4225, + "step": 810 + }, + { + "epoch": 19.31044776119403, + "grad_norm": 12.598445892333984, + "learning_rate": 5.238095238095239e-07, + "loss": 47.2514, + "step": 811 + }, + { + "epoch": 19.334328358208957, + "grad_norm": 12.203497886657715, + "learning_rate": 5.11904761904762e-07, + "loss": 47.217, + "step": 812 + }, + { + "epoch": 19.35820895522388, + "grad_norm": 12.144754409790039, + "learning_rate": 5.000000000000001e-07, + "loss": 47.191, + "step": 813 + }, + { + "epoch": 19.382089552238806, + "grad_norm": 12.585047721862793, + "learning_rate": 4.880952380952382e-07, + "loss": 48.4947, + "step": 814 + }, + { + "epoch": 19.405970149253733, + "grad_norm": 11.295561790466309, + "learning_rate": 4.7619047619047623e-07, + "loss": 46.9444, + "step": 815 + }, + { + "epoch": 19.429850746268656, + "grad_norm": 13.055256843566895, + "learning_rate": 4.642857142857143e-07, + "loss": 48.4469, + "step": 816 + }, + { + "epoch": 19.453731343283582, + "grad_norm": 12.051807403564453, + "learning_rate": 4.523809523809524e-07, + "loss": 48.0547, + "step": 817 + }, + { + "epoch": 19.47761194029851, + "grad_norm": 13.44185733795166, + "learning_rate": 4.404761904761905e-07, + "loss": 48.3155, + "step": 818 + }, + { + "epoch": 19.501492537313432, + "grad_norm": 12.405723571777344, + "learning_rate": 4.285714285714286e-07, + "loss": 48.3982, + "step": 819 + }, + { + "epoch": 19.52537313432836, + "grad_norm": 14.900402069091797, + "learning_rate": 4.1666666666666667e-07, + "loss": 48.2653, + "step": 820 + }, + { + "epoch": 19.549253731343285, + "grad_norm": 10.70801067352295, + "learning_rate": 4.047619047619048e-07, + "loss": 48.0384, + "step": 821 + }, + { + "epoch": 19.573134328358208, + "grad_norm": 12.318074226379395, + "learning_rate": 3.9285714285714286e-07, + "loss": 47.554, + "step": 822 + }, + { + "epoch": 19.597014925373134, + "grad_norm": 12.898431777954102, + "learning_rate": 3.80952380952381e-07, + "loss": 48.3586, + "step": 823 + }, + { + "epoch": 19.62089552238806, + "grad_norm": 15.45779800415039, + "learning_rate": 3.6904761904761906e-07, + "loss": 48.4193, + "step": 824 + }, + { + "epoch": 19.644776119402984, + "grad_norm": 11.230570793151855, + "learning_rate": 3.5714285714285716e-07, + "loss": 48.5294, + "step": 825 + }, + { + "epoch": 19.66865671641791, + "grad_norm": 13.647272109985352, + "learning_rate": 3.452380952380953e-07, + "loss": 47.2569, + "step": 826 + }, + { + "epoch": 19.692537313432837, + "grad_norm": 11.521178245544434, + "learning_rate": 3.3333333333333335e-07, + "loss": 47.2899, + "step": 827 + }, + { + "epoch": 19.71641791044776, + "grad_norm": 11.537907600402832, + "learning_rate": 3.214285714285714e-07, + "loss": 46.6462, + "step": 828 + }, + { + "epoch": 19.740298507462686, + "grad_norm": 11.670267105102539, + "learning_rate": 3.0952380952380955e-07, + "loss": 47.9797, + "step": 829 + }, + { + "epoch": 19.764179104477613, + "grad_norm": 11.660557746887207, + "learning_rate": 2.9761904761904765e-07, + "loss": 47.9744, + "step": 830 + }, + { + "epoch": 19.788059701492536, + "grad_norm": 12.332269668579102, + "learning_rate": 2.8571428571428575e-07, + "loss": 48.6015, + "step": 831 + }, + { + "epoch": 19.811940298507462, + "grad_norm": 12.228848457336426, + "learning_rate": 2.7380952380952385e-07, + "loss": 47.3215, + "step": 832 + }, + { + "epoch": 19.83582089552239, + "grad_norm": 13.780754089355469, + "learning_rate": 2.6190476190476194e-07, + "loss": 48.853, + "step": 833 + }, + { + "epoch": 19.85970149253731, + "grad_norm": 11.639240264892578, + "learning_rate": 2.5000000000000004e-07, + "loss": 48.8199, + "step": 834 + }, + { + "epoch": 19.883582089552238, + "grad_norm": 10.796862602233887, + "learning_rate": 2.3809523809523811e-07, + "loss": 47.5373, + "step": 835 + }, + { + "epoch": 19.907462686567165, + "grad_norm": 13.573180198669434, + "learning_rate": 2.261904761904762e-07, + "loss": 47.8368, + "step": 836 + }, + { + "epoch": 19.93134328358209, + "grad_norm": 11.497776985168457, + "learning_rate": 2.142857142857143e-07, + "loss": 47.8226, + "step": 837 + }, + { + "epoch": 19.955223880597014, + "grad_norm": 10.777889251708984, + "learning_rate": 2.023809523809524e-07, + "loss": 47.6424, + "step": 838 + }, + { + "epoch": 19.97910447761194, + "grad_norm": 10.77852725982666, + "learning_rate": 1.904761904761905e-07, + "loss": 46.379, + "step": 839 + }, + { + "epoch": 20.0, + "grad_norm": 13.582564353942871, + "learning_rate": 1.7857142857142858e-07, + "loss": 42.5239, + "step": 840 + }, + { + "epoch": 20.0, + "step": 840, + "total_flos": 4.130470305428237e+16, + "train_loss": 49.47331008002872, + "train_runtime": 26137.3223, + "train_samples_per_second": 4.095, + "train_steps_per_second": 0.032 + }, + { + "epoch": 20.023880597014927, + "grad_norm": 21.18770408630371, + "learning_rate": 1e-05, + "loss": 48.1474, + "step": 841 + }, + { + "epoch": 20.04776119402985, + "grad_norm": Infinity, + "learning_rate": 9.99404761904762e-06, + "loss": 60.2758, + "step": 842 + }, + { + "epoch": 20.071641791044776, + "grad_norm": Infinity, + "learning_rate": 9.99404761904762e-06, + "loss": 61.7211, + "step": 843 + }, + { + "epoch": 20.095522388059702, + "grad_norm": 504.4407958984375, + "learning_rate": 9.99404761904762e-06, + "loss": 60.8189, + "step": 844 + }, + { + "epoch": 20.119402985074625, + "grad_norm": 221.96849060058594, + "learning_rate": 9.988095238095239e-06, + "loss": 54.4658, + "step": 845 + }, + { + "epoch": 20.143283582089552, + "grad_norm": 110.4036865234375, + "learning_rate": 9.982142857142858e-06, + "loss": 52.4242, + "step": 846 + }, + { + "epoch": 20.16716417910448, + "grad_norm": 82.75493621826172, + "learning_rate": 9.976190476190477e-06, + "loss": 50.3129, + "step": 847 + }, + { + "epoch": 20.1910447761194, + "grad_norm": 62.56040573120117, + "learning_rate": 9.970238095238096e-06, + "loss": 49.7171, + "step": 848 + }, + { + "epoch": 20.214925373134328, + "grad_norm": 70.04007720947266, + "learning_rate": 9.964285714285714e-06, + "loss": 48.185, + "step": 849 + }, + { + "epoch": 20.238805970149254, + "grad_norm": 56.70342254638672, + "learning_rate": 9.958333333333334e-06, + "loss": 49.5787, + "step": 850 + }, + { + "epoch": 20.262686567164177, + "grad_norm": 64.66405487060547, + "learning_rate": 9.952380952380954e-06, + "loss": 49.6106, + "step": 851 + }, + { + "epoch": 20.286567164179104, + "grad_norm": 43.37612533569336, + "learning_rate": 9.946428571428572e-06, + "loss": 49.2966, + "step": 852 + }, + { + "epoch": 20.31044776119403, + "grad_norm": 42.66206359863281, + "learning_rate": 9.940476190476192e-06, + "loss": 48.7073, + "step": 853 + }, + { + "epoch": 20.334328358208957, + "grad_norm": 37.17741775512695, + "learning_rate": 9.93452380952381e-06, + "loss": 48.7592, + "step": 854 + }, + { + "epoch": 20.35820895522388, + "grad_norm": 39.27332305908203, + "learning_rate": 9.92857142857143e-06, + "loss": 48.1181, + "step": 855 + }, + { + "epoch": 20.382089552238806, + "grad_norm": 31.37261390686035, + "learning_rate": 9.922619047619048e-06, + "loss": 47.4873, + "step": 856 + }, + { + "epoch": 20.405970149253733, + "grad_norm": 41.693809509277344, + "learning_rate": 9.916666666666668e-06, + "loss": 48.9428, + "step": 857 + }, + { + "epoch": 20.429850746268656, + "grad_norm": 29.33939552307129, + "learning_rate": 9.910714285714288e-06, + "loss": 49.2928, + "step": 858 + }, + { + "epoch": 20.453731343283582, + "grad_norm": 30.606157302856445, + "learning_rate": 9.904761904761906e-06, + "loss": 49.1506, + "step": 859 + }, + { + "epoch": 20.47761194029851, + "grad_norm": 27.273784637451172, + "learning_rate": 9.898809523809525e-06, + "loss": 46.6136, + "step": 860 + }, + { + "epoch": 20.501492537313432, + "grad_norm": 24.410682678222656, + "learning_rate": 9.892857142857143e-06, + "loss": 48.3989, + "step": 861 + }, + { + "epoch": 20.52537313432836, + "grad_norm": 24.138607025146484, + "learning_rate": 9.886904761904763e-06, + "loss": 49.3858, + "step": 862 + }, + { + "epoch": 20.549253731343285, + "grad_norm": 27.50669288635254, + "learning_rate": 9.880952380952381e-06, + "loss": 48.5058, + "step": 863 + }, + { + "epoch": 20.573134328358208, + "grad_norm": 27.739347457885742, + "learning_rate": 9.875000000000001e-06, + "loss": 49.676, + "step": 864 + }, + { + "epoch": 20.597014925373134, + "grad_norm": 22.63895034790039, + "learning_rate": 9.869047619047621e-06, + "loss": 47.6998, + "step": 865 + }, + { + "epoch": 20.62089552238806, + "grad_norm": 26.80891990661621, + "learning_rate": 9.863095238095239e-06, + "loss": 47.9571, + "step": 866 + }, + { + "epoch": 20.644776119402984, + "grad_norm": 26.259008407592773, + "learning_rate": 9.857142857142859e-06, + "loss": 48.8771, + "step": 867 + }, + { + "epoch": 20.66865671641791, + "grad_norm": 23.716773986816406, + "learning_rate": 9.851190476190477e-06, + "loss": 47.1255, + "step": 868 + }, + { + "epoch": 20.692537313432837, + "grad_norm": 26.96156120300293, + "learning_rate": 9.845238095238097e-06, + "loss": 47.2227, + "step": 869 + }, + { + "epoch": 20.71641791044776, + "grad_norm": 25.1954345703125, + "learning_rate": 9.839285714285715e-06, + "loss": 47.6847, + "step": 870 + }, + { + "epoch": 20.740298507462686, + "grad_norm": 21.56642723083496, + "learning_rate": 9.833333333333333e-06, + "loss": 47.7292, + "step": 871 + }, + { + "epoch": 20.764179104477613, + "grad_norm": 25.091773986816406, + "learning_rate": 9.827380952380953e-06, + "loss": 46.6588, + "step": 872 + }, + { + "epoch": 20.788059701492536, + "grad_norm": 26.45799446105957, + "learning_rate": 9.821428571428573e-06, + "loss": 47.3963, + "step": 873 + }, + { + "epoch": 20.811940298507462, + "grad_norm": 25.865068435668945, + "learning_rate": 9.81547619047619e-06, + "loss": 48.8823, + "step": 874 + }, + { + "epoch": 20.83582089552239, + "grad_norm": 27.056106567382812, + "learning_rate": 9.80952380952381e-06, + "loss": 47.2222, + "step": 875 + }, + { + "epoch": 20.85970149253731, + "grad_norm": 27.02417755126953, + "learning_rate": 9.803571428571428e-06, + "loss": 47.4543, + "step": 876 + }, + { + "epoch": 20.883582089552238, + "grad_norm": 23.681915283203125, + "learning_rate": 9.797619047619048e-06, + "loss": 47.7518, + "step": 877 + }, + { + "epoch": 20.907462686567165, + "grad_norm": 20.77193260192871, + "learning_rate": 9.791666666666666e-06, + "loss": 46.3164, + "step": 878 + }, + { + "epoch": 20.93134328358209, + "grad_norm": 24.61642837524414, + "learning_rate": 9.785714285714286e-06, + "loss": 48.6711, + "step": 879 + }, + { + "epoch": 20.955223880597014, + "grad_norm": 20.59898567199707, + "learning_rate": 9.779761904761906e-06, + "loss": 49.114, + "step": 880 + }, + { + "epoch": 20.97910447761194, + "grad_norm": 24.815736770629883, + "learning_rate": 9.773809523809524e-06, + "loss": 48.4315, + "step": 881 + }, + { + "epoch": 21.0, + "grad_norm": 17.920352935791016, + "learning_rate": 9.767857142857144e-06, + "loss": 41.3634, + "step": 882 + }, + { + "epoch": 21.023880597014927, + "grad_norm": 26.69571876525879, + "learning_rate": 9.761904761904762e-06, + "loss": 47.8968, + "step": 883 + }, + { + "epoch": 21.04776119402985, + "grad_norm": 23.156524658203125, + "learning_rate": 9.755952380952382e-06, + "loss": 48.5914, + "step": 884 + }, + { + "epoch": 21.071641791044776, + "grad_norm": 21.612483978271484, + "learning_rate": 9.75e-06, + "loss": 47.8711, + "step": 885 + }, + { + "epoch": 21.095522388059702, + "grad_norm": 24.346399307250977, + "learning_rate": 9.74404761904762e-06, + "loss": 48.8689, + "step": 886 + }, + { + "epoch": 21.119402985074625, + "grad_norm": 21.973896026611328, + "learning_rate": 9.73809523809524e-06, + "loss": 46.7465, + "step": 887 + }, + { + "epoch": 21.143283582089552, + "grad_norm": 20.034557342529297, + "learning_rate": 9.732142857142858e-06, + "loss": 47.1505, + "step": 888 + }, + { + "epoch": 21.16716417910448, + "grad_norm": 20.113008499145508, + "learning_rate": 9.726190476190477e-06, + "loss": 47.9234, + "step": 889 + }, + { + "epoch": 21.1910447761194, + "grad_norm": 24.743249893188477, + "learning_rate": 9.720238095238095e-06, + "loss": 47.4432, + "step": 890 + }, + { + "epoch": 21.214925373134328, + "grad_norm": 25.538530349731445, + "learning_rate": 9.714285714285715e-06, + "loss": 47.0015, + "step": 891 + }, + { + "epoch": 21.238805970149254, + "grad_norm": 27.43077278137207, + "learning_rate": 9.708333333333333e-06, + "loss": 48.6757, + "step": 892 + }, + { + "epoch": 21.262686567164177, + "grad_norm": 25.34470558166504, + "learning_rate": 9.702380952380953e-06, + "loss": 46.8118, + "step": 893 + }, + { + "epoch": 21.286567164179104, + "grad_norm": 29.590490341186523, + "learning_rate": 9.696428571428573e-06, + "loss": 47.5079, + "step": 894 + }, + { + "epoch": 21.31044776119403, + "grad_norm": 16.418222427368164, + "learning_rate": 9.690476190476191e-06, + "loss": 48.7679, + "step": 895 + }, + { + "epoch": 21.334328358208957, + "grad_norm": 30.906719207763672, + "learning_rate": 9.68452380952381e-06, + "loss": 48.4926, + "step": 896 + }, + { + "epoch": 21.35820895522388, + "grad_norm": 30.252347946166992, + "learning_rate": 9.678571428571429e-06, + "loss": 48.9318, + "step": 897 + }, + { + "epoch": 21.382089552238806, + "grad_norm": 30.137592315673828, + "learning_rate": 9.672619047619049e-06, + "loss": 47.0388, + "step": 898 + }, + { + "epoch": 21.405970149253733, + "grad_norm": 25.297151565551758, + "learning_rate": 9.666666666666667e-06, + "loss": 47.334, + "step": 899 + }, + { + "epoch": 21.429850746268656, + "grad_norm": 31.72736358642578, + "learning_rate": 9.660714285714287e-06, + "loss": 47.8769, + "step": 900 + }, + { + "epoch": 21.453731343283582, + "grad_norm": 24.4852294921875, + "learning_rate": 9.654761904761906e-06, + "loss": 47.4009, + "step": 901 + }, + { + "epoch": 21.47761194029851, + "grad_norm": 31.223567962646484, + "learning_rate": 9.648809523809524e-06, + "loss": 48.4972, + "step": 902 + }, + { + "epoch": 21.501492537313432, + "grad_norm": 24.1851806640625, + "learning_rate": 9.642857142857144e-06, + "loss": 46.1818, + "step": 903 + }, + { + "epoch": 21.52537313432836, + "grad_norm": NaN, + "learning_rate": 9.636904761904762e-06, + "loss": 70.1176, + "step": 904 + }, + { + "epoch": 21.549253731343285, + "grad_norm": 29.140161514282227, + "learning_rate": 9.636904761904762e-06, + "loss": 47.1614, + "step": 905 + }, + { + "epoch": 21.573134328358208, + "grad_norm": 31.186546325683594, + "learning_rate": 9.630952380952382e-06, + "loss": 47.3643, + "step": 906 + }, + { + "epoch": 21.597014925373134, + "grad_norm": 24.395353317260742, + "learning_rate": 9.625e-06, + "loss": 48.3591, + "step": 907 + }, + { + "epoch": 21.62089552238806, + "grad_norm": 29.287492752075195, + "learning_rate": 9.61904761904762e-06, + "loss": 47.261, + "step": 908 + }, + { + "epoch": 21.644776119402984, + "grad_norm": 26.76996612548828, + "learning_rate": 9.61309523809524e-06, + "loss": 48.7017, + "step": 909 + }, + { + "epoch": 21.66865671641791, + "grad_norm": 29.820920944213867, + "learning_rate": 9.607142857142858e-06, + "loss": 48.5165, + "step": 910 + }, + { + "epoch": 21.692537313432837, + "grad_norm": 30.011823654174805, + "learning_rate": 9.601190476190478e-06, + "loss": 46.5558, + "step": 911 + }, + { + "epoch": 21.71641791044776, + "grad_norm": 32.796905517578125, + "learning_rate": 9.595238095238096e-06, + "loss": 47.276, + "step": 912 + }, + { + "epoch": 21.740298507462686, + "grad_norm": 28.798233032226562, + "learning_rate": 9.589285714285716e-06, + "loss": 47.6033, + "step": 913 + }, + { + "epoch": 21.764179104477613, + "grad_norm": 31.51072120666504, + "learning_rate": 9.583333333333335e-06, + "loss": 48.1236, + "step": 914 + }, + { + "epoch": 21.788059701492536, + "grad_norm": 20.611305236816406, + "learning_rate": 9.577380952380953e-06, + "loss": 48.2839, + "step": 915 + }, + { + "epoch": 21.811940298507462, + "grad_norm": 26.748571395874023, + "learning_rate": 9.571428571428573e-06, + "loss": 48.2225, + "step": 916 + }, + { + "epoch": 21.83582089552239, + "grad_norm": 22.262859344482422, + "learning_rate": 9.565476190476191e-06, + "loss": 46.661, + "step": 917 + }, + { + "epoch": 21.85970149253731, + "grad_norm": 34.15045166015625, + "learning_rate": 9.559523809523811e-06, + "loss": 47.3229, + "step": 918 + }, + { + "epoch": 21.883582089552238, + "grad_norm": 24.26387596130371, + "learning_rate": 9.55357142857143e-06, + "loss": 47.4686, + "step": 919 + }, + { + "epoch": 21.907462686567165, + "grad_norm": 29.463472366333008, + "learning_rate": 9.547619047619049e-06, + "loss": 47.6019, + "step": 920 + }, + { + "epoch": 21.93134328358209, + "grad_norm": 31.184497833251953, + "learning_rate": 9.541666666666669e-06, + "loss": 47.3228, + "step": 921 + }, + { + "epoch": 21.955223880597014, + "grad_norm": 26.506031036376953, + "learning_rate": 9.535714285714287e-06, + "loss": 47.9961, + "step": 922 + }, + { + "epoch": 21.97910447761194, + "grad_norm": 30.547340393066406, + "learning_rate": 9.529761904761905e-06, + "loss": 47.8973, + "step": 923 + }, + { + "epoch": 22.0, + "grad_norm": 22.91999053955078, + "learning_rate": 9.523809523809525e-06, + "loss": 41.3426, + "step": 924 + }, + { + "epoch": 22.023880597014927, + "grad_norm": 28.242450714111328, + "learning_rate": 9.517857142857143e-06, + "loss": 47.2478, + "step": 925 + }, + { + "epoch": 22.04776119402985, + "grad_norm": 33.07649612426758, + "learning_rate": 9.511904761904763e-06, + "loss": 47.6489, + "step": 926 + }, + { + "epoch": 22.071641791044776, + "grad_norm": 28.14696502685547, + "learning_rate": 9.50595238095238e-06, + "loss": 46.521, + "step": 927 + }, + { + "epoch": 22.095522388059702, + "grad_norm": 34.472206115722656, + "learning_rate": 9.5e-06, + "loss": 47.6476, + "step": 928 + }, + { + "epoch": 22.119402985074625, + "grad_norm": 25.370718002319336, + "learning_rate": 9.494047619047619e-06, + "loss": 47.7215, + "step": 929 + }, + { + "epoch": 22.143283582089552, + "grad_norm": 31.77129554748535, + "learning_rate": 9.488095238095238e-06, + "loss": 46.5566, + "step": 930 + }, + { + "epoch": 22.16716417910448, + "grad_norm": 25.42667579650879, + "learning_rate": 9.482142857142858e-06, + "loss": 47.9832, + "step": 931 + }, + { + "epoch": 22.1910447761194, + "grad_norm": 26.3134765625, + "learning_rate": 9.476190476190476e-06, + "loss": 47.9402, + "step": 932 + }, + { + "epoch": 22.214925373134328, + "grad_norm": 31.683523178100586, + "learning_rate": 9.470238095238096e-06, + "loss": 47.4404, + "step": 933 + }, + { + "epoch": 22.238805970149254, + "grad_norm": 31.90761375427246, + "learning_rate": 9.464285714285714e-06, + "loss": 47.7601, + "step": 934 + }, + { + "epoch": 22.262686567164177, + "grad_norm": 24.635921478271484, + "learning_rate": 9.458333333333334e-06, + "loss": 46.2573, + "step": 935 + }, + { + "epoch": 22.286567164179104, + "grad_norm": 25.32915496826172, + "learning_rate": 9.452380952380952e-06, + "loss": 48.4756, + "step": 936 + }, + { + "epoch": 22.31044776119403, + "grad_norm": 28.117773056030273, + "learning_rate": 9.446428571428572e-06, + "loss": 48.6971, + "step": 937 + }, + { + "epoch": 22.334328358208957, + "grad_norm": 22.504152297973633, + "learning_rate": 9.440476190476192e-06, + "loss": 47.4534, + "step": 938 + }, + { + "epoch": 22.35820895522388, + "grad_norm": 31.765676498413086, + "learning_rate": 9.43452380952381e-06, + "loss": 48.0168, + "step": 939 + }, + { + "epoch": 22.382089552238806, + "grad_norm": 27.647945404052734, + "learning_rate": 9.42857142857143e-06, + "loss": 48.0918, + "step": 940 + }, + { + "epoch": 22.405970149253733, + "grad_norm": 33.35643005371094, + "learning_rate": 9.422619047619048e-06, + "loss": 48.295, + "step": 941 + }, + { + "epoch": 22.429850746268656, + "grad_norm": 26.12603187561035, + "learning_rate": 9.416666666666667e-06, + "loss": 48.8921, + "step": 942 + }, + { + "epoch": 22.453731343283582, + "grad_norm": 23.728809356689453, + "learning_rate": 9.410714285714286e-06, + "loss": 47.3206, + "step": 943 + }, + { + "epoch": 22.47761194029851, + "grad_norm": 28.772401809692383, + "learning_rate": 9.404761904761905e-06, + "loss": 47.6536, + "step": 944 + }, + { + "epoch": 22.501492537313432, + "grad_norm": 28.205202102661133, + "learning_rate": 9.398809523809525e-06, + "loss": 47.1952, + "step": 945 + }, + { + "epoch": 22.52537313432836, + "grad_norm": 33.80730438232422, + "learning_rate": 9.392857142857143e-06, + "loss": 47.1336, + "step": 946 + }, + { + "epoch": 22.549253731343285, + "grad_norm": 25.538846969604492, + "learning_rate": 9.386904761904763e-06, + "loss": 46.4229, + "step": 947 + }, + { + "epoch": 22.573134328358208, + "grad_norm": 41.13503646850586, + "learning_rate": 9.380952380952381e-06, + "loss": 46.8325, + "step": 948 + }, + { + "epoch": 22.597014925373134, + "grad_norm": 36.823001861572266, + "learning_rate": 9.375000000000001e-06, + "loss": 47.205, + "step": 949 + }, + { + "epoch": 22.62089552238806, + "grad_norm": 29.992229461669922, + "learning_rate": 9.36904761904762e-06, + "loss": 46.683, + "step": 950 + }, + { + "epoch": 22.644776119402984, + "grad_norm": 40.20172882080078, + "learning_rate": 9.363095238095239e-06, + "loss": 48.4859, + "step": 951 + }, + { + "epoch": 22.66865671641791, + "grad_norm": 27.357097625732422, + "learning_rate": 9.357142857142859e-06, + "loss": 47.2987, + "step": 952 + }, + { + "epoch": 22.692537313432837, + "grad_norm": 40.66689682006836, + "learning_rate": 9.351190476190477e-06, + "loss": 46.3579, + "step": 953 + }, + { + "epoch": 22.71641791044776, + "grad_norm": 35.37788391113281, + "learning_rate": 9.345238095238096e-06, + "loss": 47.3369, + "step": 954 + }, + { + "epoch": 22.740298507462686, + "grad_norm": 36.279151916503906, + "learning_rate": 9.339285714285715e-06, + "loss": 47.1137, + "step": 955 + }, + { + "epoch": 22.764179104477613, + "grad_norm": 27.949628829956055, + "learning_rate": 9.333333333333334e-06, + "loss": 47.1438, + "step": 956 + }, + { + "epoch": 22.788059701492536, + "grad_norm": 45.424556732177734, + "learning_rate": 9.327380952380954e-06, + "loss": 48.3171, + "step": 957 + }, + { + "epoch": 22.811940298507462, + "grad_norm": 27.726537704467773, + "learning_rate": 9.321428571428572e-06, + "loss": 47.1718, + "step": 958 + }, + { + "epoch": 22.83582089552239, + "grad_norm": 58.36731719970703, + "learning_rate": 9.315476190476192e-06, + "loss": 47.5895, + "step": 959 + }, + { + "epoch": 22.85970149253731, + "grad_norm": 58.96028137207031, + "learning_rate": 9.30952380952381e-06, + "loss": 47.4109, + "step": 960 + }, + { + "epoch": 22.883582089552238, + "grad_norm": 24.928117752075195, + "learning_rate": 9.30357142857143e-06, + "loss": 48.1841, + "step": 961 + }, + { + "epoch": 22.907462686567165, + "grad_norm": 38.36846160888672, + "learning_rate": 9.297619047619048e-06, + "loss": 47.7438, + "step": 962 + }, + { + "epoch": 22.93134328358209, + "grad_norm": 37.60481643676758, + "learning_rate": 9.291666666666668e-06, + "loss": 46.5067, + "step": 963 + }, + { + "epoch": 22.955223880597014, + "grad_norm": NaN, + "learning_rate": 9.285714285714288e-06, + "loss": 78.3124, + "step": 964 + }, + { + "epoch": 22.97910447761194, + "grad_norm": 28.587425231933594, + "learning_rate": 9.285714285714288e-06, + "loss": 47.1599, + "step": 965 + }, + { + "epoch": 23.0, + "grad_norm": 41.493404388427734, + "learning_rate": 9.279761904761906e-06, + "loss": 41.2983, + "step": 966 + }, + { + "epoch": 23.023880597014927, + "grad_norm": 41.00606918334961, + "learning_rate": 9.273809523809525e-06, + "loss": 46.8696, + "step": 967 + }, + { + "epoch": 23.04776119402985, + "grad_norm": 31.043148040771484, + "learning_rate": 9.267857142857144e-06, + "loss": 46.4614, + "step": 968 + }, + { + "epoch": 23.071641791044776, + "grad_norm": 36.815940856933594, + "learning_rate": 9.261904761904763e-06, + "loss": 47.5987, + "step": 969 + }, + { + "epoch": 23.095522388059702, + "grad_norm": 35.73536682128906, + "learning_rate": 9.255952380952381e-06, + "loss": 47.8339, + "step": 970 + }, + { + "epoch": 23.119402985074625, + "grad_norm": 26.95656967163086, + "learning_rate": 9.250000000000001e-06, + "loss": 48.0632, + "step": 971 + }, + { + "epoch": 23.143283582089552, + "grad_norm": 40.408348083496094, + "learning_rate": 9.244047619047621e-06, + "loss": 47.5458, + "step": 972 + }, + { + "epoch": 23.16716417910448, + "grad_norm": 41.97018051147461, + "learning_rate": 9.238095238095239e-06, + "loss": 48.8528, + "step": 973 + }, + { + "epoch": 23.1910447761194, + "grad_norm": 23.809162139892578, + "learning_rate": 9.232142857142859e-06, + "loss": 47.663, + "step": 974 + }, + { + "epoch": 23.214925373134328, + "grad_norm": 36.0232048034668, + "learning_rate": 9.226190476190477e-06, + "loss": 47.496, + "step": 975 + }, + { + "epoch": 23.238805970149254, + "grad_norm": 32.06623077392578, + "learning_rate": 9.220238095238097e-06, + "loss": 47.4472, + "step": 976 + }, + { + "epoch": 23.262686567164177, + "grad_norm": 30.663307189941406, + "learning_rate": 9.214285714285715e-06, + "loss": 47.1342, + "step": 977 + }, + { + "epoch": 23.286567164179104, + "grad_norm": 39.121437072753906, + "learning_rate": 9.208333333333333e-06, + "loss": 47.9977, + "step": 978 + }, + { + "epoch": 23.31044776119403, + "grad_norm": 31.75649642944336, + "learning_rate": 9.202380952380953e-06, + "loss": 49.2196, + "step": 979 + }, + { + "epoch": 23.334328358208957, + "grad_norm": 50.10381317138672, + "learning_rate": 9.196428571428571e-06, + "loss": 47.6487, + "step": 980 + }, + { + "epoch": 23.35820895522388, + "grad_norm": 36.412906646728516, + "learning_rate": 9.19047619047619e-06, + "loss": 47.0012, + "step": 981 + }, + { + "epoch": 23.382089552238806, + "grad_norm": 40.47570037841797, + "learning_rate": 9.18452380952381e-06, + "loss": 45.4449, + "step": 982 + }, + { + "epoch": 23.405970149253733, + "grad_norm": 43.92324447631836, + "learning_rate": 9.178571428571429e-06, + "loss": 47.8727, + "step": 983 + }, + { + "epoch": 23.429850746268656, + "grad_norm": 28.896121978759766, + "learning_rate": 9.172619047619048e-06, + "loss": 47.8489, + "step": 984 + }, + { + "epoch": 23.453731343283582, + "grad_norm": 37.02536392211914, + "learning_rate": 9.166666666666666e-06, + "loss": 48.4484, + "step": 985 + }, + { + "epoch": 23.47761194029851, + "grad_norm": 26.289518356323242, + "learning_rate": 9.160714285714286e-06, + "loss": 47.0221, + "step": 986 + }, + { + "epoch": 23.501492537313432, + "grad_norm": 33.60945129394531, + "learning_rate": 9.154761904761906e-06, + "loss": 47.7454, + "step": 987 + }, + { + "epoch": 23.52537313432836, + "grad_norm": 31.284311294555664, + "learning_rate": 9.148809523809524e-06, + "loss": 47.0558, + "step": 988 + }, + { + "epoch": 23.549253731343285, + "grad_norm": 30.488906860351562, + "learning_rate": 9.142857142857144e-06, + "loss": 46.4408, + "step": 989 + }, + { + "epoch": 23.573134328358208, + "grad_norm": 34.29289245605469, + "learning_rate": 9.136904761904762e-06, + "loss": 46.2796, + "step": 990 + }, + { + "epoch": 23.597014925373134, + "grad_norm": 22.803457260131836, + "learning_rate": 9.130952380952382e-06, + "loss": 47.2684, + "step": 991 + }, + { + "epoch": 23.62089552238806, + "grad_norm": 33.18730926513672, + "learning_rate": 9.125e-06, + "loss": 47.5, + "step": 992 + }, + { + "epoch": 23.644776119402984, + "grad_norm": 28.421592712402344, + "learning_rate": 9.11904761904762e-06, + "loss": 46.7508, + "step": 993 + }, + { + "epoch": 23.66865671641791, + "grad_norm": 23.492319107055664, + "learning_rate": 9.11309523809524e-06, + "loss": 46.6042, + "step": 994 + }, + { + "epoch": 23.692537313432837, + "grad_norm": 30.10877227783203, + "learning_rate": 9.107142857142858e-06, + "loss": 46.2632, + "step": 995 + }, + { + "epoch": 23.71641791044776, + "grad_norm": 23.64444351196289, + "learning_rate": 9.101190476190477e-06, + "loss": 47.2817, + "step": 996 + }, + { + "epoch": 23.740298507462686, + "grad_norm": 28.243606567382812, + "learning_rate": 9.095238095238095e-06, + "loss": 47.1196, + "step": 997 + }, + { + "epoch": 23.764179104477613, + "grad_norm": 26.84208869934082, + "learning_rate": 9.089285714285715e-06, + "loss": 46.6631, + "step": 998 + }, + { + "epoch": 23.788059701492536, + "grad_norm": 29.558794021606445, + "learning_rate": 9.083333333333333e-06, + "loss": 45.8711, + "step": 999 + }, + { + "epoch": 23.811940298507462, + "grad_norm": 25.105928421020508, + "learning_rate": 9.077380952380953e-06, + "loss": 47.8253, + "step": 1000 + }, + { + "epoch": 23.83582089552239, + "grad_norm": NaN, + "learning_rate": 9.071428571428573e-06, + "loss": 82.5048, + "step": 1001 + }, + { + "epoch": 23.85970149253731, + "grad_norm": 25.548643112182617, + "learning_rate": 9.071428571428573e-06, + "loss": 47.5042, + "step": 1002 + }, + { + "epoch": 23.883582089552238, + "grad_norm": 28.8011531829834, + "learning_rate": 9.065476190476191e-06, + "loss": 47.0084, + "step": 1003 + }, + { + "epoch": 23.907462686567165, + "grad_norm": 31.907651901245117, + "learning_rate": 9.05952380952381e-06, + "loss": 48.287, + "step": 1004 + }, + { + "epoch": 23.93134328358209, + "grad_norm": 32.044986724853516, + "learning_rate": 9.053571428571429e-06, + "loss": 47.276, + "step": 1005 + }, + { + "epoch": 23.955223880597014, + "grad_norm": 31.224260330200195, + "learning_rate": 9.047619047619049e-06, + "loss": 47.4774, + "step": 1006 + }, + { + "epoch": 23.97910447761194, + "grad_norm": 29.830835342407227, + "learning_rate": 9.041666666666667e-06, + "loss": 47.7031, + "step": 1007 + }, + { + "epoch": 24.0, + "grad_norm": 25.12934112548828, + "learning_rate": 9.035714285714287e-06, + "loss": 41.8156, + "step": 1008 + }, + { + "epoch": 24.023880597014927, + "grad_norm": 31.172348022460938, + "learning_rate": 9.029761904761906e-06, + "loss": 48.0591, + "step": 1009 + }, + { + "epoch": 24.04776119402985, + "grad_norm": 26.59412956237793, + "learning_rate": 9.023809523809524e-06, + "loss": 47.6291, + "step": 1010 + }, + { + "epoch": 24.071641791044776, + "grad_norm": 29.16905975341797, + "learning_rate": 9.017857142857144e-06, + "loss": 47.4587, + "step": 1011 + }, + { + "epoch": 24.095522388059702, + "grad_norm": 33.05836868286133, + "learning_rate": 9.011904761904762e-06, + "loss": 47.8748, + "step": 1012 + }, + { + "epoch": 24.119402985074625, + "grad_norm": 26.13016700744629, + "learning_rate": 9.005952380952382e-06, + "loss": 48.0003, + "step": 1013 + }, + { + "epoch": 24.143283582089552, + "grad_norm": 29.883411407470703, + "learning_rate": 9e-06, + "loss": 47.0188, + "step": 1014 + }, + { + "epoch": 24.16716417910448, + "grad_norm": 29.039255142211914, + "learning_rate": 8.99404761904762e-06, + "loss": 46.8844, + "step": 1015 + }, + { + "epoch": 24.1910447761194, + "grad_norm": 26.532760620117188, + "learning_rate": 8.98809523809524e-06, + "loss": 47.1817, + "step": 1016 + }, + { + "epoch": 24.214925373134328, + "grad_norm": 30.146087646484375, + "learning_rate": 8.982142857142858e-06, + "loss": 46.4863, + "step": 1017 + }, + { + "epoch": 24.238805970149254, + "grad_norm": 27.027935028076172, + "learning_rate": 8.976190476190478e-06, + "loss": 45.7162, + "step": 1018 + }, + { + "epoch": 24.262686567164177, + "grad_norm": 27.315515518188477, + "learning_rate": 8.970238095238096e-06, + "loss": 46.6337, + "step": 1019 + }, + { + "epoch": 24.286567164179104, + "grad_norm": 25.63303565979004, + "learning_rate": 8.964285714285716e-06, + "loss": 46.5452, + "step": 1020 + }, + { + "epoch": 24.31044776119403, + "grad_norm": 22.407268524169922, + "learning_rate": 8.958333333333334e-06, + "loss": 47.3262, + "step": 1021 + }, + { + "epoch": 24.334328358208957, + "grad_norm": 19.4051513671875, + "learning_rate": 8.952380952380953e-06, + "loss": 46.8407, + "step": 1022 + }, + { + "epoch": 24.35820895522388, + "grad_norm": NaN, + "learning_rate": 8.946428571428573e-06, + "loss": 77.1735, + "step": 1023 + }, + { + "epoch": 24.382089552238806, + "grad_norm": 26.870166778564453, + "learning_rate": 8.946428571428573e-06, + "loss": 46.9395, + "step": 1024 + }, + { + "epoch": 24.405970149253733, + "grad_norm": 21.54165267944336, + "learning_rate": 8.940476190476191e-06, + "loss": 47.2505, + "step": 1025 + }, + { + "epoch": 24.429850746268656, + "grad_norm": 29.317501068115234, + "learning_rate": 8.934523809523811e-06, + "loss": 48.0554, + "step": 1026 + }, + { + "epoch": 24.453731343283582, + "grad_norm": 27.45809555053711, + "learning_rate": 8.92857142857143e-06, + "loss": 47.0495, + "step": 1027 + }, + { + "epoch": 24.47761194029851, + "grad_norm": NaN, + "learning_rate": 8.922619047619049e-06, + "loss": 66.51, + "step": 