diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,147034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.6623560952531147, + "eval_steps": 3000.0, + "global_step": 21000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 3.154076644062451e-05, + "grad_norm": 17.829654693603516, + "learning_rate": 0.0, + "loss": 1.3036, + "step": 1 + }, + { + "epoch": 6.308153288124902e-05, + "grad_norm": 136.14352416992188, + "learning_rate": 2.1008403361344538e-08, + "loss": 4.1435, + "step": 2 + }, + { + "epoch": 9.462229932187352e-05, + "grad_norm": 126.18865203857422, + "learning_rate": 4.2016806722689076e-08, + "loss": 4.1685, + "step": 3 + }, + { + "epoch": 0.00012616306576249803, + "grad_norm": 18.490768432617188, + "learning_rate": 6.302521008403361e-08, + "loss": 1.1148, + "step": 4 + }, + { + "epoch": 0.00015770383220312253, + "grad_norm": 156.1005401611328, + "learning_rate": 8.403361344537815e-08, + "loss": 4.1809, + "step": 5 + }, + { + "epoch": 0.00018924459864374705, + "grad_norm": 33.66926193237305, + "learning_rate": 1.050420168067227e-07, + "loss": 1.2829, + "step": 6 + }, + { + "epoch": 0.00022078536508437154, + "grad_norm": 185.52256774902344, + "learning_rate": 1.2605042016806723e-07, + "loss": 4.2487, + "step": 7 + }, + { + "epoch": 0.00025232613152499606, + "grad_norm": 156.76429748535156, + "learning_rate": 1.4705882352941178e-07, + "loss": 4.2101, + "step": 8 + }, + { + "epoch": 0.00028386689796562056, + "grad_norm": 179.28318786621094, + "learning_rate": 1.680672268907563e-07, + "loss": 4.1926, + "step": 9 + }, + { + "epoch": 0.00031540766440624505, + "grad_norm": 19.095914840698242, + "learning_rate": 1.8907563025210085e-07, + "loss": 1.3039, + "step": 10 + }, + { + "epoch": 0.0003469484308468696, + "grad_norm": 17.24628257751465, + "learning_rate": 2.100840336134454e-07, + "loss": 1.2815, + "step": 11 + }, + { + "epoch": 0.0003784891972874941, + "grad_norm": 138.7112579345703, + "learning_rate": 2.3109243697478996e-07, + "loss": 4.062, + "step": 12 + }, + { + "epoch": 0.0004100299637281186, + "grad_norm": 151.43966674804688, + "learning_rate": 2.5210084033613445e-07, + "loss": 3.8012, + "step": 13 + }, + { + "epoch": 0.0004415707301687431, + "grad_norm": 138.11441040039062, + "learning_rate": 2.73109243697479e-07, + "loss": 4.0031, + "step": 14 + }, + { + "epoch": 0.00047311149660936763, + "grad_norm": 18.08880615234375, + "learning_rate": 2.9411764705882356e-07, + "loss": 1.2923, + "step": 15 + }, + { + "epoch": 0.0005046522630499921, + "grad_norm": 155.21934509277344, + "learning_rate": 3.151260504201681e-07, + "loss": 4.1053, + "step": 16 + }, + { + "epoch": 0.0005361930294906167, + "grad_norm": 242.17529296875, + "learning_rate": 3.361344537815126e-07, + "loss": 3.9982, + "step": 17 + }, + { + "epoch": 0.0005677337959312411, + "grad_norm": 118.20597839355469, + "learning_rate": 3.5714285714285716e-07, + "loss": 3.9293, + "step": 18 + }, + { + "epoch": 0.0005992745623718657, + "grad_norm": 178.90406799316406, + "learning_rate": 3.781512605042017e-07, + "loss": 3.8774, + "step": 19 + }, + { + "epoch": 0.0006308153288124901, + "grad_norm": 18.795320510864258, + "learning_rate": 3.991596638655462e-07, + "loss": 1.2473, + "step": 20 + }, + { + "epoch": 0.0006623560952531147, + "grad_norm": 111.31549835205078, + "learning_rate": 4.201680672268908e-07, + "loss": 3.839, + "step": 21 + }, + { + "epoch": 0.0006938968616937392, + "grad_norm": 271.73297119140625, + "learning_rate": 4.4117647058823536e-07, + "loss": 3.6073, + "step": 22 + }, + { + "epoch": 0.0007254376281343636, + "grad_norm": 111.92586517333984, + "learning_rate": 4.621848739495799e-07, + "loss": 3.3905, + "step": 23 + }, + { + "epoch": 0.0007569783945749882, + "grad_norm": 102.03913879394531, + "learning_rate": 4.831932773109245e-07, + "loss": 3.1997, + "step": 24 + }, + { + "epoch": 0.0007885191610156126, + "grad_norm": 99.17202758789062, + "learning_rate": 5.042016806722689e-07, + "loss": 3.146, + "step": 25 + }, + { + "epoch": 0.0008200599274562372, + "grad_norm": 96.19882202148438, + "learning_rate": 5.252100840336135e-07, + "loss": 3.1491, + "step": 26 + }, + { + "epoch": 0.0008516006938968617, + "grad_norm": 112.6724624633789, + "learning_rate": 5.46218487394958e-07, + "loss": 3.119, + "step": 27 + }, + { + "epoch": 0.0008831414603374862, + "grad_norm": 136.2877655029297, + "learning_rate": 5.672268907563026e-07, + "loss": 3.0227, + "step": 28 + }, + { + "epoch": 0.0009146822267781107, + "grad_norm": 130.71543884277344, + "learning_rate": 5.882352941176471e-07, + "loss": 2.8866, + "step": 29 + }, + { + "epoch": 0.0009462229932187353, + "grad_norm": 98.2926254272461, + "learning_rate": 6.092436974789917e-07, + "loss": 2.5871, + "step": 30 + }, + { + "epoch": 0.0009777637596593597, + "grad_norm": 95.67642974853516, + "learning_rate": 6.302521008403362e-07, + "loss": 2.3773, + "step": 31 + }, + { + "epoch": 0.0010093045260999843, + "grad_norm": 14.824779510498047, + "learning_rate": 6.512605042016808e-07, + "loss": 1.2172, + "step": 32 + }, + { + "epoch": 0.0010408452925406088, + "grad_norm": 94.4736099243164, + "learning_rate": 6.722689075630252e-07, + "loss": 2.0834, + "step": 33 + }, + { + "epoch": 0.0010723860589812334, + "grad_norm": 95.37779998779297, + "learning_rate": 6.932773109243699e-07, + "loss": 1.9831, + "step": 34 + }, + { + "epoch": 0.0011039268254218577, + "grad_norm": 87.09550476074219, + "learning_rate": 7.142857142857143e-07, + "loss": 1.9913, + "step": 35 + }, + { + "epoch": 0.0011354675918624822, + "grad_norm": 89.98584747314453, + "learning_rate": 7.352941176470589e-07, + "loss": 1.8646, + "step": 36 + }, + { + "epoch": 0.0011670083583031068, + "grad_norm": 69.81629180908203, + "learning_rate": 7.563025210084034e-07, + "loss": 1.8265, + "step": 37 + }, + { + "epoch": 0.0011985491247437313, + "grad_norm": 70.90291595458984, + "learning_rate": 7.77310924369748e-07, + "loss": 1.8146, + "step": 38 + }, + { + "epoch": 0.0012300898911843559, + "grad_norm": 65.33815002441406, + "learning_rate": 7.983193277310924e-07, + "loss": 1.6884, + "step": 39 + }, + { + "epoch": 0.0012616306576249802, + "grad_norm": 14.18765926361084, + "learning_rate": 8.193277310924371e-07, + "loss": 1.1954, + "step": 40 + }, + { + "epoch": 0.0012931714240656048, + "grad_norm": 20.434709548950195, + "learning_rate": 8.403361344537816e-07, + "loss": 1.0427, + "step": 41 + }, + { + "epoch": 0.0013247121905062293, + "grad_norm": 15.858318328857422, + "learning_rate": 8.613445378151261e-07, + "loss": 1.0164, + "step": 42 + }, + { + "epoch": 0.0013562529569468539, + "grad_norm": 11.342897415161133, + "learning_rate": 8.823529411764707e-07, + "loss": 1.1202, + "step": 43 + }, + { + "epoch": 0.0013877937233874784, + "grad_norm": 71.59555053710938, + "learning_rate": 9.033613445378152e-07, + "loss": 1.4088, + "step": 44 + }, + { + "epoch": 0.0014193344898281027, + "grad_norm": 46.9373664855957, + "learning_rate": 9.243697478991598e-07, + "loss": 1.3894, + "step": 45 + }, + { + "epoch": 0.0014508752562687273, + "grad_norm": 75.53413391113281, + "learning_rate": 9.453781512605043e-07, + "loss": 1.3547, + "step": 46 + }, + { + "epoch": 0.0014824160227093518, + "grad_norm": 43.245845794677734, + "learning_rate": 9.66386554621849e-07, + "loss": 1.2923, + "step": 47 + }, + { + "epoch": 0.0015139567891499764, + "grad_norm": 57.60647964477539, + "learning_rate": 9.873949579831934e-07, + "loss": 1.2568, + "step": 48 + }, + { + "epoch": 0.001545497555590601, + "grad_norm": 59.74352264404297, + "learning_rate": 1.0084033613445378e-06, + "loss": 1.1942, + "step": 49 + }, + { + "epoch": 0.0015770383220312253, + "grad_norm": 62.66681671142578, + "learning_rate": 1.0294117647058825e-06, + "loss": 1.171, + "step": 50 + }, + { + "epoch": 0.0016085790884718498, + "grad_norm": 43.673370361328125, + "learning_rate": 1.050420168067227e-06, + "loss": 1.1496, + "step": 51 + }, + { + "epoch": 0.0016401198549124744, + "grad_norm": 38.54374313354492, + "learning_rate": 1.0714285714285714e-06, + "loss": 1.0775, + "step": 52 + }, + { + "epoch": 0.001671660621353099, + "grad_norm": 34.60348892211914, + "learning_rate": 1.092436974789916e-06, + "loss": 1.0749, + "step": 53 + }, + { + "epoch": 0.0017032013877937235, + "grad_norm": 32.78680419921875, + "learning_rate": 1.1134453781512607e-06, + "loss": 1.0561, + "step": 54 + }, + { + "epoch": 0.0017347421542343478, + "grad_norm": 51.71357727050781, + "learning_rate": 1.1344537815126051e-06, + "loss": 1.0482, + "step": 55 + }, + { + "epoch": 0.0017662829206749723, + "grad_norm": 39.96931076049805, + "learning_rate": 1.1554621848739498e-06, + "loss": 1.0323, + "step": 56 + }, + { + "epoch": 0.0017978236871155969, + "grad_norm": 37.50736999511719, + "learning_rate": 1.1764705882352942e-06, + "loss": 1.0329, + "step": 57 + }, + { + "epoch": 0.0018293644535562214, + "grad_norm": 45.65388488769531, + "learning_rate": 1.1974789915966389e-06, + "loss": 0.9941, + "step": 58 + }, + { + "epoch": 0.001860905219996846, + "grad_norm": 41.10524368286133, + "learning_rate": 1.2184873949579833e-06, + "loss": 1.003, + "step": 59 + }, + { + "epoch": 0.0018924459864374705, + "grad_norm": 35.43183898925781, + "learning_rate": 1.2394957983193278e-06, + "loss": 1.0188, + "step": 60 + }, + { + "epoch": 0.0019239867528780949, + "grad_norm": 40.335121154785156, + "learning_rate": 1.2605042016806724e-06, + "loss": 0.9942, + "step": 61 + }, + { + "epoch": 0.0019555275193187194, + "grad_norm": 33.97624969482422, + "learning_rate": 1.2815126050420169e-06, + "loss": 0.9884, + "step": 62 + }, + { + "epoch": 0.001987068285759344, + "grad_norm": 27.822330474853516, + "learning_rate": 1.3025210084033615e-06, + "loss": 0.9862, + "step": 63 + }, + { + "epoch": 0.0020186090521999685, + "grad_norm": 24.771451950073242, + "learning_rate": 1.323529411764706e-06, + "loss": 0.8882, + "step": 64 + }, + { + "epoch": 0.002050149818640593, + "grad_norm": 27.575101852416992, + "learning_rate": 1.3445378151260504e-06, + "loss": 0.9395, + "step": 65 + }, + { + "epoch": 0.0020816905850812176, + "grad_norm": 27.61154556274414, + "learning_rate": 1.3655462184873949e-06, + "loss": 0.9596, + "step": 66 + }, + { + "epoch": 0.002113231351521842, + "grad_norm": 55.6931266784668, + "learning_rate": 1.3865546218487397e-06, + "loss": 0.9482, + "step": 67 + }, + { + "epoch": 0.0021447721179624667, + "grad_norm": 33.77639389038086, + "learning_rate": 1.4075630252100842e-06, + "loss": 0.9396, + "step": 68 + }, + { + "epoch": 0.002176312884403091, + "grad_norm": 27.59514045715332, + "learning_rate": 1.4285714285714286e-06, + "loss": 0.9956, + "step": 69 + }, + { + "epoch": 0.0022078536508437154, + "grad_norm": 8.669937133789062, + "learning_rate": 1.4495798319327733e-06, + "loss": 1.0573, + "step": 70 + }, + { + "epoch": 0.00223939441728434, + "grad_norm": 44.95465087890625, + "learning_rate": 1.4705882352941177e-06, + "loss": 0.9556, + "step": 71 + }, + { + "epoch": 0.0022709351837249645, + "grad_norm": 65.41173553466797, + "learning_rate": 1.4915966386554624e-06, + "loss": 0.9505, + "step": 72 + }, + { + "epoch": 0.002302475950165589, + "grad_norm": 27.838623046875, + "learning_rate": 1.5126050420168068e-06, + "loss": 0.9365, + "step": 73 + }, + { + "epoch": 0.0023340167166062136, + "grad_norm": 39.498382568359375, + "learning_rate": 1.5336134453781515e-06, + "loss": 0.8273, + "step": 74 + }, + { + "epoch": 0.002365557483046838, + "grad_norm": 33.376495361328125, + "learning_rate": 1.554621848739496e-06, + "loss": 0.9271, + "step": 75 + }, + { + "epoch": 0.0023970982494874627, + "grad_norm": 7.855008602142334, + "learning_rate": 1.5756302521008404e-06, + "loss": 1.1351, + "step": 76 + }, + { + "epoch": 0.002428639015928087, + "grad_norm": 51.9299201965332, + "learning_rate": 1.5966386554621848e-06, + "loss": 0.931, + "step": 77 + }, + { + "epoch": 0.0024601797823687118, + "grad_norm": 35.14694595336914, + "learning_rate": 1.6176470588235297e-06, + "loss": 0.9127, + "step": 78 + }, + { + "epoch": 0.002491720548809336, + "grad_norm": 25.30535125732422, + "learning_rate": 1.6386554621848741e-06, + "loss": 0.9338, + "step": 79 + }, + { + "epoch": 0.0025232613152499604, + "grad_norm": 27.477354049682617, + "learning_rate": 1.6596638655462186e-06, + "loss": 0.9681, + "step": 80 + }, + { + "epoch": 0.002554802081690585, + "grad_norm": 25.782161712646484, + "learning_rate": 1.6806722689075632e-06, + "loss": 0.9185, + "step": 81 + }, + { + "epoch": 0.0025863428481312095, + "grad_norm": 32.20173645019531, + "learning_rate": 1.7016806722689077e-06, + "loss": 0.9147, + "step": 82 + }, + { + "epoch": 0.002617883614571834, + "grad_norm": 27.23711585998535, + "learning_rate": 1.7226890756302521e-06, + "loss": 0.9199, + "step": 83 + }, + { + "epoch": 0.0026494243810124586, + "grad_norm": 33.82099151611328, + "learning_rate": 1.7436974789915968e-06, + "loss": 0.914, + "step": 84 + }, + { + "epoch": 0.002680965147453083, + "grad_norm": 20.074338912963867, + "learning_rate": 1.7647058823529414e-06, + "loss": 0.9005, + "step": 85 + }, + { + "epoch": 0.0027125059138937077, + "grad_norm": 21.987897872924805, + "learning_rate": 1.7857142857142859e-06, + "loss": 0.9411, + "step": 86 + }, + { + "epoch": 0.0027440466803343323, + "grad_norm": 17.692150115966797, + "learning_rate": 1.8067226890756303e-06, + "loss": 0.6567, + "step": 87 + }, + { + "epoch": 0.002775587446774957, + "grad_norm": 26.075424194335938, + "learning_rate": 1.8277310924369748e-06, + "loss": 0.9159, + "step": 88 + }, + { + "epoch": 0.0028071282132155814, + "grad_norm": 45.72597885131836, + "learning_rate": 1.8487394957983196e-06, + "loss": 0.9109, + "step": 89 + }, + { + "epoch": 0.0028386689796562055, + "grad_norm": 14.051458358764648, + "learning_rate": 1.869747899159664e-06, + "loss": 0.6142, + "step": 90 + }, + { + "epoch": 0.00287020974609683, + "grad_norm": 29.187503814697266, + "learning_rate": 1.8907563025210085e-06, + "loss": 0.9181, + "step": 91 + }, + { + "epoch": 0.0029017505125374546, + "grad_norm": 16.32326889038086, + "learning_rate": 1.9117647058823528e-06, + "loss": 1.0714, + "step": 92 + }, + { + "epoch": 0.002933291278978079, + "grad_norm": 25.717138290405273, + "learning_rate": 1.932773109243698e-06, + "loss": 0.9181, + "step": 93 + }, + { + "epoch": 0.0029648320454187037, + "grad_norm": 30.00885009765625, + "learning_rate": 1.953781512605042e-06, + "loss": 0.8804, + "step": 94 + }, + { + "epoch": 0.002996372811859328, + "grad_norm": 6.440777778625488, + "learning_rate": 1.9747899159663867e-06, + "loss": 1.0558, + "step": 95 + }, + { + "epoch": 0.0030279135782999528, + "grad_norm": 29.179784774780273, + "learning_rate": 1.9957983193277314e-06, + "loss": 0.9052, + "step": 96 + }, + { + "epoch": 0.0030594543447405773, + "grad_norm": 27.012910842895508, + "learning_rate": 2.0168067226890756e-06, + "loss": 0.9011, + "step": 97 + }, + { + "epoch": 0.003090995111181202, + "grad_norm": 26.367706298828125, + "learning_rate": 2.0378151260504203e-06, + "loss": 0.8818, + "step": 98 + }, + { + "epoch": 0.0031225358776218264, + "grad_norm": 28.018714904785156, + "learning_rate": 2.058823529411765e-06, + "loss": 0.9398, + "step": 99 + }, + { + "epoch": 0.0031540766440624505, + "grad_norm": 25.49843406677246, + "learning_rate": 2.0798319327731096e-06, + "loss": 0.928, + "step": 100 + }, + { + "epoch": 0.003185617410503075, + "grad_norm": 9.962201118469238, + "learning_rate": 2.100840336134454e-06, + "loss": 1.0361, + "step": 101 + }, + { + "epoch": 0.0032171581769436996, + "grad_norm": 46.84260177612305, + "learning_rate": 2.1218487394957985e-06, + "loss": 0.8697, + "step": 102 + }, + { + "epoch": 0.003248698943384324, + "grad_norm": 33.8751106262207, + "learning_rate": 2.1428571428571427e-06, + "loss": 0.8639, + "step": 103 + }, + { + "epoch": 0.0032802397098249487, + "grad_norm": 127.12957000732422, + "learning_rate": 2.163865546218488e-06, + "loss": 0.8813, + "step": 104 + }, + { + "epoch": 0.0033117804762655733, + "grad_norm": 26.397226333618164, + "learning_rate": 2.184873949579832e-06, + "loss": 0.9152, + "step": 105 + }, + { + "epoch": 0.003343321242706198, + "grad_norm": 36.23981475830078, + "learning_rate": 2.2058823529411767e-06, + "loss": 0.9315, + "step": 106 + }, + { + "epoch": 0.0033748620091468224, + "grad_norm": 25.722124099731445, + "learning_rate": 2.2268907563025214e-06, + "loss": 0.8789, + "step": 107 + }, + { + "epoch": 0.003406402775587447, + "grad_norm": 4.514924049377441, + "learning_rate": 2.2478991596638656e-06, + "loss": 1.0288, + "step": 108 + }, + { + "epoch": 0.0034379435420280715, + "grad_norm": 29.84398078918457, + "learning_rate": 2.2689075630252102e-06, + "loss": 0.8858, + "step": 109 + }, + { + "epoch": 0.0034694843084686956, + "grad_norm": 23.011783599853516, + "learning_rate": 2.289915966386555e-06, + "loss": 0.8713, + "step": 110 + }, + { + "epoch": 0.00350102507490932, + "grad_norm": 35.55387496948242, + "learning_rate": 2.3109243697478996e-06, + "loss": 0.8741, + "step": 111 + }, + { + "epoch": 0.0035325658413499447, + "grad_norm": 6.092860221862793, + "learning_rate": 2.331932773109244e-06, + "loss": 1.0228, + "step": 112 + }, + { + "epoch": 0.0035641066077905692, + "grad_norm": 35.57551574707031, + "learning_rate": 2.3529411764705885e-06, + "loss": 0.8435, + "step": 113 + }, + { + "epoch": 0.0035956473742311938, + "grad_norm": 29.227367401123047, + "learning_rate": 2.3739495798319327e-06, + "loss": 0.8881, + "step": 114 + }, + { + "epoch": 0.0036271881406718183, + "grad_norm": 48.567237854003906, + "learning_rate": 2.3949579831932778e-06, + "loss": 0.8539, + "step": 115 + }, + { + "epoch": 0.003658728907112443, + "grad_norm": 24.709644317626953, + "learning_rate": 2.415966386554622e-06, + "loss": 0.8656, + "step": 116 + }, + { + "epoch": 0.0036902696735530674, + "grad_norm": 35.564697265625, + "learning_rate": 2.4369747899159667e-06, + "loss": 0.848, + "step": 117 + }, + { + "epoch": 0.003721810439993692, + "grad_norm": 39.81562805175781, + "learning_rate": 2.457983193277311e-06, + "loss": 0.8492, + "step": 118 + }, + { + "epoch": 0.0037533512064343165, + "grad_norm": 30.730222702026367, + "learning_rate": 2.4789915966386555e-06, + "loss": 0.8654, + "step": 119 + }, + { + "epoch": 0.003784891972874941, + "grad_norm": 22.309707641601562, + "learning_rate": 2.5e-06, + "loss": 0.8939, + "step": 120 + }, + { + "epoch": 0.003816432739315565, + "grad_norm": 22.79766845703125, + "learning_rate": 2.521008403361345e-06, + "loss": 0.8613, + "step": 121 + }, + { + "epoch": 0.0038479735057561897, + "grad_norm": 27.831905364990234, + "learning_rate": 2.542016806722689e-06, + "loss": 0.8867, + "step": 122 + }, + { + "epoch": 0.0038795142721968143, + "grad_norm": 30.341266632080078, + "learning_rate": 2.5630252100840338e-06, + "loss": 0.8962, + "step": 123 + }, + { + "epoch": 0.003911055038637439, + "grad_norm": 4.903833866119385, + "learning_rate": 2.5840336134453784e-06, + "loss": 0.5375, + "step": 124 + }, + { + "epoch": 0.003942595805078063, + "grad_norm": 27.45075225830078, + "learning_rate": 2.605042016806723e-06, + "loss": 0.8793, + "step": 125 + }, + { + "epoch": 0.003974136571518688, + "grad_norm": 41.4337043762207, + "learning_rate": 2.6260504201680677e-06, + "loss": 0.8837, + "step": 126 + }, + { + "epoch": 0.0040056773379593125, + "grad_norm": 54.159996032714844, + "learning_rate": 2.647058823529412e-06, + "loss": 0.8581, + "step": 127 + }, + { + "epoch": 0.004037218104399937, + "grad_norm": 24.742799758911133, + "learning_rate": 2.6680672268907566e-06, + "loss": 0.876, + "step": 128 + }, + { + "epoch": 0.0040687588708405616, + "grad_norm": 54.16761016845703, + "learning_rate": 2.689075630252101e-06, + "loss": 0.8801, + "step": 129 + }, + { + "epoch": 0.004100299637281186, + "grad_norm": 5.14218807220459, + "learning_rate": 2.7100840336134455e-06, + "loss": 0.9849, + "step": 130 + }, + { + "epoch": 0.004131840403721811, + "grad_norm": 82.09132385253906, + "learning_rate": 2.7310924369747897e-06, + "loss": 0.5486, + "step": 131 + }, + { + "epoch": 0.004163381170162435, + "grad_norm": 176.45826721191406, + "learning_rate": 2.752100840336135e-06, + "loss": 0.8494, + "step": 132 + }, + { + "epoch": 0.00419492193660306, + "grad_norm": 33.67400360107422, + "learning_rate": 2.7731092436974795e-06, + "loss": 0.8441, + "step": 133 + }, + { + "epoch": 0.004226462703043684, + "grad_norm": 21.21440315246582, + "learning_rate": 2.7941176470588237e-06, + "loss": 0.8288, + "step": 134 + }, + { + "epoch": 0.004258003469484309, + "grad_norm": 56.57637023925781, + "learning_rate": 2.8151260504201684e-06, + "loss": 0.8369, + "step": 135 + }, + { + "epoch": 0.004289544235924933, + "grad_norm": 3.8961353302001953, + "learning_rate": 2.8361344537815126e-06, + "loss": 0.9397, + "step": 136 + }, + { + "epoch": 0.004321085002365557, + "grad_norm": 40.76416778564453, + "learning_rate": 2.8571428571428573e-06, + "loss": 0.8611, + "step": 137 + }, + { + "epoch": 0.004352625768806182, + "grad_norm": 30.812782287597656, + "learning_rate": 2.8781512605042015e-06, + "loss": 0.8485, + "step": 138 + }, + { + "epoch": 0.004384166535246806, + "grad_norm": 75.35914611816406, + "learning_rate": 2.8991596638655466e-06, + "loss": 0.8706, + "step": 139 + }, + { + "epoch": 0.004415707301687431, + "grad_norm": 82.08843994140625, + "learning_rate": 2.9201680672268912e-06, + "loss": 0.875, + "step": 140 + }, + { + "epoch": 0.004447248068128055, + "grad_norm": 30.000667572021484, + "learning_rate": 2.9411764705882355e-06, + "loss": 0.8279, + "step": 141 + }, + { + "epoch": 0.00447878883456868, + "grad_norm": 70.68790435791016, + "learning_rate": 2.96218487394958e-06, + "loss": 0.8492, + "step": 142 + }, + { + "epoch": 0.004510329601009304, + "grad_norm": 31.08045768737793, + "learning_rate": 2.9831932773109248e-06, + "loss": 0.866, + "step": 143 + }, + { + "epoch": 0.004541870367449929, + "grad_norm": 39.32219696044922, + "learning_rate": 3.004201680672269e-06, + "loss": 0.8438, + "step": 144 + }, + { + "epoch": 0.0045734111338905535, + "grad_norm": 23.48110580444336, + "learning_rate": 3.0252100840336137e-06, + "loss": 0.8208, + "step": 145 + }, + { + "epoch": 0.004604951900331178, + "grad_norm": 12.772490501403809, + "learning_rate": 3.046218487394958e-06, + "loss": 0.5163, + "step": 146 + }, + { + "epoch": 0.004636492666771803, + "grad_norm": 38.16356658935547, + "learning_rate": 3.067226890756303e-06, + "loss": 0.8144, + "step": 147 + }, + { + "epoch": 0.004668033433212427, + "grad_norm": 9.758743286132812, + "learning_rate": 3.0882352941176476e-06, + "loss": 0.9156, + "step": 148 + }, + { + "epoch": 0.004699574199653052, + "grad_norm": 27.920076370239258, + "learning_rate": 3.109243697478992e-06, + "loss": 0.9066, + "step": 149 + }, + { + "epoch": 0.004731114966093676, + "grad_norm": 24.147111892700195, + "learning_rate": 3.1302521008403365e-06, + "loss": 0.8411, + "step": 150 + }, + { + "epoch": 0.004762655732534301, + "grad_norm": 31.514606475830078, + "learning_rate": 3.1512605042016808e-06, + "loss": 0.8399, + "step": 151 + }, + { + "epoch": 0.004794196498974925, + "grad_norm": 20.47402000427246, + "learning_rate": 3.1722689075630254e-06, + "loss": 0.8158, + "step": 152 + }, + { + "epoch": 0.00482573726541555, + "grad_norm": 87.35063171386719, + "learning_rate": 3.1932773109243696e-06, + "loss": 0.8235, + "step": 153 + }, + { + "epoch": 0.004857278031856174, + "grad_norm": 78.07725524902344, + "learning_rate": 3.2142857142857147e-06, + "loss": 0.8154, + "step": 154 + }, + { + "epoch": 0.004888818798296799, + "grad_norm": 40.38902282714844, + "learning_rate": 3.2352941176470594e-06, + "loss": 0.8329, + "step": 155 + }, + { + "epoch": 0.0049203595647374235, + "grad_norm": 33.468441009521484, + "learning_rate": 3.2563025210084036e-06, + "loss": 0.8342, + "step": 156 + }, + { + "epoch": 0.004951900331178048, + "grad_norm": 27.45941734313965, + "learning_rate": 3.2773109243697483e-06, + "loss": 0.8899, + "step": 157 + }, + { + "epoch": 0.004983441097618672, + "grad_norm": 25.41571807861328, + "learning_rate": 3.2983193277310925e-06, + "loss": 0.815, + "step": 158 + }, + { + "epoch": 0.005014981864059296, + "grad_norm": 189.85194396972656, + "learning_rate": 3.319327731092437e-06, + "loss": 0.8355, + "step": 159 + }, + { + "epoch": 0.005046522630499921, + "grad_norm": 26.509309768676758, + "learning_rate": 3.3403361344537814e-06, + "loss": 0.8552, + "step": 160 + }, + { + "epoch": 0.005078063396940545, + "grad_norm": 18.34820556640625, + "learning_rate": 3.3613445378151265e-06, + "loss": 0.829, + "step": 161 + }, + { + "epoch": 0.00510960416338117, + "grad_norm": 22.896699905395508, + "learning_rate": 3.382352941176471e-06, + "loss": 0.8591, + "step": 162 + }, + { + "epoch": 0.0051411449298217945, + "grad_norm": 27.86252784729004, + "learning_rate": 3.4033613445378154e-06, + "loss": 0.8482, + "step": 163 + }, + { + "epoch": 0.005172685696262419, + "grad_norm": 28.227724075317383, + "learning_rate": 3.42436974789916e-06, + "loss": 0.8638, + "step": 164 + }, + { + "epoch": 0.005204226462703044, + "grad_norm": 33.45288848876953, + "learning_rate": 3.4453781512605043e-06, + "loss": 0.8158, + "step": 165 + }, + { + "epoch": 0.005235767229143668, + "grad_norm": 63.427303314208984, + "learning_rate": 3.466386554621849e-06, + "loss": 0.8386, + "step": 166 + }, + { + "epoch": 0.005267307995584293, + "grad_norm": 20.12970542907715, + "learning_rate": 3.4873949579831936e-06, + "loss": 0.8045, + "step": 167 + }, + { + "epoch": 0.005298848762024917, + "grad_norm": 19.97787094116211, + "learning_rate": 3.508403361344538e-06, + "loss": 0.8137, + "step": 168 + }, + { + "epoch": 0.005330389528465542, + "grad_norm": 90.49098205566406, + "learning_rate": 3.529411764705883e-06, + "loss": 0.8352, + "step": 169 + }, + { + "epoch": 0.005361930294906166, + "grad_norm": 28.53962516784668, + "learning_rate": 3.550420168067227e-06, + "loss": 0.833, + "step": 170 + }, + { + "epoch": 0.005393471061346791, + "grad_norm": 23.355873107910156, + "learning_rate": 3.5714285714285718e-06, + "loss": 0.8872, + "step": 171 + }, + { + "epoch": 0.005425011827787415, + "grad_norm": 23.810855865478516, + "learning_rate": 3.5924369747899164e-06, + "loss": 0.8929, + "step": 172 + }, + { + "epoch": 0.00545655259422804, + "grad_norm": 31.4235897064209, + "learning_rate": 3.6134453781512607e-06, + "loss": 0.8129, + "step": 173 + }, + { + "epoch": 0.0054880933606686645, + "grad_norm": 38.28118133544922, + "learning_rate": 3.6344537815126053e-06, + "loss": 0.8576, + "step": 174 + }, + { + "epoch": 0.005519634127109289, + "grad_norm": 6.535202503204346, + "learning_rate": 3.6554621848739496e-06, + "loss": 0.4801, + "step": 175 + }, + { + "epoch": 0.005551174893549914, + "grad_norm": 21.4904842376709, + "learning_rate": 3.6764705882352946e-06, + "loss": 0.8407, + "step": 176 + }, + { + "epoch": 0.005582715659990538, + "grad_norm": 25.802291870117188, + "learning_rate": 3.6974789915966393e-06, + "loss": 0.8251, + "step": 177 + }, + { + "epoch": 0.005614256426431163, + "grad_norm": 24.022756576538086, + "learning_rate": 3.7184873949579835e-06, + "loss": 0.844, + "step": 178 + }, + { + "epoch": 0.005645797192871786, + "grad_norm": 20.742727279663086, + "learning_rate": 3.739495798319328e-06, + "loss": 0.8653, + "step": 179 + }, + { + "epoch": 0.005677337959312411, + "grad_norm": 5.086256980895996, + "learning_rate": 3.7605042016806724e-06, + "loss": 0.94, + "step": 180 + }, + { + "epoch": 0.0057088787257530355, + "grad_norm": 19.134862899780273, + "learning_rate": 3.781512605042017e-06, + "loss": 0.8048, + "step": 181 + }, + { + "epoch": 0.00574041949219366, + "grad_norm": 19.550735473632812, + "learning_rate": 3.8025210084033613e-06, + "loss": 0.8544, + "step": 182 + }, + { + "epoch": 0.005771960258634285, + "grad_norm": 30.372554779052734, + "learning_rate": 3.8235294117647055e-06, + "loss": 0.8027, + "step": 183 + }, + { + "epoch": 0.005803501025074909, + "grad_norm": 27.09385108947754, + "learning_rate": 3.844537815126051e-06, + "loss": 0.7974, + "step": 184 + }, + { + "epoch": 0.005835041791515534, + "grad_norm": 96.42764282226562, + "learning_rate": 3.865546218487396e-06, + "loss": 0.8202, + "step": 185 + }, + { + "epoch": 0.005866582557956158, + "grad_norm": 41.76380157470703, + "learning_rate": 3.8865546218487395e-06, + "loss": 0.8248, + "step": 186 + }, + { + "epoch": 0.005898123324396783, + "grad_norm": 22.68573760986328, + "learning_rate": 3.907563025210084e-06, + "loss": 0.8024, + "step": 187 + }, + { + "epoch": 0.005929664090837407, + "grad_norm": 35.804378509521484, + "learning_rate": 3.928571428571429e-06, + "loss": 0.8225, + "step": 188 + }, + { + "epoch": 0.005961204857278032, + "grad_norm": 25.79207420349121, + "learning_rate": 3.9495798319327735e-06, + "loss": 0.8275, + "step": 189 + }, + { + "epoch": 0.005992745623718656, + "grad_norm": 20.452547073364258, + "learning_rate": 3.970588235294118e-06, + "loss": 0.8196, + "step": 190 + }, + { + "epoch": 0.006024286390159281, + "grad_norm": 37.384552001953125, + "learning_rate": 3.991596638655463e-06, + "loss": 0.8091, + "step": 191 + }, + { + "epoch": 0.0060558271565999055, + "grad_norm": 20.9842586517334, + "learning_rate": 4.0126050420168075e-06, + "loss": 0.8129, + "step": 192 + }, + { + "epoch": 0.00608736792304053, + "grad_norm": 28.24593162536621, + "learning_rate": 4.033613445378151e-06, + "loss": 0.8302, + "step": 193 + }, + { + "epoch": 0.006118908689481155, + "grad_norm": 20.311885833740234, + "learning_rate": 4.054621848739496e-06, + "loss": 0.8062, + "step": 194 + }, + { + "epoch": 0.006150449455921779, + "grad_norm": 4.60398530960083, + "learning_rate": 4.075630252100841e-06, + "loss": 0.486, + "step": 195 + }, + { + "epoch": 0.006181990222362404, + "grad_norm": 19.191917419433594, + "learning_rate": 4.096638655462185e-06, + "loss": 0.8109, + "step": 196 + }, + { + "epoch": 0.006213530988803028, + "grad_norm": 28.645902633666992, + "learning_rate": 4.11764705882353e-06, + "loss": 0.8117, + "step": 197 + }, + { + "epoch": 0.006245071755243653, + "grad_norm": 3.844499349594116, + "learning_rate": 4.138655462184874e-06, + "loss": 0.9325, + "step": 198 + }, + { + "epoch": 0.0062766125216842765, + "grad_norm": 27.44171905517578, + "learning_rate": 4.159663865546219e-06, + "loss": 0.8181, + "step": 199 + }, + { + "epoch": 0.006308153288124901, + "grad_norm": 3.7000012397766113, + "learning_rate": 4.180672268907564e-06, + "loss": 0.4932, + "step": 200 + }, + { + "epoch": 0.006339694054565526, + "grad_norm": 40.338226318359375, + "learning_rate": 4.201680672268908e-06, + "loss": 0.8241, + "step": 201 + }, + { + "epoch": 0.00637123482100615, + "grad_norm": 75.82376861572266, + "learning_rate": 4.222689075630252e-06, + "loss": 0.8043, + "step": 202 + }, + { + "epoch": 0.006402775587446775, + "grad_norm": 21.27734375, + "learning_rate": 4.243697478991597e-06, + "loss": 0.8095, + "step": 203 + }, + { + "epoch": 0.006434316353887399, + "grad_norm": 16.817712783813477, + "learning_rate": 4.264705882352942e-06, + "loss": 0.8247, + "step": 204 + }, + { + "epoch": 0.006465857120328024, + "grad_norm": 90.65699768066406, + "learning_rate": 4.2857142857142855e-06, + "loss": 0.8317, + "step": 205 + }, + { + "epoch": 0.006497397886768648, + "grad_norm": 26.177085876464844, + "learning_rate": 4.306722689075631e-06, + "loss": 0.4416, + "step": 206 + }, + { + "epoch": 0.006528938653209273, + "grad_norm": 22.276323318481445, + "learning_rate": 4.327731092436976e-06, + "loss": 0.8024, + "step": 207 + }, + { + "epoch": 0.0065604794196498974, + "grad_norm": 19.594749450683594, + "learning_rate": 4.3487394957983194e-06, + "loss": 0.7822, + "step": 208 + }, + { + "epoch": 0.006592020186090522, + "grad_norm": 28.560495376586914, + "learning_rate": 4.369747899159664e-06, + "loss": 0.8216, + "step": 209 + }, + { + "epoch": 0.0066235609525311465, + "grad_norm": 33.75347137451172, + "learning_rate": 4.390756302521009e-06, + "loss": 0.8102, + "step": 210 + }, + { + "epoch": 0.006655101718971771, + "grad_norm": 39.181541442871094, + "learning_rate": 4.411764705882353e-06, + "loss": 0.8301, + "step": 211 + }, + { + "epoch": 0.006686642485412396, + "grad_norm": 22.07779312133789, + "learning_rate": 4.432773109243698e-06, + "loss": 0.8232, + "step": 212 + }, + { + "epoch": 0.00671818325185302, + "grad_norm": 25.972898483276367, + "learning_rate": 4.453781512605043e-06, + "loss": 0.8203, + "step": 213 + }, + { + "epoch": 0.006749724018293645, + "grad_norm": 40.668277740478516, + "learning_rate": 4.474789915966387e-06, + "loss": 0.8205, + "step": 214 + }, + { + "epoch": 0.006781264784734269, + "grad_norm": 25.816255569458008, + "learning_rate": 4.495798319327731e-06, + "loss": 0.8002, + "step": 215 + }, + { + "epoch": 0.006812805551174894, + "grad_norm": 87.38954162597656, + "learning_rate": 4.516806722689076e-06, + "loss": 0.8087, + "step": 216 + }, + { + "epoch": 0.006844346317615518, + "grad_norm": 27.636613845825195, + "learning_rate": 4.5378151260504205e-06, + "loss": 0.7978, + "step": 217 + }, + { + "epoch": 0.006875887084056143, + "grad_norm": 23.379953384399414, + "learning_rate": 4.558823529411765e-06, + "loss": 0.8013, + "step": 218 + }, + { + "epoch": 0.0069074278504967675, + "grad_norm": 25.465547561645508, + "learning_rate": 4.57983193277311e-06, + "loss": 0.7799, + "step": 219 + }, + { + "epoch": 0.006938968616937391, + "grad_norm": 41.498573303222656, + "learning_rate": 4.600840336134454e-06, + "loss": 0.8509, + "step": 220 + }, + { + "epoch": 0.006970509383378016, + "grad_norm": 56.100093841552734, + "learning_rate": 4.621848739495799e-06, + "loss": 0.8063, + "step": 221 + }, + { + "epoch": 0.00700205014981864, + "grad_norm": 21.222671508789062, + "learning_rate": 4.642857142857144e-06, + "loss": 0.8089, + "step": 222 + }, + { + "epoch": 0.007033590916259265, + "grad_norm": 4.75217866897583, + "learning_rate": 4.663865546218488e-06, + "loss": 0.501, + "step": 223 + }, + { + "epoch": 0.007065131682699889, + "grad_norm": 16.48567771911621, + "learning_rate": 4.684873949579832e-06, + "loss": 0.8032, + "step": 224 + }, + { + "epoch": 0.007096672449140514, + "grad_norm": 18.670562744140625, + "learning_rate": 4.705882352941177e-06, + "loss": 0.7877, + "step": 225 + }, + { + "epoch": 0.0071282132155811384, + "grad_norm": 25.6202392578125, + "learning_rate": 4.7268907563025216e-06, + "loss": 0.7864, + "step": 226 + }, + { + "epoch": 0.007159753982021763, + "grad_norm": 16.800004959106445, + "learning_rate": 4.747899159663865e-06, + "loss": 0.8172, + "step": 227 + }, + { + "epoch": 0.0071912947484623875, + "grad_norm": 3.627828359603882, + "learning_rate": 4.768907563025211e-06, + "loss": 0.4525, + "step": 228 + }, + { + "epoch": 0.007222835514903012, + "grad_norm": 19.680343627929688, + "learning_rate": 4.7899159663865555e-06, + "loss": 0.816, + "step": 229 + }, + { + "epoch": 0.007254376281343637, + "grad_norm": 20.73505401611328, + "learning_rate": 4.810924369747899e-06, + "loss": 0.8217, + "step": 230 + }, + { + "epoch": 0.007285917047784261, + "grad_norm": 19.511829376220703, + "learning_rate": 4.831932773109244e-06, + "loss": 0.8116, + "step": 231 + }, + { + "epoch": 0.007317457814224886, + "grad_norm": 6.427556991577148, + "learning_rate": 4.852941176470589e-06, + "loss": 0.4461, + "step": 232 + }, + { + "epoch": 0.00734899858066551, + "grad_norm": 22.8995304107666, + "learning_rate": 4.873949579831933e-06, + "loss": 0.7984, + "step": 233 + }, + { + "epoch": 0.007380539347106135, + "grad_norm": 16.240053176879883, + "learning_rate": 4.894957983193277e-06, + "loss": 0.8085, + "step": 234 + }, + { + "epoch": 0.007412080113546759, + "grad_norm": 18.031452178955078, + "learning_rate": 4.915966386554622e-06, + "loss": 0.773, + "step": 235 + }, + { + "epoch": 0.007443620879987384, + "grad_norm": 4.954164028167725, + "learning_rate": 4.936974789915967e-06, + "loss": 0.4797, + "step": 236 + }, + { + "epoch": 0.0074751616464280085, + "grad_norm": 28.365707397460938, + "learning_rate": 4.957983193277311e-06, + "loss": 0.8094, + "step": 237 + }, + { + "epoch": 0.007506702412868633, + "grad_norm": 29.182292938232422, + "learning_rate": 4.978991596638656e-06, + "loss": 0.8025, + "step": 238 + }, + { + "epoch": 0.007538243179309258, + "grad_norm": 28.580957412719727, + "learning_rate": 5e-06, + "loss": 0.7761, + "step": 239 + }, + { + "epoch": 0.007569783945749882, + "grad_norm": 39.18687057495117, + "learning_rate": 5.021008403361345e-06, + "loss": 0.7719, + "step": 240 + }, + { + "epoch": 0.007601324712190506, + "grad_norm": 20.657764434814453, + "learning_rate": 5.04201680672269e-06, + "loss": 0.7997, + "step": 241 + }, + { + "epoch": 0.00763286547863113, + "grad_norm": 3.9311647415161133, + "learning_rate": 5.0630252100840335e-06, + "loss": 0.9371, + "step": 242 + }, + { + "epoch": 0.007664406245071755, + "grad_norm": 20.791229248046875, + "learning_rate": 5.084033613445378e-06, + "loss": 0.7929, + "step": 243 + }, + { + "epoch": 0.0076959470115123794, + "grad_norm": 18.547060012817383, + "learning_rate": 5.105042016806723e-06, + "loss": 0.7839, + "step": 244 + }, + { + "epoch": 0.007727487777953004, + "grad_norm": 20.994077682495117, + "learning_rate": 5.1260504201680675e-06, + "loss": 0.7537, + "step": 245 + }, + { + "epoch": 0.0077590285443936285, + "grad_norm": 19.551816940307617, + "learning_rate": 5.147058823529411e-06, + "loss": 0.8206, + "step": 246 + }, + { + "epoch": 0.007790569310834253, + "grad_norm": 18.412269592285156, + "learning_rate": 5.168067226890757e-06, + "loss": 0.807, + "step": 247 + }, + { + "epoch": 0.007822110077274878, + "grad_norm": 27.20709991455078, + "learning_rate": 5.1890756302521015e-06, + "loss": 0.7816, + "step": 248 + }, + { + "epoch": 0.007853650843715502, + "grad_norm": 14.153234481811523, + "learning_rate": 5.210084033613446e-06, + "loss": 0.7727, + "step": 249 + }, + { + "epoch": 0.007885191610156127, + "grad_norm": 38.532493591308594, + "learning_rate": 5.231092436974791e-06, + "loss": 0.7911, + "step": 250 + }, + { + "epoch": 0.007916732376596751, + "grad_norm": 4.42999267578125, + "learning_rate": 5.2521008403361354e-06, + "loss": 0.9556, + "step": 251 + }, + { + "epoch": 0.007948273143037376, + "grad_norm": 21.494104385375977, + "learning_rate": 5.273109243697479e-06, + "loss": 0.7776, + "step": 252 + }, + { + "epoch": 0.007979813909478, + "grad_norm": 18.644794464111328, + "learning_rate": 5.294117647058824e-06, + "loss": 0.8086, + "step": 253 + }, + { + "epoch": 0.008011354675918625, + "grad_norm": 3.393387794494629, + "learning_rate": 5.3151260504201686e-06, + "loss": 0.9376, + "step": 254 + }, + { + "epoch": 0.00804289544235925, + "grad_norm": 19.613872528076172, + "learning_rate": 5.336134453781513e-06, + "loss": 0.7885, + "step": 255 + }, + { + "epoch": 0.008074436208799874, + "grad_norm": 3.449394464492798, + "learning_rate": 5.357142857142857e-06, + "loss": 0.8963, + "step": 256 + }, + { + "epoch": 0.008105976975240499, + "grad_norm": 31.36168098449707, + "learning_rate": 5.378151260504202e-06, + "loss": 0.7892, + "step": 257 + }, + { + "epoch": 0.008137517741681123, + "grad_norm": 24.519357681274414, + "learning_rate": 5.399159663865546e-06, + "loss": 0.801, + "step": 258 + }, + { + "epoch": 0.008169058508121748, + "grad_norm": 3.846628427505493, + "learning_rate": 5.420168067226891e-06, + "loss": 0.4529, + "step": 259 + }, + { + "epoch": 0.008200599274562372, + "grad_norm": 4.683958053588867, + "learning_rate": 5.441176470588236e-06, + "loss": 0.9005, + "step": 260 + }, + { + "epoch": 0.008232140041002997, + "grad_norm": 19.98973846435547, + "learning_rate": 5.4621848739495795e-06, + "loss": 0.7736, + "step": 261 + }, + { + "epoch": 0.008263680807443621, + "grad_norm": 17.520610809326172, + "learning_rate": 5.483193277310925e-06, + "loss": 0.795, + "step": 262 + }, + { + "epoch": 0.008295221573884246, + "grad_norm": 14.229599952697754, + "learning_rate": 5.50420168067227e-06, + "loss": 0.7691, + "step": 263 + }, + { + "epoch": 0.00832676234032487, + "grad_norm": 13.578532218933105, + "learning_rate": 5.525210084033614e-06, + "loss": 0.7842, + "step": 264 + }, + { + "epoch": 0.008358303106765495, + "grad_norm": 16.26575469970703, + "learning_rate": 5.546218487394959e-06, + "loss": 0.7829, + "step": 265 + }, + { + "epoch": 0.00838984387320612, + "grad_norm": 12.909857749938965, + "learning_rate": 5.567226890756303e-06, + "loss": 0.7703, + "step": 266 + }, + { + "epoch": 0.008421384639646744, + "grad_norm": 20.36368179321289, + "learning_rate": 5.588235294117647e-06, + "loss": 0.7797, + "step": 267 + }, + { + "epoch": 0.008452925406087369, + "grad_norm": 17.35818862915039, + "learning_rate": 5.609243697478992e-06, + "loss": 0.8088, + "step": 268 + }, + { + "epoch": 0.008484466172527993, + "grad_norm": 19.300708770751953, + "learning_rate": 5.630252100840337e-06, + "loss": 0.8046, + "step": 269 + }, + { + "epoch": 0.008516006938968618, + "grad_norm": 22.82680892944336, + "learning_rate": 5.651260504201681e-06, + "loss": 0.7868, + "step": 270 + }, + { + "epoch": 0.008547547705409242, + "grad_norm": 26.15898323059082, + "learning_rate": 5.672268907563025e-06, + "loss": 0.7724, + "step": 271 + }, + { + "epoch": 0.008579088471849867, + "grad_norm": 18.506893157958984, + "learning_rate": 5.69327731092437e-06, + "loss": 0.8051, + "step": 272 + }, + { + "epoch": 0.00861062923829049, + "grad_norm": 13.006529808044434, + "learning_rate": 5.7142857142857145e-06, + "loss": 0.7759, + "step": 273 + }, + { + "epoch": 0.008642170004731114, + "grad_norm": 18.77261734008789, + "learning_rate": 5.735294117647059e-06, + "loss": 0.7854, + "step": 274 + }, + { + "epoch": 0.008673710771171739, + "grad_norm": 20.338171005249023, + "learning_rate": 5.756302521008403e-06, + "loss": 0.758, + "step": 275 + }, + { + "epoch": 0.008705251537612363, + "grad_norm": 4.451769828796387, + "learning_rate": 5.777310924369748e-06, + "loss": 0.4163, + "step": 276 + }, + { + "epoch": 0.008736792304052988, + "grad_norm": 20.395122528076172, + "learning_rate": 5.798319327731093e-06, + "loss": 0.8263, + "step": 277 + }, + { + "epoch": 0.008768333070493612, + "grad_norm": 16.37607765197754, + "learning_rate": 5.819327731092438e-06, + "loss": 0.792, + "step": 278 + }, + { + "epoch": 0.008799873836934237, + "grad_norm": 23.0340518951416, + "learning_rate": 5.8403361344537825e-06, + "loss": 0.7588, + "step": 279 + }, + { + "epoch": 0.008831414603374861, + "grad_norm": 15.568406105041504, + "learning_rate": 5.861344537815127e-06, + "loss": 0.7695, + "step": 280 + }, + { + "epoch": 0.008862955369815486, + "grad_norm": 23.953176498413086, + "learning_rate": 5.882352941176471e-06, + "loss": 0.796, + "step": 281 + }, + { + "epoch": 0.00889449613625611, + "grad_norm": 15.70067310333252, + "learning_rate": 5.903361344537816e-06, + "loss": 0.781, + "step": 282 + }, + { + "epoch": 0.008926036902696735, + "grad_norm": 38.365760803222656, + "learning_rate": 5.92436974789916e-06, + "loss": 0.7777, + "step": 283 + }, + { + "epoch": 0.00895757766913736, + "grad_norm": 11.661864280700684, + "learning_rate": 5.945378151260505e-06, + "loss": 0.7844, + "step": 284 + }, + { + "epoch": 0.008989118435577984, + "grad_norm": 23.706085205078125, + "learning_rate": 5.9663865546218495e-06, + "loss": 0.7705, + "step": 285 + }, + { + "epoch": 0.009020659202018609, + "grad_norm": 16.721561431884766, + "learning_rate": 5.987394957983193e-06, + "loss": 0.7719, + "step": 286 + }, + { + "epoch": 0.009052199968459233, + "grad_norm": 16.0439453125, + "learning_rate": 6.008403361344538e-06, + "loss": 0.7416, + "step": 287 + }, + { + "epoch": 0.009083740734899858, + "grad_norm": 13.915446281433105, + "learning_rate": 6.029411764705883e-06, + "loss": 0.7911, + "step": 288 + }, + { + "epoch": 0.009115281501340482, + "grad_norm": 13.026446342468262, + "learning_rate": 6.050420168067227e-06, + "loss": 0.7882, + "step": 289 + }, + { + "epoch": 0.009146822267781107, + "grad_norm": 11.728982925415039, + "learning_rate": 6.071428571428571e-06, + "loss": 0.7529, + "step": 290 + }, + { + "epoch": 0.009178363034221731, + "grad_norm": 14.083760261535645, + "learning_rate": 6.092436974789916e-06, + "loss": 0.7868, + "step": 291 + }, + { + "epoch": 0.009209903800662356, + "grad_norm": 21.284709930419922, + "learning_rate": 6.113445378151261e-06, + "loss": 0.7407, + "step": 292 + }, + { + "epoch": 0.00924144456710298, + "grad_norm": 3.1844027042388916, + "learning_rate": 6.134453781512606e-06, + "loss": 0.9459, + "step": 293 + }, + { + "epoch": 0.009272985333543605, + "grad_norm": 40.13914108276367, + "learning_rate": 6.155462184873951e-06, + "loss": 0.7978, + "step": 294 + }, + { + "epoch": 0.00930452609998423, + "grad_norm": 13.233213424682617, + "learning_rate": 6.176470588235295e-06, + "loss": 0.7686, + "step": 295 + }, + { + "epoch": 0.009336066866424854, + "grad_norm": 35.70071792602539, + "learning_rate": 6.197478991596639e-06, + "loss": 0.7833, + "step": 296 + }, + { + "epoch": 0.009367607632865479, + "grad_norm": 19.559650421142578, + "learning_rate": 6.218487394957984e-06, + "loss": 0.7872, + "step": 297 + }, + { + "epoch": 0.009399148399306103, + "grad_norm": 21.009563446044922, + "learning_rate": 6.239495798319328e-06, + "loss": 0.7525, + "step": 298 + }, + { + "epoch": 0.009430689165746728, + "grad_norm": 6.027929782867432, + "learning_rate": 6.260504201680673e-06, + "loss": 0.881, + "step": 299 + }, + { + "epoch": 0.009462229932187352, + "grad_norm": 13.080622673034668, + "learning_rate": 6.281512605042017e-06, + "loss": 0.7823, + "step": 300 + }, + { + "epoch": 0.009493770698627977, + "grad_norm": 16.325252532958984, + "learning_rate": 6.3025210084033615e-06, + "loss": 0.7627, + "step": 301 + }, + { + "epoch": 0.009525311465068602, + "grad_norm": 16.671710968017578, + "learning_rate": 6.323529411764706e-06, + "loss": 0.7659, + "step": 302 + }, + { + "epoch": 0.009556852231509226, + "grad_norm": 11.597186088562012, + "learning_rate": 6.344537815126051e-06, + "loss": 0.7643, + "step": 303 + }, + { + "epoch": 0.00958839299794985, + "grad_norm": 14.073203086853027, + "learning_rate": 6.3655462184873955e-06, + "loss": 0.8049, + "step": 304 + }, + { + "epoch": 0.009619933764390475, + "grad_norm": 3.7833027839660645, + "learning_rate": 6.386554621848739e-06, + "loss": 0.8932, + "step": 305 + }, + { + "epoch": 0.0096514745308311, + "grad_norm": 36.2081413269043, + "learning_rate": 6.407563025210085e-06, + "loss": 0.7816, + "step": 306 + }, + { + "epoch": 0.009683015297271724, + "grad_norm": 16.116310119628906, + "learning_rate": 6.4285714285714295e-06, + "loss": 0.7508, + "step": 307 + }, + { + "epoch": 0.009714556063712349, + "grad_norm": 20.53990364074707, + "learning_rate": 6.449579831932774e-06, + "loss": 0.7999, + "step": 308 + }, + { + "epoch": 0.009746096830152973, + "grad_norm": 4.7028584480285645, + "learning_rate": 6.470588235294119e-06, + "loss": 0.8753, + "step": 309 + }, + { + "epoch": 0.009777637596593598, + "grad_norm": 3.549725294113159, + "learning_rate": 6.491596638655463e-06, + "loss": 0.9104, + "step": 310 + }, + { + "epoch": 0.009809178363034222, + "grad_norm": 34.350425720214844, + "learning_rate": 6.512605042016807e-06, + "loss": 0.8279, + "step": 311 + }, + { + "epoch": 0.009840719129474847, + "grad_norm": 32.634437561035156, + "learning_rate": 6.533613445378152e-06, + "loss": 0.7771, + "step": 312 + }, + { + "epoch": 0.009872259895915472, + "grad_norm": 19.84749984741211, + "learning_rate": 6.5546218487394966e-06, + "loss": 0.7921, + "step": 313 + }, + { + "epoch": 0.009903800662356096, + "grad_norm": 32.05736541748047, + "learning_rate": 6.575630252100841e-06, + "loss": 0.794, + "step": 314 + }, + { + "epoch": 0.009935341428796719, + "grad_norm": 16.79811668395996, + "learning_rate": 6.596638655462185e-06, + "loss": 0.7606, + "step": 315 + }, + { + "epoch": 0.009966882195237343, + "grad_norm": 14.114212036132812, + "learning_rate": 6.61764705882353e-06, + "loss": 0.7542, + "step": 316 + }, + { + "epoch": 0.009998422961677968, + "grad_norm": 20.307058334350586, + "learning_rate": 6.638655462184874e-06, + "loss": 0.7626, + "step": 317 + }, + { + "epoch": 0.010029963728118593, + "grad_norm": 10.990663528442383, + "learning_rate": 6.659663865546219e-06, + "loss": 0.7586, + "step": 318 + }, + { + "epoch": 0.010061504494559217, + "grad_norm": 10.486679077148438, + "learning_rate": 6.680672268907563e-06, + "loss": 0.7601, + "step": 319 + }, + { + "epoch": 0.010093045260999842, + "grad_norm": 13.805947303771973, + "learning_rate": 6.7016806722689075e-06, + "loss": 0.7705, + "step": 320 + }, + { + "epoch": 0.010124586027440466, + "grad_norm": 13.40145206451416, + "learning_rate": 6.722689075630253e-06, + "loss": 0.7457, + "step": 321 + }, + { + "epoch": 0.01015612679388109, + "grad_norm": 3.626336097717285, + "learning_rate": 6.743697478991598e-06, + "loss": 0.8966, + "step": 322 + }, + { + "epoch": 0.010187667560321715, + "grad_norm": 3.0695502758026123, + "learning_rate": 6.764705882352942e-06, + "loss": 0.4213, + "step": 323 + }, + { + "epoch": 0.01021920832676234, + "grad_norm": 17.10307502746582, + "learning_rate": 6.785714285714287e-06, + "loss": 0.7766, + "step": 324 + }, + { + "epoch": 0.010250749093202964, + "grad_norm": 20.966447830200195, + "learning_rate": 6.806722689075631e-06, + "loss": 0.7903, + "step": 325 + }, + { + "epoch": 0.010282289859643589, + "grad_norm": 5.741248607635498, + "learning_rate": 6.827731092436975e-06, + "loss": 0.9231, + "step": 326 + }, + { + "epoch": 0.010313830626084214, + "grad_norm": 19.529273986816406, + "learning_rate": 6.84873949579832e-06, + "loss": 0.7702, + "step": 327 + }, + { + "epoch": 0.010345371392524838, + "grad_norm": 5.3112053871154785, + "learning_rate": 6.869747899159665e-06, + "loss": 0.8926, + "step": 328 + }, + { + "epoch": 0.010376912158965463, + "grad_norm": 14.768327713012695, + "learning_rate": 6.8907563025210085e-06, + "loss": 0.7407, + "step": 329 + }, + { + "epoch": 0.010408452925406087, + "grad_norm": 18.78706932067871, + "learning_rate": 6.911764705882353e-06, + "loss": 0.7595, + "step": 330 + }, + { + "epoch": 0.010439993691846712, + "grad_norm": 11.735562324523926, + "learning_rate": 6.932773109243698e-06, + "loss": 0.7519, + "step": 331 + }, + { + "epoch": 0.010471534458287336, + "grad_norm": 11.264945030212402, + "learning_rate": 6.9537815126050425e-06, + "loss": 0.753, + "step": 332 + }, + { + "epoch": 0.01050307522472796, + "grad_norm": 21.1032772064209, + "learning_rate": 6.974789915966387e-06, + "loss": 0.8178, + "step": 333 + }, + { + "epoch": 0.010534615991168585, + "grad_norm": 13.072237014770508, + "learning_rate": 6.995798319327731e-06, + "loss": 0.7772, + "step": 334 + }, + { + "epoch": 0.01056615675760921, + "grad_norm": 16.557146072387695, + "learning_rate": 7.016806722689076e-06, + "loss": 0.8041, + "step": 335 + }, + { + "epoch": 0.010597697524049834, + "grad_norm": 16.054981231689453, + "learning_rate": 7.037815126050421e-06, + "loss": 0.8044, + "step": 336 + }, + { + "epoch": 0.010629238290490459, + "grad_norm": 16.82111358642578, + "learning_rate": 7.058823529411766e-06, + "loss": 0.7424, + "step": 337 + }, + { + "epoch": 0.010660779056931084, + "grad_norm": 9.708539962768555, + "learning_rate": 7.0798319327731104e-06, + "loss": 0.4574, + "step": 338 + }, + { + "epoch": 0.010692319823371708, + "grad_norm": 14.456352233886719, + "learning_rate": 7.100840336134454e-06, + "loss": 0.781, + "step": 339 + }, + { + "epoch": 0.010723860589812333, + "grad_norm": 17.215190887451172, + "learning_rate": 7.121848739495799e-06, + "loss": 0.7787, + "step": 340 + }, + { + "epoch": 0.010755401356252957, + "grad_norm": 17.784320831298828, + "learning_rate": 7.1428571428571436e-06, + "loss": 0.893, + "step": 341 + }, + { + "epoch": 0.010786942122693582, + "grad_norm": 14.018430709838867, + "learning_rate": 7.163865546218488e-06, + "loss": 0.7495, + "step": 342 + }, + { + "epoch": 0.010818482889134206, + "grad_norm": 15.077600479125977, + "learning_rate": 7.184873949579833e-06, + "loss": 0.7841, + "step": 343 + }, + { + "epoch": 0.01085002365557483, + "grad_norm": 10.303396224975586, + "learning_rate": 7.205882352941177e-06, + "loss": 0.7454, + "step": 344 + }, + { + "epoch": 0.010881564422015455, + "grad_norm": 17.379253387451172, + "learning_rate": 7.226890756302521e-06, + "loss": 0.7903, + "step": 345 + }, + { + "epoch": 0.01091310518845608, + "grad_norm": 14.070576667785645, + "learning_rate": 7.247899159663866e-06, + "loss": 0.776, + "step": 346 + }, + { + "epoch": 0.010944645954896704, + "grad_norm": 5.256331920623779, + "learning_rate": 7.268907563025211e-06, + "loss": 0.463, + "step": 347 + }, + { + "epoch": 0.010976186721337329, + "grad_norm": 17.277339935302734, + "learning_rate": 7.2899159663865545e-06, + "loss": 0.7636, + "step": 348 + }, + { + "epoch": 0.011007727487777954, + "grad_norm": 16.34393310546875, + "learning_rate": 7.310924369747899e-06, + "loss": 0.7744, + "step": 349 + }, + { + "epoch": 0.011039268254218578, + "grad_norm": 17.669750213623047, + "learning_rate": 7.331932773109244e-06, + "loss": 0.7409, + "step": 350 + }, + { + "epoch": 0.011070809020659203, + "grad_norm": 16.262861251831055, + "learning_rate": 7.352941176470589e-06, + "loss": 0.7282, + "step": 351 + }, + { + "epoch": 0.011102349787099827, + "grad_norm": 18.644977569580078, + "learning_rate": 7.373949579831934e-06, + "loss": 0.7673, + "step": 352 + }, + { + "epoch": 0.011133890553540452, + "grad_norm": 11.604530334472656, + "learning_rate": 7.394957983193279e-06, + "loss": 0.9166, + "step": 353 + }, + { + "epoch": 0.011165431319981076, + "grad_norm": 17.13448715209961, + "learning_rate": 7.415966386554622e-06, + "loss": 0.7917, + "step": 354 + }, + { + "epoch": 0.011196972086421701, + "grad_norm": 25.20997428894043, + "learning_rate": 7.436974789915967e-06, + "loss": 0.7629, + "step": 355 + }, + { + "epoch": 0.011228512852862325, + "grad_norm": 3.126866102218628, + "learning_rate": 7.457983193277312e-06, + "loss": 0.4129, + "step": 356 + }, + { + "epoch": 0.011260053619302948, + "grad_norm": 14.210492134094238, + "learning_rate": 7.478991596638656e-06, + "loss": 0.771, + "step": 357 + }, + { + "epoch": 0.011291594385743573, + "grad_norm": 25.873125076293945, + "learning_rate": 7.500000000000001e-06, + "loss": 0.7571, + "step": 358 + }, + { + "epoch": 0.011323135152184197, + "grad_norm": 11.93407154083252, + "learning_rate": 7.521008403361345e-06, + "loss": 0.7339, + "step": 359 + }, + { + "epoch": 0.011354675918624822, + "grad_norm": 14.819896697998047, + "learning_rate": 7.5420168067226895e-06, + "loss": 0.7696, + "step": 360 + }, + { + "epoch": 0.011386216685065446, + "grad_norm": 37.304481506347656, + "learning_rate": 7.563025210084034e-06, + "loss": 0.7531, + "step": 361 + }, + { + "epoch": 0.011417757451506071, + "grad_norm": 3.131352424621582, + "learning_rate": 7.584033613445379e-06, + "loss": 0.9523, + "step": 362 + }, + { + "epoch": 0.011449298217946696, + "grad_norm": 14.3023099899292, + "learning_rate": 7.605042016806723e-06, + "loss": 0.7543, + "step": 363 + }, + { + "epoch": 0.01148083898438732, + "grad_norm": 11.309164047241211, + "learning_rate": 7.626050420168067e-06, + "loss": 0.7387, + "step": 364 + }, + { + "epoch": 0.011512379750827945, + "grad_norm": 16.822776794433594, + "learning_rate": 7.647058823529411e-06, + "loss": 0.74, + "step": 365 + }, + { + "epoch": 0.01154392051726857, + "grad_norm": 14.50583553314209, + "learning_rate": 7.668067226890757e-06, + "loss": 0.4265, + "step": 366 + }, + { + "epoch": 0.011575461283709194, + "grad_norm": 16.015220642089844, + "learning_rate": 7.689075630252102e-06, + "loss": 0.76, + "step": 367 + }, + { + "epoch": 0.011607002050149818, + "grad_norm": 5.374929904937744, + "learning_rate": 7.710084033613446e-06, + "loss": 0.4299, + "step": 368 + }, + { + "epoch": 0.011638542816590443, + "grad_norm": 6.511736869812012, + "learning_rate": 7.731092436974791e-06, + "loss": 0.9028, + "step": 369 + }, + { + "epoch": 0.011670083583031067, + "grad_norm": 10.706531524658203, + "learning_rate": 7.752100840336135e-06, + "loss": 0.7869, + "step": 370 + }, + { + "epoch": 0.011701624349471692, + "grad_norm": 8.609142303466797, + "learning_rate": 7.773109243697479e-06, + "loss": 0.7843, + "step": 371 + }, + { + "epoch": 0.011733165115912316, + "grad_norm": 16.572053909301758, + "learning_rate": 7.794117647058825e-06, + "loss": 0.7876, + "step": 372 + }, + { + "epoch": 0.011764705882352941, + "grad_norm": 19.45516014099121, + "learning_rate": 7.815126050420168e-06, + "loss": 0.7245, + "step": 373 + }, + { + "epoch": 0.011796246648793566, + "grad_norm": 11.090559959411621, + "learning_rate": 7.836134453781514e-06, + "loss": 0.7494, + "step": 374 + }, + { + "epoch": 0.01182778741523419, + "grad_norm": 12.149003982543945, + "learning_rate": 7.857142857142858e-06, + "loss": 0.7414, + "step": 375 + }, + { + "epoch": 0.011859328181674815, + "grad_norm": 15.509292602539062, + "learning_rate": 7.878151260504201e-06, + "loss": 0.7781, + "step": 376 + }, + { + "epoch": 0.01189086894811544, + "grad_norm": 14.691365242004395, + "learning_rate": 7.899159663865547e-06, + "loss": 0.7723, + "step": 377 + }, + { + "epoch": 0.011922409714556064, + "grad_norm": 4.341413497924805, + "learning_rate": 7.92016806722689e-06, + "loss": 0.4181, + "step": 378 + }, + { + "epoch": 0.011953950480996688, + "grad_norm": 31.240304946899414, + "learning_rate": 7.941176470588236e-06, + "loss": 0.7339, + "step": 379 + }, + { + "epoch": 0.011985491247437313, + "grad_norm": 4.238749027252197, + "learning_rate": 7.96218487394958e-06, + "loss": 0.4172, + "step": 380 + }, + { + "epoch": 0.012017032013877937, + "grad_norm": 45.15483474731445, + "learning_rate": 7.983193277310926e-06, + "loss": 0.4533, + "step": 381 + }, + { + "epoch": 0.012048572780318562, + "grad_norm": 16.31978416442871, + "learning_rate": 8.00420168067227e-06, + "loss": 0.7406, + "step": 382 + }, + { + "epoch": 0.012080113546759186, + "grad_norm": 12.751980781555176, + "learning_rate": 8.025210084033615e-06, + "loss": 0.7517, + "step": 383 + }, + { + "epoch": 0.012111654313199811, + "grad_norm": 16.64838409423828, + "learning_rate": 8.046218487394959e-06, + "loss": 0.7488, + "step": 384 + }, + { + "epoch": 0.012143195079640436, + "grad_norm": 18.36774444580078, + "learning_rate": 8.067226890756303e-06, + "loss": 0.7375, + "step": 385 + }, + { + "epoch": 0.01217473584608106, + "grad_norm": 38.741851806640625, + "learning_rate": 8.088235294117648e-06, + "loss": 0.7614, + "step": 386 + }, + { + "epoch": 0.012206276612521685, + "grad_norm": 10.655895233154297, + "learning_rate": 8.109243697478992e-06, + "loss": 0.7367, + "step": 387 + }, + { + "epoch": 0.01223781737896231, + "grad_norm": 4.577566623687744, + "learning_rate": 8.130252100840337e-06, + "loss": 0.4094, + "step": 388 + }, + { + "epoch": 0.012269358145402934, + "grad_norm": 18.84414291381836, + "learning_rate": 8.151260504201681e-06, + "loss": 0.7621, + "step": 389 + }, + { + "epoch": 0.012300898911843558, + "grad_norm": 28.127321243286133, + "learning_rate": 8.172268907563025e-06, + "loss": 0.7666, + "step": 390 + }, + { + "epoch": 0.012332439678284183, + "grad_norm": 17.312074661254883, + "learning_rate": 8.19327731092437e-06, + "loss": 0.6949, + "step": 391 + }, + { + "epoch": 0.012363980444724807, + "grad_norm": 22.14105987548828, + "learning_rate": 8.214285714285714e-06, + "loss": 0.7901, + "step": 392 + }, + { + "epoch": 0.012395521211165432, + "grad_norm": 14.688102722167969, + "learning_rate": 8.23529411764706e-06, + "loss": 0.8017, + "step": 393 + }, + { + "epoch": 0.012427061977606057, + "grad_norm": 13.974987983703613, + "learning_rate": 8.256302521008404e-06, + "loss": 0.7259, + "step": 394 + }, + { + "epoch": 0.012458602744046681, + "grad_norm": 55.140586853027344, + "learning_rate": 8.277310924369747e-06, + "loss": 0.7469, + "step": 395 + }, + { + "epoch": 0.012490143510487306, + "grad_norm": 3.577967405319214, + "learning_rate": 8.298319327731093e-06, + "loss": 0.3792, + "step": 396 + }, + { + "epoch": 0.01252168427692793, + "grad_norm": 23.3615779876709, + "learning_rate": 8.319327731092438e-06, + "loss": 0.7688, + "step": 397 + }, + { + "epoch": 0.012553225043368553, + "grad_norm": 19.4602108001709, + "learning_rate": 8.340336134453782e-06, + "loss": 0.7787, + "step": 398 + }, + { + "epoch": 0.012584765809809178, + "grad_norm": 16.58216094970703, + "learning_rate": 8.361344537815128e-06, + "loss": 0.7553, + "step": 399 + }, + { + "epoch": 0.012616306576249802, + "grad_norm": 15.23902702331543, + "learning_rate": 8.382352941176472e-06, + "loss": 0.795, + "step": 400 + }, + { + "epoch": 0.012647847342690427, + "grad_norm": 24.715967178344727, + "learning_rate": 8.403361344537815e-06, + "loss": 0.7366, + "step": 401 + }, + { + "epoch": 0.012679388109131051, + "grad_norm": 15.568317413330078, + "learning_rate": 8.424369747899161e-06, + "loss": 0.7552, + "step": 402 + }, + { + "epoch": 0.012710928875571676, + "grad_norm": 12.279996871948242, + "learning_rate": 8.445378151260505e-06, + "loss": 0.7675, + "step": 403 + }, + { + "epoch": 0.0127424696420123, + "grad_norm": 13.177325248718262, + "learning_rate": 8.46638655462185e-06, + "loss": 0.7702, + "step": 404 + }, + { + "epoch": 0.012774010408452925, + "grad_norm": 23.844676971435547, + "learning_rate": 8.487394957983194e-06, + "loss": 0.7903, + "step": 405 + }, + { + "epoch": 0.01280555117489355, + "grad_norm": 19.44914436340332, + "learning_rate": 8.508403361344538e-06, + "loss": 0.7716, + "step": 406 + }, + { + "epoch": 0.012837091941334174, + "grad_norm": 16.88255500793457, + "learning_rate": 8.529411764705883e-06, + "loss": 0.7614, + "step": 407 + }, + { + "epoch": 0.012868632707774798, + "grad_norm": 14.611225128173828, + "learning_rate": 8.550420168067227e-06, + "loss": 0.7485, + "step": 408 + }, + { + "epoch": 0.012900173474215423, + "grad_norm": 8.723614692687988, + "learning_rate": 8.571428571428571e-06, + "loss": 0.7418, + "step": 409 + }, + { + "epoch": 0.012931714240656048, + "grad_norm": 11.743885040283203, + "learning_rate": 8.592436974789916e-06, + "loss": 0.7418, + "step": 410 + }, + { + "epoch": 0.012963255007096672, + "grad_norm": 24.54329490661621, + "learning_rate": 8.613445378151262e-06, + "loss": 0.7329, + "step": 411 + }, + { + "epoch": 0.012994795773537297, + "grad_norm": 13.88017463684082, + "learning_rate": 8.634453781512606e-06, + "loss": 0.7407, + "step": 412 + }, + { + "epoch": 0.013026336539977921, + "grad_norm": 15.26216983795166, + "learning_rate": 8.655462184873951e-06, + "loss": 0.7195, + "step": 413 + }, + { + "epoch": 0.013057877306418546, + "grad_norm": 234.9851837158203, + "learning_rate": 8.676470588235295e-06, + "loss": 0.7203, + "step": 414 + }, + { + "epoch": 0.01308941807285917, + "grad_norm": 15.380145072937012, + "learning_rate": 8.697478991596639e-06, + "loss": 0.7272, + "step": 415 + }, + { + "epoch": 0.013120958839299795, + "grad_norm": 14.486547470092773, + "learning_rate": 8.718487394957984e-06, + "loss": 0.7283, + "step": 416 + }, + { + "epoch": 0.01315249960574042, + "grad_norm": 21.10188865661621, + "learning_rate": 8.739495798319328e-06, + "loss": 0.7411, + "step": 417 + }, + { + "epoch": 0.013184040372181044, + "grad_norm": 16.821985244750977, + "learning_rate": 8.760504201680674e-06, + "loss": 0.7811, + "step": 418 + }, + { + "epoch": 0.013215581138621669, + "grad_norm": 11.21830940246582, + "learning_rate": 8.781512605042017e-06, + "loss": 0.7444, + "step": 419 + }, + { + "epoch": 0.013247121905062293, + "grad_norm": 5.6471781730651855, + "learning_rate": 8.802521008403361e-06, + "loss": 0.8788, + "step": 420 + }, + { + "epoch": 0.013278662671502918, + "grad_norm": 44.15031433105469, + "learning_rate": 8.823529411764707e-06, + "loss": 0.7882, + "step": 421 + }, + { + "epoch": 0.013310203437943542, + "grad_norm": 20.204792022705078, + "learning_rate": 8.84453781512605e-06, + "loss": 0.7137, + "step": 422 + }, + { + "epoch": 0.013341744204384167, + "grad_norm": 25.11606788635254, + "learning_rate": 8.865546218487396e-06, + "loss": 0.7607, + "step": 423 + }, + { + "epoch": 0.013373284970824791, + "grad_norm": 21.101303100585938, + "learning_rate": 8.88655462184874e-06, + "loss": 0.7242, + "step": 424 + }, + { + "epoch": 0.013404825737265416, + "grad_norm": 18.246084213256836, + "learning_rate": 8.907563025210085e-06, + "loss": 0.7683, + "step": 425 + }, + { + "epoch": 0.01343636650370604, + "grad_norm": 11.243572235107422, + "learning_rate": 8.92857142857143e-06, + "loss": 0.7153, + "step": 426 + }, + { + "epoch": 0.013467907270146665, + "grad_norm": 16.360532760620117, + "learning_rate": 8.949579831932775e-06, + "loss": 0.7442, + "step": 427 + }, + { + "epoch": 0.01349944803658729, + "grad_norm": 14.85621166229248, + "learning_rate": 8.970588235294119e-06, + "loss": 0.7465, + "step": 428 + }, + { + "epoch": 0.013530988803027914, + "grad_norm": 9.80904483795166, + "learning_rate": 8.991596638655462e-06, + "loss": 0.8891, + "step": 429 + }, + { + "epoch": 0.013562529569468539, + "grad_norm": 14.039033889770508, + "learning_rate": 9.012605042016808e-06, + "loss": 0.7611, + "step": 430 + }, + { + "epoch": 0.013594070335909163, + "grad_norm": 17.67181968688965, + "learning_rate": 9.033613445378152e-06, + "loss": 0.7853, + "step": 431 + }, + { + "epoch": 0.013625611102349788, + "grad_norm": 8.196481704711914, + "learning_rate": 9.054621848739497e-06, + "loss": 0.4015, + "step": 432 + }, + { + "epoch": 0.013657151868790412, + "grad_norm": 19.300853729248047, + "learning_rate": 9.075630252100841e-06, + "loss": 0.7681, + "step": 433 + }, + { + "epoch": 0.013688692635231037, + "grad_norm": 13.303407669067383, + "learning_rate": 9.096638655462185e-06, + "loss": 0.7463, + "step": 434 + }, + { + "epoch": 0.013720233401671661, + "grad_norm": 11.665432929992676, + "learning_rate": 9.11764705882353e-06, + "loss": 0.7621, + "step": 435 + }, + { + "epoch": 0.013751774168112286, + "grad_norm": 13.380281448364258, + "learning_rate": 9.138655462184874e-06, + "loss": 0.7581, + "step": 436 + }, + { + "epoch": 0.01378331493455291, + "grad_norm": 14.406654357910156, + "learning_rate": 9.15966386554622e-06, + "loss": 0.7593, + "step": 437 + }, + { + "epoch": 0.013814855700993535, + "grad_norm": 26.381345748901367, + "learning_rate": 9.180672268907563e-06, + "loss": 0.7352, + "step": 438 + }, + { + "epoch": 0.01384639646743416, + "grad_norm": 17.262523651123047, + "learning_rate": 9.201680672268907e-06, + "loss": 0.7798, + "step": 439 + }, + { + "epoch": 0.013877937233874782, + "grad_norm": 15.342782974243164, + "learning_rate": 9.222689075630253e-06, + "loss": 0.7695, + "step": 440 + }, + { + "epoch": 0.013909478000315407, + "grad_norm": 4.092544078826904, + "learning_rate": 9.243697478991598e-06, + "loss": 0.8885, + "step": 441 + }, + { + "epoch": 0.013941018766756031, + "grad_norm": 14.264739990234375, + "learning_rate": 9.264705882352942e-06, + "loss": 0.7493, + "step": 442 + }, + { + "epoch": 0.013972559533196656, + "grad_norm": 5.718794822692871, + "learning_rate": 9.285714285714288e-06, + "loss": 0.4117, + "step": 443 + }, + { + "epoch": 0.01400410029963728, + "grad_norm": 18.253644943237305, + "learning_rate": 9.306722689075631e-06, + "loss": 0.8012, + "step": 444 + }, + { + "epoch": 0.014035641066077905, + "grad_norm": 3.591423273086548, + "learning_rate": 9.327731092436975e-06, + "loss": 0.9327, + "step": 445 + }, + { + "epoch": 0.01406718183251853, + "grad_norm": 28.389455795288086, + "learning_rate": 9.34873949579832e-06, + "loss": 0.7615, + "step": 446 + }, + { + "epoch": 0.014098722598959154, + "grad_norm": 19.723108291625977, + "learning_rate": 9.369747899159664e-06, + "loss": 0.7562, + "step": 447 + }, + { + "epoch": 0.014130263365399779, + "grad_norm": 9.202277183532715, + "learning_rate": 9.390756302521008e-06, + "loss": 0.741, + "step": 448 + }, + { + "epoch": 0.014161804131840403, + "grad_norm": 26.42628288269043, + "learning_rate": 9.411764705882354e-06, + "loss": 0.7603, + "step": 449 + }, + { + "epoch": 0.014193344898281028, + "grad_norm": 23.68840980529785, + "learning_rate": 9.432773109243698e-06, + "loss": 0.742, + "step": 450 + }, + { + "epoch": 0.014224885664721652, + "grad_norm": 14.130125045776367, + "learning_rate": 9.453781512605043e-06, + "loss": 0.7317, + "step": 451 + }, + { + "epoch": 0.014256426431162277, + "grad_norm": 12.282410621643066, + "learning_rate": 9.474789915966387e-06, + "loss": 0.7463, + "step": 452 + }, + { + "epoch": 0.014287967197602901, + "grad_norm": 13.656505584716797, + "learning_rate": 9.49579831932773e-06, + "loss": 0.7663, + "step": 453 + }, + { + "epoch": 0.014319507964043526, + "grad_norm": 5.042065620422363, + "learning_rate": 9.516806722689076e-06, + "loss": 0.9143, + "step": 454 + }, + { + "epoch": 0.01435104873048415, + "grad_norm": 11.744863510131836, + "learning_rate": 9.537815126050422e-06, + "loss": 0.7405, + "step": 455 + }, + { + "epoch": 0.014382589496924775, + "grad_norm": 11.925531387329102, + "learning_rate": 9.558823529411766e-06, + "loss": 0.7763, + "step": 456 + }, + { + "epoch": 0.0144141302633654, + "grad_norm": 12.413290977478027, + "learning_rate": 9.579831932773111e-06, + "loss": 0.7331, + "step": 457 + }, + { + "epoch": 0.014445671029806024, + "grad_norm": 10.383286476135254, + "learning_rate": 9.600840336134455e-06, + "loss": 0.7435, + "step": 458 + }, + { + "epoch": 0.014477211796246649, + "grad_norm": 10.549772262573242, + "learning_rate": 9.621848739495799e-06, + "loss": 0.7587, + "step": 459 + }, + { + "epoch": 0.014508752562687273, + "grad_norm": 9.495292663574219, + "learning_rate": 9.642857142857144e-06, + "loss": 0.7154, + "step": 460 + }, + { + "epoch": 0.014540293329127898, + "grad_norm": 7.998931407928467, + "learning_rate": 9.663865546218488e-06, + "loss": 0.7242, + "step": 461 + }, + { + "epoch": 0.014571834095568522, + "grad_norm": 27.664466857910156, + "learning_rate": 9.684873949579834e-06, + "loss": 0.7406, + "step": 462 + }, + { + "epoch": 0.014603374862009147, + "grad_norm": 12.695233345031738, + "learning_rate": 9.705882352941177e-06, + "loss": 0.7177, + "step": 463 + }, + { + "epoch": 0.014634915628449771, + "grad_norm": 19.108118057250977, + "learning_rate": 9.726890756302521e-06, + "loss": 0.7039, + "step": 464 + }, + { + "epoch": 0.014666456394890396, + "grad_norm": 4.140760898590088, + "learning_rate": 9.747899159663867e-06, + "loss": 0.4528, + "step": 465 + }, + { + "epoch": 0.01469799716133102, + "grad_norm": 16.215728759765625, + "learning_rate": 9.76890756302521e-06, + "loss": 0.7364, + "step": 466 + }, + { + "epoch": 0.014729537927771645, + "grad_norm": 4.387180328369141, + "learning_rate": 9.789915966386554e-06, + "loss": 0.9175, + "step": 467 + }, + { + "epoch": 0.01476107869421227, + "grad_norm": 11.517229080200195, + "learning_rate": 9.8109243697479e-06, + "loss": 0.7125, + "step": 468 + }, + { + "epoch": 0.014792619460652894, + "grad_norm": 11.175978660583496, + "learning_rate": 9.831932773109244e-06, + "loss": 0.7636, + "step": 469 + }, + { + "epoch": 0.014824160227093519, + "grad_norm": 3.602457284927368, + "learning_rate": 9.852941176470589e-06, + "loss": 0.8919, + "step": 470 + }, + { + "epoch": 0.014855700993534143, + "grad_norm": 7.860671520233154, + "learning_rate": 9.873949579831935e-06, + "loss": 0.733, + "step": 471 + }, + { + "epoch": 0.014887241759974768, + "grad_norm": 10.942117691040039, + "learning_rate": 9.894957983193278e-06, + "loss": 0.7487, + "step": 472 + }, + { + "epoch": 0.014918782526415392, + "grad_norm": 12.077033042907715, + "learning_rate": 9.915966386554622e-06, + "loss": 0.765, + "step": 473 + }, + { + "epoch": 0.014950323292856017, + "grad_norm": 11.88063907623291, + "learning_rate": 9.936974789915968e-06, + "loss": 0.735, + "step": 474 + }, + { + "epoch": 0.014981864059296642, + "grad_norm": 26.330829620361328, + "learning_rate": 9.957983193277312e-06, + "loss": 0.7168, + "step": 475 + }, + { + "epoch": 0.015013404825737266, + "grad_norm": 9.837839126586914, + "learning_rate": 9.978991596638657e-06, + "loss": 0.7172, + "step": 476 + }, + { + "epoch": 0.01504494559217789, + "grad_norm": 14.9712495803833, + "learning_rate": 1e-05, + "loss": 0.7718, + "step": 477 + }, + { + "epoch": 0.015076486358618515, + "grad_norm": 11.663215637207031, + "learning_rate": 1.0021008403361345e-05, + "loss": 0.7449, + "step": 478 + }, + { + "epoch": 0.01510802712505914, + "grad_norm": 40.76372146606445, + "learning_rate": 1.004201680672269e-05, + "loss": 0.7371, + "step": 479 + }, + { + "epoch": 0.015139567891499764, + "grad_norm": 11.226447105407715, + "learning_rate": 1.0063025210084034e-05, + "loss": 0.7522, + "step": 480 + }, + { + "epoch": 0.015171108657940389, + "grad_norm": 8.776567459106445, + "learning_rate": 1.008403361344538e-05, + "loss": 0.7318, + "step": 481 + }, + { + "epoch": 0.015202649424381012, + "grad_norm": 10.93724250793457, + "learning_rate": 1.0105042016806723e-05, + "loss": 0.7724, + "step": 482 + }, + { + "epoch": 0.015234190190821636, + "grad_norm": 11.667899131774902, + "learning_rate": 1.0126050420168067e-05, + "loss": 0.7266, + "step": 483 + }, + { + "epoch": 0.01526573095726226, + "grad_norm": 66.02132415771484, + "learning_rate": 1.0147058823529413e-05, + "loss": 0.7412, + "step": 484 + }, + { + "epoch": 0.015297271723702885, + "grad_norm": 13.765091896057129, + "learning_rate": 1.0168067226890756e-05, + "loss": 0.7273, + "step": 485 + }, + { + "epoch": 0.01532881249014351, + "grad_norm": 9.002508163452148, + "learning_rate": 1.01890756302521e-05, + "loss": 0.7386, + "step": 486 + }, + { + "epoch": 0.015360353256584134, + "grad_norm": 14.92734432220459, + "learning_rate": 1.0210084033613446e-05, + "loss": 0.7472, + "step": 487 + }, + { + "epoch": 0.015391894023024759, + "grad_norm": 12.352255821228027, + "learning_rate": 1.023109243697479e-05, + "loss": 0.7397, + "step": 488 + }, + { + "epoch": 0.015423434789465383, + "grad_norm": 3.199277639389038, + "learning_rate": 1.0252100840336135e-05, + "loss": 0.9109, + "step": 489 + }, + { + "epoch": 0.015454975555906008, + "grad_norm": 10.897979736328125, + "learning_rate": 1.0273109243697479e-05, + "loss": 0.7404, + "step": 490 + }, + { + "epoch": 0.015486516322346633, + "grad_norm": 10.753725051879883, + "learning_rate": 1.0294117647058823e-05, + "loss": 0.7145, + "step": 491 + }, + { + "epoch": 0.015518057088787257, + "grad_norm": 15.342236518859863, + "learning_rate": 1.031512605042017e-05, + "loss": 0.741, + "step": 492 + }, + { + "epoch": 0.015549597855227882, + "grad_norm": 14.537572860717773, + "learning_rate": 1.0336134453781514e-05, + "loss": 0.7345, + "step": 493 + }, + { + "epoch": 0.015581138621668506, + "grad_norm": 3.581814765930176, + "learning_rate": 1.0357142857142859e-05, + "loss": 0.3704, + "step": 494 + }, + { + "epoch": 0.01561267938810913, + "grad_norm": 11.64765453338623, + "learning_rate": 1.0378151260504203e-05, + "loss": 0.7018, + "step": 495 + }, + { + "epoch": 0.015644220154549755, + "grad_norm": 13.414934158325195, + "learning_rate": 1.0399159663865548e-05, + "loss": 0.7694, + "step": 496 + }, + { + "epoch": 0.01567576092099038, + "grad_norm": 10.589452743530273, + "learning_rate": 1.0420168067226892e-05, + "loss": 0.7324, + "step": 497 + }, + { + "epoch": 0.015707301687431004, + "grad_norm": 14.338286399841309, + "learning_rate": 1.0441176470588236e-05, + "loss": 0.7279, + "step": 498 + }, + { + "epoch": 0.01573884245387163, + "grad_norm": 14.232246398925781, + "learning_rate": 1.0462184873949582e-05, + "loss": 0.7521, + "step": 499 + }, + { + "epoch": 0.015770383220312253, + "grad_norm": 12.104767799377441, + "learning_rate": 1.0483193277310925e-05, + "loss": 0.7188, + "step": 500 + }, + { + "epoch": 0.01580192398675288, + "grad_norm": 13.422708511352539, + "learning_rate": 1.0504201680672271e-05, + "loss": 0.7462, + "step": 501 + }, + { + "epoch": 0.015833464753193503, + "grad_norm": 10.307520866394043, + "learning_rate": 1.0525210084033615e-05, + "loss": 0.7548, + "step": 502 + }, + { + "epoch": 0.015865005519634125, + "grad_norm": 15.246774673461914, + "learning_rate": 1.0546218487394959e-05, + "loss": 0.7285, + "step": 503 + }, + { + "epoch": 0.01589654628607475, + "grad_norm": 11.086276054382324, + "learning_rate": 1.0567226890756304e-05, + "loss": 0.7255, + "step": 504 + }, + { + "epoch": 0.015928087052515374, + "grad_norm": 14.990659713745117, + "learning_rate": 1.0588235294117648e-05, + "loss": 0.7529, + "step": 505 + }, + { + "epoch": 0.015959627818956, + "grad_norm": 20.519676208496094, + "learning_rate": 1.0609243697478993e-05, + "loss": 0.7435, + "step": 506 + }, + { + "epoch": 0.015991168585396624, + "grad_norm": 10.32453727722168, + "learning_rate": 1.0630252100840337e-05, + "loss": 0.7377, + "step": 507 + }, + { + "epoch": 0.01602270935183725, + "grad_norm": 16.016462326049805, + "learning_rate": 1.0651260504201681e-05, + "loss": 0.7046, + "step": 508 + }, + { + "epoch": 0.016054250118277873, + "grad_norm": 12.307084083557129, + "learning_rate": 1.0672268907563026e-05, + "loss": 0.7493, + "step": 509 + }, + { + "epoch": 0.0160857908847185, + "grad_norm": 12.200705528259277, + "learning_rate": 1.069327731092437e-05, + "loss": 0.6981, + "step": 510 + }, + { + "epoch": 0.016117331651159122, + "grad_norm": 10.525723457336426, + "learning_rate": 1.0714285714285714e-05, + "loss": 0.7099, + "step": 511 + }, + { + "epoch": 0.016148872417599748, + "grad_norm": 13.770166397094727, + "learning_rate": 1.073529411764706e-05, + "loss": 0.755, + "step": 512 + }, + { + "epoch": 0.01618041318404037, + "grad_norm": 12.788710594177246, + "learning_rate": 1.0756302521008403e-05, + "loss": 0.7254, + "step": 513 + }, + { + "epoch": 0.016211953950480997, + "grad_norm": 13.525352478027344, + "learning_rate": 1.0777310924369749e-05, + "loss": 0.7297, + "step": 514 + }, + { + "epoch": 0.01624349471692162, + "grad_norm": 10.472684860229492, + "learning_rate": 1.0798319327731093e-05, + "loss": 0.7118, + "step": 515 + }, + { + "epoch": 0.016275035483362246, + "grad_norm": 27.939023971557617, + "learning_rate": 1.0819327731092437e-05, + "loss": 0.7457, + "step": 516 + }, + { + "epoch": 0.01630657624980287, + "grad_norm": 14.17068862915039, + "learning_rate": 1.0840336134453782e-05, + "loss": 0.7136, + "step": 517 + }, + { + "epoch": 0.016338117016243495, + "grad_norm": 12.397862434387207, + "learning_rate": 1.0861344537815126e-05, + "loss": 0.7383, + "step": 518 + }, + { + "epoch": 0.016369657782684118, + "grad_norm": 3.80830717086792, + "learning_rate": 1.0882352941176471e-05, + "loss": 0.378, + "step": 519 + }, + { + "epoch": 0.016401198549124744, + "grad_norm": 10.531811714172363, + "learning_rate": 1.0903361344537815e-05, + "loss": 0.7339, + "step": 520 + }, + { + "epoch": 0.016432739315565367, + "grad_norm": 10.537903785705566, + "learning_rate": 1.0924369747899159e-05, + "loss": 0.7271, + "step": 521 + }, + { + "epoch": 0.016464280082005994, + "grad_norm": 8.732210159301758, + "learning_rate": 1.0945378151260506e-05, + "loss": 0.7732, + "step": 522 + }, + { + "epoch": 0.016495820848446616, + "grad_norm": 11.597930908203125, + "learning_rate": 1.096638655462185e-05, + "loss": 0.7043, + "step": 523 + }, + { + "epoch": 0.016527361614887243, + "grad_norm": 14.626331329345703, + "learning_rate": 1.0987394957983195e-05, + "loss": 0.7275, + "step": 524 + }, + { + "epoch": 0.016558902381327865, + "grad_norm": 11.228914260864258, + "learning_rate": 1.100840336134454e-05, + "loss": 0.7506, + "step": 525 + }, + { + "epoch": 0.016590443147768492, + "grad_norm": 7.979309558868408, + "learning_rate": 1.1029411764705885e-05, + "loss": 0.7135, + "step": 526 + }, + { + "epoch": 0.016621983914209115, + "grad_norm": 12.818217277526855, + "learning_rate": 1.1050420168067229e-05, + "loss": 0.7419, + "step": 527 + }, + { + "epoch": 0.01665352468064974, + "grad_norm": 6.0516438484191895, + "learning_rate": 1.1071428571428572e-05, + "loss": 0.913, + "step": 528 + }, + { + "epoch": 0.016685065447090364, + "grad_norm": 2.7524378299713135, + "learning_rate": 1.1092436974789918e-05, + "loss": 0.9496, + "step": 529 + }, + { + "epoch": 0.01671660621353099, + "grad_norm": 8.75131607055664, + "learning_rate": 1.1113445378151262e-05, + "loss": 0.7382, + "step": 530 + }, + { + "epoch": 0.016748146979971613, + "grad_norm": 11.145548820495605, + "learning_rate": 1.1134453781512606e-05, + "loss": 0.718, + "step": 531 + }, + { + "epoch": 0.01677968774641224, + "grad_norm": 8.522745132446289, + "learning_rate": 1.1155462184873951e-05, + "loss": 0.6944, + "step": 532 + }, + { + "epoch": 0.016811228512852862, + "grad_norm": 12.669590950012207, + "learning_rate": 1.1176470588235295e-05, + "loss": 0.7092, + "step": 533 + }, + { + "epoch": 0.016842769279293488, + "grad_norm": 15.914468765258789, + "learning_rate": 1.119747899159664e-05, + "loss": 0.7445, + "step": 534 + }, + { + "epoch": 0.01687431004573411, + "grad_norm": 9.028297424316406, + "learning_rate": 1.1218487394957984e-05, + "loss": 0.7321, + "step": 535 + }, + { + "epoch": 0.016905850812174737, + "grad_norm": 16.98715591430664, + "learning_rate": 1.1239495798319328e-05, + "loss": 0.7119, + "step": 536 + }, + { + "epoch": 0.01693739157861536, + "grad_norm": 9.736994743347168, + "learning_rate": 1.1260504201680673e-05, + "loss": 0.71, + "step": 537 + }, + { + "epoch": 0.016968932345055986, + "grad_norm": 5.033062934875488, + "learning_rate": 1.1281512605042017e-05, + "loss": 0.8805, + "step": 538 + }, + { + "epoch": 0.01700047311149661, + "grad_norm": 10.590795516967773, + "learning_rate": 1.1302521008403363e-05, + "loss": 0.7717, + "step": 539 + }, + { + "epoch": 0.017032013877937235, + "grad_norm": 9.1745023727417, + "learning_rate": 1.1323529411764707e-05, + "loss": 0.7253, + "step": 540 + }, + { + "epoch": 0.017063554644377858, + "grad_norm": 9.917943000793457, + "learning_rate": 1.134453781512605e-05, + "loss": 0.725, + "step": 541 + }, + { + "epoch": 0.017095095410818485, + "grad_norm": 13.025169372558594, + "learning_rate": 1.1365546218487396e-05, + "loss": 0.7583, + "step": 542 + }, + { + "epoch": 0.017126636177259107, + "grad_norm": 10.59406566619873, + "learning_rate": 1.138655462184874e-05, + "loss": 0.7218, + "step": 543 + }, + { + "epoch": 0.017158176943699734, + "grad_norm": 8.122400283813477, + "learning_rate": 1.1407563025210085e-05, + "loss": 0.6978, + "step": 544 + }, + { + "epoch": 0.017189717710140356, + "grad_norm": 10.11878490447998, + "learning_rate": 1.1428571428571429e-05, + "loss": 0.7498, + "step": 545 + }, + { + "epoch": 0.01722125847658098, + "grad_norm": 11.424695014953613, + "learning_rate": 1.1449579831932773e-05, + "loss": 0.6717, + "step": 546 + }, + { + "epoch": 0.017252799243021606, + "grad_norm": 24.25213623046875, + "learning_rate": 1.1470588235294118e-05, + "loss": 0.7166, + "step": 547 + }, + { + "epoch": 0.01728434000946223, + "grad_norm": 9.674910545349121, + "learning_rate": 1.1491596638655462e-05, + "loss": 0.716, + "step": 548 + }, + { + "epoch": 0.017315880775902855, + "grad_norm": 12.075340270996094, + "learning_rate": 1.1512605042016806e-05, + "loss": 0.7297, + "step": 549 + }, + { + "epoch": 0.017347421542343477, + "grad_norm": 2.9317402839660645, + "learning_rate": 1.1533613445378151e-05, + "loss": 0.8808, + "step": 550 + }, + { + "epoch": 0.017378962308784104, + "grad_norm": 9.059237480163574, + "learning_rate": 1.1554621848739495e-05, + "loss": 0.7081, + "step": 551 + }, + { + "epoch": 0.017410503075224727, + "grad_norm": 11.104369163513184, + "learning_rate": 1.1575630252100842e-05, + "loss": 0.7151, + "step": 552 + }, + { + "epoch": 0.017442043841665353, + "grad_norm": 10.855880737304688, + "learning_rate": 1.1596638655462186e-05, + "loss": 0.7586, + "step": 553 + }, + { + "epoch": 0.017473584608105976, + "grad_norm": 8.111989974975586, + "learning_rate": 1.1617647058823532e-05, + "loss": 0.709, + "step": 554 + }, + { + "epoch": 0.017505125374546602, + "grad_norm": 7.6225104331970215, + "learning_rate": 1.1638655462184876e-05, + "loss": 0.7354, + "step": 555 + }, + { + "epoch": 0.017536666140987225, + "grad_norm": 9.178791046142578, + "learning_rate": 1.165966386554622e-05, + "loss": 0.7165, + "step": 556 + }, + { + "epoch": 0.01756820690742785, + "grad_norm": 7.251921653747559, + "learning_rate": 1.1680672268907565e-05, + "loss": 0.7082, + "step": 557 + }, + { + "epoch": 0.017599747673868474, + "grad_norm": 9.456345558166504, + "learning_rate": 1.1701680672268909e-05, + "loss": 0.7148, + "step": 558 + }, + { + "epoch": 0.0176312884403091, + "grad_norm": 8.780783653259277, + "learning_rate": 1.1722689075630254e-05, + "loss": 0.7446, + "step": 559 + }, + { + "epoch": 0.017662829206749723, + "grad_norm": 4.283600330352783, + "learning_rate": 1.1743697478991598e-05, + "loss": 0.4197, + "step": 560 + }, + { + "epoch": 0.01769436997319035, + "grad_norm": 9.91750431060791, + "learning_rate": 1.1764705882352942e-05, + "loss": 0.7157, + "step": 561 + }, + { + "epoch": 0.017725910739630972, + "grad_norm": 18.10777473449707, + "learning_rate": 1.1785714285714287e-05, + "loss": 0.7212, + "step": 562 + }, + { + "epoch": 0.0177574515060716, + "grad_norm": 8.340560913085938, + "learning_rate": 1.1806722689075631e-05, + "loss": 0.6945, + "step": 563 + }, + { + "epoch": 0.01778899227251222, + "grad_norm": 11.90389633178711, + "learning_rate": 1.1827731092436977e-05, + "loss": 0.7495, + "step": 564 + }, + { + "epoch": 0.017820533038952847, + "grad_norm": 9.975460052490234, + "learning_rate": 1.184873949579832e-05, + "loss": 0.735, + "step": 565 + }, + { + "epoch": 0.01785207380539347, + "grad_norm": 3.681349992752075, + "learning_rate": 1.1869747899159664e-05, + "loss": 0.9062, + "step": 566 + }, + { + "epoch": 0.017883614571834097, + "grad_norm": 11.399998664855957, + "learning_rate": 1.189075630252101e-05, + "loss": 0.7094, + "step": 567 + }, + { + "epoch": 0.01791515533827472, + "grad_norm": 3.1054255962371826, + "learning_rate": 1.1911764705882354e-05, + "loss": 0.4049, + "step": 568 + }, + { + "epoch": 0.017946696104715346, + "grad_norm": 22.3839168548584, + "learning_rate": 1.1932773109243699e-05, + "loss": 0.7195, + "step": 569 + }, + { + "epoch": 0.01797823687115597, + "grad_norm": 19.156631469726562, + "learning_rate": 1.1953781512605043e-05, + "loss": 0.7053, + "step": 570 + }, + { + "epoch": 0.018009777637596595, + "grad_norm": 9.144694328308105, + "learning_rate": 1.1974789915966387e-05, + "loss": 0.7265, + "step": 571 + }, + { + "epoch": 0.018041318404037218, + "grad_norm": 9.245054244995117, + "learning_rate": 1.1995798319327732e-05, + "loss": 0.6901, + "step": 572 + }, + { + "epoch": 0.018072859170477844, + "grad_norm": 7.877984523773193, + "learning_rate": 1.2016806722689076e-05, + "loss": 0.702, + "step": 573 + }, + { + "epoch": 0.018104399936918467, + "grad_norm": 16.199777603149414, + "learning_rate": 1.203781512605042e-05, + "loss": 0.7279, + "step": 574 + }, + { + "epoch": 0.018135940703359093, + "grad_norm": 4.027250289916992, + "learning_rate": 1.2058823529411765e-05, + "loss": 0.9119, + "step": 575 + }, + { + "epoch": 0.018167481469799716, + "grad_norm": 23.18947982788086, + "learning_rate": 1.207983193277311e-05, + "loss": 0.7317, + "step": 576 + }, + { + "epoch": 0.018199022236240342, + "grad_norm": 21.533777236938477, + "learning_rate": 1.2100840336134455e-05, + "loss": 0.7275, + "step": 577 + }, + { + "epoch": 0.018230563002680965, + "grad_norm": 5.531106948852539, + "learning_rate": 1.2121848739495798e-05, + "loss": 0.9007, + "step": 578 + }, + { + "epoch": 0.01826210376912159, + "grad_norm": 3.738276958465576, + "learning_rate": 1.2142857142857142e-05, + "loss": 0.8967, + "step": 579 + }, + { + "epoch": 0.018293644535562214, + "grad_norm": 2.811868667602539, + "learning_rate": 1.2163865546218488e-05, + "loss": 0.9092, + "step": 580 + }, + { + "epoch": 0.01832518530200284, + "grad_norm": 12.20608139038086, + "learning_rate": 1.2184873949579832e-05, + "loss": 0.7363, + "step": 581 + }, + { + "epoch": 0.018356726068443463, + "grad_norm": 14.082523345947266, + "learning_rate": 1.2205882352941179e-05, + "loss": 0.755, + "step": 582 + }, + { + "epoch": 0.01838826683488409, + "grad_norm": 9.666696548461914, + "learning_rate": 1.2226890756302523e-05, + "loss": 0.7161, + "step": 583 + }, + { + "epoch": 0.018419807601324712, + "grad_norm": 5.387965679168701, + "learning_rate": 1.2247899159663868e-05, + "loss": 0.4059, + "step": 584 + }, + { + "epoch": 0.01845134836776534, + "grad_norm": 19.75994873046875, + "learning_rate": 1.2268907563025212e-05, + "loss": 0.7677, + "step": 585 + }, + { + "epoch": 0.01848288913420596, + "grad_norm": 12.31340503692627, + "learning_rate": 1.2289915966386556e-05, + "loss": 0.7867, + "step": 586 + }, + { + "epoch": 0.018514429900646584, + "grad_norm": 13.38759994506836, + "learning_rate": 1.2310924369747901e-05, + "loss": 0.7757, + "step": 587 + }, + { + "epoch": 0.01854597066708721, + "grad_norm": 9.352699279785156, + "learning_rate": 1.2331932773109245e-05, + "loss": 0.3927, + "step": 588 + }, + { + "epoch": 0.018577511433527833, + "grad_norm": 17.18666648864746, + "learning_rate": 1.235294117647059e-05, + "loss": 0.7994, + "step": 589 + }, + { + "epoch": 0.01860905219996846, + "grad_norm": 16.388505935668945, + "learning_rate": 1.2373949579831934e-05, + "loss": 0.821, + "step": 590 + }, + { + "epoch": 0.018640592966409082, + "grad_norm": 12.7769775390625, + "learning_rate": 1.2394957983193278e-05, + "loss": 0.7423, + "step": 591 + }, + { + "epoch": 0.01867213373284971, + "grad_norm": 9.379735946655273, + "learning_rate": 1.2415966386554624e-05, + "loss": 0.7654, + "step": 592 + }, + { + "epoch": 0.01870367449929033, + "grad_norm": 9.85220718383789, + "learning_rate": 1.2436974789915967e-05, + "loss": 0.776, + "step": 593 + }, + { + "epoch": 0.018735215265730958, + "grad_norm": 10.812026023864746, + "learning_rate": 1.2457983193277311e-05, + "loss": 0.7551, + "step": 594 + }, + { + "epoch": 0.01876675603217158, + "grad_norm": 11.338605880737305, + "learning_rate": 1.2478991596638657e-05, + "loss": 0.7259, + "step": 595 + }, + { + "epoch": 0.018798296798612207, + "grad_norm": 7.274777412414551, + "learning_rate": 1.25e-05, + "loss": 0.7331, + "step": 596 + }, + { + "epoch": 0.01882983756505283, + "grad_norm": 8.285626411437988, + "learning_rate": 1.2521008403361346e-05, + "loss": 0.7631, + "step": 597 + }, + { + "epoch": 0.018861378331493456, + "grad_norm": 8.896352767944336, + "learning_rate": 1.254201680672269e-05, + "loss": 0.7467, + "step": 598 + }, + { + "epoch": 0.01889291909793408, + "grad_norm": 8.527580261230469, + "learning_rate": 1.2563025210084034e-05, + "loss": 0.7481, + "step": 599 + }, + { + "epoch": 0.018924459864374705, + "grad_norm": 10.906002044677734, + "learning_rate": 1.258403361344538e-05, + "loss": 0.761, + "step": 600 + }, + { + "epoch": 0.018956000630815328, + "grad_norm": 7.444272518157959, + "learning_rate": 1.2605042016806723e-05, + "loss": 0.74, + "step": 601 + }, + { + "epoch": 0.018987541397255954, + "grad_norm": 10.00499153137207, + "learning_rate": 1.2626050420168069e-05, + "loss": 0.7251, + "step": 602 + }, + { + "epoch": 0.019019082163696577, + "grad_norm": 8.771625518798828, + "learning_rate": 1.2647058823529412e-05, + "loss": 0.7258, + "step": 603 + }, + { + "epoch": 0.019050622930137203, + "grad_norm": 9.983410835266113, + "learning_rate": 1.2668067226890756e-05, + "loss": 0.6904, + "step": 604 + }, + { + "epoch": 0.019082163696577826, + "grad_norm": 9.208945274353027, + "learning_rate": 1.2689075630252102e-05, + "loss": 0.7548, + "step": 605 + }, + { + "epoch": 0.019113704463018452, + "grad_norm": 8.844799041748047, + "learning_rate": 1.2710084033613445e-05, + "loss": 0.7172, + "step": 606 + }, + { + "epoch": 0.019145245229459075, + "grad_norm": 7.51606559753418, + "learning_rate": 1.2731092436974791e-05, + "loss": 0.7249, + "step": 607 + }, + { + "epoch": 0.0191767859958997, + "grad_norm": 6.7394256591796875, + "learning_rate": 1.2752100840336135e-05, + "loss": 0.4034, + "step": 608 + }, + { + "epoch": 0.019208326762340324, + "grad_norm": 10.811071395874023, + "learning_rate": 1.2773109243697479e-05, + "loss": 0.7566, + "step": 609 + }, + { + "epoch": 0.01923986752878095, + "grad_norm": 7.871477127075195, + "learning_rate": 1.2794117647058824e-05, + "loss": 0.9115, + "step": 610 + }, + { + "epoch": 0.019271408295221573, + "grad_norm": 9.805269241333008, + "learning_rate": 1.281512605042017e-05, + "loss": 0.7635, + "step": 611 + }, + { + "epoch": 0.0193029490616622, + "grad_norm": 3.872938394546509, + "learning_rate": 1.2836134453781515e-05, + "loss": 0.9363, + "step": 612 + }, + { + "epoch": 0.019334489828102822, + "grad_norm": 13.10378646850586, + "learning_rate": 1.2857142857142859e-05, + "loss": 0.7333, + "step": 613 + }, + { + "epoch": 0.01936603059454345, + "grad_norm": 9.495596885681152, + "learning_rate": 1.2878151260504204e-05, + "loss": 0.7599, + "step": 614 + }, + { + "epoch": 0.01939757136098407, + "grad_norm": 8.540560722351074, + "learning_rate": 1.2899159663865548e-05, + "loss": 0.7176, + "step": 615 + }, + { + "epoch": 0.019429112127424698, + "grad_norm": 6.949551582336426, + "learning_rate": 1.2920168067226892e-05, + "loss": 0.7375, + "step": 616 + }, + { + "epoch": 0.01946065289386532, + "grad_norm": 8.460576057434082, + "learning_rate": 1.2941176470588238e-05, + "loss": 0.7606, + "step": 617 + }, + { + "epoch": 0.019492193660305947, + "grad_norm": 6.957016944885254, + "learning_rate": 1.2962184873949581e-05, + "loss": 0.7257, + "step": 618 + }, + { + "epoch": 0.01952373442674657, + "grad_norm": 7.958094120025635, + "learning_rate": 1.2983193277310925e-05, + "loss": 0.7303, + "step": 619 + }, + { + "epoch": 0.019555275193187196, + "grad_norm": 4.451076507568359, + "learning_rate": 1.300420168067227e-05, + "loss": 0.8939, + "step": 620 + }, + { + "epoch": 0.01958681595962782, + "grad_norm": 6.542793273925781, + "learning_rate": 1.3025210084033614e-05, + "loss": 0.6903, + "step": 621 + }, + { + "epoch": 0.019618356726068445, + "grad_norm": 6.999786853790283, + "learning_rate": 1.304621848739496e-05, + "loss": 0.7174, + "step": 622 + }, + { + "epoch": 0.019649897492509068, + "grad_norm": 6.190614700317383, + "learning_rate": 1.3067226890756304e-05, + "loss": 0.7246, + "step": 623 + }, + { + "epoch": 0.019681438258949694, + "grad_norm": 6.29679012298584, + "learning_rate": 1.3088235294117648e-05, + "loss": 0.7287, + "step": 624 + }, + { + "epoch": 0.019712979025390317, + "grad_norm": 4.484778881072998, + "learning_rate": 1.3109243697478993e-05, + "loss": 0.9024, + "step": 625 + }, + { + "epoch": 0.019744519791830943, + "grad_norm": 4.522465705871582, + "learning_rate": 1.3130252100840337e-05, + "loss": 0.4243, + "step": 626 + }, + { + "epoch": 0.019776060558271566, + "grad_norm": 3.360135078430176, + "learning_rate": 1.3151260504201682e-05, + "loss": 0.3909, + "step": 627 + }, + { + "epoch": 0.019807601324712192, + "grad_norm": 8.127249717712402, + "learning_rate": 1.3172268907563026e-05, + "loss": 0.7616, + "step": 628 + }, + { + "epoch": 0.019839142091152815, + "grad_norm": 9.113373756408691, + "learning_rate": 1.319327731092437e-05, + "loss": 0.7572, + "step": 629 + }, + { + "epoch": 0.019870682857593438, + "grad_norm": 5.435665607452393, + "learning_rate": 1.3214285714285716e-05, + "loss": 0.3932, + "step": 630 + }, + { + "epoch": 0.019902223624034064, + "grad_norm": 5.304680347442627, + "learning_rate": 1.323529411764706e-05, + "loss": 0.8942, + "step": 631 + }, + { + "epoch": 0.019933764390474687, + "grad_norm": 5.776784896850586, + "learning_rate": 1.3256302521008403e-05, + "loss": 0.7154, + "step": 632 + }, + { + "epoch": 0.019965305156915313, + "grad_norm": 5.93152379989624, + "learning_rate": 1.3277310924369749e-05, + "loss": 0.4128, + "step": 633 + }, + { + "epoch": 0.019996845923355936, + "grad_norm": 7.088278770446777, + "learning_rate": 1.3298319327731092e-05, + "loss": 0.7101, + "step": 634 + }, + { + "epoch": 0.020028386689796562, + "grad_norm": 7.606856346130371, + "learning_rate": 1.3319327731092438e-05, + "loss": 0.733, + "step": 635 + }, + { + "epoch": 0.020059927456237185, + "grad_norm": 11.106855392456055, + "learning_rate": 1.3340336134453782e-05, + "loss": 0.6987, + "step": 636 + }, + { + "epoch": 0.02009146822267781, + "grad_norm": 6.00390625, + "learning_rate": 1.3361344537815126e-05, + "loss": 0.7358, + "step": 637 + }, + { + "epoch": 0.020123008989118434, + "grad_norm": 42.32028579711914, + "learning_rate": 1.3382352941176471e-05, + "loss": 0.7325, + "step": 638 + }, + { + "epoch": 0.02015454975555906, + "grad_norm": 32.53657150268555, + "learning_rate": 1.3403361344537815e-05, + "loss": 0.7265, + "step": 639 + }, + { + "epoch": 0.020186090521999683, + "grad_norm": 12.576626777648926, + "learning_rate": 1.342436974789916e-05, + "loss": 0.7342, + "step": 640 + }, + { + "epoch": 0.02021763128844031, + "grad_norm": 18.388864517211914, + "learning_rate": 1.3445378151260506e-05, + "loss": 0.7324, + "step": 641 + }, + { + "epoch": 0.020249172054880932, + "grad_norm": 7.517066955566406, + "learning_rate": 1.3466386554621851e-05, + "loss": 0.7066, + "step": 642 + }, + { + "epoch": 0.02028071282132156, + "grad_norm": 7.507174968719482, + "learning_rate": 1.3487394957983195e-05, + "loss": 0.7327, + "step": 643 + }, + { + "epoch": 0.02031225358776218, + "grad_norm": 6.538325786590576, + "learning_rate": 1.3508403361344539e-05, + "loss": 0.753, + "step": 644 + }, + { + "epoch": 0.020343794354202808, + "grad_norm": 14.668622016906738, + "learning_rate": 1.3529411764705885e-05, + "loss": 0.7558, + "step": 645 + }, + { + "epoch": 0.02037533512064343, + "grad_norm": 7.157096862792969, + "learning_rate": 1.3550420168067228e-05, + "loss": 0.7162, + "step": 646 + }, + { + "epoch": 0.020406875887084057, + "grad_norm": 9.766092300415039, + "learning_rate": 1.3571428571428574e-05, + "loss": 0.7349, + "step": 647 + }, + { + "epoch": 0.02043841665352468, + "grad_norm": 7.026566982269287, + "learning_rate": 1.3592436974789918e-05, + "loss": 0.7058, + "step": 648 + }, + { + "epoch": 0.020469957419965306, + "grad_norm": 10.891143798828125, + "learning_rate": 1.3613445378151261e-05, + "loss": 0.9353, + "step": 649 + }, + { + "epoch": 0.02050149818640593, + "grad_norm": 7.424005031585693, + "learning_rate": 1.3634453781512607e-05, + "loss": 0.7241, + "step": 650 + }, + { + "epoch": 0.020533038952846555, + "grad_norm": 8.629715919494629, + "learning_rate": 1.365546218487395e-05, + "loss": 0.6772, + "step": 651 + }, + { + "epoch": 0.020564579719287178, + "grad_norm": 16.162580490112305, + "learning_rate": 1.3676470588235296e-05, + "loss": 0.7479, + "step": 652 + }, + { + "epoch": 0.020596120485727804, + "grad_norm": 6.402883529663086, + "learning_rate": 1.369747899159664e-05, + "loss": 0.7061, + "step": 653 + }, + { + "epoch": 0.020627661252168427, + "grad_norm": 7.012611389160156, + "learning_rate": 1.3718487394957984e-05, + "loss": 0.7126, + "step": 654 + }, + { + "epoch": 0.020659202018609053, + "grad_norm": 6.686916828155518, + "learning_rate": 1.373949579831933e-05, + "loss": 0.4297, + "step": 655 + }, + { + "epoch": 0.020690742785049676, + "grad_norm": 11.004369735717773, + "learning_rate": 1.3760504201680673e-05, + "loss": 0.737, + "step": 656 + }, + { + "epoch": 0.020722283551490302, + "grad_norm": 8.373412132263184, + "learning_rate": 1.3781512605042017e-05, + "loss": 0.699, + "step": 657 + }, + { + "epoch": 0.020753824317930925, + "grad_norm": 7.148489475250244, + "learning_rate": 1.3802521008403363e-05, + "loss": 0.7103, + "step": 658 + }, + { + "epoch": 0.02078536508437155, + "grad_norm": 7.009701728820801, + "learning_rate": 1.3823529411764706e-05, + "loss": 0.7061, + "step": 659 + }, + { + "epoch": 0.020816905850812174, + "grad_norm": 6.035659313201904, + "learning_rate": 1.3844537815126052e-05, + "loss": 0.9113, + "step": 660 + }, + { + "epoch": 0.0208484466172528, + "grad_norm": 9.902194023132324, + "learning_rate": 1.3865546218487396e-05, + "loss": 0.7591, + "step": 661 + }, + { + "epoch": 0.020879987383693423, + "grad_norm": 10.61933708190918, + "learning_rate": 1.388655462184874e-05, + "loss": 0.7218, + "step": 662 + }, + { + "epoch": 0.02091152815013405, + "grad_norm": 21.159019470214844, + "learning_rate": 1.3907563025210085e-05, + "loss": 0.669, + "step": 663 + }, + { + "epoch": 0.020943068916574673, + "grad_norm": 9.160653114318848, + "learning_rate": 1.3928571428571429e-05, + "loss": 0.7408, + "step": 664 + }, + { + "epoch": 0.0209746096830153, + "grad_norm": 8.072001457214355, + "learning_rate": 1.3949579831932774e-05, + "loss": 0.7064, + "step": 665 + }, + { + "epoch": 0.02100615044945592, + "grad_norm": 10.767780303955078, + "learning_rate": 1.3970588235294118e-05, + "loss": 0.7195, + "step": 666 + }, + { + "epoch": 0.021037691215896548, + "grad_norm": 9.363905906677246, + "learning_rate": 1.3991596638655462e-05, + "loss": 0.7101, + "step": 667 + }, + { + "epoch": 0.02106923198233717, + "grad_norm": 3.3684308528900146, + "learning_rate": 1.4012605042016807e-05, + "loss": 0.9545, + "step": 668 + }, + { + "epoch": 0.021100772748777797, + "grad_norm": 8.459270477294922, + "learning_rate": 1.4033613445378151e-05, + "loss": 0.4652, + "step": 669 + }, + { + "epoch": 0.02113231351521842, + "grad_norm": 3.6672306060791016, + "learning_rate": 1.4054621848739495e-05, + "loss": 0.3788, + "step": 670 + }, + { + "epoch": 0.021163854281659043, + "grad_norm": 12.43526554107666, + "learning_rate": 1.4075630252100842e-05, + "loss": 0.7927, + "step": 671 + }, + { + "epoch": 0.02119539504809967, + "grad_norm": 8.872773170471191, + "learning_rate": 1.4096638655462188e-05, + "loss": 0.6861, + "step": 672 + }, + { + "epoch": 0.02122693581454029, + "grad_norm": 24.63551139831543, + "learning_rate": 1.4117647058823532e-05, + "loss": 0.699, + "step": 673 + }, + { + "epoch": 0.021258476580980918, + "grad_norm": 8.725772857666016, + "learning_rate": 1.4138655462184875e-05, + "loss": 0.6953, + "step": 674 + }, + { + "epoch": 0.02129001734742154, + "grad_norm": 9.243935585021973, + "learning_rate": 1.4159663865546221e-05, + "loss": 0.7283, + "step": 675 + }, + { + "epoch": 0.021321558113862167, + "grad_norm": 7.179399490356445, + "learning_rate": 1.4180672268907565e-05, + "loss": 0.7089, + "step": 676 + }, + { + "epoch": 0.02135309888030279, + "grad_norm": 13.574484825134277, + "learning_rate": 1.4201680672268908e-05, + "loss": 0.7298, + "step": 677 + }, + { + "epoch": 0.021384639646743416, + "grad_norm": 5.860383033752441, + "learning_rate": 1.4222689075630254e-05, + "loss": 0.7152, + "step": 678 + }, + { + "epoch": 0.02141618041318404, + "grad_norm": 12.836163520812988, + "learning_rate": 1.4243697478991598e-05, + "loss": 0.744, + "step": 679 + }, + { + "epoch": 0.021447721179624665, + "grad_norm": 13.250812530517578, + "learning_rate": 1.4264705882352943e-05, + "loss": 0.7424, + "step": 680 + }, + { + "epoch": 0.021479261946065288, + "grad_norm": 3.214172840118408, + "learning_rate": 1.4285714285714287e-05, + "loss": 0.9363, + "step": 681 + }, + { + "epoch": 0.021510802712505914, + "grad_norm": 49.30610275268555, + "learning_rate": 1.4306722689075631e-05, + "loss": 0.7049, + "step": 682 + }, + { + "epoch": 0.021542343478946537, + "grad_norm": 14.941949844360352, + "learning_rate": 1.4327731092436976e-05, + "loss": 0.7092, + "step": 683 + }, + { + "epoch": 0.021573884245387163, + "grad_norm": 6.965025424957275, + "learning_rate": 1.434873949579832e-05, + "loss": 0.7248, + "step": 684 + }, + { + "epoch": 0.021605425011827786, + "grad_norm": 5.300570487976074, + "learning_rate": 1.4369747899159666e-05, + "loss": 0.388, + "step": 685 + }, + { + "epoch": 0.021636965778268413, + "grad_norm": 16.35323715209961, + "learning_rate": 1.439075630252101e-05, + "loss": 0.714, + "step": 686 + }, + { + "epoch": 0.021668506544709035, + "grad_norm": 4.389123439788818, + "learning_rate": 1.4411764705882353e-05, + "loss": 0.9208, + "step": 687 + }, + { + "epoch": 0.02170004731114966, + "grad_norm": 4.111270427703857, + "learning_rate": 1.4432773109243699e-05, + "loss": 0.9264, + "step": 688 + }, + { + "epoch": 0.021731588077590284, + "grad_norm": 3.2217178344726562, + "learning_rate": 1.4453781512605043e-05, + "loss": 0.3906, + "step": 689 + }, + { + "epoch": 0.02176312884403091, + "grad_norm": 8.048584938049316, + "learning_rate": 1.4474789915966388e-05, + "loss": 0.7595, + "step": 690 + }, + { + "epoch": 0.021794669610471534, + "grad_norm": 7.299341201782227, + "learning_rate": 1.4495798319327732e-05, + "loss": 0.7233, + "step": 691 + }, + { + "epoch": 0.02182621037691216, + "grad_norm": 22.964696884155273, + "learning_rate": 1.4516806722689076e-05, + "loss": 0.428, + "step": 692 + }, + { + "epoch": 0.021857751143352783, + "grad_norm": 8.804468154907227, + "learning_rate": 1.4537815126050421e-05, + "loss": 0.721, + "step": 693 + }, + { + "epoch": 0.02188929190979341, + "grad_norm": 8.20469856262207, + "learning_rate": 1.4558823529411765e-05, + "loss": 0.7131, + "step": 694 + }, + { + "epoch": 0.021920832676234032, + "grad_norm": 3.745213031768799, + "learning_rate": 1.4579831932773109e-05, + "loss": 0.9171, + "step": 695 + }, + { + "epoch": 0.021952373442674658, + "grad_norm": 8.95540714263916, + "learning_rate": 1.4600840336134454e-05, + "loss": 0.7287, + "step": 696 + }, + { + "epoch": 0.02198391420911528, + "grad_norm": 9.503332138061523, + "learning_rate": 1.4621848739495798e-05, + "loss": 0.7477, + "step": 697 + }, + { + "epoch": 0.022015454975555907, + "grad_norm": 6.156070709228516, + "learning_rate": 1.4642857142857144e-05, + "loss": 0.6926, + "step": 698 + }, + { + "epoch": 0.02204699574199653, + "grad_norm": 9.866971969604492, + "learning_rate": 1.4663865546218488e-05, + "loss": 0.7309, + "step": 699 + }, + { + "epoch": 0.022078536508437156, + "grad_norm": 5.616145610809326, + "learning_rate": 1.4684873949579831e-05, + "loss": 0.7074, + "step": 700 + }, + { + "epoch": 0.02211007727487778, + "grad_norm": 5.622080326080322, + "learning_rate": 1.4705882352941179e-05, + "loss": 0.6812, + "step": 701 + }, + { + "epoch": 0.022141618041318405, + "grad_norm": 7.33459997177124, + "learning_rate": 1.4726890756302522e-05, + "loss": 0.7164, + "step": 702 + }, + { + "epoch": 0.022173158807759028, + "grad_norm": 3.559483051300049, + "learning_rate": 1.4747899159663868e-05, + "loss": 0.8875, + "step": 703 + }, + { + "epoch": 0.022204699574199654, + "grad_norm": 7.914633274078369, + "learning_rate": 1.4768907563025212e-05, + "loss": 0.7726, + "step": 704 + }, + { + "epoch": 0.022236240340640277, + "grad_norm": 11.298602104187012, + "learning_rate": 1.4789915966386557e-05, + "loss": 0.6995, + "step": 705 + }, + { + "epoch": 0.022267781107080904, + "grad_norm": 15.686347007751465, + "learning_rate": 1.4810924369747901e-05, + "loss": 0.735, + "step": 706 + }, + { + "epoch": 0.022299321873521526, + "grad_norm": 8.483464241027832, + "learning_rate": 1.4831932773109245e-05, + "loss": 0.7233, + "step": 707 + }, + { + "epoch": 0.022330862639962153, + "grad_norm": 3.1227939128875732, + "learning_rate": 1.485294117647059e-05, + "loss": 0.4317, + "step": 708 + }, + { + "epoch": 0.022362403406402775, + "grad_norm": 2.6348321437835693, + "learning_rate": 1.4873949579831934e-05, + "loss": 0.4049, + "step": 709 + }, + { + "epoch": 0.022393944172843402, + "grad_norm": 12.748536109924316, + "learning_rate": 1.489495798319328e-05, + "loss": 0.7807, + "step": 710 + }, + { + "epoch": 0.022425484939284025, + "grad_norm": 11.334344863891602, + "learning_rate": 1.4915966386554623e-05, + "loss": 0.7196, + "step": 711 + }, + { + "epoch": 0.02245702570572465, + "grad_norm": 9.010950088500977, + "learning_rate": 1.4936974789915967e-05, + "loss": 0.7264, + "step": 712 + }, + { + "epoch": 0.022488566472165274, + "grad_norm": 8.522418022155762, + "learning_rate": 1.4957983193277313e-05, + "loss": 0.7414, + "step": 713 + }, + { + "epoch": 0.022520107238605896, + "grad_norm": 10.520930290222168, + "learning_rate": 1.4978991596638657e-05, + "loss": 0.7318, + "step": 714 + }, + { + "epoch": 0.022551648005046523, + "grad_norm": 8.683189392089844, + "learning_rate": 1.5000000000000002e-05, + "loss": 0.7186, + "step": 715 + }, + { + "epoch": 0.022583188771487146, + "grad_norm": 8.134101867675781, + "learning_rate": 1.5021008403361346e-05, + "loss": 0.7585, + "step": 716 + }, + { + "epoch": 0.022614729537927772, + "grad_norm": 12.630802154541016, + "learning_rate": 1.504201680672269e-05, + "loss": 0.7306, + "step": 717 + }, + { + "epoch": 0.022646270304368395, + "grad_norm": 7.555576801300049, + "learning_rate": 1.5063025210084035e-05, + "loss": 0.74, + "step": 718 + }, + { + "epoch": 0.02267781107080902, + "grad_norm": 30.942188262939453, + "learning_rate": 1.5084033613445379e-05, + "loss": 0.7208, + "step": 719 + }, + { + "epoch": 0.022709351837249644, + "grad_norm": 7.493412494659424, + "learning_rate": 1.5105042016806723e-05, + "loss": 0.7638, + "step": 720 + }, + { + "epoch": 0.02274089260369027, + "grad_norm": 6.474618911743164, + "learning_rate": 1.5126050420168068e-05, + "loss": 0.7199, + "step": 721 + }, + { + "epoch": 0.022772433370130893, + "grad_norm": 8.910323143005371, + "learning_rate": 1.5147058823529412e-05, + "loss": 0.7348, + "step": 722 + }, + { + "epoch": 0.02280397413657152, + "grad_norm": 7.368350028991699, + "learning_rate": 1.5168067226890758e-05, + "loss": 0.703, + "step": 723 + }, + { + "epoch": 0.022835514903012142, + "grad_norm": 7.58727502822876, + "learning_rate": 1.5189075630252101e-05, + "loss": 0.7363, + "step": 724 + }, + { + "epoch": 0.022867055669452768, + "grad_norm": 7.196917533874512, + "learning_rate": 1.5210084033613445e-05, + "loss": 0.7092, + "step": 725 + }, + { + "epoch": 0.02289859643589339, + "grad_norm": 7.270395278930664, + "learning_rate": 1.523109243697479e-05, + "loss": 0.658, + "step": 726 + }, + { + "epoch": 0.022930137202334017, + "grad_norm": 6.6760573387146, + "learning_rate": 1.5252100840336135e-05, + "loss": 0.7368, + "step": 727 + }, + { + "epoch": 0.02296167796877464, + "grad_norm": 8.430251121520996, + "learning_rate": 1.527310924369748e-05, + "loss": 0.7526, + "step": 728 + }, + { + "epoch": 0.022993218735215266, + "grad_norm": 2.9548490047454834, + "learning_rate": 1.5294117647058822e-05, + "loss": 0.4024, + "step": 729 + }, + { + "epoch": 0.02302475950165589, + "grad_norm": 11.034789085388184, + "learning_rate": 1.531512605042017e-05, + "loss": 0.7047, + "step": 730 + }, + { + "epoch": 0.023056300268096516, + "grad_norm": 9.810876846313477, + "learning_rate": 1.5336134453781513e-05, + "loss": 0.7106, + "step": 731 + }, + { + "epoch": 0.02308784103453714, + "grad_norm": 6.895846366882324, + "learning_rate": 1.535714285714286e-05, + "loss": 0.6988, + "step": 732 + }, + { + "epoch": 0.023119381800977765, + "grad_norm": 8.547887802124023, + "learning_rate": 1.5378151260504204e-05, + "loss": 0.7017, + "step": 733 + }, + { + "epoch": 0.023150922567418387, + "grad_norm": 15.871573448181152, + "learning_rate": 1.539915966386555e-05, + "loss": 0.7204, + "step": 734 + }, + { + "epoch": 0.023182463333859014, + "grad_norm": 7.721304893493652, + "learning_rate": 1.5420168067226892e-05, + "loss": 0.6965, + "step": 735 + }, + { + "epoch": 0.023214004100299637, + "grad_norm": 12.279529571533203, + "learning_rate": 1.5441176470588237e-05, + "loss": 0.7035, + "step": 736 + }, + { + "epoch": 0.023245544866740263, + "grad_norm": 10.836344718933105, + "learning_rate": 1.5462184873949583e-05, + "loss": 0.7313, + "step": 737 + }, + { + "epoch": 0.023277085633180886, + "grad_norm": 9.731264114379883, + "learning_rate": 1.5483193277310925e-05, + "loss": 0.7146, + "step": 738 + }, + { + "epoch": 0.023308626399621512, + "grad_norm": 10.158625602722168, + "learning_rate": 1.550420168067227e-05, + "loss": 0.7114, + "step": 739 + }, + { + "epoch": 0.023340167166062135, + "grad_norm": 15.818912506103516, + "learning_rate": 1.5525210084033616e-05, + "loss": 0.6919, + "step": 740 + }, + { + "epoch": 0.02337170793250276, + "grad_norm": 7.666866302490234, + "learning_rate": 1.5546218487394958e-05, + "loss": 0.6977, + "step": 741 + }, + { + "epoch": 0.023403248698943384, + "grad_norm": 8.160807609558105, + "learning_rate": 1.5567226890756304e-05, + "loss": 0.7051, + "step": 742 + }, + { + "epoch": 0.02343478946538401, + "grad_norm": 8.804266929626465, + "learning_rate": 1.558823529411765e-05, + "loss": 0.7178, + "step": 743 + }, + { + "epoch": 0.023466330231824633, + "grad_norm": 8.603964805603027, + "learning_rate": 1.5609243697478995e-05, + "loss": 0.7116, + "step": 744 + }, + { + "epoch": 0.02349787099826526, + "grad_norm": 8.538694381713867, + "learning_rate": 1.5630252100840337e-05, + "loss": 0.709, + "step": 745 + }, + { + "epoch": 0.023529411764705882, + "grad_norm": 6.132279872894287, + "learning_rate": 1.5651260504201682e-05, + "loss": 0.7122, + "step": 746 + }, + { + "epoch": 0.02356095253114651, + "grad_norm": 22.800064086914062, + "learning_rate": 1.5672268907563028e-05, + "loss": 0.7114, + "step": 747 + }, + { + "epoch": 0.02359249329758713, + "grad_norm": 9.658668518066406, + "learning_rate": 1.569327731092437e-05, + "loss": 0.6636, + "step": 748 + }, + { + "epoch": 0.023624034064027757, + "grad_norm": 7.4373579025268555, + "learning_rate": 1.5714285714285715e-05, + "loss": 0.6977, + "step": 749 + }, + { + "epoch": 0.02365557483046838, + "grad_norm": 9.80417537689209, + "learning_rate": 1.573529411764706e-05, + "loss": 0.7378, + "step": 750 + }, + { + "epoch": 0.023687115596909007, + "grad_norm": 7.252391338348389, + "learning_rate": 1.5756302521008403e-05, + "loss": 0.6719, + "step": 751 + }, + { + "epoch": 0.02371865636334963, + "grad_norm": 8.977090835571289, + "learning_rate": 1.577731092436975e-05, + "loss": 0.3575, + "step": 752 + }, + { + "epoch": 0.023750197129790256, + "grad_norm": 7.775468349456787, + "learning_rate": 1.5798319327731094e-05, + "loss": 0.6991, + "step": 753 + }, + { + "epoch": 0.02378173789623088, + "grad_norm": 9.6954345703125, + "learning_rate": 1.5819327731092436e-05, + "loss": 0.6965, + "step": 754 + }, + { + "epoch": 0.0238132786626715, + "grad_norm": 7.038577556610107, + "learning_rate": 1.584033613445378e-05, + "loss": 0.7246, + "step": 755 + }, + { + "epoch": 0.023844819429112128, + "grad_norm": 7.832708835601807, + "learning_rate": 1.5861344537815127e-05, + "loss": 0.6786, + "step": 756 + }, + { + "epoch": 0.02387636019555275, + "grad_norm": 3.3762264251708984, + "learning_rate": 1.5882352941176473e-05, + "loss": 0.3999, + "step": 757 + }, + { + "epoch": 0.023907900961993377, + "grad_norm": 7.548130512237549, + "learning_rate": 1.5903361344537815e-05, + "loss": 0.7019, + "step": 758 + }, + { + "epoch": 0.023939441728434, + "grad_norm": 8.087343215942383, + "learning_rate": 1.592436974789916e-05, + "loss": 0.9282, + "step": 759 + }, + { + "epoch": 0.023970982494874626, + "grad_norm": 10.929879188537598, + "learning_rate": 1.5945378151260506e-05, + "loss": 0.7358, + "step": 760 + }, + { + "epoch": 0.02400252326131525, + "grad_norm": 7.457026481628418, + "learning_rate": 1.596638655462185e-05, + "loss": 0.7382, + "step": 761 + }, + { + "epoch": 0.024034064027755875, + "grad_norm": 7.793738842010498, + "learning_rate": 1.5987394957983197e-05, + "loss": 0.7256, + "step": 762 + }, + { + "epoch": 0.024065604794196498, + "grad_norm": 14.46395492553711, + "learning_rate": 1.600840336134454e-05, + "loss": 0.7445, + "step": 763 + }, + { + "epoch": 0.024097145560637124, + "grad_norm": 7.753976821899414, + "learning_rate": 1.6029411764705884e-05, + "loss": 0.7093, + "step": 764 + }, + { + "epoch": 0.024128686327077747, + "grad_norm": 5.231473445892334, + "learning_rate": 1.605042016806723e-05, + "loss": 0.7275, + "step": 765 + }, + { + "epoch": 0.024160227093518373, + "grad_norm": 8.111079216003418, + "learning_rate": 1.6071428571428572e-05, + "loss": 0.697, + "step": 766 + }, + { + "epoch": 0.024191767859958996, + "grad_norm": 10.019197463989258, + "learning_rate": 1.6092436974789917e-05, + "loss": 0.7124, + "step": 767 + }, + { + "epoch": 0.024223308626399622, + "grad_norm": 11.98998737335205, + "learning_rate": 1.6113445378151263e-05, + "loss": 0.7167, + "step": 768 + }, + { + "epoch": 0.024254849392840245, + "grad_norm": 8.645854949951172, + "learning_rate": 1.6134453781512605e-05, + "loss": 0.6999, + "step": 769 + }, + { + "epoch": 0.02428639015928087, + "grad_norm": 3.1894426345825195, + "learning_rate": 1.615546218487395e-05, + "loss": 0.8733, + "step": 770 + }, + { + "epoch": 0.024317930925721494, + "grad_norm": 13.88093090057373, + "learning_rate": 1.6176470588235296e-05, + "loss": 0.6997, + "step": 771 + }, + { + "epoch": 0.02434947169216212, + "grad_norm": 18.189199447631836, + "learning_rate": 1.619747899159664e-05, + "loss": 0.6984, + "step": 772 + }, + { + "epoch": 0.024381012458602743, + "grad_norm": 17.597387313842773, + "learning_rate": 1.6218487394957984e-05, + "loss": 0.7025, + "step": 773 + }, + { + "epoch": 0.02441255322504337, + "grad_norm": 3.093254804611206, + "learning_rate": 1.623949579831933e-05, + "loss": 0.4438, + "step": 774 + }, + { + "epoch": 0.024444093991483992, + "grad_norm": 11.008865356445312, + "learning_rate": 1.6260504201680675e-05, + "loss": 0.7079, + "step": 775 + }, + { + "epoch": 0.02447563475792462, + "grad_norm": 7.918716907501221, + "learning_rate": 1.6281512605042017e-05, + "loss": 0.7327, + "step": 776 + }, + { + "epoch": 0.02450717552436524, + "grad_norm": 11.144440650939941, + "learning_rate": 1.6302521008403362e-05, + "loss": 0.7446, + "step": 777 + }, + { + "epoch": 0.024538716290805868, + "grad_norm": 12.567713737487793, + "learning_rate": 1.6323529411764708e-05, + "loss": 0.683, + "step": 778 + }, + { + "epoch": 0.02457025705724649, + "grad_norm": 11.22558307647705, + "learning_rate": 1.634453781512605e-05, + "loss": 0.6997, + "step": 779 + }, + { + "epoch": 0.024601797823687117, + "grad_norm": 11.416648864746094, + "learning_rate": 1.6365546218487395e-05, + "loss": 0.6867, + "step": 780 + }, + { + "epoch": 0.02463333859012774, + "grad_norm": 9.494429588317871, + "learning_rate": 1.638655462184874e-05, + "loss": 0.7062, + "step": 781 + }, + { + "epoch": 0.024664879356568366, + "grad_norm": 16.139745712280273, + "learning_rate": 1.6407563025210086e-05, + "loss": 0.7268, + "step": 782 + }, + { + "epoch": 0.02469642012300899, + "grad_norm": 12.9396333694458, + "learning_rate": 1.642857142857143e-05, + "loss": 0.6947, + "step": 783 + }, + { + "epoch": 0.024727960889449615, + "grad_norm": 9.700459480285645, + "learning_rate": 1.6449579831932774e-05, + "loss": 0.737, + "step": 784 + }, + { + "epoch": 0.024759501655890238, + "grad_norm": 11.644752502441406, + "learning_rate": 1.647058823529412e-05, + "loss": 0.7416, + "step": 785 + }, + { + "epoch": 0.024791042422330864, + "grad_norm": 3.2540931701660156, + "learning_rate": 1.6491596638655462e-05, + "loss": 0.4145, + "step": 786 + }, + { + "epoch": 0.024822583188771487, + "grad_norm": 3.2781124114990234, + "learning_rate": 1.6512605042016807e-05, + "loss": 0.9035, + "step": 787 + }, + { + "epoch": 0.024854123955212113, + "grad_norm": 14.338202476501465, + "learning_rate": 1.6533613445378153e-05, + "loss": 0.7436, + "step": 788 + }, + { + "epoch": 0.024885664721652736, + "grad_norm": 6.825802326202393, + "learning_rate": 1.6554621848739495e-05, + "loss": 0.7256, + "step": 789 + }, + { + "epoch": 0.024917205488093362, + "grad_norm": 7.440539836883545, + "learning_rate": 1.6575630252100844e-05, + "loss": 0.7249, + "step": 790 + }, + { + "epoch": 0.024948746254533985, + "grad_norm": 5.35239839553833, + "learning_rate": 1.6596638655462186e-05, + "loss": 0.6713, + "step": 791 + }, + { + "epoch": 0.02498028702097461, + "grad_norm": 7.269793510437012, + "learning_rate": 1.661764705882353e-05, + "loss": 0.7045, + "step": 792 + }, + { + "epoch": 0.025011827787415234, + "grad_norm": 6.839046001434326, + "learning_rate": 1.6638655462184877e-05, + "loss": 0.7046, + "step": 793 + }, + { + "epoch": 0.02504336855385586, + "grad_norm": 9.241779327392578, + "learning_rate": 1.665966386554622e-05, + "loss": 0.7133, + "step": 794 + }, + { + "epoch": 0.025074909320296483, + "grad_norm": 11.629417419433594, + "learning_rate": 1.6680672268907564e-05, + "loss": 0.6852, + "step": 795 + }, + { + "epoch": 0.025106450086737106, + "grad_norm": 8.296799659729004, + "learning_rate": 1.670168067226891e-05, + "loss": 0.7024, + "step": 796 + }, + { + "epoch": 0.025137990853177732, + "grad_norm": 3.299875497817993, + "learning_rate": 1.6722689075630255e-05, + "loss": 0.9344, + "step": 797 + }, + { + "epoch": 0.025169531619618355, + "grad_norm": 22.31218719482422, + "learning_rate": 1.6743697478991598e-05, + "loss": 0.711, + "step": 798 + }, + { + "epoch": 0.02520107238605898, + "grad_norm": 21.66029930114746, + "learning_rate": 1.6764705882352943e-05, + "loss": 0.6939, + "step": 799 + }, + { + "epoch": 0.025232613152499604, + "grad_norm": 7.497222423553467, + "learning_rate": 1.678571428571429e-05, + "loss": 0.6971, + "step": 800 + }, + { + "epoch": 0.02526415391894023, + "grad_norm": 8.954256057739258, + "learning_rate": 1.680672268907563e-05, + "loss": 0.7135, + "step": 801 + }, + { + "epoch": 0.025295694685380853, + "grad_norm": 10.050029754638672, + "learning_rate": 1.6827731092436976e-05, + "loss": 0.6669, + "step": 802 + }, + { + "epoch": 0.02532723545182148, + "grad_norm": 15.490678787231445, + "learning_rate": 1.6848739495798322e-05, + "loss": 0.7176, + "step": 803 + }, + { + "epoch": 0.025358776218262102, + "grad_norm": 17.215290069580078, + "learning_rate": 1.6869747899159664e-05, + "loss": 0.7303, + "step": 804 + }, + { + "epoch": 0.02539031698470273, + "grad_norm": 7.390389442443848, + "learning_rate": 1.689075630252101e-05, + "loss": 0.7306, + "step": 805 + }, + { + "epoch": 0.02542185775114335, + "grad_norm": 2.824692487716675, + "learning_rate": 1.6911764705882355e-05, + "loss": 0.9153, + "step": 806 + }, + { + "epoch": 0.025453398517583978, + "grad_norm": 10.222271919250488, + "learning_rate": 1.69327731092437e-05, + "loss": 0.6799, + "step": 807 + }, + { + "epoch": 0.0254849392840246, + "grad_norm": 7.346109867095947, + "learning_rate": 1.6953781512605042e-05, + "loss": 0.7142, + "step": 808 + }, + { + "epoch": 0.025516480050465227, + "grad_norm": 9.92850112915039, + "learning_rate": 1.6974789915966388e-05, + "loss": 0.7328, + "step": 809 + }, + { + "epoch": 0.02554802081690585, + "grad_norm": 12.081757545471191, + "learning_rate": 1.6995798319327733e-05, + "loss": 0.673, + "step": 810 + }, + { + "epoch": 0.025579561583346476, + "grad_norm": 4.395125865936279, + "learning_rate": 1.7016806722689076e-05, + "loss": 0.3639, + "step": 811 + }, + { + "epoch": 0.0256111023497871, + "grad_norm": 14.884145736694336, + "learning_rate": 1.703781512605042e-05, + "loss": 0.698, + "step": 812 + }, + { + "epoch": 0.025642643116227725, + "grad_norm": 15.326401710510254, + "learning_rate": 1.7058823529411767e-05, + "loss": 0.6926, + "step": 813 + }, + { + "epoch": 0.025674183882668348, + "grad_norm": 8.793776512145996, + "learning_rate": 1.707983193277311e-05, + "loss": 0.7382, + "step": 814 + }, + { + "epoch": 0.025705724649108974, + "grad_norm": 9.477044105529785, + "learning_rate": 1.7100840336134454e-05, + "loss": 0.6956, + "step": 815 + }, + { + "epoch": 0.025737265415549597, + "grad_norm": 12.119100570678711, + "learning_rate": 1.71218487394958e-05, + "loss": 0.6861, + "step": 816 + }, + { + "epoch": 0.025768806181990223, + "grad_norm": 9.220486640930176, + "learning_rate": 1.7142857142857142e-05, + "loss": 0.6856, + "step": 817 + }, + { + "epoch": 0.025800346948430846, + "grad_norm": 7.971264839172363, + "learning_rate": 1.7163865546218487e-05, + "loss": 0.7197, + "step": 818 + }, + { + "epoch": 0.025831887714871472, + "grad_norm": 10.989620208740234, + "learning_rate": 1.7184873949579833e-05, + "loss": 0.7336, + "step": 819 + }, + { + "epoch": 0.025863428481312095, + "grad_norm": 7.658380508422852, + "learning_rate": 1.720588235294118e-05, + "loss": 0.7274, + "step": 820 + }, + { + "epoch": 0.02589496924775272, + "grad_norm": 11.658668518066406, + "learning_rate": 1.7226890756302524e-05, + "loss": 0.7372, + "step": 821 + }, + { + "epoch": 0.025926510014193344, + "grad_norm": 8.738763809204102, + "learning_rate": 1.724789915966387e-05, + "loss": 0.9096, + "step": 822 + }, + { + "epoch": 0.02595805078063397, + "grad_norm": 9.205314636230469, + "learning_rate": 1.726890756302521e-05, + "loss": 0.6694, + "step": 823 + }, + { + "epoch": 0.025989591547074593, + "grad_norm": 4.098989486694336, + "learning_rate": 1.7289915966386557e-05, + "loss": 0.8762, + "step": 824 + }, + { + "epoch": 0.02602113231351522, + "grad_norm": 11.7003755569458, + "learning_rate": 1.7310924369747902e-05, + "loss": 0.7, + "step": 825 + }, + { + "epoch": 0.026052673079955842, + "grad_norm": 13.131993293762207, + "learning_rate": 1.7331932773109245e-05, + "loss": 0.677, + "step": 826 + }, + { + "epoch": 0.02608421384639647, + "grad_norm": 7.647923946380615, + "learning_rate": 1.735294117647059e-05, + "loss": 0.406, + "step": 827 + }, + { + "epoch": 0.02611575461283709, + "grad_norm": 9.247557640075684, + "learning_rate": 1.7373949579831936e-05, + "loss": 0.7091, + "step": 828 + }, + { + "epoch": 0.026147295379277718, + "grad_norm": 8.798431396484375, + "learning_rate": 1.7394957983193278e-05, + "loss": 0.7273, + "step": 829 + }, + { + "epoch": 0.02617883614571834, + "grad_norm": 8.552465438842773, + "learning_rate": 1.7415966386554623e-05, + "loss": 0.695, + "step": 830 + }, + { + "epoch": 0.026210376912158967, + "grad_norm": 15.298728942871094, + "learning_rate": 1.743697478991597e-05, + "loss": 0.6812, + "step": 831 + }, + { + "epoch": 0.02624191767859959, + "grad_norm": 14.362975120544434, + "learning_rate": 1.745798319327731e-05, + "loss": 0.7049, + "step": 832 + }, + { + "epoch": 0.026273458445040216, + "grad_norm": 4.365817546844482, + "learning_rate": 1.7478991596638656e-05, + "loss": 0.8835, + "step": 833 + }, + { + "epoch": 0.02630499921148084, + "grad_norm": 15.294150352478027, + "learning_rate": 1.7500000000000002e-05, + "loss": 0.7013, + "step": 834 + }, + { + "epoch": 0.026336539977921465, + "grad_norm": 15.8722562789917, + "learning_rate": 1.7521008403361347e-05, + "loss": 0.711, + "step": 835 + }, + { + "epoch": 0.026368080744362088, + "grad_norm": 9.134594917297363, + "learning_rate": 1.754201680672269e-05, + "loss": 0.714, + "step": 836 + }, + { + "epoch": 0.026399621510802714, + "grad_norm": 13.051192283630371, + "learning_rate": 1.7563025210084035e-05, + "loss": 0.6775, + "step": 837 + }, + { + "epoch": 0.026431162277243337, + "grad_norm": 10.801441192626953, + "learning_rate": 1.758403361344538e-05, + "loss": 0.7125, + "step": 838 + }, + { + "epoch": 0.02646270304368396, + "grad_norm": 8.554762840270996, + "learning_rate": 1.7605042016806723e-05, + "loss": 0.7148, + "step": 839 + }, + { + "epoch": 0.026494243810124586, + "grad_norm": 11.331249237060547, + "learning_rate": 1.7626050420168068e-05, + "loss": 0.7229, + "step": 840 + }, + { + "epoch": 0.02652578457656521, + "grad_norm": 10.641855239868164, + "learning_rate": 1.7647058823529414e-05, + "loss": 0.6793, + "step": 841 + }, + { + "epoch": 0.026557325343005835, + "grad_norm": 8.861638069152832, + "learning_rate": 1.7668067226890756e-05, + "loss": 0.6963, + "step": 842 + }, + { + "epoch": 0.026588866109446458, + "grad_norm": 6.989411354064941, + "learning_rate": 1.76890756302521e-05, + "loss": 0.7057, + "step": 843 + }, + { + "epoch": 0.026620406875887084, + "grad_norm": 7.490039825439453, + "learning_rate": 1.7710084033613447e-05, + "loss": 0.6769, + "step": 844 + }, + { + "epoch": 0.026651947642327707, + "grad_norm": 23.39878273010254, + "learning_rate": 1.7731092436974792e-05, + "loss": 0.3541, + "step": 845 + }, + { + "epoch": 0.026683488408768333, + "grad_norm": 5.488243103027344, + "learning_rate": 1.7752100840336134e-05, + "loss": 0.902, + "step": 846 + }, + { + "epoch": 0.026715029175208956, + "grad_norm": 9.04245376586914, + "learning_rate": 1.777310924369748e-05, + "loss": 0.7142, + "step": 847 + }, + { + "epoch": 0.026746569941649583, + "grad_norm": 7.347799301147461, + "learning_rate": 1.7794117647058825e-05, + "loss": 0.7176, + "step": 848 + }, + { + "epoch": 0.026778110708090205, + "grad_norm": 21.2207088470459, + "learning_rate": 1.781512605042017e-05, + "loss": 0.7213, + "step": 849 + }, + { + "epoch": 0.02680965147453083, + "grad_norm": 20.266326904296875, + "learning_rate": 1.7836134453781516e-05, + "loss": 0.7045, + "step": 850 + }, + { + "epoch": 0.026841192240971454, + "grad_norm": 10.498048782348633, + "learning_rate": 1.785714285714286e-05, + "loss": 0.727, + "step": 851 + }, + { + "epoch": 0.02687273300741208, + "grad_norm": 9.912834167480469, + "learning_rate": 1.7878151260504204e-05, + "loss": 0.6899, + "step": 852 + }, + { + "epoch": 0.026904273773852704, + "grad_norm": 7.0976033210754395, + "learning_rate": 1.789915966386555e-05, + "loss": 0.3719, + "step": 853 + }, + { + "epoch": 0.02693581454029333, + "grad_norm": 10.62214469909668, + "learning_rate": 1.792016806722689e-05, + "loss": 0.7166, + "step": 854 + }, + { + "epoch": 0.026967355306733953, + "grad_norm": 9.642728805541992, + "learning_rate": 1.7941176470588237e-05, + "loss": 0.6702, + "step": 855 + }, + { + "epoch": 0.02699889607317458, + "grad_norm": 11.86683464050293, + "learning_rate": 1.7962184873949583e-05, + "loss": 0.6687, + "step": 856 + }, + { + "epoch": 0.0270304368396152, + "grad_norm": 3.098804473876953, + "learning_rate": 1.7983193277310925e-05, + "loss": 0.8945, + "step": 857 + }, + { + "epoch": 0.027061977606055828, + "grad_norm": 10.789454460144043, + "learning_rate": 1.800420168067227e-05, + "loss": 0.7357, + "step": 858 + }, + { + "epoch": 0.02709351837249645, + "grad_norm": 11.506264686584473, + "learning_rate": 1.8025210084033616e-05, + "loss": 0.692, + "step": 859 + }, + { + "epoch": 0.027125059138937077, + "grad_norm": 3.3745579719543457, + "learning_rate": 1.804621848739496e-05, + "loss": 0.9, + "step": 860 + }, + { + "epoch": 0.0271565999053777, + "grad_norm": 7.906543254852295, + "learning_rate": 1.8067226890756303e-05, + "loss": 0.7215, + "step": 861 + }, + { + "epoch": 0.027188140671818326, + "grad_norm": 9.73000717163086, + "learning_rate": 1.808823529411765e-05, + "loss": 0.7093, + "step": 862 + }, + { + "epoch": 0.02721968143825895, + "grad_norm": 9.288762092590332, + "learning_rate": 1.8109243697478994e-05, + "loss": 0.6967, + "step": 863 + }, + { + "epoch": 0.027251222204699575, + "grad_norm": 6.373456954956055, + "learning_rate": 1.8130252100840336e-05, + "loss": 0.671, + "step": 864 + }, + { + "epoch": 0.027282762971140198, + "grad_norm": 13.011707305908203, + "learning_rate": 1.8151260504201682e-05, + "loss": 0.7219, + "step": 865 + }, + { + "epoch": 0.027314303737580824, + "grad_norm": 7.479901313781738, + "learning_rate": 1.8172268907563027e-05, + "loss": 0.6895, + "step": 866 + }, + { + "epoch": 0.027345844504021447, + "grad_norm": 12.561650276184082, + "learning_rate": 1.819327731092437e-05, + "loss": 0.7477, + "step": 867 + }, + { + "epoch": 0.027377385270462073, + "grad_norm": 8.267829895019531, + "learning_rate": 1.8214285714285715e-05, + "loss": 0.7653, + "step": 868 + }, + { + "epoch": 0.027408926036902696, + "grad_norm": 10.749792098999023, + "learning_rate": 1.823529411764706e-05, + "loss": 0.7228, + "step": 869 + }, + { + "epoch": 0.027440466803343323, + "grad_norm": 11.232030868530273, + "learning_rate": 1.8256302521008403e-05, + "loss": 0.6892, + "step": 870 + }, + { + "epoch": 0.027472007569783945, + "grad_norm": 10.177556991577148, + "learning_rate": 1.8277310924369748e-05, + "loss": 0.7107, + "step": 871 + }, + { + "epoch": 0.02750354833622457, + "grad_norm": 20.678451538085938, + "learning_rate": 1.8298319327731094e-05, + "loss": 0.7196, + "step": 872 + }, + { + "epoch": 0.027535089102665194, + "grad_norm": 6.54447078704834, + "learning_rate": 1.831932773109244e-05, + "loss": 0.6714, + "step": 873 + }, + { + "epoch": 0.02756662986910582, + "grad_norm": 6.174670696258545, + "learning_rate": 1.834033613445378e-05, + "loss": 0.6857, + "step": 874 + }, + { + "epoch": 0.027598170635546444, + "grad_norm": 8.851613998413086, + "learning_rate": 1.8361344537815127e-05, + "loss": 0.6862, + "step": 875 + }, + { + "epoch": 0.02762971140198707, + "grad_norm": 7.055481433868408, + "learning_rate": 1.8382352941176472e-05, + "loss": 0.7251, + "step": 876 + }, + { + "epoch": 0.027661252168427693, + "grad_norm": 7.4818878173828125, + "learning_rate": 1.8403361344537814e-05, + "loss": 0.6848, + "step": 877 + }, + { + "epoch": 0.02769279293486832, + "grad_norm": 6.7728753089904785, + "learning_rate": 1.842436974789916e-05, + "loss": 0.7183, + "step": 878 + }, + { + "epoch": 0.027724333701308942, + "grad_norm": 8.586965560913086, + "learning_rate": 1.8445378151260505e-05, + "loss": 0.7298, + "step": 879 + }, + { + "epoch": 0.027755874467749565, + "grad_norm": 15.179737091064453, + "learning_rate": 1.846638655462185e-05, + "loss": 0.7216, + "step": 880 + }, + { + "epoch": 0.02778741523419019, + "grad_norm": 12.111358642578125, + "learning_rate": 1.8487394957983196e-05, + "loss": 0.7396, + "step": 881 + }, + { + "epoch": 0.027818956000630814, + "grad_norm": 7.345612049102783, + "learning_rate": 1.850840336134454e-05, + "loss": 0.7139, + "step": 882 + }, + { + "epoch": 0.02785049676707144, + "grad_norm": 7.915726661682129, + "learning_rate": 1.8529411764705884e-05, + "loss": 0.728, + "step": 883 + }, + { + "epoch": 0.027882037533512063, + "grad_norm": 6.2250075340271, + "learning_rate": 1.855042016806723e-05, + "loss": 0.7114, + "step": 884 + }, + { + "epoch": 0.02791357829995269, + "grad_norm": 7.045414447784424, + "learning_rate": 1.8571428571428575e-05, + "loss": 0.715, + "step": 885 + }, + { + "epoch": 0.027945119066393312, + "grad_norm": 7.855751037597656, + "learning_rate": 1.8592436974789917e-05, + "loss": 0.7026, + "step": 886 + }, + { + "epoch": 0.027976659832833938, + "grad_norm": 7.734004497528076, + "learning_rate": 1.8613445378151263e-05, + "loss": 0.6612, + "step": 887 + }, + { + "epoch": 0.02800820059927456, + "grad_norm": 5.840481281280518, + "learning_rate": 1.8634453781512608e-05, + "loss": 0.4044, + "step": 888 + }, + { + "epoch": 0.028039741365715187, + "grad_norm": 16.29491424560547, + "learning_rate": 1.865546218487395e-05, + "loss": 0.7007, + "step": 889 + }, + { + "epoch": 0.02807128213215581, + "grad_norm": 9.430410385131836, + "learning_rate": 1.8676470588235296e-05, + "loss": 0.6932, + "step": 890 + }, + { + "epoch": 0.028102822898596436, + "grad_norm": 3.564699172973633, + "learning_rate": 1.869747899159664e-05, + "loss": 0.3909, + "step": 891 + }, + { + "epoch": 0.02813436366503706, + "grad_norm": 11.734607696533203, + "learning_rate": 1.8718487394957983e-05, + "loss": 0.7021, + "step": 892 + }, + { + "epoch": 0.028165904431477685, + "grad_norm": 13.915765762329102, + "learning_rate": 1.873949579831933e-05, + "loss": 0.6893, + "step": 893 + }, + { + "epoch": 0.02819744519791831, + "grad_norm": 7.234941482543945, + "learning_rate": 1.8760504201680674e-05, + "loss": 0.7077, + "step": 894 + }, + { + "epoch": 0.028228985964358935, + "grad_norm": 19.253259658813477, + "learning_rate": 1.8781512605042017e-05, + "loss": 0.7457, + "step": 895 + }, + { + "epoch": 0.028260526730799557, + "grad_norm": 6.186545372009277, + "learning_rate": 1.8802521008403362e-05, + "loss": 0.7111, + "step": 896 + }, + { + "epoch": 0.028292067497240184, + "grad_norm": 6.980288028717041, + "learning_rate": 1.8823529411764708e-05, + "loss": 0.6754, + "step": 897 + }, + { + "epoch": 0.028323608263680806, + "grad_norm": 12.99902057647705, + "learning_rate": 1.8844537815126053e-05, + "loss": 0.6711, + "step": 898 + }, + { + "epoch": 0.028355149030121433, + "grad_norm": 7.9728779792785645, + "learning_rate": 1.8865546218487395e-05, + "loss": 0.6991, + "step": 899 + }, + { + "epoch": 0.028386689796562056, + "grad_norm": 4.442030906677246, + "learning_rate": 1.888655462184874e-05, + "loss": 0.8921, + "step": 900 + }, + { + "epoch": 0.028418230563002682, + "grad_norm": 9.718960762023926, + "learning_rate": 1.8907563025210086e-05, + "loss": 0.7091, + "step": 901 + }, + { + "epoch": 0.028449771329443305, + "grad_norm": 9.95651912689209, + "learning_rate": 1.892857142857143e-05, + "loss": 0.6936, + "step": 902 + }, + { + "epoch": 0.02848131209588393, + "grad_norm": 8.299859046936035, + "learning_rate": 1.8949579831932774e-05, + "loss": 0.714, + "step": 903 + }, + { + "epoch": 0.028512852862324554, + "grad_norm": 13.619277000427246, + "learning_rate": 1.897058823529412e-05, + "loss": 0.7669, + "step": 904 + }, + { + "epoch": 0.02854439362876518, + "grad_norm": 6.935705661773682, + "learning_rate": 1.899159663865546e-05, + "loss": 0.6698, + "step": 905 + }, + { + "epoch": 0.028575934395205803, + "grad_norm": 8.68918514251709, + "learning_rate": 1.9012605042016807e-05, + "loss": 0.6859, + "step": 906 + }, + { + "epoch": 0.02860747516164643, + "grad_norm": 5.92099666595459, + "learning_rate": 1.9033613445378152e-05, + "loss": 0.6746, + "step": 907 + }, + { + "epoch": 0.028639015928087052, + "grad_norm": 10.29526138305664, + "learning_rate": 1.9054621848739495e-05, + "loss": 0.6892, + "step": 908 + }, + { + "epoch": 0.028670556694527678, + "grad_norm": 9.077865600585938, + "learning_rate": 1.9075630252100844e-05, + "loss": 0.7055, + "step": 909 + }, + { + "epoch": 0.0287020974609683, + "grad_norm": 3.164276123046875, + "learning_rate": 1.909663865546219e-05, + "loss": 0.9514, + "step": 910 + }, + { + "epoch": 0.028733638227408927, + "grad_norm": 6.2777419090271, + "learning_rate": 1.911764705882353e-05, + "loss": 0.6973, + "step": 911 + }, + { + "epoch": 0.02876517899384955, + "grad_norm": 7.724456787109375, + "learning_rate": 1.9138655462184877e-05, + "loss": 0.6699, + "step": 912 + }, + { + "epoch": 0.028796719760290176, + "grad_norm": 7.009775638580322, + "learning_rate": 1.9159663865546222e-05, + "loss": 0.6724, + "step": 913 + }, + { + "epoch": 0.0288282605267308, + "grad_norm": 3.1324124336242676, + "learning_rate": 1.9180672268907564e-05, + "loss": 0.8846, + "step": 914 + }, + { + "epoch": 0.028859801293171426, + "grad_norm": 6.902660369873047, + "learning_rate": 1.920168067226891e-05, + "loss": 0.7218, + "step": 915 + }, + { + "epoch": 0.02889134205961205, + "grad_norm": 7.354856014251709, + "learning_rate": 1.9222689075630255e-05, + "loss": 0.8944, + "step": 916 + }, + { + "epoch": 0.028922882826052675, + "grad_norm": 9.219684600830078, + "learning_rate": 1.9243697478991597e-05, + "loss": 0.6737, + "step": 917 + }, + { + "epoch": 0.028954423592493297, + "grad_norm": 18.373525619506836, + "learning_rate": 1.9264705882352943e-05, + "loss": 0.7037, + "step": 918 + }, + { + "epoch": 0.028985964358933924, + "grad_norm": 8.030267715454102, + "learning_rate": 1.928571428571429e-05, + "loss": 0.7186, + "step": 919 + }, + { + "epoch": 0.029017505125374547, + "grad_norm": 10.757308006286621, + "learning_rate": 1.930672268907563e-05, + "loss": 0.6855, + "step": 920 + }, + { + "epoch": 0.029049045891815173, + "grad_norm": 8.4976224899292, + "learning_rate": 1.9327731092436976e-05, + "loss": 0.7052, + "step": 921 + }, + { + "epoch": 0.029080586658255796, + "grad_norm": 9.719372749328613, + "learning_rate": 1.934873949579832e-05, + "loss": 0.7276, + "step": 922 + }, + { + "epoch": 0.02911212742469642, + "grad_norm": 8.078779220581055, + "learning_rate": 1.9369747899159667e-05, + "loss": 0.7071, + "step": 923 + }, + { + "epoch": 0.029143668191137045, + "grad_norm": 8.561447143554688, + "learning_rate": 1.939075630252101e-05, + "loss": 0.6927, + "step": 924 + }, + { + "epoch": 0.029175208957577668, + "grad_norm": 8.581090927124023, + "learning_rate": 1.9411764705882355e-05, + "loss": 0.6873, + "step": 925 + }, + { + "epoch": 0.029206749724018294, + "grad_norm": 8.13828182220459, + "learning_rate": 1.94327731092437e-05, + "loss": 0.7298, + "step": 926 + }, + { + "epoch": 0.029238290490458917, + "grad_norm": 6.537692546844482, + "learning_rate": 1.9453781512605042e-05, + "loss": 0.6494, + "step": 927 + }, + { + "epoch": 0.029269831256899543, + "grad_norm": 9.855411529541016, + "learning_rate": 1.9474789915966388e-05, + "loss": 0.6699, + "step": 928 + }, + { + "epoch": 0.029301372023340166, + "grad_norm": 7.414729118347168, + "learning_rate": 1.9495798319327733e-05, + "loss": 0.4023, + "step": 929 + }, + { + "epoch": 0.029332912789780792, + "grad_norm": 8.96755599975586, + "learning_rate": 1.9516806722689075e-05, + "loss": 0.6974, + "step": 930 + }, + { + "epoch": 0.029364453556221415, + "grad_norm": 7.198520183563232, + "learning_rate": 1.953781512605042e-05, + "loss": 0.7316, + "step": 931 + }, + { + "epoch": 0.02939599432266204, + "grad_norm": 9.01680850982666, + "learning_rate": 1.9558823529411766e-05, + "loss": 0.6837, + "step": 932 + }, + { + "epoch": 0.029427535089102664, + "grad_norm": 8.955812454223633, + "learning_rate": 1.957983193277311e-05, + "loss": 0.7029, + "step": 933 + }, + { + "epoch": 0.02945907585554329, + "grad_norm": 6.678302764892578, + "learning_rate": 1.9600840336134454e-05, + "loss": 0.718, + "step": 934 + }, + { + "epoch": 0.029490616621983913, + "grad_norm": 5.312445640563965, + "learning_rate": 1.96218487394958e-05, + "loss": 0.3822, + "step": 935 + }, + { + "epoch": 0.02952215738842454, + "grad_norm": 7.625219345092773, + "learning_rate": 1.9642857142857145e-05, + "loss": 0.7002, + "step": 936 + }, + { + "epoch": 0.029553698154865162, + "grad_norm": 7.269476890563965, + "learning_rate": 1.9663865546218487e-05, + "loss": 0.9137, + "step": 937 + }, + { + "epoch": 0.02958523892130579, + "grad_norm": 10.954496383666992, + "learning_rate": 1.9684873949579833e-05, + "loss": 0.7099, + "step": 938 + }, + { + "epoch": 0.02961677968774641, + "grad_norm": 6.211147785186768, + "learning_rate": 1.9705882352941178e-05, + "loss": 0.6959, + "step": 939 + }, + { + "epoch": 0.029648320454187038, + "grad_norm": 20.503786087036133, + "learning_rate": 1.9726890756302524e-05, + "loss": 0.6938, + "step": 940 + }, + { + "epoch": 0.02967986122062766, + "grad_norm": 10.232073783874512, + "learning_rate": 1.974789915966387e-05, + "loss": 0.8919, + "step": 941 + }, + { + "epoch": 0.029711401987068287, + "grad_norm": 3.6164703369140625, + "learning_rate": 1.976890756302521e-05, + "loss": 0.3765, + "step": 942 + }, + { + "epoch": 0.02974294275350891, + "grad_norm": 18.636085510253906, + "learning_rate": 1.9789915966386557e-05, + "loss": 0.7303, + "step": 943 + }, + { + "epoch": 0.029774483519949536, + "grad_norm": 5.842304229736328, + "learning_rate": 1.9810924369747902e-05, + "loss": 0.8864, + "step": 944 + }, + { + "epoch": 0.02980602428639016, + "grad_norm": 10.755277633666992, + "learning_rate": 1.9831932773109244e-05, + "loss": 0.6827, + "step": 945 + }, + { + "epoch": 0.029837565052830785, + "grad_norm": 7.158202648162842, + "learning_rate": 1.985294117647059e-05, + "loss": 0.6872, + "step": 946 + }, + { + "epoch": 0.029869105819271408, + "grad_norm": 8.313507080078125, + "learning_rate": 1.9873949579831935e-05, + "loss": 0.7112, + "step": 947 + }, + { + "epoch": 0.029900646585712034, + "grad_norm": 12.288867950439453, + "learning_rate": 1.989495798319328e-05, + "loss": 0.7023, + "step": 948 + }, + { + "epoch": 0.029932187352152657, + "grad_norm": 6.446424961090088, + "learning_rate": 1.9915966386554623e-05, + "loss": 0.7272, + "step": 949 + }, + { + "epoch": 0.029963728118593283, + "grad_norm": 8.33220386505127, + "learning_rate": 1.993697478991597e-05, + "loss": 0.7041, + "step": 950 + }, + { + "epoch": 0.029995268885033906, + "grad_norm": 9.129294395446777, + "learning_rate": 1.9957983193277314e-05, + "loss": 0.7033, + "step": 951 + }, + { + "epoch": 0.030026809651474532, + "grad_norm": 2.6840531826019287, + "learning_rate": 1.9978991596638656e-05, + "loss": 0.3699, + "step": 952 + }, + { + "epoch": 0.030058350417915155, + "grad_norm": 6.923593521118164, + "learning_rate": 2e-05, + "loss": 0.6932, + "step": 953 + }, + { + "epoch": 0.03008989118435578, + "grad_norm": 3.8326914310455322, + "learning_rate": 1.999999994782112e-05, + "loss": 0.9112, + "step": 954 + }, + { + "epoch": 0.030121431950796404, + "grad_norm": 6.26516056060791, + "learning_rate": 1.9999999791284473e-05, + "loss": 0.6621, + "step": 955 + }, + { + "epoch": 0.03015297271723703, + "grad_norm": 4.354166507720947, + "learning_rate": 1.9999999530390065e-05, + "loss": 0.7005, + "step": 956 + }, + { + "epoch": 0.030184513483677653, + "grad_norm": 4.51849889755249, + "learning_rate": 1.9999999165137895e-05, + "loss": 0.9313, + "step": 957 + }, + { + "epoch": 0.03021605425011828, + "grad_norm": 8.597533226013184, + "learning_rate": 1.999999869552797e-05, + "loss": 0.7036, + "step": 958 + }, + { + "epoch": 0.030247595016558902, + "grad_norm": 3.2627272605895996, + "learning_rate": 1.999999812156029e-05, + "loss": 0.905, + "step": 959 + }, + { + "epoch": 0.03027913578299953, + "grad_norm": 8.515935897827148, + "learning_rate": 1.9999997443234872e-05, + "loss": 0.7308, + "step": 960 + }, + { + "epoch": 0.03031067654944015, + "grad_norm": 5.353451728820801, + "learning_rate": 1.999999666055171e-05, + "loss": 0.6885, + "step": 961 + }, + { + "epoch": 0.030342217315880778, + "grad_norm": 4.053324222564697, + "learning_rate": 1.999999577351082e-05, + "loss": 0.4054, + "step": 962 + }, + { + "epoch": 0.0303737580823214, + "grad_norm": 4.905959606170654, + "learning_rate": 1.999999478211221e-05, + "loss": 0.402, + "step": 963 + }, + { + "epoch": 0.030405298848762023, + "grad_norm": 7.1532111167907715, + "learning_rate": 1.999999368635589e-05, + "loss": 0.6922, + "step": 964 + }, + { + "epoch": 0.03043683961520265, + "grad_norm": 9.333877563476562, + "learning_rate": 1.9999992486241872e-05, + "loss": 0.7015, + "step": 965 + }, + { + "epoch": 0.030468380381643272, + "grad_norm": 7.770199298858643, + "learning_rate": 1.9999991181770164e-05, + "loss": 0.7165, + "step": 966 + }, + { + "epoch": 0.0304999211480839, + "grad_norm": 12.050905227661133, + "learning_rate": 1.9999989772940785e-05, + "loss": 0.7255, + "step": 967 + }, + { + "epoch": 0.03053146191452452, + "grad_norm": 6.776571750640869, + "learning_rate": 1.9999988259753748e-05, + "loss": 0.7121, + "step": 968 + }, + { + "epoch": 0.030563002680965148, + "grad_norm": 7.504469394683838, + "learning_rate": 1.999998664220907e-05, + "loss": 0.6788, + "step": 969 + }, + { + "epoch": 0.03059454344740577, + "grad_norm": 7.475898742675781, + "learning_rate": 1.9999984920306767e-05, + "loss": 0.6938, + "step": 970 + }, + { + "epoch": 0.030626084213846397, + "grad_norm": 25.799158096313477, + "learning_rate": 1.9999983094046853e-05, + "loss": 0.6681, + "step": 971 + }, + { + "epoch": 0.03065762498028702, + "grad_norm": 7.134160995483398, + "learning_rate": 1.9999981163429352e-05, + "loss": 0.7082, + "step": 972 + }, + { + "epoch": 0.030689165746727646, + "grad_norm": 5.928959846496582, + "learning_rate": 1.9999979128454282e-05, + "loss": 0.6798, + "step": 973 + }, + { + "epoch": 0.03072070651316827, + "grad_norm": 6.735225677490234, + "learning_rate": 1.999997698912167e-05, + "loss": 0.6886, + "step": 974 + }, + { + "epoch": 0.030752247279608895, + "grad_norm": 11.305253982543945, + "learning_rate": 1.999997474543153e-05, + "loss": 0.7153, + "step": 975 + }, + { + "epoch": 0.030783788046049518, + "grad_norm": 6.582517623901367, + "learning_rate": 1.9999972397383886e-05, + "loss": 0.6755, + "step": 976 + }, + { + "epoch": 0.030815328812490144, + "grad_norm": 5.966078758239746, + "learning_rate": 1.999996994497877e-05, + "loss": 0.6906, + "step": 977 + }, + { + "epoch": 0.030846869578930767, + "grad_norm": 9.758932113647461, + "learning_rate": 1.99999673882162e-05, + "loss": 0.7477, + "step": 978 + }, + { + "epoch": 0.030878410345371393, + "grad_norm": 6.769775390625, + "learning_rate": 1.9999964727096202e-05, + "loss": 0.6832, + "step": 979 + }, + { + "epoch": 0.030909951111812016, + "grad_norm": 6.024561882019043, + "learning_rate": 1.9999961961618814e-05, + "loss": 0.6736, + "step": 980 + }, + { + "epoch": 0.030941491878252642, + "grad_norm": 6.915926933288574, + "learning_rate": 1.9999959091784056e-05, + "loss": 0.681, + "step": 981 + }, + { + "epoch": 0.030973032644693265, + "grad_norm": 68.07229614257812, + "learning_rate": 1.9999956117591962e-05, + "loss": 0.4226, + "step": 982 + }, + { + "epoch": 0.03100457341113389, + "grad_norm": 8.003413200378418, + "learning_rate": 1.9999953039042557e-05, + "loss": 0.6696, + "step": 983 + }, + { + "epoch": 0.031036114177574514, + "grad_norm": 6.990445137023926, + "learning_rate": 1.9999949856135877e-05, + "loss": 0.9269, + "step": 984 + }, + { + "epoch": 0.03106765494401514, + "grad_norm": 5.386775016784668, + "learning_rate": 1.999994656887196e-05, + "loss": 0.7091, + "step": 985 + }, + { + "epoch": 0.031099195710455763, + "grad_norm": 8.052472114562988, + "learning_rate": 1.9999943177250832e-05, + "loss": 0.6693, + "step": 986 + }, + { + "epoch": 0.03113073647689639, + "grad_norm": 7.255199432373047, + "learning_rate": 1.9999939681272535e-05, + "loss": 0.7197, + "step": 987 + }, + { + "epoch": 0.031162277243337012, + "grad_norm": 6.442477226257324, + "learning_rate": 1.99999360809371e-05, + "loss": 0.6883, + "step": 988 + }, + { + "epoch": 0.03119381800977764, + "grad_norm": 10.641590118408203, + "learning_rate": 1.999993237624457e-05, + "loss": 0.6822, + "step": 989 + }, + { + "epoch": 0.03122535877621826, + "grad_norm": 16.306015014648438, + "learning_rate": 1.999992856719498e-05, + "loss": 0.7041, + "step": 990 + }, + { + "epoch": 0.03125689954265889, + "grad_norm": 8.346558570861816, + "learning_rate": 1.9999924653788368e-05, + "loss": 0.6818, + "step": 991 + }, + { + "epoch": 0.03128844030909951, + "grad_norm": 7.541151523590088, + "learning_rate": 1.9999920636024778e-05, + "loss": 0.6737, + "step": 992 + }, + { + "epoch": 0.03131998107554013, + "grad_norm": 6.318100929260254, + "learning_rate": 1.9999916513904252e-05, + "loss": 0.6835, + "step": 993 + }, + { + "epoch": 0.03135152184198076, + "grad_norm": 9.41586685180664, + "learning_rate": 1.999991228742684e-05, + "loss": 0.6971, + "step": 994 + }, + { + "epoch": 0.031383062608421386, + "grad_norm": 6.541496276855469, + "learning_rate": 1.9999907956592567e-05, + "loss": 0.6663, + "step": 995 + }, + { + "epoch": 0.03141460337486201, + "grad_norm": 5.154834747314453, + "learning_rate": 1.9999903521401495e-05, + "loss": 0.6853, + "step": 996 + }, + { + "epoch": 0.03144614414130263, + "grad_norm": 9.25920581817627, + "learning_rate": 1.9999898981853666e-05, + "loss": 0.6583, + "step": 997 + }, + { + "epoch": 0.03147768490774326, + "grad_norm": 7.806034088134766, + "learning_rate": 1.9999894337949122e-05, + "loss": 0.7113, + "step": 998 + }, + { + "epoch": 0.031509225674183884, + "grad_norm": 5.291286945343018, + "learning_rate": 1.999988958968792e-05, + "loss": 0.6991, + "step": 999 + }, + { + "epoch": 0.03154076644062451, + "grad_norm": 9.547070503234863, + "learning_rate": 1.9999884737070108e-05, + "loss": 0.726, + "step": 1000 + }, + { + "epoch": 0.03157230720706513, + "grad_norm": 10.82068920135498, + "learning_rate": 1.999987978009573e-05, + "loss": 0.6593, + "step": 1001 + }, + { + "epoch": 0.03160384797350576, + "grad_norm": 8.005331993103027, + "learning_rate": 1.999987471876484e-05, + "loss": 0.661, + "step": 1002 + }, + { + "epoch": 0.03163538873994638, + "grad_norm": 5.858728408813477, + "learning_rate": 1.9999869553077498e-05, + "loss": 0.6963, + "step": 1003 + }, + { + "epoch": 0.031666929506387005, + "grad_norm": 5.149726390838623, + "learning_rate": 1.9999864283033748e-05, + "loss": 0.682, + "step": 1004 + }, + { + "epoch": 0.03169847027282763, + "grad_norm": 6.511922836303711, + "learning_rate": 1.9999858908633652e-05, + "loss": 0.3915, + "step": 1005 + }, + { + "epoch": 0.03173001103926825, + "grad_norm": 5.945653438568115, + "learning_rate": 1.999985342987726e-05, + "loss": 0.684, + "step": 1006 + }, + { + "epoch": 0.03176155180570888, + "grad_norm": 9.605725288391113, + "learning_rate": 1.9999847846764634e-05, + "loss": 0.6932, + "step": 1007 + }, + { + "epoch": 0.0317930925721495, + "grad_norm": 6.737133979797363, + "learning_rate": 1.9999842159295834e-05, + "loss": 0.6736, + "step": 1008 + }, + { + "epoch": 0.031824633338590126, + "grad_norm": 5.935062408447266, + "learning_rate": 1.9999836367470915e-05, + "loss": 0.9213, + "step": 1009 + }, + { + "epoch": 0.03185617410503075, + "grad_norm": 3.3410861492156982, + "learning_rate": 1.999983047128994e-05, + "loss": 0.9398, + "step": 1010 + }, + { + "epoch": 0.03188771487147138, + "grad_norm": 9.181023597717285, + "learning_rate": 1.9999824470752967e-05, + "loss": 0.7092, + "step": 1011 + }, + { + "epoch": 0.031919255637912, + "grad_norm": 6.6087799072265625, + "learning_rate": 1.9999818365860063e-05, + "loss": 0.6778, + "step": 1012 + }, + { + "epoch": 0.031950796404352624, + "grad_norm": 5.418369770050049, + "learning_rate": 1.9999812156611287e-05, + "loss": 0.6569, + "step": 1013 + }, + { + "epoch": 0.03198233717079325, + "grad_norm": 5.988561630249023, + "learning_rate": 1.999980584300671e-05, + "loss": 0.3907, + "step": 1014 + }, + { + "epoch": 0.03201387793723388, + "grad_norm": 6.76757287979126, + "learning_rate": 1.999979942504639e-05, + "loss": 0.7288, + "step": 1015 + }, + { + "epoch": 0.0320454187036745, + "grad_norm": 7.163880825042725, + "learning_rate": 1.99997929027304e-05, + "loss": 0.6828, + "step": 1016 + }, + { + "epoch": 0.03207695947011512, + "grad_norm": 6.171513080596924, + "learning_rate": 1.9999786276058808e-05, + "loss": 0.7351, + "step": 1017 + }, + { + "epoch": 0.032108500236555745, + "grad_norm": 4.900684833526611, + "learning_rate": 1.9999779545031682e-05, + "loss": 0.9119, + "step": 1018 + }, + { + "epoch": 0.032140041002996375, + "grad_norm": 5.127006530761719, + "learning_rate": 1.9999772709649094e-05, + "loss": 0.7026, + "step": 1019 + }, + { + "epoch": 0.032171581769437, + "grad_norm": 5.656378269195557, + "learning_rate": 1.9999765769911108e-05, + "loss": 0.7034, + "step": 1020 + }, + { + "epoch": 0.03220312253587762, + "grad_norm": 7.451080799102783, + "learning_rate": 1.9999758725817802e-05, + "loss": 0.6915, + "step": 1021 + }, + { + "epoch": 0.032234663302318244, + "grad_norm": 5.873846530914307, + "learning_rate": 1.999975157736925e-05, + "loss": 0.6818, + "step": 1022 + }, + { + "epoch": 0.03226620406875887, + "grad_norm": 5.308951377868652, + "learning_rate": 1.9999744324565527e-05, + "loss": 0.4268, + "step": 1023 + }, + { + "epoch": 0.032297744835199496, + "grad_norm": 5.527895927429199, + "learning_rate": 1.9999736967406706e-05, + "loss": 0.6799, + "step": 1024 + }, + { + "epoch": 0.03232928560164012, + "grad_norm": 3.285550355911255, + "learning_rate": 1.9999729505892867e-05, + "loss": 0.3589, + "step": 1025 + }, + { + "epoch": 0.03236082636808074, + "grad_norm": 4.744204044342041, + "learning_rate": 1.9999721940024085e-05, + "loss": 0.6887, + "step": 1026 + }, + { + "epoch": 0.03239236713452137, + "grad_norm": 2.799468994140625, + "learning_rate": 1.999971426980044e-05, + "loss": 0.3894, + "step": 1027 + }, + { + "epoch": 0.032423907900961994, + "grad_norm": 5.5059075355529785, + "learning_rate": 1.9999706495222008e-05, + "loss": 0.6969, + "step": 1028 + }, + { + "epoch": 0.03245544866740262, + "grad_norm": 5.473618984222412, + "learning_rate": 1.999969861628888e-05, + "loss": 0.6857, + "step": 1029 + }, + { + "epoch": 0.03248698943384324, + "grad_norm": 5.40762996673584, + "learning_rate": 1.999969063300113e-05, + "loss": 0.6966, + "step": 1030 + }, + { + "epoch": 0.03251853020028387, + "grad_norm": 5.706680774688721, + "learning_rate": 1.9999682545358844e-05, + "loss": 0.6652, + "step": 1031 + }, + { + "epoch": 0.03255007096672449, + "grad_norm": 6.3381123542785645, + "learning_rate": 1.9999674353362105e-05, + "loss": 0.6852, + "step": 1032 + }, + { + "epoch": 0.032581611733165115, + "grad_norm": 7.109625339508057, + "learning_rate": 1.9999666057011004e-05, + "loss": 0.705, + "step": 1033 + }, + { + "epoch": 0.03261315249960574, + "grad_norm": 5.300387382507324, + "learning_rate": 1.9999657656305617e-05, + "loss": 0.6742, + "step": 1034 + }, + { + "epoch": 0.03264469326604637, + "grad_norm": 4.08128023147583, + "learning_rate": 1.9999649151246042e-05, + "loss": 0.6946, + "step": 1035 + }, + { + "epoch": 0.03267623403248699, + "grad_norm": 4.717286109924316, + "learning_rate": 1.999964054183236e-05, + "loss": 0.698, + "step": 1036 + }, + { + "epoch": 0.032707774798927614, + "grad_norm": 6.716667652130127, + "learning_rate": 1.9999631828064666e-05, + "loss": 0.6868, + "step": 1037 + }, + { + "epoch": 0.032739315565368236, + "grad_norm": 5.5210347175598145, + "learning_rate": 1.9999623009943052e-05, + "loss": 0.6676, + "step": 1038 + }, + { + "epoch": 0.032770856331808866, + "grad_norm": 11.161954879760742, + "learning_rate": 1.9999614087467606e-05, + "loss": 0.9127, + "step": 1039 + }, + { + "epoch": 0.03280239709824949, + "grad_norm": 6.58017635345459, + "learning_rate": 1.9999605060638422e-05, + "loss": 0.707, + "step": 1040 + }, + { + "epoch": 0.03283393786469011, + "grad_norm": 61.63962173461914, + "learning_rate": 1.9999595929455593e-05, + "loss": 0.7137, + "step": 1041 + }, + { + "epoch": 0.032865478631130735, + "grad_norm": 6.915230751037598, + "learning_rate": 1.999958669391922e-05, + "loss": 0.6506, + "step": 1042 + }, + { + "epoch": 0.032897019397571364, + "grad_norm": 4.6773834228515625, + "learning_rate": 1.999957735402939e-05, + "loss": 0.6841, + "step": 1043 + }, + { + "epoch": 0.03292856016401199, + "grad_norm": 5.3214802742004395, + "learning_rate": 1.9999567909786207e-05, + "loss": 0.685, + "step": 1044 + }, + { + "epoch": 0.03296010093045261, + "grad_norm": 5.388648986816406, + "learning_rate": 1.9999558361189773e-05, + "loss": 0.6899, + "step": 1045 + }, + { + "epoch": 0.03299164169689323, + "grad_norm": 9.954300880432129, + "learning_rate": 1.9999548708240177e-05, + "loss": 0.9285, + "step": 1046 + }, + { + "epoch": 0.03302318246333386, + "grad_norm": 5.252139091491699, + "learning_rate": 1.999953895093753e-05, + "loss": 0.4106, + "step": 1047 + }, + { + "epoch": 0.033054723229774485, + "grad_norm": 9.043780326843262, + "learning_rate": 1.9999529089281928e-05, + "loss": 0.8978, + "step": 1048 + }, + { + "epoch": 0.03308626399621511, + "grad_norm": 3.783081531524658, + "learning_rate": 1.9999519123273478e-05, + "loss": 0.4033, + "step": 1049 + }, + { + "epoch": 0.03311780476265573, + "grad_norm": 5.031665325164795, + "learning_rate": 1.999950905291228e-05, + "loss": 0.7014, + "step": 1050 + }, + { + "epoch": 0.033149345529096354, + "grad_norm": 4.257906913757324, + "learning_rate": 1.999949887819844e-05, + "loss": 0.6802, + "step": 1051 + }, + { + "epoch": 0.033180886295536983, + "grad_norm": 9.02373218536377, + "learning_rate": 1.9999488599132063e-05, + "loss": 0.6936, + "step": 1052 + }, + { + "epoch": 0.033212427061977606, + "grad_norm": 5.141037464141846, + "learning_rate": 1.999947821571326e-05, + "loss": 0.6917, + "step": 1053 + }, + { + "epoch": 0.03324396782841823, + "grad_norm": 19.861230850219727, + "learning_rate": 1.9999467727942138e-05, + "loss": 0.8864, + "step": 1054 + }, + { + "epoch": 0.03327550859485885, + "grad_norm": 15.362019538879395, + "learning_rate": 1.999945713581881e-05, + "loss": 0.6858, + "step": 1055 + }, + { + "epoch": 0.03330704936129948, + "grad_norm": 7.230454444885254, + "learning_rate": 1.9999446439343372e-05, + "loss": 0.6688, + "step": 1056 + }, + { + "epoch": 0.033338590127740104, + "grad_norm": 5.557793617248535, + "learning_rate": 1.9999435638515953e-05, + "loss": 0.6659, + "step": 1057 + }, + { + "epoch": 0.03337013089418073, + "grad_norm": 5.285873889923096, + "learning_rate": 1.9999424733336656e-05, + "loss": 0.6815, + "step": 1058 + }, + { + "epoch": 0.03340167166062135, + "grad_norm": 6.7688069343566895, + "learning_rate": 1.99994137238056e-05, + "loss": 0.6645, + "step": 1059 + }, + { + "epoch": 0.03343321242706198, + "grad_norm": 6.476375102996826, + "learning_rate": 1.9999402609922898e-05, + "loss": 0.6903, + "step": 1060 + }, + { + "epoch": 0.0334647531935026, + "grad_norm": 5.395592212677002, + "learning_rate": 1.9999391391688664e-05, + "loss": 0.6752, + "step": 1061 + }, + { + "epoch": 0.033496293959943226, + "grad_norm": 8.328357696533203, + "learning_rate": 1.9999380069103017e-05, + "loss": 0.7029, + "step": 1062 + }, + { + "epoch": 0.03352783472638385, + "grad_norm": 8.448058128356934, + "learning_rate": 1.999936864216607e-05, + "loss": 0.9165, + "step": 1063 + }, + { + "epoch": 0.03355937549282448, + "grad_norm": 4.8428826332092285, + "learning_rate": 1.9999357110877954e-05, + "loss": 0.6662, + "step": 1064 + }, + { + "epoch": 0.0335909162592651, + "grad_norm": 7.893829345703125, + "learning_rate": 1.9999345475238782e-05, + "loss": 0.4382, + "step": 1065 + }, + { + "epoch": 0.033622457025705724, + "grad_norm": 3.3617167472839355, + "learning_rate": 1.9999333735248668e-05, + "loss": 0.9286, + "step": 1066 + }, + { + "epoch": 0.033653997792146347, + "grad_norm": 7.344726085662842, + "learning_rate": 1.999932189090775e-05, + "loss": 0.694, + "step": 1067 + }, + { + "epoch": 0.033685538558586976, + "grad_norm": 6.680184841156006, + "learning_rate": 1.999930994221614e-05, + "loss": 0.6987, + "step": 1068 + }, + { + "epoch": 0.0337170793250276, + "grad_norm": 5.221799373626709, + "learning_rate": 1.9999297889173967e-05, + "loss": 0.6366, + "step": 1069 + }, + { + "epoch": 0.03374862009146822, + "grad_norm": 5.072868347167969, + "learning_rate": 1.9999285731781358e-05, + "loss": 0.66, + "step": 1070 + }, + { + "epoch": 0.033780160857908845, + "grad_norm": 6.739128589630127, + "learning_rate": 1.9999273470038436e-05, + "loss": 0.6823, + "step": 1071 + }, + { + "epoch": 0.033811701624349474, + "grad_norm": 5.113446235656738, + "learning_rate": 1.9999261103945332e-05, + "loss": 0.6742, + "step": 1072 + }, + { + "epoch": 0.0338432423907901, + "grad_norm": 3.1313321590423584, + "learning_rate": 1.9999248633502173e-05, + "loss": 0.3987, + "step": 1073 + }, + { + "epoch": 0.03387478315723072, + "grad_norm": 6.992902755737305, + "learning_rate": 1.9999236058709092e-05, + "loss": 0.6969, + "step": 1074 + }, + { + "epoch": 0.03390632392367134, + "grad_norm": 3.0482397079467773, + "learning_rate": 1.9999223379566217e-05, + "loss": 0.8836, + "step": 1075 + }, + { + "epoch": 0.03393786469011197, + "grad_norm": 5.285612106323242, + "learning_rate": 1.9999210596073684e-05, + "loss": 0.6485, + "step": 1076 + }, + { + "epoch": 0.033969405456552595, + "grad_norm": 4.36969518661499, + "learning_rate": 1.999919770823162e-05, + "loss": 0.7187, + "step": 1077 + }, + { + "epoch": 0.03400094622299322, + "grad_norm": 8.42190170288086, + "learning_rate": 1.999918471604017e-05, + "loss": 0.6861, + "step": 1078 + }, + { + "epoch": 0.03403248698943384, + "grad_norm": 4.534999847412109, + "learning_rate": 1.999917161949946e-05, + "loss": 0.6545, + "step": 1079 + }, + { + "epoch": 0.03406402775587447, + "grad_norm": 5.476503849029541, + "learning_rate": 1.999915841860963e-05, + "loss": 0.6918, + "step": 1080 + }, + { + "epoch": 0.034095568522315094, + "grad_norm": 5.00461483001709, + "learning_rate": 1.9999145113370816e-05, + "loss": 0.6978, + "step": 1081 + }, + { + "epoch": 0.034127109288755716, + "grad_norm": 5.581603527069092, + "learning_rate": 1.9999131703783163e-05, + "loss": 0.7056, + "step": 1082 + }, + { + "epoch": 0.03415865005519634, + "grad_norm": 7.551323413848877, + "learning_rate": 1.99991181898468e-05, + "loss": 0.6988, + "step": 1083 + }, + { + "epoch": 0.03419019082163697, + "grad_norm": 2.930840253829956, + "learning_rate": 1.999910457156188e-05, + "loss": 0.9117, + "step": 1084 + }, + { + "epoch": 0.03422173158807759, + "grad_norm": 6.080114841461182, + "learning_rate": 1.9999090848928538e-05, + "loss": 0.6662, + "step": 1085 + }, + { + "epoch": 0.034253272354518215, + "grad_norm": 3.871396064758301, + "learning_rate": 1.9999077021946922e-05, + "loss": 0.6732, + "step": 1086 + }, + { + "epoch": 0.03428481312095884, + "grad_norm": 6.347407341003418, + "learning_rate": 1.999906309061717e-05, + "loss": 0.6881, + "step": 1087 + }, + { + "epoch": 0.03431635388739947, + "grad_norm": 5.286763668060303, + "learning_rate": 1.999904905493943e-05, + "loss": 0.6935, + "step": 1088 + }, + { + "epoch": 0.03434789465384009, + "grad_norm": 4.168099403381348, + "learning_rate": 1.9999034914913848e-05, + "loss": 0.3964, + "step": 1089 + }, + { + "epoch": 0.03437943542028071, + "grad_norm": 5.4693732261657715, + "learning_rate": 1.9999020670540576e-05, + "loss": 0.6762, + "step": 1090 + }, + { + "epoch": 0.034410976186721336, + "grad_norm": 16.828420639038086, + "learning_rate": 1.999900632181976e-05, + "loss": 0.8726, + "step": 1091 + }, + { + "epoch": 0.03444251695316196, + "grad_norm": 5.342063903808594, + "learning_rate": 1.9998991868751544e-05, + "loss": 0.642, + "step": 1092 + }, + { + "epoch": 0.03447405771960259, + "grad_norm": 4.9667649269104, + "learning_rate": 1.9998977311336087e-05, + "loss": 0.6938, + "step": 1093 + }, + { + "epoch": 0.03450559848604321, + "grad_norm": 4.756962776184082, + "learning_rate": 1.9998962649573537e-05, + "loss": 0.6735, + "step": 1094 + }, + { + "epoch": 0.034537139252483834, + "grad_norm": 5.386860370635986, + "learning_rate": 1.999894788346405e-05, + "loss": 0.6632, + "step": 1095 + }, + { + "epoch": 0.03456868001892446, + "grad_norm": 8.118565559387207, + "learning_rate": 1.9998933013007776e-05, + "loss": 0.685, + "step": 1096 + }, + { + "epoch": 0.034600220785365086, + "grad_norm": 5.1070942878723145, + "learning_rate": 1.999891803820487e-05, + "loss": 0.6922, + "step": 1097 + }, + { + "epoch": 0.03463176155180571, + "grad_norm": 6.163454532623291, + "learning_rate": 1.9998902959055492e-05, + "loss": 0.6737, + "step": 1098 + }, + { + "epoch": 0.03466330231824633, + "grad_norm": 4.907265663146973, + "learning_rate": 1.9998887775559803e-05, + "loss": 0.6268, + "step": 1099 + }, + { + "epoch": 0.034694843084686955, + "grad_norm": 13.61937427520752, + "learning_rate": 1.9998872487717945e-05, + "loss": 0.3452, + "step": 1100 + }, + { + "epoch": 0.034726383851127585, + "grad_norm": 7.0663933753967285, + "learning_rate": 1.99988570955301e-05, + "loss": 0.669, + "step": 1101 + }, + { + "epoch": 0.03475792461756821, + "grad_norm": 7.519038677215576, + "learning_rate": 1.999884159899641e-05, + "loss": 0.6693, + "step": 1102 + }, + { + "epoch": 0.03478946538400883, + "grad_norm": 5.904048919677734, + "learning_rate": 1.9998825998117043e-05, + "loss": 0.6761, + "step": 1103 + }, + { + "epoch": 0.03482100615044945, + "grad_norm": 4.9077253341674805, + "learning_rate": 1.9998810292892168e-05, + "loss": 0.9262, + "step": 1104 + }, + { + "epoch": 0.03485254691689008, + "grad_norm": 5.827957630157471, + "learning_rate": 1.9998794483321942e-05, + "loss": 0.3749, + "step": 1105 + }, + { + "epoch": 0.034884087683330706, + "grad_norm": 7.115464210510254, + "learning_rate": 1.999877856940653e-05, + "loss": 0.8881, + "step": 1106 + }, + { + "epoch": 0.03491562844977133, + "grad_norm": 6.820563793182373, + "learning_rate": 1.9998762551146102e-05, + "loss": 0.6865, + "step": 1107 + }, + { + "epoch": 0.03494716921621195, + "grad_norm": 5.408565521240234, + "learning_rate": 1.9998746428540824e-05, + "loss": 0.6911, + "step": 1108 + }, + { + "epoch": 0.03497870998265258, + "grad_norm": 4.544974327087402, + "learning_rate": 1.999873020159086e-05, + "loss": 0.6765, + "step": 1109 + }, + { + "epoch": 0.035010250749093204, + "grad_norm": 4.796142101287842, + "learning_rate": 1.9998713870296383e-05, + "loss": 0.8899, + "step": 1110 + }, + { + "epoch": 0.03504179151553383, + "grad_norm": 6.824923515319824, + "learning_rate": 1.9998697434657566e-05, + "loss": 0.7071, + "step": 1111 + }, + { + "epoch": 0.03507333228197445, + "grad_norm": 4.822310447692871, + "learning_rate": 1.9998680894674574e-05, + "loss": 0.6878, + "step": 1112 + }, + { + "epoch": 0.03510487304841508, + "grad_norm": 9.358375549316406, + "learning_rate": 1.9998664250347584e-05, + "loss": 0.7228, + "step": 1113 + }, + { + "epoch": 0.0351364138148557, + "grad_norm": 6.488976001739502, + "learning_rate": 1.999864750167677e-05, + "loss": 0.6509, + "step": 1114 + }, + { + "epoch": 0.035167954581296325, + "grad_norm": 5.779963493347168, + "learning_rate": 1.9998630648662304e-05, + "loss": 0.6646, + "step": 1115 + }, + { + "epoch": 0.03519949534773695, + "grad_norm": 6.190042018890381, + "learning_rate": 1.9998613691304365e-05, + "loss": 0.6955, + "step": 1116 + }, + { + "epoch": 0.03523103611417758, + "grad_norm": 4.920349597930908, + "learning_rate": 1.9998596629603128e-05, + "loss": 0.676, + "step": 1117 + }, + { + "epoch": 0.0352625768806182, + "grad_norm": 10.383318901062012, + "learning_rate": 1.9998579463558768e-05, + "loss": 0.395, + "step": 1118 + }, + { + "epoch": 0.03529411764705882, + "grad_norm": 5.726223468780518, + "learning_rate": 1.9998562193171472e-05, + "loss": 0.7008, + "step": 1119 + }, + { + "epoch": 0.035325658413499446, + "grad_norm": 8.089290618896484, + "learning_rate": 1.9998544818441417e-05, + "loss": 0.7453, + "step": 1120 + }, + { + "epoch": 0.035357199179940076, + "grad_norm": 5.292104244232178, + "learning_rate": 1.9998527339368777e-05, + "loss": 0.6627, + "step": 1121 + }, + { + "epoch": 0.0353887399463807, + "grad_norm": 4.851055145263672, + "learning_rate": 1.9998509755953747e-05, + "loss": 0.6777, + "step": 1122 + }, + { + "epoch": 0.03542028071282132, + "grad_norm": 5.88997220993042, + "learning_rate": 1.99984920681965e-05, + "loss": 0.6895, + "step": 1123 + }, + { + "epoch": 0.035451821479261944, + "grad_norm": 6.272154808044434, + "learning_rate": 1.9998474276097227e-05, + "loss": 0.6696, + "step": 1124 + }, + { + "epoch": 0.035483362245702574, + "grad_norm": 6.503087997436523, + "learning_rate": 1.999845637965611e-05, + "loss": 0.6712, + "step": 1125 + }, + { + "epoch": 0.0355149030121432, + "grad_norm": 5.88395881652832, + "learning_rate": 1.9998438378873337e-05, + "loss": 0.7016, + "step": 1126 + }, + { + "epoch": 0.03554644377858382, + "grad_norm": 5.461755275726318, + "learning_rate": 1.9998420273749095e-05, + "loss": 0.6718, + "step": 1127 + }, + { + "epoch": 0.03557798454502444, + "grad_norm": 11.470030784606934, + "learning_rate": 1.9998402064283574e-05, + "loss": 0.6708, + "step": 1128 + }, + { + "epoch": 0.03560952531146507, + "grad_norm": 3.906484365463257, + "learning_rate": 1.9998383750476966e-05, + "loss": 0.9158, + "step": 1129 + }, + { + "epoch": 0.035641066077905695, + "grad_norm": 6.244319438934326, + "learning_rate": 1.999836533232946e-05, + "loss": 0.7078, + "step": 1130 + }, + { + "epoch": 0.03567260684434632, + "grad_norm": 20.588529586791992, + "learning_rate": 1.9998346809841245e-05, + "loss": 0.7045, + "step": 1131 + }, + { + "epoch": 0.03570414761078694, + "grad_norm": 5.543274879455566, + "learning_rate": 1.999832818301252e-05, + "loss": 0.6878, + "step": 1132 + }, + { + "epoch": 0.03573568837722756, + "grad_norm": 7.209839344024658, + "learning_rate": 1.9998309451843477e-05, + "loss": 0.9119, + "step": 1133 + }, + { + "epoch": 0.03576722914366819, + "grad_norm": 3.048698663711548, + "learning_rate": 1.999829061633431e-05, + "loss": 0.9331, + "step": 1134 + }, + { + "epoch": 0.035798769910108816, + "grad_norm": 8.313304901123047, + "learning_rate": 1.999827167648522e-05, + "loss": 0.7097, + "step": 1135 + }, + { + "epoch": 0.03583031067654944, + "grad_norm": 6.467309951782227, + "learning_rate": 1.99982526322964e-05, + "loss": 0.7289, + "step": 1136 + }, + { + "epoch": 0.03586185144299006, + "grad_norm": 5.410680770874023, + "learning_rate": 1.999823348376805e-05, + "loss": 0.6994, + "step": 1137 + }, + { + "epoch": 0.03589339220943069, + "grad_norm": 4.713596343994141, + "learning_rate": 1.9998214230900368e-05, + "loss": 0.6792, + "step": 1138 + }, + { + "epoch": 0.035924932975871314, + "grad_norm": 4.594524383544922, + "learning_rate": 1.999819487369356e-05, + "loss": 0.6625, + "step": 1139 + }, + { + "epoch": 0.03595647374231194, + "grad_norm": 8.594627380371094, + "learning_rate": 1.9998175412147824e-05, + "loss": 0.6915, + "step": 1140 + }, + { + "epoch": 0.03598801450875256, + "grad_norm": 6.1704936027526855, + "learning_rate": 1.9998155846263365e-05, + "loss": 0.6889, + "step": 1141 + }, + { + "epoch": 0.03601955527519319, + "grad_norm": 7.58643913269043, + "learning_rate": 1.9998136176040385e-05, + "loss": 0.6733, + "step": 1142 + }, + { + "epoch": 0.03605109604163381, + "grad_norm": 6.85713005065918, + "learning_rate": 1.9998116401479093e-05, + "loss": 0.6932, + "step": 1143 + }, + { + "epoch": 0.036082636808074435, + "grad_norm": 4.939759731292725, + "learning_rate": 1.999809652257969e-05, + "loss": 0.6744, + "step": 1144 + }, + { + "epoch": 0.03611417757451506, + "grad_norm": 5.361473560333252, + "learning_rate": 1.999807653934239e-05, + "loss": 0.6612, + "step": 1145 + }, + { + "epoch": 0.03614571834095569, + "grad_norm": 7.429842472076416, + "learning_rate": 1.9998056451767395e-05, + "loss": 0.7204, + "step": 1146 + }, + { + "epoch": 0.03617725910739631, + "grad_norm": 6.130885601043701, + "learning_rate": 1.9998036259854918e-05, + "loss": 0.6752, + "step": 1147 + }, + { + "epoch": 0.03620879987383693, + "grad_norm": 5.052056789398193, + "learning_rate": 1.999801596360517e-05, + "loss": 0.9484, + "step": 1148 + }, + { + "epoch": 0.036240340640277556, + "grad_norm": 9.17340087890625, + "learning_rate": 1.9997995563018362e-05, + "loss": 0.6924, + "step": 1149 + }, + { + "epoch": 0.036271881406718186, + "grad_norm": 6.3060126304626465, + "learning_rate": 1.999797505809471e-05, + "loss": 0.6783, + "step": 1150 + }, + { + "epoch": 0.03630342217315881, + "grad_norm": 6.552889347076416, + "learning_rate": 1.999795444883442e-05, + "loss": 0.6422, + "step": 1151 + }, + { + "epoch": 0.03633496293959943, + "grad_norm": 6.108257293701172, + "learning_rate": 1.9997933735237714e-05, + "loss": 0.6867, + "step": 1152 + }, + { + "epoch": 0.036366503706040054, + "grad_norm": 22.168840408325195, + "learning_rate": 1.9997912917304806e-05, + "loss": 0.6691, + "step": 1153 + }, + { + "epoch": 0.036398044472480684, + "grad_norm": 7.652955532073975, + "learning_rate": 1.9997891995035914e-05, + "loss": 0.6941, + "step": 1154 + }, + { + "epoch": 0.03642958523892131, + "grad_norm": 7.9139909744262695, + "learning_rate": 1.9997870968431256e-05, + "loss": 0.6985, + "step": 1155 + }, + { + "epoch": 0.03646112600536193, + "grad_norm": 6.988741874694824, + "learning_rate": 1.999784983749105e-05, + "loss": 0.664, + "step": 1156 + }, + { + "epoch": 0.03649266677180255, + "grad_norm": 5.153792858123779, + "learning_rate": 1.999782860221552e-05, + "loss": 0.6944, + "step": 1157 + }, + { + "epoch": 0.03652420753824318, + "grad_norm": 6.111942768096924, + "learning_rate": 1.9997807262604885e-05, + "loss": 0.6756, + "step": 1158 + }, + { + "epoch": 0.036555748304683805, + "grad_norm": 7.007165908813477, + "learning_rate": 1.999778581865937e-05, + "loss": 0.6734, + "step": 1159 + }, + { + "epoch": 0.03658728907112443, + "grad_norm": 4.921291828155518, + "learning_rate": 1.9997764270379193e-05, + "loss": 0.6858, + "step": 1160 + }, + { + "epoch": 0.03661882983756505, + "grad_norm": 6.134746551513672, + "learning_rate": 1.9997742617764583e-05, + "loss": 0.6953, + "step": 1161 + }, + { + "epoch": 0.03665037060400568, + "grad_norm": 6.209624290466309, + "learning_rate": 1.9997720860815767e-05, + "loss": 0.686, + "step": 1162 + }, + { + "epoch": 0.0366819113704463, + "grad_norm": 7.483974933624268, + "learning_rate": 1.999769899953297e-05, + "loss": 0.6915, + "step": 1163 + }, + { + "epoch": 0.036713452136886926, + "grad_norm": 6.685666561126709, + "learning_rate": 1.9997677033916423e-05, + "loss": 0.6819, + "step": 1164 + }, + { + "epoch": 0.03674499290332755, + "grad_norm": 6.115634441375732, + "learning_rate": 1.9997654963966355e-05, + "loss": 0.6787, + "step": 1165 + }, + { + "epoch": 0.03677653366976818, + "grad_norm": 12.999553680419922, + "learning_rate": 1.999763278968299e-05, + "loss": 0.7309, + "step": 1166 + }, + { + "epoch": 0.0368080744362088, + "grad_norm": 4.57751989364624, + "learning_rate": 1.9997610511066562e-05, + "loss": 0.3724, + "step": 1167 + }, + { + "epoch": 0.036839615202649424, + "grad_norm": 6.282439708709717, + "learning_rate": 1.999758812811731e-05, + "loss": 0.6869, + "step": 1168 + }, + { + "epoch": 0.03687115596909005, + "grad_norm": 3.800746440887451, + "learning_rate": 1.9997565640835462e-05, + "loss": 0.3972, + "step": 1169 + }, + { + "epoch": 0.03690269673553068, + "grad_norm": 5.954683303833008, + "learning_rate": 1.9997543049221254e-05, + "loss": 0.6793, + "step": 1170 + }, + { + "epoch": 0.0369342375019713, + "grad_norm": 4.041647911071777, + "learning_rate": 1.9997520353274922e-05, + "loss": 0.6541, + "step": 1171 + }, + { + "epoch": 0.03696577826841192, + "grad_norm": 5.8372883796691895, + "learning_rate": 1.99974975529967e-05, + "loss": 0.6525, + "step": 1172 + }, + { + "epoch": 0.036997319034852545, + "grad_norm": 21.526683807373047, + "learning_rate": 1.999747464838683e-05, + "loss": 0.8952, + "step": 1173 + }, + { + "epoch": 0.03702885980129317, + "grad_norm": 3.955612897872925, + "learning_rate": 1.9997451639445547e-05, + "loss": 0.6472, + "step": 1174 + }, + { + "epoch": 0.0370604005677338, + "grad_norm": 4.168172836303711, + "learning_rate": 1.9997428526173092e-05, + "loss": 0.6346, + "step": 1175 + }, + { + "epoch": 0.03709194133417442, + "grad_norm": 5.705156326293945, + "learning_rate": 1.999740530856971e-05, + "loss": 0.6794, + "step": 1176 + }, + { + "epoch": 0.03712348210061504, + "grad_norm": 4.5772833824157715, + "learning_rate": 1.9997381986635643e-05, + "loss": 0.6354, + "step": 1177 + }, + { + "epoch": 0.037155022867055666, + "grad_norm": 6.313110828399658, + "learning_rate": 1.999735856037113e-05, + "loss": 0.6997, + "step": 1178 + }, + { + "epoch": 0.037186563633496296, + "grad_norm": 5.426894664764404, + "learning_rate": 1.9997335029776418e-05, + "loss": 0.7069, + "step": 1179 + }, + { + "epoch": 0.03721810439993692, + "grad_norm": 9.704507827758789, + "learning_rate": 1.9997311394851753e-05, + "loss": 0.8958, + "step": 1180 + }, + { + "epoch": 0.03724964516637754, + "grad_norm": 15.134077072143555, + "learning_rate": 1.999728765559738e-05, + "loss": 0.7102, + "step": 1181 + }, + { + "epoch": 0.037281185932818164, + "grad_norm": 5.671441555023193, + "learning_rate": 1.9997263812013547e-05, + "loss": 0.6804, + "step": 1182 + }, + { + "epoch": 0.037312726699258794, + "grad_norm": 7.862107276916504, + "learning_rate": 1.9997239864100504e-05, + "loss": 0.6477, + "step": 1183 + }, + { + "epoch": 0.03734426746569942, + "grad_norm": 10.753114700317383, + "learning_rate": 1.9997215811858498e-05, + "loss": 0.6798, + "step": 1184 + }, + { + "epoch": 0.03737580823214004, + "grad_norm": 4.659350395202637, + "learning_rate": 1.9997191655287786e-05, + "loss": 0.6842, + "step": 1185 + }, + { + "epoch": 0.03740734899858066, + "grad_norm": 5.005865097045898, + "learning_rate": 1.999716739438862e-05, + "loss": 0.6681, + "step": 1186 + }, + { + "epoch": 0.03743888976502129, + "grad_norm": 8.621869087219238, + "learning_rate": 1.999714302916124e-05, + "loss": 0.686, + "step": 1187 + }, + { + "epoch": 0.037470430531461915, + "grad_norm": 7.339015960693359, + "learning_rate": 1.9997118559605914e-05, + "loss": 0.6707, + "step": 1188 + }, + { + "epoch": 0.03750197129790254, + "grad_norm": 5.878515720367432, + "learning_rate": 1.9997093985722896e-05, + "loss": 0.6651, + "step": 1189 + }, + { + "epoch": 0.03753351206434316, + "grad_norm": 3.127394437789917, + "learning_rate": 1.9997069307512436e-05, + "loss": 0.8861, + "step": 1190 + }, + { + "epoch": 0.03756505283078379, + "grad_norm": 6.591914176940918, + "learning_rate": 1.9997044524974797e-05, + "loss": 0.6899, + "step": 1191 + }, + { + "epoch": 0.03759659359722441, + "grad_norm": 6.668528079986572, + "learning_rate": 1.9997019638110235e-05, + "loss": 0.6882, + "step": 1192 + }, + { + "epoch": 0.037628134363665036, + "grad_norm": 5.184406280517578, + "learning_rate": 1.9996994646919013e-05, + "loss": 0.6777, + "step": 1193 + }, + { + "epoch": 0.03765967513010566, + "grad_norm": 3.18430757522583, + "learning_rate": 1.999696955140139e-05, + "loss": 0.3799, + "step": 1194 + }, + { + "epoch": 0.03769121589654629, + "grad_norm": 6.992948055267334, + "learning_rate": 1.9996944351557625e-05, + "loss": 0.7061, + "step": 1195 + }, + { + "epoch": 0.03772275666298691, + "grad_norm": 5.604113578796387, + "learning_rate": 1.999691904738798e-05, + "loss": 0.6949, + "step": 1196 + }, + { + "epoch": 0.037754297429427534, + "grad_norm": 5.357737064361572, + "learning_rate": 1.9996893638892726e-05, + "loss": 0.6689, + "step": 1197 + }, + { + "epoch": 0.03778583819586816, + "grad_norm": 3.867499828338623, + "learning_rate": 1.9996868126072124e-05, + "loss": 0.6896, + "step": 1198 + }, + { + "epoch": 0.03781737896230879, + "grad_norm": 4.137993812561035, + "learning_rate": 1.999684250892644e-05, + "loss": 0.6627, + "step": 1199 + }, + { + "epoch": 0.03784891972874941, + "grad_norm": 4.1114373207092285, + "learning_rate": 1.9996816787455945e-05, + "loss": 0.6686, + "step": 1200 + }, + { + "epoch": 0.03788046049519003, + "grad_norm": 4.820106506347656, + "learning_rate": 1.9996790961660904e-05, + "loss": 0.6698, + "step": 1201 + }, + { + "epoch": 0.037912001261630655, + "grad_norm": 7.231561660766602, + "learning_rate": 1.9996765031541585e-05, + "loss": 0.6559, + "step": 1202 + }, + { + "epoch": 0.037943542028071285, + "grad_norm": 4.326260089874268, + "learning_rate": 1.999673899709826e-05, + "loss": 0.9153, + "step": 1203 + }, + { + "epoch": 0.03797508279451191, + "grad_norm": 7.6405205726623535, + "learning_rate": 1.9996712858331202e-05, + "loss": 0.6496, + "step": 1204 + }, + { + "epoch": 0.03800662356095253, + "grad_norm": 14.984155654907227, + "learning_rate": 1.9996686615240683e-05, + "loss": 0.6925, + "step": 1205 + }, + { + "epoch": 0.038038164327393154, + "grad_norm": 6.412452697753906, + "learning_rate": 1.9996660267826978e-05, + "loss": 0.6414, + "step": 1206 + }, + { + "epoch": 0.03806970509383378, + "grad_norm": 7.068026065826416, + "learning_rate": 1.999663381609036e-05, + "loss": 0.6663, + "step": 1207 + }, + { + "epoch": 0.038101245860274406, + "grad_norm": 4.489713191986084, + "learning_rate": 1.9996607260031105e-05, + "loss": 0.6886, + "step": 1208 + }, + { + "epoch": 0.03813278662671503, + "grad_norm": 3.132648468017578, + "learning_rate": 1.9996580599649494e-05, + "loss": 0.3772, + "step": 1209 + }, + { + "epoch": 0.03816432739315565, + "grad_norm": 3.0172207355499268, + "learning_rate": 1.99965538349458e-05, + "loss": 0.8992, + "step": 1210 + }, + { + "epoch": 0.03819586815959628, + "grad_norm": 7.255639553070068, + "learning_rate": 1.9996526965920305e-05, + "loss": 0.6699, + "step": 1211 + }, + { + "epoch": 0.038227408926036904, + "grad_norm": 9.982434272766113, + "learning_rate": 1.999649999257329e-05, + "loss": 0.6926, + "step": 1212 + }, + { + "epoch": 0.03825894969247753, + "grad_norm": 11.053763389587402, + "learning_rate": 1.9996472914905034e-05, + "loss": 0.6864, + "step": 1213 + }, + { + "epoch": 0.03829049045891815, + "grad_norm": 5.018710613250732, + "learning_rate": 1.9996445732915824e-05, + "loss": 0.653, + "step": 1214 + }, + { + "epoch": 0.03832203122535877, + "grad_norm": 5.283173561096191, + "learning_rate": 1.9996418446605938e-05, + "loss": 0.6722, + "step": 1215 + }, + { + "epoch": 0.0383535719917994, + "grad_norm": 6.0497145652771, + "learning_rate": 1.9996391055975663e-05, + "loss": 0.6531, + "step": 1216 + }, + { + "epoch": 0.038385112758240025, + "grad_norm": 3.7649712562561035, + "learning_rate": 1.9996363561025286e-05, + "loss": 0.6432, + "step": 1217 + }, + { + "epoch": 0.03841665352468065, + "grad_norm": 6.528625965118408, + "learning_rate": 1.9996335961755092e-05, + "loss": 0.6642, + "step": 1218 + }, + { + "epoch": 0.03844819429112127, + "grad_norm": 2.6026411056518555, + "learning_rate": 1.9996308258165372e-05, + "loss": 0.9095, + "step": 1219 + }, + { + "epoch": 0.0384797350575619, + "grad_norm": 12.476468086242676, + "learning_rate": 1.9996280450256415e-05, + "loss": 0.6735, + "step": 1220 + }, + { + "epoch": 0.038511275824002524, + "grad_norm": 36.68544006347656, + "learning_rate": 1.999625253802851e-05, + "loss": 0.3713, + "step": 1221 + }, + { + "epoch": 0.038542816590443146, + "grad_norm": 4.641376972198486, + "learning_rate": 1.9996224521481945e-05, + "loss": 0.66, + "step": 1222 + }, + { + "epoch": 0.03857435735688377, + "grad_norm": 8.822100639343262, + "learning_rate": 1.9996196400617015e-05, + "loss": 0.6533, + "step": 1223 + }, + { + "epoch": 0.0386058981233244, + "grad_norm": 6.165101528167725, + "learning_rate": 1.9996168175434018e-05, + "loss": 0.667, + "step": 1224 + }, + { + "epoch": 0.03863743888976502, + "grad_norm": 4.710172176361084, + "learning_rate": 1.999613984593324e-05, + "loss": 0.6382, + "step": 1225 + }, + { + "epoch": 0.038668979656205645, + "grad_norm": 4.341470241546631, + "learning_rate": 1.999611141211498e-05, + "loss": 0.6792, + "step": 1226 + }, + { + "epoch": 0.03870052042264627, + "grad_norm": 5.549160480499268, + "learning_rate": 1.999608287397954e-05, + "loss": 0.6897, + "step": 1227 + }, + { + "epoch": 0.0387320611890869, + "grad_norm": 5.95551061630249, + "learning_rate": 1.9996054231527207e-05, + "loss": 0.6647, + "step": 1228 + }, + { + "epoch": 0.03876360195552752, + "grad_norm": 7.0042338371276855, + "learning_rate": 1.9996025484758292e-05, + "loss": 0.6712, + "step": 1229 + }, + { + "epoch": 0.03879514272196814, + "grad_norm": 5.2130842208862305, + "learning_rate": 1.999599663367309e-05, + "loss": 0.7205, + "step": 1230 + }, + { + "epoch": 0.038826683488408766, + "grad_norm": 7.818285942077637, + "learning_rate": 1.9995967678271898e-05, + "loss": 0.406, + "step": 1231 + }, + { + "epoch": 0.038858224254849395, + "grad_norm": 5.454767227172852, + "learning_rate": 1.9995938618555022e-05, + "loss": 0.697, + "step": 1232 + }, + { + "epoch": 0.03888976502129002, + "grad_norm": 7.592140197753906, + "learning_rate": 1.9995909454522766e-05, + "loss": 0.7026, + "step": 1233 + }, + { + "epoch": 0.03892130578773064, + "grad_norm": 6.2525315284729, + "learning_rate": 1.9995880186175434e-05, + "loss": 0.6277, + "step": 1234 + }, + { + "epoch": 0.038952846554171264, + "grad_norm": 3.1691815853118896, + "learning_rate": 1.9995850813513325e-05, + "loss": 0.8993, + "step": 1235 + }, + { + "epoch": 0.038984387320611893, + "grad_norm": 4.800058841705322, + "learning_rate": 1.999582133653676e-05, + "loss": 0.6473, + "step": 1236 + }, + { + "epoch": 0.039015928087052516, + "grad_norm": 3.4071362018585205, + "learning_rate": 1.9995791755246032e-05, + "loss": 0.4194, + "step": 1237 + }, + { + "epoch": 0.03904746885349314, + "grad_norm": 7.089361667633057, + "learning_rate": 1.999576206964146e-05, + "loss": 0.6901, + "step": 1238 + }, + { + "epoch": 0.03907900961993376, + "grad_norm": 8.863525390625, + "learning_rate": 1.9995732279723344e-05, + "loss": 0.6732, + "step": 1239 + }, + { + "epoch": 0.03911055038637439, + "grad_norm": 6.161620140075684, + "learning_rate": 1.9995702385492e-05, + "loss": 0.6815, + "step": 1240 + }, + { + "epoch": 0.039142091152815014, + "grad_norm": 39.11034393310547, + "learning_rate": 1.9995672386947743e-05, + "loss": 0.6812, + "step": 1241 + }, + { + "epoch": 0.03917363191925564, + "grad_norm": 6.967360973358154, + "learning_rate": 1.9995642284090887e-05, + "loss": 0.6568, + "step": 1242 + }, + { + "epoch": 0.03920517268569626, + "grad_norm": 4.985154151916504, + "learning_rate": 1.9995612076921737e-05, + "loss": 0.648, + "step": 1243 + }, + { + "epoch": 0.03923671345213689, + "grad_norm": 7.524902820587158, + "learning_rate": 1.999558176544061e-05, + "loss": 0.6723, + "step": 1244 + }, + { + "epoch": 0.03926825421857751, + "grad_norm": 5.712900638580322, + "learning_rate": 1.9995551349647836e-05, + "loss": 0.6859, + "step": 1245 + }, + { + "epoch": 0.039299794985018136, + "grad_norm": 3.4193358421325684, + "learning_rate": 1.9995520829543716e-05, + "loss": 0.4068, + "step": 1246 + }, + { + "epoch": 0.03933133575145876, + "grad_norm": 4.333251953125, + "learning_rate": 1.9995490205128572e-05, + "loss": 0.3914, + "step": 1247 + }, + { + "epoch": 0.03936287651789939, + "grad_norm": 7.996802806854248, + "learning_rate": 1.9995459476402734e-05, + "loss": 0.6328, + "step": 1248 + }, + { + "epoch": 0.03939441728434001, + "grad_norm": 5.199215412139893, + "learning_rate": 1.999542864336651e-05, + "loss": 0.6777, + "step": 1249 + }, + { + "epoch": 0.039425958050780634, + "grad_norm": 7.992265224456787, + "learning_rate": 1.9995397706020227e-05, + "loss": 0.671, + "step": 1250 + }, + { + "epoch": 0.039457498817221257, + "grad_norm": 5.411418914794922, + "learning_rate": 1.999536666436421e-05, + "loss": 0.6577, + "step": 1251 + }, + { + "epoch": 0.039489039583661886, + "grad_norm": 10.209061622619629, + "learning_rate": 1.999533551839878e-05, + "loss": 0.6578, + "step": 1252 + }, + { + "epoch": 0.03952058035010251, + "grad_norm": 13.135923385620117, + "learning_rate": 1.9995304268124264e-05, + "loss": 0.6561, + "step": 1253 + }, + { + "epoch": 0.03955212111654313, + "grad_norm": 7.453579902648926, + "learning_rate": 1.9995272913540983e-05, + "loss": 0.6994, + "step": 1254 + }, + { + "epoch": 0.039583661882983755, + "grad_norm": 7.433265686035156, + "learning_rate": 1.999524145464927e-05, + "loss": 0.6848, + "step": 1255 + }, + { + "epoch": 0.039615202649424384, + "grad_norm": 4.511242389678955, + "learning_rate": 1.9995209891449453e-05, + "loss": 0.6434, + "step": 1256 + }, + { + "epoch": 0.03964674341586501, + "grad_norm": 7.332136631011963, + "learning_rate": 1.9995178223941858e-05, + "loss": 0.6499, + "step": 1257 + }, + { + "epoch": 0.03967828418230563, + "grad_norm": 7.957773685455322, + "learning_rate": 1.9995146452126818e-05, + "loss": 0.6817, + "step": 1258 + }, + { + "epoch": 0.03970982494874625, + "grad_norm": 4.032230377197266, + "learning_rate": 1.999511457600466e-05, + "loss": 0.6266, + "step": 1259 + }, + { + "epoch": 0.039741365715186876, + "grad_norm": 5.037559986114502, + "learning_rate": 1.9995082595575725e-05, + "loss": 0.6464, + "step": 1260 + }, + { + "epoch": 0.039772906481627505, + "grad_norm": 3.361104965209961, + "learning_rate": 1.999505051084034e-05, + "loss": 0.9217, + "step": 1261 + }, + { + "epoch": 0.03980444724806813, + "grad_norm": 5.008788108825684, + "learning_rate": 1.9995018321798842e-05, + "loss": 0.6613, + "step": 1262 + }, + { + "epoch": 0.03983598801450875, + "grad_norm": 8.019538879394531, + "learning_rate": 1.9994986028451565e-05, + "loss": 0.6637, + "step": 1263 + }, + { + "epoch": 0.039867528780949374, + "grad_norm": 4.401232719421387, + "learning_rate": 1.9994953630798852e-05, + "loss": 0.6549, + "step": 1264 + }, + { + "epoch": 0.039899069547390004, + "grad_norm": 4.6221723556518555, + "learning_rate": 1.9994921128841034e-05, + "loss": 0.6336, + "step": 1265 + }, + { + "epoch": 0.039930610313830626, + "grad_norm": 6.583164691925049, + "learning_rate": 1.999488852257846e-05, + "loss": 0.6618, + "step": 1266 + }, + { + "epoch": 0.03996215108027125, + "grad_norm": 15.803287506103516, + "learning_rate": 1.9994855812011456e-05, + "loss": 0.3675, + "step": 1267 + }, + { + "epoch": 0.03999369184671187, + "grad_norm": 4.559471607208252, + "learning_rate": 1.999482299714037e-05, + "loss": 0.6514, + "step": 1268 + }, + { + "epoch": 0.0400252326131525, + "grad_norm": 5.209477424621582, + "learning_rate": 1.999479007796555e-05, + "loss": 0.6735, + "step": 1269 + }, + { + "epoch": 0.040056773379593125, + "grad_norm": 4.202414512634277, + "learning_rate": 1.9994757054487332e-05, + "loss": 0.6773, + "step": 1270 + }, + { + "epoch": 0.04008831414603375, + "grad_norm": 4.600757598876953, + "learning_rate": 1.9994723926706062e-05, + "loss": 0.6691, + "step": 1271 + }, + { + "epoch": 0.04011985491247437, + "grad_norm": 3.5494163036346436, + "learning_rate": 1.999469069462209e-05, + "loss": 0.6608, + "step": 1272 + }, + { + "epoch": 0.040151395678915, + "grad_norm": 4.631522178649902, + "learning_rate": 1.999465735823576e-05, + "loss": 0.6903, + "step": 1273 + }, + { + "epoch": 0.04018293644535562, + "grad_norm": 3.276345729827881, + "learning_rate": 1.999462391754742e-05, + "loss": 0.8747, + "step": 1274 + }, + { + "epoch": 0.040214477211796246, + "grad_norm": 2.8878281116485596, + "learning_rate": 1.9994590372557416e-05, + "loss": 0.3735, + "step": 1275 + }, + { + "epoch": 0.04024601797823687, + "grad_norm": 4.661767959594727, + "learning_rate": 1.9994556723266102e-05, + "loss": 0.694, + "step": 1276 + }, + { + "epoch": 0.0402775587446775, + "grad_norm": 4.13231897354126, + "learning_rate": 1.999452296967383e-05, + "loss": 0.6587, + "step": 1277 + }, + { + "epoch": 0.04030909951111812, + "grad_norm": 2.8578672409057617, + "learning_rate": 1.999448911178095e-05, + "loss": 0.8897, + "step": 1278 + }, + { + "epoch": 0.040340640277558744, + "grad_norm": 4.175153732299805, + "learning_rate": 1.9994455149587814e-05, + "loss": 0.6583, + "step": 1279 + }, + { + "epoch": 0.04037218104399937, + "grad_norm": 3.874084234237671, + "learning_rate": 1.9994421083094776e-05, + "loss": 0.6473, + "step": 1280 + }, + { + "epoch": 0.040403721810439996, + "grad_norm": 4.520504951477051, + "learning_rate": 1.9994386912302197e-05, + "loss": 0.6933, + "step": 1281 + }, + { + "epoch": 0.04043526257688062, + "grad_norm": 5.6133880615234375, + "learning_rate": 1.999435263721043e-05, + "loss": 0.6446, + "step": 1282 + }, + { + "epoch": 0.04046680334332124, + "grad_norm": 8.935012817382812, + "learning_rate": 1.9994318257819833e-05, + "loss": 0.4209, + "step": 1283 + }, + { + "epoch": 0.040498344109761865, + "grad_norm": 5.197874069213867, + "learning_rate": 1.999428377413076e-05, + "loss": 0.6744, + "step": 1284 + }, + { + "epoch": 0.040529884876202495, + "grad_norm": 5.3234100341796875, + "learning_rate": 1.999424918614358e-05, + "loss": 0.6826, + "step": 1285 + }, + { + "epoch": 0.04056142564264312, + "grad_norm": 6.771747589111328, + "learning_rate": 1.999421449385865e-05, + "loss": 0.6703, + "step": 1286 + }, + { + "epoch": 0.04059296640908374, + "grad_norm": 6.028203964233398, + "learning_rate": 1.9994179697276328e-05, + "loss": 0.6462, + "step": 1287 + }, + { + "epoch": 0.04062450717552436, + "grad_norm": 3.4060583114624023, + "learning_rate": 1.9994144796396985e-05, + "loss": 0.4101, + "step": 1288 + }, + { + "epoch": 0.04065604794196499, + "grad_norm": 7.148763179779053, + "learning_rate": 1.9994109791220978e-05, + "loss": 0.6841, + "step": 1289 + }, + { + "epoch": 0.040687588708405616, + "grad_norm": 4.7566657066345215, + "learning_rate": 1.9994074681748675e-05, + "loss": 0.3671, + "step": 1290 + }, + { + "epoch": 0.04071912947484624, + "grad_norm": 3.0442488193511963, + "learning_rate": 1.999403946798044e-05, + "loss": 0.8897, + "step": 1291 + }, + { + "epoch": 0.04075067024128686, + "grad_norm": 7.648116111755371, + "learning_rate": 1.999400414991665e-05, + "loss": 0.6566, + "step": 1292 + }, + { + "epoch": 0.04078221100772749, + "grad_norm": 5.691143035888672, + "learning_rate": 1.999396872755766e-05, + "loss": 0.6658, + "step": 1293 + }, + { + "epoch": 0.040813751774168114, + "grad_norm": 3.851215362548828, + "learning_rate": 1.999393320090385e-05, + "loss": 0.6628, + "step": 1294 + }, + { + "epoch": 0.04084529254060874, + "grad_norm": 4.995465278625488, + "learning_rate": 1.9993897569955582e-05, + "loss": 0.6801, + "step": 1295 + }, + { + "epoch": 0.04087683330704936, + "grad_norm": 4.480391025543213, + "learning_rate": 1.999386183471324e-05, + "loss": 0.6659, + "step": 1296 + }, + { + "epoch": 0.04090837407348999, + "grad_norm": 4.808745861053467, + "learning_rate": 1.9993825995177183e-05, + "loss": 0.6551, + "step": 1297 + }, + { + "epoch": 0.04093991483993061, + "grad_norm": 4.857883930206299, + "learning_rate": 1.9993790051347795e-05, + "loss": 0.6719, + "step": 1298 + }, + { + "epoch": 0.040971455606371235, + "grad_norm": 4.433502197265625, + "learning_rate": 1.9993754003225446e-05, + "loss": 0.6848, + "step": 1299 + }, + { + "epoch": 0.04100299637281186, + "grad_norm": 3.8633131980895996, + "learning_rate": 1.999371785081052e-05, + "loss": 0.9124, + "step": 1300 + }, + { + "epoch": 0.04103453713925248, + "grad_norm": 4.04303503036499, + "learning_rate": 1.9993681594103383e-05, + "loss": 0.6536, + "step": 1301 + }, + { + "epoch": 0.04106607790569311, + "grad_norm": 5.758216857910156, + "learning_rate": 1.999364523310442e-05, + "loss": 0.4028, + "step": 1302 + }, + { + "epoch": 0.04109761867213373, + "grad_norm": 3.467033624649048, + "learning_rate": 1.9993608767814008e-05, + "loss": 0.8912, + "step": 1303 + }, + { + "epoch": 0.041129159438574356, + "grad_norm": 4.501498699188232, + "learning_rate": 1.9993572198232528e-05, + "loss": 0.6654, + "step": 1304 + }, + { + "epoch": 0.04116070020501498, + "grad_norm": 4.459884166717529, + "learning_rate": 1.9993535524360364e-05, + "loss": 0.6978, + "step": 1305 + }, + { + "epoch": 0.04119224097145561, + "grad_norm": 4.282965183258057, + "learning_rate": 1.9993498746197895e-05, + "loss": 0.6738, + "step": 1306 + }, + { + "epoch": 0.04122378173789623, + "grad_norm": 5.11058235168457, + "learning_rate": 1.999346186374551e-05, + "loss": 0.6714, + "step": 1307 + }, + { + "epoch": 0.041255322504336854, + "grad_norm": 5.023053169250488, + "learning_rate": 1.999342487700359e-05, + "loss": 0.6768, + "step": 1308 + }, + { + "epoch": 0.04128686327077748, + "grad_norm": 6.425107479095459, + "learning_rate": 1.9993387785972522e-05, + "loss": 0.6807, + "step": 1309 + }, + { + "epoch": 0.04131840403721811, + "grad_norm": 4.092491149902344, + "learning_rate": 1.999335059065269e-05, + "loss": 0.6432, + "step": 1310 + }, + { + "epoch": 0.04134994480365873, + "grad_norm": 5.109757900238037, + "learning_rate": 1.999331329104449e-05, + "loss": 0.6701, + "step": 1311 + }, + { + "epoch": 0.04138148557009935, + "grad_norm": 4.90879487991333, + "learning_rate": 1.99932758871483e-05, + "loss": 0.6369, + "step": 1312 + }, + { + "epoch": 0.041413026336539975, + "grad_norm": 5.187123775482178, + "learning_rate": 1.999323837896452e-05, + "loss": 0.6437, + "step": 1313 + }, + { + "epoch": 0.041444567102980605, + "grad_norm": 5.234122276306152, + "learning_rate": 1.9993200766493538e-05, + "loss": 0.6537, + "step": 1314 + }, + { + "epoch": 0.04147610786942123, + "grad_norm": 6.693339824676514, + "learning_rate": 1.9993163049735748e-05, + "loss": 0.7069, + "step": 1315 + }, + { + "epoch": 0.04150764863586185, + "grad_norm": 4.291707992553711, + "learning_rate": 1.9993125228691535e-05, + "loss": 0.691, + "step": 1316 + }, + { + "epoch": 0.04153918940230247, + "grad_norm": 4.795947074890137, + "learning_rate": 1.999308730336131e-05, + "loss": 0.6416, + "step": 1317 + }, + { + "epoch": 0.0415707301687431, + "grad_norm": 3.8477258682250977, + "learning_rate": 1.9993049273745454e-05, + "loss": 0.6511, + "step": 1318 + }, + { + "epoch": 0.041602270935183726, + "grad_norm": 5.133565902709961, + "learning_rate": 1.999301113984437e-05, + "loss": 0.6627, + "step": 1319 + }, + { + "epoch": 0.04163381170162435, + "grad_norm": 3.7376692295074463, + "learning_rate": 1.9992972901658458e-05, + "loss": 0.8853, + "step": 1320 + }, + { + "epoch": 0.04166535246806497, + "grad_norm": 4.037112236022949, + "learning_rate": 1.9992934559188112e-05, + "loss": 0.6639, + "step": 1321 + }, + { + "epoch": 0.0416968932345056, + "grad_norm": 3.818847894668579, + "learning_rate": 1.9992896112433736e-05, + "loss": 0.6525, + "step": 1322 + }, + { + "epoch": 0.041728434000946224, + "grad_norm": 4.901750564575195, + "learning_rate": 1.999285756139573e-05, + "loss": 0.6545, + "step": 1323 + }, + { + "epoch": 0.04175997476738685, + "grad_norm": 7.480557918548584, + "learning_rate": 1.9992818906074493e-05, + "loss": 0.6494, + "step": 1324 + }, + { + "epoch": 0.04179151553382747, + "grad_norm": 4.313000679016113, + "learning_rate": 1.9992780146470433e-05, + "loss": 0.6878, + "step": 1325 + }, + { + "epoch": 0.0418230563002681, + "grad_norm": 2.748401641845703, + "learning_rate": 1.9992741282583955e-05, + "loss": 0.9557, + "step": 1326 + }, + { + "epoch": 0.04185459706670872, + "grad_norm": 5.103557586669922, + "learning_rate": 1.9992702314415463e-05, + "loss": 0.6812, + "step": 1327 + }, + { + "epoch": 0.041886137833149345, + "grad_norm": 2.552330255508423, + "learning_rate": 1.999266324196536e-05, + "loss": 0.4364, + "step": 1328 + }, + { + "epoch": 0.04191767859958997, + "grad_norm": 4.939621925354004, + "learning_rate": 1.9992624065234057e-05, + "loss": 0.6021, + "step": 1329 + }, + { + "epoch": 0.0419492193660306, + "grad_norm": 4.594666481018066, + "learning_rate": 1.9992584784221963e-05, + "loss": 0.6635, + "step": 1330 + }, + { + "epoch": 0.04198076013247122, + "grad_norm": 3.978165864944458, + "learning_rate": 1.9992545398929488e-05, + "loss": 0.6909, + "step": 1331 + }, + { + "epoch": 0.04201230089891184, + "grad_norm": 4.929741382598877, + "learning_rate": 1.9992505909357044e-05, + "loss": 0.6859, + "step": 1332 + }, + { + "epoch": 0.042043841665352466, + "grad_norm": 4.705532073974609, + "learning_rate": 1.9992466315505044e-05, + "loss": 0.8913, + "step": 1333 + }, + { + "epoch": 0.042075382431793096, + "grad_norm": 5.464250564575195, + "learning_rate": 1.9992426617373895e-05, + "loss": 0.6894, + "step": 1334 + }, + { + "epoch": 0.04210692319823372, + "grad_norm": 2.942199945449829, + "learning_rate": 1.999238681496402e-05, + "loss": 0.8821, + "step": 1335 + }, + { + "epoch": 0.04213846396467434, + "grad_norm": 6.964365482330322, + "learning_rate": 1.9992346908275822e-05, + "loss": 0.6427, + "step": 1336 + }, + { + "epoch": 0.042170004731114964, + "grad_norm": 2.526144504547119, + "learning_rate": 1.999230689730973e-05, + "loss": 0.3905, + "step": 1337 + }, + { + "epoch": 0.042201545497555594, + "grad_norm": 5.4747090339660645, + "learning_rate": 1.9992266782066156e-05, + "loss": 0.6372, + "step": 1338 + }, + { + "epoch": 0.04223308626399622, + "grad_norm": 4.834805488586426, + "learning_rate": 1.9992226562545523e-05, + "loss": 0.6423, + "step": 1339 + }, + { + "epoch": 0.04226462703043684, + "grad_norm": 2.5780718326568604, + "learning_rate": 1.999218623874824e-05, + "loss": 0.8926, + "step": 1340 + }, + { + "epoch": 0.04229616779687746, + "grad_norm": 4.426979064941406, + "learning_rate": 1.999214581067474e-05, + "loss": 0.6528, + "step": 1341 + }, + { + "epoch": 0.042327708563318085, + "grad_norm": 3.9208686351776123, + "learning_rate": 1.9992105278325436e-05, + "loss": 0.6745, + "step": 1342 + }, + { + "epoch": 0.042359249329758715, + "grad_norm": 4.27236270904541, + "learning_rate": 1.999206464170076e-05, + "loss": 0.6775, + "step": 1343 + }, + { + "epoch": 0.04239079009619934, + "grad_norm": 4.432453632354736, + "learning_rate": 1.999202390080113e-05, + "loss": 0.647, + "step": 1344 + }, + { + "epoch": 0.04242233086263996, + "grad_norm": 4.781933784484863, + "learning_rate": 1.999198305562697e-05, + "loss": 0.6908, + "step": 1345 + }, + { + "epoch": 0.04245387162908058, + "grad_norm": 5.702025413513184, + "learning_rate": 1.9991942106178706e-05, + "loss": 0.6795, + "step": 1346 + }, + { + "epoch": 0.04248541239552121, + "grad_norm": 5.853853225708008, + "learning_rate": 1.9991901052456768e-05, + "loss": 0.91, + "step": 1347 + }, + { + "epoch": 0.042516953161961836, + "grad_norm": 5.232560634613037, + "learning_rate": 1.9991859894461587e-05, + "loss": 0.6712, + "step": 1348 + }, + { + "epoch": 0.04254849392840246, + "grad_norm": 6.892151355743408, + "learning_rate": 1.999181863219359e-05, + "loss": 0.7131, + "step": 1349 + }, + { + "epoch": 0.04258003469484308, + "grad_norm": 3.8885130882263184, + "learning_rate": 1.9991777265653206e-05, + "loss": 0.6824, + "step": 1350 + }, + { + "epoch": 0.04261157546128371, + "grad_norm": 4.169528007507324, + "learning_rate": 1.9991735794840864e-05, + "loss": 0.6765, + "step": 1351 + }, + { + "epoch": 0.042643116227724334, + "grad_norm": 5.108818531036377, + "learning_rate": 1.9991694219757007e-05, + "loss": 0.6725, + "step": 1352 + }, + { + "epoch": 0.04267465699416496, + "grad_norm": 11.639219284057617, + "learning_rate": 1.999165254040206e-05, + "loss": 0.6687, + "step": 1353 + }, + { + "epoch": 0.04270619776060558, + "grad_norm": 7.769289493560791, + "learning_rate": 1.9991610756776455e-05, + "loss": 0.6679, + "step": 1354 + }, + { + "epoch": 0.04273773852704621, + "grad_norm": 4.668744087219238, + "learning_rate": 1.999156886888064e-05, + "loss": 0.6902, + "step": 1355 + }, + { + "epoch": 0.04276927929348683, + "grad_norm": 6.912476539611816, + "learning_rate": 1.9991526876715043e-05, + "loss": 0.6669, + "step": 1356 + }, + { + "epoch": 0.042800820059927455, + "grad_norm": 5.52754020690918, + "learning_rate": 1.9991484780280106e-05, + "loss": 0.7057, + "step": 1357 + }, + { + "epoch": 0.04283236082636808, + "grad_norm": 3.7221977710723877, + "learning_rate": 1.9991442579576268e-05, + "loss": 0.3912, + "step": 1358 + }, + { + "epoch": 0.04286390159280871, + "grad_norm": 5.599247455596924, + "learning_rate": 1.9991400274603968e-05, + "loss": 0.6871, + "step": 1359 + }, + { + "epoch": 0.04289544235924933, + "grad_norm": 6.733753204345703, + "learning_rate": 1.9991357865363647e-05, + "loss": 0.6697, + "step": 1360 + }, + { + "epoch": 0.04292698312568995, + "grad_norm": 4.7127556800842285, + "learning_rate": 1.999131535185575e-05, + "loss": 0.683, + "step": 1361 + }, + { + "epoch": 0.042958523892130576, + "grad_norm": 5.325491905212402, + "learning_rate": 1.9991272734080717e-05, + "loss": 0.6605, + "step": 1362 + }, + { + "epoch": 0.042990064658571206, + "grad_norm": 5.406274318695068, + "learning_rate": 1.9991230012038992e-05, + "loss": 0.673, + "step": 1363 + }, + { + "epoch": 0.04302160542501183, + "grad_norm": 4.22442626953125, + "learning_rate": 1.9991187185731032e-05, + "loss": 0.6662, + "step": 1364 + }, + { + "epoch": 0.04305314619145245, + "grad_norm": 5.07682466506958, + "learning_rate": 1.999114425515727e-05, + "loss": 0.6504, + "step": 1365 + }, + { + "epoch": 0.043084686957893074, + "grad_norm": 9.788594245910645, + "learning_rate": 1.9991101220318164e-05, + "loss": 0.6725, + "step": 1366 + }, + { + "epoch": 0.043116227724333704, + "grad_norm": 7.198072910308838, + "learning_rate": 1.9991058081214153e-05, + "loss": 0.6667, + "step": 1367 + }, + { + "epoch": 0.04314776849077433, + "grad_norm": 3.5771310329437256, + "learning_rate": 1.99910148378457e-05, + "loss": 0.6583, + "step": 1368 + }, + { + "epoch": 0.04317930925721495, + "grad_norm": 4.1228814125061035, + "learning_rate": 1.9990971490213243e-05, + "loss": 0.6646, + "step": 1369 + }, + { + "epoch": 0.04321085002365557, + "grad_norm": 5.813457012176514, + "learning_rate": 1.9990928038317245e-05, + "loss": 0.6751, + "step": 1370 + }, + { + "epoch": 0.0432423907900962, + "grad_norm": 4.219492435455322, + "learning_rate": 1.9990884482158152e-05, + "loss": 0.6942, + "step": 1371 + }, + { + "epoch": 0.043273931556536825, + "grad_norm": 6.4036688804626465, + "learning_rate": 1.9990840821736425e-05, + "loss": 0.6316, + "step": 1372 + }, + { + "epoch": 0.04330547232297745, + "grad_norm": 4.311237335205078, + "learning_rate": 1.9990797057052514e-05, + "loss": 0.6929, + "step": 1373 + }, + { + "epoch": 0.04333701308941807, + "grad_norm": 4.664133071899414, + "learning_rate": 1.9990753188106878e-05, + "loss": 0.657, + "step": 1374 + }, + { + "epoch": 0.0433685538558587, + "grad_norm": 3.5304794311523438, + "learning_rate": 1.9990709214899976e-05, + "loss": 0.3779, + "step": 1375 + }, + { + "epoch": 0.04340009462229932, + "grad_norm": 4.216537952423096, + "learning_rate": 1.9990665137432264e-05, + "loss": 0.6529, + "step": 1376 + }, + { + "epoch": 0.043431635388739946, + "grad_norm": 4.321508884429932, + "learning_rate": 1.9990620955704204e-05, + "loss": 0.3988, + "step": 1377 + }, + { + "epoch": 0.04346317615518057, + "grad_norm": 10.688665390014648, + "learning_rate": 1.999057666971626e-05, + "loss": 0.7164, + "step": 1378 + }, + { + "epoch": 0.0434947169216212, + "grad_norm": 7.94484806060791, + "learning_rate": 1.999053227946889e-05, + "loss": 0.6864, + "step": 1379 + }, + { + "epoch": 0.04352625768806182, + "grad_norm": 11.60372257232666, + "learning_rate": 1.9990487784962555e-05, + "loss": 0.6406, + "step": 1380 + }, + { + "epoch": 0.043557798454502444, + "grad_norm": 5.925456523895264, + "learning_rate": 1.9990443186197725e-05, + "loss": 0.6838, + "step": 1381 + }, + { + "epoch": 0.04358933922094307, + "grad_norm": 8.54550838470459, + "learning_rate": 1.9990398483174863e-05, + "loss": 0.6784, + "step": 1382 + }, + { + "epoch": 0.04362087998738369, + "grad_norm": 6.481475353240967, + "learning_rate": 1.9990353675894435e-05, + "loss": 0.6786, + "step": 1383 + }, + { + "epoch": 0.04365242075382432, + "grad_norm": 5.2004499435424805, + "learning_rate": 1.999030876435691e-05, + "loss": 0.6693, + "step": 1384 + }, + { + "epoch": 0.04368396152026494, + "grad_norm": 6.752223014831543, + "learning_rate": 1.9990263748562754e-05, + "loss": 0.6649, + "step": 1385 + }, + { + "epoch": 0.043715502286705565, + "grad_norm": 4.191899299621582, + "learning_rate": 1.999021862851244e-05, + "loss": 0.6692, + "step": 1386 + }, + { + "epoch": 0.04374704305314619, + "grad_norm": 7.157429218292236, + "learning_rate": 1.9990173404206436e-05, + "loss": 0.6626, + "step": 1387 + }, + { + "epoch": 0.04377858381958682, + "grad_norm": 7.498638153076172, + "learning_rate": 1.9990128075645216e-05, + "loss": 0.633, + "step": 1388 + }, + { + "epoch": 0.04381012458602744, + "grad_norm": 14.226903915405273, + "learning_rate": 1.9990082642829253e-05, + "loss": 0.6762, + "step": 1389 + }, + { + "epoch": 0.043841665352468064, + "grad_norm": 5.266942977905273, + "learning_rate": 1.999003710575902e-05, + "loss": 0.6661, + "step": 1390 + }, + { + "epoch": 0.043873206118908686, + "grad_norm": 9.832511901855469, + "learning_rate": 1.998999146443499e-05, + "loss": 0.6289, + "step": 1391 + }, + { + "epoch": 0.043904746885349316, + "grad_norm": 4.508761405944824, + "learning_rate": 1.9989945718857648e-05, + "loss": 0.6645, + "step": 1392 + }, + { + "epoch": 0.04393628765178994, + "grad_norm": 4.0243401527404785, + "learning_rate": 1.9989899869027463e-05, + "loss": 0.6749, + "step": 1393 + }, + { + "epoch": 0.04396782841823056, + "grad_norm": 3.1836600303649902, + "learning_rate": 1.9989853914944913e-05, + "loss": 0.3764, + "step": 1394 + }, + { + "epoch": 0.043999369184671185, + "grad_norm": 5.769484996795654, + "learning_rate": 1.9989807856610483e-05, + "loss": 0.6348, + "step": 1395 + }, + { + "epoch": 0.044030909951111814, + "grad_norm": 3.4389162063598633, + "learning_rate": 1.9989761694024654e-05, + "loss": 0.679, + "step": 1396 + }, + { + "epoch": 0.04406245071755244, + "grad_norm": 4.904580593109131, + "learning_rate": 1.99897154271879e-05, + "loss": 0.6855, + "step": 1397 + }, + { + "epoch": 0.04409399148399306, + "grad_norm": 5.484129428863525, + "learning_rate": 1.998966905610071e-05, + "loss": 0.644, + "step": 1398 + }, + { + "epoch": 0.04412553225043368, + "grad_norm": 4.682531356811523, + "learning_rate": 1.998962258076357e-05, + "loss": 0.6337, + "step": 1399 + }, + { + "epoch": 0.04415707301687431, + "grad_norm": 4.557027339935303, + "learning_rate": 1.9989576001176956e-05, + "loss": 0.66, + "step": 1400 + }, + { + "epoch": 0.044188613783314935, + "grad_norm": 5.7408599853515625, + "learning_rate": 1.9989529317341363e-05, + "loss": 0.6676, + "step": 1401 + }, + { + "epoch": 0.04422015454975556, + "grad_norm": 4.090422630310059, + "learning_rate": 1.9989482529257275e-05, + "loss": 0.625, + "step": 1402 + }, + { + "epoch": 0.04425169531619618, + "grad_norm": 5.957653522491455, + "learning_rate": 1.998943563692518e-05, + "loss": 0.6403, + "step": 1403 + }, + { + "epoch": 0.04428323608263681, + "grad_norm": 4.969244956970215, + "learning_rate": 1.9989388640345567e-05, + "loss": 0.6828, + "step": 1404 + }, + { + "epoch": 0.044314776849077434, + "grad_norm": 2.054603338241577, + "learning_rate": 1.9989341539518924e-05, + "loss": 0.398, + "step": 1405 + }, + { + "epoch": 0.044346317615518056, + "grad_norm": 5.76601505279541, + "learning_rate": 1.998929433444575e-05, + "loss": 0.634, + "step": 1406 + }, + { + "epoch": 0.04437785838195868, + "grad_norm": 8.262084007263184, + "learning_rate": 1.998924702512653e-05, + "loss": 0.6544, + "step": 1407 + }, + { + "epoch": 0.04440939914839931, + "grad_norm": 4.143996238708496, + "learning_rate": 1.9989199611561764e-05, + "loss": 0.6614, + "step": 1408 + }, + { + "epoch": 0.04444093991483993, + "grad_norm": 4.049746513366699, + "learning_rate": 1.9989152093751945e-05, + "loss": 0.3564, + "step": 1409 + }, + { + "epoch": 0.044472480681280555, + "grad_norm": 6.0390305519104, + "learning_rate": 1.9989104471697563e-05, + "loss": 0.659, + "step": 1410 + }, + { + "epoch": 0.04450402144772118, + "grad_norm": 6.764182090759277, + "learning_rate": 1.9989056745399122e-05, + "loss": 0.6497, + "step": 1411 + }, + { + "epoch": 0.04453556221416181, + "grad_norm": 4.401701927185059, + "learning_rate": 1.9989008914857115e-05, + "loss": 0.6895, + "step": 1412 + }, + { + "epoch": 0.04456710298060243, + "grad_norm": 4.958024978637695, + "learning_rate": 1.9988960980072047e-05, + "loss": 0.6644, + "step": 1413 + }, + { + "epoch": 0.04459864374704305, + "grad_norm": 4.779354095458984, + "learning_rate": 1.9988912941044413e-05, + "loss": 0.6547, + "step": 1414 + }, + { + "epoch": 0.044630184513483676, + "grad_norm": 7.4023823738098145, + "learning_rate": 1.9988864797774718e-05, + "loss": 0.6329, + "step": 1415 + }, + { + "epoch": 0.044661725279924305, + "grad_norm": 2.6605093479156494, + "learning_rate": 1.9988816550263463e-05, + "loss": 0.901, + "step": 1416 + }, + { + "epoch": 0.04469326604636493, + "grad_norm": 2.4887783527374268, + "learning_rate": 1.9988768198511146e-05, + "loss": 0.8681, + "step": 1417 + }, + { + "epoch": 0.04472480681280555, + "grad_norm": 2.581831693649292, + "learning_rate": 1.998871974251828e-05, + "loss": 0.8898, + "step": 1418 + }, + { + "epoch": 0.044756347579246174, + "grad_norm": 5.532688140869141, + "learning_rate": 1.9988671182285368e-05, + "loss": 0.6698, + "step": 1419 + }, + { + "epoch": 0.044787888345686803, + "grad_norm": 10.178244590759277, + "learning_rate": 1.9988622517812915e-05, + "loss": 0.6678, + "step": 1420 + }, + { + "epoch": 0.044819429112127426, + "grad_norm": 4.991114616394043, + "learning_rate": 1.998857374910143e-05, + "loss": 0.6819, + "step": 1421 + }, + { + "epoch": 0.04485096987856805, + "grad_norm": 4.654178619384766, + "learning_rate": 1.9988524876151425e-05, + "loss": 0.6367, + "step": 1422 + }, + { + "epoch": 0.04488251064500867, + "grad_norm": 5.362782955169678, + "learning_rate": 1.9988475898963404e-05, + "loss": 0.6768, + "step": 1423 + }, + { + "epoch": 0.0449140514114493, + "grad_norm": 4.224438190460205, + "learning_rate": 1.9988426817537886e-05, + "loss": 0.6653, + "step": 1424 + }, + { + "epoch": 0.044945592177889925, + "grad_norm": 8.041529655456543, + "learning_rate": 1.9988377631875374e-05, + "loss": 0.6575, + "step": 1425 + }, + { + "epoch": 0.04497713294433055, + "grad_norm": 4.608438491821289, + "learning_rate": 1.9988328341976382e-05, + "loss": 0.6553, + "step": 1426 + }, + { + "epoch": 0.04500867371077117, + "grad_norm": 2.6114559173583984, + "learning_rate": 1.9988278947841435e-05, + "loss": 0.3462, + "step": 1427 + }, + { + "epoch": 0.04504021447721179, + "grad_norm": 7.595643043518066, + "learning_rate": 1.998822944947104e-05, + "loss": 0.7317, + "step": 1428 + }, + { + "epoch": 0.04507175524365242, + "grad_norm": 5.945670127868652, + "learning_rate": 1.998817984686571e-05, + "loss": 0.6977, + "step": 1429 + }, + { + "epoch": 0.045103296010093046, + "grad_norm": 3.4297432899475098, + "learning_rate": 1.9988130140025972e-05, + "loss": 0.3993, + "step": 1430 + }, + { + "epoch": 0.04513483677653367, + "grad_norm": 6.6978349685668945, + "learning_rate": 1.998808032895234e-05, + "loss": 0.6744, + "step": 1431 + }, + { + "epoch": 0.04516637754297429, + "grad_norm": 6.683973789215088, + "learning_rate": 1.9988030413645332e-05, + "loss": 0.6604, + "step": 1432 + }, + { + "epoch": 0.04519791830941492, + "grad_norm": 3.2776215076446533, + "learning_rate": 1.9987980394105478e-05, + "loss": 0.6317, + "step": 1433 + }, + { + "epoch": 0.045229459075855544, + "grad_norm": 4.893424987792969, + "learning_rate": 1.9987930270333287e-05, + "loss": 0.6432, + "step": 1434 + }, + { + "epoch": 0.045260999842296167, + "grad_norm": 4.336819171905518, + "learning_rate": 1.998788004232929e-05, + "loss": 0.902, + "step": 1435 + }, + { + "epoch": 0.04529254060873679, + "grad_norm": 9.723898887634277, + "learning_rate": 1.998782971009401e-05, + "loss": 0.6796, + "step": 1436 + }, + { + "epoch": 0.04532408137517742, + "grad_norm": 7.040660381317139, + "learning_rate": 1.998777927362797e-05, + "loss": 0.6839, + "step": 1437 + }, + { + "epoch": 0.04535562214161804, + "grad_norm": 4.8929901123046875, + "learning_rate": 1.99877287329317e-05, + "loss": 0.6957, + "step": 1438 + }, + { + "epoch": 0.045387162908058665, + "grad_norm": 3.5988447666168213, + "learning_rate": 1.998767808800572e-05, + "loss": 0.6677, + "step": 1439 + }, + { + "epoch": 0.04541870367449929, + "grad_norm": 4.0341010093688965, + "learning_rate": 1.998762733885057e-05, + "loss": 0.64, + "step": 1440 + }, + { + "epoch": 0.04545024444093992, + "grad_norm": 2.713111639022827, + "learning_rate": 1.998757648546677e-05, + "loss": 0.3844, + "step": 1441 + }, + { + "epoch": 0.04548178520738054, + "grad_norm": 8.606412887573242, + "learning_rate": 1.9987525527854856e-05, + "loss": 0.7001, + "step": 1442 + }, + { + "epoch": 0.04551332597382116, + "grad_norm": 6.998185634613037, + "learning_rate": 1.9987474466015356e-05, + "loss": 0.6515, + "step": 1443 + }, + { + "epoch": 0.045544866740261786, + "grad_norm": 3.9034883975982666, + "learning_rate": 1.9987423299948808e-05, + "loss": 0.6759, + "step": 1444 + }, + { + "epoch": 0.045576407506702415, + "grad_norm": 4.84599494934082, + "learning_rate": 1.998737202965574e-05, + "loss": 0.6054, + "step": 1445 + }, + { + "epoch": 0.04560794827314304, + "grad_norm": 4.3104071617126465, + "learning_rate": 1.9987320655136693e-05, + "loss": 0.9077, + "step": 1446 + }, + { + "epoch": 0.04563948903958366, + "grad_norm": 7.921546936035156, + "learning_rate": 1.99872691763922e-05, + "loss": 0.66, + "step": 1447 + }, + { + "epoch": 0.045671029806024284, + "grad_norm": 7.686931610107422, + "learning_rate": 1.9987217593422793e-05, + "loss": 0.6635, + "step": 1448 + }, + { + "epoch": 0.045702570572464914, + "grad_norm": 3.88737154006958, + "learning_rate": 1.998716590622902e-05, + "loss": 0.6287, + "step": 1449 + }, + { + "epoch": 0.045734111338905536, + "grad_norm": 7.667256832122803, + "learning_rate": 1.9987114114811415e-05, + "loss": 0.6546, + "step": 1450 + }, + { + "epoch": 0.04576565210534616, + "grad_norm": 3.0398764610290527, + "learning_rate": 1.998706221917052e-05, + "loss": 0.8786, + "step": 1451 + }, + { + "epoch": 0.04579719287178678, + "grad_norm": 4.387767791748047, + "learning_rate": 1.9987010219306873e-05, + "loss": 0.6532, + "step": 1452 + }, + { + "epoch": 0.04582873363822741, + "grad_norm": 7.696325302124023, + "learning_rate": 1.9986958115221023e-05, + "loss": 0.3708, + "step": 1453 + }, + { + "epoch": 0.045860274404668035, + "grad_norm": 6.850409984588623, + "learning_rate": 1.9986905906913508e-05, + "loss": 0.6924, + "step": 1454 + }, + { + "epoch": 0.04589181517110866, + "grad_norm": 5.713521957397461, + "learning_rate": 1.9986853594384877e-05, + "loss": 0.6874, + "step": 1455 + }, + { + "epoch": 0.04592335593754928, + "grad_norm": 8.262896537780762, + "learning_rate": 1.9986801177635673e-05, + "loss": 0.6515, + "step": 1456 + }, + { + "epoch": 0.04595489670398991, + "grad_norm": 9.118404388427734, + "learning_rate": 1.9986748656666447e-05, + "loss": 0.6612, + "step": 1457 + }, + { + "epoch": 0.04598643747043053, + "grad_norm": 7.3509650230407715, + "learning_rate": 1.998669603147774e-05, + "loss": 0.6722, + "step": 1458 + }, + { + "epoch": 0.046017978236871156, + "grad_norm": 4.758957862854004, + "learning_rate": 1.9986643302070106e-05, + "loss": 0.6708, + "step": 1459 + }, + { + "epoch": 0.04604951900331178, + "grad_norm": 4.519021511077881, + "learning_rate": 1.99865904684441e-05, + "loss": 0.915, + "step": 1460 + }, + { + "epoch": 0.04608105976975241, + "grad_norm": 4.911487102508545, + "learning_rate": 1.998653753060026e-05, + "loss": 0.6816, + "step": 1461 + }, + { + "epoch": 0.04611260053619303, + "grad_norm": 5.163975715637207, + "learning_rate": 1.9986484488539153e-05, + "loss": 0.6478, + "step": 1462 + }, + { + "epoch": 0.046144141302633654, + "grad_norm": 5.763762474060059, + "learning_rate": 1.998643134226132e-05, + "loss": 0.3948, + "step": 1463 + }, + { + "epoch": 0.04617568206907428, + "grad_norm": 4.147222518920898, + "learning_rate": 1.9986378091767328e-05, + "loss": 0.6823, + "step": 1464 + }, + { + "epoch": 0.046207222835514906, + "grad_norm": 6.059496879577637, + "learning_rate": 1.9986324737057724e-05, + "loss": 0.6682, + "step": 1465 + }, + { + "epoch": 0.04623876360195553, + "grad_norm": 5.265956878662109, + "learning_rate": 1.9986271278133066e-05, + "loss": 0.6389, + "step": 1466 + }, + { + "epoch": 0.04627030436839615, + "grad_norm": 4.806728363037109, + "learning_rate": 1.9986217714993916e-05, + "loss": 0.6798, + "step": 1467 + }, + { + "epoch": 0.046301845134836775, + "grad_norm": 5.299713134765625, + "learning_rate": 1.998616404764083e-05, + "loss": 0.8998, + "step": 1468 + }, + { + "epoch": 0.0463333859012774, + "grad_norm": 4.6153764724731445, + "learning_rate": 1.9986110276074366e-05, + "loss": 0.6884, + "step": 1469 + }, + { + "epoch": 0.04636492666771803, + "grad_norm": 5.0277099609375, + "learning_rate": 1.9986056400295087e-05, + "loss": 0.6818, + "step": 1470 + }, + { + "epoch": 0.04639646743415865, + "grad_norm": 3.9245519638061523, + "learning_rate": 1.9986002420303556e-05, + "loss": 0.8996, + "step": 1471 + }, + { + "epoch": 0.04642800820059927, + "grad_norm": 6.037971019744873, + "learning_rate": 1.9985948336100335e-05, + "loss": 0.6814, + "step": 1472 + }, + { + "epoch": 0.046459548967039896, + "grad_norm": 4.8159098625183105, + "learning_rate": 1.9985894147685993e-05, + "loss": 0.6443, + "step": 1473 + }, + { + "epoch": 0.046491089733480526, + "grad_norm": 12.358453750610352, + "learning_rate": 1.9985839855061086e-05, + "loss": 0.4266, + "step": 1474 + }, + { + "epoch": 0.04652263049992115, + "grad_norm": 3.7139790058135986, + "learning_rate": 1.998578545822619e-05, + "loss": 0.6115, + "step": 1475 + }, + { + "epoch": 0.04655417126636177, + "grad_norm": 5.504023551940918, + "learning_rate": 1.9985730957181867e-05, + "loss": 0.6488, + "step": 1476 + }, + { + "epoch": 0.046585712032802394, + "grad_norm": 5.189077854156494, + "learning_rate": 1.998567635192869e-05, + "loss": 0.6265, + "step": 1477 + }, + { + "epoch": 0.046617252799243024, + "grad_norm": 4.774097919464111, + "learning_rate": 1.9985621642467224e-05, + "loss": 0.6334, + "step": 1478 + }, + { + "epoch": 0.04664879356568365, + "grad_norm": 5.9014458656311035, + "learning_rate": 1.9985566828798043e-05, + "loss": 0.4001, + "step": 1479 + }, + { + "epoch": 0.04668033433212427, + "grad_norm": 4.79038143157959, + "learning_rate": 1.998551191092172e-05, + "loss": 0.6551, + "step": 1480 + }, + { + "epoch": 0.04671187509856489, + "grad_norm": 2.9077038764953613, + "learning_rate": 1.9985456888838824e-05, + "loss": 0.3747, + "step": 1481 + }, + { + "epoch": 0.04674341586500552, + "grad_norm": 5.432189464569092, + "learning_rate": 1.9985401762549935e-05, + "loss": 0.6721, + "step": 1482 + }, + { + "epoch": 0.046774956631446145, + "grad_norm": 4.691642761230469, + "learning_rate": 1.9985346532055624e-05, + "loss": 0.6571, + "step": 1483 + }, + { + "epoch": 0.04680649739788677, + "grad_norm": 4.58081579208374, + "learning_rate": 1.998529119735647e-05, + "loss": 0.666, + "step": 1484 + }, + { + "epoch": 0.04683803816432739, + "grad_norm": 4.638044834136963, + "learning_rate": 1.998523575845305e-05, + "loss": 0.6768, + "step": 1485 + }, + { + "epoch": 0.04686957893076802, + "grad_norm": 15.265854835510254, + "learning_rate": 1.9985180215345938e-05, + "loss": 0.6603, + "step": 1486 + }, + { + "epoch": 0.04690111969720864, + "grad_norm": 3.8056480884552, + "learning_rate": 1.998512456803572e-05, + "loss": 0.6702, + "step": 1487 + }, + { + "epoch": 0.046932660463649266, + "grad_norm": 4.216235160827637, + "learning_rate": 1.998506881652297e-05, + "loss": 0.6813, + "step": 1488 + }, + { + "epoch": 0.04696420123008989, + "grad_norm": 5.454033374786377, + "learning_rate": 1.9985012960808275e-05, + "loss": 0.6696, + "step": 1489 + }, + { + "epoch": 0.04699574199653052, + "grad_norm": 3.6954636573791504, + "learning_rate": 1.998495700089222e-05, + "loss": 0.6592, + "step": 1490 + }, + { + "epoch": 0.04702728276297114, + "grad_norm": 3.700111150741577, + "learning_rate": 1.9984900936775382e-05, + "loss": 0.6402, + "step": 1491 + }, + { + "epoch": 0.047058823529411764, + "grad_norm": 6.18770170211792, + "learning_rate": 1.9984844768458352e-05, + "loss": 0.644, + "step": 1492 + }, + { + "epoch": 0.04709036429585239, + "grad_norm": 7.385573863983154, + "learning_rate": 1.998478849594171e-05, + "loss": 0.649, + "step": 1493 + }, + { + "epoch": 0.04712190506229302, + "grad_norm": 4.947056770324707, + "learning_rate": 1.998473211922605e-05, + "loss": 0.6496, + "step": 1494 + }, + { + "epoch": 0.04715344582873364, + "grad_norm": 7.974113464355469, + "learning_rate": 1.9984675638311957e-05, + "loss": 0.644, + "step": 1495 + }, + { + "epoch": 0.04718498659517426, + "grad_norm": 4.667233943939209, + "learning_rate": 1.9984619053200022e-05, + "loss": 0.6305, + "step": 1496 + }, + { + "epoch": 0.047216527361614885, + "grad_norm": 6.720061302185059, + "learning_rate": 1.9984562363890834e-05, + "loss": 0.6566, + "step": 1497 + }, + { + "epoch": 0.047248068128055515, + "grad_norm": 5.112816333770752, + "learning_rate": 1.9984505570384982e-05, + "loss": 0.6266, + "step": 1498 + }, + { + "epoch": 0.04727960889449614, + "grad_norm": 5.00175142288208, + "learning_rate": 1.9984448672683066e-05, + "loss": 0.6568, + "step": 1499 + }, + { + "epoch": 0.04731114966093676, + "grad_norm": 6.006879806518555, + "learning_rate": 1.9984391670785672e-05, + "loss": 0.6727, + "step": 1500 + }, + { + "epoch": 0.04734269042737738, + "grad_norm": 4.877772331237793, + "learning_rate": 1.99843345646934e-05, + "loss": 0.6308, + "step": 1501 + }, + { + "epoch": 0.04737423119381801, + "grad_norm": 2.7358791828155518, + "learning_rate": 1.9984277354406844e-05, + "loss": 0.9296, + "step": 1502 + }, + { + "epoch": 0.047405771960258636, + "grad_norm": 4.651752948760986, + "learning_rate": 1.9984220039926598e-05, + "loss": 0.6539, + "step": 1503 + }, + { + "epoch": 0.04743731272669926, + "grad_norm": 4.490744113922119, + "learning_rate": 1.9984162621253266e-05, + "loss": 0.6299, + "step": 1504 + }, + { + "epoch": 0.04746885349313988, + "grad_norm": 11.908491134643555, + "learning_rate": 1.9984105098387445e-05, + "loss": 0.6464, + "step": 1505 + }, + { + "epoch": 0.04750039425958051, + "grad_norm": 4.31363582611084, + "learning_rate": 1.9984047471329738e-05, + "loss": 0.6423, + "step": 1506 + }, + { + "epoch": 0.047531935026021134, + "grad_norm": 8.419206619262695, + "learning_rate": 1.9983989740080737e-05, + "loss": 0.6828, + "step": 1507 + }, + { + "epoch": 0.04756347579246176, + "grad_norm": 4.4298858642578125, + "learning_rate": 1.9983931904641056e-05, + "loss": 0.635, + "step": 1508 + }, + { + "epoch": 0.04759501655890238, + "grad_norm": 4.70805549621582, + "learning_rate": 1.9983873965011287e-05, + "loss": 0.656, + "step": 1509 + }, + { + "epoch": 0.047626557325343, + "grad_norm": 10.81136703491211, + "learning_rate": 1.9983815921192048e-05, + "loss": 0.6264, + "step": 1510 + }, + { + "epoch": 0.04765809809178363, + "grad_norm": 3.971446990966797, + "learning_rate": 1.998375777318393e-05, + "loss": 0.6498, + "step": 1511 + }, + { + "epoch": 0.047689638858224255, + "grad_norm": 8.3983154296875, + "learning_rate": 1.9983699520987554e-05, + "loss": 0.6382, + "step": 1512 + }, + { + "epoch": 0.04772117962466488, + "grad_norm": 6.069952011108398, + "learning_rate": 1.9983641164603518e-05, + "loss": 0.8926, + "step": 1513 + }, + { + "epoch": 0.0477527203911055, + "grad_norm": 4.505650043487549, + "learning_rate": 1.9983582704032437e-05, + "loss": 0.6738, + "step": 1514 + }, + { + "epoch": 0.04778426115754613, + "grad_norm": 5.866840839385986, + "learning_rate": 1.9983524139274915e-05, + "loss": 0.6328, + "step": 1515 + }, + { + "epoch": 0.04781580192398675, + "grad_norm": 4.904523849487305, + "learning_rate": 1.9983465470331568e-05, + "loss": 0.6531, + "step": 1516 + }, + { + "epoch": 0.047847342690427376, + "grad_norm": 4.841870307922363, + "learning_rate": 1.9983406697203006e-05, + "loss": 0.6508, + "step": 1517 + }, + { + "epoch": 0.047878883456868, + "grad_norm": 5.139260292053223, + "learning_rate": 1.9983347819889845e-05, + "loss": 0.6313, + "step": 1518 + }, + { + "epoch": 0.04791042422330863, + "grad_norm": 4.147885799407959, + "learning_rate": 1.99832888383927e-05, + "loss": 0.659, + "step": 1519 + }, + { + "epoch": 0.04794196498974925, + "grad_norm": 4.226701259613037, + "learning_rate": 1.9983229752712178e-05, + "loss": 0.6329, + "step": 1520 + }, + { + "epoch": 0.047973505756189874, + "grad_norm": 6.460137844085693, + "learning_rate": 1.9983170562848908e-05, + "loss": 0.6698, + "step": 1521 + }, + { + "epoch": 0.0480050465226305, + "grad_norm": 4.311159610748291, + "learning_rate": 1.9983111268803495e-05, + "loss": 0.6495, + "step": 1522 + }, + { + "epoch": 0.04803658728907113, + "grad_norm": 8.3789701461792, + "learning_rate": 1.9983051870576568e-05, + "loss": 0.6712, + "step": 1523 + }, + { + "epoch": 0.04806812805551175, + "grad_norm": 3.549029588699341, + "learning_rate": 1.9982992368168742e-05, + "loss": 0.959, + "step": 1524 + }, + { + "epoch": 0.04809966882195237, + "grad_norm": 3.9498231410980225, + "learning_rate": 1.998293276158064e-05, + "loss": 0.6464, + "step": 1525 + }, + { + "epoch": 0.048131209588392995, + "grad_norm": 5.739779949188232, + "learning_rate": 1.9982873050812884e-05, + "loss": 0.6428, + "step": 1526 + }, + { + "epoch": 0.048162750354833625, + "grad_norm": 4.890597343444824, + "learning_rate": 1.9982813235866096e-05, + "loss": 0.6906, + "step": 1527 + }, + { + "epoch": 0.04819429112127425, + "grad_norm": 4.741819858551025, + "learning_rate": 1.9982753316740895e-05, + "loss": 0.6465, + "step": 1528 + }, + { + "epoch": 0.04822583188771487, + "grad_norm": 2.460392475128174, + "learning_rate": 1.998269329343792e-05, + "loss": 0.8832, + "step": 1529 + }, + { + "epoch": 0.04825737265415549, + "grad_norm": 8.531448364257812, + "learning_rate": 1.9982633165957784e-05, + "loss": 0.627, + "step": 1530 + }, + { + "epoch": 0.04828891342059612, + "grad_norm": 6.18242073059082, + "learning_rate": 1.998257293430112e-05, + "loss": 0.6738, + "step": 1531 + }, + { + "epoch": 0.048320454187036746, + "grad_norm": 5.146021366119385, + "learning_rate": 1.998251259846856e-05, + "loss": 0.6273, + "step": 1532 + }, + { + "epoch": 0.04835199495347737, + "grad_norm": 4.244805335998535, + "learning_rate": 1.9982452158460724e-05, + "loss": 0.6667, + "step": 1533 + }, + { + "epoch": 0.04838353571991799, + "grad_norm": 4.299344539642334, + "learning_rate": 1.9982391614278253e-05, + "loss": 0.6173, + "step": 1534 + }, + { + "epoch": 0.04841507648635862, + "grad_norm": 2.7656118869781494, + "learning_rate": 1.9982330965921776e-05, + "loss": 0.8954, + "step": 1535 + }, + { + "epoch": 0.048446617252799244, + "grad_norm": 4.528417587280273, + "learning_rate": 1.9982270213391917e-05, + "loss": 0.6453, + "step": 1536 + }, + { + "epoch": 0.04847815801923987, + "grad_norm": 3.5850722789764404, + "learning_rate": 1.9982209356689323e-05, + "loss": 0.6292, + "step": 1537 + }, + { + "epoch": 0.04850969878568049, + "grad_norm": 6.307436466217041, + "learning_rate": 1.998214839581462e-05, + "loss": 0.66, + "step": 1538 + }, + { + "epoch": 0.04854123955212112, + "grad_norm": 4.32699728012085, + "learning_rate": 1.998208733076845e-05, + "loss": 0.6508, + "step": 1539 + }, + { + "epoch": 0.04857278031856174, + "grad_norm": 4.051313877105713, + "learning_rate": 1.998202616155145e-05, + "loss": 0.6694, + "step": 1540 + }, + { + "epoch": 0.048604321085002365, + "grad_norm": 4.595232963562012, + "learning_rate": 1.9981964888164248e-05, + "loss": 0.6521, + "step": 1541 + }, + { + "epoch": 0.04863586185144299, + "grad_norm": 4.463140487670898, + "learning_rate": 1.99819035106075e-05, + "loss": 0.6465, + "step": 1542 + }, + { + "epoch": 0.04866740261788362, + "grad_norm": 4.494663238525391, + "learning_rate": 1.9981842028881835e-05, + "loss": 0.6461, + "step": 1543 + }, + { + "epoch": 0.04869894338432424, + "grad_norm": 3.6008174419403076, + "learning_rate": 1.99817804429879e-05, + "loss": 0.6483, + "step": 1544 + }, + { + "epoch": 0.04873048415076486, + "grad_norm": 14.98027515411377, + "learning_rate": 1.9981718752926328e-05, + "loss": 0.6694, + "step": 1545 + }, + { + "epoch": 0.048762024917205486, + "grad_norm": 4.3974690437316895, + "learning_rate": 1.9981656958697777e-05, + "loss": 0.6342, + "step": 1546 + }, + { + "epoch": 0.048793565683646116, + "grad_norm": 5.240350246429443, + "learning_rate": 1.9981595060302883e-05, + "loss": 0.6304, + "step": 1547 + }, + { + "epoch": 0.04882510645008674, + "grad_norm": 9.373303413391113, + "learning_rate": 1.9981533057742296e-05, + "loss": 0.6719, + "step": 1548 + }, + { + "epoch": 0.04885664721652736, + "grad_norm": 5.250371932983398, + "learning_rate": 1.998147095101666e-05, + "loss": 0.6361, + "step": 1549 + }, + { + "epoch": 0.048888187982967984, + "grad_norm": 4.283304214477539, + "learning_rate": 1.9981408740126622e-05, + "loss": 0.6662, + "step": 1550 + }, + { + "epoch": 0.04891972874940861, + "grad_norm": 8.569212913513184, + "learning_rate": 1.9981346425072836e-05, + "loss": 0.6515, + "step": 1551 + }, + { + "epoch": 0.04895126951584924, + "grad_norm": 3.784226894378662, + "learning_rate": 1.998128400585595e-05, + "loss": 0.6486, + "step": 1552 + }, + { + "epoch": 0.04898281028228986, + "grad_norm": 5.126230239868164, + "learning_rate": 1.9981221482476613e-05, + "loss": 0.6326, + "step": 1553 + }, + { + "epoch": 0.04901435104873048, + "grad_norm": 2.9173970222473145, + "learning_rate": 1.998115885493548e-05, + "loss": 0.38, + "step": 1554 + }, + { + "epoch": 0.049045891815171105, + "grad_norm": 7.224128246307373, + "learning_rate": 1.9981096123233207e-05, + "loss": 0.658, + "step": 1555 + }, + { + "epoch": 0.049077432581611735, + "grad_norm": 7.003859043121338, + "learning_rate": 1.9981033287370443e-05, + "loss": 0.6883, + "step": 1556 + }, + { + "epoch": 0.04910897334805236, + "grad_norm": 4.5963592529296875, + "learning_rate": 1.9980970347347844e-05, + "loss": 0.6425, + "step": 1557 + }, + { + "epoch": 0.04914051411449298, + "grad_norm": 12.428409576416016, + "learning_rate": 1.9980907303166075e-05, + "loss": 0.6743, + "step": 1558 + }, + { + "epoch": 0.049172054880933604, + "grad_norm": 3.941150426864624, + "learning_rate": 1.998084415482579e-05, + "loss": 0.63, + "step": 1559 + }, + { + "epoch": 0.04920359564737423, + "grad_norm": 6.221706867218018, + "learning_rate": 1.998078090232764e-05, + "loss": 0.6373, + "step": 1560 + }, + { + "epoch": 0.049235136413814856, + "grad_norm": 6.615051746368408, + "learning_rate": 1.9980717545672292e-05, + "loss": 0.6379, + "step": 1561 + }, + { + "epoch": 0.04926667718025548, + "grad_norm": 3.0864763259887695, + "learning_rate": 1.998065408486041e-05, + "loss": 0.3939, + "step": 1562 + }, + { + "epoch": 0.0492982179466961, + "grad_norm": 5.904143333435059, + "learning_rate": 1.998059051989265e-05, + "loss": 0.6439, + "step": 1563 + }, + { + "epoch": 0.04932975871313673, + "grad_norm": 5.728363513946533, + "learning_rate": 1.9980526850769678e-05, + "loss": 0.6547, + "step": 1564 + }, + { + "epoch": 0.049361299479577354, + "grad_norm": 4.857386589050293, + "learning_rate": 1.998046307749216e-05, + "loss": 0.6362, + "step": 1565 + }, + { + "epoch": 0.04939284024601798, + "grad_norm": 4.830131530761719, + "learning_rate": 1.9980399200060763e-05, + "loss": 0.6328, + "step": 1566 + }, + { + "epoch": 0.0494243810124586, + "grad_norm": 2.8538742065429688, + "learning_rate": 1.9980335218476145e-05, + "loss": 0.8968, + "step": 1567 + }, + { + "epoch": 0.04945592177889923, + "grad_norm": 3.5554730892181396, + "learning_rate": 1.9980271132738984e-05, + "loss": 0.3696, + "step": 1568 + }, + { + "epoch": 0.04948746254533985, + "grad_norm": 12.087048530578613, + "learning_rate": 1.9980206942849943e-05, + "loss": 0.681, + "step": 1569 + }, + { + "epoch": 0.049519003311780475, + "grad_norm": 9.229375839233398, + "learning_rate": 1.9980142648809694e-05, + "loss": 0.706, + "step": 1570 + }, + { + "epoch": 0.0495505440782211, + "grad_norm": 4.434086799621582, + "learning_rate": 1.9980078250618904e-05, + "loss": 0.6633, + "step": 1571 + }, + { + "epoch": 0.04958208484466173, + "grad_norm": 9.00700569152832, + "learning_rate": 1.9980013748278247e-05, + "loss": 0.6578, + "step": 1572 + }, + { + "epoch": 0.04961362561110235, + "grad_norm": 5.947846412658691, + "learning_rate": 1.99799491417884e-05, + "loss": 0.6609, + "step": 1573 + }, + { + "epoch": 0.049645166377542974, + "grad_norm": 6.394443988800049, + "learning_rate": 1.9979884431150037e-05, + "loss": 0.6294, + "step": 1574 + }, + { + "epoch": 0.049676707143983596, + "grad_norm": 9.027429580688477, + "learning_rate": 1.997981961636383e-05, + "loss": 0.9434, + "step": 1575 + }, + { + "epoch": 0.049708247910424226, + "grad_norm": 5.480592727661133, + "learning_rate": 1.9979754697430455e-05, + "loss": 0.6581, + "step": 1576 + }, + { + "epoch": 0.04973978867686485, + "grad_norm": 5.237136363983154, + "learning_rate": 1.9979689674350588e-05, + "loss": 0.6362, + "step": 1577 + }, + { + "epoch": 0.04977132944330547, + "grad_norm": 2.8706741333007812, + "learning_rate": 1.9979624547124913e-05, + "loss": 0.3441, + "step": 1578 + }, + { + "epoch": 0.049802870209746095, + "grad_norm": 4.5766282081604, + "learning_rate": 1.997955931575411e-05, + "loss": 0.6475, + "step": 1579 + }, + { + "epoch": 0.049834410976186724, + "grad_norm": 9.173072814941406, + "learning_rate": 1.9979493980238848e-05, + "loss": 0.673, + "step": 1580 + }, + { + "epoch": 0.04986595174262735, + "grad_norm": 5.300740718841553, + "learning_rate": 1.9979428540579822e-05, + "loss": 0.6781, + "step": 1581 + }, + { + "epoch": 0.04989749250906797, + "grad_norm": 5.1244001388549805, + "learning_rate": 1.9979362996777713e-05, + "loss": 0.6607, + "step": 1582 + }, + { + "epoch": 0.04992903327550859, + "grad_norm": 6.615152835845947, + "learning_rate": 1.99792973488332e-05, + "loss": 0.634, + "step": 1583 + }, + { + "epoch": 0.04996057404194922, + "grad_norm": 5.386518478393555, + "learning_rate": 1.997923159674697e-05, + "loss": 0.6401, + "step": 1584 + }, + { + "epoch": 0.049992114808389845, + "grad_norm": 3.911024570465088, + "learning_rate": 1.997916574051971e-05, + "loss": 0.8977, + "step": 1585 + }, + { + "epoch": 0.05002365557483047, + "grad_norm": 4.746402263641357, + "learning_rate": 1.9979099780152107e-05, + "loss": 0.633, + "step": 1586 + }, + { + "epoch": 0.05005519634127109, + "grad_norm": 2.5405027866363525, + "learning_rate": 1.9979033715644852e-05, + "loss": 0.3959, + "step": 1587 + }, + { + "epoch": 0.05008673710771172, + "grad_norm": 15.074835777282715, + "learning_rate": 1.9978967546998627e-05, + "loss": 0.7002, + "step": 1588 + }, + { + "epoch": 0.050118277874152344, + "grad_norm": 4.344720363616943, + "learning_rate": 1.997890127421413e-05, + "loss": 0.6204, + "step": 1589 + }, + { + "epoch": 0.050149818640592966, + "grad_norm": 4.012469291687012, + "learning_rate": 1.997883489729205e-05, + "loss": 0.6406, + "step": 1590 + }, + { + "epoch": 0.05018135940703359, + "grad_norm": 4.996235370635986, + "learning_rate": 1.9978768416233076e-05, + "loss": 0.6295, + "step": 1591 + }, + { + "epoch": 0.05021290017347421, + "grad_norm": 4.649788856506348, + "learning_rate": 1.9978701831037907e-05, + "loss": 0.6625, + "step": 1592 + }, + { + "epoch": 0.05024444093991484, + "grad_norm": 5.068787097930908, + "learning_rate": 1.9978635141707238e-05, + "loss": 0.592, + "step": 1593 + }, + { + "epoch": 0.050275981706355465, + "grad_norm": 4.056304454803467, + "learning_rate": 1.997856834824176e-05, + "loss": 0.672, + "step": 1594 + }, + { + "epoch": 0.05030752247279609, + "grad_norm": 10.433248519897461, + "learning_rate": 1.9978501450642177e-05, + "loss": 0.6581, + "step": 1595 + }, + { + "epoch": 0.05033906323923671, + "grad_norm": 6.2462921142578125, + "learning_rate": 1.997843444890918e-05, + "loss": 0.649, + "step": 1596 + }, + { + "epoch": 0.05037060400567734, + "grad_norm": 3.3605668544769287, + "learning_rate": 1.9978367343043473e-05, + "loss": 0.3637, + "step": 1597 + }, + { + "epoch": 0.05040214477211796, + "grad_norm": 3.382861614227295, + "learning_rate": 1.9978300133045754e-05, + "loss": 0.6833, + "step": 1598 + }, + { + "epoch": 0.050433685538558586, + "grad_norm": 7.092401504516602, + "learning_rate": 1.9978232818916727e-05, + "loss": 0.9243, + "step": 1599 + }, + { + "epoch": 0.05046522630499921, + "grad_norm": 4.195874214172363, + "learning_rate": 1.997816540065709e-05, + "loss": 0.6578, + "step": 1600 + }, + { + "epoch": 0.05049676707143984, + "grad_norm": 3.9126062393188477, + "learning_rate": 1.9978097878267554e-05, + "loss": 0.6382, + "step": 1601 + }, + { + "epoch": 0.05052830783788046, + "grad_norm": 3.921987295150757, + "learning_rate": 1.9978030251748813e-05, + "loss": 0.67, + "step": 1602 + }, + { + "epoch": 0.050559848604321084, + "grad_norm": 4.246300220489502, + "learning_rate": 1.9977962521101578e-05, + "loss": 0.6358, + "step": 1603 + }, + { + "epoch": 0.05059138937076171, + "grad_norm": 5.868379592895508, + "learning_rate": 1.9977894686326563e-05, + "loss": 0.6367, + "step": 1604 + }, + { + "epoch": 0.050622930137202336, + "grad_norm": 4.404080867767334, + "learning_rate": 1.9977826747424466e-05, + "loss": 0.6338, + "step": 1605 + }, + { + "epoch": 0.05065447090364296, + "grad_norm": 6.581651210784912, + "learning_rate": 1.9977758704395996e-05, + "loss": 0.6401, + "step": 1606 + }, + { + "epoch": 0.05068601167008358, + "grad_norm": 3.707873582839966, + "learning_rate": 1.997769055724187e-05, + "loss": 0.6282, + "step": 1607 + }, + { + "epoch": 0.050717552436524205, + "grad_norm": 5.661443710327148, + "learning_rate": 1.9977622305962795e-05, + "loss": 0.3732, + "step": 1608 + }, + { + "epoch": 0.050749093202964835, + "grad_norm": 4.9515228271484375, + "learning_rate": 1.9977553950559484e-05, + "loss": 0.6737, + "step": 1609 + }, + { + "epoch": 0.05078063396940546, + "grad_norm": 5.030946254730225, + "learning_rate": 1.997748549103265e-05, + "loss": 0.3694, + "step": 1610 + }, + { + "epoch": 0.05081217473584608, + "grad_norm": 5.120782852172852, + "learning_rate": 1.9977416927383008e-05, + "loss": 0.6604, + "step": 1611 + }, + { + "epoch": 0.0508437155022867, + "grad_norm": 5.185697555541992, + "learning_rate": 1.9977348259611272e-05, + "loss": 0.649, + "step": 1612 + }, + { + "epoch": 0.05087525626872733, + "grad_norm": 5.065742492675781, + "learning_rate": 1.997727948771816e-05, + "loss": 0.6735, + "step": 1613 + }, + { + "epoch": 0.050906797035167956, + "grad_norm": 4.5723772048950195, + "learning_rate": 1.997721061170439e-05, + "loss": 0.6464, + "step": 1614 + }, + { + "epoch": 0.05093833780160858, + "grad_norm": 5.468202590942383, + "learning_rate": 1.997714163157068e-05, + "loss": 0.6682, + "step": 1615 + }, + { + "epoch": 0.0509698785680492, + "grad_norm": 5.134338855743408, + "learning_rate": 1.997707254731775e-05, + "loss": 0.6415, + "step": 1616 + }, + { + "epoch": 0.05100141933448983, + "grad_norm": 9.312695503234863, + "learning_rate": 1.9977003358946322e-05, + "loss": 0.6699, + "step": 1617 + }, + { + "epoch": 0.051032960100930454, + "grad_norm": 4.024394989013672, + "learning_rate": 1.9976934066457116e-05, + "loss": 0.6511, + "step": 1618 + }, + { + "epoch": 0.051064500867371077, + "grad_norm": 3.947557210922241, + "learning_rate": 1.9976864669850853e-05, + "loss": 0.6478, + "step": 1619 + }, + { + "epoch": 0.0510960416338117, + "grad_norm": 5.271031856536865, + "learning_rate": 1.9976795169128266e-05, + "loss": 0.683, + "step": 1620 + }, + { + "epoch": 0.05112758240025233, + "grad_norm": 4.831076622009277, + "learning_rate": 1.9976725564290073e-05, + "loss": 0.6371, + "step": 1621 + }, + { + "epoch": 0.05115912316669295, + "grad_norm": 4.559508323669434, + "learning_rate": 1.9976655855337e-05, + "loss": 0.652, + "step": 1622 + }, + { + "epoch": 0.051190663933133575, + "grad_norm": 5.109986782073975, + "learning_rate": 1.9976586042269776e-05, + "loss": 0.6457, + "step": 1623 + }, + { + "epoch": 0.0512222046995742, + "grad_norm": 5.513782024383545, + "learning_rate": 1.997651612508913e-05, + "loss": 0.647, + "step": 1624 + }, + { + "epoch": 0.05125374546601483, + "grad_norm": 4.970577239990234, + "learning_rate": 1.9976446103795793e-05, + "loss": 0.3702, + "step": 1625 + }, + { + "epoch": 0.05128528623245545, + "grad_norm": 5.577104568481445, + "learning_rate": 1.9976375978390497e-05, + "loss": 0.7023, + "step": 1626 + }, + { + "epoch": 0.05131682699889607, + "grad_norm": 4.906777858734131, + "learning_rate": 1.997630574887397e-05, + "loss": 0.6677, + "step": 1627 + }, + { + "epoch": 0.051348367765336696, + "grad_norm": 2.5004403591156006, + "learning_rate": 1.997623541524694e-05, + "loss": 0.9263, + "step": 1628 + }, + { + "epoch": 0.051379908531777325, + "grad_norm": 3.7664456367492676, + "learning_rate": 1.9976164977510155e-05, + "loss": 0.6583, + "step": 1629 + }, + { + "epoch": 0.05141144929821795, + "grad_norm": 3.9081268310546875, + "learning_rate": 1.9976094435664336e-05, + "loss": 0.6441, + "step": 1630 + }, + { + "epoch": 0.05144299006465857, + "grad_norm": 5.86576509475708, + "learning_rate": 1.997602378971023e-05, + "loss": 0.6591, + "step": 1631 + }, + { + "epoch": 0.051474530831099194, + "grad_norm": 4.005279064178467, + "learning_rate": 1.9975953039648567e-05, + "loss": 0.6532, + "step": 1632 + }, + { + "epoch": 0.051506071597539824, + "grad_norm": 4.022108554840088, + "learning_rate": 1.997588218548009e-05, + "loss": 0.6413, + "step": 1633 + }, + { + "epoch": 0.051537612363980446, + "grad_norm": 3.6314499378204346, + "learning_rate": 1.9975811227205536e-05, + "loss": 0.662, + "step": 1634 + }, + { + "epoch": 0.05156915313042107, + "grad_norm": 3.43902850151062, + "learning_rate": 1.9975740164825644e-05, + "loss": 0.6662, + "step": 1635 + }, + { + "epoch": 0.05160069389686169, + "grad_norm": 6.8566694259643555, + "learning_rate": 1.9975668998341158e-05, + "loss": 0.9325, + "step": 1636 + }, + { + "epoch": 0.051632234663302315, + "grad_norm": 3.539515972137451, + "learning_rate": 1.997559772775282e-05, + "loss": 0.6748, + "step": 1637 + }, + { + "epoch": 0.051663775429742945, + "grad_norm": 4.660343170166016, + "learning_rate": 1.9975526353061374e-05, + "loss": 0.6621, + "step": 1638 + }, + { + "epoch": 0.05169531619618357, + "grad_norm": 8.882336616516113, + "learning_rate": 1.9975454874267567e-05, + "loss": 0.6506, + "step": 1639 + }, + { + "epoch": 0.05172685696262419, + "grad_norm": 3.7272510528564453, + "learning_rate": 1.9975383291372143e-05, + "loss": 0.6391, + "step": 1640 + }, + { + "epoch": 0.05175839772906481, + "grad_norm": 3.900836229324341, + "learning_rate": 1.997531160437584e-05, + "loss": 0.644, + "step": 1641 + }, + { + "epoch": 0.05178993849550544, + "grad_norm": 5.953726768493652, + "learning_rate": 1.9975239813279424e-05, + "loss": 0.6417, + "step": 1642 + }, + { + "epoch": 0.051821479261946066, + "grad_norm": 4.819821834564209, + "learning_rate": 1.9975167918083634e-05, + "loss": 0.6224, + "step": 1643 + }, + { + "epoch": 0.05185302002838669, + "grad_norm": 12.982285499572754, + "learning_rate": 1.997509591878922e-05, + "loss": 0.6611, + "step": 1644 + }, + { + "epoch": 0.05188456079482731, + "grad_norm": 2.651540756225586, + "learning_rate": 1.997502381539693e-05, + "loss": 0.908, + "step": 1645 + }, + { + "epoch": 0.05191610156126794, + "grad_norm": 4.2859787940979, + "learning_rate": 1.9974951607907527e-05, + "loss": 0.6724, + "step": 1646 + }, + { + "epoch": 0.051947642327708564, + "grad_norm": 7.1808247566223145, + "learning_rate": 1.9974879296321758e-05, + "loss": 0.6435, + "step": 1647 + }, + { + "epoch": 0.05197918309414919, + "grad_norm": 3.4232394695281982, + "learning_rate": 1.9974806880640376e-05, + "loss": 0.6261, + "step": 1648 + }, + { + "epoch": 0.05201072386058981, + "grad_norm": 4.420074462890625, + "learning_rate": 1.997473436086414e-05, + "loss": 0.6296, + "step": 1649 + }, + { + "epoch": 0.05204226462703044, + "grad_norm": 5.914814472198486, + "learning_rate": 1.9974661736993806e-05, + "loss": 0.6972, + "step": 1650 + }, + { + "epoch": 0.05207380539347106, + "grad_norm": 4.624868869781494, + "learning_rate": 1.9974589009030134e-05, + "loss": 0.6741, + "step": 1651 + }, + { + "epoch": 0.052105346159911685, + "grad_norm": 5.23649263381958, + "learning_rate": 1.9974516176973874e-05, + "loss": 0.6553, + "step": 1652 + }, + { + "epoch": 0.05213688692635231, + "grad_norm": 4.256595611572266, + "learning_rate": 1.99744432408258e-05, + "loss": 0.6516, + "step": 1653 + }, + { + "epoch": 0.05216842769279294, + "grad_norm": 4.320678234100342, + "learning_rate": 1.9974370200586663e-05, + "loss": 0.6551, + "step": 1654 + }, + { + "epoch": 0.05219996845923356, + "grad_norm": 4.191253662109375, + "learning_rate": 1.9974297056257223e-05, + "loss": 0.6251, + "step": 1655 + }, + { + "epoch": 0.05223150922567418, + "grad_norm": 2.844726800918579, + "learning_rate": 1.9974223807838255e-05, + "loss": 0.8801, + "step": 1656 + }, + { + "epoch": 0.052263049992114806, + "grad_norm": 3.881746292114258, + "learning_rate": 1.9974150455330515e-05, + "loss": 0.6127, + "step": 1657 + }, + { + "epoch": 0.052294590758555436, + "grad_norm": 5.268332481384277, + "learning_rate": 1.997407699873477e-05, + "loss": 0.6303, + "step": 1658 + }, + { + "epoch": 0.05232613152499606, + "grad_norm": 8.680971145629883, + "learning_rate": 1.9974003438051786e-05, + "loss": 0.6262, + "step": 1659 + }, + { + "epoch": 0.05235767229143668, + "grad_norm": 7.872721195220947, + "learning_rate": 1.9973929773282332e-05, + "loss": 0.6467, + "step": 1660 + }, + { + "epoch": 0.052389213057877304, + "grad_norm": 5.9080891609191895, + "learning_rate": 1.9973856004427175e-05, + "loss": 0.6376, + "step": 1661 + }, + { + "epoch": 0.052420753824317934, + "grad_norm": 4.194475173950195, + "learning_rate": 1.997378213148709e-05, + "loss": 0.6383, + "step": 1662 + }, + { + "epoch": 0.05245229459075856, + "grad_norm": 4.494786739349365, + "learning_rate": 1.997370815446284e-05, + "loss": 0.6412, + "step": 1663 + }, + { + "epoch": 0.05248383535719918, + "grad_norm": 5.364657402038574, + "learning_rate": 1.9973634073355206e-05, + "loss": 0.941, + "step": 1664 + }, + { + "epoch": 0.0525153761236398, + "grad_norm": 5.423058986663818, + "learning_rate": 1.997355988816495e-05, + "loss": 0.6529, + "step": 1665 + }, + { + "epoch": 0.05254691689008043, + "grad_norm": 5.589883327484131, + "learning_rate": 1.9973485598892857e-05, + "loss": 0.6649, + "step": 1666 + }, + { + "epoch": 0.052578457656521055, + "grad_norm": 2.4169983863830566, + "learning_rate": 1.9973411205539695e-05, + "loss": 0.9026, + "step": 1667 + }, + { + "epoch": 0.05260999842296168, + "grad_norm": 4.618585586547852, + "learning_rate": 1.9973336708106245e-05, + "loss": 0.6368, + "step": 1668 + }, + { + "epoch": 0.0526415391894023, + "grad_norm": 24.13979148864746, + "learning_rate": 1.9973262106593282e-05, + "loss": 0.4115, + "step": 1669 + }, + { + "epoch": 0.05267307995584293, + "grad_norm": 7.126575946807861, + "learning_rate": 1.997318740100158e-05, + "loss": 0.9226, + "step": 1670 + }, + { + "epoch": 0.05270462072228355, + "grad_norm": 2.606189727783203, + "learning_rate": 1.997311259133193e-05, + "loss": 0.3957, + "step": 1671 + }, + { + "epoch": 0.052736161488724176, + "grad_norm": 7.081131458282471, + "learning_rate": 1.9973037677585102e-05, + "loss": 0.6941, + "step": 1672 + }, + { + "epoch": 0.0527677022551648, + "grad_norm": 4.8304057121276855, + "learning_rate": 1.9972962659761883e-05, + "loss": 0.667, + "step": 1673 + }, + { + "epoch": 0.05279924302160543, + "grad_norm": 3.2595388889312744, + "learning_rate": 1.9972887537863056e-05, + "loss": 0.6778, + "step": 1674 + }, + { + "epoch": 0.05283078378804605, + "grad_norm": 5.4540019035339355, + "learning_rate": 1.9972812311889402e-05, + "loss": 0.6398, + "step": 1675 + }, + { + "epoch": 0.052862324554486674, + "grad_norm": 7.012209415435791, + "learning_rate": 1.9972736981841708e-05, + "loss": 0.686, + "step": 1676 + }, + { + "epoch": 0.0528938653209273, + "grad_norm": 5.370125770568848, + "learning_rate": 1.9972661547720763e-05, + "loss": 0.6306, + "step": 1677 + }, + { + "epoch": 0.05292540608736792, + "grad_norm": 4.453693389892578, + "learning_rate": 1.997258600952735e-05, + "loss": 0.6522, + "step": 1678 + }, + { + "epoch": 0.05295694685380855, + "grad_norm": 3.3184304237365723, + "learning_rate": 1.9972510367262257e-05, + "loss": 0.9195, + "step": 1679 + }, + { + "epoch": 0.05298848762024917, + "grad_norm": 2.744270086288452, + "learning_rate": 1.9972434620926276e-05, + "loss": 0.8645, + "step": 1680 + }, + { + "epoch": 0.053020028386689795, + "grad_norm": 5.192082405090332, + "learning_rate": 1.9972358770520194e-05, + "loss": 0.6685, + "step": 1681 + }, + { + "epoch": 0.05305156915313042, + "grad_norm": 6.851865291595459, + "learning_rate": 1.997228281604481e-05, + "loss": 0.6755, + "step": 1682 + }, + { + "epoch": 0.05308310991957105, + "grad_norm": 5.94407320022583, + "learning_rate": 1.9972206757500905e-05, + "loss": 0.6792, + "step": 1683 + }, + { + "epoch": 0.05311465068601167, + "grad_norm": 4.405168056488037, + "learning_rate": 1.9972130594889284e-05, + "loss": 0.6755, + "step": 1684 + }, + { + "epoch": 0.05314619145245229, + "grad_norm": 4.257384777069092, + "learning_rate": 1.9972054328210734e-05, + "loss": 0.641, + "step": 1685 + }, + { + "epoch": 0.053177732218892916, + "grad_norm": 4.8167619705200195, + "learning_rate": 1.9971977957466055e-05, + "loss": 0.6412, + "step": 1686 + }, + { + "epoch": 0.053209272985333546, + "grad_norm": 4.963869094848633, + "learning_rate": 1.9971901482656047e-05, + "loss": 0.4015, + "step": 1687 + }, + { + "epoch": 0.05324081375177417, + "grad_norm": 6.006738185882568, + "learning_rate": 1.9971824903781498e-05, + "loss": 0.66, + "step": 1688 + }, + { + "epoch": 0.05327235451821479, + "grad_norm": 22.188873291015625, + "learning_rate": 1.9971748220843216e-05, + "loss": 0.63, + "step": 1689 + }, + { + "epoch": 0.053303895284655414, + "grad_norm": 3.57649302482605, + "learning_rate": 1.9971671433842e-05, + "loss": 0.6541, + "step": 1690 + }, + { + "epoch": 0.053335436051096044, + "grad_norm": 3.6980252265930176, + "learning_rate": 1.997159454277865e-05, + "loss": 0.6273, + "step": 1691 + }, + { + "epoch": 0.05336697681753667, + "grad_norm": 4.719913959503174, + "learning_rate": 1.9971517547653963e-05, + "loss": 0.9051, + "step": 1692 + }, + { + "epoch": 0.05339851758397729, + "grad_norm": 4.303354263305664, + "learning_rate": 1.9971440448468752e-05, + "loss": 0.6748, + "step": 1693 + }, + { + "epoch": 0.05343005835041791, + "grad_norm": 4.897776126861572, + "learning_rate": 1.9971363245223813e-05, + "loss": 0.8712, + "step": 1694 + }, + { + "epoch": 0.05346159911685854, + "grad_norm": 5.898061275482178, + "learning_rate": 1.997128593791996e-05, + "loss": 0.7031, + "step": 1695 + }, + { + "epoch": 0.053493139883299165, + "grad_norm": 5.085196018218994, + "learning_rate": 1.9971208526557996e-05, + "loss": 0.6184, + "step": 1696 + }, + { + "epoch": 0.05352468064973979, + "grad_norm": 3.0960376262664795, + "learning_rate": 1.9971131011138728e-05, + "loss": 0.6369, + "step": 1697 + }, + { + "epoch": 0.05355622141618041, + "grad_norm": 4.221177101135254, + "learning_rate": 1.9971053391662963e-05, + "loss": 0.6494, + "step": 1698 + }, + { + "epoch": 0.05358776218262104, + "grad_norm": 4.499030113220215, + "learning_rate": 1.997097566813152e-05, + "loss": 0.3974, + "step": 1699 + }, + { + "epoch": 0.05361930294906166, + "grad_norm": 5.163221836090088, + "learning_rate": 1.9970897840545197e-05, + "loss": 0.6866, + "step": 1700 + }, + { + "epoch": 0.053650843715502286, + "grad_norm": 4.309240818023682, + "learning_rate": 1.9970819908904815e-05, + "loss": 0.6599, + "step": 1701 + }, + { + "epoch": 0.05368238448194291, + "grad_norm": 4.437301158905029, + "learning_rate": 1.9970741873211182e-05, + "loss": 0.6741, + "step": 1702 + }, + { + "epoch": 0.05371392524838354, + "grad_norm": 4.343757152557373, + "learning_rate": 1.997066373346512e-05, + "loss": 0.356, + "step": 1703 + }, + { + "epoch": 0.05374546601482416, + "grad_norm": 8.967626571655273, + "learning_rate": 1.997058548966744e-05, + "loss": 0.6635, + "step": 1704 + }, + { + "epoch": 0.053777006781264784, + "grad_norm": 4.352544784545898, + "learning_rate": 1.9970507141818957e-05, + "loss": 0.666, + "step": 1705 + }, + { + "epoch": 0.05380854754770541, + "grad_norm": 4.269705772399902, + "learning_rate": 1.997042868992049e-05, + "loss": 0.6684, + "step": 1706 + }, + { + "epoch": 0.05384008831414604, + "grad_norm": 4.280814170837402, + "learning_rate": 1.9970350133972858e-05, + "loss": 0.3662, + "step": 1707 + }, + { + "epoch": 0.05387162908058666, + "grad_norm": 4.953479290008545, + "learning_rate": 1.997027147397688e-05, + "loss": 0.6327, + "step": 1708 + }, + { + "epoch": 0.05390316984702728, + "grad_norm": 3.870655059814453, + "learning_rate": 1.997019270993338e-05, + "loss": 0.6264, + "step": 1709 + }, + { + "epoch": 0.053934710613467905, + "grad_norm": 4.965066909790039, + "learning_rate": 1.9970113841843176e-05, + "loss": 0.6543, + "step": 1710 + }, + { + "epoch": 0.053966251379908535, + "grad_norm": 4.6550068855285645, + "learning_rate": 1.9970034869707087e-05, + "loss": 0.393, + "step": 1711 + }, + { + "epoch": 0.05399779214634916, + "grad_norm": 6.638301372528076, + "learning_rate": 1.996995579352595e-05, + "loss": 0.6353, + "step": 1712 + }, + { + "epoch": 0.05402933291278978, + "grad_norm": 4.31451416015625, + "learning_rate": 1.996987661330058e-05, + "loss": 0.6633, + "step": 1713 + }, + { + "epoch": 0.0540608736792304, + "grad_norm": 4.402210235595703, + "learning_rate": 1.9969797329031807e-05, + "loss": 0.6409, + "step": 1714 + }, + { + "epoch": 0.05409241444567103, + "grad_norm": 4.700035572052002, + "learning_rate": 1.9969717940720456e-05, + "loss": 0.6809, + "step": 1715 + }, + { + "epoch": 0.054123955212111656, + "grad_norm": 6.370883941650391, + "learning_rate": 1.996963844836736e-05, + "loss": 0.6487, + "step": 1716 + }, + { + "epoch": 0.05415549597855228, + "grad_norm": 2.7333178520202637, + "learning_rate": 1.996955885197334e-05, + "loss": 0.4151, + "step": 1717 + }, + { + "epoch": 0.0541870367449929, + "grad_norm": 5.070029258728027, + "learning_rate": 1.9969479151539238e-05, + "loss": 0.6639, + "step": 1718 + }, + { + "epoch": 0.054218577511433524, + "grad_norm": 3.7080862522125244, + "learning_rate": 1.9969399347065878e-05, + "loss": 0.6866, + "step": 1719 + }, + { + "epoch": 0.054250118277874154, + "grad_norm": 4.079654693603516, + "learning_rate": 1.9969319438554094e-05, + "loss": 0.6441, + "step": 1720 + }, + { + "epoch": 0.05428165904431478, + "grad_norm": 4.577974319458008, + "learning_rate": 1.996923942600472e-05, + "loss": 0.648, + "step": 1721 + }, + { + "epoch": 0.0543131998107554, + "grad_norm": 6.504491806030273, + "learning_rate": 1.9969159309418594e-05, + "loss": 0.6846, + "step": 1722 + }, + { + "epoch": 0.05434474057719602, + "grad_norm": 4.941755771636963, + "learning_rate": 1.9969079088796545e-05, + "loss": 0.9163, + "step": 1723 + }, + { + "epoch": 0.05437628134363665, + "grad_norm": 5.069811820983887, + "learning_rate": 1.996899876413942e-05, + "loss": 0.6463, + "step": 1724 + }, + { + "epoch": 0.054407822110077275, + "grad_norm": 3.8987438678741455, + "learning_rate": 1.996891833544805e-05, + "loss": 0.6453, + "step": 1725 + }, + { + "epoch": 0.0544393628765179, + "grad_norm": 3.871697187423706, + "learning_rate": 1.9968837802723277e-05, + "loss": 0.634, + "step": 1726 + }, + { + "epoch": 0.05447090364295852, + "grad_norm": 3.37420654296875, + "learning_rate": 1.996875716596594e-05, + "loss": 0.6317, + "step": 1727 + }, + { + "epoch": 0.05450244440939915, + "grad_norm": 4.30679988861084, + "learning_rate": 1.9968676425176884e-05, + "loss": 0.6517, + "step": 1728 + }, + { + "epoch": 0.05453398517583977, + "grad_norm": 4.743531703948975, + "learning_rate": 1.996859558035695e-05, + "loss": 0.6807, + "step": 1729 + }, + { + "epoch": 0.054565525942280396, + "grad_norm": 4.249992847442627, + "learning_rate": 1.9968514631506978e-05, + "loss": 0.6254, + "step": 1730 + }, + { + "epoch": 0.05459706670872102, + "grad_norm": 3.1002159118652344, + "learning_rate": 1.9968433578627814e-05, + "loss": 0.4037, + "step": 1731 + }, + { + "epoch": 0.05462860747516165, + "grad_norm": 3.9501328468322754, + "learning_rate": 1.9968352421720305e-05, + "loss": 0.6692, + "step": 1732 + }, + { + "epoch": 0.05466014824160227, + "grad_norm": 31.689350128173828, + "learning_rate": 1.9968271160785302e-05, + "loss": 0.4124, + "step": 1733 + }, + { + "epoch": 0.054691689008042894, + "grad_norm": 5.1303815841674805, + "learning_rate": 1.9968189795823646e-05, + "loss": 0.6864, + "step": 1734 + }, + { + "epoch": 0.05472322977448352, + "grad_norm": 5.577346324920654, + "learning_rate": 1.9968108326836192e-05, + "loss": 0.6521, + "step": 1735 + }, + { + "epoch": 0.05475477054092415, + "grad_norm": 4.625715255737305, + "learning_rate": 1.996802675382379e-05, + "loss": 0.6395, + "step": 1736 + }, + { + "epoch": 0.05478631130736477, + "grad_norm": 6.263810634613037, + "learning_rate": 1.996794507678728e-05, + "loss": 0.639, + "step": 1737 + }, + { + "epoch": 0.05481785207380539, + "grad_norm": 19.052114486694336, + "learning_rate": 1.996786329572753e-05, + "loss": 0.6579, + "step": 1738 + }, + { + "epoch": 0.054849392840246015, + "grad_norm": 5.588168144226074, + "learning_rate": 1.9967781410645385e-05, + "loss": 0.6453, + "step": 1739 + }, + { + "epoch": 0.054880933606686645, + "grad_norm": 5.553406715393066, + "learning_rate": 1.9967699421541704e-05, + "loss": 0.6492, + "step": 1740 + }, + { + "epoch": 0.05491247437312727, + "grad_norm": 6.148271083831787, + "learning_rate": 1.9967617328417335e-05, + "loss": 0.6892, + "step": 1741 + }, + { + "epoch": 0.05494401513956789, + "grad_norm": 8.181683540344238, + "learning_rate": 1.9967535131273143e-05, + "loss": 0.6695, + "step": 1742 + }, + { + "epoch": 0.054975555906008514, + "grad_norm": 8.98608112335205, + "learning_rate": 1.996745283010998e-05, + "loss": 0.6462, + "step": 1743 + }, + { + "epoch": 0.05500709667244914, + "grad_norm": 4.00689697265625, + "learning_rate": 1.9967370424928707e-05, + "loss": 0.6113, + "step": 1744 + }, + { + "epoch": 0.055038637438889766, + "grad_norm": 4.059787750244141, + "learning_rate": 1.9967287915730184e-05, + "loss": 0.688, + "step": 1745 + }, + { + "epoch": 0.05507017820533039, + "grad_norm": 4.750704765319824, + "learning_rate": 1.9967205302515277e-05, + "loss": 0.6716, + "step": 1746 + }, + { + "epoch": 0.05510171897177101, + "grad_norm": 4.9372172355651855, + "learning_rate": 1.9967122585284836e-05, + "loss": 0.9275, + "step": 1747 + }, + { + "epoch": 0.05513325973821164, + "grad_norm": 6.070457458496094, + "learning_rate": 1.9967039764039738e-05, + "loss": 0.5953, + "step": 1748 + }, + { + "epoch": 0.055164800504652264, + "grad_norm": 4.244146823883057, + "learning_rate": 1.996695683878084e-05, + "loss": 0.6496, + "step": 1749 + }, + { + "epoch": 0.05519634127109289, + "grad_norm": 8.476703643798828, + "learning_rate": 1.9966873809509005e-05, + "loss": 0.657, + "step": 1750 + }, + { + "epoch": 0.05522788203753351, + "grad_norm": 3.58282470703125, + "learning_rate": 1.9966790676225104e-05, + "loss": 0.6473, + "step": 1751 + }, + { + "epoch": 0.05525942280397414, + "grad_norm": 2.664667844772339, + "learning_rate": 1.9966707438930006e-05, + "loss": 0.879, + "step": 1752 + }, + { + "epoch": 0.05529096357041476, + "grad_norm": 3.732595443725586, + "learning_rate": 1.996662409762457e-05, + "loss": 0.6415, + "step": 1753 + }, + { + "epoch": 0.055322504336855385, + "grad_norm": 3.331037759780884, + "learning_rate": 1.9966540652309677e-05, + "loss": 0.66, + "step": 1754 + }, + { + "epoch": 0.05535404510329601, + "grad_norm": 4.006918430328369, + "learning_rate": 1.9966457102986198e-05, + "loss": 0.6393, + "step": 1755 + }, + { + "epoch": 0.05538558586973664, + "grad_norm": 3.7907156944274902, + "learning_rate": 1.9966373449654994e-05, + "loss": 0.4194, + "step": 1756 + }, + { + "epoch": 0.05541712663617726, + "grad_norm": 3.9545655250549316, + "learning_rate": 1.9966289692316944e-05, + "loss": 0.6559, + "step": 1757 + }, + { + "epoch": 0.055448667402617884, + "grad_norm": 4.3925065994262695, + "learning_rate": 1.9966205830972927e-05, + "loss": 0.6358, + "step": 1758 + }, + { + "epoch": 0.055480208169058506, + "grad_norm": 2.82346248626709, + "learning_rate": 1.9966121865623812e-05, + "loss": 0.8881, + "step": 1759 + }, + { + "epoch": 0.05551174893549913, + "grad_norm": 3.8781025409698486, + "learning_rate": 1.9966037796270476e-05, + "loss": 0.6358, + "step": 1760 + }, + { + "epoch": 0.05554328970193976, + "grad_norm": 2.9280078411102295, + "learning_rate": 1.99659536229138e-05, + "loss": 0.8778, + "step": 1761 + }, + { + "epoch": 0.05557483046838038, + "grad_norm": 4.205440521240234, + "learning_rate": 1.9965869345554653e-05, + "loss": 0.6122, + "step": 1762 + }, + { + "epoch": 0.055606371234821005, + "grad_norm": 5.653544902801514, + "learning_rate": 1.9965784964193925e-05, + "loss": 0.6669, + "step": 1763 + }, + { + "epoch": 0.05563791200126163, + "grad_norm": 3.5088717937469482, + "learning_rate": 1.9965700478832493e-05, + "loss": 0.6453, + "step": 1764 + }, + { + "epoch": 0.05566945276770226, + "grad_norm": 3.945131540298462, + "learning_rate": 1.9965615889471238e-05, + "loss": 0.6357, + "step": 1765 + }, + { + "epoch": 0.05570099353414288, + "grad_norm": 5.3693671226501465, + "learning_rate": 1.9965531196111043e-05, + "loss": 0.6379, + "step": 1766 + }, + { + "epoch": 0.0557325343005835, + "grad_norm": 2.8142879009246826, + "learning_rate": 1.9965446398752794e-05, + "loss": 0.8942, + "step": 1767 + }, + { + "epoch": 0.055764075067024126, + "grad_norm": 3.8145487308502197, + "learning_rate": 1.996536149739737e-05, + "loss": 0.6475, + "step": 1768 + }, + { + "epoch": 0.055795615833464755, + "grad_norm": 5.012967586517334, + "learning_rate": 1.9965276492045662e-05, + "loss": 0.6487, + "step": 1769 + }, + { + "epoch": 0.05582715659990538, + "grad_norm": 2.916456460952759, + "learning_rate": 1.9965191382698558e-05, + "loss": 0.901, + "step": 1770 + }, + { + "epoch": 0.055858697366346, + "grad_norm": 2.583085060119629, + "learning_rate": 1.9965106169356946e-05, + "loss": 0.8847, + "step": 1771 + }, + { + "epoch": 0.055890238132786624, + "grad_norm": 3.572462797164917, + "learning_rate": 1.9965020852021708e-05, + "loss": 0.6178, + "step": 1772 + }, + { + "epoch": 0.055921778899227254, + "grad_norm": 4.115353107452393, + "learning_rate": 1.996493543069374e-05, + "loss": 0.6372, + "step": 1773 + }, + { + "epoch": 0.055953319665667876, + "grad_norm": 3.5372743606567383, + "learning_rate": 1.9964849905373937e-05, + "loss": 0.6511, + "step": 1774 + }, + { + "epoch": 0.0559848604321085, + "grad_norm": 4.329875469207764, + "learning_rate": 1.9964764276063183e-05, + "loss": 0.6454, + "step": 1775 + }, + { + "epoch": 0.05601640119854912, + "grad_norm": 5.666725158691406, + "learning_rate": 1.9964678542762384e-05, + "loss": 0.6455, + "step": 1776 + }, + { + "epoch": 0.05604794196498975, + "grad_norm": 7.012946605682373, + "learning_rate": 1.996459270547242e-05, + "loss": 0.9074, + "step": 1777 + }, + { + "epoch": 0.056079482731430375, + "grad_norm": 6.395144939422607, + "learning_rate": 1.9964506764194197e-05, + "loss": 0.6573, + "step": 1778 + }, + { + "epoch": 0.056111023497871, + "grad_norm": 4.399648189544678, + "learning_rate": 1.996442071892861e-05, + "loss": 0.654, + "step": 1779 + }, + { + "epoch": 0.05614256426431162, + "grad_norm": 3.2538137435913086, + "learning_rate": 1.996433456967655e-05, + "loss": 0.9151, + "step": 1780 + }, + { + "epoch": 0.05617410503075225, + "grad_norm": 2.619288921356201, + "learning_rate": 1.9964248316438928e-05, + "loss": 0.918, + "step": 1781 + }, + { + "epoch": 0.05620564579719287, + "grad_norm": 4.193357467651367, + "learning_rate": 1.9964161959216635e-05, + "loss": 0.6158, + "step": 1782 + }, + { + "epoch": 0.056237186563633496, + "grad_norm": 5.077061653137207, + "learning_rate": 1.9964075498010574e-05, + "loss": 0.6061, + "step": 1783 + }, + { + "epoch": 0.05626872733007412, + "grad_norm": 3.6463816165924072, + "learning_rate": 1.9963988932821648e-05, + "loss": 0.626, + "step": 1784 + }, + { + "epoch": 0.05630026809651475, + "grad_norm": 5.771397590637207, + "learning_rate": 1.9963902263650764e-05, + "loss": 0.6285, + "step": 1785 + }, + { + "epoch": 0.05633180886295537, + "grad_norm": 4.93966007232666, + "learning_rate": 1.996381549049882e-05, + "loss": 0.6531, + "step": 1786 + }, + { + "epoch": 0.056363349629395994, + "grad_norm": 4.950826168060303, + "learning_rate": 1.9963728613366725e-05, + "loss": 0.6417, + "step": 1787 + }, + { + "epoch": 0.05639489039583662, + "grad_norm": 5.333930015563965, + "learning_rate": 1.9963641632255386e-05, + "loss": 0.6408, + "step": 1788 + }, + { + "epoch": 0.056426431162277246, + "grad_norm": 3.7545063495635986, + "learning_rate": 1.996355454716571e-05, + "loss": 0.8932, + "step": 1789 + }, + { + "epoch": 0.05645797192871787, + "grad_norm": 4.2093400955200195, + "learning_rate": 1.9963467358098607e-05, + "loss": 0.6356, + "step": 1790 + }, + { + "epoch": 0.05648951269515849, + "grad_norm": 6.604147434234619, + "learning_rate": 1.996338006505498e-05, + "loss": 0.6358, + "step": 1791 + }, + { + "epoch": 0.056521053461599115, + "grad_norm": 4.984319686889648, + "learning_rate": 1.9963292668035753e-05, + "loss": 0.6341, + "step": 1792 + }, + { + "epoch": 0.056552594228039745, + "grad_norm": 3.0042412281036377, + "learning_rate": 1.996320516704183e-05, + "loss": 0.8636, + "step": 1793 + }, + { + "epoch": 0.05658413499448037, + "grad_norm": 3.872324228286743, + "learning_rate": 1.9963117562074118e-05, + "loss": 0.633, + "step": 1794 + }, + { + "epoch": 0.05661567576092099, + "grad_norm": 3.5554115772247314, + "learning_rate": 1.9963029853133547e-05, + "loss": 0.6204, + "step": 1795 + }, + { + "epoch": 0.05664721652736161, + "grad_norm": 6.131850242614746, + "learning_rate": 1.9962942040221017e-05, + "loss": 0.6587, + "step": 1796 + }, + { + "epoch": 0.05667875729380224, + "grad_norm": 5.24085807800293, + "learning_rate": 1.9962854123337452e-05, + "loss": 0.8896, + "step": 1797 + }, + { + "epoch": 0.056710298060242866, + "grad_norm": 6.205704212188721, + "learning_rate": 1.996276610248377e-05, + "loss": 0.8741, + "step": 1798 + }, + { + "epoch": 0.05674183882668349, + "grad_norm": 28.921939849853516, + "learning_rate": 1.9962677977660886e-05, + "loss": 0.678, + "step": 1799 + }, + { + "epoch": 0.05677337959312411, + "grad_norm": 5.295402526855469, + "learning_rate": 1.9962589748869724e-05, + "loss": 0.6293, + "step": 1800 + }, + { + "epoch": 0.056804920359564734, + "grad_norm": 4.29412841796875, + "learning_rate": 1.9962501416111203e-05, + "loss": 0.6452, + "step": 1801 + }, + { + "epoch": 0.056836461126005364, + "grad_norm": 2.724902868270874, + "learning_rate": 1.996241297938624e-05, + "loss": 0.8961, + "step": 1802 + }, + { + "epoch": 0.056868001892445987, + "grad_norm": 5.027266025543213, + "learning_rate": 1.9962324438695764e-05, + "loss": 0.6622, + "step": 1803 + }, + { + "epoch": 0.05689954265888661, + "grad_norm": 3.661609411239624, + "learning_rate": 1.9962235794040695e-05, + "loss": 0.6445, + "step": 1804 + }, + { + "epoch": 0.05693108342532723, + "grad_norm": 4.716242790222168, + "learning_rate": 1.9962147045421964e-05, + "loss": 0.6338, + "step": 1805 + }, + { + "epoch": 0.05696262419176786, + "grad_norm": 8.357508659362793, + "learning_rate": 1.996205819284049e-05, + "loss": 0.6242, + "step": 1806 + }, + { + "epoch": 0.056994164958208485, + "grad_norm": 5.71186637878418, + "learning_rate": 1.9961969236297203e-05, + "loss": 0.5843, + "step": 1807 + }, + { + "epoch": 0.05702570572464911, + "grad_norm": 3.5593314170837402, + "learning_rate": 1.9961880175793033e-05, + "loss": 0.6387, + "step": 1808 + }, + { + "epoch": 0.05705724649108973, + "grad_norm": 5.376848220825195, + "learning_rate": 1.9961791011328908e-05, + "loss": 0.6367, + "step": 1809 + }, + { + "epoch": 0.05708878725753036, + "grad_norm": 4.646358489990234, + "learning_rate": 1.996170174290576e-05, + "loss": 0.8842, + "step": 1810 + }, + { + "epoch": 0.05712032802397098, + "grad_norm": 5.654634475708008, + "learning_rate": 1.9961612370524515e-05, + "loss": 0.6526, + "step": 1811 + }, + { + "epoch": 0.057151868790411606, + "grad_norm": 4.251908779144287, + "learning_rate": 1.9961522894186115e-05, + "loss": 0.6472, + "step": 1812 + }, + { + "epoch": 0.05718340955685223, + "grad_norm": 4.135868549346924, + "learning_rate": 1.9961433313891486e-05, + "loss": 0.6127, + "step": 1813 + }, + { + "epoch": 0.05721495032329286, + "grad_norm": 4.901483535766602, + "learning_rate": 1.9961343629641568e-05, + "loss": 0.6573, + "step": 1814 + }, + { + "epoch": 0.05724649108973348, + "grad_norm": 11.53322696685791, + "learning_rate": 1.996125384143729e-05, + "loss": 0.8792, + "step": 1815 + }, + { + "epoch": 0.057278031856174104, + "grad_norm": 4.2277421951293945, + "learning_rate": 1.9961163949279595e-05, + "loss": 0.6386, + "step": 1816 + }, + { + "epoch": 0.05730957262261473, + "grad_norm": 4.791053771972656, + "learning_rate": 1.9961073953169422e-05, + "loss": 0.6562, + "step": 1817 + }, + { + "epoch": 0.057341113389055356, + "grad_norm": 3.6828958988189697, + "learning_rate": 1.9960983853107706e-05, + "loss": 0.6196, + "step": 1818 + }, + { + "epoch": 0.05737265415549598, + "grad_norm": 5.254137992858887, + "learning_rate": 1.9960893649095393e-05, + "loss": 0.6533, + "step": 1819 + }, + { + "epoch": 0.0574041949219366, + "grad_norm": 4.91532564163208, + "learning_rate": 1.996080334113341e-05, + "loss": 0.6678, + "step": 1820 + }, + { + "epoch": 0.057435735688377225, + "grad_norm": 4.740725040435791, + "learning_rate": 1.9960712929222722e-05, + "loss": 0.6137, + "step": 1821 + }, + { + "epoch": 0.057467276454817855, + "grad_norm": 6.98223352432251, + "learning_rate": 1.9960622413364255e-05, + "loss": 0.634, + "step": 1822 + }, + { + "epoch": 0.05749881722125848, + "grad_norm": 3.7692577838897705, + "learning_rate": 1.9960531793558956e-05, + "loss": 0.6545, + "step": 1823 + }, + { + "epoch": 0.0575303579876991, + "grad_norm": 3.866140604019165, + "learning_rate": 1.9960441069807778e-05, + "loss": 0.6212, + "step": 1824 + }, + { + "epoch": 0.05756189875413972, + "grad_norm": 7.527675628662109, + "learning_rate": 1.9960350242111662e-05, + "loss": 0.6574, + "step": 1825 + }, + { + "epoch": 0.05759343952058035, + "grad_norm": 5.990963935852051, + "learning_rate": 1.9960259310471555e-05, + "loss": 0.6549, + "step": 1826 + }, + { + "epoch": 0.057624980287020976, + "grad_norm": 4.636219024658203, + "learning_rate": 1.996016827488841e-05, + "loss": 0.6064, + "step": 1827 + }, + { + "epoch": 0.0576565210534616, + "grad_norm": 4.1156086921691895, + "learning_rate": 1.9960077135363172e-05, + "loss": 0.6355, + "step": 1828 + }, + { + "epoch": 0.05768806181990222, + "grad_norm": 6.181180477142334, + "learning_rate": 1.99599858918968e-05, + "loss": 0.6528, + "step": 1829 + }, + { + "epoch": 0.05771960258634285, + "grad_norm": 6.157158374786377, + "learning_rate": 1.9959894544490243e-05, + "loss": 0.6211, + "step": 1830 + }, + { + "epoch": 0.057751143352783474, + "grad_norm": 5.624359130859375, + "learning_rate": 1.995980309314445e-05, + "loss": 0.6151, + "step": 1831 + }, + { + "epoch": 0.0577826841192241, + "grad_norm": 3.7616701126098633, + "learning_rate": 1.9959711537860377e-05, + "loss": 0.6518, + "step": 1832 + }, + { + "epoch": 0.05781422488566472, + "grad_norm": 4.223861217498779, + "learning_rate": 1.995961987863898e-05, + "loss": 0.6379, + "step": 1833 + }, + { + "epoch": 0.05784576565210535, + "grad_norm": 4.428847789764404, + "learning_rate": 1.9959528115481217e-05, + "loss": 0.6065, + "step": 1834 + }, + { + "epoch": 0.05787730641854597, + "grad_norm": 4.140098571777344, + "learning_rate": 1.9959436248388044e-05, + "loss": 0.6489, + "step": 1835 + }, + { + "epoch": 0.057908847184986595, + "grad_norm": 4.434847831726074, + "learning_rate": 1.9959344277360424e-05, + "loss": 0.6509, + "step": 1836 + }, + { + "epoch": 0.05794038795142722, + "grad_norm": 4.043704986572266, + "learning_rate": 1.995925220239931e-05, + "loss": 0.6574, + "step": 1837 + }, + { + "epoch": 0.05797192871786785, + "grad_norm": 7.333907604217529, + "learning_rate": 1.9959160023505667e-05, + "loss": 0.6647, + "step": 1838 + }, + { + "epoch": 0.05800346948430847, + "grad_norm": 4.04555082321167, + "learning_rate": 1.9959067740680456e-05, + "loss": 0.648, + "step": 1839 + }, + { + "epoch": 0.05803501025074909, + "grad_norm": 8.104862213134766, + "learning_rate": 1.995897535392464e-05, + "loss": 0.6526, + "step": 1840 + }, + { + "epoch": 0.058066551017189716, + "grad_norm": 4.280680179595947, + "learning_rate": 1.9958882863239185e-05, + "loss": 0.6425, + "step": 1841 + }, + { + "epoch": 0.058098091783630346, + "grad_norm": 5.7150068283081055, + "learning_rate": 1.9958790268625053e-05, + "loss": 0.6615, + "step": 1842 + }, + { + "epoch": 0.05812963255007097, + "grad_norm": 4.881062984466553, + "learning_rate": 1.995869757008321e-05, + "loss": 0.64, + "step": 1843 + }, + { + "epoch": 0.05816117331651159, + "grad_norm": 6.569055557250977, + "learning_rate": 1.9958604767614626e-05, + "loss": 0.6541, + "step": 1844 + }, + { + "epoch": 0.058192714082952214, + "grad_norm": 7.306125164031982, + "learning_rate": 1.995851186122027e-05, + "loss": 0.6153, + "step": 1845 + }, + { + "epoch": 0.05822425484939284, + "grad_norm": 142.5227508544922, + "learning_rate": 1.995841885090111e-05, + "loss": 0.3987, + "step": 1846 + }, + { + "epoch": 0.05825579561583347, + "grad_norm": 4.1403117179870605, + "learning_rate": 1.9958325736658117e-05, + "loss": 0.5982, + "step": 1847 + }, + { + "epoch": 0.05828733638227409, + "grad_norm": 6.9329071044921875, + "learning_rate": 1.995823251849226e-05, + "loss": 0.6373, + "step": 1848 + }, + { + "epoch": 0.05831887714871471, + "grad_norm": 2.7644920349121094, + "learning_rate": 1.9958139196404516e-05, + "loss": 0.356, + "step": 1849 + }, + { + "epoch": 0.058350417915155335, + "grad_norm": 6.384450912475586, + "learning_rate": 1.9958045770395857e-05, + "loss": 0.6623, + "step": 1850 + }, + { + "epoch": 0.058381958681595965, + "grad_norm": 5.451728820800781, + "learning_rate": 1.995795224046726e-05, + "loss": 0.9155, + "step": 1851 + }, + { + "epoch": 0.05841349944803659, + "grad_norm": 3.863064765930176, + "learning_rate": 1.9957858606619697e-05, + "loss": 0.6678, + "step": 1852 + }, + { + "epoch": 0.05844504021447721, + "grad_norm": 5.073742866516113, + "learning_rate": 1.9957764868854148e-05, + "loss": 0.6443, + "step": 1853 + }, + { + "epoch": 0.05847658098091783, + "grad_norm": 3.531010866165161, + "learning_rate": 1.995767102717159e-05, + "loss": 0.8844, + "step": 1854 + }, + { + "epoch": 0.05850812174735846, + "grad_norm": 7.049334526062012, + "learning_rate": 1.9957577081573004e-05, + "loss": 0.6548, + "step": 1855 + }, + { + "epoch": 0.058539662513799086, + "grad_norm": 5.728383541107178, + "learning_rate": 1.9957483032059367e-05, + "loss": 0.6216, + "step": 1856 + }, + { + "epoch": 0.05857120328023971, + "grad_norm": 4.395351886749268, + "learning_rate": 1.9957388878631664e-05, + "loss": 0.6123, + "step": 1857 + }, + { + "epoch": 0.05860274404668033, + "grad_norm": 3.699631452560425, + "learning_rate": 1.995729462129088e-05, + "loss": 0.6647, + "step": 1858 + }, + { + "epoch": 0.05863428481312096, + "grad_norm": 3.607705593109131, + "learning_rate": 1.9957200260037993e-05, + "loss": 0.634, + "step": 1859 + }, + { + "epoch": 0.058665825579561584, + "grad_norm": 4.293314456939697, + "learning_rate": 1.9957105794873986e-05, + "loss": 0.6449, + "step": 1860 + }, + { + "epoch": 0.05869736634600221, + "grad_norm": 4.9412665367126465, + "learning_rate": 1.9957011225799853e-05, + "loss": 0.6309, + "step": 1861 + }, + { + "epoch": 0.05872890711244283, + "grad_norm": 3.4085419178009033, + "learning_rate": 1.9956916552816574e-05, + "loss": 0.6403, + "step": 1862 + }, + { + "epoch": 0.05876044787888346, + "grad_norm": 5.12522554397583, + "learning_rate": 1.995682177592514e-05, + "loss": 0.6525, + "step": 1863 + }, + { + "epoch": 0.05879198864532408, + "grad_norm": 4.007986068725586, + "learning_rate": 1.995672689512654e-05, + "loss": 0.6272, + "step": 1864 + }, + { + "epoch": 0.058823529411764705, + "grad_norm": 3.630709171295166, + "learning_rate": 1.9956631910421762e-05, + "loss": 0.6421, + "step": 1865 + }, + { + "epoch": 0.05885507017820533, + "grad_norm": 3.6970014572143555, + "learning_rate": 1.9956536821811802e-05, + "loss": 0.6396, + "step": 1866 + }, + { + "epoch": 0.05888661094464596, + "grad_norm": 3.726304054260254, + "learning_rate": 1.9956441629297646e-05, + "loss": 0.6125, + "step": 1867 + }, + { + "epoch": 0.05891815171108658, + "grad_norm": 3.0512168407440186, + "learning_rate": 1.9956346332880292e-05, + "loss": 0.4574, + "step": 1868 + }, + { + "epoch": 0.0589496924775272, + "grad_norm": 8.575797080993652, + "learning_rate": 1.995625093256073e-05, + "loss": 0.9557, + "step": 1869 + }, + { + "epoch": 0.058981233243967826, + "grad_norm": 5.371415615081787, + "learning_rate": 1.9956155428339964e-05, + "loss": 0.6479, + "step": 1870 + }, + { + "epoch": 0.059012774010408456, + "grad_norm": 3.6495347023010254, + "learning_rate": 1.9956059820218982e-05, + "loss": 0.634, + "step": 1871 + }, + { + "epoch": 0.05904431477684908, + "grad_norm": 3.4089276790618896, + "learning_rate": 1.9955964108198787e-05, + "loss": 0.6004, + "step": 1872 + }, + { + "epoch": 0.0590758555432897, + "grad_norm": 3.658210277557373, + "learning_rate": 1.9955868292280373e-05, + "loss": 0.6405, + "step": 1873 + }, + { + "epoch": 0.059107396309730324, + "grad_norm": 3.955378770828247, + "learning_rate": 1.9955772372464747e-05, + "loss": 0.6404, + "step": 1874 + }, + { + "epoch": 0.059138937076170954, + "grad_norm": 4.067502975463867, + "learning_rate": 1.9955676348752904e-05, + "loss": 0.6641, + "step": 1875 + }, + { + "epoch": 0.05917047784261158, + "grad_norm": 3.4921295642852783, + "learning_rate": 1.9955580221145847e-05, + "loss": 0.919, + "step": 1876 + }, + { + "epoch": 0.0592020186090522, + "grad_norm": 5.975368022918701, + "learning_rate": 1.995548398964458e-05, + "loss": 0.5926, + "step": 1877 + }, + { + "epoch": 0.05923355937549282, + "grad_norm": 4.69643497467041, + "learning_rate": 1.995538765425011e-05, + "loss": 0.6635, + "step": 1878 + }, + { + "epoch": 0.05926510014193345, + "grad_norm": 3.9294378757476807, + "learning_rate": 1.9955291214963437e-05, + "loss": 0.6285, + "step": 1879 + }, + { + "epoch": 0.059296640908374075, + "grad_norm": 6.0254950523376465, + "learning_rate": 1.9955194671785573e-05, + "loss": 0.6119, + "step": 1880 + }, + { + "epoch": 0.0593281816748147, + "grad_norm": 5.58022403717041, + "learning_rate": 1.9955098024717515e-05, + "loss": 0.6018, + "step": 1881 + }, + { + "epoch": 0.05935972244125532, + "grad_norm": 4.599093437194824, + "learning_rate": 1.9955001273760287e-05, + "loss": 0.6286, + "step": 1882 + }, + { + "epoch": 0.05939126320769595, + "grad_norm": 5.156805038452148, + "learning_rate": 1.9954904418914888e-05, + "loss": 0.6626, + "step": 1883 + }, + { + "epoch": 0.05942280397413657, + "grad_norm": 3.813293933868408, + "learning_rate": 1.9954807460182334e-05, + "loss": 0.6643, + "step": 1884 + }, + { + "epoch": 0.059454344740577196, + "grad_norm": 3.4141740798950195, + "learning_rate": 1.995471039756363e-05, + "loss": 0.9239, + "step": 1885 + }, + { + "epoch": 0.05948588550701782, + "grad_norm": 5.615006923675537, + "learning_rate": 1.9954613231059796e-05, + "loss": 0.6221, + "step": 1886 + }, + { + "epoch": 0.05951742627345844, + "grad_norm": 3.0947184562683105, + "learning_rate": 1.9954515960671844e-05, + "loss": 0.6122, + "step": 1887 + }, + { + "epoch": 0.05954896703989907, + "grad_norm": 4.282245635986328, + "learning_rate": 1.9954418586400787e-05, + "loss": 0.6065, + "step": 1888 + }, + { + "epoch": 0.059580507806339694, + "grad_norm": 4.494782447814941, + "learning_rate": 1.9954321108247643e-05, + "loss": 0.6137, + "step": 1889 + }, + { + "epoch": 0.05961204857278032, + "grad_norm": 4.224121570587158, + "learning_rate": 1.995422352621343e-05, + "loss": 0.6021, + "step": 1890 + }, + { + "epoch": 0.05964358933922094, + "grad_norm": 4.579341411590576, + "learning_rate": 1.9954125840299165e-05, + "loss": 0.6364, + "step": 1891 + }, + { + "epoch": 0.05967513010566157, + "grad_norm": 3.7094364166259766, + "learning_rate": 1.9954028050505866e-05, + "loss": 0.6285, + "step": 1892 + }, + { + "epoch": 0.05970667087210219, + "grad_norm": 4.441507816314697, + "learning_rate": 1.9953930156834555e-05, + "loss": 0.6302, + "step": 1893 + }, + { + "epoch": 0.059738211638542815, + "grad_norm": 5.221348285675049, + "learning_rate": 1.9953832159286256e-05, + "loss": 0.6184, + "step": 1894 + }, + { + "epoch": 0.05976975240498344, + "grad_norm": 4.550767421722412, + "learning_rate": 1.9953734057861988e-05, + "loss": 0.6713, + "step": 1895 + }, + { + "epoch": 0.05980129317142407, + "grad_norm": 5.582978248596191, + "learning_rate": 1.995363585256278e-05, + "loss": 0.6184, + "step": 1896 + }, + { + "epoch": 0.05983283393786469, + "grad_norm": 4.280006408691406, + "learning_rate": 1.995353754338965e-05, + "loss": 0.6297, + "step": 1897 + }, + { + "epoch": 0.05986437470430531, + "grad_norm": 3.685018539428711, + "learning_rate": 1.9953439130343627e-05, + "loss": 0.6343, + "step": 1898 + }, + { + "epoch": 0.059895915470745936, + "grad_norm": 3.8153512477874756, + "learning_rate": 1.9953340613425736e-05, + "loss": 0.6553, + "step": 1899 + }, + { + "epoch": 0.059927456237186566, + "grad_norm": 4.488322734832764, + "learning_rate": 1.995324199263701e-05, + "loss": 0.6111, + "step": 1900 + }, + { + "epoch": 0.05995899700362719, + "grad_norm": 2.586520195007324, + "learning_rate": 1.9953143267978475e-05, + "loss": 0.43, + "step": 1901 + }, + { + "epoch": 0.05999053777006781, + "grad_norm": 6.175080299377441, + "learning_rate": 1.995304443945116e-05, + "loss": 0.6404, + "step": 1902 + }, + { + "epoch": 0.060022078536508434, + "grad_norm": 5.396096706390381, + "learning_rate": 1.9952945507056097e-05, + "loss": 0.6195, + "step": 1903 + }, + { + "epoch": 0.060053619302949064, + "grad_norm": 4.210594654083252, + "learning_rate": 1.995284647079432e-05, + "loss": 0.6189, + "step": 1904 + }, + { + "epoch": 0.06008516006938969, + "grad_norm": 3.310014247894287, + "learning_rate": 1.9952747330666865e-05, + "loss": 0.6098, + "step": 1905 + }, + { + "epoch": 0.06011670083583031, + "grad_norm": 6.794888019561768, + "learning_rate": 1.9952648086674765e-05, + "loss": 0.7288, + "step": 1906 + }, + { + "epoch": 0.06014824160227093, + "grad_norm": 3.51838755607605, + "learning_rate": 1.9952548738819047e-05, + "loss": 0.6128, + "step": 1907 + }, + { + "epoch": 0.06017978236871156, + "grad_norm": 8.025240898132324, + "learning_rate": 1.995244928710076e-05, + "loss": 0.6196, + "step": 1908 + }, + { + "epoch": 0.060211323135152185, + "grad_norm": 4.659548759460449, + "learning_rate": 1.9952349731520936e-05, + "loss": 0.6487, + "step": 1909 + }, + { + "epoch": 0.06024286390159281, + "grad_norm": 5.323477745056152, + "learning_rate": 1.9952250072080615e-05, + "loss": 0.6334, + "step": 1910 + }, + { + "epoch": 0.06027440466803343, + "grad_norm": 4.269820690155029, + "learning_rate": 1.9952150308780835e-05, + "loss": 0.655, + "step": 1911 + }, + { + "epoch": 0.06030594543447406, + "grad_norm": 3.694920539855957, + "learning_rate": 1.995205044162264e-05, + "loss": 0.6666, + "step": 1912 + }, + { + "epoch": 0.06033748620091468, + "grad_norm": 4.233449459075928, + "learning_rate": 1.9951950470607072e-05, + "loss": 0.6377, + "step": 1913 + }, + { + "epoch": 0.060369026967355306, + "grad_norm": 3.3764262199401855, + "learning_rate": 1.9951850395735173e-05, + "loss": 0.6433, + "step": 1914 + }, + { + "epoch": 0.06040056773379593, + "grad_norm": 4.50602912902832, + "learning_rate": 1.995175021700799e-05, + "loss": 0.6583, + "step": 1915 + }, + { + "epoch": 0.06043210850023656, + "grad_norm": 6.912373065948486, + "learning_rate": 1.995164993442656e-05, + "loss": 0.6401, + "step": 1916 + }, + { + "epoch": 0.06046364926667718, + "grad_norm": 4.030069351196289, + "learning_rate": 1.995154954799194e-05, + "loss": 0.6269, + "step": 1917 + }, + { + "epoch": 0.060495190033117804, + "grad_norm": 4.959309101104736, + "learning_rate": 1.995144905770517e-05, + "loss": 0.4045, + "step": 1918 + }, + { + "epoch": 0.06052673079955843, + "grad_norm": 4.155815601348877, + "learning_rate": 1.9951348463567303e-05, + "loss": 0.638, + "step": 1919 + }, + { + "epoch": 0.06055827156599906, + "grad_norm": 4.309327125549316, + "learning_rate": 1.995124776557939e-05, + "loss": 0.6699, + "step": 1920 + }, + { + "epoch": 0.06058981233243968, + "grad_norm": 4.537395000457764, + "learning_rate": 1.995114696374248e-05, + "loss": 0.652, + "step": 1921 + }, + { + "epoch": 0.0606213530988803, + "grad_norm": 3.5936429500579834, + "learning_rate": 1.9951046058057622e-05, + "loss": 0.6505, + "step": 1922 + }, + { + "epoch": 0.060652893865320925, + "grad_norm": 3.341522216796875, + "learning_rate": 1.9950945048525873e-05, + "loss": 0.3898, + "step": 1923 + }, + { + "epoch": 0.060684434631761555, + "grad_norm": 3.9513015747070312, + "learning_rate": 1.9950843935148287e-05, + "loss": 0.6485, + "step": 1924 + }, + { + "epoch": 0.06071597539820218, + "grad_norm": 3.7066879272460938, + "learning_rate": 1.9950742717925916e-05, + "loss": 0.6328, + "step": 1925 + }, + { + "epoch": 0.0607475161646428, + "grad_norm": 4.134316921234131, + "learning_rate": 1.995064139685982e-05, + "loss": 0.6178, + "step": 1926 + }, + { + "epoch": 0.060779056931083424, + "grad_norm": 3.4930570125579834, + "learning_rate": 1.9950539971951054e-05, + "loss": 0.6256, + "step": 1927 + }, + { + "epoch": 0.060810597697524046, + "grad_norm": 4.956609725952148, + "learning_rate": 1.9950438443200674e-05, + "loss": 0.6135, + "step": 1928 + }, + { + "epoch": 0.060842138463964676, + "grad_norm": 3.421062469482422, + "learning_rate": 1.9950336810609744e-05, + "loss": 0.6153, + "step": 1929 + }, + { + "epoch": 0.0608736792304053, + "grad_norm": 5.056326866149902, + "learning_rate": 1.9950235074179324e-05, + "loss": 0.6473, + "step": 1930 + }, + { + "epoch": 0.06090521999684592, + "grad_norm": 3.6515986919403076, + "learning_rate": 1.9950133233910477e-05, + "loss": 0.6181, + "step": 1931 + }, + { + "epoch": 0.060936760763286545, + "grad_norm": 3.713503122329712, + "learning_rate": 1.9950031289804258e-05, + "loss": 0.627, + "step": 1932 + }, + { + "epoch": 0.060968301529727174, + "grad_norm": 5.028148651123047, + "learning_rate": 1.994992924186174e-05, + "loss": 0.6444, + "step": 1933 + }, + { + "epoch": 0.0609998422961678, + "grad_norm": 3.3418948650360107, + "learning_rate": 1.9949827090083987e-05, + "loss": 0.6101, + "step": 1934 + }, + { + "epoch": 0.06103138306260842, + "grad_norm": 5.0494561195373535, + "learning_rate": 1.9949724834472057e-05, + "loss": 0.6312, + "step": 1935 + }, + { + "epoch": 0.06106292382904904, + "grad_norm": 5.271204948425293, + "learning_rate": 1.9949622475027027e-05, + "loss": 0.6756, + "step": 1936 + }, + { + "epoch": 0.06109446459548967, + "grad_norm": 4.651806354522705, + "learning_rate": 1.994952001174996e-05, + "loss": 0.6247, + "step": 1937 + }, + { + "epoch": 0.061126005361930295, + "grad_norm": 4.455158233642578, + "learning_rate": 1.994941744464192e-05, + "loss": 0.6145, + "step": 1938 + }, + { + "epoch": 0.06115754612837092, + "grad_norm": 3.1404542922973633, + "learning_rate": 1.994931477370399e-05, + "loss": 0.4013, + "step": 1939 + }, + { + "epoch": 0.06118908689481154, + "grad_norm": 3.6257212162017822, + "learning_rate": 1.9949211998937236e-05, + "loss": 0.6337, + "step": 1940 + }, + { + "epoch": 0.06122062766125217, + "grad_norm": 3.5700955390930176, + "learning_rate": 1.9949109120342728e-05, + "loss": 0.9205, + "step": 1941 + }, + { + "epoch": 0.061252168427692794, + "grad_norm": 5.055789947509766, + "learning_rate": 1.994900613792154e-05, + "loss": 0.6641, + "step": 1942 + }, + { + "epoch": 0.061283709194133416, + "grad_norm": 4.575700759887695, + "learning_rate": 1.9948903051674745e-05, + "loss": 0.6046, + "step": 1943 + }, + { + "epoch": 0.06131524996057404, + "grad_norm": 5.103704452514648, + "learning_rate": 1.9948799861603423e-05, + "loss": 0.6376, + "step": 1944 + }, + { + "epoch": 0.06134679072701467, + "grad_norm": 4.0730414390563965, + "learning_rate": 1.9948696567708654e-05, + "loss": 0.6635, + "step": 1945 + }, + { + "epoch": 0.06137833149345529, + "grad_norm": 6.466216087341309, + "learning_rate": 1.9948593169991505e-05, + "loss": 0.3894, + "step": 1946 + }, + { + "epoch": 0.061409872259895915, + "grad_norm": 2.4805755615234375, + "learning_rate": 1.9948489668453066e-05, + "loss": 0.8757, + "step": 1947 + }, + { + "epoch": 0.06144141302633654, + "grad_norm": 5.702544212341309, + "learning_rate": 1.9948386063094412e-05, + "loss": 0.8978, + "step": 1948 + }, + { + "epoch": 0.06147295379277717, + "grad_norm": 5.469177722930908, + "learning_rate": 1.9948282353916624e-05, + "loss": 0.6747, + "step": 1949 + }, + { + "epoch": 0.06150449455921779, + "grad_norm": 5.9206624031066895, + "learning_rate": 1.9948178540920783e-05, + "loss": 0.6582, + "step": 1950 + }, + { + "epoch": 0.06153603532565841, + "grad_norm": 5.018537998199463, + "learning_rate": 1.9948074624107975e-05, + "loss": 0.6322, + "step": 1951 + }, + { + "epoch": 0.061567576092099036, + "grad_norm": 4.884835243225098, + "learning_rate": 1.9947970603479286e-05, + "loss": 0.4227, + "step": 1952 + }, + { + "epoch": 0.061599116858539665, + "grad_norm": 6.000855922698975, + "learning_rate": 1.9947866479035797e-05, + "loss": 0.6827, + "step": 1953 + }, + { + "epoch": 0.06163065762498029, + "grad_norm": 5.103408336639404, + "learning_rate": 1.99477622507786e-05, + "loss": 0.6795, + "step": 1954 + }, + { + "epoch": 0.06166219839142091, + "grad_norm": 2.892627000808716, + "learning_rate": 1.9947657918708775e-05, + "loss": 0.8879, + "step": 1955 + }, + { + "epoch": 0.061693739157861534, + "grad_norm": 5.822535037994385, + "learning_rate": 1.994755348282742e-05, + "loss": 0.7192, + "step": 1956 + }, + { + "epoch": 0.061725279924302164, + "grad_norm": 6.206089496612549, + "learning_rate": 1.9947448943135618e-05, + "loss": 0.6532, + "step": 1957 + }, + { + "epoch": 0.061756820690742786, + "grad_norm": 5.756224632263184, + "learning_rate": 1.9947344299634464e-05, + "loss": 0.4423, + "step": 1958 + }, + { + "epoch": 0.06178836145718341, + "grad_norm": 3.0195398330688477, + "learning_rate": 1.994723955232505e-05, + "loss": 0.9247, + "step": 1959 + }, + { + "epoch": 0.06181990222362403, + "grad_norm": 4.568525314331055, + "learning_rate": 1.994713470120846e-05, + "loss": 0.637, + "step": 1960 + }, + { + "epoch": 0.06185144299006466, + "grad_norm": 4.75670862197876, + "learning_rate": 1.99470297462858e-05, + "loss": 0.6437, + "step": 1961 + }, + { + "epoch": 0.061882983756505285, + "grad_norm": 3.3700320720672607, + "learning_rate": 1.994692468755816e-05, + "loss": 0.9116, + "step": 1962 + }, + { + "epoch": 0.06191452452294591, + "grad_norm": 5.248836040496826, + "learning_rate": 1.9946819525026642e-05, + "loss": 0.6438, + "step": 1963 + }, + { + "epoch": 0.06194606528938653, + "grad_norm": 7.702510833740234, + "learning_rate": 1.9946714258692334e-05, + "loss": 0.4261, + "step": 1964 + }, + { + "epoch": 0.06197760605582716, + "grad_norm": 5.193170070648193, + "learning_rate": 1.9946608888556344e-05, + "loss": 0.4053, + "step": 1965 + }, + { + "epoch": 0.06200914682226778, + "grad_norm": 6.684542179107666, + "learning_rate": 1.994650341461976e-05, + "loss": 0.9558, + "step": 1966 + }, + { + "epoch": 0.062040687588708406, + "grad_norm": 5.754485607147217, + "learning_rate": 1.9946397836883695e-05, + "loss": 0.7079, + "step": 1967 + }, + { + "epoch": 0.06207222835514903, + "grad_norm": 8.872687339782715, + "learning_rate": 1.994629215534924e-05, + "loss": 0.6591, + "step": 1968 + }, + { + "epoch": 0.06210376912158965, + "grad_norm": 3.9076218605041504, + "learning_rate": 1.9946186370017507e-05, + "loss": 0.6383, + "step": 1969 + }, + { + "epoch": 0.06213530988803028, + "grad_norm": 3.5052103996276855, + "learning_rate": 1.9946080480889597e-05, + "loss": 0.6682, + "step": 1970 + }, + { + "epoch": 0.062166850654470904, + "grad_norm": 4.356656074523926, + "learning_rate": 1.9945974487966615e-05, + "loss": 0.6517, + "step": 1971 + }, + { + "epoch": 0.06219839142091153, + "grad_norm": 5.698628902435303, + "learning_rate": 1.9945868391249662e-05, + "loss": 0.6563, + "step": 1972 + }, + { + "epoch": 0.06222993218735215, + "grad_norm": 2.340207576751709, + "learning_rate": 1.9945762190739854e-05, + "loss": 0.9002, + "step": 1973 + }, + { + "epoch": 0.06226147295379278, + "grad_norm": 5.969751834869385, + "learning_rate": 1.994565588643829e-05, + "loss": 0.6628, + "step": 1974 + }, + { + "epoch": 0.0622930137202334, + "grad_norm": 4.037098407745361, + "learning_rate": 1.994554947834609e-05, + "loss": 0.6568, + "step": 1975 + }, + { + "epoch": 0.062324554486674025, + "grad_norm": 3.759995460510254, + "learning_rate": 1.9945442966464354e-05, + "loss": 0.6196, + "step": 1976 + }, + { + "epoch": 0.06235609525311465, + "grad_norm": 5.268688678741455, + "learning_rate": 1.99453363507942e-05, + "loss": 0.6368, + "step": 1977 + }, + { + "epoch": 0.06238763601955528, + "grad_norm": 4.604568958282471, + "learning_rate": 1.994522963133674e-05, + "loss": 0.6153, + "step": 1978 + }, + { + "epoch": 0.0624191767859959, + "grad_norm": 7.656871318817139, + "learning_rate": 1.9945122808093088e-05, + "loss": 0.6641, + "step": 1979 + }, + { + "epoch": 0.06245071755243652, + "grad_norm": 4.691308498382568, + "learning_rate": 1.9945015881064353e-05, + "loss": 0.6439, + "step": 1980 + }, + { + "epoch": 0.062482258318877146, + "grad_norm": 5.45802640914917, + "learning_rate": 1.9944908850251657e-05, + "loss": 0.9426, + "step": 1981 + }, + { + "epoch": 0.06251379908531778, + "grad_norm": 7.4099249839782715, + "learning_rate": 1.9944801715656113e-05, + "loss": 0.662, + "step": 1982 + }, + { + "epoch": 0.06254533985175839, + "grad_norm": 5.308866024017334, + "learning_rate": 1.9944694477278843e-05, + "loss": 0.6285, + "step": 1983 + }, + { + "epoch": 0.06257688061819902, + "grad_norm": 5.090230464935303, + "learning_rate": 1.9944587135120966e-05, + "loss": 0.6087, + "step": 1984 + }, + { + "epoch": 0.06260842138463965, + "grad_norm": 4.14413595199585, + "learning_rate": 1.9944479689183602e-05, + "loss": 0.6439, + "step": 1985 + }, + { + "epoch": 0.06263996215108027, + "grad_norm": 5.893408298492432, + "learning_rate": 1.9944372139467866e-05, + "loss": 0.6249, + "step": 1986 + }, + { + "epoch": 0.0626715029175209, + "grad_norm": 5.273472785949707, + "learning_rate": 1.9944264485974884e-05, + "loss": 0.6858, + "step": 1987 + }, + { + "epoch": 0.06270304368396153, + "grad_norm": 2.900322437286377, + "learning_rate": 1.9944156728705786e-05, + "loss": 0.9018, + "step": 1988 + }, + { + "epoch": 0.06273458445040214, + "grad_norm": 5.079473972320557, + "learning_rate": 1.9944048867661688e-05, + "loss": 0.623, + "step": 1989 + }, + { + "epoch": 0.06276612521684277, + "grad_norm": 3.381896495819092, + "learning_rate": 1.9943940902843718e-05, + "loss": 0.4072, + "step": 1990 + }, + { + "epoch": 0.06279766598328339, + "grad_norm": 3.867826461791992, + "learning_rate": 1.9943832834253003e-05, + "loss": 0.6568, + "step": 1991 + }, + { + "epoch": 0.06282920674972402, + "grad_norm": 2.79996395111084, + "learning_rate": 1.9943724661890674e-05, + "loss": 0.9426, + "step": 1992 + }, + { + "epoch": 0.06286074751616465, + "grad_norm": 4.308008670806885, + "learning_rate": 1.9943616385757856e-05, + "loss": 0.6456, + "step": 1993 + }, + { + "epoch": 0.06289228828260526, + "grad_norm": 5.69791316986084, + "learning_rate": 1.994350800585568e-05, + "loss": 0.6326, + "step": 1994 + }, + { + "epoch": 0.06292382904904589, + "grad_norm": 4.522645473480225, + "learning_rate": 1.994339952218528e-05, + "loss": 0.6548, + "step": 1995 + }, + { + "epoch": 0.06295536981548652, + "grad_norm": 4.939061164855957, + "learning_rate": 1.9943290934747778e-05, + "loss": 0.6403, + "step": 1996 + }, + { + "epoch": 0.06298691058192714, + "grad_norm": 3.3947434425354004, + "learning_rate": 1.994318224354432e-05, + "loss": 0.8871, + "step": 1997 + }, + { + "epoch": 0.06301845134836777, + "grad_norm": 3.8913192749023438, + "learning_rate": 1.994307344857603e-05, + "loss": 0.6335, + "step": 1998 + }, + { + "epoch": 0.06304999211480838, + "grad_norm": 4.683254241943359, + "learning_rate": 1.9942964549844052e-05, + "loss": 0.632, + "step": 1999 + }, + { + "epoch": 0.06308153288124901, + "grad_norm": 5.526060581207275, + "learning_rate": 1.9942855547349518e-05, + "loss": 0.6165, + "step": 2000 + }, + { + "epoch": 0.06311307364768964, + "grad_norm": 3.974426507949829, + "learning_rate": 1.9942746441093567e-05, + "loss": 0.6362, + "step": 2001 + }, + { + "epoch": 0.06314461441413026, + "grad_norm": 5.210516452789307, + "learning_rate": 1.9942637231077333e-05, + "loss": 0.6316, + "step": 2002 + }, + { + "epoch": 0.06317615518057089, + "grad_norm": 3.69130802154541, + "learning_rate": 1.9942527917301962e-05, + "loss": 0.6521, + "step": 2003 + }, + { + "epoch": 0.06320769594701152, + "grad_norm": 6.183731555938721, + "learning_rate": 1.994241849976859e-05, + "loss": 0.6354, + "step": 2004 + }, + { + "epoch": 0.06323923671345213, + "grad_norm": 2.689981460571289, + "learning_rate": 1.994230897847836e-05, + "loss": 0.9173, + "step": 2005 + }, + { + "epoch": 0.06327077747989276, + "grad_norm": 5.594850063323975, + "learning_rate": 1.994219935343242e-05, + "loss": 0.6646, + "step": 2006 + }, + { + "epoch": 0.06330231824633338, + "grad_norm": 5.019543170928955, + "learning_rate": 1.9942089624631907e-05, + "loss": 0.6028, + "step": 2007 + }, + { + "epoch": 0.06333385901277401, + "grad_norm": 4.517858028411865, + "learning_rate": 1.994197979207797e-05, + "loss": 0.647, + "step": 2008 + }, + { + "epoch": 0.06336539977921464, + "grad_norm": 2.722942590713501, + "learning_rate": 1.994186985577175e-05, + "loss": 0.8957, + "step": 2009 + }, + { + "epoch": 0.06339694054565526, + "grad_norm": 2.9508156776428223, + "learning_rate": 1.9941759815714406e-05, + "loss": 0.9282, + "step": 2010 + }, + { + "epoch": 0.06342848131209589, + "grad_norm": 5.840066909790039, + "learning_rate": 1.9941649671907072e-05, + "loss": 0.6668, + "step": 2011 + }, + { + "epoch": 0.0634600220785365, + "grad_norm": 5.418063163757324, + "learning_rate": 1.9941539424350908e-05, + "loss": 0.6304, + "step": 2012 + }, + { + "epoch": 0.06349156284497713, + "grad_norm": 6.092320442199707, + "learning_rate": 1.9941429073047062e-05, + "loss": 0.645, + "step": 2013 + }, + { + "epoch": 0.06352310361141776, + "grad_norm": 5.847226142883301, + "learning_rate": 1.9941318617996683e-05, + "loss": 0.6474, + "step": 2014 + }, + { + "epoch": 0.06355464437785838, + "grad_norm": 3.1134960651397705, + "learning_rate": 1.994120805920092e-05, + "loss": 0.3975, + "step": 2015 + }, + { + "epoch": 0.063586185144299, + "grad_norm": 4.4606032371521, + "learning_rate": 1.9941097396660938e-05, + "loss": 0.6097, + "step": 2016 + }, + { + "epoch": 0.06361772591073964, + "grad_norm": 5.080752849578857, + "learning_rate": 1.994098663037788e-05, + "loss": 0.6281, + "step": 2017 + }, + { + "epoch": 0.06364926667718025, + "grad_norm": 5.950139045715332, + "learning_rate": 1.994087576035291e-05, + "loss": 0.6849, + "step": 2018 + }, + { + "epoch": 0.06368080744362088, + "grad_norm": 3.693599224090576, + "learning_rate": 1.9940764786587184e-05, + "loss": 0.4101, + "step": 2019 + }, + { + "epoch": 0.0637123482100615, + "grad_norm": 6.256530284881592, + "learning_rate": 1.994065370908186e-05, + "loss": 0.6231, + "step": 2020 + }, + { + "epoch": 0.06374388897650213, + "grad_norm": 4.477221965789795, + "learning_rate": 1.9940542527838093e-05, + "loss": 0.6481, + "step": 2021 + }, + { + "epoch": 0.06377542974294276, + "grad_norm": 4.3439178466796875, + "learning_rate": 1.9940431242857046e-05, + "loss": 0.6159, + "step": 2022 + }, + { + "epoch": 0.06380697050938337, + "grad_norm": 5.108736991882324, + "learning_rate": 1.9940319854139884e-05, + "loss": 0.6165, + "step": 2023 + }, + { + "epoch": 0.063838511275824, + "grad_norm": 5.477535247802734, + "learning_rate": 1.9940208361687762e-05, + "loss": 0.5976, + "step": 2024 + }, + { + "epoch": 0.06387005204226463, + "grad_norm": 5.328495979309082, + "learning_rate": 1.994009676550185e-05, + "loss": 0.635, + "step": 2025 + }, + { + "epoch": 0.06390159280870525, + "grad_norm": 4.461034297943115, + "learning_rate": 1.9939985065583306e-05, + "loss": 0.6119, + "step": 2026 + }, + { + "epoch": 0.06393313357514588, + "grad_norm": 7.440424919128418, + "learning_rate": 1.9939873261933306e-05, + "loss": 0.6507, + "step": 2027 + }, + { + "epoch": 0.0639646743415865, + "grad_norm": 8.196084976196289, + "learning_rate": 1.9939761354553004e-05, + "loss": 0.6566, + "step": 2028 + }, + { + "epoch": 0.06399621510802712, + "grad_norm": 6.956246376037598, + "learning_rate": 1.9939649343443578e-05, + "loss": 0.6612, + "step": 2029 + }, + { + "epoch": 0.06402775587446775, + "grad_norm": 6.424142837524414, + "learning_rate": 1.993953722860619e-05, + "loss": 0.6637, + "step": 2030 + }, + { + "epoch": 0.06405929664090837, + "grad_norm": 5.089202880859375, + "learning_rate": 1.9939425010042015e-05, + "loss": 0.6408, + "step": 2031 + }, + { + "epoch": 0.064090837407349, + "grad_norm": 4.575745582580566, + "learning_rate": 1.993931268775222e-05, + "loss": 0.659, + "step": 2032 + }, + { + "epoch": 0.06412237817378963, + "grad_norm": 5.529644012451172, + "learning_rate": 1.9939200261737986e-05, + "loss": 0.626, + "step": 2033 + }, + { + "epoch": 0.06415391894023025, + "grad_norm": 4.309192180633545, + "learning_rate": 1.9939087732000474e-05, + "loss": 0.6262, + "step": 2034 + }, + { + "epoch": 0.06418545970667087, + "grad_norm": 5.354945659637451, + "learning_rate": 1.9938975098540867e-05, + "loss": 0.6638, + "step": 2035 + }, + { + "epoch": 0.06421700047311149, + "grad_norm": 5.667020797729492, + "learning_rate": 1.9938862361360335e-05, + "loss": 0.6528, + "step": 2036 + }, + { + "epoch": 0.06424854123955212, + "grad_norm": 4.522233009338379, + "learning_rate": 1.9938749520460055e-05, + "loss": 0.5996, + "step": 2037 + }, + { + "epoch": 0.06428008200599275, + "grad_norm": 4.959874153137207, + "learning_rate": 1.993863657584121e-05, + "loss": 0.6325, + "step": 2038 + }, + { + "epoch": 0.06431162277243337, + "grad_norm": 5.210136890411377, + "learning_rate": 1.9938523527504976e-05, + "loss": 0.6379, + "step": 2039 + }, + { + "epoch": 0.064343163538874, + "grad_norm": 4.727824687957764, + "learning_rate": 1.993841037545253e-05, + "loss": 0.6094, + "step": 2040 + }, + { + "epoch": 0.06437470430531463, + "grad_norm": 4.298110485076904, + "learning_rate": 1.9938297119685054e-05, + "loss": 0.618, + "step": 2041 + }, + { + "epoch": 0.06440624507175524, + "grad_norm": 4.984130859375, + "learning_rate": 1.9938183760203734e-05, + "loss": 0.6357, + "step": 2042 + }, + { + "epoch": 0.06443778583819587, + "grad_norm": 5.766677379608154, + "learning_rate": 1.9938070297009748e-05, + "loss": 0.6291, + "step": 2043 + }, + { + "epoch": 0.06446932660463649, + "grad_norm": 6.258183002471924, + "learning_rate": 1.9937956730104278e-05, + "loss": 0.602, + "step": 2044 + }, + { + "epoch": 0.06450086737107712, + "grad_norm": 5.022643566131592, + "learning_rate": 1.9937843059488518e-05, + "loss": 0.6336, + "step": 2045 + }, + { + "epoch": 0.06453240813751775, + "grad_norm": 5.502520561218262, + "learning_rate": 1.9937729285163647e-05, + "loss": 0.6211, + "step": 2046 + }, + { + "epoch": 0.06456394890395836, + "grad_norm": 9.944141387939453, + "learning_rate": 1.9937615407130856e-05, + "loss": 0.6502, + "step": 2047 + }, + { + "epoch": 0.06459548967039899, + "grad_norm": 5.584178447723389, + "learning_rate": 1.993750142539133e-05, + "loss": 0.9416, + "step": 2048 + }, + { + "epoch": 0.06462703043683962, + "grad_norm": 5.985225200653076, + "learning_rate": 1.9937387339946263e-05, + "loss": 0.641, + "step": 2049 + }, + { + "epoch": 0.06465857120328024, + "grad_norm": 8.419351577758789, + "learning_rate": 1.993727315079684e-05, + "loss": 0.6497, + "step": 2050 + }, + { + "epoch": 0.06469011196972087, + "grad_norm": 6.160191059112549, + "learning_rate": 1.9937158857944258e-05, + "loss": 0.6437, + "step": 2051 + }, + { + "epoch": 0.06472165273616148, + "grad_norm": 2.6530001163482666, + "learning_rate": 1.9937044461389705e-05, + "loss": 0.9331, + "step": 2052 + }, + { + "epoch": 0.06475319350260211, + "grad_norm": 4.729665279388428, + "learning_rate": 1.993692996113438e-05, + "loss": 0.6251, + "step": 2053 + }, + { + "epoch": 0.06478473426904274, + "grad_norm": 13.221940040588379, + "learning_rate": 1.9936815357179472e-05, + "loss": 0.5949, + "step": 2054 + }, + { + "epoch": 0.06481627503548336, + "grad_norm": 4.558366298675537, + "learning_rate": 1.993670064952618e-05, + "loss": 0.6153, + "step": 2055 + }, + { + "epoch": 0.06484781580192399, + "grad_norm": 5.3050537109375, + "learning_rate": 1.99365858381757e-05, + "loss": 0.6063, + "step": 2056 + }, + { + "epoch": 0.0648793565683646, + "grad_norm": 4.712823867797852, + "learning_rate": 1.9936470923129234e-05, + "loss": 0.6276, + "step": 2057 + }, + { + "epoch": 0.06491089733480523, + "grad_norm": 4.932109832763672, + "learning_rate": 1.993635590438798e-05, + "loss": 0.6378, + "step": 2058 + }, + { + "epoch": 0.06494243810124586, + "grad_norm": 4.728148937225342, + "learning_rate": 1.9936240781953133e-05, + "loss": 0.9405, + "step": 2059 + }, + { + "epoch": 0.06497397886768648, + "grad_norm": 6.675343036651611, + "learning_rate": 1.99361255558259e-05, + "loss": 0.6701, + "step": 2060 + }, + { + "epoch": 0.06500551963412711, + "grad_norm": 9.697712898254395, + "learning_rate": 1.993601022600748e-05, + "loss": 0.6692, + "step": 2061 + }, + { + "epoch": 0.06503706040056774, + "grad_norm": 7.6763787269592285, + "learning_rate": 1.993589479249908e-05, + "loss": 0.6558, + "step": 2062 + }, + { + "epoch": 0.06506860116700836, + "grad_norm": 5.722901344299316, + "learning_rate": 1.9935779255301902e-05, + "loss": 0.6384, + "step": 2063 + }, + { + "epoch": 0.06510014193344899, + "grad_norm": 6.282075881958008, + "learning_rate": 1.9935663614417153e-05, + "loss": 0.6428, + "step": 2064 + }, + { + "epoch": 0.0651316826998896, + "grad_norm": 3.0482754707336426, + "learning_rate": 1.993554786984604e-05, + "loss": 0.8993, + "step": 2065 + }, + { + "epoch": 0.06516322346633023, + "grad_norm": 4.502875328063965, + "learning_rate": 1.993543202158977e-05, + "loss": 0.6262, + "step": 2066 + }, + { + "epoch": 0.06519476423277086, + "grad_norm": 3.2337775230407715, + "learning_rate": 1.993531606964955e-05, + "loss": 0.9322, + "step": 2067 + }, + { + "epoch": 0.06522630499921148, + "grad_norm": 4.218539714813232, + "learning_rate": 1.9935200014026593e-05, + "loss": 0.6435, + "step": 2068 + }, + { + "epoch": 0.0652578457656521, + "grad_norm": 3.47890567779541, + "learning_rate": 1.993508385472211e-05, + "loss": 0.3982, + "step": 2069 + }, + { + "epoch": 0.06528938653209274, + "grad_norm": 3.503009557723999, + "learning_rate": 1.9934967591737316e-05, + "loss": 0.3919, + "step": 2070 + }, + { + "epoch": 0.06532092729853335, + "grad_norm": 8.289449691772461, + "learning_rate": 1.9934851225073416e-05, + "loss": 0.6399, + "step": 2071 + }, + { + "epoch": 0.06535246806497398, + "grad_norm": 4.969714164733887, + "learning_rate": 1.9934734754731632e-05, + "loss": 0.6327, + "step": 2072 + }, + { + "epoch": 0.0653840088314146, + "grad_norm": 4.312224864959717, + "learning_rate": 1.9934618180713175e-05, + "loss": 0.6406, + "step": 2073 + }, + { + "epoch": 0.06541554959785523, + "grad_norm": 5.398203372955322, + "learning_rate": 1.9934501503019264e-05, + "loss": 0.6382, + "step": 2074 + }, + { + "epoch": 0.06544709036429586, + "grad_norm": 5.305398464202881, + "learning_rate": 1.9934384721651115e-05, + "loss": 0.6277, + "step": 2075 + }, + { + "epoch": 0.06547863113073647, + "grad_norm": 4.434843063354492, + "learning_rate": 1.993426783660995e-05, + "loss": 0.8686, + "step": 2076 + }, + { + "epoch": 0.0655101718971771, + "grad_norm": 5.588466644287109, + "learning_rate": 1.993415084789698e-05, + "loss": 0.62, + "step": 2077 + }, + { + "epoch": 0.06554171266361773, + "grad_norm": 9.812552452087402, + "learning_rate": 1.993403375551344e-05, + "loss": 0.6119, + "step": 2078 + }, + { + "epoch": 0.06557325343005835, + "grad_norm": 5.271746635437012, + "learning_rate": 1.993391655946054e-05, + "loss": 0.6618, + "step": 2079 + }, + { + "epoch": 0.06560479419649898, + "grad_norm": 4.188797473907471, + "learning_rate": 1.9933799259739512e-05, + "loss": 0.6416, + "step": 2080 + }, + { + "epoch": 0.0656363349629396, + "grad_norm": 5.85979700088501, + "learning_rate": 1.993368185635157e-05, + "loss": 0.6852, + "step": 2081 + }, + { + "epoch": 0.06566787572938022, + "grad_norm": 6.56433629989624, + "learning_rate": 1.9933564349297946e-05, + "loss": 0.6239, + "step": 2082 + }, + { + "epoch": 0.06569941649582085, + "grad_norm": 9.022429466247559, + "learning_rate": 1.9933446738579866e-05, + "loss": 0.6346, + "step": 2083 + }, + { + "epoch": 0.06573095726226147, + "grad_norm": 4.560171604156494, + "learning_rate": 1.9933329024198553e-05, + "loss": 0.6256, + "step": 2084 + }, + { + "epoch": 0.0657624980287021, + "grad_norm": 3.6774260997772217, + "learning_rate": 1.993321120615524e-05, + "loss": 0.4291, + "step": 2085 + }, + { + "epoch": 0.06579403879514273, + "grad_norm": 4.567025661468506, + "learning_rate": 1.9933093284451158e-05, + "loss": 0.6727, + "step": 2086 + }, + { + "epoch": 0.06582557956158334, + "grad_norm": 4.851759433746338, + "learning_rate": 1.9932975259087533e-05, + "loss": 0.627, + "step": 2087 + }, + { + "epoch": 0.06585712032802397, + "grad_norm": 3.2105584144592285, + "learning_rate": 1.9932857130065595e-05, + "loss": 0.8604, + "step": 2088 + }, + { + "epoch": 0.06588866109446459, + "grad_norm": 5.504612922668457, + "learning_rate": 1.9932738897386583e-05, + "loss": 0.6177, + "step": 2089 + }, + { + "epoch": 0.06592020186090522, + "grad_norm": 5.238157272338867, + "learning_rate": 1.9932620561051728e-05, + "loss": 0.6953, + "step": 2090 + }, + { + "epoch": 0.06595174262734585, + "grad_norm": 5.818564414978027, + "learning_rate": 1.9932502121062267e-05, + "loss": 0.6577, + "step": 2091 + }, + { + "epoch": 0.06598328339378647, + "grad_norm": 4.07849645614624, + "learning_rate": 1.9932383577419432e-05, + "loss": 0.6227, + "step": 2092 + }, + { + "epoch": 0.0660148241602271, + "grad_norm": 11.162467002868652, + "learning_rate": 1.993226493012446e-05, + "loss": 0.6556, + "step": 2093 + }, + { + "epoch": 0.06604636492666772, + "grad_norm": 4.270259380340576, + "learning_rate": 1.9932146179178596e-05, + "loss": 0.6175, + "step": 2094 + }, + { + "epoch": 0.06607790569310834, + "grad_norm": 4.571337699890137, + "learning_rate": 1.993202732458307e-05, + "loss": 0.6556, + "step": 2095 + }, + { + "epoch": 0.06610944645954897, + "grad_norm": 5.142566204071045, + "learning_rate": 1.993190836633913e-05, + "loss": 0.6475, + "step": 2096 + }, + { + "epoch": 0.06614098722598959, + "grad_norm": 2.9304943084716797, + "learning_rate": 1.9931789304448013e-05, + "loss": 0.9472, + "step": 2097 + }, + { + "epoch": 0.06617252799243022, + "grad_norm": 4.089705467224121, + "learning_rate": 1.9931670138910963e-05, + "loss": 0.6635, + "step": 2098 + }, + { + "epoch": 0.06620406875887085, + "grad_norm": 2.693009376525879, + "learning_rate": 1.9931550869729223e-05, + "loss": 0.8765, + "step": 2099 + }, + { + "epoch": 0.06623560952531146, + "grad_norm": 4.270763874053955, + "learning_rate": 1.9931431496904037e-05, + "loss": 0.62, + "step": 2100 + }, + { + "epoch": 0.06626715029175209, + "grad_norm": 5.494943618774414, + "learning_rate": 1.9931312020436655e-05, + "loss": 0.678, + "step": 2101 + }, + { + "epoch": 0.06629869105819271, + "grad_norm": 3.5421741008758545, + "learning_rate": 1.993119244032832e-05, + "loss": 0.6507, + "step": 2102 + }, + { + "epoch": 0.06633023182463334, + "grad_norm": 4.419683456420898, + "learning_rate": 1.9931072756580277e-05, + "loss": 0.6042, + "step": 2103 + }, + { + "epoch": 0.06636177259107397, + "grad_norm": 4.587719917297363, + "learning_rate": 1.993095296919378e-05, + "loss": 0.6515, + "step": 2104 + }, + { + "epoch": 0.06639331335751458, + "grad_norm": 4.721997261047363, + "learning_rate": 1.993083307817008e-05, + "loss": 0.6444, + "step": 2105 + }, + { + "epoch": 0.06642485412395521, + "grad_norm": 5.450900077819824, + "learning_rate": 1.9930713083510423e-05, + "loss": 0.6843, + "step": 2106 + }, + { + "epoch": 0.06645639489039584, + "grad_norm": 5.848992347717285, + "learning_rate": 1.9930592985216063e-05, + "loss": 0.6164, + "step": 2107 + }, + { + "epoch": 0.06648793565683646, + "grad_norm": 4.991812229156494, + "learning_rate": 1.9930472783288258e-05, + "loss": 0.6307, + "step": 2108 + }, + { + "epoch": 0.06651947642327709, + "grad_norm": 6.262106418609619, + "learning_rate": 1.993035247772826e-05, + "loss": 0.6391, + "step": 2109 + }, + { + "epoch": 0.0665510171897177, + "grad_norm": 4.315780162811279, + "learning_rate": 1.9930232068537317e-05, + "loss": 0.6333, + "step": 2110 + }, + { + "epoch": 0.06658255795615833, + "grad_norm": 6.329419136047363, + "learning_rate": 1.9930111555716695e-05, + "loss": 0.6641, + "step": 2111 + }, + { + "epoch": 0.06661409872259896, + "grad_norm": 4.631668567657471, + "learning_rate": 1.992999093926765e-05, + "loss": 0.6158, + "step": 2112 + }, + { + "epoch": 0.06664563948903958, + "grad_norm": 4.981015205383301, + "learning_rate": 1.9929870219191434e-05, + "loss": 0.6641, + "step": 2113 + }, + { + "epoch": 0.06667718025548021, + "grad_norm": 5.097248077392578, + "learning_rate": 1.9929749395489316e-05, + "loss": 0.6455, + "step": 2114 + }, + { + "epoch": 0.06670872102192084, + "grad_norm": 4.272308826446533, + "learning_rate": 1.9929628468162553e-05, + "loss": 0.6536, + "step": 2115 + }, + { + "epoch": 0.06674026178836145, + "grad_norm": 3.5017919540405273, + "learning_rate": 1.9929507437212405e-05, + "loss": 0.6684, + "step": 2116 + }, + { + "epoch": 0.06677180255480208, + "grad_norm": 4.6017937660217285, + "learning_rate": 1.9929386302640136e-05, + "loss": 0.6378, + "step": 2117 + }, + { + "epoch": 0.0668033433212427, + "grad_norm": 8.30997085571289, + "learning_rate": 1.9929265064447017e-05, + "loss": 0.6423, + "step": 2118 + }, + { + "epoch": 0.06683488408768333, + "grad_norm": 4.011048793792725, + "learning_rate": 1.99291437226343e-05, + "loss": 0.6064, + "step": 2119 + }, + { + "epoch": 0.06686642485412396, + "grad_norm": 6.074185371398926, + "learning_rate": 1.9929022277203264e-05, + "loss": 0.5794, + "step": 2120 + }, + { + "epoch": 0.06689796562056458, + "grad_norm": 3.5834801197052, + "learning_rate": 1.9928900728155165e-05, + "loss": 0.6604, + "step": 2121 + }, + { + "epoch": 0.0669295063870052, + "grad_norm": 12.238226890563965, + "learning_rate": 1.992877907549128e-05, + "loss": 0.5939, + "step": 2122 + }, + { + "epoch": 0.06696104715344584, + "grad_norm": 3.700932741165161, + "learning_rate": 1.9928657319212877e-05, + "loss": 0.6457, + "step": 2123 + }, + { + "epoch": 0.06699258791988645, + "grad_norm": 4.204671859741211, + "learning_rate": 1.9928535459321227e-05, + "loss": 0.5826, + "step": 2124 + }, + { + "epoch": 0.06702412868632708, + "grad_norm": 3.732499122619629, + "learning_rate": 1.9928413495817596e-05, + "loss": 0.4296, + "step": 2125 + }, + { + "epoch": 0.0670556694527677, + "grad_norm": 4.058485984802246, + "learning_rate": 1.9928291428703265e-05, + "loss": 0.6322, + "step": 2126 + }, + { + "epoch": 0.06708721021920833, + "grad_norm": 5.033778190612793, + "learning_rate": 1.99281692579795e-05, + "loss": 0.6652, + "step": 2127 + }, + { + "epoch": 0.06711875098564896, + "grad_norm": 2.8452389240264893, + "learning_rate": 1.9928046983647583e-05, + "loss": 0.9235, + "step": 2128 + }, + { + "epoch": 0.06715029175208957, + "grad_norm": 3.7794597148895264, + "learning_rate": 1.9927924605708783e-05, + "loss": 0.6107, + "step": 2129 + }, + { + "epoch": 0.0671818325185302, + "grad_norm": 6.243109703063965, + "learning_rate": 1.9927802124164383e-05, + "loss": 0.6464, + "step": 2130 + }, + { + "epoch": 0.06721337328497083, + "grad_norm": 5.503244876861572, + "learning_rate": 1.9927679539015662e-05, + "loss": 0.5908, + "step": 2131 + }, + { + "epoch": 0.06724491405141145, + "grad_norm": 3.4696578979492188, + "learning_rate": 1.9927556850263892e-05, + "loss": 0.5824, + "step": 2132 + }, + { + "epoch": 0.06727645481785208, + "grad_norm": 3.952816963195801, + "learning_rate": 1.992743405791036e-05, + "loss": 0.6207, + "step": 2133 + }, + { + "epoch": 0.06730799558429269, + "grad_norm": 3.5226986408233643, + "learning_rate": 1.9927311161956343e-05, + "loss": 0.6033, + "step": 2134 + }, + { + "epoch": 0.06733953635073332, + "grad_norm": 3.185229778289795, + "learning_rate": 1.992718816240313e-05, + "loss": 0.6126, + "step": 2135 + }, + { + "epoch": 0.06737107711717395, + "grad_norm": 3.4857494831085205, + "learning_rate": 1.9927065059251996e-05, + "loss": 0.386, + "step": 2136 + }, + { + "epoch": 0.06740261788361457, + "grad_norm": 4.548957347869873, + "learning_rate": 1.992694185250423e-05, + "loss": 0.599, + "step": 2137 + }, + { + "epoch": 0.0674341586500552, + "grad_norm": 3.366135597229004, + "learning_rate": 1.992681854216112e-05, + "loss": 0.6146, + "step": 2138 + }, + { + "epoch": 0.06746569941649581, + "grad_norm": 3.3443994522094727, + "learning_rate": 1.992669512822395e-05, + "loss": 0.8942, + "step": 2139 + }, + { + "epoch": 0.06749724018293644, + "grad_norm": 3.552417039871216, + "learning_rate": 1.9926571610694008e-05, + "loss": 0.6192, + "step": 2140 + }, + { + "epoch": 0.06752878094937707, + "grad_norm": 3.9201178550720215, + "learning_rate": 1.992644798957258e-05, + "loss": 0.6243, + "step": 2141 + }, + { + "epoch": 0.06756032171581769, + "grad_norm": 2.681299924850464, + "learning_rate": 1.9926324264860967e-05, + "loss": 0.3849, + "step": 2142 + }, + { + "epoch": 0.06759186248225832, + "grad_norm": 5.1985273361206055, + "learning_rate": 1.9926200436560448e-05, + "loss": 0.3924, + "step": 2143 + }, + { + "epoch": 0.06762340324869895, + "grad_norm": 3.4223155975341797, + "learning_rate": 1.992607650467232e-05, + "loss": 0.6517, + "step": 2144 + }, + { + "epoch": 0.06765494401513956, + "grad_norm": 3.6080849170684814, + "learning_rate": 1.992595246919788e-05, + "loss": 0.6297, + "step": 2145 + }, + { + "epoch": 0.0676864847815802, + "grad_norm": 3.873638868331909, + "learning_rate": 1.992582833013841e-05, + "loss": 0.6383, + "step": 2146 + }, + { + "epoch": 0.06771802554802081, + "grad_norm": 4.316949844360352, + "learning_rate": 1.9925704087495223e-05, + "loss": 0.6268, + "step": 2147 + }, + { + "epoch": 0.06774956631446144, + "grad_norm": 4.119110107421875, + "learning_rate": 1.99255797412696e-05, + "loss": 0.5885, + "step": 2148 + }, + { + "epoch": 0.06778110708090207, + "grad_norm": 43.12741470336914, + "learning_rate": 1.9925455291462852e-05, + "loss": 0.6482, + "step": 2149 + }, + { + "epoch": 0.06781264784734269, + "grad_norm": 9.262495040893555, + "learning_rate": 1.9925330738076266e-05, + "loss": 0.6399, + "step": 2150 + }, + { + "epoch": 0.06784418861378332, + "grad_norm": 1.9731112718582153, + "learning_rate": 1.9925206081111153e-05, + "loss": 0.9167, + "step": 2151 + }, + { + "epoch": 0.06787572938022395, + "grad_norm": 3.6015119552612305, + "learning_rate": 1.9925081320568802e-05, + "loss": 0.645, + "step": 2152 + }, + { + "epoch": 0.06790727014666456, + "grad_norm": 4.4483537673950195, + "learning_rate": 1.992495645645052e-05, + "loss": 0.6324, + "step": 2153 + }, + { + "epoch": 0.06793881091310519, + "grad_norm": 5.060041904449463, + "learning_rate": 1.9924831488757615e-05, + "loss": 0.656, + "step": 2154 + }, + { + "epoch": 0.0679703516795458, + "grad_norm": 5.050487041473389, + "learning_rate": 1.9924706417491388e-05, + "loss": 0.6287, + "step": 2155 + }, + { + "epoch": 0.06800189244598644, + "grad_norm": 4.993229389190674, + "learning_rate": 1.9924581242653137e-05, + "loss": 0.6358, + "step": 2156 + }, + { + "epoch": 0.06803343321242707, + "grad_norm": 5.17017936706543, + "learning_rate": 1.992445596424418e-05, + "loss": 0.6451, + "step": 2157 + }, + { + "epoch": 0.06806497397886768, + "grad_norm": 4.7019195556640625, + "learning_rate": 1.9924330582265817e-05, + "loss": 0.6621, + "step": 2158 + }, + { + "epoch": 0.06809651474530831, + "grad_norm": 5.90162992477417, + "learning_rate": 1.992420509671936e-05, + "loss": 0.9286, + "step": 2159 + }, + { + "epoch": 0.06812805551174894, + "grad_norm": 6.182692527770996, + "learning_rate": 1.9924079507606114e-05, + "loss": 0.628, + "step": 2160 + }, + { + "epoch": 0.06815959627818956, + "grad_norm": 3.6522481441497803, + "learning_rate": 1.9923953814927396e-05, + "loss": 0.6265, + "step": 2161 + }, + { + "epoch": 0.06819113704463019, + "grad_norm": 5.139028549194336, + "learning_rate": 1.992382801868451e-05, + "loss": 0.6404, + "step": 2162 + }, + { + "epoch": 0.0682226778110708, + "grad_norm": 9.53368091583252, + "learning_rate": 1.9923702118878773e-05, + "loss": 0.6141, + "step": 2163 + }, + { + "epoch": 0.06825421857751143, + "grad_norm": 4.923852920532227, + "learning_rate": 1.9923576115511504e-05, + "loss": 0.6178, + "step": 2164 + }, + { + "epoch": 0.06828575934395206, + "grad_norm": 4.179450511932373, + "learning_rate": 1.9923450008584008e-05, + "loss": 0.6331, + "step": 2165 + }, + { + "epoch": 0.06831730011039268, + "grad_norm": 13.691166877746582, + "learning_rate": 1.9923323798097607e-05, + "loss": 0.5825, + "step": 2166 + }, + { + "epoch": 0.06834884087683331, + "grad_norm": 3.0049314498901367, + "learning_rate": 1.9923197484053618e-05, + "loss": 0.4147, + "step": 2167 + }, + { + "epoch": 0.06838038164327394, + "grad_norm": 4.218119144439697, + "learning_rate": 1.9923071066453356e-05, + "loss": 0.6281, + "step": 2168 + }, + { + "epoch": 0.06841192240971455, + "grad_norm": 3.7785325050354004, + "learning_rate": 1.9922944545298143e-05, + "loss": 0.6353, + "step": 2169 + }, + { + "epoch": 0.06844346317615518, + "grad_norm": 4.860039710998535, + "learning_rate": 1.99228179205893e-05, + "loss": 0.6341, + "step": 2170 + }, + { + "epoch": 0.0684750039425958, + "grad_norm": 4.0147504806518555, + "learning_rate": 1.9922691192328148e-05, + "loss": 0.6617, + "step": 2171 + }, + { + "epoch": 0.06850654470903643, + "grad_norm": 3.9711644649505615, + "learning_rate": 1.9922564360516006e-05, + "loss": 0.6246, + "step": 2172 + }, + { + "epoch": 0.06853808547547706, + "grad_norm": 5.879579067230225, + "learning_rate": 1.99224374251542e-05, + "loss": 0.6423, + "step": 2173 + }, + { + "epoch": 0.06856962624191767, + "grad_norm": 8.734024047851562, + "learning_rate": 1.992231038624406e-05, + "loss": 0.617, + "step": 2174 + }, + { + "epoch": 0.0686011670083583, + "grad_norm": 5.643172740936279, + "learning_rate": 1.99221832437869e-05, + "loss": 0.5983, + "step": 2175 + }, + { + "epoch": 0.06863270777479893, + "grad_norm": 3.63088059425354, + "learning_rate": 1.9922055997784057e-05, + "loss": 0.629, + "step": 2176 + }, + { + "epoch": 0.06866424854123955, + "grad_norm": 3.4588205814361572, + "learning_rate": 1.9921928648236855e-05, + "loss": 0.6384, + "step": 2177 + }, + { + "epoch": 0.06869578930768018, + "grad_norm": 4.463915824890137, + "learning_rate": 1.9921801195146622e-05, + "loss": 0.6671, + "step": 2178 + }, + { + "epoch": 0.0687273300741208, + "grad_norm": 4.1968817710876465, + "learning_rate": 1.992167363851469e-05, + "loss": 0.635, + "step": 2179 + }, + { + "epoch": 0.06875887084056143, + "grad_norm": 23.753238677978516, + "learning_rate": 1.9921545978342392e-05, + "loss": 0.6523, + "step": 2180 + }, + { + "epoch": 0.06879041160700206, + "grad_norm": 6.046797275543213, + "learning_rate": 1.9921418214631055e-05, + "loss": 0.6434, + "step": 2181 + }, + { + "epoch": 0.06882195237344267, + "grad_norm": 6.249955654144287, + "learning_rate": 1.9921290347382016e-05, + "loss": 0.6685, + "step": 2182 + }, + { + "epoch": 0.0688534931398833, + "grad_norm": 2.960799217224121, + "learning_rate": 1.992116237659661e-05, + "loss": 0.4243, + "step": 2183 + }, + { + "epoch": 0.06888503390632392, + "grad_norm": 4.739340305328369, + "learning_rate": 1.992103430227617e-05, + "loss": 0.6341, + "step": 2184 + }, + { + "epoch": 0.06891657467276455, + "grad_norm": 4.659914016723633, + "learning_rate": 1.9920906124422032e-05, + "loss": 0.6341, + "step": 2185 + }, + { + "epoch": 0.06894811543920518, + "grad_norm": 5.068531513214111, + "learning_rate": 1.9920777843035536e-05, + "loss": 0.6522, + "step": 2186 + }, + { + "epoch": 0.06897965620564579, + "grad_norm": 19.637094497680664, + "learning_rate": 1.992064945811802e-05, + "loss": 0.6575, + "step": 2187 + }, + { + "epoch": 0.06901119697208642, + "grad_norm": 4.173631191253662, + "learning_rate": 1.9920520969670827e-05, + "loss": 0.6017, + "step": 2188 + }, + { + "epoch": 0.06904273773852705, + "grad_norm": 3.872854471206665, + "learning_rate": 1.992039237769529e-05, + "loss": 0.6398, + "step": 2189 + }, + { + "epoch": 0.06907427850496767, + "grad_norm": 4.180253505706787, + "learning_rate": 1.9920263682192758e-05, + "loss": 0.6308, + "step": 2190 + }, + { + "epoch": 0.0691058192714083, + "grad_norm": 4.481041431427002, + "learning_rate": 1.992013488316457e-05, + "loss": 0.6445, + "step": 2191 + }, + { + "epoch": 0.06913736003784891, + "grad_norm": 4.086665630340576, + "learning_rate": 1.9920005980612075e-05, + "loss": 0.6085, + "step": 2192 + }, + { + "epoch": 0.06916890080428954, + "grad_norm": 4.21210241317749, + "learning_rate": 1.991987697453661e-05, + "loss": 0.6289, + "step": 2193 + }, + { + "epoch": 0.06920044157073017, + "grad_norm": 4.903167247772217, + "learning_rate": 1.991974786493953e-05, + "loss": 0.6512, + "step": 2194 + }, + { + "epoch": 0.06923198233717079, + "grad_norm": 3.437824249267578, + "learning_rate": 1.9919618651822174e-05, + "loss": 0.6353, + "step": 2195 + }, + { + "epoch": 0.06926352310361142, + "grad_norm": 5.358554840087891, + "learning_rate": 1.99194893351859e-05, + "loss": 0.6255, + "step": 2196 + }, + { + "epoch": 0.06929506387005205, + "grad_norm": 27.681554794311523, + "learning_rate": 1.991935991503205e-05, + "loss": 0.6556, + "step": 2197 + }, + { + "epoch": 0.06932660463649266, + "grad_norm": 3.9950904846191406, + "learning_rate": 1.9919230391361977e-05, + "loss": 0.6057, + "step": 2198 + }, + { + "epoch": 0.0693581454029333, + "grad_norm": 3.3087828159332275, + "learning_rate": 1.9919100764177035e-05, + "loss": 0.6503, + "step": 2199 + }, + { + "epoch": 0.06938968616937391, + "grad_norm": 4.806223392486572, + "learning_rate": 1.9918971033478575e-05, + "loss": 0.6348, + "step": 2200 + }, + { + "epoch": 0.06942122693581454, + "grad_norm": 3.990539073944092, + "learning_rate": 1.9918841199267945e-05, + "loss": 0.6294, + "step": 2201 + }, + { + "epoch": 0.06945276770225517, + "grad_norm": 2.653846025466919, + "learning_rate": 1.991871126154651e-05, + "loss": 0.9203, + "step": 2202 + }, + { + "epoch": 0.06948430846869579, + "grad_norm": 4.854641914367676, + "learning_rate": 1.991858122031562e-05, + "loss": 0.6081, + "step": 2203 + }, + { + "epoch": 0.06951584923513641, + "grad_norm": 4.623096942901611, + "learning_rate": 1.9918451075576635e-05, + "loss": 0.6487, + "step": 2204 + }, + { + "epoch": 0.06954739000157704, + "grad_norm": 4.025958061218262, + "learning_rate": 1.9918320827330907e-05, + "loss": 0.6419, + "step": 2205 + }, + { + "epoch": 0.06957893076801766, + "grad_norm": 4.394904136657715, + "learning_rate": 1.9918190475579805e-05, + "loss": 0.6306, + "step": 2206 + }, + { + "epoch": 0.06961047153445829, + "grad_norm": 4.508657932281494, + "learning_rate": 1.991806002032468e-05, + "loss": 0.6403, + "step": 2207 + }, + { + "epoch": 0.0696420123008989, + "grad_norm": 3.6506855487823486, + "learning_rate": 1.9917929461566902e-05, + "loss": 0.6206, + "step": 2208 + }, + { + "epoch": 0.06967355306733954, + "grad_norm": 5.73282527923584, + "learning_rate": 1.9917798799307827e-05, + "loss": 0.6517, + "step": 2209 + }, + { + "epoch": 0.06970509383378017, + "grad_norm": 6.760509490966797, + "learning_rate": 1.991766803354882e-05, + "loss": 0.6596, + "step": 2210 + }, + { + "epoch": 0.06973663460022078, + "grad_norm": 4.853798866271973, + "learning_rate": 1.9917537164291244e-05, + "loss": 0.6498, + "step": 2211 + }, + { + "epoch": 0.06976817536666141, + "grad_norm": 2.8832626342773438, + "learning_rate": 1.991740619153647e-05, + "loss": 0.8792, + "step": 2212 + }, + { + "epoch": 0.06979971613310204, + "grad_norm": 9.766627311706543, + "learning_rate": 1.991727511528586e-05, + "loss": 0.6155, + "step": 2213 + }, + { + "epoch": 0.06983125689954266, + "grad_norm": 16.17064666748047, + "learning_rate": 1.9917143935540786e-05, + "loss": 0.6741, + "step": 2214 + }, + { + "epoch": 0.06986279766598329, + "grad_norm": 2.96797251701355, + "learning_rate": 1.9917012652302612e-05, + "loss": 0.3997, + "step": 2215 + }, + { + "epoch": 0.0698943384324239, + "grad_norm": 6.189297199249268, + "learning_rate": 1.991688126557271e-05, + "loss": 0.6433, + "step": 2216 + }, + { + "epoch": 0.06992587919886453, + "grad_norm": 4.273345470428467, + "learning_rate": 1.9916749775352455e-05, + "loss": 0.6192, + "step": 2217 + }, + { + "epoch": 0.06995741996530516, + "grad_norm": 7.227670192718506, + "learning_rate": 1.9916618181643215e-05, + "loss": 0.6411, + "step": 2218 + }, + { + "epoch": 0.06998896073174578, + "grad_norm": 5.485610485076904, + "learning_rate": 1.9916486484446366e-05, + "loss": 0.6463, + "step": 2219 + }, + { + "epoch": 0.07002050149818641, + "grad_norm": 4.268469333648682, + "learning_rate": 1.9916354683763276e-05, + "loss": 0.6223, + "step": 2220 + }, + { + "epoch": 0.07005204226462702, + "grad_norm": 3.6941287517547607, + "learning_rate": 1.9916222779595326e-05, + "loss": 0.4021, + "step": 2221 + }, + { + "epoch": 0.07008358303106765, + "grad_norm": 5.445878982543945, + "learning_rate": 1.9916090771943893e-05, + "loss": 0.6337, + "step": 2222 + }, + { + "epoch": 0.07011512379750828, + "grad_norm": 6.349086761474609, + "learning_rate": 1.9915958660810354e-05, + "loss": 0.6336, + "step": 2223 + }, + { + "epoch": 0.0701466645639489, + "grad_norm": 4.006677150726318, + "learning_rate": 1.9915826446196083e-05, + "loss": 0.6112, + "step": 2224 + }, + { + "epoch": 0.07017820533038953, + "grad_norm": 5.676721572875977, + "learning_rate": 1.9915694128102468e-05, + "loss": 0.6417, + "step": 2225 + }, + { + "epoch": 0.07020974609683016, + "grad_norm": 5.941351890563965, + "learning_rate": 1.9915561706530882e-05, + "loss": 0.636, + "step": 2226 + }, + { + "epoch": 0.07024128686327077, + "grad_norm": 2.6580491065979004, + "learning_rate": 1.9915429181482713e-05, + "loss": 0.3952, + "step": 2227 + }, + { + "epoch": 0.0702728276297114, + "grad_norm": 7.503307342529297, + "learning_rate": 1.991529655295934e-05, + "loss": 0.6303, + "step": 2228 + }, + { + "epoch": 0.07030436839615202, + "grad_norm": 8.425107955932617, + "learning_rate": 1.991516382096215e-05, + "loss": 0.6612, + "step": 2229 + }, + { + "epoch": 0.07033590916259265, + "grad_norm": 4.067044734954834, + "learning_rate": 1.9915030985492524e-05, + "loss": 0.6132, + "step": 2230 + }, + { + "epoch": 0.07036744992903328, + "grad_norm": 3.3317317962646484, + "learning_rate": 1.9914898046551853e-05, + "loss": 0.6442, + "step": 2231 + }, + { + "epoch": 0.0703989906954739, + "grad_norm": 2.642388105392456, + "learning_rate": 1.9914765004141522e-05, + "loss": 0.4222, + "step": 2232 + }, + { + "epoch": 0.07043053146191453, + "grad_norm": 4.697952747344971, + "learning_rate": 1.991463185826292e-05, + "loss": 0.6143, + "step": 2233 + }, + { + "epoch": 0.07046207222835515, + "grad_norm": 4.913858890533447, + "learning_rate": 1.9914498608917435e-05, + "loss": 0.615, + "step": 2234 + }, + { + "epoch": 0.07049361299479577, + "grad_norm": 8.587937355041504, + "learning_rate": 1.9914365256106457e-05, + "loss": 0.6418, + "step": 2235 + }, + { + "epoch": 0.0705251537612364, + "grad_norm": 3.4125335216522217, + "learning_rate": 1.991423179983138e-05, + "loss": 0.3733, + "step": 2236 + }, + { + "epoch": 0.07055669452767702, + "grad_norm": 10.829122543334961, + "learning_rate": 1.99140982400936e-05, + "loss": 0.6837, + "step": 2237 + }, + { + "epoch": 0.07058823529411765, + "grad_norm": 4.199316024780273, + "learning_rate": 1.9913964576894504e-05, + "loss": 0.6261, + "step": 2238 + }, + { + "epoch": 0.07061977606055828, + "grad_norm": 4.27957820892334, + "learning_rate": 1.991383081023549e-05, + "loss": 0.5828, + "step": 2239 + }, + { + "epoch": 0.07065131682699889, + "grad_norm": 4.235629558563232, + "learning_rate": 1.991369694011795e-05, + "loss": 0.6239, + "step": 2240 + }, + { + "epoch": 0.07068285759343952, + "grad_norm": 4.404016017913818, + "learning_rate": 1.991356296654329e-05, + "loss": 0.6614, + "step": 2241 + }, + { + "epoch": 0.07071439835988015, + "grad_norm": 6.430604457855225, + "learning_rate": 1.9913428889512898e-05, + "loss": 0.632, + "step": 2242 + }, + { + "epoch": 0.07074593912632077, + "grad_norm": 5.424838066101074, + "learning_rate": 1.991329470902818e-05, + "loss": 0.631, + "step": 2243 + }, + { + "epoch": 0.0707774798927614, + "grad_norm": 4.377035617828369, + "learning_rate": 1.9913160425090537e-05, + "loss": 0.6137, + "step": 2244 + }, + { + "epoch": 0.07080902065920201, + "grad_norm": 7.133978843688965, + "learning_rate": 1.9913026037701364e-05, + "loss": 0.6422, + "step": 2245 + }, + { + "epoch": 0.07084056142564264, + "grad_norm": 4.8127121925354, + "learning_rate": 1.9912891546862068e-05, + "loss": 0.6702, + "step": 2246 + }, + { + "epoch": 0.07087210219208327, + "grad_norm": 4.477261543273926, + "learning_rate": 1.9912756952574052e-05, + "loss": 0.5984, + "step": 2247 + }, + { + "epoch": 0.07090364295852389, + "grad_norm": 4.943467140197754, + "learning_rate": 1.9912622254838722e-05, + "loss": 0.5993, + "step": 2248 + }, + { + "epoch": 0.07093518372496452, + "grad_norm": 8.047342300415039, + "learning_rate": 1.9912487453657478e-05, + "loss": 0.6329, + "step": 2249 + }, + { + "epoch": 0.07096672449140515, + "grad_norm": 2.427508592605591, + "learning_rate": 1.991235254903173e-05, + "loss": 0.4182, + "step": 2250 + }, + { + "epoch": 0.07099826525784576, + "grad_norm": 4.2857465744018555, + "learning_rate": 1.991221754096289e-05, + "loss": 0.5942, + "step": 2251 + }, + { + "epoch": 0.0710298060242864, + "grad_norm": 4.307617664337158, + "learning_rate": 1.991208242945236e-05, + "loss": 0.6464, + "step": 2252 + }, + { + "epoch": 0.07106134679072701, + "grad_norm": 6.620888710021973, + "learning_rate": 1.9911947214501556e-05, + "loss": 0.6204, + "step": 2253 + }, + { + "epoch": 0.07109288755716764, + "grad_norm": 5.333327293395996, + "learning_rate": 1.9911811896111888e-05, + "loss": 0.5981, + "step": 2254 + }, + { + "epoch": 0.07112442832360827, + "grad_norm": 3.361542224884033, + "learning_rate": 1.991167647428476e-05, + "loss": 0.9015, + "step": 2255 + }, + { + "epoch": 0.07115596909004888, + "grad_norm": 4.973300933837891, + "learning_rate": 1.99115409490216e-05, + "loss": 0.6155, + "step": 2256 + }, + { + "epoch": 0.07118750985648951, + "grad_norm": 5.650325775146484, + "learning_rate": 1.991140532032381e-05, + "loss": 0.6665, + "step": 2257 + }, + { + "epoch": 0.07121905062293014, + "grad_norm": 29.525035858154297, + "learning_rate": 1.9911269588192806e-05, + "loss": 0.6275, + "step": 2258 + }, + { + "epoch": 0.07125059138937076, + "grad_norm": 4.1230974197387695, + "learning_rate": 1.991113375263001e-05, + "loss": 0.6228, + "step": 2259 + }, + { + "epoch": 0.07128213215581139, + "grad_norm": 4.1750898361206055, + "learning_rate": 1.9910997813636842e-05, + "loss": 0.6101, + "step": 2260 + }, + { + "epoch": 0.071313672922252, + "grad_norm": 5.140542984008789, + "learning_rate": 1.9910861771214712e-05, + "loss": 0.6509, + "step": 2261 + }, + { + "epoch": 0.07134521368869264, + "grad_norm": 10.629735946655273, + "learning_rate": 1.9910725625365046e-05, + "loss": 0.6358, + "step": 2262 + }, + { + "epoch": 0.07137675445513326, + "grad_norm": 2.620208978652954, + "learning_rate": 1.991058937608926e-05, + "loss": 0.4012, + "step": 2263 + }, + { + "epoch": 0.07140829522157388, + "grad_norm": 5.045866966247559, + "learning_rate": 1.9910453023388782e-05, + "loss": 0.6289, + "step": 2264 + }, + { + "epoch": 0.07143983598801451, + "grad_norm": 4.866350173950195, + "learning_rate": 1.9910316567265028e-05, + "loss": 0.6246, + "step": 2265 + }, + { + "epoch": 0.07147137675445513, + "grad_norm": 6.381232261657715, + "learning_rate": 1.9910180007719428e-05, + "loss": 0.6385, + "step": 2266 + }, + { + "epoch": 0.07150291752089576, + "grad_norm": 6.283633232116699, + "learning_rate": 1.9910043344753402e-05, + "loss": 0.9143, + "step": 2267 + }, + { + "epoch": 0.07153445828733639, + "grad_norm": 6.375021457672119, + "learning_rate": 1.990990657836838e-05, + "loss": 0.4278, + "step": 2268 + }, + { + "epoch": 0.071565999053777, + "grad_norm": 6.191483974456787, + "learning_rate": 1.990976970856579e-05, + "loss": 0.6595, + "step": 2269 + }, + { + "epoch": 0.07159753982021763, + "grad_norm": 6.82413387298584, + "learning_rate": 1.9909632735347056e-05, + "loss": 0.6247, + "step": 2270 + }, + { + "epoch": 0.07162908058665826, + "grad_norm": 6.204572677612305, + "learning_rate": 1.9909495658713612e-05, + "loss": 0.6454, + "step": 2271 + }, + { + "epoch": 0.07166062135309888, + "grad_norm": 4.494386196136475, + "learning_rate": 1.9909358478666885e-05, + "loss": 0.3323, + "step": 2272 + }, + { + "epoch": 0.07169216211953951, + "grad_norm": 3.598508358001709, + "learning_rate": 1.9909221195208307e-05, + "loss": 0.37, + "step": 2273 + }, + { + "epoch": 0.07172370288598012, + "grad_norm": 7.295068740844727, + "learning_rate": 1.9909083808339312e-05, + "loss": 0.6499, + "step": 2274 + }, + { + "epoch": 0.07175524365242075, + "grad_norm": 6.092055797576904, + "learning_rate": 1.9908946318061333e-05, + "loss": 0.6554, + "step": 2275 + }, + { + "epoch": 0.07178678441886138, + "grad_norm": 2.932814359664917, + "learning_rate": 1.9908808724375807e-05, + "loss": 0.3964, + "step": 2276 + }, + { + "epoch": 0.071818325185302, + "grad_norm": 4.886025905609131, + "learning_rate": 1.9908671027284163e-05, + "loss": 0.6045, + "step": 2277 + }, + { + "epoch": 0.07184986595174263, + "grad_norm": 7.082818031311035, + "learning_rate": 1.990853322678785e-05, + "loss": 0.6433, + "step": 2278 + }, + { + "epoch": 0.07188140671818326, + "grad_norm": 5.214577674865723, + "learning_rate": 1.9908395322888296e-05, + "loss": 0.6474, + "step": 2279 + }, + { + "epoch": 0.07191294748462387, + "grad_norm": 8.959336280822754, + "learning_rate": 1.990825731558694e-05, + "loss": 0.6057, + "step": 2280 + }, + { + "epoch": 0.0719444882510645, + "grad_norm": 2.76528263092041, + "learning_rate": 1.9908119204885227e-05, + "loss": 0.4099, + "step": 2281 + }, + { + "epoch": 0.07197602901750512, + "grad_norm": 5.8346357345581055, + "learning_rate": 1.9907980990784598e-05, + "loss": 0.6272, + "step": 2282 + }, + { + "epoch": 0.07200756978394575, + "grad_norm": 5.811941623687744, + "learning_rate": 1.9907842673286494e-05, + "loss": 0.6328, + "step": 2283 + }, + { + "epoch": 0.07203911055038638, + "grad_norm": 5.112471580505371, + "learning_rate": 1.9907704252392356e-05, + "loss": 0.6458, + "step": 2284 + }, + { + "epoch": 0.072070651316827, + "grad_norm": 7.082172870635986, + "learning_rate": 1.9907565728103636e-05, + "loss": 0.6428, + "step": 2285 + }, + { + "epoch": 0.07210219208326762, + "grad_norm": 11.607803344726562, + "learning_rate": 1.990742710042177e-05, + "loss": 0.642, + "step": 2286 + }, + { + "epoch": 0.07213373284970825, + "grad_norm": 8.479022026062012, + "learning_rate": 1.990728836934821e-05, + "loss": 0.653, + "step": 2287 + }, + { + "epoch": 0.07216527361614887, + "grad_norm": 4.587887763977051, + "learning_rate": 1.9907149534884405e-05, + "loss": 0.6531, + "step": 2288 + }, + { + "epoch": 0.0721968143825895, + "grad_norm": 9.982447624206543, + "learning_rate": 1.9907010597031798e-05, + "loss": 0.612, + "step": 2289 + }, + { + "epoch": 0.07222835514903012, + "grad_norm": 3.342390537261963, + "learning_rate": 1.9906871555791847e-05, + "loss": 0.4284, + "step": 2290 + }, + { + "epoch": 0.07225989591547075, + "grad_norm": 5.720997333526611, + "learning_rate": 1.9906732411165996e-05, + "loss": 0.6083, + "step": 2291 + }, + { + "epoch": 0.07229143668191138, + "grad_norm": 14.001333236694336, + "learning_rate": 1.9906593163155704e-05, + "loss": 0.6408, + "step": 2292 + }, + { + "epoch": 0.07232297744835199, + "grad_norm": 3.0948493480682373, + "learning_rate": 1.9906453811762415e-05, + "loss": 0.368, + "step": 2293 + }, + { + "epoch": 0.07235451821479262, + "grad_norm": 4.5572075843811035, + "learning_rate": 1.990631435698759e-05, + "loss": 0.6469, + "step": 2294 + }, + { + "epoch": 0.07238605898123325, + "grad_norm": 2.228177309036255, + "learning_rate": 1.9906174798832685e-05, + "loss": 0.8655, + "step": 2295 + }, + { + "epoch": 0.07241759974767387, + "grad_norm": 5.6606221199035645, + "learning_rate": 1.990603513729915e-05, + "loss": 0.6414, + "step": 2296 + }, + { + "epoch": 0.0724491405141145, + "grad_norm": 3.055065393447876, + "learning_rate": 1.9905895372388452e-05, + "loss": 0.3348, + "step": 2297 + }, + { + "epoch": 0.07248068128055511, + "grad_norm": 7.352939605712891, + "learning_rate": 1.990575550410204e-05, + "loss": 0.6044, + "step": 2298 + }, + { + "epoch": 0.07251222204699574, + "grad_norm": 9.903213500976562, + "learning_rate": 1.990561553244138e-05, + "loss": 0.6239, + "step": 2299 + }, + { + "epoch": 0.07254376281343637, + "grad_norm": 5.026385307312012, + "learning_rate": 1.9905475457407925e-05, + "loss": 0.3866, + "step": 2300 + }, + { + "epoch": 0.07257530357987699, + "grad_norm": 6.620054721832275, + "learning_rate": 1.990533527900315e-05, + "loss": 0.6327, + "step": 2301 + }, + { + "epoch": 0.07260684434631762, + "grad_norm": 6.06575870513916, + "learning_rate": 1.99051949972285e-05, + "loss": 0.6581, + "step": 2302 + }, + { + "epoch": 0.07263838511275825, + "grad_norm": 6.194949626922607, + "learning_rate": 1.990505461208546e-05, + "loss": 0.6547, + "step": 2303 + }, + { + "epoch": 0.07266992587919886, + "grad_norm": 6.745542049407959, + "learning_rate": 1.9904914123575478e-05, + "loss": 0.6342, + "step": 2304 + }, + { + "epoch": 0.07270146664563949, + "grad_norm": 4.059894561767578, + "learning_rate": 1.990477353170003e-05, + "loss": 0.4017, + "step": 2305 + }, + { + "epoch": 0.07273300741208011, + "grad_norm": 7.247722625732422, + "learning_rate": 1.9904632836460577e-05, + "loss": 0.6143, + "step": 2306 + }, + { + "epoch": 0.07276454817852074, + "grad_norm": 5.0805158615112305, + "learning_rate": 1.990449203785859e-05, + "loss": 0.6164, + "step": 2307 + }, + { + "epoch": 0.07279608894496137, + "grad_norm": 11.208080291748047, + "learning_rate": 1.9904351135895535e-05, + "loss": 0.6491, + "step": 2308 + }, + { + "epoch": 0.07282762971140198, + "grad_norm": 6.539576053619385, + "learning_rate": 1.990421013057289e-05, + "loss": 0.6218, + "step": 2309 + }, + { + "epoch": 0.07285917047784261, + "grad_norm": 5.7605109214782715, + "learning_rate": 1.9904069021892115e-05, + "loss": 0.6363, + "step": 2310 + }, + { + "epoch": 0.07289071124428323, + "grad_norm": 6.287595748901367, + "learning_rate": 1.9903927809854697e-05, + "loss": 0.6102, + "step": 2311 + }, + { + "epoch": 0.07292225201072386, + "grad_norm": 5.575453281402588, + "learning_rate": 1.99037864944621e-05, + "loss": 0.5994, + "step": 2312 + }, + { + "epoch": 0.07295379277716449, + "grad_norm": 4.250394344329834, + "learning_rate": 1.9903645075715798e-05, + "loss": 0.6094, + "step": 2313 + }, + { + "epoch": 0.0729853335436051, + "grad_norm": 5.962847709655762, + "learning_rate": 1.990350355361727e-05, + "loss": 0.6289, + "step": 2314 + }, + { + "epoch": 0.07301687431004573, + "grad_norm": 4.683816432952881, + "learning_rate": 1.9903361928167997e-05, + "loss": 0.6204, + "step": 2315 + }, + { + "epoch": 0.07304841507648636, + "grad_norm": 8.935823440551758, + "learning_rate": 1.990322019936945e-05, + "loss": 0.6135, + "step": 2316 + }, + { + "epoch": 0.07307995584292698, + "grad_norm": 5.006917953491211, + "learning_rate": 1.9903078367223107e-05, + "loss": 0.5903, + "step": 2317 + }, + { + "epoch": 0.07311149660936761, + "grad_norm": 4.626060962677002, + "learning_rate": 1.9902936431730454e-05, + "loss": 0.6452, + "step": 2318 + }, + { + "epoch": 0.07314303737580823, + "grad_norm": 8.783535957336426, + "learning_rate": 1.9902794392892972e-05, + "loss": 0.614, + "step": 2319 + }, + { + "epoch": 0.07317457814224886, + "grad_norm": 9.725351333618164, + "learning_rate": 1.990265225071214e-05, + "loss": 0.6239, + "step": 2320 + }, + { + "epoch": 0.07320611890868949, + "grad_norm": 4.916928291320801, + "learning_rate": 1.9902510005189444e-05, + "loss": 0.9117, + "step": 2321 + }, + { + "epoch": 0.0732376596751301, + "grad_norm": 5.4895477294921875, + "learning_rate": 1.9902367656326363e-05, + "loss": 0.6351, + "step": 2322 + }, + { + "epoch": 0.07326920044157073, + "grad_norm": 3.0656166076660156, + "learning_rate": 1.9902225204124392e-05, + "loss": 0.8755, + "step": 2323 + }, + { + "epoch": 0.07330074120801136, + "grad_norm": 5.313934803009033, + "learning_rate": 1.990208264858501e-05, + "loss": 0.6248, + "step": 2324 + }, + { + "epoch": 0.07333228197445198, + "grad_norm": 5.3869123458862305, + "learning_rate": 1.9901939989709708e-05, + "loss": 0.6337, + "step": 2325 + }, + { + "epoch": 0.0733638227408926, + "grad_norm": 3.6735191345214844, + "learning_rate": 1.990179722749997e-05, + "loss": 0.5884, + "step": 2326 + }, + { + "epoch": 0.07339536350733322, + "grad_norm": 5.047731876373291, + "learning_rate": 1.990165436195729e-05, + "loss": 0.6078, + "step": 2327 + }, + { + "epoch": 0.07342690427377385, + "grad_norm": 5.558433532714844, + "learning_rate": 1.990151139308316e-05, + "loss": 0.6303, + "step": 2328 + }, + { + "epoch": 0.07345844504021448, + "grad_norm": 8.384454727172852, + "learning_rate": 1.990136832087907e-05, + "loss": 0.6249, + "step": 2329 + }, + { + "epoch": 0.0734899858066551, + "grad_norm": 4.878937244415283, + "learning_rate": 1.990122514534651e-05, + "loss": 0.5967, + "step": 2330 + }, + { + "epoch": 0.07352152657309573, + "grad_norm": 3.105759859085083, + "learning_rate": 1.9901081866486983e-05, + "loss": 0.4072, + "step": 2331 + }, + { + "epoch": 0.07355306733953636, + "grad_norm": 10.202324867248535, + "learning_rate": 1.9900938484301975e-05, + "loss": 0.5791, + "step": 2332 + }, + { + "epoch": 0.07358460810597697, + "grad_norm": 4.4583940505981445, + "learning_rate": 1.990079499879299e-05, + "loss": 0.6801, + "step": 2333 + }, + { + "epoch": 0.0736161488724176, + "grad_norm": 9.339067459106445, + "learning_rate": 1.9900651409961517e-05, + "loss": 0.6875, + "step": 2334 + }, + { + "epoch": 0.07364768963885822, + "grad_norm": 4.222428798675537, + "learning_rate": 1.9900507717809058e-05, + "loss": 0.631, + "step": 2335 + }, + { + "epoch": 0.07367923040529885, + "grad_norm": 3.4049417972564697, + "learning_rate": 1.990036392233712e-05, + "loss": 0.6127, + "step": 2336 + }, + { + "epoch": 0.07371077117173948, + "grad_norm": 6.559947967529297, + "learning_rate": 1.990022002354719e-05, + "loss": 0.4085, + "step": 2337 + }, + { + "epoch": 0.0737423119381801, + "grad_norm": 5.008313179016113, + "learning_rate": 1.990007602144078e-05, + "loss": 0.6467, + "step": 2338 + }, + { + "epoch": 0.07377385270462072, + "grad_norm": 4.539331436157227, + "learning_rate": 1.9899931916019387e-05, + "loss": 0.9358, + "step": 2339 + }, + { + "epoch": 0.07380539347106135, + "grad_norm": 4.499982833862305, + "learning_rate": 1.9899787707284522e-05, + "loss": 0.611, + "step": 2340 + }, + { + "epoch": 0.07383693423750197, + "grad_norm": 3.9892311096191406, + "learning_rate": 1.989964339523768e-05, + "loss": 0.8855, + "step": 2341 + }, + { + "epoch": 0.0738684750039426, + "grad_norm": 6.162343502044678, + "learning_rate": 1.9899498979880378e-05, + "loss": 0.6312, + "step": 2342 + }, + { + "epoch": 0.07390001577038322, + "grad_norm": 5.3582234382629395, + "learning_rate": 1.9899354461214114e-05, + "loss": 0.633, + "step": 2343 + }, + { + "epoch": 0.07393155653682384, + "grad_norm": 5.431064605712891, + "learning_rate": 1.98992098392404e-05, + "loss": 0.6572, + "step": 2344 + }, + { + "epoch": 0.07396309730326447, + "grad_norm": 3.849703311920166, + "learning_rate": 1.9899065113960746e-05, + "loss": 0.3691, + "step": 2345 + }, + { + "epoch": 0.07399463806970509, + "grad_norm": 3.70066237449646, + "learning_rate": 1.989892028537666e-05, + "loss": 0.6228, + "step": 2346 + }, + { + "epoch": 0.07402617883614572, + "grad_norm": 8.61185073852539, + "learning_rate": 1.9898775353489653e-05, + "loss": 0.6263, + "step": 2347 + }, + { + "epoch": 0.07405771960258634, + "grad_norm": 3.419175863265991, + "learning_rate": 1.989863031830124e-05, + "loss": 0.6185, + "step": 2348 + }, + { + "epoch": 0.07408926036902697, + "grad_norm": 2.3634872436523438, + "learning_rate": 1.9898485179812933e-05, + "loss": 0.3978, + "step": 2349 + }, + { + "epoch": 0.0741208011354676, + "grad_norm": 5.471621036529541, + "learning_rate": 1.9898339938026247e-05, + "loss": 0.9221, + "step": 2350 + }, + { + "epoch": 0.07415234190190821, + "grad_norm": 6.1258440017700195, + "learning_rate": 1.98981945929427e-05, + "loss": 0.6734, + "step": 2351 + }, + { + "epoch": 0.07418388266834884, + "grad_norm": 7.185361862182617, + "learning_rate": 1.9898049144563807e-05, + "loss": 0.6487, + "step": 2352 + }, + { + "epoch": 0.07421542343478947, + "grad_norm": 4.215372085571289, + "learning_rate": 1.9897903592891082e-05, + "loss": 0.5962, + "step": 2353 + }, + { + "epoch": 0.07424696420123009, + "grad_norm": 4.613178253173828, + "learning_rate": 1.989775793792605e-05, + "loss": 0.6319, + "step": 2354 + }, + { + "epoch": 0.07427850496767072, + "grad_norm": 3.908540964126587, + "learning_rate": 1.989761217967023e-05, + "loss": 0.6262, + "step": 2355 + }, + { + "epoch": 0.07431004573411133, + "grad_norm": 9.28704833984375, + "learning_rate": 1.989746631812514e-05, + "loss": 0.6339, + "step": 2356 + }, + { + "epoch": 0.07434158650055196, + "grad_norm": 5.578834533691406, + "learning_rate": 1.9897320353292304e-05, + "loss": 0.6321, + "step": 2357 + }, + { + "epoch": 0.07437312726699259, + "grad_norm": 3.6197474002838135, + "learning_rate": 1.9897174285173244e-05, + "loss": 0.6133, + "step": 2358 + }, + { + "epoch": 0.07440466803343321, + "grad_norm": 5.7852277755737305, + "learning_rate": 1.9897028113769487e-05, + "loss": 0.6393, + "step": 2359 + }, + { + "epoch": 0.07443620879987384, + "grad_norm": 4.879981517791748, + "learning_rate": 1.9896881839082554e-05, + "loss": 0.6114, + "step": 2360 + }, + { + "epoch": 0.07446774956631447, + "grad_norm": 4.283775329589844, + "learning_rate": 1.9896735461113977e-05, + "loss": 0.6407, + "step": 2361 + }, + { + "epoch": 0.07449929033275508, + "grad_norm": 5.741538047790527, + "learning_rate": 1.989658897986528e-05, + "loss": 0.641, + "step": 2362 + }, + { + "epoch": 0.07453083109919571, + "grad_norm": 4.599820137023926, + "learning_rate": 1.9896442395337994e-05, + "loss": 0.6046, + "step": 2363 + }, + { + "epoch": 0.07456237186563633, + "grad_norm": 6.408308029174805, + "learning_rate": 1.9896295707533645e-05, + "loss": 0.4387, + "step": 2364 + }, + { + "epoch": 0.07459391263207696, + "grad_norm": 4.818927764892578, + "learning_rate": 1.9896148916453767e-05, + "loss": 0.6303, + "step": 2365 + }, + { + "epoch": 0.07462545339851759, + "grad_norm": 4.190194129943848, + "learning_rate": 1.989600202209989e-05, + "loss": 0.9278, + "step": 2366 + }, + { + "epoch": 0.0746569941649582, + "grad_norm": 6.1230034828186035, + "learning_rate": 1.9895855024473547e-05, + "loss": 0.605, + "step": 2367 + }, + { + "epoch": 0.07468853493139883, + "grad_norm": 5.822169303894043, + "learning_rate": 1.9895707923576275e-05, + "loss": 0.638, + "step": 2368 + }, + { + "epoch": 0.07472007569783946, + "grad_norm": 3.960601329803467, + "learning_rate": 1.9895560719409606e-05, + "loss": 0.6277, + "step": 2369 + }, + { + "epoch": 0.07475161646428008, + "grad_norm": 2.9575376510620117, + "learning_rate": 1.989541341197508e-05, + "loss": 0.8886, + "step": 2370 + }, + { + "epoch": 0.07478315723072071, + "grad_norm": 4.809655666351318, + "learning_rate": 1.989526600127423e-05, + "loss": 0.6125, + "step": 2371 + }, + { + "epoch": 0.07481469799716133, + "grad_norm": 6.587060928344727, + "learning_rate": 1.989511848730859e-05, + "loss": 0.6176, + "step": 2372 + }, + { + "epoch": 0.07484623876360195, + "grad_norm": 4.559661388397217, + "learning_rate": 1.9894970870079714e-05, + "loss": 0.6323, + "step": 2373 + }, + { + "epoch": 0.07487777953004258, + "grad_norm": 8.825815200805664, + "learning_rate": 1.9894823149589126e-05, + "loss": 0.586, + "step": 2374 + }, + { + "epoch": 0.0749093202964832, + "grad_norm": 3.3007984161376953, + "learning_rate": 1.989467532583838e-05, + "loss": 0.6284, + "step": 2375 + }, + { + "epoch": 0.07494086106292383, + "grad_norm": 5.360291957855225, + "learning_rate": 1.9894527398829012e-05, + "loss": 0.6538, + "step": 2376 + }, + { + "epoch": 0.07497240182936446, + "grad_norm": 6.845655918121338, + "learning_rate": 1.9894379368562568e-05, + "loss": 0.6482, + "step": 2377 + }, + { + "epoch": 0.07500394259580508, + "grad_norm": 5.007275581359863, + "learning_rate": 1.9894231235040596e-05, + "loss": 0.621, + "step": 2378 + }, + { + "epoch": 0.0750354833622457, + "grad_norm": 10.777523040771484, + "learning_rate": 1.9894082998264633e-05, + "loss": 0.6258, + "step": 2379 + }, + { + "epoch": 0.07506702412868632, + "grad_norm": 9.007742881774902, + "learning_rate": 1.9893934658236234e-05, + "loss": 0.6197, + "step": 2380 + }, + { + "epoch": 0.07509856489512695, + "grad_norm": 6.289229393005371, + "learning_rate": 1.9893786214956946e-05, + "loss": 0.6057, + "step": 2381 + }, + { + "epoch": 0.07513010566156758, + "grad_norm": 8.810538291931152, + "learning_rate": 1.9893637668428315e-05, + "loss": 0.6132, + "step": 2382 + }, + { + "epoch": 0.0751616464280082, + "grad_norm": 5.975162506103516, + "learning_rate": 1.989348901865189e-05, + "loss": 0.6243, + "step": 2383 + }, + { + "epoch": 0.07519318719444883, + "grad_norm": 2.239873170852661, + "learning_rate": 1.989334026562923e-05, + "loss": 0.3665, + "step": 2384 + }, + { + "epoch": 0.07522472796088946, + "grad_norm": 5.93299674987793, + "learning_rate": 1.989319140936188e-05, + "loss": 0.6261, + "step": 2385 + }, + { + "epoch": 0.07525626872733007, + "grad_norm": 5.3431878089904785, + "learning_rate": 1.9893042449851398e-05, + "loss": 0.3754, + "step": 2386 + }, + { + "epoch": 0.0752878094937707, + "grad_norm": 6.09681510925293, + "learning_rate": 1.989289338709933e-05, + "loss": 0.6055, + "step": 2387 + }, + { + "epoch": 0.07531935026021132, + "grad_norm": 5.822213649749756, + "learning_rate": 1.989274422110724e-05, + "loss": 0.6287, + "step": 2388 + }, + { + "epoch": 0.07535089102665195, + "grad_norm": 3.6656925678253174, + "learning_rate": 1.9892594951876687e-05, + "loss": 0.6223, + "step": 2389 + }, + { + "epoch": 0.07538243179309258, + "grad_norm": 6.537736892700195, + "learning_rate": 1.9892445579409224e-05, + "loss": 0.6511, + "step": 2390 + }, + { + "epoch": 0.0754139725595332, + "grad_norm": 4.3511834144592285, + "learning_rate": 1.9892296103706406e-05, + "loss": 0.5815, + "step": 2391 + }, + { + "epoch": 0.07544551332597382, + "grad_norm": 5.9666266441345215, + "learning_rate": 1.9892146524769797e-05, + "loss": 0.62, + "step": 2392 + }, + { + "epoch": 0.07547705409241444, + "grad_norm": 5.214536190032959, + "learning_rate": 1.989199684260096e-05, + "loss": 0.5946, + "step": 2393 + }, + { + "epoch": 0.07550859485885507, + "grad_norm": 5.175411224365234, + "learning_rate": 1.9891847057201453e-05, + "loss": 0.6149, + "step": 2394 + }, + { + "epoch": 0.0755401356252957, + "grad_norm": 8.291105270385742, + "learning_rate": 1.9891697168572845e-05, + "loss": 0.6092, + "step": 2395 + }, + { + "epoch": 0.07557167639173631, + "grad_norm": 5.453266620635986, + "learning_rate": 1.9891547176716693e-05, + "loss": 0.6147, + "step": 2396 + }, + { + "epoch": 0.07560321715817694, + "grad_norm": 3.7725303173065186, + "learning_rate": 1.9891397081634568e-05, + "loss": 0.6131, + "step": 2397 + }, + { + "epoch": 0.07563475792461757, + "grad_norm": 3.4393723011016846, + "learning_rate": 1.9891246883328034e-05, + "loss": 0.5971, + "step": 2398 + }, + { + "epoch": 0.07566629869105819, + "grad_norm": 8.178641319274902, + "learning_rate": 1.9891096581798653e-05, + "loss": 0.6073, + "step": 2399 + }, + { + "epoch": 0.07569783945749882, + "grad_norm": 4.64058780670166, + "learning_rate": 1.9890946177048004e-05, + "loss": 0.6312, + "step": 2400 + }, + { + "epoch": 0.07572938022393944, + "grad_norm": 3.7748045921325684, + "learning_rate": 1.9890795669077653e-05, + "loss": 0.6331, + "step": 2401 + }, + { + "epoch": 0.07576092099038007, + "grad_norm": 3.2949790954589844, + "learning_rate": 1.9890645057889168e-05, + "loss": 0.6003, + "step": 2402 + }, + { + "epoch": 0.0757924617568207, + "grad_norm": 4.1348371505737305, + "learning_rate": 1.989049434348412e-05, + "loss": 0.6358, + "step": 2403 + }, + { + "epoch": 0.07582400252326131, + "grad_norm": 4.137805938720703, + "learning_rate": 1.9890343525864084e-05, + "loss": 0.6361, + "step": 2404 + }, + { + "epoch": 0.07585554328970194, + "grad_norm": 7.061169147491455, + "learning_rate": 1.9890192605030633e-05, + "loss": 0.6268, + "step": 2405 + }, + { + "epoch": 0.07588708405614257, + "grad_norm": 4.363341331481934, + "learning_rate": 1.9890041580985343e-05, + "loss": 0.6359, + "step": 2406 + }, + { + "epoch": 0.07591862482258319, + "grad_norm": 4.0992631912231445, + "learning_rate": 1.9889890453729793e-05, + "loss": 0.6067, + "step": 2407 + }, + { + "epoch": 0.07595016558902382, + "grad_norm": 6.879531383514404, + "learning_rate": 1.9889739223265552e-05, + "loss": 0.6246, + "step": 2408 + }, + { + "epoch": 0.07598170635546443, + "grad_norm": 3.6823763847351074, + "learning_rate": 1.9889587889594206e-05, + "loss": 0.8989, + "step": 2409 + }, + { + "epoch": 0.07601324712190506, + "grad_norm": 4.340083122253418, + "learning_rate": 1.988943645271733e-05, + "loss": 0.6563, + "step": 2410 + }, + { + "epoch": 0.07604478788834569, + "grad_norm": 4.318861961364746, + "learning_rate": 1.988928491263651e-05, + "loss": 0.6339, + "step": 2411 + }, + { + "epoch": 0.07607632865478631, + "grad_norm": 4.2918243408203125, + "learning_rate": 1.9889133269353315e-05, + "loss": 0.6265, + "step": 2412 + }, + { + "epoch": 0.07610786942122694, + "grad_norm": 7.055503845214844, + "learning_rate": 1.9888981522869342e-05, + "loss": 0.6278, + "step": 2413 + }, + { + "epoch": 0.07613941018766757, + "grad_norm": 4.26211404800415, + "learning_rate": 1.9888829673186168e-05, + "loss": 0.6047, + "step": 2414 + }, + { + "epoch": 0.07617095095410818, + "grad_norm": 10.556702613830566, + "learning_rate": 1.9888677720305375e-05, + "loss": 0.634, + "step": 2415 + }, + { + "epoch": 0.07620249172054881, + "grad_norm": 4.925488471984863, + "learning_rate": 1.988852566422855e-05, + "loss": 0.6405, + "step": 2416 + }, + { + "epoch": 0.07623403248698943, + "grad_norm": 4.534007549285889, + "learning_rate": 1.9888373504957286e-05, + "loss": 0.6269, + "step": 2417 + }, + { + "epoch": 0.07626557325343006, + "grad_norm": 5.1684112548828125, + "learning_rate": 1.9888221242493164e-05, + "loss": 0.6505, + "step": 2418 + }, + { + "epoch": 0.07629711401987069, + "grad_norm": 5.113308906555176, + "learning_rate": 1.9888068876837776e-05, + "loss": 0.6366, + "step": 2419 + }, + { + "epoch": 0.0763286547863113, + "grad_norm": 5.029881954193115, + "learning_rate": 1.988791640799271e-05, + "loss": 0.6114, + "step": 2420 + }, + { + "epoch": 0.07636019555275193, + "grad_norm": 3.376577138900757, + "learning_rate": 1.9887763835959562e-05, + "loss": 0.8806, + "step": 2421 + }, + { + "epoch": 0.07639173631919256, + "grad_norm": 4.691312789916992, + "learning_rate": 1.9887611160739917e-05, + "loss": 0.3733, + "step": 2422 + }, + { + "epoch": 0.07642327708563318, + "grad_norm": 3.4688363075256348, + "learning_rate": 1.9887458382335376e-05, + "loss": 0.6224, + "step": 2423 + }, + { + "epoch": 0.07645481785207381, + "grad_norm": 3.733365058898926, + "learning_rate": 1.9887305500747524e-05, + "loss": 0.6287, + "step": 2424 + }, + { + "epoch": 0.07648635861851442, + "grad_norm": 2.1910977363586426, + "learning_rate": 1.9887152515977967e-05, + "loss": 0.4132, + "step": 2425 + }, + { + "epoch": 0.07651789938495505, + "grad_norm": 8.492374420166016, + "learning_rate": 1.988699942802829e-05, + "loss": 0.6466, + "step": 2426 + }, + { + "epoch": 0.07654944015139568, + "grad_norm": 2.666459083557129, + "learning_rate": 1.9886846236900102e-05, + "loss": 0.8771, + "step": 2427 + }, + { + "epoch": 0.0765809809178363, + "grad_norm": 3.2278270721435547, + "learning_rate": 1.9886692942595e-05, + "loss": 0.8502, + "step": 2428 + }, + { + "epoch": 0.07661252168427693, + "grad_norm": 5.033283233642578, + "learning_rate": 1.9886539545114572e-05, + "loss": 0.6257, + "step": 2429 + }, + { + "epoch": 0.07664406245071755, + "grad_norm": 3.8850135803222656, + "learning_rate": 1.988638604446043e-05, + "loss": 0.6474, + "step": 2430 + }, + { + "epoch": 0.07667560321715818, + "grad_norm": 6.030628681182861, + "learning_rate": 1.9886232440634176e-05, + "loss": 0.6088, + "step": 2431 + }, + { + "epoch": 0.0767071439835988, + "grad_norm": 3.463961601257324, + "learning_rate": 1.988607873363741e-05, + "loss": 0.5795, + "step": 2432 + }, + { + "epoch": 0.07673868475003942, + "grad_norm": 3.665555238723755, + "learning_rate": 1.9885924923471732e-05, + "loss": 0.926, + "step": 2433 + }, + { + "epoch": 0.07677022551648005, + "grad_norm": 4.089578628540039, + "learning_rate": 1.9885771010138753e-05, + "loss": 0.6268, + "step": 2434 + }, + { + "epoch": 0.07680176628292068, + "grad_norm": 6.625846862792969, + "learning_rate": 1.9885616993640078e-05, + "loss": 0.6183, + "step": 2435 + }, + { + "epoch": 0.0768333070493613, + "grad_norm": 8.364611625671387, + "learning_rate": 1.9885462873977313e-05, + "loss": 0.4245, + "step": 2436 + }, + { + "epoch": 0.07686484781580193, + "grad_norm": 6.533390998840332, + "learning_rate": 1.9885308651152068e-05, + "loss": 0.636, + "step": 2437 + }, + { + "epoch": 0.07689638858224254, + "grad_norm": 4.07136344909668, + "learning_rate": 1.988515432516595e-05, + "loss": 0.6294, + "step": 2438 + }, + { + "epoch": 0.07692792934868317, + "grad_norm": 5.178923606872559, + "learning_rate": 1.9884999896020573e-05, + "loss": 0.5896, + "step": 2439 + }, + { + "epoch": 0.0769594701151238, + "grad_norm": 19.524887084960938, + "learning_rate": 1.9884845363717543e-05, + "loss": 0.6412, + "step": 2440 + }, + { + "epoch": 0.07699101088156442, + "grad_norm": 3.4310462474823, + "learning_rate": 1.988469072825848e-05, + "loss": 0.6126, + "step": 2441 + }, + { + "epoch": 0.07702255164800505, + "grad_norm": 3.7964534759521484, + "learning_rate": 1.9884535989644993e-05, + "loss": 0.6102, + "step": 2442 + }, + { + "epoch": 0.07705409241444568, + "grad_norm": 4.725347518920898, + "learning_rate": 1.98843811478787e-05, + "loss": 0.6066, + "step": 2443 + }, + { + "epoch": 0.07708563318088629, + "grad_norm": 3.0900983810424805, + "learning_rate": 1.988422620296121e-05, + "loss": 0.9114, + "step": 2444 + }, + { + "epoch": 0.07711717394732692, + "grad_norm": 4.26637601852417, + "learning_rate": 1.9884071154894146e-05, + "loss": 0.6292, + "step": 2445 + }, + { + "epoch": 0.07714871471376754, + "grad_norm": 4.431549549102783, + "learning_rate": 1.9883916003679126e-05, + "loss": 0.6185, + "step": 2446 + }, + { + "epoch": 0.07718025548020817, + "grad_norm": 3.916548252105713, + "learning_rate": 1.9883760749317768e-05, + "loss": 0.6368, + "step": 2447 + }, + { + "epoch": 0.0772117962466488, + "grad_norm": 3.574462413787842, + "learning_rate": 1.9883605391811692e-05, + "loss": 0.4188, + "step": 2448 + }, + { + "epoch": 0.07724333701308941, + "grad_norm": 2.6254470348358154, + "learning_rate": 1.9883449931162517e-05, + "loss": 0.8868, + "step": 2449 + }, + { + "epoch": 0.07727487777953004, + "grad_norm": 3.0739219188690186, + "learning_rate": 1.9883294367371874e-05, + "loss": 0.5817, + "step": 2450 + }, + { + "epoch": 0.07730641854597067, + "grad_norm": 4.455557823181152, + "learning_rate": 1.988313870044137e-05, + "loss": 0.6512, + "step": 2451 + }, + { + "epoch": 0.07733795931241129, + "grad_norm": 4.05713415145874, + "learning_rate": 1.9882982930372648e-05, + "loss": 0.6177, + "step": 2452 + }, + { + "epoch": 0.07736950007885192, + "grad_norm": 3.273831367492676, + "learning_rate": 1.9882827057167324e-05, + "loss": 0.5956, + "step": 2453 + }, + { + "epoch": 0.07740104084529253, + "grad_norm": 2.4971044063568115, + "learning_rate": 1.988267108082702e-05, + "loss": 0.8503, + "step": 2454 + }, + { + "epoch": 0.07743258161173316, + "grad_norm": 3.2344374656677246, + "learning_rate": 1.9882515001353378e-05, + "loss": 0.3982, + "step": 2455 + }, + { + "epoch": 0.0774641223781738, + "grad_norm": 3.468350410461426, + "learning_rate": 1.9882358818748013e-05, + "loss": 0.6265, + "step": 2456 + }, + { + "epoch": 0.07749566314461441, + "grad_norm": 3.9163148403167725, + "learning_rate": 1.988220253301256e-05, + "loss": 0.6502, + "step": 2457 + }, + { + "epoch": 0.07752720391105504, + "grad_norm": 2.104506492614746, + "learning_rate": 1.9882046144148653e-05, + "loss": 0.4022, + "step": 2458 + }, + { + "epoch": 0.07755874467749567, + "grad_norm": 4.381267547607422, + "learning_rate": 1.988188965215792e-05, + "loss": 0.6121, + "step": 2459 + }, + { + "epoch": 0.07759028544393629, + "grad_norm": 4.905358791351318, + "learning_rate": 1.9881733057041994e-05, + "loss": 0.6018, + "step": 2460 + }, + { + "epoch": 0.07762182621037692, + "grad_norm": 4.228963851928711, + "learning_rate": 1.9881576358802513e-05, + "loss": 0.6343, + "step": 2461 + }, + { + "epoch": 0.07765336697681753, + "grad_norm": 3.206688642501831, + "learning_rate": 1.9881419557441108e-05, + "loss": 0.6275, + "step": 2462 + }, + { + "epoch": 0.07768490774325816, + "grad_norm": 3.98160982131958, + "learning_rate": 1.9881262652959415e-05, + "loss": 0.6312, + "step": 2463 + }, + { + "epoch": 0.07771644850969879, + "grad_norm": 3.53053617477417, + "learning_rate": 1.9881105645359077e-05, + "loss": 0.6334, + "step": 2464 + }, + { + "epoch": 0.0777479892761394, + "grad_norm": 4.209408760070801, + "learning_rate": 1.988094853464173e-05, + "loss": 0.5709, + "step": 2465 + }, + { + "epoch": 0.07777953004258004, + "grad_norm": 2.374502420425415, + "learning_rate": 1.9880791320809012e-05, + "loss": 0.6596, + "step": 2466 + }, + { + "epoch": 0.07781107080902067, + "grad_norm": 3.1976070404052734, + "learning_rate": 1.9880634003862565e-05, + "loss": 0.608, + "step": 2467 + }, + { + "epoch": 0.07784261157546128, + "grad_norm": 3.401745319366455, + "learning_rate": 1.9880476583804026e-05, + "loss": 0.617, + "step": 2468 + }, + { + "epoch": 0.07787415234190191, + "grad_norm": 2.5468177795410156, + "learning_rate": 1.988031906063505e-05, + "loss": 0.3623, + "step": 2469 + }, + { + "epoch": 0.07790569310834253, + "grad_norm": 3.3390133380889893, + "learning_rate": 1.9880161434357262e-05, + "loss": 0.6087, + "step": 2470 + }, + { + "epoch": 0.07793723387478316, + "grad_norm": 4.714588165283203, + "learning_rate": 1.9880003704972326e-05, + "loss": 0.6252, + "step": 2471 + }, + { + "epoch": 0.07796877464122379, + "grad_norm": 3.6935293674468994, + "learning_rate": 1.9879845872481875e-05, + "loss": 0.6099, + "step": 2472 + }, + { + "epoch": 0.0780003154076644, + "grad_norm": 3.2338123321533203, + "learning_rate": 1.987968793688756e-05, + "loss": 0.6489, + "step": 2473 + }, + { + "epoch": 0.07803185617410503, + "grad_norm": 3.143054485321045, + "learning_rate": 1.9879529898191038e-05, + "loss": 0.6006, + "step": 2474 + }, + { + "epoch": 0.07806339694054565, + "grad_norm": 2.3082876205444336, + "learning_rate": 1.9879371756393943e-05, + "loss": 0.9363, + "step": 2475 + }, + { + "epoch": 0.07809493770698628, + "grad_norm": 3.4940643310546875, + "learning_rate": 1.9879213511497935e-05, + "loss": 0.6331, + "step": 2476 + }, + { + "epoch": 0.07812647847342691, + "grad_norm": 11.952027320861816, + "learning_rate": 1.987905516350466e-05, + "loss": 0.645, + "step": 2477 + }, + { + "epoch": 0.07815801923986752, + "grad_norm": 4.193983554840088, + "learning_rate": 1.9878896712415776e-05, + "loss": 0.9111, + "step": 2478 + }, + { + "epoch": 0.07818956000630815, + "grad_norm": 3.645787477493286, + "learning_rate": 1.9878738158232935e-05, + "loss": 0.6296, + "step": 2479 + }, + { + "epoch": 0.07822110077274878, + "grad_norm": 3.1180741786956787, + "learning_rate": 1.987857950095779e-05, + "loss": 0.3841, + "step": 2480 + }, + { + "epoch": 0.0782526415391894, + "grad_norm": 2.4835498332977295, + "learning_rate": 1.9878420740591996e-05, + "loss": 0.3942, + "step": 2481 + }, + { + "epoch": 0.07828418230563003, + "grad_norm": 2.742525577545166, + "learning_rate": 1.9878261877137212e-05, + "loss": 0.9113, + "step": 2482 + }, + { + "epoch": 0.07831572307207064, + "grad_norm": 3.8710522651672363, + "learning_rate": 1.9878102910595097e-05, + "loss": 0.6255, + "step": 2483 + }, + { + "epoch": 0.07834726383851127, + "grad_norm": 4.5861101150512695, + "learning_rate": 1.9877943840967306e-05, + "loss": 0.6101, + "step": 2484 + }, + { + "epoch": 0.0783788046049519, + "grad_norm": 3.105113983154297, + "learning_rate": 1.98777846682555e-05, + "loss": 0.5814, + "step": 2485 + }, + { + "epoch": 0.07841034537139252, + "grad_norm": 26.799846649169922, + "learning_rate": 1.987762539246134e-05, + "loss": 0.8675, + "step": 2486 + }, + { + "epoch": 0.07844188613783315, + "grad_norm": 3.7351975440979004, + "learning_rate": 1.9877466013586494e-05, + "loss": 0.8608, + "step": 2487 + }, + { + "epoch": 0.07847342690427378, + "grad_norm": 3.4995718002319336, + "learning_rate": 1.987730653163262e-05, + "loss": 0.6266, + "step": 2488 + }, + { + "epoch": 0.0785049676707144, + "grad_norm": 5.603728294372559, + "learning_rate": 1.987714694660138e-05, + "loss": 0.6281, + "step": 2489 + }, + { + "epoch": 0.07853650843715503, + "grad_norm": 3.7277989387512207, + "learning_rate": 1.9876987258494443e-05, + "loss": 0.6128, + "step": 2490 + }, + { + "epoch": 0.07856804920359564, + "grad_norm": 4.7586750984191895, + "learning_rate": 1.9876827467313474e-05, + "loss": 0.6243, + "step": 2491 + }, + { + "epoch": 0.07859958997003627, + "grad_norm": 7.26305627822876, + "learning_rate": 1.987666757306014e-05, + "loss": 0.6489, + "step": 2492 + }, + { + "epoch": 0.0786311307364769, + "grad_norm": 5.977023124694824, + "learning_rate": 1.9876507575736115e-05, + "loss": 0.6107, + "step": 2493 + }, + { + "epoch": 0.07866267150291752, + "grad_norm": 4.5975751876831055, + "learning_rate": 1.9876347475343062e-05, + "loss": 0.604, + "step": 2494 + }, + { + "epoch": 0.07869421226935815, + "grad_norm": 3.5076382160186768, + "learning_rate": 1.9876187271882654e-05, + "loss": 0.6046, + "step": 2495 + }, + { + "epoch": 0.07872575303579878, + "grad_norm": 2.7547097206115723, + "learning_rate": 1.9876026965356564e-05, + "loss": 0.6075, + "step": 2496 + }, + { + "epoch": 0.07875729380223939, + "grad_norm": 5.310371398925781, + "learning_rate": 1.9875866555766463e-05, + "loss": 0.599, + "step": 2497 + }, + { + "epoch": 0.07878883456868002, + "grad_norm": 2.560279130935669, + "learning_rate": 1.9875706043114028e-05, + "loss": 0.8617, + "step": 2498 + }, + { + "epoch": 0.07882037533512064, + "grad_norm": 3.2821309566497803, + "learning_rate": 1.9875545427400932e-05, + "loss": 0.8816, + "step": 2499 + }, + { + "epoch": 0.07885191610156127, + "grad_norm": 2.5433874130249023, + "learning_rate": 1.987538470862885e-05, + "loss": 0.4308, + "step": 2500 + }, + { + "epoch": 0.0788834568680019, + "grad_norm": 4.6898512840271, + "learning_rate": 1.987522388679946e-05, + "loss": 0.9105, + "step": 2501 + }, + { + "epoch": 0.07891499763444251, + "grad_norm": 2.922309398651123, + "learning_rate": 1.9875062961914442e-05, + "loss": 0.5922, + "step": 2502 + }, + { + "epoch": 0.07894653840088314, + "grad_norm": 4.196433067321777, + "learning_rate": 1.9874901933975473e-05, + "loss": 0.6183, + "step": 2503 + }, + { + "epoch": 0.07897807916732377, + "grad_norm": 4.067763328552246, + "learning_rate": 1.9874740802984236e-05, + "loss": 0.6227, + "step": 2504 + }, + { + "epoch": 0.07900961993376439, + "grad_norm": 3.1676154136657715, + "learning_rate": 1.987457956894241e-05, + "loss": 0.6317, + "step": 2505 + }, + { + "epoch": 0.07904116070020502, + "grad_norm": 3.745601177215576, + "learning_rate": 1.987441823185168e-05, + "loss": 0.6049, + "step": 2506 + }, + { + "epoch": 0.07907270146664563, + "grad_norm": 2.909680128097534, + "learning_rate": 1.9874256791713727e-05, + "loss": 0.617, + "step": 2507 + }, + { + "epoch": 0.07910424223308626, + "grad_norm": 3.750312089920044, + "learning_rate": 1.9874095248530237e-05, + "loss": 0.6422, + "step": 2508 + }, + { + "epoch": 0.0791357829995269, + "grad_norm": 3.549450397491455, + "learning_rate": 1.98739336023029e-05, + "loss": 0.6311, + "step": 2509 + }, + { + "epoch": 0.07916732376596751, + "grad_norm": 4.634005069732666, + "learning_rate": 1.9873771853033393e-05, + "loss": 0.6077, + "step": 2510 + }, + { + "epoch": 0.07919886453240814, + "grad_norm": 3.0016696453094482, + "learning_rate": 1.9873610000723417e-05, + "loss": 0.6025, + "step": 2511 + }, + { + "epoch": 0.07923040529884877, + "grad_norm": 4.468123435974121, + "learning_rate": 1.987344804537465e-05, + "loss": 0.6527, + "step": 2512 + }, + { + "epoch": 0.07926194606528938, + "grad_norm": 4.090685844421387, + "learning_rate": 1.9873285986988784e-05, + "loss": 0.6139, + "step": 2513 + }, + { + "epoch": 0.07929348683173001, + "grad_norm": 3.763012647628784, + "learning_rate": 1.9873123825567514e-05, + "loss": 0.6121, + "step": 2514 + }, + { + "epoch": 0.07932502759817063, + "grad_norm": 5.391343116760254, + "learning_rate": 1.987296156111253e-05, + "loss": 0.641, + "step": 2515 + }, + { + "epoch": 0.07935656836461126, + "grad_norm": 4.483312129974365, + "learning_rate": 1.987279919362553e-05, + "loss": 0.6589, + "step": 2516 + }, + { + "epoch": 0.07938810913105189, + "grad_norm": 4.239063262939453, + "learning_rate": 1.98726367231082e-05, + "loss": 0.6378, + "step": 2517 + }, + { + "epoch": 0.0794196498974925, + "grad_norm": 5.833364009857178, + "learning_rate": 1.9872474149562243e-05, + "loss": 0.6132, + "step": 2518 + }, + { + "epoch": 0.07945119066393314, + "grad_norm": 3.204585552215576, + "learning_rate": 1.987231147298935e-05, + "loss": 0.6461, + "step": 2519 + }, + { + "epoch": 0.07948273143037375, + "grad_norm": 16.035198211669922, + "learning_rate": 1.9872148693391224e-05, + "loss": 0.6293, + "step": 2520 + }, + { + "epoch": 0.07951427219681438, + "grad_norm": 4.112507343292236, + "learning_rate": 1.9871985810769557e-05, + "loss": 0.5931, + "step": 2521 + }, + { + "epoch": 0.07954581296325501, + "grad_norm": 4.532461643218994, + "learning_rate": 1.9871822825126055e-05, + "loss": 0.6198, + "step": 2522 + }, + { + "epoch": 0.07957735372969563, + "grad_norm": 3.9619903564453125, + "learning_rate": 1.9871659736462418e-05, + "loss": 0.5846, + "step": 2523 + }, + { + "epoch": 0.07960889449613626, + "grad_norm": 3.574172019958496, + "learning_rate": 1.9871496544780343e-05, + "loss": 0.6385, + "step": 2524 + }, + { + "epoch": 0.07964043526257689, + "grad_norm": 4.095443248748779, + "learning_rate": 1.987133325008154e-05, + "loss": 0.617, + "step": 2525 + }, + { + "epoch": 0.0796719760290175, + "grad_norm": 3.7327935695648193, + "learning_rate": 1.9871169852367705e-05, + "loss": 0.5902, + "step": 2526 + }, + { + "epoch": 0.07970351679545813, + "grad_norm": 4.307639122009277, + "learning_rate": 1.987100635164055e-05, + "loss": 0.6337, + "step": 2527 + }, + { + "epoch": 0.07973505756189875, + "grad_norm": 3.673461437225342, + "learning_rate": 1.9870842747901782e-05, + "loss": 0.6096, + "step": 2528 + }, + { + "epoch": 0.07976659832833938, + "grad_norm": 6.963111877441406, + "learning_rate": 1.98706790411531e-05, + "loss": 0.6246, + "step": 2529 + }, + { + "epoch": 0.07979813909478001, + "grad_norm": 3.4908225536346436, + "learning_rate": 1.987051523139622e-05, + "loss": 0.4021, + "step": 2530 + }, + { + "epoch": 0.07982967986122062, + "grad_norm": 4.216615676879883, + "learning_rate": 1.987035131863285e-05, + "loss": 0.6676, + "step": 2531 + }, + { + "epoch": 0.07986122062766125, + "grad_norm": 3.388474702835083, + "learning_rate": 1.9870187302864704e-05, + "loss": 0.6047, + "step": 2532 + }, + { + "epoch": 0.07989276139410188, + "grad_norm": 4.296186923980713, + "learning_rate": 1.9870023184093484e-05, + "loss": 0.6262, + "step": 2533 + }, + { + "epoch": 0.0799243021605425, + "grad_norm": 4.294283866882324, + "learning_rate": 1.986985896232091e-05, + "loss": 0.6387, + "step": 2534 + }, + { + "epoch": 0.07995584292698313, + "grad_norm": 2.0405397415161133, + "learning_rate": 1.9869694637548693e-05, + "loss": 0.3913, + "step": 2535 + }, + { + "epoch": 0.07998738369342374, + "grad_norm": 4.300140380859375, + "learning_rate": 1.986953020977855e-05, + "loss": 0.6115, + "step": 2536 + }, + { + "epoch": 0.08001892445986437, + "grad_norm": 5.405585765838623, + "learning_rate": 1.9869365679012196e-05, + "loss": 0.6052, + "step": 2537 + }, + { + "epoch": 0.080050465226305, + "grad_norm": 5.762316703796387, + "learning_rate": 1.9869201045251345e-05, + "loss": 0.6071, + "step": 2538 + }, + { + "epoch": 0.08008200599274562, + "grad_norm": 3.079984188079834, + "learning_rate": 1.986903630849772e-05, + "loss": 0.9136, + "step": 2539 + }, + { + "epoch": 0.08011354675918625, + "grad_norm": 6.6566267013549805, + "learning_rate": 1.986887146875304e-05, + "loss": 0.6321, + "step": 2540 + }, + { + "epoch": 0.08014508752562688, + "grad_norm": 4.121115207672119, + "learning_rate": 1.9868706526019022e-05, + "loss": 0.6227, + "step": 2541 + }, + { + "epoch": 0.0801766282920675, + "grad_norm": 3.1532864570617676, + "learning_rate": 1.9868541480297385e-05, + "loss": 0.6163, + "step": 2542 + }, + { + "epoch": 0.08020816905850812, + "grad_norm": 2.6910572052001953, + "learning_rate": 1.9868376331589857e-05, + "loss": 0.9135, + "step": 2543 + }, + { + "epoch": 0.08023970982494874, + "grad_norm": 4.567962169647217, + "learning_rate": 1.9868211079898162e-05, + "loss": 0.6189, + "step": 2544 + }, + { + "epoch": 0.08027125059138937, + "grad_norm": 4.419976711273193, + "learning_rate": 1.986804572522402e-05, + "loss": 0.6225, + "step": 2545 + }, + { + "epoch": 0.08030279135783, + "grad_norm": 4.286920547485352, + "learning_rate": 1.986788026756916e-05, + "loss": 0.4252, + "step": 2546 + }, + { + "epoch": 0.08033433212427062, + "grad_norm": 6.90487003326416, + "learning_rate": 1.9867714706935305e-05, + "loss": 0.9094, + "step": 2547 + }, + { + "epoch": 0.08036587289071125, + "grad_norm": 5.105253219604492, + "learning_rate": 1.9867549043324183e-05, + "loss": 0.6263, + "step": 2548 + }, + { + "epoch": 0.08039741365715188, + "grad_norm": 7.350461006164551, + "learning_rate": 1.9867383276737528e-05, + "loss": 0.5942, + "step": 2549 + }, + { + "epoch": 0.08042895442359249, + "grad_norm": 3.9213404655456543, + "learning_rate": 1.9867217407177067e-05, + "loss": 0.6476, + "step": 2550 + }, + { + "epoch": 0.08046049519003312, + "grad_norm": 5.196112155914307, + "learning_rate": 1.9867051434644532e-05, + "loss": 0.6789, + "step": 2551 + }, + { + "epoch": 0.08049203595647374, + "grad_norm": 5.950503826141357, + "learning_rate": 1.9866885359141648e-05, + "loss": 0.6238, + "step": 2552 + }, + { + "epoch": 0.08052357672291437, + "grad_norm": 3.3354785442352295, + "learning_rate": 1.986671918067016e-05, + "loss": 0.9025, + "step": 2553 + }, + { + "epoch": 0.080555117489355, + "grad_norm": 4.5794243812561035, + "learning_rate": 1.9866552899231792e-05, + "loss": 0.6452, + "step": 2554 + }, + { + "epoch": 0.08058665825579561, + "grad_norm": 3.6169419288635254, + "learning_rate": 1.9866386514828284e-05, + "loss": 0.6149, + "step": 2555 + }, + { + "epoch": 0.08061819902223624, + "grad_norm": 3.3292694091796875, + "learning_rate": 1.9866220027461375e-05, + "loss": 0.645, + "step": 2556 + }, + { + "epoch": 0.08064973978867686, + "grad_norm": 4.520819664001465, + "learning_rate": 1.9866053437132793e-05, + "loss": 0.5947, + "step": 2557 + }, + { + "epoch": 0.08068128055511749, + "grad_norm": 4.367111682891846, + "learning_rate": 1.9865886743844287e-05, + "loss": 0.6439, + "step": 2558 + }, + { + "epoch": 0.08071282132155812, + "grad_norm": 6.152804374694824, + "learning_rate": 1.9865719947597592e-05, + "loss": 0.6239, + "step": 2559 + }, + { + "epoch": 0.08074436208799873, + "grad_norm": 4.078658580780029, + "learning_rate": 1.9865553048394445e-05, + "loss": 0.5948, + "step": 2560 + }, + { + "epoch": 0.08077590285443936, + "grad_norm": 4.44551420211792, + "learning_rate": 1.9865386046236597e-05, + "loss": 0.6605, + "step": 2561 + }, + { + "epoch": 0.08080744362087999, + "grad_norm": 5.773189544677734, + "learning_rate": 1.9865218941125782e-05, + "loss": 0.6286, + "step": 2562 + }, + { + "epoch": 0.08083898438732061, + "grad_norm": 5.498377323150635, + "learning_rate": 1.986505173306375e-05, + "loss": 0.6123, + "step": 2563 + }, + { + "epoch": 0.08087052515376124, + "grad_norm": 8.093896865844727, + "learning_rate": 1.9864884422052238e-05, + "loss": 0.6046, + "step": 2564 + }, + { + "epoch": 0.08090206592020185, + "grad_norm": 5.710133075714111, + "learning_rate": 1.9864717008093003e-05, + "loss": 0.6592, + "step": 2565 + }, + { + "epoch": 0.08093360668664248, + "grad_norm": 5.219663619995117, + "learning_rate": 1.9864549491187782e-05, + "loss": 0.6249, + "step": 2566 + }, + { + "epoch": 0.08096514745308311, + "grad_norm": 4.012958526611328, + "learning_rate": 1.986438187133833e-05, + "loss": 0.6234, + "step": 2567 + }, + { + "epoch": 0.08099668821952373, + "grad_norm": 4.05469274520874, + "learning_rate": 1.986421414854639e-05, + "loss": 0.6393, + "step": 2568 + }, + { + "epoch": 0.08102822898596436, + "grad_norm": 3.753427505493164, + "learning_rate": 1.986404632281372e-05, + "loss": 0.6294, + "step": 2569 + }, + { + "epoch": 0.08105976975240499, + "grad_norm": 3.6668519973754883, + "learning_rate": 1.986387839414207e-05, + "loss": 0.8525, + "step": 2570 + }, + { + "epoch": 0.0810913105188456, + "grad_norm": 3.5741312503814697, + "learning_rate": 1.9863710362533183e-05, + "loss": 0.6396, + "step": 2571 + }, + { + "epoch": 0.08112285128528623, + "grad_norm": 3.0118837356567383, + "learning_rate": 1.9863542227988823e-05, + "loss": 0.605, + "step": 2572 + }, + { + "epoch": 0.08115439205172685, + "grad_norm": 3.809234142303467, + "learning_rate": 1.986337399051074e-05, + "loss": 0.6182, + "step": 2573 + }, + { + "epoch": 0.08118593281816748, + "grad_norm": 3.9072513580322266, + "learning_rate": 1.986320565010069e-05, + "loss": 0.5869, + "step": 2574 + }, + { + "epoch": 0.08121747358460811, + "grad_norm": 4.2674560546875, + "learning_rate": 1.9863037206760435e-05, + "loss": 0.639, + "step": 2575 + }, + { + "epoch": 0.08124901435104873, + "grad_norm": 3.3577122688293457, + "learning_rate": 1.9862868660491723e-05, + "loss": 0.6028, + "step": 2576 + }, + { + "epoch": 0.08128055511748936, + "grad_norm": 6.730068683624268, + "learning_rate": 1.9862700011296318e-05, + "loss": 0.6433, + "step": 2577 + }, + { + "epoch": 0.08131209588392999, + "grad_norm": 3.0480899810791016, + "learning_rate": 1.9862531259175985e-05, + "loss": 0.9228, + "step": 2578 + }, + { + "epoch": 0.0813436366503706, + "grad_norm": 4.213890075683594, + "learning_rate": 1.9862362404132474e-05, + "loss": 0.6133, + "step": 2579 + }, + { + "epoch": 0.08137517741681123, + "grad_norm": 4.775226593017578, + "learning_rate": 1.986219344616756e-05, + "loss": 0.6192, + "step": 2580 + }, + { + "epoch": 0.08140671818325185, + "grad_norm": 4.996711254119873, + "learning_rate": 1.9862024385282995e-05, + "loss": 0.5975, + "step": 2581 + }, + { + "epoch": 0.08143825894969248, + "grad_norm": 8.85782527923584, + "learning_rate": 1.986185522148055e-05, + "loss": 0.8951, + "step": 2582 + }, + { + "epoch": 0.0814697997161331, + "grad_norm": 3.95892596244812, + "learning_rate": 1.9861685954761988e-05, + "loss": 0.6427, + "step": 2583 + }, + { + "epoch": 0.08150134048257372, + "grad_norm": 3.6906871795654297, + "learning_rate": 1.9861516585129076e-05, + "loss": 0.6531, + "step": 2584 + }, + { + "epoch": 0.08153288124901435, + "grad_norm": 3.066507339477539, + "learning_rate": 1.9861347112583578e-05, + "loss": 0.6382, + "step": 2585 + }, + { + "epoch": 0.08156442201545498, + "grad_norm": 6.8857879638671875, + "learning_rate": 1.986117753712727e-05, + "loss": 0.5968, + "step": 2586 + }, + { + "epoch": 0.0815959627818956, + "grad_norm": 5.785585880279541, + "learning_rate": 1.9861007858761915e-05, + "loss": 0.6415, + "step": 2587 + }, + { + "epoch": 0.08162750354833623, + "grad_norm": 6.446258544921875, + "learning_rate": 1.9860838077489286e-05, + "loss": 0.8565, + "step": 2588 + }, + { + "epoch": 0.08165904431477684, + "grad_norm": 3.9149608612060547, + "learning_rate": 1.9860668193311157e-05, + "loss": 0.6358, + "step": 2589 + }, + { + "epoch": 0.08169058508121747, + "grad_norm": 8.145187377929688, + "learning_rate": 1.98604982062293e-05, + "loss": 0.389, + "step": 2590 + }, + { + "epoch": 0.0817221258476581, + "grad_norm": 3.428645372390747, + "learning_rate": 1.986032811624548e-05, + "loss": 0.6194, + "step": 2591 + }, + { + "epoch": 0.08175366661409872, + "grad_norm": 5.446824550628662, + "learning_rate": 1.9860157923361487e-05, + "loss": 0.5955, + "step": 2592 + }, + { + "epoch": 0.08178520738053935, + "grad_norm": 5.517601490020752, + "learning_rate": 1.9859987627579088e-05, + "loss": 0.6205, + "step": 2593 + }, + { + "epoch": 0.08181674814697998, + "grad_norm": 4.530722618103027, + "learning_rate": 1.985981722890006e-05, + "loss": 0.6394, + "step": 2594 + }, + { + "epoch": 0.0818482889134206, + "grad_norm": 5.256282329559326, + "learning_rate": 1.9859646727326187e-05, + "loss": 0.6496, + "step": 2595 + }, + { + "epoch": 0.08187982967986122, + "grad_norm": 4.437375068664551, + "learning_rate": 1.9859476122859242e-05, + "loss": 0.631, + "step": 2596 + }, + { + "epoch": 0.08191137044630184, + "grad_norm": 7.9214301109313965, + "learning_rate": 1.985930541550101e-05, + "loss": 0.6146, + "step": 2597 + }, + { + "epoch": 0.08194291121274247, + "grad_norm": 8.382338523864746, + "learning_rate": 1.9859134605253268e-05, + "loss": 0.3988, + "step": 2598 + }, + { + "epoch": 0.0819744519791831, + "grad_norm": 8.02255916595459, + "learning_rate": 1.9858963692117804e-05, + "loss": 0.638, + "step": 2599 + }, + { + "epoch": 0.08200599274562372, + "grad_norm": 6.771831512451172, + "learning_rate": 1.9858792676096395e-05, + "loss": 0.6017, + "step": 2600 + }, + { + "epoch": 0.08203753351206435, + "grad_norm": 3.590074300765991, + "learning_rate": 1.985862155719083e-05, + "loss": 0.6092, + "step": 2601 + }, + { + "epoch": 0.08206907427850496, + "grad_norm": 2.2025017738342285, + "learning_rate": 1.98584503354029e-05, + "loss": 0.4152, + "step": 2602 + }, + { + "epoch": 0.08210061504494559, + "grad_norm": 4.597261905670166, + "learning_rate": 1.9858279010734376e-05, + "loss": 0.6283, + "step": 2603 + }, + { + "epoch": 0.08213215581138622, + "grad_norm": 7.486905574798584, + "learning_rate": 1.985810758318706e-05, + "loss": 0.3813, + "step": 2604 + }, + { + "epoch": 0.08216369657782684, + "grad_norm": 6.000264644622803, + "learning_rate": 1.985793605276274e-05, + "loss": 0.6519, + "step": 2605 + }, + { + "epoch": 0.08219523734426747, + "grad_norm": 6.315907001495361, + "learning_rate": 1.9857764419463197e-05, + "loss": 0.647, + "step": 2606 + }, + { + "epoch": 0.0822267781107081, + "grad_norm": 4.325610160827637, + "learning_rate": 1.9857592683290233e-05, + "loss": 0.6498, + "step": 2607 + }, + { + "epoch": 0.08225831887714871, + "grad_norm": 3.9284870624542236, + "learning_rate": 1.985742084424563e-05, + "loss": 0.6249, + "step": 2608 + }, + { + "epoch": 0.08228985964358934, + "grad_norm": 8.30870246887207, + "learning_rate": 1.9857248902331193e-05, + "loss": 0.6108, + "step": 2609 + }, + { + "epoch": 0.08232140041002996, + "grad_norm": 6.637974262237549, + "learning_rate": 1.9857076857548706e-05, + "loss": 0.619, + "step": 2610 + }, + { + "epoch": 0.08235294117647059, + "grad_norm": 5.511855602264404, + "learning_rate": 1.9856904709899967e-05, + "loss": 0.6239, + "step": 2611 + }, + { + "epoch": 0.08238448194291122, + "grad_norm": 5.095653533935547, + "learning_rate": 1.9856732459386774e-05, + "loss": 0.6308, + "step": 2612 + }, + { + "epoch": 0.08241602270935183, + "grad_norm": 5.155016899108887, + "learning_rate": 1.9856560106010926e-05, + "loss": 0.6024, + "step": 2613 + }, + { + "epoch": 0.08244756347579246, + "grad_norm": 4.965613842010498, + "learning_rate": 1.9856387649774217e-05, + "loss": 0.5819, + "step": 2614 + }, + { + "epoch": 0.08247910424223309, + "grad_norm": 3.48020339012146, + "learning_rate": 1.9856215090678454e-05, + "loss": 0.6133, + "step": 2615 + }, + { + "epoch": 0.08251064500867371, + "grad_norm": 4.766198635101318, + "learning_rate": 1.985604242872543e-05, + "loss": 0.4076, + "step": 2616 + }, + { + "epoch": 0.08254218577511434, + "grad_norm": 4.63818883895874, + "learning_rate": 1.9855869663916953e-05, + "loss": 0.5954, + "step": 2617 + }, + { + "epoch": 0.08257372654155495, + "grad_norm": 4.183529853820801, + "learning_rate": 1.9855696796254822e-05, + "loss": 0.6504, + "step": 2618 + }, + { + "epoch": 0.08260526730799558, + "grad_norm": 4.187489032745361, + "learning_rate": 1.9855523825740842e-05, + "loss": 0.5806, + "step": 2619 + }, + { + "epoch": 0.08263680807443621, + "grad_norm": 3.40747332572937, + "learning_rate": 1.9855350752376815e-05, + "loss": 0.6045, + "step": 2620 + }, + { + "epoch": 0.08266834884087683, + "grad_norm": 4.1702046394348145, + "learning_rate": 1.9855177576164557e-05, + "loss": 0.6058, + "step": 2621 + }, + { + "epoch": 0.08269988960731746, + "grad_norm": 11.835407257080078, + "learning_rate": 1.9855004297105866e-05, + "loss": 0.6067, + "step": 2622 + }, + { + "epoch": 0.08273143037375809, + "grad_norm": 3.8275818824768066, + "learning_rate": 1.985483091520255e-05, + "loss": 0.3742, + "step": 2623 + }, + { + "epoch": 0.0827629711401987, + "grad_norm": 3.687678337097168, + "learning_rate": 1.9854657430456422e-05, + "loss": 0.6196, + "step": 2624 + }, + { + "epoch": 0.08279451190663933, + "grad_norm": 3.062126874923706, + "learning_rate": 1.9854483842869294e-05, + "loss": 0.9344, + "step": 2625 + }, + { + "epoch": 0.08282605267307995, + "grad_norm": 4.103097915649414, + "learning_rate": 1.9854310152442974e-05, + "loss": 0.6462, + "step": 2626 + }, + { + "epoch": 0.08285759343952058, + "grad_norm": 4.115444183349609, + "learning_rate": 1.985413635917928e-05, + "loss": 0.6373, + "step": 2627 + }, + { + "epoch": 0.08288913420596121, + "grad_norm": 3.917208433151245, + "learning_rate": 1.9853962463080013e-05, + "loss": 0.6673, + "step": 2628 + }, + { + "epoch": 0.08292067497240183, + "grad_norm": 4.678324222564697, + "learning_rate": 1.9853788464147e-05, + "loss": 0.6304, + "step": 2629 + }, + { + "epoch": 0.08295221573884246, + "grad_norm": 5.776969909667969, + "learning_rate": 1.9853614362382058e-05, + "loss": 0.6288, + "step": 2630 + }, + { + "epoch": 0.08298375650528309, + "grad_norm": 3.7820394039154053, + "learning_rate": 1.9853440157786993e-05, + "loss": 0.6227, + "step": 2631 + }, + { + "epoch": 0.0830152972717237, + "grad_norm": 13.507676124572754, + "learning_rate": 1.9853265850363633e-05, + "loss": 0.6514, + "step": 2632 + }, + { + "epoch": 0.08304683803816433, + "grad_norm": 5.194952487945557, + "learning_rate": 1.985309144011379e-05, + "loss": 0.6055, + "step": 2633 + }, + { + "epoch": 0.08307837880460495, + "grad_norm": 5.0334696769714355, + "learning_rate": 1.9852916927039286e-05, + "loss": 0.6214, + "step": 2634 + }, + { + "epoch": 0.08310991957104558, + "grad_norm": 4.048834323883057, + "learning_rate": 1.985274231114195e-05, + "loss": 0.6244, + "step": 2635 + }, + { + "epoch": 0.0831414603374862, + "grad_norm": 4.649792194366455, + "learning_rate": 1.985256759242359e-05, + "loss": 0.6042, + "step": 2636 + }, + { + "epoch": 0.08317300110392682, + "grad_norm": 3.178823947906494, + "learning_rate": 1.9852392770886042e-05, + "loss": 0.9298, + "step": 2637 + }, + { + "epoch": 0.08320454187036745, + "grad_norm": 4.320199966430664, + "learning_rate": 1.985221784653112e-05, + "loss": 0.588, + "step": 2638 + }, + { + "epoch": 0.08323608263680807, + "grad_norm": 4.377140522003174, + "learning_rate": 1.985204281936066e-05, + "loss": 0.6068, + "step": 2639 + }, + { + "epoch": 0.0832676234032487, + "grad_norm": 4.094829082489014, + "learning_rate": 1.9851867689376482e-05, + "loss": 0.6238, + "step": 2640 + }, + { + "epoch": 0.08329916416968933, + "grad_norm": 3.644998550415039, + "learning_rate": 1.9851692456580415e-05, + "loss": 0.3897, + "step": 2641 + }, + { + "epoch": 0.08333070493612994, + "grad_norm": 5.425786972045898, + "learning_rate": 1.9851517120974283e-05, + "loss": 0.6176, + "step": 2642 + }, + { + "epoch": 0.08336224570257057, + "grad_norm": 7.991919040679932, + "learning_rate": 1.9851341682559925e-05, + "loss": 0.6509, + "step": 2643 + }, + { + "epoch": 0.0833937864690112, + "grad_norm": 3.5841567516326904, + "learning_rate": 1.9851166141339168e-05, + "loss": 0.6271, + "step": 2644 + }, + { + "epoch": 0.08342532723545182, + "grad_norm": 6.242285251617432, + "learning_rate": 1.985099049731384e-05, + "loss": 0.6497, + "step": 2645 + }, + { + "epoch": 0.08345686800189245, + "grad_norm": 6.1902689933776855, + "learning_rate": 1.985081475048578e-05, + "loss": 0.6158, + "step": 2646 + }, + { + "epoch": 0.08348840876833306, + "grad_norm": 4.890817642211914, + "learning_rate": 1.9850638900856816e-05, + "loss": 0.6181, + "step": 2647 + }, + { + "epoch": 0.0835199495347737, + "grad_norm": 5.510926723480225, + "learning_rate": 1.9850462948428786e-05, + "loss": 0.606, + "step": 2648 + }, + { + "epoch": 0.08355149030121432, + "grad_norm": 4.650176525115967, + "learning_rate": 1.985028689320353e-05, + "loss": 0.6237, + "step": 2649 + }, + { + "epoch": 0.08358303106765494, + "grad_norm": 4.9918389320373535, + "learning_rate": 1.985011073518288e-05, + "loss": 0.6324, + "step": 2650 + }, + { + "epoch": 0.08361457183409557, + "grad_norm": 4.786970138549805, + "learning_rate": 1.9849934474368673e-05, + "loss": 0.6416, + "step": 2651 + }, + { + "epoch": 0.0836461126005362, + "grad_norm": 3.192863941192627, + "learning_rate": 1.9849758110762757e-05, + "loss": 0.6158, + "step": 2652 + }, + { + "epoch": 0.08367765336697681, + "grad_norm": 3.891758441925049, + "learning_rate": 1.9849581644366964e-05, + "loss": 0.6348, + "step": 2653 + }, + { + "epoch": 0.08370919413341744, + "grad_norm": 4.796941757202148, + "learning_rate": 1.9849405075183136e-05, + "loss": 0.6262, + "step": 2654 + }, + { + "epoch": 0.08374073489985806, + "grad_norm": 3.912860870361328, + "learning_rate": 1.9849228403213124e-05, + "loss": 0.6136, + "step": 2655 + }, + { + "epoch": 0.08377227566629869, + "grad_norm": 4.487072467803955, + "learning_rate": 1.984905162845876e-05, + "loss": 0.6118, + "step": 2656 + }, + { + "epoch": 0.08380381643273932, + "grad_norm": 4.369618892669678, + "learning_rate": 1.9848874750921902e-05, + "loss": 0.6095, + "step": 2657 + }, + { + "epoch": 0.08383535719917994, + "grad_norm": 2.652750015258789, + "learning_rate": 1.9848697770604382e-05, + "loss": 0.4499, + "step": 2658 + }, + { + "epoch": 0.08386689796562057, + "grad_norm": 5.879687309265137, + "learning_rate": 1.9848520687508057e-05, + "loss": 0.6294, + "step": 2659 + }, + { + "epoch": 0.0838984387320612, + "grad_norm": 7.7289958000183105, + "learning_rate": 1.9848343501634767e-05, + "loss": 0.615, + "step": 2660 + }, + { + "epoch": 0.08392997949850181, + "grad_norm": 7.215999126434326, + "learning_rate": 1.984816621298637e-05, + "loss": 0.6141, + "step": 2661 + }, + { + "epoch": 0.08396152026494244, + "grad_norm": 3.2171082496643066, + "learning_rate": 1.984798882156471e-05, + "loss": 0.6245, + "step": 2662 + }, + { + "epoch": 0.08399306103138306, + "grad_norm": 3.281111478805542, + "learning_rate": 1.984781132737164e-05, + "loss": 0.6082, + "step": 2663 + }, + { + "epoch": 0.08402460179782369, + "grad_norm": 5.2078022956848145, + "learning_rate": 1.9847633730409012e-05, + "loss": 0.5977, + "step": 2664 + }, + { + "epoch": 0.08405614256426432, + "grad_norm": 6.108412742614746, + "learning_rate": 1.984745603067868e-05, + "loss": 0.6129, + "step": 2665 + }, + { + "epoch": 0.08408768333070493, + "grad_norm": 6.746959209442139, + "learning_rate": 1.98472782281825e-05, + "loss": 0.6404, + "step": 2666 + }, + { + "epoch": 0.08411922409714556, + "grad_norm": 5.343451499938965, + "learning_rate": 1.984710032292232e-05, + "loss": 0.3881, + "step": 2667 + }, + { + "epoch": 0.08415076486358619, + "grad_norm": 2.74922251701355, + "learning_rate": 1.9846922314900007e-05, + "loss": 0.9001, + "step": 2668 + }, + { + "epoch": 0.08418230563002681, + "grad_norm": 7.911787509918213, + "learning_rate": 1.984674420411741e-05, + "loss": 0.6505, + "step": 2669 + }, + { + "epoch": 0.08421384639646744, + "grad_norm": 4.79511022567749, + "learning_rate": 1.9846565990576394e-05, + "loss": 0.4188, + "step": 2670 + }, + { + "epoch": 0.08424538716290805, + "grad_norm": 5.851134300231934, + "learning_rate": 1.9846387674278817e-05, + "loss": 0.636, + "step": 2671 + }, + { + "epoch": 0.08427692792934868, + "grad_norm": 5.224576473236084, + "learning_rate": 1.984620925522654e-05, + "loss": 0.6336, + "step": 2672 + }, + { + "epoch": 0.08430846869578931, + "grad_norm": 7.1876349449157715, + "learning_rate": 1.984603073342142e-05, + "loss": 0.6149, + "step": 2673 + }, + { + "epoch": 0.08434000946222993, + "grad_norm": 9.137673377990723, + "learning_rate": 1.9845852108865324e-05, + "loss": 0.6683, + "step": 2674 + }, + { + "epoch": 0.08437155022867056, + "grad_norm": 6.065802097320557, + "learning_rate": 1.984567338156012e-05, + "loss": 0.6496, + "step": 2675 + }, + { + "epoch": 0.08440309099511119, + "grad_norm": 5.117461681365967, + "learning_rate": 1.9845494551507663e-05, + "loss": 0.6343, + "step": 2676 + }, + { + "epoch": 0.0844346317615518, + "grad_norm": 6.680127143859863, + "learning_rate": 1.984531561870983e-05, + "loss": 0.643, + "step": 2677 + }, + { + "epoch": 0.08446617252799243, + "grad_norm": 4.78077507019043, + "learning_rate": 1.9845136583168485e-05, + "loss": 0.6419, + "step": 2678 + }, + { + "epoch": 0.08449771329443305, + "grad_norm": 6.332810878753662, + "learning_rate": 1.9844957444885495e-05, + "loss": 0.6216, + "step": 2679 + }, + { + "epoch": 0.08452925406087368, + "grad_norm": 5.053426265716553, + "learning_rate": 1.9844778203862728e-05, + "loss": 0.6437, + "step": 2680 + }, + { + "epoch": 0.08456079482731431, + "grad_norm": 7.396470546722412, + "learning_rate": 1.9844598860102057e-05, + "loss": 0.8648, + "step": 2681 + }, + { + "epoch": 0.08459233559375492, + "grad_norm": 2.945432186126709, + "learning_rate": 1.9844419413605353e-05, + "loss": 0.9185, + "step": 2682 + }, + { + "epoch": 0.08462387636019555, + "grad_norm": 4.632440567016602, + "learning_rate": 1.984423986437449e-05, + "loss": 0.6294, + "step": 2683 + }, + { + "epoch": 0.08465541712663617, + "grad_norm": 2.3194868564605713, + "learning_rate": 1.9844060212411336e-05, + "loss": 0.8769, + "step": 2684 + }, + { + "epoch": 0.0846869578930768, + "grad_norm": 4.43363618850708, + "learning_rate": 1.9843880457717774e-05, + "loss": 0.674, + "step": 2685 + }, + { + "epoch": 0.08471849865951743, + "grad_norm": 5.505178451538086, + "learning_rate": 1.984370060029567e-05, + "loss": 0.6924, + "step": 2686 + }, + { + "epoch": 0.08475003942595805, + "grad_norm": 4.289558410644531, + "learning_rate": 1.984352064014691e-05, + "loss": 0.6568, + "step": 2687 + }, + { + "epoch": 0.08478158019239868, + "grad_norm": 4.017284393310547, + "learning_rate": 1.984334057727337e-05, + "loss": 0.6165, + "step": 2688 + }, + { + "epoch": 0.0848131209588393, + "grad_norm": 5.6302409172058105, + "learning_rate": 1.9843160411676926e-05, + "loss": 0.6292, + "step": 2689 + }, + { + "epoch": 0.08484466172527992, + "grad_norm": 3.4216625690460205, + "learning_rate": 1.9842980143359464e-05, + "loss": 0.6456, + "step": 2690 + }, + { + "epoch": 0.08487620249172055, + "grad_norm": 4.3577070236206055, + "learning_rate": 1.984279977232286e-05, + "loss": 0.6442, + "step": 2691 + }, + { + "epoch": 0.08490774325816117, + "grad_norm": 6.724470615386963, + "learning_rate": 1.9842619298568994e-05, + "loss": 0.6383, + "step": 2692 + }, + { + "epoch": 0.0849392840246018, + "grad_norm": 4.141830921173096, + "learning_rate": 1.9842438722099757e-05, + "loss": 0.6377, + "step": 2693 + }, + { + "epoch": 0.08497082479104243, + "grad_norm": 11.86594295501709, + "learning_rate": 1.9842258042917025e-05, + "loss": 0.3962, + "step": 2694 + }, + { + "epoch": 0.08500236555748304, + "grad_norm": 4.692668437957764, + "learning_rate": 1.984207726102269e-05, + "loss": 0.9407, + "step": 2695 + }, + { + "epoch": 0.08503390632392367, + "grad_norm": 7.687566757202148, + "learning_rate": 1.9841896376418643e-05, + "loss": 0.6628, + "step": 2696 + }, + { + "epoch": 0.0850654470903643, + "grad_norm": 5.379232406616211, + "learning_rate": 1.9841715389106757e-05, + "loss": 0.6868, + "step": 2697 + }, + { + "epoch": 0.08509698785680492, + "grad_norm": 11.12779426574707, + "learning_rate": 1.9841534299088932e-05, + "loss": 0.5956, + "step": 2698 + }, + { + "epoch": 0.08512852862324555, + "grad_norm": 6.129316329956055, + "learning_rate": 1.9841353106367058e-05, + "loss": 0.8883, + "step": 2699 + }, + { + "epoch": 0.08516006938968616, + "grad_norm": 7.715219974517822, + "learning_rate": 1.9841171810943018e-05, + "loss": 0.6814, + "step": 2700 + }, + { + "epoch": 0.08519161015612679, + "grad_norm": 4.636089324951172, + "learning_rate": 1.984099041281871e-05, + "loss": 0.4081, + "step": 2701 + }, + { + "epoch": 0.08522315092256742, + "grad_norm": 6.552263259887695, + "learning_rate": 1.9840808911996024e-05, + "loss": 0.6194, + "step": 2702 + }, + { + "epoch": 0.08525469168900804, + "grad_norm": 5.319258213043213, + "learning_rate": 1.9840627308476858e-05, + "loss": 0.6536, + "step": 2703 + }, + { + "epoch": 0.08528623245544867, + "grad_norm": 10.853256225585938, + "learning_rate": 1.9840445602263103e-05, + "loss": 0.4128, + "step": 2704 + }, + { + "epoch": 0.0853177732218893, + "grad_norm": 4.678876876831055, + "learning_rate": 1.984026379335666e-05, + "loss": 0.6269, + "step": 2705 + }, + { + "epoch": 0.08534931398832991, + "grad_norm": 5.917720794677734, + "learning_rate": 1.9840081881759422e-05, + "loss": 0.6119, + "step": 2706 + }, + { + "epoch": 0.08538085475477054, + "grad_norm": 3.896698474884033, + "learning_rate": 1.983989986747329e-05, + "loss": 0.3998, + "step": 2707 + }, + { + "epoch": 0.08541239552121116, + "grad_norm": 5.319193363189697, + "learning_rate": 1.983971775050016e-05, + "loss": 0.647, + "step": 2708 + }, + { + "epoch": 0.08544393628765179, + "grad_norm": 5.7632575035095215, + "learning_rate": 1.983953553084194e-05, + "loss": 0.6405, + "step": 2709 + }, + { + "epoch": 0.08547547705409242, + "grad_norm": 17.51881217956543, + "learning_rate": 1.983935320850052e-05, + "loss": 0.3559, + "step": 2710 + }, + { + "epoch": 0.08550701782053304, + "grad_norm": 8.853241920471191, + "learning_rate": 1.983917078347781e-05, + "loss": 0.6226, + "step": 2711 + }, + { + "epoch": 0.08553855858697366, + "grad_norm": 5.27386999130249, + "learning_rate": 1.9838988255775718e-05, + "loss": 0.644, + "step": 2712 + }, + { + "epoch": 0.0855700993534143, + "grad_norm": 4.144643306732178, + "learning_rate": 1.9838805625396142e-05, + "loss": 0.5888, + "step": 2713 + }, + { + "epoch": 0.08560164011985491, + "grad_norm": 4.521880626678467, + "learning_rate": 1.9838622892340987e-05, + "loss": 0.6421, + "step": 2714 + }, + { + "epoch": 0.08563318088629554, + "grad_norm": 5.422712326049805, + "learning_rate": 1.9838440056612164e-05, + "loss": 0.6222, + "step": 2715 + }, + { + "epoch": 0.08566472165273616, + "grad_norm": 4.96162223815918, + "learning_rate": 1.983825711821158e-05, + "loss": 0.6515, + "step": 2716 + }, + { + "epoch": 0.08569626241917679, + "grad_norm": 4.481231212615967, + "learning_rate": 1.9838074077141145e-05, + "loss": 0.4077, + "step": 2717 + }, + { + "epoch": 0.08572780318561742, + "grad_norm": 11.205698013305664, + "learning_rate": 1.9837890933402767e-05, + "loss": 0.5973, + "step": 2718 + }, + { + "epoch": 0.08575934395205803, + "grad_norm": 3.9921233654022217, + "learning_rate": 1.9837707686998358e-05, + "loss": 0.6161, + "step": 2719 + }, + { + "epoch": 0.08579088471849866, + "grad_norm": 6.227669715881348, + "learning_rate": 1.983752433792983e-05, + "loss": 0.6174, + "step": 2720 + }, + { + "epoch": 0.08582242548493929, + "grad_norm": 13.068532943725586, + "learning_rate": 1.9837340886199097e-05, + "loss": 0.6189, + "step": 2721 + }, + { + "epoch": 0.0858539662513799, + "grad_norm": 7.05474328994751, + "learning_rate": 1.9837157331808074e-05, + "loss": 0.6199, + "step": 2722 + }, + { + "epoch": 0.08588550701782054, + "grad_norm": 12.239441871643066, + "learning_rate": 1.9836973674758676e-05, + "loss": 0.6019, + "step": 2723 + }, + { + "epoch": 0.08591704778426115, + "grad_norm": 5.366315841674805, + "learning_rate": 1.983678991505282e-05, + "loss": 0.6648, + "step": 2724 + }, + { + "epoch": 0.08594858855070178, + "grad_norm": 4.731375694274902, + "learning_rate": 1.983660605269242e-05, + "loss": 0.9219, + "step": 2725 + }, + { + "epoch": 0.08598012931714241, + "grad_norm": 6.115241050720215, + "learning_rate": 1.98364220876794e-05, + "loss": 0.6463, + "step": 2726 + }, + { + "epoch": 0.08601167008358303, + "grad_norm": 13.019572257995605, + "learning_rate": 1.9836238020015675e-05, + "loss": 0.6525, + "step": 2727 + }, + { + "epoch": 0.08604321085002366, + "grad_norm": 5.083545207977295, + "learning_rate": 1.9836053849703174e-05, + "loss": 0.6012, + "step": 2728 + }, + { + "epoch": 0.08607475161646427, + "grad_norm": 5.134316444396973, + "learning_rate": 1.983586957674381e-05, + "loss": 0.6391, + "step": 2729 + }, + { + "epoch": 0.0861062923829049, + "grad_norm": 6.281414031982422, + "learning_rate": 1.9835685201139513e-05, + "loss": 0.6218, + "step": 2730 + }, + { + "epoch": 0.08613783314934553, + "grad_norm": 5.7819061279296875, + "learning_rate": 1.9835500722892202e-05, + "loss": 0.5952, + "step": 2731 + }, + { + "epoch": 0.08616937391578615, + "grad_norm": 5.94235372543335, + "learning_rate": 1.9835316142003803e-05, + "loss": 0.614, + "step": 2732 + }, + { + "epoch": 0.08620091468222678, + "grad_norm": 5.035347938537598, + "learning_rate": 1.983513145847624e-05, + "loss": 0.5992, + "step": 2733 + }, + { + "epoch": 0.08623245544866741, + "grad_norm": 7.407480716705322, + "learning_rate": 1.983494667231145e-05, + "loss": 0.8946, + "step": 2734 + }, + { + "epoch": 0.08626399621510802, + "grad_norm": 6.537321090698242, + "learning_rate": 1.983476178351135e-05, + "loss": 0.6231, + "step": 2735 + }, + { + "epoch": 0.08629553698154865, + "grad_norm": 7.094677448272705, + "learning_rate": 1.9834576792077876e-05, + "loss": 0.671, + "step": 2736 + }, + { + "epoch": 0.08632707774798927, + "grad_norm": 6.475344181060791, + "learning_rate": 1.9834391698012957e-05, + "loss": 0.6401, + "step": 2737 + }, + { + "epoch": 0.0863586185144299, + "grad_norm": 6.241939544677734, + "learning_rate": 1.9834206501318524e-05, + "loss": 0.6239, + "step": 2738 + }, + { + "epoch": 0.08639015928087053, + "grad_norm": 6.945199489593506, + "learning_rate": 1.9834021201996512e-05, + "loss": 0.6385, + "step": 2739 + }, + { + "epoch": 0.08642170004731115, + "grad_norm": 5.2430572509765625, + "learning_rate": 1.9833835800048852e-05, + "loss": 0.6434, + "step": 2740 + }, + { + "epoch": 0.08645324081375177, + "grad_norm": 4.698238372802734, + "learning_rate": 1.9833650295477475e-05, + "loss": 0.6116, + "step": 2741 + }, + { + "epoch": 0.0864847815801924, + "grad_norm": 4.662047386169434, + "learning_rate": 1.9833464688284327e-05, + "loss": 0.607, + "step": 2742 + }, + { + "epoch": 0.08651632234663302, + "grad_norm": 6.789970874786377, + "learning_rate": 1.9833278978471338e-05, + "loss": 0.4002, + "step": 2743 + }, + { + "epoch": 0.08654786311307365, + "grad_norm": 4.922245502471924, + "learning_rate": 1.9833093166040446e-05, + "loss": 0.5932, + "step": 2744 + }, + { + "epoch": 0.08657940387951427, + "grad_norm": 4.487638473510742, + "learning_rate": 1.9832907250993593e-05, + "loss": 0.6169, + "step": 2745 + }, + { + "epoch": 0.0866109446459549, + "grad_norm": 4.07331657409668, + "learning_rate": 1.983272123333272e-05, + "loss": 0.622, + "step": 2746 + }, + { + "epoch": 0.08664248541239553, + "grad_norm": 3.502927780151367, + "learning_rate": 1.983253511305976e-05, + "loss": 0.9408, + "step": 2747 + }, + { + "epoch": 0.08667402617883614, + "grad_norm": 5.695928573608398, + "learning_rate": 1.9832348890176666e-05, + "loss": 0.6391, + "step": 2748 + }, + { + "epoch": 0.08670556694527677, + "grad_norm": 11.399571418762207, + "learning_rate": 1.983216256468538e-05, + "loss": 0.6224, + "step": 2749 + }, + { + "epoch": 0.0867371077117174, + "grad_norm": 7.194329738616943, + "learning_rate": 1.9831976136587837e-05, + "loss": 0.6529, + "step": 2750 + }, + { + "epoch": 0.08676864847815802, + "grad_norm": 5.129217147827148, + "learning_rate": 1.983178960588599e-05, + "loss": 0.6494, + "step": 2751 + }, + { + "epoch": 0.08680018924459865, + "grad_norm": 4.643807888031006, + "learning_rate": 1.9831602972581784e-05, + "loss": 0.6364, + "step": 2752 + }, + { + "epoch": 0.08683173001103926, + "grad_norm": 9.129485130310059, + "learning_rate": 1.983141623667717e-05, + "loss": 0.6322, + "step": 2753 + }, + { + "epoch": 0.08686327077747989, + "grad_norm": 4.440095901489258, + "learning_rate": 1.983122939817409e-05, + "loss": 0.5975, + "step": 2754 + }, + { + "epoch": 0.08689481154392052, + "grad_norm": 4.983465194702148, + "learning_rate": 1.9831042457074498e-05, + "loss": 0.5827, + "step": 2755 + }, + { + "epoch": 0.08692635231036114, + "grad_norm": 4.540619373321533, + "learning_rate": 1.9830855413380346e-05, + "loss": 0.5928, + "step": 2756 + }, + { + "epoch": 0.08695789307680177, + "grad_norm": 6.135812282562256, + "learning_rate": 1.9830668267093583e-05, + "loss": 0.6094, + "step": 2757 + }, + { + "epoch": 0.0869894338432424, + "grad_norm": 3.7528929710388184, + "learning_rate": 1.9830481018216162e-05, + "loss": 0.9109, + "step": 2758 + }, + { + "epoch": 0.08702097460968301, + "grad_norm": 3.6673789024353027, + "learning_rate": 1.9830293666750045e-05, + "loss": 0.607, + "step": 2759 + }, + { + "epoch": 0.08705251537612364, + "grad_norm": 4.025293827056885, + "learning_rate": 1.9830106212697174e-05, + "loss": 0.6168, + "step": 2760 + }, + { + "epoch": 0.08708405614256426, + "grad_norm": 5.758869647979736, + "learning_rate": 1.9829918656059514e-05, + "loss": 0.6512, + "step": 2761 + }, + { + "epoch": 0.08711559690900489, + "grad_norm": 20.66312026977539, + "learning_rate": 1.982973099683902e-05, + "loss": 0.6203, + "step": 2762 + }, + { + "epoch": 0.08714713767544552, + "grad_norm": 4.57422399520874, + "learning_rate": 1.9829543235037648e-05, + "loss": 0.6584, + "step": 2763 + }, + { + "epoch": 0.08717867844188613, + "grad_norm": 4.657708644866943, + "learning_rate": 1.9829355370657363e-05, + "loss": 0.611, + "step": 2764 + }, + { + "epoch": 0.08721021920832676, + "grad_norm": 5.666173934936523, + "learning_rate": 1.9829167403700124e-05, + "loss": 0.5983, + "step": 2765 + }, + { + "epoch": 0.08724175997476738, + "grad_norm": 5.232918739318848, + "learning_rate": 1.982897933416789e-05, + "loss": 0.4008, + "step": 2766 + }, + { + "epoch": 0.08727330074120801, + "grad_norm": 4.993886470794678, + "learning_rate": 1.982879116206262e-05, + "loss": 0.6282, + "step": 2767 + }, + { + "epoch": 0.08730484150764864, + "grad_norm": 4.583425521850586, + "learning_rate": 1.9828602887386285e-05, + "loss": 0.6008, + "step": 2768 + }, + { + "epoch": 0.08733638227408926, + "grad_norm": 3.5670690536499023, + "learning_rate": 1.982841451014085e-05, + "loss": 0.6005, + "step": 2769 + }, + { + "epoch": 0.08736792304052989, + "grad_norm": 3.674342632293701, + "learning_rate": 1.9828226030328274e-05, + "loss": 0.6084, + "step": 2770 + }, + { + "epoch": 0.08739946380697051, + "grad_norm": 4.71624755859375, + "learning_rate": 1.982803744795053e-05, + "loss": 0.6243, + "step": 2771 + }, + { + "epoch": 0.08743100457341113, + "grad_norm": 6.956430435180664, + "learning_rate": 1.9827848763009584e-05, + "loss": 0.5878, + "step": 2772 + }, + { + "epoch": 0.08746254533985176, + "grad_norm": 5.298355579376221, + "learning_rate": 1.9827659975507406e-05, + "loss": 0.5946, + "step": 2773 + }, + { + "epoch": 0.08749408610629238, + "grad_norm": 5.122622489929199, + "learning_rate": 1.9827471085445963e-05, + "loss": 0.5901, + "step": 2774 + }, + { + "epoch": 0.087525626872733, + "grad_norm": 2.622389793395996, + "learning_rate": 1.9827282092827228e-05, + "loss": 0.3937, + "step": 2775 + }, + { + "epoch": 0.08755716763917364, + "grad_norm": 6.667938709259033, + "learning_rate": 1.9827092997653176e-05, + "loss": 0.6317, + "step": 2776 + }, + { + "epoch": 0.08758870840561425, + "grad_norm": 4.682895660400391, + "learning_rate": 1.9826903799925775e-05, + "loss": 0.6198, + "step": 2777 + }, + { + "epoch": 0.08762024917205488, + "grad_norm": 4.662938594818115, + "learning_rate": 1.9826714499647005e-05, + "loss": 0.6205, + "step": 2778 + }, + { + "epoch": 0.08765178993849551, + "grad_norm": 6.701237201690674, + "learning_rate": 1.982652509681884e-05, + "loss": 0.6407, + "step": 2779 + }, + { + "epoch": 0.08768333070493613, + "grad_norm": 6.0321173667907715, + "learning_rate": 1.9826335591443252e-05, + "loss": 0.5975, + "step": 2780 + }, + { + "epoch": 0.08771487147137676, + "grad_norm": 3.5273687839508057, + "learning_rate": 1.9826145983522223e-05, + "loss": 0.9259, + "step": 2781 + }, + { + "epoch": 0.08774641223781737, + "grad_norm": 5.1566481590271, + "learning_rate": 1.9825956273057733e-05, + "loss": 0.6226, + "step": 2782 + }, + { + "epoch": 0.087777953004258, + "grad_norm": 5.325843334197998, + "learning_rate": 1.982576646005176e-05, + "loss": 0.6106, + "step": 2783 + }, + { + "epoch": 0.08780949377069863, + "grad_norm": 5.406348705291748, + "learning_rate": 1.9825576544506283e-05, + "loss": 0.6158, + "step": 2784 + }, + { + "epoch": 0.08784103453713925, + "grad_norm": 4.973852634429932, + "learning_rate": 1.9825386526423282e-05, + "loss": 0.6397, + "step": 2785 + }, + { + "epoch": 0.08787257530357988, + "grad_norm": 5.603193283081055, + "learning_rate": 1.982519640580475e-05, + "loss": 0.4012, + "step": 2786 + }, + { + "epoch": 0.08790411607002051, + "grad_norm": 9.933566093444824, + "learning_rate": 1.982500618265266e-05, + "loss": 0.5974, + "step": 2787 + }, + { + "epoch": 0.08793565683646112, + "grad_norm": 5.0820207595825195, + "learning_rate": 1.9824815856969004e-05, + "loss": 0.6352, + "step": 2788 + }, + { + "epoch": 0.08796719760290175, + "grad_norm": 4.816000938415527, + "learning_rate": 1.982462542875576e-05, + "loss": 0.6009, + "step": 2789 + }, + { + "epoch": 0.08799873836934237, + "grad_norm": 8.701800346374512, + "learning_rate": 1.9824434898014928e-05, + "loss": 0.6316, + "step": 2790 + }, + { + "epoch": 0.088030279135783, + "grad_norm": 7.563843727111816, + "learning_rate": 1.982424426474849e-05, + "loss": 0.6005, + "step": 2791 + }, + { + "epoch": 0.08806181990222363, + "grad_norm": 5.747769355773926, + "learning_rate": 1.982405352895843e-05, + "loss": 0.6439, + "step": 2792 + }, + { + "epoch": 0.08809336066866424, + "grad_norm": 7.125238418579102, + "learning_rate": 1.9823862690646746e-05, + "loss": 0.607, + "step": 2793 + }, + { + "epoch": 0.08812490143510487, + "grad_norm": 7.513365745544434, + "learning_rate": 1.9823671749815424e-05, + "loss": 0.5949, + "step": 2794 + }, + { + "epoch": 0.0881564422015455, + "grad_norm": 4.36492919921875, + "learning_rate": 1.982348070646646e-05, + "loss": 0.6096, + "step": 2795 + }, + { + "epoch": 0.08818798296798612, + "grad_norm": 5.431971073150635, + "learning_rate": 1.9823289560601848e-05, + "loss": 0.5968, + "step": 2796 + }, + { + "epoch": 0.08821952373442675, + "grad_norm": 5.4707512855529785, + "learning_rate": 1.9823098312223583e-05, + "loss": 0.6418, + "step": 2797 + }, + { + "epoch": 0.08825106450086737, + "grad_norm": 7.128641605377197, + "learning_rate": 1.982290696133366e-05, + "loss": 0.6426, + "step": 2798 + }, + { + "epoch": 0.088282605267308, + "grad_norm": 5.250598907470703, + "learning_rate": 1.9822715507934074e-05, + "loss": 0.6365, + "step": 2799 + }, + { + "epoch": 0.08831414603374863, + "grad_norm": 4.5572733879089355, + "learning_rate": 1.9822523952026823e-05, + "loss": 0.6422, + "step": 2800 + }, + { + "epoch": 0.08834568680018924, + "grad_norm": 6.901379585266113, + "learning_rate": 1.982233229361391e-05, + "loss": 0.6149, + "step": 2801 + }, + { + "epoch": 0.08837722756662987, + "grad_norm": 8.608415603637695, + "learning_rate": 1.982214053269733e-05, + "loss": 0.6002, + "step": 2802 + }, + { + "epoch": 0.0884087683330705, + "grad_norm": 5.240907192230225, + "learning_rate": 1.982194866927909e-05, + "loss": 0.6021, + "step": 2803 + }, + { + "epoch": 0.08844030909951112, + "grad_norm": 11.370026588439941, + "learning_rate": 1.982175670336119e-05, + "loss": 0.6323, + "step": 2804 + }, + { + "epoch": 0.08847184986595175, + "grad_norm": 5.57560396194458, + "learning_rate": 1.982156463494563e-05, + "loss": 0.6235, + "step": 2805 + }, + { + "epoch": 0.08850339063239236, + "grad_norm": 4.402955055236816, + "learning_rate": 1.9821372464034416e-05, + "loss": 0.648, + "step": 2806 + }, + { + "epoch": 0.08853493139883299, + "grad_norm": 4.420718669891357, + "learning_rate": 1.9821180190629557e-05, + "loss": 0.5932, + "step": 2807 + }, + { + "epoch": 0.08856647216527362, + "grad_norm": 5.613361358642578, + "learning_rate": 1.9820987814733053e-05, + "loss": 0.6068, + "step": 2808 + }, + { + "epoch": 0.08859801293171424, + "grad_norm": 4.358880519866943, + "learning_rate": 1.982079533634692e-05, + "loss": 0.6048, + "step": 2809 + }, + { + "epoch": 0.08862955369815487, + "grad_norm": 7.931293964385986, + "learning_rate": 1.9820602755473158e-05, + "loss": 0.6057, + "step": 2810 + }, + { + "epoch": 0.08866109446459548, + "grad_norm": 5.118824481964111, + "learning_rate": 1.982041007211378e-05, + "loss": 0.5938, + "step": 2811 + }, + { + "epoch": 0.08869263523103611, + "grad_norm": 12.920939445495605, + "learning_rate": 1.98202172862708e-05, + "loss": 0.6345, + "step": 2812 + }, + { + "epoch": 0.08872417599747674, + "grad_norm": 7.038008213043213, + "learning_rate": 1.9820024397946232e-05, + "loss": 0.6026, + "step": 2813 + }, + { + "epoch": 0.08875571676391736, + "grad_norm": 4.596909046173096, + "learning_rate": 1.9819831407142078e-05, + "loss": 0.6187, + "step": 2814 + }, + { + "epoch": 0.08878725753035799, + "grad_norm": 4.678278923034668, + "learning_rate": 1.981963831386036e-05, + "loss": 0.6388, + "step": 2815 + }, + { + "epoch": 0.08881879829679862, + "grad_norm": 3.9946367740631104, + "learning_rate": 1.981944511810309e-05, + "loss": 0.6254, + "step": 2816 + }, + { + "epoch": 0.08885033906323923, + "grad_norm": 3.7445104122161865, + "learning_rate": 1.9819251819872288e-05, + "loss": 0.5809, + "step": 2817 + }, + { + "epoch": 0.08888187982967986, + "grad_norm": 4.030223369598389, + "learning_rate": 1.9819058419169968e-05, + "loss": 0.6402, + "step": 2818 + }, + { + "epoch": 0.08891342059612048, + "grad_norm": 10.467087745666504, + "learning_rate": 1.981886491599815e-05, + "loss": 0.6659, + "step": 2819 + }, + { + "epoch": 0.08894496136256111, + "grad_norm": 4.24562931060791, + "learning_rate": 1.981867131035885e-05, + "loss": 0.624, + "step": 2820 + }, + { + "epoch": 0.08897650212900174, + "grad_norm": 6.376105308532715, + "learning_rate": 1.981847760225409e-05, + "loss": 0.4019, + "step": 2821 + }, + { + "epoch": 0.08900804289544235, + "grad_norm": 5.608319282531738, + "learning_rate": 1.9818283791685897e-05, + "loss": 0.6035, + "step": 2822 + }, + { + "epoch": 0.08903958366188298, + "grad_norm": 3.8549089431762695, + "learning_rate": 1.9818089878656284e-05, + "loss": 0.6617, + "step": 2823 + }, + { + "epoch": 0.08907112442832361, + "grad_norm": 3.95717716217041, + "learning_rate": 1.9817895863167283e-05, + "loss": 0.9104, + "step": 2824 + }, + { + "epoch": 0.08910266519476423, + "grad_norm": 3.9558675289154053, + "learning_rate": 1.9817701745220916e-05, + "loss": 0.6291, + "step": 2825 + }, + { + "epoch": 0.08913420596120486, + "grad_norm": 4.113525867462158, + "learning_rate": 1.9817507524819206e-05, + "loss": 0.6267, + "step": 2826 + }, + { + "epoch": 0.08916574672764548, + "grad_norm": 7.389801979064941, + "learning_rate": 1.9817313201964184e-05, + "loss": 0.6347, + "step": 2827 + }, + { + "epoch": 0.0891972874940861, + "grad_norm": 4.345695972442627, + "learning_rate": 1.9817118776657872e-05, + "loss": 0.6135, + "step": 2828 + }, + { + "epoch": 0.08922882826052674, + "grad_norm": 3.2567763328552246, + "learning_rate": 1.9816924248902304e-05, + "loss": 0.616, + "step": 2829 + }, + { + "epoch": 0.08926036902696735, + "grad_norm": 4.182676315307617, + "learning_rate": 1.981672961869951e-05, + "loss": 0.6285, + "step": 2830 + }, + { + "epoch": 0.08929190979340798, + "grad_norm": 4.387113571166992, + "learning_rate": 1.9816534886051522e-05, + "loss": 0.6525, + "step": 2831 + }, + { + "epoch": 0.08932345055984861, + "grad_norm": 5.261441230773926, + "learning_rate": 1.9816340050960367e-05, + "loss": 0.5899, + "step": 2832 + }, + { + "epoch": 0.08935499132628923, + "grad_norm": 4.066127300262451, + "learning_rate": 1.9816145113428084e-05, + "loss": 0.6291, + "step": 2833 + }, + { + "epoch": 0.08938653209272986, + "grad_norm": 5.393555641174316, + "learning_rate": 1.98159500734567e-05, + "loss": 0.6553, + "step": 2834 + }, + { + "epoch": 0.08941807285917047, + "grad_norm": 7.524610996246338, + "learning_rate": 1.981575493104826e-05, + "loss": 0.6081, + "step": 2835 + }, + { + "epoch": 0.0894496136256111, + "grad_norm": 5.070773601531982, + "learning_rate": 1.9815559686204794e-05, + "loss": 0.6236, + "step": 2836 + }, + { + "epoch": 0.08948115439205173, + "grad_norm": 3.3709402084350586, + "learning_rate": 1.9815364338928336e-05, + "loss": 0.5803, + "step": 2837 + }, + { + "epoch": 0.08951269515849235, + "grad_norm": 2.523773670196533, + "learning_rate": 1.9815168889220936e-05, + "loss": 0.4455, + "step": 2838 + }, + { + "epoch": 0.08954423592493298, + "grad_norm": 3.2430410385131836, + "learning_rate": 1.9814973337084627e-05, + "loss": 0.6274, + "step": 2839 + }, + { + "epoch": 0.08957577669137361, + "grad_norm": 3.4077043533325195, + "learning_rate": 1.9814777682521445e-05, + "loss": 0.6542, + "step": 2840 + }, + { + "epoch": 0.08960731745781422, + "grad_norm": 4.021816253662109, + "learning_rate": 1.9814581925533446e-05, + "loss": 0.6271, + "step": 2841 + }, + { + "epoch": 0.08963885822425485, + "grad_norm": 3.638151168823242, + "learning_rate": 1.9814386066122658e-05, + "loss": 0.6188, + "step": 2842 + }, + { + "epoch": 0.08967039899069547, + "grad_norm": 6.816516399383545, + "learning_rate": 1.9814190104291128e-05, + "loss": 0.6011, + "step": 2843 + }, + { + "epoch": 0.0897019397571361, + "grad_norm": 3.0104317665100098, + "learning_rate": 1.9813994040040908e-05, + "loss": 0.6391, + "step": 2844 + }, + { + "epoch": 0.08973348052357673, + "grad_norm": 5.510369777679443, + "learning_rate": 1.9813797873374037e-05, + "loss": 0.6273, + "step": 2845 + }, + { + "epoch": 0.08976502129001734, + "grad_norm": 3.7674286365509033, + "learning_rate": 1.981360160429257e-05, + "loss": 0.6301, + "step": 2846 + }, + { + "epoch": 0.08979656205645797, + "grad_norm": 4.758206367492676, + "learning_rate": 1.9813405232798545e-05, + "loss": 0.6368, + "step": 2847 + }, + { + "epoch": 0.0898281028228986, + "grad_norm": 3.187709093093872, + "learning_rate": 1.981320875889402e-05, + "loss": 0.3796, + "step": 2848 + }, + { + "epoch": 0.08985964358933922, + "grad_norm": 5.890049457550049, + "learning_rate": 1.981301218258104e-05, + "loss": 0.651, + "step": 2849 + }, + { + "epoch": 0.08989118435577985, + "grad_norm": 4.595412731170654, + "learning_rate": 1.9812815503861657e-05, + "loss": 0.9224, + "step": 2850 + }, + { + "epoch": 0.08992272512222046, + "grad_norm": 3.5160279273986816, + "learning_rate": 1.9812618722737926e-05, + "loss": 0.6516, + "step": 2851 + }, + { + "epoch": 0.0899542658886611, + "grad_norm": 4.24678373336792, + "learning_rate": 1.98124218392119e-05, + "loss": 0.624, + "step": 2852 + }, + { + "epoch": 0.08998580665510172, + "grad_norm": 3.7647719383239746, + "learning_rate": 1.981222485328563e-05, + "loss": 0.6309, + "step": 2853 + }, + { + "epoch": 0.09001734742154234, + "grad_norm": 4.411267280578613, + "learning_rate": 1.981202776496118e-05, + "loss": 0.6209, + "step": 2854 + }, + { + "epoch": 0.09004888818798297, + "grad_norm": 5.866828918457031, + "learning_rate": 1.9811830574240597e-05, + "loss": 0.6249, + "step": 2855 + }, + { + "epoch": 0.09008042895442359, + "grad_norm": 2.176807403564453, + "learning_rate": 1.9811633281125945e-05, + "loss": 0.42, + "step": 2856 + }, + { + "epoch": 0.09011196972086422, + "grad_norm": 3.0352044105529785, + "learning_rate": 1.981143588561928e-05, + "loss": 0.5677, + "step": 2857 + }, + { + "epoch": 0.09014351048730485, + "grad_norm": 3.3510618209838867, + "learning_rate": 1.9811238387722665e-05, + "loss": 0.5932, + "step": 2858 + }, + { + "epoch": 0.09017505125374546, + "grad_norm": 3.7720935344696045, + "learning_rate": 1.981104078743816e-05, + "loss": 0.6277, + "step": 2859 + }, + { + "epoch": 0.09020659202018609, + "grad_norm": 2.954697847366333, + "learning_rate": 1.9810843084767823e-05, + "loss": 0.6342, + "step": 2860 + }, + { + "epoch": 0.09023813278662672, + "grad_norm": 5.340746879577637, + "learning_rate": 1.981064527971372e-05, + "loss": 0.4287, + "step": 2861 + }, + { + "epoch": 0.09026967355306734, + "grad_norm": 3.4431324005126953, + "learning_rate": 1.981044737227792e-05, + "loss": 0.6452, + "step": 2862 + }, + { + "epoch": 0.09030121431950797, + "grad_norm": 3.5904855728149414, + "learning_rate": 1.9810249362462484e-05, + "loss": 0.5918, + "step": 2863 + }, + { + "epoch": 0.09033275508594858, + "grad_norm": 4.57063627243042, + "learning_rate": 1.9810051250269475e-05, + "loss": 0.8795, + "step": 2864 + }, + { + "epoch": 0.09036429585238921, + "grad_norm": 5.019216060638428, + "learning_rate": 1.9809853035700964e-05, + "loss": 0.6262, + "step": 2865 + }, + { + "epoch": 0.09039583661882984, + "grad_norm": 2.2860727310180664, + "learning_rate": 1.980965471875902e-05, + "loss": 0.9335, + "step": 2866 + }, + { + "epoch": 0.09042737738527046, + "grad_norm": 3.175788164138794, + "learning_rate": 1.9809456299445714e-05, + "loss": 0.6402, + "step": 2867 + }, + { + "epoch": 0.09045891815171109, + "grad_norm": 3.233565092086792, + "learning_rate": 1.9809257777763114e-05, + "loss": 0.6047, + "step": 2868 + }, + { + "epoch": 0.09049045891815172, + "grad_norm": 13.474696159362793, + "learning_rate": 1.980905915371329e-05, + "loss": 0.6221, + "step": 2869 + }, + { + "epoch": 0.09052199968459233, + "grad_norm": 3.83992862701416, + "learning_rate": 1.9808860427298323e-05, + "loss": 0.6182, + "step": 2870 + }, + { + "epoch": 0.09055354045103296, + "grad_norm": 3.291292190551758, + "learning_rate": 1.9808661598520276e-05, + "loss": 0.6123, + "step": 2871 + }, + { + "epoch": 0.09058508121747358, + "grad_norm": 2.888312816619873, + "learning_rate": 1.980846266738123e-05, + "loss": 0.6239, + "step": 2872 + }, + { + "epoch": 0.09061662198391421, + "grad_norm": 3.054774284362793, + "learning_rate": 1.980826363388326e-05, + "loss": 0.6148, + "step": 2873 + }, + { + "epoch": 0.09064816275035484, + "grad_norm": 10.011621475219727, + "learning_rate": 1.9808064498028443e-05, + "loss": 0.6322, + "step": 2874 + }, + { + "epoch": 0.09067970351679545, + "grad_norm": 3.278913736343384, + "learning_rate": 1.9807865259818854e-05, + "loss": 0.6178, + "step": 2875 + }, + { + "epoch": 0.09071124428323608, + "grad_norm": 3.511819362640381, + "learning_rate": 1.980766591925658e-05, + "loss": 0.6035, + "step": 2876 + }, + { + "epoch": 0.09074278504967671, + "grad_norm": 3.6001484394073486, + "learning_rate": 1.9807466476343696e-05, + "loss": 0.6038, + "step": 2877 + }, + { + "epoch": 0.09077432581611733, + "grad_norm": 4.100569725036621, + "learning_rate": 1.980726693108228e-05, + "loss": 0.625, + "step": 2878 + }, + { + "epoch": 0.09080586658255796, + "grad_norm": 4.197580814361572, + "learning_rate": 1.9807067283474424e-05, + "loss": 0.9057, + "step": 2879 + }, + { + "epoch": 0.09083740734899858, + "grad_norm": 4.363424777984619, + "learning_rate": 1.9806867533522206e-05, + "loss": 0.6016, + "step": 2880 + }, + { + "epoch": 0.0908689481154392, + "grad_norm": 3.880742311477661, + "learning_rate": 1.9806667681227706e-05, + "loss": 0.6378, + "step": 2881 + }, + { + "epoch": 0.09090048888187983, + "grad_norm": 2.7453017234802246, + "learning_rate": 1.9806467726593018e-05, + "loss": 0.3828, + "step": 2882 + }, + { + "epoch": 0.09093202964832045, + "grad_norm": 3.3762221336364746, + "learning_rate": 1.980626766962022e-05, + "loss": 0.6255, + "step": 2883 + }, + { + "epoch": 0.09096357041476108, + "grad_norm": 4.349318981170654, + "learning_rate": 1.9806067510311408e-05, + "loss": 0.6151, + "step": 2884 + }, + { + "epoch": 0.09099511118120171, + "grad_norm": 3.8629627227783203, + "learning_rate": 1.980586724866867e-05, + "loss": 0.6135, + "step": 2885 + }, + { + "epoch": 0.09102665194764233, + "grad_norm": 4.077963352203369, + "learning_rate": 1.9805666884694088e-05, + "loss": 0.5819, + "step": 2886 + }, + { + "epoch": 0.09105819271408296, + "grad_norm": 2.838982343673706, + "learning_rate": 1.980546641838976e-05, + "loss": 0.6109, + "step": 2887 + }, + { + "epoch": 0.09108973348052357, + "grad_norm": 3.8438425064086914, + "learning_rate": 1.9805265849757777e-05, + "loss": 0.6686, + "step": 2888 + }, + { + "epoch": 0.0911212742469642, + "grad_norm": 3.3849596977233887, + "learning_rate": 1.980506517880023e-05, + "loss": 0.6015, + "step": 2889 + }, + { + "epoch": 0.09115281501340483, + "grad_norm": 2.875112533569336, + "learning_rate": 1.9804864405519215e-05, + "loss": 0.5479, + "step": 2890 + }, + { + "epoch": 0.09118435577984545, + "grad_norm": 3.167452096939087, + "learning_rate": 1.9804663529916825e-05, + "loss": 0.611, + "step": 2891 + }, + { + "epoch": 0.09121589654628608, + "grad_norm": 2.8689165115356445, + "learning_rate": 1.9804462551995162e-05, + "loss": 0.3875, + "step": 2892 + }, + { + "epoch": 0.09124743731272669, + "grad_norm": 8.040181159973145, + "learning_rate": 1.9804261471756317e-05, + "loss": 0.64, + "step": 2893 + }, + { + "epoch": 0.09127897807916732, + "grad_norm": 4.31552791595459, + "learning_rate": 1.980406028920239e-05, + "loss": 0.9655, + "step": 2894 + }, + { + "epoch": 0.09131051884560795, + "grad_norm": 5.196504592895508, + "learning_rate": 1.980385900433548e-05, + "loss": 0.6375, + "step": 2895 + }, + { + "epoch": 0.09134205961204857, + "grad_norm": 4.264398574829102, + "learning_rate": 1.9803657617157693e-05, + "loss": 0.6369, + "step": 2896 + }, + { + "epoch": 0.0913736003784892, + "grad_norm": 3.9909870624542236, + "learning_rate": 1.9803456127671122e-05, + "loss": 0.6154, + "step": 2897 + }, + { + "epoch": 0.09140514114492983, + "grad_norm": 3.251533031463623, + "learning_rate": 1.9803254535877876e-05, + "loss": 0.6033, + "step": 2898 + }, + { + "epoch": 0.09143668191137044, + "grad_norm": 4.493241786956787, + "learning_rate": 1.9803052841780058e-05, + "loss": 0.6342, + "step": 2899 + }, + { + "epoch": 0.09146822267781107, + "grad_norm": 5.022342205047607, + "learning_rate": 1.9802851045379772e-05, + "loss": 0.9075, + "step": 2900 + }, + { + "epoch": 0.09149976344425169, + "grad_norm": 3.321423053741455, + "learning_rate": 1.9802649146679123e-05, + "loss": 0.8855, + "step": 2901 + }, + { + "epoch": 0.09153130421069232, + "grad_norm": 3.0737056732177734, + "learning_rate": 1.9802447145680218e-05, + "loss": 0.6293, + "step": 2902 + }, + { + "epoch": 0.09156284497713295, + "grad_norm": 3.43882155418396, + "learning_rate": 1.9802245042385164e-05, + "loss": 0.6602, + "step": 2903 + }, + { + "epoch": 0.09159438574357356, + "grad_norm": 3.568676710128784, + "learning_rate": 1.9802042836796072e-05, + "loss": 0.6114, + "step": 2904 + }, + { + "epoch": 0.0916259265100142, + "grad_norm": 3.377164840698242, + "learning_rate": 1.9801840528915053e-05, + "loss": 0.625, + "step": 2905 + }, + { + "epoch": 0.09165746727645482, + "grad_norm": 3.819672107696533, + "learning_rate": 1.9801638118744218e-05, + "loss": 0.911, + "step": 2906 + }, + { + "epoch": 0.09168900804289544, + "grad_norm": 3.317840099334717, + "learning_rate": 1.980143560628568e-05, + "loss": 0.6259, + "step": 2907 + }, + { + "epoch": 0.09172054880933607, + "grad_norm": 3.2253177165985107, + "learning_rate": 1.980123299154155e-05, + "loss": 0.6102, + "step": 2908 + }, + { + "epoch": 0.09175208957577669, + "grad_norm": 5.04062032699585, + "learning_rate": 1.980103027451394e-05, + "loss": 0.5616, + "step": 2909 + }, + { + "epoch": 0.09178363034221731, + "grad_norm": 8.626441955566406, + "learning_rate": 1.9800827455204968e-05, + "loss": 0.3877, + "step": 2910 + }, + { + "epoch": 0.09181517110865794, + "grad_norm": 4.012566089630127, + "learning_rate": 1.9800624533616757e-05, + "loss": 0.6076, + "step": 2911 + }, + { + "epoch": 0.09184671187509856, + "grad_norm": 11.499746322631836, + "learning_rate": 1.9800421509751416e-05, + "loss": 0.6212, + "step": 2912 + }, + { + "epoch": 0.09187825264153919, + "grad_norm": 3.9195783138275146, + "learning_rate": 1.9800218383611067e-05, + "loss": 0.5832, + "step": 2913 + }, + { + "epoch": 0.09190979340797982, + "grad_norm": 2.9043338298797607, + "learning_rate": 1.980001515519783e-05, + "loss": 0.6236, + "step": 2914 + }, + { + "epoch": 0.09194133417442044, + "grad_norm": 5.669114589691162, + "learning_rate": 1.9799811824513824e-05, + "loss": 0.6097, + "step": 2915 + }, + { + "epoch": 0.09197287494086107, + "grad_norm": 14.879127502441406, + "learning_rate": 1.9799608391561174e-05, + "loss": 0.642, + "step": 2916 + }, + { + "epoch": 0.09200441570730168, + "grad_norm": 4.022953987121582, + "learning_rate": 1.9799404856342e-05, + "loss": 0.9137, + "step": 2917 + }, + { + "epoch": 0.09203595647374231, + "grad_norm": 2.6720852851867676, + "learning_rate": 1.9799201218858427e-05, + "loss": 0.3836, + "step": 2918 + }, + { + "epoch": 0.09206749724018294, + "grad_norm": 4.384119987487793, + "learning_rate": 1.9798997479112582e-05, + "loss": 0.5751, + "step": 2919 + }, + { + "epoch": 0.09209903800662356, + "grad_norm": 3.566545009613037, + "learning_rate": 1.9798793637106587e-05, + "loss": 0.6103, + "step": 2920 + }, + { + "epoch": 0.09213057877306419, + "grad_norm": 2.656848430633545, + "learning_rate": 1.979858969284257e-05, + "loss": 0.3801, + "step": 2921 + }, + { + "epoch": 0.09216211953950482, + "grad_norm": 7.7693257331848145, + "learning_rate": 1.979838564632267e-05, + "loss": 0.6454, + "step": 2922 + }, + { + "epoch": 0.09219366030594543, + "grad_norm": 4.657165050506592, + "learning_rate": 1.9798181497549e-05, + "loss": 0.6117, + "step": 2923 + }, + { + "epoch": 0.09222520107238606, + "grad_norm": 6.7317938804626465, + "learning_rate": 1.9797977246523697e-05, + "loss": 0.6126, + "step": 2924 + }, + { + "epoch": 0.09225674183882668, + "grad_norm": 5.2576751708984375, + "learning_rate": 1.97977728932489e-05, + "loss": 0.6029, + "step": 2925 + }, + { + "epoch": 0.09228828260526731, + "grad_norm": 2.742988109588623, + "learning_rate": 1.979756843772673e-05, + "loss": 0.3983, + "step": 2926 + }, + { + "epoch": 0.09231982337170794, + "grad_norm": 4.423510551452637, + "learning_rate": 1.9797363879959327e-05, + "loss": 0.8907, + "step": 2927 + }, + { + "epoch": 0.09235136413814855, + "grad_norm": 9.096057891845703, + "learning_rate": 1.9797159219948824e-05, + "loss": 0.6585, + "step": 2928 + }, + { + "epoch": 0.09238290490458918, + "grad_norm": 4.524240016937256, + "learning_rate": 1.9796954457697358e-05, + "loss": 0.6641, + "step": 2929 + }, + { + "epoch": 0.09241444567102981, + "grad_norm": 3.965543508529663, + "learning_rate": 1.9796749593207067e-05, + "loss": 0.6089, + "step": 2930 + }, + { + "epoch": 0.09244598643747043, + "grad_norm": 4.74583101272583, + "learning_rate": 1.9796544626480086e-05, + "loss": 0.6076, + "step": 2931 + }, + { + "epoch": 0.09247752720391106, + "grad_norm": 4.643465518951416, + "learning_rate": 1.9796339557518552e-05, + "loss": 0.6081, + "step": 2932 + }, + { + "epoch": 0.09250906797035167, + "grad_norm": 4.063772678375244, + "learning_rate": 1.979613438632461e-05, + "loss": 0.6124, + "step": 2933 + }, + { + "epoch": 0.0925406087367923, + "grad_norm": 5.506024360656738, + "learning_rate": 1.97959291129004e-05, + "loss": 0.6632, + "step": 2934 + }, + { + "epoch": 0.09257214950323293, + "grad_norm": 6.2220892906188965, + "learning_rate": 1.979572373724806e-05, + "loss": 0.5853, + "step": 2935 + }, + { + "epoch": 0.09260369026967355, + "grad_norm": 4.186939716339111, + "learning_rate": 1.979551825936974e-05, + "loss": 0.6132, + "step": 2936 + }, + { + "epoch": 0.09263523103611418, + "grad_norm": 3.7670071125030518, + "learning_rate": 1.979531267926758e-05, + "loss": 0.6194, + "step": 2937 + }, + { + "epoch": 0.0926667718025548, + "grad_norm": 5.682912349700928, + "learning_rate": 1.9795106996943728e-05, + "loss": 0.6261, + "step": 2938 + }, + { + "epoch": 0.09269831256899543, + "grad_norm": 4.914506912231445, + "learning_rate": 1.9794901212400327e-05, + "loss": 0.5998, + "step": 2939 + }, + { + "epoch": 0.09272985333543605, + "grad_norm": 3.9787309169769287, + "learning_rate": 1.9794695325639525e-05, + "loss": 0.6466, + "step": 2940 + }, + { + "epoch": 0.09276139410187667, + "grad_norm": 4.9087605476379395, + "learning_rate": 1.9794489336663475e-05, + "loss": 0.5891, + "step": 2941 + }, + { + "epoch": 0.0927929348683173, + "grad_norm": 4.4231367111206055, + "learning_rate": 1.979428324547432e-05, + "loss": 0.6418, + "step": 2942 + }, + { + "epoch": 0.09282447563475793, + "grad_norm": 5.362968444824219, + "learning_rate": 1.9794077052074218e-05, + "loss": 0.5698, + "step": 2943 + }, + { + "epoch": 0.09285601640119855, + "grad_norm": 8.581646919250488, + "learning_rate": 1.9793870756465313e-05, + "loss": 0.3904, + "step": 2944 + }, + { + "epoch": 0.09288755716763918, + "grad_norm": 4.205008029937744, + "learning_rate": 1.9793664358649767e-05, + "loss": 0.6444, + "step": 2945 + }, + { + "epoch": 0.09291909793407979, + "grad_norm": 4.755532741546631, + "learning_rate": 1.9793457858629725e-05, + "loss": 0.652, + "step": 2946 + }, + { + "epoch": 0.09295063870052042, + "grad_norm": 4.282418251037598, + "learning_rate": 1.9793251256407348e-05, + "loss": 0.6063, + "step": 2947 + }, + { + "epoch": 0.09298217946696105, + "grad_norm": 3.0517258644104004, + "learning_rate": 1.979304455198479e-05, + "loss": 0.6021, + "step": 2948 + }, + { + "epoch": 0.09301372023340167, + "grad_norm": 3.6174705028533936, + "learning_rate": 1.9792837745364205e-05, + "loss": 0.6078, + "step": 2949 + }, + { + "epoch": 0.0930452609998423, + "grad_norm": 2.6222121715545654, + "learning_rate": 1.9792630836547756e-05, + "loss": 0.9232, + "step": 2950 + }, + { + "epoch": 0.09307680176628293, + "grad_norm": 4.16685152053833, + "learning_rate": 1.9792423825537602e-05, + "loss": 0.6155, + "step": 2951 + }, + { + "epoch": 0.09310834253272354, + "grad_norm": 3.9732606410980225, + "learning_rate": 1.97922167123359e-05, + "loss": 0.6063, + "step": 2952 + }, + { + "epoch": 0.09313988329916417, + "grad_norm": 4.8852667808532715, + "learning_rate": 1.9792009496944812e-05, + "loss": 0.6341, + "step": 2953 + }, + { + "epoch": 0.09317142406560479, + "grad_norm": 5.282898426055908, + "learning_rate": 1.9791802179366502e-05, + "loss": 0.6209, + "step": 2954 + }, + { + "epoch": 0.09320296483204542, + "grad_norm": 4.201132774353027, + "learning_rate": 1.9791594759603137e-05, + "loss": 0.6302, + "step": 2955 + }, + { + "epoch": 0.09323450559848605, + "grad_norm": 5.1396942138671875, + "learning_rate": 1.9791387237656875e-05, + "loss": 0.6216, + "step": 2956 + }, + { + "epoch": 0.09326604636492666, + "grad_norm": 3.440955877304077, + "learning_rate": 1.9791179613529884e-05, + "loss": 0.6462, + "step": 2957 + }, + { + "epoch": 0.0932975871313673, + "grad_norm": 4.805516719818115, + "learning_rate": 1.979097188722433e-05, + "loss": 0.6268, + "step": 2958 + }, + { + "epoch": 0.09332912789780792, + "grad_norm": 3.617138385772705, + "learning_rate": 1.9790764058742384e-05, + "loss": 0.6172, + "step": 2959 + }, + { + "epoch": 0.09336066866424854, + "grad_norm": 3.3243088722229004, + "learning_rate": 1.979055612808621e-05, + "loss": 0.5929, + "step": 2960 + }, + { + "epoch": 0.09339220943068917, + "grad_norm": 3.944315195083618, + "learning_rate": 1.9790348095257983e-05, + "loss": 0.5979, + "step": 2961 + }, + { + "epoch": 0.09342375019712978, + "grad_norm": 3.164827823638916, + "learning_rate": 1.9790139960259872e-05, + "loss": 0.6267, + "step": 2962 + }, + { + "epoch": 0.09345529096357041, + "grad_norm": 12.941017150878906, + "learning_rate": 1.9789931723094046e-05, + "loss": 0.5854, + "step": 2963 + }, + { + "epoch": 0.09348683173001104, + "grad_norm": 5.106287479400635, + "learning_rate": 1.9789723383762684e-05, + "loss": 0.6177, + "step": 2964 + }, + { + "epoch": 0.09351837249645166, + "grad_norm": 2.595815420150757, + "learning_rate": 1.9789514942267954e-05, + "loss": 0.5779, + "step": 2965 + }, + { + "epoch": 0.09354991326289229, + "grad_norm": 2.905066728591919, + "learning_rate": 1.9789306398612038e-05, + "loss": 0.6133, + "step": 2966 + }, + { + "epoch": 0.09358145402933292, + "grad_norm": 3.387725353240967, + "learning_rate": 1.9789097752797105e-05, + "loss": 0.6579, + "step": 2967 + }, + { + "epoch": 0.09361299479577354, + "grad_norm": 3.489276885986328, + "learning_rate": 1.9788889004825336e-05, + "loss": 0.593, + "step": 2968 + }, + { + "epoch": 0.09364453556221417, + "grad_norm": 4.302872180938721, + "learning_rate": 1.978868015469891e-05, + "loss": 0.5788, + "step": 2969 + }, + { + "epoch": 0.09367607632865478, + "grad_norm": 6.687342166900635, + "learning_rate": 1.9788471202420006e-05, + "loss": 0.6018, + "step": 2970 + }, + { + "epoch": 0.09370761709509541, + "grad_norm": 4.108180522918701, + "learning_rate": 1.9788262147990806e-05, + "loss": 0.612, + "step": 2971 + }, + { + "epoch": 0.09373915786153604, + "grad_norm": 23.07948875427246, + "learning_rate": 1.9788052991413487e-05, + "loss": 0.6056, + "step": 2972 + }, + { + "epoch": 0.09377069862797666, + "grad_norm": 5.799622535705566, + "learning_rate": 1.978784373269024e-05, + "loss": 0.5758, + "step": 2973 + }, + { + "epoch": 0.09380223939441729, + "grad_norm": 5.525553226470947, + "learning_rate": 1.978763437182324e-05, + "loss": 0.6286, + "step": 2974 + }, + { + "epoch": 0.0938337801608579, + "grad_norm": 3.188157796859741, + "learning_rate": 1.9787424908814675e-05, + "loss": 0.6151, + "step": 2975 + }, + { + "epoch": 0.09386532092729853, + "grad_norm": 3.248171329498291, + "learning_rate": 1.9787215343666732e-05, + "loss": 0.5801, + "step": 2976 + }, + { + "epoch": 0.09389686169373916, + "grad_norm": 4.896704196929932, + "learning_rate": 1.9787005676381598e-05, + "loss": 0.6362, + "step": 2977 + }, + { + "epoch": 0.09392840246017978, + "grad_norm": 4.684083461761475, + "learning_rate": 1.978679590696146e-05, + "loss": 0.6128, + "step": 2978 + }, + { + "epoch": 0.09395994322662041, + "grad_norm": 5.421011924743652, + "learning_rate": 1.9786586035408504e-05, + "loss": 0.6134, + "step": 2979 + }, + { + "epoch": 0.09399148399306104, + "grad_norm": 3.2045397758483887, + "learning_rate": 1.978637606172493e-05, + "loss": 0.5934, + "step": 2980 + }, + { + "epoch": 0.09402302475950165, + "grad_norm": 3.5366721153259277, + "learning_rate": 1.978616598591292e-05, + "loss": 0.6126, + "step": 2981 + }, + { + "epoch": 0.09405456552594228, + "grad_norm": 4.680776596069336, + "learning_rate": 1.9785955807974666e-05, + "loss": 0.6032, + "step": 2982 + }, + { + "epoch": 0.0940861062923829, + "grad_norm": 6.996942520141602, + "learning_rate": 1.978574552791237e-05, + "loss": 0.5886, + "step": 2983 + }, + { + "epoch": 0.09411764705882353, + "grad_norm": 2.990128755569458, + "learning_rate": 1.9785535145728215e-05, + "loss": 0.5899, + "step": 2984 + }, + { + "epoch": 0.09414918782526416, + "grad_norm": 3.5269854068756104, + "learning_rate": 1.9785324661424407e-05, + "loss": 0.5907, + "step": 2985 + }, + { + "epoch": 0.09418072859170477, + "grad_norm": 3.5212502479553223, + "learning_rate": 1.9785114075003137e-05, + "loss": 0.6156, + "step": 2986 + }, + { + "epoch": 0.0942122693581454, + "grad_norm": 3.0174720287323, + "learning_rate": 1.9784903386466604e-05, + "loss": 0.6352, + "step": 2987 + }, + { + "epoch": 0.09424381012458603, + "grad_norm": 5.631509780883789, + "learning_rate": 1.9784692595817007e-05, + "loss": 0.622, + "step": 2988 + }, + { + "epoch": 0.09427535089102665, + "grad_norm": 9.230667114257812, + "learning_rate": 1.9784481703056545e-05, + "loss": 0.6076, + "step": 2989 + }, + { + "epoch": 0.09430689165746728, + "grad_norm": 5.173788070678711, + "learning_rate": 1.9784270708187416e-05, + "loss": 0.6244, + "step": 2990 + }, + { + "epoch": 0.0943384324239079, + "grad_norm": 13.625312805175781, + "learning_rate": 1.9784059611211826e-05, + "loss": 0.6383, + "step": 2991 + }, + { + "epoch": 0.09436997319034852, + "grad_norm": 3.491774559020996, + "learning_rate": 1.978384841213198e-05, + "loss": 0.6173, + "step": 2992 + }, + { + "epoch": 0.09440151395678915, + "grad_norm": 4.49334716796875, + "learning_rate": 1.9783637110950077e-05, + "loss": 0.6428, + "step": 2993 + }, + { + "epoch": 0.09443305472322977, + "grad_norm": 3.2790746688842773, + "learning_rate": 1.9783425707668322e-05, + "loss": 0.6129, + "step": 2994 + }, + { + "epoch": 0.0944645954896704, + "grad_norm": 3.7995777130126953, + "learning_rate": 1.9783214202288926e-05, + "loss": 0.5889, + "step": 2995 + }, + { + "epoch": 0.09449613625611103, + "grad_norm": 4.007745742797852, + "learning_rate": 1.978300259481409e-05, + "loss": 0.6033, + "step": 2996 + }, + { + "epoch": 0.09452767702255165, + "grad_norm": 3.457394599914551, + "learning_rate": 1.978279088524603e-05, + "loss": 0.5952, + "step": 2997 + }, + { + "epoch": 0.09455921778899228, + "grad_norm": 8.60282039642334, + "learning_rate": 1.9782579073586948e-05, + "loss": 0.5664, + "step": 2998 + }, + { + "epoch": 0.09459075855543289, + "grad_norm": 17.386756896972656, + "learning_rate": 1.9782367159839056e-05, + "loss": 0.5748, + "step": 2999 + }, + { + "epoch": 0.09462229932187352, + "grad_norm": 4.246764183044434, + "learning_rate": 1.978215514400457e-05, + "loss": 0.6308, + "step": 3000 + }, + { + "epoch": 0.09465384008831415, + "grad_norm": 3.885016679763794, + "learning_rate": 1.97819430260857e-05, + "loss": 0.9268, + "step": 3001 + }, + { + "epoch": 0.09468538085475477, + "grad_norm": 4.573757171630859, + "learning_rate": 1.9781730806084657e-05, + "loss": 0.6232, + "step": 3002 + }, + { + "epoch": 0.0947169216211954, + "grad_norm": 8.429253578186035, + "learning_rate": 1.9781518484003658e-05, + "loss": 0.6299, + "step": 3003 + }, + { + "epoch": 0.09474846238763603, + "grad_norm": 4.475785732269287, + "learning_rate": 1.9781306059844917e-05, + "loss": 0.8891, + "step": 3004 + }, + { + "epoch": 0.09478000315407664, + "grad_norm": 6.022606372833252, + "learning_rate": 1.9781093533610655e-05, + "loss": 0.6439, + "step": 3005 + }, + { + "epoch": 0.09481154392051727, + "grad_norm": 4.307050704956055, + "learning_rate": 1.9780880905303086e-05, + "loss": 0.6509, + "step": 3006 + }, + { + "epoch": 0.09484308468695789, + "grad_norm": 5.365092754364014, + "learning_rate": 1.9780668174924427e-05, + "loss": 0.6234, + "step": 3007 + }, + { + "epoch": 0.09487462545339852, + "grad_norm": 4.655923366546631, + "learning_rate": 1.9780455342476906e-05, + "loss": 0.593, + "step": 3008 + }, + { + "epoch": 0.09490616621983915, + "grad_norm": 4.471737861633301, + "learning_rate": 1.978024240796274e-05, + "loss": 0.5944, + "step": 3009 + }, + { + "epoch": 0.09493770698627976, + "grad_norm": 4.500119209289551, + "learning_rate": 1.9780029371384147e-05, + "loss": 0.6369, + "step": 3010 + }, + { + "epoch": 0.09496924775272039, + "grad_norm": 4.04485559463501, + "learning_rate": 1.9779816232743353e-05, + "loss": 0.5971, + "step": 3011 + }, + { + "epoch": 0.09500078851916102, + "grad_norm": 4.106805324554443, + "learning_rate": 1.9779602992042584e-05, + "loss": 0.6272, + "step": 3012 + }, + { + "epoch": 0.09503232928560164, + "grad_norm": 4.567190170288086, + "learning_rate": 1.9779389649284065e-05, + "loss": 0.6281, + "step": 3013 + }, + { + "epoch": 0.09506387005204227, + "grad_norm": 4.523330211639404, + "learning_rate": 1.977917620447002e-05, + "loss": 0.617, + "step": 3014 + }, + { + "epoch": 0.09509541081848288, + "grad_norm": 3.391838788986206, + "learning_rate": 1.977896265760268e-05, + "loss": 0.6134, + "step": 3015 + }, + { + "epoch": 0.09512695158492351, + "grad_norm": 3.7463457584381104, + "learning_rate": 1.9778749008684268e-05, + "loss": 0.6316, + "step": 3016 + }, + { + "epoch": 0.09515849235136414, + "grad_norm": 4.184138774871826, + "learning_rate": 1.977853525771702e-05, + "loss": 0.6216, + "step": 3017 + }, + { + "epoch": 0.09519003311780476, + "grad_norm": 6.638707160949707, + "learning_rate": 1.977832140470316e-05, + "loss": 0.8641, + "step": 3018 + }, + { + "epoch": 0.09522157388424539, + "grad_norm": 4.594690799713135, + "learning_rate": 1.9778107449644928e-05, + "loss": 0.6064, + "step": 3019 + }, + { + "epoch": 0.095253114650686, + "grad_norm": 3.2971246242523193, + "learning_rate": 1.977789339254455e-05, + "loss": 0.8979, + "step": 3020 + }, + { + "epoch": 0.09528465541712663, + "grad_norm": 3.301764726638794, + "learning_rate": 1.9777679233404267e-05, + "loss": 0.6394, + "step": 3021 + }, + { + "epoch": 0.09531619618356726, + "grad_norm": 3.0487215518951416, + "learning_rate": 1.97774649722263e-05, + "loss": 0.5931, + "step": 3022 + }, + { + "epoch": 0.09534773695000788, + "grad_norm": 2.9843602180480957, + "learning_rate": 1.97772506090129e-05, + "loss": 0.611, + "step": 3023 + }, + { + "epoch": 0.09537927771644851, + "grad_norm": 4.14968729019165, + "learning_rate": 1.9777036143766298e-05, + "loss": 0.9079, + "step": 3024 + }, + { + "epoch": 0.09541081848288914, + "grad_norm": 3.1360785961151123, + "learning_rate": 1.977682157648873e-05, + "loss": 0.4209, + "step": 3025 + }, + { + "epoch": 0.09544235924932976, + "grad_norm": 3.789788246154785, + "learning_rate": 1.977660690718244e-05, + "loss": 0.6129, + "step": 3026 + }, + { + "epoch": 0.09547390001577039, + "grad_norm": 3.945068597793579, + "learning_rate": 1.9776392135849662e-05, + "loss": 0.4023, + "step": 3027 + }, + { + "epoch": 0.095505440782211, + "grad_norm": 3.887200355529785, + "learning_rate": 1.9776177262492643e-05, + "loss": 0.6195, + "step": 3028 + }, + { + "epoch": 0.09553698154865163, + "grad_norm": 4.234487533569336, + "learning_rate": 1.9775962287113623e-05, + "loss": 0.6716, + "step": 3029 + }, + { + "epoch": 0.09556852231509226, + "grad_norm": 3.009399890899658, + "learning_rate": 1.9775747209714847e-05, + "loss": 0.9408, + "step": 3030 + }, + { + "epoch": 0.09560006308153288, + "grad_norm": 5.062989234924316, + "learning_rate": 1.9775532030298556e-05, + "loss": 0.6093, + "step": 3031 + }, + { + "epoch": 0.0956316038479735, + "grad_norm": 4.560422420501709, + "learning_rate": 1.9775316748867e-05, + "loss": 0.6013, + "step": 3032 + }, + { + "epoch": 0.09566314461441414, + "grad_norm": 5.107979774475098, + "learning_rate": 1.9775101365422423e-05, + "loss": 0.5875, + "step": 3033 + }, + { + "epoch": 0.09569468538085475, + "grad_norm": 3.9038503170013428, + "learning_rate": 1.977488587996707e-05, + "loss": 0.6024, + "step": 3034 + }, + { + "epoch": 0.09572622614729538, + "grad_norm": 2.7457034587860107, + "learning_rate": 1.9774670292503194e-05, + "loss": 0.3651, + "step": 3035 + }, + { + "epoch": 0.095757766913736, + "grad_norm": 3.5153768062591553, + "learning_rate": 1.9774454603033048e-05, + "loss": 0.4325, + "step": 3036 + }, + { + "epoch": 0.09578930768017663, + "grad_norm": 4.359099388122559, + "learning_rate": 1.9774238811558874e-05, + "loss": 0.5885, + "step": 3037 + }, + { + "epoch": 0.09582084844661726, + "grad_norm": 4.696026802062988, + "learning_rate": 1.977402291808293e-05, + "loss": 0.6169, + "step": 3038 + }, + { + "epoch": 0.09585238921305787, + "grad_norm": 4.9474005699157715, + "learning_rate": 1.9773806922607466e-05, + "loss": 0.6093, + "step": 3039 + }, + { + "epoch": 0.0958839299794985, + "grad_norm": 3.1107101440429688, + "learning_rate": 1.977359082513474e-05, + "loss": 0.8852, + "step": 3040 + }, + { + "epoch": 0.09591547074593913, + "grad_norm": 5.029510021209717, + "learning_rate": 1.9773374625667e-05, + "loss": 0.6135, + "step": 3041 + }, + { + "epoch": 0.09594701151237975, + "grad_norm": 4.711209297180176, + "learning_rate": 1.9773158324206513e-05, + "loss": 0.6136, + "step": 3042 + }, + { + "epoch": 0.09597855227882038, + "grad_norm": 4.1432905197143555, + "learning_rate": 1.9772941920755528e-05, + "loss": 0.6348, + "step": 3043 + }, + { + "epoch": 0.096010093045261, + "grad_norm": 6.445088863372803, + "learning_rate": 1.9772725415316307e-05, + "loss": 0.5907, + "step": 3044 + }, + { + "epoch": 0.09604163381170162, + "grad_norm": 4.874128818511963, + "learning_rate": 1.9772508807891104e-05, + "loss": 0.5818, + "step": 3045 + }, + { + "epoch": 0.09607317457814225, + "grad_norm": 7.019410133361816, + "learning_rate": 1.9772292098482185e-05, + "loss": 0.5857, + "step": 3046 + }, + { + "epoch": 0.09610471534458287, + "grad_norm": 6.054513931274414, + "learning_rate": 1.977207528709181e-05, + "loss": 0.6301, + "step": 3047 + }, + { + "epoch": 0.0961362561110235, + "grad_norm": 6.217750549316406, + "learning_rate": 1.9771858373722247e-05, + "loss": 0.6479, + "step": 3048 + }, + { + "epoch": 0.09616779687746413, + "grad_norm": 4.490233421325684, + "learning_rate": 1.9771641358375748e-05, + "loss": 0.563, + "step": 3049 + }, + { + "epoch": 0.09619933764390474, + "grad_norm": 6.368677616119385, + "learning_rate": 1.9771424241054583e-05, + "loss": 0.6057, + "step": 3050 + }, + { + "epoch": 0.09623087841034537, + "grad_norm": 3.663496971130371, + "learning_rate": 1.9771207021761022e-05, + "loss": 0.6353, + "step": 3051 + }, + { + "epoch": 0.09626241917678599, + "grad_norm": 5.129685878753662, + "learning_rate": 1.977098970049733e-05, + "loss": 0.6132, + "step": 3052 + }, + { + "epoch": 0.09629395994322662, + "grad_norm": 4.094722747802734, + "learning_rate": 1.977077227726577e-05, + "loss": 0.6325, + "step": 3053 + }, + { + "epoch": 0.09632550070966725, + "grad_norm": 3.008125066757202, + "learning_rate": 1.9770554752068615e-05, + "loss": 0.9178, + "step": 3054 + }, + { + "epoch": 0.09635704147610787, + "grad_norm": 4.510913848876953, + "learning_rate": 1.9770337124908138e-05, + "loss": 0.6124, + "step": 3055 + }, + { + "epoch": 0.0963885822425485, + "grad_norm": 4.092448711395264, + "learning_rate": 1.97701193957866e-05, + "loss": 0.5867, + "step": 3056 + }, + { + "epoch": 0.09642012300898913, + "grad_norm": 11.24302864074707, + "learning_rate": 1.9769901564706283e-05, + "loss": 0.8814, + "step": 3057 + }, + { + "epoch": 0.09645166377542974, + "grad_norm": 3.7733473777770996, + "learning_rate": 1.976968363166946e-05, + "loss": 0.5824, + "step": 3058 + }, + { + "epoch": 0.09648320454187037, + "grad_norm": 4.510855674743652, + "learning_rate": 1.97694655966784e-05, + "loss": 0.6179, + "step": 3059 + }, + { + "epoch": 0.09651474530831099, + "grad_norm": 11.892997741699219, + "learning_rate": 1.976924745973538e-05, + "loss": 0.6302, + "step": 3060 + }, + { + "epoch": 0.09654628607475162, + "grad_norm": 4.523686408996582, + "learning_rate": 1.9769029220842678e-05, + "loss": 0.6343, + "step": 3061 + }, + { + "epoch": 0.09657782684119225, + "grad_norm": 3.290720224380493, + "learning_rate": 1.976881088000257e-05, + "loss": 0.8824, + "step": 3062 + }, + { + "epoch": 0.09660936760763286, + "grad_norm": 4.609562397003174, + "learning_rate": 1.9768592437217337e-05, + "loss": 0.6218, + "step": 3063 + }, + { + "epoch": 0.09664090837407349, + "grad_norm": 2.0890793800354004, + "learning_rate": 1.9768373892489253e-05, + "loss": 0.8894, + "step": 3064 + }, + { + "epoch": 0.09667244914051411, + "grad_norm": 3.3025903701782227, + "learning_rate": 1.9768155245820605e-05, + "loss": 0.9103, + "step": 3065 + }, + { + "epoch": 0.09670398990695474, + "grad_norm": 4.61625337600708, + "learning_rate": 1.9767936497213672e-05, + "loss": 0.6396, + "step": 3066 + }, + { + "epoch": 0.09673553067339537, + "grad_norm": 6.074202537536621, + "learning_rate": 1.976771764667074e-05, + "loss": 0.6285, + "step": 3067 + }, + { + "epoch": 0.09676707143983598, + "grad_norm": 4.665162086486816, + "learning_rate": 1.9767498694194086e-05, + "loss": 0.6417, + "step": 3068 + }, + { + "epoch": 0.09679861220627661, + "grad_norm": 8.752592086791992, + "learning_rate": 1.9767279639786e-05, + "loss": 0.6403, + "step": 3069 + }, + { + "epoch": 0.09683015297271724, + "grad_norm": 5.194988250732422, + "learning_rate": 1.9767060483448767e-05, + "loss": 0.5942, + "step": 3070 + }, + { + "epoch": 0.09686169373915786, + "grad_norm": 8.778670310974121, + "learning_rate": 1.9766841225184675e-05, + "loss": 0.6424, + "step": 3071 + }, + { + "epoch": 0.09689323450559849, + "grad_norm": 7.7310614585876465, + "learning_rate": 1.976662186499601e-05, + "loss": 0.604, + "step": 3072 + }, + { + "epoch": 0.0969247752720391, + "grad_norm": 4.8976311683654785, + "learning_rate": 1.976640240288506e-05, + "loss": 0.5898, + "step": 3073 + }, + { + "epoch": 0.09695631603847973, + "grad_norm": 6.429174423217773, + "learning_rate": 1.976618283885412e-05, + "loss": 0.6107, + "step": 3074 + }, + { + "epoch": 0.09698785680492036, + "grad_norm": 5.476052284240723, + "learning_rate": 1.976596317290548e-05, + "loss": 0.6021, + "step": 3075 + }, + { + "epoch": 0.09701939757136098, + "grad_norm": 4.829239368438721, + "learning_rate": 1.976574340504143e-05, + "loss": 0.6669, + "step": 3076 + }, + { + "epoch": 0.09705093833780161, + "grad_norm": 3.9502944946289062, + "learning_rate": 1.9765523535264265e-05, + "loss": 0.4675, + "step": 3077 + }, + { + "epoch": 0.09708247910424224, + "grad_norm": 6.089339256286621, + "learning_rate": 1.9765303563576275e-05, + "loss": 0.6263, + "step": 3078 + }, + { + "epoch": 0.09711401987068286, + "grad_norm": 5.793859481811523, + "learning_rate": 1.9765083489979763e-05, + "loss": 0.5868, + "step": 3079 + }, + { + "epoch": 0.09714556063712348, + "grad_norm": 3.6758790016174316, + "learning_rate": 1.9764863314477024e-05, + "loss": 0.6107, + "step": 3080 + }, + { + "epoch": 0.0971771014035641, + "grad_norm": 4.465364933013916, + "learning_rate": 1.976464303707035e-05, + "loss": 0.6218, + "step": 3081 + }, + { + "epoch": 0.09720864217000473, + "grad_norm": 6.178931713104248, + "learning_rate": 1.976442265776205e-05, + "loss": 0.5999, + "step": 3082 + }, + { + "epoch": 0.09724018293644536, + "grad_norm": 3.3920624256134033, + "learning_rate": 1.9764202176554414e-05, + "loss": 0.9449, + "step": 3083 + }, + { + "epoch": 0.09727172370288598, + "grad_norm": 5.3644819259643555, + "learning_rate": 1.976398159344975e-05, + "loss": 0.6233, + "step": 3084 + }, + { + "epoch": 0.0973032644693266, + "grad_norm": 3.722536325454712, + "learning_rate": 1.976376090845035e-05, + "loss": 0.6231, + "step": 3085 + }, + { + "epoch": 0.09733480523576724, + "grad_norm": 2.951540946960449, + "learning_rate": 1.9763540121558526e-05, + "loss": 0.6202, + "step": 3086 + }, + { + "epoch": 0.09736634600220785, + "grad_norm": 43.9303092956543, + "learning_rate": 1.976331923277658e-05, + "loss": 0.6106, + "step": 3087 + }, + { + "epoch": 0.09739788676864848, + "grad_norm": 3.1642117500305176, + "learning_rate": 1.976309824210682e-05, + "loss": 0.6159, + "step": 3088 + }, + { + "epoch": 0.0974294275350891, + "grad_norm": 2.456296443939209, + "learning_rate": 1.9762877149551545e-05, + "loss": 0.8854, + "step": 3089 + }, + { + "epoch": 0.09746096830152973, + "grad_norm": 2.816539764404297, + "learning_rate": 1.9762655955113067e-05, + "loss": 0.4179, + "step": 3090 + }, + { + "epoch": 0.09749250906797036, + "grad_norm": 4.325496196746826, + "learning_rate": 1.9762434658793697e-05, + "loss": 0.5927, + "step": 3091 + }, + { + "epoch": 0.09752404983441097, + "grad_norm": 5.581847190856934, + "learning_rate": 1.9762213260595738e-05, + "loss": 0.62, + "step": 3092 + }, + { + "epoch": 0.0975555906008516, + "grad_norm": 3.7161245346069336, + "learning_rate": 1.9761991760521502e-05, + "loss": 0.6089, + "step": 3093 + }, + { + "epoch": 0.09758713136729223, + "grad_norm": 2.6418521404266357, + "learning_rate": 1.9761770158573305e-05, + "loss": 0.6103, + "step": 3094 + }, + { + "epoch": 0.09761867213373285, + "grad_norm": 4.195645809173584, + "learning_rate": 1.9761548454753455e-05, + "loss": 0.6449, + "step": 3095 + }, + { + "epoch": 0.09765021290017348, + "grad_norm": 4.512105464935303, + "learning_rate": 1.976132664906427e-05, + "loss": 0.6154, + "step": 3096 + }, + { + "epoch": 0.0976817536666141, + "grad_norm": 5.4428019523620605, + "learning_rate": 1.976110474150806e-05, + "loss": 0.6035, + "step": 3097 + }, + { + "epoch": 0.09771329443305472, + "grad_norm": 3.9944043159484863, + "learning_rate": 1.9760882732087137e-05, + "loss": 0.658, + "step": 3098 + }, + { + "epoch": 0.09774483519949535, + "grad_norm": 3.960244655609131, + "learning_rate": 1.976066062080383e-05, + "loss": 0.6405, + "step": 3099 + }, + { + "epoch": 0.09777637596593597, + "grad_norm": 4.034977912902832, + "learning_rate": 1.976043840766045e-05, + "loss": 0.6223, + "step": 3100 + }, + { + "epoch": 0.0978079167323766, + "grad_norm": 2.8991544246673584, + "learning_rate": 1.9760216092659314e-05, + "loss": 0.9115, + "step": 3101 + }, + { + "epoch": 0.09783945749881721, + "grad_norm": 6.504237174987793, + "learning_rate": 1.9759993675802748e-05, + "loss": 0.6574, + "step": 3102 + }, + { + "epoch": 0.09787099826525784, + "grad_norm": 4.652159214019775, + "learning_rate": 1.9759771157093066e-05, + "loss": 0.6235, + "step": 3103 + }, + { + "epoch": 0.09790253903169847, + "grad_norm": 3.2072179317474365, + "learning_rate": 1.9759548536532594e-05, + "loss": 0.6118, + "step": 3104 + }, + { + "epoch": 0.09793407979813909, + "grad_norm": 2.76709246635437, + "learning_rate": 1.9759325814123658e-05, + "loss": 0.6109, + "step": 3105 + }, + { + "epoch": 0.09796562056457972, + "grad_norm": 4.868649959564209, + "learning_rate": 1.9759102989868578e-05, + "loss": 0.608, + "step": 3106 + }, + { + "epoch": 0.09799716133102035, + "grad_norm": 3.973383665084839, + "learning_rate": 1.975888006376968e-05, + "loss": 0.6325, + "step": 3107 + }, + { + "epoch": 0.09802870209746097, + "grad_norm": 4.336780548095703, + "learning_rate": 1.975865703582929e-05, + "loss": 0.6249, + "step": 3108 + }, + { + "epoch": 0.0980602428639016, + "grad_norm": 3.294241189956665, + "learning_rate": 1.9758433906049738e-05, + "loss": 0.638, + "step": 3109 + }, + { + "epoch": 0.09809178363034221, + "grad_norm": 3.7521252632141113, + "learning_rate": 1.9758210674433353e-05, + "loss": 0.5795, + "step": 3110 + }, + { + "epoch": 0.09812332439678284, + "grad_norm": 3.9104819297790527, + "learning_rate": 1.9757987340982458e-05, + "loss": 0.6191, + "step": 3111 + }, + { + "epoch": 0.09815486516322347, + "grad_norm": 3.5458567142486572, + "learning_rate": 1.975776390569939e-05, + "loss": 0.5882, + "step": 3112 + }, + { + "epoch": 0.09818640592966409, + "grad_norm": 4.293820381164551, + "learning_rate": 1.975754036858648e-05, + "loss": 0.6181, + "step": 3113 + }, + { + "epoch": 0.09821794669610472, + "grad_norm": 5.0087971687316895, + "learning_rate": 1.9757316729646058e-05, + "loss": 0.6425, + "step": 3114 + }, + { + "epoch": 0.09824948746254535, + "grad_norm": 3.7201550006866455, + "learning_rate": 1.975709298888046e-05, + "loss": 0.6009, + "step": 3115 + }, + { + "epoch": 0.09828102822898596, + "grad_norm": 3.3204896450042725, + "learning_rate": 1.975686914629202e-05, + "loss": 0.6472, + "step": 3116 + }, + { + "epoch": 0.09831256899542659, + "grad_norm": 5.226902008056641, + "learning_rate": 1.975664520188308e-05, + "loss": 0.6615, + "step": 3117 + }, + { + "epoch": 0.09834410976186721, + "grad_norm": 4.164032459259033, + "learning_rate": 1.9756421155655965e-05, + "loss": 0.6339, + "step": 3118 + }, + { + "epoch": 0.09837565052830784, + "grad_norm": 3.320432186126709, + "learning_rate": 1.975619700761302e-05, + "loss": 0.6186, + "step": 3119 + }, + { + "epoch": 0.09840719129474847, + "grad_norm": 3.4331655502319336, + "learning_rate": 1.9755972757756585e-05, + "loss": 0.6501, + "step": 3120 + }, + { + "epoch": 0.09843873206118908, + "grad_norm": 3.2374751567840576, + "learning_rate": 1.9755748406089e-05, + "loss": 0.6589, + "step": 3121 + }, + { + "epoch": 0.09847027282762971, + "grad_norm": 3.426450729370117, + "learning_rate": 1.9755523952612603e-05, + "loss": 0.5625, + "step": 3122 + }, + { + "epoch": 0.09850181359407034, + "grad_norm": 2.595282793045044, + "learning_rate": 1.9755299397329743e-05, + "loss": 0.629, + "step": 3123 + }, + { + "epoch": 0.09853335436051096, + "grad_norm": 4.389976978302002, + "learning_rate": 1.9755074740242756e-05, + "loss": 0.9186, + "step": 3124 + }, + { + "epoch": 0.09856489512695159, + "grad_norm": 3.591627359390259, + "learning_rate": 1.975484998135399e-05, + "loss": 0.6026, + "step": 3125 + }, + { + "epoch": 0.0985964358933922, + "grad_norm": 3.3563575744628906, + "learning_rate": 1.9754625120665783e-05, + "loss": 0.6551, + "step": 3126 + }, + { + "epoch": 0.09862797665983283, + "grad_norm": 3.4804012775421143, + "learning_rate": 1.9754400158180498e-05, + "loss": 0.5852, + "step": 3127 + }, + { + "epoch": 0.09865951742627346, + "grad_norm": 2.8202426433563232, + "learning_rate": 1.975417509390047e-05, + "loss": 0.9316, + "step": 3128 + }, + { + "epoch": 0.09869105819271408, + "grad_norm": 2.9159932136535645, + "learning_rate": 1.975394992782805e-05, + "loss": 0.6366, + "step": 3129 + }, + { + "epoch": 0.09872259895915471, + "grad_norm": 3.1278157234191895, + "learning_rate": 1.9753724659965588e-05, + "loss": 0.6296, + "step": 3130 + }, + { + "epoch": 0.09875413972559534, + "grad_norm": 3.039308786392212, + "learning_rate": 1.975349929031544e-05, + "loss": 0.6134, + "step": 3131 + }, + { + "epoch": 0.09878568049203595, + "grad_norm": 3.8142311573028564, + "learning_rate": 1.9753273818879953e-05, + "loss": 0.5847, + "step": 3132 + }, + { + "epoch": 0.09881722125847658, + "grad_norm": 3.0745186805725098, + "learning_rate": 1.9753048245661476e-05, + "loss": 0.6012, + "step": 3133 + }, + { + "epoch": 0.0988487620249172, + "grad_norm": 4.984712600708008, + "learning_rate": 1.975282257066237e-05, + "loss": 0.6044, + "step": 3134 + }, + { + "epoch": 0.09888030279135783, + "grad_norm": 2.5653457641601562, + "learning_rate": 1.9752596793884988e-05, + "loss": 0.3799, + "step": 3135 + }, + { + "epoch": 0.09891184355779846, + "grad_norm": 4.079894542694092, + "learning_rate": 1.9752370915331686e-05, + "loss": 0.6008, + "step": 3136 + }, + { + "epoch": 0.09894338432423908, + "grad_norm": 13.528852462768555, + "learning_rate": 1.975214493500482e-05, + "loss": 0.6245, + "step": 3137 + }, + { + "epoch": 0.0989749250906797, + "grad_norm": 3.270188570022583, + "learning_rate": 1.9751918852906747e-05, + "loss": 0.6067, + "step": 3138 + }, + { + "epoch": 0.09900646585712033, + "grad_norm": 2.8690927028656006, + "learning_rate": 1.9751692669039833e-05, + "loss": 0.6054, + "step": 3139 + }, + { + "epoch": 0.09903800662356095, + "grad_norm": 2.076106309890747, + "learning_rate": 1.9751466383406434e-05, + "loss": 0.3547, + "step": 3140 + }, + { + "epoch": 0.09906954739000158, + "grad_norm": 2.672769546508789, + "learning_rate": 1.975123999600891e-05, + "loss": 0.628, + "step": 3141 + }, + { + "epoch": 0.0991010881564422, + "grad_norm": 4.8266472816467285, + "learning_rate": 1.9751013506849625e-05, + "loss": 0.6087, + "step": 3142 + }, + { + "epoch": 0.09913262892288283, + "grad_norm": 3.101212978363037, + "learning_rate": 1.975078691593094e-05, + "loss": 0.9292, + "step": 3143 + }, + { + "epoch": 0.09916416968932346, + "grad_norm": 5.914895057678223, + "learning_rate": 1.9750560223255226e-05, + "loss": 0.6095, + "step": 3144 + }, + { + "epoch": 0.09919571045576407, + "grad_norm": 3.043734073638916, + "learning_rate": 1.9750333428824843e-05, + "loss": 0.5888, + "step": 3145 + }, + { + "epoch": 0.0992272512222047, + "grad_norm": 3.619068145751953, + "learning_rate": 1.975010653264216e-05, + "loss": 0.66, + "step": 3146 + }, + { + "epoch": 0.09925879198864532, + "grad_norm": 4.359907627105713, + "learning_rate": 1.9749879534709543e-05, + "loss": 0.5795, + "step": 3147 + }, + { + "epoch": 0.09929033275508595, + "grad_norm": 4.228887557983398, + "learning_rate": 1.9749652435029366e-05, + "loss": 0.6208, + "step": 3148 + }, + { + "epoch": 0.09932187352152658, + "grad_norm": 4.509404182434082, + "learning_rate": 1.9749425233603995e-05, + "loss": 0.397, + "step": 3149 + }, + { + "epoch": 0.09935341428796719, + "grad_norm": 11.203267097473145, + "learning_rate": 1.9749197930435798e-05, + "loss": 0.6293, + "step": 3150 + }, + { + "epoch": 0.09938495505440782, + "grad_norm": 3.8355839252471924, + "learning_rate": 1.9748970525527153e-05, + "loss": 0.6149, + "step": 3151 + }, + { + "epoch": 0.09941649582084845, + "grad_norm": 2.867945432662964, + "learning_rate": 1.9748743018880433e-05, + "loss": 0.6123, + "step": 3152 + }, + { + "epoch": 0.09944803658728907, + "grad_norm": 3.12699556350708, + "learning_rate": 1.9748515410498004e-05, + "loss": 0.6111, + "step": 3153 + }, + { + "epoch": 0.0994795773537297, + "grad_norm": 5.4317498207092285, + "learning_rate": 1.974828770038225e-05, + "loss": 0.6192, + "step": 3154 + }, + { + "epoch": 0.09951111812017031, + "grad_norm": 3.0949623584747314, + "learning_rate": 1.9748059888535545e-05, + "loss": 0.6112, + "step": 3155 + }, + { + "epoch": 0.09954265888661094, + "grad_norm": 4.793638706207275, + "learning_rate": 1.9747831974960268e-05, + "loss": 0.6059, + "step": 3156 + }, + { + "epoch": 0.09957419965305157, + "grad_norm": 4.662914276123047, + "learning_rate": 1.9747603959658794e-05, + "loss": 0.6183, + "step": 3157 + }, + { + "epoch": 0.09960574041949219, + "grad_norm": 2.7734861373901367, + "learning_rate": 1.97473758426335e-05, + "loss": 0.4135, + "step": 3158 + }, + { + "epoch": 0.09963728118593282, + "grad_norm": 2.7815074920654297, + "learning_rate": 1.9747147623886774e-05, + "loss": 0.6075, + "step": 3159 + }, + { + "epoch": 0.09966882195237345, + "grad_norm": 2.9981722831726074, + "learning_rate": 1.9746919303420996e-05, + "loss": 0.5954, + "step": 3160 + }, + { + "epoch": 0.09970036271881406, + "grad_norm": 4.579103946685791, + "learning_rate": 1.974669088123854e-05, + "loss": 0.5985, + "step": 3161 + }, + { + "epoch": 0.0997319034852547, + "grad_norm": 4.094417095184326, + "learning_rate": 1.9746462357341805e-05, + "loss": 0.5867, + "step": 3162 + }, + { + "epoch": 0.09976344425169531, + "grad_norm": 2.871587038040161, + "learning_rate": 1.9746233731733163e-05, + "loss": 0.614, + "step": 3163 + }, + { + "epoch": 0.09979498501813594, + "grad_norm": 4.391870498657227, + "learning_rate": 1.9746005004415004e-05, + "loss": 0.6424, + "step": 3164 + }, + { + "epoch": 0.09982652578457657, + "grad_norm": 3.682342529296875, + "learning_rate": 1.9745776175389716e-05, + "loss": 0.6272, + "step": 3165 + }, + { + "epoch": 0.09985806655101719, + "grad_norm": 4.328029632568359, + "learning_rate": 1.9745547244659688e-05, + "loss": 0.6445, + "step": 3166 + }, + { + "epoch": 0.09988960731745782, + "grad_norm": 2.8005712032318115, + "learning_rate": 1.97453182122273e-05, + "loss": 0.4113, + "step": 3167 + }, + { + "epoch": 0.09992114808389845, + "grad_norm": 1.9822858572006226, + "learning_rate": 1.9745089078094955e-05, + "loss": 0.3717, + "step": 3168 + }, + { + "epoch": 0.09995268885033906, + "grad_norm": 3.998441457748413, + "learning_rate": 1.9744859842265042e-05, + "loss": 0.5972, + "step": 3169 + }, + { + "epoch": 0.09998422961677969, + "grad_norm": 3.2707231044769287, + "learning_rate": 1.9744630504739946e-05, + "loss": 0.4156, + "step": 3170 + }, + { + "epoch": 0.1000157703832203, + "grad_norm": 7.531773567199707, + "learning_rate": 1.9744401065522064e-05, + "loss": 0.5904, + "step": 3171 + }, + { + "epoch": 0.10004731114966094, + "grad_norm": 7.478074550628662, + "learning_rate": 1.974417152461379e-05, + "loss": 0.5965, + "step": 3172 + }, + { + "epoch": 0.10007885191610157, + "grad_norm": 4.589860439300537, + "learning_rate": 1.974394188201752e-05, + "loss": 0.6046, + "step": 3173 + }, + { + "epoch": 0.10011039268254218, + "grad_norm": 3.231721878051758, + "learning_rate": 1.974371213773565e-05, + "loss": 0.605, + "step": 3174 + }, + { + "epoch": 0.10014193344898281, + "grad_norm": 9.740652084350586, + "learning_rate": 1.974348229177058e-05, + "loss": 0.6262, + "step": 3175 + }, + { + "epoch": 0.10017347421542344, + "grad_norm": 3.7064080238342285, + "learning_rate": 1.974325234412471e-05, + "loss": 0.609, + "step": 3176 + }, + { + "epoch": 0.10020501498186406, + "grad_norm": 2.669820785522461, + "learning_rate": 1.974302229480043e-05, + "loss": 0.5959, + "step": 3177 + }, + { + "epoch": 0.10023655574830469, + "grad_norm": 5.5890583992004395, + "learning_rate": 1.9742792143800148e-05, + "loss": 0.62, + "step": 3178 + }, + { + "epoch": 0.1002680965147453, + "grad_norm": 6.237710475921631, + "learning_rate": 1.9742561891126267e-05, + "loss": 0.5853, + "step": 3179 + }, + { + "epoch": 0.10029963728118593, + "grad_norm": 3.4038960933685303, + "learning_rate": 1.9742331536781188e-05, + "loss": 0.5819, + "step": 3180 + }, + { + "epoch": 0.10033117804762656, + "grad_norm": 6.087203025817871, + "learning_rate": 1.974210108076731e-05, + "loss": 0.5852, + "step": 3181 + }, + { + "epoch": 0.10036271881406718, + "grad_norm": 3.7344980239868164, + "learning_rate": 1.974187052308705e-05, + "loss": 0.4615, + "step": 3182 + }, + { + "epoch": 0.10039425958050781, + "grad_norm": 3.422365665435791, + "learning_rate": 1.97416398637428e-05, + "loss": 0.8736, + "step": 3183 + }, + { + "epoch": 0.10042580034694842, + "grad_norm": 5.787038803100586, + "learning_rate": 1.974140910273698e-05, + "loss": 0.5965, + "step": 3184 + }, + { + "epoch": 0.10045734111338905, + "grad_norm": 4.195541858673096, + "learning_rate": 1.9741178240071987e-05, + "loss": 0.6175, + "step": 3185 + }, + { + "epoch": 0.10048888187982968, + "grad_norm": 5.002074241638184, + "learning_rate": 1.974094727575024e-05, + "loss": 0.631, + "step": 3186 + }, + { + "epoch": 0.1005204226462703, + "grad_norm": 3.5677220821380615, + "learning_rate": 1.974071620977414e-05, + "loss": 0.5903, + "step": 3187 + }, + { + "epoch": 0.10055196341271093, + "grad_norm": 4.842282772064209, + "learning_rate": 1.9740485042146104e-05, + "loss": 0.6337, + "step": 3188 + }, + { + "epoch": 0.10058350417915156, + "grad_norm": 7.583598613739014, + "learning_rate": 1.9740253772868544e-05, + "loss": 0.5944, + "step": 3189 + }, + { + "epoch": 0.10061504494559217, + "grad_norm": 3.1686320304870605, + "learning_rate": 1.9740022401943874e-05, + "loss": 0.6141, + "step": 3190 + }, + { + "epoch": 0.1006465857120328, + "grad_norm": 4.031219959259033, + "learning_rate": 1.9739790929374504e-05, + "loss": 0.8987, + "step": 3191 + }, + { + "epoch": 0.10067812647847342, + "grad_norm": 4.028090000152588, + "learning_rate": 1.9739559355162857e-05, + "loss": 0.6354, + "step": 3192 + }, + { + "epoch": 0.10070966724491405, + "grad_norm": 6.525292873382568, + "learning_rate": 1.9739327679311343e-05, + "loss": 0.6046, + "step": 3193 + }, + { + "epoch": 0.10074120801135468, + "grad_norm": 3.7867465019226074, + "learning_rate": 1.9739095901822384e-05, + "loss": 0.5768, + "step": 3194 + }, + { + "epoch": 0.1007727487777953, + "grad_norm": 3.380300283432007, + "learning_rate": 1.9738864022698396e-05, + "loss": 0.6032, + "step": 3195 + }, + { + "epoch": 0.10080428954423593, + "grad_norm": 3.8284873962402344, + "learning_rate": 1.9738632041941797e-05, + "loss": 0.6191, + "step": 3196 + }, + { + "epoch": 0.10083583031067656, + "grad_norm": 4.279428482055664, + "learning_rate": 1.973839995955501e-05, + "loss": 0.5983, + "step": 3197 + }, + { + "epoch": 0.10086737107711717, + "grad_norm": 11.744518280029297, + "learning_rate": 1.9738167775540463e-05, + "loss": 0.6137, + "step": 3198 + }, + { + "epoch": 0.1008989118435578, + "grad_norm": 3.5013840198516846, + "learning_rate": 1.9737935489900572e-05, + "loss": 0.3799, + "step": 3199 + }, + { + "epoch": 0.10093045260999842, + "grad_norm": 4.595207214355469, + "learning_rate": 1.9737703102637768e-05, + "loss": 0.6256, + "step": 3200 + }, + { + "epoch": 0.10096199337643905, + "grad_norm": 4.194145679473877, + "learning_rate": 1.9737470613754463e-05, + "loss": 0.5985, + "step": 3201 + }, + { + "epoch": 0.10099353414287968, + "grad_norm": 3.633256673812866, + "learning_rate": 1.9737238023253093e-05, + "loss": 0.6249, + "step": 3202 + }, + { + "epoch": 0.10102507490932029, + "grad_norm": 4.243246078491211, + "learning_rate": 1.9737005331136087e-05, + "loss": 0.5945, + "step": 3203 + }, + { + "epoch": 0.10105661567576092, + "grad_norm": 3.928676128387451, + "learning_rate": 1.9736772537405865e-05, + "loss": 0.636, + "step": 3204 + }, + { + "epoch": 0.10108815644220155, + "grad_norm": 3.645618438720703, + "learning_rate": 1.9736539642064865e-05, + "loss": 0.5854, + "step": 3205 + }, + { + "epoch": 0.10111969720864217, + "grad_norm": 3.1689372062683105, + "learning_rate": 1.973630664511551e-05, + "loss": 0.6036, + "step": 3206 + }, + { + "epoch": 0.1011512379750828, + "grad_norm": 2.857614278793335, + "learning_rate": 1.973607354656024e-05, + "loss": 0.605, + "step": 3207 + }, + { + "epoch": 0.10118277874152341, + "grad_norm": 5.084656715393066, + "learning_rate": 1.9735840346401477e-05, + "loss": 0.6177, + "step": 3208 + }, + { + "epoch": 0.10121431950796404, + "grad_norm": 5.703878879547119, + "learning_rate": 1.9735607044641662e-05, + "loss": 0.6045, + "step": 3209 + }, + { + "epoch": 0.10124586027440467, + "grad_norm": 4.894468307495117, + "learning_rate": 1.973537364128323e-05, + "loss": 0.6178, + "step": 3210 + }, + { + "epoch": 0.10127740104084529, + "grad_norm": 5.984530448913574, + "learning_rate": 1.9735140136328614e-05, + "loss": 0.6216, + "step": 3211 + }, + { + "epoch": 0.10130894180728592, + "grad_norm": 2.7884254455566406, + "learning_rate": 1.973490652978025e-05, + "loss": 0.4037, + "step": 3212 + }, + { + "epoch": 0.10134048257372655, + "grad_norm": 4.168567180633545, + "learning_rate": 1.9734672821640583e-05, + "loss": 0.6191, + "step": 3213 + }, + { + "epoch": 0.10137202334016716, + "grad_norm": 4.361385345458984, + "learning_rate": 1.973443901191204e-05, + "loss": 0.6284, + "step": 3214 + }, + { + "epoch": 0.1014035641066078, + "grad_norm": 4.202123165130615, + "learning_rate": 1.973420510059707e-05, + "loss": 0.6382, + "step": 3215 + }, + { + "epoch": 0.10143510487304841, + "grad_norm": 3.6014609336853027, + "learning_rate": 1.973397108769811e-05, + "loss": 0.623, + "step": 3216 + }, + { + "epoch": 0.10146664563948904, + "grad_norm": 3.1161513328552246, + "learning_rate": 1.9733736973217606e-05, + "loss": 0.6009, + "step": 3217 + }, + { + "epoch": 0.10149818640592967, + "grad_norm": 2.2281949520111084, + "learning_rate": 1.9733502757157997e-05, + "loss": 0.8795, + "step": 3218 + }, + { + "epoch": 0.10152972717237028, + "grad_norm": 3.626136302947998, + "learning_rate": 1.9733268439521726e-05, + "loss": 0.6088, + "step": 3219 + }, + { + "epoch": 0.10156126793881091, + "grad_norm": 4.806276321411133, + "learning_rate": 1.9733034020311246e-05, + "loss": 0.6168, + "step": 3220 + }, + { + "epoch": 0.10159280870525154, + "grad_norm": 2.8358218669891357, + "learning_rate": 1.9732799499528993e-05, + "loss": 0.3834, + "step": 3221 + }, + { + "epoch": 0.10162434947169216, + "grad_norm": 3.9302897453308105, + "learning_rate": 1.9732564877177426e-05, + "loss": 0.625, + "step": 3222 + }, + { + "epoch": 0.10165589023813279, + "grad_norm": 3.849076509475708, + "learning_rate": 1.9732330153258987e-05, + "loss": 0.6059, + "step": 3223 + }, + { + "epoch": 0.1016874310045734, + "grad_norm": 4.481762886047363, + "learning_rate": 1.9732095327776123e-05, + "loss": 0.6202, + "step": 3224 + }, + { + "epoch": 0.10171897177101404, + "grad_norm": 4.671913146972656, + "learning_rate": 1.9731860400731288e-05, + "loss": 0.6205, + "step": 3225 + }, + { + "epoch": 0.10175051253745467, + "grad_norm": 5.649538516998291, + "learning_rate": 1.973162537212693e-05, + "loss": 0.6102, + "step": 3226 + }, + { + "epoch": 0.10178205330389528, + "grad_norm": 3.8075504302978516, + "learning_rate": 1.973139024196551e-05, + "loss": 0.6165, + "step": 3227 + }, + { + "epoch": 0.10181359407033591, + "grad_norm": 3.8257291316986084, + "learning_rate": 1.9731155010249473e-05, + "loss": 0.6124, + "step": 3228 + }, + { + "epoch": 0.10184513483677653, + "grad_norm": 4.053287506103516, + "learning_rate": 