diff --git "a/trainer_state.json" "b/trainer_state.json" --- "a/trainer_state.json" +++ "b/trainer_state.json" @@ -10,8625 +10,8625 @@ "log_history": [ { "epoch": 0.004050632911392405, - "grad_norm": 6.411459448248905, + "grad_norm": 6.616540841176689, "learning_rate": 3.2520325203252037e-07, - "loss": 0.7982, + "loss": 0.799, "step": 1 }, { "epoch": 0.00810126582278481, - "grad_norm": 6.284441576705379, + "grad_norm": 6.474323454882261, "learning_rate": 6.504065040650407e-07, - "loss": 0.7931, + "loss": 0.7966, "step": 2 }, { "epoch": 0.012151898734177215, - "grad_norm": 6.442810424206152, + "grad_norm": 6.590202665967519, "learning_rate": 9.75609756097561e-07, - "loss": 0.8167, + "loss": 0.8157, "step": 3 }, { "epoch": 0.01620253164556962, - "grad_norm": 6.226667896495011, + "grad_norm": 6.4407197325404395, "learning_rate": 1.3008130081300815e-06, - "loss": 0.8006, + "loss": 0.8138, "step": 4 }, { "epoch": 0.020253164556962026, - "grad_norm": 5.873490463358072, + "grad_norm": 6.158479270927061, "learning_rate": 1.6260162601626018e-06, - "loss": 0.7514, + "loss": 0.7721, "step": 5 }, { "epoch": 0.02430379746835443, - "grad_norm": 5.467481784137773, + "grad_norm": 5.677102963188688, "learning_rate": 1.951219512195122e-06, - "loss": 0.7563, + "loss": 0.7531, "step": 6 }, { "epoch": 0.028354430379746835, - "grad_norm": 4.479939821002372, + "grad_norm": 4.676756291527093, "learning_rate": 2.2764227642276426e-06, - "loss": 0.7363, + "loss": 0.7373, "step": 7 }, { "epoch": 0.03240506329113924, - "grad_norm": 4.078308749114432, + "grad_norm": 4.343262501135752, "learning_rate": 2.601626016260163e-06, - "loss": 0.7044, + "loss": 0.7121, "step": 8 }, { "epoch": 0.03645569620253165, - "grad_norm": 2.2952352989540277, + "grad_norm": 2.6756698165724946, "learning_rate": 2.926829268292683e-06, - "loss": 0.6726, + "loss": 0.6824, "step": 9 }, { "epoch": 0.04050632911392405, - "grad_norm": 2.110913514798186, + "grad_norm": 2.3675911494991886, "learning_rate": 3.2520325203252037e-06, - "loss": 0.6764, + "loss": 0.6685, "step": 10 }, { "epoch": 0.044556962025316456, - "grad_norm": 2.0488549240978524, + "grad_norm": 2.3311962373874815, "learning_rate": 3.577235772357724e-06, - "loss": 0.6546, + "loss": 0.645, "step": 11 }, { "epoch": 0.04860759493670886, - "grad_norm": 3.364772695618668, + "grad_norm": 3.205846081925055, "learning_rate": 3.902439024390244e-06, - "loss": 0.637, + "loss": 0.6237, "step": 12 }, { "epoch": 0.052658227848101265, - "grad_norm": 3.682443419858485, + "grad_norm": 3.3083916500841353, "learning_rate": 4.227642276422765e-06, - "loss": 0.6665, + "loss": 0.6387, "step": 13 }, { "epoch": 0.05670886075949367, - "grad_norm": 3.281775226350443, + "grad_norm": 3.0986106516707808, "learning_rate": 4.552845528455285e-06, - "loss": 0.6211, + "loss": 0.5988, "step": 14 }, { "epoch": 0.060759493670886074, - "grad_norm": 3.408841164408115, + "grad_norm": 3.0968269938757107, "learning_rate": 4.8780487804878055e-06, - "loss": 0.6467, + "loss": 0.6203, "step": 15 }, { "epoch": 0.06481012658227848, - "grad_norm": 2.6741785799011373, + "grad_norm": 2.460478449014551, "learning_rate": 5.203252032520326e-06, - "loss": 0.5994, + "loss": 0.5746, "step": 16 }, { "epoch": 0.06886075949367089, - "grad_norm": 2.5785520997615627, + "grad_norm": 2.3493430269108955, "learning_rate": 5.528455284552846e-06, - "loss": 0.6003, + "loss": 0.5795, "step": 17 }, { "epoch": 0.0729113924050633, - "grad_norm": 2.0881611263160704, + "grad_norm": 1.9918845141289823, "learning_rate": 5.853658536585366e-06, - "loss": 0.5913, + "loss": 0.5698, "step": 18 }, { "epoch": 0.0769620253164557, - "grad_norm": 1.5525903844986657, + "grad_norm": 1.6489527830554676, "learning_rate": 6.178861788617887e-06, - "loss": 0.5824, + "loss": 0.5487, "step": 19 }, { "epoch": 0.0810126582278481, - "grad_norm": 1.4360944123723738, + "grad_norm": 1.572885949701176, "learning_rate": 6.504065040650407e-06, - "loss": 0.5594, + "loss": 0.5224, "step": 20 }, { "epoch": 0.08506329113924051, - "grad_norm": 1.460132279548988, + "grad_norm": 1.451712249307161, "learning_rate": 6.829268292682928e-06, - "loss": 0.5531, + "loss": 0.5247, "step": 21 }, { "epoch": 0.08911392405063291, - "grad_norm": 1.392918549337748, + "grad_norm": 1.4054656377299188, "learning_rate": 7.154471544715448e-06, - "loss": 0.5345, + "loss": 0.5041, "step": 22 }, { "epoch": 0.09316455696202532, - "grad_norm": 1.168615169671816, + "grad_norm": 1.2530723859668358, "learning_rate": 7.4796747967479676e-06, - "loss": 0.5411, + "loss": 0.5149, "step": 23 }, { "epoch": 0.09721518987341772, - "grad_norm": 0.9293588038073347, + "grad_norm": 1.0837888171188292, "learning_rate": 7.804878048780489e-06, - "loss": 0.5448, + "loss": 0.5091, "step": 24 }, { "epoch": 0.10126582278481013, - "grad_norm": 1.1134981947599991, + "grad_norm": 1.0771838844672532, "learning_rate": 8.130081300813009e-06, - "loss": 0.5166, + "loss": 0.4944, "step": 25 }, { "epoch": 0.10531645569620253, - "grad_norm": 0.9952500280969623, + "grad_norm": 0.9066662533989501, "learning_rate": 8.45528455284553e-06, - "loss": 0.4998, + "loss": 0.479, "step": 26 }, { "epoch": 0.10936708860759493, - "grad_norm": 0.834318647993323, + "grad_norm": 0.8248319063901174, "learning_rate": 8.78048780487805e-06, - "loss": 0.5087, + "loss": 0.4866, "step": 27 }, { "epoch": 0.11341772151898734, - "grad_norm": 0.8598633398172679, + "grad_norm": 0.8417065051344589, "learning_rate": 9.10569105691057e-06, - "loss": 0.5229, + "loss": 0.488, "step": 28 }, { "epoch": 0.11746835443037974, - "grad_norm": 0.7977296569097209, + "grad_norm": 0.7213171607490497, "learning_rate": 9.43089430894309e-06, - "loss": 0.5014, + "loss": 0.4747, "step": 29 }, { "epoch": 0.12151898734177215, - "grad_norm": 0.8547285655415781, + "grad_norm": 0.750611509191151, "learning_rate": 9.756097560975611e-06, - "loss": 0.4902, + "loss": 0.4627, "step": 30 }, { "epoch": 0.12556962025316457, - "grad_norm": 0.6350235178289126, + "grad_norm": 0.6754163606343294, "learning_rate": 1.008130081300813e-05, - "loss": 0.4839, + "loss": 0.4588, "step": 31 }, { "epoch": 0.12962025316455697, - "grad_norm": 0.6133724162075492, + "grad_norm": 0.6472243449527941, "learning_rate": 1.0406504065040652e-05, - "loss": 0.4891, + "loss": 0.4663, "step": 32 }, { "epoch": 0.13367088607594937, - "grad_norm": 0.7264612575454036, + "grad_norm": 0.6183284435767364, "learning_rate": 1.0731707317073172e-05, - "loss": 0.4861, + "loss": 0.4635, "step": 33 }, { "epoch": 0.13772151898734178, - "grad_norm": 0.6882910869808051, + "grad_norm": 0.6515160702962011, "learning_rate": 1.1056910569105692e-05, - "loss": 0.4797, + "loss": 0.4483, "step": 34 }, { "epoch": 0.14177215189873418, - "grad_norm": 0.5598383578067709, + "grad_norm": 0.5986807187059076, "learning_rate": 1.1382113821138213e-05, - "loss": 0.474, + "loss": 0.4393, "step": 35 }, { "epoch": 0.1458227848101266, - "grad_norm": 0.5796421133484014, + "grad_norm": 0.5715501569153342, "learning_rate": 1.1707317073170731e-05, - "loss": 0.4659, + "loss": 0.4402, "step": 36 }, { "epoch": 0.149873417721519, - "grad_norm": 0.6592194819641932, + "grad_norm": 0.5723777681221434, "learning_rate": 1.2032520325203254e-05, - "loss": 0.4654, + "loss": 0.4322, "step": 37 }, { "epoch": 0.1539240506329114, - "grad_norm": 0.5716217465076222, + "grad_norm": 0.5936945690667882, "learning_rate": 1.2357723577235774e-05, - "loss": 0.4635, + "loss": 0.4405, "step": 38 }, { "epoch": 0.1579746835443038, - "grad_norm": 0.6081953225329045, + "grad_norm": 1.569697710037054, "learning_rate": 1.2682926829268294e-05, - "loss": 0.4637, + "loss": 0.4337, "step": 39 }, { "epoch": 0.1620253164556962, - "grad_norm": 0.48814294005094233, + "grad_norm": 0.5830272144805612, "learning_rate": 1.3008130081300815e-05, - "loss": 0.4776, + "loss": 0.4425, "step": 40 }, { "epoch": 0.1660759493670886, - "grad_norm": 0.6692081131454619, + "grad_norm": 0.5025898006391891, "learning_rate": 1.3333333333333333e-05, - "loss": 0.4655, + "loss": 0.4403, "step": 41 }, { "epoch": 0.17012658227848101, - "grad_norm": 0.4853524170847199, + "grad_norm": 0.5187222659956625, "learning_rate": 1.3658536585365855e-05, - "loss": 0.4526, + "loss": 0.4255, "step": 42 }, { "epoch": 0.17417721518987342, - "grad_norm": 0.6224377496589653, + "grad_norm": 0.5094946140456181, "learning_rate": 1.3983739837398376e-05, - "loss": 0.4577, + "loss": 0.4284, "step": 43 }, { "epoch": 0.17822784810126582, - "grad_norm": 0.5230102899440867, + "grad_norm": 0.48752983370039016, "learning_rate": 1.4308943089430896e-05, - "loss": 0.4487, + "loss": 0.4248, "step": 44 }, { "epoch": 0.18227848101265823, - "grad_norm": 0.4810361679421865, + "grad_norm": 0.5625781112475009, "learning_rate": 1.4634146341463415e-05, - "loss": 0.4645, + "loss": 0.4347, "step": 45 }, { "epoch": 0.18632911392405063, - "grad_norm": 0.6425447642837916, + "grad_norm": 0.5130642860249767, "learning_rate": 1.4959349593495935e-05, - "loss": 0.4575, + "loss": 0.432, "step": 46 }, { "epoch": 0.19037974683544304, - "grad_norm": 0.5296793567671157, + "grad_norm": 0.559949497276752, "learning_rate": 1.528455284552846e-05, - "loss": 0.4544, + "loss": 0.4338, "step": 47 }, { "epoch": 0.19443037974683544, - "grad_norm": 0.45237005651384643, + "grad_norm": 0.47266595492741803, "learning_rate": 1.5609756097560978e-05, - "loss": 0.4386, + "loss": 0.4171, "step": 48 }, { "epoch": 0.19848101265822785, - "grad_norm": 0.6017336170425414, + "grad_norm": 0.4913030476775072, "learning_rate": 1.5934959349593496e-05, - "loss": 0.4609, + "loss": 0.4309, "step": 49 }, { "epoch": 0.20253164556962025, - "grad_norm": 0.5585041362329556, + "grad_norm": 0.47745991231812274, "learning_rate": 1.6260162601626018e-05, - "loss": 0.4511, + "loss": 0.4267, "step": 50 }, { "epoch": 0.20658227848101265, - "grad_norm": 0.5200220404307354, + "grad_norm": 0.5380315600361899, "learning_rate": 1.6585365853658537e-05, - "loss": 0.4328, + "loss": 0.4111, "step": 51 }, { "epoch": 0.21063291139240506, - "grad_norm": 0.49321697242246215, + "grad_norm": 0.5080920562171014, "learning_rate": 1.691056910569106e-05, - "loss": 0.4506, + "loss": 0.4201, "step": 52 }, { "epoch": 0.21468354430379746, - "grad_norm": 0.5751109695395125, + "grad_norm": 0.5128546367455661, "learning_rate": 1.7235772357723578e-05, - "loss": 0.4536, + "loss": 0.4278, "step": 53 }, { "epoch": 0.21873417721518987, - "grad_norm": 0.482220215925531, + "grad_norm": 0.5460161968046872, "learning_rate": 1.75609756097561e-05, - "loss": 0.4414, + "loss": 0.4199, "step": 54 }, { "epoch": 0.22278481012658227, - "grad_norm": 0.5773433130453555, + "grad_norm": 0.4745231340547568, "learning_rate": 1.788617886178862e-05, - "loss": 0.4233, + "loss": 0.4005, "step": 55 }, { "epoch": 0.22683544303797468, - "grad_norm": 0.4619573256063155, + "grad_norm": 0.5331283405788664, "learning_rate": 1.821138211382114e-05, - "loss": 0.4298, + "loss": 0.4025, "step": 56 }, { "epoch": 0.23088607594936708, - "grad_norm": 0.5663950007389541, + "grad_norm": 0.5344188550194445, "learning_rate": 1.8536585365853663e-05, - "loss": 0.4293, + "loss": 0.4074, "step": 57 }, { "epoch": 0.23493670886075949, - "grad_norm": 0.5050782889285461, + "grad_norm": 0.5651363932284691, "learning_rate": 1.886178861788618e-05, - "loss": 0.4365, + "loss": 0.4065, "step": 58 }, { "epoch": 0.2389873417721519, - "grad_norm": 0.5318743872839863, + "grad_norm": 0.5020133690315554, "learning_rate": 1.91869918699187e-05, - "loss": 0.4354, + "loss": 0.4058, "step": 59 }, { "epoch": 0.2430379746835443, - "grad_norm": 0.47296928267629484, + "grad_norm": 0.5536966308046732, "learning_rate": 1.9512195121951222e-05, - "loss": 0.4139, + "loss": 0.3977, "step": 60 }, { "epoch": 0.2470886075949367, - "grad_norm": 0.4320505993647634, + "grad_norm": 0.6457517735601038, "learning_rate": 1.983739837398374e-05, - "loss": 0.442, + "loss": 0.4089, "step": 61 }, { "epoch": 0.25113924050632913, - "grad_norm": 0.515376135721367, + "grad_norm": 0.5675906894408289, "learning_rate": 2.016260162601626e-05, - "loss": 0.4348, + "loss": 0.4093, "step": 62 }, { "epoch": 0.25518987341772154, - "grad_norm": 0.5004976889998584, + "grad_norm": 0.6612100414239813, "learning_rate": 2.048780487804878e-05, - "loss": 0.4373, + "loss": 0.4114, "step": 63 }, { "epoch": 0.25924050632911394, - "grad_norm": 0.49874326272079256, + "grad_norm": 0.5300873182013642, "learning_rate": 2.0813008130081303e-05, - "loss": 0.4398, + "loss": 0.4149, "step": 64 }, { "epoch": 0.26329113924050634, - "grad_norm": 0.5458443577827394, + "grad_norm": 0.7085856534538018, "learning_rate": 2.1138211382113822e-05, - "loss": 0.44, + "loss": 0.4163, "step": 65 }, { "epoch": 0.26734177215189875, - "grad_norm": 0.5804403681542966, + "grad_norm": 0.5413408512162725, "learning_rate": 2.1463414634146344e-05, - "loss": 0.4064, + "loss": 0.3785, "step": 66 }, { "epoch": 0.27139240506329115, - "grad_norm": 0.5294927746347444, + "grad_norm": 0.6114204625352642, "learning_rate": 2.1788617886178863e-05, - "loss": 0.4253, + "loss": 0.3987, "step": 67 }, { "epoch": 0.27544303797468356, - "grad_norm": 0.5797309355092756, + "grad_norm": 0.7978747283598397, "learning_rate": 2.2113821138211385e-05, - "loss": 0.4221, + "loss": 0.3976, "step": 68 }, { "epoch": 0.27949367088607596, - "grad_norm": 0.6600258546127993, + "grad_norm": 0.6219038558687726, "learning_rate": 2.2439024390243907e-05, - "loss": 0.4144, + "loss": 0.3942, "step": 69 }, { "epoch": 0.28354430379746837, - "grad_norm": 0.6358653004829895, + "grad_norm": 0.8831256567553301, "learning_rate": 2.2764227642276426e-05, - "loss": 0.4403, + "loss": 0.4062, "step": 70 }, { "epoch": 0.28759493670886077, - "grad_norm": 0.831018237799899, + "grad_norm": 0.8701458499016784, "learning_rate": 2.3089430894308948e-05, - "loss": 0.424, + "loss": 0.4017, "step": 71 }, { "epoch": 0.2916455696202532, - "grad_norm": 0.5949566206188223, + "grad_norm": 0.6423135673887835, "learning_rate": 2.3414634146341463e-05, - "loss": 0.398, + "loss": 0.3836, "step": 72 }, { "epoch": 0.2956962025316456, - "grad_norm": 0.7301013159590914, + "grad_norm": 0.7863473594259498, "learning_rate": 2.3739837398373985e-05, - "loss": 0.4277, + "loss": 0.3998, "step": 73 }, { "epoch": 0.299746835443038, - "grad_norm": 0.7305696540756478, + "grad_norm": 0.6856234082896128, "learning_rate": 2.4065040650406507e-05, - "loss": 0.409, + "loss": 0.3802, "step": 74 }, { "epoch": 0.3037974683544304, - "grad_norm": 0.5893960794385738, + "grad_norm": 0.6427282075061846, "learning_rate": 2.4390243902439026e-05, - "loss": 0.4294, + "loss": 0.4136, "step": 75 }, { "epoch": 0.3078481012658228, - "grad_norm": 0.6385670316947752, + "grad_norm": 0.763928377690798, "learning_rate": 2.4715447154471548e-05, - "loss": 0.4283, + "loss": 0.3932, "step": 76 }, { "epoch": 0.3118987341772152, - "grad_norm": 0.6121367243758533, + "grad_norm": 0.6659477877128168, "learning_rate": 2.5040650406504066e-05, - "loss": 0.4068, + "loss": 0.3767, "step": 77 }, { "epoch": 0.3159493670886076, - "grad_norm": 0.6135387573309041, + "grad_norm": 0.652228378642887, "learning_rate": 2.536585365853659e-05, - "loss": 0.4195, + "loss": 0.4151, "step": 78 }, { "epoch": 0.32, - "grad_norm": 0.6097803364890213, + "grad_norm": 0.7116692577132699, "learning_rate": 2.569105691056911e-05, - "loss": 0.4078, + "loss": 0.3862, "step": 79 }, { "epoch": 0.3240506329113924, - "grad_norm": 0.523742778485504, + "grad_norm": 0.6139313448609756, "learning_rate": 2.601626016260163e-05, - "loss": 0.4233, + "loss": 0.4019, "step": 80 }, { "epoch": 0.3281012658227848, - "grad_norm": 0.5770979687000392, + "grad_norm": 0.7106418148735457, "learning_rate": 2.634146341463415e-05, - "loss": 0.4057, + "loss": 0.3816, "step": 81 }, { "epoch": 0.3321518987341772, - "grad_norm": 0.5165834717822592, + "grad_norm": 0.5455098402049849, "learning_rate": 2.6666666666666667e-05, - "loss": 0.4075, + "loss": 0.388, "step": 82 }, { "epoch": 0.3362025316455696, - "grad_norm": 0.514323455515994, + "grad_norm": 0.5833625742760852, "learning_rate": 2.699186991869919e-05, - "loss": 0.4094, + "loss": 0.3926, "step": 83 }, { "epoch": 0.34025316455696203, - "grad_norm": 0.6602713254140293, + "grad_norm": 0.6976430625248523, "learning_rate": 2.731707317073171e-05, - "loss": 0.4414, + "loss": 0.4148, "step": 84 }, { "epoch": 0.34430379746835443, - "grad_norm": 0.6407738263015262, + "grad_norm": 0.5631256669723129, "learning_rate": 2.764227642276423e-05, - "loss": 0.429, + "loss": 0.4038, "step": 85 }, { "epoch": 0.34835443037974684, - "grad_norm": 0.840427905323291, + "grad_norm": 0.6297177773650035, "learning_rate": 2.796747967479675e-05, - "loss": 0.4167, + "loss": 0.3814, "step": 86 }, { "epoch": 0.35240506329113924, - "grad_norm": 0.8577955051683515, + "grad_norm": 0.6530498529887747, "learning_rate": 2.829268292682927e-05, - "loss": 0.4248, + "loss": 0.3897, "step": 87 }, { "epoch": 0.35645569620253165, - "grad_norm": 0.8143906870677902, + "grad_norm": 0.660758390996932, "learning_rate": 2.8617886178861792e-05, - "loss": 0.4248, + "loss": 0.3971, "step": 88 }, { "epoch": 0.36050632911392405, - "grad_norm": 0.9012463223551264, + "grad_norm": 0.6432981711779486, "learning_rate": 2.8943089430894314e-05, - "loss": 0.4338, + "loss": 0.4082, "step": 89 }, { "epoch": 0.36455696202531646, - "grad_norm": 0.8014605081651359, + "grad_norm": 0.6769527644089639, "learning_rate": 2.926829268292683e-05, - "loss": 0.4048, + "loss": 0.3858, "step": 90 }, { "epoch": 0.36860759493670886, - "grad_norm": 0.7926104778051809, + "grad_norm": 0.7317177578010697, "learning_rate": 2.959349593495935e-05, - "loss": 0.425, + "loss": 0.3942, "step": 91 }, { "epoch": 0.37265822784810126, - "grad_norm": 0.7945474729709876, + "grad_norm": 0.6561649562895105, "learning_rate": 2.991869918699187e-05, - "loss": 0.4325, + "loss": 0.4103, "step": 92 }, { "epoch": 0.37670886075949367, - "grad_norm": 0.6212860750732824, + "grad_norm": 0.6370332145085758, "learning_rate": 3.0243902439024392e-05, - "loss": 0.407, + "loss": 0.3763, "step": 93 }, { "epoch": 0.3807594936708861, - "grad_norm": 0.7818634140233277, + "grad_norm": 0.6994856265474081, "learning_rate": 3.056910569105692e-05, - "loss": 0.4025, + "loss": 0.379, "step": 94 }, { "epoch": 0.3848101265822785, - "grad_norm": 0.715830137232043, + "grad_norm": 0.7534532810506789, "learning_rate": 3.089430894308943e-05, - "loss": 0.4121, + "loss": 0.3967, "step": 95 }, { "epoch": 0.3888607594936709, - "grad_norm": 0.5959393845893642, + "grad_norm": 0.8013673825606735, "learning_rate": 3.1219512195121955e-05, - "loss": 0.4162, + "loss": 0.3974, "step": 96 }, { "epoch": 0.3929113924050633, - "grad_norm": 0.753759916950481, + "grad_norm": 0.6601186985117031, "learning_rate": 3.154471544715447e-05, - "loss": 0.4188, + "loss": 0.3932, "step": 97 }, { "epoch": 0.3969620253164557, - "grad_norm": 0.5779060891919074, + "grad_norm": 0.8816268253435977, "learning_rate": 3.186991869918699e-05, - "loss": 0.4005, + "loss": 0.389, "step": 98 }, { "epoch": 0.4010126582278481, - "grad_norm": 0.7395405947497233, + "grad_norm": 0.6502491759480651, "learning_rate": 3.2195121951219514e-05, - "loss": 0.4305, + "loss": 0.3931, "step": 99 }, { "epoch": 0.4050632911392405, - "grad_norm": 0.654624971420012, + "grad_norm": 0.6405538480362688, "learning_rate": 3.2520325203252037e-05, - "loss": 0.4079, + "loss": 0.3857, "step": 100 }, { "epoch": 0.4091139240506329, - "grad_norm": 0.5797198698537027, + "grad_norm": 0.7194466820059928, "learning_rate": 3.284552845528456e-05, - "loss": 0.429, + "loss": 0.4032, "step": 101 }, { "epoch": 0.4131645569620253, - "grad_norm": 0.5456432295188319, + "grad_norm": 0.5720845890175896, "learning_rate": 3.3170731707317074e-05, - "loss": 0.4182, + "loss": 0.3885, "step": 102 }, { "epoch": 0.4172151898734177, - "grad_norm": 0.5766319848636041, + "grad_norm": 0.6136400802482502, "learning_rate": 3.3495934959349596e-05, - "loss": 0.4075, + "loss": 0.3889, "step": 103 }, { "epoch": 0.4212658227848101, - "grad_norm": 0.58917293572846, + "grad_norm": 0.6483481006307124, "learning_rate": 3.382113821138212e-05, - "loss": 0.4283, + "loss": 0.4024, "step": 104 }, { "epoch": 0.4253164556962025, - "grad_norm": 0.6361081952910633, + "grad_norm": 0.7609866812145318, "learning_rate": 3.414634146341463e-05, - "loss": 0.4189, + "loss": 0.4102, "step": 105 }, { "epoch": 0.4293670886075949, - "grad_norm": 0.6290971722976698, + "grad_norm": 0.5910885647277584, "learning_rate": 3.4471544715447155e-05, - "loss": 0.402, + "loss": 0.3815, "step": 106 }, { "epoch": 0.43341772151898733, - "grad_norm": 0.7360732043700037, + "grad_norm": 0.7632752159812108, "learning_rate": 3.479674796747968e-05, - "loss": 0.4206, + "loss": 0.391, "step": 107 }, { "epoch": 0.43746835443037974, - "grad_norm": 0.5911821004970247, + "grad_norm": 0.6537989518615785, "learning_rate": 3.51219512195122e-05, - "loss": 0.4069, + "loss": 0.3857, "step": 108 }, { "epoch": 0.44151898734177214, - "grad_norm": 0.7390823321225359, + "grad_norm": 0.45132763454677677, "learning_rate": 3.544715447154472e-05, - "loss": 0.4104, + "loss": 0.3728, "step": 109 }, { "epoch": 0.44556962025316454, - "grad_norm": 0.6016659296545879, + "grad_norm": 0.5502432025399826, "learning_rate": 3.577235772357724e-05, - "loss": 0.4133, + "loss": 0.3875, "step": 110 }, { "epoch": 0.44962025316455695, - "grad_norm": 0.6178255804085709, + "grad_norm": 0.5482141122008402, "learning_rate": 3.609756097560976e-05, - "loss": 0.4065, + "loss": 0.3931, "step": 111 }, { "epoch": 0.45367088607594935, - "grad_norm": 0.5192777696411834, + "grad_norm": 0.5042579700603617, "learning_rate": 3.642276422764228e-05, - "loss": 0.412, + "loss": 0.3923, "step": 112 }, { "epoch": 0.45772151898734176, - "grad_norm": 0.6165596674272442, + "grad_norm": 0.48080252398491274, "learning_rate": 3.67479674796748e-05, - "loss": 0.4088, + "loss": 0.3939, "step": 113 }, { "epoch": 0.46177215189873416, - "grad_norm": 0.5398916538387943, + "grad_norm": 0.5819181191826066, "learning_rate": 3.7073170731707325e-05, - "loss": 0.4163, + "loss": 0.3922, "step": 114 }, { "epoch": 0.46582278481012657, - "grad_norm": 0.7416639418309549, + "grad_norm": 0.5161246323531038, "learning_rate": 3.739837398373984e-05, - "loss": 0.3936, + "loss": 0.3754, "step": 115 }, { "epoch": 0.46987341772151897, - "grad_norm": 0.48692806819532003, + "grad_norm": 0.5401719710859738, "learning_rate": 3.772357723577236e-05, - "loss": 0.4068, + "loss": 0.3877, "step": 116 }, { "epoch": 0.4739240506329114, - "grad_norm": 0.7804677886805261, + "grad_norm": 0.5850809539882289, "learning_rate": 3.804878048780488e-05, - "loss": 0.4199, + "loss": 0.3942, "step": 117 }, { "epoch": 0.4779746835443038, - "grad_norm": 0.6913214328048365, + "grad_norm": 0.5375210050579635, "learning_rate": 3.83739837398374e-05, - "loss": 0.3902, + "loss": 0.3701, "step": 118 }, { "epoch": 0.4820253164556962, - "grad_norm": 0.6432278862562701, + "grad_norm": 0.5704604164558655, "learning_rate": 3.869918699186992e-05, - "loss": 0.4062, + "loss": 0.3886, "step": 119 }, { "epoch": 0.4860759493670886, - "grad_norm": 0.9093998639224735, + "grad_norm": 0.6367835721745447, "learning_rate": 3.9024390243902444e-05, - "loss": 0.4165, + "loss": 0.3838, "step": 120 }, { "epoch": 0.490126582278481, - "grad_norm": 0.8970317436846121, + "grad_norm": 0.6276071899311296, "learning_rate": 3.9349593495934966e-05, - "loss": 0.3938, + "loss": 0.3719, "step": 121 }, { "epoch": 0.4941772151898734, - "grad_norm": 0.637751173675504, + "grad_norm": 0.5100499219774005, "learning_rate": 3.967479674796748e-05, - "loss": 0.4183, + "loss": 0.3917, "step": 122 }, { "epoch": 0.4982278481012658, - "grad_norm": 0.7695270156232722, + "grad_norm": 0.621723133019605, "learning_rate": 4e-05, - "loss": 0.4117, + "loss": 0.3955, "step": 123 }, { "epoch": 0.5022784810126583, - "grad_norm": 0.7317943550318654, + "grad_norm": 0.5629741730086855, "learning_rate": 3.999991946137476e-05, - "loss": 0.425, + "loss": 0.3922, "step": 124 }, { "epoch": 0.5063291139240507, - "grad_norm": 0.5987216777789144, + "grad_norm": 0.5406316246600198, "learning_rate": 3.999967784614766e-05, - "loss": 0.3816, + "loss": 0.3606, "step": 125 }, { "epoch": 0.5103797468354431, - "grad_norm": 0.6286702544056411, + "grad_norm": 0.5600740792620695, "learning_rate": 3.9999275156264656e-05, - "loss": 0.4014, + "loss": 0.386, "step": 126 }, { "epoch": 0.5144303797468355, - "grad_norm": 0.6730040549048478, + "grad_norm": 0.6258097987727942, "learning_rate": 3.999871139496895e-05, - "loss": 0.4057, + "loss": 0.3862, "step": 127 }, { "epoch": 0.5184810126582279, - "grad_norm": 0.7750109349487608, + "grad_norm": 0.511933762069151, "learning_rate": 3.9997986566800995e-05, - "loss": 0.4186, + "loss": 0.3948, "step": 128 }, { "epoch": 0.5225316455696203, - "grad_norm": 0.6959243199838709, + "grad_norm": 0.504332229912149, "learning_rate": 3.999710067759846e-05, - "loss": 0.4165, + "loss": 0.3932, "step": 129 }, { "epoch": 0.5265822784810127, - "grad_norm": 0.7194950776750829, + "grad_norm": 0.5105578811506033, "learning_rate": 3.999605373449617e-05, - "loss": 0.4322, + "loss": 0.4006, "step": 130 }, { "epoch": 0.5306329113924051, - "grad_norm": 0.8215692889290367, + "grad_norm": 0.5205593904868228, "learning_rate": 3.9994845745926075e-05, - "loss": 0.4099, + "loss": 0.3891, "step": 131 }, { "epoch": 0.5346835443037975, - "grad_norm": 0.8426562265208263, + "grad_norm": 0.6152102208120602, "learning_rate": 3.999347672161713e-05, - "loss": 0.403, + "loss": 0.3821, "step": 132 }, { "epoch": 0.5387341772151899, - "grad_norm": 0.5953439601178845, + "grad_norm": 0.5205473151605777, "learning_rate": 3.999194667259528e-05, - "loss": 0.3998, + "loss": 0.3855, "step": 133 }, { "epoch": 0.5427848101265823, - "grad_norm": 0.7020343178627786, + "grad_norm": 0.5838400985938534, "learning_rate": 3.999025561118334e-05, - "loss": 0.4259, + "loss": 0.393, "step": 134 }, { "epoch": 0.5468354430379747, - "grad_norm": 0.6432094482318282, + "grad_norm": 0.5605239709503395, "learning_rate": 3.998840355100086e-05, - "loss": 0.4088, + "loss": 0.3789, "step": 135 }, { "epoch": 0.5508860759493671, - "grad_norm": 0.5789807304963224, + "grad_norm": 0.49019950538882867, "learning_rate": 3.998639050696409e-05, - "loss": 0.4097, + "loss": 0.3847, "step": 136 }, { "epoch": 0.5549367088607595, - "grad_norm": 0.5966375485064512, + "grad_norm": 0.4548532481780974, "learning_rate": 3.998421649528582e-05, - "loss": 0.3985, + "loss": 0.3844, "step": 137 }, { "epoch": 0.5589873417721519, - "grad_norm": 0.52106619560223, + "grad_norm": 0.5261329592085209, "learning_rate": 3.9981881533475234e-05, - "loss": 0.4042, + "loss": 0.3825, "step": 138 }, { "epoch": 0.5630379746835443, - "grad_norm": 0.5619602508541054, + "grad_norm": 0.43722984117577496, "learning_rate": 3.997938564033779e-05, - "loss": 0.3993, + "loss": 0.3819, "step": 139 }, { "epoch": 0.5670886075949367, - "grad_norm": 0.5581802137457483, + "grad_norm": 0.5330799402756228, "learning_rate": 3.9976728835975064e-05, - "loss": 0.4053, + "loss": 0.3916, "step": 140 }, { "epoch": 0.5711392405063291, - "grad_norm": 0.6188175993465432, + "grad_norm": 0.5565583472937724, "learning_rate": 3.9973911141784605e-05, - "loss": 0.3957, + "loss": 0.3739, "step": 141 }, { "epoch": 0.5751898734177215, - "grad_norm": 0.5078857649358703, + "grad_norm": 0.6079760230334288, "learning_rate": 3.997093258045973e-05, - "loss": 0.3987, + "loss": 0.3803, "step": 142 }, { "epoch": 0.579240506329114, - "grad_norm": 0.7175749382347821, + "grad_norm": 0.494759642886606, "learning_rate": 3.996779317598936e-05, - "loss": 0.4158, + "loss": 0.3907, "step": 143 }, { "epoch": 0.5832911392405064, - "grad_norm": 0.5091440945960467, + "grad_norm": 0.5946891150411486, "learning_rate": 3.996449295365782e-05, - "loss": 0.4142, + "loss": 0.3905, "step": 144 }, { "epoch": 0.5873417721518988, - "grad_norm": 0.5563233038325771, + "grad_norm": 0.4412945363216492, "learning_rate": 3.996103194004467e-05, - "loss": 0.3978, + "loss": 0.3808, "step": 145 }, { "epoch": 0.5913924050632912, - "grad_norm": 0.5493139913445463, + "grad_norm": 0.5901528370321416, "learning_rate": 3.995741016302441e-05, - "loss": 0.4089, + "loss": 0.3897, "step": 146 }, { "epoch": 0.5954430379746836, - "grad_norm": 0.5701346424023386, + "grad_norm": 0.4899374115359715, "learning_rate": 3.9953627651766364e-05, - "loss": 0.3923, + "loss": 0.3727, "step": 147 }, { "epoch": 0.599493670886076, - "grad_norm": 0.5622567557232179, + "grad_norm": 0.6123773529216578, "learning_rate": 3.9949684436734325e-05, - "loss": 0.3878, + "loss": 0.3619, "step": 148 }, { "epoch": 0.6035443037974684, - "grad_norm": 0.7838950980499242, + "grad_norm": 0.4853453351537688, "learning_rate": 3.994558054968643e-05, - "loss": 0.4014, + "loss": 0.3732, "step": 149 }, { "epoch": 0.6075949367088608, - "grad_norm": 0.481810229817515, + "grad_norm": 0.4977697463339369, "learning_rate": 3.994131602367481e-05, - "loss": 0.3956, + "loss": 0.3752, "step": 150 }, { "epoch": 0.6116455696202532, - "grad_norm": 0.587645431284674, + "grad_norm": 0.5414166984338413, "learning_rate": 3.9936890893045376e-05, - "loss": 0.3983, + "loss": 0.3744, "step": 151 }, { "epoch": 0.6156962025316456, - "grad_norm": 0.621427847967471, + "grad_norm": 0.48749139915438855, "learning_rate": 3.993230519343752e-05, - "loss": 0.4265, + "loss": 0.3971, "step": 152 }, { "epoch": 0.619746835443038, - "grad_norm": 0.5619957639806804, + "grad_norm": 0.5976258378379328, "learning_rate": 3.992755896178383e-05, - "loss": 0.4037, + "loss": 0.3826, "step": 153 }, { "epoch": 0.6237974683544304, - "grad_norm": 0.4601235550396703, + "grad_norm": 0.42523904094458825, "learning_rate": 3.992265223630981e-05, - "loss": 0.3899, + "loss": 0.3687, "step": 154 }, { "epoch": 0.6278481012658228, - "grad_norm": 