1028 + }, + { + "epoch": 24.501492537313432, + "grad_norm": 23.169204711914062, + "learning_rate": 8.922619047619049e-06, + "loss": 47.5902, + "step": 1029 + }, + { + "epoch": 24.52537313432836, + "grad_norm": 31.986774444580078, + "learning_rate": 8.916666666666667e-06, + "loss": 47.4281, + "step": 1030 + }, + { + "epoch": 24.549253731343285, + "grad_norm": 27.15190315246582, + "learning_rate": 8.910714285714287e-06, + "loss": 46.3638, + "step": 1031 + }, + { + "epoch": 24.573134328358208, + "grad_norm": 26.88028335571289, + "learning_rate": 8.904761904761905e-06, + "loss": 45.0491, + "step": 1032 + }, + { + "epoch": 24.597014925373134, + "grad_norm": 27.693952560424805, + "learning_rate": 8.898809523809525e-06, + "loss": 47.6471, + "step": 1033 + }, + { + "epoch": 24.62089552238806, + "grad_norm": 33.45442581176758, + "learning_rate": 8.892857142857143e-06, + "loss": 47.1459, + "step": 1034 + }, + { + "epoch": 24.644776119402984, + "grad_norm": 29.933320999145508, + "learning_rate": 8.886904761904763e-06, + "loss": 46.9218, + "step": 1035 + }, + { + "epoch": 24.66865671641791, + "grad_norm": 26.401569366455078, + "learning_rate": 8.88095238095238e-06, + "loss": 47.7027, + "step": 1036 + }, + { + "epoch": 24.692537313432837, + "grad_norm": 26.92498016357422, + "learning_rate": 8.875e-06, + "loss": 47.2302, + "step": 1037 + }, + { + "epoch": 24.71641791044776, + "grad_norm": 28.368043899536133, + "learning_rate": 8.869047619047619e-06, + "loss": 47.4479, + "step": 1038 + }, + { + "epoch": 24.740298507462686, + "grad_norm": 27.319650650024414, + "learning_rate": 8.863095238095238e-06, + "loss": 47.4652, + "step": 1039 + }, + { + "epoch": 24.764179104477613, + "grad_norm": 37.10929870605469, + "learning_rate": 8.857142857142858e-06, + "loss": 47.3543, + "step": 1040 + }, + { + "epoch": 24.788059701492536, + "grad_norm": 32.430416107177734, + "learning_rate": 8.851190476190476e-06, + "loss": 46.1406, + "step": 1041 + }, + { + "epoch": 24.811940298507462, + "grad_norm": 33.29399108886719, + "learning_rate": 8.845238095238096e-06, + "loss": 47.1917, + "step": 1042 + }, + { + "epoch": 24.83582089552239, + "grad_norm": 32.72507095336914, + "learning_rate": 8.839285714285714e-06, + "loss": 47.677, + "step": 1043 + }, + { + "epoch": 24.85970149253731, + "grad_norm": 25.997148513793945, + "learning_rate": 8.833333333333334e-06, + "loss": 47.6851, + "step": 1044 + }, + { + "epoch": 24.883582089552238, + "grad_norm": 33.00047302246094, + "learning_rate": 8.827380952380952e-06, + "loss": 47.5326, + "step": 1045 + }, + { + "epoch": 24.907462686567165, + "grad_norm": 33.95719528198242, + "learning_rate": 8.821428571428572e-06, + "loss": 47.2836, + "step": 1046 + }, + { + "epoch": 24.93134328358209, + "grad_norm": 31.353008270263672, + "learning_rate": 8.815476190476192e-06, + "loss": 47.8318, + "step": 1047 + }, + { + "epoch": 24.955223880597014, + "grad_norm": 27.4250545501709, + "learning_rate": 8.80952380952381e-06, + "loss": 47.0066, + "step": 1048 + }, + { + "epoch": 24.97910447761194, + "grad_norm": 30.22010612487793, + "learning_rate": 8.80357142857143e-06, + "loss": 46.6991, + "step": 1049 + }, + { + "epoch": 25.0, + "grad_norm": 19.693180084228516, + "learning_rate": 8.797619047619048e-06, + "loss": 41.6055, + "step": 1050 + }, + { + "epoch": 25.023880597014927, + "grad_norm": 24.590662002563477, + "learning_rate": 8.791666666666667e-06, + "loss": 46.3536, + "step": 1051 + }, + { + "epoch": 25.04776119402985, + "grad_norm": 28.37199592590332, + "learning_rate": 8.785714285714286e-06, + "loss": 47.8334, + "step": 1052 + }, + { + "epoch": 25.071641791044776, + "grad_norm": 26.38755226135254, + "learning_rate": 8.779761904761905e-06, + "loss": 47.6486, + "step": 1053 + }, + { + "epoch": 25.095522388059702, + "grad_norm": 27.338485717773438, + "learning_rate": 8.773809523809525e-06, + "loss": 47.3044, + "step": 1054 + }, + { + "epoch": 25.119402985074625, + "grad_norm": 25.308486938476562, + "learning_rate": 8.767857142857143e-06, + "loss": 45.6873, + "step": 1055 + }, + { + "epoch": 25.143283582089552, + "grad_norm": 30.886962890625, + "learning_rate": 8.761904761904763e-06, + "loss": 46.8938, + "step": 1056 + }, + { + "epoch": 25.16716417910448, + "grad_norm": 25.25688934326172, + "learning_rate": 8.755952380952381e-06, + "loss": 47.4858, + "step": 1057 + }, + { + "epoch": 25.1910447761194, + "grad_norm": 30.462963104248047, + "learning_rate": 8.750000000000001e-06, + "loss": 46.6334, + "step": 1058 + }, + { + "epoch": 25.214925373134328, + "grad_norm": 22.87471580505371, + "learning_rate": 8.744047619047619e-06, + "loss": 46.0966, + "step": 1059 + }, + { + "epoch": 25.238805970149254, + "grad_norm": 23.413904190063477, + "learning_rate": 8.738095238095239e-06, + "loss": 46.8938, + "step": 1060 + }, + { + "epoch": 25.262686567164177, + "grad_norm": 26.926279067993164, + "learning_rate": 8.732142857142859e-06, + "loss": 46.3773, + "step": 1061 + }, + { + "epoch": 25.286567164179104, + "grad_norm": 27.595348358154297, + "learning_rate": 8.726190476190477e-06, + "loss": 48.0235, + "step": 1062 + }, + { + "epoch": 25.31044776119403, + "grad_norm": 26.124523162841797, + "learning_rate": 8.720238095238096e-06, + "loss": 46.6863, + "step": 1063 + }, + { + "epoch": 25.334328358208957, + "grad_norm": 28.308120727539062, + "learning_rate": 8.714285714285715e-06, + "loss": 47.7158, + "step": 1064 + }, + { + "epoch": 25.35820895522388, + "grad_norm": 23.434846878051758, + "learning_rate": 8.708333333333334e-06, + "loss": 47.2951, + "step": 1065 + }, + { + "epoch": 25.382089552238806, + "grad_norm": 26.917911529541016, + "learning_rate": 8.702380952380952e-06, + "loss": 45.7266, + "step": 1066 + }, + { + "epoch": 25.405970149253733, + "grad_norm": 19.7725772857666, + "learning_rate": 8.696428571428572e-06, + "loss": 46.5458, + "step": 1067 + }, + { + "epoch": 25.429850746268656, + "grad_norm": 27.18629264831543, + "learning_rate": 8.690476190476192e-06, + "loss": 46.3133, + "step": 1068 + }, + { + "epoch": 25.453731343283582, + "grad_norm": 26.112865447998047, + "learning_rate": 8.68452380952381e-06, + "loss": 46.6383, + "step": 1069 + }, + { + "epoch": 25.47761194029851, + "grad_norm": 19.385990142822266, + "learning_rate": 8.67857142857143e-06, + "loss": 46.5541, + "step": 1070 + }, + { + "epoch": 25.501492537313432, + "grad_norm": 26.713350296020508, + "learning_rate": 8.672619047619048e-06, + "loss": 48.045, + "step": 1071 + }, + { + "epoch": 25.52537313432836, + "grad_norm": 29.80147933959961, + "learning_rate": 8.666666666666668e-06, + "loss": 47.4443, + "step": 1072 + }, + { + "epoch": 25.549253731343285, + "grad_norm": 23.674266815185547, + "learning_rate": 8.660714285714286e-06, + "loss": 46.6662, + "step": 1073 + }, + { + "epoch": 25.573134328358208, + "grad_norm": 46.435401916503906, + "learning_rate": 8.654761904761906e-06, + "loss": 46.9276, + "step": 1074 + }, + { + "epoch": 25.597014925373134, + "grad_norm": 35.016502380371094, + "learning_rate": 8.648809523809526e-06, + "loss": 47.6811, + "step": 1075 + }, + { + "epoch": 25.62089552238806, + "grad_norm": 42.57990646362305, + "learning_rate": 8.642857142857144e-06, + "loss": 46.5684, + "step": 1076 + }, + { + "epoch": 25.644776119402984, + "grad_norm": 36.2376708984375, + "learning_rate": 8.636904761904763e-06, + "loss": 46.1807, + "step": 1077 + }, + { + "epoch": 25.66865671641791, + "grad_norm": 41.44023895263672, + "learning_rate": 8.630952380952381e-06, + "loss": 46.0823, + "step": 1078 + }, + { + "epoch": 25.692537313432837, + "grad_norm": 43.62863540649414, + "learning_rate": 8.625000000000001e-06, + "loss": 47.9958, + "step": 1079 + }, + { + "epoch": 25.71641791044776, + "grad_norm": 34.232120513916016, + "learning_rate": 8.61904761904762e-06, + "loss": 47.9585, + "step": 1080 + }, + { + "epoch": 25.740298507462686, + "grad_norm": 38.023197174072266, + "learning_rate": 8.61309523809524e-06, + "loss": 47.7344, + "step": 1081 + }, + { + "epoch": 25.764179104477613, + "grad_norm": 37.89833068847656, + "learning_rate": 8.607142857142859e-06, + "loss": 47.8956, + "step": 1082 + }, + { + "epoch": 25.788059701492536, + "grad_norm": 33.03269958496094, + "learning_rate": 8.601190476190477e-06, + "loss": 47.7408, + "step": 1083 + }, + { + "epoch": 25.811940298507462, + "grad_norm": 44.31171798706055, + "learning_rate": 8.595238095238097e-06, + "loss": 47.6232, + "step": 1084 + }, + { + "epoch": 25.83582089552239, + "grad_norm": 42.54961395263672, + "learning_rate": 8.589285714285715e-06, + "loss": 47.9757, + "step": 1085 + }, + { + "epoch": 25.85970149253731, + "grad_norm": 27.695526123046875, + "learning_rate": 8.583333333333333e-06, + "loss": 47.0934, + "step": 1086 + }, + { + "epoch": 25.883582089552238, + "grad_norm": 32.62801742553711, + "learning_rate": 8.577380952380953e-06, + "loss": 47.1186, + "step": 1087 + }, + { + "epoch": 25.907462686567165, + "grad_norm": 26.777305603027344, + "learning_rate": 8.571428571428571e-06, + "loss": 47.2931, + "step": 1088 + }, + { + "epoch": 25.93134328358209, + "grad_norm": 24.382678985595703, + "learning_rate": 8.56547619047619e-06, + "loss": 46.4698, + "step": 1089 + }, + { + "epoch": 25.955223880597014, + "grad_norm": 34.310150146484375, + "learning_rate": 8.55952380952381e-06, + "loss": 46.0509, + "step": 1090 + }, + { + "epoch": 25.97910447761194, + "grad_norm": 27.468976974487305, + "learning_rate": 8.553571428571429e-06, + "loss": 46.9123, + "step": 1091 + }, + { + "epoch": 26.0, + "grad_norm": 27.90901756286621, + "learning_rate": 8.547619047619048e-06, + "loss": 41.8265, + "step": 1092 + }, + { + "epoch": 26.023880597014927, + "grad_norm": 28.853416442871094, + "learning_rate": 8.541666666666666e-06, + "loss": 47.9612, + "step": 1093 + }, + { + "epoch": 26.04776119402985, + "grad_norm": 31.96144676208496, + "learning_rate": 8.535714285714286e-06, + "loss": 46.8167, + "step": 1094 + }, + { + "epoch": 26.071641791044776, + "grad_norm": 33.179141998291016, + "learning_rate": 8.529761904761904e-06, + "loss": 48.0464, + "step": 1095 + }, + { + "epoch": 26.095522388059702, + "grad_norm": 32.18705368041992, + "learning_rate": 8.523809523809524e-06, + "loss": 45.6743, + "step": 1096 + }, + { + "epoch": 26.119402985074625, + "grad_norm": 26.125934600830078, + "learning_rate": 8.517857142857144e-06, + "loss": 46.4944, + "step": 1097 + }, + { + "epoch": 26.143283582089552, + "grad_norm": 31.666461944580078, + "learning_rate": 8.511904761904762e-06, + "loss": 47.6152, + "step": 1098 + }, + { + "epoch": 26.16716417910448, + "grad_norm": 29.90437889099121, + "learning_rate": 8.505952380952382e-06, + "loss": 46.4497, + "step": 1099 + }, + { + "epoch": 26.1910447761194, + "grad_norm": 32.59938430786133, + "learning_rate": 8.5e-06, + "loss": 47.1877, + "step": 1100 + }, + { + "epoch": 26.214925373134328, + "grad_norm": 28.368562698364258, + "learning_rate": 8.49404761904762e-06, + "loss": 46.4898, + "step": 1101 + }, + { + "epoch": 26.238805970149254, + "grad_norm": 31.274070739746094, + "learning_rate": 8.488095238095238e-06, + "loss": 47.1507, + "step": 1102 + }, + { + "epoch": 26.262686567164177, + "grad_norm": 24.63444709777832, + "learning_rate": 8.482142857142858e-06, + "loss": 47.3659, + "step": 1103 + }, + { + "epoch": 26.286567164179104, + "grad_norm": 35.413970947265625, + "learning_rate": 8.476190476190477e-06, + "loss": 46.6459, + "step": 1104 + }, + { + "epoch": 26.31044776119403, + "grad_norm": 27.774656295776367, + "learning_rate": 8.470238095238095e-06, + "loss": 47.4369, + "step": 1105 + }, + { + "epoch": 26.334328358208957, + "grad_norm": 32.258155822753906, + "learning_rate": 8.464285714285715e-06, + "loss": 47.487, + "step": 1106 + }, + { + "epoch": 26.35820895522388, + "grad_norm": 18.22418785095215, + "learning_rate": 8.458333333333333e-06, + "loss": 46.948, + "step": 1107 + }, + { + "epoch": 26.382089552238806, + "grad_norm": 24.50945472717285, + "learning_rate": 8.452380952380953e-06, + "loss": 45.7291, + "step": 1108 + }, + { + "epoch": 26.405970149253733, + "grad_norm": 23.20486831665039, + "learning_rate": 8.446428571428571e-06, + "loss": 46.8704, + "step": 1109 + }, + { + "epoch": 26.429850746268656, + "grad_norm": 20.810514450073242, + "learning_rate": 8.440476190476191e-06, + "loss": 47.3419, + "step": 1110 + }, + { + "epoch": 26.453731343283582, + "grad_norm": 27.68440818786621, + "learning_rate": 8.434523809523811e-06, + "loss": 47.0641, + "step": 1111 + }, + { + "epoch": 26.47761194029851, + "grad_norm": 26.989046096801758, + "learning_rate": 8.428571428571429e-06, + "loss": 46.9324, + "step": 1112 + }, + { + "epoch": 26.501492537313432, + "grad_norm": 25.986888885498047, + "learning_rate": 8.422619047619049e-06, + "loss": 48.6179, + "step": 1113 + }, + { + "epoch": 26.52537313432836, + "grad_norm": 28.111356735229492, + "learning_rate": 8.416666666666667e-06, + "loss": 46.4251, + "step": 1114 + }, + { + "epoch": 26.549253731343285, + "grad_norm": 24.306228637695312, + "learning_rate": 8.410714285714287e-06, + "loss": 46.4379, + "step": 1115 + }, + { + "epoch": 26.573134328358208, + "grad_norm": 23.894895553588867, + "learning_rate": 8.404761904761905e-06, + "loss": 46.665, + "step": 1116 + }, + { + "epoch": 26.597014925373134, + "grad_norm": 25.917400360107422, + "learning_rate": 8.398809523809525e-06, + "loss": 46.6619, + "step": 1117 + }, + { + "epoch": 26.62089552238806, + "grad_norm": 21.423585891723633, + "learning_rate": 8.392857142857144e-06, + "loss": 46.3447, + "step": 1118 + }, + { + "epoch": 26.644776119402984, + "grad_norm": 29.13437271118164, + "learning_rate": 8.386904761904762e-06, + "loss": 46.4292, + "step": 1119 + }, + { + "epoch": 26.66865671641791, + "grad_norm": 25.711469650268555, + "learning_rate": 8.380952380952382e-06, + "loss": 46.6156, + "step": 1120 + }, + { + "epoch": 26.692537313432837, + "grad_norm": 26.55695915222168, + "learning_rate": 8.375e-06, + "loss": 46.7429, + "step": 1121 + }, + { + "epoch": 26.71641791044776, + "grad_norm": 27.66262435913086, + "learning_rate": 8.36904761904762e-06, + "loss": 47.1275, + "step": 1122 + }, + { + "epoch": 26.740298507462686, + "grad_norm": 33.85395050048828, + "learning_rate": 8.36309523809524e-06, + "loss": 46.7244, + "step": 1123 + }, + { + "epoch": 26.764179104477613, + "grad_norm": 29.51833152770996, + "learning_rate": 8.357142857142858e-06, + "loss": 47.2072, + "step": 1124 + }, + { + "epoch": 26.788059701492536, + "grad_norm": 26.21416664123535, + "learning_rate": 8.351190476190478e-06, + "loss": 47.8298, + "step": 1125 + }, + { + "epoch": 26.811940298507462, + "grad_norm": 31.24039649963379, + "learning_rate": 8.345238095238096e-06, + "loss": 46.8069, + "step": 1126 + }, + { + "epoch": 26.83582089552239, + "grad_norm": 32.19520568847656, + "learning_rate": 8.339285714285716e-06, + "loss": 47.1494, + "step": 1127 + }, + { + "epoch": 26.85970149253731, + "grad_norm": 29.194063186645508, + "learning_rate": 8.333333333333334e-06, + "loss": 46.1827, + "step": 1128 + }, + { + "epoch": 26.883582089552238, + "grad_norm": 28.723541259765625, + "learning_rate": 8.327380952380954e-06, + "loss": 46.8513, + "step": 1129 + }, + { + "epoch": 26.907462686567165, + "grad_norm": 26.91135597229004, + "learning_rate": 8.321428571428573e-06, + "loss": 46.561, + "step": 1130 + }, + { + "epoch": 26.93134328358209, + "grad_norm": 25.60898208618164, + "learning_rate": 8.315476190476191e-06, + "loss": 46.4706, + "step": 1131 + }, + { + "epoch": 26.955223880597014, + "grad_norm": 23.72539520263672, + "learning_rate": 8.309523809523811e-06, + "loss": 46.7871, + "step": 1132 + }, + { + "epoch": 26.97910447761194, + "grad_norm": 25.030731201171875, + "learning_rate": 8.30357142857143e-06, + "loss": 46.2433, + "step": 1133 + }, + { + "epoch": 27.0, + "grad_norm": 25.439281463623047, + "learning_rate": 8.297619047619049e-06, + "loss": 40.9101, + "step": 1134 + }, + { + "epoch": 27.023880597014927, + "grad_norm": 29.52981948852539, + "learning_rate": 8.291666666666667e-06, + "loss": 46.9385, + "step": 1135 + }, + { + "epoch": 27.04776119402985, + "grad_norm": 22.007299423217773, + "learning_rate": 8.285714285714287e-06, + "loss": 47.2751, + "step": 1136 + }, + { + "epoch": 27.071641791044776, + "grad_norm": 31.965675354003906, + "learning_rate": 8.279761904761905e-06, + "loss": 