1.973091967698128e-05, + "loss": 0.6019, + "step": 3229 + }, + { + "epoch": 0.10187667560321716, + "grad_norm": 4.503759384155273, + "learning_rate": 1.9730684242163383e-05, + "loss": 0.6319, + "step": 3230 + }, + { + "epoch": 0.10190821636965779, + "grad_norm": 5.13036584854126, + "learning_rate": 1.973044870579824e-05, + "loss": 0.641, + "step": 3231 + }, + { + "epoch": 0.1019397571360984, + "grad_norm": 4.1751508712768555, + "learning_rate": 1.973021306788831e-05, + "loss": 0.6052, + "step": 3232 + }, + { + "epoch": 0.10197129790253903, + "grad_norm": 4.888522624969482, + "learning_rate": 1.972997732843605e-05, + "loss": 0.6114, + "step": 3233 + }, + { + "epoch": 0.10200283866897966, + "grad_norm": 5.309656620025635, + "learning_rate": 1.9729741487443924e-05, + "loss": 0.6629, + "step": 3234 + }, + { + "epoch": 0.10203437943542028, + "grad_norm": 4.918057918548584, + "learning_rate": 1.972950554491439e-05, + "loss": 0.4306, + "step": 3235 + }, + { + "epoch": 0.10206592020186091, + "grad_norm": 4.211336612701416, + "learning_rate": 1.972926950084991e-05, + "loss": 0.6158, + "step": 3236 + }, + { + "epoch": 0.10209746096830152, + "grad_norm": 4.697362899780273, + "learning_rate": 1.9729033355252947e-05, + "loss": 0.6151, + "step": 3237 + }, + { + "epoch": 0.10212900173474215, + "grad_norm": 4.050116539001465, + "learning_rate": 1.9728797108125966e-05, + "loss": 0.604, + "step": 3238 + }, + { + "epoch": 0.10216054250118278, + "grad_norm": 3.0966007709503174, + "learning_rate": 1.9728560759471435e-05, + "loss": 0.575, + "step": 3239 + }, + { + "epoch": 0.1021920832676234, + "grad_norm": 4.2923197746276855, + "learning_rate": 1.972832430929182e-05, + "loss": 0.6057, + "step": 3240 + }, + { + "epoch": 0.10222362403406403, + "grad_norm": 4.724855422973633, + "learning_rate": 1.9728087757589584e-05, + "loss": 0.6092, + "step": 3241 + }, + { + "epoch": 0.10225516480050466, + "grad_norm": 2.9714674949645996, + "learning_rate": 1.97278511043672e-05, + "loss": 0.3723, + "step": 3242 + }, + { + "epoch": 0.10228670556694527, + "grad_norm": 5.260580062866211, + "learning_rate": 1.972761434962714e-05, + "loss": 0.6269, + "step": 3243 + }, + { + "epoch": 0.1023182463333859, + "grad_norm": 4.287580490112305, + "learning_rate": 1.972737749337186e-05, + "loss": 0.6276, + "step": 3244 + }, + { + "epoch": 0.10234978709982652, + "grad_norm": 3.8692238330841064, + "learning_rate": 1.972714053560385e-05, + "loss": 0.591, + "step": 3245 + }, + { + "epoch": 0.10238132786626715, + "grad_norm": 4.352402687072754, + "learning_rate": 1.9726903476325577e-05, + "loss": 0.6303, + "step": 3246 + }, + { + "epoch": 0.10241286863270778, + "grad_norm": 5.400300025939941, + "learning_rate": 1.9726666315539515e-05, + "loss": 0.6077, + "step": 3247 + }, + { + "epoch": 0.1024444093991484, + "grad_norm": 4.278470993041992, + "learning_rate": 1.972642905324813e-05, + "loss": 0.6257, + "step": 3248 + }, + { + "epoch": 0.10247595016558902, + "grad_norm": 2.5255789756774902, + "learning_rate": 1.9726191689453912e-05, + "loss": 0.3785, + "step": 3249 + }, + { + "epoch": 0.10250749093202965, + "grad_norm": 4.433877944946289, + "learning_rate": 1.972595422415933e-05, + "loss": 0.6203, + "step": 3250 + }, + { + "epoch": 0.10253903169847027, + "grad_norm": 4.542364120483398, + "learning_rate": 1.972571665736686e-05, + "loss": 0.6034, + "step": 3251 + }, + { + "epoch": 0.1025705724649109, + "grad_norm": 6.133361339569092, + "learning_rate": 1.972547898907899e-05, + "loss": 0.6127, + "step": 3252 + }, + { + "epoch": 0.10260211323135152, + "grad_norm": 4.9191107749938965, + "learning_rate": 1.9725241219298193e-05, + "loss": 0.6117, + "step": 3253 + }, + { + "epoch": 0.10263365399779215, + "grad_norm": 4.512722492218018, + "learning_rate": 1.9725003348026952e-05, + "loss": 0.6071, + "step": 3254 + }, + { + "epoch": 0.10266519476423278, + "grad_norm": 5.618770122528076, + "learning_rate": 1.9724765375267748e-05, + "loss": 0.6233, + "step": 3255 + }, + { + "epoch": 0.10269673553067339, + "grad_norm": 3.814116954803467, + "learning_rate": 1.972452730102307e-05, + "loss": 0.6068, + "step": 3256 + }, + { + "epoch": 0.10272827629711402, + "grad_norm": 12.373327255249023, + "learning_rate": 1.9724289125295397e-05, + "loss": 0.6387, + "step": 3257 + }, + { + "epoch": 0.10275981706355465, + "grad_norm": 4.020261287689209, + "learning_rate": 1.972405084808722e-05, + "loss": 0.6053, + "step": 3258 + }, + { + "epoch": 0.10279135782999527, + "grad_norm": 12.63560676574707, + "learning_rate": 1.9723812469401013e-05, + "loss": 0.5768, + "step": 3259 + }, + { + "epoch": 0.1028228985964359, + "grad_norm": 3.240635395050049, + "learning_rate": 1.972357398923928e-05, + "loss": 0.8967, + "step": 3260 + }, + { + "epoch": 0.10285443936287651, + "grad_norm": 4.339036464691162, + "learning_rate": 1.97233354076045e-05, + "loss": 0.6234, + "step": 3261 + }, + { + "epoch": 0.10288598012931714, + "grad_norm": 4.524228096008301, + "learning_rate": 1.9723096724499165e-05, + "loss": 0.6109, + "step": 3262 + }, + { + "epoch": 0.10291752089575777, + "grad_norm": 3.5003445148468018, + "learning_rate": 1.9722857939925766e-05, + "loss": 0.6549, + "step": 3263 + }, + { + "epoch": 0.10294906166219839, + "grad_norm": 3.2171928882598877, + "learning_rate": 1.9722619053886795e-05, + "loss": 0.6059, + "step": 3264 + }, + { + "epoch": 0.10298060242863902, + "grad_norm": 5.218389511108398, + "learning_rate": 1.972238006638474e-05, + "loss": 0.5976, + "step": 3265 + }, + { + "epoch": 0.10301214319507965, + "grad_norm": 5.208560466766357, + "learning_rate": 1.9722140977422107e-05, + "loss": 0.6458, + "step": 3266 + }, + { + "epoch": 0.10304368396152026, + "grad_norm": 3.984276056289673, + "learning_rate": 1.9721901787001382e-05, + "loss": 0.6177, + "step": 3267 + }, + { + "epoch": 0.10307522472796089, + "grad_norm": 6.123051166534424, + "learning_rate": 1.972166249512506e-05, + "loss": 0.5774, + "step": 3268 + }, + { + "epoch": 0.10310676549440151, + "grad_norm": 6.230579853057861, + "learning_rate": 1.9721423101795644e-05, + "loss": 0.5845, + "step": 3269 + }, + { + "epoch": 0.10313830626084214, + "grad_norm": 5.034628391265869, + "learning_rate": 1.9721183607015628e-05, + "loss": 0.6327, + "step": 3270 + }, + { + "epoch": 0.10316984702728277, + "grad_norm": 3.129819631576538, + "learning_rate": 1.9720944010787513e-05, + "loss": 0.9022, + "step": 3271 + }, + { + "epoch": 0.10320138779372338, + "grad_norm": 3.518026113510132, + "learning_rate": 1.97207043131138e-05, + "loss": 0.5913, + "step": 3272 + }, + { + "epoch": 0.10323292856016401, + "grad_norm": 4.713002681732178, + "learning_rate": 1.9720464513996987e-05, + "loss": 0.6252, + "step": 3273 + }, + { + "epoch": 0.10326446932660463, + "grad_norm": 3.758434772491455, + "learning_rate": 1.9720224613439584e-05, + "loss": 0.6433, + "step": 3274 + }, + { + "epoch": 0.10329601009304526, + "grad_norm": 3.4316327571868896, + "learning_rate": 1.9719984611444086e-05, + "loss": 0.6133, + "step": 3275 + }, + { + "epoch": 0.10332755085948589, + "grad_norm": 3.7291998863220215, + "learning_rate": 1.9719744508013e-05, + "loss": 0.6355, + "step": 3276 + }, + { + "epoch": 0.1033590916259265, + "grad_norm": 4.757803440093994, + "learning_rate": 1.9719504303148834e-05, + "loss": 0.6668, + "step": 3277 + }, + { + "epoch": 0.10339063239236713, + "grad_norm": 2.29382061958313, + "learning_rate": 1.9719263996854097e-05, + "loss": 0.4027, + "step": 3278 + }, + { + "epoch": 0.10342217315880776, + "grad_norm": 3.4879047870635986, + "learning_rate": 1.971902358913129e-05, + "loss": 0.5996, + "step": 3279 + }, + { + "epoch": 0.10345371392524838, + "grad_norm": 3.617823362350464, + "learning_rate": 1.9718783079982928e-05, + "loss": 0.5792, + "step": 3280 + }, + { + "epoch": 0.10348525469168901, + "grad_norm": 3.1894209384918213, + "learning_rate": 1.9718542469411513e-05, + "loss": 0.5993, + "step": 3281 + }, + { + "epoch": 0.10351679545812963, + "grad_norm": 2.5931754112243652, + "learning_rate": 1.9718301757419565e-05, + "loss": 0.8734, + "step": 3282 + }, + { + "epoch": 0.10354833622457026, + "grad_norm": 4.37739372253418, + "learning_rate": 1.971806094400959e-05, + "loss": 0.6226, + "step": 3283 + }, + { + "epoch": 0.10357987699101089, + "grad_norm": 4.042413711547852, + "learning_rate": 1.9717820029184108e-05, + "loss": 0.3878, + "step": 3284 + }, + { + "epoch": 0.1036114177574515, + "grad_norm": 3.017930746078491, + "learning_rate": 1.9717579012945623e-05, + "loss": 0.6318, + "step": 3285 + }, + { + "epoch": 0.10364295852389213, + "grad_norm": 5.0566086769104, + "learning_rate": 1.9717337895296657e-05, + "loss": 0.6451, + "step": 3286 + }, + { + "epoch": 0.10367449929033276, + "grad_norm": 3.036756992340088, + "learning_rate": 1.9717096676239725e-05, + "loss": 0.6513, + "step": 3287 + }, + { + "epoch": 0.10370604005677338, + "grad_norm": 3.6919548511505127, + "learning_rate": 1.9716855355777346e-05, + "loss": 0.8939, + "step": 3288 + }, + { + "epoch": 0.103737580823214, + "grad_norm": 2.369546890258789, + "learning_rate": 1.9716613933912035e-05, + "loss": 0.3787, + "step": 3289 + }, + { + "epoch": 0.10376912158965462, + "grad_norm": 2.3782196044921875, + "learning_rate": 1.971637241064631e-05, + "loss": 0.8959, + "step": 3290 + }, + { + "epoch": 0.10380066235609525, + "grad_norm": 6.318288803100586, + "learning_rate": 1.9716130785982696e-05, + "loss": 0.6699, + "step": 3291 + }, + { + "epoch": 0.10383220312253588, + "grad_norm": 4.926566123962402, + "learning_rate": 1.9715889059923714e-05, + "loss": 0.6324, + "step": 3292 + }, + { + "epoch": 0.1038637438889765, + "grad_norm": 3.318007469177246, + "learning_rate": 1.9715647232471885e-05, + "loss": 0.6168, + "step": 3293 + }, + { + "epoch": 0.10389528465541713, + "grad_norm": 2.0139918327331543, + "learning_rate": 1.9715405303629732e-05, + "loss": 0.9092, + "step": 3294 + }, + { + "epoch": 0.10392682542185776, + "grad_norm": 3.9500341415405273, + "learning_rate": 1.971516327339978e-05, + "loss": 0.6463, + "step": 3295 + }, + { + "epoch": 0.10395836618829837, + "grad_norm": 2.3498289585113525, + "learning_rate": 1.971492114178456e-05, + "loss": 0.8688, + "step": 3296 + }, + { + "epoch": 0.103989906954739, + "grad_norm": 4.061992168426514, + "learning_rate": 1.971467890878659e-05, + "loss": 0.6422, + "step": 3297 + }, + { + "epoch": 0.10402144772117962, + "grad_norm": 4.1320013999938965, + "learning_rate": 1.9714436574408408e-05, + "loss": 0.6732, + "step": 3298 + }, + { + "epoch": 0.10405298848762025, + "grad_norm": 5.0212297439575195, + "learning_rate": 1.9714194138652533e-05, + "loss": 0.6189, + "step": 3299 + }, + { + "epoch": 0.10408452925406088, + "grad_norm": 2.9227614402770996, + "learning_rate": 1.97139516015215e-05, + "loss": 0.431, + "step": 3300 + }, + { + "epoch": 0.1041160700205015, + "grad_norm": 5.286322593688965, + "learning_rate": 1.971370896301784e-05, + "loss": 0.6321, + "step": 3301 + }, + { + "epoch": 0.10414761078694212, + "grad_norm": 4.122631549835205, + "learning_rate": 1.971346622314408e-05, + "loss": 0.6491, + "step": 3302 + }, + { + "epoch": 0.10417915155338275, + "grad_norm": 4.34136962890625, + "learning_rate": 1.9713223381902764e-05, + "loss": 0.6251, + "step": 3303 + }, + { + "epoch": 0.10421069231982337, + "grad_norm": 3.032325267791748, + "learning_rate": 1.9712980439296417e-05, + "loss": 0.6304, + "step": 3304 + }, + { + "epoch": 0.104242233086264, + "grad_norm": 3.0366272926330566, + "learning_rate": 1.9712737395327576e-05, + "loss": 0.602, + "step": 3305 + }, + { + "epoch": 0.10427377385270462, + "grad_norm": 3.3221936225891113, + "learning_rate": 1.971249424999878e-05, + "loss": 0.6382, + "step": 3306 + }, + { + "epoch": 0.10430531461914525, + "grad_norm": 4.409507751464844, + "learning_rate": 1.9712251003312567e-05, + "loss": 0.6543, + "step": 3307 + }, + { + "epoch": 0.10433685538558587, + "grad_norm": 3.8882906436920166, + "learning_rate": 1.971200765527147e-05, + "loss": 0.6468, + "step": 3308 + }, + { + "epoch": 0.10436839615202649, + "grad_norm": 3.720461130142212, + "learning_rate": 1.9711764205878034e-05, + "loss": 0.5881, + "step": 3309 + }, + { + "epoch": 0.10439993691846712, + "grad_norm": 3.9312846660614014, + "learning_rate": 1.9711520655134798e-05, + "loss": 0.6148, + "step": 3310 + }, + { + "epoch": 0.10443147768490774, + "grad_norm": 3.7695586681365967, + "learning_rate": 1.97112770030443e-05, + "loss": 0.6004, + "step": 3311 + }, + { + "epoch": 0.10446301845134837, + "grad_norm": 3.3894095420837402, + "learning_rate": 1.971103324960909e-05, + "loss": 0.6342, + "step": 3312 + }, + { + "epoch": 0.104494559217789, + "grad_norm": 5.064579963684082, + "learning_rate": 1.9710789394831704e-05, + "loss": 0.6357, + "step": 3313 + }, + { + "epoch": 0.10452609998422961, + "grad_norm": 3.604858636856079, + "learning_rate": 1.971054543871469e-05, + "loss": 0.6234, + "step": 3314 + }, + { + "epoch": 0.10455764075067024, + "grad_norm": 3.473078489303589, + "learning_rate": 1.9710301381260598e-05, + "loss": 0.5975, + "step": 3315 + }, + { + "epoch": 0.10458918151711087, + "grad_norm": 3.7919328212738037, + "learning_rate": 1.971005722247197e-05, + "loss": 0.6252, + "step": 3316 + }, + { + "epoch": 0.10462072228355149, + "grad_norm": 5.087071418762207, + "learning_rate": 1.9709812962351352e-05, + "loss": 0.5641, + "step": 3317 + }, + { + "epoch": 0.10465226304999212, + "grad_norm": 11.192788124084473, + "learning_rate": 1.97095686009013e-05, + "loss": 0.5918, + "step": 3318 + }, + { + "epoch": 0.10468380381643273, + "grad_norm": 3.2787673473358154, + "learning_rate": 1.970932413812436e-05, + "loss": 0.622, + "step": 3319 + }, + { + "epoch": 0.10471534458287336, + "grad_norm": 3.405573844909668, + "learning_rate": 1.970907957402308e-05, + "loss": 0.6016, + "step": 3320 + }, + { + "epoch": 0.10474688534931399, + "grad_norm": 3.919565200805664, + "learning_rate": 1.970883490860002e-05, + "loss": 0.6196, + "step": 3321 + }, + { + "epoch": 0.10477842611575461, + "grad_norm": 2.8519022464752197, + "learning_rate": 1.970859014185773e-05, + "loss": 0.6238, + "step": 3322 + }, + { + "epoch": 0.10480996688219524, + "grad_norm": 3.550662040710449, + "learning_rate": 1.970834527379876e-05, + "loss": 0.5974, + "step": 3323 + }, + { + "epoch": 0.10484150764863587, + "grad_norm": 4.52968168258667, + "learning_rate": 1.970810030442567e-05, + "loss": 0.5987, + "step": 3324 + }, + { + "epoch": 0.10487304841507648, + "grad_norm": 2.9869117736816406, + "learning_rate": 1.9707855233741014e-05, + "loss": 0.6296, + "step": 3325 + }, + { + "epoch": 0.10490458918151711, + "grad_norm": 3.0411617755889893, + "learning_rate": 1.970761006174735e-05, + "loss": 0.6455, + "step": 3326 + }, + { + "epoch": 0.10493612994795773, + "grad_norm": 4.319265842437744, + "learning_rate": 1.970736478844724e-05, + "loss": 0.6378, + "step": 3327 + }, + { + "epoch": 0.10496767071439836, + "grad_norm": 3.213347911834717, + "learning_rate": 1.9707119413843237e-05, + "loss": 0.6119, + "step": 3328 + }, + { + "epoch": 0.10499921148083899, + "grad_norm": 3.6627893447875977, + "learning_rate": 1.9706873937937908e-05, + "loss": 0.9088, + "step": 3329 + }, + { + "epoch": 0.1050307522472796, + "grad_norm": 3.8025143146514893, + "learning_rate": 1.9706628360733814e-05, + "loss": 0.5998, + "step": 3330 + }, + { + "epoch": 0.10506229301372023, + "grad_norm": 3.1837961673736572, + "learning_rate": 1.9706382682233515e-05, + "loss": 0.6246, + "step": 3331 + }, + { + "epoch": 0.10509383378016086, + "grad_norm": 3.4935760498046875, + "learning_rate": 1.9706136902439574e-05, + "loss": 0.6103, + "step": 3332 + }, + { + "epoch": 0.10512537454660148, + "grad_norm": 5.678020477294922, + "learning_rate": 1.970589102135456e-05, + "loss": 0.6003, + "step": 3333 + }, + { + "epoch": 0.10515691531304211, + "grad_norm": 2.969203233718872, + "learning_rate": 1.9705645038981035e-05, + "loss": 0.4153, + "step": 3334 + }, + { + "epoch": 0.10518845607948273, + "grad_norm": 3.8207528591156006, + "learning_rate": 1.970539895532157e-05, + "loss": 0.6088, + "step": 3335 + }, + { + "epoch": 0.10521999684592336, + "grad_norm": 4.012478351593018, + "learning_rate": 1.970515277037873e-05, + "loss": 0.5903, + "step": 3336 + }, + { + "epoch": 0.10525153761236399, + "grad_norm": 3.2640180587768555, + "learning_rate": 1.9704906484155083e-05, + "loss": 0.6182, + "step": 3337 + }, + { + "epoch": 0.1052830783788046, + "grad_norm": 3.7155661582946777, + "learning_rate": 1.9704660096653204e-05, + "loss": 0.6543, + "step": 3338 + }, + { + "epoch": 0.10531461914524523, + "grad_norm": 3.938509464263916, + "learning_rate": 1.9704413607875657e-05, + "loss": 0.5945, + "step": 3339 + }, + { + "epoch": 0.10534615991168586, + "grad_norm": 3.800204277038574, + "learning_rate": 1.9704167017825026e-05, + "loss": 0.6149, + "step": 3340 + }, + { + "epoch": 0.10537770067812648, + "grad_norm": 2.7221879959106445, + "learning_rate": 1.970392032650387e-05, + "loss": 0.6229, + "step": 3341 + }, + { + "epoch": 0.1054092414445671, + "grad_norm": 1.820370078086853, + "learning_rate": 1.9703673533914776e-05, + "loss": 0.3593, + "step": 3342 + }, + { + "epoch": 0.10544078221100772, + "grad_norm": 5.913939952850342, + "learning_rate": 1.9703426640060307e-05, + "loss": 0.6184, + "step": 3343 + }, + { + "epoch": 0.10547232297744835, + "grad_norm": 3.5968592166900635, + "learning_rate": 1.9703179644943053e-05, + "loss": 0.5855, + "step": 3344 + }, + { + "epoch": 0.10550386374388898, + "grad_norm": 4.000395774841309, + "learning_rate": 1.9702932548565582e-05, + "loss": 0.6068, + "step": 3345 + }, + { + "epoch": 0.1055354045103296, + "grad_norm": 3.31966233253479, + "learning_rate": 1.9702685350930476e-05, + "loss": 0.5993, + "step": 3346 + }, + { + "epoch": 0.10556694527677023, + "grad_norm": 3.0945651531219482, + "learning_rate": 1.9702438052040315e-05, + "loss": 0.5893, + "step": 3347 + }, + { + "epoch": 0.10559848604321086, + "grad_norm": 3.002568483352661, + "learning_rate": 1.970219065189768e-05, + "loss": 0.5893, + "step": 3348 + }, + { + "epoch": 0.10563002680965147, + "grad_norm": 3.779029607772827, + "learning_rate": 1.970194315050515e-05, + "loss": 0.6246, + "step": 3349 + }, + { + "epoch": 0.1056615675760921, + "grad_norm": 5.001846790313721, + "learning_rate": 1.9701695547865313e-05, + "loss": 0.6147, + "step": 3350 + }, + { + "epoch": 0.10569310834253272, + "grad_norm": 3.767299175262451, + "learning_rate": 1.9701447843980748e-05, + "loss": 0.6172, + "step": 3351 + }, + { + "epoch": 0.10572464910897335, + "grad_norm": 3.2490828037261963, + "learning_rate": 1.9701200038854042e-05, + "loss": 0.5644, + "step": 3352 + }, + { + "epoch": 0.10575618987541398, + "grad_norm": 3.899658203125, + "learning_rate": 1.970095213248778e-05, + "loss": 0.5824, + "step": 3353 + }, + { + "epoch": 0.1057877306418546, + "grad_norm": 3.922585964202881, + "learning_rate": 1.9700704124884553e-05, + "loss": 0.6611, + "step": 3354 + }, + { + "epoch": 0.10581927140829522, + "grad_norm": 3.0501630306243896, + "learning_rate": 1.970045601604694e-05, + "loss": 0.9326, + "step": 3355 + }, + { + "epoch": 0.10585081217473584, + "grad_norm": 4.11614990234375, + "learning_rate": 1.970020780597754e-05, + "loss": 0.6169, + "step": 3356 + }, + { + "epoch": 0.10588235294117647, + "grad_norm": 3.6002633571624756, + "learning_rate": 1.969995949467894e-05, + "loss": 0.8966, + "step": 3357 + }, + { + "epoch": 0.1059138937076171, + "grad_norm": 4.496130466461182, + "learning_rate": 1.969971108215373e-05, + "loss": 0.6657, + "step": 3358 + }, + { + "epoch": 0.10594543447405771, + "grad_norm": 3.7954044342041016, + "learning_rate": 1.96994625684045e-05, + "loss": 0.6074, + "step": 3359 + }, + { + "epoch": 0.10597697524049834, + "grad_norm": 4.170944690704346, + "learning_rate": 1.9699213953433853e-05, + "loss": 0.6317, + "step": 3360 + }, + { + "epoch": 0.10600851600693897, + "grad_norm": 2.815412759780884, + "learning_rate": 1.9698965237244374e-05, + "loss": 0.6088, + "step": 3361 + }, + { + "epoch": 0.10604005677337959, + "grad_norm": 2.6013355255126953, + "learning_rate": 1.9698716419838662e-05, + "loss": 0.3873, + "step": 3362 + }, + { + "epoch": 0.10607159753982022, + "grad_norm": 4.29664421081543, + "learning_rate": 1.9698467501219313e-05, + "loss": 0.6241, + "step": 3363 + }, + { + "epoch": 0.10610313830626084, + "grad_norm": 4.979831695556641, + "learning_rate": 1.9698218481388925e-05, + "loss": 0.6511, + "step": 3364 + }, + { + "epoch": 0.10613467907270147, + "grad_norm": 4.430582523345947, + "learning_rate": 1.9697969360350098e-05, + "loss": 0.628, + "step": 3365 + }, + { + "epoch": 0.1061662198391421, + "grad_norm": 3.2685935497283936, + "learning_rate": 1.9697720138105426e-05, + "loss": 0.6182, + "step": 3366 + }, + { + "epoch": 0.10619776060558271, + "grad_norm": 3.1258955001831055, + "learning_rate": 1.969747081465752e-05, + "loss": 0.6001, + "step": 3367 + }, + { + "epoch": 0.10622930137202334, + "grad_norm": 6.661875247955322, + "learning_rate": 1.969722139000897e-05, + "loss": 0.586, + "step": 3368 + }, + { + "epoch": 0.10626084213846397, + "grad_norm": 3.8944790363311768, + "learning_rate": 1.969697186416239e-05, + "loss": 0.6622, + "step": 3369 + }, + { + "epoch": 0.10629238290490459, + "grad_norm": 4.391369342803955, + "learning_rate": 1.969672223712038e-05, + "loss": 0.6558, + "step": 3370 + }, + { + "epoch": 0.10632392367134522, + "grad_norm": 4.422688961029053, + "learning_rate": 1.9696472508885543e-05, + "loss": 0.6115, + "step": 3371 + }, + { + "epoch": 0.10635546443778583, + "grad_norm": 3.141519784927368, + "learning_rate": 1.9696222679460485e-05, + "loss": 0.6429, + "step": 3372 + }, + { + "epoch": 0.10638700520422646, + "grad_norm": 3.250589370727539, + "learning_rate": 1.969597274884781e-05, + "loss": 0.6205, + "step": 3373 + }, + { + "epoch": 0.10641854597066709, + "grad_norm": 3.297146797180176, + "learning_rate": 1.9695722717050136e-05, + "loss": 0.5997, + "step": 3374 + }, + { + "epoch": 0.10645008673710771, + "grad_norm": 4.910211563110352, + "learning_rate": 1.969547258407007e-05, + "loss": 0.965, + "step": 3375 + }, + { + "epoch": 0.10648162750354834, + "grad_norm": 4.757205009460449, + "learning_rate": 1.9695222349910216e-05, + "loss": 0.6289, + "step": 3376 + }, + { + "epoch": 0.10651316826998897, + "grad_norm": 2.7825944423675537, + "learning_rate": 1.9694972014573185e-05, + "loss": 0.4179, + "step": 3377 + }, + { + "epoch": 0.10654470903642958, + "grad_norm": 4.0863142013549805, + "learning_rate": 1.9694721578061596e-05, + "loss": 0.6069, + "step": 3378 + }, + { + "epoch": 0.10657624980287021, + "grad_norm": 6.848991870880127, + "learning_rate": 1.969447104037806e-05, + "loss": 0.5903, + "step": 3379 + }, + { + "epoch": 0.10660779056931083, + "grad_norm": 4.003800392150879, + "learning_rate": 1.969422040152519e-05, + "loss": 0.6052, + "step": 3380 + }, + { + "epoch": 0.10663933133575146, + "grad_norm": 2.8849167823791504, + "learning_rate": 1.9693969661505603e-05, + "loss": 0.5988, + "step": 3381 + }, + { + "epoch": 0.10667087210219209, + "grad_norm": 3.3288915157318115, + "learning_rate": 1.9693718820321916e-05, + "loss": 0.6172, + "step": 3382 + }, + { + "epoch": 0.1067024128686327, + "grad_norm": 3.527560234069824, + "learning_rate": 1.9693467877976745e-05, + "loss": 0.5825, + "step": 3383 + }, + { + "epoch": 0.10673395363507333, + "grad_norm": 4.040355205535889, + "learning_rate": 1.9693216834472713e-05, + "loss": 0.6208, + "step": 3384 + }, + { + "epoch": 0.10676549440151396, + "grad_norm": 4.188865661621094, + "learning_rate": 1.9692965689812435e-05, + "loss": 0.602, + "step": 3385 + }, + { + "epoch": 0.10679703516795458, + "grad_norm": 3.2894625663757324, + "learning_rate": 1.9692714443998534e-05, + "loss": 0.6218, + "step": 3386 + }, + { + "epoch": 0.10682857593439521, + "grad_norm": 4.110569000244141, + "learning_rate": 1.969246309703363e-05, + "loss": 0.5809, + "step": 3387 + }, + { + "epoch": 0.10686011670083582, + "grad_norm": 3.0287466049194336, + "learning_rate": 1.969221164892035e-05, + "loss": 0.5808, + "step": 3388 + }, + { + "epoch": 0.10689165746727645, + "grad_norm": 4.076018333435059, + "learning_rate": 1.9691960099661315e-05, + "loss": 0.6569, + "step": 3389 + }, + { + "epoch": 0.10692319823371708, + "grad_norm": 3.4830379486083984, + "learning_rate": 1.969170844925915e-05, + "loss": 0.6234, + "step": 3390 + }, + { + "epoch": 0.1069547390001577, + "grad_norm": 4.799561023712158, + "learning_rate": 1.9691456697716477e-05, + "loss": 0.6274, + "step": 3391 + }, + { + "epoch": 0.10698627976659833, + "grad_norm": 4.396290302276611, + "learning_rate": 1.9691204845035935e-05, + "loss": 0.9378, + "step": 3392 + }, + { + "epoch": 0.10701782053303895, + "grad_norm": 4.792314052581787, + "learning_rate": 1.9690952891220143e-05, + "loss": 0.6109, + "step": 3393 + }, + { + "epoch": 0.10704936129947958, + "grad_norm": 2.2044880390167236, + "learning_rate": 1.9690700836271734e-05, + "loss": 0.401, + "step": 3394 + }, + { + "epoch": 0.1070809020659202, + "grad_norm": 2.3290936946868896, + "learning_rate": 1.9690448680193337e-05, + "loss": 0.4326, + "step": 3395 + }, + { + "epoch": 0.10711244283236082, + "grad_norm": 4.234010219573975, + "learning_rate": 1.9690196422987582e-05, + "loss": 0.64, + "step": 3396 + }, + { + "epoch": 0.10714398359880145, + "grad_norm": 3.4707930088043213, + "learning_rate": 1.9689944064657104e-05, + "loss": 0.8801, + "step": 3397 + }, + { + "epoch": 0.10717552436524208, + "grad_norm": 3.2424581050872803, + "learning_rate": 1.9689691605204534e-05, + "loss": 0.6242, + "step": 3398 + }, + { + "epoch": 0.1072070651316827, + "grad_norm": 3.6368625164031982, + "learning_rate": 1.968943904463251e-05, + "loss": 0.6251, + "step": 3399 + }, + { + "epoch": 0.10723860589812333, + "grad_norm": 3.9670557975769043, + "learning_rate": 1.968918638294367e-05, + "loss": 0.6241, + "step": 3400 + }, + { + "epoch": 0.10727014666456394, + "grad_norm": 3.9732744693756104, + "learning_rate": 1.9688933620140638e-05, + "loss": 0.6304, + "step": 3401 + }, + { + "epoch": 0.10730168743100457, + "grad_norm": 3.942704916000366, + "learning_rate": 1.9688680756226067e-05, + "loss": 0.6329, + "step": 3402 + }, + { + "epoch": 0.1073332281974452, + "grad_norm": 3.9507391452789307, + "learning_rate": 1.9688427791202587e-05, + "loss": 0.6434, + "step": 3403 + }, + { + "epoch": 0.10736476896388582, + "grad_norm": 3.560891628265381, + "learning_rate": 1.968817472507284e-05, + "loss": 0.6313, + "step": 3404 + }, + { + "epoch": 0.10739630973032645, + "grad_norm": 3.2141170501708984, + "learning_rate": 1.9687921557839466e-05, + "loss": 0.5513, + "step": 3405 + }, + { + "epoch": 0.10742785049676708, + "grad_norm": 3.3664751052856445, + "learning_rate": 1.968766828950511e-05, + "loss": 0.6508, + "step": 3406 + }, + { + "epoch": 0.1074593912632077, + "grad_norm": 4.842483997344971, + "learning_rate": 1.968741492007241e-05, + "loss": 0.9579, + "step": 3407 + }, + { + "epoch": 0.10749093202964832, + "grad_norm": 4.981634140014648, + "learning_rate": 1.9687161449544017e-05, + "loss": 0.6026, + "step": 3408 + }, + { + "epoch": 0.10752247279608894, + "grad_norm": 6.8173828125, + "learning_rate": 1.968690787792257e-05, + "loss": 0.6324, + "step": 3409 + }, + { + "epoch": 0.10755401356252957, + "grad_norm": 2.3140146732330322, + "learning_rate": 1.968665420521072e-05, + "loss": 0.4251, + "step": 3410 + }, + { + "epoch": 0.1075855543289702, + "grad_norm": 4.548435688018799, + "learning_rate": 1.968640043141111e-05, + "loss": 0.5885, + "step": 3411 + }, + { + "epoch": 0.10761709509541081, + "grad_norm": 5.023941516876221, + "learning_rate": 1.968614655652639e-05, + "loss": 0.6635, + "step": 3412 + }, + { + "epoch": 0.10764863586185144, + "grad_norm": 3.0296759605407715, + "learning_rate": 1.968589258055921e-05, + "loss": 0.6107, + "step": 3413 + }, + { + "epoch": 0.10768017662829207, + "grad_norm": 6.37013578414917, + "learning_rate": 1.968563850351222e-05, + "loss": 0.3944, + "step": 3414 + }, + { + "epoch": 0.10771171739473269, + "grad_norm": 3.4717562198638916, + "learning_rate": 1.9685384325388072e-05, + "loss": 0.624, + "step": 3415 + }, + { + "epoch": 0.10774325816117332, + "grad_norm": 3.5299479961395264, + "learning_rate": 1.968513004618942e-05, + "loss": 0.6009, + "step": 3416 + }, + { + "epoch": 0.10777479892761394, + "grad_norm": 5.463578701019287, + "learning_rate": 1.9684875665918916e-05, + "loss": 0.6022, + "step": 3417 + }, + { + "epoch": 0.10780633969405456, + "grad_norm": 4.386422157287598, + "learning_rate": 1.9684621184579208e-05, + "loss": 0.3828, + "step": 3418 + }, + { + "epoch": 0.1078378804604952, + "grad_norm": 6.949289321899414, + "learning_rate": 1.9684366602172962e-05, + "loss": 0.6337, + "step": 3419 + }, + { + "epoch": 0.10786942122693581, + "grad_norm": 2.990694046020508, + "learning_rate": 1.968411191870283e-05, + "loss": 0.5876, + "step": 3420 + }, + { + "epoch": 0.10790096199337644, + "grad_norm": 6.494450569152832, + "learning_rate": 1.9683857134171474e-05, + "loss": 0.631, + "step": 3421 + }, + { + "epoch": 0.10793250275981707, + "grad_norm": 9.209872245788574, + "learning_rate": 1.9683602248581545e-05, + "loss": 0.4118, + "step": 3422 + }, + { + "epoch": 0.10796404352625769, + "grad_norm": 2.948955774307251, + "learning_rate": 1.9683347261935707e-05, + "loss": 0.6243, + "step": 3423 + }, + { + "epoch": 0.10799558429269832, + "grad_norm": 3.6293766498565674, + "learning_rate": 1.9683092174236628e-05, + "loss": 0.391, + "step": 3424 + }, + { + "epoch": 0.10802712505913893, + "grad_norm": 3.6258113384246826, + "learning_rate": 1.9682836985486956e-05, + "loss": 0.6335, + "step": 3425 + }, + { + "epoch": 0.10805866582557956, + "grad_norm": 4.148215293884277, + "learning_rate": 1.9682581695689364e-05, + "loss": 0.6009, + "step": 3426 + }, + { + "epoch": 0.10809020659202019, + "grad_norm": 5.294444561004639, + "learning_rate": 1.9682326304846518e-05, + "loss": 0.5896, + "step": 3427 + }, + { + "epoch": 0.1081217473584608, + "grad_norm": 5.2993245124816895, + "learning_rate": 1.9682070812961074e-05, + "loss": 0.6061, + "step": 3428 + }, + { + "epoch": 0.10815328812490144, + "grad_norm": 2.8798987865448, + "learning_rate": 1.9681815220035703e-05, + "loss": 0.5897, + "step": 3429 + }, + { + "epoch": 0.10818482889134207, + "grad_norm": 3.2866909503936768, + "learning_rate": 1.9681559526073074e-05, + "loss": 0.9303, + "step": 3430 + }, + { + "epoch": 0.10821636965778268, + "grad_norm": 3.266766309738159, + "learning_rate": 1.9681303731075854e-05, + "loss": 0.6087, + "step": 3431 + }, + { + "epoch": 0.10824791042422331, + "grad_norm": 11.885046005249023, + "learning_rate": 1.9681047835046708e-05, + "loss": 0.599, + "step": 3432 + }, + { + "epoch": 0.10827945119066393, + "grad_norm": 4.116020679473877, + "learning_rate": 1.9680791837988318e-05, + "loss": 0.6043, + "step": 3433 + }, + { + "epoch": 0.10831099195710456, + "grad_norm": 4.554950714111328, + "learning_rate": 1.9680535739903342e-05, + "loss": 0.6242, + "step": 3434 + }, + { + "epoch": 0.10834253272354519, + "grad_norm": 3.0786845684051514, + "learning_rate": 1.9680279540794465e-05, + "loss": 0.579, + "step": 3435 + }, + { + "epoch": 0.1083740734899858, + "grad_norm": 2.3896706104278564, + "learning_rate": 1.968002324066435e-05, + "loss": 0.9002, + "step": 3436 + }, + { + "epoch": 0.10840561425642643, + "grad_norm": 4.053772449493408, + "learning_rate": 1.9679766839515677e-05, + "loss": 0.6153, + "step": 3437 + }, + { + "epoch": 0.10843715502286705, + "grad_norm": 3.600041389465332, + "learning_rate": 1.967951033735112e-05, + "loss": 0.9274, + "step": 3438 + }, + { + "epoch": 0.10846869578930768, + "grad_norm": 4.3202924728393555, + "learning_rate": 1.9679253734173357e-05, + "loss": 0.5992, + "step": 3439 + }, + { + "epoch": 0.10850023655574831, + "grad_norm": 4.364053249359131, + "learning_rate": 1.967899702998507e-05, + "loss": 0.6109, + "step": 3440 + }, + { + "epoch": 0.10853177732218892, + "grad_norm": 5.452843189239502, + "learning_rate": 1.9678740224788925e-05, + "loss": 0.6079, + "step": 3441 + }, + { + "epoch": 0.10856331808862955, + "grad_norm": 3.8127777576446533, + "learning_rate": 1.9678483318587616e-05, + "loss": 0.6047, + "step": 3442 + }, + { + "epoch": 0.10859485885507018, + "grad_norm": 3.5168888568878174, + "learning_rate": 1.9678226311383818e-05, + "loss": 0.5952, + "step": 3443 + }, + { + "epoch": 0.1086263996215108, + "grad_norm": 6.15921688079834, + "learning_rate": 1.9677969203180214e-05, + "loss": 0.6, + "step": 3444 + }, + { + "epoch": 0.10865794038795143, + "grad_norm": 2.7365427017211914, + "learning_rate": 1.9677711993979485e-05, + "loss": 0.3745, + "step": 3445 + }, + { + "epoch": 0.10868948115439205, + "grad_norm": 4.121635913848877, + "learning_rate": 1.9677454683784315e-05, + "loss": 0.6177, + "step": 3446 + }, + { + "epoch": 0.10872102192083268, + "grad_norm": 4.011096477508545, + "learning_rate": 1.9677197272597395e-05, + "loss": 0.6066, + "step": 3447 + }, + { + "epoch": 0.1087525626872733, + "grad_norm": 3.5187368392944336, + "learning_rate": 1.9676939760421404e-05, + "loss": 0.583, + "step": 3448 + }, + { + "epoch": 0.10878410345371392, + "grad_norm": 5.734733581542969, + "learning_rate": 1.9676682147259038e-05, + "loss": 0.6281, + "step": 3449 + }, + { + "epoch": 0.10881564422015455, + "grad_norm": 4.090343952178955, + "learning_rate": 1.9676424433112973e-05, + "loss": 0.5833, + "step": 3450 + }, + { + "epoch": 0.10884718498659518, + "grad_norm": 2.8297951221466064, + "learning_rate": 1.9676166617985907e-05, + "loss": 0.3909, + "step": 3451 + }, + { + "epoch": 0.1088787257530358, + "grad_norm": 2.33909010887146, + "learning_rate": 1.9675908701880534e-05, + "loss": 0.8959, + "step": 3452 + }, + { + "epoch": 0.10891026651947643, + "grad_norm": 3.005246877670288, + "learning_rate": 1.9675650684799536e-05, + "loss": 0.9266, + "step": 3453 + }, + { + "epoch": 0.10894180728591704, + "grad_norm": 3.9254722595214844, + "learning_rate": 1.967539256674561e-05, + "loss": 0.6087, + "step": 3454 + }, + { + "epoch": 0.10897334805235767, + "grad_norm": 2.2145423889160156, + "learning_rate": 1.9675134347721455e-05, + "loss": 0.3893, + "step": 3455 + }, + { + "epoch": 0.1090048888187983, + "grad_norm": 28.303085327148438, + "learning_rate": 1.9674876027729756e-05, + "loss": 0.6168, + "step": 3456 + }, + { + "epoch": 0.10903642958523892, + "grad_norm": 2.5936474800109863, + "learning_rate": 1.9674617606773213e-05, + "loss": 0.8734, + "step": 3457 + }, + { + "epoch": 0.10906797035167955, + "grad_norm": 4.961180686950684, + "learning_rate": 1.9674359084854524e-05, + "loss": 0.604, + "step": 3458 + }, + { + "epoch": 0.10909951111812018, + "grad_norm": 3.5252184867858887, + "learning_rate": 1.967410046197639e-05, + "loss": 0.3553, + "step": 3459 + }, + { + "epoch": 0.10913105188456079, + "grad_norm": 3.5454859733581543, + "learning_rate": 1.96738417381415e-05, + "loss": 0.6243, + "step": 3460 + }, + { + "epoch": 0.10916259265100142, + "grad_norm": 3.4161376953125, + "learning_rate": 1.967358291335256e-05, + "loss": 0.6144, + "step": 3461 + }, + { + "epoch": 0.10919413341744204, + "grad_norm": 3.3761866092681885, + "learning_rate": 1.9673323987612278e-05, + "loss": 0.6258, + "step": 3462 + }, + { + "epoch": 0.10922567418388267, + "grad_norm": 4.2196574211120605, + "learning_rate": 1.9673064960923342e-05, + "loss": 0.3723, + "step": 3463 + }, + { + "epoch": 0.1092572149503233, + "grad_norm": 3.5599656105041504, + "learning_rate": 1.9672805833288465e-05, + "loss": 0.6016, + "step": 3464 + }, + { + "epoch": 0.10928875571676391, + "grad_norm": 3.190812110900879, + "learning_rate": 1.9672546604710344e-05, + "loss": 0.6201, + "step": 3465 + }, + { + "epoch": 0.10932029648320454, + "grad_norm": 25.416358947753906, + "learning_rate": 1.9672287275191696e-05, + "loss": 0.8842, + "step": 3466 + }, + { + "epoch": 0.10935183724964517, + "grad_norm": 4.1706414222717285, + "learning_rate": 1.9672027844735212e-05, + "loss": 0.5972, + "step": 3467 + }, + { + "epoch": 0.10938337801608579, + "grad_norm": 5.706719875335693, + "learning_rate": 1.9671768313343612e-05, + "loss": 0.626, + "step": 3468 + }, + { + "epoch": 0.10941491878252642, + "grad_norm": 4.24513053894043, + "learning_rate": 1.96715086810196e-05, + "loss": 0.5931, + "step": 3469 + }, + { + "epoch": 0.10944645954896703, + "grad_norm": 3.4157562255859375, + "learning_rate": 1.9671248947765887e-05, + "loss": 0.5902, + "step": 3470 + }, + { + "epoch": 0.10947800031540766, + "grad_norm": 3.465878963470459, + "learning_rate": 1.967098911358518e-05, + "loss": 0.5978, + "step": 3471 + }, + { + "epoch": 0.1095095410818483, + "grad_norm": 3.8083927631378174, + "learning_rate": 1.967072917848019e-05, + "loss": 0.597, + "step": 3472 + }, + { + "epoch": 0.10954108184828891, + "grad_norm": 2.6375555992126465, + "learning_rate": 1.9670469142453634e-05, + "loss": 0.908, + "step": 3473 + }, + { + "epoch": 0.10957262261472954, + "grad_norm": 3.03810453414917, + "learning_rate": 1.9670209005508223e-05, + "loss": 0.5931, + "step": 3474 + }, + { + "epoch": 0.10960416338117017, + "grad_norm": 3.121899366378784, + "learning_rate": 1.966994876764667e-05, + "loss": 0.5843, + "step": 3475 + }, + { + "epoch": 0.10963570414761079, + "grad_norm": 3.600433588027954, + "learning_rate": 1.9669688428871697e-05, + "loss": 0.9123, + "step": 3476 + }, + { + "epoch": 0.10966724491405141, + "grad_norm": 2.920789957046509, + "learning_rate": 1.9669427989186017e-05, + "loss": 0.5963, + "step": 3477 + }, + { + "epoch": 0.10969878568049203, + "grad_norm": 3.991276979446411, + "learning_rate": 1.9669167448592343e-05, + "loss": 0.6111, + "step": 3478 + }, + { + "epoch": 0.10973032644693266, + "grad_norm": 16.834287643432617, + "learning_rate": 1.9668906807093403e-05, + "loss": 0.3906, + "step": 3479 + }, + { + "epoch": 0.10976186721337329, + "grad_norm": 2.7377688884735107, + "learning_rate": 1.9668646064691914e-05, + "loss": 0.5861, + "step": 3480 + }, + { + "epoch": 0.1097934079798139, + "grad_norm": 3.934089422225952, + "learning_rate": 1.966838522139059e-05, + "loss": 0.5985, + "step": 3481 + }, + { + "epoch": 0.10982494874625454, + "grad_norm": 4.480434417724609, + "learning_rate": 1.9668124277192163e-05, + "loss": 0.8968, + "step": 3482 + }, + { + "epoch": 0.10985648951269515, + "grad_norm": 3.1653687953948975, + "learning_rate": 1.966786323209935e-05, + "loss": 0.5636, + "step": 3483 + }, + { + "epoch": 0.10988803027913578, + "grad_norm": 5.558254241943359, + "learning_rate": 1.9667602086114882e-05, + "loss": 0.6185, + "step": 3484 + }, + { + "epoch": 0.10991957104557641, + "grad_norm": 3.1517133712768555, + "learning_rate": 1.9667340839241474e-05, + "loss": 0.6238, + "step": 3485 + }, + { + "epoch": 0.10995111181201703, + "grad_norm": 3.367286443710327, + "learning_rate": 1.966707949148186e-05, + "loss": 0.6303, + "step": 3486 + }, + { + "epoch": 0.10998265257845766, + "grad_norm": 3.049639940261841, + "learning_rate": 1.9666818042838765e-05, + "loss": 0.5881, + "step": 3487 + }, + { + "epoch": 0.11001419334489829, + "grad_norm": 7.083939075469971, + "learning_rate": 1.9666556493314918e-05, + "loss": 0.5939, + "step": 3488 + }, + { + "epoch": 0.1100457341113389, + "grad_norm": 3.78189754486084, + "learning_rate": 1.966629484291305e-05, + "loss": 0.5834, + "step": 3489 + }, + { + "epoch": 0.11007727487777953, + "grad_norm": 4.234299182891846, + "learning_rate": 1.9666033091635885e-05, + "loss": 0.5856, + "step": 3490 + }, + { + "epoch": 0.11010881564422015, + "grad_norm": 3.8346054553985596, + "learning_rate": 1.9665771239486163e-05, + "loss": 0.5897, + "step": 3491 + }, + { + "epoch": 0.11014035641066078, + "grad_norm": 3.033430576324463, + "learning_rate": 1.9665509286466612e-05, + "loss": 0.6546, + "step": 3492 + }, + { + "epoch": 0.11017189717710141, + "grad_norm": 4.0028886795043945, + "learning_rate": 1.9665247232579965e-05, + "loss": 0.6019, + "step": 3493 + }, + { + "epoch": 0.11020343794354202, + "grad_norm": 5.351646900177002, + "learning_rate": 1.966498507782896e-05, + "loss": 0.6008, + "step": 3494 + }, + { + "epoch": 0.11023497870998265, + "grad_norm": 2.9248156547546387, + "learning_rate": 1.966472282221633e-05, + "loss": 0.6362, + "step": 3495 + }, + { + "epoch": 0.11026651947642328, + "grad_norm": 3.9034433364868164, + "learning_rate": 1.9664460465744812e-05, + "loss": 0.6281, + "step": 3496 + }, + { + "epoch": 0.1102980602428639, + "grad_norm": 5.174961090087891, + "learning_rate": 1.9664198008417148e-05, + "loss": 0.6186, + "step": 3497 + }, + { + "epoch": 0.11032960100930453, + "grad_norm": 3.1853525638580322, + "learning_rate": 1.9663935450236073e-05, + "loss": 0.5953, + "step": 3498 + }, + { + "epoch": 0.11036114177574514, + "grad_norm": 3.502998113632202, + "learning_rate": 1.9663672791204328e-05, + "loss": 0.6275, + "step": 3499 + }, + { + "epoch": 0.11039268254218577, + "grad_norm": 3.3711729049682617, + "learning_rate": 1.966341003132465e-05, + "loss": 0.6067, + "step": 3500 + }, + { + "epoch": 0.1104242233086264, + "grad_norm": 3.1101348400115967, + "learning_rate": 1.9663147170599788e-05, + "loss": 0.6481, + "step": 3501 + }, + { + "epoch": 0.11045576407506702, + "grad_norm": 3.8362791538238525, + "learning_rate": 1.966288420903248e-05, + "loss": 0.5974, + "step": 3502 + }, + { + "epoch": 0.11048730484150765, + "grad_norm": 2.421402931213379, + "learning_rate": 1.9662621146625474e-05, + "loss": 0.868, + "step": 3503 + }, + { + "epoch": 0.11051884560794828, + "grad_norm": 3.9696028232574463, + "learning_rate": 1.9662357983381514e-05, + "loss": 0.6139, + "step": 3504 + }, + { + "epoch": 0.1105503863743889, + "grad_norm": 3.977107048034668, + "learning_rate": 1.9662094719303343e-05, + "loss": 0.5925, + "step": 3505 + }, + { + "epoch": 0.11058192714082953, + "grad_norm": 3.5744614601135254, + "learning_rate": 1.9661831354393713e-05, + "loss": 0.609, + "step": 3506 + }, + { + "epoch": 0.11061346790727014, + "grad_norm": 3.338524103164673, + "learning_rate": 1.9661567888655372e-05, + "loss": 0.6206, + "step": 3507 + }, + { + "epoch": 0.11064500867371077, + "grad_norm": 3.028048038482666, + "learning_rate": 1.9661304322091067e-05, + "loss": 0.6062, + "step": 3508 + }, + { + "epoch": 0.1106765494401514, + "grad_norm": 4.336209774017334, + "learning_rate": 1.966104065470355e-05, + "loss": 0.5786, + "step": 3509 + }, + { + "epoch": 0.11070809020659202, + "grad_norm": 6.30415153503418, + "learning_rate": 1.966077688649557e-05, + "loss": 0.5806, + "step": 3510 + }, + { + "epoch": 0.11073963097303265, + "grad_norm": 2.9126343727111816, + "learning_rate": 1.9660513017469883e-05, + "loss": 0.5816, + "step": 3511 + }, + { + "epoch": 0.11077117173947328, + "grad_norm": 3.990845203399658, + "learning_rate": 1.966024904762924e-05, + "loss": 0.5972, + "step": 3512 + }, + { + "epoch": 0.11080271250591389, + "grad_norm": 3.3403208255767822, + "learning_rate": 1.96599849769764e-05, + "loss": 0.6243, + "step": 3513 + }, + { + "epoch": 0.11083425327235452, + "grad_norm": 2.8049187660217285, + "learning_rate": 1.9659720805514115e-05, + "loss": 0.5909, + "step": 3514 + }, + { + "epoch": 0.11086579403879514, + "grad_norm": 4.531438827514648, + "learning_rate": 1.9659456533245143e-05, + "loss": 0.6288, + "step": 3515 + }, + { + "epoch": 0.11089733480523577, + "grad_norm": 3.7370433807373047, + "learning_rate": 1.965919216017224e-05, + "loss": 0.6128, + "step": 3516 + }, + { + "epoch": 0.1109288755716764, + "grad_norm": 3.0838539600372314, + "learning_rate": 1.965892768629817e-05, + "loss": 0.5893, + "step": 3517 + }, + { + "epoch": 0.11096041633811701, + "grad_norm": 3.8805301189422607, + "learning_rate": 1.965866311162569e-05, + "loss": 0.6274, + "step": 3518 + }, + { + "epoch": 0.11099195710455764, + "grad_norm": 4.163698196411133, + "learning_rate": 1.9658398436157555e-05, + "loss": 0.5908, + "step": 3519 + }, + { + "epoch": 0.11102349787099826, + "grad_norm": 5.33357048034668, + "learning_rate": 1.9658133659896536e-05, + "loss": 0.5905, + "step": 3520 + }, + { + "epoch": 0.11105503863743889, + "grad_norm": 3.256842851638794, + "learning_rate": 1.9657868782845396e-05, + "loss": 0.6141, + "step": 3521 + }, + { + "epoch": 0.11108657940387952, + "grad_norm": 4.096503734588623, + "learning_rate": 1.965760380500689e-05, + "loss": 0.6114, + "step": 3522 + }, + { + "epoch": 0.11111812017032013, + "grad_norm": 6.158607006072998, + "learning_rate": 1.9657338726383793e-05, + "loss": 0.6018, + "step": 3523 + }, + { + "epoch": 0.11114966093676076, + "grad_norm": 3.2637650966644287, + "learning_rate": 1.965707354697887e-05, + "loss": 0.5794, + "step": 3524 + }, + { + "epoch": 0.1111812017032014, + "grad_norm": 2.713385820388794, + "learning_rate": 1.965680826679488e-05, + "loss": 0.601, + "step": 3525 + }, + { + "epoch": 0.11121274246964201, + "grad_norm": 3.151536703109741, + "learning_rate": 1.96565428858346e-05, + "loss": 0.5982, + "step": 3526 + }, + { + "epoch": 0.11124428323608264, + "grad_norm": 3.3485782146453857, + "learning_rate": 1.96562774041008e-05, + "loss": 0.6065, + "step": 3527 + }, + { + "epoch": 0.11127582400252325, + "grad_norm": 10.454240798950195, + "learning_rate": 1.9656011821596243e-05, + "loss": 0.6175, + "step": 3528 + }, + { + "epoch": 0.11130736476896388, + "grad_norm": 2.975177764892578, + "learning_rate": 1.9655746138323707e-05, + "loss": 0.4104, + "step": 3529 + }, + { + "epoch": 0.11133890553540451, + "grad_norm": 2.560823678970337, + "learning_rate": 1.9655480354285964e-05, + "loss": 0.605, + "step": 3530 + }, + { + "epoch": 0.11137044630184513, + "grad_norm": 3.5479867458343506, + "learning_rate": 1.965521446948578e-05, + "loss": 0.6024, + "step": 3531 + }, + { + "epoch": 0.11140198706828576, + "grad_norm": 2.5120205879211426, + "learning_rate": 1.9654948483925944e-05, + "loss": 0.6143, + "step": 3532 + }, + { + "epoch": 0.11143352783472639, + "grad_norm": 16.884830474853516, + "learning_rate": 1.965468239760922e-05, + "loss": 0.5938, + "step": 3533 + }, + { + "epoch": 0.111465068601167, + "grad_norm": 3.7109668254852295, + "learning_rate": 1.965441621053839e-05, + "loss": 0.5778, + "step": 3534 + }, + { + "epoch": 0.11149660936760764, + "grad_norm": 3.73516583442688, + "learning_rate": 1.9654149922716228e-05, + "loss": 0.617, + "step": 3535 + }, + { + "epoch": 0.11152815013404825, + "grad_norm": 4.1974592208862305, + "learning_rate": 1.9653883534145514e-05, + "loss": 0.9066, + "step": 3536 + }, + { + "epoch": 0.11155969090048888, + "grad_norm": 2.9584567546844482, + "learning_rate": 1.9653617044829033e-05, + "loss": 0.6191, + "step": 3537 + }, + { + "epoch": 0.11159123166692951, + "grad_norm": 3.3807759284973145, + "learning_rate": 1.965335045476956e-05, + "loss": 0.6035, + "step": 3538 + }, + { + "epoch": 0.11162277243337013, + "grad_norm": 3.427278757095337, + "learning_rate": 1.965308376396988e-05, + "loss": 0.6183, + "step": 3539 + }, + { + "epoch": 0.11165431319981076, + "grad_norm": 4.177976608276367, + "learning_rate": 1.9652816972432777e-05, + "loss": 0.6268, + "step": 3540 + }, + { + "epoch": 0.11168585396625139, + "grad_norm": 3.4829185009002686, + "learning_rate": 1.9652550080161035e-05, + "loss": 0.6409, + "step": 3541 + }, + { + "epoch": 0.111717394732692, + "grad_norm": 3.179774761199951, + "learning_rate": 1.9652283087157432e-05, + "loss": 0.5991, + "step": 3542 + }, + { + "epoch": 0.11174893549913263, + "grad_norm": 13.61487102508545, + "learning_rate": 1.9652015993424766e-05, + "loss": 0.5894, + "step": 3543 + }, + { + "epoch": 0.11178047626557325, + "grad_norm": 3.1489522457122803, + "learning_rate": 1.9651748798965815e-05, + "loss": 0.5856, + "step": 3544 + }, + { + "epoch": 0.11181201703201388, + "grad_norm": 2.6165804862976074, + "learning_rate": 1.965148150378337e-05, + "loss": 0.5839, + "step": 3545 + }, + { + "epoch": 0.11184355779845451, + "grad_norm": 3.26265025138855, + "learning_rate": 1.9651214107880224e-05, + "loss": 0.5791, + "step": 3546 + }, + { + "epoch": 0.11187509856489512, + "grad_norm": 3.9167377948760986, + "learning_rate": 1.9650946611259166e-05, + "loss": 0.5886, + "step": 3547 + }, + { + "epoch": 0.11190663933133575, + "grad_norm": 5.274738788604736, + "learning_rate": 1.9650679013922983e-05, + "loss": 0.581, + "step": 3548 + }, + { + "epoch": 0.11193818009777638, + "grad_norm": 4.488145351409912, + "learning_rate": 1.9650411315874474e-05, + "loss": 0.6419, + "step": 3549 + }, + { + "epoch": 0.111969720864217, + "grad_norm": 3.2319650650024414, + "learning_rate": 1.9650143517116428e-05, + "loss": 0.6079, + "step": 3550 + }, + { + "epoch": 0.11200126163065763, + "grad_norm": 3.248624801635742, + "learning_rate": 1.9649875617651642e-05, + "loss": 0.5858, + "step": 3551 + }, + { + "epoch": 0.11203280239709824, + "grad_norm": 3.4389822483062744, + "learning_rate": 1.964960761748291e-05, + "loss": 0.6206, + "step": 3552 + }, + { + "epoch": 0.11206434316353887, + "grad_norm": 3.20363187789917, + "learning_rate": 1.964933951661303e-05, + "loss": 0.5776, + "step": 3553 + }, + { + "epoch": 0.1120958839299795, + "grad_norm": 2.1646018028259277, + "learning_rate": 1.96490713150448e-05, + "loss": 0.376, + "step": 3554 + }, + { + "epoch": 0.11212742469642012, + "grad_norm": 2.121227741241455, + "learning_rate": 1.9648803012781018e-05, + "loss": 0.911, + "step": 3555 + }, + { + "epoch": 0.11215896546286075, + "grad_norm": 2.7850778102874756, + "learning_rate": 1.9648534609824484e-05, + "loss": 0.6085, + "step": 3556 + }, + { + "epoch": 0.11219050622930138, + "grad_norm": 5.341053485870361, + "learning_rate": 1.9648266106178e-05, + "loss": 0.6086, + "step": 3557 + }, + { + "epoch": 0.112222046995742, + "grad_norm": 6.729026794433594, + "learning_rate": 1.9647997501844366e-05, + "loss": 0.6366, + "step": 3558 + }, + { + "epoch": 0.11225358776218262, + "grad_norm": 2.0043907165527344, + "learning_rate": 1.964772879682639e-05, + "loss": 0.8661, + "step": 3559 + }, + { + "epoch": 0.11228512852862324, + "grad_norm": 4.12030553817749, + "learning_rate": 1.964745999112687e-05, + "loss": 0.6107, + "step": 3560 + }, + { + "epoch": 0.11231666929506387, + "grad_norm": 2.482457399368286, + "learning_rate": 1.9647191084748614e-05, + "loss": 0.6193, + "step": 3561 + }, + { + "epoch": 0.1123482100615045, + "grad_norm": 3.258011817932129, + "learning_rate": 1.964692207769443e-05, + "loss": 0.5998, + "step": 3562 + }, + { + "epoch": 0.11237975082794512, + "grad_norm": 3.4808685779571533, + "learning_rate": 1.9646652969967123e-05, + "loss": 0.6098, + "step": 3563 + }, + { + "epoch": 0.11241129159438575, + "grad_norm": 2.8427624702453613, + "learning_rate": 1.96463837615695e-05, + "loss": 0.6368, + "step": 3564 + }, + { + "epoch": 0.11244283236082636, + "grad_norm": 3.929466962814331, + "learning_rate": 1.9646114452504374e-05, + "loss": 0.5878, + "step": 3565 + }, + { + "epoch": 0.11247437312726699, + "grad_norm": 2.600148916244507, + "learning_rate": 1.9645845042774555e-05, + "loss": 0.6133, + "step": 3566 + }, + { + "epoch": 0.11250591389370762, + "grad_norm": 3.5975115299224854, + "learning_rate": 1.964557553238285e-05, + "loss": 0.6491, + "step": 3567 + }, + { + "epoch": 0.11253745466014824, + "grad_norm": 4.581226825714111, + "learning_rate": 1.9645305921332075e-05, + "loss": 0.6063, + "step": 3568 + }, + { + "epoch": 0.11256899542658887, + "grad_norm": 3.9976305961608887, + "learning_rate": 1.9645036209625043e-05, + "loss": 0.6133, + "step": 3569 + }, + { + "epoch": 0.1126005361930295, + "grad_norm": 6.664577484130859, + "learning_rate": 1.964476639726457e-05, + "loss": 0.6001, + "step": 3570 + }, + { + "epoch": 0.11263207695947011, + "grad_norm": 3.8292336463928223, + "learning_rate": 1.9644496484253473e-05, + "loss": 0.5716, + "step": 3571 + }, + { + "epoch": 0.11266361772591074, + "grad_norm": 3.0986926555633545, + "learning_rate": 1.9644226470594567e-05, + "loss": 0.6301, + "step": 3572 + }, + { + "epoch": 0.11269515849235136, + "grad_norm": 2.792773962020874, + "learning_rate": 1.9643956356290667e-05, + "loss": 0.6295, + "step": 3573 + }, + { + "epoch": 0.11272669925879199, + "grad_norm": 1.9297854900360107, + "learning_rate": 1.9643686141344593e-05, + "loss": 0.4323, + "step": 3574 + }, + { + "epoch": 0.11275824002523262, + "grad_norm": 2.9523885250091553, + "learning_rate": 1.9643415825759166e-05, + "loss": 0.94, + "step": 3575 + }, + { + "epoch": 0.11278978079167323, + "grad_norm": 3.030938148498535, + "learning_rate": 1.9643145409537206e-05, + "loss": 0.6137, + "step": 3576 + }, + { + "epoch": 0.11282132155811386, + "grad_norm": 4.118042945861816, + "learning_rate": 1.964287489268154e-05, + "loss": 0.6387, + "step": 3577 + }, + { + "epoch": 0.11285286232455449, + "grad_norm": 3.0214271545410156, + "learning_rate": 1.9642604275194982e-05, + "loss": 0.618, + "step": 3578 + }, + { + "epoch": 0.11288440309099511, + "grad_norm": 4.013029098510742, + "learning_rate": 1.9642333557080366e-05, + "loss": 0.6076, + "step": 3579 + }, + { + "epoch": 0.11291594385743574, + "grad_norm": 2.7411816120147705, + "learning_rate": 1.964206273834051e-05, + "loss": 0.6254, + "step": 3580 + }, + { + "epoch": 0.11294748462387635, + "grad_norm": 2.6436169147491455, + "learning_rate": 1.964179181897824e-05, + "loss": 0.6001, + "step": 3581 + }, + { + "epoch": 0.11297902539031698, + "grad_norm": 5.070707321166992, + "learning_rate": 1.964152079899639e-05, + "loss": 0.5874, + "step": 3582 + }, + { + "epoch": 0.11301056615675761, + "grad_norm": 2.8480143547058105, + "learning_rate": 1.9641249678397783e-05, + "loss": 0.6155, + "step": 3583 + }, + { + "epoch": 0.11304210692319823, + "grad_norm": 3.2272865772247314, + "learning_rate": 1.9640978457185248e-05, + "loss": 0.6203, + "step": 3584 + }, + { + "epoch": 0.11307364768963886, + "grad_norm": 388.7747497558594, + "learning_rate": 1.9640707135361617e-05, + "loss": 2.372, + "step": 3585 + }, + { + "epoch": 0.11310518845607949, + "grad_norm": 21.301170349121094, + "learning_rate": 1.964043571292972e-05, + "loss": 0.3727, + "step": 3586 + }, + { + "epoch": 0.1131367292225201, + "grad_norm": 41.16116714477539, + "learning_rate": 1.9640164189892393e-05, + "loss": 0.9928, + "step": 3587 + }, + { + "epoch": 0.11316826998896073, + "grad_norm": 171.88446044921875, + "learning_rate": 1.963989256625247e-05, + "loss": 0.6868, + "step": 3588 + }, + { + "epoch": 0.11319981075540135, + "grad_norm": 6.6595258712768555, + "learning_rate": 1.9639620842012775e-05, + "loss": 0.6241, + "step": 3589 + }, + { + "epoch": 0.11323135152184198, + "grad_norm": 4.351492404937744, + "learning_rate": 1.963934901717616e-05, + "loss": 0.6021, + "step": 3590 + }, + { + "epoch": 0.11326289228828261, + "grad_norm": 15.23533821105957, + "learning_rate": 1.9639077091745445e-05, + "loss": 0.6283, + "step": 3591 + }, + { + "epoch": 0.11329443305472323, + "grad_norm": 113.91815948486328, + "learning_rate": 1.9638805065723485e-05, + "loss": 0.7587, + "step": 3592 + }, + { + "epoch": 0.11332597382116386, + "grad_norm": 27.211637496948242, + "learning_rate": 1.9638532939113102e-05, + "loss": 0.6219, + "step": 3593 + }, + { + "epoch": 0.11335751458760449, + "grad_norm": 20.423789978027344, + "learning_rate": 1.9638260711917147e-05, + "loss": 0.704, + "step": 3594 + }, + { + "epoch": 0.1133890553540451, + "grad_norm": 16.744970321655273, + "learning_rate": 1.963798838413846e-05, + "loss": 0.6489, + "step": 3595 + }, + { + "epoch": 0.11342059612048573, + "grad_norm": 56.60093307495117, + "learning_rate": 1.9637715955779873e-05, + "loss": 1.706, + "step": 3596 + }, + { + "epoch": 0.11345213688692635, + "grad_norm": 42.76203536987305, + "learning_rate": 1.9637443426844243e-05, + "loss": 1.2454, + "step": 3597 + }, + { + "epoch": 0.11348367765336698, + "grad_norm": 56.4384880065918, + "learning_rate": 1.9637170797334407e-05, + "loss": 3.172, + "step": 3598 + }, + { + "epoch": 0.1135152184198076, + "grad_norm": 35.10319519042969, + "learning_rate": 1.963689806725321e-05, + "loss": 2.3853, + "step": 3599 + }, + { + "epoch": 0.11354675918624822, + "grad_norm": 108.93353271484375, + "learning_rate": 1.9636625236603498e-05, + "loss": 1.6228, + "step": 3600 + }, + { + "epoch": 0.11357829995268885, + "grad_norm": 29.01351547241211, + "learning_rate": 1.9636352305388116e-05, + "loss": 1.4862, + "step": 3601 + }, + { + "epoch": 0.11360984071912947, + "grad_norm": 29.816864013671875, + "learning_rate": 1.963607927360992e-05, + "loss": 1.0269, + "step": 3602 + }, + { + "epoch": 0.1136413814855701, + "grad_norm": 129.18544006347656, + "learning_rate": 1.9635806141271752e-05, + "loss": 1.0127, + "step": 3603 + }, + { + "epoch": 0.11367292225201073, + "grad_norm": 10.87602710723877, + "learning_rate": 1.9635532908376465e-05, + "loss": 0.8656, + "step": 3604 + }, + { + "epoch": 0.11370446301845134, + "grad_norm": 11.929553031921387, + "learning_rate": 1.9635259574926912e-05, + "loss": 0.815, + "step": 3605 + }, + { + "epoch": 0.11373600378489197, + "grad_norm": 8.586023330688477, + "learning_rate": 1.9634986140925946e-05, + "loss": 0.752, + "step": 3606 + }, + { + "epoch": 0.1137675445513326, + "grad_norm": 7.374168872833252, + "learning_rate": 1.9634712606376416e-05, + "loss": 0.7417, + "step": 3607 + }, + { + "epoch": 0.11379908531777322, + "grad_norm": 507.2678527832031, + "learning_rate": 1.9634438971281175e-05, + "loss": 0.8807, + "step": 3608 + }, + { + "epoch": 0.11383062608421385, + "grad_norm": 6.594489574432373, + "learning_rate": 1.9634165235643085e-05, + "loss": 0.7098, + "step": 3609 + }, + { + "epoch": 0.11386216685065446, + "grad_norm": 5.355443477630615, + "learning_rate": 1.9633891399465e-05, + "loss": 0.6847, + "step": 3610 + }, + { + "epoch": 0.1138937076170951, + "grad_norm": 6.408955097198486, + "learning_rate": 1.963361746274978e-05, + "loss": 0.845, + "step": 3611 + }, + { + "epoch": 0.11392524838353572, + "grad_norm": 5.531891345977783, + "learning_rate": 1.963334342550028e-05, + "loss": 0.7746, + "step": 3612 + }, + { + "epoch": 0.11395678914997634, + "grad_norm": 5.741943836212158, + "learning_rate": 1.963306928771936e-05, + "loss": 0.7076, + "step": 3613 + }, + { + "epoch": 0.11398832991641697, + "grad_norm": 13.627901077270508, + "learning_rate": 1.9632795049409884e-05, + "loss": 0.7486, + "step": 3614 + }, + { + "epoch": 0.1140198706828576, + "grad_norm": 135.9486846923828, + "learning_rate": 1.9632520710574713e-05, + "loss": 7.5698, + "step": 3615 + }, + { + "epoch": 0.11405141144929822, + "grad_norm": 4.6453471183776855, + "learning_rate": 1.9632246271216705e-05, + "loss": 0.7146, + "step": 3616 + }, + { + "epoch": 0.11408295221573884, + "grad_norm": 16.46793556213379, + "learning_rate": 1.963197173133873e-05, + "loss": 0.8092, + "step": 3617 + }, + { + "epoch": 0.11411449298217946, + "grad_norm": 7.586446285247803, + "learning_rate": 1.9631697090943653e-05, + "loss": 0.7366, + "step": 3618 + }, + { + "epoch": 0.11414603374862009, + "grad_norm": 5.8783698081970215, + "learning_rate": 1.9631422350034336e-05, + "loss": 0.6814, + "step": 3619 + }, + { + "epoch": 0.11417757451506072, + "grad_norm": 84.09779357910156, + "learning_rate": 1.963114750861365e-05, + "loss": 2.2079, + "step": 3620 + }, + { + "epoch": 0.11420911528150134, + "grad_norm": 5.183172225952148, + "learning_rate": 1.9630872566684454e-05, + "loss": 0.6823, + "step": 3621 + }, + { + "epoch": 0.11424065604794197, + "grad_norm": 5.665314197540283, + "learning_rate": 1.9630597524249633e-05, + "loss": 0.6715, + "step": 3622 + }, + { + "epoch": 0.1142721968143826, + "grad_norm": 4.519900321960449, + "learning_rate": 1.9630322381312045e-05, + "loss": 1.3318, + "step": 3623 + }, + { + "epoch": 0.11430373758082321, + "grad_norm": 5.770659923553467, + "learning_rate": 1.9630047137874567e-05, + "loss": 0.6835, + "step": 3624 + }, + { + "epoch": 0.11433527834726384, + "grad_norm": 4.007013320922852, + "learning_rate": 1.962977179394007e-05, + "loss": 0.652, + "step": 3625 + }, + { + "epoch": 0.11436681911370446, + "grad_norm": 4.299184799194336, + "learning_rate": 1.9629496349511423e-05, + "loss": 1.2873, + "step": 3626 + }, + { + "epoch": 0.11439835988014509, + "grad_norm": 4.7082319259643555, + "learning_rate": 1.9629220804591507e-05, + "loss": 0.6811, + "step": 3627 + }, + { + "epoch": 0.11442990064658572, + "grad_norm": 5.122767448425293, + "learning_rate": 1.9628945159183197e-05, + "loss": 0.6872, + "step": 3628 + }, + { + "epoch": 0.11446144141302633, + "grad_norm": 3.7956812381744385, + "learning_rate": 1.9628669413289363e-05, + "loss": 0.7282, + "step": 3629 + }, + { + "epoch": 0.11449298217946696, + "grad_norm": 5.102871894836426, + "learning_rate": 1.962839356691289e-05, + "loss": 0.6785, + "step": 3630 + }, + { + "epoch": 0.11452452294590759, + "grad_norm": 4.2470526695251465, + "learning_rate": 1.9628117620056655e-05, + "loss": 0.6609, + "step": 3631 + }, + { + "epoch": 0.11455606371234821, + "grad_norm": 3.122987985610962, + "learning_rate": 1.9627841572723537e-05, + "loss": 0.6659, + "step": 3632 + }, + { + "epoch": 0.11458760447878884, + "grad_norm": 4.090303421020508, + "learning_rate": 1.962756542491641e-05, + "loss": 0.6612, + "step": 3633 + }, + { + "epoch": 0.11461914524522945, + "grad_norm": 4.893101692199707, + "learning_rate": 1.9627289176638168e-05, + "loss": 0.6159, + "step": 3634 + }, + { + "epoch": 0.11465068601167008, + "grad_norm": 3.2173092365264893, + "learning_rate": 1.962701282789169e-05, + "loss": 0.6897, + "step": 3635 + }, + { + "epoch": 0.11468222677811071, + "grad_norm": 5.791793346405029, + "learning_rate": 1.9626736378679853e-05, + "loss": 0.6616, + "step": 3636 + }, + { + "epoch": 0.11471376754455133, + "grad_norm": 4.194702625274658, + "learning_rate": 1.9626459829005552e-05, + "loss": 0.7043, + "step": 3637 + }, + { + "epoch": 0.11474530831099196, + "grad_norm": 3.460827350616455, + "learning_rate": 1.9626183178871663e-05, + "loss": 0.6297, + "step": 3638 + }, + { + "epoch": 0.11477684907743259, + "grad_norm": 3.8722896575927734, + "learning_rate": 1.962590642828108e-05, + "loss": 0.639, + "step": 3639 + }, + { + "epoch": 0.1148083898438732, + "grad_norm": 23.644643783569336, + "learning_rate": 1.962562957723669e-05, + "loss": 2.1, + "step": 3640 + }, + { + "epoch": 0.11483993061031383, + "grad_norm": 5.277064323425293, + "learning_rate": 1.9625352625741377e-05, + "loss": 0.6995, + "step": 3641 + }, + { + "epoch": 0.11487147137675445, + "grad_norm": 7.1301493644714355, + "learning_rate": 1.962507557379804e-05, + "loss": 0.6492, + "step": 3642 + }, + { + "epoch": 0.11490301214319508, + "grad_norm": 6.025801658630371, + "learning_rate": 1.9624798421409564e-05, + "loss": 1.2611, + "step": 3643 + }, + { + "epoch": 0.11493455290963571, + "grad_norm": 4.034681797027588, + "learning_rate": 1.9624521168578844e-05, + "loss": 0.7152, + "step": 3644 + }, + { + "epoch": 0.11496609367607633, + "grad_norm": 105.21385955810547, + "learning_rate": 1.9624243815308773e-05, + "loss": 5.7107, + "step": 3645 + }, + { + "epoch": 0.11499763444251695, + "grad_norm": 4.598401069641113, + "learning_rate": 1.9623966361602242e-05, + "loss": 0.7032, + "step": 3646 + }, + { + "epoch": 0.11502917520895757, + "grad_norm": 5.817155838012695, + "learning_rate": 1.962368880746215e-05, + "loss": 0.7073, + "step": 3647 + }, + { + "epoch": 0.1150607159753982, + "grad_norm": 3.557535409927368, + "learning_rate": 1.9623411152891394e-05, + "loss": 1.2288, + "step": 3648 + }, + { + "epoch": 0.11509225674183883, + "grad_norm": 45.269412994384766, + "learning_rate": 1.9623133397892872e-05, + "loss": 0.765, + "step": 3649 + }, + { + "epoch": 0.11512379750827945, + "grad_norm": 4.868078708648682, + "learning_rate": 1.9622855542469476e-05, + "loss": 0.7221, + "step": 3650 + }, + { + "epoch": 0.11515533827472008, + "grad_norm": 5.027299880981445, + "learning_rate": 1.9622577586624114e-05, + "loss": 0.6978, + "step": 3651 + }, + { + "epoch": 0.1151868790411607, + "grad_norm": 8.716188430786133, + "learning_rate": 1.9622299530359682e-05, + "loss": 0.6641, + "step": 3652 + }, + { + "epoch": 0.11521841980760132, + "grad_norm": 5.163614273071289, + "learning_rate": 1.9622021373679085e-05, + "loss": 0.6451, + "step": 3653 + }, + { + "epoch": 0.11524996057404195, + "grad_norm": 4.332658767700195, + "learning_rate": 1.9621743116585218e-05, + "loss": 0.6633, + "step": 3654 + }, + { + "epoch": 0.11528150134048257, + "grad_norm": 3.33323073387146, + "learning_rate": 1.9621464759080993e-05, + "loss": 0.6636, + "step": 3655 + }, + { + "epoch": 0.1153130421069232, + "grad_norm": 3.1674914360046387, + "learning_rate": 1.9621186301169316e-05, + "loss": 0.6563, + "step": 3656 + }, + { + "epoch": 0.11534458287336383, + "grad_norm": 2.6766998767852783, + "learning_rate": 1.9620907742853085e-05, + "loss": 0.6338, + "step": 3657 + }, + { + "epoch": 0.11537612363980444, + "grad_norm": 3.515937566757202, + "learning_rate": 1.9620629084135216e-05, + "loss": 0.6463, + "step": 3658 + }, + { + "epoch": 0.11540766440624507, + "grad_norm": 4.624590873718262, + "learning_rate": 1.9620350325018606e-05, + "loss": 0.6583, + "step": 3659 + }, + { + "epoch": 0.1154392051726857, + "grad_norm": 3.4846835136413574, + "learning_rate": 1.9620071465506173e-05, + "loss": 0.6246, + "step": 3660 + }, + { + "epoch": 0.11547074593912632, + "grad_norm": 3.5097877979278564, + "learning_rate": 1.9619792505600827e-05, + "loss": 0.6595, + "step": 3661 + }, + { + "epoch": 0.11550228670556695, + "grad_norm": 3.68662428855896, + "learning_rate": 1.9619513445305474e-05, + "loss": 0.6667, + "step": 3662 + }, + { + "epoch": 0.11553382747200756, + "grad_norm": 20.354198455810547, + "learning_rate": 1.9619234284623026e-05, + "loss": 1.604, + "step": 3663 + }, + { + "epoch": 0.1155653682384482, + "grad_norm": 6.333023548126221, + "learning_rate": 1.9618955023556406e-05, + "loss": 1.0076, + "step": 3664 + }, + { + "epoch": 0.11559690900488882, + "grad_norm": 7.258790969848633, + "learning_rate": 1.961867566210852e-05, + "loss": 0.6791, + "step": 3665 + }, + { + "epoch": 0.11562844977132944, + "grad_norm": 3.3783926963806152, + "learning_rate": 1.961839620028228e-05, + "loss": 0.6203, + "step": 3666 + }, + { + "epoch": 0.11565999053777007, + "grad_norm": 19.583375930786133, + "learning_rate": 1.961811663808061e-05, + "loss": 1.0159, + "step": 3667 + }, + { + "epoch": 0.1156915313042107, + "grad_norm": 6.146492004394531, + "learning_rate": 1.9617836975506424e-05, + "loss": 0.6637, + "step": 3668 + }, + { + "epoch": 0.11572307207065131, + "grad_norm": 3.173175573348999, + "learning_rate": 1.9617557212562643e-05, + "loss": 0.628, + "step": 3669 + }, + { + "epoch": 0.11575461283709194, + "grad_norm": 8.748024940490723, + "learning_rate": 1.9617277349252186e-05, + "loss": 0.6832, + "step": 3670 + }, + { + "epoch": 0.11578615360353256, + "grad_norm": 7.08633279800415, + "learning_rate": 1.9616997385577967e-05, + "loss": 0.72, + "step": 3671 + }, + { + "epoch": 0.11581769436997319, + "grad_norm": 5.988025188446045, + "learning_rate": 1.9616717321542914e-05, + "loss": 0.64, + "step": 3672 + }, + { + "epoch": 0.11584923513641382, + "grad_norm": 3.7728374004364014, + "learning_rate": 1.961643715714995e-05, + "loss": 0.6133, + "step": 3673 + }, + { + "epoch": 0.11588077590285444, + "grad_norm": 3.882521152496338, + "learning_rate": 1.9616156892401997e-05, + "loss": 0.6424, + "step": 3674 + }, + { + "epoch": 0.11591231666929507, + "grad_norm": 4.086796760559082, + "learning_rate": 1.961587652730198e-05, + "loss": 1.1999, + "step": 3675 + }, + { + "epoch": 0.1159438574357357, + "grad_norm": 2.858827590942383, + "learning_rate": 1.9615596061852827e-05, + "loss": 0.6209, + "step": 3676 + }, + { + "epoch": 0.11597539820217631, + "grad_norm": 3.139463424682617, + "learning_rate": 1.9615315496057462e-05, + "loss": 0.5861, + "step": 3677 + }, + { + "epoch": 0.11600693896861694, + "grad_norm": 3.133419990539551, + "learning_rate": 1.961503482991881e-05, + "loss": 0.6117, + "step": 3678 + }, + { + "epoch": 0.11603847973505756, + "grad_norm": 5.112422943115234, + "learning_rate": 1.9614754063439808e-05, + "loss": 0.685, + "step": 3679 + }, + { + "epoch": 0.11607002050149819, + "grad_norm": 3.204965591430664, + "learning_rate": 1.9614473196623377e-05, + "loss": 0.6313, + "step": 3680 + }, + { + "epoch": 0.11610156126793882, + "grad_norm": 3.8351237773895264, + "learning_rate": 1.961419222947246e-05, + "loss": 1.0791, + "step": 3681 + }, + { + "epoch": 0.11613310203437943, + "grad_norm": 5.396609306335449, + "learning_rate": 1.9613911161989975e-05, + "loss": 0.6624, + "step": 3682 + }, + { + "epoch": 0.11616464280082006, + "grad_norm": 3.7206621170043945, + "learning_rate": 1.9613629994178868e-05, + "loss": 0.6918, + "step": 3683 + }, + { + "epoch": 0.11619618356726069, + "grad_norm": 5.459929466247559, + "learning_rate": 1.9613348726042062e-05, + "loss": 0.6404, + "step": 3684 + }, + { + "epoch": 0.11622772433370131, + "grad_norm": 2.8719840049743652, + "learning_rate": 1.96130673575825e-05, + "loss": 0.654, + "step": 3685 + }, + { + "epoch": 0.11625926510014194, + "grad_norm": 3.8927652835845947, + "learning_rate": 1.9612785888803115e-05, + "loss": 0.6301, + "step": 3686 + }, + { + "epoch": 0.11629080586658255, + "grad_norm": 3.024148941040039, + "learning_rate": 1.9612504319706843e-05, + "loss": 0.6028, + "step": 3687 + }, + { + "epoch": 0.11632234663302318, + "grad_norm": 2.966895341873169, + "learning_rate": 1.9612222650296625e-05, + "loss": 1.0218, + "step": 3688 + }, + { + "epoch": 0.11635388739946381, + "grad_norm": 3.864673614501953, + "learning_rate": 1.9611940880575405e-05, + "loss": 0.6499, + "step": 3689 + }, + { + "epoch": 0.11638542816590443, + "grad_norm": 3.577279567718506, + "learning_rate": 1.9611659010546114e-05, + "loss": 0.6372, + "step": 3690 + }, + { + "epoch": 0.11641696893234506, + "grad_norm": 17.788354873657227, + "learning_rate": 1.96113770402117e-05, + "loss": 1.4125, + "step": 3691 + }, + { + "epoch": 0.11644850969878567, + "grad_norm": 4.461820602416992, + "learning_rate": 1.9611094969575104e-05, + "loss": 0.6283, + "step": 3692 + }, + { + "epoch": 0.1164800504652263, + "grad_norm": 4.470247268676758, + "learning_rate": 1.9610812798639266e-05, + "loss": 0.6078, + "step": 3693 + }, + { + "epoch": 0.11651159123166693, + "grad_norm": 4.032199859619141, + "learning_rate": 1.9610530527407135e-05, + "loss": 0.632, + "step": 3694 + }, + { + "epoch": 0.11654313199810755, + "grad_norm": 3.7225594520568848, + "learning_rate": 1.9610248155881657e-05, + "loss": 0.61, + "step": 3695 + }, + { + "epoch": 0.11657467276454818, + "grad_norm": 3.845522880554199, + "learning_rate": 1.9609965684065778e-05, + "loss": 0.6123, + "step": 3696 + }, + { + "epoch": 0.11660621353098881, + "grad_norm": 6.667596817016602, + "learning_rate": 1.9609683111962444e-05, + "loss": 0.6229, + "step": 3697 + }, + { + "epoch": 0.11663775429742942, + "grad_norm": 84.74995422363281, + "learning_rate": 1.9609400439574605e-05, + "loss": 0.609, + "step": 3698 + }, + { + "epoch": 0.11666929506387005, + "grad_norm": 3.455502986907959, + "learning_rate": 1.960911766690521e-05, + "loss": 0.623, + "step": 3699 + }, + { + "epoch": 0.11670083583031067, + "grad_norm": 3.2784109115600586, + "learning_rate": 1.960883479395721e-05, + "loss": 0.6161, + "step": 3700 + }, + { + "epoch": 0.1167323765967513, + "grad_norm": 2.9068567752838135, + "learning_rate": 1.960855182073356e-05, + "loss": 0.6285, + "step": 3701 + }, + { + "epoch": 0.11676391736319193, + "grad_norm": 3.9437153339385986, + "learning_rate": 1.960826874723721e-05, + "loss": 0.6321, + "step": 3702 + }, + { + "epoch": 0.11679545812963255, + "grad_norm": 2.477358818054199, + "learning_rate": 1.9607985573471112e-05, + "loss": 1.0566, + "step": 3703 + }, + { + "epoch": 0.11682699889607318, + "grad_norm": 3.890362501144409, + "learning_rate": 1.960770229943823e-05, + "loss": 0.6283, + "step": 3704 + }, + { + "epoch": 0.1168585396625138, + "grad_norm": 4.0926289558410645, + "learning_rate": 1.9607418925141512e-05, + "loss": 0.6095, + "step": 3705 + }, + { + "epoch": 0.11689008042895442, + "grad_norm": 4.195985794067383, + "learning_rate": 1.9607135450583916e-05, + "loss": 0.6184, + "step": 3706 + }, + { + "epoch": 0.11692162119539505, + "grad_norm": 3.721707344055176, + "learning_rate": 1.9606851875768404e-05, + "loss": 0.6257, + "step": 3707 + }, + { + "epoch": 0.11695316196183567, + "grad_norm": 3.700291633605957, + "learning_rate": 1.9606568200697933e-05, + "loss": 0.6168, + "step": 3708 + }, + { + "epoch": 0.1169847027282763, + "grad_norm": 4.6507792472839355, + "learning_rate": 1.9606284425375462e-05, + "loss": 0.6382, + "step": 3709 + }, + { + "epoch": 0.11701624349471693, + "grad_norm": 3.6879007816314697, + "learning_rate": 1.9606000549803955e-05, + "loss": 0.6526, + "step": 3710 + }, + { + "epoch": 0.11704778426115754, + "grad_norm": 1.8863729238510132, + "learning_rate": 1.9605716573986372e-05, + "loss": 0.9822, + "step": 3711 + }, + { + "epoch": 0.11707932502759817, + "grad_norm": 10.343219757080078, + "learning_rate": 1.960543249792568e-05, + "loss": 1.0084, + "step": 3712 + }, + { + "epoch": 0.1171108657940388, + "grad_norm": 4.299734115600586, + "learning_rate": 1.960514832162484e-05, + "loss": 0.666, + "step": 3713 + }, + { + "epoch": 0.11714240656047942, + "grad_norm": 6.034344673156738, + "learning_rate": 1.960486404508682e-05, + "loss": 0.6184, + "step": 3714 + }, + { + "epoch": 0.11717394732692005, + "grad_norm": 2.92927885055542, + "learning_rate": 1.9604579668314584e-05, + "loss": 1.0197, + "step": 3715 + }, + { + "epoch": 0.11720548809336066, + "grad_norm": 4.000760555267334, + "learning_rate": 1.9604295191311103e-05, + "loss": 0.6356, + "step": 3716 + }, + { + "epoch": 0.11723702885980129, + "grad_norm": 3.606276273727417, + "learning_rate": 1.9604010614079342e-05, + "loss": 0.6531, + "step": 3717 + }, + { + "epoch": 0.11726856962624192, + "grad_norm": 3.981999158859253, + "learning_rate": 1.9603725936622275e-05, + "loss": 0.6653, + "step": 3718 + }, + { + "epoch": 0.11730011039268254, + "grad_norm": 3.351240873336792, + "learning_rate": 1.9603441158942875e-05, + "loss": 0.6736, + "step": 3719 + }, + { + "epoch": 0.11733165115912317, + "grad_norm": 4.409878730773926, + "learning_rate": 1.96031562810441e-05, + "loss": 0.6787, + "step": 3720 + }, + { + "epoch": 0.1173631919255638, + "grad_norm": 20.580076217651367, + "learning_rate": 1.960287130292894e-05, + "loss": 2.7267, + "step": 3721 + }, + { + "epoch": 0.11739473269200441, + "grad_norm": 2.9324769973754883, + "learning_rate": 1.9602586224600357e-05, + "loss": 1.0444, + "step": 3722 + }, + { + "epoch": 0.11742627345844504, + "grad_norm": 3.6092214584350586, + "learning_rate": 1.9602301046061333e-05, + "loss": 0.9929, + "step": 3723 + }, + { + "epoch": 0.11745781422488566, + "grad_norm": 4.6504292488098145, + "learning_rate": 1.9602015767314842e-05, + "loss": 0.6181, + "step": 3724 + }, + { + "epoch": 0.11748935499132629, + "grad_norm": 3.5242762565612793, + "learning_rate": 1.960173038836386e-05, + "loss": 0.6959, + "step": 3725 + }, + { + "epoch": 0.11752089575776692, + "grad_norm": 4.73506498336792, + "learning_rate": 1.9601444909211365e-05, + "loss": 0.6397, + "step": 3726 + }, + { + "epoch": 0.11755243652420753, + "grad_norm": 4.398555755615234, + "learning_rate": 1.9601159329860338e-05, + "loss": 0.644, + "step": 3727 + }, + { + "epoch": 0.11758397729064816, + "grad_norm": 2.7393229007720947, + "learning_rate": 1.9600873650313757e-05, + "loss": 1.0159, + "step": 3728 + }, + { + "epoch": 0.11761551805708878, + "grad_norm": 3.7836806774139404, + "learning_rate": 1.9600587870574606e-05, + "loss": 0.6794, + "step": 3729 + }, + { + "epoch": 0.11764705882352941, + "grad_norm": 4.497918128967285, + "learning_rate": 1.9600301990645864e-05, + "loss": 0.6181, + "step": 3730 + }, + { + "epoch": 0.11767859958997004, + "grad_norm": 4.241062164306641, + "learning_rate": 1.960001601053052e-05, + "loss": 0.6328, + "step": 3731 + }, + { + "epoch": 0.11771014035641066, + "grad_norm": 4.16522741317749, + "learning_rate": 1.9599729930231553e-05, + "loss": 0.6225, + "step": 3732 + }, + { + "epoch": 0.11774168112285129, + "grad_norm": 3.4409282207489014, + "learning_rate": 1.9599443749751947e-05, + "loss": 0.6321, + "step": 3733 + }, + { + "epoch": 0.11777322188929192, + "grad_norm": 3.8090250492095947, + "learning_rate": 1.9599157469094696e-05, + "loss": 0.6232, + "step": 3734 + }, + { + "epoch": 0.11780476265573253, + "grad_norm": 4.274829864501953, + "learning_rate": 1.959887108826278e-05, + "loss": 0.6499, + "step": 3735 + }, + { + "epoch": 0.11783630342217316, + "grad_norm": 3.833752155303955, + "learning_rate": 1.9598584607259196e-05, + "loss": 0.6173, + "step": 3736 + }, + { + "epoch": 0.11786784418861378, + "grad_norm": 3.6644999980926514, + "learning_rate": 1.9598298026086924e-05, + "loss": 0.6285, + "step": 3737 + }, + { + "epoch": 0.1178993849550544, + "grad_norm": 3.759671211242676, + "learning_rate": 1.959801134474896e-05, + "loss": 0.6148, + "step": 3738 + }, + { + "epoch": 0.11793092572149504, + "grad_norm": 4.105844974517822, + "learning_rate": 1.9597724563248297e-05, + "loss": 1.0237, + "step": 3739 + }, + { + "epoch": 0.11796246648793565, + "grad_norm": 6.627462387084961, + "learning_rate": 1.9597437681587922e-05, + "loss": 0.6497, + "step": 3740 + }, + { + "epoch": 0.11799400725437628, + "grad_norm": 4.226980209350586, + "learning_rate": 1.9597150699770834e-05, + "loss": 0.6389, + "step": 3741 + }, + { + "epoch": 0.11802554802081691, + "grad_norm": 3.8527684211730957, + "learning_rate": 1.9596863617800025e-05, + "loss": 0.6486, + "step": 3742 + }, + { + "epoch": 0.11805708878725753, + "grad_norm": 3.850804567337036, + "learning_rate": 1.9596576435678497e-05, + "loss": 0.6324, + "step": 3743 + }, + { + "epoch": 0.11808862955369816, + "grad_norm": 3.974146842956543, + "learning_rate": 1.959628915340924e-05, + "loss": 0.6143, + "step": 3744 + }, + { + "epoch": 0.11812017032013877, + "grad_norm": 3.350337028503418, + "learning_rate": 1.9596001770995253e-05, + "loss": 0.6244, + "step": 3745 + }, + { + "epoch": 0.1181517110865794, + "grad_norm": 2.774202346801758, + "learning_rate": 1.9595714288439538e-05, + "loss": 0.6516, + "step": 3746 + }, + { + "epoch": 0.11818325185302003, + "grad_norm": 2.2466483116149902, + "learning_rate": 1.9595426705745094e-05, + "loss": 1.0196, + "step": 3747 + }, + { + "epoch": 0.11821479261946065, + "grad_norm": 3.5459158420562744, + "learning_rate": 1.959513902291492e-05, + "loss": 0.6343, + "step": 3748 + }, + { + "epoch": 0.11824633338590128, + "grad_norm": 3.5309131145477295, + "learning_rate": 1.959485123995202e-05, + "loss": 0.6147, + "step": 3749 + }, + { + "epoch": 0.11827787415234191, + "grad_norm": 3.3555006980895996, + "learning_rate": 1.9594563356859398e-05, + "loss": 0.6583, + "step": 3750 + }, + { + "epoch": 0.11830941491878252, + "grad_norm": 6.932295322418213, + "learning_rate": 1.9594275373640058e-05, + "loss": 0.6117, + "step": 3751 + }, + { + "epoch": 0.11834095568522315, + "grad_norm": 4.356021404266357, + "learning_rate": 1.9593987290297002e-05, + "loss": 0.6414, + "step": 3752 + }, + { + "epoch": 0.11837249645166377, + "grad_norm": 3.1441102027893066, + "learning_rate": 1.9593699106833242e-05, + "loss": 0.6264, + "step": 3753 + }, + { + "epoch": 0.1184040372181044, + "grad_norm": 2.9520630836486816, + "learning_rate": 1.959341082325178e-05, + "loss": 0.5873, + "step": 3754 + }, + { + "epoch": 0.11843557798454503, + "grad_norm": 2.9399001598358154, + "learning_rate": 1.9593122439555626e-05, + "loss": 0.6402, + "step": 3755 + }, + { + "epoch": 0.11846711875098564, + "grad_norm": 3.459627151489258, + "learning_rate": 1.9592833955747793e-05, + "loss": 0.6299, + "step": 3756 + }, + { + "epoch": 0.11849865951742627, + "grad_norm": 3.92581844329834, + "learning_rate": 1.959254537183129e-05, + "loss": 0.6182, + "step": 3757 + }, + { + "epoch": 0.1185302002838669, + "grad_norm": 4.206762790679932, + "learning_rate": 1.9592256687809125e-05, + "loss": 0.5923, + "step": 3758 + }, + { + "epoch": 0.11856174105030752, + "grad_norm": 16.31416130065918, + "learning_rate": 1.9591967903684318e-05, + "loss": 2.0862, + "step": 3759 + }, + { + "epoch": 0.11859328181674815, + "grad_norm": 3.7869620323181152, + "learning_rate": 1.9591679019459874e-05, + "loss": 0.6304, + "step": 3760 + }, + { + "epoch": 0.11862482258318877, + "grad_norm": 2.969923973083496, + "learning_rate": 1.9591390035138812e-05, + "loss": 0.9746, + "step": 3761 + }, + { + "epoch": 0.1186563633496294, + "grad_norm": 4.516969203948975, + "learning_rate": 1.959110095072415e-05, + "loss": 0.6209, + "step": 3762 + }, + { + "epoch": 0.11868790411607003, + "grad_norm": 3.6692845821380615, + "learning_rate": 1.9590811766218898e-05, + "loss": 0.6156, + "step": 3763 + }, + { + "epoch": 0.11871944488251064, + "grad_norm": 6.274701118469238, + "learning_rate": 1.959052248162608e-05, + "loss": 0.6249, + "step": 3764 + }, + { + "epoch": 0.11875098564895127, + "grad_norm": 3.0353925228118896, + "learning_rate": 1.9590233096948716e-05, + "loss": 0.6161, + "step": 3765 + }, + { + "epoch": 0.1187825264153919, + "grad_norm": 5.91979455947876, + "learning_rate": 1.9589943612189818e-05, + "loss": 0.9926, + "step": 3766 + }, + { + "epoch": 0.11881406718183252, + "grad_norm": 3.082897186279297, + "learning_rate": 1.9589654027352412e-05, + "loss": 0.6067, + "step": 3767 + }, + { + "epoch": 0.11884560794827315, + "grad_norm": 2.4000742435455322, + "learning_rate": 1.9589364342439524e-05, + "loss": 1.0337, + "step": 3768 + }, + { + "epoch": 0.11887714871471376, + "grad_norm": 2.6378390789031982, + "learning_rate": 1.9589074557454172e-05, + "loss": 0.5989, + "step": 3769 + }, + { + "epoch": 0.11890868948115439, + "grad_norm": 3.9715018272399902, + "learning_rate": 1.958878467239938e-05, + "loss": 0.6073, + "step": 3770 + }, + { + "epoch": 0.11894023024759502, + "grad_norm": 3.8419413566589355, + "learning_rate": 1.9588494687278174e-05, + "loss": 0.6254, + "step": 3771 + }, + { + "epoch": 0.11897177101403564, + "grad_norm": 3.3331449031829834, + "learning_rate": 1.9588204602093585e-05, + "loss": 0.5979, + "step": 3772 + }, + { + "epoch": 0.11900331178047627, + "grad_norm": 7.8617753982543945, + "learning_rate": 1.958791441684863e-05, + "loss": 0.803, + "step": 3773 + }, + { + "epoch": 0.11903485254691688, + "grad_norm": 2.357598304748535, + "learning_rate": 1.958762413154635e-05, + "loss": 0.7062, + "step": 3774 + }, + { + "epoch": 0.11906639331335751, + "grad_norm": 2.090872049331665, + "learning_rate": 1.958733374618976e-05, + "loss": 1.0422, + "step": 3775 + }, + { + "epoch": 0.11909793407979814, + "grad_norm": 4.884191989898682, + "learning_rate": 1.95870432607819e-05, + "loss": 0.6478, + "step": 3776 + }, + { + "epoch": 0.11912947484623876, + "grad_norm": 4.630878925323486, + "learning_rate": 1.95867526753258e-05, + "loss": 0.65, + "step": 3777 + }, + { + "epoch": 0.11916101561267939, + "grad_norm": 4.095834732055664, + "learning_rate": 1.9586461989824497e-05, + "loss": 0.6537, + "step": 3778 + }, + { + "epoch": 0.11919255637912002, + "grad_norm": 4.202658176422119, + "learning_rate": 1.9586171204281012e-05, + "loss": 0.6289, + "step": 3779 + }, + { + "epoch": 0.11922409714556063, + "grad_norm": 2.8181850910186768, + "learning_rate": 1.9585880318698393e-05, + "loss": 0.5687, + "step": 3780 + }, + { + "epoch": 0.11925563791200126, + "grad_norm": 4.401491641998291, + "learning_rate": 1.9585589333079668e-05, + "loss": 0.6403, + "step": 3781 + }, + { + "epoch": 0.11928717867844188, + "grad_norm": 4.217578411102295, + "learning_rate": 1.9585298247427873e-05, + "loss": 0.6008, + "step": 3782 + }, + { + "epoch": 0.11931871944488251, + "grad_norm": 3.3676071166992188, + "learning_rate": 1.9585007061746048e-05, + "loss": 0.6261, + "step": 3783 + }, + { + "epoch": 0.11935026021132314, + "grad_norm": 7.388786792755127, + "learning_rate": 1.9584715776037234e-05, + "loss": 0.6875, + "step": 3784 + }, + { + "epoch": 0.11938180097776376, + "grad_norm": 3.254487991333008, + "learning_rate": 1.9584424390304465e-05, + "loss": 1.0406, + "step": 3785 + }, + { + "epoch": 0.11941334174420438, + "grad_norm": 4.527599811553955, + "learning_rate": 1.958413290455079e-05, + "loss": 0.6132, + "step": 3786 + }, + { + "epoch": 0.11944488251064501, + "grad_norm": 5.435885429382324, + "learning_rate": 1.9583841318779243e-05, + "loss": 0.6343, + "step": 3787 + }, + { + "epoch": 0.11947642327708563, + "grad_norm": 2.938444137573242, + "learning_rate": 1.9583549632992872e-05, + "loss": 0.6089, + "step": 3788 + }, + { + "epoch": 0.11950796404352626, + "grad_norm": 1.917851209640503, + "learning_rate": 1.958325784719472e-05, + "loss": 0.9601, + "step": 3789 + }, + { + "epoch": 0.11953950480996688, + "grad_norm": 4.152629375457764, + "learning_rate": 1.9582965961387826e-05, + "loss": 0.5945, + "step": 3790 + }, + { + "epoch": 0.1195710455764075, + "grad_norm": 6.079877853393555, + "learning_rate": 1.958267397557525e-05, + "loss": 0.6067, + "step": 3791 + }, + { + "epoch": 0.11960258634284814, + "grad_norm": 1.8050169944763184, + "learning_rate": 1.9582381889760022e-05, + "loss": 1.0013, + "step": 3792 + }, + { + "epoch": 0.11963412710928875, + "grad_norm": 3.3515357971191406, + "learning_rate": 1.9582089703945203e-05, + "loss": 0.609, + "step": 3793 + }, + { + "epoch": 0.11966566787572938, + "grad_norm": 2.7015578746795654, + "learning_rate": 1.9581797418133838e-05, + "loss": 0.5921, + "step": 3794 + }, + { + "epoch": 0.11969720864217001, + "grad_norm": 3.2285821437835693, + "learning_rate": 1.9581505032328973e-05, + "loss": 0.6136, + "step": 3795 + }, + { + "epoch": 0.11972874940861063, + "grad_norm": 3.0834851264953613, + "learning_rate": 1.9581212546533665e-05, + "loss": 0.6132, + "step": 3796 + }, + { + "epoch": 0.11976029017505126, + "grad_norm": 3.4485931396484375, + "learning_rate": 1.9580919960750965e-05, + "loss": 0.5965, + "step": 3797 + }, + { + "epoch": 0.11979183094149187, + "grad_norm": 2.7356762886047363, + "learning_rate": 1.9580627274983924e-05, + "loss": 0.5946, + "step": 3798 + }, + { + "epoch": 0.1198233717079325, + "grad_norm": 9.779808044433594, + "learning_rate": 1.95803344892356e-05, + "loss": 0.6152, + "step": 3799 + }, + { + "epoch": 0.11985491247437313, + "grad_norm": 3.4763989448547363, + "learning_rate": 1.9580041603509048e-05, + "loss": 0.6315, + "step": 3800 + }, + { + "epoch": 0.11988645324081375, + "grad_norm": 4.970693588256836, + "learning_rate": 1.9579748617807322e-05, + "loss": 0.6267, + "step": 3801 + }, + { + "epoch": 0.11991799400725438, + "grad_norm": 3.0632619857788086, + "learning_rate": 1.957945553213348e-05, + "loss": 0.5782, + "step": 3802 + }, + { + "epoch": 0.11994953477369501, + "grad_norm": 2.3821287155151367, + "learning_rate": 1.957916234649058e-05, + "loss": 0.9767, + "step": 3803 + }, + { + "epoch": 0.11998107554013562, + "grad_norm": 3.9660120010375977, + "learning_rate": 1.9578869060881684e-05, + "loss": 0.6136, + "step": 3804 + }, + { + "epoch": 0.12001261630657625, + "grad_norm": 5.326640605926514, + "learning_rate": 1.9578575675309856e-05, + "loss": 0.6148, + "step": 3805 + }, + { + "epoch": 0.12004415707301687, + "grad_norm": 2.897193670272827, + "learning_rate": 1.9578282189778148e-05, + "loss": 0.5969, + "step": 3806 + }, + { + "epoch": 0.1200756978394575, + "grad_norm": 5.440737247467041, + "learning_rate": 1.9577988604289627e-05, + "loss": 0.5959, + "step": 3807 + }, + { + "epoch": 0.12010723860589813, + "grad_norm": 3.6664929389953613, + "learning_rate": 1.957769491884736e-05, + "loss": 0.6071, + "step": 3808 + }, + { + "epoch": 0.12013877937233874, + "grad_norm": 3.200495481491089, + "learning_rate": 1.957740113345441e-05, + "loss": 0.6089, + "step": 3809 + }, + { + "epoch": 0.12017032013877937, + "grad_norm": 5.02022123336792, + "learning_rate": 1.9577107248113845e-05, + "loss": 0.5418, + "step": 3810 + }, + { + "epoch": 0.12020186090522, + "grad_norm": 3.5537824630737305, + "learning_rate": 1.9576813262828724e-05, + "loss": 0.6118, + "step": 3811 + }, + { + "epoch": 0.12023340167166062, + "grad_norm": 2.94384503364563, + "learning_rate": 1.9576519177602124e-05, + "loss": 0.5966, + "step": 3812 + }, + { + "epoch": 0.12026494243810125, + "grad_norm": 3.754328489303589, + "learning_rate": 1.957622499243711e-05, + "loss": 0.5797, + "step": 3813 + }, + { + "epoch": 0.12029648320454187, + "grad_norm": 4.197482109069824, + "learning_rate": 1.9575930707336752e-05, + "loss": 0.6242, + "step": 3814 + }, + { + "epoch": 0.1203280239709825, + "grad_norm": 2.540440320968628, + "learning_rate": 1.9575636322304122e-05, + "loss": 0.594, + "step": 3815 + }, + { + "epoch": 0.12035956473742312, + "grad_norm": 2.565169334411621, + "learning_rate": 1.957534183734229e-05, + "loss": 0.601, + "step": 3816 + }, + { + "epoch": 0.12039110550386374, + "grad_norm": 4.0102972984313965, + "learning_rate": 1.9575047252454335e-05, + "loss": 0.6175, + "step": 3817 + }, + { + "epoch": 0.12042264627030437, + "grad_norm": 3.701070785522461, + "learning_rate": 1.9574752567643324e-05, + "loss": 0.5734, + "step": 3818 + }, + { + "epoch": 0.12045418703674499, + "grad_norm": 2.7590718269348145, + "learning_rate": 1.9574457782912332e-05, + "loss": 0.6084, + "step": 3819 + }, + { + "epoch": 0.12048572780318562, + "grad_norm": 5.262868404388428, + "learning_rate": 1.9574162898264443e-05, + "loss": 0.6034, + "step": 3820 + }, + { + "epoch": 0.12051726856962625, + "grad_norm": 6.132518768310547, + "learning_rate": 1.957386791370273e-05, + "loss": 0.5815, + "step": 3821 + }, + { + "epoch": 0.12054880933606686, + "grad_norm": 2.347574472427368, + "learning_rate": 1.9573572829230268e-05, + "loss": 0.9548, + "step": 3822 + }, + { + "epoch": 0.12058035010250749, + "grad_norm": 4.603287220001221, + "learning_rate": 1.9573277644850145e-05, + "loss": 0.6134, + "step": 3823 + }, + { + "epoch": 0.12061189086894812, + "grad_norm": 9.321380615234375, + "learning_rate": 1.957298236056543e-05, + "loss": 0.5716, + "step": 3824 + }, + { + "epoch": 0.12064343163538874, + "grad_norm": 4.5286545753479, + "learning_rate": 1.9572686976379217e-05, + "loss": 0.6065, + "step": 3825 + }, + { + "epoch": 0.12067497240182937, + "grad_norm": 3.0510847568511963, + "learning_rate": 1.957239149229458e-05, + "loss": 0.5718, + "step": 3826 + }, + { + "epoch": 0.12070651316826998, + "grad_norm": 5.224523544311523, + "learning_rate": 1.9572095908314607e-05, + "loss": 0.5672, + "step": 3827 + }, + { + "epoch": 0.12073805393471061, + "grad_norm": 3.08856463432312, + "learning_rate": 1.9571800224442377e-05, + "loss": 0.6105, + "step": 3828 + }, + { + "epoch": 0.12076959470115124, + "grad_norm": 6.882884979248047, + "learning_rate": 1.957150444068098e-05, + "loss": 0.5894, + "step": 3829 + }, + { + "epoch": 0.12080113546759186, + "grad_norm": 8.517573356628418, + "learning_rate": 1.9571208557033503e-05, + "loss": 0.6196, + "step": 3830 + }, + { + "epoch": 0.12083267623403249, + "grad_norm": 3.606618881225586, + "learning_rate": 1.9570912573503032e-05, + "loss": 0.6263, + "step": 3831 + }, + { + "epoch": 0.12086421700047312, + "grad_norm": 3.183398962020874, + "learning_rate": 1.957061649009266e-05, + "loss": 0.575, + "step": 3832 + }, + { + "epoch": 0.12089575776691373, + "grad_norm": 3.4156510829925537, + "learning_rate": 1.957032030680547e-05, + "loss": 0.9468, + "step": 3833 + }, + { + "epoch": 0.12092729853335436, + "grad_norm": 21.258129119873047, + "learning_rate": 1.957002402364456e-05, + "loss": 0.6094, + "step": 3834 + }, + { + "epoch": 0.12095883929979498, + "grad_norm": 3.1507821083068848, + "learning_rate": 1.9569727640613014e-05, + "loss": 0.583, + "step": 3835 + }, + { + "epoch": 0.12099038006623561, + "grad_norm": 3.4573285579681396, + "learning_rate": 1.956943115771393e-05, + "loss": 0.6348, + "step": 3836 + }, + { + "epoch": 0.12102192083267624, + "grad_norm": 3.5927674770355225, + "learning_rate": 1.9569134574950405e-05, + "loss": 0.6054, + "step": 3837 + }, + { + "epoch": 0.12105346159911685, + "grad_norm": 3.456313371658325, + "learning_rate": 1.9568837892325526e-05, + "loss": 0.6143, + "step": 3838 + }, + { + "epoch": 0.12108500236555748, + "grad_norm": 4.349405288696289, + "learning_rate": 1.9568541109842395e-05, + "loss": 0.611, + "step": 3839 + }, + { + "epoch": 0.12111654313199811, + "grad_norm": 3.4663896560668945, + "learning_rate": 1.956824422750411e-05, + "loss": 0.6158, + "step": 3840 + }, + { + "epoch": 0.12114808389843873, + "grad_norm": 6.671550750732422, + "learning_rate": 1.9567947245313767e-05, + "loss": 0.576, + "step": 3841 + }, + { + "epoch": 0.12117962466487936, + "grad_norm": 3.564598798751831, + "learning_rate": 1.956765016327446e-05, + "loss": 0.5755, + "step": 3842 + }, + { + "epoch": 0.12121116543131998, + "grad_norm": 2.106168270111084, + "learning_rate": 1.9567352981389297e-05, + "loss": 0.9843, + "step": 3843 + }, + { + "epoch": 0.1212427061977606, + "grad_norm": 4.558315753936768, + "learning_rate": 1.956705569966138e-05, + "loss": 0.6368, + "step": 3844 + }, + { + "epoch": 0.12127424696420123, + "grad_norm": 3.8751699924468994, + "learning_rate": 1.9566758318093807e-05, + "loss": 0.6575, + "step": 3845 + }, + { + "epoch": 0.12130578773064185, + "grad_norm": 3.198014736175537, + "learning_rate": 1.956646083668968e-05, + "loss": 0.645, + "step": 3846 + }, + { + "epoch": 0.12133732849708248, + "grad_norm": 3.1776604652404785, + "learning_rate": 1.9566163255452107e-05, + "loss": 0.6018, + "step": 3847 + }, + { + "epoch": 0.12136886926352311, + "grad_norm": 2.563570976257324, + "learning_rate": 1.9565865574384195e-05, + "loss": 0.6015, + "step": 3848 + }, + { + "epoch": 0.12140041002996373, + "grad_norm": 8.400947570800781, + "learning_rate": 1.9565567793489048e-05, + "loss": 0.6046, + "step": 3849 + }, + { + "epoch": 0.12143195079640436, + "grad_norm": 1.7859163284301758, + "learning_rate": 1.9565269912769767e-05, + "loss": 0.9824, + "step": 3850 + }, + { + "epoch": 0.12146349156284497, + "grad_norm": 3.4490864276885986, + "learning_rate": 1.9564971932229473e-05, + "loss": 0.6217, + "step": 3851 + }, + { + "epoch": 0.1214950323292856, + "grad_norm": 3.6086204051971436, + "learning_rate": 1.956467385187127e-05, + "loss": 0.6599, + "step": 3852 + }, + { + "epoch": 0.12152657309572623, + "grad_norm": 3.5008742809295654, + "learning_rate": 1.9564375671698266e-05, + "loss": 0.6061, + "step": 3853 + }, + { + "epoch": 0.12155811386216685, + "grad_norm": 2.9967286586761475, + "learning_rate": 1.9564077391713576e-05, + "loss": 0.6011, + "step": 3854 + }, + { + "epoch": 0.12158965462860748, + "grad_norm": 6.894935607910156, + "learning_rate": 1.9563779011920315e-05, + "loss": 0.5804, + "step": 3855 + }, + { + "epoch": 0.12162119539504809, + "grad_norm": 4.475103378295898, + "learning_rate": 1.956348053232159e-05, + "loss": 0.9227, + "step": 3856 + }, + { + "epoch": 0.12165273616148872, + "grad_norm": 5.996613025665283, + "learning_rate": 1.9563181952920523e-05, + "loss": 0.6215, + "step": 3857 + }, + { + "epoch": 0.12168427692792935, + "grad_norm": 4.091640472412109, + "learning_rate": 1.9562883273720224e-05, + "loss": 0.5646, + "step": 3858 + }, + { + "epoch": 0.12171581769436997, + "grad_norm": 5.539577484130859, + "learning_rate": 1.9562584494723814e-05, + "loss": 0.5795, + "step": 3859 + }, + { + "epoch": 0.1217473584608106, + "grad_norm": 3.009808301925659, + "learning_rate": 1.956228561593441e-05, + "loss": 0.6382, + "step": 3860 + }, + { + "epoch": 0.12177889922725123, + "grad_norm": 6.1815924644470215, + "learning_rate": 1.956198663735513e-05, + "loss": 0.566, + "step": 3861 + }, + { + "epoch": 0.12181043999369184, + "grad_norm": 13.221380233764648, + "learning_rate": 1.9561687558989094e-05, + "loss": 0.5273, + "step": 3862 + }, + { + "epoch": 0.12184198076013247, + "grad_norm": 3.994936466217041, + "learning_rate": 1.9561388380839426e-05, + "loss": 0.6306, + "step": 3863 + }, + { + "epoch": 0.12187352152657309, + "grad_norm": 7.018301010131836, + "learning_rate": 1.9561089102909246e-05, + "loss": 0.6332, + "step": 3864 + }, + { + "epoch": 0.12190506229301372, + "grad_norm": 3.4688498973846436, + "learning_rate": 1.9560789725201675e-05, + "loss": 0.5668, + "step": 3865 + }, + { + "epoch": 0.12193660305945435, + "grad_norm": 4.742527961730957, + "learning_rate": 1.956049024771984e-05, + "loss": 0.6139, + "step": 3866 + }, + { + "epoch": 0.12196814382589496, + "grad_norm": 3.844604253768921, + "learning_rate": 1.9560190670466865e-05, + "loss": 0.6213, + "step": 3867 + }, + { + "epoch": 0.1219996845923356, + "grad_norm": 2.900442600250244, + "learning_rate": 1.955989099344588e-05, + "loss": 0.6319, + "step": 3868 + }, + { + "epoch": 0.12203122535877622, + "grad_norm": 9.527104377746582, + "learning_rate": 1.9559591216660007e-05, + "loss": 0.5777, + "step": 3869 + }, + { + "epoch": 0.12206276612521684, + "grad_norm": 2.011383533477783, + "learning_rate": 1.9559291340112378e-05, + "loss": 0.9526, + "step": 3870 + }, + { + "epoch": 0.12209430689165747, + "grad_norm": 2.908942461013794, + "learning_rate": 1.9558991363806118e-05, + "loss": 0.5962, + "step": 3871 + }, + { + "epoch": 0.12212584765809809, + "grad_norm": 5.165131568908691, + "learning_rate": 1.9558691287744365e-05, + "loss": 0.5808, + "step": 3872 + }, + { + "epoch": 0.12215738842453872, + "grad_norm": 1.7090743780136108, + "learning_rate": 1.9558391111930245e-05, + "loss": 0.9272, + "step": 3873 + }, + { + "epoch": 0.12218892919097935, + "grad_norm": 3.4498093128204346, + "learning_rate": 1.9558090836366893e-05, + "loss": 0.5797, + "step": 3874 + }, + { + "epoch": 0.12222046995741996, + "grad_norm": 3.840846300125122, + "learning_rate": 1.955779046105744e-05, + "loss": 0.6192, + "step": 3875 + }, + { + "epoch": 0.12225201072386059, + "grad_norm": 3.634138345718384, + "learning_rate": 1.9557489986005023e-05, + "loss": 0.638, + "step": 3876 + }, + { + "epoch": 0.12228355149030122, + "grad_norm": 2.7607626914978027, + "learning_rate": 1.9557189411212774e-05, + "loss": 0.5854, + "step": 3877 + }, + { + "epoch": 0.12231509225674184, + "grad_norm": 2.688655138015747, + "learning_rate": 1.9556888736683836e-05, + "loss": 0.6014, + "step": 3878 + }, + { + "epoch": 0.12234663302318247, + "grad_norm": 7.313781261444092, + "learning_rate": 1.955658796242134e-05, + "loss": 0.5948, + "step": 3879 + }, + { + "epoch": 0.12237817378962308, + "grad_norm": 8.261552810668945, + "learning_rate": 1.9556287088428433e-05, + "loss": 0.5924, + "step": 3880 + }, + { + "epoch": 0.12240971455606371, + "grad_norm": 3.4415853023529053, + "learning_rate": 1.9555986114708246e-05, + "loss": 0.5545, + "step": 3881 + }, + { + "epoch": 0.12244125532250434, + "grad_norm": 3.301738977432251, + "learning_rate": 1.9555685041263924e-05, + "loss": 0.5908, + "step": 3882 + }, + { + "epoch": 0.12247279608894496, + "grad_norm": 3.831775426864624, + "learning_rate": 1.955538386809861e-05, + "loss": 0.4831, + "step": 3883 + }, + { + "epoch": 0.12250433685538559, + "grad_norm": 3.120333671569824, + "learning_rate": 1.9555082595215445e-05, + "loss": 0.5984, + "step": 3884 + }, + { + "epoch": 0.12253587762182622, + "grad_norm": 14.545559883117676, + "learning_rate": 1.955478122261757e-05, + "loss": 0.4771, + "step": 3885 + }, + { + "epoch": 0.12256741838826683, + "grad_norm": 2.364999294281006, + "learning_rate": 1.955447975030814e-05, + "loss": 0.9696, + "step": 3886 + }, + { + "epoch": 0.12259895915470746, + "grad_norm": 4.150489330291748, + "learning_rate": 1.9554178178290294e-05, + "loss": 0.6247, + "step": 3887 + }, + { + "epoch": 0.12263049992114808, + "grad_norm": 3.2260031700134277, + "learning_rate": 1.955387650656718e-05, + "loss": 0.6275, + "step": 3888 + }, + { + "epoch": 0.12266204068758871, + "grad_norm": 3.9737708568573, + "learning_rate": 1.9553574735141946e-05, + "loss": 0.6106, + "step": 3889 + }, + { + "epoch": 0.12269358145402934, + "grad_norm": 1.8615566492080688, + "learning_rate": 1.9553272864017738e-05, + "loss": 0.9708, + "step": 3890 + }, + { + "epoch": 0.12272512222046995, + "grad_norm": 3.1075961589813232, + "learning_rate": 1.9552970893197712e-05, + "loss": 0.6221, + "step": 3891 + }, + { + "epoch": 0.12275666298691058, + "grad_norm": 3.9900460243225098, + "learning_rate": 1.9552668822685018e-05, + "loss": 0.6001, + "step": 3892 + }, + { + "epoch": 0.12278820375335121, + "grad_norm": 3.327033042907715, + "learning_rate": 1.9552366652482808e-05, + "loss": 0.6284, + "step": 3893 + }, + { + "epoch": 0.12281974451979183, + "grad_norm": 4.236921787261963, + "learning_rate": 1.9552064382594232e-05, + "loss": 0.5853, + "step": 3894 + }, + { + "epoch": 0.12285128528623246, + "grad_norm": 3.6070733070373535, + "learning_rate": 1.955176201302245e-05, + "loss": 0.5896, + "step": 3895 + }, + { + "epoch": 0.12288282605267307, + "grad_norm": 6.80300760269165, + "learning_rate": 1.9551459543770614e-05, + "loss": 0.5083, + "step": 3896 + }, + { + "epoch": 0.1229143668191137, + "grad_norm": 5.17471981048584, + "learning_rate": 1.955115697484188e-05, + "loss": 0.5909, + "step": 3897 + }, + { + "epoch": 0.12294590758555433, + "grad_norm": 3.1016488075256348, + "learning_rate": 1.9550854306239406e-05, + "loss": 0.6273, + "step": 3898 + }, + { + "epoch": 0.12297744835199495, + "grad_norm": 4.371035099029541, + "learning_rate": 1.9550551537966354e-05, + "loss": 0.6034, + "step": 3899 + }, + { + "epoch": 0.12300898911843558, + "grad_norm": 1.9317653179168701, + "learning_rate": 1.955024867002588e-05, + "loss": 0.9825, + "step": 3900 + }, + { + "epoch": 0.1230405298848762, + "grad_norm": 1.8262300491333008, + "learning_rate": 1.9549945702421144e-05, + "loss": 0.9465, + "step": 3901 + }, + { + "epoch": 0.12307207065131683, + "grad_norm": 2.6463513374328613, + "learning_rate": 1.954964263515531e-05, + "loss": 0.5853, + "step": 3902 + }, + { + "epoch": 0.12310361141775746, + "grad_norm": 2.3052196502685547, + "learning_rate": 1.9549339468231542e-05, + "loss": 0.9553, + "step": 3903 + }, + { + "epoch": 0.12313515218419807, + "grad_norm": 3.122018337249756, + "learning_rate": 1.9549036201653e-05, + "loss": 0.6123, + "step": 3904 + }, + { + "epoch": 0.1231666929506387, + "grad_norm": 2.2045516967773438, + "learning_rate": 1.954873283542285e-05, + "loss": 0.5778, + "step": 3905 + }, + { + "epoch": 0.12319823371707933, + "grad_norm": 1.7400434017181396, + "learning_rate": 1.954842936954426e-05, + "loss": 0.9728, + "step": 3906 + }, + { + "epoch": 0.12322977448351995, + "grad_norm": 8.624736785888672, + "learning_rate": 1.9548125804020392e-05, + "loss": 0.6225, + "step": 3907 + }, + { + "epoch": 0.12326131524996058, + "grad_norm": 3.5499088764190674, + "learning_rate": 1.9547822138854422e-05, + "loss": 0.6031, + "step": 3908 + }, + { + "epoch": 0.12329285601640119, + "grad_norm": 9.72619342803955, + "learning_rate": 1.9547518374049514e-05, + "loss": 0.5855, + "step": 3909 + }, + { + "epoch": 0.12332439678284182, + "grad_norm": 4.35174560546875, + "learning_rate": 1.954721450960883e-05, + "loss": 0.6117, + "step": 3910 + }, + { + "epoch": 0.12335593754928245, + "grad_norm": 2.314366579055786, + "learning_rate": 1.954691054553556e-05, + "loss": 0.5937, + "step": 3911 + }, + { + "epoch": 0.12338747831572307, + "grad_norm": 3.7443289756774902, + "learning_rate": 1.954660648183286e-05, + "loss": 0.5807, + "step": 3912 + }, + { + "epoch": 0.1234190190821637, + "grad_norm": 6.167291164398193, + "learning_rate": 1.954630231850391e-05, + "loss": 0.6062, + "step": 3913 + }, + { + "epoch": 0.12345055984860433, + "grad_norm": 5.652634620666504, + "learning_rate": 1.9545998055551883e-05, + "loss": 0.5668, + "step": 3914 + }, + { + "epoch": 0.12348210061504494, + "grad_norm": 3.454998016357422, + "learning_rate": 1.9545693692979954e-05, + "loss": 0.4923, + "step": 3915 + }, + { + "epoch": 0.12351364138148557, + "grad_norm": 2.8649208545684814, + "learning_rate": 1.9545389230791295e-05, + "loss": 0.6194, + "step": 3916 + }, + { + "epoch": 0.12354518214792619, + "grad_norm": 3.1697371006011963, + "learning_rate": 1.9545084668989095e-05, + "loss": 0.6069, + "step": 3917 + }, + { + "epoch": 0.12357672291436682, + "grad_norm": 2.3298425674438477, + "learning_rate": 1.954478000757652e-05, + "loss": 0.6259, + "step": 3918 + }, + { + "epoch": 0.12360826368080745, + "grad_norm": 2.4747447967529297, + "learning_rate": 1.9544475246556756e-05, + "loss": 0.6313, + "step": 3919 + }, + { + "epoch": 0.12363980444724806, + "grad_norm": 1.7410452365875244, + "learning_rate": 1.9544170385932984e-05, + "loss": 0.9326, + "step": 3920 + }, + { + "epoch": 0.1236713452136887, + "grad_norm": 3.946232795715332, + "learning_rate": 1.9543865425708377e-05, + "loss": 0.6132, + "step": 3921 + }, + { + "epoch": 0.12370288598012932, + "grad_norm": 3.45658016204834, + "learning_rate": 1.9543560365886128e-05, + "loss": 0.5923, + "step": 3922 + }, + { + "epoch": 0.12373442674656994, + "grad_norm": 2.1030113697052, + "learning_rate": 1.9543255206469417e-05, + "loss": 0.5887, + "step": 3923 + }, + { + "epoch": 0.12376596751301057, + "grad_norm": 2.2657828330993652, + "learning_rate": 1.9542949947461425e-05, + "loss": 0.6052, + "step": 3924 + }, + { + "epoch": 0.12379750827945118, + "grad_norm": 2.376772880554199, + "learning_rate": 1.954264458886534e-05, + "loss": 0.604, + "step": 3925 + }, + { + "epoch": 0.12382904904589181, + "grad_norm": 1.7040245532989502, + "learning_rate": 1.954233913068435e-05, + "loss": 0.9681, + "step": 3926 + }, + { + "epoch": 0.12386058981233244, + "grad_norm": 1.8743681907653809, + "learning_rate": 1.9542033572921644e-05, + "loss": 0.9637, + "step": 3927 + }, + { + "epoch": 0.12389213057877306, + "grad_norm": 8.437690734863281, + "learning_rate": 1.954172791558041e-05, + "loss": 0.5863, + "step": 3928 + }, + { + "epoch": 0.12392367134521369, + "grad_norm": 3.447131395339966, + "learning_rate": 1.9541422158663832e-05, + "loss": 0.5938, + "step": 3929 + }, + { + "epoch": 0.12395521211165432, + "grad_norm": 3.394407033920288, + "learning_rate": 1.9541116302175106e-05, + "loss": 0.6035, + "step": 3930 + }, + { + "epoch": 0.12398675287809494, + "grad_norm": 1.9978218078613281, + "learning_rate": 1.954081034611742e-05, + "loss": 0.9399, + "step": 3931 + }, + { + "epoch": 0.12401829364453557, + "grad_norm": 2.471787452697754, + "learning_rate": 1.9540504290493976e-05, + "loss": 0.5514, + "step": 3932 + }, + { + "epoch": 0.12404983441097618, + "grad_norm": 3.4226205348968506, + "learning_rate": 1.9540198135307958e-05, + "loss": 0.6075, + "step": 3933 + }, + { + "epoch": 0.12408137517741681, + "grad_norm": 2.993842601776123, + "learning_rate": 1.9539891880562565e-05, + "loss": 0.5834, + "step": 3934 + }, + { + "epoch": 0.12411291594385744, + "grad_norm": 2.9092676639556885, + "learning_rate": 1.9539585526260995e-05, + "loss": 0.5828, + "step": 3935 + }, + { + "epoch": 0.12414445671029806, + "grad_norm": 2.5512659549713135, + "learning_rate": 1.953927907240644e-05, + "loss": 0.605, + "step": 3936 + }, + { + "epoch": 0.12417599747673869, + "grad_norm": 2.4339041709899902, + "learning_rate": 1.95389725190021e-05, + "loss": 0.5608, + "step": 3937 + }, + { + "epoch": 0.1242075382431793, + "grad_norm": 2.739859104156494, + "learning_rate": 1.9538665866051175e-05, + "loss": 0.5645, + "step": 3938 + }, + { + "epoch": 0.12423907900961993, + "grad_norm": 3.4686896800994873, + "learning_rate": 1.953835911355687e-05, + "loss": 0.6083, + "step": 3939 + }, + { + "epoch": 0.12427061977606056, + "grad_norm": 5.101952075958252, + "learning_rate": 1.9538052261522375e-05, + "loss": 0.5132, + "step": 3940 + }, + { + "epoch": 0.12430216054250118, + "grad_norm": 4.20035982131958, + "learning_rate": 1.95377453099509e-05, + "loss": 0.4733, + "step": 3941 + }, + { + "epoch": 0.12433370130894181, + "grad_norm": 3.336825370788574, + "learning_rate": 1.953743825884565e-05, + "loss": 0.6064, + "step": 3942 + }, + { + "epoch": 0.12436524207538244, + "grad_norm": 2.404437303543091, + "learning_rate": 1.9537131108209822e-05, + "loss": 0.5975, + "step": 3943 + }, + { + "epoch": 0.12439678284182305, + "grad_norm": 3.120166778564453, + "learning_rate": 1.953682385804663e-05, + "loss": 0.6274, + "step": 3944 + }, + { + "epoch": 0.12442832360826368, + "grad_norm": 3.0709116458892822, + "learning_rate": 1.9536516508359273e-05, + "loss": 0.5996, + "step": 3945 + }, + { + "epoch": 0.1244598643747043, + "grad_norm": 7.106463432312012, + "learning_rate": 1.9536209059150963e-05, + "loss": 0.4944, + "step": 3946 + }, + { + "epoch": 0.12449140514114493, + "grad_norm": 2.7913074493408203, + "learning_rate": 1.9535901510424903e-05, + "loss": 0.594, + "step": 3947 + }, + { + "epoch": 0.12452294590758556, + "grad_norm": 2.686375856399536, + "learning_rate": 1.953559386218431e-05, + "loss": 0.6034, + "step": 3948 + }, + { + "epoch": 0.12455448667402617, + "grad_norm": 2.6742048263549805, + "learning_rate": 1.9535286114432394e-05, + "loss": 0.6017, + "step": 3949 + }, + { + "epoch": 0.1245860274404668, + "grad_norm": 3.0578956604003906, + "learning_rate": 1.9534978267172357e-05, + "loss": 0.582, + "step": 3950 + }, + { + "epoch": 0.12461756820690743, + "grad_norm": 3.209693193435669, + "learning_rate": 1.953467032040742e-05, + "loss": 0.6127, + "step": 3951 + }, + { + "epoch": 0.12464910897334805, + "grad_norm": 2.4801688194274902, + "learning_rate": 1.9534362274140797e-05, + "loss": 0.6099, + "step": 3952 + }, + { + "epoch": 0.12468064973978868, + "grad_norm": 5.596282482147217, + "learning_rate": 1.9534054128375698e-05, + "loss": 0.4575, + "step": 3953 + }, + { + "epoch": 0.1247121905062293, + "grad_norm": 2.1423163414001465, + "learning_rate": 1.953374588311534e-05, + "loss": 0.5341, + "step": 3954 + }, + { + "epoch": 0.12474373127266992, + "grad_norm": 2.8999030590057373, + "learning_rate": 1.9533437538362943e-05, + "loss": 0.6186, + "step": 3955 + }, + { + "epoch": 0.12477527203911055, + "grad_norm": 4.202732563018799, + "learning_rate": 1.9533129094121722e-05, + "loss": 0.6315, + "step": 3956 + }, + { + "epoch": 0.12480681280555117, + "grad_norm": 12.913784980773926, + "learning_rate": 1.9532820550394896e-05, + "loss": 0.6239, + "step": 3957 + }, + { + "epoch": 0.1248383535719918, + "grad_norm": 2.6100926399230957, + "learning_rate": 1.953251190718569e-05, + "loss": 0.5872, + "step": 3958 + }, + { + "epoch": 0.12486989433843243, + "grad_norm": 2.503495454788208, + "learning_rate": 1.9532203164497316e-05, + "loss": 0.5868, + "step": 3959 + }, + { + "epoch": 0.12490143510487305, + "grad_norm": 2.503694534301758, + "learning_rate": 1.9531894322333e-05, + "loss": 0.5898, + "step": 3960 + }, + { + "epoch": 0.12493297587131368, + "grad_norm": 3.166774034500122, + "learning_rate": 1.9531585380695967e-05, + "loss": 0.6008, + "step": 3961 + }, + { + "epoch": 0.12496451663775429, + "grad_norm": 2.5987465381622314, + "learning_rate": 1.953127633958944e-05, + "loss": 0.595, + "step": 3962 + }, + { + "epoch": 0.12499605740419492, + "grad_norm": 2.1980836391448975, + "learning_rate": 1.9530967199016645e-05, + "loss": 0.6073, + "step": 3963 + }, + { + "epoch": 0.12502759817063555, + "grad_norm": 3.1509084701538086, + "learning_rate": 1.9530657958980804e-05, + "loss": 0.6069, + "step": 3964 + }, + { + "epoch": 0.12505913893707618, + "grad_norm": 3.053011655807495, + "learning_rate": 1.9530348619485145e-05, + "loss": 0.6032, + "step": 3965 + }, + { + "epoch": 0.12509067970351678, + "grad_norm": 2.9610040187835693, + "learning_rate": 1.9530039180532904e-05, + "loss": 0.5901, + "step": 3966 + }, + { + "epoch": 0.1251222204699574, + "grad_norm": 2.513859748840332, + "learning_rate": 1.9529729642127298e-05, + "loss": 0.5929, + "step": 3967 + }, + { + "epoch": 0.12515376123639804, + "grad_norm": 3.5827884674072266, + "learning_rate": 1.9529420004271568e-05, + "loss": 0.5899, + "step": 3968 + }, + { + "epoch": 0.12518530200283867, + "grad_norm": 3.003756046295166, + "learning_rate": 1.952911026696894e-05, + "loss": 0.6164, + "step": 3969 + }, + { + "epoch": 0.1252168427692793, + "grad_norm": 1.8120872974395752, + "learning_rate": 1.9528800430222645e-05, + "loss": 0.9064, + "step": 3970 + }, + { + "epoch": 0.12524838353571993, + "grad_norm": 4.15649938583374, + "learning_rate": 1.952849049403592e-05, + "loss": 0.6075, + "step": 3971 + }, + { + "epoch": 0.12527992430216053, + "grad_norm": 2.61372709274292, + "learning_rate": 1.9528180458411997e-05, + "loss": 0.6015, + "step": 3972 + }, + { + "epoch": 0.12531146506860116, + "grad_norm": 2.8963654041290283, + "learning_rate": 1.9527870323354114e-05, + "loss": 0.613, + "step": 3973 + }, + { + "epoch": 0.1253430058350418, + "grad_norm": 2.583305835723877, + "learning_rate": 1.9527560088865506e-05, + "loss": 0.5993, + "step": 3974 + }, + { + "epoch": 0.12537454660148242, + "grad_norm": 2.876589059829712, + "learning_rate": 1.952724975494941e-05, + "loss": 0.6227, + "step": 3975 + }, + { + "epoch": 0.12540608736792305, + "grad_norm": 3.0188775062561035, + "learning_rate": 1.9526939321609063e-05, + "loss": 0.6132, + "step": 3976 + }, + { + "epoch": 0.12543762813436365, + "grad_norm": 2.832472562789917, + "learning_rate": 1.952662878884771e-05, + "loss": 0.551, + "step": 3977 + }, + { + "epoch": 0.12546916890080428, + "grad_norm": 2.984382390975952, + "learning_rate": 1.9526318156668585e-05, + "loss": 0.6203, + "step": 3978 + }, + { + "epoch": 0.12550070966724491, + "grad_norm": 4.308155536651611, + "learning_rate": 1.9526007425074938e-05, + "loss": 0.5867, + "step": 3979 + }, + { + "epoch": 0.12553225043368554, + "grad_norm": 2.5650129318237305, + "learning_rate": 1.9525696594070004e-05, + "loss": 0.5818, + "step": 3980 + }, + { + "epoch": 0.12556379120012617, + "grad_norm": 3.1876494884490967, + "learning_rate": 1.952538566365703e-05, + "loss": 0.5745, + "step": 3981 + }, + { + "epoch": 0.12559533196656678, + "grad_norm": 3.2186686992645264, + "learning_rate": 1.9525074633839257e-05, + "loss": 0.6029, + "step": 3982 + }, + { + "epoch": 0.1256268727330074, + "grad_norm": 5.102095127105713, + "learning_rate": 1.9524763504619938e-05, + "loss": 0.5732, + "step": 3983 + }, + { + "epoch": 0.12565841349944804, + "grad_norm": 2.904669761657715, + "learning_rate": 1.9524452276002316e-05, + "loss": 0.6009, + "step": 3984 + }, + { + "epoch": 0.12568995426588866, + "grad_norm": 3.2606871128082275, + "learning_rate": 1.9524140947989636e-05, + "loss": 0.6154, + "step": 3985 + }, + { + "epoch": 0.1257214950323293, + "grad_norm": 4.923027038574219, + "learning_rate": 1.9523829520585152e-05, + "loss": 0.6412, + "step": 3986 + }, + { + "epoch": 0.1257530357987699, + "grad_norm": 3.123669147491455, + "learning_rate": 1.9523517993792112e-05, + "loss": 0.6082, + "step": 3987 + }, + { + "epoch": 0.12578457656521053, + "grad_norm": 3.821441650390625, + "learning_rate": 1.9523206367613766e-05, + "loss": 0.6468, + "step": 3988 + }, + { + "epoch": 0.12581611733165116, + "grad_norm": 3.761082172393799, + "learning_rate": 1.952289464205337e-05, + "loss": 0.5949, + "step": 3989 + }, + { + "epoch": 0.12584765809809179, + "grad_norm": 6.85695743560791, + "learning_rate": 1.952258281711417e-05, + "loss": 0.5956, + "step": 3990 + }, + { + "epoch": 0.12587919886453242, + "grad_norm": 2.8126180171966553, + "learning_rate": 1.9522270892799424e-05, + "loss": 0.5997, + "step": 3991 + }, + { + "epoch": 0.12591073963097305, + "grad_norm": 5.416510105133057, + "learning_rate": 1.9521958869112388e-05, + "loss": 0.478, + "step": 3992 + }, + { + "epoch": 0.12594228039741365, + "grad_norm": 3.8621068000793457, + "learning_rate": 1.9521646746056322e-05, + "loss": 0.551, + "step": 3993 + }, + { + "epoch": 0.12597382116385428, + "grad_norm": 3.5245957374572754, + "learning_rate": 1.9521334523634476e-05, + "loss": 0.6061, + "step": 3994 + }, + { + "epoch": 0.1260053619302949, + "grad_norm": 3.139739751815796, + "learning_rate": 1.9521022201850108e-05, + "loss": 0.5886, + "step": 3995 + }, + { + "epoch": 0.12603690269673554, + "grad_norm": 2.8429927825927734, + "learning_rate": 1.9520709780706485e-05, + "loss": 0.6005, + "step": 3996 + }, + { + "epoch": 0.12606844346317617, + "grad_norm": 2.6000072956085205, + "learning_rate": 1.952039726020686e-05, + "loss": 0.5824, + "step": 3997 + }, + { + "epoch": 0.12609998422961677, + "grad_norm": 6.669317722320557, + "learning_rate": 1.95200846403545e-05, + "loss": 0.4455, + "step": 3998 + }, + { + "epoch": 0.1261315249960574, + "grad_norm": 17.615602493286133, + "learning_rate": 1.9519771921152663e-05, + "loss": 0.9338, + "step": 3999 + }, + { + "epoch": 0.12616306576249803, + "grad_norm": 3.514356851577759, + "learning_rate": 1.951945910260461e-05, + "loss": 0.6173, + "step": 4000 + }, + { + "epoch": 0.12619460652893866, + "grad_norm": 3.0936648845672607, + "learning_rate": 1.9519146184713617e-05, + "loss": 0.5606, + "step": 4001 + }, + { + "epoch": 0.1262261472953793, + "grad_norm": 3.144794464111328, + "learning_rate": 1.9518833167482937e-05, + "loss": 0.6262, + "step": 4002 + }, + { + "epoch": 0.1262576880618199, + "grad_norm": 3.116441011428833, + "learning_rate": 1.9518520050915843e-05, + "loss": 0.5819, + "step": 4003 + }, + { + "epoch": 0.12628922882826052, + "grad_norm": 1.8181068897247314, + "learning_rate": 1.95182068350156e-05, + "loss": 0.4253, + "step": 4004 + }, + { + "epoch": 0.12632076959470115, + "grad_norm": 4.010478496551514, + "learning_rate": 1.9517893519785484e-05, + "loss": 0.4319, + "step": 4005 + }, + { + "epoch": 0.12635231036114178, + "grad_norm": 3.285918712615967, + "learning_rate": 1.9517580105228753e-05, + "loss": 0.639, + "step": 4006 + }, + { + "epoch": 0.1263838511275824, + "grad_norm": 3.4926857948303223, + "learning_rate": 1.9517266591348684e-05, + "loss": 0.5745, + "step": 4007 + }, + { + "epoch": 0.12641539189402304, + "grad_norm": 2.6336355209350586, + "learning_rate": 1.951695297814855e-05, + "loss": 0.6035, + "step": 4008 + }, + { + "epoch": 0.12644693266046364, + "grad_norm": 2.875861406326294, + "learning_rate": 1.951663926563162e-05, + "loss": 0.6015, + "step": 4009 + }, + { + "epoch": 0.12647847342690427, + "grad_norm": 2.332951307296753, + "learning_rate": 1.9516325453801174e-05, + "loss": 0.5839, + "step": 4010 + }, + { + "epoch": 0.1265100141933449, + "grad_norm": 2.347815990447998, + "learning_rate": 1.9516011542660478e-05, + "loss": 0.6539, + "step": 4011 + }, + { + "epoch": 0.12654155495978553, + "grad_norm": 3.869731903076172, + "learning_rate": 1.9515697532212812e-05, + "loss": 0.5907, + "step": 4012 + }, + { + "epoch": 0.12657309572622616, + "grad_norm": 3.0076892375946045, + "learning_rate": 1.9515383422461457e-05, + "loss": 0.5387, + "step": 4013 + }, + { + "epoch": 0.12660463649266676, + "grad_norm": 2.7471485137939453, + "learning_rate": 1.9515069213409684e-05, + "loss": 0.601, + "step": 4014 + }, + { + "epoch": 0.1266361772591074, + "grad_norm": 2.642251968383789, + "learning_rate": 1.9514754905060778e-05, + "loss": 0.568, + "step": 4015 + }, + { + "epoch": 0.12666771802554802, + "grad_norm": 2.802473306655884, + "learning_rate": 1.9514440497418015e-05, + "loss": 0.6107, + "step": 4016 + }, + { + "epoch": 0.12669925879198865, + "grad_norm": 2.797393560409546, + "learning_rate": 1.9514125990484677e-05, + "loss": 0.589, + "step": 4017 + }, + { + "epoch": 0.12673079955842928, + "grad_norm": 3.150066614151001, + "learning_rate": 1.9513811384264048e-05, + "loss": 0.9624, + "step": 4018 + }, + { + "epoch": 0.12676234032486988, + "grad_norm": 2.9072251319885254, + "learning_rate": 1.9513496678759415e-05, + "loss": 0.6071, + "step": 4019 + }, + { + "epoch": 0.1267938810913105, + "grad_norm": 2.4266676902770996, + "learning_rate": 1.951318187397405e-05, + "loss": 0.4337, + "step": 4020 + }, + { + "epoch": 0.12682542185775114, + "grad_norm": 2.313995599746704, + "learning_rate": 1.9512866969911246e-05, + "loss": 0.6099, + "step": 4021 + }, + { + "epoch": 0.12685696262419177, + "grad_norm": 2.3160336017608643, + "learning_rate": 1.951255196657429e-05, + "loss": 0.5724, + "step": 4022 + }, + { + "epoch": 0.1268885033906324, + "grad_norm": 1.9547613859176636, + "learning_rate": 1.951223686396647e-05, + "loss": 0.9255, + "step": 4023 + }, + { + "epoch": 0.126920044157073, + "grad_norm": 3.3260231018066406, + "learning_rate": 1.9511921662091068e-05, + "loss": 0.5551, + "step": 4024 + }, + { + "epoch": 0.12695158492351363, + "grad_norm": 52.447181701660156, + "learning_rate": 1.9511606360951382e-05, + "loss": 0.919, + "step": 4025 + }, + { + "epoch": 0.12698312568995426, + "grad_norm": 2.3654093742370605, + "learning_rate": 1.9511290960550697e-05, + "loss": 0.5795, + "step": 4026 + }, + { + "epoch": 0.1270146664563949, + "grad_norm": 2.8001081943511963, + "learning_rate": 1.9510975460892302e-05, + "loss": 0.5992, + "step": 4027 + }, + { + "epoch": 0.12704620722283552, + "grad_norm": 2.4352405071258545, + "learning_rate": 1.9510659861979498e-05, + "loss": 0.605, + "step": 4028 + }, + { + "epoch": 0.12707774798927615, + "grad_norm": 6.172600746154785, + "learning_rate": 1.951034416381557e-05, + "loss": 0.5894, + "step": 4029 + }, + { + "epoch": 0.12710928875571675, + "grad_norm": 2.6104042530059814, + "learning_rate": 1.9510028366403818e-05, + "loss": 0.6101, + "step": 4030 + }, + { + "epoch": 0.12714082952215738, + "grad_norm": 3.1142685413360596, + "learning_rate": 1.9509712469747535e-05, + "loss": 0.6211, + "step": 4031 + }, + { + "epoch": 0.127172370288598, + "grad_norm": 3.5348498821258545, + "learning_rate": 1.9509396473850016e-05, + "loss": 0.5779, + "step": 4032 + }, + { + "epoch": 0.12720391105503864, + "grad_norm": 3.0237159729003906, + "learning_rate": 1.9509080378714564e-05, + "loss": 0.6166, + "step": 4033 + }, + { + "epoch": 0.12723545182147927, + "grad_norm": 3.670351982116699, + "learning_rate": 1.9508764184344473e-05, + "loss": 0.5065, + "step": 4034 + }, + { + "epoch": 0.12726699258791987, + "grad_norm": 2.444959878921509, + "learning_rate": 1.950844789074305e-05, + "loss": 0.599, + "step": 4035 + }, + { + "epoch": 0.1272985333543605, + "grad_norm": 2.805448532104492, + "learning_rate": 1.950813149791358e-05, + "loss": 0.6281, + "step": 4036 + }, + { + "epoch": 0.12733007412080113, + "grad_norm": 3.8829283714294434, + "learning_rate": 1.9507815005859382e-05, + "loss": 0.6078, + "step": 4037 + }, + { + "epoch": 0.12736161488724176, + "grad_norm": 2.652451992034912, + "learning_rate": 1.950749841458375e-05, + "loss": 0.6029, + "step": 4038 + }, + { + "epoch": 0.1273931556536824, + "grad_norm": 15.668305397033691, + "learning_rate": 1.950718172408999e-05, + "loss": 0.4336, + "step": 4039 + }, + { + "epoch": 0.127424696420123, + "grad_norm": 2.348478078842163, + "learning_rate": 1.9506864934381403e-05, + "loss": 0.5726, + "step": 4040 + }, + { + "epoch": 0.12745623718656363, + "grad_norm": 3.518960475921631, + "learning_rate": 1.9506548045461303e-05, + "loss": 0.6585, + "step": 4041 + }, + { + "epoch": 0.12748777795300426, + "grad_norm": 2.6818385124206543, + "learning_rate": 1.9506231057332995e-05, + "loss": 0.5747, + "step": 4042 + }, + { + "epoch": 0.12751931871944489, + "grad_norm": 3.163466453552246, + "learning_rate": 1.9505913969999778e-05, + "loss": 0.4081, + "step": 4043 + }, + { + "epoch": 0.12755085948588551, + "grad_norm": 2.849989414215088, + "learning_rate": 1.9505596783464972e-05, + "loss": 0.5598, + "step": 4044 + }, + { + "epoch": 0.12758240025232614, + "grad_norm": 3.319054365158081, + "learning_rate": 1.9505279497731878e-05, + "loss": 0.6062, + "step": 4045 + }, + { + "epoch": 0.12761394101876675, + "grad_norm": 4.155129909515381, + "learning_rate": 1.9504962112803816e-05, + "loss": 0.6164, + "step": 4046 + }, + { + "epoch": 0.12764548178520738, + "grad_norm": 2.819620370864868, + "learning_rate": 1.9504644628684093e-05, + "loss": 0.5947, + "step": 4047 + }, + { + "epoch": 0.127677022551648, + "grad_norm": 2.5323266983032227, + "learning_rate": 1.9504327045376022e-05, + "loss": 0.6147, + "step": 4048 + }, + { + "epoch": 0.12770856331808864, + "grad_norm": 3.215818166732788, + "learning_rate": 1.950400936288292e-05, + "loss": 0.5926, + "step": 4049 + }, + { + "epoch": 0.12774010408452927, + "grad_norm": 2.883509635925293, + "learning_rate": 1.95036915812081e-05, + "loss": 0.5779, + "step": 4050 + }, + { + "epoch": 0.12777164485096987, + "grad_norm": 2.7416610717773438, + "learning_rate": 1.950337370035488e-05, + "loss": 0.6206, + "step": 4051 + }, + { + "epoch": 0.1278031856174105, + "grad_norm": 2.673997163772583, + "learning_rate": 1.9503055720326573e-05, + "loss": 0.5962, + "step": 4052 + }, + { + "epoch": 0.12783472638385113, + "grad_norm": 2.508253574371338, + "learning_rate": 1.9502737641126502e-05, + "loss": 0.9361, + "step": 4053 + }, + { + "epoch": 0.12786626715029176, + "grad_norm": 5.596494674682617, + "learning_rate": 1.9502419462757986e-05, + "loss": 0.635, + "step": 4054 + }, + { + "epoch": 0.1278978079167324, + "grad_norm": 2.417268753051758, + "learning_rate": 1.9502101185224344e-05, + "loss": 0.6462, + "step": 4055 + }, + { + "epoch": 0.127929348683173, + "grad_norm": 2.547915458679199, + "learning_rate": 1.9501782808528897e-05, + "loss": 0.5945, + "step": 4056 + }, + { + "epoch": 0.12796088944961362, + "grad_norm": 2.4110283851623535, + "learning_rate": 1.950146433267497e-05, + "loss": 0.5992, + "step": 4057 + }, + { + "epoch": 0.12799243021605425, + "grad_norm": 3.227529525756836, + "learning_rate": 1.9501145757665884e-05, + "loss": 0.6087, + "step": 4058 + }, + { + "epoch": 0.12802397098249488, + "grad_norm": 3.024714469909668, + "learning_rate": 1.950082708350497e-05, + "loss": 0.5948, + "step": 4059 + }, + { + "epoch": 0.1280555117489355, + "grad_norm": 2.5803773403167725, + "learning_rate": 1.950050831019554e-05, + "loss": 0.5857, + "step": 4060 + }, + { + "epoch": 0.12808705251537614, + "grad_norm": 2.4909353256225586, + "learning_rate": 1.950018943774093e-05, + "loss": 0.5773, + "step": 4061 + }, + { + "epoch": 0.12811859328181674, + "grad_norm": 3.0730648040771484, + "learning_rate": 1.949987046614447e-05, + "loss": 0.5979, + "step": 4062 + }, + { + "epoch": 0.12815013404825737, + "grad_norm": 5.3990397453308105, + "learning_rate": 1.9499551395409482e-05, + "loss": 0.576, + "step": 4063 + }, + { + "epoch": 0.128181674814698, + "grad_norm": 8.175182342529297, + "learning_rate": 1.9499232225539302e-05, + "loss": 0.4358, + "step": 4064 + }, + { + "epoch": 0.12821321558113863, + "grad_norm": 2.7657229900360107, + "learning_rate": 1.949891295653726e-05, + "loss": 0.5458, + "step": 4065 + }, + { + "epoch": 0.12824475634757926, + "grad_norm": 3.3208296298980713, + "learning_rate": 1.949859358840668e-05, + "loss": 0.5872, + "step": 4066 + }, + { + "epoch": 0.12827629711401986, + "grad_norm": 3.28814435005188, + "learning_rate": 1.94982741211509e-05, + "loss": 0.5912, + "step": 4067 + }, + { + "epoch": 0.1283078378804605, + "grad_norm": 2.7358644008636475, + "learning_rate": 1.9497954554773258e-05, + "loss": 0.571, + "step": 4068 + }, + { + "epoch": 0.12833937864690112, + "grad_norm": 2.757894992828369, + "learning_rate": 1.9497634889277087e-05, + "loss": 0.5992, + "step": 4069 + }, + { + "epoch": 0.12837091941334175, + "grad_norm": 2.4117558002471924, + "learning_rate": 1.9497315124665718e-05, + "loss": 0.585, + "step": 4070 + }, + { + "epoch": 0.12840246017978238, + "grad_norm": 1.667640209197998, + "learning_rate": 1.9496995260942492e-05, + "loss": 0.9532, + "step": 4071 + }, + { + "epoch": 0.12843400094622298, + "grad_norm": 3.182732582092285, + "learning_rate": 1.9496675298110743e-05, + "loss": 0.5846, + "step": 4072 + }, + { + "epoch": 0.1284655417126636, + "grad_norm": 3.421252489089966, + "learning_rate": 1.9496355236173815e-05, + "loss": 0.5804, + "step": 4073 + }, + { + "epoch": 0.12849708247910424, + "grad_norm": 3.2724502086639404, + "learning_rate": 1.9496035075135048e-05, + "loss": 0.6008, + "step": 4074 + }, + { + "epoch": 0.12852862324554487, + "grad_norm": 2.725229501724243, + "learning_rate": 1.949571481499778e-05, + "loss": 0.5635, + "step": 4075 + }, + { + "epoch": 0.1285601640119855, + "grad_norm": 3.5685455799102783, + "learning_rate": 1.9495394455765356e-05, + "loss": 0.6215, + "step": 4076 + }, + { + "epoch": 0.1285917047784261, + "grad_norm": 3.314225435256958, + "learning_rate": 1.9495073997441118e-05, + "loss": 0.6007, + "step": 4077 + }, + { + "epoch": 0.12862324554486673, + "grad_norm": 2.0947823524475098, + "learning_rate": 1.9494753440028412e-05, + "loss": 0.5787, + "step": 4078 + }, + { + "epoch": 0.12865478631130736, + "grad_norm": 2.4325027465820312, + "learning_rate": 1.9494432783530577e-05, + "loss": 0.6113, + "step": 4079 + }, + { + "epoch": 0.128686327077748, + "grad_norm": 2.0295727252960205, + "learning_rate": 1.9494112027950966e-05, + "loss": 0.9829, + "step": 4080 + }, + { + "epoch": 0.12871786784418862, + "grad_norm": 3.00991153717041, + "learning_rate": 1.9493791173292924e-05, + "loss": 0.6351, + "step": 4081 + }, + { + "epoch": 0.12874940861062925, + "grad_norm": 3.068849802017212, + "learning_rate": 1.9493470219559797e-05, + "loss": 0.5729, + "step": 4082 + }, + { + "epoch": 0.12878094937706985, + "grad_norm": 3.512467622756958, + "learning_rate": 1.949314916675494e-05, + "loss": 0.5719, + "step": 4083 + }, + { + "epoch": 0.12881249014351048, + "grad_norm": 2.3008830547332764, + "learning_rate": 1.94928280148817e-05, + "loss": 0.6002, + "step": 4084 + }, + { + "epoch": 0.1288440309099511, + "grad_norm": 4.358293533325195, + "learning_rate": 1.9492506763943426e-05, + "loss": 0.5904, + "step": 4085 + }, + { + "epoch": 0.12887557167639174, + "grad_norm": 1.8463102579116821, + "learning_rate": 1.9492185413943475e-05, + "loss": 0.9573, + "step": 4086 + }, + { + "epoch": 0.12890711244283237, + "grad_norm": 2.2774007320404053, + "learning_rate": 1.94918639648852e-05, + "loss": 0.6032, + "step": 4087 + }, + { + "epoch": 0.12893865320927297, + "grad_norm": 2.7257485389709473, + "learning_rate": 1.9491542416771953e-05, + "loss": 0.5877, + "step": 4088 + }, + { + "epoch": 0.1289701939757136, + "grad_norm": 3.367952346801758, + "learning_rate": 1.949122076960709e-05, + "loss": 0.6034, + "step": 4089 + }, + { + "epoch": 0.12900173474215423, + "grad_norm": 2.6143815517425537, + "learning_rate": 1.949089902339397e-05, + "loss": 0.5852, + "step": 4090 + }, + { + "epoch": 0.12903327550859486, + "grad_norm": 2.5483591556549072, + "learning_rate": 1.9490577178135948e-05, + "loss": 0.6174, + "step": 4091 + }, + { + "epoch": 0.1290648162750355, + "grad_norm": 3.2525062561035156, + "learning_rate": 1.9490255233836383e-05, + "loss": 0.6504, + "step": 4092 + }, + { + "epoch": 0.1290963570414761, + "grad_norm": 3.156797170639038, + "learning_rate": 1.9489933190498636e-05, + "loss": 0.5817, + "step": 4093 + }, + { + "epoch": 0.12912789780791673, + "grad_norm": 2.3973677158355713, + "learning_rate": 1.9489611048126068e-05, + "loss": 0.6172, + "step": 4094 + }, + { + "epoch": 0.12915943857435735, + "grad_norm": 3.1882619857788086, + "learning_rate": 1.9489288806722043e-05, + "loss": 0.5991, + "step": 4095 + }, + { + "epoch": 0.12919097934079798, + "grad_norm": 3.559926748275757, + "learning_rate": 1.948896646628992e-05, + "loss": 0.599, + "step": 4096 + }, + { + "epoch": 0.12922252010723861, + "grad_norm": 2.275947332382202, + "learning_rate": 1.948864402683306e-05, + "loss": 0.5814, + "step": 4097 + }, + { + "epoch": 0.12925406087367924, + "grad_norm": 2.786180257797241, + "learning_rate": 1.9488321488354835e-05, + "loss": 0.5749, + "step": 4098 + }, + { + "epoch": 0.12928560164011985, + "grad_norm": 2.406839370727539, + "learning_rate": 1.948799885085861e-05, + "loss": 0.5624, + "step": 4099 + }, + { + "epoch": 0.12931714240656048, + "grad_norm": 4.569635391235352, + "learning_rate": 1.9487676114347744e-05, + "loss": 0.5719, + "step": 4100 + }, + { + "epoch": 0.1293486831730011, + "grad_norm": 3.146819591522217, + "learning_rate": 1.9487353278825615e-05, + "loss": 0.6083, + "step": 4101 + }, + { + "epoch": 0.12938022393944174, + "grad_norm": 2.555123805999756, + "learning_rate": 1.9487030344295586e-05, + "loss": 0.563, + "step": 4102 + }, + { + "epoch": 0.12941176470588237, + "grad_norm": 3.819948196411133, + "learning_rate": 1.9486707310761027e-05, + "loss": 0.5817, + "step": 4103 + }, + { + "epoch": 0.12944330547232297, + "grad_norm": 2.3608360290527344, + "learning_rate": 1.9486384178225317e-05, + "loss": 0.9105, + "step": 4104 + }, + { + "epoch": 0.1294748462387636, + "grad_norm": 3.6187853813171387, + "learning_rate": 1.948606094669182e-05, + "loss": 0.6083, + "step": 4105 + }, + { + "epoch": 0.12950638700520423, + "grad_norm": 2.886950731277466, + "learning_rate": 1.9485737616163906e-05, + "loss": 0.6075, + "step": 4106 + }, + { + "epoch": 0.12953792777164486, + "grad_norm": 2.0485875606536865, + "learning_rate": 1.9485414186644957e-05, + "loss": 0.5864, + "step": 4107 + }, + { + "epoch": 0.12956946853808549, + "grad_norm": 2.6149256229400635, + "learning_rate": 1.948509065813835e-05, + "loss": 0.6061, + "step": 4108 + }, + { + "epoch": 0.1296010093045261, + "grad_norm": 2.7819957733154297, + "learning_rate": 1.9484767030647454e-05, + "loss": 0.9439, + "step": 4109 + }, + { + "epoch": 0.12963255007096672, + "grad_norm": 2.8237907886505127, + "learning_rate": 1.9484443304175647e-05, + "loss": 0.4702, + "step": 4110 + }, + { + "epoch": 0.12966409083740735, + "grad_norm": 3.202178955078125, + "learning_rate": 1.9484119478726312e-05, + "loss": 0.5851, + "step": 4111 + }, + { + "epoch": 0.12969563160384798, + "grad_norm": 5.779841423034668, + "learning_rate": 1.9483795554302827e-05, + "loss": 0.6051, + "step": 4112 + }, + { + "epoch": 0.1297271723702886, + "grad_norm": 2.725231409072876, + "learning_rate": 1.948347153090857e-05, + "loss": 0.6057, + "step": 4113 + }, + { + "epoch": 0.1297587131367292, + "grad_norm": 2.8519210815429688, + "learning_rate": 1.9483147408546923e-05, + "loss": 0.5859, + "step": 4114 + }, + { + "epoch": 0.12979025390316984, + "grad_norm": 2.906904458999634, + "learning_rate": 1.948282318722127e-05, + "loss": 0.9482, + "step": 4115 + }, + { + "epoch": 0.12982179466961047, + "grad_norm": 3.6698055267333984, + "learning_rate": 1.9482498866935e-05, + "loss": 0.5959, + "step": 4116 + }, + { + "epoch": 0.1298533354360511, + "grad_norm": 3.512207508087158, + "learning_rate": 1.9482174447691482e-05, + "loss": 0.6002, + "step": 4117 + }, + { + "epoch": 0.12988487620249173, + "grad_norm": 2.00150465965271, + "learning_rate": 1.9481849929494113e-05, + "loss": 0.9331, + "step": 4118 + }, + { + "epoch": 0.12991641696893236, + "grad_norm": 2.94787335395813, + "learning_rate": 1.948152531234628e-05, + "loss": 0.5858, + "step": 4119 + }, + { + "epoch": 0.12994795773537296, + "grad_norm": 1.7984521389007568, + "learning_rate": 1.948120059625137e-05, + "loss": 0.4525, + "step": 4120 + }, + { + "epoch": 0.1299794985018136, + "grad_norm": 3.7671127319335938, + "learning_rate": 1.9480875781212765e-05, + "loss": 0.5971, + "step": 4121 + }, + { + "epoch": 0.13001103926825422, + "grad_norm": 2.3638041019439697, + "learning_rate": 1.948055086723386e-05, + "loss": 0.586, + "step": 4122 + }, + { + "epoch": 0.13004258003469485, + "grad_norm": 2.880347967147827, + "learning_rate": 1.948022585431805e-05, + "loss": 0.5886, + "step": 4123 + }, + { + "epoch": 0.13007412080113548, + "grad_norm": 4.968620300292969, + "learning_rate": 1.947990074246872e-05, + "loss": 0.5524, + "step": 4124 + }, + { + "epoch": 0.13010566156757608, + "grad_norm": 4.098604679107666, + "learning_rate": 1.947957553168926e-05, + "loss": 0.419, + "step": 4125 + }, + { + "epoch": 0.1301372023340167, + "grad_norm": 2.766617774963379, + "learning_rate": 1.9479250221983075e-05, + "loss": 0.6008, + "step": 4126 + }, + { + "epoch": 0.13016874310045734, + "grad_norm": 1.9805216789245605, + "learning_rate": 1.9478924813353548e-05, + "loss": 0.4325, + "step": 4127 + }, + { + "epoch": 0.13020028386689797, + "grad_norm": 2.4580600261688232, + "learning_rate": 1.9478599305804085e-05, + "loss": 0.6164, + "step": 4128 + }, + { + "epoch": 0.1302318246333386, + "grad_norm": 2.1995911598205566, + "learning_rate": 1.9478273699338074e-05, + "loss": 0.5789, + "step": 4129 + }, + { + "epoch": 0.1302633653997792, + "grad_norm": 2.2874903678894043, + "learning_rate": 1.947794799395892e-05, + "loss": 0.6126, + "step": 4130 + }, + { + "epoch": 0.13029490616621983, + "grad_norm": 3.73634672164917, + "learning_rate": 1.947762218967002e-05, + "loss": 0.5963, + "step": 4131 + }, + { + "epoch": 0.13032644693266046, + "grad_norm": 3.100569009780884, + "learning_rate": 1.9477296286474772e-05, + "loss": 0.5633, + "step": 4132 + }, + { + "epoch": 0.1303579876991011, + "grad_norm": 2.6490824222564697, + "learning_rate": 1.947697028437658e-05, + "loss": 0.5924, + "step": 4133 + }, + { + "epoch": 0.13038952846554172, + "grad_norm": 2.3496932983398438, + "learning_rate": 1.947664418337884e-05, + "loss": 0.6092, + "step": 4134 + }, + { + "epoch": 0.13042106923198235, + "grad_norm": 2.5581467151641846, + "learning_rate": 1.9476317983484966e-05, + "loss": 0.6038, + "step": 4135 + }, + { + "epoch": 0.13045260999842295, + "grad_norm": 2.7418148517608643, + "learning_rate": 1.947599168469835e-05, + "loss": 0.6138, + "step": 4136 + }, + { + "epoch": 0.13048415076486358, + "grad_norm": 2.3376100063323975, + "learning_rate": 1.9475665287022405e-05, + "loss": 0.5784, + "step": 4137 + }, + { + "epoch": 0.1305156915313042, + "grad_norm": 2.2730789184570312, + "learning_rate": 1.9475338790460535e-05, + "loss": 0.5672, + "step": 4138 + }, + { + "epoch": 0.13054723229774484, + "grad_norm": 2.5283005237579346, + "learning_rate": 1.9475012195016148e-05, + "loss": 0.6141, + "step": 4139 + }, + { + "epoch": 0.13057877306418547, + "grad_norm": 2.835526943206787, + "learning_rate": 1.947468550069265e-05, + "loss": 0.5989, + "step": 4140 + }, + { + "epoch": 0.13061031383062607, + "grad_norm": 3.7161061763763428, + "learning_rate": 1.9474358707493454e-05, + "loss": 0.5909, + "step": 4141 + }, + { + "epoch": 0.1306418545970667, + "grad_norm": 8.15815258026123, + "learning_rate": 1.9474031815421964e-05, + "loss": 0.4072, + "step": 4142 + }, + { + "epoch": 0.13067339536350733, + "grad_norm": 2.377345085144043, + "learning_rate": 1.94737048244816e-05, + "loss": 0.4426, + "step": 4143 + }, + { + "epoch": 0.13070493612994796, + "grad_norm": 9.872472763061523, + "learning_rate": 1.9473377734675768e-05, + "loss": 0.6296, + "step": 4144 + }, + { + "epoch": 0.1307364768963886, + "grad_norm": 6.056309223175049, + "learning_rate": 1.9473050546007885e-05, + "loss": 0.9405, + "step": 4145 + }, + { + "epoch": 0.1307680176628292, + "grad_norm": 2.94309139251709, + "learning_rate": 1.9472723258481364e-05, + "loss": 0.9458, + "step": 4146 + }, + { + "epoch": 0.13079955842926982, + "grad_norm": 1.9775782823562622, + "learning_rate": 1.947239587209962e-05, + "loss": 0.952, + "step": 4147 + }, + { + "epoch": 0.13083109919571045, + "grad_norm": 3.286999225616455, + "learning_rate": 1.947206838686607e-05, + "loss": 0.6322, + "step": 4148 + }, + { + "epoch": 0.13086263996215108, + "grad_norm": 2.5472631454467773, + "learning_rate": 1.947174080278413e-05, + "loss": 0.587, + "step": 4149 + }, + { + "epoch": 0.1308941807285917, + "grad_norm": 2.378054141998291, + "learning_rate": 1.9471413119857222e-05, + "loss": 0.6278, + "step": 4150 + }, + { + "epoch": 0.13092572149503232, + "grad_norm": 2.5380303859710693, + "learning_rate": 1.9471085338088762e-05, + "loss": 0.5629, + "step": 4151 + }, + { + "epoch": 0.13095726226147295, + "grad_norm": 4.281765460968018, + "learning_rate": 1.9470757457482175e-05, + "loss": 0.6025, + "step": 4152 + }, + { + "epoch": 0.13098880302791358, + "grad_norm": 1.9603960514068604, + "learning_rate": 1.9470429478040874e-05, + "loss": 0.5842, + "step": 4153 + }, + { + "epoch": 0.1310203437943542, + "grad_norm": 23.990463256835938, + "learning_rate": 1.9470101399768297e-05, + "loss": 0.5898, + "step": 4154 + }, + { + "epoch": 0.13105188456079483, + "grad_norm": 2.4214651584625244, + "learning_rate": 1.9469773222667852e-05, + "loss": 0.5469, + "step": 4155 + }, + { + "epoch": 0.13108342532723546, + "grad_norm": 2.9473681449890137, + "learning_rate": 1.946944494674297e-05, + "loss": 0.5953, + "step": 4156 + }, + { + "epoch": 0.13111496609367607, + "grad_norm": 2.005140781402588, + "learning_rate": 1.946911657199708e-05, + "loss": 0.5886, + "step": 4157 + }, + { + "epoch": 0.1311465068601167, + "grad_norm": 2.645476818084717, + "learning_rate": 1.9468788098433603e-05, + "loss": 0.5642, + "step": 4158 + }, + { + "epoch": 0.13117804762655733, + "grad_norm": 2.019465684890747, + "learning_rate": 1.9468459526055974e-05, + "loss": 0.9556, + "step": 4159 + }, + { + "epoch": 0.13120958839299796, + "grad_norm": 3.8380937576293945, + "learning_rate": 1.9468130854867615e-05, + "loss": 0.6112, + "step": 4160 + }, + { + "epoch": 0.13124112915943859, + "grad_norm": 3.1515724658966064, + "learning_rate": 1.946780208487196e-05, + "loss": 0.6036, + "step": 4161 + }, + { + "epoch": 0.1312726699258792, + "grad_norm": 2.67860746383667, + "learning_rate": 1.946747321607244e-05, + "loss": 0.5895, + "step": 4162 + }, + { + "epoch": 0.13130421069231982, + "grad_norm": 3.4760076999664307, + "learning_rate": 1.9467144248472483e-05, + "loss": 0.5967, + "step": 4163 + }, + { + "epoch": 0.13133575145876045, + "grad_norm": 3.255692958831787, + "learning_rate": 1.9466815182075525e-05, + "loss": 0.6062, + "step": 4164 + }, + { + "epoch": 0.13136729222520108, + "grad_norm": 3.091742515563965, + "learning_rate": 1.9466486016885e-05, + "loss": 0.593, + "step": 4165 + }, + { + "epoch": 0.1313988329916417, + "grad_norm": 3.2561910152435303, + "learning_rate": 1.9466156752904344e-05, + "loss": 0.6181, + "step": 4166 + }, + { + "epoch": 0.1314303737580823, + "grad_norm": 6.006330490112305, + "learning_rate": 1.9465827390136992e-05, + "loss": 0.5768, + "step": 4167 + }, + { + "epoch": 0.13146191452452294, + "grad_norm": 2.1131339073181152, + "learning_rate": 1.9465497928586378e-05, + "loss": 0.6097, + "step": 4168 + }, + { + "epoch": 0.13149345529096357, + "grad_norm": 2.2569570541381836, + "learning_rate": 1.9465168368255946e-05, + "loss": 0.6173, + "step": 4169 + }, + { + "epoch": 0.1315249960574042, + "grad_norm": 1.755247712135315, + "learning_rate": 1.9464838709149134e-05, + "loss": 0.9542, + "step": 4170 + }, + { + "epoch": 0.13155653682384483, + "grad_norm": 8.330572128295898, + "learning_rate": 1.9464508951269376e-05, + "loss": 0.5873, + "step": 4171 + }, + { + "epoch": 0.13158807759028546, + "grad_norm": 3.2801573276519775, + "learning_rate": 1.9464179094620124e-05, + "loss": 0.558, + "step": 4172 + }, + { + "epoch": 0.13161961835672606, + "grad_norm": 1.604791522026062, + "learning_rate": 1.946384913920481e-05, + "loss": 0.8927, + "step": 4173 + }, + { + "epoch": 0.1316511591231667, + "grad_norm": 2.865582227706909, + "learning_rate": 1.9463519085026885e-05, + "loss": 0.6086, + "step": 4174 + }, + { + "epoch": 0.13168269988960732, + "grad_norm": 2.8855953216552734, + "learning_rate": 1.9463188932089788e-05, + "loss": 0.6196, + "step": 4175 + }, + { + "epoch": 0.13171424065604795, + "grad_norm": 2.415933847427368, + "learning_rate": 1.9462858680396967e-05, + "loss": 0.6006, + "step": 4176 + }, + { + "epoch": 0.13174578142248858, + "grad_norm": 2.0023772716522217, + "learning_rate": 1.9462528329951868e-05, + "loss": 0.5983, + "step": 4177 + }, + { + "epoch": 0.13177732218892918, + "grad_norm": 3.9725592136383057, + "learning_rate": 1.9462197880757937e-05, + "loss": 0.6071, + "step": 4178 + }, + { + "epoch": 0.1318088629553698, + "grad_norm": 2.450451135635376, + "learning_rate": 1.9461867332818626e-05, + "loss": 0.607, + "step": 4179 + }, + { + "epoch": 0.13184040372181044, + "grad_norm": 2.2817044258117676, + "learning_rate": 1.946153668613738e-05, + "loss": 0.5763, + "step": 4180 + }, + { + "epoch": 0.13187194448825107, + "grad_norm": 2.088613271713257, + "learning_rate": 1.9461205940717654e-05, + "loss": 0.5952, + "step": 4181 + }, + { + "epoch": 0.1319034852546917, + "grad_norm": 2.74070143699646, + "learning_rate": 1.9460875096562898e-05, + "loss": 0.6117, + "step": 4182 + }, + { + "epoch": 0.1319350260211323, + "grad_norm": 3.4004859924316406, + "learning_rate": 1.9460544153676562e-05, + "loss": 0.5717, + "step": 4183 + }, + { + "epoch": 0.13196656678757293, + "grad_norm": 2.6179263591766357, + "learning_rate": 1.9460213112062103e-05, + "loss": 0.6086, + "step": 4184 + }, + { + "epoch": 0.13199810755401356, + "grad_norm": 2.9501287937164307, + "learning_rate": 1.9459881971722975e-05, + "loss": 0.6312, + "step": 4185 + }, + { + "epoch": 0.1320296483204542, + "grad_norm": 3.102257490158081, + "learning_rate": 1.9459550732662634e-05, + "loss": 0.6195, + "step": 4186 + }, + { + "epoch": 0.13206118908689482, + "grad_norm": 7.108062267303467, + "learning_rate": 1.9459219394884538e-05, + "loss": 0.6206, + "step": 4187 + }, + { + "epoch": 0.13209272985333545, + "grad_norm": 2.0504133701324463, + "learning_rate": 1.9458887958392137e-05, + "loss": 0.6065, + "step": 4188 + }, + { + "epoch": 0.13212427061977605, + "grad_norm": 2.8085036277770996, + "learning_rate": 1.9458556423188895e-05, + "loss": 0.5713, + "step": 4189 + }, + { + "epoch": 0.13215581138621668, + "grad_norm": 6.900393486022949, + "learning_rate": 1.945822478927828e-05, + "loss": 0.5994, + "step": 4190 + }, + { + "epoch": 0.1321873521526573, + "grad_norm": 2.051724433898926, + "learning_rate": 1.945789305666374e-05, + "loss": 0.5911, + "step": 4191 + }, + { + "epoch": 0.13221889291909794, + "grad_norm": 2.745999336242676, + "learning_rate": 1.9457561225348743e-05, + "loss": 0.5833, + "step": 4192 + }, + { + "epoch": 0.13225043368553857, + "grad_norm": 2.228388547897339, + "learning_rate": 1.945722929533675e-05, + "loss": 0.5321, + "step": 4193 + }, + { + "epoch": 0.13228197445197917, + "grad_norm": 2.5240566730499268, + "learning_rate": 1.9456897266631225e-05, + "loss": 0.5822, + "step": 4194 + }, + { + "epoch": 0.1323135152184198, + "grad_norm": 2.4216525554656982, + "learning_rate": 1.9456565139235635e-05, + "loss": 0.5856, + "step": 4195 + }, + { + "epoch": 0.13234505598486043, + "grad_norm": 2.7949275970458984, + "learning_rate": 1.9456232913153446e-05, + "loss": 0.6218, + "step": 4196 + }, + { + "epoch": 0.13237659675130106, + "grad_norm": 2.930541753768921, + "learning_rate": 1.9455900588388124e-05, + "loss": 0.601, + "step": 4197 + }, + { + "epoch": 0.1324081375177417, + "grad_norm": 2.6412343978881836, + "learning_rate": 1.945556816494314e-05, + "loss": 0.5754, + "step": 4198 + }, + { + "epoch": 0.1324396782841823, + "grad_norm": 2.3851125240325928, + "learning_rate": 1.9455235642821956e-05, + "loss": 0.61, + "step": 4199 + }, + { + "epoch": 0.13247121905062292, + "grad_norm": 3.2933576107025146, + "learning_rate": 1.9454903022028048e-05, + "loss": 0.578, + "step": 4200 + }, + { + "epoch": 0.13250275981706355, + "grad_norm": 2.050661563873291, + "learning_rate": 1.9454570302564885e-05, + "loss": 0.6127, + "step": 4201 + }, + { + "epoch": 0.13253430058350418, + "grad_norm": 2.5608177185058594, + "learning_rate": 1.9454237484435936e-05, + "loss": 0.5824, + "step": 4202 + }, + { + "epoch": 0.1325658413499448, + "grad_norm": 3.2630045413970947, + "learning_rate": 1.945390456764468e-05, + "loss": 0.629, + "step": 4203 + }, + { + "epoch": 0.13259738211638541, + "grad_norm": 3.2858874797821045, + "learning_rate": 1.9453571552194592e-05, + "loss": 0.5989, + "step": 4204 + }, + { + "epoch": 0.13262892288282604, + "grad_norm": 3.8173649311065674, + "learning_rate": 1.9453238438089143e-05, + "loss": 0.5925, + "step": 4205 + }, + { + "epoch": 0.13266046364926667, + "grad_norm": 2.1875483989715576, + "learning_rate": 1.945290522533181e-05, + "loss": 0.9151, + "step": 4206 + }, + { + "epoch": 0.1326920044157073, + "grad_norm": 2.108792304992676, + "learning_rate": 1.945257191392607e-05, + "loss": 0.5811, + "step": 4207 + }, + { + "epoch": 0.13272354518214793, + "grad_norm": 3.025662422180176, + "learning_rate": 1.94522385038754e-05, + "loss": 0.5898, + "step": 4208 + }, + { + "epoch": 0.13275508594858856, + "grad_norm": 3.5279955863952637, + "learning_rate": 1.945190499518329e-05, + "loss": 0.6137, + "step": 4209 + }, + { + "epoch": 0.13278662671502917, + "grad_norm": 2.8292300701141357, + "learning_rate": 1.9451571387853204e-05, + "loss": 0.5624, + "step": 4210 + }, + { + "epoch": 0.1328181674814698, + "grad_norm": 3.004573345184326, + "learning_rate": 1.9451237681888637e-05, + "loss": 0.5951, + "step": 4211 + }, + { + "epoch": 0.13284970824791043, + "grad_norm": 2.2864012718200684, + "learning_rate": 1.9450903877293062e-05, + "loss": 0.5968, + "step": 4212 + }, + { + "epoch": 0.13288124901435105, + "grad_norm": 7.402925491333008, + "learning_rate": 1.9450569974069968e-05, + "loss": 0.6109, + "step": 4213 + }, + { + "epoch": 0.13291278978079168, + "grad_norm": 2.558150053024292, + "learning_rate": 1.945023597222284e-05, + "loss": 0.5788, + "step": 4214 + }, + { + "epoch": 0.1329443305472323, + "grad_norm": 2.6321768760681152, + "learning_rate": 1.944990187175516e-05, + "loss": 0.4552, + "step": 4215 + }, + { + "epoch": 0.13297587131367292, + "grad_norm": 2.353091239929199, + "learning_rate": 1.9449567672670416e-05, + "loss": 0.6005, + "step": 4216 + }, + { + "epoch": 0.13300741208011355, + "grad_norm": 2.5014853477478027, + "learning_rate": 1.9449233374972098e-05, + "loss": 0.5748, + "step": 4217 + }, + { + "epoch": 0.13303895284655418, + "grad_norm": 3.153031826019287, + "learning_rate": 1.9448898978663687e-05, + "loss": 0.587, + "step": 4218 + }, + { + "epoch": 0.1330704936129948, + "grad_norm": 2.463059186935425, + "learning_rate": 1.9448564483748685e-05, + "loss": 0.5823, + "step": 4219 + }, + { + "epoch": 0.1331020343794354, + "grad_norm": 1.9607787132263184, + "learning_rate": 1.9448229890230574e-05, + "loss": 0.9469, + "step": 4220 + }, + { + "epoch": 0.13313357514587604, + "grad_norm": 2.2965149879455566, + "learning_rate": 1.9447895198112848e-05, + "loss": 0.573, + "step": 4221 + }, + { + "epoch": 0.13316511591231667, + "grad_norm": 2.648927927017212, + "learning_rate": 1.9447560407399e-05, + "loss": 0.5797, + "step": 4222 + }, + { + "epoch": 0.1331966566787573, + "grad_norm": 3.494765520095825, + "learning_rate": 1.9447225518092518e-05, + "loss": 0.5613, + "step": 4223 + }, + { + "epoch": 0.13322819744519793, + "grad_norm": 2.023784637451172, + "learning_rate": 1.944689053019691e-05, + "loss": 0.9216, + "step": 4224 + }, + { + "epoch": 0.13325973821163856, + "grad_norm": 2.852809429168701, + "learning_rate": 1.944655544371566e-05, + "loss": 0.6162, + "step": 4225 + }, + { + "epoch": 0.13329127897807916, + "grad_norm": 2.682377338409424, + "learning_rate": 1.944622025865227e-05, + "loss": 0.5632, + "step": 4226 + }, + { + "epoch": 0.1333228197445198, + "grad_norm": 2.1771743297576904, + "learning_rate": 1.9445884975010233e-05, + "loss": 0.566, + "step": 4227 + }, + { + "epoch": 0.13335436051096042, + "grad_norm": 2.2826266288757324, + "learning_rate": 1.9445549592793057e-05, + "loss": 0.5529, + "step": 4228 + }, + { + "epoch": 0.13338590127740105, + "grad_norm": 2.7922685146331787, + "learning_rate": 1.944521411200423e-05, + "loss": 0.589, + "step": 4229 + }, + { + "epoch": 0.13341744204384168, + "grad_norm": 2.1719110012054443, + "learning_rate": 1.9444878532647267e-05, + "loss": 0.5442, + "step": 4230 + }, + { + "epoch": 0.13344898281028228, + "grad_norm": 2.1752851009368896, + "learning_rate": 1.9444542854725658e-05, + "loss": 0.9516, + "step": 4231 + }, + { + "epoch": 0.1334805235767229, + "grad_norm": 2.645653247833252, + "learning_rate": 1.9444207078242915e-05, + "loss": 0.483, + "step": 4232 + }, + { + "epoch": 0.13351206434316354, + "grad_norm": 2.7710514068603516, + "learning_rate": 1.9443871203202532e-05, + "loss": 0.5887, + "step": 4233 + }, + { + "epoch": 0.13354360510960417, + "grad_norm": 2.437399387359619, + "learning_rate": 1.9443535229608023e-05, + "loss": 0.6382, + "step": 4234 + }, + { + "epoch": 0.1335751458760448, + "grad_norm": 3.1304726600646973, + "learning_rate": 1.9443199157462894e-05, + "loss": 0.6027, + "step": 4235 + }, + { + "epoch": 0.1336066866424854, + "grad_norm": 3.3847389221191406, + "learning_rate": 1.9442862986770645e-05, + "loss": 0.588, + "step": 4236 + }, + { + "epoch": 0.13363822740892603, + "grad_norm": 2.958951234817505, + "learning_rate": 1.9442526717534793e-05, + "loss": 0.5794, + "step": 4237 + }, + { + "epoch": 0.13366976817536666, + "grad_norm": 3.0688793659210205, + "learning_rate": 1.944219034975884e-05, + "loss": 0.6411, + "step": 4238 + }, + { + "epoch": 0.1337013089418073, + "grad_norm": 2.052825927734375, + "learning_rate": 1.94418538834463e-05, + "loss": 0.9664, + "step": 4239 + }, + { + "epoch": 0.13373284970824792, + "grad_norm": 5.236640453338623, + "learning_rate": 1.944151731860068e-05, + "loss": 0.5813, + "step": 4240 + }, + { + "epoch": 0.13376439047468852, + "grad_norm": 2.311568260192871, + "learning_rate": 1.94411806552255e-05, + "loss": 0.572, + "step": 4241 + }, + { + "epoch": 0.13379593124112915, + "grad_norm": 2.986274480819702, + "learning_rate": 1.9440843893324266e-05, + "loss": 0.5873, + "step": 4242 + }, + { + "epoch": 0.13382747200756978, + "grad_norm": 2.4648914337158203, + "learning_rate": 1.9440507032900498e-05, + "loss": 0.589, + "step": 4243 + }, + { + "epoch": 0.1338590127740104, + "grad_norm": 3.1214098930358887, + "learning_rate": 1.9440170073957708e-05, + "loss": 0.6112, + "step": 4244 + }, + { + "epoch": 0.13389055354045104, + "grad_norm": 3.008241891860962, + "learning_rate": 1.943983301649941e-05, + "loss": 0.5951, + "step": 4245 + }, + { + "epoch": 0.13392209430689167, + "grad_norm": 5.275271892547607, + "learning_rate": 1.943949586052913e-05, + "loss": 0.6028, + "step": 4246 + }, + { + "epoch": 0.13395363507333227, + "grad_norm": 2.29318904876709, + "learning_rate": 1.9439158606050377e-05, + "loss": 0.5752, + "step": 4247 + }, + { + "epoch": 0.1339851758397729, + "grad_norm": 4.139581680297852, + "learning_rate": 1.9438821253066677e-05, + "loss": 0.6073, + "step": 4248 + }, + { + "epoch": 0.13401671660621353, + "grad_norm": 3.162275791168213, + "learning_rate": 1.9438483801581546e-05, + "loss": 0.5952, + "step": 4249 + }, + { + "epoch": 0.13404825737265416, + "grad_norm": 2.46108341217041, + "learning_rate": 1.9438146251598508e-05, + "loss": 0.5824, + "step": 4250 + }, + { + "epoch": 0.1340797981390948, + "grad_norm": 2.6085305213928223, + "learning_rate": 1.9437808603121086e-05, + "loss": 0.5828, + "step": 4251 + }, + { + "epoch": 0.1341113389055354, + "grad_norm": 2.4328315258026123, + "learning_rate": 1.9437470856152805e-05, + "loss": 0.5984, + "step": 4252 + }, + { + "epoch": 0.13414287967197602, + "grad_norm": 1.8795504570007324, + "learning_rate": 1.9437133010697183e-05, + "loss": 0.9659, + "step": 4253 + }, + { + "epoch": 0.13417442043841665, + "grad_norm": 3.418816328048706, + "learning_rate": 1.9436795066757753e-05, + "loss": 0.59, + "step": 4254 + }, + { + "epoch": 0.13420596120485728, + "grad_norm": 3.3314857482910156, + "learning_rate": 1.943645702433804e-05, + "loss": 0.5894, + "step": 4255 + }, + { + "epoch": 0.1342375019712979, + "grad_norm": 4.20680046081543, + "learning_rate": 1.9436118883441572e-05, + "loss": 0.6002, + "step": 4256 + }, + { + "epoch": 0.13426904273773851, + "grad_norm": 2.119678258895874, + "learning_rate": 1.9435780644071872e-05, + "loss": 0.5332, + "step": 4257 + }, + { + "epoch": 0.13430058350417914, + "grad_norm": 2.5917561054229736, + "learning_rate": 1.943544230623248e-05, + "loss": 0.5815, + "step": 4258 + }, + { + "epoch": 0.13433212427061977, + "grad_norm": 56.07221603393555, + "learning_rate": 1.9435103869926914e-05, + "loss": 0.5704, + "step": 4259 + }, + { + "epoch": 0.1343636650370604, + "grad_norm": 2.64570689201355, + "learning_rate": 1.9434765335158716e-05, + "loss": 0.5858, + "step": 4260 + }, + { + "epoch": 0.13439520580350103, + "grad_norm": 7.001703262329102, + "learning_rate": 1.943442670193142e-05, + "loss": 0.5899, + "step": 4261 + }, + { + "epoch": 0.13442674656994166, + "grad_norm": 28.413745880126953, + "learning_rate": 1.943408797024855e-05, + "loss": 0.5914, + "step": 4262 + }, + { + "epoch": 0.13445828733638227, + "grad_norm": 3.336930751800537, + "learning_rate": 1.943374914011365e-05, + "loss": 0.6189, + "step": 4263 + }, + { + "epoch": 0.1344898281028229, + "grad_norm": 2.816824436187744, + "learning_rate": 1.943341021153025e-05, + "loss": 0.6048, + "step": 4264 + }, + { + "epoch": 0.13452136886926352, + "grad_norm": 4.222872734069824, + "learning_rate": 1.943307118450189e-05, + "loss": 0.5591, + "step": 4265 + }, + { + "epoch": 0.13455290963570415, + "grad_norm": 2.355396270751953, + "learning_rate": 1.943273205903211e-05, + "loss": 0.6087, + "step": 4266 + }, + { + "epoch": 0.13458445040214478, + "grad_norm": 3.511584520339966, + "learning_rate": 1.9432392835124445e-05, + "loss": 0.5836, + "step": 4267 + }, + { + "epoch": 0.13461599116858539, + "grad_norm": 2.5927226543426514, + "learning_rate": 1.9432053512782435e-05, + "loss": 0.6241, + "step": 4268 + }, + { + "epoch": 0.13464753193502602, + "grad_norm": 4.315495491027832, + "learning_rate": 1.9431714092009625e-05, + "loss": 0.6068, + "step": 4269 + }, + { + "epoch": 0.13467907270146665, + "grad_norm": 2.4919402599334717, + "learning_rate": 1.9431374572809555e-05, + "loss": 0.6382, + "step": 4270 + }, + { + "epoch": 0.13471061346790728, + "grad_norm": 2.374497413635254, + "learning_rate": 1.9431034955185767e-05, + "loss": 0.5751, + "step": 4271 + }, + { + "epoch": 0.1347421542343479, + "grad_norm": 3.54911732673645, + "learning_rate": 1.9430695239141806e-05, + "loss": 0.613, + "step": 4272 + }, + { + "epoch": 0.1347736950007885, + "grad_norm": 2.9013748168945312, + "learning_rate": 1.9430355424681218e-05, + "loss": 0.6044, + "step": 4273 + }, + { + "epoch": 0.13480523576722914, + "grad_norm": 2.561495304107666, + "learning_rate": 1.9430015511807552e-05, + "loss": 0.595, + "step": 4274 + }, + { + "epoch": 0.13483677653366977, + "grad_norm": 2.5762410163879395, + "learning_rate": 1.9429675500524348e-05, + "loss": 0.6008, + "step": 4275 + }, + { + "epoch": 0.1348683173001104, + "grad_norm": 2.243745803833008, + "learning_rate": 1.9429335390835157e-05, + "loss": 0.6356, + "step": 4276 + }, + { + "epoch": 0.13489985806655103, + "grad_norm": 2.19427752494812, + "learning_rate": 1.9428995182743533e-05, + "loss": 0.6403, + "step": 4277 + }, + { + "epoch": 0.13493139883299163, + "grad_norm": 2.1190717220306396, + "learning_rate": 1.9428654876253023e-05, + "loss": 0.5827, + "step": 4278 + }, + { + "epoch": 0.13496293959943226, + "grad_norm": 3.4292190074920654, + "learning_rate": 1.9428314471367174e-05, + "loss": 0.5907, + "step": 4279 + }, + { + "epoch": 0.1349944803658729, + "grad_norm": 3.546494960784912, + "learning_rate": 1.9427973968089545e-05, + "loss": 0.6172, + "step": 4280 + }, + { + "epoch": 0.13502602113231352, + "grad_norm": 2.083320140838623, + "learning_rate": 1.942763336642369e-05, + "loss": 0.9433, + "step": 4281 + }, + { + "epoch": 0.13505756189875415, + "grad_norm": 2.5349552631378174, + "learning_rate": 1.9427292666373157e-05, + "loss": 0.5952, + "step": 4282 + }, + { + "epoch": 0.13508910266519478, + "grad_norm": 2.8902931213378906, + "learning_rate": 1.9426951867941506e-05, + "loss": 0.5972, + "step": 4283 + }, + { + "epoch": 0.13512064343163538, + "grad_norm": 3.855405569076538, + "learning_rate": 1.9426610971132292e-05, + "loss": 0.4633, + "step": 4284 + }, + { + "epoch": 0.135152184198076, + "grad_norm": 2.742950677871704, + "learning_rate": 1.9426269975949074e-05, + "loss": 0.6178, + "step": 4285 + }, + { + "epoch": 0.13518372496451664, + "grad_norm": 1.9593955278396606, + "learning_rate": 1.9425928882395407e-05, + "loss": 0.9332, + "step": 4286 + }, + { + "epoch": 0.13521526573095727, + "grad_norm": 5.992365837097168, + "learning_rate": 1.9425587690474855e-05, + "loss": 0.4383, + "step": 4287 + }, + { + "epoch": 0.1352468064973979, + "grad_norm": 4.169990539550781, + "learning_rate": 1.9425246400190976e-05, + "loss": 0.6238, + "step": 4288 + }, + { + "epoch": 0.1352783472638385, + "grad_norm": 2.1000075340270996, + "learning_rate": 1.9424905011547334e-05, + "loss": 0.9105, + "step": 4289 + }, + { + "epoch": 0.13530988803027913, + "grad_norm": 3.2399704456329346, + "learning_rate": 1.9424563524547486e-05, + "loss": 0.6015, + "step": 4290 + }, + { + "epoch": 0.13534142879671976, + "grad_norm": 3.1494619846343994, + "learning_rate": 1.9424221939195004e-05, + "loss": 0.5985, + "step": 4291 + }, + { + "epoch": 0.1353729695631604, + "grad_norm": 2.585618257522583, + "learning_rate": 1.9423880255493442e-05, + "loss": 0.598, + "step": 4292 + }, + { + "epoch": 0.13540451032960102, + "grad_norm": 3.2006490230560303, + "learning_rate": 1.942353847344638e-05, + "loss": 0.6105, + "step": 4293 + }, + { + "epoch": 0.13543605109604162, + "grad_norm": 4.199006080627441, + "learning_rate": 1.9423196593057372e-05, + "loss": 0.5845, + "step": 4294 + }, + { + "epoch": 0.13546759186248225, + "grad_norm": 3.0284957885742188, + "learning_rate": 1.9422854614329992e-05, + "loss": 0.5799, + "step": 4295 + }, + { + "epoch": 0.13549913262892288, + "grad_norm": 3.1515073776245117, + "learning_rate": 1.9422512537267806e-05, + "loss": 0.6315, + "step": 4296 + }, + { + "epoch": 0.1355306733953635, + "grad_norm": 3.1354737281799316, + "learning_rate": 1.942217036187439e-05, + "loss": 0.6063, + "step": 4297 + }, + { + "epoch": 0.13556221416180414, + "grad_norm": 2.0584287643432617, + "learning_rate": 1.9421828088153305e-05, + "loss": 0.9477, + "step": 4298 + }, + { + "epoch": 0.13559375492824477, + "grad_norm": 2.531747579574585, + "learning_rate": 1.942148571610813e-05, + "loss": 0.4492, + "step": 4299 + }, + { + "epoch": 0.13562529569468537, + "grad_norm": 2.5971169471740723, + "learning_rate": 1.942114324574244e-05, + "loss": 0.5976, + "step": 4300 + }, + { + "epoch": 0.135656836461126, + "grad_norm": 3.491889715194702, + "learning_rate": 1.9420800677059797e-05, + "loss": 0.593, + "step": 4301 + }, + { + "epoch": 0.13568837722756663, + "grad_norm": 3.2912776470184326, + "learning_rate": 1.9420458010063788e-05, + "loss": 0.6042, + "step": 4302 + }, + { + "epoch": 0.13571991799400726, + "grad_norm": 2.3964157104492188, + "learning_rate": 1.9420115244757985e-05, + "loss": 0.5727, + "step": 4303 + }, + { + "epoch": 0.1357514587604479, + "grad_norm": 6.548861026763916, + "learning_rate": 1.941977238114596e-05, + "loss": 0.5634, + "step": 4304 + }, + { + "epoch": 0.1357829995268885, + "grad_norm": 2.14249587059021, + "learning_rate": 1.9419429419231305e-05, + "loss": 0.9501, + "step": 4305 + }, + { + "epoch": 0.13581454029332912, + "grad_norm": 4.543886661529541, + "learning_rate": 1.941908635901758e-05, + "loss": 0.5961, + "step": 4306 + }, + { + "epoch": 0.13584608105976975, + "grad_norm": 3.6497700214385986, + "learning_rate": 1.9418743200508384e-05, + "loss": 0.5732, + "step": 4307 + }, + { + "epoch": 0.13587762182621038, + "grad_norm": 21.345502853393555, + "learning_rate": 1.9418399943707285e-05, + "loss": 0.5876, + "step": 4308 + }, + { + "epoch": 0.135909162592651, + "grad_norm": 2.9400410652160645, + "learning_rate": 1.9418056588617872e-05, + "loss": 0.5851, + "step": 4309 + }, + { + "epoch": 0.1359407033590916, + "grad_norm": 4.054344654083252, + "learning_rate": 1.9417713135243722e-05, + "loss": 0.6065, + "step": 4310 + }, + { + "epoch": 0.13597224412553224, + "grad_norm": 3.5064806938171387, + "learning_rate": 1.9417369583588425e-05, + "loss": 0.5666, + "step": 4311 + }, + { + "epoch": 0.13600378489197287, + "grad_norm": 3.063236951828003, + "learning_rate": 1.9417025933655563e-05, + "loss": 0.6362, + "step": 4312 + }, + { + "epoch": 0.1360353256584135, + "grad_norm": 3.629277229309082, + "learning_rate": 1.9416682185448727e-05, + "loss": 0.5915, + "step": 4313 + }, + { + "epoch": 0.13606686642485413, + "grad_norm": 3.0159544944763184, + "learning_rate": 1.9416338338971496e-05, + "loss": 0.6002, + "step": 4314 + }, + { + "epoch": 0.13609840719129473, + "grad_norm": 2.3840553760528564, + "learning_rate": 1.9415994394227462e-05, + "loss": 0.6173, + "step": 4315 + }, + { + "epoch": 0.13612994795773536, + "grad_norm": 2.114293098449707, + "learning_rate": 1.9415650351220217e-05, + "loss": 0.5937, + "step": 4316 + }, + { + "epoch": 0.136161488724176, + "grad_norm": 3.6860830783843994, + "learning_rate": 1.9415306209953352e-05, + "loss": 0.5947, + "step": 4317 + }, + { + "epoch": 0.13619302949061662, + "grad_norm": 2.6307990550994873, + "learning_rate": 1.941496197043045e-05, + "loss": 0.6071, + "step": 4318 + }, + { + "epoch": 0.13622457025705725, + "grad_norm": 2.5476436614990234, + "learning_rate": 1.9414617632655114e-05, + "loss": 0.5691, + "step": 4319 + }, + { + "epoch": 0.13625611102349788, + "grad_norm": 2.842930793762207, + "learning_rate": 1.9414273196630934e-05, + "loss": 0.5904, + "step": 4320 + }, + { + "epoch": 0.13628765178993849, + "grad_norm": 2.4976534843444824, + "learning_rate": 1.94139286623615e-05, + "loss": 0.5954, + "step": 4321 + }, + { + "epoch": 0.13631919255637912, + "grad_norm": 2.4254860877990723, + "learning_rate": 1.9413584029850413e-05, + "loss": 0.569, + "step": 4322 + }, + { + "epoch": 0.13635073332281974, + "grad_norm": 4.6013689041137695, + "learning_rate": 1.9413239299101265e-05, + "loss": 0.5679, + "step": 4323 + }, + { + "epoch": 0.13638227408926037, + "grad_norm": 3.468674421310425, + "learning_rate": 1.9412894470117653e-05, + "loss": 0.5555, + "step": 4324 + }, + { + "epoch": 0.136413814855701, + "grad_norm": 2.712679624557495, + "learning_rate": 1.9412549542903184e-05, + "loss": 0.604, + "step": 4325 + }, + { + "epoch": 0.1364453556221416, + "grad_norm": 3.015449047088623, + "learning_rate": 1.941220451746145e-05, + "loss": 0.5709, + "step": 4326 + }, + { + "epoch": 0.13647689638858224, + "grad_norm": 3.011504888534546, + "learning_rate": 1.9411859393796053e-05, + "loss": 0.9127, + "step": 4327 + }, + { + "epoch": 0.13650843715502287, + "grad_norm": 3.631298780441284, + "learning_rate": 1.9411514171910595e-05, + "loss": 0.61, + "step": 4328 + }, + { + "epoch": 0.1365399779214635, + "grad_norm": 4.600008487701416, + "learning_rate": 1.941116885180868e-05, + "loss": 0.6002, + "step": 4329 + }, + { + "epoch": 0.13657151868790413, + "grad_norm": 3.200148582458496, + "learning_rate": 1.941082343349391e-05, + "loss": 0.5681, + "step": 4330 + }, + { + "epoch": 0.13660305945434473, + "grad_norm": 3.026008129119873, + "learning_rate": 1.9410477916969886e-05, + "loss": 0.5771, + "step": 4331 + }, + { + "epoch": 0.13663460022078536, + "grad_norm": 4.558535099029541, + "learning_rate": 1.9410132302240224e-05, + "loss": 0.5775, + "step": 4332 + }, + { + "epoch": 0.136666140987226, + "grad_norm": 2.81691575050354, + "learning_rate": 1.9409786589308523e-05, + "loss": 0.4037, + "step": 4333 + }, + { + "epoch": 0.13669768175366662, + "grad_norm": 1.8181794881820679, + "learning_rate": 1.940944077817839e-05, + "loss": 0.9566, + "step": 4334 + }, + { + "epoch": 0.13672922252010725, + "grad_norm": 2.989142894744873, + "learning_rate": 1.9409094868853438e-05, + "loss": 0.588, + "step": 4335 + }, + { + "epoch": 0.13676076328654788, + "grad_norm": 3.3320515155792236, + "learning_rate": 1.9408748861337274e-05, + "loss": 0.5975, + "step": 4336 + }, + { + "epoch": 0.13679230405298848, + "grad_norm": 2.6021978855133057, + "learning_rate": 1.9408402755633512e-05, + "loss": 0.587, + "step": 4337 + }, + { + "epoch": 0.1368238448194291, + "grad_norm": 4.637216567993164, + "learning_rate": 1.940805655174576e-05, + "loss": 0.6191, + "step": 4338 + }, + { + "epoch": 0.13685538558586974, + "grad_norm": 2.5581483840942383, + "learning_rate": 1.9407710249677633e-05, + "loss": 0.6142, + "step": 4339 + }, + { + "epoch": 0.13688692635231037, + "grad_norm": 3.489470958709717, + "learning_rate": 1.9407363849432746e-05, + "loss": 0.608, + "step": 4340 + }, + { + "epoch": 0.136918467118751, + "grad_norm": 3.3933136463165283, + "learning_rate": 1.940701735101471e-05, + "loss": 0.6209, + "step": 4341 + }, + { + "epoch": 0.1369500078851916, + "grad_norm": 2.2824299335479736, + "learning_rate": 1.9406670754427143e-05, + "loss": 0.5563, + "step": 4342 + }, + { + "epoch": 0.13698154865163223, + "grad_norm": 3.3064980506896973, + "learning_rate": 1.9406324059673667e-05, + "loss": 0.6008, + "step": 4343 + }, + { + "epoch": 0.13701308941807286, + "grad_norm": 9.108086585998535, + "learning_rate": 1.9405977266757893e-05, + "loss": 0.6119, + "step": 4344 + }, + { + "epoch": 0.1370446301845135, + "grad_norm": 3.3063061237335205, + "learning_rate": 1.940563037568344e-05, + "loss": 0.5607, + "step": 4345 + }, + { + "epoch": 0.13707617095095412, + "grad_norm": 8.520140647888184, + "learning_rate": 1.9405283386453936e-05, + "loss": 0.4978, + "step": 4346 + }, + { + "epoch": 0.13710771171739472, + "grad_norm": 3.5687103271484375, + "learning_rate": 1.9404936299072995e-05, + "loss": 0.6124, + "step": 4347 + }, + { + "epoch": 0.13713925248383535, + "grad_norm": 3.0760691165924072, + "learning_rate": 1.9404589113544242e-05, + "loss": 0.5645, + "step": 4348 + }, + { + "epoch": 0.13717079325027598, + "grad_norm": 2.3285229206085205, + "learning_rate": 1.94042418298713e-05, + "loss": 0.9334, + "step": 4349 + }, + { + "epoch": 0.1372023340167166, + "grad_norm": 1.8361515998840332, + "learning_rate": 1.940389444805779e-05, + "loss": 0.9852, + "step": 4350 + }, + { + "epoch": 0.13723387478315724, + "grad_norm": 2.4053711891174316, + "learning_rate": 1.9403546968107336e-05, + "loss": 0.5877, + "step": 4351 + }, + { + "epoch": 0.13726541554959787, + "grad_norm": 2.4048404693603516, + "learning_rate": 1.940319939002357e-05, + "loss": 0.5681, + "step": 4352 + }, + { + "epoch": 0.13729695631603847, + "grad_norm": 2.927034616470337, + "learning_rate": 1.940285171381012e-05, + "loss": 0.6096, + "step": 4353 + }, + { + "epoch": 0.1373284970824791, + "grad_norm": 7.671116352081299, + "learning_rate": 1.9402503939470606e-05, + "loss": 0.4864, + "step": 4354 + }, + { + "epoch": 0.13736003784891973, + "grad_norm": 2.90744948387146, + "learning_rate": 1.940215606700867e-05, + "loss": 0.6076, + "step": 4355 + }, + { + "epoch": 0.13739157861536036, + "grad_norm": 2.521393060684204, + "learning_rate": 1.940180809642793e-05, + "loss": 0.6175, + "step": 4356 + }, + { + "epoch": 0.137423119381801, + "grad_norm": 2.8042280673980713, + "learning_rate": 1.9401460027732022e-05, + "loss": 0.5991, + "step": 4357 + }, + { + "epoch": 0.1374546601482416, + "grad_norm": 3.615346670150757, + "learning_rate": 1.940111186092458e-05, + "loss": 0.6344, + "step": 4358 + }, + { + "epoch": 0.13748620091468222, + "grad_norm": 2.433274507522583, + "learning_rate": 1.9400763596009236e-05, + "loss": 0.623, + "step": 4359 + }, + { + "epoch": 0.13751774168112285, + "grad_norm": 2.1560733318328857, + "learning_rate": 1.9400415232989625e-05, + "loss": 0.5671, + "step": 4360 + }, + { + "epoch": 0.13754928244756348, + "grad_norm": 3.037848711013794, + "learning_rate": 1.940006677186938e-05, + "loss": 0.5909, + "step": 4361 + }, + { + "epoch": 0.1375808232140041, + "grad_norm": 2.836215019226074, + "learning_rate": 1.939971821265214e-05, + "loss": 0.5962, + "step": 4362 + }, + { + "epoch": 0.1376123639804447, + "grad_norm": 2.7751781940460205, + "learning_rate": 1.939936955534154e-05, + "loss": 0.5784, + "step": 4363 + }, + { + "epoch": 0.13764390474688534, + "grad_norm": 2.3450160026550293, + "learning_rate": 1.9399020799941222e-05, + "loss": 0.5832, + "step": 4364 + }, + { + "epoch": 0.13767544551332597, + "grad_norm": 2.9513068199157715, + "learning_rate": 1.9398671946454827e-05, + "loss": 0.6069, + "step": 4365 + }, + { + "epoch": 0.1377069862797666, + "grad_norm": 2.14342999458313, + "learning_rate": 1.9398322994885984e-05, + "loss": 0.5896, + "step": 4366 + }, + { + "epoch": 0.13773852704620723, + "grad_norm": 2.588669538497925, + "learning_rate": 1.939797394523835e-05, + "loss": 0.617, + "step": 4367 + }, + { + "epoch": 0.13777006781264783, + "grad_norm": 3.3775429725646973, + "learning_rate": 1.9397624797515558e-05, + "loss": 0.5839, + "step": 4368 + }, + { + "epoch": 0.13780160857908846, + "grad_norm": 3.354935646057129, + "learning_rate": 1.9397275551721256e-05, + "loss": 0.5882, + "step": 4369 + }, + { + "epoch": 0.1378331493455291, + "grad_norm": 2.9108359813690186, + "learning_rate": 1.9396926207859085e-05, + "loss": 0.618, + "step": 4370 + }, + { + "epoch": 0.13786469011196972, + "grad_norm": 3.6389551162719727, + "learning_rate": 1.9396576765932694e-05, + "loss": 0.6074, + "step": 4371 + }, + { + "epoch": 0.13789623087841035, + "grad_norm": 2.830496072769165, + "learning_rate": 1.9396227225945727e-05, + "loss": 0.5817, + "step": 4372 + }, + { + "epoch": 0.13792777164485098, + "grad_norm": 2.7275240421295166, + "learning_rate": 1.9395877587901834e-05, + "loss": 0.5797, + "step": 4373 + }, + { + "epoch": 0.13795931241129158, + "grad_norm": 2.530501127243042, + "learning_rate": 1.939552785180466e-05, + "loss": 0.5872, + "step": 4374 + }, + { + "epoch": 0.13799085317773221, + "grad_norm": 2.969075918197632, + "learning_rate": 1.939517801765786e-05, + "loss": 0.5747, + "step": 4375 + }, + { + "epoch": 0.13802239394417284, + "grad_norm": 3.2589592933654785, + "learning_rate": 1.939482808546508e-05, + "loss": 0.9397, + "step": 4376 + }, + { + "epoch": 0.13805393471061347, + "grad_norm": 3.306469678878784, + "learning_rate": 1.939447805522998e-05, + "loss": 0.6034, + "step": 4377 + }, + { + "epoch": 0.1380854754770541, + "grad_norm": 3.952378988265991, + "learning_rate": 1.93941279269562e-05, + "loss": 0.62, + "step": 4378 + }, + { + "epoch": 0.1381170162434947, + "grad_norm": 3.3587403297424316, + "learning_rate": 1.9393777700647403e-05, + "loss": 0.6157, + "step": 4379 + }, + { + "epoch": 0.13814855700993534, + "grad_norm": 2.9316136837005615, + "learning_rate": 1.939342737630724e-05, + "loss": 0.5844, + "step": 4380 + }, + { + "epoch": 0.13818009777637597, + "grad_norm": 2.559535026550293, + "learning_rate": 1.939307695393937e-05, + "loss": 0.6046, + "step": 4381 + }, + { + "epoch": 0.1382116385428166, + "grad_norm": 4.421999454498291, + "learning_rate": 1.939272643354745e-05, + "loss": 0.6363, + "step": 4382 + }, + { + "epoch": 0.13824317930925722, + "grad_norm": 3.8170812129974365, + "learning_rate": 1.9392375815135136e-05, + "loss": 0.5854, + "step": 4383 + }, + { + "epoch": 0.13827472007569783, + "grad_norm": 2.680070161819458, + "learning_rate": 1.9392025098706086e-05, + "loss": 0.5973, + "step": 4384 + }, + { + "epoch": 0.13830626084213846, + "grad_norm": 3.9344046115875244, + "learning_rate": 1.939167428426396e-05, + "loss": 0.5718, + "step": 4385 + }, + { + "epoch": 0.1383378016085791, + "grad_norm": 2.892819404602051, + "learning_rate": 1.9391323371812425e-05, + "loss": 0.6128, + "step": 4386 + }, + { + "epoch": 0.13836934237501972, + "grad_norm": 2.6283719539642334, + "learning_rate": 1.9390972361355132e-05, + "loss": 0.5597, + "step": 4387 + }, + { + "epoch": 0.13840088314146035, + "grad_norm": 4.689798831939697, + "learning_rate": 1.9390621252895755e-05, + "loss": 0.4648, + "step": 4388 + }, + { + "epoch": 0.13843242390790098, + "grad_norm": 2.067275047302246, + "learning_rate": 1.9390270046437955e-05, + "loss": 0.9106, + "step": 4389 + }, + { + "epoch": 0.13846396467434158, + "grad_norm": 4.186642646789551, + "learning_rate": 1.9389918741985393e-05, + "loss": 0.6129, + "step": 4390 + }, + { + "epoch": 0.1384955054407822, + "grad_norm": 5.684109687805176, + "learning_rate": 1.938956733954174e-05, + "loss": 0.5675, + "step": 4391 + }, + { + "epoch": 0.13852704620722284, + "grad_norm": 5.104311466217041, + "learning_rate": 1.938921583911066e-05, + "loss": 0.6104, + "step": 4392 + }, + { + "epoch": 0.13855858697366347, + "grad_norm": 3.8904011249542236, + "learning_rate": 1.9388864240695818e-05, + "loss": 0.5937, + "step": 4393 + }, + { + "epoch": 0.1385901277401041, + "grad_norm": 4.799550533294678, + "learning_rate": 1.9388512544300894e-05, + "loss": 0.5991, + "step": 4394 + }, + { + "epoch": 0.1386216685065447, + "grad_norm": 6.211132526397705, + "learning_rate": 1.9388160749929547e-05, + "loss": 0.6128, + "step": 4395 + }, + { + "epoch": 0.13865320927298533, + "grad_norm": 3.5383923053741455, + "learning_rate": 1.9387808857585456e-05, + "loss": 0.5837, + "step": 4396 + }, + { + "epoch": 0.13868475003942596, + "grad_norm": 2.765167713165283, + "learning_rate": 1.9387456867272288e-05, + "loss": 0.5828, + "step": 4397 + }, + { + "epoch": 0.1387162908058666, + "grad_norm": 1.9577823877334595, + "learning_rate": 1.938710477899372e-05, + "loss": 0.9303, + "step": 4398 + }, + { + "epoch": 0.13874783157230722, + "grad_norm": 27.44800567626953, + "learning_rate": 1.938675259275342e-05, + "loss": 0.6121, + "step": 4399 + }, + { + "epoch": 0.13877937233874782, + "grad_norm": 2.734877347946167, + "learning_rate": 1.9386400308555076e-05, + "loss": 0.5504, + "step": 4400 + }, + { + "epoch": 0.13881091310518845, + "grad_norm": 3.060702323913574, + "learning_rate": 1.9386047926402352e-05, + "loss": 0.592, + "step": 4401 + }, + { + "epoch": 0.13884245387162908, + "grad_norm": 3.5059258937835693, + "learning_rate": 1.9385695446298925e-05, + "loss": 0.6108, + "step": 4402 + }, + { + "epoch": 0.1388739946380697, + "grad_norm": 3.116598129272461, + "learning_rate": 1.9385342868248487e-05, + "loss": 0.5969, + "step": 4403 + }, + { + "epoch": 0.13890553540451034, + "grad_norm": 3.2935729026794434, + "learning_rate": 1.9384990192254704e-05, + "loss": 0.591, + "step": 4404 + }, + { + "epoch": 0.13893707617095094, + "grad_norm": 3.428786277770996, + "learning_rate": 1.938463741832126e-05, + "loss": 0.5943, + "step": 4405 + }, + { + "epoch": 0.13896861693739157, + "grad_norm": 2.861354351043701, + "learning_rate": 1.938428454645184e-05, + "loss": 0.573, + "step": 4406 + }, + { + "epoch": 0.1390001577038322, + "grad_norm": 7.014374732971191, + "learning_rate": 1.9383931576650124e-05, + "loss": 0.4341, + "step": 4407 + }, + { + "epoch": 0.13903169847027283, + "grad_norm": 3.285902738571167, + "learning_rate": 1.9383578508919795e-05, + "loss": 0.5997, + "step": 4408 + }, + { + "epoch": 0.13906323923671346, + "grad_norm": 5.340433597564697, + "learning_rate": 1.938322534326454e-05, + "loss": 0.5946, + "step": 4409 + }, + { + "epoch": 0.1390947800031541, + "grad_norm": 2.8603999614715576, + "learning_rate": 1.938287207968804e-05, + "loss": 0.5844, + "step": 4410 + }, + { + "epoch": 0.1391263207695947, + "grad_norm": 2.9842729568481445, + "learning_rate": 1.9382518718193986e-05, + "loss": 0.6069, + "step": 4411 + }, + { + "epoch": 0.13915786153603532, + "grad_norm": 2.762352705001831, + "learning_rate": 1.938216525878606e-05, + "loss": 0.6203, + "step": 4412 + }, + { + "epoch": 0.13918940230247595, + "grad_norm": 4.751638412475586, + "learning_rate": 1.938181170146796e-05, + "loss": 0.6013, + "step": 4413 + }, + { + "epoch": 0.13922094306891658, + "grad_norm": 3.8208818435668945, + "learning_rate": 1.9381458046243366e-05, + "loss": 0.9134, + "step": 4414 + }, + { + "epoch": 0.1392524838353572, + "grad_norm": 3.32558536529541, + "learning_rate": 1.9381104293115976e-05, + "loss": 0.5907, + "step": 4415 + }, + { + "epoch": 0.1392840246017978, + "grad_norm": 5.212766170501709, + "learning_rate": 1.938075044208948e-05, + "loss": 0.5888, + "step": 4416 + }, + { + "epoch": 0.13931556536823844, + "grad_norm": 2.6951687335968018, + "learning_rate": 1.9380396493167563e-05, + "loss": 0.624, + "step": 4417 + }, + { + "epoch": 0.13934710613467907, + "grad_norm": 3.553300619125366, + "learning_rate": 1.9380042446353928e-05, + "loss": 0.588, + "step": 4418 + }, + { + "epoch": 0.1393786469011197, + "grad_norm": 2.708501100540161, + "learning_rate": 1.9379688301652265e-05, + "loss": 0.5808, + "step": 4419 + }, + { + "epoch": 0.13941018766756033, + "grad_norm": 2.841461658477783, + "learning_rate": 1.9379334059066273e-05, + "loss": 0.6066, + "step": 4420 + }, + { + "epoch": 0.13944172843400093, + "grad_norm": 4.10732364654541, + "learning_rate": 1.9378979718599647e-05, + "loss": 0.5852, + "step": 4421 + }, + { + "epoch": 0.13947326920044156, + "grad_norm": 2.8439741134643555, + "learning_rate": 1.9378625280256087e-05, + "loss": 0.8999, + "step": 4422 + }, + { + "epoch": 0.1395048099668822, + "grad_norm": 4.9054951667785645, + "learning_rate": 1.9378270744039286e-05, + "loss": 0.589, + "step": 4423 + }, + { + "epoch": 0.13953635073332282, + "grad_norm": 4.015968322753906, + "learning_rate": 1.9377916109952954e-05, + "loss": 0.5695, + "step": 4424 + }, + { + "epoch": 0.13956789149976345, + "grad_norm": 2.4677557945251465, + "learning_rate": 1.937756137800078e-05, + "loss": 0.5542, + "step": 4425 + }, + { + "epoch": 0.13959943226620408, + "grad_norm": 3.7696900367736816, + "learning_rate": 1.937720654818647e-05, + "loss": 0.5861, + "step": 4426 + }, + { + "epoch": 0.13963097303264468, + "grad_norm": 14.074225425720215, + "learning_rate": 1.937685162051373e-05, + "loss": 0.6509, + "step": 4427 + }, + { + "epoch": 0.1396625137990853, + "grad_norm": 3.7839531898498535, + "learning_rate": 1.9376496594986265e-05, + "loss": 0.5702, + "step": 4428 + }, + { + "epoch": 0.13969405456552594, + "grad_norm": 2.809305191040039, + "learning_rate": 1.937614147160778e-05, + "loss": 0.5794, + "step": 4429 + }, + { + "epoch": 0.13972559533196657, + "grad_norm": 2.4908623695373535, + "learning_rate": 1.937578625038197e-05, + "loss": 0.5817, + "step": 4430 + }, + { + "epoch": 0.1397571360984072, + "grad_norm": 2.6353037357330322, + "learning_rate": 1.937543093131256e-05, + "loss": 0.5757, + "step": 4431 + }, + { + "epoch": 0.1397886768648478, + "grad_norm": 3.0570216178894043, + "learning_rate": 1.937507551440324e-05, + "loss": 0.5965, + "step": 4432 + }, + { + "epoch": 0.13982021763128843, + "grad_norm": 3.381089210510254, + "learning_rate": 1.937471999965773e-05, + "loss": 0.5763, + "step": 4433 + }, + { + "epoch": 0.13985175839772906, + "grad_norm": 5.777001857757568, + "learning_rate": 1.937436438707974e-05, + "loss": 0.576, + "step": 4434 + }, + { + "epoch": 0.1398832991641697, + "grad_norm": 2.980950117111206, + "learning_rate": 1.9374008676672975e-05, + "loss": 0.5892, + "step": 4435 + }, + { + "epoch": 0.13991483993061032, + "grad_norm": 3.197906970977783, + "learning_rate": 1.9373652868441155e-05, + "loss": 0.5677, + "step": 4436 + }, + { + "epoch": 0.13994638069705093, + "grad_norm": 3.7464072704315186, + "learning_rate": 1.9373296962387988e-05, + "loss": 0.5612, + "step": 4437 + }, + { + "epoch": 0.13997792146349156, + "grad_norm": 3.8260748386383057, + "learning_rate": 1.9372940958517186e-05, + "loss": 0.5493, + "step": 4438 + }, + { + "epoch": 0.14000946222993219, + "grad_norm": 3.1047494411468506, + "learning_rate": 1.937258485683247e-05, + "loss": 0.5773, + "step": 4439 + }, + { + "epoch": 0.14004100299637282, + "grad_norm": 3.2582809925079346, + "learning_rate": 1.937222865733755e-05, + "loss": 0.636, + "step": 4440 + }, + { + "epoch": 0.14007254376281345, + "grad_norm": 4.336559772491455, + "learning_rate": 1.937187236003615e-05, + "loss": 0.6088, + "step": 4441 + }, + { + "epoch": 0.14010408452925405, + "grad_norm": 6.14829683303833, + "learning_rate": 1.9371515964931982e-05, + "loss": 0.5855, + "step": 4442 + }, + { + "epoch": 0.14013562529569468, + "grad_norm": 26.12579917907715, + "learning_rate": 1.937115947202877e-05, + "loss": 0.5758, + "step": 4443 + }, + { + "epoch": 0.1401671660621353, + "grad_norm": 3.14318585395813, + "learning_rate": 1.9370802881330233e-05, + "loss": 0.5995, + "step": 4444 + }, + { + "epoch": 0.14019870682857594, + "grad_norm": 6.958794116973877, + "learning_rate": 1.9370446192840087e-05, + "loss": 0.5975, + "step": 4445 + }, + { + "epoch": 0.14023024759501657, + "grad_norm": 4.629293441772461, + "learning_rate": 1.9370089406562065e-05, + "loss": 0.5799, + "step": 4446 + }, + { + "epoch": 0.1402617883614572, + "grad_norm": 3.480706214904785, + "learning_rate": 1.9369732522499883e-05, + "loss": 0.5736, + "step": 4447 + }, + { + "epoch": 0.1402933291278978, + "grad_norm": 4.175063610076904, + "learning_rate": 1.9369375540657267e-05, + "loss": 0.5649, + "step": 4448 + }, + { + "epoch": 0.14032486989433843, + "grad_norm": 3.5910215377807617, + "learning_rate": 1.936901846103794e-05, + "loss": 0.5739, + "step": 4449 + }, + { + "epoch": 0.14035641066077906, + "grad_norm": 5.0239763259887695, + "learning_rate": 1.936866128364563e-05, + "loss": 0.6275, + "step": 4450 + }, + { + "epoch": 0.1403879514272197, + "grad_norm": 6.986630916595459, + "learning_rate": 1.9368304008484068e-05, + "loss": 0.4249, + "step": 4451 + }, + { + "epoch": 0.14041949219366032, + "grad_norm": 2.992119550704956, + "learning_rate": 1.9367946635556976e-05, + "loss": 0.5641, + "step": 4452 + }, + { + "epoch": 0.14045103296010092, + "grad_norm": 4.0857391357421875, + "learning_rate": 1.936758916486809e-05, + "loss": 0.5811, + "step": 4453 + }, + { + "epoch": 0.14048257372654155, + "grad_norm": 2.183324098587036, + "learning_rate": 1.9367231596421138e-05, + "loss": 0.9312, + "step": 4454 + }, + { + "epoch": 0.14051411449298218, + "grad_norm": 3.3397321701049805, + "learning_rate": 1.936687393021985e-05, + "loss": 0.6009, + "step": 4455 + }, + { + "epoch": 0.1405456552594228, + "grad_norm": 3.6155219078063965, + "learning_rate": 1.9366516166267955e-05, + "loss": 0.5789, + "step": 4456 + }, + { + "epoch": 0.14057719602586344, + "grad_norm": 3.6568961143493652, + "learning_rate": 1.9366158304569195e-05, + "loss": 0.5595, + "step": 4457 + }, + { + "epoch": 0.14060873679230404, + "grad_norm": 3.8436715602874756, + "learning_rate": 1.9365800345127298e-05, + "loss": 0.5967, + "step": 4458 + }, + { + "epoch": 0.14064027755874467, + "grad_norm": 2.9390580654144287, + "learning_rate": 1.9365442287946005e-05, + "loss": 0.5937, + "step": 4459 + }, + { + "epoch": 0.1406718183251853, + "grad_norm": 3.6921567916870117, + "learning_rate": 1.936508413302905e-05, + "loss": 0.6058, + "step": 4460 + }, + { + "epoch": 0.14070335909162593, + "grad_norm": 3.1974680423736572, + "learning_rate": 1.9364725880380163e-05, + "loss": 0.5912, + "step": 4461 + }, + { + "epoch": 0.14073489985806656, + "grad_norm": 2.5467255115509033, + "learning_rate": 1.9364367530003098e-05, + "loss": 0.5926, + "step": 4462 + }, + { + "epoch": 0.1407664406245072, + "grad_norm": 8.589229583740234, + "learning_rate": 1.936400908190158e-05, + "loss": 0.6132, + "step": 4463 + }, + { + "epoch": 0.1407979813909478, + "grad_norm": 2.7767062187194824, + "learning_rate": 1.9363650536079365e-05, + "loss": 0.599, + "step": 4464 + }, + { + "epoch": 0.14082952215738842, + "grad_norm": 3.334972381591797, + "learning_rate": 1.9363291892540182e-05, + "loss": 0.584, + "step": 4465 + }, + { + "epoch": 0.14086106292382905, + "grad_norm": 3.267894983291626, + "learning_rate": 1.9362933151287776e-05, + "loss": 0.5594, + "step": 4466 + }, + { + "epoch": 0.14089260369026968, + "grad_norm": 2.6477973461151123, + "learning_rate": 1.936257431232589e-05, + "loss": 0.558, + "step": 4467 + }, + { + "epoch": 0.1409241444567103, + "grad_norm": 3.471048593521118, + "learning_rate": 1.936221537565828e-05, + "loss": 0.5486, + "step": 4468 + }, + { + "epoch": 0.1409556852231509, + "grad_norm": 5.475439071655273, + "learning_rate": 1.9361856341288675e-05, + "loss": 0.5705, + "step": 4469 + }, + { + "epoch": 0.14098722598959154, + "grad_norm": 2.7123358249664307, + "learning_rate": 1.936149720922083e-05, + "loss": 0.603, + "step": 4470 + }, + { + "epoch": 0.14101876675603217, + "grad_norm": 2.4781253337860107, + "learning_rate": 1.9361137979458496e-05, + "loss": 0.5579, + "step": 4471 + }, + { + "epoch": 0.1410503075224728, + "grad_norm": 4.013701438903809, + "learning_rate": 1.9360778652005416e-05, + "loss": 0.5569, + "step": 4472 + }, + { + "epoch": 0.14108184828891343, + "grad_norm": 3.3133416175842285, + "learning_rate": 1.9360419226865344e-05, + "loss": 0.5983, + "step": 4473 + }, + { + "epoch": 0.14111338905535403, + "grad_norm": 2.85267972946167, + "learning_rate": 1.936005970404203e-05, + "loss": 0.6101, + "step": 4474 + }, + { + "epoch": 0.14114492982179466, + "grad_norm": 3.6411163806915283, + "learning_rate": 1.9359700083539224e-05, + "loss": 0.5819, + "step": 4475 + }, + { + "epoch": 0.1411764705882353, + "grad_norm": 3.2211661338806152, + "learning_rate": 1.9359340365360678e-05, + "loss": 0.4703, + "step": 4476 + }, + { + "epoch": 0.14120801135467592, + "grad_norm": 3.297743082046509, + "learning_rate": 1.9358980549510152e-05, + "loss": 0.6035, + "step": 4477 + }, + { + "epoch": 0.14123955212111655, + "grad_norm": 3.1980271339416504, + "learning_rate": 1.9358620635991394e-05, + "loss": 0.6025, + "step": 4478 + }, + { + "epoch": 0.14127109288755718, + "grad_norm": 2.5237536430358887, + "learning_rate": 1.9358260624808163e-05, + "loss": 0.5816, + "step": 4479 + }, + { + "epoch": 0.14130263365399778, + "grad_norm": 11.751012802124023, + "learning_rate": 1.9357900515964217e-05, + "loss": 0.5825, + "step": 4480 + }, + { + "epoch": 0.1413341744204384, + "grad_norm": 2.5012049674987793, + "learning_rate": 1.9357540309463313e-05, + "loss": 0.9446, + "step": 4481 + }, + { + "epoch": 0.14136571518687904, + "grad_norm": 3.4880757331848145, + "learning_rate": 1.935718000530921e-05, + "loss": 0.5748, + "step": 4482 + }, + { + "epoch": 0.14139725595331967, + "grad_norm": 5.809312343597412, + "learning_rate": 1.9356819603505665e-05, + "loss": 0.5976, + "step": 4483 + }, + { + "epoch": 0.1414287967197603, + "grad_norm": 3.3570244312286377, + "learning_rate": 1.9356459104056444e-05, + "loss": 0.5618, + "step": 4484 + }, + { + "epoch": 0.1414603374862009, + "grad_norm": 2.1401147842407227, + "learning_rate": 1.935609850696531e-05, + "loss": 0.5764, + "step": 4485 + }, + { + "epoch": 0.14149187825264153, + "grad_norm": 2.354172468185425, + "learning_rate": 1.935573781223602e-05, + "loss": 0.8833, + "step": 4486 + }, + { + "epoch": 0.14152341901908216, + "grad_norm": 4.932677268981934, + "learning_rate": 1.9355377019872343e-05, + "loss": 0.8867, + "step": 4487 + }, + { + "epoch": 0.1415549597855228, + "grad_norm": 3.7598161697387695, + "learning_rate": 1.9355016129878043e-05, + "loss": 0.5899, + "step": 4488 + }, + { + "epoch": 0.14158650055196342, + "grad_norm": 6.031693458557129, + "learning_rate": 1.9354655142256883e-05, + "loss": 0.6379, + "step": 4489 + }, + { + "epoch": 0.14161804131840403, + "grad_norm": 2.9558866024017334, + "learning_rate": 1.9354294057012635e-05, + "loss": 0.5817, + "step": 4490 + }, + { + "epoch": 0.14164958208484466, + "grad_norm": 2.8480889797210693, + "learning_rate": 1.9353932874149062e-05, + "loss": 0.5584, + "step": 4491 + }, + { + "epoch": 0.14168112285128528, + "grad_norm": 2.4356558322906494, + "learning_rate": 1.935357159366994e-05, + "loss": 0.5923, + "step": 4492 + }, + { + "epoch": 0.14171266361772591, + "grad_norm": 4.728663921356201, + "learning_rate": 1.9353210215579033e-05, + "loss": 0.6151, + "step": 4493 + }, + { + "epoch": 0.14174420438416654, + "grad_norm": 3.068356513977051, + "learning_rate": 1.935284873988012e-05, + "loss": 0.5584, + "step": 4494 + }, + { + "epoch": 0.14177574515060715, + "grad_norm": 3.647186756134033, + "learning_rate": 1.935248716657696e-05, + "loss": 0.5933, + "step": 4495 + }, + { + "epoch": 0.14180728591704778, + "grad_norm": 3.532515048980713, + "learning_rate": 1.935212549567334e-05, + "loss": 0.5595, + "step": 4496 + }, + { + "epoch": 0.1418388266834884, + "grad_norm": 3.6735992431640625, + "learning_rate": 1.9351763727173027e-05, + "loss": 0.5475, + "step": 4497 + }, + { + "epoch": 0.14187036744992904, + "grad_norm": 2.264633893966675, + "learning_rate": 1.9351401861079798e-05, + "loss": 0.8963, + "step": 4498 + }, + { + "epoch": 0.14190190821636967, + "grad_norm": 3.198639154434204, + "learning_rate": 1.935103989739743e-05, + "loss": 0.5421, + "step": 4499 + }, + { + "epoch": 0.1419334489828103, + "grad_norm": 3.366716146469116, + "learning_rate": 1.93506778361297e-05, + "loss": 0.5911, + "step": 4500 + }, + { + "epoch": 0.1419649897492509, + "grad_norm": 3.518004894256592, + "learning_rate": 1.9350315677280383e-05, + "loss": 0.5835, + "step": 4501 + }, + { + "epoch": 0.14199653051569153, + "grad_norm": 5.262356758117676, + "learning_rate": 1.9349953420853267e-05, + "loss": 0.4235, + "step": 4502 + }, + { + "epoch": 0.14202807128213216, + "grad_norm": 2.582728862762451, + "learning_rate": 1.9349591066852123e-05, + "loss": 0.6036, + "step": 4503 + }, + { + "epoch": 0.1420596120485728, + "grad_norm": 3.5789198875427246, + "learning_rate": 1.9349228615280736e-05, + "loss": 0.5926, + "step": 4504 + }, + { + "epoch": 0.14209115281501342, + "grad_norm": 3.166055917739868, + "learning_rate": 1.9348866066142893e-05, + "loss": 0.5997, + "step": 4505 + }, + { + "epoch": 0.14212269358145402, + "grad_norm": 2.9597485065460205, + "learning_rate": 1.934850341944237e-05, + "loss": 0.4438, + "step": 4506 + }, + { + "epoch": 0.14215423434789465, + "grad_norm": 2.652303457260132, + "learning_rate": 1.9348140675182957e-05, + "loss": 0.5868, + "step": 4507 + }, + { + "epoch": 0.14218577511433528, + "grad_norm": 2.3177621364593506, + "learning_rate": 1.9347777833368434e-05, + "loss": 0.9396, + "step": 4508 + }, + { + "epoch": 0.1422173158807759, + "grad_norm": 2.452523708343506, + "learning_rate": 1.9347414894002593e-05, + "loss": 0.5822, + "step": 4509 + }, + { + "epoch": 0.14224885664721654, + "grad_norm": 3.05607533454895, + "learning_rate": 1.9347051857089223e-05, + "loss": 0.6123, + "step": 4510 + }, + { + "epoch": 0.14228039741365714, + "grad_norm": 2.5014374256134033, + "learning_rate": 1.9346688722632104e-05, + "loss": 0.5716, + "step": 4511 + }, + { + "epoch": 0.14231193818009777, + "grad_norm": 4.046553611755371, + "learning_rate": 1.9346325490635036e-05, + "loss": 0.4362, + "step": 4512 + }, + { + "epoch": 0.1423434789465384, + "grad_norm": 2.8549933433532715, + "learning_rate": 1.9345962161101797e-05, + "loss": 0.5837, + "step": 4513 + }, + { + "epoch": 0.14237501971297903, + "grad_norm": 14.617403030395508, + "learning_rate": 1.9345598734036192e-05, + "loss": 0.6007, + "step": 4514 + }, + { + "epoch": 0.14240656047941966, + "grad_norm": 2.22765851020813, + "learning_rate": 1.9345235209442003e-05, + "loss": 0.9619, + "step": 4515 + }, + { + "epoch": 0.1424381012458603, + "grad_norm": 3.0678000450134277, + "learning_rate": 1.9344871587323033e-05, + "loss": 0.5837, + "step": 4516 + }, + { + "epoch": 0.1424696420123009, + "grad_norm": 3.3914544582366943, + "learning_rate": 1.9344507867683065e-05, + "loss": 0.5881, + "step": 4517 + }, + { + "epoch": 0.14250118277874152, + "grad_norm": 2.922086715698242, + "learning_rate": 1.934414405052591e-05, + "loss": 0.4432, + "step": 4518 + }, + { + "epoch": 0.14253272354518215, + "grad_norm": 5.754217624664307, + "learning_rate": 1.934378013585535e-05, + "loss": 0.5869, + "step": 4519 + }, + { + "epoch": 0.14256426431162278, + "grad_norm": 2.547619342803955, + "learning_rate": 1.934341612367519e-05, + "loss": 0.3843, + "step": 4520 + }, + { + "epoch": 0.1425958050780634, + "grad_norm": 2.5760247707366943, + "learning_rate": 1.9343052013989234e-05, + "loss": 0.5952, + "step": 4521 + }, + { + "epoch": 0.142627345844504, + "grad_norm": 2.8536465167999268, + "learning_rate": 1.934268780680127e-05, + "loss": 0.5576, + "step": 4522 + }, + { + "epoch": 0.14265888661094464, + "grad_norm": 2.519244432449341, + "learning_rate": 1.9342323502115103e-05, + "loss": 0.5451, + "step": 4523 + }, + { + "epoch": 0.14269042737738527, + "grad_norm": 7.040013790130615, + "learning_rate": 1.9341959099934536e-05, + "loss": 0.5911, + "step": 4524 + }, + { + "epoch": 0.1427219681438259, + "grad_norm": 3.1916191577911377, + "learning_rate": 1.9341594600263374e-05, + "loss": 0.5976, + "step": 4525 + }, + { + "epoch": 0.14275350891026653, + "grad_norm": 3.367593288421631, + "learning_rate": 1.934123000310542e-05, + "loss": 0.5926, + "step": 4526 + }, + { + "epoch": 0.14278504967670713, + "grad_norm": 7.231448173522949, + "learning_rate": 1.9340865308464477e-05, + "loss": 0.5797, + "step": 4527 + }, + { + "epoch": 0.14281659044314776, + "grad_norm": 2.646397113800049, + "learning_rate": 1.934050051634435e-05, + "loss": 0.5898, + "step": 4528 + }, + { + "epoch": 0.1428481312095884, + "grad_norm": 2.4427008628845215, + "learning_rate": 1.934013562674885e-05, + "loss": 0.6004, + "step": 4529 + }, + { + "epoch": 0.14287967197602902, + "grad_norm": 2.48079514503479, + "learning_rate": 1.9339770639681777e-05, + "loss": 0.8962, + "step": 4530 + }, + { + "epoch": 0.14291121274246965, + "grad_norm": 3.128120183944702, + "learning_rate": 1.933940555514695e-05, + "loss": 0.5833, + "step": 4531 + }, + { + "epoch": 0.14294275350891025, + "grad_norm": 3.7487752437591553, + "learning_rate": 1.9339040373148173e-05, + "loss": 0.6174, + "step": 4532 + }, + { + "epoch": 0.14297429427535088, + "grad_norm": 2.9118080139160156, + "learning_rate": 1.9338675093689257e-05, + "loss": 0.5728, + "step": 4533 + }, + { + "epoch": 0.1430058350417915, + "grad_norm": 2.378549098968506, + "learning_rate": 1.9338309716774017e-05, + "loss": 0.9695, + "step": 4534 + }, + { + "epoch": 0.14303737580823214, + "grad_norm": 2.617636203765869, + "learning_rate": 1.9337944242406264e-05, + "loss": 0.5915, + "step": 4535 + }, + { + "epoch": 0.14306891657467277, + "grad_norm": 4.602554798126221, + "learning_rate": 1.933757867058981e-05, + "loss": 0.4394, + "step": 4536 + }, + { + "epoch": 0.1431004573411134, + "grad_norm": 3.1438560485839844, + "learning_rate": 1.9337213001328474e-05, + "loss": 0.5911, + "step": 4537 + }, + { + "epoch": 0.143131998107554, + "grad_norm": 2.3707826137542725, + "learning_rate": 1.933684723462607e-05, + "loss": 0.91, + "step": 4538 + }, + { + "epoch": 0.14316353887399463, + "grad_norm": 7.87752628326416, + "learning_rate": 1.9336481370486416e-05, + "loss": 0.5654, + "step": 4539 + }, + { + "epoch": 0.14319507964043526, + "grad_norm": 3.144428253173828, + "learning_rate": 1.9336115408913328e-05, + "loss": 0.5714, + "step": 4540 + }, + { + "epoch": 0.1432266204068759, + "grad_norm": 2.4621756076812744, + "learning_rate": 1.9335749349910628e-05, + "loss": 0.5725, + "step": 4541 + }, + { + "epoch": 0.14325816117331652, + "grad_norm": 7.898487567901611, + "learning_rate": 1.933538319348213e-05, + "loss": 0.592, + "step": 4542 + }, + { + "epoch": 0.14328970193975712, + "grad_norm": 2.4243037700653076, + "learning_rate": 1.9335016939631664e-05, + "loss": 0.8724, + "step": 4543 + }, + { + "epoch": 0.14332124270619775, + "grad_norm": 3.023855686187744, + "learning_rate": 1.9334650588363047e-05, + "loss": 0.4004, + "step": 4544 + }, + { + "epoch": 0.14335278347263838, + "grad_norm": 2.7206482887268066, + "learning_rate": 1.9334284139680102e-05, + "loss": 0.5553, + "step": 4545 + }, + { + "epoch": 0.14338432423907901, + "grad_norm": 4.142518043518066, + "learning_rate": 1.9333917593586654e-05, + "loss": 0.566, + "step": 4546 + }, + { + "epoch": 0.14341586500551964, + "grad_norm": 3.668588876724243, + "learning_rate": 1.9333550950086527e-05, + "loss": 0.5941, + "step": 4547 + }, + { + "epoch": 0.14344740577196025, + "grad_norm": 2.7604568004608154, + "learning_rate": 1.9333184209183554e-05, + "loss": 0.6019, + "step": 4548 + }, + { + "epoch": 0.14347894653840088, + "grad_norm": 4.311107158660889, + "learning_rate": 1.933281737088155e-05, + "loss": 0.5812, + "step": 4549 + }, + { + "epoch": 0.1435104873048415, + "grad_norm": 16.16173553466797, + "learning_rate": 1.9332450435184354e-05, + "loss": 0.5798, + "step": 4550 + }, + { + "epoch": 0.14354202807128214, + "grad_norm": 3.3196215629577637, + "learning_rate": 1.933208340209579e-05, + "loss": 0.6159, + "step": 4551 + }, + { + "epoch": 0.14357356883772276, + "grad_norm": 11.82988166809082, + "learning_rate": 1.9331716271619692e-05, + "loss": 0.5773, + "step": 4552 + }, + { + "epoch": 0.1436051096041634, + "grad_norm": 2.6678335666656494, + "learning_rate": 1.9331349043759887e-05, + "loss": 0.577, + "step": 4553 + }, + { + "epoch": 0.143636650370604, + "grad_norm": 4.491864204406738, + "learning_rate": 1.933098171852021e-05, + "loss": 0.6138, + "step": 4554 + }, + { + "epoch": 0.14366819113704463, + "grad_norm": 3.036017656326294, + "learning_rate": 1.933061429590449e-05, + "loss": 0.6065, + "step": 4555 + }, + { + "epoch": 0.14369973190348526, + "grad_norm": 3.3548080921173096, + "learning_rate": 1.933024677591657e-05, + "loss": 0.5615, + "step": 4556 + }, + { + "epoch": 0.14373127266992589, + "grad_norm": 3.2293426990509033, + "learning_rate": 1.9329879158560274e-05, + "loss": 0.5399, + "step": 4557 + }, + { + "epoch": 0.14376281343636652, + "grad_norm": 4.626341819763184, + "learning_rate": 1.932951144383945e-05, + "loss": 0.8961, + "step": 4558 + }, + { + "epoch": 0.14379435420280712, + "grad_norm": 2.8061723709106445, + "learning_rate": 1.9329143631757926e-05, + "loss": 0.5631, + "step": 4559 + }, + { + "epoch": 0.14382589496924775, + "grad_norm": 4.374220371246338, + "learning_rate": 1.932877572231955e-05, + "loss": 0.4141, + "step": 4560 + }, + { + "epoch": 0.14385743573568838, + "grad_norm": 2.9715733528137207, + "learning_rate": 1.9328407715528147e-05, + "loss": 0.5433, + "step": 4561 + }, + { + "epoch": 0.143888976502129, + "grad_norm": 2.263739824295044, + "learning_rate": 1.9328039611387574e-05, + "loss": 0.9271, + "step": 4562 + }, + { + "epoch": 0.14392051726856964, + "grad_norm": 3.7791945934295654, + "learning_rate": 1.932767140990166e-05, + "loss": 0.5596, + "step": 4563 + }, + { + "epoch": 0.14395205803501024, + "grad_norm": 5.565816879272461, + "learning_rate": 1.9327303111074255e-05, + "loss": 0.6013, + "step": 4564 + }, + { + "epoch": 0.14398359880145087, + "grad_norm": 3.6306607723236084, + "learning_rate": 1.9326934714909197e-05, + "loss": 0.5876, + "step": 4565 + }, + { + "epoch": 0.1440151395678915, + "grad_norm": 5.500343322753906, + "learning_rate": 1.9326566221410332e-05, + "loss": 0.5868, + "step": 4566 + }, + { + "epoch": 0.14404668033433213, + "grad_norm": 3.670175790786743, + "learning_rate": 1.932619763058151e-05, + "loss": 0.5915, + "step": 4567 + }, + { + "epoch": 0.14407822110077276, + "grad_norm": 3.8808999061584473, + "learning_rate": 1.9325828942426574e-05, + "loss": 0.5947, + "step": 4568 + }, + { + "epoch": 0.14410976186721336, + "grad_norm": 5.897028923034668, + "learning_rate": 1.9325460156949368e-05, + "loss": 0.5853, + "step": 4569 + }, + { + "epoch": 0.144141302633654, + "grad_norm": 2.937437057495117, + "learning_rate": 1.9325091274153748e-05, + "loss": 0.5607, + "step": 4570 + }, + { + "epoch": 0.14417284340009462, + "grad_norm": 3.709728479385376, + "learning_rate": 1.932472229404356e-05, + "loss": 0.6015, + "step": 4571 + }, + { + "epoch": 0.14420438416653525, + "grad_norm": 2.374786853790283, + "learning_rate": 1.9324353216622654e-05, + "loss": 0.9072, + "step": 4572 + }, + { + "epoch": 0.14423592493297588, + "grad_norm": 3.721223831176758, + "learning_rate": 1.932398404189488e-05, + "loss": 0.5884, + "step": 4573 + }, + { + "epoch": 0.1442674656994165, + "grad_norm": 4.526257038116455, + "learning_rate": 1.9323614769864093e-05, + "loss": 0.5991, + "step": 4574 + }, + { + "epoch": 0.1442990064658571, + "grad_norm": 3.06355619430542, + "learning_rate": 1.9323245400534147e-05, + "loss": 0.6121, + "step": 4575 + }, + { + "epoch": 0.14433054723229774, + "grad_norm": 4.102237701416016, + "learning_rate": 1.9322875933908894e-05, + "loss": 0.567, + "step": 4576 + }, + { + "epoch": 0.14436208799873837, + "grad_norm": 3.2817578315734863, + "learning_rate": 1.9322506369992194e-05, + "loss": 0.5852, + "step": 4577 + }, + { + "epoch": 0.144393628765179, + "grad_norm": 5.174901962280273, + "learning_rate": 1.9322136708787905e-05, + "loss": 0.5939, + "step": 4578 + }, + { + "epoch": 0.14442516953161963, + "grad_norm": 12.058923721313477, + "learning_rate": 1.9321766950299876e-05, + "loss": 0.5746, + "step": 4579 + }, + { + "epoch": 0.14445671029806023, + "grad_norm": 3.9779698848724365, + "learning_rate": 1.932139709453197e-05, + "loss": 0.5872, + "step": 4580 + }, + { + "epoch": 0.14448825106450086, + "grad_norm": 3.5969655513763428, + "learning_rate": 1.932102714148805e-05, + "loss": 0.5621, + "step": 4581 + }, + { + "epoch": 0.1445197918309415, + "grad_norm": 5.391377925872803, + "learning_rate": 1.9320657091171972e-05, + "loss": 0.4157, + "step": 4582 + }, + { + "epoch": 0.14455133259738212, + "grad_norm": 13.74290657043457, + "learning_rate": 1.9320286943587602e-05, + "loss": 0.5913, + "step": 4583 + }, + { + "epoch": 0.14458287336382275, + "grad_norm": 3.1019301414489746, + "learning_rate": 1.9319916698738798e-05, + "loss": 0.5919, + "step": 4584 + }, + { + "epoch": 0.14461441413026335, + "grad_norm": 5.564927101135254, + "learning_rate": 1.9319546356629432e-05, + "loss": 0.6012, + "step": 4585 + }, + { + "epoch": 0.14464595489670398, + "grad_norm": 3.645996570587158, + "learning_rate": 1.9319175917263354e-05, + "loss": 0.5799, + "step": 4586 + }, + { + "epoch": 0.1446774956631446, + "grad_norm": 3.0941107273101807, + "learning_rate": 1.931880538064445e-05, + "loss": 0.4079, + "step": 4587 + }, + { + "epoch": 0.14470903642958524, + "grad_norm": 5.280252933502197, + "learning_rate": 1.931843474677657e-05, + "loss": 0.5836, + "step": 4588 + }, + { + "epoch": 0.14474057719602587, + "grad_norm": 3.307636260986328, + "learning_rate": 1.9318064015663586e-05, + "loss": 0.5904, + "step": 4589 + }, + { + "epoch": 0.1447721179624665, + "grad_norm": 3.193345308303833, + "learning_rate": 1.9317693187309373e-05, + "loss": 0.6039, + "step": 4590 + }, + { + "epoch": 0.1448036587289071, + "grad_norm": 2.3427135944366455, + "learning_rate": 1.9317322261717794e-05, + "loss": 0.568, + "step": 4591 + }, + { + "epoch": 0.14483519949534773, + "grad_norm": 1.9920495748519897, + "learning_rate": 1.931695123889272e-05, + "loss": 0.5739, + "step": 4592 + }, + { + "epoch": 0.14486674026178836, + "grad_norm": 2.2162926197052, + "learning_rate": 1.9316580118838034e-05, + "loss": 0.9224, + "step": 4593 + }, + { + "epoch": 0.144898281028229, + "grad_norm": 5.11832332611084, + "learning_rate": 1.9316208901557595e-05, + "loss": 0.5692, + "step": 4594 + }, + { + "epoch": 0.14492982179466962, + "grad_norm": 4.063121795654297, + "learning_rate": 1.9315837587055283e-05, + "loss": 0.5669, + "step": 4595 + }, + { + "epoch": 0.14496136256111022, + "grad_norm": 4.309508323669434, + "learning_rate": 1.9315466175334968e-05, + "loss": 0.566, + "step": 4596 + }, + { + "epoch": 0.14499290332755085, + "grad_norm": 3.900391101837158, + "learning_rate": 1.9315094666400536e-05, + "loss": 0.5762, + "step": 4597 + }, + { + "epoch": 0.14502444409399148, + "grad_norm": 4.065573692321777, + "learning_rate": 1.9314723060255857e-05, + "loss": 0.5758, + "step": 4598 + }, + { + "epoch": 0.1450559848604321, + "grad_norm": 3.7648277282714844, + "learning_rate": 1.931435135690481e-05, + "loss": 0.6, + "step": 4599 + }, + { + "epoch": 0.14508752562687274, + "grad_norm": 2.6042850017547607, + "learning_rate": 1.931397955635127e-05, + "loss": 0.9094, + "step": 4600 + }, + { + "epoch": 0.14511906639331335, + "grad_norm": 3.109785795211792, + "learning_rate": 1.9313607658599128e-05, + "loss": 0.5765, + "step": 4601 + }, + { + "epoch": 0.14515060715975397, + "grad_norm": 2.7768678665161133, + "learning_rate": 1.931323566365225e-05, + "loss": 0.9117, + "step": 4602 + }, + { + "epoch": 0.1451821479261946, + "grad_norm": 3.5362401008605957, + "learning_rate": 1.9312863571514534e-05, + "loss": 0.5611, + "step": 4603 + }, + { + "epoch": 0.14521368869263523, + "grad_norm": 4.399355411529541, + "learning_rate": 1.9312491382189852e-05, + "loss": 0.5903, + "step": 4604 + }, + { + "epoch": 0.14524522945907586, + "grad_norm": 3.4052069187164307, + "learning_rate": 1.931211909568209e-05, + "loss": 0.5557, + "step": 4605 + }, + { + "epoch": 0.1452767702255165, + "grad_norm": 3.116990089416504, + "learning_rate": 1.931174671199514e-05, + "loss": 0.5873, + "step": 4606 + }, + { + "epoch": 0.1453083109919571, + "grad_norm": 3.503039598464966, + "learning_rate": 1.9311374231132877e-05, + "loss": 0.5856, + "step": 4607 + }, + { + "epoch": 0.14533985175839773, + "grad_norm": 3.1082663536071777, + "learning_rate": 1.9311001653099195e-05, + "loss": 0.5945, + "step": 4608 + }, + { + "epoch": 0.14537139252483836, + "grad_norm": 4.380859375, + "learning_rate": 1.9310628977897983e-05, + "loss": 0.5939, + "step": 4609 + }, + { + "epoch": 0.14540293329127899, + "grad_norm": 4.187594890594482, + "learning_rate": 1.9310256205533127e-05, + "loss": 0.6143, + "step": 4610 + }, + { + "epoch": 0.14543447405771961, + "grad_norm": 3.461625576019287, + "learning_rate": 1.930988333600852e-05, + "loss": 0.5968, + "step": 4611 + }, + { + "epoch": 0.14546601482416022, + "grad_norm": 2.7598986625671387, + "learning_rate": 1.930951036932805e-05, + "loss": 0.5811, + "step": 4612 + }, + { + "epoch": 0.14549755559060085, + "grad_norm": 3.550346612930298, + "learning_rate": 1.9309137305495608e-05, + "loss": 0.5742, + "step": 4613 + }, + { + "epoch": 0.14552909635704148, + "grad_norm": 3.5575382709503174, + "learning_rate": 1.9308764144515092e-05, + "loss": 0.4351, + "step": 4614 + }, + { + "epoch": 0.1455606371234821, + "grad_norm": 3.5310311317443848, + "learning_rate": 1.9308390886390396e-05, + "loss": 0.5834, + "step": 4615 + }, + { + "epoch": 0.14559217788992274, + "grad_norm": 3.0340356826782227, + "learning_rate": 1.930801753112541e-05, + "loss": 0.549, + "step": 4616 + }, + { + "epoch": 0.14562371865636334, + "grad_norm": 2.3488996028900146, + "learning_rate": 1.9307644078724037e-05, + "loss": 0.5593, + "step": 4617 + }, + { + "epoch": 0.14565525942280397, + "grad_norm": 3.0969154834747314, + "learning_rate": 1.9307270529190168e-05, + "loss": 0.6094, + "step": 4618 + }, + { + "epoch": 0.1456868001892446, + "grad_norm": 3.0962278842926025, + "learning_rate": 1.9306896882527706e-05, + "loss": 0.4126, + "step": 4619 + }, + { + "epoch": 0.14571834095568523, + "grad_norm": 5.975498676300049, + "learning_rate": 1.9306523138740546e-05, + "loss": 0.6007, + "step": 4620 + }, + { + "epoch": 0.14574988172212586, + "grad_norm": 2.795109987258911, + "learning_rate": 1.9306149297832592e-05, + "loss": 0.5897, + "step": 4621 + }, + { + "epoch": 0.14578142248856646, + "grad_norm": 3.996246099472046, + "learning_rate": 1.9305775359807747e-05, + "loss": 0.5672, + "step": 4622 + }, + { + "epoch": 0.1458129632550071, + "grad_norm": 2.303553819656372, + "learning_rate": 1.9305401324669906e-05, + "loss": 0.5745, + "step": 4623 + }, + { + "epoch": 0.14584450402144772, + "grad_norm": 4.017536163330078, + "learning_rate": 1.930502719242298e-05, + "loss": 0.4572, + "step": 4624 + }, + { + "epoch": 0.14587604478788835, + "grad_norm": 5.039464473724365, + "learning_rate": 1.9304652963070868e-05, + "loss": 0.5639, + "step": 4625 + }, + { + "epoch": 0.14590758555432898, + "grad_norm": 3.0558760166168213, + "learning_rate": 1.930427863661748e-05, + "loss": 0.5685, + "step": 4626 + }, + { + "epoch": 0.1459391263207696, + "grad_norm": 7.41541862487793, + "learning_rate": 1.9303904213066714e-05, + "loss": 0.5848, + "step": 4627 + }, + { + "epoch": 0.1459706670872102, + "grad_norm": 4.063368320465088, + "learning_rate": 1.930352969242249e-05, + "loss": 0.9514, + "step": 4628 + }, + { + "epoch": 0.14600220785365084, + "grad_norm": 2.289787530899048, + "learning_rate": 1.9303155074688704e-05, + "loss": 0.8471, + "step": 4629 + }, + { + "epoch": 0.14603374862009147, + "grad_norm": 3.7404074668884277, + "learning_rate": 1.9302780359869276e-05, + "loss": 0.607, + "step": 4630 + }, + { + "epoch": 0.1460652893865321, + "grad_norm": 3.0173563957214355, + "learning_rate": 1.9302405547968108e-05, + "loss": 0.5875, + "step": 4631 + }, + { + "epoch": 0.14609683015297273, + "grad_norm": 3.061530351638794, + "learning_rate": 1.9302030638989117e-05, + "loss": 0.5672, + "step": 4632 + }, + { + "epoch": 0.14612837091941333, + "grad_norm": 4.348293304443359, + "learning_rate": 1.9301655632936214e-05, + "loss": 0.5671, + "step": 4633 + }, + { + "epoch": 0.14615991168585396, + "grad_norm": 3.5569465160369873, + "learning_rate": 1.9301280529813308e-05, + "loss": 0.5852, + "step": 4634 + }, + { + "epoch": 0.1461914524522946, + "grad_norm": 3.6292107105255127, + "learning_rate": 1.930090532962432e-05, + "loss": 0.5866, + "step": 4635 + }, + { + "epoch": 0.14622299321873522, + "grad_norm": 3.5552978515625, + "learning_rate": 1.9300530032373164e-05, + "loss": 0.6443, + "step": 4636 + }, + { + "epoch": 0.14625453398517585, + "grad_norm": 2.8650708198547363, + "learning_rate": 1.9300154638063756e-05, + "loss": 0.5894, + "step": 4637 + }, + { + "epoch": 0.14628607475161645, + "grad_norm": 4.03895378112793, + "learning_rate": 1.929977914670001e-05, + "loss": 0.3844, + "step": 4638 + }, + { + "epoch": 0.14631761551805708, + "grad_norm": 7.269891262054443, + "learning_rate": 1.9299403558285848e-05, + "loss": 0.4164, + "step": 4639 + }, + { + "epoch": 0.1463491562844977, + "grad_norm": 3.7432808876037598, + "learning_rate": 1.9299027872825192e-05, + "loss": 0.5511, + "step": 4640 + }, + { + "epoch": 0.14638069705093834, + "grad_norm": 3.786872625350952, + "learning_rate": 1.9298652090321956e-05, + "loss": 0.5549, + "step": 4641 + }, + { + "epoch": 0.14641223781737897, + "grad_norm": 2.738598585128784, + "learning_rate": 1.929827621078007e-05, + "loss": 0.6155, + "step": 4642 + }, + { + "epoch": 0.1464437785838196, + "grad_norm": 2.8829312324523926, + "learning_rate": 1.9297900234203445e-05, + "loss": 0.5871, + "step": 4643 + }, + { + "epoch": 0.1464753193502602, + "grad_norm": 2.963005304336548, + "learning_rate": 1.929752416059602e-05, + "loss": 0.5487, + "step": 4644 + }, + { + "epoch": 0.14650686011670083, + "grad_norm": 3.557544469833374, + "learning_rate": 1.92971479899617e-05, + "loss": 0.6094, + "step": 4645 + }, + { + "epoch": 0.14653840088314146, + "grad_norm": 3.0918591022491455, + "learning_rate": 1.929677172230443e-05, + "loss": 0.5912, + "step": 4646 + }, + { + "epoch": 0.1465699416495821, + "grad_norm": 3.875828266143799, + "learning_rate": 1.9296395357628125e-05, + "loss": 0.5635, + "step": 4647 + }, + { + "epoch": 0.14660148241602272, + "grad_norm": 4.272403240203857, + "learning_rate": 1.929601889593672e-05, + "loss": 0.5576, + "step": 4648 + }, + { + "epoch": 0.14663302318246332, + "grad_norm": 2.5368075370788574, + "learning_rate": 1.9295642337234134e-05, + "loss": 0.5683, + "step": 4649 + }, + { + "epoch": 0.14666456394890395, + "grad_norm": 2.3071751594543457, + "learning_rate": 1.9295265681524308e-05, + "loss": 0.9372, + "step": 4650 + }, + { + "epoch": 0.14669610471534458, + "grad_norm": 2.806068181991577, + "learning_rate": 1.9294888928811164e-05, + "loss": 0.5526, + "step": 4651 + }, + { + "epoch": 0.1467276454817852, + "grad_norm": 2.8063721656799316, + "learning_rate": 1.9294512079098636e-05, + "loss": 0.6124, + "step": 4652 + }, + { + "epoch": 0.14675918624822584, + "grad_norm": 2.823410749435425, + "learning_rate": 1.9294135132390658e-05, + "loss": 0.5659, + "step": 4653 + }, + { + "epoch": 0.14679072701466644, + "grad_norm": 2.2720470428466797, + "learning_rate": 1.9293758088691163e-05, + "loss": 0.5577, + "step": 4654 + }, + { + "epoch": 0.14682226778110707, + "grad_norm": 3.875908136367798, + "learning_rate": 1.9293380948004085e-05, + "loss": 0.6145, + "step": 4655 + }, + { + "epoch": 0.1468538085475477, + "grad_norm": 3.255223274230957, + "learning_rate": 1.929300371033336e-05, + "loss": 0.4572, + "step": 4656 + }, + { + "epoch": 0.14688534931398833, + "grad_norm": 3.0212254524230957, + "learning_rate": 1.929262637568293e-05, + "loss": 0.5501, + "step": 4657 + }, + { + "epoch": 0.14691689008042896, + "grad_norm": 3.0984573364257812, + "learning_rate": 1.929224894405673e-05, + "loss": 0.5897, + "step": 4658 + }, + { + "epoch": 0.14694843084686957, + "grad_norm": 2.7030727863311768, + "learning_rate": 1.9291871415458688e-05, + "loss": 0.5855, + "step": 4659 + }, + { + "epoch": 0.1469799716133102, + "grad_norm": 3.0244197845458984, + "learning_rate": 1.929149378989276e-05, + "loss": 0.6394, + "step": 4660 + }, + { + "epoch": 0.14701151237975082, + "grad_norm": 3.461334228515625, + "learning_rate": 1.9291116067362876e-05, + "loss": 0.4126, + "step": 4661 + }, + { + "epoch": 0.14704305314619145, + "grad_norm": 2.527129650115967, + "learning_rate": 1.9290738247872983e-05, + "loss": 0.5482, + "step": 4662 + }, + { + "epoch": 0.14707459391263208, + "grad_norm": 4.573460578918457, + "learning_rate": 1.9290360331427023e-05, + "loss": 0.5785, + "step": 4663 + }, + { + "epoch": 0.14710613467907271, + "grad_norm": 2.920423746109009, + "learning_rate": 1.9289982318028936e-05, + "loss": 0.4079, + "step": 4664 + }, + { + "epoch": 0.14713767544551332, + "grad_norm": 2.5609045028686523, + "learning_rate": 1.9289604207682673e-05, + "loss": 0.598, + "step": 4665 + }, + { + "epoch": 0.14716921621195395, + "grad_norm": 3.461416244506836, + "learning_rate": 1.9289226000392176e-05, + "loss": 0.5671, + "step": 4666 + }, + { + "epoch": 0.14720075697839458, + "grad_norm": 2.774594783782959, + "learning_rate": 1.9288847696161396e-05, + "loss": 0.5827, + "step": 4667 + }, + { + "epoch": 0.1472322977448352, + "grad_norm": 3.4154152870178223, + "learning_rate": 1.9288469294994273e-05, + "loss": 0.6009, + "step": 4668 + }, + { + "epoch": 0.14726383851127584, + "grad_norm": 3.148378372192383, + "learning_rate": 1.928809079689476e-05, + "loss": 0.5859, + "step": 4669 + }, + { + "epoch": 0.14729537927771644, + "grad_norm": 2.7673561573028564, + "learning_rate": 1.9287712201866814e-05, + "loss": 0.6379, + "step": 4670 + }, + { + "epoch": 0.14732692004415707, + "grad_norm": 3.8814830780029297, + "learning_rate": 1.928733350991437e-05, + "loss": 0.5907, + "step": 4671 + }, + { + "epoch": 0.1473584608105977, + "grad_norm": 2.731327533721924, + "learning_rate": 1.92869547210414e-05, + "loss": 0.5623, + "step": 4672 + }, + { + "epoch": 0.14739000157703833, + "grad_norm": 3.223153829574585, + "learning_rate": 1.9286575835251838e-05, + "loss": 0.5813, + "step": 4673 + }, + { + "epoch": 0.14742154234347896, + "grad_norm": 4.774909496307373, + "learning_rate": 1.9286196852549647e-05, + "loss": 0.5573, + "step": 4674 + }, + { + "epoch": 0.14745308310991956, + "grad_norm": 3.600865125656128, + "learning_rate": 1.9285817772938785e-05, + "loss": 0.5677, + "step": 4675 + }, + { + "epoch": 0.1474846238763602, + "grad_norm": 3.7421281337738037, + "learning_rate": 1.9285438596423204e-05, + "loss": 0.5616, + "step": 4676 + }, + { + "epoch": 0.14751616464280082, + "grad_norm": 3.020419120788574, + "learning_rate": 1.928505932300686e-05, + "loss": 0.5837, + "step": 4677 + }, + { + "epoch": 0.14754770540924145, + "grad_norm": 3.7563886642456055, + "learning_rate": 1.9284679952693712e-05, + "loss": 0.5807, + "step": 4678 + }, + { + "epoch": 0.14757924617568208, + "grad_norm": 3.5854458808898926, + "learning_rate": 1.928430048548772e-05, + "loss": 0.5693, + "step": 4679 + }, + { + "epoch": 0.1476107869421227, + "grad_norm": 6.912154197692871, + "learning_rate": 1.9283920921392843e-05, + "loss": 0.6115, + "step": 4680 + }, + { + "epoch": 0.1476423277085633, + "grad_norm": 3.6270875930786133, + "learning_rate": 1.9283541260413044e-05, + "loss": 0.6037, + "step": 4681 + }, + { + "epoch": 0.14767386847500394, + "grad_norm": 3.3269028663635254, + "learning_rate": 1.928316150255228e-05, + "loss": 0.5694, + "step": 4682 + }, + { + "epoch": 0.14770540924144457, + "grad_norm": 11.333059310913086, + "learning_rate": 1.928278164781452e-05, + "loss": 0.5705, + "step": 4683 + }, + { + "epoch": 0.1477369500078852, + "grad_norm": 5.259132385253906, + "learning_rate": 1.9282401696203725e-05, + "loss": 0.621, + "step": 4684 + }, + { + "epoch": 0.14776849077432583, + "grad_norm": 4.869450092315674, + "learning_rate": 1.9282021647723862e-05, + "loss": 0.6261, + "step": 4685 + }, + { + "epoch": 0.14780003154076643, + "grad_norm": 4.914072036743164, + "learning_rate": 1.9281641502378894e-05, + "loss": 0.5889, + "step": 4686 + }, + { + "epoch": 0.14783157230720706, + "grad_norm": 3.4143753051757812, + "learning_rate": 1.928126126017279e-05, + "loss": 0.6002, + "step": 4687 + }, + { + "epoch": 0.1478631130736477, + "grad_norm": 3.352673053741455, + "learning_rate": 1.9280880921109518e-05, + "loss": 0.5995, + "step": 4688 + }, + { + "epoch": 0.14789465384008832, + "grad_norm": 5.043631076812744, + "learning_rate": 1.928050048519305e-05, + "loss": 0.6034, + "step": 4689 + }, + { + "epoch": 0.14792619460652895, + "grad_norm": 3.69852614402771, + "learning_rate": 1.928011995242735e-05, + "loss": 0.5852, + "step": 4690 + }, + { + "epoch": 0.14795773537296955, + "grad_norm": 3.2758822441101074, + "learning_rate": 1.9279739322816394e-05, + "loss": 0.566, + "step": 4691 + }, + { + "epoch": 0.14798927613941018, + "grad_norm": 5.190277099609375, + "learning_rate": 1.9279358596364154e-05, + "loss": 0.5614, + "step": 4692 + }, + { + "epoch": 0.1480208169058508, + "grad_norm": 3.515305995941162, + "learning_rate": 1.9278977773074597e-05, + "loss": 0.5772, + "step": 4693 + }, + { + "epoch": 0.14805235767229144, + "grad_norm": 3.472820997238159, + "learning_rate": 1.9278596852951706e-05, + "loss": 0.5759, + "step": 4694 + }, + { + "epoch": 0.14808389843873207, + "grad_norm": 3.615250825881958, + "learning_rate": 1.927821583599945e-05, + "loss": 0.5964, + "step": 4695 + }, + { + "epoch": 0.14811543920517267, + "grad_norm": 4.542019367218018, + "learning_rate": 1.927783472222181e-05, + "loss": 0.5664, + "step": 4696 + }, + { + "epoch": 0.1481469799716133, + "grad_norm": 3.0207765102386475, + "learning_rate": 1.9277453511622758e-05, + "loss": 0.5958, + "step": 4697 + }, + { + "epoch": 0.14817852073805393, + "grad_norm": 3.578406810760498, + "learning_rate": 1.9277072204206276e-05, + "loss": 0.9497, + "step": 4698 + }, + { + "epoch": 0.14821006150449456, + "grad_norm": 3.4104855060577393, + "learning_rate": 1.9276690799976344e-05, + "loss": 0.5872, + "step": 4699 + }, + { + "epoch": 0.1482416022709352, + "grad_norm": 7.265411853790283, + "learning_rate": 1.9276309298936936e-05, + "loss": 0.5345, + "step": 4700 + }, + { + "epoch": 0.14827314303737582, + "grad_norm": 3.2427525520324707, + "learning_rate": 1.927592770109204e-05, + "loss": 0.5922, + "step": 4701 + }, + { + "epoch": 0.14830468380381642, + "grad_norm": 4.10505485534668, + "learning_rate": 1.9275546006445637e-05, + "loss": 0.5661, + "step": 4702 + }, + { + "epoch": 0.14833622457025705, + "grad_norm": 3.047884941101074, + "learning_rate": 1.927516421500171e-05, + "loss": 0.5787, + "step": 4703 + }, + { + "epoch": 0.14836776533669768, + "grad_norm": 3.5128653049468994, + "learning_rate": 1.9274782326764238e-05, + "loss": 0.418, + "step": 4704 + }, + { + "epoch": 0.1483993061031383, + "grad_norm": 4.081721305847168, + "learning_rate": 1.9274400341737214e-05, + "loss": 0.5808, + "step": 4705 + }, + { + "epoch": 0.14843084686957894, + "grad_norm": 7.182386875152588, + "learning_rate": 1.9274018259924618e-05, + "loss": 0.5639, + "step": 4706 + }, + { + "epoch": 0.14846238763601954, + "grad_norm": 3.386509895324707, + "learning_rate": 1.9273636081330443e-05, + "loss": 0.5482, + "step": 4707 + }, + { + "epoch": 0.14849392840246017, + "grad_norm": 2.503131628036499, + "learning_rate": 1.9273253805958675e-05, + "loss": 0.5927, + "step": 4708 + }, + { + "epoch": 0.1485254691689008, + "grad_norm": 4.387040138244629, + "learning_rate": 1.92728714338133e-05, + "loss": 0.5551, + "step": 4709 + }, + { + "epoch": 0.14855700993534143, + "grad_norm": 3.3665614128112793, + "learning_rate": 1.9272488964898316e-05, + "loss": 0.6024, + "step": 4710 + }, + { + "epoch": 0.14858855070178206, + "grad_norm": 6.920803070068359, + "learning_rate": 1.927210639921771e-05, + "loss": 0.588, + "step": 4711 + }, + { + "epoch": 0.14862009146822266, + "grad_norm": 3.3206677436828613, + "learning_rate": 1.9271723736775468e-05, + "loss": 0.5675, + "step": 4712 + }, + { + "epoch": 0.1486516322346633, + "grad_norm": 4.883582592010498, + "learning_rate": 1.9271340977575597e-05, + "loss": 0.6111, + "step": 4713 + }, + { + "epoch": 0.14868317300110392, + "grad_norm": 4.303486347198486, + "learning_rate": 1.927095812162208e-05, + "loss": 0.5672, + "step": 4714 + }, + { + "epoch": 0.14871471376754455, + "grad_norm": 2.4242091178894043, + "learning_rate": 1.9270575168918915e-05, + "loss": 0.5665, + "step": 4715 + }, + { + "epoch": 0.14874625453398518, + "grad_norm": 2.9389848709106445, + "learning_rate": 1.9270192119470103e-05, + "loss": 0.5661, + "step": 4716 + }, + { + "epoch": 0.1487777953004258, + "grad_norm": 2.4976611137390137, + "learning_rate": 1.9269808973279635e-05, + "loss": 0.9086, + "step": 4717 + }, + { + "epoch": 0.14880933606686642, + "grad_norm": 2.5311155319213867, + "learning_rate": 1.9269425730351513e-05, + "loss": 0.5902, + "step": 4718 + }, + { + "epoch": 0.14884087683330705, + "grad_norm": 2.794510841369629, + "learning_rate": 1.9269042390689735e-05, + "loss": 0.557, + "step": 4719 + }, + { + "epoch": 0.14887241759974768, + "grad_norm": 13.885738372802734, + "learning_rate": 1.926865895429831e-05, + "loss": 0.5402, + "step": 4720 + }, + { + "epoch": 0.1489039583661883, + "grad_norm": 2.982762336730957, + "learning_rate": 1.9268275421181222e-05, + "loss": 0.5796, + "step": 4721 + }, + { + "epoch": 0.14893549913262893, + "grad_norm": 4.791544437408447, + "learning_rate": 1.9267891791342487e-05, + "loss": 0.4068, + "step": 4722 + }, + { + "epoch": 0.14896703989906954, + "grad_norm": 3.0559682846069336, + "learning_rate": 1.9267508064786104e-05, + "loss": 0.562, + "step": 4723 + }, + { + "epoch": 0.14899858066551017, + "grad_norm": 2.4574530124664307, + "learning_rate": 1.926712424151608e-05, + "loss": 0.5932, + "step": 4724 + }, + { + "epoch": 0.1490301214319508, + "grad_norm": 2.739664077758789, + "learning_rate": 1.926674032153642e-05, + "loss": 0.561, + "step": 4725 + }, + { + "epoch": 0.14906166219839143, + "grad_norm": 2.8262205123901367, + "learning_rate": 1.9266356304851127e-05, + "loss": 0.5593, + "step": 4726 + }, + { + "epoch": 0.14909320296483206, + "grad_norm": 3.703155040740967, + "learning_rate": 1.9265972191464213e-05, + "loss": 0.5896, + "step": 4727 + }, + { + "epoch": 0.14912474373127266, + "grad_norm": 2.768228530883789, + "learning_rate": 1.926558798137968e-05, + "loss": 0.5583, + "step": 4728 + }, + { + "epoch": 0.1491562844977133, + "grad_norm": 4.166100978851318, + "learning_rate": 1.9265203674601544e-05, + "loss": 0.8945, + "step": 4729 + }, + { + "epoch": 0.14918782526415392, + "grad_norm": 2.8311002254486084, + "learning_rate": 1.9264819271133818e-05, + "loss": 0.5319, + "step": 4730 + }, + { + "epoch": 0.14921936603059455, + "grad_norm": 3.8657617568969727, + "learning_rate": 1.9264434770980503e-05, + "loss": 0.6106, + "step": 4731 + }, + { + "epoch": 0.14925090679703518, + "grad_norm": 6.143987655639648, + "learning_rate": 1.9264050174145617e-05, + "loss": 0.5853, + "step": 4732 + }, + { + "epoch": 0.14928244756347578, + "grad_norm": 3.278864622116089, + "learning_rate": 1.9263665480633177e-05, + "loss": 0.6174, + "step": 4733 + }, + { + "epoch": 0.1493139883299164, + "grad_norm": 3.380457639694214, + "learning_rate": 1.92632806904472e-05, + "loss": 0.5794, + "step": 4734 + }, + { + "epoch": 0.14934552909635704, + "grad_norm": 3.8538050651550293, + "learning_rate": 1.9262895803591685e-05, + "loss": 0.8566, + "step": 4735 + }, + { + "epoch": 0.14937706986279767, + "grad_norm": 4.957706451416016, + "learning_rate": 1.9262510820070667e-05, + "loss": 0.5995, + "step": 4736 + }, + { + "epoch": 0.1494086106292383, + "grad_norm": 2.691786766052246, + "learning_rate": 1.9262125739888153e-05, + "loss": 0.5744, + "step": 4737 + }, + { + "epoch": 0.14944015139567893, + "grad_norm": 4.219600677490234, + "learning_rate": 1.926174056304817e-05, + "loss": 0.586, + "step": 4738 + }, + { + "epoch": 0.14947169216211953, + "grad_norm": 3.349841594696045, + "learning_rate": 1.926135528955473e-05, + "loss": 0.5944, + "step": 4739 + }, + { + "epoch": 0.14950323292856016, + "grad_norm": 3.3121230602264404, + "learning_rate": 1.9260969919411854e-05, + "loss": 0.5854, + "step": 4740 + }, + { + "epoch": 0.1495347736950008, + "grad_norm": 2.8161966800689697, + "learning_rate": 1.9260584452623568e-05, + "loss": 0.5883, + "step": 4741 + }, + { + "epoch": 0.14956631446144142, + "grad_norm": 2.1796343326568604, + "learning_rate": 1.926019888919389e-05, + "loss": 0.9039, + "step": 4742 + }, + { + "epoch": 0.14959785522788205, + "grad_norm": 4.4655022621154785, + "learning_rate": 1.9259813229126848e-05, + "loss": 0.5835, + "step": 4743 + }, + { + "epoch": 0.14962939599432265, + "grad_norm": 3.6946020126342773, + "learning_rate": 1.9259427472426468e-05, + "loss": 0.5986, + "step": 4744 + }, + { + "epoch": 0.14966093676076328, + "grad_norm": 3.6314697265625, + "learning_rate": 1.925904161909677e-05, + "loss": 0.6238, + "step": 4745 + }, + { + "epoch": 0.1496924775272039, + "grad_norm": 2.230069875717163, + "learning_rate": 1.9258655669141784e-05, + "loss": 0.5829, + "step": 4746 + }, + { + "epoch": 0.14972401829364454, + "grad_norm": 4.341128826141357, + "learning_rate": 1.9258269622565537e-05, + "loss": 0.3936, + "step": 4747 + }, + { + "epoch": 0.14975555906008517, + "grad_norm": 3.729440450668335, + "learning_rate": 1.9257883479372056e-05, + "loss": 0.5723, + "step": 4748 + }, + { + "epoch": 0.14978709982652577, + "grad_norm": 2.475118398666382, + "learning_rate": 1.925749723956537e-05, + "loss": 0.4182, + "step": 4749 + }, + { + "epoch": 0.1498186405929664, + "grad_norm": 2.559643268585205, + "learning_rate": 1.925711090314952e-05, + "loss": 0.5996, + "step": 4750 + }, + { + "epoch": 0.14985018135940703, + "grad_norm": 8.985841751098633, + "learning_rate": 1.9256724470128524e-05, + "loss": 0.6009, + "step": 4751 + }, + { + "epoch": 0.14988172212584766, + "grad_norm": 2.7321252822875977, + "learning_rate": 1.925633794050642e-05, + "loss": 0.5907, + "step": 4752 + }, + { + "epoch": 0.1499132628922883, + "grad_norm": 2.341341733932495, + "learning_rate": 1.9255951314287244e-05, + "loss": 0.558, + "step": 4753 + }, + { + "epoch": 0.14994480365872892, + "grad_norm": 2.463015079498291, + "learning_rate": 1.9255564591475027e-05, + "loss": 0.5865, + "step": 4754 + }, + { + "epoch": 0.14997634442516952, + "grad_norm": 2.067533016204834, + "learning_rate": 1.925517777207381e-05, + "loss": 0.5903, + "step": 4755 + }, + { + "epoch": 0.15000788519161015, + "grad_norm": 2.7010178565979004, + "learning_rate": 1.9254790856087624e-05, + "loss": 0.5873, + "step": 4756 + }, + { + "epoch": 0.15003942595805078, + "grad_norm": 27.03782844543457, + "learning_rate": 1.925440384352051e-05, + "loss": 0.6137, + "step": 4757 + }, + { + "epoch": 0.1500709667244914, + "grad_norm": 2.5766592025756836, + "learning_rate": 1.9254016734376508e-05, + "loss": 0.5529, + "step": 4758 + }, + { + "epoch": 0.15010250749093204, + "grad_norm": 2.9068984985351562, + "learning_rate": 1.9253629528659653e-05, + "loss": 0.5842, + "step": 4759 + }, + { + "epoch": 0.15013404825737264, + "grad_norm": 3.279568672180176, + "learning_rate": 1.9253242226373986e-05, + "loss": 0.611, + "step": 4760 + }, + { + "epoch": 0.15016558902381327, + "grad_norm": 2.9895732402801514, + "learning_rate": 1.9252854827523557e-05, + "loss": 0.5711, + "step": 4761 + }, + { + "epoch": 0.1501971297902539, + "grad_norm": 3.0179872512817383, + "learning_rate": 1.9252467332112405e-05, + "loss": 0.5977, + "step": 4762 + }, + { + "epoch": 0.15022867055669453, + "grad_norm": 2.7726998329162598, + "learning_rate": 1.9252079740144566e-05, + "loss": 0.5647, + "step": 4763 + }, + { + "epoch": 0.15026021132313516, + "grad_norm": 3.7241883277893066, + "learning_rate": 1.9251692051624092e-05, + "loss": 0.5802, + "step": 4764 + }, + { + "epoch": 0.15029175208957576, + "grad_norm": 2.9103970527648926, + "learning_rate": 1.925130426655503e-05, + "loss": 0.5752, + "step": 4765 + }, + { + "epoch": 0.1503232928560164, + "grad_norm": 3.447075605392456, + "learning_rate": 1.9250916384941424e-05, + "loss": 0.5792, + "step": 4766 + }, + { + "epoch": 0.15035483362245702, + "grad_norm": 2.6498799324035645, + "learning_rate": 1.925052840678732e-05, + "loss": 0.5802, + "step": 4767 + }, + { + "epoch": 0.15038637438889765, + "grad_norm": 2.480290174484253, + "learning_rate": 1.925014033209677e-05, + "loss": 0.611, + "step": 4768 + }, + { + "epoch": 0.15041791515533828, + "grad_norm": 3.163900136947632, + "learning_rate": 1.9249752160873825e-05, + "loss": 0.5948, + "step": 4769 + }, + { + "epoch": 0.1504494559217789, + "grad_norm": 4.304858207702637, + "learning_rate": 1.9249363893122532e-05, + "loss": 0.6206, + "step": 4770 + }, + { + "epoch": 0.15048099668821951, + "grad_norm": 5.054910182952881, + "learning_rate": 1.9248975528846948e-05, + "loss": 0.6073, + "step": 4771 + }, + { + "epoch": 0.15051253745466014, + "grad_norm": 6.732681751251221, + "learning_rate": 1.924858706805112e-05, + "loss": 0.5977, + "step": 4772 + }, + { + "epoch": 0.15054407822110077, + "grad_norm": 4.348542213439941, + "learning_rate": 1.9248198510739103e-05, + "loss": 0.576, + "step": 4773 + }, + { + "epoch": 0.1505756189875414, + "grad_norm": 4.109795570373535, + "learning_rate": 1.9247809856914956e-05, + "loss": 0.5791, + "step": 4774 + }, + { + "epoch": 0.15060715975398203, + "grad_norm": 3.3753037452697754, + "learning_rate": 1.924742110658273e-05, + "loss": 0.543, + "step": 4775 + }, + { + "epoch": 0.15063870052042264, + "grad_norm": 3.4075400829315186, + "learning_rate": 1.924703225974649e-05, + "loss": 0.5686, + "step": 4776 + }, + { + "epoch": 0.15067024128686327, + "grad_norm": 3.247370958328247, + "learning_rate": 1.9246643316410286e-05, + "loss": 0.5906, + "step": 4777 + }, + { + "epoch": 0.1507017820533039, + "grad_norm": 2.7910313606262207, + "learning_rate": 1.9246254276578175e-05, + "loss": 0.6059, + "step": 4778 + }, + { + "epoch": 0.15073332281974453, + "grad_norm": 3.585728645324707, + "learning_rate": 1.9245865140254225e-05, + "loss": 0.5819, + "step": 4779 + }, + { + "epoch": 0.15076486358618515, + "grad_norm": 3.0546066761016846, + "learning_rate": 1.9245475907442492e-05, + "loss": 0.6268, + "step": 4780 + }, + { + "epoch": 0.15079640435262576, + "grad_norm": 2.7681522369384766, + "learning_rate": 1.9245086578147036e-05, + "loss": 0.5475, + "step": 4781 + }, + { + "epoch": 0.1508279451190664, + "grad_norm": 2.9861843585968018, + "learning_rate": 1.9244697152371928e-05, + "loss": 0.5705, + "step": 4782 + }, + { + "epoch": 0.15085948588550702, + "grad_norm": 3.7630112171173096, + "learning_rate": 1.9244307630121224e-05, + "loss": 0.5948, + "step": 4783 + }, + { + "epoch": 0.15089102665194765, + "grad_norm": 3.1148531436920166, + "learning_rate": 1.9243918011398993e-05, + "loss": 0.5606, + "step": 4784 + }, + { + "epoch": 0.15092256741838828, + "grad_norm": 11.261624336242676, + "learning_rate": 1.9243528296209298e-05, + "loss": 0.5694, + "step": 4785 + }, + { + "epoch": 0.15095410818482888, + "grad_norm": 3.091651678085327, + "learning_rate": 1.9243138484556208e-05, + "loss": 0.5713, + "step": 4786 + }, + { + "epoch": 0.1509856489512695, + "grad_norm": 6.470444679260254, + "learning_rate": 1.9242748576443792e-05, + "loss": 0.5812, + "step": 4787 + }, + { + "epoch": 0.15101718971771014, + "grad_norm": 3.5036277770996094, + "learning_rate": 1.924235857187612e-05, + "loss": 0.5653, + "step": 4788 + }, + { + "epoch": 0.15104873048415077, + "grad_norm": 9.744293212890625, + "learning_rate": 1.9241968470857256e-05, + "loss": 0.597, + "step": 4789 + }, + { + "epoch": 0.1510802712505914, + "grad_norm": 4.678783416748047, + "learning_rate": 1.9241578273391277e-05, + "loss": 0.412, + "step": 4790 + }, + { + "epoch": 0.15111181201703203, + "grad_norm": 3.852717399597168, + "learning_rate": 1.924118797948225e-05, + "loss": 0.9758, + "step": 4791 + }, + { + "epoch": 0.15114335278347263, + "grad_norm": 3.33343505859375, + "learning_rate": 1.9240797589134253e-05, + "loss": 0.5914, + "step": 4792 + }, + { + "epoch": 0.15117489354991326, + "grad_norm": 3.1913533210754395, + "learning_rate": 1.924040710235136e-05, + "loss": 0.5883, + "step": 4793 + }, + { + "epoch": 0.1512064343163539, + "grad_norm": 4.235106468200684, + "learning_rate": 1.924001651913764e-05, + "loss": 0.5782, + "step": 4794 + }, + { + "epoch": 0.15123797508279452, + "grad_norm": 3.305588722229004, + "learning_rate": 1.9239625839497176e-05, + "loss": 0.595, + "step": 4795 + }, + { + "epoch": 0.15126951584923515, + "grad_norm": 3.550518274307251, + "learning_rate": 1.923923506343404e-05, + "loss": 0.5711, + "step": 4796 + }, + { + "epoch": 0.15130105661567575, + "grad_norm": 3.253761053085327, + "learning_rate": 1.9238844190952312e-05, + "loss": 0.6125, + "step": 4797 + }, + { + "epoch": 0.15133259738211638, + "grad_norm": 2.9558064937591553, + "learning_rate": 1.9238453222056074e-05, + "loss": 0.56, + "step": 4798 + }, + { + "epoch": 0.151364138148557, + "grad_norm": 7.14475679397583, + "learning_rate": 1.92380621567494e-05, + "loss": 0.8953, + "step": 4799 + }, + { + "epoch": 0.15139567891499764, + "grad_norm": 5.252791881561279, + "learning_rate": 1.9237670995036378e-05, + "loss": 0.4313, + "step": 4800 + }, + { + "epoch": 0.15142721968143827, + "grad_norm": 3.5677490234375, + "learning_rate": 1.9237279736921082e-05, + "loss": 0.5736, + "step": 4801 + }, + { + "epoch": 0.15145876044787887, + "grad_norm": 3.756397247314453, + "learning_rate": 1.92368883824076e-05, + "loss": 0.5311, + "step": 4802 + }, + { + "epoch": 0.1514903012143195, + "grad_norm": 3.059981346130371, + "learning_rate": 1.923649693150002e-05, + "loss": 0.5544, + "step": 4803 + }, + { + "epoch": 0.15152184198076013, + "grad_norm": 6.237941741943359, + "learning_rate": 1.9236105384202416e-05, + "loss": 0.5859, + "step": 4804 + }, + { + "epoch": 0.15155338274720076, + "grad_norm": 2.941239595413208, + "learning_rate": 1.923571374051888e-05, + "loss": 0.5531, + "step": 4805 + }, + { + "epoch": 0.1515849235136414, + "grad_norm": 4.462599277496338, + "learning_rate": 1.9235322000453507e-05, + "loss": 0.57, + "step": 4806 + }, + { + "epoch": 0.15161646428008202, + "grad_norm": 3.8347418308258057, + "learning_rate": 1.9234930164010373e-05, + "loss": 0.5745, + "step": 4807 + }, + { + "epoch": 0.15164800504652262, + "grad_norm": 5.452023029327393, + "learning_rate": 1.9234538231193575e-05, + "loss": 0.6245, + "step": 4808 + }, + { + "epoch": 0.15167954581296325, + "grad_norm": 2.863508939743042, + "learning_rate": 1.92341462020072e-05, + "loss": 0.899, + "step": 4809 + }, + { + "epoch": 0.15171108657940388, + "grad_norm": 3.89034104347229, + "learning_rate": 1.9233754076455335e-05, + "loss": 0.5694, + "step": 4810 + }, + { + "epoch": 0.1517426273458445, + "grad_norm": 17.35304069519043, + "learning_rate": 1.923336185454208e-05, + "loss": 0.6026, + "step": 4811 + }, + { + "epoch": 0.15177416811228514, + "grad_norm": 2.955266237258911, + "learning_rate": 1.9232969536271523e-05, + "loss": 0.6121, + "step": 4812 + }, + { + "epoch": 0.15180570887872574, + "grad_norm": 4.318248748779297, + "learning_rate": 1.9232577121647758e-05, + "loss": 0.5901, + "step": 4813 + }, + { + "epoch": 0.15183724964516637, + "grad_norm": 3.1727256774902344, + "learning_rate": 1.9232184610674886e-05, + "loss": 0.5666, + "step": 4814 + }, + { + "epoch": 0.151868790411607, + "grad_norm": 3.2648379802703857, + "learning_rate": 1.9231792003356994e-05, + "loss": 0.5704, + "step": 4815 + }, + { + "epoch": 0.15190033117804763, + "grad_norm": 2.5742673873901367, + "learning_rate": 1.923139929969819e-05, + "loss": 0.5912, + "step": 4816 + }, + { + "epoch": 0.15193187194448826, + "grad_norm": 25.67828369140625, + "learning_rate": 1.9231006499702563e-05, + "loss": 0.412, + "step": 4817 + }, + { + "epoch": 0.15196341271092886, + "grad_norm": 2.943013906478882, + "learning_rate": 1.9230613603374214e-05, + "loss": 0.5819, + "step": 4818 + }, + { + "epoch": 0.1519949534773695, + "grad_norm": 3.484879732131958, + "learning_rate": 1.923022061071725e-05, + "loss": 0.5749, + "step": 4819 + }, + { + "epoch": 0.15202649424381012, + "grad_norm": 3.1050446033477783, + "learning_rate": 1.9229827521735762e-05, + "loss": 0.5963, + "step": 4820 + }, + { + "epoch": 0.15205803501025075, + "grad_norm": 1.8894978761672974, + "learning_rate": 1.922943433643386e-05, + "loss": 0.9117, + "step": 4821 + }, + { + "epoch": 0.15208957577669138, + "grad_norm": 3.8328726291656494, + "learning_rate": 1.922904105481564e-05, + "loss": 0.5693, + "step": 4822 + }, + { + "epoch": 0.15212111654313198, + "grad_norm": 2.7415788173675537, + "learning_rate": 1.9228647676885214e-05, + "loss": 0.5879, + "step": 4823 + }, + { + "epoch": 0.15215265730957261, + "grad_norm": 3.191957950592041, + "learning_rate": 1.922825420264668e-05, + "loss": 0.5893, + "step": 4824 + }, + { + "epoch": 0.15218419807601324, + "grad_norm": 2.4650402069091797, + "learning_rate": 1.9227860632104154e-05, + "loss": 0.4051, + "step": 4825 + }, + { + "epoch": 0.15221573884245387, + "grad_norm": 4.397758483886719, + "learning_rate": 1.9227466965261734e-05, + "loss": 0.5968, + "step": 4826 + }, + { + "epoch": 0.1522472796088945, + "grad_norm": 2.8918628692626953, + "learning_rate": 1.922707320212353e-05, + "loss": 0.5783, + "step": 4827 + }, + { + "epoch": 0.15227882037533513, + "grad_norm": 1.8203362226486206, + "learning_rate": 1.9226679342693653e-05, + "loss": 0.9046, + "step": 4828 + }, + { + "epoch": 0.15231036114177574, + "grad_norm": 3.1137893199920654, + "learning_rate": 1.9226285386976212e-05, + "loss": 0.6079, + "step": 4829 + }, + { + "epoch": 0.15234190190821637, + "grad_norm": 3.0267961025238037, + "learning_rate": 1.922589133497532e-05, + "loss": 0.5838, + "step": 4830 + }, + { + "epoch": 0.152373442674657, + "grad_norm": 3.2229223251342773, + "learning_rate": 1.922549718669509e-05, + "loss": 0.4436, + "step": 4831 + }, + { + "epoch": 0.15240498344109762, + "grad_norm": 3.6994898319244385, + "learning_rate": 1.922510294213963e-05, + "loss": 0.5566, + "step": 4832 + }, + { + "epoch": 0.15243652420753825, + "grad_norm": 2.6233816146850586, + "learning_rate": 1.922470860131306e-05, + "loss": 0.6036, + "step": 4833 + }, + { + "epoch": 0.15246806497397886, + "grad_norm": 3.1367995738983154, + "learning_rate": 1.922431416421949e-05, + "loss": 0.5855, + "step": 4834 + }, + { + "epoch": 0.15249960574041949, + "grad_norm": 3.1680610179901123, + "learning_rate": 1.9223919630863044e-05, + "loss": 0.5544, + "step": 4835 + }, + { + "epoch": 0.15253114650686012, + "grad_norm": 4.5799078941345215, + "learning_rate": 1.9223525001247834e-05, + "loss": 0.5898, + "step": 4836 + }, + { + "epoch": 0.15256268727330075, + "grad_norm": 6.065136909484863, + "learning_rate": 1.9223130275377975e-05, + "loss": 0.5441, + "step": 4837 + }, + { + "epoch": 0.15259422803974138, + "grad_norm": 2.3213305473327637, + "learning_rate": 1.9222735453257593e-05, + "loss": 0.5792, + "step": 4838 + }, + { + "epoch": 0.15262576880618198, + "grad_norm": 4.847630977630615, + "learning_rate": 1.9222340534890803e-05, + "loss": 0.5939, + "step": 4839 + }, + { + "epoch": 0.1526573095726226, + "grad_norm": 5.424413204193115, + "learning_rate": 1.922194552028173e-05, + "loss": 0.5902, + "step": 4840 + }, + { + "epoch": 0.15268885033906324, + "grad_norm": 2.861696720123291, + "learning_rate": 1.9221550409434496e-05, + "loss": 0.5941, + "step": 4841 + }, + { + "epoch": 0.15272039110550387, + "grad_norm": 3.2108707427978516, + "learning_rate": 1.922115520235322e-05, + "loss": 0.4434, + "step": 4842 + }, + { + "epoch": 0.1527519318719445, + "grad_norm": 2.216343402862549, + "learning_rate": 1.9220759899042032e-05, + "loss": 