diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,9555 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 500, + "global_step": 1359, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002207505518763797, + "grad_norm": 2.7993483543395996, + "learning_rate": 7.352941176470589e-08, + "loss": 0.7615, + "step": 1 + }, + { + "epoch": 0.004415011037527594, + "grad_norm": 2.814159393310547, + "learning_rate": 1.4705882352941178e-07, + "loss": 0.7759, + "step": 2 + }, + { + "epoch": 0.006622516556291391, + "grad_norm": 2.932206392288208, + "learning_rate": 2.2058823529411768e-07, + "loss": 0.7897, + "step": 3 + }, + { + "epoch": 0.008830022075055188, + "grad_norm": 2.909721851348877, + "learning_rate": 2.9411764705882356e-07, + "loss": 0.7698, + "step": 4 + }, + { + "epoch": 0.011037527593818985, + "grad_norm": 2.855281114578247, + "learning_rate": 3.6764705882352943e-07, + "loss": 0.7673, + "step": 5 + }, + { + "epoch": 0.013245033112582781, + "grad_norm": 2.930614709854126, + "learning_rate": 4.4117647058823536e-07, + "loss": 0.7914, + "step": 6 + }, + { + "epoch": 0.01545253863134658, + "grad_norm": 2.8653368949890137, + "learning_rate": 5.147058823529412e-07, + "loss": 0.7713, + "step": 7 + }, + { + "epoch": 0.017660044150110375, + "grad_norm": 2.7474284172058105, + "learning_rate": 5.882352941176471e-07, + "loss": 0.7713, + "step": 8 + }, + { + "epoch": 0.019867549668874173, + "grad_norm": 2.6992416381835938, + "learning_rate": 6.61764705882353e-07, + "loss": 0.7549, + "step": 9 + }, + { + "epoch": 0.02207505518763797, + "grad_norm": 2.718104362487793, + "learning_rate": 7.352941176470589e-07, + "loss": 0.757, + "step": 10 + }, + { + "epoch": 0.024282560706401765, + "grad_norm": 2.670048952102661, + "learning_rate": 8.088235294117648e-07, + "loss": 0.7367, + "step": 11 + }, + { + "epoch": 0.026490066225165563, + "grad_norm": 2.205970287322998, + "learning_rate": 8.823529411764707e-07, + "loss": 0.7397, + "step": 12 + }, + { + "epoch": 0.02869757174392936, + "grad_norm": 2.2243752479553223, + "learning_rate": 9.558823529411764e-07, + "loss": 0.7565, + "step": 13 + }, + { + "epoch": 0.03090507726269316, + "grad_norm": 2.2006163597106934, + "learning_rate": 1.0294117647058825e-06, + "loss": 0.744, + "step": 14 + }, + { + "epoch": 0.033112582781456956, + "grad_norm": 2.0692009925842285, + "learning_rate": 1.1029411764705884e-06, + "loss": 0.7282, + "step": 15 + }, + { + "epoch": 0.03532008830022075, + "grad_norm": 1.755692958831787, + "learning_rate": 1.1764705882352942e-06, + "loss": 0.7027, + "step": 16 + }, + { + "epoch": 0.037527593818984545, + "grad_norm": 1.429651141166687, + "learning_rate": 1.25e-06, + "loss": 0.7028, + "step": 17 + }, + { + "epoch": 0.039735099337748346, + "grad_norm": 1.4149061441421509, + "learning_rate": 1.323529411764706e-06, + "loss": 0.7027, + "step": 18 + }, + { + "epoch": 0.04194260485651214, + "grad_norm": 1.4027491807937622, + "learning_rate": 1.3970588235294119e-06, + "loss": 0.7034, + "step": 19 + }, + { + "epoch": 0.04415011037527594, + "grad_norm": 1.3382784128189087, + "learning_rate": 1.4705882352941177e-06, + "loss": 0.6838, + "step": 20 + }, + { + "epoch": 0.046357615894039736, + "grad_norm": 1.2848469018936157, + "learning_rate": 1.5441176470588238e-06, + "loss": 0.6904, + "step": 21 + }, + { + "epoch": 0.04856512141280353, + "grad_norm": 1.0581680536270142, + "learning_rate": 1.6176470588235297e-06, + "loss": 0.6658, + "step": 22 + }, + { + "epoch": 0.05077262693156733, + "grad_norm": 1.0082190036773682, + "learning_rate": 1.6911764705882356e-06, + "loss": 0.653, + "step": 23 + }, + { + "epoch": 0.052980132450331126, + "grad_norm": 0.979928195476532, + "learning_rate": 1.7647058823529414e-06, + "loss": 0.6501, + "step": 24 + }, + { + "epoch": 0.05518763796909492, + "grad_norm": 0.9646239876747131, + "learning_rate": 1.8382352941176473e-06, + "loss": 0.6415, + "step": 25 + }, + { + "epoch": 0.05739514348785872, + "grad_norm": 0.8932761549949646, + "learning_rate": 1.9117647058823528e-06, + "loss": 0.6523, + "step": 26 + }, + { + "epoch": 0.059602649006622516, + "grad_norm": 0.8707468509674072, + "learning_rate": 1.985294117647059e-06, + "loss": 0.6434, + "step": 27 + }, + { + "epoch": 0.06181015452538632, + "grad_norm": 0.8291523456573486, + "learning_rate": 2.058823529411765e-06, + "loss": 0.6333, + "step": 28 + }, + { + "epoch": 0.0640176600441501, + "grad_norm": 0.7280705571174622, + "learning_rate": 2.132352941176471e-06, + "loss": 0.6198, + "step": 29 + }, + { + "epoch": 0.06622516556291391, + "grad_norm": 0.6107202172279358, + "learning_rate": 2.2058823529411767e-06, + "loss": 0.6057, + "step": 30 + }, + { + "epoch": 0.0684326710816777, + "grad_norm": 0.6679992079734802, + "learning_rate": 2.2794117647058826e-06, + "loss": 0.5983, + "step": 31 + }, + { + "epoch": 0.0706401766004415, + "grad_norm": 0.6695526242256165, + "learning_rate": 2.3529411764705885e-06, + "loss": 0.5946, + "step": 32 + }, + { + "epoch": 0.0728476821192053, + "grad_norm": 0.6435050964355469, + "learning_rate": 2.4264705882352943e-06, + "loss": 0.5975, + "step": 33 + }, + { + "epoch": 0.07505518763796909, + "grad_norm": 0.6036580204963684, + "learning_rate": 2.5e-06, + "loss": 0.592, + "step": 34 + }, + { + "epoch": 0.0772626931567329, + "grad_norm": 0.5135894417762756, + "learning_rate": 2.5735294117647057e-06, + "loss": 0.596, + "step": 35 + }, + { + "epoch": 0.07947019867549669, + "grad_norm": 0.4571033716201782, + "learning_rate": 2.647058823529412e-06, + "loss": 0.5764, + "step": 36 + }, + { + "epoch": 0.08167770419426049, + "grad_norm": 0.48609447479248047, + "learning_rate": 2.720588235294118e-06, + "loss": 0.5593, + "step": 37 + }, + { + "epoch": 0.08388520971302428, + "grad_norm": 0.49421966075897217, + "learning_rate": 2.7941176470588237e-06, + "loss": 0.5634, + "step": 38 + }, + { + "epoch": 0.08609271523178808, + "grad_norm": 0.48713281750679016, + "learning_rate": 2.8676470588235296e-06, + "loss": 0.5577, + "step": 39 + }, + { + "epoch": 0.08830022075055188, + "grad_norm": 0.42998674511909485, + "learning_rate": 2.9411764705882355e-06, + "loss": 0.5602, + "step": 40 + }, + { + "epoch": 0.09050772626931568, + "grad_norm": 0.39199528098106384, + "learning_rate": 3.0147058823529413e-06, + "loss": 0.5509, + "step": 41 + }, + { + "epoch": 0.09271523178807947, + "grad_norm": 0.3977169692516327, + "learning_rate": 3.0882352941176476e-06, + "loss": 0.5408, + "step": 42 + }, + { + "epoch": 0.09492273730684327, + "grad_norm": 0.3659592866897583, + "learning_rate": 3.161764705882353e-06, + "loss": 0.5468, + "step": 43 + }, + { + "epoch": 0.09713024282560706, + "grad_norm": 0.41232988238334656, + "learning_rate": 3.2352941176470594e-06, + "loss": 0.5312, + "step": 44 + }, + { + "epoch": 0.09933774834437085, + "grad_norm": 0.41818928718566895, + "learning_rate": 3.308823529411765e-06, + "loss": 0.542, + "step": 45 + }, + { + "epoch": 0.10154525386313466, + "grad_norm": 0.38174012303352356, + "learning_rate": 3.382352941176471e-06, + "loss": 0.5341, + "step": 46 + }, + { + "epoch": 0.10375275938189846, + "grad_norm": 0.3903500735759735, + "learning_rate": 3.4558823529411766e-06, + "loss": 0.5215, + "step": 47 + }, + { + "epoch": 0.10596026490066225, + "grad_norm": 0.3658589720726013, + "learning_rate": 3.529411764705883e-06, + "loss": 0.5277, + "step": 48 + }, + { + "epoch": 0.10816777041942605, + "grad_norm": 0.3160182237625122, + "learning_rate": 3.6029411764705883e-06, + "loss": 0.5193, + "step": 49 + }, + { + "epoch": 0.11037527593818984, + "grad_norm": 0.29906004667282104, + "learning_rate": 3.6764705882352946e-06, + "loss": 0.525, + "step": 50 + }, + { + "epoch": 0.11258278145695365, + "grad_norm": 0.28197285532951355, + "learning_rate": 3.7500000000000005e-06, + "loss": 0.5136, + "step": 51 + }, + { + "epoch": 0.11479028697571744, + "grad_norm": 0.2733807861804962, + "learning_rate": 3.8235294117647055e-06, + "loss": 0.5126, + "step": 52 + }, + { + "epoch": 0.11699779249448124, + "grad_norm": 0.2581369876861572, + "learning_rate": 3.897058823529412e-06, + "loss": 0.5102, + "step": 53 + }, + { + "epoch": 0.11920529801324503, + "grad_norm": 0.24509315192699432, + "learning_rate": 3.970588235294118e-06, + "loss": 0.4959, + "step": 54 + }, + { + "epoch": 0.12141280353200883, + "grad_norm": 0.23736661672592163, + "learning_rate": 4.044117647058824e-06, + "loss": 0.5024, + "step": 55 + }, + { + "epoch": 0.12362030905077263, + "grad_norm": 0.2297072857618332, + "learning_rate": 4.11764705882353e-06, + "loss": 0.516, + "step": 56 + }, + { + "epoch": 0.12582781456953643, + "grad_norm": 0.21641261875629425, + "learning_rate": 4.191176470588236e-06, + "loss": 0.5045, + "step": 57 + }, + { + "epoch": 0.1280353200883002, + "grad_norm": 0.2174406796693802, + "learning_rate": 4.264705882352942e-06, + "loss": 0.4954, + "step": 58 + }, + { + "epoch": 0.13024282560706402, + "grad_norm": 0.2182844579219818, + "learning_rate": 4.3382352941176475e-06, + "loss": 0.5073, + "step": 59 + }, + { + "epoch": 0.13245033112582782, + "grad_norm": 0.18668220937252045, + "learning_rate": 4.411764705882353e-06, + "loss": 0.4927, + "step": 60 + }, + { + "epoch": 0.1346578366445916, + "grad_norm": 0.20196162164211273, + "learning_rate": 4.485294117647059e-06, + "loss": 0.4962, + "step": 61 + }, + { + "epoch": 0.1368653421633554, + "grad_norm": 0.20648355782032013, + "learning_rate": 4.558823529411765e-06, + "loss": 0.4965, + "step": 62 + }, + { + "epoch": 0.1390728476821192, + "grad_norm": 0.1998893916606903, + "learning_rate": 4.632352941176471e-06, + "loss": 0.4857, + "step": 63 + }, + { + "epoch": 0.141280353200883, + "grad_norm": 0.20052312314510345, + "learning_rate": 4.705882352941177e-06, + "loss": 0.4919, + "step": 64 + }, + { + "epoch": 0.1434878587196468, + "grad_norm": 0.18653374910354614, + "learning_rate": 4.779411764705883e-06, + "loss": 0.4895, + "step": 65 + }, + { + "epoch": 0.1456953642384106, + "grad_norm": 0.17638395726680756, + "learning_rate": 4.852941176470589e-06, + "loss": 0.4831, + "step": 66 + }, + { + "epoch": 0.1479028697571744, + "grad_norm": 0.17102564871311188, + "learning_rate": 4.9264705882352945e-06, + "loss": 0.4817, + "step": 67 + }, + { + "epoch": 0.15011037527593818, + "grad_norm": 0.15633539855480194, + "learning_rate": 5e-06, + "loss": 0.4882, + "step": 68 + }, + { + "epoch": 0.152317880794702, + "grad_norm": 0.17420779168605804, + "learning_rate": 5.073529411764706e-06, + "loss": 0.4853, + "step": 69 + }, + { + "epoch": 0.1545253863134658, + "grad_norm": 0.1659373641014099, + "learning_rate": 5.147058823529411e-06, + "loss": 0.492, + "step": 70 + }, + { + "epoch": 0.15673289183222958, + "grad_norm": 0.17147059738636017, + "learning_rate": 5.220588235294118e-06, + "loss": 0.4744, + "step": 71 + }, + { + "epoch": 0.15894039735099338, + "grad_norm": 0.15868496894836426, + "learning_rate": 5.294117647058824e-06, + "loss": 0.4845, + "step": 72 + }, + { + "epoch": 0.16114790286975716, + "grad_norm": 0.16405610740184784, + "learning_rate": 5.36764705882353e-06, + "loss": 0.4806, + "step": 73 + }, + { + "epoch": 0.16335540838852097, + "grad_norm": 0.16762660443782806, + "learning_rate": 5.441176470588236e-06, + "loss": 0.465, + "step": 74 + }, + { + "epoch": 0.16556291390728478, + "grad_norm": 0.209846630692482, + "learning_rate": 5.514705882352942e-06, + "loss": 0.486, + "step": 75 + }, + { + "epoch": 0.16777041942604856, + "grad_norm": 0.17751334607601166, + "learning_rate": 5.588235294117647e-06, + "loss": 0.4761, + "step": 76 + }, + { + "epoch": 0.16997792494481237, + "grad_norm": 0.16121278703212738, + "learning_rate": 5.661764705882353e-06, + "loss": 0.4774, + "step": 77 + }, + { + "epoch": 0.17218543046357615, + "grad_norm": 0.2625029385089874, + "learning_rate": 5.735294117647059e-06, + "loss": 0.4663, + "step": 78 + }, + { + "epoch": 0.17439293598233996, + "grad_norm": 0.15058760344982147, + "learning_rate": 5.808823529411766e-06, + "loss": 0.4644, + "step": 79 + }, + { + "epoch": 0.17660044150110377, + "grad_norm": 0.16306105256080627, + "learning_rate": 5.882352941176471e-06, + "loss": 0.4745, + "step": 80 + }, + { + "epoch": 0.17880794701986755, + "grad_norm": 0.1497834473848343, + "learning_rate": 5.955882352941177e-06, + "loss": 0.4614, + "step": 81 + }, + { + "epoch": 0.18101545253863136, + "grad_norm": 0.15127182006835938, + "learning_rate": 6.029411764705883e-06, + "loss": 0.468, + "step": 82 + }, + { + "epoch": 0.18322295805739514, + "grad_norm": 0.14110144972801208, + "learning_rate": 6.102941176470589e-06, + "loss": 0.4709, + "step": 83 + }, + { + "epoch": 0.18543046357615894, + "grad_norm": 0.150424063205719, + "learning_rate": 6.176470588235295e-06, + "loss": 0.4614, + "step": 84 + }, + { + "epoch": 0.18763796909492272, + "grad_norm": 0.14802858233451843, + "learning_rate": 6.25e-06, + "loss": 0.4549, + "step": 85 + }, + { + "epoch": 0.18984547461368653, + "grad_norm": 0.1567091941833496, + "learning_rate": 6.323529411764706e-06, + "loss": 0.4487, + "step": 86 + }, + { + "epoch": 0.19205298013245034, + "grad_norm": 0.14341039955615997, + "learning_rate": 6.397058823529412e-06, + "loss": 0.4453, + "step": 87 + }, + { + "epoch": 0.19426048565121412, + "grad_norm": 0.20370961725711823, + "learning_rate": 6.470588235294119e-06, + "loss": 0.4626, + "step": 88 + }, + { + "epoch": 0.19646799116997793, + "grad_norm": 0.14623787999153137, + "learning_rate": 6.544117647058824e-06, + "loss": 0.4559, + "step": 89 + }, + { + "epoch": 0.1986754966887417, + "grad_norm": 0.1428503692150116, + "learning_rate": 6.61764705882353e-06, + "loss": 0.4464, + "step": 90 + }, + { + "epoch": 0.20088300220750552, + "grad_norm": 0.14742067456245422, + "learning_rate": 6.6911764705882356e-06, + "loss": 0.4663, + "step": 91 + }, + { + "epoch": 0.20309050772626933, + "grad_norm": 0.16871479153633118, + "learning_rate": 6.764705882352942e-06, + "loss": 0.4549, + "step": 92 + }, + { + "epoch": 0.2052980132450331, + "grad_norm": 0.14448532462120056, + "learning_rate": 6.838235294117648e-06, + "loss": 0.4544, + "step": 93 + }, + { + "epoch": 0.20750551876379691, + "grad_norm": 0.17373333871364594, + "learning_rate": 6.911764705882353e-06, + "loss": 0.4557, + "step": 94 + }, + { + "epoch": 0.2097130242825607, + "grad_norm": 0.15801453590393066, + "learning_rate": 6.985294117647059e-06, + "loss": 0.4507, + "step": 95 + }, + { + "epoch": 0.2119205298013245, + "grad_norm": 0.15145470201969147, + "learning_rate": 7.058823529411766e-06, + "loss": 0.4499, + "step": 96 + }, + { + "epoch": 0.2141280353200883, + "grad_norm": 0.14600904285907745, + "learning_rate": 7.132352941176472e-06, + "loss": 0.4548, + "step": 97 + }, + { + "epoch": 0.2163355408388521, + "grad_norm": 0.15834525227546692, + "learning_rate": 7.205882352941177e-06, + "loss": 0.4505, + "step": 98 + }, + { + "epoch": 0.2185430463576159, + "grad_norm": 0.15612734854221344, + "learning_rate": 7.2794117647058826e-06, + "loss": 0.4598, + "step": 99 + }, + { + "epoch": 0.22075055187637968, + "grad_norm": 0.18165510892868042, + "learning_rate": 7.352941176470589e-06, + "loss": 0.4479, + "step": 100 + }, + { + "epoch": 0.2229580573951435, + "grad_norm": 0.16820134222507477, + "learning_rate": 7.426470588235295e-06, + "loss": 0.4497, + "step": 101 + }, + { + "epoch": 0.2251655629139073, + "grad_norm": 0.16453172266483307, + "learning_rate": 7.500000000000001e-06, + "loss": 0.4532, + "step": 102 + }, + { + "epoch": 0.22737306843267108, + "grad_norm": 0.15813149511814117, + "learning_rate": 7.573529411764706e-06, + "loss": 0.4513, + "step": 103 + }, + { + "epoch": 0.22958057395143489, + "grad_norm": 0.1652165949344635, + "learning_rate": 7.647058823529411e-06, + "loss": 0.445, + "step": 104 + }, + { + "epoch": 0.23178807947019867, + "grad_norm": 0.14849768579006195, + "learning_rate": 7.720588235294119e-06, + "loss": 0.435, + "step": 105 + }, + { + "epoch": 0.23399558498896247, + "grad_norm": 0.17478714883327484, + "learning_rate": 7.794117647058825e-06, + "loss": 0.4343, + "step": 106 + }, + { + "epoch": 0.23620309050772628, + "grad_norm": 0.16727301478385925, + "learning_rate": 7.86764705882353e-06, + "loss": 0.4483, + "step": 107 + }, + { + "epoch": 0.23841059602649006, + "grad_norm": 0.1674540638923645, + "learning_rate": 7.941176470588236e-06, + "loss": 0.4452, + "step": 108 + }, + { + "epoch": 0.24061810154525387, + "grad_norm": 0.16588866710662842, + "learning_rate": 8.014705882352942e-06, + "loss": 0.4517, + "step": 109 + }, + { + "epoch": 0.24282560706401765, + "grad_norm": 0.176283061504364, + "learning_rate": 8.088235294117648e-06, + "loss": 0.4434, + "step": 110 + }, + { + "epoch": 0.24503311258278146, + "grad_norm": 0.17021676898002625, + "learning_rate": 8.161764705882354e-06, + "loss": 0.442, + "step": 111 + }, + { + "epoch": 0.24724061810154527, + "grad_norm": 0.15938061475753784, + "learning_rate": 8.23529411764706e-06, + "loss": 0.4436, + "step": 112 + }, + { + "epoch": 0.24944812362030905, + "grad_norm": 0.18990886211395264, + "learning_rate": 8.308823529411766e-06, + "loss": 0.439, + "step": 113 + }, + { + "epoch": 0.25165562913907286, + "grad_norm": 0.16168883442878723, + "learning_rate": 8.382352941176472e-06, + "loss": 0.4392, + "step": 114 + }, + { + "epoch": 0.25386313465783666, + "grad_norm": 0.2176610231399536, + "learning_rate": 8.455882352941177e-06, + "loss": 0.4488, + "step": 115 + }, + { + "epoch": 0.2560706401766004, + "grad_norm": 0.19106176495552063, + "learning_rate": 8.529411764705883e-06, + "loss": 0.436, + "step": 116 + }, + { + "epoch": 0.2582781456953642, + "grad_norm": 0.1480225920677185, + "learning_rate": 8.60294117647059e-06, + "loss": 0.4386, + "step": 117 + }, + { + "epoch": 0.26048565121412803, + "grad_norm": 0.20528610050678253, + "learning_rate": 8.676470588235295e-06, + "loss": 0.4441, + "step": 118 + }, + { + "epoch": 0.26269315673289184, + "grad_norm": 0.18629467487335205, + "learning_rate": 8.750000000000001e-06, + "loss": 0.4322, + "step": 119 + }, + { + "epoch": 0.26490066225165565, + "grad_norm": 0.1764117330312729, + "learning_rate": 8.823529411764707e-06, + "loss": 0.4261, + "step": 120 + }, + { + "epoch": 0.2671081677704194, + "grad_norm": 0.1975659281015396, + "learning_rate": 8.897058823529413e-06, + "loss": 0.4401, + "step": 121 + }, + { + "epoch": 0.2693156732891832, + "grad_norm": 0.2136935442686081, + "learning_rate": 8.970588235294119e-06, + "loss": 0.4416, + "step": 122 + }, + { + "epoch": 0.271523178807947, + "grad_norm": 0.18505676090717316, + "learning_rate": 9.044117647058824e-06, + "loss": 0.4423, + "step": 123 + }, + { + "epoch": 0.2737306843267108, + "grad_norm": 0.176743283867836, + "learning_rate": 9.11764705882353e-06, + "loss": 0.4435, + "step": 124 + }, + { + "epoch": 0.27593818984547464, + "grad_norm": 0.17542196810245514, + "learning_rate": 9.191176470588236e-06, + "loss": 0.4255, + "step": 125 + }, + { + "epoch": 0.2781456953642384, + "grad_norm": 0.15908770263195038, + "learning_rate": 9.264705882352942e-06, + "loss": 0.438, + "step": 126 + }, + { + "epoch": 0.2803532008830022, + "grad_norm": 0.16695120930671692, + "learning_rate": 9.338235294117648e-06, + "loss": 0.4418, + "step": 127 + }, + { + "epoch": 0.282560706401766, + "grad_norm": 0.1714697927236557, + "learning_rate": 9.411764705882354e-06, + "loss": 0.4293, + "step": 128 + }, + { + "epoch": 0.2847682119205298, + "grad_norm": 0.1601938009262085, + "learning_rate": 9.48529411764706e-06, + "loss": 0.4402, + "step": 129 + }, + { + "epoch": 0.2869757174392936, + "grad_norm": 0.16425947844982147, + "learning_rate": 9.558823529411766e-06, + "loss": 0.4328, + "step": 130 + }, + { + "epoch": 0.2891832229580574, + "grad_norm": 0.1931643933057785, + "learning_rate": 9.632352941176471e-06, + "loss": 0.4307, + "step": 131 + }, + { + "epoch": 0.2913907284768212, + "grad_norm": 0.18675902485847473, + "learning_rate": 9.705882352941177e-06, + "loss": 0.4554, + "step": 132 + }, + { + "epoch": 0.293598233995585, + "grad_norm": 0.16391406953334808, + "learning_rate": 9.779411764705883e-06, + "loss": 0.4272, + "step": 133 + }, + { + "epoch": 0.2958057395143488, + "grad_norm": 0.20189572870731354, + "learning_rate": 9.852941176470589e-06, + "loss": 0.4258, + "step": 134 + }, + { + "epoch": 0.2980132450331126, + "grad_norm": 0.16828037798404694, + "learning_rate": 9.926470588235295e-06, + "loss": 0.4287, + "step": 135 + }, + { + "epoch": 0.30022075055187636, + "grad_norm": 0.1638776957988739, + "learning_rate": 1e-05, + "loss": 0.4326, + "step": 136 + }, + { + "epoch": 0.30242825607064017, + "grad_norm": 0.20775483548641205, + "learning_rate": 9.999983503697906e-06, + "loss": 0.4274, + "step": 137 + }, + { + "epoch": 0.304635761589404, + "grad_norm": 0.17668417096138, + "learning_rate": 9.999934014900475e-06, + "loss": 0.4277, + "step": 138 + }, + { + "epoch": 0.3068432671081678, + "grad_norm": 0.20949822664260864, + "learning_rate": 9.999851533934259e-06, + "loss": 0.4277, + "step": 139 + }, + { + "epoch": 0.3090507726269316, + "grad_norm": 0.13797180354595184, + "learning_rate": 9.999736061343512e-06, + "loss": 0.4072, + "step": 140 + }, + { + "epoch": 0.31125827814569534, + "grad_norm": 0.20374132692813873, + "learning_rate": 9.99958759789018e-06, + "loss": 0.4302, + "step": 141 + }, + { + "epoch": 0.31346578366445915, + "grad_norm": 0.1946762204170227, + "learning_rate": 9.999406144553905e-06, + "loss": 0.4213, + "step": 142 + }, + { + "epoch": 0.31567328918322296, + "grad_norm": 0.21811267733573914, + "learning_rate": 9.999191702532008e-06, + "loss": 0.4285, + "step": 143 + }, + { + "epoch": 0.31788079470198677, + "grad_norm": 0.20533326268196106, + "learning_rate": 9.99894427323949e-06, + "loss": 0.4251, + "step": 144 + }, + { + "epoch": 0.3200883002207506, + "grad_norm": 0.20235486328601837, + "learning_rate": 9.99866385830902e-06, + "loss": 0.4237, + "step": 145 + }, + { + "epoch": 0.32229580573951433, + "grad_norm": 0.20962080359458923, + "learning_rate": 9.99835045959092e-06, + "loss": 0.4266, + "step": 146 + }, + { + "epoch": 0.32450331125827814, + "grad_norm": 0.22475510835647583, + "learning_rate": 9.998004079153156e-06, + "loss": 0.4263, + "step": 147 + }, + { + "epoch": 0.32671081677704195, + "grad_norm": 0.20724737644195557, + "learning_rate": 9.997624719281332e-06, + "loss": 0.416, + "step": 148 + }, + { + "epoch": 0.32891832229580575, + "grad_norm": 0.1801760494709015, + "learning_rate": 9.997212382478658e-06, + "loss": 0.4233, + "step": 149 + }, + { + "epoch": 0.33112582781456956, + "grad_norm": 0.23205707967281342, + "learning_rate": 9.996767071465947e-06, + "loss": 0.4277, + "step": 150 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 0.18537338078022003, + "learning_rate": 9.996288789181595e-06, + "loss": 0.4317, + "step": 151 + }, + { + "epoch": 0.3355408388520971, + "grad_norm": 0.17498917877674103, + "learning_rate": 9.995777538781556e-06, + "loss": 0.4288, + "step": 152 + }, + { + "epoch": 0.33774834437086093, + "grad_norm": 0.18248897790908813, + "learning_rate": 9.995233323639326e-06, + "loss": 0.4261, + "step": 153 + }, + { + "epoch": 0.33995584988962474, + "grad_norm": 0.1849048137664795, + "learning_rate": 9.994656147345922e-06, + "loss": 0.4216, + "step": 154 + }, + { + "epoch": 0.34216335540838855, + "grad_norm": 0.19169744849205017, + "learning_rate": 9.994046013709852e-06, + "loss": 0.423, + "step": 155 + }, + { + "epoch": 0.3443708609271523, + "grad_norm": 0.18524128198623657, + "learning_rate": 9.993402926757098e-06, + "loss": 0.4213, + "step": 156 + }, + { + "epoch": 0.3465783664459161, + "grad_norm": 0.16218866407871246, + "learning_rate": 9.99272689073108e-06, + "loss": 0.4252, + "step": 157 + }, + { + "epoch": 0.3487858719646799, + "grad_norm": 0.17451681196689606, + "learning_rate": 9.992017910092636e-06, + "loss": 0.4251, + "step": 158 + }, + { + "epoch": 0.3509933774834437, + "grad_norm": 0.15944437682628632, + "learning_rate": 9.991275989519991e-06, + "loss": 0.4123, + "step": 159 + }, + { + "epoch": 0.35320088300220753, + "grad_norm": 0.17371642589569092, + "learning_rate": 9.990501133908722e-06, + "loss": 0.4234, + "step": 160 + }, + { + "epoch": 0.3554083885209713, + "grad_norm": 0.1783660650253296, + "learning_rate": 9.98969334837173e-06, + "loss": 0.4234, + "step": 161 + }, + { + "epoch": 0.3576158940397351, + "grad_norm": 0.1762082278728485, + "learning_rate": 9.988852638239206e-06, + "loss": 0.418, + "step": 162 + }, + { + "epoch": 0.3598233995584989, + "grad_norm": 0.16491912305355072, + "learning_rate": 9.987979009058593e-06, + "loss": 0.4248, + "step": 163 + }, + { + "epoch": 0.3620309050772627, + "grad_norm": 0.2407284379005432, + "learning_rate": 9.98707246659455e-06, + "loss": 0.4307, + "step": 164 + }, + { + "epoch": 0.36423841059602646, + "grad_norm": 0.19042451679706573, + "learning_rate": 9.986133016828916e-06, + "loss": 0.4231, + "step": 165 + }, + { + "epoch": 0.36644591611479027, + "grad_norm": 0.18169504404067993, + "learning_rate": 9.985160665960672e-06, + "loss": 0.4266, + "step": 166 + }, + { + "epoch": 0.3686534216335541, + "grad_norm": 0.18646620213985443, + "learning_rate": 9.984155420405895e-06, + "loss": 0.4231, + "step": 167 + }, + { + "epoch": 0.3708609271523179, + "grad_norm": 0.19154079258441925, + "learning_rate": 9.983117286797718e-06, + "loss": 0.4308, + "step": 168 + }, + { + "epoch": 0.3730684326710817, + "grad_norm": 0.17594484984874725, + "learning_rate": 9.982046271986287e-06, + "loss": 0.4115, + "step": 169 + }, + { + "epoch": 0.37527593818984545, + "grad_norm": 0.18167531490325928, + "learning_rate": 9.980942383038717e-06, + "loss": 