47.7763, + "step": 1137 + }, + { + "epoch": 27.095522388059702, + "grad_norm": 23.38637351989746, + "learning_rate": 8.273809523809523e-06, + "loss": 46.0412, + "step": 1138 + }, + { + "epoch": 27.119402985074625, + "grad_norm": 24.295711517333984, + "learning_rate": 8.267857142857143e-06, + "loss": 46.8552, + "step": 1139 + }, + { + "epoch": 27.143283582089552, + "grad_norm": 20.915624618530273, + "learning_rate": 8.261904761904763e-06, + "loss": 46.3041, + "step": 1140 + }, + { + "epoch": 27.16716417910448, + "grad_norm": 28.25569725036621, + "learning_rate": 8.25595238095238e-06, + "loss": 45.8736, + "step": 1141 + }, + { + "epoch": 27.1910447761194, + "grad_norm": 24.8399658203125, + "learning_rate": 8.25e-06, + "loss": 46.0629, + "step": 1142 + }, + { + "epoch": 27.214925373134328, + "grad_norm": 21.237272262573242, + "learning_rate": 8.244047619047619e-06, + "loss": 47.1252, + "step": 1143 + }, + { + "epoch": 27.238805970149254, + "grad_norm": 24.35887336730957, + "learning_rate": 8.238095238095239e-06, + "loss": 46.9687, + "step": 1144 + }, + { + "epoch": 27.262686567164177, + "grad_norm": 26.583545684814453, + "learning_rate": 8.232142857142857e-06, + "loss": 47.1486, + "step": 1145 + }, + { + "epoch": 27.286567164179104, + "grad_norm": 23.712989807128906, + "learning_rate": 8.226190476190476e-06, + "loss": 46.4769, + "step": 1146 + }, + { + "epoch": 27.31044776119403, + "grad_norm": 18.466094970703125, + "learning_rate": 8.220238095238096e-06, + "loss": 46.6911, + "step": 1147 + }, + { + "epoch": 27.334328358208957, + "grad_norm": 31.812236785888672, + "learning_rate": 8.214285714285714e-06, + "loss": 47.6261, + "step": 1148 + }, + { + "epoch": 27.35820895522388, + "grad_norm": 23.437780380249023, + "learning_rate": 8.208333333333334e-06, + "loss": 46.1776, + "step": 1149 + }, + { + "epoch": 27.382089552238806, + "grad_norm": 27.252187728881836, + "learning_rate": 8.202380952380952e-06, + "loss": 44.8499, + "step": 1150 + }, + { + "epoch": 27.405970149253733, + "grad_norm": 24.425500869750977, + "learning_rate": 8.196428571428572e-06, + "loss": 46.5397, + "step": 1151 + }, + { + "epoch": 27.429850746268656, + "grad_norm": 28.237712860107422, + "learning_rate": 8.190476190476192e-06, + "loss": 46.8922, + "step": 1152 + }, + { + "epoch": 27.453731343283582, + "grad_norm": 23.262300491333008, + "learning_rate": 8.18452380952381e-06, + "loss": 47.4204, + "step": 1153 + }, + { + "epoch": 27.47761194029851, + "grad_norm": 20.69318389892578, + "learning_rate": 8.17857142857143e-06, + "loss": 47.1507, + "step": 1154 + }, + { + "epoch": 27.501492537313432, + "grad_norm": 30.182701110839844, + "learning_rate": 8.172619047619048e-06, + "loss": 46.198, + "step": 1155 + }, + { + "epoch": 27.52537313432836, + "grad_norm": 28.804855346679688, + "learning_rate": 8.166666666666668e-06, + "loss": 46.1366, + "step": 1156 + }, + { + "epoch": 27.549253731343285, + "grad_norm": 26.992097854614258, + "learning_rate": 8.160714285714286e-06, + "loss": 47.3639, + "step": 1157 + }, + { + "epoch": 27.573134328358208, + "grad_norm": 27.22978401184082, + "learning_rate": 8.154761904761905e-06, + "loss": 46.7295, + "step": 1158 + }, + { + "epoch": 27.597014925373134, + "grad_norm": 24.036380767822266, + "learning_rate": 8.148809523809525e-06, + "loss": 46.7824, + "step": 1159 + }, + { + "epoch": 27.62089552238806, + "grad_norm": 28.67648696899414, + "learning_rate": 8.142857142857143e-06, + "loss": 46.9712, + "step": 1160 + }, + { + "epoch": 27.644776119402984, + "grad_norm": 27.389991760253906, + "learning_rate": 8.136904761904763e-06, + "loss": 47.6263, + "step": 1161 + }, + { + "epoch": 27.66865671641791, + "grad_norm": 28.887022018432617, + "learning_rate": 8.130952380952381e-06, + "loss": 47.7553, + "step": 1162 + }, + { + "epoch": 27.692537313432837, + "grad_norm": 17.354753494262695, + "learning_rate": 8.125000000000001e-06, + "loss": 46.1359, + "step": 1163 + }, + { + "epoch": 27.71641791044776, + "grad_norm": 26.72220230102539, + "learning_rate": 8.119047619047619e-06, + "loss": 46.3653, + "step": 1164 + }, + { + "epoch": 27.740298507462686, + "grad_norm": 22.09579849243164, + "learning_rate": 8.113095238095239e-06, + "loss": 46.856, + "step": 1165 + }, + { + "epoch": 27.764179104477613, + "grad_norm": 30.197189331054688, + "learning_rate": 8.107142857142859e-06, + "loss": 46.4435, + "step": 1166 + }, + { + "epoch": 27.788059701492536, + "grad_norm": 28.597610473632812, + "learning_rate": 8.101190476190477e-06, + "loss": 48.3097, + "step": 1167 + }, + { + "epoch": 27.811940298507462, + "grad_norm": 22.391801834106445, + "learning_rate": 8.095238095238097e-06, + "loss": 47.2598, + "step": 1168 + }, + { + "epoch": 27.83582089552239, + "grad_norm": 28.523584365844727, + "learning_rate": 8.089285714285715e-06, + "loss": 46.3123, + "step": 1169 + }, + { + "epoch": 27.85970149253731, + "grad_norm": 21.646997451782227, + "learning_rate": 8.083333333333334e-06, + "loss": 46.2853, + "step": 1170 + }, + { + "epoch": 27.883582089552238, + "grad_norm": 22.68369483947754, + "learning_rate": 8.077380952380953e-06, + "loss": 46.1355, + "step": 1171 + }, + { + "epoch": 27.907462686567165, + "grad_norm": 25.581987380981445, + "learning_rate": 8.071428571428572e-06, + "loss": 45.5431, + "step": 1172 + }, + { + "epoch": 27.93134328358209, + "grad_norm": 26.512523651123047, + "learning_rate": 8.065476190476192e-06, + "loss": 47.1898, + "step": 1173 + }, + { + "epoch": 27.955223880597014, + "grad_norm": 26.89809226989746, + "learning_rate": 8.05952380952381e-06, + "loss": 46.8108, + "step": 1174 + }, + { + "epoch": 27.97910447761194, + "grad_norm": 23.638704299926758, + "learning_rate": 8.05357142857143e-06, + "loss": 47.1454, + "step": 1175 + }, + { + "epoch": 28.0, + "grad_norm": 28.61042022705078, + "learning_rate": 8.047619047619048e-06, + "loss": 41.7161, + "step": 1176 + }, + { + "epoch": 28.023880597014927, + "grad_norm": 31.2153377532959, + "learning_rate": 8.041666666666668e-06, + "loss": 46.8925, + "step": 1177 + }, + { + "epoch": 28.04776119402985, + "grad_norm": 26.284482955932617, + "learning_rate": 8.035714285714286e-06, + "loss": 45.7065, + "step": 1178 + }, + { + "epoch": 28.071641791044776, + "grad_norm": 30.96581268310547, + "learning_rate": 8.029761904761906e-06, + "loss": 46.5612, + "step": 1179 + }, + { + "epoch": 28.095522388059702, + "grad_norm": 24.686336517333984, + "learning_rate": 8.023809523809526e-06, + "loss": 46.4376, + "step": 1180 + }, + { + "epoch": 28.119402985074625, + "grad_norm": 27.783416748046875, + "learning_rate": 8.017857142857144e-06, + "loss": 47.247, + "step": 1181 + }, + { + "epoch": 28.143283582089552, + "grad_norm": 33.3108024597168, + "learning_rate": 8.011904761904763e-06, + "loss": 47.3171, + "step": 1182 + }, + { + "epoch": 28.16716417910448, + "grad_norm": 30.010536193847656, + "learning_rate": 8.005952380952382e-06, + "loss": 45.9761, + "step": 1183 + }, + { + "epoch": 28.1910447761194, + "grad_norm": 29.399965286254883, + "learning_rate": 8.000000000000001e-06, + "loss": 47.3345, + "step": 1184 + }, + { + "epoch": 28.214925373134328, + "grad_norm": 25.835142135620117, + "learning_rate": 7.99404761904762e-06, + "loss": 46.3395, + "step": 1185 + }, + { + "epoch": 28.238805970149254, + "grad_norm": 26.06570053100586, + "learning_rate": 7.98809523809524e-06, + "loss": 45.842, + "step": 1186 + }, + { + "epoch": 28.262686567164177, + "grad_norm": 28.64603042602539, + "learning_rate": 7.982142857142859e-06, + "loss": 46.4802, + "step": 1187 + }, + { + "epoch": 28.286567164179104, + "grad_norm": 27.157583236694336, + "learning_rate": 7.976190476190477e-06, + "loss": 45.7376, + "step": 1188 + }, + { + "epoch": 28.31044776119403, + "grad_norm": 22.073328018188477, + "learning_rate": 7.970238095238097e-06, + "loss": 47.0787, + "step": 1189 + }, + { + "epoch": 28.334328358208957, + "grad_norm": 21.545568466186523, + "learning_rate": 7.964285714285715e-06, + "loss": 45.6691, + "step": 1190 + }, + { + "epoch": 28.35820895522388, + "grad_norm": 26.17327880859375, + "learning_rate": 7.958333333333333e-06, + "loss": 46.2058, + "step": 1191 + }, + { + "epoch": 28.382089552238806, + "grad_norm": 24.443920135498047, + "learning_rate": 7.952380952380953e-06, + "loss": 45.3531, + "step": 1192 + }, + { + "epoch": 28.405970149253733, + "grad_norm": 27.207778930664062, + "learning_rate": 7.946428571428571e-06, + "loss": 46.5519, + "step": 1193 + }, + { + "epoch": 28.429850746268656, + "grad_norm": 23.15156364440918, + "learning_rate": 7.94047619047619e-06, + "loss": 46.9284, + "step": 1194 + }, + { + "epoch": 28.453731343283582, + "grad_norm": 27.96567153930664, + "learning_rate": 7.93452380952381e-06, + "loss": 46.3696, + "step": 1195 + }, + { + "epoch": 28.47761194029851, + "grad_norm": 25.828689575195312, + "learning_rate": 7.928571428571429e-06, + "loss": 46.2933, + "step": 1196 + }, + { + "epoch": 28.501492537313432, + "grad_norm": 30.69676971435547, + "learning_rate": 7.922619047619048e-06, + "loss": 46.7471, + "step": 1197 + }, + { + "epoch": 28.52537313432836, + "grad_norm": 24.977018356323242, + "learning_rate": 7.916666666666667e-06, + "loss": 47.045, + "step": 1198 + }, + { + "epoch": 28.549253731343285, + "grad_norm": 26.286821365356445, + "learning_rate": 7.910714285714286e-06, + "loss": 47.0185, + "step": 1199 + }, + { + "epoch": 28.573134328358208, + "grad_norm": 25.324783325195312, + "learning_rate": 7.904761904761904e-06, + "loss": 46.4036, + "step": 1200 + }, + { + "epoch": 28.597014925373134, + "grad_norm": 34.25847625732422, + "learning_rate": 7.898809523809524e-06, + "loss": 46.8307, + "step": 1201 + }, + { + "epoch": 28.62089552238806, + "grad_norm": 24.739521026611328, + "learning_rate": 7.892857142857144e-06, + "loss": 46.515, + "step": 1202 + }, + { + "epoch": 28.644776119402984, + "grad_norm": 36.694252014160156, + "learning_rate": 7.886904761904762e-06, + "loss": 45.7078, + "step": 1203 + }, + { + "epoch": 28.66865671641791, + "grad_norm": 34.95314025878906, + "learning_rate": 7.880952380952382e-06, + "loss": 47.6302, + "step": 1204 + }, + { + "epoch": 28.692537313432837, + "grad_norm": 26.474821090698242, + "learning_rate": 7.875e-06, + "loss": 47.2158, + "step": 1205 + }, + { + "epoch": 28.71641791044776, + "grad_norm": 30.19892692565918, + "learning_rate": 7.86904761904762e-06, + "loss": 46.1515, + "step": 1206 + }, + { + "epoch": 28.740298507462686, + "grad_norm": 32.16860580444336, + "learning_rate": 7.863095238095238e-06, + "loss": 46.3963, + "step": 1207 + }, + { + "epoch": 28.764179104477613, + "grad_norm": 27.323444366455078, + "learning_rate": 7.857142857142858e-06, + "loss": 47.6511, + "step": 1208 + }, + { + "epoch": 28.788059701492536, + "grad_norm": 24.073701858520508, + "learning_rate": 7.851190476190477e-06, + "loss": 46.7283, + "step": 1209 + }, + { + "epoch": 28.811940298507462, + "grad_norm": 25.73206329345703, + "learning_rate": 7.845238095238096e-06, + "loss": 47.3957, + "step": 1210 + }, + { + "epoch": 28.83582089552239, + "grad_norm": 23.368709564208984, + "learning_rate": 7.839285714285715e-06, + "loss": 46.1844, + "step": 1211 + }, + { + "epoch": 28.85970149253731, + "grad_norm": 24.563371658325195, + "learning_rate": 7.833333333333333e-06, + "loss": 47.0752, + "step": 1212 + }, + { + "epoch": 28.883582089552238, + "grad_norm": 20.747081756591797, + "learning_rate": 7.827380952380953e-06, + "loss": 45.9425, + "step": 1213 + }, + { + "epoch": 28.907462686567165, + "grad_norm": 22.27573013305664, + "learning_rate": 7.821428571428571e-06, + "loss": 46.746, + "step": 1214 + }, + { + "epoch": 28.93134328358209, + "grad_norm": 23.162179946899414, + "learning_rate": 7.815476190476191e-06, + "loss": 46.808, + "step": 1215 + }, + { + "epoch": 28.955223880597014, + "grad_norm": 23.585325241088867, + "learning_rate": 7.809523809523811e-06, + "loss": 47.2584, + "step": 1216 + }, + { + "epoch": 28.97910447761194, + "grad_norm": 29.979564666748047, + "learning_rate": 7.803571428571429e-06, + "loss": 47.3433, + "step": 1217 + }, + { + "epoch": 29.0, + "grad_norm": 25.872072219848633, + "learning_rate": 7.797619047619049e-06, + "loss": 41.1844, + "step": 1218 + }, + { + "epoch": 29.023880597014927, + "grad_norm": 25.673351287841797, + "learning_rate": 7.791666666666667e-06, + "loss": 46.4546, + "step": 1219 + }, + { + "epoch": 29.04776119402985, + "grad_norm": 29.831058502197266, + "learning_rate": 7.785714285714287e-06, + "loss": 46.8122, + "step": 1220 + }, + { + "epoch": 29.071641791044776, + "grad_norm": 27.548316955566406, + "learning_rate": 7.779761904761905e-06, + "loss": 47.748, + "step": 1221 + }, + { + "epoch": 29.095522388059702, + "grad_norm": 26.399370193481445, + "learning_rate": 7.773809523809525e-06, + "loss": 46.736, + "step": 1222 + }, + { + "epoch": 29.119402985074625, + "grad_norm": 25.127031326293945, + "learning_rate": 7.767857142857144e-06, + "loss": 46.8307, + "step": 1223 + }, + { + "epoch": 29.143283582089552, + "grad_norm": 26.624732971191406, + "learning_rate": 7.761904761904762e-06, + "loss": 46.2401, + "step": 1224 + }, + { + "epoch": 29.16716417910448, + "grad_norm": 30.770824432373047, + "learning_rate": 7.755952380952382e-06, + "loss": 46.7194, + "step": 1225 + }, + { + "epoch": 29.1910447761194, + "grad_norm": 23.830007553100586, + "learning_rate": 7.75e-06, + "loss": 46.4737, + "step": 1226 + }, + { + "epoch": 29.214925373134328, + "grad_norm": 32.90129470825195, + "learning_rate": 7.74404761904762e-06, + "loss": 47.4361, + "step": 1227 + }, + { + "epoch": 29.238805970149254, + "grad_norm": 23.381397247314453, + "learning_rate": 7.738095238095238e-06, + "loss": 45.3297, + "step": 1228 + }, + { + "epoch": 29.262686567164177, + "grad_norm": 32.836387634277344, + "learning_rate": 7.732142857142858e-06, + "loss": 46.0574, + "step": 1229 + }, + { + "epoch": 29.286567164179104, + "grad_norm": 25.803264617919922, + "learning_rate": 7.726190476190478e-06, + "loss": 45.757, + "step": 1230 + }, + { + "epoch": 29.31044776119403, + "grad_norm": 29.38982391357422, + "learning_rate": 7.720238095238096e-06, + "loss": 46.7099, + "step": 1231 + }, + { + "epoch": 29.334328358208957, + "grad_norm": 26.39947509765625, + "learning_rate": 7.714285714285716e-06, + "loss": 47.5944, + "step": 1232 + }, + { + "epoch": 29.35820895522388, + "grad_norm": 25.958354949951172, + "learning_rate": 7.708333333333334e-06, + "loss": 46.1395, + "step": 1233 + }, + { + "epoch": 29.382089552238806, + "grad_norm": 28.697542190551758, + "learning_rate": 7.702380952380954e-06, + "loss": 46.1713, + "step": 1234 + }, + { + "epoch": 29.405970149253733, + "grad_norm": 19.471586227416992, + "learning_rate": 7.696428571428572e-06, + "loss": 47.7724, + "step": 1235 + }, + { + "epoch": 29.429850746268656, + "grad_norm": 29.924991607666016, + "learning_rate": 7.690476190476191e-06, + "loss": 47.2203, + "step": 1236 + }, + { + "epoch": 29.453731343283582, + "grad_norm": 20.516891479492188, + "learning_rate": 7.684523809523811e-06, + "loss": 46.2945, + "step": 1237 + }, + { + "epoch": 29.47761194029851, + "grad_norm": 30.605262756347656, + "learning_rate": 7.67857142857143e-06, + "loss": 47.1786, + "step": 1238 + }, + { + "epoch": 29.501492537313432, + "grad_norm": 16.288013458251953, + "learning_rate": 7.672619047619049e-06, + "loss": 45.3413, + "step": 1239 + }, + { + "epoch": 29.52537313432836, + "grad_norm": 23.54091453552246, + "learning_rate": 7.666666666666667e-06, + "loss": 45.9196, + "step": 1240 + }, + { + "epoch": 29.549253731343285, + "grad_norm": 20.33724021911621, + "learning_rate": 7.660714285714287e-06, + "loss": 47.0275, + "step": 1241 + }, + { + "epoch": 29.573134328358208, + "grad_norm": 27.460975646972656, + "learning_rate": 7.654761904761905e-06, + "loss": 44.8995, + "step": 1242 + }, + { + "epoch": 29.597014925373134, + "grad_norm": 25.58623695373535, + "learning_rate": 7.648809523809523e-06, + "loss": 46.0706, + "step": 1243 + }, + { + "epoch": 29.62089552238806, + "grad_norm": 27.997203826904297, + "learning_rate": 7.642857142857143e-06, + "loss": 47.2368, + "step": 1244 + }, + { + "epoch": 29.644776119402984, + "grad_norm": 31.361181259155273, + "learning_rate": 7.636904761904763e-06, + "loss": 46.8056, + "step": 1245 + }, + { + "epoch": 29.66865671641791, + "grad_norm": 