0.6041, + "step": 4843 + }, + { + "epoch": 0.15278347263838513, + "grad_norm": 2.2216060161590576, + "learning_rate": 1.9220364499505054e-05, + "loss": 0.5905, + "step": 4844 + }, + { + "epoch": 0.15281501340482573, + "grad_norm": 2.53216814994812, + "learning_rate": 1.9219969003746414e-05, + "loss": 0.5946, + "step": 4845 + }, + { + "epoch": 0.15284655417126636, + "grad_norm": 2.5876455307006836, + "learning_rate": 1.9219573411770235e-05, + "loss": 0.5636, + "step": 4846 + }, + { + "epoch": 0.152878094937707, + "grad_norm": 3.4579715728759766, + "learning_rate": 1.921917772358065e-05, + "loss": 0.6368, + "step": 4847 + }, + { + "epoch": 0.15290963570414762, + "grad_norm": 3.321911334991455, + "learning_rate": 1.9218781939181788e-05, + "loss": 0.6108, + "step": 4848 + }, + { + "epoch": 0.15294117647058825, + "grad_norm": 4.35638952255249, + "learning_rate": 1.9218386058577777e-05, + "loss": 0.601, + "step": 4849 + }, + { + "epoch": 0.15297271723702885, + "grad_norm": 2.5618746280670166, + "learning_rate": 1.921799008177275e-05, + "loss": 0.5929, + "step": 4850 + }, + { + "epoch": 0.15300425800346948, + "grad_norm": 3.231499433517456, + "learning_rate": 1.921759400877084e-05, + "loss": 0.5781, + "step": 4851 + }, + { + "epoch": 0.1530357987699101, + "grad_norm": 2.712585210800171, + "learning_rate": 1.921719783957618e-05, + "loss": 0.5668, + "step": 4852 + }, + { + "epoch": 0.15306733953635074, + "grad_norm": 2.978250741958618, + "learning_rate": 1.92168015741929e-05, + "loss": 0.598, + "step": 4853 + }, + { + "epoch": 0.15309888030279137, + "grad_norm": 2.6029889583587646, + "learning_rate": 1.9216405212625143e-05, + "loss": 0.4226, + "step": 4854 + }, + { + "epoch": 0.15313042106923197, + "grad_norm": 3.9152984619140625, + "learning_rate": 1.921600875487704e-05, + "loss": 0.5996, + "step": 4855 + }, + { + "epoch": 0.1531619618356726, + "grad_norm": 2.2888662815093994, + "learning_rate": 1.9215612200952728e-05, + "loss": 0.5768, + "step": 4856 + }, + { + "epoch": 0.15319350260211323, + "grad_norm": 2.8317089080810547, + "learning_rate": 1.9215215550856347e-05, + "loss": 0.5802, + "step": 4857 + }, + { + "epoch": 0.15322504336855386, + "grad_norm": 2.161702871322632, + "learning_rate": 1.921481880459204e-05, + "loss": 0.5709, + "step": 4858 + }, + { + "epoch": 0.1532565841349945, + "grad_norm": 3.0369873046875, + "learning_rate": 1.921442196216394e-05, + "loss": 0.587, + "step": 4859 + }, + { + "epoch": 0.1532881249014351, + "grad_norm": 2.7860491275787354, + "learning_rate": 1.9214025023576194e-05, + "loss": 0.5804, + "step": 4860 + }, + { + "epoch": 0.15331966566787572, + "grad_norm": 1.922361135482788, + "learning_rate": 1.9213627988832943e-05, + "loss": 0.98, + "step": 4861 + }, + { + "epoch": 0.15335120643431635, + "grad_norm": 1.8855723142623901, + "learning_rate": 1.9213230857938327e-05, + "loss": 0.4058, + "step": 4862 + }, + { + "epoch": 0.15338274720075698, + "grad_norm": 2.340980291366577, + "learning_rate": 1.921283363089649e-05, + "loss": 0.9529, + "step": 4863 + }, + { + "epoch": 0.1534142879671976, + "grad_norm": 2.5605356693267822, + "learning_rate": 1.9212436307711587e-05, + "loss": 0.5996, + "step": 4864 + }, + { + "epoch": 0.15344582873363824, + "grad_norm": 1.6584280729293823, + "learning_rate": 1.9212038888387754e-05, + "loss": 0.9124, + "step": 4865 + }, + { + "epoch": 0.15347736950007884, + "grad_norm": 2.6064579486846924, + "learning_rate": 1.9211641372929145e-05, + "loss": 0.6058, + "step": 4866 + }, + { + "epoch": 0.15350891026651947, + "grad_norm": 3.5449442863464355, + "learning_rate": 1.9211243761339902e-05, + "loss": 0.5556, + "step": 4867 + }, + { + "epoch": 0.1535404510329601, + "grad_norm": 3.9365344047546387, + "learning_rate": 1.921084605362418e-05, + "loss": 0.5939, + "step": 4868 + }, + { + "epoch": 0.15357199179940073, + "grad_norm": 6.400400638580322, + "learning_rate": 1.921044824978613e-05, + "loss": 0.4185, + "step": 4869 + }, + { + "epoch": 0.15360353256584136, + "grad_norm": 2.217824935913086, + "learning_rate": 1.9210050349829897e-05, + "loss": 0.9308, + "step": 4870 + }, + { + "epoch": 0.15363507333228196, + "grad_norm": 4.213483810424805, + "learning_rate": 1.920965235375964e-05, + "loss": 0.4314, + "step": 4871 + }, + { + "epoch": 0.1536666140987226, + "grad_norm": 1.9698834419250488, + "learning_rate": 1.920925426157951e-05, + "loss": 0.9198, + "step": 4872 + }, + { + "epoch": 0.15369815486516322, + "grad_norm": 2.75481915473938, + "learning_rate": 1.9208856073293657e-05, + "loss": 0.562, + "step": 4873 + }, + { + "epoch": 0.15372969563160385, + "grad_norm": 2.247653007507324, + "learning_rate": 1.9208457788906243e-05, + "loss": 0.4126, + "step": 4874 + }, + { + "epoch": 0.15376123639804448, + "grad_norm": 3.8915748596191406, + "learning_rate": 1.9208059408421423e-05, + "loss": 0.6219, + "step": 4875 + }, + { + "epoch": 0.15379277716448508, + "grad_norm": 3.3410086631774902, + "learning_rate": 1.920766093184335e-05, + "loss": 0.6006, + "step": 4876 + }, + { + "epoch": 0.1538243179309257, + "grad_norm": 3.1935348510742188, + "learning_rate": 1.9207262359176187e-05, + "loss": 0.5943, + "step": 4877 + }, + { + "epoch": 0.15385585869736634, + "grad_norm": 3.3296313285827637, + "learning_rate": 1.920686369042409e-05, + "loss": 0.6028, + "step": 4878 + }, + { + "epoch": 0.15388739946380697, + "grad_norm": 3.431225299835205, + "learning_rate": 1.9206464925591226e-05, + "loss": 0.5909, + "step": 4879 + }, + { + "epoch": 0.1539189402302476, + "grad_norm": 7.653417587280273, + "learning_rate": 1.920606606468175e-05, + "loss": 0.397, + "step": 4880 + }, + { + "epoch": 0.15395048099668823, + "grad_norm": 3.159905195236206, + "learning_rate": 1.9205667107699827e-05, + "loss": 0.5825, + "step": 4881 + }, + { + "epoch": 0.15398202176312883, + "grad_norm": 3.0111968517303467, + "learning_rate": 1.9205268054649617e-05, + "loss": 0.5758, + "step": 4882 + }, + { + "epoch": 0.15401356252956946, + "grad_norm": 3.307443857192993, + "learning_rate": 1.920486890553529e-05, + "loss": 0.6011, + "step": 4883 + }, + { + "epoch": 0.1540451032960101, + "grad_norm": 2.0797200202941895, + "learning_rate": 1.9204469660361008e-05, + "loss": 0.3472, + "step": 4884 + }, + { + "epoch": 0.15407664406245072, + "grad_norm": 2.354367256164551, + "learning_rate": 1.9204070319130938e-05, + "loss": 0.4207, + "step": 4885 + }, + { + "epoch": 0.15410818482889135, + "grad_norm": 3.052694320678711, + "learning_rate": 1.9203670881849246e-05, + "loss": 0.5599, + "step": 4886 + }, + { + "epoch": 0.15413972559533196, + "grad_norm": 4.608438014984131, + "learning_rate": 1.9203271348520102e-05, + "loss": 0.5714, + "step": 4887 + }, + { + "epoch": 0.15417126636177259, + "grad_norm": 1.8679701089859009, + "learning_rate": 1.9202871719147675e-05, + "loss": 0.8829, + "step": 4888 + }, + { + "epoch": 0.15420280712821322, + "grad_norm": 3.2602646350860596, + "learning_rate": 1.9202471993736135e-05, + "loss": 0.5866, + "step": 4889 + }, + { + "epoch": 0.15423434789465384, + "grad_norm": 2.4992716312408447, + "learning_rate": 1.9202072172289656e-05, + "loss": 0.5761, + "step": 4890 + }, + { + "epoch": 0.15426588866109447, + "grad_norm": 3.0750043392181396, + "learning_rate": 1.920167225481241e-05, + "loss": 0.6177, + "step": 4891 + }, + { + "epoch": 0.15429742942753508, + "grad_norm": 3.173391342163086, + "learning_rate": 1.9201272241308565e-05, + "loss": 0.5913, + "step": 4892 + }, + { + "epoch": 0.1543289701939757, + "grad_norm": 2.8302605152130127, + "learning_rate": 1.9200872131782305e-05, + "loss": 0.4134, + "step": 4893 + }, + { + "epoch": 0.15436051096041634, + "grad_norm": 2.275123119354248, + "learning_rate": 1.9200471926237796e-05, + "loss": 0.5663, + "step": 4894 + }, + { + "epoch": 0.15439205172685697, + "grad_norm": 5.4131245613098145, + "learning_rate": 1.920007162467922e-05, + "loss": 0.8926, + "step": 4895 + }, + { + "epoch": 0.1544235924932976, + "grad_norm": 2.7919323444366455, + "learning_rate": 1.9199671227110756e-05, + "loss": 0.5715, + "step": 4896 + }, + { + "epoch": 0.15445513325973823, + "grad_norm": 2.437500476837158, + "learning_rate": 1.9199270733536572e-05, + "loss": 0.5635, + "step": 4897 + }, + { + "epoch": 0.15448667402617883, + "grad_norm": 2.73703670501709, + "learning_rate": 1.919887014396086e-05, + "loss": 0.8991, + "step": 4898 + }, + { + "epoch": 0.15451821479261946, + "grad_norm": 2.8884246349334717, + "learning_rate": 1.9198469458387795e-05, + "loss": 0.5793, + "step": 4899 + }, + { + "epoch": 0.1545497555590601, + "grad_norm": 1.863141655921936, + "learning_rate": 1.919806867682156e-05, + "loss": 0.9029, + "step": 4900 + }, + { + "epoch": 0.15458129632550072, + "grad_norm": 2.333913564682007, + "learning_rate": 1.919766779926633e-05, + "loss": 0.5718, + "step": 4901 + }, + { + "epoch": 0.15461283709194135, + "grad_norm": 3.4970052242279053, + "learning_rate": 1.9197266825726298e-05, + "loss": 0.5558, + "step": 4902 + }, + { + "epoch": 0.15464437785838195, + "grad_norm": 4.741415023803711, + "learning_rate": 1.919686575620565e-05, + "loss": 0.6064, + "step": 4903 + }, + { + "epoch": 0.15467591862482258, + "grad_norm": 3.1357967853546143, + "learning_rate": 1.919646459070856e-05, + "loss": 0.5827, + "step": 4904 + }, + { + "epoch": 0.1547074593912632, + "grad_norm": 3.116927146911621, + "learning_rate": 1.9196063329239226e-05, + "loss": 0.5868, + "step": 4905 + }, + { + "epoch": 0.15473900015770384, + "grad_norm": 3.719252824783325, + "learning_rate": 1.9195661971801825e-05, + "loss": 0.5947, + "step": 4906 + }, + { + "epoch": 0.15477054092414447, + "grad_norm": 3.574671983718872, + "learning_rate": 1.919526051840056e-05, + "loss": 0.6012, + "step": 4907 + }, + { + "epoch": 0.15480208169058507, + "grad_norm": 2.61110782623291, + "learning_rate": 1.9194858969039603e-05, + "loss": 0.4284, + "step": 4908 + }, + { + "epoch": 0.1548336224570257, + "grad_norm": 3.6886887550354004, + "learning_rate": 1.9194457323723154e-05, + "loss": 0.6064, + "step": 4909 + }, + { + "epoch": 0.15486516322346633, + "grad_norm": 2.598935604095459, + "learning_rate": 1.919405558245541e-05, + "loss": 0.5602, + "step": 4910 + }, + { + "epoch": 0.15489670398990696, + "grad_norm": 2.692960500717163, + "learning_rate": 1.919365374524055e-05, + "loss": 0.6022, + "step": 4911 + }, + { + "epoch": 0.1549282447563476, + "grad_norm": 4.1523308753967285, + "learning_rate": 1.9193251812082776e-05, + "loss": 0.5774, + "step": 4912 + }, + { + "epoch": 0.1549597855227882, + "grad_norm": 2.124166965484619, + "learning_rate": 1.9192849782986283e-05, + "loss": 0.6112, + "step": 4913 + }, + { + "epoch": 0.15499132628922882, + "grad_norm": 2.4695956707000732, + "learning_rate": 1.9192447657955266e-05, + "loss": 0.5714, + "step": 4914 + }, + { + "epoch": 0.15502286705566945, + "grad_norm": 3.6600728034973145, + "learning_rate": 1.9192045436993915e-05, + "loss": 0.5757, + "step": 4915 + }, + { + "epoch": 0.15505440782211008, + "grad_norm": 4.355360507965088, + "learning_rate": 1.9191643120106435e-05, + "loss": 0.5984, + "step": 4916 + }, + { + "epoch": 0.1550859485885507, + "grad_norm": 3.433077335357666, + "learning_rate": 1.9191240707297024e-05, + "loss": 0.5574, + "step": 4917 + }, + { + "epoch": 0.15511748935499134, + "grad_norm": 4.375749111175537, + "learning_rate": 1.9190838198569873e-05, + "loss": 0.5524, + "step": 4918 + }, + { + "epoch": 0.15514903012143194, + "grad_norm": 2.813568115234375, + "learning_rate": 1.9190435593929197e-05, + "loss": 0.6059, + "step": 4919 + }, + { + "epoch": 0.15518057088787257, + "grad_norm": 2.7040979862213135, + "learning_rate": 1.9190032893379183e-05, + "loss": 0.5697, + "step": 4920 + }, + { + "epoch": 0.1552121116543132, + "grad_norm": 2.88849139213562, + "learning_rate": 1.9189630096924044e-05, + "loss": 0.588, + "step": 4921 + }, + { + "epoch": 0.15524365242075383, + "grad_norm": 2.5696239471435547, + "learning_rate": 1.9189227204567977e-05, + "loss": 0.608, + "step": 4922 + }, + { + "epoch": 0.15527519318719446, + "grad_norm": 3.3654630184173584, + "learning_rate": 1.918882421631519e-05, + "loss": 0.5779, + "step": 4923 + }, + { + "epoch": 0.15530673395363506, + "grad_norm": 3.673642158508301, + "learning_rate": 1.9188421132169884e-05, + "loss": 0.5728, + "step": 4924 + }, + { + "epoch": 0.1553382747200757, + "grad_norm": 2.2861311435699463, + "learning_rate": 1.9188017952136275e-05, + "loss": 0.5676, + "step": 4925 + }, + { + "epoch": 0.15536981548651632, + "grad_norm": 2.832831382751465, + "learning_rate": 1.918761467621856e-05, + "loss": 0.5603, + "step": 4926 + }, + { + "epoch": 0.15540135625295695, + "grad_norm": 2.138071060180664, + "learning_rate": 1.9187211304420953e-05, + "loss": 0.5721, + "step": 4927 + }, + { + "epoch": 0.15543289701939758, + "grad_norm": 3.866720676422119, + "learning_rate": 1.9186807836747663e-05, + "loss": 0.5916, + "step": 4928 + }, + { + "epoch": 0.15546443778583818, + "grad_norm": 2.196183204650879, + "learning_rate": 1.91864042732029e-05, + "loss": 0.4322, + "step": 4929 + }, + { + "epoch": 0.1554959785522788, + "grad_norm": 2.7438724040985107, + "learning_rate": 1.9186000613790874e-05, + "loss": 0.5965, + "step": 4930 + }, + { + "epoch": 0.15552751931871944, + "grad_norm": 2.210083246231079, + "learning_rate": 1.9185596858515797e-05, + "loss": 0.4102, + "step": 4931 + }, + { + "epoch": 0.15555906008516007, + "grad_norm": 2.4842336177825928, + "learning_rate": 1.918519300738189e-05, + "loss": 0.57, + "step": 4932 + }, + { + "epoch": 0.1555906008516007, + "grad_norm": 2.6519339084625244, + "learning_rate": 1.9184789060393357e-05, + "loss": 0.5979, + "step": 4933 + }, + { + "epoch": 0.15562214161804133, + "grad_norm": 1.8927143812179565, + "learning_rate": 1.918438501755442e-05, + "loss": 0.5665, + "step": 4934 + }, + { + "epoch": 0.15565368238448193, + "grad_norm": 1.9965872764587402, + "learning_rate": 1.9183980878869294e-05, + "loss": 0.596, + "step": 4935 + }, + { + "epoch": 0.15568522315092256, + "grad_norm": 2.1654233932495117, + "learning_rate": 1.9183576644342196e-05, + "loss": 0.5838, + "step": 4936 + }, + { + "epoch": 0.1557167639173632, + "grad_norm": 2.010772466659546, + "learning_rate": 1.9183172313977346e-05, + "loss": 0.5885, + "step": 4937 + }, + { + "epoch": 0.15574830468380382, + "grad_norm": 3.236344814300537, + "learning_rate": 1.918276788777896e-05, + "loss": 0.5849, + "step": 4938 + }, + { + "epoch": 0.15577984545024445, + "grad_norm": 44.18450164794922, + "learning_rate": 1.9182363365751263e-05, + "loss": 0.4087, + "step": 4939 + }, + { + "epoch": 0.15581138621668505, + "grad_norm": 3.104612350463867, + "learning_rate": 1.9181958747898473e-05, + "loss": 0.5732, + "step": 4940 + }, + { + "epoch": 0.15584292698312568, + "grad_norm": 2.751380205154419, + "learning_rate": 1.9181554034224817e-05, + "loss": 0.5619, + "step": 4941 + }, + { + "epoch": 0.15587446774956631, + "grad_norm": 3.0122573375701904, + "learning_rate": 1.9181149224734513e-05, + "loss": 0.5586, + "step": 4942 + }, + { + "epoch": 0.15590600851600694, + "grad_norm": 2.646289587020874, + "learning_rate": 1.9180744319431786e-05, + "loss": 0.5804, + "step": 4943 + }, + { + "epoch": 0.15593754928244757, + "grad_norm": 2.4963526725769043, + "learning_rate": 1.9180339318320867e-05, + "loss": 0.5709, + "step": 4944 + }, + { + "epoch": 0.15596909004888818, + "grad_norm": 3.310276746749878, + "learning_rate": 1.9179934221405975e-05, + "loss": 0.5929, + "step": 4945 + }, + { + "epoch": 0.1560006308153288, + "grad_norm": 5.954658031463623, + "learning_rate": 1.9179529028691345e-05, + "loss": 0.5914, + "step": 4946 + }, + { + "epoch": 0.15603217158176944, + "grad_norm": 2.323235034942627, + "learning_rate": 1.91791237401812e-05, + "loss": 0.5879, + "step": 4947 + }, + { + "epoch": 0.15606371234821007, + "grad_norm": 2.6804141998291016, + "learning_rate": 1.917871835587977e-05, + "loss": 0.5967, + "step": 4948 + }, + { + "epoch": 0.1560952531146507, + "grad_norm": 1.924511194229126, + "learning_rate": 1.917831287579129e-05, + "loss": 0.5947, + "step": 4949 + }, + { + "epoch": 0.1561267938810913, + "grad_norm": 1.9431331157684326, + "learning_rate": 1.9177907299919988e-05, + "loss": 0.6001, + "step": 4950 + }, + { + "epoch": 0.15615833464753193, + "grad_norm": 2.8955342769622803, + "learning_rate": 1.9177501628270094e-05, + "loss": 0.5613, + "step": 4951 + }, + { + "epoch": 0.15618987541397256, + "grad_norm": 2.1849899291992188, + "learning_rate": 1.9177095860845845e-05, + "loss": 0.6106, + "step": 4952 + }, + { + "epoch": 0.1562214161804132, + "grad_norm": 2.0087249279022217, + "learning_rate": 1.9176689997651477e-05, + "loss": 0.5861, + "step": 4953 + }, + { + "epoch": 0.15625295694685382, + "grad_norm": 2.018141508102417, + "learning_rate": 1.9176284038691223e-05, + "loss": 0.4151, + "step": 4954 + }, + { + "epoch": 0.15628449771329445, + "grad_norm": 2.069730043411255, + "learning_rate": 1.917587798396932e-05, + "loss": 0.5904, + "step": 4955 + }, + { + "epoch": 0.15631603847973505, + "grad_norm": 2.643126964569092, + "learning_rate": 1.917547183349e-05, + "loss": 0.9451, + "step": 4956 + }, + { + "epoch": 0.15634757924617568, + "grad_norm": 2.408372163772583, + "learning_rate": 1.9175065587257513e-05, + "loss": 0.5806, + "step": 4957 + }, + { + "epoch": 0.1563791200126163, + "grad_norm": 2.3781578540802, + "learning_rate": 1.9174659245276092e-05, + "loss": 0.586, + "step": 4958 + }, + { + "epoch": 0.15641066077905694, + "grad_norm": 2.8537583351135254, + "learning_rate": 1.9174252807549976e-05, + "loss": 0.5919, + "step": 4959 + }, + { + "epoch": 0.15644220154549757, + "grad_norm": 6.2439117431640625, + "learning_rate": 1.9173846274083407e-05, + "loss": 0.5842, + "step": 4960 + }, + { + "epoch": 0.15647374231193817, + "grad_norm": 2.200645685195923, + "learning_rate": 1.917343964488063e-05, + "loss": 0.9106, + "step": 4961 + }, + { + "epoch": 0.1565052830783788, + "grad_norm": 3.1650965213775635, + "learning_rate": 1.917303291994589e-05, + "loss": 0.6088, + "step": 4962 + }, + { + "epoch": 0.15653682384481943, + "grad_norm": 3.3606228828430176, + "learning_rate": 1.9172626099283426e-05, + "loss": 0.5532, + "step": 4963 + }, + { + "epoch": 0.15656836461126006, + "grad_norm": 3.303736925125122, + "learning_rate": 1.9172219182897488e-05, + "loss": 0.5332, + "step": 4964 + }, + { + "epoch": 0.1565999053777007, + "grad_norm": 2.0180654525756836, + "learning_rate": 1.917181217079232e-05, + "loss": 0.5816, + "step": 4965 + }, + { + "epoch": 0.1566314461441413, + "grad_norm": 2.2350871562957764, + "learning_rate": 1.917140506297217e-05, + "loss": 0.5649, + "step": 4966 + }, + { + "epoch": 0.15666298691058192, + "grad_norm": 2.022568464279175, + "learning_rate": 1.9170997859441284e-05, + "loss": 0.4358, + "step": 4967 + }, + { + "epoch": 0.15669452767702255, + "grad_norm": 2.209726333618164, + "learning_rate": 1.917059056020392e-05, + "loss": 0.573, + "step": 4968 + }, + { + "epoch": 0.15672606844346318, + "grad_norm": 2.620110273361206, + "learning_rate": 1.917018316526432e-05, + "loss": 0.5758, + "step": 4969 + }, + { + "epoch": 0.1567576092099038, + "grad_norm": 2.1586267948150635, + "learning_rate": 1.9169775674626735e-05, + "loss": 0.569, + "step": 4970 + }, + { + "epoch": 0.15678914997634444, + "grad_norm": 2.189218521118164, + "learning_rate": 1.9169368088295425e-05, + "loss": 0.6, + "step": 4971 + }, + { + "epoch": 0.15682069074278504, + "grad_norm": 2.335115909576416, + "learning_rate": 1.916896040627464e-05, + "loss": 0.5645, + "step": 4972 + }, + { + "epoch": 0.15685223150922567, + "grad_norm": 2.7504565715789795, + "learning_rate": 1.9168552628568632e-05, + "loss": 0.5849, + "step": 4973 + }, + { + "epoch": 0.1568837722756663, + "grad_norm": 2.3211829662323, + "learning_rate": 1.9168144755181654e-05, + "loss": 0.5865, + "step": 4974 + }, + { + "epoch": 0.15691531304210693, + "grad_norm": 2.7486000061035156, + "learning_rate": 1.916773678611797e-05, + "loss": 0.5944, + "step": 4975 + }, + { + "epoch": 0.15694685380854756, + "grad_norm": 2.284242868423462, + "learning_rate": 1.9167328721381835e-05, + "loss": 0.6002, + "step": 4976 + }, + { + "epoch": 0.15697839457498816, + "grad_norm": 2.58205509185791, + "learning_rate": 1.916692056097751e-05, + "loss": 0.5833, + "step": 4977 + }, + { + "epoch": 0.1570099353414288, + "grad_norm": 2.3623759746551514, + "learning_rate": 1.9166512304909245e-05, + "loss": 0.5806, + "step": 4978 + }, + { + "epoch": 0.15704147610786942, + "grad_norm": 1.908484935760498, + "learning_rate": 1.9166103953181308e-05, + "loss": 0.9494, + "step": 4979 + }, + { + "epoch": 0.15707301687431005, + "grad_norm": 2.7099263668060303, + "learning_rate": 1.916569550579796e-05, + "loss": 0.6019, + "step": 4980 + }, + { + "epoch": 0.15710455764075068, + "grad_norm": 2.74349045753479, + "learning_rate": 1.916528696276346e-05, + "loss": 0.5735, + "step": 4981 + }, + { + "epoch": 0.15713609840719128, + "grad_norm": 2.51916766166687, + "learning_rate": 1.9164878324082073e-05, + "loss": 0.5943, + "step": 4982 + }, + { + "epoch": 0.1571676391736319, + "grad_norm": 2.575486421585083, + "learning_rate": 1.916446958975807e-05, + "loss": 0.4323, + "step": 4983 + }, + { + "epoch": 0.15719917994007254, + "grad_norm": 2.6232826709747314, + "learning_rate": 1.9164060759795705e-05, + "loss": 0.5668, + "step": 4984 + }, + { + "epoch": 0.15723072070651317, + "grad_norm": 3.6173436641693115, + "learning_rate": 1.9163651834199256e-05, + "loss": 0.5722, + "step": 4985 + }, + { + "epoch": 0.1572622614729538, + "grad_norm": 3.8867690563201904, + "learning_rate": 1.916324281297298e-05, + "loss": 0.5381, + "step": 4986 + }, + { + "epoch": 0.1572938022393944, + "grad_norm": 2.0677945613861084, + "learning_rate": 1.9162833696121152e-05, + "loss": 0.921, + "step": 4987 + }, + { + "epoch": 0.15732534300583503, + "grad_norm": 2.1637628078460693, + "learning_rate": 1.916242448364804e-05, + "loss": 0.3844, + "step": 4988 + }, + { + "epoch": 0.15735688377227566, + "grad_norm": 3.6271119117736816, + "learning_rate": 1.9162015175557914e-05, + "loss": 0.5796, + "step": 4989 + }, + { + "epoch": 0.1573884245387163, + "grad_norm": 2.896975517272949, + "learning_rate": 1.9161605771855046e-05, + "loss": 0.61, + "step": 4990 + }, + { + "epoch": 0.15741996530515692, + "grad_norm": 1.6517807245254517, + "learning_rate": 1.9161196272543704e-05, + "loss": 0.8796, + "step": 4991 + }, + { + "epoch": 0.15745150607159755, + "grad_norm": 3.887352466583252, + "learning_rate": 1.916078667762817e-05, + "loss": 0.5562, + "step": 4992 + }, + { + "epoch": 0.15748304683803815, + "grad_norm": 2.5358335971832275, + "learning_rate": 1.9160376987112713e-05, + "loss": 0.566, + "step": 4993 + }, + { + "epoch": 0.15751458760447878, + "grad_norm": 3.139160394668579, + "learning_rate": 1.915996720100161e-05, + "loss": 0.4339, + "step": 4994 + }, + { + "epoch": 0.1575461283709194, + "grad_norm": 3.474188804626465, + "learning_rate": 1.9159557319299137e-05, + "loss": 0.9529, + "step": 4995 + }, + { + "epoch": 0.15757766913736004, + "grad_norm": 4.535307884216309, + "learning_rate": 1.9159147342009566e-05, + "loss": 0.5587, + "step": 4996 + }, + { + "epoch": 0.15760920990380067, + "grad_norm": 2.66784930229187, + "learning_rate": 1.9158737269137185e-05, + "loss": 0.5807, + "step": 4997 + }, + { + "epoch": 0.15764075067024128, + "grad_norm": 2.6025502681732178, + "learning_rate": 1.9158327100686267e-05, + "loss": 0.5817, + "step": 4998 + }, + { + "epoch": 0.1576722914366819, + "grad_norm": 4.878999710083008, + "learning_rate": 1.9157916836661095e-05, + "loss": 0.907, + "step": 4999 + }, + { + "epoch": 0.15770383220312253, + "grad_norm": 3.3408641815185547, + "learning_rate": 1.915750647706595e-05, + "loss": 0.5847, + "step": 5000 + }, + { + "epoch": 0.15773537296956316, + "grad_norm": 2.7538318634033203, + "learning_rate": 1.9157096021905118e-05, + "loss": 0.9642, + "step": 5001 + }, + { + "epoch": 0.1577669137360038, + "grad_norm": 2.34601092338562, + "learning_rate": 1.9156685471182875e-05, + "loss": 0.5808, + "step": 5002 + }, + { + "epoch": 0.1577984545024444, + "grad_norm": 3.1696701049804688, + "learning_rate": 1.915627482490351e-05, + "loss": 0.5655, + "step": 5003 + }, + { + "epoch": 0.15782999526888503, + "grad_norm": 2.588193416595459, + "learning_rate": 1.915586408307131e-05, + "loss": 0.5748, + "step": 5004 + }, + { + "epoch": 0.15786153603532566, + "grad_norm": 2.5688371658325195, + "learning_rate": 1.915545324569055e-05, + "loss": 0.6078, + "step": 5005 + }, + { + "epoch": 0.15789307680176629, + "grad_norm": 2.47332763671875, + "learning_rate": 1.9155042312765534e-05, + "loss": 0.9031, + "step": 5006 + }, + { + "epoch": 0.15792461756820692, + "grad_norm": 2.973313093185425, + "learning_rate": 1.9154631284300543e-05, + "loss": 0.5826, + "step": 5007 + }, + { + "epoch": 0.15795615833464755, + "grad_norm": 2.7898616790771484, + "learning_rate": 1.9154220160299862e-05, + "loss": 0.5495, + "step": 5008 + }, + { + "epoch": 0.15798769910108815, + "grad_norm": 3.33120059967041, + "learning_rate": 1.915380894076779e-05, + "loss": 0.5847, + "step": 5009 + }, + { + "epoch": 0.15801923986752878, + "grad_norm": 1.8023124933242798, + "learning_rate": 1.9153397625708612e-05, + "loss": 0.3953, + "step": 5010 + }, + { + "epoch": 0.1580507806339694, + "grad_norm": 3.326401710510254, + "learning_rate": 1.9152986215126622e-05, + "loss": 0.5777, + "step": 5011 + }, + { + "epoch": 0.15808232140041004, + "grad_norm": 3.03523325920105, + "learning_rate": 1.9152574709026112e-05, + "loss": 0.5645, + "step": 5012 + }, + { + "epoch": 0.15811386216685067, + "grad_norm": 3.1485373973846436, + "learning_rate": 1.9152163107411382e-05, + "loss": 0.6056, + "step": 5013 + }, + { + "epoch": 0.15814540293329127, + "grad_norm": 3.207791566848755, + "learning_rate": 1.915175141028672e-05, + "loss": 0.5859, + "step": 5014 + }, + { + "epoch": 0.1581769436997319, + "grad_norm": 3.625908136367798, + "learning_rate": 1.915133961765643e-05, + "loss": 0.564, + "step": 5015 + }, + { + "epoch": 0.15820848446617253, + "grad_norm": 3.027904987335205, + "learning_rate": 1.91509277295248e-05, + "loss": 0.3956, + "step": 5016 + }, + { + "epoch": 0.15824002523261316, + "grad_norm": 2.4553685188293457, + "learning_rate": 1.915051574589614e-05, + "loss": 0.566, + "step": 5017 + }, + { + "epoch": 0.1582715659990538, + "grad_norm": 2.759150505065918, + "learning_rate": 1.9150103666774734e-05, + "loss": 0.5682, + "step": 5018 + }, + { + "epoch": 0.1583031067654944, + "grad_norm": 4.033566951751709, + "learning_rate": 1.91496914921649e-05, + "loss": 0.5903, + "step": 5019 + }, + { + "epoch": 0.15833464753193502, + "grad_norm": 5.119388580322266, + "learning_rate": 1.9149279222070926e-05, + "loss": 0.5433, + "step": 5020 + }, + { + "epoch": 0.15836618829837565, + "grad_norm": 2.5290639400482178, + "learning_rate": 1.914886685649712e-05, + "loss": 0.5729, + "step": 5021 + }, + { + "epoch": 0.15839772906481628, + "grad_norm": 2.1629528999328613, + "learning_rate": 1.9148454395447786e-05, + "loss": 0.5668, + "step": 5022 + }, + { + "epoch": 0.1584292698312569, + "grad_norm": 3.42170786857605, + "learning_rate": 1.9148041838927227e-05, + "loss": 0.5991, + "step": 5023 + }, + { + "epoch": 0.15846081059769754, + "grad_norm": 2.631276845932007, + "learning_rate": 1.9147629186939743e-05, + "loss": 0.6016, + "step": 5024 + }, + { + "epoch": 0.15849235136413814, + "grad_norm": 3.5300629138946533, + "learning_rate": 1.914721643948965e-05, + "loss": 0.5866, + "step": 5025 + }, + { + "epoch": 0.15852389213057877, + "grad_norm": 3.070882797241211, + "learning_rate": 1.9146803596581255e-05, + "loss": 0.5808, + "step": 5026 + }, + { + "epoch": 0.1585554328970194, + "grad_norm": 1.782670259475708, + "learning_rate": 1.9146390658218857e-05, + "loss": 0.5801, + "step": 5027 + }, + { + "epoch": 0.15858697366346003, + "grad_norm": 8.440407752990723, + "learning_rate": 1.914597762440677e-05, + "loss": 0.5638, + "step": 5028 + }, + { + "epoch": 0.15861851442990066, + "grad_norm": 3.030862808227539, + "learning_rate": 1.9145564495149305e-05, + "loss": 0.5729, + "step": 5029 + }, + { + "epoch": 0.15865005519634126, + "grad_norm": 2.788724899291992, + "learning_rate": 1.9145151270450773e-05, + "loss": 0.5483, + "step": 5030 + }, + { + "epoch": 0.1586815959627819, + "grad_norm": 2.025850534439087, + "learning_rate": 1.9144737950315487e-05, + "loss": 0.555, + "step": 5031 + }, + { + "epoch": 0.15871313672922252, + "grad_norm": 3.5802934169769287, + "learning_rate": 1.914432453474776e-05, + "loss": 0.5655, + "step": 5032 + }, + { + "epoch": 0.15874467749566315, + "grad_norm": 1.857330083847046, + "learning_rate": 1.9143911023751907e-05, + "loss": 0.9296, + "step": 5033 + }, + { + "epoch": 0.15877621826210378, + "grad_norm": 2.5762040615081787, + "learning_rate": 1.914349741733224e-05, + "loss": 0.6167, + "step": 5034 + }, + { + "epoch": 0.15880775902854438, + "grad_norm": 2.546536922454834, + "learning_rate": 1.914308371549308e-05, + "loss": 0.5948, + "step": 5035 + }, + { + "epoch": 0.158839299794985, + "grad_norm": 2.512620449066162, + "learning_rate": 1.9142669918238742e-05, + "loss": 0.5395, + "step": 5036 + }, + { + "epoch": 0.15887084056142564, + "grad_norm": 2.5057497024536133, + "learning_rate": 1.9142256025573545e-05, + "loss": 0.5806, + "step": 5037 + }, + { + "epoch": 0.15890238132786627, + "grad_norm": 2.7319202423095703, + "learning_rate": 1.9141842037501806e-05, + "loss": 0.5605, + "step": 5038 + }, + { + "epoch": 0.1589339220943069, + "grad_norm": 2.5238380432128906, + "learning_rate": 1.914142795402785e-05, + "loss": 0.5616, + "step": 5039 + }, + { + "epoch": 0.1589654628607475, + "grad_norm": 2.389063835144043, + "learning_rate": 1.914101377515599e-05, + "loss": 0.5586, + "step": 5040 + }, + { + "epoch": 0.15899700362718813, + "grad_norm": 2.8583052158355713, + "learning_rate": 1.9140599500890553e-05, + "loss": 0.6113, + "step": 5041 + }, + { + "epoch": 0.15902854439362876, + "grad_norm": 2.3372104167938232, + "learning_rate": 1.9140185131235865e-05, + "loss": 0.5699, + "step": 5042 + }, + { + "epoch": 0.1590600851600694, + "grad_norm": 2.598681688308716, + "learning_rate": 1.913977066619625e-05, + "loss": 0.5933, + "step": 5043 + }, + { + "epoch": 0.15909162592651002, + "grad_norm": 5.704716205596924, + "learning_rate": 1.913935610577603e-05, + "loss": 0.57, + "step": 5044 + }, + { + "epoch": 0.15912316669295065, + "grad_norm": 2.691952705383301, + "learning_rate": 1.913894144997953e-05, + "loss": 0.5914, + "step": 5045 + }, + { + "epoch": 0.15915470745939125, + "grad_norm": 3.2981555461883545, + "learning_rate": 1.9138526698811083e-05, + "loss": 0.5686, + "step": 5046 + }, + { + "epoch": 0.15918624822583188, + "grad_norm": 3.349775552749634, + "learning_rate": 1.9138111852275013e-05, + "loss": 0.5539, + "step": 5047 + }, + { + "epoch": 0.1592177889922725, + "grad_norm": 8.133042335510254, + "learning_rate": 1.913769691037565e-05, + "loss": 0.5876, + "step": 5048 + }, + { + "epoch": 0.15924932975871314, + "grad_norm": 3.2101545333862305, + "learning_rate": 1.9137281873117325e-05, + "loss": 0.5923, + "step": 5049 + }, + { + "epoch": 0.15928087052515377, + "grad_norm": 2.8702783584594727, + "learning_rate": 1.9136866740504368e-05, + "loss": 0.5671, + "step": 5050 + }, + { + "epoch": 0.15931241129159437, + "grad_norm": 6.290864944458008, + "learning_rate": 1.9136451512541112e-05, + "loss": 0.6141, + "step": 5051 + }, + { + "epoch": 0.159343952058035, + "grad_norm": 3.064589023590088, + "learning_rate": 1.913603618923189e-05, + "loss": 0.5668, + "step": 5052 + }, + { + "epoch": 0.15937549282447563, + "grad_norm": 2.6864326000213623, + "learning_rate": 1.913562077058104e-05, + "loss": 0.5868, + "step": 5053 + }, + { + "epoch": 0.15940703359091626, + "grad_norm": 3.329864978790283, + "learning_rate": 1.9135205256592886e-05, + "loss": 0.5657, + "step": 5054 + }, + { + "epoch": 0.1594385743573569, + "grad_norm": 2.8795976638793945, + "learning_rate": 1.913478964727178e-05, + "loss": 0.5757, + "step": 5055 + }, + { + "epoch": 0.1594701151237975, + "grad_norm": 2.858027696609497, + "learning_rate": 1.9134373942622046e-05, + "loss": 0.5923, + "step": 5056 + }, + { + "epoch": 0.15950165589023813, + "grad_norm": 6.794103622436523, + "learning_rate": 1.9133958142648028e-05, + "loss": 0.5852, + "step": 5057 + }, + { + "epoch": 0.15953319665667876, + "grad_norm": 2.4414749145507812, + "learning_rate": 1.9133542247354065e-05, + "loss": 0.6228, + "step": 5058 + }, + { + "epoch": 0.15956473742311938, + "grad_norm": 2.5118632316589355, + "learning_rate": 1.9133126256744495e-05, + "loss": 0.9236, + "step": 5059 + }, + { + "epoch": 0.15959627818956001, + "grad_norm": 3.306664228439331, + "learning_rate": 1.9132710170823667e-05, + "loss": 0.5833, + "step": 5060 + }, + { + "epoch": 0.15962781895600064, + "grad_norm": 3.2954580783843994, + "learning_rate": 1.913229398959591e-05, + "loss": 0.5992, + "step": 5061 + }, + { + "epoch": 0.15965935972244125, + "grad_norm": 2.7322206497192383, + "learning_rate": 1.913187771306558e-05, + "loss": 0.9431, + "step": 5062 + }, + { + "epoch": 0.15969090048888188, + "grad_norm": 5.098703384399414, + "learning_rate": 1.913146134123701e-05, + "loss": 0.4078, + "step": 5063 + }, + { + "epoch": 0.1597224412553225, + "grad_norm": 3.1162712574005127, + "learning_rate": 1.9131044874114553e-05, + "loss": 0.59, + "step": 5064 + }, + { + "epoch": 0.15975398202176314, + "grad_norm": 3.080657482147217, + "learning_rate": 1.9130628311702554e-05, + "loss": 0.5817, + "step": 5065 + }, + { + "epoch": 0.15978552278820377, + "grad_norm": 3.1345841884613037, + "learning_rate": 1.9130211654005355e-05, + "loss": 0.5849, + "step": 5066 + }, + { + "epoch": 0.15981706355464437, + "grad_norm": 3.690284252166748, + "learning_rate": 1.912979490102731e-05, + "loss": 0.5735, + "step": 5067 + }, + { + "epoch": 0.159848604321085, + "grad_norm": 5.195720195770264, + "learning_rate": 1.9129378052772768e-05, + "loss": 0.6048, + "step": 5068 + }, + { + "epoch": 0.15988014508752563, + "grad_norm": 4.9911112785339355, + "learning_rate": 1.9128961109246074e-05, + "loss": 0.8934, + "step": 5069 + }, + { + "epoch": 0.15991168585396626, + "grad_norm": 3.4602625370025635, + "learning_rate": 1.9128544070451584e-05, + "loss": 0.5754, + "step": 5070 + }, + { + "epoch": 0.1599432266204069, + "grad_norm": 3.326467514038086, + "learning_rate": 1.912812693639365e-05, + "loss": 0.6083, + "step": 5071 + }, + { + "epoch": 0.1599747673868475, + "grad_norm": 2.8114562034606934, + "learning_rate": 1.912770970707662e-05, + "loss": 0.9191, + "step": 5072 + }, + { + "epoch": 0.16000630815328812, + "grad_norm": 4.8544416427612305, + "learning_rate": 1.9127292382504853e-05, + "loss": 0.597, + "step": 5073 + }, + { + "epoch": 0.16003784891972875, + "grad_norm": 6.580648422241211, + "learning_rate": 1.9126874962682703e-05, + "loss": 0.5855, + "step": 5074 + }, + { + "epoch": 0.16006938968616938, + "grad_norm": 6.389339923858643, + "learning_rate": 1.9126457447614527e-05, + "loss": 0.5676, + "step": 5075 + }, + { + "epoch": 0.16010093045261, + "grad_norm": 2.95796275138855, + "learning_rate": 1.9126039837304682e-05, + "loss": 0.5761, + "step": 5076 + }, + { + "epoch": 0.1601324712190506, + "grad_norm": 11.632195472717285, + "learning_rate": 1.9125622131757522e-05, + "loss": 0.5444, + "step": 5077 + }, + { + "epoch": 0.16016401198549124, + "grad_norm": 3.574458599090576, + "learning_rate": 1.912520433097741e-05, + "loss": 0.5814, + "step": 5078 + }, + { + "epoch": 0.16019555275193187, + "grad_norm": 3.935697555541992, + "learning_rate": 1.9124786434968707e-05, + "loss": 0.5865, + "step": 5079 + }, + { + "epoch": 0.1602270935183725, + "grad_norm": 4.25323486328125, + "learning_rate": 1.9124368443735772e-05, + "loss": 0.5851, + "step": 5080 + }, + { + "epoch": 0.16025863428481313, + "grad_norm": 2.6888880729675293, + "learning_rate": 1.9123950357282964e-05, + "loss": 0.5887, + "step": 5081 + }, + { + "epoch": 0.16029017505125376, + "grad_norm": 3.301939010620117, + "learning_rate": 1.9123532175614655e-05, + "loss": 0.5799, + "step": 5082 + }, + { + "epoch": 0.16032171581769436, + "grad_norm": 2.878058433532715, + "learning_rate": 1.91231138987352e-05, + "loss": 0.5898, + "step": 5083 + }, + { + "epoch": 0.160353256584135, + "grad_norm": 3.4906420707702637, + "learning_rate": 1.9122695526648968e-05, + "loss": 0.5656, + "step": 5084 + }, + { + "epoch": 0.16038479735057562, + "grad_norm": 3.9729552268981934, + "learning_rate": 1.9122277059360325e-05, + "loss": 0.5706, + "step": 5085 + }, + { + "epoch": 0.16041633811701625, + "grad_norm": 2.772900104522705, + "learning_rate": 1.912185849687364e-05, + "loss": 0.593, + "step": 5086 + }, + { + "epoch": 0.16044787888345688, + "grad_norm": 3.177272319793701, + "learning_rate": 1.9121439839193274e-05, + "loss": 0.5696, + "step": 5087 + }, + { + "epoch": 0.16047941964989748, + "grad_norm": 4.5333452224731445, + "learning_rate": 1.9121021086323604e-05, + "loss": 0.6088, + "step": 5088 + }, + { + "epoch": 0.1605109604163381, + "grad_norm": 2.970458745956421, + "learning_rate": 1.9120602238268996e-05, + "loss": 0.5651, + "step": 5089 + }, + { + "epoch": 0.16054250118277874, + "grad_norm": 2.55179500579834, + "learning_rate": 1.912018329503382e-05, + "loss": 0.5647, + "step": 5090 + }, + { + "epoch": 0.16057404194921937, + "grad_norm": 2.7100470066070557, + "learning_rate": 1.911976425662245e-05, + "loss": 0.5794, + "step": 5091 + }, + { + "epoch": 0.16060558271566, + "grad_norm": 3.0321414470672607, + "learning_rate": 1.9119345123039262e-05, + "loss": 0.6104, + "step": 5092 + }, + { + "epoch": 0.1606371234821006, + "grad_norm": 4.247925281524658, + "learning_rate": 1.9118925894288623e-05, + "loss": 0.613, + "step": 5093 + }, + { + "epoch": 0.16066866424854123, + "grad_norm": 2.7621562480926514, + "learning_rate": 1.9118506570374914e-05, + "loss": 0.5865, + "step": 5094 + }, + { + "epoch": 0.16070020501498186, + "grad_norm": 2.8562965393066406, + "learning_rate": 1.9118087151302504e-05, + "loss": 0.5792, + "step": 5095 + }, + { + "epoch": 0.1607317457814225, + "grad_norm": 3.3750627040863037, + "learning_rate": 1.9117667637075778e-05, + "loss": 0.549, + "step": 5096 + }, + { + "epoch": 0.16076328654786312, + "grad_norm": 3.28963041305542, + "learning_rate": 1.911724802769911e-05, + "loss": 0.5493, + "step": 5097 + }, + { + "epoch": 0.16079482731430375, + "grad_norm": 3.978052854537964, + "learning_rate": 1.9116828323176883e-05, + "loss": 0.5943, + "step": 5098 + }, + { + "epoch": 0.16082636808074435, + "grad_norm": 4.830804824829102, + "learning_rate": 1.911640852351347e-05, + "loss": 0.6043, + "step": 5099 + }, + { + "epoch": 0.16085790884718498, + "grad_norm": 3.769892930984497, + "learning_rate": 1.9115988628713256e-05, + "loss": 0.5973, + "step": 5100 + }, + { + "epoch": 0.1608894496136256, + "grad_norm": 3.2399659156799316, + "learning_rate": 1.911556863878062e-05, + "loss": 0.5803, + "step": 5101 + }, + { + "epoch": 0.16092099038006624, + "grad_norm": 5.607853412628174, + "learning_rate": 1.911514855371995e-05, + "loss": 0.4097, + "step": 5102 + }, + { + "epoch": 0.16095253114650687, + "grad_norm": 3.9205892086029053, + "learning_rate": 1.9114728373535626e-05, + "loss": 0.5924, + "step": 5103 + }, + { + "epoch": 0.16098407191294747, + "grad_norm": 11.541608810424805, + "learning_rate": 1.9114308098232037e-05, + "loss": 0.5644, + "step": 5104 + }, + { + "epoch": 0.1610156126793881, + "grad_norm": 3.31773042678833, + "learning_rate": 1.911388772781356e-05, + "loss": 0.5695, + "step": 5105 + }, + { + "epoch": 0.16104715344582873, + "grad_norm": 5.70050573348999, + "learning_rate": 1.911346726228459e-05, + "loss": 0.5749, + "step": 5106 + }, + { + "epoch": 0.16107869421226936, + "grad_norm": 2.3202245235443115, + "learning_rate": 1.9113046701649517e-05, + "loss": 0.6038, + "step": 5107 + }, + { + "epoch": 0.16111023497871, + "grad_norm": 3.788243293762207, + "learning_rate": 1.911262604591272e-05, + "loss": 0.9259, + "step": 5108 + }, + { + "epoch": 0.1611417757451506, + "grad_norm": 3.621447801589966, + "learning_rate": 1.9112205295078598e-05, + "loss": 0.9213, + "step": 5109 + }, + { + "epoch": 0.16117331651159122, + "grad_norm": 3.061953067779541, + "learning_rate": 1.9111784449151538e-05, + "loss": 0.6084, + "step": 5110 + }, + { + "epoch": 0.16120485727803185, + "grad_norm": 3.5663390159606934, + "learning_rate": 1.911136350813593e-05, + "loss": 0.5898, + "step": 5111 + }, + { + "epoch": 0.16123639804447248, + "grad_norm": 8.504860877990723, + "learning_rate": 1.9110942472036173e-05, + "loss": 0.4011, + "step": 5112 + }, + { + "epoch": 0.16126793881091311, + "grad_norm": 3.708416700363159, + "learning_rate": 1.9110521340856653e-05, + "loss": 0.5664, + "step": 5113 + }, + { + "epoch": 0.16129947957735372, + "grad_norm": 2.6396987438201904, + "learning_rate": 1.9110100114601767e-05, + "loss": 0.5908, + "step": 5114 + }, + { + "epoch": 0.16133102034379435, + "grad_norm": 3.458510398864746, + "learning_rate": 1.9109678793275917e-05, + "loss": 0.5894, + "step": 5115 + }, + { + "epoch": 0.16136256111023498, + "grad_norm": 3.3231701850891113, + "learning_rate": 1.9109257376883494e-05, + "loss": 0.5509, + "step": 5116 + }, + { + "epoch": 0.1613941018766756, + "grad_norm": 4.435111045837402, + "learning_rate": 1.9108835865428895e-05, + "loss": 0.5819, + "step": 5117 + }, + { + "epoch": 0.16142564264311624, + "grad_norm": 2.4204814434051514, + "learning_rate": 1.9108414258916525e-05, + "loss": 0.5734, + "step": 5118 + }, + { + "epoch": 0.16145718340955686, + "grad_norm": 2.352877616882324, + "learning_rate": 1.9107992557350775e-05, + "loss": 0.5824, + "step": 5119 + }, + { + "epoch": 0.16148872417599747, + "grad_norm": 3.4553403854370117, + "learning_rate": 1.9107570760736053e-05, + "loss": 0.5723, + "step": 5120 + }, + { + "epoch": 0.1615202649424381, + "grad_norm": 2.3634092807769775, + "learning_rate": 1.910714886907676e-05, + "loss": 0.5968, + "step": 5121 + }, + { + "epoch": 0.16155180570887873, + "grad_norm": 3.1282429695129395, + "learning_rate": 1.9106726882377293e-05, + "loss": 0.5857, + "step": 5122 + }, + { + "epoch": 0.16158334647531936, + "grad_norm": 6.907857894897461, + "learning_rate": 1.9106304800642062e-05, + "loss": 0.5613, + "step": 5123 + }, + { + "epoch": 0.16161488724175999, + "grad_norm": 2.496361255645752, + "learning_rate": 1.910588262387547e-05, + "loss": 0.5887, + "step": 5124 + }, + { + "epoch": 0.1616464280082006, + "grad_norm": 3.103210210800171, + "learning_rate": 1.9105460352081922e-05, + "loss": 0.5422, + "step": 5125 + }, + { + "epoch": 0.16167796877464122, + "grad_norm": 2.395054340362549, + "learning_rate": 1.9105037985265823e-05, + "loss": 0.6028, + "step": 5126 + }, + { + "epoch": 0.16170950954108185, + "grad_norm": 2.3498709201812744, + "learning_rate": 1.9104615523431587e-05, + "loss": 0.5633, + "step": 5127 + }, + { + "epoch": 0.16174105030752248, + "grad_norm": 2.6383261680603027, + "learning_rate": 1.9104192966583615e-05, + "loss": 0.5693, + "step": 5128 + }, + { + "epoch": 0.1617725910739631, + "grad_norm": 6.899106502532959, + "learning_rate": 1.910377031472632e-05, + "loss": 1.1174, + "step": 5129 + }, + { + "epoch": 0.1618041318404037, + "grad_norm": 2.614548921585083, + "learning_rate": 1.9103347567864114e-05, + "loss": 0.6086, + "step": 5130 + }, + { + "epoch": 0.16183567260684434, + "grad_norm": 4.900826930999756, + "learning_rate": 1.9102924726001408e-05, + "loss": 0.9861, + "step": 5131 + }, + { + "epoch": 0.16186721337328497, + "grad_norm": 3.8710763454437256, + "learning_rate": 1.9102501789142614e-05, + "loss": 0.9502, + "step": 5132 + }, + { + "epoch": 0.1618987541397256, + "grad_norm": 2.4418442249298096, + "learning_rate": 1.9102078757292145e-05, + "loss": 0.5857, + "step": 5133 + }, + { + "epoch": 0.16193029490616623, + "grad_norm": 3.51599383354187, + "learning_rate": 1.9101655630454416e-05, + "loss": 0.5766, + "step": 5134 + }, + { + "epoch": 0.16196183567260686, + "grad_norm": 2.693899154663086, + "learning_rate": 1.9101232408633844e-05, + "loss": 0.5953, + "step": 5135 + }, + { + "epoch": 0.16199337643904746, + "grad_norm": 2.35205078125, + "learning_rate": 1.9100809091834848e-05, + "loss": 0.5739, + "step": 5136 + }, + { + "epoch": 0.1620249172054881, + "grad_norm": 3.661376714706421, + "learning_rate": 1.910038568006184e-05, + "loss": 0.5825, + "step": 5137 + }, + { + "epoch": 0.16205645797192872, + "grad_norm": 2.7909700870513916, + "learning_rate": 1.909996217331924e-05, + "loss": 0.3913, + "step": 5138 + }, + { + "epoch": 0.16208799873836935, + "grad_norm": 4.107274532318115, + "learning_rate": 1.909953857161147e-05, + "loss": 0.5769, + "step": 5139 + }, + { + "epoch": 0.16211953950480998, + "grad_norm": 3.594683885574341, + "learning_rate": 1.909911487494295e-05, + "loss": 0.4451, + "step": 5140 + }, + { + "epoch": 0.16215108027125058, + "grad_norm": 4.100887298583984, + "learning_rate": 1.90986910833181e-05, + "loss": 0.5908, + "step": 5141 + }, + { + "epoch": 0.1621826210376912, + "grad_norm": 2.935671329498291, + "learning_rate": 1.9098267196741345e-05, + "loss": 0.5829, + "step": 5142 + }, + { + "epoch": 0.16221416180413184, + "grad_norm": 4.127333164215088, + "learning_rate": 1.9097843215217106e-05, + "loss": 0.5799, + "step": 5143 + }, + { + "epoch": 0.16224570257057247, + "grad_norm": 2.8564090728759766, + "learning_rate": 1.9097419138749807e-05, + "loss": 0.5455, + "step": 5144 + }, + { + "epoch": 0.1622772433370131, + "grad_norm": 3.3370769023895264, + "learning_rate": 1.9096994967343878e-05, + "loss": 0.6259, + "step": 5145 + }, + { + "epoch": 0.1623087841034537, + "grad_norm": 3.0716049671173096, + "learning_rate": 1.909657070100374e-05, + "loss": 0.5695, + "step": 5146 + }, + { + "epoch": 0.16234032486989433, + "grad_norm": 2.229719638824463, + "learning_rate": 1.9096146339733827e-05, + "loss": 0.5807, + "step": 5147 + }, + { + "epoch": 0.16237186563633496, + "grad_norm": 2.528427839279175, + "learning_rate": 1.909572188353856e-05, + "loss": 0.5935, + "step": 5148 + }, + { + "epoch": 0.1624034064027756, + "grad_norm": 2.258831262588501, + "learning_rate": 1.9095297332422378e-05, + "loss": 0.5996, + "step": 5149 + }, + { + "epoch": 0.16243494716921622, + "grad_norm": 2.7408175468444824, + "learning_rate": 1.9094872686389702e-05, + "loss": 0.6079, + "step": 5150 + }, + { + "epoch": 0.16246648793565685, + "grad_norm": 2.546107530593872, + "learning_rate": 1.909444794544497e-05, + "loss": 0.5424, + "step": 5151 + }, + { + "epoch": 0.16249802870209745, + "grad_norm": 2.524461269378662, + "learning_rate": 1.909402310959261e-05, + "loss": 0.5999, + "step": 5152 + }, + { + "epoch": 0.16252956946853808, + "grad_norm": 2.0135419368743896, + "learning_rate": 1.9093598178837055e-05, + "loss": 0.5795, + "step": 5153 + }, + { + "epoch": 0.1625611102349787, + "grad_norm": 2.906998634338379, + "learning_rate": 1.9093173153182746e-05, + "loss": 0.5752, + "step": 5154 + }, + { + "epoch": 0.16259265100141934, + "grad_norm": 2.118028402328491, + "learning_rate": 1.9092748032634115e-05, + "loss": 0.5757, + "step": 5155 + }, + { + "epoch": 0.16262419176785997, + "grad_norm": 3.6161930561065674, + "learning_rate": 1.9092322817195595e-05, + "loss": 0.5972, + "step": 5156 + }, + { + "epoch": 0.16265573253430057, + "grad_norm": 4.439910888671875, + "learning_rate": 1.909189750687163e-05, + "loss": 0.5829, + "step": 5157 + }, + { + "epoch": 0.1626872733007412, + "grad_norm": 2.193821430206299, + "learning_rate": 1.909147210166665e-05, + "loss": 0.5721, + "step": 5158 + }, + { + "epoch": 0.16271881406718183, + "grad_norm": 3.516564130783081, + "learning_rate": 1.9091046601585103e-05, + "loss": 0.5735, + "step": 5159 + }, + { + "epoch": 0.16275035483362246, + "grad_norm": 2.8079516887664795, + "learning_rate": 1.9090621006631425e-05, + "loss": 0.5917, + "step": 5160 + }, + { + "epoch": 0.1627818956000631, + "grad_norm": 2.3187761306762695, + "learning_rate": 1.909019531681006e-05, + "loss": 0.5696, + "step": 5161 + }, + { + "epoch": 0.1628134363665037, + "grad_norm": 2.1882643699645996, + "learning_rate": 1.908976953212544e-05, + "loss": 0.5857, + "step": 5162 + }, + { + "epoch": 0.16284497713294432, + "grad_norm": 12.458009719848633, + "learning_rate": 1.9089343652582028e-05, + "loss": 1.3099, + "step": 5163 + }, + { + "epoch": 0.16287651789938495, + "grad_norm": 2.7862813472747803, + "learning_rate": 1.908891767818425e-05, + "loss": 0.5869, + "step": 5164 + }, + { + "epoch": 0.16290805866582558, + "grad_norm": 2.814365863800049, + "learning_rate": 1.9088491608936564e-05, + "loss": 0.598, + "step": 5165 + }, + { + "epoch": 0.1629395994322662, + "grad_norm": 2.599350690841675, + "learning_rate": 1.9088065444843407e-05, + "loss": 0.6102, + "step": 5166 + }, + { + "epoch": 0.16297114019870682, + "grad_norm": 1.9070543050765991, + "learning_rate": 1.908763918590923e-05, + "loss": 0.5605, + "step": 5167 + }, + { + "epoch": 0.16300268096514745, + "grad_norm": 2.982609748840332, + "learning_rate": 1.9087212832138485e-05, + "loss": 0.5839, + "step": 5168 + }, + { + "epoch": 0.16303422173158807, + "grad_norm": 5.576913833618164, + "learning_rate": 1.9086786383535614e-05, + "loss": 0.5659, + "step": 5169 + }, + { + "epoch": 0.1630657624980287, + "grad_norm": 2.7069029808044434, + "learning_rate": 1.9086359840105077e-05, + "loss": 0.5837, + "step": 5170 + }, + { + "epoch": 0.16309730326446933, + "grad_norm": 2.295731782913208, + "learning_rate": 1.9085933201851316e-05, + "loss": 0.5889, + "step": 5171 + }, + { + "epoch": 0.16312884403090996, + "grad_norm": 2.425942897796631, + "learning_rate": 1.9085506468778786e-05, + "loss": 0.5721, + "step": 5172 + }, + { + "epoch": 0.16316038479735057, + "grad_norm": 2.030714511871338, + "learning_rate": 1.9085079640891942e-05, + "loss": 0.5586, + "step": 5173 + }, + { + "epoch": 0.1631919255637912, + "grad_norm": 2.1147031784057617, + "learning_rate": 1.9084652718195237e-05, + "loss": 0.5819, + "step": 5174 + }, + { + "epoch": 0.16322346633023183, + "grad_norm": 2.472848415374756, + "learning_rate": 1.908422570069313e-05, + "loss": 0.5981, + "step": 5175 + }, + { + "epoch": 0.16325500709667246, + "grad_norm": 3.3114020824432373, + "learning_rate": 1.908379858839007e-05, + "loss": 0.5676, + "step": 5176 + }, + { + "epoch": 0.16328654786311309, + "grad_norm": 2.687126636505127, + "learning_rate": 1.9083371381290524e-05, + "loss": 0.5964, + "step": 5177 + }, + { + "epoch": 0.1633180886295537, + "grad_norm": 2.9938011169433594, + "learning_rate": 1.908294407939894e-05, + "loss": 0.5887, + "step": 5178 + }, + { + "epoch": 0.16334962939599432, + "grad_norm": 3.5451791286468506, + "learning_rate": 1.9082516682719783e-05, + "loss": 0.5863, + "step": 5179 + }, + { + "epoch": 0.16338117016243495, + "grad_norm": 2.572409152984619, + "learning_rate": 1.908208919125751e-05, + "loss": 0.9627, + "step": 5180 + }, + { + "epoch": 0.16341271092887558, + "grad_norm": 20.913278579711914, + "learning_rate": 1.9081661605016587e-05, + "loss": 0.5931, + "step": 5181 + }, + { + "epoch": 0.1634442516953162, + "grad_norm": 2.4577720165252686, + "learning_rate": 1.9081233924001472e-05, + "loss": 0.5949, + "step": 5182 + }, + { + "epoch": 0.1634757924617568, + "grad_norm": 2.6134629249572754, + "learning_rate": 1.9080806148216632e-05, + "loss": 0.5819, + "step": 5183 + }, + { + "epoch": 0.16350733322819744, + "grad_norm": 2.327219009399414, + "learning_rate": 1.9080378277666525e-05, + "loss": 0.9417, + "step": 5184 + }, + { + "epoch": 0.16353887399463807, + "grad_norm": 2.114867687225342, + "learning_rate": 1.9079950312355624e-05, + "loss": 0.9707, + "step": 5185 + }, + { + "epoch": 0.1635704147610787, + "grad_norm": 2.5663418769836426, + "learning_rate": 1.9079522252288387e-05, + "loss": 0.5781, + "step": 5186 + }, + { + "epoch": 0.16360195552751933, + "grad_norm": 3.8084239959716797, + "learning_rate": 1.9079094097469286e-05, + "loss": 0.5385, + "step": 5187 + }, + { + "epoch": 0.16363349629395996, + "grad_norm": 2.757359266281128, + "learning_rate": 1.9078665847902794e-05, + "loss": 0.6001, + "step": 5188 + }, + { + "epoch": 0.16366503706040056, + "grad_norm": 2.055025100708008, + "learning_rate": 1.9078237503593367e-05, + "loss": 0.5723, + "step": 5189 + }, + { + "epoch": 0.1636965778268412, + "grad_norm": 2.4602227210998535, + "learning_rate": 1.9077809064545486e-05, + "loss": 0.6041, + "step": 5190 + }, + { + "epoch": 0.16372811859328182, + "grad_norm": 2.7288737297058105, + "learning_rate": 1.9077380530763618e-05, + "loss": 0.5817, + "step": 5191 + }, + { + "epoch": 0.16375965935972245, + "grad_norm": 2.5361576080322266, + "learning_rate": 1.9076951902252238e-05, + "loss": 0.573, + "step": 5192 + }, + { + "epoch": 0.16379120012616308, + "grad_norm": 2.388950824737549, + "learning_rate": 1.9076523179015815e-05, + "loss": 0.5931, + "step": 5193 + }, + { + "epoch": 0.16382274089260368, + "grad_norm": 2.3635330200195312, + "learning_rate": 1.9076094361058824e-05, + "loss": 0.5679, + "step": 5194 + }, + { + "epoch": 0.1638542816590443, + "grad_norm": 2.1796927452087402, + "learning_rate": 1.907566544838574e-05, + "loss": 0.5604, + "step": 5195 + }, + { + "epoch": 0.16388582242548494, + "grad_norm": 2.120636224746704, + "learning_rate": 1.9075236441001045e-05, + "loss": 0.5957, + "step": 5196 + }, + { + "epoch": 0.16391736319192557, + "grad_norm": 3.854348659515381, + "learning_rate": 1.9074807338909204e-05, + "loss": 0.9229, + "step": 5197 + }, + { + "epoch": 0.1639489039583662, + "grad_norm": 2.5880072116851807, + "learning_rate": 1.9074378142114708e-05, + "loss": 0.6185, + "step": 5198 + }, + { + "epoch": 0.1639804447248068, + "grad_norm": 2.178737163543701, + "learning_rate": 1.9073948850622028e-05, + "loss": 0.6098, + "step": 5199 + }, + { + "epoch": 0.16401198549124743, + "grad_norm": 2.8972907066345215, + "learning_rate": 1.9073519464435644e-05, + "loss": 0.6024, + "step": 5200 + }, + { + "epoch": 0.16404352625768806, + "grad_norm": 1.8961021900177002, + "learning_rate": 1.9073089983560044e-05, + "loss": 0.5687, + "step": 5201 + }, + { + "epoch": 0.1640750670241287, + "grad_norm": 4.193877696990967, + "learning_rate": 1.90726604079997e-05, + "loss": 0.5706, + "step": 5202 + }, + { + "epoch": 0.16410660779056932, + "grad_norm": 2.6126086711883545, + "learning_rate": 1.9072230737759103e-05, + "loss": 0.5765, + "step": 5203 + }, + { + "epoch": 0.16413814855700992, + "grad_norm": 2.1526217460632324, + "learning_rate": 1.907180097284273e-05, + "loss": 0.5722, + "step": 5204 + }, + { + "epoch": 0.16416968932345055, + "grad_norm": 2.0410149097442627, + "learning_rate": 1.9071371113255074e-05, + "loss": 0.916, + "step": 5205 + }, + { + "epoch": 0.16420123008989118, + "grad_norm": 2.4688451290130615, + "learning_rate": 1.907094115900061e-05, + "loss": 0.5681, + "step": 5206 + }, + { + "epoch": 0.1642327708563318, + "grad_norm": 2.1460442543029785, + "learning_rate": 1.907051111008384e-05, + "loss": 0.5629, + "step": 5207 + }, + { + "epoch": 0.16426431162277244, + "grad_norm": 2.174999952316284, + "learning_rate": 1.907008096650924e-05, + "loss": 0.5648, + "step": 5208 + }, + { + "epoch": 0.16429585238921307, + "grad_norm": 2.5704195499420166, + "learning_rate": 1.90696507282813e-05, + "loss": 0.548, + "step": 5209 + }, + { + "epoch": 0.16432739315565367, + "grad_norm": 2.453192710876465, + "learning_rate": 1.906922039540452e-05, + "loss": 0.615, + "step": 5210 + }, + { + "epoch": 0.1643589339220943, + "grad_norm": 4.048333644866943, + "learning_rate": 1.9068789967883372e-05, + "loss": 0.5542, + "step": 5211 + }, + { + "epoch": 0.16439047468853493, + "grad_norm": 2.4219446182250977, + "learning_rate": 1.9068359445722367e-05, + "loss": 0.5689, + "step": 5212 + }, + { + "epoch": 0.16442201545497556, + "grad_norm": 3.4027984142303467, + "learning_rate": 1.9067928828925986e-05, + "loss": 0.5383, + "step": 5213 + }, + { + "epoch": 0.1644535562214162, + "grad_norm": 2.8154172897338867, + "learning_rate": 1.9067498117498728e-05, + "loss": 0.5653, + "step": 5214 + }, + { + "epoch": 0.1644850969878568, + "grad_norm": 2.0924558639526367, + "learning_rate": 1.906706731144509e-05, + "loss": 0.5718, + "step": 5215 + }, + { + "epoch": 0.16451663775429742, + "grad_norm": 3.0304667949676514, + "learning_rate": 1.9066636410769556e-05, + "loss": 0.4471, + "step": 5216 + }, + { + "epoch": 0.16454817852073805, + "grad_norm": 2.3734498023986816, + "learning_rate": 1.9066205415476638e-05, + "loss": 0.5756, + "step": 5217 + }, + { + "epoch": 0.16457971928717868, + "grad_norm": 2.4128658771514893, + "learning_rate": 1.9065774325570824e-05, + "loss": 0.5761, + "step": 5218 + }, + { + "epoch": 0.1646112600536193, + "grad_norm": 2.771165132522583, + "learning_rate": 1.9065343141056616e-05, + "loss": 0.567, + "step": 5219 + }, + { + "epoch": 0.16464280082005991, + "grad_norm": 1.9619263410568237, + "learning_rate": 1.9064911861938512e-05, + "loss": 0.9572, + "step": 5220 + }, + { + "epoch": 0.16467434158650054, + "grad_norm": 1.8557617664337158, + "learning_rate": 1.9064480488221018e-05, + "loss": 0.5591, + "step": 5221 + }, + { + "epoch": 0.16470588235294117, + "grad_norm": 2.0244030952453613, + "learning_rate": 1.906404901990863e-05, + "loss": 0.5821, + "step": 5222 + }, + { + "epoch": 0.1647374231193818, + "grad_norm": 2.0603864192962646, + "learning_rate": 1.906361745700585e-05, + "loss": 0.5684, + "step": 5223 + }, + { + "epoch": 0.16476896388582243, + "grad_norm": 2.6898741722106934, + "learning_rate": 1.9063185799517187e-05, + "loss": 0.58, + "step": 5224 + }, + { + "epoch": 0.16480050465226306, + "grad_norm": 4.446715831756592, + "learning_rate": 1.9062754047447137e-05, + "loss": 0.9369, + "step": 5225 + }, + { + "epoch": 0.16483204541870367, + "grad_norm": 3.3686039447784424, + "learning_rate": 1.9062322200800215e-05, + "loss": 0.6045, + "step": 5226 + }, + { + "epoch": 0.1648635861851443, + "grad_norm": 3.0783467292785645, + "learning_rate": 1.9061890259580925e-05, + "loss": 0.6119, + "step": 5227 + }, + { + "epoch": 0.16489512695158492, + "grad_norm": 1.6341643333435059, + "learning_rate": 1.9061458223793778e-05, + "loss": 0.9346, + "step": 5228 + }, + { + "epoch": 0.16492666771802555, + "grad_norm": 1.734567403793335, + "learning_rate": 1.9061026093443272e-05, + "loss": 0.87, + "step": 5229 + }, + { + "epoch": 0.16495820848446618, + "grad_norm": 1.9806628227233887, + "learning_rate": 1.9060593868533926e-05, + "loss": 0.4485, + "step": 5230 + }, + { + "epoch": 0.1649897492509068, + "grad_norm": 9.994633674621582, + "learning_rate": 1.9060161549070244e-05, + "loss": 0.5931, + "step": 5231 + }, + { + "epoch": 0.16502129001734742, + "grad_norm": 1.916032314300537, + "learning_rate": 1.9059729135056746e-05, + "loss": 0.4441, + "step": 5232 + }, + { + "epoch": 0.16505283078378805, + "grad_norm": 6.780179977416992, + "learning_rate": 1.905929662649794e-05, + "loss": 0.5839, + "step": 5233 + }, + { + "epoch": 0.16508437155022868, + "grad_norm": 2.6539931297302246, + "learning_rate": 1.9058864023398333e-05, + "loss": 0.5739, + "step": 5234 + }, + { + "epoch": 0.1651159123166693, + "grad_norm": 3.2139620780944824, + "learning_rate": 1.9058431325762453e-05, + "loss": 0.5664, + "step": 5235 + }, + { + "epoch": 0.1651474530831099, + "grad_norm": 1.7413698434829712, + "learning_rate": 1.9057998533594803e-05, + "loss": 0.5507, + "step": 5236 + }, + { + "epoch": 0.16517899384955054, + "grad_norm": 1.835086464881897, + "learning_rate": 1.905756564689991e-05, + "loss": 0.5464, + "step": 5237 + }, + { + "epoch": 0.16521053461599117, + "grad_norm": 2.2034833431243896, + "learning_rate": 1.905713266568228e-05, + "loss": 0.5989, + "step": 5238 + }, + { + "epoch": 0.1652420753824318, + "grad_norm": 2.608813762664795, + "learning_rate": 1.9056699589946444e-05, + "loss": 0.5574, + "step": 5239 + }, + { + "epoch": 0.16527361614887243, + "grad_norm": 1.8143872022628784, + "learning_rate": 1.905626641969691e-05, + "loss": 0.58, + "step": 5240 + }, + { + "epoch": 0.16530515691531303, + "grad_norm": 2.6316189765930176, + "learning_rate": 1.9055833154938208e-05, + "loss": 0.5209, + "step": 5241 + }, + { + "epoch": 0.16533669768175366, + "grad_norm": 6.798625469207764, + "learning_rate": 1.9055399795674857e-05, + "loss": 0.5545, + "step": 5242 + }, + { + "epoch": 0.1653682384481943, + "grad_norm": 2.146449089050293, + "learning_rate": 1.9054966341911375e-05, + "loss": 0.5667, + "step": 5243 + }, + { + "epoch": 0.16539977921463492, + "grad_norm": 2.167837142944336, + "learning_rate": 1.9054532793652287e-05, + "loss": 0.5726, + "step": 5244 + }, + { + "epoch": 0.16543131998107555, + "grad_norm": 3.230726480484009, + "learning_rate": 1.905409915090212e-05, + "loss": 0.5707, + "step": 5245 + }, + { + "epoch": 0.16546286074751618, + "grad_norm": 3.294827461242676, + "learning_rate": 1.9053665413665397e-05, + "loss": 0.6123, + "step": 5246 + }, + { + "epoch": 0.16549440151395678, + "grad_norm": 2.287404775619507, + "learning_rate": 1.905323158194665e-05, + "loss": 0.5907, + "step": 5247 + }, + { + "epoch": 0.1655259422803974, + "grad_norm": 2.9126172065734863, + "learning_rate": 1.9052797655750394e-05, + "loss": 0.5491, + "step": 5248 + }, + { + "epoch": 0.16555748304683804, + "grad_norm": 2.403320550918579, + "learning_rate": 1.905236363508117e-05, + "loss": 0.5835, + "step": 5249 + }, + { + "epoch": 0.16558902381327867, + "grad_norm": 1.879048228263855, + "learning_rate": 1.9051929519943503e-05, + "loss": 0.6008, + "step": 5250 + }, + { + "epoch": 0.1656205645797193, + "grad_norm": 2.7454681396484375, + "learning_rate": 1.905149531034192e-05, + "loss": 0.6089, + "step": 5251 + }, + { + "epoch": 0.1656521053461599, + "grad_norm": 2.4899377822875977, + "learning_rate": 1.9051061006280957e-05, + "loss": 0.5537, + "step": 5252 + }, + { + "epoch": 0.16568364611260053, + "grad_norm": 3.109208106994629, + "learning_rate": 1.9050626607765145e-05, + "loss": 0.5552, + "step": 5253 + }, + { + "epoch": 0.16571518687904116, + "grad_norm": 2.0285303592681885, + "learning_rate": 1.9050192114799016e-05, + "loss": 0.9576, + "step": 5254 + }, + { + "epoch": 0.1657467276454818, + "grad_norm": 2.3052620887756348, + "learning_rate": 1.9049757527387105e-05, + "loss": 0.5986, + "step": 5255 + }, + { + "epoch": 0.16577826841192242, + "grad_norm": 2.4886293411254883, + "learning_rate": 1.9049322845533947e-05, + "loss": 0.5938, + "step": 5256 + }, + { + "epoch": 0.16580980917836302, + "grad_norm": 1.802199125289917, + "learning_rate": 1.9048888069244076e-05, + "loss": 0.9311, + "step": 5257 + }, + { + "epoch": 0.16584134994480365, + "grad_norm": 2.6496572494506836, + "learning_rate": 1.9048453198522036e-05, + "loss": 0.5739, + "step": 5258 + }, + { + "epoch": 0.16587289071124428, + "grad_norm": 2.0652272701263428, + "learning_rate": 1.904801823337236e-05, + "loss": 0.5665, + "step": 5259 + }, + { + "epoch": 0.1659044314776849, + "grad_norm": 2.915303945541382, + "learning_rate": 1.9047583173799587e-05, + "loss": 0.5792, + "step": 5260 + }, + { + "epoch": 0.16593597224412554, + "grad_norm": 2.2645115852355957, + "learning_rate": 1.904714801980826e-05, + "loss": 0.5573, + "step": 5261 + }, + { + "epoch": 0.16596751301056617, + "grad_norm": 3.5027658939361572, + "learning_rate": 1.9046712771402918e-05, + "loss": 0.439, + "step": 5262 + }, + { + "epoch": 0.16599905377700677, + "grad_norm": 1.7174893617630005, + "learning_rate": 1.90462774285881e-05, + "loss": 0.5769, + "step": 5263 + }, + { + "epoch": 0.1660305945434474, + "grad_norm": 1.5380065441131592, + "learning_rate": 1.9045841991368356e-05, + "loss": 0.3839, + "step": 5264 + }, + { + "epoch": 0.16606213530988803, + "grad_norm": 2.1546826362609863, + "learning_rate": 1.904540645974823e-05, + "loss": 0.5605, + "step": 5265 + }, + { + "epoch": 0.16609367607632866, + "grad_norm": 3.2179152965545654, + "learning_rate": 1.904497083373226e-05, + "loss": 0.5537, + "step": 5266 + }, + { + "epoch": 0.1661252168427693, + "grad_norm": 2.121579170227051, + "learning_rate": 1.9044535113324995e-05, + "loss": 0.5795, + "step": 5267 + }, + { + "epoch": 0.1661567576092099, + "grad_norm": 2.473846673965454, + "learning_rate": 1.9044099298530986e-05, + "loss": 0.5746, + "step": 5268 + }, + { + "epoch": 0.16618829837565052, + "grad_norm": 2.1892905235290527, + "learning_rate": 1.9043663389354775e-05, + "loss": 0.5863, + "step": 5269 + }, + { + "epoch": 0.16621983914209115, + "grad_norm": 2.0441620349884033, + "learning_rate": 1.9043227385800917e-05, + "loss": 0.4296, + "step": 5270 + }, + { + "epoch": 0.16625137990853178, + "grad_norm": 2.3905184268951416, + "learning_rate": 1.9042791287873958e-05, + "loss": 0.5566, + "step": 5271 + }, + { + "epoch": 0.1662829206749724, + "grad_norm": 2.7917075157165527, + "learning_rate": 1.9042355095578452e-05, + "loss": 0.596, + "step": 5272 + }, + { + "epoch": 0.16631446144141301, + "grad_norm": 3.652669668197632, + "learning_rate": 1.904191880891895e-05, + "loss": 0.5783, + "step": 5273 + }, + { + "epoch": 0.16634600220785364, + "grad_norm": 2.0195538997650146, + "learning_rate": 1.9041482427900003e-05, + "loss": 0.8826, + "step": 5274 + }, + { + "epoch": 0.16637754297429427, + "grad_norm": 2.62011981010437, + "learning_rate": 1.9041045952526163e-05, + "loss": 0.5864, + "step": 5275 + }, + { + "epoch": 0.1664090837407349, + "grad_norm": 2.5866689682006836, + "learning_rate": 1.9040609382801992e-05, + "loss": 0.5878, + "step": 5276 + }, + { + "epoch": 0.16644062450717553, + "grad_norm": 2.010509729385376, + "learning_rate": 1.904017271873204e-05, + "loss": 0.5819, + "step": 5277 + }, + { + "epoch": 0.16647216527361614, + "grad_norm": 2.5355615615844727, + "learning_rate": 1.9039735960320872e-05, + "loss": 0.6108, + "step": 5278 + }, + { + "epoch": 0.16650370604005676, + "grad_norm": 2.7313780784606934, + "learning_rate": 1.903929910757304e-05, + "loss": 0.6134, + "step": 5279 + }, + { + "epoch": 0.1665352468064974, + "grad_norm": 4.067015171051025, + "learning_rate": 1.9038862160493093e-05, + "loss": 0.5814, + "step": 5280 + }, + { + "epoch": 0.16656678757293802, + "grad_norm": 2.270120143890381, + "learning_rate": 1.9038425119085608e-05, + "loss": 0.5917, + "step": 5281 + }, + { + "epoch": 0.16659832833937865, + "grad_norm": 2.059863328933716, + "learning_rate": 1.903798798335514e-05, + "loss": 0.5695, + "step": 5282 + }, + { + "epoch": 0.16662986910581928, + "grad_norm": 2.062089204788208, + "learning_rate": 1.9037550753306243e-05, + "loss": 0.9222, + "step": 5283 + }, + { + "epoch": 0.16666140987225989, + "grad_norm": 2.4433107376098633, + "learning_rate": 1.9037113428943495e-05, + "loss": 0.5793, + "step": 5284 + }, + { + "epoch": 0.16669295063870052, + "grad_norm": 3.373746395111084, + "learning_rate": 1.9036676010271445e-05, + "loss": 0.5835, + "step": 5285 + }, + { + "epoch": 0.16672449140514115, + "grad_norm": 2.212773084640503, + "learning_rate": 1.9036238497294666e-05, + "loss": 0.4088, + "step": 5286 + }, + { + "epoch": 0.16675603217158178, + "grad_norm": 2.3485610485076904, + "learning_rate": 1.9035800890017723e-05, + "loss": 0.5447, + "step": 5287 + }, + { + "epoch": 0.1667875729380224, + "grad_norm": 2.5881969928741455, + "learning_rate": 1.903536318844518e-05, + "loss": 0.5904, + "step": 5288 + }, + { + "epoch": 0.166819113704463, + "grad_norm": 2.7199478149414062, + "learning_rate": 1.9034925392581605e-05, + "loss": 0.6205, + "step": 5289 + }, + { + "epoch": 0.16685065447090364, + "grad_norm": 3.038517475128174, + "learning_rate": 1.903448750243157e-05, + "loss": 0.5691, + "step": 5290 + }, + { + "epoch": 0.16688219523734427, + "grad_norm": 1.9248241186141968, + "learning_rate": 1.9034049517999643e-05, + "loss": 0.593, + "step": 5291 + }, + { + "epoch": 0.1669137360037849, + "grad_norm": 3.828073024749756, + "learning_rate": 1.9033611439290392e-05, + "loss": 0.5673, + "step": 5292 + }, + { + "epoch": 0.16694527677022553, + "grad_norm": 2.606973648071289, + "learning_rate": 1.9033173266308394e-05, + "loss": 0.5919, + "step": 5293 + }, + { + "epoch": 0.16697681753666613, + "grad_norm": 3.9576172828674316, + "learning_rate": 1.903273499905822e-05, + "loss": 0.5912, + "step": 5294 + }, + { + "epoch": 0.16700835830310676, + "grad_norm": 2.309753894805908, + "learning_rate": 1.903229663754444e-05, + "loss": 0.5783, + "step": 5295 + }, + { + "epoch": 0.1670398990695474, + "grad_norm": 2.2829360961914062, + "learning_rate": 1.9031858181771632e-05, + "loss": 0.5761, + "step": 5296 + }, + { + "epoch": 0.16707143983598802, + "grad_norm": 2.892585277557373, + "learning_rate": 1.903141963174437e-05, + "loss": 0.5728, + "step": 5297 + }, + { + "epoch": 0.16710298060242865, + "grad_norm": 2.4209675788879395, + "learning_rate": 1.9030980987467232e-05, + "loss": 0.6236, + "step": 5298 + }, + { + "epoch": 0.16713452136886928, + "grad_norm": 3.019749402999878, + "learning_rate": 1.9030542248944794e-05, + "loss": 0.6108, + "step": 5299 + }, + { + "epoch": 0.16716606213530988, + "grad_norm": 2.4164416790008545, + "learning_rate": 1.9030103416181637e-05, + "loss": 0.5666, + "step": 5300 + }, + { + "epoch": 0.1671976029017505, + "grad_norm": 2.6214828491210938, + "learning_rate": 1.902966448918234e-05, + "loss": 0.6007, + "step": 5301 + }, + { + "epoch": 0.16722914366819114, + "grad_norm": 2.539414405822754, + "learning_rate": 1.9029225467951482e-05, + "loss": 0.4416, + "step": 5302 + }, + { + "epoch": 0.16726068443463177, + "grad_norm": 2.611039876937866, + "learning_rate": 1.9028786352493643e-05, + "loss": 0.5566, + "step": 5303 + }, + { + "epoch": 0.1672922252010724, + "grad_norm": 2.893799066543579, + "learning_rate": 1.902834714281341e-05, + "loss": 0.5908, + "step": 5304 + }, + { + "epoch": 0.167323765967513, + "grad_norm": 2.21822190284729, + "learning_rate": 1.9027907838915363e-05, + "loss": 0.4622, + "step": 5305 + }, + { + "epoch": 0.16735530673395363, + "grad_norm": 2.6424882411956787, + "learning_rate": 1.902746844080409e-05, + "loss": 0.5736, + "step": 5306 + }, + { + "epoch": 0.16738684750039426, + "grad_norm": 2.70534348487854, + "learning_rate": 1.9027028948484173e-05, + "loss": 0.5429, + "step": 5307 + }, + { + "epoch": 0.1674183882668349, + "grad_norm": 2.339423656463623, + "learning_rate": 1.90265893619602e-05, + "loss": 0.933, + "step": 5308 + }, + { + "epoch": 0.16744992903327552, + "grad_norm": 2.3342020511627197, + "learning_rate": 1.9026149681236756e-05, + "loss": 0.5744, + "step": 5309 + }, + { + "epoch": 0.16748146979971612, + "grad_norm": 2.4962403774261475, + "learning_rate": 1.902570990631843e-05, + "loss": 0.5884, + "step": 5310 + }, + { + "epoch": 0.16751301056615675, + "grad_norm": 3.2818763256073, + "learning_rate": 1.902527003720982e-05, + "loss": 0.8946, + "step": 5311 + }, + { + "epoch": 0.16754455133259738, + "grad_norm": 2.327913999557495, + "learning_rate": 1.9024830073915503e-05, + "loss": 0.558, + "step": 5312 + }, + { + "epoch": 0.167576092099038, + "grad_norm": 2.442582607269287, + "learning_rate": 1.902439001644008e-05, + "loss": 0.5545, + "step": 5313 + }, + { + "epoch": 0.16760763286547864, + "grad_norm": 2.813396453857422, + "learning_rate": 1.9023949864788137e-05, + "loss": 0.5785, + "step": 5314 + }, + { + "epoch": 0.16763917363191927, + "grad_norm": 4.287252902984619, + "learning_rate": 1.9023509618964274e-05, + "loss": 0.5871, + "step": 5315 + }, + { + "epoch": 0.16767071439835987, + "grad_norm": 3.2509002685546875, + "learning_rate": 1.9023069278973075e-05, + "loss": 0.5833, + "step": 5316 + }, + { + "epoch": 0.1677022551648005, + "grad_norm": 2.788161277770996, + "learning_rate": 1.902262884481915e-05, + "loss": 0.6214, + "step": 5317 + }, + { + "epoch": 0.16773379593124113, + "grad_norm": 1.798336386680603, + "learning_rate": 1.902218831650708e-05, + "loss": 0.8964, + "step": 5318 + }, + { + "epoch": 0.16776533669768176, + "grad_norm": 2.7746541500091553, + "learning_rate": 1.9021747694041473e-05, + "loss": 0.595, + "step": 5319 + }, + { + "epoch": 0.1677968774641224, + "grad_norm": 2.4494359493255615, + "learning_rate": 1.9021306977426924e-05, + "loss": 0.9058, + "step": 5320 + }, + { + "epoch": 0.167828418230563, + "grad_norm": 5.718324184417725, + "learning_rate": 1.902086616666803e-05, + "loss": 0.5866, + "step": 5321 + }, + { + "epoch": 0.16785995899700362, + "grad_norm": 5.6397881507873535, + "learning_rate": 1.9020425261769393e-05, + "loss": 0.5576, + "step": 5322 + }, + { + "epoch": 0.16789149976344425, + "grad_norm": 2.1888670921325684, + "learning_rate": 1.9019984262735614e-05, + "loss": 0.5629, + "step": 5323 + }, + { + "epoch": 0.16792304052988488, + "grad_norm": 1.9248781204223633, + "learning_rate": 1.90195431695713e-05, + "loss": 0.5639, + "step": 5324 + }, + { + "epoch": 0.1679545812963255, + "grad_norm": 1.9750218391418457, + "learning_rate": 1.9019101982281044e-05, + "loss": 0.5931, + "step": 5325 + }, + { + "epoch": 0.1679861220627661, + "grad_norm": 3.3637142181396484, + "learning_rate": 1.901866070086946e-05, + "loss": 0.6215, + "step": 5326 + }, + { + "epoch": 0.16801766282920674, + "grad_norm": 2.4646453857421875, + "learning_rate": 1.9018219325341146e-05, + "loss": 0.5421, + "step": 5327 + }, + { + "epoch": 0.16804920359564737, + "grad_norm": 3.1465086936950684, + "learning_rate": 1.901777785570071e-05, + "loss": 0.6047, + "step": 5328 + }, + { + "epoch": 0.168080744362088, + "grad_norm": 2.8652539253234863, + "learning_rate": 1.9017336291952762e-05, + "loss": 0.5692, + "step": 5329 + }, + { + "epoch": 0.16811228512852863, + "grad_norm": 2.9704246520996094, + "learning_rate": 1.9016894634101906e-05, + "loss": 0.5723, + "step": 5330 + }, + { + "epoch": 0.16814382589496923, + "grad_norm": 3.346189260482788, + "learning_rate": 1.9016452882152755e-05, + "loss": 0.5744, + "step": 5331 + }, + { + "epoch": 0.16817536666140986, + "grad_norm": 1.9714106321334839, + "learning_rate": 1.9016011036109918e-05, + "loss": 0.8756, + "step": 5332 + }, + { + "epoch": 0.1682069074278505, + "grad_norm": 3.3605411052703857, + "learning_rate": 1.9015569095978e-05, + "loss": 0.602, + "step": 5333 + }, + { + "epoch": 0.16823844819429112, + "grad_norm": 2.9400389194488525, + "learning_rate": 1.9015127061761625e-05, + "loss": 0.5738, + "step": 5334 + }, + { + "epoch": 0.16826998896073175, + "grad_norm": 3.8815410137176514, + "learning_rate": 1.9014684933465394e-05, + "loss": 0.5495, + "step": 5335 + }, + { + "epoch": 0.16830152972717238, + "grad_norm": 2.8987550735473633, + "learning_rate": 1.9014242711093927e-05, + "loss": 0.5558, + "step": 5336 + }, + { + "epoch": 0.16833307049361299, + "grad_norm": 3.128819704055786, + "learning_rate": 1.901380039465184e-05, + "loss": 0.5749, + "step": 5337 + }, + { + "epoch": 0.16836461126005361, + "grad_norm": 2.3433947563171387, + "learning_rate": 1.9013357984143745e-05, + "loss": 0.5682, + "step": 5338 + }, + { + "epoch": 0.16839615202649424, + "grad_norm": 2.567394256591797, + "learning_rate": 1.9012915479574263e-05, + "loss": 0.5667, + "step": 5339 + }, + { + "epoch": 0.16842769279293487, + "grad_norm": 2.156235933303833, + "learning_rate": 1.901247288094801e-05, + "loss": 0.5458, + "step": 5340 + }, + { + "epoch": 0.1684592335593755, + "grad_norm": 3.2366039752960205, + "learning_rate": 1.90120301882696e-05, + "loss": 0.5851, + "step": 5341 + }, + { + "epoch": 0.1684907743258161, + "grad_norm": 3.315239906311035, + "learning_rate": 1.901158740154366e-05, + "loss": 0.6199, + "step": 5342 + }, + { + "epoch": 0.16852231509225674, + "grad_norm": 2.3710227012634277, + "learning_rate": 1.9011144520774808e-05, + "loss": 0.5834, + "step": 5343 + }, + { + "epoch": 0.16855385585869737, + "grad_norm": 2.2128145694732666, + "learning_rate": 1.9010701545967663e-05, + "loss": 0.5906, + "step": 5344 + }, + { + "epoch": 0.168585396625138, + "grad_norm": 2.8135311603546143, + "learning_rate": 1.9010258477126854e-05, + "loss": 0.5846, + "step": 5345 + }, + { + "epoch": 0.16861693739157863, + "grad_norm": 2.241569757461548, + "learning_rate": 1.9009815314257002e-05, + "loss": 0.5527, + "step": 5346 + }, + { + "epoch": 0.16864847815801923, + "grad_norm": 2.770151138305664, + "learning_rate": 1.900937205736273e-05, + "loss": 0.5494, + "step": 5347 + }, + { + "epoch": 0.16868001892445986, + "grad_norm": 2.591076612472534, + "learning_rate": 1.9008928706448663e-05, + "loss": 0.5553, + "step": 5348 + }, + { + "epoch": 0.1687115596909005, + "grad_norm": 3.0893542766571045, + "learning_rate": 1.9008485261519435e-05, + "loss": 0.4093, + "step": 5349 + }, + { + "epoch": 0.16874310045734112, + "grad_norm": 2.2668511867523193, + "learning_rate": 1.9008041722579664e-05, + "loss": 0.5625, + "step": 5350 + }, + { + "epoch": 0.16877464122378175, + "grad_norm": 2.4072418212890625, + "learning_rate": 1.9007598089633984e-05, + "loss": 0.5831, + "step": 5351 + }, + { + "epoch": 0.16880618199022238, + "grad_norm": 2.1554622650146484, + "learning_rate": 1.9007154362687027e-05, + "loss": 0.5706, + "step": 5352 + }, + { + "epoch": 0.16883772275666298, + "grad_norm": 3.809018611907959, + "learning_rate": 1.9006710541743418e-05, + "loss": 0.6176, + "step": 5353 + }, + { + "epoch": 0.1688692635231036, + "grad_norm": 3.0899577140808105, + "learning_rate": 1.9006266626807788e-05, + "loss": 0.5751, + "step": 5354 + }, + { + "epoch": 0.16890080428954424, + "grad_norm": 2.055899143218994, + "learning_rate": 1.9005822617884775e-05, + "loss": 0.5413, + "step": 5355 + }, + { + "epoch": 0.16893234505598487, + "grad_norm": 7.626235008239746, + "learning_rate": 1.900537851497901e-05, + "loss": 0.574, + "step": 5356 + }, + { + "epoch": 0.1689638858224255, + "grad_norm": 2.019932270050049, + "learning_rate": 1.9004934318095127e-05, + "loss": 0.864, + "step": 5357 + }, + { + "epoch": 0.1689954265888661, + "grad_norm": 2.816866874694824, + "learning_rate": 1.9004490027237764e-05, + "loss": 0.5849, + "step": 5358 + }, + { + "epoch": 0.16902696735530673, + "grad_norm": 2.4557337760925293, + "learning_rate": 1.9004045642411552e-05, + "loss": 0.5649, + "step": 5359 + }, + { + "epoch": 0.16905850812174736, + "grad_norm": 1.6864069700241089, + "learning_rate": 1.9003601163621136e-05, + "loss": 0.8824, + "step": 5360 + }, + { + "epoch": 0.169090048888188, + "grad_norm": 2.6876273155212402, + "learning_rate": 1.9003156590871154e-05, + "loss": 0.5833, + "step": 5361 + }, + { + "epoch": 0.16912158965462862, + "grad_norm": 2.5657854080200195, + "learning_rate": 1.9002711924166234e-05, + "loss": 0.6113, + "step": 5362 + }, + { + "epoch": 0.16915313042106922, + "grad_norm": 2.5073068141937256, + "learning_rate": 1.900226716351103e-05, + "loss": 0.5609, + "step": 5363 + }, + { + "epoch": 0.16918467118750985, + "grad_norm": 3.419414758682251, + "learning_rate": 1.9001822308910177e-05, + "loss": 0.4101, + "step": 5364 + }, + { + "epoch": 0.16921621195395048, + "grad_norm": 2.2426891326904297, + "learning_rate": 1.9001377360368315e-05, + "loss": 0.5652, + "step": 5365 + }, + { + "epoch": 0.1692477527203911, + "grad_norm": 4.148472785949707, + "learning_rate": 1.9000932317890093e-05, + "loss": 0.8818, + "step": 5366 + }, + { + "epoch": 0.16927929348683174, + "grad_norm": 4.2634968757629395, + "learning_rate": 1.900048718148016e-05, + "loss": 0.8923, + "step": 5367 + }, + { + "epoch": 0.16931083425327234, + "grad_norm": 2.9579074382781982, + "learning_rate": 1.9000041951143146e-05, + "loss": 0.5703, + "step": 5368 + }, + { + "epoch": 0.16934237501971297, + "grad_norm": 3.1828153133392334, + "learning_rate": 1.899959662688371e-05, + "loss": 0.6099, + "step": 5369 + }, + { + "epoch": 0.1693739157861536, + "grad_norm": 2.156693935394287, + "learning_rate": 1.899915120870649e-05, + "loss": 0.5716, + "step": 5370 + }, + { + "epoch": 0.16940545655259423, + "grad_norm": 2.3076467514038086, + "learning_rate": 1.8998705696616147e-05, + "loss": 0.5844, + "step": 5371 + }, + { + "epoch": 0.16943699731903486, + "grad_norm": 3.9634528160095215, + "learning_rate": 1.899826009061732e-05, + "loss": 0.5945, + "step": 5372 + }, + { + "epoch": 0.1694685380854755, + "grad_norm": 1.8944621086120605, + "learning_rate": 1.899781439071466e-05, + "loss": 0.4061, + "step": 5373 + }, + { + "epoch": 0.1695000788519161, + "grad_norm": 3.017303705215454, + "learning_rate": 1.8997368596912823e-05, + "loss": 0.5689, + "step": 5374 + }, + { + "epoch": 0.16953161961835672, + "grad_norm": 1.513225793838501, + "learning_rate": 1.8996922709216456e-05, + "loss": 0.4044, + "step": 5375 + }, + { + "epoch": 0.16956316038479735, + "grad_norm": 2.3515889644622803, + "learning_rate": 1.8996476727630214e-05, + "loss": 0.6117, + "step": 5376 + }, + { + "epoch": 0.16959470115123798, + "grad_norm": 2.3517353534698486, + "learning_rate": 1.8996030652158755e-05, + "loss": 0.5518, + "step": 5377 + }, + { + "epoch": 0.1696262419176786, + "grad_norm": 3.1282598972320557, + "learning_rate": 1.8995584482806724e-05, + "loss": 0.5856, + "step": 5378 + }, + { + "epoch": 0.1696577826841192, + "grad_norm": 2.051043748855591, + "learning_rate": 1.899513821957879e-05, + "loss": 0.5733, + "step": 5379 + }, + { + "epoch": 0.16968932345055984, + "grad_norm": 2.7429518699645996, + "learning_rate": 1.8994691862479608e-05, + "loss": 0.5739, + "step": 5380 + }, + { + "epoch": 0.16972086421700047, + "grad_norm": 2.916271924972534, + "learning_rate": 1.8994245411513825e-05, + "loss": 0.6153, + "step": 5381 + }, + { + "epoch": 0.1697524049834411, + "grad_norm": 2.466395139694214, + "learning_rate": 1.899379886668611e-05, + "loss": 0.55, + "step": 5382 + }, + { + "epoch": 0.16978394574988173, + "grad_norm": 6.813971042633057, + "learning_rate": 1.899335222800112e-05, + "loss": 0.5872, + "step": 5383 + }, + { + "epoch": 0.16981548651632233, + "grad_norm": 2.404571056365967, + "learning_rate": 1.8992905495463516e-05, + "loss": 0.5692, + "step": 5384 + }, + { + "epoch": 0.16984702728276296, + "grad_norm": 5.130895614624023, + "learning_rate": 1.899245866907796e-05, + "loss": 0.5614, + "step": 5385 + }, + { + "epoch": 0.1698785680492036, + "grad_norm": 2.4926934242248535, + "learning_rate": 1.8992011748849117e-05, + "loss": 0.5852, + "step": 5386 + }, + { + "epoch": 0.16991010881564422, + "grad_norm": 3.9908878803253174, + "learning_rate": 1.899156473478165e-05, + "loss": 0.5766, + "step": 5387 + }, + { + "epoch": 0.16994164958208485, + "grad_norm": 5.370936393737793, + "learning_rate": 1.899111762688022e-05, + "loss": 0.582, + "step": 5388 + }, + { + "epoch": 0.16997319034852548, + "grad_norm": 2.586725950241089, + "learning_rate": 1.8990670425149494e-05, + "loss": 0.5694, + "step": 5389 + }, + { + "epoch": 0.17000473111496608, + "grad_norm": 3.5914158821105957, + "learning_rate": 1.8990223129594146e-05, + "loss": 0.5809, + "step": 5390 + }, + { + "epoch": 0.17003627188140671, + "grad_norm": 2.822237253189087, + "learning_rate": 1.8989775740218837e-05, + "loss": 0.6201, + "step": 5391 + }, + { + "epoch": 0.17006781264784734, + "grad_norm": 2.3299593925476074, + "learning_rate": 1.8989328257028243e-05, + "loss": 0.5668, + "step": 5392 + }, + { + "epoch": 0.17009935341428797, + "grad_norm": 2.3649187088012695, + "learning_rate": 1.898888068002702e-05, + "loss": 0.5903, + "step": 5393 + }, + { + "epoch": 0.1701308941807286, + "grad_norm": 1.9539872407913208, + "learning_rate": 1.898843300921985e-05, + "loss": 0.5685, + "step": 5394 + }, + { + "epoch": 0.1701624349471692, + "grad_norm": 2.3395485877990723, + "learning_rate": 1.8987985244611407e-05, + "loss": 0.5588, + "step": 5395 + }, + { + "epoch": 0.17019397571360984, + "grad_norm": 4.802759170532227, + "learning_rate": 1.8987537386206355e-05, + "loss": 0.5848, + "step": 5396 + }, + { + "epoch": 0.17022551648005046, + "grad_norm": 1.9878201484680176, + "learning_rate": 1.8987089434009372e-05, + "loss": 0.5202, + "step": 5397 + }, + { + "epoch": 0.1702570572464911, + "grad_norm": 2.6069693565368652, + "learning_rate": 1.8986641388025133e-05, + "loss": 0.5783, + "step": 5398 + }, + { + "epoch": 0.17028859801293172, + "grad_norm": 2.6344926357269287, + "learning_rate": 1.898619324825831e-05, + "loss": 0.5875, + "step": 5399 + }, + { + "epoch": 0.17032013877937233, + "grad_norm": 2.8299002647399902, + "learning_rate": 1.898574501471359e-05, + "loss": 0.5811, + "step": 5400 + }, + { + "epoch": 0.17035167954581296, + "grad_norm": 2.3876938819885254, + "learning_rate": 1.8985296687395638e-05, + "loss": 0.9371, + "step": 5401 + }, + { + "epoch": 0.17038322031225359, + "grad_norm": 3.0241925716400146, + "learning_rate": 1.898484826630914e-05, + "loss": 0.5632, + "step": 5402 + }, + { + "epoch": 0.17041476107869422, + "grad_norm": 2.60986590385437, + "learning_rate": 1.8984399751458772e-05, + "loss": 0.5367, + "step": 5403 + }, + { + "epoch": 0.17044630184513485, + "grad_norm": 2.1456830501556396, + "learning_rate": 1.898395114284922e-05, + "loss": 0.5774, + "step": 5404 + }, + { + "epoch": 0.17047784261157545, + "grad_norm": 2.959012508392334, + "learning_rate": 1.898350244048516e-05, + "loss": 0.6017, + "step": 5405 + }, + { + "epoch": 0.17050938337801608, + "grad_norm": 1.6003361940383911, + "learning_rate": 1.8983053644371274e-05, + "loss": 0.3656, + "step": 5406 + }, + { + "epoch": 0.1705409241444567, + "grad_norm": 2.2805333137512207, + "learning_rate": 1.898260475451225e-05, + "loss": 0.8665, + "step": 5407 + }, + { + "epoch": 0.17057246491089734, + "grad_norm": 3.017091751098633, + "learning_rate": 1.8982155770912772e-05, + "loss": 0.5833, + "step": 5408 + }, + { + "epoch": 0.17060400567733797, + "grad_norm": 3.157573938369751, + "learning_rate": 1.898170669357752e-05, + "loss": 0.6058, + "step": 5409 + }, + { + "epoch": 0.1706355464437786, + "grad_norm": 2.1664493083953857, + "learning_rate": 1.898125752251119e-05, + "loss": 0.5872, + "step": 5410 + }, + { + "epoch": 0.1706670872102192, + "grad_norm": 2.5878188610076904, + "learning_rate": 1.8980808257718463e-05, + "loss": 0.3865, + "step": 5411 + }, + { + "epoch": 0.17069862797665983, + "grad_norm": 2.233633279800415, + "learning_rate": 1.8980358899204024e-05, + "loss": 0.5608, + "step": 5412 + }, + { + "epoch": 0.17073016874310046, + "grad_norm": 3.164433479309082, + "learning_rate": 1.897990944697257e-05, + "loss": 0.6099, + "step": 5413 + }, + { + "epoch": 0.1707617095095411, + "grad_norm": 3.0813286304473877, + "learning_rate": 1.897945990102879e-05, + "loss": 0.5777, + "step": 5414 + }, + { + "epoch": 0.17079325027598172, + "grad_norm": 4.777221202850342, + "learning_rate": 1.8979010261377372e-05, + "loss": 0.5384, + "step": 5415 + }, + { + "epoch": 0.17082479104242232, + "grad_norm": 3.254169464111328, + "learning_rate": 1.8978560528023008e-05, + "loss": 0.5816, + "step": 5416 + }, + { + "epoch": 0.17085633180886295, + "grad_norm": 1.9008816480636597, + "learning_rate": 1.8978110700970394e-05, + "loss": 0.576, + "step": 5417 + }, + { + "epoch": 0.17088787257530358, + "grad_norm": 1.9969550371170044, + "learning_rate": 1.8977660780224223e-05, + "loss": 0.3727, + "step": 5418 + }, + { + "epoch": 0.1709194133417442, + "grad_norm": 5.330421447753906, + "learning_rate": 1.8977210765789192e-05, + "loss": 0.5751, + "step": 5419 + }, + { + "epoch": 0.17095095410818484, + "grad_norm": 3.266503095626831, + "learning_rate": 1.8976760657669995e-05, + "loss": 0.6011, + "step": 5420 + }, + { + "epoch": 0.17098249487462544, + "grad_norm": 2.84000563621521, + "learning_rate": 1.897631045587133e-05, + "loss": 0.5767, + "step": 5421 + }, + { + "epoch": 0.17101403564106607, + "grad_norm": 3.9997310638427734, + "learning_rate": 1.89758601603979e-05, + "loss": 0.5828, + "step": 5422 + }, + { + "epoch": 0.1710455764075067, + "grad_norm": 3.9557480812072754, + "learning_rate": 1.8975409771254397e-05, + "loss": 0.5886, + "step": 5423 + }, + { + "epoch": 0.17107711717394733, + "grad_norm": 2.6606714725494385, + "learning_rate": 1.897495928844552e-05, + "loss": 0.5681, + "step": 5424 + }, + { + "epoch": 0.17110865794038796, + "grad_norm": 3.0278539657592773, + "learning_rate": 1.8974508711975978e-05, + "loss": 0.6042, + "step": 5425 + }, + { + "epoch": 0.1711401987068286, + "grad_norm": 2.5508925914764404, + "learning_rate": 1.897405804185047e-05, + "loss": 0.6014, + "step": 5426 + }, + { + "epoch": 0.1711717394732692, + "grad_norm": 5.850919246673584, + "learning_rate": 1.8973607278073696e-05, + "loss": 0.904, + "step": 5427 + }, + { + "epoch": 0.17120328023970982, + "grad_norm": 2.2558155059814453, + "learning_rate": 1.897315642065036e-05, + "loss": 0.574, + "step": 5428 + }, + { + "epoch": 0.17123482100615045, + "grad_norm": 3.822828769683838, + "learning_rate": 1.8972705469585172e-05, + "loss": 0.5852, + "step": 5429 + }, + { + "epoch": 0.17126636177259108, + "grad_norm": 3.1345419883728027, + "learning_rate": 1.8972254424882838e-05, + "loss": 0.5749, + "step": 5430 + }, + { + "epoch": 0.1712979025390317, + "grad_norm": 3.200054168701172, + "learning_rate": 1.8971803286548056e-05, + "loss": 0.5829, + "step": 5431 + }, + { + "epoch": 0.1713294433054723, + "grad_norm": 3.119941234588623, + "learning_rate": 1.8971352054585542e-05, + "loss": 0.589, + "step": 5432 + }, + { + "epoch": 0.17136098407191294, + "grad_norm": 3.5189507007598877, + "learning_rate": 1.897090072900001e-05, + "loss": 0.5699, + "step": 5433 + }, + { + "epoch": 0.17139252483835357, + "grad_norm": 2.5126962661743164, + "learning_rate": 1.8970449309796156e-05, + "loss": 0.5524, + "step": 5434 + }, + { + "epoch": 0.1714240656047942, + "grad_norm": 2.908229112625122, + "learning_rate": 1.89699977969787e-05, + "loss": 0.5483, + "step": 5435 + }, + { + "epoch": 0.17145560637123483, + "grad_norm": 7.930446147918701, + "learning_rate": 1.896954619055235e-05, + "loss": 0.4026, + "step": 5436 + }, + { + "epoch": 0.17148714713767543, + "grad_norm": 4.617363929748535, + "learning_rate": 1.8969094490521823e-05, + "loss": 0.5872, + "step": 5437 + }, + { + "epoch": 0.17151868790411606, + "grad_norm": 3.2122857570648193, + "learning_rate": 1.896864269689183e-05, + "loss": 0.5573, + "step": 5438 + }, + { + "epoch": 0.1715502286705567, + "grad_norm": 3.7475967407226562, + "learning_rate": 1.896819080966709e-05, + "loss": 0.5861, + "step": 5439 + }, + { + "epoch": 0.17158176943699732, + "grad_norm": 2.872791051864624, + "learning_rate": 1.896773882885231e-05, + "loss": 0.5872, + "step": 5440 + }, + { + "epoch": 0.17161331020343795, + "grad_norm": 2.3677845001220703, + "learning_rate": 1.8967286754452214e-05, + "loss": 0.582, + "step": 5441 + }, + { + "epoch": 0.17164485096987858, + "grad_norm": 2.5236051082611084, + "learning_rate": 1.8966834586471517e-05, + "loss": 0.5839, + "step": 5442 + }, + { + "epoch": 0.17167639173631918, + "grad_norm": 2.1347711086273193, + "learning_rate": 1.896638232491494e-05, + "loss": 0.5686, + "step": 5443 + }, + { + "epoch": 0.1717079325027598, + "grad_norm": 2.96783447265625, + "learning_rate": 1.89659299697872e-05, + "loss": 0.5616, + "step": 5444 + }, + { + "epoch": 0.17173947326920044, + "grad_norm": 3.7984910011291504, + "learning_rate": 1.896547752109302e-05, + "loss": 0.5665, + "step": 5445 + }, + { + "epoch": 0.17177101403564107, + "grad_norm": 3.7851579189300537, + "learning_rate": 1.896502497883712e-05, + "loss": 0.5782, + "step": 5446 + }, + { + "epoch": 0.1718025548020817, + "grad_norm": 3.2696480751037598, + "learning_rate": 1.896457234302423e-05, + "loss": 0.6048, + "step": 5447 + }, + { + "epoch": 0.1718340955685223, + "grad_norm": 2.268434762954712, + "learning_rate": 1.896411961365906e-05, + "loss": 0.557, + "step": 5448 + }, + { + "epoch": 0.17186563633496293, + "grad_norm": 2.5560462474823, + "learning_rate": 1.8963666790746337e-05, + "loss": 0.5705, + "step": 5449 + }, + { + "epoch": 0.17189717710140356, + "grad_norm": 2.472290515899658, + "learning_rate": 1.8963213874290797e-05, + "loss": 0.5692, + "step": 5450 + }, + { + "epoch": 0.1719287178678442, + "grad_norm": 2.939544439315796, + "learning_rate": 1.8962760864297157e-05, + "loss": 0.5748, + "step": 5451 + }, + { + "epoch": 0.17196025863428482, + "grad_norm": 3.0538392066955566, + "learning_rate": 1.8962307760770152e-05, + "loss": 0.5565, + "step": 5452 + }, + { + "epoch": 0.17199179940072543, + "grad_norm": 2.794520378112793, + "learning_rate": 1.8961854563714507e-05, + "loss": 0.913, + "step": 5453 + }, + { + "epoch": 0.17202334016716606, + "grad_norm": 3.422398805618286, + "learning_rate": 1.8961401273134946e-05, + "loss": 0.5873, + "step": 5454 + }, + { + "epoch": 0.17205488093360669, + "grad_norm": 1.9732929468154907, + "learning_rate": 1.8960947889036205e-05, + "loss": 0.9125, + "step": 5455 + }, + { + "epoch": 0.17208642170004732, + "grad_norm": 2.320221185684204, + "learning_rate": 1.8960494411423018e-05, + "loss": 0.8939, + "step": 5456 + }, + { + "epoch": 0.17211796246648794, + "grad_norm": 2.9920668601989746, + "learning_rate": 1.896004084030011e-05, + "loss": 0.6034, + "step": 5457 + }, + { + "epoch": 0.17214950323292855, + "grad_norm": 1.8242799043655396, + "learning_rate": 1.8959587175672226e-05, + "loss": 0.9051, + "step": 5458 + }, + { + "epoch": 0.17218104399936918, + "grad_norm": 2.401092529296875, + "learning_rate": 1.8959133417544088e-05, + "loss": 0.5499, + "step": 5459 + }, + { + "epoch": 0.1722125847658098, + "grad_norm": 3.4910190105438232, + "learning_rate": 1.8958679565920433e-05, + "loss": 0.573, + "step": 5460 + }, + { + "epoch": 0.17224412553225044, + "grad_norm": 3.0999577045440674, + "learning_rate": 1.8958225620806005e-05, + "loss": 0.9052, + "step": 5461 + }, + { + "epoch": 0.17227566629869107, + "grad_norm": 2.4958529472351074, + "learning_rate": 1.8957771582205536e-05, + "loss": 0.5533, + "step": 5462 + }, + { + "epoch": 0.1723072070651317, + "grad_norm": 1.9181935787200928, + "learning_rate": 1.8957317450123766e-05, + "loss": 0.5634, + "step": 5463 + }, + { + "epoch": 0.1723387478315723, + "grad_norm": 2.704655647277832, + "learning_rate": 1.895686322456543e-05, + "loss": 0.5511, + "step": 5464 + }, + { + "epoch": 0.17237028859801293, + "grad_norm": 15.62679672241211, + "learning_rate": 1.8956408905535272e-05, + "loss": 0.566, + "step": 5465 + }, + { + "epoch": 0.17240182936445356, + "grad_norm": 3.0706632137298584, + "learning_rate": 1.8955954493038035e-05, + "loss": 0.562, + "step": 5466 + }, + { + "epoch": 0.1724333701308942, + "grad_norm": 2.7344040870666504, + "learning_rate": 1.895549998707846e-05, + "loss": 0.5462, + "step": 5467 + }, + { + "epoch": 0.17246491089733482, + "grad_norm": 3.3160555362701416, + "learning_rate": 1.8955045387661285e-05, + "loss": 0.5581, + "step": 5468 + }, + { + "epoch": 0.17249645166377542, + "grad_norm": 4.177845478057861, + "learning_rate": 1.895459069479126e-05, + "loss": 0.5714, + "step": 5469 + }, + { + "epoch": 0.17252799243021605, + "grad_norm": 5.4935197830200195, + "learning_rate": 1.8954135908473127e-05, + "loss": 0.5637, + "step": 5470 + }, + { + "epoch": 0.17255953319665668, + "grad_norm": 2.500507116317749, + "learning_rate": 1.895368102871163e-05, + "loss": 0.9209, + "step": 5471 + }, + { + "epoch": 0.1725910739630973, + "grad_norm": 5.341735363006592, + "learning_rate": 1.8953226055511525e-05, + "loss": 0.5499, + "step": 5472 + }, + { + "epoch": 0.17262261472953794, + "grad_norm": 3.4127583503723145, + "learning_rate": 1.8952770988877553e-05, + "loss": 0.555, + "step": 5473 + }, + { + "epoch": 0.17265415549597854, + "grad_norm": 3.290567636489868, + "learning_rate": 1.8952315828814464e-05, + "loss": 0.5932, + "step": 5474 + }, + { + "epoch": 0.17268569626241917, + "grad_norm": 2.463017702102661, + "learning_rate": 1.895186057532701e-05, + "loss": 0.5898, + "step": 5475 + }, + { + "epoch": 0.1727172370288598, + "grad_norm": 2.9938719272613525, + "learning_rate": 1.8951405228419934e-05, + "loss": 0.5779, + "step": 5476 + }, + { + "epoch": 0.17274877779530043, + "grad_norm": 2.7873852252960205, + "learning_rate": 1.8950949788097997e-05, + "loss": 0.6032, + "step": 5477 + }, + { + "epoch": 0.17278031856174106, + "grad_norm": 2.4368231296539307, + "learning_rate": 1.8950494254365947e-05, + "loss": 0.5323, + "step": 5478 + }, + { + "epoch": 0.1728118593281817, + "grad_norm": 3.554675340652466, + "learning_rate": 1.895003862722854e-05, + "loss": 0.5816, + "step": 5479 + }, + { + "epoch": 0.1728434000946223, + "grad_norm": 2.106593608856201, + "learning_rate": 1.8949582906690535e-05, + "loss": 0.4642, + "step": 5480 + }, + { + "epoch": 0.17287494086106292, + "grad_norm": 3.2096853256225586, + "learning_rate": 1.894912709275668e-05, + "loss": 0.5591, + "step": 5481 + }, + { + "epoch": 0.17290648162750355, + "grad_norm": 2.6990318298339844, + "learning_rate": 1.8948671185431732e-05, + "loss": 0.5947, + "step": 5482 + }, + { + "epoch": 0.17293802239394418, + "grad_norm": 1.7697352170944214, + "learning_rate": 1.8948215184720454e-05, + "loss": 0.8781, + "step": 5483 + }, + { + "epoch": 0.1729695631603848, + "grad_norm": 2.4215266704559326, + "learning_rate": 1.8947759090627605e-05, + "loss": 0.5799, + "step": 5484 + }, + { + "epoch": 0.1730011039268254, + "grad_norm": 2.4603328704833984, + "learning_rate": 1.894730290315794e-05, + "loss": 0.5729, + "step": 5485 + }, + { + "epoch": 0.17303264469326604, + "grad_norm": 3.0181212425231934, + "learning_rate": 1.894684662231622e-05, + "loss": 0.5672, + "step": 5486 + }, + { + "epoch": 0.17306418545970667, + "grad_norm": 1.998921513557434, + "learning_rate": 1.894639024810721e-05, + "loss": 0.5531, + "step": 5487 + }, + { + "epoch": 0.1730957262261473, + "grad_norm": 2.0158326625823975, + "learning_rate": 1.894593378053567e-05, + "loss": 0.5528, + "step": 5488 + }, + { + "epoch": 0.17312726699258793, + "grad_norm": 5.758883953094482, + "learning_rate": 1.8945477219606367e-05, + "loss": 0.4049, + "step": 5489 + }, + { + "epoch": 0.17315880775902853, + "grad_norm": 2.7088844776153564, + "learning_rate": 1.894502056532406e-05, + "loss": 0.5548, + "step": 5490 + }, + { + "epoch": 0.17319034852546916, + "grad_norm": 2.6695916652679443, + "learning_rate": 1.8944563817693522e-05, + "loss": 0.5186, + "step": 5491 + }, + { + "epoch": 0.1732218892919098, + "grad_norm": 2.4089977741241455, + "learning_rate": 1.8944106976719512e-05, + "loss": 0.5593, + "step": 5492 + }, + { + "epoch": 0.17325343005835042, + "grad_norm": 3.022958755493164, + "learning_rate": 1.8943650042406806e-05, + "loss": 0.594, + "step": 5493 + }, + { + "epoch": 0.17328497082479105, + "grad_norm": 4.993942737579346, + "learning_rate": 1.8943193014760162e-05, + "loss": 0.5748, + "step": 5494 + }, + { + "epoch": 0.17331651159123165, + "grad_norm": 2.3340229988098145, + "learning_rate": 1.8942735893784358e-05, + "loss": 0.9171, + "step": 5495 + }, + { + "epoch": 0.17334805235767228, + "grad_norm": 8.183518409729004, + "learning_rate": 1.8942278679484155e-05, + "loss": 0.447, + "step": 5496 + }, + { + "epoch": 0.1733795931241129, + "grad_norm": 4.053654670715332, + "learning_rate": 1.8941821371864336e-05, + "loss": 0.5808, + "step": 5497 + }, + { + "epoch": 0.17341113389055354, + "grad_norm": 5.0253729820251465, + "learning_rate": 1.8941363970929666e-05, + "loss": 0.5421, + "step": 5498 + }, + { + "epoch": 0.17344267465699417, + "grad_norm": 2.5993404388427734, + "learning_rate": 1.894090647668492e-05, + "loss": 0.5719, + "step": 5499 + }, + { + "epoch": 0.1734742154234348, + "grad_norm": 4.674483299255371, + "learning_rate": 1.8940448889134873e-05, + "loss": 0.598, + "step": 5500 + }, + { + "epoch": 0.1735057561898754, + "grad_norm": 3.5280399322509766, + "learning_rate": 1.89399912082843e-05, + "loss": 0.417, + "step": 5501 + }, + { + "epoch": 0.17353729695631603, + "grad_norm": 3.7512590885162354, + "learning_rate": 1.8939533434137976e-05, + "loss": 0.3889, + "step": 5502 + }, + { + "epoch": 0.17356883772275666, + "grad_norm": 2.508490562438965, + "learning_rate": 1.893907556670068e-05, + "loss": 0.5565, + "step": 5503 + }, + { + "epoch": 0.1736003784891973, + "grad_norm": 3.9972920417785645, + "learning_rate": 1.893861760597719e-05, + "loss": 0.5686, + "step": 5504 + }, + { + "epoch": 0.17363191925563792, + "grad_norm": 2.9104597568511963, + "learning_rate": 1.8938159551972285e-05, + "loss": 0.591, + "step": 5505 + }, + { + "epoch": 0.17366346002207853, + "grad_norm": 3.0501210689544678, + "learning_rate": 1.8937701404690742e-05, + "loss": 0.5919, + "step": 5506 + }, + { + "epoch": 0.17369500078851915, + "grad_norm": 1.8680543899536133, + "learning_rate": 1.8937243164137346e-05, + "loss": 0.9049, + "step": 5507 + }, + { + "epoch": 0.17372654155495978, + "grad_norm": 4.004834175109863, + "learning_rate": 1.8936784830316882e-05, + "loss": 0.559, + "step": 5508 + }, + { + "epoch": 0.17375808232140041, + "grad_norm": 2.205561399459839, + "learning_rate": 1.8936326403234125e-05, + "loss": 0.5823, + "step": 5509 + }, + { + "epoch": 0.17378962308784104, + "grad_norm": 2.6108899116516113, + "learning_rate": 1.8935867882893865e-05, + "loss": 0.5965, + "step": 5510 + }, + { + "epoch": 0.17382116385428165, + "grad_norm": 3.414098024368286, + "learning_rate": 1.893540926930088e-05, + "loss": 0.5862, + "step": 5511 + }, + { + "epoch": 0.17385270462072228, + "grad_norm": 3.104675531387329, + "learning_rate": 1.8934950562459968e-05, + "loss": 0.5825, + "step": 5512 + }, + { + "epoch": 0.1738842453871629, + "grad_norm": 5.947774410247803, + "learning_rate": 1.8934491762375902e-05, + "loss": 0.4176, + "step": 5513 + }, + { + "epoch": 0.17391578615360354, + "grad_norm": 3.247823715209961, + "learning_rate": 1.8934032869053486e-05, + "loss": 0.5798, + "step": 5514 + }, + { + "epoch": 0.17394732692004417, + "grad_norm": 5.863234519958496, + "learning_rate": 1.8933573882497492e-05, + "loss": 0.6132, + "step": 5515 + }, + { + "epoch": 0.1739788676864848, + "grad_norm": 2.20123553276062, + "learning_rate": 1.8933114802712726e-05, + "loss": 0.5245, + "step": 5516 + }, + { + "epoch": 0.1740104084529254, + "grad_norm": 2.4893181324005127, + "learning_rate": 1.893265562970396e-05, + "loss": 0.5542, + "step": 5517 + }, + { + "epoch": 0.17404194921936603, + "grad_norm": 2.3705837726593018, + "learning_rate": 1.8932196363476004e-05, + "loss": 0.5572, + "step": 5518 + }, + { + "epoch": 0.17407348998580666, + "grad_norm": 2.680161952972412, + "learning_rate": 1.893173700403364e-05, + "loss": 0.9412, + "step": 5519 + }, + { + "epoch": 0.1741050307522473, + "grad_norm": 2.806061029434204, + "learning_rate": 1.8931277551381667e-05, + "loss": 0.5953, + "step": 5520 + }, + { + "epoch": 0.17413657151868792, + "grad_norm": 2.8569860458374023, + "learning_rate": 1.893081800552488e-05, + "loss": 0.5666, + "step": 5521 + }, + { + "epoch": 0.17416811228512852, + "grad_norm": 2.8511056900024414, + "learning_rate": 1.893035836646807e-05, + "loss": 0.5632, + "step": 5522 + }, + { + "epoch": 0.17419965305156915, + "grad_norm": 6.4566216468811035, + "learning_rate": 1.8929898634216032e-05, + "loss": 0.8792, + "step": 5523 + }, + { + "epoch": 0.17423119381800978, + "grad_norm": 2.5433883666992188, + "learning_rate": 1.8929438808773573e-05, + "loss": 0.9224, + "step": 5524 + }, + { + "epoch": 0.1742627345844504, + "grad_norm": 3.1549177169799805, + "learning_rate": 1.8928978890145485e-05, + "loss": 0.5758, + "step": 5525 + }, + { + "epoch": 0.17429427535089104, + "grad_norm": 2.836448907852173, + "learning_rate": 1.892851887833657e-05, + "loss": 0.5833, + "step": 5526 + }, + { + "epoch": 0.17432581611733164, + "grad_norm": 2.9408602714538574, + "learning_rate": 1.8928058773351626e-05, + "loss": 0.5797, + "step": 5527 + }, + { + "epoch": 0.17435735688377227, + "grad_norm": 2.9969677925109863, + "learning_rate": 1.892759857519546e-05, + "loss": 0.5729, + "step": 5528 + }, + { + "epoch": 0.1743888976502129, + "grad_norm": 2.5925705432891846, + "learning_rate": 1.8927138283872865e-05, + "loss": 0.5455, + "step": 5529 + }, + { + "epoch": 0.17442043841665353, + "grad_norm": 6.2068915367126465, + "learning_rate": 1.8926677899388653e-05, + "loss": 0.9326, + "step": 5530 + }, + { + "epoch": 0.17445197918309416, + "grad_norm": 2.9016950130462646, + "learning_rate": 1.8926217421747627e-05, + "loss": 0.8832, + "step": 5531 + }, + { + "epoch": 0.17448351994953476, + "grad_norm": 3.3654441833496094, + "learning_rate": 1.8925756850954592e-05, + "loss": 0.5669, + "step": 5532 + }, + { + "epoch": 0.1745150607159754, + "grad_norm": 3.2551956176757812, + "learning_rate": 1.892529618701435e-05, + "loss": 0.5853, + "step": 5533 + }, + { + "epoch": 0.17454660148241602, + "grad_norm": 8.077231407165527, + "learning_rate": 1.8924835429931713e-05, + "loss": 0.5969, + "step": 5534 + }, + { + "epoch": 0.17457814224885665, + "grad_norm": 2.3292315006256104, + "learning_rate": 1.8924374579711487e-05, + "loss": 0.6044, + "step": 5535 + }, + { + "epoch": 0.17460968301529728, + "grad_norm": 2.571709394454956, + "learning_rate": 1.8923913636358485e-05, + "loss": 0.5716, + "step": 5536 + }, + { + "epoch": 0.1746412237817379, + "grad_norm": 3.959526300430298, + "learning_rate": 1.8923452599877517e-05, + "loss": 0.6012, + "step": 5537 + }, + { + "epoch": 0.1746727645481785, + "grad_norm": 2.8080387115478516, + "learning_rate": 1.892299147027339e-05, + "loss": 0.5758, + "step": 5538 + }, + { + "epoch": 0.17470430531461914, + "grad_norm": 2.6590147018432617, + "learning_rate": 1.8922530247550918e-05, + "loss": 0.5962, + "step": 5539 + }, + { + "epoch": 0.17473584608105977, + "grad_norm": 14.381475448608398, + "learning_rate": 1.8922068931714916e-05, + "loss": 0.9264, + "step": 5540 + }, + { + "epoch": 0.1747673868475004, + "grad_norm": 2.901237726211548, + "learning_rate": 1.8921607522770196e-05, + "loss": 0.5808, + "step": 5541 + }, + { + "epoch": 0.17479892761394103, + "grad_norm": 2.322772264480591, + "learning_rate": 1.8921146020721573e-05, + "loss": 0.5373, + "step": 5542 + }, + { + "epoch": 0.17483046838038163, + "grad_norm": 3.3507676124572754, + "learning_rate": 1.8920684425573865e-05, + "loss": 0.5913, + "step": 5543 + }, + { + "epoch": 0.17486200914682226, + "grad_norm": 2.522913932800293, + "learning_rate": 1.892022273733189e-05, + "loss": 0.5711, + "step": 5544 + }, + { + "epoch": 0.1748935499132629, + "grad_norm": 2.3331727981567383, + "learning_rate": 1.8919760956000465e-05, + "loss": 0.6073, + "step": 5545 + }, + { + "epoch": 0.17492509067970352, + "grad_norm": 2.3650474548339844, + "learning_rate": 1.8919299081584405e-05, + "loss": 0.5601, + "step": 5546 + }, + { + "epoch": 0.17495663144614415, + "grad_norm": 2.330075263977051, + "learning_rate": 1.891883711408853e-05, + "loss": 0.6017, + "step": 5547 + }, + { + "epoch": 0.17498817221258475, + "grad_norm": 5.358626842498779, + "learning_rate": 1.8918375053517676e-05, + "loss": 0.5882, + "step": 5548 + }, + { + "epoch": 0.17501971297902538, + "grad_norm": 1.921399712562561, + "learning_rate": 1.8917912899876643e-05, + "loss": 0.5706, + "step": 5549 + }, + { + "epoch": 0.175051253745466, + "grad_norm": 2.4164674282073975, + "learning_rate": 1.891745065317027e-05, + "loss": 0.6058, + "step": 5550 + }, + { + "epoch": 0.17508279451190664, + "grad_norm": 3.5850021839141846, + "learning_rate": 1.8916988313403372e-05, + "loss": 0.5834, + "step": 5551 + }, + { + "epoch": 0.17511433527834727, + "grad_norm": 2.680055618286133, + "learning_rate": 1.8916525880580777e-05, + "loss": 0.4028, + "step": 5552 + }, + { + "epoch": 0.1751458760447879, + "grad_norm": 2.5619678497314453, + "learning_rate": 1.8916063354707314e-05, + "loss": 0.6061, + "step": 5553 + }, + { + "epoch": 0.1751774168112285, + "grad_norm": 2.702662229537964, + "learning_rate": 1.8915600735787803e-05, + "loss": 0.5759, + "step": 5554 + }, + { + "epoch": 0.17520895757766913, + "grad_norm": 2.2689573764801025, + "learning_rate": 1.891513802382708e-05, + "loss": 0.5818, + "step": 5555 + }, + { + "epoch": 0.17524049834410976, + "grad_norm": 2.429702043533325, + "learning_rate": 1.8914675218829964e-05, + "loss": 0.562, + "step": 5556 + }, + { + "epoch": 0.1752720391105504, + "grad_norm": 2.2232651710510254, + "learning_rate": 1.8914212320801293e-05, + "loss": 0.5428, + "step": 5557 + }, + { + "epoch": 0.17530357987699102, + "grad_norm": 2.4288055896759033, + "learning_rate": 1.8913749329745897e-05, + "loss": 0.5635, + "step": 5558 + }, + { + "epoch": 0.17533512064343162, + "grad_norm": 2.35992169380188, + "learning_rate": 1.89132862456686e-05, + "loss": 0.5679, + "step": 5559 + }, + { + "epoch": 0.17536666140987225, + "grad_norm": 2.4575412273406982, + "learning_rate": 1.8912823068574245e-05, + "loss": 0.5827, + "step": 5560 + }, + { + "epoch": 0.17539820217631288, + "grad_norm": 2.318786859512329, + "learning_rate": 1.8912359798467656e-05, + "loss": 0.5971, + "step": 5561 + }, + { + "epoch": 0.1754297429427535, + "grad_norm": 1.7302322387695312, + "learning_rate": 1.8911896435353674e-05, + "loss": 0.8501, + "step": 5562 + }, + { + "epoch": 0.17546128370919414, + "grad_norm": 2.9319818019866943, + "learning_rate": 1.8911432979237136e-05, + "loss": 0.5692, + "step": 5563 + }, + { + "epoch": 0.17549282447563475, + "grad_norm": 3.1406800746917725, + "learning_rate": 1.891096943012287e-05, + "loss": 0.5879, + "step": 5564 + }, + { + "epoch": 0.17552436524207538, + "grad_norm": 1.7175486087799072, + "learning_rate": 1.8910505788015725e-05, + "loss": 0.4305, + "step": 5565 + }, + { + "epoch": 0.175555906008516, + "grad_norm": 2.0727169513702393, + "learning_rate": 1.891004205292053e-05, + "loss": 0.5555, + "step": 5566 + }, + { + "epoch": 0.17558744677495663, + "grad_norm": 3.050563097000122, + "learning_rate": 1.8909578224842127e-05, + "loss": 0.5629, + "step": 5567 + }, + { + "epoch": 0.17561898754139726, + "grad_norm": 2.4774279594421387, + "learning_rate": 1.8909114303785358e-05, + "loss": 0.5532, + "step": 5568 + }, + { + "epoch": 0.1756505283078379, + "grad_norm": 2.689631938934326, + "learning_rate": 1.8908650289755064e-05, + "loss": 0.5662, + "step": 5569 + }, + { + "epoch": 0.1756820690742785, + "grad_norm": 2.844172239303589, + "learning_rate": 1.8908186182756084e-05, + "loss": 0.5671, + "step": 5570 + }, + { + "epoch": 0.17571360984071913, + "grad_norm": 1.7160584926605225, + "learning_rate": 1.8907721982793266e-05, + "loss": 0.5934, + "step": 5571 + }, + { + "epoch": 0.17574515060715976, + "grad_norm": 1.7880513668060303, + "learning_rate": 1.8907257689871456e-05, + "loss": 0.5495, + "step": 5572 + }, + { + "epoch": 0.17577669137360039, + "grad_norm": 2.2663652896881104, + "learning_rate": 1.890679330399549e-05, + "loss": 0.9394, + "step": 5573 + }, + { + "epoch": 0.17580823214004102, + "grad_norm": 2.072310447692871, + "learning_rate": 1.8906328825170223e-05, + "loss": 0.5824, + "step": 5574 + }, + { + "epoch": 0.17583977290648162, + "grad_norm": 2.3203282356262207, + "learning_rate": 1.89058642534005e-05, + "loss": 0.5902, + "step": 5575 + }, + { + "epoch": 0.17587131367292225, + "grad_norm": 2.4390954971313477, + "learning_rate": 1.8905399588691165e-05, + "loss": 0.5819, + "step": 5576 + }, + { + "epoch": 0.17590285443936288, + "grad_norm": 16.45229721069336, + "learning_rate": 1.8904934831047073e-05, + "loss": 0.4282, + "step": 5577 + }, + { + "epoch": 0.1759343952058035, + "grad_norm": 1.7748463153839111, + "learning_rate": 1.890446998047307e-05, + "loss": 0.5977, + "step": 5578 + }, + { + "epoch": 0.17596593597224414, + "grad_norm": 2.2984440326690674, + "learning_rate": 1.890400503697401e-05, + "loss": 0.5733, + "step": 5579 + }, + { + "epoch": 0.17599747673868474, + "grad_norm": 2.2806389331817627, + "learning_rate": 1.8903540000554745e-05, + "loss": 0.5948, + "step": 5580 + }, + { + "epoch": 0.17602901750512537, + "grad_norm": 2.1191470623016357, + "learning_rate": 1.8903074871220123e-05, + "loss": 0.5846, + "step": 5581 + }, + { + "epoch": 0.176060558271566, + "grad_norm": 2.004152774810791, + "learning_rate": 1.8902609648975e-05, + "loss": 0.5722, + "step": 5582 + }, + { + "epoch": 0.17609209903800663, + "grad_norm": 3.2141976356506348, + "learning_rate": 1.8902144333824236e-05, + "loss": 0.6006, + "step": 5583 + }, + { + "epoch": 0.17612363980444726, + "grad_norm": 2.4169974327087402, + "learning_rate": 1.8901678925772685e-05, + "loss": 0.5886, + "step": 5584 + }, + { + "epoch": 0.17615518057088786, + "grad_norm": 1.9161432981491089, + "learning_rate": 1.89012134248252e-05, + "loss": 0.5588, + "step": 5585 + }, + { + "epoch": 0.1761867213373285, + "grad_norm": 5.523727893829346, + "learning_rate": 1.8900747830986642e-05, + "loss": 0.9107, + "step": 5586 + }, + { + "epoch": 0.17621826210376912, + "grad_norm": 1.8153198957443237, + "learning_rate": 1.890028214426187e-05, + "loss": 0.5694, + "step": 5587 + }, + { + "epoch": 0.17624980287020975, + "grad_norm": 1.6076115369796753, + "learning_rate": 1.8899816364655743e-05, + "loss": 0.5597, + "step": 5588 + }, + { + "epoch": 0.17628134363665038, + "grad_norm": 2.0396153926849365, + "learning_rate": 1.889935049217312e-05, + "loss": 0.5758, + "step": 5589 + }, + { + "epoch": 0.176312884403091, + "grad_norm": 2.490780830383301, + "learning_rate": 1.8898884526818866e-05, + "loss": 0.5721, + "step": 5590 + }, + { + "epoch": 0.1763444251695316, + "grad_norm": 2.1541895866394043, + "learning_rate": 1.889841846859784e-05, + "loss": 0.5725, + "step": 5591 + }, + { + "epoch": 0.17637596593597224, + "grad_norm": 1.8698948621749878, + "learning_rate": 1.8897952317514908e-05, + "loss": 0.5859, + "step": 5592 + }, + { + "epoch": 0.17640750670241287, + "grad_norm": 1.6688363552093506, + "learning_rate": 1.8897486073574937e-05, + "loss": 0.5488, + "step": 5593 + }, + { + "epoch": 0.1764390474688535, + "grad_norm": 1.9513882398605347, + "learning_rate": 1.889701973678279e-05, + "loss": 0.5913, + "step": 5594 + }, + { + "epoch": 0.17647058823529413, + "grad_norm": 2.3970718383789062, + "learning_rate": 1.889655330714333e-05, + "loss": 0.5663, + "step": 5595 + }, + { + "epoch": 0.17650212900173473, + "grad_norm": 4.704119682312012, + "learning_rate": 1.8896086784661434e-05, + "loss": 0.4206, + "step": 5596 + }, + { + "epoch": 0.17653366976817536, + "grad_norm": 2.424496650695801, + "learning_rate": 1.889562016934196e-05, + "loss": 0.54, + "step": 5597 + }, + { + "epoch": 0.176565210534616, + "grad_norm": 2.546172857284546, + "learning_rate": 1.8895153461189785e-05, + "loss": 0.5173, + "step": 5598 + }, + { + "epoch": 0.17659675130105662, + "grad_norm": 2.3628346920013428, + "learning_rate": 1.8894686660209777e-05, + "loss": 0.5758, + "step": 5599 + }, + { + "epoch": 0.17662829206749725, + "grad_norm": 2.4477219581604004, + "learning_rate": 1.8894219766406806e-05, + "loss": 0.5907, + "step": 5600 + }, + { + "epoch": 0.17665983283393785, + "grad_norm": 2.1191086769104004, + "learning_rate": 1.8893752779785742e-05, + "loss": 0.564, + "step": 5601 + }, + { + "epoch": 0.17669137360037848, + "grad_norm": 1.817258596420288, + "learning_rate": 1.8893285700351466e-05, + "loss": 0.8756, + "step": 5602 + }, + { + "epoch": 0.1767229143668191, + "grad_norm": 2.438451051712036, + "learning_rate": 1.889281852810885e-05, + "loss": 0.5862, + "step": 5603 + }, + { + "epoch": 0.17675445513325974, + "grad_norm": 2.615851879119873, + "learning_rate": 1.8892351263062763e-05, + "loss": 0.5663, + "step": 5604 + }, + { + "epoch": 0.17678599589970037, + "grad_norm": 1.8850277662277222, + "learning_rate": 1.8891883905218088e-05, + "loss": 0.4169, + "step": 5605 + }, + { + "epoch": 0.176817536666141, + "grad_norm": 2.6847023963928223, + "learning_rate": 1.88914164545797e-05, + "loss": 0.5802, + "step": 5606 + }, + { + "epoch": 0.1768490774325816, + "grad_norm": 2.6802473068237305, + "learning_rate": 1.889094891115248e-05, + "loss": 0.891, + "step": 5607 + }, + { + "epoch": 0.17688061819902223, + "grad_norm": 2.04276180267334, + "learning_rate": 1.88904812749413e-05, + "loss": 0.5569, + "step": 5608 + }, + { + "epoch": 0.17691215896546286, + "grad_norm": 2.080615282058716, + "learning_rate": 1.8890013545951047e-05, + "loss": 0.5729, + "step": 5609 + }, + { + "epoch": 0.1769436997319035, + "grad_norm": 2.389608860015869, + "learning_rate": 1.88895457241866e-05, + "loss": 0.5682, + "step": 5610 + }, + { + "epoch": 0.17697524049834412, + "grad_norm": 1.9466127157211304, + "learning_rate": 1.8889077809652837e-05, + "loss": 0.5902, + "step": 5611 + }, + { + "epoch": 0.17700678126478472, + "grad_norm": 1.8799519538879395, + "learning_rate": 1.888860980235465e-05, + "loss": 0.6024, + "step": 5612 + }, + { + "epoch": 0.17703832203122535, + "grad_norm": 2.680399179458618, + "learning_rate": 1.8888141702296912e-05, + "loss": 0.3443, + "step": 5613 + }, + { + "epoch": 0.17706986279766598, + "grad_norm": 2.1890008449554443, + "learning_rate": 1.8887673509484522e-05, + "loss": 0.5757, + "step": 5614 + }, + { + "epoch": 0.1771014035641066, + "grad_norm": 2.5528764724731445, + "learning_rate": 1.8887205223922353e-05, + "loss": 0.5878, + "step": 5615 + }, + { + "epoch": 0.17713294433054724, + "grad_norm": 1.5922633409500122, + "learning_rate": 1.8886736845615297e-05, + "loss": 0.4173, + "step": 5616 + }, + { + "epoch": 0.17716448509698784, + "grad_norm": 1.8301912546157837, + "learning_rate": 1.888626837456824e-05, + "loss": 0.5819, + "step": 5617 + }, + { + "epoch": 0.17719602586342847, + "grad_norm": 2.2070260047912598, + "learning_rate": 1.8885799810786075e-05, + "loss": 0.581, + "step": 5618 + }, + { + "epoch": 0.1772275666298691, + "grad_norm": 3.9083304405212402, + "learning_rate": 1.888533115427369e-05, + "loss": 0.5753, + "step": 5619 + }, + { + "epoch": 0.17725910739630973, + "grad_norm": 1.7386866807937622, + "learning_rate": 1.8884862405035972e-05, + "loss": 0.5829, + "step": 5620 + }, + { + "epoch": 0.17729064816275036, + "grad_norm": 1.9223746061325073, + "learning_rate": 1.888439356307782e-05, + "loss": 0.5768, + "step": 5621 + }, + { + "epoch": 0.17732218892919097, + "grad_norm": 1.9165472984313965, + "learning_rate": 1.888392462840412e-05, + "loss": 0.5899, + "step": 5622 + }, + { + "epoch": 0.1773537296956316, + "grad_norm": 2.402157783508301, + "learning_rate": 1.8883455601019766e-05, + "loss": 0.5732, + "step": 5623 + }, + { + "epoch": 0.17738527046207223, + "grad_norm": 2.336158037185669, + "learning_rate": 1.8882986480929658e-05, + "loss": 0.5635, + "step": 5624 + }, + { + "epoch": 0.17741681122851286, + "grad_norm": 10.517950057983398, + "learning_rate": 1.888251726813869e-05, + "loss": 0.4237, + "step": 5625 + }, + { + "epoch": 0.17744835199495348, + "grad_norm": 2.4670422077178955, + "learning_rate": 1.8882047962651754e-05, + "loss": 0.9291, + "step": 5626 + }, + { + "epoch": 0.17747989276139411, + "grad_norm": 2.1131935119628906, + "learning_rate": 1.8881578564473753e-05, + "loss": 0.5865, + "step": 5627 + }, + { + "epoch": 0.17751143352783472, + "grad_norm": 3.111973762512207, + "learning_rate": 1.888110907360958e-05, + "loss": 0.603, + "step": 5628 + }, + { + "epoch": 0.17754297429427535, + "grad_norm": 3.3748350143432617, + "learning_rate": 1.8880639490064142e-05, + "loss": 0.5558, + "step": 5629 + }, + { + "epoch": 0.17757451506071598, + "grad_norm": 1.790619969367981, + "learning_rate": 1.8880169813842335e-05, + "loss": 0.8568, + "step": 5630 + }, + { + "epoch": 0.1776060558271566, + "grad_norm": 2.3218977451324463, + "learning_rate": 1.887970004494906e-05, + "loss": 0.5647, + "step": 5631 + }, + { + "epoch": 0.17763759659359724, + "grad_norm": 3.0088677406311035, + "learning_rate": 1.887923018338922e-05, + "loss": 0.5856, + "step": 5632 + }, + { + "epoch": 0.17766913736003784, + "grad_norm": 2.067457914352417, + "learning_rate": 1.8878760229167715e-05, + "loss": 0.5629, + "step": 5633 + }, + { + "epoch": 0.17770067812647847, + "grad_norm": 2.323772430419922, + "learning_rate": 1.8878290182289457e-05, + "loss": 0.5625, + "step": 5634 + }, + { + "epoch": 0.1777322188929191, + "grad_norm": 1.916513442993164, + "learning_rate": 1.8877820042759343e-05, + "loss": 0.5726, + "step": 5635 + }, + { + "epoch": 0.17776375965935973, + "grad_norm": 2.420011043548584, + "learning_rate": 1.8877349810582287e-05, + "loss": 0.576, + "step": 5636 + }, + { + "epoch": 0.17779530042580036, + "grad_norm": 2.361992835998535, + "learning_rate": 1.887687948576319e-05, + "loss": 0.5707, + "step": 5637 + }, + { + "epoch": 0.17782684119224096, + "grad_norm": 2.389298915863037, + "learning_rate": 1.8876409068306965e-05, + "loss": 0.5828, + "step": 5638 + }, + { + "epoch": 0.1778583819586816, + "grad_norm": 2.421722650527954, + "learning_rate": 1.887593855821852e-05, + "loss": 0.8966, + "step": 5639 + }, + { + "epoch": 0.17788992272512222, + "grad_norm": 2.237729549407959, + "learning_rate": 1.8875467955502766e-05, + "loss": 0.8594, + "step": 5640 + }, + { + "epoch": 0.17792146349156285, + "grad_norm": 1.907015323638916, + "learning_rate": 1.8874997260164608e-05, + "loss": 0.8619, + "step": 5641 + }, + { + "epoch": 0.17795300425800348, + "grad_norm": 2.9518492221832275, + "learning_rate": 1.8874526472208964e-05, + "loss": 0.5498, + "step": 5642 + }, + { + "epoch": 0.1779845450244441, + "grad_norm": 2.5570616722106934, + "learning_rate": 1.8874055591640746e-05, + "loss": 0.6048, + "step": 5643 + }, + { + "epoch": 0.1780160857908847, + "grad_norm": 2.568134069442749, + "learning_rate": 1.8873584618464865e-05, + "loss": 0.6196, + "step": 5644 + }, + { + "epoch": 0.17804762655732534, + "grad_norm": 2.4281694889068604, + "learning_rate": 1.887311355268624e-05, + "loss": 0.553, + "step": 5645 + }, + { + "epoch": 0.17807916732376597, + "grad_norm": 2.8497931957244873, + "learning_rate": 1.8872642394309783e-05, + "loss": 0.5496, + "step": 5646 + }, + { + "epoch": 0.1781107080902066, + "grad_norm": 2.652735948562622, + "learning_rate": 1.8872171143340417e-05, + "loss": 0.9374, + "step": 5647 + }, + { + "epoch": 0.17814224885664723, + "grad_norm": 1.9451528787612915, + "learning_rate": 1.8871699799783056e-05, + "loss": 0.5434, + "step": 5648 + }, + { + "epoch": 0.17817378962308783, + "grad_norm": 2.4194600582122803, + "learning_rate": 1.8871228363642615e-05, + "loss": 0.5658, + "step": 5649 + }, + { + "epoch": 0.17820533038952846, + "grad_norm": 1.9913873672485352, + "learning_rate": 1.8870756834924017e-05, + "loss": 0.3962, + "step": 5650 + }, + { + "epoch": 0.1782368711559691, + "grad_norm": 2.0892252922058105, + "learning_rate": 1.887028521363219e-05, + "loss": 0.6082, + "step": 5651 + }, + { + "epoch": 0.17826841192240972, + "grad_norm": 5.215782642364502, + "learning_rate": 1.8869813499772044e-05, + "loss": 0.8961, + "step": 5652 + }, + { + "epoch": 0.17829995268885035, + "grad_norm": 1.7121026515960693, + "learning_rate": 1.8869341693348507e-05, + "loss": 0.9243, + "step": 5653 + }, + { + "epoch": 0.17833149345529095, + "grad_norm": 2.594343423843384, + "learning_rate": 1.8868869794366505e-05, + "loss": 0.5681, + "step": 5654 + }, + { + "epoch": 0.17836303422173158, + "grad_norm": 2.712965250015259, + "learning_rate": 1.886839780283096e-05, + "loss": 0.5484, + "step": 5655 + }, + { + "epoch": 0.1783945749881722, + "grad_norm": 2.0195887088775635, + "learning_rate": 1.8867925718746796e-05, + "loss": 0.5869, + "step": 5656 + }, + { + "epoch": 0.17842611575461284, + "grad_norm": 2.18707013130188, + "learning_rate": 1.886745354211894e-05, + "loss": 0.595, + "step": 5657 + }, + { + "epoch": 0.17845765652105347, + "grad_norm": 2.0696041584014893, + "learning_rate": 1.8866981272952324e-05, + "loss": 0.572, + "step": 5658 + }, + { + "epoch": 0.17848919728749407, + "grad_norm": 2.6951916217803955, + "learning_rate": 1.8866508911251873e-05, + "loss": 0.5329, + "step": 5659 + }, + { + "epoch": 0.1785207380539347, + "grad_norm": 1.8641797304153442, + "learning_rate": 1.8866036457022516e-05, + "loss": 0.572, + "step": 5660 + }, + { + "epoch": 0.17855227882037533, + "grad_norm": 2.3043222427368164, + "learning_rate": 1.8865563910269185e-05, + "loss": 0.5365, + "step": 5661 + }, + { + "epoch": 0.17858381958681596, + "grad_norm": 2.43369722366333, + "learning_rate": 1.886509127099681e-05, + "loss": 0.5586, + "step": 5662 + }, + { + "epoch": 0.1786153603532566, + "grad_norm": 2.2658610343933105, + "learning_rate": 1.8864618539210324e-05, + "loss": 0.5441, + "step": 5663 + }, + { + "epoch": 0.17864690111969722, + "grad_norm": 3.125455379486084, + "learning_rate": 1.886414571491466e-05, + "loss": 0.9183, + "step": 5664 + }, + { + "epoch": 0.17867844188613782, + "grad_norm": 2.2126426696777344, + "learning_rate": 1.8863672798114755e-05, + "loss": 0.5723, + "step": 5665 + }, + { + "epoch": 0.17870998265257845, + "grad_norm": 1.9485664367675781, + "learning_rate": 1.886319978881554e-05, + "loss": 0.5987, + "step": 5666 + }, + { + "epoch": 0.17874152341901908, + "grad_norm": 2.0680441856384277, + "learning_rate": 1.886272668702195e-05, + "loss": 0.8968, + "step": 5667 + }, + { + "epoch": 0.1787730641854597, + "grad_norm": 1.7826825380325317, + "learning_rate": 1.8862253492738932e-05, + "loss": 0.5848, + "step": 5668 + }, + { + "epoch": 0.17880460495190034, + "grad_norm": 1.9121090173721313, + "learning_rate": 1.886178020597141e-05, + "loss": 0.8808, + "step": 5669 + }, + { + "epoch": 0.17883614571834094, + "grad_norm": 2.093369245529175, + "learning_rate": 1.8861306826724338e-05, + "loss": 0.5594, + "step": 5670 + }, + { + "epoch": 0.17886768648478157, + "grad_norm": 3.0298142433166504, + "learning_rate": 1.8860833355002643e-05, + "loss": 0.5493, + "step": 5671 + }, + { + "epoch": 0.1788992272512222, + "grad_norm": 2.347367763519287, + "learning_rate": 1.8860359790811275e-05, + "loss": 0.5739, + "step": 5672 + }, + { + "epoch": 0.17893076801766283, + "grad_norm": 3.838369607925415, + "learning_rate": 1.8859886134155173e-05, + "loss": 0.5676, + "step": 5673 + }, + { + "epoch": 0.17896230878410346, + "grad_norm": 2.0130908489227295, + "learning_rate": 1.8859412385039277e-05, + "loss": 0.4222, + "step": 5674 + }, + { + "epoch": 0.17899384955054407, + "grad_norm": 2.441819190979004, + "learning_rate": 1.8858938543468533e-05, + "loss": 0.599, + "step": 5675 + }, + { + "epoch": 0.1790253903169847, + "grad_norm": 2.237339973449707, + "learning_rate": 1.885846460944789e-05, + "loss": 0.558, + "step": 5676 + }, + { + "epoch": 0.17905693108342532, + "grad_norm": 2.4262032508850098, + "learning_rate": 1.8857990582982292e-05, + "loss": 0.5741, + "step": 5677 + }, + { + "epoch": 0.17908847184986595, + "grad_norm": 2.7266576290130615, + "learning_rate": 1.885751646407668e-05, + "loss": 0.5758, + "step": 5678 + }, + { + "epoch": 0.17912001261630658, + "grad_norm": 2.344672918319702, + "learning_rate": 1.8857042252736004e-05, + "loss": 0.5977, + "step": 5679 + }, + { + "epoch": 0.17915155338274721, + "grad_norm": 2.9186606407165527, + "learning_rate": 1.885656794896522e-05, + "loss": 0.5545, + "step": 5680 + }, + { + "epoch": 0.17918309414918782, + "grad_norm": 1.7610913515090942, + "learning_rate": 1.8856093552769268e-05, + "loss": 0.5677, + "step": 5681 + }, + { + "epoch": 0.17921463491562845, + "grad_norm": 7.01786470413208, + "learning_rate": 1.8855619064153105e-05, + "loss": 0.5551, + "step": 5682 + }, + { + "epoch": 0.17924617568206908, + "grad_norm": 2.658799409866333, + "learning_rate": 1.8855144483121683e-05, + "loss": 0.5837, + "step": 5683 + }, + { + "epoch": 0.1792777164485097, + "grad_norm": 3.2408671379089355, + "learning_rate": 1.885466980967995e-05, + "loss": 0.5523, + "step": 5684 + }, + { + "epoch": 0.17930925721495033, + "grad_norm": 2.3920021057128906, + "learning_rate": 1.885419504383286e-05, + "loss": 0.394, + "step": 5685 + }, + { + "epoch": 0.17934079798139094, + "grad_norm": 2.1673011779785156, + "learning_rate": 1.8853720185585376e-05, + "loss": 0.5663, + "step": 5686 + }, + { + "epoch": 0.17937233874783157, + "grad_norm": 2.4509871006011963, + "learning_rate": 1.8853245234942444e-05, + "loss": 0.5751, + "step": 5687 + }, + { + "epoch": 0.1794038795142722, + "grad_norm": 2.2678050994873047, + "learning_rate": 1.885277019190902e-05, + "loss": 0.5639, + "step": 5688 + }, + { + "epoch": 0.17943542028071283, + "grad_norm": 2.100416660308838, + "learning_rate": 1.885229505649007e-05, + "loss": 0.5816, + "step": 5689 + }, + { + "epoch": 0.17946696104715346, + "grad_norm": 2.267833948135376, + "learning_rate": 1.8851819828690544e-05, + "loss": 0.3947, + "step": 5690 + }, + { + "epoch": 0.17949850181359406, + "grad_norm": 2.0468413829803467, + "learning_rate": 1.8851344508515407e-05, + "loss": 0.5838, + "step": 5691 + }, + { + "epoch": 0.1795300425800347, + "grad_norm": 3.7069873809814453, + "learning_rate": 1.8850869095969617e-05, + "loss": 0.546, + "step": 5692 + }, + { + "epoch": 0.17956158334647532, + "grad_norm": 2.459092378616333, + "learning_rate": 1.8850393591058136e-05, + "loss": 0.5874, + "step": 5693 + }, + { + "epoch": 0.17959312411291595, + "grad_norm": 3.1094932556152344, + "learning_rate": 1.8849917993785926e-05, + "loss": 0.547, + "step": 5694 + }, + { + "epoch": 0.17962466487935658, + "grad_norm": 1.9834359884262085, + "learning_rate": 1.884944230415795e-05, + "loss": 0.3868, + "step": 5695 + }, + { + "epoch": 0.1796562056457972, + "grad_norm": 2.1168816089630127, + "learning_rate": 1.884896652217917e-05, + "loss": 0.6045, + "step": 5696 + }, + { + "epoch": 0.1796877464122378, + "grad_norm": 3.1166980266571045, + "learning_rate": 1.8848490647854553e-05, + "loss": 0.5845, + "step": 5697 + }, + { + "epoch": 0.17971928717867844, + "grad_norm": 2.7329671382904053, + "learning_rate": 1.8848014681189068e-05, + "loss": 0.5869, + "step": 5698 + }, + { + "epoch": 0.17975082794511907, + "grad_norm": 2.035183906555176, + "learning_rate": 1.8847538622187675e-05, + "loss": 0.5634, + "step": 5699 + }, + { + "epoch": 0.1797823687115597, + "grad_norm": 3.271754503250122, + "learning_rate": 1.884706247085535e-05, + "loss": 0.9342, + "step": 5700 + }, + { + "epoch": 0.17981390947800033, + "grad_norm": 2.572021484375, + "learning_rate": 1.8846586227197057e-05, + "loss": 0.908, + "step": 5701 + }, + { + "epoch": 0.17984545024444093, + "grad_norm": 2.802492380142212, + "learning_rate": 1.884610989121777e-05, + "loss": 0.5786, + "step": 5702 + }, + { + "epoch": 0.17987699101088156, + "grad_norm": 1.927873969078064, + "learning_rate": 1.8845633462922457e-05, + "loss": 0.5545, + "step": 5703 + }, + { + "epoch": 0.1799085317773222, + "grad_norm": 3.006526470184326, + "learning_rate": 1.884515694231609e-05, + "loss": 0.5696, + "step": 5704 + }, + { + "epoch": 0.17994007254376282, + "grad_norm": 2.5813586711883545, + "learning_rate": 1.884468032940364e-05, + "loss": 0.5741, + "step": 5705 + }, + { + "epoch": 0.17997161331020345, + "grad_norm": 2.9601211547851562, + "learning_rate": 1.8844203624190088e-05, + "loss": 0.5583, + "step": 5706 + }, + { + "epoch": 0.18000315407664405, + "grad_norm": 2.0539917945861816, + "learning_rate": 1.88437268266804e-05, + "loss": 0.5611, + "step": 5707 + }, + { + "epoch": 0.18003469484308468, + "grad_norm": 2.095416307449341, + "learning_rate": 1.8843249936879555e-05, + "loss": 0.5447, + "step": 5708 + }, + { + "epoch": 0.1800662356095253, + "grad_norm": 2.1049551963806152, + "learning_rate": 1.884277295479253e-05, + "loss": 0.5736, + "step": 5709 + }, + { + "epoch": 0.18009777637596594, + "grad_norm": 2.5105388164520264, + "learning_rate": 1.8842295880424305e-05, + "loss": 0.575, + "step": 5710 + }, + { + "epoch": 0.18012931714240657, + "grad_norm": 4.156827926635742, + "learning_rate": 1.8841818713779857e-05, + "loss": 0.594, + "step": 5711 + }, + { + "epoch": 0.18016085790884717, + "grad_norm": 3.5369160175323486, + "learning_rate": 1.8841341454864164e-05, + "loss": 0.5523, + "step": 5712 + }, + { + "epoch": 0.1801923986752878, + "grad_norm": 1.9762134552001953, + "learning_rate": 1.884086410368221e-05, + "loss": 0.5693, + "step": 5713 + }, + { + "epoch": 0.18022393944172843, + "grad_norm": 2.2525827884674072, + "learning_rate": 1.884038666023897e-05, + "loss": 0.4547, + "step": 5714 + }, + { + "epoch": 0.18025548020816906, + "grad_norm": 2.391709804534912, + "learning_rate": 1.8839909124539434e-05, + "loss": 0.4203, + "step": 5715 + }, + { + "epoch": 0.1802870209746097, + "grad_norm": 2.253753900527954, + "learning_rate": 1.8839431496588582e-05, + "loss": 0.5876, + "step": 5716 + }, + { + "epoch": 0.18031856174105032, + "grad_norm": 2.8160250186920166, + "learning_rate": 1.88389537763914e-05, + "loss": 0.5857, + "step": 5717 + }, + { + "epoch": 0.18035010250749092, + "grad_norm": 2.1156187057495117, + "learning_rate": 1.883847596395287e-05, + "loss": 0.5704, + "step": 5718 + }, + { + "epoch": 0.18038164327393155, + "grad_norm": 4.4089884757995605, + "learning_rate": 1.8837998059277977e-05, + "loss": 0.9754, + "step": 5719 + }, + { + "epoch": 0.18041318404037218, + "grad_norm": 3.8474793434143066, + "learning_rate": 1.8837520062371717e-05, + "loss": 0.599, + "step": 5720 + }, + { + "epoch": 0.1804447248068128, + "grad_norm": 2.8812801837921143, + "learning_rate": 1.8837041973239072e-05, + "loss": 0.5527, + "step": 5721 + }, + { + "epoch": 0.18047626557325344, + "grad_norm": 2.1580357551574707, + "learning_rate": 1.883656379188503e-05, + "loss": 0.5389, + "step": 5722 + }, + { + "epoch": 0.18050780633969404, + "grad_norm": 2.176288604736328, + "learning_rate": 1.8836085518314585e-05, + "loss": 0.5458, + "step": 5723 + }, + { + "epoch": 0.18053934710613467, + "grad_norm": 2.008875846862793, + "learning_rate": 1.883560715253273e-05, + "loss": 0.596, + "step": 5724 + }, + { + "epoch": 0.1805708878725753, + "grad_norm": 2.3352785110473633, + "learning_rate": 1.883512869454445e-05, + "loss": 0.5692, + "step": 5725 + }, + { + "epoch": 0.18060242863901593, + "grad_norm": 2.2785749435424805, + "learning_rate": 1.883465014435474e-05, + "loss": 0.5862, + "step": 5726 + }, + { + "epoch": 0.18063396940545656, + "grad_norm": 2.28619122505188, + "learning_rate": 1.88341715019686e-05, + "loss": 0.4228, + "step": 5727 + }, + { + "epoch": 0.18066551017189716, + "grad_norm": 2.212445020675659, + "learning_rate": 1.8833692767391017e-05, + "loss": 0.5395, + "step": 5728 + }, + { + "epoch": 0.1806970509383378, + "grad_norm": 2.6169893741607666, + "learning_rate": 1.8833213940626995e-05, + "loss": 0.5778, + "step": 5729 + }, + { + "epoch": 0.18072859170477842, + "grad_norm": 3.0154805183410645, + "learning_rate": 1.8832735021681522e-05, + "loss": 0.556, + "step": 5730 + }, + { + "epoch": 0.18076013247121905, + "grad_norm": 2.351909875869751, + "learning_rate": 1.8832256010559603e-05, + "loss": 0.5756, + "step": 5731 + }, + { + "epoch": 0.18079167323765968, + "grad_norm": 2.561110258102417, + "learning_rate": 1.8831776907266235e-05, + "loss": 0.575, + "step": 5732 + }, + { + "epoch": 0.1808232140041003, + "grad_norm": 2.3774399757385254, + "learning_rate": 1.8831297711806416e-05, + "loss": 0.5866, + "step": 5733 + }, + { + "epoch": 0.18085475477054092, + "grad_norm": 2.2087819576263428, + "learning_rate": 1.883081842418515e-05, + "loss": 0.5898, + "step": 5734 + }, + { + "epoch": 0.18088629553698155, + "grad_norm": 5.867150783538818, + "learning_rate": 1.8830339044407435e-05, + "loss": 0.5393, + "step": 5735 + }, + { + "epoch": 0.18091783630342217, + "grad_norm": 10.883454322814941, + "learning_rate": 1.8829859572478275e-05, + "loss": 0.5894, + "step": 5736 + }, + { + "epoch": 0.1809493770698628, + "grad_norm": 2.2985076904296875, + "learning_rate": 1.8829380008402677e-05, + "loss": 0.5957, + "step": 5737 + }, + { + "epoch": 0.18098091783630343, + "grad_norm": 2.2731406688690186, + "learning_rate": 1.882890035218564e-05, + "loss": 0.5645, + "step": 5738 + }, + { + "epoch": 0.18101245860274404, + "grad_norm": 2.51628041267395, + "learning_rate": 1.8828420603832172e-05, + "loss": 0.5426, + "step": 5739 + }, + { + "epoch": 0.18104399936918467, + "grad_norm": 3.6878514289855957, + "learning_rate": 1.882794076334728e-05, + "loss": 0.938, + "step": 5740 + }, + { + "epoch": 0.1810755401356253, + "grad_norm": 2.978127956390381, + "learning_rate": 1.8827460830735975e-05, + "loss": 0.4667, + "step": 5741 + }, + { + "epoch": 0.18110708090206593, + "grad_norm": 3.910335063934326, + "learning_rate": 1.882698080600326e-05, + "loss": 0.6004, + "step": 5742 + }, + { + "epoch": 0.18113862166850656, + "grad_norm": 6.008777141571045, + "learning_rate": 1.8826500689154144e-05, + "loss": 0.5902, + "step": 5743 + }, + { + "epoch": 0.18117016243494716, + "grad_norm": 4.704282283782959, + "learning_rate": 1.8826020480193644e-05, + "loss": 0.5417, + "step": 5744 + }, + { + "epoch": 0.1812017032013878, + "grad_norm": 2.6428678035736084, + "learning_rate": 1.882554017912676e-05, + "loss": 0.573, + "step": 5745 + }, + { + "epoch": 0.18123324396782842, + "grad_norm": 3.3007571697235107, + "learning_rate": 1.882505978595852e-05, + "loss": 0.5413, + "step": 5746 + }, + { + "epoch": 0.18126478473426905, + "grad_norm": 2.570298194885254, + "learning_rate": 1.8824579300693923e-05, + "loss": 0.5559, + "step": 5747 + }, + { + "epoch": 0.18129632550070968, + "grad_norm": 2.877561092376709, + "learning_rate": 1.882409872333799e-05, + "loss": 0.5628, + "step": 5748 + }, + { + "epoch": 0.18132786626715028, + "grad_norm": 2.2711880207061768, + "learning_rate": 1.8823618053895737e-05, + "loss": 0.451, + "step": 5749 + }, + { + "epoch": 0.1813594070335909, + "grad_norm": 2.4688351154327393, + "learning_rate": 1.882313729237218e-05, + "loss": 0.5902, + "step": 5750 + }, + { + "epoch": 0.18139094780003154, + "grad_norm": 2.011759042739868, + "learning_rate": 1.8822656438772328e-05, + "loss": 0.5722, + "step": 5751 + }, + { + "epoch": 0.18142248856647217, + "grad_norm": 3.038864850997925, + "learning_rate": 1.882217549310121e-05, + "loss": 0.5972, + "step": 5752 + }, + { + "epoch": 0.1814540293329128, + "grad_norm": 4.740351676940918, + "learning_rate": 1.882169445536384e-05, + "loss": 0.5597, + "step": 5753 + }, + { + "epoch": 0.18148557009935343, + "grad_norm": 2.7494962215423584, + "learning_rate": 1.8821213325565236e-05, + "loss": 0.5794, + "step": 5754 + }, + { + "epoch": 0.18151711086579403, + "grad_norm": 3.2926218509674072, + "learning_rate": 1.882073210371042e-05, + "loss": 0.5673, + "step": 5755 + }, + { + "epoch": 0.18154865163223466, + "grad_norm": 1.787379264831543, + "learning_rate": 1.882025078980442e-05, + "loss": 0.5421, + "step": 5756 + }, + { + "epoch": 0.1815801923986753, + "grad_norm": 3.173095703125, + "learning_rate": 1.8819769383852253e-05, + "loss": 0.5871, + "step": 5757 + }, + { + "epoch": 0.18161173316511592, + "grad_norm": 2.2917332649230957, + "learning_rate": 1.8819287885858944e-05, + "loss": 0.6008, + "step": 5758 + }, + { + "epoch": 0.18164327393155655, + "grad_norm": 20.724708557128906, + "learning_rate": 1.8818806295829516e-05, + "loss": 0.9616, + "step": 5759 + }, + { + "epoch": 0.18167481469799715, + "grad_norm": 2.964980363845825, + "learning_rate": 1.8818324613769003e-05, + "loss": 0.5691, + "step": 5760 + }, + { + "epoch": 0.18170635546443778, + "grad_norm": 3.7042975425720215, + "learning_rate": 1.8817842839682415e-05, + "loss": 0.562, + "step": 5761 + }, + { + "epoch": 0.1817378962308784, + "grad_norm": 2.858046531677246, + "learning_rate": 1.88173609735748e-05, + "loss": 0.5606, + "step": 5762 + }, + { + "epoch": 0.18176943699731904, + "grad_norm": 3.2827932834625244, + "learning_rate": 1.881687901545117e-05, + "loss": 0.5666, + "step": 5763 + }, + { + "epoch": 0.18180097776375967, + "grad_norm": 5.070435047149658, + "learning_rate": 1.8816396965316562e-05, + "loss": 0.4199, + "step": 5764 + }, + { + "epoch": 0.18183251853020027, + "grad_norm": 3.0273473262786865, + "learning_rate": 1.881591482317601e-05, + "loss": 0.6002, + "step": 5765 + }, + { + "epoch": 0.1818640592966409, + "grad_norm": 3.4758591651916504, + "learning_rate": 1.8815432589034542e-05, + "loss": 0.5627, + "step": 5766 + }, + { + "epoch": 0.18189560006308153, + "grad_norm": 2.657707452774048, + "learning_rate": 1.8814950262897185e-05, + "loss": 0.582, + "step": 5767 + }, + { + "epoch": 0.18192714082952216, + "grad_norm": 3.0507993698120117, + "learning_rate": 1.881446784476898e-05, + "loss": 0.5747, + "step": 5768 + }, + { + "epoch": 0.1819586815959628, + "grad_norm": 2.9889299869537354, + "learning_rate": 1.8813985334654957e-05, + "loss": 0.5418, + "step": 5769 + }, + { + "epoch": 0.18199022236240342, + "grad_norm": 5.259953022003174, + "learning_rate": 1.8813502732560155e-05, + "loss": 0.5984, + "step": 5770 + }, + { + "epoch": 0.18202176312884402, + "grad_norm": 2.782200813293457, + "learning_rate": 1.8813020038489606e-05, + "loss": 0.587, + "step": 5771 + }, + { + "epoch": 0.18205330389528465, + "grad_norm": 3.514207124710083, + "learning_rate": 1.8812537252448356e-05, + "loss": 0.5723, + "step": 5772 + }, + { + "epoch": 0.18208484466172528, + "grad_norm": 6.742508411407471, + "learning_rate": 1.881205437444143e-05, + "loss": 0.4036, + "step": 5773 + }, + { + "epoch": 0.1821163854281659, + "grad_norm": 3.6416823863983154, + "learning_rate": 1.8811571404473877e-05, + "loss": 0.5535, + "step": 5774 + }, + { + "epoch": 0.18214792619460654, + "grad_norm": 9.620870590209961, + "learning_rate": 1.8811088342550733e-05, + "loss": 0.5754, + "step": 5775 + }, + { + "epoch": 0.18217946696104714, + "grad_norm": 3.1726224422454834, + "learning_rate": 1.8810605188677046e-05, + "loss": 0.595, + "step": 5776 + }, + { + "epoch": 0.18221100772748777, + "grad_norm": 4.167644500732422, + "learning_rate": 1.8810121942857848e-05, + "loss": 0.594, + "step": 5777 + }, + { + "epoch": 0.1822425484939284, + "grad_norm": 3.326977252960205, + "learning_rate": 1.8809638605098187e-05, + "loss": 0.596, + "step": 5778 + }, + { + "epoch": 0.18227408926036903, + "grad_norm": 4.7832255363464355, + "learning_rate": 1.8809155175403108e-05, + "loss": 0.6078, + "step": 5779 + }, + { + "epoch": 0.18230563002680966, + "grad_norm": 3.297356605529785, + "learning_rate": 1.8808671653777653e-05, + "loss": 0.6051, + "step": 5780 + }, + { + "epoch": 0.18233717079325026, + "grad_norm": 3.659367799758911, + "learning_rate": 1.880818804022687e-05, + "loss": 0.6162, + "step": 5781 + }, + { + "epoch": 0.1823687115596909, + "grad_norm": 7.155792713165283, + "learning_rate": 1.8807704334755803e-05, + "loss": 0.5561, + "step": 5782 + }, + { + "epoch": 0.18240025232613152, + "grad_norm": 3.810332775115967, + "learning_rate": 1.8807220537369506e-05, + "loss": 0.5781, + "step": 5783 + }, + { + "epoch": 0.18243179309257215, + "grad_norm": 3.154193639755249, + "learning_rate": 1.880673664807302e-05, + "loss": 0.5716, + "step": 5784 + }, + { + "epoch": 0.18246333385901278, + "grad_norm": 3.892932653427124, + "learning_rate": 1.8806252666871403e-05, + "loss": 0.8968, + "step": 5785 + }, + { + "epoch": 0.18249487462545338, + "grad_norm": 2.2401621341705322, + "learning_rate": 1.88057685937697e-05, + "loss": 0.9042, + "step": 5786 + }, + { + "epoch": 0.18252641539189401, + "grad_norm": 3.5892226696014404, + "learning_rate": 1.8805284428772964e-05, + "loss": 0.5685, + "step": 5787 + }, + { + "epoch": 0.18255795615833464, + "grad_norm": 3.2752621173858643, + "learning_rate": 1.8804800171886248e-05, + "loss": 0.5714, + "step": 5788 + }, + { + "epoch": 0.18258949692477527, + "grad_norm": 2.51296329498291, + "learning_rate": 1.8804315823114607e-05, + "loss": 0.613, + "step": 5789 + }, + { + "epoch": 0.1826210376912159, + "grad_norm": 2.3819169998168945, + "learning_rate": 1.880383138246309e-05, + "loss": 0.9237, + "step": 5790 + }, + { + "epoch": 0.18265257845765653, + "grad_norm": 3.594210624694824, + "learning_rate": 1.8803346849936762e-05, + "loss": 0.4077, + "step": 5791 + }, + { + "epoch": 0.18268411922409714, + "grad_norm": 4.320341110229492, + "learning_rate": 1.880286222554067e-05, + "loss": 0.5737, + "step": 5792 + }, + { + "epoch": 0.18271565999053777, + "grad_norm": 2.756925106048584, + "learning_rate": 1.8802377509279876e-05, + "loss": 0.5927, + "step": 5793 + }, + { + "epoch": 0.1827472007569784, + "grad_norm": 3.0142199993133545, + "learning_rate": 1.880189270115944e-05, + "loss": 0.5951, + "step": 5794 + }, + { + "epoch": 0.18277874152341902, + "grad_norm": 3.6111249923706055, + "learning_rate": 1.8801407801184414e-05, + "loss": 0.5446, + "step": 5795 + }, + { + "epoch": 0.18281028228985965, + "grad_norm": 4.408742427825928, + "learning_rate": 1.8800922809359868e-05, + "loss": 0.3929, + "step": 5796 + }, + { + "epoch": 0.18284182305630026, + "grad_norm": 2.4567759037017822, + "learning_rate": 1.8800437725690858e-05, + "loss": 0.9468, + "step": 5797 + }, + { + "epoch": 0.1828733638227409, + "grad_norm": 3.803858757019043, + "learning_rate": 1.8799952550182446e-05, + "loss": 0.5501, + "step": 5798 + }, + { + "epoch": 0.18290490458918152, + "grad_norm": 3.393916368484497, + "learning_rate": 1.87994672828397e-05, + "loss": 0.5843, + "step": 5799 + }, + { + "epoch": 0.18293644535562215, + "grad_norm": 33.100921630859375, + "learning_rate": 1.8798981923667673e-05, + "loss": 0.5586, + "step": 5800 + }, + { + "epoch": 0.18296798612206278, + "grad_norm": 3.5127198696136475, + "learning_rate": 1.8798496472671444e-05, + "loss": 0.5611, + "step": 5801 + }, + { + "epoch": 0.18299952688850338, + "grad_norm": 2.4151124954223633, + "learning_rate": 1.879801092985607e-05, + "loss": 0.5774, + "step": 5802 + }, + { + "epoch": 0.183031067654944, + "grad_norm": 2.727644920349121, + "learning_rate": 1.879752529522662e-05, + "loss": 0.6172, + "step": 5803 + }, + { + "epoch": 0.18306260842138464, + "grad_norm": 3.4612069129943848, + "learning_rate": 1.8797039568788163e-05, + "loss": 0.5691, + "step": 5804 + }, + { + "epoch": 0.18309414918782527, + "grad_norm": 3.2613003253936768, + "learning_rate": 1.8796553750545768e-05, + "loss": 0.5708, + "step": 5805 + }, + { + "epoch": 0.1831256899542659, + "grad_norm": 2.0206305980682373, + "learning_rate": 1.8796067840504503e-05, + "loss": 0.5969, + "step": 5806 + }, + { + "epoch": 0.18315723072070653, + "grad_norm": 2.261101245880127, + "learning_rate": 1.879558183866944e-05, + "loss": 0.5734, + "step": 5807 + }, + { + "epoch": 0.18318877148714713, + "grad_norm": 3.112593412399292, + "learning_rate": 1.8795095745045653e-05, + "loss": 0.6121, + "step": 5808 + }, + { + "epoch": 0.18322031225358776, + "grad_norm": 2.491278886795044, + "learning_rate": 1.879460955963821e-05, + "loss": 0.6188, + "step": 5809 + }, + { + "epoch": 0.1832518530200284, + "grad_norm": 3.173007011413574, + "learning_rate": 1.879412328245219e-05, + "loss": 0.583, + "step": 5810 + }, + { + "epoch": 0.18328339378646902, + "grad_norm": 2.446089744567871, + "learning_rate": 1.8793636913492664e-05, + "loss": 0.598, + "step": 5811 + }, + { + "epoch": 0.18331493455290965, + "grad_norm": 5.421658992767334, + "learning_rate": 1.879315045276471e-05, + "loss": 0.5722, + "step": 5812 + }, + { + "epoch": 0.18334647531935025, + "grad_norm": 1.7984511852264404, + "learning_rate": 1.87926639002734e-05, + "loss": 0.5726, + "step": 5813 + }, + { + "epoch": 0.18337801608579088, + "grad_norm": 3.9558396339416504, + "learning_rate": 1.8792177256023816e-05, + "loss": 0.3898, + "step": 5814 + }, + { + "epoch": 0.1834095568522315, + "grad_norm": 3.2739531993865967, + "learning_rate": 1.879169052002104e-05, + "loss": 0.5922, + "step": 5815 + }, + { + "epoch": 0.18344109761867214, + "grad_norm": 2.9559645652770996, + "learning_rate": 1.879120369227014e-05, + "loss": 0.5564, + "step": 5816 + }, + { + "epoch": 0.18347263838511277, + "grad_norm": 14.188807487487793, + "learning_rate": 1.8790716772776205e-05, + "loss": 0.5322, + "step": 5817 + }, + { + "epoch": 0.18350417915155337, + "grad_norm": 1.9014546871185303, + "learning_rate": 1.8790229761544315e-05, + "loss": 0.5906, + "step": 5818 + }, + { + "epoch": 0.183535719917994, + "grad_norm": 2.8587050437927246, + "learning_rate": 1.8789742658579554e-05, + "loss": 0.5699, + "step": 5819 + }, + { + "epoch": 0.18356726068443463, + "grad_norm": 2.4867992401123047, + "learning_rate": 1.8789255463887002e-05, + "loss": 0.5467, + "step": 5820 + }, + { + "epoch": 0.18359880145087526, + "grad_norm": 2.7080163955688477, + "learning_rate": 1.8788768177471747e-05, + "loss": 0.5663, + "step": 5821 + }, + { + "epoch": 0.1836303422173159, + "grad_norm": 2.166862964630127, + "learning_rate": 1.8788280799338866e-05, + "loss": 0.9335, + "step": 5822 + }, + { + "epoch": 0.1836618829837565, + "grad_norm": 1.8260449171066284, + "learning_rate": 1.8787793329493456e-05, + "loss": 0.925, + "step": 5823 + }, + { + "epoch": 0.18369342375019712, + "grad_norm": 2.0094470977783203, + "learning_rate": 1.8787305767940598e-05, + "loss": 0.5438, + "step": 5824 + }, + { + "epoch": 0.18372496451663775, + "grad_norm": 2.4029998779296875, + "learning_rate": 1.878681811468538e-05, + "loss": 0.5864, + "step": 5825 + }, + { + "epoch": 0.18375650528307838, + "grad_norm": 2.8704142570495605, + "learning_rate": 1.8786330369732892e-05, + "loss": 0.5677, + "step": 5826 + }, + { + "epoch": 0.183788046049519, + "grad_norm": 2.677215099334717, + "learning_rate": 1.8785842533088226e-05, + "loss": 0.5281, + "step": 5827 + }, + { + "epoch": 0.18381958681595964, + "grad_norm": 6.231016635894775, + "learning_rate": 1.8785354604756468e-05, + "loss": 0.6334, + "step": 5828 + }, + { + "epoch": 0.18385112758240024, + "grad_norm": 2.464576005935669, + "learning_rate": 1.8784866584742718e-05, + "loss": 0.5843, + "step": 5829 + }, + { + "epoch": 0.18388266834884087, + "grad_norm": 2.6420485973358154, + "learning_rate": 1.878437847305206e-05, + "loss": 0.5739, + "step": 5830 + }, + { + "epoch": 0.1839142091152815, + "grad_norm": 2.4984707832336426, + "learning_rate": 1.8783890269689592e-05, + "loss": 0.5975, + "step": 5831 + }, + { + "epoch": 0.18394574988172213, + "grad_norm": 2.3999600410461426, + "learning_rate": 1.8783401974660412e-05, + "loss": 0.5694, + "step": 5832 + }, + { + "epoch": 0.18397729064816276, + "grad_norm": 4.183218479156494, + "learning_rate": 1.8782913587969606e-05, + "loss": 0.5599, + "step": 5833 + }, + { + "epoch": 0.18400883141460336, + "grad_norm": 2.531395435333252, + "learning_rate": 1.8782425109622282e-05, + "loss": 0.5721, + "step": 5834 + }, + { + "epoch": 0.184040372181044, + "grad_norm": 2.1858279705047607, + "learning_rate": 1.8781936539623534e-05, + "loss": 0.5723, + "step": 5835 + }, + { + "epoch": 0.18407191294748462, + "grad_norm": 1.7210484743118286, + "learning_rate": 1.8781447877978454e-05, + "loss": 0.5418, + "step": 5836 + }, + { + "epoch": 0.18410345371392525, + "grad_norm": 1.9150327444076538, + "learning_rate": 1.878095912469215e-05, + "loss": 0.563, + "step": 5837 + }, + { + "epoch": 0.18413499448036588, + "grad_norm": 2.1218385696411133, + "learning_rate": 1.8780470279769716e-05, + "loss": 0.5856, + "step": 5838 + }, + { + "epoch": 0.18416653524680648, + "grad_norm": 2.1158840656280518, + "learning_rate": 1.877998134321626e-05, + "loss": 0.5746, + "step": 5839 + }, + { + "epoch": 0.18419807601324711, + "grad_norm": 2.2348313331604004, + "learning_rate": 1.877949231503688e-05, + "loss": 0.5907, + "step": 5840 + }, + { + "epoch": 0.18422961677968774, + "grad_norm": 2.9264626502990723, + "learning_rate": 1.8779003195236682e-05, + "loss": 0.5637, + "step": 5841 + }, + { + "epoch": 0.18426115754612837, + "grad_norm": 4.802423000335693, + "learning_rate": 1.877851398382077e-05, + "loss": 0.5783, + "step": 5842 + }, + { + "epoch": 0.184292698312569, + "grad_norm": 2.919119119644165, + "learning_rate": 1.8778024680794245e-05, + "loss": 0.5806, + "step": 5843 + }, + { + "epoch": 0.18432423907900963, + "grad_norm": 2.2632787227630615, + "learning_rate": 1.8777535286162217e-05, + "loss": 0.5691, + "step": 5844 + }, + { + "epoch": 0.18435577984545023, + "grad_norm": 3.4122233390808105, + "learning_rate": 1.8777045799929794e-05, + "loss": 0.5833, + "step": 5845 + }, + { + "epoch": 0.18438732061189086, + "grad_norm": 2.432349443435669, + "learning_rate": 1.8776556222102083e-05, + "loss": 0.5961, + "step": 5846 + }, + { + "epoch": 0.1844188613783315, + "grad_norm": 2.696739435195923, + "learning_rate": 1.8776066552684192e-05, + "loss": 0.5847, + "step": 5847 + }, + { + "epoch": 0.18445040214477212, + "grad_norm": 3.086099624633789, + "learning_rate": 1.8775576791681233e-05, + "loss": 0.6141, + "step": 5848 + }, + { + "epoch": 0.18448194291121275, + "grad_norm": 2.1432578563690186, + "learning_rate": 1.877508693909831e-05, + "loss": 0.5687, + "step": 5849 + }, + { + "epoch": 0.18451348367765336, + "grad_norm": 2.5212836265563965, + "learning_rate": 1.877459699494055e-05, + "loss": 0.6166, + "step": 5850 + }, + { + "epoch": 0.18454502444409399, + "grad_norm": 2.6767783164978027, + "learning_rate": 1.8774106959213056e-05, + "loss": 0.5872, + "step": 5851 + }, + { + "epoch": 0.18457656521053462, + "grad_norm": 4.902268886566162, + "learning_rate": 1.8773616831920937e-05, + "loss": 0.5599, + "step": 5852 + }, + { + "epoch": 0.18460810597697525, + "grad_norm": 3.8919551372528076, + "learning_rate": 1.8773126613069318e-05, + "loss": 0.4094, + "step": 5853 + }, + { + "epoch": 0.18463964674341588, + "grad_norm": 2.494567394256592, + "learning_rate": 1.877263630266331e-05, + "loss": 0.9035, + "step": 5854 + }, + { + "epoch": 0.18467118750985648, + "grad_norm": 2.731011390686035, + "learning_rate": 1.8772145900708032e-05, + "loss": 0.5868, + "step": 5855 + }, + { + "epoch": 0.1847027282762971, + "grad_norm": 2.8272945880889893, + "learning_rate": 1.87716554072086e-05, + "loss": 0.4267, + "step": 5856 + }, + { + "epoch": 0.18473426904273774, + "grad_norm": 3.1421103477478027, + "learning_rate": 1.877116482217013e-05, + "loss": 0.9209, + "step": 5857 + }, + { + "epoch": 0.18476580980917837, + "grad_norm": 2.130977153778076, + "learning_rate": 1.8770674145597745e-05, + "loss": 0.5795, + "step": 5858 + }, + { + "epoch": 0.184797350575619, + "grad_norm": 1.6433261632919312, + "learning_rate": 1.8770183377496566e-05, + "loss": 0.5567, + "step": 5859 + }, + { + "epoch": 0.18482889134205963, + "grad_norm": 2.226576566696167, + "learning_rate": 1.8769692517871715e-05, + "loss": 0.5822, + "step": 5860 + }, + { + "epoch": 0.18486043210850023, + "grad_norm": 2.3795480728149414, + "learning_rate": 1.876920156672831e-05, + "loss": 0.609, + "step": 5861 + }, + { + "epoch": 0.18489197287494086, + "grad_norm": 2.5873124599456787, + "learning_rate": 1.876871052407148e-05, + "loss": 0.5503, + "step": 5862 + }, + { + "epoch": 0.1849235136413815, + "grad_norm": 2.674715518951416, + "learning_rate": 1.8768219389906345e-05, + "loss": 0.5572, + "step": 5863 + }, + { + "epoch": 0.18495505440782212, + "grad_norm": 2.343304395675659, + "learning_rate": 1.876772816423803e-05, + "loss": 0.5495, + "step": 5864 + }, + { + "epoch": 0.18498659517426275, + "grad_norm": 3.745985746383667, + "learning_rate": 1.8767236847071674e-05, + "loss": 0.5795, + "step": 5865 + }, + { + "epoch": 0.18501813594070335, + "grad_norm": 2.8872456550598145, + "learning_rate": 1.8766745438412382e-05, + "loss": 0.5609, + "step": 5866 + }, + { + "epoch": 0.18504967670714398, + "grad_norm": 3.491236925125122, + "learning_rate": 1.87662539382653e-05, + "loss": 0.5699, + "step": 5867 + }, + { + "epoch": 0.1850812174735846, + "grad_norm": 3.387545347213745, + "learning_rate": 1.876576234663555e-05, + "loss": 0.9197, + "step": 5868 + }, + { + "epoch": 0.18511275824002524, + "grad_norm": 3.1930270195007324, + "learning_rate": 1.8765270663528266e-05, + "loss": 0.582, + "step": 5869 + }, + { + "epoch": 0.18514429900646587, + "grad_norm": 1.8582870960235596, + "learning_rate": 1.8764778888948573e-05, + "loss": 0.5972, + "step": 5870 + }, + { + "epoch": 0.18517583977290647, + "grad_norm": 2.7053134441375732, + "learning_rate": 1.876428702290161e-05, + "loss": 0.542, + "step": 5871 + }, + { + "epoch": 0.1852073805393471, + "grad_norm": 6.92692232131958, + "learning_rate": 1.8763795065392503e-05, + "loss": 0.579, + "step": 5872 + }, + { + "epoch": 0.18523892130578773, + "grad_norm": 2.297301769256592, + "learning_rate": 1.8763303016426393e-05, + "loss": 0.5712, + "step": 5873 + }, + { + "epoch": 0.18527046207222836, + "grad_norm": 3.213775396347046, + "learning_rate": 1.876281087600841e-05, + "loss": 0.9209, + "step": 5874 + }, + { + "epoch": 0.185302002838669, + "grad_norm": 2.430924654006958, + "learning_rate": 1.876231864414369e-05, + "loss": 0.5812, + "step": 5875 + }, + { + "epoch": 0.1853335436051096, + "grad_norm": 2.937680244445801, + "learning_rate": 1.8761826320837373e-05, + "loss": 0.5858, + "step": 5876 + }, + { + "epoch": 0.18536508437155022, + "grad_norm": 2.838921546936035, + "learning_rate": 1.8761333906094593e-05, + "loss": 0.6009, + "step": 5877 + }, + { + "epoch": 0.18539662513799085, + "grad_norm": 2.347559690475464, + "learning_rate": 1.8760841399920494e-05, + "loss": 0.5714, + "step": 5878 + }, + { + "epoch": 0.18542816590443148, + "grad_norm": 2.4781970977783203, + "learning_rate": 1.876034880232021e-05, + "loss": 0.5817, + "step": 5879 + }, + { + "epoch": 0.1854597066708721, + "grad_norm": 3.4034817218780518, + "learning_rate": 1.8759856113298882e-05, + "loss": 0.5932, + "step": 5880 + }, + { + "epoch": 0.18549124743731274, + "grad_norm": 3.239762306213379, + "learning_rate": 1.875936333286166e-05, + "loss": 0.5699, + "step": 5881 + }, + { + "epoch": 0.18552278820375334, + "grad_norm": 1.9361460208892822, + "learning_rate": 1.8758870461013672e-05, + "loss": 0.5838, + "step": 5882 + }, + { + "epoch": 0.18555432897019397, + "grad_norm": 2.228325366973877, + "learning_rate": 1.8758377497760074e-05, + "loss": 0.5719, + "step": 5883 + }, + { + "epoch": 0.1855858697366346, + "grad_norm": 2.4096460342407227, + "learning_rate": 1.8757884443106006e-05, + "loss": 0.5675, + "step": 5884 + }, + { + "epoch": 0.18561741050307523, + "grad_norm": 1.9631234407424927, + "learning_rate": 1.8757391297056616e-05, + "loss": 0.5589, + "step": 5885 + }, + { + "epoch": 0.18564895126951586, + "grad_norm": 57.42019271850586, + "learning_rate": 1.8756898059617043e-05, + "loss": 0.9223, + "step": 5886 + }, + { + "epoch": 0.18568049203595646, + "grad_norm": 2.9815027713775635, + "learning_rate": 1.875640473079244e-05, + "loss": 0.5669, + "step": 5887 + }, + { + "epoch": 0.1857120328023971, + "grad_norm": 4.404913902282715, + "learning_rate": 1.875591131058796e-05, + "loss": 0.5841, + "step": 5888 + }, + { + "epoch": 0.18574357356883772, + "grad_norm": 2.4154741764068604, + "learning_rate": 1.875541779900874e-05, + "loss": 0.6136, + "step": 5889 + }, + { + "epoch": 0.18577511433527835, + "grad_norm": 2.83685302734375, + "learning_rate": 1.875492419605994e-05, + "loss": 0.4449, + "step": 5890 + }, + { + "epoch": 0.18580665510171898, + "grad_norm": 3.838313341140747, + "learning_rate": 1.875443050174671e-05, + "loss": 0.5839, + "step": 5891 + }, + { + "epoch": 0.18583819586815958, + "grad_norm": 3.136934995651245, + "learning_rate": 1.8753936716074195e-05, + "loss": 0.6279, + "step": 5892 + }, + { + "epoch": 0.1858697366346002, + "grad_norm": 3.048215389251709, + "learning_rate": 1.8753442839047555e-05, + "loss": 0.584, + "step": 5893 + }, + { + "epoch": 0.18590127740104084, + "grad_norm": 2.4197778701782227, + "learning_rate": 1.8752948870671942e-05, + "loss": 0.5927, + "step": 5894 + }, + { + "epoch": 0.18593281816748147, + "grad_norm": 1.823521375656128, + "learning_rate": 1.8752454810952513e-05, + "loss": 0.5732, + "step": 5895 + }, + { + "epoch": 0.1859643589339221, + "grad_norm": 3.3619613647460938, + "learning_rate": 1.875196065989442e-05, + "loss": 0.6015, + "step": 5896 + }, + { + "epoch": 0.18599589970036273, + "grad_norm": 2.8333892822265625, + "learning_rate": 1.8751466417502823e-05, + "loss": 0.5775, + "step": 5897 + }, + { + "epoch": 0.18602744046680333, + "grad_norm": 1.824493408203125, + "learning_rate": 1.875097208378288e-05, + "loss": 0.9021, + "step": 5898 + }, + { + "epoch": 0.18605898123324396, + "grad_norm": 2.439018964767456, + "learning_rate": 1.8750477658739746e-05, + "loss": 0.5702, + "step": 5899 + }, + { + "epoch": 0.1860905219996846, + "grad_norm": 2.3545844554901123, + "learning_rate": 1.8749983142378583e-05, + "loss": 0.5493, + "step": 5900 + }, + { + "epoch": 0.18612206276612522, + "grad_norm": 3.623250722885132, + "learning_rate": 1.8749488534704552e-05, + "loss": 0.5812, + "step": 5901 + }, + { + "epoch": 0.18615360353256585, + "grad_norm": 4.513586044311523, + "learning_rate": 1.8748993835722815e-05, + "loss": 0.5769, + "step": 5902 + }, + { + "epoch": 0.18618514429900646, + "grad_norm": 2.1525440216064453, + "learning_rate": 1.8748499045438533e-05, + "loss": 0.5566, + "step": 5903 + }, + { + "epoch": 0.18621668506544709, + "grad_norm": 2.257185697555542, + "learning_rate": 1.874800416385687e-05, + "loss": 0.5478, + "step": 5904 + }, + { + "epoch": 0.18624822583188771, + "grad_norm": 2.491421699523926, + "learning_rate": 1.8747509190982993e-05, + "loss": 0.5449, + "step": 5905 + }, + { + "epoch": 0.18627976659832834, + "grad_norm": 2.7861075401306152, + "learning_rate": 1.8747014126822062e-05, + "loss": 0.5892, + "step": 5906 + }, + { + "epoch": 0.18631130736476897, + "grad_norm": 2.4199576377868652, + "learning_rate": 1.874651897137925e-05, + "loss": 0.5647, + "step": 5907 + }, + { + "epoch": 0.18634284813120958, + "grad_norm": 2.684025764465332, + "learning_rate": 1.874602372465972e-05, + "loss": 0.6131, + "step": 5908 + }, + { + "epoch": 0.1863743888976502, + "grad_norm": 2.4279563426971436, + "learning_rate": 1.874552838666864e-05, + "loss": 0.5634, + "step": 5909 + }, + { + "epoch": 0.18640592966409084, + "grad_norm": 3.0384535789489746, + "learning_rate": 1.874503295741118e-05, + "loss": 0.5707, + "step": 5910 + }, + { + "epoch": 0.18643747043053147, + "grad_norm": 2.2703845500946045, + "learning_rate": 1.8744537436892517e-05, + "loss": 0.5667, + "step": 5911 + }, + { + "epoch": 0.1864690111969721, + "grad_norm": 2.9794888496398926, + "learning_rate": 1.874404182511781e-05, + "loss": 0.5829, + "step": 5912 + }, + { + "epoch": 0.1865005519634127, + "grad_norm": 2.4499893188476562, + "learning_rate": 1.8743546122092237e-05, + "loss": 0.8797, + "step": 5913 + }, + { + "epoch": 0.18653209272985333, + "grad_norm": 4.4063849449157715, + "learning_rate": 1.874305032782097e-05, + "loss": 0.5856, + "step": 5914 + }, + { + "epoch": 0.18656363349629396, + "grad_norm": 1.9742192029953003, + "learning_rate": 1.874255444230919e-05, + "loss": 0.6156, + "step": 5915 + }, + { + "epoch": 0.1865951742627346, + "grad_norm": 2.201780080795288, + "learning_rate": 1.8742058465562064e-05, + "loss": 0.5652, + "step": 5916 + }, + { + "epoch": 0.18662671502917522, + "grad_norm": 2.13741397857666, + "learning_rate": 1.874156239758477e-05, + "loss": 0.5641, + "step": 5917 + }, + { + "epoch": 0.18665825579561585, + "grad_norm": 2.1330270767211914, + "learning_rate": 1.874106623838248e-05, + "loss": 0.549, + "step": 5918 + }, + { + "epoch": 0.18668979656205645, + "grad_norm": 1.5190587043762207, + "learning_rate": 1.8740569987960378e-05, + "loss": 0.9335, + "step": 5919 + }, + { + "epoch": 0.18672133732849708, + "grad_norm": 3.3041810989379883, + "learning_rate": 1.8740073646323645e-05, + "loss": 0.5842, + "step": 5920 + }, + { + "epoch": 0.1867528780949377, + "grad_norm": 2.44262957572937, + "learning_rate": 1.8739577213477456e-05, + "loss": 0.5587, + "step": 5921 + }, + { + "epoch": 0.18678441886137834, + "grad_norm": 2.0953426361083984, + "learning_rate": 1.873908068942699e-05, + "loss": 0.5393, + "step": 5922 + }, + { + "epoch": 0.18681595962781897, + "grad_norm": 1.9225530624389648, + "learning_rate": 1.8738584074177437e-05, + "loss": 0.6059, + "step": 5923 + }, + { + "epoch": 0.18684750039425957, + "grad_norm": 3.071197748184204, + "learning_rate": 1.8738087367733967e-05, + "loss": 0.5793, + "step": 5924 + }, + { + "epoch": 0.1868790411607002, + "grad_norm": 2.7756059169769287, + "learning_rate": 1.8737590570101774e-05, + "loss": 0.5716, + "step": 5925 + }, + { + "epoch": 0.18691058192714083, + "grad_norm": 2.407569408416748, + "learning_rate": 1.8737093681286036e-05, + "loss": 0.5527, + "step": 5926 + }, + { + "epoch": 0.18694212269358146, + "grad_norm": 2.9577770233154297, + "learning_rate": 1.8736596701291945e-05, + "loss": 0.5961, + "step": 5927 + }, + { + "epoch": 0.1869736634600221, + "grad_norm": 2.387122392654419, + "learning_rate": 1.873609963012468e-05, + "loss": 0.5503, + "step": 5928 + }, + { + "epoch": 0.1870052042264627, + "grad_norm": 2.0033586025238037, + "learning_rate": 1.8735602467789435e-05, + "loss": 0.5599, + "step": 5929 + }, + { + "epoch": 0.18703674499290332, + "grad_norm": 2.3308801651000977, + "learning_rate": 1.873510521429139e-05, + "loss": 0.9295, + "step": 5930 + }, + { + "epoch": 0.18706828575934395, + "grad_norm": 2.109679698944092, + "learning_rate": 1.8734607869635745e-05, + "loss": 0.5758, + "step": 5931 + }, + { + "epoch": 0.18709982652578458, + "grad_norm": 3.2415921688079834, + "learning_rate": 1.8734110433827685e-05, + "loss": 0.5785, + "step": 5932 + }, + { + "epoch": 0.1871313672922252, + "grad_norm": 2.5114552974700928, + "learning_rate": 1.87336129068724e-05, + "loss": 0.4305, + "step": 5933 + }, + { + "epoch": 0.18716290805866584, + "grad_norm": 3.0638015270233154, + "learning_rate": 1.873311528877508e-05, + "loss": 0.5661, + "step": 5934 + }, + { + "epoch": 0.18719444882510644, + "grad_norm": 2.0338704586029053, + "learning_rate": 1.8732617579540922e-05, + "loss": 0.4569, + "step": 5935 + }, + { + "epoch": 0.18722598959154707, + "grad_norm": 2.831662654876709, + "learning_rate": 1.873211977917512e-05, + "loss": 0.5806, + "step": 5936 + }, + { + "epoch": 0.1872575303579877, + "grad_norm": 2.8604965209960938, + "learning_rate": 1.8731621887682868e-05, + "loss": 0.5867, + "step": 5937 + }, + { + "epoch": 0.18728907112442833, + "grad_norm": 1.8018860816955566, + "learning_rate": 1.873112390506936e-05, + "loss": 0.539, + "step": 5938 + }, + { + "epoch": 0.18732061189086896, + "grad_norm": 1.8312712907791138, + "learning_rate": 1.8730625831339793e-05, + "loss": 0.5806, + "step": 5939 + }, + { + "epoch": 0.18735215265730956, + "grad_norm": 1.80681574344635, + "learning_rate": 1.873012766649937e-05, + "loss": 0.5661, + "step": 5940 + }, + { + "epoch": 0.1873836934237502, + "grad_norm": 2.8190743923187256, + "learning_rate": 1.872962941055329e-05, + "loss": 0.5516, + "step": 5941 + }, + { + "epoch": 0.18741523419019082, + "grad_norm": 1.927761435508728, + "learning_rate": 1.8729131063506737e-05, + "loss": 0.922, + "step": 5942 + }, + { + "epoch": 0.18744677495663145, + "grad_norm": 2.011341094970703, + "learning_rate": 1.8728632625364933e-05, + "loss": 0.5805, + "step": 5943 + }, + { + "epoch": 0.18747831572307208, + "grad_norm": 2.4520435333251953, + "learning_rate": 1.8728134096133063e-05, + "loss": 0.5875, + "step": 5944 + }, + { + "epoch": 0.18750985648951268, + "grad_norm": 2.763502597808838, + "learning_rate": 1.8727635475816343e-05, + "loss": 0.5955, + "step": 5945 + }, + { + "epoch": 0.1875413972559533, + "grad_norm": 2.3607985973358154, + "learning_rate": 1.8727136764419968e-05, + "loss": 0.5858, + "step": 5946 + }, + { + "epoch": 0.18757293802239394, + "grad_norm": 2.0436270236968994, + "learning_rate": 1.8726637961949144e-05, + "loss": 0.9538, + "step": 5947 + }, + { + "epoch": 0.18760447878883457, + "grad_norm": 2.8630926609039307, + "learning_rate": 1.8726139068409075e-05, + "loss": 0.5804, + "step": 5948 + }, + { + "epoch": 0.1876360195552752, + "grad_norm": 1.8386045694351196, + "learning_rate": 1.872564008380497e-05, + "loss": 0.526, + "step": 5949 + }, + { + "epoch": 0.1876675603217158, + "grad_norm": 1.7866488695144653, + "learning_rate": 1.8725141008142037e-05, + "loss": 0.5437, + "step": 5950 + }, + { + "epoch": 0.18769910108815643, + "grad_norm": 1.9579846858978271, + "learning_rate": 1.872464184142548e-05, + "loss": 0.5623, + "step": 5951 + }, + { + "epoch": 0.18773064185459706, + "grad_norm": 2.4138994216918945, + "learning_rate": 1.8724142583660515e-05, + "loss": 0.5841, + "step": 5952 + }, + { + "epoch": 0.1877621826210377, + "grad_norm": 2.246837615966797, + "learning_rate": 1.8723643234852345e-05, + "loss": 0.5381, + "step": 5953 + }, + { + "epoch": 0.18779372338747832, + "grad_norm": 3.2590057849884033, + "learning_rate": 1.8723143795006186e-05, + "loss": 0.5763, + "step": 5954 + }, + { + "epoch": 0.18782526415391895, + "grad_norm": 3.662635087966919, + "learning_rate": 1.8722644264127244e-05, + "loss": 0.5626, + "step": 5955 + }, + { + "epoch": 0.18785680492035955, + "grad_norm": 2.5741024017333984, + "learning_rate": 1.8722144642220738e-05, + "loss": 0.5803, + "step": 5956 + }, + { + "epoch": 0.18788834568680018, + "grad_norm": 1.779341459274292, + "learning_rate": 1.8721644929291882e-05, + "loss": 0.5324, + "step": 5957 + }, + { + "epoch": 0.18791988645324081, + "grad_norm": 2.041552782058716, + "learning_rate": 1.872114512534589e-05, + "loss": 0.5626, + "step": 5958 + }, + { + "epoch": 0.18795142721968144, + "grad_norm": 1.845535397529602, + "learning_rate": 1.8720645230387972e-05, + "loss": 0.9117, + "step": 5959 + }, + { + "epoch": 0.18798296798612207, + "grad_norm": 5.065467357635498, + "learning_rate": 1.8720145244423355e-05, + "loss": 0.563, + "step": 5960 + }, + { + "epoch": 0.18801450875256268, + "grad_norm": 2.323697090148926, + "learning_rate": 1.8719645167457247e-05, + "loss": 0.5975, + "step": 5961 + }, + { + "epoch": 0.1880460495190033, + "grad_norm": 2.0234079360961914, + "learning_rate": 1.8719144999494875e-05, + "loss": 0.896, + "step": 5962 + }, + { + "epoch": 0.18807759028544394, + "grad_norm": 4.197727680206299, + "learning_rate": 1.871864474054145e-05, + "loss": 0.872, + "step": 5963 + }, + { + "epoch": 0.18810913105188456, + "grad_norm": 2.0438692569732666, + "learning_rate": 1.87181443906022e-05, + "loss": 0.5396, + "step": 5964 + }, + { + "epoch": 0.1881406718183252, + "grad_norm": 4.174474716186523, + "learning_rate": 1.8717643949682345e-05, + "loss": 0.5929, + "step": 5965 + }, + { + "epoch": 0.1881722125847658, + "grad_norm": 3.23115873336792, + "learning_rate": 1.8717143417787104e-05, + "loss": 0.5917, + "step": 5966 + }, + { + "epoch": 0.18820375335120643, + "grad_norm": 2.088151454925537, + "learning_rate": 1.8716642794921703e-05, + "loss": 0.6051, + "step": 5967 + }, + { + "epoch": 0.18823529411764706, + "grad_norm": 1.989108681678772, + "learning_rate": 1.871614208109137e-05, + "loss": 0.9035, + "step": 5968 + }, + { + "epoch": 0.18826683488408769, + "grad_norm": 2.7460994720458984, + "learning_rate": 1.871564127630132e-05, + "loss": 0.5679, + "step": 5969 + }, + { + "epoch": 0.18829837565052832, + "grad_norm": 2.5097975730895996, + "learning_rate": 1.871514038055679e-05, + "loss": 0.5812, + "step": 5970 + }, + { + "epoch": 0.18832991641696895, + "grad_norm": 2.56976056098938, + "learning_rate": 1.8714639393863004e-05, + "loss": 0.5784, + "step": 5971 + }, + { + "epoch": 0.18836145718340955, + "grad_norm": 2.9230270385742188, + "learning_rate": 1.8714138316225187e-05, + "loss": 0.5852, + "step": 5972 + }, + { + "epoch": 0.18839299794985018, + "grad_norm": 3.6019504070281982, + "learning_rate": 1.871363714764857e-05, + "loss": 0.59, + "step": 5973 + }, + { + "epoch": 0.1884245387162908, + "grad_norm": 2.08579683303833, + "learning_rate": 1.8713135888138387e-05, + "loss": 0.5815, + "step": 5974 + }, + { + "epoch": 0.18845607948273144, + "grad_norm": 2.9808273315429688, + "learning_rate": 1.871263453769986e-05, + "loss": 0.5822, + "step": 5975 + }, + { + "epoch": 0.18848762024917207, + "grad_norm": 3.0097811222076416, + "learning_rate": 1.8712133096338228e-05, + "loss": 0.564, + "step": 5976 + }, + { + "epoch": 0.18851916101561267, + "grad_norm": 3.7159321308135986, + "learning_rate": 1.871163156405872e-05, + "loss": 0.5661, + "step": 5977 + }, + { + "epoch": 0.1885507017820533, + "grad_norm": 2.3060302734375, + "learning_rate": 1.8711129940866577e-05, + "loss": 0.5939, + "step": 5978 + }, + { + "epoch": 0.18858224254849393, + "grad_norm": 2.3427364826202393, + "learning_rate": 1.871062822676703e-05, + "loss": 0.5572, + "step": 5979 + }, + { + "epoch": 0.18861378331493456, + "grad_norm": 2.0587706565856934, + "learning_rate": 1.871012642176531e-05, + "loss": 0.5552, + "step": 5980 + }, + { + "epoch": 0.1886453240813752, + "grad_norm": 2.4578657150268555, + "learning_rate": 1.870962452586666e-05, + "loss": 0.5606, + "step": 5981 + }, + { + "epoch": 0.1886768648478158, + "grad_norm": 2.7446742057800293, + "learning_rate": 1.870912253907631e-05, + "loss": 0.5512, + "step": 5982 + }, + { + "epoch": 0.18870840561425642, + "grad_norm": 2.564786672592163, + "learning_rate": 1.8708620461399508e-05, + "loss": 0.595, + "step": 5983 + }, + { + "epoch": 0.18873994638069705, + "grad_norm": 12.11506462097168, + "learning_rate": 1.8708118292841493e-05, + "loss": 0.5973, + "step": 5984 + }, + { + "epoch": 0.18877148714713768, + "grad_norm": 3.0029759407043457, + "learning_rate": 1.87076160334075e-05, + "loss": 0.5658, + "step": 5985 + }, + { + "epoch": 0.1888030279135783, + "grad_norm": 2.430318832397461, + "learning_rate": 1.8707113683102766e-05, + "loss": 0.5755, + "step": 5986 + }, + { + "epoch": 0.18883456868001894, + "grad_norm": 1.7699034214019775, + "learning_rate": 1.8706611241932547e-05, + "loss": 0.888, + "step": 5987 + }, + { + "epoch": 0.18886610944645954, + "grad_norm": 2.565864086151123, + "learning_rate": 1.8706108709902076e-05, + "loss": 0.5772, + "step": 5988 + }, + { + "epoch": 0.18889765021290017, + "grad_norm": 2.6346144676208496, + "learning_rate": 1.87056060870166e-05, + "loss": 0.569, + "step": 5989 + }, + { + "epoch": 0.1889291909793408, + "grad_norm": 3.2378029823303223, + "learning_rate": 1.8705103373281365e-05, + "loss": 0.5732, + "step": 5990 + }, + { + "epoch": 0.18896073174578143, + "grad_norm": 2.528045177459717, + "learning_rate": 1.870460056870162e-05, + "loss": 0.5773, + "step": 5991 + }, + { + "epoch": 0.18899227251222206, + "grad_norm": 3.6550819873809814, + "learning_rate": 1.8704097673282605e-05, + "loss": 0.4833, + "step": 5992 + }, + { + "epoch": 0.18902381327866266, + "grad_norm": 4.73289680480957, + "learning_rate": 1.8703594687029577e-05, + "loss": 0.5689, + "step": 5993 + }, + { + "epoch": 0.1890553540451033, + "grad_norm": 2.1031124591827393, + "learning_rate": 1.8703091609947777e-05, + "loss": 0.414, + "step": 5994 + }, + { + "epoch": 0.18908689481154392, + "grad_norm": 3.975323438644409, + "learning_rate": 1.8702588442042456e-05, + "loss": 0.5879, + "step": 5995 + }, + { + "epoch": 0.18911843557798455, + "grad_norm": 5.582812309265137, + "learning_rate": 1.8702085183318873e-05, + "loss": 0.5761, + "step": 5996 + }, + { + "epoch": 0.18914997634442518, + "grad_norm": 1.9250967502593994, + "learning_rate": 1.870158183378227e-05, + "loss": 0.5822, + "step": 5997 + }, + { + "epoch": 0.18918151711086578, + "grad_norm": 2.4410622119903564, + "learning_rate": 1.8701078393437906e-05, + "loss": 0.5481, + "step": 5998 + }, + { + "epoch": 0.1892130578773064, + "grad_norm": 4.5802435874938965, + "learning_rate": 1.870057486229103e-05, + "loss": 0.5841, + "step": 5999 + }, + { + "epoch": 0.18924459864374704, + "grad_norm": 2.112340211868286, + "learning_rate": 1.87000712403469e-05, + "loss": 0.8941, + "step": 6000 + }, + { + "epoch": 0.18927613941018767, + "grad_norm": 2.971442461013794, + "learning_rate": 1.8699567527610774e-05, + "loss": 0.5554, + "step": 6001 + }, + { + "epoch": 0.1893076801766283, + "grad_norm": 5.237555503845215, + "learning_rate": 1.8699063724087905e-05, + "loss": 0.5789, + "step": 6002 + }, + { + "epoch": 0.1893392209430689, + "grad_norm": 3.25650954246521, + "learning_rate": 1.869855982978355e-05, + "loss": 0.5838, + "step": 6003 + }, + { + "epoch": 0.18937076170950953, + "grad_norm": 1.557605266571045, + "learning_rate": 1.8698055844702973e-05, + "loss": 0.9058, + "step": 6004 + }, + { + "epoch": 0.18940230247595016, + "grad_norm": 1.6964911222457886, + "learning_rate": 1.869755176885142e-05, + "loss": 0.8934, + "step": 6005 + }, + { + "epoch": 0.1894338432423908, + "grad_norm": 7.360222339630127, + "learning_rate": 1.8697047602234167e-05, + "loss": 0.5611, + "step": 6006 + }, + { + "epoch": 0.18946538400883142, + "grad_norm": 3.5027310848236084, + "learning_rate": 1.869654334485647e-05, + "loss": 0.5951, + "step": 6007 + }, + { + "epoch": 0.18949692477527205, + "grad_norm": 3.531411647796631, + "learning_rate": 1.869603899672359e-05, + "loss": 0.5999, + "step": 6008 + }, + { + "epoch": 0.18952846554171265, + "grad_norm": 2.560389995574951, + "learning_rate": 1.8695534557840792e-05, + "loss": 0.573, + "step": 6009 + }, + { + "epoch": 0.18956000630815328, + "grad_norm": 1.88282310962677, + "learning_rate": 1.8695030028213338e-05, + "loss": 0.5507, + "step": 6010 + }, + { + "epoch": 0.1895915470745939, + "grad_norm": 2.2881336212158203, + "learning_rate": 1.869452540784649e-05, + "loss": 0.5502, + "step": 6011 + }, + { + "epoch": 0.18962308784103454, + "grad_norm": 1.9921241998672485, + "learning_rate": 1.8694020696745523e-05, + "loss": 0.5879, + "step": 6012 + }, + { + "epoch": 0.18965462860747517, + "grad_norm": 2.725849151611328, + "learning_rate": 1.86935158949157e-05, + "loss": 0.5665, + "step": 6013 + }, + { + "epoch": 0.18968616937391578, + "grad_norm": 3.7439887523651123, + "learning_rate": 1.8693011002362286e-05, + "loss": 0.5432, + "step": 6014 + }, + { + "epoch": 0.1897177101403564, + "grad_norm": 3.1978650093078613, + "learning_rate": 1.869250601909055e-05, + "loss": 0.4574, + "step": 6015 + }, + { + "epoch": 0.18974925090679703, + "grad_norm": 2.4444549083709717, + "learning_rate": 1.8692000945105766e-05, + "loss": 0.6211, + "step": 6016 + }, + { + "epoch": 0.18978079167323766, + "grad_norm": 3.1750714778900146, + "learning_rate": 1.8691495780413205e-05, + "loss": 0.5752, + "step": 6017 + }, + { + "epoch": 0.1898123324396783, + "grad_norm": 3.412607192993164, + "learning_rate": 1.8690990525018136e-05, + "loss": 0.5711, + "step": 6018 + }, + { + "epoch": 0.1898438732061189, + "grad_norm": 4.042147636413574, + "learning_rate": 1.869048517892583e-05, + "loss": 0.5833, + "step": 6019 + }, + { + "epoch": 0.18987541397255953, + "grad_norm": 2.37540602684021, + "learning_rate": 1.8689979742141566e-05, + "loss": 0.3597, + "step": 6020 + }, + { + "epoch": 0.18990695473900016, + "grad_norm": 2.2399466037750244, + "learning_rate": 1.8689474214670614e-05, + "loss": 0.5898, + "step": 6021 + }, + { + "epoch": 0.18993849550544079, + "grad_norm": 2.2914226055145264, + "learning_rate": 1.8688968596518253e-05, + "loss": 0.5815, + "step": 6022 + }, + { + "epoch": 0.18997003627188142, + "grad_norm": 4.088912487030029, + "learning_rate": 1.8688462887689758e-05, + "loss": 0.5974, + "step": 6023 + }, + { + "epoch": 0.19000157703832204, + "grad_norm": 3.2226932048797607, + "learning_rate": 1.8687957088190404e-05, + "loss": 0.582, + "step": 6024 + }, + { + "epoch": 0.19003311780476265, + "grad_norm": 1.6268174648284912, + "learning_rate": 1.8687451198025474e-05, + "loss": 0.4458, + "step": 6025 + }, + { + "epoch": 0.19006465857120328, + "grad_norm": 1.7650243043899536, + "learning_rate": 1.8686945217200243e-05, + "loss": 0.5662, + "step": 6026 + }, + { + "epoch": 0.1900961993376439, + "grad_norm": 2.303189516067505, + "learning_rate": 1.8686439145719995e-05, + "loss": 0.5691, + "step": 6027 + }, + { + "epoch": 0.19012774010408454, + "grad_norm": 2.2261040210723877, + "learning_rate": 1.868593298359001e-05, + "loss": 0.5623, + "step": 6028 + }, + { + "epoch": 0.19015928087052517, + "grad_norm": 5.224728584289551, + "learning_rate": 1.868542673081557e-05, + "loss": 0.8944, + "step": 6029 + }, + { + "epoch": 0.19019082163696577, + "grad_norm": 2.920222282409668, + "learning_rate": 1.8684920387401956e-05, + "loss": 0.5488, + "step": 6030 + }, + { + "epoch": 0.1902223624034064, + "grad_norm": 2.2084784507751465, + "learning_rate": 1.8684413953354456e-05, + "loss": 0.5717, + "step": 6031 + }, + { + "epoch": 0.19025390316984703, + "grad_norm": 2.823137044906616, + "learning_rate": 1.868390742867835e-05, + "loss": 0.9065, + "step": 6032 + }, + { + "epoch": 0.19028544393628766, + "grad_norm": 2.4006075859069824, + "learning_rate": 1.868340081337893e-05, + "loss": 0.6235, + "step": 6033 + }, + { + "epoch": 0.1903169847027283, + "grad_norm": 3.0638058185577393, + "learning_rate": 1.868289410746148e-05, + "loss": 0.6052, + "step": 6034 + }, + { + "epoch": 0.1903485254691689, + "grad_norm": 2.49587345123291, + "learning_rate": 1.8682387310931287e-05, + "loss": 0.5617, + "step": 6035 + }, + { + "epoch": 0.19038006623560952, + "grad_norm": 1.7950794696807861, + "learning_rate": 1.868188042379364e-05, + "loss": 0.5506, + "step": 6036 + }, + { + "epoch": 0.19041160700205015, + "grad_norm": 2.2068111896514893, + "learning_rate": 1.8681373446053835e-05, + "loss": 0.5704, + "step": 6037 + }, + { + "epoch": 0.19044314776849078, + "grad_norm": 2.0630404949188232, + "learning_rate": 1.868086637771715e-05, + "loss": 0.5476, + "step": 6038 + }, + { + "epoch": 0.1904746885349314, + "grad_norm": 1.8132132291793823, + "learning_rate": 1.8680359218788888e-05, + "loss": 0.5658, + "step": 6039 + }, + { + "epoch": 0.190506229301372, + "grad_norm": 3.618834972381592, + "learning_rate": 1.8679851969274336e-05, + "loss": 0.5863, + "step": 6040 + }, + { + "epoch": 0.19053777006781264, + "grad_norm": 1.7804977893829346, + "learning_rate": 1.8679344629178787e-05, + "loss": 0.5361, + "step": 6041 + }, + { + "epoch": 0.19056931083425327, + "grad_norm": 2.1630585193634033, + "learning_rate": 1.867883719850754e-05, + "loss": 0.5378, + "step": 6042 + }, + { + "epoch": 0.1906008516006939, + "grad_norm": 2.7587804794311523, + "learning_rate": 1.867832967726589e-05, + "loss": 0.5699, + "step": 6043 + }, + { + "epoch": 0.19063239236713453, + "grad_norm": 2.9399755001068115, + "learning_rate": 1.867782206545913e-05, + "loss": 0.5922, + "step": 6044 + }, + { + "epoch": 0.19066393313357516, + "grad_norm": 1.7532591819763184, + "learning_rate": 1.8677314363092555e-05, + "loss": 0.5697, + "step": 6045 + }, + { + "epoch": 0.19069547390001576, + "grad_norm": 2.1022071838378906, + "learning_rate": 1.8676806570171467e-05, + "loss": 0.5594, + "step": 6046 + }, + { + "epoch": 0.1907270146664564, + "grad_norm": 3.250007390975952, + "learning_rate": 1.8676298686701168e-05, + "loss": 0.5745, + "step": 6047 + }, + { + "epoch": 0.19075855543289702, + "grad_norm": 1.9563289880752563, + "learning_rate": 1.8675790712686957e-05, + "loss": 0.5852, + "step": 6048 + }, + { + "epoch": 0.19079009619933765, + "grad_norm": 2.5635781288146973, + "learning_rate": 1.867528264813413e-05, + "loss": 0.9109, + "step": 6049 + }, + { + "epoch": 0.19082163696577828, + "grad_norm": 1.8832125663757324, + "learning_rate": 1.8674774493047994e-05, + "loss": 0.5636, + "step": 6050 + }, + { + "epoch": 0.19085317773221888, + "grad_norm": 2.8443472385406494, + "learning_rate": 1.8674266247433847e-05, + "loss": 0.5671, + "step": 6051 + }, + { + "epoch": 0.1908847184986595, + "grad_norm": 2.5002994537353516, + "learning_rate": 1.8673757911297e-05, + "loss": 0.5659, + "step": 6052 + }, + { + "epoch": 0.19091625926510014, + "grad_norm": 2.1824259757995605, + "learning_rate": 1.8673249484642752e-05, + "loss": 0.5406, + "step": 6053 + }, + { + "epoch": 0.19094780003154077, + "grad_norm": 2.1608335971832275, + "learning_rate": 1.8672740967476412e-05, + "loss": 0.5786, + "step": 6054 + }, + { + "epoch": 0.1909793407979814, + "grad_norm": 1.8513509035110474, + "learning_rate": 1.8672232359803287e-05, + "loss": 0.5318, + "step": 6055 + }, + { + "epoch": 0.191010881564422, + "grad_norm": 2.868067979812622, + "learning_rate": 1.8671723661628683e-05, + "loss": 0.3932, + "step": 6056 + }, + { + "epoch": 0.19104242233086263, + "grad_norm": 3.0698742866516113, + "learning_rate": 1.867121487295791e-05, + "loss": 0.5637, + "step": 6057 + }, + { + "epoch": 0.19107396309730326, + "grad_norm": 6.041662693023682, + "learning_rate": 1.8670705993796273e-05, + "loss": 0.6025, + "step": 6058 + }, + { + "epoch": 0.1911055038637439, + "grad_norm": 2.594372034072876, + "learning_rate": 1.8670197024149087e-05, + "loss": 0.5444, + "step": 6059 + }, + { + "epoch": 0.19113704463018452, + "grad_norm": 1.5938209295272827, + "learning_rate": 1.8669687964021665e-05, + "loss": 0.8815, + "step": 6060 + }, + { + "epoch": 0.19116858539662515, + "grad_norm": 2.5389440059661865, + "learning_rate": 1.8669178813419316e-05, + "loss": 0.6014, + "step": 6061 + }, + { + "epoch": 0.19120012616306575, + "grad_norm": 1.9817326068878174, + "learning_rate": 1.8668669572347357e-05, + "loss": 0.548, + "step": 6062 + }, + { + "epoch": 0.19123166692950638, + "grad_norm": 2.699014186859131, + "learning_rate": 1.8668160240811097e-05, + "loss": 0.5951, + "step": 6063 + }, + { + "epoch": 0.191263207695947, + "grad_norm": 1.8517357110977173, + "learning_rate": 1.8667650818815858e-05, + "loss": 0.5432, + "step": 6064 + }, + { + "epoch": 0.19129474846238764, + "grad_norm": 2.53574538230896, + "learning_rate": 1.866714130636695e-05, + "loss": 0.544, + "step": 6065 + }, + { + "epoch": 0.19132628922882827, + "grad_norm": 3.1919491291046143, + "learning_rate": 1.8666631703469692e-05, + "loss": 0.5543, + "step": 6066 + }, + { + "epoch": 0.19135782999526887, + "grad_norm": 2.7364392280578613, + "learning_rate": 1.86661220101294e-05, + "loss": 0.5879, + "step": 6067 + }, + { + "epoch": 0.1913893707617095, + "grad_norm": 6.465374946594238, + "learning_rate": 1.86656122263514e-05, + "loss": 0.5565, + "step": 6068 + }, + { + "epoch": 0.19142091152815013, + "grad_norm": 1.9668506383895874, + "learning_rate": 1.8665102352141006e-05, + "loss": 0.5568, + "step": 6069 + }, + { + "epoch": 0.19145245229459076, + "grad_norm": 2.298959970474243, + "learning_rate": 1.8664592387503545e-05, + "loss": 0.8789, + "step": 6070 + }, + { + "epoch": 0.1914839930610314, + "grad_norm": 2.3655242919921875, + "learning_rate": 1.8664082332444332e-05, + "loss": 0.5337, + "step": 6071 + }, + { + "epoch": 0.191515533827472, + "grad_norm": 2.553969144821167, + "learning_rate": 1.866357218696869e-05, + "loss": 0.5965, + "step": 6072 + }, + { + "epoch": 0.19154707459391263, + "grad_norm": 2.734684705734253, + "learning_rate": 1.8663061951081943e-05, + "loss": 0.5693, + "step": 6073 + }, + { + "epoch": 0.19157861536035325, + "grad_norm": 3.94311785697937, + "learning_rate": 1.866255162478942e-05, + "loss": 0.5862, + "step": 6074 + }, + { + "epoch": 0.19161015612679388, + "grad_norm": 4.411649703979492, + "learning_rate": 1.866204120809645e-05, + "loss": 0.377, + "step": 6075 + }, + { + "epoch": 0.19164169689323451, + "grad_norm": 3.111626148223877, + "learning_rate": 1.866153070100835e-05, + "loss": 0.6153, + "step": 6076 + }, + { + "epoch": 0.19167323765967512, + "grad_norm": 3.9472010135650635, + "learning_rate": 1.8661020103530453e-05, + "loss": 0.6179, + "step": 6077 + }, + { + "epoch": 0.19170477842611575, + "grad_norm": 2.8721506595611572, + "learning_rate": 1.8660509415668083e-05, + "loss": 0.5505, + "step": 6078 + }, + { + "epoch": 0.19173631919255638, + "grad_norm": 3.2384250164031982, + "learning_rate": 1.8659998637426576e-05, + "loss": 0.5574, + "step": 6079 + }, + { + "epoch": 0.191767859958997, + "grad_norm": 3.6555469036102295, + "learning_rate": 1.8659487768811255e-05, + "loss": 0.5633, + "step": 6080 + }, + { + "epoch": 0.19179940072543764, + "grad_norm": 2.7416672706604004, + "learning_rate": 1.865897680982746e-05, + "loss": 0.5616, + "step": 6081 + }, + { + "epoch": 0.19183094149187827, + "grad_norm": 4.878819942474365, + "learning_rate": 1.8658465760480514e-05, + "loss": 0.5683, + "step": 6082 + }, + { + "epoch": 0.19186248225831887, + "grad_norm": 4.3835320472717285, + "learning_rate": 1.8657954620775757e-05, + "loss": 0.5668, + "step": 6083 + }, + { + "epoch": 0.1918940230247595, + "grad_norm": 2.603762626647949, + "learning_rate": 1.865744339071852e-05, + "loss": 0.6094, + "step": 6084 + }, + { + "epoch": 0.19192556379120013, + "grad_norm": 1.8240375518798828, + "learning_rate": 1.8656932070314138e-05, + "loss": 0.8654, + "step": 6085 + }, + { + "epoch": 0.19195710455764076, + "grad_norm": 2.953071117401123, + "learning_rate": 1.8656420659567952e-05, + "loss": 0.5447, + "step": 6086 + }, + { + "epoch": 0.1919886453240814, + "grad_norm": 3.041304349899292, + "learning_rate": 1.865590915848529e-05, + "loss": 0.5543, + "step": 6087 + }, + { + "epoch": 0.192020186090522, + "grad_norm": 3.2451236248016357, + "learning_rate": 1.8655397567071495e-05, + "loss": 0.5626, + "step": 6088 + }, + { + "epoch": 0.19205172685696262, + "grad_norm": 7.203774452209473, + "learning_rate": 1.8654885885331907e-05, + "loss": 0.5708, + "step": 6089 + }, + { + "epoch": 0.19208326762340325, + "grad_norm": 4.361639976501465, + "learning_rate": 1.8654374113271867e-05, + "loss": 0.5773, + "step": 6090 + }, + { + "epoch": 0.19211480838984388, + "grad_norm": 3.059753179550171, + "learning_rate": 1.865386225089671e-05, + "loss": 0.5764, + "step": 6091 + }, + { + "epoch": 0.1921463491562845, + "grad_norm": 1.9463798999786377, + "learning_rate": 1.8653350298211783e-05, + "loss": 0.8824, + "step": 6092 + }, + { + "epoch": 0.1921778899227251, + "grad_norm": 2.681190252304077, + "learning_rate": 1.8652838255222424e-05, + "loss": 0.6104, + "step": 6093 + }, + { + "epoch": 0.19220943068916574, + "grad_norm": 2.648643732070923, + "learning_rate": 1.8652326121933983e-05, + "loss": 0.588, + "step": 6094 + }, + { + "epoch": 0.19224097145560637, + "grad_norm": 7.45781946182251, + "learning_rate": 1.8651813898351796e-05, + "loss": 0.577, + "step": 6095 + }, + { + "epoch": 0.192272512222047, + "grad_norm": 1.5346462726593018, + "learning_rate": 1.8651301584481216e-05, + "loss": 0.8975, + "step": 6096 + }, + { + "epoch": 0.19230405298848763, + "grad_norm": 2.241776466369629, + "learning_rate": 1.8650789180327583e-05, + "loss": 0.5598, + "step": 6097 + }, + { + "epoch": 0.19233559375492826, + "grad_norm": 2.6247942447662354, + "learning_rate": 1.865027668589625e-05, + "loss": 0.5466, + "step": 6098 + }, + { + "epoch": 0.19236713452136886, + "grad_norm": 3.1980106830596924, + "learning_rate": 1.8649764101192563e-05, + "loss": 0.5834, + "step": 6099 + }, + { + "epoch": 0.1923986752878095, + "grad_norm": 2.697915554046631, + "learning_rate": 1.8649251426221874e-05, + "loss": 0.5496, + "step": 6100 + }, + { + "epoch": 0.19243021605425012, + "grad_norm": 4.403506278991699, + "learning_rate": 1.864873866098953e-05, + "loss": 0.5892, + "step": 6101 + }, + { + "epoch": 0.19246175682069075, + "grad_norm": 4.870048522949219, + "learning_rate": 1.864822580550088e-05, + "loss": 0.5569, + "step": 6102 + }, + { + "epoch": 0.19249329758713138, + "grad_norm": 2.6908488273620605, + "learning_rate": 1.8647712859761275e-05, + "loss": 0.5927, + "step": 6103 + }, + { + "epoch": 0.19252483835357198, + "grad_norm": 3.023332357406616, + "learning_rate": 1.8647199823776076e-05, + "loss": 0.5511, + "step": 6104 + }, + { + "epoch": 0.1925563791200126, + "grad_norm": 2.1249568462371826, + "learning_rate": 1.864668669755063e-05, + "loss": 0.5994, + "step": 6105 + }, + { + "epoch": 0.19258791988645324, + "grad_norm": 2.87420916557312, + "learning_rate": 1.86461734810903e-05, + "loss": 0.587, + "step": 6106 + }, + { + "epoch": 0.19261946065289387, + "grad_norm": 3.610460042953491, + "learning_rate": 1.8645660174400432e-05, + "loss": 0.5824, + "step": 6107 + }, + { + "epoch": 0.1926510014193345, + "grad_norm": 2.6071150302886963, + "learning_rate": 1.8645146777486383e-05, + "loss": 0.545, + "step": 6108 + }, + { + "epoch": 0.1926825421857751, + "grad_norm": 3.5820188522338867, + "learning_rate": 1.864463329035352e-05, + "loss": 0.5873, + "step": 6109 + }, + { + "epoch": 0.19271408295221573, + "grad_norm": 3.4587998390197754, + "learning_rate": 1.8644119713007197e-05, + "loss": 0.5913, + "step": 6110 + }, + { + "epoch": 0.19274562371865636, + "grad_norm": 3.5924875736236572, + "learning_rate": 1.8643606045452772e-05, + "loss": 0.5803, + "step": 6111 + }, + { + "epoch": 0.192777164485097, + "grad_norm": 7.6037068367004395, + "learning_rate": 1.8643092287695604e-05, + "loss": 0.4623, + "step": 6112 + }, + { + "epoch": 0.19280870525153762, + "grad_norm": 5.161551475524902, + "learning_rate": 1.8642578439741058e-05, + "loss": 0.6513, + "step": 6113 + }, + { + "epoch": 0.19284024601797825, + "grad_norm": 3.2860162258148193, + "learning_rate": 1.8642064501594494e-05, + "loss": 0.5849, + "step": 6114 + }, + { + "epoch": 0.19287178678441885, + "grad_norm": 3.3076610565185547, + "learning_rate": 1.864155047326128e-05, + "loss": 0.5488, + "step": 6115 + }, + { + "epoch": 0.19290332755085948, + "grad_norm": 2.920980453491211, + "learning_rate": 1.8641036354746775e-05, + "loss": 0.5686, + "step": 6116 + }, + { + "epoch": 0.1929348683173001, + "grad_norm": 3.8568336963653564, + "learning_rate": 1.8640522146056343e-05, + "loss": 0.3879, + "step": 6117 + }, + { + "epoch": 0.19296640908374074, + "grad_norm": 1.5406743288040161, + "learning_rate": 1.8640007847195358e-05, + "loss": 0.3931, + "step": 6118 + }, + { + "epoch": 0.19299794985018137, + "grad_norm": 1.91502046585083, + "learning_rate": 1.863949345816918e-05, + "loss": 0.951, + "step": 6119 + }, + { + "epoch": 0.19302949061662197, + "grad_norm": 2.349215030670166, + "learning_rate": 1.8638978978983176e-05, + "loss": 0.5982, + "step": 6120 + }, + { + "epoch": 0.1930610313830626, + "grad_norm": 2.985450267791748, + "learning_rate": 1.8638464409642724e-05, + "loss": 0.5722, + "step": 6121 + }, + { + "epoch": 0.19309257214950323, + "grad_norm": 1.7519477605819702, + "learning_rate": 1.8637949750153185e-05, + "loss": 0.8876, + "step": 6122 + }, + { + "epoch": 0.19312411291594386, + "grad_norm": 3.592782497406006, + "learning_rate": 1.8637435000519937e-05, + "loss": 0.546, + "step": 6123 + }, + { + "epoch": 0.1931556536823845, + "grad_norm": 2.390937089920044, + "learning_rate": 1.8636920160748343e-05, + "loss": 0.5884, + "step": 6124 + }, + { + "epoch": 0.1931871944488251, + "grad_norm": 2.92399001121521, + "learning_rate": 1.8636405230843783e-05, + "loss": 0.5874, + "step": 6125 + }, + { + "epoch": 0.19321873521526572, + "grad_norm": 2.7432403564453125, + "learning_rate": 1.8635890210811628e-05, + "loss": 0.5862, + "step": 6126 + }, + { + "epoch": 0.19325027598170635, + "grad_norm": 3.631263017654419, + "learning_rate": 1.8635375100657255e-05, + "loss": 0.5595, + "step": 6127 + }, + { + "epoch": 0.19328181674814698, + "grad_norm": 2.40810489654541, + "learning_rate": 1.8634859900386036e-05, + "loss": 0.5841, + "step": 6128 + }, + { + "epoch": 0.1933133575145876, + "grad_norm": 1.8090327978134155, + "learning_rate": 1.8634344610003346e-05, + "loss": 0.911, + "step": 6129 + }, + { + "epoch": 0.19334489828102822, + "grad_norm": 2.5818612575531006, + "learning_rate": 1.8633829229514572e-05, + "loss": 0.5664, + "step": 6130 + }, + { + "epoch": 0.19337643904746885, + "grad_norm": 3.2491660118103027, + "learning_rate": 1.8633313758925084e-05, + "loss": 0.5797, + "step": 6131 + }, + { + "epoch": 0.19340797981390948, + "grad_norm": 2.8285071849823, + "learning_rate": 1.8632798198240262e-05, + "loss": 0.5303, + "step": 6132 + }, + { + "epoch": 0.1934395205803501, + "grad_norm": 2.650676965713501, + "learning_rate": 1.8632282547465487e-05, + "loss": 0.5677, + "step": 6133 + }, + { + "epoch": 0.19347106134679073, + "grad_norm": 2.287914514541626, + "learning_rate": 1.863176680660614e-05, + "loss": 0.8827, + "step": 6134 + }, + { + "epoch": 0.19350260211323136, + "grad_norm": 2.910663604736328, + "learning_rate": 1.8631250975667608e-05, + "loss": 0.568, + "step": 6135 + }, + { + "epoch": 0.19353414287967197, + "grad_norm": 2.5710151195526123, + "learning_rate": 1.8630735054655265e-05, + "loss": 0.5713, + "step": 6136 + }, + { + "epoch": 0.1935656836461126, + "grad_norm": 11.031912803649902, + "learning_rate": 1.8630219043574504e-05, + "loss": 0.5619, + "step": 6137 + }, + { + "epoch": 0.19359722441255323, + "grad_norm": 2.8942816257476807, + "learning_rate": 1.8629702942430705e-05, + "loss": 0.5358, + "step": 6138 + }, + { + "epoch": 0.19362876517899386, + "grad_norm": 6.47468900680542, + "learning_rate": 1.8629186751229255e-05, + "loss": 0.5789, + "step": 6139 + }, + { + "epoch": 0.19366030594543449, + "grad_norm": 2.479954957962036, + "learning_rate": 1.862867046997554e-05, + "loss": 0.5469, + "step": 6140 + }, + { + "epoch": 0.1936918467118751, + "grad_norm": 4.470119953155518, + "learning_rate": 1.862815409867495e-05, + "loss": 0.5673, + "step": 6141 + }, + { + "epoch": 0.19372338747831572, + "grad_norm": 1.930204153060913, + "learning_rate": 1.862763763733287e-05, + "loss": 0.5381, + "step": 6142 + }, + { + "epoch": 0.19375492824475635, + "grad_norm": 3.165278196334839, + "learning_rate": 1.8627121085954695e-05, + "loss": 0.6115, + "step": 6143 + }, + { + "epoch": 0.19378646901119698, + "grad_norm": 2.195070743560791, + "learning_rate": 1.862660444454581e-05, + "loss": 0.5866, + "step": 6144 + }, + { + "epoch": 0.1938180097776376, + "grad_norm": 3.7760322093963623, + "learning_rate": 1.8626087713111613e-05, + "loss": 0.5871, + "step": 6145 + }, + { + "epoch": 0.1938495505440782, + "grad_norm": 2.3996057510375977, + "learning_rate": 1.862557089165749e-05, + "loss": 0.559, + "step": 6146 + }, + { + "epoch": 0.19388109131051884, + "grad_norm": 2.1472630500793457, + "learning_rate": 1.8625053980188838e-05, + "loss": 0.5782, + "step": 6147 + }, + { + "epoch": 0.19391263207695947, + "grad_norm": 2.615499258041382, + "learning_rate": 1.862453697871105e-05, + "loss": 0.5465, + "step": 6148 + }, + { + "epoch": 0.1939441728434001, + "grad_norm": 2.1515700817108154, + "learning_rate": 1.8624019887229524e-05, + "loss": 0.5653, + "step": 6149 + }, + { + "epoch": 0.19397571360984073, + "grad_norm": 1.7732945680618286, + "learning_rate": 1.862350270574965e-05, + "loss": 0.562, + "step": 6150 + }, + { + "epoch": 0.19400725437628136, + "grad_norm": 2.710406541824341, + "learning_rate": 1.862298543427683e-05, + "loss": 0.5706, + "step": 6151 + }, + { + "epoch": 0.19403879514272196, + "grad_norm": 3.0570485591888428, + "learning_rate": 1.8622468072816464e-05, + "loss": 0.5686, + "step": 6152 + }, + { + "epoch": 0.1940703359091626, + "grad_norm": 2.7815616130828857, + "learning_rate": 1.8621950621373947e-05, + "loss": 0.5724, + "step": 6153 + }, + { + "epoch": 0.19410187667560322, + "grad_norm": 2.1505298614501953, + "learning_rate": 1.862143307995468e-05, + "loss": 0.5861, + "step": 6154 + }, + { + "epoch": 0.19413341744204385, + "grad_norm": 8.014962196350098, + "learning_rate": 1.862091544856407e-05, + "loss": 0.5671, + "step": 6155 + }, + { + "epoch": 0.19416495820848448, + "grad_norm": 2.2119007110595703, + "learning_rate": 1.8620397727207503e-05, + "loss": 0.5526, + "step": 6156 + }, + { + "epoch": 0.19419649897492508, + "grad_norm": 2.358050584793091, + "learning_rate": 1.86198799158904e-05, + "loss": 0.5876, + "step": 6157 + }, + { + "epoch": 0.1942280397413657, + "grad_norm": 2.7445156574249268, + "learning_rate": 1.8619362014618155e-05, + "loss": 0.5832, + "step": 6158 + }, + { + "epoch": 0.19425958050780634, + "grad_norm": 2.659391403198242, + "learning_rate": 1.861884402339617e-05, + "loss": 0.577, + "step": 6159 + }, + { + "epoch": 0.19429112127424697, + "grad_norm": 4.358825206756592, + "learning_rate": 1.861832594222986e-05, + "loss": 0.562, + "step": 6160 + }, + { + "epoch": 0.1943226620406876, + "grad_norm": 3.310258626937866, + "learning_rate": 1.861780777112463e-05, + "loss": 0.8969, + "step": 6161 + }, + { + "epoch": 0.1943542028071282, + "grad_norm": 3.1531739234924316, + "learning_rate": 1.861728951008588e-05, + "loss": 0.6072, + "step": 6162 + }, + { + "epoch": 0.19438574357356883, + "grad_norm": 5.216645240783691, + "learning_rate": 1.8616771159119024e-05, + "loss": 0.4055, + "step": 6163 + }, + { + "epoch": 0.19441728434000946, + "grad_norm": 1.9870848655700684, + "learning_rate": 1.861625271822947e-05, + "loss": 0.3966, + "step": 6164 + }, + { + "epoch": 0.1944488251064501, + "grad_norm": 2.9901201725006104, + "learning_rate": 1.8615734187422626e-05, + "loss": 0.5661, + "step": 6165 + }, + { + "epoch": 0.19448036587289072, + "grad_norm": 3.5335817337036133, + "learning_rate": 1.861521556670391e-05, + "loss": 0.6118, + "step": 6166 + }, + { + "epoch": 0.19451190663933132, + "grad_norm": 2.582751512527466, + "learning_rate": 1.861469685607873e-05, + "loss": 0.5411, + "step": 6167 + }, + { + "epoch": 0.19454344740577195, + "grad_norm": 2.3450701236724854, + "learning_rate": 1.8614178055552495e-05, + "loss": 0.536, + "step": 6168 + }, + { + "epoch": 0.19457498817221258, + "grad_norm": 2.5561625957489014, + "learning_rate": 1.8613659165130627e-05, + "loss": 0.5586, + "step": 6169 + }, + { + "epoch": 0.1946065289386532, + "grad_norm": 2.4276466369628906, + "learning_rate": 1.8613140184818536e-05, + "loss": 0.5629, + "step": 6170 + }, + { + "epoch": 0.19463806970509384, + "grad_norm": 3.84981632232666, + "learning_rate": 1.8612621114621636e-05, + "loss": 0.6024, + "step": 6171 + }, + { + "epoch": 0.19466961047153447, + "grad_norm": 2.588944911956787, + "learning_rate": 1.8612101954545352e-05, + "loss": 0.5804, + "step": 6172 + }, + { + "epoch": 0.19470115123797507, + "grad_norm": 1.859272837638855, + "learning_rate": 1.8611582704595095e-05, + "loss": 0.9099, + "step": 6173 + }, + { + "epoch": 0.1947326920044157, + "grad_norm": 2.731053590774536, + "learning_rate": 1.8611063364776285e-05, + "loss": 0.5782, + "step": 6174 + }, + { + "epoch": 0.19476423277085633, + "grad_norm": 1.9374312162399292, + "learning_rate": 1.8610543935094343e-05, + "loss": 0.5623, + "step": 6175 + }, + { + "epoch": 0.19479577353729696, + "grad_norm": 2.536395788192749, + "learning_rate": 1.8610024415554688e-05, + "loss": 0.59, + "step": 6176 + }, + { + "epoch": 0.1948273143037376, + "grad_norm": 5.371738910675049, + "learning_rate": 1.8609504806162747e-05, + "loss": 0.5887, + "step": 6177 + }, + { + "epoch": 0.1948588550701782, + "grad_norm": 2.394620895385742, + "learning_rate": 1.8608985106923937e-05, + "loss": 0.574, + "step": 6178 + }, + { + "epoch": 0.19489039583661882, + "grad_norm": 2.1820991039276123, + "learning_rate": 1.860846531784368e-05, + "loss": 0.5547, + "step": 6179 + }, + { + "epoch": 0.19492193660305945, + "grad_norm": 2.491528272628784, + "learning_rate": 1.8607945438927403e-05, + "loss": 0.5631, + "step": 6180 + }, + { + "epoch": 0.19495347736950008, + "grad_norm": 2.0627243518829346, + "learning_rate": 1.8607425470180534e-05, + "loss": 0.5334, + "step": 6181 + }, + { + "epoch": 0.1949850181359407, + "grad_norm": 2.632352352142334, + "learning_rate": 1.8606905411608495e-05, + "loss": 0.5375, + "step": 6182 + }, + { + "epoch": 0.19501655890238132, + "grad_norm": 3.056126594543457, + "learning_rate": 1.8606385263216716e-05, + "loss": 0.5643, + "step": 6183 + }, + { + "epoch": 0.19504809966882194, + "grad_norm": 2.8116302490234375, + "learning_rate": 1.8605865025010625e-05, + "loss": 0.5596, + "step": 6184 + }, + { + "epoch": 0.19507964043526257, + "grad_norm": 2.0620529651641846, + "learning_rate": 1.8605344696995646e-05, + "loss": 0.9148, + "step": 6185 + }, + { + "epoch": 0.1951111812017032, + "grad_norm": 2.4144701957702637, + "learning_rate": 1.8604824279177215e-05, + "loss": 0.5303, + "step": 6186 + }, + { + "epoch": 0.19514272196814383, + "grad_norm": 3.1577234268188477, + "learning_rate": 1.8604303771560762e-05, + "loss": 0.5458, + "step": 6187 + }, + { + "epoch": 0.19517426273458446, + "grad_norm": 2.8730762004852295, + "learning_rate": 1.860378317415172e-05, + "loss": 0.5514, + "step": 6188 + }, + { + "epoch": 0.19520580350102507, + "grad_norm": 2.5900678634643555, + "learning_rate": 1.8603262486955516e-05, + "loss": 0.5704, + "step": 6189 + }, + { + "epoch": 0.1952373442674657, + "grad_norm": 2.3500988483428955, + "learning_rate": 1.860274170997759e-05, + "loss": 0.5294, + "step": 6190 + }, + { + "epoch": 0.19526888503390633, + "grad_norm": 2.5026001930236816, + "learning_rate": 1.8602220843223377e-05, + "loss": 0.5597, + "step": 6191 + }, + { + "epoch": 0.19530042580034696, + "grad_norm": 2.8767385482788086, + "learning_rate": 1.8601699886698307e-05, + "loss": 0.5686, + "step": 6192 + }, + { + "epoch": 0.19533196656678758, + "grad_norm": 4.1132283210754395, + "learning_rate": 1.8601178840407815e-05, + "loss": 0.5604, + "step": 6193 + }, + { + "epoch": 0.1953635073332282, + "grad_norm": 4.006298542022705, + "learning_rate": 1.860065770435735e-05, + "loss": 0.5424, + "step": 6194 + }, + { + "epoch": 0.19539504809966882, + "grad_norm": 3.4406754970550537, + "learning_rate": 1.860013647855234e-05, + "loss": 0.3985, + "step": 6195 + }, + { + "epoch": 0.19542658886610945, + "grad_norm": 2.1723475456237793, + "learning_rate": 1.859961516299823e-05, + "loss": 0.5743, + "step": 6196 + }, + { + "epoch": 0.19545812963255008, + "grad_norm": 2.612299919128418, + "learning_rate": 1.859909375770046e-05, + "loss": 0.5541, + "step": 6197 + }, + { + "epoch": 0.1954896703989907, + "grad_norm": 7.647508144378662, + "learning_rate": 1.8598572262664465e-05, + "loss": 0.3965, + "step": 6198 + }, + { + "epoch": 0.1955212111654313, + "grad_norm": 2.351306676864624, + "learning_rate": 1.8598050677895695e-05, + "loss": 0.5566, + "step": 6199 + }, + { + "epoch": 0.19555275193187194, + "grad_norm": 3.0358402729034424, + "learning_rate": 1.859752900339959e-05, + "loss": 0.5602, + "step": 6200 + }, + { + "epoch": 0.19558429269831257, + "grad_norm": 2.690641164779663, + "learning_rate": 1.8597007239181593e-05, + "loss": 0.607, + "step": 6201 + }, + { + "epoch": 0.1956158334647532, + "grad_norm": 2.836129665374756, + "learning_rate": 1.8596485385247152e-05, + "loss": 0.6111, + "step": 6202 + }, + { + "epoch": 0.19564737423119383, + "grad_norm": 2.2803592681884766, + "learning_rate": 1.8595963441601706e-05, + "loss": 0.5468, + "step": 6203 + }, + { + "epoch": 0.19567891499763443, + "grad_norm": 6.908331871032715, + "learning_rate": 1.8595441408250714e-05, + "loss": 0.5849, + "step": 6204 + }, + { + "epoch": 0.19571045576407506, + "grad_norm": 2.2508323192596436, + "learning_rate": 1.8594919285199614e-05, + "loss": 0.5305, + "step": 6205 + }, + { + "epoch": 0.1957419965305157, + "grad_norm": 2.3157026767730713, + "learning_rate": 1.8594397072453854e-05, + "loss": 0.5749, + "step": 6206 + }, + { + "epoch": 0.19577353729695632, + "grad_norm": 2.5361013412475586, + "learning_rate": 1.8593874770018893e-05, + "loss": 0.4479, + "step": 6207 + }, + { + "epoch": 0.19580507806339695, + "grad_norm": 3.03580379486084, + "learning_rate": 1.8593352377900167e-05, + "loss": 0.5583, + "step": 6208 + }, + { + "epoch": 0.19583661882983758, + "grad_norm": 1.849034309387207, + "learning_rate": 1.8592829896103147e-05, + "loss": 0.9109, + "step": 6209 + }, + { + "epoch": 0.19586815959627818, + "grad_norm": 2.584385633468628, + "learning_rate": 1.859230732463327e-05, + "loss": 0.3953, + "step": 6210 + }, + { + "epoch": 0.1958997003627188, + "grad_norm": 2.2561144828796387, + "learning_rate": 1.859178466349599e-05, + "loss": 0.5887, + "step": 6211 + }, + { + "epoch": 0.19593124112915944, + "grad_norm": 2.1053578853607178, + "learning_rate": 1.859126191269677e-05, + "loss": 0.8959, + "step": 6212 + }, + { + "epoch": 0.19596278189560007, + "grad_norm": 2.6234002113342285, + "learning_rate": 1.8590739072241063e-05, + "loss": 0.5987, + "step": 6213 + }, + { + "epoch": 0.1959943226620407, + "grad_norm": 2.538100481033325, + "learning_rate": 1.859021614213432e-05, + "loss": 0.5793, + "step": 6214 + }, + { + "epoch": 0.1960258634284813, + "grad_norm": 2.597109317779541, + "learning_rate": 1.8589693122382004e-05, + "loss": 0.6075, + "step": 6215 + }, + { + "epoch": 0.19605740419492193, + "grad_norm": 2.425053596496582, + "learning_rate": 1.8589170012989565e-05, + "loss": 0.8995, + "step": 6216 + }, + { + "epoch": 0.19608894496136256, + "grad_norm": 1.7085355520248413, + "learning_rate": 1.8588646813962473e-05, + "loss": 0.5602, + "step": 6217 + }, + { + "epoch": 0.1961204857278032, + "grad_norm": 2.1277074813842773, + "learning_rate": 1.858812352530618e-05, + "loss": 0.5621, + "step": 6218 + }, + { + "epoch": 0.19615202649424382, + "grad_norm": 2.9055514335632324, + "learning_rate": 1.858760014702615e-05, + "loss": 0.8797, + "step": 6219 + }, + { + "epoch": 0.19618356726068442, + "grad_norm": 1.9758721590042114, + "learning_rate": 1.8587076679127847e-05, + "loss": 0.5805, + "step": 6220 + }, + { + "epoch": 0.19621510802712505, + "grad_norm": 3.2163279056549072, + "learning_rate": 1.8586553121616727e-05, + "loss": 0.5863, + "step": 6221 + }, + { + "epoch": 0.19624664879356568, + "grad_norm": 2.9011974334716797, + "learning_rate": 1.8586029474498262e-05, + "loss": 0.5859, + "step": 6222 + }, + { + "epoch": 0.1962781895600063, + "grad_norm": 2.1801297664642334, + "learning_rate": 1.858550573777791e-05, + "loss": 0.5624, + "step": 6223 + }, + { + "epoch": 0.19630973032644694, + "grad_norm": 3.88484263420105, + "learning_rate": 1.8584981911461143e-05, + "loss": 0.5472, + "step": 6224 + }, + { + "epoch": 0.19634127109288757, + "grad_norm": 3.2331676483154297, + "learning_rate": 1.858445799555342e-05, + "loss": 0.5645, + "step": 6225 + }, + { + "epoch": 0.19637281185932817, + "grad_norm": 2.1493237018585205, + "learning_rate": 1.858393399006021e-05, + "loss": 0.5705, + "step": 6226 + }, + { + "epoch": 0.1964043526257688, + "grad_norm": 1.9761022329330444, + "learning_rate": 1.8583409894986986e-05, + "loss": 0.5699, + "step": 6227 + }, + { + "epoch": 0.19643589339220943, + "grad_norm": 2.3643767833709717, + "learning_rate": 1.8582885710339214e-05, + "loss": 0.5484, + "step": 6228 + }, + { + "epoch": 0.19646743415865006, + "grad_norm": 2.0829970836639404, + "learning_rate": 1.8582361436122367e-05, + "loss": 0.606, + "step": 6229 + }, + { + "epoch": 0.1964989749250907, + "grad_norm": 2.0128238201141357, + "learning_rate": 1.8581837072341913e-05, + "loss": 0.581, + "step": 6230 + }, + { + "epoch": 0.1965305156915313, + "grad_norm": 3.641944169998169, + "learning_rate": 1.858131261900333e-05, + "loss": 0.5626, + "step": 6231 + }, + { + "epoch": 0.19656205645797192, + "grad_norm": 2.119180679321289, + "learning_rate": 1.858078807611208e-05, + "loss": 0.5403, + "step": 6232 + }, + { + "epoch": 0.19659359722441255, + "grad_norm": 3.024754524230957, + "learning_rate": 1.8580263443673646e-05, + "loss": 0.5743, + "step": 6233 + }, + { + "epoch": 0.19662513799085318, + "grad_norm": 3.0674448013305664, + "learning_rate": 1.8579738721693503e-05, + "loss": 0.5879, + "step": 6234 + }, + { + "epoch": 0.1966566787572938, + "grad_norm": 1.9847136735916138, + "learning_rate": 1.857921391017712e-05, + "loss": 0.5698, + "step": 6235 + }, + { + "epoch": 0.19668821952373441, + "grad_norm": 2.618522882461548, + "learning_rate": 1.857868900912998e-05, + "loss": 0.5588, + "step": 6236 + }, + { + "epoch": 0.19671976029017504, + "grad_norm": 2.406998634338379, + "learning_rate": 1.857816401855756e-05, + "loss": 0.5989, + "step": 6237 + }, + { + "epoch": 0.19675130105661567, + "grad_norm": 2.2389330863952637, + "learning_rate": 1.8577638938465336e-05, + "loss": 0.5589, + "step": 6238 + }, + { + "epoch": 0.1967828418230563, + "grad_norm": 2.1189403533935547, + "learning_rate": 1.8577113768858794e-05, + "loss": 0.5943, + "step": 6239 + }, + { + "epoch": 0.19681438258949693, + "grad_norm": 1.8783679008483887, + "learning_rate": 1.8576588509743406e-05, + "loss": 0.573, + "step": 6240 + }, + { + "epoch": 0.19684592335593754, + "grad_norm": 6.048691749572754, + "learning_rate": 1.857606316112466e-05, + "loss": 0.5954, + "step": 6241 + }, + { + "epoch": 0.19687746412237817, + "grad_norm": 3.3741953372955322, + "learning_rate": 1.857553772300803e-05, + "loss": 0.5827, + "step": 6242 + }, + { + "epoch": 0.1969090048888188, + "grad_norm": 2.3605949878692627, + "learning_rate": 1.857501219539901e-05, + "loss": 0.4057, + "step": 6243 + }, + { + "epoch": 0.19694054565525942, + "grad_norm": 3.483444929122925, + "learning_rate": 1.857448657830308e-05, + "loss": 0.5757, + "step": 6244 + }, + { + "epoch": 0.19697208642170005, + "grad_norm": 2.5832698345184326, + "learning_rate": 1.857396087172572e-05, + "loss": 0.5979, + "step": 6245 + }, + { + "epoch": 0.19700362718814068, + "grad_norm": 2.263256549835205, + "learning_rate": 1.8573435075672422e-05, + "loss": 0.561, + "step": 6246 + }, + { + "epoch": 0.1970351679545813, + "grad_norm": 5.538475036621094, + "learning_rate": 1.8572909190148676e-05, + "loss": 0.5865, + "step": 6247 + }, + { + "epoch": 0.19706670872102192, + "grad_norm": 2.734834671020508, + "learning_rate": 1.8572383215159964e-05, + "loss": 0.5766, + "step": 6248 + }, + { + "epoch": 0.19709824948746255, + "grad_norm": 3.4175894260406494, + "learning_rate": 1.8571857150711773e-05, + "loss": 0.5801, + "step": 6249 + }, + { + "epoch": 0.19712979025390318, + "grad_norm": 2.897380828857422, + "learning_rate": 1.85713309968096e-05, + "loss": 0.3909, + "step": 6250 + }, + { + "epoch": 0.1971613310203438, + "grad_norm": 3.5497775077819824, + "learning_rate": 1.8570804753458933e-05, + "loss": 0.558, + "step": 6251 + }, + { + "epoch": 0.1971928717867844, + "grad_norm": 2.211667537689209, + "learning_rate": 1.8570278420665263e-05, + "loss": 0.5474, + "step": 6252 + }, + { + "epoch": 0.19722441255322504, + "grad_norm": 2.3954501152038574, + "learning_rate": 1.8569751998434085e-05, + "loss": 0.5827, + "step": 6253 + }, + { + "epoch": 0.19725595331966567, + "grad_norm": 3.417677640914917, + "learning_rate": 1.856922548677089e-05, + "loss": 0.5644, + "step": 6254 + }, + { + "epoch": 0.1972874940861063, + "grad_norm": 2.700676918029785, + "learning_rate": 1.8568698885681172e-05, + "loss": 0.5662, + "step": 6255 + }, + { + "epoch": 0.19731903485254693, + "grad_norm": 4.114185333251953, + "learning_rate": 1.8568172195170424e-05, + "loss": 0.5942, + "step": 6256 + }, + { + "epoch": 0.19735057561898753, + "grad_norm": 1.8206816911697388, + "learning_rate": 1.856764541524415e-05, + "loss": 0.931, + "step": 6257 + }, + { + "epoch": 0.19738211638542816, + "grad_norm": 3.0051136016845703, + "learning_rate": 1.8567118545907845e-05, + "loss": 0.5882, + "step": 6258 + }, + { + "epoch": 0.1974136571518688, + "grad_norm": 5.425649166107178, + "learning_rate": 1.8566591587167005e-05, + "loss": 0.5842, + "step": 6259 + }, + { + "epoch": 0.19744519791830942, + "grad_norm": 2.7144737243652344, + "learning_rate": 1.8566064539027126e-05, + "loss": 0.5883, + "step": 6260 + }, + { + "epoch": 0.19747673868475005, + "grad_norm": 3.669083833694458, + "learning_rate": 1.8565537401493717e-05, + "loss": 0.61, + "step": 6261 + }, + { + "epoch": 0.19750827945119068, + "grad_norm": 2.556612730026245, + "learning_rate": 1.856501017457227e-05, + "loss": 0.5694, + "step": 6262 + }, + { + "epoch": 0.19753982021763128, + "grad_norm": 3.4685258865356445, + "learning_rate": 1.8564482858268296e-05, + "loss": 0.5718, + "step": 6263 + }, + { + "epoch": 0.1975713609840719, + "grad_norm": 2.101456880569458, + "learning_rate": 1.856395545258729e-05, + "loss": 0.5681, + "step": 6264 + }, + { + "epoch": 0.19760290175051254, + "grad_norm": 2.4975743293762207, + "learning_rate": 1.8563427957534758e-05, + "loss": 0.5496, + "step": 6265 + }, + { + "epoch": 0.19763444251695317, + "grad_norm": 3.8908371925354004, + "learning_rate": 1.856290037311621e-05, + "loss": 0.5537, + "step": 6266 + }, + { + "epoch": 0.1976659832833938, + "grad_norm": 7.810050010681152, + "learning_rate": 1.8562372699337144e-05, + "loss": 0.5922, + "step": 6267 + }, + { + "epoch": 0.1976975240498344, + "grad_norm": 2.5134036540985107, + "learning_rate": 1.8561844936203073e-05, + "loss": 0.5706, + "step": 6268 + }, + { + "epoch": 0.19772906481627503, + "grad_norm": 2.6809749603271484, + "learning_rate": 1.85613170837195e-05, + "loss": 0.5658, + "step": 6269 + }, + { + "epoch": 0.19776060558271566, + "grad_norm": 2.9711568355560303, + "learning_rate": 1.8560789141891937e-05, + "loss": 0.5905, + "step": 6270 + }, + { + "epoch": 0.1977921463491563, + "grad_norm": 3.0017130374908447, + "learning_rate": 1.856026111072589e-05, + "loss": 0.5644, + "step": 6271 + }, + { + "epoch": 0.19782368711559692, + "grad_norm": 3.9417316913604736, + "learning_rate": 1.8559732990226874e-05, + "loss": 0.5464, + "step": 6272 + }, + { + "epoch": 0.19785522788203752, + "grad_norm": 2.750772714614868, + "learning_rate": 1.85592047804004e-05, + "loss": 0.6249, + "step": 6273 + }, + { + "epoch": 0.19788676864847815, + "grad_norm": 3.4859793186187744, + "learning_rate": 1.855867648125197e-05, + "loss": 0.5409, + "step": 6274 + }, + { + "epoch": 0.19791830941491878, + "grad_norm": 2.286681890487671, + "learning_rate": 1.8558148092787115e-05, + "loss": 0.569, + "step": 6275 + }, + { + "epoch": 0.1979498501813594, + "grad_norm": 3.465344190597534, + "learning_rate": 1.8557619615011333e-05, + "loss": 0.5439, + "step": 6276 + }, + { + "epoch": 0.19798139094780004, + "grad_norm": 3.158324718475342, + "learning_rate": 1.8557091047930146e-05, + "loss": 0.5427, + "step": 6277 + }, + { + "epoch": 0.19801293171424067, + "grad_norm": 4.189657211303711, + "learning_rate": 1.855656239154907e-05, + "loss": 0.5816, + "step": 6278 + }, + { + "epoch": 0.19804447248068127, + "grad_norm": 2.274535655975342, + "learning_rate": 1.8556033645873624e-05, + "loss": 0.5721, + "step": 6279 + }, + { + "epoch": 0.1980760132471219, + "grad_norm": 2.306346893310547, + "learning_rate": 1.855550481090932e-05, + "loss": 0.9411, + "step": 6280 + }, + { + "epoch": 0.19810755401356253, + "grad_norm": 2.628539800643921, + "learning_rate": 1.855497588666168e-05, + "loss": 0.5413, + "step": 6281 + }, + { + "epoch": 0.19813909478000316, + "grad_norm": 5.583720684051514, + "learning_rate": 1.8554446873136228e-05, + "loss": 0.5639, + "step": 6282 + }, + { + "epoch": 0.1981706355464438, + "grad_norm": 3.72397518157959, + "learning_rate": 1.8553917770338476e-05, + "loss": 0.5854, + "step": 6283 + }, + { + "epoch": 0.1982021763128844, + "grad_norm": 3.4091973304748535, + "learning_rate": 1.855338857827395e-05, + "loss": 0.5658, + "step": 6284 + }, + { + "epoch": 0.19823371707932502, + "grad_norm": 3.8947556018829346, + "learning_rate": 1.8552859296948177e-05, + "loss": 0.5957, + "step": 6285 + }, + { + "epoch": 0.19826525784576565, + "grad_norm": 2.9942147731781006, + "learning_rate": 1.855232992636667e-05, + "loss": 0.5799, + "step": 6286 + }, + { + "epoch": 0.19829679861220628, + "grad_norm": 2.2430059909820557, + "learning_rate": 1.8551800466534967e-05, + "loss": 0.8939, + "step": 6287 + }, + { + "epoch": 0.1983283393786469, + "grad_norm": 5.7316060066223145, + "learning_rate": 1.855127091745858e-05, + "loss": 0.584, + "step": 6288 + }, + { + "epoch": 0.1983598801450875, + "grad_norm": 3.042085886001587, + "learning_rate": 1.8550741279143042e-05, + "loss": 0.5506, + "step": 6289 + }, + { + "epoch": 0.19839142091152814, + "grad_norm": 2.476465940475464, + "learning_rate": 1.8550211551593884e-05, + "loss": 0.5599, + "step": 6290 + }, + { + "epoch": 0.19842296167796877, + "grad_norm": 3.114607572555542, + "learning_rate": 1.8549681734816624e-05, + "loss": 0.5966, + "step": 6291 + }, + { + "epoch": 0.1984545024444094, + "grad_norm": 3.01440167427063, + "learning_rate": 1.8549151828816797e-05, + "loss": 0.5903, + "step": 6292 + }, + { + "epoch": 0.19848604321085003, + "grad_norm": 2.8171937465667725, + "learning_rate": 1.8548621833599934e-05, + "loss": 0.562, + "step": 6293 + }, + { + "epoch": 0.19851758397729063, + "grad_norm": 2.313491106033325, + "learning_rate": 1.8548091749171563e-05, + "loss": 0.8891, + "step": 6294 + }, + { + "epoch": 0.19854912474373126, + "grad_norm": 3.6350104808807373, + "learning_rate": 1.854756157553722e-05, + "loss": 0.3968, + "step": 6295 + }, + { + "epoch": 0.1985806655101719, + "grad_norm": 2.2087061405181885, + "learning_rate": 1.854703131270243e-05, + "loss": 0.567, + "step": 6296 + }, + { + "epoch": 0.19861220627661252, + "grad_norm": 3.164982795715332, + "learning_rate": 1.8546500960672737e-05, + "loss": 0.5867, + "step": 6297 + }, + { + "epoch": 0.19864374704305315, + "grad_norm": 2.6455001831054688, + "learning_rate": 1.8545970519453668e-05, + "loss": 0.5628, + "step": 6298 + }, + { + "epoch": 0.19867528780949378, + "grad_norm": 2.9619743824005127, + "learning_rate": 1.854543998905076e-05, + "loss": 0.5463, + "step": 6299 + }, + { + "epoch": 0.19870682857593439, + "grad_norm": 2.357942581176758, + "learning_rate": 1.854490936946955e-05, + "loss": 0.5707, + "step": 6300 + }, + { + "epoch": 0.19873836934237502, + "grad_norm": 2.3767457008361816, + "learning_rate": 1.854437866071558e-05, + "loss": 0.5819, + "step": 6301 + }, + { + "epoch": 0.19876991010881565, + "grad_norm": 4.665820121765137, + "learning_rate": 1.8543847862794383e-05, + "loss": 0.5485, + "step": 6302 + }, + { + "epoch": 0.19880145087525627, + "grad_norm": 3.061347723007202, + "learning_rate": 1.8543316975711495e-05, + "loss": 0.5415, + "step": 6303 + }, + { + "epoch": 0.1988329916416969, + "grad_norm": 2.5856637954711914, + "learning_rate": 1.8542785999472464e-05, + "loss": 0.5492, + "step": 6304 + }, + { + "epoch": 0.1988645324081375, + "grad_norm": 2.5066888332366943, + "learning_rate": 1.8542254934082826e-05, + "loss": 0.4107, + "step": 6305 + }, + { + "epoch": 0.19889607317457814, + "grad_norm": 3.427156925201416, + "learning_rate": 1.8541723779548127e-05, + "loss": 0.5516, + "step": 6306 + }, + { + "epoch": 0.19892761394101877, + "grad_norm": 3.3668112754821777, + "learning_rate": 1.8541192535873906e-05, + "loss": 0.5735, + "step": 6307 + }, + { + "epoch": 0.1989591547074594, + "grad_norm": 2.8600406646728516, + "learning_rate": 1.854066120306571e-05, + "loss": 0.5249, + "step": 6308 + }, + { + "epoch": 0.19899069547390003, + "grad_norm": 3.061807155609131, + "learning_rate": 1.8540129781129083e-05, + "loss": 0.5876, + "step": 6309 + }, + { + "epoch": 0.19902223624034063, + "grad_norm": 2.8675153255462646, + "learning_rate": 1.853959827006957e-05, + "loss": 0.5866, + "step": 6310 + }, + { + "epoch": 0.19905377700678126, + "grad_norm": 2.2617416381835938, + "learning_rate": 1.8539066669892717e-05, + "loss": 0.5745, + "step": 6311 + }, + { + "epoch": 0.1990853177732219, + "grad_norm": 4.829782962799072, + "learning_rate": 1.853853498060407e-05, + "loss": 0.5783, + "step": 6312 + }, + { + "epoch": 0.19911685853966252, + "grad_norm": 2.4087367057800293, + "learning_rate": 1.8538003202209186e-05, + "loss": 0.5748, + "step": 6313 + }, + { + "epoch": 0.19914839930610315, + "grad_norm": 3.2623980045318604, + "learning_rate": 1.853747133471361e-05, + "loss": 0.4167, + "step": 6314 + }, + { + "epoch": 0.19917994007254378, + "grad_norm": 1.952483057975769, + "learning_rate": 1.8536939378122886e-05, + "loss": 0.4192, + "step": 6315 + }, + { + "epoch": 0.19921148083898438, + "grad_norm": 2.2705235481262207, + "learning_rate": 1.8536407332442573e-05, + "loss": 0.8562, + "step": 6316 + }, + { + "epoch": 0.199243021605425, + "grad_norm": 3.325225353240967, + "learning_rate": 1.8535875197678223e-05, + "loss": 0.5775, + "step": 6317 + }, + { + "epoch": 0.19927456237186564, + "grad_norm": 2.654099225997925, + "learning_rate": 1.8535342973835386e-05, + "loss": 0.5791, + "step": 6318 + }, + { + "epoch": 0.19930610313830627, + "grad_norm": 2.336137294769287, + "learning_rate": 1.8534810660919617e-05, + "loss": 0.5671, + "step": 6319 + }, + { + "epoch": 0.1993376439047469, + "grad_norm": 1.581402063369751, + "learning_rate": 1.8534278258936474e-05, + "loss": 0.8936, + "step": 6320 + }, + { + "epoch": 0.1993691846711875, + "grad_norm": 2.0056912899017334, + "learning_rate": 1.853374576789151e-05, + "loss": 0.5601, + "step": 6321 + }, + { + "epoch": 0.19940072543762813, + "grad_norm": 2.2995803356170654, + "learning_rate": 1.853321318779028e-05, + "loss": 0.5487, + "step": 6322 + }, + { + "epoch": 0.19943226620406876, + "grad_norm": 2.4301888942718506, + "learning_rate": 1.8532680518638346e-05, + "loss": 0.606, + "step": 6323 + }, + { + "epoch": 0.1994638069705094, + "grad_norm": 3.248619794845581, + "learning_rate": 1.8532147760441267e-05, + "loss": 0.5699, + "step": 6324 + }, + { + "epoch": 0.19949534773695002, + "grad_norm": 3.1647846698760986, + "learning_rate": 1.85316149132046e-05, + "loss": 0.5568, + "step": 6325 + }, + { + "epoch": 0.19952688850339062, + "grad_norm": 2.016409158706665, + "learning_rate": 1.8531081976933904e-05, + "loss": 0.5608, + "step": 6326 + }, + { + "epoch": 0.19955842926983125, + "grad_norm": 2.3818089962005615, + "learning_rate": 1.853054895163475e-05, + "loss": 0.585, + "step": 6327 + }, + { + "epoch": 0.19958997003627188, + "grad_norm": 2.8259377479553223, + "learning_rate": 1.8530015837312687e-05, + "loss": 0.5607, + "step": 6328 + }, + { + "epoch": 0.1996215108027125, + "grad_norm": 3.5254087448120117, + "learning_rate": 1.8529482633973285e-05, + "loss": 0.5657, + "step": 6329 + }, + { + "epoch": 0.19965305156915314, + "grad_norm": 2.588325023651123, + "learning_rate": 1.8528949341622112e-05, + "loss": 0.6012, + "step": 6330 + }, + { + "epoch": 0.19968459233559374, + "grad_norm": 2.5120255947113037, + "learning_rate": 1.852841596026473e-05, + "loss": 0.5367, + "step": 6331 + }, + { + "epoch": 0.19971613310203437, + "grad_norm": 1.7711738348007202, + "learning_rate": 1.8527882489906706e-05, + "loss": 0.5522, + "step": 6332 + }, + { + "epoch": 0.199747673868475, + "grad_norm": 2.3539741039276123, + "learning_rate": 1.8527348930553607e-05, + "loss": 0.574, + "step": 6333 + }, + { + "epoch": 0.19977921463491563, + "grad_norm": 2.0306015014648438, + "learning_rate": 1.8526815282211e-05, + "loss": 0.5344, + "step": 6334 + }, + { + "epoch": 0.19981075540135626, + "grad_norm": 2.7450740337371826, + "learning_rate": 1.852628154488445e-05, + "loss": 0.5785, + "step": 6335 + }, + { + "epoch": 0.1998422961677969, + "grad_norm": 2.510629892349243, + "learning_rate": 1.8525747718579537e-05, + "loss": 0.5821, + "step": 6336 + }, + { + "epoch": 0.1998738369342375, + "grad_norm": 1.9515361785888672, + "learning_rate": 1.8525213803301824e-05, + "loss": 0.5699, + "step": 6337 + }, + { + "epoch": 0.19990537770067812, + "grad_norm": 2.297170639038086, + "learning_rate": 1.8524679799056888e-05, + "loss": 0.5396, + "step": 6338 + }, + { + "epoch": 0.19993691846711875, + "grad_norm": 2.399383544921875, + "learning_rate": 1.8524145705850296e-05, + "loss": 0.5996, + "step": 6339 + }, + { + "epoch": 0.19996845923355938, + "grad_norm": 2.9128048419952393, + "learning_rate": 1.8523611523687623e-05, + "loss": 0.4566, + "step": 6340 + }, + { + "epoch": 0.2, + "grad_norm": 2.438721179962158, + "learning_rate": 1.852307725257445e-05, + "loss": 0.5939, + "step": 6341 + }, + { + "epoch": 0.2000315407664406, + "grad_norm": 2.1698877811431885, + "learning_rate": 1.8522542892516346e-05, + "loss": 0.5671, + "step": 6342 + }, + { + "epoch": 0.20006308153288124, + "grad_norm": 2.7116262912750244, + "learning_rate": 1.8522008443518887e-05, + "loss": 0.5646, + "step": 6343 + }, + { + "epoch": 0.20009462229932187, + "grad_norm": 1.9866117238998413, + "learning_rate": 1.852147390558766e-05, + "loss": 0.8849, + "step": 6344 + }, + { + "epoch": 0.2001261630657625, + "grad_norm": 2.0324058532714844, + "learning_rate": 1.852093927872823e-05, + "loss": 0.5607, + "step": 6345 + }, + { + "epoch": 0.20015770383220313, + "grad_norm": 4.920982360839844, + "learning_rate": 1.8520404562946185e-05, + "loss": 0.5445, + "step": 6346 + }, + { + "epoch": 0.20018924459864373, + "grad_norm": 1.696325659751892, + "learning_rate": 1.8519869758247097e-05, + "loss": 0.5575, + "step": 6347 + }, + { + "epoch": 0.20022078536508436, + "grad_norm": 2.439997434616089, + "learning_rate": 1.851933486463656e-05, + "loss": 0.5651, + "step": 6348 + }, + { + "epoch": 0.200252326131525, + "grad_norm": 2.2827064990997314, + "learning_rate": 1.8518799882120146e-05, + "loss": 0.5658, + "step": 6349 + }, + { + "epoch": 0.20028386689796562, + "grad_norm": 1.8621550798416138, + "learning_rate": 1.8518264810703437e-05, + "loss": 0.5439, + "step": 6350 + }, + { + "epoch": 0.20031540766440625, + "grad_norm": 3.463343858718872, + "learning_rate": 1.8517729650392022e-05, + "loss": 0.5739, + "step": 6351 + }, + { + "epoch": 0.20034694843084688, + "grad_norm": 3.171424627304077, + "learning_rate": 1.8517194401191483e-05, + "loss": 0.556, + "step": 6352 + }, + { + "epoch": 0.20037848919728748, + "grad_norm": 2.7698752880096436, + "learning_rate": 1.8516659063107413e-05, + "loss": 0.5439, + "step": 6353 + }, + { + "epoch": 0.20041002996372811, + "grad_norm": 2.1024551391601562, + "learning_rate": 1.851612363614539e-05, + "loss": 0.921, + "step": 6354 + }, + { + "epoch": 0.20044157073016874, + "grad_norm": 1.8210679292678833, + "learning_rate": 1.8515588120311e-05, + "loss": 0.9048, + "step": 6355 + }, + { + "epoch": 0.20047311149660937, + "grad_norm": 11.785433769226074, + "learning_rate": 1.8515052515609843e-05, + "loss": 0.5622, + "step": 6356 + }, + { + "epoch": 0.20050465226305, + "grad_norm": 4.289825916290283, + "learning_rate": 1.8514516822047495e-05, + "loss": 0.5949, + "step": 6357 + }, + { + "epoch": 0.2005361930294906, + "grad_norm": 1.9475111961364746, + "learning_rate": 1.8513981039629555e-05, + "loss": 0.5469, + "step": 6358 + }, + { + "epoch": 0.20056773379593124, + "grad_norm": 2.0356976985931396, + "learning_rate": 1.8513445168361613e-05, + "loss": 0.565, + "step": 6359 + }, + { + "epoch": 0.20059927456237187, + "grad_norm": 2.341898202896118, + "learning_rate": 1.8512909208249255e-05, + "loss": 0.5556, + "step": 6360 + }, + { + "epoch": 0.2006308153288125, + "grad_norm": 2.614562749862671, + "learning_rate": 1.8512373159298083e-05, + "loss": 0.8631, + "step": 6361 + }, + { + "epoch": 0.20066235609525312, + "grad_norm": 2.4738378524780273, + "learning_rate": 1.851183702151369e-05, + "loss": 0.5817, + "step": 6362 + }, + { + "epoch": 0.20069389686169373, + "grad_norm": 5.024857521057129, + "learning_rate": 1.8511300794901664e-05, + "loss": 0.5474, + "step": 6363 + }, + { + "epoch": 0.20072543762813436, + "grad_norm": 2.712676525115967, + "learning_rate": 1.8510764479467608e-05, + "loss": 0.5616, + "step": 6364 + }, + { + "epoch": 0.200756978394575, + "grad_norm": 1.989434838294983, + "learning_rate": 1.851022807521712e-05, + "loss": 0.5756, + "step": 6365 + }, + { + "epoch": 0.20078851916101562, + "grad_norm": 2.0902516841888428, + "learning_rate": 1.8509691582155786e-05, + "loss": 0.5533, + "step": 6366 + }, + { + "epoch": 0.20082005992745625, + "grad_norm": 4.118203163146973, + "learning_rate": 1.8509155000289217e-05, + "loss": 0.4316, + "step": 6367 + }, + { + "epoch": 0.20085160069389685, + "grad_norm": 2.510035514831543, + "learning_rate": 1.8508618329623007e-05, + "loss": 0.5624, + "step": 6368 + }, + { + "epoch": 0.20088314146033748, + "grad_norm": 2.932861566543579, + "learning_rate": 1.850808157016276e-05, + "loss": 0.5656, + "step": 6369 + }, + { + "epoch": 0.2009146822267781, + "grad_norm": 1.915759563446045, + "learning_rate": 1.8507544721914074e-05, + "loss": 0.5513, + "step": 6370 + }, + { + "epoch": 0.20094622299321874, + "grad_norm": 2.6470930576324463, + "learning_rate": 1.8507007784882552e-05, + "loss": 0.5687, + "step": 6371 + }, + { + "epoch": 0.20097776375965937, + "grad_norm": 2.3530828952789307, + "learning_rate": 1.85064707590738e-05, + "loss": 0.8836, + "step": 6372 + }, + { + "epoch": 0.2010093045261, + "grad_norm": 4.184006690979004, + "learning_rate": 1.850593364449342e-05, + "loss": 0.5659, + "step": 6373 + }, + { + "epoch": 0.2010408452925406, + "grad_norm": 2.020746946334839, + "learning_rate": 1.8505396441147017e-05, + "loss": 0.3943, + "step": 6374 + }, + { + "epoch": 0.20107238605898123, + "grad_norm": 2.198843240737915, + "learning_rate": 1.8504859149040198e-05, + "loss": 0.534, + "step": 6375 + }, + { + "epoch": 0.20110392682542186, + "grad_norm": 3.085946559906006, + "learning_rate": 1.850432176817857e-05, + "loss": 0.5682, + "step": 6376 + }, + { + "epoch": 0.2011354675918625, + "grad_norm": 4.68485689163208, + "learning_rate": 1.8503784298567744e-05, + "loss": 0.8547, + "step": 6377 + }, + { + "epoch": 0.20116700835830312, + "grad_norm": 3.7143168449401855, + "learning_rate": 1.8503246740213323e-05, + "loss": 0.5576, + "step": 6378 + }, + { + "epoch": 0.20119854912474372, + "grad_norm": 2.171377658843994, + "learning_rate": 1.8502709093120918e-05, + "loss": 0.5701, + "step": 6379 + }, + { + "epoch": 0.20123008989118435, + "grad_norm": 2.944169044494629, + "learning_rate": 1.8502171357296144e-05, + "loss": 0.5694, + "step": 6380 + }, + { + "epoch": 0.20126163065762498, + "grad_norm": 2.239287853240967, + "learning_rate": 1.8501633532744607e-05, + "loss": 0.5431, + "step": 6381 + }, + { + "epoch": 0.2012931714240656, + "grad_norm": 2.3582980632781982, + "learning_rate": 1.8501095619471926e-05, + "loss": 0.3963, + "step": 6382 + }, + { + "epoch": 0.20132471219050624, + "grad_norm": 3.2289369106292725, + "learning_rate": 1.850055761748371e-05, + "loss": 0.5721, + "step": 6383 + }, + { + "epoch": 0.20135625295694684, + "grad_norm": 2.607476234436035, + "learning_rate": 1.8500019526785576e-05, + "loss": 0.581, + "step": 6384 + }, + { + "epoch": 0.20138779372338747, + "grad_norm": 2.1132543087005615, + "learning_rate": 1.8499481347383136e-05, + "loss": 0.5717, + "step": 6385 + }, + { + "epoch": 0.2014193344898281, + "grad_norm": 2.211068630218506, + "learning_rate": 1.8498943079282014e-05, + "loss": 0.5846, + "step": 6386 + }, + { + "epoch": 0.20145087525626873, + "grad_norm": 3.577618360519409, + "learning_rate": 1.8498404722487815e-05, + "loss": 0.5785, + "step": 6387 + }, + { + "epoch": 0.20148241602270936, + "grad_norm": 2.6560943126678467, + "learning_rate": 1.8497866277006165e-05, + "loss": 0.582, + "step": 6388 + }, + { + "epoch": 0.20151395678915, + "grad_norm": 2.52964186668396, + "learning_rate": 1.8497327742842684e-05, + "loss": 0.5622, + "step": 6389 + }, + { + "epoch": 0.2015454975555906, + "grad_norm": 3.473381519317627, + "learning_rate": 1.849678912000299e-05, + "loss": 0.5663, + "step": 6390 + }, + { + "epoch": 0.20157703832203122, + "grad_norm": 2.306671619415283, + "learning_rate": 1.8496250408492702e-05, + "loss": 0.5613, + "step": 6391 + }, + { + "epoch": 0.20160857908847185, + "grad_norm": 3.501688003540039, + "learning_rate": 1.8495711608317446e-05, + "loss": 0.5734, + "step": 6392 + }, + { + "epoch": 0.20164011985491248, + "grad_norm": 3.2661163806915283, + "learning_rate": 1.8495172719482842e-05, + "loss": 0.5796, + "step": 6393 + }, + { + "epoch": 0.2016716606213531, + "grad_norm": 2.023251533508301, + "learning_rate": 1.8494633741994513e-05, + "loss": 0.6, + "step": 6394 + }, + { + "epoch": 0.2017032013877937, + "grad_norm": 2.2858633995056152, + "learning_rate": 1.8494094675858085e-05, + "loss": 0.5466, + "step": 6395 + }, + { + "epoch": 0.20173474215423434, + "grad_norm": 3.2170205116271973, + "learning_rate": 1.8493555521079184e-05, + "loss": 0.5825, + "step": 6396 + }, + { + "epoch": 0.20176628292067497, + "grad_norm": 1.8994373083114624, + "learning_rate": 1.8493016277663435e-05, + "loss": 0.8705, + "step": 6397 + }, + { + "epoch": 0.2017978236871156, + "grad_norm": 2.643179178237915, + "learning_rate": 1.849247694561647e-05, + "loss": 0.5773, + "step": 6398 + }, + { + "epoch": 0.20182936445355623, + "grad_norm": 2.7894034385681152, + "learning_rate": 1.849193752494391e-05, + "loss": 0.5673, + "step": 6399 + }, + { + "epoch": 0.20186090521999683, + "grad_norm": 1.7454512119293213, + "learning_rate": 1.8491398015651395e-05, + "loss": 0.5404, + "step": 6400 + }, + { + "epoch": 0.20189244598643746, + "grad_norm": 2.2088067531585693, + "learning_rate": 1.8490858417744538e-05, + "loss": 0.5654, + "step": 6401 + }, + { + "epoch": 0.2019239867528781, + "grad_norm": 2.013326644897461, + "learning_rate": 1.8490318731228985e-05, + "loss": 0.5798, + "step": 6402 + }, + { + "epoch": 0.20195552751931872, + "grad_norm": 3.658478260040283, + "learning_rate": 1.8489778956110367e-05, + "loss": 0.5427, + "step": 6403 + }, + { + "epoch": 0.20198706828575935, + "grad_norm": 3.2432353496551514, + "learning_rate": 1.8489239092394312e-05, + "loss": 0.6059, + "step": 6404 + }, + { + "epoch": 0.20201860905219998, + "grad_norm": 1.576206922531128, + "learning_rate": 1.8488699140086454e-05, + "loss": 0.8976, + "step": 6405 + }, + { + "epoch": 0.20205014981864058, + "grad_norm": 1.9578009843826294, + "learning_rate": 1.848815909919243e-05, + "loss": 0.561, + "step": 6406 + }, + { + "epoch": 0.20208169058508121, + "grad_norm": 1.9822306632995605, + "learning_rate": 1.8487618969717872e-05, + "loss": 0.5709, + "step": 6407 + }, + { + "epoch": 0.20211323135152184, + "grad_norm": 2.1414308547973633, + "learning_rate": 1.8487078751668423e-05, + "loss": 0.6019, + "step": 6408 + }, + { + "epoch": 0.20214477211796247, + "grad_norm": 1.8525891304016113, + "learning_rate": 1.8486538445049715e-05, + "loss": 0.5874, + "step": 6409 + }, + { + "epoch": 0.2021763128844031, + "grad_norm": 2.434260129928589, + "learning_rate": 1.8485998049867388e-05, + "loss": 0.5323, + "step": 6410 + }, + { + "epoch": 0.2022078536508437, + "grad_norm": 2.857943058013916, + "learning_rate": 1.8485457566127084e-05, + "loss": 0.5349, + "step": 6411 + }, + { + "epoch": 0.20223939441728433, + "grad_norm": 2.4517366886138916, + "learning_rate": 1.8484916993834443e-05, + "loss": 0.5577, + "step": 6412 + }, + { + "epoch": 0.20227093518372496, + "grad_norm": 2.687114715576172, + "learning_rate": 1.8484376332995102e-05, + "loss": 0.5656, + "step": 6413 + }, + { + "epoch": 0.2023024759501656, + "grad_norm": 1.847161889076233, + "learning_rate": 1.8483835583614706e-05, + "loss": 0.8703, + "step": 6414 + }, + { + "epoch": 0.20233401671660622, + "grad_norm": 2.664820909500122, + "learning_rate": 1.8483294745698903e-05, + "loss": 0.5862, + "step": 6415 + }, + { + "epoch": 0.20236555748304683, + "grad_norm": 2.8472228050231934, + "learning_rate": 1.8482753819253326e-05, + "loss": 0.9106, + "step": 6416 + }, + { + "epoch": 0.20239709824948746, + "grad_norm": 2.674410343170166, + "learning_rate": 1.848221280428363e-05, + "loss": 0.5775, + "step": 6417 + }, + { + "epoch": 0.20242863901592809, + "grad_norm": 2.266702651977539, + "learning_rate": 1.8481671700795454e-05, + "loss": 0.5664, + "step": 6418 + }, + { + "epoch": 0.20246017978236872, + "grad_norm": 2.633641242980957, + "learning_rate": 1.848113050879445e-05, + "loss": 0.5521, + "step": 6419 + }, + { + "epoch": 0.20249172054880935, + "grad_norm": 4.10727071762085, + "learning_rate": 1.8480589228286266e-05, + "loss": 0.4036, + "step": 6420 + }, + { + "epoch": 0.20252326131524995, + "grad_norm": 2.802588939666748, + "learning_rate": 1.8480047859276542e-05, + "loss": 0.6028, + "step": 6421 + }, + { + "epoch": 0.20255480208169058, + "grad_norm": 2.3260462284088135, + "learning_rate": 1.847950640177094e-05, + "loss": 0.558, + "step": 6422 + }, + { + "epoch": 0.2025863428481312, + "grad_norm": 2.2095861434936523, + "learning_rate": 1.8478964855775102e-05, + "loss": 0.3922, + "step": 6423 + }, + { + "epoch": 0.20261788361457184, + "grad_norm": 1.8903852701187134, + "learning_rate": 1.847842322129468e-05, + "loss": 0.5558, + "step": 6424 + }, + { + "epoch": 0.20264942438101247, + "grad_norm": 2.1737518310546875, + "learning_rate": 1.8477881498335334e-05, + "loss": 0.873, + "step": 6425 + }, + { + "epoch": 0.2026809651474531, + "grad_norm": 2.5463101863861084, + "learning_rate": 1.8477339686902707e-05, + "loss": 0.5871, + "step": 6426 + }, + { + "epoch": 0.2027125059138937, + "grad_norm": 1.9685707092285156, + "learning_rate": 1.847679778700246e-05, + "loss": 0.5675, + "step": 6427 + }, + { + "epoch": 0.20274404668033433, + "grad_norm": 2.7070600986480713, + "learning_rate": 1.8476255798640242e-05, + "loss": 0.5828, + "step": 6428 + }, + { + "epoch": 0.20277558744677496, + "grad_norm": 4.488592147827148, + "learning_rate": 1.8475713721821717e-05, + "loss": 0.884, + "step": 6429 + }, + { + "epoch": 0.2028071282132156, + "grad_norm": 2.204909086227417, + "learning_rate": 1.8475171556552542e-05, + "loss": 0.5648, + "step": 6430 + }, + { + "epoch": 0.20283866897965622, + "grad_norm": 3.0965235233306885, + "learning_rate": 1.8474629302838362e-05, + "loss": 0.6103, + "step": 6431 + }, + { + "epoch": 0.20287020974609682, + "grad_norm": 5.768806457519531, + "learning_rate": 1.8474086960684848e-05, + "loss": 0.6065, + "step": 6432 + }, + { + "epoch": 0.20290175051253745, + "grad_norm": 1.8462382555007935, + "learning_rate": 1.847354453009766e-05, + "loss": 0.5739, + "step": 6433 + }, + { + "epoch": 0.20293329127897808, + "grad_norm": 2.3677008152008057, + "learning_rate": 1.8473002011082453e-05, + "loss": 0.5546, + "step": 6434 + }, + { + "epoch": 0.2029648320454187, + "grad_norm": 1.9065955877304077, + "learning_rate": 1.8472459403644892e-05, + "loss": 0.5774, + "step": 6435 + }, + { + "epoch": 0.20299637281185934, + "grad_norm": 3.0339715480804443, + "learning_rate": 1.8471916707790636e-05, + "loss": 0.5673, + "step": 6436 + }, + { + "epoch": 0.20302791357829994, + "grad_norm": 1.5028760433197021, + "learning_rate": 1.8471373923525355e-05, + "loss": 0.5481, + "step": 6437 + }, + { + "epoch": 0.20305945434474057, + "grad_norm": 2.510496139526367, + "learning_rate": 1.8470831050854704e-05, + "loss": 0.897, + "step": 6438 + }, + { + "epoch": 0.2030909951111812, + "grad_norm": 2.5406432151794434, + "learning_rate": 1.847028808978436e-05, + "loss": 0.5589, + "step": 6439 + }, + { + "epoch": 0.20312253587762183, + "grad_norm": 2.7102773189544678, + "learning_rate": 1.8469745040319975e-05, + "loss": 0.5679, + "step": 6440 + }, + { + "epoch": 0.20315407664406246, + "grad_norm": 2.594153881072998, + "learning_rate": 1.8469201902467227e-05, + "loss": 0.5856, + "step": 6441 + }, + { + "epoch": 0.2031856174105031, + "grad_norm": 3.165316104888916, + "learning_rate": 1.8468658676231786e-05, + "loss": 0.5628, + "step": 6442 + }, + { + "epoch": 0.2032171581769437, + "grad_norm": 2.288804531097412, + "learning_rate": 1.846811536161931e-05, + "loss": 0.5787, + "step": 6443 + }, + { + "epoch": 0.20324869894338432, + "grad_norm": 1.6109294891357422, + "learning_rate": 1.8467571958635474e-05, + "loss": 0.5398, + "step": 6444 + }, + { + "epoch": 0.20328023970982495, + "grad_norm": 1.771280288696289, + "learning_rate": 1.8467028467285954e-05, + "loss": 0.9004, + "step": 6445 + }, + { + "epoch": 0.20331178047626558, + "grad_norm": 2.794593334197998, + "learning_rate": 1.8466484887576416e-05, + "loss": 0.538, + "step": 6446 + }, + { + "epoch": 0.2033433212427062, + "grad_norm": 2.63944935798645, + "learning_rate": 1.8465941219512533e-05, + "loss": 0.5683, + "step": 6447 + }, + { + "epoch": 0.2033748620091468, + "grad_norm": 1.7015012502670288, + "learning_rate": 1.8465397463099983e-05, + "loss": 0.8798, + "step": 6448 + }, + { + "epoch": 0.20340640277558744, + "grad_norm": 2.1291487216949463, + "learning_rate": 1.8464853618344434e-05, + "loss": 0.5333, + "step": 6449 + }, + { + "epoch": 0.20343794354202807, + "grad_norm": 2.1081953048706055, + "learning_rate": 1.8464309685251565e-05, + "loss": 0.5592, + "step": 6450 + }, + { + "epoch": 0.2034694843084687, + "grad_norm": 2.627621650695801, + "learning_rate": 1.8463765663827055e-05, + "loss": 0.548, + "step": 6451 + }, + { + "epoch": 0.20350102507490933, + "grad_norm": 2.192521572113037, + "learning_rate": 1.8463221554076574e-05, + "loss": 0.5631, + "step": 6452 + }, + { + "epoch": 0.20353256584134993, + "grad_norm": 2.9657933712005615, + "learning_rate": 1.8462677356005805e-05, + "loss": 0.5812, + "step": 6453 + }, + { + "epoch": 0.20356410660779056, + "grad_norm": 2.458981513977051, + "learning_rate": 1.846213306962043e-05, + "loss": 0.5578, + "step": 6454 + }, + { + "epoch": 0.2035956473742312, + "grad_norm": 3.1785666942596436, + "learning_rate": 1.8461588694926128e-05, + "loss": 0.5469, + "step": 6455 + }, + { + "epoch": 0.20362718814067182, + "grad_norm": 2.0842995643615723, + "learning_rate": 1.8461044231928573e-05, + "loss": 0.6067, + "step": 6456 + }, + { + "epoch": 0.20365872890711245, + "grad_norm": 2.019272565841675, + "learning_rate": 1.846049968063345e-05, + "loss": 0.5717, + "step": 6457 + }, + { + "epoch": 0.20369026967355305, + "grad_norm": 2.8592119216918945, + "learning_rate": 1.8459955041046447e-05, + "loss": 0.5707, + "step": 6458 + }, + { + "epoch": 0.20372181043999368, + "grad_norm": 2.351926803588867, + "learning_rate": 1.8459410313173243e-05, + "loss": 0.5536, + "step": 6459 + }, + { + "epoch": 0.2037533512064343, + "grad_norm": 2.5835278034210205, + "learning_rate": 1.845886549701952e-05, + "loss": 0.6029, + "step": 6460 + }, + { + "epoch": 0.20378489197287494, + "grad_norm": 2.2618610858917236, + "learning_rate": 1.8458320592590976e-05, + "loss": 0.8604, + "step": 6461 + }, + { + "epoch": 0.20381643273931557, + "grad_norm": 3.282008647918701, + "learning_rate": 1.8457775599893283e-05, + "loss": 0.5629, + "step": 6462 + }, + { + "epoch": 0.2038479735057562, + "grad_norm": 2.3161652088165283, + "learning_rate": 1.8457230518932135e-05, + "loss": 0.5279, + "step": 6463 + }, + { + "epoch": 0.2038795142721968, + "grad_norm": 2.081555128097534, + "learning_rate": 1.845668534971322e-05, + "loss": 0.5632, + "step": 6464 + }, + { + "epoch": 0.20391105503863743, + "grad_norm": 1.6531282663345337, + "learning_rate": 1.8456140092242227e-05, + "loss": 0.5615, + "step": 6465 + }, + { + "epoch": 0.20394259580507806, + "grad_norm": 1.994537353515625, + "learning_rate": 1.8455594746524848e-05, + "loss": 0.4366, + "step": 6466 + }, + { + "epoch": 0.2039741365715187, + "grad_norm": 2.2309370040893555, + "learning_rate": 1.8455049312566768e-05, + "loss": 0.5455, + "step": 6467 + }, + { + "epoch": 0.20400567733795932, + "grad_norm": 3.1458890438079834, + "learning_rate": 1.845450379037369e-05, + "loss": 0.5548, + "step": 6468 + }, + { + "epoch": 0.20403721810439993, + "grad_norm": 3.2519261837005615, + "learning_rate": 1.8453958179951294e-05, + "loss": 0.549, + "step": 6469 + }, + { + "epoch": 0.20406875887084056, + "grad_norm": 2.5798332691192627, + "learning_rate": 1.845341248130528e-05, + "loss": 0.6009, + "step": 6470 + }, + { + "epoch": 0.20410029963728119, + "grad_norm": 2.1118741035461426, + "learning_rate": 1.845286669444135e-05, + "loss": 0.5806, + "step": 6471 + }, + { + "epoch": 0.20413184040372181, + "grad_norm": 1.899053692817688, + "learning_rate": 1.8452320819365183e-05, + "loss": 0.8909, + "step": 6472 + }, + { + "epoch": 0.20416338117016244, + "grad_norm": 3.1176259517669678, + "learning_rate": 1.845177485608249e-05, + "loss": 0.5474, + "step": 6473 + }, + { + "epoch": 0.20419492193660305, + "grad_norm": 4.7380218505859375, + "learning_rate": 1.8451228804598962e-05, + "loss": 0.5661, + "step": 6474 + }, + { + "epoch": 0.20422646270304368, + "grad_norm": 3.202702045440674, + "learning_rate": 1.8450682664920305e-05, + "loss": 0.5707, + "step": 6475 + }, + { + "epoch": 0.2042580034694843, + "grad_norm": 4.764935493469238, + "learning_rate": 1.8450136437052206e-05, + "loss": 0.5723, + "step": 6476 + }, + { + "epoch": 0.20428954423592494, + "grad_norm": 2.0208311080932617, + "learning_rate": 1.8449590121000374e-05, + "loss": 0.5937, + "step": 6477 + }, + { + "epoch": 0.20432108500236557, + "grad_norm": 4.733363151550293, + "learning_rate": 1.8449043716770508e-05, + "loss": 0.3646, + "step": 6478 + }, + { + "epoch": 0.2043526257688062, + "grad_norm": 7.073383331298828, + "learning_rate": 1.844849722436831e-05, + "loss": 0.5693, + "step": 6479 + }, + { + "epoch": 0.2043841665352468, + "grad_norm": 5.475374221801758, + "learning_rate": 1.8447950643799484e-05, + "loss": 0.6223, + "step": 6480 + }, + { + "epoch": 0.20441570730168743, + "grad_norm": 3.5298893451690674, + "learning_rate": 1.844740397506973e-05, + "loss": 0.5559, + "step": 6481 + }, + { + "epoch": 0.20444724806812806, + "grad_norm": 2.012925863265991, + "learning_rate": 1.844685721818476e-05, + "loss": 0.5595, + "step": 6482 + }, + { + "epoch": 0.2044787888345687, + "grad_norm": 4.209769248962402, + "learning_rate": 1.844631037315028e-05, + "loss": 0.5565, + "step": 6483 + }, + { + "epoch": 0.20451032960100932, + "grad_norm": 3.068040609359741, + "learning_rate": 1.8445763439971987e-05, + "loss": 0.626, + "step": 6484 + }, + { + "epoch": 0.20454187036744992, + "grad_norm": 2.5093727111816406, + "learning_rate": 1.8445216418655597e-05, + "loss": 0.5849, + "step": 6485 + }, + { + "epoch": 0.20457341113389055, + "grad_norm": 2.645991802215576, + "learning_rate": 1.8444669309206814e-05, + "loss": 0.5669, + "step": 6486 + }, + { + "epoch": 0.20460495190033118, + "grad_norm": 2.849940061569214, + "learning_rate": 1.8444122111631348e-05, + "loss": 0.5644, + "step": 6487 + }, + { + "epoch": 0.2046364926667718, + "grad_norm": 3.4549317359924316, + "learning_rate": 1.8443574825934915e-05, + "loss": 0.4086, + "step": 6488 + }, + { + "epoch": 0.20466803343321244, + "grad_norm": 2.174098491668701, + "learning_rate": 1.8443027452123222e-05, + "loss": 0.5799, + "step": 6489 + }, + { + "epoch": 0.20469957419965304, + "grad_norm": 2.864108085632324, + "learning_rate": 1.844247999020198e-05, + "loss": 0.5913, + "step": 6490 + }, + { + "epoch": 0.20473111496609367, + "grad_norm": 1.7718502283096313, + "learning_rate": 1.8441932440176905e-05, + "loss": 0.544, + "step": 6491 + }, + { + "epoch": 0.2047626557325343, + "grad_norm": 2.107004404067993, + "learning_rate": 1.844138480205371e-05, + "loss": 0.6071, + "step": 6492 + }, + { + "epoch": 0.20479419649897493, + "grad_norm": 9.949335098266602, + "learning_rate": 1.844083707583811e-05, + "loss": 0.3958, + "step": 6493 + }, + { + "epoch": 0.20482573726541556, + "grad_norm": 4.907886028289795, + "learning_rate": 1.8440289261535822e-05, + "loss": 0.5542, + "step": 6494 + }, + { + "epoch": 0.20485727803185616, + "grad_norm": 2.2809276580810547, + "learning_rate": 1.843974135915256e-05, + "loss": 0.8877, + "step": 6495 + }, + { + "epoch": 0.2048888187982968, + "grad_norm": 2.580515146255493, + "learning_rate": 1.8439193368694047e-05, + "loss": 0.5854, + "step": 6496 + }, + { + "epoch": 0.20492035956473742, + "grad_norm": 2.211732864379883, + "learning_rate": 1.8438645290165993e-05, + "loss": 0.5899, + "step": 6497 + }, + { + "epoch": 0.20495190033117805, + "grad_norm": 2.174370288848877, + "learning_rate": 1.8438097123574125e-05, + "loss": 0.4319, + "step": 6498 + }, + { + "epoch": 0.20498344109761868, + "grad_norm": 1.8062188625335693, + "learning_rate": 1.8437548868924164e-05, + "loss": 0.5702, + "step": 6499 + }, + { + "epoch": 0.2050149818640593, + "grad_norm": 1.9357885122299194, + "learning_rate": 1.8437000526221827e-05, + "loss": 0.5996, + "step": 6500 + }, + { + "epoch": 0.2050465226304999, + "grad_norm": 3.5905323028564453, + "learning_rate": 1.8436452095472835e-05, + "loss": 0.8531, + "step": 6501 + }, + { + "epoch": 0.20507806339694054, + "grad_norm": 2.4059226512908936, + "learning_rate": 1.843590357668292e-05, + "loss": 0.5786, + "step": 6502 + }, + { + "epoch": 0.20510960416338117, + "grad_norm": 1.999337911605835, + "learning_rate": 1.84353549698578e-05, + "loss": 0.5644, + "step": 6503 + }, + { + "epoch": 0.2051411449298218, + "grad_norm": 2.51875638961792, + "learning_rate": 1.84348062750032e-05, + "loss": 0.5839, + "step": 6504 + }, + { + "epoch": 0.20517268569626243, + "grad_norm": 1.990050196647644, + "learning_rate": 1.8434257492124844e-05, + "loss": 0.5532, + "step": 6505 + }, + { + "epoch": 0.20520422646270303, + "grad_norm": 5.704040050506592, + "learning_rate": 1.843370862122847e-05, + "loss": 0.8403, + "step": 6506 + }, + { + "epoch": 0.20523576722914366, + "grad_norm": 2.5653634071350098, + "learning_rate": 1.843315966231979e-05, + "loss": 0.5676, + "step": 6507 + }, + { + "epoch": 0.2052673079955843, + "grad_norm": 1.8841650485992432, + "learning_rate": 1.8432610615404545e-05, + "loss": 0.5471, + "step": 6508 + }, + { + "epoch": 0.20529884876202492, + "grad_norm": 1.971738576889038, + "learning_rate": 1.8432061480488458e-05, + "loss": 0.4562, + "step": 6509 + }, + { + "epoch": 0.20533038952846555, + "grad_norm": 2.5074353218078613, + "learning_rate": 1.8431512257577265e-05, + "loss": 0.5739, + "step": 6510 + }, + { + "epoch": 0.20536193029490615, + "grad_norm": 3.5753180980682373, + "learning_rate": 1.843096294667669e-05, + "loss": 0.5915, + "step": 6511 + }, + { + "epoch": 0.20539347106134678, + "grad_norm": 2.175743341445923, + "learning_rate": 1.8430413547792475e-05, + "loss": 0.5917, + "step": 6512 + }, + { + "epoch": 0.2054250118277874, + "grad_norm": 2.3786964416503906, + "learning_rate": 1.8429864060930345e-05, + "loss": 0.5658, + "step": 6513 + }, + { + "epoch": 0.20545655259422804, + "grad_norm": 2.6803016662597656, + "learning_rate": 1.8429314486096042e-05, + "loss": 0.5815, + "step": 6514 + }, + { + "epoch": 0.20548809336066867, + "grad_norm": 2.95127534866333, + "learning_rate": 1.8428764823295293e-05, + "loss": 0.5712, + "step": 6515 + }, + { + "epoch": 0.2055196341271093, + "grad_norm": 4.139904499053955, + "learning_rate": 1.8428215072533842e-05, + "loss": 0.5556, + "step": 6516 + }, + { + "epoch": 0.2055511748935499, + "grad_norm": 2.0713233947753906, + "learning_rate": 1.8427665233817422e-05, + "loss": 0.9185, + "step": 6517 + }, + { + "epoch": 0.20558271565999053, + "grad_norm": 2.6077282428741455, + "learning_rate": 1.8427115307151768e-05, + "loss": 0.5333, + "step": 6518 + }, + { + "epoch": 0.20561425642643116, + "grad_norm": 2.3845486640930176, + "learning_rate": 1.8426565292542623e-05, + "loss": 0.4146, + "step": 6519 + }, + { + "epoch": 0.2056457971928718, + "grad_norm": 2.81085205078125, + "learning_rate": 1.8426015189995728e-05, + "loss": 0.5528, + "step": 6520 + }, + { + "epoch": 0.20567733795931242, + "grad_norm": 2.133586883544922, + "learning_rate": 1.842546499951682e-05, + "loss": 0.5335, + "step": 6521 + }, + { + "epoch": 0.20570887872575302, + "grad_norm": 3.1787381172180176, + "learning_rate": 1.8424914721111646e-05, + "loss": 0.5565, + "step": 6522 + }, + { + "epoch": 0.20574041949219365, + "grad_norm": 2.9045581817626953, + "learning_rate": 1.842436435478594e-05, + "loss": 0.5272, + "step": 6523 + }, + { + "epoch": 0.20577196025863428, + "grad_norm": 2.792973756790161, + "learning_rate": 1.8423813900545457e-05, + "loss": 0.566, + "step": 6524 + }, + { + "epoch": 0.20580350102507491, + "grad_norm": 2.1599526405334473, + "learning_rate": 1.842326335839593e-05, + "loss": 0.5808, + "step": 6525 + }, + { + "epoch": 0.20583504179151554, + "grad_norm": 2.4006855487823486, + "learning_rate": 1.842271272834311e-05, + "loss": 0.5677, + "step": 6526 + }, + { + "epoch": 0.20586658255795615, + "grad_norm": 2.775487184524536, + "learning_rate": 1.8422162010392743e-05, + "loss": 0.5647, + "step": 6527 + }, + { + "epoch": 0.20589812332439678, + "grad_norm": 3.7357099056243896, + "learning_rate": 1.842161120455058e-05, + "loss": 0.5545, + "step": 6528 + }, + { + "epoch": 0.2059296640908374, + "grad_norm": 1.6298047304153442, + "learning_rate": 1.8421060310822358e-05, + "loss": 0.3881, + "step": 6529 + }, + { + "epoch": 0.20596120485727804, + "grad_norm": 2.7072057723999023, + "learning_rate": 1.8420509329213835e-05, + "loss": 0.5373, + "step": 6530 + }, + { + "epoch": 0.20599274562371866, + "grad_norm": 3.639707088470459, + "learning_rate": 1.8419958259730763e-05, + "loss": 0.5753, + "step": 6531 + }, + { + "epoch": 0.2060242863901593, + "grad_norm": 2.697120189666748, + "learning_rate": 1.8419407102378885e-05, + "loss": 0.5833, + "step": 6532 + }, + { + "epoch": 0.2060558271565999, + "grad_norm": 2.6432104110717773, + "learning_rate": 1.8418855857163954e-05, + "loss": 0.5504, + "step": 6533 + }, + { + "epoch": 0.20608736792304053, + "grad_norm": 5.06777811050415, + "learning_rate": 1.841830452409173e-05, + "loss": 0.4195, + "step": 6534 + }, + { + "epoch": 0.20611890868948116, + "grad_norm": 3.3967435359954834, + "learning_rate": 1.8417753103167956e-05, + "loss": 0.5446, + "step": 6535 + }, + { + "epoch": 0.20615044945592179, + "grad_norm": 3.0411577224731445, + "learning_rate": 1.8417201594398398e-05, + "loss": 0.5179, + "step": 6536 + }, + { + "epoch": 0.20618199022236242, + "grad_norm": 2.797720432281494, + "learning_rate": 1.8416649997788803e-05, + "loss": 0.5713, + "step": 6537 + }, + { + "epoch": 0.20621353098880302, + "grad_norm": 3.360440731048584, + "learning_rate": 1.841609831334493e-05, + "loss": 0.5861, + "step": 6538 + }, + { + "epoch": 0.20624507175524365, + "grad_norm": 3.186134099960327, + "learning_rate": 1.8415546541072537e-05, + "loss": 0.3568, + "step": 6539 + }, + { + "epoch": 0.20627661252168428, + "grad_norm": 5.408341884613037, + "learning_rate": 1.841499468097738e-05, + "loss": 0.9189, + "step": 6540 + }, + { + "epoch": 0.2063081532881249, + "grad_norm": 2.656442642211914, + "learning_rate": 1.841444273306522e-05, + "loss": 0.9091, + "step": 6541 + }, + { + "epoch": 0.20633969405456554, + "grad_norm": 3.429513454437256, + "learning_rate": 1.8413890697341815e-05, + "loss": 0.546, + "step": 6542 + }, + { + "epoch": 0.20637123482100614, + "grad_norm": 3.1337549686431885, + "learning_rate": 1.841333857381293e-05, + "loss": 0.5479, + "step": 6543 + }, + { + "epoch": 0.20640277558744677, + "grad_norm": 2.3562276363372803, + "learning_rate": 1.8412786362484324e-05, + "loss": 0.8604, + "step": 6544 + }, + { + "epoch": 0.2064343163538874, + "grad_norm": 2.161402702331543, + "learning_rate": 1.8412234063361757e-05, + "loss": 0.5433, + "step": 6545 + }, + { + "epoch": 0.20646585712032803, + "grad_norm": 3.3058524131774902, + "learning_rate": 1.8411681676450998e-05, + "loss": 0.5717, + "step": 6546 + }, + { + "epoch": 0.20649739788676866, + "grad_norm": 3.4828779697418213, + "learning_rate": 1.841112920175781e-05, + "loss": 0.6034, + "step": 6547 + }, + { + "epoch": 0.20652893865320926, + "grad_norm": 2.5487112998962402, + "learning_rate": 1.8410576639287953e-05, + "loss": 0.5754, + "step": 6548 + }, + { + "epoch": 0.2065604794196499, + "grad_norm": 1.801105260848999, + "learning_rate": 1.8410023989047204e-05, + "loss": 0.5589, + "step": 6549 + }, + { + "epoch": 0.20659202018609052, + "grad_norm": 2.494838237762451, + "learning_rate": 1.8409471251041322e-05, + "loss": 0.6055, + "step": 6550 + }, + { + "epoch": 0.20662356095253115, + "grad_norm": 3.9534428119659424, + "learning_rate": 1.8408918425276078e-05, + "loss": 0.5894, + "step": 6551 + }, + { + "epoch": 0.20665510171897178, + "grad_norm": 2.0425517559051514, + "learning_rate": 1.840836551175724e-05, + "loss": 0.5679, + "step": 6552 + }, + { + "epoch": 0.2066866424854124, + "grad_norm": 2.629063129425049, + "learning_rate": 1.840781251049058e-05, + "loss": 0.6165, + "step": 6553 + }, + { + "epoch": 0.206718183251853, + "grad_norm": 2.98811936378479, + "learning_rate": 1.8407259421481867e-05, + "loss": 0.582, + "step": 6554 + }, + { + "epoch": 0.20674972401829364, + "grad_norm": 2.3535072803497314, + "learning_rate": 1.8406706244736873e-05, + "loss": 0.6148, + "step": 6555 + }, + { + "epoch": 0.20678126478473427, + "grad_norm": 2.7346854209899902, + "learning_rate": 1.8406152980261373e-05, + "loss": 0.9273, + "step": 6556 + }, + { + "epoch": 0.2068128055511749, + "grad_norm": 1.8193968534469604, + "learning_rate": 1.840559962806114e-05, + "loss": 0.5948, + "step": 6557 + }, + { + "epoch": 0.20684434631761553, + "grad_norm": 2.016230344772339, + "learning_rate": 1.840504618814195e-05, + "loss": 0.574, + "step": 6558 + }, + { + "epoch": 0.20687588708405613, + "grad_norm": 2.5940732955932617, + "learning_rate": 1.8404492660509575e-05, + "loss": 0.5895, + "step": 6559 + }, + { + "epoch": 0.20690742785049676, + "grad_norm": 2.0143890380859375, + "learning_rate": 1.840393904516979e-05, + "loss": 0.5741, + "step": 6560 + }, + { + "epoch": 0.2069389686169374, + "grad_norm": 2.0917580127716064, + "learning_rate": 1.840338534212838e-05, + "loss": 0.6034, + "step": 6561 + }, + { + "epoch": 0.20697050938337802, + "grad_norm": 1.6128507852554321, + "learning_rate": 1.8402831551391117e-05, + "loss": 0.5475, + "step": 6562 + }, + { + "epoch": 0.20700205014981865, + "grad_norm": 2.5103759765625, + "learning_rate": 1.8402277672963782e-05, + "loss": 0.5472, + "step": 6563 + }, + { + "epoch": 0.20703359091625925, + "grad_norm": 1.789245367050171, + "learning_rate": 1.8401723706852156e-05, + "loss": 0.5815, + "step": 6564 + }, + { + "epoch": 0.20706513168269988, + "grad_norm": 1.6675082445144653, + "learning_rate": 1.8401169653062017e-05, + "loss": 0.5935, + "step": 6565 + }, + { + "epoch": 0.2070966724491405, + "grad_norm": 2.0230579376220703, + "learning_rate": 1.840061551159915e-05, + "loss": 0.5719, + "step": 6566 + }, + { + "epoch": 0.20712821321558114, + "grad_norm": 2.6075189113616943, + "learning_rate": 1.840006128246934e-05, + "loss": 0.5457, + "step": 6567 + }, + { + "epoch": 0.20715975398202177, + "grad_norm": 1.778749704360962, + "learning_rate": 1.8399506965678365e-05, + "loss": 0.553, + "step": 6568 + }, + { + "epoch": 0.2071912947484624, + "grad_norm": 2.0407052040100098, + "learning_rate": 1.8398952561232016e-05, + "loss": 0.5481, + "step": 6569 + }, + { + "epoch": 0.207222835514903, + "grad_norm": 4.711043357849121, + "learning_rate": 1.839839806913607e-05, + "loss": 0.5554, + "step": 6570 + }, + { + "epoch": 0.20725437628134363, + "grad_norm": 2.3169126510620117, + "learning_rate": 1.8397843489396324e-05, + "loss": 0.5401, + "step": 6571 + }, + { + "epoch": 0.20728591704778426, + "grad_norm": 3.0433526039123535, + "learning_rate": 1.839728882201856e-05, + "loss": 0.5373, + "step": 6572 + }, + { + "epoch": 0.2073174578142249, + "grad_norm": 2.0604588985443115, + "learning_rate": 1.8396734067008566e-05, + "loss": 0.549, + "step": 6573 + }, + { + "epoch": 0.20734899858066552, + "grad_norm": 2.4726436138153076, + "learning_rate": 1.839617922437213e-05, + "loss": 0.4627, + "step": 6574 + }, + { + "epoch": 0.20738053934710612, + "grad_norm": 1.9642115831375122, + "learning_rate": 1.8395624294115046e-05, + "loss": 0.6052, + "step": 6575 + }, + { + "epoch": 0.20741208011354675, + "grad_norm": 2.2640388011932373, + "learning_rate": 1.8395069276243104e-05, + "loss": 0.564, + "step": 6576 + }, + { + "epoch": 0.20744362087998738, + "grad_norm": 2.104398488998413, + "learning_rate": 1.8394514170762096e-05, + "loss": 0.5766, + "step": 6577 + }, + { + "epoch": 0.207475161646428, + "grad_norm": 2.174219846725464, + "learning_rate": 1.8393958977677814e-05, + "loss": 0.5884, + "step": 6578 + }, + { + "epoch": 0.20750670241286864, + "grad_norm": 2.2183289527893066, + "learning_rate": 1.839340369699605e-05, + "loss": 0.5477, + "step": 6579 + }, + { + "epoch": 0.20753824317930925, + "grad_norm": 2.727656364440918, + "learning_rate": 1.8392848328722603e-05, + "loss": 0.8809, + "step": 6580 + }, + { + "epoch": 0.20756978394574988, + "grad_norm": 4.47168493270874, + "learning_rate": 1.839229287286327e-05, + "loss": 0.5891, + "step": 6581 + }, + { + "epoch": 0.2076013247121905, + "grad_norm": 2.257497549057007, + "learning_rate": 1.839173732942384e-05, + "loss": 0.5703, + "step": 6582 + }, + { + "epoch": 0.20763286547863113, + "grad_norm": 2.1303341388702393, + "learning_rate": 1.8391181698410116e-05, + "loss": 0.5297, + "step": 6583 + }, + { + "epoch": 0.20766440624507176, + "grad_norm": 3.645106554031372, + "learning_rate": 1.8390625979827894e-05, + "loss": 0.5727, + "step": 6584 + }, + { + "epoch": 0.20769594701151237, + "grad_norm": 2.6791529655456543, + "learning_rate": 1.8390070173682978e-05, + "loss": 0.5537, + "step": 6585 + }, + { + "epoch": 0.207727487777953, + "grad_norm": 3.1817517280578613, + "learning_rate": 1.8389514279981163e-05, + "loss": 0.4283, + "step": 6586 + }, + { + "epoch": 0.20775902854439363, + "grad_norm": 3.048966884613037, + "learning_rate": 1.838895829872825e-05, + "loss": 0.5542, + "step": 6587 + }, + { + "epoch": 0.20779056931083426, + "grad_norm": 2.1497507095336914, + "learning_rate": 1.838840222993005e-05, + "loss": 0.5689, + "step": 6588 + }, + { + "epoch": 0.20782211007727489, + "grad_norm": 2.83357310295105, + "learning_rate": 1.8387846073592353e-05, + "loss": 0.5386, + "step": 6589 + }, + { + "epoch": 0.20785365084371552, + "grad_norm": 2.351982831954956, + "learning_rate": 1.8387289829720972e-05, + "loss": 0.5611, + "step": 6590 + }, + { + "epoch": 0.20788519161015612, + "grad_norm": 1.9061381816864014, + "learning_rate": 1.838673349832171e-05, + "loss": 0.5351, + "step": 6591 + }, + { + "epoch": 0.20791673237659675, + "grad_norm": 2.602712392807007, + "learning_rate": 1.8386177079400368e-05, + "loss": 0.5717, + "step": 6592 + }, + { + "epoch": 0.20794827314303738, + "grad_norm": 3.2194173336029053, + "learning_rate": 1.838562057296276e-05, + "loss": 0.579, + "step": 6593 + }, + { + "epoch": 0.207979813909478, + "grad_norm": 2.895324230194092, + "learning_rate": 1.8385063979014687e-05, + "loss": 0.5816, + "step": 6594 + }, + { + "epoch": 0.20801135467591864, + "grad_norm": 2.6414647102355957, + "learning_rate": 1.8384507297561965e-05, + "loss": 0.5452, + "step": 6595 + }, + { + "epoch": 0.20804289544235924, + "grad_norm": 1.9957243204116821, + "learning_rate": 1.8383950528610395e-05, + "loss": 0.5508, + "step": 6596 + }, + { + "epoch": 0.20807443620879987, + "grad_norm": 2.86738920211792, + "learning_rate": 1.8383393672165792e-05, + "loss": 0.5471, + "step": 6597 + }, + { + "epoch": 0.2081059769752405, + "grad_norm": 3.127293586730957, + "learning_rate": 1.8382836728233966e-05, + "loss": 0.8775, + "step": 6598 + }, + { + "epoch": 0.20813751774168113, + "grad_norm": 3.1549673080444336, + "learning_rate": 1.8382279696820732e-05, + "loss": 0.5415, + "step": 6599 + }, + { + "epoch": 0.20816905850812176, + "grad_norm": 2.2858047485351562, + "learning_rate": 1.83817225779319e-05, + "loss": 0.5635, + "step": 6600 + }, + { + "epoch": 0.20820059927456236, + "grad_norm": 2.3268816471099854, + "learning_rate": 1.8381165371573283e-05, + "loss": 0.5748, + "step": 6601 + }, + { + "epoch": 0.208232140041003, + "grad_norm": 2.368797540664673, + "learning_rate": 1.83806080777507e-05, + "loss": 0.5259, + "step": 6602 + }, + { + "epoch": 0.20826368080744362, + "grad_norm": 2.602254867553711, + "learning_rate": 1.838005069646996e-05, + "loss": 0.5692, + "step": 6603 + }, + { + "epoch": 0.20829522157388425, + "grad_norm": 2.373018741607666, + "learning_rate": 1.837949322773689e-05, + "loss": 0.9105, + "step": 6604 + }, + { + "epoch": 0.20832676234032488, + "grad_norm": 2.926422357559204, + "learning_rate": 1.8378935671557296e-05, + "loss": 0.5459, + "step": 6605 + }, + { + "epoch": 0.2083583031067655, + "grad_norm": 2.5811636447906494, + "learning_rate": 1.8378378027937007e-05, + "loss": 0.5855, + "step": 6606 + }, + { + "epoch": 0.2083898438732061, + "grad_norm": 8.047904968261719, + "learning_rate": 1.8377820296881833e-05, + "loss": 0.5808, + "step": 6607 + }, + { + "epoch": 0.20842138463964674, + "grad_norm": 4.178526401519775, + "learning_rate": 1.8377262478397602e-05, + "loss": 0.4601, + "step": 6608 + }, + { + "epoch": 0.20845292540608737, + "grad_norm": 3.428457498550415, + "learning_rate": 1.8376704572490132e-05, + "loss": 0.8951, + "step": 6609 + }, + { + "epoch": 0.208484466172528, + "grad_norm": 2.4802052974700928, + "learning_rate": 1.8376146579165243e-05, + "loss": 0.5887, + "step": 6610 + }, + { + "epoch": 0.20851600693896863, + "grad_norm": 2.672485113143921, + "learning_rate": 1.8375588498428763e-05, + "loss": 0.5875, + "step": 6611 + }, + { + "epoch": 0.20854754770540923, + "grad_norm": 2.9388747215270996, + "learning_rate": 1.8375030330286513e-05, + "loss": 0.5674, + "step": 6612 + }, + { + "epoch": 0.20857908847184986, + "grad_norm": 2.272367000579834, + "learning_rate": 1.837447207474432e-05, + "loss": 0.5578, + "step": 6613 + }, + { + "epoch": 0.2086106292382905, + "grad_norm": 2.2291367053985596, + "learning_rate": 1.837391373180801e-05, + "loss": 0.5657, + "step": 6614 + }, + { + "epoch": 0.20864217000473112, + "grad_norm": 3.0116491317749023, + "learning_rate": 1.8373355301483403e-05, + "loss": 0.5596, + "step": 6615 + }, + { + "epoch": 0.20867371077117175, + "grad_norm": 2.200167417526245, + "learning_rate": 1.8372796783776334e-05, + "loss": 0.558, + "step": 6616 + }, + { + "epoch": 0.20870525153761235, + "grad_norm": 1.8241167068481445, + "learning_rate": 1.837223817869263e-05, + "loss": 0.5429, + "step": 6617 + }, + { + "epoch": 0.20873679230405298, + "grad_norm": 2.149446964263916, + "learning_rate": 1.8371679486238116e-05, + "loss": 0.9029, + "step": 6618 + }, + { + "epoch": 0.2087683330704936, + "grad_norm": 2.259770154953003, + "learning_rate": 1.837112070641863e-05, + "loss": 0.5759, + "step": 6619 + }, + { + "epoch": 0.20879987383693424, + "grad_norm": 2.227287530899048, + "learning_rate": 1.8370561839239997e-05, + "loss": 0.5597, + "step": 6620 + }, + { + "epoch": 0.20883141460337487, + "grad_norm": 4.239623069763184, + "learning_rate": 1.8370002884708052e-05, + "loss": 0.5707, + "step": 6621 + }, + { + "epoch": 0.20886295536981547, + "grad_norm": 2.4334211349487305, + "learning_rate": 1.836944384282863e-05, + "loss": 0.5491, + "step": 6622 + }, + { + "epoch": 0.2088944961362561, + "grad_norm": 1.9986205101013184, + "learning_rate": 1.836888471360756e-05, + "loss": 0.5395, + "step": 6623 + }, + { + "epoch": 0.20892603690269673, + "grad_norm": 2.3491146564483643, + "learning_rate": 1.836832549705068e-05, + "loss": 0.5619, + "step": 6624 + }, + { + "epoch": 0.20895757766913736, + "grad_norm": 1.9359591007232666, + "learning_rate": 1.8367766193163828e-05, + "loss": 0.5633, + "step": 6625 + }, + { + "epoch": 0.208989118435578, + "grad_norm": 1.9560213088989258, + "learning_rate": 1.836720680195284e-05, + "loss": 0.5617, + "step": 6626 + }, + { + "epoch": 0.20902065920201862, + "grad_norm": 8.168313026428223, + "learning_rate": 1.836664732342355e-05, + "loss": 0.5613, + "step": 6627 + }, + { + "epoch": 0.20905219996845922, + "grad_norm": 2.9465134143829346, + "learning_rate": 1.8366087757581797e-05, + "loss": 0.5547, + "step": 6628 + }, + { + "epoch": 0.20908374073489985, + "grad_norm": 2.755906343460083, + "learning_rate": 1.8365528104433422e-05, + "loss": 0.5458, + "step": 6629 + }, + { + "epoch": 0.20911528150134048, + "grad_norm": 3.250515937805176, + "learning_rate": 1.8364968363984264e-05, + "loss": 0.5602, + "step": 6630 + }, + { + "epoch": 0.2091468222677811, + "grad_norm": 2.0720913410186768, + "learning_rate": 1.836440853624017e-05, + "loss": 0.5311, + "step": 6631 + }, + { + "epoch": 0.20917836303422174, + "grad_norm": 3.763087749481201, + "learning_rate": 1.8363848621206977e-05, + "loss": 0.538, + "step": 6632 + }, + { + "epoch": 0.20920990380066234, + "grad_norm": 2.3436779975891113, + "learning_rate": 1.8363288618890526e-05, + "loss": 0.5501, + "step": 6633 + }, + { + "epoch": 0.20924144456710297, + "grad_norm": 5.52656888961792, + "learning_rate": 1.8362728529296668e-05, + "loss": 0.5421, + "step": 6634 + }, + { + "epoch": 0.2092729853335436, + "grad_norm": 2.7542436122894287, + "learning_rate": 1.8362168352431244e-05, + "loss": 0.5579, + "step": 6635 + }, + { + "epoch": 0.20930452609998423, + "grad_norm": 4.854994297027588, + "learning_rate": 1.8361608088300104e-05, + "loss": 0.5723, + "step": 6636 + }, + { + "epoch": 0.20933606686642486, + "grad_norm": 1.8223899602890015, + "learning_rate": 1.8361047736909084e-05, + "loss": 0.875, + "step": 6637 + }, + { + "epoch": 0.20936760763286547, + "grad_norm": 2.288733959197998, + "learning_rate": 1.836048729826404e-05, + "loss": 0.5875, + "step": 6638 + }, + { + "epoch": 0.2093991483993061, + "grad_norm": 2.434857130050659, + "learning_rate": 1.8359926772370823e-05, + "loss": 0.5674, + "step": 6639 + }, + { + "epoch": 0.20943068916574673, + "grad_norm": 6.367259979248047, + "learning_rate": 1.835936615923527e-05, + "loss": 0.5513, + "step": 6640 + }, + { + "epoch": 0.20946222993218735, + "grad_norm": 2.9925076961517334, + "learning_rate": 1.835880545886325e-05, + "loss": 0.5457, + "step": 6641 + }, + { + "epoch": 0.20949377069862798, + "grad_norm": 2.3031868934631348, + "learning_rate": 1.8358244671260604e-05, + "loss": 0.904, + "step": 6642 + }, + { + "epoch": 0.20952531146506861, + "grad_norm": 2.6957242488861084, + "learning_rate": 1.835768379643318e-05, + "loss": 0.5801, + "step": 6643 + }, + { + "epoch": 0.20955685223150922, + "grad_norm": 1.8711313009262085, + "learning_rate": 1.8357122834386838e-05, + "loss": 0.8892, + "step": 6644 + }, + { + "epoch": 0.20958839299794985, + "grad_norm": 1.8174365758895874, + "learning_rate": 1.8356561785127432e-05, + "loss": 0.8808, + "step": 6645 + }, + { + "epoch": 0.20961993376439048, + "grad_norm": 2.526663303375244, + "learning_rate": 1.835600064866081e-05, + "loss": 0.5156, + "step": 6646 + }, + { + "epoch": 0.2096514745308311, + "grad_norm": 2.320354461669922, + "learning_rate": 1.835543942499284e-05, + "loss": 0.5601, + "step": 6647 + }, + { + "epoch": 0.20968301529727174, + "grad_norm": 19.078737258911133, + "learning_rate": 1.8354878114129368e-05, + "loss": 0.5468, + "step": 6648 + }, + { + "epoch": 0.20971455606371234, + "grad_norm": 2.643645763397217, + "learning_rate": 1.835431671607625e-05, + "loss": 0.6084, + "step": 6649 + }, + { + "epoch": 0.20974609683015297, + "grad_norm": 4.107604503631592, + "learning_rate": 1.835375523083936e-05, + "loss": 0.5844, + "step": 6650 + }, + { + "epoch": 0.2097776375965936, + "grad_norm": 5.041844844818115, + "learning_rate": 1.835319365842454e-05, + "loss": 0.5549, + "step": 6651 + }, + { + "epoch": 0.20980917836303423, + "grad_norm": 2.648952007293701, + "learning_rate": 1.8352631998837666e-05, + "loss": 0.8997, + "step": 6652 + }, + { + "epoch": 0.20984071912947486, + "grad_norm": 2.4964935779571533, + "learning_rate": 1.8352070252084582e-05, + "loss": 0.5588, + "step": 6653 + }, + { + "epoch": 0.20987225989591546, + "grad_norm": 4.01652193069458, + "learning_rate": 1.8351508418171168e-05, + "loss": 0.5799, + "step": 6654 + }, + { + "epoch": 0.2099038006623561, + "grad_norm": 2.8962862491607666, + "learning_rate": 1.8350946497103274e-05, + "loss": 0.5796, + "step": 6655 + }, + { + "epoch": 0.20993534142879672, + "grad_norm": 2.8691494464874268, + "learning_rate": 1.835038448888677e-05, + "loss": 0.6152, + "step": 6656 + }, + { + "epoch": 0.20996688219523735, + "grad_norm": 2.846327066421509, + "learning_rate": 1.834982239352752e-05, + "loss": 0.5473, + "step": 6657 + }, + { + "epoch": 0.20999842296167798, + "grad_norm": 3.1719162464141846, + "learning_rate": 1.834926021103139e-05, + "loss": 0.5573, + "step": 6658 + }, + { + "epoch": 0.2100299637281186, + "grad_norm": 10.084249496459961, + "learning_rate": 1.8348697941404247e-05, + "loss": 0.401, + "step": 6659 + }, + { + "epoch": 0.2100615044945592, + "grad_norm": 6.022458553314209, + "learning_rate": 1.834813558465196e-05, + "loss": 0.5653, + "step": 6660 + }, + { + "epoch": 0.21009304526099984, + "grad_norm": 6.423914432525635, + "learning_rate": 1.8347573140780392e-05, + "loss": 0.5639, + "step": 6661 + }, + { + "epoch": 0.21012458602744047, + "grad_norm": 3.7374002933502197, + "learning_rate": 1.834701060979542e-05, + "loss": 0.5472, + "step": 6662 + }, + { + "epoch": 0.2101561267938811, + "grad_norm": 4.120702743530273, + "learning_rate": 1.8346447991702908e-05, + "loss": 0.5718, + "step": 6663 + }, + { + "epoch": 0.21018766756032173, + "grad_norm": 5.194669723510742, + "learning_rate": 1.834588528650873e-05, + "loss": 0.3882, + "step": 6664 + }, + { + "epoch": 0.21021920832676233, + "grad_norm": 4.584064960479736, + "learning_rate": 1.8345322494218763e-05, + "loss": 0.5769, + "step": 6665 + }, + { + "epoch": 0.21025074909320296, + "grad_norm": 2.4933996200561523, + "learning_rate": 1.8344759614838874e-05, + "loss": 0.5524, + "step": 6666 + }, + { + "epoch": 0.2102822898596436, + "grad_norm": 2.8657288551330566, + "learning_rate": 1.8344196648374938e-05, + "loss": 0.5704, + "step": 6667 + }, + { + "epoch": 0.21031383062608422, + "grad_norm": 2.9230308532714844, + "learning_rate": 1.834363359483283e-05, + "loss": 0.6051, + "step": 6668 + }, + { + "epoch": 0.21034537139252485, + "grad_norm": 3.5035593509674072, + "learning_rate": 1.8343070454218428e-05, + "loss": 0.4511, + "step": 6669 + }, + { + "epoch": 0.21037691215896545, + "grad_norm": 2.5276663303375244, + "learning_rate": 1.834250722653761e-05, + "loss": 0.5339, + "step": 6670 + }, + { + "epoch": 0.21040845292540608, + "grad_norm": 3.2271556854248047, + "learning_rate": 1.834194391179625e-05, + "loss": 0.5299, + "step": 6671 + }, + { + "epoch": 0.2104399936918467, + "grad_norm": 3.256270408630371, + "learning_rate": 1.8341380510000224e-05, + "loss": 0.6025, + "step": 6672 + }, + { + "epoch": 0.21047153445828734, + "grad_norm": 3.3643417358398438, + "learning_rate": 1.834081702115542e-05, + "loss": 0.5717, + "step": 6673 + }, + { + "epoch": 0.21050307522472797, + "grad_norm": 2.7245960235595703, + "learning_rate": 1.8340253445267713e-05, + "loss": 0.5692, + "step": 6674 + }, + { + "epoch": 0.21053461599116857, + "grad_norm": 2.2161455154418945, + "learning_rate": 1.833968978234298e-05, + "loss": 0.5516, + "step": 6675 + }, + { + "epoch": 0.2105661567576092, + "grad_norm": 5.312060832977295, + "learning_rate": 1.8339126032387114e-05, + "loss": 0.5702, + "step": 6676 + }, + { + "epoch": 0.21059769752404983, + "grad_norm": 3.5642261505126953, + "learning_rate": 1.8338562195405995e-05, + "loss": 0.5951, + "step": 6677 + }, + { + "epoch": 0.21062923829049046, + "grad_norm": 2.918226718902588, + "learning_rate": 1.8337998271405498e-05, + "loss": 0.5756, + "step": 6678 + }, + { + "epoch": 0.2106607790569311, + "grad_norm": 3.351016044616699, + "learning_rate": 1.833743426039152e-05, + "loss": 0.5888, + "step": 6679 + }, + { + "epoch": 0.21069231982337172, + "grad_norm": 4.155823707580566, + "learning_rate": 1.833687016236994e-05, + "loss": 0.5508, + "step": 6680 + }, + { + "epoch": 0.21072386058981232, + "grad_norm": 2.1013097763061523, + "learning_rate": 1.833630597734665e-05, + "loss": 0.5736, + "step": 6681 + }, + { + "epoch": 0.21075540135625295, + "grad_norm": 11.513503074645996, + "learning_rate": 1.8335741705327527e-05, + "loss": 0.4257, + "step": 6682 + }, + { + "epoch": 0.21078694212269358, + "grad_norm": 3.0454773902893066, + "learning_rate": 1.8335177346318467e-05, + "loss": 0.6169, + "step": 6683 + }, + { + "epoch": 0.2108184828891342, + "grad_norm": 4.0589165687561035, + "learning_rate": 1.8334612900325365e-05, + "loss": 0.5756, + "step": 6684 + }, + { + "epoch": 0.21085002365557484, + "grad_norm": 2.6825480461120605, + "learning_rate": 1.8334048367354102e-05, + "loss": 0.5332, + "step": 6685 + }, + { + "epoch": 0.21088156442201544, + "grad_norm": 2.871746778488159, + "learning_rate": 1.8333483747410573e-05, + "loss": 0.5822, + "step": 6686 + }, + { + "epoch": 0.21091310518845607, + "grad_norm": 2.6608211994171143, + "learning_rate": 1.8332919040500674e-05, + "loss": 0.5601, + "step": 6687 + }, + { + "epoch": 0.2109446459548967, + "grad_norm": 2.7962515354156494, + "learning_rate": 1.833235424663029e-05, + "loss": 0.5511, + "step": 6688 + }, + { + "epoch": 0.21097618672133733, + "grad_norm": 2.224421739578247, + "learning_rate": 1.8331789365805323e-05, + "loss": 0.5704, + "step": 6689 + }, + { + "epoch": 0.21100772748777796, + "grad_norm": 2.4909749031066895, + "learning_rate": 1.8331224398031665e-05, + "loss": 0.5398, + "step": 6690 + }, + { + "epoch": 0.21103926825421856, + "grad_norm": 3.0725462436676025, + "learning_rate": 1.8330659343315208e-05, + "loss": 0.5756, + "step": 6691 + }, + { + "epoch": 0.2110708090206592, + "grad_norm": 2.139578342437744, + "learning_rate": 1.833009420166185e-05, + "loss": 0.9266, + "step": 6692 + }, + { + "epoch": 0.21110234978709982, + "grad_norm": 3.42990779876709, + "learning_rate": 1.83295289730775e-05, + "loss": 0.5706, + "step": 6693 + }, + { + "epoch": 0.21113389055354045, + "grad_norm": 3.152384042739868, + "learning_rate": 1.832896365756804e-05, + "loss": 0.9292, + "step": 6694 + }, + { + "epoch": 0.21116543131998108, + "grad_norm": 2.8193140029907227, + "learning_rate": 1.8328398255139378e-05, + "loss": 0.5309, + "step": 6695 + }, + { + "epoch": 0.2111969720864217, + "grad_norm": 2.441204071044922, + "learning_rate": 1.8327832765797414e-05, + "loss": 0.5948, + "step": 6696 + }, + { + "epoch": 0.21122851285286232, + "grad_norm": 2.128626823425293, + "learning_rate": 1.832726718954805e-05, + "loss": 0.5757, + "step": 6697 + }, + { + "epoch": 0.21126005361930295, + "grad_norm": 2.1619553565979004, + "learning_rate": 1.8326701526397185e-05, + "loss": 0.4364, + "step": 6698 + }, + { + "epoch": 0.21129159438574358, + "grad_norm": 2.9184911251068115, + "learning_rate": 1.8326135776350727e-05, + "loss": 0.5785, + "step": 6699 + }, + { + "epoch": 0.2113231351521842, + "grad_norm": 3.375999689102173, + "learning_rate": 1.8325569939414577e-05, + "loss": 0.5899, + "step": 6700 + }, + { + "epoch": 0.21135467591862483, + "grad_norm": 2.079286813735962, + "learning_rate": 1.8325004015594635e-05, + "loss": 0.5717, + "step": 6701 + }, + { + "epoch": 0.21138621668506544, + "grad_norm": 1.995955467224121, + "learning_rate": 1.8324438004896814e-05, + "loss": 0.8806, + "step": 6702 + }, + { + "epoch": 0.21141775745150607, + "grad_norm": 2.6044633388519287, + "learning_rate": 1.832387190732702e-05, + "loss": 0.4472, + "step": 6703 + }, + { + "epoch": 0.2114492982179467, + "grad_norm": 3.6909019947052, + "learning_rate": 1.832330572289116e-05, + "loss": 0.5709, + "step": 6704 + }, + { + "epoch": 0.21148083898438733, + "grad_norm": 2.470165491104126, + "learning_rate": 1.832273945159514e-05, + "loss": 0.564, + "step": 6705 + }, + { + "epoch": 0.21151237975082796, + "grad_norm": 2.559404134750366, + "learning_rate": 1.8322173093444873e-05, + "loss": 0.4471, + "step": 6706 + }, + { + "epoch": 0.21154392051726856, + "grad_norm": 3.535562515258789, + "learning_rate": 1.8321606648446267e-05, + "loss": 0.3898, + "step": 6707 + }, + { + "epoch": 0.2115754612837092, + "grad_norm": 3.5176634788513184, + "learning_rate": 1.8321040116605235e-05, + "loss": 0.5581, + "step": 6708 + }, + { + "epoch": 0.21160700205014982, + "grad_norm": 9.676607131958008, + "learning_rate": 1.8320473497927686e-05, + "loss": 0.4519, + "step": 6709 + }, + { + "epoch": 0.21163854281659045, + "grad_norm": 2.643329381942749, + "learning_rate": 1.8319906792419536e-05, + "loss": 0.5536, + "step": 6710 + }, + { + "epoch": 0.21167008358303108, + "grad_norm": 2.163996696472168, + "learning_rate": 1.83193400000867e-05, + "loss": 0.406, + "step": 6711 + }, + { + "epoch": 0.21170162434947168, + "grad_norm": 2.4508752822875977, + "learning_rate": 1.831877312093509e-05, + "loss": 0.919, + "step": 6712 + }, + { + "epoch": 0.2117331651159123, + "grad_norm": 2.9959616661071777, + "learning_rate": 1.8318206154970624e-05, + "loss": 0.5938, + "step": 6713 + }, + { + "epoch": 0.21176470588235294, + "grad_norm": 3.699314832687378, + "learning_rate": 1.831763910219922e-05, + "loss": 0.5965, + "step": 6714 + }, + { + "epoch": 0.21179624664879357, + "grad_norm": 2.3963799476623535, + "learning_rate": 1.831707196262679e-05, + "loss": 0.5855, + "step": 6715 + }, + { + "epoch": 0.2118277874152342, + "grad_norm": 3.6668338775634766, + "learning_rate": 1.8316504736259257e-05, + "loss": 0.5772, + "step": 6716 + }, + { + "epoch": 0.21185932818167483, + "grad_norm": 2.1184771060943604, + "learning_rate": 1.8315937423102535e-05, + "loss": 0.5642, + "step": 6717 + }, + { + "epoch": 0.21189086894811543, + "grad_norm": 3.141071081161499, + "learning_rate": 1.8315370023162555e-05, + "loss": 0.8865, + "step": 6718 + }, + { + "epoch": 0.21192240971455606, + "grad_norm": 2.764538526535034, + "learning_rate": 1.8314802536445232e-05, + "loss": 0.5636, + "step": 6719 + }, + { + "epoch": 0.2119539504809967, + "grad_norm": 2.0721163749694824, + "learning_rate": 1.8314234962956485e-05, + "loss": 0.5543, + "step": 6720 + }, + { + "epoch": 0.21198549124743732, + "grad_norm": 2.425971031188965, + "learning_rate": 1.831366730270224e-05, + "loss": 0.5638, + "step": 6721 + }, + { + "epoch": 0.21201703201387795, + "grad_norm": 1.8699504137039185, + "learning_rate": 1.8313099555688426e-05, + "loss": 0.5636, + "step": 6722 + }, + { + "epoch": 0.21204857278031855, + "grad_norm": 2.930262565612793, + "learning_rate": 1.831253172192096e-05, + "loss": 0.5704, + "step": 6723 + }, + { + "epoch": 0.21208011354675918, + "grad_norm": 1.8507194519042969, + "learning_rate": 1.831196380140577e-05, + "loss": 0.5858, + "step": 6724 + }, + { + "epoch": 0.2121116543131998, + "grad_norm": 1.7899006605148315, + "learning_rate": 1.8311395794148786e-05, + "loss": 0.5397, + "step": 6725 + }, + { + "epoch": 0.21214319507964044, + "grad_norm": 2.1026670932769775, + "learning_rate": 1.831082770015593e-05, + "loss": 0.5436, + "step": 6726 + }, + { + "epoch": 0.21217473584608107, + "grad_norm": 1.8518115282058716, + "learning_rate": 1.8310259519433137e-05, + "loss": 0.5757, + "step": 6727 + }, + { + "epoch": 0.21220627661252167, + "grad_norm": 7.963903903961182, + "learning_rate": 1.8309691251986333e-05, + "loss": 0.5649, + "step": 6728 + }, + { + "epoch": 0.2122378173789623, + "grad_norm": 2.159794807434082, + "learning_rate": 1.8309122897821447e-05, + "loss": 0.5578, + "step": 6729 + }, + { + "epoch": 0.21226935814540293, + "grad_norm": 1.9325724840164185, + "learning_rate": 1.8308554456944412e-05, + "loss": 0.9259, + "step": 6730 + }, + { + "epoch": 0.21230089891184356, + "grad_norm": 1.9354139566421509, + "learning_rate": 1.8307985929361162e-05, + "loss": 0.5551, + "step": 6731 + }, + { + "epoch": 0.2123324396782842, + "grad_norm": 2.6858818531036377, + "learning_rate": 1.8307417315077625e-05, + "loss": 0.5873, + "step": 6732 + }, + { + "epoch": 0.21236398044472482, + "grad_norm": 1.8879314661026, + "learning_rate": 1.830684861409974e-05, + "loss": 0.5796, + "step": 6733 + }, + { + "epoch": 0.21239552121116542, + "grad_norm": 2.4961774349212646, + "learning_rate": 1.830627982643344e-05, + "loss": 0.4441, + "step": 6734 + }, + { + "epoch": 0.21242706197760605, + "grad_norm": 2.047642230987549, + "learning_rate": 1.8305710952084657e-05, + "loss": 0.5926, + "step": 6735 + }, + { + "epoch": 0.21245860274404668, + "grad_norm": 1.8902968168258667, + "learning_rate": 1.8305141991059334e-05, + "loss": 0.5487, + "step": 6736 + }, + { + "epoch": 0.2124901435104873, + "grad_norm": 1.712627649307251, + "learning_rate": 1.83045729433634e-05, + "loss": 0.542, + "step": 6737 + }, + { + "epoch": 0.21252168427692794, + "grad_norm": 2.274207830429077, + "learning_rate": 1.8304003809002802e-05, + "loss": 0.5711, + "step": 6738 + }, + { + "epoch": 0.21255322504336854, + "grad_norm": 1.6806182861328125, + "learning_rate": 1.830343458798348e-05, + "loss": 0.5875, + "step": 6739 + }, + { + "epoch": 0.21258476580980917, + "grad_norm": 3.002157688140869, + "learning_rate": 1.830286528031137e-05, + "loss": 0.5724, + "step": 6740 + }, + { + "epoch": 0.2126163065762498, + "grad_norm": 1.9167064428329468, + "learning_rate": 1.8302295885992413e-05, + "loss": 0.5779, + "step": 6741 + }, + { + "epoch": 0.21264784734269043, + "grad_norm": 1.9632012844085693, + "learning_rate": 1.8301726405032547e-05, + "loss": 0.5549, + "step": 6742 + }, + { + "epoch": 0.21267938810913106, + "grad_norm": 1.6790757179260254, + "learning_rate": 1.8301156837437727e-05, + "loss": 0.5737, + "step": 6743 + }, + { + "epoch": 0.21271092887557166, + "grad_norm": 1.9326742887496948, + "learning_rate": 1.8300587183213884e-05, + "loss": 0.5573, + "step": 6744 + }, + { + "epoch": 0.2127424696420123, + "grad_norm": 1.6974254846572876, + "learning_rate": 1.8300017442366972e-05, + "loss": 0.5814, + "step": 6745 + }, + { + "epoch": 0.21277401040845292, + "grad_norm": 2.288694143295288, + "learning_rate": 1.8299447614902933e-05, + "loss": 0.5747, + "step": 6746 + }, + { + "epoch": 0.21280555117489355, + "grad_norm": 2.831690549850464, + "learning_rate": 1.829887770082771e-05, + "loss": 0.5927, + "step": 6747 + }, + { + "epoch": 0.21283709194133418, + "grad_norm": 1.7349807024002075, + "learning_rate": 1.8298307700147256e-05, + "loss": 0.5461, + "step": 6748 + }, + { + "epoch": 0.21286863270777479, + "grad_norm": 4.289912223815918, + "learning_rate": 1.829773761286752e-05, + "loss": 0.5765, + "step": 6749 + }, + { + "epoch": 0.21290017347421542, + "grad_norm": 2.226809024810791, + "learning_rate": 1.829716743899445e-05, + "loss": 0.5676, + "step": 6750 + }, + { + "epoch": 0.21293171424065604, + "grad_norm": 1.8176181316375732, + "learning_rate": 1.829659717853399e-05, + "loss": 0.5555, + "step": 6751 + }, + { + "epoch": 0.21296325500709667, + "grad_norm": 1.8319103717803955, + "learning_rate": 1.8296026831492102e-05, + "loss": 0.5345, + "step": 6752 + }, + { + "epoch": 0.2129947957735373, + "grad_norm": 1.8606023788452148, + "learning_rate": 1.829545639787473e-05, + "loss": 0.6007, + "step": 6753 + }, + { + "epoch": 0.21302633653997793, + "grad_norm": 3.186624765396118, + "learning_rate": 1.8294885877687827e-05, + "loss": 0.5729, + "step": 6754 + }, + { + "epoch": 0.21305787730641854, + "grad_norm": 2.8214099407196045, + "learning_rate": 1.8294315270937352e-05, + "loss": 0.5456, + "step": 6755 + }, + { + "epoch": 0.21308941807285917, + "grad_norm": 2.208658456802368, + "learning_rate": 1.8293744577629254e-05, + "loss": 0.5534, + "step": 6756 + }, + { + "epoch": 0.2131209588392998, + "grad_norm": 4.598232746124268, + "learning_rate": 1.8293173797769497e-05, + "loss": 0.897, + "step": 6757 + }, + { + "epoch": 0.21315249960574043, + "grad_norm": 3.7338149547576904, + "learning_rate": 1.8292602931364026e-05, + "loss": 0.5412, + "step": 6758 + }, + { + "epoch": 0.21318404037218106, + "grad_norm": 2.389033079147339, + "learning_rate": 1.8292031978418808e-05, + "loss": 0.5832, + "step": 6759 + }, + { + "epoch": 0.21321558113862166, + "grad_norm": 3.7435286045074463, + "learning_rate": 1.8291460938939796e-05, + "loss": 0.5683, + "step": 6760 + }, + { + "epoch": 0.2132471219050623, + "grad_norm": 2.056281566619873, + "learning_rate": 1.8290889812932947e-05, + "loss": 0.5967, + "step": 6761 + }, + { + "epoch": 0.21327866267150292, + "grad_norm": 1.9530764818191528, + "learning_rate": 1.829031860040423e-05, + "loss": 0.5674, + "step": 6762 + }, + { + "epoch": 0.21331020343794355, + "grad_norm": 2.368314266204834, + "learning_rate": 1.8289747301359605e-05, + "loss": 0.5856, + "step": 6763 + }, + { + "epoch": 0.21334174420438418, + "grad_norm": 1.9595520496368408, + "learning_rate": 1.8289175915805022e-05, + "loss": 0.564, + "step": 6764 + }, + { + "epoch": 0.21337328497082478, + "grad_norm": 3.371703863143921, + "learning_rate": 1.8288604443746457e-05, + "loss": 0.8723, + "step": 6765 + }, + { + "epoch": 0.2134048257372654, + "grad_norm": 4.424230098724365, + "learning_rate": 1.8288032885189866e-05, + "loss": 0.5224, + "step": 6766 + }, + { + "epoch": 0.21343636650370604, + "grad_norm": 2.233358860015869, + "learning_rate": 1.8287461240141217e-05, + "loss": 0.5918, + "step": 6767 + }, + { + "epoch": 0.21346790727014667, + "grad_norm": 2.2510440349578857, + "learning_rate": 1.8286889508606475e-05, + "loss": 0.5577, + "step": 6768 + }, + { + "epoch": 0.2134994480365873, + "grad_norm": 1.8413259983062744, + "learning_rate": 1.8286317690591608e-05, + "loss": 0.5649, + "step": 6769 + }, + { + "epoch": 0.21353098880302793, + "grad_norm": 2.177706003189087, + "learning_rate": 1.828574578610258e-05, + "loss": 0.562, + "step": 6770 + }, + { + "epoch": 0.21356252956946853, + "grad_norm": 2.3112196922302246, + "learning_rate": 1.828517379514536e-05, + "loss": 0.4417, + "step": 6771 + }, + { + "epoch": 0.21359407033590916, + "grad_norm": 1.6618027687072754, + "learning_rate": 1.828460171772592e-05, + "loss": 0.8746, + "step": 6772 + }, + { + "epoch": 0.2136256111023498, + "grad_norm": 3.4803974628448486, + "learning_rate": 1.828402955385023e-05, + "loss": 0.5722, + "step": 6773 + }, + { + "epoch": 0.21365715186879042, + "grad_norm": 3.079660177230835, + "learning_rate": 1.8283457303524256e-05, + "loss": 0.5925, + "step": 6774 + }, + { + "epoch": 0.21368869263523105, + "grad_norm": 1.943420171737671, + "learning_rate": 1.8282884966753977e-05, + "loss": 0.8961, + "step": 6775 + }, + { + "epoch": 0.21372023340167165, + "grad_norm": 8.124831199645996, + "learning_rate": 1.8282312543545354e-05, + "loss": 0.5781, + "step": 6776 + }, + { + "epoch": 0.21375177416811228, + "grad_norm": 2.2111093997955322, + "learning_rate": 1.8281740033904375e-05, + "loss": 0.5288, + "step": 6777 + }, + { + "epoch": 0.2137833149345529, + "grad_norm": 3.5235090255737305, + "learning_rate": 1.828116743783701e-05, + "loss": 0.5872, + "step": 6778 + }, + { + "epoch": 0.21381485570099354, + "grad_norm": 2.260303497314453, + "learning_rate": 1.828059475534923e-05, + "loss": 0.5781, + "step": 6779 + }, + { + "epoch": 0.21384639646743417, + "grad_norm": 1.9864475727081299, + "learning_rate": 1.828002198644701e-05, + "loss": 0.5632, + "step": 6780 + }, + { + "epoch": 0.21387793723387477, + "grad_norm": 4.216651439666748, + "learning_rate": 1.8279449131136337e-05, + "loss": 0.5636, + "step": 6781 + }, + { + "epoch": 0.2139094780003154, + "grad_norm": 1.8676018714904785, + "learning_rate": 1.827887618942318e-05, + "loss": 0.5639, + "step": 6782 + }, + { + "epoch": 0.21394101876675603, + "grad_norm": 2.8029274940490723, + "learning_rate": 1.827830316131352e-05, + "loss": 0.5683, + "step": 6783 + }, + { + "epoch": 0.21397255953319666, + "grad_norm": 1.990625023841858, + "learning_rate": 1.8277730046813344e-05, + "loss": 0.5861, + "step": 6784 + }, + { + "epoch": 0.2140041002996373, + "grad_norm": 2.3836772441864014, + "learning_rate": 1.8277156845928624e-05, + "loss": 0.5799, + "step": 6785 + }, + { + "epoch": 0.2140356410660779, + "grad_norm": 4.208896160125732, + "learning_rate": 1.8276583558665345e-05, + "loss": 0.927, + "step": 6786 + }, + { + "epoch": 0.21406718183251852, + "grad_norm": 2.3794190883636475, + "learning_rate": 1.8276010185029492e-05, + "loss": 0.5703, + "step": 6787 + }, + { + "epoch": 0.21409872259895915, + "grad_norm": 2.501871347427368, + "learning_rate": 1.8275436725027046e-05, + "loss": 0.546, + "step": 6788 + }, + { + "epoch": 0.21413026336539978, + "grad_norm": 2.6455254554748535, + "learning_rate": 1.827486317866399e-05, + "loss": 0.5664, + "step": 6789 + }, + { + "epoch": 0.2141618041318404, + "grad_norm": 1.916340947151184, + "learning_rate": 1.8274289545946313e-05, + "loss": 0.5712, + "step": 6790 + }, + { + "epoch": 0.21419334489828104, + "grad_norm": 2.254605770111084, + "learning_rate": 1.827371582688e-05, + "loss": 0.5661, + "step": 6791 + }, + { + "epoch": 0.21422488566472164, + "grad_norm": 1.6737706661224365, + "learning_rate": 1.827314202147104e-05, + "loss": 0.3994, + "step": 6792 + }, + { + "epoch": 0.21425642643116227, + "grad_norm": 2.0808374881744385, + "learning_rate": 1.8272568129725418e-05, + "loss": 0.5791, + "step": 6793 + }, + { + "epoch": 0.2142879671976029, + "grad_norm": 1.8865216970443726, + "learning_rate": 1.827199415164912e-05, + "loss": 0.8834, + "step": 6794 + }, + { + "epoch": 0.21431950796404353, + "grad_norm": 2.8676061630249023, + "learning_rate": 1.8271420087248145e-05, + "loss": 0.5858, + "step": 6795 + }, + { + "epoch": 0.21435104873048416, + "grad_norm": 2.2573320865631104, + "learning_rate": 1.8270845936528478e-05, + "loss": 0.5639, + "step": 6796 + }, + { + "epoch": 0.21438258949692476, + "grad_norm": 2.3049700260162354, + "learning_rate": 1.827027169949611e-05, + "loss": 0.4774, + "step": 6797 + }, + { + "epoch": 0.2144141302633654, + "grad_norm": 2.2865657806396484, + "learning_rate": 1.8269697376157033e-05, + "loss": 0.5466, + "step": 6798 + }, + { + "epoch": 0.21444567102980602, + "grad_norm": 2.9224424362182617, + "learning_rate": 1.826912296651725e-05, + "loss": 0.564, + "step": 6799 + }, + { + "epoch": 0.21447721179624665, + "grad_norm": 1.9478847980499268, + "learning_rate": 1.8268548470582738e-05, + "loss": 0.6022, + "step": 6800 + }, + { + "epoch": 0.21450875256268728, + "grad_norm": 1.9955731630325317, + "learning_rate": 1.826797388835951e-05, + "loss": 0.5397, + "step": 6801 + }, + { + "epoch": 0.21454029332912788, + "grad_norm": 3.8638038635253906, + "learning_rate": 1.8267399219853552e-05, + "loss": 0.567, + "step": 6802 + }, + { + "epoch": 0.21457183409556851, + "grad_norm": 2.5961005687713623, + "learning_rate": 1.826682446507086e-05, + "loss": 0.4009, + "step": 6803 + }, + { + "epoch": 0.21460337486200914, + "grad_norm": 2.376195192337036, + "learning_rate": 1.826624962401744e-05, + "loss": 0.5461, + "step": 6804 + }, + { + "epoch": 0.21463491562844977, + "grad_norm": 2.208913564682007, + "learning_rate": 1.8265674696699286e-05, + "loss": 0.9223, + "step": 6805 + }, + { + "epoch": 0.2146664563948904, + "grad_norm": 2.45992374420166, + "learning_rate": 1.8265099683122402e-05, + "loss": 0.5598, + "step": 6806 + }, + { + "epoch": 0.21469799716133103, + "grad_norm": 2.532285451889038, + "learning_rate": 1.826452458329278e-05, + "loss": 0.5661, + "step": 6807 + }, + { + "epoch": 0.21472953792777164, + "grad_norm": 2.433109760284424, + "learning_rate": 1.826394939721643e-05, + "loss": 0.5837, + "step": 6808 + }, + { + "epoch": 0.21476107869421227, + "grad_norm": 2.5550711154937744, + "learning_rate": 1.826337412489935e-05, + "loss": 0.5644, + "step": 6809 + }, + { + "epoch": 0.2147926194606529, + "grad_norm": 2.8974156379699707, + "learning_rate": 1.8262798766347542e-05, + "loss": 0.602, + "step": 6810 + }, + { + "epoch": 0.21482416022709352, + "grad_norm": 2.0946903228759766, + "learning_rate": 1.8262223321567014e-05, + "loss": 0.4087, + "step": 6811 + }, + { + "epoch": 0.21485570099353415, + "grad_norm": 2.0758187770843506, + "learning_rate": 1.826164779056377e-05, + "loss": 0.567, + "step": 6812 + }, + { + "epoch": 0.21488724175997476, + "grad_norm": 2.1714401245117188, + "learning_rate": 1.8261072173343823e-05, + "loss": 0.5563, + "step": 6813 + }, + { + "epoch": 0.2149187825264154, + "grad_norm": 1.88141930103302, + "learning_rate": 1.8260496469913166e-05, + "loss": 0.5393, + "step": 6814 + }, + { + "epoch": 0.21495032329285602, + "grad_norm": 2.448253870010376, + "learning_rate": 1.8259920680277817e-05, + "loss": 0.5787, + "step": 6815 + }, + { + "epoch": 0.21498186405929665, + "grad_norm": 2.0717976093292236, + "learning_rate": 1.825934480444378e-05, + "loss": 0.5745, + "step": 6816 + }, + { + "epoch": 0.21501340482573728, + "grad_norm": 2.1668174266815186, + "learning_rate": 1.8258768842417066e-05, + "loss": 0.5789, + "step": 6817 + }, + { + "epoch": 0.21504494559217788, + "grad_norm": 1.8281491994857788, + "learning_rate": 1.8258192794203688e-05, + "loss": 0.5127, + "step": 6818 + }, + { + "epoch": 0.2150764863586185, + "grad_norm": 2.1096348762512207, + "learning_rate": 1.8257616659809657e-05, + "loss": 0.5778, + "step": 6819 + }, + { + "epoch": 0.21510802712505914, + "grad_norm": 2.016234874725342, + "learning_rate": 1.8257040439240984e-05, + "loss": 0.5501, + "step": 6820 + }, + { + "epoch": 0.21513956789149977, + "grad_norm": 2.023848056793213, + "learning_rate": 1.8256464132503683e-05, + "loss": 0.4128, + "step": 6821 + }, + { + "epoch": 0.2151711086579404, + "grad_norm": 2.297861337661743, + "learning_rate": 1.8255887739603766e-05, + "loss": 0.5567, + "step": 6822 + }, + { + "epoch": 0.21520264942438103, + "grad_norm": 3.209291458129883, + "learning_rate": 1.825531126054725e-05, + "loss": 0.5953, + "step": 6823 + }, + { + "epoch": 0.21523419019082163, + "grad_norm": 4.206735134124756, + "learning_rate": 1.8254734695340154e-05, + "loss": 0.8928, + "step": 6824 + }, + { + "epoch": 0.21526573095726226, + "grad_norm": 2.213750123977661, + "learning_rate": 1.825415804398849e-05, + "loss": 0.5738, + "step": 6825 + }, + { + "epoch": 0.2152972717237029, + "grad_norm": 3.6412999629974365, + "learning_rate": 1.8253581306498275e-05, + "loss": 0.5752, + "step": 6826 + }, + { + "epoch": 0.21532881249014352, + "grad_norm": 1.9442416429519653, + "learning_rate": 1.8253004482875533e-05, + "loss": 0.5771, + "step": 6827 + }, + { + "epoch": 0.21536035325658415, + "grad_norm": 8.289064407348633, + "learning_rate": 1.8252427573126282e-05, + "loss": 0.4022, + "step": 6828 + }, + { + "epoch": 0.21539189402302475, + "grad_norm": 2.5890352725982666, + "learning_rate": 1.825185057725654e-05, + "loss": 0.5786, + "step": 6829 + }, + { + "epoch": 0.21542343478946538, + "grad_norm": 1.8557473421096802, + "learning_rate": 1.825127349527233e-05, + "loss": 0.5755, + "step": 6830 + }, + { + "epoch": 0.215454975555906, + "grad_norm": 3.9805924892425537, + "learning_rate": 1.825069632717967e-05, + "loss": 0.5597, + "step": 6831 + }, + { + "epoch": 0.21548651632234664, + "grad_norm": 1.695425033569336, + "learning_rate": 1.8250119072984592e-05, + "loss": 0.887, + "step": 6832 + }, + { + "epoch": 0.21551805708878727, + "grad_norm": 1.7286558151245117, + "learning_rate": 1.8249541732693116e-05, + "loss": 0.8901, + "step": 6833 + }, + { + "epoch": 0.21554959785522787, + "grad_norm": 2.485725164413452, + "learning_rate": 1.8248964306311263e-05, + "loss": 0.8969, + "step": 6834 + }, + { + "epoch": 0.2155811386216685, + "grad_norm": 2.093273878097534, + "learning_rate": 1.8248386793845063e-05, + "loss": 0.5676, + "step": 6835 + }, + { + "epoch": 0.21561267938810913, + "grad_norm": 2.045389175415039, + "learning_rate": 1.8247809195300546e-05, + "loss": 0.5764, + "step": 6836 + }, + { + "epoch": 0.21564422015454976, + "grad_norm": 3.012702703475952, + "learning_rate": 1.824723151068373e-05, + "loss": 0.5901, + "step": 6837 + }, + { + "epoch": 0.2156757609209904, + "grad_norm": 1.9577893018722534, + "learning_rate": 1.8246653740000655e-05, + "loss": 0.5645, + "step": 6838 + }, + { + "epoch": 0.215707301687431, + "grad_norm": 1.8344343900680542, + "learning_rate": 1.824607588325734e-05, + "loss": 0.5494, + "step": 6839 + }, + { + "epoch": 0.21573884245387162, + "grad_norm": 3.7513949871063232, + "learning_rate": 1.8245497940459823e-05, + "loss": 0.5436, + "step": 6840 + }, + { + "epoch": 0.21577038322031225, + "grad_norm": 2.1276957988739014, + "learning_rate": 1.824491991161413e-05, + "loss": 0.5939, + "step": 6841 + }, + { + "epoch": 0.21580192398675288, + "grad_norm": 1.7745970487594604, + "learning_rate": 1.82443417967263e-05, + "loss": 0.4002, + "step": 6842 + }, + { + "epoch": 0.2158334647531935, + "grad_norm": 2.4037606716156006, + "learning_rate": 1.824376359580236e-05, + "loss": 0.5833, + "step": 6843 + }, + { + "epoch": 0.21586500551963414, + "grad_norm": 5.466787338256836, + "learning_rate": 1.824318530884835e-05, + "loss": 0.5758, + "step": 6844 + }, + { + "epoch": 0.21589654628607474, + "grad_norm": 2.4153997898101807, + "learning_rate": 1.824260693587029e-05, + "loss": 0.5575, + "step": 6845 + }, + { + "epoch": 0.21592808705251537, + "grad_norm": 3.061108112335205, + "learning_rate": 1.8242028476874236e-05, + "loss": 0.5634, + "step": 6846 + }, + { + "epoch": 0.215959627818956, + "grad_norm": 2.4745564460754395, + "learning_rate": 1.824144993186621e-05, + "loss": 0.5659, + "step": 6847 + }, + { + "epoch": 0.21599116858539663, + "grad_norm": 3.5945920944213867, + "learning_rate": 1.8240871300852256e-05, + "loss": 0.5888, + "step": 6848 + }, + { + "epoch": 0.21602270935183726, + "grad_norm": 2.46536922454834, + "learning_rate": 1.824029258383841e-05, + "loss": 0.5635, + "step": 6849 + }, + { + "epoch": 0.21605425011827786, + "grad_norm": 3.5609588623046875, + "learning_rate": 1.8239713780830715e-05, + "loss": 0.5596, + "step": 6850 + }, + { + "epoch": 0.2160857908847185, + "grad_norm": 11.661822319030762, + "learning_rate": 1.8239134891835206e-05, + "loss": 0.9016, + "step": 6851 + }, + { + "epoch": 0.21611733165115912, + "grad_norm": 2.7448678016662598, + "learning_rate": 1.823855591685793e-05, + "loss": 0.572, + "step": 6852 + }, + { + "epoch": 0.21614887241759975, + "grad_norm": 2.1380295753479004, + "learning_rate": 1.8237976855904924e-05, + "loss": 0.56, + "step": 6853 + }, + { + "epoch": 0.21618041318404038, + "grad_norm": 1.9065744876861572, + "learning_rate": 1.8237397708982234e-05, + "loss": 0.4041, + "step": 6854 + }, + { + "epoch": 0.21621195395048098, + "grad_norm": 3.2937424182891846, + "learning_rate": 1.8236818476095902e-05, + "loss": 0.5806, + "step": 6855 + }, + { + "epoch": 0.2162434947169216, + "grad_norm": 2.416888475418091, + "learning_rate": 1.823623915725197e-05, + "loss": 0.5741, + "step": 6856 + }, + { + "epoch": 0.21627503548336224, + "grad_norm": 2.938460111618042, + "learning_rate": 1.823565975245649e-05, + "loss": 0.5794, + "step": 6857 + }, + { + "epoch": 0.21630657624980287, + "grad_norm": 1.9168381690979004, + "learning_rate": 1.8235080261715508e-05, + "loss": 0.4332, + "step": 6858 + }, + { + "epoch": 0.2163381170162435, + "grad_norm": 2.4519762992858887, + "learning_rate": 1.8234500685035067e-05, + "loss": 0.5916, + "step": 6859 + }, + { + "epoch": 0.21636965778268413, + "grad_norm": 1.673237681388855, + "learning_rate": 1.823392102242122e-05, + "loss": 0.3938, + "step": 6860 + }, + { + "epoch": 0.21640119854912473, + "grad_norm": 3.424506902694702, + "learning_rate": 1.823334127388001e-05, + "loss": 0.5977, + "step": 6861 + }, + { + "epoch": 0.21643273931556536, + "grad_norm": 2.492581367492676, + "learning_rate": 1.8232761439417494e-05, + "loss": 0.5513, + "step": 6862 + }, + { + "epoch": 0.216464280082006, + "grad_norm": 3.1001291275024414, + "learning_rate": 1.8232181519039716e-05, + "loss": 0.5932, + "step": 6863 + }, + { + "epoch": 0.21649582084844662, + "grad_norm": 2.366194009780884, + "learning_rate": 1.8231601512752736e-05, + "loss": 0.5747, + "step": 6864 + }, + { + "epoch": 0.21652736161488725, + "grad_norm": 2.3583593368530273, + "learning_rate": 1.8231021420562598e-05, + "loss": 0.5872, + "step": 6865 + }, + { + "epoch": 0.21655890238132786, + "grad_norm": 3.3086321353912354, + "learning_rate": 1.8230441242475365e-05, + "loss": 0.9307, + "step": 6866 + }, + { + "epoch": 0.21659044314776849, + "grad_norm": 2.1990978717803955, + "learning_rate": 1.8229860978497082e-05, + "loss": 0.5709, + "step": 6867 + }, + { + "epoch": 0.21662198391420912, + "grad_norm": 2.661220073699951, + "learning_rate": 1.822928062863381e-05, + "loss": 0.5578, + "step": 6868 + }, + { + "epoch": 0.21665352468064975, + "grad_norm": 2.481905221939087, + "learning_rate": 1.8228700192891605e-05, + "loss": 0.5517, + "step": 6869 + }, + { + "epoch": 0.21668506544709037, + "grad_norm": 3.232391357421875, + "learning_rate": 1.822811967127653e-05, + "loss": 0.5934, + "step": 6870 + }, + { + "epoch": 0.21671660621353098, + "grad_norm": 2.3144571781158447, + "learning_rate": 1.822753906379463e-05, + "loss": 0.585, + "step": 6871 + }, + { + "epoch": 0.2167481469799716, + "grad_norm": 2.457057476043701, + "learning_rate": 1.8226958370451976e-05, + "loss": 0.5617, + "step": 6872 + }, + { + "epoch": 0.21677968774641224, + "grad_norm": 3.8812949657440186, + "learning_rate": 1.822637759125462e-05, + "loss": 0.5615, + "step": 6873 + }, + { + "epoch": 0.21681122851285287, + "grad_norm": 4.325790882110596, + "learning_rate": 1.822579672620863e-05, + "loss": 0.8866, + "step": 6874 + }, + { + "epoch": 0.2168427692792935, + "grad_norm": 2.6446335315704346, + "learning_rate": 1.822521577532006e-05, + "loss": 0.6065, + "step": 6875 + }, + { + "epoch": 0.2168743100457341, + "grad_norm": 3.445925235748291, + "learning_rate": 1.8224634738594978e-05, + "loss": 0.5752, + "step": 6876 + }, + { + "epoch": 0.21690585081217473, + "grad_norm": 1.914482831954956, + "learning_rate": 1.8224053616039445e-05, + "loss": 0.8994, + "step": 6877 + }, + { + "epoch": 0.21693739157861536, + "grad_norm": 2.607299566268921, + "learning_rate": 1.8223472407659528e-05, + "loss": 0.5427, + "step": 6878 + }, + { + "epoch": 0.216968932345056, + "grad_norm": 2.38691782951355, + "learning_rate": 1.8222891113461293e-05, + "loss": 0.5325, + "step": 6879 + }, + { + "epoch": 0.21700047311149662, + "grad_norm": 2.3067123889923096, + "learning_rate": 1.8222309733450804e-05, + "loss": 0.5533, + "step": 6880 + }, + { + "epoch": 0.21703201387793725, + "grad_norm": 2.394726514816284, + "learning_rate": 1.8221728267634126e-05, + "loss": 0.8684, + "step": 6881 + }, + { + "epoch": 0.21706355464437785, + "grad_norm": 3.412292957305908, + "learning_rate": 1.822114671601733e-05, + "loss": 0.5929, + "step": 6882 + }, + { + "epoch": 0.21709509541081848, + "grad_norm": 3.5294997692108154, + "learning_rate": 1.8220565078606488e-05, + "loss": 0.581, + "step": 6883 + }, + { + "epoch": 0.2171266361772591, + "grad_norm": 2.5534141063690186, + "learning_rate": 1.821998335540766e-05, + "loss": 0.5854, + "step": 6884 + }, + { + "epoch": 0.21715817694369974, + "grad_norm": 3.9497218132019043, + "learning_rate": 1.8219401546426927e-05, + "loss": 0.5633, + "step": 6885 + }, + { + "epoch": 0.21718971771014037, + "grad_norm": 2.0119144916534424, + "learning_rate": 1.8218819651670356e-05, + "loss": 0.5662, + "step": 6886 + }, + { + "epoch": 0.21722125847658097, + "grad_norm": 3.033705234527588, + "learning_rate": 1.821823767114402e-05, + "loss": 0.5546, + "step": 6887 + }, + { + "epoch": 0.2172527992430216, + "grad_norm": 2.601264476776123, + "learning_rate": 1.8217655604853993e-05, + "loss": 0.5589, + "step": 6888 + }, + { + "epoch": 0.21728434000946223, + "grad_norm": 4.4673895835876465, + "learning_rate": 1.821707345280635e-05, + "loss": 0.8802, + "step": 6889 + }, + { + "epoch": 0.21731588077590286, + "grad_norm": 2.0699052810668945, + "learning_rate": 1.821649121500716e-05, + "loss": 0.5421, + "step": 6890 + }, + { + "epoch": 0.2173474215423435, + "grad_norm": 3.6801834106445312, + "learning_rate": 1.821590889146251e-05, + "loss": 0.5951, + "step": 6891 + }, + { + "epoch": 0.2173789623087841, + "grad_norm": 2.0458834171295166, + "learning_rate": 1.821532648217847e-05, + "loss": 0.579, + "step": 6892 + }, + { + "epoch": 0.21741050307522472, + "grad_norm": 1.8865114450454712, + "learning_rate": 1.821474398716112e-05, + "loss": 0.9197, + "step": 6893 + }, + { + "epoch": 0.21744204384166535, + "grad_norm": 3.3609495162963867, + "learning_rate": 1.8214161406416534e-05, + "loss": 0.8737, + "step": 6894 + }, + { + "epoch": 0.21747358460810598, + "grad_norm": 3.0231282711029053, + "learning_rate": 1.82135787399508e-05, + "loss": 0.4277, + "step": 6895 + }, + { + "epoch": 0.2175051253745466, + "grad_norm": 2.412224769592285, + "learning_rate": 1.8212995987769992e-05, + "loss": 0.5522, + "step": 6896 + }, + { + "epoch": 0.21753666614098724, + "grad_norm": 1.9973623752593994, + "learning_rate": 1.8212413149880194e-05, + "loss": 0.5619, + "step": 6897 + }, + { + "epoch": 0.21756820690742784, + "grad_norm": 4.715005397796631, + "learning_rate": 1.8211830226287487e-05, + "loss": 0.8484, + "step": 6898 + }, + { + "epoch": 0.21759974767386847, + "grad_norm": 2.3387320041656494, + "learning_rate": 1.8211247216997957e-05, + "loss": 0.5757, + "step": 6899 + }, + { + "epoch": 0.2176312884403091, + "grad_norm": 1.8520575761795044, + "learning_rate": 1.8210664122017684e-05, + "loss": 0.8635, + "step": 6900 + }, + { + "epoch": 0.21766282920674973, + "grad_norm": 1.6090952157974243, + "learning_rate": 1.821008094135276e-05, + "loss": 0.5406, + "step": 6901 + }, + { + "epoch": 0.21769436997319036, + "grad_norm": 2.361970901489258, + "learning_rate": 1.820949767500926e-05, + "loss": 0.5405, + "step": 6902 + }, + { + "epoch": 0.21772591073963096, + "grad_norm": 1.9860600233078003, + "learning_rate": 1.820891432299328e-05, + "loss": 0.4118, + "step": 6903 + }, + { + "epoch": 0.2177574515060716, + "grad_norm": 1.5762460231781006, + "learning_rate": 1.8208330885310907e-05, + "loss": 0.5655, + "step": 6904 + }, + { + "epoch": 0.21778899227251222, + "grad_norm": 2.6632626056671143, + "learning_rate": 1.8207747361968228e-05, + "loss": 0.536, + "step": 6905 + }, + { + "epoch": 0.21782053303895285, + "grad_norm": 2.5808658599853516, + "learning_rate": 1.8207163752971327e-05, + "loss": 0.5481, + "step": 6906 + }, + { + "epoch": 0.21785207380539348, + "grad_norm": 3.6729238033294678, + "learning_rate": 1.8206580058326304e-05, + "loss": 0.5775, + "step": 6907 + }, + { + "epoch": 0.21788361457183408, + "grad_norm": 1.793377161026001, + "learning_rate": 1.8205996278039247e-05, + "loss": 0.5171, + "step": 6908 + }, + { + "epoch": 0.2179151553382747, + "grad_norm": 3.5779263973236084, + "learning_rate": 1.8205412412116243e-05, + "loss": 0.5713, + "step": 6909 + }, + { + "epoch": 0.21794669610471534, + "grad_norm": 2.618910312652588, + "learning_rate": 1.820482846056339e-05, + "loss": 0.5956, + "step": 6910 + }, + { + "epoch": 0.21797823687115597, + "grad_norm": 2.2820186614990234, + "learning_rate": 1.8204244423386784e-05, + "loss": 0.5489, + "step": 6911 + }, + { + "epoch": 0.2180097776375966, + "grad_norm": 1.9982281923294067, + "learning_rate": 1.8203660300592515e-05, + "loss": 0.5499, + "step": 6912 + }, + { + "epoch": 0.2180413184040372, + "grad_norm": 2.1166858673095703, + "learning_rate": 1.820307609218668e-05, + "loss": 0.5317, + "step": 6913 + }, + { + "epoch": 0.21807285917047783, + "grad_norm": 2.7905406951904297, + "learning_rate": 1.8202491798175378e-05, + "loss": 0.5488, + "step": 6914 + }, + { + "epoch": 0.21810439993691846, + "grad_norm": 3.3015902042388916, + "learning_rate": 1.8201907418564704e-05, + "loss": 0.4131, + "step": 6915 + }, + { + "epoch": 0.2181359407033591, + "grad_norm": 2.0599260330200195, + "learning_rate": 1.8201322953360758e-05, + "loss": 0.5539, + "step": 6916 + }, + { + "epoch": 0.21816748146979972, + "grad_norm": 2.5995209217071533, + "learning_rate": 1.8200738402569638e-05, + "loss": 0.5649, + "step": 6917 + }, + { + "epoch": 0.21819902223624035, + "grad_norm": 2.2364346981048584, + "learning_rate": 1.8200153766197446e-05, + "loss": 0.5595, + "step": 6918 + }, + { + "epoch": 0.21823056300268096, + "grad_norm": 2.0599842071533203, + "learning_rate": 1.819956904425028e-05, + "loss": 0.5683, + "step": 6919 + }, + { + "epoch": 0.21826210376912158, + "grad_norm": 3.314469814300537, + "learning_rate": 1.8198984236734246e-05, + "loss": 0.591, + "step": 6920 + }, + { + "epoch": 0.21829364453556221, + "grad_norm": 4.795950889587402, + "learning_rate": 1.8198399343655445e-05, + "loss": 0.5641, + "step": 6921 + }, + { + "epoch": 0.21832518530200284, + "grad_norm": 2.45804500579834, + "learning_rate": 1.819781436501998e-05, + "loss": 0.5773, + "step": 6922 + }, + { + "epoch": 0.21835672606844347, + "grad_norm": 2.430284023284912, + "learning_rate": 1.819722930083396e-05, + "loss": 0.5454, + "step": 6923 + }, + { + "epoch": 0.21838826683488408, + "grad_norm": 3.0067827701568604, + "learning_rate": 1.8196644151103486e-05, + "loss": 0.5302, + "step": 6924 + }, + { + "epoch": 0.2184198076013247, + "grad_norm": 1.9030717611312866, + "learning_rate": 1.8196058915834664e-05, + "loss": 0.5845, + "step": 6925 + }, + { + "epoch": 0.21845134836776534, + "grad_norm": 2.9222207069396973, + "learning_rate": 1.8195473595033607e-05, + "loss": 0.5639, + "step": 6926 + }, + { + "epoch": 0.21848288913420597, + "grad_norm": 3.5828468799591064, + "learning_rate": 1.8194888188706413e-05, + "loss": 0.5645, + "step": 6927 + }, + { + "epoch": 0.2185144299006466, + "grad_norm": 2.086477756500244, + "learning_rate": 1.81943026968592e-05, + "loss": 0.5261, + "step": 6928 + }, + { + "epoch": 0.2185459706670872, + "grad_norm": 2.0715277194976807, + "learning_rate": 1.8193717119498078e-05, + "loss": 0.5686, + "step": 6929 + }, + { + "epoch": 0.21857751143352783, + "grad_norm": 1.8198615312576294, + "learning_rate": 1.8193131456629155e-05, + "loss": 0.5298, + "step": 6930 + }, + { + "epoch": 0.21860905219996846, + "grad_norm": 1.810465693473816, + "learning_rate": 1.819254570825854e-05, + "loss": 0.5242, + "step": 6931 + }, + { + "epoch": 0.2186405929664091, + "grad_norm": 2.3698980808258057, + "learning_rate": 1.8191959874392355e-05, + "loss": 0.3877, + "step": 6932 + }, + { + "epoch": 0.21867213373284972, + "grad_norm": 2.190279722213745, + "learning_rate": 1.8191373955036705e-05, + "loss": 0.5397, + "step": 6933 + }, + { + "epoch": 0.21870367449929035, + "grad_norm": 2.015702247619629, + "learning_rate": 1.8190787950197704e-05, + "loss": 0.5399, + "step": 6934 + }, + { + "epoch": 0.21873521526573095, + "grad_norm": 2.3528101444244385, + "learning_rate": 1.819020185988148e-05, + "loss": 0.562, + "step": 6935 + }, + { + "epoch": 0.21876675603217158, + "grad_norm": 4.405192852020264, + "learning_rate": 1.818961568409413e-05, + "loss": 0.5504, + "step": 6936 + }, + { + "epoch": 0.2187982967986122, + "grad_norm": 4.513250350952148, + "learning_rate": 1.8189029422841786e-05, + "loss": 0.5858, + "step": 6937 + }, + { + "epoch": 0.21882983756505284, + "grad_norm": 2.1365137100219727, + "learning_rate": 1.818844307613056e-05, + "loss": 0.5574, + "step": 6938 + }, + { + "epoch": 0.21886137833149347, + "grad_norm": 4.391490936279297, + "learning_rate": 1.818785664396657e-05, + "loss": 0.5441, + "step": 6939 + }, + { + "epoch": 0.21889291909793407, + "grad_norm": 3.8610665798187256, + "learning_rate": 1.8187270126355946e-05, + "loss": 0.5699, + "step": 6940 + }, + { + "epoch": 0.2189244598643747, + "grad_norm": 4.585278034210205, + "learning_rate": 1.8186683523304795e-05, + "loss": 0.5736, + "step": 6941 + }, + { + "epoch": 0.21895600063081533, + "grad_norm": 2.1994354724884033, + "learning_rate": 1.8186096834819245e-05, + "loss": 0.5525, + "step": 6942 + }, + { + "epoch": 0.21898754139725596, + "grad_norm": 2.6934826374053955, + "learning_rate": 1.8185510060905418e-05, + "loss": 0.528, + "step": 6943 + }, + { + "epoch": 0.2190190821636966, + "grad_norm": 2.453362464904785, + "learning_rate": 1.818492320156944e-05, + "loss": 0.5369, + "step": 6944 + }, + { + "epoch": 0.2190506229301372, + "grad_norm": 1.9513728618621826, + "learning_rate": 1.8184336256817428e-05, + "loss": 0.5575, + "step": 6945 + }, + { + "epoch": 0.21908216369657782, + "grad_norm": 2.1383960247039795, + "learning_rate": 1.8183749226655518e-05, + "loss": 0.5434, + "step": 6946 + }, + { + "epoch": 0.21911370446301845, + "grad_norm": 2.5035502910614014, + "learning_rate": 1.8183162111089826e-05, + "loss": 0.5687, + "step": 6947 + }, + { + "epoch": 0.21914524522945908, + "grad_norm": 1.9716044664382935, + "learning_rate": 1.8182574910126487e-05, + "loss": 0.561, + "step": 6948 + }, + { + "epoch": 0.2191767859958997, + "grad_norm": 5.75026798248291, + "learning_rate": 1.818198762377162e-05, + "loss": 0.5172, + "step": 6949 + }, + { + "epoch": 0.21920832676234034, + "grad_norm": 2.6127164363861084, + "learning_rate": 1.8181400252031365e-05, + "loss": 0.604, + "step": 6950 + }, + { + "epoch": 0.21923986752878094, + "grad_norm": 1.76158607006073, + "learning_rate": 1.8180812794911845e-05, + "loss": 0.5359, + "step": 6951 + }, + { + "epoch": 0.21927140829522157, + "grad_norm": 2.210156202316284, + "learning_rate": 1.8180225252419187e-05, + "loss": 0.5753, + "step": 6952 + }, + { + "epoch": 0.2193029490616622, + "grad_norm": 2.3363256454467773, + "learning_rate": 1.817963762455953e-05, + "loss": 0.5629, + "step": 6953 + }, + { + "epoch": 0.21933448982810283, + "grad_norm": 4.23211669921875, + "learning_rate": 1.8179049911339e-05, + "loss": 0.5448, + "step": 6954 + }, + { + "epoch": 0.21936603059454346, + "grad_norm": 2.0333104133605957, + "learning_rate": 1.8178462112763736e-05, + "loss": 0.5762, + "step": 6955 + }, + { + "epoch": 0.21939757136098406, + "grad_norm": 2.047410488128662, + "learning_rate": 1.8177874228839872e-05, + "loss": 0.5715, + "step": 6956 + }, + { + "epoch": 0.2194291121274247, + "grad_norm": 1.712843418121338, + "learning_rate": 1.8177286259573536e-05, + "loss": 0.5755, + "step": 6957 + }, + { + "epoch": 0.21946065289386532, + "grad_norm": 3.347313404083252, + "learning_rate": 1.817669820497087e-05, + "loss": 0.5809, + "step": 6958 + }, + { + "epoch": 0.21949219366030595, + "grad_norm": 2.0959572792053223, + "learning_rate": 1.817611006503801e-05, + "loss": 0.5292, + "step": 6959 + }, + { + "epoch": 0.21952373442674658, + "grad_norm": 2.1421968936920166, + "learning_rate": 1.8175521839781088e-05, + "loss": 0.5686, + "step": 6960 + }, + { + "epoch": 0.21955527519318718, + "grad_norm": 2.4010114669799805, + "learning_rate": 1.8174933529206254e-05, + "loss": 0.588, + "step": 6961 + }, + { + "epoch": 0.2195868159596278, + "grad_norm": 2.2711985111236572, + "learning_rate": 1.8174345133319638e-05, + "loss": 0.5444, + "step": 6962 + }, + { + "epoch": 0.21961835672606844, + "grad_norm": 3.6699769496917725, + "learning_rate": 1.8173756652127385e-05, + "loss": 0.5743, + "step": 6963 + }, + { + "epoch": 0.21964989749250907, + "grad_norm": 1.7937334775924683, + "learning_rate": 1.8173168085635634e-05, + "loss": 0.5439, + "step": 6964 + }, + { + "epoch": 0.2196814382589497, + "grad_norm": 7.282869815826416, + "learning_rate": 1.8172579433850525e-05, + "loss": 0.889, + "step": 6965 + }, + { + "epoch": 0.2197129790253903, + "grad_norm": 2.8446590900421143, + "learning_rate": 1.8171990696778207e-05, + "loss": 0.5866, + "step": 6966 + }, + { + "epoch": 0.21974451979183093, + "grad_norm": 2.5205092430114746, + "learning_rate": 1.817140187442482e-05, + "loss": 0.5281, + "step": 6967 + }, + { + "epoch": 0.21977606055827156, + "grad_norm": 3.10962176322937, + "learning_rate": 1.8170812966796506e-05, + "loss": 0.9171, + "step": 6968 + }, + { + "epoch": 0.2198076013247122, + "grad_norm": 4.1949663162231445, + "learning_rate": 1.8170223973899418e-05, + "loss": 0.574, + "step": 6969 + }, + { + "epoch": 0.21983914209115282, + "grad_norm": 3.4974398612976074, + "learning_rate": 1.8169634895739696e-05, + "loss": 0.5431, + "step": 6970 + }, + { + "epoch": 0.21987068285759345, + "grad_norm": 2.6622467041015625, + "learning_rate": 1.8169045732323495e-05, + "loss": 0.5834, + "step": 6971 + }, + { + "epoch": 0.21990222362403405, + "grad_norm": 2.180077075958252, + "learning_rate": 1.8168456483656955e-05, + "loss": 0.5402, + "step": 6972 + }, + { + "epoch": 0.21993376439047468, + "grad_norm": 2.1792166233062744, + "learning_rate": 1.816786714974623e-05, + "loss": 0.5681, + "step": 6973 + }, + { + "epoch": 0.21996530515691531, + "grad_norm": 1.8891538381576538, + "learning_rate": 1.816727773059747e-05, + "loss": 0.5636, + "step": 6974 + }, + { + "epoch": 0.21999684592335594, + "grad_norm": 1.9710547924041748, + "learning_rate": 1.8166688226216822e-05, + "loss": 0.584, + "step": 6975 + }, + { + "epoch": 0.22002838668979657, + "grad_norm": 3.761920928955078, + "learning_rate": 1.8166098636610444e-05, + "loss": 0.5795, + "step": 6976 + }, + { + "epoch": 0.22005992745623718, + "grad_norm": 2.1699819564819336, + "learning_rate": 1.8165508961784484e-05, + "loss": 0.3809, + "step": 6977 + }, + { + "epoch": 0.2200914682226778, + "grad_norm": 2.5881552696228027, + "learning_rate": 1.8164919201745096e-05, + "loss": 0.4345, + "step": 6978 + }, + { + "epoch": 0.22012300898911843, + "grad_norm": 6.619387626647949, + "learning_rate": 1.816432935649844e-05, + "loss": 0.412, + "step": 6979 + }, + { + "epoch": 0.22015454975555906, + "grad_norm": 7.769380569458008, + "learning_rate": 1.8163739426050668e-05, + "loss": 0.5577, + "step": 6980 + }, + { + "epoch": 0.2201860905219997, + "grad_norm": 2.1511409282684326, + "learning_rate": 1.8163149410407937e-05, + "loss": 0.5464, + "step": 6981 + }, + { + "epoch": 0.2202176312884403, + "grad_norm": 1.7707232236862183, + "learning_rate": 1.81625593095764e-05, + "loss": 0.9371, + "step": 6982 + }, + { + "epoch": 0.22024917205488093, + "grad_norm": 2.290881633758545, + "learning_rate": 1.816196912356222e-05, + "loss": 0.5768, + "step": 6983 + }, + { + "epoch": 0.22028071282132156, + "grad_norm": 2.2356951236724854, + "learning_rate": 1.816137885237155e-05, + "loss": 0.5659, + "step": 6984 + }, + { + "epoch": 0.22031225358776219, + "grad_norm": 3.4283130168914795, + "learning_rate": 1.8160788496010562e-05, + "loss": 0.5656, + "step": 6985 + }, + { + "epoch": 0.22034379435420282, + "grad_norm": 1.830977201461792, + "learning_rate": 1.816019805448541e-05, + "loss": 0.5656, + "step": 6986 + }, + { + "epoch": 0.22037533512064345, + "grad_norm": 2.4208033084869385, + "learning_rate": 1.8159607527802246e-05, + "loss": 0.5614, + "step": 6987 + }, + { + "epoch": 0.22040687588708405, + "grad_norm": 2.5171637535095215, + "learning_rate": 1.8159016915967247e-05, + "loss": 0.5847, + "step": 6988 + }, + { + "epoch": 0.22043841665352468, + "grad_norm": 6.171496391296387, + "learning_rate": 1.8158426218986572e-05, + "loss": 0.451, + "step": 6989 + }, + { + "epoch": 0.2204699574199653, + "grad_norm": 2.5808351039886475, + "learning_rate": 1.8157835436866384e-05, + "loss": 0.5604, + "step": 6990 + }, + { + "epoch": 0.22050149818640594, + "grad_norm": 3.265437602996826, + "learning_rate": 1.815724456961285e-05, + "loss": 0.5794, + "step": 6991 + }, + { + "epoch": 0.22053303895284657, + "grad_norm": 2.0522913932800293, + "learning_rate": 1.815665361723213e-05, + "loss": 0.8787, + "step": 6992 + }, + { + "epoch": 0.22056457971928717, + "grad_norm": 2.2218968868255615, + "learning_rate": 1.81560625797304e-05, + "loss": 0.902, + "step": 6993 + }, + { + "epoch": 0.2205961204857278, + "grad_norm": 2.6476082801818848, + "learning_rate": 1.8155471457113825e-05, + "loss": 0.4031, + "step": 6994 + }, + { + "epoch": 0.22062766125216843, + "grad_norm": 2.2259368896484375, + "learning_rate": 1.815488024938857e-05, + "loss": 0.5772, + "step": 6995 + }, + { + "epoch": 0.22065920201860906, + "grad_norm": 2.2741854190826416, + "learning_rate": 1.815428895656081e-05, + "loss": 0.4377, + "step": 6996 + }, + { + "epoch": 0.2206907427850497, + "grad_norm": 2.8409855365753174, + "learning_rate": 1.815369757863671e-05, + "loss": 0.5622, + "step": 6997 + }, + { + "epoch": 0.2207222835514903, + "grad_norm": 2.381453514099121, + "learning_rate": 1.8153106115622447e-05, + "loss": 0.5604, + "step": 6998 + }, + { + "epoch": 0.22075382431793092, + "grad_norm": 2.766216516494751, + "learning_rate": 1.8152514567524192e-05, + "loss": 0.5544, + "step": 6999 + }, + { + "epoch": 0.22078536508437155, + "grad_norm": 2.9577014446258545, + "learning_rate": 1.8151922934348114e-05, + "loss": 0.5786, + "step": 7000 + }, + { + "epoch": 0.22081690585081218, + "grad_norm": 2.9407026767730713, + "learning_rate": 1.815133121610039e-05, + "loss": 0.5695, + "step": 7001 + }, + { + "epoch": 0.2208484466172528, + "grad_norm": 2.0884382724761963, + "learning_rate": 1.81507394127872e-05, + "loss": 0.5346, + "step": 7002 + }, + { + "epoch": 0.2208799873836934, + "grad_norm": 3.61143159866333, + "learning_rate": 1.8150147524414713e-05, + "loss": 0.5729, + "step": 7003 + }, + { + "epoch": 0.22091152815013404, + "grad_norm": 2.720743417739868, + "learning_rate": 1.8149555550989108e-05, + "loss": 0.5362, + "step": 7004 + }, + { + "epoch": 0.22094306891657467, + "grad_norm": 2.864933729171753, + "learning_rate": 1.814896349251656e-05, + "loss": 0.5422, + "step": 7005 + }, + { + "epoch": 0.2209746096830153, + "grad_norm": 2.9981696605682373, + "learning_rate": 1.8148371349003254e-05, + "loss": 0.5409, + "step": 7006 + }, + { + "epoch": 0.22100615044945593, + "grad_norm": 2.650913953781128, + "learning_rate": 1.8147779120455365e-05, + "loss": 0.5501, + "step": 7007 + }, + { + "epoch": 0.22103769121589656, + "grad_norm": 2.7431046962738037, + "learning_rate": 1.8147186806879076e-05, + "loss": 0.5677, + "step": 7008 + }, + { + "epoch": 0.22106923198233716, + "grad_norm": 4.148238182067871, + "learning_rate": 1.8146594408280564e-05, + "loss": 0.9338, + "step": 7009 + }, + { + "epoch": 0.2211007727487778, + "grad_norm": 4.457897663116455, + "learning_rate": 1.8146001924666015e-05, + "loss": 0.5185, + "step": 7010 + }, + { + "epoch": 0.22113231351521842, + "grad_norm": 3.304265022277832, + "learning_rate": 1.814540935604161e-05, + "loss": 0.5491, + "step": 7011 + }, + { + "epoch": 0.22116385428165905, + "grad_norm": 3.6549644470214844, + "learning_rate": 1.8144816702413533e-05, + "loss": 0.5839, + "step": 7012 + }, + { + "epoch": 0.22119539504809968, + "grad_norm": 2.6239559650421143, + "learning_rate": 1.814422396378797e-05, + "loss": 0.5659, + "step": 7013 + }, + { + "epoch": 0.22122693581454028, + "grad_norm": 2.2585995197296143, + "learning_rate": 1.8143631140171107e-05, + "loss": 0.5799, + "step": 7014 + }, + { + "epoch": 0.2212584765809809, + "grad_norm": 2.2962417602539062, + "learning_rate": 1.814303823156913e-05, + "loss": 0.569, + "step": 7015 + }, + { + "epoch": 0.22129001734742154, + "grad_norm": 2.575317621231079, + "learning_rate": 1.8142445237988226e-05, + "loss": 0.533, + "step": 7016 + }, + { + "epoch": 0.22132155811386217, + "grad_norm": 3.583483934402466, + "learning_rate": 1.814185215943458e-05, + "loss": 0.57, + "step": 7017 + }, + { + "epoch": 0.2213530988803028, + "grad_norm": 6.263027667999268, + "learning_rate": 1.814125899591439e-05, + "loss": 0.5673, + "step": 7018 + }, + { + "epoch": 0.2213846396467434, + "grad_norm": 2.820742130279541, + "learning_rate": 1.8140665747433838e-05, + "loss": 0.4504, + "step": 7019 + }, + { + "epoch": 0.22141618041318403, + "grad_norm": 3.2687339782714844, + "learning_rate": 1.814007241399912e-05, + "loss": 0.5401, + "step": 7020 + }, + { + "epoch": 0.22144772117962466, + "grad_norm": 2.9906399250030518, + "learning_rate": 1.8139478995616422e-05, + "loss": 0.56, + "step": 7021 + }, + { + "epoch": 0.2214792619460653, + "grad_norm": 2.5002920627593994, + "learning_rate": 1.8138885492291943e-05, + "loss": 0.5474, + "step": 7022 + }, + { + "epoch": 0.22151080271250592, + "grad_norm": 2.5601134300231934, + "learning_rate": 1.8138291904031872e-05, + "loss": 0.5218, + "step": 7023 + }, + { + "epoch": 0.22154234347894655, + "grad_norm": 13.167017936706543, + "learning_rate": 1.813769823084241e-05, + "loss": 0.5471, + "step": 7024 + }, + { + "epoch": 0.22157388424538715, + "grad_norm": 3.0334582328796387, + "learning_rate": 1.8137104472729742e-05, + "loss": 0.574, + "step": 7025 + }, + { + "epoch": 0.22160542501182778, + "grad_norm": 3.118638038635254, + "learning_rate": 1.8136510629700077e-05, + "loss": 0.5617, + "step": 7026 + }, + { + "epoch": 0.2216369657782684, + "grad_norm": 3.7837677001953125, + "learning_rate": 1.8135916701759605e-05, + "loss": 0.9247, + "step": 7027 + }, + { + "epoch": 0.22166850654470904, + "grad_norm": 3.4002132415771484, + "learning_rate": 1.813532268891452e-05, + "loss": 0.5555, + "step": 7028 + }, + { + "epoch": 0.22170004731114967, + "grad_norm": 3.1969034671783447, + "learning_rate": 1.813472859117103e-05, + "loss": 0.5301, + "step": 7029 + }, + { + "epoch": 0.22173158807759027, + "grad_norm": 2.381080389022827, + "learning_rate": 1.813413440853533e-05, + "loss": 0.8884, + "step": 7030 + }, + { + "epoch": 0.2217631288440309, + "grad_norm": 7.1258673667907715, + "learning_rate": 1.813354014101362e-05, + "loss": 0.5887, + "step": 7031 + }, + { + "epoch": 0.22179466961047153, + "grad_norm": 2.870314121246338, + "learning_rate": 1.8132945788612103e-05, + "loss": 0.5648, + "step": 7032 + }, + { + "epoch": 0.22182621037691216, + "grad_norm": 2.8441762924194336, + "learning_rate": 1.8132351351336983e-05, + "loss": 0.5756, + "step": 7033 + }, + { + "epoch": 0.2218577511433528, + "grad_norm": 5.229165077209473, + "learning_rate": 1.813175682919446e-05, + "loss": 0.5588, + "step": 7034 + }, + { + "epoch": 0.2218892919097934, + "grad_norm": 3.2011494636535645, + "learning_rate": 1.8131162222190745e-05, + "loss": 0.5459, + "step": 7035 + }, + { + "epoch": 0.22192083267623403, + "grad_norm": 3.6994681358337402, + "learning_rate": 1.8130567530332036e-05, + "loss": 0.5887, + "step": 7036 + }, + { + "epoch": 0.22195237344267466, + "grad_norm": 3.031538248062134, + "learning_rate": 1.812997275362454e-05, + "loss": 0.5552, + "step": 7037 + }, + { + "epoch": 0.22198391420911529, + "grad_norm": 2.130115032196045, + "learning_rate": 1.8129377892074463e-05, + "loss": 0.5419, + "step": 7038 + }, + { + "epoch": 0.22201545497555591, + "grad_norm": 3.2427594661712646, + "learning_rate": 1.8128782945688022e-05, + "loss": 0.5766, + "step": 7039 + }, + { + "epoch": 0.22204699574199652, + "grad_norm": 2.921034574508667, + "learning_rate": 1.8128187914471415e-05, + "loss": 0.5734, + "step": 7040 + }, + { + "epoch": 0.22207853650843715, + "grad_norm": 3.7750706672668457, + "learning_rate": 1.8127592798430858e-05, + "loss": 0.5203, + "step": 7041 + }, + { + "epoch": 0.22211007727487778, + "grad_norm": 2.1956732273101807, + "learning_rate": 1.812699759757256e-05, + "loss": 0.5569, + "step": 7042 + }, + { + "epoch": 0.2221416180413184, + "grad_norm": 5.010038375854492, + "learning_rate": 1.8126402311902727e-05, + "loss": 0.9867, + "step": 7043 + }, + { + "epoch": 0.22217315880775904, + "grad_norm": 2.612067699432373, + "learning_rate": 1.8125806941427578e-05, + "loss": 0.5696, + "step": 7044 + }, + { + "epoch": 0.22220469957419967, + "grad_norm": 3.1980133056640625, + "learning_rate": 1.8125211486153324e-05, + "loss": 0.6067, + "step": 7045 + }, + { + "epoch": 0.22223624034064027, + "grad_norm": 2.304736614227295, + "learning_rate": 1.812461594608618e-05, + "loss": 0.521, + "step": 7046 + }, + { + "epoch": 0.2222677811070809, + "grad_norm": 2.741262912750244, + "learning_rate": 1.8124020321232362e-05, + "loss": 0.4656, + "step": 7047 + }, + { + "epoch": 0.22229932187352153, + "grad_norm": 3.6503052711486816, + "learning_rate": 1.812342461159808e-05, + "loss": 0.564, + "step": 7048 + }, + { + "epoch": 0.22233086263996216, + "grad_norm": 2.098836898803711, + "learning_rate": 1.8122828817189556e-05, + "loss": 0.414, + "step": 7049 + }, + { + "epoch": 0.2223624034064028, + "grad_norm": 5.493417263031006, + "learning_rate": 1.812223293801301e-05, + "loss": 0.5633, + "step": 7050 + }, + { + "epoch": 0.2223939441728434, + "grad_norm": 3.274336338043213, + "learning_rate": 1.8121636974074646e-05, + "loss": 0.5562, + "step": 7051 + }, + { + "epoch": 0.22242548493928402, + "grad_norm": 3.3754262924194336, + "learning_rate": 1.8121040925380704e-05, + "loss": 0.5931, + "step": 7052 + }, + { + "epoch": 0.22245702570572465, + "grad_norm": 2.6750993728637695, + "learning_rate": 1.812044479193739e-05, + "loss": 0.3732, + "step": 7053 + }, + { + "epoch": 0.22248856647216528, + "grad_norm": 2.4629716873168945, + "learning_rate": 1.811984857375093e-05, + "loss": 0.5646, + "step": 7054 + }, + { + "epoch": 0.2225201072386059, + "grad_norm": 1.6507506370544434, + "learning_rate": 1.8119252270827546e-05, + "loss": 0.3951, + "step": 7055 + }, + { + "epoch": 0.2225516480050465, + "grad_norm": 3.0907764434814453, + "learning_rate": 1.8118655883173458e-05, + "loss": 0.5681, + "step": 7056 + }, + { + "epoch": 0.22258318877148714, + "grad_norm": 11.125265121459961, + "learning_rate": 1.8118059410794892e-05, + "loss": 0.5761, + "step": 7057 + }, + { + "epoch": 0.22261472953792777, + "grad_norm": 3.4140431880950928, + "learning_rate": 1.8117462853698073e-05, + "loss": 0.5548, + "step": 7058 + }, + { + "epoch": 0.2226462703043684, + "grad_norm": 2.5325961112976074, + "learning_rate": 1.811686621188923e-05, + "loss": 0.5614, + "step": 7059 + }, + { + "epoch": 0.22267781107080903, + "grad_norm": 5.391631603240967, + "learning_rate": 1.811626948537458e-05, + "loss": 0.5473, + "step": 7060 + }, + { + "epoch": 0.22270935183724966, + "grad_norm": 3.5645384788513184, + "learning_rate": 1.8115672674160357e-05, + "loss": 0.564, + "step": 7061 + }, + { + "epoch": 0.22274089260369026, + "grad_norm": 3.14737606048584, + "learning_rate": 1.811507577825279e-05, + "loss": 0.561, + "step": 7062 + }, + { + "epoch": 0.2227724333701309, + "grad_norm": 3.176445484161377, + "learning_rate": 1.8114478797658106e-05, + "loss": 0.4905, + "step": 7063 + }, + { + "epoch": 0.22280397413657152, + "grad_norm": 3.1003565788269043, + "learning_rate": 1.8113881732382532e-05, + "loss": 0.5549, + "step": 7064 + }, + { + "epoch": 0.22283551490301215, + "grad_norm": 3.1951379776000977, + "learning_rate": 1.8113284582432307e-05, + "loss": 0.5581, + "step": 7065 + }, + { + "epoch": 0.22286705566945278, + "grad_norm": 11.00485610961914, + "learning_rate": 1.8112687347813654e-05, + "loss": 0.4329, + "step": 7066 + }, + { + "epoch": 0.22289859643589338, + "grad_norm": 4.589970111846924, + "learning_rate": 1.8112090028532808e-05, + "loss": 0.5846, + "step": 7067 + }, + { + "epoch": 0.222930137202334, + "grad_norm": 3.202220916748047, + "learning_rate": 1.8111492624596006e-05, + "loss": 0.5337, + "step": 7068 + }, + { + "epoch": 0.22296167796877464, + "grad_norm": 3.0122978687286377, + "learning_rate": 1.8110895136009482e-05, + "loss": 0.4353, + "step": 7069 + }, + { + "epoch": 0.22299321873521527, + "grad_norm": 3.284325122833252, + "learning_rate": 1.8110297562779465e-05, + "loss": 0.5838, + "step": 7070 + }, + { + "epoch": 0.2230247595016559, + "grad_norm": 3.56113600730896, + "learning_rate": 1.8109699904912194e-05, + "loss": 0.4552, + "step": 7071 + }, + { + "epoch": 0.2230563002680965, + "grad_norm": 2.4620699882507324, + "learning_rate": 1.8109102162413914e-05, + "loss": 0.5859, + "step": 7072 + }, + { + "epoch": 0.22308784103453713, + "grad_norm": 7.242951393127441, + "learning_rate": 1.8108504335290852e-05, + "loss": 0.6079, + "step": 7073 + }, + { + "epoch": 0.22311938180097776, + "grad_norm": 3.754560947418213, + "learning_rate": 1.8107906423549253e-05, + "loss": 0.5493, + "step": 7074 + }, + { + "epoch": 0.2231509225674184, + "grad_norm": 3.8782835006713867, + "learning_rate": 1.8107308427195355e-05, + "loss": 0.5809, + "step": 7075 + }, + { + "epoch": 0.22318246333385902, + "grad_norm": 2.6151442527770996, + "learning_rate": 1.8106710346235395e-05, + "loss": 0.5579, + "step": 7076 + }, + { + "epoch": 0.22321400410029965, + "grad_norm": 2.884159803390503, + "learning_rate": 1.810611218067562e-05, + "loss": 0.5522, + "step": 7077 + }, + { + "epoch": 0.22324554486674025, + "grad_norm": 2.3308775424957275, + "learning_rate": 1.810551393052227e-05, + "loss": 0.5892, + "step": 7078 + }, + { + "epoch": 0.22327708563318088, + "grad_norm": 4.132657527923584, + "learning_rate": 1.810491559578159e-05, + "loss": 0.5683, + "step": 7079 + }, + { + "epoch": 0.2233086263996215, + "grad_norm": 2.899259567260742, + "learning_rate": 1.810431717645982e-05, + "loss": 0.5113, + "step": 7080 + }, + { + "epoch": 0.22334016716606214, + "grad_norm": 4.781321048736572, + "learning_rate": 1.810371867256321e-05, + "loss": 0.5836, + "step": 7081 + }, + { + "epoch": 0.22337170793250277, + "grad_norm": 3.9161951541900635, + "learning_rate": 1.8103120084098003e-05, + "loss": 0.5717, + "step": 7082 + }, + { + "epoch": 0.22340324869894337, + "grad_norm": 3.2854695320129395, + "learning_rate": 1.8102521411070445e-05, + "loss": 0.5514, + "step": 7083 + }, + { + "epoch": 0.223434789465384, + "grad_norm": 4.497298240661621, + "learning_rate": 1.8101922653486787e-05, + "loss": 0.5294, + "step": 7084 + }, + { + "epoch": 0.22346633023182463, + "grad_norm": 2.8005385398864746, + "learning_rate": 1.810132381135327e-05, + "loss": 0.9645, + "step": 7085 + }, + { + "epoch": 0.22349787099826526, + "grad_norm": 5.158459663391113, + "learning_rate": 1.8100724884676152e-05, + "loss": 0.5637, + "step": 7086 + }, + { + "epoch": 0.2235294117647059, + "grad_norm": 3.2676072120666504, + "learning_rate": 1.8100125873461678e-05, + "loss": 0.5773, + "step": 7087 + }, + { + "epoch": 0.2235609525311465, + "grad_norm": 2.7327213287353516, + "learning_rate": 1.80995267777161e-05, + "loss": 0.893, + "step": 7088 + }, + { + "epoch": 0.22359249329758712, + "grad_norm": 2.994246006011963, + "learning_rate": 1.8098927597445675e-05, + "loss": 0.5731, + "step": 7089 + }, + { + "epoch": 0.22362403406402775, + "grad_norm": 4.073358058929443, + "learning_rate": 1.809832833265665e-05, + "loss": 0.5173, + "step": 7090 + }, + { + "epoch": 0.22365557483046838, + "grad_norm": 6.586395740509033, + "learning_rate": 1.809772898335528e-05, + "loss": 0.5825, + "step": 7091 + }, + { + "epoch": 0.22368711559690901, + "grad_norm": 3.614251136779785, + "learning_rate": 1.809712954954782e-05, + "loss": 0.5626, + "step": 7092 + }, + { + "epoch": 0.22371865636334962, + "grad_norm": 3.024585008621216, + "learning_rate": 1.8096530031240527e-05, + "loss": 0.543, + "step": 7093 + }, + { + "epoch": 0.22375019712979025, + "grad_norm": 3.4032342433929443, + "learning_rate": 1.8095930428439656e-05, + "loss": 0.6081, + "step": 7094 + }, + { + "epoch": 0.22378173789623088, + "grad_norm": 2.9109761714935303, + "learning_rate": 1.8095330741151465e-05, + "loss": 0.5524, + "step": 7095 + }, + { + "epoch": 0.2238132786626715, + "grad_norm": 10.081742286682129, + "learning_rate": 1.809473096938221e-05, + "loss": 0.5856, + "step": 7096 + }, + { + "epoch": 0.22384481942911214, + "grad_norm": 3.1622753143310547, + "learning_rate": 1.8094131113138155e-05, + "loss": 0.5568, + "step": 7097 + }, + { + "epoch": 0.22387636019555276, + "grad_norm": 24.66636085510254, + "learning_rate": 1.8093531172425557e-05, + "loss": 0.4543, + "step": 7098 + }, + { + "epoch": 0.22390790096199337, + "grad_norm": 3.0019843578338623, + "learning_rate": 1.8092931147250673e-05, + "loss": 0.5578, + "step": 7099 + }, + { + "epoch": 0.223939441728434, + "grad_norm": 3.040146589279175, + "learning_rate": 1.809233103761977e-05, + "loss": 0.5739, + "step": 7100 + }, + { + "epoch": 0.22397098249487463, + "grad_norm": 3.2526357173919678, + "learning_rate": 1.809173084353911e-05, + "loss": 0.5703, + "step": 7101 + }, + { + "epoch": 0.22400252326131526, + "grad_norm": 7.457939624786377, + "learning_rate": 1.8091130565014956e-05, + "loss": 0.6126, + "step": 7102 + }, + { + "epoch": 0.22403406402775589, + "grad_norm": 5.02000617980957, + "learning_rate": 1.809053020205357e-05, + "loss": 0.9398, + "step": 7103 + }, + { + "epoch": 0.2240656047941965, + "grad_norm": 2.790511131286621, + "learning_rate": 1.8089929754661223e-05, + "loss": 0.8797, + "step": 7104 + }, + { + "epoch": 0.22409714556063712, + "grad_norm": 6.147787570953369, + "learning_rate": 1.8089329222844173e-05, + "loss": 0.5742, + "step": 7105 + }, + { + "epoch": 0.22412868632707775, + "grad_norm": 9.6300687789917, + "learning_rate": 1.8088728606608695e-05, + "loss": 0.5953, + "step": 7106 + }, + { + "epoch": 0.22416022709351838, + "grad_norm": 6.180251598358154, + "learning_rate": 1.8088127905961047e-05, + "loss": 0.9199, + "step": 7107 + }, + { + "epoch": 0.224191767859959, + "grad_norm": 3.041832685470581, + "learning_rate": 1.8087527120907512e-05, + "loss": 0.5793, + "step": 7108 + }, + { + "epoch": 0.2242233086263996, + "grad_norm": 3.2656593322753906, + "learning_rate": 1.8086926251454347e-05, + "loss": 0.5519, + "step": 7109 + }, + { + "epoch": 0.22425484939284024, + "grad_norm": 2.495262861251831, + "learning_rate": 1.8086325297607826e-05, + "loss": 0.5794, + "step": 7110 + }, + { + "epoch": 0.22428639015928087, + "grad_norm": 2.6106789112091064, + "learning_rate": 1.8085724259374225e-05, + "loss": 0.5733, + "step": 7111 + }, + { + "epoch": 0.2243179309257215, + "grad_norm": 3.215632438659668, + "learning_rate": 1.8085123136759812e-05, + "loss": 0.9351, + "step": 7112 + }, + { + "epoch": 0.22434947169216213, + "grad_norm": 2.8665385246276855, + "learning_rate": 1.808452192977086e-05, + "loss": 0.8997, + "step": 7113 + }, + { + "epoch": 0.22438101245860276, + "grad_norm": 2.6132426261901855, + "learning_rate": 1.8083920638413648e-05, + "loss": 0.5856, + "step": 7114 + }, + { + "epoch": 0.22441255322504336, + "grad_norm": 3.207975149154663, + "learning_rate": 1.8083319262694443e-05, + "loss": 0.5615, + "step": 7115 + }, + { + "epoch": 0.224444093991484, + "grad_norm": 2.5608649253845215, + "learning_rate": 1.8082717802619525e-05, + "loss": 0.5577, + "step": 7116 + }, + { + "epoch": 0.22447563475792462, + "grad_norm": 2.097315549850464, + "learning_rate": 1.8082116258195173e-05, + "loss": 0.5547, + "step": 7117 + }, + { + "epoch": 0.22450717552436525, + "grad_norm": 2.5990190505981445, + "learning_rate": 1.808151462942766e-05, + "loss": 0.5605, + "step": 7118 + }, + { + "epoch": 0.22453871629080588, + "grad_norm": 2.217923641204834, + "learning_rate": 1.808091291632327e-05, + "loss": 0.5502, + "step": 7119 + }, + { + "epoch": 0.22457025705724648, + "grad_norm": 2.4985783100128174, + "learning_rate": 1.8080311118888278e-05, + "loss": 0.5599, + "step": 7120 + }, + { + "epoch": 0.2246017978236871, + "grad_norm": 3.031991481781006, + "learning_rate": 1.8079709237128964e-05, + "loss": 0.5696, + "step": 7121 + }, + { + "epoch": 0.22463333859012774, + "grad_norm": 2.8241777420043945, + "learning_rate": 1.807910727105161e-05, + "loss": 0.5439, + "step": 7122 + }, + { + "epoch": 0.22466487935656837, + "grad_norm": 2.5590081214904785, + "learning_rate": 1.8078505220662498e-05, + "loss": 0.5481, + "step": 7123 + }, + { + "epoch": 0.224696420123009, + "grad_norm": 1.800658941268921, + "learning_rate": 1.807790308596791e-05, + "loss": 0.5367, + "step": 7124 + }, + { + "epoch": 0.2247279608894496, + "grad_norm": 2.226078987121582, + "learning_rate": 1.8077300866974136e-05, + "loss": 0.5702, + "step": 7125 + }, + { + "epoch": 0.22475950165589023, + "grad_norm": 2.803926467895508, + "learning_rate": 1.8076698563687454e-05, + "loss": 0.5504, + "step": 7126 + }, + { + "epoch": 0.22479104242233086, + "grad_norm": 9.304357528686523, + "learning_rate": 1.807609617611415e-05, + "loss": 0.5325, + "step": 7127 + }, + { + "epoch": 0.2248225831887715, + "grad_norm": 2.281403064727783, + "learning_rate": 1.807549370426051e-05, + "loss": 0.5393, + "step": 7128 + }, + { + "epoch": 0.22485412395521212, + "grad_norm": 3.712322950363159, + "learning_rate": 1.8074891148132826e-05, + "loss": 0.9153, + "step": 7129 + }, + { + "epoch": 0.22488566472165272, + "grad_norm": 2.6492321491241455, + "learning_rate": 1.807428850773738e-05, + "loss": 0.5582, + "step": 7130 + }, + { + "epoch": 0.22491720548809335, + "grad_norm": 2.0739285945892334, + "learning_rate": 1.8073685783080464e-05, + "loss": 0.5533, + "step": 7131 + }, + { + "epoch": 0.22494874625453398, + "grad_norm": 1.6040732860565186, + "learning_rate": 1.807308297416837e-05, + "loss": 0.4408, + "step": 7132 + }, + { + "epoch": 0.2249802870209746, + "grad_norm": 2.2843666076660156, + "learning_rate": 1.8072480081007384e-05, + "loss": 0.912, + "step": 7133 + }, + { + "epoch": 0.22501182778741524, + "grad_norm": 4.047179222106934, + "learning_rate": 1.8071877103603804e-05, + "loss": 0.5886, + "step": 7134 + }, + { + "epoch": 0.22504336855385587, + "grad_norm": 3.308736562728882, + "learning_rate": 1.8071274041963916e-05, + "loss": 0.5438, + "step": 7135 + }, + { + "epoch": 0.22507490932029647, + "grad_norm": 2.6515886783599854, + "learning_rate": 1.8070670896094015e-05, + "loss": 0.5553, + "step": 7136 + }, + { + "epoch": 0.2251064500867371, + "grad_norm": 2.177349090576172, + "learning_rate": 1.80700676660004e-05, + "loss": 0.5537, + "step": 7137 + }, + { + "epoch": 0.22513799085317773, + "grad_norm": 1.5266591310501099, + "learning_rate": 1.8069464351689356e-05, + "loss": 0.4272, + "step": 7138 + }, + { + "epoch": 0.22516953161961836, + "grad_norm": 3.0100505352020264, + "learning_rate": 1.806886095316719e-05, + "loss": 0.5715, + "step": 7139 + }, + { + "epoch": 0.225201072386059, + "grad_norm": 3.6694493293762207, + "learning_rate": 1.8068257470440195e-05, + "loss": 0.6248, + "step": 7140 + }, + { + "epoch": 0.2252326131524996, + "grad_norm": 1.934720516204834, + "learning_rate": 1.8067653903514674e-05, + "loss": 0.4149, + "step": 7141 + }, + { + "epoch": 0.22526415391894022, + "grad_norm": 3.190886974334717, + "learning_rate": 1.806705025239691e-05, + "loss": 0.5839, + "step": 7142 + }, + { + "epoch": 0.22529569468538085, + "grad_norm": 10.357500076293945, + "learning_rate": 1.8066446517093217e-05, + "loss": 0.5665, + "step": 7143 + }, + { + "epoch": 0.22532723545182148, + "grad_norm": 3.09527587890625, + "learning_rate": 1.8065842697609892e-05, + "loss": 0.5933, + "step": 7144 + }, + { + "epoch": 0.2253587762182621, + "grad_norm": 2.2985100746154785, + "learning_rate": 1.8065238793953236e-05, + "loss": 0.5547, + "step": 7145 + }, + { + "epoch": 0.22539031698470272, + "grad_norm": 2.5676190853118896, + "learning_rate": 1.806463480612955e-05, + "loss": 0.5723, + "step": 7146 + }, + { + "epoch": 0.22542185775114335, + "grad_norm": 2.0705881118774414, + "learning_rate": 1.8064030734145137e-05, + "loss": 0.5659, + "step": 7147 + }, + { + "epoch": 0.22545339851758397, + "grad_norm": 2.963630199432373, + "learning_rate": 1.8063426578006303e-05, + "loss": 0.5822, + "step": 7148 + }, + { + "epoch": 0.2254849392840246, + "grad_norm": 2.292179822921753, + "learning_rate": 1.806282233771935e-05, + "loss": 0.576, + "step": 7149 + }, + { + "epoch": 0.22551648005046523, + "grad_norm": 3.3720703125, + "learning_rate": 1.8062218013290588e-05, + "loss": 0.5797, + "step": 7150 + }, + { + "epoch": 0.22554802081690586, + "grad_norm": 2.343465566635132, + "learning_rate": 1.8061613604726317e-05, + "loss": 0.5819, + "step": 7151 + }, + { + "epoch": 0.22557956158334647, + "grad_norm": 2.6501402854919434, + "learning_rate": 1.8061009112032847e-05, + "loss": 0.5934, + "step": 7152 + }, + { + "epoch": 0.2256111023497871, + "grad_norm": 1.7203969955444336, + "learning_rate": 1.8060404535216492e-05, + "loss": 0.569, + "step": 7153 + }, + { + "epoch": 0.22564264311622773, + "grad_norm": 1.8080071210861206, + "learning_rate": 1.805979987428356e-05, + "loss": 0.5523, + "step": 7154 + }, + { + "epoch": 0.22567418388266836, + "grad_norm": 2.8603463172912598, + "learning_rate": 1.8059195129240353e-05, + "loss": 0.5913, + "step": 7155 + }, + { + "epoch": 0.22570572464910899, + "grad_norm": 2.741952419281006, + "learning_rate": 1.8058590300093185e-05, + "loss": 0.5413, + "step": 7156 + }, + { + "epoch": 0.2257372654155496, + "grad_norm": 1.6597719192504883, + "learning_rate": 1.8057985386848375e-05, + "loss": 0.5188, + "step": 7157 + }, + { + "epoch": 0.22576880618199022, + "grad_norm": 2.1097984313964844, + "learning_rate": 1.8057380389512227e-05, + "loss": 0.5864, + "step": 7158 + }, + { + "epoch": 0.22580034694843085, + "grad_norm": 2.0329177379608154, + "learning_rate": 1.8056775308091058e-05, + "loss": 0.5696, + "step": 7159 + }, + { + "epoch": 0.22583188771487148, + "grad_norm": 2.360025644302368, + "learning_rate": 1.8056170142591186e-05, + "loss": 0.5844, + "step": 7160 + }, + { + "epoch": 0.2258634284813121, + "grad_norm": 2.5857088565826416, + "learning_rate": 1.8055564893018924e-05, + "loss": 0.5943, + "step": 7161 + }, + { + "epoch": 0.2258949692477527, + "grad_norm": 2.795111656188965, + "learning_rate": 1.8054959559380582e-05, + "loss": 0.5416, + "step": 7162 + }, + { + "epoch": 0.22592651001419334, + "grad_norm": 3.530346393585205, + "learning_rate": 1.8054354141682485e-05, + "loss": 0.5555, + "step": 7163 + }, + { + "epoch": 0.22595805078063397, + "grad_norm": 1.8197673559188843, + "learning_rate": 1.805374863993095e-05, + "loss": 0.5623, + "step": 7164 + }, + { + "epoch": 0.2259895915470746, + "grad_norm": 1.8910143375396729, + "learning_rate": 1.8053143054132293e-05, + "loss": 0.5623, + "step": 7165 + }, + { + "epoch": 0.22602113231351523, + "grad_norm": 2.004382848739624, + "learning_rate": 1.8052537384292836e-05, + "loss": 0.5448, + "step": 7166 + }, + { + "epoch": 0.22605267307995583, + "grad_norm": 2.12833833694458, + "learning_rate": 1.8051931630418898e-05, + "loss": 0.5749, + "step": 7167 + }, + { + "epoch": 0.22608421384639646, + "grad_norm": 2.184751033782959, + "learning_rate": 1.8051325792516802e-05, + "loss": 0.5312, + "step": 7168 + }, + { + "epoch": 0.2261157546128371, + "grad_norm": 1.8455746173858643, + "learning_rate": 1.805071987059287e-05, + "loss": 0.579, + "step": 7169 + }, + { + "epoch": 0.22614729537927772, + "grad_norm": 4.119250774383545, + "learning_rate": 1.805011386465342e-05, + "loss": 0.9782, + "step": 7170 + }, + { + "epoch": 0.22617883614571835, + "grad_norm": 1.7310761213302612, + "learning_rate": 1.8049507774704787e-05, + "loss": 0.5516, + "step": 7171 + }, + { + "epoch": 0.22621037691215898, + "grad_norm": 1.9954038858413696, + "learning_rate": 1.804890160075329e-05, + "loss": 0.5567, + "step": 7172 + }, + { + "epoch": 0.22624191767859958, + "grad_norm": 1.6690415143966675, + "learning_rate": 1.804829534280525e-05, + "loss": 0.5609, + "step": 7173 + }, + { + "epoch": 0.2262734584450402, + "grad_norm": 2.2910683155059814, + "learning_rate": 1.8047689000867003e-05, + "loss": 0.5732, + "step": 7174 + }, + { + "epoch": 0.22630499921148084, + "grad_norm": 4.450465202331543, + "learning_rate": 1.804708257494487e-05, + "loss": 0.5856, + "step": 7175 + }, + { + "epoch": 0.22633653997792147, + "grad_norm": 2.057849407196045, + "learning_rate": 1.8046476065045183e-05, + "loss": 0.5395, + "step": 7176 + }, + { + "epoch": 0.2263680807443621, + "grad_norm": 2.3516364097595215, + "learning_rate": 1.8045869471174272e-05, + "loss": 0.5451, + "step": 7177 + }, + { + "epoch": 0.2263996215108027, + "grad_norm": 2.154726266860962, + "learning_rate": 1.8045262793338466e-05, + "loss": 0.5646, + "step": 7178 + }, + { + "epoch": 0.22643116227724333, + "grad_norm": 2.0033366680145264, + "learning_rate": 1.8044656031544093e-05, + "loss": 0.5687, + "step": 7179 + }, + { + "epoch": 0.22646270304368396, + "grad_norm": 1.9998482465744019, + "learning_rate": 1.8044049185797486e-05, + "loss": 0.5557, + "step": 7180 + }, + { + "epoch": 0.2264942438101246, + "grad_norm": 1.945158839225769, + "learning_rate": 1.8043442256104985e-05, + "loss": 0.3943, + "step": 7181 + }, + { + "epoch": 0.22652578457656522, + "grad_norm": 2.3692376613616943, + "learning_rate": 1.8042835242472914e-05, + "loss": 0.5299, + "step": 7182 + }, + { + "epoch": 0.22655732534300582, + "grad_norm": 2.244891881942749, + "learning_rate": 1.8042228144907613e-05, + "loss": 0.5444, + "step": 7183 + }, + { + "epoch": 0.22658886610944645, + "grad_norm": 1.8019354343414307, + "learning_rate": 1.8041620963415418e-05, + "loss": 0.5536, + "step": 7184 + }, + { + "epoch": 0.22662040687588708, + "grad_norm": 2.0098018646240234, + "learning_rate": 1.8041013698002665e-05, + "loss": 0.579, + "step": 7185 + }, + { + "epoch": 0.2266519476423277, + "grad_norm": 2.1495230197906494, + "learning_rate": 1.804040634867569e-05, + "loss": 0.5612, + "step": 7186 + }, + { + "epoch": 0.22668348840876834, + "grad_norm": 1.7719730138778687, + "learning_rate": 1.803979891544083e-05, + "loss": 0.4123, + "step": 7187 + }, + { + "epoch": 0.22671502917520897, + "grad_norm": 2.1856493949890137, + "learning_rate": 1.8039191398304427e-05, + "loss": 0.5909, + "step": 7188 + }, + { + "epoch": 0.22674656994164957, + "grad_norm": 3.5395145416259766, + "learning_rate": 1.8038583797272818e-05, + "loss": 0.5547, + "step": 7189 + }, + { + "epoch": 0.2267781107080902, + "grad_norm": 1.9237722158432007, + "learning_rate": 1.8037976112352344e-05, + "loss": 0.5549, + "step": 7190 + }, + { + "epoch": 0.22680965147453083, + "grad_norm": 1.951852798461914, + "learning_rate": 1.8037368343549354e-05, + "loss": 0.556, + "step": 7191 + }, + { + "epoch": 0.22684119224097146, + "grad_norm": 2.341661214828491, + "learning_rate": 1.803676049087018e-05, + "loss": 0.5736, + "step": 7192 + }, + { + "epoch": 0.2268727330074121, + "grad_norm": 2.087937593460083, + "learning_rate": 1.8036152554321167e-05, + "loss": 0.5518, + "step": 7193 + }, + { + "epoch": 0.2269042737738527, + "grad_norm": 4.321962833404541, + "learning_rate": 1.8035544533908666e-05, + "loss": 0.5647, + "step": 7194 + }, + { + "epoch": 0.22693581454029332, + "grad_norm": 1.7903791666030884, + "learning_rate": 1.803493642963902e-05, + "loss": 0.577, + "step": 7195 + }, + { + "epoch": 0.22696735530673395, + "grad_norm": 1.8305003643035889, + "learning_rate": 1.803432824151857e-05, + "loss": 0.5714, + "step": 7196 + }, + { + "epoch": 0.22699889607317458, + "grad_norm": 2.7747888565063477, + "learning_rate": 1.8033719969553666e-05, + "loss": 0.5794, + "step": 7197 + }, + { + "epoch": 0.2270304368396152, + "grad_norm": 2.182438850402832, + "learning_rate": 1.8033111613750657e-05, + "loss": 0.5337, + "step": 7198 + }, + { + "epoch": 0.22706197760605581, + "grad_norm": 2.328131914138794, + "learning_rate": 1.8032503174115895e-05, + "loss": 0.5686, + "step": 7199 + }, + { + "epoch": 0.22709351837249644, + "grad_norm": 2.8012335300445557, + "learning_rate": 1.803189465065572e-05, + "loss": 0.5498, + "step": 7200 + }, + { + "epoch": 0.22712505913893707, + "grad_norm": 3.7239372730255127, + "learning_rate": 1.803128604337649e-05, + "loss": 0.9333, + "step": 7201 + }, + { + "epoch": 0.2271565999053777, + "grad_norm": 2.0027804374694824, + "learning_rate": 1.8030677352284554e-05, + "loss": 0.595, + "step": 7202 + }, + { + "epoch": 0.22718814067181833, + "grad_norm": 8.507844924926758, + "learning_rate": 1.8030068577386264e-05, + "loss": 0.5483, + "step": 7203 + }, + { + "epoch": 0.22721968143825894, + "grad_norm": 2.0742483139038086, + "learning_rate": 1.8029459718687973e-05, + "loss": 0.5597, + "step": 7204 + }, + { + "epoch": 0.22725122220469957, + "grad_norm": 2.514953136444092, + "learning_rate": 1.8028850776196035e-05, + "loss": 0.5908, + "step": 7205 + }, + { + "epoch": 0.2272827629711402, + "grad_norm": 7.699317932128906, + "learning_rate": 1.8028241749916805e-05, + "loss": 0.3951, + "step": 7206 + }, + { + "epoch": 0.22731430373758083, + "grad_norm": 2.117806911468506, + "learning_rate": 1.8027632639856638e-05, + "loss": 0.566, + "step": 7207 + }, + { + "epoch": 0.22734584450402145, + "grad_norm": 2.390096426010132, + "learning_rate": 1.8027023446021893e-05, + "loss": 0.5433, + "step": 7208 + }, + { + "epoch": 0.22737738527046208, + "grad_norm": 1.461653470993042, + "learning_rate": 1.802641416841892e-05, + "loss": 0.8975, + "step": 7209 + }, + { + "epoch": 0.2274089260369027, + "grad_norm": 2.244276285171509, + "learning_rate": 1.8025804807054087e-05, + "loss": 0.5857, + "step": 7210 + }, + { + "epoch": 0.22744046680334332, + "grad_norm": 1.7935162782669067, + "learning_rate": 1.802519536193375e-05, + "loss": 0.5602, + "step": 7211 + }, + { + "epoch": 0.22747200756978395, + "grad_norm": 4.58599328994751, + "learning_rate": 1.8024585833064265e-05, + "loss": 0.5438, + "step": 7212 + }, + { + "epoch": 0.22750354833622458, + "grad_norm": 2.0985167026519775, + "learning_rate": 1.8023976220452e-05, + "loss": 0.5842, + "step": 7213 + }, + { + "epoch": 0.2275350891026652, + "grad_norm": 1.6083357334136963, + "learning_rate": 1.802336652410331e-05, + "loss": 0.5497, + "step": 7214 + }, + { + "epoch": 0.2275666298691058, + "grad_norm": 1.9778425693511963, + "learning_rate": 1.802275674402456e-05, + "loss": 0.5397, + "step": 7215 + }, + { + "epoch": 0.22759817063554644, + "grad_norm": 2.749390125274658, + "learning_rate": 1.8022146880222114e-05, + "loss": 0.5456, + "step": 7216 + }, + { + "epoch": 0.22762971140198707, + "grad_norm": 1.8434338569641113, + "learning_rate": 1.802153693270234e-05, + "loss": 0.5796, + "step": 7217 + }, + { + "epoch": 0.2276612521684277, + "grad_norm": 2.0792903900146484, + "learning_rate": 1.8020926901471596e-05, + "loss": 0.5892, + "step": 7218 + }, + { + "epoch": 0.22769279293486833, + "grad_norm": 1.9238494634628296, + "learning_rate": 1.802031678653625e-05, + "loss": 0.5584, + "step": 7219 + }, + { + "epoch": 0.22772433370130893, + "grad_norm": 2.554710865020752, + "learning_rate": 1.801970658790267e-05, + "loss": 0.5356, + "step": 7220 + }, + { + "epoch": 0.22775587446774956, + "grad_norm": 1.803959846496582, + "learning_rate": 1.8019096305577233e-05, + "loss": 0.5701, + "step": 7221 + }, + { + "epoch": 0.2277874152341902, + "grad_norm": 1.5709881782531738, + "learning_rate": 1.8018485939566292e-05, + "loss": 0.5795, + "step": 7222 + }, + { + "epoch": 0.22781895600063082, + "grad_norm": 1.6558758020401, + "learning_rate": 1.8017875489876227e-05, + "loss": 0.5309, + "step": 7223 + }, + { + "epoch": 0.22785049676707145, + "grad_norm": 2.654341459274292, + "learning_rate": 1.8017264956513402e-05, + "loss": 0.5419, + "step": 7224 + }, + { + "epoch": 0.22788203753351208, + "grad_norm": 2.7159252166748047, + "learning_rate": 1.8016654339484196e-05, + "loss": 0.581, + "step": 7225 + }, + { + "epoch": 0.22791357829995268, + "grad_norm": 3.0579946041107178, + "learning_rate": 1.8016043638794975e-05, + "loss": 0.565, + "step": 7226 + }, + { + "epoch": 0.2279451190663933, + "grad_norm": 1.5536967515945435, + "learning_rate": 1.8015432854452113e-05, + "loss": 0.5592, + "step": 7227 + }, + { + "epoch": 0.22797665983283394, + "grad_norm": 3.170227289199829, + "learning_rate": 1.8014821986461987e-05, + "loss": 0.5914, + "step": 7228 + }, + { + "epoch": 0.22800820059927457, + "grad_norm": 2.375746726989746, + "learning_rate": 1.8014211034830972e-05, + "loss": 0.4132, + "step": 7229 + }, + { + "epoch": 0.2280397413657152, + "grad_norm": 2.2362873554229736, + "learning_rate": 1.801359999956544e-05, + "loss": 0.5562, + "step": 7230 + }, + { + "epoch": 0.2280712821321558, + "grad_norm": 2.1135287284851074, + "learning_rate": 1.801298888067177e-05, + "loss": 0.5535, + "step": 7231 + }, + { + "epoch": 0.22810282289859643, + "grad_norm": 2.867269992828369, + "learning_rate": 1.8012377678156337e-05, + "loss": 0.9128, + "step": 7232 + }, + { + "epoch": 0.22813436366503706, + "grad_norm": 2.2824079990386963, + "learning_rate": 1.8011766392025526e-05, + "loss": 0.5509, + "step": 7233 + }, + { + "epoch": 0.2281659044314777, + "grad_norm": 2.3251214027404785, + "learning_rate": 1.8011155022285713e-05, + "loss": 0.5519, + "step": 7234 + }, + { + "epoch": 0.22819744519791832, + "grad_norm": 3.2579495906829834, + "learning_rate": 1.801054356894327e-05, + "loss": 0.5795, + "step": 7235 + }, + { + "epoch": 0.22822898596435892, + "grad_norm": 2.5268445014953613, + "learning_rate": 1.800993203200459e-05, + "loss": 0.5676, + "step": 7236 + }, + { + "epoch": 0.22826052673079955, + "grad_norm": 1.9291960000991821, + "learning_rate": 1.8009320411476045e-05, + "loss": 0.5212, + "step": 7237 + }, + { + "epoch": 0.22829206749724018, + "grad_norm": 2.182103157043457, + "learning_rate": 1.8008708707364026e-05, + "loss": 0.5573, + "step": 7238 + }, + { + "epoch": 0.2283236082636808, + "grad_norm": 1.8340773582458496, + "learning_rate": 1.800809691967491e-05, + "loss": 0.5658, + "step": 7239 + }, + { + "epoch": 0.22835514903012144, + "grad_norm": 1.7665399312973022, + "learning_rate": 1.8007485048415088e-05, + "loss": 0.8713, + "step": 7240 + }, + { + "epoch": 0.22838668979656207, + "grad_norm": 1.9656147956848145, + "learning_rate": 1.800687309359094e-05, + "loss": 0.5659, + "step": 7241 + }, + { + "epoch": 0.22841823056300267, + "grad_norm": 1.5907223224639893, + "learning_rate": 1.8006261055208856e-05, + "loss": 0.5624, + "step": 7242 + }, + { + "epoch": 0.2284497713294433, + "grad_norm": 4.652174472808838, + "learning_rate": 1.800564893327522e-05, + "loss": 0.8913, + "step": 7243 + }, + { + "epoch": 0.22848131209588393, + "grad_norm": 1.971590518951416, + "learning_rate": 1.800503672779642e-05, + "loss": 0.5681, + "step": 7244 + }, + { + "epoch": 0.22851285286232456, + "grad_norm": 2.1646783351898193, + "learning_rate": 1.8004424438778848e-05, + "loss": 0.5483, + "step": 7245 + }, + { + "epoch": 0.2285443936287652, + "grad_norm": 2.783267021179199, + "learning_rate": 1.800381206622889e-05, + "loss": 0.5985, + "step": 7246 + }, + { + "epoch": 0.2285759343952058, + "grad_norm": 3.163954973220825, + "learning_rate": 1.800319961015294e-05, + "loss": 0.5761, + "step": 7247 + }, + { + "epoch": 0.22860747516164642, + "grad_norm": 1.607134461402893, + "learning_rate": 1.8002587070557384e-05, + "loss": 0.5456, + "step": 7248 + }, + { + "epoch": 0.22863901592808705, + "grad_norm": 2.282604694366455, + "learning_rate": 1.800197444744862e-05, + "loss": 0.5629, + "step": 7249 + }, + { + "epoch": 0.22867055669452768, + "grad_norm": 5.001379013061523, + "learning_rate": 1.8001361740833044e-05, + "loss": 0.8825, + "step": 7250 + }, + { + "epoch": 0.2287020974609683, + "grad_norm": 1.9438124895095825, + "learning_rate": 1.800074895071704e-05, + "loss": 0.4138, + "step": 7251 + }, + { + "epoch": 0.22873363822740891, + "grad_norm": 1.8598227500915527, + "learning_rate": 1.8000136077107008e-05, + "loss": 0.5571, + "step": 7252 + }, + { + "epoch": 0.22876517899384954, + "grad_norm": 2.4966418743133545, + "learning_rate": 1.7999523120009348e-05, + "loss": 0.8898, + "step": 7253 + }, + { + "epoch": 0.22879671976029017, + "grad_norm": 1.9585449695587158, + "learning_rate": 1.7998910079430452e-05, + "loss": 0.5868, + "step": 7254 + }, + { + "epoch": 0.2288282605267308, + "grad_norm": 2.156559467315674, + "learning_rate": 1.7998296955376714e-05, + "loss": 0.5569, + "step": 7255 + }, + { + "epoch": 0.22885980129317143, + "grad_norm": 2.6333556175231934, + "learning_rate": 1.799768374785454e-05, + "loss": 0.5475, + "step": 7256 + }, + { + "epoch": 0.22889134205961204, + "grad_norm": 2.422119140625, + "learning_rate": 1.7997070456870326e-05, + "loss": 0.5586, + "step": 7257 + }, + { + "epoch": 0.22892288282605266, + "grad_norm": 1.6713409423828125, + "learning_rate": 1.7996457082430474e-05, + "loss": 0.3927, + "step": 7258 + }, + { + "epoch": 0.2289544235924933, + "grad_norm": 2.722557544708252, + "learning_rate": 1.799584362454138e-05, + "loss": 0.5309, + "step": 7259 + }, + { + "epoch": 0.22898596435893392, + "grad_norm": 2.480414628982544, + "learning_rate": 1.7995230083209446e-05, + "loss": 0.5359, + "step": 7260 + }, + { + "epoch": 0.22901750512537455, + "grad_norm": 7.830502033233643, + "learning_rate": 1.7994616458441085e-05, + "loss": 0.5667, + "step": 7261 + }, + { + "epoch": 0.22904904589181518, + "grad_norm": 1.6704063415527344, + "learning_rate": 1.7994002750242692e-05, + "loss": 0.5396, + "step": 7262 + }, + { + "epoch": 0.22908058665825579, + "grad_norm": 3.110318422317505, + "learning_rate": 1.799338895862067e-05, + "loss": 0.5881, + "step": 7263 + }, + { + "epoch": 0.22911212742469642, + "grad_norm": 1.9670013189315796, + "learning_rate": 1.799277508358143e-05, + "loss": 0.5795, + "step": 7264 + }, + { + "epoch": 0.22914366819113705, + "grad_norm": 2.8519437313079834, + "learning_rate": 1.7992161125131375e-05, + "loss": 0.5718, + "step": 7265 + }, + { + "epoch": 0.22917520895757768, + "grad_norm": 2.6682724952697754, + "learning_rate": 1.7991547083276916e-05, + "loss": 0.5483, + "step": 7266 + }, + { + "epoch": 0.2292067497240183, + "grad_norm": 4.161194801330566, + "learning_rate": 1.7990932958024453e-05, + "loss": 0.5655, + "step": 7267 + }, + { + "epoch": 0.2292382904904589, + "grad_norm": 2.2736222743988037, + "learning_rate": 1.79903187493804e-05, + "loss": 0.5577, + "step": 7268 + }, + { + "epoch": 0.22926983125689954, + "grad_norm": 1.607399821281433, + "learning_rate": 1.798970445735117e-05, + "loss": 0.8918, + "step": 7269 + }, + { + "epoch": 0.22930137202334017, + "grad_norm": 2.060765027999878, + "learning_rate": 1.7989090081943166e-05, + "loss": 0.9023, + "step": 7270 + }, + { + "epoch": 0.2293329127897808, + "grad_norm": 2.636786699295044, + "learning_rate": 1.7988475623162807e-05, + "loss": 0.5729, + "step": 7271 + }, + { + "epoch": 0.22936445355622143, + "grad_norm": 2.5040082931518555, + "learning_rate": 1.7987861081016504e-05, + "loss": 0.5729, + "step": 7272 + }, + { + "epoch": 0.22939599432266203, + "grad_norm": 3.647862434387207, + "learning_rate": 1.7987246455510665e-05, + "loss": 0.4153, + "step": 7273 + }, + { + "epoch": 0.22942753508910266, + "grad_norm": 2.4054038524627686, + "learning_rate": 1.798663174665171e-05, + "loss": 0.5716, + "step": 7274 + }, + { + "epoch": 0.2294590758555433, + "grad_norm": 1.9876744747161865, + "learning_rate": 1.7986016954446046e-05, + "loss": 0.5505, + "step": 7275 + }, + { + "epoch": 0.22949061662198392, + "grad_norm": 2.8174660205841064, + "learning_rate": 1.7985402078900097e-05, + "loss": 0.592, + "step": 7276 + }, + { + "epoch": 0.22952215738842455, + "grad_norm": 1.8585233688354492, + "learning_rate": 1.7984787120020275e-05, + "loss": 0.544, + "step": 7277 + }, + { + "epoch": 0.22955369815486518, + "grad_norm": 2.5825533866882324, + "learning_rate": 1.7984172077813003e-05, + "loss": 0.552, + "step": 7278 + }, + { + "epoch": 0.22958523892130578, + "grad_norm": 2.2326817512512207, + "learning_rate": 1.7983556952284693e-05, + "loss": 0.5635, + "step": 7279 + }, + { + "epoch": 0.2296167796877464, + "grad_norm": 2.0966076850891113, + "learning_rate": 1.798294174344177e-05, + "loss": 0.5383, + "step": 7280 + }, + { + "epoch": 0.22964832045418704, + "grad_norm": 2.5197675228118896, + "learning_rate": 1.7982326451290647e-05, + "loss": 0.5662, + "step": 7281 + }, + { + "epoch": 0.22967986122062767, + "grad_norm": 2.4246840476989746, + "learning_rate": 1.7981711075837754e-05, + "loss": 0.5659, + "step": 7282 + }, + { + "epoch": 0.2297114019870683, + "grad_norm": 1.8921979665756226, + "learning_rate": 1.7981095617089505e-05, + "loss": 0.5254, + "step": 7283 + }, + { + "epoch": 0.2297429427535089, + "grad_norm": 1.6847283840179443, + "learning_rate": 1.7980480075052327e-05, + "loss": 0.5468, + "step": 7284 + }, + { + "epoch": 0.22977448351994953, + "grad_norm": 1.5562511682510376, + "learning_rate": 1.7979864449732643e-05, + "loss": 0.545, + "step": 7285 + }, + { + "epoch": 0.22980602428639016, + "grad_norm": 1.649473786354065, + "learning_rate": 1.7979248741136878e-05, + "loss": 0.5548, + "step": 7286 + }, + { + "epoch": 0.2298375650528308, + "grad_norm": 4.3100738525390625, + "learning_rate": 1.7978632949271456e-05, + "loss": 0.5452, + "step": 7287 + }, + { + "epoch": 0.22986910581927142, + "grad_norm": 2.2061171531677246, + "learning_rate": 1.7978017074142803e-05, + "loss": 0.5199, + "step": 7288 + }, + { + "epoch": 0.22990064658571202, + "grad_norm": 2.0779216289520264, + "learning_rate": 1.7977401115757347e-05, + "loss": 0.5567, + "step": 7289 + }, + { + "epoch": 0.22993218735215265, + "grad_norm": 2.237701892852783, + "learning_rate": 1.7976785074121518e-05, + "loss": 0.5537, + "step": 7290 + }, + { + "epoch": 0.22996372811859328, + "grad_norm": 2.393104076385498, + "learning_rate": 1.797616894924174e-05, + "loss": 0.5453, + "step": 7291 + }, + { + "epoch": 0.2299952688850339, + "grad_norm": 3.8168036937713623, + "learning_rate": 1.797555274112445e-05, + "loss": 0.4187, + "step": 7292 + }, + { + "epoch": 0.23002680965147454, + "grad_norm": 8.84260368347168, + "learning_rate": 1.7974936449776072e-05, + "loss": 0.5852, + "step": 7293 + }, + { + "epoch": 0.23005835041791514, + "grad_norm": 1.894027590751648, + "learning_rate": 1.7974320075203037e-05, + "loss": 0.5681, + "step": 7294 + }, + { + "epoch": 0.23008989118435577, + "grad_norm": 2.1526145935058594, + "learning_rate": 1.7973703617411784e-05, + "loss": 0.5657, + "step": 7295 + }, + { + "epoch": 0.2301214319507964, + "grad_norm": 2.123210906982422, + "learning_rate": 1.797308707640874e-05, + "loss": 0.5569, + "step": 7296 + }, + { + "epoch": 0.23015297271723703, + "grad_norm": 2.179248094558716, + "learning_rate": 1.7972470452200342e-05, + "loss": 0.5463, + "step": 7297 + }, + { + "epoch": 0.23018451348367766, + "grad_norm": 4.173264980316162, + "learning_rate": 1.7971853744793026e-05, + "loss": 0.5598, + "step": 7298 + }, + { + "epoch": 0.2302160542501183, + "grad_norm": 2.3487894535064697, + "learning_rate": 1.7971236954193223e-05, + "loss": 0.5325, + "step": 7299 + }, + { + "epoch": 0.2302475950165589, + "grad_norm": 2.0144295692443848, + "learning_rate": 1.7970620080407376e-05, + "loss": 0.5484, + "step": 7300 + }, + { + "epoch": 0.23027913578299952, + "grad_norm": 4.650521278381348, + "learning_rate": 1.7970003123441917e-05, + "loss": 0.4612, + "step": 7301 + }, + { + "epoch": 0.23031067654944015, + "grad_norm": 1.843595266342163, + "learning_rate": 1.7969386083303288e-05, + "loss": 0.5779, + "step": 7302 + }, + { + "epoch": 0.23034221731588078, + "grad_norm": 1.931010127067566, + "learning_rate": 1.796876895999793e-05, + "loss": 0.566, + "step": 7303 + }, + { + "epoch": 0.2303737580823214, + "grad_norm": 2.8404054641723633, + "learning_rate": 1.7968151753532278e-05, + "loss": 0.5502, + "step": 7304 + }, + { + "epoch": 0.230405298848762, + "grad_norm": 1.671299934387207, + "learning_rate": 1.7967534463912773e-05, + "loss": 0.3886, + "step": 7305 + }, + { + "epoch": 0.23043683961520264, + "grad_norm": 1.8562182188034058, + "learning_rate": 1.796691709114586e-05, + "loss": 0.5257, + "step": 7306 + }, + { + "epoch": 0.23046838038164327, + "grad_norm": 2.033928394317627, + "learning_rate": 1.7966299635237985e-05, + "loss": 0.5701, + "step": 7307 + }, + { + "epoch": 0.2304999211480839, + "grad_norm": 1.6981350183486938, + "learning_rate": 1.7965682096195584e-05, + "loss": 0.5385, + "step": 7308 + }, + { + "epoch": 0.23053146191452453, + "grad_norm": 2.2303316593170166, + "learning_rate": 1.7965064474025106e-05, + "loss": 0.5593, + "step": 7309 + }, + { + "epoch": 0.23056300268096513, + "grad_norm": 2.15995717048645, + "learning_rate": 1.7964446768732994e-05, + "loss": 0.5764, + "step": 7310 + }, + { + "epoch": 0.23059454344740576, + "grad_norm": 2.1739325523376465, + "learning_rate": 1.7963828980325696e-05, + "loss": 0.515, + "step": 7311 + }, + { + "epoch": 0.2306260842138464, + "grad_norm": 2.382481813430786, + "learning_rate": 1.796321110880966e-05, + "loss": 0.5451, + "step": 7312 + }, + { + "epoch": 0.23065762498028702, + "grad_norm": 2.0487911701202393, + "learning_rate": 1.7962593154191334e-05, + "loss": 0.5225, + "step": 7313 + }, + { + "epoch": 0.23068916574672765, + "grad_norm": 5.546956539154053, + "learning_rate": 1.7961975116477168e-05, + "loss": 0.5689, + "step": 7314 + }, + { + "epoch": 0.23072070651316828, + "grad_norm": 1.7472238540649414, + "learning_rate": 1.7961356995673603e-05, + "loss": 0.5824, + "step": 7315 + }, + { + "epoch": 0.23075224727960889, + "grad_norm": 1.8466668128967285, + "learning_rate": 1.7960738791787097e-05, + "loss": 0.925, + "step": 7316 + }, + { + "epoch": 0.23078378804604952, + "grad_norm": 2.405247449874878, + "learning_rate": 1.7960120504824104e-05, + "loss": 0.9262, + "step": 7317 + }, + { + "epoch": 0.23081532881249014, + "grad_norm": 2.202197551727295, + "learning_rate": 1.795950213479107e-05, + "loss": 0.5619, + "step": 7318 + }, + { + "epoch": 0.23084686957893077, + "grad_norm": 4.262872695922852, + "learning_rate": 1.7958883681694452e-05, + "loss": 0.5854, + "step": 7319 + }, + { + "epoch": 0.2308784103453714, + "grad_norm": 2.1533620357513428, + "learning_rate": 1.7958265145540702e-05, + "loss": 0.5916, + "step": 7320 + }, + { + "epoch": 0.230909951111812, + "grad_norm": 2.3931591510772705, + "learning_rate": 1.7957646526336277e-05, + "loss": 0.56, + "step": 7321 + }, + { + "epoch": 0.23094149187825264, + "grad_norm": 1.8387928009033203, + "learning_rate": 1.7957027824087633e-05, + "loss": 0.5888, + "step": 7322 + }, + { + "epoch": 0.23097303264469327, + "grad_norm": 2.337435245513916, + "learning_rate": 1.7956409038801222e-05, + "loss": 0.5446, + "step": 7323 + }, + { + "epoch": 0.2310045734111339, + "grad_norm": 2.3530750274658203, + "learning_rate": 1.7955790170483505e-05, + "loss": 0.5807, + "step": 7324 + }, + { + "epoch": 0.23103611417757453, + "grad_norm": 2.568427085876465, + "learning_rate": 1.795517121914094e-05, + "loss": 0.5766, + "step": 7325 + }, + { + "epoch": 0.23106765494401513, + "grad_norm": 2.0649967193603516, + "learning_rate": 1.7954552184779985e-05, + "loss": 0.5643, + "step": 7326 + }, + { + "epoch": 0.23109919571045576, + "grad_norm": 4.756355285644531, + "learning_rate": 1.7953933067407107e-05, + "loss": 0.45, + "step": 7327 + }, + { + "epoch": 0.2311307364768964, + "grad_norm": 2.010225772857666, + "learning_rate": 1.7953313867028758e-05, + "loss": 0.5777, + "step": 7328 + }, + { + "epoch": 0.23116227724333702, + "grad_norm": 3.149738311767578, + "learning_rate": 1.79526945836514e-05, + "loss": 0.4087, + "step": 7329 + }, + { + "epoch": 0.23119381800977765, + "grad_norm": 6.233059406280518, + "learning_rate": 1.79520752172815e-05, + "loss": 0.8993, + "step": 7330 + }, + { + "epoch": 0.23122535877621825, + "grad_norm": 2.297729969024658, + "learning_rate": 1.7951455767925524e-05, + "loss": 0.5611, + "step": 7331 + }, + { + "epoch": 0.23125689954265888, + "grad_norm": 2.393968105316162, + "learning_rate": 1.7950836235589932e-05, + "loss": 0.5462, + "step": 7332 + }, + { + "epoch": 0.2312884403090995, + "grad_norm": 2.468482494354248, + "learning_rate": 1.795021662028119e-05, + "loss": 0.5513, + "step": 7333 + }, + { + "epoch": 0.23131998107554014, + "grad_norm": 1.670810341835022, + "learning_rate": 1.7949596922005762e-05, + "loss": 0.8605, + "step": 7334 + }, + { + "epoch": 0.23135152184198077, + "grad_norm": 1.6927722692489624, + "learning_rate": 1.794897714077012e-05, + "loss": 0.9008, + "step": 7335 + }, + { + "epoch": 0.2313830626084214, + "grad_norm": 1.889897346496582, + "learning_rate": 1.794835727658073e-05, + "loss": 0.6081, + "step": 7336 + }, + { + "epoch": 0.231414603374862, + "grad_norm": 2.562199354171753, + "learning_rate": 1.7947737329444058e-05, + "loss": 0.5743, + "step": 7337 + }, + { + "epoch": 0.23144614414130263, + "grad_norm": 2.2357380390167236, + "learning_rate": 1.7947117299366576e-05, + "loss": 0.5402, + "step": 7338 + }, + { + "epoch": 0.23147768490774326, + "grad_norm": 3.4163999557495117, + "learning_rate": 1.7946497186354753e-05, + "loss": 0.5647, + "step": 7339 + }, + { + "epoch": 0.2315092256741839, + "grad_norm": 1.7042648792266846, + "learning_rate": 1.7945876990415064e-05, + "loss": 0.5584, + "step": 7340 + }, + { + "epoch": 0.23154076644062452, + "grad_norm": 3.235001802444458, + "learning_rate": 1.7945256711553977e-05, + "loss": 0.569, + "step": 7341 + }, + { + "epoch": 0.23157230720706512, + "grad_norm": 2.3349478244781494, + "learning_rate": 1.7944636349777972e-05, + "loss": 0.5753, + "step": 7342 + }, + { + "epoch": 0.23160384797350575, + "grad_norm": 1.757590889930725, + "learning_rate": 1.794401590509351e-05, + "loss": 0.5586, + "step": 7343 + }, + { + "epoch": 0.23163538873994638, + "grad_norm": 2.105178117752075, + "learning_rate": 1.794339537750708e-05, + "loss": 0.4698, + "step": 7344 + }, + { + "epoch": 0.231666929506387, + "grad_norm": 2.3729517459869385, + "learning_rate": 1.7942774767025145e-05, + "loss": 0.8861, + "step": 7345 + }, + { + "epoch": 0.23169847027282764, + "grad_norm": 2.1266653537750244, + "learning_rate": 1.7942154073654194e-05, + "loss": 0.5311, + "step": 7346 + }, + { + "epoch": 0.23173001103926824, + "grad_norm": 2.1626124382019043, + "learning_rate": 1.7941533297400696e-05, + "loss": 0.5577, + "step": 7347 + }, + { + "epoch": 0.23176155180570887, + "grad_norm": 2.5677740573883057, + "learning_rate": 1.7940912438271133e-05, + "loss": 0.5642, + "step": 7348 + }, + { + "epoch": 0.2317930925721495, + "grad_norm": 2.388603687286377, + "learning_rate": 1.794029149627198e-05, + "loss": 0.54, + "step": 7349 + }, + { + "epoch": 0.23182463333859013, + "grad_norm": 2.4518227577209473, + "learning_rate": 1.793967047140972e-05, + "loss": 0.5613, + "step": 7350 + }, + { + "epoch": 0.23185617410503076, + "grad_norm": 2.003835439682007, + "learning_rate": 1.7939049363690834e-05, + "loss": 0.558, + "step": 7351 + }, + { + "epoch": 0.2318877148714714, + "grad_norm": 2.2637836933135986, + "learning_rate": 1.7938428173121806e-05, + "loss": 0.845, + "step": 7352 + }, + { + "epoch": 0.231919255637912, + "grad_norm": 1.8179324865341187, + "learning_rate": 1.7937806899709112e-05, + "loss": 0.5835, + "step": 7353 + }, + { + "epoch": 0.23195079640435262, + "grad_norm": 1.667626142501831, + "learning_rate": 1.793718554345924e-05, + "loss": 0.5394, + "step": 7354 + }, + { + "epoch": 0.23198233717079325, + "grad_norm": 2.827930450439453, + "learning_rate": 1.7936564104378676e-05, + "loss": 0.5666, + "step": 7355 + }, + { + "epoch": 0.23201387793723388, + "grad_norm": 1.7480719089508057, + "learning_rate": 1.7935942582473906e-05, + "loss": 0.9075, + "step": 7356 + }, + { + "epoch": 0.2320454187036745, + "grad_norm": 2.315833806991577, + "learning_rate": 1.7935320977751407e-05, + "loss": 0.5713, + "step": 7357 + }, + { + "epoch": 0.2320769594701151, + "grad_norm": 1.9855250120162964, + "learning_rate": 1.7934699290217676e-05, + "loss": 0.4208, + "step": 7358 + }, + { + "epoch": 0.23210850023655574, + "grad_norm": 2.7681422233581543, + "learning_rate": 1.7934077519879192e-05, + "loss": 0.5855, + "step": 7359 + }, + { + "epoch": 0.23214004100299637, + "grad_norm": 12.501077651977539, + "learning_rate": 1.7933455666742453e-05, + "loss": 0.581, + "step": 7360 + }, + { + "epoch": 0.232171581769437, + "grad_norm": 2.820342540740967, + "learning_rate": 1.793283373081394e-05, + "loss": 0.5365, + "step": 7361 + }, + { + "epoch": 0.23220312253587763, + "grad_norm": 2.3062446117401123, + "learning_rate": 1.7932211712100152e-05, + "loss": 0.522, + "step": 7362 + }, + { + "epoch": 0.23223466330231823, + "grad_norm": 1.7356176376342773, + "learning_rate": 1.7931589610607572e-05, + "loss": 0.8694, + "step": 7363 + }, + { + "epoch": 0.23226620406875886, + "grad_norm": 1.6011905670166016, + "learning_rate": 1.79309674263427e-05, + "loss": 0.8682, + "step": 7364 + }, + { + "epoch": 0.2322977448351995, + "grad_norm": 2.285557508468628, + "learning_rate": 1.793034515931202e-05, + "loss": 0.5754, + "step": 7365 + }, + { + "epoch": 0.23232928560164012, + "grad_norm": 2.287351131439209, + "learning_rate": 1.792972280952203e-05, + "loss": 0.5809, + "step": 7366 + }, + { + "epoch": 0.23236082636808075, + "grad_norm": 2.9047136306762695, + "learning_rate": 1.792910037697923e-05, + "loss": 0.5819, + "step": 7367 + }, + { + "epoch": 0.23239236713452138, + "grad_norm": 1.790505290031433, + "learning_rate": 1.7928477861690107e-05, + "loss": 0.5585, + "step": 7368 + }, + { + "epoch": 0.23242390790096198, + "grad_norm": 1.495437741279602, + "learning_rate": 1.792785526366116e-05, + "loss": 0.5591, + "step": 7369 + }, + { + "epoch": 0.23245544866740261, + "grad_norm": 3.0831894874572754, + "learning_rate": 1.7927232582898893e-05, + "loss": 0.5527, + "step": 7370 + }, + { + "epoch": 0.23248698943384324, + "grad_norm": 2.110016345977783, + "learning_rate": 1.7926609819409793e-05, + "loss": 0.5694, + "step": 7371 + }, + { + "epoch": 0.23251853020028387, + "grad_norm": 2.538043737411499, + "learning_rate": 1.7925986973200367e-05, + "loss": 0.4021, + "step": 7372 + }, + { + "epoch": 0.2325500709667245, + "grad_norm": 1.4780852794647217, + "learning_rate": 1.7925364044277117e-05, + "loss": 0.4122, + "step": 7373 + }, + { + "epoch": 0.2325816117331651, + "grad_norm": 2.0709896087646484, + "learning_rate": 1.7924741032646534e-05, + "loss": 0.579, + "step": 7374 + }, + { + "epoch": 0.23261315249960574, + "grad_norm": 2.4691245555877686, + "learning_rate": 1.7924117938315126e-05, + "loss": 0.5839, + "step": 7375 + }, + { + "epoch": 0.23264469326604637, + "grad_norm": 2.2083916664123535, + "learning_rate": 1.7923494761289393e-05, + "loss": 0.5581, + "step": 7376 + }, + { + "epoch": 0.232676234032487, + "grad_norm": 2.2747323513031006, + "learning_rate": 1.7922871501575843e-05, + "loss": 0.5822, + "step": 7377 + }, + { + "epoch": 0.23270777479892762, + "grad_norm": 1.9469261169433594, + "learning_rate": 1.792224815918097e-05, + "loss": 0.5803, + "step": 7378 + }, + { + "epoch": 0.23273931556536823, + "grad_norm": 1.9971351623535156, + "learning_rate": 1.7921624734111292e-05, + "loss": 0.5585, + "step": 7379 + }, + { + "epoch": 0.23277085633180886, + "grad_norm": 2.5308148860931396, + "learning_rate": 1.7921001226373312e-05, + "loss": 0.5841, + "step": 7380 + }, + { + "epoch": 0.2328023970982495, + "grad_norm": 1.8567783832550049, + "learning_rate": 1.792037763597353e-05, + "loss": 0.5504, + "step": 7381 + }, + { + "epoch": 0.23283393786469012, + "grad_norm": 1.6903486251831055, + "learning_rate": 1.7919753962918454e-05, + "loss": 0.5595, + "step": 7382 + }, + { + "epoch": 0.23286547863113075, + "grad_norm": 2.1447176933288574, + "learning_rate": 1.79191302072146e-05, + "loss": 0.5784, + "step": 7383 + }, + { + "epoch": 0.23289701939757135, + "grad_norm": 2.1530568599700928, + "learning_rate": 1.7918506368868474e-05, + "loss": 0.5387, + "step": 7384 + }, + { + "epoch": 0.23292856016401198, + "grad_norm": 2.3867132663726807, + "learning_rate": 1.7917882447886585e-05, + "loss": 0.9042, + "step": 7385 + }, + { + "epoch": 0.2329601009304526, + "grad_norm": 2.188201904296875, + "learning_rate": 1.791725844427544e-05, + "loss": 0.5434, + "step": 7386 + }, + { + "epoch": 0.23299164169689324, + "grad_norm": 5.208747863769531, + "learning_rate": 1.791663435804156e-05, + "loss": 0.5507, + "step": 7387 + }, + { + "epoch": 0.23302318246333387, + "grad_norm": 1.9301891326904297, + "learning_rate": 1.791601018919145e-05, + "loss": 0.5621, + "step": 7388 + }, + { + "epoch": 0.2330547232297745, + "grad_norm": 1.7732335329055786, + "learning_rate": 1.791538593773163e-05, + "loss": 0.5733, + "step": 7389 + }, + { + "epoch": 0.2330862639962151, + "grad_norm": 1.7321637868881226, + "learning_rate": 1.791476160366861e-05, + "loss": 0.57, + "step": 7390 + }, + { + "epoch": 0.23311780476265573, + "grad_norm": 1.8680498600006104, + "learning_rate": 1.791413718700891e-05, + "loss": 0.5621, + "step": 7391 + }, + { + "epoch": 0.23314934552909636, + "grad_norm": 1.6408063173294067, + "learning_rate": 1.791351268775904e-05, + "loss": 0.8755, + "step": 7392 + }, + { + "epoch": 0.233180886295537, + "grad_norm": 1.6673297882080078, + "learning_rate": 1.7912888105925522e-05, + "loss": 0.5794, + "step": 7393 + }, + { + "epoch": 0.23321242706197762, + "grad_norm": 1.7327592372894287, + "learning_rate": 1.791226344151487e-05, + "loss": 0.561, + "step": 7394 + }, + { + "epoch": 0.23324396782841822, + "grad_norm": 2.3266167640686035, + "learning_rate": 1.791163869453361e-05, + "loss": 0.5912, + "step": 7395 + }, + { + "epoch": 0.23327550859485885, + "grad_norm": 1.8056998252868652, + "learning_rate": 1.7911013864988254e-05, + "loss": 0.5664, + "step": 7396 + }, + { + "epoch": 0.23330704936129948, + "grad_norm": 2.4570016860961914, + "learning_rate": 1.7910388952885324e-05, + "loss": 0.5697, + "step": 7397 + }, + { + "epoch": 0.2333385901277401, + "grad_norm": 1.7533284425735474, + "learning_rate": 1.7909763958231346e-05, + "loss": 0.5625, + "step": 7398 + }, + { + "epoch": 0.23337013089418074, + "grad_norm": 1.9607526063919067, + "learning_rate": 1.7909138881032838e-05, + "loss": 0.5511, + "step": 7399 + }, + { + "epoch": 0.23340167166062134, + "grad_norm": 2.182992935180664, + "learning_rate": 1.7908513721296326e-05, + "loss": 0.5478, + "step": 7400 + }, + { + "epoch": 0.23343321242706197, + "grad_norm": 2.5909013748168945, + "learning_rate": 1.7907888479028332e-05, + "loss": 0.5861, + "step": 7401 + }, + { + "epoch": 0.2334647531935026, + "grad_norm": 1.5816460847854614, + "learning_rate": 1.7907263154235378e-05, + "loss": 0.8943, + "step": 7402 + }, + { + "epoch": 0.23349629395994323, + "grad_norm": 2.1465814113616943, + "learning_rate": 1.7906637746923997e-05, + "loss": 0.5825, + "step": 7403 + }, + { + "epoch": 0.23352783472638386, + "grad_norm": 2.1074955463409424, + "learning_rate": 1.790601225710071e-05, + "loss": 0.5805, + "step": 7404 + }, + { + "epoch": 0.2335593754928245, + "grad_norm": 1.9164317846298218, + "learning_rate": 1.7905386684772044e-05, + "loss": 0.5525, + "step": 7405 + }, + { + "epoch": 0.2335909162592651, + "grad_norm": 2.0356662273406982, + "learning_rate": 1.7904761029944534e-05, + "loss": 0.5295, + "step": 7406 + }, + { + "epoch": 0.23362245702570572, + "grad_norm": 1.8487995862960815, + "learning_rate": 1.7904135292624702e-05, + "loss": 0.5552, + "step": 7407 + }, + { + "epoch": 0.23365399779214635, + "grad_norm": 2.111416816711426, + "learning_rate": 1.7903509472819084e-05, + "loss": 0.578, + "step": 7408 + }, + { + "epoch": 0.23368553855858698, + "grad_norm": 2.009361982345581, + "learning_rate": 1.7902883570534205e-05, + "loss": 0.5345, + "step": 7409 + }, + { + "epoch": 0.2337170793250276, + "grad_norm": 1.7460671663284302, + "learning_rate": 1.7902257585776597e-05, + "loss": 0.5411, + "step": 7410 + }, + { + "epoch": 0.2337486200914682, + "grad_norm": 1.640189290046692, + "learning_rate": 1.79016315185528e-05, + "loss": 0.8997, + "step": 7411 + }, + { + "epoch": 0.23378016085790884, + "grad_norm": 1.9426889419555664, + "learning_rate": 1.790100536886934e-05, + "loss": 0.4023, + "step": 7412 + }, + { + "epoch": 0.23381170162434947, + "grad_norm": 2.65559458732605, + "learning_rate": 1.7900379136732756e-05, + "loss": 0.546, + "step": 7413 + }, + { + "epoch": 0.2338432423907901, + "grad_norm": 1.9017771482467651, + "learning_rate": 1.789975282214958e-05, + "loss": 0.5324, + "step": 7414 + }, + { + "epoch": 0.23387478315723073, + "grad_norm": 2.0384931564331055, + "learning_rate": 1.789912642512635e-05, + "loss": 0.5482, + "step": 7415 + }, + { + "epoch": 0.23390632392367133, + "grad_norm": 1.8410704135894775, + "learning_rate": 1.78984999456696e-05, + "loss": 0.5603, + "step": 7416 + }, + { + "epoch": 0.23393786469011196, + "grad_norm": 1.916532039642334, + "learning_rate": 1.7897873383785875e-05, + "loss": 0.5494, + "step": 7417 + }, + { + "epoch": 0.2339694054565526, + "grad_norm": 2.199944019317627, + "learning_rate": 1.7897246739481702e-05, + "loss": 0.5519, + "step": 7418 + }, + { + "epoch": 0.23400094622299322, + "grad_norm": 2.0503664016723633, + "learning_rate": 1.789662001276363e-05, + "loss": 0.5636, + "step": 7419 + }, + { + "epoch": 0.23403248698943385, + "grad_norm": 2.41585111618042, + "learning_rate": 1.78959932036382e-05, + "loss": 0.5574, + "step": 7420 + }, + { + "epoch": 0.23406402775587445, + "grad_norm": 1.9983587265014648, + "learning_rate": 1.7895366312111944e-05, + "loss": 0.5475, + "step": 7421 + }, + { + "epoch": 0.23409556852231508, + "grad_norm": 2.012535572052002, + "learning_rate": 1.7894739338191413e-05, + "loss": 0.5626, + "step": 7422 + }, + { + "epoch": 0.2341271092887557, + "grad_norm": 2.277390241622925, + "learning_rate": 1.789411228188315e-05, + "loss": 0.5425, + "step": 7423 + }, + { + "epoch": 0.23415865005519634, + "grad_norm": 6.267063617706299, + "learning_rate": 1.789348514319369e-05, + "loss": 0.5768, + "step": 7424 + }, + { + "epoch": 0.23419019082163697, + "grad_norm": 1.9471842050552368, + "learning_rate": 1.7892857922129586e-05, + "loss": 0.5588, + "step": 7425 + }, + { + "epoch": 0.2342217315880776, + "grad_norm": 1.8034666776657104, + "learning_rate": 1.789223061869738e-05, + "loss": 0.553, + "step": 7426 + }, + { + "epoch": 0.2342532723545182, + "grad_norm": 2.1211395263671875, + "learning_rate": 1.7891603232903623e-05, + "loss": 0.5788, + "step": 7427 + }, + { + "epoch": 0.23428481312095883, + "grad_norm": 1.695216417312622, + "learning_rate": 1.7890975764754854e-05, + "loss": 0.5419, + "step": 7428 + }, + { + "epoch": 0.23431635388739946, + "grad_norm": 1.715461254119873, + "learning_rate": 1.789034821425763e-05, + "loss": 0.5282, + "step": 7429 + }, + { + "epoch": 0.2343478946538401, + "grad_norm": 2.8645942211151123, + "learning_rate": 1.788972058141849e-05, + "loss": 0.56, + "step": 7430 + }, + { + "epoch": 0.23437943542028072, + "grad_norm": 1.2506433725357056, + "learning_rate": 1.7889092866243993e-05, + "loss": 0.4277, + "step": 7431 + }, + { + "epoch": 0.23441097618672133, + "grad_norm": 2.6075022220611572, + "learning_rate": 1.788846506874069e-05, + "loss": 0.9055, + "step": 7432 + }, + { + "epoch": 0.23444251695316196, + "grad_norm": 2.2295122146606445, + "learning_rate": 1.7887837188915123e-05, + "loss": 0.5642, + "step": 7433 + }, + { + "epoch": 0.23447405771960259, + "grad_norm": 2.0962469577789307, + "learning_rate": 1.788720922677385e-05, + "loss": 0.5599, + "step": 7434 + }, + { + "epoch": 0.23450559848604322, + "grad_norm": 2.0779852867126465, + "learning_rate": 1.788658118232343e-05, + "loss": 0.5355, + "step": 7435 + }, + { + "epoch": 0.23453713925248384, + "grad_norm": 4.999734878540039, + "learning_rate": 1.7885953055570406e-05, + "loss": 0.5341, + "step": 7436 + }, + { + "epoch": 0.23456868001892445, + "grad_norm": 1.6426944732666016, + "learning_rate": 1.788532484652134e-05, + "loss": 0.578, + "step": 7437 + }, + { + "epoch": 0.23460022078536508, + "grad_norm": 1.7802828550338745, + "learning_rate": 1.7884696555182786e-05, + "loss": 0.5281, + "step": 7438 + }, + { + "epoch": 0.2346317615518057, + "grad_norm": 2.4436869621276855, + "learning_rate": 1.7884068181561302e-05, + "loss": 0.5661, + "step": 7439 + }, + { + "epoch": 0.23466330231824634, + "grad_norm": 2.242827892303467, + "learning_rate": 1.788343972566345e-05, + "loss": 0.5629, + "step": 7440 + }, + { + "epoch": 0.23469484308468697, + "grad_norm": 2.4765665531158447, + "learning_rate": 1.7882811187495775e-05, + "loss": 0.5734, + "step": 7441 + }, + { + "epoch": 0.2347263838511276, + "grad_norm": 1.6864391565322876, + "learning_rate": 1.788218256706485e-05, + "loss": 0.8902, + "step": 7442 + }, + { + "epoch": 0.2347579246175682, + "grad_norm": 1.7249943017959595, + "learning_rate": 1.7881553864377224e-05, + "loss": 0.5708, + "step": 7443 + }, + { + "epoch": 0.23478946538400883, + "grad_norm": 2.4792850017547607, + "learning_rate": 1.7880925079439465e-05, + "loss": 0.5432, + "step": 7444 + }, + { + "epoch": 0.23482100615044946, + "grad_norm": 2.9244208335876465, + "learning_rate": 1.7880296212258133e-05, + "loss": 0.5802, + "step": 7445 + }, + { + "epoch": 0.2348525469168901, + "grad_norm": 5.183645248413086, + "learning_rate": 1.7879667262839796e-05, + "loss": 0.5741, + "step": 7446 + }, + { + "epoch": 0.23488408768333072, + "grad_norm": 2.3685142993927, + "learning_rate": 1.787903823119101e-05, + "loss": 0.5657, + "step": 7447 + }, + { + "epoch": 0.23491562844977132, + "grad_norm": 1.9362136125564575, + "learning_rate": 1.7878409117318342e-05, + "loss": 0.5663, + "step": 7448 + }, + { + "epoch": 0.23494716921621195, + "grad_norm": 1.805367112159729, + "learning_rate": 1.7877779921228357e-05, + "loss": 0.5355, + "step": 7449 + }, + { + "epoch": 0.23497870998265258, + "grad_norm": 1.950238585472107, + "learning_rate": 1.7877150642927625e-05, + "loss": 0.5778, + "step": 7450 + }, + { + "epoch": 0.2350102507490932, + "grad_norm": 2.1034178733825684, + "learning_rate": 1.7876521282422704e-05, + "loss": 0.5632, + "step": 7451 + }, + { + "epoch": 0.23504179151553384, + "grad_norm": 1.8976341485977173, + "learning_rate": 1.787589183972017e-05, + "loss": 0.6092, + "step": 7452 + }, + { + "epoch": 0.23507333228197444, + "grad_norm": 1.9545283317565918, + "learning_rate": 1.7875262314826586e-05, + "loss": 0.5544, + "step": 7453 + }, + { + "epoch": 0.23510487304841507, + "grad_norm": 1.9999582767486572, + "learning_rate": 1.787463270774853e-05, + "loss": 0.3987, + "step": 7454 + }, + { + "epoch": 0.2351364138148557, + "grad_norm": 2.465904951095581, + "learning_rate": 1.7874003018492566e-05, + "loss": 0.5738, + "step": 7455 + }, + { + "epoch": 0.23516795458129633, + "grad_norm": 1.5312467813491821, + "learning_rate": 1.7873373247065267e-05, + "loss": 0.5804, + "step": 7456 + }, + { + "epoch": 0.23519949534773696, + "grad_norm": 1.5827710628509521, + "learning_rate": 1.78727433934732e-05, + "loss": 0.5413, + "step": 7457 + }, + { + "epoch": 0.23523103611417756, + "grad_norm": 1.908293604850769, + "learning_rate": 1.7872113457722947e-05, + "loss": 0.5451, + "step": 7458 + }, + { + "epoch": 0.2352625768806182, + "grad_norm": 1.7557897567749023, + "learning_rate": 1.7871483439821075e-05, + "loss": 0.9103, + "step": 7459 + }, + { + "epoch": 0.23529411764705882, + "grad_norm": 1.58220374584198, + "learning_rate": 1.7870853339774165e-05, + "loss": 0.5904, + "step": 7460 + }, + { + "epoch": 0.23532565841349945, + "grad_norm": 1.6366499662399292, + "learning_rate": 1.7870223157588786e-05, + "loss": 0.5624, + "step": 7461 + }, + { + "epoch": 0.23535719917994008, + "grad_norm": 2.4646389484405518, + "learning_rate": 1.7869592893271518e-05, + "loss": 0.5614, + "step": 7462 + }, + { + "epoch": 0.2353887399463807, + "grad_norm": 1.5986566543579102, + "learning_rate": 1.7868962546828937e-05, + "loss": 0.5442, + "step": 7463 + }, + { + "epoch": 0.2354202807128213, + "grad_norm": 1.6387715339660645, + "learning_rate": 1.7868332118267623e-05, + "loss": 0.9082, + "step": 7464 + }, + { + "epoch": 0.23545182147926194, + "grad_norm": 1.9027913808822632, + "learning_rate": 1.7867701607594153e-05, + "loss": 0.5337, + "step": 7465 + }, + { + "epoch": 0.23548336224570257, + "grad_norm": 2.284926176071167, + "learning_rate": 1.786707101481511e-05, + "loss": 0.5703, + "step": 7466 + }, + { + "epoch": 0.2355149030121432, + "grad_norm": 1.8475658893585205, + "learning_rate": 1.7866440339937065e-05, + "loss": 0.5736, + "step": 7467 + }, + { + "epoch": 0.23554644377858383, + "grad_norm": 2.261148691177368, + "learning_rate": 1.7865809582966613e-05, + "loss": 0.5639, + "step": 7468 + }, + { + "epoch": 0.23557798454502443, + "grad_norm": 1.7090492248535156, + "learning_rate": 1.786517874391033e-05, + "loss": 0.547, + "step": 7469 + }, + { + "epoch": 0.23560952531146506, + "grad_norm": 1.5062782764434814, + "learning_rate": 1.7864547822774798e-05, + "loss": 0.5419, + "step": 7470 + }, + { + "epoch": 0.2356410660779057, + "grad_norm": 1.9940508604049683, + "learning_rate": 1.7863916819566604e-05, + "loss": 0.5066, + "step": 7471 + }, + { + "epoch": 0.23567260684434632, + "grad_norm": 2.1543068885803223, + "learning_rate": 1.786328573429233e-05, + "loss": 0.5515, + "step": 7472 + }, + { + "epoch": 0.23570414761078695, + "grad_norm": 2.531899929046631, + "learning_rate": 1.7862654566958568e-05, + "loss": 0.8467, + "step": 7473 + }, + { + "epoch": 0.23573568837722755, + "grad_norm": 2.292431116104126, + "learning_rate": 1.7862023317571895e-05, + "loss": 0.5897, + "step": 7474 + }, + { + "epoch": 0.23576722914366818, + "grad_norm": 1.9281357526779175, + "learning_rate": 1.786139198613891e-05, + "loss": 0.5461, + "step": 7475 + }, + { + "epoch": 0.2357987699101088, + "grad_norm": 1.485550045967102, + "learning_rate": 1.7860760572666186e-05, + "loss": 0.5677, + "step": 7476 + }, + { + "epoch": 0.23583031067654944, + "grad_norm": 1.7622798681259155, + "learning_rate": 1.7860129077160332e-05, + "loss": 0.5376, + "step": 7477 + }, + { + "epoch": 0.23586185144299007, + "grad_norm": 2.0127673149108887, + "learning_rate": 1.785949749962792e-05, + "loss": 0.5131, + "step": 7478 + }, + { + "epoch": 0.2358933922094307, + "grad_norm": 1.1931227445602417, + "learning_rate": 1.7858865840075554e-05, + "loss": 0.386, + "step": 7479 + }, + { + "epoch": 0.2359249329758713, + "grad_norm": 1.681113362312317, + "learning_rate": 1.7858234098509817e-05, + "loss": 0.5451, + "step": 7480 + }, + { + "epoch": 0.23595647374231193, + "grad_norm": 1.842545509338379, + "learning_rate": 1.785760227493731e-05, + "loss": 0.5622, + "step": 7481 + }, + { + "epoch": 0.23598801450875256, + "grad_norm": 2.1206862926483154, + "learning_rate": 1.785697036936462e-05, + "loss": 0.5263, + "step": 7482 + }, + { + "epoch": 0.2360195552751932, + "grad_norm": 1.542089819908142, + "learning_rate": 1.785633838179834e-05, + "loss": 0.5594, + "step": 7483 + }, + { + "epoch": 0.23605109604163382, + "grad_norm": 2.2313766479492188, + "learning_rate": 1.7855706312245075e-05, + "loss": 0.5434, + "step": 7484 + }, + { + "epoch": 0.23608263680807443, + "grad_norm": 2.153261184692383, + "learning_rate": 1.7855074160711407e-05, + "loss": 0.5459, + "step": 7485 + }, + { + "epoch": 0.23611417757451506, + "grad_norm": 2.0535459518432617, + "learning_rate": 1.7854441927203947e-05, + "loss": 0.5455, + "step": 7486 + }, + { + "epoch": 0.23614571834095568, + "grad_norm": 2.5952117443084717, + "learning_rate": 1.7853809611729285e-05, + "loss": 0.5535, + "step": 7487 + }, + { + "epoch": 0.23617725910739631, + "grad_norm": 1.8070815801620483, + "learning_rate": 1.785317721429402e-05, + "loss": 0.5504, + "step": 7488 + }, + { + "epoch": 0.23620879987383694, + "grad_norm": 1.883534550666809, + "learning_rate": 1.7852544734904754e-05, + "loss": 0.5251, + "step": 7489 + }, + { + "epoch": 0.23624034064027755, + "grad_norm": 1.9127732515335083, + "learning_rate": 1.7851912173568086e-05, + "loss": 0.5568, + "step": 7490 + }, + { + "epoch": 0.23627188140671818, + "grad_norm": 1.5823924541473389, + "learning_rate": 1.785127953029062e-05, + "loss": 0.5239, + "step": 7491 + }, + { + "epoch": 0.2363034221731588, + "grad_norm": 2.041130781173706, + "learning_rate": 1.7850646805078952e-05, + "loss": 0.5241, + "step": 7492 + }, + { + "epoch": 0.23633496293959944, + "grad_norm": 1.928021788597107, + "learning_rate": 1.785001399793969e-05, + "loss": 0.5924, + "step": 7493 + }, + { + "epoch": 0.23636650370604007, + "grad_norm": 1.6528637409210205, + "learning_rate": 1.7849381108879438e-05, + "loss": 0.5698, + "step": 7494 + }, + { + "epoch": 0.2363980444724807, + "grad_norm": 2.2602760791778564, + "learning_rate": 1.78487481379048e-05, + "loss": 0.5523, + "step": 7495 + }, + { + "epoch": 0.2364295852389213, + "grad_norm": 5.5924272537231445, + "learning_rate": 1.7848115085022376e-05, + "loss": 0.5366, + "step": 7496 + }, + { + "epoch": 0.23646112600536193, + "grad_norm": 1.4795217514038086, + "learning_rate": 1.7847481950238782e-05, + "loss": 0.5644, + "step": 7497 + }, + { + "epoch": 0.23649266677180256, + "grad_norm": 2.129195213317871, + "learning_rate": 1.7846848733560617e-05, + "loss": 0.5364, + "step": 7498 + }, + { + "epoch": 0.2365242075382432, + "grad_norm": 2.4157967567443848, + "learning_rate": 1.7846215434994495e-05, + "loss": 0.5597, + "step": 7499 + }, + { + "epoch": 0.23655574830468382, + "grad_norm": 1.5126570463180542, + "learning_rate": 1.7845582054547023e-05, + "loss": 0.5438, + "step": 7500 + }, + { + "epoch": 0.23658728907112442, + "grad_norm": 2.223973274230957, + "learning_rate": 1.784494859222481e-05, + "loss": 0.5454, + "step": 7501 + }, + { + "epoch": 0.23661882983756505, + "grad_norm": 1.6348583698272705, + "learning_rate": 1.7844315048034465e-05, + "loss": 0.5525, + "step": 7502 + }, + { + "epoch": 0.23665037060400568, + "grad_norm": 2.7514548301696777, + "learning_rate": 1.78436814219826e-05, + "loss": 0.5333, + "step": 7503 + }, + { + "epoch": 0.2366819113704463, + "grad_norm": 1.6663402318954468, + "learning_rate": 1.7843047714075832e-05, + "loss": 0.5611, + "step": 7504 + }, + { + "epoch": 0.23671345213688694, + "grad_norm": 1.4275709390640259, + "learning_rate": 1.7842413924320773e-05, + "loss": 0.5235, + "step": 7505 + }, + { + "epoch": 0.23674499290332754, + "grad_norm": 1.6137263774871826, + "learning_rate": 1.7841780052724035e-05, + "loss": 0.5534, + "step": 7506 + }, + { + "epoch": 0.23677653366976817, + "grad_norm": 1.8087120056152344, + "learning_rate": 1.7841146099292228e-05, + "loss": 0.573, + "step": 7507 + }, + { + "epoch": 0.2368080744362088, + "grad_norm": 1.7981199026107788, + "learning_rate": 1.7840512064031977e-05, + "loss": 0.5564, + "step": 7508 + }, + { + "epoch": 0.23683961520264943, + "grad_norm": 2.1655635833740234, + "learning_rate": 1.7839877946949894e-05, + "loss": 0.5617, + "step": 7509 + }, + { + "epoch": 0.23687115596909006, + "grad_norm": 2.371670961380005, + "learning_rate": 1.78392437480526e-05, + "loss": 0.5497, + "step": 7510 + }, + { + "epoch": 0.23690269673553066, + "grad_norm": 1.6978586912155151, + "learning_rate": 1.7838609467346705e-05, + "loss": 0.5653, + "step": 7511 + }, + { + "epoch": 0.2369342375019713, + "grad_norm": 1.648612380027771, + "learning_rate": 1.7837975104838832e-05, + "loss": 0.5859, + "step": 7512 + }, + { + "epoch": 0.23696577826841192, + "grad_norm": 1.8786429166793823, + "learning_rate": 1.783734066053561e-05, + "loss": 0.8799, + "step": 7513 + }, + { + "epoch": 0.23699731903485255, + "grad_norm": 3.017174005508423, + "learning_rate": 1.783670613444365e-05, + "loss": 0.5394, + "step": 7514 + }, + { + "epoch": 0.23702885980129318, + "grad_norm": 2.0630457401275635, + "learning_rate": 1.7836071526569573e-05, + "loss": 0.5564, + "step": 7515 + }, + { + "epoch": 0.2370604005677338, + "grad_norm": 2.5932505130767822, + "learning_rate": 1.7835436836920008e-05, + "loss": 0.5695, + "step": 7516 + }, + { + "epoch": 0.2370919413341744, + "grad_norm": 1.8701282739639282, + "learning_rate": 1.7834802065501576e-05, + "loss": 0.5473, + "step": 7517 + }, + { + "epoch": 0.23712348210061504, + "grad_norm": 2.1475672721862793, + "learning_rate": 1.78341672123209e-05, + "loss": 0.4314, + "step": 7518 + }, + { + "epoch": 0.23715502286705567, + "grad_norm": 2.6093177795410156, + "learning_rate": 1.7833532277384607e-05, + "loss": 0.5467, + "step": 7519 + }, + { + "epoch": 0.2371865636334963, + "grad_norm": 2.5207786560058594, + "learning_rate": 1.783289726069932e-05, + "loss": 0.5577, + "step": 7520 + }, + { + "epoch": 0.23721810439993693, + "grad_norm": 1.668636679649353, + "learning_rate": 1.783226216227167e-05, + "loss": 0.555, + "step": 7521 + }, + { + "epoch": 0.23724964516637753, + "grad_norm": 1.7168896198272705, + "learning_rate": 1.7831626982108283e-05, + "loss": 0.8557, + "step": 7522 + }, + { + "epoch": 0.23728118593281816, + "grad_norm": 2.0138442516326904, + "learning_rate": 1.7830991720215786e-05, + "loss": 0.541, + "step": 7523 + }, + { + "epoch": 0.2373127266992588, + "grad_norm": 2.4989981651306152, + "learning_rate": 1.783035637660081e-05, + "loss": 0.5932, + "step": 7524 + }, + { + "epoch": 0.23734426746569942, + "grad_norm": 2.970391273498535, + "learning_rate": 1.7829720951269985e-05, + "loss": 0.5785, + "step": 7525 + }, + { + "epoch": 0.23737580823214005, + "grad_norm": 2.2103540897369385, + "learning_rate": 1.7829085444229943e-05, + "loss": 0.5856, + "step": 7526 + }, + { + "epoch": 0.23740734899858065, + "grad_norm": 2.149707317352295, + "learning_rate": 1.7828449855487316e-05, + "loss": 0.5699, + "step": 7527 + }, + { + "epoch": 0.23743888976502128, + "grad_norm": 1.606632113456726, + "learning_rate": 1.7827814185048734e-05, + "loss": 0.5355, + "step": 7528 + }, + { + "epoch": 0.2374704305314619, + "grad_norm": 2.2539899349212646, + "learning_rate": 1.7827178432920834e-05, + "loss": 0.5762, + "step": 7529 + }, + { + "epoch": 0.23750197129790254, + "grad_norm": 1.8473730087280273, + "learning_rate": 1.782654259911025e-05, + "loss": 0.5654, + "step": 7530 + }, + { + "epoch": 0.23753351206434317, + "grad_norm": 6.0869975090026855, + "learning_rate": 1.7825906683623617e-05, + "loss": 0.8757, + "step": 7531 + }, + { + "epoch": 0.2375650528307838, + "grad_norm": 1.9464075565338135, + "learning_rate": 1.7825270686467567e-05, + "loss": 0.5366, + "step": 7532 + }, + { + "epoch": 0.2375965935972244, + "grad_norm": 2.6449368000030518, + "learning_rate": 1.7824634607648746e-05, + "loss": 0.5737, + "step": 7533 + }, + { + "epoch": 0.23762813436366503, + "grad_norm": 1.6742500066757202, + "learning_rate": 1.7823998447173787e-05, + "loss": 0.4259, + "step": 7534 + }, + { + "epoch": 0.23765967513010566, + "grad_norm": 2.353454351425171, + "learning_rate": 1.7823362205049325e-05, + "loss": 0.5776, + "step": 7535 + }, + { + "epoch": 0.2376912158965463, + "grad_norm": 1.7959951162338257, + "learning_rate": 1.782272588128201e-05, + "loss": 0.3992, + "step": 7536 + }, + { + "epoch": 0.23772275666298692, + "grad_norm": 1.9036751985549927, + "learning_rate": 1.782208947587847e-05, + "loss": 0.5202, + "step": 7537 + }, + { + "epoch": 0.23775429742942752, + "grad_norm": 2.424788475036621, + "learning_rate": 1.7821452988845356e-05, + "loss": 0.5677, + "step": 7538 + }, + { + "epoch": 0.23778583819586815, + "grad_norm": 2.611229419708252, + "learning_rate": 1.7820816420189305e-05, + "loss": 0.5475, + "step": 7539 + }, + { + "epoch": 0.23781737896230878, + "grad_norm": 2.639796495437622, + "learning_rate": 1.7820179769916962e-05, + "loss": 0.5662, + "step": 7540 + }, + { + "epoch": 0.23784891972874941, + "grad_norm": 2.9862382411956787, + "learning_rate": 1.7819543038034973e-05, + "loss": 0.5304, + "step": 7541 + }, + { + "epoch": 0.23788046049519004, + "grad_norm": 2.2380526065826416, + "learning_rate": 1.7818906224549982e-05, + "loss": 0.5529, + "step": 7542 + }, + { + "epoch": 0.23791200126163065, + "grad_norm": 4.8113226890563965, + "learning_rate": 1.781826932946863e-05, + "loss": 0.5277, + "step": 7543 + }, + { + "epoch": 0.23794354202807128, + "grad_norm": 3.048297166824341, + "learning_rate": 1.7817632352797565e-05, + "loss": 0.9064, + "step": 7544 + }, + { + "epoch": 0.2379750827945119, + "grad_norm": 2.453402280807495, + "learning_rate": 1.781699529454344e-05, + "loss": 0.5591, + "step": 7545 + }, + { + "epoch": 0.23800662356095253, + "grad_norm": 2.338012933731079, + "learning_rate": 1.78163581547129e-05, + "loss": 0.5425, + "step": 7546 + }, + { + "epoch": 0.23803816432739316, + "grad_norm": 1.7374682426452637, + "learning_rate": 1.781572093331259e-05, + "loss": 0.5367, + "step": 7547 + }, + { + "epoch": 0.23806970509383377, + "grad_norm": 1.6676089763641357, + "learning_rate": 1.7815083630349165e-05, + "loss": 0.9197, + "step": 7548 + }, + { + "epoch": 0.2381012458602744, + "grad_norm": 1.6535682678222656, + "learning_rate": 1.7814446245829272e-05, + "loss": 0.4087, + "step": 7549 + }, + { + "epoch": 0.23813278662671503, + "grad_norm": 2.7353174686431885, + "learning_rate": 1.781380877975957e-05, + "loss": 0.5507, + "step": 7550 + }, + { + "epoch": 0.23816432739315566, + "grad_norm": 2.095764636993408, + "learning_rate": 1.78131712321467e-05, + "loss": 0.5934, + "step": 7551 + }, + { + "epoch": 0.23819586815959629, + "grad_norm": 2.42132830619812, + "learning_rate": 1.7812533602997327e-05, + "loss": 0.854, + "step": 7552 + }, + { + "epoch": 0.23822740892603692, + "grad_norm": 1.9665889739990234, + "learning_rate": 1.7811895892318098e-05, + "loss": 0.584, + "step": 7553 + }, + { + "epoch": 0.23825894969247752, + "grad_norm": 2.1037755012512207, + "learning_rate": 1.7811258100115668e-05, + "loss": 0.5744, + "step": 7554 + }, + { + "epoch": 0.23829049045891815, + "grad_norm": 2.1833434104919434, + "learning_rate": 1.7810620226396694e-05, + "loss": 0.5595, + "step": 7555 + }, + { + "epoch": 0.23832203122535878, + "grad_norm": 3.467421054840088, + "learning_rate": 1.780998227116783e-05, + "loss": 0.5271, + "step": 7556 + }, + { + "epoch": 0.2383535719917994, + "grad_norm": 2.0766491889953613, + "learning_rate": 1.7809344234435744e-05, + "loss": 0.5323, + "step": 7557 + }, + { + "epoch": 0.23838511275824004, + "grad_norm": 2.2749061584472656, + "learning_rate": 1.7808706116207083e-05, + "loss": 0.5917, + "step": 7558 + }, + { + "epoch": 0.23841665352468064, + "grad_norm": 2.204500198364258, + "learning_rate": 1.7808067916488513e-05, + "loss": 0.5695, + "step": 7559 + }, + { + "epoch": 0.23844819429112127, + "grad_norm": 1.6063296794891357, + "learning_rate": 1.780742963528669e-05, + "loss": 0.5525, + "step": 7560 + }, + { + "epoch": 0.2384797350575619, + "grad_norm": 1.6455187797546387, + "learning_rate": 1.7806791272608276e-05, + "loss": 0.5833, + "step": 7561 + }, + { + "epoch": 0.23851127582400253, + "grad_norm": 1.7892093658447266, + "learning_rate": 1.780615282845993e-05, + "loss": 0.5445, + "step": 7562 + }, + { + "epoch": 0.23854281659044316, + "grad_norm": 2.200371265411377, + "learning_rate": 1.7805514302848325e-05, + "loss": 0.528, + "step": 7563 + }, + { + "epoch": 0.23857435735688376, + "grad_norm": 1.9229700565338135, + "learning_rate": 1.780487569578011e-05, + "loss": 0.584, + "step": 7564 + }, + { + "epoch": 0.2386058981233244, + "grad_norm": 2.900214433670044, + "learning_rate": 1.7804237007261962e-05, + "loss": 0.5776, + "step": 7565 + }, + { + "epoch": 0.23863743888976502, + "grad_norm": 2.098271369934082, + "learning_rate": 1.780359823730054e-05, + "loss": 0.5605, + "step": 7566 + }, + { + "epoch": 0.23866897965620565, + "grad_norm": 1.7454988956451416, + "learning_rate": 1.7802959385902512e-05, + "loss": 0.5858, + "step": 7567 + }, + { + "epoch": 0.23870052042264628, + "grad_norm": 1.4813904762268066, + "learning_rate": 1.780232045307454e-05, + "loss": 0.5692, + "step": 7568 + }, + { + "epoch": 0.2387320611890869, + "grad_norm": 1.9412654638290405, + "learning_rate": 1.7801681438823298e-05, + "loss": 0.5597, + "step": 7569 + }, + { + "epoch": 0.2387636019555275, + "grad_norm": 2.298454523086548, + "learning_rate": 1.7801042343155455e-05, + "loss": 0.5287, + "step": 7570 + }, + { + "epoch": 0.23879514272196814, + "grad_norm": 2.0105628967285156, + "learning_rate": 1.7800403166077673e-05, + "loss": 0.4145, + "step": 7571 + }, + { + "epoch": 0.23882668348840877, + "grad_norm": 1.7149430513381958, + "learning_rate": 1.779976390759663e-05, + "loss": 0.5319, + "step": 7572 + }, + { + "epoch": 0.2388582242548494, + "grad_norm": 2.1601192951202393, + "learning_rate": 1.7799124567718994e-05, + "loss": 0.5713, + "step": 7573 + }, + { + "epoch": 0.23888976502129003, + "grad_norm": 2.0146567821502686, + "learning_rate": 1.7798485146451436e-05, + "loss": 0.5357, + "step": 7574 + }, + { + "epoch": 0.23892130578773063, + "grad_norm": 1.7866477966308594, + "learning_rate": 1.779784564380063e-05, + "loss": 0.5641, + "step": 7575 + }, + { + "epoch": 0.23895284655417126, + "grad_norm": 1.9414012432098389, + "learning_rate": 1.7797206059773254e-05, + "loss": 0.5029, + "step": 7576 + }, + { + "epoch": 0.2389843873206119, + "grad_norm": 2.4220340251922607, + "learning_rate": 1.7796566394375974e-05, + "loss": 0.5873, + "step": 7577 + }, + { + "epoch": 0.23901592808705252, + "grad_norm": 1.7579927444458008, + "learning_rate": 1.779592664761547e-05, + "loss": 0.5642, + "step": 7578 + }, + { + "epoch": 0.23904746885349315, + "grad_norm": 2.138303756713867, + "learning_rate": 1.779528681949842e-05, + "loss": 0.5517, + "step": 7579 + }, + { + "epoch": 0.23907900961993375, + "grad_norm": 1.7731865644454956, + "learning_rate": 1.77946469100315e-05, + "loss": 0.8884, + "step": 7580 + }, + { + "epoch": 0.23911055038637438, + "grad_norm": 2.0085413455963135, + "learning_rate": 1.7794006919221384e-05, + "loss": 0.569, + "step": 7581 + }, + { + "epoch": 0.239142091152815, + "grad_norm": 1.9764504432678223, + "learning_rate": 1.7793366847074755e-05, + "loss": 0.5704, + "step": 7582 + }, + { + "epoch": 0.23917363191925564, + "grad_norm": 1.826699137687683, + "learning_rate": 1.7792726693598293e-05, + "loss": 0.5548, + "step": 7583 + }, + { + "epoch": 0.23920517268569627, + "grad_norm": 1.649519681930542, + "learning_rate": 1.7792086458798676e-05, + "loss": 0.5384, + "step": 7584 + }, + { + "epoch": 0.23923671345213687, + "grad_norm": 2.4263486862182617, + "learning_rate": 1.7791446142682588e-05, + "loss": 0.5654, + "step": 7585 + }, + { + "epoch": 0.2392682542185775, + "grad_norm": 9.801230430603027, + "learning_rate": 1.7790805745256703e-05, + "loss": 0.8411, + "step": 7586 + }, + { + "epoch": 0.23929979498501813, + "grad_norm": 2.7991368770599365, + "learning_rate": 1.7790165266527718e-05, + "loss": 0.6006, + "step": 7587 + }, + { + "epoch": 0.23933133575145876, + "grad_norm": 2.1097919940948486, + "learning_rate": 1.778952470650231e-05, + "loss": 0.5457, + "step": 7588 + }, + { + "epoch": 0.2393628765178994, + "grad_norm": 1.96624755859375, + "learning_rate": 1.778888406518716e-05, + "loss": 0.5715, + "step": 7589 + }, + { + "epoch": 0.23939441728434002, + "grad_norm": 2.520171642303467, + "learning_rate": 1.7788243342588955e-05, + "loss": 0.5682, + "step": 7590 + }, + { + "epoch": 0.23942595805078062, + "grad_norm": 2.530559778213501, + "learning_rate": 1.7787602538714385e-05, + "loss": 0.4477, + "step": 7591 + }, + { + "epoch": 0.23945749881722125, + "grad_norm": 1.9519025087356567, + "learning_rate": 1.7786961653570135e-05, + "loss": 0.5617, + "step": 7592 + }, + { + "epoch": 0.23948903958366188, + "grad_norm": 1.7902652025222778, + "learning_rate": 1.7786320687162896e-05, + "loss": 0.5499, + "step": 7593 + }, + { + "epoch": 0.2395205803501025, + "grad_norm": 1.3179339170455933, + "learning_rate": 1.7785679639499352e-05, + "loss": 0.3702, + "step": 7594 + }, + { + "epoch": 0.23955212111654314, + "grad_norm": 1.90031099319458, + "learning_rate": 1.77850385105862e-05, + "loss": 0.5686, + "step": 7595 + }, + { + "epoch": 0.23958366188298374, + "grad_norm": 2.0706303119659424, + "learning_rate": 1.7784397300430122e-05, + "loss": 0.5532, + "step": 7596 + }, + { + "epoch": 0.23961520264942437, + "grad_norm": 2.2816946506500244, + "learning_rate": 1.7783756009037815e-05, + "loss": 0.5558, + "step": 7597 + }, + { + "epoch": 0.239646743415865, + "grad_norm": 1.9841976165771484, + "learning_rate": 1.7783114636415973e-05, + "loss": 0.5659, + "step": 7598 + }, + { + "epoch": 0.23967828418230563, + "grad_norm": 1.996941089630127, + "learning_rate": 1.778247318257128e-05, + "loss": 0.5346, + "step": 7599 + }, + { + "epoch": 0.23970982494874626, + "grad_norm": 1.7867324352264404, + "learning_rate": 1.7781831647510445e-05, + "loss": 0.8741, + "step": 7600 + }, + { + "epoch": 0.23974136571518687, + "grad_norm": 2.1039204597473145, + "learning_rate": 1.7781190031240148e-05, + "loss": 0.5836, + "step": 7601 + }, + { + "epoch": 0.2397729064816275, + "grad_norm": 1.6062761545181274, + "learning_rate": 1.778054833376709e-05, + "loss": 0.5532, + "step": 7602 + }, + { + "epoch": 0.23980444724806813, + "grad_norm": 1.6072887182235718, + "learning_rate": 1.7779906555097972e-05, + "loss": 0.5407, + "step": 7603 + }, + { + "epoch": 0.23983598801450876, + "grad_norm": 2.5474820137023926, + "learning_rate": 1.7779264695239486e-05, + "loss": 0.5658, + "step": 7604 + }, + { + "epoch": 0.23986752878094939, + "grad_norm": 1.6094805002212524, + "learning_rate": 1.7778622754198336e-05, + "loss": 0.5874, + "step": 7605 + }, + { + "epoch": 0.23989906954739001, + "grad_norm": 2.021195411682129, + "learning_rate": 1.777798073198121e-05, + "loss": 0.5358, + "step": 7606 + }, + { + "epoch": 0.23993061031383062, + "grad_norm": 2.408769369125366, + "learning_rate": 1.7777338628594826e-05, + "loss": 0.5824, + "step": 7607 + }, + { + "epoch": 0.23996215108027125, + "grad_norm": 1.7442902326583862, + "learning_rate": 1.777669644404587e-05, + "loss": 0.5762, + "step": 7608 + }, + { + "epoch": 0.23999369184671188, + "grad_norm": 1.6403828859329224, + "learning_rate": 1.7776054178341047e-05, + "loss": 0.9075, + "step": 7609 + }, + { + "epoch": 0.2400252326131525, + "grad_norm": 1.906215786933899, + "learning_rate": 1.777541183148706e-05, + "loss": 0.5642, + "step": 7610 + }, + { + "epoch": 0.24005677337959314, + "grad_norm": 1.8300422430038452, + "learning_rate": 1.7774769403490616e-05, + "loss": 0.5418, + "step": 7611 + }, + { + "epoch": 0.24008831414603374, + "grad_norm": 2.4759013652801514, + "learning_rate": 1.7774126894358414e-05, + "loss": 0.57, + "step": 7612 + }, + { + "epoch": 0.24011985491247437, + "grad_norm": 1.614804744720459, + "learning_rate": 1.7773484304097162e-05, + "loss": 0.555, + "step": 7613 + }, + { + "epoch": 0.240151395678915, + "grad_norm": 2.6141910552978516, + "learning_rate": 1.7772841632713563e-05, + "loss": 0.5806, + "step": 7614 + }, + { + "epoch": 0.24018293644535563, + "grad_norm": 2.0469329357147217, + "learning_rate": 1.777219888021433e-05, + "loss": 0.5442, + "step": 7615 + }, + { + "epoch": 0.24021447721179626, + "grad_norm": 1.2674307823181152, + "learning_rate": 1.7771556046606166e-05, + "loss": 0.4361, + "step": 7616 + }, + { + "epoch": 0.24024601797823686, + "grad_norm": 1.8114863634109497, + "learning_rate": 1.7770913131895778e-05, + "loss": 0.5639, + "step": 7617 + }, + { + "epoch": 0.2402775587446775, + "grad_norm": 1.6786577701568604, + "learning_rate": 1.777027013608988e-05, + "loss": 0.3995, + "step": 7618 + }, + { + "epoch": 0.24030909951111812, + "grad_norm": 1.5193606615066528, + "learning_rate": 1.776962705919518e-05, + "loss": 0.8589, + "step": 7619 + }, + { + "epoch": 0.24034064027755875, + "grad_norm": 1.9040254354476929, + "learning_rate": 1.7768983901218386e-05, + "loss": 0.8612, + "step": 7620 + }, + { + "epoch": 0.24037218104399938, + "grad_norm": 1.7723478078842163, + "learning_rate": 1.776834066216621e-05, + "loss": 0.5451, + "step": 7621 + }, + { + "epoch": 0.24040372181044, + "grad_norm": 2.375779151916504, + "learning_rate": 1.7767697342045376e-05, + "loss": 0.5903, + "step": 7622 + }, + { + "epoch": 0.2404352625768806, + "grad_norm": 1.926307201385498, + "learning_rate": 1.776705394086258e-05, + "loss": 0.5615, + "step": 7623 + }, + { + "epoch": 0.24046680334332124, + "grad_norm": 1.918005347251892, + "learning_rate": 1.7766410458624553e-05, + "loss": 0.5569, + "step": 7624 + }, + { + "epoch": 0.24049834410976187, + "grad_norm": 1.531197190284729, + "learning_rate": 1.7765766895338e-05, + "loss": 0.8654, + "step": 7625 + }, + { + "epoch": 0.2405298848762025, + "grad_norm": 1.8777015209197998, + "learning_rate": 1.7765123251009635e-05, + "loss": 0.5512, + "step": 7626 + }, + { + "epoch": 0.24056142564264313, + "grad_norm": 1.7359014749526978, + "learning_rate": 1.7764479525646182e-05, + "loss": 0.4098, + "step": 7627 + }, + { + "epoch": 0.24059296640908373, + "grad_norm": 4.518047332763672, + "learning_rate": 1.7763835719254358e-05, + "loss": 0.5739, + "step": 7628 + }, + { + "epoch": 0.24062450717552436, + "grad_norm": 2.0030758380889893, + "learning_rate": 1.7763191831840874e-05, + "loss": 0.5777, + "step": 7629 + }, + { + "epoch": 0.240656047941965, + "grad_norm": 1.7798492908477783, + "learning_rate": 1.7762547863412464e-05, + "loss": 0.8653, + "step": 7630 + }, + { + "epoch": 0.24068758870840562, + "grad_norm": 2.876857042312622, + "learning_rate": 1.7761903813975835e-05, + "loss": 0.8661, + "step": 7631 + }, + { + "epoch": 0.24071912947484625, + "grad_norm": 1.9772282838821411, + "learning_rate": 1.776125968353771e-05, + "loss": 0.8994, + "step": 7632 + }, + { + "epoch": 0.24075067024128685, + "grad_norm": 1.7052255868911743, + "learning_rate": 1.776061547210482e-05, + "loss": 0.5587, + "step": 7633 + }, + { + "epoch": 0.24078221100772748, + "grad_norm": 1.7399749755859375, + "learning_rate": 1.7759971179683874e-05, + "loss": 0.5461, + "step": 7634 + }, + { + "epoch": 0.2408137517741681, + "grad_norm": 2.831836700439453, + "learning_rate": 1.775932680628161e-05, + "loss": 0.5726, + "step": 7635 + }, + { + "epoch": 0.24084529254060874, + "grad_norm": 2.2921981811523438, + "learning_rate": 1.7758682351904745e-05, + "loss": 0.5623, + "step": 7636 + }, + { + "epoch": 0.24087683330704937, + "grad_norm": 1.7051656246185303, + "learning_rate": 1.775803781656e-05, + "loss": 0.9001, + "step": 7637 + }, + { + "epoch": 0.24090837407348997, + "grad_norm": 2.515727996826172, + "learning_rate": 1.775739320025411e-05, + "loss": 0.5641, + "step": 7638 + }, + { + "epoch": 0.2409399148399306, + "grad_norm": 1.5469732284545898, + "learning_rate": 1.77567485029938e-05, + "loss": 0.906, + "step": 7639 + }, + { + "epoch": 0.24097145560637123, + "grad_norm": 2.429395914077759, + "learning_rate": 1.7756103724785797e-05, + "loss": 0.5368, + "step": 7640 + }, + { + "epoch": 0.24100299637281186, + "grad_norm": 3.9875779151916504, + "learning_rate": 1.7755458865636825e-05, + "loss": 0.5492, + "step": 7641 + }, + { + "epoch": 0.2410345371392525, + "grad_norm": 2.6423237323760986, + "learning_rate": 1.775481392555362e-05, + "loss": 0.5815, + "step": 7642 + }, + { + "epoch": 0.24106607790569312, + "grad_norm": 2.02535343170166, + "learning_rate": 1.775416890454291e-05, + "loss": 0.5126, + "step": 7643 + }, + { + "epoch": 0.24109761867213372, + "grad_norm": 1.6276297569274902, + "learning_rate": 1.7753523802611424e-05, + "loss": 0.5428, + "step": 7644 + }, + { + "epoch": 0.24112915943857435, + "grad_norm": 2.653106451034546, + "learning_rate": 1.77528786197659e-05, + "loss": 0.3722, + "step": 7645 + }, + { + "epoch": 0.24116070020501498, + "grad_norm": 1.8348585367202759, + "learning_rate": 1.775223335601307e-05, + "loss": 0.3819, + "step": 7646 + }, + { + "epoch": 0.2411922409714556, + "grad_norm": 2.4592506885528564, + "learning_rate": 1.775158801135966e-05, + "loss": 0.5731, + "step": 7647 + }, + { + "epoch": 0.24122378173789624, + "grad_norm": 1.945088505744934, + "learning_rate": 1.7750942585812413e-05, + "loss": 0.5481, + "step": 7648 + }, + { + "epoch": 0.24125532250433684, + "grad_norm": 2.41514253616333, + "learning_rate": 1.775029707937806e-05, + "loss": 0.5557, + "step": 7649 + }, + { + "epoch": 0.24128686327077747, + "grad_norm": 2.041881799697876, + "learning_rate": 1.774965149206334e-05, + "loss": 0.5531, + "step": 7650 + }, + { + "epoch": 0.2413184040372181, + "grad_norm": 2.004340410232544, + "learning_rate": 1.774900582387499e-05, + "loss": 0.544, + "step": 7651 + }, + { + "epoch": 0.24134994480365873, + "grad_norm": 2.0685300827026367, + "learning_rate": 1.7748360074819747e-05, + "loss": 0.5502, + "step": 7652 + }, + { + "epoch": 0.24138148557009936, + "grad_norm": 8.169626235961914, + "learning_rate": 1.7747714244904348e-05, + "loss": 0.5841, + "step": 7653 + }, + { + "epoch": 0.24141302633653997, + "grad_norm": 2.763098955154419, + "learning_rate": 1.774706833413554e-05, + "loss": 0.4059, + "step": 7654 + }, + { + "epoch": 0.2414445671029806, + "grad_norm": 2.237963914871216, + "learning_rate": 1.7746422342520056e-05, + "loss": 0.5809, + "step": 7655 + }, + { + "epoch": 0.24147610786942122, + "grad_norm": 2.942135810852051, + "learning_rate": 1.7745776270064637e-05, + "loss": 0.5573, + "step": 7656 + }, + { + "epoch": 0.24150764863586185, + "grad_norm": 1.841801643371582, + "learning_rate": 1.7745130116776034e-05, + "loss": 0.5625, + "step": 7657 + }, + { + "epoch": 0.24153918940230248, + "grad_norm": 2.6253621578216553, + "learning_rate": 1.7744483882660975e-05, + "loss": 0.5551, + "step": 7658 + }, + { + "epoch": 0.24157073016874311, + "grad_norm": 2.3995094299316406, + "learning_rate": 1.774383756772622e-05, + "loss": 0.3962, + "step": 7659 + }, + { + "epoch": 0.24160227093518372, + "grad_norm": 2.1537973880767822, + "learning_rate": 1.7743191171978505e-05, + "loss": 0.5582, + "step": 7660 + }, + { + "epoch": 0.24163381170162435, + "grad_norm": 2.4545395374298096, + "learning_rate": 1.774254469542458e-05, + "loss": 0.5777, + "step": 7661 + }, + { + "epoch": 0.24166535246806498, + "grad_norm": 1.7466769218444824, + "learning_rate": 1.774189813807119e-05, + "loss": 0.8878, + "step": 7662 + }, + { + "epoch": 0.2416968932345056, + "grad_norm": 2.9159717559814453, + "learning_rate": 1.7741251499925075e-05, + "loss": 0.5356, + "step": 7663 + }, + { + "epoch": 0.24172843400094624, + "grad_norm": 2.2466306686401367, + "learning_rate": 1.7740604780993e-05, + "loss": 0.5516, + "step": 7664 + }, + { + "epoch": 0.24175997476738684, + "grad_norm": 2.5420966148376465, + "learning_rate": 1.7739957981281695e-05, + "loss": 0.5446, + "step": 7665 + }, + { + "epoch": 0.24179151553382747, + "grad_norm": 2.2358875274658203, + "learning_rate": 1.7739311100797922e-05, + "loss": 0.5227, + "step": 7666 + }, + { + "epoch": 0.2418230563002681, + "grad_norm": 4.703837871551514, + "learning_rate": 1.7738664139548426e-05, + "loss": 0.5605, + "step": 7667 + }, + { + "epoch": 0.24185459706670873, + "grad_norm": 2.316227436065674, + "learning_rate": 1.7738017097539962e-05, + "loss": 0.5663, + "step": 7668 + }, + { + "epoch": 0.24188613783314936, + "grad_norm": 3.636686086654663, + "learning_rate": 1.773736997477928e-05, + "loss": 0.5735, + "step": 7669 + }, + { + "epoch": 0.24191767859958996, + "grad_norm": 2.6475958824157715, + "learning_rate": 1.7736722771273137e-05, + "loss": 0.5706, + "step": 7670 + }, + { + "epoch": 0.2419492193660306, + "grad_norm": 3.0840489864349365, + "learning_rate": 1.7736075487028284e-05, + "loss": 0.5747, + "step": 7671 + }, + { + "epoch": 0.24198076013247122, + "grad_norm": 2.2621371746063232, + "learning_rate": 1.7735428122051477e-05, + "loss": 0.5434, + "step": 7672 + }, + { + "epoch": 0.24201230089891185, + "grad_norm": 1.922281265258789, + "learning_rate": 1.773478067634947e-05, + "loss": 0.5692, + "step": 7673 + }, + { + "epoch": 0.24204384166535248, + "grad_norm": 2.2098705768585205, + "learning_rate": 1.7734133149929016e-05, + "loss": 0.9095, + "step": 7674 + }, + { + "epoch": 0.24207538243179308, + "grad_norm": 1.9983198642730713, + "learning_rate": 1.7733485542796884e-05, + "loss": 0.5617, + "step": 7675 + }, + { + "epoch": 0.2421069231982337, + "grad_norm": 2.3646347522735596, + "learning_rate": 1.7732837854959822e-05, + "loss": 0.5296, + "step": 7676 + }, + { + "epoch": 0.24213846396467434, + "grad_norm": 2.420684576034546, + "learning_rate": 1.7732190086424593e-05, + "loss": 0.5722, + "step": 7677 + }, + { + "epoch": 0.24217000473111497, + "grad_norm": 2.107970952987671, + "learning_rate": 1.773154223719796e-05, + "loss": 0.5737, + "step": 7678 + }, + { + "epoch": 0.2422015454975556, + "grad_norm": 2.1304123401641846, + "learning_rate": 1.7730894307286677e-05, + "loss": 0.5308, + "step": 7679 + }, + { + "epoch": 0.24223308626399623, + "grad_norm": 2.186567783355713, + "learning_rate": 1.7730246296697512e-05, + "loss": 0.5887, + "step": 7680 + }, + { + "epoch": 0.24226462703043683, + "grad_norm": 3.0258994102478027, + "learning_rate": 1.772959820543722e-05, + "loss": 0.5464, + "step": 7681 + }, + { + "epoch": 0.24229616779687746, + "grad_norm": 2.3332104682922363, + "learning_rate": 1.772895003351257e-05, + "loss": 0.5651, + "step": 7682 + }, + { + "epoch": 0.2423277085633181, + "grad_norm": 2.5194718837738037, + "learning_rate": 1.7728301780930327e-05, + "loss": 0.5746, + "step": 7683 + }, + { + "epoch": 0.24235924932975872, + "grad_norm": 2.8911240100860596, + "learning_rate": 1.7727653447697256e-05, + "loss": 0.5623, + "step": 7684 + }, + { + "epoch": 0.24239079009619935, + "grad_norm": 2.531796932220459, + "learning_rate": 1.7727005033820117e-05, + "loss": 0.5405, + "step": 7685 + }, + { + "epoch": 0.24242233086263995, + "grad_norm": 2.4644980430603027, + "learning_rate": 1.772635653930568e-05, + "loss": 0.56, + "step": 7686 + }, + { + "epoch": 0.24245387162908058, + "grad_norm": 2.0818748474121094, + "learning_rate": 1.7725707964160715e-05, + "loss": 0.5266, + "step": 7687 + }, + { + "epoch": 0.2424854123955212, + "grad_norm": 2.1019630432128906, + "learning_rate": 1.772505930839199e-05, + "loss": 0.5457, + "step": 7688 + }, + { + "epoch": 0.24251695316196184, + "grad_norm": 3.0359747409820557, + "learning_rate": 1.772441057200627e-05, + "loss": 0.5665, + "step": 7689 + }, + { + "epoch": 0.24254849392840247, + "grad_norm": 2.0764763355255127, + "learning_rate": 1.772376175501033e-05, + "loss": 0.5277, + "step": 7690 + }, + { + "epoch": 0.24258003469484307, + "grad_norm": 2.248363971710205, + "learning_rate": 1.7723112857410936e-05, + "loss": 0.5456, + "step": 7691 + }, + { + "epoch": 0.2426115754612837, + "grad_norm": 2.780001163482666, + "learning_rate": 1.7722463879214864e-05, + "loss": 0.5586, + "step": 7692 + }, + { + "epoch": 0.24264311622772433, + "grad_norm": 2.0818636417388916, + "learning_rate": 1.7721814820428888e-05, + "loss": 0.5672, + "step": 7693 + }, + { + "epoch": 0.24267465699416496, + "grad_norm": 2.6801857948303223, + "learning_rate": 1.7721165681059772e-05, + "loss": 0.5697, + "step": 7694 + }, + { + "epoch": 0.2427061977606056, + "grad_norm": 4.331172466278076, + "learning_rate": 1.7720516461114302e-05, + "loss": 0.5642, + "step": 7695 + }, + { + "epoch": 0.24273773852704622, + "grad_norm": 2.076033592224121, + "learning_rate": 1.7719867160599245e-05, + "loss": 0.5733, + "step": 7696 + }, + { + "epoch": 0.24276927929348682, + "grad_norm": 1.859954833984375, + "learning_rate": 1.7719217779521383e-05, + "loss": 0.5401, + "step": 7697 + }, + { + "epoch": 0.24280082005992745, + "grad_norm": 2.074827194213867, + "learning_rate": 1.7718568317887487e-05, + "loss": 0.3888, + "step": 7698 + }, + { + "epoch": 0.24283236082636808, + "grad_norm": 5.656703948974609, + "learning_rate": 1.7717918775704338e-05, + "loss": 0.8931, + "step": 7699 + }, + { + "epoch": 0.2428639015928087, + "grad_norm": 2.5106372833251953, + "learning_rate": 1.7717269152978717e-05, + "loss": 0.5275, + "step": 7700 + }, + { + "epoch": 0.24289544235924934, + "grad_norm": 1.9490312337875366, + "learning_rate": 1.7716619449717396e-05, + "loss": 0.5802, + "step": 7701 + }, + { + "epoch": 0.24292698312568994, + "grad_norm": 2.255788564682007, + "learning_rate": 1.771596966592716e-05, + "loss": 0.5239, + "step": 7702 + }, + { + "epoch": 0.24295852389213057, + "grad_norm": 5.874750137329102, + "learning_rate": 1.7715319801614788e-05, + "loss": 0.4939, + "step": 7703 + }, + { + "epoch": 0.2429900646585712, + "grad_norm": 3.291675567626953, + "learning_rate": 1.7714669856787067e-05, + "loss": 0.5865, + "step": 7704 + }, + { + "epoch": 0.24302160542501183, + "grad_norm": 2.4166643619537354, + "learning_rate": 1.7714019831450772e-05, + "loss": 0.5733, + "step": 7705 + }, + { + "epoch": 0.24305314619145246, + "grad_norm": 3.4902989864349365, + "learning_rate": 1.7713369725612693e-05, + "loss": 0.5745, + "step": 7706 + }, + { + "epoch": 0.24308468695789306, + "grad_norm": 2.3956313133239746, + "learning_rate": 1.771271953927961e-05, + "loss": 0.554, + "step": 7707 + }, + { + "epoch": 0.2431162277243337, + "grad_norm": 2.7458178997039795, + "learning_rate": 1.7712069272458306e-05, + "loss": 0.586, + "step": 7708 + }, + { + "epoch": 0.24314776849077432, + "grad_norm": 2.1721367835998535, + "learning_rate": 1.7711418925155576e-05, + "loss": 0.8573, + "step": 7709 + }, + { + "epoch": 0.24317930925721495, + "grad_norm": 2.3089189529418945, + "learning_rate": 1.7710768497378203e-05, + "loss": 0.5492, + "step": 7710 + }, + { + "epoch": 0.24321085002365558, + "grad_norm": 4.6520490646362305, + "learning_rate": 1.771011798913297e-05, + "loss": 0.5763, + "step": 7711 + }, + { + "epoch": 0.24324239079009619, + "grad_norm": 3.5116031169891357, + "learning_rate": 1.770946740042667e-05, + "loss": 0.5515, + "step": 7712 + }, + { + "epoch": 0.24327393155653682, + "grad_norm": 2.91115403175354, + "learning_rate": 1.7708816731266094e-05, + "loss": 0.5723, + "step": 7713 + }, + { + "epoch": 0.24330547232297745, + "grad_norm": 2.583162784576416, + "learning_rate": 1.7708165981658028e-05, + "loss": 0.5931, + "step": 7714 + }, + { + "epoch": 0.24333701308941807, + "grad_norm": 3.038163900375366, + "learning_rate": 1.7707515151609262e-05, + "loss": 0.5585, + "step": 7715 + }, + { + "epoch": 0.2433685538558587, + "grad_norm": 2.448251724243164, + "learning_rate": 1.7706864241126593e-05, + "loss": 0.5653, + "step": 7716 + }, + { + "epoch": 0.24340009462229933, + "grad_norm": 2.5954174995422363, + "learning_rate": 1.770621325021681e-05, + "loss": 0.5863, + "step": 7717 + }, + { + "epoch": 0.24343163538873994, + "grad_norm": 2.517263650894165, + "learning_rate": 1.770556217888671e-05, + "loss": 0.5562, + "step": 7718 + }, + { + "epoch": 0.24346317615518057, + "grad_norm": 1.715929388999939, + "learning_rate": 1.7704911027143087e-05, + "loss": 0.5729, + "step": 7719 + }, + { + "epoch": 0.2434947169216212, + "grad_norm": 2.740107297897339, + "learning_rate": 1.7704259794992734e-05, + "loss": 0.5463, + "step": 7720 + }, + { + "epoch": 0.24352625768806183, + "grad_norm": 3.313544511795044, + "learning_rate": 1.7703608482442445e-05, + "loss": 0.4329, + "step": 7721 + }, + { + "epoch": 0.24355779845450246, + "grad_norm": 5.848851680755615, + "learning_rate": 1.770295708949902e-05, + "loss": 0.5629, + "step": 7722 + }, + { + "epoch": 0.24358933922094306, + "grad_norm": 3.3441646099090576, + "learning_rate": 1.7702305616169257e-05, + "loss": 0.5474, + "step": 7723 + }, + { + "epoch": 0.2436208799873837, + "grad_norm": 2.278095006942749, + "learning_rate": 1.7701654062459953e-05, + "loss": 0.5402, + "step": 7724 + }, + { + "epoch": 0.24365242075382432, + "grad_norm": 1.9822264909744263, + "learning_rate": 1.7701002428377912e-05, + "loss": 0.8893, + "step": 7725 + }, + { + "epoch": 0.24368396152026495, + "grad_norm": 1.9615963697433472, + "learning_rate": 1.770035071392993e-05, + "loss": 0.5151, + "step": 7726 + }, + { + "epoch": 0.24371550228670558, + "grad_norm": 1.8791429996490479, + "learning_rate": 1.7699698919122807e-05, + "loss": 0.5426, + "step": 7727 + }, + { + "epoch": 0.24374704305314618, + "grad_norm": 2.4740700721740723, + "learning_rate": 1.769904704396335e-05, + "loss": 0.554, + "step": 7728 + }, + { + "epoch": 0.2437785838195868, + "grad_norm": 2.1510632038116455, + "learning_rate": 1.7698395088458357e-05, + "loss": 0.5786, + "step": 7729 + }, + { + "epoch": 0.24381012458602744, + "grad_norm": 2.5893144607543945, + "learning_rate": 1.7697743052614637e-05, + "loss": 0.5832, + "step": 7730 + }, + { + "epoch": 0.24384166535246807, + "grad_norm": 2.1836137771606445, + "learning_rate": 1.769709093643899e-05, + "loss": 0.5657, + "step": 7731 + }, + { + "epoch": 0.2438732061189087, + "grad_norm": 2.9508919715881348, + "learning_rate": 1.7696438739938217e-05, + "loss": 0.5534, + "step": 7732 + }, + { + "epoch": 0.24390474688534933, + "grad_norm": 2.2212376594543457, + "learning_rate": 1.7695786463119132e-05, + "loss": 0.552, + "step": 7733 + }, + { + "epoch": 0.24393628765178993, + "grad_norm": 2.337574005126953, + "learning_rate": 1.7695134105988544e-05, + "loss": 0.577, + "step": 7734 + }, + { + "epoch": 0.24396782841823056, + "grad_norm": 2.0025291442871094, + "learning_rate": 1.7694481668553255e-05, + "loss": 0.5703, + "step": 7735 + }, + { + "epoch": 0.2439993691846712, + "grad_norm": 2.636854887008667, + "learning_rate": 1.769382915082007e-05, + "loss": 0.4332, + "step": 7736 + }, + { + "epoch": 0.24403090995111182, + "grad_norm": 2.4047906398773193, + "learning_rate": 1.769317655279581e-05, + "loss": 0.5437, + "step": 7737 + }, + { + "epoch": 0.24406245071755245, + "grad_norm": 1.744888186454773, + "learning_rate": 1.7692523874487274e-05, + "loss": 0.8807, + "step": 7738 + }, + { + "epoch": 0.24409399148399305, + "grad_norm": 1.727378487586975, + "learning_rate": 1.7691871115901283e-05, + "loss": 0.5392, + "step": 7739 + }, + { + "epoch": 0.24412553225043368, + "grad_norm": 2.4270544052124023, + "learning_rate": 1.7691218277044643e-05, + "loss": 0.5416, + "step": 7740 + }, + { + "epoch": 0.2441570730168743, + "grad_norm": 3.2748665809631348, + "learning_rate": 1.7690565357924164e-05, + "loss": 0.5556, + "step": 7741 + }, + { + "epoch": 0.24418861378331494, + "grad_norm": 2.8195738792419434, + "learning_rate": 1.768991235854667e-05, + "loss": 0.4287, + "step": 7742 + }, + { + "epoch": 0.24422015454975557, + "grad_norm": 2.168311357498169, + "learning_rate": 1.7689259278918966e-05, + "loss": 0.8722, + "step": 7743 + }, + { + "epoch": 0.24425169531619617, + "grad_norm": 2.8269739151000977, + "learning_rate": 1.768860611904787e-05, + "loss": 0.5985, + "step": 7744 + }, + { + "epoch": 0.2442832360826368, + "grad_norm": 4.58597993850708, + "learning_rate": 1.7687952878940203e-05, + "loss": 0.8944, + "step": 7745 + }, + { + "epoch": 0.24431477684907743, + "grad_norm": 3.209151029586792, + "learning_rate": 1.7687299558602775e-05, + "loss": 0.565, + "step": 7746 + }, + { + "epoch": 0.24434631761551806, + "grad_norm": 2.9719743728637695, + "learning_rate": 1.768664615804241e-05, + "loss": 0.5558, + "step": 7747 + }, + { + "epoch": 0.2443778583819587, + "grad_norm": 2.052536725997925, + "learning_rate": 1.7685992677265916e-05, + "loss": 0.5687, + "step": 7748 + }, + { + "epoch": 0.2444093991483993, + "grad_norm": 2.199345350265503, + "learning_rate": 1.768533911628013e-05, + "loss": 0.558, + "step": 7749 + }, + { + "epoch": 0.24444093991483992, + "grad_norm": 2.0603439807891846, + "learning_rate": 1.768468547509186e-05, + "loss": 0.5535, + "step": 7750 + }, + { + "epoch": 0.24447248068128055, + "grad_norm": 2.1479036808013916, + "learning_rate": 1.7684031753707928e-05, + "loss": 0.5561, + "step": 7751 + }, + { + "epoch": 0.24450402144772118, + "grad_norm": 2.8395113945007324, + "learning_rate": 1.7683377952135158e-05, + "loss": 0.543, + "step": 7752 + }, + { + "epoch": 0.2445355622141618, + "grad_norm": 2.5376648902893066, + "learning_rate": 1.7682724070380376e-05, + "loss": 0.553, + "step": 7753 + }, + { + "epoch": 0.24456710298060244, + "grad_norm": 2.6489737033843994, + "learning_rate": 1.76820701084504e-05, + "loss": 0.5497, + "step": 7754 + }, + { + "epoch": 0.24459864374704304, + "grad_norm": 3.1171669960021973, + "learning_rate": 1.768141606635206e-05, + "loss": 0.553, + "step": 7755 + }, + { + "epoch": 0.24463018451348367, + "grad_norm": 2.4292712211608887, + "learning_rate": 1.7680761944092174e-05, + "loss": 0.5846, + "step": 7756 + }, + { + "epoch": 0.2446617252799243, + "grad_norm": 2.522878885269165, + "learning_rate": 1.768010774167758e-05, + "loss": 0.5264, + "step": 7757 + }, + { + "epoch": 0.24469326604636493, + "grad_norm": 4.2511305809021, + "learning_rate": 1.7679453459115094e-05, + "loss": 0.5607, + "step": 7758 + }, + { + "epoch": 0.24472480681280556, + "grad_norm": 3.683677911758423, + "learning_rate": 1.7678799096411548e-05, + "loss": 0.5753, + "step": 7759 + }, + { + "epoch": 0.24475634757924616, + "grad_norm": 3.2745590209960938, + "learning_rate": 1.7678144653573773e-05, + "loss": 0.5491, + "step": 7760 + }, + { + "epoch": 0.2447878883456868, + "grad_norm": 6.490307331085205, + "learning_rate": 1.7677490130608595e-05, + "loss": 0.5451, + "step": 7761 + }, + { + "epoch": 0.24481942911212742, + "grad_norm": 2.8515968322753906, + "learning_rate": 1.7676835527522845e-05, + "loss": 0.6166, + "step": 7762 + }, + { + "epoch": 0.24485096987856805, + "grad_norm": 1.8099087476730347, + "learning_rate": 1.7676180844323357e-05, + "loss": 0.9254, + "step": 7763 + }, + { + "epoch": 0.24488251064500868, + "grad_norm": 1.9075862169265747, + "learning_rate": 1.7675526081016965e-05, + "loss": 0.5613, + "step": 7764 + }, + { + "epoch": 0.24491405141144929, + "grad_norm": 2.3555948734283447, + "learning_rate": 1.7674871237610492e-05, + "loss": 0.5324, + "step": 7765 + }, + { + "epoch": 0.24494559217788991, + "grad_norm": 2.2098584175109863, + "learning_rate": 1.7674216314110782e-05, + "loss": 0.8721, + "step": 7766 + }, + { + "epoch": 0.24497713294433054, + "grad_norm": 3.8147759437561035, + "learning_rate": 1.7673561310524667e-05, + "loss": 0.5675, + "step": 7767 + }, + { + "epoch": 0.24500867371077117, + "grad_norm": 2.323934316635132, + "learning_rate": 1.7672906226858975e-05, + "loss": 0.5322, + "step": 7768 + }, + { + "epoch": 0.2450402144772118, + "grad_norm": 2.1173086166381836, + "learning_rate": 1.7672251063120556e-05, + "loss": 0.5652, + "step": 7769 + }, + { + "epoch": 0.24507175524365243, + "grad_norm": 2.811889171600342, + "learning_rate": 1.767159581931624e-05, + "loss": 0.5337, + "step": 7770 + }, + { + "epoch": 0.24510329601009304, + "grad_norm": 2.202597141265869, + "learning_rate": 1.767094049545286e-05, + "loss": 0.5697, + "step": 7771 + }, + { + "epoch": 0.24513483677653367, + "grad_norm": 2.391878366470337, + "learning_rate": 1.767028509153726e-05, + "loss": 0.5745, + "step": 7772 + }, + { + "epoch": 0.2451663775429743, + "grad_norm": 2.499040365219116, + "learning_rate": 1.7669629607576283e-05, + "loss": 0.5468, + "step": 7773 + }, + { + "epoch": 0.24519791830941493, + "grad_norm": 2.6477904319763184, + "learning_rate": 1.766897404357676e-05, + "loss": 0.5232, + "step": 7774 + }, + { + "epoch": 0.24522945907585555, + "grad_norm": 2.3695688247680664, + "learning_rate": 1.7668318399545544e-05, + "loss": 0.5587, + "step": 7775 + }, + { + "epoch": 0.24526099984229616, + "grad_norm": 2.3853580951690674, + "learning_rate": 1.7667662675489466e-05, + "loss": 0.5564, + "step": 7776 + }, + { + "epoch": 0.2452925406087368, + "grad_norm": 4.571869373321533, + "learning_rate": 1.766700687141538e-05, + "loss": 0.5611, + "step": 7777 + }, + { + "epoch": 0.24532408137517742, + "grad_norm": 3.084425687789917, + "learning_rate": 1.7666350987330122e-05, + "loss": 0.5688, + "step": 7778 + }, + { + "epoch": 0.24535562214161805, + "grad_norm": 2.373335599899292, + "learning_rate": 1.7665695023240537e-05, + "loss": 0.5868, + "step": 7779 + }, + { + "epoch": 0.24538716290805868, + "grad_norm": 2.9410293102264404, + "learning_rate": 1.766503897915347e-05, + "loss": 0.5288, + "step": 7780 + }, + { + "epoch": 0.24541870367449928, + "grad_norm": 2.1675052642822266, + "learning_rate": 1.7664382855075773e-05, + "loss": 0.5598, + "step": 7781 + }, + { + "epoch": 0.2454502444409399, + "grad_norm": 2.3826379776000977, + "learning_rate": 1.766372665101429e-05, + "loss": 0.5572, + "step": 7782 + }, + { + "epoch": 0.24548178520738054, + "grad_norm": 2.0371124744415283, + "learning_rate": 1.7663070366975867e-05, + "loss": 0.5767, + "step": 7783 + }, + { + "epoch": 0.24551332597382117, + "grad_norm": 3.4381585121154785, + "learning_rate": 1.7662414002967353e-05, + "loss": 0.5991, + "step": 7784 + }, + { + "epoch": 0.2455448667402618, + "grad_norm": 2.28572940826416, + "learning_rate": 1.7661757558995602e-05, + "loss": 0.9269, + "step": 7785 + }, + { + "epoch": 0.24557640750670243, + "grad_norm": 1.9917744398117065, + "learning_rate": 1.766110103506746e-05, + "loss": 0.5307, + "step": 7786 + }, + { + "epoch": 0.24560794827314303, + "grad_norm": 2.1920416355133057, + "learning_rate": 1.766044443118978e-05, + "loss": 0.5769, + "step": 7787 + }, + { + "epoch": 0.24563948903958366, + "grad_norm": 2.301436424255371, + "learning_rate": 1.7659787747369415e-05, + "loss": 0.513, + "step": 7788 + }, + { + "epoch": 0.2456710298060243, + "grad_norm": 1.9986010789871216, + "learning_rate": 1.765913098361322e-05, + "loss": 0.8531, + "step": 7789 + }, + { + "epoch": 0.24570257057246492, + "grad_norm": 2.4034993648529053, + "learning_rate": 1.765847413992804e-05, + "loss": 0.5867, + "step": 7790 + }, + { + "epoch": 0.24573411133890555, + "grad_norm": 2.790219306945801, + "learning_rate": 1.765781721632074e-05, + "loss": 0.8802, + "step": 7791 + }, + { + "epoch": 0.24576565210534615, + "grad_norm": 3.9757328033447266, + "learning_rate": 1.765716021279817e-05, + "loss": 0.566, + "step": 7792 + }, + { + "epoch": 0.24579719287178678, + "grad_norm": 2.5215847492218018, + "learning_rate": 1.7656503129367187e-05, + "loss": 0.5899, + "step": 7793 + }, + { + "epoch": 0.2458287336382274, + "grad_norm": 3.621583938598633, + "learning_rate": 1.7655845966034653e-05, + "loss": 0.5455, + "step": 7794 + }, + { + "epoch": 0.24586027440466804, + "grad_norm": 6.2868757247924805, + "learning_rate": 1.7655188722807418e-05, + "loss": 0.5424, + "step": 7795 + }, + { + "epoch": 0.24589181517110867, + "grad_norm": 2.5811767578125, + "learning_rate": 1.7654531399692344e-05, + "loss": 0.5403, + "step": 7796 + }, + { + "epoch": 0.24592335593754927, + "grad_norm": 3.588564872741699, + "learning_rate": 1.765387399669629e-05, + "loss": 0.5653, + "step": 7797 + }, + { + "epoch": 0.2459548967039899, + "grad_norm": 2.192315101623535, + "learning_rate": 1.7653216513826123e-05, + "loss": 0.5512, + "step": 7798 + }, + { + "epoch": 0.24598643747043053, + "grad_norm": 2.587761163711548, + "learning_rate": 1.7652558951088698e-05, + "loss": 0.559, + "step": 7799 + }, + { + "epoch": 0.24601797823687116, + "grad_norm": 2.498892307281494, + "learning_rate": 1.7651901308490872e-05, + "loss": 0.5814, + "step": 7800 + }, + { + "epoch": 0.2460495190033118, + "grad_norm": 3.1401114463806152, + "learning_rate": 1.765124358603952e-05, + "loss": 0.5594, + "step": 7801 + }, + { + "epoch": 0.2460810597697524, + "grad_norm": 2.0049569606781006, + "learning_rate": 1.76505857837415e-05, + "loss": 0.5916, + "step": 7802 + }, + { + "epoch": 0.24611260053619302, + "grad_norm": 6.595622539520264, + "learning_rate": 1.7649927901603674e-05, + "loss": 0.5693, + "step": 7803 + }, + { + "epoch": 0.24614414130263365, + "grad_norm": 2.0114989280700684, + "learning_rate": 1.764926993963291e-05, + "loss": 0.5566, + "step": 7804 + }, + { + "epoch": 0.24617568206907428, + "grad_norm": 2.526902914047241, + "learning_rate": 1.764861189783608e-05, + "loss": 0.5871, + "step": 7805 + }, + { + "epoch": 0.2462072228355149, + "grad_norm": 1.926603078842163, + "learning_rate": 1.764795377622004e-05, + "loss": 0.5634, + "step": 7806 + }, + { + "epoch": 0.24623876360195554, + "grad_norm": 3.543552875518799, + "learning_rate": 1.7647295574791665e-05, + "loss": 0.5904, + "step": 7807 + }, + { + "epoch": 0.24627030436839614, + "grad_norm": 3.2157835960388184, + "learning_rate": 1.7646637293557824e-05, + "loss": 0.5752, + "step": 7808 + }, + { + "epoch": 0.24630184513483677, + "grad_norm": 1.8955910205841064, + "learning_rate": 1.7645978932525385e-05, + "loss": 0.592, + "step": 7809 + }, + { + "epoch": 0.2463333859012774, + "grad_norm": 3.2936861515045166, + "learning_rate": 1.764532049170122e-05, + "loss": 0.5686, + "step": 7810 + }, + { + "epoch": 0.24636492666771803, + "grad_norm": 1.907185673713684, + "learning_rate": 1.76446619710922e-05, + "loss": 0.5782, + "step": 7811 + }, + { + "epoch": 0.24639646743415866, + "grad_norm": 5.443737506866455, + "learning_rate": 1.7644003370705193e-05, + "loss": 0.4801, + "step": 7812 + }, + { + "epoch": 0.24642800820059926, + "grad_norm": 1.9754401445388794, + "learning_rate": 1.7643344690547078e-05, + "loss": 0.5667, + "step": 7813 + }, + { + "epoch": 0.2464595489670399, + "grad_norm": 3.3639440536499023, + "learning_rate": 1.7642685930624727e-05, + "loss": 0.5714, + "step": 7814 + }, + { + "epoch": 0.24649108973348052, + "grad_norm": 2.9618191719055176, + "learning_rate": 1.764202709094501e-05, + "loss": 0.5614, + "step": 7815 + }, + { + "epoch": 0.24652263049992115, + "grad_norm": 1.9057848453521729, + "learning_rate": 1.7641368171514806e-05, + "loss": 0.5553, + "step": 7816 + }, + { + "epoch": 0.24655417126636178, + "grad_norm": 1.8917930126190186, + "learning_rate": 1.7640709172340996e-05, + "loss": 0.518, + "step": 7817 + }, + { + "epoch": 0.24658571203280238, + "grad_norm": 2.4900386333465576, + "learning_rate": 1.7640050093430452e-05, + "loss": 0.556, + "step": 7818 + }, + { + "epoch": 0.24661725279924301, + "grad_norm": 2.2838375568389893, + "learning_rate": 1.7639390934790052e-05, + "loss": 0.5683, + "step": 7819 + }, + { + "epoch": 0.24664879356568364, + "grad_norm": 2.089371919631958, + "learning_rate": 1.763873169642668e-05, + "loss": 0.5592, + "step": 7820 + }, + { + "epoch": 0.24668033433212427, + "grad_norm": 2.698275566101074, + "learning_rate": 1.7638072378347205e-05, + "loss": 0.5809, + "step": 7821 + }, + { + "epoch": 0.2467118750985649, + "grad_norm": 2.740889310836792, + "learning_rate": 1.7637412980558517e-05, + "loss": 0.5382, + "step": 7822 + }, + { + "epoch": 0.24674341586500553, + "grad_norm": 1.9438927173614502, + "learning_rate": 1.7636753503067492e-05, + "loss": 0.5629, + "step": 7823 + }, + { + "epoch": 0.24677495663144614, + "grad_norm": 2.7208268642425537, + "learning_rate": 1.7636093945881018e-05, + "loss": 0.9296, + "step": 7824 + }, + { + "epoch": 0.24680649739788676, + "grad_norm": 3.52590274810791, + "learning_rate": 1.763543430900597e-05, + "loss": 0.5216, + "step": 7825 + }, + { + "epoch": 0.2468380381643274, + "grad_norm": 2.0935070514678955, + "learning_rate": 1.763477459244924e-05, + "loss": 0.5598, + "step": 7826 + }, + { + "epoch": 0.24686957893076802, + "grad_norm": 3.2878050804138184, + "learning_rate": 1.763411479621771e-05, + "loss": 0.5212, + "step": 7827 + }, + { + "epoch": 0.24690111969720865, + "grad_norm": 2.2814128398895264, + "learning_rate": 1.763345492031826e-05, + "loss": 0.533, + "step": 7828 + }, + { + "epoch": 0.24693266046364926, + "grad_norm": 2.8760790824890137, + "learning_rate": 1.7632794964757783e-05, + "loss": 0.5823, + "step": 7829 + }, + { + "epoch": 0.24696420123008989, + "grad_norm": 1.8530217409133911, + "learning_rate": 1.763213492954316e-05, + "loss": 0.5512, + "step": 7830 + }, + { + "epoch": 0.24699574199653052, + "grad_norm": 1.9317028522491455, + "learning_rate": 1.7631474814681287e-05, + "loss": 0.9065, + "step": 7831 + }, + { + "epoch": 0.24702728276297115, + "grad_norm": 1.6334772109985352, + "learning_rate": 1.763081462017905e-05, + "loss": 0.8855, + "step": 7832 + }, + { + "epoch": 0.24705882352941178, + "grad_norm": 1.602992296218872, + "learning_rate": 1.7630154346043337e-05, + "loss": 0.8699, + "step": 7833 + }, + { + "epoch": 0.24709036429585238, + "grad_norm": 2.692005157470703, + "learning_rate": 1.7629493992281034e-05, + "loss": 0.553, + "step": 7834 + }, + { + "epoch": 0.247121905062293, + "grad_norm": 2.1179633140563965, + "learning_rate": 1.7628833558899044e-05, + "loss": 0.5698, + "step": 7835 + }, + { + "epoch": 0.24715344582873364, + "grad_norm": 2.0194509029388428, + "learning_rate": 1.7628173045904247e-05, + "loss": 0.5464, + "step": 7836 + }, + { + "epoch": 0.24718498659517427, + "grad_norm": 1.9989811182022095, + "learning_rate": 1.7627512453303542e-05, + "loss": 0.584, + "step": 7837 + }, + { + "epoch": 0.2472165273616149, + "grad_norm": 1.9670565128326416, + "learning_rate": 1.762685178110382e-05, + "loss": 0.4374, + "step": 7838 + }, + { + "epoch": 0.2472480681280555, + "grad_norm": 2.5555260181427, + "learning_rate": 1.762619102931198e-05, + "loss": 0.5682, + "step": 7839 + }, + { + "epoch": 0.24727960889449613, + "grad_norm": 2.1701176166534424, + "learning_rate": 1.7625530197934915e-05, + "loss": 0.5211, + "step": 7840 + }, + { + "epoch": 0.24731114966093676, + "grad_norm": 4.796583652496338, + "learning_rate": 1.7624869286979523e-05, + "loss": 0.5239, + "step": 7841 + }, + { + "epoch": 0.2473426904273774, + "grad_norm": 1.8789920806884766, + "learning_rate": 1.7624208296452697e-05, + "loss": 0.5505, + "step": 7842 + }, + { + "epoch": 0.24737423119381802, + "grad_norm": 2.158489465713501, + "learning_rate": 1.762354722636134e-05, + "loss": 0.5903, + "step": 7843 + }, + { + "epoch": 0.24740577196025865, + "grad_norm": 1.9883041381835938, + "learning_rate": 1.7622886076712344e-05, + "loss": 0.5674, + "step": 7844 + }, + { + "epoch": 0.24743731272669925, + "grad_norm": 3.987381935119629, + "learning_rate": 1.7622224847512613e-05, + "loss": 0.8939, + "step": 7845 + }, + { + "epoch": 0.24746885349313988, + "grad_norm": 1.7378002405166626, + "learning_rate": 1.762156353876905e-05, + "loss": 0.5337, + "step": 7846 + }, + { + "epoch": 0.2475003942595805, + "grad_norm": 1.8051700592041016, + "learning_rate": 1.7620902150488554e-05, + "loss": 0.5629, + "step": 7847 + }, + { + "epoch": 0.24753193502602114, + "grad_norm": 1.899153470993042, + "learning_rate": 1.7620240682678026e-05, + "loss": 0.5592, + "step": 7848 + }, + { + "epoch": 0.24756347579246177, + "grad_norm": 2.0200517177581787, + "learning_rate": 1.7619579135344368e-05, + "loss": 0.5447, + "step": 7849 + }, + { + "epoch": 0.24759501655890237, + "grad_norm": 1.748901605606079, + "learning_rate": 1.7618917508494483e-05, + "loss": 0.5373, + "step": 7850 + }, + { + "epoch": 0.247626557325343, + "grad_norm": 1.5670145750045776, + "learning_rate": 1.7618255802135285e-05, + "loss": 0.9109, + "step": 7851 + }, + { + "epoch": 0.24765809809178363, + "grad_norm": 3.036254405975342, + "learning_rate": 1.7617594016273667e-05, + "loss": 0.5533, + "step": 7852 + }, + { + "epoch": 0.24768963885822426, + "grad_norm": 1.94063401222229, + "learning_rate": 1.7616932150916543e-05, + "loss": 0.5676, + "step": 7853 + }, + { + "epoch": 0.2477211796246649, + "grad_norm": 2.4512009620666504, + "learning_rate": 1.7616270206070814e-05, + "loss": 0.8686, + "step": 7854 + }, + { + "epoch": 0.2477527203911055, + "grad_norm": 1.6822023391723633, + "learning_rate": 1.7615608181743394e-05, + "loss": 0.5567, + "step": 7855 + }, + { + "epoch": 0.24778426115754612, + "grad_norm": 2.2710325717926025, + "learning_rate": 1.7614946077941192e-05, + "loss": 0.5868, + "step": 7856 + }, + { + "epoch": 0.24781580192398675, + "grad_norm": 1.7215986251831055, + "learning_rate": 1.761428389467111e-05, + "loss": 0.866, + "step": 7857 + }, + { + "epoch": 0.24784734269042738, + "grad_norm": 3.27828311920166, + "learning_rate": 1.7613621631940065e-05, + "loss": 0.5643, + "step": 7858 + }, + { + "epoch": 0.247878883456868, + "grad_norm": 2.0179221630096436, + "learning_rate": 1.761295928975497e-05, + "loss": 0.5557, + "step": 7859 + }, + { + "epoch": 0.24791042422330864, + "grad_norm": 2.1634488105773926, + "learning_rate": 1.7612296868122726e-05, + "loss": 0.5681, + "step": 7860 + }, + { + "epoch": 0.24794196498974924, + "grad_norm": 1.7826013565063477, + "learning_rate": 1.7611634367050258e-05, + "loss": 0.6, + "step": 7861 + }, + { + "epoch": 0.24797350575618987, + "grad_norm": 2.3334593772888184, + "learning_rate": 1.7610971786544473e-05, + "loss": 0.5618, + "step": 7862 + }, + { + "epoch": 0.2480050465226305, + "grad_norm": 14.931379318237305, + "learning_rate": 1.7610309126612287e-05, + "loss": 0.5403, + "step": 7863 + }, + { + "epoch": 0.24803658728907113, + "grad_norm": 2.007114887237549, + "learning_rate": 1.7609646387260618e-05, + "loss": 0.8824, + "step": 7864 + }, + { + "epoch": 0.24806812805551176, + "grad_norm": 2.3736162185668945, + "learning_rate": 1.760898356849638e-05, + "loss": 0.5469, + "step": 7865 + }, + { + "epoch": 0.24809966882195236, + "grad_norm": 1.7664973735809326, + "learning_rate": 1.7608320670326496e-05, + "loss": 0.8561, + "step": 7866 + }, + { + "epoch": 0.248131209588393, + "grad_norm": 2.9371461868286133, + "learning_rate": 1.760765769275787e-05, + "loss": 0.5692, + "step": 7867 + }, + { + "epoch": 0.24816275035483362, + "grad_norm": 2.3671653270721436, + "learning_rate": 1.760699463579743e-05, + "loss": 0.8968, + "step": 7868 + }, + { + "epoch": 0.24819429112127425, + "grad_norm": 1.9567281007766724, + "learning_rate": 1.7606331499452096e-05, + "loss": 0.5595, + "step": 7869 + }, + { + "epoch": 0.24822583188771488, + "grad_norm": 1.9482144117355347, + "learning_rate": 1.760566828372879e-05, + "loss": 0.5535, + "step": 7870 + }, + { + "epoch": 0.24825737265415548, + "grad_norm": 2.5588817596435547, + "learning_rate": 1.7605004988634426e-05, + "loss": 0.424, + "step": 7871 + }, + { + "epoch": 0.2482889134205961, + "grad_norm": 3.5225706100463867, + "learning_rate": 1.760434161417593e-05, + "loss": 0.5824, + "step": 7872 + }, + { + "epoch": 0.24832045418703674, + "grad_norm": 2.552374839782715, + "learning_rate": 1.7603678160360225e-05, + "loss": 0.5506, + "step": 7873 + }, + { + "epoch": 0.24835199495347737, + "grad_norm": 2.095529556274414, + "learning_rate": 1.7603014627194232e-05, + "loss": 0.8756, + "step": 7874 + }, + { + "epoch": 0.248383535719918, + "grad_norm": 1.7343894243240356, + "learning_rate": 1.760235101468488e-05, + "loss": 0.5333, + "step": 7875 + }, + { + "epoch": 0.2484150764863586, + "grad_norm": 1.7977219820022583, + "learning_rate": 1.7601687322839092e-05, + "loss": 0.5881, + "step": 7876 + }, + { + "epoch": 0.24844661725279923, + "grad_norm": 1.9231065511703491, + "learning_rate": 1.7601023551663793e-05, + "loss": 0.5315, + "step": 7877 + }, + { + "epoch": 0.24847815801923986, + "grad_norm": 2.1227493286132812, + "learning_rate": 1.7600359701165915e-05, + "loss": 0.5421, + "step": 7878 + }, + { + "epoch": 0.2485096987856805, + "grad_norm": 1.501905083656311, + "learning_rate": 1.759969577135238e-05, + "loss": 0.5289, + "step": 7879 + }, + { + "epoch": 0.24854123955212112, + "grad_norm": 1.5283923149108887, + "learning_rate": 1.7599031762230113e-05, + "loss": 0.5214, + "step": 7880 + }, + { + "epoch": 0.24857278031856175, + "grad_norm": 2.9396026134490967, + "learning_rate": 1.7598367673806057e-05, + "loss": 0.5567, + "step": 7881 + }, + { + "epoch": 0.24860432108500236, + "grad_norm": 2.0317366123199463, + "learning_rate": 1.759770350608713e-05, + "loss": 0.5757, + "step": 7882 + }, + { + "epoch": 0.24863586185144299, + "grad_norm": 1.6875720024108887, + "learning_rate": 1.7597039259080268e-05, + "loss": 0.5493, + "step": 7883 + }, + { + "epoch": 0.24866740261788361, + "grad_norm": 2.124046564102173, + "learning_rate": 1.7596374932792403e-05, + "loss": 0.5745, + "step": 7884 + }, + { + "epoch": 0.24869894338432424, + "grad_norm": 2.7994608879089355, + "learning_rate": 1.7595710527230466e-05, + "loss": 0.5565, + "step": 7885 + }, + { + "epoch": 0.24873048415076487, + "grad_norm": 1.7373437881469727, + "learning_rate": 1.7595046042401395e-05, + "loss": 0.5624, + "step": 7886 + }, + { + "epoch": 0.24876202491720548, + "grad_norm": 1.8541232347488403, + "learning_rate": 1.7594381478312123e-05, + "loss": 0.5636, + "step": 7887 + }, + { + "epoch": 0.2487935656836461, + "grad_norm": 2.113865375518799, + "learning_rate": 1.7593716834969576e-05, + "loss": 0.5601, + "step": 7888 + }, + { + "epoch": 0.24882510645008674, + "grad_norm": 2.0588040351867676, + "learning_rate": 1.75930521123807e-05, + "loss": 0.5211, + "step": 7889 + }, + { + "epoch": 0.24885664721652737, + "grad_norm": 6.992254257202148, + "learning_rate": 1.7592387310552432e-05, + "loss": 0.4228, + "step": 7890 + }, + { + "epoch": 0.248888187982968, + "grad_norm": 2.5579192638397217, + "learning_rate": 1.7591722429491708e-05, + "loss": 0.5656, + "step": 7891 + }, + { + "epoch": 0.2489197287494086, + "grad_norm": 2.091280937194824, + "learning_rate": 1.759105746920546e-05, + "loss": 0.5588, + "step": 7892 + }, + { + "epoch": 0.24895126951584923, + "grad_norm": 2.361844301223755, + "learning_rate": 1.7590392429700636e-05, + "loss": 0.5503, + "step": 7893 + }, + { + "epoch": 0.24898281028228986, + "grad_norm": 1.8383209705352783, + "learning_rate": 1.7589727310984175e-05, + "loss": 0.8624, + "step": 7894 + }, + { + "epoch": 0.2490143510487305, + "grad_norm": 1.8730307817459106, + "learning_rate": 1.758906211306301e-05, + "loss": 0.5167, + "step": 7895 + }, + { + "epoch": 0.24904589181517112, + "grad_norm": 1.8990861177444458, + "learning_rate": 1.75883968359441e-05, + "loss": 0.5584, + "step": 7896 + }, + { + "epoch": 0.24907743258161175, + "grad_norm": 2.4185187816619873, + "learning_rate": 1.7587731479634367e-05, + "loss": 0.5529, + "step": 7897 + }, + { + "epoch": 0.24910897334805235, + "grad_norm": 2.234544277191162, + "learning_rate": 1.7587066044140767e-05, + "loss": 0.4346, + "step": 7898 + }, + { + "epoch": 0.24914051411449298, + "grad_norm": 2.1478500366210938, + "learning_rate": 1.758640052947024e-05, + "loss": 0.5764, + "step": 7899 + }, + { + "epoch": 0.2491720548809336, + "grad_norm": 5.466667175292969, + "learning_rate": 1.7585734935629737e-05, + "loss": 0.5538, + "step": 7900 + }, + { + "epoch": 0.24920359564737424, + "grad_norm": 2.4368298053741455, + "learning_rate": 1.7585069262626198e-05, + "loss": 0.5233, + "step": 7901 + }, + { + "epoch": 0.24923513641381487, + "grad_norm": 2.1277151107788086, + "learning_rate": 1.7584403510466573e-05, + "loss": 0.5336, + "step": 7902 + }, + { + "epoch": 0.24926667718025547, + "grad_norm": 1.8184175491333008, + "learning_rate": 1.7583737679157807e-05, + "loss": 0.5763, + "step": 7903 + }, + { + "epoch": 0.2492982179466961, + "grad_norm": 1.7067854404449463, + "learning_rate": 1.7583071768706847e-05, + "loss": 0.8914, + "step": 7904 + }, + { + "epoch": 0.24932975871313673, + "grad_norm": 1.8958501815795898, + "learning_rate": 1.7582405779120647e-05, + "loss": 0.8727, + "step": 7905 + }, + { + "epoch": 0.24936129947957736, + "grad_norm": 3.1509156227111816, + "learning_rate": 1.7581739710406158e-05, + "loss": 0.545, + "step": 7906 + }, + { + "epoch": 0.249392840246018, + "grad_norm": 2.7854926586151123, + "learning_rate": 1.7581073562570325e-05, + "loss": 0.5579, + "step": 7907 + }, + { + "epoch": 0.2494243810124586, + "grad_norm": 2.2300240993499756, + "learning_rate": 1.7580407335620103e-05, + "loss": 0.5811, + "step": 7908 + }, + { + "epoch": 0.24945592177889922, + "grad_norm": 1.9047785997390747, + "learning_rate": 1.7579741029562445e-05, + "loss": 0.5773, + "step": 7909 + }, + { + "epoch": 0.24948746254533985, + "grad_norm": 1.7634618282318115, + "learning_rate": 1.75790746444043e-05, + "loss": 0.5842, + "step": 7910 + }, + { + "epoch": 0.24951900331178048, + "grad_norm": 2.0117077827453613, + "learning_rate": 1.757840818015263e-05, + "loss": 0.567, + "step": 7911 + }, + { + "epoch": 0.2495505440782211, + "grad_norm": 2.6267333030700684, + "learning_rate": 1.7577741636814388e-05, + "loss": 0.545, + "step": 7912 + }, + { + "epoch": 0.24958208484466174, + "grad_norm": 2.656409502029419, + "learning_rate": 1.7577075014396524e-05, + "loss": 0.539, + "step": 7913 + }, + { + "epoch": 0.24961362561110234, + "grad_norm": 1.4663283824920654, + "learning_rate": 1.7576408312906004e-05, + "loss": 0.5279, + "step": 7914 + }, + { + "epoch": 0.24964516637754297, + "grad_norm": 2.493518829345703, + "learning_rate": 1.7575741532349776e-05, + "loss": 0.8875, + "step": 7915 + }, + { + "epoch": 0.2496767071439836, + "grad_norm": 3.1516222953796387, + "learning_rate": 1.7575074672734803e-05, + "loss": 0.55, + "step": 7916 + }, + { + "epoch": 0.24970824791042423, + "grad_norm": 2.7383100986480713, + "learning_rate": 1.7574407734068044e-05, + "loss": 0.5826, + "step": 7917 + }, + { + "epoch": 0.24973978867686486, + "grad_norm": 1.8581160306930542, + "learning_rate": 1.757374071635646e-05, + "loss": 0.5492, + "step": 7918 + }, + { + "epoch": 0.24977132944330546, + "grad_norm": 1.8651597499847412, + "learning_rate": 1.7573073619607012e-05, + "loss": 0.8282, + "step": 7919 + }, + { + "epoch": 0.2498028702097461, + "grad_norm": 1.5022170543670654, + "learning_rate": 1.7572406443826658e-05, + "loss": 0.5551, + "step": 7920 + }, + { + "epoch": 0.24983441097618672, + "grad_norm": 1.7310703992843628, + "learning_rate": 1.7571739189022365e-05, + "loss": 0.5449, + "step": 7921 + }, + { + "epoch": 0.24986595174262735, + "grad_norm": 2.397320508956909, + "learning_rate": 1.7571071855201092e-05, + "loss": 0.5523, + "step": 7922 + }, + { + "epoch": 0.24989749250906798, + "grad_norm": 1.9942313432693481, + "learning_rate": 1.7570404442369805e-05, + "loss": 0.5308, + "step": 7923 + }, + { + "epoch": 0.24992903327550858, + "grad_norm": 2.8499228954315186, + "learning_rate": 1.756973695053547e-05, + "loss": 0.5658, + "step": 7924 + }, + { + "epoch": 0.2499605740419492, + "grad_norm": 1.987012267112732, + "learning_rate": 1.7569069379705052e-05, + "loss": 0.5834, + "step": 7925 + }, + { + "epoch": 0.24999211480838984, + "grad_norm": 1.7639706134796143, + "learning_rate": 1.756840172988552e-05, + "loss": 0.5313, + "step": 7926 + }, + { + "epoch": 0.2500236555748305, + "grad_norm": 2.5861926078796387, + "learning_rate": 1.7567734001083837e-05, + "loss": 0.8918, + "step": 7927 + }, + { + "epoch": 0.2500551963412711, + "grad_norm": 2.2949295043945312, + "learning_rate": 1.7567066193306976e-05, + "loss": 0.5787, + "step": 7928 + }, + { + "epoch": 0.25008673710771173, + "grad_norm": 1.7439602613449097, + "learning_rate": 1.75663983065619e-05, + "loss": 0.5844, + "step": 7929 + }, + { + "epoch": 0.25011827787415236, + "grad_norm": 1.419713020324707, + "learning_rate": 1.7565730340855588e-05, + "loss": 0.561, + "step": 7930 + }, + { + "epoch": 0.250149818640593, + "grad_norm": 1.314281702041626, + "learning_rate": 1.7565062296195e-05, + "loss": 0.5372, + "step": 7931 + }, + { + "epoch": 0.25018135940703357, + "grad_norm": 1.9010460376739502, + "learning_rate": 1.7564394172587114e-05, + "loss": 0.5927, + "step": 7932 + }, + { + "epoch": 0.2502129001734742, + "grad_norm": 1.5828142166137695, + "learning_rate": 1.7563725970038903e-05, + "loss": 0.4173, + "step": 7933 + }, + { + "epoch": 0.2502444409399148, + "grad_norm": 1.8157919645309448, + "learning_rate": 1.7563057688557335e-05, + "loss": 0.5571, + "step": 7934 + }, + { + "epoch": 0.25027598170635545, + "grad_norm": 1.9971143007278442, + "learning_rate": 1.7562389328149393e-05, + "loss": 0.5623, + "step": 7935 + }, + { + "epoch": 0.2503075224727961, + "grad_norm": 1.4570509195327759, + "learning_rate": 1.7561720888822042e-05, + "loss": 0.5568, + "step": 7936 + }, + { + "epoch": 0.2503390632392367, + "grad_norm": 1.8208541870117188, + "learning_rate": 1.7561052370582263e-05, + "loss": 0.4221, + "step": 7937 + }, + { + "epoch": 0.25037060400567734, + "grad_norm": 1.990247368812561, + "learning_rate": 1.756038377343703e-05, + "loss": 0.5471, + "step": 7938 + }, + { + "epoch": 0.250402144772118, + "grad_norm": 1.7287778854370117, + "learning_rate": 1.7559715097393324e-05, + "loss": 0.5957, + "step": 7939 + }, + { + "epoch": 0.2504336855385586, + "grad_norm": 1.6941821575164795, + "learning_rate": 1.7559046342458123e-05, + "loss": 0.5271, + "step": 7940 + }, + { + "epoch": 0.25046522630499923, + "grad_norm": 1.9524052143096924, + "learning_rate": 1.7558377508638397e-05, + "loss": 0.5794, + "step": 7941 + }, + { + "epoch": 0.25049676707143986, + "grad_norm": 3.2585623264312744, + "learning_rate": 1.7557708595941138e-05, + "loss": 0.5435, + "step": 7942 + }, + { + "epoch": 0.25052830783788044, + "grad_norm": 2.9212677478790283, + "learning_rate": 1.755703960437332e-05, + "loss": 0.8812, + "step": 7943 + }, + { + "epoch": 0.25055984860432107, + "grad_norm": 2.0435328483581543, + "learning_rate": 1.7556370533941928e-05, + "loss": 0.3876, + "step": 7944 + }, + { + "epoch": 0.2505913893707617, + "grad_norm": 2.229644298553467, + "learning_rate": 1.755570138465394e-05, + "loss": 0.5544, + "step": 7945 + }, + { + "epoch": 0.2506229301372023, + "grad_norm": 2.6843416690826416, + "learning_rate": 1.7555032156516342e-05, + "loss": 0.5999, + "step": 7946 + }, + { + "epoch": 0.25065447090364296, + "grad_norm": 1.637668490409851, + "learning_rate": 1.7554362849536118e-05, + "loss": 0.5504, + "step": 7947 + }, + { + "epoch": 0.2506860116700836, + "grad_norm": 1.6370540857315063, + "learning_rate": 1.755369346372025e-05, + "loss": 0.5425, + "step": 7948 + }, + { + "epoch": 0.2507175524365242, + "grad_norm": 1.9947468042373657, + "learning_rate": 1.7553023999075723e-05, + "loss": 0.549, + "step": 7949 + }, + { + "epoch": 0.25074909320296485, + "grad_norm": 1.8170369863510132, + "learning_rate": 1.755235445560953e-05, + "loss": 0.8854, + "step": 7950 + }, + { + "epoch": 0.2507806339694055, + "grad_norm": 2.8774735927581787, + "learning_rate": 1.7551684833328652e-05, + "loss": 0.5913, + "step": 7951 + }, + { + "epoch": 0.2508121747358461, + "grad_norm": 1.5471630096435547, + "learning_rate": 1.755101513224008e-05, + "loss": 0.5273, + "step": 7952 + }, + { + "epoch": 0.2508437155022867, + "grad_norm": 2.0927608013153076, + "learning_rate": 1.7550345352350805e-05, + "loss": 0.5796, + "step": 7953 + }, + { + "epoch": 0.2508752562687273, + "grad_norm": 2.5297656059265137, + "learning_rate": 1.7549675493667807e-05, + "loss": 0.5562, + "step": 7954 + }, + { + "epoch": 0.25090679703516794, + "grad_norm": 1.9615368843078613, + "learning_rate": 1.7549005556198087e-05, + "loss": 0.5543, + "step": 7955 + }, + { + "epoch": 0.25093833780160857, + "grad_norm": 2.824714422225952, + "learning_rate": 1.754833553994863e-05, + "loss": 0.5548, + "step": 7956 + }, + { + "epoch": 0.2509698785680492, + "grad_norm": 1.819399356842041, + "learning_rate": 1.7547665444926437e-05, + "loss": 0.5339, + "step": 7957 + }, + { + "epoch": 0.25100141933448983, + "grad_norm": 3.1541876792907715, + "learning_rate": 1.754699527113849e-05, + "loss": 0.5682, + "step": 7958 + }, + { + "epoch": 0.25103296010093046, + "grad_norm": 1.8616242408752441, + "learning_rate": 1.7546325018591784e-05, + "loss": 0.5588, + "step": 7959 + }, + { + "epoch": 0.2510645008673711, + "grad_norm": 1.990215539932251, + "learning_rate": 1.7545654687293322e-05, + "loss": 0.4232, + "step": 7960 + }, + { + "epoch": 0.2510960416338117, + "grad_norm": 3.643890380859375, + "learning_rate": 1.754498427725009e-05, + "loss": 0.551, + "step": 7961 + }, + { + "epoch": 0.25112758240025235, + "grad_norm": 2.222639799118042, + "learning_rate": 1.7544313788469092e-05, + "loss": 0.5074, + "step": 7962 + }, + { + "epoch": 0.251159123166693, + "grad_norm": 4.711174488067627, + "learning_rate": 1.754364322095732e-05, + "loss": 0.5832, + "step": 7963 + }, + { + "epoch": 0.25119066393313355, + "grad_norm": 1.6757149696350098, + "learning_rate": 1.7542972574721773e-05, + "loss": 0.8698, + "step": 7964 + }, + { + "epoch": 0.2512222046995742, + "grad_norm": 3.8035202026367188, + "learning_rate": 1.7542301849769452e-05, + "loss": 0.5521, + "step": 7965 + }, + { + "epoch": 0.2512537454660148, + "grad_norm": 2.040242910385132, + "learning_rate": 1.7541631046107354e-05, + "loss": 0.5655, + "step": 7966 + }, + { + "epoch": 0.25128528623245544, + "grad_norm": 1.8793193101882935, + "learning_rate": 1.754096016374248e-05, + "loss": 0.537, + "step": 7967 + }, + { + "epoch": 0.25131682699889607, + "grad_norm": 1.832893967628479, + "learning_rate": 1.754028920268183e-05, + "loss": 0.538, + "step": 7968 + }, + { + "epoch": 0.2513483677653367, + "grad_norm": 2.0228939056396484, + "learning_rate": 1.7539618162932408e-05, + "loss": 0.583, + "step": 7969 + }, + { + "epoch": 0.25137990853177733, + "grad_norm": 2.114424705505371, + "learning_rate": 1.7538947044501218e-05, + "loss": 0.8862, + "step": 7970 + }, + { + "epoch": 0.25141144929821796, + "grad_norm": 3.546727418899536, + "learning_rate": 1.753827584739526e-05, + "loss": 0.5816, + "step": 7971 + }, + { + "epoch": 0.2514429900646586, + "grad_norm": 2.245659112930298, + "learning_rate": 1.7537604571621537e-05, + "loss": 0.5404, + "step": 7972 + }, + { + "epoch": 0.2514745308310992, + "grad_norm": 3.366640567779541, + "learning_rate": 1.7536933217187062e-05, + "loss": 0.5353, + "step": 7973 + }, + { + "epoch": 0.2515060715975398, + "grad_norm": 2.113994836807251, + "learning_rate": 1.7536261784098836e-05, + "loss": 0.5565, + "step": 7974 + }, + { + "epoch": 0.2515376123639804, + "grad_norm": 1.7829407453536987, + "learning_rate": 1.7535590272363866e-05, + "loss": 0.5445, + "step": 7975 + }, + { + "epoch": 0.25156915313042105, + "grad_norm": 1.5493440628051758, + "learning_rate": 1.753491868198916e-05, + "loss": 0.5505, + "step": 7976 + }, + { + "epoch": 0.2516006938968617, + "grad_norm": 1.8867406845092773, + "learning_rate": 1.753424701298173e-05, + "loss": 0.5276, + "step": 7977 + }, + { + "epoch": 0.2516322346633023, + "grad_norm": 2.73425030708313, + "learning_rate": 1.7533575265348577e-05, + "loss": 0.5549, + "step": 7978 + }, + { + "epoch": 0.25166377542974294, + "grad_norm": 1.9582401514053345, + "learning_rate": 1.753290343909672e-05, + "loss": 0.9031, + "step": 7979 + }, + { + "epoch": 0.25169531619618357, + "grad_norm": 2.1195712089538574, + "learning_rate": 1.7532231534233166e-05, + "loss": 0.5515, + "step": 7980 + }, + { + "epoch": 0.2517268569626242, + "grad_norm": 2.605334997177124, + "learning_rate": 1.7531559550764927e-05, + "loss": 0.5653, + "step": 7981 + }, + { + "epoch": 0.25175839772906483, + "grad_norm": 2.1576836109161377, + "learning_rate": 1.7530887488699013e-05, + "loss": 0.5494, + "step": 7982 + }, + { + "epoch": 0.25178993849550546, + "grad_norm": 1.738994836807251, + "learning_rate": 1.7530215348042444e-05, + "loss": 0.8832, + "step": 7983 + }, + { + "epoch": 0.2518214792619461, + "grad_norm": 2.030984878540039, + "learning_rate": 1.7529543128802232e-05, + "loss": 0.5351, + "step": 7984 + }, + { + "epoch": 0.25185302002838666, + "grad_norm": 1.5893371105194092, + "learning_rate": 1.752887083098539e-05, + "loss": 0.5651, + "step": 7985 + }, + { + "epoch": 0.2518845607948273, + "grad_norm": 1.906062126159668, + "learning_rate": 1.752819845459893e-05, + "loss": 0.5621, + "step": 7986 + }, + { + "epoch": 0.2519161015612679, + "grad_norm": 1.6308828592300415, + "learning_rate": 1.7527525999649876e-05, + "loss": 0.5327, + "step": 7987 + }, + { + "epoch": 0.25194764232770855, + "grad_norm": 1.9252352714538574, + "learning_rate": 1.7526853466145248e-05, + "loss": 0.5737, + "step": 7988 + }, + { + "epoch": 0.2519791830941492, + "grad_norm": 1.7468147277832031, + "learning_rate": 1.7526180854092056e-05, + "loss": 0.5797, + "step": 7989 + }, + { + "epoch": 0.2520107238605898, + "grad_norm": 5.182703971862793, + "learning_rate": 1.7525508163497324e-05, + "loss": 0.547, + "step": 7990 + }, + { + "epoch": 0.25204226462703044, + "grad_norm": 1.6359127759933472, + "learning_rate": 1.752483539436807e-05, + "loss": 0.8246, + "step": 7991 + }, + { + "epoch": 0.2520738053934711, + "grad_norm": 2.819481372833252, + "learning_rate": 1.752416254671131e-05, + "loss": 0.5619, + "step": 7992 + }, + { + "epoch": 0.2521053461599117, + "grad_norm": 1.5967578887939453, + "learning_rate": 1.752348962053408e-05, + "loss": 0.554, + "step": 7993 + }, + { + "epoch": 0.25213688692635233, + "grad_norm": 1.6618365049362183, + "learning_rate": 1.752281661584339e-05, + "loss": 0.5553, + "step": 7994 + }, + { + "epoch": 0.25216842769279296, + "grad_norm": 3.2762527465820312, + "learning_rate": 1.752214353264627e-05, + "loss": 0.3883, + "step": 7995 + }, + { + "epoch": 0.25219996845923354, + "grad_norm": 1.7268532514572144, + "learning_rate": 1.7521470370949737e-05, + "loss": 0.5048, + "step": 7996 + }, + { + "epoch": 0.25223150922567417, + "grad_norm": 3.5139193534851074, + "learning_rate": 1.7520797130760825e-05, + "loss": 0.5389, + "step": 7997 + }, + { + "epoch": 0.2522630499921148, + "grad_norm": 1.6010996103286743, + "learning_rate": 1.7520123812086555e-05, + "loss": 0.5578, + "step": 7998 + }, + { + "epoch": 0.2522945907585554, + "grad_norm": 1.7817984819412231, + "learning_rate": 1.7519450414933953e-05, + "loss": 0.5637, + "step": 7999 + }, + { + "epoch": 0.25232613152499606, + "grad_norm": 1.9408628940582275, + "learning_rate": 1.7518776939310045e-05, + "loss": 0.5335, + "step": 8000 + }, + { + "epoch": 0.2523576722914367, + "grad_norm": 1.4435768127441406, + "learning_rate": 1.7518103385221862e-05, + "loss": 0.5583, + "step": 8001 + }, + { + "epoch": 0.2523892130578773, + "grad_norm": 1.6002776622772217, + "learning_rate": 1.7517429752676433e-05, + "loss": 0.877, + "step": 8002 + }, + { + "epoch": 0.25242075382431794, + "grad_norm": 2.4029908180236816, + "learning_rate": 1.7516756041680788e-05, + "loss": 0.5643, + "step": 8003 + }, + { + "epoch": 0.2524522945907586, + "grad_norm": 1.9497637748718262, + "learning_rate": 1.751608225224196e-05, + "loss": 0.5414, + "step": 8004 + }, + { + "epoch": 0.2524838353571992, + "grad_norm": 3.0950284004211426, + "learning_rate": 1.7515408384366974e-05, + "loss": 0.5517, + "step": 8005 + }, + { + "epoch": 0.2525153761236398, + "grad_norm": 1.6634093523025513, + "learning_rate": 1.7514734438062865e-05, + "loss": 0.5484, + "step": 8006 + }, + { + "epoch": 0.2525469168900804, + "grad_norm": 1.457532525062561, + "learning_rate": 1.751406041333667e-05, + "loss": 0.5331, + "step": 8007 + }, + { + "epoch": 0.25257845765652104, + "grad_norm": 1.8350120782852173, + "learning_rate": 1.7513386310195418e-05, + "loss": 0.5562, + "step": 8008 + }, + { + "epoch": 0.25260999842296167, + "grad_norm": 1.809683918952942, + "learning_rate": 1.7512712128646146e-05, + "loss": 0.578, + "step": 8009 + }, + { + "epoch": 0.2526415391894023, + "grad_norm": 1.5435971021652222, + "learning_rate": 1.7512037868695892e-05, + "loss": 0.5481, + "step": 8010 + }, + { + "epoch": 0.2526730799558429, + "grad_norm": 1.664350152015686, + "learning_rate": 1.7511363530351687e-05, + "loss": 0.5717, + "step": 8011 + }, + { + "epoch": 0.25270462072228356, + "grad_norm": 1.8371429443359375, + "learning_rate": 1.751068911362057e-05, + "loss": 0.5557, + "step": 8012 + }, + { + "epoch": 0.2527361614887242, + "grad_norm": 1.4434609413146973, + "learning_rate": 1.7510014618509582e-05, + "loss": 0.5429, + "step": 8013 + }, + { + "epoch": 0.2527677022551648, + "grad_norm": 2.3835182189941406, + "learning_rate": 1.7509340045025762e-05, + "loss": 0.5432, + "step": 8014 + }, + { + "epoch": 0.25279924302160545, + "grad_norm": 1.8245930671691895, + "learning_rate": 1.7508665393176145e-05, + "loss": 0.5282, + "step": 8015 + }, + { + "epoch": 0.2528307837880461, + "grad_norm": 1.7956154346466064, + "learning_rate": 1.7507990662967777e-05, + "loss": 0.5523, + "step": 8016 + }, + { + "epoch": 0.25286232455448665, + "grad_norm": 1.66368567943573, + "learning_rate": 1.7507315854407692e-05, + "loss": 0.5757, + "step": 8017 + }, + { + "epoch": 0.2528938653209273, + "grad_norm": 1.5014009475708008, + "learning_rate": 1.7506640967502942e-05, + "loss": 0.8759, + "step": 8018 + }, + { + "epoch": 0.2529254060873679, + "grad_norm": 2.1435866355895996, + "learning_rate": 1.750596600226056e-05, + "loss": 0.5484, + "step": 8019 + }, + { + "epoch": 0.25295694685380854, + "grad_norm": 1.885305643081665, + "learning_rate": 1.75052909586876e-05, + "loss": 0.5435, + "step": 8020 + }, + { + "epoch": 0.25298848762024917, + "grad_norm": 2.2006947994232178, + "learning_rate": 1.7504615836791095e-05, + "loss": 0.5601, + "step": 8021 + }, + { + "epoch": 0.2530200283866898, + "grad_norm": 1.9347296953201294, + "learning_rate": 1.7503940636578104e-05, + "loss": 0.5793, + "step": 8022 + }, + { + "epoch": 0.25305156915313043, + "grad_norm": 1.8487764596939087, + "learning_rate": 1.7503265358055657e-05, + "loss": 0.5173, + "step": 8023 + }, + { + "epoch": 0.25308310991957106, + "grad_norm": 2.677635908126831, + "learning_rate": 1.7502590001230817e-05, + "loss": 0.5457, + "step": 8024 + }, + { + "epoch": 0.2531146506860117, + "grad_norm": 2.0695290565490723, + "learning_rate": 1.750191456611062e-05, + "loss": 0.5513, + "step": 8025 + }, + { + "epoch": 0.2531461914524523, + "grad_norm": 1.6868820190429688, + "learning_rate": 1.750123905270212e-05, + "loss": 0.5756, + "step": 8026 + }, + { + "epoch": 0.2531777322188929, + "grad_norm": 1.730625033378601, + "learning_rate": 1.750056346101237e-05, + "loss": 0.5412, + "step": 8027 + }, + { + "epoch": 0.2532092729853335, + "grad_norm": 1.7485347986221313, + "learning_rate": 1.7499887791048412e-05, + "loss": 0.8811, + "step": 8028 + }, + { + "epoch": 0.25324081375177415, + "grad_norm": 1.8311165571212769, + "learning_rate": 1.7499212042817305e-05, + "loss": 0.539, + "step": 8029 + }, + { + "epoch": 0.2532723545182148, + "grad_norm": 2.3485300540924072, + "learning_rate": 1.7498536216326093e-05, + "loss": 0.5707, + "step": 8030 + }, + { + "epoch": 0.2533038952846554, + "grad_norm": 1.72393000125885, + "learning_rate": 1.7497860311581837e-05, + "loss": 0.8856, + "step": 8031 + }, + { + "epoch": 0.25333543605109604, + "grad_norm": 2.0698354244232178, + "learning_rate": 1.7497184328591584e-05, + "loss": 0.5787, + "step": 8032 + }, + { + "epoch": 0.25336697681753667, + "grad_norm": 1.8670589923858643, + "learning_rate": 1.7496508267362395e-05, + "loss": 0.542, + "step": 8033 + }, + { + "epoch": 0.2533985175839773, + "grad_norm": 4.8064284324646, + "learning_rate": 1.7495832127901318e-05, + "loss": 0.5301, + "step": 8034 + }, + { + "epoch": 0.25343005835041793, + "grad_norm": 2.0525829792022705, + "learning_rate": 1.7495155910215413e-05, + "loss": 0.5692, + "step": 8035 + }, + { + "epoch": 0.25346159911685856, + "grad_norm": 1.5867249965667725, + "learning_rate": 1.7494479614311735e-05, + "loss": 0.5814, + "step": 8036 + }, + { + "epoch": 0.2534931398832992, + "grad_norm": 3.6995651721954346, + "learning_rate": 1.7493803240197348e-05, + "loss": 0.5255, + "step": 8037 + }, + { + "epoch": 0.25352468064973976, + "grad_norm": 2.586592674255371, + "learning_rate": 1.7493126787879298e-05, + "loss": 0.5471, + "step": 8038 + }, + { + "epoch": 0.2535562214161804, + "grad_norm": 2.2966935634613037, + "learning_rate": 1.7492450257364658e-05, + "loss": 0.5517, + "step": 8039 + }, + { + "epoch": 0.253587762182621, + "grad_norm": 2.4093611240386963, + "learning_rate": 1.7491773648660477e-05, + "loss": 0.5489, + "step": 8040 + }, + { + "epoch": 0.25361930294906165, + "grad_norm": 1.6718566417694092, + "learning_rate": 1.7491096961773824e-05, + "loss": 0.5658, + "step": 8041 + }, + { + "epoch": 0.2536508437155023, + "grad_norm": 2.0138401985168457, + "learning_rate": 1.7490420196711756e-05, + "loss": 0.5593, + "step": 8042 + }, + { + "epoch": 0.2536823844819429, + "grad_norm": 1.8065022230148315, + "learning_rate": 1.7489743353481337e-05, + "loss": 0.859, + "step": 8043 + }, + { + "epoch": 0.25371392524838354, + "grad_norm": 1.740176796913147, + "learning_rate": 1.748906643208963e-05, + "loss": 0.9236, + "step": 8044 + }, + { + "epoch": 0.2537454660148242, + "grad_norm": 2.864596128463745, + "learning_rate": 1.7488389432543704e-05, + "loss": 0.5665, + "step": 8045 + }, + { + "epoch": 0.2537770067812648, + "grad_norm": 3.3238065242767334, + "learning_rate": 1.7487712354850614e-05, + "loss": 0.5775, + "step": 8046 + }, + { + "epoch": 0.25380854754770543, + "grad_norm": 2.412726402282715, + "learning_rate": 1.7487035199017434e-05, + "loss": 0.5456, + "step": 8047 + }, + { + "epoch": 0.253840088314146, + "grad_norm": 1.8138688802719116, + "learning_rate": 1.7486357965051232e-05, + "loss": 0.5421, + "step": 8048 + }, + { + "epoch": 0.25387162908058664, + "grad_norm": 2.1770009994506836, + "learning_rate": 1.7485680652959064e-05, + "loss": 0.5615, + "step": 8049 + }, + { + "epoch": 0.25390316984702727, + "grad_norm": 2.1999905109405518, + "learning_rate": 1.7485003262748012e-05, + "loss": 0.5505, + "step": 8050 + }, + { + "epoch": 0.2539347106134679, + "grad_norm": 1.9572266340255737, + "learning_rate": 1.7484325794425133e-05, + "loss": 0.572, + "step": 8051 + }, + { + "epoch": 0.2539662513799085, + "grad_norm": 1.5336567163467407, + "learning_rate": 1.748364824799751e-05, + "loss": 0.561, + "step": 8052 + }, + { + "epoch": 0.25399779214634916, + "grad_norm": 2.336376190185547, + "learning_rate": 1.74829706234722e-05, + "loss": 0.5824, + "step": 8053 + }, + { + "epoch": 0.2540293329127898, + "grad_norm": 1.6212786436080933, + "learning_rate": 1.7482292920856285e-05, + "loss": 0.5383, + "step": 8054 + }, + { + "epoch": 0.2540608736792304, + "grad_norm": 2.087266445159912, + "learning_rate": 1.7481615140156837e-05, + "loss": 0.5335, + "step": 8055 + }, + { + "epoch": 0.25409241444567104, + "grad_norm": 1.8354969024658203, + "learning_rate": 1.7480937281380915e-05, + "loss": 0.5766, + "step": 8056 + }, + { + "epoch": 0.2541239552121117, + "grad_norm": 1.6424682140350342, + "learning_rate": 1.7480259344535612e-05, + "loss": 0.572, + "step": 8057 + }, + { + "epoch": 0.2541554959785523, + "grad_norm": 1.7763242721557617, + "learning_rate": 1.7479581329627993e-05, + "loss": 0.5765, + "step": 8058 + }, + { + "epoch": 0.2541870367449929, + "grad_norm": 3.820254325866699, + "learning_rate": 1.7478903236665137e-05, + "loss": 0.8641, + "step": 8059 + }, + { + "epoch": 0.2542185775114335, + "grad_norm": 1.7005819082260132, + "learning_rate": 1.7478225065654117e-05, + "loss": 0.5526, + "step": 8060 + }, + { + "epoch": 0.25425011827787414, + "grad_norm": 3.40905499458313, + "learning_rate": 1.747754681660201e-05, + "loss": 0.5612, + "step": 8061 + }, + { + "epoch": 0.25428165904431477, + "grad_norm": 1.8000376224517822, + "learning_rate": 1.7476868489515896e-05, + "loss": 0.5522, + "step": 8062 + }, + { + "epoch": 0.2543131998107554, + "grad_norm": 7.512241840362549, + "learning_rate": 1.7476190084402852e-05, + "loss": 0.527, + "step": 8063 + }, + { + "epoch": 0.254344740577196, + "grad_norm": 2.1009621620178223, + "learning_rate": 1.7475511601269962e-05, + "loss": 0.5482, + "step": 8064 + }, + { + "epoch": 0.25437628134363666, + "grad_norm": 4.011864185333252, + "learning_rate": 1.7474833040124308e-05, + "loss": 0.3988, + "step": 8065 + }, + { + "epoch": 0.2544078221100773, + "grad_norm": 1.8151577711105347, + "learning_rate": 1.747415440097296e-05, + "loss": 0.6045, + "step": 8066 + }, + { + "epoch": 0.2544393628765179, + "grad_norm": 1.8210681676864624, + "learning_rate": 1.747347568382301e-05, + "loss": 0.5381, + "step": 8067 + }, + { + "epoch": 0.25447090364295855, + "grad_norm": 3.7596235275268555, + "learning_rate": 1.747279688868154e-05, + "loss": 0.5463, + "step": 8068 + }, + { + "epoch": 0.2545024444093992, + "grad_norm": 2.4117047786712646, + "learning_rate": 1.747211801555563e-05, + "loss": 0.53, + "step": 8069 + }, + { + "epoch": 0.25453398517583975, + "grad_norm": 2.1709132194519043, + "learning_rate": 1.7471439064452367e-05, + "loss": 0.8872, + "step": 8070 + }, + { + "epoch": 0.2545655259422804, + "grad_norm": 