0.424, + "step": 170 + }, + { + "epoch": 0.37748344370860926, + "grad_norm": 0.1535561978816986, + "learning_rate": 9.97980562723904e-06, + "loss": 0.4296, + "step": 171 + }, + { + "epoch": 0.37969094922737306, + "grad_norm": 0.15756377577781677, + "learning_rate": 9.978636012088165e-06, + "loss": 0.4169, + "step": 172 + }, + { + "epoch": 0.3818984547461369, + "grad_norm": 0.15670788288116455, + "learning_rate": 9.97743354530382e-06, + "loss": 0.4394, + "step": 173 + }, + { + "epoch": 0.3841059602649007, + "grad_norm": 0.16224409639835358, + "learning_rate": 9.976198234820509e-06, + "loss": 0.4228, + "step": 174 + }, + { + "epoch": 0.38631346578366443, + "grad_norm": 0.14743737876415253, + "learning_rate": 9.974930088789452e-06, + "loss": 0.4144, + "step": 175 + }, + { + "epoch": 0.38852097130242824, + "grad_norm": 0.1594422310590744, + "learning_rate": 9.97362911557854e-06, + "loss": 0.4168, + "step": 176 + }, + { + "epoch": 0.39072847682119205, + "grad_norm": 0.16173714399337769, + "learning_rate": 9.972295323772268e-06, + "loss": 0.4166, + "step": 177 + }, + { + "epoch": 0.39293598233995586, + "grad_norm": 0.1668204814195633, + "learning_rate": 9.970928722171691e-06, + "loss": 0.4252, + "step": 178 + }, + { + "epoch": 0.39514348785871967, + "grad_norm": 0.15836164355278015, + "learning_rate": 9.96952931979436e-06, + "loss": 0.4209, + "step": 179 + }, + { + "epoch": 0.3973509933774834, + "grad_norm": 0.1634080708026886, + "learning_rate": 9.968097125874258e-06, + "loss": 0.4076, + "step": 180 + }, + { + "epoch": 0.3995584988962472, + "grad_norm": 0.16134855151176453, + "learning_rate": 9.966632149861748e-06, + "loss": 0.4276, + "step": 181 + }, + { + "epoch": 0.40176600441501104, + "grad_norm": 0.15993578732013702, + "learning_rate": 9.965134401423503e-06, + "loss": 0.4308, + "step": 182 + }, + { + "epoch": 0.40397350993377484, + "grad_norm": 0.19045297801494598, + "learning_rate": 9.963603890442448e-06, + "loss": 0.4185, + "step": 183 + }, + { + "epoch": 0.40618101545253865, + "grad_norm": 0.16455209255218506, + "learning_rate": 9.962040627017693e-06, + "loss": 0.4232, + "step": 184 + }, + { + "epoch": 0.4083885209713024, + "grad_norm": 0.16577620804309845, + "learning_rate": 9.960444621464462e-06, + "loss": 0.4149, + "step": 185 + }, + { + "epoch": 0.4105960264900662, + "grad_norm": 0.25177431106567383, + "learning_rate": 9.958815884314033e-06, + "loss": 0.4172, + "step": 186 + }, + { + "epoch": 0.41280353200883, + "grad_norm": 0.18712477385997772, + "learning_rate": 9.957154426313662e-06, + "loss": 0.4205, + "step": 187 + }, + { + "epoch": 0.41501103752759383, + "grad_norm": 0.1667563021183014, + "learning_rate": 9.955460258426512e-06, + "loss": 0.4207, + "step": 188 + }, + { + "epoch": 0.41721854304635764, + "grad_norm": 0.22865413129329681, + "learning_rate": 9.953733391831586e-06, + "loss": 0.4109, + "step": 189 + }, + { + "epoch": 0.4194260485651214, + "grad_norm": 0.18536990880966187, + "learning_rate": 9.951973837923652e-06, + "loss": 0.4187, + "step": 190 + }, + { + "epoch": 0.4216335540838852, + "grad_norm": 0.19504587352275848, + "learning_rate": 9.950181608313158e-06, + "loss": 0.4142, + "step": 191 + }, + { + "epoch": 0.423841059602649, + "grad_norm": 0.1892482340335846, + "learning_rate": 9.948356714826172e-06, + "loss": 0.4142, + "step": 192 + }, + { + "epoch": 0.4260485651214128, + "grad_norm": 0.1839127093553543, + "learning_rate": 9.946499169504294e-06, + "loss": 0.4161, + "step": 193 + }, + { + "epoch": 0.4282560706401766, + "grad_norm": 0.20385828614234924, + "learning_rate": 9.944608984604569e-06, + "loss": 0.4124, + "step": 194 + }, + { + "epoch": 0.4304635761589404, + "grad_norm": 0.1948205530643463, + "learning_rate": 9.942686172599425e-06, + "loss": 0.4251, + "step": 195 + }, + { + "epoch": 0.4326710816777042, + "grad_norm": 0.19438982009887695, + "learning_rate": 9.940730746176578e-06, + "loss": 0.4158, + "step": 196 + }, + { + "epoch": 0.434878587196468, + "grad_norm": 0.17213338613510132, + "learning_rate": 9.93874271823895e-06, + "loss": 0.4175, + "step": 197 + }, + { + "epoch": 0.4370860927152318, + "grad_norm": 0.22870118916034698, + "learning_rate": 9.936722101904582e-06, + "loss": 0.4267, + "step": 198 + }, + { + "epoch": 0.4392935982339956, + "grad_norm": 0.20383016765117645, + "learning_rate": 9.934668910506555e-06, + "loss": 0.422, + "step": 199 + }, + { + "epoch": 0.44150110375275936, + "grad_norm": 0.16936808824539185, + "learning_rate": 9.932583157592896e-06, + "loss": 0.4144, + "step": 200 + }, + { + "epoch": 0.44370860927152317, + "grad_norm": 0.19149592518806458, + "learning_rate": 9.930464856926488e-06, + "loss": 0.4077, + "step": 201 + }, + { + "epoch": 0.445916114790287, + "grad_norm": 0.15204112231731415, + "learning_rate": 9.928314022484982e-06, + "loss": 0.4207, + "step": 202 + }, + { + "epoch": 0.4481236203090508, + "grad_norm": 0.19697798788547516, + "learning_rate": 9.926130668460702e-06, + "loss": 0.4159, + "step": 203 + }, + { + "epoch": 0.4503311258278146, + "grad_norm": 0.1751161813735962, + "learning_rate": 9.92391480926056e-06, + "loss": 0.4179, + "step": 204 + }, + { + "epoch": 0.45253863134657835, + "grad_norm": 0.17122775316238403, + "learning_rate": 9.921666459505944e-06, + "loss": 0.4104, + "step": 205 + }, + { + "epoch": 0.45474613686534215, + "grad_norm": 0.18898116052150726, + "learning_rate": 9.91938563403264e-06, + "loss": 0.4183, + "step": 206 + }, + { + "epoch": 0.45695364238410596, + "grad_norm": 0.17628274857997894, + "learning_rate": 9.917072347890721e-06, + "loss": 0.411, + "step": 207 + }, + { + "epoch": 0.45916114790286977, + "grad_norm": 0.19427300989627838, + "learning_rate": 9.914726616344454e-06, + "loss": 0.4144, + "step": 208 + }, + { + "epoch": 0.4613686534216336, + "grad_norm": 0.19164274632930756, + "learning_rate": 9.912348454872196e-06, + "loss": 0.4067, + "step": 209 + }, + { + "epoch": 0.46357615894039733, + "grad_norm": 0.17385149002075195, + "learning_rate": 9.909937879166298e-06, + "loss": 0.408, + "step": 210 + }, + { + "epoch": 0.46578366445916114, + "grad_norm": 0.18648236989974976, + "learning_rate": 9.907494905132994e-06, + "loss": 0.4273, + "step": 211 + }, + { + "epoch": 0.46799116997792495, + "grad_norm": 0.1911754459142685, + "learning_rate": 9.905019548892296e-06, + "loss": 0.4127, + "step": 212 + }, + { + "epoch": 0.47019867549668876, + "grad_norm": 0.15749873220920563, + "learning_rate": 9.902511826777895e-06, + "loss": 0.4174, + "step": 213 + }, + { + "epoch": 0.47240618101545256, + "grad_norm": 0.19831134378910065, + "learning_rate": 9.899971755337049e-06, + "loss": 0.4045, + "step": 214 + }, + { + "epoch": 0.4746136865342163, + "grad_norm": 0.18536338210105896, + "learning_rate": 9.897399351330471e-06, + "loss": 0.4134, + "step": 215 + }, + { + "epoch": 0.4768211920529801, + "grad_norm": 0.17010532319545746, + "learning_rate": 9.894794631732223e-06, + "loss": 0.4076, + "step": 216 + }, + { + "epoch": 0.47902869757174393, + "grad_norm": 0.17602422833442688, + "learning_rate": 9.8921576137296e-06, + "loss": 0.4079, + "step": 217 + }, + { + "epoch": 0.48123620309050774, + "grad_norm": 0.1983213573694229, + "learning_rate": 9.889488314723024e-06, + "loss": 0.413, + "step": 218 + }, + { + "epoch": 0.48344370860927155, + "grad_norm": 0.17187613248825073, + "learning_rate": 9.886786752325917e-06, + "loss": 0.4229, + "step": 219 + }, + { + "epoch": 0.4856512141280353, + "grad_norm": 0.2029353827238083, + "learning_rate": 9.884052944364595e-06, + "loss": 0.4242, + "step": 220 + }, + { + "epoch": 0.4878587196467991, + "grad_norm": 0.18107262253761292, + "learning_rate": 9.881286908878148e-06, + "loss": 0.4157, + "step": 221 + }, + { + "epoch": 0.4900662251655629, + "grad_norm": 0.18853497505187988, + "learning_rate": 9.878488664118316e-06, + "loss": 0.4153, + "step": 222 + }, + { + "epoch": 0.4922737306843267, + "grad_norm": 0.1775682419538498, + "learning_rate": 9.875658228549379e-06, + "loss": 0.4109, + "step": 223 + }, + { + "epoch": 0.49448123620309054, + "grad_norm": 0.18237102031707764, + "learning_rate": 9.872795620848024e-06, + "loss": 0.407, + "step": 224 + }, + { + "epoch": 0.4966887417218543, + "grad_norm": 0.20489446818828583, + "learning_rate": 9.869900859903225e-06, + "loss": 0.4198, + "step": 225 + }, + { + "epoch": 0.4988962472406181, + "grad_norm": 0.1481681913137436, + "learning_rate": 9.866973964816126e-06, + "loss": 0.4085, + "step": 226 + }, + { + "epoch": 0.5011037527593819, + "grad_norm": 0.18315470218658447, + "learning_rate": 9.864014954899905e-06, + "loss": 0.4079, + "step": 227 + }, + { + "epoch": 0.5033112582781457, + "grad_norm": 0.18866921961307526, + "learning_rate": 9.861023849679648e-06, + "loss": 0.4249, + "step": 228 + }, + { + "epoch": 0.5055187637969095, + "grad_norm": 0.15121376514434814, + "learning_rate": 9.858000668892226e-06, + "loss": 0.4179, + "step": 229 + }, + { + "epoch": 0.5077262693156733, + "grad_norm": 0.19818085432052612, + "learning_rate": 9.85494543248616e-06, + "loss": 0.4218, + "step": 230 + }, + { + "epoch": 0.5099337748344371, + "grad_norm": 0.15964657068252563, + "learning_rate": 9.851858160621496e-06, + "loss": 0.4074, + "step": 231 + }, + { + "epoch": 0.5121412803532008, + "grad_norm": 0.18371723592281342, + "learning_rate": 9.848738873669653e-06, + "loss": 0.4103, + "step": 232 + }, + { + "epoch": 0.5143487858719646, + "grad_norm": 0.1854197084903717, + "learning_rate": 9.845587592213318e-06, + "loss": 0.4092, + "step": 233 + }, + { + "epoch": 0.5165562913907285, + "grad_norm": 0.15952259302139282, + "learning_rate": 9.842404337046284e-06, + "loss": 0.4139, + "step": 234 + }, + { + "epoch": 0.5187637969094923, + "grad_norm": 0.1830589473247528, + "learning_rate": 9.839189129173328e-06, + "loss": 0.4143, + "step": 235 + }, + { + "epoch": 0.5209713024282561, + "grad_norm": 0.1684870719909668, + "learning_rate": 9.835941989810065e-06, + "loss": 0.4088, + "step": 236 + }, + { + "epoch": 0.5231788079470199, + "grad_norm": 0.17676231265068054, + "learning_rate": 9.832662940382813e-06, + "loss": 0.4123, + "step": 237 + }, + { + "epoch": 0.5253863134657837, + "grad_norm": 0.17995555698871613, + "learning_rate": 9.829352002528449e-06, + "loss": 0.4086, + "step": 238 + }, + { + "epoch": 0.5275938189845475, + "grad_norm": 0.1694101095199585, + "learning_rate": 9.826009198094262e-06, + "loss": 0.4137, + "step": 239 + }, + { + "epoch": 0.5298013245033113, + "grad_norm": 0.16819199919700623, + "learning_rate": 9.822634549137819e-06, + "loss": 0.4072, + "step": 240 + }, + { + "epoch": 0.5320088300220751, + "grad_norm": 0.17408284544944763, + "learning_rate": 9.81922807792681e-06, + "loss": 0.4052, + "step": 241 + }, + { + "epoch": 0.5342163355408388, + "grad_norm": 0.1843319982290268, + "learning_rate": 9.815789806938909e-06, + "loss": 0.3966, + "step": 242 + }, + { + "epoch": 0.5364238410596026, + "grad_norm": 0.18125468492507935, + "learning_rate": 9.812319758861616e-06, + "loss": 0.4044, + "step": 243 + }, + { + "epoch": 0.5386313465783664, + "grad_norm": 0.16861899197101593, + "learning_rate": 9.808817956592115e-06, + "loss": 0.4092, + "step": 244 + }, + { + "epoch": 0.5408388520971302, + "grad_norm": 0.22598163783550262, + "learning_rate": 9.805284423237126e-06, + "loss": 0.4137, + "step": 245 + }, + { + "epoch": 0.543046357615894, + "grad_norm": 0.22511421144008636, + "learning_rate": 9.801719182112738e-06, + "loss": 0.4073, + "step": 246 + }, + { + "epoch": 0.5452538631346578, + "grad_norm": 0.1690152883529663, + "learning_rate": 9.798122256744269e-06, + "loss": 0.4097, + "step": 247 + }, + { + "epoch": 0.5474613686534217, + "grad_norm": 0.22893239557743073, + "learning_rate": 9.794493670866108e-06, + "loss": 0.4054, + "step": 248 + }, + { + "epoch": 0.5496688741721855, + "grad_norm": 0.17257164418697357, + "learning_rate": 9.790833448421554e-06, + "loss": 0.4109, + "step": 249 + }, + { + "epoch": 0.5518763796909493, + "grad_norm": 0.1636303812265396, + "learning_rate": 9.787141613562661e-06, + "loss": 0.3995, + "step": 250 + }, + { + "epoch": 0.5540838852097131, + "grad_norm": 0.17622527480125427, + "learning_rate": 9.783418190650079e-06, + "loss": 0.409, + "step": 251 + }, + { + "epoch": 0.5562913907284768, + "grad_norm": 0.20090465247631073, + "learning_rate": 9.779663204252887e-06, + "loss": 0.4015, + "step": 252 + }, + { + "epoch": 0.5584988962472406, + "grad_norm": 0.15368807315826416, + "learning_rate": 9.775876679148449e-06, + "loss": 0.4009, + "step": 253 + }, + { + "epoch": 0.5607064017660044, + "grad_norm": 0.2135976403951645, + "learning_rate": 9.772058640322221e-06, + "loss": 0.407, + "step": 254 + }, + { + "epoch": 0.5629139072847682, + "grad_norm": 0.17839573323726654, + "learning_rate": 9.768209112967619e-06, + "loss": 0.404, + "step": 255 + }, + { + "epoch": 0.565121412803532, + "grad_norm": 0.1953095942735672, + "learning_rate": 9.764328122485827e-06, + "loss": 0.402, + "step": 256 + }, + { + "epoch": 0.5673289183222958, + "grad_norm": 0.16529282927513123, + "learning_rate": 9.76041569448564e-06, + "loss": 0.4117, + "step": 257 + }, + { + "epoch": 0.5695364238410596, + "grad_norm": 0.17458973824977875, + "learning_rate": 9.756471854783297e-06, + "loss": 0.4144, + "step": 258 + }, + { + "epoch": 0.5717439293598234, + "grad_norm": 0.1776653230190277, + "learning_rate": 9.752496629402307e-06, + "loss": 0.4123, + "step": 259 + }, + { + "epoch": 0.5739514348785872, + "grad_norm": 0.17428044974803925, + "learning_rate": 9.748490044573275e-06, + "loss": 0.3975, + "step": 260 + }, + { + "epoch": 0.5761589403973509, + "grad_norm": 0.17150405049324036, + "learning_rate": 9.744452126733739e-06, + "loss": 0.4023, + "step": 261 + }, + { + "epoch": 0.5783664459161147, + "grad_norm": 0.1668493002653122, + "learning_rate": 9.740382902527981e-06, + "loss": 0.409, + "step": 262 + }, + { + "epoch": 0.5805739514348786, + "grad_norm": 0.17727597057819366, + "learning_rate": 9.736282398806862e-06, + "loss": 0.4061, + "step": 263 + }, + { + "epoch": 0.5827814569536424, + "grad_norm": 0.15944141149520874, + "learning_rate": 9.73215064262764e-06, + "loss": 0.402, + "step": 264 + }, + { + "epoch": 0.5849889624724062, + "grad_norm": 0.17232947051525116, + "learning_rate": 9.727987661253796e-06, + "loss": 0.4102, + "step": 265 + }, + { + "epoch": 0.58719646799117, + "grad_norm": 0.2054961770772934, + "learning_rate": 9.72379348215485e-06, + "loss": 0.3994, + "step": 266 + }, + { + "epoch": 0.5894039735099338, + "grad_norm": 0.1817290335893631, + "learning_rate": 9.719568133006177e-06, + "loss": 0.4033, + "step": 267 + }, + { + "epoch": 0.5916114790286976, + "grad_norm": 0.16904965043067932, + "learning_rate": 9.715311641688835e-06, + "loss": 0.4082, + "step": 268 + }, + { + "epoch": 0.5938189845474614, + "grad_norm": 0.17628441751003265, + "learning_rate": 9.71102403628937e-06, + "loss": 0.3969, + "step": 269 + }, + { + "epoch": 0.5960264900662252, + "grad_norm": 0.20128114521503448, + "learning_rate": 9.706705345099632e-06, + "loss": 0.4081, + "step": 270 + }, + { + "epoch": 0.5982339955849889, + "grad_norm": 0.17412817478179932, + "learning_rate": 9.7023555966166e-06, + "loss": 0.4076, + "step": 271 + }, + { + "epoch": 0.6004415011037527, + "grad_norm": 0.18813695013523102, + "learning_rate": 9.697974819542178e-06, + "loss": 0.4042, + "step": 272 + }, + { + "epoch": 0.6026490066225165, + "grad_norm": 0.17176660895347595, + "learning_rate": 9.693563042783011e-06, + "loss": 0.4099, + "step": 273 + }, + { + "epoch": 0.6048565121412803, + "grad_norm": 0.17056933045387268, + "learning_rate": 9.689120295450308e-06, + "loss": 0.4045, + "step": 274 + }, + { + "epoch": 0.6070640176600441, + "grad_norm": 0.20970946550369263, + "learning_rate": 9.684646606859621e-06, + "loss": 0.3944, + "step": 275 + }, + { + "epoch": 0.609271523178808, + "grad_norm": 0.19319604337215424, + "learning_rate": 9.680142006530684e-06, + "loss": 0.4062, + "step": 276 + }, + { + "epoch": 0.6114790286975718, + "grad_norm": 0.17683015763759613, + "learning_rate": 9.675606524187192e-06, + "loss": 0.4077, + "step": 277 + }, + { + "epoch": 0.6136865342163356, + "grad_norm": 0.16389036178588867, + "learning_rate": 9.671040189756623e-06, + "loss": 0.412, + "step": 278 + }, + { + "epoch": 0.6158940397350994, + "grad_norm": 0.15970537066459656, + "learning_rate": 9.666443033370026e-06, + "loss": 0.4068, + "step": 279 + }, + { + "epoch": 0.6181015452538632, + "grad_norm": 0.17420579493045807, + "learning_rate": 9.661815085361836e-06, + "loss": 0.4046, + "step": 280 + }, + { + "epoch": 0.6203090507726269, + "grad_norm": 0.16548947989940643, + "learning_rate": 9.657156376269665e-06, + "loss": 0.4122, + "step": 281 + }, + { + "epoch": 0.6225165562913907, + "grad_norm": 0.147056445479393, + "learning_rate": 9.652466936834101e-06, + "loss": 0.4085, + "step": 282 + }, + { + "epoch": 0.6247240618101545, + "grad_norm": 0.17280390858650208, + "learning_rate": 9.647746797998508e-06, + "loss": 0.4142, + "step": 283 + }, + { + "epoch": 0.6269315673289183, + "grad_norm": 0.16386403143405914, + "learning_rate": 9.642995990908817e-06, + "loss": 0.4103, + "step": 284 + }, + { + "epoch": 0.6291390728476821, + "grad_norm": 0.16252438724040985, + "learning_rate": 9.638214546913333e-06, + "loss": 0.4023, + "step": 285 + }, + { + "epoch": 0.6313465783664459, + "grad_norm": 0.16047868132591248, + "learning_rate": 9.633402497562512e-06, + "loss": 0.4032, + "step": 286 + }, + { + "epoch": 0.6335540838852097, + "grad_norm": 0.18797720968723297, + "learning_rate": 9.628559874608761e-06, + "loss": 0.4014, + "step": 287 + }, + { + "epoch": 0.6357615894039735, + "grad_norm": 0.15865157544612885, + "learning_rate": 9.62368671000623e-06, + "loss": 0.4022, + "step": 288 + }, + { + "epoch": 0.6379690949227373, + "grad_norm": 0.15757879614830017, + "learning_rate": 9.618783035910596e-06, + "loss": 0.4019, + "step": 289 + }, + { + "epoch": 0.6401766004415012, + "grad_norm": 0.15986581146717072, + "learning_rate": 9.613848884678851e-06, + "loss": 0.3981, + "step": 290 + }, + { + "epoch": 0.6423841059602649, + "grad_norm": 0.1824173629283905, + "learning_rate": 9.608884288869103e-06, + "loss": 0.407, + "step": 291 + }, + { + "epoch": 0.6445916114790287, + "grad_norm": 0.185842826962471, + "learning_rate": 9.603889281240334e-06, + "loss": 0.4016, + "step": 292 + }, + { + "epoch": 0.6467991169977925, + "grad_norm": 0.17280805110931396, + "learning_rate": 9.59886389475221e-06, + "loss": 0.4081, + "step": 293 + }, + { + "epoch": 0.6490066225165563, + "grad_norm": 0.1815565824508667, + "learning_rate": 9.593808162564845e-06, + "loss": 0.4141, + "step": 294 + }, + { + "epoch": 0.6512141280353201, + "grad_norm": 0.19602055847644806, + "learning_rate": 9.588722118038595e-06, + "loss": 0.4048, + "step": 295 + }, + { + "epoch": 0.6534216335540839, + "grad_norm": 0.1916995495557785, + "learning_rate": 9.583605794733833e-06, + "loss": 0.3953, + "step": 296 + }, + { + "epoch": 0.6556291390728477, + "grad_norm": 0.17578016221523285, + "learning_rate": 9.578459226410722e-06, + "loss": 0.4133, + "step": 297 + }, + { + "epoch": 0.6578366445916115, + "grad_norm": 0.18694248795509338, + "learning_rate": 9.573282447029e-06, + "loss": 0.4023, + "step": 298 + }, + { + "epoch": 0.6600441501103753, + "grad_norm": 0.20029006898403168, + "learning_rate": 9.568075490747756e-06, + "loss": 0.3967, + "step": 299 + }, + { + "epoch": 0.6622516556291391, + "grad_norm": 0.15852802991867065, + "learning_rate": 9.562838391925197e-06, + "loss": 0.4053, + "step": 300 + }, + { + "epoch": 0.6644591611479028, + "grad_norm": 0.18394002318382263, + "learning_rate": 9.557571185118431e-06, + "loss": 0.4001, + "step": 301 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 0.1864984780550003, + "learning_rate": 9.55227390508323e-06, + "loss": 0.3964, + "step": 302 + }, + { + "epoch": 0.6688741721854304, + "grad_norm": 0.15841950476169586, + "learning_rate": 9.546946586773808e-06, + "loss": 0.4045, + "step": 303 + }, + { + "epoch": 0.6710816777041942, + "grad_norm": 0.19293592870235443, + "learning_rate": 9.541589265342585e-06, + "loss": 0.405, + "step": 304 + }, + { + "epoch": 0.673289183222958, + "grad_norm": 0.16064338386058807, + "learning_rate": 9.536201976139958e-06, + "loss": 0.4098, + "step": 305 + }, + { + "epoch": 0.6754966887417219, + "grad_norm": 0.16002054512500763, + "learning_rate": 9.530784754714069e-06, + "loss": 0.3837, + "step": 306 + }, + { + "epoch": 0.6777041942604857, + "grad_norm": 0.16032235324382782, + "learning_rate": 9.525337636810564e-06, + "loss": 0.4095, + "step": 307 + }, + { + "epoch": 0.6799116997792495, + "grad_norm": 0.19734519720077515, + "learning_rate": 9.519860658372364e-06, + "loss": 0.4049, + "step": 308 + }, + { + "epoch": 0.6821192052980133, + "grad_norm": 0.1894775629043579, + "learning_rate": 9.514353855539428e-06, + "loss": 0.3926, + "step": 309 + }, + { + "epoch": 0.6843267108167771, + "grad_norm": 0.2021895796060562, + "learning_rate": 9.508817264648506e-06, + "loss": 0.3983, + "step": 310 + }, + { + "epoch": 0.6865342163355408, + "grad_norm": 0.20270881056785583, + "learning_rate": 9.503250922232911e-06, + "loss": 0.4105, + "step": 311 + }, + { + "epoch": 0.6887417218543046, + "grad_norm": 0.17310801148414612, + "learning_rate": 9.497654865022268e-06, + "loss": 0.4026, + "step": 312 + }, + { + "epoch": 0.6909492273730684, + "grad_norm": 0.22883597016334534, + "learning_rate": 9.492029129942277e-06, + "loss": 0.4062, + "step": 313 + }, + { + "epoch": 0.6931567328918322, + "grad_norm": 0.1955188363790512, + "learning_rate": 9.48637375411447e-06, + "loss": 0.4044, + "step": 314 + }, + { + "epoch": 0.695364238410596, + "grad_norm": 0.18373191356658936, + "learning_rate": 9.48068877485596e-06, + "loss": 0.4032, + "step": 315 + }, + { + "epoch": 0.6975717439293598, + "grad_norm": 0.21621759235858917, + "learning_rate": 9.474974229679201e-06, + "loss": 0.3904, + "step": 316 + }, + { + "epoch": 0.6997792494481236, + "grad_norm": 0.17569488286972046, + "learning_rate": 9.469230156291742e-06, + "loss": 0.4087, + "step": 317 + }, + { + "epoch": 0.7019867549668874, + "grad_norm": 0.2064937800168991, + "learning_rate": 9.463456592595966e-06, + "loss": 0.396, + "step": 318 + }, + { + "epoch": 0.7041942604856513, + "grad_norm": 0.24051177501678467, + "learning_rate": 9.457653576688857e-06, + "loss": 0.4062, + "step": 319 + }, + { + "epoch": 0.7064017660044151, + "grad_norm": 0.16446927189826965, + "learning_rate": 9.451821146861734e-06, + "loss": 0.3987, + "step": 320 + }, + { + "epoch": 0.7086092715231788, + "grad_norm": 0.20139139890670776, + "learning_rate": 9.445959341600009e-06, + "loss": 0.4036, + "step": 321 + }, + { + "epoch": 0.7108167770419426, + "grad_norm": 0.18157215416431427, + "learning_rate": 9.440068199582923e-06, + "loss": 0.4124, + "step": 322 + }, + { + "epoch": 0.7130242825607064, + "grad_norm": 0.15641391277313232, + "learning_rate": 9.434147759683303e-06, + "loss": 0.405, + "step": 323 + }, + { + "epoch": 0.7152317880794702, + "grad_norm": 0.16391552984714508, + "learning_rate": 9.428198060967294e-06, + "loss": 0.4005, + "step": 324 + }, + { + "epoch": 0.717439293598234, + "grad_norm": 0.179102823138237, + "learning_rate": 9.422219142694104e-06, + "loss": 0.3968, + "step": 325 + }, + { + "epoch": 0.7196467991169978, + "grad_norm": 0.17109854519367218, + "learning_rate": 9.416211044315754e-06, + "loss": 0.4049, + "step": 326 + }, + { + "epoch": 0.7218543046357616, + "grad_norm": 0.20250354707241058, + "learning_rate": 9.410173805476804e-06, + "loss": 0.4186, + "step": 327 + }, + { + "epoch": 0.7240618101545254, + "grad_norm": 0.18576852977275848, + "learning_rate": 9.404107466014101e-06, + "loss": 0.4022, + "step": 328 + }, + { + "epoch": 0.7262693156732892, + "grad_norm": 0.16930828988552094, + "learning_rate": 9.398012065956512e-06, + "loss": 0.3949, + "step": 329 + }, + { + "epoch": 0.7284768211920529, + "grad_norm": 0.1966543048620224, + "learning_rate": 9.39188764552466e-06, + "loss": 0.4103, + "step": 330 + }, + { + "epoch": 0.7306843267108167, + "grad_norm": 0.18573778867721558, + "learning_rate": 9.385734245130664e-06, + "loss": 0.4069, + "step": 331 + }, + { + "epoch": 0.7328918322295805, + "grad_norm": 0.16225308179855347, + "learning_rate": 9.379551905377863e-06, + "loss": 0.4049, + "step": 332 + }, + { + "epoch": 0.7350993377483444, + "grad_norm": 0.1815934032201767, + "learning_rate": 9.373340667060553e-06, + "loss": 0.3927, + "step": 333 + }, + { + "epoch": 0.7373068432671082, + "grad_norm": 0.1688205897808075, + "learning_rate": 9.367100571163722e-06, + "loss": 0.4019, + "step": 334 + }, + { + "epoch": 0.739514348785872, + "grad_norm": 0.18315370380878448, + "learning_rate": 9.360831658862774e-06, + "loss": 0.3989, + "step": 335 + }, + { + "epoch": 0.7417218543046358, + "grad_norm": 0.17102378606796265, + "learning_rate": 9.354533971523253e-06, + "loss": 0.4122, + "step": 336 + }, + { + "epoch": 0.7439293598233996, + "grad_norm": 0.17149822413921356, + "learning_rate": 9.348207550700584e-06, + "loss": 0.4087, + "step": 337 + }, + { + "epoch": 0.7461368653421634, + "grad_norm": 0.19260001182556152, + "learning_rate": 9.341852438139784e-06, + "loss": 0.4064, + "step": 338 + }, + { + "epoch": 0.7483443708609272, + "grad_norm": 0.18973691761493683, + "learning_rate": 9.335468675775196e-06, + "loss": 0.399, + "step": 339 + }, + { + "epoch": 0.7505518763796909, + "grad_norm": 0.19365577399730682, + "learning_rate": 9.329056305730211e-06, + "loss": 0.3959, + "step": 340 + }, + { + "epoch": 0.7527593818984547, + "grad_norm": 