0.5599801109734166, + "grad_norm": 0.528902294404794, "learning_rate": 3.991758505653355e-05, - "loss": 0.4184, + "loss": 0.3967, "step": 155 }, { "epoch": 0.6318987341772152, - "grad_norm": 0.48716310257237083, + "grad_norm": 0.48770937311698753, "learning_rate": 3.991235746326543e-05, - "loss": 0.4022, + "loss": 0.3815, "step": 156 }, { "epoch": 0.6359493670886076, - "grad_norm": 0.3701750847308048, + "grad_norm": 0.445358901306249, "learning_rate": 3.9906969498607745e-05, - "loss": 0.3827, + "loss": 0.3717, "step": 157 }, { "epoch": 0.64, - "grad_norm": 0.4161570673769205, + "grad_norm": 0.5388773039788232, "learning_rate": 3.990142120595444e-05, - "loss": 0.405, + "loss": 0.382, "step": 158 }, { "epoch": 0.6440506329113924, - "grad_norm": 0.5400900975014964, + "grad_norm": 0.4974551837610294, "learning_rate": 3.98957126299907e-05, - "loss": 0.3993, + "loss": 0.3763, "step": 159 }, { "epoch": 0.6481012658227848, - "grad_norm": 0.5090251491537814, + "grad_norm": 0.48145696573304353, "learning_rate": 3.9889843816692596e-05, - "loss": 0.403, + "loss": 0.3885, "step": 160 }, { "epoch": 0.6521518987341772, - "grad_norm": 0.4829764109881706, + "grad_norm": 0.4852697094770394, "learning_rate": 3.9883814813326766e-05, - "loss": 0.4004, + "loss": 0.3857, "step": 161 }, { "epoch": 0.6562025316455696, - "grad_norm": 0.6668089082489538, + "grad_norm": 0.46635533311708754, "learning_rate": 3.9877625668449956e-05, - "loss": 0.411, + "loss": 0.3776, "step": 162 }, { "epoch": 0.660253164556962, - "grad_norm": 0.5966344494900105, + "grad_norm": 0.5151311130783601, "learning_rate": 3.98712764319087e-05, - "loss": 0.4083, + "loss": 0.3804, "step": 163 }, { "epoch": 0.6643037974683544, - "grad_norm": 0.4835981514753116, + "grad_norm": 0.4264588202621475, "learning_rate": 3.9864767154838864e-05, - "loss": 0.391, + "loss": 0.384, "step": 164 }, { "epoch": 0.6683544303797468, - "grad_norm": 0.5807471314012177, + "grad_norm": 0.46551804296424865, "learning_rate": 3.9858097889665277e-05, - "loss": 0.3914, + "loss": 0.3713, "step": 165 }, { "epoch": 0.6724050632911392, - "grad_norm": 0.4920868095455476, + "grad_norm": 0.4744091958623959, "learning_rate": 3.985126869010129e-05, - "loss": 0.3946, + "loss": 0.3803, "step": 166 }, { "epoch": 0.6764556962025317, - "grad_norm": 0.5759238260735862, + "grad_norm": 0.4732573885033216, "learning_rate": 3.984427961114833e-05, - "loss": 0.394, + "loss": 0.3665, "step": 167 }, { "epoch": 0.6805063291139241, - "grad_norm": 0.4739984995986994, + "grad_norm": 0.4655474646655494, "learning_rate": 3.9837130709095475e-05, - "loss": 0.3947, + "loss": 0.3677, "step": 168 }, { "epoch": 0.6845569620253165, - "grad_norm": 0.4250149821655739, + "grad_norm": 0.498662203183722, "learning_rate": 3.982982204151901e-05, - "loss": 0.4095, + "loss": 0.3801, "step": 169 }, { "epoch": 0.6886075949367089, - "grad_norm": 0.4623058033175209, + "grad_norm": 0.4548444564791788, "learning_rate": 3.982235366728193e-05, - "loss": 0.394, + "loss": 0.3714, "step": 170 }, { "epoch": 0.6926582278481013, - "grad_norm": 0.5213039627437589, + "grad_norm": 0.5674550155900002, "learning_rate": 3.9814725646533505e-05, - "loss": 0.3994, + "loss": 0.373, "step": 171 }, { "epoch": 0.6967088607594937, - "grad_norm": 0.47394032388905616, + "grad_norm": 0.4568990388963496, "learning_rate": 3.9806938040708746e-05, - "loss": 0.3734, + "loss": 0.3545, "step": 172 }, { "epoch": 0.7007594936708861, - "grad_norm": 0.49046945165538175, + "grad_norm": 0.4854290944250977, "learning_rate": 3.9798990912527976e-05, - "loss": 0.3987, + "loss": 0.3664, "step": 173 }, { "epoch": 0.7048101265822785, - "grad_norm": 0.5849537790730299, + "grad_norm": 0.5068358545778741, "learning_rate": 3.979088432599627e-05, - "loss": 0.3991, + "loss": 0.3791, "step": 174 }, { "epoch": 0.7088607594936709, - "grad_norm": 0.41067896529263853, + "grad_norm": 0.44246109952716284, "learning_rate": 3.9782618346402964e-05, - "loss": 0.413, + "loss": 0.3971, "step": 175 }, { "epoch": 0.7129113924050633, - "grad_norm": 0.6719791969738832, + "grad_norm": 0.5105198118923095, "learning_rate": 3.977419304032111e-05, - "loss": 0.3895, + "loss": 0.3769, "step": 176 }, { "epoch": 0.7169620253164557, - "grad_norm": 0.4932442355971398, + "grad_norm": 0.4504255417270314, "learning_rate": 3.976560847560697e-05, - "loss": 0.4023, + "loss": 0.3839, "step": 177 }, { "epoch": 0.7210126582278481, - "grad_norm": 0.4293708014495667, + "grad_norm": 0.47083549172542233, "learning_rate": 3.9756864721399456e-05, - "loss": 0.3997, + "loss": 0.3825, "step": 178 }, { "epoch": 0.7250632911392405, - "grad_norm": 0.4528462207947421, + "grad_norm": 0.4938812526861007, "learning_rate": 3.974796184811956e-05, - "loss": 0.3923, + "loss": 0.3655, "step": 179 }, { "epoch": 0.7291139240506329, - "grad_norm": 0.4519739740804986, + "grad_norm": 0.6121123595102624, "learning_rate": 3.973889992746979e-05, - "loss": 0.4048, + "loss": 0.3867, "step": 180 }, { "epoch": 0.7331645569620253, - "grad_norm": 0.47376669948725775, + "grad_norm": 0.3935469525704463, "learning_rate": 3.972967903243361e-05, - "loss": 0.3995, + "loss": 0.3812, "step": 181 }, { "epoch": 0.7372151898734177, - "grad_norm": 0.3721711883495072, + "grad_norm": 0.5665964439606438, "learning_rate": 3.972029923727486e-05, - "loss": 0.3984, + "loss": 0.3835, "step": 182 }, { "epoch": 0.7412658227848101, - "grad_norm": 0.416832352950399, + "grad_norm": 0.40364607683557086, "learning_rate": 3.971076061753709e-05, - "loss": 0.3958, + "loss": 0.3605, "step": 183 }, { "epoch": 0.7453164556962025, - "grad_norm": 0.3455581901414806, + "grad_norm": 0.5981224928226465, "learning_rate": 3.9701063250043066e-05, - "loss": 0.3821, + "loss": 0.3588, "step": 184 }, { "epoch": 0.7493670886075949, - "grad_norm": 0.4636177824417175, + "grad_norm": 0.3856239194304908, "learning_rate": 3.969120721289402e-05, - "loss": 0.3861, + "loss": 0.3734, "step": 185 }, { "epoch": 0.7534177215189873, - "grad_norm": 0.3718257716590197, + "grad_norm": 0.5908369710632956, "learning_rate": 3.9681192585469146e-05, - "loss": 0.3868, + "loss": 0.3685, "step": 186 }, { "epoch": 0.7574683544303797, - "grad_norm": 0.3747485055944841, + "grad_norm": 0.3894179164448138, "learning_rate": 3.9671019448424865e-05, - "loss": 0.3729, + "loss": 0.3593, "step": 187 }, { "epoch": 0.7615189873417721, - "grad_norm": 0.4614708003377788, + "grad_norm": 0.5215780390811954, "learning_rate": 3.966068788369422e-05, - "loss": 0.4069, + "loss": 0.3824, "step": 188 }, { "epoch": 0.7655696202531646, - "grad_norm": 0.40847100811106896, + "grad_norm": 0.4154544733655475, "learning_rate": 3.965019797448622e-05, - "loss": 0.3998, + "loss": 0.3722, "step": 189 }, { "epoch": 0.769620253164557, - "grad_norm": 0.45898575730439506, + "grad_norm": 0.48384942940680387, "learning_rate": 3.963954980528515e-05, - "loss": 0.3968, + "loss": 0.371, "step": 190 }, { "epoch": 0.7736708860759494, - "grad_norm": 0.4259869936695245, + "grad_norm": 0.4151020053728562, "learning_rate": 3.9628743461849905e-05, - "loss": 0.3913, + "loss": 0.3628, "step": 191 }, { "epoch": 0.7777215189873418, - "grad_norm": 0.46795812250649077, + "grad_norm": 0.5605697944605167, "learning_rate": 3.961777903121329e-05, - "loss": 0.3937, + "loss": 0.3725, "step": 192 }, { "epoch": 0.7817721518987342, - "grad_norm": 0.41779176950798863, + "grad_norm": 0.43406064369137337, "learning_rate": 3.960665660168131e-05, - "loss": 0.4003, + "loss": 0.3762, "step": 193 }, { "epoch": 0.7858227848101266, - "grad_norm": 0.40411804759614206, + "grad_norm": 0.4543925214390367, "learning_rate": 3.9595376262832485e-05, - "loss": 0.3953, + "loss": 0.3716, "step": 194 }, { "epoch": 0.789873417721519, - "grad_norm": 0.39504255930540255, + "grad_norm": 0.42761526807896044, "learning_rate": 3.9583938105517127e-05, - "loss": 0.3878, + "loss": 0.3656, "step": 195 }, { "epoch": 0.7939240506329114, - "grad_norm": 0.4298612884396961, + "grad_norm": 0.425195946163417, "learning_rate": 3.957234222185657e-05, - "loss": 0.402, + "loss": 0.3685, "step": 196 }, { "epoch": 0.7979746835443038, - "grad_norm": 0.4424061234631262, + "grad_norm": 0.4620078808266855, "learning_rate": 3.9560588705242474e-05, - "loss": 0.3898, + "loss": 0.3658, "step": 197 }, { "epoch": 0.8020253164556962, - "grad_norm": 0.4303080944053748, + "grad_norm": 0.5274449762131294, "learning_rate": 3.954867765033605e-05, - "loss": 0.3854, + "loss": 0.3596, "step": 198 }, { "epoch": 0.8060759493670886, - "grad_norm": 0.4448263630713307, + "grad_norm": 0.37957278413317347, "learning_rate": 3.953660915306728e-05, - "loss": 0.3722, + "loss": 0.3479, "step": 199 }, { "epoch": 0.810126582278481, - "grad_norm": 0.3939579336970092, + "grad_norm": 0.4799711325269261, "learning_rate": 3.952438331063419e-05, - "loss": 0.3754, + "loss": 0.3669, "step": 200 }, { "epoch": 0.8141772151898734, - "grad_norm": 0.43924733624335843, + "grad_norm": 0.4514457644026668, "learning_rate": 3.951200022150205e-05, - "loss": 0.3953, + "loss": 0.3726, "step": 201 }, { "epoch": 0.8182278481012658, - "grad_norm": 0.4332047834777511, + "grad_norm": 0.39922943671244554, "learning_rate": 3.949945998540253e-05, - "loss": 0.3771, + "loss": 0.36, "step": 202 }, { "epoch": 0.8222784810126582, - "grad_norm": 0.42852471851423357, + "grad_norm": 0.5341218610008884, "learning_rate": 3.9486762703332993e-05, - "loss": 0.3677, + "loss": 0.3626, "step": 203 }, { "epoch": 0.8263291139240506, - "grad_norm": 0.4363581624596705, + "grad_norm": 0.39994061049090884, "learning_rate": 3.947390847755559e-05, - "loss": 0.3811, + "loss": 0.3616, "step": 204 }, { "epoch": 0.830379746835443, - "grad_norm": 0.41853531859988113, + "grad_norm": 0.5087763598480363, "learning_rate": 3.946089741159648e-05, - "loss": 0.3889, + "loss": 0.3723, "step": 205 }, { "epoch": 0.8344303797468354, - "grad_norm": 0.4872328168565743, + "grad_norm": 0.571082065233194, "learning_rate": 3.944772961024501e-05, - "loss": 0.3905, + "loss": 0.3647, "step": 206 }, { "epoch": 0.8384810126582278, - "grad_norm": 0.48135401668668604, + "grad_norm": 0.40769972678351213, "learning_rate": 3.943440517955285e-05, - "loss": 0.3847, + "loss": 0.3635, "step": 207 }, { "epoch": 0.8425316455696202, - "grad_norm": 0.40178319750932817, + "grad_norm": 0.49410582685700866, "learning_rate": 3.9420924226833126e-05, - "loss": 0.3849, + "loss": 0.3645, "step": 208 }, { "epoch": 0.8465822784810126, - "grad_norm": 0.467274374993188, + "grad_norm": 0.4387525444885117, "learning_rate": 3.9407286860659566e-05, - "loss": 0.3956, + "loss": 0.3598, "step": 209 }, { "epoch": 0.850632911392405, - "grad_norm": 0.378615821088885, + "grad_norm": 0.44273304130244956, "learning_rate": 3.9393493190865657e-05, - "loss": 0.379, + "loss": 0.3611, "step": 210 }, { "epoch": 0.8546835443037974, - "grad_norm": 0.42738525611803163, + "grad_norm": 0.43175110862933125, "learning_rate": 3.937954332854371e-05, - "loss": 0.3804, + "loss": 0.3584, "step": 211 }, { "epoch": 0.8587341772151899, - "grad_norm": 0.4718498749896371, + "grad_norm": 0.4564656033730629, "learning_rate": 3.9365437386044016e-05, - "loss": 0.3999, + "loss": 0.371, "step": 212 }, { "epoch": 0.8627848101265823, - "grad_norm": 0.4378766261082102, + "grad_norm": 0.40961026413253804, "learning_rate": 3.935117547697387e-05, - "loss": 0.3872, + "loss": 0.3767, "step": 213 }, { "epoch": 0.8668354430379747, - "grad_norm": 0.46092131018072824, + "grad_norm": 0.4406667742487598, "learning_rate": 3.933675771619675e-05, - "loss": 0.3946, + "loss": 0.3724, "step": 214 }, { "epoch": 0.8708860759493671, - "grad_norm": 0.4254142306164415, + "grad_norm": 0.4322305722373633, "learning_rate": 3.932218421983131e-05, - "loss": 0.3998, + "loss": 0.3745, "step": 215 }, { "epoch": 0.8749367088607595, - "grad_norm": 0.4537959753487451, + "grad_norm": 0.39692139508630897, "learning_rate": 3.9307455105250484e-05, - "loss": 0.4034, + "loss": 0.3792, "step": 216 }, { "epoch": 0.8789873417721519, - "grad_norm": 0.40849112757029404, + "grad_norm": 0.39211956930153335, "learning_rate": 3.929257049108054e-05, - "loss": 0.3862, + "loss": 0.3592, "step": 217 }, { "epoch": 0.8830379746835443, - "grad_norm": 0.4816020330539227, + "grad_norm": 0.4168765456638921, "learning_rate": 3.927753049720011e-05, - "loss": 0.3846, + "loss": 0.3661, "step": 218 }, { "epoch": 0.8870886075949367, - "grad_norm": 0.4136974287677684, + "grad_norm": 0.4261728514980525, "learning_rate": 3.9262335244739234e-05, - "loss": 0.3763, + "loss": 0.3611, "step": 219 }, { "epoch": 0.8911392405063291, - "grad_norm": 0.4021135273076994, + "grad_norm": 0.42900491455321205, "learning_rate": 3.92469848560784e-05, - "loss": 0.3746, + "loss": 0.3561, "step": 220 }, { "epoch": 0.8951898734177215, - "grad_norm": 0.45585194083479996, + "grad_norm": 0.46058162414644876, "learning_rate": 3.923147945484751e-05, - "loss": 0.3957, + "loss": 0.3694, "step": 221 }, { "epoch": 0.8992405063291139, - "grad_norm": 0.40482649957828903, + "grad_norm": 0.44165633680367516, "learning_rate": 3.9215819165924956e-05, - "loss": 0.3777, + "loss": 0.3549, "step": 222 }, { "epoch": 0.9032911392405063, - "grad_norm": 0.525496205751538, + "grad_norm": 0.5460218821199697, "learning_rate": 3.920000411543654e-05, - "loss": 0.3844, + "loss": 0.3731, "step": 223 }, { "epoch": 0.9073417721518987, - "grad_norm": 0.43820945156493263, + "grad_norm": 0.4938065359019653, "learning_rate": 3.9184034430754495e-05, - "loss": 0.3727, + "loss": 0.3505, "step": 224 }, { "epoch": 0.9113924050632911, - "grad_norm": 0.47776961254005973, + "grad_norm": 0.4508355784135828, "learning_rate": 3.916791024049648e-05, - "loss": 0.3752, + "loss": 0.3553, "step": 225 }, { "epoch": 0.9154430379746835, - "grad_norm": 0.478364896238862, + "grad_norm": 0.48359403997241646, "learning_rate": 3.91516316745245e-05, - "loss": 0.3794, + "loss": 0.3644, "step": 226 }, { "epoch": 0.9194936708860759, - "grad_norm": 0.5060336583228564, + "grad_norm": 0.544720421516036, "learning_rate": 3.913519886394389e-05, - "loss": 0.4071, + "loss": 0.3876, "step": 227 }, { "epoch": 0.9235443037974683, - "grad_norm": 0.46803759026595343, + "grad_norm": 0.4118678349885707, "learning_rate": 3.911861194110225e-05, - "loss": 0.3833, + "loss": 0.3713, "step": 228 }, { "epoch": 0.9275949367088607, - "grad_norm": 0.4689199143066036, + "grad_norm": 0.5112105119981865, "learning_rate": 3.910187103958837e-05, - "loss": 0.3925, + "loss": 0.3681, "step": 229 }, { "epoch": 0.9316455696202531, - "grad_norm": 0.42662769099154557, + "grad_norm": 0.4105441927327187, "learning_rate": 3.908497629423117e-05, - "loss": 0.3907, + "loss": 0.3715, "step": 230 }, { "epoch": 0.9356962025316455, - "grad_norm": 0.4546223554906684, + "grad_norm": 0.4644045076064998, "learning_rate": 3.9067927841098614e-05, - "loss": 0.392, + "loss": 0.3597, "step": 231 }, { "epoch": 0.9397468354430379, - "grad_norm": 0.39666956297455297, + "grad_norm": 0.3989283424653455, "learning_rate": 3.9050725817496594e-05, - "loss": 0.3779, + "loss": 0.355, "step": 232 }, { "epoch": 0.9437974683544303, - "grad_norm": 0.455823613228658, + "grad_norm": 0.45332897262360705, "learning_rate": 3.9033370361967844e-05, - "loss": 0.3954, + "loss": 0.3668, "step": 233 }, { "epoch": 0.9478481012658228, - "grad_norm": 0.45474813856454466, + "grad_norm": 0.41886635544985723, "learning_rate": 3.901586161429081e-05, - "loss": 0.3912, + "loss": 0.3713, "step": 234 }, { "epoch": 0.9518987341772152, - "grad_norm": 0.5135615538127752, + "grad_norm": 0.40715278318792036, "learning_rate": 3.8998199715478545e-05, - "loss": 0.3813, + "loss": 0.3625, "step": 235 }, { "epoch": 0.9559493670886076, - "grad_norm": 0.47940427295247023, + "grad_norm": 0.4729078180980691, "learning_rate": 3.8980384807777564e-05, - "loss": 0.3802, + "loss": 0.36, "step": 236 }, { "epoch": 0.96, - "grad_norm": 0.44046844955213627, + "grad_norm": 0.4102015941268841, "learning_rate": 3.896241703466667e-05, - "loss": 0.3974, + "loss": 0.3814, "step": 237 }, { "epoch": 0.9640506329113924, - "grad_norm": 0.4667241418191417, + "grad_norm": 0.4305575571401939, "learning_rate": 3.894429654085585e-05, - "loss": 0.3957, + "loss": 0.3653, "step": 238 }, { "epoch": 0.9681012658227848, - "grad_norm": 0.40510812159755255, + "grad_norm": 0.433837803321305, "learning_rate": 3.892602347228505e-05, - "loss": 0.3966, + "loss": 0.3794, "step": 239 }, { "epoch": 0.9721518987341772, - "grad_norm": 0.42835989218395815, + "grad_norm": 0.43435638402327614, "learning_rate": 3.890759797612307e-05, - "loss": 0.3745, + "loss": 0.3502, "step": 240 }, { "epoch": 0.9762025316455696, - "grad_norm": 0.4188429309485113, + "grad_norm": 0.38073355766199574, "learning_rate": 3.888902020076632e-05, - "loss": 0.3911, + "loss": 0.3728, "step": 241 }, { "epoch": 0.980253164556962, - "grad_norm": 0.44206891521706126, + "grad_norm": 0.4242913532591757, "learning_rate": 3.887029029583764e-05, - "loss": 0.3869, + "loss": 0.36, "step": 242 }, { "epoch": 0.9843037974683544, - "grad_norm": 0.41570856902679265, + "grad_norm": 0.36490913133345676, "learning_rate": 3.8851408412185125e-05, - "loss": 0.3994, + "loss": 0.3746, "step": 243 }, { "epoch": 0.9883544303797468, - "grad_norm": 0.4281721436806353, + "grad_norm": 0.4115648846497593, "learning_rate": 3.8832374701880855e-05, - "loss": 0.3844, + "loss": 0.3732, "step": 244 }, { "epoch": 0.9924050632911392, - "grad_norm": 0.4232614593453341, + "grad_norm": 0.382223367664726, "learning_rate": 3.881318931821972e-05, - "loss": 0.3839, + "loss": 0.3631, "step": 245 }, { "epoch": 0.9964556962025316, - "grad_norm": 0.4818227012289353, + "grad_norm": 0.4269495405065219, "learning_rate": 3.879385241571817e-05, - "loss": 0.3893, + "loss": 0.3651, "step": 246 }, { "epoch": 1.000506329113924, - "grad_norm": 0.5476037889029386, + "grad_norm": 0.3857743605396771, "learning_rate": 3.8774364150112955e-05, - "loss": 0.4187, + "loss": 0.3439, "step": 247 }, { "epoch": 1.0045569620253165, - "grad_norm": 0.4571389615563016, + "grad_norm": 0.5182363861417547, "learning_rate": 3.8754724678359884e-05, - "loss": 0.3251, + "loss": 0.3144, "step": 248 }, { "epoch": 1.0086075949367088, - "grad_norm": 0.46823466820055304, + "grad_norm": 0.40471323867861103, "learning_rate": 3.873493415863256e-05, - "loss": 0.3362, + "loss": 0.3177, "step": 249 }, { "epoch": 1.0126582278481013, - "grad_norm": 0.46435337862494125, + "grad_norm": 0.5153235955639972, "learning_rate": 3.871499275032111e-05, - "loss": 0.3548, + "loss": 0.3223, "step": 250 }, { "epoch": 1.0167088607594936, - "grad_norm": 0.43414490739128436, + "grad_norm": 0.38156584074959976, "learning_rate": 3.869490061403091e-05, - "loss": 0.3269, + "loss": 0.2978, "step": 251 }, { "epoch": 1.0207594936708861, - "grad_norm": 0.3910397142833239, + "grad_norm": 0.4694601899432551, "learning_rate": 3.867465791158124e-05, - "loss": 0.3164, + "loss": 0.3117, "step": 252 }, { "epoch": 1.0248101265822784, - "grad_norm": 0.47364535731869795, + "grad_norm": 0.4472606457664077, "learning_rate": 3.865426480600407e-05, - "loss": 0.347, + "loss": 0.3052, "step": 253 }, { "epoch": 1.028860759493671, - "grad_norm": 0.4424228615731336, + "grad_norm": 0.4213830764206453, "learning_rate": 3.863372146154264e-05, - "loss": 0.3417, + "loss": 0.3055, "step": 254 }, { "epoch": 1.0329113924050632, - "grad_norm": 0.5284000077188254, + "grad_norm": 0.4520073084621344, "learning_rate": 3.861302804365024e-05, - "loss": 0.3487, + "loss": 0.3175, "step": 255 }, { "epoch": 1.0369620253164558, - "grad_norm": 0.40233649188530907, + "grad_norm": 0.39518451005870114, "learning_rate": 3.85921847189888e-05, - "loss": 0.3344, + "loss": 0.3141, "step": 256 }, { "epoch": 1.041012658227848, - "grad_norm": 0.5083760690559533, + "grad_norm": 0.4593623563226846, "learning_rate": 3.85711916554276e-05, - "loss": 0.3379, + "loss": 0.3093, "step": 257 }, { "epoch": 1.0450632911392406, - "grad_norm": 0.43532870949132046, + "grad_norm": 0.4548290497711387, "learning_rate": 3.85500490220419e-05, - "loss": 0.3309, + "loss": 0.3079, "step": 258 }, { "epoch": 1.0491139240506329, - "grad_norm": 0.5167161778541496, + "grad_norm": 0.4029758807372885, "learning_rate": 3.852875698911154e-05, - "loss": 0.3472, + "loss": 0.294, "step": 259 }, { "epoch": 1.0531645569620254, - "grad_norm": 0.48043788332596066, + "grad_norm": 0.44801762962474895, "learning_rate": 3.850731572811963e-05, - "loss": 0.3243, + "loss": 0.311, "step": 260 }, { "epoch": 1.0572151898734177, - "grad_norm": 0.5148059171374179, + "grad_norm": 0.42512004600221576, "learning_rate": 3.848572541175116e-05, - "loss": 0.3573, + "loss": 0.3213, "step": 261 }, { "epoch": 1.0612658227848102, - "grad_norm": 0.4722636498027736, + "grad_norm": 0.42043840299184576, "learning_rate": 3.846398621389154e-05, - "loss": 0.3489, + "loss": 0.3183, "step": 262 }, { "epoch": 1.0653164556962025, - "grad_norm": 0.46825262237470155, + "grad_norm": 0.3892962415472906, "learning_rate": 3.84420983096253e-05, - "loss": 0.3248, + "loss": 0.3051, "step": 263 }, { "epoch": 1.069367088607595, - "grad_norm": 0.4814150006967104, + "grad_norm": 0.42278867827373684, "learning_rate": 3.8420061875234606e-05, - "loss": 0.3405, + "loss": 0.3004, "step": 264 }, { "epoch": 1.0734177215189873, - "grad_norm": 0.47718846830732237, + "grad_norm": 0.3988295654730013, "learning_rate": 3.839787708819787e-05, - "loss": 0.3361, + "loss": 0.3154, "step": 265 }, { "epoch": 1.0774683544303798, - "grad_norm": 0.4859044231945241, + "grad_norm": 0.3992434575808868, "learning_rate": 3.8375544127188325e-05, - "loss": 0.3258, + "loss": 0.3136, "step": 266 }, { "epoch": 1.081518987341772, - "grad_norm": 0.462969044192099, + "grad_norm": 0.44220015426617215, "learning_rate": 3.8353063172072564e-05, - "loss": 0.3312, + "loss": 0.316, "step": 267 }, { "epoch": 1.0855696202531646, - "grad_norm": 0.5399400348299159, + "grad_norm": 0.4187428826568106, "learning_rate": 3.8330434403909105e-05, - "loss": 0.3446, + "loss": 0.3168, "step": 268 }, { "epoch": 1.089620253164557, - "grad_norm": 0.4074846037974673, + "grad_norm": 0.37810889035188694, "learning_rate": 3.8307658004946934e-05, - "loss": 0.3118, + "loss": 0.289, "step": 269 }, { "epoch": 1.0936708860759494, - "grad_norm": 0.4293388707068114, + "grad_norm": 0.3623707862537442, "learning_rate": 3.8284734158624046e-05, - "loss": 0.3267, + "loss": 0.3028, "step": 270 }, { "epoch": 1.0977215189873417, - "grad_norm": 0.46230221830015045, + "grad_norm": 0.373036044593057, "learning_rate": 3.826166304956594e-05, - "loss": 0.328, + "loss": 0.3024, "step": 271 }, { "epoch": 1.1017721518987342, - "grad_norm": 0.36579262073629076, + "grad_norm": 0.3579610036865689, "learning_rate": 3.8238444863584164e-05, - "loss": 0.3255, + "loss": 0.3063, "step": 272 }, { "epoch": 1.1058227848101265, - "grad_norm": 0.38816461945716824, + "grad_norm": 0.373698605628248, "learning_rate": 3.821507978767479e-05, - "loss": 0.3499, + "loss": 0.3107, "step": 273 }, { "epoch": 1.109873417721519, - "grad_norm": 0.41053072610266017, + "grad_norm": 0.36314227762444573, "learning_rate": 3.819156801001693e-05, - "loss": 0.3087, + "loss": 0.2985, "step": 274 }, { "epoch": 1.1139240506329113, - "grad_norm": 0.38134737053208123, + "grad_norm": 0.388679686159093, "learning_rate": 3.816790971997121e-05, - "loss": 0.3264, + "loss": 0.2904, "step": 275 }, { "epoch": 1.1179746835443038, - "grad_norm": 0.41259304090422777, + "grad_norm": 0.40794205230160135, "learning_rate": 3.8144105108078246e-05, - "loss": 0.3251, + "loss": 0.2998, "step": 276 }, { "epoch": 1.1220253164556961, - "grad_norm": 0.37614992489649796, + "grad_norm": 0.39930400463981613, "learning_rate": 3.81201543660571e-05, - "loss": 0.3281, + "loss": 0.3118, "step": 277 }, { "epoch": 1.1260759493670887, - "grad_norm": 0.37682342647462186, + "grad_norm": 0.4255774371377535, "learning_rate": 3.809605768680377e-05, - "loss": 0.327, + "loss": 0.3202, "step": 278 }, { "epoch": 1.130126582278481, - "grad_norm": 0.42631104449448914, + "grad_norm": 0.38735615562552056, "learning_rate": 3.807181526438958e-05, - "loss": 0.3471, + "loss": 0.309, "step": 279 }, { "epoch": 1.1341772151898735, - "grad_norm": 0.4908811399794211, + "grad_norm": 0.4466492250913194, "learning_rate": 3.8047427294059697e-05, - "loss": 0.3333, + "loss": 0.3043, "step": 280 }, { "epoch": 1.1382278481012658, - "grad_norm": 0.380930285424604, + "grad_norm": 0.42640365815915254, "learning_rate": 3.802289397223145e-05, - "loss": 0.309, + "loss": 0.3031, "step": 281 }, { "epoch": 1.1422784810126583, - "grad_norm": 0.4231096878122168, + "grad_norm": 0.4409857000647316, "learning_rate": 3.7998215496492854e-05, - "loss": 0.3279, + "loss": 0.3055, "step": 282 }, { "epoch": 1.1463291139240506, - "grad_norm": 0.4369348406284585, + "grad_norm": 0.39427645212429807, "learning_rate": 3.797339206560096e-05, - "loss": 0.3342, + "loss": 0.3043, "step": 283 }, { "epoch": 1.150379746835443, - "grad_norm": 0.38904724260357476, + "grad_norm": 0.48099247250782884, "learning_rate": 3.794842387948027e-05, - "loss": 0.3553, + "loss": 0.3201, "step": 284 }, { "epoch": 1.1544303797468354, - "grad_norm": 0.3798742938249508, + "grad_norm": 0.4683806591681051, "learning_rate": 3.7923311139221114e-05, - "loss": 0.3254, + "loss": 0.3194, "step": 285 }, { "epoch": 1.158481012658228, - "grad_norm": 0.43645498467058075, + "grad_norm": 0.3718883562363268, "learning_rate": 3.7898054047078054e-05, - "loss": 0.3454, + "loss": 0.2945, "step": 286 }, { "epoch": 1.1625316455696202, - "grad_norm": 0.4092509254405474, + "grad_norm": 0.48855101618719643, "learning_rate": 3.787265280646825e-05, - "loss": 0.3337, + "loss": 0.3074, "step": 287 }, { "epoch": 1.1665822784810127, - "grad_norm": 0.35574599408301616, + "grad_norm": 0.488001904655089, "learning_rate": 3.7847107621969786e-05, - "loss": 0.3213, + "loss": 0.3098, "step": 288 }, { "epoch": 1.170632911392405, - "grad_norm": 0.45758980644545105, + "grad_norm": 0.41149185666387744, "learning_rate": 3.7821418699320064e-05, - "loss": 0.3375, + "loss": 0.3018, "step": 289 }, { "epoch": 1.1746835443037975, - "grad_norm": 0.38896714962605344, + "grad_norm": 0.5049317655930331, "learning_rate": 3.7795586245414145e-05, - "loss": 0.3308, + "loss": 0.3198, "step": 290 }, { "epoch": 1.1787341772151898, - "grad_norm": 0.4184792537667025, + "grad_norm": 0.3638509943809286, "learning_rate": 3.776961046830306e-05, - "loss": 0.357, + "loss": 0.3128, "step": 291 }, { "epoch": 1.1827848101265823, - "grad_norm": 0.4441596666543161, + "grad_norm": 0.512119104956546, "learning_rate": 3.774349157719215e-05, - "loss": 0.318, + "loss": 0.3047, "step": 292 }, { "epoch": 1.1868354430379746, - "grad_norm": 0.35978079076160885, + "grad_norm": 0.41507224029396195, "learning_rate": 3.7717229782439365e-05, - "loss": 0.32, + "loss": 0.2938, "step": 293 }, { "epoch": 1.1908860759493671, - "grad_norm": 0.4597179754566955, + "grad_norm": 0.4458602180586056, "learning_rate": 3.769082529555359e-05, - "loss": 0.3503, + "loss": 0.3143, "step": 294 }, { "epoch": 1.1949367088607594, - "grad_norm": 0.40207987004173007, + "grad_norm": 0.45200586849839086, "learning_rate": 3.766427832919294e-05, - "loss": 0.3138, + "loss": 0.2907, "step": 295 }, { "epoch": 1.198987341772152, - "grad_norm": 0.3951548566707041, + "grad_norm": 0.4185441323047734, "learning_rate": 3.7637589097163024e-05, - "loss": 0.3161, + "loss": 0.2996, "step": 296 }, { "epoch": 1.2030379746835442, - "grad_norm": 0.46099866436741455, + "grad_norm": 0.47620238078848565, "learning_rate": 3.761075781441526e-05, - "loss": 0.3468, + "loss": 0.3204, "step": 297 }, { "epoch": 1.2070886075949367, - "grad_norm": 0.35469594975824653, + "grad_norm": 0.37681597816293283, "learning_rate": 3.75837846970451e-05, - "loss": 0.3186, + "loss": 0.295, "step": 298 }, { "epoch": 1.211139240506329, - "grad_norm": 0.49723954741966225, + "grad_norm": 0.44045998419606053, "learning_rate": 3.755666996229032e-05, - "loss": 0.3225, + "loss": 0.3156, "step": 299 }, { "epoch": 1.2151898734177216, - "grad_norm": 0.37054938225048767, + "grad_norm": 0.42622523087891967, "learning_rate": 3.752941382852927e-05, - "loss": 0.3419, + "loss": 0.3056, "step": 300 }, { "epoch": 1.2192405063291138, - "grad_norm": 0.4892147902377198, + "grad_norm": 0.39145017259592085, "learning_rate": 3.7502016515279115e-05, - "loss": 0.3455, + "loss": 0.32, "step": 301 }, { "epoch": 1.2232911392405064, - "grad_norm": 0.3358253627430396, + "grad_norm": 0.38906613244202576, "learning_rate": 3.7474478243194043e-05, - "loss": 0.3561, + "loss": 0.3301, "step": 302 }, { "epoch": 1.2273417721518987, - "grad_norm": 0.4059763400849329, + "grad_norm": 0.3784912113621531, "learning_rate": 3.744679923406351e-05, - "loss": 0.3411, + "loss": 0.3111, "step": 303 }, { "epoch": 1.2313924050632912, - "grad_norm": 0.31960273045218307, + "grad_norm": 0.4216892242894796, "learning_rate": 3.741897971081043e-05, - "loss": 0.3337, + "loss": 0.3203, "step": 304 }, { "epoch": 1.2354430379746835, - "grad_norm": 0.40456523025447644, + "grad_norm": 0.3768836360285952, "learning_rate": 3.739101989748946e-05, - "loss": 0.3424, + "loss": 0.3102, "step": 305 }, { "epoch": 1.239493670886076, - "grad_norm": 0.38126069728437595, + "grad_norm": 0.4008715996205904, "learning_rate": 3.7362920019285066e-05, - "loss": 0.3139, + "loss": 0.2882, "step": 306 }, { "epoch": 1.2435443037974683, - "grad_norm": 0.4093334757871732, + "grad_norm": 0.4169500727114222, "learning_rate": 3.73346803025098e-05, - "loss": 0.3511, + "loss": 0.3196, "step": 307 }, { "epoch": 1.2475949367088608, - "grad_norm": 0.3378852694473554, + "grad_norm": 0.3596736817631642, "learning_rate": 3.730630097460247e-05, - "loss": 0.3292, + "loss": 0.3169, "step": 308 }, { "epoch": 1.251645569620253, - "grad_norm": 0.4106796022819733, + "grad_norm": 0.45383372875838074, "learning_rate": 