29.266433715820312, + "learning_rate": 7.630952380952381e-06, + "loss": 45.323, + "step": 1246 + }, + { + "epoch": 29.692537313432837, + "grad_norm": 24.066415786743164, + "learning_rate": 7.625e-06, + "loss": 46.9221, + "step": 1247 + }, + { + "epoch": 29.71641791044776, + "grad_norm": 25.790491104125977, + "learning_rate": 7.61904761904762e-06, + "loss": 45.7051, + "step": 1248 + }, + { + "epoch": 29.740298507462686, + "grad_norm": 24.202716827392578, + "learning_rate": 7.6130952380952386e-06, + "loss": 47.1067, + "step": 1249 + }, + { + "epoch": 29.764179104477613, + "grad_norm": 27.302003860473633, + "learning_rate": 7.6071428571428575e-06, + "loss": 46.6039, + "step": 1250 + }, + { + "epoch": 29.788059701492536, + "grad_norm": 22.75196075439453, + "learning_rate": 7.6011904761904765e-06, + "loss": 46.3265, + "step": 1251 + }, + { + "epoch": 29.811940298507462, + "grad_norm": 30.963153839111328, + "learning_rate": 7.595238095238095e-06, + "loss": 46.3283, + "step": 1252 + }, + { + "epoch": 29.83582089552239, + "grad_norm": 21.538162231445312, + "learning_rate": 7.589285714285714e-06, + "loss": 46.5345, + "step": 1253 + }, + { + "epoch": 29.85970149253731, + "grad_norm": 28.09955596923828, + "learning_rate": 7.583333333333333e-06, + "loss": 47.1, + "step": 1254 + }, + { + "epoch": 29.883582089552238, + "grad_norm": 20.88216781616211, + "learning_rate": 7.577380952380953e-06, + "loss": 45.3354, + "step": 1255 + }, + { + "epoch": 29.907462686567165, + "grad_norm": 24.15240478515625, + "learning_rate": 7.571428571428572e-06, + "loss": 46.299, + "step": 1256 + }, + { + "epoch": 29.93134328358209, + "grad_norm": 22.839298248291016, + "learning_rate": 7.565476190476191e-06, + "loss": 46.3436, + "step": 1257 + }, + { + "epoch": 29.955223880597014, + "grad_norm": 26.582752227783203, + "learning_rate": 7.55952380952381e-06, + "loss": 45.9107, + "step": 1258 + }, + { + "epoch": 29.97910447761194, + "grad_norm": 24.98562240600586, + "learning_rate": 7.553571428571429e-06, + "loss": 46.7134, + "step": 1259 + }, + { + "epoch": 30.0, + "grad_norm": 23.327436447143555, + "learning_rate": 7.547619047619048e-06, + "loss": 41.2325, + "step": 1260 + }, + { + "epoch": 30.023880597014927, + "grad_norm": 20.400623321533203, + "learning_rate": 7.541666666666667e-06, + "loss": 46.9564, + "step": 1261 + }, + { + "epoch": 30.04776119402985, + "grad_norm": NaN, + "learning_rate": 7.5357142857142865e-06, + "loss": 68.2215, + "step": 1262 + }, + { + "epoch": 30.071641791044776, + "grad_norm": 22.870811462402344, + "learning_rate": 7.5357142857142865e-06, + "loss": 46.5712, + "step": 1263 + }, + { + "epoch": 30.095522388059702, + "grad_norm": 24.057098388671875, + "learning_rate": 7.5297619047619055e-06, + "loss": 46.8943, + "step": 1264 + }, + { + "epoch": 30.119402985074625, + "grad_norm": 25.820720672607422, + "learning_rate": 7.523809523809524e-06, + "loss": 46.0747, + "step": 1265 + }, + { + "epoch": 30.143283582089552, + "grad_norm": 28.460693359375, + "learning_rate": 7.517857142857143e-06, + "loss": 47.0127, + "step": 1266 + }, + { + "epoch": 30.16716417910448, + "grad_norm": 21.60432243347168, + "learning_rate": 7.511904761904762e-06, + "loss": 45.8081, + "step": 1267 + }, + { + "epoch": 30.1910447761194, + "grad_norm": 29.013648986816406, + "learning_rate": 7.505952380952381e-06, + "loss": 46.6712, + "step": 1268 + }, + { + "epoch": 30.214925373134328, + "grad_norm": 24.865493774414062, + "learning_rate": 7.500000000000001e-06, + "loss": 46.6816, + "step": 1269 + }, + { + "epoch": 30.238805970149254, + "grad_norm": 23.676206588745117, + "learning_rate": 7.49404761904762e-06, + "loss": 46.2663, + "step": 1270 + }, + { + "epoch": 30.262686567164177, + "grad_norm": 27.889135360717773, + "learning_rate": 7.488095238095239e-06, + "loss": 45.7052, + "step": 1271 + }, + { + "epoch": 30.286567164179104, + "grad_norm": 29.024211883544922, + "learning_rate": 7.482142857142858e-06, + "loss": 45.5005, + "step": 1272 + }, + { + "epoch": 30.31044776119403, + "grad_norm": 25.8428955078125, + "learning_rate": 7.476190476190477e-06, + "loss": 46.788, + "step": 1273 + }, + { + "epoch": 30.334328358208957, + "grad_norm": 26.765539169311523, + "learning_rate": 7.470238095238096e-06, + "loss": 46.454, + "step": 1274 + }, + { + "epoch": 30.35820895522388, + "grad_norm": 34.80079650878906, + "learning_rate": 7.464285714285715e-06, + "loss": 47.6929, + "step": 1275 + }, + { + "epoch": 30.382089552238806, + "grad_norm": 25.589618682861328, + "learning_rate": 7.4583333333333345e-06, + "loss": 46.0104, + "step": 1276 + }, + { + "epoch": 30.405970149253733, + "grad_norm": 27.0733699798584, + "learning_rate": 7.4523809523809534e-06, + "loss": 45.4742, + "step": 1277 + }, + { + "epoch": 30.429850746268656, + "grad_norm": 26.662338256835938, + "learning_rate": 7.446428571428572e-06, + "loss": 46.8066, + "step": 1278 + }, + { + "epoch": 30.453731343283582, + "grad_norm": 28.389951705932617, + "learning_rate": 7.440476190476191e-06, + "loss": 46.9716, + "step": 1279 + }, + { + "epoch": 30.47761194029851, + "grad_norm": NaN, + "learning_rate": 7.43452380952381e-06, + "loss": 52.2915, + "step": 1280 + }, + { + "epoch": 30.501492537313432, + "grad_norm": 26.77708625793457, + "learning_rate": 7.43452380952381e-06, + "loss": 44.919, + "step": 1281 + }, + { + "epoch": 30.52537313432836, + "grad_norm": 25.423444747924805, + "learning_rate": 7.428571428571429e-06, + "loss": 46.5057, + "step": 1282 + }, + { + "epoch": 30.549253731343285, + "grad_norm": 24.04167366027832, + "learning_rate": 7.422619047619048e-06, + "loss": 46.3685, + "step": 1283 + }, + { + "epoch": 30.573134328358208, + "grad_norm": 23.51607894897461, + "learning_rate": 7.416666666666668e-06, + "loss": 45.9694, + "step": 1284 + }, + { + "epoch": 30.597014925373134, + "grad_norm": 26.216157913208008, + "learning_rate": 7.410714285714287e-06, + "loss": 47.2582, + "step": 1285 + }, + { + "epoch": 30.62089552238806, + "grad_norm": 24.339780807495117, + "learning_rate": 7.404761904761906e-06, + "loss": 44.8052, + "step": 1286 + }, + { + "epoch": 30.644776119402984, + "grad_norm": 19.203577041625977, + "learning_rate": 7.398809523809525e-06, + "loss": 47.0301, + "step": 1287 + }, + { + "epoch": 30.66865671641791, + "grad_norm": 22.252805709838867, + "learning_rate": 7.392857142857144e-06, + "loss": 45.5993, + "step": 1288 + }, + { + "epoch": 30.692537313432837, + "grad_norm": 25.316205978393555, + "learning_rate": 7.386904761904763e-06, + "loss": 46.1157, + "step": 1289 + }, + { + "epoch": 30.71641791044776, + "grad_norm": 18.311643600463867, + "learning_rate": 7.380952380952382e-06, + "loss": 46.5986, + "step": 1290 + }, + { + "epoch": 30.740298507462686, + "grad_norm": 31.84505271911621, + "learning_rate": 7.375000000000001e-06, + "loss": 46.9177, + "step": 1291 + }, + { + "epoch": 30.764179104477613, + "grad_norm": 26.221525192260742, + "learning_rate": 7.36904761904762e-06, + "loss": 47.0897, + "step": 1292 + }, + { + "epoch": 30.788059701492536, + "grad_norm": 27.029104232788086, + "learning_rate": 7.363095238095239e-06, + "loss": 45.3724, + "step": 1293 + }, + { + "epoch": 30.811940298507462, + "grad_norm": 33.51012420654297, + "learning_rate": 7.357142857142858e-06, + "loss": 46.7046, + "step": 1294 + }, + { + "epoch": 30.83582089552239, + "grad_norm": 26.42972183227539, + "learning_rate": 7.351190476190477e-06, + "loss": 46.7606, + "step": 1295 + }, + { + "epoch": 30.85970149253731, + "grad_norm": 30.91115951538086, + "learning_rate": 7.345238095238096e-06, + "loss": 47.5485, + "step": 1296 + }, + { + "epoch": 30.883582089552238, + "grad_norm": 28.296560287475586, + "learning_rate": 7.339285714285714e-06, + "loss": 46.4997, + "step": 1297 + }, + { + "epoch": 30.907462686567165, + "grad_norm": 32.054561614990234, + "learning_rate": 7.333333333333333e-06, + "loss": 46.4953, + "step": 1298 + }, + { + "epoch": 30.93134328358209, + "grad_norm": 31.635595321655273, + "learning_rate": 7.327380952380952e-06, + "loss": 46.5325, + "step": 1299 + }, + { + "epoch": 30.955223880597014, + "grad_norm": 25.557523727416992, + "learning_rate": 7.321428571428572e-06, + "loss": 45.56, + "step": 1300 + }, + { + "epoch": 30.97910447761194, + "grad_norm": 30.01810073852539, + "learning_rate": 7.315476190476191e-06, + "loss": 46.6149, + "step": 1301 + }, + { + "epoch": 31.0, + "grad_norm": 24.6826114654541, + "learning_rate": 7.30952380952381e-06, + "loss": 40.8651, + "step": 1302 + }, + { + "epoch": 31.023880597014927, + "grad_norm": 24.378164291381836, + "learning_rate": 7.303571428571429e-06, + "loss": 46.0721, + "step": 1303 + }, + { + "epoch": 31.04776119402985, + "grad_norm": 20.247482299804688, + "learning_rate": 7.297619047619048e-06, + "loss": 45.8819, + "step": 1304 + }, + { + "epoch": 31.071641791044776, + "grad_norm": 25.636112213134766, + "learning_rate": 7.291666666666667e-06, + "loss": 47.1987, + "step": 1305 + }, + { + "epoch": 31.095522388059702, + "grad_norm": 30.428096771240234, + "learning_rate": 7.285714285714286e-06, + "loss": 46.6961, + "step": 1306 + }, + { + "epoch": 31.119402985074625, + "grad_norm": 21.404991149902344, + "learning_rate": 7.279761904761905e-06, + "loss": 46.6841, + "step": 1307 + }, + { + "epoch": 31.143283582089552, + "grad_norm": 31.655052185058594, + "learning_rate": 7.273809523809524e-06, + "loss": 47.7781, + "step": 1308 + }, + { + "epoch": 31.16716417910448, + "grad_norm": 24.327327728271484, + "learning_rate": 7.267857142857143e-06, + "loss": 46.002, + "step": 1309 + }, + { + "epoch": 31.1910447761194, + "grad_norm": 26.230745315551758, + "learning_rate": 7.261904761904762e-06, + "loss": 47.3903, + "step": 1310 + }, + { + "epoch": 31.214925373134328, + "grad_norm": 27.337961196899414, + "learning_rate": 7.255952380952381e-06, + "loss": 46.0999, + "step": 1311 + }, + { + "epoch": 31.238805970149254, + "grad_norm": 35.14864730834961, + "learning_rate": 7.25e-06, + "loss": 46.5187, + "step": 1312 + }, + { + "epoch": 31.262686567164177, + "grad_norm": 26.60109519958496, + "learning_rate": 7.24404761904762e-06, + "loss": 44.5864, + "step": 1313 + }, + { + "epoch": 31.286567164179104, + "grad_norm": 33.15165710449219, + "learning_rate": 7.238095238095239e-06, + "loss": 46.4779, + "step": 1314 + }, + { + "epoch": 31.31044776119403, + "grad_norm": 26.3510684967041, + "learning_rate": 7.232142857142858e-06, + "loss": 47.0845, + "step": 1315 + }, + { + "epoch": 31.334328358208957, + "grad_norm": 47.12569046020508, + "learning_rate": 7.226190476190477e-06, + "loss": 47.2947, + "step": 1316 + }, + { + "epoch": 31.35820895522388, + "grad_norm": 40.15263748168945, + "learning_rate": 7.220238095238096e-06, + "loss": 45.8788, + "step": 1317 + }, + { + "epoch": 31.382089552238806, + "grad_norm": 36.59072494506836, + "learning_rate": 7.2142857142857145e-06, + "loss": 45.991, + "step": 1318 + }, + { + "epoch": 31.405970149253733, + "grad_norm": 36.895408630371094, + "learning_rate": 7.2083333333333335e-06, + "loss": 46.197, + "step": 1319 + }, + { + "epoch": 31.429850746268656, + "grad_norm": NaN, + "learning_rate": 7.202380952380953e-06, + "loss": 38.9024, + "step": 1320 + }, + { + "epoch": 31.453731343283582, + "grad_norm": 27.446247100830078, + "learning_rate": 7.202380952380953e-06, + "loss": 45.5293, + "step": 1321 + }, + { + "epoch": 31.47761194029851, + "grad_norm": 27.48939323425293, + "learning_rate": 7.196428571428572e-06, + "loss": 46.8754, + "step": 1322 + }, + { + "epoch": 31.501492537313432, + "grad_norm": 22.736833572387695, + "learning_rate": 7.190476190476191e-06, + "loss": 44.4905, + "step": 1323 + }, + { + "epoch": 31.52537313432836, + "grad_norm": 23.413612365722656, + "learning_rate": 7.18452380952381e-06, + "loss": 47.0714, + "step": 1324 + }, + { + "epoch": 31.549253731343285, + "grad_norm": 29.154848098754883, + "learning_rate": 7.178571428571429e-06, + "loss": 46.393, + "step": 1325 + }, + { + "epoch": 31.573134328358208, + "grad_norm": 28.130638122558594, + "learning_rate": 7.172619047619048e-06, + "loss": 46.1857, + "step": 1326 + }, + { + "epoch": 31.597014925373134, + "grad_norm": 19.745920181274414, + "learning_rate": 7.166666666666667e-06, + "loss": 45.2873, + "step": 1327 + }, + { + "epoch": 31.62089552238806, + "grad_norm": 27.630279541015625, + "learning_rate": 7.160714285714287e-06, + "loss": 46.5475, + "step": 1328 + }, + { + "epoch": 31.644776119402984, + "grad_norm": 20.568862915039062, + "learning_rate": 7.154761904761906e-06, + "loss": 46.231, + "step": 1329 + }, + { + "epoch": 31.66865671641791, + "grad_norm": 17.769695281982422, + "learning_rate": 7.148809523809525e-06, + "loss": 46.8431, + "step": 1330 + }, + { + "epoch": 31.692537313432837, + "grad_norm": 29.941057205200195, + "learning_rate": 7.1428571428571436e-06, + "loss": 44.842, + "step": 1331 + }, + { + "epoch": 31.71641791044776, + "grad_norm": 21.054975509643555, + "learning_rate": 7.1369047619047625e-06, + "loss": 45.1147, + "step": 1332 + }, + { + "epoch": 31.740298507462686, + "grad_norm": 23.80388069152832, + "learning_rate": 7.1309523809523814e-06, + "loss": 46.1839, + "step": 1333 + }, + { + "epoch": 31.764179104477613, + "grad_norm": 30.561933517456055, + "learning_rate": 7.125e-06, + "loss": 46.2703, + "step": 1334 + }, + { + "epoch": 31.788059701492536, + "grad_norm": 23.752151489257812, + "learning_rate": 7.11904761904762e-06, + "loss": 46.7347, + "step": 1335 + }, + { + "epoch": 31.811940298507462, + "grad_norm": 32.00548553466797, + "learning_rate": 7.113095238095239e-06, + "loss": 46.1236, + "step": 1336 + }, + { + "epoch": 31.83582089552239, + "grad_norm": 26.685504913330078, + "learning_rate": 7.107142857142858e-06, + "loss": 47.8881, + "step": 1337 + }, + { + "epoch": 31.85970149253731, + "grad_norm": 26.5799503326416, + "learning_rate": 7.101190476190477e-06, + "loss": 46.1187, + "step": 1338 + }, + { + "epoch": 31.883582089552238, + "grad_norm": 28.78062629699707, + "learning_rate": 7.095238095238096e-06, + "loss": 46.8058, + "step": 1339 + }, + { + "epoch": 31.907462686567165, + "grad_norm": 26.98428726196289, + "learning_rate": 7.089285714285715e-06, + "loss": 46.3602, + "step": 1340 + }, + { + "epoch": 31.93134328358209, + "grad_norm": 32.5291633605957, + "learning_rate": 7.083333333333335e-06, + "loss": 46.464, + "step": 1341 + }, + { + "epoch": 31.955223880597014, + "grad_norm": 25.088685989379883, + "learning_rate": 7.077380952380954e-06, + "loss": 47.0542, + "step": 1342 + }, + { + "epoch": 31.97910447761194, + "grad_norm": 32.58052444458008, + "learning_rate": 7.0714285714285726e-06, + "loss": 46.3364, + "step": 1343 + }, + { + "epoch": 32.0, + "grad_norm": 22.65249252319336, + "learning_rate": 7.0654761904761915e-06, + "loss": 39.8201, + "step": 1344 + }, + { + "epoch": 32.02388059701492, + "grad_norm": 27.03556251525879, + "learning_rate": 7.0595238095238105e-06, + "loss": 47.7819, + "step": 1345 + }, + { + "epoch": 32.04776119402985, + "grad_norm": 25.712047576904297, + "learning_rate": 7.053571428571429e-06, + "loss": 46.1116, + "step": 1346 + }, + { + "epoch": 32.071641791044776, + "grad_norm": 21.99336051940918, + "learning_rate": 7.047619047619048e-06, + "loss": 46.3745, + "step": 1347 + }, + { + "epoch": 32.0955223880597, + "grad_norm": 28.53151512145996, + "learning_rate": 7.041666666666668e-06, + "loss": 46.5998, + "step": 1348 + }, + { + "epoch": 32.11940298507463, + "grad_norm": 20.151912689208984, + "learning_rate": 7.035714285714287e-06, + "loss": 45.4197, + "step": 1349 + }, + { + "epoch": 32.14328358208955, + "grad_norm": 21.491193771362305, + "learning_rate": 7.029761904761905e-06, + "loss": 46.0246, + "step": 1350 + }, + { + "epoch": 32.167164179104475, + "grad_norm": 20.057588577270508, + "learning_rate": 7.023809523809524e-06, + "loss": 46.2149, + "step": 1351 + }, + { + "epoch": 32.191044776119405, + "grad_norm": 16.675336837768555, + "learning_rate": 7.017857142857143e-06, + "loss": 46.5231, + "step": 1352 + }, + { + "epoch": 32.21492537313433, + "grad_norm": 22.007305145263672, + "learning_rate": 7.011904761904762e-06, + "loss": 44.8665, + "step": 1353 + }, + { + "epoch": 32.23880597014925, + "grad_norm": 22.947837829589844, + "learning_rate": 7.005952380952381e-06, + "loss": 45.0394, + "step": 1354 + }, + { + "epoch": 32.26268656716418, + "grad_norm": 25.444522857666016, + "learning_rate": 