0.21033021807670593, + "learning_rate": 9.322615370316986e-06, + "loss": 0.4027, + "step": 341 + }, + { + "epoch": 0.7549668874172185, + "grad_norm": 0.1825421005487442, + "learning_rate": 9.316145912036165e-06, + "loss": 0.3971, + "step": 342 + }, + { + "epoch": 0.7571743929359823, + "grad_norm": 0.1825910061597824, + "learning_rate": 9.309647973576605e-06, + "loss": 0.3959, + "step": 343 + }, + { + "epoch": 0.7593818984547461, + "grad_norm": 0.1806352436542511, + "learning_rate": 9.30312159781509e-06, + "loss": 0.4047, + "step": 344 + }, + { + "epoch": 0.7615894039735099, + "grad_norm": 0.15455348789691925, + "learning_rate": 9.296566827816044e-06, + "loss": 0.3958, + "step": 345 + }, + { + "epoch": 0.7637969094922737, + "grad_norm": 0.19140276312828064, + "learning_rate": 9.289983706831254e-06, + "loss": 0.3955, + "step": 346 + }, + { + "epoch": 0.7660044150110376, + "grad_norm": 0.1777879148721695, + "learning_rate": 9.28337227829958e-06, + "loss": 0.3971, + "step": 347 + }, + { + "epoch": 0.7682119205298014, + "grad_norm": 0.16399559378623962, + "learning_rate": 9.276732585846673e-06, + "loss": 0.3998, + "step": 348 + }, + { + "epoch": 0.7704194260485652, + "grad_norm": 0.1594182848930359, + "learning_rate": 9.270064673284681e-06, + "loss": 0.3898, + "step": 349 + }, + { + "epoch": 0.7726269315673289, + "grad_norm": 0.16792874038219452, + "learning_rate": 9.263368584611965e-06, + "loss": 0.4021, + "step": 350 + }, + { + "epoch": 0.7748344370860927, + "grad_norm": 0.18502795696258545, + "learning_rate": 9.256644364012803e-06, + "loss": 0.3987, + "step": 351 + }, + { + "epoch": 0.7770419426048565, + "grad_norm": 0.18240559101104736, + "learning_rate": 9.249892055857107e-06, + "loss": 0.4074, + "step": 352 + }, + { + "epoch": 0.7792494481236203, + "grad_norm": 0.15263855457305908, + "learning_rate": 9.243111704700126e-06, + "loss": 0.3928, + "step": 353 + }, + { + "epoch": 0.7814569536423841, + "grad_norm": 0.18149858713150024, + "learning_rate": 9.236303355282142e-06, + "loss": 0.4015, + "step": 354 + }, + { + "epoch": 0.7836644591611479, + "grad_norm": 0.1617615818977356, + "learning_rate": 9.229467052528191e-06, + "loss": 0.4024, + "step": 355 + }, + { + "epoch": 0.7858719646799117, + "grad_norm": 0.19588352739810944, + "learning_rate": 9.222602841547766e-06, + "loss": 0.3946, + "step": 356 + }, + { + "epoch": 0.7880794701986755, + "grad_norm": 0.1633402556180954, + "learning_rate": 9.2157107676345e-06, + "loss": 0.4027, + "step": 357 + }, + { + "epoch": 0.7902869757174393, + "grad_norm": 0.16059032082557678, + "learning_rate": 9.208790876265887e-06, + "loss": 0.3981, + "step": 358 + }, + { + "epoch": 0.7924944812362031, + "grad_norm": 0.1558162271976471, + "learning_rate": 9.201843213102976e-06, + "loss": 0.3969, + "step": 359 + }, + { + "epoch": 0.7947019867549668, + "grad_norm": 0.1607562005519867, + "learning_rate": 9.194867823990069e-06, + "loss": 0.3972, + "step": 360 + }, + { + "epoch": 0.7969094922737306, + "grad_norm": 0.1428382843732834, + "learning_rate": 9.187864754954412e-06, + "loss": 0.3928, + "step": 361 + }, + { + "epoch": 0.7991169977924945, + "grad_norm": 0.15147483348846436, + "learning_rate": 9.180834052205903e-06, + "loss": 0.392, + "step": 362 + }, + { + "epoch": 0.8013245033112583, + "grad_norm": 0.14845693111419678, + "learning_rate": 9.173775762136783e-06, + "loss": 0.3989, + "step": 363 + }, + { + "epoch": 0.8035320088300221, + "grad_norm": 0.1665990948677063, + "learning_rate": 9.166689931321326e-06, + "loss": 0.3928, + "step": 364 + }, + { + "epoch": 0.8057395143487859, + "grad_norm": 0.14729240536689758, + "learning_rate": 9.159576606515532e-06, + "loss": 0.3953, + "step": 365 + }, + { + "epoch": 0.8079470198675497, + "grad_norm": 0.14690490067005157, + "learning_rate": 9.152435834656823e-06, + "loss": 0.4023, + "step": 366 + }, + { + "epoch": 0.8101545253863135, + "grad_norm": 0.15939104557037354, + "learning_rate": 9.145267662863732e-06, + "loss": 0.4013, + "step": 367 + }, + { + "epoch": 0.8123620309050773, + "grad_norm": 0.1510556936264038, + "learning_rate": 9.13807213843559e-06, + "loss": 0.3992, + "step": 368 + }, + { + "epoch": 0.8145695364238411, + "grad_norm": 0.1646500676870346, + "learning_rate": 9.130849308852217e-06, + "loss": 0.3925, + "step": 369 + }, + { + "epoch": 0.8167770419426048, + "grad_norm": 0.17254091799259186, + "learning_rate": 9.123599221773601e-06, + "loss": 0.4071, + "step": 370 + }, + { + "epoch": 0.8189845474613686, + "grad_norm": 0.1497507095336914, + "learning_rate": 9.116321925039591e-06, + "loss": 0.3883, + "step": 371 + }, + { + "epoch": 0.8211920529801324, + "grad_norm": 0.16002339124679565, + "learning_rate": 9.109017466669587e-06, + "loss": 0.3953, + "step": 372 + }, + { + "epoch": 0.8233995584988962, + "grad_norm": 0.14812716841697693, + "learning_rate": 9.101685894862206e-06, + "loss": 0.4021, + "step": 373 + }, + { + "epoch": 0.82560706401766, + "grad_norm": 0.15898163616657257, + "learning_rate": 9.094327257994978e-06, + "loss": 0.4102, + "step": 374 + }, + { + "epoch": 0.8278145695364238, + "grad_norm": 0.15488837659358978, + "learning_rate": 9.086941604624022e-06, + "loss": 0.3912, + "step": 375 + }, + { + "epoch": 0.8300220750551877, + "grad_norm": 0.1613994836807251, + "learning_rate": 9.079528983483726e-06, + "loss": 0.4029, + "step": 376 + }, + { + "epoch": 0.8322295805739515, + "grad_norm": 0.1536788046360016, + "learning_rate": 9.072089443486425e-06, + "loss": 0.3956, + "step": 377 + }, + { + "epoch": 0.8344370860927153, + "grad_norm": 0.1693611443042755, + "learning_rate": 9.064623033722077e-06, + "loss": 0.3984, + "step": 378 + }, + { + "epoch": 0.8366445916114791, + "grad_norm": 0.18005254864692688, + "learning_rate": 9.057129803457943e-06, + "loss": 0.4022, + "step": 379 + }, + { + "epoch": 0.8388520971302428, + "grad_norm": 0.15852075815200806, + "learning_rate": 9.049609802138262e-06, + "loss": 0.3816, + "step": 380 + }, + { + "epoch": 0.8410596026490066, + "grad_norm": 0.1933150738477707, + "learning_rate": 9.042063079383916e-06, + "loss": 0.4028, + "step": 381 + }, + { + "epoch": 0.8432671081677704, + "grad_norm": 0.18579082190990448, + "learning_rate": 9.034489684992112e-06, + "loss": 0.4057, + "step": 382 + }, + { + "epoch": 0.8454746136865342, + "grad_norm": 0.1803402453660965, + "learning_rate": 9.026889668936054e-06, + "loss": 0.3976, + "step": 383 + }, + { + "epoch": 0.847682119205298, + "grad_norm": 0.17872066795825958, + "learning_rate": 9.019263081364605e-06, + "loss": 0.3908, + "step": 384 + }, + { + "epoch": 0.8498896247240618, + "grad_norm": 0.18206505477428436, + "learning_rate": 9.01160997260196e-06, + "loss": 0.4072, + "step": 385 + }, + { + "epoch": 0.8520971302428256, + "grad_norm": 0.1718129813671112, + "learning_rate": 9.00393039314732e-06, + "loss": 0.4036, + "step": 386 + }, + { + "epoch": 0.8543046357615894, + "grad_norm": 0.21254006028175354, + "learning_rate": 8.996224393674545e-06, + "loss": 0.4097, + "step": 387 + }, + { + "epoch": 0.8565121412803532, + "grad_norm": 0.15457090735435486, + "learning_rate": 8.988492025031838e-06, + "loss": 0.4001, + "step": 388 + }, + { + "epoch": 0.8587196467991169, + "grad_norm": 0.18473166227340698, + "learning_rate": 8.980733338241395e-06, + "loss": 0.3965, + "step": 389 + }, + { + "epoch": 0.8609271523178808, + "grad_norm": 0.19067919254302979, + "learning_rate": 8.972948384499068e-06, + "loss": 0.3996, + "step": 390 + }, + { + "epoch": 0.8631346578366446, + "grad_norm": 0.199691504240036, + "learning_rate": 8.965137215174037e-06, + "loss": 0.4003, + "step": 391 + }, + { + "epoch": 0.8653421633554084, + "grad_norm": 0.22586044669151306, + "learning_rate": 8.957299881808471e-06, + "loss": 0.393, + "step": 392 + }, + { + "epoch": 0.8675496688741722, + "grad_norm": 0.1807902604341507, + "learning_rate": 8.949436436117172e-06, + "loss": 0.4039, + "step": 393 + }, + { + "epoch": 0.869757174392936, + "grad_norm": 0.22098712623119354, + "learning_rate": 8.941546929987253e-06, + "loss": 0.39, + "step": 394 + }, + { + "epoch": 0.8719646799116998, + "grad_norm": 0.19912059605121613, + "learning_rate": 8.933631415477785e-06, + "loss": 0.396, + "step": 395 + }, + { + "epoch": 0.8741721854304636, + "grad_norm": 0.16147476434707642, + "learning_rate": 8.925689944819452e-06, + "loss": 0.4011, + "step": 396 + }, + { + "epoch": 0.8763796909492274, + "grad_norm": 0.19136802852153778, + "learning_rate": 8.917722570414217e-06, + "loss": 0.3896, + "step": 397 + }, + { + "epoch": 0.8785871964679912, + "grad_norm": 0.20598876476287842, + "learning_rate": 8.909729344834965e-06, + "loss": 0.3995, + "step": 398 + }, + { + "epoch": 0.8807947019867549, + "grad_norm": 0.1776047945022583, + "learning_rate": 8.901710320825161e-06, + "loss": 0.4001, + "step": 399 + }, + { + "epoch": 0.8830022075055187, + "grad_norm": 0.18616576492786407, + "learning_rate": 8.893665551298502e-06, + "loss": 0.3995, + "step": 400 + }, + { + "epoch": 0.8852097130242825, + "grad_norm": 0.18072794377803802, + "learning_rate": 8.885595089338567e-06, + "loss": 0.3978, + "step": 401 + }, + { + "epoch": 0.8874172185430463, + "grad_norm": 0.181128591299057, + "learning_rate": 8.877498988198471e-06, + "loss": 0.4112, + "step": 402 + }, + { + "epoch": 0.8896247240618101, + "grad_norm": 0.17399437725543976, + "learning_rate": 8.869377301300501e-06, + "loss": 0.3937, + "step": 403 + }, + { + "epoch": 0.891832229580574, + "grad_norm": 0.17823876440525055, + "learning_rate": 8.86123008223578e-06, + "loss": 0.4055, + "step": 404 + }, + { + "epoch": 0.8940397350993378, + "grad_norm": 0.176737442612648, + "learning_rate": 8.853057384763904e-06, + "loss": 0.3991, + "step": 405 + }, + { + "epoch": 0.8962472406181016, + "grad_norm": 0.1630028337240219, + "learning_rate": 8.844859262812584e-06, + "loss": 0.3903, + "step": 406 + }, + { + "epoch": 0.8984547461368654, + "grad_norm": 0.18285702168941498, + "learning_rate": 8.8366357704773e-06, + "loss": 0.3904, + "step": 407 + }, + { + "epoch": 0.9006622516556292, + "grad_norm": 0.1716790795326233, + "learning_rate": 8.82838696202094e-06, + "loss": 0.4033, + "step": 408 + }, + { + "epoch": 0.9028697571743929, + "grad_norm": 0.15402917563915253, + "learning_rate": 8.820112891873433e-06, + "loss": 0.39, + "step": 409 + }, + { + "epoch": 0.9050772626931567, + "grad_norm": 0.16390980780124664, + "learning_rate": 8.811813614631411e-06, + "loss": 0.3993, + "step": 410 + }, + { + "epoch": 0.9072847682119205, + "grad_norm": 0.15821807086467743, + "learning_rate": 8.803489185057822e-06, + "loss": 0.3953, + "step": 411 + }, + { + "epoch": 0.9094922737306843, + "grad_norm": 0.17032112181186676, + "learning_rate": 8.795139658081586e-06, + "loss": 0.4035, + "step": 412 + }, + { + "epoch": 0.9116997792494481, + "grad_norm": 0.15845684707164764, + "learning_rate": 8.786765088797238e-06, + "loss": 0.4013, + "step": 413 + }, + { + "epoch": 0.9139072847682119, + "grad_norm": 0.17293021082878113, + "learning_rate": 8.778365532464543e-06, + "loss": 0.3965, + "step": 414 + }, + { + "epoch": 0.9161147902869757, + "grad_norm": 0.14891566336154938, + "learning_rate": 8.76994104450815e-06, + "loss": 0.392, + "step": 415 + }, + { + "epoch": 0.9183222958057395, + "grad_norm": 0.15123558044433594, + "learning_rate": 8.761491680517218e-06, + "loss": 0.396, + "step": 416 + }, + { + "epoch": 0.9205298013245033, + "grad_norm": 0.1682642549276352, + "learning_rate": 8.75301749624505e-06, + "loss": 0.3838, + "step": 417 + }, + { + "epoch": 0.9227373068432672, + "grad_norm": 0.1535079926252365, + "learning_rate": 8.744518547608732e-06, + "loss": 0.3921, + "step": 418 + }, + { + "epoch": 0.9249448123620309, + "grad_norm": 0.16250640153884888, + "learning_rate": 8.735994890688749e-06, + "loss": 0.3896, + "step": 419 + }, + { + "epoch": 0.9271523178807947, + "grad_norm": 0.16567584872245789, + "learning_rate": 8.72744658172863e-06, + "loss": 0.3928, + "step": 420 + }, + { + "epoch": 0.9293598233995585, + "grad_norm": 0.16492141783237457, + "learning_rate": 8.718873677134569e-06, + "loss": 0.405, + "step": 421 + }, + { + "epoch": 0.9315673289183223, + "grad_norm": 0.14573417603969574, + "learning_rate": 8.710276233475058e-06, + "loss": 0.3955, + "step": 422 + }, + { + "epoch": 0.9337748344370861, + "grad_norm": 0.15780314803123474, + "learning_rate": 8.701654307480508e-06, + "loss": 0.3868, + "step": 423 + }, + { + "epoch": 0.9359823399558499, + "grad_norm": 0.14735905826091766, + "learning_rate": 8.693007956042874e-06, + "loss": 0.3924, + "step": 424 + }, + { + "epoch": 0.9381898454746137, + "grad_norm": 0.16305825114250183, + "learning_rate": 8.684337236215289e-06, + "loss": 0.3866, + "step": 425 + }, + { + "epoch": 0.9403973509933775, + "grad_norm": 0.1656455248594284, + "learning_rate": 8.675642205211679e-06, + "loss": 0.3965, + "step": 426 + }, + { + "epoch": 0.9426048565121413, + "grad_norm": 0.15034431219100952, + "learning_rate": 8.666922920406384e-06, + "loss": 0.3982, + "step": 427 + }, + { + "epoch": 0.9448123620309051, + "grad_norm": 0.161673903465271, + "learning_rate": 8.65817943933379e-06, + "loss": 0.3933, + "step": 428 + }, + { + "epoch": 0.9470198675496688, + "grad_norm": 0.1589784324169159, + "learning_rate": 8.649411819687936e-06, + "loss": 0.3976, + "step": 429 + }, + { + "epoch": 0.9492273730684326, + "grad_norm": 0.166986882686615, + "learning_rate": 8.640620119322146e-06, + "loss": 0.4003, + "step": 430 + }, + { + "epoch": 0.9514348785871964, + "grad_norm": 0.15880529582500458, + "learning_rate": 8.631804396248637e-06, + "loss": 0.3926, + "step": 431 + }, + { + "epoch": 0.9536423841059603, + "grad_norm": 0.1741640269756317, + "learning_rate": 8.62296470863814e-06, + "loss": 0.3978, + "step": 432 + }, + { + "epoch": 0.9558498896247241, + "grad_norm": 0.15339982509613037, + "learning_rate": 8.61410111481952e-06, + "loss": 0.4059, + "step": 433 + }, + { + "epoch": 0.9580573951434879, + "grad_norm": 0.1605585813522339, + "learning_rate": 8.605213673279382e-06, + "loss": 0.3897, + "step": 434 + }, + { + "epoch": 0.9602649006622517, + "grad_norm": 0.154588520526886, + "learning_rate": 8.5963024426617e-06, + "loss": 0.3853, + "step": 435 + }, + { + "epoch": 0.9624724061810155, + "grad_norm": 0.16428066790103912, + "learning_rate": 8.587367481767409e-06, + "loss": 0.3944, + "step": 436 + }, + { + "epoch": 0.9646799116997793, + "grad_norm": 0.15117953717708588, + "learning_rate": 8.578408849554037e-06, + "loss": 0.3924, + "step": 437 + }, + { + "epoch": 0.9668874172185431, + "grad_norm": 0.15928852558135986, + "learning_rate": 8.569426605135307e-06, + "loss": 0.399, + "step": 438 + }, + { + "epoch": 0.9690949227373068, + "grad_norm": 0.16322654485702515, + "learning_rate": 8.560420807780742e-06, + "loss": 0.3863, + "step": 439 + }, + { + "epoch": 0.9713024282560706, + "grad_norm": 0.1749914139509201, + "learning_rate": 8.551391516915288e-06, + "loss": 0.4018, + "step": 440 + }, + { + "epoch": 0.9735099337748344, + "grad_norm": 0.15128456056118011, + "learning_rate": 8.542338792118907e-06, + "loss": 0.3784, + "step": 441 + }, + { + "epoch": 0.9757174392935982, + "grad_norm": 0.19126002490520477, + "learning_rate": 8.533262693126191e-06, + "loss": 0.3923, + "step": 442 + }, + { + "epoch": 0.977924944812362, + "grad_norm": 0.16031384468078613, + "learning_rate": 8.52416327982597e-06, + "loss": 0.4036, + "step": 443 + }, + { + "epoch": 0.9801324503311258, + "grad_norm": 0.16994954645633698, + "learning_rate": 8.515040612260912e-06, + "loss": 0.3865, + "step": 444 + }, + { + "epoch": 0.9823399558498896, + "grad_norm": 0.17667335271835327, + "learning_rate": 8.505894750627128e-06, + "loss": 0.3884, + "step": 445 + }, + { + "epoch": 0.9845474613686535, + "grad_norm": 0.1745615154504776, + "learning_rate": 8.496725755273778e-06, + "loss": 0.3926, + "step": 446 + }, + { + "epoch": 0.9867549668874173, + "grad_norm": 0.16568712890148163, + "learning_rate": 8.487533686702668e-06, + "loss": 0.393, + "step": 447 + }, + { + "epoch": 0.9889624724061811, + "grad_norm": 0.1762251853942871, + "learning_rate": 8.478318605567853e-06, + "loss": 0.3868, + "step": 448 + }, + { + "epoch": 0.9911699779249448, + "grad_norm": 0.18370142579078674, + "learning_rate": 8.46908057267524e-06, + "loss": 0.3939, + "step": 449 + }, + { + "epoch": 0.9933774834437086, + "grad_norm": 0.1815333366394043, + "learning_rate": 8.459819648982182e-06, + "loss": 0.3849, + "step": 450 + }, + { + "epoch": 0.9955849889624724, + "grad_norm": 0.19721707701683044, + "learning_rate": 8.450535895597074e-06, + "loss": 0.3953, + "step": 451 + }, + { + "epoch": 0.9977924944812362, + "grad_norm": 0.16212257742881775, + "learning_rate": 8.441229373778957e-06, + "loss": 0.3933, + "step": 452 + }, + { + "epoch": 1.0, + "grad_norm": 0.2004193216562271, + "learning_rate": 8.43190014493711e-06, + "loss": 0.3914, + "step": 453 + }, + { + "epoch": 1.0022075055187638, + "grad_norm": 0.15856173634529114, + "learning_rate": 8.422548270630646e-06, + "loss": 0.3806, + "step": 454 + }, + { + "epoch": 1.0044150110375276, + "grad_norm": 0.18225938081741333, + "learning_rate": 8.413173812568099e-06, + "loss": 0.3714, + "step": 455 + }, + { + "epoch": 1.0066225165562914, + "grad_norm": 0.17915207147598267, + "learning_rate": 8.403776832607028e-06, + "loss": 0.3834, + "step": 456 + }, + { + "epoch": 1.0088300220750552, + "grad_norm": 0.1916593611240387, + "learning_rate": 8.394357392753599e-06, + "loss": 0.3666, + "step": 457 + }, + { + "epoch": 1.011037527593819, + "grad_norm": 0.166362464427948, + "learning_rate": 8.384915555162183e-06, + "loss": 0.3839, + "step": 458 + }, + { + "epoch": 1.0132450331125828, + "grad_norm": 0.20503659546375275, + "learning_rate": 8.375451382134942e-06, + "loss": 0.381, + "step": 459 + }, + { + "epoch": 1.0154525386313467, + "grad_norm": 0.18665875494480133, + "learning_rate": 8.365964936121422e-06, + "loss": 0.3893, + "step": 460 + }, + { + "epoch": 1.0176600441501105, + "grad_norm": 0.15370848774909973, + "learning_rate": 8.35645627971813e-06, + "loss": 0.3796, + "step": 461 + }, + { + "epoch": 1.0198675496688743, + "grad_norm": 0.17442859709262848, + "learning_rate": 8.346925475668138e-06, + "loss": 0.3633, + "step": 462 + }, + { + "epoch": 1.022075055187638, + "grad_norm": 0.18989215791225433, + "learning_rate": 8.337372586860651e-06, + "loss": 0.3743, + "step": 463 + }, + { + "epoch": 1.0242825607064017, + "grad_norm": 0.16941452026367188, + "learning_rate": 8.327797676330604e-06, + "loss": 0.369, + "step": 464 + }, + { + "epoch": 1.0264900662251655, + "grad_norm": 0.14053016901016235, + "learning_rate": 8.31820080725825e-06, + "loss": 0.3724, + "step": 465 + }, + { + "epoch": 1.0286975717439293, + "grad_norm": 0.17336276173591614, + "learning_rate": 8.308582042968726e-06, + "loss": 0.3903, + "step": 466 + }, + { + "epoch": 1.030905077262693, + "grad_norm": 0.15840588510036469, + "learning_rate": 8.298941446931646e-06, + "loss": 0.3847, + "step": 467 + }, + { + "epoch": 1.033112582781457, + "grad_norm": 0.16392916440963745, + "learning_rate": 8.289279082760685e-06, + "loss": 0.3893, + "step": 468 + }, + { + "epoch": 1.0353200883002207, + "grad_norm": 0.17757326364517212, + "learning_rate": 8.279595014213158e-06, + "loss": 0.3876, + "step": 469 + }, + { + "epoch": 1.0375275938189845, + "grad_norm": 0.1557520627975464, + "learning_rate": 8.26988930518959e-06, + "loss": 0.3782, + "step": 470 + }, + { + "epoch": 1.0397350993377483, + "grad_norm": 0.16154958307743073, + "learning_rate": 8.260162019733305e-06, + "loss": 0.3933, + "step": 471 + }, + { + "epoch": 1.0419426048565121, + "grad_norm": 0.18913333117961884, + "learning_rate": 8.250413222029997e-06, + "loss": 0.3888, + "step": 472 + }, + { + "epoch": 1.044150110375276, + "grad_norm": 0.1763067990541458, + "learning_rate": 8.240642976407313e-06, + "loss": 0.3875, + "step": 473 + }, + { + "epoch": 1.0463576158940397, + "grad_norm": 0.16598297655582428, + "learning_rate": 8.230851347334424e-06, + "loss": 0.3894, + "step": 474 + }, + { + "epoch": 1.0485651214128036, + "grad_norm": 0.17947614192962646, + "learning_rate": 8.221038399421592e-06, + "loss": 0.3853, + "step": 475 + }, + { + "epoch": 1.0507726269315674, + "grad_norm": 0.16084261238574982, + "learning_rate": 8.211204197419766e-06, + "loss": 0.3762, + "step": 476 + }, + { + "epoch": 1.0529801324503312, + "grad_norm": 0.21188747882843018, + "learning_rate": 8.201348806220127e-06, + "loss": 0.3791, + "step": 477 + }, + { + "epoch": 1.055187637969095, + "grad_norm": 0.15903352200984955, + "learning_rate": 8.191472290853683e-06, + "loss": 0.3799, + "step": 478 + }, + { + "epoch": 1.0573951434878588, + "grad_norm": 0.2036534547805786, + "learning_rate": 8.181574716490823e-06, + "loss": 0.3909, + "step": 479 + }, + { + "epoch": 1.0596026490066226, + "grad_norm": 0.16643071174621582, + "learning_rate": 8.171656148440902e-06, + "loss": 0.3842, + "step": 480 + }, + { + "epoch": 1.0618101545253864, + "grad_norm": 0.16434621810913086, + "learning_rate": 8.161716652151795e-06, + "loss": 0.3748, + "step": 481 + }, + { + "epoch": 1.0640176600441502, + "grad_norm": 0.18316681683063507, + "learning_rate": 8.151756293209476e-06, + "loss": 0.3795, + "step": 482 + }, + { + "epoch": 1.0662251655629138, + "grad_norm": 0.14608311653137207, + "learning_rate": 8.14177513733758e-06, + "loss": 0.3749, + "step": 483 + }, + { + "epoch": 1.0684326710816776, + "grad_norm": 0.1656610518693924, + "learning_rate": 8.131773250396973e-06, + "loss": 0.3794, + "step": 484 + }, + { + "epoch": 1.0706401766004414, + "grad_norm": 0.14669524133205414, + "learning_rate": 8.121750698385315e-06, + "loss": 0.3825, + "step": 485 + }, + { + "epoch": 1.0728476821192052, + "grad_norm": 0.17891530692577362, + "learning_rate": 8.111707547436623e-06, + "loss": 0.373, + "step": 486 + }, + { + "epoch": 1.075055187637969, + "grad_norm": 0.15400467813014984, + "learning_rate": 8.10164386382084e-06, + "loss": 0.3792, + "step": 487 + }, + { + "epoch": 1.0772626931567328, + "grad_norm": 0.17551693320274353, + "learning_rate": 8.091559713943388e-06, + "loss": 0.3908, + "step": 488 + }, + { + "epoch": 1.0794701986754967, + "grad_norm": 0.15460826456546783, + "learning_rate": 8.081455164344745e-06, + "loss": 0.3838, + "step": 489 + }, + { + "epoch": 1.0816777041942605, + "grad_norm": 0.15716975927352905, + "learning_rate": 8.071330281699989e-06, + "loss": 0.3934, + "step": 490 + }, + { + "epoch": 1.0838852097130243, + "grad_norm": 0.15411889553070068, + "learning_rate": 8.06118513281837e-06, + "loss": 0.3792, + "step": 491 + }, + { + "epoch": 1.086092715231788, + "grad_norm": 0.18962320685386658, + "learning_rate": 8.051019784642864e-06, + "loss": 0.3888, + "step": 492 + }, + { + "epoch": 1.0883002207505519, + "grad_norm": 0.15896014869213104, + "learning_rate": 8.040834304249733e-06, + "loss": 0.3861, + "step": 493 + }, + { + "epoch": 1.0905077262693157, + "grad_norm": 0.1523059904575348, + "learning_rate": 8.03062875884808e-06, + "loss": 0.3727, + "step": 494 + }, + { + "epoch": 1.0927152317880795, + "grad_norm": 0.13958971202373505, + "learning_rate": 8.02040321577941e-06, + "loss": 0.3718, + "step": 495 + }, + { + "epoch": 1.0949227373068433, + "grad_norm": 0.1423843950033188, + "learning_rate": 8.010157742517185e-06, + "loss": 0.3746, + "step": 496 + }, + { + "epoch": 1.0971302428256071, + "grad_norm": 0.13685709238052368, + "learning_rate": 7.99989240666637e-06, + "loss": 0.3727, + "step": 497 + }, + { + "epoch": 1.099337748344371, + "grad_norm": 0.15423151850700378, + "learning_rate": 7.989607275963e-06, + "loss": 0.3778, + "step": 498 + }, + { + "epoch": 1.1015452538631347, + "grad_norm": 0.15014959871768951, + "learning_rate": 7.979302418273723e-06, + "loss": 0.3747, + "step": 499 + }, + { + "epoch": 1.1037527593818985, + "grad_norm": 0.16919955611228943, + "learning_rate": 7.968977901595355e-06, + "loss": 0.3782, + "step": 500 + }, + { + "epoch": 1.1059602649006623, + "grad_norm": 0.1481235921382904, + "learning_rate": 7.958633794054439e-06, + "loss": 0.3808, + "step": 501 + }, + { + "epoch": 1.108167770419426, + "grad_norm": 0.17575468122959137, + "learning_rate": 7.94827016390678e-06, + "loss": 0.3708, + "step": 502 + }, + { + "epoch": 1.1103752759381897, + "grad_norm": 0.139452263712883, + "learning_rate": 7.93788707953701e-06, + "loss": 0.376, + "step": 503 + }, + { + "epoch": 1.1125827814569536, + "grad_norm": 0.14127641916275024, + "learning_rate": 7.927484609458128e-06, + "loss": 0.3847, + "step": 504 + }, + { + "epoch": 1.1147902869757174, + "grad_norm": 0.15063825249671936, + "learning_rate": 7.917062822311047e-06, + "loss": 0.3842, + "step": 505 + }, + { + "epoch": 1.1169977924944812, + "grad_norm": 0.15780918300151825, + "learning_rate": 7.90662178686415e-06, + "loss": 0.3737, + "step": 506 + }, + { + "epoch": 1.119205298013245, + "grad_norm": 0.14620929956436157, + "learning_rate": 7.896161572012824e-06, + "loss": 0.3862, + "step": 507 + }, + { + "epoch": 1.1214128035320088, + "grad_norm": 0.1493159383535385, + "learning_rate": 7.885682246779016e-06, + "loss": 0.3809, + "step": 508 + }, + { + "epoch": 1.1236203090507726, + "grad_norm": 0.1471249759197235, + "learning_rate": 7.875183880310772e-06, + "loss": 0.3885, + "step": 509 + }, + { + "epoch": 1.1258278145695364, + "grad_norm": 0.1509741246700287, + "learning_rate": 7.86466654188178e-06, + "loss": 0.3778, + "step": 510 + }, + { + "epoch": 1.1280353200883002, + "grad_norm": 0.14812204241752625, + "learning_rate": 7.854130300890921e-06, + "loss": 0.3731, + "step": 511 + }, + { + "epoch": 1.130242825607064, + "grad_norm": 