3.727778226412628e-05, - "loss": 0.3515, + "loss": 0.3188, "step": 309 }, { "epoch": 1.2556962025316456, - "grad_norm": 0.3195122613616194, + "grad_norm": 0.3708696452154138, "learning_rate": 3.7249124400767006e-05, - "loss": 0.3083, + "loss": 0.2932, "step": 310 }, { "epoch": 1.259746835443038, - "grad_norm": 0.40899496764146936, + "grad_norm": 0.38017714206275194, "learning_rate": 3.722032761533114e-05, - "loss": 0.3427, + "loss": 0.3044, "step": 311 }, { "epoch": 1.2637974683544304, - "grad_norm": 0.3495918987821639, + "grad_norm": 0.38903080040893345, "learning_rate": 3.719139213974403e-05, - "loss": 0.2907, + "loss": 0.2859, "step": 312 }, { "epoch": 1.2678481012658227, - "grad_norm": 0.48812581260219656, + "grad_norm": 0.4237171809133419, "learning_rate": 3.7162318207048006e-05, - "loss": 0.3747, + "loss": 0.3218, "step": 313 }, { "epoch": 1.2718987341772152, - "grad_norm": 0.3370626687877601, + "grad_norm": 0.3911093364068046, "learning_rate": 3.713310605140055e-05, - "loss": 0.31, + "loss": 0.3003, "step": 314 }, { "epoch": 1.2759493670886077, - "grad_norm": 0.46106036110463877, + "grad_norm": 0.40094745664913267, "learning_rate": 3.710375590807233e-05, - "loss": 0.3383, + "loss": 0.3167, "step": 315 }, { "epoch": 1.28, - "grad_norm": 0.39860004680354066, + "grad_norm": 0.388554840588802, "learning_rate": 3.7074268013445365e-05, - "loss": 0.3477, + "loss": 0.3072, "step": 316 }, { "epoch": 1.2840506329113923, - "grad_norm": 0.37332196871903506, + "grad_norm": 0.3528133424702583, "learning_rate": 3.7044642605011114e-05, - "loss": 0.3272, + "loss": 0.2975, "step": 317 }, { "epoch": 1.2881012658227848, - "grad_norm": 0.3660542344299313, + "grad_norm": 0.37292530385294026, "learning_rate": 3.701487992136854e-05, - "loss": 0.316, + "loss": 0.3025, "step": 318 }, { "epoch": 1.2921518987341774, - "grad_norm": 0.42233959606691723, + "grad_norm": 0.3517577174738942, "learning_rate": 3.69849802022222e-05, - "loss": 0.3678, + "loss": 0.3191, "step": 319 }, { "epoch": 1.2962025316455696, - "grad_norm": 0.3675905031434426, + "grad_norm": 0.3683039864288579, "learning_rate": 3.6954943688380334e-05, - "loss": 0.3335, + "loss": 0.306, "step": 320 }, { "epoch": 1.300253164556962, - "grad_norm": 0.4136559907433035, + "grad_norm": 0.3934537659862702, "learning_rate": 3.692477062175289e-05, - "loss": 0.3209, + "loss": 0.3101, "step": 321 }, { "epoch": 1.3043037974683545, - "grad_norm": 0.36062476835644686, + "grad_norm": 0.3709759260295349, "learning_rate": 3.689446124534958e-05, - "loss": 0.3342, + "loss": 0.3115, "step": 322 }, { "epoch": 1.308354430379747, - "grad_norm": 0.39149956567777666, + "grad_norm": 0.4078998532827138, "learning_rate": 3.686401580327799e-05, - "loss": 0.3519, + "loss": 0.3138, "step": 323 }, { "epoch": 1.3124050632911393, - "grad_norm": 0.3486675570840393, + "grad_norm": 0.38497107622568194, "learning_rate": 3.683343454074149e-05, - "loss": 0.3184, + "loss": 0.3081, "step": 324 }, { "epoch": 1.3164556962025316, - "grad_norm": 0.3897817283596259, + "grad_norm": 0.39117870137121996, "learning_rate": 3.6802717704037386e-05, - "loss": 0.3419, + "loss": 0.3115, "step": 325 }, { "epoch": 1.320506329113924, - "grad_norm": 0.3921985724242067, + "grad_norm": 0.3573059313663357, "learning_rate": 3.6771865540554855e-05, - "loss": 0.3307, + "loss": 0.2992, "step": 326 }, { "epoch": 1.3245569620253166, - "grad_norm": 0.36272450024832337, + "grad_norm": 0.40203363003379106, "learning_rate": 3.674087829877297e-05, - "loss": 0.3338, + "loss": 0.3107, "step": 327 }, { "epoch": 1.3286075949367089, - "grad_norm": 0.38645424570698883, + "grad_norm": 0.38962655382917494, "learning_rate": 3.6709756228258735e-05, - "loss": 0.3434, + "loss": 0.3114, "step": 328 }, { "epoch": 1.3326582278481012, - "grad_norm": 0.36084312352620335, + "grad_norm": 0.37511970994837557, "learning_rate": 3.667849957966501e-05, - "loss": 0.318, + "loss": 0.3047, "step": 329 }, { "epoch": 1.3367088607594937, - "grad_norm": 0.35777497825650123, + "grad_norm": 0.36457850547592774, "learning_rate": 3.6647108604728546e-05, - "loss": 0.3212, + "loss": 0.2971, "step": 330 }, { "epoch": 1.3407594936708862, - "grad_norm": 0.384847059415587, + "grad_norm": 0.33774521726771906, "learning_rate": 3.661558355626795e-05, - "loss": 0.3313, + "loss": 0.2932, "step": 331 }, { "epoch": 1.3448101265822785, - "grad_norm": 0.3556399390039954, + "grad_norm": 0.3636656105219239, "learning_rate": 3.658392468818163e-05, - "loss": 0.3355, + "loss": 0.3002, "step": 332 }, { "epoch": 1.3488607594936708, - "grad_norm": 0.4052226827344288, + "grad_norm": 0.31321411197159726, "learning_rate": 3.655213225544574e-05, - "loss": 0.3385, + "loss": 0.3083, "step": 333 }, { "epoch": 1.3529113924050633, - "grad_norm": 0.3237089326570128, + "grad_norm": 0.32657355911635233, "learning_rate": 3.652020651411218e-05, - "loss": 0.3144, + "loss": 0.298, "step": 334 }, { "epoch": 1.3569620253164558, - "grad_norm": 0.38232899071178317, + "grad_norm": 0.3380579487426661, "learning_rate": 3.6488147721306474e-05, - "loss": 0.3444, + "loss": 0.3124, "step": 335 }, { "epoch": 1.3610126582278481, - "grad_norm": 0.31408018363206536, + "grad_norm": 0.35873855755908063, "learning_rate": 3.645595613522574e-05, - "loss": 0.343, + "loss": 0.3173, "step": 336 }, { "epoch": 1.3650632911392404, - "grad_norm": 0.33903010701019765, + "grad_norm": 0.3582562735927334, "learning_rate": 3.642363201513657e-05, - "loss": 0.316, + "loss": 0.3115, "step": 337 }, { "epoch": 1.369113924050633, - "grad_norm": 0.37572760308166736, + "grad_norm": 0.3687602502711562, "learning_rate": 3.6391175621373006e-05, - "loss": 0.3441, + "loss": 0.3045, "step": 338 }, { "epoch": 1.3731645569620254, - "grad_norm": 0.348764271848724, + "grad_norm": 0.41965199699997474, "learning_rate": 3.6358587215334355e-05, - "loss": 0.3158, + "loss": 0.308, "step": 339 }, { "epoch": 1.3772151898734177, - "grad_norm": 0.3978885840749925, + "grad_norm": 0.3368401903867559, "learning_rate": 3.632586705948318e-05, - "loss": 0.3487, + "loss": 0.3025, "step": 340 }, { "epoch": 1.38126582278481, - "grad_norm": 0.3955376928530547, + "grad_norm": 0.3496207224304805, "learning_rate": 3.629301541734311e-05, - "loss": 0.3367, + "loss": 0.3242, "step": 341 }, { "epoch": 1.3853164556962025, - "grad_norm": 0.3817920671430099, + "grad_norm": 0.35937586417012435, "learning_rate": 3.626003255349676e-05, - "loss": 0.3549, + "loss": 0.3218, "step": 342 }, { "epoch": 1.389367088607595, - "grad_norm": 0.36228058553423714, + "grad_norm": 0.30954197599652233, "learning_rate": 3.622691873358357e-05, - "loss": 0.3174, + "loss": 0.3151, "step": 343 }, { "epoch": 1.3934177215189874, - "grad_norm": 0.383083511752387, + "grad_norm": 0.3796695377233597, "learning_rate": 3.61936742242977e-05, - "loss": 0.3471, + "loss": 0.3171, "step": 344 }, { "epoch": 1.3974683544303796, - "grad_norm": 0.3664973769399827, + "grad_norm": 0.33671418084622534, "learning_rate": 3.6160299293385864e-05, - "loss": 0.3467, + "loss": 0.3087, "step": 345 }, { "epoch": 1.4015189873417722, - "grad_norm": 0.3969190023340561, + "grad_norm": 0.34008801775337266, "learning_rate": 3.612679420964516e-05, - "loss": 0.3437, + "loss": 0.3141, "step": 346 }, { "epoch": 1.4055696202531647, - "grad_norm": 0.35321778050750025, + "grad_norm": 0.35438680409491535, "learning_rate": 3.609315924292092e-05, - "loss": 0.3211, + "loss": 0.3047, "step": 347 }, { "epoch": 1.409620253164557, - "grad_norm": 0.3376643767698355, + "grad_norm": 0.3309026533894053, "learning_rate": 3.6059394664104554e-05, - "loss": 0.3434, + "loss": 0.3046, "step": 348 }, { "epoch": 1.4136708860759493, - "grad_norm": 0.36057355547651787, + "grad_norm": 0.37041288531262945, "learning_rate": 3.602550074513133e-05, - "loss": 0.3204, + "loss": 0.3022, "step": 349 }, { "epoch": 1.4177215189873418, - "grad_norm": 0.37836020669196235, + "grad_norm": 0.36671584499785154, "learning_rate": 3.599147775897822e-05, - "loss": 0.3469, + "loss": 0.3076, "step": 350 }, { "epoch": 1.4217721518987343, - "grad_norm": 0.33491672553200263, + "grad_norm": 0.3683879520241095, "learning_rate": 3.595732597966167e-05, - "loss": 0.329, + "loss": 0.3083, "step": 351 }, { "epoch": 1.4258227848101266, - "grad_norm": 0.35162479084071546, + "grad_norm": 0.33413301296680736, "learning_rate": 3.592304568223542e-05, - "loss": 0.3279, + "loss": 0.3062, "step": 352 }, { "epoch": 1.4298734177215189, - "grad_norm": 0.3619927092449756, + "grad_norm": 0.37380631139938114, "learning_rate": 3.588863714278826e-05, - "loss": 0.3328, + "loss": 0.2958, "step": 353 }, { "epoch": 1.4339240506329114, - "grad_norm": 0.3055889371856443, + "grad_norm": 0.36062326488543883, "learning_rate": 3.585410063844186e-05, - "loss": 0.3152, + "loss": 0.2931, "step": 354 }, { "epoch": 1.437974683544304, - "grad_norm": 0.38355449336774566, + "grad_norm": 0.3622229125164926, "learning_rate": 3.581943644734846e-05, - "loss": 0.327, + "loss": 0.2996, "step": 355 }, { "epoch": 1.4420253164556962, - "grad_norm": 0.32190665434227606, + "grad_norm": 0.3819719196404846, "learning_rate": 3.578464484868869e-05, - "loss": 0.3207, + "loss": 0.309, "step": 356 }, { "epoch": 1.4460759493670885, - "grad_norm": 0.3818830997656326, + "grad_norm": 0.352125928219145, "learning_rate": 3.5749726122669316e-05, - "loss": 0.3277, + "loss": 0.3036, "step": 357 }, { "epoch": 1.450126582278481, - "grad_norm": 0.39625653605889494, + "grad_norm": 0.3726055203774872, "learning_rate": 3.5714680550520943e-05, - "loss": 0.3319, + "loss": 0.31, "step": 358 }, { "epoch": 1.4541772151898735, - "grad_norm": 0.3529986113755309, + "grad_norm": 0.3776051700137096, "learning_rate": 3.5679508414495794e-05, - "loss": 0.3465, + "loss": 0.314, "step": 359 }, { "epoch": 1.4582278481012658, - "grad_norm": 0.41790418966877346, + "grad_norm": 0.3767577519707251, "learning_rate": 3.564420999786543e-05, - "loss": 0.3208, + "loss": 0.3094, "step": 360 }, { "epoch": 1.4622784810126581, - "grad_norm": 0.44553059887031055, + "grad_norm": 0.4074339338683601, "learning_rate": 3.560878558491842e-05, - "loss": 0.3425, + "loss": 0.3162, "step": 361 }, { "epoch": 1.4663291139240506, - "grad_norm": 0.39105649954405497, + "grad_norm": 0.33414754474626074, "learning_rate": 3.5573235460958145e-05, - "loss": 0.3503, + "loss": 0.3087, "step": 362 }, { "epoch": 1.4703797468354431, - "grad_norm": 0.41908657101626307, + "grad_norm": 0.40276270930067204, "learning_rate": 3.553755991230039e-05, - "loss": 0.3087, + "loss": 0.3019, "step": 363 }, { "epoch": 1.4744303797468354, - "grad_norm": 0.40064495829319396, + "grad_norm": 0.3569737364499927, "learning_rate": 3.5501759226271144e-05, - "loss": 0.3481, + "loss": 0.3095, "step": 364 }, { "epoch": 1.4784810126582277, - "grad_norm": 0.4510329626769198, + "grad_norm": 0.3951990904713934, "learning_rate": 3.546583369120419e-05, - "loss": 0.346, + "loss": 0.3149, "step": 365 }, { "epoch": 1.4825316455696202, - "grad_norm": 0.3862953120033996, + "grad_norm": 0.42006591448621794, "learning_rate": 3.5429783596438864e-05, - "loss": 0.3145, + "loss": 0.3036, "step": 366 }, { "epoch": 1.4865822784810128, - "grad_norm": 0.3862342613650019, + "grad_norm": 0.33040874615111787, "learning_rate": 3.539360923231766e-05, - "loss": 0.3072, + "loss": 0.292, "step": 367 }, { "epoch": 1.490632911392405, - "grad_norm": 0.34526771928904004, + "grad_norm": 0.3778169988859503, "learning_rate": 3.535731089018394e-05, - "loss": 0.3182, + "loss": 0.3022, "step": 368 }, { "epoch": 1.4946835443037974, - "grad_norm": 0.3617630232819585, + "grad_norm": 0.37768212944799073, "learning_rate": 3.532088886237956e-05, - "loss": 0.3255, + "loss": 0.2984, "step": 369 }, { "epoch": 1.4987341772151899, - "grad_norm": 0.3857896726305055, + "grad_norm": 0.35620827314210374, "learning_rate": 3.528434344224253e-05, - "loss": 0.3432, + "loss": 0.3002, "step": 370 }, { "epoch": 1.5027848101265824, - "grad_norm": 0.3519294600765587, + "grad_norm": 0.43037923315563587, "learning_rate": 3.524767492410464e-05, - "loss": 0.3458, + "loss": 0.3211, "step": 371 }, { "epoch": 1.5068354430379747, - "grad_norm": 0.3520878618788712, + "grad_norm": 0.35195909697612277, "learning_rate": 3.521088360328908e-05, - "loss": 0.3274, + "loss": 0.311, "step": 372 }, { "epoch": 1.510886075949367, - "grad_norm": 0.3333783260241572, + "grad_norm": 0.3711450229276612, "learning_rate": 3.517396977610811e-05, - "loss": 0.3415, + "loss": 0.3001, "step": 373 }, { "epoch": 1.5149367088607595, - "grad_norm": 0.31735120028003116, + "grad_norm": 0.36729472410650504, "learning_rate": 3.5136933739860595e-05, - "loss": 0.3078, + "loss": 0.2988, "step": 374 }, { "epoch": 1.518987341772152, - "grad_norm": 0.3244822983112285, + "grad_norm": 0.40554013217426027, "learning_rate": 3.509977579282971e-05, - "loss": 0.349, + "loss": 0.3219, "step": 375 }, { "epoch": 1.5230379746835443, - "grad_norm": 0.33686589702818753, + "grad_norm": 0.4495547020923126, "learning_rate": 3.5062496234280424e-05, - "loss": 0.3322, + "loss": 0.3081, "step": 376 }, { "epoch": 1.5270886075949366, - "grad_norm": 0.33322541670094125, + "grad_norm": 0.4153316219102672, "learning_rate": 3.502509536445719e-05, - "loss": 0.3298, + "loss": 0.3124, "step": 377 }, { "epoch": 1.531139240506329, - "grad_norm": 0.38110540522192, + "grad_norm": 0.3846414856431148, "learning_rate": 3.498757348458147e-05, - "loss": 0.3261, + "loss": 0.3036, "step": 378 }, { "epoch": 1.5351898734177216, - "grad_norm": 0.32827175300849076, + "grad_norm": 0.3967641070642658, "learning_rate": 3.4949930896849324e-05, - "loss": 0.3418, + "loss": 0.3075, "step": 379 }, { "epoch": 1.539240506329114, - "grad_norm": 0.3159773512356361, + "grad_norm": 0.3554207096813955, "learning_rate": 3.491216790442899e-05, - "loss": 0.2992, + "loss": 0.2922, "step": 380 }, { "epoch": 1.5432911392405062, - "grad_norm": 0.32284251511346596, + "grad_norm": 0.37608270198564364, "learning_rate": 3.487428481145839e-05, - "loss": 0.3298, + "loss": 0.299, "step": 381 }, { "epoch": 1.5473417721518987, - "grad_norm": 0.33054605861687214, + "grad_norm": 0.4242786975157559, "learning_rate": 3.483628192304278e-05, - "loss": 0.3275, + "loss": 0.3075, "step": 382 }, { "epoch": 1.5513924050632912, - "grad_norm": 0.3459227415980159, + "grad_norm": 0.35626465601849155, "learning_rate": 3.479815954525219e-05, - "loss": 0.368, + "loss": 0.3248, "step": 383 }, { "epoch": 1.5554430379746835, - "grad_norm": 0.31898984955718246, + "grad_norm": 0.37477137297898305, "learning_rate": 3.475991798511899e-05, - "loss": 0.3202, + "loss": 0.2997, "step": 384 }, { "epoch": 1.5594936708860758, - "grad_norm": 0.3409771484413442, + "grad_norm": 0.3640621366264976, "learning_rate": 3.4721557550635464e-05, - "loss": 0.3265, + "loss": 0.316, "step": 385 }, { "epoch": 1.5635443037974683, - "grad_norm": 0.29620879496814184, + "grad_norm": 0.3484506899595033, "learning_rate": 3.468307855075128e-05, - "loss": 0.3185, + "loss": 0.3102, "step": 386 }, { "epoch": 1.5675949367088609, - "grad_norm": 0.36903468367768383, + "grad_norm": 0.3708071599479605, "learning_rate": 3.4644481295371005e-05, - "loss": 0.3726, + "loss": 0.3229, "step": 387 }, { "epoch": 1.5716455696202531, - "grad_norm": 0.29913922533727344, + "grad_norm": 0.3288023071698783, "learning_rate": 3.460576609535163e-05, - "loss": 0.3094, + "loss": 0.3, "step": 388 }, { "epoch": 1.5756962025316454, - "grad_norm": 0.3285546100078742, + "grad_norm": 0.3518968564337706, "learning_rate": 3.456693326250006e-05, - "loss": 0.324, + "loss": 0.3111, "step": 389 }, { "epoch": 1.579746835443038, - "grad_norm": 0.307154168850344, + "grad_norm": 0.3245852088532655, "learning_rate": 3.452798310957058e-05, - "loss": 0.3149, + "loss": 0.2918, "step": 390 }, { "epoch": 1.5837974683544305, - "grad_norm": 0.3254629306542, + "grad_norm": 0.3358093803637686, "learning_rate": 3.4488915950262386e-05, - "loss": 0.3384, + "loss": 0.2993, "step": 391 }, { "epoch": 1.5878481012658228, - "grad_norm": 0.3322895800767244, + "grad_norm": 0.35882399915393237, "learning_rate": 3.4449732099216985e-05, - "loss": 0.3162, + "loss": 0.3145, "step": 392 }, { "epoch": 1.591898734177215, - "grad_norm": 0.3559725169324527, + "grad_norm": 0.33067413016730346, "learning_rate": 3.441043187201574e-05, - "loss": 0.344, + "loss": 0.3115, "step": 393 }, { "epoch": 1.5959493670886076, - "grad_norm": 0.3412711161835948, + "grad_norm": 0.32778330362095626, "learning_rate": 3.437101558517728e-05, - "loss": 0.341, + "loss": 0.3071, "step": 394 }, { "epoch": 1.6, - "grad_norm": 0.3172166605851028, + "grad_norm": 0.31413419589106195, "learning_rate": 3.433148355615496e-05, - "loss": 0.333, + "loss": 0.3073, "step": 395 }, { "epoch": 1.6040506329113924, - "grad_norm": 0.2858804054198202, + "grad_norm": 0.31471642804486183, "learning_rate": 3.4291836103334294e-05, - "loss": 0.3163, + "loss": 0.3011, "step": 396 }, { "epoch": 1.6081012658227847, - "grad_norm": 0.29505337153270805, + "grad_norm": 0.34047956205692265, "learning_rate": 3.425207354603043e-05, - "loss": 0.3264, + "loss": 0.3069, "step": 397 }, { "epoch": 1.6121518987341772, - "grad_norm": 0.3147589014672285, + "grad_norm": 0.3626339218829372, "learning_rate": 3.421219620448553e-05, - "loss": 0.3362, + "loss": 0.3017, "step": 398 }, { "epoch": 1.6162025316455697, - "grad_norm": 0.3117430657784261, + "grad_norm": 0.34868182854954943, "learning_rate": 3.417220439986623e-05, - "loss": 0.3181, + "loss": 0.3012, "step": 399 }, { "epoch": 1.620253164556962, - "grad_norm": 0.3185278223807896, + "grad_norm": 0.3480734460766342, "learning_rate": 3.4132098454261024e-05, - "loss": 0.3211, + "loss": 0.3037, "step": 400 }, { "epoch": 1.6243037974683543, - "grad_norm": 0.350164183409633, + "grad_norm": 0.33681019616820673, "learning_rate": 3.4091878690677676e-05, - "loss": 0.35, + "loss": 0.312, "step": 401 }, { "epoch": 1.6283544303797468, - "grad_norm": 0.3179429674610946, + "grad_norm": 0.32409164592458173, "learning_rate": 3.405154543304065e-05, - "loss": 0.3278, + "loss": 0.2994, "step": 402 }, { "epoch": 1.6324050632911393, - "grad_norm": 0.38294064455282006, + "grad_norm": 0.3351476462826444, "learning_rate": 3.401109900618843e-05, - "loss": 0.3224, + "loss": 0.3088, "step": 403 }, { "epoch": 1.6364556962025316, - "grad_norm": 0.3717737566074266, + "grad_norm": 0.3722931928591337, "learning_rate": 3.3970539735870996e-05, - "loss": 0.3325, + "loss": 0.2945, "step": 404 }, { "epoch": 1.640506329113924, - "grad_norm": 0.3415077979022879, + "grad_norm": 0.32738075064384836, "learning_rate": 3.392986794874714e-05, - "loss": 0.3399, + "loss": 0.3104, "step": 405 }, { "epoch": 1.6445569620253164, - "grad_norm": 0.33992049590237494, + "grad_norm": 0.3807225132103434, "learning_rate": 3.388908397238184e-05, - "loss": 0.3149, + "loss": 0.3016, "step": 406 }, { "epoch": 1.648607594936709, - "grad_norm": 0.3632969054506239, + "grad_norm": 0.3882093765887405, "learning_rate": 3.384818813524362e-05, - "loss": 0.3485, + "loss": 0.3099, "step": 407 }, { "epoch": 1.6526582278481012, - "grad_norm": 0.30792995290699265, + "grad_norm": 0.32245071553202787, "learning_rate": 3.380718076670195e-05, - "loss": 0.3379, + "loss": 0.3099, "step": 408 }, { "epoch": 1.6567088607594935, - "grad_norm": 0.3317032913044713, + "grad_norm": 0.38340179945418557, "learning_rate": 3.376606219702454e-05, - "loss": 0.331, + "loss": 0.3125, "step": 409 }, { "epoch": 1.660759493670886, - "grad_norm": 0.3481674328483524, + "grad_norm": 0.3249376884421805, "learning_rate": 3.372483275737468e-05, - "loss": 0.3388, + "loss": 0.3064, "step": 410 }, { "epoch": 1.6648101265822786, - "grad_norm": 0.38632041234437425, + "grad_norm": 0.35158598633902166, "learning_rate": 3.368349277980861e-05, - "loss": 0.3481, + "loss": 0.3177, "step": 411 }, { "epoch": 1.6688607594936709, - "grad_norm": 0.3545136227687343, + "grad_norm": 0.3538849132777769, "learning_rate": 3.3642042597272844e-05, - "loss": 0.3327, + "loss": 0.3101, "step": 412 }, { "epoch": 1.6729113924050631, - "grad_norm": 0.38671374169221295, + "grad_norm": 0.33348812206396083, "learning_rate": 3.360048254360144e-05, - "loss": 0.3357, + "loss": 0.3107, "step": 413 }, { "epoch": 1.6769620253164557, - "grad_norm": 0.37113913723765113, + "grad_norm": 0.38120476907145756, "learning_rate": 3.355881295351336e-05, - "loss": 0.3359, + "loss": 0.3045, "step": 414 }, { "epoch": 1.6810126582278482, - "grad_norm": 0.3668347641925283, + "grad_norm": 0.3186755410913132, "learning_rate": 3.351703416260975e-05, - "loss": 0.3354, + "loss": 0.3184, "step": 415 }, { "epoch": 1.6850632911392405, - "grad_norm": 0.3820636401810705, + "grad_norm": 0.4028436539339753, "learning_rate": 3.347514650737126e-05, - "loss": 0.3458, + "loss": 0.3163, "step": 416 }, { "epoch": 1.6891139240506328, - "grad_norm": 0.3553048840149249, + "grad_norm": 0.35998718747286695, "learning_rate": 3.3433150325155295e-05, - "loss": 0.3266, + "loss": 0.3154, "step": 417 }, { "epoch": 1.6931645569620253, - "grad_norm": 0.31465812648404634, + "grad_norm": 0.3419052451854561, "learning_rate": 3.339104595419334e-05, - "loss": 0.3286, + "loss": 0.3072, "step": 418 }, { "epoch": 1.6972151898734178, - "grad_norm": 0.3431869089625049, + "grad_norm": 0.32571003218071815, "learning_rate": 3.3348833733588204e-05, - "loss": 0.3356, + "loss": 0.3221, "step": 419 }, { "epoch": 1.70126582278481, - "grad_norm": 0.31714917874814047, + "grad_norm": 0.37665502467793605, "learning_rate": 3.3306514003311305e-05, - "loss": 0.3528, + "loss": 0.3134, "step": 420 }, { "epoch": 1.7053164556962024, - "grad_norm": 0.34031238912325995, + "grad_norm": 0.32960721576607344, "learning_rate": 3.326408710419996e-05, - "loss": 0.3314, + "loss": 0.3118, "step": 421 }, { "epoch": 1.709367088607595, - "grad_norm": 0.31688610038179754, + "grad_norm": 0.31986035746301433, "learning_rate": 3.322155337795454e-05, - "loss": 0.3142, + "loss": 0.3035, "step": 422 }, { "epoch": 1.7134177215189874, - "grad_norm": 0.38201964080829454, + "grad_norm": 0.29131743585380454, "learning_rate": 3.317891316713587e-05, - "loss": 0.3394, + "loss": 0.309, "step": 423 }, { "epoch": 1.7174683544303797, - "grad_norm": 0.3969057464779143, + "grad_norm": 0.34580285455860643, "learning_rate": 3.313616681516231e-05, - "loss": 0.3374, + "loss": 0.3045, "step": 424 }, { "epoch": 1.721518987341772, - "grad_norm": 0.33596885231027346, + "grad_norm": 0.3065576659134362, "learning_rate": 3.309331466630713e-05, - "loss": 0.3343, + "loss": 0.3182, "step": 425 }, { "epoch": 1.7255696202531645, - "grad_norm": 0.3436168727378085, + "grad_norm": 0.34846853798629507, "learning_rate": 3.305035706569563e-05, - "loss": 0.3364, + "loss": 0.3047, "step": 426 }, { "epoch": 1.729620253164557, - "grad_norm": 0.3331852639216001, + "grad_norm": 0.30321531766671955, "learning_rate": 3.3007294359302433e-05, - "loss": 0.3185, + "loss": 0.3044, "step": 427 }, { "epoch": 1.7336708860759493, - "grad_norm": 0.3735426187787576, + "grad_norm": 0.41512873547163487, "learning_rate": 3.296412689394864e-05, - "loss": 0.3199, + "loss": 0.3043, "step": 428 }, { "epoch": 1.7377215189873416, - "grad_norm": 0.3301482145252355, + "grad_norm": 0.3425970937025628, "learning_rate": 3.292085501729909e-05, - "loss": 0.3423, + "loss": 0.3083, "step": 429 }, { "epoch": 1.7417721518987341, - "grad_norm": 0.3775241122271627, + "grad_norm": 0.3600679162185109, "learning_rate": 3.2877479077859534e-05, - "loss": 0.3407, + "loss": 0.3036, "step": 430 }, { "epoch": 1.7458227848101266, - "grad_norm": 0.34253674754352764, + "grad_norm": 0.3455753775334914, "learning_rate": 3.283399942497381e-05, - "loss": 0.2917, + "loss": 0.288, "step": 431 }, { "epoch": 1.749873417721519, - "grad_norm": 0.35465977904277834, + "grad_norm": 0.3560226655218737, "learning_rate": 3.279041640882108e-05, - "loss": 0.3315, + "loss": 0.3027, "step": 432 }, { "epoch": 1.7539240506329112, - "grad_norm": 0.36913823786838457, + "grad_norm": 0.3508361229146194, "learning_rate": 3.2746730380412964e-05, - "loss": 0.3062, + "loss": 0.2919, "step": 433 }, { "epoch": 1.7579746835443038, - "grad_norm": 0.3653456114674793, + "grad_norm": 0.3320865940991662, "learning_rate": 3.2702941691590726e-05, - "loss": 0.3299, + "loss": 0.2953, "step": 434 }, { "epoch": 1.7620253164556963, - "grad_norm": 0.39409060251095357, + "grad_norm": 0.36987211858137564, "learning_rate": 3.265905069502244e-05, - "loss": 0.3455, + "loss": 0.3203, "step": 435 }, { "epoch": 1.7660759493670886, - "grad_norm": 0.361648102477994, + "grad_norm": 0.33831485507546677, "learning_rate": 3.261505774420016e-05, - "loss": 0.3024, + "loss": 0.2977, "step": 436 }, { "epoch": 1.7701265822784809, - "grad_norm": 0.39111921410550105, + "grad_norm": 0.4040583165106916, "learning_rate": 3.257096319343707e-05, - "loss": 0.3228, + "loss": 0.301, "step": 437 }, { "epoch": 1.7741772151898734, - "grad_norm": 0.37331404254985623, + "grad_norm": 0.35598305455242285, "learning_rate": 3.2526767397864614e-05, - "loss": 0.3192, + "loss": 0.2935, "step": 438 }, { "epoch": 1.7782278481012659, - "grad_norm": 0.36663476237322856, + "grad_norm": 0.4059701179135238, "learning_rate": 3.248247071342966e-05, - "loss": 0.3456, + "loss": 0.3219, "step": 439 }, { "epoch": 1.7822784810126582, - "grad_norm": 0.3665614497728774, + "grad_norm": 0.3510217153208696, "learning_rate": 3.243807349689161e-05, - "loss": 0.3195, + "loss": 0.3043, "step": 440 }, { "epoch": 1.7863291139240505, - "grad_norm": 0.3384815109657021, + "grad_norm": 0.3512658956120526, "learning_rate": 3.2393576105819544e-05, - "loss": 0.3538, + "loss": 0.3192, "step": 441 }, { "epoch": 1.790379746835443, - "grad_norm": 0.41509264530990103, + "grad_norm": 0.35895811010040746, "learning_rate": 3.2348978898589333e-05, - "loss": 0.3377, + "loss": 0.307, "step": 442 }, { "epoch": 1.7944303797468355, - "grad_norm": 0.3565041148165934, + "grad_norm": 0.4249513099091344, "learning_rate": 3.230428223438075e-05, - "loss": 0.3274, + "loss": 0.3013, "step": 443 }, { "epoch": 1.7984810126582278, - "grad_norm": 0.3662706369198632, + "grad_norm": 0.3352076104746956, "learning_rate": 3.225948647317459e-05, - "loss": 0.3197, + "loss": 0.2987, "step": 444 }, { "epoch": 1.80253164556962, - "grad_norm": 0.410486741399871, + "grad_norm": 0.4289409058939938, "learning_rate": 3.2214591975749745e-05, - "loss": 0.3378, + "loss": 0.3181, "step": 445 }, { "epoch": 1.8065822784810126, - "grad_norm": 0.354681308324123, + "grad_norm": 0.39341694347057665, "learning_rate": 3.216959910368034e-05, - "loss": 0.3235, + "loss": 0.311, "step": 446 }, { "epoch": 1.8106329113924051, - "grad_norm": 0.41077402427089005, + "grad_norm": 0.417025530824109, "learning_rate": 3.212450821933277e-05, - "loss": 0.3325, + "loss": 0.3049, "step": 447 }, { "epoch": 1.8146835443037974, - "grad_norm": 0.40990328689070793, + "grad_norm": 0.3887723170385768, "learning_rate": 3.207931968586281e-05, - "loss": 0.3574, + "loss": 0.3179, "step": 448 }, { "epoch": 1.81873417721519, - "grad_norm": 0.32703530686197374, + "grad_norm": 0.38974756267927496, "learning_rate": 3.203403386721272e-05, - "loss": 0.2994, + "loss": 0.298, "step": 449 }, { "epoch": 1.8227848101265822, - "grad_norm": 0.4006787750219712, + "grad_norm": 0.4040405016056989, "learning_rate": 3.1988651128108245e-05, - "loss": 0.3243, + "loss": 0.2983, "step": 450 }, { "epoch": 1.8268354430379747, - "grad_norm": 0.3812329269246029, + "grad_norm": 0.4082509500711886, "learning_rate": 3.194317183405573e-05, - "loss": 0.3371, + "loss": 0.3091, "step": 451 }, { "epoch": 1.8308860759493673, - "grad_norm": 0.34590931472325726, + "grad_norm": 0.36798266778857186, "learning_rate": 3.189759635133914e-05, - "loss": 0.3614, + "loss": 0.3316, "step": 452 }, { "epoch": 1.8349367088607595, - "grad_norm": 0.34757907252811465, + "grad_norm": 0.3532281353860484, "learning_rate": 3.185192504701718e-05, - "loss": 0.3373, + "loss": 0.3108, "step": 453 }, { "epoch": 1.8389873417721518, - "grad_norm": 0.30807077415055145, + "grad_norm": 0.347162741022233, "learning_rate": 3.1806158288920234e-05, - "loss": 0.3308, + "loss": 0.3136, "step": 454 }, { "epoch": 1.8430379746835444, - "grad_norm": 0.3772964713553934, + "grad_norm": 0.346705743547188, "learning_rate": 3.1760296445647477e-05, - "loss": 0.3278, + "loss": 0.3, "step": 455 }, { "epoch": 1.8470886075949369, - "grad_norm": 0.37805011229388197, + "grad_norm": 0.33622617828683116, "learning_rate": 3.1714339886563896e-05, - "loss": 0.3437, + "loss": 0.3128, "step": 456 }, { "epoch": 1.8511392405063292, - "grad_norm": 0.30488558568913077, + "grad_norm": 0.3441087707959669, "learning_rate": 3.166828898179731e-05, - "loss": 0.3265, + "loss": 0.3036, "step": 457 }, { "epoch": 1.8551898734177215, - "grad_norm": 0.3365758984247796, + "grad_norm": 0.3280173255061695, "learning_rate": 3.162214410223536e-05, - "loss": 0.3588, + "loss": 0.3289, "step": 458 }, { "epoch": 1.859240506329114, - "grad_norm": 0.3042101841717299, + "grad_norm": 0.3478782448564192, "learning_rate": 3.157590561952257e-05, - "loss": 0.3011, + "loss": 0.2899, "step": 459 }, { "epoch": 1.8632911392405065, - "grad_norm": 0.31735240582853663, + "grad_norm": 0.31582728388215037, "learning_rate": 3.152957390605732e-05, - "loss": 0.3356, + "loss": 0.2951, "step": 460 }, { "epoch": 1.8673417721518988, - "grad_norm": 0.2919914634523959, + "grad_norm": 0.31975948851908287, "learning_rate": 3.148314933498886e-05, - "loss": 0.3115, + "loss": 0.3005, "step": 461 }, { "epoch": 1.871392405063291, - "grad_norm": 0.34626579767075943, + "grad_norm": 0.33305479142753014, "learning_rate": 3.143663228021431e-05, - "loss": 0.3274, + "loss": 0.3019, "step": 462 }, { "epoch": 1.8754430379746836, - "grad_norm": 0.3085553046191359, + "grad_norm": 0.29771055186034223, "learning_rate": 3.1390023116375624e-05, - "loss": 0.3389, + "loss": 0.3114, "step": 