7e-06, + "loss": 46.0367, + "step": 1355 + }, + { + "epoch": 32.286567164179104, + "grad_norm": 22.319833755493164, + "learning_rate": 6.994047619047619e-06, + "loss": 47.0455, + "step": 1356 + }, + { + "epoch": 32.31044776119403, + "grad_norm": 20.41710090637207, + "learning_rate": 6.988095238095239e-06, + "loss": 45.5119, + "step": 1357 + }, + { + "epoch": 32.33432835820896, + "grad_norm": 29.03120994567871, + "learning_rate": 6.9821428571428576e-06, + "loss": 45.1962, + "step": 1358 + }, + { + "epoch": 32.35820895522388, + "grad_norm": 22.10372543334961, + "learning_rate": 6.9761904761904765e-06, + "loss": 47.379, + "step": 1359 + }, + { + "epoch": 32.3820895522388, + "grad_norm": 29.49492073059082, + "learning_rate": 6.9702380952380955e-06, + "loss": 48.2375, + "step": 1360 + }, + { + "epoch": 32.40597014925373, + "grad_norm": 26.655149459838867, + "learning_rate": 6.964285714285714e-06, + "loss": 45.8468, + "step": 1361 + }, + { + "epoch": 32.429850746268656, + "grad_norm": 27.994979858398438, + "learning_rate": 6.958333333333333e-06, + "loss": 46.4883, + "step": 1362 + }, + { + "epoch": 32.45373134328358, + "grad_norm": 25.787900924682617, + "learning_rate": 6.952380952380952e-06, + "loss": 47.0159, + "step": 1363 + }, + { + "epoch": 32.47761194029851, + "grad_norm": 29.429485321044922, + "learning_rate": 6.946428571428572e-06, + "loss": 45.182, + "step": 1364 + }, + { + "epoch": 32.50149253731343, + "grad_norm": 21.825122833251953, + "learning_rate": 6.940476190476191e-06, + "loss": 47.4224, + "step": 1365 + }, + { + "epoch": 32.525373134328355, + "grad_norm": 26.284622192382812, + "learning_rate": 6.93452380952381e-06, + "loss": 45.7025, + "step": 1366 + }, + { + "epoch": 32.549253731343285, + "grad_norm": 21.384979248046875, + "learning_rate": 6.928571428571429e-06, + "loss": 45.6267, + "step": 1367 + }, + { + "epoch": 32.57313432835821, + "grad_norm": 21.64442253112793, + "learning_rate": 6.922619047619048e-06, + "loss": 46.8577, + "step": 1368 + }, + { + "epoch": 32.59701492537313, + "grad_norm": 22.377302169799805, + "learning_rate": 6.916666666666667e-06, + "loss": 46.5022, + "step": 1369 + }, + { + "epoch": 32.62089552238806, + "grad_norm": 18.1933536529541, + "learning_rate": 6.910714285714286e-06, + "loss": 46.7098, + "step": 1370 + }, + { + "epoch": 32.644776119402984, + "grad_norm": NaN, + "learning_rate": 6.9047619047619055e-06, + "loss": 59.6159, + "step": 1371 + }, + { + "epoch": 32.668656716417914, + "grad_norm": 20.35690689086914, + "learning_rate": 6.9047619047619055e-06, + "loss": 47.4638, + "step": 1372 + }, + { + "epoch": 32.69253731343284, + "grad_norm": 29.140775680541992, + "learning_rate": 6.8988095238095245e-06, + "loss": 46.242, + "step": 1373 + }, + { + "epoch": 32.71641791044776, + "grad_norm": 25.27906608581543, + "learning_rate": 6.892857142857143e-06, + "loss": 45.7122, + "step": 1374 + }, + { + "epoch": 32.74029850746269, + "grad_norm": 19.000076293945312, + "learning_rate": 6.886904761904762e-06, + "loss": 46.4813, + "step": 1375 + }, + { + "epoch": 32.76417910447761, + "grad_norm": 25.048797607421875, + "learning_rate": 6.880952380952381e-06, + "loss": 45.5569, + "step": 1376 + }, + { + "epoch": 32.788059701492536, + "grad_norm": 24.078060150146484, + "learning_rate": 6.875e-06, + "loss": 45.9708, + "step": 1377 + }, + { + "epoch": 32.811940298507466, + "grad_norm": 23.822643280029297, + "learning_rate": 6.86904761904762e-06, + "loss": 47.5914, + "step": 1378 + }, + { + "epoch": 32.83582089552239, + "grad_norm": 29.267864227294922, + "learning_rate": 6.863095238095239e-06, + "loss": 45.2741, + "step": 1379 + }, + { + "epoch": 32.85970149253731, + "grad_norm": 19.477649688720703, + "learning_rate": 6.857142857142858e-06, + "loss": 46.3849, + "step": 1380 + }, + { + "epoch": 32.88358208955224, + "grad_norm": 33.31391525268555, + "learning_rate": 6.851190476190477e-06, + "loss": 44.9609, + "step": 1381 + }, + { + "epoch": 32.907462686567165, + "grad_norm": 23.064956665039062, + "learning_rate": 6.845238095238096e-06, + "loss": 45.8295, + "step": 1382 + }, + { + "epoch": 32.93134328358209, + "grad_norm": 30.366653442382812, + "learning_rate": 6.839285714285715e-06, + "loss": 44.3142, + "step": 1383 + }, + { + "epoch": 32.95522388059702, + "grad_norm": 25.059572219848633, + "learning_rate": 6.833333333333334e-06, + "loss": 46.5768, + "step": 1384 + }, + { + "epoch": 32.97910447761194, + "grad_norm": 23.186697006225586, + "learning_rate": 6.8273809523809535e-06, + "loss": 45.185, + "step": 1385 + }, + { + "epoch": 33.0, + "grad_norm": 21.550168991088867, + "learning_rate": 6.8214285714285724e-06, + "loss": 39.1732, + "step": 1386 + }, + { + "epoch": 33.02388059701492, + "grad_norm": 22.417282104492188, + "learning_rate": 6.815476190476191e-06, + "loss": 47.6667, + "step": 1387 + }, + { + "epoch": 33.04776119402985, + "grad_norm": 26.805702209472656, + "learning_rate": 6.80952380952381e-06, + "loss": 46.4091, + "step": 1388 + }, + { + "epoch": 33.071641791044776, + "grad_norm": 23.723695755004883, + "learning_rate": 6.803571428571429e-06, + "loss": 46.3798, + "step": 1389 + }, + { + "epoch": 33.0955223880597, + "grad_norm": 30.029897689819336, + "learning_rate": 6.797619047619048e-06, + "loss": 45.9736, + "step": 1390 + }, + { + "epoch": 33.11940298507463, + "grad_norm": 19.387653350830078, + "learning_rate": 6.791666666666667e-06, + "loss": 45.1998, + "step": 1391 + }, + { + "epoch": 33.14328358208955, + "grad_norm": 33.68477249145508, + "learning_rate": 6.785714285714287e-06, + "loss": 45.4435, + "step": 1392 + }, + { + "epoch": 33.167164179104475, + "grad_norm": 26.001699447631836, + "learning_rate": 6.779761904761906e-06, + "loss": 45.6725, + "step": 1393 + }, + { + "epoch": 33.191044776119405, + "grad_norm": 34.19535827636719, + "learning_rate": 6.773809523809525e-06, + "loss": 46.6387, + "step": 1394 + }, + { + "epoch": 33.21492537313433, + "grad_norm": 24.243515014648438, + "learning_rate": 6.767857142857144e-06, + "loss": 46.4235, + "step": 1395 + }, + { + "epoch": 33.23880597014925, + "grad_norm": 33.013675689697266, + "learning_rate": 6.761904761904763e-06, + "loss": 46.7151, + "step": 1396 + }, + { + "epoch": 33.26268656716418, + "grad_norm": 30.15135955810547, + "learning_rate": 6.755952380952382e-06, + "loss": 46.3002, + "step": 1397 + }, + { + "epoch": 33.286567164179104, + "grad_norm": 31.58100128173828, + "learning_rate": 6.750000000000001e-06, + "loss": 46.6084, + "step": 1398 + }, + { + "epoch": 33.31044776119403, + "grad_norm": 26.23592185974121, + "learning_rate": 6.74404761904762e-06, + "loss": 45.5745, + "step": 1399 + }, + { + "epoch": 33.33432835820896, + "grad_norm": 32.273311614990234, + "learning_rate": 6.738095238095239e-06, + "loss": 45.1131, + "step": 1400 + }, + { + "epoch": 33.35820895522388, + "grad_norm": 29.7532958984375, + "learning_rate": 6.732142857142858e-06, + "loss": 45.9739, + "step": 1401 + }, + { + "epoch": 33.3820895522388, + "grad_norm": 32.648704528808594, + "learning_rate": 6.726190476190477e-06, + "loss": 46.6293, + "step": 1402 + }, + { + "epoch": 33.40597014925373, + "grad_norm": 26.455778121948242, + "learning_rate": 6.720238095238096e-06, + "loss": 46.5187, + "step": 1403 + }, + { + "epoch": 33.429850746268656, + "grad_norm": 30.5809326171875, + "learning_rate": 6.714285714285714e-06, + "loss": 46.5477, + "step": 1404 + }, + { + "epoch": 33.45373134328358, + "grad_norm": 29.604442596435547, + "learning_rate": 6.708333333333333e-06, + "loss": 45.462, + "step": 1405 + }, + { + "epoch": 33.47761194029851, + "grad_norm": 36.19733428955078, + "learning_rate": 6.702380952380952e-06, + "loss": 46.7046, + "step": 1406 + }, + { + "epoch": 33.50149253731343, + "grad_norm": 37.733619689941406, + "learning_rate": 6.696428571428571e-06, + "loss": 46.2156, + "step": 1407 + }, + { + "epoch": 33.525373134328355, + "grad_norm": 26.49405288696289, + "learning_rate": 6.690476190476191e-06, + "loss": 45.373, + "step": 1408 + }, + { + "epoch": 33.549253731343285, + "grad_norm": 30.09432601928711, + "learning_rate": 6.68452380952381e-06, + "loss": 46.3868, + "step": 1409 + }, + { + "epoch": 33.57313432835821, + "grad_norm": 25.85702896118164, + "learning_rate": 6.678571428571429e-06, + "loss": 45.805, + "step": 1410 + }, + { + "epoch": 33.59701492537313, + "grad_norm": 28.564380645751953, + "learning_rate": 6.672619047619048e-06, + "loss": 46.4158, + "step": 1411 + }, + { + "epoch": 33.62089552238806, + "grad_norm": 19.878551483154297, + "learning_rate": 6.666666666666667e-06, + "loss": 46.5922, + "step": 1412 + }, + { + "epoch": 33.644776119402984, + "grad_norm": 22.83441734313965, + "learning_rate": 6.660714285714286e-06, + "loss": 45.1216, + "step": 1413 + }, + { + "epoch": 33.668656716417914, + "grad_norm": 31.372957229614258, + "learning_rate": 6.654761904761905e-06, + "loss": 47.111, + "step": 1414 + }, + { + "epoch": 33.69253731343284, + "grad_norm": 23.98666763305664, + "learning_rate": 6.648809523809524e-06, + "loss": 47.1762, + "step": 1415 + }, + { + "epoch": 33.71641791044776, + "grad_norm": 27.895401000976562, + "learning_rate": 6.642857142857143e-06, + "loss": 45.6151, + "step": 1416 + }, + { + "epoch": 33.74029850746269, + "grad_norm": 21.776100158691406, + "learning_rate": 6.636904761904762e-06, + "loss": 45.7198, + "step": 1417 + }, + { + "epoch": 33.76417910447761, + "grad_norm": 30.373878479003906, + "learning_rate": 6.630952380952381e-06, + "loss": 45.2212, + "step": 1418 + }, + { + "epoch": 33.788059701492536, + "grad_norm": 26.604324340820312, + "learning_rate": 6.625e-06, + "loss": 45.2001, + "step": 1419 + }, + { + "epoch": 33.811940298507466, + "grad_norm": 29.38104248046875, + "learning_rate": 6.619047619047619e-06, + "loss": 46.711, + "step": 1420 + }, + { + "epoch": 33.83582089552239, + "grad_norm": 24.36806869506836, + "learning_rate": 6.613095238095239e-06, + "loss": 46.3608, + "step": 1421 + }, + { + "epoch": 33.85970149253731, + "grad_norm": 33.40534210205078, + "learning_rate": 6.607142857142858e-06, + "loss": 45.5189, + "step": 1422 + }, + { + "epoch": 33.88358208955224, + "grad_norm": 25.91522789001465, + "learning_rate": 6.601190476190477e-06, + "loss": 47.3604, + "step": 1423 + }, + { + "epoch": 33.907462686567165, + "grad_norm": 25.26549530029297, + "learning_rate": 6.595238095238096e-06, + "loss": 46.483, + "step": 1424 + }, + { + "epoch": 33.93134328358209, + "grad_norm": 26.101816177368164, + "learning_rate": 6.589285714285715e-06, + "loss": 45.7998, + "step": 1425 + }, + { + "epoch": 33.95522388059702, + "grad_norm": 27.942903518676758, + "learning_rate": 6.5833333333333335e-06, + "loss": 46.4593, + "step": 1426 + }, + { + "epoch": 33.97910447761194, + "grad_norm": 21.551429748535156, + "learning_rate": 6.5773809523809525e-06, + "loss": 45.458, + "step": 1427 + }, + { + "epoch": 34.0, + "grad_norm": 32.26907730102539, + "learning_rate": 6.571428571428572e-06, + "loss": 38.5718, + "step": 1428 + }, + { + "epoch": 34.02388059701492, + "grad_norm": 32.16934585571289, + "learning_rate": 6.565476190476191e-06, + "loss": 45.5812, + "step": 1429 + }, + { + "epoch": 34.04776119402985, + "grad_norm": 19.646459579467773, + "learning_rate": 6.55952380952381e-06, + "loss": 44.9032, + "step": 1430 + }, + { + "epoch": 34.071641791044776, + "grad_norm": 28.886430740356445, + "learning_rate": 6.553571428571429e-06, + "loss": 45.4187, + "step": 1431 + }, + { + "epoch": 34.0955223880597, + "grad_norm": 22.722471237182617, + "learning_rate": 6.547619047619048e-06, + "loss": 45.468, + "step": 1432 + }, + { + "epoch": 34.11940298507463, + "grad_norm": 25.334766387939453, + "learning_rate": 6.541666666666667e-06, + "loss": 47.3534, + "step": 1433 + }, + { + "epoch": 34.14328358208955, + "grad_norm": 28.49740982055664, + "learning_rate": 6.535714285714286e-06, + "loss": 47.4733, + "step": 1434 + }, + { + "epoch": 34.167164179104475, + "grad_norm": 27.773820877075195, + "learning_rate": 6.529761904761906e-06, + "loss": 45.3215, + "step": 1435 + }, + { + "epoch": 34.191044776119405, + "grad_norm": 24.25234031677246, + "learning_rate": 6.523809523809525e-06, + "loss": 46.0011, + "step": 1436 + }, + { + "epoch": 34.21492537313433, + "grad_norm": 28.666475296020508, + "learning_rate": 6.517857142857144e-06, + "loss": 45.9091, + "step": 1437 + }, + { + "epoch": 34.23880597014925, + "grad_norm": 24.367712020874023, + "learning_rate": 6.5119047619047626e-06, + "loss": 46.5004, + "step": 1438 + }, + { + "epoch": 34.26268656716418, + "grad_norm": 23.11983299255371, + "learning_rate": 6.5059523809523815e-06, + "loss": 47.3335, + "step": 1439 + }, + { + "epoch": 34.286567164179104, + "grad_norm": 20.672304153442383, + "learning_rate": 6.5000000000000004e-06, + "loss": 47.1491, + "step": 1440 + }, + { + "epoch": 34.31044776119403, + "grad_norm": 23.815290451049805, + "learning_rate": 6.49404761904762e-06, + "loss": 46.7084, + "step": 1441 + }, + { + "epoch": 34.33432835820896, + "grad_norm": 20.582489013671875, + "learning_rate": 6.488095238095239e-06, + "loss": 46.9707, + "step": 1442 + }, + { + "epoch": 34.35820895522388, + "grad_norm": 18.315673828125, + "learning_rate": 6.482142857142858e-06, + "loss": 47.5359, + "step": 1443 + }, + { + "epoch": 34.3820895522388, + "grad_norm": 24.396499633789062, + "learning_rate": 6.476190476190477e-06, + "loss": 46.052, + "step": 1444 + }, + { + "epoch": 34.40597014925373, + "grad_norm": 21.200523376464844, + "learning_rate": 6.470238095238096e-06, + "loss": 46.5843, + "step": 1445 + }, + { + "epoch": 34.429850746268656, + "grad_norm": 17.59020233154297, + "learning_rate": 6.464285714285715e-06, + "loss": 46.0017, + "step": 1446 + }, + { + "epoch": 34.45373134328358, + "grad_norm": 21.810382843017578, + "learning_rate": 6.458333333333334e-06, + "loss": 46.4232, + "step": 1447 + }, + { + "epoch": 34.47761194029851, + "grad_norm": 27.78464126586914, + "learning_rate": 6.452380952380954e-06, + "loss": 46.0973, + "step": 1448 + }, + { + "epoch": 34.50149253731343, + "grad_norm": 29.360275268554688, + "learning_rate": 6.446428571428573e-06, + "loss": 45.4821, + "step": 1449 + }, + { + "epoch": 34.525373134328355, + "grad_norm": 26.914587020874023, + "learning_rate": 6.4404761904761916e-06, + "loss": 45.2982, + "step": 1450 + }, + { + "epoch": 34.549253731343285, + "grad_norm": 22.19925880432129, + "learning_rate": 6.4345238095238105e-06, + "loss": 46.6693, + "step": 1451 + }, + { + "epoch": 34.57313432835821, + "grad_norm": 25.39541244506836, + "learning_rate": 6.4285714285714295e-06, + "loss": 45.8936, + "step": 1452 + }, + { + "epoch": 34.59701492537313, + "grad_norm": 20.633222579956055, + "learning_rate": 6.422619047619048e-06, + "loss": 44.6061, + "step": 1453 + }, + { + "epoch": 34.62089552238806, + "grad_norm": 22.513790130615234, + "learning_rate": 6.416666666666667e-06, + "loss": 45.5503, + "step": 1454 + }, + { + "epoch": 34.644776119402984, + "grad_norm": 25.715484619140625, + "learning_rate": 6.410714285714287e-06, + "loss": 45.7485, + "step": 1455 + }, + { + "epoch": 34.668656716417914, + "grad_norm": 21.964609146118164, + "learning_rate": 6.404761904761904e-06, + "loss": 46.3223, + "step": 1456 + }, + { + "epoch": 34.69253731343284, + "grad_norm": 20.32435417175293, + "learning_rate": 6.398809523809524e-06, + "loss": 45.1507, + "step": 1457 + }, + { + "epoch": 34.71641791044776, + "grad_norm": 24.32924461364746, + "learning_rate": 6.392857142857143e-06, + "loss": 45.8221, + "step": 1458 + }, + { + "epoch": 34.74029850746269, + "grad_norm": 19.200895309448242, + "learning_rate": 6.386904761904762e-06, + "loss": 45.0915, + "step": 1459 + }, + { + "epoch": 34.76417910447761, + "grad_norm": 24.436569213867188, + "learning_rate": 6.380952380952381e-06, + "loss": 45.5892, + "step": 1460 + }, + { + "epoch": 34.788059701492536, + "grad_norm": 24.381568908691406, + "learning_rate": 6.375e-06, + "loss": 45.5295, + "step": 1461 + }, + { + "epoch": 34.811940298507466, + "grad_norm": 19.64159393310547, + "learning_rate": 6.369047619047619e-06, + "loss": 46.244, + "step": 1462 + }, + { + "epoch": 34.83582089552239, + "grad_norm": 27.420351028442383, + "learning_rate": 6.363095238095238e-06, + "loss": 45.9723, + "step": 1463 + }, + { + "epoch": 34.85970149253731, + "grad_norm": 18.136165618896484, + "learning_rate": 