0.1415746510028839, + "learning_rate": 7.843575226861795e-06, + "loss": 0.3798, + "step": 512 + }, + { + "epoch": 1.1324503311258278, + "grad_norm": 0.14002346992492676, + "learning_rate": 7.833001389442283e-06, + "loss": 0.3833, + "step": 513 + }, + { + "epoch": 1.1346578366445916, + "grad_norm": 0.15462878346443176, + "learning_rate": 7.82240885840407e-06, + "loss": 0.3785, + "step": 514 + }, + { + "epoch": 1.1368653421633554, + "grad_norm": 0.15146006643772125, + "learning_rate": 7.811797703642193e-06, + "loss": 0.3767, + "step": 515 + }, + { + "epoch": 1.1390728476821192, + "grad_norm": 0.1545468419790268, + "learning_rate": 7.801167995174575e-06, + "loss": 0.3739, + "step": 516 + }, + { + "epoch": 1.141280353200883, + "grad_norm": 0.1515691876411438, + "learning_rate": 7.790519803141572e-06, + "loss": 0.394, + "step": 517 + }, + { + "epoch": 1.1434878587196469, + "grad_norm": 0.1403704136610031, + "learning_rate": 7.7798531978055e-06, + "loss": 0.3844, + "step": 518 + }, + { + "epoch": 1.1456953642384107, + "grad_norm": 0.12919290363788605, + "learning_rate": 7.769168249550176e-06, + "loss": 0.3746, + "step": 519 + }, + { + "epoch": 1.1479028697571745, + "grad_norm": 0.137840136885643, + "learning_rate": 7.758465028880455e-06, + "loss": 0.3748, + "step": 520 + }, + { + "epoch": 1.150110375275938, + "grad_norm": 0.13867942988872528, + "learning_rate": 7.747743606421761e-06, + "loss": 0.3813, + "step": 521 + }, + { + "epoch": 1.152317880794702, + "grad_norm": 0.1563754677772522, + "learning_rate": 7.737004052919623e-06, + "loss": 0.3673, + "step": 522 + }, + { + "epoch": 1.1545253863134657, + "grad_norm": 0.15125791728496552, + "learning_rate": 7.726246439239209e-06, + "loss": 0.3814, + "step": 523 + }, + { + "epoch": 1.1567328918322295, + "grad_norm": 0.15860167145729065, + "learning_rate": 7.715470836364857e-06, + "loss": 0.3962, + "step": 524 + }, + { + "epoch": 1.1589403973509933, + "grad_norm": 0.14636121690273285, + "learning_rate": 7.704677315399607e-06, + "loss": 0.3786, + "step": 525 + }, + { + "epoch": 1.161147902869757, + "grad_norm": 0.1441173404455185, + "learning_rate": 7.693865947564733e-06, + "loss": 0.3737, + "step": 526 + }, + { + "epoch": 1.163355408388521, + "grad_norm": 0.1591232568025589, + "learning_rate": 7.68303680419927e-06, + "loss": 0.3721, + "step": 527 + }, + { + "epoch": 1.1655629139072847, + "grad_norm": 0.1598597913980484, + "learning_rate": 7.672189956759546e-06, + "loss": 0.3793, + "step": 528 + }, + { + "epoch": 1.1677704194260485, + "grad_norm": 0.16569367051124573, + "learning_rate": 7.661325476818708e-06, + "loss": 0.3799, + "step": 529 + }, + { + "epoch": 1.1699779249448123, + "grad_norm": 0.1879713088274002, + "learning_rate": 7.65044343606626e-06, + "loss": 0.3831, + "step": 530 + }, + { + "epoch": 1.1721854304635762, + "grad_norm": 0.14313288033008575, + "learning_rate": 7.639543906307565e-06, + "loss": 0.3774, + "step": 531 + }, + { + "epoch": 1.17439293598234, + "grad_norm": 0.20557354390621185, + "learning_rate": 7.628626959463405e-06, + "loss": 0.3766, + "step": 532 + }, + { + "epoch": 1.1766004415011038, + "grad_norm": 0.14320634305477142, + "learning_rate": 7.6176926675694786e-06, + "loss": 0.3754, + "step": 533 + }, + { + "epoch": 1.1788079470198676, + "grad_norm": 0.18957629799842834, + "learning_rate": 7.606741102775936e-06, + "loss": 0.3799, + "step": 534 + }, + { + "epoch": 1.1810154525386314, + "grad_norm": 0.15119552612304688, + "learning_rate": 7.595772337346912e-06, + "loss": 0.3863, + "step": 535 + }, + { + "epoch": 1.1832229580573952, + "grad_norm": 0.16690769791603088, + "learning_rate": 7.584786443660028e-06, + "loss": 0.371, + "step": 536 + }, + { + "epoch": 1.185430463576159, + "grad_norm": 0.1685992032289505, + "learning_rate": 7.573783494205936e-06, + "loss": 0.3741, + "step": 537 + }, + { + "epoch": 1.1876379690949228, + "grad_norm": 0.14501997828483582, + "learning_rate": 7.562763561587824e-06, + "loss": 0.3701, + "step": 538 + }, + { + "epoch": 1.1898454746136866, + "grad_norm": 0.16112159192562103, + "learning_rate": 7.55172671852095e-06, + "loss": 0.3781, + "step": 539 + }, + { + "epoch": 1.1920529801324504, + "grad_norm": 0.13081905245780945, + "learning_rate": 7.5406730378321506e-06, + "loss": 0.39, + "step": 540 + }, + { + "epoch": 1.1942604856512142, + "grad_norm": 0.15677790343761444, + "learning_rate": 7.5296025924593705e-06, + "loss": 0.3814, + "step": 541 + }, + { + "epoch": 1.1964679911699778, + "grad_norm": 0.1494779884815216, + "learning_rate": 7.518515455451172e-06, + "loss": 0.3757, + "step": 542 + }, + { + "epoch": 1.1986754966887416, + "grad_norm": 0.1597052365541458, + "learning_rate": 7.50741169996626e-06, + "loss": 0.3772, + "step": 543 + }, + { + "epoch": 1.2008830022075054, + "grad_norm": 0.1445734202861786, + "learning_rate": 7.496291399273e-06, + "loss": 0.3779, + "step": 544 + }, + { + "epoch": 1.2030905077262692, + "grad_norm": 0.14188243448734283, + "learning_rate": 7.485154626748924e-06, + "loss": 0.3801, + "step": 545 + }, + { + "epoch": 1.205298013245033, + "grad_norm": 0.15295171737670898, + "learning_rate": 7.474001455880258e-06, + "loss": 0.3786, + "step": 546 + }, + { + "epoch": 1.2075055187637969, + "grad_norm": 0.12851738929748535, + "learning_rate": 7.4628319602614315e-06, + "loss": 0.3759, + "step": 547 + }, + { + "epoch": 1.2097130242825607, + "grad_norm": 0.12803953886032104, + "learning_rate": 7.451646213594597e-06, + "loss": 0.3807, + "step": 548 + }, + { + "epoch": 1.2119205298013245, + "grad_norm": 0.15474575757980347, + "learning_rate": 7.440444289689135e-06, + "loss": 0.3761, + "step": 549 + }, + { + "epoch": 1.2141280353200883, + "grad_norm": 0.14079326391220093, + "learning_rate": 7.429226262461175e-06, + "loss": 0.3744, + "step": 550 + }, + { + "epoch": 1.216335540838852, + "grad_norm": 0.13553716242313385, + "learning_rate": 7.417992205933104e-06, + "loss": 0.3813, + "step": 551 + }, + { + "epoch": 1.218543046357616, + "grad_norm": 0.14058908820152283, + "learning_rate": 7.406742194233074e-06, + "loss": 0.3722, + "step": 552 + }, + { + "epoch": 1.2207505518763797, + "grad_norm": 0.15452724695205688, + "learning_rate": 7.3954763015945266e-06, + "loss": 0.3628, + "step": 553 + }, + { + "epoch": 1.2229580573951435, + "grad_norm": 0.14715242385864258, + "learning_rate": 7.384194602355685e-06, + "loss": 0.3779, + "step": 554 + }, + { + "epoch": 1.2251655629139073, + "grad_norm": 0.15119391679763794, + "learning_rate": 7.37289717095908e-06, + "loss": 0.3787, + "step": 555 + }, + { + "epoch": 1.2273730684326711, + "grad_norm": 0.1800910234451294, + "learning_rate": 7.361584081951046e-06, + "loss": 0.3781, + "step": 556 + }, + { + "epoch": 1.229580573951435, + "grad_norm": 0.1489747166633606, + "learning_rate": 7.350255409981237e-06, + "loss": 0.3843, + "step": 557 + }, + { + "epoch": 1.2317880794701987, + "grad_norm": 0.14379991590976715, + "learning_rate": 7.338911229802133e-06, + "loss": 0.3832, + "step": 558 + }, + { + "epoch": 1.2339955849889626, + "grad_norm": 0.13160887360572815, + "learning_rate": 7.327551616268541e-06, + "loss": 0.3858, + "step": 559 + }, + { + "epoch": 1.2362030905077264, + "grad_norm": 0.1353141814470291, + "learning_rate": 7.316176644337107e-06, + "loss": 0.3688, + "step": 560 + }, + { + "epoch": 1.23841059602649, + "grad_norm": 0.16255126893520355, + "learning_rate": 7.304786389065823e-06, + "loss": 0.377, + "step": 561 + }, + { + "epoch": 1.240618101545254, + "grad_norm": 0.13924263417720795, + "learning_rate": 7.293380925613524e-06, + "loss": 0.376, + "step": 562 + }, + { + "epoch": 1.2428256070640176, + "grad_norm": 0.1347806751728058, + "learning_rate": 7.281960329239398e-06, + "loss": 0.3881, + "step": 563 + }, + { + "epoch": 1.2450331125827814, + "grad_norm": 0.1494341343641281, + "learning_rate": 7.270524675302491e-06, + "loss": 0.3843, + "step": 564 + }, + { + "epoch": 1.2472406181015452, + "grad_norm": 0.16562990844249725, + "learning_rate": 7.259074039261199e-06, + "loss": 0.3843, + "step": 565 + }, + { + "epoch": 1.249448123620309, + "grad_norm": 0.1605864018201828, + "learning_rate": 7.247608496672786e-06, + "loss": 0.3741, + "step": 566 + }, + { + "epoch": 1.2516556291390728, + "grad_norm": 0.1619981825351715, + "learning_rate": 7.2361281231928725e-06, + "loss": 0.392, + "step": 567 + }, + { + "epoch": 1.2538631346578366, + "grad_norm": 0.13447479903697968, + "learning_rate": 7.2246329945749425e-06, + "loss": 0.3755, + "step": 568 + }, + { + "epoch": 1.2560706401766004, + "grad_norm": 0.1463773399591446, + "learning_rate": 7.213123186669842e-06, + "loss": 0.3913, + "step": 569 + }, + { + "epoch": 1.2582781456953642, + "grad_norm": 0.14680610597133636, + "learning_rate": 7.201598775425278e-06, + "loss": 0.3776, + "step": 570 + }, + { + "epoch": 1.260485651214128, + "grad_norm": 0.14192719757556915, + "learning_rate": 7.190059836885318e-06, + "loss": 0.3829, + "step": 571 + }, + { + "epoch": 1.2626931567328918, + "grad_norm": 0.1412820965051651, + "learning_rate": 7.178506447189887e-06, + "loss": 0.3782, + "step": 572 + }, + { + "epoch": 1.2649006622516556, + "grad_norm": 0.1515144258737564, + "learning_rate": 7.166938682574272e-06, + "loss": 0.3779, + "step": 573 + }, + { + "epoch": 1.2671081677704195, + "grad_norm": 0.148550346493721, + "learning_rate": 7.155356619368604e-06, + "loss": 0.376, + "step": 574 + }, + { + "epoch": 1.2693156732891833, + "grad_norm": 0.14767727255821228, + "learning_rate": 7.14376033399737e-06, + "loss": 0.3768, + "step": 575 + }, + { + "epoch": 1.271523178807947, + "grad_norm": 0.1577431708574295, + "learning_rate": 7.132149902978902e-06, + "loss": 0.373, + "step": 576 + }, + { + "epoch": 1.2737306843267109, + "grad_norm": 0.15225905179977417, + "learning_rate": 7.120525402924871e-06, + "loss": 0.365, + "step": 577 + }, + { + "epoch": 1.2759381898454747, + "grad_norm": 0.16508492827415466, + "learning_rate": 7.108886910539781e-06, + "loss": 0.3866, + "step": 578 + }, + { + "epoch": 1.2781456953642385, + "grad_norm": 0.16345573961734772, + "learning_rate": 7.097234502620468e-06, + "loss": 0.3819, + "step": 579 + }, + { + "epoch": 1.280353200883002, + "grad_norm": 0.15904562175273895, + "learning_rate": 7.085568256055589e-06, + "loss": 0.3877, + "step": 580 + }, + { + "epoch": 1.2825607064017661, + "grad_norm": 0.17447535693645477, + "learning_rate": 7.073888247825111e-06, + "loss": 0.3902, + "step": 581 + }, + { + "epoch": 1.2847682119205297, + "grad_norm": 0.1613752692937851, + "learning_rate": 7.062194554999817e-06, + "loss": 0.3765, + "step": 582 + }, + { + "epoch": 1.2869757174392937, + "grad_norm": 0.1705675572156906, + "learning_rate": 7.05048725474078e-06, + "loss": 0.3699, + "step": 583 + }, + { + "epoch": 1.2891832229580573, + "grad_norm": 0.20963701605796814, + "learning_rate": 7.038766424298865e-06, + "loss": 0.3909, + "step": 584 + }, + { + "epoch": 1.2913907284768211, + "grad_norm": 0.16403751075267792, + "learning_rate": 7.027032141014216e-06, + "loss": 0.3753, + "step": 585 + }, + { + "epoch": 1.293598233995585, + "grad_norm": 0.18310756981372833, + "learning_rate": 7.0152844823157474e-06, + "loss": 0.3807, + "step": 586 + }, + { + "epoch": 1.2958057395143487, + "grad_norm": 0.16576433181762695, + "learning_rate": 7.003523525720626e-06, + "loss": 0.3659, + "step": 587 + }, + { + "epoch": 1.2980132450331126, + "grad_norm": 0.1746741682291031, + "learning_rate": 6.991749348833773e-06, + "loss": 0.3783, + "step": 588 + }, + { + "epoch": 1.3002207505518764, + "grad_norm": 0.16022977232933044, + "learning_rate": 6.979962029347338e-06, + "loss": 0.3745, + "step": 589 + }, + { + "epoch": 1.3024282560706402, + "grad_norm": 0.1607898473739624, + "learning_rate": 6.9681616450401936e-06, + "loss": 0.3732, + "step": 590 + }, + { + "epoch": 1.304635761589404, + "grad_norm": 0.1624252200126648, + "learning_rate": 6.956348273777424e-06, + "loss": 0.3815, + "step": 591 + }, + { + "epoch": 1.3068432671081678, + "grad_norm": 0.14729204773902893, + "learning_rate": 6.944521993509803e-06, + "loss": 0.3734, + "step": 592 + }, + { + "epoch": 1.3090507726269316, + "grad_norm": 0.18559885025024414, + "learning_rate": 6.9326828822732885e-06, + "loss": 0.386, + "step": 593 + }, + { + "epoch": 1.3112582781456954, + "grad_norm": 0.15120132267475128, + "learning_rate": 6.920831018188502e-06, + "loss": 0.3702, + "step": 594 + }, + { + "epoch": 1.3134657836644592, + "grad_norm": 0.18332970142364502, + "learning_rate": 6.908966479460219e-06, + "loss": 0.3807, + "step": 595 + }, + { + "epoch": 1.315673289183223, + "grad_norm": 0.17234960198402405, + "learning_rate": 6.89708934437684e-06, + "loss": 0.3644, + "step": 596 + }, + { + "epoch": 1.3178807947019868, + "grad_norm": 0.18331332504749298, + "learning_rate": 6.885199691309892e-06, + "loss": 0.3841, + "step": 597 + }, + { + "epoch": 1.3200883002207506, + "grad_norm": 0.15819165110588074, + "learning_rate": 6.873297598713497e-06, + "loss": 0.3725, + "step": 598 + }, + { + "epoch": 1.3222958057395142, + "grad_norm": 0.1651681512594223, + "learning_rate": 6.8613831451238636e-06, + "loss": 0.3762, + "step": 599 + }, + { + "epoch": 1.3245033112582782, + "grad_norm": 0.1659359186887741, + "learning_rate": 6.84945640915876e-06, + "loss": 0.3757, + "step": 600 + }, + { + "epoch": 1.3267108167770418, + "grad_norm": 0.15379931032657623, + "learning_rate": 6.837517469517001e-06, + "loss": 0.3735, + "step": 601 + }, + { + "epoch": 1.3289183222958059, + "grad_norm": 0.1771930605173111, + "learning_rate": 6.82556640497793e-06, + "loss": 0.3798, + "step": 602 + }, + { + "epoch": 1.3311258278145695, + "grad_norm": 0.16866551339626312, + "learning_rate": 6.813603294400895e-06, + "loss": 0.3724, + "step": 603 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 0.16047048568725586, + "learning_rate": 6.8016282167247325e-06, + "loss": 0.3744, + "step": 604 + }, + { + "epoch": 1.335540838852097, + "grad_norm": 0.13133041560649872, + "learning_rate": 6.7896412509672385e-06, + "loss": 0.3775, + "step": 605 + }, + { + "epoch": 1.3377483443708609, + "grad_norm": 0.16929350793361664, + "learning_rate": 6.777642476224658e-06, + "loss": 0.3724, + "step": 606 + }, + { + "epoch": 1.3399558498896247, + "grad_norm": 0.14370004832744598, + "learning_rate": 6.765631971671156e-06, + "loss": 0.3795, + "step": 607 + }, + { + "epoch": 1.3421633554083885, + "grad_norm": 0.18519333004951477, + "learning_rate": 6.753609816558297e-06, + "loss": 0.3789, + "step": 608 + }, + { + "epoch": 1.3443708609271523, + "grad_norm": 0.15061011910438538, + "learning_rate": 6.741576090214526e-06, + "loss": 0.3779, + "step": 609 + }, + { + "epoch": 1.346578366445916, + "grad_norm": 0.16702063381671906, + "learning_rate": 6.729530872044629e-06, + "loss": 0.3791, + "step": 610 + }, + { + "epoch": 1.34878587196468, + "grad_norm": 0.1492423117160797, + "learning_rate": 6.717474241529235e-06, + "loss": 0.3821, + "step": 611 + }, + { + "epoch": 1.3509933774834437, + "grad_norm": 0.15660499036312103, + "learning_rate": 6.705406278224269e-06, + "loss": 0.3764, + "step": 612 + }, + { + "epoch": 1.3532008830022075, + "grad_norm": 0.13613760471343994, + "learning_rate": 6.69332706176044e-06, + "loss": 0.3842, + "step": 613 + }, + { + "epoch": 1.3554083885209713, + "grad_norm": 0.1655375063419342, + "learning_rate": 6.681236671842709e-06, + "loss": 0.3861, + "step": 614 + }, + { + "epoch": 1.3576158940397351, + "grad_norm": 0.15369166433811188, + "learning_rate": 6.669135188249767e-06, + "loss": 0.3753, + "step": 615 + }, + { + "epoch": 1.359823399558499, + "grad_norm": 0.1611892580986023, + "learning_rate": 6.657022690833503e-06, + "loss": 0.3732, + "step": 616 + }, + { + "epoch": 1.3620309050772628, + "grad_norm": 0.15144729614257812, + "learning_rate": 6.644899259518485e-06, + "loss": 0.3756, + "step": 617 + }, + { + "epoch": 1.3642384105960264, + "grad_norm": 0.1591017246246338, + "learning_rate": 6.632764974301429e-06, + "loss": 0.3643, + "step": 618 + }, + { + "epoch": 1.3664459161147904, + "grad_norm": 0.15775880217552185, + "learning_rate": 6.620619915250666e-06, + "loss": 0.3773, + "step": 619 + }, + { + "epoch": 1.368653421633554, + "grad_norm": 0.14554527401924133, + "learning_rate": 6.608464162505621e-06, + "loss": 0.3788, + "step": 620 + }, + { + "epoch": 1.370860927152318, + "grad_norm": 0.15284626185894012, + "learning_rate": 6.596297796276284e-06, + "loss": 0.3743, + "step": 621 + }, + { + "epoch": 1.3730684326710816, + "grad_norm": 0.14736317098140717, + "learning_rate": 6.584120896842675e-06, + "loss": 0.3763, + "step": 622 + }, + { + "epoch": 1.3752759381898454, + "grad_norm": 0.15955647826194763, + "learning_rate": 6.571933544554319e-06, + "loss": 0.3664, + "step": 623 + }, + { + "epoch": 1.3774834437086092, + "grad_norm": 0.16919533908367157, + "learning_rate": 6.559735819829713e-06, + "loss": 0.3755, + "step": 624 + }, + { + "epoch": 1.379690949227373, + "grad_norm": 0.15306392312049866, + "learning_rate": 6.5475278031558e-06, + "loss": 0.3772, + "step": 625 + }, + { + "epoch": 1.3818984547461368, + "grad_norm": 0.1504804790019989, + "learning_rate": 6.5353095750874295e-06, + "loss": 0.3709, + "step": 626 + }, + { + "epoch": 1.3841059602649006, + "grad_norm": 0.14429201185703278, + "learning_rate": 6.523081216246838e-06, + "loss": 0.3709, + "step": 627 + }, + { + "epoch": 1.3863134657836644, + "grad_norm": 0.14882300794124603, + "learning_rate": 6.5108428073231e-06, + "loss": 0.3768, + "step": 628 + }, + { + "epoch": 1.3885209713024282, + "grad_norm": 0.1360517293214798, + "learning_rate": 6.498594429071618e-06, + "loss": 0.3768, + "step": 629 + }, + { + "epoch": 1.390728476821192, + "grad_norm": 0.13942170143127441, + "learning_rate": 6.486336162313568e-06, + "loss": 0.3796, + "step": 630 + }, + { + "epoch": 1.3929359823399559, + "grad_norm": 0.14323705434799194, + "learning_rate": 6.474068087935379e-06, + "loss": 0.379, + "step": 631 + }, + { + "epoch": 1.3951434878587197, + "grad_norm": 0.15375126898288727, + "learning_rate": 6.461790286888196e-06, + "loss": 0.3721, + "step": 632 + }, + { + "epoch": 1.3973509933774835, + "grad_norm": 0.15478521585464478, + "learning_rate": 6.449502840187344e-06, + "loss": 0.3726, + "step": 633 + }, + { + "epoch": 1.3995584988962473, + "grad_norm": 0.16403432190418243, + "learning_rate": 6.437205828911797e-06, + "loss": 0.3765, + "step": 634 + }, + { + "epoch": 1.401766004415011, + "grad_norm": 0.1496008038520813, + "learning_rate": 6.424899334203641e-06, + "loss": 0.3785, + "step": 635 + }, + { + "epoch": 1.403973509933775, + "grad_norm": 0.152191162109375, + "learning_rate": 6.4125834372675355e-06, + "loss": 0.3752, + "step": 636 + }, + { + "epoch": 1.4061810154525387, + "grad_norm": 0.15051637589931488, + "learning_rate": 6.400258219370183e-06, + "loss": 0.3743, + "step": 637 + }, + { + "epoch": 1.4083885209713025, + "grad_norm": 0.15709619224071503, + "learning_rate": 6.387923761839794e-06, + "loss": 0.3697, + "step": 638 + }, + { + "epoch": 1.410596026490066, + "grad_norm": 0.14248456060886383, + "learning_rate": 6.375580146065539e-06, + "loss": 0.373, + "step": 639 + }, + { + "epoch": 1.4128035320088301, + "grad_norm": 0.1371423304080963, + "learning_rate": 6.363227453497029e-06, + "loss": 0.3819, + "step": 640 + }, + { + "epoch": 1.4150110375275937, + "grad_norm": 0.16379839181900024, + "learning_rate": 6.350865765643759e-06, + "loss": 0.3793, + "step": 641 + }, + { + "epoch": 1.4172185430463577, + "grad_norm": 0.14373619854450226, + "learning_rate": 6.3384951640745865e-06, + "loss": 0.3779, + "step": 642 + }, + { + "epoch": 1.4194260485651213, + "grad_norm": 0.14477743208408356, + "learning_rate": 6.32611573041718e-06, + "loss": 0.3786, + "step": 643 + }, + { + "epoch": 1.4216335540838851, + "grad_norm": 0.17920951545238495, + "learning_rate": 6.313727546357498e-06, + "loss": 0.3901, + "step": 644 + }, + { + "epoch": 1.423841059602649, + "grad_norm": 0.14486798644065857, + "learning_rate": 6.301330693639224e-06, + "loss": 0.3786, + "step": 645 + }, + { + "epoch": 1.4260485651214128, + "grad_norm": 0.13817182183265686, + "learning_rate": 6.288925254063249e-06, + "loss": 0.3725, + "step": 646 + }, + { + "epoch": 1.4282560706401766, + "grad_norm": 0.14681169390678406, + "learning_rate": 6.27651130948713e-06, + "loss": 0.3851, + "step": 647 + }, + { + "epoch": 1.4304635761589404, + "grad_norm": 0.14171209931373596, + "learning_rate": 6.264088941824534e-06, + "loss": 0.3802, + "step": 648 + }, + { + "epoch": 1.4326710816777042, + "grad_norm": 0.14188838005065918, + "learning_rate": 6.251658233044715e-06, + "loss": 0.3764, + "step": 649 + }, + { + "epoch": 1.434878587196468, + "grad_norm": 0.15146797895431519, + "learning_rate": 6.239219265171964e-06, + "loss": 0.3791, + "step": 650 + }, + { + "epoch": 1.4370860927152318, + "grad_norm": 0.1602582484483719, + "learning_rate": 6.22677212028507e-06, + "loss": 0.3773, + "step": 651 + }, + { + "epoch": 1.4392935982339956, + "grad_norm": 0.14558373391628265, + "learning_rate": 6.214316880516775e-06, + "loss": 0.3631, + "step": 652 + }, + { + "epoch": 1.4415011037527594, + "grad_norm": 0.17893047630786896, + "learning_rate": 6.201853628053243e-06, + "loss": 0.374, + "step": 653 + }, + { + "epoch": 1.4437086092715232, + "grad_norm": 0.14639593660831451, + "learning_rate": 6.189382445133502e-06, + "loss": 0.3753, + "step": 654 + }, + { + "epoch": 1.445916114790287, + "grad_norm": 0.14472749829292297, + "learning_rate": 6.17690341404891e-06, + "loss": 0.3796, + "step": 655 + }, + { + "epoch": 1.4481236203090508, + "grad_norm": 0.1566278040409088, + "learning_rate": 6.16441661714262e-06, + "loss": 0.3741, + "step": 656 + }, + { + "epoch": 1.4503311258278146, + "grad_norm": 0.14471283555030823, + "learning_rate": 6.151922136809017e-06, + "loss": 0.3808, + "step": 657 + }, + { + "epoch": 1.4525386313465782, + "grad_norm": 0.15734338760375977, + "learning_rate": 6.13942005549319e-06, + "loss": 0.3781, + "step": 658 + }, + { + "epoch": 1.4547461368653423, + "grad_norm": 0.13064594566822052, + "learning_rate": 6.126910455690384e-06, + "loss": 0.377, + "step": 659 + }, + { + "epoch": 1.4569536423841059, + "grad_norm": 0.15152592957019806, + "learning_rate": 6.114393419945455e-06, + "loss": 0.3784, + "step": 660 + }, + { + "epoch": 1.4591611479028699, + "grad_norm": 0.14837685227394104, + "learning_rate": 6.101869030852321e-06, + "loss": 0.3889, + "step": 661 + }, + { + "epoch": 1.4613686534216335, + "grad_norm": 0.14178897440433502, + "learning_rate": 6.089337371053429e-06, + "loss": 0.3714, + "step": 662 + }, + { + "epoch": 1.4635761589403973, + "grad_norm": 0.15140818059444427, + "learning_rate": 6.076798523239194e-06, + "loss": 0.3816, + "step": 663 + }, + { + "epoch": 1.465783664459161, + "grad_norm": 0.14195586740970612, + "learning_rate": 6.064252570147464e-06, + "loss": 0.3738, + "step": 664 + }, + { + "epoch": 1.467991169977925, + "grad_norm": 0.13242661952972412, + "learning_rate": 6.051699594562976e-06, + "loss": 0.3669, + "step": 665 + }, + { + "epoch": 1.4701986754966887, + "grad_norm": 0.1379902958869934, + "learning_rate": 6.039139679316797e-06, + "loss": 0.362, + "step": 666 + }, + { + "epoch": 1.4724061810154525, + "grad_norm": 0.1443612426519394, + "learning_rate": 6.026572907285791e-06, + "loss": 0.3711, + "step": 667 + }, + { + "epoch": 1.4746136865342163, + "grad_norm": 0.14281077682971954, + "learning_rate": 6.013999361392064e-06, + "loss": 0.3753, + "step": 668 + }, + { + "epoch": 1.4768211920529801, + "grad_norm": 0.14133213460445404, + "learning_rate": 6.001419124602422e-06, + "loss": 0.377, + "step": 669 + }, + { + "epoch": 1.479028697571744, + "grad_norm": 0.14672201871871948, + "learning_rate": 5.988832279927818e-06, + "loss": 0.3763, + "step": 670 + }, + { + "epoch": 1.4812362030905077, + "grad_norm": 0.13457442820072174, + "learning_rate": 5.976238910422811e-06, + "loss": 0.381, + "step": 671 + }, + { + "epoch": 1.4834437086092715, + "grad_norm": 0.15843527019023895, + "learning_rate": 5.963639099185011e-06, + "loss": 0.3773, + "step": 672 + }, + { + "epoch": 1.4856512141280354, + "grad_norm": 0.15779507160186768, + "learning_rate": 5.9510329293545356e-06, + "loss": 0.3789, + "step": 673 + }, + { + "epoch": 1.4878587196467992, + "grad_norm": 0.14860793948173523, + "learning_rate": 5.938420484113458e-06, + "loss": 0.3765, + "step": 674 + }, + { + "epoch": 1.490066225165563, + "grad_norm": 0.13150885701179504, + "learning_rate": 5.925801846685262e-06, + "loss": 0.37, + "step": 675 + }, + { + "epoch": 1.4922737306843268, + "grad_norm": 0.13564042747020721, + "learning_rate": 5.913177100334287e-06, + "loss": 0.3788, + "step": 676 + }, + { + "epoch": 1.4944812362030906, + "grad_norm": 0.14382179081439972, + "learning_rate": 5.900546328365189e-06, + "loss": 0.381, + "step": 677 + }, + { + "epoch": 1.4966887417218544, + "grad_norm": 0.14150989055633545, + "learning_rate": 5.887909614122378e-06, + "loss": 0.3683, + "step": 678 + }, + { + "epoch": 1.498896247240618, + "grad_norm": 0.13479964435100555, + "learning_rate": 5.875267040989475e-06, + "loss": 0.3651, + "step": 679 + }, + { + "epoch": 1.501103752759382, + "grad_norm": 0.14743578433990479, + "learning_rate": 5.862618692388765e-06, + "loss": 0.3676, + "step": 680 + }, + { + "epoch": 1.5033112582781456, + "grad_norm": 0.1419716477394104, + "learning_rate": 5.849964651780637e-06, + "loss": 0.367, + "step": 681 + }, + { + "epoch": 1.5055187637969096, + "grad_norm": 0.15260837972164154, + "learning_rate": 5.837305002663043e-06, + "loss": 0.3814, + "step": 682 + }, + { + "epoch": 