463 }, { "epoch": 1.879493670886076, - "grad_norm": 0.31805739412600625, + "grad_norm": 0.34254152813331346, "learning_rate": 3.134332221885661e-05, - "loss": 0.3187, + "loss": 0.3072, "step": 464 }, { "epoch": 1.8835443037974684, - "grad_norm": 0.32067986384618635, + "grad_norm": 0.3030677988687901, "learning_rate": 3.129652996377987e-05, - "loss": 0.3409, + "loss": 0.3063, "step": 465 }, { "epoch": 1.8875949367088607, - "grad_norm": 0.3893099144976499, + "grad_norm": 0.33846526093021156, "learning_rate": 3.12496467280038e-05, - "loss": 0.3651, + "loss": 0.314, "step": 466 }, { "epoch": 1.8916455696202532, - "grad_norm": 0.4054164649315256, + "grad_norm": 0.3558736795532902, "learning_rate": 3.120267288911952e-05, - "loss": 0.3329, + "loss": 0.3082, "step": 467 }, { "epoch": 1.8956962025316457, - "grad_norm": 0.33537159841877334, + "grad_norm": 0.3182751704662135, "learning_rate": 3.11556088254479e-05, - "loss": 0.3437, + "loss": 0.3147, "step": 468 }, { "epoch": 1.899746835443038, - "grad_norm": 0.27403601334771244, + "grad_norm": 0.3285375429365072, "learning_rate": 3.11084549160364e-05, - "loss": 0.3042, + "loss": 0.2885, "step": 469 }, { "epoch": 1.9037974683544303, - "grad_norm": 0.3810516539530209, + "grad_norm": 0.361452802984596, "learning_rate": 3.106121154065615e-05, - "loss": 0.3591, + "loss": 0.3234, "step": 470 }, { "epoch": 1.9078481012658228, - "grad_norm": 0.2938894478582794, + "grad_norm": 0.33561875479342007, "learning_rate": 3.1013879079798805e-05, - "loss": 0.3258, + "loss": 0.3093, "step": 471 }, { "epoch": 1.9118987341772153, - "grad_norm": 0.34720796941143417, + "grad_norm": 0.37255876812371796, "learning_rate": 3.096645791467348e-05, - "loss": 0.3438, + "loss": 0.3123, "step": 472 }, { "epoch": 1.9159493670886076, - "grad_norm": 0.3545735422843274, + "grad_norm": 0.3499580467284806, "learning_rate": 3.091894842720373e-05, - "loss": 0.3417, + "loss": 0.3043, "step": 473 }, { "epoch": 1.92, - "grad_norm": 0.2933154176028958, + "grad_norm": 0.33157066558634335, "learning_rate": 3.0871351000024425e-05, - "loss": 0.3003, + "loss": 0.2966, "step": 474 }, { "epoch": 1.9240506329113924, - "grad_norm": 0.40814301387123464, + "grad_norm": 0.40534424699598226, "learning_rate": 3.0823666016478716e-05, - "loss": 0.3553, + "loss": 0.3073, "step": 475 }, { "epoch": 1.928101265822785, - "grad_norm": 0.330747547136201, + "grad_norm": 0.30211307801700543, "learning_rate": 3.0775893860614896e-05, - "loss": 0.3159, + "loss": 0.3041, "step": 476 }, { "epoch": 1.9321518987341773, - "grad_norm": 0.42336903508137785, + "grad_norm": 0.37154144186317534, "learning_rate": 3.0728034917183336e-05, - "loss": 0.346, + "loss": 0.3075, "step": 477 }, { "epoch": 1.9362025316455695, - "grad_norm": 0.3424873388746941, + "grad_norm": 0.3975354357050741, "learning_rate": 3.06800895716334e-05, - "loss": 0.3177, + "loss": 0.3076, "step": 478 }, { "epoch": 1.940253164556962, - "grad_norm": 0.3757586121439783, + "grad_norm": 0.375381675859087, "learning_rate": 3.063205821011029e-05, - "loss": 0.3389, + "loss": 0.3109, "step": 479 }, { "epoch": 1.9443037974683546, - "grad_norm": 0.3494257465196283, + "grad_norm": 0.35385791064965955, "learning_rate": 3.0583941219452016e-05, - "loss": 0.344, + "loss": 0.314, "step": 480 }, { "epoch": 1.9483544303797469, - "grad_norm": 0.3607717714411562, + "grad_norm": 0.35131466246152804, "learning_rate": 3.053573898718618e-05, - "loss": 0.3187, + "loss": 0.2942, "step": 481 }, { "epoch": 1.9524050632911392, - "grad_norm": 0.32051492219297506, + "grad_norm": 0.3882910092558706, "learning_rate": 3.0487451901526956e-05, - "loss": 0.3413, + "loss": 0.3218, "step": 482 }, { "epoch": 1.9564556962025317, - "grad_norm": 0.30019651476693404, + "grad_norm": 0.3427505221427456, "learning_rate": 3.0439080351371875e-05, - "loss": 0.3129, + "loss": 0.2968, "step": 483 }, { "epoch": 1.9605063291139242, - "grad_norm": 0.3036529423978121, + "grad_norm": 0.3819014089455782, "learning_rate": 3.0390624726298764e-05, - "loss": 0.3237, + "loss": 0.3068, "step": 484 }, { "epoch": 1.9645569620253165, - "grad_norm": 0.30334234230045953, + "grad_norm": 0.33721238680975085, "learning_rate": 3.034208541656255e-05, - "loss": 0.3334, + "loss": 0.2962, "step": 485 }, { "epoch": 1.9686075949367088, - "grad_norm": 0.32781482542978363, + "grad_norm": 0.3391141070069093, "learning_rate": 3.029346281309218e-05, - "loss": 0.3358, + "loss": 0.3039, "step": 486 }, { "epoch": 1.9726582278481013, - "grad_norm": 0.3101702669425864, + "grad_norm": 0.3272795261940452, "learning_rate": 3.0244757307487415e-05, - "loss": 0.3467, + "loss": 0.3013, "step": 487 }, { "epoch": 1.9767088607594938, - "grad_norm": 0.2953995458591134, + "grad_norm": 0.3062386641155375, "learning_rate": 3.019596929201569e-05, - "loss": 0.2903, + "loss": 0.2863, "step": 488 }, { "epoch": 1.980759493670886, - "grad_norm": 0.32037284079919737, + "grad_norm": 0.3475874933998762, "learning_rate": 3.0147099159608985e-05, - "loss": 0.3388, + "loss": 0.3092, "step": 489 }, { "epoch": 1.9848101265822784, - "grad_norm": 0.3132139783897028, + "grad_norm": 0.32914917810414906, "learning_rate": 3.0098147303860616e-05, - "loss": 0.3258, + "loss": 0.301, "step": 490 }, { "epoch": 1.988860759493671, - "grad_norm": 0.30546925968051764, + "grad_norm": 0.32867483006849474, "learning_rate": 3.0049114119022117e-05, - "loss": 0.3349, + "loss": 0.3022, "step": 491 }, { "epoch": 1.9929113924050634, - "grad_norm": 0.2967705001941645, + "grad_norm": 0.32880097025368576, "learning_rate": 3.0000000000000004e-05, - "loss": 0.3201, + "loss": 0.2939, "step": 492 }, { "epoch": 1.9969620253164557, - "grad_norm": 0.3372610700491939, + "grad_norm": 0.3381942447692117, "learning_rate": 2.995080534235264e-05, - "loss": 0.3239, + "loss": 0.2969, "step": 493 }, { "epoch": 2.001012658227848, - "grad_norm": 0.4046252245255199, + "grad_norm": 0.3186874182696045, "learning_rate": 2.9901530542287044e-05, - "loss": 0.3451, + "loss": 0.2839, "step": 494 }, { "epoch": 2.0050632911392405, - "grad_norm": 0.4150050845873609, + "grad_norm": 0.4202143661111709, "learning_rate": 2.9852175996655676e-05, - "loss": 0.2456, + "loss": 0.2299, "step": 495 }, { "epoch": 2.009113924050633, - "grad_norm": 0.38471604712194507, + "grad_norm": 0.421940030363095, "learning_rate": 2.980274210295326e-05, - "loss": 0.2694, + "loss": 0.2529, "step": 496 }, { "epoch": 2.013164556962025, - "grad_norm": 0.5798377514090283, + "grad_norm": 0.5029481786190206, "learning_rate": 2.9753229259313578e-05, - "loss": 0.254, + "loss": 0.2316, "step": 497 }, { "epoch": 2.0172151898734176, - "grad_norm": 0.33327827195844034, + "grad_norm": 0.38965584880344234, "learning_rate": 2.9703637864506274e-05, - "loss": 0.2435, + "loss": 0.225, "step": 498 }, { "epoch": 2.02126582278481, - "grad_norm": 0.48878210284318513, + "grad_norm": 0.36436934740828836, "learning_rate": 2.965396831793362e-05, - "loss": 0.2554, + "loss": 0.2322, "step": 499 }, { "epoch": 2.0253164556962027, - "grad_norm": 0.3586394446978556, + "grad_norm": 0.36999122350211794, "learning_rate": 2.9604221019627316e-05, - "loss": 0.2436, + "loss": 0.2298, "step": 500 }, { "epoch": 2.0293670886075947, - "grad_norm": 0.5331203883263865, + "grad_norm": 0.3716459623187574, "learning_rate": 2.955439637024526e-05, - "loss": 0.2674, + "loss": 0.2237, "step": 501 }, { "epoch": 2.0334177215189873, - "grad_norm": 0.367956953832359, + "grad_norm": 0.40076888138912475, "learning_rate": 2.9504494771068334e-05, - "loss": 0.2504, + "loss": 0.2233, "step": 502 }, { "epoch": 2.0374683544303798, - "grad_norm": 0.42082173533029543, + "grad_norm": 0.3540927619182084, "learning_rate": 2.9454516623997156e-05, - "loss": 0.2444, + "loss": 0.2154, "step": 503 }, { "epoch": 2.0415189873417723, - "grad_norm": 0.36879493097660254, + "grad_norm": 0.442313865919713, "learning_rate": 2.9404462331548847e-05, - "loss": 0.2562, + "loss": 0.2224, "step": 504 }, { "epoch": 2.0455696202531644, - "grad_norm": 0.35551548418544754, + "grad_norm": 0.3261808547505268, "learning_rate": 2.93543322968538e-05, - "loss": 0.2453, + "loss": 0.2343, "step": 505 }, { "epoch": 2.049620253164557, - "grad_norm": 0.36385628342647847, + "grad_norm": 0.422051468390874, "learning_rate": 2.9304126923652428e-05, - "loss": 0.2559, + "loss": 0.2241, "step": 506 }, { "epoch": 2.0536708860759494, - "grad_norm": 0.3193896948361773, + "grad_norm": 0.35311825171712996, "learning_rate": 2.9253846616291896e-05, - "loss": 0.2585, + "loss": 0.2354, "step": 507 }, { "epoch": 2.057721518987342, - "grad_norm": 0.34539734216155527, + "grad_norm": 0.3912566361021335, "learning_rate": 2.9203491779722896e-05, - "loss": 0.2411, + "loss": 0.2166, "step": 508 }, { "epoch": 2.061772151898734, - "grad_norm": 0.30890737038928234, + "grad_norm": 0.35866533690807656, "learning_rate": 2.9153062819496357e-05, - "loss": 0.2499, + "loss": 0.2198, "step": 509 }, { "epoch": 2.0658227848101265, - "grad_norm": 0.37854916678973605, + "grad_norm": 0.41608012138400047, "learning_rate": 2.9102560141760178e-05, - "loss": 0.2627, + "loss": 0.2375, "step": 510 }, { "epoch": 2.069873417721519, - "grad_norm": 0.34653154523701657, + "grad_norm": 0.3077940784724969, "learning_rate": 2.9051984153256004e-05, - "loss": 0.2579, + "loss": 0.2362, "step": 511 }, { "epoch": 2.0739240506329115, - "grad_norm": 0.39835237280930447, + "grad_norm": 0.42065743908520087, "learning_rate": 2.900133526131588e-05, - "loss": 0.2822, + "loss": 0.2403, "step": 512 }, { "epoch": 2.0779746835443036, - "grad_norm": 0.3060241854645988, + "grad_norm": 0.33933903281834593, "learning_rate": 2.8950613873859025e-05, - "loss": 0.2331, + "loss": 0.2303, "step": 513 }, { "epoch": 2.082025316455696, - "grad_norm": 0.3511646008877787, + "grad_norm": 0.342172208506659, "learning_rate": 2.8899820399388515e-05, - "loss": 0.2651, + "loss": 0.2423, "step": 514 }, { "epoch": 2.0860759493670886, - "grad_norm": 0.33216249979251017, + "grad_norm": 0.382036033143454, "learning_rate": 2.8848955246988012e-05, - "loss": 0.2613, + "loss": 0.2305, "step": 515 }, { "epoch": 2.090126582278481, - "grad_norm": 0.2994351289211745, + "grad_norm": 0.32557128773846483, "learning_rate": 2.879801882631847e-05, - "loss": 0.2525, + "loss": 0.2218, "step": 516 }, { "epoch": 2.094177215189873, - "grad_norm": 0.37025121666194, + "grad_norm": 0.3767529172465701, "learning_rate": 2.8747011547614808e-05, - "loss": 0.2744, + "loss": 0.2394, "step": 517 }, { "epoch": 2.0982278481012657, - "grad_norm": 0.2961171760511582, + "grad_norm": 0.3867110994320598, "learning_rate": 2.8695933821682635e-05, - "loss": 0.2264, + "loss": 0.2211, "step": 518 }, { "epoch": 2.1022784810126582, - "grad_norm": 0.3462065704666768, + "grad_norm": 0.3108020881301015, "learning_rate": 2.864478605989494e-05, - "loss": 0.2494, + "loss": 0.2334, "step": 519 }, { "epoch": 2.1063291139240508, - "grad_norm": 0.32502605391332046, + "grad_norm": 0.343336721151281, "learning_rate": 2.8593568674188765e-05, - "loss": 0.2616, + "loss": 0.236, "step": 520 }, { "epoch": 2.110379746835443, - "grad_norm": 0.3635631513050799, + "grad_norm": 0.3667039556169178, "learning_rate": 2.8542282077061892e-05, - "loss": 0.2719, + "loss": 0.2318, "step": 521 }, { "epoch": 2.1144303797468353, - "grad_norm": 0.35671048059125665, + "grad_norm": 0.34807687715788505, "learning_rate": 2.8490926681569523e-05, - "loss": 0.2591, + "loss": 0.2289, "step": 522 }, { "epoch": 2.118481012658228, - "grad_norm": 0.33499761803781986, + "grad_norm": 0.35076105289609055, "learning_rate": 2.8439502901320956e-05, - "loss": 0.2697, + "loss": 0.2443, "step": 523 }, { "epoch": 2.1225316455696204, - "grad_norm": 0.39690317102329165, + "grad_norm": 0.34075041057557043, "learning_rate": 2.8388011150476237e-05, - "loss": 0.2413, + "loss": 0.2205, "step": 524 }, { "epoch": 2.1265822784810124, - "grad_norm": 0.36784487645574165, + "grad_norm": 0.3818341449276432, "learning_rate": 2.8336451843742866e-05, - "loss": 0.267, + "loss": 0.2326, "step": 525 }, { "epoch": 2.130632911392405, - "grad_norm": 0.3677822697005792, + "grad_norm": 0.3003569304567432, "learning_rate": 2.8284825396372387e-05, - "loss": 0.2436, + "loss": 0.2191, "step": 526 }, { "epoch": 2.1346835443037975, - "grad_norm": 0.3451056263641965, + "grad_norm": 0.370402792510317, "learning_rate": 2.8233132224157132e-05, - "loss": 0.2395, + "loss": 0.2317, "step": 527 }, { "epoch": 2.13873417721519, - "grad_norm": 0.3590220837203755, + "grad_norm": 0.3279957066011223, "learning_rate": 2.8181372743426805e-05, - "loss": 0.2609, + "loss": 0.2131, "step": 528 }, { "epoch": 2.1427848101265825, - "grad_norm": 0.36523627929303726, + "grad_norm": 0.33356932643747456, "learning_rate": 2.8129547371045128e-05, - "loss": 0.2666, + "loss": 0.2322, "step": 529 }, { "epoch": 2.1468354430379746, - "grad_norm": 0.3183410967156122, + "grad_norm": 0.3583470956277111, "learning_rate": 2.8077656524406534e-05, - "loss": 0.2443, + "loss": 0.2182, "step": 530 }, { "epoch": 2.150886075949367, - "grad_norm": 0.3368079067797023, + "grad_norm": 0.33005145912038103, "learning_rate": 2.802570062143278e-05, - "loss": 0.2322, + "loss": 0.2147, "step": 531 }, { "epoch": 2.1549367088607596, - "grad_norm": 0.3388024983429746, + "grad_norm": 0.33952235565565364, "learning_rate": 2.7973680080569555e-05, - "loss": 0.2731, + "loss": 0.2445, "step": 532 }, { "epoch": 2.1589873417721517, - "grad_norm": 0.31093277972451266, + "grad_norm": 0.3982486914977662, "learning_rate": 2.792159532078314e-05, - "loss": 0.2515, + "loss": 0.235, "step": 533 }, { "epoch": 2.163037974683544, - "grad_norm": 0.343578061692914, + "grad_norm": 0.31376903305711296, "learning_rate": 2.7869446761557033e-05, - "loss": 0.2725, + "loss": 0.2305, "step": 534 }, { "epoch": 2.1670886075949367, - "grad_norm": 0.29726541936162626, + "grad_norm": 0.40324962441140294, "learning_rate": 2.781723482288857e-05, - "loss": 0.2381, + "loss": 0.2231, "step": 535 }, { "epoch": 2.1711392405063292, - "grad_norm": 0.32238160001797006, + "grad_norm": 0.30543611061645193, "learning_rate": 2.7764959925285517e-05, - "loss": 0.2662, + "loss": 0.2304, "step": 536 }, { "epoch": 2.1751898734177217, - "grad_norm": 0.283151907370888, + "grad_norm": 0.36582683865338333, "learning_rate": 2.771262248976272e-05, - "loss": 0.2446, + "loss": 0.2247, "step": 537 }, { "epoch": 2.179240506329114, - "grad_norm": 0.3268722556848409, + "grad_norm": 0.30045849876187175, "learning_rate": 2.7660222937838677e-05, - "loss": 0.2721, + "loss": 0.2365, "step": 538 }, { "epoch": 2.1832911392405063, - "grad_norm": 0.29882391331952946, + "grad_norm": 0.3310106943877157, "learning_rate": 2.7607761691532186e-05, - "loss": 0.2416, + "loss": 0.2296, "step": 539 }, { "epoch": 2.187341772151899, - "grad_norm": 0.33162542993889255, + "grad_norm": 0.3222424600684214, "learning_rate": 2.7555239173358916e-05, - "loss": 0.2609, + "loss": 0.2253, "step": 540 }, { "epoch": 2.191392405063291, - "grad_norm": 0.29017685368595025, + "grad_norm": 0.3331263485269118, "learning_rate": 2.7502655806328e-05, - "loss": 0.2578, + "loss": 0.2177, "step": 541 }, { "epoch": 2.1954430379746834, - "grad_norm": 0.29424523232667593, + "grad_norm": 0.3226078619705582, "learning_rate": 2.7450012013938648e-05, - "loss": 0.2302, + "loss": 0.2181, "step": 542 }, { "epoch": 2.199493670886076, - "grad_norm": 0.30423303487207193, + "grad_norm": 0.34647964811167686, "learning_rate": 2.739730822017673e-05, - "loss": 0.2387, + "loss": 0.2174, "step": 543 }, { "epoch": 2.2035443037974685, - "grad_norm": 0.32799250435044763, + "grad_norm": 0.3100039945841827, "learning_rate": 2.7344544849511355e-05, - "loss": 0.2615, + "loss": 0.227, "step": 544 }, { "epoch": 2.207594936708861, - "grad_norm": 0.3353697519741972, + "grad_norm": 0.35351912932360496, "learning_rate": 2.7291722326891456e-05, - "loss": 0.2683, + "loss": 0.2406, "step": 545 }, { "epoch": 2.211645569620253, - "grad_norm": 0.3170241881136372, + "grad_norm": 0.34145910345911307, "learning_rate": 2.723884107774236e-05, - "loss": 0.2448, + "loss": 0.2246, "step": 546 }, { "epoch": 2.2156962025316456, - "grad_norm": 0.34504988860437663, + "grad_norm": 0.34103007269174834, "learning_rate": 2.718590152796239e-05, - "loss": 0.2658, + "loss": 0.238, "step": 547 }, { "epoch": 2.219746835443038, - "grad_norm": 0.2908427804284906, + "grad_norm": 0.37752222785639417, "learning_rate": 2.71329041039194e-05, - "loss": 0.2359, + "loss": 0.2146, "step": 548 }, { "epoch": 2.22379746835443, - "grad_norm": 0.3174868953434826, + "grad_norm": 0.3106685632570866, "learning_rate": 2.7079849232447357e-05, - "loss": 0.2447, + "loss": 0.2255, "step": 549 }, { "epoch": 2.2278481012658227, - "grad_norm": 0.3057338624112918, + "grad_norm": 0.3565144379699296, "learning_rate": 2.7026737340842895e-05, - "loss": 0.273, + "loss": 0.2468, "step": 550 }, { "epoch": 2.231898734177215, - "grad_norm": 0.28375422484871393, + "grad_norm": 0.32066486779989345, "learning_rate": 2.697356885686189e-05, - "loss": 0.2429, + "loss": 0.2213, "step": 551 }, { "epoch": 2.2359493670886077, - "grad_norm": 0.30731724652711795, + "grad_norm": 0.3571471990643203, "learning_rate": 2.6920344208716014e-05, - "loss": 0.2554, + "loss": 0.234, "step": 552 }, { "epoch": 2.24, - "grad_norm": 0.3725433655939576, + "grad_norm": 0.34387486472214907, "learning_rate": 2.6867063825069252e-05, - "loss": 0.2898, + "loss": 0.244, "step": 553 }, { "epoch": 2.2440506329113923, - "grad_norm": 0.29080583776887126, + "grad_norm": 0.3183468678976038, "learning_rate": 2.6813728135034494e-05, - "loss": 0.2517, + "loss": 0.2292, "step": 554 }, { "epoch": 2.248101265822785, - "grad_norm": 0.3707071512669043, + "grad_norm": 0.3517469907430876, "learning_rate": 2.6760337568170056e-05, - "loss": 0.2555, + "loss": 0.2334, "step": 555 }, { "epoch": 2.2521518987341773, - "grad_norm": 0.32721071565123966, + "grad_norm": 0.3139780522459836, "learning_rate": 2.6706892554476226e-05, - "loss": 0.2599, + "loss": 0.2319, "step": 556 }, { "epoch": 2.2562025316455694, - "grad_norm": 0.29910260257848165, + "grad_norm": 0.3319448150834173, "learning_rate": 2.6653393524391795e-05, - "loss": 0.2542, + "loss": 0.2319, "step": 557 }, { "epoch": 2.260253164556962, - "grad_norm": 0.3547135342557573, + "grad_norm": 0.3474796980649211, "learning_rate": 2.6599840908790592e-05, - "loss": 0.2681, + "loss": 0.2391, "step": 558 }, { "epoch": 2.2643037974683544, - "grad_norm": 0.2870982018583257, + "grad_norm": 0.31059297441143163, "learning_rate": 2.6546235138978028e-05, - "loss": 0.2195, + "loss": 0.2015, "step": 559 }, { "epoch": 2.268354430379747, - "grad_norm": 0.35216881703424185, + "grad_norm": 0.3358312632552793, "learning_rate": 2.6492576646687597e-05, - "loss": 0.2675, + "loss": 0.2276, "step": 560 }, { "epoch": 2.2724050632911394, - "grad_norm": 0.32691910655303635, + "grad_norm": 0.37064924468074395, "learning_rate": 2.6438865864077425e-05, - "loss": 0.2445, + "loss": 0.2224, "step": 561 }, { "epoch": 2.2764556962025315, - "grad_norm": 0.3353940648982948, + "grad_norm": 0.3003172244276441, "learning_rate": 2.6385103223726766e-05, - "loss": 0.2591, + "loss": 0.2394, "step": 562 }, { "epoch": 2.280506329113924, - "grad_norm": 0.33139857272772094, + "grad_norm": 0.3540759021212988, "learning_rate": 2.6331289158632537e-05, - "loss": 0.2595, + "loss": 0.232, "step": 563 }, { "epoch": 2.2845569620253166, - "grad_norm": 0.29143584764577435, + "grad_norm": 0.3343498062647225, "learning_rate": 2.6277424102205817e-05, - "loss": 0.2572, + "loss": 0.2349, "step": 564 }, { "epoch": 2.2886075949367086, - "grad_norm": 0.33792995899768175, + "grad_norm": 0.3205503466207593, "learning_rate": 2.6223508488268374e-05, - "loss": 0.2423, + "loss": 0.2348, "step": 565 }, { "epoch": 2.292658227848101, - "grad_norm": 0.3041558336336573, + "grad_norm": 0.34886547389591654, "learning_rate": 2.6169542751049148e-05, - "loss": 0.2563, + "loss": 0.2217, "step": 566 }, { "epoch": 2.2967088607594937, - "grad_norm": 0.3332343052878762, + "grad_norm": 0.30657664739681834, "learning_rate": 2.6115527325180754e-05, - "loss": 0.2819, + "loss": 0.2411, "step": 567 }, { "epoch": 2.300759493670886, - "grad_norm": 0.29530331886078354, + "grad_norm": 0.36195707990008286, "learning_rate": 2.606146264569603e-05, - "loss": 0.2181, + "loss": 0.2075, "step": 568 }, { "epoch": 2.3048101265822787, - "grad_norm": 0.34018510431561616, + "grad_norm": 0.3312578144682574, "learning_rate": 2.6007349148024447e-05, - "loss": 0.2576, + "loss": 0.2279, "step": 569 }, { "epoch": 2.3088607594936708, - "grad_norm": 0.3371237014164785, + "grad_norm": 0.32303200348689703, "learning_rate": 2.5953187267988694e-05, - "loss": 0.2637, + "loss": 0.2263, "step": 570 }, { "epoch": 2.3129113924050633, - "grad_norm": 0.3287329666594749, + "grad_norm": 0.3539564156273888, "learning_rate": 2.5898977441801097e-05, - "loss": 0.2514, + "loss": 0.2238, "step": 571 }, { "epoch": 2.316962025316456, - "grad_norm": 0.2949771936382736, + "grad_norm": 0.3175429448707806, "learning_rate": 2.584472010606015e-05, - "loss": 0.2363, + "loss": 0.2196, "step": 572 }, { "epoch": 2.321012658227848, - "grad_norm": 0.3516586499800208, + "grad_norm": 0.3427743621699918, "learning_rate": 2.5790415697746976e-05, - "loss": 0.2503, + "loss": 0.2231, "step": 573 }, { "epoch": 2.3250632911392404, - "grad_norm": 0.31837159727025294, + "grad_norm": 0.33047355749585616, "learning_rate": 2.5736064654221808e-05, - "loss": 0.2749, + "loss": 0.2343, "step": 574 }, { "epoch": 2.329113924050633, - "grad_norm": 0.3627220008963462, + "grad_norm": 0.3084207181707558, "learning_rate": 2.568166741322048e-05, - "loss": 0.2669, + "loss": 0.2313, "step": 575 }, { "epoch": 2.3331645569620254, - "grad_norm": 0.25534528847832944, + "grad_norm": 0.29696309945301413, "learning_rate": 2.56272244128509e-05, - "loss": 0.2455, + "loss": 0.2247, "step": 576 }, { "epoch": 2.337215189873418, - "grad_norm": 0.33794578222795, + "grad_norm": 0.33442276560415224, "learning_rate": 2.55727360915895e-05, - "loss": 0.2491, + "loss": 0.2307, "step": 577 }, { "epoch": 2.34126582278481, - "grad_norm": 0.28970914869478004, + "grad_norm": 0.33916975183181786, "learning_rate": 2.5518202888277734e-05, - "loss": 0.2493, + "loss": 0.2263, "step": 578 }, { "epoch": 2.3453164556962025, - "grad_norm": 0.32705408572972394, + "grad_norm": 0.32041529521824186, "learning_rate": 2.5463625242118523e-05, - "loss": 0.265, + "loss": 0.2378, "step": 579 }, { "epoch": 2.349367088607595, - "grad_norm": 0.3255449600003331, + "grad_norm": 0.3363192682354492, "learning_rate": 2.5409003592672723e-05, - "loss": 0.2688, + "loss": 0.232, "step": 580 }, { "epoch": 2.353417721518987, - "grad_norm": 0.30211849745016767, + "grad_norm": 0.3125051262879329, "learning_rate": 2.535433837985559e-05, - "loss": 0.2478, + "loss": 0.229, "step": 581 }, { "epoch": 2.3574683544303796, - "grad_norm": 0.35291408663161034, + "grad_norm": 0.35146861684417946, "learning_rate": 2.529963004393324e-05, - "loss": 0.2655, + "loss": 0.2295, "step": 582 }, { "epoch": 2.361518987341772, - "grad_norm": 0.2962013484628534, + "grad_norm": 0.3343526977980359, "learning_rate": 2.524487902551908e-05, - "loss": 0.2664, + "loss": 0.2456, "step": 583 }, { "epoch": 2.3655696202531646, - "grad_norm": 0.33442287897133444, + "grad_norm": 0.3344390038543155, "learning_rate": 2.519008576557029e-05, - "loss": 0.2508, + "loss": 0.2331, "step": 584 }, { "epoch": 2.369620253164557, - "grad_norm": 0.2861604866490563, + "grad_norm": 0.37276725143202716, "learning_rate": 2.5135250705384254e-05, - "loss": 0.2511, + "loss": 0.2388, "step": 585 }, { "epoch": 2.3736708860759492, - "grad_norm": 0.33168305947616183, + "grad_norm": 0.3038000050999679, "learning_rate": 2.5080374286595007e-05, - "loss": 0.2759, + "loss": 0.2429, "step": 586 }, { "epoch": 2.3777215189873417, - "grad_norm": 0.3011371416253627, + "grad_norm": 0.3837563815303415, "learning_rate": 2.5025456951169677e-05, - "loss": 0.2799, + "loss": 0.246, "step": 587 }, { "epoch": 2.3817721518987343, - "grad_norm": 0.30372447018537707, + "grad_norm": 0.3052384603062137, "learning_rate": 2.4970499141404942e-05, - "loss": 0.2461, + "loss": 0.2313, "step": 588 }, { "epoch": 2.3858227848101268, - "grad_norm": 0.3103315984870202, + "grad_norm": 0.3427864640613424, "learning_rate": 2.491550129992345e-05, - "loss": 0.2828, + "loss": 0.2378, "step": 589 }, { "epoch": 2.389873417721519, - "grad_norm": 0.2840288940854976, + "grad_norm": 0.3377584439736947, "learning_rate": 2.486046386967024e-05, - "loss": 0.2331, + "loss": 0.23, "step": 590 }, { "epoch": 2.3939240506329114, - "grad_norm": 0.39594622336656315, + "grad_norm": 0.3209817438957892, "learning_rate": 2.4805387293909214e-05, - "loss": 0.2843, + "loss": 0.2407, "step": 591 }, { "epoch": 2.397974683544304, - "grad_norm": 0.3067480014546156, + "grad_norm": 0.3396306654692784, "learning_rate": 2.4750272016219552e-05, - "loss": 0.2391, + "loss": 0.2245, "step": 592 }, { "epoch": 2.4020253164556964, - "grad_norm": 0.3565936674977987, + "grad_norm": 0.3044799039387763, "learning_rate": 2.4695118480492114e-05, - "loss": 0.2875, + "loss": 0.2395, "step": 593 }, { "epoch": 2.4060759493670885, - "grad_norm": 0.3253647340257518, + "grad_norm": 0.3244853017704809, "learning_rate": 2.4639927130925898e-05, - "loss": 0.2414, + "loss": 0.2267, "step": 594 }, { "epoch": 2.410126582278481, - "grad_norm": 0.3223980277187895, + "grad_norm": 0.3242814335344009, "learning_rate": 2.458469841202444e-05, - "loss": 0.2407, + "loss": 0.2186, "step": 595 }, { "epoch": 2.4141772151898735, - "grad_norm": 0.3394308689906547, + "grad_norm": 0.29210813570125793, "learning_rate": 2.452943276859226e-05, - "loss": 0.278, + "loss": 0.2329, "step": 596 }, { "epoch": 2.418227848101266, - "grad_norm": 0.3023923384775553, + "grad_norm": 0.3667952132266445, "learning_rate": 2.447413064573125e-05, - "loss": 0.2589, + "loss": 0.2389, "step": 597 }, { "epoch": 2.422278481012658, - "grad_norm": 0.35179888366770046, + "grad_norm": 0.3236306693329527, "learning_rate": 2.4418792488837095e-05, - "loss": 0.2714, + "loss": 0.2373, "step": 598 }, { "epoch": 2.4263291139240506, - "grad_norm": 0.2986972363503083, + "grad_norm": 0.34364296794999205, "learning_rate": 2.4363418743595713e-05, - "loss": 0.2406, + "loss": 0.2265, "step": 599 }, { "epoch": 2.430379746835443, - "grad_norm": 0.31759201514464247, + "grad_norm": 0.3178166876200261, "learning_rate": 2.430800985597963e-05, - "loss": 0.26, + "loss": 0.2324, "step": 600 }, { "epoch": 2.4344303797468356, - "grad_norm": 0.35619379754231556, + "grad_norm": 0.33390152246493937, "learning_rate": 2.4252566272244415e-05, - "loss": 0.2656, + "loss": 0.2327, "step": 601 }, { "epoch": 2.4384810126582277, - "grad_norm": 0.3128710940616226, + "grad_norm": 0.3132143786510946, "learning_rate": 2.4197088438925063e-05, - "loss": 0.2496, + "loss": 0.2249, "step": 602 }, { "epoch": 2.44253164556962, - "grad_norm": 0.3356968194686584, + "grad_norm": 0.3144053814144751, "learning_rate": 2.4141576802832417e-05, - "loss": 0.2673, + "loss": 0.2316, "step": 603 }, { "epoch": 2.4465822784810127, - "grad_norm": 0.27777988173298446, + "grad_norm": 0.3066758354289081, "learning_rate": 2.408603181104957e-05, - "loss": 0.2247, + "loss": 0.2127, "step": 604 }, { "epoch": 2.4506329113924052, - "grad_norm": 0.3260273346966271, + "grad_norm": 0.33590284205673726, "learning_rate": 2.4030453910928245e-05, - "loss": 0.2573, + "loss": 0.2283, "step": 605 }, { "epoch": 2.4546835443037973, - "grad_norm": 0.32035663678079257, + "grad_norm": 0.2929689449460151, "learning_rate": 2.397484355008521e-05, - "loss": 0.2821, + "loss": 0.2275, "step": 606 }, { "epoch": 2.45873417721519, - "grad_norm": 0.31105737723006815, + "grad_norm": 0.3317807036650442, "learning_rate": 2.3919201176398662e-05, - "loss": 0.2392, + "loss": 0.2226, "step": 607 }, { "epoch": 2.4627848101265823, - "grad_norm": 0.334896418273189, + "grad_norm": 0.32008657883466285, "learning_rate": 2.3863527238004633e-05, - "loss": 0.2627, + "loss": 0.2316, "step": 608 }, { "epoch": 2.466835443037975, - "grad_norm": 0.27698900441462093, + "grad_norm": 0.3114178326090441, "learning_rate": 2.380782218329337e-05, - "loss": 0.2491, + "loss": 0.234, "step": 609 }, { "epoch": 2.470886075949367, - "grad_norm": 0.3525181380254263, + "grad_norm": 0.3233869789345488, "learning_rate": 2.3752086460905725e-05, - "loss": 0.2797, + "loss": 0.234, "step": 610 }, { "epoch": 2.4749367088607594, - "grad_norm": 0.32636669987925004, + "grad_norm": 0.3309863242340352, "learning_rate": 2.3696320519729544e-05, - "loss": 0.2537, + "loss": 0.2417, "step": 611 }, { "epoch": 2.478987341772152, - "grad_norm": 0.32264162249146944, + "grad_norm": 0.32844562271316874, "learning_rate": 2.3640524808896045e-05, - "loss": 0.2602, + "loss": 0.2348, "step": 612 }, { "epoch": 2.4830379746835445, - "grad_norm": 0.3449844150369798, + "grad_norm": 0.3369904678729884, "learning_rate": 2.3584699777776222e-05, - "loss": 0.2679, + "loss": 0.2398, "step": 613 }, { "epoch": 2.4870886075949366, - "grad_norm": 0.3324959093905074, + "grad_norm": 0.2860776951333733, "learning_rate": 2.3528845875977195e-05, - "loss": 0.2648, + "loss": 0.2324, "step": 614 }, { "epoch": 2.491139240506329, - "grad_norm": 0.30074641396376595, + "grad_norm": 0.32477149257158855, "learning_rate": 2.3472963553338614e-05, - "loss": 0.2453, + "loss": 0.2261, "step": 615 }, { "epoch": 2.4951898734177216, - "grad_norm": 0.29845749150847506, + "grad_norm": 0.29667596682905584, "learning_rate": 2.341705325992901e-05, - "loss": 0.2609, + "loss": 0.2368, "step": 616 }, { "epoch": 2.499240506329114, - "grad_norm": 0.3054496699497279, + "grad_norm": 0.3024219913294697, "learning_rate": 2.336111544604222e-05, - "loss": 0.2591, + "loss": 0.2323, "step": 617 }, { "epoch": 2.503291139240506, - "grad_norm": 