6.357142857142858e-06, + "loss": 45.5106, + "step": 1464 + }, + { + "epoch": 34.88358208955224, + "grad_norm": 21.70622444152832, + "learning_rate": 6.3511904761904766e-06, + "loss": 46.4965, + "step": 1465 + }, + { + "epoch": 34.907462686567165, + "grad_norm": 23.573131561279297, + "learning_rate": 6.3452380952380955e-06, + "loss": 46.0698, + "step": 1466 + }, + { + "epoch": 34.93134328358209, + "grad_norm": 21.20003890991211, + "learning_rate": 6.3392857142857145e-06, + "loss": 45.6992, + "step": 1467 + }, + { + "epoch": 34.95522388059702, + "grad_norm": 23.745859146118164, + "learning_rate": 6.333333333333333e-06, + "loss": 45.8431, + "step": 1468 + }, + { + "epoch": 34.97910447761194, + "grad_norm": 21.26241683959961, + "learning_rate": 6.327380952380952e-06, + "loss": 45.6577, + "step": 1469 + }, + { + "epoch": 35.0, + "grad_norm": 22.033447265625, + "learning_rate": 6.321428571428571e-06, + "loss": 39.8491, + "step": 1470 + }, + { + "epoch": 35.02388059701492, + "grad_norm": NaN, + "learning_rate": 6.315476190476191e-06, + "loss": 68.4405, + "step": 1471 + }, + { + "epoch": 35.04776119402985, + "grad_norm": 22.06501007080078, + "learning_rate": 6.315476190476191e-06, + "loss": 44.971, + "step": 1472 + }, + { + "epoch": 35.071641791044776, + "grad_norm": 23.923011779785156, + "learning_rate": 6.30952380952381e-06, + "loss": 45.4865, + "step": 1473 + }, + { + "epoch": 35.0955223880597, + "grad_norm": 18.272428512573242, + "learning_rate": 6.303571428571429e-06, + "loss": 46.6551, + "step": 1474 + }, + { + "epoch": 35.11940298507463, + "grad_norm": 23.046764373779297, + "learning_rate": 6.297619047619048e-06, + "loss": 46.3486, + "step": 1475 + }, + { + "epoch": 35.14328358208955, + "grad_norm": 23.790733337402344, + "learning_rate": 6.291666666666667e-06, + "loss": 46.7032, + "step": 1476 + }, + { + "epoch": 35.167164179104475, + "grad_norm": 23.891183853149414, + "learning_rate": 6.285714285714286e-06, + "loss": 44.9916, + "step": 1477 + }, + { + "epoch": 35.191044776119405, + "grad_norm": 25.107316970825195, + "learning_rate": 6.279761904761906e-06, + "loss": 46.2358, + "step": 1478 + }, + { + "epoch": 35.21492537313433, + "grad_norm": 20.48590660095215, + "learning_rate": 6.2738095238095245e-06, + "loss": 46.0048, + "step": 1479 + }, + { + "epoch": 35.23880597014925, + "grad_norm": 25.425119400024414, + "learning_rate": 6.2678571428571435e-06, + "loss": 44.0941, + "step": 1480 + }, + { + "epoch": 35.26268656716418, + "grad_norm": 28.264352798461914, + "learning_rate": 6.261904761904762e-06, + "loss": 46.5301, + "step": 1481 + }, + { + "epoch": 35.286567164179104, + "grad_norm": 23.869232177734375, + "learning_rate": 6.255952380952381e-06, + "loss": 45.681, + "step": 1482 + }, + { + "epoch": 35.31044776119403, + "grad_norm": 28.840408325195312, + "learning_rate": 6.25e-06, + "loss": 43.7517, + "step": 1483 + }, + { + "epoch": 35.33432835820896, + "grad_norm": 26.768037796020508, + "learning_rate": 6.244047619047619e-06, + "loss": 46.1423, + "step": 1484 + }, + { + "epoch": 35.35820895522388, + "grad_norm": 23.532470703125, + "learning_rate": 6.238095238095239e-06, + "loss": 45.6669, + "step": 1485 + }, + { + "epoch": 35.3820895522388, + "grad_norm": 25.94774055480957, + "learning_rate": 6.232142857142858e-06, + "loss": 45.7672, + "step": 1486 + }, + { + "epoch": 35.40597014925373, + "grad_norm": 23.215801239013672, + "learning_rate": 6.226190476190477e-06, + "loss": 45.6991, + "step": 1487 + }, + { + "epoch": 35.429850746268656, + "grad_norm": 22.13661003112793, + "learning_rate": 6.220238095238096e-06, + "loss": 44.5214, + "step": 1488 + }, + { + "epoch": 35.45373134328358, + "grad_norm": 24.596481323242188, + "learning_rate": 6.214285714285715e-06, + "loss": 46.1515, + "step": 1489 + }, + { + "epoch": 35.47761194029851, + "grad_norm": 19.416872024536133, + "learning_rate": 6.208333333333334e-06, + "loss": 45.7596, + "step": 1490 + }, + { + "epoch": 35.50149253731343, + "grad_norm": 23.993833541870117, + "learning_rate": 6.202380952380953e-06, + "loss": 46.1668, + "step": 1491 + }, + { + "epoch": 35.525373134328355, + "grad_norm": 21.481637954711914, + "learning_rate": 6.1964285714285725e-06, + "loss": 45.1812, + "step": 1492 + }, + { + "epoch": 35.549253731343285, + "grad_norm": 19.26917839050293, + "learning_rate": 6.1904761904761914e-06, + "loss": 45.9316, + "step": 1493 + }, + { + "epoch": 35.57313432835821, + "grad_norm": 22.80115509033203, + "learning_rate": 6.18452380952381e-06, + "loss": 45.9088, + "step": 1494 + }, + { + "epoch": 35.59701492537313, + "grad_norm": 21.33648109436035, + "learning_rate": 6.178571428571429e-06, + "loss": 46.7602, + "step": 1495 + }, + { + "epoch": 35.62089552238806, + "grad_norm": 28.059947967529297, + "learning_rate": 6.172619047619048e-06, + "loss": 46.1767, + "step": 1496 + }, + { + "epoch": 35.644776119402984, + "grad_norm": 21.1577205657959, + "learning_rate": 6.166666666666667e-06, + "loss": 45.6847, + "step": 1497 + }, + { + "epoch": 35.668656716417914, + "grad_norm": 23.277509689331055, + "learning_rate": 6.160714285714286e-06, + "loss": 45.6145, + "step": 1498 + }, + { + "epoch": 35.69253731343284, + "grad_norm": 16.815677642822266, + "learning_rate": 6.154761904761906e-06, + "loss": 45.515, + "step": 1499 + }, + { + "epoch": 35.71641791044776, + "grad_norm": 24.218280792236328, + "learning_rate": 6.148809523809525e-06, + "loss": 47.6329, + "step": 1500 + }, + { + "epoch": 35.74029850746269, + "grad_norm": 20.943737030029297, + "learning_rate": 6.142857142857144e-06, + "loss": 45.7388, + "step": 1501 + }, + { + "epoch": 35.76417910447761, + "grad_norm": 20.344369888305664, + "learning_rate": 6.136904761904763e-06, + "loss": 45.9404, + "step": 1502 + }, + { + "epoch": 35.788059701492536, + "grad_norm": 25.980487823486328, + "learning_rate": 6.130952380952382e-06, + "loss": 46.6928, + "step": 1503 + }, + { + "epoch": 35.811940298507466, + "grad_norm": 19.285552978515625, + "learning_rate": 6.125000000000001e-06, + "loss": 46.4614, + "step": 1504 + }, + { + "epoch": 35.83582089552239, + "grad_norm": 27.701011657714844, + "learning_rate": 6.11904761904762e-06, + "loss": 45.258, + "step": 1505 + }, + { + "epoch": 35.85970149253731, + "grad_norm": 24.963760375976562, + "learning_rate": 6.113095238095239e-06, + "loss": 47.0721, + "step": 1506 + }, + { + "epoch": 35.88358208955224, + "grad_norm": 25.08616828918457, + "learning_rate": 6.107142857142858e-06, + "loss": 45.9668, + "step": 1507 + }, + { + "epoch": 35.907462686567165, + "grad_norm": 18.00580406188965, + "learning_rate": 6.101190476190477e-06, + "loss": 46.1049, + "step": 1508 + }, + { + "epoch": 35.93134328358209, + "grad_norm": 24.686004638671875, + "learning_rate": 6.095238095238096e-06, + "loss": 46.6996, + "step": 1509 + }, + { + "epoch": 35.95522388059702, + "grad_norm": 18.304157257080078, + "learning_rate": 6.089285714285714e-06, + "loss": 46.694, + "step": 1510 + }, + { + "epoch": 35.97910447761194, + "grad_norm": 23.10132598876953, + "learning_rate": 6.083333333333333e-06, + "loss": 46.3807, + "step": 1511 + }, + { + "epoch": 36.0, + "grad_norm": 19.077655792236328, + "learning_rate": 6.077380952380952e-06, + "loss": 41.1702, + "step": 1512 + }, + { + "epoch": 36.02388059701492, + "grad_norm": 26.49584197998047, + "learning_rate": 6.071428571428571e-06, + "loss": 45.382, + "step": 1513 + }, + { + "epoch": 36.04776119402985, + "grad_norm": 24.438323974609375, + "learning_rate": 6.065476190476191e-06, + "loss": 45.9433, + "step": 1514 + }, + { + "epoch": 36.071641791044776, + "grad_norm": 30.8107852935791, + "learning_rate": 6.05952380952381e-06, + "loss": 45.6688, + "step": 1515 + }, + { + "epoch": 36.0955223880597, + "grad_norm": 31.754154205322266, + "learning_rate": 6.053571428571429e-06, + "loss": 45.9768, + "step": 1516 + }, + { + "epoch": 36.11940298507463, + "grad_norm": 26.034778594970703, + "learning_rate": 6.047619047619048e-06, + "loss": 46.022, + "step": 1517 + }, + { + "epoch": 36.14328358208955, + "grad_norm": 31.643035888671875, + "learning_rate": 6.041666666666667e-06, + "loss": 44.5987, + "step": 1518 + }, + { + "epoch": 36.167164179104475, + "grad_norm": 24.322874069213867, + "learning_rate": 6.035714285714286e-06, + "loss": 45.3774, + "step": 1519 + }, + { + "epoch": 36.191044776119405, + "grad_norm": 29.067466735839844, + "learning_rate": 6.029761904761905e-06, + "loss": 46.1784, + "step": 1520 + }, + { + "epoch": 36.21492537313433, + "grad_norm": 30.415788650512695, + "learning_rate": 6.023809523809524e-06, + "loss": 46.7259, + "step": 1521 + }, + { + "epoch": 36.23880597014925, + "grad_norm": 19.417943954467773, + "learning_rate": 6.017857142857143e-06, + "loss": 46.0544, + "step": 1522 + }, + { + "epoch": 36.26268656716418, + "grad_norm": 27.239500045776367, + "learning_rate": 6.011904761904762e-06, + "loss": 46.9344, + "step": 1523 + }, + { + "epoch": 36.286567164179104, + "grad_norm": 27.671018600463867, + "learning_rate": 6.005952380952381e-06, + "loss": 45.78, + "step": 1524 + }, + { + "epoch": 36.31044776119403, + "grad_norm": 25.103811264038086, + "learning_rate": 6e-06, + "loss": 45.7153, + "step": 1525 + }, + { + "epoch": 36.33432835820896, + "grad_norm": 26.25937843322754, + "learning_rate": 5.994047619047619e-06, + "loss": 45.3151, + "step": 1526 + }, + { + "epoch": 36.35820895522388, + "grad_norm": 18.400033950805664, + "learning_rate": 5.988095238095238e-06, + "loss": 46.5614, + "step": 1527 + }, + { + "epoch": 36.3820895522388, + "grad_norm": 35.505374908447266, + "learning_rate": 5.982142857142858e-06, + "loss": 45.8805, + "step": 1528 + }, + { + "epoch": 36.40597014925373, + "grad_norm": 31.476438522338867, + "learning_rate": 5.976190476190477e-06, + "loss": 46.189, + "step": 1529 + }, + { + "epoch": 36.429850746268656, + "grad_norm": 26.192047119140625, + "learning_rate": 5.970238095238096e-06, + "loss": 45.7026, + "step": 1530 + }, + { + "epoch": 36.45373134328358, + "grad_norm": 29.712961196899414, + "learning_rate": 5.964285714285715e-06, + "loss": 44.86, + "step": 1531 + }, + { + "epoch": 36.47761194029851, + "grad_norm": 28.22374153137207, + "learning_rate": 5.958333333333334e-06, + "loss": 45.7644, + "step": 1532 + }, + { + "epoch": 36.50149253731343, + "grad_norm": 23.614940643310547, + "learning_rate": 5.9523809523809525e-06, + "loss": 45.0373, + "step": 1533 + }, + { + "epoch": 36.525373134328355, + "grad_norm": 27.78896141052246, + "learning_rate": 5.9464285714285715e-06, + "loss": 46.9277, + "step": 1534 + }, + { + "epoch": 36.549253731343285, + "grad_norm": 18.64702606201172, + "learning_rate": 5.940476190476191e-06, + "loss": 45.277, + "step": 1535 + }, + { + "epoch": 36.57313432835821, + "grad_norm": 27.2061710357666, + "learning_rate": 5.93452380952381e-06, + "loss": 46.8394, + "step": 1536 + }, + { + "epoch": 36.59701492537313, + "grad_norm": 26.296287536621094, + "learning_rate": 5.928571428571429e-06, + "loss": 44.8519, + "step": 1537 + }, + { + "epoch": 36.62089552238806, + "grad_norm": 26.594314575195312, + "learning_rate": 5.922619047619048e-06, + "loss": 45.1743, + "step": 1538 + }, + { + "epoch": 36.644776119402984, + "grad_norm": 24.076461791992188, + "learning_rate": 5.916666666666667e-06, + "loss": 45.4145, + "step": 1539 + }, + { + "epoch": 36.668656716417914, + "grad_norm": 23.31978416442871, + "learning_rate": 5.910714285714286e-06, + "loss": 45.7526, + "step": 1540 + }, + { + "epoch": 36.69253731343284, + "grad_norm": 22.630998611450195, + "learning_rate": 5.904761904761905e-06, + "loss": 46.4197, + "step": 1541 + }, + { + "epoch": 36.71641791044776, + "grad_norm": 32.66592025756836, + "learning_rate": 5.898809523809525e-06, + "loss": 45.0123, + "step": 1542 + }, + { + "epoch": 36.74029850746269, + "grad_norm": 24.478839874267578, + "learning_rate": 5.892857142857144e-06, + "loss": 46.1418, + "step": 1543 + }, + { + "epoch": 36.76417910447761, + "grad_norm": 33.325775146484375, + "learning_rate": 5.886904761904763e-06, + "loss": 45.8228, + "step": 1544 + }, + { + "epoch": 36.788059701492536, + "grad_norm": 29.264528274536133, + "learning_rate": 5.8809523809523816e-06, + "loss": 46.1921, + "step": 1545 + }, + { + "epoch": 36.811940298507466, + "grad_norm": 31.78297233581543, + "learning_rate": 5.8750000000000005e-06, + "loss": 45.4564, + "step": 1546 + }, + { + "epoch": 36.83582089552239, + "grad_norm": 27.223127365112305, + "learning_rate": 5.8690476190476194e-06, + "loss": 45.5277, + "step": 1547 + }, + { + "epoch": 36.85970149253731, + "grad_norm": 26.29422950744629, + "learning_rate": 5.863095238095239e-06, + "loss": 46.2285, + "step": 1548 + }, + { + "epoch": 36.88358208955224, + "grad_norm": 27.933652877807617, + "learning_rate": 5.857142857142858e-06, + "loss": 46.4441, + "step": 1549 + }, + { + "epoch": 36.907462686567165, + "grad_norm": 25.306129455566406, + "learning_rate": 5.851190476190477e-06, + "loss": 45.9724, + "step": 1550 + }, + { + "epoch": 36.93134328358209, + "grad_norm": 23.481304168701172, + "learning_rate": 5.845238095238096e-06, + "loss": 46.2544, + "step": 1551 + }, + { + "epoch": 36.95522388059702, + "grad_norm": 20.86615562438965, + "learning_rate": 5.839285714285715e-06, + "loss": 47.4502, + "step": 1552 + }, + { + "epoch": 36.97910447761194, + "grad_norm": 21.519290924072266, + "learning_rate": 5.833333333333334e-06, + "loss": 45.0165, + "step": 1553 + }, + { + "epoch": 37.0, + "grad_norm": 22.031705856323242, + "learning_rate": 5.827380952380953e-06, + "loss": 40.6199, + "step": 1554 + }, + { + "epoch": 37.02388059701492, + "grad_norm": 29.273820877075195, + "learning_rate": 5.821428571428573e-06, + "loss": 46.5836, + "step": 1555 + }, + { + "epoch": 37.04776119402985, + "grad_norm": 24.417945861816406, + "learning_rate": 5.815476190476192e-06, + "loss": 44.9549, + "step": 1556 + }, + { + "epoch": 37.071641791044776, + "grad_norm": 24.60706901550293, + "learning_rate": 5.8095238095238106e-06, + "loss": 44.8607, + "step": 1557 + }, + { + "epoch": 37.0955223880597, + "grad_norm": 24.76397132873535, + "learning_rate": 5.8035714285714295e-06, + "loss": 44.9875, + "step": 1558 + }, + { + "epoch": 37.11940298507463, + "grad_norm": 24.380352020263672, + "learning_rate": 5.7976190476190485e-06, + "loss": 45.4835, + "step": 1559 + }, + { + "epoch": 37.14328358208955, + "grad_norm": 19.852746963500977, + "learning_rate": 5.791666666666667e-06, + "loss": 45.1303, + "step": 1560 + }, + { + "epoch": 37.167164179104475, + "grad_norm": 23.550888061523438, + "learning_rate": 5.785714285714286e-06, + "loss": 46.1086, + "step": 1561 + }, + { + "epoch": 37.191044776119405, + "grad_norm": 24.31315803527832, + "learning_rate": 5.7797619047619044e-06, + "loss": 45.8181, + "step": 1562 + }, + { + "epoch": 37.21492537313433, + "grad_norm": 19.324602127075195, + "learning_rate": 5.773809523809523e-06, + "loss": 44.8606, + "step": 1563 + }, + { + "epoch": 37.23880597014925, + "grad_norm": 26.747098922729492, + "learning_rate": 5.767857142857143e-06, + "loss": 45.753, + "step": 1564 + }, + { + "epoch": 37.26268656716418, + "grad_norm": 22.472572326660156, + "learning_rate": 5.761904761904762e-06, + "loss": 46.0156, + "step": 1565 + }, + { + "epoch": 37.286567164179104, + "grad_norm": 20.813426971435547, + "learning_rate": 5.755952380952381e-06, + "loss": 46.7466, + "step": 1566 + }, + { + "epoch": 37.31044776119403, + "grad_norm": 27.869413375854492, + "learning_rate": 5.75e-06, + "loss": 46.287, + "step": 1567 + }, + { + "epoch": 37.33432835820896, + "grad_norm": 23.257444381713867, + "learning_rate": 5.744047619047619e-06, + "loss": 45.9862, + "step": 1568 + }, + { + "epoch": 37.35820895522388, + "grad_norm": 24.715946197509766, + "learning_rate": 5.738095238095238e-06, + "loss": 47.3128, + "step": 1569 + }, + { + "epoch": 37.3820895522388, + "grad_norm": 21.670385360717773, + "learning_rate": 5.732142857142857e-06, + "loss": 46.121, + "step": 1570 + }, + { + "epoch": 37.40597014925373, + "grad_norm": 24.53063201904297, + "learning_rate": 5.726190476190477e-06, + "loss": 46.5441, + "step": 1571 + }, + { + "epoch": 37.429850746268656, + "grad_norm": 19.584630966186523, + "learning_rate": 5.7202380952380956e-06, + "loss": 46.0683, + "step": 1572 + }, + { + "epoch": 37.45373134328358, + "grad_norm": 26.179149627685547, + "learning_rate": 5.7142857142857145e-06, + "loss": 