1.7053691148757935, + "learning_rate": 1.7470760035378834e-05, + "loss": 0.5611, + "step": 8071 + }, + { + "epoch": 0.254597066708721, + "grad_norm": 1.4070097208023071, + "learning_rate": 1.7470080928342123e-05, + "loss": 0.408, + "step": 8072 + }, + { + "epoch": 0.25462860747516164, + "grad_norm": 1.8785183429718018, + "learning_rate": 1.7469401743349313e-05, + "loss": 0.5302, + "step": 8073 + }, + { + "epoch": 0.25466014824160227, + "grad_norm": 1.8610323667526245, + "learning_rate": 1.74687224804075e-05, + "loss": 0.5673, + "step": 8074 + }, + { + "epoch": 0.2546916890080429, + "grad_norm": 2.067408561706543, + "learning_rate": 1.7468043139523766e-05, + "loss": 0.5369, + "step": 8075 + }, + { + "epoch": 0.25472322977448353, + "grad_norm": 1.742169737815857, + "learning_rate": 1.7467363720705204e-05, + "loss": 0.8399, + "step": 8076 + }, + { + "epoch": 0.25475477054092416, + "grad_norm": 1.9543219804763794, + "learning_rate": 1.7466684223958902e-05, + "loss": 0.5574, + "step": 8077 + }, + { + "epoch": 0.2547863113073648, + "grad_norm": 1.5596015453338623, + "learning_rate": 1.7466004649291953e-05, + "loss": 0.5206, + "step": 8078 + }, + { + "epoch": 0.2548178520738054, + "grad_norm": 2.101959705352783, + "learning_rate": 1.746532499671145e-05, + "loss": 0.5313, + "step": 8079 + }, + { + "epoch": 0.254849392840246, + "grad_norm": 1.9787410497665405, + "learning_rate": 1.7464645266224483e-05, + "loss": 0.5509, + "step": 8080 + }, + { + "epoch": 0.2548809336066866, + "grad_norm": 3.8107943534851074, + "learning_rate": 1.7463965457838145e-05, + "loss": 0.5372, + "step": 8081 + }, + { + "epoch": 0.25491247437312725, + "grad_norm": 2.280897378921509, + "learning_rate": 1.7463285571559535e-05, + "loss": 0.5608, + "step": 8082 + }, + { + "epoch": 0.2549440151395679, + "grad_norm": 1.7454391717910767, + "learning_rate": 1.7462605607395744e-05, + "loss": 0.5215, + "step": 8083 + }, + { + "epoch": 0.2549755559060085, + "grad_norm": 1.7371673583984375, + "learning_rate": 1.7461925565353867e-05, + "loss": 0.5629, + "step": 8084 + }, + { + "epoch": 0.25500709667244914, + "grad_norm": 1.6727961301803589, + "learning_rate": 1.7461245445441e-05, + "loss": 0.5479, + "step": 8085 + }, + { + "epoch": 0.25503863743888977, + "grad_norm": 1.612295389175415, + "learning_rate": 1.746056524766425e-05, + "loss": 0.5576, + "step": 8086 + }, + { + "epoch": 0.2550701782053304, + "grad_norm": 2.444690704345703, + "learning_rate": 1.7459884972030705e-05, + "loss": 0.555, + "step": 8087 + }, + { + "epoch": 0.25510171897177103, + "grad_norm": 1.7460863590240479, + "learning_rate": 1.7459204618547468e-05, + "loss": 0.5738, + "step": 8088 + }, + { + "epoch": 0.25513325973821166, + "grad_norm": 4.906073570251465, + "learning_rate": 1.745852418722164e-05, + "loss": 0.5298, + "step": 8089 + }, + { + "epoch": 0.2551648005046523, + "grad_norm": 2.855862855911255, + "learning_rate": 1.745784367806032e-05, + "loss": 0.5754, + "step": 8090 + }, + { + "epoch": 0.25519634127109286, + "grad_norm": 3.462924003601074, + "learning_rate": 1.745716309107061e-05, + "loss": 0.9337, + "step": 8091 + }, + { + "epoch": 0.2552278820375335, + "grad_norm": 2.60203218460083, + "learning_rate": 1.7456482426259612e-05, + "loss": 0.5565, + "step": 8092 + }, + { + "epoch": 0.2552594228039741, + "grad_norm": 2.207404851913452, + "learning_rate": 1.7455801683634433e-05, + "loss": 0.4211, + "step": 8093 + }, + { + "epoch": 0.25529096357041475, + "grad_norm": 2.115206003189087, + "learning_rate": 1.7455120863202172e-05, + "loss": 0.8946, + "step": 8094 + }, + { + "epoch": 0.2553225043368554, + "grad_norm": 2.9590892791748047, + "learning_rate": 1.745443996496994e-05, + "loss": 0.5546, + "step": 8095 + }, + { + "epoch": 0.255354045103296, + "grad_norm": 1.8834681510925293, + "learning_rate": 1.745375898894483e-05, + "loss": 0.5591, + "step": 8096 + }, + { + "epoch": 0.25538558586973664, + "grad_norm": 2.291170358657837, + "learning_rate": 1.7453077935133963e-05, + "loss": 0.5697, + "step": 8097 + }, + { + "epoch": 0.25541712663617727, + "grad_norm": 2.55552077293396, + "learning_rate": 1.745239680354444e-05, + "loss": 0.5275, + "step": 8098 + }, + { + "epoch": 0.2554486674026179, + "grad_norm": 1.8215837478637695, + "learning_rate": 1.7451715594183372e-05, + "loss": 0.5686, + "step": 8099 + }, + { + "epoch": 0.25548020816905853, + "grad_norm": 2.0555853843688965, + "learning_rate": 1.7451034307057862e-05, + "loss": 0.52, + "step": 8100 + }, + { + "epoch": 0.2555117489354991, + "grad_norm": 3.04034423828125, + "learning_rate": 1.7450352942175025e-05, + "loss": 0.9083, + "step": 8101 + }, + { + "epoch": 0.25554328970193974, + "grad_norm": 2.536306142807007, + "learning_rate": 1.744967149954197e-05, + "loss": 0.8921, + "step": 8102 + }, + { + "epoch": 0.25557483046838037, + "grad_norm": 3.232175827026367, + "learning_rate": 1.744898997916581e-05, + "loss": 0.5439, + "step": 8103 + }, + { + "epoch": 0.255606371234821, + "grad_norm": 2.633195638656616, + "learning_rate": 1.744830838105365e-05, + "loss": 0.5432, + "step": 8104 + }, + { + "epoch": 0.2556379120012616, + "grad_norm": 2.6414127349853516, + "learning_rate": 1.7447626705212615e-05, + "loss": 0.5649, + "step": 8105 + }, + { + "epoch": 0.25566945276770225, + "grad_norm": 5.763829231262207, + "learning_rate": 1.744694495164981e-05, + "loss": 0.5943, + "step": 8106 + }, + { + "epoch": 0.2557009935341429, + "grad_norm": 2.085299491882324, + "learning_rate": 1.744626312037235e-05, + "loss": 0.5469, + "step": 8107 + }, + { + "epoch": 0.2557325343005835, + "grad_norm": 1.7809113264083862, + "learning_rate": 1.7445581211387358e-05, + "loss": 0.3976, + "step": 8108 + }, + { + "epoch": 0.25576407506702414, + "grad_norm": 1.8808196783065796, + "learning_rate": 1.7444899224701937e-05, + "loss": 0.569, + "step": 8109 + }, + { + "epoch": 0.2557956158334648, + "grad_norm": 2.12770414352417, + "learning_rate": 1.744421716032322e-05, + "loss": 0.5416, + "step": 8110 + }, + { + "epoch": 0.2558271565999054, + "grad_norm": 3.9844024181365967, + "learning_rate": 1.744353501825831e-05, + "loss": 0.9638, + "step": 8111 + }, + { + "epoch": 0.255858697366346, + "grad_norm": 3.315093517303467, + "learning_rate": 1.7442852798514335e-05, + "loss": 0.938, + "step": 8112 + }, + { + "epoch": 0.2558902381327866, + "grad_norm": 2.5105972290039062, + "learning_rate": 1.7442170501098412e-05, + "loss": 0.5614, + "step": 8113 + }, + { + "epoch": 0.25592177889922724, + "grad_norm": 2.9332666397094727, + "learning_rate": 1.744148812601766e-05, + "loss": 0.5749, + "step": 8114 + }, + { + "epoch": 0.25595331966566787, + "grad_norm": 3.282339096069336, + "learning_rate": 1.7440805673279203e-05, + "loss": 0.5682, + "step": 8115 + }, + { + "epoch": 0.2559848604321085, + "grad_norm": 2.4816160202026367, + "learning_rate": 1.744012314289016e-05, + "loss": 0.5477, + "step": 8116 + }, + { + "epoch": 0.2560164011985491, + "grad_norm": 2.631711721420288, + "learning_rate": 1.7439440534857654e-05, + "loss": 0.8942, + "step": 8117 + }, + { + "epoch": 0.25604794196498976, + "grad_norm": 2.5559401512145996, + "learning_rate": 1.743875784918881e-05, + "loss": 0.5695, + "step": 8118 + }, + { + "epoch": 0.2560794827314304, + "grad_norm": 1.8056890964508057, + "learning_rate": 1.7438075085890754e-05, + "loss": 0.5512, + "step": 8119 + }, + { + "epoch": 0.256111023497871, + "grad_norm": 5.423292636871338, + "learning_rate": 1.7437392244970607e-05, + "loss": 0.5946, + "step": 8120 + }, + { + "epoch": 0.25614256426431165, + "grad_norm": 1.9321153163909912, + "learning_rate": 1.7436709326435493e-05, + "loss": 0.4001, + "step": 8121 + }, + { + "epoch": 0.2561741050307523, + "grad_norm": 1.815592646598816, + "learning_rate": 1.743602633029255e-05, + "loss": 0.5249, + "step": 8122 + }, + { + "epoch": 0.25620564579719285, + "grad_norm": 2.0245096683502197, + "learning_rate": 1.7435343256548895e-05, + "loss": 0.5399, + "step": 8123 + }, + { + "epoch": 0.2562371865636335, + "grad_norm": 3.5149130821228027, + "learning_rate": 1.743466010521166e-05, + "loss": 0.9174, + "step": 8124 + }, + { + "epoch": 0.2562687273300741, + "grad_norm": 2.1406033039093018, + "learning_rate": 1.7433976876287975e-05, + "loss": 0.5445, + "step": 8125 + }, + { + "epoch": 0.25630026809651474, + "grad_norm": 2.146378755569458, + "learning_rate": 1.7433293569784968e-05, + "loss": 0.5863, + "step": 8126 + }, + { + "epoch": 0.25633180886295537, + "grad_norm": 2.5198299884796143, + "learning_rate": 1.743261018570977e-05, + "loss": 0.5959, + "step": 8127 + }, + { + "epoch": 0.256363349629396, + "grad_norm": 2.8563125133514404, + "learning_rate": 1.7431926724069512e-05, + "loss": 0.5447, + "step": 8128 + }, + { + "epoch": 0.2563948903958366, + "grad_norm": 3.238025188446045, + "learning_rate": 1.7431243184871332e-05, + "loss": 0.5487, + "step": 8129 + }, + { + "epoch": 0.25642643116227726, + "grad_norm": 2.7217209339141846, + "learning_rate": 1.7430559568122355e-05, + "loss": 0.5536, + "step": 8130 + }, + { + "epoch": 0.2564579719287179, + "grad_norm": 1.5475656986236572, + "learning_rate": 1.742987587382972e-05, + "loss": 0.5653, + "step": 8131 + }, + { + "epoch": 0.2564895126951585, + "grad_norm": 2.0653202533721924, + "learning_rate": 1.742919210200057e-05, + "loss": 0.5366, + "step": 8132 + }, + { + "epoch": 0.2565210534615991, + "grad_norm": 2.7889692783355713, + "learning_rate": 1.742850825264202e-05, + "loss": 0.5708, + "step": 8133 + }, + { + "epoch": 0.2565525942280397, + "grad_norm": 1.6410341262817383, + "learning_rate": 1.7427824325761223e-05, + "loss": 0.5646, + "step": 8134 + }, + { + "epoch": 0.25658413499448035, + "grad_norm": 2.4092636108398438, + "learning_rate": 1.742714032136531e-05, + "loss": 0.5389, + "step": 8135 + }, + { + "epoch": 0.256615675760921, + "grad_norm": 3.1398308277130127, + "learning_rate": 1.742645623946142e-05, + "loss": 0.8906, + "step": 8136 + }, + { + "epoch": 0.2566472165273616, + "grad_norm": 2.9267115592956543, + "learning_rate": 1.7425772080056694e-05, + "loss": 0.5415, + "step": 8137 + }, + { + "epoch": 0.25667875729380224, + "grad_norm": 2.3271477222442627, + "learning_rate": 1.7425087843158274e-05, + "loss": 0.4187, + "step": 8138 + }, + { + "epoch": 0.25671029806024287, + "grad_norm": 2.6860289573669434, + "learning_rate": 1.7424403528773294e-05, + "loss": 0.5779, + "step": 8139 + }, + { + "epoch": 0.2567418388266835, + "grad_norm": 2.9159626960754395, + "learning_rate": 1.7423719136908896e-05, + "loss": 0.5703, + "step": 8140 + }, + { + "epoch": 0.25677337959312413, + "grad_norm": 1.755019187927246, + "learning_rate": 1.7423034667572228e-05, + "loss": 0.5413, + "step": 8141 + }, + { + "epoch": 0.25680492035956476, + "grad_norm": 1.7695106267929077, + "learning_rate": 1.7422350120770426e-05, + "loss": 0.535, + "step": 8142 + }, + { + "epoch": 0.2568364611260054, + "grad_norm": 1.9592339992523193, + "learning_rate": 1.742166549651064e-05, + "loss": 0.8682, + "step": 8143 + }, + { + "epoch": 0.25686800189244596, + "grad_norm": 4.924685001373291, + "learning_rate": 1.7420980794800013e-05, + "loss": 0.3887, + "step": 8144 + }, + { + "epoch": 0.2568995426588866, + "grad_norm": 1.6896436214447021, + "learning_rate": 1.7420296015645687e-05, + "loss": 0.548, + "step": 8145 + }, + { + "epoch": 0.2569310834253272, + "grad_norm": 1.8284655809402466, + "learning_rate": 1.741961115905481e-05, + "loss": 0.874, + "step": 8146 + }, + { + "epoch": 0.25696262419176785, + "grad_norm": 2.2422642707824707, + "learning_rate": 1.7418926225034528e-05, + "loss": 0.568, + "step": 8147 + }, + { + "epoch": 0.2569941649582085, + "grad_norm": 2.38613224029541, + "learning_rate": 1.7418241213591993e-05, + "loss": 0.5497, + "step": 8148 + }, + { + "epoch": 0.2570257057246491, + "grad_norm": 2.534881591796875, + "learning_rate": 1.741755612473435e-05, + "loss": 0.5408, + "step": 8149 + }, + { + "epoch": 0.25705724649108974, + "grad_norm": 1.9655946493148804, + "learning_rate": 1.7416870958468748e-05, + "loss": 0.5196, + "step": 8150 + }, + { + "epoch": 0.25708878725753037, + "grad_norm": 1.6695460081100464, + "learning_rate": 1.741618571480234e-05, + "loss": 0.5202, + "step": 8151 + }, + { + "epoch": 0.257120328023971, + "grad_norm": 1.6221081018447876, + "learning_rate": 1.7415500393742276e-05, + "loss": 0.5473, + "step": 8152 + }, + { + "epoch": 0.25715186879041163, + "grad_norm": 2.037471294403076, + "learning_rate": 1.7414814995295707e-05, + "loss": 0.5227, + "step": 8153 + }, + { + "epoch": 0.2571834095568522, + "grad_norm": 2.2338101863861084, + "learning_rate": 1.741412951946978e-05, + "loss": 0.5689, + "step": 8154 + }, + { + "epoch": 0.25721495032329283, + "grad_norm": 1.8131076097488403, + "learning_rate": 1.7413443966271665e-05, + "loss": 0.5486, + "step": 8155 + }, + { + "epoch": 0.25724649108973346, + "grad_norm": 2.0309674739837646, + "learning_rate": 1.74127583357085e-05, + "loss": 0.5652, + "step": 8156 + }, + { + "epoch": 0.2572780318561741, + "grad_norm": 2.321820020675659, + "learning_rate": 1.741207262778745e-05, + "loss": 0.5718, + "step": 8157 + }, + { + "epoch": 0.2573095726226147, + "grad_norm": 1.8566606044769287, + "learning_rate": 1.7411386842515662e-05, + "loss": 0.5739, + "step": 8158 + }, + { + "epoch": 0.25734111338905535, + "grad_norm": 1.7202856540679932, + "learning_rate": 1.7410700979900296e-05, + "loss": 0.5565, + "step": 8159 + }, + { + "epoch": 0.257372654155496, + "grad_norm": 8.232400894165039, + "learning_rate": 1.7410015039948515e-05, + "loss": 0.4162, + "step": 8160 + }, + { + "epoch": 0.2574041949219366, + "grad_norm": 2.4538135528564453, + "learning_rate": 1.740932902266747e-05, + "loss": 0.5742, + "step": 8161 + }, + { + "epoch": 0.25743573568837724, + "grad_norm": 2.4399538040161133, + "learning_rate": 1.7408642928064327e-05, + "loss": 0.398, + "step": 8162 + }, + { + "epoch": 0.2574672764548179, + "grad_norm": 2.2322258949279785, + "learning_rate": 1.7407956756146244e-05, + "loss": 0.5454, + "step": 8163 + }, + { + "epoch": 0.2574988172212585, + "grad_norm": 1.8812767267227173, + "learning_rate": 1.7407270506920376e-05, + "loss": 0.5504, + "step": 8164 + }, + { + "epoch": 0.2575303579876991, + "grad_norm": 1.9029345512390137, + "learning_rate": 1.740658418039389e-05, + "loss": 0.5668, + "step": 8165 + }, + { + "epoch": 0.2575618987541397, + "grad_norm": 2.0197203159332275, + "learning_rate": 1.740589777657395e-05, + "loss": 0.8735, + "step": 8166 + }, + { + "epoch": 0.25759343952058034, + "grad_norm": 1.9588783979415894, + "learning_rate": 1.740521129546771e-05, + "loss": 0.5294, + "step": 8167 + }, + { + "epoch": 0.25762498028702097, + "grad_norm": 1.841134786605835, + "learning_rate": 1.7404524737082343e-05, + "loss": 0.5598, + "step": 8168 + }, + { + "epoch": 0.2576565210534616, + "grad_norm": 1.8803315162658691, + "learning_rate": 1.740383810142501e-05, + "loss": 0.5426, + "step": 8169 + }, + { + "epoch": 0.2576880618199022, + "grad_norm": 1.7691580057144165, + "learning_rate": 1.7403151388502882e-05, + "loss": 0.5384, + "step": 8170 + }, + { + "epoch": 0.25771960258634286, + "grad_norm": 1.7942430973052979, + "learning_rate": 1.7402464598323118e-05, + "loss": 0.872, + "step": 8171 + }, + { + "epoch": 0.2577511433527835, + "grad_norm": 1.918691635131836, + "learning_rate": 1.740177773089289e-05, + "loss": 0.5802, + "step": 8172 + }, + { + "epoch": 0.2577826841192241, + "grad_norm": 1.5706909894943237, + "learning_rate": 1.7401090786219357e-05, + "loss": 0.8913, + "step": 8173 + }, + { + "epoch": 0.25781422488566474, + "grad_norm": 1.6749905347824097, + "learning_rate": 1.7400403764309702e-05, + "loss": 0.5291, + "step": 8174 + }, + { + "epoch": 0.2578457656521053, + "grad_norm": 1.7312620878219604, + "learning_rate": 1.7399716665171083e-05, + "loss": 0.5633, + "step": 8175 + }, + { + "epoch": 0.25787730641854595, + "grad_norm": 2.2148923873901367, + "learning_rate": 1.739902948881068e-05, + "loss": 0.5399, + "step": 8176 + }, + { + "epoch": 0.2579088471849866, + "grad_norm": 1.791763424873352, + "learning_rate": 1.7398342235235653e-05, + "loss": 0.5656, + "step": 8177 + }, + { + "epoch": 0.2579403879514272, + "grad_norm": 1.8707232475280762, + "learning_rate": 1.7397654904453184e-05, + "loss": 0.5599, + "step": 8178 + }, + { + "epoch": 0.25797192871786784, + "grad_norm": 2.5249578952789307, + "learning_rate": 1.739696749647044e-05, + "loss": 0.4056, + "step": 8179 + }, + { + "epoch": 0.25800346948430847, + "grad_norm": 1.62041437625885, + "learning_rate": 1.7396280011294598e-05, + "loss": 0.5572, + "step": 8180 + }, + { + "epoch": 0.2580350102507491, + "grad_norm": 1.945928692817688, + "learning_rate": 1.7395592448932834e-05, + "loss": 0.5544, + "step": 8181 + }, + { + "epoch": 0.2580665510171897, + "grad_norm": 1.782224416732788, + "learning_rate": 1.7394904809392317e-05, + "loss": 0.5222, + "step": 8182 + }, + { + "epoch": 0.25809809178363036, + "grad_norm": 1.9981932640075684, + "learning_rate": 1.7394217092680224e-05, + "loss": 0.5479, + "step": 8183 + }, + { + "epoch": 0.258129632550071, + "grad_norm": 1.7301965951919556, + "learning_rate": 1.7393529298803735e-05, + "loss": 0.8882, + "step": 8184 + }, + { + "epoch": 0.2581611733165116, + "grad_norm": 1.9006208181381226, + "learning_rate": 1.739284142777003e-05, + "loss": 0.503, + "step": 8185 + }, + { + "epoch": 0.2581927140829522, + "grad_norm": 1.9050800800323486, + "learning_rate": 1.739215347958628e-05, + "loss": 0.557, + "step": 8186 + }, + { + "epoch": 0.2582242548493928, + "grad_norm": 2.1312081813812256, + "learning_rate": 1.7391465454259675e-05, + "loss": 0.5653, + "step": 8187 + }, + { + "epoch": 0.25825579561583345, + "grad_norm": 1.4295860528945923, + "learning_rate": 1.7390777351797384e-05, + "loss": 0.9057, + "step": 8188 + }, + { + "epoch": 0.2582873363822741, + "grad_norm": 1.896990180015564, + "learning_rate": 1.7390089172206594e-05, + "loss": 0.5489, + "step": 8189 + }, + { + "epoch": 0.2583188771487147, + "grad_norm": 3.0178771018981934, + "learning_rate": 1.7389400915494483e-05, + "loss": 0.5806, + "step": 8190 + }, + { + "epoch": 0.25835041791515534, + "grad_norm": 1.7538377046585083, + "learning_rate": 1.7388712581668236e-05, + "loss": 0.5418, + "step": 8191 + }, + { + "epoch": 0.25838195868159597, + "grad_norm": 1.379908800125122, + "learning_rate": 1.738802417073504e-05, + "loss": 0.3949, + "step": 8192 + }, + { + "epoch": 0.2584134994480366, + "grad_norm": 1.679268479347229, + "learning_rate": 1.7387335682702072e-05, + "loss": 0.567, + "step": 8193 + }, + { + "epoch": 0.25844504021447723, + "grad_norm": 2.130614995956421, + "learning_rate": 1.738664711757652e-05, + "loss": 0.5787, + "step": 8194 + }, + { + "epoch": 0.25847658098091786, + "grad_norm": 2.025974750518799, + "learning_rate": 1.738595847536557e-05, + "loss": 0.544, + "step": 8195 + }, + { + "epoch": 0.2585081217473585, + "grad_norm": 1.9281558990478516, + "learning_rate": 1.7385269756076408e-05, + "loss": 0.5895, + "step": 8196 + }, + { + "epoch": 0.25853966251379906, + "grad_norm": 1.6526554822921753, + "learning_rate": 1.7384580959716225e-05, + "loss": 0.5338, + "step": 8197 + }, + { + "epoch": 0.2585712032802397, + "grad_norm": 2.4081666469573975, + "learning_rate": 1.73838920862922e-05, + "loss": 0.5514, + "step": 8198 + }, + { + "epoch": 0.2586027440466803, + "grad_norm": 1.7987213134765625, + "learning_rate": 1.738320313581153e-05, + "loss": 0.5128, + "step": 8199 + }, + { + "epoch": 0.25863428481312095, + "grad_norm": 1.4586246013641357, + "learning_rate": 1.7382514108281405e-05, + "loss": 0.4014, + "step": 8200 + }, + { + "epoch": 0.2586658255795616, + "grad_norm": 1.69736647605896, + "learning_rate": 1.738182500370901e-05, + "loss": 0.5396, + "step": 8201 + }, + { + "epoch": 0.2586973663460022, + "grad_norm": 2.258662223815918, + "learning_rate": 1.7381135822101536e-05, + "loss": 0.5471, + "step": 8202 + }, + { + "epoch": 0.25872890711244284, + "grad_norm": 2.0908899307250977, + "learning_rate": 1.7380446563466184e-05, + "loss": 0.5412, + "step": 8203 + }, + { + "epoch": 0.25876044787888347, + "grad_norm": 1.3211361169815063, + "learning_rate": 1.737975722781014e-05, + "loss": 0.5536, + "step": 8204 + }, + { + "epoch": 0.2587919886453241, + "grad_norm": 1.66392183303833, + "learning_rate": 1.73790678151406e-05, + "loss": 0.5445, + "step": 8205 + }, + { + "epoch": 0.25882352941176473, + "grad_norm": 2.0289695262908936, + "learning_rate": 1.737837832546476e-05, + "loss": 0.9146, + "step": 8206 + }, + { + "epoch": 0.2588550701782053, + "grad_norm": 1.8797633647918701, + "learning_rate": 1.7377688758789805e-05, + "loss": 0.5712, + "step": 8207 + }, + { + "epoch": 0.25888661094464593, + "grad_norm": 2.056478500366211, + "learning_rate": 1.7376999115122942e-05, + "loss": 0.5382, + "step": 8208 + }, + { + "epoch": 0.25891815171108656, + "grad_norm": 1.7134485244750977, + "learning_rate": 1.7376309394471368e-05, + "loss": 0.8561, + "step": 8209 + }, + { + "epoch": 0.2589496924775272, + "grad_norm": 1.838043212890625, + "learning_rate": 1.7375619596842276e-05, + "loss": 0.5349, + "step": 8210 + }, + { + "epoch": 0.2589812332439678, + "grad_norm": 1.736246109008789, + "learning_rate": 1.737492972224287e-05, + "loss": 0.5366, + "step": 8211 + }, + { + "epoch": 0.25901277401040845, + "grad_norm": 1.6729857921600342, + "learning_rate": 1.7374239770680344e-05, + "loss": 0.5281, + "step": 8212 + }, + { + "epoch": 0.2590443147768491, + "grad_norm": 1.461666226387024, + "learning_rate": 1.7373549742161896e-05, + "loss": 0.8501, + "step": 8213 + }, + { + "epoch": 0.2590758555432897, + "grad_norm": 1.8949846029281616, + "learning_rate": 1.7372859636694733e-05, + "loss": 0.5613, + "step": 8214 + }, + { + "epoch": 0.25910739630973034, + "grad_norm": 2.3960812091827393, + "learning_rate": 1.7372169454286058e-05, + "loss": 0.524, + "step": 8215 + }, + { + "epoch": 0.25913893707617097, + "grad_norm": 1.6741580963134766, + "learning_rate": 1.7371479194943068e-05, + "loss": 0.5348, + "step": 8216 + }, + { + "epoch": 0.2591704778426116, + "grad_norm": 1.8147798776626587, + "learning_rate": 1.737078885867297e-05, + "loss": 0.5356, + "step": 8217 + }, + { + "epoch": 0.2592020186090522, + "grad_norm": 2.052562713623047, + "learning_rate": 1.7370098445482965e-05, + "loss": 0.5451, + "step": 8218 + }, + { + "epoch": 0.2592335593754928, + "grad_norm": 1.9690288305282593, + "learning_rate": 1.736940795538026e-05, + "loss": 0.5427, + "step": 8219 + }, + { + "epoch": 0.25926510014193344, + "grad_norm": 1.8127590417861938, + "learning_rate": 1.7368717388372065e-05, + "loss": 0.5298, + "step": 8220 + }, + { + "epoch": 0.25929664090837407, + "grad_norm": 1.9194706678390503, + "learning_rate": 1.7368026744465577e-05, + "loss": 0.5574, + "step": 8221 + }, + { + "epoch": 0.2593281816748147, + "grad_norm": 2.6929380893707275, + "learning_rate": 1.7367336023668012e-05, + "loss": 0.5299, + "step": 8222 + }, + { + "epoch": 0.2593597224412553, + "grad_norm": 1.8176599740982056, + "learning_rate": 1.7366645225986575e-05, + "loss": 0.5412, + "step": 8223 + }, + { + "epoch": 0.25939126320769595, + "grad_norm": 1.5971604585647583, + "learning_rate": 1.736595435142847e-05, + "loss": 0.3905, + "step": 8224 + }, + { + "epoch": 0.2594228039741366, + "grad_norm": 2.3927547931671143, + "learning_rate": 1.7365263400000918e-05, + "loss": 0.553, + "step": 8225 + }, + { + "epoch": 0.2594543447405772, + "grad_norm": 1.7024004459381104, + "learning_rate": 1.7364572371711122e-05, + "loss": 0.5487, + "step": 8226 + }, + { + "epoch": 0.25948588550701784, + "grad_norm": 1.8459203243255615, + "learning_rate": 1.7363881266566296e-05, + "loss": 0.5329, + "step": 8227 + }, + { + "epoch": 0.2595174262734584, + "grad_norm": 2.39003324508667, + "learning_rate": 1.7363190084573647e-05, + "loss": 0.5589, + "step": 8228 + }, + { + "epoch": 0.25954896703989905, + "grad_norm": 5.6166558265686035, + "learning_rate": 1.7362498825740395e-05, + "loss": 0.9093, + "step": 8229 + }, + { + "epoch": 0.2595805078063397, + "grad_norm": 1.799070119857788, + "learning_rate": 1.736180749007375e-05, + "loss": 0.5572, + "step": 8230 + }, + { + "epoch": 0.2596120485727803, + "grad_norm": 6.724463939666748, + "learning_rate": 1.736111607758093e-05, + "loss": 0.5369, + "step": 8231 + }, + { + "epoch": 0.25964358933922094, + "grad_norm": 1.4585366249084473, + "learning_rate": 1.7360424588269142e-05, + "loss": 0.5653, + "step": 8232 + }, + { + "epoch": 0.25967513010566157, + "grad_norm": 1.717145562171936, + "learning_rate": 1.7359733022145615e-05, + "loss": 0.5904, + "step": 8233 + }, + { + "epoch": 0.2597066708721022, + "grad_norm": 1.4591717720031738, + "learning_rate": 1.7359041379217555e-05, + "loss": 0.5167, + "step": 8234 + }, + { + "epoch": 0.2597382116385428, + "grad_norm": 1.660102128982544, + "learning_rate": 1.7358349659492186e-05, + "loss": 0.5503, + "step": 8235 + }, + { + "epoch": 0.25976975240498346, + "grad_norm": 1.6848856210708618, + "learning_rate": 1.7357657862976724e-05, + "loss": 0.5514, + "step": 8236 + }, + { + "epoch": 0.2598012931714241, + "grad_norm": 1.7219310998916626, + "learning_rate": 1.735696598967839e-05, + "loss": 0.572, + "step": 8237 + }, + { + "epoch": 0.2598328339378647, + "grad_norm": 1.6387529373168945, + "learning_rate": 1.73562740396044e-05, + "loss": 0.5098, + "step": 8238 + }, + { + "epoch": 0.2598643747043053, + "grad_norm": 1.5570940971374512, + "learning_rate": 1.7355582012761982e-05, + "loss": 0.52, + "step": 8239 + }, + { + "epoch": 0.2598959154707459, + "grad_norm": 2.510401487350464, + "learning_rate": 1.7354889909158354e-05, + "loss": 0.5447, + "step": 8240 + }, + { + "epoch": 0.25992745623718655, + "grad_norm": 1.763014793395996, + "learning_rate": 1.7354197728800737e-05, + "loss": 0.5438, + "step": 8241 + }, + { + "epoch": 0.2599589970036272, + "grad_norm": 3.3852202892303467, + "learning_rate": 1.7353505471696358e-05, + "loss": 0.6034, + "step": 8242 + }, + { + "epoch": 0.2599905377700678, + "grad_norm": 1.3299168348312378, + "learning_rate": 1.7352813137852437e-05, + "loss": 0.5363, + "step": 8243 + }, + { + "epoch": 0.26002207853650844, + "grad_norm": 1.5534762144088745, + "learning_rate": 1.73521207272762e-05, + "loss": 0.5092, + "step": 8244 + }, + { + "epoch": 0.26005361930294907, + "grad_norm": 1.996166706085205, + "learning_rate": 1.7351428239974884e-05, + "loss": 0.5364, + "step": 8245 + }, + { + "epoch": 0.2600851600693897, + "grad_norm": 1.8001388311386108, + "learning_rate": 1.7350735675955696e-05, + "loss": 0.5616, + "step": 8246 + }, + { + "epoch": 0.26011670083583033, + "grad_norm": 1.8435947895050049, + "learning_rate": 1.735004303522588e-05, + "loss": 0.5319, + "step": 8247 + }, + { + "epoch": 0.26014824160227096, + "grad_norm": 1.9938316345214844, + "learning_rate": 1.7349350317792654e-05, + "loss": 0.5539, + "step": 8248 + }, + { + "epoch": 0.2601797823687116, + "grad_norm": 2.5696020126342773, + "learning_rate": 1.7348657523663254e-05, + "loss": 0.5316, + "step": 8249 + }, + { + "epoch": 0.26021132313515216, + "grad_norm": 2.011624574661255, + "learning_rate": 1.7347964652844903e-05, + "loss": 0.5352, + "step": 8250 + }, + { + "epoch": 0.2602428639015928, + "grad_norm": 3.911817789077759, + "learning_rate": 1.734727170534484e-05, + "loss": 0.8685, + "step": 8251 + }, + { + "epoch": 0.2602744046680334, + "grad_norm": 2.556647777557373, + "learning_rate": 1.7346578681170287e-05, + "loss": 0.433, + "step": 8252 + }, + { + "epoch": 0.26030594543447405, + "grad_norm": 2.188424587249756, + "learning_rate": 1.734588558032848e-05, + "loss": 0.5599, + "step": 8253 + }, + { + "epoch": 0.2603374862009147, + "grad_norm": 1.7777155637741089, + "learning_rate": 1.734519240282666e-05, + "loss": 0.5386, + "step": 8254 + }, + { + "epoch": 0.2603690269673553, + "grad_norm": 2.2380895614624023, + "learning_rate": 1.7344499148672047e-05, + "loss": 0.5548, + "step": 8255 + }, + { + "epoch": 0.26040056773379594, + "grad_norm": 1.482012152671814, + "learning_rate": 1.7343805817871885e-05, + "loss": 0.5215, + "step": 8256 + }, + { + "epoch": 0.26043210850023657, + "grad_norm": 1.7593228816986084, + "learning_rate": 1.734311241043341e-05, + "loss": 0.5518, + "step": 8257 + }, + { + "epoch": 0.2604636492666772, + "grad_norm": 1.313977837562561, + "learning_rate": 1.734241892636385e-05, + "loss": 0.5184, + "step": 8258 + }, + { + "epoch": 0.26049519003311783, + "grad_norm": 2.0060317516326904, + "learning_rate": 1.734172536567045e-05, + "loss": 0.5398, + "step": 8259 + }, + { + "epoch": 0.2605267307995584, + "grad_norm": 1.4875891208648682, + "learning_rate": 1.7341031728360443e-05, + "loss": 0.5468, + "step": 8260 + }, + { + "epoch": 0.26055827156599903, + "grad_norm": 1.8071942329406738, + "learning_rate": 1.7340338014441073e-05, + "loss": 0.5789, + "step": 8261 + }, + { + "epoch": 0.26058981233243966, + "grad_norm": 3.444323778152466, + "learning_rate": 1.7339644223919573e-05, + "loss": 0.5461, + "step": 8262 + }, + { + "epoch": 0.2606213530988803, + "grad_norm": 3.686602830886841, + "learning_rate": 1.733895035680319e-05, + "loss": 0.5513, + "step": 8263 + }, + { + "epoch": 0.2606528938653209, + "grad_norm": 1.76663339138031, + "learning_rate": 1.733825641309916e-05, + "loss": 0.5591, + "step": 8264 + }, + { + "epoch": 0.26068443463176155, + "grad_norm": 1.442857027053833, + "learning_rate": 1.7337562392814725e-05, + "loss": 0.5287, + "step": 8265 + }, + { + "epoch": 0.2607159753982022, + "grad_norm": 6.101930618286133, + "learning_rate": 1.7336868295957133e-05, + "loss": 0.552, + "step": 8266 + }, + { + "epoch": 0.2607475161646428, + "grad_norm": 1.370273232460022, + "learning_rate": 1.7336174122533616e-05, + "loss": 0.5549, + "step": 8267 + }, + { + "epoch": 0.26077905693108344, + "grad_norm": 1.628576636314392, + "learning_rate": 1.7335479872551434e-05, + "loss": 0.5818, + "step": 8268 + }, + { + "epoch": 0.26081059769752407, + "grad_norm": 3.1067492961883545, + "learning_rate": 1.733478554601782e-05, + "loss": 0.9058, + "step": 8269 + }, + { + "epoch": 0.2608421384639647, + "grad_norm": 1.9081429243087769, + "learning_rate": 1.7334091142940018e-05, + "loss": 0.6053, + "step": 8270 + }, + { + "epoch": 0.2608736792304053, + "grad_norm": 1.4351178407669067, + "learning_rate": 1.733339666332529e-05, + "loss": 0.5304, + "step": 8271 + }, + { + "epoch": 0.2609052199968459, + "grad_norm": 1.7026495933532715, + "learning_rate": 1.733270210718087e-05, + "loss": 0.5602, + "step": 8272 + }, + { + "epoch": 0.26093676076328653, + "grad_norm": 1.5397002696990967, + "learning_rate": 1.7332007474514005e-05, + "loss": 0.5456, + "step": 8273 + }, + { + "epoch": 0.26096830152972716, + "grad_norm": 1.6471494436264038, + "learning_rate": 1.7331312765331955e-05, + "loss": 0.5685, + "step": 8274 + }, + { + "epoch": 0.2609998422961678, + "grad_norm": 1.5339614152908325, + "learning_rate": 1.733061797964196e-05, + "loss": 0.5363, + "step": 8275 + }, + { + "epoch": 0.2610313830626084, + "grad_norm": 1.8392894268035889, + "learning_rate": 1.7329923117451274e-05, + "loss": 0.424, + "step": 8276 + }, + { + "epoch": 0.26106292382904905, + "grad_norm": 1.6650638580322266, + "learning_rate": 1.732922817876715e-05, + "loss": 0.5354, + "step": 8277 + }, + { + "epoch": 0.2610944645954897, + "grad_norm": 1.6976667642593384, + "learning_rate": 1.732853316359684e-05, + "loss": 0.5554, + "step": 8278 + }, + { + "epoch": 0.2611260053619303, + "grad_norm": 1.81255042552948, + "learning_rate": 1.7327838071947596e-05, + "loss": 0.5411, + "step": 8279 + }, + { + "epoch": 0.26115754612837094, + "grad_norm": 1.557135820388794, + "learning_rate": 1.732714290382667e-05, + "loss": 0.8851, + "step": 8280 + }, + { + "epoch": 0.2611890868948115, + "grad_norm": 2.0301787853240967, + "learning_rate": 1.732644765924132e-05, + "loss": 0.5649, + "step": 8281 + }, + { + "epoch": 0.26122062766125215, + "grad_norm": 1.7686368227005005, + "learning_rate": 1.7325752338198797e-05, + "loss": 0.5143, + "step": 8282 + }, + { + "epoch": 0.2612521684276928, + "grad_norm": 1.6997731924057007, + "learning_rate": 1.7325056940706362e-05, + "loss": 0.5649, + "step": 8283 + }, + { + "epoch": 0.2612837091941334, + "grad_norm": 2.0131919384002686, + "learning_rate": 1.7324361466771275e-05, + "loss": 0.4342, + "step": 8284 + }, + { + "epoch": 0.26131524996057404, + "grad_norm": 1.9686863422393799, + "learning_rate": 1.7323665916400784e-05, + "loss": 0.5261, + "step": 8285 + }, + { + "epoch": 0.26134679072701467, + "grad_norm": 1.532086968421936, + "learning_rate": 1.732297028960215e-05, + "loss": 0.5235, + "step": 8286 + }, + { + "epoch": 0.2613783314934553, + "grad_norm": 2.2886760234832764, + "learning_rate": 1.7322274586382638e-05, + "loss": 0.5845, + "step": 8287 + }, + { + "epoch": 0.2614098722598959, + "grad_norm": 1.4166510105133057, + "learning_rate": 1.7321578806749504e-05, + "loss": 0.5381, + "step": 8288 + }, + { + "epoch": 0.26144141302633656, + "grad_norm": 1.9178231954574585, + "learning_rate": 1.7320882950710013e-05, + "loss": 0.5673, + "step": 8289 + }, + { + "epoch": 0.2614729537927772, + "grad_norm": 2.2859127521514893, + "learning_rate": 1.732018701827142e-05, + "loss": 0.5291, + "step": 8290 + }, + { + "epoch": 0.2615044945592178, + "grad_norm": 1.8916682004928589, + "learning_rate": 1.7319491009440993e-05, + "loss": 0.5441, + "step": 8291 + }, + { + "epoch": 0.2615360353256584, + "grad_norm": 1.4574238061904907, + "learning_rate": 1.7318794924225995e-05, + "loss": 0.5602, + "step": 8292 + }, + { + "epoch": 0.261567576092099, + "grad_norm": 1.8488571643829346, + "learning_rate": 1.7318098762633687e-05, + "loss": 0.5448, + "step": 8293 + }, + { + "epoch": 0.26159911685853965, + "grad_norm": 1.7037147283554077, + "learning_rate": 1.7317402524671337e-05, + "loss": 0.8838, + "step": 8294 + }, + { + "epoch": 0.2616306576249803, + "grad_norm": 2.2392568588256836, + "learning_rate": 1.731670621034621e-05, + "loss": 0.8619, + "step": 8295 + }, + { + "epoch": 0.2616621983914209, + "grad_norm": 2.0061542987823486, + "learning_rate": 1.7316009819665573e-05, + "loss": 0.5348, + "step": 8296 + }, + { + "epoch": 0.26169373915786154, + "grad_norm": 2.55086612701416, + "learning_rate": 1.731531335263669e-05, + "loss": 0.5429, + "step": 8297 + }, + { + "epoch": 0.26172527992430217, + "grad_norm": 1.5568190813064575, + "learning_rate": 1.7314616809266837e-05, + "loss": 0.5448, + "step": 8298 + }, + { + "epoch": 0.2617568206907428, + "grad_norm": 1.8800503015518188, + "learning_rate": 1.7313920189563275e-05, + "loss": 0.5465, + "step": 8299 + }, + { + "epoch": 0.2617883614571834, + "grad_norm": 1.5122146606445312, + "learning_rate": 1.7313223493533276e-05, + "loss": 0.542, + "step": 8300 + }, + { + "epoch": 0.26181990222362406, + "grad_norm": 2.5064914226531982, + "learning_rate": 1.7312526721184114e-05, + "loss": 0.5771, + "step": 8301 + }, + { + "epoch": 0.26185144299006463, + "grad_norm": 2.2238059043884277, + "learning_rate": 1.7311829872523054e-05, + "loss": 0.5599, + "step": 8302 + }, + { + "epoch": 0.26188298375650526, + "grad_norm": 1.8058749437332153, + "learning_rate": 1.7311132947557373e-05, + "loss": 0.5647, + "step": 8303 + }, + { + "epoch": 0.2619145245229459, + "grad_norm": 1.9659992456436157, + "learning_rate": 1.7310435946294345e-05, + "loss": 0.527, + "step": 8304 + }, + { + "epoch": 0.2619460652893865, + "grad_norm": 2.1659739017486572, + "learning_rate": 1.7309738868741238e-05, + "loss": 0.4191, + "step": 8305 + }, + { + "epoch": 0.26197760605582715, + "grad_norm": 1.8791435956954956, + "learning_rate": 1.7309041714905333e-05, + "loss": 0.59, + "step": 8306 + }, + { + "epoch": 0.2620091468222678, + "grad_norm": 3.0882601737976074, + "learning_rate": 1.73083444847939e-05, + "loss": 0.5347, + "step": 8307 + }, + { + "epoch": 0.2620406875887084, + "grad_norm": 2.3114285469055176, + "learning_rate": 1.7307647178414218e-05, + "loss": 0.546, + "step": 8308 + }, + { + "epoch": 0.26207222835514904, + "grad_norm": 2.7298660278320312, + "learning_rate": 1.7306949795773565e-05, + "loss": 0.5158, + "step": 8309 + }, + { + "epoch": 0.26210376912158967, + "grad_norm": 3.2373311519622803, + "learning_rate": 1.7306252336879216e-05, + "loss": 0.5423, + "step": 8310 + }, + { + "epoch": 0.2621353098880303, + "grad_norm": 1.5553100109100342, + "learning_rate": 1.730555480173845e-05, + "loss": 0.5542, + "step": 8311 + }, + { + "epoch": 0.26216685065447093, + "grad_norm": 1.9964778423309326, + "learning_rate": 1.730485719035855e-05, + "loss": 0.5432, + "step": 8312 + }, + { + "epoch": 0.2621983914209115, + "grad_norm": 2.0645902156829834, + "learning_rate": 1.7304159502746788e-05, + "loss": 0.5614, + "step": 8313 + }, + { + "epoch": 0.26222993218735213, + "grad_norm": 1.5554814338684082, + "learning_rate": 1.7303461738910457e-05, + "loss": 0.5386, + "step": 8314 + }, + { + "epoch": 0.26226147295379276, + "grad_norm": 2.2856247425079346, + "learning_rate": 1.7302763898856824e-05, + "loss": 0.5438, + "step": 8315 + }, + { + "epoch": 0.2622930137202334, + "grad_norm": 2.437453269958496, + "learning_rate": 1.7302065982593187e-05, + "loss": 0.5395, + "step": 8316 + }, + { + "epoch": 0.262324554486674, + "grad_norm": 2.0269052982330322, + "learning_rate": 1.7301367990126817e-05, + "loss": 0.8955, + "step": 8317 + }, + { + "epoch": 0.26235609525311465, + "grad_norm": 1.9824695587158203, + "learning_rate": 1.7300669921465002e-05, + "loss": 0.5873, + "step": 8318 + }, + { + "epoch": 0.2623876360195553, + "grad_norm": 3.4315857887268066, + "learning_rate": 1.729997177661503e-05, + "loss": 0.5143, + "step": 8319 + }, + { + "epoch": 0.2624191767859959, + "grad_norm": 1.7830357551574707, + "learning_rate": 1.7299273555584186e-05, + "loss": 0.5122, + "step": 8320 + }, + { + "epoch": 0.26245071755243654, + "grad_norm": 1.9267648458480835, + "learning_rate": 1.729857525837975e-05, + "loss": 0.5666, + "step": 8321 + }, + { + "epoch": 0.26248225831887717, + "grad_norm": 1.764571189880371, + "learning_rate": 1.7297876885009016e-05, + "loss": 0.5072, + "step": 8322 + }, + { + "epoch": 0.2625137990853178, + "grad_norm": 2.522796154022217, + "learning_rate": 1.729717843547927e-05, + "loss": 0.5555, + "step": 8323 + }, + { + "epoch": 0.2625453398517584, + "grad_norm": 2.045708179473877, + "learning_rate": 1.72964799097978e-05, + "loss": 0.5461, + "step": 8324 + }, + { + "epoch": 0.262576880618199, + "grad_norm": 2.5301551818847656, + "learning_rate": 1.72957813079719e-05, + "loss": 0.57, + "step": 8325 + }, + { + "epoch": 0.26260842138463963, + "grad_norm": 2.2793548107147217, + "learning_rate": 1.7295082630008853e-05, + "loss": 0.5336, + "step": 8326 + }, + { + "epoch": 0.26263996215108026, + "grad_norm": 1.8763326406478882, + "learning_rate": 1.7294383875915956e-05, + "loss": 0.5498, + "step": 8327 + }, + { + "epoch": 0.2626715029175209, + "grad_norm": 1.8943419456481934, + "learning_rate": 1.72936850457005e-05, + "loss": 0.5238, + "step": 8328 + }, + { + "epoch": 0.2627030436839615, + "grad_norm": 1.918520212173462, + "learning_rate": 1.7292986139369775e-05, + "loss": 0.5401, + "step": 8329 + }, + { + "epoch": 0.26273458445040215, + "grad_norm": 1.8295363187789917, + "learning_rate": 1.7292287156931076e-05, + "loss": 0.5421, + "step": 8330 + }, + { + "epoch": 0.2627661252168428, + "grad_norm": 2.365753412246704, + "learning_rate": 1.72915880983917e-05, + "loss": 0.5595, + "step": 8331 + }, + { + "epoch": 0.2627976659832834, + "grad_norm": 3.063760757446289, + "learning_rate": 1.7290888963758944e-05, + "loss": 0.5395, + "step": 8332 + }, + { + "epoch": 0.26282920674972404, + "grad_norm": 1.4747767448425293, + "learning_rate": 1.7290189753040096e-05, + "loss": 0.5367, + "step": 8333 + }, + { + "epoch": 0.2628607475161646, + "grad_norm": 2.0228090286254883, + "learning_rate": 1.728949046624246e-05, + "loss": 0.5744, + "step": 8334 + }, + { + "epoch": 0.26289228828260525, + "grad_norm": 2.0745303630828857, + "learning_rate": 1.728879110337333e-05, + "loss": 0.527, + "step": 8335 + }, + { + "epoch": 0.2629238290490459, + "grad_norm": 1.7256284952163696, + "learning_rate": 1.728809166444e-05, + "loss": 0.5516, + "step": 8336 + }, + { + "epoch": 0.2629553698154865, + "grad_norm": 2.361514091491699, + "learning_rate": 1.728739214944978e-05, + "loss": 0.5711, + "step": 8337 + }, + { + "epoch": 0.26298691058192714, + "grad_norm": 2.1369431018829346, + "learning_rate": 1.7286692558409967e-05, + "loss": 0.5675, + "step": 8338 + }, + { + "epoch": 0.26301845134836777, + "grad_norm": 2.1603212356567383, + "learning_rate": 1.7285992891327852e-05, + "loss": 0.5344, + "step": 8339 + }, + { + "epoch": 0.2630499921148084, + "grad_norm": 1.6661583185195923, + "learning_rate": 1.7285293148210748e-05, + "loss": 0.416, + "step": 8340 + }, + { + "epoch": 0.263081532881249, + "grad_norm": 2.085557222366333, + "learning_rate": 1.7284593329065952e-05, + "loss": 0.5491, + "step": 8341 + }, + { + "epoch": 0.26311307364768965, + "grad_norm": 1.964309811592102, + "learning_rate": 1.728389343390077e-05, + "loss": 0.5499, + "step": 8342 + }, + { + "epoch": 0.2631446144141303, + "grad_norm": 4.999919891357422, + "learning_rate": 1.72831934627225e-05, + "loss": 0.5274, + "step": 8343 + }, + { + "epoch": 0.2631761551805709, + "grad_norm": 1.7248473167419434, + "learning_rate": 1.7282493415538453e-05, + "loss": 0.8975, + "step": 8344 + }, + { + "epoch": 0.2632076959470115, + "grad_norm": 1.9329206943511963, + "learning_rate": 1.7281793292355934e-05, + "loss": 0.566, + "step": 8345 + }, + { + "epoch": 0.2632392367134521, + "grad_norm": 1.7375373840332031, + "learning_rate": 1.7281093093182247e-05, + "loss": 0.5548, + "step": 8346 + }, + { + "epoch": 0.26327077747989275, + "grad_norm": 1.4507712125778198, + "learning_rate": 1.7280392818024704e-05, + "loss": 0.5958, + "step": 8347 + }, + { + "epoch": 0.2633023182463334, + "grad_norm": 1.9816951751708984, + "learning_rate": 1.7279692466890603e-05, + "loss": 0.5397, + "step": 8348 + }, + { + "epoch": 0.263333859012774, + "grad_norm": 1.3096635341644287, + "learning_rate": 1.727899203978726e-05, + "loss": 0.5154, + "step": 8349 + }, + { + "epoch": 0.26336539977921464, + "grad_norm": 1.799622893333435, + "learning_rate": 1.7278291536721982e-05, + "loss": 0.5551, + "step": 8350 + }, + { + "epoch": 0.26339694054565527, + "grad_norm": 1.6753346920013428, + "learning_rate": 1.7277590957702082e-05, + "loss": 0.8458, + "step": 8351 + }, + { + "epoch": 0.2634284813120959, + "grad_norm": 1.604994297027588, + "learning_rate": 1.727689030273487e-05, + "loss": 0.5346, + "step": 8352 + }, + { + "epoch": 0.2634600220785365, + "grad_norm": 2.250757932662964, + "learning_rate": 1.7276189571827656e-05, + "loss": 0.5868, + "step": 8353 + }, + { + "epoch": 0.26349156284497716, + "grad_norm": 2.229444980621338, + "learning_rate": 1.7275488764987755e-05, + "loss": 0.5221, + "step": 8354 + }, + { + "epoch": 0.26352310361141773, + "grad_norm": 3.8391053676605225, + "learning_rate": 1.727478788222248e-05, + "loss": 0.5527, + "step": 8355 + }, + { + "epoch": 0.26355464437785836, + "grad_norm": 2.733729839324951, + "learning_rate": 1.7274086923539145e-05, + "loss": 0.5542, + "step": 8356 + }, + { + "epoch": 0.263586185144299, + "grad_norm": 1.3977100849151611, + "learning_rate": 1.727338588894506e-05, + "loss": 0.5825, + "step": 8357 + }, + { + "epoch": 0.2636177259107396, + "grad_norm": 1.4961730241775513, + "learning_rate": 1.727268477844755e-05, + "loss": 0.5599, + "step": 8358 + }, + { + "epoch": 0.26364926667718025, + "grad_norm": 1.9812755584716797, + "learning_rate": 1.7271983592053926e-05, + "loss": 0.5472, + "step": 8359 + }, + { + "epoch": 0.2636808074436209, + "grad_norm": 1.4761871099472046, + "learning_rate": 1.727128232977151e-05, + "loss": 0.5285, + "step": 8360 + }, + { + "epoch": 0.2637123482100615, + "grad_norm": 1.3068314790725708, + "learning_rate": 1.7270580991607613e-05, + "loss": 0.4208, + "step": 8361 + }, + { + "epoch": 0.26374388897650214, + "grad_norm": 2.1876330375671387, + "learning_rate": 1.726987957756956e-05, + "loss": 0.5176, + "step": 8362 + }, + { + "epoch": 0.26377542974294277, + "grad_norm": 1.6910245418548584, + "learning_rate": 1.7269178087664668e-05, + "loss": 0.575, + "step": 8363 + }, + { + "epoch": 0.2638069705093834, + "grad_norm": 4.788868427276611, + "learning_rate": 1.726847652190026e-05, + "loss": 0.5517, + "step": 8364 + }, + { + "epoch": 0.26383851127582403, + "grad_norm": 3.8935739994049072, + "learning_rate": 1.726777488028365e-05, + "loss": 0.5575, + "step": 8365 + }, + { + "epoch": 0.2638700520422646, + "grad_norm": 1.5111198425292969, + "learning_rate": 1.7267073162822168e-05, + "loss": 0.5764, + "step": 8366 + }, + { + "epoch": 0.26390159280870523, + "grad_norm": 1.991891622543335, + "learning_rate": 1.726637136952314e-05, + "loss": 0.5473, + "step": 8367 + }, + { + "epoch": 0.26393313357514586, + "grad_norm": 2.1397387981414795, + "learning_rate": 1.7265669500393877e-05, + "loss": 0.559, + "step": 8368 + }, + { + "epoch": 0.2639646743415865, + "grad_norm": 1.7007765769958496, + "learning_rate": 1.7264967555441718e-05, + "loss": 0.8551, + "step": 8369 + }, + { + "epoch": 0.2639962151080271, + "grad_norm": 2.1759471893310547, + "learning_rate": 1.7264265534673977e-05, + "loss": 0.5138, + "step": 8370 + }, + { + "epoch": 0.26402775587446775, + "grad_norm": 2.637361526489258, + "learning_rate": 1.7263563438097984e-05, + "loss": 0.5546, + "step": 8371 + }, + { + "epoch": 0.2640592966409084, + "grad_norm": 8.695575714111328, + "learning_rate": 1.7262861265721073e-05, + "loss": 0.4233, + "step": 8372 + }, + { + "epoch": 0.264090837407349, + "grad_norm": 1.688001275062561, + "learning_rate": 1.726215901755056e-05, + "loss": 0.5596, + "step": 8373 + }, + { + "epoch": 0.26412237817378964, + "grad_norm": 1.5825262069702148, + "learning_rate": 1.7261456693593776e-05, + "loss": 0.5559, + "step": 8374 + }, + { + "epoch": 0.26415391894023027, + "grad_norm": 2.232239246368408, + "learning_rate": 1.7260754293858057e-05, + "loss": 0.5741, + "step": 8375 + }, + { + "epoch": 0.2641854597066709, + "grad_norm": 1.7457319498062134, + "learning_rate": 1.7260051818350728e-05, + "loss": 0.5488, + "step": 8376 + }, + { + "epoch": 0.2642170004731115, + "grad_norm": 2.1623928546905518, + "learning_rate": 1.7259349267079124e-05, + "loss": 0.5138, + "step": 8377 + }, + { + "epoch": 0.2642485412395521, + "grad_norm": 1.8443827629089355, + "learning_rate": 1.725864664005057e-05, + "loss": 0.519, + "step": 8378 + }, + { + "epoch": 0.26428008200599273, + "grad_norm": 1.827744722366333, + "learning_rate": 1.7257943937272402e-05, + "loss": 0.5342, + "step": 8379 + }, + { + "epoch": 0.26431162277243336, + "grad_norm": 1.7329450845718384, + "learning_rate": 1.7257241158751955e-05, + "loss": 0.8205, + "step": 8380 + }, + { + "epoch": 0.264343163538874, + "grad_norm": 1.5876030921936035, + "learning_rate": 1.725653830449656e-05, + "loss": 0.5644, + "step": 8381 + }, + { + "epoch": 0.2643747043053146, + "grad_norm": 2.2122464179992676, + "learning_rate": 1.7255835374513555e-05, + "loss": 0.5518, + "step": 8382 + }, + { + "epoch": 0.26440624507175525, + "grad_norm": 1.8820992708206177, + "learning_rate": 1.7255132368810276e-05, + "loss": 0.5433, + "step": 8383 + }, + { + "epoch": 0.2644377858381959, + "grad_norm": 2.0802013874053955, + "learning_rate": 1.7254429287394052e-05, + "loss": 0.559, + "step": 8384 + }, + { + "epoch": 0.2644693266046365, + "grad_norm": 3.8119957447052, + "learning_rate": 1.725372613027223e-05, + "loss": 0.54, + "step": 8385 + }, + { + "epoch": 0.26450086737107714, + "grad_norm": 1.6717355251312256, + "learning_rate": 1.7253022897452142e-05, + "loss": 0.5511, + "step": 8386 + }, + { + "epoch": 0.2645324081375177, + "grad_norm": 1.6855535507202148, + "learning_rate": 1.7252319588941127e-05, + "loss": 0.8599, + "step": 8387 + }, + { + "epoch": 0.26456394890395835, + "grad_norm": 2.1490981578826904, + "learning_rate": 1.7251616204746525e-05, + "loss": 0.5262, + "step": 8388 + }, + { + "epoch": 0.264595489670399, + "grad_norm": 4.499973297119141, + "learning_rate": 1.725091274487568e-05, + "loss": 0.5583, + "step": 8389 + }, + { + "epoch": 0.2646270304368396, + "grad_norm": 2.3106560707092285, + "learning_rate": 1.725020920933593e-05, + "loss": 0.5564, + "step": 8390 + }, + { + "epoch": 0.26465857120328024, + "grad_norm": 2.1333630084991455, + "learning_rate": 1.7249505598134615e-05, + "loss": 0.5572, + "step": 8391 + }, + { + "epoch": 0.26469011196972086, + "grad_norm": 4.266940593719482, + "learning_rate": 1.724880191127908e-05, + "loss": 0.9068, + "step": 8392 + }, + { + "epoch": 0.2647216527361615, + "grad_norm": 1.676161527633667, + "learning_rate": 1.724809814877667e-05, + "loss": 0.5519, + "step": 8393 + }, + { + "epoch": 0.2647531935026021, + "grad_norm": 1.5602794885635376, + "learning_rate": 1.7247394310634728e-05, + "loss": 0.8541, + "step": 8394 + }, + { + "epoch": 0.26478473426904275, + "grad_norm": 2.1859073638916016, + "learning_rate": 1.72466903968606e-05, + "loss": 0.421, + "step": 8395 + }, + { + "epoch": 0.2648162750354834, + "grad_norm": 2.3400630950927734, + "learning_rate": 1.7245986407461627e-05, + "loss": 0.591, + "step": 8396 + }, + { + "epoch": 0.264847815801924, + "grad_norm": 4.2918853759765625, + "learning_rate": 1.724528234244516e-05, + "loss": 0.5513, + "step": 8397 + }, + { + "epoch": 0.2648793565683646, + "grad_norm": 1.7682064771652222, + "learning_rate": 1.724457820181855e-05, + "loss": 0.5377, + "step": 8398 + }, + { + "epoch": 0.2649108973348052, + "grad_norm": 2.129096746444702, + "learning_rate": 1.7243873985589135e-05, + "loss": 0.5622, + "step": 8399 + }, + { + "epoch": 0.26494243810124585, + "grad_norm": 3.074846029281616, + "learning_rate": 1.7243169693764277e-05, + "loss": 0.5527, + "step": 8400 + }, + { + "epoch": 0.2649739788676865, + "grad_norm": 1.4590034484863281, + "learning_rate": 1.7242465326351316e-05, + "loss": 0.8768, + "step": 8401 + }, + { + "epoch": 0.2650055196341271, + "grad_norm": 1.575154423713684, + "learning_rate": 1.7241760883357605e-05, + "loss": 0.39, + "step": 8402 + }, + { + "epoch": 0.26503706040056774, + "grad_norm": 3.1799254417419434, + "learning_rate": 1.72410563647905e-05, + "loss": 0.5571, + "step": 8403 + }, + { + "epoch": 0.26506860116700837, + "grad_norm": 2.4771933555603027, + "learning_rate": 1.7240351770657343e-05, + "loss": 0.5636, + "step": 8404 + }, + { + "epoch": 0.265100141933449, + "grad_norm": 1.8796049356460571, + "learning_rate": 1.72396471009655e-05, + "loss": 0.5769, + "step": 8405 + }, + { + "epoch": 0.2651316826998896, + "grad_norm": 1.7841306924819946, + "learning_rate": 1.723894235572231e-05, + "loss": 0.5278, + "step": 8406 + }, + { + "epoch": 0.26516322346633026, + "grad_norm": 2.9118645191192627, + "learning_rate": 1.7238237534935142e-05, + "loss": 0.5189, + "step": 8407 + }, + { + "epoch": 0.26519476423277083, + "grad_norm": 2.0602667331695557, + "learning_rate": 1.7237532638611342e-05, + "loss": 0.5662, + "step": 8408 + }, + { + "epoch": 0.26522630499921146, + "grad_norm": 2.5194709300994873, + "learning_rate": 1.7236827666758272e-05, + "loss": 0.5428, + "step": 8409 + }, + { + "epoch": 0.2652578457656521, + "grad_norm": 4.523859977722168, + "learning_rate": 1.7236122619383284e-05, + "loss": 0.5638, + "step": 8410 + }, + { + "epoch": 0.2652893865320927, + "grad_norm": 1.9592033624649048, + "learning_rate": 1.723541749649374e-05, + "loss": 0.5491, + "step": 8411 + }, + { + "epoch": 0.26532092729853335, + "grad_norm": 6.597458839416504, + "learning_rate": 1.7234712298096995e-05, + "loss": 0.573, + "step": 8412 + }, + { + "epoch": 0.265352468064974, + "grad_norm": 2.750115394592285, + "learning_rate": 1.7234007024200406e-05, + "loss": 0.5329, + "step": 8413 + }, + { + "epoch": 0.2653840088314146, + "grad_norm": 1.553728461265564, + "learning_rate": 1.723330167481134e-05, + "loss": 0.552, + "step": 8414 + }, + { + "epoch": 0.26541554959785524, + "grad_norm": 1.604196548461914, + "learning_rate": 1.7232596249937153e-05, + "loss": 0.5488, + "step": 8415 + }, + { + "epoch": 0.26544709036429587, + "grad_norm": 1.763942003250122, + "learning_rate": 1.7231890749585208e-05, + "loss": 0.5339, + "step": 8416 + }, + { + "epoch": 0.2654786311307365, + "grad_norm": 3.402832508087158, + "learning_rate": 1.723118517376287e-05, + "loss": 0.575, + "step": 8417 + }, + { + "epoch": 0.2655101718971771, + "grad_norm": 1.7106677293777466, + "learning_rate": 1.7230479522477498e-05, + "loss": 0.86, + "step": 8418 + }, + { + "epoch": 0.2655417126636177, + "grad_norm": 2.0470287799835205, + "learning_rate": 1.722977379573646e-05, + "loss": 0.5557, + "step": 8419 + }, + { + "epoch": 0.26557325343005833, + "grad_norm": 3.0286788940429688, + "learning_rate": 1.722906799354712e-05, + "loss": 0.5462, + "step": 8420 + }, + { + "epoch": 0.26560479419649896, + "grad_norm": 1.7459577322006226, + "learning_rate": 1.722836211591684e-05, + "loss": 0.8852, + "step": 8421 + }, + { + "epoch": 0.2656363349629396, + "grad_norm": 1.753676176071167, + "learning_rate": 1.7227656162852984e-05, + "loss": 0.5295, + "step": 8422 + }, + { + "epoch": 0.2656678757293802, + "grad_norm": 4.147209167480469, + "learning_rate": 1.722695013436293e-05, + "loss": 0.5544, + "step": 8423 + }, + { + "epoch": 0.26569941649582085, + "grad_norm": 1.8387376070022583, + "learning_rate": 1.7226244030454034e-05, + "loss": 0.5595, + "step": 8424 + }, + { + "epoch": 0.2657309572622615, + "grad_norm": 2.8917975425720215, + "learning_rate": 1.722553785113368e-05, + "loss": 0.5621, + "step": 8425 + }, + { + "epoch": 0.2657624980287021, + "grad_norm": 2.0543925762176514, + "learning_rate": 1.722483159640922e-05, + "loss": 0.5165, + "step": 8426 + }, + { + "epoch": 0.26579403879514274, + "grad_norm": 7.152740955352783, + "learning_rate": 1.7224125266288038e-05, + "loss": 0.5447, + "step": 8427 + }, + { + "epoch": 0.26582557956158337, + "grad_norm": 1.540380835533142, + "learning_rate": 1.722341886077749e-05, + "loss": 0.5305, + "step": 8428 + }, + { + "epoch": 0.26585712032802394, + "grad_norm": 1.6918392181396484, + "learning_rate": 1.7222712379884966e-05, + "loss": 0.5263, + "step": 8429 + }, + { + "epoch": 0.2658886610944646, + "grad_norm": 2.128417730331421, + "learning_rate": 1.722200582361783e-05, + "loss": 0.5561, + "step": 8430 + }, + { + "epoch": 0.2659202018609052, + "grad_norm": 1.9845960140228271, + "learning_rate": 1.722129919198345e-05, + "loss": 0.5442, + "step": 8431 + }, + { + "epoch": 0.26595174262734583, + "grad_norm": 1.7295844554901123, + "learning_rate": 1.722059248498921e-05, + "loss": 0.8909, + "step": 8432 + }, + { + "epoch": 0.26598328339378646, + "grad_norm": 2.03440260887146, + "learning_rate": 1.721988570264248e-05, + "loss": 0.5674, + "step": 8433 + }, + { + "epoch": 0.2660148241602271, + "grad_norm": 13.759145736694336, + "learning_rate": 1.7219178844950635e-05, + "loss": 0.4067, + "step": 8434 + }, + { + "epoch": 0.2660463649266677, + "grad_norm": 3.6622722148895264, + "learning_rate": 1.7218471911921053e-05, + "loss": 0.5282, + "step": 8435 + }, + { + "epoch": 0.26607790569310835, + "grad_norm": 1.5117524862289429, + "learning_rate": 1.7217764903561115e-05, + "loss": 0.5379, + "step": 8436 + }, + { + "epoch": 0.266109446459549, + "grad_norm": 1.8728605508804321, + "learning_rate": 1.721705781987819e-05, + "loss": 0.5288, + "step": 8437 + }, + { + "epoch": 0.2661409872259896, + "grad_norm": 3.079024314880371, + "learning_rate": 1.721635066087967e-05, + "loss": 0.5511, + "step": 8438 + }, + { + "epoch": 0.26617252799243024, + "grad_norm": 1.683858871459961, + "learning_rate": 1.7215643426572924e-05, + "loss": 0.544, + "step": 8439 + }, + { + "epoch": 0.2662040687588708, + "grad_norm": 1.542323112487793, + "learning_rate": 1.721493611696533e-05, + "loss": 0.896, + "step": 8440 + }, + { + "epoch": 0.26623560952531145, + "grad_norm": 4.291437149047852, + "learning_rate": 1.7214228732064286e-05, + "loss": 0.8991, + "step": 8441 + }, + { + "epoch": 0.2662671502917521, + "grad_norm": 1.5199123620986938, + "learning_rate": 1.7213521271877157e-05, + "loss": 0.8539, + "step": 8442 + }, + { + "epoch": 0.2662986910581927, + "grad_norm": 2.072359323501587, + "learning_rate": 1.7212813736411337e-05, + "loss": 0.5102, + "step": 8443 + }, + { + "epoch": 0.26633023182463333, + "grad_norm": 2.125553607940674, + "learning_rate": 1.72121061256742e-05, + "loss": 0.5807, + "step": 8444 + }, + { + "epoch": 0.26636177259107396, + "grad_norm": 1.8778189420700073, + "learning_rate": 1.721139843967314e-05, + "loss": 0.5183, + "step": 8445 + }, + { + "epoch": 0.2663933133575146, + "grad_norm": 2.672844171524048, + "learning_rate": 1.7210690678415532e-05, + "loss": 0.4049, + "step": 8446 + }, + { + "epoch": 0.2664248541239552, + "grad_norm": 2.2759504318237305, + "learning_rate": 1.720998284190877e-05, + "loss": 0.5407, + "step": 8447 + }, + { + "epoch": 0.26645639489039585, + "grad_norm": 2.298233985900879, + "learning_rate": 1.7209274930160243e-05, + "loss": 0.574, + "step": 8448 + }, + { + "epoch": 0.2664879356568365, + "grad_norm": 2.7467212677001953, + "learning_rate": 1.7208566943177333e-05, + "loss": 0.5396, + "step": 8449 + }, + { + "epoch": 0.2665194764232771, + "grad_norm": 1.8654788732528687, + "learning_rate": 1.720785888096743e-05, + "loss": 0.5316, + "step": 8450 + }, + { + "epoch": 0.2665510171897177, + "grad_norm": 1.9224543571472168, + "learning_rate": 1.720715074353792e-05, + "loss": 0.5561, + "step": 8451 + }, + { + "epoch": 0.2665825579561583, + "grad_norm": 3.159149408340454, + "learning_rate": 1.7206442530896192e-05, + "loss": 0.5439, + "step": 8452 + }, + { + "epoch": 0.26661409872259895, + "grad_norm": 2.116903781890869, + "learning_rate": 1.7205734243049646e-05, + "loss": 0.5671, + "step": 8453 + }, + { + "epoch": 0.2666456394890396, + "grad_norm": 2.0340659618377686, + "learning_rate": 1.7205025880005667e-05, + "loss": 0.4376, + "step": 8454 + }, + { + "epoch": 0.2666771802554802, + "grad_norm": 2.0265917778015137, + "learning_rate": 1.720431744177165e-05, + "loss": 0.532, + "step": 8455 + }, + { + "epoch": 0.26670872102192084, + "grad_norm": 3.4764387607574463, + "learning_rate": 1.7203608928354986e-05, + "loss": 0.524, + "step": 8456 + }, + { + "epoch": 0.26674026178836147, + "grad_norm": 2.21304988861084, + "learning_rate": 1.7202900339763066e-05, + "loss": 0.5623, + "step": 8457 + }, + { + "epoch": 0.2667718025548021, + "grad_norm": 2.7473552227020264, + "learning_rate": 1.720219167600329e-05, + "loss": 0.9177, + "step": 8458 + }, + { + "epoch": 0.2668033433212427, + "grad_norm": 2.5138092041015625, + "learning_rate": 1.7201482937083048e-05, + "loss": 0.5423, + "step": 8459 + }, + { + "epoch": 0.26683488408768335, + "grad_norm": 2.5387651920318604, + "learning_rate": 1.720077412300975e-05, + "loss": 0.5476, + "step": 8460 + }, + { + "epoch": 0.26686642485412393, + "grad_norm": 1.9558541774749756, + "learning_rate": 1.7200065233790772e-05, + "loss": 0.5498, + "step": 8461 + }, + { + "epoch": 0.26689796562056456, + "grad_norm": 2.3128552436828613, + "learning_rate": 1.719935626943353e-05, + "loss": 0.5657, + "step": 8462 + }, + { + "epoch": 0.2669295063870052, + "grad_norm": 1.6004085540771484, + "learning_rate": 1.719864722994541e-05, + "loss": 0.8468, + "step": 8463 + }, + { + "epoch": 0.2669610471534458, + "grad_norm": 3.3129019737243652, + "learning_rate": 1.719793811533382e-05, + "loss": 0.5455, + "step": 8464 + }, + { + "epoch": 0.26699258791988645, + "grad_norm": 2.4654641151428223, + "learning_rate": 1.7197228925606155e-05, + "loss": 0.5448, + "step": 8465 + }, + { + "epoch": 0.2670241286863271, + "grad_norm": 2.056013584136963, + "learning_rate": 1.719651966076982e-05, + "loss": 0.5449, + "step": 8466 + }, + { + "epoch": 0.2670556694527677, + "grad_norm": 2.307459831237793, + "learning_rate": 1.719581032083221e-05, + "loss": 0.5835, + "step": 8467 + }, + { + "epoch": 0.26708721021920834, + "grad_norm": 1.9082086086273193, + "learning_rate": 1.7195100905800734e-05, + "loss": 0.5418, + "step": 8468 + }, + { + "epoch": 0.26711875098564897, + "grad_norm": 2.119614362716675, + "learning_rate": 1.7194391415682792e-05, + "loss": 0.4165, + "step": 8469 + }, + { + "epoch": 0.2671502917520896, + "grad_norm": 3.5978188514709473, + "learning_rate": 1.719368185048579e-05, + "loss": 0.5486, + "step": 8470 + }, + { + "epoch": 0.2671818325185302, + "grad_norm": 2.4459006786346436, + "learning_rate": 1.7192972210217135e-05, + "loss": 0.5643, + "step": 8471 + }, + { + "epoch": 0.2672133732849708, + "grad_norm": 2.125305652618408, + "learning_rate": 1.7192262494884228e-05, + "loss": 0.5815, + "step": 8472 + }, + { + "epoch": 0.26724491405141143, + "grad_norm": 2.437995195388794, + "learning_rate": 1.7191552704494475e-05, + "loss": 0.5646, + "step": 8473 + }, + { + "epoch": 0.26727645481785206, + "grad_norm": 2.0472052097320557, + "learning_rate": 1.719084283905529e-05, + "loss": 0.5705, + "step": 8474 + }, + { + "epoch": 0.2673079955842927, + "grad_norm": 2.061509370803833, + "learning_rate": 1.7190132898574073e-05, + "loss": 0.8488, + "step": 8475 + }, + { + "epoch": 0.2673395363507333, + "grad_norm": 2.1205241680145264, + "learning_rate": 1.7189422883058236e-05, + "loss": 0.5426, + "step": 8476 + }, + { + "epoch": 0.26737107711717395, + "grad_norm": 2.5074269771575928, + "learning_rate": 1.718871279251519e-05, + "loss": 0.5647, + "step": 8477 + }, + { + "epoch": 0.2674026178836146, + "grad_norm": 2.478722333908081, + "learning_rate": 1.7188002626952345e-05, + "loss": 0.5726, + "step": 8478 + }, + { + "epoch": 0.2674341586500552, + "grad_norm": 2.5593209266662598, + "learning_rate": 1.7187292386377107e-05, + "loss": 0.5586, + "step": 8479 + }, + { + "epoch": 0.26746569941649584, + "grad_norm": 1.671086072921753, + "learning_rate": 1.7186582070796897e-05, + "loss": 0.3744, + "step": 8480 + }, + { + "epoch": 0.26749724018293647, + "grad_norm": 2.1896374225616455, + "learning_rate": 1.718587168021912e-05, + "loss": 0.5438, + "step": 8481 + }, + { + "epoch": 0.26752878094937704, + "grad_norm": 1.3801313638687134, + "learning_rate": 1.7185161214651196e-05, + "loss": 0.8522, + "step": 8482 + }, + { + "epoch": 0.2675603217158177, + "grad_norm": 2.561091423034668, + "learning_rate": 1.7184450674100533e-05, + "loss": 0.5695, + "step": 8483 + }, + { + "epoch": 0.2675918624822583, + "grad_norm": 2.96368408203125, + "learning_rate": 1.7183740058574546e-05, + "loss": 0.5593, + "step": 8484 + }, + { + "epoch": 0.26762340324869893, + "grad_norm": 1.9023610353469849, + "learning_rate": 1.7183029368080658e-05, + "loss": 0.5577, + "step": 8485 + }, + { + "epoch": 0.26765494401513956, + "grad_norm": 2.6619882583618164, + "learning_rate": 1.718231860262628e-05, + "loss": 0.5869, + "step": 8486 + }, + { + "epoch": 0.2676864847815802, + "grad_norm": 2.5086193084716797, + "learning_rate": 1.718160776221883e-05, + "loss": 0.8977, + "step": 8487 + }, + { + "epoch": 0.2677180255480208, + "grad_norm": 2.2576208114624023, + "learning_rate": 1.7180896846865725e-05, + "loss": 0.5689, + "step": 8488 + }, + { + "epoch": 0.26774956631446145, + "grad_norm": 2.3284828662872314, + "learning_rate": 1.7180185856574385e-05, + "loss": 0.5153, + "step": 8489 + }, + { + "epoch": 0.2677811070809021, + "grad_norm": 2.054291248321533, + "learning_rate": 1.7179474791352233e-05, + "loss": 0.5414, + "step": 8490 + }, + { + "epoch": 0.2678126478473427, + "grad_norm": 1.7953063249588013, + "learning_rate": 1.7178763651206686e-05, + "loss": 0.5459, + "step": 8491 + }, + { + "epoch": 0.26784418861378334, + "grad_norm": 1.7706714868545532, + "learning_rate": 1.717805243614517e-05, + "loss": 0.4935, + "step": 8492 + }, + { + "epoch": 0.2678757293802239, + "grad_norm": 1.7368160486221313, + "learning_rate": 1.7177341146175096e-05, + "loss": 0.5513, + "step": 8493 + }, + { + "epoch": 0.26790727014666454, + "grad_norm": 1.7827774286270142, + "learning_rate": 1.71766297813039e-05, + "loss": 0.5394, + "step": 8494 + }, + { + "epoch": 0.2679388109131052, + "grad_norm": 2.382394790649414, + "learning_rate": 1.7175918341538994e-05, + "loss": 0.5363, + "step": 8495 + }, + { + "epoch": 0.2679703516795458, + "grad_norm": 1.7496267557144165, + "learning_rate": 1.717520682688781e-05, + "loss": 0.5222, + "step": 8496 + }, + { + "epoch": 0.26800189244598643, + "grad_norm": 1.897981882095337, + "learning_rate": 1.7174495237357774e-05, + "loss": 0.5283, + "step": 8497 + }, + { + "epoch": 0.26803343321242706, + "grad_norm": 2.180158853530884, + "learning_rate": 1.717378357295631e-05, + "loss": 0.5614, + "step": 8498 + }, + { + "epoch": 0.2680649739788677, + "grad_norm": 1.8134269714355469, + "learning_rate": 1.717307183369084e-05, + "loss": 0.5406, + "step": 8499 + }, + { + "epoch": 0.2680965147453083, + "grad_norm": 1.5639241933822632, + "learning_rate": 1.7172360019568798e-05, + "loss": 0.8738, + "step": 8500 + }, + { + "epoch": 0.26812805551174895, + "grad_norm": 2.007192850112915, + "learning_rate": 1.717164813059761e-05, + "loss": 0.5597, + "step": 8501 + }, + { + "epoch": 0.2681595962781896, + "grad_norm": 1.8805861473083496, + "learning_rate": 1.7170936166784704e-05, + "loss": 0.4089, + "step": 8502 + }, + { + "epoch": 0.26819113704463016, + "grad_norm": 2.432129144668579, + "learning_rate": 1.7170224128137514e-05, + "loss": 0.5672, + "step": 8503 + }, + { + "epoch": 0.2682226778110708, + "grad_norm": 3.212786912918091, + "learning_rate": 1.7169512014663467e-05, + "loss": 0.5662, + "step": 8504 + }, + { + "epoch": 0.2682542185775114, + "grad_norm": 2.647183418273926, + "learning_rate": 1.716879982636999e-05, + "loss": 0.556, + "step": 8505 + }, + { + "epoch": 0.26828575934395205, + "grad_norm": 1.9378286600112915, + "learning_rate": 1.7168087563264527e-05, + "loss": 0.5371, + "step": 8506 + }, + { + "epoch": 0.2683173001103927, + "grad_norm": 1.5436666011810303, + "learning_rate": 1.71673752253545e-05, + "loss": 0.5192, + "step": 8507 + }, + { + "epoch": 0.2683488408768333, + "grad_norm": 2.0657758712768555, + "learning_rate": 1.716666281264735e-05, + "loss": 0.5672, + "step": 8508 + }, + { + "epoch": 0.26838038164327394, + "grad_norm": 2.846400499343872, + "learning_rate": 1.7165950325150508e-05, + "loss": 0.5422, + "step": 8509 + }, + { + "epoch": 0.26841192240971457, + "grad_norm": 1.9759387969970703, + "learning_rate": 1.716523776287141e-05, + "loss": 0.5728, + "step": 8510 + }, + { + "epoch": 0.2684434631761552, + "grad_norm": 2.0810835361480713, + "learning_rate": 1.716452512581749e-05, + "loss": 0.5546, + "step": 8511 + }, + { + "epoch": 0.2684750039425958, + "grad_norm": 1.8243342638015747, + "learning_rate": 1.716381241399619e-05, + "loss": 0.549, + "step": 8512 + }, + { + "epoch": 0.26850654470903645, + "grad_norm": 4.123532295227051, + "learning_rate": 1.716309962741494e-05, + "loss": 0.5644, + "step": 8513 + }, + { + "epoch": 0.26853808547547703, + "grad_norm": 2.558276653289795, + "learning_rate": 1.716238676608119e-05, + "loss": 0.5631, + "step": 8514 + }, + { + "epoch": 0.26856962624191766, + "grad_norm": 1.6997554302215576, + "learning_rate": 1.716167383000237e-05, + "loss": 0.4172, + "step": 8515 + }, + { + "epoch": 0.2686011670083583, + "grad_norm": 1.4790903329849243, + "learning_rate": 1.7160960819185924e-05, + "loss": 0.5277, + "step": 8516 + }, + { + "epoch": 0.2686327077747989, + "grad_norm": 2.577902317047119, + "learning_rate": 1.716024773363929e-05, + "loss": 0.5596, + "step": 8517 + }, + { + "epoch": 0.26866424854123955, + "grad_norm": 1.8493444919586182, + "learning_rate": 1.715953457336991e-05, + "loss": 0.5386, + "step": 8518 + }, + { + "epoch": 0.2686957893076802, + "grad_norm": 1.9362602233886719, + "learning_rate": 1.715882133838523e-05, + "loss": 0.5487, + "step": 8519 + }, + { + "epoch": 0.2687273300741208, + "grad_norm": 1.7710520029067993, + "learning_rate": 1.715810802869269e-05, + "loss": 0.4059, + "step": 8520 + }, + { + "epoch": 0.26875887084056144, + "grad_norm": 1.6514866352081299, + "learning_rate": 1.7157394644299734e-05, + "loss": 0.4051, + "step": 8521 + }, + { + "epoch": 0.26879041160700207, + "grad_norm": 1.822918176651001, + "learning_rate": 1.715668118521381e-05, + "loss": 0.5184, + "step": 8522 + }, + { + "epoch": 0.2688219523734427, + "grad_norm": 1.813090443611145, + "learning_rate": 1.7155967651442356e-05, + "loss": 0.5449, + "step": 8523 + }, + { + "epoch": 0.2688534931398833, + "grad_norm": 1.9380087852478027, + "learning_rate": 1.7155254042992827e-05, + "loss": 0.5216, + "step": 8524 + }, + { + "epoch": 0.2688850339063239, + "grad_norm": 1.564749836921692, + "learning_rate": 1.7154540359872664e-05, + "loss": 0.4981, + "step": 8525 + }, + { + "epoch": 0.26891657467276453, + "grad_norm": 1.7493853569030762, + "learning_rate": 1.7153826602089318e-05, + "loss": 0.536, + "step": 8526 + }, + { + "epoch": 0.26894811543920516, + "grad_norm": 1.832444190979004, + "learning_rate": 1.7153112769650233e-05, + "loss": 0.5293, + "step": 8527 + }, + { + "epoch": 0.2689796562056458, + "grad_norm": 2.1979775428771973, + "learning_rate": 1.7152398862562866e-05, + "loss": 0.5649, + "step": 8528 + }, + { + "epoch": 0.2690111969720864, + "grad_norm": 2.075533151626587, + "learning_rate": 1.715168488083466e-05, + "loss": 0.5344, + "step": 8529 + }, + { + "epoch": 0.26904273773852705, + "grad_norm": 2.068662405014038, + "learning_rate": 1.7150970824473075e-05, + "loss": 0.5653, + "step": 8530 + }, + { + "epoch": 0.2690742785049677, + "grad_norm": 2.221172332763672, + "learning_rate": 1.7150256693485553e-05, + "loss": 0.5591, + "step": 8531 + }, + { + "epoch": 0.2691058192714083, + "grad_norm": 2.364182710647583, + "learning_rate": 1.714954248787955e-05, + "loss": 0.5412, + "step": 8532 + }, + { + "epoch": 0.26913736003784894, + "grad_norm": 1.9468008279800415, + "learning_rate": 1.714882820766252e-05, + "loss": 0.5425, + "step": 8533 + }, + { + "epoch": 0.26916890080428957, + "grad_norm": 2.1444032192230225, + "learning_rate": 1.7148113852841918e-05, + "loss": 0.5414, + "step": 8534 + }, + { + "epoch": 0.26920044157073014, + "grad_norm": 1.9872764348983765, + "learning_rate": 1.7147399423425196e-05, + "loss": 0.5556, + "step": 8535 + }, + { + "epoch": 0.26923198233717077, + "grad_norm": 2.920369863510132, + "learning_rate": 1.7146684919419813e-05, + "loss": 0.5487, + "step": 8536 + }, + { + "epoch": 0.2692635231036114, + "grad_norm": 1.4551376104354858, + "learning_rate": 1.714597034083322e-05, + "loss": 0.528, + "step": 8537 + }, + { + "epoch": 0.26929506387005203, + "grad_norm": 1.905037522315979, + "learning_rate": 1.714525568767288e-05, + "loss": 0.5408, + "step": 8538 + }, + { + "epoch": 0.26932660463649266, + "grad_norm": 2.049788475036621, + "learning_rate": 1.7144540959946252e-05, + "loss": 0.5496, + "step": 8539 + }, + { + "epoch": 0.2693581454029333, + "grad_norm": 1.6414955854415894, + "learning_rate": 1.7143826157660788e-05, + "loss": 0.5297, + "step": 8540 + }, + { + "epoch": 0.2693896861693739, + "grad_norm": 2.777175188064575, + "learning_rate": 1.714311128082395e-05, + "loss": 0.4304, + "step": 8541 + }, + { + "epoch": 0.26942122693581455, + "grad_norm": 8.024937629699707, + "learning_rate": 1.71423963294432e-05, + "loss": 0.5537, + "step": 8542 + }, + { + "epoch": 0.2694527677022552, + "grad_norm": 2.9772987365722656, + "learning_rate": 1.7141681303526e-05, + "loss": 0.5422, + "step": 8543 + }, + { + "epoch": 0.2694843084686958, + "grad_norm": 1.6708319187164307, + "learning_rate": 1.714096620307981e-05, + "loss": 0.5018, + "step": 8544 + }, + { + "epoch": 0.26951584923513644, + "grad_norm": 1.4242569208145142, + "learning_rate": 1.714025102811209e-05, + "loss": 0.5225, + "step": 8545 + }, + { + "epoch": 0.269547390001577, + "grad_norm": 2.0135042667388916, + "learning_rate": 1.713953577863031e-05, + "loss": 0.5466, + "step": 8546 + }, + { + "epoch": 0.26957893076801764, + "grad_norm": 1.9012612104415894, + "learning_rate": 1.713882045464193e-05, + "loss": 0.5971, + "step": 8547 + }, + { + "epoch": 0.2696104715344583, + "grad_norm": 1.6549935340881348, + "learning_rate": 1.7138105056154413e-05, + "loss": 0.5101, + "step": 8548 + }, + { + "epoch": 0.2696420123008989, + "grad_norm": 2.104048013687134, + "learning_rate": 1.713738958317523e-05, + "loss": 0.5383, + "step": 8549 + }, + { + "epoch": 0.26967355306733953, + "grad_norm": 1.5729931592941284, + "learning_rate": 1.713667403571184e-05, + "loss": 0.5481, + "step": 8550 + }, + { + "epoch": 0.26970509383378016, + "grad_norm": 1.7408536672592163, + "learning_rate": 1.7135958413771717e-05, + "loss": 0.5654, + "step": 8551 + }, + { + "epoch": 0.2697366346002208, + "grad_norm": 1.8447563648223877, + "learning_rate": 1.713524271736233e-05, + "loss": 0.5239, + "step": 8552 + }, + { + "epoch": 0.2697681753666614, + "grad_norm": 2.115460157394409, + "learning_rate": 1.7134526946491147e-05, + "loss": 0.3951, + "step": 8553 + }, + { + "epoch": 0.26979971613310205, + "grad_norm": 5.093709468841553, + "learning_rate": 1.713381110116563e-05, + "loss": 0.5778, + "step": 8554 + }, + { + "epoch": 0.2698312568995427, + "grad_norm": 1.9761427640914917, + "learning_rate": 1.7133095181393254e-05, + "loss": 0.571, + "step": 8555 + }, + { + "epoch": 0.26986279766598326, + "grad_norm": 1.643876552581787, + "learning_rate": 1.7132379187181493e-05, + "loss": 0.5442, + "step": 8556 + }, + { + "epoch": 0.2698943384324239, + "grad_norm": 3.221508026123047, + "learning_rate": 1.713166311853782e-05, + "loss": 0.5337, + "step": 8557 + }, + { + "epoch": 0.2699258791988645, + "grad_norm": 1.7022844552993774, + "learning_rate": 1.71309469754697e-05, + "loss": 0.8601, + "step": 8558 + }, + { + "epoch": 0.26995741996530515, + "grad_norm": 1.4821126461029053, + "learning_rate": 1.7130230757984617e-05, + "loss": 0.5785, + "step": 8559 + }, + { + "epoch": 0.2699889607317458, + "grad_norm": 2.285259485244751, + "learning_rate": 1.7129514466090034e-05, + "loss": 0.548, + "step": 8560 + }, + { + "epoch": 0.2700205014981864, + "grad_norm": 2.294874668121338, + "learning_rate": 1.7128798099793433e-05, + "loss": 0.533, + "step": 8561 + }, + { + "epoch": 0.27005204226462703, + "grad_norm": 1.4151235818862915, + "learning_rate": 1.712808165910229e-05, + "loss": 0.5574, + "step": 8562 + }, + { + "epoch": 0.27008358303106766, + "grad_norm": 1.8551304340362549, + "learning_rate": 1.712736514402408e-05, + "loss": 0.5312, + "step": 8563 + }, + { + "epoch": 0.2701151237975083, + "grad_norm": 1.6762974262237549, + "learning_rate": 1.7126648554566275e-05, + "loss": 0.5584, + "step": 8564 + }, + { + "epoch": 0.2701466645639489, + "grad_norm": 1.8098150491714478, + "learning_rate": 1.7125931890736367e-05, + "loss": 0.5635, + "step": 8565 + }, + { + "epoch": 0.27017820533038955, + "grad_norm": 1.590552568435669, + "learning_rate": 1.7125215152541822e-05, + "loss": 0.4421, + "step": 8566 + }, + { + "epoch": 0.27020974609683013, + "grad_norm": 1.992849349975586, + "learning_rate": 1.7124498339990124e-05, + "loss": 0.5516, + "step": 8567 + }, + { + "epoch": 0.27024128686327076, + "grad_norm": 1.622179627418518, + "learning_rate": 1.7123781453088757e-05, + "loss": 0.5629, + "step": 8568 + }, + { + "epoch": 0.2702728276297114, + "grad_norm": 1.6208611726760864, + "learning_rate": 1.7123064491845195e-05, + "loss": 0.858, + "step": 8569 + }, + { + "epoch": 0.270304368396152, + "grad_norm": 1.6090482473373413, + "learning_rate": 1.7122347456266927e-05, + "loss": 0.5559, + "step": 8570 + }, + { + "epoch": 0.27033590916259265, + "grad_norm": 2.1788034439086914, + "learning_rate": 1.7121630346361433e-05, + "loss": 0.5315, + "step": 8571 + }, + { + "epoch": 0.2703674499290333, + "grad_norm": 2.394991397857666, + "learning_rate": 1.7120913162136194e-05, + "loss": 0.3913, + "step": 8572 + }, + { + "epoch": 0.2703989906954739, + "grad_norm": 1.8254896402359009, + "learning_rate": 1.71201959035987e-05, + "loss": 0.5317, + "step": 8573 + }, + { + "epoch": 0.27043053146191454, + "grad_norm": 2.041081428527832, + "learning_rate": 1.7119478570756432e-05, + "loss": 0.522, + "step": 8574 + }, + { + "epoch": 0.27046207222835517, + "grad_norm": 2.0515999794006348, + "learning_rate": 1.7118761163616878e-05, + "loss": 0.5908, + "step": 8575 + }, + { + "epoch": 0.2704936129947958, + "grad_norm": 3.1855716705322266, + "learning_rate": 1.711804368218752e-05, + "loss": 0.5517, + "step": 8576 + }, + { + "epoch": 0.2705251537612364, + "grad_norm": 2.2380595207214355, + "learning_rate": 1.7117326126475853e-05, + "loss": 0.563, + "step": 8577 + }, + { + "epoch": 0.270556694527677, + "grad_norm": 2.2879881858825684, + "learning_rate": 1.7116608496489358e-05, + "loss": 0.5172, + "step": 8578 + }, + { + "epoch": 0.27058823529411763, + "grad_norm": 2.1721677780151367, + "learning_rate": 1.711589079223553e-05, + "loss": 0.5756, + "step": 8579 + }, + { + "epoch": 0.27061977606055826, + "grad_norm": 2.502406120300293, + "learning_rate": 1.7115173013721854e-05, + "loss": 0.5577, + "step": 8580 + }, + { + "epoch": 0.2706513168269989, + "grad_norm": 2.634007453918457, + "learning_rate": 1.7114455160955824e-05, + "loss": 0.5634, + "step": 8581 + }, + { + "epoch": 0.2706828575934395, + "grad_norm": 2.293612480163574, + "learning_rate": 1.7113737233944933e-05, + "loss": 0.5239, + "step": 8582 + }, + { + "epoch": 0.27071439835988015, + "grad_norm": 1.9117108583450317, + "learning_rate": 1.7113019232696667e-05, + "loss": 0.5225, + "step": 8583 + }, + { + "epoch": 0.2707459391263208, + "grad_norm": 1.768381953239441, + "learning_rate": 1.7112301157218522e-05, + "loss": 0.3967, + "step": 8584 + }, + { + "epoch": 0.2707774798927614, + "grad_norm": 1.5507001876831055, + "learning_rate": 1.7111583007517994e-05, + "loss": 0.9182, + "step": 8585 + }, + { + "epoch": 0.27080902065920204, + "grad_norm": 2.5026497840881348, + "learning_rate": 1.711086478360257e-05, + "loss": 0.5411, + "step": 8586 + }, + { + "epoch": 0.27084056142564267, + "grad_norm": 2.7274935245513916, + "learning_rate": 1.711014648547976e-05, + "loss": 0.6041, + "step": 8587 + }, + { + "epoch": 0.27087210219208324, + "grad_norm": 3.0223352909088135, + "learning_rate": 1.710942811315704e-05, + "loss": 0.5407, + "step": 8588 + }, + { + "epoch": 0.27090364295852387, + "grad_norm": 2.886413335800171, + "learning_rate": 1.710870966664193e-05, + "loss": 0.5766, + "step": 8589 + }, + { + "epoch": 0.2709351837249645, + "grad_norm": 2.1542375087738037, + "learning_rate": 1.7107991145941905e-05, + "loss": 0.5183, + "step": 8590 + }, + { + "epoch": 0.27096672449140513, + "grad_norm": 1.398371934890747, + "learning_rate": 1.710727255106447e-05, + "loss": 0.812, + "step": 8591 + }, + { + "epoch": 0.27099826525784576, + "grad_norm": 3.8534743785858154, + "learning_rate": 1.7106553882017137e-05, + "loss": 0.5402, + "step": 8592 + }, + { + "epoch": 0.2710298060242864, + "grad_norm": 2.1689014434814453, + "learning_rate": 1.7105835138807393e-05, + "loss": 0.5623, + "step": 8593 + }, + { + "epoch": 0.271061346790727, + "grad_norm": 2.6497743129730225, + "learning_rate": 1.710511632144274e-05, + "loss": 0.5706, + "step": 8594 + }, + { + "epoch": 0.27109288755716765, + "grad_norm": 2.458695411682129, + "learning_rate": 1.710439742993068e-05, + "loss": 0.5755, + "step": 8595 + }, + { + "epoch": 0.2711244283236083, + "grad_norm": 1.7189977169036865, + "learning_rate": 1.710367846427872e-05, + "loss": 0.5283, + "step": 8596 + }, + { + "epoch": 0.2711559690900489, + "grad_norm": 1.8953273296356201, + "learning_rate": 1.710295942449436e-05, + "loss": 0.5605, + "step": 8597 + }, + { + "epoch": 0.27118750985648954, + "grad_norm": 1.9709348678588867, + "learning_rate": 1.7102240310585098e-05, + "loss": 0.5647, + "step": 8598 + }, + { + "epoch": 0.2712190506229301, + "grad_norm": 3.0504872798919678, + "learning_rate": 1.7101521122558447e-05, + "loss": 0.5146, + "step": 8599 + }, + { + "epoch": 0.27125059138937074, + "grad_norm": 2.8723866939544678, + "learning_rate": 1.7100801860421906e-05, + "loss": 0.5513, + "step": 8600 + }, + { + "epoch": 0.2712821321558114, + "grad_norm": 2.8033759593963623, + "learning_rate": 1.710008252418299e-05, + "loss": 0.5284, + "step": 8601 + }, + { + "epoch": 0.271313672922252, + "grad_norm": 2.3220531940460205, + "learning_rate": 1.7099363113849193e-05, + "loss": 0.5413, + "step": 8602 + }, + { + "epoch": 0.27134521368869263, + "grad_norm": 2.14111590385437, + "learning_rate": 1.7098643629428035e-05, + "loss": 0.5408, + "step": 8603 + }, + { + "epoch": 0.27137675445513326, + "grad_norm": 1.3203885555267334, + "learning_rate": 1.7097924070927015e-05, + "loss": 0.4209, + "step": 8604 + }, + { + "epoch": 0.2714082952215739, + "grad_norm": 1.4213783740997314, + "learning_rate": 1.709720443835365e-05, + "loss": 0.389, + "step": 8605 + }, + { + "epoch": 0.2714398359880145, + "grad_norm": 2.6303842067718506, + "learning_rate": 1.709648473171544e-05, + "loss": 0.5875, + "step": 8606 + }, + { + "epoch": 0.27147137675445515, + "grad_norm": 3.043844699859619, + "learning_rate": 1.7095764951019904e-05, + "loss": 0.5763, + "step": 8607 + }, + { + "epoch": 0.2715029175208958, + "grad_norm": 2.739858627319336, + "learning_rate": 1.7095045096274555e-05, + "loss": 0.5723, + "step": 8608 + }, + { + "epoch": 0.27153445828733636, + "grad_norm": 2.514479875564575, + "learning_rate": 1.7094325167486896e-05, + "loss": 0.5315, + "step": 8609 + }, + { + "epoch": 0.271565999053777, + "grad_norm": 2.098212957382202, + "learning_rate": 1.7093605164664447e-05, + "loss": 0.5624, + "step": 8610 + }, + { + "epoch": 0.2715975398202176, + "grad_norm": 2.0419633388519287, + "learning_rate": 1.709288508781472e-05, + "loss": 0.5676, + "step": 8611 + }, + { + "epoch": 0.27162908058665824, + "grad_norm": 2.9502084255218506, + "learning_rate": 1.709216493694523e-05, + "loss": 0.5258, + "step": 8612 + }, + { + "epoch": 0.2716606213530989, + "grad_norm": 2.386331796646118, + "learning_rate": 1.7091444712063493e-05, + "loss": 0.5454, + "step": 8613 + }, + { + "epoch": 0.2716921621195395, + "grad_norm": 1.6821658611297607, + "learning_rate": 1.709072441317702e-05, + "loss": 0.5398, + "step": 8614 + }, + { + "epoch": 0.27172370288598013, + "grad_norm": 3.3441755771636963, + "learning_rate": 1.7090004040293335e-05, + "loss": 0.5442, + "step": 8615 + }, + { + "epoch": 0.27175524365242076, + "grad_norm": 2.324758291244507, + "learning_rate": 1.7089283593419952e-05, + "loss": 0.5344, + "step": 8616 + }, + { + "epoch": 0.2717867844188614, + "grad_norm": 2.017373561859131, + "learning_rate": 1.708856307256439e-05, + "loss": 0.5452, + "step": 8617 + }, + { + "epoch": 0.271818325185302, + "grad_norm": 1.7565200328826904, + "learning_rate": 1.7087842477734173e-05, + "loss": 0.5511, + "step": 8618 + }, + { + "epoch": 0.27184986595174265, + "grad_norm": 1.888620376586914, + "learning_rate": 1.708712180893681e-05, + "loss": 0.5134, + "step": 8619 + }, + { + "epoch": 0.2718814067181832, + "grad_norm": 6.208207607269287, + "learning_rate": 1.708640106617983e-05, + "loss": 0.5332, + "step": 8620 + }, + { + "epoch": 0.27191294748462386, + "grad_norm": 1.6602693796157837, + "learning_rate": 1.708568024947075e-05, + "loss": 0.5408, + "step": 8621 + }, + { + "epoch": 0.2719444882510645, + "grad_norm": 1.6836320161819458, + "learning_rate": 1.7084959358817096e-05, + "loss": 0.5362, + "step": 8622 + }, + { + "epoch": 0.2719760290175051, + "grad_norm": 3.084947109222412, + "learning_rate": 1.7084238394226395e-05, + "loss": 0.8797, + "step": 8623 + }, + { + "epoch": 0.27200756978394575, + "grad_norm": 5.735466480255127, + "learning_rate": 1.708351735570616e-05, + "loss": 0.5351, + "step": 8624 + }, + { + "epoch": 0.2720391105503864, + "grad_norm": 1.8711051940917969, + "learning_rate": 1.7082796243263923e-05, + "loss": 0.5425, + "step": 8625 + }, + { + "epoch": 0.272070651316827, + "grad_norm": 1.584682583808899, + "learning_rate": 1.7082075056907202e-05, + "loss": 0.8671, + "step": 8626 + }, + { + "epoch": 0.27210219208326764, + "grad_norm": 1.979872703552246, + "learning_rate": 1.7081353796643537e-05, + "loss": 0.5278, + "step": 8627 + }, + { + "epoch": 0.27213373284970827, + "grad_norm": 1.8059769868850708, + "learning_rate": 1.7080632462480445e-05, + "loss": 0.5486, + "step": 8628 + }, + { + "epoch": 0.2721652736161489, + "grad_norm": 2.78820538520813, + "learning_rate": 1.707991105442545e-05, + "loss": 0.5236, + "step": 8629 + }, + { + "epoch": 0.27219681438258947, + "grad_norm": 1.609540343284607, + "learning_rate": 1.707918957248609e-05, + "loss": 0.5678, + "step": 8630 + }, + { + "epoch": 0.2722283551490301, + "grad_norm": 2.1129019260406494, + "learning_rate": 1.7078468016669888e-05, + "loss": 0.5584, + "step": 8631 + }, + { + "epoch": 0.27225989591547073, + "grad_norm": 2.040102481842041, + "learning_rate": 1.707774638698438e-05, + "loss": 0.5356, + "step": 8632 + }, + { + "epoch": 0.27229143668191136, + "grad_norm": 1.6908221244812012, + "learning_rate": 1.707702468343709e-05, + "loss": 0.5554, + "step": 8633 + }, + { + "epoch": 0.272322977448352, + "grad_norm": 2.261949300765991, + "learning_rate": 1.707630290603555e-05, + "loss": 0.5461, + "step": 8634 + }, + { + "epoch": 0.2723545182147926, + "grad_norm": 1.6739259958267212, + "learning_rate": 1.7075581054787295e-05, + "loss": 0.5299, + "step": 8635 + }, + { + "epoch": 0.27238605898123325, + "grad_norm": 2.172159194946289, + "learning_rate": 1.7074859129699863e-05, + "loss": 0.5496, + "step": 8636 + }, + { + "epoch": 0.2724175997476739, + "grad_norm": 2.3405404090881348, + "learning_rate": 1.707413713078078e-05, + "loss": 0.5575, + "step": 8637 + }, + { + "epoch": 0.2724491405141145, + "grad_norm": 1.8808331489562988, + "learning_rate": 1.7073415058037582e-05, + "loss": 0.5388, + "step": 8638 + }, + { + "epoch": 0.27248068128055514, + "grad_norm": 2.2573001384735107, + "learning_rate": 1.7072692911477804e-05, + "loss": 0.5375, + "step": 8639 + }, + { + "epoch": 0.27251222204699577, + "grad_norm": 1.5974650382995605, + "learning_rate": 1.7071970691108984e-05, + "loss": 0.4346, + "step": 8640 + }, + { + "epoch": 0.27254376281343634, + "grad_norm": 2.0631842613220215, + "learning_rate": 1.707124839693866e-05, + "loss": 0.5568, + "step": 8641 + }, + { + "epoch": 0.27257530357987697, + "grad_norm": 2.179165840148926, + "learning_rate": 1.7070526028974368e-05, + "loss": 0.869, + "step": 8642 + }, + { + "epoch": 0.2726068443463176, + "grad_norm": 2.184647560119629, + "learning_rate": 1.7069803587223645e-05, + "loss": 0.5406, + "step": 8643 + }, + { + "epoch": 0.27263838511275823, + "grad_norm": 1.7351436614990234, + "learning_rate": 1.7069081071694034e-05, + "loss": 0.8643, + "step": 8644 + }, + { + "epoch": 0.27266992587919886, + "grad_norm": 1.8527439832687378, + "learning_rate": 1.7068358482393076e-05, + "loss": 0.8712, + "step": 8645 + }, + { + "epoch": 0.2727014666456395, + "grad_norm": 2.8393096923828125, + "learning_rate": 1.7067635819328307e-05, + "loss": 0.5499, + "step": 8646 + }, + { + "epoch": 0.2727330074120801, + "grad_norm": 2.1076414585113525, + "learning_rate": 1.7066913082507267e-05, + "loss": 0.5767, + "step": 8647 + }, + { + "epoch": 0.27276454817852075, + "grad_norm": 2.214926242828369, + "learning_rate": 1.7066190271937504e-05, + "loss": 0.5419, + "step": 8648 + }, + { + "epoch": 0.2727960889449614, + "grad_norm": 1.9205716848373413, + "learning_rate": 1.7065467387626557e-05, + "loss": 0.5409, + "step": 8649 + }, + { + "epoch": 0.272827629711402, + "grad_norm": 2.7475075721740723, + "learning_rate": 1.706474442958197e-05, + "loss": 0.518, + "step": 8650 + }, + { + "epoch": 0.27285917047784264, + "grad_norm": 2.7558233737945557, + "learning_rate": 1.7064021397811296e-05, + "loss": 0.5521, + "step": 8651 + }, + { + "epoch": 0.2728907112442832, + "grad_norm": 2.4885032176971436, + "learning_rate": 1.706329829232207e-05, + "loss": 0.5292, + "step": 8652 + }, + { + "epoch": 0.27292225201072384, + "grad_norm": 2.042773962020874, + "learning_rate": 1.7062575113121843e-05, + "loss": 0.5124, + "step": 8653 + }, + { + "epoch": 0.2729537927771645, + "grad_norm": 1.919157862663269, + "learning_rate": 1.7061851860218156e-05, + "loss": 0.5292, + "step": 8654 + }, + { + "epoch": 0.2729853335436051, + "grad_norm": 2.4824697971343994, + "learning_rate": 1.7061128533618565e-05, + "loss": 0.555, + "step": 8655 + }, + { + "epoch": 0.27301687431004573, + "grad_norm": 1.9711822271347046, + "learning_rate": 1.7060405133330616e-05, + "loss": 0.5317, + "step": 8656 + }, + { + "epoch": 0.27304841507648636, + "grad_norm": 4.544112682342529, + "learning_rate": 1.7059681659361854e-05, + "loss": 0.8989, + "step": 8657 + }, + { + "epoch": 0.273079955842927, + "grad_norm": 4.091040134429932, + "learning_rate": 1.7058958111719836e-05, + "loss": 0.5776, + "step": 8658 + }, + { + "epoch": 0.2731114966093676, + "grad_norm": 2.0845279693603516, + "learning_rate": 1.7058234490412105e-05, + "loss": 0.5746, + "step": 8659 + }, + { + "epoch": 0.27314303737580825, + "grad_norm": 2.232231616973877, + "learning_rate": 1.705751079544622e-05, + "loss": 0.5428, + "step": 8660 + }, + { + "epoch": 0.2731745781422489, + "grad_norm": 2.2192702293395996, + "learning_rate": 1.705678702682973e-05, + "loss": 0.547, + "step": 8661 + }, + { + "epoch": 0.27320611890868945, + "grad_norm": 2.311744213104248, + "learning_rate": 1.7056063184570186e-05, + "loss": 0.4957, + "step": 8662 + }, + { + "epoch": 0.2732376596751301, + "grad_norm": 1.5896170139312744, + "learning_rate": 1.7055339268675143e-05, + "loss": 0.4293, + "step": 8663 + }, + { + "epoch": 0.2732692004415707, + "grad_norm": 2.1521613597869873, + "learning_rate": 1.7054615279152158e-05, + "loss": 0.888, + "step": 8664 + }, + { + "epoch": 0.27330074120801134, + "grad_norm": 1.615056037902832, + "learning_rate": 1.7053891216008785e-05, + "loss": 0.5551, + "step": 8665 + }, + { + "epoch": 0.273332281974452, + "grad_norm": 2.2498269081115723, + "learning_rate": 1.7053167079252578e-05, + "loss": 0.542, + "step": 8666 + }, + { + "epoch": 0.2733638227408926, + "grad_norm": 2.093165874481201, + "learning_rate": 1.7052442868891098e-05, + "loss": 0.567, + "step": 8667 + }, + { + "epoch": 0.27339536350733323, + "grad_norm": 1.8288853168487549, + "learning_rate": 1.70517185849319e-05, + "loss": 0.5812, + "step": 8668 + }, + { + "epoch": 0.27342690427377386, + "grad_norm": 2.321265935897827, + "learning_rate": 1.7050994227382542e-05, + "loss": 0.5192, + "step": 8669 + }, + { + "epoch": 0.2734584450402145, + "grad_norm": 2.341381072998047, + "learning_rate": 1.7050269796250588e-05, + "loss": 0.5581, + "step": 8670 + }, + { + "epoch": 0.2734899858066551, + "grad_norm": 3.2492311000823975, + "learning_rate": 1.704954529154359e-05, + "loss": 0.568, + "step": 8671 + }, + { + "epoch": 0.27352152657309575, + "grad_norm": 3.213606595993042, + "learning_rate": 1.7048820713269114e-05, + "loss": 0.5701, + "step": 8672 + }, + { + "epoch": 0.2735530673395363, + "grad_norm": 1.5465866327285767, + "learning_rate": 1.704809606143472e-05, + "loss": 0.5641, + "step": 8673 + }, + { + "epoch": 0.27358460810597696, + "grad_norm": 1.7401337623596191, + "learning_rate": 1.704737133604797e-05, + "loss": 0.5558, + "step": 8674 + }, + { + "epoch": 0.2736161488724176, + "grad_norm": 2.4853603839874268, + "learning_rate": 1.7046646537116432e-05, + "loss": 0.545, + "step": 8675 + }, + { + "epoch": 0.2736476896388582, + "grad_norm": 2.352755308151245, + "learning_rate": 1.7045921664647665e-05, + "loss": 0.9455, + "step": 8676 + }, + { + "epoch": 0.27367923040529885, + "grad_norm": 1.8150410652160645, + "learning_rate": 1.7045196718649234e-05, + "loss": 0.573, + "step": 8677 + }, + { + "epoch": 0.2737107711717395, + "grad_norm": 2.3746891021728516, + "learning_rate": 1.7044471699128702e-05, + "loss": 0.5402, + "step": 8678 + }, + { + "epoch": 0.2737423119381801, + "grad_norm": 2.0243453979492188, + "learning_rate": 1.704374660609364e-05, + "loss": 0.55, + "step": 8679 + }, + { + "epoch": 0.27377385270462073, + "grad_norm": 1.662299633026123, + "learning_rate": 1.7043021439551614e-05, + "loss": 0.5826, + "step": 8680 + }, + { + "epoch": 0.27380539347106136, + "grad_norm": 2.3690600395202637, + "learning_rate": 1.7042296199510188e-05, + "loss": 0.5661, + "step": 8681 + }, + { + "epoch": 0.273836934237502, + "grad_norm": 2.451799154281616, + "learning_rate": 1.7041570885976935e-05, + "loss": 0.545, + "step": 8682 + }, + { + "epoch": 0.27386847500394257, + "grad_norm": 1.5463563203811646, + "learning_rate": 1.7040845498959422e-05, + "loss": 0.54, + "step": 8683 + }, + { + "epoch": 0.2739000157703832, + "grad_norm": 2.5237483978271484, + "learning_rate": 1.7040120038465214e-05, + "loss": 0.5241, + "step": 8684 + }, + { + "epoch": 0.27393155653682383, + "grad_norm": 1.6550053358078003, + "learning_rate": 1.7039394504501895e-05, + "loss": 0.5461, + "step": 8685 + }, + { + "epoch": 0.27396309730326446, + "grad_norm": 2.467958450317383, + "learning_rate": 1.7038668897077025e-05, + "loss": 0.5661, + "step": 8686 + }, + { + "epoch": 0.2739946380697051, + "grad_norm": 1.7280912399291992, + "learning_rate": 1.703794321619818e-05, + "loss": 0.545, + "step": 8687 + }, + { + "epoch": 0.2740261788361457, + "grad_norm": 1.9074695110321045, + "learning_rate": 1.703721746187293e-05, + "loss": 0.4244, + "step": 8688 + }, + { + "epoch": 0.27405771960258635, + "grad_norm": 2.0158417224884033, + "learning_rate": 1.7036491634108855e-05, + "loss": 0.5429, + "step": 8689 + }, + { + "epoch": 0.274089260369027, + "grad_norm": 2.024786949157715, + "learning_rate": 1.7035765732913524e-05, + "loss": 0.8568, + "step": 8690 + }, + { + "epoch": 0.2741208011354676, + "grad_norm": 2.331233501434326, + "learning_rate": 1.7035039758294515e-05, + "loss": 0.5725, + "step": 8691 + }, + { + "epoch": 0.27415234190190824, + "grad_norm": 2.4006142616271973, + "learning_rate": 1.7034313710259406e-05, + "loss": 0.5724, + "step": 8692 + }, + { + "epoch": 0.27418388266834887, + "grad_norm": 1.516426682472229, + "learning_rate": 1.7033587588815767e-05, + "loss": 0.5057, + "step": 8693 + }, + { + "epoch": 0.27421542343478944, + "grad_norm": 1.6902843713760376, + "learning_rate": 1.7032861393971183e-05, + "loss": 0.5844, + "step": 8694 + }, + { + "epoch": 0.27424696420123007, + "grad_norm": 2.5122976303100586, + "learning_rate": 1.703213512573323e-05, + "loss": 0.5408, + "step": 8695 + }, + { + "epoch": 0.2742785049676707, + "grad_norm": 3.661827802658081, + "learning_rate": 1.7031408784109485e-05, + "loss": 0.5371, + "step": 8696 + }, + { + "epoch": 0.27431004573411133, + "grad_norm": 2.99776029586792, + "learning_rate": 1.703068236910753e-05, + "loss": 0.4512, + "step": 8697 + }, + { + "epoch": 0.27434158650055196, + "grad_norm": 2.8224074840545654, + "learning_rate": 1.7029955880734946e-05, + "loss": 0.5308, + "step": 8698 + }, + { + "epoch": 0.2743731272669926, + "grad_norm": 1.5979341268539429, + "learning_rate": 1.7029229318999312e-05, + "loss": 0.9233, + "step": 8699 + }, + { + "epoch": 0.2744046680334332, + "grad_norm": 2.0652718544006348, + "learning_rate": 1.7028502683908212e-05, + "loss": 0.5506, + "step": 8700 + }, + { + "epoch": 0.27443620879987385, + "grad_norm": 2.0460970401763916, + "learning_rate": 1.702777597546923e-05, + "loss": 0.5038, + "step": 8701 + }, + { + "epoch": 0.2744677495663145, + "grad_norm": 1.7612366676330566, + "learning_rate": 1.702704919368995e-05, + "loss": 0.5513, + "step": 8702 + }, + { + "epoch": 0.2744992903327551, + "grad_norm": 2.00567889213562, + "learning_rate": 1.7026322338577955e-05, + "loss": 0.5416, + "step": 8703 + }, + { + "epoch": 0.27453083109919574, + "grad_norm": 2.407680034637451, + "learning_rate": 1.702559541014083e-05, + "loss": 0.5643, + "step": 8704 + }, + { + "epoch": 0.2745623718656363, + "grad_norm": 2.0442962646484375, + "learning_rate": 1.702486840838616e-05, + "loss": 0.5636, + "step": 8705 + }, + { + "epoch": 0.27459391263207694, + "grad_norm": 2.2619752883911133, + "learning_rate": 1.7024141333321536e-05, + "loss": 0.5427, + "step": 8706 + }, + { + "epoch": 0.27462545339851757, + "grad_norm": 1.5747156143188477, + "learning_rate": 1.7023414184954542e-05, + "loss": 0.8734, + "step": 8707 + }, + { + "epoch": 0.2746569941649582, + "grad_norm": 2.253070116043091, + "learning_rate": 1.7022686963292763e-05, + "loss": 0.5536, + "step": 8708 + }, + { + "epoch": 0.27468853493139883, + "grad_norm": 2.2152371406555176, + "learning_rate": 1.70219596683438e-05, + "loss": 0.5415, + "step": 8709 + }, + { + "epoch": 0.27472007569783946, + "grad_norm": 2.8567440509796143, + "learning_rate": 1.702123230011523e-05, + "loss": 0.5309, + "step": 8710 + }, + { + "epoch": 0.2747516164642801, + "grad_norm": 2.164777994155884, + "learning_rate": 1.7020504858614654e-05, + "loss": 0.5384, + "step": 8711 + }, + { + "epoch": 0.2747831572307207, + "grad_norm": 2.2428336143493652, + "learning_rate": 1.7019777343849653e-05, + "loss": 0.5245, + "step": 8712 + }, + { + "epoch": 0.27481469799716135, + "grad_norm": 2.7576863765716553, + "learning_rate": 1.701904975582783e-05, + "loss": 0.555, + "step": 8713 + }, + { + "epoch": 0.274846238763602, + "grad_norm": 3.104313611984253, + "learning_rate": 1.701832209455677e-05, + "loss": 0.5503, + "step": 8714 + }, + { + "epoch": 0.27487777953004255, + "grad_norm": 1.9820276498794556, + "learning_rate": 1.7017594360044068e-05, + "loss": 0.5507, + "step": 8715 + }, + { + "epoch": 0.2749093202964832, + "grad_norm": 2.460216522216797, + "learning_rate": 1.7016866552297323e-05, + "loss": 0.5384, + "step": 8716 + }, + { + "epoch": 0.2749408610629238, + "grad_norm": 2.0752341747283936, + "learning_rate": 1.7016138671324126e-05, + "loss": 0.8251, + "step": 8717 + }, + { + "epoch": 0.27497240182936444, + "grad_norm": 2.0718448162078857, + "learning_rate": 1.7015410717132073e-05, + "loss": 0.5418, + "step": 8718 + }, + { + "epoch": 0.2750039425958051, + "grad_norm": 2.808933734893799, + "learning_rate": 1.7014682689728764e-05, + "loss": 0.5841, + "step": 8719 + }, + { + "epoch": 0.2750354833622457, + "grad_norm": 4.559696197509766, + "learning_rate": 1.7013954589121794e-05, + "loss": 0.4402, + "step": 8720 + }, + { + "epoch": 0.27506702412868633, + "grad_norm": 2.5628440380096436, + "learning_rate": 1.701322641531876e-05, + "loss": 0.5154, + "step": 8721 + }, + { + "epoch": 0.27509856489512696, + "grad_norm": 3.5830187797546387, + "learning_rate": 1.7012498168327264e-05, + "loss": 0.5617, + "step": 8722 + }, + { + "epoch": 0.2751301056615676, + "grad_norm": 2.1213419437408447, + "learning_rate": 1.7011769848154904e-05, + "loss": 0.8673, + "step": 8723 + }, + { + "epoch": 0.2751616464280082, + "grad_norm": 2.356118679046631, + "learning_rate": 1.7011041454809278e-05, + "loss": 0.561, + "step": 8724 + }, + { + "epoch": 0.27519318719444885, + "grad_norm": 2.7644028663635254, + "learning_rate": 1.7010312988297993e-05, + "loss": 0.5465, + "step": 8725 + }, + { + "epoch": 0.2752247279608894, + "grad_norm": 1.7563978433609009, + "learning_rate": 1.700958444862865e-05, + "loss": 0.5469, + "step": 8726 + }, + { + "epoch": 0.27525626872733006, + "grad_norm": 1.7074248790740967, + "learning_rate": 1.7008855835808852e-05, + "loss": 0.5642, + "step": 8727 + }, + { + "epoch": 0.2752878094937707, + "grad_norm": 3.241865873336792, + "learning_rate": 1.7008127149846198e-05, + "loss": 0.5565, + "step": 8728 + }, + { + "epoch": 0.2753193502602113, + "grad_norm": 2.887376070022583, + "learning_rate": 1.7007398390748296e-05, + "loss": 0.5715, + "step": 8729 + }, + { + "epoch": 0.27535089102665194, + "grad_norm": 1.9224612712860107, + "learning_rate": 1.700666955852275e-05, + "loss": 0.5358, + "step": 8730 + }, + { + "epoch": 0.2753824317930926, + "grad_norm": 2.8741044998168945, + "learning_rate": 1.7005940653177167e-05, + "loss": 0.5356, + "step": 8731 + }, + { + "epoch": 0.2754139725595332, + "grad_norm": 1.6720556020736694, + "learning_rate": 1.7005211674719152e-05, + "loss": 0.4352, + "step": 8732 + }, + { + "epoch": 0.27544551332597383, + "grad_norm": 2.0927445888519287, + "learning_rate": 1.700448262315632e-05, + "loss": 0.5442, + "step": 8733 + }, + { + "epoch": 0.27547705409241446, + "grad_norm": 1.7901802062988281, + "learning_rate": 1.7003753498496267e-05, + "loss": 0.5377, + "step": 8734 + }, + { + "epoch": 0.2755085948588551, + "grad_norm": 2.255826711654663, + "learning_rate": 1.700302430074661e-05, + "loss": 0.5598, + "step": 8735 + }, + { + "epoch": 0.27554013562529567, + "grad_norm": 1.965234637260437, + "learning_rate": 1.7002295029914955e-05, + "loss": 0.5723, + "step": 8736 + }, + { + "epoch": 0.2755716763917363, + "grad_norm": 3.6632843017578125, + "learning_rate": 1.7001565686008915e-05, + "loss": 0.4947, + "step": 8737 + }, + { + "epoch": 0.2756032171581769, + "grad_norm": 1.917040228843689, + "learning_rate": 1.7000836269036102e-05, + "loss": 0.5516, + "step": 8738 + }, + { + "epoch": 0.27563475792461756, + "grad_norm": 2.0218822956085205, + "learning_rate": 1.7000106779004126e-05, + "loss": 0.5591, + "step": 8739 + }, + { + "epoch": 0.2756662986910582, + "grad_norm": 1.7379114627838135, + "learning_rate": 1.69993772159206e-05, + "loss": 0.5341, + "step": 8740 + }, + { + "epoch": 0.2756978394574988, + "grad_norm": 2.8848023414611816, + "learning_rate": 1.6998647579793136e-05, + "loss": 0.5201, + "step": 8741 + }, + { + "epoch": 0.27572938022393945, + "grad_norm": 1.5685316324234009, + "learning_rate": 1.6997917870629352e-05, + "loss": 0.8585, + "step": 8742 + }, + { + "epoch": 0.2757609209903801, + "grad_norm": 2.2342276573181152, + "learning_rate": 1.699718808843686e-05, + "loss": 0.546, + "step": 8743 + }, + { + "epoch": 0.2757924617568207, + "grad_norm": 1.9961435794830322, + "learning_rate": 1.6996458233223282e-05, + "loss": 0.5273, + "step": 8744 + }, + { + "epoch": 0.27582400252326134, + "grad_norm": 2.0613441467285156, + "learning_rate": 1.6995728304996222e-05, + "loss": 0.5534, + "step": 8745 + }, + { + "epoch": 0.27585554328970197, + "grad_norm": 2.16137433052063, + "learning_rate": 1.699499830376331e-05, + "loss": 0.5159, + "step": 8746 + }, + { + "epoch": 0.27588708405614254, + "grad_norm": 1.5044625997543335, + "learning_rate": 1.699426822953216e-05, + "loss": 0.8724, + "step": 8747 + }, + { + "epoch": 0.27591862482258317, + "grad_norm": 1.933415412902832, + "learning_rate": 1.699353808231039e-05, + "loss": 0.5642, + "step": 8748 + }, + { + "epoch": 0.2759501655890238, + "grad_norm": 2.5842480659484863, + "learning_rate": 1.6992807862105617e-05, + "loss": 0.5167, + "step": 8749 + }, + { + "epoch": 0.27598170635546443, + "grad_norm": 2.1300559043884277, + "learning_rate": 1.6992077568925464e-05, + "loss": 0.5437, + "step": 8750 + }, + { + "epoch": 0.27601324712190506, + "grad_norm": 2.0519912242889404, + "learning_rate": 1.6991347202777556e-05, + "loss": 0.5574, + "step": 8751 + }, + { + "epoch": 0.2760447878883457, + "grad_norm": 2.1243395805358887, + "learning_rate": 1.6990616763669506e-05, + "loss": 0.5625, + "step": 8752 + }, + { + "epoch": 0.2760763286547863, + "grad_norm": 1.6102039813995361, + "learning_rate": 1.6989886251608948e-05, + "loss": 0.5695, + "step": 8753 + }, + { + "epoch": 0.27610786942122695, + "grad_norm": 2.0031678676605225, + "learning_rate": 1.6989155666603495e-05, + "loss": 0.5091, + "step": 8754 + }, + { + "epoch": 0.2761394101876676, + "grad_norm": 1.7238287925720215, + "learning_rate": 1.6988425008660776e-05, + "loss": 0.5652, + "step": 8755 + }, + { + "epoch": 0.2761709509541082, + "grad_norm": 2.438053846359253, + "learning_rate": 1.698769427778842e-05, + "loss": 0.5579, + "step": 8756 + }, + { + "epoch": 0.2762024917205488, + "grad_norm": 1.8941128253936768, + "learning_rate": 1.6986963473994044e-05, + "loss": 0.5595, + "step": 8757 + }, + { + "epoch": 0.2762340324869894, + "grad_norm": 1.5042126178741455, + "learning_rate": 1.698623259728528e-05, + "loss": 0.5693, + "step": 8758 + }, + { + "epoch": 0.27626557325343004, + "grad_norm": 1.8976881504058838, + "learning_rate": 1.6985501647669755e-05, + "loss": 0.5343, + "step": 8759 + }, + { + "epoch": 0.27629711401987067, + "grad_norm": 5.150027275085449, + "learning_rate": 1.6984770625155094e-05, + "loss": 0.4403, + "step": 8760 + }, + { + "epoch": 0.2763286547863113, + "grad_norm": 2.448930025100708, + "learning_rate": 1.698403952974893e-05, + "loss": 0.5568, + "step": 8761 + }, + { + "epoch": 0.27636019555275193, + "grad_norm": 2.1852121353149414, + "learning_rate": 1.698330836145889e-05, + "loss": 0.5466, + "step": 8762 + }, + { + "epoch": 0.27639173631919256, + "grad_norm": 2.0360124111175537, + "learning_rate": 1.6982577120292607e-05, + "loss": 0.5722, + "step": 8763 + }, + { + "epoch": 0.2764232770856332, + "grad_norm": 1.8473334312438965, + "learning_rate": 1.6981845806257707e-05, + "loss": 0.5274, + "step": 8764 + }, + { + "epoch": 0.2764548178520738, + "grad_norm": 2.683528423309326, + "learning_rate": 1.6981114419361825e-05, + "loss": 0.5436, + "step": 8765 + }, + { + "epoch": 0.27648635861851445, + "grad_norm": 5.140604496002197, + "learning_rate": 1.6980382959612593e-05, + "loss": 0.5452, + "step": 8766 + }, + { + "epoch": 0.2765178993849551, + "grad_norm": 3.7666449546813965, + "learning_rate": 1.6979651427017646e-05, + "loss": 0.5796, + "step": 8767 + }, + { + "epoch": 0.27654944015139565, + "grad_norm": 2.2275757789611816, + "learning_rate": 1.6978919821584618e-05, + "loss": 0.8994, + "step": 8768 + }, + { + "epoch": 0.2765809809178363, + "grad_norm": 2.4942636489868164, + "learning_rate": 1.697818814332114e-05, + "loss": 0.5486, + "step": 8769 + }, + { + "epoch": 0.2766125216842769, + "grad_norm": 1.7689130306243896, + "learning_rate": 1.697745639223485e-05, + "loss": 0.5408, + "step": 8770 + }, + { + "epoch": 0.27664406245071754, + "grad_norm": 2.284383773803711, + "learning_rate": 1.6976724568333387e-05, + "loss": 0.5466, + "step": 8771 + }, + { + "epoch": 0.2766756032171582, + "grad_norm": 2.877485990524292, + "learning_rate": 1.6975992671624384e-05, + "loss": 0.5544, + "step": 8772 + }, + { + "epoch": 0.2767071439835988, + "grad_norm": 1.6227407455444336, + "learning_rate": 1.697526070211548e-05, + "loss": 0.5051, + "step": 8773 + }, + { + "epoch": 0.27673868475003943, + "grad_norm": 1.7755388021469116, + "learning_rate": 1.6974528659814317e-05, + "loss": 0.4298, + "step": 8774 + }, + { + "epoch": 0.27677022551648006, + "grad_norm": 1.9659463167190552, + "learning_rate": 1.6973796544728534e-05, + "loss": 0.5491, + "step": 8775 + }, + { + "epoch": 0.2768017662829207, + "grad_norm": 2.171741008758545, + "learning_rate": 1.6973064356865765e-05, + "loss": 0.5365, + "step": 8776 + }, + { + "epoch": 0.2768333070493613, + "grad_norm": 2.387242555618286, + "learning_rate": 1.6972332096233654e-05, + "loss": 0.538, + "step": 8777 + }, + { + "epoch": 0.27686484781580195, + "grad_norm": 3.1182260513305664, + "learning_rate": 1.6971599762839847e-05, + "loss": 0.5423, + "step": 8778 + }, + { + "epoch": 0.2768963885822425, + "grad_norm": 2.0985159873962402, + "learning_rate": 1.697086735669198e-05, + "loss": 0.5299, + "step": 8779 + }, + { + "epoch": 0.27692792934868316, + "grad_norm": 2.5069968700408936, + "learning_rate": 1.69701348777977e-05, + "loss": 0.5392, + "step": 8780 + }, + { + "epoch": 0.2769594701151238, + "grad_norm": 1.9394901990890503, + "learning_rate": 1.6969402326164655e-05, + "loss": 0.5649, + "step": 8781 + }, + { + "epoch": 0.2769910108815644, + "grad_norm": 2.1867787837982178, + "learning_rate": 1.6968669701800483e-05, + "loss": 0.532, + "step": 8782 + }, + { + "epoch": 0.27702255164800504, + "grad_norm": 2.1577329635620117, + "learning_rate": 1.696793700471283e-05, + "loss": 0.5632, + "step": 8783 + }, + { + "epoch": 0.2770540924144457, + "grad_norm": 1.9328359365463257, + "learning_rate": 1.6967204234909345e-05, + "loss": 0.5525, + "step": 8784 + }, + { + "epoch": 0.2770856331808863, + "grad_norm": 2.6589460372924805, + "learning_rate": 1.696647139239767e-05, + "loss": 0.5582, + "step": 8785 + }, + { + "epoch": 0.27711717394732693, + "grad_norm": 2.1949570178985596, + "learning_rate": 1.696573847718546e-05, + "loss": 0.8664, + "step": 8786 + }, + { + "epoch": 0.27714871471376756, + "grad_norm": 2.0954179763793945, + "learning_rate": 1.696500548928036e-05, + "loss": 0.5645, + "step": 8787 + }, + { + "epoch": 0.2771802554802082, + "grad_norm": 3.0279760360717773, + "learning_rate": 1.6964272428690022e-05, + "loss": 0.5169, + "step": 8788 + }, + { + "epoch": 0.27721179624664877, + "grad_norm": 5.1482367515563965, + "learning_rate": 1.696353929542209e-05, + "loss": 0.5022, + "step": 8789 + }, + { + "epoch": 0.2772433370130894, + "grad_norm": 1.6129511594772339, + "learning_rate": 1.696280608948422e-05, + "loss": 0.8498, + "step": 8790 + }, + { + "epoch": 0.27727487777953, + "grad_norm": 2.364786386489868, + "learning_rate": 1.6962072810884063e-05, + "loss": 0.5264, + "step": 8791 + }, + { + "epoch": 0.27730641854597066, + "grad_norm": 2.173137903213501, + "learning_rate": 1.696133945962927e-05, + "loss": 0.5377, + "step": 8792 + }, + { + "epoch": 0.2773379593124113, + "grad_norm": 2.2623112201690674, + "learning_rate": 1.696060603572749e-05, + "loss": 0.5603, + "step": 8793 + }, + { + "epoch": 0.2773695000788519, + "grad_norm": 1.763086199760437, + "learning_rate": 1.6959872539186385e-05, + "loss": 0.5475, + "step": 8794 + }, + { + "epoch": 0.27740104084529255, + "grad_norm": 1.9462939500808716, + "learning_rate": 1.6959138970013606e-05, + "loss": 0.5698, + "step": 8795 + }, + { + "epoch": 0.2774325816117332, + "grad_norm": 1.7937911748886108, + "learning_rate": 1.6958405328216807e-05, + "loss": 0.5209, + "step": 8796 + }, + { + "epoch": 0.2774641223781738, + "grad_norm": 1.6739089488983154, + "learning_rate": 1.6957671613803648e-05, + "loss": 0.5514, + "step": 8797 + }, + { + "epoch": 0.27749566314461444, + "grad_norm": 1.7230968475341797, + "learning_rate": 1.695693782678178e-05, + "loss": 0.5327, + "step": 8798 + }, + { + "epoch": 0.27752720391105506, + "grad_norm": 1.9570016860961914, + "learning_rate": 1.6956203967158865e-05, + "loss": 0.5796, + "step": 8799 + }, + { + "epoch": 0.27755874467749564, + "grad_norm": 1.7864623069763184, + "learning_rate": 1.695547003494256e-05, + "loss": 0.5648, + "step": 8800 + }, + { + "epoch": 0.27759028544393627, + "grad_norm": 2.334247589111328, + "learning_rate": 1.6954736030140523e-05, + "loss": 0.3712, + "step": 8801 + }, + { + "epoch": 0.2776218262103769, + "grad_norm": 2.5795350074768066, + "learning_rate": 1.695400195276042e-05, + "loss": 0.5094, + "step": 8802 + }, + { + "epoch": 0.27765336697681753, + "grad_norm": 3.158586263656616, + "learning_rate": 1.69532678028099e-05, + "loss": 0.5456, + "step": 8803 + }, + { + "epoch": 0.27768490774325816, + "grad_norm": 1.8165616989135742, + "learning_rate": 1.695253358029664e-05, + "loss": 0.5519, + "step": 8804 + }, + { + "epoch": 0.2777164485096988, + "grad_norm": 3.5501549243927, + "learning_rate": 1.6951799285228285e-05, + "loss": 0.5192, + "step": 8805 + }, + { + "epoch": 0.2777479892761394, + "grad_norm": 2.063694715499878, + "learning_rate": 1.6951064917612516e-05, + "loss": 0.5724, + "step": 8806 + }, + { + "epoch": 0.27777953004258005, + "grad_norm": 2.030113935470581, + "learning_rate": 1.6950330477456976e-05, + "loss": 0.5472, + "step": 8807 + }, + { + "epoch": 0.2778110708090207, + "grad_norm": 1.1185134649276733, + "learning_rate": 1.694959596476935e-05, + "loss": 0.4089, + "step": 8808 + }, + { + "epoch": 0.2778426115754613, + "grad_norm": 1.9837830066680908, + "learning_rate": 1.694886137955729e-05, + "loss": 0.5235, + "step": 8809 + }, + { + "epoch": 0.2778741523419019, + "grad_norm": 1.8852522373199463, + "learning_rate": 1.6948126721828466e-05, + "loss": 0.5647, + "step": 8810 + }, + { + "epoch": 0.2779056931083425, + "grad_norm": 2.0018534660339355, + "learning_rate": 1.6947391991590546e-05, + "loss": 0.5452, + "step": 8811 + }, + { + "epoch": 0.27793723387478314, + "grad_norm": 1.7781481742858887, + "learning_rate": 1.6946657188851194e-05, + "loss": 0.5167, + "step": 8812 + }, + { + "epoch": 0.27796877464122377, + "grad_norm": 2.128249168395996, + "learning_rate": 1.694592231361808e-05, + "loss": 0.5345, + "step": 8813 + }, + { + "epoch": 0.2780003154076644, + "grad_norm": 1.1386678218841553, + "learning_rate": 1.6945187365898873e-05, + "loss": 0.4265, + "step": 8814 + }, + { + "epoch": 0.27803185617410503, + "grad_norm": 1.6926600933074951, + "learning_rate": 1.6944452345701246e-05, + "loss": 0.5776, + "step": 8815 + }, + { + "epoch": 0.27806339694054566, + "grad_norm": 1.3657747507095337, + "learning_rate": 1.6943717253032866e-05, + "loss": 0.579, + "step": 8816 + }, + { + "epoch": 0.2780949377069863, + "grad_norm": 1.6074857711791992, + "learning_rate": 1.6942982087901404e-05, + "loss": 0.5549, + "step": 8817 + }, + { + "epoch": 0.2781264784734269, + "grad_norm": 2.1610257625579834, + "learning_rate": 1.6942246850314533e-05, + "loss": 0.5478, + "step": 8818 + }, + { + "epoch": 0.27815801923986755, + "grad_norm": 3.477694034576416, + "learning_rate": 1.6941511540279925e-05, + "loss": 0.5038, + "step": 8819 + }, + { + "epoch": 0.2781895600063082, + "grad_norm": 1.7012379169464111, + "learning_rate": 1.6940776157805253e-05, + "loss": 0.8613, + "step": 8820 + }, + { + "epoch": 0.27822110077274875, + "grad_norm": 1.4970641136169434, + "learning_rate": 1.6940040702898192e-05, + "loss": 0.5492, + "step": 8821 + }, + { + "epoch": 0.2782526415391894, + "grad_norm": 2.3599531650543213, + "learning_rate": 1.693930517556642e-05, + "loss": 0.5199, + "step": 8822 + }, + { + "epoch": 0.27828418230563, + "grad_norm": 2.0272159576416016, + "learning_rate": 1.6938569575817608e-05, + "loss": 0.5239, + "step": 8823 + }, + { + "epoch": 0.27831572307207064, + "grad_norm": 1.7502130270004272, + "learning_rate": 1.693783390365944e-05, + "loss": 0.5439, + "step": 8824 + }, + { + "epoch": 0.27834726383851127, + "grad_norm": 1.8251121044158936, + "learning_rate": 1.693709815909958e-05, + "loss": 0.5623, + "step": 8825 + }, + { + "epoch": 0.2783788046049519, + "grad_norm": 1.8761669397354126, + "learning_rate": 1.6936362342145724e-05, + "loss": 0.5313, + "step": 8826 + }, + { + "epoch": 0.27841034537139253, + "grad_norm": 1.689082384109497, + "learning_rate": 1.693562645280553e-05, + "loss": 0.5735, + "step": 8827 + }, + { + "epoch": 0.27844188613783316, + "grad_norm": 1.5603758096694946, + "learning_rate": 1.6934890491086696e-05, + "loss": 0.5638, + "step": 8828 + }, + { + "epoch": 0.2784734269042738, + "grad_norm": 1.6189345121383667, + "learning_rate": 1.6934154456996893e-05, + "loss": 0.541, + "step": 8829 + }, + { + "epoch": 0.2785049676707144, + "grad_norm": 2.9997541904449463, + "learning_rate": 1.6933418350543808e-05, + "loss": 0.4405, + "step": 8830 + }, + { + "epoch": 0.27853650843715505, + "grad_norm": 1.880354642868042, + "learning_rate": 1.6932682171735112e-05, + "loss": 0.5601, + "step": 8831 + }, + { + "epoch": 0.2785680492035956, + "grad_norm": 1.6001577377319336, + "learning_rate": 1.6931945920578497e-05, + "loss": 0.8271, + "step": 8832 + }, + { + "epoch": 0.27859958997003625, + "grad_norm": 1.5315853357315063, + "learning_rate": 1.6931209597081645e-05, + "loss": 0.5214, + "step": 8833 + }, + { + "epoch": 0.2786311307364769, + "grad_norm": 2.085747718811035, + "learning_rate": 1.693047320125224e-05, + "loss": 0.5272, + "step": 8834 + }, + { + "epoch": 0.2786626715029175, + "grad_norm": 1.5175681114196777, + "learning_rate": 1.6929736733097963e-05, + "loss": 0.5548, + "step": 8835 + }, + { + "epoch": 0.27869421226935814, + "grad_norm": 2.993905544281006, + "learning_rate": 1.6929000192626504e-05, + "loss": 0.5667, + "step": 8836 + }, + { + "epoch": 0.2787257530357988, + "grad_norm": 1.8784602880477905, + "learning_rate": 1.6928263579845546e-05, + "loss": 0.4283, + "step": 8837 + }, + { + "epoch": 0.2787572938022394, + "grad_norm": 2.1526408195495605, + "learning_rate": 1.692752689476278e-05, + "loss": 0.5274, + "step": 8838 + }, + { + "epoch": 0.27878883456868003, + "grad_norm": 1.8961046934127808, + "learning_rate": 1.692679013738589e-05, + "loss": 0.5546, + "step": 8839 + }, + { + "epoch": 0.27882037533512066, + "grad_norm": 2.567683458328247, + "learning_rate": 1.692605330772257e-05, + "loss": 0.5084, + "step": 8840 + }, + { + "epoch": 0.2788519161015613, + "grad_norm": 1.8065208196640015, + "learning_rate": 1.69253164057805e-05, + "loss": 0.5044, + "step": 8841 + }, + { + "epoch": 0.27888345686800187, + "grad_norm": 1.9839600324630737, + "learning_rate": 1.692457943156738e-05, + "loss": 0.5403, + "step": 8842 + }, + { + "epoch": 0.2789149976344425, + "grad_norm": 1.4321388006210327, + "learning_rate": 1.6923842385090896e-05, + "loss": 0.5292, + "step": 8843 + }, + { + "epoch": 0.2789465384008831, + "grad_norm": 2.1481573581695557, + "learning_rate": 1.6923105266358737e-05, + "loss": 0.5656, + "step": 8844 + }, + { + "epoch": 0.27897807916732376, + "grad_norm": 2.203932285308838, + "learning_rate": 1.6922368075378602e-05, + "loss": 0.5287, + "step": 8845 + }, + { + "epoch": 0.2790096199337644, + "grad_norm": 2.7004637718200684, + "learning_rate": 1.692163081215818e-05, + "loss": 0.5525, + "step": 8846 + }, + { + "epoch": 0.279041160700205, + "grad_norm": 3.4273746013641357, + "learning_rate": 1.692089347670517e-05, + "loss": 0.4023, + "step": 8847 + }, + { + "epoch": 0.27907270146664565, + "grad_norm": 1.8615822792053223, + "learning_rate": 1.692015606902726e-05, + "loss": 0.5492, + "step": 8848 + }, + { + "epoch": 0.2791042422330863, + "grad_norm": 1.9737274646759033, + "learning_rate": 1.6919418589132145e-05, + "loss": 0.5579, + "step": 8849 + }, + { + "epoch": 0.2791357829995269, + "grad_norm": 1.858440637588501, + "learning_rate": 1.691868103702753e-05, + "loss": 0.55, + "step": 8850 + }, + { + "epoch": 0.27916732376596753, + "grad_norm": 1.9892019033432007, + "learning_rate": 1.69179434127211e-05, + "loss": 0.5346, + "step": 8851 + }, + { + "epoch": 0.27919886453240816, + "grad_norm": 1.8199435472488403, + "learning_rate": 1.6917205716220565e-05, + "loss": 0.5221, + "step": 8852 + }, + { + "epoch": 0.27923040529884874, + "grad_norm": 1.6714849472045898, + "learning_rate": 1.6916467947533613e-05, + "loss": 0.5353, + "step": 8853 + }, + { + "epoch": 0.27926194606528937, + "grad_norm": 2.8080055713653564, + "learning_rate": 1.691573010666795e-05, + "loss": 0.5499, + "step": 8854 + }, + { + "epoch": 0.27929348683173, + "grad_norm": 2.335024118423462, + "learning_rate": 1.6914992193631272e-05, + "loss": 0.5363, + "step": 8855 + }, + { + "epoch": 0.2793250275981706, + "grad_norm": 2.052950143814087, + "learning_rate": 1.691425420843128e-05, + "loss": 0.5246, + "step": 8856 + }, + { + "epoch": 0.27935656836461126, + "grad_norm": 1.4415864944458008, + "learning_rate": 1.691351615107568e-05, + "loss": 0.5337, + "step": 8857 + }, + { + "epoch": 0.2793881091310519, + "grad_norm": 1.712086796760559, + "learning_rate": 1.6912778021572166e-05, + "loss": 0.8852, + "step": 8858 + }, + { + "epoch": 0.2794196498974925, + "grad_norm": 1.7999048233032227, + "learning_rate": 1.691203981992845e-05, + "loss": 0.5598, + "step": 8859 + }, + { + "epoch": 0.27945119066393315, + "grad_norm": 2.054523468017578, + "learning_rate": 1.691130154615223e-05, + "loss": 0.5474, + "step": 8860 + }, + { + "epoch": 0.2794827314303738, + "grad_norm": 1.792357325553894, + "learning_rate": 1.6910563200251212e-05, + "loss": 0.5507, + "step": 8861 + }, + { + "epoch": 0.2795142721968144, + "grad_norm": 2.009383201599121, + "learning_rate": 1.6909824782233098e-05, + "loss": 0.5382, + "step": 8862 + }, + { + "epoch": 0.279545812963255, + "grad_norm": 1.543275237083435, + "learning_rate": 1.6909086292105602e-05, + "loss": 0.5614, + "step": 8863 + }, + { + "epoch": 0.2795773537296956, + "grad_norm": 1.6290552616119385, + "learning_rate": 1.6908347729876423e-05, + "loss": 0.8515, + "step": 8864 + }, + { + "epoch": 0.27960889449613624, + "grad_norm": 1.647796869277954, + "learning_rate": 1.6907609095553274e-05, + "loss": 0.518, + "step": 8865 + }, + { + "epoch": 0.27964043526257687, + "grad_norm": 1.6166229248046875, + "learning_rate": 1.6906870389143853e-05, + "loss": 0.5499, + "step": 8866 + }, + { + "epoch": 0.2796719760290175, + "grad_norm": 1.7470592260360718, + "learning_rate": 1.6906131610655886e-05, + "loss": 0.8603, + "step": 8867 + }, + { + "epoch": 0.27970351679545813, + "grad_norm": 2.068655014038086, + "learning_rate": 1.6905392760097067e-05, + "loss": 0.5616, + "step": 8868 + }, + { + "epoch": 0.27973505756189876, + "grad_norm": 1.5746164321899414, + "learning_rate": 1.6904653837475115e-05, + "loss": 0.5558, + "step": 8869 + }, + { + "epoch": 0.2797665983283394, + "grad_norm": 2.3835155963897705, + "learning_rate": 1.6903914842797736e-05, + "loss": 0.551, + "step": 8870 + }, + { + "epoch": 0.27979813909478, + "grad_norm": 2.0097382068634033, + "learning_rate": 1.690317577607265e-05, + "loss": 0.5543, + "step": 8871 + }, + { + "epoch": 0.27982967986122065, + "grad_norm": 1.5570924282073975, + "learning_rate": 1.690243663730756e-05, + "loss": 0.5001, + "step": 8872 + }, + { + "epoch": 0.2798612206276613, + "grad_norm": 1.5277173519134521, + "learning_rate": 1.6901697426510186e-05, + "loss": 0.8677, + "step": 8873 + }, + { + "epoch": 0.27989276139410185, + "grad_norm": 1.7424732446670532, + "learning_rate": 1.690095814368824e-05, + "loss": 0.5453, + "step": 8874 + }, + { + "epoch": 0.2799243021605425, + "grad_norm": 1.4983164072036743, + "learning_rate": 1.6900218788849438e-05, + "loss": 0.8546, + "step": 8875 + }, + { + "epoch": 0.2799558429269831, + "grad_norm": 1.817829966545105, + "learning_rate": 1.6899479362001498e-05, + "loss": 0.5154, + "step": 8876 + }, + { + "epoch": 0.27998738369342374, + "grad_norm": 1.7496671676635742, + "learning_rate": 1.6898739863152128e-05, + "loss": 0.5633, + "step": 8877 + }, + { + "epoch": 0.28001892445986437, + "grad_norm": 2.607194662094116, + "learning_rate": 1.6898000292309054e-05, + "loss": 0.8314, + "step": 8878 + }, + { + "epoch": 0.280050465226305, + "grad_norm": 2.2374107837677, + "learning_rate": 1.689726064947999e-05, + "loss": 0.5754, + "step": 8879 + }, + { + "epoch": 0.28008200599274563, + "grad_norm": 1.9900529384613037, + "learning_rate": 1.6896520934672652e-05, + "loss": 0.9051, + "step": 8880 + }, + { + "epoch": 0.28011354675918626, + "grad_norm": 2.121316909790039, + "learning_rate": 1.6895781147894768e-05, + "loss": 0.5478, + "step": 8881 + }, + { + "epoch": 0.2801450875256269, + "grad_norm": 1.7456082105636597, + "learning_rate": 1.6895041289154052e-05, + "loss": 0.5543, + "step": 8882 + }, + { + "epoch": 0.2801766282920675, + "grad_norm": 2.681425094604492, + "learning_rate": 1.6894301358458226e-05, + "loss": 0.5405, + "step": 8883 + }, + { + "epoch": 0.2802081690585081, + "grad_norm": 1.520898461341858, + "learning_rate": 1.689356135581501e-05, + "loss": 0.5377, + "step": 8884 + }, + { + "epoch": 0.2802397098249487, + "grad_norm": 2.4075942039489746, + "learning_rate": 1.6892821281232133e-05, + "loss": 0.5388, + "step": 8885 + }, + { + "epoch": 0.28027125059138935, + "grad_norm": 2.1284430027008057, + "learning_rate": 1.6892081134717312e-05, + "loss": 0.5527, + "step": 8886 + }, + { + "epoch": 0.28030279135783, + "grad_norm": 1.7295427322387695, + "learning_rate": 1.6891340916278268e-05, + "loss": 0.5176, + "step": 8887 + }, + { + "epoch": 0.2803343321242706, + "grad_norm": 2.240882635116577, + "learning_rate": 1.689060062592274e-05, + "loss": 0.5206, + "step": 8888 + }, + { + "epoch": 0.28036587289071124, + "grad_norm": 1.640700340270996, + "learning_rate": 1.688986026365844e-05, + "loss": 0.4057, + "step": 8889 + }, + { + "epoch": 0.2803974136571519, + "grad_norm": 2.617180824279785, + "learning_rate": 1.6889119829493093e-05, + "loss": 0.9061, + "step": 8890 + }, + { + "epoch": 0.2804289544235925, + "grad_norm": 2.032600164413452, + "learning_rate": 1.6888379323434433e-05, + "loss": 0.5717, + "step": 8891 + }, + { + "epoch": 0.28046049519003313, + "grad_norm": 1.6616394519805908, + "learning_rate": 1.688763874549019e-05, + "loss": 0.8672, + "step": 8892 + }, + { + "epoch": 0.28049203595647376, + "grad_norm": 2.413271188735962, + "learning_rate": 1.6886898095668086e-05, + "loss": 0.523, + "step": 8893 + }, + { + "epoch": 0.2805235767229144, + "grad_norm": 2.7391128540039062, + "learning_rate": 1.6886157373975854e-05, + "loss": 0.5652, + "step": 8894 + }, + { + "epoch": 0.28055511748935497, + "grad_norm": 3.182950973510742, + "learning_rate": 1.6885416580421226e-05, + "loss": 0.522, + "step": 8895 + }, + { + "epoch": 0.2805866582557956, + "grad_norm": 2.199956178665161, + "learning_rate": 1.6884675715011925e-05, + "loss": 0.5075, + "step": 8896 + }, + { + "epoch": 0.2806181990222362, + "grad_norm": 1.7840275764465332, + "learning_rate": 1.6883934777755687e-05, + "loss": 0.4165, + "step": 8897 + }, + { + "epoch": 0.28064973978867686, + "grad_norm": 2.1306169033050537, + "learning_rate": 1.6883193768660247e-05, + "loss": 0.5602, + "step": 8898 + }, + { + "epoch": 0.2806812805551175, + "grad_norm": 2.1502525806427, + "learning_rate": 1.6882452687733333e-05, + "loss": 0.5344, + "step": 8899 + }, + { + "epoch": 0.2807128213215581, + "grad_norm": 3.1537375450134277, + "learning_rate": 1.688171153498268e-05, + "loss": 0.5395, + "step": 8900 + }, + { + "epoch": 0.28074436208799874, + "grad_norm": 1.4599312543869019, + "learning_rate": 1.6880970310416028e-05, + "loss": 0.5413, + "step": 8901 + }, + { + "epoch": 0.2807759028544394, + "grad_norm": 3.594083309173584, + "learning_rate": 1.6880229014041107e-05, + "loss": 0.5533, + "step": 8902 + }, + { + "epoch": 0.28080744362088, + "grad_norm": 1.4515703916549683, + "learning_rate": 1.6879487645865652e-05, + "loss": 0.3999, + "step": 8903 + }, + { + "epoch": 0.28083898438732063, + "grad_norm": 2.086235523223877, + "learning_rate": 1.68787462058974e-05, + "loss": 0.5654, + "step": 8904 + }, + { + "epoch": 0.28087052515376126, + "grad_norm": 1.6152809858322144, + "learning_rate": 1.6878004694144094e-05, + "loss": 0.5247, + "step": 8905 + }, + { + "epoch": 0.28090206592020184, + "grad_norm": 1.4626179933547974, + "learning_rate": 1.6877263110613467e-05, + "loss": 0.5197, + "step": 8906 + }, + { + "epoch": 0.28093360668664247, + "grad_norm": 2.094265937805176, + "learning_rate": 1.6876521455313258e-05, + "loss": 0.5666, + "step": 8907 + }, + { + "epoch": 0.2809651474530831, + "grad_norm": 1.9324899911880493, + "learning_rate": 1.687577972825121e-05, + "loss": 0.8512, + "step": 8908 + }, + { + "epoch": 0.2809966882195237, + "grad_norm": 1.8093880414962769, + "learning_rate": 1.687503792943506e-05, + "loss": 0.4232, + "step": 8909 + }, + { + "epoch": 0.28102822898596436, + "grad_norm": 2.5142667293548584, + "learning_rate": 1.6874296058872552e-05, + "loss": 0.828, + "step": 8910 + }, + { + "epoch": 0.281059769752405, + "grad_norm": 1.912563681602478, + "learning_rate": 1.6873554116571427e-05, + "loss": 0.5628, + "step": 8911 + }, + { + "epoch": 0.2810913105188456, + "grad_norm": 1.3934365510940552, + "learning_rate": 1.687281210253943e-05, + "loss": 0.4079, + "step": 8912 + }, + { + "epoch": 0.28112285128528625, + "grad_norm": 1.8284673690795898, + "learning_rate": 1.6872070016784295e-05, + "loss": 0.5778, + "step": 8913 + }, + { + "epoch": 0.2811543920517269, + "grad_norm": 1.842100739479065, + "learning_rate": 1.687132785931378e-05, + "loss": 0.5137, + "step": 8914 + }, + { + "epoch": 0.2811859328181675, + "grad_norm": 2.024345636367798, + "learning_rate": 1.6870585630135622e-05, + "loss": 0.5681, + "step": 8915 + }, + { + "epoch": 0.2812174735846081, + "grad_norm": 1.466995358467102, + "learning_rate": 1.6869843329257566e-05, + "loss": 0.5371, + "step": 8916 + }, + { + "epoch": 0.2812490143510487, + "grad_norm": 3.934852123260498, + "learning_rate": 1.686910095668736e-05, + "loss": 0.3928, + "step": 8917 + }, + { + "epoch": 0.28128055511748934, + "grad_norm": 2.411628246307373, + "learning_rate": 1.6868358512432755e-05, + "loss": 0.5779, + "step": 8918 + }, + { + "epoch": 0.28131209588392997, + "grad_norm": 2.0724635124206543, + "learning_rate": 1.6867615996501494e-05, + "loss": 0.5565, + "step": 8919 + }, + { + "epoch": 0.2813436366503706, + "grad_norm": 2.069687843322754, + "learning_rate": 1.686687340890133e-05, + "loss": 0.5508, + "step": 8920 + }, + { + "epoch": 0.28137517741681123, + "grad_norm": 2.4466006755828857, + "learning_rate": 1.6866130749640004e-05, + "loss": 0.5738, + "step": 8921 + }, + { + "epoch": 0.28140671818325186, + "grad_norm": 1.8577433824539185, + "learning_rate": 1.6865388018725276e-05, + "loss": 0.552, + "step": 8922 + }, + { + "epoch": 0.2814382589496925, + "grad_norm": 1.9127320051193237, + "learning_rate": 1.6864645216164893e-05, + "loss": 0.5555, + "step": 8923 + }, + { + "epoch": 0.2814697997161331, + "grad_norm": 2.3710289001464844, + "learning_rate": 1.6863902341966606e-05, + "loss": 0.548, + "step": 8924 + }, + { + "epoch": 0.28150134048257375, + "grad_norm": 1.928346872329712, + "learning_rate": 1.6863159396138173e-05, + "loss": 0.5561, + "step": 8925 + }, + { + "epoch": 0.2815328812490144, + "grad_norm": 1.946969747543335, + "learning_rate": 1.686241637868734e-05, + "loss": 0.5544, + "step": 8926 + }, + { + "epoch": 0.28156442201545495, + "grad_norm": 1.648935317993164, + "learning_rate": 1.686167328962186e-05, + "loss": 0.5301, + "step": 8927 + }, + { + "epoch": 0.2815959627818956, + "grad_norm": 3.939547061920166, + "learning_rate": 1.6860930128949492e-05, + "loss": 0.5506, + "step": 8928 + }, + { + "epoch": 0.2816275035483362, + "grad_norm": 1.4482405185699463, + "learning_rate": 1.6860186896677996e-05, + "loss": 0.5501, + "step": 8929 + }, + { + "epoch": 0.28165904431477684, + "grad_norm": 1.2490042448043823, + "learning_rate": 1.6859443592815118e-05, + "loss": 0.5637, + "step": 8930 + }, + { + "epoch": 0.28169058508121747, + "grad_norm": 1.4193676710128784, + "learning_rate": 1.6858700217368622e-05, + "loss": 0.5687, + "step": 8931 + }, + { + "epoch": 0.2817221258476581, + "grad_norm": 1.437089204788208, + "learning_rate": 1.6857956770346264e-05, + "loss": 0.5365, + "step": 8932 + }, + { + "epoch": 0.28175366661409873, + "grad_norm": 1.3980143070220947, + "learning_rate": 1.68572132517558e-05, + "loss": 0.5266, + "step": 8933 + }, + { + "epoch": 0.28178520738053936, + "grad_norm": 1.334481120109558, + "learning_rate": 1.685646966160499e-05, + "loss": 0.4297, + "step": 8934 + }, + { + "epoch": 0.28181674814698, + "grad_norm": 1.5209932327270508, + "learning_rate": 1.6855725999901598e-05, + "loss": 0.5403, + "step": 8935 + }, + { + "epoch": 0.2818482889134206, + "grad_norm": 1.940477728843689, + "learning_rate": 1.6854982266653383e-05, + "loss": 0.5587, + "step": 8936 + }, + { + "epoch": 0.2818798296798612, + "grad_norm": 1.6660157442092896, + "learning_rate": 1.6854238461868102e-05, + "loss": 0.5189, + "step": 8937 + }, + { + "epoch": 0.2819113704463018, + "grad_norm": 1.8820455074310303, + "learning_rate": 1.685349458555352e-05, + "loss": 0.5247, + "step": 8938 + }, + { + "epoch": 0.28194291121274245, + "grad_norm": 1.662661075592041, + "learning_rate": 1.6852750637717407e-05, + "loss": 0.3936, + "step": 8939 + }, + { + "epoch": 0.2819744519791831, + "grad_norm": 1.4947295188903809, + "learning_rate": 1.6852006618367513e-05, + "loss": 0.5303, + "step": 8940 + }, + { + "epoch": 0.2820059927456237, + "grad_norm": 2.223937749862671, + "learning_rate": 1.6851262527511613e-05, + "loss": 0.5555, + "step": 8941 + }, + { + "epoch": 0.28203753351206434, + "grad_norm": 1.9304291009902954, + "learning_rate": 1.6850518365157465e-05, + "loss": 0.385, + "step": 8942 + }, + { + "epoch": 0.28206907427850497, + "grad_norm": 1.7532728910446167, + "learning_rate": 1.6849774131312843e-05, + "loss": 0.5369, + "step": 8943 + }, + { + "epoch": 0.2821006150449456, + "grad_norm": 1.5504226684570312, + "learning_rate": 1.6849029825985507e-05, + "loss": 0.5302, + "step": 8944 + }, + { + "epoch": 0.28213215581138623, + "grad_norm": 2.180985450744629, + "learning_rate": 1.6848285449183227e-05, + "loss": 0.4224, + "step": 8945 + }, + { + "epoch": 0.28216369657782686, + "grad_norm": 1.8062245845794678, + "learning_rate": 1.6847541000913772e-05, + "loss": 0.5375, + "step": 8946 + }, + { + "epoch": 0.2821952373442675, + "grad_norm": 1.9786919355392456, + "learning_rate": 1.684679648118491e-05, + "loss": 0.5374, + "step": 8947 + }, + { + "epoch": 0.28222677811070807, + "grad_norm": 2.101066827774048, + "learning_rate": 1.6846051890004406e-05, + "loss": 0.5657, + "step": 8948 + }, + { + "epoch": 0.2822583188771487, + "grad_norm": 1.6928035020828247, + "learning_rate": 1.684530722738004e-05, + "loss": 0.4983, + "step": 8949 + }, + { + "epoch": 0.2822898596435893, + "grad_norm": 2.0784921646118164, + "learning_rate": 1.6844562493319575e-05, + "loss": 0.5322, + "step": 8950 + }, + { + "epoch": 0.28232140041002995, + "grad_norm": 1.6984295845031738, + "learning_rate": 1.6843817687830787e-05, + "loss": 0.5302, + "step": 8951 + }, + { + "epoch": 0.2823529411764706, + "grad_norm": 2.1598079204559326, + "learning_rate": 1.6843072810921444e-05, + "loss": 0.5228, + "step": 8952 + }, + { + "epoch": 0.2823844819429112, + "grad_norm": 2.0475544929504395, + "learning_rate": 1.6842327862599326e-05, + "loss": 0.4225, + "step": 8953 + }, + { + "epoch": 0.28241602270935184, + "grad_norm": 1.8144642114639282, + "learning_rate": 1.6841582842872202e-05, + "loss": 0.5483, + "step": 8954 + }, + { + "epoch": 0.2824475634757925, + "grad_norm": 2.5778520107269287, + "learning_rate": 1.684083775174785e-05, + "loss": 0.8887, + "step": 8955 + }, + { + "epoch": 0.2824791042422331, + "grad_norm": 2.11324143409729, + "learning_rate": 1.684009258923404e-05, + "loss": 0.5329, + "step": 8956 + }, + { + "epoch": 0.28251064500867373, + "grad_norm": 1.2658717632293701, + "learning_rate": 1.6839347355338563e-05, + "loss": 0.4357, + "step": 8957 + }, + { + "epoch": 0.28254218577511436, + "grad_norm": 1.6345539093017578, + "learning_rate": 1.683860205006918e-05, + "loss": 0.5826, + "step": 8958 + }, + { + "epoch": 0.28257372654155494, + "grad_norm": 4.754882335662842, + "learning_rate": 1.6837856673433673e-05, + "loss": 0.563, + "step": 8959 + }, + { + "epoch": 0.28260526730799557, + "grad_norm": 1.9515670537948608, + "learning_rate": 1.6837111225439823e-05, + "loss": 0.5003, + "step": 8960 + }, + { + "epoch": 0.2826368080744362, + "grad_norm": 10.589337348937988, + "learning_rate": 1.6836365706095407e-05, + "loss": 0.8957, + "step": 8961 + }, + { + "epoch": 0.2826683488408768, + "grad_norm": 1.7992888689041138, + "learning_rate": 1.6835620115408205e-05, + "loss": 0.5498, + "step": 8962 + }, + { + "epoch": 0.28269988960731746, + "grad_norm": 1.811360478401184, + "learning_rate": 1.683487445338601e-05, + "loss": 0.5841, + "step": 8963 + }, + { + "epoch": 0.2827314303737581, + "grad_norm": 1.7331104278564453, + "learning_rate": 1.683412872003658e-05, + "loss": 0.5481, + "step": 8964 + }, + { + "epoch": 0.2827629711401987, + "grad_norm": 2.7974133491516113, + "learning_rate": 1.6833382915367718e-05, + "loss": 0.5583, + "step": 8965 + }, + { + "epoch": 0.28279451190663935, + "grad_norm": 2.7899882793426514, + "learning_rate": 1.6832637039387195e-05, + "loss": 0.557, + "step": 8966 + }, + { + "epoch": 0.28282605267308, + "grad_norm": 1.6588447093963623, + "learning_rate": 1.68318910921028e-05, + "loss": 0.5352, + "step": 8967 + }, + { + "epoch": 0.2828575934395206, + "grad_norm": 2.2859742641448975, + "learning_rate": 1.683114507352232e-05, + "loss": 0.5459, + "step": 8968 + }, + { + "epoch": 0.2828891342059612, + "grad_norm": 1.6896637678146362, + "learning_rate": 1.6830398983653533e-05, + "loss": 0.553, + "step": 8969 + }, + { + "epoch": 0.2829206749724018, + "grad_norm": 1.8773486614227295, + "learning_rate": 1.6829652822504233e-05, + "loss": 0.8993, + "step": 8970 + }, + { + "epoch": 0.28295221573884244, + "grad_norm": 1.5341570377349854, + "learning_rate": 1.68289065900822e-05, + "loss": 0.5596, + "step": 8971 + }, + { + "epoch": 0.28298375650528307, + "grad_norm": 2.201106309890747, + "learning_rate": 1.6828160286395222e-05, + "loss": 0.3849, + "step": 8972 + }, + { + "epoch": 0.2830152972717237, + "grad_norm": 3.8059420585632324, + "learning_rate": 1.6827413911451094e-05, + "loss": 0.5762, + "step": 8973 + }, + { + "epoch": 0.28304683803816433, + "grad_norm": 1.9354612827301025, + "learning_rate": 1.68266674652576e-05, + "loss": 0.538, + "step": 8974 + }, + { + "epoch": 0.28307837880460496, + "grad_norm": 1.5802593231201172, + "learning_rate": 1.6825920947822525e-05, + "loss": 0.3994, + "step": 8975 + }, + { + "epoch": 0.2831099195710456, + "grad_norm": 2.2182133197784424, + "learning_rate": 1.6825174359153667e-05, + "loss": 0.5025, + "step": 8976 + }, + { + "epoch": 0.2831414603374862, + "grad_norm": 1.8596686124801636, + "learning_rate": 1.682442769925882e-05, + "loss": 0.5366, + "step": 8977 + }, + { + "epoch": 0.28317300110392685, + "grad_norm": 1.8024709224700928, + "learning_rate": 1.6823680968145765e-05, + "loss": 0.5375, + "step": 8978 + }, + { + "epoch": 0.2832045418703675, + "grad_norm": 1.656270980834961, + "learning_rate": 1.6822934165822303e-05, + "loss": 0.8268, + "step": 8979 + }, + { + "epoch": 0.28323608263680805, + "grad_norm": 1.4834262132644653, + "learning_rate": 1.6822187292296227e-05, + "loss": 0.5405, + "step": 8980 + }, + { + "epoch": 0.2832676234032487, + "grad_norm": 1.9455515146255493, + "learning_rate": 1.6821440347575325e-05, + "loss": 0.8284, + "step": 8981 + }, + { + "epoch": 0.2832991641696893, + "grad_norm": 1.4703317880630493, + "learning_rate": 1.68206933316674e-05, + "loss": 0.524, + "step": 8982 + }, + { + "epoch": 0.28333070493612994, + "grad_norm": 1.7398521900177002, + "learning_rate": 1.6819946244580242e-05, + "loss": 0.5068, + "step": 8983 + }, + { + "epoch": 0.28336224570257057, + "grad_norm": 2.079953908920288, + "learning_rate": 1.681919908632165e-05, + "loss": 0.547, + "step": 8984 + }, + { + "epoch": 0.2833937864690112, + "grad_norm": 2.478269100189209, + "learning_rate": 1.6818451856899422e-05, + "loss": 0.5335, + "step": 8985 + }, + { + "epoch": 0.28342532723545183, + "grad_norm": 1.8409333229064941, + "learning_rate": 1.681770455632135e-05, + "loss": 0.4935, + "step": 8986 + }, + { + "epoch": 0.28345686800189246, + "grad_norm": 1.8743600845336914, + "learning_rate": 1.6816957184595243e-05, + "loss": 0.5095, + "step": 8987 + }, + { + "epoch": 0.2834884087683331, + "grad_norm": 2.754469156265259, + "learning_rate": 1.681620974172889e-05, + "loss": 0.5517, + "step": 8988 + }, + { + "epoch": 0.2835199495347737, + "grad_norm": 3.2199859619140625, + "learning_rate": 1.6815462227730098e-05, + "loss": 0.5632, + "step": 8989 + }, + { + "epoch": 0.2835514903012143, + "grad_norm": 1.8539655208587646, + "learning_rate": 1.6814714642606663e-05, + "loss": 0.5413, + "step": 8990 + }, + { + "epoch": 0.2835830310676549, + "grad_norm": 2.442668914794922, + "learning_rate": 1.6813966986366393e-05, + "loss": 0.5641, + "step": 8991 + }, + { + "epoch": 0.28361457183409555, + "grad_norm": 1.6279553174972534, + "learning_rate": 1.6813219259017084e-05, + "loss": 0.5149, + "step": 8992 + }, + { + "epoch": 0.2836461126005362, + "grad_norm": 2.4200994968414307, + "learning_rate": 1.681247146056654e-05, + "loss": 0.5396, + "step": 8993 + }, + { + "epoch": 0.2836776533669768, + "grad_norm": 1.9047380685806274, + "learning_rate": 1.6811723591022572e-05, + "loss": 0.5574, + "step": 8994 + }, + { + "epoch": 0.28370919413341744, + "grad_norm": 2.494518518447876, + "learning_rate": 1.6810975650392975e-05, + "loss": 0.8565, + "step": 8995 + }, + { + "epoch": 0.28374073489985807, + "grad_norm": 2.405066728591919, + "learning_rate": 1.681022763868556e-05, + "loss": 0.5664, + "step": 8996 + }, + { + "epoch": 0.2837722756662987, + "grad_norm": 1.8512521982192993, + "learning_rate": 1.680947955590813e-05, + "loss": 0.5476, + "step": 8997 + }, + { + "epoch": 0.28380381643273933, + "grad_norm": 5.954269886016846, + "learning_rate": 1.6808731402068497e-05, + "loss": 0.4379, + "step": 8998 + }, + { + "epoch": 0.28383535719917996, + "grad_norm": 1.7200958728790283, + "learning_rate": 1.680798317717446e-05, + "loss": 0.5745, + "step": 8999 + }, + { + "epoch": 0.2838668979656206, + "grad_norm": 2.4487040042877197, + "learning_rate": 1.6807234881233837e-05, + "loss": 0.3758, + "step": 9000 + }, + { + "epoch": 0.28389843873206116, + "grad_norm": 2.3271381855010986, + "learning_rate": 1.680648651425443e-05, + "loss": 0.5491, + "step": 9001 + }, + { + "epoch": 0.2839299794985018, + "grad_norm": 2.8655447959899902, + "learning_rate": 1.6805738076244053e-05, + "loss": 0.5498, + "step": 9002 + }, + { + "epoch": 0.2839615202649424, + "grad_norm": 1.8760117292404175, + "learning_rate": 1.680498956721051e-05, + "loss": 0.5421, + "step": 9003 + }, + { + "epoch": 0.28399306103138305, + "grad_norm": 1.8217540979385376, + "learning_rate": 1.680424098716162e-05, + "loss": 0.5571, + "step": 9004 + }, + { + "epoch": 0.2840246017978237, + "grad_norm": 1.7562216520309448, + "learning_rate": 1.6803492336105195e-05, + "loss": 0.5266, + "step": 9005 + }, + { + "epoch": 0.2840561425642643, + "grad_norm": 1.7628432512283325, + "learning_rate": 1.680274361404904e-05, + "loss": 0.5268, + "step": 9006 + }, + { + "epoch": 0.28408768333070494, + "grad_norm": 2.1887097358703613, + "learning_rate": 1.680199482100097e-05, + "loss": 0.845, + "step": 9007 + }, + { + "epoch": 0.2841192240971456, + "grad_norm": 1.5447156429290771, + "learning_rate": 1.6801245956968812e-05, + "loss": 0.4547, + "step": 9008 + }, + { + "epoch": 0.2841507648635862, + "grad_norm": 2.29720401763916, + "learning_rate": 1.6800497021960364e-05, + "loss": 0.5524, + "step": 9009 + }, + { + "epoch": 0.28418230563002683, + "grad_norm": 2.5709609985351562, + "learning_rate": 1.6799748015983453e-05, + "loss": 0.519, + "step": 9010 + }, + { + "epoch": 0.2842138463964674, + "grad_norm": 2.773650884628296, + "learning_rate": 1.6798998939045893e-05, + "loss": 0.5066, + "step": 9011 + }, + { + "epoch": 0.28424538716290804, + "grad_norm": 4.0754899978637695, + "learning_rate": 1.6798249791155497e-05, + "loss": 0.5498, + "step": 9012 + }, + { + "epoch": 0.28427692792934867, + "grad_norm": 1.6870105266571045, + "learning_rate": 1.6797500572320084e-05, + "loss": 0.5488, + "step": 9013 + }, + { + "epoch": 0.2843084686957893, + "grad_norm": 1.673525333404541, + "learning_rate": 1.6796751282547478e-05, + "loss": 0.5451, + "step": 9014 + }, + { + "epoch": 0.2843400094622299, + "grad_norm": 1.581194519996643, + "learning_rate": 1.679600192184549e-05, + "loss": 0.5359, + "step": 9015 + }, + { + "epoch": 0.28437155022867056, + "grad_norm": 2.3751866817474365, + "learning_rate": 1.6795252490221948e-05, + "loss": 0.5301, + "step": 9016 + }, + { + "epoch": 0.2844030909951112, + "grad_norm": 1.8819090127944946, + "learning_rate": 1.679450298768467e-05, + "loss": 0.5587, + "step": 9017 + }, + { + "epoch": 0.2844346317615518, + "grad_norm": 1.6801707744598389, + "learning_rate": 1.679375341424148e-05, + "loss": 0.5209, + "step": 9018 + }, + { + "epoch": 0.28446617252799244, + "grad_norm": 1.8306474685668945, + "learning_rate": 1.6793003769900194e-05, + "loss": 0.5334, + "step": 9019 + }, + { + "epoch": 0.2844977132944331, + "grad_norm": 1.6223974227905273, + "learning_rate": 1.679225405466864e-05, + "loss": 0.5307, + "step": 9020 + }, + { + "epoch": 0.2845292540608737, + "grad_norm": 1.8884316682815552, + "learning_rate": 1.6791504268554643e-05, + "loss": 0.5244, + "step": 9021 + }, + { + "epoch": 0.2845607948273143, + "grad_norm": 1.9289735555648804, + "learning_rate": 1.6790754411566024e-05, + "loss": 0.5593, + "step": 9022 + }, + { + "epoch": 0.2845923355937549, + "grad_norm": 2.03302264213562, + "learning_rate": 1.679000448371061e-05, + "loss": 0.5535, + "step": 9023 + }, + { + "epoch": 0.28462387636019554, + "grad_norm": 2.8753395080566406, + "learning_rate": 1.6789254484996225e-05, + "loss": 0.5297, + "step": 9024 + }, + { + "epoch": 0.28465541712663617, + "grad_norm": 2.241300344467163, + "learning_rate": 1.67885044154307e-05, + "loss": 0.5514, + "step": 9025 + }, + { + "epoch": 0.2846869578930768, + "grad_norm": 2.3438429832458496, + "learning_rate": 1.678775427502186e-05, + "loss": 0.547, + "step": 9026 + }, + { + "epoch": 0.2847184986595174, + "grad_norm": 2.0772461891174316, + "learning_rate": 1.6787004063777533e-05, + "loss": 0.8695, + "step": 9027 + }, + { + "epoch": 0.28475003942595806, + "grad_norm": 1.9492770433425903, + "learning_rate": 1.678625378170555e-05, + "loss": 0.5591, + "step": 9028 + }, + { + "epoch": 0.2847815801923987, + "grad_norm": 1.9408369064331055, + "learning_rate": 1.6785503428813735e-05, + "loss": 0.5281, + "step": 9029 + }, + { + "epoch": 0.2848131209588393, + "grad_norm": 2.5019168853759766, + "learning_rate": 1.678475300510993e-05, + "loss": 0.5301, + "step": 9030 + }, + { + "epoch": 0.28484466172527995, + "grad_norm": 1.5042743682861328, + "learning_rate": 1.6784002510601953e-05, + "loss": 0.5299, + "step": 9031 + }, + { + "epoch": 0.2848762024917206, + "grad_norm": 2.433593511581421, + "learning_rate": 1.6783251945297644e-05, + "loss": 0.4155, + "step": 9032 + }, + { + "epoch": 0.28490774325816115, + "grad_norm": 2.0574214458465576, + "learning_rate": 1.6782501309204834e-05, + "loss": 0.5524, + "step": 9033 + }, + { + "epoch": 0.2849392840246018, + "grad_norm": 1.8042798042297363, + "learning_rate": 1.678175060233136e-05, + "loss": 0.5333, + "step": 9034 + }, + { + "epoch": 0.2849708247910424, + "grad_norm": 3.3794748783111572, + "learning_rate": 1.6780999824685048e-05, + "loss": 0.5299, + "step": 9035 + }, + { + "epoch": 0.28500236555748304, + "grad_norm": 1.6090761423110962, + "learning_rate": 1.6780248976273736e-05, + "loss": 0.5432, + "step": 9036 + }, + { + "epoch": 0.28503390632392367, + "grad_norm": 1.6178237199783325, + "learning_rate": 1.6779498057105263e-05, + "loss": 0.5364, + "step": 9037 + }, + { + "epoch": 0.2850654470903643, + "grad_norm": 3.1890318393707275, + "learning_rate": 1.6778747067187468e-05, + "loss": 0.5256, + "step": 9038 + }, + { + "epoch": 0.28509698785680493, + "grad_norm": 1.5537879467010498, + "learning_rate": 1.6777996006528178e-05, + "loss": 0.3692, + "step": 9039 + }, + { + "epoch": 0.28512852862324556, + "grad_norm": 1.8805339336395264, + "learning_rate": 1.677724487513524e-05, + "loss": 0.5395, + "step": 9040 + }, + { + "epoch": 0.2851600693896862, + "grad_norm": 1.759102463722229, + "learning_rate": 1.6776493673016488e-05, + "loss": 0.5583, + "step": 9041 + }, + { + "epoch": 0.2851916101561268, + "grad_norm": 1.5688140392303467, + "learning_rate": 1.6775742400179765e-05, + "loss": 0.5917, + "step": 9042 + }, + { + "epoch": 0.2852231509225674, + "grad_norm": 1.5496883392333984, + "learning_rate": 1.6774991056632904e-05, + "loss": 0.5469, + "step": 9043 + }, + { + "epoch": 0.285254691689008, + "grad_norm": 2.3168163299560547, + "learning_rate": 1.6774239642383756e-05, + "loss": 0.8817, + "step": 9044 + }, + { + "epoch": 0.28528623245544865, + "grad_norm": 2.312460422515869, + "learning_rate": 1.6773488157440154e-05, + "loss": 0.5486, + "step": 9045 + }, + { + "epoch": 0.2853177732218893, + "grad_norm": 1.509342908859253, + "learning_rate": 1.6772736601809944e-05, + "loss": 0.5533, + "step": 9046 + }, + { + "epoch": 0.2853493139883299, + "grad_norm": 1.6991956233978271, + "learning_rate": 1.677198497550097e-05, + "loss": 0.5257, + "step": 9047 + }, + { + "epoch": 0.28538085475477054, + "grad_norm": 1.6335759162902832, + "learning_rate": 1.6771233278521072e-05, + "loss": 0.547, + "step": 9048 + }, + { + "epoch": 0.28541239552121117, + "grad_norm": 1.5892280340194702, + "learning_rate": 1.67704815108781e-05, + "loss": 0.868, + "step": 9049 + }, + { + "epoch": 0.2854439362876518, + "grad_norm": 1.7917975187301636, + "learning_rate": 1.6769729672579894e-05, + "loss": 0.5391, + "step": 9050 + }, + { + "epoch": 0.28547547705409243, + "grad_norm": 1.7033934593200684, + "learning_rate": 1.6768977763634305e-05, + "loss": 0.5488, + "step": 9051 + }, + { + "epoch": 0.28550701782053306, + "grad_norm": 1.8712671995162964, + "learning_rate": 1.676822578404917e-05, + "loss": 0.5276, + "step": 9052 + }, + { + "epoch": 0.2855385585869737, + "grad_norm": 1.9250866174697876, + "learning_rate": 1.6767473733832355e-05, + "loss": 0.5491, + "step": 9053 + }, + { + "epoch": 0.28557009935341426, + "grad_norm": 1.360823154449463, + "learning_rate": 1.6766721612991687e-05, + "loss": 0.5868, + "step": 9054 + }, + { + "epoch": 0.2856016401198549, + "grad_norm": 2.1880908012390137, + "learning_rate": 1.676596942153503e-05, + "loss": 0.8525, + "step": 9055 + }, + { + "epoch": 0.2856331808862955, + "grad_norm": 2.221759557723999, + "learning_rate": 1.676521715947023e-05, + "loss": 0.3874, + "step": 9056 + }, + { + "epoch": 0.28566472165273615, + "grad_norm": 2.0763766765594482, + "learning_rate": 1.676446482680513e-05, + "loss": 0.5566, + "step": 9057 + }, + { + "epoch": 0.2856962624191768, + "grad_norm": 1.7152948379516602, + "learning_rate": 1.676371242354759e-05, + "loss": 0.5465, + "step": 9058 + }, + { + "epoch": 0.2857278031856174, + "grad_norm": 2.159756898880005, + "learning_rate": 1.6762959949705456e-05, + "loss": 0.5781, + "step": 9059 + }, + { + "epoch": 0.28575934395205804, + "grad_norm": 1.507399559020996, + "learning_rate": 1.676220740528659e-05, + "loss": 0.5772, + "step": 9060 + }, + { + "epoch": 0.2857908847184987, + "grad_norm": 1.6670986413955688, + "learning_rate": 1.6761454790298836e-05, + "loss": 0.5358, + "step": 9061 + }, + { + "epoch": 0.2858224254849393, + "grad_norm": 1.9964245557785034, + "learning_rate": 1.6760702104750046e-05, + "loss": 0.365, + "step": 9062 + }, + { + "epoch": 0.28585396625137993, + "grad_norm": 2.4036366939544678, + "learning_rate": 1.6759949348648085e-05, + "loss": 0.539, + "step": 9063 + }, + { + "epoch": 0.2858855070178205, + "grad_norm": 1.5788401365280151, + "learning_rate": 1.6759196522000804e-05, + "loss": 0.5642, + "step": 9064 + }, + { + "epoch": 0.28591704778426114, + "grad_norm": 1.9304673671722412, + "learning_rate": 1.675844362481606e-05, + "loss": 0.5401, + "step": 9065 + }, + { + "epoch": 0.28594858855070177, + "grad_norm": 1.8209619522094727, + "learning_rate": 1.6757690657101706e-05, + "loss": 0.5553, + "step": 9066 + }, + { + "epoch": 0.2859801293171424, + "grad_norm": 1.7110788822174072, + "learning_rate": 1.6756937618865602e-05, + "loss": 0.5373, + "step": 9067 + }, + { + "epoch": 0.286011670083583, + "grad_norm": 1.5871986150741577, + "learning_rate": 1.675618451011561e-05, + "loss": 0.552, + "step": 9068 + }, + { + "epoch": 0.28604321085002365, + "grad_norm": 1.6405643224716187, + "learning_rate": 1.6755431330859587e-05, + "loss": 0.5337, + "step": 9069 + }, + { + "epoch": 0.2860747516164643, + "grad_norm": 1.9403774738311768, + "learning_rate": 1.6754678081105387e-05, + "loss": 0.5645, + "step": 9070 + }, + { + "epoch": 0.2861062923829049, + "grad_norm": 1.386564016342163, + "learning_rate": 1.6753924760860885e-05, + "loss": 0.5361, + "step": 9071 + }, + { + "epoch": 0.28613783314934554, + "grad_norm": 1.5155683755874634, + "learning_rate": 1.6753171370133928e-05, + "loss": 0.53, + "step": 9072 + }, + { + "epoch": 0.2861693739157862, + "grad_norm": 2.394850969314575, + "learning_rate": 1.6752417908932386e-05, + "loss": 0.555, + "step": 9073 + }, + { + "epoch": 0.2862009146822268, + "grad_norm": 1.9176725149154663, + "learning_rate": 1.675166437726412e-05, + "loss": 0.5295, + "step": 9074 + }, + { + "epoch": 0.2862324554486674, + "grad_norm": 1.2979567050933838, + "learning_rate": 1.6750910775136994e-05, + "loss": 0.5672, + "step": 9075 + }, + { + "epoch": 0.286263996215108, + "grad_norm": 1.8815786838531494, + "learning_rate": 1.6750157102558868e-05, + "loss": 0.5403, + "step": 9076 + }, + { + "epoch": 0.28629553698154864, + "grad_norm": 1.724144458770752, + "learning_rate": 1.6749403359537616e-05, + "loss": 0.5539, + "step": 9077 + }, + { + "epoch": 0.28632707774798927, + "grad_norm": 1.5333856344223022, + "learning_rate": 1.6748649546081096e-05, + "loss": 0.5212, + "step": 9078 + }, + { + "epoch": 0.2863586185144299, + "grad_norm": 1.9637104272842407, + "learning_rate": 1.674789566219718e-05, + "loss": 0.5555, + "step": 9079 + }, + { + "epoch": 0.2863901592808705, + "grad_norm": 2.289175033569336, + "learning_rate": 1.674714170789373e-05, + "loss": 0.8795, + "step": 9080 + }, + { + "epoch": 0.28642170004731116, + "grad_norm": 2.190659999847412, + "learning_rate": 1.6746387683178622e-05, + "loss": 0.8657, + "step": 9081 + }, + { + "epoch": 0.2864532408137518, + "grad_norm": 2.3288960456848145, + "learning_rate": 1.6745633588059716e-05, + "loss": 0.5288, + "step": 9082 + }, + { + "epoch": 0.2864847815801924, + "grad_norm": 3.0653834342956543, + "learning_rate": 1.6744879422544892e-05, + "loss": 0.5532, + "step": 9083 + }, + { + "epoch": 0.28651632234663305, + "grad_norm": 1.747379183769226, + "learning_rate": 1.674412518664201e-05, + "loss": 0.5456, + "step": 9084 + }, + { + "epoch": 0.2865478631130737, + "grad_norm": 2.3558712005615234, + "learning_rate": 1.674337088035894e-05, + "loss": 0.5246, + "step": 9085 + }, + { + "epoch": 0.28657940387951425, + "grad_norm": 2.574937343597412, + "learning_rate": 1.6742616503703567e-05, + "loss": 0.4104, + "step": 9086 + }, + { + "epoch": 0.2866109446459549, + "grad_norm": 1.6107068061828613, + "learning_rate": 1.6741862056683752e-05, + "loss": 0.5381, + "step": 9087 + }, + { + "epoch": 0.2866424854123955, + "grad_norm": 2.263679027557373, + "learning_rate": 1.6741107539307372e-05, + "loss": 0.3762, + "step": 9088 + }, + { + "epoch": 0.28667402617883614, + "grad_norm": 1.9367135763168335, + "learning_rate": 1.67403529515823e-05, + "loss": 0.5412, + "step": 9089 + }, + { + "epoch": 0.28670556694527677, + "grad_norm": 1.4197633266448975, + "learning_rate": 1.673959829351641e-05, + "loss": 0.541, + "step": 9090 + }, + { + "epoch": 0.2867371077117174, + "grad_norm": 1.570115327835083, + "learning_rate": 1.673884356511758e-05, + "loss": 0.5431, + "step": 9091 + }, + { + "epoch": 0.28676864847815803, + "grad_norm": 2.96016263961792, + "learning_rate": 1.673808876639368e-05, + "loss": 0.8653, + "step": 9092 + }, + { + "epoch": 0.28680018924459866, + "grad_norm": 2.1563124656677246, + "learning_rate": 1.67373338973526e-05, + "loss": 0.522, + "step": 9093 + }, + { + "epoch": 0.2868317300110393, + "grad_norm": 2.253978729248047, + "learning_rate": 1.6736578958002207e-05, + "loss": 0.8937, + "step": 9094 + }, + { + "epoch": 0.2868632707774799, + "grad_norm": 1.462328314781189, + "learning_rate": 1.673582394835038e-05, + "loss": 0.5325, + "step": 9095 + }, + { + "epoch": 0.2868948115439205, + "grad_norm": 4.491025924682617, + "learning_rate": 1.6735068868405e-05, + "loss": 0.5685, + "step": 9096 + }, + { + "epoch": 0.2869263523103611, + "grad_norm": 2.097702741622925, + "learning_rate": 1.6734313718173946e-05, + "loss": 0.5484, + "step": 9097 + }, + { + "epoch": 0.28695789307680175, + "grad_norm": 1.9290776252746582, + "learning_rate": 1.67335584976651e-05, + "loss": 0.5358, + "step": 9098 + }, + { + "epoch": 0.2869894338432424, + "grad_norm": 1.6168657541275024, + "learning_rate": 1.6732803206886342e-05, + "loss": 0.5177, + "step": 9099 + }, + { + "epoch": 0.287020974609683, + "grad_norm": 2.9583704471588135, + "learning_rate": 1.6732047845845553e-05, + "loss": 0.9015, + "step": 9100 + }, + { + "epoch": 0.28705251537612364, + "grad_norm": 2.2599666118621826, + "learning_rate": 1.673129241455062e-05, + "loss": 0.5556, + "step": 9101 + }, + { + "epoch": 0.28708405614256427, + "grad_norm": 1.5950393676757812, + "learning_rate": 1.6730536913009425e-05, + "loss": 0.5517, + "step": 9102 + }, + { + "epoch": 0.2871155969090049, + "grad_norm": 1.599971055984497, + "learning_rate": 1.672978134122985e-05, + "loss": 0.534, + "step": 9103 + }, + { + "epoch": 0.28714713767544553, + "grad_norm": 1.4389744997024536, + "learning_rate": 1.6729025699219783e-05, + "loss": 0.5431, + "step": 9104 + }, + { + "epoch": 0.28717867844188616, + "grad_norm": 2.5493969917297363, + "learning_rate": 1.6728269986987106e-05, + "loss": 0.8473, + "step": 9105 + }, + { + "epoch": 0.2872102192083268, + "grad_norm": 1.8832684755325317, + "learning_rate": 1.6727514204539704e-05, + "loss": 0.5385, + "step": 9106 + }, + { + "epoch": 0.28724175997476736, + "grad_norm": 2.3641774654388428, + "learning_rate": 1.672675835188547e-05, + "loss": 0.8262, + "step": 9107 + }, + { + "epoch": 0.287273300741208, + "grad_norm": 1.6418546438217163, + "learning_rate": 1.6726002429032292e-05, + "loss": 0.5244, + "step": 9108 + }, + { + "epoch": 0.2873048415076486, + "grad_norm": 1.6424825191497803, + "learning_rate": 1.6725246435988058e-05, + "loss": 0.4058, + "step": 9109 + }, + { + "epoch": 0.28733638227408925, + "grad_norm": 1.6196995973587036, + "learning_rate": 1.6724490372760648e-05, + "loss": 0.5377, + "step": 9110 + }, + { + "epoch": 0.2873679230405299, + "grad_norm": 1.6210023164749146, + "learning_rate": 1.6723734239357965e-05, + "loss": 0.5305, + "step": 9111 + }, + { + "epoch": 0.2873994638069705, + "grad_norm": 1.876457691192627, + "learning_rate": 1.672297803578789e-05, + "loss": 0.5464, + "step": 9112 + }, + { + "epoch": 0.28743100457341114, + "grad_norm": 1.3378510475158691, + "learning_rate": 1.6722221762058322e-05, + "loss": 0.5402, + "step": 9113 + }, + { + "epoch": 0.28746254533985177, + "grad_norm": 2.499396324157715, + "learning_rate": 1.672146541817715e-05, + "loss": 0.5536, + "step": 9114 + }, + { + "epoch": 0.2874940861062924, + "grad_norm": 1.4458624124526978, + "learning_rate": 1.6720709004152266e-05, + "loss": 0.5386, + "step": 9115 + }, + { + "epoch": 0.28752562687273303, + "grad_norm": 1.5067341327667236, + "learning_rate": 1.6719952519991566e-05, + "loss": 0.5392, + "step": 9116 + }, + { + "epoch": 0.2875571676391736, + "grad_norm": 2.1391263008117676, + "learning_rate": 1.6719195965702945e-05, + "loss": 0.4333, + "step": 9117 + }, + { + "epoch": 0.28758870840561424, + "grad_norm": 2.274399518966675, + "learning_rate": 1.6718439341294297e-05, + "loss": 0.5272, + "step": 9118 + }, + { + "epoch": 0.28762024917205486, + "grad_norm": 1.8824442625045776, + "learning_rate": 1.6717682646773513e-05, + "loss": 0.5174, + "step": 9119 + }, + { + "epoch": 0.2876517899384955, + "grad_norm": 1.754992127418518, + "learning_rate": 1.67169258821485e-05, + "loss": 0.5126, + "step": 9120 + }, + { + "epoch": 0.2876833307049361, + "grad_norm": 1.8188354969024658, + "learning_rate": 1.671616904742715e-05, + "loss": 0.537, + "step": 9121 + }, + { + "epoch": 0.28771487147137675, + "grad_norm": 1.8122066259384155, + "learning_rate": 1.6715412142617355e-05, + "loss": 0.5289, + "step": 9122 + }, + { + "epoch": 0.2877464122378174, + "grad_norm": 1.9538099765777588, + "learning_rate": 1.6714655167727026e-05, + "loss": 0.5496, + "step": 9123 + }, + { + "epoch": 0.287777953004258, + "grad_norm": 1.614912748336792, + "learning_rate": 1.671389812276405e-05, + "loss": 0.5422, + "step": 9124 + }, + { + "epoch": 0.28780949377069864, + "grad_norm": 1.9527267217636108, + "learning_rate": 1.671314100773634e-05, + "loss": 0.545, + "step": 9125 + }, + { + "epoch": 0.2878410345371393, + "grad_norm": 3.5344691276550293, + "learning_rate": 1.6712383822651788e-05, + "loss": 0.4612, + "step": 9126 + }, + { + "epoch": 0.2878725753035799, + "grad_norm": 1.5426322221755981, + "learning_rate": 1.67116265675183e-05, + "loss": 0.5422, + "step": 9127 + }, + { + "epoch": 0.2879041160700205, + "grad_norm": 1.7026985883712769, + "learning_rate": 1.6710869242343778e-05, + "loss": 0.5312, + "step": 9128 + }, + { + "epoch": 0.2879356568364611, + "grad_norm": 1.6548943519592285, + "learning_rate": 1.671011184713612e-05, + "loss": 0.5446, + "step": 9129 + }, + { + "epoch": 0.28796719760290174, + "grad_norm": 2.8941147327423096, + "learning_rate": 1.6709354381903236e-05, + "loss": 0.5276, + "step": 9130 + }, + { + "epoch": 0.28799873836934237, + "grad_norm": 1.6502091884613037, + "learning_rate": 1.6708596846653033e-05, + "loss": 0.5152, + "step": 9131 + }, + { + "epoch": 0.288030279135783, + "grad_norm": 8.495779991149902, + "learning_rate": 1.6707839241393412e-05, + "loss": 0.5697, + "step": 9132 + }, + { + "epoch": 0.2880618199022236, + "grad_norm": 1.7708934545516968, + "learning_rate": 1.670708156613228e-05, + "loss": 0.5452, + "step": 9133 + }, + { + "epoch": 0.28809336066866426, + "grad_norm": 2.3605377674102783, + "learning_rate": 1.6706323820877545e-05, + "loss": 0.5714, + "step": 9134 + }, + { + "epoch": 0.2881249014351049, + "grad_norm": 2.3200912475585938, + "learning_rate": 1.670556600563711e-05, + "loss": 0.5819, + "step": 9135 + }, + { + "epoch": 0.2881564422015455, + "grad_norm": 2.2301859855651855, + "learning_rate": 1.6704808120418888e-05, + "loss": 0.5324, + "step": 9136 + }, + { + "epoch": 0.28818798296798614, + "grad_norm": 1.5194530487060547, + "learning_rate": 1.6704050165230786e-05, + "loss": 0.522, + "step": 9137 + }, + { + "epoch": 0.2882195237344267, + "grad_norm": 1.6022223234176636, + "learning_rate": 1.670329214008072e-05, + "loss": 0.5089, + "step": 9138 + }, + { + "epoch": 0.28825106450086735, + "grad_norm": 3.3238637447357178, + "learning_rate": 1.6702534044976593e-05, + "loss": 0.5527, + "step": 9139 + }, + { + "epoch": 0.288282605267308, + "grad_norm": 1.780572772026062, + "learning_rate": 1.6701775879926318e-05, + "loss": 0.5202, + "step": 9140 + }, + { + "epoch": 0.2883141460337486, + "grad_norm": 3.500318765640259, + "learning_rate": 1.6701017644937806e-05, + "loss": 0.9318, + "step": 9141 + }, + { + "epoch": 0.28834568680018924, + "grad_norm": 2.177665948867798, + "learning_rate": 1.6700259340018976e-05, + "loss": 0.5281, + "step": 9142 + }, + { + "epoch": 0.28837722756662987, + "grad_norm": 6.185132026672363, + "learning_rate": 1.6699500965177735e-05, + "loss": 0.5125, + "step": 9143 + }, + { + "epoch": 0.2884087683330705, + "grad_norm": 2.1263678073883057, + "learning_rate": 1.6698742520421998e-05, + "loss": 0.8651, + "step": 9144 + }, + { + "epoch": 0.2884403090995111, + "grad_norm": 1.650347352027893, + "learning_rate": 1.6697984005759683e-05, + "loss": 0.5423, + "step": 9145 + }, + { + "epoch": 0.28847184986595176, + "grad_norm": 1.8649001121520996, + "learning_rate": 1.6697225421198705e-05, + "loss": 0.5359, + "step": 9146 + }, + { + "epoch": 0.2885033906323924, + "grad_norm": 2.009719133377075, + "learning_rate": 1.6696466766746976e-05, + "loss": 0.54, + "step": 9147 + }, + { + "epoch": 0.288534931398833, + "grad_norm": 2.679546356201172, + "learning_rate": 1.669570804241242e-05, + "loss": 0.5443, + "step": 9148 + }, + { + "epoch": 0.2885664721652736, + "grad_norm": 1.5007351636886597, + "learning_rate": 1.669494924820295e-05, + "loss": 0.5557, + "step": 9149 + }, + { + "epoch": 0.2885980129317142, + "grad_norm": 1.6950730085372925, + "learning_rate": 1.6694190384126487e-05, + "loss": 0.5098, + "step": 9150 + }, + { + "epoch": 0.28862955369815485, + "grad_norm": 2.4344916343688965, + "learning_rate": 1.669343145019095e-05, + "loss": 0.573, + "step": 9151 + }, + { + "epoch": 0.2886610944645955, + "grad_norm": 1.5298070907592773, + "learning_rate": 1.6692672446404256e-05, + "loss": 0.5065, + "step": 9152 + }, + { + "epoch": 0.2886926352310361, + "grad_norm": 2.354832410812378, + "learning_rate": 1.669191337277433e-05, + "loss": 0.5755, + "step": 9153 + }, + { + "epoch": 0.28872417599747674, + "grad_norm": 2.643301248550415, + "learning_rate": 1.6691154229309093e-05, + "loss": 0.5458, + "step": 9154 + }, + { + "epoch": 0.28875571676391737, + "grad_norm": 1.902624487876892, + "learning_rate": 1.6690395016016463e-05, + "loss": 0.5429, + "step": 9155 + }, + { + "epoch": 0.288787257530358, + "grad_norm": 5.27330207824707, + "learning_rate": 1.6689635732904368e-05, + "loss": 0.4425, + "step": 9156 + }, + { + "epoch": 0.28881879829679863, + "grad_norm": 1.7779797315597534, + "learning_rate": 1.6688876379980728e-05, + "loss": 0.5477, + "step": 9157 + }, + { + "epoch": 0.28885033906323926, + "grad_norm": 1.3707610368728638, + "learning_rate": 1.668811695725347e-05, + "loss": 0.5107, + "step": 9158 + }, + { + "epoch": 0.2888818798296799, + "grad_norm": 2.486720085144043, + "learning_rate": 1.6687357464730523e-05, + "loss": 0.5536, + "step": 9159 + }, + { + "epoch": 0.28891342059612046, + "grad_norm": 1.4942991733551025, + "learning_rate": 1.66865979024198e-05, + "loss": 0.5323, + "step": 9160 + }, + { + "epoch": 0.2889449613625611, + "grad_norm": 3.27107834815979, + "learning_rate": 1.6685838270329243e-05, + "loss": 0.519, + "step": 9161 + }, + { + "epoch": 0.2889765021290017, + "grad_norm": 2.2162816524505615, + "learning_rate": 1.668507856846677e-05, + "loss": 0.4165, + "step": 9162 + }, + { + "epoch": 0.28900804289544235, + "grad_norm": 2.1628036499023438, + "learning_rate": 1.6684318796840307e-05, + "loss": 0.5385, + "step": 9163 + }, + { + "epoch": 0.289039583661883, + "grad_norm": 2.8752548694610596, + "learning_rate": 1.6683558955457792e-05, + "loss": 0.9229, + "step": 9164 + }, + { + "epoch": 0.2890711244283236, + "grad_norm": 1.7965813875198364, + "learning_rate": 1.6682799044327148e-05, + "loss": 0.5202, + "step": 9165 + }, + { + "epoch": 0.28910266519476424, + "grad_norm": 1.5873929262161255, + "learning_rate": 1.6682039063456307e-05, + "loss": 0.5335, + "step": 9166 + }, + { + "epoch": 0.28913420596120487, + "grad_norm": 2.6952126026153564, + "learning_rate": 1.66812790128532e-05, + "loss": 0.5177, + "step": 9167 + }, + { + "epoch": 0.2891657467276455, + "grad_norm": 1.4913678169250488, + "learning_rate": 1.668051889252576e-05, + "loss": 0.5037, + "step": 9168 + }, + { + "epoch": 0.28919728749408613, + "grad_norm": 1.4144481420516968, + "learning_rate": 1.6679758702481917e-05, + "loss": 0.5541, + "step": 9169 + }, + { + "epoch": 0.2892288282605267, + "grad_norm": 2.0435237884521484, + "learning_rate": 1.6678998442729605e-05, + "loss": 0.5321, + "step": 9170 + }, + { + "epoch": 0.28926036902696733, + "grad_norm": 1.811039924621582, + "learning_rate": 1.6678238113276758e-05, + "loss": 0.5114, + "step": 9171 + }, + { + "epoch": 0.28929190979340796, + "grad_norm": 1.7621638774871826, + "learning_rate": 1.667747771413131e-05, + "loss": 0.8679, + "step": 9172 + }, + { + "epoch": 0.2893234505598486, + "grad_norm": 1.7720122337341309, + "learning_rate": 1.6676717245301198e-05, + "loss": 0.5138, + "step": 9173 + }, + { + "epoch": 0.2893549913262892, + "grad_norm": 1.8127696514129639, + "learning_rate": 1.6675956706794358e-05, + "loss": 0.5402, + "step": 9174 + }, + { + "epoch": 0.28938653209272985, + "grad_norm": 2.657942533493042, + "learning_rate": 1.6675196098618726e-05, + "loss": 0.5535, + "step": 9175 + }, + { + "epoch": 0.2894180728591705, + "grad_norm": 1.8066555261611938, + "learning_rate": 1.6674435420782238e-05, + "loss": 0.896, + "step": 9176 + }, + { + "epoch": 0.2894496136256111, + "grad_norm": 1.5411392450332642, + "learning_rate": 1.6673674673292837e-05, + "loss": 0.5505, + "step": 9177 + }, + { + "epoch": 0.28948115439205174, + "grad_norm": 1.870774507522583, + "learning_rate": 1.6672913856158453e-05, + "loss": 0.5871, + "step": 9178 + }, + { + "epoch": 0.2895126951584924, + "grad_norm": 1.7665166854858398, + "learning_rate": 1.6672152969387037e-05, + "loss": 0.5402, + "step": 9179 + }, + { + "epoch": 0.289544235924933, + "grad_norm": 1.4077037572860718, + "learning_rate": 1.6671392012986522e-05, + "loss": 0.5379, + "step": 9180 + }, + { + "epoch": 0.2895757766913736, + "grad_norm": 2.247255563735962, + "learning_rate": 1.667063098696485e-05, + "loss": 0.9052, + "step": 9181 + }, + { + "epoch": 0.2896073174578142, + "grad_norm": 1.5929399728775024, + "learning_rate": 1.6669869891329965e-05, + "loss": 0.5613, + "step": 9182 + }, + { + "epoch": 0.28963885822425484, + "grad_norm": 1.2928608655929565, + "learning_rate": 1.6669108726089813e-05, + "loss": 0.537, + "step": 9183 + }, + { + "epoch": 0.28967039899069547, + "grad_norm": 1.7430952787399292, + "learning_rate": 1.666834749125233e-05, + "loss": 0.5226, + "step": 9184 + }, + { + "epoch": 0.2897019397571361, + "grad_norm": 1.5087367296218872, + "learning_rate": 1.666758618682546e-05, + "loss": 0.4739, + "step": 9185 + }, + { + "epoch": 0.2897334805235767, + "grad_norm": 2.247727870941162, + "learning_rate": 1.6666824812817153e-05, + "loss": 0.5642, + "step": 9186 + }, + { + "epoch": 0.28976502129001735, + "grad_norm": 2.6429953575134277, + "learning_rate": 1.666606336923535e-05, + "loss": 0.4514, + "step": 9187 + }, + { + "epoch": 0.289796562056458, + "grad_norm": 3.1364572048187256, + "learning_rate": 1.6665301856088004e-05, + "loss": 0.5606, + "step": 9188 + }, + { + "epoch": 0.2898281028228986, + "grad_norm": 1.8569904565811157, + "learning_rate": 1.6664540273383054e-05, + "loss": 0.544, + "step": 9189 + }, + { + "epoch": 0.28985964358933924, + "grad_norm": 1.721754550933838, + "learning_rate": 1.6663778621128455e-05, + "loss": 0.5516, + "step": 9190 + }, + { + "epoch": 0.2898911843557798, + "grad_norm": 1.6408116817474365, + "learning_rate": 1.666301689933215e-05, + "loss": 0.5658, + "step": 9191 + }, + { + "epoch": 0.28992272512222045, + "grad_norm": 1.9472041130065918, + "learning_rate": 1.6662255108002086e-05, + "loss": 0.5184, + "step": 9192 + }, + { + "epoch": 0.2899542658886611, + "grad_norm": 1.5337345600128174, + "learning_rate": 1.6661493247146224e-05, + "loss": 0.3959, + "step": 9193 + }, + { + "epoch": 0.2899858066551017, + "grad_norm": 2.189401626586914, + "learning_rate": 1.6660731316772503e-05, + "loss": 0.8705, + "step": 9194 + }, + { + "epoch": 0.29001734742154234, + "grad_norm": 1.6884691715240479, + "learning_rate": 1.665996931688888e-05, + "loss": 0.5503, + "step": 9195 + }, + { + "epoch": 0.29004888818798297, + "grad_norm": 1.476967453956604, + "learning_rate": 1.6659207247503306e-05, + "loss": 0.5527, + "step": 9196 + }, + { + "epoch": 0.2900804289544236, + "grad_norm": 1.7336032390594482, + "learning_rate": 1.665844510862373e-05, + "loss": 0.4082, + "step": 9197 + }, + { + "epoch": 0.2901119697208642, + "grad_norm": 1.758501410484314, + "learning_rate": 1.6657682900258115e-05, + "loss": 0.4943, + "step": 9198 + }, + { + "epoch": 0.29014351048730486, + "grad_norm": 1.5420490503311157, + "learning_rate": 1.6656920622414407e-05, + "loss": 0.5395, + "step": 9199 + }, + { + "epoch": 0.2901750512537455, + "grad_norm": 1.5220426321029663, + "learning_rate": 1.6656158275100563e-05, + "loss": 0.5309, + "step": 9200 + }, + { + "epoch": 0.2902065920201861, + "grad_norm": 1.529678463935852, + "learning_rate": 1.665539585832454e-05, + "loss": 0.5642, + "step": 9201 + }, + { + "epoch": 0.2902381327866267, + "grad_norm": 1.6320806741714478, + "learning_rate": 1.6654633372094294e-05, + "loss": 0.5826, + "step": 9202 + }, + { + "epoch": 0.2902696735530673, + "grad_norm": 1.3494168519973755, + "learning_rate": 1.665387081641778e-05, + "loss": 0.5235, + "step": 9203 + }, + { + "epoch": 0.29030121431950795, + "grad_norm": 1.7101168632507324, + "learning_rate": 1.6653108191302957e-05, + "loss": 0.8741, + "step": 9204 + }, + { + "epoch": 0.2903327550859486, + "grad_norm": 1.5512802600860596, + "learning_rate": 1.6652345496757782e-05, + "loss": 0.5367, + "step": 9205 + }, + { + "epoch": 0.2903642958523892, + "grad_norm": 1.5031119585037231, + "learning_rate": 1.665158273279022e-05, + "loss": 0.5146, + "step": 9206 + }, + { + "epoch": 0.29039583661882984, + "grad_norm": 1.8462742567062378, + "learning_rate": 1.665081989940823e-05, + "loss": 0.528, + "step": 9207 + }, + { + "epoch": 0.29042737738527047, + "grad_norm": 1.961246371269226, + "learning_rate": 1.6650056996619763e-05, + "loss": 0.884, + "step": 9208 + }, + { + "epoch": 0.2904589181517111, + "grad_norm": 1.527402400970459, + "learning_rate": 1.664929402443279e-05, + "loss": 0.5503, + "step": 9209 + }, + { + "epoch": 0.29049045891815173, + "grad_norm": 2.310675859451294, + "learning_rate": 1.6648530982855274e-05, + "loss": 0.5675, + "step": 9210 + }, + { + "epoch": 0.29052199968459236, + "grad_norm": 1.5658544301986694, + "learning_rate": 1.6647767871895173e-05, + "loss": 0.5585, + "step": 9211 + }, + { + "epoch": 0.290553540451033, + "grad_norm": 1.698067307472229, + "learning_rate": 1.6647004691560452e-05, + "loss": 0.5569, + "step": 9212 + }, + { + "epoch": 0.29058508121747356, + "grad_norm": 1.4430956840515137, + "learning_rate": 1.6646241441859076e-05, + "loss": 0.5674, + "step": 9213 + }, + { + "epoch": 0.2906166219839142, + "grad_norm": 1.60844087600708, + "learning_rate": 1.664547812279901e-05, + "loss": 0.5584, + "step": 9214 + }, + { + "epoch": 0.2906481627503548, + "grad_norm": 1.3018587827682495, + "learning_rate": 1.664471473438822e-05, + "loss": 0.5461, + "step": 9215 + }, + { + "epoch": 0.29067970351679545, + "grad_norm": 2.8200020790100098, + "learning_rate": 1.664395127663467e-05, + "loss": 0.536, + "step": 9216 + }, + { + "epoch": 0.2907112442832361, + "grad_norm": 1.5542837381362915, + "learning_rate": 1.664318774954633e-05, + "loss": 0.5638, + "step": 9217 + }, + { + "epoch": 0.2907427850496767, + "grad_norm": 1.6195447444915771, + "learning_rate": 1.664242415313117e-05, + "loss": 0.3913, + "step": 9218 + }, + { + "epoch": 0.29077432581611734, + "grad_norm": 2.5424234867095947, + "learning_rate": 1.6641660487397155e-05, + "loss": 0.5375, + "step": 9219 + }, + { + "epoch": 0.29080586658255797, + "grad_norm": 2.209378480911255, + "learning_rate": 1.6640896752352256e-05, + "loss": 0.5134, + "step": 9220 + }, + { + "epoch": 0.2908374073489986, + "grad_norm": 6.857260227203369, + "learning_rate": 1.6640132948004442e-05, + "loss": 0.5411, + "step": 9221 + }, + { + "epoch": 0.29086894811543923, + "grad_norm": 2.0607125759124756, + "learning_rate": 1.6639369074361687e-05, + "loss": 0.8447, + "step": 9222 + }, + { + "epoch": 0.2909004888818798, + "grad_norm": 2.1078383922576904, + "learning_rate": 1.663860513143196e-05, + "loss": 0.5359, + "step": 9223 + }, + { + "epoch": 0.29093202964832043, + "grad_norm": 4.741464614868164, + "learning_rate": 1.663784111922323e-05, + "loss": 0.5404, + "step": 9224 + }, + { + "epoch": 0.29096357041476106, + "grad_norm": 1.598520278930664, + "learning_rate": 1.6637077037743477e-05, + "loss": 0.5518, + "step": 9225 + }, + { + "epoch": 0.2909951111812017, + "grad_norm": 3.2032036781311035, + "learning_rate": 1.663631288700067e-05, + "loss": 0.4431, + "step": 9226 + }, + { + "epoch": 0.2910266519476423, + "grad_norm": 1.8698244094848633, + "learning_rate": 1.6635548667002784e-05, + "loss": 0.555, + "step": 9227 + }, + { + "epoch": 0.29105819271408295, + "grad_norm": 1.7807306051254272, + "learning_rate": 1.6634784377757797e-05, + "loss": 0.5386, + "step": 9228 + }, + { + "epoch": 0.2910897334805236, + "grad_norm": 1.8601042032241821, + "learning_rate": 1.6634020019273688e-05, + "loss": 0.543, + "step": 9229 + }, + { + "epoch": 0.2911212742469642, + "grad_norm": 1.5603877305984497, + "learning_rate": 1.663325559155842e-05, + "loss": 0.5173, + "step": 9230 + }, + { + "epoch": 0.29115281501340484, + "grad_norm": 1.7581168413162231, + "learning_rate": 1.6632491094619986e-05, + "loss": 0.5383, + "step": 9231 + }, + { + "epoch": 0.29118435577984547, + "grad_norm": 1.4873120784759521, + "learning_rate": 1.6631726528466353e-05, + "loss": 0.5375, + "step": 9232 + }, + { + "epoch": 0.2912158965462861, + "grad_norm": 2.0449564456939697, + "learning_rate": 1.663096189310551e-05, + "loss": 0.54, + "step": 9233 + }, + { + "epoch": 0.2912474373127267, + "grad_norm": 2.0352282524108887, + "learning_rate": 1.6630197188545425e-05, + "loss": 0.5133, + "step": 9234 + }, + { + "epoch": 0.2912789780791673, + "grad_norm": 2.6163158416748047, + "learning_rate": 1.6629432414794085e-05, + "loss": 0.5453, + "step": 9235 + }, + { + "epoch": 0.29131051884560794, + "grad_norm": 2.284289836883545, + "learning_rate": 1.662866757185947e-05, + "loss": 0.5167, + "step": 9236 + }, + { + "epoch": 0.29134205961204857, + "grad_norm": 3.258622407913208, + "learning_rate": 1.6627902659749567e-05, + "loss": 0.514, + "step": 9237 + }, + { + "epoch": 0.2913736003784892, + "grad_norm": 1.9197133779525757, + "learning_rate": 1.662713767847235e-05, + "loss": 0.5458, + "step": 9238 + }, + { + "epoch": 0.2914051411449298, + "grad_norm": 1.9603288173675537, + "learning_rate": 1.66263726280358e-05, + "loss": 0.5587, + "step": 9239 + }, + { + "epoch": 0.29143668191137045, + "grad_norm": 1.8292304277420044, + "learning_rate": 1.6625607508447913e-05, + "loss": 0.551, + "step": 9240 + }, + { + "epoch": 0.2914682226778111, + "grad_norm": 1.5623304843902588, + "learning_rate": 1.6624842319716667e-05, + "loss": 0.5361, + "step": 9241 + }, + { + "epoch": 0.2914997634442517, + "grad_norm": 1.539121150970459, + "learning_rate": 1.6624077061850047e-05, + "loss": 0.5315, + "step": 9242 + }, + { + "epoch": 0.29153130421069234, + "grad_norm": 1.424766182899475, + "learning_rate": 1.6623311734856035e-05, + "loss": 0.5261, + "step": 9243 + }, + { + "epoch": 0.2915628449771329, + "grad_norm": 1.6963618993759155, + "learning_rate": 1.6622546338742626e-05, + "loss": 0.8998, + "step": 9244 + }, + { + "epoch": 0.29159438574357355, + "grad_norm": 2.204437255859375, + "learning_rate": 1.6621780873517803e-05, + "loss": 0.5448, + "step": 9245 + }, + { + "epoch": 0.2916259265100142, + "grad_norm": 2.2134411334991455, + "learning_rate": 1.6621015339189556e-05, + "loss": 0.5591, + "step": 9246 + }, + { + "epoch": 0.2916574672764548, + "grad_norm": 1.8087875843048096, + "learning_rate": 1.662024973576587e-05, + "loss": 0.5323, + "step": 9247 + }, + { + "epoch": 0.29168900804289544, + "grad_norm": 2.433875560760498, + "learning_rate": 1.6619484063254742e-05, + "loss": 0.8872, + "step": 9248 + }, + { + "epoch": 0.29172054880933607, + "grad_norm": 1.6492445468902588, + "learning_rate": 1.6618718321664153e-05, + "loss": 0.5468, + "step": 9249 + }, + { + "epoch": 0.2917520895757767, + "grad_norm": 2.0857529640197754, + "learning_rate": 1.6617952511002103e-05, + "loss": 0.5428, + "step": 9250 + }, + { + "epoch": 0.2917836303422173, + "grad_norm": 2.052320718765259, + "learning_rate": 1.6617186631276575e-05, + "loss": 0.5628, + "step": 9251 + }, + { + "epoch": 0.29181517110865796, + "grad_norm": 1.865122675895691, + "learning_rate": 1.6616420682495575e-05, + "loss": 0.5576, + "step": 9252 + }, + { + "epoch": 0.2918467118750986, + "grad_norm": 1.5515000820159912, + "learning_rate": 1.6615654664667077e-05, + "loss": 0.5503, + "step": 9253 + }, + { + "epoch": 0.2918782526415392, + "grad_norm": 1.9929407835006714, + "learning_rate": 1.6614888577799093e-05, + "loss": 0.5559, + "step": 9254 + }, + { + "epoch": 0.2919097934079798, + "grad_norm": 1.9087097644805908, + "learning_rate": 1.6614122421899607e-05, + "loss": 0.5674, + "step": 9255 + }, + { + "epoch": 0.2919413341744204, + "grad_norm": 2.8619377613067627, + "learning_rate": 1.661335619697662e-05, + "loss": 0.5607, + "step": 9256 + }, + { + "epoch": 0.29197287494086105, + "grad_norm": 1.343275785446167, + "learning_rate": 1.6612589903038123e-05, + "loss": 0.5293, + "step": 9257 + }, + { + "epoch": 0.2920044157073017, + "grad_norm": 1.442591905593872, + "learning_rate": 1.661182354009212e-05, + "loss": 0.5155, + "step": 9258 + }, + { + "epoch": 0.2920359564737423, + "grad_norm": 1.4810072183609009, + "learning_rate": 1.66110571081466e-05, + "loss": 0.5354, + "step": 9259 + }, + { + "epoch": 0.29206749724018294, + "grad_norm": 1.4697213172912598, + "learning_rate": 1.661029060720957e-05, + "loss": 0.5293, + "step": 9260 + }, + { + "epoch": 0.29209903800662357, + "grad_norm": 1.567935585975647, + "learning_rate": 1.660952403728902e-05, + "loss": 0.5109, + "step": 9261 + }, + { + "epoch": 0.2921305787730642, + "grad_norm": 1.581913709640503, + "learning_rate": 1.660875739839296e-05, + "loss": 0.5289, + "step": 9262 + }, + { + "epoch": 0.2921621195395048, + "grad_norm": 1.7458072900772095, + "learning_rate": 1.660799069052938e-05, + "loss": 0.5178, + "step": 9263 + }, + { + "epoch": 0.29219366030594546, + "grad_norm": 1.34100341796875, + "learning_rate": 1.6607223913706287e-05, + "loss": 0.525, + "step": 9264 + }, + { + "epoch": 0.29222520107238603, + "grad_norm": 2.0704309940338135, + "learning_rate": 1.6606457067931683e-05, + "loss": 0.8933, + "step": 9265 + }, + { + "epoch": 0.29225674183882666, + "grad_norm": 1.6803516149520874, + "learning_rate": 1.660569015321357e-05, + "loss": 0.5588, + "step": 9266 + }, + { + "epoch": 0.2922882826052673, + "grad_norm": 2.5321974754333496, + "learning_rate": 1.6604923169559954e-05, + "loss": 0.3745, + "step": 9267 + }, + { + "epoch": 0.2923198233717079, + "grad_norm": 1.646507978439331, + "learning_rate": 1.660415611697883e-05, + "loss": 0.5295, + "step": 9268 + }, + { + "epoch": 0.29235136413814855, + "grad_norm": 1.6198225021362305, + "learning_rate": 1.6603388995478214e-05, + "loss": 0.5351, + "step": 9269 + }, + { + "epoch": 0.2923829049045892, + "grad_norm": 1.5580443143844604, + "learning_rate": 1.6602621805066104e-05, + "loss": 0.5598, + "step": 9270 + }, + { + "epoch": 0.2924144456710298, + "grad_norm": 1.5014727115631104, + "learning_rate": 1.660185454575051e-05, + "loss": 0.5607, + "step": 9271 + }, + { + "epoch": 0.29244598643747044, + "grad_norm": 1.6769187450408936, + "learning_rate": 1.6601087217539437e-05, + "loss": 0.526, + "step": 9272 + }, + { + "epoch": 0.29247752720391107, + "grad_norm": 1.582152009010315, + "learning_rate": 1.6600319820440893e-05, + "loss": 0.5639, + "step": 9273 + }, + { + "epoch": 0.2925090679703517, + "grad_norm": 1.5759360790252686, + "learning_rate": 1.6599552354462883e-05, + "loss": 0.552, + "step": 9274 + }, + { + "epoch": 0.29254060873679233, + "grad_norm": 1.5051162242889404, + "learning_rate": 1.6598784819613424e-05, + "loss": 0.5035, + "step": 9275 + }, + { + "epoch": 0.2925721495032329, + "grad_norm": 2.0924811363220215, + "learning_rate": 1.6598017215900525e-05, + "loss": 0.8875, + "step": 9276 + }, + { + "epoch": 0.29260369026967353, + "grad_norm": 1.7725051641464233, + "learning_rate": 1.6597249543332186e-05, + "loss": 0.5218, + "step": 9277 + }, + { + "epoch": 0.29263523103611416, + "grad_norm": 1.920898199081421, + "learning_rate": 1.659648180191643e-05, + "loss": 0.5704, + "step": 9278 + }, + { + "epoch": 0.2926667718025548, + "grad_norm": 1.6775885820388794, + "learning_rate": 1.659571399166126e-05, + "loss": 0.5717, + "step": 9279 + }, + { + "epoch": 0.2926983125689954, + "grad_norm": 1.9193185567855835, + "learning_rate": 1.65949461125747e-05, + "loss": 0.5104, + "step": 9280 + }, + { + "epoch": 0.29272985333543605, + "grad_norm": 2.0135700702667236, + "learning_rate": 1.659417816466475e-05, + "loss": 0.5473, + "step": 9281 + }, + { + "epoch": 0.2927613941018767, + "grad_norm": 1.6403154134750366, + "learning_rate": 1.6593410147939434e-05, + "loss": 0.5158, + "step": 9282 + }, + { + "epoch": 0.2927929348683173, + "grad_norm": 2.293503761291504, + "learning_rate": 1.6592642062406764e-05, + "loss": 0.5435, + "step": 9283 + }, + { + "epoch": 0.29282447563475794, + "grad_norm": 1.4130278825759888, + "learning_rate": 1.659187390807475e-05, + "loss": 0.4305, + "step": 9284 + }, + { + "epoch": 0.29285601640119857, + "grad_norm": 2.5100340843200684, + "learning_rate": 1.659110568495142e-05, + "loss": 0.5433, + "step": 9285 + }, + { + "epoch": 0.2928875571676392, + "grad_norm": 1.8430607318878174, + "learning_rate": 1.6590337393044782e-05, + "loss": 0.3884, + "step": 9286 + }, + { + "epoch": 0.2929190979340798, + "grad_norm": 1.9860316514968872, + "learning_rate": 1.6589569032362852e-05, + "loss": 0.8409, + "step": 9287 + }, + { + "epoch": 0.2929506387005204, + "grad_norm": 2.2264151573181152, + "learning_rate": 1.6588800602913657e-05, + "loss": 0.5717, + "step": 9288 + }, + { + "epoch": 0.29298217946696103, + "grad_norm": 2.65512752532959, + "learning_rate": 1.658803210470521e-05, + "loss": 0.4036, + "step": 9289 + }, + { + "epoch": 0.29301372023340166, + "grad_norm": 1.6207010746002197, + "learning_rate": 1.6587263537745536e-05, + "loss": 0.383, + "step": 9290 + }, + { + "epoch": 0.2930452609998423, + "grad_norm": 2.4588029384613037, + "learning_rate": 1.6586494902042646e-05, + "loss": 0.5117, + "step": 9291 + }, + { + "epoch": 0.2930768017662829, + "grad_norm": 1.8540375232696533, + "learning_rate": 1.658572619760457e-05, + "loss": 0.5457, + "step": 9292 + }, + { + "epoch": 0.29310834253272355, + "grad_norm": 2.612896680831909, + "learning_rate": 1.658495742443933e-05, + "loss": 0.5426, + "step": 9293 + }, + { + "epoch": 0.2931398832991642, + "grad_norm": 1.6688597202301025, + "learning_rate": 1.6584188582554943e-05, + "loss": 0.5556, + "step": 9294 + }, + { + "epoch": 0.2931714240656048, + "grad_norm": 1.7098733186721802, + "learning_rate": 1.6583419671959436e-05, + "loss": 0.5392, + "step": 9295 + }, + { + "epoch": 0.29320296483204544, + "grad_norm": 2.1328930854797363, + "learning_rate": 1.6582650692660833e-05, + "loss": 0.5406, + "step": 9296 + }, + { + "epoch": 0.293234505598486, + "grad_norm": 2.201063394546509, + "learning_rate": 1.658188164466716e-05, + "loss": 0.5491, + "step": 9297 + }, + { + "epoch": 0.29326604636492665, + "grad_norm": 1.5660442113876343, + "learning_rate": 1.658111252798644e-05, + "loss": 0.5654, + "step": 9298 + }, + { + "epoch": 0.2932975871313673, + "grad_norm": 1.7276636362075806, + "learning_rate": 1.6580343342626703e-05, + "loss": 0.5203, + "step": 9299 + }, + { + "epoch": 0.2933291278978079, + "grad_norm": 1.7276811599731445, + "learning_rate": 1.6579574088595972e-05, + "loss": 0.502, + "step": 9300 + }, + { + "epoch": 0.29336066866424854, + "grad_norm": 2.034869432449341, + "learning_rate": 1.6578804765902274e-05, + "loss": 0.5146, + "step": 9301 + }, + { + "epoch": 0.29339220943068917, + "grad_norm": 1.5348443984985352, + "learning_rate": 1.657803537455364e-05, + "loss": 0.5256, + "step": 9302 + }, + { + "epoch": 0.2934237501971298, + "grad_norm": 1.621331810951233, + "learning_rate": 1.6577265914558104e-05, + "loss": 0.5232, + "step": 9303 + }, + { + "epoch": 0.2934552909635704, + "grad_norm": 1.949758529663086, + "learning_rate": 1.657649638592369e-05, + "loss": 0.5447, + "step": 9304 + }, + { + "epoch": 0.29348683173001106, + "grad_norm": 2.0686066150665283, + "learning_rate": 1.6575726788658427e-05, + "loss": 0.5746, + "step": 9305 + }, + { + "epoch": 0.2935183724964517, + "grad_norm": 2.8170907497406006, + "learning_rate": 1.657495712277035e-05, + "loss": 0.5439, + "step": 9306 + }, + { + "epoch": 0.2935499132628923, + "grad_norm": 2.003526210784912, + "learning_rate": 1.657418738826749e-05, + "loss": 0.5409, + "step": 9307 + }, + { + "epoch": 0.2935814540293329, + "grad_norm": 1.6320284605026245, + "learning_rate": 1.657341758515788e-05, + "loss": 0.586, + "step": 9308 + }, + { + "epoch": 0.2936129947957735, + "grad_norm": 3.4390385150909424, + "learning_rate": 1.6572647713449556e-05, + "loss": 0.5126, + "step": 9309 + }, + { + "epoch": 0.29364453556221415, + "grad_norm": 3.2266597747802734, + "learning_rate": 1.6571877773150546e-05, + "loss": 0.5372, + "step": 9310 + }, + { + "epoch": 0.2936760763286548, + "grad_norm": 10.336581230163574, + "learning_rate": 1.657110776426889e-05, + "loss": 0.5557, + "step": 9311 + }, + { + "epoch": 0.2937076170950954, + "grad_norm": 1.7886786460876465, + "learning_rate": 1.657033768681262e-05, + "loss": 0.5344, + "step": 9312 + }, + { + "epoch": 0.29373915786153604, + "grad_norm": 1.5374690294265747, + "learning_rate": 1.656956754078978e-05, + "loss": 0.5569, + "step": 9313 + }, + { + "epoch": 0.29377069862797667, + "grad_norm": 2.653400182723999, + "learning_rate": 1.65687973262084e-05, + "loss": 0.5501, + "step": 9314 + }, + { + "epoch": 0.2938022393944173, + "grad_norm": 1.8467375040054321, + "learning_rate": 1.6568027043076513e-05, + "loss": 0.5336, + "step": 9315 + }, + { + "epoch": 0.2938337801608579, + "grad_norm": 3.685703992843628, + "learning_rate": 1.656725669140217e-05, + "loss": 0.4592, + "step": 9316 + }, + { + "epoch": 0.29386532092729856, + "grad_norm": 1.9752800464630127, + "learning_rate": 1.6566486271193403e-05, + "loss": 0.5314, + "step": 9317 + }, + { + "epoch": 0.29389686169373913, + "grad_norm": 2.916969060897827, + "learning_rate": 1.6565715782458253e-05, + "loss": 0.7995, + "step": 9318 + }, + { + "epoch": 0.29392840246017976, + "grad_norm": 1.765961766242981, + "learning_rate": 1.6564945225204757e-05, + "loss": 0.5552, + "step": 9319 + }, + { + "epoch": 0.2939599432266204, + "grad_norm": 1.4810378551483154, + "learning_rate": 1.6564174599440967e-05, + "loss": 0.5667, + "step": 9320 + }, + { + "epoch": 0.293991483993061, + "grad_norm": 1.9605793952941895, + "learning_rate": 1.656340390517491e-05, + "loss": 0.548, + "step": 9321 + }, + { + "epoch": 0.29402302475950165, + "grad_norm": 2.47269868850708, + "learning_rate": 1.656263314241464e-05, + "loss": 0.5387, + "step": 9322 + }, + { + "epoch": 0.2940545655259423, + "grad_norm": 1.767849326133728, + "learning_rate": 1.65618623111682e-05, + "loss": 0.566, + "step": 9323 + }, + { + "epoch": 0.2940861062923829, + "grad_norm": 1.4612473249435425, + "learning_rate": 1.6561091411443634e-05, + "loss": 0.559, + "step": 9324 + }, + { + "epoch": 0.29411764705882354, + "grad_norm": 1.6764169931411743, + "learning_rate": 1.656032044324898e-05, + "loss": 0.5391, + "step": 9325 + }, + { + "epoch": 0.29414918782526417, + "grad_norm": 1.2755621671676636, + "learning_rate": 1.6559549406592285e-05, + "loss": 0.5278, + "step": 9326 + }, + { + "epoch": 0.2941807285917048, + "grad_norm": 1.4771044254302979, + "learning_rate": 1.6558778301481602e-05, + "loss": 0.546, + "step": 9327 + }, + { + "epoch": 0.29421226935814543, + "grad_norm": 2.9030141830444336, + "learning_rate": 1.655800712792498e-05, + "loss": 0.5459, + "step": 9328 + }, + { + "epoch": 0.294243810124586, + "grad_norm": 1.894031286239624, + "learning_rate": 1.6557235885930452e-05, + "loss": 0.5108, + "step": 9329 + }, + { + "epoch": 0.29427535089102663, + "grad_norm": 2.0260705947875977, + "learning_rate": 1.655646457550608e-05, + "loss": 0.8945, + "step": 9330 + }, + { + "epoch": 0.29430689165746726, + "grad_norm": 1.7615718841552734, + "learning_rate": 1.655569319665991e-05, + "loss": 0.5471, + "step": 9331 + }, + { + "epoch": 0.2943384324239079, + "grad_norm": 1.7306153774261475, + "learning_rate": 1.6554921749399993e-05, + "loss": 0.5324, + "step": 9332 + }, + { + "epoch": 0.2943699731903485, + "grad_norm": 1.7988260984420776, + "learning_rate": 1.6554150233734373e-05, + "loss": 0.519, + "step": 9333 + }, + { + "epoch": 0.29440151395678915, + "grad_norm": 1.4721624851226807, + "learning_rate": 1.6553378649671112e-05, + "loss": 0.5305, + "step": 9334 + }, + { + "epoch": 0.2944330547232298, + "grad_norm": 1.5940099954605103, + "learning_rate": 1.6552606997218253e-05, + "loss": 0.549, + "step": 9335 + }, + { + "epoch": 0.2944645954896704, + "grad_norm": 1.4209729433059692, + "learning_rate": 1.655183527638385e-05, + "loss": 0.4915, + "step": 9336 + }, + { + "epoch": 0.29449613625611104, + "grad_norm": 1.4549925327301025, + "learning_rate": 1.6551063487175963e-05, + "loss": 0.5362, + "step": 9337 + }, + { + "epoch": 0.29452767702255167, + "grad_norm": 1.695330262184143, + "learning_rate": 1.6550291629602637e-05, + "loss": 0.5514, + "step": 9338 + }, + { + "epoch": 0.2945592177889923, + "grad_norm": 1.7037317752838135, + "learning_rate": 1.6549519703671933e-05, + "loss": 0.5364, + "step": 9339 + }, + { + "epoch": 0.2945907585554329, + "grad_norm": 1.7607386112213135, + "learning_rate": 1.6548747709391908e-05, + "loss": 0.5009, + "step": 9340 + }, + { + "epoch": 0.2946222993218735, + "grad_norm": 1.8622827529907227, + "learning_rate": 1.6547975646770614e-05, + "loss": 0.5463, + "step": 9341 + }, + { + "epoch": 0.29465384008831413, + "grad_norm": 1.5807814598083496, + "learning_rate": 1.6547203515816107e-05, + "loss": 0.5308, + "step": 9342 + }, + { + "epoch": 0.29468538085475476, + "grad_norm": 1.9238685369491577, + "learning_rate": 1.6546431316536452e-05, + "loss": 0.5647, + "step": 9343 + }, + { + "epoch": 0.2947169216211954, + "grad_norm": 1.5702111721038818, + "learning_rate": 1.65456590489397e-05, + "loss": 0.5451, + "step": 9344 + }, + { + "epoch": 0.294748462387636, + "grad_norm": 1.3391573429107666, + "learning_rate": 1.6544886713033917e-05, + "loss": 0.5286, + "step": 9345 + }, + { + "epoch": 0.29478000315407665, + "grad_norm": 1.2602428197860718, + "learning_rate": 1.6544114308827153e-05, + "loss": 0.5165, + "step": 9346 + }, + { + "epoch": 0.2948115439205173, + "grad_norm": 1.7318589687347412, + "learning_rate": 1.654334183632748e-05, + "loss": 0.568, + "step": 9347 + }, + { + "epoch": 0.2948430846869579, + "grad_norm": 1.5280141830444336, + "learning_rate": 1.6542569295542953e-05, + "loss": 0.5246, + "step": 9348 + }, + { + "epoch": 0.29487462545339854, + "grad_norm": 2.08170485496521, + "learning_rate": 1.6541796686481637e-05, + "loss": 0.5403, + "step": 9349 + }, + { + "epoch": 0.2949061662198391, + "grad_norm": 1.9519144296646118, + "learning_rate": 1.654102400915159e-05, + "loss": 0.5129, + "step": 9350 + }, + { + "epoch": 0.29493770698627975, + "grad_norm": 1.5499210357666016, + "learning_rate": 1.654025126356088e-05, + "loss": 0.3906, + "step": 9351 + }, + { + "epoch": 0.2949692477527204, + "grad_norm": 2.3011972904205322, + "learning_rate": 1.6539478449717567e-05, + "loss": 0.525, + "step": 9352 + }, + { + "epoch": 0.295000788519161, + "grad_norm": 1.6074097156524658, + "learning_rate": 1.6538705567629723e-05, + "loss": 0.5366, + "step": 9353 + }, + { + "epoch": 0.29503232928560164, + "grad_norm": 1.903331995010376, + "learning_rate": 1.65379326173054e-05, + "loss": 0.4894, + "step": 9354 + }, + { + "epoch": 0.29506387005204227, + "grad_norm": 2.1079885959625244, + "learning_rate": 1.6537159598752685e-05, + "loss": 0.5607, + "step": 9355 + }, + { + "epoch": 0.2950954108184829, + "grad_norm": 1.8393970727920532, + "learning_rate": 1.653638651197963e-05, + "loss": 0.5444, + "step": 9356 + }, + { + "epoch": 0.2951269515849235, + "grad_norm": 1.3578829765319824, + "learning_rate": 1.6535613356994304e-05, + "loss": 0.4404, + "step": 9357 + }, + { + "epoch": 0.29515849235136415, + "grad_norm": 1.6771942377090454, + "learning_rate": 1.653484013380478e-05, + "loss": 0.5409, + "step": 9358 + }, + { + "epoch": 0.2951900331178048, + "grad_norm": 1.34040367603302, + "learning_rate": 1.6534066842419127e-05, + "loss": 0.5359, + "step": 9359 + }, + { + "epoch": 0.2952215738842454, + "grad_norm": 1.9470527172088623, + "learning_rate": 1.6533293482845406e-05, + "loss": 0.87, + "step": 9360 + }, + { + "epoch": 0.295253114650686, + "grad_norm": 1.7488447427749634, + "learning_rate": 1.65325200550917e-05, + "loss": 0.5464, + "step": 9361 + }, + { + "epoch": 0.2952846554171266, + "grad_norm": 1.6178534030914307, + "learning_rate": 1.653174655916607e-05, + "loss": 0.5508, + "step": 9362 + }, + { + "epoch": 0.29531619618356725, + "grad_norm": 1.4640804529190063, + "learning_rate": 1.6530972995076598e-05, + "loss": 0.3792, + "step": 9363 + }, + { + "epoch": 0.2953477369500079, + "grad_norm": 1.350914716720581, + "learning_rate": 1.653019936283135e-05, + "loss": 0.5176, + "step": 9364 + }, + { + "epoch": 0.2953792777164485, + "grad_norm": 1.568254828453064, + "learning_rate": 1.6529425662438406e-05, + "loss": 0.5447, + "step": 9365 + }, + { + "epoch": 0.29541081848288914, + "grad_norm": 1.3277239799499512, + "learning_rate": 1.652865189390583e-05, + "loss": 0.5214, + "step": 9366 + }, + { + "epoch": 0.29544235924932977, + "grad_norm": 1.536524772644043, + "learning_rate": 1.6527878057241702e-05, + "loss": 0.5345, + "step": 9367 + }, + { + "epoch": 0.2954739000157704, + "grad_norm": 5.668264865875244, + "learning_rate": 1.6527104152454096e-05, + "loss": 0.533, + "step": 9368 + }, + { + "epoch": 0.295505440782211, + "grad_norm": 1.6122568845748901, + "learning_rate": 1.6526330179551093e-05, + "loss": 0.8268, + "step": 9369 + }, + { + "epoch": 0.29553698154865166, + "grad_norm": 2.1982686519622803, + "learning_rate": 1.6525556138540766e-05, + "loss": 0.3816, + "step": 9370 + }, + { + "epoch": 0.29556852231509223, + "grad_norm": 1.9071801900863647, + "learning_rate": 1.6524782029431195e-05, + "loss": 0.843, + "step": 9371 + }, + { + "epoch": 0.29560006308153286, + "grad_norm": 1.2143206596374512, + "learning_rate": 1.6524007852230455e-05, + "loss": 0.5616, + "step": 9372 + }, + { + "epoch": 0.2956316038479735, + "grad_norm": 2.43302059173584, + "learning_rate": 1.652323360694663e-05, + "loss": 0.5553, + "step": 9373 + }, + { + "epoch": 0.2956631446144141, + "grad_norm": 1.5875983238220215, + "learning_rate": 1.6522459293587797e-05, + "loss": 0.5608, + "step": 9374 + }, + { + "epoch": 0.29569468538085475, + "grad_norm": 1.5100467205047607, + "learning_rate": 1.6521684912162033e-05, + "loss": 0.5112, + "step": 9375 + }, + { + "epoch": 0.2957262261472954, + "grad_norm": 1.5097568035125732, + "learning_rate": 1.6520910462677426e-05, + "loss": 0.4114, + "step": 9376 + }, + { + "epoch": 0.295757766913736, + "grad_norm": 2.198265314102173, + "learning_rate": 1.6520135945142057e-05, + "loss": 0.3969, + "step": 9377 + }, + { + "epoch": 0.29578930768017664, + "grad_norm": 1.7870979309082031, + "learning_rate": 1.6519361359564002e-05, + "loss": 0.5411, + "step": 9378 + }, + { + "epoch": 0.29582084844661727, + "grad_norm": 2.112583875656128, + "learning_rate": 1.6518586705951353e-05, + "loss": 0.5368, + "step": 9379 + }, + { + "epoch": 0.2958523892130579, + "grad_norm": 1.6590218544006348, + "learning_rate": 1.6517811984312187e-05, + "loss": 0.5527, + "step": 9380 + }, + { + "epoch": 0.29588392997949853, + "grad_norm": 1.7356832027435303, + "learning_rate": 1.6517037194654594e-05, + "loss": 0.5544, + "step": 9381 + }, + { + "epoch": 0.2959154707459391, + "grad_norm": 1.4395129680633545, + "learning_rate": 1.6516262336986656e-05, + "loss": 0.5487, + "step": 9382 + }, + { + "epoch": 0.29594701151237973, + "grad_norm": 1.7412364482879639, + "learning_rate": 1.651548741131646e-05, + "loss": 0.8556, + "step": 9383 + }, + { + "epoch": 0.29597855227882036, + "grad_norm": 1.6852805614471436, + "learning_rate": 1.6514712417652098e-05, + "loss": 0.36, + "step": 9384 + }, + { + "epoch": 0.296010093045261, + "grad_norm": 2.5194554328918457, + "learning_rate": 1.651393735600165e-05, + "loss": 0.5599, + "step": 9385 + }, + { + "epoch": 0.2960416338117016, + "grad_norm": 2.7525997161865234, + "learning_rate": 1.6513162226373207e-05, + "loss": 0.5487, + "step": 9386 + }, + { + "epoch": 0.29607317457814225, + "grad_norm": 2.182396173477173, + "learning_rate": 1.651238702877486e-05, + "loss": 0.5165, + "step": 9387 + }, + { + "epoch": 0.2961047153445829, + "grad_norm": 1.4515620470046997, + "learning_rate": 1.6511611763214697e-05, + "loss": 0.5467, + "step": 9388 + }, + { + "epoch": 0.2961362561110235, + "grad_norm": 1.623067021369934, + "learning_rate": 1.6510836429700813e-05, + "loss": 0.5324, + "step": 9389 + }, + { + "epoch": 0.29616779687746414, + "grad_norm": 2.2204129695892334, + "learning_rate": 1.651006102824129e-05, + "loss": 0.5453, + "step": 9390 + }, + { + "epoch": 0.29619933764390477, + "grad_norm": 2.872488021850586, + "learning_rate": 1.6509285558844224e-05, + "loss": 0.5487, + "step": 9391 + }, + { + "epoch": 0.29623087841034534, + "grad_norm": 2.4485599994659424, + "learning_rate": 1.6508510021517717e-05, + "loss": 0.8218, + "step": 9392 + }, + { + "epoch": 0.296262419176786, + "grad_norm": 1.826754093170166, + "learning_rate": 1.650773441626985e-05, + "loss": 0.5753, + "step": 9393 + }, + { + "epoch": 0.2962939599432266, + "grad_norm": 2.0359323024749756, + "learning_rate": 1.6506958743108718e-05, + "loss": 0.5595, + "step": 9394 + }, + { + "epoch": 0.29632550070966723, + "grad_norm": 1.2772912979125977, + "learning_rate": 1.650618300204242e-05, + "loss": 0.5446, + "step": 9395 + }, + { + "epoch": 0.29635704147610786, + "grad_norm": 1.9186415672302246, + "learning_rate": 1.6505407193079055e-05, + "loss": 0.5537, + "step": 9396 + }, + { + "epoch": 0.2963885822425485, + "grad_norm": 2.7779996395111084, + "learning_rate": 1.6504631316226713e-05, + "loss": 0.5297, + "step": 9397 + }, + { + "epoch": 0.2964201230089891, + "grad_norm": 2.3525211811065674, + "learning_rate": 1.6503855371493492e-05, + "loss": 0.567, + "step": 9398 + }, + { + "epoch": 0.29645166377542975, + "grad_norm": 1.348686933517456, + "learning_rate": 1.6503079358887488e-05, + "loss": 0.5239, + "step": 9399 + }, + { + "epoch": 0.2964832045418704, + "grad_norm": 1.8515344858169556, + "learning_rate": 1.6502303278416804e-05, + "loss": 0.5526, + "step": 9400 + }, + { + "epoch": 0.296514745308311, + "grad_norm": 1.9714603424072266, + "learning_rate": 1.650152713008953e-05, + "loss": 0.5567, + "step": 9401 + }, + { + "epoch": 0.29654628607475164, + "grad_norm": 1.644823670387268, + "learning_rate": 1.650075091391378e-05, + "loss": 0.5371, + "step": 9402 + }, + { + "epoch": 0.2965778268411922, + "grad_norm": 1.5630245208740234, + "learning_rate": 1.6499974629897644e-05, + "loss": 0.5742, + "step": 9403 + }, + { + "epoch": 0.29660936760763285, + "grad_norm": 1.4963382482528687, + "learning_rate": 1.6499198278049228e-05, + "loss": 0.4883, + "step": 9404 + }, + { + "epoch": 0.2966409083740735, + "grad_norm": 1.3619928359985352, + "learning_rate": 1.6498421858376628e-05, + "loss": 0.53, + "step": 9405 + }, + { + "epoch": 0.2966724491405141, + "grad_norm": 1.4101060628890991, + "learning_rate": 1.649764537088795e-05, + "loss": 0.5342, + "step": 9406 + }, + { + "epoch": 0.29670398990695473, + "grad_norm": 1.6552801132202148, + "learning_rate": 1.6496868815591298e-05, + "loss": 0.5622, + "step": 9407 + }, + { + "epoch": 0.29673553067339536, + "grad_norm": 4.747129917144775, + "learning_rate": 1.6496092192494774e-05, + "loss": 0.5542, + "step": 9408 + }, + { + "epoch": 0.296767071439836, + "grad_norm": 1.9743345975875854, + "learning_rate": 1.649531550160649e-05, + "loss": 0.543, + "step": 9409 + }, + { + "epoch": 0.2967986122062766, + "grad_norm": 1.7544766664505005, + "learning_rate": 1.6494538742934536e-05, + "loss": 0.3905, + "step": 9410 + }, + { + "epoch": 0.29683015297271725, + "grad_norm": 1.4749407768249512, + "learning_rate": 1.649376191648703e-05, + "loss": 0.5666, + "step": 9411 + }, + { + "epoch": 0.2968616937391579, + "grad_norm": 2.0528910160064697, + "learning_rate": 1.6492985022272083e-05, + "loss": 0.5708, + "step": 9412 + }, + { + "epoch": 0.2968932345055985, + "grad_norm": 1.841170072555542, + "learning_rate": 1.6492208060297788e-05, + "loss": 0.8973, + "step": 9413 + }, + { + "epoch": 0.2969247752720391, + "grad_norm": 2.0995328426361084, + "learning_rate": 1.6491431030572266e-05, + "loss": 0.5276, + "step": 9414 + }, + { + "epoch": 0.2969563160384797, + "grad_norm": 1.4833070039749146, + "learning_rate": 1.6490653933103617e-05, + "loss": 0.5052, + "step": 9415 + }, + { + "epoch": 0.29698785680492035, + "grad_norm": 3.9878644943237305, + "learning_rate": 1.6489876767899955e-05, + "loss": 0.5386, + "step": 9416 + }, + { + "epoch": 0.297019397571361, + "grad_norm": 2.5029075145721436, + "learning_rate": 1.648909953496939e-05, + "loss": 0.5215, + "step": 9417 + }, + { + "epoch": 0.2970509383378016, + "grad_norm": 1.689795970916748, + "learning_rate": 1.6488322234320033e-05, + "loss": 0.5413, + "step": 9418 + }, + { + "epoch": 0.29708247910424224, + "grad_norm": 1.6414297819137573, + "learning_rate": 1.6487544865959995e-05, + "loss": 0.54, + "step": 9419 + }, + { + "epoch": 0.29711401987068287, + "grad_norm": 1.7084474563598633, + "learning_rate": 1.648676742989739e-05, + "loss": 0.5459, + "step": 9420 + }, + { + "epoch": 0.2971455606371235, + "grad_norm": 1.5408929586410522, + "learning_rate": 1.6485989926140327e-05, + "loss": 0.5402, + "step": 9421 + }, + { + "epoch": 0.2971771014035641, + "grad_norm": 1.7798022031784058, + "learning_rate": 1.6485212354696926e-05, + "loss": 0.5264, + "step": 9422 + }, + { + "epoch": 0.29720864217000476, + "grad_norm": 1.6508113145828247, + "learning_rate": 1.6484434715575298e-05, + "loss": 0.5345, + "step": 9423 + }, + { + "epoch": 0.29724018293644533, + "grad_norm": 1.6044384241104126, + "learning_rate": 1.648365700878356e-05, + "loss": 0.5485, + "step": 9424 + }, + { + "epoch": 0.29727172370288596, + "grad_norm": 1.817351222038269, + "learning_rate": 1.6482879234329822e-05, + "loss": 0.5373, + "step": 9425 + }, + { + "epoch": 0.2973032644693266, + "grad_norm": 2.1243865489959717, + "learning_rate": 1.648210139222221e-05, + "loss": 0.5405, + "step": 9426 + }, + { + "epoch": 0.2973348052357672, + "grad_norm": 1.5687670707702637, + "learning_rate": 1.6481323482468836e-05, + "loss": 0.5578, + "step": 9427 + }, + { + "epoch": 0.29736634600220785, + "grad_norm": 1.4544142484664917, + "learning_rate": 1.6480545505077817e-05, + "loss": 0.5339, + "step": 9428 + }, + { + "epoch": 0.2973978867686485, + "grad_norm": 1.975899338722229, + "learning_rate": 1.6479767460057276e-05, + "loss": 0.5726, + "step": 9429 + }, + { + "epoch": 0.2974294275350891, + "grad_norm": 1.7039514780044556, + "learning_rate": 1.647898934741533e-05, + "loss": 0.5526, + "step": 9430 + }, + { + "epoch": 0.29746096830152974, + "grad_norm": 1.3644444942474365, + "learning_rate": 1.6478211167160093e-05, + "loss": 0.5099, + "step": 9431 + }, + { + "epoch": 0.29749250906797037, + "grad_norm": 1.5749365091323853, + "learning_rate": 1.6477432919299698e-05, + "loss": 0.5819, + "step": 9432 + }, + { + "epoch": 0.297524049834411, + "grad_norm": 2.209876775741577, + "learning_rate": 1.6476654603842265e-05, + "loss": 0.5422, + "step": 9433 + }, + { + "epoch": 0.2975555906008516, + "grad_norm": 2.0395538806915283, + "learning_rate": 1.6475876220795902e-05, + "loss": 0.5371, + "step": 9434 + }, + { + "epoch": 0.2975871313672922, + "grad_norm": 2.478391647338867, + "learning_rate": 1.647509777016875e-05, + "loss": 0.5452, + "step": 9435 + }, + { + "epoch": 0.29761867213373283, + "grad_norm": 1.7661893367767334, + "learning_rate": 1.647431925196892e-05, + "loss": 0.8456, + "step": 9436 + }, + { + "epoch": 0.29765021290017346, + "grad_norm": 1.7756644487380981, + "learning_rate": 1.6473540666204546e-05, + "loss": 0.5473, + "step": 9437 + }, + { + "epoch": 0.2976817536666141, + "grad_norm": 1.3563628196716309, + "learning_rate": 1.6472762012883746e-05, + "loss": 0.5155, + "step": 9438 + }, + { + "epoch": 0.2977132944330547, + "grad_norm": 1.65590238571167, + "learning_rate": 1.6471983292014647e-05, + "loss": 0.5644, + "step": 9439 + }, + { + "epoch": 0.29774483519949535, + "grad_norm": 1.3010421991348267, + "learning_rate": 1.6471204503605384e-05, + "loss": 0.372, + "step": 9440 + }, + { + "epoch": 0.297776375965936, + "grad_norm": 1.743193507194519, + "learning_rate": 1.647042564766407e-05, + "loss": 0.5763, + "step": 9441 + }, + { + "epoch": 0.2978079167323766, + "grad_norm": 1.6601769924163818, + "learning_rate": 1.6469646724198843e-05, + "loss": 0.4241, + "step": 9442 + }, + { + "epoch": 0.29783945749881724, + "grad_norm": 1.9537171125411987, + "learning_rate": 1.6468867733217825e-05, + "loss": 0.5188, + "step": 9443 + }, + { + "epoch": 0.29787099826525787, + "grad_norm": 1.3698513507843018, + "learning_rate": 1.6468088674729156e-05, + "loss": 0.5367, + "step": 9444 + }, + { + "epoch": 0.29790253903169844, + "grad_norm": 1.9329767227172852, + "learning_rate": 1.6467309548740956e-05, + "loss": 0.8596, + "step": 9445 + }, + { + "epoch": 0.2979340797981391, + "grad_norm": 1.5634840726852417, + "learning_rate": 1.6466530355261357e-05, + "loss": 0.8727, + "step": 9446 + }, + { + "epoch": 0.2979656205645797, + "grad_norm": 1.9704747200012207, + "learning_rate": 1.6465751094298498e-05, + "loss": 0.5435, + "step": 9447 + }, + { + "epoch": 0.29799716133102033, + "grad_norm": 2.9779982566833496, + "learning_rate": 1.64649717658605e-05, + "loss": 0.5319, + "step": 9448 + }, + { + "epoch": 0.29802870209746096, + "grad_norm": 2.126389741897583, + "learning_rate": 1.6464192369955506e-05, + "loss": 0.5067, + "step": 9449 + }, + { + "epoch": 0.2980602428639016, + "grad_norm": 4.69340705871582, + "learning_rate": 1.6463412906591644e-05, + "loss": 0.9068, + "step": 9450 + }, + { + "epoch": 0.2980917836303422, + "grad_norm": 5.329592227935791, + "learning_rate": 1.6462633375777047e-05, + "loss": 0.5351, + "step": 9451 + }, + { + "epoch": 0.29812332439678285, + "grad_norm": 1.870339274406433, + "learning_rate": 1.6461853777519854e-05, + "loss": 0.5326, + "step": 9452 + }, + { + "epoch": 0.2981548651632235, + "grad_norm": 2.129439115524292, + "learning_rate": 1.6461074111828203e-05, + "loss": 0.5264, + "step": 9453 + }, + { + "epoch": 0.2981864059296641, + "grad_norm": 1.8230664730072021, + "learning_rate": 1.6460294378710222e-05, + "loss": 0.565, + "step": 9454 + }, + { + "epoch": 0.29821794669610474, + "grad_norm": 1.5160455703735352, + "learning_rate": 1.6459514578174057e-05, + "loss": 0.8833, + "step": 9455 + }, + { + "epoch": 0.2982494874625453, + "grad_norm": 1.27036714553833, + "learning_rate": 1.6458734710227838e-05, + "loss": 0.4936, + "step": 9456 + }, + { + "epoch": 0.29828102822898594, + "grad_norm": 1.700321912765503, + "learning_rate": 1.6457954774879708e-05, + "loss": 0.5398, + "step": 9457 + }, + { + "epoch": 0.2983125689954266, + "grad_norm": 1.5163451433181763, + "learning_rate": 1.6457174772137804e-05, + "loss": 0.505, + "step": 9458 + }, + { + "epoch": 0.2983441097618672, + "grad_norm": 2.364405870437622, + "learning_rate": 1.6456394702010272e-05, + "loss": 0.532, + "step": 9459 + }, + { + "epoch": 0.29837565052830783, + "grad_norm": 1.3793995380401611, + "learning_rate": 1.6455614564505244e-05, + "loss": 0.5271, + "step": 9460 + }, + { + "epoch": 0.29840719129474846, + "grad_norm": 1.9863375425338745, + "learning_rate": 1.6454834359630867e-05, + "loss": 0.5425, + "step": 9461 + }, + { + "epoch": 0.2984387320611891, + "grad_norm": 1.6072546243667603, + "learning_rate": 1.6454054087395284e-05, + "loss": 0.5118, + "step": 9462 + }, + { + "epoch": 0.2984702728276297, + "grad_norm": 2.2320175170898438, + "learning_rate": 1.6453273747806632e-05, + "loss": 0.5419, + "step": 9463 + }, + { + "epoch": 0.29850181359407035, + "grad_norm": 1.5980308055877686, + "learning_rate": 1.6452493340873058e-05, + "loss": 0.5315, + "step": 9464 + }, + { + "epoch": 0.298533354360511, + "grad_norm": 1.4845205545425415, + "learning_rate": 1.64517128666027e-05, + "loss": 0.531, + "step": 9465 + }, + { + "epoch": 0.29856489512695156, + "grad_norm": 1.4370745420455933, + "learning_rate": 1.645093232500372e-05, + "loss": 0.522, + "step": 9466 + }, + { + "epoch": 0.2985964358933922, + "grad_norm": 1.6330455541610718, + "learning_rate": 1.6450151716084245e-05, + "loss": 0.5253, + "step": 9467 + }, + { + "epoch": 0.2986279766598328, + "grad_norm": 1.686122179031372, + "learning_rate": 1.6449371039852432e-05, + "loss": 0.845, + "step": 9468 + }, + { + "epoch": 0.29865951742627345, + "grad_norm": 1.5872384309768677, + "learning_rate": 1.644859029631642e-05, + "loss": 0.5562, + "step": 9469 + }, + { + "epoch": 0.2986910581927141, + "grad_norm": 1.6111215353012085, + "learning_rate": 1.6447809485484364e-05, + "loss": 0.5438, + "step": 9470 + }, + { + "epoch": 0.2987225989591547, + "grad_norm": 3.730861186981201, + "learning_rate": 1.644702860736441e-05, + "loss": 0.8798, + "step": 9471 + }, + { + "epoch": 0.29875413972559534, + "grad_norm": 2.1264097690582275, + "learning_rate": 1.6446247661964705e-05, + "loss": 0.5439, + "step": 9472 + }, + { + "epoch": 0.29878568049203597, + "grad_norm": 1.533493995666504, + "learning_rate": 1.6445466649293398e-05, + "loss": 0.5373, + "step": 9473 + }, + { + "epoch": 0.2988172212584766, + "grad_norm": 1.8083624839782715, + "learning_rate": 1.6444685569358645e-05, + "loss": 0.5236, + "step": 9474 + }, + { + "epoch": 0.2988487620249172, + "grad_norm": 1.5115314722061157, + "learning_rate": 1.644390442216859e-05, + "loss": 0.5481, + "step": 9475 + }, + { + "epoch": 0.29888030279135785, + "grad_norm": 1.3648710250854492, + "learning_rate": 1.6443123207731396e-05, + "loss": 0.5223, + "step": 9476 + }, + { + "epoch": 0.29891184355779843, + "grad_norm": 1.4228066205978394, + "learning_rate": 1.6442341926055204e-05, + "loss": 0.5249, + "step": 9477 + }, + { + "epoch": 0.29894338432423906, + "grad_norm": 1.9731365442276, + "learning_rate": 1.644156057714817e-05, + "loss": 0.8584, + "step": 9478 + }, + { + "epoch": 0.2989749250906797, + "grad_norm": 1.4348450899124146, + "learning_rate": 1.6440779161018455e-05, + "loss": 0.5363, + "step": 9479 + }, + { + "epoch": 0.2990064658571203, + "grad_norm": 1.4078348875045776, + "learning_rate": 1.6439997677674205e-05, + "loss": 0.5255, + "step": 9480 + }, + { + "epoch": 0.29903800662356095, + "grad_norm": 1.719821572303772, + "learning_rate": 1.6439216127123576e-05, + "loss": 0.5441, + "step": 9481 + }, + { + "epoch": 0.2990695473900016, + "grad_norm": 1.7218594551086426, + "learning_rate": 1.6438434509374733e-05, + "loss": 0.5141, + "step": 9482 + }, + { + "epoch": 0.2991010881564422, + "grad_norm": 1.5581104755401611, + "learning_rate": 1.643765282443582e-05, + "loss": 0.5482, + "step": 9483 + }, + { + "epoch": 0.29913262892288284, + "grad_norm": 1.4837602376937866, + "learning_rate": 1.6436871072315007e-05, + "loss": 0.5309, + "step": 9484 + }, + { + "epoch": 0.29916416968932347, + "grad_norm": 1.7852293252944946, + "learning_rate": 1.6436089253020444e-05, + "loss": 0.517, + "step": 9485 + }, + { + "epoch": 0.2991957104557641, + "grad_norm": 3.070032835006714, + "learning_rate": 1.643530736656029e-05, + "loss": 0.8387, + "step": 9486 + }, + { + "epoch": 0.2992272512222047, + "grad_norm": 1.7272151708602905, + "learning_rate": 1.643452541294271e-05, + "loss": 0.5256, + "step": 9487 + }, + { + "epoch": 0.2992587919886453, + "grad_norm": 1.7100355625152588, + "learning_rate": 1.6433743392175857e-05, + "loss": 0.5185, + "step": 9488 + }, + { + "epoch": 0.29929033275508593, + "grad_norm": 2.300767660140991, + "learning_rate": 1.64329613042679e-05, + "loss": 0.5475, + "step": 9489 + }, + { + "epoch": 0.29932187352152656, + "grad_norm": 1.3812121152877808, + "learning_rate": 1.6432179149226996e-05, + "loss": 0.536, + "step": 9490 + }, + { + "epoch": 0.2993534142879672, + "grad_norm": 1.5557212829589844, + "learning_rate": 1.6431396927061303e-05, + "loss": 0.5764, + "step": 9491 + }, + { + "epoch": 0.2993849550544078, + "grad_norm": 1.3784468173980713, + "learning_rate": 1.6430614637778997e-05, + "loss": 0.5679, + "step": 9492 + }, + { + "epoch": 0.29941649582084845, + "grad_norm": 1.4025776386260986, + "learning_rate": 1.642983228138823e-05, + "loss": 0.5291, + "step": 9493 + }, + { + "epoch": 0.2994480365872891, + "grad_norm": 1.4968816041946411, + "learning_rate": 1.6429049857897167e-05, + "loss": 0.5038, + "step": 9494 + }, + { + "epoch": 0.2994795773537297, + "grad_norm": 1.495286226272583, + "learning_rate": 1.6428267367313987e-05, + "loss": 0.5358, + "step": 9495 + }, + { + "epoch": 0.29951111812017034, + "grad_norm": 1.8951772451400757, + "learning_rate": 1.6427484809646835e-05, + "loss": 0.5276, + "step": 9496 + }, + { + "epoch": 0.29954265888661097, + "grad_norm": 1.2988388538360596, + "learning_rate": 1.6426702184903893e-05, + "loss": 0.5165, + "step": 9497 + }, + { + "epoch": 0.29957419965305154, + "grad_norm": 1.4039020538330078, + "learning_rate": 1.6425919493093322e-05, + "loss": 0.5237, + "step": 9498 + }, + { + "epoch": 0.2996057404194922, + "grad_norm": 1.4678832292556763, + "learning_rate": 1.642513673422329e-05, + "loss": 0.5969, + "step": 9499 + }, + { + "epoch": 0.2996372811859328, + "grad_norm": 1.442994236946106, + "learning_rate": 1.6424353908301974e-05, + "loss": 0.5475, + "step": 9500 + }, + { + "epoch": 0.29966882195237343, + "grad_norm": 1.55736243724823, + "learning_rate": 1.642357101533753e-05, + "loss": 0.5648, + "step": 9501 + }, + { + "epoch": 0.29970036271881406, + "grad_norm": 1.4533991813659668, + "learning_rate": 1.6422788055338138e-05, + "loss": 0.5269, + "step": 9502 + }, + { + "epoch": 0.2997319034852547, + "grad_norm": 1.5270777940750122, + "learning_rate": 1.6422005028311964e-05, + "loss": 0.5324, + "step": 9503 + }, + { + "epoch": 0.2997634442516953, + "grad_norm": 1.6769822835922241, + "learning_rate": 1.6421221934267182e-05, + "loss": 0.5238, + "step": 9504 + }, + { + "epoch": 0.29979498501813595, + "grad_norm": 1.5094457864761353, + "learning_rate": 1.6420438773211963e-05, + "loss": 0.5512, + "step": 9505 + }, + { + "epoch": 0.2998265257845766, + "grad_norm": 1.854255199432373, + "learning_rate": 1.641965554515448e-05, + "loss": 0.5125, + "step": 9506 + }, + { + "epoch": 0.2998580665510172, + "grad_norm": 1.5349453687667847, + "learning_rate": 1.6418872250102906e-05, + "loss": 0.5243, + "step": 9507 + }, + { + "epoch": 0.29988960731745784, + "grad_norm": 1.934357762336731, + "learning_rate": 1.6418088888065417e-05, + "loss": 0.5188, + "step": 9508 + }, + { + "epoch": 0.2999211480838984, + "grad_norm": 1.9041353464126587, + "learning_rate": 1.6417305459050187e-05, + "loss": 0.5192, + "step": 9509 + }, + { + "epoch": 0.29995268885033904, + "grad_norm": 1.9696810245513916, + "learning_rate": 1.6416521963065387e-05, + "loss": 0.8579, + "step": 9510 + }, + { + "epoch": 0.2999842296167797, + "grad_norm": 2.296473741531372, + "learning_rate": 1.6415738400119204e-05, + "loss": 0.527, + "step": 9511 + }, + { + "epoch": 0.3000157703832203, + "grad_norm": 1.5960979461669922, + "learning_rate": 1.6414954770219807e-05, + "loss": 0.5556, + "step": 9512 + }, + { + "epoch": 0.30004731114966093, + "grad_norm": 1.3356658220291138, + "learning_rate": 1.6414171073375374e-05, + "loss": 0.5724, + "step": 9513 + }, + { + "epoch": 0.30007885191610156, + "grad_norm": 1.8096126317977905, + "learning_rate": 1.641338730959409e-05, + "loss": 0.4443, + "step": 9514 + }, + { + "epoch": 0.3001103926825422, + "grad_norm": 2.0380945205688477, + "learning_rate": 1.6412603478884125e-05, + "loss": 0.4711, + "step": 9515 + }, + { + "epoch": 0.3001419334489828, + "grad_norm": 1.648329496383667, + "learning_rate": 1.6411819581253666e-05, + "loss": 0.8582, + "step": 9516 + }, + { + "epoch": 0.30017347421542345, + "grad_norm": 2.217092514038086, + "learning_rate": 1.641103561671089e-05, + "loss": 0.5144, + "step": 9517 + }, + { + "epoch": 0.3002050149818641, + "grad_norm": 2.275676727294922, + "learning_rate": 1.6410251585263978e-05, + "loss": 0.5588, + "step": 9518 + }, + { + "epoch": 0.30023655574830466, + "grad_norm": 1.5352368354797363, + "learning_rate": 1.6409467486921117e-05, + "loss": 0.5624, + "step": 9519 + }, + { + "epoch": 0.3002680965147453, + "grad_norm": 1.3006795644760132, + "learning_rate": 1.640868332169048e-05, + "loss": 0.5328, + "step": 9520 + }, + { + "epoch": 0.3002996372811859, + "grad_norm": 1.2771129608154297, + "learning_rate": 1.6407899089580263e-05, + "loss": 0.5408, + "step": 9521 + }, + { + "epoch": 0.30033117804762655, + "grad_norm": 2.316279172897339, + "learning_rate": 1.640711479059864e-05, + "loss": 0.5603, + "step": 9522 + }, + { + "epoch": 0.3003627188140672, + "grad_norm": 1.3983389139175415, + "learning_rate": 1.64063304247538e-05, + "loss": 0.519, + "step": 9523 + }, + { + "epoch": 0.3003942595805078, + "grad_norm": 1.5860871076583862, + "learning_rate": 1.6405545992053926e-05, + "loss": 0.5082, + "step": 9524 + }, + { + "epoch": 0.30042580034694844, + "grad_norm": 2.7019667625427246, + "learning_rate": 1.6404761492507208e-05, + "loss": 0.5063, + "step": 9525 + }, + { + "epoch": 0.30045734111338906, + "grad_norm": 1.4386579990386963, + "learning_rate": 1.6403976926121828e-05, + "loss": 0.5241, + "step": 9526 + }, + { + "epoch": 0.3004888818798297, + "grad_norm": 1.9717178344726562, + "learning_rate": 1.640319229290598e-05, + "loss": 0.5299, + "step": 9527 + }, + { + "epoch": 0.3005204226462703, + "grad_norm": 2.5414841175079346, + "learning_rate": 1.6402407592867847e-05, + "loss": 0.5588, + "step": 9528 + }, + { + "epoch": 0.30055196341271095, + "grad_norm": 1.9587397575378418, + "learning_rate": 1.6401622826015616e-05, + "loss": 0.5099, + "step": 9529 + }, + { + "epoch": 0.30058350417915153, + "grad_norm": 1.7901941537857056, + "learning_rate": 1.6400837992357487e-05, + "loss": 0.85, + "step": 9530 + }, + { + "epoch": 0.30061504494559216, + "grad_norm": 1.3505958318710327, + "learning_rate": 1.6400053091901635e-05, + "loss": 0.5147, + "step": 9531 + }, + { + "epoch": 0.3006465857120328, + "grad_norm": 2.0009894371032715, + "learning_rate": 1.6399268124656263e-05, + "loss": 0.5604, + "step": 9532 + }, + { + "epoch": 0.3006781264784734, + "grad_norm": 2.5299456119537354, + "learning_rate": 1.6398483090629563e-05, + "loss": 0.5231, + "step": 9533 + }, + { + "epoch": 0.30070966724491405, + "grad_norm": 2.6034913063049316, + "learning_rate": 1.6397697989829718e-05, + "loss": 0.4698, + "step": 9534 + }, + { + "epoch": 0.3007412080113547, + "grad_norm": 1.4394299983978271, + "learning_rate": 1.639691282226493e-05, + "loss": 0.5426, + "step": 9535 + }, + { + "epoch": 0.3007727487777953, + "grad_norm": 1.7566624879837036, + "learning_rate": 1.639612758794339e-05, + "loss": 0.5666, + "step": 9536 + }, + { + "epoch": 0.30080428954423594, + "grad_norm": 1.8362098932266235, + "learning_rate": 1.639534228687329e-05, + "loss": 0.5191, + "step": 9537 + }, + { + "epoch": 0.30083583031067657, + "grad_norm": 1.5930207967758179, + "learning_rate": 1.639455691906283e-05, + "loss": 0.5282, + "step": 9538 + }, + { + "epoch": 0.3008673710771172, + "grad_norm": 2.0007998943328857, + "learning_rate": 1.6393771484520203e-05, + "loss": 0.5139, + "step": 9539 + }, + { + "epoch": 0.3008989118435578, + "grad_norm": 1.6636470556259155, + "learning_rate": 1.6392985983253602e-05, + "loss": 0.5664, + "step": 9540 + }, + { + "epoch": 0.3009304526099984, + "grad_norm": 1.7463128566741943, + "learning_rate": 1.639220041527123e-05, + "loss": 0.5478, + "step": 9541 + }, + { + "epoch": 0.30096199337643903, + "grad_norm": 1.844229817390442, + "learning_rate": 1.6391414780581287e-05, + "loss": 0.5452, + "step": 9542 + }, + { + "epoch": 0.30099353414287966, + "grad_norm": 1.7968195676803589, + "learning_rate": 1.6390629079191964e-05, + "loss": 0.5588, + "step": 9543 + }, + { + "epoch": 0.3010250749093203, + "grad_norm": 1.9648126363754272, + "learning_rate": 1.6389843311111466e-05, + "loss": 0.5334, + "step": 9544 + }, + { + "epoch": 0.3010566156757609, + "grad_norm": 1.429336428642273, + "learning_rate": 1.6389057476347993e-05, + "loss": 0.5558, + "step": 9545 + }, + { + "epoch": 0.30108815644220155, + "grad_norm": 1.6634185314178467, + "learning_rate": 1.638827157490974e-05, + "loss": 0.539, + "step": 9546 + }, + { + "epoch": 0.3011196972086422, + "grad_norm": 2.494330883026123, + "learning_rate": 1.6387485606804918e-05, + "loss": 0.5569, + "step": 9547 + }, + { + "epoch": 0.3011512379750828, + "grad_norm": 1.5748353004455566, + "learning_rate": 1.6386699572041716e-05, + "loss": 0.5475, + "step": 9548 + }, + { + "epoch": 0.30118277874152344, + "grad_norm": 1.834816813468933, + "learning_rate": 1.6385913470628352e-05, + "loss": 0.5216, + "step": 9549 + }, + { + "epoch": 0.30121431950796407, + "grad_norm": 1.9238158464431763, + "learning_rate": 1.6385127302573024e-05, + "loss": 0.515, + "step": 9550 + }, + { + "epoch": 0.30124586027440464, + "grad_norm": 2.692567825317383, + "learning_rate": 1.6384341067883928e-05, + "loss": 0.5473, + "step": 9551 + }, + { + "epoch": 0.30127740104084527, + "grad_norm": 1.887901782989502, + "learning_rate": 1.6383554766569277e-05, + "loss": 0.8695, + "step": 9552 + }, + { + "epoch": 0.3013089418072859, + "grad_norm": 2.1872379779815674, + "learning_rate": 1.6382768398637277e-05, + "loss": 0.4922, + "step": 9553 + }, + { + "epoch": 0.30134048257372653, + "grad_norm": 2.1970858573913574, + "learning_rate": 1.638198196409613e-05, + "loss": 0.5758, + "step": 9554 + }, + { + "epoch": 0.30137202334016716, + "grad_norm": 1.8214107751846313, + "learning_rate": 1.638119546295405e-05, + "loss": 0.5139, + "step": 9555 + }, + { + "epoch": 0.3014035641066078, + "grad_norm": 1.8697625398635864, + "learning_rate": 1.638040889521924e-05, + "loss": 0.5427, + "step": 9556 + }, + { + "epoch": 0.3014351048730484, + "grad_norm": 2.4658138751983643, + "learning_rate": 1.637962226089991e-05, + "loss": 0.5649, + "step": 9557 + }, + { + "epoch": 0.30146664563948905, + "grad_norm": 1.5437642335891724, + "learning_rate": 1.637883556000426e-05, + "loss": 0.8644, + "step": 9558 + }, + { + "epoch": 0.3014981864059297, + "grad_norm": 1.5412144660949707, + "learning_rate": 1.637804879254052e-05, + "loss": 0.5488, + "step": 9559 + }, + { + "epoch": 0.3015297271723703, + "grad_norm": 1.6719990968704224, + "learning_rate": 1.6377261958516883e-05, + "loss": 0.546, + "step": 9560 + }, + { + "epoch": 0.30156126793881094, + "grad_norm": 2.988933801651001, + "learning_rate": 1.637647505794156e-05, + "loss": 0.5775, + "step": 9561 + }, + { + "epoch": 0.3015928087052515, + "grad_norm": 2.2459380626678467, + "learning_rate": 1.6375688090822773e-05, + "loss": 0.5264, + "step": 9562 + }, + { + "epoch": 0.30162434947169214, + "grad_norm": 1.7238863706588745, + "learning_rate": 1.637490105716873e-05, + "loss": 0.5432, + "step": 9563 + }, + { + "epoch": 0.3016558902381328, + "grad_norm": 1.6174293756484985, + "learning_rate": 1.6374113956987648e-05, + "loss": 0.5419, + "step": 9564 + }, + { + "epoch": 0.3016874310045734, + "grad_norm": 1.7992533445358276, + "learning_rate": 1.6373326790287732e-05, + "loss": 0.5539, + "step": 9565 + }, + { + "epoch": 0.30171897177101403, + "grad_norm": 1.6068544387817383, + "learning_rate": 1.6372539557077204e-05, + "loss": 0.8604, + "step": 9566 + }, + { + "epoch": 0.30175051253745466, + "grad_norm": 3.1048269271850586, + "learning_rate": 1.637175225736428e-05, + "loss": 0.542, + "step": 9567 + }, + { + "epoch": 0.3017820533038953, + "grad_norm": 1.3669906854629517, + "learning_rate": 1.637096489115717e-05, + "loss": 0.5458, + "step": 9568 + }, + { + "epoch": 0.3018135940703359, + "grad_norm": 1.7323634624481201, + "learning_rate": 1.63701774584641e-05, + "loss": 0.5241, + "step": 9569 + }, + { + "epoch": 0.30184513483677655, + "grad_norm": 1.609863042831421, + "learning_rate": 1.6369389959293272e-05, + "loss": 0.5124, + "step": 9570 + }, + { + "epoch": 0.3018766756032172, + "grad_norm": 1.7717106342315674, + "learning_rate": 1.6368602393652924e-05, + "loss": 0.5169, + "step": 9571 + }, + { + "epoch": 0.30190821636965776, + "grad_norm": 1.9404491186141968, + "learning_rate": 1.636781476155126e-05, + "loss": 0.5649, + "step": 9572 + }, + { + "epoch": 0.3019397571360984, + "grad_norm": 1.8797258138656616, + "learning_rate": 1.6367027062996505e-05, + "loss": 0.5489, + "step": 9573 + }, + { + "epoch": 0.301971297902539, + "grad_norm": 1.604876160621643, + "learning_rate": 1.6366239297996883e-05, + "loss": 0.8531, + "step": 9574 + }, + { + "epoch": 0.30200283866897965, + "grad_norm": 1.8480472564697266, + "learning_rate": 1.6365451466560608e-05, + "loss": 0.5396, + "step": 9575 + }, + { + "epoch": 0.3020343794354203, + "grad_norm": 2.1477465629577637, + "learning_rate": 1.63646635686959e-05, + "loss": 0.4961, + "step": 9576 + }, + { + "epoch": 0.3020659202018609, + "grad_norm": 1.6309021711349487, + "learning_rate": 1.6363875604410987e-05, + "loss": 0.5301, + "step": 9577 + }, + { + "epoch": 0.30209746096830153, + "grad_norm": 1.8786295652389526, + "learning_rate": 1.6363087573714094e-05, + "loss": 0.5288, + "step": 9578 + }, + { + "epoch": 0.30212900173474216, + "grad_norm": 2.629058361053467, + "learning_rate": 1.636229947661344e-05, + "loss": 0.5036, + "step": 9579 + }, + { + "epoch": 0.3021605425011828, + "grad_norm": 1.7657564878463745, + "learning_rate": 1.636151131311725e-05, + "loss": 0.5434, + "step": 9580 + }, + { + "epoch": 0.3021920832676234, + "grad_norm": 1.7036386728286743, + "learning_rate": 1.6360723083233754e-05, + "loss": 0.5191, + "step": 9581 + }, + { + "epoch": 0.30222362403406405, + "grad_norm": 1.980628490447998, + "learning_rate": 1.6359934786971168e-05, + "loss": 0.5728, + "step": 9582 + }, + { + "epoch": 0.3022551648005046, + "grad_norm": 1.9738191366195679, + "learning_rate": 1.6359146424337726e-05, + "loss": 0.5312, + "step": 9583 + }, + { + "epoch": 0.30228670556694526, + "grad_norm": 2.3443899154663086, + "learning_rate": 1.6358357995341655e-05, + "loss": 0.551, + "step": 9584 + }, + { + "epoch": 0.3023182463333859, + "grad_norm": 1.7639942169189453, + "learning_rate": 1.6357569499991178e-05, + "loss": 0.5351, + "step": 9585 + }, + { + "epoch": 0.3023497870998265, + "grad_norm": 3.211604356765747, + "learning_rate": 1.635678093829453e-05, + "loss": 0.5405, + "step": 9586 + }, + { + "epoch": 0.30238132786626715, + "grad_norm": 1.562912940979004, + "learning_rate": 1.6355992310259934e-05, + "loss": 0.5409, + "step": 9587 + }, + { + "epoch": 0.3024128686327078, + "grad_norm": 1.4959930181503296, + "learning_rate": 1.6355203615895624e-05, + "loss": 0.5337, + "step": 9588 + }, + { + "epoch": 0.3024444093991484, + "grad_norm": 3.242285966873169, + "learning_rate": 1.6354414855209833e-05, + "loss": 0.5388, + "step": 9589 + }, + { + "epoch": 0.30247595016558904, + "grad_norm": 1.7003581523895264, + "learning_rate": 1.635362602821078e-05, + "loss": 0.547, + "step": 9590 + }, + { + "epoch": 0.30250749093202967, + "grad_norm": 2.3871443271636963, + "learning_rate": 1.635283713490672e-05, + "loss": 0.5075, + "step": 9591 + }, + { + "epoch": 0.3025390316984703, + "grad_norm": 5.601263523101807, + "learning_rate": 1.635204817530586e-05, + "loss": 0.5269, + "step": 9592 + }, + { + "epoch": 0.30257057246491087, + "grad_norm": 1.719478964805603, + "learning_rate": 1.635125914941645e-05, + "loss": 0.5196, + "step": 9593 + }, + { + "epoch": 0.3026021132313515, + "grad_norm": 2.239854574203491, + "learning_rate": 1.635047005724672e-05, + "loss": 0.5272, + "step": 9594 + }, + { + "epoch": 0.30263365399779213, + "grad_norm": 2.1945641040802, + "learning_rate": 1.63496808988049e-05, + "loss": 0.4686, + "step": 9595 + }, + { + "epoch": 0.30266519476423276, + "grad_norm": 1.6164239645004272, + "learning_rate": 1.634889167409923e-05, + "loss": 0.5191, + "step": 9596 + }, + { + "epoch": 0.3026967355306734, + "grad_norm": 2.8681962490081787, + "learning_rate": 1.6348102383137944e-05, + "loss": 0.5494, + "step": 9597 + }, + { + "epoch": 0.302728276297114, + "grad_norm": 1.4390901327133179, + "learning_rate": 1.6347313025929285e-05, + "loss": 0.5233, + "step": 9598 + }, + { + "epoch": 0.30275981706355465, + "grad_norm": 5.421645164489746, + "learning_rate": 1.6346523602481484e-05, + "loss": 0.5315, + "step": 9599 + }, + { + "epoch": 0.3027913578299953, + "grad_norm": 1.7751861810684204, + "learning_rate": 1.6345734112802784e-05, + "loss": 0.5627, + "step": 9600 + }, + { + "epoch": 0.3028228985964359, + "grad_norm": 2.851832628250122, + "learning_rate": 1.6344944556901415e-05, + "loss": 0.529, + "step": 9601 + }, + { + "epoch": 0.30285443936287654, + "grad_norm": 1.9817888736724854, + "learning_rate": 1.6344154934785628e-05, + "loss": 0.5355, + "step": 9602 + }, + { + "epoch": 0.30288598012931717, + "grad_norm": 1.751670002937317, + "learning_rate": 1.6343365246463658e-05, + "loss": 0.5186, + "step": 9603 + }, + { + "epoch": 0.30291752089575774, + "grad_norm": 1.9237946271896362, + "learning_rate": 1.6342575491943746e-05, + "loss": 0.5413, + "step": 9604 + }, + { + "epoch": 0.30294906166219837, + "grad_norm": 2.1070730686187744, + "learning_rate": 1.634178567123413e-05, + "loss": 0.5193, + "step": 9605 + }, + { + "epoch": 0.302980602428639, + "grad_norm": 2.1493046283721924, + "learning_rate": 1.6340995784343058e-05, + "loss": 0.5237, + "step": 9606 + }, + { + "epoch": 0.30301214319507963, + "grad_norm": 1.912309169769287, + "learning_rate": 1.634020583127877e-05, + "loss": 0.5321, + "step": 9607 + }, + { + "epoch": 0.30304368396152026, + "grad_norm": 1.7608919143676758, + "learning_rate": 1.6339415812049517e-05, + "loss": 0.5377, + "step": 9608 + }, + { + "epoch": 0.3030752247279609, + "grad_norm": 1.7178261280059814, + "learning_rate": 1.6338625726663536e-05, + "loss": 0.5763, + "step": 9609 + }, + { + "epoch": 0.3031067654944015, + "grad_norm": 1.6264631748199463, + "learning_rate": 1.633783557512907e-05, + "loss": 0.5588, + "step": 9610 + }, + { + "epoch": 0.30313830626084215, + "grad_norm": 2.201308488845825, + "learning_rate": 1.6337045357454368e-05, + "loss": 0.511, + "step": 9611 + }, + { + "epoch": 0.3031698470272828, + "grad_norm": 1.9816001653671265, + "learning_rate": 1.6336255073647683e-05, + "loss": 0.5477, + "step": 9612 + }, + { + "epoch": 0.3032013877937234, + "grad_norm": 2.0152878761291504, + "learning_rate": 1.633546472371725e-05, + "loss": 0.8787, + "step": 9613 + }, + { + "epoch": 0.30323292856016404, + "grad_norm": 1.797839641571045, + "learning_rate": 1.633467430767133e-05, + "loss": 0.852, + "step": 9614 + }, + { + "epoch": 0.3032644693266046, + "grad_norm": 1.8051202297210693, + "learning_rate": 1.6333883825518158e-05, + "loss": 0.5271, + "step": 9615 + }, + { + "epoch": 0.30329601009304524, + "grad_norm": 1.8976478576660156, + "learning_rate": 1.633309327726599e-05, + "loss": 0.4208, + "step": 9616 + }, + { + "epoch": 0.3033275508594859, + "grad_norm": 1.879549503326416, + "learning_rate": 1.633230266292308e-05, + "loss": 0.554, + "step": 9617 + }, + { + "epoch": 0.3033590916259265, + "grad_norm": 2.427772283554077, + "learning_rate": 1.6331511982497677e-05, + "loss": 0.8487, + "step": 9618 + }, + { + "epoch": 0.30339063239236713, + "grad_norm": 1.9066342115402222, + "learning_rate": 1.6330721235998026e-05, + "loss": 0.53, + "step": 9619 + }, + { + "epoch": 0.30342217315880776, + "grad_norm": 1.6501272916793823, + "learning_rate": 1.6329930423432383e-05, + "loss": 0.5146, + "step": 9620 + }, + { + "epoch": 0.3034537139252484, + "grad_norm": 1.7304210662841797, + "learning_rate": 1.6329139544809002e-05, + "loss": 0.5481, + "step": 9621 + }, + { + "epoch": 0.303485254691689, + "grad_norm": 1.6328808069229126, + "learning_rate": 1.6328348600136137e-05, + "loss": 0.8444, + "step": 9622 + }, + { + "epoch": 0.30351679545812965, + "grad_norm": 1.6590968370437622, + "learning_rate": 1.6327557589422037e-05, + "loss": 0.5556, + "step": 9623 + }, + { + "epoch": 0.3035483362245703, + "grad_norm": 2.1573030948638916, + "learning_rate": 1.6326766512674966e-05, + "loss": 0.564, + "step": 9624 + }, + { + "epoch": 0.30357987699101086, + "grad_norm": 1.7597304582595825, + "learning_rate": 1.632597536990317e-05, + "loss": 0.5527, + "step": 9625 + }, + { + "epoch": 0.3036114177574515, + "grad_norm": 1.5981886386871338, + "learning_rate": 1.632518416111491e-05, + "loss": 0.5112, + "step": 9626 + }, + { + "epoch": 0.3036429585238921, + "grad_norm": 1.8513494729995728, + "learning_rate": 1.6324392886318438e-05, + "loss": 0.5463, + "step": 9627 + }, + { + "epoch": 0.30367449929033274, + "grad_norm": 1.4478716850280762, + "learning_rate": 1.6323601545522022e-05, + "loss": 0.5474, + "step": 9628 + }, + { + "epoch": 0.3037060400567734, + "grad_norm": 1.4996613264083862, + "learning_rate": 1.632281013873391e-05, + "loss": 0.557, + "step": 9629 + }, + { + "epoch": 0.303737580823214, + "grad_norm": 1.6501911878585815, + "learning_rate": 1.6322018665962365e-05, + "loss": 0.5373, + "step": 9630 + }, + { + "epoch": 0.30376912158965463, + "grad_norm": 2.4956705570220947, + "learning_rate": 1.6321227127215644e-05, + "loss": 0.5481, + "step": 9631 + }, + { + "epoch": 0.30380066235609526, + "grad_norm": 1.7068742513656616, + "learning_rate": 1.632043552250201e-05, + "loss": 0.5352, + "step": 9632 + }, + { + "epoch": 0.3038322031225359, + "grad_norm": 1.7633106708526611, + "learning_rate": 1.631964385182973e-05, + "loss": 0.8571, + "step": 9633 + }, + { + "epoch": 0.3038637438889765, + "grad_norm": 1.2547129392623901, + "learning_rate": 1.6318852115207054e-05, + "loss": 0.5322, + "step": 9634 + }, + { + "epoch": 0.30389528465541715, + "grad_norm": 1.6764726638793945, + "learning_rate": 1.631806031264225e-05, + "loss": 0.8788, + "step": 9635 + }, + { + "epoch": 0.3039268254218577, + "grad_norm": 2.0212349891662598, + "learning_rate": 1.6317268444143582e-05, + "loss": 0.5683, + "step": 9636 + }, + { + "epoch": 0.30395836618829836, + "grad_norm": 2.472132921218872, + "learning_rate": 1.631647650971931e-05, + "loss": 0.5646, + "step": 9637 + }, + { + "epoch": 0.303989906954739, + "grad_norm": 2.3307077884674072, + "learning_rate": 1.6315684509377706e-05, + "loss": 0.8621, + "step": 9638 + }, + { + "epoch": 0.3040214477211796, + "grad_norm": 11.27590560913086, + "learning_rate": 1.6314892443127025e-05, + "loss": 0.3738, + "step": 9639 + }, + { + "epoch": 0.30405298848762025, + "grad_norm": 1.6892865896224976, + "learning_rate": 1.6314100310975543e-05, + "loss": 0.8685, + "step": 9640 + }, + { + "epoch": 0.3040845292540609, + "grad_norm": 1.620226263999939, + "learning_rate": 1.6313308112931515e-05, + "loss": 0.5083, + "step": 9641 + }, + { + "epoch": 0.3041160700205015, + "grad_norm": 1.8134959936141968, + "learning_rate": 1.6312515849003218e-05, + "loss": 0.5285, + "step": 9642 + }, + { + "epoch": 0.30414761078694214, + "grad_norm": 1.3374594449996948, + "learning_rate": 1.631172351919892e-05, + "loss": 0.5037, + "step": 9643 + }, + { + "epoch": 0.30417915155338277, + "grad_norm": 2.2253780364990234, + "learning_rate": 1.6310931123526882e-05, + "loss": 0.5397, + "step": 9644 + }, + { + "epoch": 0.3042106923198234, + "grad_norm": 1.8494551181793213, + "learning_rate": 1.6310138661995377e-05, + "loss": 0.5322, + "step": 9645 + }, + { + "epoch": 0.30424223308626397, + "grad_norm": 2.1048290729522705, + "learning_rate": 1.6309346134612677e-05, + "loss": 0.5267, + "step": 9646 + }, + { + "epoch": 0.3042737738527046, + "grad_norm": 1.8452316522598267, + "learning_rate": 1.630855354138705e-05, + "loss": 0.5113, + "step": 9647 + }, + { + "epoch": 0.30430531461914523, + "grad_norm": 1.9317611455917358, + "learning_rate": 1.630776088232677e-05, + "loss": 0.5454, + "step": 9648 + }, + { + "epoch": 0.30433685538558586, + "grad_norm": 1.970542550086975, + "learning_rate": 1.6306968157440105e-05, + "loss": 0.5568, + "step": 9649 + }, + { + "epoch": 0.3043683961520265, + "grad_norm": 1.9658513069152832, + "learning_rate": 1.6306175366735333e-05, + "loss": 0.5208, + "step": 9650 + }, + { + "epoch": 0.3043999369184671, + "grad_norm": 2.181666851043701, + "learning_rate": 1.6305382510220723e-05, + "loss": 0.5164, + "step": 9651 + }, + { + "epoch": 0.30443147768490775, + "grad_norm": 2.3110859394073486, + "learning_rate": 1.6304589587904552e-05, + "loss": 0.5813, + "step": 9652 + }, + { + "epoch": 0.3044630184513484, + "grad_norm": 1.9453643560409546, + "learning_rate": 1.6303796599795093e-05, + "loss": 0.545, + "step": 9653 + }, + { + "epoch": 0.304494559217789, + "grad_norm": 2.683568000793457, + "learning_rate": 1.630300354590062e-05, + "loss": 0.5194, + "step": 9654 + }, + { + "epoch": 0.30452609998422964, + "grad_norm": 1.6104722023010254, + "learning_rate": 1.630221042622941e-05, + "loss": 0.5276, + "step": 9655 + }, + { + "epoch": 0.30455764075067027, + "grad_norm": 1.9418210983276367, + "learning_rate": 1.630141724078974e-05, + "loss": 0.5348, + "step": 9656 + }, + { + "epoch": 0.30458918151711084, + "grad_norm": 1.546397089958191, + "learning_rate": 1.630062398958989e-05, + "loss": 0.5202, + "step": 9657 + }, + { + "epoch": 0.30462072228355147, + "grad_norm": 2.1824305057525635, + "learning_rate": 1.6299830672638138e-05, + "loss": 0.5628, + "step": 9658 + }, + { + "epoch": 0.3046522630499921, + "grad_norm": 1.5645970106124878, + "learning_rate": 1.6299037289942758e-05, + "loss": 0.5493, + "step": 9659 + }, + { + "epoch": 0.30468380381643273, + "grad_norm": 1.7020739316940308, + "learning_rate": 1.6298243841512038e-05, + "loss": 0.5476, + "step": 9660 + }, + { + "epoch": 0.30471534458287336, + "grad_norm": 1.8643876314163208, + "learning_rate": 1.629745032735425e-05, + "loss": 0.5512, + "step": 9661 + }, + { + "epoch": 0.304746885349314, + "grad_norm": 2.8586535453796387, + "learning_rate": 1.6296656747477678e-05, + "loss": 0.5195, + "step": 9662 + }, + { + "epoch": 0.3047784261157546, + "grad_norm": 1.5588105916976929, + "learning_rate": 1.6295863101890603e-05, + "loss": 0.5457, + "step": 9663 + }, + { + "epoch": 0.30480996688219525, + "grad_norm": 1.4351924657821655, + "learning_rate": 1.6295069390601306e-05, + "loss": 0.5663, + "step": 9664 + }, + { + "epoch": 0.3048415076486359, + "grad_norm": 1.7874120473861694, + "learning_rate": 1.6294275613618073e-05, + "loss": 0.5465, + "step": 9665 + }, + { + "epoch": 0.3048730484150765, + "grad_norm": 2.3766720294952393, + "learning_rate": 1.629348177094919e-05, + "loss": 0.5388, + "step": 9666 + }, + { + "epoch": 0.30490458918151714, + "grad_norm": 1.491125464439392, + "learning_rate": 1.6292687862602936e-05, + "loss": 0.5296, + "step": 9667 + }, + { + "epoch": 0.3049361299479577, + "grad_norm": 2.0913755893707275, + "learning_rate": 1.6291893888587594e-05, + "loss": 0.5116, + "step": 9668 + }, + { + "epoch": 0.30496767071439834, + "grad_norm": 1.7810813188552856, + "learning_rate": 1.6291099848911457e-05, + "loss": 0.5259, + "step": 9669 + }, + { + "epoch": 0.30499921148083897, + "grad_norm": 1.655593752861023, + "learning_rate": 1.629030574358281e-05, + "loss": 0.5655, + "step": 9670 + }, + { + "epoch": 0.3050307522472796, + "grad_norm": 2.208707332611084, + "learning_rate": 1.6289511572609936e-05, + "loss": 0.4407, + "step": 9671 + }, + { + "epoch": 0.30506229301372023, + "grad_norm": 1.4697338342666626, + "learning_rate": 1.6288717336001126e-05, + "loss": 0.5611, + "step": 9672 + }, + { + "epoch": 0.30509383378016086, + "grad_norm": 1.7492119073867798, + "learning_rate": 1.6287923033764668e-05, + "loss": 0.5298, + "step": 9673 + }, + { + "epoch": 0.3051253745466015, + "grad_norm": 1.4105716943740845, + "learning_rate": 1.628712866590885e-05, + "loss": 0.5578, + "step": 9674 + }, + { + "epoch": 0.3051569153130421, + "grad_norm": 1.6153225898742676, + "learning_rate": 1.628633423244196e-05, + "loss": 0.5882, + "step": 9675 + }, + { + "epoch": 0.30518845607948275, + "grad_norm": 1.6042858362197876, + "learning_rate": 1.6285539733372298e-05, + "loss": 0.5067, + "step": 9676 + }, + { + "epoch": 0.3052199968459234, + "grad_norm": 2.0498616695404053, + "learning_rate": 1.6284745168708146e-05, + "loss": 0.9003, + "step": 9677 + }, + { + "epoch": 0.30525153761236395, + "grad_norm": 1.6386010646820068, + "learning_rate": 1.6283950538457793e-05, + "loss": 0.5454, + "step": 9678 + }, + { + "epoch": 0.3052830783788046, + "grad_norm": 1.7903015613555908, + "learning_rate": 1.628315584262954e-05, + "loss": 0.5542, + "step": 9679 + }, + { + "epoch": 0.3053146191452452, + "grad_norm": 1.6410975456237793, + "learning_rate": 1.6282361081231678e-05, + "loss": 0.5426, + "step": 9680 + }, + { + "epoch": 0.30534615991168584, + "grad_norm": 1.6048082113265991, + "learning_rate": 1.6281566254272498e-05, + "loss": 0.527, + "step": 9681 + }, + { + "epoch": 0.3053777006781265, + "grad_norm": 1.8354042768478394, + "learning_rate": 1.6280771361760297e-05, + "loss": 0.5439, + "step": 9682 + }, + { + "epoch": 0.3054092414445671, + "grad_norm": 2.3231332302093506, + "learning_rate": 1.627997640370337e-05, + "loss": 0.5549, + "step": 9683 + }, + { + "epoch": 0.30544078221100773, + "grad_norm": 2.1825525760650635, + "learning_rate": 1.6279181380110012e-05, + "loss": 0.5218, + "step": 9684 + }, + { + "epoch": 0.30547232297744836, + "grad_norm": 1.5700721740722656, + "learning_rate": 1.6278386290988525e-05, + "loss": 0.5135, + "step": 9685 + }, + { + "epoch": 0.305503863743889, + "grad_norm": 2.151787757873535, + "learning_rate": 1.62775911363472e-05, + "loss": 0.5345, + "step": 9686 + }, + { + "epoch": 