0.34078679299867914, + "grad_norm": 0.3136508071085885, "learning_rate": 2.33051505621937e-05, - "loss": 0.2801, + "loss": 0.2395, "step": 618 }, { "epoch": 2.5073417721518987, - "grad_norm": 0.2703066553108464, + "grad_norm": 0.29577653614805455, "learning_rate": 2.324915905911693e-05, - "loss": 0.2466, + "loss": 0.2338, "step": 619 }, { "epoch": 2.511392405063291, - "grad_norm": 0.3079796966037467, + "grad_norm": 0.30366453062617665, "learning_rate": 2.319314138775977e-05, - "loss": 0.2673, + "loss": 0.2435, "step": 620 }, { "epoch": 2.5154430379746833, - "grad_norm": 0.3208622250808163, + "grad_norm": 0.31821408416911096, "learning_rate": 2.3137097999280856e-05, - "loss": 0.2771, + "loss": 0.2581, "step": 621 }, { "epoch": 2.519493670886076, - "grad_norm": 0.2788353636909791, + "grad_norm": 0.28811472658974696, "learning_rate": 2.308102934504593e-05, - "loss": 0.2487, + "loss": 0.2273, "step": 622 }, { "epoch": 2.5235443037974683, - "grad_norm": 0.33726732315132163, + "grad_norm": 0.32020178339840244, "learning_rate": 2.3024935876624222e-05, - "loss": 0.2599, + "loss": 0.2391, "step": 623 }, { "epoch": 2.527594936708861, - "grad_norm": 0.30914415723886735, + "grad_norm": 0.307084993447976, "learning_rate": 2.2968818045784813e-05, - "loss": 0.2544, + "loss": 0.233, "step": 624 }, { "epoch": 2.5316455696202533, - "grad_norm": 0.32286450144101736, + "grad_norm": 0.29015370749820446, "learning_rate": 2.2912676304493006e-05, - "loss": 0.2794, + "loss": 0.2489, "step": 625 }, { "epoch": 2.5356962025316454, - "grad_norm": 0.30270556944694954, + "grad_norm": 0.2999431888526454, "learning_rate": 2.2856511104906668e-05, - "loss": 0.2515, + "loss": 0.2259, "step": 626 }, { "epoch": 2.539746835443038, - "grad_norm": 0.3317981804953664, + "grad_norm": 0.3317474664144101, "learning_rate": 2.2800322899372586e-05, - "loss": 0.2442, + "loss": 0.2194, "step": 627 }, { "epoch": 2.5437974683544304, - "grad_norm": 0.3058430454231997, + "grad_norm": 0.2834874033188405, "learning_rate": 2.2744112140422844e-05, - "loss": 0.2644, + "loss": 0.2338, "step": 628 }, { "epoch": 2.547848101265823, - "grad_norm": 0.33300172744223905, + "grad_norm": 0.29730279435675405, "learning_rate": 2.2687879280771177e-05, - "loss": 0.2552, + "loss": 0.2324, "step": 629 }, { "epoch": 2.5518987341772155, - "grad_norm": 0.3164389820295178, + "grad_norm": 0.3114087763338251, "learning_rate": 2.26316247733093e-05, - "loss": 0.2436, + "loss": 0.2335, "step": 630 }, { "epoch": 2.5559493670886075, - "grad_norm": 0.29422334988584836, + "grad_norm": 0.28827782752869485, "learning_rate": 2.257534907110328e-05, - "loss": 0.2479, + "loss": 0.2294, "step": 631 }, { "epoch": 2.56, - "grad_norm": 0.3363439072154995, + "grad_norm": 0.28708836411935695, "learning_rate": 2.2519052627389882e-05, - "loss": 0.2717, + "loss": 0.2395, "step": 632 }, { "epoch": 2.5640506329113926, - "grad_norm": 0.2793247405154207, + "grad_norm": 0.2887297457314167, "learning_rate": 2.246273589557294e-05, - "loss": 0.269, + "loss": 0.2416, "step": 633 }, { "epoch": 2.5681012658227846, - "grad_norm": 0.2658250459219405, + "grad_norm": 0.286262059121319, "learning_rate": 2.240639932921966e-05, - "loss": 0.2413, + "loss": 0.238, "step": 634 }, { "epoch": 2.572151898734177, - "grad_norm": 0.3036044032741859, + "grad_norm": 0.27024567578020364, "learning_rate": 2.2350043382056995e-05, - "loss": 0.2766, + "loss": 0.2466, "step": 635 }, { "epoch": 2.5762025316455697, - "grad_norm": 0.2717102916362659, + "grad_norm": 0.3025562806557872, "learning_rate": 2.2293668507968015e-05, - "loss": 0.2464, + "loss": 0.2422, "step": 636 }, { "epoch": 2.580253164556962, - "grad_norm": 0.3507821993589926, + "grad_norm": 0.28661160948762593, "learning_rate": 2.2237275160988186e-05, - "loss": 0.267, + "loss": 0.2358, "step": 637 }, { "epoch": 2.5843037974683547, - "grad_norm": 0.26318985553996144, + "grad_norm": 0.2929293274627698, "learning_rate": 2.2180863795301787e-05, - "loss": 0.2462, + "loss": 0.2236, "step": 638 }, { "epoch": 2.5883544303797468, - "grad_norm": 0.29478252487085727, + "grad_norm": 0.28877826219946695, "learning_rate": 2.212443486523819e-05, - "loss": 0.2543, + "loss": 0.2248, "step": 639 }, { "epoch": 2.5924050632911393, - "grad_norm": 0.35442328706929843, + "grad_norm": 0.2923062100120105, "learning_rate": 2.2067988825268243e-05, - "loss": 0.2713, + "loss": 0.2433, "step": 640 }, { "epoch": 2.596455696202532, - "grad_norm": 0.25188740449307107, + "grad_norm": 0.2927163765388562, "learning_rate": 2.2011526130000596e-05, - "loss": 0.245, + "loss": 0.2345, "step": 641 }, { "epoch": 2.600506329113924, - "grad_norm": 0.3606681706074205, + "grad_norm": 0.3057658365762034, "learning_rate": 2.1955047234178038e-05, - "loss": 0.2809, + "loss": 0.2389, "step": 642 }, { "epoch": 2.6045569620253164, - "grad_norm": 0.28859294577384703, + "grad_norm": 0.2972381289521406, "learning_rate": 2.1898552592673825e-05, - "loss": 0.2666, + "loss": 0.2393, "step": 643 }, { "epoch": 2.608607594936709, - "grad_norm": 0.3195754776259867, + "grad_norm": 0.3264507687670856, "learning_rate": 2.184204266048803e-05, - "loss": 0.2645, + "loss": 0.238, "step": 644 }, { "epoch": 2.6126582278481014, - "grad_norm": 0.24586424633252446, + "grad_norm": 0.3081470129365603, "learning_rate": 2.1785517892743887e-05, - "loss": 0.2256, + "loss": 0.2139, "step": 645 }, { "epoch": 2.616708860759494, - "grad_norm": 0.32451063463078383, + "grad_norm": 0.3464173287109854, "learning_rate": 2.17289787446841e-05, - "loss": 0.2797, + "loss": 0.2341, "step": 646 }, { "epoch": 2.620759493670886, - "grad_norm": 0.2957753278443552, + "grad_norm": 0.3060514959264648, "learning_rate": 2.1672425671667198e-05, - "loss": 0.253, + "loss": 0.2225, "step": 647 }, { "epoch": 2.6248101265822785, - "grad_norm": 0.29394899836350324, + "grad_norm": 0.30792753435196085, "learning_rate": 2.161585912916385e-05, - "loss": 0.2655, + "loss": 0.2463, "step": 648 }, { "epoch": 2.628860759493671, - "grad_norm": 0.27467309321030836, + "grad_norm": 0.3396525801751568, "learning_rate": 2.1559279572753214e-05, - "loss": 0.2572, + "loss": 0.2351, "step": 649 }, { "epoch": 2.632911392405063, - "grad_norm": 0.2993278596918924, + "grad_norm": 0.2911893715014043, "learning_rate": 2.1502687458119268e-05, - "loss": 0.2494, + "loss": 0.2185, "step": 650 }, { "epoch": 2.6369620253164556, - "grad_norm": 0.30155529079159765, + "grad_norm": 0.3514074632282774, "learning_rate": 2.1446083241047116e-05, - "loss": 0.2601, + "loss": 0.2239, "step": 651 }, { "epoch": 2.641012658227848, - "grad_norm": 0.2806569137233707, + "grad_norm": 0.3077437004270069, "learning_rate": 2.1389467377419333e-05, - "loss": 0.2535, + "loss": 0.2274, "step": 652 }, { "epoch": 2.6450632911392407, - "grad_norm": 0.2707567767411227, + "grad_norm": 0.29631837947427786, "learning_rate": 2.133284032321232e-05, - "loss": 0.2483, + "loss": 0.2372, "step": 653 }, { "epoch": 2.649113924050633, - "grad_norm": 0.32848236432077127, + "grad_norm": 0.3019119801912528, "learning_rate": 2.1276202534492566e-05, - "loss": 0.3001, + "loss": 0.2479, "step": 654 }, { "epoch": 2.6531645569620252, - "grad_norm": 0.27116134968834354, + "grad_norm": 0.31960189583156856, "learning_rate": 2.121955446741306e-05, - "loss": 0.2588, + "loss": 0.2338, "step": 655 }, { "epoch": 2.6572151898734178, - "grad_norm": 0.27974360662587316, + "grad_norm": 0.3187735220545649, "learning_rate": 2.1162896578209517e-05, - "loss": 0.2317, + "loss": 0.2185, "step": 656 }, { "epoch": 2.6612658227848103, - "grad_norm": 0.2823508927144383, + "grad_norm": 0.3305830311295424, "learning_rate": 2.1106229323196813e-05, - "loss": 0.251, + "loss": 0.2286, "step": 657 }, { "epoch": 2.6653164556962023, - "grad_norm": 0.2964171921577135, + "grad_norm": 0.33401267312879923, "learning_rate": 2.1049553158765214e-05, - "loss": 0.2465, + "loss": 0.2182, "step": 658 }, { "epoch": 2.669367088607595, - "grad_norm": 0.3017504392993927, + "grad_norm": 0.3173485119264029, "learning_rate": 2.0992868541376764e-05, - "loss": 0.266, + "loss": 0.2325, "step": 659 }, { "epoch": 2.6734177215189874, - "grad_norm": 0.30520598344936245, + "grad_norm": 0.3221766612292656, "learning_rate": 2.093617592756158e-05, - "loss": 0.2588, + "loss": 0.2382, "step": 660 }, { "epoch": 2.67746835443038, - "grad_norm": 0.2974226028063475, + "grad_norm": 0.3317191399796401, "learning_rate": 2.0879475773914167e-05, - "loss": 0.2621, + "loss": 0.2284, "step": 661 }, { "epoch": 2.6815189873417724, - "grad_norm": 0.3248863200182017, + "grad_norm": 0.3070225612509448, "learning_rate": 2.082276853708978e-05, - "loss": 0.2476, + "loss": 0.2298, "step": 662 }, { "epoch": 2.6855696202531645, - "grad_norm": 0.28526669745650723, + "grad_norm": 0.31677634285509093, "learning_rate": 2.076605467380071e-05, - "loss": 0.2515, + "loss": 0.2277, "step": 663 }, { "epoch": 2.689620253164557, - "grad_norm": 0.29741599133606517, + "grad_norm": 0.31824172182379085, "learning_rate": 2.0709334640812613e-05, - "loss": 0.2672, + "loss": 0.2404, "step": 664 }, { "epoch": 2.6936708860759495, - "grad_norm": 0.28402158096624375, + "grad_norm": 0.2942177303440855, "learning_rate": 2.0652608894940824e-05, - "loss": 0.2562, + "loss": 0.2314, "step": 665 }, { "epoch": 2.6977215189873416, - "grad_norm": 0.3219578959262951, + "grad_norm": 0.29885634544731843, "learning_rate": 2.0595877893046722e-05, - "loss": 0.2732, + "loss": 0.2455, "step": 666 }, { "epoch": 2.701772151898734, - "grad_norm": 0.2790950669071131, + "grad_norm": 0.304670070991131, "learning_rate": 2.0539142092033985e-05, - "loss": 0.2469, + "loss": 0.2272, "step": 667 }, { "epoch": 2.7058227848101266, - "grad_norm": 0.30015650711255587, + "grad_norm": 0.2959052841778297, "learning_rate": 2.048240194884496e-05, - "loss": 0.2525, + "loss": 0.2243, "step": 668 }, { "epoch": 2.709873417721519, - "grad_norm": 0.3257975640149779, + "grad_norm": 0.3341247923500945, "learning_rate": 2.042565792045695e-05, - "loss": 0.2669, + "loss": 0.2292, "step": 669 }, { "epoch": 2.7139240506329116, - "grad_norm": 0.29265832436832667, + "grad_norm": 0.3141099074496931, "learning_rate": 2.036891046387857e-05, - "loss": 0.2477, + "loss": 0.2354, "step": 670 }, { "epoch": 2.7179746835443037, - "grad_norm": 0.3162637745701073, + "grad_norm": 0.3242743822642043, "learning_rate": 2.0312160036146036e-05, - "loss": 0.2546, + "loss": 0.2301, "step": 671 }, { "epoch": 2.7220253164556962, - "grad_norm": 0.2905955152774222, + "grad_norm": 0.3332099266280377, "learning_rate": 2.025540709431948e-05, - "loss": 0.259, + "loss": 0.2438, "step": 672 }, { "epoch": 2.7260759493670887, - "grad_norm": 0.33203717311046016, + "grad_norm": 0.3295532005466524, "learning_rate": 2.0198652095479298e-05, - "loss": 0.2732, + "loss": 0.2536, "step": 673 }, { "epoch": 2.730126582278481, - "grad_norm": 0.30572666880079036, + "grad_norm": 0.33331789931730993, "learning_rate": 2.014189549672245e-05, - "loss": 0.2644, + "loss": 0.2385, "step": 674 }, { "epoch": 2.7341772151898733, - "grad_norm": 0.29816253173814344, + "grad_norm": 0.3324722495958164, "learning_rate": 2.0085137755158776e-05, - "loss": 0.2532, + "loss": 0.2314, "step": 675 }, { "epoch": 2.738227848101266, - "grad_norm": 0.31228487940160127, + "grad_norm": 0.3236170103075277, "learning_rate": 2.0028379327907327e-05, - "loss": 0.2658, + "loss": 0.2412, "step": 676 }, { "epoch": 2.7422784810126584, - "grad_norm": 0.30924288006334416, + "grad_norm": 0.31790139845186743, "learning_rate": 1.9971620672092676e-05, - "loss": 0.2524, + "loss": 0.2241, "step": 677 }, { "epoch": 2.746329113924051, - "grad_norm": 0.27253770440983854, + "grad_norm": 0.30257758385673583, "learning_rate": 1.991486224484123e-05, - "loss": 0.2375, + "loss": 0.2187, "step": 678 }, { "epoch": 2.750379746835443, - "grad_norm": 0.31509635581193124, + "grad_norm": 0.30373666544695865, "learning_rate": 1.985810450327756e-05, - "loss": 0.2615, + "loss": 0.2287, "step": 679 }, { "epoch": 2.7544303797468355, - "grad_norm": 0.3097795439142612, + "grad_norm": 0.3085964878334211, "learning_rate": 1.9801347904520706e-05, - "loss": 0.2641, + "loss": 0.2401, "step": 680 }, { "epoch": 2.758481012658228, - "grad_norm": 0.2568358436277696, + "grad_norm": 0.27497435294521877, "learning_rate": 1.974459290568053e-05, - "loss": 0.2247, + "loss": 0.2092, "step": 681 }, { "epoch": 2.76253164556962, - "grad_norm": 0.31728766572738426, + "grad_norm": 0.3215781053713376, "learning_rate": 1.968783996385397e-05, - "loss": 0.2681, + "loss": 0.23, "step": 682 }, { "epoch": 2.7665822784810126, - "grad_norm": 0.2740366579688183, + "grad_norm": 0.3074045343659693, "learning_rate": 1.963108953612143e-05, - "loss": 0.2564, + "loss": 0.2359, "step": 683 }, { "epoch": 2.770632911392405, - "grad_norm": 0.3030158420829032, + "grad_norm": 0.2919988041419382, "learning_rate": 1.9574342079543056e-05, - "loss": 0.2735, + "loss": 0.2412, "step": 684 }, { "epoch": 2.7746835443037976, - "grad_norm": 0.3096293881914137, + "grad_norm": 0.32384774822913615, "learning_rate": 1.9517598051155046e-05, - "loss": 0.2666, + "loss": 0.2269, "step": 685 }, { "epoch": 2.77873417721519, - "grad_norm": 0.25357502690710215, + "grad_norm": 0.31168931913535425, "learning_rate": 1.9460857907966025e-05, - "loss": 0.2408, + "loss": 0.2288, "step": 686 }, { "epoch": 2.782784810126582, - "grad_norm": 0.3176681709458035, + "grad_norm": 0.27446813238880874, "learning_rate": 1.9404122106953285e-05, - "loss": 0.2597, + "loss": 0.2366, "step": 687 }, { "epoch": 2.7868354430379747, - "grad_norm": 0.299883409081254, + "grad_norm": 0.338853314742463, "learning_rate": 1.9347391105059176e-05, - "loss": 0.2514, + "loss": 0.2222, "step": 688 }, { "epoch": 2.790886075949367, - "grad_norm": 0.3097905544152895, + "grad_norm": 0.31577529708831825, "learning_rate": 1.92906653591874e-05, - "loss": 0.2628, + "loss": 0.2372, "step": 689 }, { "epoch": 2.7949367088607593, - "grad_norm": 0.27683516167581357, + "grad_norm": 0.2858206562054224, "learning_rate": 1.9233945326199295e-05, - "loss": 0.2552, + "loss": 0.2357, "step": 690 }, { "epoch": 2.798987341772152, - "grad_norm": 0.29318931984544283, + "grad_norm": 0.31468871310816077, "learning_rate": 1.917723146291022e-05, - "loss": 0.2408, + "loss": 0.2356, "step": 691 }, { "epoch": 2.8030379746835443, - "grad_norm": 0.31118305343127434, + "grad_norm": 0.3220685990666963, "learning_rate": 1.912052422608584e-05, - "loss": 0.25, + "loss": 0.2367, "step": 692 }, { "epoch": 2.807088607594937, - "grad_norm": 0.3010635310379988, + "grad_norm": 0.3049555792340489, "learning_rate": 1.9063824072438428e-05, - "loss": 0.2969, + "loss": 0.258, "step": 693 }, { "epoch": 2.8111392405063294, - "grad_norm": 0.31085152737466465, + "grad_norm": 0.32314734654617416, "learning_rate": 1.9007131458623246e-05, - "loss": 0.2387, + "loss": 0.2266, "step": 694 }, { "epoch": 2.8151898734177214, - "grad_norm": 0.3171687062275103, + "grad_norm": 0.33811331139430184, "learning_rate": 1.895044684123479e-05, - "loss": 0.2669, + "loss": 0.2377, "step": 695 }, { "epoch": 2.819240506329114, - "grad_norm": 0.2565487304249103, + "grad_norm": 0.2901367320286777, "learning_rate": 1.8893770676803194e-05, - "loss": 0.2344, + "loss": 0.221, "step": 696 }, { "epoch": 2.8232911392405065, - "grad_norm": 0.2788542807226049, + "grad_norm": 0.2923434569382237, "learning_rate": 1.8837103421790486e-05, - "loss": 0.2452, + "loss": 0.229, "step": 697 }, { "epoch": 2.8273417721518985, - "grad_norm": 0.3151126817640209, + "grad_norm": 0.34319938178192727, "learning_rate": 1.8780445532586952e-05, - "loss": 0.2697, + "loss": 0.235, "step": 698 }, { "epoch": 2.831392405063291, - "grad_norm": 0.2917313584905453, + "grad_norm": 0.29899766300741853, "learning_rate": 1.872379746550743e-05, - "loss": 0.2478, + "loss": 0.2236, "step": 699 }, { "epoch": 2.8354430379746836, - "grad_norm": 0.31104219489330304, + "grad_norm": 0.29920529165177634, "learning_rate": 1.866715967678769e-05, - "loss": 0.259, + "loss": 0.2284, "step": 700 }, { "epoch": 2.839493670886076, - "grad_norm": 0.29826191329313684, + "grad_norm": 0.3203336061991541, "learning_rate": 1.861053262258067e-05, - "loss": 0.263, + "loss": 0.2265, "step": 701 }, { "epoch": 2.8435443037974686, - "grad_norm": 0.2788486445515333, + "grad_norm": 0.2927777622315442, "learning_rate": 1.8553916758952897e-05, - "loss": 0.2383, + "loss": 0.2205, "step": 702 }, { "epoch": 2.8475949367088607, - "grad_norm": 0.29107718462365745, + "grad_norm": 0.29665631189719466, "learning_rate": 1.8497312541880735e-05, - "loss": 0.2593, + "loss": 0.2244, "step": 703 }, { "epoch": 2.851645569620253, - "grad_norm": 0.25895963697741403, + "grad_norm": 0.278546753617254, "learning_rate": 1.8440720427246786e-05, - "loss": 0.2375, + "loss": 0.2215, "step": 704 }, { "epoch": 2.8556962025316457, - "grad_norm": 0.2967771029872844, + "grad_norm": 0.27576053637433146, "learning_rate": 1.8384140870836157e-05, - "loss": 0.251, + "loss": 0.2311, "step": 705 }, { "epoch": 2.8597468354430378, - "grad_norm": 0.28456386384269605, + "grad_norm": 0.28842207805624825, "learning_rate": 1.8327574328332806e-05, - "loss": 0.2608, + "loss": 0.2279, "step": 706 }, { "epoch": 2.8637974683544303, - "grad_norm": 0.26451771820947834, + "grad_norm": 0.2853800299665492, "learning_rate": 1.8271021255315906e-05, - "loss": 0.2539, + "loss": 0.2269, "step": 707 }, { "epoch": 2.867848101265823, - "grad_norm": 0.29696702502878214, + "grad_norm": 0.29992224522148414, "learning_rate": 1.8214482107256117e-05, - "loss": 0.2473, + "loss": 0.2327, "step": 708 }, { "epoch": 2.8718987341772153, - "grad_norm": 0.28698760203104484, + "grad_norm": 0.28498835940422096, "learning_rate": 1.8157957339511968e-05, - "loss": 0.2472, + "loss": 0.2166, "step": 709 }, { "epoch": 2.875949367088608, - "grad_norm": 0.27153912038474975, + "grad_norm": 0.28990475967898827, "learning_rate": 1.8101447407326182e-05, - "loss": 0.2517, + "loss": 0.2175, "step": 710 }, { "epoch": 2.88, - "grad_norm": 0.30062307576217456, + "grad_norm": 0.296267780247893, "learning_rate": 1.8044952765821966e-05, - "loss": 0.2522, + "loss": 0.2306, "step": 711 }, { "epoch": 2.8840506329113924, - "grad_norm": 0.2712192827868264, + "grad_norm": 0.27978590112411217, "learning_rate": 1.7988473869999407e-05, - "loss": 0.2495, + "loss": 0.2345, "step": 712 }, { "epoch": 2.888101265822785, - "grad_norm": 0.2675348599933565, + "grad_norm": 0.29646088061093234, "learning_rate": 1.7932011174731764e-05, - "loss": 0.2508, + "loss": 0.2295, "step": 713 }, { "epoch": 2.892151898734177, - "grad_norm": 0.28877782306858446, + "grad_norm": 0.28546996263123725, "learning_rate": 1.7875565134761817e-05, - "loss": 0.2534, + "loss": 0.2212, "step": 714 }, { "epoch": 2.8962025316455695, - "grad_norm": 0.25667666034576436, + "grad_norm": 0.2931315302261362, "learning_rate": 1.7819136204698226e-05, - "loss": 0.241, + "loss": 0.2266, "step": 715 }, { "epoch": 2.900253164556962, - "grad_norm": 0.2779140246385452, + "grad_norm": 0.30361105735540095, "learning_rate": 1.776272483901182e-05, - "loss": 0.2741, + "loss": 0.2431, "step": 716 }, { "epoch": 2.9043037974683545, - "grad_norm": 0.2806104963789347, + "grad_norm": 0.287661548824662, "learning_rate": 1.7706331492031995e-05, - "loss": 0.2658, + "loss": 0.2248, "step": 717 }, { "epoch": 2.908354430379747, - "grad_norm": 0.28018336722189985, + "grad_norm": 0.2835093836474247, "learning_rate": 1.764995661794301e-05, - "loss": 0.2465, + "loss": 0.2291, "step": 718 }, { "epoch": 2.912405063291139, - "grad_norm": 0.2880810661676387, + "grad_norm": 0.29842256431288633, "learning_rate": 1.759360067078035e-05, - "loss": 0.2486, + "loss": 0.2149, "step": 719 }, { "epoch": 2.9164556962025316, - "grad_norm": 0.28419945305682315, + "grad_norm": 0.3030425200800445, "learning_rate": 1.7537264104427064e-05, - "loss": 0.263, + "loss": 0.229, "step": 720 }, { "epoch": 2.920506329113924, - "grad_norm": 0.28221134528606107, + "grad_norm": 0.2854031957945021, "learning_rate": 1.748094737261012e-05, - "loss": 0.2504, + "loss": 0.2364, "step": 721 }, { "epoch": 2.9245569620253162, - "grad_norm": 0.2806088050219559, + "grad_norm": 0.2894403790882635, "learning_rate": 1.7424650928896726e-05, - "loss": 0.2585, + "loss": 0.2246, "step": 722 }, { "epoch": 2.9286075949367087, - "grad_norm": 0.2953207407999672, + "grad_norm": 0.3235362848666229, "learning_rate": 1.7368375226690712e-05, - "loss": 0.2744, + "loss": 0.2442, "step": 723 }, { "epoch": 2.9326582278481013, - "grad_norm": 0.28632817180103165, + "grad_norm": 0.2869789964507192, "learning_rate": 1.731212071922883e-05, - "loss": 0.2466, + "loss": 0.219, "step": 724 }, { "epoch": 2.9367088607594938, - "grad_norm": 0.26717300730685334, + "grad_norm": 0.2924959355001303, "learning_rate": 1.7255887859577156e-05, - "loss": 0.2575, + "loss": 0.2337, "step": 725 }, { "epoch": 2.9407594936708863, - "grad_norm": 0.2733710782492506, + "grad_norm": 0.3237147366649734, "learning_rate": 1.7199677100627427e-05, - "loss": 0.2583, + "loss": 0.2404, "step": 726 }, { "epoch": 2.9448101265822784, - "grad_norm": 0.3054811480289982, + "grad_norm": 0.2942224292902647, "learning_rate": 1.7143488895093343e-05, - "loss": 0.284, + "loss": 0.2458, "step": 727 }, { "epoch": 2.948860759493671, - "grad_norm": 0.28148411072798324, + "grad_norm": 0.31212122955483235, "learning_rate": 1.7087323695506994e-05, - "loss": 0.2482, + "loss": 0.2306, "step": 728 }, { "epoch": 2.9529113924050634, - "grad_norm": 0.2629974167578059, + "grad_norm": 0.2792245722798537, "learning_rate": 1.7031181954215194e-05, - "loss": 0.2461, + "loss": 0.2237, "step": 729 }, { "epoch": 2.9569620253164555, - "grad_norm": 0.2858357297042613, + "grad_norm": 0.29860037140954665, "learning_rate": 1.6975064123375788e-05, - "loss": 0.2687, + "loss": 0.2352, "step": 730 }, { "epoch": 2.961012658227848, - "grad_norm": 0.2536940245362183, + "grad_norm": 0.2782302264236621, "learning_rate": 1.6918970654954084e-05, - "loss": 0.2333, + "loss": 0.2201, "step": 731 }, { "epoch": 2.9650632911392405, - "grad_norm": 0.2784753633368692, + "grad_norm": 0.28914524601595876, "learning_rate": 1.686290200071915e-05, - "loss": 0.2758, + "loss": 0.2422, "step": 732 }, { "epoch": 2.969113924050633, - "grad_norm": 0.27509228737076613, + "grad_norm": 0.28470459164380324, "learning_rate": 1.6806858612240234e-05, - "loss": 0.2422, + "loss": 0.222, "step": 733 }, { "epoch": 2.9731645569620255, - "grad_norm": 0.25536164647463827, + "grad_norm": 0.2679451075393547, "learning_rate": 1.6750840940883078e-05, - "loss": 0.2412, + "loss": 0.2119, "step": 734 }, { "epoch": 2.9772151898734176, - "grad_norm": 0.28173483652103254, + "grad_norm": 0.28370523303601225, "learning_rate": 1.6694849437806305e-05, - "loss": 0.2625, + "loss": 0.227, "step": 735 }, { "epoch": 2.98126582278481, - "grad_norm": 0.26241616021772557, + "grad_norm": 0.2690847647682115, "learning_rate": 1.663888455395778e-05, - "loss": 0.2382, + "loss": 0.2242, "step": 736 }, { "epoch": 2.9853164556962026, - "grad_norm": 0.29409910752688534, + "grad_norm": 0.2840641582274788, "learning_rate": 1.6582946740070995e-05, - "loss": 0.2897, + "loss": 0.2372, "step": 737 }, { "epoch": 2.9893670886075947, - "grad_norm": 0.23638321995677702, + "grad_norm": 0.28313134940882745, "learning_rate": 1.6527036446661396e-05, - "loss": 0.2184, + "loss": 0.2073, "step": 738 }, { "epoch": 2.993417721518987, - "grad_norm": 0.27171051610387736, + "grad_norm": 0.28042052444691495, "learning_rate": 1.6471154124022818e-05, - "loss": 0.2715, + "loss": 0.2292, "step": 739 }, { "epoch": 2.9974683544303797, - "grad_norm": 0.29716053915058704, + "grad_norm": 0.2780981778825185, "learning_rate": 1.6415300222223788e-05, - "loss": 0.2793, + "loss": 0.2199, "step": 740 }, { "epoch": 3.0015189873417722, - "grad_norm": 0.32527273174935706, + "grad_norm": 0.3060274509454366, "learning_rate": 1.6359475191103958e-05, - "loss": 0.2486, + "loss": 0.2063, "step": 741 }, { "epoch": 3.0055696202531648, - "grad_norm": 0.36135712788754476, + "grad_norm": 0.36200049117108607, "learning_rate": 1.6303679480270466e-05, - "loss": 0.1779, + "loss": 0.1735, "step": 742 }, { "epoch": 3.009620253164557, - "grad_norm": 0.3259458982767883, + "grad_norm": 0.29139065567740846, "learning_rate": 1.624791353909428e-05, - "loss": 0.1864, + "loss": 0.1651, "step": 743 }, { "epoch": 3.0136708860759494, - "grad_norm": 0.5222317215109146, + "grad_norm": 0.552809595589079, "learning_rate": 1.619217781670663e-05, - "loss": 0.177, + "loss": 0.1622, "step": 744 }, { "epoch": 3.017721518987342, - "grad_norm": 0.37109715950565414, + "grad_norm": 0.3077459681145641, "learning_rate": 1.6136472761995373e-05, - "loss": 0.179, + "loss": 0.1536, "step": 745 }, { "epoch": 3.0217721518987344, - "grad_norm": 0.3607186753422075, + "grad_norm": 0.39515053362386904, "learning_rate": 1.608079882360134e-05, - "loss": 0.1808, + "loss": 0.1654, "step": 746 }, { "epoch": 3.0258227848101265, - "grad_norm": 0.34169247132664815, + "grad_norm": 0.3667221489412283, "learning_rate": 1.60251564499148e-05, - "loss": 0.1646, + "loss": 0.145, "step": 747 }, { "epoch": 3.029873417721519, - "grad_norm": 0.33350696682379993, + "grad_norm": 0.30166240937442923, "learning_rate": 1.596954608907176e-05, - "loss": 0.1592, + "loss": 0.1504, "step": 748 }, { "epoch": 3.0339240506329115, - "grad_norm": 0.3373411011498317, + "grad_norm": 0.34557963011605447, "learning_rate": 1.591396818895043e-05, - "loss": 0.1937, + "loss": 0.178, "step": 749 }, { "epoch": 3.037974683544304, - "grad_norm": 0.29922297974326695, + "grad_norm": 0.32585853695194056, "learning_rate": 1.585842319716759e-05, - "loss": 0.1654, + "loss": 0.1623, "step": 750 }, { "epoch": 3.042025316455696, - "grad_norm": 0.39086530947327014, + "grad_norm": 0.2905305194110086, "learning_rate": 1.5802911561074944e-05, - "loss": 0.1908, + "loss": 0.1722, "step": 751 }, { "epoch": 3.0460759493670886, - "grad_norm": 0.35286823567318104, + "grad_norm": 0.30064416636759667, "learning_rate": 1.5747433727755595e-05, - "loss": 0.1883, + "loss": 0.1743, "step": 752 }, { "epoch": 3.050126582278481, - "grad_norm": 0.2984306965363546, + "grad_norm": 0.2866589520138042, "learning_rate": 1.5691990144020376e-05, - "loss": 0.1543, + "loss": 0.1601, "step": 753 }, { "epoch": 3.0541772151898736, - "grad_norm": 0.3551313594309716, + "grad_norm": 0.28816265537993224, "learning_rate": 1.5636581256404297e-05, - "loss": 0.1809, + "loss": 0.1737, "step": 754 }, { "epoch": 3.0582278481012657, - "grad_norm": 0.36280020025490184, + "grad_norm": 0.29241602740225925, "learning_rate": 1.558120751116291e-05, - "loss": 0.1939, + "loss": 0.1774, "step": 755 }, { "epoch": 3.062278481012658, - "grad_norm": 0.3513233271876707, + "grad_norm": 0.32084543061779264, "learning_rate": 1.552586935426876e-05, - "loss": 0.182, + "loss": 0.1664, "step": 756 }, { "epoch": 3.0663291139240507, - "grad_norm": 0.29645755904667553, + "grad_norm": 0.3005010703992188, "learning_rate": 1.547056723140774e-05, - "loss": 0.1504, + "loss": 0.1351, "step": 757 }, { "epoch": 3.0703797468354432, - "grad_norm": 0.3326220620180981, + "grad_norm": 0.29300814112506124, "learning_rate": 1.5415301587975565e-05, - "loss": 0.1841, + "loss": 0.1658, "step": 758 }, { "epoch": 3.0744303797468353, - "grad_norm": 0.30749560337302506, + "grad_norm": 0.27870219956972886, "learning_rate": 1.536007286907411e-05, - "loss": 0.1882, + "loss": 0.1711, "step": 759 }, { "epoch": 3.078481012658228, - "grad_norm": 0.28242429411761555, + "grad_norm": 0.2720793934517818, "learning_rate": 1.5304881519507896e-05, - "loss": 0.1598, + "loss": 0.1596, "step": 760 }, { "epoch": 3.0825316455696203, - "grad_norm": 0.2895141093963346, + "grad_norm": 0.2899767816627878, "learning_rate": 1.5249727983780453e-05, - "loss": 0.1784, + "loss": 0.1657, "step": 761 }, { "epoch": 3.086582278481013, - "grad_norm": 0.3089257371323482, + "grad_norm": 0.2637903993928107, "learning_rate": 1.5194612706090786e-05, - "loss": 0.1934, + "loss": 0.1833, "step": 762 }, { "epoch": 3.090632911392405, - "grad_norm": 0.2848249263962955, + "grad_norm": 0.27524005418064934, "learning_rate": 1.5139536130329771e-05, - "loss": 0.1626, + "loss": 0.1585, "step": 763 }, { "epoch": 3.0946835443037974, - "grad_norm": 0.2971176049652067, + "grad_norm": 0.2749368536956698, "learning_rate": 1.508449870007656e-05, - "loss": 0.1745, + "loss": 0.159, "step": 764 }, { "epoch": 3.09873417721519, - "grad_norm": 0.30916577584387533, + "grad_norm": 0.29371154109453923, "learning_rate": 1.5029500858595056e-05, - "loss": 0.1882, + "loss": 0.1688, "step": 765 }, { "epoch": 3.1027848101265825, - "grad_norm": 0.288802839023431, + "grad_norm": 0.2715792930644393, "learning_rate": 1.4974543048830328e-05, - "loss": 0.1691, + "loss": 0.1601, "step": 766 }, { "epoch": 3.1068354430379745, - "grad_norm": 0.3198413745877133, + "grad_norm": 0.2760931252285285, "learning_rate": 1.4919625713405e-05, - "loss": 0.1911, + "loss": 0.1683, "step": 767 }, { "epoch": 3.110886075949367, - "grad_norm": 0.3133639987745797, + "grad_norm": 0.2862316730292663, "learning_rate": 1.4864749294615756e-05, - "loss": 0.1711, + "loss": 0.1589, "step": 768 }, { "epoch": 3.1149367088607596, - "grad_norm": 0.2769958955282199, + "grad_norm": 0.2813630153689391, "learning_rate": 1.4809914234429716e-05, - "loss": 0.1739, + "loss": 0.167, "step": 769 }, { "epoch": 3.118987341772152, - "grad_norm": 0.3376171501122422, + "grad_norm": 0.2936114861512913, "learning_rate": 1.4755120974480923e-05, - "loss": 0.2065, + "loss": 0.1938, "step": 770 }, { "epoch": 3.123037974683544, - "grad_norm": 0.2864331501003532, + "grad_norm": 0.26514449715469973, "learning_rate": 1.4700369956066771e-05, - "loss": 0.1775, + "loss": 0.1719, "step": 771 }, { "epoch": 3.1270886075949367, - "grad_norm": 0.333573669603318, + "grad_norm": 