1.5077262693156732, + "grad_norm": 0.12932439148426056, + "learning_rate": 5.824639828570943e-06, + "loss": 0.3868, + "step": 683 + }, + { + "epoch": 1.5099337748344372, + "grad_norm": 0.15234258770942688, + "learning_rate": 5.811969213075749e-06, + "loss": 0.3695, + "step": 684 + }, + { + "epoch": 1.5121412803532008, + "grad_norm": 0.1671096384525299, + "learning_rate": 5.799293239784783e-06, + "loss": 0.3821, + "step": 685 + }, + { + "epoch": 1.5143487858719646, + "grad_norm": 0.14161601662635803, + "learning_rate": 5.786611992340719e-06, + "loss": 0.3657, + "step": 686 + }, + { + "epoch": 1.5165562913907285, + "grad_norm": 0.16359050571918488, + "learning_rate": 5.773925554421033e-06, + "loss": 0.3788, + "step": 687 + }, + { + "epoch": 1.5187637969094923, + "grad_norm": 0.14034205675125122, + "learning_rate": 5.761234009737449e-06, + "loss": 0.3779, + "step": 688 + }, + { + "epoch": 1.520971302428256, + "grad_norm": 0.14073529839515686, + "learning_rate": 5.74853744203539e-06, + "loss": 0.3756, + "step": 689 + }, + { + "epoch": 1.5231788079470199, + "grad_norm": 0.14516787230968475, + "learning_rate": 5.735835935093421e-06, + "loss": 0.3886, + "step": 690 + }, + { + "epoch": 1.5253863134657837, + "grad_norm": 0.1415787935256958, + "learning_rate": 5.723129572722701e-06, + "loss": 0.375, + "step": 691 + }, + { + "epoch": 1.5275938189845475, + "grad_norm": 0.13726189732551575, + "learning_rate": 5.710418438766428e-06, + "loss": 0.3735, + "step": 692 + }, + { + "epoch": 1.5298013245033113, + "grad_norm": 0.14881715178489685, + "learning_rate": 5.6977026170992834e-06, + "loss": 0.3735, + "step": 693 + }, + { + "epoch": 1.532008830022075, + "grad_norm": 0.1551184505224228, + "learning_rate": 5.68498219162688e-06, + "loss": 0.3815, + "step": 694 + }, + { + "epoch": 1.534216335540839, + "grad_norm": 0.15148243308067322, + "learning_rate": 5.672257246285212e-06, + "loss": 0.3736, + "step": 695 + }, + { + "epoch": 1.5364238410596025, + "grad_norm": 0.15206126868724823, + "learning_rate": 5.659527865040097e-06, + "loss": 0.3771, + "step": 696 + }, + { + "epoch": 1.5386313465783665, + "grad_norm": 0.14559738337993622, + "learning_rate": 5.6467941318866214e-06, + "loss": 0.3742, + "step": 697 + }, + { + "epoch": 1.5408388520971301, + "grad_norm": 0.13176165521144867, + "learning_rate": 5.634056130848589e-06, + "loss": 0.3715, + "step": 698 + }, + { + "epoch": 1.5430463576158941, + "grad_norm": 0.14761999249458313, + "learning_rate": 5.621313945977964e-06, + "loss": 0.377, + "step": 699 + }, + { + "epoch": 1.5452538631346577, + "grad_norm": 0.12837889790534973, + "learning_rate": 5.60856766135432e-06, + "loss": 0.371, + "step": 700 + }, + { + "epoch": 1.5474613686534218, + "grad_norm": 0.1346467137336731, + "learning_rate": 5.595817361084283e-06, + "loss": 0.3757, + "step": 701 + }, + { + "epoch": 1.5496688741721854, + "grad_norm": 0.129667267203331, + "learning_rate": 5.583063129300971e-06, + "loss": 0.3821, + "step": 702 + }, + { + "epoch": 1.5518763796909494, + "grad_norm": 0.1456080973148346, + "learning_rate": 5.570305050163453e-06, + "loss": 0.3659, + "step": 703 + }, + { + "epoch": 1.554083885209713, + "grad_norm": 0.13144616782665253, + "learning_rate": 5.557543207856175e-06, + "loss": 0.3755, + "step": 704 + }, + { + "epoch": 1.5562913907284768, + "grad_norm": 0.13695232570171356, + "learning_rate": 5.544777686588424e-06, + "loss": 0.3704, + "step": 705 + }, + { + "epoch": 1.5584988962472406, + "grad_norm": 0.1521306186914444, + "learning_rate": 5.532008570593756e-06, + "loss": 0.3829, + "step": 706 + }, + { + "epoch": 1.5607064017660044, + "grad_norm": 0.1419302225112915, + "learning_rate": 5.519235944129448e-06, + "loss": 0.371, + "step": 707 + }, + { + "epoch": 1.5629139072847682, + "grad_norm": 0.14487622678279877, + "learning_rate": 5.506459891475944e-06, + "loss": 0.3785, + "step": 708 + }, + { + "epoch": 1.565121412803532, + "grad_norm": 0.15528716146945953, + "learning_rate": 5.4936804969362924e-06, + "loss": 0.3853, + "step": 709 + }, + { + "epoch": 1.5673289183222958, + "grad_norm": 0.14845651388168335, + "learning_rate": 5.480897844835595e-06, + "loss": 0.3853, + "step": 710 + }, + { + "epoch": 1.5695364238410596, + "grad_norm": 0.13595354557037354, + "learning_rate": 5.4681120195204466e-06, + "loss": 0.3718, + "step": 711 + }, + { + "epoch": 1.5717439293598234, + "grad_norm": 0.15339161455631256, + "learning_rate": 5.455323105358384e-06, + "loss": 0.3791, + "step": 712 + }, + { + "epoch": 1.5739514348785872, + "grad_norm": 0.1527385264635086, + "learning_rate": 5.442531186737322e-06, + "loss": 0.3703, + "step": 713 + }, + { + "epoch": 1.576158940397351, + "grad_norm": 0.13763274252414703, + "learning_rate": 5.429736348065003e-06, + "loss": 0.3638, + "step": 714 + }, + { + "epoch": 1.5783664459161146, + "grad_norm": 0.16397210955619812, + "learning_rate": 5.416938673768437e-06, + "loss": 0.3761, + "step": 715 + }, + { + "epoch": 1.5805739514348787, + "grad_norm": 0.14585243165493011, + "learning_rate": 5.404138248293346e-06, + "loss": 0.3762, + "step": 716 + }, + { + "epoch": 1.5827814569536423, + "grad_norm": 0.13906413316726685, + "learning_rate": 5.3913351561036e-06, + "loss": 0.3669, + "step": 717 + }, + { + "epoch": 1.5849889624724063, + "grad_norm": 0.13377633690834045, + "learning_rate": 5.378529481680672e-06, + "loss": 0.3721, + "step": 718 + }, + { + "epoch": 1.5871964679911699, + "grad_norm": 0.14944018423557281, + "learning_rate": 5.365721309523072e-06, + "loss": 0.3739, + "step": 719 + }, + { + "epoch": 1.589403973509934, + "grad_norm": 0.14180952310562134, + "learning_rate": 5.352910724145789e-06, + "loss": 0.373, + "step": 720 + }, + { + "epoch": 1.5916114790286975, + "grad_norm": 0.1462966352701187, + "learning_rate": 5.340097810079741e-06, + "loss": 0.3768, + "step": 721 + }, + { + "epoch": 1.5938189845474615, + "grad_norm": 0.12976975739002228, + "learning_rate": 5.327282651871205e-06, + "loss": 0.3638, + "step": 722 + }, + { + "epoch": 1.596026490066225, + "grad_norm": 0.15462516248226166, + "learning_rate": 5.3144653340812715e-06, + "loss": 0.3854, + "step": 723 + }, + { + "epoch": 1.598233995584989, + "grad_norm": 0.13364708423614502, + "learning_rate": 5.301645941285278e-06, + "loss": 0.3862, + "step": 724 + }, + { + "epoch": 1.6004415011037527, + "grad_norm": 0.14875876903533936, + "learning_rate": 5.288824558072257e-06, + "loss": 0.3847, + "step": 725 + }, + { + "epoch": 1.6026490066225165, + "grad_norm": 0.13750705122947693, + "learning_rate": 5.276001269044369e-06, + "loss": 0.3792, + "step": 726 + }, + { + "epoch": 1.6048565121412803, + "grad_norm": 0.13552795350551605, + "learning_rate": 5.263176158816355e-06, + "loss": 0.3824, + "step": 727 + }, + { + "epoch": 1.6070640176600441, + "grad_norm": 0.13398821651935577, + "learning_rate": 5.250349312014976e-06, + "loss": 0.3728, + "step": 728 + }, + { + "epoch": 1.609271523178808, + "grad_norm": 0.14047983288764954, + "learning_rate": 5.237520813278443e-06, + "loss": 0.3833, + "step": 729 + }, + { + "epoch": 1.6114790286975718, + "grad_norm": 0.1419335901737213, + "learning_rate": 5.224690747255875e-06, + "loss": 0.3738, + "step": 730 + }, + { + "epoch": 1.6136865342163356, + "grad_norm": 0.1423855870962143, + "learning_rate": 5.211859198606729e-06, + "loss": 0.3702, + "step": 731 + }, + { + "epoch": 1.6158940397350994, + "grad_norm": 0.14920298755168915, + "learning_rate": 5.199026252000245e-06, + "loss": 0.3719, + "step": 732 + }, + { + "epoch": 1.6181015452538632, + "grad_norm": 0.13715742528438568, + "learning_rate": 5.186191992114892e-06, + "loss": 0.3766, + "step": 733 + }, + { + "epoch": 1.6203090507726268, + "grad_norm": 0.142774298787117, + "learning_rate": 5.173356503637799e-06, + "loss": 0.3693, + "step": 734 + }, + { + "epoch": 1.6225165562913908, + "grad_norm": 0.12615923583507538, + "learning_rate": 5.160519871264204e-06, + "loss": 0.375, + "step": 735 + }, + { + "epoch": 1.6247240618101544, + "grad_norm": 0.14882232248783112, + "learning_rate": 5.147682179696893e-06, + "loss": 0.3746, + "step": 736 + }, + { + "epoch": 1.6269315673289184, + "grad_norm": 0.12539036571979523, + "learning_rate": 5.134843513645642e-06, + "loss": 0.3786, + "step": 737 + }, + { + "epoch": 1.629139072847682, + "grad_norm": 0.14418825507164001, + "learning_rate": 5.122003957826657e-06, + "loss": 0.3693, + "step": 738 + }, + { + "epoch": 1.631346578366446, + "grad_norm": 0.14610478281974792, + "learning_rate": 5.109163596962013e-06, + "loss": 0.3759, + "step": 739 + }, + { + "epoch": 1.6335540838852096, + "grad_norm": 0.13532301783561707, + "learning_rate": 5.096322515779101e-06, + "loss": 0.373, + "step": 740 + }, + { + "epoch": 1.6357615894039736, + "grad_norm": 0.14194665849208832, + "learning_rate": 5.083480799010061e-06, + "loss": 0.3762, + "step": 741 + }, + { + "epoch": 1.6379690949227372, + "grad_norm": 0.12951509654521942, + "learning_rate": 5.070638531391229e-06, + "loss": 0.3723, + "step": 742 + }, + { + "epoch": 1.6401766004415013, + "grad_norm": 0.14456267654895782, + "learning_rate": 5.057795797662577e-06, + "loss": 0.3666, + "step": 743 + }, + { + "epoch": 1.6423841059602649, + "grad_norm": 0.13711507618427277, + "learning_rate": 5.044952682567148e-06, + "loss": 0.3749, + "step": 744 + }, + { + "epoch": 1.6445916114790287, + "grad_norm": 0.12987838685512543, + "learning_rate": 5.032109270850507e-06, + "loss": 0.3779, + "step": 745 + }, + { + "epoch": 1.6467991169977925, + "grad_norm": 0.1495376080274582, + "learning_rate": 5.019265647260173e-06, + "loss": 0.3704, + "step": 746 + }, + { + "epoch": 1.6490066225165563, + "grad_norm": 0.13476036489009857, + "learning_rate": 5.006421896545064e-06, + "loss": 0.3663, + "step": 747 + }, + { + "epoch": 1.65121412803532, + "grad_norm": 0.1335316300392151, + "learning_rate": 4.993578103454938e-06, + "loss": 0.3843, + "step": 748 + }, + { + "epoch": 1.653421633554084, + "grad_norm": 0.13593193888664246, + "learning_rate": 4.980734352739829e-06, + "loss": 0.3707, + "step": 749 + }, + { + "epoch": 1.6556291390728477, + "grad_norm": 0.13324475288391113, + "learning_rate": 4.967890729149494e-06, + "loss": 0.3746, + "step": 750 + }, + { + "epoch": 1.6578366445916115, + "grad_norm": 0.12398661673069, + "learning_rate": 4.955047317432854e-06, + "loss": 0.3783, + "step": 751 + }, + { + "epoch": 1.6600441501103753, + "grad_norm": 0.12614096701145172, + "learning_rate": 4.942204202337425e-06, + "loss": 0.3806, + "step": 752 + }, + { + "epoch": 1.6622516556291391, + "grad_norm": 0.14331743121147156, + "learning_rate": 4.929361468608773e-06, + "loss": 0.3833, + "step": 753 + }, + { + "epoch": 1.664459161147903, + "grad_norm": 0.1342383772134781, + "learning_rate": 4.9165192009899414e-06, + "loss": 0.3799, + "step": 754 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 0.13297194242477417, + "learning_rate": 4.9036774842209e-06, + "loss": 0.3702, + "step": 755 + }, + { + "epoch": 1.6688741721854305, + "grad_norm": 0.12575410306453705, + "learning_rate": 4.890836403037988e-06, + "loss": 0.364, + "step": 756 + }, + { + "epoch": 1.6710816777041941, + "grad_norm": 0.1317722052335739, + "learning_rate": 4.877996042173345e-06, + "loss": 0.3782, + "step": 757 + }, + { + "epoch": 1.6732891832229582, + "grad_norm": 0.13363225758075714, + "learning_rate": 4.865156486354359e-06, + "loss": 0.3804, + "step": 758 + }, + { + "epoch": 1.6754966887417218, + "grad_norm": 0.14083701372146606, + "learning_rate": 4.852317820303107e-06, + "loss": 0.38, + "step": 759 + }, + { + "epoch": 1.6777041942604858, + "grad_norm": 0.1284651756286621, + "learning_rate": 4.839480128735798e-06, + "loss": 0.3655, + "step": 760 + }, + { + "epoch": 1.6799116997792494, + "grad_norm": 0.1379299908876419, + "learning_rate": 4.826643496362202e-06, + "loss": 0.3779, + "step": 761 + }, + { + "epoch": 1.6821192052980134, + "grad_norm": 0.13508039712905884, + "learning_rate": 4.81380800788511e-06, + "loss": 0.3785, + "step": 762 + }, + { + "epoch": 1.684326710816777, + "grad_norm": 0.14211615920066833, + "learning_rate": 4.800973747999757e-06, + "loss": 0.3648, + "step": 763 + }, + { + "epoch": 1.6865342163355408, + "grad_norm": 0.1293489634990692, + "learning_rate": 4.788140801393273e-06, + "loss": 0.3663, + "step": 764 + }, + { + "epoch": 1.6887417218543046, + "grad_norm": 0.13812755048274994, + "learning_rate": 4.775309252744126e-06, + "loss": 0.3643, + "step": 765 + }, + { + "epoch": 1.6909492273730684, + "grad_norm": 0.13112276792526245, + "learning_rate": 4.762479186721559e-06, + "loss": 0.3703, + "step": 766 + }, + { + "epoch": 1.6931567328918322, + "grad_norm": 0.12749050557613373, + "learning_rate": 4.7496506879850264e-06, + "loss": 0.3844, + "step": 767 + }, + { + "epoch": 1.695364238410596, + "grad_norm": 0.1374199390411377, + "learning_rate": 4.736823841183645e-06, + "loss": 0.3689, + "step": 768 + }, + { + "epoch": 1.6975717439293598, + "grad_norm": 0.12659025192260742, + "learning_rate": 4.723998730955633e-06, + "loss": 0.3606, + "step": 769 + }, + { + "epoch": 1.6997792494481236, + "grad_norm": 0.11276005953550339, + "learning_rate": 4.711175441927746e-06, + "loss": 0.3722, + "step": 770 + }, + { + "epoch": 1.7019867549668874, + "grad_norm": 0.12801696360111237, + "learning_rate": 4.698354058714724e-06, + "loss": 0.3789, + "step": 771 + }, + { + "epoch": 1.7041942604856513, + "grad_norm": 0.12910126149654388, + "learning_rate": 4.685534665918731e-06, + "loss": 0.3669, + "step": 772 + }, + { + "epoch": 1.706401766004415, + "grad_norm": 0.12605293095111847, + "learning_rate": 4.672717348128796e-06, + "loss": 0.377, + "step": 773 + }, + { + "epoch": 1.7086092715231787, + "grad_norm": 0.12655854225158691, + "learning_rate": 4.65990218992026e-06, + "loss": 0.3738, + "step": 774 + }, + { + "epoch": 1.7108167770419427, + "grad_norm": 0.1326448917388916, + "learning_rate": 4.647089275854212e-06, + "loss": 0.3847, + "step": 775 + }, + { + "epoch": 1.7130242825607063, + "grad_norm": 0.12570975720882416, + "learning_rate": 4.63427869047693e-06, + "loss": 0.3652, + "step": 776 + }, + { + "epoch": 1.7152317880794703, + "grad_norm": 0.13106000423431396, + "learning_rate": 4.621470518319329e-06, + "loss": 0.3781, + "step": 777 + }, + { + "epoch": 1.7174392935982339, + "grad_norm": 0.12707674503326416, + "learning_rate": 4.608664843896402e-06, + "loss": 0.367, + "step": 778 + }, + { + "epoch": 1.719646799116998, + "grad_norm": 0.1342884600162506, + "learning_rate": 4.595861751706656e-06, + "loss": 0.3827, + "step": 779 + }, + { + "epoch": 1.7218543046357615, + "grad_norm": 0.15550127625465393, + "learning_rate": 4.583061326231564e-06, + "loss": 0.384, + "step": 780 + }, + { + "epoch": 1.7240618101545255, + "grad_norm": 0.1267244517803192, + "learning_rate": 4.570263651934998e-06, + "loss": 0.38, + "step": 781 + }, + { + "epoch": 1.7262693156732891, + "grad_norm": 0.13444717228412628, + "learning_rate": 4.557468813262678e-06, + "loss": 0.3856, + "step": 782 + }, + { + "epoch": 1.728476821192053, + "grad_norm": 0.12585113942623138, + "learning_rate": 4.544676894641617e-06, + "loss": 0.3603, + "step": 783 + }, + { + "epoch": 1.7306843267108167, + "grad_norm": 0.14303074777126312, + "learning_rate": 4.531887980479555e-06, + "loss": 0.3769, + "step": 784 + }, + { + "epoch": 1.7328918322295805, + "grad_norm": 0.14289291203022003, + "learning_rate": 4.519102155164407e-06, + "loss": 0.3679, + "step": 785 + }, + { + "epoch": 1.7350993377483444, + "grad_norm": 0.13349799811840057, + "learning_rate": 4.506319503063708e-06, + "loss": 0.3702, + "step": 786 + }, + { + "epoch": 1.7373068432671082, + "grad_norm": 0.15247680246829987, + "learning_rate": 4.493540108524057e-06, + "loss": 0.3713, + "step": 787 + }, + { + "epoch": 1.739514348785872, + "grad_norm": 0.14266639947891235, + "learning_rate": 4.480764055870553e-06, + "loss": 0.3742, + "step": 788 + }, + { + "epoch": 1.7417218543046358, + "grad_norm": 0.13838379085063934, + "learning_rate": 4.467991429406247e-06, + "loss": 0.3857, + "step": 789 + }, + { + "epoch": 1.7439293598233996, + "grad_norm": 0.13494673371315002, + "learning_rate": 4.455222313411578e-06, + "loss": 0.3783, + "step": 790 + }, + { + "epoch": 1.7461368653421634, + "grad_norm": 0.12723763287067413, + "learning_rate": 4.442456792143825e-06, + "loss": 0.3657, + "step": 791 + }, + { + "epoch": 1.7483443708609272, + "grad_norm": 0.1302330642938614, + "learning_rate": 4.4296949498365496e-06, + "loss": 0.3671, + "step": 792 + }, + { + "epoch": 1.7505518763796908, + "grad_norm": 0.12404655665159225, + "learning_rate": 4.416936870699031e-06, + "loss": 0.3755, + "step": 793 + }, + { + "epoch": 1.7527593818984548, + "grad_norm": 0.13801135122776031, + "learning_rate": 4.40418263891572e-06, + "loss": 0.377, + "step": 794 + }, + { + "epoch": 1.7549668874172184, + "grad_norm": 0.1327562779188156, + "learning_rate": 4.391432338645681e-06, + "loss": 0.3649, + "step": 795 + }, + { + "epoch": 1.7571743929359824, + "grad_norm": 0.13369016349315643, + "learning_rate": 4.378686054022037e-06, + "loss": 0.3882, + "step": 796 + }, + { + "epoch": 1.759381898454746, + "grad_norm": 0.14249202609062195, + "learning_rate": 4.365943869151412e-06, + "loss": 0.3763, + "step": 797 + }, + { + "epoch": 1.76158940397351, + "grad_norm": 0.1298934519290924, + "learning_rate": 4.35320586811338e-06, + "loss": 0.3742, + "step": 798 + }, + { + "epoch": 1.7637969094922736, + "grad_norm": 0.14034999907016754, + "learning_rate": 4.3404721349599044e-06, + "loss": 0.3685, + "step": 799 + }, + { + "epoch": 1.7660044150110377, + "grad_norm": 0.12729597091674805, + "learning_rate": 4.327742753714788e-06, + "loss": 0.3742, + "step": 800 + }, + { + "epoch": 1.7682119205298013, + "grad_norm": 0.13787223398685455, + "learning_rate": 4.315017808373121e-06, + "loss": 0.3718, + "step": 801 + }, + { + "epoch": 1.7704194260485653, + "grad_norm": 0.13648386299610138, + "learning_rate": 4.302297382900718e-06, + "loss": 0.3722, + "step": 802 + }, + { + "epoch": 1.7726269315673289, + "grad_norm": 0.12655648589134216, + "learning_rate": 4.289581561233574e-06, + "loss": 0.3667, + "step": 803 + }, + { + "epoch": 1.7748344370860927, + "grad_norm": 0.12654156982898712, + "learning_rate": 4.276870427277299e-06, + "loss": 0.3666, + "step": 804 + }, + { + "epoch": 1.7770419426048565, + "grad_norm": 0.15781165659427643, + "learning_rate": 4.264164064906581e-06, + "loss": 0.3787, + "step": 805 + }, + { + "epoch": 1.7792494481236203, + "grad_norm": 0.12070343643426895, + "learning_rate": 4.251462557964612e-06, + "loss": 0.3897, + "step": 806 + }, + { + "epoch": 1.781456953642384, + "grad_norm": 0.1305495798587799, + "learning_rate": 4.238765990262554e-06, + "loss": 0.3622, + "step": 807 + }, + { + "epoch": 1.783664459161148, + "grad_norm": 0.12424585223197937, + "learning_rate": 4.226074445578969e-06, + "loss": 0.3732, + "step": 808 + }, + { + "epoch": 1.7858719646799117, + "grad_norm": 0.12352015823125839, + "learning_rate": 4.213388007659281e-06, + "loss": 0.367, + "step": 809 + }, + { + "epoch": 1.7880794701986755, + "grad_norm": 0.1388520896434784, + "learning_rate": 4.200706760215219e-06, + "loss": 0.3723, + "step": 810 + }, + { + "epoch": 1.7902869757174393, + "grad_norm": 0.13200843334197998, + "learning_rate": 4.188030786924252e-06, + "loss": 0.3765, + "step": 811 + }, + { + "epoch": 1.7924944812362031, + "grad_norm": 0.11911409348249435, + "learning_rate": 4.17536017142906e-06, + "loss": 0.3694, + "step": 812 + }, + { + "epoch": 1.794701986754967, + "grad_norm": 0.1263428032398224, + "learning_rate": 4.162694997336957e-06, + "loss": 0.3656, + "step": 813 + }, + { + "epoch": 1.7969094922737305, + "grad_norm": 0.125799298286438, + "learning_rate": 4.1500353482193646e-06, + "loss": 0.3686, + "step": 814 + }, + { + "epoch": 1.7991169977924946, + "grad_norm": 0.13606040179729462, + "learning_rate": 4.137381307611236e-06, + "loss": 0.3579, + "step": 815 + }, + { + "epoch": 1.8013245033112582, + "grad_norm": 0.13091742992401123, + "learning_rate": 4.124732959010526e-06, + "loss": 0.369, + "step": 816 + }, + { + "epoch": 1.8035320088300222, + "grad_norm": 0.13721388578414917, + "learning_rate": 4.112090385877624e-06, + "loss": 0.3673, + "step": 817 + }, + { + "epoch": 1.8057395143487858, + "grad_norm": 0.12669645249843597, + "learning_rate": 4.099453671634811e-06, + "loss": 0.3778, + "step": 818 + }, + { + "epoch": 1.8079470198675498, + "grad_norm": 0.12563163042068481, + "learning_rate": 4.086822899665713e-06, + "loss": 0.3687, + "step": 819 + }, + { + "epoch": 1.8101545253863134, + "grad_norm": 0.1383393555879593, + "learning_rate": 4.07419815331474e-06, + "loss": 0.3787, + "step": 820 + }, + { + "epoch": 1.8123620309050774, + "grad_norm": 0.1277354508638382, + "learning_rate": 4.061579515886544e-06, + "loss": 0.3752, + "step": 821 + }, + { + "epoch": 1.814569536423841, + "grad_norm": 0.12579599022865295, + "learning_rate": 4.048967070645465e-06, + "loss": 0.3767, + "step": 822 + }, + { + "epoch": 1.8167770419426048, + "grad_norm": 0.1388178914785385, + "learning_rate": 4.03636090081499e-06, + "loss": 0.3795, + "step": 823 + }, + { + "epoch": 1.8189845474613686, + "grad_norm": 0.13451789319515228, + "learning_rate": 4.0237610895771895e-06, + "loss": 0.3636, + "step": 824 + }, + { + "epoch": 1.8211920529801324, + "grad_norm": 0.1298052817583084, + "learning_rate": 4.011167720072183e-06, + "loss": 0.3831, + "step": 825 + }, + { + "epoch": 1.8233995584988962, + "grad_norm": 0.13578468561172485, + "learning_rate": 3.99858087539758e-06, + "loss": 0.3787, + "step": 826 + }, + { + "epoch": 1.82560706401766, + "grad_norm": 0.1320340931415558, + "learning_rate": 3.986000638607937e-06, + "loss": 0.3697, + "step": 827 + }, + { + "epoch": 1.8278145695364238, + "grad_norm": 0.13334733247756958, + "learning_rate": 3.973427092714212e-06, + "loss": 0.361, + "step": 828 + }, + { + "epoch": 1.8300220750551877, + "grad_norm": 0.12365536391735077, + "learning_rate": 3.960860320683205e-06, + "loss": 0.3716, + "step": 829 + }, + { + "epoch": 1.8322295805739515, + "grad_norm": 0.1276884227991104, + "learning_rate": 3.948300405437026e-06, + "loss": 0.3717, + "step": 830 + }, + { + "epoch": 1.8344370860927153, + "grad_norm": 0.13205379247665405, + "learning_rate": 3.9357474298525356e-06, + "loss": 0.3632, + "step": 831 + }, + { + "epoch": 1.836644591611479, + "grad_norm": 0.12293750792741776, + "learning_rate": 3.9232014767608065e-06, + "loss": 0.3809, + "step": 832 + }, + { + "epoch": 1.8388520971302427, + "grad_norm": 0.1293838620185852, + "learning_rate": 3.910662628946573e-06, + "loss": 0.3782, + "step": 833 + }, + { + "epoch": 1.8410596026490067, + "grad_norm": 0.1325875222682953, + "learning_rate": 3.89813096914768e-06, + "loss": 0.3734, + "step": 834 + }, + { + "epoch": 1.8432671081677703, + "grad_norm": 0.12951777875423431, + "learning_rate": 3.8856065800545475e-06, + "loss": 0.3828, + "step": 835 + }, + { + "epoch": 1.8454746136865343, + "grad_norm": 0.12658190727233887, + "learning_rate": 3.873089544309616e-06, + "loss": 0.3753, + "step": 836 + }, + { + "epoch": 1.847682119205298, + "grad_norm": 0.12527629733085632, + "learning_rate": 3.860579944506811e-06, + "loss": 0.3716, + "step": 837 + }, + { + "epoch": 1.849889624724062, + "grad_norm": 0.12750259041786194, + "learning_rate": 3.848077863190985e-06, + "loss": 0.3727, + "step": 838 + }, + { + "epoch": 1.8520971302428255, + "grad_norm": 0.13881564140319824, + "learning_rate": 3.835583382857382e-06, + "loss": 0.3743, + "step": 839 + }, + { + "epoch": 1.8543046357615895, + "grad_norm": 0.12845860421657562, + "learning_rate": 3.82309658595109e-06, + "loss": 0.3692, + "step": 840 + }, + { + "epoch": 1.8565121412803531, + "grad_norm": 0.12954264879226685, + "learning_rate": 3.8106175548664994e-06, + "loss": 0.3778, + "step": 841 + }, + { + "epoch": 1.858719646799117, + "grad_norm": 0.13028046488761902, + "learning_rate": 3.7981463719467587e-06, + "loss": 0.3825, + "step": 842 + }, + { + "epoch": 1.8609271523178808, + "grad_norm": 0.1258542686700821, + "learning_rate": 3.7856831194832262e-06, + "loss": 0.3817, + "step": 843 + }, + { + "epoch": 1.8631346578366446, + "grad_norm": 0.1192869246006012, + "learning_rate": 3.7732278797149324e-06, + "loss": 0.3706, + "step": 844 + }, + { + "epoch": 1.8653421633554084, + "grad_norm": 0.13153821229934692, + "learning_rate": 3.7607807348280364e-06, + "loss": 0.3626, + "step": 845 + }, + { + "epoch": 1.8675496688741722, + "grad_norm": 0.12072479724884033, + "learning_rate": 3.7483417669552857e-06, + "loss": 0.3557, + "step": 846 + }, + { + "epoch": 1.869757174392936, + "grad_norm": 0.12467250972986221, + "learning_rate": 3.735911058175467e-06, + "loss": 0.3621, + "step": 847 + }, + { + "epoch": 1.8719646799116998, + "grad_norm": 0.12176632136106491, + "learning_rate": 3.7234886905128724e-06, + "loss": 0.3726, + "step": 848 + }, + { + "epoch": 1.8741721854304636, + "grad_norm": 0.1291275918483734, + "learning_rate": 3.711074745936751e-06, + "loss": 0.3663, + "step": 849 + }, + { + "epoch": 1.8763796909492274, + "grad_norm": 0.1401110738515854, + "learning_rate": 3.698669306360778e-06, + "loss": 0.3763, + "step": 850 + }, + { + "epoch": 1.8785871964679912, + "grad_norm": 0.12075633555650711, + "learning_rate": 3.6862724536425042e-06, + "loss": 0.3761, + "step": 851 + }, + { + "epoch": 1.8807947019867548, + "grad_norm": 0.12913060188293457, + "learning_rate": 3.67388426958282e-06, + "loss": 0.3735, + "step": 852 + }, + { + "epoch": 1.8830022075055188, + "grad_norm": 0.13538530468940735, + "learning_rate": 3.6615048359254155e-06, + "loss": 