0.3055354045103296, + "grad_norm": 1.7704325914382935, + "learning_rate": 1.6276795916194335e-05, + "loss": 0.5269, + "step": 9687 + }, + { + "epoch": 0.30556694527677025, + "grad_norm": 2.731267213821411, + "learning_rate": 1.6276000630538232e-05, + "loss": 0.4127, + "step": 9688 + }, + { + "epoch": 0.3055984860432108, + "grad_norm": 9.893000602722168, + "learning_rate": 1.6275205279387187e-05, + "loss": 0.5121, + "step": 9689 + }, + { + "epoch": 0.30563002680965146, + "grad_norm": 1.8622685670852661, + "learning_rate": 1.6274409862749504e-05, + "loss": 0.5104, + "step": 9690 + }, + { + "epoch": 0.3056615675760921, + "grad_norm": 2.325511932373047, + "learning_rate": 1.6273614380633484e-05, + "loss": 0.8583, + "step": 9691 + }, + { + "epoch": 0.3056931083425327, + "grad_norm": 1.7451951503753662, + "learning_rate": 1.6272818833047427e-05, + "loss": 0.5401, + "step": 9692 + }, + { + "epoch": 0.30572464910897335, + "grad_norm": 1.6726638078689575, + "learning_rate": 1.627202321999963e-05, + "loss": 0.5157, + "step": 9693 + }, + { + "epoch": 0.305756189875414, + "grad_norm": 1.6774767637252808, + "learning_rate": 1.6271227541498404e-05, + "loss": 0.519, + "step": 9694 + }, + { + "epoch": 0.3057877306418546, + "grad_norm": 1.3001620769500732, + "learning_rate": 1.627043179755205e-05, + "loss": 0.5404, + "step": 9695 + }, + { + "epoch": 0.30581927140829523, + "grad_norm": 1.3799433708190918, + "learning_rate": 1.626963598816887e-05, + "loss": 0.5172, + "step": 9696 + }, + { + "epoch": 0.30585081217473586, + "grad_norm": 2.443756103515625, + "learning_rate": 1.626884011335717e-05, + "loss": 0.4061, + "step": 9697 + }, + { + "epoch": 0.3058823529411765, + "grad_norm": 1.6386220455169678, + "learning_rate": 1.626804417312526e-05, + "loss": 0.5295, + "step": 9698 + }, + { + "epoch": 0.30591389370761707, + "grad_norm": 2.067565441131592, + "learning_rate": 1.6267248167481433e-05, + "loss": 0.3942, + "step": 9699 + }, + { + "epoch": 0.3059454344740577, + "grad_norm": 1.834854245185852, + "learning_rate": 1.6266452096434013e-05, + "loss": 0.5679, + "step": 9700 + }, + { + "epoch": 0.30597697524049833, + "grad_norm": 1.6483234167099, + "learning_rate": 1.6265655959991295e-05, + "loss": 0.5122, + "step": 9701 + }, + { + "epoch": 0.30600851600693896, + "grad_norm": 1.9710067510604858, + "learning_rate": 1.6264859758161592e-05, + "loss": 0.5427, + "step": 9702 + }, + { + "epoch": 0.3060400567733796, + "grad_norm": 1.6334609985351562, + "learning_rate": 1.6264063490953216e-05, + "loss": 0.868, + "step": 9703 + }, + { + "epoch": 0.3060715975398202, + "grad_norm": 1.8214716911315918, + "learning_rate": 1.626326715837447e-05, + "loss": 0.5411, + "step": 9704 + }, + { + "epoch": 0.30610313830626085, + "grad_norm": 1.5262147188186646, + "learning_rate": 1.6262470760433674e-05, + "loss": 0.507, + "step": 9705 + }, + { + "epoch": 0.3061346790727015, + "grad_norm": 2.1891531944274902, + "learning_rate": 1.6261674297139127e-05, + "loss": 0.5294, + "step": 9706 + }, + { + "epoch": 0.3061662198391421, + "grad_norm": 1.5961582660675049, + "learning_rate": 1.626087776849915e-05, + "loss": 0.5537, + "step": 9707 + }, + { + "epoch": 0.30619776060558274, + "grad_norm": 1.5574368238449097, + "learning_rate": 1.626008117452205e-05, + "loss": 0.5808, + "step": 9708 + }, + { + "epoch": 0.30622930137202337, + "grad_norm": 2.1524734497070312, + "learning_rate": 1.6259284515216143e-05, + "loss": 0.5428, + "step": 9709 + }, + { + "epoch": 0.30626084213846394, + "grad_norm": 1.6274505853652954, + "learning_rate": 1.625848779058974e-05, + "loss": 0.5608, + "step": 9710 + }, + { + "epoch": 0.30629238290490457, + "grad_norm": 1.8950647115707397, + "learning_rate": 1.625769100065116e-05, + "loss": 0.5415, + "step": 9711 + }, + { + "epoch": 0.3063239236713452, + "grad_norm": 1.3662434816360474, + "learning_rate": 1.6256894145408713e-05, + "loss": 0.5499, + "step": 9712 + }, + { + "epoch": 0.30635546443778583, + "grad_norm": 1.8378310203552246, + "learning_rate": 1.625609722487072e-05, + "loss": 0.5013, + "step": 9713 + }, + { + "epoch": 0.30638700520422646, + "grad_norm": 1.4015440940856934, + "learning_rate": 1.625530023904549e-05, + "loss": 0.5396, + "step": 9714 + }, + { + "epoch": 0.3064185459706671, + "grad_norm": 1.4245139360427856, + "learning_rate": 1.625450318794135e-05, + "loss": 0.515, + "step": 9715 + }, + { + "epoch": 0.3064500867371077, + "grad_norm": 1.35922110080719, + "learning_rate": 1.6253706071566613e-05, + "loss": 0.5443, + "step": 9716 + }, + { + "epoch": 0.30648162750354835, + "grad_norm": 1.8232940435409546, + "learning_rate": 1.6252908889929596e-05, + "loss": 0.8432, + "step": 9717 + }, + { + "epoch": 0.306513168269989, + "grad_norm": 3.528548240661621, + "learning_rate": 1.6252111643038618e-05, + "loss": 0.4176, + "step": 9718 + }, + { + "epoch": 0.3065447090364296, + "grad_norm": 1.4010967016220093, + "learning_rate": 1.6251314330902e-05, + "loss": 0.5344, + "step": 9719 + }, + { + "epoch": 0.3065762498028702, + "grad_norm": 1.8910366296768188, + "learning_rate": 1.6250516953528065e-05, + "loss": 0.587, + "step": 9720 + }, + { + "epoch": 0.3066077905693108, + "grad_norm": 1.4587604999542236, + "learning_rate": 1.6249719510925127e-05, + "loss": 0.5365, + "step": 9721 + }, + { + "epoch": 0.30663933133575144, + "grad_norm": 1.3681278228759766, + "learning_rate": 1.6248922003101523e-05, + "loss": 0.5286, + "step": 9722 + }, + { + "epoch": 0.30667087210219207, + "grad_norm": 2.938122034072876, + "learning_rate": 1.6248124430065556e-05, + "loss": 0.3689, + "step": 9723 + }, + { + "epoch": 0.3067024128686327, + "grad_norm": 1.3338277339935303, + "learning_rate": 1.6247326791825563e-05, + "loss": 0.5523, + "step": 9724 + }, + { + "epoch": 0.30673395363507333, + "grad_norm": 1.4551403522491455, + "learning_rate": 1.6246529088389867e-05, + "loss": 0.5176, + "step": 9725 + }, + { + "epoch": 0.30676549440151396, + "grad_norm": 1.5998971462249756, + "learning_rate": 1.6245731319766783e-05, + "loss": 0.509, + "step": 9726 + }, + { + "epoch": 0.3067970351679546, + "grad_norm": 2.210531234741211, + "learning_rate": 1.624493348596465e-05, + "loss": 0.557, + "step": 9727 + }, + { + "epoch": 0.3068285759343952, + "grad_norm": 1.877239465713501, + "learning_rate": 1.6244135586991784e-05, + "loss": 0.5552, + "step": 9728 + }, + { + "epoch": 0.30686011670083585, + "grad_norm": 1.5026521682739258, + "learning_rate": 1.624333762285651e-05, + "loss": 0.5516, + "step": 9729 + }, + { + "epoch": 0.3068916574672765, + "grad_norm": 1.6891698837280273, + "learning_rate": 1.624253959356717e-05, + "loss": 0.5196, + "step": 9730 + }, + { + "epoch": 0.30692319823371705, + "grad_norm": 1.6118276119232178, + "learning_rate": 1.6241741499132073e-05, + "loss": 0.5238, + "step": 9731 + }, + { + "epoch": 0.3069547390001577, + "grad_norm": 2.0449788570404053, + "learning_rate": 1.6240943339559564e-05, + "loss": 0.8735, + "step": 9732 + }, + { + "epoch": 0.3069862797665983, + "grad_norm": 2.6261203289031982, + "learning_rate": 1.624014511485796e-05, + "loss": 0.8547, + "step": 9733 + }, + { + "epoch": 0.30701782053303894, + "grad_norm": 1.7307924032211304, + "learning_rate": 1.6239346825035594e-05, + "loss": 0.5569, + "step": 9734 + }, + { + "epoch": 0.3070493612994796, + "grad_norm": 1.9040426015853882, + "learning_rate": 1.6238548470100804e-05, + "loss": 0.5289, + "step": 9735 + }, + { + "epoch": 0.3070809020659202, + "grad_norm": 1.712203025817871, + "learning_rate": 1.623775005006192e-05, + "loss": 0.5365, + "step": 9736 + }, + { + "epoch": 0.30711244283236083, + "grad_norm": 2.119551658630371, + "learning_rate": 1.6236951564927265e-05, + "loss": 0.4099, + "step": 9737 + }, + { + "epoch": 0.30714398359880146, + "grad_norm": 1.9311070442199707, + "learning_rate": 1.6236153014705177e-05, + "loss": 0.5378, + "step": 9738 + }, + { + "epoch": 0.3071755243652421, + "grad_norm": 1.6179356575012207, + "learning_rate": 1.623535439940399e-05, + "loss": 0.5088, + "step": 9739 + }, + { + "epoch": 0.3072070651316827, + "grad_norm": 1.8850456476211548, + "learning_rate": 1.6234555719032044e-05, + "loss": 0.5383, + "step": 9740 + }, + { + "epoch": 0.30723860589812335, + "grad_norm": 2.783438205718994, + "learning_rate": 1.6233756973597664e-05, + "loss": 0.5638, + "step": 9741 + }, + { + "epoch": 0.3072701466645639, + "grad_norm": 2.582524299621582, + "learning_rate": 1.623295816310919e-05, + "loss": 0.8857, + "step": 9742 + }, + { + "epoch": 0.30730168743100456, + "grad_norm": 1.6186132431030273, + "learning_rate": 1.623215928757496e-05, + "loss": 0.5261, + "step": 9743 + }, + { + "epoch": 0.3073332281974452, + "grad_norm": 2.001508951187134, + "learning_rate": 1.6231360347003306e-05, + "loss": 0.5249, + "step": 9744 + }, + { + "epoch": 0.3073647689638858, + "grad_norm": 3.549907922744751, + "learning_rate": 1.623056134140257e-05, + "loss": 0.8996, + "step": 9745 + }, + { + "epoch": 0.30739630973032644, + "grad_norm": 2.2934556007385254, + "learning_rate": 1.6229762270781086e-05, + "loss": 0.5349, + "step": 9746 + }, + { + "epoch": 0.3074278504967671, + "grad_norm": 1.7043894529342651, + "learning_rate": 1.6228963135147195e-05, + "loss": 0.5316, + "step": 9747 + }, + { + "epoch": 0.3074593912632077, + "grad_norm": 2.23087739944458, + "learning_rate": 1.622816393450924e-05, + "loss": 0.5225, + "step": 9748 + }, + { + "epoch": 0.30749093202964833, + "grad_norm": 1.6130660772323608, + "learning_rate": 1.6227364668875557e-05, + "loss": 0.5146, + "step": 9749 + }, + { + "epoch": 0.30752247279608896, + "grad_norm": 1.999011754989624, + "learning_rate": 1.622656533825449e-05, + "loss": 0.8507, + "step": 9750 + }, + { + "epoch": 0.3075540135625296, + "grad_norm": 2.0797922611236572, + "learning_rate": 1.6225765942654375e-05, + "loss": 0.5567, + "step": 9751 + }, + { + "epoch": 0.30758555432897017, + "grad_norm": 1.8906430006027222, + "learning_rate": 1.622496648208356e-05, + "loss": 0.5138, + "step": 9752 + }, + { + "epoch": 0.3076170950954108, + "grad_norm": 2.1011760234832764, + "learning_rate": 1.622416695655039e-05, + "loss": 0.559, + "step": 9753 + }, + { + "epoch": 0.3076486358618514, + "grad_norm": 1.8401398658752441, + "learning_rate": 1.62233673660632e-05, + "loss": 0.421, + "step": 9754 + }, + { + "epoch": 0.30768017662829206, + "grad_norm": 2.268023729324341, + "learning_rate": 1.622256771063034e-05, + "loss": 0.5477, + "step": 9755 + }, + { + "epoch": 0.3077117173947327, + "grad_norm": 1.5566518306732178, + "learning_rate": 1.6221767990260155e-05, + "loss": 0.551, + "step": 9756 + }, + { + "epoch": 0.3077432581611733, + "grad_norm": 1.7979713678359985, + "learning_rate": 1.622096820496099e-05, + "loss": 0.5566, + "step": 9757 + }, + { + "epoch": 0.30777479892761395, + "grad_norm": 1.8776437044143677, + "learning_rate": 1.622016835474119e-05, + "loss": 0.5433, + "step": 9758 + }, + { + "epoch": 0.3078063396940546, + "grad_norm": 2.504209041595459, + "learning_rate": 1.6219368439609105e-05, + "loss": 0.5384, + "step": 9759 + }, + { + "epoch": 0.3078378804604952, + "grad_norm": 2.212517023086548, + "learning_rate": 1.621856845957308e-05, + "loss": 0.5697, + "step": 9760 + }, + { + "epoch": 0.30786942122693584, + "grad_norm": 1.9896661043167114, + "learning_rate": 1.6217768414641467e-05, + "loss": 0.5355, + "step": 9761 + }, + { + "epoch": 0.30790096199337647, + "grad_norm": 2.983022928237915, + "learning_rate": 1.621696830482261e-05, + "loss": 0.5395, + "step": 9762 + }, + { + "epoch": 0.30793250275981704, + "grad_norm": 1.9770618677139282, + "learning_rate": 1.6216168130124862e-05, + "loss": 0.5447, + "step": 9763 + }, + { + "epoch": 0.30796404352625767, + "grad_norm": 1.5524656772613525, + "learning_rate": 1.6215367890556573e-05, + "loss": 0.5178, + "step": 9764 + }, + { + "epoch": 0.3079955842926983, + "grad_norm": 3.3349497318267822, + "learning_rate": 1.6214567586126097e-05, + "loss": 0.5463, + "step": 9765 + }, + { + "epoch": 0.30802712505913893, + "grad_norm": 1.7130221128463745, + "learning_rate": 1.6213767216841776e-05, + "loss": 0.5245, + "step": 9766 + }, + { + "epoch": 0.30805866582557956, + "grad_norm": 1.5354474782943726, + "learning_rate": 1.6212966782711975e-05, + "loss": 0.5576, + "step": 9767 + }, + { + "epoch": 0.3080902065920202, + "grad_norm": 2.103148937225342, + "learning_rate": 1.621216628374504e-05, + "loss": 0.883, + "step": 9768 + }, + { + "epoch": 0.3081217473584608, + "grad_norm": 1.796555757522583, + "learning_rate": 1.6211365719949324e-05, + "loss": 0.5177, + "step": 9769 + }, + { + "epoch": 0.30815328812490145, + "grad_norm": 2.24181866645813, + "learning_rate": 1.6210565091333185e-05, + "loss": 0.56, + "step": 9770 + }, + { + "epoch": 0.3081848288913421, + "grad_norm": 1.7161743640899658, + "learning_rate": 1.6209764397904976e-05, + "loss": 0.5279, + "step": 9771 + }, + { + "epoch": 0.3082163696577827, + "grad_norm": 1.750534176826477, + "learning_rate": 1.6208963639673057e-05, + "loss": 0.5449, + "step": 9772 + }, + { + "epoch": 0.3082479104242233, + "grad_norm": 1.8890929222106934, + "learning_rate": 1.620816281664578e-05, + "loss": 0.5579, + "step": 9773 + }, + { + "epoch": 0.3082794511906639, + "grad_norm": 2.3053441047668457, + "learning_rate": 1.6207361928831502e-05, + "loss": 0.506, + "step": 9774 + }, + { + "epoch": 0.30831099195710454, + "grad_norm": 1.772137999534607, + "learning_rate": 1.6206560976238583e-05, + "loss": 0.5648, + "step": 9775 + }, + { + "epoch": 0.30834253272354517, + "grad_norm": 2.076995611190796, + "learning_rate": 1.620575995887538e-05, + "loss": 0.5211, + "step": 9776 + }, + { + "epoch": 0.3083740734899858, + "grad_norm": 1.7448983192443848, + "learning_rate": 1.6204958876750258e-05, + "loss": 0.5325, + "step": 9777 + }, + { + "epoch": 0.30840561425642643, + "grad_norm": 1.9707800149917603, + "learning_rate": 1.6204157729871566e-05, + "loss": 0.5401, + "step": 9778 + }, + { + "epoch": 0.30843715502286706, + "grad_norm": 1.7325570583343506, + "learning_rate": 1.6203356518247674e-05, + "loss": 0.9077, + "step": 9779 + }, + { + "epoch": 0.3084686957893077, + "grad_norm": 1.9425346851348877, + "learning_rate": 1.6202555241886938e-05, + "loss": 0.5435, + "step": 9780 + }, + { + "epoch": 0.3085002365557483, + "grad_norm": 1.7836105823516846, + "learning_rate": 1.6201753900797723e-05, + "loss": 0.494, + "step": 9781 + }, + { + "epoch": 0.30853177732218895, + "grad_norm": 1.637963056564331, + "learning_rate": 1.620095249498839e-05, + "loss": 0.5134, + "step": 9782 + }, + { + "epoch": 0.3085633180886296, + "grad_norm": 1.6360682249069214, + "learning_rate": 1.6200151024467303e-05, + "loss": 0.5449, + "step": 9783 + }, + { + "epoch": 0.30859485885507015, + "grad_norm": 1.557848572731018, + "learning_rate": 1.6199349489242825e-05, + "loss": 0.5349, + "step": 9784 + }, + { + "epoch": 0.3086263996215108, + "grad_norm": 2.208428382873535, + "learning_rate": 1.6198547889323323e-05, + "loss": 0.4135, + "step": 9785 + }, + { + "epoch": 0.3086579403879514, + "grad_norm": 1.5479627847671509, + "learning_rate": 1.619774622471716e-05, + "loss": 0.5188, + "step": 9786 + }, + { + "epoch": 0.30868948115439204, + "grad_norm": 2.066714286804199, + "learning_rate": 1.6196944495432703e-05, + "loss": 0.5685, + "step": 9787 + }, + { + "epoch": 0.3087210219208327, + "grad_norm": 2.199260711669922, + "learning_rate": 1.6196142701478318e-05, + "loss": 0.5505, + "step": 9788 + }, + { + "epoch": 0.3087525626872733, + "grad_norm": 3.280972957611084, + "learning_rate": 1.6195340842862374e-05, + "loss": 0.5465, + "step": 9789 + }, + { + "epoch": 0.30878410345371393, + "grad_norm": 1.6492602825164795, + "learning_rate": 1.6194538919593238e-05, + "loss": 0.5033, + "step": 9790 + }, + { + "epoch": 0.30881564422015456, + "grad_norm": 2.113356590270996, + "learning_rate": 1.6193736931679276e-05, + "loss": 0.5438, + "step": 9791 + }, + { + "epoch": 0.3088471849865952, + "grad_norm": 1.7403932809829712, + "learning_rate": 1.619293487912886e-05, + "loss": 0.8686, + "step": 9792 + }, + { + "epoch": 0.3088787257530358, + "grad_norm": 2.0928711891174316, + "learning_rate": 1.6192132761950363e-05, + "loss": 0.5791, + "step": 9793 + }, + { + "epoch": 0.30891026651947645, + "grad_norm": 2.9994497299194336, + "learning_rate": 1.6191330580152148e-05, + "loss": 0.4068, + "step": 9794 + }, + { + "epoch": 0.308941807285917, + "grad_norm": 1.5387893915176392, + "learning_rate": 1.6190528333742596e-05, + "loss": 0.4927, + "step": 9795 + }, + { + "epoch": 0.30897334805235765, + "grad_norm": 2.13014554977417, + "learning_rate": 1.6189726022730068e-05, + "loss": 0.506, + "step": 9796 + }, + { + "epoch": 0.3090048888187983, + "grad_norm": 1.3743560314178467, + "learning_rate": 1.6188923647122946e-05, + "loss": 0.5678, + "step": 9797 + }, + { + "epoch": 0.3090364295852389, + "grad_norm": 4.615204334259033, + "learning_rate": 1.61881212069296e-05, + "loss": 0.5023, + "step": 9798 + }, + { + "epoch": 0.30906797035167954, + "grad_norm": 1.9976918697357178, + "learning_rate": 1.6187318702158406e-05, + "loss": 0.5411, + "step": 9799 + }, + { + "epoch": 0.3090995111181202, + "grad_norm": 1.9505752325057983, + "learning_rate": 1.6186516132817733e-05, + "loss": 0.5396, + "step": 9800 + }, + { + "epoch": 0.3091310518845608, + "grad_norm": 1.5710467100143433, + "learning_rate": 1.6185713498915964e-05, + "loss": 0.5229, + "step": 9801 + }, + { + "epoch": 0.30916259265100143, + "grad_norm": 1.6339911222457886, + "learning_rate": 1.6184910800461472e-05, + "loss": 0.5335, + "step": 9802 + }, + { + "epoch": 0.30919413341744206, + "grad_norm": 1.732454538345337, + "learning_rate": 1.618410803746263e-05, + "loss": 0.5306, + "step": 9803 + }, + { + "epoch": 0.3092256741838827, + "grad_norm": 1.5631489753723145, + "learning_rate": 1.618330520992782e-05, + "loss": 0.5049, + "step": 9804 + }, + { + "epoch": 0.30925721495032327, + "grad_norm": 2.0824146270751953, + "learning_rate": 1.6182502317865422e-05, + "loss": 0.5686, + "step": 9805 + }, + { + "epoch": 0.3092887557167639, + "grad_norm": 1.8206058740615845, + "learning_rate": 1.6181699361283806e-05, + "loss": 0.87, + "step": 9806 + }, + { + "epoch": 0.3093202964832045, + "grad_norm": 1.9106383323669434, + "learning_rate": 1.6180896340191364e-05, + "loss": 0.5545, + "step": 9807 + }, + { + "epoch": 0.30935183724964516, + "grad_norm": 1.675495982170105, + "learning_rate": 1.6180093254596463e-05, + "loss": 0.5409, + "step": 9808 + }, + { + "epoch": 0.3093833780160858, + "grad_norm": 1.912187099456787, + "learning_rate": 1.6179290104507493e-05, + "loss": 0.5545, + "step": 9809 + }, + { + "epoch": 0.3094149187825264, + "grad_norm": 2.244285821914673, + "learning_rate": 1.6178486889932834e-05, + "loss": 0.5668, + "step": 9810 + }, + { + "epoch": 0.30944645954896705, + "grad_norm": 2.2525858879089355, + "learning_rate": 1.6177683610880866e-05, + "loss": 0.5504, + "step": 9811 + }, + { + "epoch": 0.3094780003154077, + "grad_norm": 1.437774419784546, + "learning_rate": 1.6176880267359973e-05, + "loss": 0.5394, + "step": 9812 + }, + { + "epoch": 0.3095095410818483, + "grad_norm": 1.3102452754974365, + "learning_rate": 1.6176076859378534e-05, + "loss": 0.4092, + "step": 9813 + }, + { + "epoch": 0.30954108184828893, + "grad_norm": 1.6020011901855469, + "learning_rate": 1.6175273386944942e-05, + "loss": 0.5374, + "step": 9814 + }, + { + "epoch": 0.30957262261472956, + "grad_norm": 1.9556185007095337, + "learning_rate": 1.617446985006758e-05, + "loss": 0.8672, + "step": 9815 + }, + { + "epoch": 0.30960416338117014, + "grad_norm": 2.0842678546905518, + "learning_rate": 1.617366624875482e-05, + "loss": 0.435, + "step": 9816 + }, + { + "epoch": 0.30963570414761077, + "grad_norm": 1.7192326784133911, + "learning_rate": 1.6172862583015066e-05, + "loss": 0.5181, + "step": 9817 + }, + { + "epoch": 0.3096672449140514, + "grad_norm": 3.125339984893799, + "learning_rate": 1.61720588528567e-05, + "loss": 0.5316, + "step": 9818 + }, + { + "epoch": 0.30969878568049203, + "grad_norm": 1.6663014888763428, + "learning_rate": 1.6171255058288107e-05, + "loss": 0.8417, + "step": 9819 + }, + { + "epoch": 0.30973032644693266, + "grad_norm": 1.903653621673584, + "learning_rate": 1.617045119931767e-05, + "loss": 0.5373, + "step": 9820 + }, + { + "epoch": 0.3097618672133733, + "grad_norm": 1.3850337266921997, + "learning_rate": 1.616964727595379e-05, + "loss": 0.5287, + "step": 9821 + }, + { + "epoch": 0.3097934079798139, + "grad_norm": 1.7239091396331787, + "learning_rate": 1.6168843288204845e-05, + "loss": 0.547, + "step": 9822 + }, + { + "epoch": 0.30982494874625455, + "grad_norm": 1.6568347215652466, + "learning_rate": 1.6168039236079234e-05, + "loss": 0.8729, + "step": 9823 + }, + { + "epoch": 0.3098564895126952, + "grad_norm": 1.7978942394256592, + "learning_rate": 1.6167235119585344e-05, + "loss": 0.5146, + "step": 9824 + }, + { + "epoch": 0.3098880302791358, + "grad_norm": 1.5796325206756592, + "learning_rate": 1.6166430938731562e-05, + "loss": 0.5225, + "step": 9825 + }, + { + "epoch": 0.3099195710455764, + "grad_norm": 1.7043921947479248, + "learning_rate": 1.6165626693526292e-05, + "loss": 0.5066, + "step": 9826 + }, + { + "epoch": 0.309951111812017, + "grad_norm": 1.8859556913375854, + "learning_rate": 1.6164822383977912e-05, + "loss": 0.5313, + "step": 9827 + }, + { + "epoch": 0.30998265257845764, + "grad_norm": 1.6471163034439087, + "learning_rate": 1.616401801009483e-05, + "loss": 0.5258, + "step": 9828 + }, + { + "epoch": 0.31001419334489827, + "grad_norm": 1.6027603149414062, + "learning_rate": 1.616321357188543e-05, + "loss": 0.5345, + "step": 9829 + }, + { + "epoch": 0.3100457341113389, + "grad_norm": 2.1535627841949463, + "learning_rate": 1.6162409069358115e-05, + "loss": 0.5509, + "step": 9830 + }, + { + "epoch": 0.31007727487777953, + "grad_norm": 1.7958723306655884, + "learning_rate": 1.6161604502521275e-05, + "loss": 0.4952, + "step": 9831 + }, + { + "epoch": 0.31010881564422016, + "grad_norm": 1.5869446992874146, + "learning_rate": 1.6160799871383305e-05, + "loss": 0.5098, + "step": 9832 + }, + { + "epoch": 0.3101403564106608, + "grad_norm": 1.5954331159591675, + "learning_rate": 1.6159995175952607e-05, + "loss": 0.5302, + "step": 9833 + }, + { + "epoch": 0.3101718971771014, + "grad_norm": 2.0531415939331055, + "learning_rate": 1.6159190416237573e-05, + "loss": 0.5141, + "step": 9834 + }, + { + "epoch": 0.31020343794354205, + "grad_norm": 1.7724895477294922, + "learning_rate": 1.615838559224661e-05, + "loss": 0.546, + "step": 9835 + }, + { + "epoch": 0.3102349787099827, + "grad_norm": 2.3646304607391357, + "learning_rate": 1.6157580703988105e-05, + "loss": 0.5275, + "step": 9836 + }, + { + "epoch": 0.31026651947642325, + "grad_norm": 1.7624218463897705, + "learning_rate": 1.615677575147047e-05, + "loss": 0.5291, + "step": 9837 + }, + { + "epoch": 0.3102980602428639, + "grad_norm": 1.6653330326080322, + "learning_rate": 1.6155970734702098e-05, + "loss": 0.4134, + "step": 9838 + }, + { + "epoch": 0.3103296010093045, + "grad_norm": 2.026078939437866, + "learning_rate": 1.615516565369139e-05, + "loss": 0.5451, + "step": 9839 + }, + { + "epoch": 0.31036114177574514, + "grad_norm": 1.919809341430664, + "learning_rate": 1.6154360508446746e-05, + "loss": 0.5358, + "step": 9840 + }, + { + "epoch": 0.31039268254218577, + "grad_norm": 1.9585262537002563, + "learning_rate": 1.6153555298976577e-05, + "loss": 0.5133, + "step": 9841 + }, + { + "epoch": 0.3104242233086264, + "grad_norm": 1.694050908088684, + "learning_rate": 1.615275002528928e-05, + "loss": 0.8802, + "step": 9842 + }, + { + "epoch": 0.31045576407506703, + "grad_norm": 1.8207268714904785, + "learning_rate": 1.6151944687393255e-05, + "loss": 0.5045, + "step": 9843 + }, + { + "epoch": 0.31048730484150766, + "grad_norm": 1.6126223802566528, + "learning_rate": 1.615113928529691e-05, + "loss": 0.4016, + "step": 9844 + }, + { + "epoch": 0.3105188456079483, + "grad_norm": 1.7980225086212158, + "learning_rate": 1.6150333819008656e-05, + "loss": 0.5567, + "step": 9845 + }, + { + "epoch": 0.3105503863743889, + "grad_norm": 2.5401954650878906, + "learning_rate": 1.614952828853689e-05, + "loss": 0.8727, + "step": 9846 + }, + { + "epoch": 0.3105819271408295, + "grad_norm": 1.6545522212982178, + "learning_rate": 1.614872269389002e-05, + "loss": 0.5304, + "step": 9847 + }, + { + "epoch": 0.3106134679072701, + "grad_norm": 1.7353864908218384, + "learning_rate": 1.6147917035076456e-05, + "loss": 0.5299, + "step": 9848 + }, + { + "epoch": 0.31064500867371075, + "grad_norm": 1.8418445587158203, + "learning_rate": 1.61471113121046e-05, + "loss": 0.5479, + "step": 9849 + }, + { + "epoch": 0.3106765494401514, + "grad_norm": 1.449502944946289, + "learning_rate": 1.6146305524982867e-05, + "loss": 0.5574, + "step": 9850 + }, + { + "epoch": 0.310708090206592, + "grad_norm": 1.9276397228240967, + "learning_rate": 1.6145499673719666e-05, + "loss": 0.5234, + "step": 9851 + }, + { + "epoch": 0.31073963097303264, + "grad_norm": 1.6397285461425781, + "learning_rate": 1.6144693758323402e-05, + "loss": 0.5383, + "step": 9852 + }, + { + "epoch": 0.3107711717394733, + "grad_norm": 1.6530390977859497, + "learning_rate": 1.614388777880249e-05, + "loss": 0.5115, + "step": 9853 + }, + { + "epoch": 0.3108027125059139, + "grad_norm": 3.9742400646209717, + "learning_rate": 1.6143081735165336e-05, + "loss": 0.917, + "step": 9854 + }, + { + "epoch": 0.31083425327235453, + "grad_norm": 5.188697814941406, + "learning_rate": 1.614227562742035e-05, + "loss": 0.5385, + "step": 9855 + }, + { + "epoch": 0.31086579403879516, + "grad_norm": 1.7457276582717896, + "learning_rate": 1.6141469455575956e-05, + "loss": 0.5541, + "step": 9856 + }, + { + "epoch": 0.3108973348052358, + "grad_norm": 1.635706901550293, + "learning_rate": 1.6140663219640554e-05, + "loss": 0.5444, + "step": 9857 + }, + { + "epoch": 0.31092887557167637, + "grad_norm": 3.3759844303131104, + "learning_rate": 1.6139856919622564e-05, + "loss": 0.5187, + "step": 9858 + }, + { + "epoch": 0.310960416338117, + "grad_norm": 13.692225456237793, + "learning_rate": 1.6139050555530403e-05, + "loss": 0.4182, + "step": 9859 + }, + { + "epoch": 0.3109919571045576, + "grad_norm": 2.7580015659332275, + "learning_rate": 1.6138244127372482e-05, + "loss": 0.5476, + "step": 9860 + }, + { + "epoch": 0.31102349787099826, + "grad_norm": 1.901973009109497, + "learning_rate": 1.6137437635157214e-05, + "loss": 0.4824, + "step": 9861 + }, + { + "epoch": 0.3110550386374389, + "grad_norm": 2.3134498596191406, + "learning_rate": 1.613663107889302e-05, + "loss": 0.5398, + "step": 9862 + }, + { + "epoch": 0.3110865794038795, + "grad_norm": 2.762723684310913, + "learning_rate": 1.6135824458588318e-05, + "loss": 0.5773, + "step": 9863 + }, + { + "epoch": 0.31111812017032014, + "grad_norm": 1.7676337957382202, + "learning_rate": 1.613501777425152e-05, + "loss": 0.5169, + "step": 9864 + }, + { + "epoch": 0.3111496609367608, + "grad_norm": 1.696390151977539, + "learning_rate": 1.6134211025891048e-05, + "loss": 0.5332, + "step": 9865 + }, + { + "epoch": 0.3111812017032014, + "grad_norm": 1.861243486404419, + "learning_rate": 1.6133404213515323e-05, + "loss": 0.3971, + "step": 9866 + }, + { + "epoch": 0.31121274246964203, + "grad_norm": 2.1392829418182373, + "learning_rate": 1.6132597337132764e-05, + "loss": 0.5071, + "step": 9867 + }, + { + "epoch": 0.31124428323608266, + "grad_norm": 1.6950633525848389, + "learning_rate": 1.6131790396751784e-05, + "loss": 0.5361, + "step": 9868 + }, + { + "epoch": 0.31127582400252324, + "grad_norm": 1.8814095258712769, + "learning_rate": 1.6130983392380815e-05, + "loss": 0.5069, + "step": 9869 + }, + { + "epoch": 0.31130736476896387, + "grad_norm": 2.045262575149536, + "learning_rate": 1.6130176324028274e-05, + "loss": 0.5405, + "step": 9870 + }, + { + "epoch": 0.3113389055354045, + "grad_norm": 2.378178119659424, + "learning_rate": 1.612936919170258e-05, + "loss": 0.5754, + "step": 9871 + }, + { + "epoch": 0.3113704463018451, + "grad_norm": 1.818124771118164, + "learning_rate": 1.612856199541216e-05, + "loss": 0.534, + "step": 9872 + }, + { + "epoch": 0.31140198706828576, + "grad_norm": 2.449275493621826, + "learning_rate": 1.612775473516544e-05, + "loss": 0.853, + "step": 9873 + }, + { + "epoch": 0.3114335278347264, + "grad_norm": 1.873281717300415, + "learning_rate": 1.6126947410970837e-05, + "loss": 0.5793, + "step": 9874 + }, + { + "epoch": 0.311465068601167, + "grad_norm": 2.3533382415771484, + "learning_rate": 1.612614002283678e-05, + "loss": 0.5494, + "step": 9875 + }, + { + "epoch": 0.31149660936760765, + "grad_norm": 1.6320351362228394, + "learning_rate": 1.6125332570771702e-05, + "loss": 0.5483, + "step": 9876 + }, + { + "epoch": 0.3115281501340483, + "grad_norm": 2.3757519721984863, + "learning_rate": 1.6124525054784017e-05, + "loss": 0.5339, + "step": 9877 + }, + { + "epoch": 0.3115596909004889, + "grad_norm": 1.402794599533081, + "learning_rate": 1.6123717474882157e-05, + "loss": 0.8851, + "step": 9878 + }, + { + "epoch": 0.3115912316669295, + "grad_norm": 1.1646391153335571, + "learning_rate": 1.612290983107455e-05, + "loss": 0.3757, + "step": 9879 + }, + { + "epoch": 0.3116227724333701, + "grad_norm": 2.091095209121704, + "learning_rate": 1.612210212336963e-05, + "loss": 0.5408, + "step": 9880 + }, + { + "epoch": 0.31165431319981074, + "grad_norm": 1.7937294244766235, + "learning_rate": 1.6121294351775813e-05, + "loss": 0.8445, + "step": 9881 + }, + { + "epoch": 0.31168585396625137, + "grad_norm": 1.6342824697494507, + "learning_rate": 1.612048651630154e-05, + "loss": 0.5361, + "step": 9882 + }, + { + "epoch": 0.311717394732692, + "grad_norm": 1.6648621559143066, + "learning_rate": 1.611967861695524e-05, + "loss": 0.5678, + "step": 9883 + }, + { + "epoch": 0.31174893549913263, + "grad_norm": 1.4336254596710205, + "learning_rate": 1.6118870653745338e-05, + "loss": 0.837, + "step": 9884 + }, + { + "epoch": 0.31178047626557326, + "grad_norm": 2.375343084335327, + "learning_rate": 1.611806262668027e-05, + "loss": 0.5389, + "step": 9885 + }, + { + "epoch": 0.3118120170320139, + "grad_norm": 1.3234533071517944, + "learning_rate": 1.611725453576847e-05, + "loss": 0.5365, + "step": 9886 + }, + { + "epoch": 0.3118435577984545, + "grad_norm": 1.913821816444397, + "learning_rate": 1.611644638101837e-05, + "loss": 0.5421, + "step": 9887 + }, + { + "epoch": 0.31187509856489515, + "grad_norm": 1.6825366020202637, + "learning_rate": 1.6115638162438398e-05, + "loss": 0.5616, + "step": 9888 + }, + { + "epoch": 0.3119066393313358, + "grad_norm": 1.7106417417526245, + "learning_rate": 1.6114829880036996e-05, + "loss": 0.5437, + "step": 9889 + }, + { + "epoch": 0.31193818009777635, + "grad_norm": 1.8070602416992188, + "learning_rate": 1.6114021533822595e-05, + "loss": 0.5463, + "step": 9890 + }, + { + "epoch": 0.311969720864217, + "grad_norm": 1.4175695180892944, + "learning_rate": 1.6113213123803632e-05, + "loss": 0.5479, + "step": 9891 + }, + { + "epoch": 0.3120012616306576, + "grad_norm": 1.6700314283370972, + "learning_rate": 1.6112404649988546e-05, + "loss": 0.4709, + "step": 9892 + }, + { + "epoch": 0.31203280239709824, + "grad_norm": 2.546457290649414, + "learning_rate": 1.611159611238577e-05, + "loss": 0.5449, + "step": 9893 + }, + { + "epoch": 0.31206434316353887, + "grad_norm": 1.4849793910980225, + "learning_rate": 1.611078751100374e-05, + "loss": 0.5136, + "step": 9894 + }, + { + "epoch": 0.3120958839299795, + "grad_norm": 1.3965966701507568, + "learning_rate": 1.6109978845850902e-05, + "loss": 0.5523, + "step": 9895 + }, + { + "epoch": 0.31212742469642013, + "grad_norm": 1.6340019702911377, + "learning_rate": 1.610917011693569e-05, + "loss": 0.5361, + "step": 9896 + }, + { + "epoch": 0.31215896546286076, + "grad_norm": 1.3458307981491089, + "learning_rate": 1.6108361324266543e-05, + "loss": 0.5582, + "step": 9897 + }, + { + "epoch": 0.3121905062293014, + "grad_norm": 1.5787237882614136, + "learning_rate": 1.61075524678519e-05, + "loss": 0.5193, + "step": 9898 + }, + { + "epoch": 0.312222046995742, + "grad_norm": 1.4424140453338623, + "learning_rate": 1.6106743547700205e-05, + "loss": 0.5453, + "step": 9899 + }, + { + "epoch": 0.3122535877621826, + "grad_norm": 1.7170906066894531, + "learning_rate": 1.6105934563819903e-05, + "loss": 0.3984, + "step": 9900 + }, + { + "epoch": 0.3122851285286232, + "grad_norm": 1.9890291690826416, + "learning_rate": 1.610512551621943e-05, + "loss": 0.5122, + "step": 9901 + }, + { + "epoch": 0.31231666929506385, + "grad_norm": 2.1856942176818848, + "learning_rate": 1.6104316404907236e-05, + "loss": 0.5293, + "step": 9902 + }, + { + "epoch": 0.3123482100615045, + "grad_norm": 1.6288573741912842, + "learning_rate": 1.6103507229891754e-05, + "loss": 0.5171, + "step": 9903 + }, + { + "epoch": 0.3123797508279451, + "grad_norm": 1.2965062856674194, + "learning_rate": 1.6102697991181437e-05, + "loss": 0.5414, + "step": 9904 + }, + { + "epoch": 0.31241129159438574, + "grad_norm": 2.3376734256744385, + "learning_rate": 1.610188868878473e-05, + "loss": 0.515, + "step": 9905 + }, + { + "epoch": 0.3124428323608264, + "grad_norm": 1.4180909395217896, + "learning_rate": 1.6101079322710073e-05, + "loss": 0.5098, + "step": 9906 + }, + { + "epoch": 0.312474373127267, + "grad_norm": 2.6275737285614014, + "learning_rate": 1.610026989296592e-05, + "loss": 0.8698, + "step": 9907 + }, + { + "epoch": 0.31250591389370763, + "grad_norm": 1.4835916757583618, + "learning_rate": 1.6099460399560713e-05, + "loss": 0.5457, + "step": 9908 + }, + { + "epoch": 0.31253745466014826, + "grad_norm": 1.9917758703231812, + "learning_rate": 1.6098650842502895e-05, + "loss": 0.5189, + "step": 9909 + }, + { + "epoch": 0.3125689954265889, + "grad_norm": 1.685550570487976, + "learning_rate": 1.6097841221800926e-05, + "loss": 0.5097, + "step": 9910 + }, + { + "epoch": 0.31260053619302947, + "grad_norm": 1.6196272373199463, + "learning_rate": 1.609703153746325e-05, + "loss": 0.5552, + "step": 9911 + }, + { + "epoch": 0.3126320769594701, + "grad_norm": 1.3874495029449463, + "learning_rate": 1.6096221789498313e-05, + "loss": 0.5495, + "step": 9912 + }, + { + "epoch": 0.3126636177259107, + "grad_norm": 3.060136318206787, + "learning_rate": 1.609541197791457e-05, + "loss": 0.5284, + "step": 9913 + }, + { + "epoch": 0.31269515849235135, + "grad_norm": 1.8676588535308838, + "learning_rate": 1.6094602102720464e-05, + "loss": 0.528, + "step": 9914 + }, + { + "epoch": 0.312726699258792, + "grad_norm": 2.067026376724243, + "learning_rate": 1.6093792163924456e-05, + "loss": 0.5099, + "step": 9915 + }, + { + "epoch": 0.3127582400252326, + "grad_norm": 1.9116630554199219, + "learning_rate": 1.6092982161534998e-05, + "loss": 0.522, + "step": 9916 + }, + { + "epoch": 0.31278978079167324, + "grad_norm": 2.1364595890045166, + "learning_rate": 1.6092172095560538e-05, + "loss": 0.5131, + "step": 9917 + }, + { + "epoch": 0.3128213215581139, + "grad_norm": 2.655447244644165, + "learning_rate": 1.6091361966009533e-05, + "loss": 0.5388, + "step": 9918 + }, + { + "epoch": 0.3128528623245545, + "grad_norm": 1.6325984001159668, + "learning_rate": 1.6090551772890433e-05, + "loss": 0.5094, + "step": 9919 + }, + { + "epoch": 0.31288440309099513, + "grad_norm": 1.6711046695709229, + "learning_rate": 1.60897415162117e-05, + "loss": 0.5122, + "step": 9920 + }, + { + "epoch": 0.31291594385743576, + "grad_norm": 2.6321520805358887, + "learning_rate": 1.6088931195981783e-05, + "loss": 0.5419, + "step": 9921 + }, + { + "epoch": 0.31294748462387634, + "grad_norm": 1.9851678609848022, + "learning_rate": 1.608812081220914e-05, + "loss": 0.5057, + "step": 9922 + }, + { + "epoch": 0.31297902539031697, + "grad_norm": 1.6802725791931152, + "learning_rate": 1.6087310364902232e-05, + "loss": 0.5283, + "step": 9923 + }, + { + "epoch": 0.3130105661567576, + "grad_norm": 2.025886297225952, + "learning_rate": 1.6086499854069514e-05, + "loss": 0.4955, + "step": 9924 + }, + { + "epoch": 0.3130421069231982, + "grad_norm": 2.2668628692626953, + "learning_rate": 1.608568927971944e-05, + "loss": 0.5528, + "step": 9925 + }, + { + "epoch": 0.31307364768963886, + "grad_norm": 1.9087249040603638, + "learning_rate": 1.6084878641860475e-05, + "loss": 0.5431, + "step": 9926 + }, + { + "epoch": 0.3131051884560795, + "grad_norm": 1.5330452919006348, + "learning_rate": 1.6084067940501075e-05, + "loss": 0.4882, + "step": 9927 + }, + { + "epoch": 0.3131367292225201, + "grad_norm": 1.4225118160247803, + "learning_rate": 1.6083257175649708e-05, + "loss": 0.4995, + "step": 9928 + }, + { + "epoch": 0.31316826998896075, + "grad_norm": 1.4816062450408936, + "learning_rate": 1.608244634731482e-05, + "loss": 0.5512, + "step": 9929 + }, + { + "epoch": 0.3131998107554014, + "grad_norm": 2.1904141902923584, + "learning_rate": 1.6081635455504887e-05, + "loss": 0.876, + "step": 9930 + }, + { + "epoch": 0.313231351521842, + "grad_norm": 1.9398926496505737, + "learning_rate": 1.6080824500228367e-05, + "loss": 0.5427, + "step": 9931 + }, + { + "epoch": 0.3132628922882826, + "grad_norm": 1.9630126953125, + "learning_rate": 1.6080013481493718e-05, + "loss": 0.564, + "step": 9932 + }, + { + "epoch": 0.3132944330547232, + "grad_norm": 2.1636695861816406, + "learning_rate": 1.607920239930941e-05, + "loss": 0.4156, + "step": 9933 + }, + { + "epoch": 0.31332597382116384, + "grad_norm": 1.9496387243270874, + "learning_rate": 1.6078391253683905e-05, + "loss": 0.4058, + "step": 9934 + }, + { + "epoch": 0.31335751458760447, + "grad_norm": 1.5515156984329224, + "learning_rate": 1.6077580044625665e-05, + "loss": 0.8776, + "step": 9935 + }, + { + "epoch": 0.3133890553540451, + "grad_norm": 1.5113877058029175, + "learning_rate": 1.607676877214316e-05, + "loss": 0.5351, + "step": 9936 + }, + { + "epoch": 0.31342059612048573, + "grad_norm": 1.537325143814087, + "learning_rate": 1.607595743624485e-05, + "loss": 0.5524, + "step": 9937 + }, + { + "epoch": 0.31345213688692636, + "grad_norm": 1.613961935043335, + "learning_rate": 1.6075146036939212e-05, + "loss": 0.5254, + "step": 9938 + }, + { + "epoch": 0.313483677653367, + "grad_norm": 1.3782916069030762, + "learning_rate": 1.6074334574234707e-05, + "loss": 0.5379, + "step": 9939 + }, + { + "epoch": 0.3135152184198076, + "grad_norm": 1.5734760761260986, + "learning_rate": 1.60735230481398e-05, + "loss": 0.5359, + "step": 9940 + }, + { + "epoch": 0.31354675918624825, + "grad_norm": 1.5507276058197021, + "learning_rate": 1.6072711458662967e-05, + "loss": 0.5358, + "step": 9941 + }, + { + "epoch": 0.3135782999526889, + "grad_norm": 1.4744608402252197, + "learning_rate": 1.6071899805812677e-05, + "loss": 0.5254, + "step": 9942 + }, + { + "epoch": 0.31360984071912945, + "grad_norm": 1.3650579452514648, + "learning_rate": 1.6071088089597397e-05, + "loss": 0.5133, + "step": 9943 + }, + { + "epoch": 0.3136413814855701, + "grad_norm": 13.704471588134766, + "learning_rate": 1.607027631002559e-05, + "loss": 0.5523, + "step": 9944 + }, + { + "epoch": 0.3136729222520107, + "grad_norm": 2.4363691806793213, + "learning_rate": 1.6069464467105745e-05, + "loss": 0.4771, + "step": 9945 + }, + { + "epoch": 0.31370446301845134, + "grad_norm": 1.504418969154358, + "learning_rate": 1.6068652560846328e-05, + "loss": 0.5369, + "step": 9946 + }, + { + "epoch": 0.31373600378489197, + "grad_norm": 2.126922130584717, + "learning_rate": 1.6067840591255802e-05, + "loss": 0.5445, + "step": 9947 + }, + { + "epoch": 0.3137675445513326, + "grad_norm": 1.9465303421020508, + "learning_rate": 1.6067028558342652e-05, + "loss": 0.5278, + "step": 9948 + }, + { + "epoch": 0.31379908531777323, + "grad_norm": 1.4572057723999023, + "learning_rate": 1.6066216462115347e-05, + "loss": 0.5259, + "step": 9949 + }, + { + "epoch": 0.31383062608421386, + "grad_norm": 2.2455317974090576, + "learning_rate": 1.6065404302582362e-05, + "loss": 0.5362, + "step": 9950 + }, + { + "epoch": 0.3138621668506545, + "grad_norm": 1.8257255554199219, + "learning_rate": 1.6064592079752174e-05, + "loss": 0.5044, + "step": 9951 + }, + { + "epoch": 0.3138937076170951, + "grad_norm": 7.075316905975342, + "learning_rate": 1.606377979363326e-05, + "loss": 0.5131, + "step": 9952 + }, + { + "epoch": 0.3139252483835357, + "grad_norm": 1.6777900457382202, + "learning_rate": 1.606296744423409e-05, + "loss": 0.5253, + "step": 9953 + }, + { + "epoch": 0.3139567891499763, + "grad_norm": 1.450944185256958, + "learning_rate": 1.6062155031563155e-05, + "loss": 0.5327, + "step": 9954 + }, + { + "epoch": 0.31398832991641695, + "grad_norm": 1.6648468971252441, + "learning_rate": 1.6061342555628924e-05, + "loss": 0.5313, + "step": 9955 + }, + { + "epoch": 0.3140198706828576, + "grad_norm": 1.5460721254348755, + "learning_rate": 1.606053001643987e-05, + "loss": 0.5327, + "step": 9956 + }, + { + "epoch": 0.3140514114492982, + "grad_norm": 1.499727487564087, + "learning_rate": 1.6059717414004485e-05, + "loss": 0.409, + "step": 9957 + }, + { + "epoch": 0.31408295221573884, + "grad_norm": 2.213042736053467, + "learning_rate": 1.6058904748331244e-05, + "loss": 0.8855, + "step": 9958 + }, + { + "epoch": 0.31411449298217947, + "grad_norm": 1.8905161619186401, + "learning_rate": 1.6058092019428625e-05, + "loss": 0.5526, + "step": 9959 + }, + { + "epoch": 0.3141460337486201, + "grad_norm": 1.6418633460998535, + "learning_rate": 1.6057279227305112e-05, + "loss": 0.5561, + "step": 9960 + }, + { + "epoch": 0.31417757451506073, + "grad_norm": 1.5498093366622925, + "learning_rate": 1.6056466371969187e-05, + "loss": 0.5241, + "step": 9961 + }, + { + "epoch": 0.31420911528150136, + "grad_norm": 1.1852842569351196, + "learning_rate": 1.6055653453429335e-05, + "loss": 0.4069, + "step": 9962 + }, + { + "epoch": 0.314240656047942, + "grad_norm": 1.8086433410644531, + "learning_rate": 1.6054840471694034e-05, + "loss": 0.8955, + "step": 9963 + }, + { + "epoch": 0.31427219681438257, + "grad_norm": 1.2938225269317627, + "learning_rate": 1.605402742677177e-05, + "loss": 0.5342, + "step": 9964 + }, + { + "epoch": 0.3143037375808232, + "grad_norm": 2.6815185546875, + "learning_rate": 1.6053214318671033e-05, + "loss": 0.4536, + "step": 9965 + }, + { + "epoch": 0.3143352783472638, + "grad_norm": 1.346151351928711, + "learning_rate": 1.6052401147400303e-05, + "loss": 0.5222, + "step": 9966 + }, + { + "epoch": 0.31436681911370445, + "grad_norm": 1.64398193359375, + "learning_rate": 1.6051587912968067e-05, + "loss": 0.5191, + "step": 9967 + }, + { + "epoch": 0.3143983598801451, + "grad_norm": 1.2710930109024048, + "learning_rate": 1.605077461538281e-05, + "loss": 0.5632, + "step": 9968 + }, + { + "epoch": 0.3144299006465857, + "grad_norm": 1.9160497188568115, + "learning_rate": 1.6049961254653026e-05, + "loss": 0.539, + "step": 9969 + }, + { + "epoch": 0.31446144141302634, + "grad_norm": 1.3068512678146362, + "learning_rate": 1.6049147830787195e-05, + "loss": 0.5331, + "step": 9970 + }, + { + "epoch": 0.314492982179467, + "grad_norm": 1.3061891794204712, + "learning_rate": 1.6048334343793814e-05, + "loss": 0.5191, + "step": 9971 + }, + { + "epoch": 0.3145245229459076, + "grad_norm": 1.3016866445541382, + "learning_rate": 1.604752079368136e-05, + "loss": 0.4984, + "step": 9972 + }, + { + "epoch": 0.31455606371234823, + "grad_norm": 1.4688955545425415, + "learning_rate": 1.6046707180458336e-05, + "loss": 0.5466, + "step": 9973 + }, + { + "epoch": 0.3145876044787888, + "grad_norm": 1.4336193799972534, + "learning_rate": 1.6045893504133227e-05, + "loss": 0.5101, + "step": 9974 + }, + { + "epoch": 0.31461914524522944, + "grad_norm": 1.5340967178344727, + "learning_rate": 1.6045079764714527e-05, + "loss": 0.5396, + "step": 9975 + }, + { + "epoch": 0.31465068601167007, + "grad_norm": 1.7634589672088623, + "learning_rate": 1.6044265962210725e-05, + "loss": 0.5194, + "step": 9976 + }, + { + "epoch": 0.3146822267781107, + "grad_norm": 1.6346538066864014, + "learning_rate": 1.604345209663031e-05, + "loss": 0.561, + "step": 9977 + }, + { + "epoch": 0.3147137675445513, + "grad_norm": 1.474086046218872, + "learning_rate": 1.604263816798178e-05, + "loss": 0.5746, + "step": 9978 + }, + { + "epoch": 0.31474530831099196, + "grad_norm": 1.5059961080551147, + "learning_rate": 1.6041824176273633e-05, + "loss": 0.4992, + "step": 9979 + }, + { + "epoch": 0.3147768490774326, + "grad_norm": 2.2876334190368652, + "learning_rate": 1.604101012151436e-05, + "loss": 0.843, + "step": 9980 + }, + { + "epoch": 0.3148083898438732, + "grad_norm": 1.709599494934082, + "learning_rate": 1.604019600371245e-05, + "loss": 0.5636, + "step": 9981 + }, + { + "epoch": 0.31483993061031385, + "grad_norm": 1.3204818964004517, + "learning_rate": 1.603938182287641e-05, + "loss": 0.5734, + "step": 9982 + }, + { + "epoch": 0.3148714713767545, + "grad_norm": 3.265292167663574, + "learning_rate": 1.6038567579014727e-05, + "loss": 0.519, + "step": 9983 + }, + { + "epoch": 0.3149030121431951, + "grad_norm": 1.6009689569473267, + "learning_rate": 1.6037753272135903e-05, + "loss": 0.5295, + "step": 9984 + }, + { + "epoch": 0.3149345529096357, + "grad_norm": 1.5319427251815796, + "learning_rate": 1.603693890224844e-05, + "loss": 0.5227, + "step": 9985 + }, + { + "epoch": 0.3149660936760763, + "grad_norm": 1.3751393556594849, + "learning_rate": 1.603612446936083e-05, + "loss": 0.837, + "step": 9986 + }, + { + "epoch": 0.31499763444251694, + "grad_norm": 1.8132798671722412, + "learning_rate": 1.6035309973481573e-05, + "loss": 0.563, + "step": 9987 + }, + { + "epoch": 0.31502917520895757, + "grad_norm": 1.8701468706130981, + "learning_rate": 1.6034495414619165e-05, + "loss": 0.5398, + "step": 9988 + }, + { + "epoch": 0.3150607159753982, + "grad_norm": 1.4746836423873901, + "learning_rate": 1.6033680792782124e-05, + "loss": 0.5307, + "step": 9989 + }, + { + "epoch": 0.3150922567418388, + "grad_norm": 1.5824999809265137, + "learning_rate": 1.6032866107978927e-05, + "loss": 0.5449, + "step": 9990 + }, + { + "epoch": 0.31512379750827946, + "grad_norm": 1.4488881826400757, + "learning_rate": 1.6032051360218096e-05, + "loss": 0.5229, + "step": 9991 + }, + { + "epoch": 0.3151553382747201, + "grad_norm": 1.3974721431732178, + "learning_rate": 1.603123654950812e-05, + "loss": 0.5079, + "step": 9992 + }, + { + "epoch": 0.3151868790411607, + "grad_norm": 3.1313467025756836, + "learning_rate": 1.603042167585751e-05, + "loss": 0.5089, + "step": 9993 + }, + { + "epoch": 0.31521841980760135, + "grad_norm": 1.7922197580337524, + "learning_rate": 1.6029606739274767e-05, + "loss": 0.4984, + "step": 9994 + }, + { + "epoch": 0.315249960574042, + "grad_norm": 1.8886451721191406, + "learning_rate": 1.6028791739768395e-05, + "loss": 0.5418, + "step": 9995 + }, + { + "epoch": 0.31528150134048255, + "grad_norm": 1.5463557243347168, + "learning_rate": 1.6027976677346902e-05, + "loss": 0.5148, + "step": 9996 + }, + { + "epoch": 0.3153130421069232, + "grad_norm": 1.708709478378296, + "learning_rate": 1.602716155201879e-05, + "loss": 0.5234, + "step": 9997 + }, + { + "epoch": 0.3153445828733638, + "grad_norm": 1.5445854663848877, + "learning_rate": 1.6026346363792565e-05, + "loss": 0.5268, + "step": 9998 + }, + { + "epoch": 0.31537612363980444, + "grad_norm": 1.8160651922225952, + "learning_rate": 1.6025531112676736e-05, + "loss": 0.546, + "step": 9999 + }, + { + "epoch": 0.31540766440624507, + "grad_norm": 1.4117776155471802, + "learning_rate": 1.6024715798679818e-05, + "loss": 0.5197, + "step": 10000 + }, + { + "epoch": 0.3154392051726857, + "grad_norm": 1.3520290851593018, + "learning_rate": 1.6023900421810306e-05, + "loss": 0.5135, + "step": 10001 + }, + { + "epoch": 0.31547074593912633, + "grad_norm": 1.643619179725647, + "learning_rate": 1.602308498207672e-05, + "loss": 0.5338, + "step": 10002 + }, + { + "epoch": 0.31550228670556696, + "grad_norm": 1.8958909511566162, + "learning_rate": 1.6022269479487563e-05, + "loss": 0.5148, + "step": 10003 + }, + { + "epoch": 0.3155338274720076, + "grad_norm": 1.9994267225265503, + "learning_rate": 1.6021453914051346e-05, + "loss": 0.5153, + "step": 10004 + }, + { + "epoch": 0.3155653682384482, + "grad_norm": 2.0231943130493164, + "learning_rate": 1.6020638285776588e-05, + "loss": 0.8725, + "step": 10005 + }, + { + "epoch": 0.3155969090048888, + "grad_norm": 1.5068023204803467, + "learning_rate": 1.601982259467179e-05, + "loss": 0.4896, + "step": 10006 + }, + { + "epoch": 0.3156284497713294, + "grad_norm": 3.040560722351074, + "learning_rate": 1.601900684074547e-05, + "loss": 0.8574, + "step": 10007 + }, + { + "epoch": 0.31565999053777005, + "grad_norm": 2.9930825233459473, + "learning_rate": 1.6018191024006137e-05, + "loss": 0.5436, + "step": 10008 + }, + { + "epoch": 0.3156915313042107, + "grad_norm": 1.4679447412490845, + "learning_rate": 1.601737514446231e-05, + "loss": 0.5702, + "step": 10009 + }, + { + "epoch": 0.3157230720706513, + "grad_norm": 1.663633942604065, + "learning_rate": 1.6016559202122505e-05, + "loss": 0.5391, + "step": 10010 + }, + { + "epoch": 0.31575461283709194, + "grad_norm": 1.759950041770935, + "learning_rate": 1.601574319699523e-05, + "loss": 0.5353, + "step": 10011 + }, + { + "epoch": 0.31578615360353257, + "grad_norm": 1.6311259269714355, + "learning_rate": 1.6014927129089002e-05, + "loss": 0.552, + "step": 10012 + }, + { + "epoch": 0.3158176943699732, + "grad_norm": 1.3035869598388672, + "learning_rate": 1.601411099841234e-05, + "loss": 0.5123, + "step": 10013 + }, + { + "epoch": 0.31584923513641383, + "grad_norm": 1.8146532773971558, + "learning_rate": 1.601329480497376e-05, + "loss": 0.5096, + "step": 10014 + }, + { + "epoch": 0.31588077590285446, + "grad_norm": 1.4437743425369263, + "learning_rate": 1.601247854878178e-05, + "loss": 0.5463, + "step": 10015 + }, + { + "epoch": 0.3159123166692951, + "grad_norm": 1.793583869934082, + "learning_rate": 1.6011662229844915e-05, + "loss": 0.5461, + "step": 10016 + }, + { + "epoch": 0.31594385743573566, + "grad_norm": 2.5717506408691406, + "learning_rate": 1.601084584817169e-05, + "loss": 0.9087, + "step": 10017 + }, + { + "epoch": 0.3159753982021763, + "grad_norm": 1.550194263458252, + "learning_rate": 1.6010029403770618e-05, + "loss": 0.4986, + "step": 10018 + }, + { + "epoch": 0.3160069389686169, + "grad_norm": 1.5527377128601074, + "learning_rate": 1.600921289665023e-05, + "loss": 0.53, + "step": 10019 + }, + { + "epoch": 0.31603847973505755, + "grad_norm": 1.4528356790542603, + "learning_rate": 1.600839632681903e-05, + "loss": 0.5407, + "step": 10020 + }, + { + "epoch": 0.3160700205014982, + "grad_norm": 1.7708442211151123, + "learning_rate": 1.6007579694285556e-05, + "loss": 0.5369, + "step": 10021 + }, + { + "epoch": 0.3161015612679388, + "grad_norm": 1.7242860794067383, + "learning_rate": 1.600676299905832e-05, + "loss": 0.5213, + "step": 10022 + }, + { + "epoch": 0.31613310203437944, + "grad_norm": 1.911960244178772, + "learning_rate": 1.6005946241145846e-05, + "loss": 0.5399, + "step": 10023 + }, + { + "epoch": 0.3161646428008201, + "grad_norm": 1.3498660326004028, + "learning_rate": 1.6005129420556662e-05, + "loss": 0.5602, + "step": 10024 + }, + { + "epoch": 0.3161961835672607, + "grad_norm": 1.5788767337799072, + "learning_rate": 1.600431253729929e-05, + "loss": 0.5245, + "step": 10025 + }, + { + "epoch": 0.31622772433370133, + "grad_norm": 1.7538120746612549, + "learning_rate": 1.6003495591382252e-05, + "loss": 0.5421, + "step": 10026 + }, + { + "epoch": 0.3162592651001419, + "grad_norm": 1.5728963613510132, + "learning_rate": 1.600267858281408e-05, + "loss": 0.5485, + "step": 10027 + }, + { + "epoch": 0.31629080586658254, + "grad_norm": 1.6827702522277832, + "learning_rate": 1.6001861511603292e-05, + "loss": 0.5053, + "step": 10028 + }, + { + "epoch": 0.31632234663302317, + "grad_norm": 1.5608896017074585, + "learning_rate": 1.6001044377758417e-05, + "loss": 0.5314, + "step": 10029 + }, + { + "epoch": 0.3163538873994638, + "grad_norm": 1.4529364109039307, + "learning_rate": 1.6000227181287992e-05, + "loss": 0.503, + "step": 10030 + }, + { + "epoch": 0.3163854281659044, + "grad_norm": 1.917528748512268, + "learning_rate": 1.5999409922200534e-05, + "loss": 0.557, + "step": 10031 + }, + { + "epoch": 0.31641696893234506, + "grad_norm": 1.9073442220687866, + "learning_rate": 1.599859260050457e-05, + "loss": 0.5177, + "step": 10032 + }, + { + "epoch": 0.3164485096987857, + "grad_norm": 1.7133837938308716, + "learning_rate": 1.599777521620864e-05, + "loss": 0.557, + "step": 10033 + }, + { + "epoch": 0.3164800504652263, + "grad_norm": 2.2005116939544678, + "learning_rate": 1.5996957769321268e-05, + "loss": 0.5765, + "step": 10034 + }, + { + "epoch": 0.31651159123166694, + "grad_norm": 1.8326566219329834, + "learning_rate": 1.5996140259850982e-05, + "loss": 0.5488, + "step": 10035 + }, + { + "epoch": 0.3165431319981076, + "grad_norm": 1.6891448497772217, + "learning_rate": 1.599532268780632e-05, + "loss": 0.5365, + "step": 10036 + }, + { + "epoch": 0.3165746727645482, + "grad_norm": 1.5708211660385132, + "learning_rate": 1.599450505319581e-05, + "loss": 0.5479, + "step": 10037 + }, + { + "epoch": 0.3166062135309888, + "grad_norm": 1.8736727237701416, + "learning_rate": 1.5993687356027982e-05, + "loss": 0.4626, + "step": 10038 + }, + { + "epoch": 0.3166377542974294, + "grad_norm": 2.4108431339263916, + "learning_rate": 1.5992869596311377e-05, + "loss": 0.5258, + "step": 10039 + }, + { + "epoch": 0.31666929506387004, + "grad_norm": 2.024076461791992, + "learning_rate": 1.599205177405452e-05, + "loss": 0.523, + "step": 10040 + }, + { + "epoch": 0.31670083583031067, + "grad_norm": 1.757797360420227, + "learning_rate": 1.5991233889265953e-05, + "loss": 0.5558, + "step": 10041 + }, + { + "epoch": 0.3167323765967513, + "grad_norm": 2.266413450241089, + "learning_rate": 1.5990415941954207e-05, + "loss": 0.5273, + "step": 10042 + }, + { + "epoch": 0.3167639173631919, + "grad_norm": 2.1615500450134277, + "learning_rate": 1.5989597932127817e-05, + "loss": 0.8782, + "step": 10043 + }, + { + "epoch": 0.31679545812963256, + "grad_norm": 2.272050142288208, + "learning_rate": 1.5988779859795326e-05, + "loss": 0.5338, + "step": 10044 + }, + { + "epoch": 0.3168269988960732, + "grad_norm": 1.8969985246658325, + "learning_rate": 1.5987961724965267e-05, + "loss": 0.5322, + "step": 10045 + }, + { + "epoch": 0.3168585396625138, + "grad_norm": 1.9604666233062744, + "learning_rate": 1.5987143527646173e-05, + "loss": 0.5236, + "step": 10046 + }, + { + "epoch": 0.31689008042895445, + "grad_norm": 1.9041049480438232, + "learning_rate": 1.598632526784659e-05, + "loss": 0.5613, + "step": 10047 + }, + { + "epoch": 0.3169216211953951, + "grad_norm": 1.3763020038604736, + "learning_rate": 1.5985506945575057e-05, + "loss": 0.5321, + "step": 10048 + }, + { + "epoch": 0.31695316196183565, + "grad_norm": 3.3881444931030273, + "learning_rate": 1.598468856084011e-05, + "loss": 0.5265, + "step": 10049 + }, + { + "epoch": 0.3169847027282763, + "grad_norm": 1.9340547323226929, + "learning_rate": 1.598387011365029e-05, + "loss": 0.5518, + "step": 10050 + }, + { + "epoch": 0.3170162434947169, + "grad_norm": 1.646757960319519, + "learning_rate": 1.5983051604014142e-05, + "loss": 0.5379, + "step": 10051 + }, + { + "epoch": 0.31704778426115754, + "grad_norm": 1.9182554483413696, + "learning_rate": 1.59822330319402e-05, + "loss": 0.3932, + "step": 10052 + }, + { + "epoch": 0.31707932502759817, + "grad_norm": 1.584681510925293, + "learning_rate": 1.5981414397437014e-05, + "loss": 0.5327, + "step": 10053 + }, + { + "epoch": 0.3171108657940388, + "grad_norm": 3.303770065307617, + "learning_rate": 1.5980595700513124e-05, + "loss": 0.5288, + "step": 10054 + }, + { + "epoch": 0.31714240656047943, + "grad_norm": 1.7124685049057007, + "learning_rate": 1.5979776941177075e-05, + "loss": 0.5392, + "step": 10055 + }, + { + "epoch": 0.31717394732692006, + "grad_norm": 2.579712152481079, + "learning_rate": 1.5978958119437408e-05, + "loss": 0.5229, + "step": 10056 + }, + { + "epoch": 0.3172054880933607, + "grad_norm": 1.6791473627090454, + "learning_rate": 1.5978139235302668e-05, + "loss": 0.8767, + "step": 10057 + }, + { + "epoch": 0.3172370288598013, + "grad_norm": 1.8372446298599243, + "learning_rate": 1.597732028878141e-05, + "loss": 0.5283, + "step": 10058 + }, + { + "epoch": 0.3172685696262419, + "grad_norm": 1.8720502853393555, + "learning_rate": 1.5976501279882174e-05, + "loss": 0.5351, + "step": 10059 + }, + { + "epoch": 0.3173001103926825, + "grad_norm": 1.992740273475647, + "learning_rate": 1.5975682208613503e-05, + "loss": 0.8414, + "step": 10060 + }, + { + "epoch": 0.31733165115912315, + "grad_norm": 2.140758752822876, + "learning_rate": 1.5974863074983945e-05, + "loss": 0.5629, + "step": 10061 + }, + { + "epoch": 0.3173631919255638, + "grad_norm": 2.111313819885254, + "learning_rate": 1.5974043879002055e-05, + "loss": 0.5532, + "step": 10062 + }, + { + "epoch": 0.3173947326920044, + "grad_norm": 2.1533524990081787, + "learning_rate": 1.5973224620676376e-05, + "loss": 0.5398, + "step": 10063 + }, + { + "epoch": 0.31742627345844504, + "grad_norm": 1.6939624547958374, + "learning_rate": 1.597240530001546e-05, + "loss": 0.5228, + "step": 10064 + }, + { + "epoch": 0.31745781422488567, + "grad_norm": 2.1259422302246094, + "learning_rate": 1.5971585917027864e-05, + "loss": 0.4075, + "step": 10065 + }, + { + "epoch": 0.3174893549913263, + "grad_norm": 2.014450788497925, + "learning_rate": 1.5970766471722128e-05, + "loss": 0.5025, + "step": 10066 + }, + { + "epoch": 0.31752089575776693, + "grad_norm": 3.5109987258911133, + "learning_rate": 1.5969946964106802e-05, + "loss": 0.5048, + "step": 10067 + }, + { + "epoch": 0.31755243652420756, + "grad_norm": 1.7506663799285889, + "learning_rate": 1.5969127394190455e-05, + "loss": 0.5918, + "step": 10068 + }, + { + "epoch": 0.3175839772906482, + "grad_norm": 1.9276405572891235, + "learning_rate": 1.596830776198162e-05, + "loss": 0.5206, + "step": 10069 + }, + { + "epoch": 0.31761551805708876, + "grad_norm": 1.9010154008865356, + "learning_rate": 1.5967488067488864e-05, + "loss": 0.8556, + "step": 10070 + }, + { + "epoch": 0.3176470588235294, + "grad_norm": 1.8897992372512817, + "learning_rate": 1.596666831072073e-05, + "loss": 0.541, + "step": 10071 + }, + { + "epoch": 0.31767859958997, + "grad_norm": 2.53879451751709, + "learning_rate": 1.5965848491685787e-05, + "loss": 0.5442, + "step": 10072 + }, + { + "epoch": 0.31771014035641065, + "grad_norm": 1.6524507999420166, + "learning_rate": 1.596502861039258e-05, + "loss": 0.5371, + "step": 10073 + }, + { + "epoch": 0.3177416811228513, + "grad_norm": 2.2799065113067627, + "learning_rate": 1.5964208666849663e-05, + "loss": 0.3833, + "step": 10074 + }, + { + "epoch": 0.3177732218892919, + "grad_norm": 2.66758394241333, + "learning_rate": 1.5963388661065604e-05, + "loss": 0.5418, + "step": 10075 + }, + { + "epoch": 0.31780476265573254, + "grad_norm": 1.8871512413024902, + "learning_rate": 1.596256859304895e-05, + "loss": 0.5132, + "step": 10076 + }, + { + "epoch": 0.31783630342217317, + "grad_norm": 1.753023386001587, + "learning_rate": 1.5961748462808263e-05, + "loss": 0.5586, + "step": 10077 + }, + { + "epoch": 0.3178678441886138, + "grad_norm": 1.4933942556381226, + "learning_rate": 1.59609282703521e-05, + "loss": 0.8162, + "step": 10078 + }, + { + "epoch": 0.31789938495505443, + "grad_norm": 1.9718319177627563, + "learning_rate": 1.5960108015689027e-05, + "loss": 0.5414, + "step": 10079 + }, + { + "epoch": 0.317930925721495, + "grad_norm": 2.6901509761810303, + "learning_rate": 1.59592876988276e-05, + "loss": 0.5439, + "step": 10080 + }, + { + "epoch": 0.31796246648793564, + "grad_norm": 3.0585155487060547, + "learning_rate": 1.595846731977637e-05, + "loss": 0.5491, + "step": 10081 + }, + { + "epoch": 0.31799400725437627, + "grad_norm": 1.7337838411331177, + "learning_rate": 1.5957646878543915e-05, + "loss": 0.552, + "step": 10082 + }, + { + "epoch": 0.3180255480208169, + "grad_norm": 2.3282155990600586, + "learning_rate": 1.5956826375138787e-05, + "loss": 0.5337, + "step": 10083 + }, + { + "epoch": 0.3180570887872575, + "grad_norm": 1.5721726417541504, + "learning_rate": 1.595600580956955e-05, + "loss": 0.5372, + "step": 10084 + }, + { + "epoch": 0.31808862955369815, + "grad_norm": 5.14393949508667, + "learning_rate": 1.5955185181844765e-05, + "loss": 0.5115, + "step": 10085 + }, + { + "epoch": 0.3181201703201388, + "grad_norm": 2.032482385635376, + "learning_rate": 1.5954364491973e-05, + "loss": 0.5297, + "step": 10086 + }, + { + "epoch": 0.3181517110865794, + "grad_norm": 1.9715919494628906, + "learning_rate": 1.595354373996282e-05, + "loss": 0.5394, + "step": 10087 + }, + { + "epoch": 0.31818325185302004, + "grad_norm": 2.014065742492676, + "learning_rate": 1.5952722925822787e-05, + "loss": 0.5128, + "step": 10088 + }, + { + "epoch": 0.3182147926194607, + "grad_norm": 1.904327630996704, + "learning_rate": 1.595190204956147e-05, + "loss": 0.5486, + "step": 10089 + }, + { + "epoch": 0.3182463333859013, + "grad_norm": 1.4682931900024414, + "learning_rate": 1.595108111118743e-05, + "loss": 0.5492, + "step": 10090 + }, + { + "epoch": 0.3182778741523419, + "grad_norm": 1.5123023986816406, + "learning_rate": 1.595026011070924e-05, + "loss": 0.5002, + "step": 10091 + }, + { + "epoch": 0.3183094149187825, + "grad_norm": 1.8426458835601807, + "learning_rate": 1.5949439048135463e-05, + "loss": 0.5128, + "step": 10092 + }, + { + "epoch": 0.31834095568522314, + "grad_norm": 1.716080904006958, + "learning_rate": 1.594861792347467e-05, + "loss": 0.4981, + "step": 10093 + }, + { + "epoch": 0.31837249645166377, + "grad_norm": 1.8860529661178589, + "learning_rate": 1.5947796736735433e-05, + "loss": 0.5236, + "step": 10094 + }, + { + "epoch": 0.3184040372181044, + "grad_norm": 1.958280086517334, + "learning_rate": 1.5946975487926316e-05, + "loss": 0.5449, + "step": 10095 + }, + { + "epoch": 0.318435577984545, + "grad_norm": 2.1178183555603027, + "learning_rate": 1.5946154177055894e-05, + "loss": 0.5034, + "step": 10096 + }, + { + "epoch": 0.31846711875098566, + "grad_norm": 1.8337022066116333, + "learning_rate": 1.5945332804132736e-05, + "loss": 0.5246, + "step": 10097 + }, + { + "epoch": 0.3184986595174263, + "grad_norm": 3.244900941848755, + "learning_rate": 1.5944511369165415e-05, + "loss": 0.5506, + "step": 10098 + }, + { + "epoch": 0.3185302002838669, + "grad_norm": 1.7475664615631104, + "learning_rate": 1.59436898721625e-05, + "loss": 0.5687, + "step": 10099 + }, + { + "epoch": 0.31856174105030755, + "grad_norm": 1.4913749694824219, + "learning_rate": 1.5942868313132565e-05, + "loss": 0.4797, + "step": 10100 + }, + { + "epoch": 0.3185932818167481, + "grad_norm": 1.7552131414413452, + "learning_rate": 1.5942046692084187e-05, + "loss": 0.5526, + "step": 10101 + }, + { + "epoch": 0.31862482258318875, + "grad_norm": 1.788928747177124, + "learning_rate": 1.5941225009025936e-05, + "loss": 0.4997, + "step": 10102 + }, + { + "epoch": 0.3186563633496294, + "grad_norm": 1.3724929094314575, + "learning_rate": 1.594040326396639e-05, + "loss": 0.5223, + "step": 10103 + }, + { + "epoch": 0.31868790411607, + "grad_norm": 1.7684816122055054, + "learning_rate": 1.5939581456914123e-05, + "loss": 0.4891, + "step": 10104 + }, + { + "epoch": 0.31871944488251064, + "grad_norm": 1.8491030931472778, + "learning_rate": 1.5938759587877713e-05, + "loss": 0.5198, + "step": 10105 + }, + { + "epoch": 0.31875098564895127, + "grad_norm": 1.4924544095993042, + "learning_rate": 1.5937937656865733e-05, + "loss": 0.5167, + "step": 10106 + }, + { + "epoch": 0.3187825264153919, + "grad_norm": 1.6872442960739136, + "learning_rate": 1.5937115663886763e-05, + "loss": 0.5368, + "step": 10107 + }, + { + "epoch": 0.31881406718183253, + "grad_norm": 2.2954092025756836, + "learning_rate": 1.5936293608949383e-05, + "loss": 0.523, + "step": 10108 + }, + { + "epoch": 0.31884560794827316, + "grad_norm": 1.723466157913208, + "learning_rate": 1.593547149206217e-05, + "loss": 0.5352, + "step": 10109 + }, + { + "epoch": 0.3188771487147138, + "grad_norm": 1.8705803155899048, + "learning_rate": 1.59346493132337e-05, + "loss": 0.5049, + "step": 10110 + }, + { + "epoch": 0.3189086894811544, + "grad_norm": 4.667625904083252, + "learning_rate": 1.5933827072472556e-05, + "loss": 0.5486, + "step": 10111 + }, + { + "epoch": 0.318940230247595, + "grad_norm": 1.4488774538040161, + "learning_rate": 1.5933004769787323e-05, + "loss": 0.5437, + "step": 10112 + }, + { + "epoch": 0.3189717710140356, + "grad_norm": 2.899685859680176, + "learning_rate": 1.5932182405186576e-05, + "loss": 0.5156, + "step": 10113 + }, + { + "epoch": 0.31900331178047625, + "grad_norm": 1.5740649700164795, + "learning_rate": 1.59313599786789e-05, + "loss": 0.5403, + "step": 10114 + }, + { + "epoch": 0.3190348525469169, + "grad_norm": 1.440720796585083, + "learning_rate": 1.5930537490272876e-05, + "loss": 0.4986, + "step": 10115 + }, + { + "epoch": 0.3190663933133575, + "grad_norm": 1.876039981842041, + "learning_rate": 1.592971493997709e-05, + "loss": 0.5307, + "step": 10116 + }, + { + "epoch": 0.31909793407979814, + "grad_norm": 1.9613099098205566, + "learning_rate": 1.592889232780012e-05, + "loss": 0.8461, + "step": 10117 + }, + { + "epoch": 0.31912947484623877, + "grad_norm": 6.6908674240112305, + "learning_rate": 1.5928069653750562e-05, + "loss": 0.5519, + "step": 10118 + }, + { + "epoch": 0.3191610156126794, + "grad_norm": 1.777037262916565, + "learning_rate": 1.592724691783699e-05, + "loss": 0.5469, + "step": 10119 + }, + { + "epoch": 0.31919255637912003, + "grad_norm": 2.1271426677703857, + "learning_rate": 1.5926424120067997e-05, + "loss": 0.5112, + "step": 10120 + }, + { + "epoch": 0.31922409714556066, + "grad_norm": 1.784792423248291, + "learning_rate": 1.592560126045216e-05, + "loss": 0.5167, + "step": 10121 + }, + { + "epoch": 0.3192556379120013, + "grad_norm": 2.2517025470733643, + "learning_rate": 1.592477833899808e-05, + "loss": 0.3861, + "step": 10122 + }, + { + "epoch": 0.31928717867844186, + "grad_norm": 1.7136026620864868, + "learning_rate": 1.5923955355714335e-05, + "loss": 0.5267, + "step": 10123 + }, + { + "epoch": 0.3193187194448825, + "grad_norm": 1.9355037212371826, + "learning_rate": 1.5923132310609514e-05, + "loss": 0.8357, + "step": 10124 + }, + { + "epoch": 0.3193502602113231, + "grad_norm": 3.046842336654663, + "learning_rate": 1.592230920369221e-05, + "loss": 0.402, + "step": 10125 + }, + { + "epoch": 0.31938180097776375, + "grad_norm": 2.333110809326172, + "learning_rate": 1.5921486034971007e-05, + "loss": 0.5868, + "step": 10126 + }, + { + "epoch": 0.3194133417442044, + "grad_norm": 1.7388843297958374, + "learning_rate": 1.5920662804454506e-05, + "loss": 0.5549, + "step": 10127 + }, + { + "epoch": 0.319444882510645, + "grad_norm": 1.5639772415161133, + "learning_rate": 1.591983951215129e-05, + "loss": 0.5564, + "step": 10128 + }, + { + "epoch": 0.31947642327708564, + "grad_norm": 1.4926880598068237, + "learning_rate": 1.591901615806995e-05, + "loss": 0.4037, + "step": 10129 + }, + { + "epoch": 0.31950796404352627, + "grad_norm": 1.5601946115493774, + "learning_rate": 1.591819274221908e-05, + "loss": 0.5479, + "step": 10130 + }, + { + "epoch": 0.3195395048099669, + "grad_norm": 1.6071038246154785, + "learning_rate": 1.591736926460727e-05, + "loss": 0.5397, + "step": 10131 + }, + { + "epoch": 0.31957104557640753, + "grad_norm": 1.5558480024337769, + "learning_rate": 1.5916545725243124e-05, + "loss": 0.5119, + "step": 10132 + }, + { + "epoch": 0.3196025863428481, + "grad_norm": 1.8496789932250977, + "learning_rate": 1.5915722124135227e-05, + "loss": 0.563, + "step": 10133 + }, + { + "epoch": 0.31963412710928873, + "grad_norm": 1.7044291496276855, + "learning_rate": 1.5914898461292175e-05, + "loss": 0.5668, + "step": 10134 + }, + { + "epoch": 0.31966566787572936, + "grad_norm": 2.919006824493408, + "learning_rate": 1.5914074736722565e-05, + "loss": 0.8756, + "step": 10135 + }, + { + "epoch": 0.31969720864217, + "grad_norm": 1.51656973361969, + "learning_rate": 1.591325095043499e-05, + "loss": 0.5587, + "step": 10136 + }, + { + "epoch": 0.3197287494086106, + "grad_norm": 1.3550642728805542, + "learning_rate": 1.5912427102438055e-05, + "loss": 0.5247, + "step": 10137 + }, + { + "epoch": 0.31976029017505125, + "grad_norm": 1.8045037984848022, + "learning_rate": 1.591160319274035e-05, + "loss": 0.5156, + "step": 10138 + }, + { + "epoch": 0.3197918309414919, + "grad_norm": 1.679896593093872, + "learning_rate": 1.5910779221350478e-05, + "loss": 0.5358, + "step": 10139 + }, + { + "epoch": 0.3198233717079325, + "grad_norm": 2.7840442657470703, + "learning_rate": 1.5909955188277032e-05, + "loss": 0.5112, + "step": 10140 + }, + { + "epoch": 0.31985491247437314, + "grad_norm": 1.7059688568115234, + "learning_rate": 1.5909131093528617e-05, + "loss": 0.5397, + "step": 10141 + }, + { + "epoch": 0.3198864532408138, + "grad_norm": 1.848497986793518, + "learning_rate": 1.5908306937113825e-05, + "loss": 0.5372, + "step": 10142 + }, + { + "epoch": 0.3199179940072544, + "grad_norm": 1.8623058795928955, + "learning_rate": 1.590748271904127e-05, + "loss": 0.5572, + "step": 10143 + }, + { + "epoch": 0.319949534773695, + "grad_norm": 2.5999836921691895, + "learning_rate": 1.5906658439319543e-05, + "loss": 0.8344, + "step": 10144 + }, + { + "epoch": 0.3199810755401356, + "grad_norm": 1.8204840421676636, + "learning_rate": 1.5905834097957245e-05, + "loss": 0.5515, + "step": 10145 + }, + { + "epoch": 0.32001261630657624, + "grad_norm": 1.5428895950317383, + "learning_rate": 1.5905009694962987e-05, + "loss": 0.537, + "step": 10146 + }, + { + "epoch": 0.32004415707301687, + "grad_norm": 1.9690964221954346, + "learning_rate": 1.5904185230345367e-05, + "loss": 0.5288, + "step": 10147 + }, + { + "epoch": 0.3200756978394575, + "grad_norm": 1.7416702508926392, + "learning_rate": 1.5903360704112988e-05, + "loss": 0.5151, + "step": 10148 + }, + { + "epoch": 0.3201072386058981, + "grad_norm": 1.8308799266815186, + "learning_rate": 1.590253611627446e-05, + "loss": 0.5625, + "step": 10149 + }, + { + "epoch": 0.32013877937233876, + "grad_norm": 2.035442590713501, + "learning_rate": 1.5901711466838378e-05, + "loss": 0.5524, + "step": 10150 + }, + { + "epoch": 0.3201703201387794, + "grad_norm": 1.5389662981033325, + "learning_rate": 1.590088675581336e-05, + "loss": 0.5297, + "step": 10151 + }, + { + "epoch": 0.32020186090522, + "grad_norm": 1.4211550951004028, + "learning_rate": 1.5900061983208005e-05, + "loss": 0.5521, + "step": 10152 + }, + { + "epoch": 0.32023340167166064, + "grad_norm": 1.9417184591293335, + "learning_rate": 1.589923714903092e-05, + "loss": 0.5624, + "step": 10153 + }, + { + "epoch": 0.3202649424381012, + "grad_norm": 1.5250426530838013, + "learning_rate": 1.5898412253290718e-05, + "loss": 0.5559, + "step": 10154 + }, + { + "epoch": 0.32029648320454185, + "grad_norm": 1.3019018173217773, + "learning_rate": 1.5897587295996002e-05, + "loss": 0.536, + "step": 10155 + }, + { + "epoch": 0.3203280239709825, + "grad_norm": 1.699517846107483, + "learning_rate": 1.5896762277155383e-05, + "loss": 0.4352, + "step": 10156 + }, + { + "epoch": 0.3203595647374231, + "grad_norm": 1.7762466669082642, + "learning_rate": 1.5895937196777472e-05, + "loss": 0.5181, + "step": 10157 + }, + { + "epoch": 0.32039110550386374, + "grad_norm": 2.0226478576660156, + "learning_rate": 1.5895112054870878e-05, + "loss": 0.5275, + "step": 10158 + }, + { + "epoch": 0.32042264627030437, + "grad_norm": 1.0916271209716797, + "learning_rate": 1.5894286851444213e-05, + "loss": 0.3811, + "step": 10159 + }, + { + "epoch": 0.320454187036745, + "grad_norm": 1.354839563369751, + "learning_rate": 1.5893461586506087e-05, + "loss": 0.5258, + "step": 10160 + }, + { + "epoch": 0.3204857278031856, + "grad_norm": 1.6689543724060059, + "learning_rate": 1.5892636260065115e-05, + "loss": 0.5366, + "step": 10161 + }, + { + "epoch": 0.32051726856962626, + "grad_norm": 2.362971305847168, + "learning_rate": 1.5891810872129907e-05, + "loss": 0.5592, + "step": 10162 + }, + { + "epoch": 0.3205488093360669, + "grad_norm": 1.7323213815689087, + "learning_rate": 1.5890985422709075e-05, + "loss": 0.5309, + "step": 10163 + }, + { + "epoch": 0.3205803501025075, + "grad_norm": 1.4594902992248535, + "learning_rate": 1.589015991181124e-05, + "loss": 0.5651, + "step": 10164 + }, + { + "epoch": 0.3206118908689481, + "grad_norm": 1.6050195693969727, + "learning_rate": 1.5889334339445008e-05, + "loss": 0.5369, + "step": 10165 + }, + { + "epoch": 0.3206434316353887, + "grad_norm": 1.4203308820724487, + "learning_rate": 1.5888508705619005e-05, + "loss": 0.5274, + "step": 10166 + }, + { + "epoch": 0.32067497240182935, + "grad_norm": 1.9822449684143066, + "learning_rate": 1.5887683010341836e-05, + "loss": 0.5077, + "step": 10167 + }, + { + "epoch": 0.32070651316827, + "grad_norm": 1.6680136919021606, + "learning_rate": 1.5886857253622126e-05, + "loss": 0.5448, + "step": 10168 + }, + { + "epoch": 0.3207380539347106, + "grad_norm": 1.6450748443603516, + "learning_rate": 1.5886031435468486e-05, + "loss": 0.5255, + "step": 10169 + }, + { + "epoch": 0.32076959470115124, + "grad_norm": 2.7981581687927246, + "learning_rate": 1.588520555588954e-05, + "loss": 0.5009, + "step": 10170 + }, + { + "epoch": 0.32080113546759187, + "grad_norm": 1.7789952754974365, + "learning_rate": 1.5884379614893908e-05, + "loss": 0.504, + "step": 10171 + }, + { + "epoch": 0.3208326762340325, + "grad_norm": 1.6464422941207886, + "learning_rate": 1.58835536124902e-05, + "loss": 0.4247, + "step": 10172 + }, + { + "epoch": 0.32086421700047313, + "grad_norm": 3.4420857429504395, + "learning_rate": 1.5882727548687045e-05, + "loss": 0.8973, + "step": 10173 + }, + { + "epoch": 0.32089575776691376, + "grad_norm": 1.7962517738342285, + "learning_rate": 1.588190142349306e-05, + "loss": 0.5422, + "step": 10174 + }, + { + "epoch": 0.3209272985333544, + "grad_norm": 2.07438063621521, + "learning_rate": 1.5881075236916865e-05, + "loss": 0.867, + "step": 10175 + }, + { + "epoch": 0.32095883929979496, + "grad_norm": 1.5259844064712524, + "learning_rate": 1.5880248988967086e-05, + "loss": 0.4875, + "step": 10176 + }, + { + "epoch": 0.3209903800662356, + "grad_norm": 1.612173318862915, + "learning_rate": 1.587942267965234e-05, + "loss": 0.5399, + "step": 10177 + }, + { + "epoch": 0.3210219208326762, + "grad_norm": 1.5006422996520996, + "learning_rate": 1.5878596308981254e-05, + "loss": 0.8694, + "step": 10178 + }, + { + "epoch": 0.32105346159911685, + "grad_norm": 1.5403928756713867, + "learning_rate": 1.587776987696245e-05, + "loss": 0.5076, + "step": 10179 + }, + { + "epoch": 0.3210850023655575, + "grad_norm": 1.5325055122375488, + "learning_rate": 1.5876943383604554e-05, + "loss": 0.5482, + "step": 10180 + }, + { + "epoch": 0.3211165431319981, + "grad_norm": 2.0588924884796143, + "learning_rate": 1.5876116828916192e-05, + "loss": 0.542, + "step": 10181 + }, + { + "epoch": 0.32114808389843874, + "grad_norm": 1.5674957036972046, + "learning_rate": 1.5875290212905986e-05, + "loss": 0.5059, + "step": 10182 + }, + { + "epoch": 0.32117962466487937, + "grad_norm": 1.6663780212402344, + "learning_rate": 1.587446353558257e-05, + "loss": 0.4903, + "step": 10183 + }, + { + "epoch": 0.32121116543132, + "grad_norm": 1.7903283834457397, + "learning_rate": 1.5873636796954555e-05, + "loss": 0.5356, + "step": 10184 + }, + { + "epoch": 0.32124270619776063, + "grad_norm": 1.6691049337387085, + "learning_rate": 1.587280999703059e-05, + "loss": 0.5253, + "step": 10185 + }, + { + "epoch": 0.3212742469642012, + "grad_norm": 2.0771961212158203, + "learning_rate": 1.5871983135819283e-05, + "loss": 0.4196, + "step": 10186 + }, + { + "epoch": 0.32130578773064183, + "grad_norm": 1.63034188747406, + "learning_rate": 1.587115621332928e-05, + "loss": 0.5768, + "step": 10187 + }, + { + "epoch": 0.32133732849708246, + "grad_norm": 1.5468312501907349, + "learning_rate": 1.58703292295692e-05, + "loss": 0.5559, + "step": 10188 + }, + { + "epoch": 0.3213688692635231, + "grad_norm": 1.8745362758636475, + "learning_rate": 1.586950218454768e-05, + "loss": 0.5369, + "step": 10189 + }, + { + "epoch": 0.3214004100299637, + "grad_norm": 1.7318648099899292, + "learning_rate": 1.5868675078273343e-05, + "loss": 0.5332, + "step": 10190 + }, + { + "epoch": 0.32143195079640435, + "grad_norm": 1.792204737663269, + "learning_rate": 1.5867847910754822e-05, + "loss": 0.5496, + "step": 10191 + }, + { + "epoch": 0.321463491562845, + "grad_norm": 1.5394178628921509, + "learning_rate": 1.5867020682000758e-05, + "loss": 0.5194, + "step": 10192 + }, + { + "epoch": 0.3214950323292856, + "grad_norm": 1.34026300907135, + "learning_rate": 1.5866193392019775e-05, + "loss": 0.5288, + "step": 10193 + }, + { + "epoch": 0.32152657309572624, + "grad_norm": 2.19506573677063, + "learning_rate": 1.586536604082051e-05, + "loss": 0.5286, + "step": 10194 + }, + { + "epoch": 0.32155811386216687, + "grad_norm": 1.9919379949569702, + "learning_rate": 1.5864538628411596e-05, + "loss": 0.5303, + "step": 10195 + }, + { + "epoch": 0.3215896546286075, + "grad_norm": 1.790393590927124, + "learning_rate": 1.5863711154801667e-05, + "loss": 0.5385, + "step": 10196 + }, + { + "epoch": 0.3216211953950481, + "grad_norm": 1.5292762517929077, + "learning_rate": 1.5862883619999356e-05, + "loss": 0.5113, + "step": 10197 + }, + { + "epoch": 0.3216527361614887, + "grad_norm": 2.1143980026245117, + "learning_rate": 1.5862056024013304e-05, + "loss": 0.5133, + "step": 10198 + }, + { + "epoch": 0.32168427692792934, + "grad_norm": 2.2604050636291504, + "learning_rate": 1.5861228366852148e-05, + "loss": 0.4965, + "step": 10199 + }, + { + "epoch": 0.32171581769436997, + "grad_norm": 1.9401930570602417, + "learning_rate": 1.586040064852452e-05, + "loss": 0.5217, + "step": 10200 + }, + { + "epoch": 0.3217473584608106, + "grad_norm": 4.91261100769043, + "learning_rate": 1.5859572869039063e-05, + "loss": 0.8748, + "step": 10201 + }, + { + "epoch": 0.3217788992272512, + "grad_norm": 1.9139213562011719, + "learning_rate": 1.5858745028404414e-05, + "loss": 0.3988, + "step": 10202 + }, + { + "epoch": 0.32181043999369185, + "grad_norm": 1.45791757106781, + "learning_rate": 1.585791712662921e-05, + "loss": 0.5107, + "step": 10203 + }, + { + "epoch": 0.3218419807601325, + "grad_norm": 1.543752908706665, + "learning_rate": 1.585708916372209e-05, + "loss": 0.5482, + "step": 10204 + }, + { + "epoch": 0.3218735215265731, + "grad_norm": 1.6451826095581055, + "learning_rate": 1.58562611396917e-05, + "loss": 0.5081, + "step": 10205 + }, + { + "epoch": 0.32190506229301374, + "grad_norm": 1.920560359954834, + "learning_rate": 1.5855433054546673e-05, + "loss": 0.5487, + "step": 10206 + }, + { + "epoch": 0.3219366030594543, + "grad_norm": 1.4545544385910034, + "learning_rate": 1.5854604908295662e-05, + "loss": 0.5058, + "step": 10207 + }, + { + "epoch": 0.32196814382589495, + "grad_norm": 2.126784563064575, + "learning_rate": 1.5853776700947297e-05, + "loss": 0.5626, + "step": 10208 + }, + { + "epoch": 0.3219996845923356, + "grad_norm": 3.0116727352142334, + "learning_rate": 1.5852948432510232e-05, + "loss": 0.5252, + "step": 10209 + }, + { + "epoch": 0.3220312253587762, + "grad_norm": 1.3917726278305054, + "learning_rate": 1.5852120102993103e-05, + "loss": 0.5202, + "step": 10210 + }, + { + "epoch": 0.32206276612521684, + "grad_norm": 1.4478764533996582, + "learning_rate": 1.5851291712404555e-05, + "loss": 0.5377, + "step": 10211 + }, + { + "epoch": 0.32209430689165747, + "grad_norm": 1.6118371486663818, + "learning_rate": 1.5850463260753237e-05, + "loss": 0.5631, + "step": 10212 + }, + { + "epoch": 0.3221258476580981, + "grad_norm": 1.9247026443481445, + "learning_rate": 1.584963474804779e-05, + "loss": 0.5551, + "step": 10213 + }, + { + "epoch": 0.3221573884245387, + "grad_norm": 1.7362550497055054, + "learning_rate": 1.5848806174296863e-05, + "loss": 0.4891, + "step": 10214 + }, + { + "epoch": 0.32218892919097936, + "grad_norm": 1.7245380878448486, + "learning_rate": 1.5847977539509103e-05, + "loss": 0.5354, + "step": 10215 + }, + { + "epoch": 0.32222046995742, + "grad_norm": 2.360882520675659, + "learning_rate": 1.5847148843693158e-05, + "loss": 0.8472, + "step": 10216 + }, + { + "epoch": 0.3222520107238606, + "grad_norm": 1.7139270305633545, + "learning_rate": 1.584632008685767e-05, + "loss": 0.8785, + "step": 10217 + }, + { + "epoch": 0.3222835514903012, + "grad_norm": 1.5601744651794434, + "learning_rate": 1.5845491269011296e-05, + "loss": 0.5508, + "step": 10218 + }, + { + "epoch": 0.3223150922567418, + "grad_norm": 1.7819374799728394, + "learning_rate": 1.584466239016268e-05, + "loss": 0.5642, + "step": 10219 + }, + { + "epoch": 0.32234663302318245, + "grad_norm": 1.6125186681747437, + "learning_rate": 1.5843833450320475e-05, + "loss": 0.538, + "step": 10220 + }, + { + "epoch": 0.3223781737896231, + "grad_norm": 1.6025797128677368, + "learning_rate": 1.584300444949333e-05, + "loss": 0.5323, + "step": 10221 + }, + { + "epoch": 0.3224097145560637, + "grad_norm": 1.605505347251892, + "learning_rate": 1.5842175387689898e-05, + "loss": 0.5626, + "step": 10222 + }, + { + "epoch": 0.32244125532250434, + "grad_norm": 1.3966875076293945, + "learning_rate": 1.5841346264918827e-05, + "loss": 0.5106, + "step": 10223 + }, + { + "epoch": 0.32247279608894497, + "grad_norm": 1.9667916297912598, + "learning_rate": 1.584051708118877e-05, + "loss": 0.4955, + "step": 10224 + }, + { + "epoch": 0.3225043368553856, + "grad_norm": 1.4886523485183716, + "learning_rate": 1.5839687836508385e-05, + "loss": 0.5297, + "step": 10225 + }, + { + "epoch": 0.32253587762182623, + "grad_norm": 1.5536434650421143, + "learning_rate": 1.5838858530886326e-05, + "loss": 0.5224, + "step": 10226 + }, + { + "epoch": 0.32256741838826686, + "grad_norm": 1.3650249242782593, + "learning_rate": 1.5838029164331242e-05, + "loss": 0.525, + "step": 10227 + }, + { + "epoch": 0.32259895915470743, + "grad_norm": 1.4346458911895752, + "learning_rate": 1.5837199736851788e-05, + "loss": 0.5191, + "step": 10228 + }, + { + "epoch": 0.32263049992114806, + "grad_norm": 1.628716230392456, + "learning_rate": 1.5836370248456625e-05, + "loss": 0.5207, + "step": 10229 + }, + { + "epoch": 0.3226620406875887, + "grad_norm": 1.4537169933319092, + "learning_rate": 1.5835540699154405e-05, + "loss": 0.5036, + "step": 10230 + }, + { + "epoch": 0.3226935814540293, + "grad_norm": 1.5651929378509521, + "learning_rate": 1.583471108895379e-05, + "loss": 0.5348, + "step": 10231 + }, + { + "epoch": 0.32272512222046995, + "grad_norm": 1.3915501832962036, + "learning_rate": 1.583388141786343e-05, + "loss": 0.5169, + "step": 10232 + }, + { + "epoch": 0.3227566629869106, + "grad_norm": 1.8604705333709717, + "learning_rate": 1.5833051685891992e-05, + "loss": 0.517, + "step": 10233 + }, + { + "epoch": 0.3227882037533512, + "grad_norm": 1.6897889375686646, + "learning_rate": 1.5832221893048125e-05, + "loss": 0.3905, + "step": 10234 + }, + { + "epoch": 0.32281974451979184, + "grad_norm": 1.68332839012146, + "learning_rate": 1.58313920393405e-05, + "loss": 0.5135, + "step": 10235 + }, + { + "epoch": 0.32285128528623247, + "grad_norm": 1.4849518537521362, + "learning_rate": 1.5830562124777765e-05, + "loss": 0.5648, + "step": 10236 + }, + { + "epoch": 0.3228828260526731, + "grad_norm": 1.5386090278625488, + "learning_rate": 1.582973214936859e-05, + "loss": 0.5763, + "step": 10237 + }, + { + "epoch": 0.32291436681911373, + "grad_norm": 3.5376625061035156, + "learning_rate": 1.5828902113121632e-05, + "loss": 0.9265, + "step": 10238 + }, + { + "epoch": 0.3229459075855543, + "grad_norm": 2.3388612270355225, + "learning_rate": 1.5828072016045557e-05, + "loss": 0.5357, + "step": 10239 + }, + { + "epoch": 0.32297744835199493, + "grad_norm": 1.7687913179397583, + "learning_rate": 1.5827241858149023e-05, + "loss": 0.5357, + "step": 10240 + }, + { + "epoch": 0.32300898911843556, + "grad_norm": 1.3574093580245972, + "learning_rate": 1.5826411639440698e-05, + "loss": 0.537, + "step": 10241 + }, + { + "epoch": 0.3230405298848762, + "grad_norm": 1.46947181224823, + "learning_rate": 1.5825581359929244e-05, + "loss": 0.506, + "step": 10242 + }, + { + "epoch": 0.3230720706513168, + "grad_norm": 1.5498186349868774, + "learning_rate": 1.582475101962332e-05, + "loss": 0.4974, + "step": 10243 + }, + { + "epoch": 0.32310361141775745, + "grad_norm": 1.5820337533950806, + "learning_rate": 1.58239206185316e-05, + "loss": 0.5114, + "step": 10244 + }, + { + "epoch": 0.3231351521841981, + "grad_norm": 2.0950615406036377, + "learning_rate": 1.5823090156662745e-05, + "loss": 0.5328, + "step": 10245 + }, + { + "epoch": 0.3231666929506387, + "grad_norm": 2.1766693592071533, + "learning_rate": 1.5822259634025422e-05, + "loss": 0.5302, + "step": 10246 + }, + { + "epoch": 0.32319823371707934, + "grad_norm": 3.227414608001709, + "learning_rate": 1.58214290506283e-05, + "loss": 0.8748, + "step": 10247 + }, + { + "epoch": 0.32322977448351997, + "grad_norm": 1.620792269706726, + "learning_rate": 1.5820598406480046e-05, + "loss": 0.5214, + "step": 10248 + }, + { + "epoch": 0.3232613152499606, + "grad_norm": 1.5497857332229614, + "learning_rate": 1.5819767701589327e-05, + "loss": 0.5724, + "step": 10249 + }, + { + "epoch": 0.3232928560164012, + "grad_norm": 1.8785266876220703, + "learning_rate": 1.5818936935964814e-05, + "loss": 0.5179, + "step": 10250 + }, + { + "epoch": 0.3233243967828418, + "grad_norm": 1.5507192611694336, + "learning_rate": 1.5818106109615174e-05, + "loss": 0.5654, + "step": 10251 + }, + { + "epoch": 0.32335593754928244, + "grad_norm": 1.5509039163589478, + "learning_rate": 1.581727522254908e-05, + "loss": 0.5301, + "step": 10252 + }, + { + "epoch": 0.32338747831572306, + "grad_norm": 1.4800647497177124, + "learning_rate": 1.5816444274775205e-05, + "loss": 0.5382, + "step": 10253 + }, + { + "epoch": 0.3234190190821637, + "grad_norm": 1.8687232732772827, + "learning_rate": 1.5815613266302216e-05, + "loss": 0.5317, + "step": 10254 + }, + { + "epoch": 0.3234505598486043, + "grad_norm": 1.544561743736267, + "learning_rate": 1.5814782197138784e-05, + "loss": 0.5423, + "step": 10255 + }, + { + "epoch": 0.32348210061504495, + "grad_norm": 2.155958414077759, + "learning_rate": 1.5813951067293583e-05, + "loss": 0.4111, + "step": 10256 + }, + { + "epoch": 0.3235136413814856, + "grad_norm": 1.6319688558578491, + "learning_rate": 1.5813119876775294e-05, + "loss": 0.4023, + "step": 10257 + }, + { + "epoch": 0.3235451821479262, + "grad_norm": 2.070793390274048, + "learning_rate": 1.5812288625592583e-05, + "loss": 0.5157, + "step": 10258 + }, + { + "epoch": 0.32357672291436684, + "grad_norm": 4.055390357971191, + "learning_rate": 1.5811457313754128e-05, + "loss": 0.5451, + "step": 10259 + }, + { + "epoch": 0.3236082636808074, + "grad_norm": 1.6020255088806152, + "learning_rate": 1.5810625941268604e-05, + "loss": 0.5595, + "step": 10260 + }, + { + "epoch": 0.32363980444724805, + "grad_norm": 2.3473713397979736, + "learning_rate": 1.5809794508144687e-05, + "loss": 0.5609, + "step": 10261 + }, + { + "epoch": 0.3236713452136887, + "grad_norm": 1.7784560918807983, + "learning_rate": 1.580896301439105e-05, + "loss": 0.5366, + "step": 10262 + }, + { + "epoch": 0.3237028859801293, + "grad_norm": 1.5834064483642578, + "learning_rate": 1.5808131460016374e-05, + "loss": 0.5541, + "step": 10263 + }, + { + "epoch": 0.32373442674656994, + "grad_norm": 1.9322917461395264, + "learning_rate": 1.580729984502934e-05, + "loss": 0.539, + "step": 10264 + }, + { + "epoch": 0.32376596751301057, + "grad_norm": 1.4201979637145996, + "learning_rate": 1.580646816943862e-05, + "loss": 0.527, + "step": 10265 + }, + { + "epoch": 0.3237975082794512, + "grad_norm": 2.838111400604248, + "learning_rate": 1.5805636433252892e-05, + "loss": 0.8583, + "step": 10266 + }, + { + "epoch": 0.3238290490458918, + "grad_norm": 1.5084234476089478, + "learning_rate": 1.5804804636480847e-05, + "loss": 0.5415, + "step": 10267 + }, + { + "epoch": 0.32386058981233246, + "grad_norm": 1.5770457983016968, + "learning_rate": 1.5803972779131154e-05, + "loss": 0.4989, + "step": 10268 + }, + { + "epoch": 0.3238921305787731, + "grad_norm": 1.7415587902069092, + "learning_rate": 1.58031408612125e-05, + "loss": 0.8472, + "step": 10269 + }, + { + "epoch": 0.3239236713452137, + "grad_norm": 1.5517215728759766, + "learning_rate": 1.580230888273356e-05, + "loss": 0.5176, + "step": 10270 + }, + { + "epoch": 0.3239552121116543, + "grad_norm": 2.3048830032348633, + "learning_rate": 1.580147684370303e-05, + "loss": 0.4309, + "step": 10271 + }, + { + "epoch": 0.3239867528780949, + "grad_norm": 1.4364445209503174, + "learning_rate": 1.580064474412958e-05, + "loss": 0.505, + "step": 10272 + }, + { + "epoch": 0.32401829364453555, + "grad_norm": 2.2505173683166504, + "learning_rate": 1.5799812584021896e-05, + "loss": 0.5072, + "step": 10273 + }, + { + "epoch": 0.3240498344109762, + "grad_norm": 1.700307846069336, + "learning_rate": 1.579898036338867e-05, + "loss": 0.5061, + "step": 10274 + }, + { + "epoch": 0.3240813751774168, + "grad_norm": 1.5321520566940308, + "learning_rate": 1.5798148082238576e-05, + "loss": 0.5175, + "step": 10275 + }, + { + "epoch": 0.32411291594385744, + "grad_norm": 1.766347885131836, + "learning_rate": 1.579731574058031e-05, + "loss": 0.5487, + "step": 10276 + }, + { + "epoch": 0.32414445671029807, + "grad_norm": 1.7909550666809082, + "learning_rate": 1.5796483338422545e-05, + "loss": 0.5207, + "step": 10277 + }, + { + "epoch": 0.3241759974767387, + "grad_norm": 1.6745597124099731, + "learning_rate": 1.579565087577398e-05, + "loss": 0.5142, + "step": 10278 + }, + { + "epoch": 0.3242075382431793, + "grad_norm": 2.629910469055176, + "learning_rate": 1.5794818352643298e-05, + "loss": 0.8699, + "step": 10279 + }, + { + "epoch": 0.32423907900961996, + "grad_norm": 1.4663710594177246, + "learning_rate": 1.5793985769039187e-05, + "loss": 0.5233, + "step": 10280 + }, + { + "epoch": 0.32427061977606053, + "grad_norm": 1.507713794708252, + "learning_rate": 1.5793153124970334e-05, + "loss": 0.5531, + "step": 10281 + }, + { + "epoch": 0.32430216054250116, + "grad_norm": 2.1877126693725586, + "learning_rate": 1.5792320420445433e-05, + "loss": 0.8514, + "step": 10282 + }, + { + "epoch": 0.3243337013089418, + "grad_norm": 1.829972267150879, + "learning_rate": 1.5791487655473167e-05, + "loss": 0.5191, + "step": 10283 + }, + { + "epoch": 0.3243652420753824, + "grad_norm": 2.7978782653808594, + "learning_rate": 1.5790654830062234e-05, + "loss": 0.9076, + "step": 10284 + }, + { + "epoch": 0.32439678284182305, + "grad_norm": 2.5903215408325195, + "learning_rate": 1.5789821944221318e-05, + "loss": 0.5574, + "step": 10285 + }, + { + "epoch": 0.3244283236082637, + "grad_norm": 2.9604644775390625, + "learning_rate": 1.5788988997959115e-05, + "loss": 0.4684, + "step": 10286 + }, + { + "epoch": 0.3244598643747043, + "grad_norm": 1.5335254669189453, + "learning_rate": 1.578815599128432e-05, + "loss": 0.4941, + "step": 10287 + }, + { + "epoch": 0.32449140514114494, + "grad_norm": 2.103334903717041, + "learning_rate": 1.578732292420562e-05, + "loss": 0.5354, + "step": 10288 + }, + { + "epoch": 0.32452294590758557, + "grad_norm": 1.7063536643981934, + "learning_rate": 1.578648979673171e-05, + "loss": 0.5318, + "step": 10289 + }, + { + "epoch": 0.3245544866740262, + "grad_norm": 2.149902820587158, + "learning_rate": 1.578565660887129e-05, + "loss": 0.528, + "step": 10290 + }, + { + "epoch": 0.32458602744046683, + "grad_norm": 2.701986312866211, + "learning_rate": 1.578482336063305e-05, + "loss": 0.8655, + "step": 10291 + }, + { + "epoch": 0.3246175682069074, + "grad_norm": 1.6524587869644165, + "learning_rate": 1.5783990052025684e-05, + "loss": 0.5569, + "step": 10292 + }, + { + "epoch": 0.32464910897334803, + "grad_norm": 1.5286946296691895, + "learning_rate": 1.578315668305789e-05, + "loss": 0.5349, + "step": 10293 + }, + { + "epoch": 0.32468064973978866, + "grad_norm": 1.5488625764846802, + "learning_rate": 1.5782323253738368e-05, + "loss": 0.5592, + "step": 10294 + }, + { + "epoch": 0.3247121905062293, + "grad_norm": 1.697389841079712, + "learning_rate": 1.5781489764075816e-05, + "loss": 0.5264, + "step": 10295 + }, + { + "epoch": 0.3247437312726699, + "grad_norm": 2.341729164123535, + "learning_rate": 1.5780656214078923e-05, + "loss": 0.5344, + "step": 10296 + }, + { + "epoch": 0.32477527203911055, + "grad_norm": 1.3749516010284424, + "learning_rate": 1.5779822603756395e-05, + "loss": 0.5444, + "step": 10297 + }, + { + "epoch": 0.3248068128055512, + "grad_norm": 1.5317246913909912, + "learning_rate": 1.5778988933116936e-05, + "loss": 0.5392, + "step": 10298 + }, + { + "epoch": 0.3248383535719918, + "grad_norm": 1.8546983003616333, + "learning_rate": 1.5778155202169232e-05, + "loss": 0.5653, + "step": 10299 + }, + { + "epoch": 0.32486989433843244, + "grad_norm": 1.8110374212265015, + "learning_rate": 1.5777321410921995e-05, + "loss": 0.8654, + "step": 10300 + }, + { + "epoch": 0.32490143510487307, + "grad_norm": 1.2999628782272339, + "learning_rate": 1.5776487559383924e-05, + "loss": 0.5284, + "step": 10301 + }, + { + "epoch": 0.3249329758713137, + "grad_norm": 1.4878016710281372, + "learning_rate": 1.5775653647563722e-05, + "loss": 0.5171, + "step": 10302 + }, + { + "epoch": 0.3249645166377543, + "grad_norm": 3.356412887573242, + "learning_rate": 1.5774819675470085e-05, + "loss": 0.5406, + "step": 10303 + }, + { + "epoch": 0.3249960574041949, + "grad_norm": 1.7245351076126099, + "learning_rate": 1.577398564311172e-05, + "loss": 0.5511, + "step": 10304 + }, + { + "epoch": 0.32502759817063553, + "grad_norm": 3.066279411315918, + "learning_rate": 1.577315155049734e-05, + "loss": 0.8303, + "step": 10305 + }, + { + "epoch": 0.32505913893707616, + "grad_norm": 1.610529899597168, + "learning_rate": 1.5772317397635634e-05, + "loss": 0.5017, + "step": 10306 + }, + { + "epoch": 0.3250906797035168, + "grad_norm": 1.974104881286621, + "learning_rate": 1.5771483184535316e-05, + "loss": 0.5154, + "step": 10307 + }, + { + "epoch": 0.3251222204699574, + "grad_norm": 1.4401237964630127, + "learning_rate": 1.5770648911205087e-05, + "loss": 0.5357, + "step": 10308 + }, + { + "epoch": 0.32515376123639805, + "grad_norm": 1.5148602724075317, + "learning_rate": 1.576981457765366e-05, + "loss": 0.5121, + "step": 10309 + }, + { + "epoch": 0.3251853020028387, + "grad_norm": 1.6534340381622314, + "learning_rate": 1.5768980183889734e-05, + "loss": 0.5292, + "step": 10310 + }, + { + "epoch": 0.3252168427692793, + "grad_norm": 1.5051950216293335, + "learning_rate": 1.5768145729922018e-05, + "loss": 0.5248, + "step": 10311 + }, + { + "epoch": 0.32524838353571994, + "grad_norm": 1.5819804668426514, + "learning_rate": 1.5767311215759228e-05, + "loss": 0.5264, + "step": 10312 + }, + { + "epoch": 0.3252799243021605, + "grad_norm": 1.9071675539016724, + "learning_rate": 1.5766476641410065e-05, + "loss": 0.5065, + "step": 10313 + }, + { + "epoch": 0.32531146506860115, + "grad_norm": 1.4338905811309814, + "learning_rate": 1.5765642006883242e-05, + "loss": 0.5413, + "step": 10314 + }, + { + "epoch": 0.3253430058350418, + "grad_norm": 1.2536953687667847, + "learning_rate": 1.576480731218747e-05, + "loss": 0.5125, + "step": 10315 + }, + { + "epoch": 0.3253745466014824, + "grad_norm": 1.9939265251159668, + "learning_rate": 1.576397255733145e-05, + "loss": 0.5457, + "step": 10316 + }, + { + "epoch": 0.32540608736792304, + "grad_norm": 1.414168119430542, + "learning_rate": 1.5763137742323903e-05, + "loss": 0.5168, + "step": 10317 + }, + { + "epoch": 0.32543762813436367, + "grad_norm": 2.2036094665527344, + "learning_rate": 1.5762302867173543e-05, + "loss": 0.5071, + "step": 10318 + }, + { + "epoch": 0.3254691689008043, + "grad_norm": 1.5519347190856934, + "learning_rate": 1.5761467931889075e-05, + "loss": 0.5134, + "step": 10319 + }, + { + "epoch": 0.3255007096672449, + "grad_norm": 1.6953098773956299, + "learning_rate": 1.5760632936479217e-05, + "loss": 0.5236, + "step": 10320 + }, + { + "epoch": 0.32553225043368555, + "grad_norm": 1.4725887775421143, + "learning_rate": 1.5759797880952676e-05, + "loss": 0.513, + "step": 10321 + }, + { + "epoch": 0.3255637912001262, + "grad_norm": 1.841469168663025, + "learning_rate": 1.5758962765318177e-05, + "loss": 0.5199, + "step": 10322 + }, + { + "epoch": 0.3255953319665668, + "grad_norm": 1.390004277229309, + "learning_rate": 1.575812758958443e-05, + "loss": 0.5298, + "step": 10323 + }, + { + "epoch": 0.3256268727330074, + "grad_norm": 1.5547456741333008, + "learning_rate": 1.5757292353760153e-05, + "loss": 0.5283, + "step": 10324 + }, + { + "epoch": 0.325658413499448, + "grad_norm": 1.4641034603118896, + "learning_rate": 1.5756457057854054e-05, + "loss": 0.525, + "step": 10325 + }, + { + "epoch": 0.32568995426588865, + "grad_norm": 5.388288497924805, + "learning_rate": 1.5755621701874857e-05, + "loss": 0.7551, + "step": 10326 + }, + { + "epoch": 0.3257214950323293, + "grad_norm": 1.6560347080230713, + "learning_rate": 1.575478628583128e-05, + "loss": 0.5061, + "step": 10327 + }, + { + "epoch": 0.3257530357987699, + "grad_norm": 4.285520553588867, + "learning_rate": 1.575395080973204e-05, + "loss": 0.5371, + "step": 10328 + }, + { + "epoch": 0.32578457656521054, + "grad_norm": 1.470699667930603, + "learning_rate": 1.5753115273585856e-05, + "loss": 0.5431, + "step": 10329 + }, + { + "epoch": 0.32581611733165117, + "grad_norm": 1.6562670469284058, + "learning_rate": 1.5752279677401444e-05, + "loss": 0.871, + "step": 10330 + }, + { + "epoch": 0.3258476580980918, + "grad_norm": 1.4532833099365234, + "learning_rate": 1.575144402118753e-05, + "loss": 0.5296, + "step": 10331 + }, + { + "epoch": 0.3258791988645324, + "grad_norm": 1.4287463426589966, + "learning_rate": 1.575060830495283e-05, + "loss": 0.5504, + "step": 10332 + }, + { + "epoch": 0.32591073963097306, + "grad_norm": 1.640559196472168, + "learning_rate": 1.574977252870607e-05, + "loss": 0.541, + "step": 10333 + }, + { + "epoch": 0.32594228039741363, + "grad_norm": 1.6068768501281738, + "learning_rate": 1.5748936692455963e-05, + "loss": 0.5401, + "step": 10334 + }, + { + "epoch": 0.32597382116385426, + "grad_norm": 1.3935215473175049, + "learning_rate": 1.5748100796211244e-05, + "loss": 0.5339, + "step": 10335 + }, + { + "epoch": 0.3260053619302949, + "grad_norm": 1.8486701250076294, + "learning_rate": 1.5747264839980628e-05, + "loss": 0.5075, + "step": 10336 + }, + { + "epoch": 0.3260369026967355, + "grad_norm": 1.6450599431991577, + "learning_rate": 1.574642882377284e-05, + "loss": 0.8981, + "step": 10337 + }, + { + "epoch": 0.32606844346317615, + "grad_norm": 1.7615635395050049, + "learning_rate": 1.5745592747596604e-05, + "loss": 0.5498, + "step": 10338 + }, + { + "epoch": 0.3260999842296168, + "grad_norm": 1.2956039905548096, + "learning_rate": 1.574475661146065e-05, + "loss": 0.5393, + "step": 10339 + }, + { + "epoch": 0.3261315249960574, + "grad_norm": 1.3848624229431152, + "learning_rate": 1.5743920415373693e-05, + "loss": 0.5132, + "step": 10340 + }, + { + "epoch": 0.32616306576249804, + "grad_norm": 1.4014158248901367, + "learning_rate": 1.5743084159344473e-05, + "loss": 0.5212, + "step": 10341 + }, + { + "epoch": 0.32619460652893867, + "grad_norm": 1.6139800548553467, + "learning_rate": 1.5742247843381705e-05, + "loss": 0.5008, + "step": 10342 + }, + { + "epoch": 0.3262261472953793, + "grad_norm": 1.7703418731689453, + "learning_rate": 1.5741411467494128e-05, + "loss": 0.527, + "step": 10343 + }, + { + "epoch": 0.32625768806181993, + "grad_norm": 1.74611234664917, + "learning_rate": 1.5740575031690465e-05, + "loss": 0.5222, + "step": 10344 + }, + { + "epoch": 0.3262892288282605, + "grad_norm": 1.4977781772613525, + "learning_rate": 1.5739738535979435e-05, + "loss": 0.5037, + "step": 10345 + }, + { + "epoch": 0.32632076959470113, + "grad_norm": 1.5666024684906006, + "learning_rate": 1.5738901980369786e-05, + "loss": 0.5143, + "step": 10346 + }, + { + "epoch": 0.32635231036114176, + "grad_norm": 1.4635215997695923, + "learning_rate": 1.5738065364870235e-05, + "loss": 0.8473, + "step": 10347 + }, + { + "epoch": 0.3263838511275824, + "grad_norm": 1.8099197149276733, + "learning_rate": 1.5737228689489517e-05, + "loss": 0.533, + "step": 10348 + }, + { + "epoch": 0.326415391894023, + "grad_norm": 1.3869081735610962, + "learning_rate": 1.573639195423636e-05, + "loss": 0.495, + "step": 10349 + }, + { + "epoch": 0.32644693266046365, + "grad_norm": 1.768218755722046, + "learning_rate": 1.5735555159119506e-05, + "loss": 0.8276, + "step": 10350 + }, + { + "epoch": 0.3264784734269043, + "grad_norm": 1.681673526763916, + "learning_rate": 1.5734718304147674e-05, + "loss": 0.5314, + "step": 10351 + }, + { + "epoch": 0.3265100141933449, + "grad_norm": 1.850890040397644, + "learning_rate": 1.5733881389329604e-05, + "loss": 0.5432, + "step": 10352 + }, + { + "epoch": 0.32654155495978554, + "grad_norm": 1.7625428438186646, + "learning_rate": 1.573304441467403e-05, + "loss": 0.5335, + "step": 10353 + }, + { + "epoch": 0.32657309572622617, + "grad_norm": 1.5441159009933472, + "learning_rate": 1.573220738018969e-05, + "loss": 0.8118, + "step": 10354 + }, + { + "epoch": 0.32660463649266674, + "grad_norm": 1.4678235054016113, + "learning_rate": 1.5731370285885307e-05, + "loss": 0.5408, + "step": 10355 + }, + { + "epoch": 0.3266361772591074, + "grad_norm": 1.8590198755264282, + "learning_rate": 1.573053313176963e-05, + "loss": 0.5677, + "step": 10356 + }, + { + "epoch": 0.326667718025548, + "grad_norm": 5.613908290863037, + "learning_rate": 1.5729695917851392e-05, + "loss": 0.7089, + "step": 10357 + }, + { + "epoch": 0.32669925879198863, + "grad_norm": 1.4612224102020264, + "learning_rate": 1.5728858644139323e-05, + "loss": 0.5392, + "step": 10358 + }, + { + "epoch": 0.32673079955842926, + "grad_norm": 1.6345365047454834, + "learning_rate": 1.5728021310642167e-05, + "loss": 0.5454, + "step": 10359 + }, + { + "epoch": 0.3267623403248699, + "grad_norm": 1.87114417552948, + "learning_rate": 1.5727183917368663e-05, + "loss": 0.523, + "step": 10360 + }, + { + "epoch": 0.3267938810913105, + "grad_norm": 1.3763728141784668, + "learning_rate": 1.572634646432755e-05, + "loss": 0.551, + "step": 10361 + }, + { + "epoch": 0.32682542185775115, + "grad_norm": 1.6526265144348145, + "learning_rate": 1.572550895152756e-05, + "loss": 0.5076, + "step": 10362 + }, + { + "epoch": 0.3268569626241918, + "grad_norm": 1.4319201707839966, + "learning_rate": 1.572467137897744e-05, + "loss": 0.5315, + "step": 10363 + }, + { + "epoch": 0.3268885033906324, + "grad_norm": 1.6514618396759033, + "learning_rate": 1.572383374668593e-05, + "loss": 0.5311, + "step": 10364 + }, + { + "epoch": 0.32692004415707304, + "grad_norm": 1.453654170036316, + "learning_rate": 1.572299605466177e-05, + "loss": 0.5145, + "step": 10365 + }, + { + "epoch": 0.3269515849235136, + "grad_norm": 1.7499935626983643, + "learning_rate": 1.5722158302913705e-05, + "loss": 0.5725, + "step": 10366 + }, + { + "epoch": 0.32698312568995425, + "grad_norm": 2.1815226078033447, + "learning_rate": 1.572132049145047e-05, + "loss": 0.845, + "step": 10367 + }, + { + "epoch": 0.3270146664563949, + "grad_norm": 1.8927017450332642, + "learning_rate": 1.5720482620280818e-05, + "loss": 0.5831, + "step": 10368 + }, + { + "epoch": 0.3270462072228355, + "grad_norm": 1.661436915397644, + "learning_rate": 1.5719644689413483e-05, + "loss": 0.5786, + "step": 10369 + }, + { + "epoch": 0.32707774798927614, + "grad_norm": 3.6459710597991943, + "learning_rate": 1.5718806698857217e-05, + "loss": 0.5408, + "step": 10370 + }, + { + "epoch": 0.32710928875571676, + "grad_norm": 1.9308252334594727, + "learning_rate": 1.571796864862076e-05, + "loss": 0.547, + "step": 10371 + }, + { + "epoch": 0.3271408295221574, + "grad_norm": 1.4570738077163696, + "learning_rate": 1.5717130538712867e-05, + "loss": 0.5425, + "step": 10372 + }, + { + "epoch": 0.327172370288598, + "grad_norm": 1.4563102722167969, + "learning_rate": 1.5716292369142273e-05, + "loss": 0.5697, + "step": 10373 + }, + { + "epoch": 0.32720391105503865, + "grad_norm": 1.3919121026992798, + "learning_rate": 1.571545413991773e-05, + "loss": 0.5385, + "step": 10374 + }, + { + "epoch": 0.3272354518214793, + "grad_norm": 1.4725719690322876, + "learning_rate": 1.5714615851047986e-05, + "loss": 0.5562, + "step": 10375 + }, + { + "epoch": 0.3272669925879199, + "grad_norm": 1.6258639097213745, + "learning_rate": 1.5713777502541784e-05, + "loss": 0.5379, + "step": 10376 + }, + { + "epoch": 0.3272985333543605, + "grad_norm": 1.927556037902832, + "learning_rate": 1.5712939094407883e-05, + "loss": 0.5385, + "step": 10377 + }, + { + "epoch": 0.3273300741208011, + "grad_norm": 1.706718921661377, + "learning_rate": 1.5712100626655024e-05, + "loss": 0.5206, + "step": 10378 + }, + { + "epoch": 0.32736161488724175, + "grad_norm": 1.5289783477783203, + "learning_rate": 1.571126209929196e-05, + "loss": 0.5087, + "step": 10379 + }, + { + "epoch": 0.3273931556536824, + "grad_norm": 7.547660827636719, + "learning_rate": 1.5710423512327442e-05, + "loss": 1.3688, + "step": 10380 + }, + { + "epoch": 0.327424696420123, + "grad_norm": 4.250632286071777, + "learning_rate": 1.570958486577022e-05, + "loss": 0.7924, + "step": 10381 + }, + { + "epoch": 0.32745623718656364, + "grad_norm": 3.598278522491455, + "learning_rate": 1.5708746159629044e-05, + "loss": 0.6153, + "step": 10382 + }, + { + "epoch": 0.32748777795300427, + "grad_norm": 2.2880635261535645, + "learning_rate": 1.570790739391267e-05, + "loss": 0.5433, + "step": 10383 + }, + { + "epoch": 0.3275193187194449, + "grad_norm": 4.189234256744385, + "learning_rate": 1.5707068568629856e-05, + "loss": 0.5259, + "step": 10384 + }, + { + "epoch": 0.3275508594858855, + "grad_norm": 1.6748090982437134, + "learning_rate": 1.5706229683789344e-05, + "loss": 0.8501, + "step": 10385 + }, + { + "epoch": 0.32758240025232616, + "grad_norm": 1.6634012460708618, + "learning_rate": 1.5705390739399894e-05, + "loss": 0.531, + "step": 10386 + }, + { + "epoch": 0.32761394101876673, + "grad_norm": 1.7685208320617676, + "learning_rate": 1.5704551735470266e-05, + "loss": 0.8925, + "step": 10387 + }, + { + "epoch": 0.32764548178520736, + "grad_norm": 1.5365893840789795, + "learning_rate": 1.5703712672009208e-05, + "loss": 0.5417, + "step": 10388 + }, + { + "epoch": 0.327677022551648, + "grad_norm": 1.2277469635009766, + "learning_rate": 1.570287354902548e-05, + "loss": 0.4915, + "step": 10389 + }, + { + "epoch": 0.3277085633180886, + "grad_norm": 10.116520881652832, + "learning_rate": 1.570203436652784e-05, + "loss": 1.0838, + "step": 10390 + }, + { + "epoch": 0.32774010408452925, + "grad_norm": 2.7166218757629395, + "learning_rate": 1.5701195124525042e-05, + "loss": 0.5714, + "step": 10391 + }, + { + "epoch": 0.3277716448509699, + "grad_norm": 1.8479228019714355, + "learning_rate": 1.5700355823025845e-05, + "loss": 0.5105, + "step": 10392 + }, + { + "epoch": 0.3278031856174105, + "grad_norm": 2.030036211013794, + "learning_rate": 1.5699516462039012e-05, + "loss": 0.5471, + "step": 10393 + }, + { + "epoch": 0.32783472638385114, + "grad_norm": 2.26749849319458, + "learning_rate": 1.56986770415733e-05, + "loss": 0.5328, + "step": 10394 + }, + { + "epoch": 0.32786626715029177, + "grad_norm": 1.2221533060073853, + "learning_rate": 1.5697837561637463e-05, + "loss": 0.5008, + "step": 10395 + }, + { + "epoch": 0.3278978079167324, + "grad_norm": 1.4708038568496704, + "learning_rate": 1.569699802224027e-05, + "loss": 0.5151, + "step": 10396 + }, + { + "epoch": 0.327929348683173, + "grad_norm": 1.607167363166809, + "learning_rate": 1.5696158423390478e-05, + "loss": 0.544, + "step": 10397 + }, + { + "epoch": 0.3279608894496136, + "grad_norm": 3.265759229660034, + "learning_rate": 1.569531876509685e-05, + "loss": 0.9039, + "step": 10398 + }, + { + "epoch": 0.32799243021605423, + "grad_norm": 1.7487454414367676, + "learning_rate": 1.569447904736815e-05, + "loss": 0.5434, + "step": 10399 + }, + { + "epoch": 0.32802397098249486, + "grad_norm": 2.0664615631103516, + "learning_rate": 1.5693639270213138e-05, + "loss": 0.5285, + "step": 10400 + }, + { + "epoch": 0.3280555117489355, + "grad_norm": 5.727591037750244, + "learning_rate": 1.569279943364058e-05, + "loss": 0.8332, + "step": 10401 + }, + { + "epoch": 0.3280870525153761, + "grad_norm": 1.6016162633895874, + "learning_rate": 1.5691959537659238e-05, + "loss": 0.524, + "step": 10402 + }, + { + "epoch": 0.32811859328181675, + "grad_norm": 1.6751878261566162, + "learning_rate": 1.569111958227788e-05, + "loss": 0.4969, + "step": 10403 + }, + { + "epoch": 0.3281501340482574, + "grad_norm": 1.712382197380066, + "learning_rate": 1.5690279567505272e-05, + "loss": 0.5215, + "step": 10404 + }, + { + "epoch": 0.328181674814698, + "grad_norm": 1.3538814783096313, + "learning_rate": 1.5689439493350175e-05, + "loss": 0.5539, + "step": 10405 + }, + { + "epoch": 0.32821321558113864, + "grad_norm": 1.3212345838546753, + "learning_rate": 1.5688599359821356e-05, + "loss": 0.531, + "step": 10406 + }, + { + "epoch": 0.32824475634757927, + "grad_norm": 1.5846235752105713, + "learning_rate": 1.5687759166927592e-05, + "loss": 0.5215, + "step": 10407 + }, + { + "epoch": 0.32827629711401984, + "grad_norm": 1.957971453666687, + "learning_rate": 1.5686918914677642e-05, + "loss": 0.5269, + "step": 10408 + }, + { + "epoch": 0.3283078378804605, + "grad_norm": 1.4965004920959473, + "learning_rate": 1.5686078603080278e-05, + "loss": 0.5265, + "step": 10409 + }, + { + "epoch": 0.3283393786469011, + "grad_norm": 1.6281119585037231, + "learning_rate": 1.5685238232144267e-05, + "loss": 0.53, + "step": 10410 + }, + { + "epoch": 0.32837091941334173, + "grad_norm": 1.6724002361297607, + "learning_rate": 1.5684397801878383e-05, + "loss": 0.5062, + "step": 10411 + }, + { + "epoch": 0.32840246017978236, + "grad_norm": 1.428649663925171, + "learning_rate": 1.5683557312291394e-05, + "loss": 0.5139, + "step": 10412 + }, + { + "epoch": 0.328434000946223, + "grad_norm": 2.9413092136383057, + "learning_rate": 1.568271676339207e-05, + "loss": 0.5361, + "step": 10413 + }, + { + "epoch": 0.3284655417126636, + "grad_norm": 1.5795543193817139, + "learning_rate": 1.5681876155189183e-05, + "loss": 0.5232, + "step": 10414 + }, + { + "epoch": 0.32849708247910425, + "grad_norm": 2.627969980239868, + "learning_rate": 1.568103548769151e-05, + "loss": 0.8679, + "step": 10415 + }, + { + "epoch": 0.3285286232455449, + "grad_norm": 1.8560614585876465, + "learning_rate": 1.5680194760907816e-05, + "loss": 0.5372, + "step": 10416 + }, + { + "epoch": 0.3285601640119855, + "grad_norm": 2.029914140701294, + "learning_rate": 1.5679353974846884e-05, + "loss": 0.5173, + "step": 10417 + }, + { + "epoch": 0.32859170477842614, + "grad_norm": 1.9266724586486816, + "learning_rate": 1.5678513129517478e-05, + "loss": 0.8693, + "step": 10418 + }, + { + "epoch": 0.3286232455448667, + "grad_norm": 1.6851431131362915, + "learning_rate": 1.5677672224928384e-05, + "loss": 0.537, + "step": 10419 + }, + { + "epoch": 0.32865478631130735, + "grad_norm": 1.8712321519851685, + "learning_rate": 1.5676831261088365e-05, + "loss": 0.5285, + "step": 10420 + }, + { + "epoch": 0.328686327077748, + "grad_norm": 1.631922960281372, + "learning_rate": 1.567599023800621e-05, + "loss": 0.4984, + "step": 10421 + }, + { + "epoch": 0.3287178678441886, + "grad_norm": 2.2408814430236816, + "learning_rate": 1.567514915569069e-05, + "loss": 0.4604, + "step": 10422 + }, + { + "epoch": 0.32874940861062923, + "grad_norm": 1.2853622436523438, + "learning_rate": 1.5674308014150575e-05, + "loss": 0.499, + "step": 10423 + }, + { + "epoch": 0.32878094937706986, + "grad_norm": 2.8975882530212402, + "learning_rate": 1.5673466813394655e-05, + "loss": 0.517, + "step": 10424 + }, + { + "epoch": 0.3288124901435105, + "grad_norm": 2.3502037525177, + "learning_rate": 1.5672625553431702e-05, + "loss": 0.8722, + "step": 10425 + }, + { + "epoch": 0.3288440309099511, + "grad_norm": 1.5162420272827148, + "learning_rate": 1.5671784234270496e-05, + "loss": 0.5093, + "step": 10426 + }, + { + "epoch": 0.32887557167639175, + "grad_norm": 2.0917415618896484, + "learning_rate": 1.567094285591982e-05, + "loss": 0.871, + "step": 10427 + }, + { + "epoch": 0.3289071124428324, + "grad_norm": 1.6579433679580688, + "learning_rate": 1.5670101418388452e-05, + "loss": 0.5503, + "step": 10428 + }, + { + "epoch": 0.32893865320927296, + "grad_norm": 1.8593392372131348, + "learning_rate": 1.566925992168517e-05, + "loss": 0.8447, + "step": 10429 + }, + { + "epoch": 0.3289701939757136, + "grad_norm": 1.697512149810791, + "learning_rate": 1.566841836581876e-05, + "loss": 0.5595, + "step": 10430 + }, + { + "epoch": 0.3290017347421542, + "grad_norm": 1.721709966659546, + "learning_rate": 1.5667576750798004e-05, + "loss": 0.5193, + "step": 10431 + }, + { + "epoch": 0.32903327550859485, + "grad_norm": 2.092031955718994, + "learning_rate": 1.5666735076631683e-05, + "loss": 0.8772, + "step": 10432 + }, + { + "epoch": 0.3290648162750355, + "grad_norm": 1.7891931533813477, + "learning_rate": 1.5665893343328582e-05, + "loss": 0.5139, + "step": 10433 + }, + { + "epoch": 0.3290963570414761, + "grad_norm": 1.78189218044281, + "learning_rate": 1.5665051550897482e-05, + "loss": 0.5357, + "step": 10434 + }, + { + "epoch": 0.32912789780791674, + "grad_norm": 1.7588375806808472, + "learning_rate": 1.5664209699347176e-05, + "loss": 0.5073, + "step": 10435 + }, + { + "epoch": 0.32915943857435737, + "grad_norm": 1.2880429029464722, + "learning_rate": 1.5663367788686437e-05, + "loss": 0.5364, + "step": 10436 + }, + { + "epoch": 0.329190979340798, + "grad_norm": 2.3088910579681396, + "learning_rate": 1.566252581892406e-05, + "loss": 0.5297, + "step": 10437 + }, + { + "epoch": 0.3292225201072386, + "grad_norm": 1.7280720472335815, + "learning_rate": 1.566168379006883e-05, + "loss": 0.4864, + "step": 10438 + }, + { + "epoch": 0.32925406087367926, + "grad_norm": 1.6441600322723389, + "learning_rate": 1.5660841702129533e-05, + "loss": 0.5377, + "step": 10439 + }, + { + "epoch": 0.32928560164011983, + "grad_norm": 3.322307586669922, + "learning_rate": 1.5659999555114958e-05, + "loss": 0.8632, + "step": 10440 + }, + { + "epoch": 0.32931714240656046, + "grad_norm": 1.775944709777832, + "learning_rate": 1.565915734903389e-05, + "loss": 0.4968, + "step": 10441 + }, + { + "epoch": 0.3293486831730011, + "grad_norm": 1.9807274341583252, + "learning_rate": 1.5658315083895127e-05, + "loss": 0.4858, + "step": 10442 + }, + { + "epoch": 0.3293802239394417, + "grad_norm": 3.621051549911499, + "learning_rate": 1.5657472759707446e-05, + "loss": 0.5334, + "step": 10443 + }, + { + "epoch": 0.32941176470588235, + "grad_norm": 1.7285994291305542, + "learning_rate": 1.5656630376479647e-05, + "loss": 0.5308, + "step": 10444 + }, + { + "epoch": 0.329443305472323, + "grad_norm": 1.9697667360305786, + "learning_rate": 1.5655787934220516e-05, + "loss": 0.4761, + "step": 10445 + }, + { + "epoch": 0.3294748462387636, + "grad_norm": 1.4591376781463623, + "learning_rate": 1.565494543293885e-05, + "loss": 0.5237, + "step": 10446 + }, + { + "epoch": 0.32950638700520424, + "grad_norm": 1.8489124774932861, + "learning_rate": 1.5654102872643434e-05, + "loss": 0.5241, + "step": 10447 + }, + { + "epoch": 0.32953792777164487, + "grad_norm": 1.8059865236282349, + "learning_rate": 1.5653260253343064e-05, + "loss": 0.5094, + "step": 10448 + }, + { + "epoch": 0.3295694685380855, + "grad_norm": 1.6702241897583008, + "learning_rate": 1.5652417575046537e-05, + "loss": 0.5472, + "step": 10449 + }, + { + "epoch": 0.3296010093045261, + "grad_norm": 1.2914204597473145, + "learning_rate": 1.565157483776264e-05, + "loss": 0.5285, + "step": 10450 + }, + { + "epoch": 0.3296325500709667, + "grad_norm": 1.5619456768035889, + "learning_rate": 1.5650732041500173e-05, + "loss": 0.5046, + "step": 10451 + }, + { + "epoch": 0.32966409083740733, + "grad_norm": 1.4763940572738647, + "learning_rate": 1.564988918626793e-05, + "loss": 0.5636, + "step": 10452 + }, + { + "epoch": 0.32969563160384796, + "grad_norm": 1.5626620054244995, + "learning_rate": 1.5649046272074705e-05, + "loss": 0.5399, + "step": 10453 + }, + { + "epoch": 0.3297271723702886, + "grad_norm": 1.3794541358947754, + "learning_rate": 1.5648203298929296e-05, + "loss": 0.4842, + "step": 10454 + }, + { + "epoch": 0.3297587131367292, + "grad_norm": 1.5117337703704834, + "learning_rate": 1.5647360266840497e-05, + "loss": 0.4923, + "step": 10455 + }, + { + "epoch": 0.32979025390316985, + "grad_norm": 1.7325340509414673, + "learning_rate": 1.5646517175817114e-05, + "loss": 0.5475, + "step": 10456 + }, + { + "epoch": 0.3298217946696105, + "grad_norm": 1.4980690479278564, + "learning_rate": 1.5645674025867936e-05, + "loss": 0.5274, + "step": 10457 + }, + { + "epoch": 0.3298533354360511, + "grad_norm": 1.2751948833465576, + "learning_rate": 1.564483081700177e-05, + "loss": 0.5082, + "step": 10458 + }, + { + "epoch": 0.32988487620249174, + "grad_norm": 1.4131619930267334, + "learning_rate": 1.5643987549227407e-05, + "loss": 0.5311, + "step": 10459 + }, + { + "epoch": 0.32991641696893237, + "grad_norm": 1.2843576669692993, + "learning_rate": 1.5643144222553655e-05, + "loss": 0.5321, + "step": 10460 + }, + { + "epoch": 0.32994795773537294, + "grad_norm": 1.6046584844589233, + "learning_rate": 1.564230083698931e-05, + "loss": 0.537, + "step": 10461 + }, + { + "epoch": 0.3299794985018136, + "grad_norm": 1.674691915512085, + "learning_rate": 1.564145739254317e-05, + "loss": 0.4984, + "step": 10462 + }, + { + "epoch": 0.3300110392682542, + "grad_norm": 18.1663875579834, + "learning_rate": 1.5640613889224048e-05, + "loss": 1.1777, + "step": 10463 + }, + { + "epoch": 0.33004258003469483, + "grad_norm": 1.8202491998672485, + "learning_rate": 1.563977032704074e-05, + "loss": 0.542, + "step": 10464 + }, + { + "epoch": 0.33007412080113546, + "grad_norm": 1.7917509078979492, + "learning_rate": 1.5638926706002044e-05, + "loss": 0.5218, + "step": 10465 + }, + { + "epoch": 0.3301056615675761, + "grad_norm": 1.5989265441894531, + "learning_rate": 1.5638083026116772e-05, + "loss": 0.5313, + "step": 10466 + }, + { + "epoch": 0.3301372023340167, + "grad_norm": 2.4345412254333496, + "learning_rate": 1.5637239287393725e-05, + "loss": 0.8637, + "step": 10467 + }, + { + "epoch": 0.33016874310045735, + "grad_norm": 2.0432910919189453, + "learning_rate": 1.5636395489841713e-05, + "loss": 0.8683, + "step": 10468 + }, + { + "epoch": 0.330200283866898, + "grad_norm": 1.509250521659851, + "learning_rate": 1.5635551633469533e-05, + "loss": 0.5121, + "step": 10469 + }, + { + "epoch": 0.3302318246333386, + "grad_norm": 1.6847656965255737, + "learning_rate": 1.5634707718286e-05, + "loss": 0.5522, + "step": 10470 + }, + { + "epoch": 0.33026336539977924, + "grad_norm": 11.989360809326172, + "learning_rate": 1.5633863744299914e-05, + "loss": 1.1149, + "step": 10471 + }, + { + "epoch": 0.3302949061662198, + "grad_norm": 1.8164440393447876, + "learning_rate": 1.5633019711520083e-05, + "loss": 0.8687, + "step": 10472 + }, + { + "epoch": 0.33032644693266044, + "grad_norm": 1.6207791566848755, + "learning_rate": 1.563217561995532e-05, + "loss": 0.5002, + "step": 10473 + }, + { + "epoch": 0.3303579876991011, + "grad_norm": 1.9506367444992065, + "learning_rate": 1.563133146961443e-05, + "loss": 0.8884, + "step": 10474 + }, + { + "epoch": 0.3303895284655417, + "grad_norm": 2.9755334854125977, + "learning_rate": 1.5630487260506226e-05, + "loss": 0.8786, + "step": 10475 + }, + { + "epoch": 0.33042106923198233, + "grad_norm": 6.397338390350342, + "learning_rate": 1.5629642992639512e-05, + "loss": 0.6413, + "step": 10476 + }, + { + "epoch": 0.33045260999842296, + "grad_norm": 1.3165826797485352, + "learning_rate": 1.5628798666023107e-05, + "loss": 0.4975, + "step": 10477 + }, + { + "epoch": 0.3304841507648636, + "grad_norm": 1.7762216329574585, + "learning_rate": 1.5627954280665815e-05, + "loss": 0.5261, + "step": 10478 + }, + { + "epoch": 0.3305156915313042, + "grad_norm": 1.5384297370910645, + "learning_rate": 1.5627109836576447e-05, + "loss": 0.5202, + "step": 10479 + }, + { + "epoch": 0.33054723229774485, + "grad_norm": 1.5821648836135864, + "learning_rate": 1.5626265333763823e-05, + "loss": 0.5383, + "step": 10480 + }, + { + "epoch": 0.3305787730641855, + "grad_norm": 1.3708038330078125, + "learning_rate": 1.562542077223675e-05, + "loss": 0.5188, + "step": 10481 + }, + { + "epoch": 0.33061031383062606, + "grad_norm": 1.8250888586044312, + "learning_rate": 1.562457615200404e-05, + "loss": 0.5388, + "step": 10482 + }, + { + "epoch": 0.3306418545970667, + "grad_norm": 1.5285457372665405, + "learning_rate": 1.5623731473074514e-05, + "loss": 0.533, + "step": 10483 + }, + { + "epoch": 0.3306733953635073, + "grad_norm": 1.6415460109710693, + "learning_rate": 1.562288673545698e-05, + "loss": 0.5536, + "step": 10484 + }, + { + "epoch": 0.33070493612994795, + "grad_norm": 2.085577964782715, + "learning_rate": 1.562204193916026e-05, + "loss": 0.5206, + "step": 10485 + }, + { + "epoch": 0.3307364768963886, + "grad_norm": 1.3987140655517578, + "learning_rate": 1.5621197084193168e-05, + "loss": 0.5369, + "step": 10486 + }, + { + "epoch": 0.3307680176628292, + "grad_norm": 1.983276605606079, + "learning_rate": 1.562035217056452e-05, + "loss": 0.844, + "step": 10487 + }, + { + "epoch": 0.33079955842926984, + "grad_norm": 1.7341476678848267, + "learning_rate": 1.561950719828313e-05, + "loss": 0.5356, + "step": 10488 + }, + { + "epoch": 0.33083109919571047, + "grad_norm": 3.9859397411346436, + "learning_rate": 1.561866216735782e-05, + "loss": 0.7913, + "step": 10489 + }, + { + "epoch": 0.3308626399621511, + "grad_norm": 1.7878013849258423, + "learning_rate": 1.5617817077797406e-05, + "loss": 0.8694, + "step": 10490 + }, + { + "epoch": 0.3308941807285917, + "grad_norm": 1.4928232431411743, + "learning_rate": 1.561697192961071e-05, + "loss": 0.5502, + "step": 10491 + }, + { + "epoch": 0.33092572149503235, + "grad_norm": 1.9133788347244263, + "learning_rate": 1.5616126722806552e-05, + "loss": 0.5309, + "step": 10492 + }, + { + "epoch": 0.33095726226147293, + "grad_norm": 2.1405029296875, + "learning_rate": 1.561528145739375e-05, + "loss": 0.5165, + "step": 10493 + }, + { + "epoch": 0.33098880302791356, + "grad_norm": 1.522963523864746, + "learning_rate": 1.5614436133381125e-05, + "loss": 0.5275, + "step": 10494 + }, + { + "epoch": 0.3310203437943542, + "grad_norm": 1.6257390975952148, + "learning_rate": 1.5613590750777497e-05, + "loss": 0.4782, + "step": 10495 + }, + { + "epoch": 0.3310518845607948, + "grad_norm": 1.9671250581741333, + "learning_rate": 1.5612745309591698e-05, + "loss": 0.5227, + "step": 10496 + }, + { + "epoch": 0.33108342532723545, + "grad_norm": 1.4776943922042847, + "learning_rate": 1.5611899809832537e-05, + "loss": 0.5112, + "step": 10497 + }, + { + "epoch": 0.3311149660936761, + "grad_norm": 1.8223903179168701, + "learning_rate": 1.5611054251508847e-05, + "loss": 0.5188, + "step": 10498 + }, + { + "epoch": 0.3311465068601167, + "grad_norm": 1.427954912185669, + "learning_rate": 1.5610208634629444e-05, + "loss": 0.5029, + "step": 10499 + }, + { + "epoch": 0.33117804762655734, + "grad_norm": 1.687726378440857, + "learning_rate": 1.5609362959203162e-05, + "loss": 0.5297, + "step": 10500 + }, + { + "epoch": 0.33120958839299797, + "grad_norm": 1.4072015285491943, + "learning_rate": 1.5608517225238825e-05, + "loss": 0.5178, + "step": 10501 + }, + { + "epoch": 0.3312411291594386, + "grad_norm": 1.658516764640808, + "learning_rate": 1.560767143274525e-05, + "loss": 0.501, + "step": 10502 + }, + { + "epoch": 0.3312726699258792, + "grad_norm": 1.532775640487671, + "learning_rate": 1.5606825581731274e-05, + "loss": 0.5472, + "step": 10503 + }, + { + "epoch": 0.3313042106923198, + "grad_norm": 1.7437448501586914, + "learning_rate": 1.5605979672205718e-05, + "loss": 0.5247, + "step": 10504 + }, + { + "epoch": 0.33133575145876043, + "grad_norm": 2.3545892238616943, + "learning_rate": 1.560513370417741e-05, + "loss": 0.8785, + "step": 10505 + }, + { + "epoch": 0.33136729222520106, + "grad_norm": 1.5549638271331787, + "learning_rate": 1.5604287677655177e-05, + "loss": 0.5373, + "step": 10506 + }, + { + "epoch": 0.3313988329916417, + "grad_norm": 2.4261231422424316, + "learning_rate": 1.5603441592647858e-05, + "loss": 0.4791, + "step": 10507 + }, + { + "epoch": 0.3314303737580823, + "grad_norm": 2.0294692516326904, + "learning_rate": 1.5602595449164267e-05, + "loss": 0.5547, + "step": 10508 + }, + { + "epoch": 0.33146191452452295, + "grad_norm": 1.8943651914596558, + "learning_rate": 1.5601749247213244e-05, + "loss": 0.5104, + "step": 10509 + }, + { + "epoch": 0.3314934552909636, + "grad_norm": 2.6714632511138916, + "learning_rate": 1.560090298680362e-05, + "loss": 0.5222, + "step": 10510 + }, + { + "epoch": 0.3315249960574042, + "grad_norm": 1.4363698959350586, + "learning_rate": 1.5600056667944225e-05, + "loss": 0.5588, + "step": 10511 + }, + { + "epoch": 0.33155653682384484, + "grad_norm": 1.457722544670105, + "learning_rate": 1.5599210290643888e-05, + "loss": 0.5418, + "step": 10512 + }, + { + "epoch": 0.33158807759028547, + "grad_norm": 2.0035226345062256, + "learning_rate": 1.5598363854911446e-05, + "loss": 0.4885, + "step": 10513 + }, + { + "epoch": 0.33161961835672604, + "grad_norm": 1.291238784790039, + "learning_rate": 1.559751736075573e-05, + "loss": 0.5006, + "step": 10514 + }, + { + "epoch": 0.33165115912316667, + "grad_norm": 1.405013084411621, + "learning_rate": 1.5596670808185574e-05, + "loss": 0.5232, + "step": 10515 + }, + { + "epoch": 0.3316826998896073, + "grad_norm": 4.590391159057617, + "learning_rate": 1.559582419720981e-05, + "loss": 0.4167, + "step": 10516 + }, + { + "epoch": 0.33171424065604793, + "grad_norm": 2.4216620922088623, + "learning_rate": 1.5594977527837278e-05, + "loss": 0.5007, + "step": 10517 + }, + { + "epoch": 0.33174578142248856, + "grad_norm": 1.8126823902130127, + "learning_rate": 1.559413080007681e-05, + "loss": 0.8745, + "step": 10518 + }, + { + "epoch": 0.3317773221889292, + "grad_norm": 2.12021541595459, + "learning_rate": 1.5593284013937244e-05, + "loss": 0.558, + "step": 10519 + }, + { + "epoch": 0.3318088629553698, + "grad_norm": 1.3704313039779663, + "learning_rate": 1.5592437169427418e-05, + "loss": 0.5297, + "step": 10520 + }, + { + "epoch": 0.33184040372181045, + "grad_norm": 7.476769924163818, + "learning_rate": 1.5591590266556164e-05, + "loss": 0.4241, + "step": 10521 + }, + { + "epoch": 0.3318719444882511, + "grad_norm": 1.9154154062271118, + "learning_rate": 1.5590743305332325e-05, + "loss": 0.578, + "step": 10522 + }, + { + "epoch": 0.3319034852546917, + "grad_norm": 1.635193943977356, + "learning_rate": 1.558989628576474e-05, + "loss": 0.5053, + "step": 10523 + }, + { + "epoch": 0.33193502602113234, + "grad_norm": 1.9311000108718872, + "learning_rate": 1.5589049207862244e-05, + "loss": 0.5338, + "step": 10524 + }, + { + "epoch": 0.3319665667875729, + "grad_norm": 2.4345860481262207, + "learning_rate": 1.558820207163368e-05, + "loss": 0.5419, + "step": 10525 + }, + { + "epoch": 0.33199810755401354, + "grad_norm": 1.8035153150558472, + "learning_rate": 1.5587354877087892e-05, + "loss": 0.4435, + "step": 10526 + }, + { + "epoch": 0.3320296483204542, + "grad_norm": 1.3514187335968018, + "learning_rate": 1.5586507624233714e-05, + "loss": 0.4995, + "step": 10527 + }, + { + "epoch": 0.3320611890868948, + "grad_norm": 1.771273136138916, + "learning_rate": 1.5585660313079992e-05, + "loss": 0.5105, + "step": 10528 + }, + { + "epoch": 0.33209272985333543, + "grad_norm": 1.816629409790039, + "learning_rate": 1.5584812943635567e-05, + "loss": 0.5184, + "step": 10529 + }, + { + "epoch": 0.33212427061977606, + "grad_norm": 1.7690175771713257, + "learning_rate": 1.5583965515909285e-05, + "loss": 0.5196, + "step": 10530 + }, + { + "epoch": 0.3321558113862167, + "grad_norm": 10.174489974975586, + "learning_rate": 1.5583118029909986e-05, + "loss": 0.4073, + "step": 10531 + }, + { + "epoch": 0.3321873521526573, + "grad_norm": 1.433706283569336, + "learning_rate": 1.5582270485646516e-05, + "loss": 0.5347, + "step": 10532 + }, + { + "epoch": 0.33221889291909795, + "grad_norm": 1.467451572418213, + "learning_rate": 1.5581422883127714e-05, + "loss": 0.3804, + "step": 10533 + }, + { + "epoch": 0.3322504336855386, + "grad_norm": 2.11985182762146, + "learning_rate": 1.5580575222362435e-05, + "loss": 0.5265, + "step": 10534 + }, + { + "epoch": 0.33228197445197916, + "grad_norm": 1.631434440612793, + "learning_rate": 1.557972750335952e-05, + "loss": 0.5558, + "step": 10535 + }, + { + "epoch": 0.3323135152184198, + "grad_norm": 1.6640150547027588, + "learning_rate": 1.5578879726127817e-05, + "loss": 0.5333, + "step": 10536 + }, + { + "epoch": 0.3323450559848604, + "grad_norm": 1.733594536781311, + "learning_rate": 1.5578031890676168e-05, + "loss": 0.5014, + "step": 10537 + }, + { + "epoch": 0.33237659675130105, + "grad_norm": 2.0096921920776367, + "learning_rate": 1.5577183997013426e-05, + "loss": 0.4973, + "step": 10538 + }, + { + "epoch": 0.3324081375177417, + "grad_norm": 18.739500045776367, + "learning_rate": 1.5576336045148442e-05, + "loss": 0.4311, + "step": 10539 + }, + { + "epoch": 0.3324396782841823, + "grad_norm": 1.8125483989715576, + "learning_rate": 1.557548803509006e-05, + "loss": 0.551, + "step": 10540 + }, + { + "epoch": 0.33247121905062293, + "grad_norm": 1.3893569707870483, + "learning_rate": 1.5574639966847128e-05, + "loss": 0.5133, + "step": 10541 + }, + { + "epoch": 0.33250275981706356, + "grad_norm": 1.6327168941497803, + "learning_rate": 1.55737918404285e-05, + "loss": 0.5296, + "step": 10542 + }, + { + "epoch": 0.3325343005835042, + "grad_norm": 1.542576789855957, + "learning_rate": 1.5572943655843027e-05, + "loss": 0.5365, + "step": 10543 + }, + { + "epoch": 0.3325658413499448, + "grad_norm": 1.429848074913025, + "learning_rate": 1.557209541309956e-05, + "loss": 0.5356, + "step": 10544 + }, + { + "epoch": 0.33259738211638545, + "grad_norm": 1.4287859201431274, + "learning_rate": 1.5571247112206952e-05, + "loss": 0.5211, + "step": 10545 + }, + { + "epoch": 0.33262892288282603, + "grad_norm": 1.4116131067276, + "learning_rate": 1.5570398753174053e-05, + "loss": 0.8897, + "step": 10546 + }, + { + "epoch": 0.33266046364926666, + "grad_norm": 1.4875006675720215, + "learning_rate": 1.5569550336009717e-05, + "loss": 0.518, + "step": 10547 + }, + { + "epoch": 0.3326920044157073, + "grad_norm": 1.7652541399002075, + "learning_rate": 1.55687018607228e-05, + "loss": 0.5199, + "step": 10548 + }, + { + "epoch": 0.3327235451821479, + "grad_norm": 1.482079267501831, + "learning_rate": 1.556785332732216e-05, + "loss": 0.5527, + "step": 10549 + }, + { + "epoch": 0.33275508594858855, + "grad_norm": 1.5544534921646118, + "learning_rate": 1.556700473581664e-05, + "loss": 0.519, + "step": 10550 + }, + { + "epoch": 0.3327866267150292, + "grad_norm": 1.9269633293151855, + "learning_rate": 1.5566156086215106e-05, + "loss": 0.4435, + "step": 10551 + }, + { + "epoch": 0.3328181674814698, + "grad_norm": 1.4031472206115723, + "learning_rate": 1.5565307378526407e-05, + "loss": 0.5482, + "step": 10552 + }, + { + "epoch": 0.33284970824791044, + "grad_norm": 2.1224424839019775, + "learning_rate": 1.556445861275941e-05, + "loss": 0.4295, + "step": 10553 + }, + { + "epoch": 0.33288124901435107, + "grad_norm": 1.573742389678955, + "learning_rate": 1.5563609788922966e-05, + "loss": 0.399, + "step": 10554 + }, + { + "epoch": 0.3329127897807917, + "grad_norm": 1.5249112844467163, + "learning_rate": 1.556276090702593e-05, + "loss": 0.5376, + "step": 10555 + }, + { + "epoch": 0.33294433054723227, + "grad_norm": 1.529524564743042, + "learning_rate": 1.5561911967077165e-05, + "loss": 0.5481, + "step": 10556 + }, + { + "epoch": 0.3329758713136729, + "grad_norm": 1.5923882722854614, + "learning_rate": 1.556106296908553e-05, + "loss": 0.5341, + "step": 10557 + }, + { + "epoch": 0.33300741208011353, + "grad_norm": 2.527345895767212, + "learning_rate": 1.5560213913059887e-05, + "loss": 0.533, + "step": 10558 + }, + { + "epoch": 0.33303895284655416, + "grad_norm": 1.4876295328140259, + "learning_rate": 1.5559364799009097e-05, + "loss": 0.5503, + "step": 10559 + }, + { + "epoch": 0.3330704936129948, + "grad_norm": 1.426844596862793, + "learning_rate": 1.555851562694201e-05, + "loss": 0.5518, + "step": 10560 + }, + { + "epoch": 0.3331020343794354, + "grad_norm": 1.4544488191604614, + "learning_rate": 1.5557666396867505e-05, + "loss": 0.5326, + "step": 10561 + }, + { + "epoch": 0.33313357514587605, + "grad_norm": 1.5252628326416016, + "learning_rate": 1.5556817108794433e-05, + "loss": 0.5468, + "step": 10562 + }, + { + "epoch": 0.3331651159123167, + "grad_norm": 1.6620824337005615, + "learning_rate": 1.555596776273166e-05, + "loss": 0.5185, + "step": 10563 + }, + { + "epoch": 0.3331966566787573, + "grad_norm": 1.4908488988876343, + "learning_rate": 1.5555118358688047e-05, + "loss": 0.5148, + "step": 10564 + }, + { + "epoch": 0.33322819744519794, + "grad_norm": 1.4539947509765625, + "learning_rate": 1.5554268896672463e-05, + "loss": 0.5573, + "step": 10565 + }, + { + "epoch": 0.33325973821163857, + "grad_norm": 1.4122611284255981, + "learning_rate": 1.5553419376693767e-05, + "loss": 0.5586, + "step": 10566 + }, + { + "epoch": 0.33329127897807914, + "grad_norm": 1.4987592697143555, + "learning_rate": 1.555256979876083e-05, + "loss": 0.5173, + "step": 10567 + }, + { + "epoch": 0.33332281974451977, + "grad_norm": 1.381272792816162, + "learning_rate": 1.5551720162882515e-05, + "loss": 0.3988, + "step": 10568 + }, + { + "epoch": 0.3333543605109604, + "grad_norm": 1.4218072891235352, + "learning_rate": 1.555087046906769e-05, + "loss": 0.5017, + "step": 10569 + }, + { + "epoch": 0.33338590127740103, + "grad_norm": 1.662571907043457, + "learning_rate": 1.5550020717325223e-05, + "loss": 0.521, + "step": 10570 + }, + { + "epoch": 0.33341744204384166, + "grad_norm": 1.7561546564102173, + "learning_rate": 1.5549170907663977e-05, + "loss": 0.5195, + "step": 10571 + }, + { + "epoch": 0.3334489828102823, + "grad_norm": 1.3889737129211426, + "learning_rate": 1.5548321040092826e-05, + "loss": 0.4902, + "step": 10572 + }, + { + "epoch": 0.3334805235767229, + "grad_norm": 5.294985771179199, + "learning_rate": 1.5547471114620634e-05, + "loss": 0.5116, + "step": 10573 + }, + { + "epoch": 0.33351206434316355, + "grad_norm": 1.6240116357803345, + "learning_rate": 1.554662113125628e-05, + "loss": 0.4791, + "step": 10574 + }, + { + "epoch": 0.3335436051096042, + "grad_norm": 1.9511752128601074, + "learning_rate": 1.5545771090008625e-05, + "loss": 0.5648, + "step": 10575 + }, + { + "epoch": 0.3335751458760448, + "grad_norm": 1.8497953414916992, + "learning_rate": 1.5544920990886537e-05, + "loss": 0.5283, + "step": 10576 + }, + { + "epoch": 0.33360668664248544, + "grad_norm": 1.8665165901184082, + "learning_rate": 1.5544070833898895e-05, + "loss": 0.523, + "step": 10577 + }, + { + "epoch": 0.333638227408926, + "grad_norm": 1.8418372869491577, + "learning_rate": 1.5543220619054574e-05, + "loss": 0.5461, + "step": 10578 + }, + { + "epoch": 0.33366976817536664, + "grad_norm": 2.448674440383911, + "learning_rate": 1.5542370346362438e-05, + "loss": 0.4374, + "step": 10579 + }, + { + "epoch": 0.3337013089418073, + "grad_norm": 2.2703182697296143, + "learning_rate": 1.5541520015831366e-05, + "loss": 0.5374, + "step": 10580 + }, + { + "epoch": 0.3337328497082479, + "grad_norm": 2.7588112354278564, + "learning_rate": 1.5540669627470227e-05, + "loss": 0.5472, + "step": 10581 + }, + { + "epoch": 0.33376439047468853, + "grad_norm": 1.8536869287490845, + "learning_rate": 1.55398191812879e-05, + "loss": 0.5453, + "step": 10582 + }, + { + "epoch": 0.33379593124112916, + "grad_norm": 1.7088872194290161, + "learning_rate": 1.5538968677293254e-05, + "loss": 0.5561, + "step": 10583 + }, + { + "epoch": 0.3338274720075698, + "grad_norm": 1.382057547569275, + "learning_rate": 1.5538118115495174e-05, + "loss": 0.4749, + "step": 10584 + }, + { + "epoch": 0.3338590127740104, + "grad_norm": 1.7828532457351685, + "learning_rate": 1.5537267495902528e-05, + "loss": 0.572, + "step": 10585 + }, + { + "epoch": 0.33389055354045105, + "grad_norm": 2.0942599773406982, + "learning_rate": 1.55364168185242e-05, + "loss": 0.5364, + "step": 10586 + }, + { + "epoch": 0.3339220943068917, + "grad_norm": 1.6983362436294556, + "learning_rate": 1.5535566083369063e-05, + "loss": 0.5406, + "step": 10587 + }, + { + "epoch": 0.33395363507333226, + "grad_norm": 1.8362526893615723, + "learning_rate": 1.553471529044599e-05, + "loss": 0.5092, + "step": 10588 + }, + { + "epoch": 0.3339851758397729, + "grad_norm": 2.083268642425537, + "learning_rate": 1.5533864439763872e-05, + "loss": 0.5348, + "step": 10589 + }, + { + "epoch": 0.3340167166062135, + "grad_norm": 1.50175940990448, + "learning_rate": 1.553301353133158e-05, + "loss": 0.5099, + "step": 10590 + }, + { + "epoch": 0.33404825737265414, + "grad_norm": 2.0515291690826416, + "learning_rate": 1.5532162565157993e-05, + "loss": 0.4945, + "step": 10591 + }, + { + "epoch": 0.3340797981390948, + "grad_norm": 1.5433292388916016, + "learning_rate": 1.5531311541251995e-05, + "loss": 0.4725, + "step": 10592 + }, + { + "epoch": 0.3341113389055354, + "grad_norm": 2.3818819522857666, + "learning_rate": 1.5530460459622466e-05, + "loss": 0.5415, + "step": 10593 + }, + { + "epoch": 0.33414287967197603, + "grad_norm": 1.668082356452942, + "learning_rate": 1.5529609320278288e-05, + "loss": 0.5095, + "step": 10594 + }, + { + "epoch": 0.33417442043841666, + "grad_norm": 1.5107084512710571, + "learning_rate": 1.5528758123228347e-05, + "loss": 0.5404, + "step": 10595 + }, + { + "epoch": 0.3342059612048573, + "grad_norm": 1.3001071214675903, + "learning_rate": 1.5527906868481516e-05, + "loss": 0.5111, + "step": 10596 + }, + { + "epoch": 0.3342375019712979, + "grad_norm": 1.5383195877075195, + "learning_rate": 1.5527055556046687e-05, + "loss": 0.5207, + "step": 10597 + }, + { + "epoch": 0.33426904273773855, + "grad_norm": 1.5189733505249023, + "learning_rate": 1.5526204185932742e-05, + "loss": 0.5201, + "step": 10598 + }, + { + "epoch": 0.3343005835041791, + "grad_norm": 1.7185382843017578, + "learning_rate": 1.5525352758148567e-05, + "loss": 0.5297, + "step": 10599 + }, + { + "epoch": 0.33433212427061976, + "grad_norm": 1.4072433710098267, + "learning_rate": 1.5524501272703046e-05, + "loss": 0.5137, + "step": 10600 + }, + { + "epoch": 0.3343636650370604, + "grad_norm": 1.3548105955123901, + "learning_rate": 1.552364972960506e-05, + "loss": 0.5248, + "step": 10601 + }, + { + "epoch": 0.334395205803501, + "grad_norm": 1.976363182067871, + "learning_rate": 1.55227981288635e-05, + "loss": 0.5096, + "step": 10602 + }, + { + "epoch": 0.33442674656994165, + "grad_norm": 2.270641565322876, + "learning_rate": 1.552194647048726e-05, + "loss": 0.5333, + "step": 10603 + }, + { + "epoch": 0.3344582873363823, + "grad_norm": 2.1299407482147217, + "learning_rate": 1.5521094754485216e-05, + "loss": 0.4479, + "step": 10604 + }, + { + "epoch": 0.3344898281028229, + "grad_norm": 1.6971454620361328, + "learning_rate": 1.552024298086626e-05, + "loss": 0.5363, + "step": 10605 + }, + { + "epoch": 0.33452136886926354, + "grad_norm": 1.6520905494689941, + "learning_rate": 1.551939114963928e-05, + "loss": 0.5374, + "step": 10606 + }, + { + "epoch": 0.33455290963570417, + "grad_norm": 1.665865182876587, + "learning_rate": 1.5518539260813175e-05, + "loss": 0.5271, + "step": 10607 + }, + { + "epoch": 0.3345844504021448, + "grad_norm": 1.9564826488494873, + "learning_rate": 1.5517687314396824e-05, + "loss": 0.5191, + "step": 10608 + }, + { + "epoch": 0.33461599116858537, + "grad_norm": 2.163348436355591, + "learning_rate": 1.5516835310399123e-05, + "loss": 0.3919, + "step": 10609 + }, + { + "epoch": 0.334647531935026, + "grad_norm": 1.4496361017227173, + "learning_rate": 1.551598324882896e-05, + "loss": 0.3802, + "step": 10610 + }, + { + "epoch": 0.33467907270146663, + "grad_norm": 1.5559948682785034, + "learning_rate": 1.551513112969523e-05, + "loss": 0.5083, + "step": 10611 + }, + { + "epoch": 0.33471061346790726, + "grad_norm": 2.026461124420166, + "learning_rate": 1.551427895300682e-05, + "loss": 0.5356, + "step": 10612 + }, + { + "epoch": 0.3347421542343479, + "grad_norm": 1.7508909702301025, + "learning_rate": 1.5513426718772634e-05, + "loss": 0.5288, + "step": 10613 + }, + { + "epoch": 0.3347736950007885, + "grad_norm": 1.716176986694336, + "learning_rate": 1.5512574427001557e-05, + "loss": 0.526, + "step": 10614 + }, + { + "epoch": 0.33480523576722915, + "grad_norm": 1.3934955596923828, + "learning_rate": 1.5511722077702486e-05, + "loss": 0.8315, + "step": 10615 + }, + { + "epoch": 0.3348367765336698, + "grad_norm": 1.462982416152954, + "learning_rate": 1.551086967088431e-05, + "loss": 0.5337, + "step": 10616 + }, + { + "epoch": 0.3348683173001104, + "grad_norm": 2.1789534091949463, + "learning_rate": 1.5510017206555935e-05, + "loss": 0.5348, + "step": 10617 + }, + { + "epoch": 0.33489985806655104, + "grad_norm": 2.496229410171509, + "learning_rate": 1.550916468472625e-05, + "loss": 0.5487, + "step": 10618 + }, + { + "epoch": 0.33493139883299167, + "grad_norm": 1.6329981088638306, + "learning_rate": 1.5508312105404155e-05, + "loss": 0.5033, + "step": 10619 + }, + { + "epoch": 0.33496293959943224, + "grad_norm": 2.350545644760132, + "learning_rate": 1.5507459468598547e-05, + "loss": 0.5727, + "step": 10620 + }, + { + "epoch": 0.33499448036587287, + "grad_norm": 1.490217685699463, + "learning_rate": 1.550660677431832e-05, + "loss": 0.5122, + "step": 10621 + }, + { + "epoch": 0.3350260211323135, + "grad_norm": 1.371787667274475, + "learning_rate": 1.550575402257238e-05, + "loss": 0.5376, + "step": 10622 + }, + { + "epoch": 0.33505756189875413, + "grad_norm": 1.441479206085205, + "learning_rate": 1.550490121336962e-05, + "loss": 0.5303, + "step": 10623 + }, + { + "epoch": 0.33508910266519476, + "grad_norm": 1.425869345664978, + "learning_rate": 1.5504048346718938e-05, + "loss": 0.5134, + "step": 10624 + }, + { + "epoch": 0.3351206434316354, + "grad_norm": 1.552415132522583, + "learning_rate": 1.550319542262924e-05, + "loss": 0.5182, + "step": 10625 + }, + { + "epoch": 0.335152184198076, + "grad_norm": 1.5332525968551636, + "learning_rate": 1.5502342441109423e-05, + "loss": 0.5413, + "step": 10626 + }, + { + "epoch": 0.33518372496451665, + "grad_norm": 2.2415194511413574, + "learning_rate": 1.550148940216839e-05, + "loss": 0.5455, + "step": 10627 + }, + { + "epoch": 0.3352152657309573, + "grad_norm": 1.587587594985962, + "learning_rate": 1.550063630581505e-05, + "loss": 0.5134, + "step": 10628 + }, + { + "epoch": 0.3352468064973979, + "grad_norm": 1.9881870746612549, + "learning_rate": 1.5499783152058295e-05, + "loss": 0.5437, + "step": 10629 + }, + { + "epoch": 0.33527834726383854, + "grad_norm": 1.264998197555542, + "learning_rate": 1.549892994090703e-05, + "loss": 0.5026, + "step": 10630 + }, + { + "epoch": 0.3353098880302791, + "grad_norm": 4.593145847320557, + "learning_rate": 1.5498076672370164e-05, + "loss": 0.5513, + "step": 10631 + }, + { + "epoch": 0.33534142879671974, + "grad_norm": 1.7125152349472046, + "learning_rate": 1.5497223346456595e-05, + "loss": 0.4987, + "step": 10632 + }, + { + "epoch": 0.3353729695631604, + "grad_norm": 1.5225279331207275, + "learning_rate": 1.5496369963175237e-05, + "loss": 0.5318, + "step": 10633 + }, + { + "epoch": 0.335404510329601, + "grad_norm": 1.597045660018921, + "learning_rate": 1.549551652253499e-05, + "loss": 0.5061, + "step": 10634 + }, + { + "epoch": 0.33543605109604163, + "grad_norm": 1.5313332080841064, + "learning_rate": 1.5494663024544756e-05, + "loss": 0.5174, + "step": 10635 + }, + { + "epoch": 0.33546759186248226, + "grad_norm": 1.5107423067092896, + "learning_rate": 1.549380946921345e-05, + "loss": 0.5054, + "step": 10636 + }, + { + "epoch": 0.3354991326289229, + "grad_norm": 1.9093620777130127, + "learning_rate": 1.5492955856549975e-05, + "loss": 0.5212, + "step": 10637 + }, + { + "epoch": 0.3355306733953635, + "grad_norm": 1.776673674583435, + "learning_rate": 1.5492102186563244e-05, + "loss": 0.4447, + "step": 10638 + }, + { + "epoch": 0.33556221416180415, + "grad_norm": 1.8728969097137451, + "learning_rate": 1.5491248459262158e-05, + "loss": 0.8768, + "step": 10639 + }, + { + "epoch": 0.3355937549282448, + "grad_norm": 1.4708433151245117, + "learning_rate": 1.5490394674655634e-05, + "loss": 0.8767, + "step": 10640 + }, + { + "epoch": 0.33562529569468535, + "grad_norm": 2.2644307613372803, + "learning_rate": 1.5489540832752576e-05, + "loss": 0.5403, + "step": 10641 + }, + { + "epoch": 0.335656836461126, + "grad_norm": 1.8183863162994385, + "learning_rate": 1.5488686933561897e-05, + "loss": 0.5546, + "step": 10642 + }, + { + "epoch": 0.3356883772275666, + "grad_norm": 1.581255555152893, + "learning_rate": 1.548783297709251e-05, + "loss": 0.5277, + "step": 10643 + }, + { + "epoch": 0.33571991799400724, + "grad_norm": 1.626944661140442, + "learning_rate": 1.548697896335332e-05, + "loss": 0.5038, + "step": 10644 + }, + { + "epoch": 0.3357514587604479, + "grad_norm": 1.5344979763031006, + "learning_rate": 1.5486124892353248e-05, + "loss": 0.5258, + "step": 10645 + }, + { + "epoch": 0.3357829995268885, + "grad_norm": 1.3917869329452515, + "learning_rate": 1.5485270764101202e-05, + "loss": 0.5432, + "step": 10646 + }, + { + "epoch": 0.33581454029332913, + "grad_norm": 1.7017126083374023, + "learning_rate": 1.5484416578606096e-05, + "loss": 0.5485, + "step": 10647 + }, + { + "epoch": 0.33584608105976976, + "grad_norm": 1.715828776359558, + "learning_rate": 1.5483562335876846e-05, + "loss": 0.5761, + "step": 10648 + }, + { + "epoch": 0.3358776218262104, + "grad_norm": 1.386440396308899, + "learning_rate": 1.5482708035922364e-05, + "loss": 0.5141, + "step": 10649 + }, + { + "epoch": 0.335909162592651, + "grad_norm": 1.908083438873291, + "learning_rate": 1.548185367875157e-05, + "loss": 0.5151, + "step": 10650 + }, + { + "epoch": 0.33594070335909165, + "grad_norm": 1.7639917135238647, + "learning_rate": 1.548099926437337e-05, + "loss": 0.5049, + "step": 10651 + }, + { + "epoch": 0.3359722441255322, + "grad_norm": 1.7645723819732666, + "learning_rate": 1.548014479279669e-05, + "loss": 0.5323, + "step": 10652 + }, + { + "epoch": 0.33600378489197286, + "grad_norm": 1.5767371654510498, + "learning_rate": 1.547929026403044e-05, + "loss": 0.5255, + "step": 10653 + }, + { + "epoch": 0.3360353256584135, + "grad_norm": 1.4229896068572998, + "learning_rate": 1.547843567808355e-05, + "loss": 0.5182, + "step": 10654 + }, + { + "epoch": 0.3360668664248541, + "grad_norm": 1.5072957277297974, + "learning_rate": 1.5477581034964924e-05, + "loss": 0.5171, + "step": 10655 + }, + { + "epoch": 0.33609840719129475, + "grad_norm": 1.673652172088623, + "learning_rate": 1.5476726334683485e-05, + "loss": 0.5264, + "step": 10656 + }, + { + "epoch": 0.3361299479577354, + "grad_norm": 2.375107526779175, + "learning_rate": 1.547587157724816e-05, + "loss": 0.5301, + "step": 10657 + }, + { + "epoch": 0.336161488724176, + "grad_norm": 1.412239909172058, + "learning_rate": 1.547501676266786e-05, + "loss": 0.5321, + "step": 10658 + }, + { + "epoch": 0.33619302949061663, + "grad_norm": 1.6774624586105347, + "learning_rate": 1.547416189095151e-05, + "loss": 0.5407, + "step": 10659 + }, + { + "epoch": 0.33622457025705726, + "grad_norm": 2.7015180587768555, + "learning_rate": 1.5473306962108024e-05, + "loss": 0.5093, + "step": 10660 + }, + { + "epoch": 0.3362561110234979, + "grad_norm": 1.3587427139282227, + "learning_rate": 1.5472451976146336e-05, + "loss": 0.5147, + "step": 10661 + }, + { + "epoch": 0.33628765178993847, + "grad_norm": 1.7407951354980469, + "learning_rate": 1.5471596933075362e-05, + "loss": 0.5462, + "step": 10662 + }, + { + "epoch": 0.3363191925563791, + "grad_norm": 1.3145153522491455, + "learning_rate": 1.5470741832904024e-05, + "loss": 0.5283, + "step": 10663 + }, + { + "epoch": 0.33635073332281973, + "grad_norm": 1.3342201709747314, + "learning_rate": 1.546988667564125e-05, + "loss": 0.527, + "step": 10664 + }, + { + "epoch": 0.33638227408926036, + "grad_norm": 1.7148034572601318, + "learning_rate": 1.5469031461295958e-05, + "loss": 0.5337, + "step": 10665 + }, + { + "epoch": 0.336413814855701, + "grad_norm": 1.6697840690612793, + "learning_rate": 1.5468176189877075e-05, + "loss": 0.519, + "step": 10666 + }, + { + "epoch": 0.3364453556221416, + "grad_norm": 1.6107176542282104, + "learning_rate": 1.5467320861393532e-05, + "loss": 0.5038, + "step": 10667 + }, + { + "epoch": 0.33647689638858225, + "grad_norm": 1.7418242692947388, + "learning_rate": 1.5466465475854246e-05, + "loss": 0.5285, + "step": 10668 + }, + { + "epoch": 0.3365084371550229, + "grad_norm": 1.566055178642273, + "learning_rate": 1.546561003326815e-05, + "loss": 0.5598, + "step": 10669 + }, + { + "epoch": 0.3365399779214635, + "grad_norm": 1.3039659261703491, + "learning_rate": 1.546475453364417e-05, + "loss": 0.5503, + "step": 10670 + }, + { + "epoch": 0.33657151868790414, + "grad_norm": 1.837551474571228, + "learning_rate": 1.546389897699123e-05, + "loss": 0.5162, + "step": 10671 + }, + { + "epoch": 0.33660305945434477, + "grad_norm": 2.7158403396606445, + "learning_rate": 1.5463043363318267e-05, + "loss": 0.9161, + "step": 10672 + }, + { + "epoch": 0.33663460022078534, + "grad_norm": 1.6406190395355225, + "learning_rate": 1.5462187692634203e-05, + "loss": 0.5371, + "step": 10673 + }, + { + "epoch": 0.33666614098722597, + "grad_norm": 2.0051138401031494, + "learning_rate": 1.546133196494797e-05, + "loss": 0.536, + "step": 10674 + }, + { + "epoch": 0.3366976817536666, + "grad_norm": 1.653372049331665, + "learning_rate": 1.5460476180268493e-05, + "loss": 0.5294, + "step": 10675 + }, + { + "epoch": 0.33672922252010723, + "grad_norm": 1.6688305139541626, + "learning_rate": 1.545962033860471e-05, + "loss": 0.546, + "step": 10676 + }, + { + "epoch": 0.33676076328654786, + "grad_norm": 1.7641681432724, + "learning_rate": 1.545876443996555e-05, + "loss": 0.4985, + "step": 10677 + }, + { + "epoch": 0.3367923040529885, + "grad_norm": 1.7988531589508057, + "learning_rate": 1.545790848435994e-05, + "loss": 0.5116, + "step": 10678 + }, + { + "epoch": 0.3368238448194291, + "grad_norm": 1.5592864751815796, + "learning_rate": 1.5457052471796823e-05, + "loss": 0.5189, + "step": 10679 + }, + { + "epoch": 0.33685538558586975, + "grad_norm": 1.8566749095916748, + "learning_rate": 1.545619640228512e-05, + "loss": 0.5105, + "step": 10680 + }, + { + "epoch": 0.3368869263523104, + "grad_norm": 1.6705938577651978, + "learning_rate": 1.5455340275833776e-05, + "loss": 0.8515, + "step": 10681 + }, + { + "epoch": 0.336918467118751, + "grad_norm": 1.5782254934310913, + "learning_rate": 1.545448409245172e-05, + "loss": 0.5291, + "step": 10682 + }, + { + "epoch": 0.3369500078851916, + "grad_norm": 2.1534204483032227, + "learning_rate": 1.5453627852147886e-05, + "loss": 0.5533, + "step": 10683 + }, + { + "epoch": 0.3369815486516322, + "grad_norm": 2.051147699356079, + "learning_rate": 1.545277155493121e-05, + "loss": 0.5349, + "step": 10684 + }, + { + "epoch": 0.33701308941807284, + "grad_norm": 3.425163507461548, + "learning_rate": 1.545191520081063e-05, + "loss": 0.423, + "step": 10685 + }, + { + "epoch": 0.33704463018451347, + "grad_norm": 2.196729898452759, + "learning_rate": 1.5451058789795082e-05, + "loss": 0.8625, + "step": 10686 + }, + { + "epoch": 0.3370761709509541, + "grad_norm": 1.4073694944381714, + "learning_rate": 1.54502023218935e-05, + "loss": 0.8645, + "step": 10687 + }, + { + "epoch": 0.33710771171739473, + "grad_norm": 1.3327149152755737, + "learning_rate": 1.5449345797114827e-05, + "loss": 0.5407, + "step": 10688 + }, + { + "epoch": 0.33713925248383536, + "grad_norm": 1.3571646213531494, + "learning_rate": 1.5448489215468e-05, + "loss": 0.5258, + "step": 10689 + }, + { + "epoch": 0.337170793250276, + "grad_norm": 1.5269618034362793, + "learning_rate": 1.544763257696196e-05, + "loss": 0.5013, + "step": 10690 + }, + { + "epoch": 0.3372023340167166, + "grad_norm": 1.4235386848449707, + "learning_rate": 1.5446775881605635e-05, + "loss": 0.5355, + "step": 10691 + }, + { + "epoch": 0.33723387478315725, + "grad_norm": 1.4632022380828857, + "learning_rate": 1.5445919129407982e-05, + "loss": 0.4808, + "step": 10692 + }, + { + "epoch": 0.3372654155495979, + "grad_norm": 2.061629056930542, + "learning_rate": 1.5445062320377932e-05, + "loss": 0.423, + "step": 10693 + }, + { + "epoch": 0.33729695631603845, + "grad_norm": 1.8768534660339355, + "learning_rate": 1.544420545452443e-05, + "loss": 0.5058, + "step": 10694 + }, + { + "epoch": 0.3373284970824791, + "grad_norm": 1.6015220880508423, + "learning_rate": 1.5443348531856416e-05, + "loss": 0.5334, + "step": 10695 + }, + { + "epoch": 0.3373600378489197, + "grad_norm": 35.76470184326172, + "learning_rate": 1.544249155238283e-05, + "loss": 0.5334, + "step": 10696 + }, + { + "epoch": 0.33739157861536034, + "grad_norm": 1.689634084701538, + "learning_rate": 1.5441634516112626e-05, + "loss": 0.5017, + "step": 10697 + }, + { + "epoch": 0.337423119381801, + "grad_norm": 2.2500646114349365, + "learning_rate": 1.5440777423054732e-05, + "loss": 0.8458, + "step": 10698 + }, + { + "epoch": 0.3374546601482416, + "grad_norm": 1.3852956295013428, + "learning_rate": 1.5439920273218106e-05, + "loss": 0.5046, + "step": 10699 + }, + { + "epoch": 0.33748620091468223, + "grad_norm": 1.9626657962799072, + "learning_rate": 1.543906306661169e-05, + "loss": 0.5093, + "step": 10700 + }, + { + "epoch": 0.33751774168112286, + "grad_norm": 1.9736114740371704, + "learning_rate": 1.543820580324442e-05, + "loss": 0.5549, + "step": 10701 + }, + { + "epoch": 0.3375492824475635, + "grad_norm": 1.7759524583816528, + "learning_rate": 1.5437348483125255e-05, + "loss": 0.5279, + "step": 10702 + }, + { + "epoch": 0.3375808232140041, + "grad_norm": 1.6738159656524658, + "learning_rate": 1.543649110626314e-05, + "loss": 0.5536, + "step": 10703 + }, + { + "epoch": 0.33761236398044475, + "grad_norm": 1.3938636779785156, + "learning_rate": 1.5435633672667016e-05, + "loss": 0.3875, + "step": 10704 + }, + { + "epoch": 0.3376439047468853, + "grad_norm": 1.8413909673690796, + "learning_rate": 1.543477618234583e-05, + "loss": 0.5433, + "step": 10705 + }, + { + "epoch": 0.33767544551332596, + "grad_norm": 2.7034454345703125, + "learning_rate": 1.5433918635308538e-05, + "loss": 0.5525, + "step": 10706 + }, + { + "epoch": 0.3377069862797666, + "grad_norm": 1.6281750202178955, + "learning_rate": 1.5433061031564086e-05, + "loss": 0.5357, + "step": 10707 + }, + { + "epoch": 0.3377385270462072, + "grad_norm": 2.2897839546203613, + "learning_rate": 1.543220337112142e-05, + "loss": 0.4381, + "step": 10708 + }, + { + "epoch": 0.33777006781264785, + "grad_norm": 1.3962477445602417, + "learning_rate": 1.54313456539895e-05, + "loss": 0.5373, + "step": 10709 + }, + { + "epoch": 0.3378016085790885, + "grad_norm": 1.8764843940734863, + "learning_rate": 1.5430487880177264e-05, + "loss": 0.4239, + "step": 10710 + }, + { + "epoch": 0.3378331493455291, + "grad_norm": 1.907392144203186, + "learning_rate": 1.5429630049693676e-05, + "loss": 0.4376, + "step": 10711 + }, + { + "epoch": 0.33786469011196973, + "grad_norm": 1.8333563804626465, + "learning_rate": 1.542877216254768e-05, + "loss": 0.8983, + "step": 10712 + }, + { + "epoch": 0.33789623087841036, + "grad_norm": 2.616304874420166, + "learning_rate": 1.542791421874823e-05, + "loss": 0.5309, + "step": 10713 + }, + { + "epoch": 0.337927771644851, + "grad_norm": 1.6508262157440186, + "learning_rate": 1.5427056218304283e-05, + "loss": 0.8796, + "step": 10714 + }, + { + "epoch": 0.33795931241129157, + "grad_norm": 2.2179296016693115, + "learning_rate": 1.542619816122479e-05, + "loss": 0.5551, + "step": 10715 + }, + { + "epoch": 0.3379908531777322, + "grad_norm": 2.1058449745178223, + "learning_rate": 1.5425340047518703e-05, + "loss": 0.5089, + "step": 10716 + }, + { + "epoch": 0.3380223939441728, + "grad_norm": 2.0037357807159424, + "learning_rate": 1.542448187719498e-05, + "loss": 0.5163, + "step": 10717 + }, + { + "epoch": 0.33805393471061346, + "grad_norm": 1.8875551223754883, + "learning_rate": 1.542362365026258e-05, + "loss": 0.8242, + "step": 10718 + }, + { + "epoch": 0.3380854754770541, + "grad_norm": 1.8241935968399048, + "learning_rate": 1.5422765366730455e-05, + "loss": 0.5261, + "step": 10719 + }, + { + "epoch": 0.3381170162434947, + "grad_norm": 1.526937484741211, + "learning_rate": 1.542190702660756e-05, + "loss": 0.8665, + "step": 10720 + }, + { + "epoch": 0.33814855700993535, + "grad_norm": 2.042249917984009, + "learning_rate": 1.5421048629902856e-05, + "loss": 0.5041, + "step": 10721 + }, + { + "epoch": 0.338180097776376, + "grad_norm": 2.074537754058838, + "learning_rate": 1.5420190176625304e-05, + "loss": 0.5327, + "step": 10722 + }, + { + "epoch": 0.3382116385428166, + "grad_norm": 1.485841155052185, + "learning_rate": 1.5419331666783856e-05, + "loss": 0.8214, + "step": 10723 + }, + { + "epoch": 0.33824317930925724, + "grad_norm": 2.3492584228515625, + "learning_rate": 1.5418473100387473e-05, + "loss": 0.5369, + "step": 10724 + }, + { + "epoch": 0.33827472007569787, + "grad_norm": 2.066711664199829, + "learning_rate": 1.541761447744512e-05, + "loss": 0.5354, + "step": 10725 + }, + { + "epoch": 0.33830626084213844, + "grad_norm": 1.6580363512039185, + "learning_rate": 1.541675579796575e-05, + "loss": 0.561, + "step": 10726 + }, + { + "epoch": 0.33833780160857907, + "grad_norm": 1.7452242374420166, + "learning_rate": 1.5415897061958327e-05, + "loss": 0.424, + "step": 10727 + }, + { + "epoch": 0.3383693423750197, + "grad_norm": 1.8532941341400146, + "learning_rate": 1.5415038269431814e-05, + "loss": 0.5359, + "step": 10728 + }, + { + "epoch": 0.33840088314146033, + "grad_norm": 1.851524829864502, + "learning_rate": 1.5414179420395172e-05, + "loss": 0.5073, + "step": 10729 + }, + { + "epoch": 0.33843242390790096, + "grad_norm": 1.5985783338546753, + "learning_rate": 1.5413320514857364e-05, + "loss": 0.5226, + "step": 10730 + }, + { + "epoch": 0.3384639646743416, + "grad_norm": 1.792625069618225, + "learning_rate": 1.541246155282735e-05, + "loss": 0.5529, + "step": 10731 + }, + { + "epoch": 0.3384955054407822, + "grad_norm": 1.518830418586731, + "learning_rate": 1.54116025343141e-05, + "loss": 0.5403, + "step": 10732 + }, + { + "epoch": 0.33852704620722285, + "grad_norm": 1.935701847076416, + "learning_rate": 1.5410743459326573e-05, + "loss": 0.5278, + "step": 10733 + }, + { + "epoch": 0.3385585869736635, + "grad_norm": 1.6866943836212158, + "learning_rate": 1.540988432787374e-05, + "loss": 0.5205, + "step": 10734 + }, + { + "epoch": 0.3385901277401041, + "grad_norm": 1.8348857164382935, + "learning_rate": 1.540902513996456e-05, + "loss": 0.4879, + "step": 10735 + }, + { + "epoch": 0.3386216685065447, + "grad_norm": 1.968409538269043, + "learning_rate": 1.5408165895608007e-05, + "loss": 0.5319, + "step": 10736 + }, + { + "epoch": 0.3386532092729853, + "grad_norm": 1.5546363592147827, + "learning_rate": 1.540730659481304e-05, + "loss": 0.5078, + "step": 10737 + }, + { + "epoch": 0.33868475003942594, + "grad_norm": 1.9932862520217896, + "learning_rate": 1.5406447237588633e-05, + "loss": 0.4462, + "step": 10738 + }, + { + "epoch": 0.33871629080586657, + "grad_norm": 1.779764175415039, + "learning_rate": 1.540558782394375e-05, + "loss": 0.5157, + "step": 10739 + }, + { + "epoch": 0.3387478315723072, + "grad_norm": 1.9111005067825317, + "learning_rate": 1.540472835388736e-05, + "loss": 0.3998, + "step": 10740 + }, + { + "epoch": 0.33877937233874783, + "grad_norm": 1.828980565071106, + "learning_rate": 1.5403868827428433e-05, + "loss": 0.5378, + "step": 10741 + }, + { + "epoch": 0.33881091310518846, + "grad_norm": 1.7426925897598267, + "learning_rate": 1.5403009244575938e-05, + "loss": 0.5005, + "step": 10742 + }, + { + "epoch": 0.3388424538716291, + "grad_norm": 2.289336681365967, + "learning_rate": 1.540214960533885e-05, + "loss": 0.5291, + "step": 10743 + }, + { + "epoch": 0.3388739946380697, + "grad_norm": 1.9530370235443115, + "learning_rate": 1.5401289909726133e-05, + "loss": 0.5349, + "step": 10744 + }, + { + "epoch": 0.33890553540451035, + "grad_norm": 2.3826682567596436, + "learning_rate": 1.540043015774676e-05, + "loss": 0.4998, + "step": 10745 + }, + { + "epoch": 0.338937076170951, + "grad_norm": 2.1695444583892822, + "learning_rate": 1.5399570349409705e-05, + "loss": 0.5332, + "step": 10746 + }, + { + "epoch": 0.33896861693739155, + "grad_norm": 2.2306244373321533, + "learning_rate": 1.5398710484723945e-05, + "loss": 0.5379, + "step": 10747 + }, + { + "epoch": 0.3390001577038322, + "grad_norm": 4.575160026550293, + "learning_rate": 1.5397850563698448e-05, + "loss": 0.4242, + "step": 10748 + }, + { + "epoch": 0.3390316984702728, + "grad_norm": 1.7739306688308716, + "learning_rate": 1.5396990586342187e-05, + "loss": 0.4789, + "step": 10749 + }, + { + "epoch": 0.33906323923671344, + "grad_norm": 1.6344727277755737, + "learning_rate": 1.5396130552664138e-05, + "loss": 0.4847, + "step": 10750 + }, + { + "epoch": 0.3390947800031541, + "grad_norm": 2.1155576705932617, + "learning_rate": 1.5395270462673278e-05, + "loss": 0.526, + "step": 10751 + }, + { + "epoch": 0.3391263207695947, + "grad_norm": 1.7846300601959229, + "learning_rate": 1.5394410316378583e-05, + "loss": 0.5336, + "step": 10752 + }, + { + "epoch": 0.33915786153603533, + "grad_norm": 1.485452651977539, + "learning_rate": 1.5393550113789025e-05, + "loss": 0.5144, + "step": 10753 + }, + { + "epoch": 0.33918940230247596, + "grad_norm": 1.6435319185256958, + "learning_rate": 1.5392689854913585e-05, + "loss": 0.5358, + "step": 10754 + }, + { + "epoch": 0.3392209430689166, + "grad_norm": 2.195749282836914, + "learning_rate": 1.539182953976124e-05, + "loss": 0.5133, + "step": 10755 + }, + { + "epoch": 0.3392524838353572, + "grad_norm": 1.7672008275985718, + "learning_rate": 1.5390969168340962e-05, + "loss": 0.5468, + "step": 10756 + }, + { + "epoch": 0.33928402460179785, + "grad_norm": 1.343839406967163, + "learning_rate": 1.539010874066174e-05, + "loss": 0.5342, + "step": 10757 + }, + { + "epoch": 0.3393155653682384, + "grad_norm": 1.8056567907333374, + "learning_rate": 1.538924825673255e-05, + "loss": 0.4794, + "step": 10758 + }, + { + "epoch": 0.33934710613467906, + "grad_norm": 1.9520193338394165, + "learning_rate": 1.5388387716562366e-05, + "loss": 0.5454, + "step": 10759 + }, + { + "epoch": 0.3393786469011197, + "grad_norm": 1.5023796558380127, + "learning_rate": 1.538752712016017e-05, + "loss": 0.5558, + "step": 10760 + }, + { + "epoch": 0.3394101876675603, + "grad_norm": 2.0224130153656006, + "learning_rate": 1.538666646753495e-05, + "loss": 0.4757, + "step": 10761 + }, + { + "epoch": 0.33944172843400094, + "grad_norm": 1.5401726961135864, + "learning_rate": 1.538580575869568e-05, + "loss": 0.5083, + "step": 10762 + }, + { + "epoch": 0.3394732692004416, + "grad_norm": 1.688484787940979, + "learning_rate": 1.5384944993651345e-05, + "loss": 0.5526, + "step": 10763 + }, + { + "epoch": 0.3395048099668822, + "grad_norm": 2.0378880500793457, + "learning_rate": 1.5384084172410927e-05, + "loss": 0.5221, + "step": 10764 + }, + { + "epoch": 0.33953635073332283, + "grad_norm": 1.7794302701950073, + "learning_rate": 1.538322329498341e-05, + "loss": 0.5468, + "step": 10765 + }, + { + "epoch": 0.33956789149976346, + "grad_norm": 1.7027781009674072, + "learning_rate": 1.538236236137778e-05, + "loss": 0.4776, + "step": 10766 + }, + { + "epoch": 0.3395994322662041, + "grad_norm": 1.6955385208129883, + "learning_rate": 1.538150137160302e-05, + "loss": 0.5366, + "step": 10767 + }, + { + "epoch": 0.33963097303264467, + "grad_norm": 1.6899371147155762, + "learning_rate": 1.5380640325668114e-05, + "loss": 0.4976, + "step": 10768 + }, + { + "epoch": 0.3396625137990853, + "grad_norm": 6.587172985076904, + "learning_rate": 1.5379779223582048e-05, + "loss": 0.5609, + "step": 10769 + }, + { + "epoch": 0.3396940545655259, + "grad_norm": 2.10701060295105, + "learning_rate": 1.5378918065353808e-05, + "loss": 0.5202, + "step": 10770 + }, + { + "epoch": 0.33972559533196656, + "grad_norm": 1.9027597904205322, + "learning_rate": 1.537805685099238e-05, + "loss": 0.4816, + "step": 10771 + }, + { + "epoch": 0.3397571360984072, + "grad_norm": 1.6574907302856445, + "learning_rate": 1.5377195580506755e-05, + "loss": 0.5323, + "step": 10772 + }, + { + "epoch": 0.3397886768648478, + "grad_norm": 1.7733376026153564, + "learning_rate": 1.537633425390592e-05, + "loss": 0.5276, + "step": 10773 + }, + { + "epoch": 0.33982021763128845, + "grad_norm": 1.3542156219482422, + "learning_rate": 1.5375472871198857e-05, + "loss": 0.3926, + "step": 10774 + }, + { + "epoch": 0.3398517583977291, + "grad_norm": 1.3671951293945312, + "learning_rate": 1.5374611432394565e-05, + "loss": 0.531, + "step": 10775 + }, + { + "epoch": 0.3398832991641697, + "grad_norm": 1.5637253522872925, + "learning_rate": 1.5373749937502028e-05, + "loss": 0.5418, + "step": 10776 + }, + { + "epoch": 0.33991483993061034, + "grad_norm": 2.0950844287872314, + "learning_rate": 1.537288838653024e-05, + "loss": 0.8578, + "step": 10777 + }, + { + "epoch": 0.33994638069705096, + "grad_norm": 1.8038661479949951, + "learning_rate": 1.5372026779488188e-05, + "loss": 0.5482, + "step": 10778 + }, + { + "epoch": 0.33997792146349154, + "grad_norm": 2.4198901653289795, + "learning_rate": 1.5371165116384865e-05, + "loss": 0.5364, + "step": 10779 + }, + { + "epoch": 0.34000946222993217, + "grad_norm": 1.781990885734558, + "learning_rate": 1.5370303397229266e-05, + "loss": 0.517, + "step": 10780 + }, + { + "epoch": 0.3400410029963728, + "grad_norm": 1.7045210599899292, + "learning_rate": 1.5369441622030376e-05, + "loss": 0.5355, + "step": 10781 + }, + { + "epoch": 0.34007254376281343, + "grad_norm": 1.3770290613174438, + "learning_rate": 1.5368579790797196e-05, + "loss": 0.4959, + "step": 10782 + }, + { + "epoch": 0.34010408452925406, + "grad_norm": 1.9199295043945312, + "learning_rate": 1.5367717903538717e-05, + "loss": 0.5311, + "step": 10783 + }, + { + "epoch": 0.3401356252956947, + "grad_norm": 1.5754770040512085, + "learning_rate": 1.5366855960263936e-05, + "loss": 0.5293, + "step": 10784 + }, + { + "epoch": 0.3401671660621353, + "grad_norm": 1.703206181526184, + "learning_rate": 1.5365993960981846e-05, + "loss": 0.8556, + "step": 10785 + }, + { + "epoch": 0.34019870682857595, + "grad_norm": 2.0869855880737305, + "learning_rate": 1.536513190570144e-05, + "loss": 0.5338, + "step": 10786 + }, + { + "epoch": 0.3402302475950166, + "grad_norm": 1.7628732919692993, + "learning_rate": 1.5364269794431717e-05, + "loss": 0.5503, + "step": 10787 + }, + { + "epoch": 0.3402617883614572, + "grad_norm": 1.7631661891937256, + "learning_rate": 1.536340762718167e-05, + "loss": 0.4903, + "step": 10788 + }, + { + "epoch": 0.3402933291278978, + "grad_norm": 1.837078332901001, + "learning_rate": 1.5362545403960306e-05, + "loss": 0.41, + "step": 10789 + }, + { + "epoch": 0.3403248698943384, + "grad_norm": 2.1897592544555664, + "learning_rate": 1.5361683124776616e-05, + "loss": 0.4919, + "step": 10790 + }, + { + "epoch": 0.34035641066077904, + "grad_norm": 1.5417258739471436, + "learning_rate": 1.5360820789639596e-05, + "loss": 0.5109, + "step": 10791 + }, + { + "epoch": 0.34038795142721967, + "grad_norm": 2.023865222930908, + "learning_rate": 1.5359958398558252e-05, + "loss": 0.5131, + "step": 10792 + }, + { + "epoch": 0.3404194921936603, + "grad_norm": 2.382563591003418, + "learning_rate": 1.5359095951541576e-05, + "loss": 0.5224, + "step": 10793 + }, + { + "epoch": 0.34045103296010093, + "grad_norm": 1.6858421564102173, + "learning_rate": 1.5358233448598574e-05, + "loss": 0.5016, + "step": 10794 + }, + { + "epoch": 0.34048257372654156, + "grad_norm": 2.0017757415771484, + "learning_rate": 1.5357370889738245e-05, + "loss": 0.535, + "step": 10795 + }, + { + "epoch": 0.3405141144929822, + "grad_norm": 1.5124651193618774, + "learning_rate": 1.5356508274969595e-05, + "loss": 0.8361, + "step": 10796 + }, + { + "epoch": 0.3405456552594228, + "grad_norm": 2.1894845962524414, + "learning_rate": 1.5355645604301618e-05, + "loss": 0.4651, + "step": 10797 + }, + { + "epoch": 0.34057719602586345, + "grad_norm": 2.1701693534851074, + "learning_rate": 1.5354782877743324e-05, + "loss": 0.5475, + "step": 10798 + }, + { + "epoch": 0.3406087367923041, + "grad_norm": 2.389295816421509, + "learning_rate": 1.5353920095303712e-05, + "loss": 0.5223, + "step": 10799 + }, + { + "epoch": 0.34064027755874465, + "grad_norm": 1.6532396078109741, + "learning_rate": 1.5353057256991784e-05, + "loss": 0.3947, + "step": 10800 + }, + { + "epoch": 0.3406718183251853, + "grad_norm": 2.107966423034668, + "learning_rate": 1.5352194362816546e-05, + "loss": 0.5308, + "step": 10801 + }, + { + "epoch": 0.3407033590916259, + "grad_norm": 2.135749340057373, + "learning_rate": 1.5351331412787004e-05, + "loss": 0.5123, + "step": 10802 + }, + { + "epoch": 0.34073489985806654, + "grad_norm": 1.5825408697128296, + "learning_rate": 1.5350468406912165e-05, + "loss": 0.5395, + "step": 10803 + }, + { + "epoch": 0.34076644062450717, + "grad_norm": 1.4734431505203247, + "learning_rate": 1.5349605345201034e-05, + "loss": 0.5222, + "step": 10804 + }, + { + "epoch": 0.3407979813909478, + "grad_norm": 1.6302049160003662, + "learning_rate": 1.5348742227662618e-05, + "loss": 0.5558, + "step": 10805 + }, + { + "epoch": 0.34082952215738843, + "grad_norm": 1.4663329124450684, + "learning_rate": 1.5347879054305927e-05, + "loss": 0.4994, + "step": 10806 + }, + { + "epoch": 0.34086106292382906, + "grad_norm": 2.0736308097839355, + "learning_rate": 1.534701582513996e-05, + "loss": 0.5141, + "step": 10807 + }, + { + "epoch": 0.3408926036902697, + "grad_norm": 1.9619494676589966, + "learning_rate": 1.534615254017373e-05, + "loss": 0.5335, + "step": 10808 + }, + { + "epoch": 0.3409241444567103, + "grad_norm": 2.105738878250122, + "learning_rate": 1.5345289199416252e-05, + "loss": 0.4953, + "step": 10809 + }, + { + "epoch": 0.3409556852231509, + "grad_norm": 1.8239569664001465, + "learning_rate": 1.534442580287653e-05, + "loss": 0.5342, + "step": 10810 + }, + { + "epoch": 0.3409872259895915, + "grad_norm": 1.7361512184143066, + "learning_rate": 1.534356235056357e-05, + "loss": 0.4911, + "step": 10811 + }, + { + "epoch": 0.34101876675603215, + "grad_norm": 2.1507408618927, + "learning_rate": 1.5342698842486392e-05, + "loss": 0.515, + "step": 10812 + }, + { + "epoch": 0.3410503075224728, + "grad_norm": 2.280733346939087, + "learning_rate": 1.5341835278654004e-05, + "loss": 0.5131, + "step": 10813 + }, + { + "epoch": 0.3410818482889134, + "grad_norm": 1.7495830059051514, + "learning_rate": 1.5340971659075413e-05, + "loss": 0.5239, + "step": 10814 + }, + { + "epoch": 0.34111338905535404, + "grad_norm": 1.4515652656555176, + "learning_rate": 1.534010798375964e-05, + "loss": 0.5187, + "step": 10815 + }, + { + "epoch": 0.3411449298217947, + "grad_norm": 1.4653797149658203, + "learning_rate": 1.533924425271569e-05, + "loss": 0.5193, + "step": 10816 + }, + { + "epoch": 0.3411764705882353, + "grad_norm": 1.468480110168457, + "learning_rate": 1.533838046595258e-05, + "loss": 0.4156, + "step": 10817 + }, + { + "epoch": 0.34120801135467593, + "grad_norm": 2.908642053604126, + "learning_rate": 1.533751662347933e-05, + "loss": 0.5192, + "step": 10818 + }, + { + "epoch": 0.34123955212111656, + "grad_norm": 2.350618362426758, + "learning_rate": 1.5336652725304945e-05, + "loss": 0.5398, + "step": 10819 + }, + { + "epoch": 0.3412710928875572, + "grad_norm": 1.3931243419647217, + "learning_rate": 1.5335788771438445e-05, + "loss": 0.5078, + "step": 10820 + }, + { + "epoch": 0.34130263365399777, + "grad_norm": 2.336479902267456, + "learning_rate": 1.5334924761888847e-05, + "loss": 0.8167, + "step": 10821 + }, + { + "epoch": 0.3413341744204384, + "grad_norm": 1.9132159948349, + "learning_rate": 1.5334060696665168e-05, + "loss": 0.5146, + "step": 10822 + }, + { + "epoch": 0.341365715186879, + "grad_norm": 1.727124810218811, + "learning_rate": 1.5333196575776423e-05, + "loss": 0.8425, + "step": 10823 + }, + { + "epoch": 0.34139725595331966, + "grad_norm": 2.3725838661193848, + "learning_rate": 1.533233239923163e-05, + "loss": 0.5327, + "step": 10824 + }, + { + "epoch": 0.3414287967197603, + "grad_norm": 1.9897736310958862, + "learning_rate": 1.5331468167039806e-05, + "loss": 0.512, + "step": 10825 + }, + { + "epoch": 0.3414603374862009, + "grad_norm": 2.2791876792907715, + "learning_rate": 1.5330603879209974e-05, + "loss": 0.8234, + "step": 10826 + }, + { + "epoch": 0.34149187825264155, + "grad_norm": 1.6386878490447998, + "learning_rate": 1.5329739535751155e-05, + "loss": 0.5054, + "step": 10827 + }, + { + "epoch": 0.3415234190190822, + "grad_norm": 1.5564789772033691, + "learning_rate": 1.532887513667236e-05, + "loss": 0.5318, + "step": 10828 + }, + { + "epoch": 0.3415549597855228, + "grad_norm": 1.6253612041473389, + "learning_rate": 1.5328010681982618e-05, + "loss": 0.515, + "step": 10829 + }, + { + "epoch": 0.34158650055196343, + "grad_norm": 1.8396003246307373, + "learning_rate": 1.5327146171690946e-05, + "loss": 0.5578, + "step": 10830 + }, + { + "epoch": 0.34161804131840406, + "grad_norm": 2.198861837387085, + "learning_rate": 1.5326281605806367e-05, + "loss": 0.5246, + "step": 10831 + }, + { + "epoch": 0.34164958208484464, + "grad_norm": 2.015825033187866, + "learning_rate": 1.5325416984337907e-05, + "loss": 0.5186, + "step": 10832 + }, + { + "epoch": 0.34168112285128527, + "grad_norm": 1.4124990701675415, + "learning_rate": 1.5324552307294586e-05, + "loss": 0.511, + "step": 10833 + }, + { + "epoch": 0.3417126636177259, + "grad_norm": 3.223118305206299, + "learning_rate": 1.5323687574685422e-05, + "loss": 0.5267, + "step": 10834 + }, + { + "epoch": 0.34174420438416653, + "grad_norm": 1.6342114210128784, + "learning_rate": 1.532282278651945e-05, + "loss": 0.5158, + "step": 10835 + }, + { + "epoch": 0.34177574515060716, + "grad_norm": 1.4038851261138916, + "learning_rate": 1.5321957942805684e-05, + "loss": 0.5017, + "step": 10836 + }, + { + "epoch": 0.3418072859170478, + "grad_norm": 2.257185459136963, + "learning_rate": 1.5321093043553156e-05, + "loss": 0.5285, + "step": 10837 + }, + { + "epoch": 0.3418388266834884, + "grad_norm": 1.6144100427627563, + "learning_rate": 1.532022808877089e-05, + "loss": 0.5284, + "step": 10838 + }, + { + "epoch": 0.34187036744992905, + "grad_norm": 1.9911694526672363, + "learning_rate": 1.5319363078467912e-05, + "loss": 0.5247, + "step": 10839 + }, + { + "epoch": 0.3419019082163697, + "grad_norm": 1.4295400381088257, + "learning_rate": 1.5318498012653256e-05, + "loss": 0.5107, + "step": 10840 + }, + { + "epoch": 0.3419334489828103, + "grad_norm": 4.203090667724609, + "learning_rate": 1.531763289133594e-05, + "loss": 0.5212, + "step": 10841 + }, + { + "epoch": 0.3419649897492509, + "grad_norm": 1.4928960800170898, + "learning_rate": 1.5316767714524992e-05, + "loss": 0.5341, + "step": 10842 + }, + { + "epoch": 0.3419965305156915, + "grad_norm": 5.688877582550049, + "learning_rate": 1.5315902482229446e-05, + "loss": 0.5193, + "step": 10843 + }, + { + "epoch": 0.34202807128213214, + "grad_norm": 1.634124994277954, + "learning_rate": 1.5315037194458334e-05, + "loss": 0.3947, + "step": 10844 + }, + { + "epoch": 0.34205961204857277, + "grad_norm": 1.633955478668213, + "learning_rate": 1.531417185122068e-05, + "loss": 0.5031, + "step": 10845 + }, + { + "epoch": 0.3420911528150134, + "grad_norm": 1.710634469985962, + "learning_rate": 1.5313306452525517e-05, + "loss": 0.5223, + "step": 10846 + }, + { + "epoch": 0.34212269358145403, + "grad_norm": 2.1661651134490967, + "learning_rate": 1.5312440998381874e-05, + "loss": 0.5146, + "step": 10847 + }, + { + "epoch": 0.34215423434789466, + "grad_norm": 2.678969144821167, + "learning_rate": 1.5311575488798783e-05, + "loss": 0.5519, + "step": 10848 + }, + { + "epoch": 0.3421857751143353, + "grad_norm": 2.4102258682250977, + "learning_rate": 1.5310709923785276e-05, + "loss": 0.5498, + "step": 10849 + }, + { + "epoch": 0.3422173158807759, + "grad_norm": 1.9538702964782715, + "learning_rate": 1.530984430335039e-05, + "loss": 0.5491, + "step": 10850 + }, + { + "epoch": 0.34224885664721655, + "grad_norm": 3.606581211090088, + "learning_rate": 1.5308978627503158e-05, + "loss": 0.496, + "step": 10851 + }, + { + "epoch": 0.3422803974136572, + "grad_norm": 5.143319606781006, + "learning_rate": 1.530811289625261e-05, + "loss": 0.4135, + "step": 10852 + }, + { + "epoch": 0.34231193818009775, + "grad_norm": 1.934077501296997, + "learning_rate": 1.530724710960778e-05, + "loss": 0.5351, + "step": 10853 + }, + { + "epoch": 0.3423434789465384, + "grad_norm": 2.4121205806732178, + "learning_rate": 1.530638126757771e-05, + "loss": 0.5218, + "step": 10854 + }, + { + "epoch": 0.342375019712979, + "grad_norm": 1.6156543493270874, + "learning_rate": 1.5305515370171425e-05, + "loss": 0.8429, + "step": 10855 + }, + { + "epoch": 0.34240656047941964, + "grad_norm": 2.2506635189056396, + "learning_rate": 1.530464941739797e-05, + "loss": 0.8946, + "step": 10856 + }, + { + "epoch": 0.34243810124586027, + "grad_norm": 2.1523995399475098, + "learning_rate": 1.5303783409266385e-05, + "loss": 0.5584, + "step": 10857 + }, + { + "epoch": 0.3424696420123009, + "grad_norm": 2.4677059650421143, + "learning_rate": 1.53029173457857e-05, + "loss": 0.5259, + "step": 10858 + }, + { + "epoch": 0.34250118277874153, + "grad_norm": 2.431400775909424, + "learning_rate": 1.5302051226964948e-05, + "loss": 0.5137, + "step": 10859 + }, + { + "epoch": 0.34253272354518216, + "grad_norm": 2.7055227756500244, + "learning_rate": 1.5301185052813176e-05, + "loss": 0.4973, + "step": 10860 + }, + { + "epoch": 0.3425642643116228, + "grad_norm": 1.735756516456604, + "learning_rate": 1.5300318823339426e-05, + "loss": 0.5161, + "step": 10861 + }, + { + "epoch": 0.3425958050780634, + "grad_norm": 2.3396317958831787, + "learning_rate": 1.529945253855273e-05, + "loss": 0.5247, + "step": 10862 + }, + { + "epoch": 0.342627345844504, + "grad_norm": 2.389777183532715, + "learning_rate": 1.5298586198462135e-05, + "loss": 0.5129, + "step": 10863 + }, + { + "epoch": 0.3426588866109446, + "grad_norm": 2.0293216705322266, + "learning_rate": 1.5297719803076677e-05, + "loss": 0.8429, + "step": 10864 + }, + { + "epoch": 0.34269042737738525, + "grad_norm": 2.088425636291504, + "learning_rate": 1.5296853352405402e-05, + "loss": 0.546, + "step": 10865 + }, + { + "epoch": 0.3427219681438259, + "grad_norm": 1.7209664583206177, + "learning_rate": 1.5295986846457343e-05, + "loss": 0.4916, + "step": 10866 + }, + { + "epoch": 0.3427535089102665, + "grad_norm": 1.8117265701293945, + "learning_rate": 1.5295120285241557e-05, + "loss": 0.8062, + "step": 10867 + }, + { + "epoch": 0.34278504967670714, + "grad_norm": 2.223870038986206, + "learning_rate": 1.5294253668767073e-05, + "loss": 0.8191, + "step": 10868 + }, + { + "epoch": 0.3428165904431478, + "grad_norm": 1.394720196723938, + "learning_rate": 1.5293386997042943e-05, + "loss": 0.5424, + "step": 10869 + }, + { + "epoch": 0.3428481312095884, + "grad_norm": 1.378980278968811, + "learning_rate": 1.529252027007821e-05, + "loss": 0.8164, + "step": 10870 + }, + { + "epoch": 0.34287967197602903, + "grad_norm": 1.729403018951416, + "learning_rate": 1.529165348788192e-05, + "loss": 0.5291, + "step": 10871 + }, + { + "epoch": 0.34291121274246966, + "grad_norm": 1.722753643989563, + "learning_rate": 1.5290786650463113e-05, + "loss": 0.548, + "step": 10872 + }, + { + "epoch": 0.3429427535089103, + "grad_norm": 1.3378446102142334, + "learning_rate": 1.5289919757830843e-05, + "loss": 0.5503, + "step": 10873 + }, + { + "epoch": 0.34297429427535087, + "grad_norm": 8.790058135986328, + "learning_rate": 1.528905280999415e-05, + "loss": 0.4013, + "step": 10874 + }, + { + "epoch": 0.3430058350417915, + "grad_norm": 2.510432004928589, + "learning_rate": 1.5288185806962085e-05, + "loss": 0.502, + "step": 10875 + }, + { + "epoch": 0.3430373758082321, + "grad_norm": 2.8493852615356445, + "learning_rate": 1.5287318748743696e-05, + "loss": 0.5171, + "step": 10876 + }, + { + "epoch": 0.34306891657467276, + "grad_norm": 2.3387691974639893, + "learning_rate": 1.528645163534803e-05, + "loss": 0.8573, + "step": 10877 + }, + { + "epoch": 0.3431004573411134, + "grad_norm": 1.561781406402588, + "learning_rate": 1.5285584466784137e-05, + "loss": 0.533, + "step": 10878 + }, + { + "epoch": 0.343131998107554, + "grad_norm": 1.4469213485717773, + "learning_rate": 1.5284717243061064e-05, + "loss": 0.5249, + "step": 10879 + }, + { + "epoch": 0.34316353887399464, + "grad_norm": 1.8363960981369019, + "learning_rate": 1.528384996418786e-05, + "loss": 0.505, + "step": 10880 + }, + { + "epoch": 0.3431950796404353, + "grad_norm": 1.5613173246383667, + "learning_rate": 1.5282982630173587e-05, + "loss": 0.5238, + "step": 10881 + }, + { + "epoch": 0.3432266204068759, + "grad_norm": 1.7359306812286377, + "learning_rate": 1.5282115241027282e-05, + "loss": 0.8632, + "step": 10882 + }, + { + "epoch": 0.34325816117331653, + "grad_norm": 1.8842219114303589, + "learning_rate": 1.5281247796758005e-05, + "loss": 0.5282, + "step": 10883 + }, + { + "epoch": 0.34328970193975716, + "grad_norm": 1.4867814779281616, + "learning_rate": 1.5280380297374805e-05, + "loss": 0.5158, + "step": 10884 + }, + { + "epoch": 0.34332124270619774, + "grad_norm": 1.6663845777511597, + "learning_rate": 1.527951274288674e-05, + "loss": 0.5422, + "step": 10885 + }, + { + "epoch": 0.34335278347263837, + "grad_norm": 1.5369185209274292, + "learning_rate": 1.527864513330286e-05, + "loss": 0.5154, + "step": 10886 + }, + { + "epoch": 0.343384324239079, + "grad_norm": 1.7470414638519287, + "learning_rate": 1.5277777468632215e-05, + "loss": 0.577, + "step": 10887 + }, + { + "epoch": 0.3434158650055196, + "grad_norm": 2.039661407470703, + "learning_rate": 1.527690974888387e-05, + "loss": 0.8682, + "step": 10888 + }, + { + "epoch": 0.34344740577196026, + "grad_norm": 1.8257960081100464, + "learning_rate": 1.527604197406687e-05, + "loss": 0.5537, + "step": 10889 + }, + { + "epoch": 0.3434789465384009, + "grad_norm": 1.5687098503112793, + "learning_rate": 1.527517414419027e-05, + "loss": 0.8762, + "step": 10890 + }, + { + "epoch": 0.3435104873048415, + "grad_norm": 1.4788480997085571, + "learning_rate": 1.527430625926314e-05, + "loss": 0.3983, + "step": 10891 + }, + { + "epoch": 0.34354202807128215, + "grad_norm": 1.4319435358047485, + "learning_rate": 1.5273438319294527e-05, + "loss": 0.5084, + "step": 10892 + }, + { + "epoch": 0.3435735688377228, + "grad_norm": 2.0793020725250244, + "learning_rate": 1.527257032429349e-05, + "loss": 0.5037, + "step": 10893 + }, + { + "epoch": 0.3436051096041634, + "grad_norm": 1.7386213541030884, + "learning_rate": 1.5271702274269086e-05, + "loss": 0.5099, + "step": 10894 + }, + { + "epoch": 0.343636650370604, + "grad_norm": 1.5094168186187744, + "learning_rate": 1.527083416923038e-05, + "loss": 0.8409, + "step": 10895 + }, + { + "epoch": 0.3436681911370446, + "grad_norm": 1.7501161098480225, + "learning_rate": 1.5269966009186423e-05, + "loss": 0.5044, + "step": 10896 + }, + { + "epoch": 0.34369973190348524, + "grad_norm": 1.9138767719268799, + "learning_rate": 1.5269097794146278e-05, + "loss": 0.5187, + "step": 10897 + }, + { + "epoch": 0.34373127266992587, + "grad_norm": 1.6970278024673462, + "learning_rate": 1.5268229524119007e-05, + "loss": 0.5241, + "step": 10898 + }, + { + "epoch": 0.3437628134363665, + "grad_norm": 1.9782130718231201, + "learning_rate": 1.5267361199113674e-05, + "loss": 0.5182, + "step": 10899 + }, + { + "epoch": 0.34379435420280713, + "grad_norm": 2.101475715637207, + "learning_rate": 1.526649281913933e-05, + "loss": 0.5262, + "step": 10900 + }, + { + "epoch": 0.34382589496924776, + "grad_norm": 1.7215222120285034, + "learning_rate": 1.5265624384205045e-05, + "loss": 0.3973, + "step": 10901 + }, + { + "epoch": 0.3438574357356884, + "grad_norm": 2.551565170288086, + "learning_rate": 1.5264755894319886e-05, + "loss": 0.5375, + "step": 10902 + }, + { + "epoch": 0.343888976502129, + "grad_norm": 2.3973677158355713, + "learning_rate": 1.5263887349492908e-05, + "loss": 0.4987, + "step": 10903 + }, + { + "epoch": 0.34392051726856965, + "grad_norm": 1.807876706123352, + "learning_rate": 1.526301874973318e-05, + "loss": 0.4937, + "step": 10904 + }, + { + "epoch": 0.3439520580350103, + "grad_norm": 2.2442712783813477, + "learning_rate": 1.5262150095049766e-05, + "loss": 0.5349, + "step": 10905 + }, + { + "epoch": 0.34398359880145085, + "grad_norm": 1.9619721174240112, + "learning_rate": 1.5261281385451728e-05, + "loss": 0.503, + "step": 10906 + }, + { + "epoch": 0.3440151395678915, + "grad_norm": 1.6805171966552734, + "learning_rate": 1.5260412620948134e-05, + "loss": 0.5059, + "step": 10907 + }, + { + "epoch": 0.3440466803343321, + "grad_norm": 1.8267024755477905, + "learning_rate": 1.5259543801548047e-05, + "loss": 0.5084, + "step": 10908 + }, + { + "epoch": 0.34407822110077274, + "grad_norm": 1.6897099018096924, + "learning_rate": 1.5258674927260538e-05, + "loss": 0.5112, + "step": 10909 + }, + { + "epoch": 0.34410976186721337, + "grad_norm": 1.5605711936950684, + "learning_rate": 1.5257805998094672e-05, + "loss": 0.5106, + "step": 10910 + }, + { + "epoch": 0.344141302633654, + "grad_norm": 1.6674970388412476, + "learning_rate": 1.5256937014059518e-05, + "loss": 0.5305, + "step": 10911 + }, + { + "epoch": 0.34417284340009463, + "grad_norm": 1.6028411388397217, + "learning_rate": 1.5256067975164146e-05, + "loss": 0.5343, + "step": 10912 + }, + { + "epoch": 0.34420438416653526, + "grad_norm": 1.56158447265625, + "learning_rate": 1.5255198881417623e-05, + "loss": 0.5245, + "step": 10913 + }, + { + "epoch": 0.3442359249329759, + "grad_norm": 2.019047498703003, + "learning_rate": 1.5254329732829018e-05, + "loss": 0.4995, + "step": 10914 + }, + { + "epoch": 0.3442674656994165, + "grad_norm": 1.9186716079711914, + "learning_rate": 1.5253460529407404e-05, + "loss": 0.5266, + "step": 10915 + }, + { + "epoch": 0.3442990064658571, + "grad_norm": 3.117175579071045, + "learning_rate": 1.5252591271161852e-05, + "loss": 0.568, + "step": 10916 + }, + { + "epoch": 0.3443305472322977, + "grad_norm": 1.9411367177963257, + "learning_rate": 1.5251721958101429e-05, + "loss": 0.5144, + "step": 10917 + }, + { + "epoch": 0.34436208799873835, + "grad_norm": 1.3927565813064575, + "learning_rate": 1.5250852590235208e-05, + "loss": 0.5194, + "step": 10918 + }, + { + "epoch": 0.344393628765179, + "grad_norm": 2.4503133296966553, + "learning_rate": 1.5249983167572265e-05, + "loss": 0.5212, + "step": 10919 + }, + { + "epoch": 0.3444251695316196, + "grad_norm": 1.7002619504928589, + "learning_rate": 1.5249113690121671e-05, + "loss": 0.5381, + "step": 10920 + }, + { + "epoch": 0.34445671029806024, + "grad_norm": 2.3195993900299072, + "learning_rate": 1.5248244157892497e-05, + "loss": 0.5052, + "step": 10921 + }, + { + "epoch": 0.34448825106450087, + "grad_norm": 1.57675039768219, + "learning_rate": 1.5247374570893823e-05, + "loss": 0.5206, + "step": 10922 + }, + { + "epoch": 0.3445197918309415, + "grad_norm": 2.3020644187927246, + "learning_rate": 1.5246504929134722e-05, + "loss": 0.5572, + "step": 10923 + }, + { + "epoch": 0.34455133259738213, + "grad_norm": 1.8518821001052856, + "learning_rate": 1.5245635232624266e-05, + "loss": 0.5335, + "step": 10924 + }, + { + "epoch": 0.34458287336382276, + "grad_norm": 1.961378574371338, + "learning_rate": 1.5244765481371536e-05, + "loss": 0.5315, + "step": 10925 + }, + { + "epoch": 0.3446144141302634, + "grad_norm": 1.964531660079956, + "learning_rate": 1.5243895675385603e-05, + "loss": 0.5137, + "step": 10926 + }, + { + "epoch": 0.34464595489670397, + "grad_norm": 2.095294713973999, + "learning_rate": 1.5243025814675548e-05, + "loss": 0.5083, + "step": 10927 + }, + { + "epoch": 0.3446774956631446, + "grad_norm": 1.7644964456558228, + "learning_rate": 1.5242155899250445e-05, + "loss": 0.8542, + "step": 10928 + }, + { + "epoch": 0.3447090364295852, + "grad_norm": 1.5527732372283936, + "learning_rate": 1.5241285929119376e-05, + "loss": 0.5336, + "step": 10929 + }, + { + "epoch": 0.34474057719602585, + "grad_norm": 1.4502198696136475, + "learning_rate": 1.5240415904291419e-05, + "loss": 0.3462, + "step": 10930 + }, + { + "epoch": 0.3447721179624665, + "grad_norm": 1.9574253559112549, + "learning_rate": 1.5239545824775652e-05, + "loss": 0.5089, + "step": 10931 + }, + { + "epoch": 0.3448036587289071, + "grad_norm": 2.028795003890991, + "learning_rate": 1.5238675690581158e-05, + "loss": 0.5342, + "step": 10932 + }, + { + "epoch": 0.34483519949534774, + "grad_norm": 2.1880321502685547, + "learning_rate": 1.5237805501717015e-05, + "loss": 0.5401, + "step": 10933 + }, + { + "epoch": 0.3448667402617884, + "grad_norm": 2.0349714756011963, + "learning_rate": 1.5236935258192302e-05, + "loss": 0.5367, + "step": 10934 + }, + { + "epoch": 0.344898281028229, + "grad_norm": 1.9712235927581787, + "learning_rate": 1.5236064960016109e-05, + "loss": 0.5216, + "step": 10935 + }, + { + "epoch": 0.34492982179466963, + "grad_norm": 1.6788005828857422, + "learning_rate": 1.5235194607197508e-05, + "loss": 0.5219, + "step": 10936 + }, + { + "epoch": 0.3449613625611102, + "grad_norm": 1.7666772603988647, + "learning_rate": 1.5234324199745587e-05, + "loss": 0.8291, + "step": 10937 + }, + { + "epoch": 0.34499290332755084, + "grad_norm": 1.4976626634597778, + "learning_rate": 1.5233453737669427e-05, + "loss": 0.4834, + "step": 10938 + }, + { + "epoch": 0.34502444409399147, + "grad_norm": 1.6284664869308472, + "learning_rate": 1.5232583220978115e-05, + "loss": 0.8544, + "step": 10939 + }, + { + "epoch": 0.3450559848604321, + "grad_norm": 2.1882259845733643, + "learning_rate": 1.5231712649680736e-05, + "loss": 0.532, + "step": 10940 + }, + { + "epoch": 0.3450875256268727, + "grad_norm": 2.2122342586517334, + "learning_rate": 1.5230842023786369e-05, + "loss": 0.4976, + "step": 10941 + }, + { + "epoch": 0.34511906639331336, + "grad_norm": 1.9167500734329224, + "learning_rate": 1.5229971343304106e-05, + "loss": 0.4997, + "step": 10942 + }, + { + "epoch": 0.345150607159754, + "grad_norm": 1.4991334676742554, + "learning_rate": 1.5229100608243034e-05, + "loss": 0.509, + "step": 10943 + }, + { + "epoch": 0.3451821479261946, + "grad_norm": 1.841333270072937, + "learning_rate": 1.5228229818612233e-05, + "loss": 0.5353, + "step": 10944 + }, + { + "epoch": 0.34521368869263525, + "grad_norm": 1.3218990564346313, + "learning_rate": 1.5227358974420798e-05, + "loss": 0.8627, + "step": 10945 + }, + { + "epoch": 0.3452452294590759, + "grad_norm": 1.954505443572998, + "learning_rate": 1.522648807567781e-05, + "loss": 0.5203, + "step": 10946 + }, + { + "epoch": 0.3452767702255165, + "grad_norm": 3.473853826522827, + "learning_rate": 1.5225617122392361e-05, + "loss": 0.5429, + "step": 10947 + }, + { + "epoch": 0.3453083109919571, + "grad_norm": 1.8875375986099243, + "learning_rate": 1.5224746114573542e-05, + "loss": 0.4889, + "step": 10948 + }, + { + "epoch": 0.3453398517583977, + "grad_norm": 1.7772598266601562, + "learning_rate": 1.5223875052230437e-05, + "loss": 0.3983, + "step": 10949 + }, + { + "epoch": 0.34537139252483834, + "grad_norm": 2.668680429458618, + "learning_rate": 1.522300393537214e-05, + "loss": 0.3822, + "step": 10950 + }, + { + "epoch": 0.34540293329127897, + "grad_norm": 2.1036036014556885, + "learning_rate": 1.5222132764007742e-05, + "loss": 0.5264, + "step": 10951 + }, + { + "epoch": 0.3454344740577196, + "grad_norm": 2.216341495513916, + "learning_rate": 1.5221261538146333e-05, + "loss": 0.5034, + "step": 10952 + }, + { + "epoch": 0.34546601482416023, + "grad_norm": 1.474410057067871, + "learning_rate": 1.5220390257797009e-05, + "loss": 0.8663, + "step": 10953 + }, + { + "epoch": 0.34549755559060086, + "grad_norm": 2.6591527462005615, + "learning_rate": 1.5219518922968856e-05, + "loss": 0.5395, + "step": 10954 + }, + { + "epoch": 0.3455290963570415, + "grad_norm": 2.8411471843719482, + "learning_rate": 1.5218647533670971e-05, + "loss": 0.5811, + "step": 10955 + }, + { + "epoch": 0.3455606371234821, + "grad_norm": 4.177398681640625, + "learning_rate": 1.5217776089912449e-05, + "loss": 0.5379, + "step": 10956 + }, + { + "epoch": 0.34559217788992275, + "grad_norm": 2.2917606830596924, + "learning_rate": 1.521690459170238e-05, + "loss": 0.5136, + "step": 10957 + }, + { + "epoch": 0.3456237186563634, + "grad_norm": 2.1092917919158936, + "learning_rate": 1.521603303904986e-05, + "loss": 0.4931, + "step": 10958 + }, + { + "epoch": 0.34565525942280395, + "grad_norm": 1.835003137588501, + "learning_rate": 1.5215161431963985e-05, + "loss": 0.5115, + "step": 10959 + }, + { + "epoch": 0.3456868001892446, + "grad_norm": 2.0763840675354004, + "learning_rate": 1.5214289770453853e-05, + "loss": 0.5515, + "step": 10960 + }, + { + "epoch": 0.3457183409556852, + "grad_norm": 1.5517992973327637, + "learning_rate": 1.5213418054528557e-05, + "loss": 0.8591, + "step": 10961 + }, + { + "epoch": 0.34574988172212584, + "grad_norm": 2.3507089614868164, + "learning_rate": 1.5212546284197194e-05, + "loss": 0.512, + "step": 10962 + }, + { + "epoch": 0.34578142248856647, + "grad_norm": 1.6009469032287598, + "learning_rate": 1.5211674459468866e-05, + "loss": 0.5281, + "step": 10963 + }, + { + "epoch": 0.3458129632550071, + "grad_norm": 1.4761691093444824, + "learning_rate": 1.5210802580352666e-05, + "loss": 0.5185, + "step": 10964 + }, + { + "epoch": 0.34584450402144773, + "grad_norm": 1.3305493593215942, + "learning_rate": 1.5209930646857698e-05, + "loss": 0.5209, + "step": 10965 + }, + { + "epoch": 0.34587604478788836, + "grad_norm": 1.3059581518173218, + "learning_rate": 1.5209058658993056e-05, + "loss": 0.8176, + "step": 10966 + }, + { + "epoch": 0.345907585554329, + "grad_norm": 2.0336272716522217, + "learning_rate": 1.5208186616767843e-05, + "loss": 0.5117, + "step": 10967 + }, + { + "epoch": 0.3459391263207696, + "grad_norm": 1.4247796535491943, + "learning_rate": 1.5207314520191157e-05, + "loss": 0.4978, + "step": 10968 + }, + { + "epoch": 0.3459706670872102, + "grad_norm": 3.5805282592773438, + "learning_rate": 1.5206442369272101e-05, + "loss": 0.5033, + "step": 10969 + }, + { + "epoch": 0.3460022078536508, + "grad_norm": 1.9473358392715454, + "learning_rate": 1.5205570164019777e-05, + "loss": 0.5354, + "step": 10970 + }, + { + "epoch": 0.34603374862009145, + "grad_norm": 2.6663312911987305, + "learning_rate": 1.5204697904443286e-05, + "loss": 0.5309, + "step": 10971 + }, + { + "epoch": 0.3460652893865321, + "grad_norm": 1.8804044723510742, + "learning_rate": 1.520382559055173e-05, + "loss": 0.5355, + "step": 10972 + }, + { + "epoch": 0.3460968301529727, + "grad_norm": 1.5287388563156128, + "learning_rate": 1.5202953222354216e-05, + "loss": 0.5294, + "step": 10973 + }, + { + "epoch": 0.34612837091941334, + "grad_norm": 1.7836534976959229, + "learning_rate": 1.5202080799859844e-05, + "loss": 0.5087, + "step": 10974 + }, + { + "epoch": 0.34615991168585397, + "grad_norm": 1.7240039110183716, + "learning_rate": 1.520120832307772e-05, + "loss": 0.5193, + "step": 10975 + }, + { + "epoch": 0.3461914524522946, + "grad_norm": 1.670929193496704, + "learning_rate": 1.5200335792016948e-05, + "loss": 0.5384, + "step": 10976 + }, + { + "epoch": 0.34622299321873523, + "grad_norm": 1.9944230318069458, + "learning_rate": 1.5199463206686633e-05, + "loss": 0.4791, + "step": 10977 + }, + { + "epoch": 0.34625453398517586, + "grad_norm": 1.3641080856323242, + "learning_rate": 1.5198590567095883e-05, + "loss": 0.5357, + "step": 10978 + }, + { + "epoch": 0.3462860747516165, + "grad_norm": 2.85294508934021, + "learning_rate": 1.5197717873253803e-05, + "loss": 0.3989, + "step": 10979 + }, + { + "epoch": 0.34631761551805706, + "grad_norm": 1.973551869392395, + "learning_rate": 1.5196845125169502e-05, + "loss": 0.5196, + "step": 10980 + }, + { + "epoch": 0.3463491562844977, + "grad_norm": 2.1372156143188477, + "learning_rate": 1.5195972322852086e-05, + "loss": 0.5333, + "step": 10981 + }, + { + "epoch": 0.3463806970509383, + "grad_norm": 1.3037495613098145, + "learning_rate": 1.5195099466310665e-05, + "loss": 0.4912, + "step": 10982 + }, + { + "epoch": 0.34641223781737895, + "grad_norm": 1.618131399154663, + "learning_rate": 1.519422655555435e-05, + "loss": 0.5024, + "step": 10983 + }, + { + "epoch": 0.3464437785838196, + "grad_norm": 2.090416193008423, + "learning_rate": 1.5193353590592241e-05, + "loss": 0.5121, + "step": 10984 + }, + { + "epoch": 0.3464753193502602, + "grad_norm": 2.624037981033325, + "learning_rate": 1.519248057143346e-05, + "loss": 0.5404, + "step": 10985 + }, + { + "epoch": 0.34650686011670084, + "grad_norm": 3.539874792098999, + "learning_rate": 1.519160749808711e-05, + "loss": 0.5429, + "step": 10986 + }, + { + "epoch": 0.3465384008831415, + "grad_norm": 2.603062629699707, + "learning_rate": 1.5190734370562306e-05, + "loss": 0.5319, + "step": 10987 + }, + { + "epoch": 0.3465699416495821, + "grad_norm": 1.7572256326675415, + "learning_rate": 1.518986118886816e-05, + "loss": 0.5288, + "step": 10988 + }, + { + "epoch": 0.34660148241602273, + "grad_norm": 2.3416898250579834, + "learning_rate": 1.5188987953013778e-05, + "loss": 0.5198, + "step": 10989 + }, + { + "epoch": 0.3466330231824633, + "grad_norm": 1.8457906246185303, + "learning_rate": 1.5188114663008281e-05, + "loss": 0.5385, + "step": 10990 + }, + { + "epoch": 0.34666456394890394, + "grad_norm": 1.611089825630188, + "learning_rate": 1.5187241318860779e-05, + "loss": 0.8201, + "step": 10991 + }, + { + "epoch": 0.34669610471534457, + "grad_norm": 1.6236134767532349, + "learning_rate": 1.5186367920580383e-05, + "loss": 0.5087, + "step": 10992 + }, + { + "epoch": 0.3467276454817852, + "grad_norm": 1.3433667421340942, + "learning_rate": 1.5185494468176214e-05, + "loss": 0.5417, + "step": 10993 + }, + { + "epoch": 0.3467591862482258, + "grad_norm": 1.4336376190185547, + "learning_rate": 1.5184620961657381e-05, + "loss": 0.4702, + "step": 10994 + }, + { + "epoch": 0.34679072701466646, + "grad_norm": 2.799142360687256, + "learning_rate": 1.5183747401033006e-05, + "loss": 0.4902, + "step": 10995 + }, + { + "epoch": 0.3468222677811071, + "grad_norm": 2.053232431411743, + "learning_rate": 1.51828737863122e-05, + "loss": 0.5305, + "step": 10996 + }, + { + "epoch": 0.3468538085475477, + "grad_norm": 1.3146363496780396, + "learning_rate": 1.5182000117504081e-05, + "loss": 0.8263, + "step": 10997 + }, + { + "epoch": 0.34688534931398834, + "grad_norm": 1.4521418809890747, + "learning_rate": 1.5181126394617768e-05, + "loss": 0.5149, + "step": 10998 + }, + { + "epoch": 0.346916890080429, + "grad_norm": 1.6480295658111572, + "learning_rate": 1.518025261766238e-05, + "loss": 0.5164, + "step": 10999 + }, + { + "epoch": 0.3469484308468696, + "grad_norm": 1.868434190750122, + "learning_rate": 1.5179378786647027e-05, + "loss": 0.5387, + "step": 11000 + }, + { + "epoch": 0.3469799716133102, + "grad_norm": 1.7014294862747192, + "learning_rate": 1.5178504901580842e-05, + "loss": 0.5276, + "step": 11001 + }, + { + "epoch": 0.3470115123797508, + "grad_norm": 1.8096320629119873, + "learning_rate": 1.5177630962472934e-05, + "loss": 0.5338, + "step": 11002 + }, + { + "epoch": 0.34704305314619144, + "grad_norm": 1.4002649784088135, + "learning_rate": 1.5176756969332428e-05, + "loss": 0.5177, + "step": 11003 + }, + { + "epoch": 0.34707459391263207, + "grad_norm": 2.013371229171753, + "learning_rate": 1.5175882922168441e-05, + "loss": 0.5159, + "step": 11004 + }, + { + "epoch": 0.3471061346790727, + "grad_norm": 1.546478509902954, + "learning_rate": 1.5175008820990097e-05, + "loss": 0.4251, + "step": 11005 + }, + { + "epoch": 0.3471376754455133, + "grad_norm": 1.549560785293579, + "learning_rate": 1.5174134665806518e-05, + "loss": 0.5346, + "step": 11006 + }, + { + "epoch": 0.34716921621195396, + "grad_norm": 1.845116138458252, + "learning_rate": 1.5173260456626829e-05, + "loss": 0.5255, + "step": 11007 + }, + { + "epoch": 0.3472007569783946, + "grad_norm": 1.4111793041229248, + "learning_rate": 1.5172386193460147e-05, + "loss": 0.5283, + "step": 11008 + }, + { + "epoch": 0.3472322977448352, + "grad_norm": 1.612521767616272, + "learning_rate": 1.5171511876315598e-05, + "loss": 0.5277, + "step": 11009 + }, + { + "epoch": 0.34726383851127585, + "grad_norm": 1.962688684463501, + "learning_rate": 1.5170637505202308e-05, + "loss": 0.5265, + "step": 11010 + }, + { + "epoch": 0.3472953792777165, + "grad_norm": 1.3823554515838623, + "learning_rate": 1.5169763080129402e-05, + "loss": 0.5409, + "step": 11011 + }, + { + "epoch": 0.34732692004415705, + "grad_norm": 1.8896747827529907, + "learning_rate": 1.5168888601106005e-05, + "loss": 0.5497, + "step": 11012 + }, + { + "epoch": 0.3473584608105977, + "grad_norm": 1.6890010833740234, + "learning_rate": 1.5168014068141242e-05, + "loss": 0.5415, + "step": 11013 + }, + { + "epoch": 0.3473900015770383, + "grad_norm": 1.4826825857162476, + "learning_rate": 1.5167139481244238e-05, + "loss": 0.5027, + "step": 11014 + }, + { + "epoch": 0.34742154234347894, + "grad_norm": 1.5499664545059204, + "learning_rate": 1.516626484042412e-05, + "loss": 0.5167, + "step": 11015 + }, + { + "epoch": 0.34745308310991957, + "grad_norm": 1.5734708309173584, + "learning_rate": 1.5165390145690017e-05, + "loss": 0.5579, + "step": 11016 + }, + { + "epoch": 0.3474846238763602, + "grad_norm": 1.6940094232559204, + "learning_rate": 1.5164515397051058e-05, + "loss": 0.5595, + "step": 11017 + }, + { + "epoch": 0.34751616464280083, + "grad_norm": 1.768413782119751, + "learning_rate": 1.5163640594516369e-05, + "loss": 0.5217, + "step": 11018 + }, + { + "epoch": 0.34754770540924146, + "grad_norm": 2.230421304702759, + "learning_rate": 1.5162765738095084e-05, + "loss": 0.5394, + "step": 11019 + }, + { + "epoch": 0.3475792461756821, + "grad_norm": 1.5095330476760864, + "learning_rate": 1.5161890827796328e-05, + "loss": 0.5297, + "step": 11020 + }, + { + "epoch": 0.3476107869421227, + "grad_norm": 1.2361387014389038, + "learning_rate": 1.5161015863629232e-05, + "loss": 0.4967, + "step": 11021 + }, + { + "epoch": 0.3476423277085633, + "grad_norm": 1.5831910371780396, + "learning_rate": 1.5160140845602934e-05, + "loss": 0.5528, + "step": 11022 + }, + { + "epoch": 0.3476738684750039, + "grad_norm": 1.6039233207702637, + "learning_rate": 1.5159265773726555e-05, + "loss": 0.5265, + "step": 11023 + }, + { + "epoch": 0.34770540924144455, + "grad_norm": 2.4359755516052246, + "learning_rate": 1.515839064800923e-05, + "loss": 0.5248, + "step": 11024 + }, + { + "epoch": 0.3477369500078852, + "grad_norm": 1.4812878370285034, + "learning_rate": 1.5157515468460094e-05, + "loss": 0.5266, + "step": 11025 + }, + { + "epoch": 0.3477684907743258, + "grad_norm": 1.4652109146118164, + "learning_rate": 1.515664023508828e-05, + "loss": 0.5463, + "step": 11026 + }, + { + "epoch": 0.34780003154076644, + "grad_norm": 1.3431223630905151, + "learning_rate": 1.5155764947902923e-05, + "loss": 0.3852, + "step": 11027 + }, + { + "epoch": 0.34783157230720707, + "grad_norm": 1.5935266017913818, + "learning_rate": 1.5154889606913152e-05, + "loss": 0.5067, + "step": 11028 + }, + { + "epoch": 0.3478631130736477, + "grad_norm": 1.5219225883483887, + "learning_rate": 1.5154014212128108e-05, + "loss": 0.4974, + "step": 11029 + }, + { + "epoch": 0.34789465384008833, + "grad_norm": 1.6997699737548828, + "learning_rate": 1.515313876355692e-05, + "loss": 0.5132, + "step": 11030 + }, + { + "epoch": 0.34792619460652896, + "grad_norm": 1.9760304689407349, + "learning_rate": 1.5152263261208733e-05, + "loss": 0.5375, + "step": 11031 + }, + { + "epoch": 0.3479577353729696, + "grad_norm": 1.7476248741149902, + "learning_rate": 1.5151387705092675e-05, + "loss": 0.5295, + "step": 11032 + }, + { + "epoch": 0.34798927613941016, + "grad_norm": 1.4033222198486328, + "learning_rate": 1.5150512095217889e-05, + "loss": 0.8515, + "step": 11033 + }, + { + "epoch": 0.3480208169058508, + "grad_norm": 1.5249875783920288, + "learning_rate": 1.5149636431593507e-05, + "loss": 0.526, + "step": 11034 + }, + { + "epoch": 0.3480523576722914, + "grad_norm": 1.4846912622451782, + "learning_rate": 1.514876071422867e-05, + "loss": 0.5484, + "step": 11035 + }, + { + "epoch": 0.34808389843873205, + "grad_norm": 1.7444521188735962, + "learning_rate": 1.514788494313252e-05, + "loss": 0.5294, + "step": 11036 + }, + { + "epoch": 0.3481154392051727, + "grad_norm": 1.48554527759552, + "learning_rate": 1.5147009118314191e-05, + "loss": 0.5159, + "step": 11037 + }, + { + "epoch": 0.3481469799716133, + "grad_norm": 1.78166925907135, + "learning_rate": 1.5146133239782826e-05, + "loss": 0.478, + "step": 11038 + }, + { + "epoch": 0.34817852073805394, + "grad_norm": 1.3714251518249512, + "learning_rate": 1.5145257307547564e-05, + "loss": 0.7998, + "step": 11039 + }, + { + "epoch": 0.3482100615044946, + "grad_norm": 1.5340068340301514, + "learning_rate": 1.5144381321617548e-05, + "loss": 0.8041, + "step": 11040 + }, + { + "epoch": 0.3482416022709352, + "grad_norm": 2.40862774848938, + "learning_rate": 1.5143505282001915e-05, + "loss": 0.4995, + "step": 11041 + }, + { + "epoch": 0.34827314303737583, + "grad_norm": 1.9121482372283936, + "learning_rate": 1.5142629188709817e-05, + "loss": 0.5123, + "step": 11042 + }, + { + "epoch": 0.3483046838038164, + "grad_norm": 1.73802649974823, + "learning_rate": 1.5141753041750389e-05, + "loss": 0.5109, + "step": 11043 + }, + { + "epoch": 0.34833622457025704, + "grad_norm": 2.033416509628296, + "learning_rate": 1.5140876841132772e-05, + "loss": 0.526, + "step": 11044 + }, + { + "epoch": 0.34836776533669767, + "grad_norm": 1.780104160308838, + "learning_rate": 1.5140000586866114e-05, + "loss": 0.8208, + "step": 11045 + }, + { + "epoch": 0.3483993061031383, + "grad_norm": 1.855386734008789, + "learning_rate": 1.513912427895956e-05, + "loss": 0.8147, + "step": 11046 + }, + { + "epoch": 0.3484308468695789, + "grad_norm": 1.5229549407958984, + "learning_rate": 1.5138247917422253e-05, + "loss": 0.5234, + "step": 11047 + }, + { + "epoch": 0.34846238763601955, + "grad_norm": 1.5817221403121948, + "learning_rate": 1.5137371502263339e-05, + "loss": 0.5142, + "step": 11048 + }, + { + "epoch": 0.3484939284024602, + "grad_norm": 1.8969100713729858, + "learning_rate": 1.5136495033491962e-05, + "loss": 0.4986, + "step": 11049 + }, + { + "epoch": 0.3485254691689008, + "grad_norm": 4.27705192565918, + "learning_rate": 1.5135618511117275e-05, + "loss": 0.5402, + "step": 11050 + }, + { + "epoch": 0.34855700993534144, + "grad_norm": 2.967705249786377, + "learning_rate": 1.513474193514842e-05, + "loss": 0.4865, + "step": 11051 + }, + { + "epoch": 0.3485885507017821, + "grad_norm": 1.4744693040847778, + "learning_rate": 1.5133865305594548e-05, + "loss": 0.5426, + "step": 11052 + }, + { + "epoch": 0.3486200914682227, + "grad_norm": 1.713639497756958, + "learning_rate": 1.5132988622464804e-05, + "loss": 0.541, + "step": 11053 + }, + { + "epoch": 0.3486516322346633, + "grad_norm": 2.218216896057129, + "learning_rate": 1.5132111885768335e-05, + "loss": 0.509, + "step": 11054 + }, + { + "epoch": 0.3486831730011039, + "grad_norm": 2.320793628692627, + "learning_rate": 1.5131235095514294e-05, + "loss": 0.5146, + "step": 11055 + }, + { + "epoch": 0.34871471376754454, + "grad_norm": 1.7382268905639648, + "learning_rate": 1.5130358251711834e-05, + "loss": 0.5141, + "step": 11056 + }, + { + "epoch": 0.34874625453398517, + "grad_norm": 1.3993003368377686, + "learning_rate": 1.5129481354370099e-05, + "loss": 0.5224, + "step": 11057 + }, + { + "epoch": 0.3487777953004258, + "grad_norm": 1.1810001134872437, + "learning_rate": 1.5128604403498245e-05, + "loss": 0.4977, + "step": 11058 + }, + { + "epoch": 0.3488093360668664, + "grad_norm": 1.9825823307037354, + "learning_rate": 1.512772739910542e-05, + "loss": 0.4868, + "step": 11059 + }, + { + "epoch": 0.34884087683330706, + "grad_norm": 1.9493699073791504, + "learning_rate": 1.5126850341200779e-05, + "loss": 0.5449, + "step": 11060 + }, + { + "epoch": 0.3488724175997477, + "grad_norm": 1.580614447593689, + "learning_rate": 1.5125973229793474e-05, + "loss": 0.4969, + "step": 11061 + }, + { + "epoch": 0.3489039583661883, + "grad_norm": 1.531175971031189, + "learning_rate": 1.5125096064892657e-05, + "loss": 0.4918, + "step": 11062 + }, + { + "epoch": 0.34893549913262895, + "grad_norm": 1.2756718397140503, + "learning_rate": 1.5124218846507486e-05, + "loss": 0.3913, + "step": 11063 + }, + { + "epoch": 0.3489670398990695, + "grad_norm": 1.846732497215271, + "learning_rate": 1.5123341574647108e-05, + "loss": 0.5053, + "step": 11064 + }, + { + "epoch": 0.34899858066551015, + "grad_norm": 1.500337839126587, + "learning_rate": 1.5122464249320685e-05, + "loss": 0.5246, + "step": 11065 + }, + { + "epoch": 0.3490301214319508, + "grad_norm": 1.7260500192642212, + "learning_rate": 1.512158687053737e-05, + "loss": 0.5204, + "step": 11066 + }, + { + "epoch": 0.3490616621983914, + "grad_norm": 1.8076636791229248, + "learning_rate": 1.512070943830632e-05, + "loss": 0.8746, + "step": 11067 + }, + { + "epoch": 0.34909320296483204, + "grad_norm": 2.2693638801574707, + "learning_rate": 1.5119831952636687e-05, + "loss": 0.4737, + "step": 11068 + }, + { + "epoch": 0.34912474373127267, + "grad_norm": 2.2545509338378906, + "learning_rate": 1.5118954413537636e-05, + "loss": 0.5295, + "step": 11069 + }, + { + "epoch": 0.3491562844977133, + "grad_norm": 1.6288158893585205, + "learning_rate": 1.5118076821018322e-05, + "loss": 0.3957, + "step": 11070 + }, + { + "epoch": 0.34918782526415393, + "grad_norm": 1.367996096611023, + "learning_rate": 1.5117199175087897e-05, + "loss": 0.5442, + "step": 11071 + }, + { + "epoch": 0.34921936603059456, + "grad_norm": 1.2655082941055298, + "learning_rate": 1.5116321475755534e-05, + "loss": 0.5276, + "step": 11072 + }, + { + "epoch": 0.3492509067970352, + "grad_norm": 2.02911376953125, + "learning_rate": 1.5115443723030377e-05, + "loss": 0.5364, + "step": 11073 + }, + { + "epoch": 0.3492824475634758, + "grad_norm": 1.7218772172927856, + "learning_rate": 1.5114565916921594e-05, + "loss": 0.4975, + "step": 11074 + }, + { + "epoch": 0.3493139883299164, + "grad_norm": 1.4803190231323242, + "learning_rate": 1.5113688057438342e-05, + "loss": 0.5119, + "step": 11075 + }, + { + "epoch": 0.349345529096357, + "grad_norm": 1.6626156568527222, + "learning_rate": 1.5112810144589787e-05, + "loss": 0.5093, + "step": 11076 + }, + { + "epoch": 0.34937706986279765, + "grad_norm": 1.3330625295639038, + "learning_rate": 1.5111932178385089e-05, + "loss": 0.4926, + "step": 11077 + }, + { + "epoch": 0.3494086106292383, + "grad_norm": 1.3030927181243896, + "learning_rate": 1.5111054158833407e-05, + "loss": 0.5471, + "step": 11078 + }, + { + "epoch": 0.3494401513956789, + "grad_norm": 1.449388027191162, + "learning_rate": 1.5110176085943909e-05, + "loss": 0.5639, + "step": 11079 + }, + { + "epoch": 0.34947169216211954, + "grad_norm": 1.539125680923462, + "learning_rate": 1.5109297959725752e-05, + "loss": 0.5477, + "step": 11080 + }, + { + "epoch": 0.34950323292856017, + "grad_norm": 1.7939066886901855, + "learning_rate": 1.5108419780188107e-05, + "loss": 0.523, + "step": 11081 + }, + { + "epoch": 0.3495347736950008, + "grad_norm": 1.7367364168167114, + "learning_rate": 1.5107541547340132e-05, + "loss": 0.4869, + "step": 11082 + }, + { + "epoch": 0.34956631446144143, + "grad_norm": 1.591400146484375, + "learning_rate": 1.5106663261191e-05, + "loss": 0.5489, + "step": 11083 + }, + { + "epoch": 0.34959785522788206, + "grad_norm": 1.7214469909667969, + "learning_rate": 1.5105784921749866e-05, + "loss": 0.5152, + "step": 11084 + }, + { + "epoch": 0.3496293959943227, + "grad_norm": 2.120126485824585, + "learning_rate": 1.5104906529025904e-05, + "loss": 0.522, + "step": 11085 + }, + { + "epoch": 0.34966093676076326, + "grad_norm": 1.423924446105957, + "learning_rate": 1.510402808302828e-05, + "loss": 0.5161, + "step": 11086 + }, + { + "epoch": 0.3496924775272039, + "grad_norm": 1.5331281423568726, + "learning_rate": 1.5103149583766156e-05, + "loss": 0.5299, + "step": 11087 + }, + { + "epoch": 0.3497240182936445, + "grad_norm": 1.815454125404358, + "learning_rate": 1.5102271031248706e-05, + "loss": 0.539, + "step": 11088 + }, + { + "epoch": 0.34975555906008515, + "grad_norm": 1.3616328239440918, + "learning_rate": 1.5101392425485096e-05, + "loss": 0.5203, + "step": 11089 + }, + { + "epoch": 0.3497870998265258, + "grad_norm": 1.241912603378296, + "learning_rate": 1.5100513766484493e-05, + "loss": 0.3654, + "step": 11090 + }, + { + "epoch": 0.3498186405929664, + "grad_norm": 1.7201298475265503, + "learning_rate": 1.5099635054256071e-05, + "loss": 0.8692, + "step": 11091 + }, + { + "epoch": 0.34985018135940704, + "grad_norm": 1.6676820516586304, + "learning_rate": 1.5098756288809e-05, + "loss": 0.531, + "step": 11092 + }, + { + "epoch": 0.34988172212584767, + "grad_norm": 2.0300185680389404, + "learning_rate": 1.5097877470152443e-05, + "loss": 0.5032, + "step": 11093 + }, + { + "epoch": 0.3499132628922883, + "grad_norm": 1.596695065498352, + "learning_rate": 1.5096998598295579e-05, + "loss": 0.8542, + "step": 11094 + }, + { + "epoch": 0.34994480365872893, + "grad_norm": 4.423717021942139, + "learning_rate": 1.5096119673247575e-05, + "loss": 0.5079, + "step": 11095 + }, + { + "epoch": 0.3499763444251695, + "grad_norm": 1.5722330808639526, + "learning_rate": 1.5095240695017606e-05, + "loss": 0.5109, + "step": 11096 + }, + { + "epoch": 0.35000788519161014, + "grad_norm": 2.8642475605010986, + "learning_rate": 1.5094361663614843e-05, + "loss": 0.5293, + "step": 11097 + }, + { + "epoch": 0.35003942595805076, + "grad_norm": 1.8287289142608643, + "learning_rate": 1.5093482579048461e-05, + "loss": 0.5165, + "step": 11098 + }, + { + "epoch": 0.3500709667244914, + "grad_norm": 2.037470579147339, + "learning_rate": 1.5092603441327635e-05, + "loss": 0.4943, + "step": 11099 + }, + { + "epoch": 0.350102507490932, + "grad_norm": 1.790403962135315, + "learning_rate": 1.5091724250461534e-05, + "loss": 0.5415, + "step": 11100 + }, + { + "epoch": 0.35013404825737265, + "grad_norm": 1.4916988611221313, + "learning_rate": 1.5090845006459342e-05, + "loss": 0.53, + "step": 11101 + }, + { + "epoch": 0.3501655890238133, + "grad_norm": 1.4573423862457275, + "learning_rate": 1.5089965709330227e-05, + "loss": 0.5383, + "step": 11102 + }, + { + "epoch": 0.3501971297902539, + "grad_norm": 1.8010034561157227, + "learning_rate": 1.5089086359083368e-05, + "loss": 0.5105, + "step": 11103 + }, + { + "epoch": 0.35022867055669454, + "grad_norm": 5.472100734710693, + "learning_rate": 1.5088206955727942e-05, + "loss": 0.5044, + "step": 11104 + }, + { + "epoch": 0.3502602113231352, + "grad_norm": 1.8617546558380127, + "learning_rate": 1.5087327499273121e-05, + "loss": 0.5312, + "step": 11105 + }, + { + "epoch": 0.3502917520895758, + "grad_norm": 1.7209938764572144, + "learning_rate": 1.508644798972809e-05, + "loss": 0.4844, + "step": 11106 + }, + { + "epoch": 0.3503232928560164, + "grad_norm": 1.9960297346115112, + "learning_rate": 1.5085568427102026e-05, + "loss": 0.5414, + "step": 11107 + }, + { + "epoch": 0.350354833622457, + "grad_norm": 1.5207982063293457, + "learning_rate": 1.5084688811404107e-05, + "loss": 0.5363, + "step": 11108 + }, + { + "epoch": 0.35038637438889764, + "grad_norm": 1.6514760255813599, + "learning_rate": 1.5083809142643509e-05, + "loss": 0.518, + "step": 11109 + }, + { + "epoch": 0.35041791515533827, + "grad_norm": 1.54685378074646, + "learning_rate": 1.5082929420829416e-05, + "loss": 0.5033, + "step": 11110 + }, + { + "epoch": 0.3504494559217789, + "grad_norm": 1.5591816902160645, + "learning_rate": 1.5082049645971012e-05, + "loss": 0.5236, + "step": 11111 + }, + { + "epoch": 0.3504809966882195, + "grad_norm": 1.6216124296188354, + "learning_rate": 1.508116981807747e-05, + "loss": 0.4817, + "step": 11112 + }, + { + "epoch": 0.35051253745466016, + "grad_norm": 1.7065119743347168, + "learning_rate": 1.5080289937157975e-05, + "loss": 0.5163, + "step": 11113 + }, + { + "epoch": 0.3505440782211008, + "grad_norm": 1.4020928144454956, + "learning_rate": 1.507941000322171e-05, + "loss": 0.4832, + "step": 11114 + }, + { + "epoch": 0.3505756189875414, + "grad_norm": 1.4636527299880981, + "learning_rate": 1.507853001627786e-05, + "loss": 0.5032, + "step": 11115 + }, + { + "epoch": 0.35060715975398205, + "grad_norm": 1.9027504920959473, + "learning_rate": 1.5077649976335602e-05, + "loss": 0.5198, + "step": 11116 + }, + { + "epoch": 0.3506387005204226, + "grad_norm": 3.0564329624176025, + "learning_rate": 1.5076769883404128e-05, + "loss": 0.5175, + "step": 11117 + }, + { + "epoch": 0.35067024128686325, + "grad_norm": 1.7536132335662842, + "learning_rate": 1.5075889737492614e-05, + "loss": 0.4846, + "step": 11118 + }, + { + "epoch": 0.3507017820533039, + "grad_norm": 1.811607837677002, + "learning_rate": 1.5075009538610252e-05, + "loss": 0.509, + "step": 11119 + }, + { + "epoch": 0.3507333228197445, + "grad_norm": 1.7482056617736816, + "learning_rate": 1.5074129286766224e-05, + "loss": 0.5258, + "step": 11120 + }, + { + "epoch": 0.35076486358618514, + "grad_norm": 1.628178358078003, + "learning_rate": 1.5073248981969717e-05, + "loss": 0.5076, + "step": 11121 + }, + { + "epoch": 0.35079640435262577, + "grad_norm": 2.042001247406006, + "learning_rate": 1.507236862422992e-05, + "loss": 0.4978, + "step": 11122 + }, + { + "epoch": 0.3508279451190664, + "grad_norm": 3.297020673751831, + "learning_rate": 1.5071488213556014e-05, + "loss": 0.9087, + "step": 11123 + }, + { + "epoch": 0.350859485885507, + "grad_norm": 1.6200414896011353, + "learning_rate": 1.5070607749957193e-05, + "loss": 0.5309, + "step": 11124 + }, + { + "epoch": 0.35089102665194766, + "grad_norm": 3.91961669921875, + "learning_rate": 1.5069727233442644e-05, + "loss": 0.4686, + "step": 11125 + }, + { + "epoch": 0.3509225674183883, + "grad_norm": 2.841543197631836, + "learning_rate": 1.5068846664021553e-05, + "loss": 0.5278, + "step": 11126 + }, + { + "epoch": 0.3509541081848289, + "grad_norm": 1.6256861686706543, + "learning_rate": 1.5067966041703109e-05, + "loss": 0.56, + "step": 11127 + }, + { + "epoch": 0.3509856489512695, + "grad_norm": 2.385496139526367, + "learning_rate": 1.5067085366496508e-05, + "loss": 0.5499, + "step": 11128 + }, + { + "epoch": 0.3510171897177101, + "grad_norm": 1.7379150390625, + "learning_rate": 1.5066204638410935e-05, + "loss": 0.5339, + "step": 11129 + }, + { + "epoch": 0.35104873048415075, + "grad_norm": 1.9948972463607788, + "learning_rate": 1.506532385745558e-05, + "loss": 0.5077, + "step": 11130 + }, + { + "epoch": 0.3510802712505914, + "grad_norm": 1.7834504842758179, + "learning_rate": 1.5064443023639643e-05, + "loss": 0.5265, + "step": 11131 + }, + { + "epoch": 0.351111812017032, + "grad_norm": 2.160663366317749, + "learning_rate": 1.5063562136972309e-05, + "loss": 0.5384, + "step": 11132 + }, + { + "epoch": 0.35114335278347264, + "grad_norm": 2.1310877799987793, + "learning_rate": 1.5062681197462769e-05, + "loss": 0.5294, + "step": 11133 + }, + { + "epoch": 0.35117489354991327, + "grad_norm": 1.931631326675415, + "learning_rate": 1.5061800205120221e-05, + "loss": 0.521, + "step": 11134 + }, + { + "epoch": 0.3512064343163539, + "grad_norm": 1.636881947517395, + "learning_rate": 1.506091915995386e-05, + "loss": 0.5241, + "step": 11135 + }, + { + "epoch": 0.35123797508279453, + "grad_norm": 1.433558702468872, + "learning_rate": 1.5060038061972875e-05, + "loss": 0.4483, + "step": 11136 + }, + { + "epoch": 0.35126951584923516, + "grad_norm": 1.8150358200073242, + "learning_rate": 1.5059156911186465e-05, + "loss": 0.5164, + "step": 11137 + }, + { + "epoch": 0.3513010566156758, + "grad_norm": 1.7618058919906616, + "learning_rate": 1.5058275707603822e-05, + "loss": 0.5441, + "step": 11138 + }, + { + "epoch": 0.35133259738211636, + "grad_norm": 2.1317062377929688, + "learning_rate": 1.5057394451234149e-05, + "loss": 0.5144, + "step": 11139 + }, + { + "epoch": 0.351364138148557, + "grad_norm": 1.6321601867675781, + "learning_rate": 1.5056513142086636e-05, + "loss": 0.5218, + "step": 11140 + }, + { + "epoch": 0.3513956789149976, + "grad_norm": 1.6564399003982544, + "learning_rate": 1.5055631780170484e-05, + "loss": 0.5172, + "step": 11141 + }, + { + "epoch": 0.35142721968143825, + "grad_norm": 1.488629937171936, + "learning_rate": 1.5054750365494886e-05, + "loss": 0.5101, + "step": 11142 + }, + { + "epoch": 0.3514587604478789, + "grad_norm": 2.121896743774414, + "learning_rate": 1.5053868898069045e-05, + "loss": 0.4925, + "step": 11143 + }, + { + "epoch": 0.3514903012143195, + "grad_norm": 1.5127103328704834, + "learning_rate": 1.5052987377902153e-05, + "loss": 0.4141, + "step": 11144 + }, + { + "epoch": 0.35152184198076014, + "grad_norm": 2.002048969268799, + "learning_rate": 1.5052105805003416e-05, + "loss": 0.5065, + "step": 11145 + }, + { + "epoch": 0.35155338274720077, + "grad_norm": 1.6787430047988892, + "learning_rate": 1.5051224179382035e-05, + "loss": 0.5484, + "step": 11146 + }, + { + "epoch": 0.3515849235136414, + "grad_norm": 2.9636409282684326, + "learning_rate": 1.5050342501047206e-05, + "loss": 0.8753, + "step": 11147 + }, + { + "epoch": 0.35161646428008203, + "grad_norm": 1.9615504741668701, + "learning_rate": 1.5049460770008131e-05, + "loss": 0.5501, + "step": 11148 + }, + { + "epoch": 0.3516480050465226, + "grad_norm": 1.8848451375961304, + "learning_rate": 1.5048578986274013e-05, + "loss": 0.5434, + "step": 11149 + }, + { + "epoch": 0.35167954581296323, + "grad_norm": 1.6318769454956055, + "learning_rate": 1.5047697149854052e-05, + "loss": 0.8309, + "step": 11150 + }, + { + "epoch": 0.35171108657940386, + "grad_norm": 1.4639606475830078, + "learning_rate": 1.5046815260757455e-05, + "loss": 0.5407, + "step": 11151 + }, + { + "epoch": 0.3517426273458445, + "grad_norm": 2.3747849464416504, + "learning_rate": 1.5045933318993419e-05, + "loss": 0.5249, + "step": 11152 + }, + { + "epoch": 0.3517741681122851, + "grad_norm": 2.923966407775879, + "learning_rate": 1.5045051324571152e-05, + "loss": 0.5325, + "step": 11153 + }, + { + "epoch": 0.35180570887872575, + "grad_norm": 2.0951058864593506, + "learning_rate": 1.5044169277499855e-05, + "loss": 0.5317, + "step": 11154 + }, + { + "epoch": 0.3518372496451664, + "grad_norm": 1.8713607788085938, + "learning_rate": 1.5043287177788734e-05, + "loss": 0.4957, + "step": 11155 + }, + { + "epoch": 0.351868790411607, + "grad_norm": 1.5833380222320557, + "learning_rate": 1.5042405025447e-05, + "loss": 0.5362, + "step": 11156 + }, + { + "epoch": 0.35190033117804764, + "grad_norm": 1.3431355953216553, + "learning_rate": 1.5041522820483852e-05, + "loss": 0.5064, + "step": 11157 + }, + { + "epoch": 0.3519318719444883, + "grad_norm": 3.355858087539673, + "learning_rate": 1.5040640562908497e-05, + "loss": 0.4002, + "step": 11158 + }, + { + "epoch": 0.3519634127109289, + "grad_norm": 1.6055247783660889, + "learning_rate": 1.5039758252730145e-05, + "loss": 0.5062, + "step": 11159 + }, + { + "epoch": 0.3519949534773695, + "grad_norm": 1.4347057342529297, + "learning_rate": 1.5038875889958005e-05, + "loss": 0.4994, + "step": 11160 + }, + { + "epoch": 0.3520264942438101, + "grad_norm": 1.8455196619033813, + "learning_rate": 1.503799347460128e-05, + "loss": 0.5307, + "step": 11161 + }, + { + "epoch": 0.35205803501025074, + "grad_norm": 1.670097827911377, + "learning_rate": 1.503711100666918e-05, + "loss": 0.4822, + "step": 11162 + }, + { + "epoch": 0.35208957577669137, + "grad_norm": 1.4384753704071045, + "learning_rate": 1.5036228486170918e-05, + "loss": 0.428, + "step": 11163 + }, + { + "epoch": 0.352121116543132, + "grad_norm": 1.6394140720367432, + "learning_rate": 1.50353459131157e-05, + "loss": 0.5016, + "step": 11164 + }, + { + "epoch": 0.3521526573095726, + "grad_norm": 1.685172200202942, + "learning_rate": 1.5034463287512737e-05, + "loss": 0.5208, + "step": 11165 + }, + { + "epoch": 0.35218419807601326, + "grad_norm": 5.6306562423706055, + "learning_rate": 1.5033580609371242e-05, + "loss": 0.4128, + "step": 11166 + }, + { + "epoch": 0.3522157388424539, + "grad_norm": 1.5179656744003296, + "learning_rate": 1.5032697878700422e-05, + "loss": 0.531, + "step": 11167 + }, + { + "epoch": 0.3522472796088945, + "grad_norm": 1.5506839752197266, + "learning_rate": 1.5031815095509492e-05, + "loss": 0.5183, + "step": 11168 + }, + { + "epoch": 0.35227882037533514, + "grad_norm": 1.71352219581604, + "learning_rate": 1.503093225980767e-05, + "loss": 0.5035, + "step": 11169 + }, + { + "epoch": 0.3523103611417757, + "grad_norm": 2.709826946258545, + "learning_rate": 1.5030049371604158e-05, + "loss": 0.8559, + "step": 11170 + }, + { + "epoch": 0.35234190190821635, + "grad_norm": 1.8628374338150024, + "learning_rate": 1.5029166430908178e-05, + "loss": 0.3976, + "step": 11171 + }, + { + "epoch": 0.352373442674657, + "grad_norm": 4.420161724090576, + "learning_rate": 1.502828343772894e-05, + "loss": 0.5092, + "step": 11172 + }, + { + "epoch": 0.3524049834410976, + "grad_norm": 2.057760000228882, + "learning_rate": 1.5027400392075659e-05, + "loss": 0.527, + "step": 11173 + }, + { + "epoch": 0.35243652420753824, + "grad_norm": 1.5540894269943237, + "learning_rate": 1.5026517293957552e-05, + "loss": 0.8365, + "step": 11174 + }, + { + "epoch": 0.35246806497397887, + "grad_norm": 1.9038211107254028, + "learning_rate": 1.5025634143383832e-05, + "loss": 0.4794, + "step": 11175 + }, + { + "epoch": 0.3524996057404195, + "grad_norm": 1.462317943572998, + "learning_rate": 1.5024750940363722e-05, + "loss": 0.8413, + "step": 11176 + }, + { + "epoch": 0.3525311465068601, + "grad_norm": 2.8263518810272217, + "learning_rate": 1.502386768490643e-05, + "loss": 0.8796, + "step": 11177 + }, + { + "epoch": 0.35256268727330076, + "grad_norm": 2.3478620052337646, + "learning_rate": 1.502298437702118e-05, + "loss": 0.5225, + "step": 11178 + }, + { + "epoch": 0.3525942280397414, + "grad_norm": 1.7709707021713257, + "learning_rate": 1.5022101016717187e-05, + "loss": 0.5125, + "step": 11179 + }, + { + "epoch": 0.352625768806182, + "grad_norm": 5.254807949066162, + "learning_rate": 1.5021217604003671e-05, + "loss": 0.5261, + "step": 11180 + }, + { + "epoch": 0.3526573095726226, + "grad_norm": 1.5266919136047363, + "learning_rate": 1.5020334138889851e-05, + "loss": 0.5126, + "step": 11181 + }, + { + "epoch": 0.3526888503390632, + "grad_norm": 2.201605796813965, + "learning_rate": 1.5019450621384946e-05, + "loss": 0.5016, + "step": 11182 + }, + { + "epoch": 0.35272039110550385, + "grad_norm": 1.8341736793518066, + "learning_rate": 1.5018567051498174e-05, + "loss": 0.5026, + "step": 11183 + }, + { + "epoch": 0.3527519318719445, + "grad_norm": 2.207920789718628, + "learning_rate": 1.501768342923876e-05, + "loss": 0.4641, + "step": 11184 + }, + { + "epoch": 0.3527834726383851, + "grad_norm": 2.049182891845703, + "learning_rate": 1.5016799754615922e-05, + "loss": 0.5141, + "step": 11185 + }, + { + "epoch": 0.35281501340482574, + "grad_norm": 9.728013038635254, + "learning_rate": 1.5015916027638882e-05, + "loss": 0.5322, + "step": 11186 + }, + { + "epoch": 0.35284655417126637, + "grad_norm": 2.3452775478363037, + "learning_rate": 1.5015032248316867e-05, + "loss": 0.8568, + "step": 11187 + }, + { + "epoch": 0.352878094937707, + "grad_norm": 4.861794948577881, + "learning_rate": 1.5014148416659095e-05, + "loss": 0.5247, + "step": 11188 + }, + { + "epoch": 0.35290963570414763, + "grad_norm": 1.8310774564743042, + "learning_rate": 1.5013264532674791e-05, + "loss": 0.5289, + "step": 11189 + }, + { + "epoch": 0.35294117647058826, + "grad_norm": 2.7586803436279297, + "learning_rate": 1.5012380596373176e-05, + "loss": 0.5284, + "step": 11190 + }, + { + "epoch": 0.35297271723702883, + "grad_norm": 2.4665987491607666, + "learning_rate": 1.5011496607763483e-05, + "loss": 0.8301, + "step": 11191 + }, + { + "epoch": 0.35300425800346946, + "grad_norm": 1.6853387355804443, + "learning_rate": 1.5010612566854929e-05, + "loss": 0.3718, + "step": 11192 + }, + { + "epoch": 0.3530357987699101, + "grad_norm": 1.9907671213150024, + "learning_rate": 1.5009728473656741e-05, + "loss": 0.5207, + "step": 11193 + }, + { + "epoch": 0.3530673395363507, + "grad_norm": 2.7208919525146484, + "learning_rate": 1.5008844328178146e-05, + "loss": 0.5126, + "step": 11194 + }, + { + "epoch": 0.35309888030279135, + "grad_norm": 3.3332831859588623, + "learning_rate": 1.5007960130428374e-05, + "loss": 0.5252, + "step": 11195 + }, + { + "epoch": 0.353130421069232, + "grad_norm": 2.2209720611572266, + "learning_rate": 1.5007075880416646e-05, + "loss": 0.5761, + "step": 11196 + }, + { + "epoch": 0.3531619618356726, + "grad_norm": 2.566041946411133, + "learning_rate": 1.5006191578152197e-05, + "loss": 0.5216, + "step": 11197 + }, + { + "epoch": 0.35319350260211324, + "grad_norm": 1.8063733577728271, + "learning_rate": 1.5005307223644248e-05, + "loss": 0.5325, + "step": 11198 + }, + { + "epoch": 0.35322504336855387, + "grad_norm": 1.8653961420059204, + "learning_rate": 1.5004422816902035e-05, + "loss": 0.526, + "step": 11199 + }, + { + "epoch": 0.3532565841349945, + "grad_norm": 2.2240495681762695, + "learning_rate": 1.5003538357934782e-05, + "loss": 0.5324, + "step": 11200 + }, + { + "epoch": 0.35328812490143513, + "grad_norm": 2.28714919090271, + "learning_rate": 1.5002653846751723e-05, + "loss": 0.5532, + "step": 11201 + }, + { + "epoch": 0.3533196656678757, + "grad_norm": 2.4935367107391357, + "learning_rate": 1.5001769283362083e-05, + "loss": 0.5385, + "step": 11202 + }, + { + "epoch": 0.35335120643431633, + "grad_norm": 1.869188666343689, + "learning_rate": 1.5000884667775099e-05, + "loss": 0.5296, + "step": 11203 + }, + { + "epoch": 0.35338274720075696, + "grad_norm": 2.10680890083313, + "learning_rate": 1.5000000000000002e-05, + "loss": 0.5251, + "step": 11204 + }, + { + "epoch": 0.3534142879671976, + "grad_norm": 2.5213117599487305, + "learning_rate": 1.4999115280046019e-05, + "loss": 0.5127, + "step": 11205 + }, + { + "epoch": 0.3534458287336382, + "grad_norm": 1.753695011138916, + "learning_rate": 1.4998230507922389e-05, + "loss": 0.5219, + "step": 11206 + }, + { + "epoch": 0.35347736950007885, + "grad_norm": 4.299899578094482, + "learning_rate": 1.4997345683638343e-05, + "loss": 0.8633, + "step": 11207 + }, + { + "epoch": 0.3535089102665195, + "grad_norm": 1.7054047584533691, + "learning_rate": 1.4996460807203112e-05, + "loss": 0.5044, + "step": 11208 + }, + { + "epoch": 0.3535404510329601, + "grad_norm": 2.5488758087158203, + "learning_rate": 1.4995575878625937e-05, + "loss": 0.5315, + "step": 11209 + }, + { + "epoch": 0.35357199179940074, + "grad_norm": 1.988586187362671, + "learning_rate": 1.4994690897916044e-05, + "loss": 0.4828, + "step": 11210 + }, + { + "epoch": 0.35360353256584137, + "grad_norm": 1.9045060873031616, + "learning_rate": 1.4993805865082678e-05, + "loss": 0.527, + "step": 11211 + }, + { + "epoch": 0.353635073332282, + "grad_norm": 2.0402963161468506, + "learning_rate": 1.4992920780135066e-05, + "loss": 0.8401, + "step": 11212 + }, + { + "epoch": 0.3536666140987226, + "grad_norm": 1.5260281562805176, + "learning_rate": 1.4992035643082447e-05, + "loss": 0.5141, + "step": 11213 + }, + { + "epoch": 0.3536981548651632, + "grad_norm": 1.89044988155365, + "learning_rate": 1.4991150453934067e-05, + "loss": 0.5007, + "step": 11214 + }, + { + "epoch": 0.35372969563160384, + "grad_norm": 2.456176996231079, + "learning_rate": 1.499026521269915e-05, + "loss": 0.5139, + "step": 11215 + }, + { + "epoch": 0.35376123639804447, + "grad_norm": 2.0754075050354004, + "learning_rate": 1.4989379919386944e-05, + "loss": 0.5455, + "step": 11216 + }, + { + "epoch": 0.3537927771644851, + "grad_norm": 1.8480463027954102, + "learning_rate": 1.4988494574006682e-05, + "loss": 0.5118, + "step": 11217 + }, + { + "epoch": 0.3538243179309257, + "grad_norm": 1.7014154195785522, + "learning_rate": 1.4987609176567607e-05, + "loss": 0.5208, + "step": 11218 + }, + { + "epoch": 0.35385585869736635, + "grad_norm": 1.743956446647644, + "learning_rate": 1.498672372707896e-05, + "loss": 0.4864, + "step": 11219 + }, + { + "epoch": 0.353887399463807, + "grad_norm": 2.3238511085510254, + "learning_rate": 1.4985838225549975e-05, + "loss": 0.5315, + "step": 11220 + }, + { + "epoch": 0.3539189402302476, + "grad_norm": 1.7482447624206543, + "learning_rate": 1.49849526719899e-05, + "loss": 0.5287, + "step": 11221 + }, + { + "epoch": 0.35395048099668824, + "grad_norm": 2.390885829925537, + "learning_rate": 1.498406706640797e-05, + "loss": 0.5304, + "step": 11222 + }, + { + "epoch": 0.3539820217631288, + "grad_norm": 2.0645861625671387, + "learning_rate": 1.4983181408813432e-05, + "loss": 0.4706, + "step": 11223 + }, + { + "epoch": 0.35401356252956945, + "grad_norm": 1.6454110145568848, + "learning_rate": 1.4982295699215525e-05, + "loss": 0.5233, + "step": 11224 + }, + { + "epoch": 0.3540451032960101, + "grad_norm": 2.1297717094421387, + "learning_rate": 1.4981409937623497e-05, + "loss": 0.5027, + "step": 11225 + }, + { + "epoch": 0.3540766440624507, + "grad_norm": 1.6794904470443726, + "learning_rate": 1.4980524124046585e-05, + "loss": 0.514, + "step": 11226 + }, + { + "epoch": 0.35410818482889134, + "grad_norm": 1.8034917116165161, + "learning_rate": 1.4979638258494036e-05, + "loss": 0.5298, + "step": 11227 + }, + { + "epoch": 0.35413972559533197, + "grad_norm": 1.9180134534835815, + "learning_rate": 1.4978752340975098e-05, + "loss": 0.5009, + "step": 11228 + }, + { + "epoch": 0.3541712663617726, + "grad_norm": 1.3081343173980713, + "learning_rate": 1.4977866371499015e-05, + "loss": 0.5137, + "step": 11229 + }, + { + "epoch": 0.3542028071282132, + "grad_norm": 1.726974368095398, + "learning_rate": 1.4976980350075025e-05, + "loss": 0.5304, + "step": 11230 + }, + { + "epoch": 0.35423434789465386, + "grad_norm": 1.798707127571106, + "learning_rate": 1.4976094276712387e-05, + "loss": 0.5151, + "step": 11231 + }, + { + "epoch": 0.3542658886610945, + "grad_norm": 1.4928693771362305, + "learning_rate": 1.4975208151420337e-05, + "loss": 0.5395, + "step": 11232 + }, + { + "epoch": 0.3542974294275351, + "grad_norm": 1.5281556844711304, + "learning_rate": 1.4974321974208128e-05, + "loss": 0.5032, + "step": 11233 + }, + { + "epoch": 0.3543289701939757, + "grad_norm": 1.634488821029663, + "learning_rate": 1.4973435745085006e-05, + "loss": 0.5685, + "step": 11234 + }, + { + "epoch": 0.3543605109604163, + "grad_norm": 1.9138891696929932, + "learning_rate": 1.4972549464060222e-05, + "loss": 0.553, + "step": 11235 + }, + { + "epoch": 0.35439205172685695, + "grad_norm": 1.4746044874191284, + "learning_rate": 1.4971663131143021e-05, + "loss": 0.5339, + "step": 11236 + }, + { + "epoch": 0.3544235924932976, + "grad_norm": 1.5804295539855957, + "learning_rate": 1.4970776746342656e-05, + "loss": 0.5361, + "step": 11237 + }, + { + "epoch": 0.3544551332597382, + "grad_norm": 1.9773396253585815, + "learning_rate": 1.4969890309668378e-05, + "loss": 0.499, + "step": 11238 + }, + { + "epoch": 0.35448667402617884, + "grad_norm": 1.5092990398406982, + "learning_rate": 1.4969003821129435e-05, + "loss": 0.54, + "step": 11239 + }, + { + "epoch": 0.35451821479261947, + "grad_norm": 2.872514486312866, + "learning_rate": 1.4968117280735077e-05, + "loss": 0.8955, + "step": 11240 + }, + { + "epoch": 0.3545497555590601, + "grad_norm": 1.466500163078308, + "learning_rate": 1.4967230688494556e-05, + "loss": 0.5288, + "step": 11241 + }, + { + "epoch": 0.3545812963255007, + "grad_norm": 1.6439063549041748, + "learning_rate": 1.496634404441713e-05, + "loss": 0.5139, + "step": 11242 + }, + { + "epoch": 0.35461283709194136, + "grad_norm": 1.5504683256149292, + "learning_rate": 1.4965457348512045e-05, + "loss": 0.5117, + "step": 11243 + }, + { + "epoch": 0.35464437785838193, + "grad_norm": 1.2849059104919434, + "learning_rate": 1.4964570600788557e-05, + "loss": 0.388, + "step": 11244 + }, + { + "epoch": 0.35467591862482256, + "grad_norm": 4.045576572418213, + "learning_rate": 1.4963683801255921e-05, + "loss": 0.5589, + "step": 11245 + }, + { + "epoch": 0.3547074593912632, + "grad_norm": 1.5576914548873901, + "learning_rate": 1.4962796949923389e-05, + "loss": 0.5026, + "step": 11246 + }, + { + "epoch": 0.3547390001577038, + "grad_norm": 1.7531532049179077, + "learning_rate": 1.496191004680022e-05, + "loss": 0.526, + "step": 11247 + }, + { + "epoch": 0.35477054092414445, + "grad_norm": 1.3308686017990112, + "learning_rate": 1.4961023091895664e-05, + "loss": 0.8034, + "step": 11248 + }, + { + "epoch": 0.3548020816905851, + "grad_norm": 1.3788129091262817, + "learning_rate": 1.4960136085218981e-05, + "loss": 0.507, + "step": 11249 + }, + { + "epoch": 0.3548336224570257, + "grad_norm": 1.6054474115371704, + "learning_rate": 1.4959249026779426e-05, + "loss": 0.5208, + "step": 11250 + }, + { + "epoch": 0.35486516322346634, + "grad_norm": 1.8512574434280396, + "learning_rate": 1.4958361916586257e-05, + "loss": 0.8197, + "step": 11251 + }, + { + "epoch": 0.35489670398990697, + "grad_norm": 1.7707983255386353, + "learning_rate": 1.4957474754648732e-05, + "loss": 0.3871, + "step": 11252 + }, + { + "epoch": 0.3549282447563476, + "grad_norm": 1.6962138414382935, + "learning_rate": 1.4956587540976106e-05, + "loss": 0.4899, + "step": 11253 + }, + { + "epoch": 0.35495978552278823, + "grad_norm": 1.5647106170654297, + "learning_rate": 1.4955700275577643e-05, + "loss": 0.8047, + "step": 11254 + }, + { + "epoch": 0.3549913262892288, + "grad_norm": 1.9236425161361694, + "learning_rate": 1.4954812958462599e-05, + "loss": 0.4849, + "step": 11255 + }, + { + "epoch": 0.35502286705566943, + "grad_norm": 1.6759458780288696, + "learning_rate": 1.4953925589640235e-05, + "loss": 0.4857, + "step": 11256 + }, + { + "epoch": 0.35505440782211006, + "grad_norm": 4.02466344833374, + "learning_rate": 1.4953038169119809e-05, + "loss": 0.4023, + "step": 11257 + }, + { + "epoch": 0.3550859485885507, + "grad_norm": 1.5984758138656616, + "learning_rate": 1.4952150696910586e-05, + "loss": 0.5241, + "step": 11258 + }, + { + "epoch": 0.3551174893549913, + "grad_norm": 1.7802201509475708, + "learning_rate": 1.4951263173021823e-05, + "loss": 0.5114, + "step": 11259 + }, + { + "epoch": 0.35514903012143195, + "grad_norm": 1.3091264963150024, + "learning_rate": 1.4950375597462787e-05, + "loss": 0.4938, + "step": 11260 + }, + { + "epoch": 0.3551805708878726, + "grad_norm": 1.4038817882537842, + "learning_rate": 1.4949487970242733e-05, + "loss": 0.5278, + "step": 11261 + }, + { + "epoch": 0.3552121116543132, + "grad_norm": 1.9943859577178955, + "learning_rate": 1.4948600291370934e-05, + "loss": 0.54, + "step": 11262 + }, + { + "epoch": 0.35524365242075384, + "grad_norm": 1.6943449974060059, + "learning_rate": 1.4947712560856646e-05, + "loss": 0.5269, + "step": 11263 + }, + { + "epoch": 0.35527519318719447, + "grad_norm": 3.012526273727417, + "learning_rate": 1.4946824778709136e-05, + "loss": 0.4, + "step": 11264 + }, + { + "epoch": 0.3553067339536351, + "grad_norm": 2.47921085357666, + "learning_rate": 1.4945936944937666e-05, + "loss": 0.5485, + "step": 11265 + }, + { + "epoch": 0.3553382747200757, + "grad_norm": 4.801315784454346, + "learning_rate": 1.494504905955151e-05, + "loss": 0.4841, + "step": 11266 + }, + { + "epoch": 0.3553698154865163, + "grad_norm": 1.934852123260498, + "learning_rate": 1.4944161122559921e-05, + "loss": 0.5178, + "step": 11267 + }, + { + "epoch": 0.35540135625295693, + "grad_norm": 1.5411778688430786, + "learning_rate": 1.4943273133972173e-05, + "loss": 0.5599, + "step": 11268 + }, + { + "epoch": 0.35543289701939756, + "grad_norm": 1.7392387390136719, + "learning_rate": 1.4942385093797535e-05, + "loss": 0.5268, + "step": 11269 + }, + { + "epoch": 0.3554644377858382, + "grad_norm": 1.8097853660583496, + "learning_rate": 1.4941497002045269e-05, + "loss": 0.8326, + "step": 11270 + }, + { + "epoch": 0.3554959785522788, + "grad_norm": 4.063976287841797, + "learning_rate": 1.494060885872464e-05, + "loss": 0.5454, + "step": 11271 + }, + { + "epoch": 0.35552751931871945, + "grad_norm": 1.926897406578064, + "learning_rate": 1.4939720663844928e-05, + "loss": 0.5183, + "step": 11272 + }, + { + "epoch": 0.3555590600851601, + "grad_norm": 1.778882622718811, + "learning_rate": 1.4938832417415393e-05, + "loss": 0.5242, + "step": 11273 + }, + { + "epoch": 0.3555906008516007, + "grad_norm": 1.9477512836456299, + "learning_rate": 1.4937944119445305e-05, + "loss": 0.4318, + "step": 11274 + }, + { + "epoch": 0.35562214161804134, + "grad_norm": 1.5682768821716309, + "learning_rate": 1.4937055769943938e-05, + "loss": 0.5297, + "step": 11275 + }, + { + "epoch": 0.3556536823844819, + "grad_norm": 1.4356625080108643, + "learning_rate": 1.493616736892056e-05, + "loss": 0.5086, + "step": 11276 + }, + { + "epoch": 0.35568522315092255, + "grad_norm": 1.2898503541946411, + "learning_rate": 1.4935278916384443e-05, + "loss": 0.4782, + "step": 11277 + }, + { + "epoch": 0.3557167639173632, + "grad_norm": 1.7510379552841187, + "learning_rate": 1.493439041234486e-05, + "loss": 0.4931, + "step": 11278 + }, + { + "epoch": 0.3557483046838038, + "grad_norm": 1.6757014989852905, + "learning_rate": 1.4933501856811078e-05, + "loss": 0.5113, + "step": 11279 + }, + { + "epoch": 0.35577984545024444, + "grad_norm": 1.550889253616333, + "learning_rate": 1.4932613249792377e-05, + "loss": 0.8228, + "step": 11280 + }, + { + "epoch": 0.35581138621668507, + "grad_norm": 1.8149751424789429, + "learning_rate": 1.4931724591298022e-05, + "loss": 0.4838, + "step": 11281 + }, + { + "epoch": 0.3558429269831257, + "grad_norm": 1.667015552520752, + "learning_rate": 1.4930835881337294e-05, + "loss": 0.4746, + "step": 11282 + }, + { + "epoch": 0.3558744677495663, + "grad_norm": 2.3408546447753906, + "learning_rate": 1.4929947119919466e-05, + "loss": 0.531, + "step": 11283 + }, + { + "epoch": 0.35590600851600696, + "grad_norm": 1.5217839479446411, + "learning_rate": 1.492905830705381e-05, + "loss": 0.5485, + "step": 11284 + }, + { + "epoch": 0.3559375492824476, + "grad_norm": 2.170809030532837, + "learning_rate": 1.4928169442749603e-05, + "loss": 0.5247, + "step": 11285 + }, + { + "epoch": 0.3559690900488882, + "grad_norm": 1.244773030281067, + "learning_rate": 1.4927280527016123e-05, + "loss": 0.5101, + "step": 11286 + }, + { + "epoch": 0.3560006308153288, + "grad_norm": 1.4044703245162964, + "learning_rate": 1.4926391559862642e-05, + "loss": 0.5247, + "step": 11287 + }, + { + "epoch": 0.3560321715817694, + "grad_norm": 2.0338807106018066, + "learning_rate": 1.4925502541298445e-05, + "loss": 0.4758, + "step": 11288 + }, + { + "epoch": 0.35606371234821005, + "grad_norm": 2.5606517791748047, + "learning_rate": 1.4924613471332801e-05, + "loss": 0.4496, + "step": 11289 + }, + { + "epoch": 0.3560952531146507, + "grad_norm": 1.4163219928741455, + "learning_rate": 1.4923724349974991e-05, + "loss": 0.5353, + "step": 11290 + }, + { + "epoch": 0.3561267938810913, + "grad_norm": 1.5019817352294922, + "learning_rate": 1.4922835177234297e-05, + "loss": 0.5178, + "step": 11291 + }, + { + "epoch": 0.35615833464753194, + "grad_norm": 3.6738080978393555, + "learning_rate": 1.492194595311999e-05, + "loss": 0.5065, + "step": 11292 + }, + { + "epoch": 