0.2970807274290257, "learning_rate": 1.4645661620144413e-05, - "loss": 0.1902, + "loss": 0.1677, "step": 772 }, { "epoch": 3.131139240506329, - "grad_norm": 0.25908165693002533, + "grad_norm": 0.28503235828734624, "learning_rate": 1.4590996407327284e-05, - "loss": 0.1445, + "loss": 0.1378, "step": 773 }, { "epoch": 3.1351898734177217, - "grad_norm": 0.3056553492378959, + "grad_norm": 0.2797121676564152, "learning_rate": 1.4536374757881487e-05, - "loss": 0.1792, + "loss": 0.1637, "step": 774 }, { "epoch": 3.1392405063291138, - "grad_norm": 0.2707660208942242, + "grad_norm": 0.2771464596220541, "learning_rate": 1.4481797111722271e-05, - "loss": 0.1594, + "loss": 0.1568, "step": 775 }, { "epoch": 3.1432911392405063, - "grad_norm": 0.2805504528819787, + "grad_norm": 0.2623454910146439, "learning_rate": 1.4427263908410507e-05, - "loss": 0.1826, + "loss": 0.1664, "step": 776 }, { "epoch": 3.147341772151899, - "grad_norm": 0.2754940958891165, + "grad_norm": 0.2688475800669211, "learning_rate": 1.4372775587149108e-05, - "loss": 0.1771, + "loss": 0.1648, "step": 777 }, { "epoch": 3.1513924050632913, - "grad_norm": 0.30203333756602413, + "grad_norm": 0.26394204865784615, "learning_rate": 1.4318332586779522e-05, - "loss": 0.1824, + "loss": 0.1523, "step": 778 }, { "epoch": 3.1554430379746834, - "grad_norm": 0.2664644791145807, + "grad_norm": 0.2843726582000157, "learning_rate": 1.4263935345778202e-05, - "loss": 0.1575, + "loss": 0.1476, "step": 779 }, { "epoch": 3.159493670886076, - "grad_norm": 0.2909327495437149, + "grad_norm": 0.2813034614077088, "learning_rate": 1.420958430225303e-05, - "loss": 0.1903, + "loss": 0.1754, "step": 780 }, { "epoch": 3.1635443037974684, - "grad_norm": 0.29330453853468796, + "grad_norm": 0.28225866238992053, "learning_rate": 1.415527989393985e-05, - "loss": 0.1855, + "loss": 0.1762, "step": 781 }, { "epoch": 3.167594936708861, - "grad_norm": 0.2732038954013241, + "grad_norm": 0.29136345317324325, "learning_rate": 1.410102255819891e-05, - "loss": 0.169, + "loss": 0.1508, "step": 782 }, { "epoch": 3.171645569620253, - "grad_norm": 0.31391152004149225, + "grad_norm": 0.28828969148301964, "learning_rate": 1.404681273201131e-05, - "loss": 0.178, + "loss": 0.1615, "step": 783 }, { "epoch": 3.1756962025316455, - "grad_norm": 0.2753534221359672, + "grad_norm": 0.2824882045613792, "learning_rate": 1.399265085197556e-05, - "loss": 0.1677, + "loss": 0.1549, "step": 784 }, { "epoch": 3.179746835443038, - "grad_norm": 0.27483604523734384, + "grad_norm": 0.28164846694763584, "learning_rate": 1.393853735430398e-05, - "loss": 0.1651, + "loss": 0.158, "step": 785 }, { "epoch": 3.1837974683544306, - "grad_norm": 0.3151573300751667, + "grad_norm": 0.2715068507902323, "learning_rate": 1.3884472674819246e-05, - "loss": 0.1897, + "loss": 0.1784, "step": 786 }, { "epoch": 3.1878481012658226, - "grad_norm": 0.26066202378714476, + "grad_norm": 0.2826962543702162, "learning_rate": 1.3830457248950864e-05, - "loss": 0.1676, + "loss": 0.1574, "step": 787 }, { "epoch": 3.191898734177215, - "grad_norm": 0.30411021363008695, + "grad_norm": 0.2771333748317222, "learning_rate": 1.377649151173163e-05, - "loss": 0.1746, + "loss": 0.1586, "step": 788 }, { "epoch": 3.1959493670886077, - "grad_norm": 0.2861710579676362, + "grad_norm": 0.2923828090801464, "learning_rate": 1.3722575897794181e-05, - "loss": 0.1699, + "loss": 0.1542, "step": 789 }, { "epoch": 3.2, - "grad_norm": 0.2973241489226898, + "grad_norm": 0.2721248174056244, "learning_rate": 1.3668710841367472e-05, - "loss": 0.1676, + "loss": 0.1532, "step": 790 }, { "epoch": 3.2040506329113922, - "grad_norm": 0.2967646232072694, + "grad_norm": 0.28059024731526755, "learning_rate": 1.361489677627324e-05, - "loss": 0.1646, + "loss": 0.1435, "step": 791 }, { "epoch": 3.2081012658227848, - "grad_norm": 0.31492011570982953, + "grad_norm": 0.29150260293484526, "learning_rate": 1.3561134135922585e-05, - "loss": 0.1949, + "loss": 0.1692, "step": 792 }, { "epoch": 3.2121518987341773, - "grad_norm": 0.28225018773976973, + "grad_norm": 0.274049392809611, "learning_rate": 1.350742335331241e-05, - "loss": 0.1723, + "loss": 0.1606, "step": 793 }, { "epoch": 3.21620253164557, - "grad_norm": 0.2801538795014913, + "grad_norm": 0.288976421924057, "learning_rate": 1.345376486102198e-05, - "loss": 0.1746, + "loss": 0.1674, "step": 794 }, { "epoch": 3.220253164556962, - "grad_norm": 0.2917528642624306, + "grad_norm": 0.2667104566333963, "learning_rate": 1.3400159091209414e-05, - "loss": 0.1667, + "loss": 0.1561, "step": 795 }, { "epoch": 3.2243037974683544, - "grad_norm": 0.2800236981541603, + "grad_norm": 0.26712839050702614, "learning_rate": 1.3346606475608216e-05, - "loss": 0.1736, + "loss": 0.1639, "step": 796 }, { "epoch": 3.228354430379747, - "grad_norm": 0.28937472209999104, + "grad_norm": 0.27315109482276695, "learning_rate": 1.3293107445523781e-05, - "loss": 0.1753, + "loss": 0.1688, "step": 797 }, { "epoch": 3.2324050632911394, - "grad_norm": 0.2847070429858721, + "grad_norm": 0.26405267634975127, "learning_rate": 1.3239662431829949e-05, - "loss": 0.1754, + "loss": 0.1566, "step": 798 }, { "epoch": 3.2364556962025315, - "grad_norm": 0.2758751506174337, + "grad_norm": 0.26998958631664255, "learning_rate": 1.3186271864965509e-05, - "loss": 0.1546, + "loss": 0.1483, "step": 799 }, { "epoch": 3.240506329113924, - "grad_norm": 0.3036228987227513, + "grad_norm": 0.26731679471473246, "learning_rate": 1.3132936174930756e-05, - "loss": 0.1898, + "loss": 0.1643, "step": 800 }, { "epoch": 3.2445569620253165, - "grad_norm": 0.2927158768146142, + "grad_norm": 0.2710451129219873, "learning_rate": 1.3079655791283995e-05, - "loss": 0.1768, + "loss": 0.1542, "step": 801 }, { "epoch": 3.248607594936709, - "grad_norm": 0.26153651216248697, + "grad_norm": 0.26708794900151694, "learning_rate": 1.3026431143138108e-05, - "loss": 0.1733, + "loss": 0.169, "step": 802 }, { "epoch": 3.252658227848101, - "grad_norm": 0.29172981923598856, + "grad_norm": 0.2824829343783079, "learning_rate": 1.2973262659157114e-05, - "loss": 0.1821, + "loss": 0.1709, "step": 803 }, { "epoch": 3.2567088607594936, - "grad_norm": 0.280937206458443, + "grad_norm": 0.26994341552354434, "learning_rate": 1.2920150767552651e-05, - "loss": 0.173, + "loss": 0.1576, "step": 804 }, { "epoch": 3.260759493670886, - "grad_norm": 0.29665223902408094, + "grad_norm": 0.28109100373791535, "learning_rate": 1.2867095896080607e-05, - "loss": 0.1902, + "loss": 0.18, "step": 805 }, { "epoch": 3.2648101265822786, - "grad_norm": 0.2772143172474141, + "grad_norm": 0.2702272439526788, "learning_rate": 1.2814098472037612e-05, - "loss": 0.1652, + "loss": 0.1572, "step": 806 }, { "epoch": 3.2688607594936707, - "grad_norm": 0.2839590168259018, + "grad_norm": 0.26364607289168573, "learning_rate": 1.276115892225764e-05, - "loss": 0.1866, + "loss": 0.1753, "step": 807 }, { "epoch": 3.2729113924050632, - "grad_norm": 0.2812761244159481, + "grad_norm": 0.25186717760782223, "learning_rate": 1.2708277673108555e-05, - "loss": 0.1725, + "loss": 0.1636, "step": 808 }, { "epoch": 3.2769620253164558, - "grad_norm": 0.286164761751961, + "grad_norm": 0.28016137440435945, "learning_rate": 1.2655455150488649e-05, - "loss": 0.1766, + "loss": 0.1607, "step": 809 }, { "epoch": 3.2810126582278483, - "grad_norm": 0.2978466271009404, + "grad_norm": 0.25492211703388984, "learning_rate": 1.2602691779823272e-05, - "loss": 0.1831, + "loss": 0.1637, "step": 810 }, { "epoch": 3.2850632911392403, - "grad_norm": 0.2933572150186941, + "grad_norm": 0.26982822566380793, "learning_rate": 1.2549987986061355e-05, - "loss": 0.1796, + "loss": 0.1705, "step": 811 }, { "epoch": 3.289113924050633, - "grad_norm": 0.26097559334171827, + "grad_norm": 0.2765062465349065, "learning_rate": 1.2497344193672005e-05, - "loss": 0.1915, + "loss": 0.1864, "step": 812 }, { "epoch": 3.2931645569620254, - "grad_norm": 0.2846181138568646, + "grad_norm": 0.2832962810055447, "learning_rate": 1.2444760826641092e-05, - "loss": 0.1648, + "loss": 0.1603, "step": 813 }, { "epoch": 3.297215189873418, - "grad_norm": 0.2800538326618217, + "grad_norm": 0.27482883677138903, "learning_rate": 1.2392238308467817e-05, - "loss": 0.1811, + "loss": 0.1673, "step": 814 }, { "epoch": 3.30126582278481, - "grad_norm": 0.2734354214229583, + "grad_norm": 0.2754201565091245, "learning_rate": 1.2339777062161326e-05, - "loss": 0.1798, + "loss": 0.1734, "step": 815 }, { "epoch": 3.3053164556962025, - "grad_norm": 0.2747173573990617, + "grad_norm": 0.2727530905418694, "learning_rate": 1.2287377510237293e-05, - "loss": 0.1766, + "loss": 0.1676, "step": 816 }, { "epoch": 3.309367088607595, - "grad_norm": 0.272612498231348, + "grad_norm": 0.2632611201992197, "learning_rate": 1.2235040074714488e-05, - "loss": 0.1723, + "loss": 0.1608, "step": 817 }, { "epoch": 3.3134177215189875, - "grad_norm": 0.36568375488129046, + "grad_norm": 0.2746162876177735, "learning_rate": 1.2182765177111434e-05, - "loss": 0.1865, + "loss": 0.1715, "step": 818 }, { "epoch": 3.3174683544303796, - "grad_norm": 0.27888840797422965, + "grad_norm": 0.2664981827021515, "learning_rate": 1.213055323844297e-05, - "loss": 0.1796, + "loss": 0.1604, "step": 819 }, { "epoch": 3.321518987341772, - "grad_norm": 0.2769702426156551, + "grad_norm": 0.26182640199339563, "learning_rate": 1.2078404679216864e-05, - "loss": 0.1654, + "loss": 0.1509, "step": 820 }, { "epoch": 3.3255696202531646, - "grad_norm": 0.25436562932336737, + "grad_norm": 0.27011924857934344, "learning_rate": 1.2026319919430458e-05, - "loss": 0.1605, + "loss": 0.1559, "step": 821 }, { "epoch": 3.329620253164557, - "grad_norm": 0.27929153133575246, + "grad_norm": 0.2665968876031193, "learning_rate": 1.1974299378567227e-05, - "loss": 0.1786, + "loss": 0.1617, "step": 822 }, { "epoch": 3.333670886075949, - "grad_norm": 0.2772940023908925, + "grad_norm": 0.2531737345534062, "learning_rate": 1.1922343475593462e-05, - "loss": 0.1788, + "loss": 0.1547, "step": 823 }, { "epoch": 3.3377215189873417, - "grad_norm": 0.29159276102710496, + "grad_norm": 0.26813349453528385, "learning_rate": 1.187045262895488e-05, - "loss": 0.1935, + "loss": 0.1755, "step": 824 }, { "epoch": 3.3417721518987342, - "grad_norm": 0.24207182378294564, + "grad_norm": 0.2612534758453089, "learning_rate": 1.1818627256573203e-05, - "loss": 0.1555, + "loss": 0.1591, "step": 825 }, { "epoch": 3.3458227848101267, - "grad_norm": 0.27879559205244875, + "grad_norm": 0.2532972604472537, "learning_rate": 1.1766867775842864e-05, - "loss": 0.164, + "loss": 0.1467, "step": 826 }, { "epoch": 3.349873417721519, - "grad_norm": 0.29090593250728386, + "grad_norm": 0.2617337455859381, "learning_rate": 1.1715174603627615e-05, - "loss": 0.1737, + "loss": 0.1505, "step": 827 }, { "epoch": 3.3539240506329113, - "grad_norm": 0.2849640477549279, + "grad_norm": 0.2734240076400232, "learning_rate": 1.1663548156257147e-05, - "loss": 0.1683, + "loss": 0.151, "step": 828 }, { "epoch": 3.357974683544304, - "grad_norm": 0.283551982499017, + "grad_norm": 0.2605088936753551, "learning_rate": 1.161198884952377e-05, - "loss": 0.1661, + "loss": 0.1549, "step": 829 }, { "epoch": 3.3620253164556964, - "grad_norm": 0.262693224661225, + "grad_norm": 0.25811178491616976, "learning_rate": 1.1560497098679056e-05, - "loss": 0.1686, + "loss": 0.153, "step": 830 }, { "epoch": 3.3660759493670884, - "grad_norm": 0.2898189298820896, + "grad_norm": 0.264060797707371, "learning_rate": 1.1509073318430479e-05, - "loss": 0.1751, + "loss": 0.1572, "step": 831 }, { "epoch": 3.370126582278481, - "grad_norm": 0.2926398817823663, + "grad_norm": 0.2637885910448637, "learning_rate": 1.1457717922938116e-05, - "loss": 0.1815, + "loss": 0.1582, "step": 832 }, { "epoch": 3.3741772151898735, - "grad_norm": 0.26870660611504615, + "grad_norm": 0.26511174310555086, "learning_rate": 1.1406431325811233e-05, - "loss": 0.164, + "loss": 0.1489, "step": 833 }, { "epoch": 3.378227848101266, - "grad_norm": 0.2798023611926712, + "grad_norm": 0.2569367447847199, "learning_rate": 1.135521394010506e-05, - "loss": 0.1698, + "loss": 0.1553, "step": 834 }, { "epoch": 3.382278481012658, - "grad_norm": 0.28207932083894044, + "grad_norm": 0.25883756922850437, "learning_rate": 1.1304066178317367e-05, - "loss": 0.1725, + "loss": 0.1566, "step": 835 }, { "epoch": 3.3863291139240506, - "grad_norm": 0.2847946843287544, + "grad_norm": 0.26305626535383103, "learning_rate": 1.1252988452385199e-05, - "loss": 0.1861, + "loss": 0.1741, "step": 836 }, { "epoch": 3.390379746835443, - "grad_norm": 0.28339184113741833, + "grad_norm": 0.2476710870101896, "learning_rate": 1.1201981173681536e-05, - "loss": 0.194, + "loss": 0.1863, "step": 837 }, { "epoch": 3.3944303797468356, - "grad_norm": 0.2757792119513553, + "grad_norm": 0.2795290027009559, "learning_rate": 1.1151044753011991e-05, - "loss": 0.1569, + "loss": 0.1443, "step": 838 }, { "epoch": 3.3984810126582277, - "grad_norm": 0.29185355064815566, + "grad_norm": 0.2697790172731661, "learning_rate": 1.1100179600611491e-05, - "loss": 0.1784, + "loss": 0.1656, "step": 839 }, { "epoch": 3.40253164556962, - "grad_norm": 0.27502757577849074, + "grad_norm": 0.26452246824122355, "learning_rate": 1.1049386126140985e-05, - "loss": 0.1806, + "loss": 0.1714, "step": 840 }, { "epoch": 3.4065822784810127, - "grad_norm": 0.2711718642780975, + "grad_norm": 0.2666929959990581, "learning_rate": 1.0998664738684128e-05, - "loss": 0.1876, + "loss": 0.1706, "step": 841 }, { "epoch": 3.410632911392405, - "grad_norm": 0.28100520151306607, + "grad_norm": 0.25790067564700403, "learning_rate": 1.0948015846744e-05, - "loss": 0.1735, + "loss": 0.1547, "step": 842 }, { "epoch": 3.4146835443037973, - "grad_norm": 0.2800380881397214, + "grad_norm": 0.27596472166554403, "learning_rate": 1.0897439858239832e-05, - "loss": 0.1759, + "loss": 0.1525, "step": 843 }, { "epoch": 3.41873417721519, - "grad_norm": 0.27305295632617466, + "grad_norm": 0.2632876978898246, "learning_rate": 1.0846937180503652e-05, - "loss": 0.1694, + "loss": 0.1603, "step": 844 }, { "epoch": 3.4227848101265823, - "grad_norm": 0.27921156899452887, + "grad_norm": 0.2616944596796812, "learning_rate": 1.0796508220277117e-05, - "loss": 0.1946, + "loss": 0.1604, "step": 845 }, { "epoch": 3.426835443037975, - "grad_norm": 0.2700924226968216, + "grad_norm": 0.26895280100145114, "learning_rate": 1.0746153383708107e-05, - "loss": 0.1697, + "loss": 0.15, "step": 846 }, { "epoch": 3.430886075949367, - "grad_norm": 0.29050169570697376, + "grad_norm": 0.2603983195953071, "learning_rate": 1.0695873076347579e-05, - "loss": 0.1883, + "loss": 0.1683, "step": 847 }, { "epoch": 3.4349367088607594, - "grad_norm": 0.2644277159204431, + "grad_norm": 0.25783974892111267, "learning_rate": 1.0645667703146205e-05, - "loss": 0.1694, + "loss": 0.1621, "step": 848 }, { "epoch": 3.438987341772152, - "grad_norm": 0.2656940970671332, + "grad_norm": 0.2664377200624605, "learning_rate": 1.0595537668451161e-05, - "loss": 0.1627, + "loss": 0.1646, "step": 849 }, { "epoch": 3.4430379746835444, - "grad_norm": 0.2756184922835925, + "grad_norm": 0.2507921000791203, "learning_rate": 1.0545483376002854e-05, - "loss": 0.1886, + "loss": 0.1761, "step": 850 }, { "epoch": 3.4470886075949365, - "grad_norm": 0.2622183496614422, + "grad_norm": 0.24727966359970321, "learning_rate": 1.0495505228931676e-05, - "loss": 0.1665, + "loss": 0.1575, "step": 851 }, { "epoch": 3.451139240506329, - "grad_norm": 0.2836665251246406, + "grad_norm": 0.2503863859844233, "learning_rate": 1.044560362975474e-05, - "loss": 0.1841, + "loss": 0.1638, "step": 852 }, { "epoch": 3.4551898734177215, - "grad_norm": 0.24546214502089908, + "grad_norm": 0.25425801714150653, "learning_rate": 1.0395778980372695e-05, - "loss": 0.1617, + "loss": 0.1704, "step": 853 }, { "epoch": 3.459240506329114, - "grad_norm": 0.273733010882519, + "grad_norm": 0.25099140954760923, "learning_rate": 1.0346031682066381e-05, - "loss": 0.1897, + "loss": 0.179, "step": 854 }, { "epoch": 3.463291139240506, - "grad_norm": 0.2621836416901058, + "grad_norm": 0.2562649829244321, "learning_rate": 1.0296362135493724e-05, - "loss": 0.1655, + "loss": 0.1534, "step": 855 }, { "epoch": 3.4673417721518986, - "grad_norm": 0.271077860743379, + "grad_norm": 0.2504488814660078, "learning_rate": 1.0246770740686422e-05, - "loss": 0.1719, + "loss": 0.1508, "step": 856 }, { "epoch": 3.471392405063291, - "grad_norm": 0.27552682173680665, + "grad_norm": 0.24980955902959262, "learning_rate": 1.0197257897046743e-05, - "loss": 0.1797, + "loss": 0.1645, "step": 857 }, { "epoch": 3.4754430379746837, - "grad_norm": 0.2461692862934291, + "grad_norm": 0.25644277016623085, "learning_rate": 1.014782400334433e-05, - "loss": 0.161, + "loss": 0.1602, "step": 858 }, { "epoch": 3.479493670886076, - "grad_norm": 0.272929417208825, + "grad_norm": 0.2736284721150555, "learning_rate": 1.009846945771296e-05, - "loss": 0.1686, + "loss": 0.1468, "step": 859 }, { "epoch": 3.4835443037974683, - "grad_norm": 0.2797164380612947, + "grad_norm": 0.26431620043307835, "learning_rate": 1.0049194657647363e-05, - "loss": 0.1841, + "loss": 0.161, "step": 860 }, { "epoch": 3.487594936708861, - "grad_norm": 0.2604860161258587, + "grad_norm": 0.25157764784583264, "learning_rate": 1.0000000000000006e-05, - "loss": 0.1835, + "loss": 0.1769, "step": 861 }, { "epoch": 3.4916455696202533, - "grad_norm": 0.2580530328531284, + "grad_norm": 0.2513244077822873, "learning_rate": 9.950885880977891e-06, - "loss": 0.1647, + "loss": 0.1495, "step": 862 }, { "epoch": 3.4956962025316454, - "grad_norm": 0.2839850619379385, + "grad_norm": 0.2600404661140625, "learning_rate": 9.901852696139382e-06, - "loss": 0.1701, + "loss": 0.1499, "step": 863 }, { "epoch": 3.499746835443038, - "grad_norm": 0.2622703054491846, + "grad_norm": 0.24766211299666183, "learning_rate": 9.852900840391027e-06, - "loss": 0.1744, + "loss": 0.1601, "step": 864 }, { "epoch": 3.5037974683544304, - "grad_norm": 0.274382061904978, + "grad_norm": 0.25459478267005264, "learning_rate": 9.804030707984313e-06, - "loss": 0.1853, + "loss": 0.1666, "step": 865 }, { "epoch": 3.507848101265823, - "grad_norm": 0.27785955857140693, + "grad_norm": 0.25551337650680533, "learning_rate": 9.755242692512599e-06, - "loss": 0.1675, + "loss": 0.1522, "step": 866 }, { "epoch": 3.5118987341772154, - "grad_norm": 0.2877662827131889, + "grad_norm": 0.2686448292917107, "learning_rate": 9.70653718690782e-06, - "loss": 0.1815, + "loss": 0.1537, "step": 867 }, { "epoch": 3.5159493670886075, - "grad_norm": 0.28467569193225123, + "grad_norm": 0.2637220992910812, "learning_rate": 9.657914583437454e-06, - "loss": 0.1823, + "loss": 0.1583, "step": 868 }, { "epoch": 3.52, - "grad_norm": 0.28179667285347954, + "grad_norm": 0.27194163884586703, "learning_rate": 9.609375273701246e-06, - "loss": 0.1706, + "loss": 0.1616, "step": 869 }, { "epoch": 3.5240506329113925, - "grad_norm": 0.2599978095395558, + "grad_norm": 0.26587082551037833, "learning_rate": 9.560919648628133e-06, - "loss": 0.1757, + "loss": 0.1609, "step": 870 }, { "epoch": 3.5281012658227846, - "grad_norm": 0.2769412564500931, + "grad_norm": 0.2598823215000778, "learning_rate": 9.512548098473047e-06, - "loss": 0.183, + "loss": 0.1714, "step": 871 }, { "epoch": 3.532151898734177, - "grad_norm": 0.2785397435807348, + "grad_norm": 0.26568758375422563, "learning_rate": 9.464261012813825e-06, - "loss": 0.1758, + "loss": 0.1572, "step": 872 }, { "epoch": 3.5362025316455696, - "grad_norm": 0.2861943585008685, + "grad_norm": 0.2580110536237401, "learning_rate": 9.416058780547987e-06, - "loss": 0.1856, + "loss": 0.1604, "step": 873 }, { "epoch": 3.540253164556962, - "grad_norm": 0.2775129342240785, + "grad_norm": 0.26724647140895236, "learning_rate": 9.367941789889714e-06, - "loss": 0.1743, + "loss": 0.1565, "step": 874 }, { "epoch": 3.5443037974683547, - "grad_norm": 0.27172320558716895, + "grad_norm": 0.2554052505751437, "learning_rate": 9.319910428366607e-06, - "loss": 0.1719, + "loss": 0.1583, "step": 875 }, { "epoch": 3.5483544303797467, - "grad_norm": 0.2672926943602523, + "grad_norm": 0.27077633235334864, "learning_rate": 9.271965082816667e-06, - "loss": 0.1778, + "loss": 0.1667, "step": 876 }, { "epoch": 3.5524050632911393, - "grad_norm": 0.2620089042688723, + "grad_norm": 0.24902228668988838, "learning_rate": 9.224106139385111e-06, - "loss": 0.1696, + "loss": 0.1617, "step": 877 }, { "epoch": 3.5564556962025318, - "grad_norm": 0.26388015500105605, + "grad_norm": 0.25829664660334706, "learning_rate": 9.176333983521291e-06, - "loss": 0.1712, + "loss": 0.163, "step": 878 }, { "epoch": 3.560506329113924, - "grad_norm": 0.27577770505837446, + "grad_norm": 0.25982362872171905, "learning_rate": 9.12864899997558e-06, - "loss": 0.1876, + "loss": 0.1645, "step": 879 }, { "epoch": 3.5645569620253164, - "grad_norm": 0.2841579087868122, + "grad_norm": 0.27312349793829543, "learning_rate": 9.08105157279628e-06, - "loss": 0.1671, + "loss": 0.1492, "step": 880 }, { "epoch": 3.568607594936709, - "grad_norm": 0.2664009187039243, + "grad_norm": 0.2629289373979584, "learning_rate": 9.03354208532653e-06, - "loss": 0.1744, + "loss": 0.1611, "step": 881 }, { "epoch": 3.5726582278481014, - "grad_norm": 0.2766858378565487, + "grad_norm": 0.26342271900968806, "learning_rate": 8.986120920201205e-06, - "loss": 0.1794, + "loss": 0.1725, "step": 882 }, { "epoch": 3.576708860759494, - "grad_norm": 0.27626494447612526, + "grad_norm": 0.2573326384384912, "learning_rate": 8.938788459343852e-06, - "loss": 0.1775, + "loss": 0.1503, "step": 883 }, { "epoch": 3.580759493670886, - "grad_norm": 0.26401179859005786, + "grad_norm": 0.254933662034643, "learning_rate": 8.8915450839636e-06, - "loss": 0.1654, + "loss": 0.1447, "step": 884 }, { "epoch": 3.5848101265822785, - "grad_norm": 0.2581243097796981, + "grad_norm": 0.26528432513607414, "learning_rate": 8.844391174552116e-06, - "loss": 0.1712, + "loss": 0.1471, "step": 885 }, { "epoch": 3.588860759493671, - "grad_norm": 0.2798792246656419, + "grad_norm": 0.24412207701039962, "learning_rate": 8.797327110880479e-06, - "loss": 0.1867, + "loss": 0.1616, "step": 886 }, { "epoch": 3.592911392405063, - "grad_norm": 0.2615538609964748, + "grad_norm": 0.25765974263196645, "learning_rate": 8.750353271996206e-06, - "loss": 0.1682, + "loss": 0.157, "step": 887 }, { "epoch": 3.5969620253164556, - "grad_norm": 0.27272477504792425, + "grad_norm": 0.26300833085224196, "learning_rate": 8.703470036220132e-06, - "loss": 0.1881, + "loss": 0.1571, "step": 888 }, { "epoch": 3.601012658227848, - "grad_norm": 0.2824415375894017, + "grad_norm": 0.2543082166393287, "learning_rate": 8.656677781143394e-06, - "loss": 0.1705, + "loss": 0.154, "step": 889 }, { "epoch": 3.6050632911392406, - "grad_norm": 0.2553539802475999, + "grad_norm": 0.2646333150767995, "learning_rate": 8.609976883624377e-06, - "loss": 0.1706, + "loss": 0.1627, "step": 890 }, { "epoch": 3.609113924050633, - "grad_norm": 0.25920444118737596, + "grad_norm": 0.26294403036811204, "learning_rate": 8.563367719785698e-06, - "loss": 0.1599, + "loss": 0.1562, "step": 891 }, { "epoch": 3.613164556962025, - "grad_norm": 0.30114969649669077, + "grad_norm": 0.2537339884786861, "learning_rate": 8.516850665011138e-06, - "loss": 0.2014, + "loss": 0.1703, "step": 892 }, { "epoch": 3.6172151898734177, - "grad_norm": 0.2803108737069555, + "grad_norm": 0.2503776872498334, "learning_rate": 8.47042609394269e-06, - "loss": 0.1886, + "loss": 0.1782, "step": 893 }, { "epoch": 3.6212658227848102, - "grad_norm": 0.25612933829994483, + "grad_norm": 0.24695938707334175, "learning_rate": 8.424094380477432e-06, - "loss": 0.1755, + "loss": 0.1745, "step": 894 }, { "epoch": 3.6253164556962023, - "grad_norm": 0.2466284725874976, + "grad_norm": 0.2585438113825923, "learning_rate": 8.37785589776465e-06, - "loss": 0.1569, + "loss": 0.1553, "step": 895 }, { "epoch": 3.629367088607595, - "grad_norm": 0.25167215718405234, + "grad_norm": 0.25144708706099395, "learning_rate": 8.331711018202694e-06, - "loss": 0.1647, + "loss": 0.1536, "step": 896 }, { "epoch": 3.6334177215189873, - "grad_norm": 0.2840133545538615, + "grad_norm": 0.25498137052430847, "learning_rate": 8.285660113436104e-06, - "loss": 0.1909, + "loss": 0.164, "step": 897 }, { "epoch": 3.63746835443038, - "grad_norm": 0.24814959510505255, + "grad_norm": 0.24662897995461333, "learning_rate": 8.239703554352527e-06, - "loss": 0.1647, + "loss": 0.1558, "step": 898 }, { "epoch": 3.6415189873417724, - "grad_norm": 0.2580613770852761, + "grad_norm": 0.25294462267597984, "learning_rate": 8.193841711079775e-06, - "loss": 0.177, + "loss": 0.1656, "step": 899 }, { "epoch": 3.6455696202531644, - "grad_norm": 0.24859624443428102, + "grad_norm": 0.26255345053131307, "learning_rate": 8.148074952982828e-06, - "loss": 0.1692, + "loss": 0.1566, "step": 900 }, { "epoch": 3.649620253164557, - "grad_norm": 0.2541657902861798, + "grad_norm": 0.25137851755900076, "learning_rate": 8.102403648660859e-06, - "loss": 0.1677, + "loss": 0.1537, "step": 901 }, { "epoch": 3.6536708860759495, - "grad_norm": 0.2758728021992847, + "grad_norm": 0.25173240400442926, "learning_rate": 8.056828165944282e-06, - "loss": 0.1923, + "loss": 0.1683, "step": 902 }, { "epoch": 3.6577215189873415, - "grad_norm": 0.2698634112226578, + "grad_norm": 0.26893119901402424, "learning_rate": 8.011348871891762e-06, - "loss": 0.1687, + "loss": 0.1476, "step": 903 }, { "epoch": 3.661772151898734, - "grad_norm": 0.2511251804602074, + "grad_norm": 0.25370461352083684, "learning_rate": 7.965966132787287e-06, - "loss": 0.1643, + "loss": 0.1517, "step": 904 }, { "epoch": 3.6658227848101266, - "grad_norm": 0.25905569595025596, + "grad_norm": 0.2592657674137614, "learning_rate": 7.920680314137189e-06, - "loss": 0.1673, + "loss": 0.1543, "step": 905 }, { "epoch": 3.669873417721519, - "grad_norm": 0.26946307349356297, + "grad_norm": 0.26904350172396935, "learning_rate": 7.875491780667246e-06, - "loss": 0.1805, + "loss": 0.1579, "step": 906 }, { "epoch": 3.6739240506329116, - "grad_norm": 0.2719557249031326, + "grad_norm": 0.2563642222287851, "learning_rate": 7.830400896319667e-06, - "loss": 0.1832, + "loss": 0.1634, "step": 907 }, { "epoch": 3.6779746835443037, - "grad_norm": 0.27135402532138925, + "grad_norm": 0.252001696487572, "learning_rate": 7.785408024250259e-06, - "loss": 0.1896, + "loss": 0.172, "step": 908 }, { "epoch": 3.682025316455696, - "grad_norm": 0.26225458987782446, + "grad_norm": 0.25501007372280865, "learning_rate": 7.74051352682542e-06, - "loss": 0.1639, + "loss": 0.145, "step": 909 }, { "epoch": 3.6860759493670887, - "grad_norm": 0.24394235555316082, + "grad_norm": 0.25574545342675387, "learning_rate": 7.695717765619257e-06, - "loss": 0.1631, + "loss": 0.1576, "step": 910 }, { "epoch": 3.690126582278481, - "grad_norm": 0.2808079237821931, + "grad_norm": 0.2689826734109508, "learning_rate": 7.651021101410673e-06, - "loss": 0.188, + "loss": 0.1681, "step": 911 }, { "epoch": 3.6941772151898733, - "grad_norm": 0.26354604975498974, + "grad_norm": 0.26280527011809796, "learning_rate": 7.606423894180464e-06, - "loss": 0.1724, + "loss": 0.152, "step": 912 }, { "epoch": 3.698227848101266, - "grad_norm": 0.25640687341404617, + "grad_norm": 0.24575540960862094, "learning_rate": 7.56192650310839e-06, - "loss": 0.1844, + "loss": 0.1738, "step": 913 }, { "epoch": 3.7022784810126583, - "grad_norm": 0.2692035263371163, + "grad_norm": 0.26201779693822813, "learning_rate": 7.517529286570349e-06, - "loss": 0.1809, + "loss": 0.1829, "step": 914 }, { "epoch": 3.706329113924051, - "grad_norm": 0.2636233383786978, + "grad_norm": 0.25715506327145493, "learning_rate": 7.473232602135387e-06, - "loss": 0.17, + "loss": 0.1571, "step": 915 }, { "epoch": 3.710379746835443, - "grad_norm": 0.25041654055614876, + "grad_norm": 0.25651673921895973, "learning_rate": 7.429036806562935e-06, - "loss": 0.1652, + "loss": 0.1542, "step": 916 }, { "epoch": 3.7144303797468354, - "grad_norm": 0.2692343654214029, + "grad_norm": 0.26890002819348385, "learning_rate": 7.3849422557998455e-06, - "loss": 0.1605, + "loss": 0.151, "step": 917 }, { "epoch": 3.718481012658228, - "grad_norm": 0.27184847587461597, + "grad_norm": 0.23887823313256512, "learning_rate": 7.340949304977567e-06, - "loss": 0.1848, + "loss": 0.1638, "step": 918 }, { "epoch": 3.72253164556962, - "grad_norm": 0.24623729259086247, + "grad_norm": 0.2609412277247185, "learning_rate": 7.297058308409282e-06, - "loss": 0.169, + "loss": 0.1687, "step": 919 }, { "epoch": 3.7265822784810125, - "grad_norm": 0.2483980077338682, + "grad_norm": 0.24915967002903086, "learning_rate": 7.25326961958704e-06, - "loss": 0.1608, + "loss": 0.1519, "step": 920 }, { "epoch": 3.730632911392405, - "grad_norm": 0.3053357858902545, + "grad_norm": 0.24613298106351236, "learning_rate": 7.209583591178921e-06, - "loss": 0.1948, + "loss": 0.1785, "step": 921 }, { "epoch": 3.7346835443037976, - "grad_norm": 0.25878033145825885, + "grad_norm": 0.2549832671818719, "learning_rate": 7.1660005750261925e-06, - "loss": 0.1616, + "loss": 0.1534, "step": 922 }, { "epoch": 3.73873417721519, - "grad_norm": 0.26132364756563636, + "grad_norm": 0.2588558511375635, "learning_rate": 7.1225209221404765e-06, - "loss": 0.1774, + "loss": 0.1678, "step": 923 }, { "epoch": 3.742784810126582, - "grad_norm": 0.2700490061140025, + "grad_norm": 0.2595025463453933, "learning_rate": 7.079144982700909e-06, - "loss": 0.1736, + "loss": 0.164, "step": 924 }, { "epoch": 3.7468354430379747, - "grad_norm": 0.277883521581307, + "grad_norm": 0.25301945521507274, "learning_rate": 7.0358731060513695e-06, - "loss": 0.1729, + "loss": 0.1468, "step": 925 }, { "epoch": 3.750886075949367, - "grad_norm": 0.2664777256375004, + "grad_norm": 0.2682988450021254, "learning_rate": 6.99270564069757e-06, - "loss": 