0.3737, + "step": 853 + }, + { + "epoch": 1.8852097130242824, + "grad_norm": 0.1396217793226242, + "learning_rate": 3.649134234356242e-06, + "loss": 0.3765, + "step": 854 + }, + { + "epoch": 1.8874172185430464, + "grad_norm": 0.12726326286792755, + "learning_rate": 3.636772546502973e-06, + "loss": 0.3706, + "step": 855 + }, + { + "epoch": 1.88962472406181, + "grad_norm": 0.12331200391054153, + "learning_rate": 3.624419853934461e-06, + "loss": 0.3698, + "step": 856 + }, + { + "epoch": 1.891832229580574, + "grad_norm": 0.12955856323242188, + "learning_rate": 3.612076238160209e-06, + "loss": 0.3779, + "step": 857 + }, + { + "epoch": 1.8940397350993377, + "grad_norm": 0.1281704604625702, + "learning_rate": 3.5997417806298172e-06, + "loss": 0.3711, + "step": 858 + }, + { + "epoch": 1.8962472406181017, + "grad_norm": 0.13094061613082886, + "learning_rate": 3.5874165627324665e-06, + "loss": 0.3706, + "step": 859 + }, + { + "epoch": 1.8984547461368653, + "grad_norm": 0.1244734600186348, + "learning_rate": 3.575100665796362e-06, + "loss": 0.3701, + "step": 860 + }, + { + "epoch": 1.9006622516556293, + "grad_norm": 0.13025392591953278, + "learning_rate": 3.5627941710882042e-06, + "loss": 0.3774, + "step": 861 + }, + { + "epoch": 1.9028697571743929, + "grad_norm": 0.13566212356090546, + "learning_rate": 3.550497159812658e-06, + "loss": 0.376, + "step": 862 + }, + { + "epoch": 1.9050772626931567, + "grad_norm": 0.13212575018405914, + "learning_rate": 3.5382097131118052e-06, + "loss": 0.3759, + "step": 863 + }, + { + "epoch": 1.9072847682119205, + "grad_norm": 0.13214178383350372, + "learning_rate": 3.525931912064623e-06, + "loss": 0.3706, + "step": 864 + }, + { + "epoch": 1.9094922737306843, + "grad_norm": 0.1290368288755417, + "learning_rate": 3.5136638376864333e-06, + "loss": 0.3767, + "step": 865 + }, + { + "epoch": 1.9116997792494481, + "grad_norm": 0.1320725828409195, + "learning_rate": 3.5014055709283847e-06, + "loss": 0.3704, + "step": 866 + }, + { + "epoch": 1.913907284768212, + "grad_norm": 0.13395248353481293, + "learning_rate": 3.4891571926769e-06, + "loss": 0.3774, + "step": 867 + }, + { + "epoch": 1.9161147902869757, + "grad_norm": 0.12245456129312515, + "learning_rate": 3.4769187837531637e-06, + "loss": 0.3872, + "step": 868 + }, + { + "epoch": 1.9183222958057395, + "grad_norm": 0.1327778398990631, + "learning_rate": 3.4646904249125713e-06, + "loss": 0.3666, + "step": 869 + }, + { + "epoch": 1.9205298013245033, + "grad_norm": 0.13339614868164062, + "learning_rate": 3.4524721968442017e-06, + "loss": 0.3782, + "step": 870 + }, + { + "epoch": 1.9227373068432672, + "grad_norm": 0.12259134650230408, + "learning_rate": 3.4402641801702883e-06, + "loss": 0.38, + "step": 871 + }, + { + "epoch": 1.924944812362031, + "grad_norm": 0.12124523520469666, + "learning_rate": 3.4280664554456823e-06, + "loss": 0.3746, + "step": 872 + }, + { + "epoch": 1.9271523178807946, + "grad_norm": 0.12291016429662704, + "learning_rate": 3.415879103157327e-06, + "loss": 0.3697, + "step": 873 + }, + { + "epoch": 1.9293598233995586, + "grad_norm": 0.12792356312274933, + "learning_rate": 3.403702203723718e-06, + "loss": 0.3755, + "step": 874 + }, + { + "epoch": 1.9315673289183222, + "grad_norm": 0.1225101426243782, + "learning_rate": 3.3915358374943813e-06, + "loss": 0.3821, + "step": 875 + }, + { + "epoch": 1.9337748344370862, + "grad_norm": 0.1265992522239685, + "learning_rate": 3.379380084749336e-06, + "loss": 0.3672, + "step": 876 + }, + { + "epoch": 1.9359823399558498, + "grad_norm": 0.12633873522281647, + "learning_rate": 3.3672350256985727e-06, + "loss": 0.3766, + "step": 877 + }, + { + "epoch": 1.9381898454746138, + "grad_norm": 0.1319660246372223, + "learning_rate": 3.3551007404815162e-06, + "loss": 0.374, + "step": 878 + }, + { + "epoch": 1.9403973509933774, + "grad_norm": 0.12871594727039337, + "learning_rate": 3.3429773091664985e-06, + "loss": 0.3673, + "step": 879 + }, + { + "epoch": 1.9426048565121414, + "grad_norm": 0.13455092906951904, + "learning_rate": 3.330864811750235e-06, + "loss": 0.3805, + "step": 880 + }, + { + "epoch": 1.944812362030905, + "grad_norm": 0.13253872096538544, + "learning_rate": 3.3187633281572913e-06, + "loss": 0.3695, + "step": 881 + }, + { + "epoch": 1.9470198675496688, + "grad_norm": 0.11975626647472382, + "learning_rate": 3.306672938239561e-06, + "loss": 0.3767, + "step": 882 + }, + { + "epoch": 1.9492273730684326, + "grad_norm": 0.12552706897258759, + "learning_rate": 3.2945937217757324e-06, + "loss": 0.363, + "step": 883 + }, + { + "epoch": 1.9514348785871964, + "grad_norm": 0.13104894757270813, + "learning_rate": 3.282525758470768e-06, + "loss": 0.3699, + "step": 884 + }, + { + "epoch": 1.9536423841059603, + "grad_norm": 0.12783929705619812, + "learning_rate": 3.2704691279553725e-06, + "loss": 0.3681, + "step": 885 + }, + { + "epoch": 1.955849889624724, + "grad_norm": 0.12647448480129242, + "learning_rate": 3.2584239097854763e-06, + "loss": 0.3873, + "step": 886 + }, + { + "epoch": 1.9580573951434879, + "grad_norm": 0.11794891208410263, + "learning_rate": 3.246390183441703e-06, + "loss": 0.3689, + "step": 887 + }, + { + "epoch": 1.9602649006622517, + "grad_norm": 0.1255064755678177, + "learning_rate": 3.234368028328845e-06, + "loss": 0.3874, + "step": 888 + }, + { + "epoch": 1.9624724061810155, + "grad_norm": 0.12461333721876144, + "learning_rate": 3.2223575237753448e-06, + "loss": 0.3719, + "step": 889 + }, + { + "epoch": 1.9646799116997793, + "grad_norm": 0.13075792789459229, + "learning_rate": 3.2103587490327624e-06, + "loss": 0.3746, + "step": 890 + }, + { + "epoch": 1.966887417218543, + "grad_norm": 0.13413941860198975, + "learning_rate": 3.198371783275269e-06, + "loss": 0.3723, + "step": 891 + }, + { + "epoch": 1.9690949227373067, + "grad_norm": 0.13534380495548248, + "learning_rate": 3.186396705599106e-06, + "loss": 0.3762, + "step": 892 + }, + { + "epoch": 1.9713024282560707, + "grad_norm": 0.12702573835849762, + "learning_rate": 3.1744335950220716e-06, + "loss": 0.3763, + "step": 893 + }, + { + "epoch": 1.9735099337748343, + "grad_norm": 0.131902277469635, + "learning_rate": 3.162482530483e-06, + "loss": 0.3647, + "step": 894 + }, + { + "epoch": 1.9757174392935983, + "grad_norm": 0.12826719880104065, + "learning_rate": 3.1505435908412415e-06, + "loss": 0.3705, + "step": 895 + }, + { + "epoch": 1.977924944812362, + "grad_norm": 0.1235160231590271, + "learning_rate": 3.1386168548761377e-06, + "loss": 0.3806, + "step": 896 + }, + { + "epoch": 1.980132450331126, + "grad_norm": 0.13026514649391174, + "learning_rate": 3.126702401286503e-06, + "loss": 0.3714, + "step": 897 + }, + { + "epoch": 1.9823399558498895, + "grad_norm": 0.12384995073080063, + "learning_rate": 3.11480030869011e-06, + "loss": 0.3775, + "step": 898 + }, + { + "epoch": 1.9845474613686536, + "grad_norm": 0.11622393131256104, + "learning_rate": 3.1029106556231615e-06, + "loss": 0.3685, + "step": 899 + }, + { + "epoch": 1.9867549668874172, + "grad_norm": 0.12467402219772339, + "learning_rate": 3.0910335205397834e-06, + "loss": 0.3753, + "step": 900 + }, + { + "epoch": 1.9889624724061812, + "grad_norm": 0.12134755402803421, + "learning_rate": 3.0791689818114988e-06, + "loss": 0.3763, + "step": 901 + }, + { + "epoch": 1.9911699779249448, + "grad_norm": 0.118636354804039, + "learning_rate": 3.067317117726715e-06, + "loss": 0.3779, + "step": 902 + }, + { + "epoch": 1.9933774834437086, + "grad_norm": 0.12621286511421204, + "learning_rate": 3.055478006490199e-06, + "loss": 0.3755, + "step": 903 + }, + { + "epoch": 1.9955849889624724, + "grad_norm": 0.13427864015102386, + "learning_rate": 3.043651726222578e-06, + "loss": 0.3677, + "step": 904 + }, + { + "epoch": 1.9977924944812362, + "grad_norm": 0.1349465250968933, + "learning_rate": 3.0318383549598085e-06, + "loss": 0.3821, + "step": 905 + }, + { + "epoch": 2.0, + "grad_norm": 0.1383267641067505, + "learning_rate": 3.020037970652664e-06, + "loss": 0.3605, + "step": 906 + }, + { + "epoch": 2.0022075055187636, + "grad_norm": 0.1400349736213684, + "learning_rate": 3.00825065116623e-06, + "loss": 0.3587, + "step": 907 + }, + { + "epoch": 2.0044150110375276, + "grad_norm": 0.12129946798086166, + "learning_rate": 2.996476474279375e-06, + "loss": 0.3564, + "step": 908 + }, + { + "epoch": 2.006622516556291, + "grad_norm": 0.13566596806049347, + "learning_rate": 2.9847155176842547e-06, + "loss": 0.3628, + "step": 909 + }, + { + "epoch": 2.0088300220750552, + "grad_norm": 0.12823748588562012, + "learning_rate": 2.9729678589857852e-06, + "loss": 0.3589, + "step": 910 + }, + { + "epoch": 2.011037527593819, + "grad_norm": 0.1251477748155594, + "learning_rate": 2.9612335757011362e-06, + "loss": 0.3583, + "step": 911 + }, + { + "epoch": 2.013245033112583, + "grad_norm": 0.1308368593454361, + "learning_rate": 2.94951274525922e-06, + "loss": 0.354, + "step": 912 + }, + { + "epoch": 2.0154525386313464, + "grad_norm": 0.1357547789812088, + "learning_rate": 2.9378054450001836e-06, + "loss": 0.3664, + "step": 913 + }, + { + "epoch": 2.0176600441501105, + "grad_norm": 0.13018201291561127, + "learning_rate": 2.9261117521748904e-06, + "loss": 0.3627, + "step": 914 + }, + { + "epoch": 2.019867549668874, + "grad_norm": 0.12550939619541168, + "learning_rate": 2.914431743944414e-06, + "loss": 0.3518, + "step": 915 + }, + { + "epoch": 2.022075055187638, + "grad_norm": 0.12784212827682495, + "learning_rate": 2.902765497379534e-06, + "loss": 0.3547, + "step": 916 + }, + { + "epoch": 2.0242825607064017, + "grad_norm": 0.13502328097820282, + "learning_rate": 2.8911130894602198e-06, + "loss": 0.3648, + "step": 917 + }, + { + "epoch": 2.0264900662251657, + "grad_norm": 0.12056614458560944, + "learning_rate": 2.8794745970751308e-06, + "loss": 0.3612, + "step": 918 + }, + { + "epoch": 2.0286975717439293, + "grad_norm": 0.12976831197738647, + "learning_rate": 2.8678500970210977e-06, + "loss": 0.3619, + "step": 919 + }, + { + "epoch": 2.0309050772626933, + "grad_norm": 0.13974756002426147, + "learning_rate": 2.8562396660026304e-06, + "loss": 0.3585, + "step": 920 + }, + { + "epoch": 2.033112582781457, + "grad_norm": 0.1298479437828064, + "learning_rate": 2.8446433806313966e-06, + "loss": 0.3533, + "step": 921 + }, + { + "epoch": 2.035320088300221, + "grad_norm": 0.13592906296253204, + "learning_rate": 2.83306131742573e-06, + "loss": 0.36, + "step": 922 + }, + { + "epoch": 2.0375275938189845, + "grad_norm": 0.13170795142650604, + "learning_rate": 2.821493552810114e-06, + "loss": 0.3535, + "step": 923 + }, + { + "epoch": 2.0397350993377485, + "grad_norm": 0.14562806487083435, + "learning_rate": 2.809940163114685e-06, + "loss": 0.3737, + "step": 924 + }, + { + "epoch": 2.041942604856512, + "grad_norm": 0.12969909608364105, + "learning_rate": 2.7984012245747237e-06, + "loss": 0.3597, + "step": 925 + }, + { + "epoch": 2.044150110375276, + "grad_norm": 0.13986288011074066, + "learning_rate": 2.786876813330158e-06, + "loss": 0.3608, + "step": 926 + }, + { + "epoch": 2.0463576158940397, + "grad_norm": 0.13949252665042877, + "learning_rate": 2.7753670054250583e-06, + "loss": 0.3507, + "step": 927 + }, + { + "epoch": 2.0485651214128033, + "grad_norm": 0.12819348275661469, + "learning_rate": 2.763871876807129e-06, + "loss": 0.356, + "step": 928 + }, + { + "epoch": 2.0507726269315674, + "grad_norm": 0.13626167178153992, + "learning_rate": 2.7523915033272163e-06, + "loss": 0.3534, + "step": 929 + }, + { + "epoch": 2.052980132450331, + "grad_norm": 0.14214330911636353, + "learning_rate": 2.740925960738802e-06, + "loss": 0.3627, + "step": 930 + }, + { + "epoch": 2.055187637969095, + "grad_norm": 0.1371086835861206, + "learning_rate": 2.72947532469751e-06, + "loss": 0.3594, + "step": 931 + }, + { + "epoch": 2.0573951434878586, + "grad_norm": 0.15132765471935272, + "learning_rate": 2.7180396707606023e-06, + "loss": 0.3487, + "step": 932 + }, + { + "epoch": 2.0596026490066226, + "grad_norm": 0.1387225240468979, + "learning_rate": 2.7066190743864774e-06, + "loss": 0.3625, + "step": 933 + }, + { + "epoch": 2.061810154525386, + "grad_norm": 0.13822317123413086, + "learning_rate": 2.695213610934179e-06, + "loss": 0.3542, + "step": 934 + }, + { + "epoch": 2.06401766004415, + "grad_norm": 0.1481810212135315, + "learning_rate": 2.6838233556628932e-06, + "loss": 0.3543, + "step": 935 + }, + { + "epoch": 2.066225165562914, + "grad_norm": 0.1475333869457245, + "learning_rate": 2.672448383731461e-06, + "loss": 0.3587, + "step": 936 + }, + { + "epoch": 2.068432671081678, + "grad_norm": 0.13259293138980865, + "learning_rate": 2.661088770197868e-06, + "loss": 0.3531, + "step": 937 + }, + { + "epoch": 2.0706401766004414, + "grad_norm": 0.129195436835289, + "learning_rate": 2.6497445900187635e-06, + "loss": 0.3557, + "step": 938 + }, + { + "epoch": 2.0728476821192054, + "grad_norm": 0.12847961485385895, + "learning_rate": 2.6384159180489542e-06, + "loss": 0.3477, + "step": 939 + }, + { + "epoch": 2.075055187637969, + "grad_norm": 0.13196007907390594, + "learning_rate": 2.6271028290409216e-06, + "loss": 0.3584, + "step": 940 + }, + { + "epoch": 2.077262693156733, + "grad_norm": 0.1364007443189621, + "learning_rate": 2.6158053976443164e-06, + "loss": 0.3601, + "step": 941 + }, + { + "epoch": 2.0794701986754967, + "grad_norm": 0.1247749775648117, + "learning_rate": 2.604523698405477e-06, + "loss": 0.3647, + "step": 942 + }, + { + "epoch": 2.0816777041942607, + "grad_norm": 0.12793247401714325, + "learning_rate": 2.5932578057669273e-06, + "loss": 0.3656, + "step": 943 + }, + { + "epoch": 2.0838852097130243, + "grad_norm": 0.12500843405723572, + "learning_rate": 2.5820077940668975e-06, + "loss": 0.3511, + "step": 944 + }, + { + "epoch": 2.0860927152317883, + "grad_norm": 0.13672660291194916, + "learning_rate": 2.570773737538825e-06, + "loss": 0.3709, + "step": 945 + }, + { + "epoch": 2.088300220750552, + "grad_norm": 0.1374640017747879, + "learning_rate": 2.559555710310866e-06, + "loss": 0.3528, + "step": 946 + }, + { + "epoch": 2.0905077262693155, + "grad_norm": 0.13424260914325714, + "learning_rate": 2.5483537864054055e-06, + "loss": 0.354, + "step": 947 + }, + { + "epoch": 2.0927152317880795, + "grad_norm": 0.13996686041355133, + "learning_rate": 2.537168039738569e-06, + "loss": 0.3655, + "step": 948 + }, + { + "epoch": 2.094922737306843, + "grad_norm": 0.12471663951873779, + "learning_rate": 2.525998544119743e-06, + "loss": 0.3606, + "step": 949 + }, + { + "epoch": 2.097130242825607, + "grad_norm": 0.13321642577648163, + "learning_rate": 2.514845373251078e-06, + "loss": 0.3477, + "step": 950 + }, + { + "epoch": 2.0993377483443707, + "grad_norm": 0.1407744139432907, + "learning_rate": 2.5037086007270017e-06, + "loss": 0.3566, + "step": 951 + }, + { + "epoch": 2.1015452538631347, + "grad_norm": 0.1319989413022995, + "learning_rate": 2.4925883000337407e-06, + "loss": 0.3511, + "step": 952 + }, + { + "epoch": 2.1037527593818983, + "grad_norm": 0.12347118556499481, + "learning_rate": 2.4814845445488288e-06, + "loss": 0.3609, + "step": 953 + }, + { + "epoch": 2.1059602649006623, + "grad_norm": 0.15142813324928284, + "learning_rate": 2.4703974075406316e-06, + "loss": 0.3581, + "step": 954 + }, + { + "epoch": 2.108167770419426, + "grad_norm": 0.12996384501457214, + "learning_rate": 2.4593269621678502e-06, + "loss": 0.3586, + "step": 955 + }, + { + "epoch": 2.11037527593819, + "grad_norm": 0.12407524138689041, + "learning_rate": 2.448273281479052e-06, + "loss": 0.3561, + "step": 956 + }, + { + "epoch": 2.1125827814569536, + "grad_norm": 0.12946733832359314, + "learning_rate": 2.437236438412177e-06, + "loss": 0.3637, + "step": 957 + }, + { + "epoch": 2.1147902869757176, + "grad_norm": 0.14091919362545013, + "learning_rate": 2.426216505794066e-06, + "loss": 0.3572, + "step": 958 + }, + { + "epoch": 2.116997792494481, + "grad_norm": 0.13770008087158203, + "learning_rate": 2.4152135563399743e-06, + "loss": 0.3617, + "step": 959 + }, + { + "epoch": 2.119205298013245, + "grad_norm": 0.1215134859085083, + "learning_rate": 2.404227662653089e-06, + "loss": 0.3613, + "step": 960 + }, + { + "epoch": 2.121412803532009, + "grad_norm": 0.13981877267360687, + "learning_rate": 2.3932588972240642e-06, + "loss": 0.3601, + "step": 961 + }, + { + "epoch": 2.123620309050773, + "grad_norm": 0.13256755471229553, + "learning_rate": 2.3823073324305222e-06, + "loss": 0.3551, + "step": 962 + }, + { + "epoch": 2.1258278145695364, + "grad_norm": 0.14095091819763184, + "learning_rate": 2.3713730405365965e-06, + "loss": 0.3671, + "step": 963 + }, + { + "epoch": 2.1280353200883004, + "grad_norm": 0.11968285590410233, + "learning_rate": 2.3604560936924363e-06, + "loss": 0.3558, + "step": 964 + }, + { + "epoch": 2.130242825607064, + "grad_norm": 0.13734926283359528, + "learning_rate": 2.3495565639337446e-06, + "loss": 0.3653, + "step": 965 + }, + { + "epoch": 2.1324503311258276, + "grad_norm": 0.13910116255283356, + "learning_rate": 2.3386745231812923e-06, + "loss": 0.3559, + "step": 966 + }, + { + "epoch": 2.1346578366445916, + "grad_norm": 0.13142208755016327, + "learning_rate": 2.3278100432404554e-06, + "loss": 0.3571, + "step": 967 + }, + { + "epoch": 2.136865342163355, + "grad_norm": 0.12986791133880615, + "learning_rate": 2.3169631958007316e-06, + "loss": 0.3613, + "step": 968 + }, + { + "epoch": 2.1390728476821192, + "grad_norm": 0.12400130182504654, + "learning_rate": 2.3061340524352687e-06, + "loss": 0.3596, + "step": 969 + }, + { + "epoch": 2.141280353200883, + "grad_norm": 0.12488456070423126, + "learning_rate": 2.295322684600395e-06, + "loss": 0.3571, + "step": 970 + }, + { + "epoch": 2.143487858719647, + "grad_norm": 0.13525961339473724, + "learning_rate": 2.284529163635144e-06, + "loss": 0.352, + "step": 971 + }, + { + "epoch": 2.1456953642384105, + "grad_norm": 0.14086809754371643, + "learning_rate": 2.273753560760793e-06, + "loss": 0.3529, + "step": 972 + }, + { + "epoch": 2.1479028697571745, + "grad_norm": 0.13002382218837738, + "learning_rate": 2.262995947080378e-06, + "loss": 0.3628, + "step": 973 + }, + { + "epoch": 2.150110375275938, + "grad_norm": 0.13714134693145752, + "learning_rate": 2.252256393578241e-06, + "loss": 0.3602, + "step": 974 + }, + { + "epoch": 2.152317880794702, + "grad_norm": 0.1353500783443451, + "learning_rate": 2.2415349711195455e-06, + "loss": 0.3562, + "step": 975 + }, + { + "epoch": 2.1545253863134657, + "grad_norm": 0.14100465178489685, + "learning_rate": 2.230831750449825e-06, + "loss": 0.3562, + "step": 976 + }, + { + "epoch": 2.1567328918322297, + "grad_norm": 0.11746090650558472, + "learning_rate": 2.2201468021945024e-06, + "loss": 0.3679, + "step": 977 + }, + { + "epoch": 2.1589403973509933, + "grad_norm": 0.13000161945819855, + "learning_rate": 2.209480196858429e-06, + "loss": 0.3579, + "step": 978 + }, + { + "epoch": 2.1611479028697573, + "grad_norm": 0.14291274547576904, + "learning_rate": 2.198832004825427e-06, + "loss": 0.366, + "step": 979 + }, + { + "epoch": 2.163355408388521, + "grad_norm": 0.1327233463525772, + "learning_rate": 2.1882022963578087e-06, + "loss": 0.3576, + "step": 980 + }, + { + "epoch": 2.165562913907285, + "grad_norm": 0.12592221796512604, + "learning_rate": 2.177591141595931e-06, + "loss": 0.3554, + "step": 981 + }, + { + "epoch": 2.1677704194260485, + "grad_norm": 0.13044218719005585, + "learning_rate": 2.166998610557718e-06, + "loss": 0.3489, + "step": 982 + }, + { + "epoch": 2.1699779249448126, + "grad_norm": 0.14413036406040192, + "learning_rate": 2.1564247731382063e-06, + "loss": 0.3653, + "step": 983 + }, + { + "epoch": 2.172185430463576, + "grad_norm": 0.12830859422683716, + "learning_rate": 2.145869699109081e-06, + "loss": 0.3566, + "step": 984 + }, + { + "epoch": 2.1743929359823397, + "grad_norm": 0.12629762291908264, + "learning_rate": 2.1353334581182193e-06, + "loss": 0.3581, + "step": 985 + }, + { + "epoch": 2.1766004415011038, + "grad_norm": 0.12835480272769928, + "learning_rate": 2.1248161196892295e-06, + "loss": 0.3655, + "step": 986 + }, + { + "epoch": 2.1788079470198674, + "grad_norm": 0.12383104115724564, + "learning_rate": 2.1143177532209855e-06, + "loss": 0.3564, + "step": 987 + }, + { + "epoch": 2.1810154525386314, + "grad_norm": 0.1490948498249054, + "learning_rate": 2.1038384279871786e-06, + "loss": 0.37, + "step": 988 + }, + { + "epoch": 2.183222958057395, + "grad_norm": 0.12230035662651062, + "learning_rate": 2.0933782131358516e-06, + "loss": 0.3549, + "step": 989 + }, + { + "epoch": 2.185430463576159, + "grad_norm": 0.13511480391025543, + "learning_rate": 2.082937177688952e-06, + "loss": 0.3576, + "step": 990 + }, + { + "epoch": 2.1876379690949226, + "grad_norm": 0.135064497590065, + "learning_rate": 2.0725153905418726e-06, + "loss": 0.3544, + "step": 991 + }, + { + "epoch": 2.1898454746136866, + "grad_norm": 0.12062691897153854, + "learning_rate": 2.0621129204629907e-06, + "loss": 0.3523, + "step": 992 + }, + { + "epoch": 2.19205298013245, + "grad_norm": 0.14147672057151794, + "learning_rate": 2.0517298360932202e-06, + "loss": 0.3583, + "step": 993 + }, + { + "epoch": 2.1942604856512142, + "grad_norm": 0.12477768212556839, + "learning_rate": 2.041366205945563e-06, + "loss": 0.3604, + "step": 994 + }, + { + "epoch": 2.196467991169978, + "grad_norm": 0.14072567224502563, + "learning_rate": 2.0310220984046467e-06, + "loss": 0.3693, + "step": 995 + }, + { + "epoch": 2.198675496688742, + "grad_norm": 0.13882459700107574, + "learning_rate": 2.020697581726279e-06, + "loss": 0.3604, + "step": 996 + }, + { + "epoch": 2.2008830022075054, + "grad_norm": 0.12994110584259033, + "learning_rate": 2.010392724037002e-06, + "loss": 0.3589, + "step": 997 + }, + { + "epoch": 2.2030905077262695, + "grad_norm": 0.13540343940258026, + "learning_rate": 2.0001075933336302e-06, + "loss": 0.3664, + "step": 998 + }, + { + "epoch": 2.205298013245033, + "grad_norm": 0.12710383534431458, + "learning_rate": 1.9898422574828163e-06, + "loss": 0.3606, + "step": 999 + }, + { + "epoch": 2.207505518763797, + "grad_norm": 0.13436554372310638, + "learning_rate": 1.979596784220591e-06, + "loss": 0.3583, + "step": 1000 + }, + { + "epoch": 2.2097130242825607, + "grad_norm": 0.1312997192144394, + "learning_rate": 1.969371241151923e-06, + "loss": 0.3485, + "step": 1001 + }, + { + "epoch": 2.2119205298013247, + "grad_norm": 0.12719684839248657, + "learning_rate": 1.9591656957502696e-06, + "loss": 0.3646, + "step": 1002 + }, + { + "epoch": 2.2141280353200883, + "grad_norm": 0.11936060339212418, + "learning_rate": 1.9489802153571373e-06, + "loss": 0.3578, + "step": 1003 + }, + { + "epoch": 2.216335540838852, + "grad_norm": 0.11757774651050568, + "learning_rate": 1.938814867181632e-06, + "loss": 0.3633, + "step": 1004 + }, + { + "epoch": 2.218543046357616, + "grad_norm": 0.12061762064695358, + "learning_rate": 1.928669718300013e-06, + "loss": 0.3539, + "step": 1005 + }, + { + "epoch": 2.2207505518763795, + "grad_norm": 0.11802493780851364, + "learning_rate": 1.9185448356552575e-06, + "loss": 0.3519, + "step": 1006 + }, + { + "epoch": 2.2229580573951435, + "grad_norm": 0.12381359189748764, + "learning_rate": 1.9084402860566128e-06, + "loss": 0.3528, + "step": 1007 + }, + { + "epoch": 2.225165562913907, + "grad_norm": 0.12773284316062927, + "learning_rate": 1.8983561361791608e-06, + "loss": 0.3618, + "step": 1008 + }, + { + "epoch": 2.227373068432671, + "grad_norm": 0.1377752125263214, + "learning_rate": 1.8882924525633778e-06, + "loss": 0.36, + "step": 1009 + }, + { + "epoch": 2.2295805739514347, + "grad_norm": 0.13043536245822906, + "learning_rate": 1.8782493016146868e-06, + "loss": 0.3588, + "step": 1010 + }, + { + "epoch": 2.2317880794701987, + "grad_norm": 0.12282190471887589, + "learning_rate": 1.8682267496030276e-06, + "loss": 0.3661, + "step": 1011 + }, + { + "epoch": 2.2339955849889623, + "grad_norm": 0.11887135356664658, + "learning_rate": 1.8582248626624217e-06, + "loss": 0.3585, + "step": 1012 + }, + { + "epoch": 2.2362030905077264, + "grad_norm": 0.13011059165000916, + "learning_rate": 1.8482437067905268e-06, + "loss": 0.3648, + "step": 1013 + }, + { + "epoch": 2.23841059602649, + "grad_norm": 0.12632228434085846, + "learning_rate": 1.8382833478482066e-06, + "loss": 0.3711, + "step": 1014 + }, + { + "epoch": 2.240618101545254, + "grad_norm": 0.13252626359462738, + "learning_rate": 1.8283438515590996e-06, + "loss": 0.3691, + "step": 1015 + }, + { + "epoch": 2.2428256070640176, + "grad_norm": 0.12210851162672043, + "learning_rate": 1.8184252835091764e-06, + "loss": 0.36, + "step": 1016 + }, + { + "epoch": 2.2450331125827816, + "grad_norm": 0.1259019374847412, + "learning_rate": 1.8085277091463188e-06, + "loss": 0.3507, + "step": 1017 + }, + { + "epoch": 2.247240618101545, + "grad_norm": 0.11658161133527756, + "learning_rate": 1.798651193779875e-06, + "loss": 0.3512, + "step": 1018 + }, + { + "epoch": 2.249448123620309, + "grad_norm": 0.12686073780059814, + "learning_rate": 1.788795802580236e-06, + "loss": 0.3554, + "step": 1019 + }, + { + "epoch": 2.251655629139073, + "grad_norm": 0.13531404733657837, + "learning_rate": 1.7789616005784077e-06, + "loss": 0.3569, + "step": 1020 + }, + { + "epoch": 2.253863134657837, + "grad_norm": 0.12338366359472275, + "learning_rate": 1.7691486526655782e-06, + "loss": 0.361, + "step": 1021 + }, + { + "epoch": 2.2560706401766004, + "grad_norm": 0.1204574853181839, + "learning_rate": 1.7593570235926883e-06, + "loss": 0.3665, + "step": 1022 + }, + { + "epoch": 2.258278145695364, + "grad_norm": 0.12831008434295654, + "learning_rate": 1.7495867779700053e-06, + "loss": 0.3617, + "step": 1023 + }, + { + "epoch": 2.260485651214128, + "grad_norm": 0.1243155300617218, + "learning_rate": 1.7398379802666993e-06, + "loss": 0.3485, + "step": 1024 + }, + { + "epoch": 2.262693156732892, + "grad_norm": 0.12370171397924423, + "learning_rate": 1.7301106948104123e-06, + "loss": 0.363, + "step": 1025 + }, + { + "epoch": 2.2649006622516556, + "grad_norm": 0.12362192571163177, + "learning_rate": 1.7204049857868433e-06, + "loss": 0.3553, + "step": 1026 + }, + { + "epoch": 2.2671081677704192, + "grad_norm": 0.1303797960281372, + "learning_rate": 1.7107209172393158e-06, + "loss": 0.3582, + "step": 1027 + }, + { + "epoch": 2.2693156732891833, + "grad_norm": 0.12913931906223297, + "learning_rate": 1.701058553068357e-06, + "loss": 0.3621, + "step": 1028 + }, + { + "epoch": 2.271523178807947, + "grad_norm": 0.13231627643108368, + "learning_rate": 1.6914179570312767e-06, + "loss": 0.3529, + "step": 1029 + }, + { + "epoch": 2.273730684326711, + "grad_norm": 0.12977519631385803, + "learning_rate": 1.6817991927417516e-06, + "loss": 0.3619, + "step": 1030 + }, + { + "epoch": 2.2759381898454745, + "grad_norm": 0.12845906615257263, + "learning_rate": 1.672202323669397e-06, + "loss": 0.3613, + "step": 1031 + }, + { + "epoch": 2.2781456953642385, + "grad_norm": 0.1263485550880432, + "learning_rate": 1.662627413139351e-06, + "loss": 0.3608, + "step": 1032 + }, + { + "epoch": 2.280353200883002, + "grad_norm": 0.12105315178632736, + "learning_rate": 1.6530745243318646e-06, + "loss": 0.3694, + "step": 1033 + }, + { + "epoch": 2.282560706401766, + "grad_norm": 0.12585538625717163, + "learning_rate": 1.64354372028187e-06, + "loss": 0.3674, + "step": 1034 + }, + { + "epoch": 2.2847682119205297, + "grad_norm": 0.13143764436244965, + "learning_rate": 1.634035063878579e-06, + "loss": 0.3654, + "step": 1035 + }, + { + "epoch": 2.2869757174392937, + "grad_norm": 0.1330743134021759, + "learning_rate": 1.6245486178650582e-06, + "loss": 0.3579, + "step": 1036 + }, + { + "epoch": 2.2891832229580573, + "grad_norm": 0.12009730935096741, + "learning_rate": 1.6150844448378178e-06, + "loss": 0.3531, + "step": 1037 + }, + { + "epoch": 2.2913907284768213, + "grad_norm": 0.13864392042160034, + "learning_rate": 1.6056426072464015e-06, + "loss": 0.3628, + "step": 1038 + }, + { + "epoch": 2.293598233995585, + "grad_norm": 0.13580797612667084, + "learning_rate": 1.5962231673929735e-06, + "loss": 0.3722, + "step": 1039 + }, + { + "epoch": 2.295805739514349, + "grad_norm": 0.12375160306692123, + "learning_rate": 1.586826187431902e-06, + "loss": 0.364, + "step": 1040 + }, + { + "epoch": 2.2980132450331126, + "grad_norm": 0.12487831711769104, + "learning_rate": 1.5774517293693558e-06, + "loss": 0.3592, + "step": 1041 + }, + { + "epoch": 2.300220750551876, + "grad_norm": 0.14687076210975647, + "learning_rate": 1.5680998550628912e-06, + "loss": 0.3506, + "step": 1042 + }, + { + "epoch": 2.30242825607064, + "grad_norm": 0.12587517499923706, + "learning_rate": 1.558770626221044e-06, + "loss": 0.3577, + "step": 1043 + }, + { + "epoch": 2.304635761589404, + "grad_norm": 0.13661305606365204, + "learning_rate": 1.5494641044029268e-06, + "loss": 0.3516, + "step": 1044 + }, + { + "epoch": 2.306843267108168, + "grad_norm": 0.1377173811197281, + "learning_rate": 1.5401803510178197e-06, + "loss": 0.3603, + "step": 1045 + }, + { + "epoch": 2.3090507726269314, + "grad_norm": 0.13497798144817352, + "learning_rate": 1.5309194273247612e-06, + "loss": 0.3545, + "step": 1046 + }, + { + "epoch": 2.3112582781456954, + "grad_norm": 0.12721391022205353, + "learning_rate": 1.5216813944321473e-06, + "loss": 0.3447, + "step": 1047 + }, + { + "epoch": 2.313465783664459, + "grad_norm": 0.12541592121124268, + "learning_rate": 1.5124663132973338e-06, + "loss": 0.3637, + "step": 1048 + }, + { + "epoch": 2.315673289183223, + "grad_norm": 0.13365043699741364, + "learning_rate": 1.5032742447262228e-06, + "loss": 0.3535, + "step": 1049 + }, + { + "epoch": 2.3178807947019866, + "grad_norm": 0.13111183047294617, + "learning_rate": 1.4941052493728731e-06, + "loss": 0.3528, + "step": 1050 + }, + { + "epoch": 2.3200883002207506, + "grad_norm": 0.1259627640247345, + "learning_rate": 1.48495938773909e-06, + "loss": 0.3599, + "step": 1051 + }, + { + "epoch": 2.322295805739514, + "grad_norm": 0.12294916808605194, + "learning_rate": 1.4758367201740303e-06, + "loss": 0.3604, + "step": 1052 + }, + { + "epoch": 2.3245033112582782, + "grad_norm": 0.12867586314678192, + "learning_rate": 1.46673730687381e-06, + "loss": 0.3649, + "step": 1053 + }, + { + "epoch": 2.326710816777042, + "grad_norm": 0.12218500673770905, + "learning_rate": 1.4576612078810953e-06, + "loss": 0.3637, + "step": 1054 + }, + { + "epoch": 2.328918322295806, + "grad_norm": 0.12343169003725052, + "learning_rate": 1.448608483084713e-06, + "loss": 0.3552, + "step": 1055 + }, + { + "epoch": 2.3311258278145695, + "grad_norm": 0.12273656576871872, + "learning_rate": 1.4395791922192575e-06, + "loss": 0.3613, + "step": 1056 + }, + { + "epoch": 2.3333333333333335, + "grad_norm": 0.1312197744846344, + "learning_rate": 1.430573394864695e-06, + "loss": 0.3603, + "step": 1057 + }, + { + "epoch": 2.335540838852097, + "grad_norm": 0.13838644325733185, + "learning_rate": 1.4215911504459645e-06, + "loss": 0.3642, + "step": 1058 + }, + { + "epoch": 2.337748344370861, + "grad_norm": 0.11605527251958847, + "learning_rate": 1.412632518232594e-06, + "loss": 0.3626, + "step": 1059 + }, + { + "epoch": 2.3399558498896247, + "grad_norm": 0.13165433704853058, + "learning_rate": 1.4036975573383028e-06, + "loss": 0.3636, + "step": 1060 + }, + { + "epoch": 2.3421633554083887, + "grad_norm": 0.1277674436569214, + "learning_rate": 1.3947863267206174e-06, + "loss": 0.3517, + "step": 1061 + }, + { + "epoch": 2.3443708609271523, + "grad_norm": 0.13928262889385223, + "learning_rate": 1.3858988851804816e-06, + "loss": 0.3649, + "step": 1062 + }, + { + "epoch": 2.3465783664459163, + "grad_norm": 0.13180677592754364, + "learning_rate": 1.3770352913618613e-06, + "loss": 0.3506, + "step": 1063 + }, + { + "epoch": 2.34878587196468, + "grad_norm": 0.13746923208236694, + "learning_rate": 1.3681956037513656e-06, + "loss": 0.3583, + "step": 1064 + }, + { + "epoch": 2.3509933774834435, + "grad_norm": 0.13763363659381866, + "learning_rate": 1.3593798806778546e-06, + "loss": 0.354, + "step": 1065 + }, + { + "epoch": 2.3532008830022075, + "grad_norm": 0.12797822058200836, + "learning_rate": 1.3505881803120647e-06, + "loss": 0.3719, + "step": 1066 + }, + { + "epoch": 2.355408388520971, + "grad_norm": 0.12430882453918457, + "learning_rate": 1.341820560666211e-06, + "loss": 0.3622, + "step": 1067 + }, + { + "epoch": 2.357615894039735, + "grad_norm": 0.13003040850162506, + "learning_rate": 1.3330770795936172e-06, + "loss": 0.3644, + "step": 1068 + }, + { + "epoch": 2.3598233995584987, + "grad_norm": 0.13834795355796814, + "learning_rate": 1.3243577947883223e-06, + "loss": 0.3539, + "step": 1069 + }, + { + "epoch": 2.3620309050772628, + "grad_norm": 0.12837594747543335, + "learning_rate": 1.315662763784712e-06, + "loss": 0.366, + "step": 1070 + }, + { + "epoch": 2.3642384105960264, + "grad_norm": 0.12611278891563416, + "learning_rate": 1.3069920439571277e-06, + "loss": 0.3567, + "step": 1071 + }, + { + "epoch": 2.3664459161147904, + "grad_norm": 0.1504361927509308, + "learning_rate": 1.2983456925194953e-06, + "loss": 0.3539, + "step": 1072 + }, + { + "epoch": 2.368653421633554, + "grad_norm": 0.13299378752708435, + "learning_rate": 1.2897237665249429e-06, + "loss": 0.3612, + "step": 1073 + }, + { + "epoch": 2.370860927152318, + "grad_norm": 0.13157807290554047, + "learning_rate": 1.2811263228654308e-06, + "loss": 0.3652, + "step": 1074 + }, + { + "epoch": 2.3730684326710816, + "grad_norm": 0.11843698471784592, + "learning_rate": 1.2725534182713717e-06, + "loss": 0.3664, + "step": 1075 + }, + { + "epoch": 2.3752759381898456, + "grad_norm": 0.1275041550397873, + "learning_rate": 1.2640051093112532e-06, + "loss": 0.3595, + "step": 1076 + }, + { + "epoch": 2.377483443708609, + "grad_norm": 0.12319236993789673, + "learning_rate": 1.25548145239127e-06, + "loss": 0.3458, + "step": 1077 + }, + { + "epoch": 2.3796909492273732, + "grad_norm": 0.1297508329153061, + "learning_rate": 1.2469825037549493e-06, + "loss": 0.3599, + "step": 1078 + }, + { + "epoch": 2.381898454746137, + "grad_norm": 0.13835854828357697, + "learning_rate": 1.2385083194827818e-06, + "loss": 0.3631, + "step": 1079 + }, + { + "epoch": 2.384105960264901, + "grad_norm": 0.14326325058937073, + "learning_rate": 1.2300589554918502e-06, + "loss": 0.3612, + "step": 1080 + }, + { + "epoch": 2.3863134657836644, + "grad_norm": 0.13561151921749115, + "learning_rate": 1.221634467535458e-06, + "loss": 0.3623, + "step": 1081 + }, + { + "epoch": 2.3885209713024285, + "grad_norm": 0.12151821702718735, + "learning_rate": 1.2132349112027636e-06, + "loss": 0.347, + "step": 1082 + }, + { + "epoch": 2.390728476821192, + "grad_norm": 0.12261461466550827, + "learning_rate": 1.204860341918414e-06, + "loss": 0.3615, + "step": 1083 + }, + { + "epoch": 2.3929359823399556, + "grad_norm": 0.1293371468782425, + "learning_rate": 1.1965108149421812e-06, + "loss": 0.3611, + "step": 1084 + }, + { + "epoch": 2.3951434878587197, + "grad_norm": 0.12331968545913696, + "learning_rate": 1.1881863853685904e-06, + "loss": 0.3635, + "step": 1085 + }, + { + "epoch": 2.3973509933774833, + "grad_norm": 0.13170567154884338, + "learning_rate": 1.1798871081265672e-06, + "loss": 0.3698, + "step": 1086 + }, + { + "epoch": 2.3995584988962473, + "grad_norm": 0.1309044063091278, + "learning_rate": 1.1716130379790613e-06, + "loss": 0.3474, + "step": 1087 + }, + { + "epoch": 2.401766004415011, + "grad_norm": 0.12268619239330292, + "learning_rate": 1.1633642295227005e-06, + "loss": 0.3634, + "step": 1088 + }, + { + "epoch": 2.403973509933775, + "grad_norm": 0.12474343925714493, + "learning_rate": 1.155140737187418e-06, + "loss": 0.3651, + "step": 1089 + }, + { + "epoch": 2.4061810154525385, + "grad_norm": 0.13622671365737915, + "learning_rate": 1.1469426152360974e-06, + "loss": 0.3626, + "step": 1090 + }, + { + "epoch": 2.4083885209713025, + "grad_norm": 0.1363033503293991, + "learning_rate": 1.138769917764221e-06, + "loss": 0.3608, + "step": 1091 + }, + { + "epoch": 2.410596026490066, + "grad_norm": 0.12666958570480347, + "learning_rate": 1.1306226986994989e-06, + "loss": 0.3571, + "step": 1092 + }, + { + "epoch": 2.41280353200883, + "grad_norm": 0.13340440392494202, + "learning_rate": 1.1225010118015306e-06, + "loss": 0.3602, + "step": 1093 + }, + { + "epoch": 2.4150110375275937, + "grad_norm": 0.14224812388420105, + "learning_rate": 1.1144049106614335e-06, + "loss": 0.3525, + "step": 1094 + }, + { + "epoch": 2.4172185430463577, + "grad_norm": 0.13309240341186523, + "learning_rate": 1.1063344487015e-06, + "loss": 0.3595, + "step": 1095 + }, + { + "epoch": 2.4194260485651213, + "grad_norm": 0.1310422122478485, + "learning_rate": 1.098289679174841e-06, + "loss": 0.3674, + "step": 1096 + }, + { + "epoch": 2.4216335540838854, + "grad_norm": 0.12578153610229492, + "learning_rate": 1.090270655165036e-06, + "loss": 0.3543, + "step": 1097 + }, + { + "epoch": 2.423841059602649, + "grad_norm": 0.13076968491077423, + "learning_rate": 1.082277429585784e-06, + "loss": 0.3724, + "step": 1098 + }, + { + "epoch": 2.426048565121413, + "grad_norm": 0.14538316428661346, + "learning_rate": 1.074310055180549e-06, + "loss": 0.3539, + "step": 1099 + }, + { + "epoch": 2.4282560706401766, + "grad_norm": 0.13270893692970276, + "learning_rate": 1.0663685845222177e-06, + "loss": 0.3637, + "step": 1100 + }, + { + "epoch": 2.4304635761589406, + "grad_norm": 0.13810580968856812, + "learning_rate": 1.0584530700127478e-06, + "loss": 0.3526, + "step": 1101 + }, + { + "epoch": 2.432671081677704, + "grad_norm": 0.1298057585954666, + "learning_rate": 1.0505635638828288e-06, + "loss": 0.3551, + "step": 1102 + }, + { + "epoch": 2.4348785871964678, + "grad_norm": 0.13741736114025116, + "learning_rate": 1.0427001181915298e-06, + "loss": 0.3512, + "step": 1103 + }, + { + "epoch": 2.437086092715232, + "grad_norm": 0.1301170289516449, + "learning_rate": 1.034862784825963e-06, + "loss": 0.3568, + "step": 1104 + }, + { + "epoch": 2.439293598233996, + "grad_norm": 0.12320411205291748, + "learning_rate": 1.0270516155009336e-06, + "loss": 0.3701, + "step": 1105 + }, + { + "epoch": 2.4415011037527594, + "grad_norm": 0.1281418800354004, + "learning_rate": 1.0192666617586072e-06, + "loss": 0.3657, + "step": 1106 + }, + { + "epoch": 2.443708609271523, + "grad_norm": 0.14360150694847107, + "learning_rate": 1.0115079749681628e-06, + "loss": 0.3598, + "step": 1107 + }, + { + "epoch": 2.445916114790287, + "grad_norm": 0.13790085911750793, + "learning_rate": 1.0037756063254555e-06, + "loss": 0.3612, + "step": 1108 + }, + { + "epoch": 2.4481236203090506, + "grad_norm": 0.1285860389471054, + "learning_rate": 9.960696068526826e-07, + "loss": 0.3615, + "step": 1109 + }, + { + "epoch": 2.4503311258278146, + "grad_norm": 0.1286444067955017, + "learning_rate": 9.883900273980408e-07, + "loss": 0.3587, + "step": 1110 + }, + { + "epoch": 2.4525386313465782, + "grad_norm": 0.12481298297643661, + "learning_rate": 9.807369186353965e-07, + "loss": 0.356, + "step": 1111 + }, + { + "epoch": 2.4547461368653423, + "grad_norm": 0.1254754513502121, + "learning_rate": 9.731103310639473e-07, + "loss": 0.3575, + "step": 1112 + }, + { + "epoch": 2.456953642384106, + "grad_norm": 0.12286875396966934, + "learning_rate": 9.655103150078892e-07, + "loss": 0.3608, + "step": 1113 + }, + { + "epoch": 2.45916114790287, + "grad_norm": 0.11990831047296524, + "learning_rate": 9.57936920616086e-07, + "loss": 0.3586, + "step": 1114 + }, + { + "epoch": 2.4613686534216335, + "grad_norm": 0.11828765273094177, + "learning_rate": 9.503901978617392e-07, + "loss": 0.3621, + "step": 1115 + }, + { + "epoch": 2.4635761589403975, + "grad_norm": 0.13093435764312744, + "learning_rate": 9.428701965420572e-07, + "loss": 0.361, + "step": 1116 + }, + { + "epoch": 2.465783664459161, + "grad_norm": 0.12574456632137299, + "learning_rate": 9.353769662779249e-07, + "loss": 0.3556, + "step": 1117 + }, + { + "epoch": 2.467991169977925, + "grad_norm": 0.136729896068573, + "learning_rate": 9.279105565135777e-07, + "loss": 0.3661, + "step": 1118 + }, + { + "epoch": 2.4701986754966887, + "grad_norm": 0.12174253910779953, + "learning_rate": 9.204710165162751e-07, + "loss": 0.3487, + "step": 1119 + }, + { + "epoch": 2.4724061810154527, + "grad_norm": 0.13417866826057434, + "learning_rate": 9.130583953759781e-07, + "loss": 0.358, + "step": 1120 + }, + { + "epoch": 2.4746136865342163, + "grad_norm": 0.14360179007053375, + "learning_rate": 9.056727420050227e-07, + "loss": 0.3628, + "step": 1121 + }, + { + "epoch": 2.47682119205298, + "grad_norm": 0.12731443345546722, + "learning_rate": 8.983141051377953e-07, + "loss": 0.3603, + "step": 1122 + }, + { + "epoch": 2.479028697571744, + "grad_norm": 0.13204246759414673, + "learning_rate": 8.909825333304134e-07, + "loss": 0.355, + "step": 1123 + }, + { + "epoch": 2.481236203090508, + "grad_norm": 0.1328221708536148, + "learning_rate": 8.836780749604096e-07, + "loss": 0.3541, + "step": 1124 + }, + { + "epoch": 2.4834437086092715, + "grad_norm": 0.13670921325683594, + "learning_rate": 8.764007782264022e-07, + "loss": 0.3649, + "step": 1125 + }, + { + "epoch": 2.485651214128035, + "grad_norm": 0.124913290143013, + "learning_rate": 8.691506911477848e-07, + "loss": 0.3474, + "step": 1126 + }, + { + "epoch": 2.487858719646799, + "grad_norm": 0.12466172128915787, + "learning_rate": 8.619278615644106e-07, + "loss": 0.3559, + "step": 1127 + }, + { + "epoch": 2.4900662251655628, + "grad_norm": 0.13907819986343384, + "learning_rate": 8.547323371362682e-07, + "loss": 0.3544, + "step": 1128 + }, + { + "epoch": 2.492273730684327, + "grad_norm": 0.13403339684009552, + "learning_rate": 8.475641653431782e-07, + "loss": 0.3621, + "step": 1129 + }, + { + "epoch": 2.4944812362030904, + "grad_norm": 0.13107538223266602, + "learning_rate": 8.404233934844707e-07, + "loss": 0.3629, + "step": 1130 + }, + { + "epoch": 2.4966887417218544, + "grad_norm": 0.12382876127958298, + "learning_rate": 8.333100686786766e-07, + "loss": 0.3425, + "step": 1131 + }, + { + "epoch": 2.498896247240618, + "grad_norm": 0.1340240240097046, + "learning_rate": 8.262242378632179e-07, + "loss": 0.3637, + "step": 1132 + }, + { + "epoch": 2.501103752759382, + "grad_norm": 0.12560687959194183, + "learning_rate": 8.191659477940972e-07, + "loss": 0.3694, + "step": 1133 + }, + { + "epoch": 2.5033112582781456, + "grad_norm": 0.13509927690029144, + "learning_rate": 8.121352450455899e-07, + "loss": 0.3588, + "step": 1134 + }, + { + "epoch": 2.5055187637969096, + "grad_norm": 0.13506671786308289, + "learning_rate": 8.051321760099334e-07, + "loss": 0.3657, + "step": 1135 + }, + { + "epoch": 2.507726269315673, + "grad_norm": 0.12824004888534546, + "learning_rate": 7.981567868970252e-07, + "loss": 0.3541, + "step": 1136 + }, + { + "epoch": 2.5099337748344372, + "grad_norm": 0.12654347717761993, + "learning_rate": 7.91209123734114e-07, + "loss": 0.3628, + "step": 1137 + }, + { + "epoch": 2.512141280353201, + "grad_norm": 0.13879314064979553, + "learning_rate": 7.84289232365501e-07, + "loss": 0.3631, + "step": 1138 + }, + { + "epoch": 2.514348785871965, + "grad_norm": 0.11704026162624359, + "learning_rate": 7.773971584522355e-07, + "loss": 0.3589, + "step": 1139 + }, + { + "epoch": 2.5165562913907285, + "grad_norm": 0.12439953535795212, + "learning_rate": 7.705329474718093e-07, + "loss": 0.3537, + "step": 1140 + }, + { + "epoch": 2.518763796909492, + "grad_norm": 0.13436546921730042, + "learning_rate": 7.636966447178601e-07, + "loss": 0.356, + "step": 1141 + }, + { + "epoch": 2.520971302428256, + "grad_norm": 0.12255270779132843, + "learning_rate": 7.568882952998762e-07, + "loss": 0.3568, + "step": 1142 + }, + { + "epoch": 2.52317880794702, + "grad_norm": 0.12233200669288635, + "learning_rate": 7.501079441428927e-07, + "loss": 0.3538, + "step": 1143 + }, + { + "epoch": 2.5253863134657837, + "grad_norm": 0.12473509460687637, + "learning_rate": 7.433556359871968e-07, + "loss": 0.3711, + "step": 1144 + }, + { + "epoch": 2.5275938189845473, + "grad_norm": 0.13673733174800873, + "learning_rate": 7.366314153880361e-07, + "loss": 0.3613, + "step": 1145 + }, + { + "epoch": 2.5298013245033113, + "grad_norm": 0.1345418095588684, + "learning_rate": 7.299353267153192e-07, + "loss": 0.3632, + "step": 1146 + }, + { + "epoch": 2.5320088300220753, + "grad_norm": 0.1333194524049759, + "learning_rate": 7.232674141533274e-07, + "loss": 0.3653, + "step": 1147 + }, + { + "epoch": 2.534216335540839, + "grad_norm": 0.13697415590286255, + "learning_rate": 7.166277217004214e-07, + "loss": 0.3571, + "step": 1148 + }, + { + "epoch": 2.5364238410596025, + "grad_norm": 0.13772854208946228, + "learning_rate": 7.100162931687476e-07, + "loss": 0.3737, + "step": 1149 + }, + { + "epoch": 2.5386313465783665, + "grad_norm": 0.1227809488773346, + "learning_rate": 7.034331721839566e-07, + "loss": 0.3622, + "step": 1150 + }, + { + "epoch": 2.54083885209713, + "grad_norm": 0.12854185700416565, + "learning_rate": 6.968784021849106e-07, + "loss": 0.3612, + "step": 1151 + }, + { + "epoch": 2.543046357615894, + "grad_norm": 0.11819867044687271, + "learning_rate": 6.903520264233954e-07, + "loss": 0.3625, + "step": 1152 + }, + { + "epoch": 2.5452538631346577, + "grad_norm": 0.1528642773628235, + "learning_rate": 6.838540879638367e-07, + "loss": 0.3644, + "step": 1153 + }, + { + "epoch": 2.5474613686534218, + "grad_norm": 0.1466141641139984, + "learning_rate": 6.773846296830167e-07, + "loss": 0.3681, + "step": 1154 + }, + { + "epoch": 2.5496688741721854, + "grad_norm": 0.13698391616344452, + "learning_rate": 6.7094369426979e-07, + "loss": 0.3569, + "step": 1155 + }, + { + "epoch": 2.5518763796909494, + "grad_norm": 0.1320303976535797, + "learning_rate": 6.645313242248042e-07, + "loss": 0.3531, + "step": 1156 + }, + { + "epoch": 2.554083885209713, + "grad_norm": 0.14429907500743866, + "learning_rate": 6.581475618602174e-07, + "loss": 0.3587, + "step": 1157 + }, + { + "epoch": 2.556291390728477, + "grad_norm": 0.12878787517547607, + "learning_rate": 6.517924492994182e-07, + "loss": 0.3648, + "step": 1158 + }, + { + "epoch": 2.5584988962472406, + "grad_norm": 0.13368941843509674, + "learning_rate": 6.454660284767477e-07, + "loss": 0.3569, + "step": 1159 + }, + { + "epoch": 2.560706401766004, + "grad_norm": 0.13522258400917053, + "learning_rate": 6.391683411372279e-07, + "loss": 0.3624, + "step": 1160 + }, + { + "epoch": 2.562913907284768, + "grad_norm": 0.1288643628358841, + "learning_rate": 6.328994288362783e-07, + "loss": 0.3571, + "step": 1161 + }, + { + "epoch": 2.5651214128035322, + "grad_norm": 0.13528789579868317, + "learning_rate": 6.266593329394471e-07, + "loss": 0.3614, + "step": 1162 + }, + { + "epoch": 2.567328918322296, + "grad_norm": 0.11771933734416962, + "learning_rate": 6.204480946221386e-07, + "loss": 0.3503, + "step": 1163 + }, + { + "epoch": 2.5695364238410594, + "grad_norm": 0.13160669803619385, + "learning_rate": 6.142657548693364e-07, + "loss": 0.3662, + "step": 1164 + }, + { + "epoch": 2.5717439293598234, + "grad_norm": 0.13713331520557404, + "learning_rate": 6.081123544753404e-07, + "loss": 0.3474, + "step": 1165 + }, + { + "epoch": 2.5739514348785875, + "grad_norm": 0.12216666340827942, + "learning_rate": 6.019879340434904e-07, + "loss": 0.3414, + "step": 1166 + }, + { + "epoch": 2.576158940397351, + "grad_norm": 0.12497063726186752, + "learning_rate": 5.958925339859001e-07, + "loss": 0.3501, + "step": 1167 + }, + { + "epoch": 2.5783664459161146, + "grad_norm": 0.1288682073354721, + "learning_rate": 5.898261945231965e-07, + "loss": 0.3518, + "step": 1168 + }, + { + "epoch": 2.5805739514348787, + "grad_norm": 0.13969705998897552, + "learning_rate": 5.837889556842469e-07, + "loss": 0.3595, + "step": 1169 + }, + { + "epoch": 2.5827814569536423, + "grad_norm": 0.13971775770187378, + "learning_rate": 5.777808573058969e-07, + "loss": 0.3635, + "step": 1170 + }, + { + "epoch": 2.5849889624724063, + "grad_norm": 0.12668545544147491, + "learning_rate": 5.718019390327084e-07, + "loss": 0.3519, + "step": 1171 + }, + { + "epoch": 2.58719646799117, + "grad_norm": 0.13675172626972198, + "learning_rate": 5.658522403166989e-07, + "loss": 0.363, + "step": 1172 + }, + { + "epoch": 2.589403973509934, + "grad_norm": 0.145385280251503, + "learning_rate": 5.599318004170778e-07, + "loss": 0.3561, + "step": 1173 + }, + { + "epoch": 2.5916114790286975, + "grad_norm": 0.14911803603172302, + "learning_rate": 5.540406583999925e-07, + "loss": 0.3578, + "step": 1174 + }, + { + "epoch": 2.5938189845474615, + "grad_norm": 0.12694287300109863, + "learning_rate": 5.481788531382671e-07, + "loss": 0.364, + "step": 1175 + }, + { + "epoch": 2.596026490066225, + "grad_norm": 0.13262110948562622, + "learning_rate": 5.423464233111448e-07, + "loss": 0.3718, + "step": 1176 + }, + { + "epoch": 2.598233995584989, + "grad_norm": 0.13582631945610046, + "learning_rate": 5.365434074040343e-07, + "loss": 0.3654, + "step": 1177 + }, + { + "epoch": 2.6004415011037527, + "grad_norm": 0.13755717873573303, + "learning_rate": 5.307698437082598e-07, + "loss": 0.3522, + "step": 1178 + }, + { + "epoch": 2.6026490066225163, + "grad_norm": 0.1361360102891922, + "learning_rate": 5.250257703207984e-07, + "loss": 0.3683, + "step": 1179 + }, + { + "epoch": 2.6048565121412803, + "grad_norm": 0.13162481784820557, + "learning_rate": 5.193112251440407e-07, + "loss": 0.3559, + "step": 1180 + }, + { + "epoch": 2.6070640176600444, + "grad_norm": 0.1336871236562729, + "learning_rate": 5.136262458855312e-07, + "loss": 0.3539, + "step": 1181 + }, + { + "epoch": 2.609271523178808, + "grad_norm": 0.12142772227525711, + "learning_rate": 5.079708700577229e-07, + "loss": 0.3633, + "step": 1182 + }, + { + "epoch": 2.6114790286975715, + "grad_norm": 0.12398800998926163, + "learning_rate": 5.023451349777331e-07, + "loss": 0.3666, + "step": 1183 + }, + { + "epoch": 2.6136865342163356, + "grad_norm": 0.1366734653711319, + "learning_rate": 4.967490777670903e-07, + "loss": 0.3552, + "step": 1184 + }, + { + "epoch": 2.6158940397350996, + "grad_norm": 0.13601034879684448, + "learning_rate": 4.911827353514947e-07, + "loss": 0.3606, + "step": 1185 + }, + { + "epoch": 2.618101545253863, + "grad_norm": 0.13659435510635376, + "learning_rate": 4.856461444605732e-07, + "loss": 0.356, + "step": 1186 + }, + { + "epoch": 2.6203090507726268, + "grad_norm": 0.13636133074760437, + "learning_rate": 4.801393416276368e-07, + "loss": 0.3591, + "step": 1187 + }, + { + "epoch": 2.622516556291391, + "grad_norm": 0.1251705437898636, + "learning_rate": 4.7466236318943816e-07, + "loss": 0.3557, + "step": 1188 + }, + { + "epoch": 2.6247240618101544, + "grad_norm": 0.14416959881782532, + "learning_rate": 4.692152452859333e-07, + "loss": 0.3587, + "step": 1189 + }, + { + "epoch": 2.6269315673289184, + "grad_norm": 0.15531152486801147, + "learning_rate": 4.637980238600437e-07, + "loss": 0.3494, + "step": 1190 + }, + { + "epoch": 2.629139072847682, + "grad_norm": 0.1295311003923416, + "learning_rate": 4.584107346574168e-07, + "loss": 0.3595, + "step": 1191 + }, + { + "epoch": 2.631346578366446, + "grad_norm": 0.13844533264636993, + "learning_rate": 4.530534132261932e-07, + "loss": 0.3589, + "step": 1192 + }, + { + "epoch": 2.6335540838852096, + "grad_norm": 