46.3294, + "step": 1573 + }, + { + "epoch": 37.47761194029851, + "grad_norm": 21.13595199584961, + "learning_rate": 5.7083333333333335e-06, + "loss": 45.7853, + "step": 1574 + }, + { + "epoch": 37.50149253731343, + "grad_norm": 28.440006256103516, + "learning_rate": 5.702380952380952e-06, + "loss": 46.5029, + "step": 1575 + }, + { + "epoch": 37.525373134328355, + "grad_norm": 27.941879272460938, + "learning_rate": 5.696428571428571e-06, + "loss": 45.6132, + "step": 1576 + }, + { + "epoch": 37.549253731343285, + "grad_norm": 25.952688217163086, + "learning_rate": 5.690476190476191e-06, + "loss": 45.6803, + "step": 1577 + }, + { + "epoch": 37.57313432835821, + "grad_norm": 23.551633834838867, + "learning_rate": 5.68452380952381e-06, + "loss": 45.1563, + "step": 1578 + }, + { + "epoch": 37.59701492537313, + "grad_norm": 23.119415283203125, + "learning_rate": 5.678571428571429e-06, + "loss": 47.2717, + "step": 1579 + }, + { + "epoch": 37.62089552238806, + "grad_norm": 27.995214462280273, + "learning_rate": 5.672619047619048e-06, + "loss": 46.1847, + "step": 1580 + }, + { + "epoch": 37.644776119402984, + "grad_norm": 28.0698299407959, + "learning_rate": 5.666666666666667e-06, + "loss": 46.4639, + "step": 1581 + }, + { + "epoch": 37.668656716417914, + "grad_norm": 23.09457015991211, + "learning_rate": 5.660714285714286e-06, + "loss": 45.0939, + "step": 1582 + }, + { + "epoch": 37.69253731343284, + "grad_norm": 25.94692611694336, + "learning_rate": 5.654761904761905e-06, + "loss": 45.216, + "step": 1583 + }, + { + "epoch": 37.71641791044776, + "grad_norm": 20.192176818847656, + "learning_rate": 5.648809523809525e-06, + "loss": 45.7997, + "step": 1584 + }, + { + "epoch": 37.74029850746269, + "grad_norm": 26.115283966064453, + "learning_rate": 5.6428571428571435e-06, + "loss": 44.8405, + "step": 1585 + }, + { + "epoch": 37.76417910447761, + "grad_norm": 24.431346893310547, + "learning_rate": 5.6369047619047625e-06, + "loss": 46.5067, + "step": 1586 + }, + { + "epoch": 37.788059701492536, + "grad_norm": 25.838623046875, + "learning_rate": 5.630952380952381e-06, + "loss": 46.1806, + "step": 1587 + }, + { + "epoch": 37.811940298507466, + "grad_norm": 20.44222640991211, + "learning_rate": 5.625e-06, + "loss": 45.7445, + "step": 1588 + }, + { + "epoch": 37.83582089552239, + "grad_norm": 19.459331512451172, + "learning_rate": 5.619047619047619e-06, + "loss": 45.7875, + "step": 1589 + }, + { + "epoch": 37.85970149253731, + "grad_norm": 17.49920082092285, + "learning_rate": 5.613095238095238e-06, + "loss": 44.2889, + "step": 1590 + }, + { + "epoch": 37.88358208955224, + "grad_norm": 18.541828155517578, + "learning_rate": 5.607142857142858e-06, + "loss": 46.7668, + "step": 1591 + }, + { + "epoch": 37.907462686567165, + "grad_norm": 16.22308349609375, + "learning_rate": 5.601190476190477e-06, + "loss": 45.0406, + "step": 1592 + }, + { + "epoch": 37.93134328358209, + "grad_norm": 21.068069458007812, + "learning_rate": 5.595238095238096e-06, + "loss": 44.0997, + "step": 1593 + }, + { + "epoch": 37.95522388059702, + "grad_norm": 18.877992630004883, + "learning_rate": 5.589285714285715e-06, + "loss": 46.5816, + "step": 1594 + }, + { + "epoch": 37.97910447761194, + "grad_norm": 20.14031410217285, + "learning_rate": 5.583333333333334e-06, + "loss": 44.8537, + "step": 1595 + }, + { + "epoch": 38.0, + "grad_norm": 19.989953994750977, + "learning_rate": 5.577380952380953e-06, + "loss": 39.8501, + "step": 1596 + }, + { + "epoch": 38.02388059701492, + "grad_norm": 23.484283447265625, + "learning_rate": 5.571428571428572e-06, + "loss": 46.3864, + "step": 1597 + }, + { + "epoch": 38.04776119402985, + "grad_norm": 20.579587936401367, + "learning_rate": 5.5654761904761915e-06, + "loss": 46.1473, + "step": 1598 + }, + { + "epoch": 38.071641791044776, + "grad_norm": 19.48423194885254, + "learning_rate": 5.5595238095238104e-06, + "loss": 45.3255, + "step": 1599 + }, + { + "epoch": 38.0955223880597, + "grad_norm": 23.766077041625977, + "learning_rate": 5.553571428571429e-06, + "loss": 45.4387, + "step": 1600 + }, + { + "epoch": 38.11940298507463, + "grad_norm": 17.605247497558594, + "learning_rate": 5.547619047619048e-06, + "loss": 46.1065, + "step": 1601 + }, + { + "epoch": 38.14328358208955, + "grad_norm": 20.179826736450195, + "learning_rate": 5.541666666666667e-06, + "loss": 45.974, + "step": 1602 + }, + { + "epoch": 38.167164179104475, + "grad_norm": 28.50605583190918, + "learning_rate": 5.535714285714286e-06, + "loss": 46.0505, + "step": 1603 + }, + { + "epoch": 38.191044776119405, + "grad_norm": 16.770771026611328, + "learning_rate": 5.529761904761905e-06, + "loss": 46.4403, + "step": 1604 + }, + { + "epoch": 38.21492537313433, + "grad_norm": NaN, + "learning_rate": 5.523809523809525e-06, + "loss": 69.3153, + "step": 1605 + }, + { + "epoch": 38.23880597014925, + "grad_norm": 25.01431655883789, + "learning_rate": 5.523809523809525e-06, + "loss": 46.8119, + "step": 1606 + }, + { + "epoch": 38.26268656716418, + "grad_norm": 20.459747314453125, + "learning_rate": 5.517857142857144e-06, + "loss": 47.7687, + "step": 1607 + }, + { + "epoch": 38.286567164179104, + "grad_norm": 21.603086471557617, + "learning_rate": 5.511904761904763e-06, + "loss": 44.6093, + "step": 1608 + }, + { + "epoch": 38.31044776119403, + "grad_norm": 25.284805297851562, + "learning_rate": 5.505952380952382e-06, + "loss": 45.0834, + "step": 1609 + }, + { + "epoch": 38.33432835820896, + "grad_norm": 21.638917922973633, + "learning_rate": 5.500000000000001e-06, + "loss": 45.3904, + "step": 1610 + }, + { + "epoch": 38.35820895522388, + "grad_norm": 22.443374633789062, + "learning_rate": 5.49404761904762e-06, + "loss": 43.7163, + "step": 1611 + }, + { + "epoch": 38.3820895522388, + "grad_norm": 23.427288055419922, + "learning_rate": 5.4880952380952394e-06, + "loss": 44.7692, + "step": 1612 + }, + { + "epoch": 38.40597014925373, + "grad_norm": 22.346813201904297, + "learning_rate": 5.482142857142858e-06, + "loss": 45.0674, + "step": 1613 + }, + { + "epoch": 38.429850746268656, + "grad_norm": 20.567325592041016, + "learning_rate": 5.476190476190477e-06, + "loss": 45.5367, + "step": 1614 + }, + { + "epoch": 38.45373134328358, + "grad_norm": 23.872394561767578, + "learning_rate": 5.470238095238096e-06, + "loss": 46.2728, + "step": 1615 + }, + { + "epoch": 38.47761194029851, + "grad_norm": 23.790176391601562, + "learning_rate": 5.464285714285714e-06, + "loss": 46.3734, + "step": 1616 + }, + { + "epoch": 38.50149253731343, + "grad_norm": 22.707136154174805, + "learning_rate": 5.458333333333333e-06, + "loss": 44.577, + "step": 1617 + }, + { + "epoch": 38.525373134328355, + "grad_norm": 26.203781127929688, + "learning_rate": 5.452380952380952e-06, + "loss": 45.6794, + "step": 1618 + }, + { + "epoch": 38.549253731343285, + "grad_norm": 22.935991287231445, + "learning_rate": 5.446428571428571e-06, + "loss": 45.7815, + "step": 1619 + }, + { + "epoch": 38.57313432835821, + "grad_norm": 28.275053024291992, + "learning_rate": 5.44047619047619e-06, + "loss": 45.0312, + "step": 1620 + }, + { + "epoch": 38.59701492537313, + "grad_norm": 23.848264694213867, + "learning_rate": 5.43452380952381e-06, + "loss": 46.7093, + "step": 1621 + }, + { + "epoch": 38.62089552238806, + "grad_norm": 25.240819931030273, + "learning_rate": 5.428571428571429e-06, + "loss": 46.6751, + "step": 1622 + }, + { + "epoch": 38.644776119402984, + "grad_norm": 26.2618350982666, + "learning_rate": 5.422619047619048e-06, + "loss": 47.5501, + "step": 1623 + }, + { + "epoch": 38.668656716417914, + "grad_norm": 23.986392974853516, + "learning_rate": 5.416666666666667e-06, + "loss": 45.6208, + "step": 1624 + }, + { + "epoch": 38.69253731343284, + "grad_norm": 22.11539077758789, + "learning_rate": 5.410714285714286e-06, + "loss": 44.4163, + "step": 1625 + }, + { + "epoch": 38.71641791044776, + "grad_norm": 22.9071044921875, + "learning_rate": 5.404761904761905e-06, + "loss": 45.5715, + "step": 1626 + }, + { + "epoch": 38.74029850746269, + "grad_norm": 22.759733200073242, + "learning_rate": 5.398809523809524e-06, + "loss": 45.1706, + "step": 1627 + }, + { + "epoch": 38.76417910447761, + "grad_norm": 23.66644287109375, + "learning_rate": 5.392857142857143e-06, + "loss": 45.4343, + "step": 1628 + }, + { + "epoch": 38.788059701492536, + "grad_norm": 20.179203033447266, + "learning_rate": 5.386904761904762e-06, + "loss": 45.9163, + "step": 1629 + }, + { + "epoch": 38.811940298507466, + "grad_norm": 22.327817916870117, + "learning_rate": 5.380952380952381e-06, + "loss": 44.1558, + "step": 1630 + }, + { + "epoch": 38.83582089552239, + "grad_norm": 22.10496711730957, + "learning_rate": 5.375e-06, + "loss": 45.764, + "step": 1631 + }, + { + "epoch": 38.85970149253731, + "grad_norm": 24.25627326965332, + "learning_rate": 5.369047619047619e-06, + "loss": 46.6394, + "step": 1632 + }, + { + "epoch": 38.88358208955224, + "grad_norm": 20.797740936279297, + "learning_rate": 5.363095238095238e-06, + "loss": 45.6251, + "step": 1633 + }, + { + "epoch": 38.907462686567165, + "grad_norm": 24.14659309387207, + "learning_rate": 5.357142857142857e-06, + "loss": 45.5603, + "step": 1634 + }, + { + "epoch": 38.93134328358209, + "grad_norm": 23.259584426879883, + "learning_rate": 5.351190476190477e-06, + "loss": 46.055, + "step": 1635 + }, + { + "epoch": 38.95522388059702, + "grad_norm": 23.72128677368164, + "learning_rate": 5.345238095238096e-06, + "loss": 45.9729, + "step": 1636 + }, + { + "epoch": 38.97910447761194, + "grad_norm": 22.746183395385742, + "learning_rate": 5.339285714285715e-06, + "loss": 46.1893, + "step": 1637 + }, + { + "epoch": 39.0, + "grad_norm": 22.067306518554688, + "learning_rate": 5.333333333333334e-06, + "loss": 39.8095, + "step": 1638 + }, + { + "epoch": 39.02388059701492, + "grad_norm": 22.888097763061523, + "learning_rate": 5.327380952380953e-06, + "loss": 45.3095, + "step": 1639 + }, + { + "epoch": 39.04776119402985, + "grad_norm": 23.86408233642578, + "learning_rate": 5.3214285714285715e-06, + "loss": 46.3774, + "step": 1640 + }, + { + "epoch": 39.071641791044776, + "grad_norm": 21.418088912963867, + "learning_rate": 5.3154761904761905e-06, + "loss": 45.6404, + "step": 1641 + }, + { + "epoch": 39.0955223880597, + "grad_norm": 21.521831512451172, + "learning_rate": 5.30952380952381e-06, + "loss": 46.4895, + "step": 1642 + }, + { + "epoch": 39.11940298507463, + "grad_norm": 20.189105987548828, + "learning_rate": 5.303571428571429e-06, + "loss": 44.6538, + "step": 1643 + }, + { + "epoch": 39.14328358208955, + "grad_norm": 19.73761558532715, + "learning_rate": 5.297619047619048e-06, + "loss": 45.5941, + "step": 1644 + }, + { + "epoch": 39.167164179104475, + "grad_norm": 25.631227493286133, + "learning_rate": 5.291666666666667e-06, + "loss": 44.4105, + "step": 1645 + }, + { + "epoch": 39.191044776119405, + "grad_norm": 19.47798728942871, + "learning_rate": 5.285714285714286e-06, + "loss": 46.3286, + "step": 1646 + }, + { + "epoch": 39.21492537313433, + "grad_norm": 19.627609252929688, + "learning_rate": 5.279761904761905e-06, + "loss": 46.0707, + "step": 1647 + }, + { + "epoch": 39.23880597014925, + "grad_norm": 19.668777465820312, + "learning_rate": 5.273809523809525e-06, + "loss": 44.8447, + "step": 1648 + }, + { + "epoch": 39.26268656716418, + "grad_norm": 23.311546325683594, + "learning_rate": 5.267857142857144e-06, + "loss": 45.9081, + "step": 1649 + }, + { + "epoch": 39.286567164179104, + "grad_norm": 21.426624298095703, + "learning_rate": 5.261904761904763e-06, + "loss": 45.4256, + "step": 1650 + }, + { + "epoch": 39.31044776119403, + "grad_norm": 19.545969009399414, + "learning_rate": 5.255952380952382e-06, + "loss": 45.4264, + "step": 1651 + }, + { + "epoch": 39.33432835820896, + "grad_norm": 22.78704833984375, + "learning_rate": 5.2500000000000006e-06, + "loss": 47.6822, + "step": 1652 + }, + { + "epoch": 39.35820895522388, + "grad_norm": 18.759178161621094, + "learning_rate": 5.2440476190476195e-06, + "loss": 44.5254, + "step": 1653 + }, + { + "epoch": 39.3820895522388, + "grad_norm": 19.855981826782227, + "learning_rate": 5.2380952380952384e-06, + "loss": 46.249, + "step": 1654 + }, + { + "epoch": 39.40597014925373, + "grad_norm": 18.817089080810547, + "learning_rate": 5.232142857142858e-06, + "loss": 45.2813, + "step": 1655 + }, + { + "epoch": 39.429850746268656, + "grad_norm": 19.587581634521484, + "learning_rate": 5.226190476190477e-06, + "loss": 45.0445, + "step": 1656 + }, + { + "epoch": 39.45373134328358, + "grad_norm": 19.9105167388916, + "learning_rate": 5.220238095238096e-06, + "loss": 46.8658, + "step": 1657 + }, + { + "epoch": 39.47761194029851, + "grad_norm": 19.529748916625977, + "learning_rate": 5.214285714285715e-06, + "loss": 46.6175, + "step": 1658 + }, + { + "epoch": 39.50149253731343, + "grad_norm": 18.63764762878418, + "learning_rate": 5.208333333333334e-06, + "loss": 46.3122, + "step": 1659 + }, + { + "epoch": 39.525373134328355, + "grad_norm": 19.58228874206543, + "learning_rate": 5.202380952380953e-06, + "loss": 44.6263, + "step": 1660 + }, + { + "epoch": 39.549253731343285, + "grad_norm": 21.451528549194336, + "learning_rate": 5.196428571428572e-06, + "loss": 46.2707, + "step": 1661 + }, + { + "epoch": 39.57313432835821, + "grad_norm": 22.756628036499023, + "learning_rate": 5.190476190476192e-06, + "loss": 45.0001, + "step": 1662 + }, + { + "epoch": 39.59701492537313, + "grad_norm": 24.481945037841797, + "learning_rate": 5.184523809523811e-06, + "loss": 45.3038, + "step": 1663 + }, + { + "epoch": 39.62089552238806, + "grad_norm": 19.3010196685791, + "learning_rate": 5.1785714285714296e-06, + "loss": 46.1894, + "step": 1664 + }, + { + "epoch": 39.644776119402984, + "grad_norm": 24.840822219848633, + "learning_rate": 5.1726190476190485e-06, + "loss": 46.6593, + "step": 1665 + }, + { + "epoch": 39.668656716417914, + "grad_norm": 20.712875366210938, + "learning_rate": 5.1666666666666675e-06, + "loss": 46.7594, + "step": 1666 + }, + { + "epoch": 39.69253731343284, + "grad_norm": 20.431598663330078, + "learning_rate": 5.160714285714286e-06, + "loss": 46.4969, + "step": 1667 + }, + { + "epoch": 39.71641791044776, + "grad_norm": 21.094484329223633, + "learning_rate": 5.1547619047619045e-06, + "loss": 44.5114, + "step": 1668 + }, + { + "epoch": 39.74029850746269, + "grad_norm": 22.929946899414062, + "learning_rate": 5.1488095238095234e-06, + "loss": 44.5251, + "step": 1669 + }, + { + "epoch": 39.76417910447761, + "grad_norm": 17.285877227783203, + "learning_rate": 5.142857142857142e-06, + "loss": 45.312, + "step": 1670 + }, + { + "epoch": 39.788059701492536, + "grad_norm": 18.29960823059082, + "learning_rate": 5.136904761904762e-06, + "loss": 44.984, + "step": 1671 + }, + { + "epoch": 39.811940298507466, + "grad_norm": 25.79044532775879, + "learning_rate": 5.130952380952381e-06, + "loss": 44.8192, + "step": 1672 + }, + { + "epoch": 39.83582089552239, + "grad_norm": 21.014759063720703, + "learning_rate": 5.125e-06, + "loss": 46.9319, + "step": 1673 + }, + { + "epoch": 39.85970149253731, + "grad_norm": 24.91911506652832, + "learning_rate": 5.119047619047619e-06, + "loss": 46.7778, + "step": 1674 + }, + { + "epoch": 39.88358208955224, + "grad_norm": 25.174942016601562, + "learning_rate": 5.113095238095238e-06, + "loss": 44.948, + "step": 1675 + }, + { + "epoch": 39.907462686567165, + "grad_norm": 22.642148971557617, + "learning_rate": 5.107142857142857e-06, + "loss": 45.5964, + "step": 1676 + }, + { + "epoch": 39.93134328358209, + "grad_norm": 24.867389678955078, + "learning_rate": 5.101190476190476e-06, + "loss": 45.446, + "step": 1677 + }, + { + "epoch": 39.95522388059702, + "grad_norm": 21.888269424438477, + "learning_rate": 5.095238095238096e-06, + "loss": 45.414, + "step": 1678 + }, + { + "epoch": 39.97910447761194, + "grad_norm": 25.071487426757812, + "learning_rate": 5.0892857142857146e-06, + "loss": 44.9464, + "step": 1679 + }, + { + "epoch": 40.0, + "grad_norm": 19.389556884765625, + "learning_rate": 5.0833333333333335e-06, + "loss": 39.5515, + "step": 1680 + } + ], + "logging_steps": 1.0, + "max_steps": 1680, + "num_input_tokens_seen": 0, + "num_train_epochs": 40, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 8.26172747445074e+16, + "train_batch_size": 16, + "trial_name": null, + "trial_params": null +}