0.176, + "loss": 0.165, "step": 926 }, { "epoch": 3.7549367088607593, - "grad_norm": 0.26012996164634505, + "grad_norm": 0.2651277878082875, "learning_rate": 6.949642934304375e-06, - "loss": 0.1747, + "loss": 0.1611, "step": 927 }, { "epoch": 3.7589873417721518, - "grad_norm": 0.25582568666719535, + "grad_norm": 0.27212629687356005, "learning_rate": 6.906685333692871e-06, - "loss": 0.1637, + "loss": 0.1559, "step": 928 }, { "epoch": 3.7630379746835443, - "grad_norm": 0.2914149192336742, + "grad_norm": 0.25918895431616457, "learning_rate": 6.86383318483769e-06, - "loss": 0.1891, + "loss": 0.175, "step": 929 }, { "epoch": 3.767088607594937, - "grad_norm": 0.2672007493398608, + "grad_norm": 0.2513515923164042, "learning_rate": 6.821086832864139e-06, - "loss": 0.1676, + "loss": 0.1643, "step": 930 }, { "epoch": 3.7711392405063293, - "grad_norm": 0.26673006208443095, + "grad_norm": 0.2636534796012303, "learning_rate": 6.77844662204546e-06, - "loss": 0.1701, + "loss": 0.1539, "step": 931 }, { "epoch": 3.7751898734177214, - "grad_norm": 0.27711310252488663, + "grad_norm": 0.2590482063373185, "learning_rate": 6.7359128958000455e-06, - "loss": 0.1854, + "loss": 0.1675, "step": 932 }, { "epoch": 3.779240506329114, - "grad_norm": 0.27130234120748753, + "grad_norm": 0.24214244261885493, "learning_rate": 6.693485996688695e-06, - "loss": 0.1857, + "loss": 0.1708, "step": 933 }, { "epoch": 3.7832911392405064, - "grad_norm": 0.25836813785853663, + "grad_norm": 0.24885286145042437, "learning_rate": 6.651166266411801e-06, - "loss": 0.1553, + "loss": 0.1454, "step": 934 }, { "epoch": 3.7873417721518985, - "grad_norm": 0.2616551181189434, + "grad_norm": 0.25639296535596184, "learning_rate": 6.6089540458066725e-06, - "loss": 0.1689, + "loss": 0.1534, "step": 935 }, { "epoch": 3.791392405063291, - "grad_norm": 0.241425719061084, + "grad_norm": 0.2454823023479455, "learning_rate": 6.566849674844711e-06, - "loss": 0.1511, + "loss": 0.1487, "step": 936 }, { "epoch": 3.7954430379746835, - "grad_norm": 0.2771350886830721, + "grad_norm": 0.24995289547525043, "learning_rate": 6.524853492628747e-06, - "loss": 0.1825, + "loss": 0.1661, "step": 937 }, { "epoch": 3.799493670886076, - "grad_norm": 0.2528942051405548, + "grad_norm": 0.24871013416158672, "learning_rate": 6.4829658373902536e-06, - "loss": 0.1613, + "loss": 0.1483, "step": 938 }, { "epoch": 3.8035443037974686, - "grad_norm": 0.2605462615762724, + "grad_norm": 0.2564676971680992, "learning_rate": 6.441187046486648e-06, - "loss": 0.1704, + "loss": 0.1449, "step": 939 }, { "epoch": 3.8075949367088606, - "grad_norm": 0.2625533221653404, + "grad_norm": 0.25408841558086237, "learning_rate": 6.399517456398567e-06, - "loss": 0.1734, + "loss": 0.151, "step": 940 }, { "epoch": 3.811645569620253, - "grad_norm": 0.2606673645322864, + "grad_norm": 0.24553866719178202, "learning_rate": 6.357957402727164e-06, - "loss": 0.1667, + "loss": 0.1386, "step": 941 }, { "epoch": 3.8156962025316457, - "grad_norm": 0.24870029154572304, + "grad_norm": 0.25238766164802734, "learning_rate": 6.316507220191395e-06, - "loss": 0.1608, + "loss": 0.1491, "step": 942 }, { "epoch": 3.8197468354430377, - "grad_norm": 0.27320146318116934, + "grad_norm": 0.24249730208629064, "learning_rate": 6.275167242625331e-06, - "loss": 0.1924, + "loss": 0.1653, "step": 943 }, { "epoch": 3.8237974683544302, - "grad_norm": 0.25066524156899334, + "grad_norm": 0.25595514239465644, "learning_rate": 6.233937802975471e-06, - "loss": 0.1727, + "loss": 0.1615, "step": 944 }, { "epoch": 3.8278481012658228, - "grad_norm": 0.23551245062856346, + "grad_norm": 0.25168076676429946, "learning_rate": 6.192819233298046e-06, - "loss": 0.1611, + "loss": 0.1636, "step": 945 }, { "epoch": 3.8318987341772153, - "grad_norm": 0.2695843320453921, + "grad_norm": 0.24912997906881862, "learning_rate": 6.151811864756383e-06, - "loss": 0.1963, + "loss": 0.1681, "step": 946 }, { "epoch": 3.835949367088608, - "grad_norm": 0.2509083869300616, + "grad_norm": 0.2586493335469311, "learning_rate": 6.1109160276181655e-06, - "loss": 0.1442, + "loss": 0.1292, "step": 947 }, { "epoch": 3.84, - "grad_norm": 0.26456906100433486, + "grad_norm": 0.2510417615265298, "learning_rate": 6.070132051252868e-06, - "loss": 0.1797, + "loss": 0.1637, "step": 948 }, { "epoch": 3.8440506329113924, - "grad_norm": 0.2507321468969918, + "grad_norm": 0.24791258276792308, "learning_rate": 6.0294602641290034e-06, - "loss": 0.1762, + "loss": 0.1771, "step": 949 }, { "epoch": 3.848101265822785, - "grad_norm": 0.25842791508062973, + "grad_norm": 0.25194700112199614, "learning_rate": 5.988900993811575e-06, - "loss": 0.1907, + "loss": 0.1741, "step": 950 }, { "epoch": 3.852151898734177, - "grad_norm": 0.2476416829399894, + "grad_norm": 0.23820939426186255, "learning_rate": 5.948454566959363e-06, - "loss": 0.1673, + "loss": 0.164, "step": 951 }, { "epoch": 3.8562025316455695, - "grad_norm": 0.24758462707970846, + "grad_norm": 0.2408436276487035, "learning_rate": 5.908121309322328e-06, - "loss": 0.1657, + "loss": 0.1511, "step": 952 }, { "epoch": 3.860253164556962, - "grad_norm": 0.24772726137023449, + "grad_norm": 0.24745522516161628, "learning_rate": 5.867901545738976e-06, - "loss": 0.1629, + "loss": 0.1545, "step": 953 }, { "epoch": 3.8643037974683545, - "grad_norm": 0.2718695219257469, + "grad_norm": 0.2488410445991274, "learning_rate": 5.827795600133774e-06, - "loss": 0.1949, + "loss": 0.1735, "step": 954 }, { "epoch": 3.868354430379747, - "grad_norm": 0.2538875539710115, + "grad_norm": 0.24700430953241254, "learning_rate": 5.787803795514466e-06, - "loss": 0.1674, + "loss": 0.156, "step": 955 }, { "epoch": 3.872405063291139, - "grad_norm": 0.2538420980870467, + "grad_norm": 0.24172489200686792, "learning_rate": 5.747926453969576e-06, - "loss": 0.16, + "loss": 0.156, "step": 956 }, { "epoch": 3.8764556962025316, - "grad_norm": 0.26947186598862055, + "grad_norm": 0.25016605363013683, "learning_rate": 5.708163896665708e-06, - "loss": 0.1721, + "loss": 0.1503, "step": 957 }, { "epoch": 3.880506329113924, - "grad_norm": 0.2734328227130373, + "grad_norm": 0.2505268401010105, "learning_rate": 5.668516443845047e-06, - "loss": 0.1825, + "loss": 0.1514, "step": 958 }, { "epoch": 3.884556962025316, - "grad_norm": 0.2515614453718243, + "grad_norm": 0.23397148882392907, "learning_rate": 5.6289844148227225e-06, - "loss": 0.1681, + "loss": 0.1547, "step": 959 }, { "epoch": 3.8886075949367087, - "grad_norm": 0.24885378529402208, + "grad_norm": 0.24839991533585884, "learning_rate": 5.5895681279842615e-06, - "loss": 0.1691, + "loss": 0.166, "step": 960 }, { "epoch": 3.8926582278481012, - "grad_norm": 0.2638288539044266, + "grad_norm": 0.2515908714602789, "learning_rate": 5.550267900783019e-06, - "loss": 0.174, + "loss": 0.1564, "step": 961 }, { "epoch": 3.8967088607594937, - "grad_norm": 0.2616179432618583, + "grad_norm": 0.24054860826873142, "learning_rate": 5.511084049737623e-06, - "loss": 0.1759, + "loss": 0.1506, "step": 962 }, { "epoch": 3.9007594936708863, - "grad_norm": 0.24762213279636877, + "grad_norm": 0.24243524633870642, "learning_rate": 5.4720168904294215e-06, - "loss": 0.1709, + "loss": 0.1647, "step": 963 }, { "epoch": 3.9048101265822783, - "grad_norm": 0.27391465615722305, + "grad_norm": 0.23721814393073262, "learning_rate": 5.433066737499948e-06, - "loss": 0.1737, + "loss": 0.164, "step": 964 }, { "epoch": 3.908860759493671, - "grad_norm": 0.26621664338361856, + "grad_norm": 0.25360314027534153, "learning_rate": 5.394233904648376e-06, - "loss": 0.1619, + "loss": 0.1466, "step": 965 }, { "epoch": 3.9129113924050634, - "grad_norm": 0.25354767216696183, + "grad_norm": 0.24223205317647217, "learning_rate": 5.355518704628997e-06, - "loss": 0.178, + "loss": 0.1626, "step": 966 }, { "epoch": 3.9169620253164554, - "grad_norm": 0.2850456775913234, + "grad_norm": 0.24803890780467708, "learning_rate": 5.316921449248731e-06, - "loss": 0.1926, + "loss": 0.1772, "step": 967 }, { "epoch": 3.921012658227848, - "grad_norm": 0.2570894550399128, + "grad_norm": 0.24828158524718116, "learning_rate": 5.278442449364538e-06, - "loss": 0.1548, + "loss": 0.1464, "step": 968 }, { "epoch": 3.9250632911392405, - "grad_norm": 0.2717648858005556, + "grad_norm": 0.25132187353390767, "learning_rate": 5.240082014881016e-06, - "loss": 0.1804, + "loss": 0.1663, "step": 969 }, { "epoch": 3.929113924050633, - "grad_norm": 0.2424117812026223, + "grad_norm": 0.23833499831406454, "learning_rate": 5.201840454747822e-06, - "loss": 0.1599, + "loss": 0.1534, "step": 970 }, { "epoch": 3.9331645569620255, - "grad_norm": 0.2698580767180666, + "grad_norm": 0.24977835699120865, "learning_rate": 5.163718076957223e-06, - "loss": 0.1718, + "loss": 0.1483, "step": 971 }, { "epoch": 3.9372151898734176, - "grad_norm": 0.277912504226624, + "grad_norm": 0.24809800617318303, "learning_rate": 5.125715188541609e-06, - "loss": 0.1784, + "loss": 0.1631, "step": 972 }, { "epoch": 3.94126582278481, - "grad_norm": 0.2382762863364678, + "grad_norm": 0.25573283701244554, "learning_rate": 5.087832095571021e-06, - "loss": 0.1501, + "loss": 0.1417, "step": 973 }, { "epoch": 3.9453164556962026, - "grad_norm": 0.273404508738298, + "grad_norm": 0.25998208286045466, "learning_rate": 5.0500691031506766e-06, - "loss": 0.1703, + "loss": 0.1655, "step": 974 }, { "epoch": 3.9493670886075947, - "grad_norm": 0.281964985498285, + "grad_norm": 0.25074200303320515, "learning_rate": 5.01242651541854e-06, - "loss": 0.1893, + "loss": 0.1661, "step": 975 }, { "epoch": 3.953417721518987, - "grad_norm": 0.2528154594761226, + "grad_norm": 0.2595153696780746, "learning_rate": 4.974904635542815e-06, - "loss": 0.1522, + "loss": 0.1462, "step": 976 }, { "epoch": 3.9574683544303797, - "grad_norm": 0.267808544974997, + "grad_norm": 0.2463018573381016, "learning_rate": 4.937503765719582e-06, - "loss": 0.1723, + "loss": 0.1589, "step": 977 }, { "epoch": 3.961518987341772, - "grad_norm": 0.26112015418063333, + "grad_norm": 0.2521897661099928, "learning_rate": 4.900224207170299e-06, - "loss": 0.1864, + "loss": 0.1698, "step": 978 }, { "epoch": 3.9655696202531647, - "grad_norm": 0.2371986410638994, + "grad_norm": 0.27432207410824794, "learning_rate": 4.8630662601394065e-06, - "loss": 0.1496, + "loss": 0.1445, "step": 979 }, { "epoch": 3.969620253164557, - "grad_norm": 0.263158653636524, + "grad_norm": 0.24196881622190164, "learning_rate": 4.8260302238918995e-06, - "loss": 0.175, + "loss": 0.1576, "step": 980 }, { "epoch": 3.9736708860759493, - "grad_norm": 0.24308850121124262, + "grad_norm": 0.24618525514853487, "learning_rate": 4.789116396710924e-06, - "loss": 0.1533, + "loss": 0.1472, "step": 981 }, { "epoch": 3.977721518987342, - "grad_norm": 0.2650979446263137, + "grad_norm": 0.25540689964575236, "learning_rate": 4.752325075895368e-06, - "loss": 0.1714, + "loss": 0.1523, "step": 982 }, { "epoch": 3.981772151898734, - "grad_norm": 0.2627984529755924, + "grad_norm": 0.259029620573096, "learning_rate": 4.715656557757473e-06, - "loss": 0.1829, + "loss": 0.164, "step": 983 }, { "epoch": 3.9858227848101264, - "grad_norm": 0.25371328157933154, + "grad_norm": 0.24416716324394266, "learning_rate": 4.679111137620442e-06, - "loss": 0.1787, + "loss": 0.1624, "step": 984 }, { "epoch": 3.989873417721519, - "grad_norm": 0.24140365101653927, + "grad_norm": 0.24455449312611538, "learning_rate": 4.6426891098160585e-06, - "loss": 0.1692, + "loss": 0.1567, "step": 985 }, { "epoch": 3.9939240506329114, - "grad_norm": 0.2506655531342222, + "grad_norm": 0.2672936255859228, "learning_rate": 4.6063907676823474e-06, - "loss": 0.1703, + "loss": 0.1658, "step": 986 }, { "epoch": 3.997974683544304, - "grad_norm": 0.2775980739507107, + "grad_norm": 0.25292377519422765, "learning_rate": 4.570216403561141e-06, - "loss": 0.1829, + "loss": 0.1537, "step": 987 }, { "epoch": 4.002025316455696, - "grad_norm": 0.32013055362757475, + "grad_norm": 0.2630643821793862, "learning_rate": 4.534166308795815e-06, - "loss": 0.1677, + "loss": 0.1314, "step": 988 }, { "epoch": 4.006075949367089, - "grad_norm": 0.40322867158532866, + "grad_norm": 0.33880314687057583, "learning_rate": 4.498240773728859e-06, - "loss": 0.1218, + "loss": 0.1235, "step": 989 }, { "epoch": 4.010126582278481, - "grad_norm": 0.3213244180588823, + "grad_norm": 0.27208439979546833, "learning_rate": 4.462440087699609e-06, - "loss": 0.1176, + "loss": 0.1166, "step": 990 }, { "epoch": 4.014177215189873, - "grad_norm": 0.26286401377216234, + "grad_norm": 0.20534963200806275, "learning_rate": 4.426764539041861e-06, - "loss": 0.1197, + "loss": 0.1135, "step": 991 }, { "epoch": 4.018227848101266, - "grad_norm": 0.2948935238717514, + "grad_norm": 0.2704665900696553, "learning_rate": 4.391214415081582e-06, - "loss": 0.1153, + "loss": 0.1124, "step": 992 }, { "epoch": 4.022278481012658, - "grad_norm": 0.3964310414429152, + "grad_norm": 0.3754796314299987, "learning_rate": 4.355790002134579e-06, - "loss": 0.1237, + "loss": 0.118, "step": 993 }, { "epoch": 4.02632911392405, - "grad_norm": 0.3957273374260363, + "grad_norm": 0.36722700935941166, "learning_rate": 4.320491585504207e-06, - "loss": 0.1199, + "loss": 0.1195, "step": 994 }, { "epoch": 4.030379746835443, - "grad_norm": 0.40850484834500655, + "grad_norm": 0.2691072819760026, "learning_rate": 4.2853194494790615e-06, - "loss": 0.1301, + "loss": 0.1232, "step": 995 }, { "epoch": 4.034430379746835, - "grad_norm": 0.30208889819549795, + "grad_norm": 0.24776192744305042, "learning_rate": 4.250273877330691e-06, - "loss": 0.1167, + "loss": 0.1093, "step": 996 }, { "epoch": 4.038481012658228, - "grad_norm": 0.2784869644870103, + "grad_norm": 0.3282723338526353, "learning_rate": 4.215355151311313e-06, - "loss": 0.1136, + "loss": 0.1228, "step": 997 }, { "epoch": 4.04253164556962, - "grad_norm": 0.3330359878307397, + "grad_norm": 0.3248037904612766, "learning_rate": 4.180563552651542e-06, - "loss": 0.1231, + "loss": 0.1283, "step": 998 }, { "epoch": 4.046582278481012, - "grad_norm": 0.29411586845349713, + "grad_norm": 0.2676506944728411, "learning_rate": 4.145899361558147e-06, - "loss": 0.1082, + "loss": 0.1128, "step": 999 }, { "epoch": 4.050632911392405, - "grad_norm": 0.2927731446147977, + "grad_norm": 0.21153746241567573, "learning_rate": 4.111362857211738e-06, - "loss": 0.1245, + "loss": 0.1231, "step": 1000 }, { "epoch": 4.054683544303797, - "grad_norm": 0.2798948320082335, + "grad_norm": 0.2216142981544128, "learning_rate": 4.076954317764592e-06, - "loss": 0.1212, + "loss": 0.1181, "step": 1001 }, { "epoch": 4.0587341772151895, - "grad_norm": 0.30426169053053376, + "grad_norm": 0.25186408890903295, "learning_rate": 4.042674020338335e-06, - "loss": 0.1357, + "loss": 0.1272, "step": 1002 }, { "epoch": 4.062784810126582, - "grad_norm": 0.27163766038669124, + "grad_norm": 0.2862467211159429, "learning_rate": 4.0085222410217835e-06, - "loss": 0.115, + "loss": 0.1178, "step": 1003 }, { "epoch": 4.0668354430379745, - "grad_norm": 0.29010289549492724, + "grad_norm": 0.2584655362217925, "learning_rate": 3.974499254868674e-06, - "loss": 0.116, + "loss": 0.1202, "step": 1004 }, { "epoch": 4.0708860759493675, - "grad_norm": 0.27691614798304554, + "grad_norm": 0.21287397999311464, "learning_rate": 3.940605335895451e-06, - "loss": 0.1128, + "loss": 0.1158, "step": 1005 }, { "epoch": 4.0749367088607595, - "grad_norm": 0.2819397539223702, + "grad_norm": 0.2272606799300811, "learning_rate": 3.90684075707908e-06, - "loss": 0.1132, + "loss": 0.1167, "step": 1006 }, { "epoch": 4.078987341772152, - "grad_norm": 0.29742961362429976, + "grad_norm": 0.23460983905126037, "learning_rate": 3.8732057903548505e-06, - "loss": 0.1221, + "loss": 0.1183, "step": 1007 }, { "epoch": 4.083037974683545, - "grad_norm": 0.26755589764762033, + "grad_norm": 0.24355711245616457, "learning_rate": 3.8397007066141375e-06, - "loss": 0.1152, + "loss": 0.1174, "step": 1008 }, { "epoch": 4.087088607594937, - "grad_norm": 0.26504408919529965, + "grad_norm": 0.22953984106875466, "learning_rate": 3.806325775702304e-06, - "loss": 0.1192, + "loss": 0.1235, "step": 1009 }, { "epoch": 4.091139240506329, - "grad_norm": 0.2579109954237344, + "grad_norm": 0.21988789633880235, "learning_rate": 3.773081266416434e-06, - "loss": 0.1233, + "loss": 0.1255, "step": 1010 }, { "epoch": 4.095189873417722, - "grad_norm": 0.26054626437218986, + "grad_norm": 0.22299107697306728, "learning_rate": 3.739967446503245e-06, - "loss": 0.1228, + "loss": 0.126, "step": 1011 }, { "epoch": 4.099240506329114, - "grad_norm": 0.25960879435935574, + "grad_norm": 0.22849862539071805, "learning_rate": 3.706984582656894e-06, - "loss": 0.1126, + "loss": 0.1111, "step": 1012 }, { "epoch": 4.103291139240507, - "grad_norm": 0.25193460932370526, + "grad_norm": 0.23103160714784965, "learning_rate": 3.6741329405168237e-06, - "loss": 0.1121, + "loss": 0.1152, "step": 1013 }, { "epoch": 4.107341772151899, - "grad_norm": 0.24775683253307748, + "grad_norm": 0.225094717670364, "learning_rate": 3.641412784665648e-06, - "loss": 0.1049, + "loss": 0.1038, "step": 1014 }, { "epoch": 4.111392405063291, - "grad_norm": 0.2710808901090634, + "grad_norm": 0.22321826082979954, "learning_rate": 3.608824378627005e-06, - "loss": 0.1282, + "loss": 0.1292, "step": 1015 }, { "epoch": 4.115443037974684, - "grad_norm": 0.26787893379264927, + "grad_norm": 0.2230738656578773, "learning_rate": 3.5763679848634337e-06, - "loss": 0.1152, + "loss": 0.1172, "step": 1016 }, { "epoch": 4.119493670886076, - "grad_norm": 0.2638658516993552, + "grad_norm": 0.21839372704933116, "learning_rate": 3.544043864774269e-06, - "loss": 0.1138, + "loss": 0.1059, "step": 1017 }, { "epoch": 4.123544303797468, - "grad_norm": 0.2588848519610793, + "grad_norm": 0.2286119411270651, "learning_rate": 3.5118522786935282e-06, - "loss": 0.12, + "loss": 0.1237, "step": 1018 }, { "epoch": 4.127594936708861, - "grad_norm": 0.25132531512532824, + "grad_norm": 0.21380841486349964, "learning_rate": 3.479793485887819e-06, - "loss": 0.1176, + "loss": 0.1183, "step": 1019 }, { "epoch": 4.131645569620253, - "grad_norm": 0.26287302135535723, + "grad_norm": 0.2081825307671998, "learning_rate": 3.4478677445542653e-06, - "loss": 0.1305, + "loss": 0.1283, "step": 1020 }, { "epoch": 4.135696202531646, - "grad_norm": 0.22478334503884323, + "grad_norm": 0.20569889756285226, "learning_rate": 3.4160753118183767e-06, - "loss": 0.1065, + "loss": 0.1188, "step": 1021 }, { "epoch": 4.139746835443038, - "grad_norm": 0.25328545089014176, + "grad_norm": 0.2170529063060265, "learning_rate": 3.3844164437320527e-06, - "loss": 0.1182, + "loss": 0.1165, "step": 1022 }, { "epoch": 4.14379746835443, - "grad_norm": 0.2591205539082008, + "grad_norm": 0.2117667452477069, "learning_rate": 3.3528913952714558e-06, - "loss": 0.1115, + "loss": 0.1015, "step": 1023 }, { "epoch": 4.147848101265823, - "grad_norm": 0.23972551309405832, + "grad_norm": 0.21430017179735353, "learning_rate": 3.321500420335e-06, - "loss": 0.1067, + "loss": 0.1059, "step": 1024 }, { "epoch": 4.151898734177215, - "grad_norm": 0.24671950027617306, + "grad_norm": 0.21470745995743878, "learning_rate": 3.290243771741275e-06, - "loss": 0.1178, + "loss": 0.1283, "step": 1025 }, { "epoch": 4.155949367088607, - "grad_norm": 0.2614516259972278, + "grad_norm": 0.2030013839294294, "learning_rate": 3.2591217012270325e-06, - "loss": 0.1247, + "loss": 0.1183, "step": 1026 }, { "epoch": 4.16, - "grad_norm": 0.2520985919856274, + "grad_norm": 0.22519001699224542, "learning_rate": 3.228134459445149e-06, - "loss": 0.1026, + "loss": 0.1001, "step": 1027 }, { "epoch": 4.164050632911392, - "grad_norm": 0.25249169293165064, + "grad_norm": 0.20670809055762013, "learning_rate": 3.1972822959626205e-06, - "loss": 0.1205, + "loss": 0.1151, "step": 1028 }, { "epoch": 4.168101265822785, - "grad_norm": 0.26048946003845386, + "grad_norm": 0.21987863043333009, "learning_rate": 3.166565459258513e-06, - "loss": 0.12, + "loss": 0.1167, "step": 1029 }, { "epoch": 4.172151898734177, - "grad_norm": 0.2603289533907277, + "grad_norm": 0.21189346202964568, "learning_rate": 3.1359841967220193e-06, - "loss": 0.1233, + "loss": 0.1158, "step": 1030 }, { "epoch": 4.176202531645569, - "grad_norm": 0.24298433972164693, + "grad_norm": 0.21345713130052868, "learning_rate": 3.105538754650419e-06, - "loss": 0.111, + "loss": 0.1082, "step": 1031 }, { "epoch": 4.180253164556962, - "grad_norm": 0.23688323076825393, + "grad_norm": 0.22010486394357115, "learning_rate": 3.07522937824712e-06, - "loss": 0.103, + "loss": 0.1058, "step": 1032 }, { "epoch": 4.184303797468354, - "grad_norm": 0.25711148774428777, + "grad_norm": 0.21336919310878416, "learning_rate": 3.0450563116196697e-06, - "loss": 0.1354, + "loss": 0.1348, "step": 1033 }, { "epoch": 4.188354430379746, - "grad_norm": 0.22880704919710712, + "grad_norm": 0.21401173426009146, "learning_rate": 3.0150197977778008e-06, - "loss": 0.1011, + "loss": 0.107, "step": 1034 }, { "epoch": 4.192405063291139, - "grad_norm": 0.2620620525114609, + "grad_norm": 0.2105273688417726, "learning_rate": 2.985120078631465e-06, - "loss": 0.1269, + "loss": 0.1274, "step": 1035 }, { "epoch": 4.1964556962025314, - "grad_norm": 0.2359855562228777, + "grad_norm": 0.21352221483537187, "learning_rate": 2.9553573949888893e-06, - "loss": 0.1109, + "loss": 0.1172, "step": 1036 }, { "epoch": 4.200506329113924, - "grad_norm": 0.2362843871300956, + "grad_norm": 0.2130753836403766, "learning_rate": 2.9257319865546384e-06, - "loss": 0.1163, + "loss": 0.121, "step": 1037 }, { "epoch": 4.2045569620253165, - "grad_norm": 0.24257006291260302, + "grad_norm": 0.1985612497120337, "learning_rate": 2.896244091927678e-06, - "loss": 0.1201, + "loss": 0.1271, "step": 1038 }, { "epoch": 4.2086075949367086, - "grad_norm": 0.23824969722732522, + "grad_norm": 0.2089819131698675, "learning_rate": 2.8668939485994584e-06, - "loss": 0.1079, + "loss": 0.1125, "step": 1039 }, { "epoch": 4.2126582278481015, - "grad_norm": 0.25283743310876056, + "grad_norm": 0.20674932442009689, "learning_rate": 2.837681792951994e-06, - "loss": 0.1235, + "loss": 0.1339, "step": 1040 }, { "epoch": 4.216708860759494, - "grad_norm": 0.23690980723695132, + "grad_norm": 0.20694356951579712, "learning_rate": 2.808607860255981e-06, - "loss": 0.1126, + "loss": 0.1165, "step": 1041 }, { "epoch": 4.220759493670886, - "grad_norm": 0.2331886388863026, + "grad_norm": 0.20817915565327672, "learning_rate": 2.7796723846688634e-06, - "loss": 0.1109, + "loss": 0.1122, "step": 1042 }, { "epoch": 4.224810126582279, - "grad_norm": 0.24805227673368532, + "grad_norm": 0.21245203026818976, "learning_rate": 2.7508755992329937e-06, - "loss": 0.1218, + "loss": 0.1237, "step": 1043 }, { "epoch": 4.228860759493671, - "grad_norm": 0.2597054742588775, + "grad_norm": 0.20551464168371134, "learning_rate": 2.722217735873718e-06, - "loss": 0.127, + "loss": 0.132, "step": 1044 }, { "epoch": 4.232911392405064, - "grad_norm": 0.23687982775852348, + "grad_norm": 0.21495236065423326, "learning_rate": 2.6936990253975315e-06, - "loss": 0.1112, + "loss": 0.1183, "step": 1045 }, { "epoch": 4.236962025316456, - "grad_norm": 0.23251743797966518, + "grad_norm": 0.20062676615951858, "learning_rate": 2.665319697490205e-06, - "loss": 0.1038, + "loss": 0.1064, "step": 1046 }, { "epoch": 4.241012658227848, - "grad_norm": 0.25539803459721894, + "grad_norm": 0.2123075948988442, "learning_rate": 2.637079980714945e-06, - "loss": 0.1255, + "loss": 0.1329, "step": 1047 }, { "epoch": 4.245063291139241, - "grad_norm": 0.23494620381513848, + "grad_norm": 0.20341629001007777, "learning_rate": 2.6089801025105453e-06, - "loss": 0.1061, + "loss": 0.1119, "step": 1048 }, { "epoch": 4.249113924050633, - "grad_norm": 0.24495593172240096, + "grad_norm": 0.20713083837574972, "learning_rate": 2.581020289189571e-06, - "loss": 0.1191, + "loss": 0.1121, "step": 1049 }, { "epoch": 4.253164556962025, - "grad_norm": 0.24430021074722647, + "grad_norm": 0.2157678977321124, "learning_rate": 2.553200765936501e-06, - "loss": 0.1155, + "loss": 0.1237, "step": 1050 }, { "epoch": 4.257215189873418, - "grad_norm": 0.23191935413175543, + "grad_norm": 0.2217107081643587, "learning_rate": 2.525521756805962e-06, - "loss": 0.1151, + "loss": 0.133, "step": 1051 }, { "epoch": 4.26126582278481, - "grad_norm": 0.2439948984219711, + "grad_norm": 0.20744164058612227, "learning_rate": 2.497983484720885e-06, - "loss": 0.1231, + "loss": 0.1297, "step": 1052 }, { "epoch": 4.265316455696203, - "grad_norm": 0.2396130011158247, + "grad_norm": 0.21715298158620164, "learning_rate": 2.470586171470728e-06, - "loss": 0.1115, + "loss": 0.1105, "step": 1053 }, { "epoch": 4.269367088607595, - "grad_norm": 0.2447869304774896, + "grad_norm": 0.21088198002146902, "learning_rate": 2.4433300377096836e-06, - "loss": 0.116, + "loss": 0.1217, "step": 1054 }, { "epoch": 4.273417721518987, - "grad_norm": 0.26334400059721647, + "grad_norm": 0.213936525667129, "learning_rate": 2.4162153029549073e-06, - "loss": 0.1343, + "loss": 0.1316, "step": 1055 }, { "epoch": 4.27746835443038, - "grad_norm": 0.2303130314351784, + "grad_norm": 0.21696396428842615, "learning_rate": 2.3892421855847458e-06, - "loss": 0.1052, + "loss": 0.1123, "step": 1056 }, { "epoch": 4.281518987341772, - "grad_norm": 0.24144900472982161, + "grad_norm": 0.20443173310555152, "learning_rate": 2.362410902836978e-06, - "loss": 0.1099, + "loss": 0.118, "step": 1057 }, { "epoch": 4.285569620253165, - "grad_norm": 0.23821932259608225, + "grad_norm": 0.21461082139433513, "learning_rate": 2.3357216708070653e-06, - "loss": 0.106, + "loss": 0.1078, "step": 1058 }, { "epoch": 4.289620253164557, - "grad_norm": 0.2422304138494312, + "grad_norm": 0.2087392070133489, "learning_rate": 2.309174704446411e-06, - "loss": 0.1114, + "loss": 0.1219, "step": 1059 }, { "epoch": 4.293670886075949, - "grad_norm": 0.23886253773817032, + "grad_norm": 0.20754734076427592, "learning_rate": 2.2827702175606437e-06, - "loss": 0.117, + "loss": 0.1206, "step": 1060 }, { "epoch": 4.297721518987342, - "grad_norm": 0.25906441811632364, + "grad_norm": 0.20075152583335018, "learning_rate": 2.256508422807855e-06, - "loss": 0.1305, + "loss": 0.1223, "step": 1061 }, { "epoch": 4.301772151898734, - "grad_norm": 0.23279658823927288, + "grad_norm": 0.21696826977527567, "learning_rate": 2.230389531696946e-06, - "loss": 0.1017, + "loss": 0.1047, "step": 1062 }, { "epoch": 4.305822784810126, - "grad_norm": 0.23170026027479287, + "grad_norm": 0.21041400641960695, "learning_rate": 2.204413754585857e-06, - "loss": 0.0971, + "loss": 0.0975, "step": 1063 }, { "epoch": 4.309873417721519, - "grad_norm": 0.2651159490649785, + "grad_norm": 0.21178496521510046, "learning_rate": 2.1785813006799406e-06, - "loss": 0.1209, + "loss": 0.1167, "step": 1064 }, { "epoch": 4.313924050632911, - "grad_norm": 0.24860343617864072, + "grad_norm": 0.2330828758422294, "learning_rate": 2.1528923780302224e-06, - "loss": 0.1104, + "loss": 0.1119, "step": 1065 }, { "epoch": 4.317974683544303, - "grad_norm": 0.24190066611697628, + "grad_norm": 0.22699012410415995, "learning_rate": 2.127347193531757e-06, "loss": 0.116, "step": 1066 }, { "epoch": 4.322025316455696, - "grad_norm": 0.2552339755960148, + "grad_norm": 0.21138016027001472, "learning_rate": 2.101945952921942e-06, - "loss": 0.1232, + "loss": 0.1158, "step": 1067 }, { "epoch": 4.326075949367088, - "grad_norm": 0.26279140088710506, + "grad_norm": 0.20151910317051508, "learning_rate": 2.0766888607788906e-06, - "loss": 0.1301, + "loss": 0.1168, "step": 1068 }, { "epoch": 4.330126582278481, - "grad_norm": 0.2367959775589517, + "grad_norm": 0.2125072515748105, "learning_rate": 2.0515761205197337e-06, - "loss": 0.1018, + "loss": 0.0979, "step": 1069 }, { "epoch": 4.334177215189873, - "grad_norm": 0.24680630396699663, + "grad_norm": 0.2107698407842126, "learning_rate": 2.0266079343990453e-06, - "loss": 0.1236, + "loss": 0.1244, "step": 1070 }, { "epoch": 4.3382278481012655, - "grad_norm": 0.24048685345806675, + "grad_norm": 0.21027002659203328, "learning_rate": 2.0017845035071494e-06, - "loss": 0.109, + "loss": 0.1116, "step": 1071 }, { "epoch": 4.3422784810126585, - "grad_norm": 0.2575900235480316, + "grad_norm": 0.21996085057991177, "learning_rate": 1.9771060277685537e-06, - "loss": 0.1183, + "loss": 0.1186, "step": 1072 }, { "epoch": 4.3463291139240505, - "grad_norm": 0.2279774032538057, + "grad_norm": 0.2103809729797862, "learning_rate": 1.95257270594031e-06, - "loss": 0.106, + "loss": 0.1019, "step": 1073 }, { "epoch": 4.3503797468354435, - "grad_norm": 0.23038403899675514, + "grad_norm": 0.20627715439090055, "learning_rate": 1.9281847356104188e-06, - "loss": 0.1136, + "loss": 0.1183, "step": 1074 }, { "epoch": 4.3544303797468356, - "grad_norm": 0.259233532712715, + "grad_norm": 0.20589562887736468, "learning_rate": 1.9039423131962365e-06, - "loss": 0.114, + "loss": 0.1227, "step": 1075 }, { "epoch": 4.358481012658228, - "grad_norm": 0.21786481964765095, + "grad_norm": 0.21336148947063796, "learning_rate": 1.8798456339429027e-06, - "loss": 0.0867, + "loss": 0.0859, "step": 1076 }, { "epoch": 4.362531645569621, - "grad_norm": 0.2573601037302012, + "grad_norm": 0.20831422816400133, "learning_rate": 1.8558948919217612e-06, - "loss": 0.1235, + "loss": 0.1162, "step": 1077 }, { "epoch": 4.366582278481013, - "grad_norm": 0.25706337570308346, + "grad_norm": 0.21145219119477301, "learning_rate": 1.8320902800287954e-06, - "loss": 0.14, + "loss": 0.1267, "step": 1078 }, { "epoch": 4.370632911392405, - "grad_norm": 0.2371137723073174, + "grad_norm": 0.2136018809141807, "learning_rate": 1.8084319899830726e-06, - "loss": 0.1104, + "loss": 0.1143, "step": 1079 }, { "epoch": 4.374683544303798, - "grad_norm": 0.2486842998702614, + "grad_norm": 0.21949543447597544, "learning_rate": 1.7849202123252097e-06, - "loss": 0.1275, + "loss": 0.1257, "step": 