0.13652655482292175, + "learning_rate": 4.477260949167711e-07, + "loss": 0.3492, + "step": 1193 + }, + { + "epoch": 2.6357615894039736, + "grad_norm": 0.14151211082935333, + "learning_rate": 4.4242881488157083e-07, + "loss": 0.3565, + "step": 1194 + }, + { + "epoch": 2.6379690949227372, + "grad_norm": 0.13526977598667145, + "learning_rate": 4.371616080748037e-07, + "loss": 0.355, + "step": 1195 + }, + { + "epoch": 2.6401766004415013, + "grad_norm": 0.13757188618183136, + "learning_rate": 4.319245092522456e-07, + "loss": 0.3547, + "step": 1196 + }, + { + "epoch": 2.642384105960265, + "grad_norm": 0.1314619481563568, + "learning_rate": 4.2671755297100047e-07, + "loss": 0.36, + "step": 1197 + }, + { + "epoch": 2.6445916114790284, + "grad_norm": 0.13309867680072784, + "learning_rate": 4.215407735892796e-07, + "loss": 0.3578, + "step": 1198 + }, + { + "epoch": 2.6467991169977925, + "grad_norm": 0.14237742125988007, + "learning_rate": 4.1639420526616845e-07, + "loss": 0.3567, + "step": 1199 + }, + { + "epoch": 2.6490066225165565, + "grad_norm": 0.13557758927345276, + "learning_rate": 4.1127788196140437e-07, + "loss": 0.3627, + "step": 1200 + }, + { + "epoch": 2.65121412803532, + "grad_norm": 0.14701923727989197, + "learning_rate": 4.061918374351559e-07, + "loss": 0.3518, + "step": 1201 + }, + { + "epoch": 2.6534216335540837, + "grad_norm": 0.1429191380739212, + "learning_rate": 4.0113610524779246e-07, + "loss": 0.3579, + "step": 1202 + }, + { + "epoch": 2.6556291390728477, + "grad_norm": 0.13261815905570984, + "learning_rate": 3.961107187596669e-07, + "loss": 0.3558, + "step": 1203 + }, + { + "epoch": 2.6578366445916117, + "grad_norm": 0.15215714275836945, + "learning_rate": 3.911157111308983e-07, + "loss": 0.3582, + "step": 1204 + }, + { + "epoch": 2.6600441501103753, + "grad_norm": 0.134977787733078, + "learning_rate": 3.861511153211489e-07, + "loss": 0.3484, + "step": 1205 + }, + { + "epoch": 2.662251655629139, + "grad_norm": 0.13735589385032654, + "learning_rate": 3.81216964089407e-07, + "loss": 0.3633, + "step": 1206 + }, + { + "epoch": 2.664459161147903, + "grad_norm": 0.13418880105018616, + "learning_rate": 3.763132899937721e-07, + "loss": 0.3646, + "step": 1207 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 0.13318879902362823, + "learning_rate": 3.7144012539123973e-07, + "loss": 0.3449, + "step": 1208 + }, + { + "epoch": 2.6688741721854305, + "grad_norm": 0.14231936633586884, + "learning_rate": 3.665975024374879e-07, + "loss": 0.3661, + "step": 1209 + }, + { + "epoch": 2.671081677704194, + "grad_norm": 0.13167575001716614, + "learning_rate": 3.617854530866671e-07, + "loss": 0.3627, + "step": 1210 + }, + { + "epoch": 2.673289183222958, + "grad_norm": 0.13379094004631042, + "learning_rate": 3.5700400909118306e-07, + "loss": 0.3585, + "step": 1211 + }, + { + "epoch": 2.6754966887417218, + "grad_norm": 0.1270161271095276, + "learning_rate": 3.522532020014946e-07, + "loss": 0.3571, + "step": 1212 + }, + { + "epoch": 2.677704194260486, + "grad_norm": 0.1238294392824173, + "learning_rate": 3.475330631659002e-07, + "loss": 0.3653, + "step": 1213 + }, + { + "epoch": 2.6799116997792494, + "grad_norm": 0.12466707080602646, + "learning_rate": 3.4284362373033566e-07, + "loss": 0.3535, + "step": 1214 + }, + { + "epoch": 2.6821192052980134, + "grad_norm": 0.12770843505859375, + "learning_rate": 3.3818491463816385e-07, + "loss": 0.3573, + "step": 1215 + }, + { + "epoch": 2.684326710816777, + "grad_norm": 0.1300242394208908, + "learning_rate": 3.335569666299748e-07, + "loss": 0.3566, + "step": 1216 + }, + { + "epoch": 2.6865342163355406, + "grad_norm": 0.1325581669807434, + "learning_rate": 3.289598102433794e-07, + "loss": 0.3575, + "step": 1217 + }, + { + "epoch": 2.6887417218543046, + "grad_norm": 0.1356075257062912, + "learning_rate": 3.2439347581280865e-07, + "loss": 0.3614, + "step": 1218 + }, + { + "epoch": 2.6909492273730686, + "grad_norm": 0.12728764116764069, + "learning_rate": 3.19857993469318e-07, + "loss": 0.3586, + "step": 1219 + }, + { + "epoch": 2.693156732891832, + "grad_norm": 0.1246216669678688, + "learning_rate": 3.1535339314038015e-07, + "loss": 0.3581, + "step": 1220 + }, + { + "epoch": 2.695364238410596, + "grad_norm": 0.1401343196630478, + "learning_rate": 3.108797045496942e-07, + "loss": 0.3611, + "step": 1221 + }, + { + "epoch": 2.69757174392936, + "grad_norm": 0.1308797299861908, + "learning_rate": 3.0643695721698783e-07, + "loss": 0.3664, + "step": 1222 + }, + { + "epoch": 2.699779249448124, + "grad_norm": 0.12849809229373932, + "learning_rate": 3.0202518045782337e-07, + "loss": 0.3605, + "step": 1223 + }, + { + "epoch": 2.7019867549668874, + "grad_norm": 0.128408744931221, + "learning_rate": 2.9764440338340083e-07, + "loss": 0.3583, + "step": 1224 + }, + { + "epoch": 2.704194260485651, + "grad_norm": 0.1296570599079132, + "learning_rate": 2.9329465490036844e-07, + "loss": 0.3528, + "step": 1225 + }, + { + "epoch": 2.706401766004415, + "grad_norm": 0.14551834762096405, + "learning_rate": 2.8897596371063153e-07, + "loss": 0.3497, + "step": 1226 + }, + { + "epoch": 2.7086092715231787, + "grad_norm": 0.13115178048610687, + "learning_rate": 2.846883583111648e-07, + "loss": 0.3594, + "step": 1227 + }, + { + "epoch": 2.7108167770419427, + "grad_norm": 0.12846940755844116, + "learning_rate": 2.804318669938233e-07, + "loss": 0.3583, + "step": 1228 + }, + { + "epoch": 2.7130242825607063, + "grad_norm": 0.1380581110715866, + "learning_rate": 2.762065178451517e-07, + "loss": 0.3585, + "step": 1229 + }, + { + "epoch": 2.7152317880794703, + "grad_norm": 0.12574802339076996, + "learning_rate": 2.7201233874620534e-07, + "loss": 0.3598, + "step": 1230 + }, + { + "epoch": 2.717439293598234, + "grad_norm": 0.13319121301174164, + "learning_rate": 2.678493573723612e-07, + "loss": 0.3669, + "step": 1231 + }, + { + "epoch": 2.719646799116998, + "grad_norm": 0.12967029213905334, + "learning_rate": 2.6371760119314026e-07, + "loss": 0.3637, + "step": 1232 + }, + { + "epoch": 2.7218543046357615, + "grad_norm": 0.13314829766750336, + "learning_rate": 2.596170974720202e-07, + "loss": 0.3505, + "step": 1233 + }, + { + "epoch": 2.7240618101545255, + "grad_norm": 0.13859589397907257, + "learning_rate": 2.5554787326626194e-07, + "loss": 0.3508, + "step": 1234 + }, + { + "epoch": 2.726269315673289, + "grad_norm": 0.13054342567920685, + "learning_rate": 2.515099554267247e-07, + "loss": 0.3459, + "step": 1235 + }, + { + "epoch": 2.7284768211920527, + "grad_norm": 0.13782578706741333, + "learning_rate": 2.4750337059769425e-07, + "loss": 0.3653, + "step": 1236 + }, + { + "epoch": 2.7306843267108167, + "grad_norm": 0.1294957846403122, + "learning_rate": 2.4352814521670375e-07, + "loss": 0.3614, + "step": 1237 + }, + { + "epoch": 2.7328918322295808, + "grad_norm": 0.1325952708721161, + "learning_rate": 2.3958430551436095e-07, + "loss": 0.3638, + "step": 1238 + }, + { + "epoch": 2.7350993377483444, + "grad_norm": 0.12079501897096634, + "learning_rate": 2.3567187751417475e-07, + "loss": 0.3605, + "step": 1239 + }, + { + "epoch": 2.737306843267108, + "grad_norm": 0.13967706263065338, + "learning_rate": 2.3179088703238096e-07, + "loss": 0.3721, + "step": 1240 + }, + { + "epoch": 2.739514348785872, + "grad_norm": 0.133636474609375, + "learning_rate": 2.2794135967777908e-07, + "loss": 0.3655, + "step": 1241 + }, + { + "epoch": 2.741721854304636, + "grad_norm": 0.13323010504245758, + "learning_rate": 2.2412332085155364e-07, + "loss": 0.3531, + "step": 1242 + }, + { + "epoch": 2.7439293598233996, + "grad_norm": 0.13812537491321564, + "learning_rate": 2.2033679574711365e-07, + "loss": 0.3453, + "step": 1243 + }, + { + "epoch": 2.746136865342163, + "grad_norm": 0.12806718051433563, + "learning_rate": 2.1658180934992333e-07, + "loss": 0.3632, + "step": 1244 + }, + { + "epoch": 2.748344370860927, + "grad_norm": 0.13933822512626648, + "learning_rate": 2.1285838643733958e-07, + "loss": 0.3595, + "step": 1245 + }, + { + "epoch": 2.750551876379691, + "grad_norm": 0.12600216269493103, + "learning_rate": 2.0916655157844634e-07, + "loss": 0.352, + "step": 1246 + }, + { + "epoch": 2.752759381898455, + "grad_norm": 0.12791283428668976, + "learning_rate": 2.0550632913389213e-07, + "loss": 0.3556, + "step": 1247 + }, + { + "epoch": 2.7549668874172184, + "grad_norm": 0.13090188801288605, + "learning_rate": 2.0187774325573174e-07, + "loss": 0.3632, + "step": 1248 + }, + { + "epoch": 2.7571743929359824, + "grad_norm": 0.1253674477338791, + "learning_rate": 1.9828081788726307e-07, + "loss": 0.3499, + "step": 1249 + }, + { + "epoch": 2.759381898454746, + "grad_norm": 0.13384896516799927, + "learning_rate": 1.9471557676287501e-07, + "loss": 0.3556, + "step": 1250 + }, + { + "epoch": 2.76158940397351, + "grad_norm": 0.13008341193199158, + "learning_rate": 1.9118204340788426e-07, + "loss": 0.3521, + "step": 1251 + }, + { + "epoch": 2.7637969094922736, + "grad_norm": 0.14128050208091736, + "learning_rate": 1.8768024113838546e-07, + "loss": 0.3629, + "step": 1252 + }, + { + "epoch": 2.7660044150110377, + "grad_norm": 0.13070915639400482, + "learning_rate": 1.8421019306109288e-07, + "loss": 0.3576, + "step": 1253 + }, + { + "epoch": 2.7682119205298013, + "grad_norm": 0.13155661523342133, + "learning_rate": 1.8077192207319072e-07, + "loss": 0.3614, + "step": 1254 + }, + { + "epoch": 2.7704194260485653, + "grad_norm": 0.14038477838039398, + "learning_rate": 1.773654508621825e-07, + "loss": 0.354, + "step": 1255 + }, + { + "epoch": 2.772626931567329, + "grad_norm": 0.13292455673217773, + "learning_rate": 1.7399080190573903e-07, + "loss": 0.3635, + "step": 1256 + }, + { + "epoch": 2.774834437086093, + "grad_norm": 0.1393938958644867, + "learning_rate": 1.7064799747155248e-07, + "loss": 0.3502, + "step": 1257 + }, + { + "epoch": 2.7770419426048565, + "grad_norm": 0.13566245138645172, + "learning_rate": 1.6733705961718694e-07, + "loss": 0.353, + "step": 1258 + }, + { + "epoch": 2.77924944812362, + "grad_norm": 0.1313687562942505, + "learning_rate": 1.640580101899353e-07, + "loss": 0.3519, + "step": 1259 + }, + { + "epoch": 2.781456953642384, + "grad_norm": 0.13328541815280914, + "learning_rate": 1.6081087082667314e-07, + "loss": 0.3513, + "step": 1260 + }, + { + "epoch": 2.783664459161148, + "grad_norm": 0.13640496134757996, + "learning_rate": 1.575956629537173e-07, + "loss": 0.3559, + "step": 1261 + }, + { + "epoch": 2.7858719646799117, + "grad_norm": 0.13169561326503754, + "learning_rate": 1.5441240778668321e-07, + "loss": 0.3589, + "step": 1262 + }, + { + "epoch": 2.7880794701986753, + "grad_norm": 0.1369987428188324, + "learning_rate": 1.5126112633034761e-07, + "loss": 0.368, + "step": 1263 + }, + { + "epoch": 2.7902869757174393, + "grad_norm": 0.12467605620622635, + "learning_rate": 1.4814183937850668e-07, + "loss": 0.3563, + "step": 1264 + }, + { + "epoch": 2.7924944812362034, + "grad_norm": 0.1307917833328247, + "learning_rate": 1.4505456751383985e-07, + "loss": 0.3516, + "step": 1265 + }, + { + "epoch": 2.794701986754967, + "grad_norm": 0.14058281481266022, + "learning_rate": 1.4199933110777553e-07, + "loss": 0.3546, + "step": 1266 + }, + { + "epoch": 2.7969094922737305, + "grad_norm": 0.12925602495670319, + "learning_rate": 1.389761503203535e-07, + "loss": 0.3615, + "step": 1267 + }, + { + "epoch": 2.7991169977924946, + "grad_norm": 0.1409684419631958, + "learning_rate": 1.3598504510009602e-07, + "loss": 0.3611, + "step": 1268 + }, + { + "epoch": 2.801324503311258, + "grad_norm": 0.13529111444950104, + "learning_rate": 1.3302603518387358e-07, + "loss": 0.3556, + "step": 1269 + }, + { + "epoch": 2.803532008830022, + "grad_norm": 0.12611360847949982, + "learning_rate": 1.3009914009677493e-07, + "loss": 0.3695, + "step": 1270 + }, + { + "epoch": 2.8057395143487858, + "grad_norm": 0.1310908943414688, + "learning_rate": 1.272043791519778e-07, + "loss": 0.3571, + "step": 1271 + }, + { + "epoch": 2.80794701986755, + "grad_norm": 0.13993841409683228, + "learning_rate": 1.2434177145062177e-07, + "loss": 0.3572, + "step": 1272 + }, + { + "epoch": 2.8101545253863134, + "grad_norm": 0.1199953481554985, + "learning_rate": 1.215113358816844e-07, + "loss": 0.3523, + "step": 1273 + }, + { + "epoch": 2.8123620309050774, + "grad_norm": 0.12030813843011856, + "learning_rate": 1.187130911218537e-07, + "loss": 0.3533, + "step": 1274 + }, + { + "epoch": 2.814569536423841, + "grad_norm": 0.1288137286901474, + "learning_rate": 1.1594705563540642e-07, + "loss": 0.3587, + "step": 1275 + }, + { + "epoch": 2.816777041942605, + "grad_norm": 0.1391681730747223, + "learning_rate": 1.1321324767408382e-07, + "loss": 0.3592, + "step": 1276 + }, + { + "epoch": 2.8189845474613686, + "grad_norm": 0.12392991036176682, + "learning_rate": 1.1051168527697665e-07, + "loss": 0.3605, + "step": 1277 + }, + { + "epoch": 2.821192052980132, + "grad_norm": 0.12354668974876404, + "learning_rate": 1.0784238627039977e-07, + "loss": 0.3593, + "step": 1278 + }, + { + "epoch": 2.8233995584988962, + "grad_norm": 0.13385632634162903, + "learning_rate": 1.0520536826777783e-07, + "loss": 0.3573, + "step": 1279 + }, + { + "epoch": 2.8256070640176603, + "grad_norm": 0.12551288306713104, + "learning_rate": 1.0260064866952968e-07, + "loss": 0.3563, + "step": 1280 + }, + { + "epoch": 2.827814569536424, + "grad_norm": 0.119255430996418, + "learning_rate": 1.0002824466295191e-07, + "loss": 0.3553, + "step": 1281 + }, + { + "epoch": 2.8300220750551874, + "grad_norm": 0.12346068769693375, + "learning_rate": 9.748817322210558e-08, + "loss": 0.3547, + "step": 1282 + }, + { + "epoch": 2.8322295805739515, + "grad_norm": 0.12130031734704971, + "learning_rate": 9.498045110770571e-08, + "loss": 0.3476, + "step": 1283 + }, + { + "epoch": 2.8344370860927155, + "grad_norm": 0.13486534357070923, + "learning_rate": 9.250509486700809e-08, + "loss": 0.3652, + "step": 1284 + }, + { + "epoch": 2.836644591611479, + "grad_norm": 0.12220917642116547, + "learning_rate": 9.006212083370213e-08, + "loss": 0.3548, + "step": 1285 + }, + { + "epoch": 2.8388520971302427, + "grad_norm": 0.12486063688993454, + "learning_rate": 8.765154512780428e-08, + "loss": 0.3502, + "step": 1286 + }, + { + "epoch": 2.8410596026490067, + "grad_norm": 0.11943277716636658, + "learning_rate": 8.527338365554749e-08, + "loss": 0.3462, + "step": 1287 + }, + { + "epoch": 2.8432671081677703, + "grad_norm": 0.13910432159900665, + "learning_rate": 8.292765210928089e-08, + "loss": 0.3585, + "step": 1288 + }, + { + "epoch": 2.8454746136865343, + "grad_norm": 0.12948773801326752, + "learning_rate": 8.061436596736139e-08, + "loss": 0.3589, + "step": 1289 + }, + { + "epoch": 2.847682119205298, + "grad_norm": 0.13374410569667816, + "learning_rate": 7.833354049405717e-08, + "loss": 0.3505, + "step": 1290 + }, + { + "epoch": 2.849889624724062, + "grad_norm": 0.1347481608390808, + "learning_rate": 7.608519073944165e-08, + "loss": 0.3687, + "step": 1291 + }, + { + "epoch": 2.8520971302428255, + "grad_norm": 0.13521930575370789, + "learning_rate": 7.386933153929798e-08, + "loss": 0.366, + "step": 1292 + }, + { + "epoch": 2.8543046357615895, + "grad_norm": 0.13502496480941772, + "learning_rate": 7.168597751501972e-08, + "loss": 0.366, + "step": 1293 + }, + { + "epoch": 2.856512141280353, + "grad_norm": 0.1404918134212494, + "learning_rate": 6.953514307351306e-08, + "loss": 0.357, + "step": 1294 + }, + { + "epoch": 2.858719646799117, + "grad_norm": 0.12767677009105682, + "learning_rate": 6.741684240710477e-08, + "loss": 0.3517, + "step": 1295 + }, + { + "epoch": 2.8609271523178808, + "grad_norm": 0.13126912713050842, + "learning_rate": 6.533108949344558e-08, + "loss": 0.3564, + "step": 1296 + }, + { + "epoch": 2.8631346578366443, + "grad_norm": 0.13056641817092896, + "learning_rate": 6.32778980954185e-08, + "loss": 0.3587, + "step": 1297 + }, + { + "epoch": 2.8653421633554084, + "grad_norm": 0.13267594575881958, + "learning_rate": 6.125728176105129e-08, + "loss": 0.3686, + "step": 1298 + }, + { + "epoch": 2.8675496688741724, + "grad_norm": 0.12520645558834076, + "learning_rate": 5.9269253823421855e-08, + "loss": 0.3703, + "step": 1299 + }, + { + "epoch": 2.869757174392936, + "grad_norm": 0.1308923214673996, + "learning_rate": 5.731382740057523e-08, + "loss": 0.3549, + "step": 1300 + }, + { + "epoch": 2.8719646799116996, + "grad_norm": 0.13203322887420654, + "learning_rate": 5.5391015395432346e-08, + "loss": 0.353, + "step": 1301 + }, + { + "epoch": 2.8741721854304636, + "grad_norm": 0.13061153888702393, + "learning_rate": 5.350083049570853e-08, + "loss": 0.3556, + "step": 1302 + }, + { + "epoch": 2.8763796909492276, + "grad_norm": 0.13314051926136017, + "learning_rate": 5.164328517382744e-08, + "loss": 0.3578, + "step": 1303 + }, + { + "epoch": 2.878587196467991, + "grad_norm": 0.13481754064559937, + "learning_rate": 4.9818391686842214e-08, + "loss": 0.362, + "step": 1304 + }, + { + "epoch": 2.880794701986755, + "grad_norm": 0.12934917211532593, + "learning_rate": 4.802616207634947e-08, + "loss": 0.3489, + "step": 1305 + }, + { + "epoch": 2.883002207505519, + "grad_norm": 0.1276032030582428, + "learning_rate": 4.626660816841433e-08, + "loss": 0.3554, + "step": 1306 + }, + { + "epoch": 2.8852097130242824, + "grad_norm": 0.12852588295936584, + "learning_rate": 4.4539741573489395e-08, + "loss": 0.3519, + "step": 1307 + }, + { + "epoch": 2.8874172185430464, + "grad_norm": 0.1251831203699112, + "learning_rate": 4.2845573686339235e-08, + "loss": 0.3586, + "step": 1308 + }, + { + "epoch": 2.88962472406181, + "grad_norm": 0.12635353207588196, + "learning_rate": 4.118411568596714e-08, + "loss": 0.3521, + "step": 1309 + }, + { + "epoch": 2.891832229580574, + "grad_norm": 0.13549165427684784, + "learning_rate": 3.9555378535537925e-08, + "loss": 0.3573, + "step": 1310 + }, + { + "epoch": 2.8940397350993377, + "grad_norm": 0.12129613012075424, + "learning_rate": 3.795937298230801e-08, + "loss": 0.3644, + "step": 1311 + }, + { + "epoch": 2.8962472406181017, + "grad_norm": 0.12583400309085846, + "learning_rate": 3.639610955755213e-08, + "loss": 0.3628, + "step": 1312 + }, + { + "epoch": 2.8984547461368653, + "grad_norm": 0.11590792238712311, + "learning_rate": 3.486559857649785e-08, + "loss": 0.3541, + "step": 1313 + }, + { + "epoch": 2.9006622516556293, + "grad_norm": 0.1254376322031021, + "learning_rate": 3.336785013825339e-08, + "loss": 0.3591, + "step": 1314 + }, + { + "epoch": 2.902869757174393, + "grad_norm": 0.12448057532310486, + "learning_rate": 3.190287412574267e-08, + "loss": 0.3685, + "step": 1315 + }, + { + "epoch": 2.9050772626931565, + "grad_norm": 0.12795038521289825, + "learning_rate": 3.047068020564037e-08, + "loss": 0.3661, + "step": 1316 + }, + { + "epoch": 2.9072847682119205, + "grad_norm": 0.1275874376296997, + "learning_rate": 2.9071277828308654e-08, + "loss": 0.3547, + "step": 1317 + }, + { + "epoch": 2.9094922737306845, + "grad_norm": 0.13140250742435455, + "learning_rate": 2.7704676227732764e-08, + "loss": 0.3476, + "step": 1318 + }, + { + "epoch": 2.911699779249448, + "grad_norm": 0.13054263591766357, + "learning_rate": 2.637088442146163e-08, + "loss": 0.3555, + "step": 1319 + }, + { + "epoch": 2.9139072847682117, + "grad_norm": 0.14265048503875732, + "learning_rate": 2.506991121054847e-08, + "loss": 0.3552, + "step": 1320 + }, + { + "epoch": 2.9161147902869757, + "grad_norm": 0.12008768320083618, + "learning_rate": 2.380176517949251e-08, + "loss": 0.3555, + "step": 1321 + }, + { + "epoch": 2.9183222958057398, + "grad_norm": 0.13723327219486237, + "learning_rate": 2.256645469618124e-08, + "loss": 0.3578, + "step": 1322 + }, + { + "epoch": 2.9205298013245033, + "grad_norm": 0.13760052621364594, + "learning_rate": 2.136398791183658e-08, + "loss": 0.3543, + "step": 1323 + }, + { + "epoch": 2.922737306843267, + "grad_norm": 0.129390150308609, + "learning_rate": 2.0194372760961034e-08, + "loss": 0.362, + "step": 1324 + }, + { + "epoch": 2.924944812362031, + "grad_norm": 0.15059252083301544, + "learning_rate": 1.905761696128494e-08, + "loss": 0.3611, + "step": 1325 + }, + { + "epoch": 2.9271523178807946, + "grad_norm": 0.13671807944774628, + "learning_rate": 1.795372801371431e-08, + "loss": 0.3601, + "step": 1326 + }, + { + "epoch": 2.9293598233995586, + "grad_norm": 0.138703852891922, + "learning_rate": 1.6882713202283076e-08, + "loss": 0.3525, + "step": 1327 + }, + { + "epoch": 2.931567328918322, + "grad_norm": 0.12975488603115082, + "learning_rate": 1.5844579594105904e-08, + "loss": 0.3636, + "step": 1328 + }, + { + "epoch": 2.933774834437086, + "grad_norm": 0.15756423771381378, + "learning_rate": 1.48393340393288e-08, + "loss": 0.3586, + "step": 1329 + }, + { + "epoch": 2.93598233995585, + "grad_norm": 0.13145272433757782, + "learning_rate": 1.3866983171084703e-08, + "loss": 0.3563, + "step": 1330 + }, + { + "epoch": 2.938189845474614, + "grad_norm": 0.15646247565746307, + "learning_rate": 1.292753340545183e-08, + "loss": 0.3567, + "step": 1331 + }, + { + "epoch": 2.9403973509933774, + "grad_norm": 0.1439492553472519, + "learning_rate": 1.2020990941408739e-08, + "loss": 0.3678, + "step": 1332 + }, + { + "epoch": 2.9426048565121414, + "grad_norm": 0.1377602368593216, + "learning_rate": 1.1147361760794895e-08, + "loss": 0.3547, + "step": 1333 + }, + { + "epoch": 2.944812362030905, + "grad_norm": 0.14668287336826324, + "learning_rate": 1.0306651628270715e-08, + "loss": 0.3672, + "step": 1334 + }, + { + "epoch": 2.9470198675496686, + "grad_norm": 0.12802091240882874, + "learning_rate": 9.498866091278702e-09, + "loss": 0.3596, + "step": 1335 + }, + { + "epoch": 2.9492273730684326, + "grad_norm": 0.15803688764572144, + "learning_rate": 8.724010480010147e-09, + "loss": 0.3586, + "step": 1336 + }, + { + "epoch": 2.9514348785871967, + "grad_norm": 0.13727080821990967, + "learning_rate": 7.982089907364598e-09, + "loss": 0.3515, + "step": 1337 + }, + { + "epoch": 2.9536423841059603, + "grad_norm": 0.14836713671684265, + "learning_rate": 7.273109268920997e-09, + "loss": 0.3572, + "step": 1338 + }, + { + "epoch": 2.955849889624724, + "grad_norm": 0.15154053270816803, + "learning_rate": 6.597073242902707e-09, + "loss": 0.3632, + "step": 1339 + }, + { + "epoch": 2.958057395143488, + "grad_norm": 0.14268819987773895, + "learning_rate": 5.95398629014754e-09, + "loss": 0.3737, + "step": 1340 + }, + { + "epoch": 2.960264900662252, + "grad_norm": 0.13592885434627533, + "learning_rate": 5.3438526540777745e-09, + "loss": 0.3647, + "step": 1341 + }, + { + "epoch": 2.9624724061810155, + "grad_norm": 0.15282437205314636, + "learning_rate": 4.766676360674072e-09, + "loss": 0.3666, + "step": 1342 + }, + { + "epoch": 2.964679911699779, + "grad_norm": 0.14543361961841583, + "learning_rate": 4.22246121844494e-09, + "loss": 0.3576, + "step": 1343 + }, + { + "epoch": 2.966887417218543, + "grad_norm": 0.1356423795223236, + "learning_rate": 3.7112108184061966e-09, + "loss": 0.3646, + "step": 1344 + }, + { + "epoch": 2.9690949227373067, + "grad_norm": 0.13516850769519806, + "learning_rate": 3.2329285340537696e-09, + "loss": 0.3577, + "step": 1345 + }, + { + "epoch": 2.9713024282560707, + "grad_norm": 0.12816190719604492, + "learning_rate": 2.7876175213431557e-09, + "loss": 0.3568, + "step": 1346 + }, + { + "epoch": 2.9735099337748343, + "grad_norm": 0.13490551710128784, + "learning_rate": 2.375280718668882e-09, + "loss": 0.3548, + "step": 1347 + }, + { + "epoch": 2.9757174392935983, + "grad_norm": 0.14126001298427582, + "learning_rate": 1.995920846843968e-09, + "loss": 0.3666, + "step": 1348 + }, + { + "epoch": 2.977924944812362, + "grad_norm": 0.12279074639081955, + "learning_rate": 1.649540409081607e-09, + "loss": 0.3562, + "step": 1349 + }, + { + "epoch": 2.980132450331126, + "grad_norm": 0.12752176821231842, + "learning_rate": 1.3361416909812852e-09, + "loss": 0.365, + "step": 1350 + }, + { + "epoch": 2.9823399558498895, + "grad_norm": 0.13065557181835175, + "learning_rate": 1.055726760510467e-09, + "loss": 0.3615, + "step": 1351 + }, + { + "epoch": 2.9845474613686536, + "grad_norm": 0.11860226094722748, + "learning_rate": 8.082974679929357e-10, + "loss": 0.3564, + "step": 1352 + }, + { + "epoch": 2.986754966887417, + "grad_norm": 0.1291607767343521, + "learning_rate": 5.938554460965807e-10, + "loss": 0.359, + "step": 1353 + }, + { + "epoch": 2.988962472406181, + "grad_norm": 0.146859273314476, + "learning_rate": 4.12402109820631e-10, + "loss": 0.3614, + "step": 1354 + }, + { + "epoch": 2.9911699779249448, + "grad_norm": 0.12649129331111908, + "learning_rate": 2.639386564889934e-10, + "loss": 0.3566, + "step": 1355 + }, + { + "epoch": 2.993377483443709, + "grad_norm": 0.14126034080982208, + "learning_rate": 1.4846606574137058e-10, + "loss": 0.3608, + "step": 1356 + }, + { + "epoch": 2.9955849889624724, + "grad_norm": 0.13480675220489502, + "learning_rate": 6.59850995254896e-11, + "loss": 0.3555, + "step": 1357 + }, + { + "epoch": 2.997792494481236, + "grad_norm": 0.14647357165813446, + "learning_rate": 1.649630209432651e-11, + "loss": 0.3643, + "step": 1358 + }, + { + "epoch": 3.0, + "grad_norm": 0.13931505382061005, + "learning_rate": 0.0, + "loss": 0.3486, + "step": 1359 + }, + { + "epoch": 3.0, + "step": 1359, + "total_flos": 2.7571726585954304e+16, + "train_loss": 0.13500037013037053, + "train_runtime": 42654.1727, + "train_samples_per_second": 12.208, + "train_steps_per_second": 0.032 + } + ], + "logging_steps": 1, + "max_steps": 1359, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 50, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 2.7571726585954304e+16, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}