1080 }, { "epoch": 4.37873417721519, - "grad_norm": 0.25409674042079944, + "grad_norm": 0.2140850337924572, "learning_rate": 1.7615551364158401e-06, - "loss": 0.1276, + "loss": 0.1228, "step": 1081 }, { "epoch": 4.382784810126582, - "grad_norm": 0.21462208438040503, + "grad_norm": 0.21026595117652644, "learning_rate": 1.738336950434061e-06, - "loss": 0.1026, + "loss": 0.11, "step": 1082 }, { "epoch": 4.386835443037975, - "grad_norm": 0.26076380941498906, + "grad_norm": 0.20738756716663165, "learning_rate": 1.715265841375957e-06, - "loss": 0.1307, + "loss": 0.1358, "step": 1083 }, { "epoch": 4.390886075949367, - "grad_norm": 0.2423955998242404, + "grad_norm": 0.21865162793635812, "learning_rate": 1.6923419950530684e-06, - "loss": 0.1088, + "loss": 0.1051, "step": 1084 }, { "epoch": 4.39493670886076, - "grad_norm": 0.2304954846906631, + "grad_norm": 0.2054525486858317, "learning_rate": 1.6695655960909008e-06, - "loss": 0.1105, + "loss": 0.1184, "step": 1085 }, { "epoch": 4.398987341772152, - "grad_norm": 0.23716089185528708, + "grad_norm": 0.21244269696954227, "learning_rate": 1.646936827927441e-06, - "loss": 0.1179, + "loss": 0.1316, "step": 1086 }, { "epoch": 4.403037974683544, - "grad_norm": 0.22946946961836873, + "grad_norm": 0.20580273370728408, "learning_rate": 1.6244558728116766e-06, - "loss": 0.113, + "loss": 0.1222, "step": 1087 }, { "epoch": 4.407088607594937, - "grad_norm": 0.2343155039200944, + "grad_norm": 0.21336703945479918, "learning_rate": 1.6021229118021265e-06, - "loss": 0.1103, + "loss": 0.1209, "step": 1088 }, { "epoch": 4.411139240506329, - "grad_norm": 0.24917108797789872, + "grad_norm": 0.2012888307312296, "learning_rate": 1.5799381247653967e-06, - "loss": 0.1188, + "loss": 0.11, "step": 1089 }, { "epoch": 4.415189873417722, - "grad_norm": 0.24874644502802415, + "grad_norm": 0.20989183896767596, "learning_rate": 1.5579016903747013e-06, - "loss": 0.118, + "loss": 0.1123, "step": 1090 }, { "epoch": 4.419240506329114, - "grad_norm": 0.24919834116918899, + "grad_norm": 0.20811171237500875, "learning_rate": 1.5360137861084656e-06, - "loss": 0.1103, + "loss": 0.1066, "step": 1091 }, { "epoch": 4.423291139240506, - "grad_norm": 0.24655159727237513, + "grad_norm": 0.20996562964240167, "learning_rate": 1.5142745882488475e-06, - "loss": 0.1206, + "loss": 0.1172, "step": 1092 }, { "epoch": 4.427341772151899, - "grad_norm": 0.23431749833371734, + "grad_norm": 0.21055298826319746, "learning_rate": 1.4926842718803691e-06, - "loss": 0.1149, + "loss": 0.1227, "step": 1093 }, { "epoch": 4.431392405063291, - "grad_norm": 0.2320902257704476, + "grad_norm": 0.21826211484484484, "learning_rate": 1.4712430108884657e-06, - "loss": 0.1052, + "loss": 0.1129, "step": 1094 }, { "epoch": 4.435443037974683, - "grad_norm": 0.24059802357561275, + "grad_norm": 0.20575334368513748, "learning_rate": 1.4499509779581078e-06, - "loss": 0.1142, + "loss": 0.1147, "step": 1095 }, { "epoch": 4.439493670886076, - "grad_norm": 0.24590480342834167, + "grad_norm": 0.20683016805102505, "learning_rate": 1.4288083445723988e-06, - "loss": 0.1262, + "loss": 0.1291, "step": 1096 }, { "epoch": 4.443544303797468, - "grad_norm": 0.2353670326696283, + "grad_norm": 0.21522887137057378, "learning_rate": 1.4078152810112045e-06, - "loss": 0.1173, + "loss": 0.1203, "step": 1097 }, { "epoch": 4.44759493670886, - "grad_norm": 0.23548752323785965, + "grad_norm": 0.21198371267282254, "learning_rate": 1.3869719563497697e-06, - "loss": 0.1091, + "loss": 0.111, "step": 1098 }, { "epoch": 4.451645569620253, - "grad_norm": 0.24692128866482752, + "grad_norm": 0.20300825250127186, "learning_rate": 1.3662785384573663e-06, - "loss": 0.1167, + "loss": 0.1175, "step": 1099 }, { "epoch": 4.455696202531645, - "grad_norm": 0.24448142756074148, + "grad_norm": 0.20681093665476818, "learning_rate": 1.3457351939959383e-06, - "loss": 0.1181, + "loss": 0.1143, "step": 1100 }, { "epoch": 4.459746835443038, - "grad_norm": 0.2545052058095343, + "grad_norm": 0.20407933733347589, "learning_rate": 1.3253420884187551e-06, - "loss": 0.1305, + "loss": 0.1347, "step": 1101 }, { "epoch": 4.46379746835443, - "grad_norm": 0.24480725533889608, + "grad_norm": 0.20555664919896124, "learning_rate": 1.3050993859690953e-06, - "loss": 0.1138, + "loss": 0.1131, "step": 1102 }, { "epoch": 4.467848101265822, - "grad_norm": 0.22977802019846974, + "grad_norm": 0.21494197075788835, "learning_rate": 1.2850072496788869e-06, - "loss": 0.0938, + "loss": 0.0946, "step": 1103 }, { "epoch": 4.471898734177215, - "grad_norm": 0.251628861102355, + "grad_norm": 0.20360098681010508, "learning_rate": 1.2650658413674434e-06, - "loss": 0.1293, + "loss": 0.1201, "step": 1104 }, { "epoch": 4.4759493670886075, - "grad_norm": 0.23852345567412667, + "grad_norm": 0.2040934099996908, "learning_rate": 1.2452753216401226e-06, - "loss": 0.1092, + "loss": 0.1163, "step": 1105 }, { "epoch": 4.48, - "grad_norm": 0.23056855744857324, + "grad_norm": 0.211373206426806, "learning_rate": 1.2256358498870503e-06, - "loss": 0.1141, + "loss": 0.1176, "step": 1106 }, { "epoch": 4.4840506329113925, - "grad_norm": 0.2476852625599391, + "grad_norm": 0.20926912822238875, "learning_rate": 1.2061475842818337e-06, - "loss": 0.1204, + "loss": 0.1142, "step": 1107 }, { "epoch": 4.488101265822785, - "grad_norm": 0.221755319703585, + "grad_norm": 0.2159979372098844, "learning_rate": 1.1868106817802816e-06, - "loss": 0.0999, + "loss": 0.1086, "step": 1108 }, { "epoch": 4.4921518987341775, - "grad_norm": 0.2632791615402449, + "grad_norm": 0.2105402047851378, "learning_rate": 1.1676252981191482e-06, - "loss": 0.121, + "loss": 0.1128, "step": 1109 }, { "epoch": 4.49620253164557, - "grad_norm": 0.241314097167227, + "grad_norm": 0.20316782582784526, "learning_rate": 1.1485915878148823e-06, - "loss": 0.119, + "loss": 0.1157, "step": 1110 }, { "epoch": 4.500253164556962, - "grad_norm": 0.2422880935099222, + "grad_norm": 0.22405189785088125, "learning_rate": 1.1297097041623584e-06, - "loss": 0.1089, + "loss": 0.1101, "step": 1111 }, { "epoch": 4.504303797468355, - "grad_norm": 0.2569432782294898, + "grad_norm": 0.21294812237958868, "learning_rate": 1.1109797992336847e-06, - "loss": 0.1222, + "loss": 0.1154, "step": 1112 }, { "epoch": 4.508354430379747, - "grad_norm": 0.2469856938511531, + "grad_norm": 0.21814642926193412, "learning_rate": 1.092402023876933e-06, - "loss": 0.1055, + "loss": 0.104, "step": 1113 }, { "epoch": 4.512405063291139, - "grad_norm": 0.2551701113277187, + "grad_norm": 0.20926351251232184, "learning_rate": 1.0739765277149527e-06, - "loss": 0.1128, + "loss": 0.0972, "step": 1114 }, { "epoch": 4.516455696202532, - "grad_norm": 0.25020302833349345, + "grad_norm": 0.21393826956472128, "learning_rate": 1.0557034591441596e-06, - "loss": 0.1232, + "loss": 0.1242, "step": 1115 }, { "epoch": 4.520506329113924, - "grad_norm": 0.23333914138307743, + "grad_norm": 0.19864420070866448, "learning_rate": 1.0375829653333324e-06, - "loss": 0.1088, + "loss": 0.1149, "step": 1116 }, { "epoch": 4.524556962025317, - "grad_norm": 0.23619402296134515, + "grad_norm": 0.20980522179760241, "learning_rate": 1.0196151922224385e-06, - "loss": 0.1216, + "loss": 0.1207, "step": 1117 }, { "epoch": 4.528607594936709, - "grad_norm": 0.24625664762597585, + "grad_norm": 0.20963704558047147, "learning_rate": 1.0018002845214526e-06, - "loss": 0.1151, + "loss": 0.1158, "step": 1118 }, { "epoch": 4.532658227848101, - "grad_norm": 0.23832620699624682, + "grad_norm": 0.21167444116505832, "learning_rate": 9.841383857091947e-07, - "loss": 0.1163, + "loss": 0.1185, "step": 1119 }, { "epoch": 4.536708860759494, - "grad_norm": 0.23042558963158336, + "grad_norm": 0.2025059229430548, "learning_rate": 9.666296380321616e-07, - "loss": 0.1118, + "loss": 0.1165, "step": 1120 }, { "epoch": 4.540759493670886, - "grad_norm": 0.23676738344956927, + "grad_norm": 0.19894280074565418, "learning_rate": 9.492741825034124e-07, - "loss": 0.1098, + "loss": 0.1099, "step": 1121 }, { "epoch": 4.544810126582279, - "grad_norm": 0.2318742261715398, + "grad_norm": 0.21171253336046708, "learning_rate": 9.320721589013892e-07, - "loss": 0.1143, + "loss": 0.1177, "step": 1122 }, { "epoch": 4.548860759493671, - "grad_norm": 0.23289116145624666, + "grad_norm": 0.20105006257948016, "learning_rate": 9.150237057688339e-07, - "loss": 0.1074, + "loss": 0.1047, "step": 1123 }, { "epoch": 4.552911392405063, - "grad_norm": 0.2580692022965623, + "grad_norm": 0.2032328366210336, "learning_rate": 8.981289604116328e-07, - "loss": 0.1247, + "loss": 0.1213, "step": 1124 }, { "epoch": 4.556962025316456, - "grad_norm": 0.24234598365903134, + "grad_norm": 0.19853846094696845, "learning_rate": 8.813880588977542e-07, - "loss": 0.122, + "loss": 0.1195, "step": 1125 }, { "epoch": 4.561012658227848, - "grad_norm": 0.22226680008874727, + "grad_norm": 0.2074935286115484, "learning_rate": 8.648011360561126e-07, - "loss": 0.1069, + "loss": 0.114, "step": 1126 }, { "epoch": 4.56506329113924, - "grad_norm": 0.2519571050776005, + "grad_norm": 0.22013640458218453, "learning_rate": 8.483683254755037e-07, - "loss": 0.1247, + "loss": 0.1281, "step": 1127 }, { "epoch": 4.569113924050633, - "grad_norm": 0.22265406572701163, + "grad_norm": 0.21339634951818434, "learning_rate": 8.320897595035227e-07, - "loss": 0.1027, + "loss": 0.1123, "step": 1128 }, { "epoch": 4.573164556962025, - "grad_norm": 0.24252522826003195, + "grad_norm": 0.19926989858502814, "learning_rate": 8.159655692455093e-07, - "loss": 0.1264, + "loss": 0.1224, "step": 1129 }, { "epoch": 4.577215189873417, - "grad_norm": 0.24460918273907392, + "grad_norm": 0.21057777132930375, "learning_rate": 7.999958845634648e-07, - "loss": 0.1196, + "loss": 0.1199, "step": 1130 }, { "epoch": 4.58126582278481, - "grad_norm": 0.23659092871490672, + "grad_norm": 0.21230670196649526, "learning_rate": 7.841808340750478e-07, - "loss": 0.1144, + "loss": 0.1138, "step": 1131 }, { "epoch": 4.585316455696202, - "grad_norm": 0.24351647413226774, + "grad_norm": 0.21718564894895476, "learning_rate": 7.685205451524869e-07, - "loss": 0.1242, + "loss": 0.125, "step": 1132 }, { "epoch": 4.589367088607595, - "grad_norm": 0.23862291365231075, + "grad_norm": 0.20942737386401283, "learning_rate": 7.530151439216027e-07, - "loss": 0.1063, + "loss": 0.1106, "step": 1133 }, { "epoch": 4.593417721518987, - "grad_norm": 0.25455211984276244, + "grad_norm": 0.20594666622404534, "learning_rate": 7.376647552607675e-07, - "loss": 0.1227, + "loss": 0.1174, "step": 1134 }, { "epoch": 4.597468354430379, - "grad_norm": 0.24669481287395503, + "grad_norm": 0.20526587440873395, "learning_rate": 7.224695027998963e-07, - "loss": 0.1216, + "loss": 0.1212, "step": 1135 }, { "epoch": 4.601518987341772, - "grad_norm": 0.22275314150387743, + "grad_norm": 0.21472834819077183, "learning_rate": 7.07429508919466e-07, - "loss": 0.1003, + "loss": 0.1019, "step": 1136 }, { "epoch": 4.605569620253164, - "grad_norm": 0.2690682227464728, + "grad_norm": 0.20581667163388773, "learning_rate": 6.925448947495206e-07, - "loss": 0.1389, + "loss": 0.1404, "step": 1137 }, { "epoch": 4.609620253164557, - "grad_norm": 0.24036765338414862, + "grad_norm": 0.21338707700272144, "learning_rate": 6.778157801686936e-07, - "loss": 0.1158, + "loss": 0.1152, "step": 1138 }, { "epoch": 4.613670886075949, - "grad_norm": 0.23722350051672159, + "grad_norm": 0.2053462512645686, "learning_rate": 6.632422838032515e-07, - "loss": 0.1053, + "loss": 0.1051, "step": 1139 }, { "epoch": 4.6177215189873415, - "grad_norm": 0.2555945168377162, + "grad_norm": 0.2041771303195351, "learning_rate": 6.488245230261281e-07, - "loss": 0.1316, + "loss": 0.1259, "step": 1140 }, { "epoch": 4.6217721518987345, - "grad_norm": 0.24876205520465955, + "grad_norm": 0.20335070817851897, "learning_rate": 6.345626139559868e-07, - "loss": 0.1221, + "loss": 0.115, "step": 1141 }, { "epoch": 4.6258227848101265, - "grad_norm": 0.22605418485053172, + "grad_norm": 0.20335588641121002, "learning_rate": 6.204566714562866e-07, - "loss": 0.1033, + "loss": 0.1105, "step": 1142 }, { "epoch": 4.629873417721519, - "grad_norm": 0.25222222933772276, + "grad_norm": 0.1978284104826405, "learning_rate": 6.06506809134344e-07, - "loss": 0.1263, + "loss": 0.1256, "step": 1143 }, { "epoch": 4.633924050632912, - "grad_norm": 0.24684027200432848, + "grad_norm": 0.19588668465530068, "learning_rate": 5.927131393404373e-07, - "loss": 0.1281, + "loss": 0.1229, "step": 1144 }, { "epoch": 4.637974683544304, - "grad_norm": 0.21883930553081657, + "grad_norm": 0.2112026075669013, "learning_rate": 5.790757731668817e-07, - "loss": 0.0972, + "loss": 0.1048, "step": 1145 }, { "epoch": 4.642025316455696, - "grad_norm": 0.23297061406959818, + "grad_norm": 0.1988884909158249, "learning_rate": 5.655948204471507e-07, - "loss": 0.1167, + "loss": 0.1252, "step": 1146 }, { "epoch": 4.646075949367089, - "grad_norm": 0.24262137667778935, + "grad_norm": 0.2052698975996843, "learning_rate": 5.522703897549875e-07, - "loss": 0.1222, + "loss": 0.13, "step": 1147 }, { "epoch": 4.650126582278481, - "grad_norm": 0.24610559178408167, + "grad_norm": 0.2113471441983246, "learning_rate": 5.391025884035239e-07, - "loss": 0.1107, + "loss": 0.118, "step": 1148 }, { "epoch": 4.654177215189874, - "grad_norm": 0.24245511269928224, + "grad_norm": 0.20811912693391285, "learning_rate": 5.260915224444207e-07, - "loss": 0.1057, + "loss": 0.0994, "step": 1149 }, { "epoch": 4.658227848101266, - "grad_norm": 0.2625208354185958, + "grad_norm": 0.21073259866741254, "learning_rate": 5.132372966670129e-07, - "loss": 0.1279, + "loss": 0.1184, "step": 1150 }, { "epoch": 4.662278481012658, - "grad_norm": 0.24017540585212674, + "grad_norm": 0.2122659373264562, "learning_rate": 5.005400145974704e-07, - "loss": 0.1175, + "loss": 0.1233, "step": 1151 }, { "epoch": 4.666329113924051, - "grad_norm": 0.2471224948578742, + "grad_norm": 0.21095871472177763, "learning_rate": 4.879997784979562e-07, - "loss": 0.1324, + "loss": 0.1309, "step": 1152 }, { "epoch": 4.670379746835443, - "grad_norm": 0.23567664757892884, + "grad_norm": 0.20313814160332547, "learning_rate": 4.7561668936580984e-07, - "loss": 0.1103, + "loss": 0.1074, "step": 1153 }, { "epoch": 4.674430379746836, - "grad_norm": 0.24297496373091534, + "grad_norm": 0.2085712200415366, "learning_rate": 4.6339084693272306e-07, - "loss": 0.1127, + "loss": 0.117, "step": 1154 }, { "epoch": 4.678481012658228, - "grad_norm": 0.2242049913364936, + "grad_norm": 0.19007347926883442, "learning_rate": 4.5132234966395847e-07, - "loss": 0.1133, + "loss": 0.1174, "step": 1155 }, { "epoch": 4.68253164556962, - "grad_norm": 0.2382501654068166, + "grad_norm": 0.21950506053306648, "learning_rate": 4.3941129475752795e-07, - "loss": 0.1165, + "loss": 0.1201, "step": 1156 }, { "epoch": 4.686582278481013, - "grad_norm": 0.25027009377600545, + "grad_norm": 0.20730175876913146, "learning_rate": 4.27657778143431e-07, - "loss": 0.1257, + "loss": 0.1279, "step": 1157 }, { "epoch": 4.690632911392405, - "grad_norm": 0.24729443161029124, + "grad_norm": 0.20838929170563114, "learning_rate": 4.1606189448287757e-07, - "loss": 0.126, + "loss": 0.1283, "step": 1158 }, { "epoch": 4.694683544303797, - "grad_norm": 0.22063427319483517, + "grad_norm": 0.20839511874075492, "learning_rate": 4.046237371675177e-07, - "loss": 0.0991, + "loss": 0.1059, "step": 1159 }, { "epoch": 4.69873417721519, - "grad_norm": 0.23872317153101974, + "grad_norm": 0.20355928247494684, "learning_rate": 3.9334339831869963e-07, - "loss": 0.1171, + "loss": 0.1246, "step": 1160 }, { "epoch": 4.702784810126582, - "grad_norm": 0.2454232215665374, + "grad_norm": 0.20633094453618736, "learning_rate": 3.8222096878671955e-07, - "loss": 0.113, + "loss": 0.1166, "step": 1161 }, { "epoch": 4.706835443037974, - "grad_norm": 0.2566607962088178, + "grad_norm": 0.20431024970942696, "learning_rate": 3.7125653815009545e-07, - "loss": 0.1275, + "loss": 0.1282, "step": 1162 }, { "epoch": 4.710886075949367, - "grad_norm": 0.23223551794879096, + "grad_norm": 0.20623206191042928, "learning_rate": 3.6045019471484974e-07, - "loss": 0.1099, + "loss": 0.1177, "step": 1163 }, { "epoch": 4.714936708860759, - "grad_norm": 0.2335554106817479, + "grad_norm": 0.2144716107826845, "learning_rate": 3.498020255137813e-07, - "loss": 0.1102, + "loss": 0.1192, "step": 1164 }, { "epoch": 4.718987341772152, - "grad_norm": 0.23068362089979244, + "grad_norm": 0.21757555870967651, "learning_rate": 3.393121163057811e-07, - "loss": 0.111, + "loss": 0.1155, "step": 1165 }, { "epoch": 4.723037974683544, - "grad_norm": 0.2345192803075931, + "grad_norm": 0.2129426113170404, "learning_rate": 3.289805515751399e-07, - "loss": 0.1194, + "loss": 0.13, "step": 1166 }, { "epoch": 4.727088607594936, - "grad_norm": 0.23895336546569676, + "grad_norm": 0.20367334408369145, "learning_rate": 3.188074145308573e-07, - "loss": 0.1136, + "loss": 0.1072, "step": 1167 }, { "epoch": 4.731139240506329, - "grad_norm": 0.25491084627490324, + "grad_norm": 0.20341722591197034, "learning_rate": 3.087927871059804e-07, - "loss": 0.1237, + "loss": 0.1174, "step": 1168 }, { "epoch": 4.735189873417721, - "grad_norm": 0.23696884696908171, + "grad_norm": 0.20920904027456105, "learning_rate": 2.989367499569418e-07, - "loss": 0.1167, + "loss": 0.12, "step": 1169 }, { "epoch": 4.739240506329114, - "grad_norm": 0.23284577548682825, + "grad_norm": 0.20504474871296832, "learning_rate": 2.8923938246290917e-07, - "loss": 0.1269, + "loss": 0.1405, "step": 1170 }, { "epoch": 4.743291139240506, - "grad_norm": 0.22898652254877466, + "grad_norm": 0.2042104154739712, "learning_rate": 2.7970076272514804e-07, - "loss": 0.11, + "loss": 0.1175, "step": 1171 }, { "epoch": 4.7473417721518985, - "grad_norm": 0.2542478803679144, + "grad_norm": 0.2017294394627823, "learning_rate": 2.703209675663887e-07, - "loss": 0.1259, + "loss": 0.1224, "step": 1172 }, { "epoch": 4.751392405063291, - "grad_norm": 0.23530288530161478, + "grad_norm": 0.1980570388527049, "learning_rate": 2.6110007253021374e-07, - "loss": 0.1124, + "loss": 0.1159, "step": 1173 }, { "epoch": 4.7554430379746835, - "grad_norm": 0.22947761538508535, + "grad_norm": 0.20035308721355388, "learning_rate": 2.520381518804471e-07, - "loss": 0.1013, + "loss": 0.0996, "step": 1174 }, { "epoch": 4.759493670886076, - "grad_norm": 0.2441932009892703, + "grad_norm": 0.20241333870484432, "learning_rate": 2.4313527860054585e-07, - "loss": 0.1165, + "loss": 0.1152, "step": 1175 }, { "epoch": 4.7635443037974685, - "grad_norm": 0.23880575834789772, + "grad_norm": 0.20014095559607262, "learning_rate": 2.343915243930317e-07, - "loss": 0.1082, + "loss": 0.108, "step": 1176 }, { "epoch": 4.767594936708861, - "grad_norm": 0.23188226840846451, + "grad_norm": 0.2128426050120164, "learning_rate": 2.2580695967889367e-07, - "loss": 0.1081, + "loss": 0.1127, "step": 1177 }, { "epoch": 4.7716455696202535, - "grad_norm": 0.24801353237414867, + "grad_norm": 0.19967549439905458, "learning_rate": 2.1738165359704189e-07, - "loss": 0.119, + "loss": 0.1178, "step": 1178 }, { "epoch": 4.775696202531646, - "grad_norm": 0.23533511058057707, + "grad_norm": 0.21115778749209888, "learning_rate": 2.0911567400373257e-07, - "loss": 0.1157, + "loss": 0.1168, "step": 1179 }, { "epoch": 4.779746835443038, - "grad_norm": 0.2326885456962792, + "grad_norm": 0.21039004378596235, "learning_rate": 2.0100908747202607e-07, - "loss": 0.1012, + "loss": 0.0961, "step": 1180 }, { "epoch": 4.783797468354431, - "grad_norm": 0.2459231496153141, + "grad_norm": 0.19953766867394895, "learning_rate": 1.9306195929125638e-07, - "loss": 0.1182, + "loss": 0.11, "step": 1181 }, { "epoch": 4.787848101265823, - "grad_norm": 0.24502708561528558, + "grad_norm": 0.20760586525440083, "learning_rate": 1.8527435346650247e-07, - "loss": 0.1251, + "loss": 0.1224, "step": 1182 }, { "epoch": 4.791898734177215, - "grad_norm": 0.23971410190487166, + "grad_norm": 0.2060238614534191, "learning_rate": 1.7764633271807108e-07, - "loss": 0.1166, + "loss": 0.12, "step": 1183 }, { "epoch": 4.795949367088608, - "grad_norm": 0.2306153114747138, + "grad_norm": 0.21099171015572574, "learning_rate": 1.7017795848099262e-07, - "loss": 0.1098, + "loss": 0.1163, "step": 1184 }, { "epoch": 4.8, - "grad_norm": 0.24082278360563614, + "grad_norm": 0.20425209892239116, "learning_rate": 1.6286929090452596e-07, - "loss": 0.1117, + "loss": 0.107, "step": 1185 }, { "epoch": 4.804050632911393, - "grad_norm": 0.2404790327783987, + "grad_norm": 0.20896229654474527, "learning_rate": 1.557203888516745e-07, - "loss": 0.1139, + "loss": 0.1118, "step": 1186 }, { "epoch": 4.808101265822785, - "grad_norm": 0.24314980164187977, + "grad_norm": 0.20505541101005698, "learning_rate": 1.487313098987131e-07, - "loss": 0.119, + "loss": 0.1188, "step": 1187 }, { "epoch": 4.812151898734177, - "grad_norm": 0.2467760739525281, + "grad_norm": 0.20922577673677953, "learning_rate": 1.4190211033472402e-07, - "loss": 0.1261, + "loss": 0.1274, "step": 1188 }, { "epoch": 4.81620253164557, - "grad_norm": 0.23968572790840473, + "grad_norm": 0.21116939541376195, "learning_rate": 1.3523284516113955e-07, - "loss": 0.1102, + "loss": 0.106, "step": 1189 }, { "epoch": 4.820253164556962, - "grad_norm": 0.23865311575641185, + "grad_norm": 0.20768343006102155, "learning_rate": 1.2872356809130682e-07, - "loss": 0.1079, + "loss": 0.1076, "step": 1190 }, { "epoch": 4.824303797468355, - "grad_norm": 0.24955542761382052, + "grad_norm": 0.20468417782779227, "learning_rate": 1.2237433155004807e-07, - "loss": 0.1228, + "loss": 0.1223, "step": 1191 }, { "epoch": 4.828354430379747, - "grad_norm": 0.23329409983410573, + "grad_norm": 0.2098082124146172, "learning_rate": 1.1618518667323886e-07, - "loss": 0.1093, + "loss": 0.1039, "step": 1192 }, { "epoch": 4.832405063291139, - "grad_norm": 0.2351900937944146, + "grad_norm": 0.21051427840192627, "learning_rate": 1.1015618330740385e-07, - "loss": 0.1195, + "loss": 0.1292, "step": 1193 }, { "epoch": 4.836455696202532, - "grad_norm": 0.22078521394236744, + "grad_norm": 0.19749426940217815, "learning_rate": 1.042873700093061e-07, - "loss": 0.1175, + "loss": 0.1318, "step": 1194 }, { "epoch": 4.840506329113924, - "grad_norm": 0.25050782682166095, + "grad_norm": 0.19820803691814487, "learning_rate": 9.857879404556291e-08, - "loss": 0.1261, + "loss": 0.1231, "step": 1195 }, { "epoch": 4.844556962025316, - "grad_norm": 0.23462662535457032, + "grad_norm": 0.19930057255770592, "learning_rate": 9.303050139225722e-08, - "loss": 0.1054, + "loss": 0.1049, "step": 1196 }, { "epoch": 4.848607594936709, - "grad_norm": 0.23267483877640302, + "grad_norm": 0.20497167366532656, "learning_rate": 8.76425367345779e-08, - "loss": 0.1111, + "loss": 0.1171, "step": 1197 }, { "epoch": 4.852658227848101, - "grad_norm": 0.23081232926077547, + "grad_norm": 0.2087849578411844, "learning_rate": 8.241494346644897e-08, - "loss": 0.122, + "loss": 0.1256, "step": 1198 }, { "epoch": 4.856708860759493, - "grad_norm": 0.2336547692204078, + "grad_norm": 0.20960047207966875, "learning_rate": 7.734776369019204e-08, - "loss": 0.1114, + "loss": 0.1145, "step": 1199 }, { "epoch": 4.860759493670886, - "grad_norm": 0.2435732315044985, + "grad_norm": 0.20709045703359283, "learning_rate": 7.244103821617332e-08, - "loss": 0.1301, + "loss": 0.1276, "step": 1200 }, { "epoch": 4.864810126582278, - "grad_norm": 0.24331432487228732, + "grad_norm": 0.20764505959342205, "learning_rate": 6.769480656248606e-08, - "loss": 0.1209, + "loss": 0.1127, "step": 1201 }, { "epoch": 4.868860759493671, - "grad_norm": 0.24182292100878547, + "grad_norm": 0.20732384965327175, "learning_rate": 6.310910695462635e-08, - "loss": 0.1149, + "loss": 0.1133, "step": 1202 }, { "epoch": 4.872911392405063, - "grad_norm": 0.2436669236921596, + "grad_norm": 0.20349684428929318, "learning_rate": 5.8683976325191185e-08, - "loss": 0.1186, + "loss": 0.1142, "step": 1203 }, { "epoch": 4.876962025316455, - "grad_norm": 0.2365856211364889, + "grad_norm": 0.20097560676224976, "learning_rate": 5.4419450313571984e-08, - "loss": 0.1101, + "loss": 0.1149, "step": 1204 }, { "epoch": 4.881012658227848, - "grad_norm": 0.24306052267481215, + "grad_norm": 0.2023305140232061, "learning_rate": 5.031556326567488e-08, - "loss": 0.1224, + "loss": 0.1242, "step": 1205 }, { "epoch": 4.88506329113924, - "grad_norm": 0.24013256127386048, + "grad_norm": 0.20571007500161037, "learning_rate": 4.637234823364312e-08, - "loss": 0.1171, + "loss": 0.1094, "step": 1206 }, { "epoch": 4.889113924050633, - "grad_norm": 0.23281385145559444, + "grad_norm": 0.2025404660661232, "learning_rate": 4.258983697558838e-08, - "loss": 0.106, + "loss": 0.1087, "step": 1207 }, { "epoch": 4.8931645569620255, - "grad_norm": 0.2388970278631686, + "grad_norm": 0.20330971685864385, "learning_rate": 3.896805995533548e-08, - "loss": 0.1158, + "loss": 0.1147, "step": 1208 }, { "epoch": 4.8972151898734175, - "grad_norm": 0.25062653421138625, + "grad_norm": 0.21059128788978063, "learning_rate": 3.550704634218028e-08, - "loss": 0.1151, + "loss": 0.1117, "step": 1209 }, { "epoch": 4.9012658227848105, - "grad_norm": 0.24402303887641025, + "grad_norm": 0.20481831153392321, "learning_rate": 3.2206824010647676e-08, - "loss": 0.1182, + "loss": 0.1172, "step": 1210 }, { "epoch": 4.905316455696203, - "grad_norm": 0.2354873472393346, + "grad_norm": 0.20538447024850046, "learning_rate": 2.9067419540278476e-08, - "loss": 0.1168, + "loss": 0.122, "step": 1211 }, { "epoch": 4.909367088607595, - "grad_norm": 0.2355432138541809, + "grad_norm": 0.20292283079286413, "learning_rate": 2.6088858215400638e-08, - "loss": 0.1182, + "loss": 0.1245, "step": 1212 }, { "epoch": 4.913417721518988, - "grad_norm": 0.23785954471040072, + "grad_norm": 0.21145456513897448, "learning_rate": 2.3271164024940564e-08, - "loss": 0.1206, + "loss": 0.1318, "step": 1213 }, { "epoch": 4.91746835443038, - "grad_norm": 0.23618835283865625, + "grad_norm": 0.210870267562045, "learning_rate": 2.061435966221881e-08, - "loss": 0.1119, + "loss": 0.1167, "step": 1214 }, { "epoch": 4.921518987341772, - "grad_norm": 0.24287886786881124, + "grad_norm": 0.21228156298120943, "learning_rate": 1.811846652477245e-08, - "loss": 0.112, + "loss": 0.1123, "step": 1215 }, { "epoch": 4.925569620253165, - "grad_norm": 0.24330247700789032, + "grad_norm": 0.20151352196850916, "learning_rate": 1.5783504714184106e-08, - "loss": 0.1133, + "loss": 0.1129, "step": 1216 }, { "epoch": 4.929620253164557, - "grad_norm": 0.2453681957949906, + "grad_norm": 0.2099959540784521, "learning_rate": 1.360949303591097e-08, - "loss": 0.1204, + "loss": 0.1202, "step": 1217 }, { "epoch": 4.93367088607595, - "grad_norm": 0.2241707463640441, + "grad_norm": 0.20824123991142346, "learning_rate": 1.1596448999144916e-08, - "loss": 0.1118, + "loss": 0.1212, "step": 1218 }, { "epoch": 4.937721518987342, - "grad_norm": 0.24397427799132193, + "grad_norm": 0.20295686813461908, "learning_rate": 9.744388816668172e-09, - "loss": 0.1228, + "loss": 0.1275, "step": 1219 }, { "epoch": 4.941772151898734, - "grad_norm": 0.22901886768635465, + "grad_norm": 0.2005956711554205, "learning_rate": 8.05332740472009e-09, - "loss": 0.1076, + "loss": 0.1122, "step": 1220 }, { "epoch": 4.945822784810127, - "grad_norm": 0.2429907277645274, + "grad_norm": 0.20202159349472282, "learning_rate": 6.523278382872811e-09, - "loss": 0.1231, + "loss": 0.1275, "step": 1221 }, { "epoch": 4.949873417721519, - "grad_norm": 0.22358904309323535, + "grad_norm": 0.21763566583306004, "learning_rate": 5.15425407393133e-09, - "loss": 0.1063, + "loss": 0.1111, "step": 1222 }, { "epoch": 4.953924050632912, - "grad_norm": 0.23174779658416808, + "grad_norm": 0.20255252844405391, "learning_rate": 3.94626550383137e-09, - "loss": 0.1146, + "loss": 0.1253, "step": 1223 }, { "epoch": 4.957974683544304, - "grad_norm": 0.23273157676292552, + "grad_norm": 0.21164634900002333, "learning_rate": 2.899322401546112e-09, - "loss": 0.1035, + "loss": 0.1102, "step": 1224 }, { "epoch": 4.962025316455696, - "grad_norm": 0.24719957924676503, + "grad_norm": 0.2106747639379142, "learning_rate": 2.013433199010706e-09, - "loss": 0.1141, + "loss": 0.1129, "step": 1225 }, { "epoch": 4.966075949367089, - "grad_norm": 0.24741891904081048, + "grad_norm": 0.20158062763575213, "learning_rate": 1.2886050310556563e-09, - "loss": 0.1223, + "loss": 0.1321, "step": 1226 }, { "epoch": 4.970126582278481, - "grad_norm": 0.21305228504484663, + "grad_norm": 0.21237574487414587, "learning_rate": 7.248437353468695e-10, - "loss": 0.0946, + "loss": 0.1054, "step": 1227 }, { "epoch": 4.974177215189873, - "grad_norm": 0.23467727674677274, + "grad_norm": 0.19762341059700764, "learning_rate": 3.221538523412449e-10, - "loss": 0.1142, + "loss": 0.1168, "step": 1228 }, { "epoch": 4.978227848101266, - "grad_norm": 0.23906109869252376, + "grad_norm": 0.20565669764660643, "learning_rate": 8.053862524670663e-11, - "loss": 0.1218, + "loss": 0.1198, "step": 1229 }, { "epoch": 4.982278481012658, - "grad_norm": 0.23981752304627496, + "grad_norm": 0.20290470136362806, "learning_rate": 0.0, - "loss": 0.1264, + "loss": 0.1357, "step": 1230 }, { "epoch": 4.982278481012658, "step": 1230, - "total_flos": 2.2387422629367644e+18, - "train_loss": 0.26373260683887373, - "train_runtime": 64181.8261, - "train_samples_per_second": 2.462, - "train_steps_per_second": 0.019 + "total_flos": 2.090332464318251e+18, + "train_loss": 0.24650180314977962, + "train_runtime": 35345.5491, + "train_samples_per_second": 4.47, + "train_steps_per_second": 0.035 } ], - "logging_steps": 1.0, + "logging_steps": 1, "max_steps": 1230, "num_input_tokens_seen": 0, "num_train_epochs": 5, @@ -8645,7 +8645,7 @@ "attributes": {} } }, - "total_flos": 2.2387422629367644e+18, + "total_flos": 2.090332464318251e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null