0.35618987541397257, + "grad_norm": 1.9256829023361206, + "learning_rate": 1.4921056677641362e-05, + "loss": 0.5447, + "step": 11293 + }, + { + "epoch": 0.3562214161804132, + "grad_norm": 1.482672929763794, + "learning_rate": 1.4920167350807682e-05, + "loss": 0.508, + "step": 11294 + }, + { + "epoch": 0.3562529569468538, + "grad_norm": 1.455244541168213, + "learning_rate": 1.4919277972628235e-05, + "loss": 0.5118, + "step": 11295 + }, + { + "epoch": 0.35628449771329446, + "grad_norm": 1.5138299465179443, + "learning_rate": 1.4918388543112303e-05, + "loss": 0.5199, + "step": 11296 + }, + { + "epoch": 0.35631603847973503, + "grad_norm": 1.7415043115615845, + "learning_rate": 1.491749906226917e-05, + "loss": 0.5338, + "step": 11297 + }, + { + "epoch": 0.35634757924617566, + "grad_norm": 1.2495027780532837, + "learning_rate": 1.4916609530108115e-05, + "loss": 0.5231, + "step": 11298 + }, + { + "epoch": 0.3563791200126163, + "grad_norm": 1.7600947618484497, + "learning_rate": 1.491571994663842e-05, + "loss": 0.4039, + "step": 11299 + }, + { + "epoch": 0.3564106607790569, + "grad_norm": 1.7325412034988403, + "learning_rate": 1.4914830311869375e-05, + "loss": 0.518, + "step": 11300 + }, + { + "epoch": 0.35644220154549755, + "grad_norm": 1.3763360977172852, + "learning_rate": 1.4913940625810255e-05, + "loss": 0.5127, + "step": 11301 + }, + { + "epoch": 0.3564737423119382, + "grad_norm": 1.9212366342544556, + "learning_rate": 1.4913050888470349e-05, + "loss": 0.4905, + "step": 11302 + }, + { + "epoch": 0.3565052830783788, + "grad_norm": 1.4930540323257446, + "learning_rate": 1.4912161099858945e-05, + "loss": 0.486, + "step": 11303 + }, + { + "epoch": 0.35653682384481944, + "grad_norm": 2.044088840484619, + "learning_rate": 1.4911271259985323e-05, + "loss": 0.551, + "step": 11304 + }, + { + "epoch": 0.35656836461126007, + "grad_norm": 1.85774564743042, + "learning_rate": 1.4910381368858773e-05, + "loss": 0.4516, + "step": 11305 + }, + { + "epoch": 0.3565999053777007, + "grad_norm": 1.4290813207626343, + "learning_rate": 1.4909491426488579e-05, + "loss": 0.5263, + "step": 11306 + }, + { + "epoch": 0.35663144614414133, + "grad_norm": 1.6761707067489624, + "learning_rate": 1.4908601432884034e-05, + "loss": 0.5334, + "step": 11307 + }, + { + "epoch": 0.3566629869105819, + "grad_norm": 1.46330726146698, + "learning_rate": 1.4907711388054419e-05, + "loss": 0.5112, + "step": 11308 + }, + { + "epoch": 0.35669452767702253, + "grad_norm": 1.788717269897461, + "learning_rate": 1.4906821292009023e-05, + "loss": 0.5013, + "step": 11309 + }, + { + "epoch": 0.35672606844346316, + "grad_norm": 1.8814650774002075, + "learning_rate": 1.490593114475714e-05, + "loss": 0.5213, + "step": 11310 + }, + { + "epoch": 0.3567576092099038, + "grad_norm": 1.6242049932479858, + "learning_rate": 1.4905040946308056e-05, + "loss": 0.5109, + "step": 11311 + }, + { + "epoch": 0.3567891499763444, + "grad_norm": 1.1998872756958008, + "learning_rate": 1.4904150696671059e-05, + "loss": 0.511, + "step": 11312 + }, + { + "epoch": 0.35682069074278505, + "grad_norm": 1.561694860458374, + "learning_rate": 1.4903260395855441e-05, + "loss": 0.4996, + "step": 11313 + }, + { + "epoch": 0.3568522315092257, + "grad_norm": 1.3782143592834473, + "learning_rate": 1.4902370043870495e-05, + "loss": 0.5266, + "step": 11314 + }, + { + "epoch": 0.3568837722756663, + "grad_norm": 2.727936029434204, + "learning_rate": 1.4901479640725512e-05, + "loss": 0.4162, + "step": 11315 + }, + { + "epoch": 0.35691531304210694, + "grad_norm": 1.8945733308792114, + "learning_rate": 1.490058918642978e-05, + "loss": 0.5496, + "step": 11316 + }, + { + "epoch": 0.35694685380854757, + "grad_norm": 1.9408310651779175, + "learning_rate": 1.4899698680992597e-05, + "loss": 0.5078, + "step": 11317 + }, + { + "epoch": 0.35697839457498814, + "grad_norm": 1.510504126548767, + "learning_rate": 1.4898808124423257e-05, + "loss": 0.8475, + "step": 11318 + }, + { + "epoch": 0.3570099353414288, + "grad_norm": 1.2595046758651733, + "learning_rate": 1.4897917516731043e-05, + "loss": 0.466, + "step": 11319 + }, + { + "epoch": 0.3570414761078694, + "grad_norm": 1.9170011281967163, + "learning_rate": 1.489702685792526e-05, + "loss": 0.5049, + "step": 11320 + }, + { + "epoch": 0.35707301687431003, + "grad_norm": 1.5658551454544067, + "learning_rate": 1.4896136148015204e-05, + "loss": 0.8479, + "step": 11321 + }, + { + "epoch": 0.35710455764075066, + "grad_norm": 1.5428580045700073, + "learning_rate": 1.4895245387010157e-05, + "loss": 0.5234, + "step": 11322 + }, + { + "epoch": 0.3571360984071913, + "grad_norm": 1.3609477281570435, + "learning_rate": 1.489435457491943e-05, + "loss": 0.5152, + "step": 11323 + }, + { + "epoch": 0.3571676391736319, + "grad_norm": 1.630083441734314, + "learning_rate": 1.489346371175231e-05, + "loss": 0.841, + "step": 11324 + }, + { + "epoch": 0.35719917994007255, + "grad_norm": 3.096264362335205, + "learning_rate": 1.4892572797518097e-05, + "loss": 0.5201, + "step": 11325 + }, + { + "epoch": 0.3572307207065132, + "grad_norm": 1.589161992073059, + "learning_rate": 1.4891681832226086e-05, + "loss": 0.5166, + "step": 11326 + }, + { + "epoch": 0.3572622614729538, + "grad_norm": 2.008333206176758, + "learning_rate": 1.4890790815885578e-05, + "loss": 0.4844, + "step": 11327 + }, + { + "epoch": 0.35729380223939444, + "grad_norm": 1.3294157981872559, + "learning_rate": 1.4889899748505872e-05, + "loss": 0.5215, + "step": 11328 + }, + { + "epoch": 0.357325343005835, + "grad_norm": 1.420180320739746, + "learning_rate": 1.4889008630096262e-05, + "loss": 0.5495, + "step": 11329 + }, + { + "epoch": 0.35735688377227565, + "grad_norm": 1.5836026668548584, + "learning_rate": 1.4888117460666053e-05, + "loss": 0.4956, + "step": 11330 + }, + { + "epoch": 0.3573884245387163, + "grad_norm": 1.7581919431686401, + "learning_rate": 1.4887226240224544e-05, + "loss": 0.5566, + "step": 11331 + }, + { + "epoch": 0.3574199653051569, + "grad_norm": 1.668523907661438, + "learning_rate": 1.4886334968781031e-05, + "loss": 0.5059, + "step": 11332 + }, + { + "epoch": 0.35745150607159754, + "grad_norm": 1.4406205415725708, + "learning_rate": 1.4885443646344819e-05, + "loss": 0.5348, + "step": 11333 + }, + { + "epoch": 0.35748304683803817, + "grad_norm": 2.016190528869629, + "learning_rate": 1.4884552272925212e-05, + "loss": 0.5116, + "step": 11334 + }, + { + "epoch": 0.3575145876044788, + "grad_norm": 1.5059665441513062, + "learning_rate": 1.4883660848531508e-05, + "loss": 0.4939, + "step": 11335 + }, + { + "epoch": 0.3575461283709194, + "grad_norm": 1.5431890487670898, + "learning_rate": 1.4882769373173007e-05, + "loss": 0.4917, + "step": 11336 + }, + { + "epoch": 0.35757766913736005, + "grad_norm": 2.0567917823791504, + "learning_rate": 1.488187784685902e-05, + "loss": 0.5311, + "step": 11337 + }, + { + "epoch": 0.3576092099038007, + "grad_norm": 1.5380661487579346, + "learning_rate": 1.488098626959885e-05, + "loss": 0.5545, + "step": 11338 + }, + { + "epoch": 0.3576407506702413, + "grad_norm": 2.4045822620391846, + "learning_rate": 1.4880094641401795e-05, + "loss": 0.4801, + "step": 11339 + }, + { + "epoch": 0.3576722914366819, + "grad_norm": 2.3965582847595215, + "learning_rate": 1.4879202962277162e-05, + "loss": 0.5314, + "step": 11340 + }, + { + "epoch": 0.3577038322031225, + "grad_norm": 2.658017873764038, + "learning_rate": 1.487831123223426e-05, + "loss": 0.5459, + "step": 11341 + }, + { + "epoch": 0.35773537296956315, + "grad_norm": 1.594659447669983, + "learning_rate": 1.487741945128239e-05, + "loss": 0.5064, + "step": 11342 + }, + { + "epoch": 0.3577669137360038, + "grad_norm": 1.8501577377319336, + "learning_rate": 1.487652761943086e-05, + "loss": 0.502, + "step": 11343 + }, + { + "epoch": 0.3577984545024444, + "grad_norm": 1.6204662322998047, + "learning_rate": 1.4875635736688979e-05, + "loss": 0.5153, + "step": 11344 + }, + { + "epoch": 0.35782999526888504, + "grad_norm": 1.5656254291534424, + "learning_rate": 1.4874743803066055e-05, + "loss": 0.5401, + "step": 11345 + }, + { + "epoch": 0.35786153603532567, + "grad_norm": 2.1544854640960693, + "learning_rate": 1.4873851818571393e-05, + "loss": 0.5002, + "step": 11346 + }, + { + "epoch": 0.3578930768017663, + "grad_norm": 2.3603382110595703, + "learning_rate": 1.4872959783214303e-05, + "loss": 0.8315, + "step": 11347 + }, + { + "epoch": 0.3579246175682069, + "grad_norm": 1.5282948017120361, + "learning_rate": 1.4872067697004096e-05, + "loss": 0.8445, + "step": 11348 + }, + { + "epoch": 0.35795615833464756, + "grad_norm": 1.6466258764266968, + "learning_rate": 1.4871175559950075e-05, + "loss": 0.5336, + "step": 11349 + }, + { + "epoch": 0.35798769910108813, + "grad_norm": 1.732338547706604, + "learning_rate": 1.4870283372061556e-05, + "loss": 0.5118, + "step": 11350 + }, + { + "epoch": 0.35801923986752876, + "grad_norm": 1.4716540575027466, + "learning_rate": 1.486939113334785e-05, + "loss": 0.5112, + "step": 11351 + }, + { + "epoch": 0.3580507806339694, + "grad_norm": 1.4258272647857666, + "learning_rate": 1.4868498843818267e-05, + "loss": 0.4859, + "step": 11352 + }, + { + "epoch": 0.35808232140041, + "grad_norm": 1.3200182914733887, + "learning_rate": 1.4867606503482114e-05, + "loss": 0.54, + "step": 11353 + }, + { + "epoch": 0.35811386216685065, + "grad_norm": 2.1947474479675293, + "learning_rate": 1.486671411234871e-05, + "loss": 0.5232, + "step": 11354 + }, + { + "epoch": 0.3581454029332913, + "grad_norm": 1.6462697982788086, + "learning_rate": 1.4865821670427369e-05, + "loss": 0.4779, + "step": 11355 + }, + { + "epoch": 0.3581769436997319, + "grad_norm": 1.5975974798202515, + "learning_rate": 1.4864929177727396e-05, + "loss": 0.5333, + "step": 11356 + }, + { + "epoch": 0.35820848446617254, + "grad_norm": 1.7902803421020508, + "learning_rate": 1.4864036634258112e-05, + "loss": 0.5216, + "step": 11357 + }, + { + "epoch": 0.35824002523261317, + "grad_norm": 1.2054824829101562, + "learning_rate": 1.486314404002883e-05, + "loss": 0.5503, + "step": 11358 + }, + { + "epoch": 0.3582715659990538, + "grad_norm": 1.1383559703826904, + "learning_rate": 1.4862251395048861e-05, + "loss": 0.5106, + "step": 11359 + }, + { + "epoch": 0.35830310676549443, + "grad_norm": 1.4743828773498535, + "learning_rate": 1.4861358699327521e-05, + "loss": 0.5317, + "step": 11360 + }, + { + "epoch": 0.358334647531935, + "grad_norm": 1.7962400913238525, + "learning_rate": 1.4860465952874133e-05, + "loss": 0.5087, + "step": 11361 + }, + { + "epoch": 0.35836618829837563, + "grad_norm": 1.343260407447815, + "learning_rate": 1.4859573155698005e-05, + "loss": 0.52, + "step": 11362 + }, + { + "epoch": 0.35839772906481626, + "grad_norm": 1.5865885019302368, + "learning_rate": 1.4858680307808462e-05, + "loss": 0.543, + "step": 11363 + }, + { + "epoch": 0.3584292698312569, + "grad_norm": 1.791762113571167, + "learning_rate": 1.4857787409214815e-05, + "loss": 0.5155, + "step": 11364 + }, + { + "epoch": 0.3584608105976975, + "grad_norm": 1.6257110834121704, + "learning_rate": 1.4856894459926383e-05, + "loss": 0.514, + "step": 11365 + }, + { + "epoch": 0.35849235136413815, + "grad_norm": 2.018980026245117, + "learning_rate": 1.4856001459952488e-05, + "loss": 0.5406, + "step": 11366 + }, + { + "epoch": 0.3585238921305788, + "grad_norm": 1.4005407094955444, + "learning_rate": 1.4855108409302447e-05, + "loss": 0.5302, + "step": 11367 + }, + { + "epoch": 0.3585554328970194, + "grad_norm": 2.062361001968384, + "learning_rate": 1.485421530798558e-05, + "loss": 0.5207, + "step": 11368 + }, + { + "epoch": 0.35858697366346004, + "grad_norm": 1.49112868309021, + "learning_rate": 1.4853322156011208e-05, + "loss": 0.5172, + "step": 11369 + }, + { + "epoch": 0.35861851442990067, + "grad_norm": 1.5378899574279785, + "learning_rate": 1.4852428953388649e-05, + "loss": 0.4838, + "step": 11370 + }, + { + "epoch": 0.35865005519634124, + "grad_norm": 1.8776708841323853, + "learning_rate": 1.4851535700127225e-05, + "loss": 0.4777, + "step": 11371 + }, + { + "epoch": 0.3586815959627819, + "grad_norm": 4.064377307891846, + "learning_rate": 1.4850642396236261e-05, + "loss": 0.4831, + "step": 11372 + }, + { + "epoch": 0.3587131367292225, + "grad_norm": 2.2817018032073975, + "learning_rate": 1.4849749041725077e-05, + "loss": 0.4973, + "step": 11373 + }, + { + "epoch": 0.35874467749566313, + "grad_norm": 1.9654923677444458, + "learning_rate": 1.4848855636602995e-05, + "loss": 0.5414, + "step": 11374 + }, + { + "epoch": 0.35877621826210376, + "grad_norm": 1.4754973649978638, + "learning_rate": 1.4847962180879341e-05, + "loss": 0.4977, + "step": 11375 + }, + { + "epoch": 0.3588077590285444, + "grad_norm": 1.7153326272964478, + "learning_rate": 1.4847068674563434e-05, + "loss": 0.5251, + "step": 11376 + }, + { + "epoch": 0.358839299794985, + "grad_norm": 1.5382344722747803, + "learning_rate": 1.4846175117664605e-05, + "loss": 0.5313, + "step": 11377 + }, + { + "epoch": 0.35887084056142565, + "grad_norm": 1.8929523229599, + "learning_rate": 1.4845281510192174e-05, + "loss": 0.4964, + "step": 11378 + }, + { + "epoch": 0.3589023813278663, + "grad_norm": 1.7547202110290527, + "learning_rate": 1.4844387852155471e-05, + "loss": 0.506, + "step": 11379 + }, + { + "epoch": 0.3589339220943069, + "grad_norm": 1.414832592010498, + "learning_rate": 1.4843494143563816e-05, + "loss": 0.489, + "step": 11380 + }, + { + "epoch": 0.35896546286074754, + "grad_norm": 1.482171893119812, + "learning_rate": 1.4842600384426538e-05, + "loss": 0.5267, + "step": 11381 + }, + { + "epoch": 0.3589970036271881, + "grad_norm": 1.6889808177947998, + "learning_rate": 1.4841706574752968e-05, + "loss": 0.512, + "step": 11382 + }, + { + "epoch": 0.35902854439362875, + "grad_norm": 1.5883259773254395, + "learning_rate": 1.4840812714552429e-05, + "loss": 0.5085, + "step": 11383 + }, + { + "epoch": 0.3590600851600694, + "grad_norm": 1.933937668800354, + "learning_rate": 1.4839918803834251e-05, + "loss": 0.5445, + "step": 11384 + }, + { + "epoch": 0.35909162592651, + "grad_norm": 1.6257259845733643, + "learning_rate": 1.4839024842607761e-05, + "loss": 0.536, + "step": 11385 + }, + { + "epoch": 0.35912316669295063, + "grad_norm": 1.5734121799468994, + "learning_rate": 1.4838130830882289e-05, + "loss": 0.5107, + "step": 11386 + }, + { + "epoch": 0.35915470745939126, + "grad_norm": 1.5888395309448242, + "learning_rate": 1.4837236768667168e-05, + "loss": 0.5529, + "step": 11387 + }, + { + "epoch": 0.3591862482258319, + "grad_norm": 1.9710010290145874, + "learning_rate": 1.4836342655971722e-05, + "loss": 0.5163, + "step": 11388 + }, + { + "epoch": 0.3592177889922725, + "grad_norm": 1.6978124380111694, + "learning_rate": 1.4835448492805287e-05, + "loss": 0.5309, + "step": 11389 + }, + { + "epoch": 0.35924932975871315, + "grad_norm": 2.449537992477417, + "learning_rate": 1.4834554279177192e-05, + "loss": 0.4958, + "step": 11390 + }, + { + "epoch": 0.3592808705251538, + "grad_norm": 2.2599666118621826, + "learning_rate": 1.4833660015096767e-05, + "loss": 0.512, + "step": 11391 + }, + { + "epoch": 0.3593124112915944, + "grad_norm": 1.6717396974563599, + "learning_rate": 1.4832765700573348e-05, + "loss": 0.5057, + "step": 11392 + }, + { + "epoch": 0.359343952058035, + "grad_norm": 3.070244073867798, + "learning_rate": 1.4831871335616268e-05, + "loss": 0.5379, + "step": 11393 + }, + { + "epoch": 0.3593754928244756, + "grad_norm": 1.5149707794189453, + "learning_rate": 1.4830976920234856e-05, + "loss": 0.5363, + "step": 11394 + }, + { + "epoch": 0.35940703359091625, + "grad_norm": 1.6523505449295044, + "learning_rate": 1.4830082454438453e-05, + "loss": 0.5369, + "step": 11395 + }, + { + "epoch": 0.3594385743573569, + "grad_norm": 1.4578098058700562, + "learning_rate": 1.4829187938236384e-05, + "loss": 0.5075, + "step": 11396 + }, + { + "epoch": 0.3594701151237975, + "grad_norm": 2.210150957107544, + "learning_rate": 1.4828293371637993e-05, + "loss": 0.5184, + "step": 11397 + }, + { + "epoch": 0.35950165589023814, + "grad_norm": 1.6947466135025024, + "learning_rate": 1.482739875465261e-05, + "loss": 0.4951, + "step": 11398 + }, + { + "epoch": 0.35953319665667877, + "grad_norm": 1.5024769306182861, + "learning_rate": 1.4826504087289573e-05, + "loss": 0.524, + "step": 11399 + }, + { + "epoch": 0.3595647374231194, + "grad_norm": 1.6147834062576294, + "learning_rate": 1.4825609369558219e-05, + "loss": 0.4926, + "step": 11400 + }, + { + "epoch": 0.35959627818956, + "grad_norm": 1.322823405265808, + "learning_rate": 1.4824714601467883e-05, + "loss": 0.3893, + "step": 11401 + }, + { + "epoch": 0.35962781895600066, + "grad_norm": 1.4911251068115234, + "learning_rate": 1.4823819783027901e-05, + "loss": 0.541, + "step": 11402 + }, + { + "epoch": 0.35965935972244123, + "grad_norm": 2.1171364784240723, + "learning_rate": 1.4822924914247618e-05, + "loss": 0.5187, + "step": 11403 + }, + { + "epoch": 0.35969090048888186, + "grad_norm": 1.9799286127090454, + "learning_rate": 1.4822029995136367e-05, + "loss": 0.4986, + "step": 11404 + }, + { + "epoch": 0.3597224412553225, + "grad_norm": 1.8384426832199097, + "learning_rate": 1.4821135025703491e-05, + "loss": 0.5197, + "step": 11405 + }, + { + "epoch": 0.3597539820217631, + "grad_norm": 1.4813563823699951, + "learning_rate": 1.4820240005958325e-05, + "loss": 0.4924, + "step": 11406 + }, + { + "epoch": 0.35978552278820375, + "grad_norm": 1.8770731687545776, + "learning_rate": 1.4819344935910213e-05, + "loss": 0.5076, + "step": 11407 + }, + { + "epoch": 0.3598170635546444, + "grad_norm": 3.174705982208252, + "learning_rate": 1.4818449815568495e-05, + "loss": 0.5342, + "step": 11408 + }, + { + "epoch": 0.359848604321085, + "grad_norm": 1.3867689371109009, + "learning_rate": 1.4817554644942508e-05, + "loss": 0.501, + "step": 11409 + }, + { + "epoch": 0.35988014508752564, + "grad_norm": 1.7806705236434937, + "learning_rate": 1.4816659424041601e-05, + "loss": 0.8606, + "step": 11410 + }, + { + "epoch": 0.35991168585396627, + "grad_norm": 1.6296677589416504, + "learning_rate": 1.4815764152875111e-05, + "loss": 0.8247, + "step": 11411 + }, + { + "epoch": 0.3599432266204069, + "grad_norm": 1.5811200141906738, + "learning_rate": 1.4814868831452385e-05, + "loss": 0.4145, + "step": 11412 + }, + { + "epoch": 0.3599747673868475, + "grad_norm": 2.036412000656128, + "learning_rate": 1.4813973459782764e-05, + "loss": 0.4855, + "step": 11413 + }, + { + "epoch": 0.3600063081532881, + "grad_norm": 2.3294308185577393, + "learning_rate": 1.481307803787559e-05, + "loss": 0.552, + "step": 11414 + }, + { + "epoch": 0.36003784891972873, + "grad_norm": 4.100742816925049, + "learning_rate": 1.4812182565740212e-05, + "loss": 0.5185, + "step": 11415 + }, + { + "epoch": 0.36006938968616936, + "grad_norm": 1.7294827699661255, + "learning_rate": 1.4811287043385969e-05, + "loss": 0.8585, + "step": 11416 + }, + { + "epoch": 0.36010093045261, + "grad_norm": 3.9165401458740234, + "learning_rate": 1.4810391470822213e-05, + "loss": 0.5375, + "step": 11417 + }, + { + "epoch": 0.3601324712190506, + "grad_norm": 2.0369555950164795, + "learning_rate": 1.4809495848058286e-05, + "loss": 0.534, + "step": 11418 + }, + { + "epoch": 0.36016401198549125, + "grad_norm": 2.074563503265381, + "learning_rate": 1.4808600175103534e-05, + "loss": 0.5354, + "step": 11419 + }, + { + "epoch": 0.3601955527519319, + "grad_norm": 1.3173723220825195, + "learning_rate": 1.4807704451967306e-05, + "loss": 0.3945, + "step": 11420 + }, + { + "epoch": 0.3602270935183725, + "grad_norm": 1.4411653280258179, + "learning_rate": 1.4806808678658949e-05, + "loss": 0.4931, + "step": 11421 + }, + { + "epoch": 0.36025863428481314, + "grad_norm": 2.0273313522338867, + "learning_rate": 1.4805912855187809e-05, + "loss": 0.5119, + "step": 11422 + }, + { + "epoch": 0.36029017505125377, + "grad_norm": 1.8937363624572754, + "learning_rate": 1.4805016981563238e-05, + "loss": 0.531, + "step": 11423 + }, + { + "epoch": 0.36032171581769434, + "grad_norm": 2.1121721267700195, + "learning_rate": 1.4804121057794586e-05, + "loss": 0.4966, + "step": 11424 + }, + { + "epoch": 0.360353256584135, + "grad_norm": 2.5857512950897217, + "learning_rate": 1.4803225083891201e-05, + "loss": 0.5286, + "step": 11425 + }, + { + "epoch": 0.3603847973505756, + "grad_norm": 3.0720129013061523, + "learning_rate": 1.480232905986243e-05, + "loss": 0.8594, + "step": 11426 + }, + { + "epoch": 0.36041633811701623, + "grad_norm": 1.9908868074417114, + "learning_rate": 1.4801432985717627e-05, + "loss": 0.5094, + "step": 11427 + }, + { + "epoch": 0.36044787888345686, + "grad_norm": 2.2484099864959717, + "learning_rate": 1.4800536861466141e-05, + "loss": 0.4997, + "step": 11428 + }, + { + "epoch": 0.3604794196498975, + "grad_norm": 1.7968155145645142, + "learning_rate": 1.4799640687117327e-05, + "loss": 0.4841, + "step": 11429 + }, + { + "epoch": 0.3605109604163381, + "grad_norm": 2.2112255096435547, + "learning_rate": 1.479874446268054e-05, + "loss": 0.5365, + "step": 11430 + }, + { + "epoch": 0.36054250118277875, + "grad_norm": 1.4122942686080933, + "learning_rate": 1.479784818816512e-05, + "loss": 0.4301, + "step": 11431 + }, + { + "epoch": 0.3605740419492194, + "grad_norm": 1.7262495756149292, + "learning_rate": 1.4796951863580434e-05, + "loss": 0.5266, + "step": 11432 + }, + { + "epoch": 0.36060558271566, + "grad_norm": 1.6334054470062256, + "learning_rate": 1.4796055488935827e-05, + "loss": 0.8524, + "step": 11433 + }, + { + "epoch": 0.36063712348210064, + "grad_norm": 2.337653636932373, + "learning_rate": 1.479515906424066e-05, + "loss": 0.5318, + "step": 11434 + }, + { + "epoch": 0.3606686642485412, + "grad_norm": 2.7841854095458984, + "learning_rate": 1.4794262589504285e-05, + "loss": 0.4843, + "step": 11435 + }, + { + "epoch": 0.36070020501498185, + "grad_norm": 1.673039197921753, + "learning_rate": 1.4793366064736057e-05, + "loss": 0.5089, + "step": 11436 + }, + { + "epoch": 0.3607317457814225, + "grad_norm": 1.682087779045105, + "learning_rate": 1.4792469489945332e-05, + "loss": 0.4984, + "step": 11437 + }, + { + "epoch": 0.3607632865478631, + "grad_norm": 2.1680638790130615, + "learning_rate": 1.4791572865141466e-05, + "loss": 0.8471, + "step": 11438 + }, + { + "epoch": 0.36079482731430373, + "grad_norm": 1.7825002670288086, + "learning_rate": 1.4790676190333814e-05, + "loss": 0.5246, + "step": 11439 + }, + { + "epoch": 0.36082636808074436, + "grad_norm": 2.1567182540893555, + "learning_rate": 1.478977946553174e-05, + "loss": 0.5197, + "step": 11440 + }, + { + "epoch": 0.360857908847185, + "grad_norm": 1.7848643064498901, + "learning_rate": 1.4788882690744597e-05, + "loss": 0.531, + "step": 11441 + }, + { + "epoch": 0.3608894496136256, + "grad_norm": 1.75265371799469, + "learning_rate": 1.4787985865981744e-05, + "loss": 0.3655, + "step": 11442 + }, + { + "epoch": 0.36092099038006625, + "grad_norm": 2.172529697418213, + "learning_rate": 1.4787088991252541e-05, + "loss": 0.5486, + "step": 11443 + }, + { + "epoch": 0.3609525311465069, + "grad_norm": 1.7724570035934448, + "learning_rate": 1.4786192066566348e-05, + "loss": 0.5279, + "step": 11444 + }, + { + "epoch": 0.36098407191294746, + "grad_norm": 2.1384551525115967, + "learning_rate": 1.4785295091932526e-05, + "loss": 0.5343, + "step": 11445 + }, + { + "epoch": 0.3610156126793881, + "grad_norm": 5.274745941162109, + "learning_rate": 1.478439806736043e-05, + "loss": 0.5475, + "step": 11446 + }, + { + "epoch": 0.3610471534458287, + "grad_norm": 1.6855560541152954, + "learning_rate": 1.4783500992859425e-05, + "loss": 0.4766, + "step": 11447 + }, + { + "epoch": 0.36107869421226935, + "grad_norm": 1.9610512256622314, + "learning_rate": 1.4782603868438879e-05, + "loss": 0.5252, + "step": 11448 + }, + { + "epoch": 0.36111023497871, + "grad_norm": 1.913326621055603, + "learning_rate": 1.4781706694108143e-05, + "loss": 0.5635, + "step": 11449 + }, + { + "epoch": 0.3611417757451506, + "grad_norm": 1.444624900817871, + "learning_rate": 1.4780809469876585e-05, + "loss": 0.5048, + "step": 11450 + }, + { + "epoch": 0.36117331651159124, + "grad_norm": 2.664754629135132, + "learning_rate": 1.477991219575357e-05, + "loss": 0.5124, + "step": 11451 + }, + { + "epoch": 0.36120485727803187, + "grad_norm": 1.847169041633606, + "learning_rate": 1.4779014871748456e-05, + "loss": 0.529, + "step": 11452 + }, + { + "epoch": 0.3612363980444725, + "grad_norm": 1.8968195915222168, + "learning_rate": 1.4778117497870613e-05, + "loss": 0.4851, + "step": 11453 + }, + { + "epoch": 0.3612679388109131, + "grad_norm": 2.5868475437164307, + "learning_rate": 1.4777220074129406e-05, + "loss": 0.5139, + "step": 11454 + }, + { + "epoch": 0.36129947957735375, + "grad_norm": 1.8389804363250732, + "learning_rate": 1.4776322600534196e-05, + "loss": 0.4879, + "step": 11455 + }, + { + "epoch": 0.36133102034379433, + "grad_norm": 1.845396876335144, + "learning_rate": 1.4775425077094351e-05, + "loss": 0.5086, + "step": 11456 + }, + { + "epoch": 0.36136256111023496, + "grad_norm": 2.1026458740234375, + "learning_rate": 1.4774527503819238e-05, + "loss": 0.5054, + "step": 11457 + }, + { + "epoch": 0.3613941018766756, + "grad_norm": 3.620626926422119, + "learning_rate": 1.4773629880718224e-05, + "loss": 0.5127, + "step": 11458 + }, + { + "epoch": 0.3614256426431162, + "grad_norm": 1.714141845703125, + "learning_rate": 1.4772732207800673e-05, + "loss": 0.5401, + "step": 11459 + }, + { + "epoch": 0.36145718340955685, + "grad_norm": 1.7446223497390747, + "learning_rate": 1.4771834485075956e-05, + "loss": 0.8398, + "step": 11460 + }, + { + "epoch": 0.3614887241759975, + "grad_norm": 2.075052261352539, + "learning_rate": 1.4770936712553443e-05, + "loss": 0.5102, + "step": 11461 + }, + { + "epoch": 0.3615202649424381, + "grad_norm": 1.9486520290374756, + "learning_rate": 1.4770038890242498e-05, + "loss": 0.5182, + "step": 11462 + }, + { + "epoch": 0.36155180570887874, + "grad_norm": 2.08510422706604, + "learning_rate": 1.4769141018152495e-05, + "loss": 0.5364, + "step": 11463 + }, + { + "epoch": 0.36158334647531937, + "grad_norm": 1.822143793106079, + "learning_rate": 1.4768243096292802e-05, + "loss": 0.5374, + "step": 11464 + }, + { + "epoch": 0.36161488724176, + "grad_norm": 1.5012218952178955, + "learning_rate": 1.4767345124672792e-05, + "loss": 0.5242, + "step": 11465 + }, + { + "epoch": 0.3616464280082006, + "grad_norm": 1.9719998836517334, + "learning_rate": 1.4766447103301831e-05, + "loss": 0.5247, + "step": 11466 + }, + { + "epoch": 0.3616779687746412, + "grad_norm": 2.0429747104644775, + "learning_rate": 1.4765549032189292e-05, + "loss": 0.4932, + "step": 11467 + }, + { + "epoch": 0.36170950954108183, + "grad_norm": 1.916969895362854, + "learning_rate": 1.4764650911344553e-05, + "loss": 0.5226, + "step": 11468 + }, + { + "epoch": 0.36174105030752246, + "grad_norm": 1.9815990924835205, + "learning_rate": 1.476375274077698e-05, + "loss": 0.524, + "step": 11469 + }, + { + "epoch": 0.3617725910739631, + "grad_norm": 2.1318066120147705, + "learning_rate": 1.4762854520495945e-05, + "loss": 0.5216, + "step": 11470 + }, + { + "epoch": 0.3618041318404037, + "grad_norm": 1.9458250999450684, + "learning_rate": 1.476195625051083e-05, + "loss": 0.5042, + "step": 11471 + }, + { + "epoch": 0.36183567260684435, + "grad_norm": 1.6027607917785645, + "learning_rate": 1.4761057930831002e-05, + "loss": 0.5005, + "step": 11472 + }, + { + "epoch": 0.361867213373285, + "grad_norm": 2.4101574420928955, + "learning_rate": 1.4760159561465836e-05, + "loss": 0.552, + "step": 11473 + }, + { + "epoch": 0.3618987541397256, + "grad_norm": 1.7469671964645386, + "learning_rate": 1.4759261142424712e-05, + "loss": 0.5302, + "step": 11474 + }, + { + "epoch": 0.36193029490616624, + "grad_norm": 1.8023855686187744, + "learning_rate": 1.4758362673717005e-05, + "loss": 0.5127, + "step": 11475 + }, + { + "epoch": 0.36196183567260687, + "grad_norm": 3.3649139404296875, + "learning_rate": 1.4757464155352082e-05, + "loss": 0.8186, + "step": 11476 + }, + { + "epoch": 0.36199337643904744, + "grad_norm": 1.5693955421447754, + "learning_rate": 1.4756565587339328e-05, + "loss": 0.4963, + "step": 11477 + }, + { + "epoch": 0.3620249172054881, + "grad_norm": 3.2356772422790527, + "learning_rate": 1.4755666969688123e-05, + "loss": 0.4887, + "step": 11478 + }, + { + "epoch": 0.3620564579719287, + "grad_norm": 1.6195093393325806, + "learning_rate": 1.4754768302407834e-05, + "loss": 0.4614, + "step": 11479 + }, + { + "epoch": 0.36208799873836933, + "grad_norm": 1.8218841552734375, + "learning_rate": 1.475386958550785e-05, + "loss": 0.4376, + "step": 11480 + }, + { + "epoch": 0.36211953950480996, + "grad_norm": 1.8043535947799683, + "learning_rate": 1.4752970818997543e-05, + "loss": 0.5054, + "step": 11481 + }, + { + "epoch": 0.3621510802712506, + "grad_norm": 1.5598814487457275, + "learning_rate": 1.4752072002886298e-05, + "loss": 0.492, + "step": 11482 + }, + { + "epoch": 0.3621826210376912, + "grad_norm": 1.659883737564087, + "learning_rate": 1.4751173137183489e-05, + "loss": 0.8541, + "step": 11483 + }, + { + "epoch": 0.36221416180413185, + "grad_norm": 1.981368899345398, + "learning_rate": 1.4750274221898503e-05, + "loss": 0.5235, + "step": 11484 + }, + { + "epoch": 0.3622457025705725, + "grad_norm": 1.7856727838516235, + "learning_rate": 1.4749375257040714e-05, + "loss": 0.8212, + "step": 11485 + }, + { + "epoch": 0.3622772433370131, + "grad_norm": 3.3981382846832275, + "learning_rate": 1.4748476242619507e-05, + "loss": 0.405, + "step": 11486 + }, + { + "epoch": 0.36230878410345374, + "grad_norm": 1.9490454196929932, + "learning_rate": 1.4747577178644261e-05, + "loss": 0.5266, + "step": 11487 + }, + { + "epoch": 0.3623403248698943, + "grad_norm": 1.859244465827942, + "learning_rate": 1.4746678065124362e-05, + "loss": 0.5081, + "step": 11488 + }, + { + "epoch": 0.36237186563633494, + "grad_norm": 2.444080352783203, + "learning_rate": 1.4745778902069194e-05, + "loss": 0.5442, + "step": 11489 + }, + { + "epoch": 0.3624034064027756, + "grad_norm": 2.0065486431121826, + "learning_rate": 1.4744879689488137e-05, + "loss": 0.5123, + "step": 11490 + }, + { + "epoch": 0.3624349471692162, + "grad_norm": 1.4216586351394653, + "learning_rate": 1.4743980427390575e-05, + "loss": 0.4931, + "step": 11491 + }, + { + "epoch": 0.36246648793565683, + "grad_norm": 2.117227792739868, + "learning_rate": 1.4743081115785895e-05, + "loss": 0.8534, + "step": 11492 + }, + { + "epoch": 0.36249802870209746, + "grad_norm": 1.860024333000183, + "learning_rate": 1.474218175468348e-05, + "loss": 0.5155, + "step": 11493 + }, + { + "epoch": 0.3625295694685381, + "grad_norm": 2.0611701011657715, + "learning_rate": 1.4741282344092718e-05, + "loss": 0.5345, + "step": 11494 + }, + { + "epoch": 0.3625611102349787, + "grad_norm": 1.6343131065368652, + "learning_rate": 1.474038288402299e-05, + "loss": 0.5114, + "step": 11495 + }, + { + "epoch": 0.36259265100141935, + "grad_norm": 3.011406898498535, + "learning_rate": 1.473948337448369e-05, + "loss": 0.55, + "step": 11496 + }, + { + "epoch": 0.36262419176786, + "grad_norm": 3.1093833446502686, + "learning_rate": 1.4738583815484197e-05, + "loss": 0.5058, + "step": 11497 + }, + { + "epoch": 0.36265573253430056, + "grad_norm": 2.1853585243225098, + "learning_rate": 1.4737684207033903e-05, + "loss": 0.4947, + "step": 11498 + }, + { + "epoch": 0.3626872733007412, + "grad_norm": 2.554616689682007, + "learning_rate": 1.4736784549142197e-05, + "loss": 0.4919, + "step": 11499 + }, + { + "epoch": 0.3627188140671818, + "grad_norm": 1.6989868879318237, + "learning_rate": 1.4735884841818467e-05, + "loss": 0.5216, + "step": 11500 + }, + { + "epoch": 0.36275035483362245, + "grad_norm": 1.5933499336242676, + "learning_rate": 1.4734985085072097e-05, + "loss": 0.4996, + "step": 11501 + }, + { + "epoch": 0.3627818956000631, + "grad_norm": 1.73661208152771, + "learning_rate": 1.4734085278912485e-05, + "loss": 0.8285, + "step": 11502 + }, + { + "epoch": 0.3628134363665037, + "grad_norm": 2.2831015586853027, + "learning_rate": 1.4733185423349014e-05, + "loss": 0.5151, + "step": 11503 + }, + { + "epoch": 0.36284497713294434, + "grad_norm": 2.5432016849517822, + "learning_rate": 1.4732285518391082e-05, + "loss": 0.5137, + "step": 11504 + }, + { + "epoch": 0.36287651789938496, + "grad_norm": 2.046700954437256, + "learning_rate": 1.4731385564048072e-05, + "loss": 0.5352, + "step": 11505 + }, + { + "epoch": 0.3629080586658256, + "grad_norm": 2.157160997390747, + "learning_rate": 1.4730485560329384e-05, + "loss": 0.4843, + "step": 11506 + }, + { + "epoch": 0.3629395994322662, + "grad_norm": 3.1627860069274902, + "learning_rate": 1.4729585507244404e-05, + "loss": 0.4832, + "step": 11507 + }, + { + "epoch": 0.36297114019870685, + "grad_norm": 1.8717403411865234, + "learning_rate": 1.4728685404802527e-05, + "loss": 0.4332, + "step": 11508 + }, + { + "epoch": 0.36300268096514743, + "grad_norm": 1.9067978858947754, + "learning_rate": 1.4727785253013145e-05, + "loss": 0.5633, + "step": 11509 + }, + { + "epoch": 0.36303422173158806, + "grad_norm": 3.397590160369873, + "learning_rate": 1.4726885051885654e-05, + "loss": 0.4917, + "step": 11510 + }, + { + "epoch": 0.3630657624980287, + "grad_norm": 1.8066800832748413, + "learning_rate": 1.4725984801429448e-05, + "loss": 0.5346, + "step": 11511 + }, + { + "epoch": 0.3630973032644693, + "grad_norm": 1.77766752243042, + "learning_rate": 1.4725084501653916e-05, + "loss": 0.5303, + "step": 11512 + }, + { + "epoch": 0.36312884403090995, + "grad_norm": 7.540788650512695, + "learning_rate": 1.4724184152568464e-05, + "loss": 0.5036, + "step": 11513 + }, + { + "epoch": 0.3631603847973506, + "grad_norm": 1.9036794900894165, + "learning_rate": 1.4723283754182482e-05, + "loss": 0.5085, + "step": 11514 + }, + { + "epoch": 0.3631919255637912, + "grad_norm": 1.5472607612609863, + "learning_rate": 1.4722383306505364e-05, + "loss": 0.8373, + "step": 11515 + }, + { + "epoch": 0.36322346633023184, + "grad_norm": 2.279578685760498, + "learning_rate": 1.472148280954651e-05, + "loss": 0.5345, + "step": 11516 + }, + { + "epoch": 0.36325500709667247, + "grad_norm": 1.9433619976043701, + "learning_rate": 1.4720582263315318e-05, + "loss": 0.5763, + "step": 11517 + }, + { + "epoch": 0.3632865478631131, + "grad_norm": 3.0034968852996826, + "learning_rate": 1.4719681667821181e-05, + "loss": 0.5316, + "step": 11518 + }, + { + "epoch": 0.36331808862955367, + "grad_norm": 2.1576082706451416, + "learning_rate": 1.4718781023073503e-05, + "loss": 0.4168, + "step": 11519 + }, + { + "epoch": 0.3633496293959943, + "grad_norm": 1.771942138671875, + "learning_rate": 1.4717880329081684e-05, + "loss": 0.505, + "step": 11520 + }, + { + "epoch": 0.36338117016243493, + "grad_norm": 1.5384036302566528, + "learning_rate": 1.4716979585855116e-05, + "loss": 0.5268, + "step": 11521 + }, + { + "epoch": 0.36341271092887556, + "grad_norm": 1.5020381212234497, + "learning_rate": 1.4716078793403203e-05, + "loss": 0.489, + "step": 11522 + }, + { + "epoch": 0.3634442516953162, + "grad_norm": 1.3155536651611328, + "learning_rate": 1.471517795173535e-05, + "loss": 0.8308, + "step": 11523 + }, + { + "epoch": 0.3634757924617568, + "grad_norm": 2.977905511856079, + "learning_rate": 1.4714277060860953e-05, + "loss": 0.8435, + "step": 11524 + }, + { + "epoch": 0.36350733322819745, + "grad_norm": 3.28285551071167, + "learning_rate": 1.4713376120789414e-05, + "loss": 0.5029, + "step": 11525 + }, + { + "epoch": 0.3635388739946381, + "grad_norm": 1.934801459312439, + "learning_rate": 1.4712475131530136e-05, + "loss": 0.3713, + "step": 11526 + }, + { + "epoch": 0.3635704147610787, + "grad_norm": 1.4781633615493774, + "learning_rate": 1.4711574093092518e-05, + "loss": 0.5109, + "step": 11527 + }, + { + "epoch": 0.36360195552751934, + "grad_norm": 2.2296905517578125, + "learning_rate": 1.4710673005485967e-05, + "loss": 0.524, + "step": 11528 + }, + { + "epoch": 0.36363349629395997, + "grad_norm": 3.5720136165618896, + "learning_rate": 1.4709771868719884e-05, + "loss": 0.4995, + "step": 11529 + }, + { + "epoch": 0.36366503706040054, + "grad_norm": 2.507251262664795, + "learning_rate": 1.4708870682803678e-05, + "loss": 0.5029, + "step": 11530 + }, + { + "epoch": 0.36369657782684117, + "grad_norm": 1.697668433189392, + "learning_rate": 1.4707969447746748e-05, + "loss": 0.4963, + "step": 11531 + }, + { + "epoch": 0.3637281185932818, + "grad_norm": 1.5654200315475464, + "learning_rate": 1.47070681635585e-05, + "loss": 0.8308, + "step": 11532 + }, + { + "epoch": 0.36375965935972243, + "grad_norm": 1.6280466318130493, + "learning_rate": 1.4706166830248346e-05, + "loss": 0.5166, + "step": 11533 + }, + { + "epoch": 0.36379120012616306, + "grad_norm": 1.9525465965270996, + "learning_rate": 1.4705265447825683e-05, + "loss": 0.5226, + "step": 11534 + }, + { + "epoch": 0.3638227408926037, + "grad_norm": 1.9577456712722778, + "learning_rate": 1.4704364016299922e-05, + "loss": 0.5019, + "step": 11535 + }, + { + "epoch": 0.3638542816590443, + "grad_norm": 1.5011699199676514, + "learning_rate": 1.4703462535680468e-05, + "loss": 0.5159, + "step": 11536 + }, + { + "epoch": 0.36388582242548495, + "grad_norm": 3.4498965740203857, + "learning_rate": 1.4702561005976733e-05, + "loss": 0.5341, + "step": 11537 + }, + { + "epoch": 0.3639173631919256, + "grad_norm": 1.4902070760726929, + "learning_rate": 1.4701659427198119e-05, + "loss": 0.5429, + "step": 11538 + }, + { + "epoch": 0.3639489039583662, + "grad_norm": 1.414708137512207, + "learning_rate": 1.470075779935404e-05, + "loss": 0.5005, + "step": 11539 + }, + { + "epoch": 0.36398044472480684, + "grad_norm": 1.8491334915161133, + "learning_rate": 1.4699856122453904e-05, + "loss": 0.5079, + "step": 11540 + }, + { + "epoch": 0.3640119854912474, + "grad_norm": 1.3916493654251099, + "learning_rate": 1.4698954396507119e-05, + "loss": 0.5435, + "step": 11541 + }, + { + "epoch": 0.36404352625768804, + "grad_norm": 1.722014307975769, + "learning_rate": 1.4698052621523098e-05, + "loss": 0.4827, + "step": 11542 + }, + { + "epoch": 0.3640750670241287, + "grad_norm": 1.8832811117172241, + "learning_rate": 1.4697150797511248e-05, + "loss": 0.5344, + "step": 11543 + }, + { + "epoch": 0.3641066077905693, + "grad_norm": 1.5901503562927246, + "learning_rate": 1.4696248924480984e-05, + "loss": 0.4981, + "step": 11544 + }, + { + "epoch": 0.36413814855700993, + "grad_norm": 1.3624671697616577, + "learning_rate": 1.4695347002441715e-05, + "loss": 0.5256, + "step": 11545 + }, + { + "epoch": 0.36416968932345056, + "grad_norm": 1.7912209033966064, + "learning_rate": 1.4694445031402853e-05, + "loss": 0.5229, + "step": 11546 + }, + { + "epoch": 0.3642012300898912, + "grad_norm": 2.2835330963134766, + "learning_rate": 1.4693543011373815e-05, + "loss": 0.5544, + "step": 11547 + }, + { + "epoch": 0.3642327708563318, + "grad_norm": 1.9264826774597168, + "learning_rate": 1.469264094236401e-05, + "loss": 0.5232, + "step": 11548 + }, + { + "epoch": 0.36426431162277245, + "grad_norm": 1.9662394523620605, + "learning_rate": 1.469173882438285e-05, + "loss": 0.8571, + "step": 11549 + }, + { + "epoch": 0.3642958523892131, + "grad_norm": 1.7866013050079346, + "learning_rate": 1.4690836657439755e-05, + "loss": 0.4236, + "step": 11550 + }, + { + "epoch": 0.36432739315565366, + "grad_norm": 2.053820848464966, + "learning_rate": 1.4689934441544136e-05, + "loss": 0.5116, + "step": 11551 + }, + { + "epoch": 0.3643589339220943, + "grad_norm": 1.9509342908859253, + "learning_rate": 1.4689032176705408e-05, + "loss": 0.5403, + "step": 11552 + }, + { + "epoch": 0.3643904746885349, + "grad_norm": 1.7528846263885498, + "learning_rate": 1.4688129862932994e-05, + "loss": 0.3516, + "step": 11553 + }, + { + "epoch": 0.36442201545497555, + "grad_norm": 1.4803755283355713, + "learning_rate": 1.4687227500236303e-05, + "loss": 0.5355, + "step": 11554 + }, + { + "epoch": 0.3644535562214162, + "grad_norm": 1.8051596879959106, + "learning_rate": 1.4686325088624751e-05, + "loss": 0.5115, + "step": 11555 + }, + { + "epoch": 0.3644850969878568, + "grad_norm": 2.1387858390808105, + "learning_rate": 1.4685422628107757e-05, + "loss": 0.5366, + "step": 11556 + }, + { + "epoch": 0.36451663775429743, + "grad_norm": 1.8662909269332886, + "learning_rate": 1.4684520118694743e-05, + "loss": 0.5417, + "step": 11557 + }, + { + "epoch": 0.36454817852073806, + "grad_norm": 1.8875819444656372, + "learning_rate": 1.468361756039512e-05, + "loss": 0.4922, + "step": 11558 + }, + { + "epoch": 0.3645797192871787, + "grad_norm": 1.5371372699737549, + "learning_rate": 1.4682714953218313e-05, + "loss": 0.5234, + "step": 11559 + }, + { + "epoch": 0.3646112600536193, + "grad_norm": 3.195774555206299, + "learning_rate": 1.468181229717374e-05, + "loss": 0.5165, + "step": 11560 + }, + { + "epoch": 0.36464280082005995, + "grad_norm": 1.4964252710342407, + "learning_rate": 1.468090959227082e-05, + "loss": 0.512, + "step": 11561 + }, + { + "epoch": 0.36467434158650053, + "grad_norm": 2.5758895874023438, + "learning_rate": 1.4680006838518975e-05, + "loss": 0.4188, + "step": 11562 + }, + { + "epoch": 0.36470588235294116, + "grad_norm": 2.026383399963379, + "learning_rate": 1.4679104035927625e-05, + "loss": 0.5057, + "step": 11563 + }, + { + "epoch": 0.3647374231193818, + "grad_norm": 2.049708127975464, + "learning_rate": 1.467820118450619e-05, + "loss": 0.4986, + "step": 11564 + }, + { + "epoch": 0.3647689638858224, + "grad_norm": 1.7635594606399536, + "learning_rate": 1.4677298284264092e-05, + "loss": 0.5234, + "step": 11565 + }, + { + "epoch": 0.36480050465226305, + "grad_norm": 1.619071125984192, + "learning_rate": 1.4676395335210757e-05, + "loss": 0.815, + "step": 11566 + }, + { + "epoch": 0.3648320454187037, + "grad_norm": 1.5822123289108276, + "learning_rate": 1.4675492337355601e-05, + "loss": 0.5283, + "step": 11567 + }, + { + "epoch": 0.3648635861851443, + "grad_norm": 1.844348430633545, + "learning_rate": 1.4674589290708057e-05, + "loss": 0.3622, + "step": 11568 + }, + { + "epoch": 0.36489512695158494, + "grad_norm": 1.796522617340088, + "learning_rate": 1.467368619527754e-05, + "loss": 0.5323, + "step": 11569 + }, + { + "epoch": 0.36492666771802557, + "grad_norm": 2.3981223106384277, + "learning_rate": 1.4672783051073481e-05, + "loss": 0.8228, + "step": 11570 + }, + { + "epoch": 0.3649582084844662, + "grad_norm": 1.9483457803726196, + "learning_rate": 1.4671879858105303e-05, + "loss": 0.5426, + "step": 11571 + }, + { + "epoch": 0.36498974925090677, + "grad_norm": 1.8877373933792114, + "learning_rate": 1.467097661638243e-05, + "loss": 0.5266, + "step": 11572 + }, + { + "epoch": 0.3650212900173474, + "grad_norm": 1.878258466720581, + "learning_rate": 1.4670073325914292e-05, + "loss": 0.5185, + "step": 11573 + }, + { + "epoch": 0.36505283078378803, + "grad_norm": 1.4094042778015137, + "learning_rate": 1.466916998671031e-05, + "loss": 0.4709, + "step": 11574 + }, + { + "epoch": 0.36508437155022866, + "grad_norm": 2.0720341205596924, + "learning_rate": 1.4668266598779911e-05, + "loss": 0.5281, + "step": 11575 + }, + { + "epoch": 0.3651159123166693, + "grad_norm": 1.3341469764709473, + "learning_rate": 1.4667363162132528e-05, + "loss": 0.5314, + "step": 11576 + }, + { + "epoch": 0.3651474530831099, + "grad_norm": 2.3118491172790527, + "learning_rate": 1.4666459676777587e-05, + "loss": 0.4836, + "step": 11577 + }, + { + "epoch": 0.36517899384955055, + "grad_norm": 1.829724669456482, + "learning_rate": 1.4665556142724515e-05, + "loss": 0.5201, + "step": 11578 + }, + { + "epoch": 0.3652105346159912, + "grad_norm": 1.6328988075256348, + "learning_rate": 1.4664652559982742e-05, + "loss": 0.509, + "step": 11579 + }, + { + "epoch": 0.3652420753824318, + "grad_norm": 1.5933070182800293, + "learning_rate": 1.4663748928561696e-05, + "loss": 0.5046, + "step": 11580 + }, + { + "epoch": 0.36527361614887244, + "grad_norm": 1.5987600088119507, + "learning_rate": 1.4662845248470813e-05, + "loss": 0.514, + "step": 11581 + }, + { + "epoch": 0.36530515691531307, + "grad_norm": 1.5107210874557495, + "learning_rate": 1.4661941519719514e-05, + "loss": 0.5763, + "step": 11582 + }, + { + "epoch": 0.36533669768175364, + "grad_norm": 1.6449629068374634, + "learning_rate": 1.4661037742317239e-05, + "loss": 0.5494, + "step": 11583 + }, + { + "epoch": 0.36536823844819427, + "grad_norm": 1.6165248155593872, + "learning_rate": 1.4660133916273414e-05, + "loss": 0.5254, + "step": 11584 + }, + { + "epoch": 0.3653997792146349, + "grad_norm": 1.500887155532837, + "learning_rate": 1.4659230041597474e-05, + "loss": 0.5298, + "step": 11585 + }, + { + "epoch": 0.36543131998107553, + "grad_norm": 1.3393349647521973, + "learning_rate": 1.4658326118298852e-05, + "loss": 0.5263, + "step": 11586 + }, + { + "epoch": 0.36546286074751616, + "grad_norm": 1.8405749797821045, + "learning_rate": 1.4657422146386976e-05, + "loss": 0.5074, + "step": 11587 + }, + { + "epoch": 0.3654944015139568, + "grad_norm": 1.9240587949752808, + "learning_rate": 1.4656518125871287e-05, + "loss": 0.516, + "step": 11588 + }, + { + "epoch": 0.3655259422803974, + "grad_norm": 1.6131564378738403, + "learning_rate": 1.4655614056761213e-05, + "loss": 0.8449, + "step": 11589 + }, + { + "epoch": 0.36555748304683805, + "grad_norm": 1.9474166631698608, + "learning_rate": 1.4654709939066192e-05, + "loss": 0.5166, + "step": 11590 + }, + { + "epoch": 0.3655890238132787, + "grad_norm": 1.8646011352539062, + "learning_rate": 1.4653805772795662e-05, + "loss": 0.8591, + "step": 11591 + }, + { + "epoch": 0.3656205645797193, + "grad_norm": 1.3605077266693115, + "learning_rate": 1.4652901557959052e-05, + "loss": 0.5033, + "step": 11592 + }, + { + "epoch": 0.36565210534615994, + "grad_norm": 1.7238309383392334, + "learning_rate": 1.4651997294565802e-05, + "loss": 0.8335, + "step": 11593 + }, + { + "epoch": 0.3656836461126005, + "grad_norm": 1.858999490737915, + "learning_rate": 1.4651092982625347e-05, + "loss": 0.51, + "step": 11594 + }, + { + "epoch": 0.36571518687904114, + "grad_norm": 1.5477160215377808, + "learning_rate": 1.4650188622147127e-05, + "loss": 0.513, + "step": 11595 + }, + { + "epoch": 0.3657467276454818, + "grad_norm": 1.5656150579452515, + "learning_rate": 1.4649284213140576e-05, + "loss": 0.5239, + "step": 11596 + }, + { + "epoch": 0.3657782684119224, + "grad_norm": 1.4954190254211426, + "learning_rate": 1.4648379755615137e-05, + "loss": 0.5101, + "step": 11597 + }, + { + "epoch": 0.36580980917836303, + "grad_norm": 1.5004711151123047, + "learning_rate": 1.4647475249580242e-05, + "loss": 0.5369, + "step": 11598 + }, + { + "epoch": 0.36584134994480366, + "grad_norm": 1.5545064210891724, + "learning_rate": 1.4646570695045337e-05, + "loss": 0.8233, + "step": 11599 + }, + { + "epoch": 0.3658728907112443, + "grad_norm": 1.4500592947006226, + "learning_rate": 1.4645666092019857e-05, + "loss": 0.5063, + "step": 11600 + }, + { + "epoch": 0.3659044314776849, + "grad_norm": 1.429632306098938, + "learning_rate": 1.4644761440513246e-05, + "loss": 0.5379, + "step": 11601 + }, + { + "epoch": 0.36593597224412555, + "grad_norm": 1.5179272890090942, + "learning_rate": 1.4643856740534943e-05, + "loss": 0.8142, + "step": 11602 + }, + { + "epoch": 0.3659675130105662, + "grad_norm": 2.185117721557617, + "learning_rate": 1.4642951992094388e-05, + "loss": 0.5276, + "step": 11603 + }, + { + "epoch": 0.36599905377700676, + "grad_norm": 1.8986988067626953, + "learning_rate": 1.4642047195201023e-05, + "loss": 0.4911, + "step": 11604 + }, + { + "epoch": 0.3660305945434474, + "grad_norm": 1.7012218236923218, + "learning_rate": 1.4641142349864292e-05, + "loss": 0.5026, + "step": 11605 + }, + { + "epoch": 0.366062135309888, + "grad_norm": 1.8588956594467163, + "learning_rate": 1.4640237456093636e-05, + "loss": 0.4893, + "step": 11606 + }, + { + "epoch": 0.36609367607632864, + "grad_norm": 2.1876437664031982, + "learning_rate": 1.4639332513898499e-05, + "loss": 0.538, + "step": 11607 + }, + { + "epoch": 0.3661252168427693, + "grad_norm": 1.6687569618225098, + "learning_rate": 1.4638427523288326e-05, + "loss": 0.5169, + "step": 11608 + }, + { + "epoch": 0.3661567576092099, + "grad_norm": 1.8211275339126587, + "learning_rate": 1.463752248427256e-05, + "loss": 0.4977, + "step": 11609 + }, + { + "epoch": 0.36618829837565053, + "grad_norm": 1.7009034156799316, + "learning_rate": 1.4636617396860646e-05, + "loss": 0.8213, + "step": 11610 + }, + { + "epoch": 0.36621983914209116, + "grad_norm": 1.613950252532959, + "learning_rate": 1.463571226106203e-05, + "loss": 0.4979, + "step": 11611 + }, + { + "epoch": 0.3662513799085318, + "grad_norm": 2.6604599952697754, + "learning_rate": 1.4634807076886155e-05, + "loss": 0.533, + "step": 11612 + }, + { + "epoch": 0.3662829206749724, + "grad_norm": 1.5472629070281982, + "learning_rate": 1.463390184434247e-05, + "loss": 0.5386, + "step": 11613 + }, + { + "epoch": 0.36631446144141305, + "grad_norm": 1.444386601448059, + "learning_rate": 1.463299656344042e-05, + "loss": 0.5102, + "step": 11614 + }, + { + "epoch": 0.3663460022078536, + "grad_norm": 1.4812990427017212, + "learning_rate": 1.4632091234189457e-05, + "loss": 0.3644, + "step": 11615 + }, + { + "epoch": 0.36637754297429426, + "grad_norm": 1.5684493780136108, + "learning_rate": 1.4631185856599021e-05, + "loss": 0.4946, + "step": 11616 + }, + { + "epoch": 0.3664090837407349, + "grad_norm": 3.042147636413574, + "learning_rate": 1.4630280430678567e-05, + "loss": 0.5498, + "step": 11617 + }, + { + "epoch": 0.3664406245071755, + "grad_norm": 1.7414366006851196, + "learning_rate": 1.462937495643754e-05, + "loss": 0.4937, + "step": 11618 + }, + { + "epoch": 0.36647216527361615, + "grad_norm": 1.8620424270629883, + "learning_rate": 1.4628469433885393e-05, + "loss": 0.4977, + "step": 11619 + }, + { + "epoch": 0.3665037060400568, + "grad_norm": 1.7021366357803345, + "learning_rate": 1.4627563863031573e-05, + "loss": 0.5169, + "step": 11620 + }, + { + "epoch": 0.3665352468064974, + "grad_norm": 2.143167495727539, + "learning_rate": 1.462665824388553e-05, + "loss": 0.5331, + "step": 11621 + }, + { + "epoch": 0.36656678757293804, + "grad_norm": 1.7917377948760986, + "learning_rate": 1.4625752576456716e-05, + "loss": 0.509, + "step": 11622 + }, + { + "epoch": 0.36659832833937867, + "grad_norm": 1.6246885061264038, + "learning_rate": 1.4624846860754583e-05, + "loss": 0.5048, + "step": 11623 + }, + { + "epoch": 0.3666298691058193, + "grad_norm": 1.512393593788147, + "learning_rate": 1.4623941096788579e-05, + "loss": 0.506, + "step": 11624 + }, + { + "epoch": 0.36666140987225987, + "grad_norm": 1.5568844079971313, + "learning_rate": 1.4623035284568162e-05, + "loss": 0.5013, + "step": 11625 + }, + { + "epoch": 0.3666929506387005, + "grad_norm": 1.6107813119888306, + "learning_rate": 1.462212942410278e-05, + "loss": 0.5178, + "step": 11626 + }, + { + "epoch": 0.36672449140514113, + "grad_norm": 1.5337178707122803, + "learning_rate": 1.462122351540189e-05, + "loss": 0.5174, + "step": 11627 + }, + { + "epoch": 0.36675603217158176, + "grad_norm": 1.6254727840423584, + "learning_rate": 1.4620317558474944e-05, + "loss": 0.5013, + "step": 11628 + }, + { + "epoch": 0.3667875729380224, + "grad_norm": 2.71573805809021, + "learning_rate": 1.4619411553331397e-05, + "loss": 0.5146, + "step": 11629 + }, + { + "epoch": 0.366819113704463, + "grad_norm": 1.5178883075714111, + "learning_rate": 1.4618505499980702e-05, + "loss": 0.4962, + "step": 11630 + }, + { + "epoch": 0.36685065447090365, + "grad_norm": 1.5904539823532104, + "learning_rate": 1.4617599398432319e-05, + "loss": 0.4968, + "step": 11631 + }, + { + "epoch": 0.3668821952373443, + "grad_norm": 1.9174976348876953, + "learning_rate": 1.4616693248695696e-05, + "loss": 0.5061, + "step": 11632 + }, + { + "epoch": 0.3669137360037849, + "grad_norm": 2.6832282543182373, + "learning_rate": 1.46157870507803e-05, + "loss": 0.4916, + "step": 11633 + }, + { + "epoch": 0.36694527677022554, + "grad_norm": 1.8121298551559448, + "learning_rate": 1.4614880804695579e-05, + "loss": 0.448, + "step": 11634 + }, + { + "epoch": 0.36697681753666617, + "grad_norm": 2.2524499893188477, + "learning_rate": 1.4613974510450992e-05, + "loss": 0.5308, + "step": 11635 + }, + { + "epoch": 0.36700835830310674, + "grad_norm": 1.7423449754714966, + "learning_rate": 1.4613068168056002e-05, + "loss": 0.5397, + "step": 11636 + }, + { + "epoch": 0.36703989906954737, + "grad_norm": 2.019608736038208, + "learning_rate": 1.4612161777520059e-05, + "loss": 0.4725, + "step": 11637 + }, + { + "epoch": 0.367071439835988, + "grad_norm": 1.989352822303772, + "learning_rate": 1.461125533885263e-05, + "loss": 0.5292, + "step": 11638 + }, + { + "epoch": 0.36710298060242863, + "grad_norm": 1.6350022554397583, + "learning_rate": 1.461034885206317e-05, + "loss": 0.5127, + "step": 11639 + }, + { + "epoch": 0.36713452136886926, + "grad_norm": 2.8837995529174805, + "learning_rate": 1.460944231716114e-05, + "loss": 0.8293, + "step": 11640 + }, + { + "epoch": 0.3671660621353099, + "grad_norm": 2.2868711948394775, + "learning_rate": 1.4608535734156002e-05, + "loss": 0.52, + "step": 11641 + }, + { + "epoch": 0.3671976029017505, + "grad_norm": 1.798834204673767, + "learning_rate": 1.4607629103057214e-05, + "loss": 0.5277, + "step": 11642 + }, + { + "epoch": 0.36722914366819115, + "grad_norm": 1.7142317295074463, + "learning_rate": 1.4606722423874237e-05, + "loss": 0.5334, + "step": 11643 + }, + { + "epoch": 0.3672606844346318, + "grad_norm": 1.7108049392700195, + "learning_rate": 1.4605815696616536e-05, + "loss": 0.5013, + "step": 11644 + }, + { + "epoch": 0.3672922252010724, + "grad_norm": 1.8871502876281738, + "learning_rate": 1.4604908921293573e-05, + "loss": 0.5322, + "step": 11645 + }, + { + "epoch": 0.367323765967513, + "grad_norm": 1.4174916744232178, + "learning_rate": 1.4604002097914806e-05, + "loss": 0.5364, + "step": 11646 + }, + { + "epoch": 0.3673553067339536, + "grad_norm": 1.5138977766036987, + "learning_rate": 1.4603095226489706e-05, + "loss": 0.5026, + "step": 11647 + }, + { + "epoch": 0.36738684750039424, + "grad_norm": 2.41336989402771, + "learning_rate": 1.4602188307027732e-05, + "loss": 0.5197, + "step": 11648 + }, + { + "epoch": 0.36741838826683487, + "grad_norm": 1.8350942134857178, + "learning_rate": 1.4601281339538347e-05, + "loss": 0.5093, + "step": 11649 + }, + { + "epoch": 0.3674499290332755, + "grad_norm": 2.281937837600708, + "learning_rate": 1.4600374324031021e-05, + "loss": 0.5291, + "step": 11650 + }, + { + "epoch": 0.36748146979971613, + "grad_norm": 1.9294240474700928, + "learning_rate": 1.459946726051522e-05, + "loss": 0.521, + "step": 11651 + }, + { + "epoch": 0.36751301056615676, + "grad_norm": 2.1544697284698486, + "learning_rate": 1.45985601490004e-05, + "loss": 0.5058, + "step": 11652 + }, + { + "epoch": 0.3675445513325974, + "grad_norm": 2.479003667831421, + "learning_rate": 1.4597652989496038e-05, + "loss": 0.5167, + "step": 11653 + }, + { + "epoch": 0.367576092099038, + "grad_norm": 2.0761125087738037, + "learning_rate": 1.4596745782011597e-05, + "loss": 0.5241, + "step": 11654 + }, + { + "epoch": 0.36760763286547865, + "grad_norm": 1.660919189453125, + "learning_rate": 1.4595838526556543e-05, + "loss": 0.8353, + "step": 11655 + }, + { + "epoch": 0.3676391736319193, + "grad_norm": 1.9638746976852417, + "learning_rate": 1.4594931223140345e-05, + "loss": 0.5336, + "step": 11656 + }, + { + "epoch": 0.36767071439835985, + "grad_norm": 1.5952167510986328, + "learning_rate": 1.4594023871772477e-05, + "loss": 0.5519, + "step": 11657 + }, + { + "epoch": 0.3677022551648005, + "grad_norm": 1.5623376369476318, + "learning_rate": 1.4593116472462396e-05, + "loss": 0.5059, + "step": 11658 + }, + { + "epoch": 0.3677337959312411, + "grad_norm": 1.7847027778625488, + "learning_rate": 1.459220902521958e-05, + "loss": 0.518, + "step": 11659 + }, + { + "epoch": 0.36776533669768174, + "grad_norm": 4.421815395355225, + "learning_rate": 1.45913015300535e-05, + "loss": 0.5344, + "step": 11660 + }, + { + "epoch": 0.3677968774641224, + "grad_norm": 1.3945013284683228, + "learning_rate": 1.4590393986973621e-05, + "loss": 0.8646, + "step": 11661 + }, + { + "epoch": 0.367828418230563, + "grad_norm": 2.0329067707061768, + "learning_rate": 1.4589486395989419e-05, + "loss": 0.5232, + "step": 11662 + }, + { + "epoch": 0.36785995899700363, + "grad_norm": 2.178433656692505, + "learning_rate": 1.4588578757110359e-05, + "loss": 0.5512, + "step": 11663 + }, + { + "epoch": 0.36789149976344426, + "grad_norm": 1.5973495244979858, + "learning_rate": 1.4587671070345919e-05, + "loss": 0.5098, + "step": 11664 + }, + { + "epoch": 0.3679230405298849, + "grad_norm": 2.236726760864258, + "learning_rate": 1.4586763335705567e-05, + "loss": 0.5426, + "step": 11665 + }, + { + "epoch": 0.3679545812963255, + "grad_norm": 1.7981683015823364, + "learning_rate": 1.458585555319878e-05, + "loss": 0.5008, + "step": 11666 + }, + { + "epoch": 0.36798612206276615, + "grad_norm": 1.8907692432403564, + "learning_rate": 1.458494772283503e-05, + "loss": 0.5159, + "step": 11667 + }, + { + "epoch": 0.3680176628292067, + "grad_norm": 1.2522798776626587, + "learning_rate": 1.458403984462379e-05, + "loss": 0.8317, + "step": 11668 + }, + { + "epoch": 0.36804920359564736, + "grad_norm": 1.6443674564361572, + "learning_rate": 1.4583131918574532e-05, + "loss": 0.5192, + "step": 11669 + }, + { + "epoch": 0.368080744362088, + "grad_norm": 1.6342732906341553, + "learning_rate": 1.4582223944696736e-05, + "loss": 0.5118, + "step": 11670 + }, + { + "epoch": 0.3681122851285286, + "grad_norm": 1.3630255460739136, + "learning_rate": 1.4581315922999876e-05, + "loss": 0.5439, + "step": 11671 + }, + { + "epoch": 0.36814382589496925, + "grad_norm": 1.9585697650909424, + "learning_rate": 1.4580407853493426e-05, + "loss": 0.393, + "step": 11672 + }, + { + "epoch": 0.3681753666614099, + "grad_norm": 1.839520812034607, + "learning_rate": 1.4579499736186864e-05, + "loss": 0.5582, + "step": 11673 + }, + { + "epoch": 0.3682069074278505, + "grad_norm": 1.7802379131317139, + "learning_rate": 1.4578591571089666e-05, + "loss": 0.5215, + "step": 11674 + }, + { + "epoch": 0.36823844819429113, + "grad_norm": 1.6593425273895264, + "learning_rate": 1.457768335821131e-05, + "loss": 0.4106, + "step": 11675 + }, + { + "epoch": 0.36826998896073176, + "grad_norm": 2.072328805923462, + "learning_rate": 1.4576775097561273e-05, + "loss": 0.4994, + "step": 11676 + }, + { + "epoch": 0.3683015297271724, + "grad_norm": 1.7708455324172974, + "learning_rate": 1.4575866789149036e-05, + "loss": 0.5154, + "step": 11677 + }, + { + "epoch": 0.36833307049361297, + "grad_norm": 1.8379710912704468, + "learning_rate": 1.4574958432984075e-05, + "loss": 0.5383, + "step": 11678 + }, + { + "epoch": 0.3683646112600536, + "grad_norm": 1.7336238622665405, + "learning_rate": 1.4574050029075868e-05, + "loss": 0.5207, + "step": 11679 + }, + { + "epoch": 0.36839615202649423, + "grad_norm": 2.3506598472595215, + "learning_rate": 1.45731415774339e-05, + "loss": 0.5175, + "step": 11680 + }, + { + "epoch": 0.36842769279293486, + "grad_norm": 2.027564287185669, + "learning_rate": 1.4572233078067648e-05, + "loss": 0.5053, + "step": 11681 + }, + { + "epoch": 0.3684592335593755, + "grad_norm": 1.8724273443222046, + "learning_rate": 1.4571324530986592e-05, + "loss": 0.5169, + "step": 11682 + }, + { + "epoch": 0.3684907743258161, + "grad_norm": 1.8478198051452637, + "learning_rate": 1.4570415936200217e-05, + "loss": 0.5562, + "step": 11683 + }, + { + "epoch": 0.36852231509225675, + "grad_norm": 1.8212144374847412, + "learning_rate": 1.4569507293718005e-05, + "loss": 0.5011, + "step": 11684 + }, + { + "epoch": 0.3685538558586974, + "grad_norm": 5.711288928985596, + "learning_rate": 1.4568598603549431e-05, + "loss": 0.5028, + "step": 11685 + }, + { + "epoch": 0.368585396625138, + "grad_norm": 2.2138872146606445, + "learning_rate": 1.4567689865703987e-05, + "loss": 0.3894, + "step": 11686 + }, + { + "epoch": 0.36861693739157864, + "grad_norm": 1.789583444595337, + "learning_rate": 1.4566781080191147e-05, + "loss": 0.4965, + "step": 11687 + }, + { + "epoch": 0.36864847815801927, + "grad_norm": 2.268496513366699, + "learning_rate": 1.4565872247020407e-05, + "loss": 0.807, + "step": 11688 + }, + { + "epoch": 0.36868001892445984, + "grad_norm": 1.572744607925415, + "learning_rate": 1.4564963366201243e-05, + "loss": 0.4719, + "step": 11689 + }, + { + "epoch": 0.36871155969090047, + "grad_norm": 2.1911377906799316, + "learning_rate": 1.4564054437743143e-05, + "loss": 0.5197, + "step": 11690 + }, + { + "epoch": 0.3687431004573411, + "grad_norm": 1.5990374088287354, + "learning_rate": 1.456314546165559e-05, + "loss": 0.5168, + "step": 11691 + }, + { + "epoch": 0.36877464122378173, + "grad_norm": 1.7869689464569092, + "learning_rate": 1.456223643794807e-05, + "loss": 0.5293, + "step": 11692 + }, + { + "epoch": 0.36880618199022236, + "grad_norm": 1.630486011505127, + "learning_rate": 1.4561327366630069e-05, + "loss": 0.4813, + "step": 11693 + }, + { + "epoch": 0.368837722756663, + "grad_norm": 1.5466469526290894, + "learning_rate": 1.4560418247711075e-05, + "loss": 0.5384, + "step": 11694 + }, + { + "epoch": 0.3688692635231036, + "grad_norm": 8.905324935913086, + "learning_rate": 1.4559509081200578e-05, + "loss": 0.5484, + "step": 11695 + }, + { + "epoch": 0.36890080428954425, + "grad_norm": 1.8955999612808228, + "learning_rate": 1.4558599867108062e-05, + "loss": 0.3809, + "step": 11696 + }, + { + "epoch": 0.3689323450559849, + "grad_norm": 1.514533519744873, + "learning_rate": 1.4557690605443015e-05, + "loss": 0.5255, + "step": 11697 + }, + { + "epoch": 0.3689638858224255, + "grad_norm": 1.3699368238449097, + "learning_rate": 1.4556781296214932e-05, + "loss": 0.511, + "step": 11698 + }, + { + "epoch": 0.3689954265888661, + "grad_norm": 1.4723931550979614, + "learning_rate": 1.4555871939433294e-05, + "loss": 0.3823, + "step": 11699 + }, + { + "epoch": 0.3690269673553067, + "grad_norm": 1.8757895231246948, + "learning_rate": 1.4554962535107597e-05, + "loss": 0.51, + "step": 11700 + }, + { + "epoch": 0.36905850812174734, + "grad_norm": 1.5830235481262207, + "learning_rate": 1.4554053083247333e-05, + "loss": 0.5104, + "step": 11701 + }, + { + "epoch": 0.36909004888818797, + "grad_norm": 1.9077779054641724, + "learning_rate": 1.4553143583861981e-05, + "loss": 0.5392, + "step": 11702 + }, + { + "epoch": 0.3691215896546286, + "grad_norm": 1.736991047859192, + "learning_rate": 1.4552234036961044e-05, + "loss": 0.4891, + "step": 11703 + }, + { + "epoch": 0.36915313042106923, + "grad_norm": 1.649229645729065, + "learning_rate": 1.455132444255401e-05, + "loss": 0.5304, + "step": 11704 + }, + { + "epoch": 0.36918467118750986, + "grad_norm": 1.8422852754592896, + "learning_rate": 1.4550414800650372e-05, + "loss": 0.5143, + "step": 11705 + }, + { + "epoch": 0.3692162119539505, + "grad_norm": 1.5309735536575317, + "learning_rate": 1.4549505111259621e-05, + "loss": 0.8525, + "step": 11706 + }, + { + "epoch": 0.3692477527203911, + "grad_norm": 1.7132575511932373, + "learning_rate": 1.454859537439125e-05, + "loss": 0.5153, + "step": 11707 + }, + { + "epoch": 0.36927929348683175, + "grad_norm": 1.4997320175170898, + "learning_rate": 1.4547685590054759e-05, + "loss": 0.513, + "step": 11708 + }, + { + "epoch": 0.3693108342532724, + "grad_norm": 1.9620100259780884, + "learning_rate": 1.4546775758259635e-05, + "loss": 0.5267, + "step": 11709 + }, + { + "epoch": 0.36934237501971295, + "grad_norm": 1.7639809846878052, + "learning_rate": 1.4545865879015375e-05, + "loss": 0.5405, + "step": 11710 + }, + { + "epoch": 0.3693739157861536, + "grad_norm": 1.7631670236587524, + "learning_rate": 1.4544955952331474e-05, + "loss": 0.8328, + "step": 11711 + }, + { + "epoch": 0.3694054565525942, + "grad_norm": 2.058519124984741, + "learning_rate": 1.4544045978217432e-05, + "loss": 0.5249, + "step": 11712 + }, + { + "epoch": 0.36943699731903484, + "grad_norm": 1.8908456563949585, + "learning_rate": 1.4543135956682738e-05, + "loss": 0.5067, + "step": 11713 + }, + { + "epoch": 0.3694685380854755, + "grad_norm": 2.9475040435791016, + "learning_rate": 1.4542225887736896e-05, + "loss": 0.4912, + "step": 11714 + }, + { + "epoch": 0.3695000788519161, + "grad_norm": 1.977508783340454, + "learning_rate": 1.4541315771389399e-05, + "loss": 0.4398, + "step": 11715 + }, + { + "epoch": 0.36953161961835673, + "grad_norm": 1.9834948778152466, + "learning_rate": 1.4540405607649743e-05, + "loss": 0.5112, + "step": 11716 + }, + { + "epoch": 0.36956316038479736, + "grad_norm": 1.9610717296600342, + "learning_rate": 1.4539495396527432e-05, + "loss": 0.5046, + "step": 11717 + }, + { + "epoch": 0.369594701151238, + "grad_norm": 1.7225431203842163, + "learning_rate": 1.453858513803196e-05, + "loss": 0.547, + "step": 11718 + }, + { + "epoch": 0.3696262419176786, + "grad_norm": 2.0135271549224854, + "learning_rate": 1.453767483217283e-05, + "loss": 0.5073, + "step": 11719 + }, + { + "epoch": 0.36965778268411925, + "grad_norm": 2.1836583614349365, + "learning_rate": 1.453676447895954e-05, + "loss": 0.5164, + "step": 11720 + }, + { + "epoch": 0.3696893234505598, + "grad_norm": 1.6307761669158936, + "learning_rate": 1.4535854078401585e-05, + "loss": 0.5209, + "step": 11721 + }, + { + "epoch": 0.36972086421700046, + "grad_norm": 1.6241519451141357, + "learning_rate": 1.4534943630508478e-05, + "loss": 0.4978, + "step": 11722 + }, + { + "epoch": 0.3697524049834411, + "grad_norm": 1.9184081554412842, + "learning_rate": 1.4534033135289707e-05, + "loss": 0.5129, + "step": 11723 + }, + { + "epoch": 0.3697839457498817, + "grad_norm": 1.6222777366638184, + "learning_rate": 1.453312259275478e-05, + "loss": 0.5004, + "step": 11724 + }, + { + "epoch": 0.36981548651632234, + "grad_norm": 3.0382118225097656, + "learning_rate": 1.4532212002913201e-05, + "loss": 0.8288, + "step": 11725 + }, + { + "epoch": 0.369847027282763, + "grad_norm": 2.2677102088928223, + "learning_rate": 1.453130136577447e-05, + "loss": 0.5358, + "step": 11726 + }, + { + "epoch": 0.3698785680492036, + "grad_norm": 1.976231336593628, + "learning_rate": 1.453039068134809e-05, + "loss": 0.5289, + "step": 11727 + }, + { + "epoch": 0.36991010881564423, + "grad_norm": 2.1495327949523926, + "learning_rate": 1.4529479949643568e-05, + "loss": 0.502, + "step": 11728 + }, + { + "epoch": 0.36994164958208486, + "grad_norm": 2.001357316970825, + "learning_rate": 1.4528569170670404e-05, + "loss": 0.4724, + "step": 11729 + }, + { + "epoch": 0.3699731903485255, + "grad_norm": 2.3218047618865967, + "learning_rate": 1.4527658344438104e-05, + "loss": 0.5102, + "step": 11730 + }, + { + "epoch": 0.37000473111496607, + "grad_norm": 1.933022379875183, + "learning_rate": 1.4526747470956175e-05, + "loss": 0.5246, + "step": 11731 + }, + { + "epoch": 0.3700362718814067, + "grad_norm": 1.6841922998428345, + "learning_rate": 1.452583655023412e-05, + "loss": 0.4968, + "step": 11732 + }, + { + "epoch": 0.3700678126478473, + "grad_norm": 1.9541306495666504, + "learning_rate": 1.4524925582281446e-05, + "loss": 0.5338, + "step": 11733 + }, + { + "epoch": 0.37009935341428796, + "grad_norm": 2.0578532218933105, + "learning_rate": 1.452401456710766e-05, + "loss": 0.5339, + "step": 11734 + }, + { + "epoch": 0.3701308941807286, + "grad_norm": 2.335038423538208, + "learning_rate": 1.4523103504722269e-05, + "loss": 0.5161, + "step": 11735 + }, + { + "epoch": 0.3701624349471692, + "grad_norm": 1.9820891618728638, + "learning_rate": 1.4522192395134784e-05, + "loss": 0.5083, + "step": 11736 + }, + { + "epoch": 0.37019397571360985, + "grad_norm": 2.0301320552825928, + "learning_rate": 1.4521281238354704e-05, + "loss": 0.5169, + "step": 11737 + }, + { + "epoch": 0.3702255164800505, + "grad_norm": 2.0947682857513428, + "learning_rate": 1.452037003439155e-05, + "loss": 0.5212, + "step": 11738 + }, + { + "epoch": 0.3702570572464911, + "grad_norm": 2.1319220066070557, + "learning_rate": 1.4519458783254821e-05, + "loss": 0.532, + "step": 11739 + }, + { + "epoch": 0.37028859801293174, + "grad_norm": 1.8637837171554565, + "learning_rate": 1.4518547484954033e-05, + "loss": 0.8506, + "step": 11740 + }, + { + "epoch": 0.37032013877937237, + "grad_norm": 1.8521395921707153, + "learning_rate": 1.451763613949869e-05, + "loss": 0.5527, + "step": 11741 + }, + { + "epoch": 0.37035167954581294, + "grad_norm": 1.5799920558929443, + "learning_rate": 1.4516724746898308e-05, + "loss": 0.8201, + "step": 11742 + }, + { + "epoch": 0.37038322031225357, + "grad_norm": 1.4986798763275146, + "learning_rate": 1.4515813307162397e-05, + "loss": 0.5281, + "step": 11743 + }, + { + "epoch": 0.3704147610786942, + "grad_norm": 1.8490885496139526, + "learning_rate": 1.4514901820300464e-05, + "loss": 0.5256, + "step": 11744 + }, + { + "epoch": 0.37044630184513483, + "grad_norm": 2.5390939712524414, + "learning_rate": 1.4513990286322028e-05, + "loss": 0.5478, + "step": 11745 + }, + { + "epoch": 0.37047784261157546, + "grad_norm": 2.082658529281616, + "learning_rate": 1.4513078705236598e-05, + "loss": 0.8563, + "step": 11746 + }, + { + "epoch": 0.3705093833780161, + "grad_norm": 1.7025612592697144, + "learning_rate": 1.4512167077053687e-05, + "loss": 0.5296, + "step": 11747 + }, + { + "epoch": 0.3705409241444567, + "grad_norm": 1.8842167854309082, + "learning_rate": 1.4511255401782809e-05, + "loss": 0.5205, + "step": 11748 + }, + { + "epoch": 0.37057246491089735, + "grad_norm": 1.7520806789398193, + "learning_rate": 1.4510343679433479e-05, + "loss": 0.5083, + "step": 11749 + }, + { + "epoch": 0.370604005677338, + "grad_norm": 1.8376705646514893, + "learning_rate": 1.450943191001521e-05, + "loss": 0.5263, + "step": 11750 + }, + { + "epoch": 0.3706355464437786, + "grad_norm": 1.3683881759643555, + "learning_rate": 1.4508520093537514e-05, + "loss": 0.4819, + "step": 11751 + }, + { + "epoch": 0.3706670872102192, + "grad_norm": 3.02789568901062, + "learning_rate": 1.4507608230009913e-05, + "loss": 0.4225, + "step": 11752 + }, + { + "epoch": 0.3706986279766598, + "grad_norm": 1.51878023147583, + "learning_rate": 1.4506696319441922e-05, + "loss": 0.8037, + "step": 11753 + }, + { + "epoch": 0.37073016874310044, + "grad_norm": 1.8568580150604248, + "learning_rate": 1.4505784361843053e-05, + "loss": 0.5284, + "step": 11754 + }, + { + "epoch": 0.37076170950954107, + "grad_norm": 2.5475618839263916, + "learning_rate": 1.4504872357222826e-05, + "loss": 0.5664, + "step": 11755 + }, + { + "epoch": 0.3707932502759817, + "grad_norm": 2.0024685859680176, + "learning_rate": 1.4503960305590759e-05, + "loss": 0.5323, + "step": 11756 + }, + { + "epoch": 0.37082479104242233, + "grad_norm": 1.9197932481765747, + "learning_rate": 1.4503048206956368e-05, + "loss": 0.5316, + "step": 11757 + }, + { + "epoch": 0.37085633180886296, + "grad_norm": 1.7231287956237793, + "learning_rate": 1.4502136061329173e-05, + "loss": 0.5, + "step": 11758 + }, + { + "epoch": 0.3708878725753036, + "grad_norm": 2.1642069816589355, + "learning_rate": 1.450122386871869e-05, + "loss": 0.3863, + "step": 11759 + }, + { + "epoch": 0.3709194133417442, + "grad_norm": 1.9650280475616455, + "learning_rate": 1.4500311629134444e-05, + "loss": 0.8274, + "step": 11760 + }, + { + "epoch": 0.37095095410818485, + "grad_norm": 1.5104553699493408, + "learning_rate": 1.4499399342585951e-05, + "loss": 0.3945, + "step": 11761 + }, + { + "epoch": 0.3709824948746255, + "grad_norm": 2.065443515777588, + "learning_rate": 1.449848700908273e-05, + "loss": 0.8279, + "step": 11762 + }, + { + "epoch": 0.37101403564106605, + "grad_norm": 1.8817230463027954, + "learning_rate": 1.4497574628634308e-05, + "loss": 0.5274, + "step": 11763 + }, + { + "epoch": 0.3710455764075067, + "grad_norm": 1.7779078483581543, + "learning_rate": 1.4496662201250197e-05, + "loss": 0.515, + "step": 11764 + }, + { + "epoch": 0.3710771171739473, + "grad_norm": 2.0217206478118896, + "learning_rate": 1.4495749726939928e-05, + "loss": 0.4409, + "step": 11765 + }, + { + "epoch": 0.37110865794038794, + "grad_norm": 1.5896872282028198, + "learning_rate": 1.4494837205713019e-05, + "loss": 0.4994, + "step": 11766 + }, + { + "epoch": 0.3711401987068286, + "grad_norm": 1.6699151992797852, + "learning_rate": 1.4493924637578995e-05, + "loss": 0.5222, + "step": 11767 + }, + { + "epoch": 0.3711717394732692, + "grad_norm": 2.328969717025757, + "learning_rate": 1.4493012022547376e-05, + "loss": 0.5072, + "step": 11768 + }, + { + "epoch": 0.37120328023970983, + "grad_norm": 1.6280081272125244, + "learning_rate": 1.449209936062769e-05, + "loss": 0.8338, + "step": 11769 + }, + { + "epoch": 0.37123482100615046, + "grad_norm": 1.5911474227905273, + "learning_rate": 1.4491186651829459e-05, + "loss": 0.5418, + "step": 11770 + }, + { + "epoch": 0.3712663617725911, + "grad_norm": 1.6300197839736938, + "learning_rate": 1.4490273896162205e-05, + "loss": 0.5398, + "step": 11771 + }, + { + "epoch": 0.3712979025390317, + "grad_norm": 1.7864623069763184, + "learning_rate": 1.448936109363546e-05, + "loss": 0.5264, + "step": 11772 + }, + { + "epoch": 0.3713294433054723, + "grad_norm": 1.5664489269256592, + "learning_rate": 1.4488448244258743e-05, + "loss": 0.5184, + "step": 11773 + }, + { + "epoch": 0.3713609840719129, + "grad_norm": 1.5704748630523682, + "learning_rate": 1.4487535348041583e-05, + "loss": 0.5004, + "step": 11774 + }, + { + "epoch": 0.37139252483835355, + "grad_norm": 1.6541852951049805, + "learning_rate": 1.448662240499351e-05, + "loss": 0.5156, + "step": 11775 + }, + { + "epoch": 0.3714240656047942, + "grad_norm": 1.900161623954773, + "learning_rate": 1.4485709415124045e-05, + "loss": 0.5022, + "step": 11776 + }, + { + "epoch": 0.3714556063712348, + "grad_norm": 1.638720154762268, + "learning_rate": 1.4484796378442723e-05, + "loss": 0.8491, + "step": 11777 + }, + { + "epoch": 0.37148714713767544, + "grad_norm": 1.859724998474121, + "learning_rate": 1.4483883294959067e-05, + "loss": 0.411, + "step": 11778 + }, + { + "epoch": 0.3715186879041161, + "grad_norm": 2.0269269943237305, + "learning_rate": 1.4482970164682608e-05, + "loss": 0.5502, + "step": 11779 + }, + { + "epoch": 0.3715502286705567, + "grad_norm": 2.1576945781707764, + "learning_rate": 1.4482056987622874e-05, + "loss": 0.4845, + "step": 11780 + }, + { + "epoch": 0.37158176943699733, + "grad_norm": 1.8589015007019043, + "learning_rate": 1.4481143763789394e-05, + "loss": 0.5424, + "step": 11781 + }, + { + "epoch": 0.37161331020343796, + "grad_norm": 1.740381121635437, + "learning_rate": 1.4480230493191697e-05, + "loss": 0.5133, + "step": 11782 + }, + { + "epoch": 0.3716448509698786, + "grad_norm": 1.7591830492019653, + "learning_rate": 1.447931717583932e-05, + "loss": 0.5431, + "step": 11783 + }, + { + "epoch": 0.37167639173631917, + "grad_norm": 1.7007296085357666, + "learning_rate": 1.447840381174179e-05, + "loss": 0.5489, + "step": 11784 + }, + { + "epoch": 0.3717079325027598, + "grad_norm": 1.682403564453125, + "learning_rate": 1.4477490400908637e-05, + "loss": 0.5294, + "step": 11785 + }, + { + "epoch": 0.3717394732692004, + "grad_norm": 1.4901256561279297, + "learning_rate": 1.4476576943349395e-05, + "loss": 0.8005, + "step": 11786 + }, + { + "epoch": 0.37177101403564106, + "grad_norm": 1.7955067157745361, + "learning_rate": 1.4475663439073597e-05, + "loss": 0.49, + "step": 11787 + }, + { + "epoch": 0.3718025548020817, + "grad_norm": 1.409817099571228, + "learning_rate": 1.4474749888090778e-05, + "loss": 0.4813, + "step": 11788 + }, + { + "epoch": 0.3718340955685223, + "grad_norm": 2.483816623687744, + "learning_rate": 1.4473836290410467e-05, + "loss": 0.5456, + "step": 11789 + }, + { + "epoch": 0.37186563633496295, + "grad_norm": 1.5307906866073608, + "learning_rate": 1.44729226460422e-05, + "loss": 0.5306, + "step": 11790 + }, + { + "epoch": 0.3718971771014036, + "grad_norm": 1.6382888555526733, + "learning_rate": 1.4472008954995516e-05, + "loss": 0.51, + "step": 11791 + }, + { + "epoch": 0.3719287178678442, + "grad_norm": 1.6819430589675903, + "learning_rate": 1.447109521727994e-05, + "loss": 0.5372, + "step": 11792 + }, + { + "epoch": 0.37196025863428483, + "grad_norm": 1.4844703674316406, + "learning_rate": 1.4470181432905018e-05, + "loss": 0.5088, + "step": 11793 + }, + { + "epoch": 0.37199179940072546, + "grad_norm": 3.4865410327911377, + "learning_rate": 1.4469267601880281e-05, + "loss": 0.5137, + "step": 11794 + }, + { + "epoch": 0.37202334016716604, + "grad_norm": 1.4363765716552734, + "learning_rate": 1.4468353724215265e-05, + "loss": 0.5103, + "step": 11795 + }, + { + "epoch": 0.37205488093360667, + "grad_norm": 1.883522868156433, + "learning_rate": 1.4467439799919509e-05, + "loss": 0.5359, + "step": 11796 + }, + { + "epoch": 0.3720864217000473, + "grad_norm": 1.6021040678024292, + "learning_rate": 1.4466525829002551e-05, + "loss": 0.5249, + "step": 11797 + }, + { + "epoch": 0.37211796246648793, + "grad_norm": 1.7038859128952026, + "learning_rate": 1.4465611811473927e-05, + "loss": 0.5162, + "step": 11798 + }, + { + "epoch": 0.37214950323292856, + "grad_norm": 1.6724766492843628, + "learning_rate": 1.4464697747343176e-05, + "loss": 0.5159, + "step": 11799 + }, + { + "epoch": 0.3721810439993692, + "grad_norm": 1.734269380569458, + "learning_rate": 1.4463783636619835e-05, + "loss": 0.49, + "step": 11800 + }, + { + "epoch": 0.3722125847658098, + "grad_norm": 1.6272213459014893, + "learning_rate": 1.4462869479313451e-05, + "loss": 0.5482, + "step": 11801 + }, + { + "epoch": 0.37224412553225045, + "grad_norm": 1.5498113632202148, + "learning_rate": 1.4461955275433554e-05, + "loss": 0.5217, + "step": 11802 + }, + { + "epoch": 0.3722756662986911, + "grad_norm": 1.683668851852417, + "learning_rate": 1.4461041024989691e-05, + "loss": 0.5211, + "step": 11803 + }, + { + "epoch": 0.3723072070651317, + "grad_norm": 1.603402018547058, + "learning_rate": 1.4460126727991401e-05, + "loss": 0.5654, + "step": 11804 + }, + { + "epoch": 0.3723387478315723, + "grad_norm": 1.7272921800613403, + "learning_rate": 1.4459212384448224e-05, + "loss": 0.5403, + "step": 11805 + }, + { + "epoch": 0.3723702885980129, + "grad_norm": 1.9945775270462036, + "learning_rate": 1.4458297994369704e-05, + "loss": 0.3768, + "step": 11806 + }, + { + "epoch": 0.37240182936445354, + "grad_norm": 1.9312989711761475, + "learning_rate": 1.4457383557765385e-05, + "loss": 0.4964, + "step": 11807 + }, + { + "epoch": 0.37243337013089417, + "grad_norm": 2.928967237472534, + "learning_rate": 1.4456469074644805e-05, + "loss": 0.4978, + "step": 11808 + }, + { + "epoch": 0.3724649108973348, + "grad_norm": 1.377001166343689, + "learning_rate": 1.4455554545017512e-05, + "loss": 0.5384, + "step": 11809 + }, + { + "epoch": 0.37249645166377543, + "grad_norm": 1.7591631412506104, + "learning_rate": 1.4454639968893045e-05, + "loss": 0.5098, + "step": 11810 + }, + { + "epoch": 0.37252799243021606, + "grad_norm": 1.8069249391555786, + "learning_rate": 1.4453725346280952e-05, + "loss": 0.5109, + "step": 11811 + }, + { + "epoch": 0.3725595331966567, + "grad_norm": 1.691007375717163, + "learning_rate": 1.4452810677190777e-05, + "loss": 0.5098, + "step": 11812 + }, + { + "epoch": 0.3725910739630973, + "grad_norm": 1.842098355293274, + "learning_rate": 1.4451895961632062e-05, + "loss": 0.54, + "step": 11813 + }, + { + "epoch": 0.37262261472953795, + "grad_norm": 1.7842973470687866, + "learning_rate": 1.445098119961436e-05, + "loss": 0.5129, + "step": 11814 + }, + { + "epoch": 0.3726541554959786, + "grad_norm": 2.213731288909912, + "learning_rate": 1.4450066391147212e-05, + "loss": 0.4965, + "step": 11815 + }, + { + "epoch": 0.37268569626241915, + "grad_norm": 1.2646323442459106, + "learning_rate": 1.4449151536240167e-05, + "loss": 0.4742, + "step": 11816 + }, + { + "epoch": 0.3727172370288598, + "grad_norm": 1.9002854824066162, + "learning_rate": 1.444823663490277e-05, + "loss": 0.5054, + "step": 11817 + }, + { + "epoch": 0.3727487777953004, + "grad_norm": 1.5771914720535278, + "learning_rate": 1.4447321687144571e-05, + "loss": 0.4903, + "step": 11818 + }, + { + "epoch": 0.37278031856174104, + "grad_norm": 1.5664458274841309, + "learning_rate": 1.4446406692975114e-05, + "loss": 0.8525, + "step": 11819 + }, + { + "epoch": 0.37281185932818167, + "grad_norm": 2.320594072341919, + "learning_rate": 1.444549165240395e-05, + "loss": 0.4932, + "step": 11820 + }, + { + "epoch": 0.3728434000946223, + "grad_norm": 1.6535075902938843, + "learning_rate": 1.4444576565440632e-05, + "loss": 0.511, + "step": 11821 + }, + { + "epoch": 0.37287494086106293, + "grad_norm": 1.777086615562439, + "learning_rate": 1.4443661432094707e-05, + "loss": 0.548, + "step": 11822 + }, + { + "epoch": 0.37290648162750356, + "grad_norm": 1.9939919710159302, + "learning_rate": 1.444274625237572e-05, + "loss": 0.7992, + "step": 11823 + }, + { + "epoch": 0.3729380223939442, + "grad_norm": 1.465320348739624, + "learning_rate": 1.444183102629323e-05, + "loss": 0.5301, + "step": 11824 + }, + { + "epoch": 0.3729695631603848, + "grad_norm": 2.2463669776916504, + "learning_rate": 1.4440915753856784e-05, + "loss": 0.5127, + "step": 11825 + }, + { + "epoch": 0.3730011039268254, + "grad_norm": 1.7449111938476562, + "learning_rate": 1.4440000435075931e-05, + "loss": 0.4993, + "step": 11826 + }, + { + "epoch": 0.373032644693266, + "grad_norm": 1.9416886568069458, + "learning_rate": 1.443908506996023e-05, + "loss": 0.529, + "step": 11827 + }, + { + "epoch": 0.37306418545970665, + "grad_norm": 1.8358150720596313, + "learning_rate": 1.4438169658519228e-05, + "loss": 0.8212, + "step": 11828 + }, + { + "epoch": 0.3730957262261473, + "grad_norm": 1.6231826543807983, + "learning_rate": 1.4437254200762478e-05, + "loss": 0.8262, + "step": 11829 + }, + { + "epoch": 0.3731272669925879, + "grad_norm": 1.8062793016433716, + "learning_rate": 1.4436338696699534e-05, + "loss": 0.5228, + "step": 11830 + }, + { + "epoch": 0.37315880775902854, + "grad_norm": 1.9991447925567627, + "learning_rate": 1.4435423146339953e-05, + "loss": 0.5489, + "step": 11831 + }, + { + "epoch": 0.3731903485254692, + "grad_norm": 2.7113492488861084, + "learning_rate": 1.4434507549693288e-05, + "loss": 0.5519, + "step": 11832 + }, + { + "epoch": 0.3732218892919098, + "grad_norm": 2.3868813514709473, + "learning_rate": 1.4433591906769091e-05, + "loss": 0.4157, + "step": 11833 + }, + { + "epoch": 0.37325343005835043, + "grad_norm": 1.8939529657363892, + "learning_rate": 1.4432676217576922e-05, + "loss": 0.5019, + "step": 11834 + }, + { + "epoch": 0.37328497082479106, + "grad_norm": 2.2724809646606445, + "learning_rate": 1.4431760482126334e-05, + "loss": 0.5303, + "step": 11835 + }, + { + "epoch": 0.3733165115912317, + "grad_norm": 2.4931671619415283, + "learning_rate": 1.4430844700426885e-05, + "loss": 0.521, + "step": 11836 + }, + { + "epoch": 0.37334805235767227, + "grad_norm": 1.8468170166015625, + "learning_rate": 1.4429928872488131e-05, + "loss": 0.5445, + "step": 11837 + }, + { + "epoch": 0.3733795931241129, + "grad_norm": 2.0307343006134033, + "learning_rate": 1.442901299831963e-05, + "loss": 0.5285, + "step": 11838 + }, + { + "epoch": 0.3734111338905535, + "grad_norm": 1.697882056236267, + "learning_rate": 1.4428097077930938e-05, + "loss": 0.5229, + "step": 11839 + }, + { + "epoch": 0.37344267465699416, + "grad_norm": 1.4208248853683472, + "learning_rate": 1.4427181111331615e-05, + "loss": 0.5307, + "step": 11840 + }, + { + "epoch": 0.3734742154234348, + "grad_norm": 1.6153736114501953, + "learning_rate": 1.4426265098531218e-05, + "loss": 0.5348, + "step": 11841 + }, + { + "epoch": 0.3735057561898754, + "grad_norm": 1.4473382234573364, + "learning_rate": 1.4425349039539311e-05, + "loss": 0.5016, + "step": 11842 + }, + { + "epoch": 0.37353729695631605, + "grad_norm": 1.991956353187561, + "learning_rate": 1.4424432934365447e-05, + "loss": 0.5348, + "step": 11843 + }, + { + "epoch": 0.3735688377227567, + "grad_norm": 2.064678907394409, + "learning_rate": 1.4423516783019192e-05, + "loss": 0.4645, + "step": 11844 + }, + { + "epoch": 0.3736003784891973, + "grad_norm": 1.4675230979919434, + "learning_rate": 1.4422600585510103e-05, + "loss": 0.5085, + "step": 11845 + }, + { + "epoch": 0.37363191925563793, + "grad_norm": 1.7739622592926025, + "learning_rate": 1.4421684341847746e-05, + "loss": 0.5214, + "step": 11846 + }, + { + "epoch": 0.37366346002207856, + "grad_norm": 3.120745897293091, + "learning_rate": 1.4420768052041679e-05, + "loss": 0.4963, + "step": 11847 + }, + { + "epoch": 0.37369500078851914, + "grad_norm": 1.8011953830718994, + "learning_rate": 1.441985171610146e-05, + "loss": 0.5432, + "step": 11848 + }, + { + "epoch": 0.37372654155495977, + "grad_norm": 1.3616279363632202, + "learning_rate": 1.4418935334036661e-05, + "loss": 0.4954, + "step": 11849 + }, + { + "epoch": 0.3737580823214004, + "grad_norm": 2.0866169929504395, + "learning_rate": 1.4418018905856838e-05, + "loss": 0.5097, + "step": 11850 + }, + { + "epoch": 0.373789623087841, + "grad_norm": 2.0858571529388428, + "learning_rate": 1.4417102431571559e-05, + "loss": 0.5099, + "step": 11851 + }, + { + "epoch": 0.37382116385428166, + "grad_norm": 1.576167106628418, + "learning_rate": 1.4416185911190387e-05, + "loss": 0.4875, + "step": 11852 + }, + { + "epoch": 0.3738527046207223, + "grad_norm": 1.622829794883728, + "learning_rate": 1.4415269344722885e-05, + "loss": 0.5192, + "step": 11853 + }, + { + "epoch": 0.3738842453871629, + "grad_norm": 1.6480661630630493, + "learning_rate": 1.4414352732178618e-05, + "loss": 0.5327, + "step": 11854 + }, + { + "epoch": 0.37391578615360355, + "grad_norm": 1.821798324584961, + "learning_rate": 1.4413436073567153e-05, + "loss": 0.5152, + "step": 11855 + }, + { + "epoch": 0.3739473269200442, + "grad_norm": 1.3143000602722168, + "learning_rate": 1.4412519368898056e-05, + "loss": 0.5004, + "step": 11856 + }, + { + "epoch": 0.3739788676864848, + "grad_norm": 1.374544620513916, + "learning_rate": 1.4411602618180896e-05, + "loss": 0.5246, + "step": 11857 + }, + { + "epoch": 0.3740104084529254, + "grad_norm": 1.5367790460586548, + "learning_rate": 1.4410685821425232e-05, + "loss": 0.4854, + "step": 11858 + }, + { + "epoch": 0.374041949219366, + "grad_norm": 1.567171335220337, + "learning_rate": 1.4409768978640642e-05, + "loss": 0.4766, + "step": 11859 + }, + { + "epoch": 0.37407348998580664, + "grad_norm": 1.4757529497146606, + "learning_rate": 1.4408852089836684e-05, + "loss": 0.5167, + "step": 11860 + }, + { + "epoch": 0.37410503075224727, + "grad_norm": 1.579651117324829, + "learning_rate": 1.4407935155022932e-05, + "loss": 0.5137, + "step": 11861 + }, + { + "epoch": 0.3741365715186879, + "grad_norm": 1.8893094062805176, + "learning_rate": 1.4407018174208953e-05, + "loss": 0.505, + "step": 11862 + }, + { + "epoch": 0.37416811228512853, + "grad_norm": 1.776879072189331, + "learning_rate": 1.4406101147404321e-05, + "loss": 0.5447, + "step": 11863 + }, + { + "epoch": 0.37419965305156916, + "grad_norm": 2.5915889739990234, + "learning_rate": 1.4405184074618599e-05, + "loss": 0.519, + "step": 11864 + }, + { + "epoch": 0.3742311938180098, + "grad_norm": 1.7711583375930786, + "learning_rate": 1.440426695586136e-05, + "loss": 0.4944, + "step": 11865 + }, + { + "epoch": 0.3742627345844504, + "grad_norm": 2.7200767993927, + "learning_rate": 1.440334979114218e-05, + "loss": 0.4167, + "step": 11866 + }, + { + "epoch": 0.37429427535089105, + "grad_norm": 2.210369348526001, + "learning_rate": 1.4402432580470624e-05, + "loss": 0.5293, + "step": 11867 + }, + { + "epoch": 0.3743258161173317, + "grad_norm": 4.451773643493652, + "learning_rate": 1.4401515323856266e-05, + "loss": 0.4669, + "step": 11868 + }, + { + "epoch": 0.37435735688377225, + "grad_norm": 1.336647868156433, + "learning_rate": 1.4400598021308674e-05, + "loss": 0.5098, + "step": 11869 + }, + { + "epoch": 0.3743888976502129, + "grad_norm": 1.5436835289001465, + "learning_rate": 1.439968067283743e-05, + "loss": 0.5336, + "step": 11870 + }, + { + "epoch": 0.3744204384166535, + "grad_norm": 1.8514162302017212, + "learning_rate": 1.4398763278452099e-05, + "loss": 0.5341, + "step": 11871 + }, + { + "epoch": 0.37445197918309414, + "grad_norm": 1.4521898031234741, + "learning_rate": 1.4397845838162257e-05, + "loss": 0.354, + "step": 11872 + }, + { + "epoch": 0.37448351994953477, + "grad_norm": 1.6506139039993286, + "learning_rate": 1.439692835197748e-05, + "loss": 0.5281, + "step": 11873 + }, + { + "epoch": 0.3745150607159754, + "grad_norm": 1.5844498872756958, + "learning_rate": 1.439601081990734e-05, + "loss": 0.4281, + "step": 11874 + }, + { + "epoch": 0.37454660148241603, + "grad_norm": 1.419861078262329, + "learning_rate": 1.4395093241961415e-05, + "loss": 0.4974, + "step": 11875 + }, + { + "epoch": 0.37457814224885666, + "grad_norm": 2.1101415157318115, + "learning_rate": 1.4394175618149282e-05, + "loss": 0.8522, + "step": 11876 + }, + { + "epoch": 0.3746096830152973, + "grad_norm": 1.6345648765563965, + "learning_rate": 1.4393257948480512e-05, + "loss": 0.5168, + "step": 11877 + }, + { + "epoch": 0.3746412237817379, + "grad_norm": 1.3111662864685059, + "learning_rate": 1.4392340232964683e-05, + "loss": 0.5082, + "step": 11878 + }, + { + "epoch": 0.3746727645481785, + "grad_norm": 1.5906808376312256, + "learning_rate": 1.4391422471611373e-05, + "loss": 0.4638, + "step": 11879 + }, + { + "epoch": 0.3747043053146191, + "grad_norm": 1.800207495689392, + "learning_rate": 1.4390504664430162e-05, + "loss": 0.527, + "step": 11880 + }, + { + "epoch": 0.37473584608105975, + "grad_norm": 2.198791742324829, + "learning_rate": 1.4389586811430625e-05, + "loss": 0.5398, + "step": 11881 + }, + { + "epoch": 0.3747673868475004, + "grad_norm": 9.901216506958008, + "learning_rate": 1.438866891262234e-05, + "loss": 0.807, + "step": 11882 + }, + { + "epoch": 0.374798927613941, + "grad_norm": 1.9075208902359009, + "learning_rate": 1.4387750968014888e-05, + "loss": 0.5083, + "step": 11883 + }, + { + "epoch": 0.37483046838038164, + "grad_norm": 1.7250840663909912, + "learning_rate": 1.4386832977617847e-05, + "loss": 0.5386, + "step": 11884 + }, + { + "epoch": 0.3748620091468223, + "grad_norm": 1.8816819190979004, + "learning_rate": 1.4385914941440797e-05, + "loss": 0.5055, + "step": 11885 + }, + { + "epoch": 0.3748935499132629, + "grad_norm": 1.6602195501327515, + "learning_rate": 1.4384996859493321e-05, + "loss": 0.5204, + "step": 11886 + }, + { + "epoch": 0.37492509067970353, + "grad_norm": 1.703078031539917, + "learning_rate": 1.4384078731784999e-05, + "loss": 0.5085, + "step": 11887 + }, + { + "epoch": 0.37495663144614416, + "grad_norm": 1.5056555271148682, + "learning_rate": 1.4383160558325406e-05, + "loss": 0.8256, + "step": 11888 + }, + { + "epoch": 0.3749881722125848, + "grad_norm": 1.26071298122406, + "learning_rate": 1.4382242339124132e-05, + "loss": 0.5428, + "step": 11889 + }, + { + "epoch": 0.37501971297902537, + "grad_norm": 1.7773531675338745, + "learning_rate": 1.4381324074190758e-05, + "loss": 0.5163, + "step": 11890 + }, + { + "epoch": 0.375051253745466, + "grad_norm": 1.3683977127075195, + "learning_rate": 1.4380405763534864e-05, + "loss": 0.5179, + "step": 11891 + }, + { + "epoch": 0.3750827945119066, + "grad_norm": 1.5085846185684204, + "learning_rate": 1.4379487407166034e-05, + "loss": 0.5046, + "step": 11892 + }, + { + "epoch": 0.37511433527834726, + "grad_norm": 1.6888223886489868, + "learning_rate": 1.4378569005093851e-05, + "loss": 0.5334, + "step": 11893 + }, + { + "epoch": 0.3751458760447879, + "grad_norm": 1.7868858575820923, + "learning_rate": 1.4377650557327904e-05, + "loss": 0.8165, + "step": 11894 + }, + { + "epoch": 0.3751774168112285, + "grad_norm": 1.4728319644927979, + "learning_rate": 1.4376732063877772e-05, + "loss": 0.4993, + "step": 11895 + }, + { + "epoch": 0.37520895757766914, + "grad_norm": 1.6453287601470947, + "learning_rate": 1.4375813524753044e-05, + "loss": 0.4969, + "step": 11896 + }, + { + "epoch": 0.3752404983441098, + "grad_norm": 1.5853607654571533, + "learning_rate": 1.4374894939963304e-05, + "loss": 0.5277, + "step": 11897 + }, + { + "epoch": 0.3752720391105504, + "grad_norm": 1.877909541130066, + "learning_rate": 1.4373976309518139e-05, + "loss": 0.8385, + "step": 11898 + }, + { + "epoch": 0.37530357987699103, + "grad_norm": 2.420689821243286, + "learning_rate": 1.4373057633427132e-05, + "loss": 0.513, + "step": 11899 + }, + { + "epoch": 0.3753351206434316, + "grad_norm": 2.136885166168213, + "learning_rate": 1.4372138911699876e-05, + "loss": 0.8161, + "step": 11900 + }, + { + "epoch": 0.37536666140987224, + "grad_norm": 1.4709835052490234, + "learning_rate": 1.4371220144345954e-05, + "loss": 0.5129, + "step": 11901 + }, + { + "epoch": 0.37539820217631287, + "grad_norm": 1.4651187658309937, + "learning_rate": 1.4370301331374954e-05, + "loss": 0.5174, + "step": 11902 + }, + { + "epoch": 0.3754297429427535, + "grad_norm": 1.5142813920974731, + "learning_rate": 1.4369382472796467e-05, + "loss": 0.4145, + "step": 11903 + }, + { + "epoch": 0.3754612837091941, + "grad_norm": 1.3477901220321655, + "learning_rate": 1.4368463568620084e-05, + "loss": 0.5609, + "step": 11904 + }, + { + "epoch": 0.37549282447563476, + "grad_norm": 1.456714153289795, + "learning_rate": 1.436754461885539e-05, + "loss": 0.524, + "step": 11905 + }, + { + "epoch": 0.3755243652420754, + "grad_norm": 1.452974557876587, + "learning_rate": 1.4366625623511976e-05, + "loss": 0.4924, + "step": 11906 + }, + { + "epoch": 0.375555906008516, + "grad_norm": 1.8749229907989502, + "learning_rate": 1.4365706582599434e-05, + "loss": 0.5342, + "step": 11907 + }, + { + "epoch": 0.37558744677495665, + "grad_norm": 2.7266156673431396, + "learning_rate": 1.4364787496127353e-05, + "loss": 0.5028, + "step": 11908 + }, + { + "epoch": 0.3756189875413973, + "grad_norm": 1.4182541370391846, + "learning_rate": 1.4363868364105324e-05, + "loss": 0.518, + "step": 11909 + }, + { + "epoch": 0.3756505283078379, + "grad_norm": 1.5776221752166748, + "learning_rate": 1.4362949186542942e-05, + "loss": 0.5346, + "step": 11910 + }, + { + "epoch": 0.3756820690742785, + "grad_norm": 1.6743534803390503, + "learning_rate": 1.4362029963449798e-05, + "loss": 0.5185, + "step": 11911 + }, + { + "epoch": 0.3757136098407191, + "grad_norm": 2.5612781047821045, + "learning_rate": 1.4361110694835483e-05, + "loss": 0.498, + "step": 11912 + }, + { + "epoch": 0.37574515060715974, + "grad_norm": 1.5132145881652832, + "learning_rate": 1.4360191380709592e-05, + "loss": 0.4834, + "step": 11913 + }, + { + "epoch": 0.37577669137360037, + "grad_norm": 1.4126917123794556, + "learning_rate": 1.4359272021081718e-05, + "loss": 0.4876, + "step": 11914 + }, + { + "epoch": 0.375808232140041, + "grad_norm": 1.9430047273635864, + "learning_rate": 1.4358352615961458e-05, + "loss": 0.8596, + "step": 11915 + }, + { + "epoch": 0.37583977290648163, + "grad_norm": 1.7131938934326172, + "learning_rate": 1.4357433165358402e-05, + "loss": 0.5015, + "step": 11916 + }, + { + "epoch": 0.37587131367292226, + "grad_norm": 1.6119873523712158, + "learning_rate": 1.4356513669282146e-05, + "loss": 0.5481, + "step": 11917 + }, + { + "epoch": 0.3759028544393629, + "grad_norm": 1.6125231981277466, + "learning_rate": 1.4355594127742291e-05, + "loss": 0.5187, + "step": 11918 + }, + { + "epoch": 0.3759343952058035, + "grad_norm": 1.5284793376922607, + "learning_rate": 1.4354674540748426e-05, + "loss": 0.8129, + "step": 11919 + }, + { + "epoch": 0.37596593597224415, + "grad_norm": 1.2937943935394287, + "learning_rate": 1.435375490831015e-05, + "loss": 0.5108, + "step": 11920 + }, + { + "epoch": 0.3759974767386848, + "grad_norm": 1.9849879741668701, + "learning_rate": 1.4352835230437065e-05, + "loss": 0.3683, + "step": 11921 + }, + { + "epoch": 0.37602901750512535, + "grad_norm": 2.228902578353882, + "learning_rate": 1.4351915507138761e-05, + "loss": 0.5129, + "step": 11922 + }, + { + "epoch": 0.376060558271566, + "grad_norm": 1.2467548847198486, + "learning_rate": 1.435099573842484e-05, + "loss": 0.3864, + "step": 11923 + }, + { + "epoch": 0.3760920990380066, + "grad_norm": 1.48666512966156, + "learning_rate": 1.4350075924304903e-05, + "loss": 0.5019, + "step": 11924 + }, + { + "epoch": 0.37612363980444724, + "grad_norm": 4.189208030700684, + "learning_rate": 1.4349156064788543e-05, + "loss": 0.8563, + "step": 11925 + }, + { + "epoch": 0.37615518057088787, + "grad_norm": 1.582457423210144, + "learning_rate": 1.4348236159885364e-05, + "loss": 0.4894, + "step": 11926 + }, + { + "epoch": 0.3761867213373285, + "grad_norm": 1.8026195764541626, + "learning_rate": 1.4347316209604962e-05, + "loss": 0.5133, + "step": 11927 + }, + { + "epoch": 0.37621826210376913, + "grad_norm": 1.447094440460205, + "learning_rate": 1.4346396213956944e-05, + "loss": 0.4961, + "step": 11928 + }, + { + "epoch": 0.37624980287020976, + "grad_norm": 1.366905927658081, + "learning_rate": 1.4345476172950903e-05, + "loss": 0.4896, + "step": 11929 + }, + { + "epoch": 0.3762813436366504, + "grad_norm": 1.8490338325500488, + "learning_rate": 1.4344556086596443e-05, + "loss": 0.8595, + "step": 11930 + }, + { + "epoch": 0.376312884403091, + "grad_norm": 1.4724565744400024, + "learning_rate": 1.4343635954903171e-05, + "loss": 0.5097, + "step": 11931 + }, + { + "epoch": 0.3763444251695316, + "grad_norm": 1.7044481039047241, + "learning_rate": 1.4342715777880683e-05, + "loss": 0.5034, + "step": 11932 + }, + { + "epoch": 0.3763759659359722, + "grad_norm": 1.5718655586242676, + "learning_rate": 1.4341795555538583e-05, + "loss": 0.5278, + "step": 11933 + }, + { + "epoch": 0.37640750670241285, + "grad_norm": 1.492850422859192, + "learning_rate": 1.4340875287886475e-05, + "loss": 0.5308, + "step": 11934 + }, + { + "epoch": 0.3764390474688535, + "grad_norm": 1.5356724262237549, + "learning_rate": 1.4339954974933963e-05, + "loss": 0.8181, + "step": 11935 + }, + { + "epoch": 0.3764705882352941, + "grad_norm": 1.6045091152191162, + "learning_rate": 1.4339034616690653e-05, + "loss": 0.5193, + "step": 11936 + }, + { + "epoch": 0.37650212900173474, + "grad_norm": 1.5756524801254272, + "learning_rate": 1.4338114213166145e-05, + "loss": 0.807, + "step": 11937 + }, + { + "epoch": 0.37653366976817537, + "grad_norm": 1.903879165649414, + "learning_rate": 1.433719376437005e-05, + "loss": 0.5311, + "step": 11938 + }, + { + "epoch": 0.376565210534616, + "grad_norm": 2.0543177127838135, + "learning_rate": 1.4336273270311967e-05, + "loss": 0.4871, + "step": 11939 + }, + { + "epoch": 0.37659675130105663, + "grad_norm": 1.537170648574829, + "learning_rate": 1.4335352731001506e-05, + "loss": 0.5341, + "step": 11940 + }, + { + "epoch": 0.37662829206749726, + "grad_norm": 1.7071614265441895, + "learning_rate": 1.4334432146448272e-05, + "loss": 0.5724, + "step": 11941 + }, + { + "epoch": 0.3766598328339379, + "grad_norm": 1.9137014150619507, + "learning_rate": 1.4333511516661876e-05, + "loss": 0.5036, + "step": 11942 + }, + { + "epoch": 0.37669137360037847, + "grad_norm": 1.3750683069229126, + "learning_rate": 1.433259084165192e-05, + "loss": 0.814, + "step": 11943 + }, + { + "epoch": 0.3767229143668191, + "grad_norm": 1.951766848564148, + "learning_rate": 1.4331670121428016e-05, + "loss": 0.525, + "step": 11944 + }, + { + "epoch": 0.3767544551332597, + "grad_norm": 2.4835824966430664, + "learning_rate": 1.433074935599977e-05, + "loss": 0.5422, + "step": 11945 + }, + { + "epoch": 0.37678599589970035, + "grad_norm": 1.5268712043762207, + "learning_rate": 1.4329828545376794e-05, + "loss": 0.4878, + "step": 11946 + }, + { + "epoch": 0.376817536666141, + "grad_norm": 1.3713968992233276, + "learning_rate": 1.4328907689568689e-05, + "loss": 0.807, + "step": 11947 + }, + { + "epoch": 0.3768490774325816, + "grad_norm": 1.9985698461532593, + "learning_rate": 1.4327986788585078e-05, + "loss": 0.8279, + "step": 11948 + }, + { + "epoch": 0.37688061819902224, + "grad_norm": 1.8359769582748413, + "learning_rate": 1.4327065842435559e-05, + "loss": 0.4115, + "step": 11949 + }, + { + "epoch": 0.3769121589654629, + "grad_norm": 1.3101385831832886, + "learning_rate": 1.4326144851129751e-05, + "loss": 0.8315, + "step": 11950 + }, + { + "epoch": 0.3769436997319035, + "grad_norm": 1.6576141119003296, + "learning_rate": 1.432522381467726e-05, + "loss": 0.5508, + "step": 11951 + }, + { + "epoch": 0.37697524049834413, + "grad_norm": 2.1144909858703613, + "learning_rate": 1.4324302733087703e-05, + "loss": 0.4678, + "step": 11952 + }, + { + "epoch": 0.3770067812647847, + "grad_norm": 1.6406959295272827, + "learning_rate": 1.4323381606370687e-05, + "loss": 0.5157, + "step": 11953 + }, + { + "epoch": 0.37703832203122534, + "grad_norm": 2.2995293140411377, + "learning_rate": 1.4322460434535828e-05, + "loss": 0.5019, + "step": 11954 + }, + { + "epoch": 0.37706986279766597, + "grad_norm": 1.7158679962158203, + "learning_rate": 1.4321539217592737e-05, + "loss": 0.5609, + "step": 11955 + }, + { + "epoch": 0.3771014035641066, + "grad_norm": 1.4080612659454346, + "learning_rate": 1.432061795555103e-05, + "loss": 0.5214, + "step": 11956 + }, + { + "epoch": 0.3771329443305472, + "grad_norm": 2.1150741577148438, + "learning_rate": 1.4319696648420318e-05, + "loss": 0.5101, + "step": 11957 + }, + { + "epoch": 0.37716448509698786, + "grad_norm": 2.088914394378662, + "learning_rate": 1.431877529621022e-05, + "loss": 0.8484, + "step": 11958 + }, + { + "epoch": 0.3771960258634285, + "grad_norm": 1.5122628211975098, + "learning_rate": 1.4317853898930348e-05, + "loss": 0.8279, + "step": 11959 + }, + { + "epoch": 0.3772275666298691, + "grad_norm": 2.4915542602539062, + "learning_rate": 1.4316932456590316e-05, + "loss": 0.4831, + "step": 11960 + }, + { + "epoch": 0.37725910739630975, + "grad_norm": 1.9734126329421997, + "learning_rate": 1.431601096919974e-05, + "loss": 0.519, + "step": 11961 + }, + { + "epoch": 0.3772906481627504, + "grad_norm": 2.274028778076172, + "learning_rate": 1.4315089436768244e-05, + "loss": 0.5388, + "step": 11962 + }, + { + "epoch": 0.377322188929191, + "grad_norm": 1.5603975057601929, + "learning_rate": 1.4314167859305434e-05, + "loss": 0.8127, + "step": 11963 + }, + { + "epoch": 0.3773537296956316, + "grad_norm": 2.0467090606689453, + "learning_rate": 1.4313246236820935e-05, + "loss": 0.5361, + "step": 11964 + }, + { + "epoch": 0.3773852704620722, + "grad_norm": 1.758751630783081, + "learning_rate": 1.431232456932436e-05, + "loss": 0.5304, + "step": 11965 + }, + { + "epoch": 0.37741681122851284, + "grad_norm": 1.6593825817108154, + "learning_rate": 1.4311402856825335e-05, + "loss": 0.4992, + "step": 11966 + }, + { + "epoch": 0.37744835199495347, + "grad_norm": 1.8782920837402344, + "learning_rate": 1.4310481099333469e-05, + "loss": 0.4106, + "step": 11967 + }, + { + "epoch": 0.3774798927613941, + "grad_norm": 2.0615148544311523, + "learning_rate": 1.4309559296858384e-05, + "loss": 0.5224, + "step": 11968 + }, + { + "epoch": 0.3775114335278347, + "grad_norm": 3.7631092071533203, + "learning_rate": 1.4308637449409705e-05, + "loss": 0.4951, + "step": 11969 + }, + { + "epoch": 0.37754297429427536, + "grad_norm": 1.6393808126449585, + "learning_rate": 1.4307715556997048e-05, + "loss": 0.4861, + "step": 11970 + }, + { + "epoch": 0.377574515060716, + "grad_norm": 2.0672850608825684, + "learning_rate": 1.4306793619630033e-05, + "loss": 0.5021, + "step": 11971 + }, + { + "epoch": 0.3776060558271566, + "grad_norm": 2.2193410396575928, + "learning_rate": 1.4305871637318285e-05, + "loss": 0.5316, + "step": 11972 + }, + { + "epoch": 0.37763759659359725, + "grad_norm": 1.6493287086486816, + "learning_rate": 1.4304949610071421e-05, + "loss": 0.4854, + "step": 11973 + }, + { + "epoch": 0.3776691373600379, + "grad_norm": 1.4291660785675049, + "learning_rate": 1.4304027537899065e-05, + "loss": 0.8069, + "step": 11974 + }, + { + "epoch": 0.37770067812647845, + "grad_norm": 2.1385841369628906, + "learning_rate": 1.4303105420810841e-05, + "loss": 0.4854, + "step": 11975 + }, + { + "epoch": 0.3777322188929191, + "grad_norm": 2.4752440452575684, + "learning_rate": 1.430218325881637e-05, + "loss": 0.527, + "step": 11976 + }, + { + "epoch": 0.3777637596593597, + "grad_norm": 1.983851432800293, + "learning_rate": 1.4301261051925278e-05, + "loss": 0.4992, + "step": 11977 + }, + { + "epoch": 0.37779530042580034, + "grad_norm": 1.9758085012435913, + "learning_rate": 1.4300338800147182e-05, + "loss": 0.5207, + "step": 11978 + }, + { + "epoch": 0.37782684119224097, + "grad_norm": 1.2733044624328613, + "learning_rate": 1.4299416503491716e-05, + "loss": 0.4978, + "step": 11979 + }, + { + "epoch": 0.3778583819586816, + "grad_norm": 1.682163119316101, + "learning_rate": 1.4298494161968499e-05, + "loss": 0.4855, + "step": 11980 + }, + { + "epoch": 0.37788992272512223, + "grad_norm": 1.9034022092819214, + "learning_rate": 1.4297571775587157e-05, + "loss": 0.5422, + "step": 11981 + }, + { + "epoch": 0.37792146349156286, + "grad_norm": 1.603746771812439, + "learning_rate": 1.4296649344357317e-05, + "loss": 0.5366, + "step": 11982 + }, + { + "epoch": 0.3779530042580035, + "grad_norm": 1.532109022140503, + "learning_rate": 1.4295726868288604e-05, + "loss": 0.5007, + "step": 11983 + }, + { + "epoch": 0.3779845450244441, + "grad_norm": 1.8045309782028198, + "learning_rate": 1.429480434739065e-05, + "loss": 0.4889, + "step": 11984 + }, + { + "epoch": 0.3780160857908847, + "grad_norm": 1.7451345920562744, + "learning_rate": 1.4293881781673073e-05, + "loss": 0.5223, + "step": 11985 + }, + { + "epoch": 0.3780476265573253, + "grad_norm": 1.6702314615249634, + "learning_rate": 1.4292959171145509e-05, + "loss": 0.5065, + "step": 11986 + }, + { + "epoch": 0.37807916732376595, + "grad_norm": 7.122173309326172, + "learning_rate": 1.4292036515817578e-05, + "loss": 0.5133, + "step": 11987 + }, + { + "epoch": 0.3781107080902066, + "grad_norm": 1.5931003093719482, + "learning_rate": 1.4291113815698917e-05, + "loss": 0.8134, + "step": 11988 + }, + { + "epoch": 0.3781422488566472, + "grad_norm": 1.438891053199768, + "learning_rate": 1.4290191070799148e-05, + "loss": 0.4935, + "step": 11989 + }, + { + "epoch": 0.37817378962308784, + "grad_norm": 2.112243175506592, + "learning_rate": 1.4289268281127909e-05, + "loss": 0.805, + "step": 11990 + }, + { + "epoch": 0.37820533038952847, + "grad_norm": 1.919837474822998, + "learning_rate": 1.428834544669482e-05, + "loss": 0.5115, + "step": 11991 + }, + { + "epoch": 0.3782368711559691, + "grad_norm": 1.7471545934677124, + "learning_rate": 1.4287422567509518e-05, + "loss": 0.5335, + "step": 11992 + }, + { + "epoch": 0.37826841192240973, + "grad_norm": 3.2118964195251465, + "learning_rate": 1.4286499643581636e-05, + "loss": 0.4971, + "step": 11993 + }, + { + "epoch": 0.37829995268885036, + "grad_norm": 1.5466526746749878, + "learning_rate": 1.4285576674920798e-05, + "loss": 0.4892, + "step": 11994 + }, + { + "epoch": 0.378331493455291, + "grad_norm": 1.6303346157073975, + "learning_rate": 1.428465366153664e-05, + "loss": 0.5054, + "step": 11995 + }, + { + "epoch": 0.37836303422173156, + "grad_norm": 2.2908506393432617, + "learning_rate": 1.4283730603438795e-05, + "loss": 0.5126, + "step": 11996 + }, + { + "epoch": 0.3783945749881722, + "grad_norm": 1.6254817247390747, + "learning_rate": 1.4282807500636895e-05, + "loss": 0.5255, + "step": 11997 + }, + { + "epoch": 0.3784261157546128, + "grad_norm": 1.855526328086853, + "learning_rate": 1.4281884353140571e-05, + "loss": 0.4957, + "step": 11998 + }, + { + "epoch": 0.37845765652105345, + "grad_norm": 2.0366179943084717, + "learning_rate": 1.4280961160959461e-05, + "loss": 0.5228, + "step": 11999 + }, + { + "epoch": 0.3784891972874941, + "grad_norm": 1.931212306022644, + "learning_rate": 1.4280037924103198e-05, + "loss": 0.5143, + "step": 12000 + }, + { + "epoch": 0.3785207380539347, + "grad_norm": 2.059412717819214, + "learning_rate": 1.4279114642581415e-05, + "loss": 0.5436, + "step": 12001 + }, + { + "epoch": 0.37855227882037534, + "grad_norm": 1.4441049098968506, + "learning_rate": 1.4278191316403747e-05, + "loss": 0.5206, + "step": 12002 + }, + { + "epoch": 0.378583819586816, + "grad_norm": 1.400641679763794, + "learning_rate": 1.4277267945579833e-05, + "loss": 0.5322, + "step": 12003 + }, + { + "epoch": 0.3786153603532566, + "grad_norm": 1.4941030740737915, + "learning_rate": 1.4276344530119307e-05, + "loss": 0.4827, + "step": 12004 + }, + { + "epoch": 0.37864690111969723, + "grad_norm": 1.9712748527526855, + "learning_rate": 1.4275421070031803e-05, + "loss": 0.4842, + "step": 12005 + }, + { + "epoch": 0.3786784418861378, + "grad_norm": 2.140021324157715, + "learning_rate": 1.4274497565326959e-05, + "loss": 0.4982, + "step": 12006 + }, + { + "epoch": 0.37870998265257844, + "grad_norm": 1.28803288936615, + "learning_rate": 1.4273574016014417e-05, + "loss": 0.4954, + "step": 12007 + }, + { + "epoch": 0.37874152341901907, + "grad_norm": 1.5862317085266113, + "learning_rate": 1.427265042210381e-05, + "loss": 0.5142, + "step": 12008 + }, + { + "epoch": 0.3787730641854597, + "grad_norm": 1.5724135637283325, + "learning_rate": 1.4271726783604778e-05, + "loss": 0.5055, + "step": 12009 + }, + { + "epoch": 0.3788046049519003, + "grad_norm": 1.3950048685073853, + "learning_rate": 1.4270803100526962e-05, + "loss": 0.4874, + "step": 12010 + }, + { + "epoch": 0.37883614571834096, + "grad_norm": 1.610412836074829, + "learning_rate": 1.4269879372879998e-05, + "loss": 0.4954, + "step": 12011 + }, + { + "epoch": 0.3788676864847816, + "grad_norm": 1.3593428134918213, + "learning_rate": 1.4268955600673528e-05, + "loss": 0.5158, + "step": 12012 + }, + { + "epoch": 0.3788992272512222, + "grad_norm": 2.601422071456909, + "learning_rate": 1.4268031783917193e-05, + "loss": 0.4886, + "step": 12013 + }, + { + "epoch": 0.37893076801766284, + "grad_norm": 2.0545814037323, + "learning_rate": 1.4267107922620631e-05, + "loss": 0.8511, + "step": 12014 + }, + { + "epoch": 0.3789623087841035, + "grad_norm": 1.396885633468628, + "learning_rate": 1.4266184016793486e-05, + "loss": 0.5267, + "step": 12015 + }, + { + "epoch": 0.3789938495505441, + "grad_norm": 2.022106409072876, + "learning_rate": 1.4265260066445392e-05, + "loss": 0.5526, + "step": 12016 + }, + { + "epoch": 0.3790253903169847, + "grad_norm": 5.069474220275879, + "learning_rate": 1.4264336071586005e-05, + "loss": 0.5435, + "step": 12017 + }, + { + "epoch": 0.3790569310834253, + "grad_norm": 1.4372801780700684, + "learning_rate": 1.4263412032224954e-05, + "loss": 0.5012, + "step": 12018 + }, + { + "epoch": 0.37908847184986594, + "grad_norm": 1.718032717704773, + "learning_rate": 1.426248794837189e-05, + "loss": 0.534, + "step": 12019 + }, + { + "epoch": 0.37912001261630657, + "grad_norm": 1.6438437700271606, + "learning_rate": 1.4261563820036455e-05, + "loss": 0.5333, + "step": 12020 + }, + { + "epoch": 0.3791515533827472, + "grad_norm": 1.9380371570587158, + "learning_rate": 1.4260639647228294e-05, + "loss": 0.4748, + "step": 12021 + }, + { + "epoch": 0.3791830941491878, + "grad_norm": 1.44586181640625, + "learning_rate": 1.4259715429957049e-05, + "loss": 0.5344, + "step": 12022 + }, + { + "epoch": 0.37921463491562846, + "grad_norm": 2.2344634532928467, + "learning_rate": 1.4258791168232366e-05, + "loss": 0.5052, + "step": 12023 + }, + { + "epoch": 0.3792461756820691, + "grad_norm": 1.6461378335952759, + "learning_rate": 1.4257866862063892e-05, + "loss": 0.424, + "step": 12024 + }, + { + "epoch": 0.3792777164485097, + "grad_norm": 1.3722410202026367, + "learning_rate": 1.4256942511461267e-05, + "loss": 0.4861, + "step": 12025 + }, + { + "epoch": 0.37930925721495035, + "grad_norm": 3.2565853595733643, + "learning_rate": 1.4256018116434143e-05, + "loss": 0.5368, + "step": 12026 + }, + { + "epoch": 0.3793407979813909, + "grad_norm": 1.7470777034759521, + "learning_rate": 1.4255093676992167e-05, + "loss": 0.4044, + "step": 12027 + }, + { + "epoch": 0.37937233874783155, + "grad_norm": 2.0358569622039795, + "learning_rate": 1.4254169193144983e-05, + "loss": 0.5157, + "step": 12028 + }, + { + "epoch": 0.3794038795142722, + "grad_norm": 2.299102544784546, + "learning_rate": 1.4253244664902237e-05, + "loss": 0.8223, + "step": 12029 + }, + { + "epoch": 0.3794354202807128, + "grad_norm": 2.0222835540771484, + "learning_rate": 1.4252320092273584e-05, + "loss": 0.3757, + "step": 12030 + }, + { + "epoch": 0.37946696104715344, + "grad_norm": 1.7295398712158203, + "learning_rate": 1.425139547526867e-05, + "loss": 0.5153, + "step": 12031 + }, + { + "epoch": 0.37949850181359407, + "grad_norm": 1.6806581020355225, + "learning_rate": 1.4250470813897142e-05, + "loss": 0.5405, + "step": 12032 + }, + { + "epoch": 0.3795300425800347, + "grad_norm": 1.3177728652954102, + "learning_rate": 1.424954610816865e-05, + "loss": 0.5595, + "step": 12033 + }, + { + "epoch": 0.37956158334647533, + "grad_norm": 1.9263463020324707, + "learning_rate": 1.4248621358092847e-05, + "loss": 0.4922, + "step": 12034 + }, + { + "epoch": 0.37959312411291596, + "grad_norm": 1.2546993494033813, + "learning_rate": 1.424769656367938e-05, + "loss": 0.5162, + "step": 12035 + }, + { + "epoch": 0.3796246648793566, + "grad_norm": 1.4746487140655518, + "learning_rate": 1.4246771724937898e-05, + "loss": 0.5024, + "step": 12036 + }, + { + "epoch": 0.3796562056457972, + "grad_norm": 5.135655403137207, + "learning_rate": 1.4245846841878056e-05, + "loss": 0.3847, + "step": 12037 + }, + { + "epoch": 0.3796877464122378, + "grad_norm": 1.8209565877914429, + "learning_rate": 1.4244921914509508e-05, + "loss": 0.5222, + "step": 12038 + }, + { + "epoch": 0.3797192871786784, + "grad_norm": 1.6167688369750977, + "learning_rate": 1.4243996942841905e-05, + "loss": 0.5068, + "step": 12039 + }, + { + "epoch": 0.37975082794511905, + "grad_norm": 1.7664340734481812, + "learning_rate": 1.4243071926884895e-05, + "loss": 0.5121, + "step": 12040 + }, + { + "epoch": 0.3797823687115597, + "grad_norm": 1.6567703485488892, + "learning_rate": 1.4242146866648136e-05, + "loss": 0.501, + "step": 12041 + }, + { + "epoch": 0.3798139094780003, + "grad_norm": 2.0465199947357178, + "learning_rate": 1.424122176214128e-05, + "loss": 0.5234, + "step": 12042 + }, + { + "epoch": 0.37984545024444094, + "grad_norm": 2.0578689575195312, + "learning_rate": 1.4240296613373985e-05, + "loss": 0.801, + "step": 12043 + }, + { + "epoch": 0.37987699101088157, + "grad_norm": 2.1160638332366943, + "learning_rate": 1.42393714203559e-05, + "loss": 0.4665, + "step": 12044 + }, + { + "epoch": 0.3799085317773222, + "grad_norm": 2.182375907897949, + "learning_rate": 1.4238446183096681e-05, + "loss": 0.5458, + "step": 12045 + }, + { + "epoch": 0.37994007254376283, + "grad_norm": 1.633445382118225, + "learning_rate": 1.4237520901605986e-05, + "loss": 0.5159, + "step": 12046 + }, + { + "epoch": 0.37997161331020346, + "grad_norm": 1.6528230905532837, + "learning_rate": 1.423659557589347e-05, + "loss": 0.4926, + "step": 12047 + }, + { + "epoch": 0.3800031540766441, + "grad_norm": 1.6006653308868408, + "learning_rate": 1.4235670205968792e-05, + "loss": 0.4862, + "step": 12048 + }, + { + "epoch": 0.38003469484308466, + "grad_norm": 1.6838924884796143, + "learning_rate": 1.4234744791841602e-05, + "loss": 0.5315, + "step": 12049 + }, + { + "epoch": 0.3800662356095253, + "grad_norm": 1.6504981517791748, + "learning_rate": 1.4233819333521564e-05, + "loss": 0.4948, + "step": 12050 + }, + { + "epoch": 0.3800977763759659, + "grad_norm": 1.8703546524047852, + "learning_rate": 1.4232893831018336e-05, + "loss": 0.5128, + "step": 12051 + }, + { + "epoch": 0.38012931714240655, + "grad_norm": 1.8647432327270508, + "learning_rate": 1.423196828434157e-05, + "loss": 0.5122, + "step": 12052 + }, + { + "epoch": 0.3801608579088472, + "grad_norm": 1.8089107275009155, + "learning_rate": 1.4231042693500934e-05, + "loss": 0.516, + "step": 12053 + }, + { + "epoch": 0.3801923986752878, + "grad_norm": 1.907871961593628, + "learning_rate": 1.4230117058506076e-05, + "loss": 0.5338, + "step": 12054 + }, + { + "epoch": 0.38022393944172844, + "grad_norm": 2.175873279571533, + "learning_rate": 1.4229191379366666e-05, + "loss": 0.3763, + "step": 12055 + }, + { + "epoch": 0.38025548020816907, + "grad_norm": 1.8851643800735474, + "learning_rate": 1.422826565609236e-05, + "loss": 0.4999, + "step": 12056 + }, + { + "epoch": 0.3802870209746097, + "grad_norm": 3.724558115005493, + "learning_rate": 1.4227339888692818e-05, + "loss": 0.8297, + "step": 12057 + }, + { + "epoch": 0.38031856174105033, + "grad_norm": 1.7542119026184082, + "learning_rate": 1.4226414077177702e-05, + "loss": 0.5171, + "step": 12058 + }, + { + "epoch": 0.3803501025074909, + "grad_norm": 1.690457820892334, + "learning_rate": 1.4225488221556672e-05, + "loss": 0.4851, + "step": 12059 + }, + { + "epoch": 0.38038164327393154, + "grad_norm": 1.3800956010818481, + "learning_rate": 1.4224562321839393e-05, + "loss": 0.5021, + "step": 12060 + }, + { + "epoch": 0.38041318404037217, + "grad_norm": 2.2345001697540283, + "learning_rate": 1.4223636378035523e-05, + "loss": 0.5425, + "step": 12061 + }, + { + "epoch": 0.3804447248068128, + "grad_norm": 2.468665838241577, + "learning_rate": 1.4222710390154732e-05, + "loss": 0.492, + "step": 12062 + }, + { + "epoch": 0.3804762655732534, + "grad_norm": 1.3884605169296265, + "learning_rate": 1.4221784358206679e-05, + "loss": 0.5152, + "step": 12063 + }, + { + "epoch": 0.38050780633969405, + "grad_norm": 2.0692625045776367, + "learning_rate": 1.4220858282201025e-05, + "loss": 0.5415, + "step": 12064 + }, + { + "epoch": 0.3805393471061347, + "grad_norm": 1.5027198791503906, + "learning_rate": 1.4219932162147441e-05, + "loss": 0.5123, + "step": 12065 + }, + { + "epoch": 0.3805708878725753, + "grad_norm": 1.7411936521530151, + "learning_rate": 1.4219005998055585e-05, + "loss": 0.4958, + "step": 12066 + }, + { + "epoch": 0.38060242863901594, + "grad_norm": 1.6828707456588745, + "learning_rate": 1.4218079789935127e-05, + "loss": 0.5325, + "step": 12067 + }, + { + "epoch": 0.3806339694054566, + "grad_norm": 1.937387228012085, + "learning_rate": 1.421715353779573e-05, + "loss": 0.5269, + "step": 12068 + }, + { + "epoch": 0.3806655101718972, + "grad_norm": 1.6372511386871338, + "learning_rate": 1.4216227241647063e-05, + "loss": 0.5031, + "step": 12069 + }, + { + "epoch": 0.3806970509383378, + "grad_norm": 1.351663589477539, + "learning_rate": 1.4215300901498788e-05, + "loss": 0.5263, + "step": 12070 + }, + { + "epoch": 0.3807285917047784, + "grad_norm": 1.5771201848983765, + "learning_rate": 1.4214374517360576e-05, + "loss": 0.4726, + "step": 12071 + }, + { + "epoch": 0.38076013247121904, + "grad_norm": 5.591776371002197, + "learning_rate": 1.4213448089242095e-05, + "loss": 0.5058, + "step": 12072 + }, + { + "epoch": 0.38079167323765967, + "grad_norm": 1.3442044258117676, + "learning_rate": 1.4212521617153012e-05, + "loss": 0.3819, + "step": 12073 + }, + { + "epoch": 0.3808232140041003, + "grad_norm": 2.098184585571289, + "learning_rate": 1.4211595101102994e-05, + "loss": 0.5284, + "step": 12074 + }, + { + "epoch": 0.3808547547705409, + "grad_norm": 2.0565924644470215, + "learning_rate": 1.4210668541101713e-05, + "loss": 0.5348, + "step": 12075 + }, + { + "epoch": 0.38088629553698156, + "grad_norm": 2.1382858753204346, + "learning_rate": 1.4209741937158833e-05, + "loss": 0.5057, + "step": 12076 + }, + { + "epoch": 0.3809178363034222, + "grad_norm": 1.9393094778060913, + "learning_rate": 1.4208815289284026e-05, + "loss": 0.5281, + "step": 12077 + }, + { + "epoch": 0.3809493770698628, + "grad_norm": 2.05837345123291, + "learning_rate": 1.4207888597486967e-05, + "loss": 0.4773, + "step": 12078 + }, + { + "epoch": 0.38098091783630345, + "grad_norm": 1.4899080991744995, + "learning_rate": 1.4206961861777322e-05, + "loss": 0.5037, + "step": 12079 + }, + { + "epoch": 0.381012458602744, + "grad_norm": 1.4528892040252686, + "learning_rate": 1.4206035082164762e-05, + "loss": 0.5044, + "step": 12080 + }, + { + "epoch": 0.38104399936918465, + "grad_norm": 1.6594178676605225, + "learning_rate": 1.420510825865896e-05, + "loss": 0.8144, + "step": 12081 + }, + { + "epoch": 0.3810755401356253, + "grad_norm": 2.1770739555358887, + "learning_rate": 1.4204181391269592e-05, + "loss": 0.5277, + "step": 12082 + }, + { + "epoch": 0.3811070809020659, + "grad_norm": 2.1731116771698, + "learning_rate": 1.4203254480006323e-05, + "loss": 0.5268, + "step": 12083 + }, + { + "epoch": 0.38113862166850654, + "grad_norm": 1.7386939525604248, + "learning_rate": 1.4202327524878831e-05, + "loss": 0.503, + "step": 12084 + }, + { + "epoch": 0.38117016243494717, + "grad_norm": 1.7367440462112427, + "learning_rate": 1.420140052589679e-05, + "loss": 0.5379, + "step": 12085 + }, + { + "epoch": 0.3812017032013878, + "grad_norm": 1.8858979940414429, + "learning_rate": 1.420047348306987e-05, + "loss": 0.5102, + "step": 12086 + }, + { + "epoch": 0.38123324396782843, + "grad_norm": 1.496999979019165, + "learning_rate": 1.4199546396407746e-05, + "loss": 0.4765, + "step": 12087 + }, + { + "epoch": 0.38126478473426906, + "grad_norm": 1.6354608535766602, + "learning_rate": 1.4198619265920095e-05, + "loss": 0.5144, + "step": 12088 + }, + { + "epoch": 0.3812963255007097, + "grad_norm": 1.5233551263809204, + "learning_rate": 1.4197692091616596e-05, + "loss": 0.518, + "step": 12089 + }, + { + "epoch": 0.3813278662671503, + "grad_norm": 1.7489458322525024, + "learning_rate": 1.4196764873506917e-05, + "loss": 0.5086, + "step": 12090 + }, + { + "epoch": 0.3813594070335909, + "grad_norm": 1.5670884847640991, + "learning_rate": 1.4195837611600736e-05, + "loss": 0.5251, + "step": 12091 + }, + { + "epoch": 0.3813909478000315, + "grad_norm": 1.6877580881118774, + "learning_rate": 1.4194910305907737e-05, + "loss": 0.5223, + "step": 12092 + }, + { + "epoch": 0.38142248856647215, + "grad_norm": 1.9876190423965454, + "learning_rate": 1.4193982956437589e-05, + "loss": 0.5475, + "step": 12093 + }, + { + "epoch": 0.3814540293329128, + "grad_norm": 1.3791683912277222, + "learning_rate": 1.419305556319997e-05, + "loss": 0.515, + "step": 12094 + }, + { + "epoch": 0.3814855700993534, + "grad_norm": 2.120680570602417, + "learning_rate": 1.4192128126204562e-05, + "loss": 0.8166, + "step": 12095 + }, + { + "epoch": 0.38151711086579404, + "grad_norm": 1.4849233627319336, + "learning_rate": 1.4191200645461045e-05, + "loss": 0.5111, + "step": 12096 + }, + { + "epoch": 0.38154865163223467, + "grad_norm": 1.6443874835968018, + "learning_rate": 1.4190273120979092e-05, + "loss": 0.484, + "step": 12097 + }, + { + "epoch": 0.3815801923986753, + "grad_norm": 1.8303377628326416, + "learning_rate": 1.4189345552768385e-05, + "loss": 0.4922, + "step": 12098 + }, + { + "epoch": 0.38161173316511593, + "grad_norm": 2.209115505218506, + "learning_rate": 1.4188417940838606e-05, + "loss": 0.8344, + "step": 12099 + }, + { + "epoch": 0.38164327393155656, + "grad_norm": 1.300163984298706, + "learning_rate": 1.4187490285199431e-05, + "loss": 0.5136, + "step": 12100 + }, + { + "epoch": 0.3816748146979972, + "grad_norm": 1.530582308769226, + "learning_rate": 1.4186562585860543e-05, + "loss": 0.5021, + "step": 12101 + }, + { + "epoch": 0.38170635546443776, + "grad_norm": 1.339748740196228, + "learning_rate": 1.4185634842831628e-05, + "loss": 0.5288, + "step": 12102 + }, + { + "epoch": 0.3817378962308784, + "grad_norm": 1.4258372783660889, + "learning_rate": 1.4184707056122362e-05, + "loss": 0.519, + "step": 12103 + }, + { + "epoch": 0.381769436997319, + "grad_norm": 1.632839560508728, + "learning_rate": 1.4183779225742426e-05, + "loss": 0.848, + "step": 12104 + }, + { + "epoch": 0.38180097776375965, + "grad_norm": 2.1109516620635986, + "learning_rate": 1.4182851351701505e-05, + "loss": 0.4317, + "step": 12105 + }, + { + "epoch": 0.3818325185302003, + "grad_norm": 1.7071340084075928, + "learning_rate": 1.4181923434009286e-05, + "loss": 0.5017, + "step": 12106 + }, + { + "epoch": 0.3818640592966409, + "grad_norm": 1.9053618907928467, + "learning_rate": 1.4180995472675445e-05, + "loss": 0.5041, + "step": 12107 + }, + { + "epoch": 0.38189560006308154, + "grad_norm": 1.5888444185256958, + "learning_rate": 1.418006746770967e-05, + "loss": 0.5057, + "step": 12108 + }, + { + "epoch": 0.38192714082952217, + "grad_norm": 1.6078603267669678, + "learning_rate": 1.4179139419121648e-05, + "loss": 0.4709, + "step": 12109 + }, + { + "epoch": 0.3819586815959628, + "grad_norm": 1.602287769317627, + "learning_rate": 1.417821132692106e-05, + "loss": 0.524, + "step": 12110 + }, + { + "epoch": 0.38199022236240343, + "grad_norm": 1.506387710571289, + "learning_rate": 1.417728319111759e-05, + "loss": 0.513, + "step": 12111 + }, + { + "epoch": 0.382021763128844, + "grad_norm": 1.367925763130188, + "learning_rate": 1.417635501172093e-05, + "loss": 0.5182, + "step": 12112 + }, + { + "epoch": 0.38205330389528463, + "grad_norm": 1.292794942855835, + "learning_rate": 1.4175426788740763e-05, + "loss": 0.5021, + "step": 12113 + }, + { + "epoch": 0.38208484466172526, + "grad_norm": 1.412054419517517, + "learning_rate": 1.4174498522186773e-05, + "loss": 0.5067, + "step": 12114 + }, + { + "epoch": 0.3821163854281659, + "grad_norm": 1.7353594303131104, + "learning_rate": 1.4173570212068648e-05, + "loss": 0.5236, + "step": 12115 + }, + { + "epoch": 0.3821479261946065, + "grad_norm": 1.6334881782531738, + "learning_rate": 1.4172641858396078e-05, + "loss": 0.5186, + "step": 12116 + }, + { + "epoch": 0.38217946696104715, + "grad_norm": 1.4523578882217407, + "learning_rate": 1.4171713461178753e-05, + "loss": 0.5353, + "step": 12117 + }, + { + "epoch": 0.3822110077274878, + "grad_norm": 1.5628312826156616, + "learning_rate": 1.4170785020426356e-05, + "loss": 0.5377, + "step": 12118 + }, + { + "epoch": 0.3822425484939284, + "grad_norm": 1.4995933771133423, + "learning_rate": 1.4169856536148578e-05, + "loss": 0.5221, + "step": 12119 + }, + { + "epoch": 0.38227408926036904, + "grad_norm": 1.5924692153930664, + "learning_rate": 1.4168928008355113e-05, + "loss": 0.5251, + "step": 12120 + }, + { + "epoch": 0.3823056300268097, + "grad_norm": 1.6636528968811035, + "learning_rate": 1.4167999437055646e-05, + "loss": 0.8135, + "step": 12121 + }, + { + "epoch": 0.3823371707932503, + "grad_norm": 1.3673306703567505, + "learning_rate": 1.4167070822259868e-05, + "loss": 0.4821, + "step": 12122 + }, + { + "epoch": 0.3823687115596909, + "grad_norm": 1.6215509176254272, + "learning_rate": 1.4166142163977472e-05, + "loss": 0.507, + "step": 12123 + }, + { + "epoch": 0.3824002523261315, + "grad_norm": 1.4301338195800781, + "learning_rate": 1.4165213462218144e-05, + "loss": 0.4995, + "step": 12124 + }, + { + "epoch": 0.38243179309257214, + "grad_norm": 1.7728203535079956, + "learning_rate": 1.4164284716991581e-05, + "loss": 0.5421, + "step": 12125 + }, + { + "epoch": 0.38246333385901277, + "grad_norm": 1.9981143474578857, + "learning_rate": 1.4163355928307472e-05, + "loss": 0.5479, + "step": 12126 + }, + { + "epoch": 0.3824948746254534, + "grad_norm": 1.7608222961425781, + "learning_rate": 1.4162427096175515e-05, + "loss": 0.5104, + "step": 12127 + }, + { + "epoch": 0.382526415391894, + "grad_norm": 1.4038258790969849, + "learning_rate": 1.4161498220605397e-05, + "loss": 0.4925, + "step": 12128 + }, + { + "epoch": 0.38255795615833466, + "grad_norm": 1.7333964109420776, + "learning_rate": 1.4160569301606811e-05, + "loss": 0.445, + "step": 12129 + }, + { + "epoch": 0.3825894969247753, + "grad_norm": 2.0674469470977783, + "learning_rate": 1.4159640339189458e-05, + "loss": 0.4996, + "step": 12130 + }, + { + "epoch": 0.3826210376912159, + "grad_norm": 1.7262558937072754, + "learning_rate": 1.4158711333363024e-05, + "loss": 0.3945, + "step": 12131 + }, + { + "epoch": 0.38265257845765654, + "grad_norm": 1.3342671394348145, + "learning_rate": 1.4157782284137213e-05, + "loss": 0.4921, + "step": 12132 + }, + { + "epoch": 0.3826841192240971, + "grad_norm": 1.607978343963623, + "learning_rate": 1.4156853191521713e-05, + "loss": 0.8067, + "step": 12133 + }, + { + "epoch": 0.38271565999053775, + "grad_norm": 1.8705992698669434, + "learning_rate": 1.4155924055526222e-05, + "loss": 0.4793, + "step": 12134 + }, + { + "epoch": 0.3827472007569784, + "grad_norm": 1.8809293508529663, + "learning_rate": 1.4154994876160438e-05, + "loss": 0.5353, + "step": 12135 + }, + { + "epoch": 0.382778741523419, + "grad_norm": 1.5798556804656982, + "learning_rate": 1.4154065653434053e-05, + "loss": 0.4755, + "step": 12136 + }, + { + "epoch": 0.38281028228985964, + "grad_norm": 1.2368214130401611, + "learning_rate": 1.415313638735677e-05, + "loss": 0.4913, + "step": 12137 + }, + { + "epoch": 0.38284182305630027, + "grad_norm": 1.5333222150802612, + "learning_rate": 1.4152207077938282e-05, + "loss": 0.4992, + "step": 12138 + }, + { + "epoch": 0.3828733638227409, + "grad_norm": 1.3609719276428223, + "learning_rate": 1.4151277725188291e-05, + "loss": 0.3892, + "step": 12139 + }, + { + "epoch": 0.3829049045891815, + "grad_norm": 1.8695745468139648, + "learning_rate": 1.4150348329116492e-05, + "loss": 0.5057, + "step": 12140 + }, + { + "epoch": 0.38293644535562216, + "grad_norm": 1.4839001893997192, + "learning_rate": 1.4149418889732588e-05, + "loss": 0.4943, + "step": 12141 + }, + { + "epoch": 0.3829679861220628, + "grad_norm": 1.444700837135315, + "learning_rate": 1.4148489407046274e-05, + "loss": 0.377, + "step": 12142 + }, + { + "epoch": 0.3829995268885034, + "grad_norm": 1.3541475534439087, + "learning_rate": 1.414755988106725e-05, + "loss": 0.3937, + "step": 12143 + }, + { + "epoch": 0.383031067654944, + "grad_norm": 2.3018734455108643, + "learning_rate": 1.4146630311805223e-05, + "loss": 0.5218, + "step": 12144 + }, + { + "epoch": 0.3830626084213846, + "grad_norm": 1.879582405090332, + "learning_rate": 1.4145700699269883e-05, + "loss": 0.5036, + "step": 12145 + }, + { + "epoch": 0.38309414918782525, + "grad_norm": 1.614256739616394, + "learning_rate": 1.4144771043470942e-05, + "loss": 0.5155, + "step": 12146 + }, + { + "epoch": 0.3831256899542659, + "grad_norm": 1.196631908416748, + "learning_rate": 1.4143841344418096e-05, + "loss": 0.4093, + "step": 12147 + }, + { + "epoch": 0.3831572307207065, + "grad_norm": 2.1057393550872803, + "learning_rate": 1.4142911602121046e-05, + "loss": 0.5164, + "step": 12148 + }, + { + "epoch": 0.38318877148714714, + "grad_norm": 1.560903787612915, + "learning_rate": 1.4141981816589496e-05, + "loss": 0.5242, + "step": 12149 + }, + { + "epoch": 0.38322031225358777, + "grad_norm": 2.3781049251556396, + "learning_rate": 1.4141051987833154e-05, + "loss": 0.5274, + "step": 12150 + }, + { + "epoch": 0.3832518530200284, + "grad_norm": 1.2004461288452148, + "learning_rate": 1.4140122115861714e-05, + "loss": 0.3911, + "step": 12151 + }, + { + "epoch": 0.38328339378646903, + "grad_norm": 1.2021489143371582, + "learning_rate": 1.4139192200684889e-05, + "loss": 0.4918, + "step": 12152 + }, + { + "epoch": 0.38331493455290966, + "grad_norm": 1.4888585805892944, + "learning_rate": 1.4138262242312378e-05, + "loss": 0.521, + "step": 12153 + }, + { + "epoch": 0.38334647531935023, + "grad_norm": 1.4572217464447021, + "learning_rate": 1.4137332240753887e-05, + "loss": 0.4834, + "step": 12154 + }, + { + "epoch": 0.38337801608579086, + "grad_norm": 1.2835906744003296, + "learning_rate": 1.413640219601912e-05, + "loss": 0.5156, + "step": 12155 + }, + { + "epoch": 0.3834095568522315, + "grad_norm": 1.8516076803207397, + "learning_rate": 1.4135472108117786e-05, + "loss": 0.8413, + "step": 12156 + }, + { + "epoch": 0.3834410976186721, + "grad_norm": 1.7762833833694458, + "learning_rate": 1.4134541977059588e-05, + "loss": 0.4931, + "step": 12157 + }, + { + "epoch": 0.38347263838511275, + "grad_norm": 1.6619404554367065, + "learning_rate": 1.4133611802854236e-05, + "loss": 0.5145, + "step": 12158 + }, + { + "epoch": 0.3835041791515534, + "grad_norm": 2.8835866451263428, + "learning_rate": 1.4132681585511434e-05, + "loss": 0.503, + "step": 12159 + }, + { + "epoch": 0.383535719917994, + "grad_norm": 1.4420381784439087, + "learning_rate": 1.4131751325040893e-05, + "loss": 0.3937, + "step": 12160 + }, + { + "epoch": 0.38356726068443464, + "grad_norm": 1.8448566198349, + "learning_rate": 1.4130821021452317e-05, + "loss": 0.8582, + "step": 12161 + }, + { + "epoch": 0.38359880145087527, + "grad_norm": 1.9187054634094238, + "learning_rate": 1.4129890674755415e-05, + "loss": 0.4989, + "step": 12162 + }, + { + "epoch": 0.3836303422173159, + "grad_norm": 1.4472622871398926, + "learning_rate": 1.4128960284959898e-05, + "loss": 0.4998, + "step": 12163 + }, + { + "epoch": 0.38366188298375653, + "grad_norm": 1.5825856924057007, + "learning_rate": 1.4128029852075474e-05, + "loss": 0.517, + "step": 12164 + }, + { + "epoch": 0.3836934237501971, + "grad_norm": 1.6044645309448242, + "learning_rate": 1.4127099376111854e-05, + "loss": 0.5314, + "step": 12165 + }, + { + "epoch": 0.38372496451663773, + "grad_norm": 1.3887687921524048, + "learning_rate": 1.4126168857078746e-05, + "loss": 0.5198, + "step": 12166 + }, + { + "epoch": 0.38375650528307836, + "grad_norm": 1.6959892511367798, + "learning_rate": 1.4125238294985862e-05, + "loss": 0.5216, + "step": 12167 + }, + { + "epoch": 0.383788046049519, + "grad_norm": 1.4055359363555908, + "learning_rate": 1.4124307689842916e-05, + "loss": 0.5143, + "step": 12168 + }, + { + "epoch": 0.3838195868159596, + "grad_norm": 1.5480101108551025, + "learning_rate": 1.4123377041659614e-05, + "loss": 0.4999, + "step": 12169 + }, + { + "epoch": 0.38385112758240025, + "grad_norm": 1.2535326480865479, + "learning_rate": 1.4122446350445673e-05, + "loss": 0.5451, + "step": 12170 + }, + { + "epoch": 0.3838826683488409, + "grad_norm": 1.11228346824646, + "learning_rate": 1.4121515616210802e-05, + "loss": 0.4038, + "step": 12171 + }, + { + "epoch": 0.3839142091152815, + "grad_norm": 1.262247920036316, + "learning_rate": 1.4120584838964716e-05, + "loss": 0.5256, + "step": 12172 + }, + { + "epoch": 0.38394574988172214, + "grad_norm": 1.3251012563705444, + "learning_rate": 1.4119654018717126e-05, + "loss": 0.5197, + "step": 12173 + }, + { + "epoch": 0.3839772906481628, + "grad_norm": 1.8451762199401855, + "learning_rate": 1.411872315547775e-05, + "loss": 0.3939, + "step": 12174 + }, + { + "epoch": 0.3840088314146034, + "grad_norm": 1.7041960954666138, + "learning_rate": 1.4117792249256298e-05, + "loss": 0.5403, + "step": 12175 + }, + { + "epoch": 0.384040372181044, + "grad_norm": 1.523829698562622, + "learning_rate": 1.4116861300062486e-05, + "loss": 0.502, + "step": 12176 + }, + { + "epoch": 0.3840719129474846, + "grad_norm": 6.20953893661499, + "learning_rate": 1.411593030790603e-05, + "loss": 0.5029, + "step": 12177 + }, + { + "epoch": 0.38410345371392524, + "grad_norm": 1.5457018613815308, + "learning_rate": 1.4114999272796647e-05, + "loss": 0.5144, + "step": 12178 + }, + { + "epoch": 0.38413499448036587, + "grad_norm": 1.442305088043213, + "learning_rate": 1.4114068194744049e-05, + "loss": 0.4944, + "step": 12179 + }, + { + "epoch": 0.3841665352468065, + "grad_norm": 1.3517216444015503, + "learning_rate": 1.4113137073757957e-05, + "loss": 0.5343, + "step": 12180 + }, + { + "epoch": 0.3841980760132471, + "grad_norm": 1.4367271661758423, + "learning_rate": 1.4112205909848084e-05, + "loss": 0.8106, + "step": 12181 + }, + { + "epoch": 0.38422961677968775, + "grad_norm": 1.3851770162582397, + "learning_rate": 1.4111274703024152e-05, + "loss": 0.5075, + "step": 12182 + }, + { + "epoch": 0.3842611575461284, + "grad_norm": 1.8507957458496094, + "learning_rate": 1.4110343453295874e-05, + "loss": 0.5014, + "step": 12183 + }, + { + "epoch": 0.384292698312569, + "grad_norm": 1.896435022354126, + "learning_rate": 1.4109412160672969e-05, + "loss": 0.5116, + "step": 12184 + }, + { + "epoch": 0.38432423907900964, + "grad_norm": 1.9646035432815552, + "learning_rate": 1.4108480825165161e-05, + "loss": 0.8137, + "step": 12185 + }, + { + "epoch": 0.3843557798454502, + "grad_norm": 1.4712601900100708, + "learning_rate": 1.4107549446782162e-05, + "loss": 0.4817, + "step": 12186 + }, + { + "epoch": 0.38438732061189085, + "grad_norm": 1.6782972812652588, + "learning_rate": 1.4106618025533695e-05, + "loss": 0.5361, + "step": 12187 + }, + { + "epoch": 0.3844188613783315, + "grad_norm": 1.4266806840896606, + "learning_rate": 1.4105686561429481e-05, + "loss": 0.4629, + "step": 12188 + }, + { + "epoch": 0.3844504021447721, + "grad_norm": 1.8239033222198486, + "learning_rate": 1.4104755054479244e-05, + "loss": 0.4801, + "step": 12189 + }, + { + "epoch": 0.38448194291121274, + "grad_norm": 1.5905718803405762, + "learning_rate": 1.4103823504692698e-05, + "loss": 0.5089, + "step": 12190 + }, + { + "epoch": 0.38451348367765337, + "grad_norm": 1.527572751045227, + "learning_rate": 1.4102891912079566e-05, + "loss": 0.5202, + "step": 12191 + }, + { + "epoch": 0.384545024444094, + "grad_norm": 1.5276658535003662, + "learning_rate": 1.4101960276649572e-05, + "loss": 0.3667, + "step": 12192 + }, + { + "epoch": 0.3845765652105346, + "grad_norm": 1.5139912366867065, + "learning_rate": 1.4101028598412436e-05, + "loss": 0.5, + "step": 12193 + }, + { + "epoch": 0.38460810597697526, + "grad_norm": 1.632599115371704, + "learning_rate": 1.4100096877377882e-05, + "loss": 0.4743, + "step": 12194 + }, + { + "epoch": 0.3846396467434159, + "grad_norm": 1.5416125059127808, + "learning_rate": 1.4099165113555635e-05, + "loss": 0.838, + "step": 12195 + }, + { + "epoch": 0.3846711875098565, + "grad_norm": 4.378067970275879, + "learning_rate": 1.4098233306955418e-05, + "loss": 0.4056, + "step": 12196 + }, + { + "epoch": 0.3847027282762971, + "grad_norm": 1.948890209197998, + "learning_rate": 1.4097301457586952e-05, + "loss": 0.4956, + "step": 12197 + }, + { + "epoch": 0.3847342690427377, + "grad_norm": 1.7065004110336304, + "learning_rate": 1.4096369565459964e-05, + "loss": 0.4702, + "step": 12198 + }, + { + "epoch": 0.38476580980917835, + "grad_norm": 1.798027753829956, + "learning_rate": 1.4095437630584181e-05, + "loss": 0.8306, + "step": 12199 + }, + { + "epoch": 0.384797350575619, + "grad_norm": 1.6589605808258057, + "learning_rate": 1.4094505652969326e-05, + "loss": 0.4872, + "step": 12200 + }, + { + "epoch": 0.3848288913420596, + "grad_norm": 1.7928102016448975, + "learning_rate": 1.4093573632625123e-05, + "loss": 0.5012, + "step": 12201 + }, + { + "epoch": 0.38486043210850024, + "grad_norm": 2.1259262561798096, + "learning_rate": 1.4092641569561303e-05, + "loss": 0.5471, + "step": 12202 + }, + { + "epoch": 0.38489197287494087, + "grad_norm": 1.5354877710342407, + "learning_rate": 1.4091709463787588e-05, + "loss": 0.503, + "step": 12203 + }, + { + "epoch": 0.3849235136413815, + "grad_norm": 1.795877456665039, + "learning_rate": 1.4090777315313708e-05, + "loss": 0.5468, + "step": 12204 + }, + { + "epoch": 0.38495505440782213, + "grad_norm": 1.5437346696853638, + "learning_rate": 1.408984512414939e-05, + "loss": 0.5265, + "step": 12205 + }, + { + "epoch": 0.38498659517426276, + "grad_norm": 1.4460883140563965, + "learning_rate": 1.4088912890304367e-05, + "loss": 0.5179, + "step": 12206 + }, + { + "epoch": 0.38501813594070333, + "grad_norm": 2.340120315551758, + "learning_rate": 1.408798061378836e-05, + "loss": 0.8296, + "step": 12207 + }, + { + "epoch": 0.38504967670714396, + "grad_norm": 1.4845510721206665, + "learning_rate": 1.4087048294611099e-05, + "loss": 0.8555, + "step": 12208 + }, + { + "epoch": 0.3850812174735846, + "grad_norm": 1.4816086292266846, + "learning_rate": 1.4086115932782316e-05, + "loss": 0.4914, + "step": 12209 + }, + { + "epoch": 0.3851127582400252, + "grad_norm": 1.43892502784729, + "learning_rate": 1.4085183528311743e-05, + "loss": 0.5299, + "step": 12210 + }, + { + "epoch": 0.38514429900646585, + "grad_norm": 1.6401432752609253, + "learning_rate": 1.4084251081209105e-05, + "loss": 0.526, + "step": 12211 + }, + { + "epoch": 0.3851758397729065, + "grad_norm": 1.9484260082244873, + "learning_rate": 1.408331859148414e-05, + "loss": 0.484, + "step": 12212 + }, + { + "epoch": 0.3852073805393471, + "grad_norm": 1.5107530355453491, + "learning_rate": 1.4082386059146572e-05, + "loss": 0.5212, + "step": 12213 + }, + { + "epoch": 0.38523892130578774, + "grad_norm": 1.2458895444869995, + "learning_rate": 1.4081453484206134e-05, + "loss": 0.495, + "step": 12214 + }, + { + "epoch": 0.38527046207222837, + "grad_norm": 2.234898090362549, + "learning_rate": 1.408052086667256e-05, + "loss": 0.392, + "step": 12215 + }, + { + "epoch": 0.385302002838669, + "grad_norm": 1.5847870111465454, + "learning_rate": 1.4079588206555586e-05, + "loss": 0.5503, + "step": 12216 + }, + { + "epoch": 0.38533354360510963, + "grad_norm": 10.054447174072266, + "learning_rate": 1.4078655503864935e-05, + "loss": 0.8351, + "step": 12217 + }, + { + "epoch": 0.3853650843715502, + "grad_norm": 1.326490044593811, + "learning_rate": 1.407772275861035e-05, + "loss": 0.4932, + "step": 12218 + }, + { + "epoch": 0.38539662513799083, + "grad_norm": 1.970439076423645, + "learning_rate": 1.4076789970801566e-05, + "loss": 0.5344, + "step": 12219 + }, + { + "epoch": 0.38542816590443146, + "grad_norm": 1.3057570457458496, + "learning_rate": 1.4075857140448311e-05, + "loss": 0.49, + "step": 12220 + }, + { + "epoch": 0.3854597066708721, + "grad_norm": 1.4052493572235107, + "learning_rate": 1.4074924267560319e-05, + "loss": 0.4932, + "step": 12221 + }, + { + "epoch": 0.3854912474373127, + "grad_norm": 2.500067949295044, + "learning_rate": 1.4073991352147329e-05, + "loss": 0.5251, + "step": 12222 + }, + { + "epoch": 0.38552278820375335, + "grad_norm": 1.7771856784820557, + "learning_rate": 1.4073058394219077e-05, + "loss": 0.5334, + "step": 12223 + }, + { + "epoch": 0.385554328970194, + "grad_norm": 1.9071717262268066, + "learning_rate": 1.4072125393785296e-05, + "loss": 0.5108, + "step": 12224 + }, + { + "epoch": 0.3855858697366346, + "grad_norm": 1.7567747831344604, + "learning_rate": 1.4071192350855727e-05, + "loss": 0.5083, + "step": 12225 + }, + { + "epoch": 0.38561741050307524, + "grad_norm": 1.5745187997817993, + "learning_rate": 1.4070259265440104e-05, + "loss": 0.5133, + "step": 12226 + }, + { + "epoch": 0.38564895126951587, + "grad_norm": 1.4660149812698364, + "learning_rate": 1.4069326137548165e-05, + "loss": 0.51, + "step": 12227 + }, + { + "epoch": 0.3856804920359565, + "grad_norm": 1.9728847742080688, + "learning_rate": 1.4068392967189645e-05, + "loss": 0.5254, + "step": 12228 + }, + { + "epoch": 0.3857120328023971, + "grad_norm": 2.231476068496704, + "learning_rate": 1.406745975437429e-05, + "loss": 0.4943, + "step": 12229 + }, + { + "epoch": 0.3857435735688377, + "grad_norm": 1.3706883192062378, + "learning_rate": 1.4066526499111834e-05, + "loss": 0.4983, + "step": 12230 + }, + { + "epoch": 0.38577511433527834, + "grad_norm": 1.7037158012390137, + "learning_rate": 1.4065593201412013e-05, + "loss": 0.5577, + "step": 12231 + }, + { + "epoch": 0.38580665510171896, + "grad_norm": 1.589223861694336, + "learning_rate": 1.4064659861284574e-05, + "loss": 0.5113, + "step": 12232 + }, + { + "epoch": 0.3858381958681596, + "grad_norm": 1.3228040933609009, + "learning_rate": 1.406372647873925e-05, + "loss": 0.483, + "step": 12233 + }, + { + "epoch": 0.3858697366346002, + "grad_norm": 1.335626244544983, + "learning_rate": 1.4062793053785785e-05, + "loss": 0.5037, + "step": 12234 + }, + { + "epoch": 0.38590127740104085, + "grad_norm": 2.6202785968780518, + "learning_rate": 1.4061859586433921e-05, + "loss": 0.8556, + "step": 12235 + }, + { + "epoch": 0.3859328181674815, + "grad_norm": 1.4128960371017456, + "learning_rate": 1.4060926076693398e-05, + "loss": 0.4909, + "step": 12236 + }, + { + "epoch": 0.3859643589339221, + "grad_norm": 1.9561741352081299, + "learning_rate": 1.4059992524573958e-05, + "loss": 0.5113, + "step": 12237 + }, + { + "epoch": 0.38599589970036274, + "grad_norm": 1.5352369546890259, + "learning_rate": 1.4059058930085343e-05, + "loss": 0.5197, + "step": 12238 + }, + { + "epoch": 0.3860274404668033, + "grad_norm": 1.703596591949463, + "learning_rate": 1.4058125293237299e-05, + "loss": 0.4768, + "step": 12239 + }, + { + "epoch": 0.38605898123324395, + "grad_norm": 1.8761720657348633, + "learning_rate": 1.4057191614039564e-05, + "loss": 0.3788, + "step": 12240 + }, + { + "epoch": 0.3860905219996846, + "grad_norm": 1.3509156703948975, + "learning_rate": 1.4056257892501886e-05, + "loss": 0.5122, + "step": 12241 + }, + { + "epoch": 0.3861220627661252, + "grad_norm": 1.4995777606964111, + "learning_rate": 1.4055324128634004e-05, + "loss": 0.5208, + "step": 12242 + }, + { + "epoch": 0.38615360353256584, + "grad_norm": 1.6785390377044678, + "learning_rate": 1.4054390322445666e-05, + "loss": 0.5046, + "step": 12243 + }, + { + "epoch": 0.38618514429900647, + "grad_norm": 1.519964337348938, + "learning_rate": 1.4053456473946621e-05, + "loss": 0.5204, + "step": 12244 + }, + { + "epoch": 0.3862166850654471, + "grad_norm": 1.644553303718567, + "learning_rate": 1.4052522583146605e-05, + "loss": 0.8042, + "step": 12245 + }, + { + "epoch": 0.3862482258318877, + "grad_norm": 1.613461971282959, + "learning_rate": 1.4051588650055373e-05, + "loss": 0.5005, + "step": 12246 + }, + { + "epoch": 0.38627976659832836, + "grad_norm": 1.3026504516601562, + "learning_rate": 1.4050654674682666e-05, + "loss": 0.5044, + "step": 12247 + }, + { + "epoch": 0.386311307364769, + "grad_norm": 2.230633020401001, + "learning_rate": 1.4049720657038233e-05, + "loss": 0.5368, + "step": 12248 + }, + { + "epoch": 0.3863428481312096, + "grad_norm": 1.6316298246383667, + "learning_rate": 1.404878659713182e-05, + "loss": 0.5404, + "step": 12249 + }, + { + "epoch": 0.3863743888976502, + "grad_norm": 2.01277232170105, + "learning_rate": 1.4047852494973178e-05, + "loss": 0.4982, + "step": 12250 + }, + { + "epoch": 0.3864059296640908, + "grad_norm": 2.1750950813293457, + "learning_rate": 1.4046918350572048e-05, + "loss": 0.5114, + "step": 12251 + }, + { + "epoch": 0.38643747043053145, + "grad_norm": 1.5889731645584106, + "learning_rate": 1.4045984163938182e-05, + "loss": 0.5596, + "step": 12252 + }, + { + "epoch": 0.3864690111969721, + "grad_norm": 5.094799995422363, + "learning_rate": 1.4045049935081332e-05, + "loss": 0.4951, + "step": 12253 + }, + { + "epoch": 0.3865005519634127, + "grad_norm": 1.364967942237854, + "learning_rate": 1.4044115664011244e-05, + "loss": 0.5013, + "step": 12254 + }, + { + "epoch": 0.38653209272985334, + "grad_norm": 1.577090859413147, + "learning_rate": 1.4043181350737669e-05, + "loss": 0.5204, + "step": 12255 + }, + { + "epoch": 0.38656363349629397, + "grad_norm": 1.2202448844909668, + "learning_rate": 1.4042246995270357e-05, + "loss": 0.54, + "step": 12256 + }, + { + "epoch": 0.3865951742627346, + "grad_norm": 1.4896446466445923, + "learning_rate": 1.4041312597619061e-05, + "loss": 0.5402, + "step": 12257 + }, + { + "epoch": 0.3866267150291752, + "grad_norm": 1.5850659608840942, + "learning_rate": 1.4040378157793529e-05, + "loss": 0.497, + "step": 12258 + }, + { + "epoch": 0.38665825579561586, + "grad_norm": 1.9860259294509888, + "learning_rate": 1.4039443675803513e-05, + "loss": 0.4005, + "step": 12259 + }, + { + "epoch": 0.38668979656205643, + "grad_norm": 1.472230315208435, + "learning_rate": 1.4038509151658767e-05, + "loss": 0.5048, + "step": 12260 + }, + { + "epoch": 0.38672133732849706, + "grad_norm": 1.8668138980865479, + "learning_rate": 1.403757458536904e-05, + "loss": 0.5156, + "step": 12261 + }, + { + "epoch": 0.3867528780949377, + "grad_norm": 1.6272464990615845, + "learning_rate": 1.4036639976944088e-05, + "loss": 0.5103, + "step": 12262 + }, + { + "epoch": 0.3867844188613783, + "grad_norm": 1.561583399772644, + "learning_rate": 1.4035705326393662e-05, + "loss": 0.4978, + "step": 12263 + }, + { + "epoch": 0.38681595962781895, + "grad_norm": 2.0112898349761963, + "learning_rate": 1.4034770633727521e-05, + "loss": 0.5164, + "step": 12264 + }, + { + "epoch": 0.3868475003942596, + "grad_norm": 1.7786061763763428, + "learning_rate": 1.4033835898955413e-05, + "loss": 0.3784, + "step": 12265 + }, + { + "epoch": 0.3868790411607002, + "grad_norm": 1.3772634267807007, + "learning_rate": 1.4032901122087094e-05, + "loss": 0.5101, + "step": 12266 + }, + { + "epoch": 0.38691058192714084, + "grad_norm": 1.427772879600525, + "learning_rate": 1.4031966303132323e-05, + "loss": 0.5172, + "step": 12267 + }, + { + "epoch": 0.38694212269358147, + "grad_norm": 1.5309693813323975, + "learning_rate": 1.403103144210085e-05, + "loss": 0.4965, + "step": 12268 + }, + { + "epoch": 0.3869736634600221, + "grad_norm": 1.756460189819336, + "learning_rate": 1.4030096539002439e-05, + "loss": 0.5109, + "step": 12269 + }, + { + "epoch": 0.38700520422646273, + "grad_norm": 1.4283679723739624, + "learning_rate": 1.4029161593846837e-05, + "loss": 0.4963, + "step": 12270 + }, + { + "epoch": 0.3870367449929033, + "grad_norm": 1.3920632600784302, + "learning_rate": 1.4028226606643806e-05, + "loss": 0.4998, + "step": 12271 + }, + { + "epoch": 0.38706828575934393, + "grad_norm": 1.9193893671035767, + "learning_rate": 1.4027291577403104e-05, + "loss": 0.5195, + "step": 12272 + }, + { + "epoch": 0.38709982652578456, + "grad_norm": 1.506105661392212, + "learning_rate": 1.4026356506134485e-05, + "loss": 0.4564, + "step": 12273 + }, + { + "epoch": 0.3871313672922252, + "grad_norm": 1.8999062776565552, + "learning_rate": 1.4025421392847712e-05, + "loss": 0.4753, + "step": 12274 + }, + { + "epoch": 0.3871629080586658, + "grad_norm": 1.9947118759155273, + "learning_rate": 1.402448623755254e-05, + "loss": 0.3988, + "step": 12275 + }, + { + "epoch": 0.38719444882510645, + "grad_norm": 1.733036994934082, + "learning_rate": 1.4023551040258726e-05, + "loss": 0.8507, + "step": 12276 + }, + { + "epoch": 0.3872259895915471, + "grad_norm": 2.0845155715942383, + "learning_rate": 1.4022615800976033e-05, + "loss": 0.458, + "step": 12277 + }, + { + "epoch": 0.3872575303579877, + "grad_norm": 2.430889368057251, + "learning_rate": 1.4021680519714224e-05, + "loss": 0.512, + "step": 12278 + }, + { + "epoch": 0.38728907112442834, + "grad_norm": 1.4988652467727661, + "learning_rate": 1.4020745196483059e-05, + "loss": 0.5081, + "step": 12279 + }, + { + "epoch": 0.38732061189086897, + "grad_norm": 1.774492621421814, + "learning_rate": 1.401980983129229e-05, + "loss": 0.5083, + "step": 12280 + }, + { + "epoch": 0.38735215265730955, + "grad_norm": 2.157759666442871, + "learning_rate": 1.4018874424151687e-05, + "loss": 0.391, + "step": 12281 + }, + { + "epoch": 0.3873836934237502, + "grad_norm": 1.472447395324707, + "learning_rate": 1.4017938975071009e-05, + "loss": 0.5043, + "step": 12282 + }, + { + "epoch": 0.3874152341901908, + "grad_norm": 1.433921217918396, + "learning_rate": 1.4017003484060015e-05, + "loss": 0.4746, + "step": 12283 + }, + { + "epoch": 0.38744677495663143, + "grad_norm": 1.823323369026184, + "learning_rate": 1.4016067951128471e-05, + "loss": 0.8057, + "step": 12284 + }, + { + "epoch": 0.38747831572307206, + "grad_norm": 1.4224441051483154, + "learning_rate": 1.4015132376286142e-05, + "loss": 0.4943, + "step": 12285 + }, + { + "epoch": 0.3875098564895127, + "grad_norm": 1.6642370223999023, + "learning_rate": 1.4014196759542787e-05, + "loss": 0.5254, + "step": 12286 + }, + { + "epoch": 0.3875413972559533, + "grad_norm": 1.6488043069839478, + "learning_rate": 1.4013261100908174e-05, + "loss": 0.5225, + "step": 12287 + }, + { + "epoch": 0.38757293802239395, + "grad_norm": 1.585338830947876, + "learning_rate": 1.4012325400392064e-05, + "loss": 0.4712, + "step": 12288 + }, + { + "epoch": 0.3876044787888346, + "grad_norm": 1.4063750505447388, + "learning_rate": 1.4011389658004224e-05, + "loss": 0.5144, + "step": 12289 + }, + { + "epoch": 0.3876360195552752, + "grad_norm": 1.7123161554336548, + "learning_rate": 1.4010453873754415e-05, + "loss": 0.5153, + "step": 12290 + }, + { + "epoch": 0.38766756032171584, + "grad_norm": 1.7939460277557373, + "learning_rate": 1.4009518047652408e-05, + "loss": 0.5037, + "step": 12291 + }, + { + "epoch": 0.3876991010881564, + "grad_norm": 1.7697370052337646, + "learning_rate": 1.4008582179707968e-05, + "loss": 0.8664, + "step": 12292 + }, + { + "epoch": 0.38773064185459705, + "grad_norm": 1.4758663177490234, + "learning_rate": 1.4007646269930859e-05, + "loss": 0.537, + "step": 12293 + }, + { + "epoch": 0.3877621826210377, + "grad_norm": 1.8045802116394043, + "learning_rate": 1.4006710318330846e-05, + "loss": 0.5133, + "step": 12294 + }, + { + "epoch": 0.3877937233874783, + "grad_norm": 1.645462989807129, + "learning_rate": 1.4005774324917706e-05, + "loss": 0.8549, + "step": 12295 + }, + { + "epoch": 0.38782526415391894, + "grad_norm": 2.0532870292663574, + "learning_rate": 1.4004838289701196e-05, + "loss": 0.8204, + "step": 12296 + }, + { + "epoch": 0.38785680492035957, + "grad_norm": 1.8060208559036255, + "learning_rate": 1.4003902212691091e-05, + "loss": 0.5027, + "step": 12297 + }, + { + "epoch": 0.3878883456868002, + "grad_norm": 1.4466643333435059, + "learning_rate": 1.4002966093897156e-05, + "loss": 0.5053, + "step": 12298 + }, + { + "epoch": 0.3879198864532408, + "grad_norm": 1.5381274223327637, + "learning_rate": 1.4002029933329167e-05, + "loss": 0.5557, + "step": 12299 + }, + { + "epoch": 0.38795142721968146, + "grad_norm": 1.1909953355789185, + "learning_rate": 1.4001093730996882e-05, + "loss": 0.5176, + "step": 12300 + }, + { + "epoch": 0.3879829679861221, + "grad_norm": 1.3693773746490479, + "learning_rate": 1.4000157486910077e-05, + "loss": 0.5283, + "step": 12301 + }, + { + "epoch": 0.3880145087525627, + "grad_norm": 1.807316780090332, + "learning_rate": 1.3999221201078526e-05, + "loss": 0.4674, + "step": 12302 + }, + { + "epoch": 0.3880460495190033, + "grad_norm": 1.5406816005706787, + "learning_rate": 1.3998284873511994e-05, + "loss": 0.5217, + "step": 12303 + }, + { + "epoch": 0.3880775902854439, + "grad_norm": 1.628138542175293, + "learning_rate": 1.3997348504220254e-05, + "loss": 0.5116, + "step": 12304 + }, + { + "epoch": 0.38810913105188455, + "grad_norm": 2.016026258468628, + "learning_rate": 1.3996412093213084e-05, + "loss": 0.5109, + "step": 12305 + }, + { + "epoch": 0.3881406718183252, + "grad_norm": 1.4386111497879028, + "learning_rate": 1.3995475640500243e-05, + "loss": 0.5311, + "step": 12306 + }, + { + "epoch": 0.3881722125847658, + "grad_norm": 5.5341796875, + "learning_rate": 1.3994539146091514e-05, + "loss": 0.5059, + "step": 12307 + }, + { + "epoch": 0.38820375335120644, + "grad_norm": 1.4523985385894775, + "learning_rate": 1.399360260999667e-05, + "loss": 0.5191, + "step": 12308 + }, + { + "epoch": 0.38823529411764707, + "grad_norm": 3.562936544418335, + "learning_rate": 1.3992666032225481e-05, + "loss": 0.5169, + "step": 12309 + }, + { + "epoch": 0.3882668348840877, + "grad_norm": 1.6233196258544922, + "learning_rate": 1.399172941278772e-05, + "loss": 0.5107, + "step": 12310 + }, + { + "epoch": 0.3882983756505283, + "grad_norm": 1.4369385242462158, + "learning_rate": 1.3990792751693163e-05, + "loss": 0.4774, + "step": 12311 + }, + { + "epoch": 0.38832991641696896, + "grad_norm": 5.984892845153809, + "learning_rate": 1.3989856048951587e-05, + "loss": 0.8435, + "step": 12312 + }, + { + "epoch": 0.38836145718340953, + "grad_norm": 2.192505121231079, + "learning_rate": 1.3988919304572763e-05, + "loss": 0.3797, + "step": 12313 + }, + { + "epoch": 0.38839299794985016, + "grad_norm": 1.5250661373138428, + "learning_rate": 1.398798251856647e-05, + "loss": 0.4891, + "step": 12314 + }, + { + "epoch": 0.3884245387162908, + "grad_norm": 1.7394959926605225, + "learning_rate": 1.3987045690942483e-05, + "loss": 0.512, + "step": 12315 + }, + { + "epoch": 0.3884560794827314, + "grad_norm": 1.5143253803253174, + "learning_rate": 1.3986108821710577e-05, + "loss": 0.3786, + "step": 12316 + }, + { + "epoch": 0.38848762024917205, + "grad_norm": 1.9525713920593262, + "learning_rate": 1.3985171910880531e-05, + "loss": 0.518, + "step": 12317 + }, + { + "epoch": 0.3885191610156127, + "grad_norm": 1.5744597911834717, + "learning_rate": 1.3984234958462124e-05, + "loss": 0.4654, + "step": 12318 + }, + { + "epoch": 0.3885507017820533, + "grad_norm": 1.3366901874542236, + "learning_rate": 1.398329796446513e-05, + "loss": 0.487, + "step": 12319 + }, + { + "epoch": 0.38858224254849394, + "grad_norm": 1.6010063886642456, + "learning_rate": 1.3982360928899326e-05, + "loss": 0.5375, + "step": 12320 + }, + { + "epoch": 0.38861378331493457, + "grad_norm": 1.659265160560608, + "learning_rate": 1.3981423851774496e-05, + "loss": 0.4505, + "step": 12321 + }, + { + "epoch": 0.3886453240813752, + "grad_norm": 1.3976867198944092, + "learning_rate": 1.3980486733100419e-05, + "loss": 0.5107, + "step": 12322 + }, + { + "epoch": 0.38867686484781583, + "grad_norm": 1.6834452152252197, + "learning_rate": 1.3979549572886866e-05, + "loss": 0.5187, + "step": 12323 + }, + { + "epoch": 0.3887084056142564, + "grad_norm": 1.9453297853469849, + "learning_rate": 1.3978612371143627e-05, + "loss": 0.5377, + "step": 12324 + }, + { + "epoch": 0.38873994638069703, + "grad_norm": 1.7024815082550049, + "learning_rate": 1.3977675127880477e-05, + "loss": 0.4701, + "step": 12325 + }, + { + "epoch": 0.38877148714713766, + "grad_norm": 2.2060675621032715, + "learning_rate": 1.3976737843107203e-05, + "loss": 0.4899, + "step": 12326 + }, + { + "epoch": 0.3888030279135783, + "grad_norm": 1.5394076108932495, + "learning_rate": 1.3975800516833576e-05, + "loss": 0.5217, + "step": 12327 + }, + { + "epoch": 0.3888345686800189, + "grad_norm": 1.7170542478561401, + "learning_rate": 1.3974863149069387e-05, + "loss": 0.5097, + "step": 12328 + }, + { + "epoch": 0.38886610944645955, + "grad_norm": 1.4099289178848267, + "learning_rate": 1.3973925739824414e-05, + "loss": 0.8644, + "step": 12329 + }, + { + "epoch": 0.3888976502129002, + "grad_norm": 1.8614698648452759, + "learning_rate": 1.3972988289108437e-05, + "loss": 0.5017, + "step": 12330 + }, + { + "epoch": 0.3889291909793408, + "grad_norm": 1.6974619626998901, + "learning_rate": 1.3972050796931246e-05, + "loss": 0.4933, + "step": 12331 + }, + { + "epoch": 0.38896073174578144, + "grad_norm": 1.5705790519714355, + "learning_rate": 1.3971113263302619e-05, + "loss": 0.5521, + "step": 12332 + }, + { + "epoch": 0.38899227251222207, + "grad_norm": 1.7514493465423584, + "learning_rate": 1.3970175688232344e-05, + "loss": 0.5167, + "step": 12333 + }, + { + "epoch": 0.38902381327866264, + "grad_norm": 2.4634346961975098, + "learning_rate": 1.39692380717302e-05, + "loss": 0.5241, + "step": 12334 + }, + { + "epoch": 0.3890553540451033, + "grad_norm": 1.4380520582199097, + "learning_rate": 1.3968300413805976e-05, + "loss": 0.5015, + "step": 12335 + }, + { + "epoch": 0.3890868948115439, + "grad_norm": 1.4376873970031738, + "learning_rate": 1.3967362714469457e-05, + "loss": 0.8364, + "step": 12336 + }, + { + "epoch": 0.38911843557798453, + "grad_norm": 1.6121957302093506, + "learning_rate": 1.3966424973730424e-05, + "loss": 0.8693, + "step": 12337 + }, + { + "epoch": 0.38914997634442516, + "grad_norm": 1.9188061952590942, + "learning_rate": 1.396548719159867e-05, + "loss": 0.4913, + "step": 12338 + }, + { + "epoch": 0.3891815171108658, + "grad_norm": 1.7235963344573975, + "learning_rate": 1.3964549368083977e-05, + "loss": 0.5409, + "step": 12339 + }, + { + "epoch": 0.3892130578773064, + "grad_norm": 2.073237180709839, + "learning_rate": 1.3963611503196134e-05, + "loss": 0.4759, + "step": 12340 + }, + { + "epoch": 0.38924459864374705, + "grad_norm": 1.3870189189910889, + "learning_rate": 1.3962673596944925e-05, + "loss": 0.5187, + "step": 12341 + }, + { + "epoch": 0.3892761394101877, + "grad_norm": 1.6022741794586182, + "learning_rate": 1.396173564934014e-05, + "loss": 0.4948, + "step": 12342 + }, + { + "epoch": 0.3893076801766283, + "grad_norm": 2.8160927295684814, + "learning_rate": 1.396079766039157e-05, + "loss": 0.3851, + "step": 12343 + }, + { + "epoch": 0.38933922094306894, + "grad_norm": 1.2723512649536133, + "learning_rate": 1.3959859630108998e-05, + "loss": 0.5079, + "step": 12344 + }, + { + "epoch": 0.3893707617095095, + "grad_norm": 1.606709599494934, + "learning_rate": 1.3958921558502216e-05, + "loss": 0.5319, + "step": 12345 + }, + { + "epoch": 0.38940230247595015, + "grad_norm": 1.9054110050201416, + "learning_rate": 1.3957983445581019e-05, + "loss": 0.3477, + "step": 12346 + }, + { + "epoch": 0.3894338432423908, + "grad_norm": 1.4158706665039062, + "learning_rate": 1.3957045291355185e-05, + "loss": 0.5256, + "step": 12347 + }, + { + "epoch": 0.3894653840088314, + "grad_norm": 1.3385685682296753, + "learning_rate": 1.3956107095834515e-05, + "loss": 0.5132, + "step": 12348 + }, + { + "epoch": 0.38949692477527204, + "grad_norm": 1.2741711139678955, + "learning_rate": 1.3955168859028794e-05, + "loss": 0.554, + "step": 12349 + }, + { + "epoch": 0.38952846554171267, + "grad_norm": 1.4587135314941406, + "learning_rate": 1.3954230580947814e-05, + "loss": 0.5054, + "step": 12350 + }, + { + "epoch": 0.3895600063081533, + "grad_norm": 1.6298145055770874, + "learning_rate": 1.3953292261601368e-05, + "loss": 0.5442, + "step": 12351 + }, + { + "epoch": 0.3895915470745939, + "grad_norm": 1.9874025583267212, + "learning_rate": 1.3952353900999248e-05, + "loss": 0.8417, + "step": 12352 + }, + { + "epoch": 0.38962308784103455, + "grad_norm": 1.4277242422103882, + "learning_rate": 1.3951415499151247e-05, + "loss": 0.4968, + "step": 12353 + }, + { + "epoch": 0.3896546286074752, + "grad_norm": 1.7708101272583008, + "learning_rate": 1.3950477056067154e-05, + "loss": 0.5024, + "step": 12354 + }, + { + "epoch": 0.3896861693739158, + "grad_norm": 1.4671753644943237, + "learning_rate": 1.3949538571756767e-05, + "loss": 0.4968, + "step": 12355 + }, + { + "epoch": 0.3897177101403564, + "grad_norm": 1.4059113264083862, + "learning_rate": 1.3948600046229878e-05, + "loss": 0.4936, + "step": 12356 + }, + { + "epoch": 0.389749250906797, + "grad_norm": 1.8278756141662598, + "learning_rate": 1.3947661479496283e-05, + "loss": 0.4944, + "step": 12357 + }, + { + "epoch": 0.38978079167323765, + "grad_norm": 1.739396572113037, + "learning_rate": 1.3946722871565776e-05, + "loss": 0.5077, + "step": 12358 + }, + { + "epoch": 0.3898123324396783, + "grad_norm": 1.4277278184890747, + "learning_rate": 1.3945784222448149e-05, + "loss": 0.5007, + "step": 12359 + }, + { + "epoch": 0.3898438732061189, + "grad_norm": 1.6283166408538818, + "learning_rate": 1.3944845532153201e-05, + "loss": 0.5044, + "step": 12360 + }, + { + "epoch": 0.38987541397255954, + "grad_norm": 1.3402281999588013, + "learning_rate": 1.3943906800690728e-05, + "loss": 0.4585, + "step": 12361 + }, + { + "epoch": 0.38990695473900017, + "grad_norm": 1.722828984260559, + "learning_rate": 1.394296802807052e-05, + "loss": 0.4819, + "step": 12362 + }, + { + "epoch": 0.3899384955054408, + "grad_norm": 1.9074821472167969, + "learning_rate": 1.3942029214302382e-05, + "loss": 0.5025, + "step": 12363 + }, + { + "epoch": 0.3899700362718814, + "grad_norm": 1.2752280235290527, + "learning_rate": 1.394109035939611e-05, + "loss": 0.4872, + "step": 12364 + }, + { + "epoch": 0.39000157703832206, + "grad_norm": 1.4736343622207642, + "learning_rate": 1.3940151463361498e-05, + "loss": 0.4779, + "step": 12365 + }, + { + "epoch": 0.39003311780476263, + "grad_norm": 1.9110453128814697, + "learning_rate": 1.3939212526208347e-05, + "loss": 0.5338, + "step": 12366 + }, + { + "epoch": 0.39006465857120326, + "grad_norm": 1.8985934257507324, + "learning_rate": 1.3938273547946455e-05, + "loss": 0.5243, + "step": 12367 + }, + { + "epoch": 0.3900961993376439, + "grad_norm": 1.7016035318374634, + "learning_rate": 1.393733452858562e-05, + "loss": 0.5087, + "step": 12368 + }, + { + "epoch": 0.3901277401040845, + "grad_norm": 1.532899022102356, + "learning_rate": 1.3936395468135642e-05, + "loss": 0.4902, + "step": 12369 + }, + { + "epoch": 0.39015928087052515, + "grad_norm": 2.4462406635284424, + "learning_rate": 1.3935456366606319e-05, + "loss": 0.5126, + "step": 12370 + }, + { + "epoch": 0.3901908216369658, + "grad_norm": 2.257459878921509, + "learning_rate": 1.3934517224007455e-05, + "loss": 0.8668, + "step": 12371 + }, + { + "epoch": 0.3902223624034064, + "grad_norm": 1.5594992637634277, + "learning_rate": 1.3933578040348848e-05, + "loss": 0.5329, + "step": 12372 + }, + { + "epoch": 0.39025390316984704, + "grad_norm": 1.7864221334457397, + "learning_rate": 1.39326388156403e-05, + "loss": 0.4899, + "step": 12373 + }, + { + "epoch": 0.39028544393628767, + "grad_norm": 1.4497934579849243, + "learning_rate": 1.3931699549891612e-05, + "loss": 0.4928, + "step": 12374 + }, + { + "epoch": 0.3903169847027283, + "grad_norm": 1.5514533519744873, + "learning_rate": 1.3930760243112587e-05, + "loss": 0.5322, + "step": 12375 + }, + { + "epoch": 0.3903485254691689, + "grad_norm": 1.563397765159607, + "learning_rate": 1.3929820895313027e-05, + "loss": 0.5187, + "step": 12376 + }, + { + "epoch": 0.3903800662356095, + "grad_norm": 1.3257962465286255, + "learning_rate": 1.3928881506502734e-05, + "loss": 0.5311, + "step": 12377 + }, + { + "epoch": 0.39041160700205013, + "grad_norm": 1.3146780729293823, + "learning_rate": 1.3927942076691512e-05, + "loss": 0.5092, + "step": 12378 + }, + { + "epoch": 0.39044314776849076, + "grad_norm": 1.6337143182754517, + "learning_rate": 1.3927002605889163e-05, + "loss": 0.7897, + "step": 12379 + }, + { + "epoch": 0.3904746885349314, + "grad_norm": 1.292171835899353, + "learning_rate": 1.3926063094105493e-05, + "loss": 0.5313, + "step": 12380 + }, + { + "epoch": 0.390506229301372, + "grad_norm": 1.4661751985549927, + "learning_rate": 1.3925123541350308e-05, + "loss": 0.7772, + "step": 12381 + }, + { + "epoch": 0.39053777006781265, + "grad_norm": 1.366384506225586, + "learning_rate": 1.392418394763341e-05, + "loss": 0.5194, + "step": 12382 + }, + { + "epoch": 0.3905693108342533, + "grad_norm": 1.9755579233169556, + "learning_rate": 1.3923244312964603e-05, + "loss": 0.5077, + "step": 12383 + }, + { + "epoch": 0.3906008516006939, + "grad_norm": 1.6731387376785278, + "learning_rate": 1.3922304637353698e-05, + "loss": 0.4999, + "step": 12384 + }, + { + "epoch": 0.39063239236713454, + "grad_norm": 1.7359728813171387, + "learning_rate": 1.3921364920810498e-05, + "loss": 0.5115, + "step": 12385 + }, + { + "epoch": 0.39066393313357517, + "grad_norm": 1.889565348625183, + "learning_rate": 1.392042516334481e-05, + "loss": 0.5533, + "step": 12386 + }, + { + "epoch": 0.39069547390001574, + "grad_norm": 1.2643636465072632, + "learning_rate": 1.3919485364966443e-05, + "loss": 0.4827, + "step": 12387 + }, + { + "epoch": 0.3907270146664564, + "grad_norm": 3.4724442958831787, + "learning_rate": 1.3918545525685204e-05, + "loss": 0.8088, + "step": 12388 + }, + { + "epoch": 0.390758555432897, + "grad_norm": 1.6351913213729858, + "learning_rate": 1.3917605645510894e-05, + "loss": 0.5293, + "step": 12389 + }, + { + "epoch": 0.39079009619933763, + "grad_norm": 1.5020956993103027, + "learning_rate": 1.3916665724453328e-05, + "loss": 0.4916, + "step": 12390 + }, + { + "epoch": 0.39082163696577826, + "grad_norm": 1.5165437459945679, + "learning_rate": 1.3915725762522318e-05, + "loss": 0.5022, + "step": 12391 + }, + { + "epoch": 0.3908531777322189, + "grad_norm": 1.7633681297302246, + "learning_rate": 1.3914785759727667e-05, + "loss": 0.508, + "step": 12392 + }, + { + "epoch": 0.3908847184986595, + "grad_norm": 1.7643730640411377, + "learning_rate": 1.3913845716079189e-05, + "loss": 0.5169, + "step": 12393 + }, + { + "epoch": 0.39091625926510015, + "grad_norm": 1.3455837965011597, + "learning_rate": 1.3912905631586689e-05, + "loss": 0.5383, + "step": 12394 + }, + { + "epoch": 0.3909478000315408, + "grad_norm": 1.2075495719909668, + "learning_rate": 1.391196550625998e-05, + "loss": 0.5243, + "step": 12395 + }, + { + "epoch": 0.3909793407979814, + "grad_norm": 2.5207486152648926, + "learning_rate": 1.3911025340108876e-05, + "loss": 0.5045, + "step": 12396 + }, + { + "epoch": 0.39101088156442204, + "grad_norm": 1.471681833267212, + "learning_rate": 1.3910085133143186e-05, + "loss": 0.5143, + "step": 12397 + }, + { + "epoch": 0.3910424223308626, + "grad_norm": 1.8378548622131348, + "learning_rate": 1.390914488537272e-05, + "loss": 0.5239, + "step": 12398 + }, + { + "epoch": 0.39107396309730325, + "grad_norm": 1.3337888717651367, + "learning_rate": 1.3908204596807293e-05, + "loss": 0.4874, + "step": 12399 + }, + { + "epoch": 0.3911055038637439, + "grad_norm": 1.4943896532058716, + "learning_rate": 1.3907264267456714e-05, + "loss": 0.5219, + "step": 12400 + }, + { + "epoch": 0.3911370446301845, + "grad_norm": 1.513304591178894, + "learning_rate": 1.3906323897330803e-05, + "loss": 0.497, + "step": 12401 + }, + { + "epoch": 0.39116858539662513, + "grad_norm": 4.248316764831543, + "learning_rate": 1.3905383486439364e-05, + "loss": 0.5, + "step": 12402 + }, + { + "epoch": 0.39120012616306576, + "grad_norm": 1.8111629486083984, + "learning_rate": 1.3904443034792217e-05, + "loss": 0.4721, + "step": 12403 + }, + { + "epoch": 0.3912316669295064, + "grad_norm": 1.372909426689148, + "learning_rate": 1.3903502542399175e-05, + "loss": 0.5033, + "step": 12404 + }, + { + "epoch": 0.391263207695947, + "grad_norm": 1.2965846061706543, + "learning_rate": 1.3902562009270053e-05, + "loss": 0.5001, + "step": 12405 + }, + { + "epoch": 0.39129474846238765, + "grad_norm": 2.9308552742004395, + "learning_rate": 1.390162143541467e-05, + "loss": 0.8233, + "step": 12406 + }, + { + "epoch": 0.3913262892288283, + "grad_norm": 1.768820881843567, + "learning_rate": 1.3900680820842833e-05, + "loss": 0.4973, + "step": 12407 + }, + { + "epoch": 0.39135782999526886, + "grad_norm": 2.1543729305267334, + "learning_rate": 1.3899740165564366e-05, + "loss": 0.4415, + "step": 12408 + }, + { + "epoch": 0.3913893707617095, + "grad_norm": 1.8939570188522339, + "learning_rate": 1.389879946958908e-05, + "loss": 0.5038, + "step": 12409 + }, + { + "epoch": 0.3914209115281501, + "grad_norm": 1.8511303663253784, + "learning_rate": 1.3897858732926794e-05, + "loss": 0.5315, + "step": 12410 + }, + { + "epoch": 0.39145245229459075, + "grad_norm": 1.6280107498168945, + "learning_rate": 1.3896917955587328e-05, + "loss": 0.5343, + "step": 12411 + }, + { + "epoch": 0.3914839930610314, + "grad_norm": 1.3928121328353882, + "learning_rate": 1.3895977137580496e-05, + "loss": 0.4687, + "step": 12412 + }, + { + "epoch": 0.391515533827472, + "grad_norm": 1.5762205123901367, + "learning_rate": 1.3895036278916117e-05, + "loss": 0.4609, + "step": 12413 + }, + { + "epoch": 0.39154707459391264, + "grad_norm": 2.307025909423828, + "learning_rate": 1.389409537960401e-05, + "loss": 0.5228, + "step": 12414 + }, + { + "epoch": 0.39157861536035327, + "grad_norm": 1.7643567323684692, + "learning_rate": 1.3893154439653995e-05, + "loss": 0.3898, + "step": 12415 + }, + { + "epoch": 0.3916101561267939, + "grad_norm": 2.7770910263061523, + "learning_rate": 1.3892213459075892e-05, + "loss": 0.3774, + "step": 12416 + }, + { + "epoch": 0.3916416968932345, + "grad_norm": 1.5896669626235962, + "learning_rate": 1.3891272437879515e-05, + "loss": 0.5291, + "step": 12417 + }, + { + "epoch": 0.39167323765967516, + "grad_norm": 1.3745293617248535, + "learning_rate": 1.389033137607469e-05, + "loss": 0.5351, + "step": 12418 + }, + { + "epoch": 0.39170477842611573, + "grad_norm": 1.7640784978866577, + "learning_rate": 1.388939027367124e-05, + "loss": 0.5215, + "step": 12419 + }, + { + "epoch": 0.39173631919255636, + "grad_norm": 2.2148818969726562, + "learning_rate": 1.3888449130678977e-05, + "loss": 0.4967, + "step": 12420 + }, + { + "epoch": 0.391767859958997, + "grad_norm": 9.423383712768555, + "learning_rate": 1.388750794710773e-05, + "loss": 0.5191, + "step": 12421 + }, + { + "epoch": 0.3917994007254376, + "grad_norm": 1.1477291584014893, + "learning_rate": 1.388656672296732e-05, + "loss": 0.5299, + "step": 12422 + }, + { + "epoch": 0.39183094149187825, + "grad_norm": 1.238606333732605, + "learning_rate": 1.3885625458267569e-05, + "loss": 0.5194, + "step": 12423 + }, + { + "epoch": 0.3918624822583189, + "grad_norm": 1.6727170944213867, + "learning_rate": 1.3884684153018297e-05, + "loss": 0.5116, + "step": 12424 + }, + { + "epoch": 0.3918940230247595, + "grad_norm": 2.498108386993408, + "learning_rate": 1.3883742807229333e-05, + "loss": 0.5175, + "step": 12425 + }, + { + "epoch": 0.39192556379120014, + "grad_norm": 1.4993215799331665, + "learning_rate": 1.3882801420910495e-05, + "loss": 0.7944, + "step": 12426 + }, + { + "epoch": 0.39195710455764077, + "grad_norm": 1.4705829620361328, + "learning_rate": 1.3881859994071607e-05, + "loss": 0.5082, + "step": 12427 + }, + { + "epoch": 0.3919886453240814, + "grad_norm": 1.9031108617782593, + "learning_rate": 1.3880918526722497e-05, + "loss": 0.5647, + "step": 12428 + }, + { + "epoch": 0.392020186090522, + "grad_norm": 1.9147840738296509, + "learning_rate": 1.3879977018872991e-05, + "loss": 0.4256, + "step": 12429 + }, + { + "epoch": 0.3920517268569626, + "grad_norm": 1.943358302116394, + "learning_rate": 1.387903547053291e-05, + "loss": 0.4939, + "step": 12430 + }, + { + "epoch": 0.39208326762340323, + "grad_norm": 1.4651998281478882, + "learning_rate": 1.3878093881712083e-05, + "loss": 0.5315, + "step": 12431 + }, + { + "epoch": 0.39211480838984386, + "grad_norm": 1.4776771068572998, + "learning_rate": 1.3877152252420337e-05, + "loss": 0.4949, + "step": 12432 + }, + { + "epoch": 0.3921463491562845, + "grad_norm": 2.122972249984741, + "learning_rate": 1.3876210582667493e-05, + "loss": 0.5299, + "step": 12433 + }, + { + "epoch": 0.3921778899227251, + "grad_norm": 1.7504441738128662, + "learning_rate": 1.3875268872463382e-05, + "loss": 0.406, + "step": 12434 + }, + { + "epoch": 0.39220943068916575, + "grad_norm": 1.6611876487731934, + "learning_rate": 1.3874327121817835e-05, + "loss": 0.506, + "step": 12435 + }, + { + "epoch": 0.3922409714556064, + "grad_norm": 1.8443118333816528, + "learning_rate": 1.3873385330740674e-05, + "loss": 0.4938, + "step": 12436 + }, + { + "epoch": 0.392272512222047, + "grad_norm": 1.5085152387619019, + "learning_rate": 1.3872443499241728e-05, + "loss": 0.5189, + "step": 12437 + }, + { + "epoch": 0.39230405298848764, + "grad_norm": 1.441650152206421, + "learning_rate": 1.3871501627330828e-05, + "loss": 0.4929, + "step": 12438 + }, + { + "epoch": 0.39233559375492827, + "grad_norm": 1.6548805236816406, + "learning_rate": 1.3870559715017802e-05, + "loss": 0.5447, + "step": 12439 + }, + { + "epoch": 0.39236713452136884, + "grad_norm": 1.5175114870071411, + "learning_rate": 1.3869617762312481e-05, + "loss": 0.5002, + "step": 12440 + }, + { + "epoch": 0.3923986752878095, + "grad_norm": 2.3396995067596436, + "learning_rate": 1.386867576922469e-05, + "loss": 0.4879, + "step": 12441 + }, + { + "epoch": 0.3924302160542501, + "grad_norm": 1.5847529172897339, + "learning_rate": 1.3867733735764268e-05, + "loss": 0.5047, + "step": 12442 + }, + { + "epoch": 0.39246175682069073, + "grad_norm": 1.7844491004943848, + "learning_rate": 1.3866791661941042e-05, + "loss": 0.5024, + "step": 12443 + }, + { + "epoch": 0.39249329758713136, + "grad_norm": 1.5947697162628174, + "learning_rate": 1.386584954776484e-05, + "loss": 0.488, + "step": 12444 + }, + { + "epoch": 0.392524838353572, + "grad_norm": 1.398177146911621, + "learning_rate": 1.3864907393245499e-05, + "loss": 0.536, + "step": 12445 + }, + { + "epoch": 0.3925563791200126, + "grad_norm": 1.7000932693481445, + "learning_rate": 1.3863965198392845e-05, + "loss": 0.5564, + "step": 12446 + }, + { + "epoch": 0.39258791988645325, + "grad_norm": 1.8024547100067139, + "learning_rate": 1.3863022963216714e-05, + "loss": 0.4876, + "step": 12447 + }, + { + "epoch": 0.3926194606528939, + "grad_norm": 1.4170958995819092, + "learning_rate": 1.3862080687726942e-05, + "loss": 0.5242, + "step": 12448 + }, + { + "epoch": 0.3926510014193345, + "grad_norm": 1.83929443359375, + "learning_rate": 1.3861138371933355e-05, + "loss": 0.5003, + "step": 12449 + }, + { + "epoch": 0.39268254218577514, + "grad_norm": 1.5302072763442993, + "learning_rate": 1.3860196015845795e-05, + "loss": 0.4646, + "step": 12450 + }, + { + "epoch": 0.3927140829522157, + "grad_norm": 2.367669105529785, + "learning_rate": 1.385925361947409e-05, + "loss": 0.5123, + "step": 12451 + }, + { + "epoch": 0.39274562371865634, + "grad_norm": 2.6742334365844727, + "learning_rate": 1.3858311182828075e-05, + "loss": 0.4838, + "step": 12452 + }, + { + "epoch": 0.392777164485097, + "grad_norm": 1.5501421689987183, + "learning_rate": 1.3857368705917592e-05, + "loss": 0.5097, + "step": 12453 + }, + { + "epoch": 0.3928087052515376, + "grad_norm": 1.8286004066467285, + "learning_rate": 1.3856426188752466e-05, + "loss": 0.8485, + "step": 12454 + }, + { + "epoch": 0.39284024601797823, + "grad_norm": 1.591304898262024, + "learning_rate": 1.3855483631342543e-05, + "loss": 0.485, + "step": 12455 + }, + { + "epoch": 0.39287178678441886, + "grad_norm": 1.3931139707565308, + "learning_rate": 1.3854541033697652e-05, + "loss": 0.5128, + "step": 12456 + }, + { + "epoch": 0.3929033275508595, + "grad_norm": 3.696664333343506, + "learning_rate": 1.3853598395827633e-05, + "loss": 0.5151, + "step": 12457 + }, + { + "epoch": 0.3929348683173001, + "grad_norm": 2.4738035202026367, + "learning_rate": 1.385265571774232e-05, + "loss": 0.5409, + "step": 12458 + }, + { + "epoch": 0.39296640908374075, + "grad_norm": 1.4196876287460327, + "learning_rate": 1.3851712999451557e-05, + "loss": 0.4799, + "step": 12459 + }, + { + "epoch": 0.3929979498501814, + "grad_norm": 2.7828378677368164, + "learning_rate": 1.3850770240965177e-05, + "loss": 0.5294, + "step": 12460 + }, + { + "epoch": 0.39302949061662196, + "grad_norm": 3.007871150970459, + "learning_rate": 1.3849827442293019e-05, + "loss": 0.5058, + "step": 12461 + }, + { + "epoch": 0.3930610313830626, + "grad_norm": 2.189709186553955, + "learning_rate": 1.3848884603444922e-05, + "loss": 0.5242, + "step": 12462 + }, + { + "epoch": 0.3930925721495032, + "grad_norm": 2.5790345668792725, + "learning_rate": 1.3847941724430724e-05, + "loss": 0.4272, + "step": 12463 + }, + { + "epoch": 0.39312411291594385, + "grad_norm": 1.4642431735992432, + "learning_rate": 1.3846998805260268e-05, + "loss": 0.5183, + "step": 12464 + }, + { + "epoch": 0.3931556536823845, + "grad_norm": 1.766913652420044, + "learning_rate": 1.3846055845943393e-05, + "loss": 0.5114, + "step": 12465 + }, + { + "epoch": 0.3931871944488251, + "grad_norm": 1.8041424751281738, + "learning_rate": 1.3845112846489938e-05, + "loss": 0.4949, + "step": 12466 + }, + { + "epoch": 0.39321873521526574, + "grad_norm": 3.3222196102142334, + "learning_rate": 1.3844169806909746e-05, + "loss": 0.5053, + "step": 12467 + }, + { + "epoch": 0.39325027598170637, + "grad_norm": 1.4865564107894897, + "learning_rate": 1.3843226727212655e-05, + "loss": 0.5373, + "step": 12468 + }, + { + "epoch": 0.393281816748147, + "grad_norm": 1.4338854551315308, + "learning_rate": 1.3842283607408509e-05, + "loss": 0.5004, + "step": 12469 + }, + { + "epoch": 0.3933133575145876, + "grad_norm": 1.4420909881591797, + "learning_rate": 1.3841340447507152e-05, + "loss": 0.5431, + "step": 12470 + }, + { + "epoch": 0.39334489828102825, + "grad_norm": 1.4379594326019287, + "learning_rate": 1.3840397247518425e-05, + "loss": 0.8509, + "step": 12471 + }, + { + "epoch": 0.39337643904746883, + "grad_norm": 2.7084615230560303, + "learning_rate": 1.3839454007452168e-05, + "loss": 0.5035, + "step": 12472 + }, + { + "epoch": 0.39340797981390946, + "grad_norm": 1.7869526147842407, + "learning_rate": 1.383851072731823e-05, + "loss": 0.5246, + "step": 12473 + }, + { + "epoch": 0.3934395205803501, + "grad_norm": 1.5251038074493408, + "learning_rate": 1.3837567407126451e-05, + "loss": 0.3694, + "step": 12474 + }, + { + "epoch": 0.3934710613467907, + "grad_norm": 1.5072625875473022, + "learning_rate": 1.3836624046886678e-05, + "loss": 0.4973, + "step": 12475 + }, + { + "epoch": 0.39350260211323135, + "grad_norm": 1.6818082332611084, + "learning_rate": 1.3835680646608749e-05, + "loss": 0.541, + "step": 12476 + }, + { + "epoch": 0.393534142879672, + "grad_norm": 1.7490010261535645, + "learning_rate": 1.3834737206302519e-05, + "loss": 0.505, + "step": 12477 + }, + { + "epoch": 0.3935656836461126, + "grad_norm": 2.074500560760498, + "learning_rate": 1.3833793725977826e-05, + "loss": 0.5113, + "step": 12478 + }, + { + "epoch": 0.39359722441255324, + "grad_norm": 2.272930383682251, + "learning_rate": 1.383285020564452e-05, + "loss": 0.5296, + "step": 12479 + }, + { + "epoch": 0.39362876517899387, + "grad_norm": 2.290822982788086, + "learning_rate": 1.3831906645312446e-05, + "loss": 0.5349, + "step": 12480 + }, + { + "epoch": 0.3936603059454345, + "grad_norm": 1.7067227363586426, + "learning_rate": 1.3830963044991448e-05, + "loss": 0.5322, + "step": 12481 + }, + { + "epoch": 0.39369184671187507, + "grad_norm": 1.6201881170272827, + "learning_rate": 1.383001940469138e-05, + "loss": 0.849, + "step": 12482 + }, + { + "epoch": 0.3937233874783157, + "grad_norm": 1.5266591310501099, + "learning_rate": 1.3829075724422081e-05, + "loss": 0.5046, + "step": 12483 + }, + { + "epoch": 0.39375492824475633, + "grad_norm": 1.7103009223937988, + "learning_rate": 1.382813200419341e-05, + "loss": 0.5536, + "step": 12484 + }, + { + "epoch": 0.39378646901119696, + "grad_norm": 1.831491231918335, + "learning_rate": 1.3827188244015206e-05, + "loss": 0.5538, + "step": 12485 + }, + { + "epoch": 0.3938180097776376, + "grad_norm": 2.3965227603912354, + "learning_rate": 1.3826244443897318e-05, + "loss": 0.5336, + "step": 12486 + }, + { + "epoch": 0.3938495505440782, + "grad_norm": 1.2659194469451904, + "learning_rate": 1.3825300603849603e-05, + "loss": 0.8491, + "step": 12487 + }, + { + "epoch": 0.39388109131051885, + "grad_norm": 2.590583086013794, + "learning_rate": 1.3824356723881903e-05, + "loss": 0.4828, + "step": 12488 + }, + { + "epoch": 0.3939126320769595, + "grad_norm": 1.8722593784332275, + "learning_rate": 1.382341280400407e-05, + "loss": 0.4926, + "step": 12489 + }, + { + "epoch": 0.3939441728434001, + "grad_norm": 1.6683528423309326, + "learning_rate": 1.3822468844225956e-05, + "loss": 0.4945, + "step": 12490 + }, + { + "epoch": 0.39397571360984074, + "grad_norm": 1.8300620317459106, + "learning_rate": 1.3821524844557414e-05, + "loss": 0.5047, + "step": 12491 + }, + { + "epoch": 0.39400725437628137, + "grad_norm": 1.7962629795074463, + "learning_rate": 1.382058080500829e-05, + "loss": 0.5148, + "step": 12492 + }, + { + "epoch": 0.39403879514272194, + "grad_norm": 1.6339328289031982, + "learning_rate": 1.381963672558844e-05, + "loss": 0.4922, + "step": 12493 + }, + { + "epoch": 0.3940703359091626, + "grad_norm": 2.0414013862609863, + "learning_rate": 1.3818692606307717e-05, + "loss": 0.5258, + "step": 12494 + }, + { + "epoch": 0.3941018766756032, + "grad_norm": 1.4622944593429565, + "learning_rate": 1.3817748447175972e-05, + "loss": 0.505, + "step": 12495 + }, + { + "epoch": 0.39413341744204383, + "grad_norm": 1.5178897380828857, + "learning_rate": 1.3816804248203053e-05, + "loss": 0.5381, + "step": 12496 + }, + { + "epoch": 0.39416495820848446, + "grad_norm": 1.6879125833511353, + "learning_rate": 1.3815860009398823e-05, + "loss": 0.5111, + "step": 12497 + }, + { + "epoch": 0.3941964989749251, + "grad_norm": 1.5704494714736938, + "learning_rate": 1.3814915730773127e-05, + "loss": 0.4575, + "step": 12498 + }, + { + "epoch": 0.3942280397413657, + "grad_norm": 1.5937690734863281, + "learning_rate": 1.3813971412335823e-05, + "loss": 0.5061, + "step": 12499 + }, + { + "epoch": 0.39425958050780635, + "grad_norm": 1.6759337186813354, + "learning_rate": 1.3813027054096766e-05, + "loss": 0.3698, + "step": 12500 + }, + { + "epoch": 0.394291121274247, + "grad_norm": 1.4929413795471191, + "learning_rate": 1.3812082656065812e-05, + "loss": 0.5128, + "step": 12501 + }, + { + "epoch": 0.3943226620406876, + "grad_norm": 2.175351619720459, + "learning_rate": 1.3811138218252817e-05, + "loss": 0.8396, + "step": 12502 + }, + { + "epoch": 0.39435420280712824, + "grad_norm": 1.4826937913894653, + "learning_rate": 1.3810193740667631e-05, + "loss": 0.8247, + "step": 12503 + }, + { + "epoch": 0.3943857435735688, + "grad_norm": 2.080617666244507, + "learning_rate": 1.380924922332012e-05, + "loss": 0.5164, + "step": 12504 + }, + { + "epoch": 0.39441728434000944, + "grad_norm": 1.5416364669799805, + "learning_rate": 1.3808304666220136e-05, + "loss": 0.391, + "step": 12505 + }, + { + "epoch": 0.3944488251064501, + "grad_norm": 3.1290671825408936, + "learning_rate": 1.3807360069377531e-05, + "loss": 0.5056, + "step": 12506 + }, + { + "epoch": 0.3944803658728907, + "grad_norm": 1.6692452430725098, + "learning_rate": 1.3806415432802169e-05, + "loss": 0.4786, + "step": 12507 + }, + { + "epoch": 0.39451190663933133, + "grad_norm": 1.5277645587921143, + "learning_rate": 1.3805470756503909e-05, + "loss": 0.5226, + "step": 12508 + }, + { + "epoch": 0.39454344740577196, + "grad_norm": 1.820713758468628, + "learning_rate": 1.3804526040492602e-05, + "loss": 0.5082, + "step": 12509 + }, + { + "epoch": 0.3945749881722126, + "grad_norm": 1.4762110710144043, + "learning_rate": 1.3803581284778114e-05, + "loss": 0.3886, + "step": 12510 + }, + { + "epoch": 0.3946065289386532, + "grad_norm": 1.8147941827774048, + "learning_rate": 1.3802636489370303e-05, + "loss": 0.514, + "step": 12511 + }, + { + "epoch": 0.39463806970509385, + "grad_norm": 5.169741630554199, + "learning_rate": 1.3801691654279025e-05, + "loss": 0.4954, + "step": 12512 + }, + { + "epoch": 0.3946696104715345, + "grad_norm": 1.7734582424163818, + "learning_rate": 1.3800746779514144e-05, + "loss": 0.5245, + "step": 12513 + }, + { + "epoch": 0.39470115123797506, + "grad_norm": 1.5970767736434937, + "learning_rate": 1.379980186508552e-05, + "loss": 0.5161, + "step": 12514 + }, + { + "epoch": 0.3947326920044157, + "grad_norm": 1.5557875633239746, + "learning_rate": 1.3798856911003016e-05, + "loss": 0.5039, + "step": 12515 + }, + { + "epoch": 0.3947642327708563, + "grad_norm": 1.4898219108581543, + "learning_rate": 1.3797911917276486e-05, + "loss": 0.5358, + "step": 12516 + }, + { + "epoch": 0.39479577353729695, + "grad_norm": 1.5309375524520874, + "learning_rate": 1.3796966883915795e-05, + "loss": 0.5029, + "step": 12517 + }, + { + "epoch": 0.3948273143037376, + "grad_norm": 1.393518090248108, + "learning_rate": 1.379602181093081e-05, + "loss": 0.5082, + "step": 12518 + }, + { + "epoch": 0.3948588550701782, + "grad_norm": 1.9389210939407349, + "learning_rate": 1.3795076698331386e-05, + "loss": 0.5168, + "step": 12519 + }, + { + "epoch": 0.39489039583661883, + "grad_norm": 3.4419591426849365, + "learning_rate": 1.3794131546127392e-05, + "loss": 0.3892, + "step": 12520 + }, + { + "epoch": 0.39492193660305946, + "grad_norm": 1.9114927053451538, + "learning_rate": 1.3793186354328692e-05, + "loss": 0.5248, + "step": 12521 + }, + { + "epoch": 0.3949534773695001, + "grad_norm": 1.7885828018188477, + "learning_rate": 1.3792241122945142e-05, + "loss": 0.3998, + "step": 12522 + }, + { + "epoch": 0.3949850181359407, + "grad_norm": 1.4404377937316895, + "learning_rate": 1.3791295851986612e-05, + "loss": 0.5274, + "step": 12523 + }, + { + "epoch": 0.39501655890238135, + "grad_norm": 1.9400110244750977, + "learning_rate": 1.379035054146297e-05, + "loss": 0.5301, + "step": 12524 + }, + { + "epoch": 0.39504809966882193, + "grad_norm": 1.8999826908111572, + "learning_rate": 1.3789405191384075e-05, + "loss": 0.5174, + "step": 12525 + }, + { + "epoch": 0.39507964043526256, + "grad_norm": 1.4837497472763062, + "learning_rate": 1.3788459801759793e-05, + "loss": 0.509, + "step": 12526 + }, + { + "epoch": 0.3951111812017032, + "grad_norm": 1.6665037870407104, + "learning_rate": 1.3787514372599993e-05, + "loss": 0.8243, + "step": 12527 + }, + { + "epoch": 0.3951427219681438, + "grad_norm": 1.5956422090530396, + "learning_rate": 1.3786568903914539e-05, + "loss": 0.5243, + "step": 12528 + }, + { + "epoch": 0.39517426273458445, + "grad_norm": 1.6584724187850952, + "learning_rate": 1.3785623395713298e-05, + "loss": 0.8438, + "step": 12529 + }, + { + "epoch": 0.3952058035010251, + "grad_norm": 1.7406084537506104, + "learning_rate": 1.3784677848006136e-05, + "loss": 0.488, + "step": 12530 + }, + { + "epoch": 0.3952373442674657, + "grad_norm": 1.8185012340545654, + "learning_rate": 1.3783732260802922e-05, + "loss": 0.5072, + "step": 12531 + }, + { + "epoch": 0.39526888503390634, + "grad_norm": 1.906829833984375, + "learning_rate": 1.3782786634113528e-05, + "loss": 0.8591, + "step": 12532 + }, + { + "epoch": 0.39530042580034697, + "grad_norm": 1.4867095947265625, + "learning_rate": 1.3781840967947815e-05, + "loss": 0.5099, + "step": 12533 + }, + { + "epoch": 0.3953319665667876, + "grad_norm": 1.7667182683944702, + "learning_rate": 1.3780895262315657e-05, + "loss": 0.5018, + "step": 12534 + }, + { + "epoch": 0.39536350733322817, + "grad_norm": 2.2139034271240234, + "learning_rate": 1.3779949517226921e-05, + "loss": 0.4738, + "step": 12535 + }, + { + "epoch": 0.3953950480996688, + "grad_norm": 1.7369784116744995, + "learning_rate": 1.3779003732691475e-05, + "loss": 0.5241, + "step": 12536 + }, + { + "epoch": 0.39542658886610943, + "grad_norm": 1.5192445516586304, + "learning_rate": 1.3778057908719189e-05, + "loss": 0.5037, + "step": 12537 + }, + { + "epoch": 0.39545812963255006, + "grad_norm": 1.7611562013626099, + "learning_rate": 1.3777112045319936e-05, + "loss": 0.5136, + "step": 12538 + }, + { + "epoch": 0.3954896703989907, + "grad_norm": 1.571912169456482, + "learning_rate": 1.3776166142503589e-05, + "loss": 0.4789, + "step": 12539 + }, + { + "epoch": 0.3955212111654313, + "grad_norm": 1.6118360757827759, + "learning_rate": 1.3775220200280015e-05, + "loss": 0.4648, + "step": 12540 + }, + { + "epoch": 0.39555275193187195, + "grad_norm": 1.5102561712265015, + "learning_rate": 1.3774274218659086e-05, + "loss": 0.5021, + "step": 12541 + }, + { + "epoch": 0.3955842926983126, + "grad_norm": 2.497863531112671, + "learning_rate": 1.3773328197650675e-05, + "loss": 0.8589, + "step": 12542 + }, + { + "epoch": 0.3956158334647532, + "grad_norm": 1.7079728841781616, + "learning_rate": 1.3772382137264653e-05, + "loss": 0.4735, + "step": 12543 + }, + { + "epoch": 0.39564737423119384, + "grad_norm": 2.5017757415771484, + "learning_rate": 1.3771436037510897e-05, + "loss": 0.4663, + "step": 12544 + }, + { + "epoch": 0.39567891499763447, + "grad_norm": 2.2184321880340576, + "learning_rate": 1.3770489898399278e-05, + "loss": 0.5233, + "step": 12545 + }, + { + "epoch": 0.39571045576407504, + "grad_norm": 1.5938491821289062, + "learning_rate": 1.3769543719939667e-05, + "loss": 0.5038, + "step": 12546 + }, + { + "epoch": 0.39574199653051567, + "grad_norm": 1.8693163394927979, + "learning_rate": 1.3768597502141938e-05, + "loss": 0.5548, + "step": 12547 + }, + { + "epoch": 0.3957735372969563, + "grad_norm": 2.6793532371520996, + "learning_rate": 1.3767651245015969e-05, + "loss": 0.3795, + "step": 12548 + }, + { + "epoch": 0.39580507806339693, + "grad_norm": 1.4962284564971924, + "learning_rate": 1.376670494857164e-05, + "loss": 0.481, + "step": 12549 + }, + { + "epoch": 0.39583661882983756, + "grad_norm": 1.4836138486862183, + "learning_rate": 1.3765758612818813e-05, + "loss": 0.4994, + "step": 12550 + }, + { + "epoch": 0.3958681595962782, + "grad_norm": 1.537824034690857, + "learning_rate": 1.3764812237767373e-05, + "loss": 0.5208, + "step": 12551 + }, + { + "epoch": 0.3958997003627188, + "grad_norm": 1.846458911895752, + "learning_rate": 1.3763865823427194e-05, + "loss": 0.5251, + "step": 12552 + }, + { + "epoch": 0.39593124112915945, + "grad_norm": 1.3910818099975586, + "learning_rate": 1.3762919369808153e-05, + "loss": 0.5007, + "step": 12553 + }, + { + "epoch": 0.3959627818956001, + "grad_norm": 1.8997182846069336, + "learning_rate": 1.3761972876920127e-05, + "loss": 0.8355, + "step": 12554 + }, + { + "epoch": 0.3959943226620407, + "grad_norm": 1.5890958309173584, + "learning_rate": 1.3761026344772991e-05, + "loss": 0.5294, + "step": 12555 + }, + { + "epoch": 0.39602586342848134, + "grad_norm": 1.706575632095337, + "learning_rate": 1.3760079773376629e-05, + "loss": 0.5098, + "step": 12556 + }, + { + "epoch": 0.3960574041949219, + "grad_norm": 1.9098174571990967, + "learning_rate": 1.375913316274091e-05, + "loss": 0.5282, + "step": 12557 + }, + { + "epoch": 0.39608894496136254, + "grad_norm": 1.6163804531097412, + "learning_rate": 1.3758186512875717e-05, + "loss": 0.5123, + "step": 12558 + }, + { + "epoch": 0.3961204857278032, + "grad_norm": 1.672432541847229, + "learning_rate": 1.3757239823790935e-05, + "loss": 0.5045, + "step": 12559 + }, + { + "epoch": 0.3961520264942438, + "grad_norm": 1.8352702856063843, + "learning_rate": 1.3756293095496434e-05, + "loss": 0.4952, + "step": 12560 + }, + { + "epoch": 0.39618356726068443, + "grad_norm": 1.2476489543914795, + "learning_rate": 1.3755346328002097e-05, + "loss": 0.5045, + "step": 12561 + }, + { + "epoch": 0.39621510802712506, + "grad_norm": 1.6455470323562622, + "learning_rate": 1.3754399521317809e-05, + "loss": 0.8386, + "step": 12562 + }, + { + "epoch": 0.3962466487935657, + "grad_norm": 1.636009931564331, + "learning_rate": 1.3753452675453444e-05, + "loss": 0.5074, + "step": 12563 + }, + { + "epoch": 0.3962781895600063, + "grad_norm": 2.0923478603363037, + "learning_rate": 1.375250579041889e-05, + "loss": 0.4826, + "step": 12564 + }, + { + "epoch": 0.39630973032644695, + "grad_norm": 1.8532260656356812, + "learning_rate": 1.3751558866224018e-05, + "loss": 0.5035, + "step": 12565 + }, + { + "epoch": 0.3963412710928876, + "grad_norm": 1.839982032775879, + "learning_rate": 1.375061190287872e-05, + "loss": 0.4838, + "step": 12566 + }, + { + "epoch": 0.39637281185932816, + "grad_norm": 1.688146948814392, + "learning_rate": 1.3749664900392873e-05, + "loss": 0.8225, + "step": 12567 + }, + { + "epoch": 0.3964043526257688, + "grad_norm": 1.4524208307266235, + "learning_rate": 1.3748717858776362e-05, + "loss": 0.5122, + "step": 12568 + }, + { + "epoch": 0.3964358933922094, + "grad_norm": 1.4569734334945679, + "learning_rate": 1.374777077803907e-05, + "loss": 0.5157, + "step": 12569 + }, + { + "epoch": 0.39646743415865005, + "grad_norm": 1.6318241357803345, + "learning_rate": 1.374682365819088e-05, + "loss": 0.5273, + "step": 12570 + }, + { + "epoch": 0.3964989749250907, + "grad_norm": 1.6824396848678589, + "learning_rate": 1.3745876499241672e-05, + "loss": 0.524, + "step": 12571 + }, + { + "epoch": 0.3965305156915313, + "grad_norm": 1.6881024837493896, + "learning_rate": 1.374492930120134e-05, + "loss": 0.5063, + "step": 12572 + }, + { + "epoch": 0.39656205645797193, + "grad_norm": 1.5314644575119019, + "learning_rate": 1.374398206407976e-05, + "loss": 0.4019, + "step": 12573 + }, + { + "epoch": 0.39659359722441256, + "grad_norm": 1.7586486339569092, + "learning_rate": 1.374303478788682e-05, + "loss": 0.4765, + "step": 12574 + }, + { + "epoch": 0.3966251379908532, + "grad_norm": 1.5635979175567627, + "learning_rate": 1.3742087472632405e-05, + "loss": 0.4662, + "step": 12575 + }, + { + "epoch": 0.3966566787572938, + "grad_norm": 1.708595871925354, + "learning_rate": 1.3741140118326404e-05, + "loss": 0.5127, + "step": 12576 + }, + { + "epoch": 0.39668821952373445, + "grad_norm": 1.6211967468261719, + "learning_rate": 1.3740192724978699e-05, + "loss": 0.4974, + "step": 12577 + }, + { + "epoch": 0.396719760290175, + "grad_norm": 1.5589807033538818, + "learning_rate": 1.3739245292599177e-05, + "loss": 0.5192, + "step": 12578 + }, + { + "epoch": 0.39675130105661566, + "grad_norm": 1.6363859176635742, + "learning_rate": 1.3738297821197728e-05, + "loss": 0.475, + "step": 12579 + }, + { + "epoch": 0.3967828418230563, + "grad_norm": 1.3473094701766968, + "learning_rate": 1.373735031078424e-05, + "loss": 0.5309, + "step": 12580 + }, + { + "epoch": 0.3968143825894969, + "grad_norm": 1.9543519020080566, + "learning_rate": 1.3736402761368597e-05, + "loss": 0.4685, + "step": 12581 + }, + { + "epoch": 0.39684592335593755, + "grad_norm": 2.2528672218322754, + "learning_rate": 1.3735455172960695e-05, + "loss": 0.5244, + "step": 12582 + }, + { + "epoch": 0.3968774641223782, + "grad_norm": 1.6929820775985718, + "learning_rate": 1.3734507545570412e-05, + "loss": 0.4935, + "step": 12583 + }, + { + "epoch": 0.3969090048888188, + "grad_norm": 1.5390633344650269, + "learning_rate": 1.3733559879207648e-05, + "loss": 0.5276, + "step": 12584 + }, + { + "epoch": 0.39694054565525944, + "grad_norm": 1.5267413854599, + "learning_rate": 1.3732612173882285e-05, + "loss": 0.4884, + "step": 12585 + }, + { + "epoch": 0.39697208642170007, + "grad_norm": 2.3130922317504883, + "learning_rate": 1.3731664429604214e-05, + "loss": 0.3702, + "step": 12586 + }, + { + "epoch": 0.3970036271881407, + "grad_norm": 1.8025792837142944, + "learning_rate": 1.373071664638333e-05, + "loss": 0.5027, + "step": 12587 + }, + { + "epoch": 0.39703516795458127, + "grad_norm": 2.230625867843628, + "learning_rate": 1.3729768824229519e-05, + "loss": 0.3897, + "step": 12588 + }, + { + "epoch": 0.3970667087210219, + "grad_norm": 1.6752846240997314, + "learning_rate": 1.3728820963152677e-05, + "loss": 0.4461, + "step": 12589 + }, + { + "epoch": 0.39709824948746253, + "grad_norm": 2.049468517303467, + "learning_rate": 1.3727873063162692e-05, + "loss": 0.508, + "step": 12590 + }, + { + "epoch": 0.39712979025390316, + "grad_norm": 1.7486156225204468, + "learning_rate": 1.3726925124269453e-05, + "loss": 0.5237, + "step": 12591 + }, + { + "epoch": 0.3971613310203438, + "grad_norm": 1.7756047248840332, + "learning_rate": 1.372597714648286e-05, + "loss": 0.4678, + "step": 12592 + }, + { + "epoch": 0.3971928717867844, + "grad_norm": 1.4935866594314575, + "learning_rate": 1.3725029129812801e-05, + "loss": 0.5146, + "step": 12593 + }, + { + "epoch": 0.39722441255322505, + "grad_norm": 1.424317479133606, + "learning_rate": 1.3724081074269173e-05, + "loss": 0.363, + "step": 12594 + }, + { + "epoch": 0.3972559533196657, + "grad_norm": 1.6761113405227661, + "learning_rate": 1.3723132979861866e-05, + "loss": 0.8246, + "step": 12595 + }, + { + "epoch": 0.3972874940861063, + "grad_norm": 1.6682655811309814, + "learning_rate": 1.3722184846600772e-05, + "loss": 0.8245, + "step": 12596 + }, + { + "epoch": 0.39731903485254694, + "grad_norm": 2.204333543777466, + "learning_rate": 1.3721236674495795e-05, + "loss": 0.5671, + "step": 12597 + }, + { + "epoch": 0.39735057561898757, + "grad_norm": 2.2079460620880127, + "learning_rate": 1.3720288463556819e-05, + "loss": 0.5218, + "step": 12598 + }, + { + "epoch": 0.39738211638542814, + "grad_norm": 1.6942318677902222, + "learning_rate": 1.3719340213793745e-05, + "loss": 0.5066, + "step": 12599 + }, + { + "epoch": 0.39741365715186877, + "grad_norm": 1.6905447244644165, + "learning_rate": 1.3718391925216473e-05, + "loss": 0.5049, + "step": 12600 + }, + { + "epoch": 0.3974451979183094, + "grad_norm": 2.8446357250213623, + "learning_rate": 1.3717443597834888e-05, + "loss": 0.4968, + "step": 12601 + }, + { + "epoch": 0.39747673868475003, + "grad_norm": 1.450355052947998, + "learning_rate": 1.3716495231658894e-05, + "loss": 0.5086, + "step": 12602 + }, + { + "epoch": 0.39750827945119066, + "grad_norm": 1.8769906759262085, + "learning_rate": 1.3715546826698387e-05, + "loss": 0.8397, + "step": 12603 + }, + { + "epoch": 0.3975398202176313, + "grad_norm": 1.738867163658142, + "learning_rate": 1.3714598382963264e-05, + "loss": 0.5027, + "step": 12604 + }, + { + "epoch": 0.3975713609840719, + "grad_norm": 1.4999786615371704, + "learning_rate": 1.3713649900463422e-05, + "loss": 0.5449, + "step": 12605 + }, + { + "epoch": 0.39760290175051255, + "grad_norm": 1.8500417470932007, + "learning_rate": 1.3712701379208759e-05, + "loss": 0.8627, + "step": 12606 + }, + { + "epoch": 0.3976344425169532, + "grad_norm": 1.5979329347610474, + "learning_rate": 1.3711752819209179e-05, + "loss": 0.4976, + "step": 12607 + }, + { + "epoch": 0.3976659832833938, + "grad_norm": 1.6613173484802246, + "learning_rate": 1.371080422047457e-05, + "loss": 0.5068, + "step": 12608 + }, + { + "epoch": 0.3976975240498344, + "grad_norm": 1.6394795179367065, + "learning_rate": 1.370985558301484e-05, + "loss": 0.5293, + "step": 12609 + }, + { + "epoch": 0.397729064816275, + "grad_norm": 1.3428186178207397, + "learning_rate": 1.3708906906839885e-05, + "loss": 0.5187, + "step": 12610 + }, + { + "epoch": 0.39776060558271564, + "grad_norm": 1.5966800451278687, + "learning_rate": 1.3707958191959609e-05, + "loss": 0.5016, + "step": 12611 + }, + { + "epoch": 0.3977921463491563, + "grad_norm": 1.861409306526184, + "learning_rate": 1.3707009438383912e-05, + "loss": 0.4848, + "step": 12612 + }, + { + "epoch": 0.3978236871155969, + "grad_norm": 1.5454869270324707, + "learning_rate": 1.3706060646122688e-05, + "loss": 0.4761, + "step": 12613 + }, + { + "epoch": 0.39785522788203753, + "grad_norm": 1.8522796630859375, + "learning_rate": 1.3705111815185847e-05, + "loss": 0.5186, + "step": 12614 + }, + { + "epoch": 0.39788676864847816, + "grad_norm": 1.6521199941635132, + "learning_rate": 1.3704162945583285e-05, + "loss": 0.4694, + "step": 12615 + }, + { + "epoch": 0.3979183094149188, + "grad_norm": 1.8582637310028076, + "learning_rate": 1.3703214037324908e-05, + "loss": 0.5166, + "step": 12616 + }, + { + "epoch": 0.3979498501813594, + "grad_norm": 3.1084089279174805, + "learning_rate": 1.3702265090420615e-05, + "loss": 0.5183, + "step": 12617 + }, + { + "epoch": 0.39798139094780005, + "grad_norm": 2.4876275062561035, + "learning_rate": 1.3701316104880313e-05, + "loss": 0.5083, + "step": 12618 + }, + { + "epoch": 0.3980129317142407, + "grad_norm": 1.8546456098556519, + "learning_rate": 1.3700367080713903e-05, + "loss": 0.5334, + "step": 12619 + }, + { + "epoch": 0.39804447248068126, + "grad_norm": 1.7450613975524902, + "learning_rate": 1.3699418017931288e-05, + "loss": 0.5059, + "step": 12620 + }, + { + "epoch": 0.3980760132471219, + "grad_norm": 1.8619283437728882, + "learning_rate": 1.3698468916542375e-05, + "loss": 0.5058, + "step": 12621 + }, + { + "epoch": 0.3981075540135625, + "grad_norm": 1.8291970491409302, + "learning_rate": 1.369751977655707e-05, + "loss": 0.5072, + "step": 12622 + }, + { + "epoch": 0.39813909478000314, + "grad_norm": 1.7883726358413696, + "learning_rate": 1.369657059798527e-05, + "loss": 0.4845, + "step": 12623 + }, + { + "epoch": 0.3981706355464438, + "grad_norm": 1.6218442916870117, + "learning_rate": 1.3695621380836888e-05, + "loss": 0.481, + "step": 12624 + }, + { + "epoch": 0.3982021763128844, + "grad_norm": 1.6751877069473267, + "learning_rate": 1.3694672125121826e-05, + "loss": 0.5183, + "step": 12625 + }, + { + "epoch": 0.39823371707932503, + "grad_norm": 1.5550130605697632, + "learning_rate": 1.369372283084999e-05, + "loss": 0.5306, + "step": 12626 + }, + { + "epoch": 0.39826525784576566, + "grad_norm": 1.4960529804229736, + "learning_rate": 1.3692773498031292e-05, + "loss": 0.5131, + "step": 12627 + }, + { + "epoch": 0.3982967986122063, + "grad_norm": 2.2455530166625977, + "learning_rate": 1.3691824126675633e-05, + "loss": 0.4844, + "step": 12628 + }, + { + "epoch": 0.3983283393786469, + "grad_norm": 1.6384046077728271, + "learning_rate": 1.3690874716792923e-05, + "loss": 0.5038, + "step": 12629 + }, + { + "epoch": 0.39835988014508755, + "grad_norm": 1.695850133895874, + "learning_rate": 1.368992526839307e-05, + "loss": 0.4894, + "step": 12630 + }, + { + "epoch": 0.3983914209115281, + "grad_norm": 2.5429348945617676, + "learning_rate": 1.368897578148598e-05, + "loss": 0.8833, + "step": 12631 + }, + { + "epoch": 0.39842296167796876, + "grad_norm": 1.9899214506149292, + "learning_rate": 1.3688026256081566e-05, + "loss": 0.5204, + "step": 12632 + }, + { + "epoch": 0.3984545024444094, + "grad_norm": 1.8335161209106445, + "learning_rate": 1.3687076692189733e-05, + "loss": 0.5308, + "step": 12633 + }, + { + "epoch": 0.39848604321085, + "grad_norm": 1.490891456604004, + "learning_rate": 1.3686127089820392e-05, + "loss": 0.5098, + "step": 12634 + }, + { + "epoch": 0.39851758397729065, + "grad_norm": 1.7036409378051758, + "learning_rate": 1.368517744898345e-05, + "loss": 0.5035, + "step": 12635 + }, + { + "epoch": 0.3985491247437313, + "grad_norm": 2.22522234916687, + "learning_rate": 1.3684227769688825e-05, + "loss": 0.5112, + "step": 12636 + }, + { + "epoch": 0.3985806655101719, + "grad_norm": 1.376826286315918, + "learning_rate": 1.3683278051946418e-05, + "loss": 0.5234, + "step": 12637 + }, + { + "epoch": 0.39861220627661254, + "grad_norm": 1.5189789533615112, + "learning_rate": 1.3682328295766147e-05, + "loss": 0.4613, + "step": 12638 + }, + { + "epoch": 0.39864374704305316, + "grad_norm": 1.4287604093551636, + "learning_rate": 1.368137850115792e-05, + "loss": 0.5128, + "step": 12639 + }, + { + "epoch": 0.3986752878094938, + "grad_norm": 1.4983510971069336, + "learning_rate": 1.3680428668131649e-05, + "loss": 0.4868, + "step": 12640 + }, + { + "epoch": 0.39870682857593437, + "grad_norm": 9.921305656433105, + "learning_rate": 1.3679478796697247e-05, + "loss": 0.5225, + "step": 12641 + }, + { + "epoch": 0.398738369342375, + "grad_norm": 2.278465986251831, + "learning_rate": 1.367852888686463e-05, + "loss": 0.5456, + "step": 12642 + }, + { + "epoch": 0.39876991010881563, + "grad_norm": 2.6174509525299072, + "learning_rate": 1.3677578938643705e-05, + "loss": 0.4895, + "step": 12643 + }, + { + "epoch": 0.39880145087525626, + "grad_norm": 1.890452265739441, + "learning_rate": 1.367662895204439e-05, + "loss": 0.5087, + "step": 12644 + }, + { + "epoch": 0.3988329916416969, + "grad_norm": 2.3310036659240723, + "learning_rate": 1.3675678927076595e-05, + "loss": 0.5348, + "step": 12645 + }, + { + "epoch": 0.3988645324081375, + "grad_norm": 1.686635136604309, + "learning_rate": 1.3674728863750235e-05, + "loss": 0.5183, + "step": 12646 + }, + { + "epoch": 0.39889607317457815, + "grad_norm": 1.8231209516525269, + "learning_rate": 1.367377876207523e-05, + "loss": 0.5418, + "step": 12647 + }, + { + "epoch": 0.3989276139410188, + "grad_norm": 2.2630183696746826, + "learning_rate": 1.367282862206149e-05, + "loss": 0.5032, + "step": 12648 + }, + { + "epoch": 0.3989591547074594, + "grad_norm": 1.2715104818344116, + "learning_rate": 1.367187844371893e-05, + "loss": 0.4968, + "step": 12649 + }, + { + "epoch": 0.39899069547390004, + "grad_norm": 1.5083208084106445, + "learning_rate": 1.3670928227057464e-05, + "loss": 0.5016, + "step": 12650 + }, + { + "epoch": 0.39902223624034067, + "grad_norm": 1.5162205696105957, + "learning_rate": 1.3669977972087018e-05, + "loss": 0.5141, + "step": 12651 + }, + { + "epoch": 0.39905377700678124, + "grad_norm": 1.5111515522003174, + "learning_rate": 1.3669027678817501e-05, + "loss": 0.4785, + "step": 12652 + }, + { + "epoch": 0.39908531777322187, + "grad_norm": 1.5067239999771118, + "learning_rate": 1.3668077347258828e-05, + "loss": 0.5054, + "step": 12653 + }, + { + "epoch": 0.3991168585396625, + "grad_norm": 1.8079880475997925, + "learning_rate": 1.3667126977420921e-05, + "loss": 0.5197, + "step": 12654 + }, + { + "epoch": 0.39914839930610313, + "grad_norm": 3.5630648136138916, + "learning_rate": 1.3666176569313698e-05, + "loss": 0.8645, + "step": 12655 + }, + { + "epoch": 0.39917994007254376, + "grad_norm": 1.6690783500671387, + "learning_rate": 1.3665226122947075e-05, + "loss": 0.5056, + "step": 12656 + }, + { + "epoch": 0.3992114808389844, + "grad_norm": 1.5352753400802612, + "learning_rate": 1.366427563833097e-05, + "loss": 0.5168, + "step": 12657 + }, + { + "epoch": 0.399243021605425, + "grad_norm": 1.6073992252349854, + "learning_rate": 1.3663325115475304e-05, + "loss": 0.5305, + "step": 12658 + }, + { + "epoch": 0.39927456237186565, + "grad_norm": 2.1406333446502686, + "learning_rate": 1.3662374554389998e-05, + "loss": 0.5386, + "step": 12659 + }, + { + "epoch": 0.3993061031383063, + "grad_norm": 1.4992892742156982, + "learning_rate": 1.3661423955084967e-05, + "loss": 0.515, + "step": 12660 + }, + { + "epoch": 0.3993376439047469, + "grad_norm": 1.2818819284439087, + "learning_rate": 1.3660473317570138e-05, + "loss": 0.5044, + "step": 12661 + }, + { + "epoch": 0.3993691846711875, + "grad_norm": 1.5857281684875488, + "learning_rate": 1.3659522641855425e-05, + "loss": 0.5388, + "step": 12662 + }, + { + "epoch": 0.3994007254376281, + "grad_norm": 1.692716121673584, + "learning_rate": 1.3658571927950752e-05, + "loss": 0.5247, + "step": 12663 + }, + { + "epoch": 0.39943226620406874, + "grad_norm": 3.217113971710205, + "learning_rate": 1.3657621175866038e-05, + "loss": 0.5056, + "step": 12664 + }, + { + "epoch": 0.39946380697050937, + "grad_norm": 1.7250957489013672, + "learning_rate": 1.3656670385611208e-05, + "loss": 0.5106, + "step": 12665 + }, + { + "epoch": 0.39949534773695, + "grad_norm": 3.9081742763519287, + "learning_rate": 1.3655719557196185e-05, + "loss": 0.5097, + "step": 12666 + }, + { + "epoch": 0.39952688850339063, + "grad_norm": 1.5692076683044434, + "learning_rate": 1.3654768690630888e-05, + "loss": 0.5038, + "step": 12667 + }, + { + "epoch": 0.39955842926983126, + "grad_norm": 1.3615785837173462, + "learning_rate": 1.365381778592524e-05, + "loss": 0.5225, + "step": 12668 + }, + { + "epoch": 0.3995899700362719, + "grad_norm": 3.382143974304199, + "learning_rate": 1.3652866843089172e-05, + "loss": 0.5171, + "step": 12669 + }, + { + "epoch": 0.3996215108027125, + "grad_norm": 1.5970178842544556, + "learning_rate": 1.3651915862132599e-05, + "loss": 0.5034, + "step": 12670 + }, + { + "epoch": 0.39965305156915315, + "grad_norm": 1.3848587274551392, + "learning_rate": 1.3650964843065448e-05, + "loss": 0.518, + "step": 12671 + }, + { + "epoch": 0.3996845923355938, + "grad_norm": 1.4194614887237549, + "learning_rate": 1.3650013785897647e-05, + "loss": 0.5048, + "step": 12672 + }, + { + "epoch": 0.39971613310203435, + "grad_norm": 1.4743751287460327, + "learning_rate": 1.3649062690639114e-05, + "loss": 0.5115, + "step": 12673 + }, + { + "epoch": 0.399747673868475, + "grad_norm": 2.302347421646118, + "learning_rate": 1.3648111557299779e-05, + "loss": 0.465, + "step": 12674 + }, + { + "epoch": 0.3997792146349156, + "grad_norm": 1.5376149415969849, + "learning_rate": 1.3647160385889567e-05, + "loss": 0.5052, + "step": 12675 + }, + { + "epoch": 0.39981075540135624, + "grad_norm": 1.6085423231124878, + "learning_rate": 1.3646209176418407e-05, + "loss": 0.4504, + "step": 12676 + }, + { + "epoch": 0.3998422961677969, + "grad_norm": 1.3321516513824463, + "learning_rate": 1.364525792889622e-05, + "loss": 0.5137, + "step": 12677 + }, + { + "epoch": 0.3998738369342375, + "grad_norm": 2.7042236328125, + "learning_rate": 1.3644306643332939e-05, + "loss": 0.8635, + "step": 12678 + }, + { + "epoch": 0.39990537770067813, + "grad_norm": 1.9814475774765015, + "learning_rate": 1.3643355319738487e-05, + "loss": 0.5073, + "step": 12679 + }, + { + "epoch": 0.39993691846711876, + "grad_norm": 1.5623061656951904, + "learning_rate": 1.3642403958122792e-05, + "loss": 0.4746, + "step": 12680 + }, + { + "epoch": 0.3999684592335594, + "grad_norm": 1.7291572093963623, + "learning_rate": 1.3641452558495787e-05, + "loss": 0.5329, + "step": 12681 + }, + { + "epoch": 0.4, + "grad_norm": 1.706836462020874, + "learning_rate": 1.3640501120867394e-05, + "loss": 0.5286, + "step": 12682 + }, + { + "epoch": 0.40003154076644065, + "grad_norm": 1.4122016429901123, + "learning_rate": 1.3639549645247545e-05, + "loss": 0.8184, + "step": 12683 + }, + { + "epoch": 0.4000630815328812, + "grad_norm": 1.3099365234375, + "learning_rate": 1.363859813164617e-05, + "loss": 0.4814, + "step": 12684 + }, + { + "epoch": 0.40009462229932186, + "grad_norm": 1.512078881263733, + "learning_rate": 1.3637646580073196e-05, + "loss": 0.5157, + "step": 12685 + }, + { + "epoch": 0.4001261630657625, + "grad_norm": 2.1114563941955566, + "learning_rate": 1.363669499053856e-05, + "loss": 0.848, + "step": 12686 + }, + { + "epoch": 0.4001577038322031, + "grad_norm": 1.4953478574752808, + "learning_rate": 1.3635743363052185e-05, + "loss": 0.5036, + "step": 12687 + }, + { + "epoch": 0.40018924459864375, + "grad_norm": 2.5848770141601562, + "learning_rate": 1.3634791697624005e-05, + "loss": 0.5227, + "step": 12688 + }, + { + "epoch": 0.4002207853650844, + "grad_norm": 1.5396761894226074, + "learning_rate": 1.3633839994263951e-05, + "loss": 0.8234, + "step": 12689 + }, + { + "epoch": 0.400252326131525, + "grad_norm": 1.7246211767196655, + "learning_rate": 1.3632888252981957e-05, + "loss": 0.5098, + "step": 12690 + }, + { + "epoch": 0.40028386689796563, + "grad_norm": 1.5483888387680054, + "learning_rate": 1.3631936473787954e-05, + "loss": 0.5223, + "step": 12691 + }, + { + "epoch": 0.40031540766440626, + "grad_norm": 3.8666927814483643, + "learning_rate": 1.363098465669187e-05, + "loss": 0.4506, + "step": 12692 + }, + { + "epoch": 0.4003469484308469, + "grad_norm": 1.7086454629898071, + "learning_rate": 1.3630032801703646e-05, + "loss": 0.398, + "step": 12693 + }, + { + "epoch": 0.40037848919728747, + "grad_norm": 1.6640063524246216, + "learning_rate": 1.3629080908833208e-05, + "loss": 0.4757, + "step": 12694 + }, + { + "epoch": 0.4004100299637281, + "grad_norm": 1.6507161855697632, + "learning_rate": 1.3628128978090492e-05, + "loss": 0.5166, + "step": 12695 + }, + { + "epoch": 0.4004415707301687, + "grad_norm": 1.6083192825317383, + "learning_rate": 1.3627177009485435e-05, + "loss": 0.5393, + "step": 12696 + }, + { + "epoch": 0.40047311149660936, + "grad_norm": 1.5259371995925903, + "learning_rate": 1.3626225003027968e-05, + "loss": 0.5029, + "step": 12697 + }, + { + "epoch": 0.40050465226305, + "grad_norm": 1.5863382816314697, + "learning_rate": 1.362527295872803e-05, + "loss": 0.5131, + "step": 12698 + }, + { + "epoch": 0.4005361930294906, + "grad_norm": 1.6912283897399902, + "learning_rate": 1.362432087659555e-05, + "loss": 0.5138, + "step": 12699 + }, + { + "epoch": 0.40056773379593125, + "grad_norm": 1.2297049760818481, + "learning_rate": 1.362336875664047e-05, + "loss": 0.5108, + "step": 12700 + }, + { + "epoch": 0.4005992745623719, + "grad_norm": 1.4234988689422607, + "learning_rate": 1.3622416598872725e-05, + "loss": 0.5205, + "step": 12701 + }, + { + "epoch": 0.4006308153288125, + "grad_norm": 1.6012881994247437, + "learning_rate": 1.3621464403302247e-05, + "loss": 0.5262, + "step": 12702 + }, + { + "epoch": 0.40066235609525314, + "grad_norm": 1.7155559062957764, + "learning_rate": 1.3620512169938977e-05, + "loss": 0.5432, + "step": 12703 + }, + { + "epoch": 0.40069389686169377, + "grad_norm": 1.6311849355697632, + "learning_rate": 1.361955989879285e-05, + "loss": 0.4703, + "step": 12704 + }, + { + "epoch": 0.40072543762813434, + "grad_norm": 1.642824411392212, + "learning_rate": 1.3618607589873807e-05, + "loss": 0.5016, + "step": 12705 + }, + { + "epoch": 0.40075697839457497, + "grad_norm": 1.6832540035247803, + "learning_rate": 1.361765524319178e-05, + "loss": 0.4903, + "step": 12706 + }, + { + "epoch": 0.4007885191610156, + "grad_norm": 2.446223735809326, + "learning_rate": 1.3616702858756718e-05, + "loss": 0.824, + "step": 12707 + }, + { + "epoch": 0.40082005992745623, + "grad_norm": 1.5339303016662598, + "learning_rate": 1.361575043657855e-05, + "loss": 0.5356, + "step": 12708 + }, + { + "epoch": 0.40085160069389686, + "grad_norm": 1.431501865386963, + "learning_rate": 1.3614797976667217e-05, + "loss": 0.8449, + "step": 12709 + }, + { + "epoch": 0.4008831414603375, + "grad_norm": 1.4773844480514526, + "learning_rate": 1.3613845479032663e-05, + "loss": 0.5293, + "step": 12710 + }, + { + "epoch": 0.4009146822267781, + "grad_norm": 1.544260859489441, + "learning_rate": 1.3612892943684825e-05, + "loss": 0.4561, + "step": 12711 + }, + { + "epoch": 0.40094622299321875, + "grad_norm": 1.4744781255722046, + "learning_rate": 1.3611940370633642e-05, + "loss": 0.4884, + "step": 12712 + }, + { + "epoch": 0.4009777637596594, + "grad_norm": 1.9384688138961792, + "learning_rate": 1.3610987759889058e-05, + "loss": 0.5358, + "step": 12713 + }, + { + "epoch": 0.4010093045261, + "grad_norm": 1.524806261062622, + "learning_rate": 1.3610035111461016e-05, + "loss": 0.526, + "step": 12714 + }, + { + "epoch": 0.4010408452925406, + "grad_norm": 1.90337073802948, + "learning_rate": 1.360908242535945e-05, + "loss": 0.481, + "step": 12715 + }, + { + "epoch": 0.4010723860589812, + "grad_norm": 1.9155941009521484, + "learning_rate": 1.3608129701594306e-05, + "loss": 0.5187, + "step": 12716 + }, + { + "epoch": 0.40110392682542184, + "grad_norm": 1.4846841096878052, + "learning_rate": 1.3607176940175531e-05, + "loss": 0.4832, + "step": 12717 + }, + { + "epoch": 0.40113546759186247, + "grad_norm": 1.3138244152069092, + "learning_rate": 1.360622414111306e-05, + "loss": 0.5363, + "step": 12718 + }, + { + "epoch": 0.4011670083583031, + "grad_norm": 1.5804890394210815, + "learning_rate": 1.360527130441684e-05, + "loss": 0.5096, + "step": 12719 + }, + { + "epoch": 0.40119854912474373, + "grad_norm": 1.5039372444152832, + "learning_rate": 1.3604318430096817e-05, + "loss": 0.5246, + "step": 12720 + }, + { + "epoch": 0.40123008989118436, + "grad_norm": 1.4221296310424805, + "learning_rate": 1.3603365518162932e-05, + "loss": 0.4935, + "step": 12721 + }, + { + "epoch": 0.401261630657625, + "grad_norm": 1.4178799390792847, + "learning_rate": 1.360241256862513e-05, + "loss": 0.4819, + "step": 12722 + }, + { + "epoch": 0.4012931714240656, + "grad_norm": 1.5620505809783936, + "learning_rate": 1.3601459581493353e-05, + "loss": 0.5041, + "step": 12723 + }, + { + "epoch": 0.40132471219050625, + "grad_norm": 1.7384551763534546, + "learning_rate": 1.3600506556777551e-05, + "loss": 0.4758, + "step": 12724 + }, + { + "epoch": 0.4013562529569469, + "grad_norm": 1.6007734537124634, + "learning_rate": 1.3599553494487668e-05, + "loss": 0.4725, + "step": 12725 + }, + { + "epoch": 0.40138779372338745, + "grad_norm": 1.3246865272521973, + "learning_rate": 1.3598600394633645e-05, + "loss": 0.4704, + "step": 12726 + }, + { + "epoch": 0.4014193344898281, + "grad_norm": 1.7313611507415771, + "learning_rate": 1.3597647257225438e-05, + "loss": 0.4844, + "step": 12727 + }, + { + "epoch": 0.4014508752562687, + "grad_norm": 7.619290351867676, + "learning_rate": 1.3596694082272986e-05, + "loss": 0.8347, + "step": 12728 + }, + { + "epoch": 0.40148241602270934, + "grad_norm": 1.4796422719955444, + "learning_rate": 1.3595740869786238e-05, + "loss": 0.5238, + "step": 12729 + }, + { + "epoch": 0.40151395678915, + "grad_norm": 1.5187194347381592, + "learning_rate": 1.3594787619775143e-05, + "loss": 0.5534, + "step": 12730 + }, + { + "epoch": 0.4015454975555906, + "grad_norm": 1.6676088571548462, + "learning_rate": 1.359383433224965e-05, + "loss": 0.5469, + "step": 12731 + }, + { + "epoch": 0.40157703832203123, + "grad_norm": 1.7805968523025513, + "learning_rate": 1.3592881007219702e-05, + "loss": 0.4508, + "step": 12732 + }, + { + "epoch": 0.40160857908847186, + "grad_norm": 1.7296620607376099, + "learning_rate": 1.3591927644695252e-05, + "loss": 0.5133, + "step": 12733 + }, + { + "epoch": 0.4016401198549125, + "grad_norm": 1.5649420022964478, + "learning_rate": 1.3590974244686248e-05, + "loss": 0.5052, + "step": 12734 + }, + { + "epoch": 0.4016716606213531, + "grad_norm": 1.5966044664382935, + "learning_rate": 1.3590020807202637e-05, + "loss": 0.5077, + "step": 12735 + }, + { + "epoch": 0.4017032013877937, + "grad_norm": 1.5155677795410156, + "learning_rate": 1.3589067332254374e-05, + "loss": 0.5235, + "step": 12736 + }, + { + "epoch": 0.4017347421542343, + "grad_norm": 1.7490025758743286, + "learning_rate": 1.3588113819851404e-05, + "loss": 0.4938, + "step": 12737 + }, + { + "epoch": 0.40176628292067496, + "grad_norm": 1.5658884048461914, + "learning_rate": 1.3587160270003685e-05, + "loss": 0.4986, + "step": 12738 + }, + { + "epoch": 0.4017978236871156, + "grad_norm": 1.7760244607925415, + "learning_rate": 1.358620668272116e-05, + "loss": 0.5177, + "step": 12739 + }, + { + "epoch": 0.4018293644535562, + "grad_norm": 1.5980008840560913, + "learning_rate": 1.3585253058013785e-05, + "loss": 0.4886, + "step": 12740 + }, + { + "epoch": 0.40186090521999684, + "grad_norm": 1.5539426803588867, + "learning_rate": 1.358429939589151e-05, + "loss": 0.4986, + "step": 12741 + }, + { + "epoch": 0.4018924459864375, + "grad_norm": 1.5026706457138062, + "learning_rate": 1.3583345696364287e-05, + "loss": 0.5048, + "step": 12742 + }, + { + "epoch": 0.4019239867528781, + "grad_norm": 1.2803462743759155, + "learning_rate": 1.3582391959442069e-05, + "loss": 0.4947, + "step": 12743 + }, + { + "epoch": 0.40195552751931873, + "grad_norm": 1.7540169954299927, + "learning_rate": 1.358143818513481e-05, + "loss": 0.5475, + "step": 12744 + }, + { + "epoch": 0.40198706828575936, + "grad_norm": 1.9290536642074585, + "learning_rate": 1.3580484373452462e-05, + "loss": 0.5171, + "step": 12745 + }, + { + "epoch": 0.4020186090522, + "grad_norm": 1.4262431859970093, + "learning_rate": 1.357953052440498e-05, + "loss": 0.5075, + "step": 12746 + }, + { + "epoch": 0.40205014981864057, + "grad_norm": 1.7203863859176636, + "learning_rate": 1.3578576638002314e-05, + "loss": 0.4871, + "step": 12747 + }, + { + "epoch": 0.4020816905850812, + "grad_norm": 1.7050483226776123, + "learning_rate": 1.3577622714254428e-05, + "loss": 0.4995, + "step": 12748 + }, + { + "epoch": 0.4021132313515218, + "grad_norm": 1.8729687929153442, + "learning_rate": 1.3576668753171267e-05, + "loss": 0.4955, + "step": 12749 + }, + { + "epoch": 0.40214477211796246, + "grad_norm": 1.7574199438095093, + "learning_rate": 1.3575714754762795e-05, + "loss": 0.5004, + "step": 12750 + }, + { + "epoch": 0.4021763128844031, + "grad_norm": 1.528692603111267, + "learning_rate": 1.3574760719038959e-05, + "loss": 0.484, + "step": 12751 + }, + { + "epoch": 0.4022078536508437, + "grad_norm": 1.4155970811843872, + "learning_rate": 1.357380664600972e-05, + "loss": 0.4988, + "step": 12752 + }, + { + "epoch": 0.40223939441728435, + "grad_norm": 1.5835165977478027, + "learning_rate": 1.3572852535685032e-05, + "loss": 0.4946, + "step": 12753 + }, + { + "epoch": 0.402270935183725, + "grad_norm": 3.4797046184539795, + "learning_rate": 1.3571898388074856e-05, + "loss": 0.5001, + "step": 12754 + }, + { + "epoch": 0.4023024759501656, + "grad_norm": 1.2830806970596313, + "learning_rate": 1.3570944203189146e-05, + "loss": 0.4963, + "step": 12755 + }, + { + "epoch": 0.40233401671660624, + "grad_norm": 1.5227489471435547, + "learning_rate": 1.356998998103786e-05, + "loss": 0.5003, + "step": 12756 + }, + { + "epoch": 0.40236555748304687, + "grad_norm": 2.128345012664795, + "learning_rate": 1.3569035721630957e-05, + "loss": 0.8616, + "step": 12757 + }, + { + "epoch": 0.40239709824948744, + "grad_norm": 1.8614040613174438, + "learning_rate": 1.3568081424978394e-05, + "loss": 0.4967, + "step": 12758 + }, + { + "epoch": 0.40242863901592807, + "grad_norm": 1.735836386680603, + "learning_rate": 1.3567127091090133e-05, + "loss": 0.5117, + "step": 12759 + }, + { + "epoch": 0.4024601797823687, + "grad_norm": 2.0707592964172363, + "learning_rate": 1.3566172719976129e-05, + "loss": 0.5361, + "step": 12760 + }, + { + "epoch": 0.40249172054880933, + "grad_norm": 1.8435848951339722, + "learning_rate": 1.3565218311646345e-05, + "loss": 0.5069, + "step": 12761 + }, + { + "epoch": 0.40252326131524996, + "grad_norm": 1.2476329803466797, + "learning_rate": 1.356426386611074e-05, + "loss": 0.8278, + "step": 12762 + }, + { + "epoch": 0.4025548020816906, + "grad_norm": 1.757185697555542, + "learning_rate": 1.3563309383379273e-05, + "loss": 0.4721, + "step": 12763 + }, + { + "epoch": 0.4025863428481312, + "grad_norm": 1.5174959897994995, + "learning_rate": 1.3562354863461903e-05, + "loss": 0.8482, + "step": 12764 + }, + { + "epoch": 0.40261788361457185, + "grad_norm": 1.8077462911605835, + "learning_rate": 1.35614003063686e-05, + "loss": 0.5058, + "step": 12765 + }, + { + "epoch": 0.4026494243810125, + "grad_norm": 1.509103536605835, + "learning_rate": 1.3560445712109314e-05, + "loss": 0.5022, + "step": 12766 + }, + { + "epoch": 0.4026809651474531, + "grad_norm": 1.7444729804992676, + "learning_rate": 1.3559491080694013e-05, + "loss": 0.4922, + "step": 12767 + }, + { + "epoch": 0.4027125059138937, + "grad_norm": 1.5711324214935303, + "learning_rate": 1.3558536412132663e-05, + "loss": 0.5351, + "step": 12768 + }, + { + "epoch": 0.4027440466803343, + "grad_norm": 2.020146369934082, + "learning_rate": 1.3557581706435219e-05, + "loss": 0.8248, + "step": 12769 + }, + { + "epoch": 0.40277558744677494, + "grad_norm": 2.103179693222046, + "learning_rate": 1.3556626963611648e-05, + "loss": 0.5018, + "step": 12770 + }, + { + "epoch": 0.40280712821321557, + "grad_norm": 1.897020697593689, + "learning_rate": 1.3555672183671913e-05, + "loss": 0.4941, + "step": 12771 + }, + { + "epoch": 0.4028386689796562, + "grad_norm": 1.624422311782837, + "learning_rate": 1.3554717366625977e-05, + "loss": 0.5095, + "step": 12772 + }, + { + "epoch": 0.40287020974609683, + "grad_norm": 1.8873037099838257, + "learning_rate": 1.3553762512483805e-05, + "loss": 0.5077, + "step": 12773 + }, + { + "epoch": 0.40290175051253746, + "grad_norm": 1.9859898090362549, + "learning_rate": 1.3552807621255364e-05, + "loss": 0.4761, + "step": 12774 + }, + { + "epoch": 0.4029332912789781, + "grad_norm": 1.4789578914642334, + "learning_rate": 1.3551852692950615e-05, + "loss": 0.4993, + "step": 12775 + }, + { + "epoch": 0.4029648320454187, + "grad_norm": 1.720956802368164, + "learning_rate": 1.3550897727579526e-05, + "loss": 0.5444, + "step": 12776 + }, + { + "epoch": 0.40299637281185935, + "grad_norm": 1.549331545829773, + "learning_rate": 1.354994272515206e-05, + "loss": 0.5316, + "step": 12777 + }, + { + "epoch": 0.4030279135783, + "grad_norm": 1.3113762140274048, + "learning_rate": 1.354898768567819e-05, + "loss": 0.7837, + "step": 12778 + }, + { + "epoch": 0.40305945434474055, + "grad_norm": 1.6082514524459839, + "learning_rate": 1.3548032609167873e-05, + "loss": 0.5331, + "step": 12779 + }, + { + "epoch": 0.4030909951111812, + "grad_norm": 1.5456348657608032, + "learning_rate": 1.3547077495631085e-05, + "loss": 0.3901, + "step": 12780 + }, + { + "epoch": 0.4031225358776218, + "grad_norm": 1.823625922203064, + "learning_rate": 1.3546122345077786e-05, + "loss": 0.4954, + "step": 12781 + }, + { + "epoch": 0.40315407664406244, + "grad_norm": 1.7948981523513794, + "learning_rate": 1.3545167157517947e-05, + "loss": 0.492, + "step": 12782 + }, + { + "epoch": 0.40318561741050307, + "grad_norm": 1.6788595914840698, + "learning_rate": 1.3544211932961534e-05, + "loss": 0.5192, + "step": 12783 + }, + { + "epoch": 0.4032171581769437, + "grad_norm": 1.8038337230682373, + "learning_rate": 1.354325667141852e-05, + "loss": 0.5082, + "step": 12784 + }, + { + "epoch": 0.40324869894338433, + "grad_norm": 1.705905795097351, + "learning_rate": 1.3542301372898872e-05, + "loss": 0.5243, + "step": 12785 + }, + { + "epoch": 0.40328023970982496, + "grad_norm": 1.3152129650115967, + "learning_rate": 1.3541346037412559e-05, + "loss": 0.5453, + "step": 12786 + }, + { + "epoch": 0.4033117804762656, + "grad_norm": 2.5231807231903076, + "learning_rate": 1.3540390664969547e-05, + "loss": 0.4031, + "step": 12787 + }, + { + "epoch": 0.4033433212427062, + "grad_norm": 1.8648182153701782, + "learning_rate": 1.3539435255579814e-05, + "loss": 0.4794, + "step": 12788 + }, + { + "epoch": 0.4033748620091468, + "grad_norm": 1.3883821964263916, + "learning_rate": 1.3538479809253321e-05, + "loss": 0.491, + "step": 12789 + }, + { + "epoch": 0.4034064027755874, + "grad_norm": 1.4226911067962646, + "learning_rate": 1.3537524326000048e-05, + "loss": 0.821, + "step": 12790 + }, + { + "epoch": 0.40343794354202805, + "grad_norm": 1.6437944173812866, + "learning_rate": 1.3536568805829959e-05, + "loss": 0.5257, + "step": 12791 + }, + { + "epoch": 0.4034694843084687, + "grad_norm": 1.5119256973266602, + "learning_rate": 1.3535613248753028e-05, + "loss": 0.5104, + "step": 12792 + }, + { + "epoch": 0.4035010250749093, + "grad_norm": 1.6501353979110718, + "learning_rate": 1.3534657654779231e-05, + "loss": 0.535, + "step": 12793 + }, + { + "epoch": 0.40353256584134994, + "grad_norm": 1.4515734910964966, + "learning_rate": 1.3533702023918534e-05, + "loss": 0.4927, + "step": 12794 + }, + { + "epoch": 0.4035641066077906, + "grad_norm": 1.4704004526138306, + "learning_rate": 1.3532746356180911e-05, + "loss": 0.5439, + "step": 12795 + }, + { + "epoch": 0.4035956473742312, + "grad_norm": 1.5755447149276733, + "learning_rate": 1.353179065157634e-05, + "loss": 0.3607, + "step": 12796 + }, + { + "epoch": 0.40362718814067183, + "grad_norm": 1.3708720207214355, + "learning_rate": 1.353083491011479e-05, + "loss": 0.4892, + "step": 12797 + }, + { + "epoch": 0.40365872890711246, + "grad_norm": 1.7816462516784668, + "learning_rate": 1.3529879131806236e-05, + "loss": 0.3612, + "step": 12798 + }, + { + "epoch": 0.4036902696735531, + "grad_norm": 1.7772603034973145, + "learning_rate": 1.3528923316660651e-05, + "loss": 0.5, + "step": 12799 + }, + { + "epoch": 0.40372181043999367, + "grad_norm": 1.5938432216644287, + "learning_rate": 1.3527967464688014e-05, + "loss": 0.4678, + "step": 12800 + }, + { + "epoch": 0.4037533512064343, + "grad_norm": 1.9800634384155273, + "learning_rate": 1.3527011575898295e-05, + "loss": 0.3765, + "step": 12801 + }, + { + "epoch": 0.4037848919728749, + "grad_norm": 1.3224085569381714, + "learning_rate": 1.352605565030147e-05, + "loss": 0.8264, + "step": 12802 + }, + { + "epoch": 0.40381643273931556, + "grad_norm": 1.3757110834121704, + "learning_rate": 1.3525099687907521e-05, + "loss": 0.3349, + "step": 12803 + }, + { + "epoch": 0.4038479735057562, + "grad_norm": 1.4558191299438477, + "learning_rate": 1.3524143688726417e-05, + "loss": 0.4993, + "step": 12804 + }, + { + "epoch": 0.4038795142721968, + "grad_norm": 1.7597044706344604, + "learning_rate": 1.3523187652768135e-05, + "loss": 0.5078, + "step": 12805 + }, + { + "epoch": 0.40391105503863745, + "grad_norm": 2.476242780685425, + "learning_rate": 1.3522231580042657e-05, + "loss": 0.4797, + "step": 12806 + }, + { + "epoch": 0.4039425958050781, + "grad_norm": 1.6186676025390625, + "learning_rate": 1.3521275470559955e-05, + "loss": 0.5325, + "step": 12807 + }, + { + "epoch": 0.4039741365715187, + "grad_norm": 4.714893817901611, + "learning_rate": 1.3520319324330014e-05, + "loss": 0.4754, + "step": 12808 + }, + { + "epoch": 0.40400567733795933, + "grad_norm": 1.7645537853240967, + "learning_rate": 1.3519363141362804e-05, + "loss": 0.478, + "step": 12809 + }, + { + "epoch": 0.40403721810439996, + "grad_norm": 1.6554585695266724, + "learning_rate": 1.3518406921668307e-05, + "loss": 0.5175, + "step": 12810 + }, + { + "epoch": 0.40406875887084054, + "grad_norm": 3.5147039890289307, + "learning_rate": 1.3517450665256503e-05, + "loss": 0.4976, + "step": 12811 + }, + { + "epoch": 0.40410029963728117, + "grad_norm": 1.5984646081924438, + "learning_rate": 1.3516494372137368e-05, + "loss": 0.5066, + "step": 12812 + }, + { + "epoch": 0.4041318404037218, + "grad_norm": 1.9909173250198364, + "learning_rate": 1.3515538042320887e-05, + "loss": 0.7983, + "step": 12813 + }, + { + "epoch": 0.40416338117016243, + "grad_norm": 1.6402003765106201, + "learning_rate": 1.3514581675817033e-05, + "loss": 0.468, + "step": 12814 + }, + { + "epoch": 0.40419492193660306, + "grad_norm": 2.227426767349243, + "learning_rate": 1.3513625272635791e-05, + "loss": 0.5307, + "step": 12815 + }, + { + "epoch": 0.4042264627030437, + "grad_norm": 1.7633975744247437, + "learning_rate": 1.3512668832787142e-05, + "loss": 0.5165, + "step": 12816 + }, + { + "epoch": 0.4042580034694843, + "grad_norm": 1.4005217552185059, + "learning_rate": 1.3511712356281068e-05, + "loss": 0.5201, + "step": 12817 + }, + { + "epoch": 0.40428954423592495, + "grad_norm": 1.5796911716461182, + "learning_rate": 1.3510755843127547e-05, + "loss": 0.5147, + "step": 12818 + }, + { + "epoch": 0.4043210850023656, + "grad_norm": 1.3303427696228027, + "learning_rate": 1.3509799293336562e-05, + "loss": 0.5431, + "step": 12819 + }, + { + "epoch": 0.4043526257688062, + "grad_norm": 1.7272474765777588, + "learning_rate": 1.3508842706918099e-05, + "loss": 0.4989, + "step": 12820 + }, + { + "epoch": 0.4043841665352468, + "grad_norm": 5.54616641998291, + "learning_rate": 1.3507886083882133e-05, + "loss": 0.4202, + "step": 12821 + }, + { + "epoch": 0.4044157073016874, + "grad_norm": 1.3492209911346436, + "learning_rate": 1.3506929424238654e-05, + "loss": 0.5173, + "step": 12822 + }, + { + "epoch": 0.40444724806812804, + "grad_norm": 5.676778793334961, + "learning_rate": 1.3505972727997648e-05, + "loss": 0.7995, + "step": 12823 + }, + { + "epoch": 0.40447878883456867, + "grad_norm": 1.7909010648727417, + "learning_rate": 1.3505015995169089e-05, + "loss": 0.5061, + "step": 12824 + }, + { + "epoch": 0.4045103296010093, + "grad_norm": 2.4155964851379395, + "learning_rate": 1.3504059225762967e-05, + "loss": 0.5272, + "step": 12825 + }, + { + "epoch": 0.40454187036744993, + "grad_norm": 1.5102488994598389, + "learning_rate": 1.3503102419789264e-05, + "loss": 0.8276, + "step": 12826 + }, + { + "epoch": 0.40457341113389056, + "grad_norm": 1.6382637023925781, + "learning_rate": 1.3502145577257972e-05, + "loss": 0.4736, + "step": 12827 + }, + { + "epoch": 0.4046049519003312, + "grad_norm": 1.4814743995666504, + "learning_rate": 1.3501188698179071e-05, + "loss": 0.4801, + "step": 12828 + }, + { + "epoch": 0.4046364926667718, + "grad_norm": 1.6660822629928589, + "learning_rate": 1.3500231782562545e-05, + "loss": 0.5334, + "step": 12829 + }, + { + "epoch": 0.40466803343321245, + "grad_norm": 1.7653160095214844, + "learning_rate": 1.3499274830418384e-05, + "loss": 0.514, + "step": 12830 + }, + { + "epoch": 0.4046995741996531, + "grad_norm": 1.7205389738082886, + "learning_rate": 1.3498317841756575e-05, + "loss": 0.506, + "step": 12831 + }, + { + "epoch": 0.40473111496609365, + "grad_norm": 1.9282245635986328, + "learning_rate": 1.3497360816587098e-05, + "loss": 0.4823, + "step": 12832 + }, + { + "epoch": 0.4047626557325343, + "grad_norm": 1.5485026836395264, + "learning_rate": 1.3496403754919947e-05, + "loss": 0.4855, + "step": 12833 + }, + { + "epoch": 0.4047941964989749, + "grad_norm": 1.484824776649475, + "learning_rate": 1.349544665676511e-05, + "loss": 0.5172, + "step": 12834 + }, + { + "epoch": 0.40482573726541554, + "grad_norm": 1.813965916633606, + "learning_rate": 1.349448952213257e-05, + "loss": 0.5101, + "step": 12835 + }, + { + "epoch": 0.40485727803185617, + "grad_norm": 5.131689071655273, + "learning_rate": 1.3493532351032318e-05, + "loss": 0.3622, + "step": 12836 + }, + { + "epoch": 0.4048888187982968, + "grad_norm": 1.427498698234558, + "learning_rate": 1.3492575143474347e-05, + "loss": 0.8398, + "step": 12837 + }, + { + "epoch": 0.40492035956473743, + "grad_norm": 1.7360422611236572, + "learning_rate": 1.3491617899468643e-05, + "loss": 0.5332, + "step": 12838 + }, + { + "epoch": 0.40495190033117806, + "grad_norm": 1.7563388347625732, + "learning_rate": 1.3490660619025191e-05, + "loss": 0.4477, + "step": 12839 + }, + { + "epoch": 0.4049834410976187, + "grad_norm": 1.7805999517440796, + "learning_rate": 1.3489703302153986e-05, + "loss": 0.5125, + "step": 12840 + }, + { + "epoch": 0.4050149818640593, + "grad_norm": 5.7946085929870605, + "learning_rate": 1.348874594886502e-05, + "loss": 0.5194, + "step": 12841 + }, + { + "epoch": 0.4050465226304999, + "grad_norm": 1.6559332609176636, + "learning_rate": 1.3487788559168277e-05, + "loss": 0.5483, + "step": 12842 + }, + { + "epoch": 0.4050780633969405, + "grad_norm": 1.5116013288497925, + "learning_rate": 1.3486831133073754e-05, + "loss": 0.4912, + "step": 12843 + }, + { + "epoch": 0.40510960416338115, + "grad_norm": 2.121800422668457, + "learning_rate": 1.3485873670591443e-05, + "loss": 0.4993, + "step": 12844 + }, + { + "epoch": 0.4051411449298218, + "grad_norm": 1.4514312744140625, + "learning_rate": 1.3484916171731333e-05, + "loss": 0.4254, + "step": 12845 + }, + { + "epoch": 0.4051726856962624, + "grad_norm": 1.7818008661270142, + "learning_rate": 1.3483958636503413e-05, + "loss": 0.5189, + "step": 12846 + }, + { + "epoch": 0.40520422646270304, + "grad_norm": 1.7330803871154785, + "learning_rate": 1.3483001064917683e-05, + "loss": 0.4871, + "step": 12847 + }, + { + "epoch": 0.4052357672291437, + "grad_norm": 1.4632916450500488, + "learning_rate": 1.3482043456984133e-05, + "loss": 0.5089, + "step": 12848 + }, + { + "epoch": 0.4052673079955843, + "grad_norm": 1.419620156288147, + "learning_rate": 1.3481085812712754e-05, + "loss": 0.501, + "step": 12849 + }, + { + "epoch": 0.40529884876202493, + "grad_norm": 2.0391898155212402, + "learning_rate": 1.348012813211354e-05, + "loss": 0.5033, + "step": 12850 + }, + { + "epoch": 0.40533038952846556, + "grad_norm": 2.0298211574554443, + "learning_rate": 1.347917041519649e-05, + "loss": 0.5153, + "step": 12851 + }, + { + "epoch": 0.4053619302949062, + "grad_norm": 1.8647487163543701, + "learning_rate": 1.3478212661971592e-05, + "loss": 0.8024, + "step": 12852 + }, + { + "epoch": 0.40539347106134677, + "grad_norm": 1.3023656606674194, + "learning_rate": 1.3477254872448843e-05, + "loss": 0.4917, + "step": 12853 + }, + { + "epoch": 0.4054250118277874, + "grad_norm": 2.3705790042877197, + "learning_rate": 1.3476297046638244e-05, + "loss": 0.5044, + "step": 12854 + }, + { + "epoch": 0.405456552594228, + "grad_norm": 1.7419824600219727, + "learning_rate": 1.3475339184549783e-05, + "loss": 0.5293, + "step": 12855 + }, + { + "epoch": 0.40548809336066866, + "grad_norm": 1.6674635410308838, + "learning_rate": 1.3474381286193462e-05, + "loss": 0.5101, + "step": 12856 + }, + { + "epoch": 0.4055196341271093, + "grad_norm": 2.0482940673828125, + "learning_rate": 1.3473423351579272e-05, + "loss": 0.4877, + "step": 12857 + }, + { + "epoch": 0.4055511748935499, + "grad_norm": 3.6782968044281006, + "learning_rate": 1.3472465380717215e-05, + "loss": 0.3644, + "step": 12858 + }, + { + "epoch": 0.40558271565999054, + "grad_norm": 1.969359040260315, + "learning_rate": 1.347150737361728e-05, + "loss": 0.5193, + "step": 12859 + }, + { + "epoch": 0.4056142564264312, + "grad_norm": 1.4335517883300781, + "learning_rate": 1.3470549330289474e-05, + "loss": 0.809, + "step": 12860 + }, + { + "epoch": 0.4056457971928718, + "grad_norm": 2.480778694152832, + "learning_rate": 1.3469591250743791e-05, + "loss": 0.5268, + "step": 12861 + }, + { + "epoch": 0.40567733795931243, + "grad_norm": 1.9341102838516235, + "learning_rate": 1.3468633134990228e-05, + "loss": 0.5017, + "step": 12862 + }, + { + "epoch": 0.405708878725753, + "grad_norm": 1.5918793678283691, + "learning_rate": 1.3467674983038785e-05, + "loss": 0.4949, + "step": 12863 + }, + { + "epoch": 0.40574041949219364, + "grad_norm": 2.2502081394195557, + "learning_rate": 1.346671679489946e-05, + "loss": 0.4803, + "step": 12864 + }, + { + "epoch": 0.40577196025863427, + "grad_norm": 1.472274899482727, + "learning_rate": 1.3465758570582257e-05, + "loss": 0.5259, + "step": 12865 + }, + { + "epoch": 0.4058035010250749, + "grad_norm": 1.2947849035263062, + "learning_rate": 1.3464800310097168e-05, + "loss": 0.4861, + "step": 12866 + }, + { + "epoch": 0.4058350417915155, + "grad_norm": 1.7547857761383057, + "learning_rate": 1.34638420134542e-05, + "loss": 0.8379, + "step": 12867 + }, + { + "epoch": 0.40586658255795616, + "grad_norm": 1.6042433977127075, + "learning_rate": 1.3462883680663352e-05, + "loss": 0.5054, + "step": 12868 + }, + { + "epoch": 0.4058981233243968, + "grad_norm": 1.8063815832138062, + "learning_rate": 1.3461925311734622e-05, + "loss": 0.5227, + "step": 12869 + }, + { + "epoch": 0.4059296640908374, + "grad_norm": 1.7858920097351074, + "learning_rate": 1.3460966906678013e-05, + "loss": 0.4897, + "step": 12870 + }, + { + "epoch": 0.40596120485727805, + "grad_norm": 5.687320709228516, + "learning_rate": 1.346000846550353e-05, + "loss": 0.3639, + "step": 12871 + }, + { + "epoch": 0.4059927456237187, + "grad_norm": 3.456059217453003, + "learning_rate": 1.3459049988221166e-05, + "loss": 0.5064, + "step": 12872 + }, + { + "epoch": 0.4060242863901593, + "grad_norm": 1.534090280532837, + "learning_rate": 1.3458091474840935e-05, + "loss": 0.8711, + "step": 12873 + }, + { + "epoch": 0.4060558271565999, + "grad_norm": 1.825630784034729, + "learning_rate": 1.3457132925372832e-05, + "loss": 0.5069, + "step": 12874 + }, + { + "epoch": 0.4060873679230405, + "grad_norm": 2.0841503143310547, + "learning_rate": 1.3456174339826861e-05, + "loss": 0.4765, + "step": 12875 + }, + { + "epoch": 0.40611890868948114, + "grad_norm": 1.5057380199432373, + "learning_rate": 1.3455215718213031e-05, + "loss": 0.544, + "step": 12876 + }, + { + "epoch": 0.40615044945592177, + "grad_norm": 1.7459722757339478, + "learning_rate": 1.3454257060541341e-05, + "loss": 0.5022, + "step": 12877 + }, + { + "epoch": 0.4061819902223624, + "grad_norm": 1.3245738744735718, + "learning_rate": 1.3453298366821797e-05, + "loss": 0.497, + "step": 12878 + }, + { + "epoch": 0.40621353098880303, + "grad_norm": 1.34184992313385, + "learning_rate": 1.34523396370644e-05, + "loss": 0.4847, + "step": 12879 + }, + { + "epoch": 0.40624507175524366, + "grad_norm": 1.3466169834136963, + "learning_rate": 1.3451380871279159e-05, + "loss": 0.4929, + "step": 12880 + }, + { + "epoch": 0.4062766125216843, + "grad_norm": 1.4138635396957397, + "learning_rate": 1.345042206947608e-05, + "loss": 0.4958, + "step": 12881 + }, + { + "epoch": 0.4063081532881249, + "grad_norm": 1.5870765447616577, + "learning_rate": 1.3449463231665167e-05, + "loss": 0.5175, + "step": 12882 + }, + { + "epoch": 0.40633969405456555, + "grad_norm": 1.234911322593689, + "learning_rate": 1.3448504357856427e-05, + "loss": 0.4853, + "step": 12883 + }, + { + "epoch": 0.4063712348210062, + "grad_norm": 1.3569415807724, + "learning_rate": 1.3447545448059864e-05, + "loss": 0.5151, + "step": 12884 + }, + { + "epoch": 0.40640277558744675, + "grad_norm": 1.3363468647003174, + "learning_rate": 1.344658650228549e-05, + "loss": 0.5152, + "step": 12885 + }, + { + "epoch": 0.4064343163538874, + "grad_norm": 3.4663398265838623, + "learning_rate": 1.344562752054331e-05, + "loss": 0.3609, + "step": 12886 + }, + { + "epoch": 0.406465857120328, + "grad_norm": 1.4744824171066284, + "learning_rate": 1.344466850284333e-05, + "loss": 0.5017, + "step": 12887 + }, + { + "epoch": 0.40649739788676864, + "grad_norm": 1.5558463335037231, + "learning_rate": 1.3443709449195561e-05, + "loss": 0.5257, + "step": 12888 + }, + { + "epoch": 0.40652893865320927, + "grad_norm": 1.4694637060165405, + "learning_rate": 1.3442750359610009e-05, + "loss": 0.5119, + "step": 12889 + }, + { + "epoch": 0.4065604794196499, + "grad_norm": 1.7623379230499268, + "learning_rate": 1.3441791234096684e-05, + "loss": 0.3591, + "step": 12890 + }, + { + "epoch": 0.40659202018609053, + "grad_norm": 1.9256658554077148, + "learning_rate": 1.3440832072665592e-05, + "loss": 0.4732, + "step": 12891 + }, + { + "epoch": 0.40662356095253116, + "grad_norm": 1.6258748769760132, + "learning_rate": 1.343987287532675e-05, + "loss": 0.5126, + "step": 12892 + }, + { + "epoch": 0.4066551017189718, + "grad_norm": 1.4010120630264282, + "learning_rate": 1.3438913642090161e-05, + "loss": 0.5258, + "step": 12893 + }, + { + "epoch": 0.4066866424854124, + "grad_norm": 2.1791434288024902, + "learning_rate": 1.3437954372965839e-05, + "loss": 0.4809, + "step": 12894 + }, + { + "epoch": 0.406718183251853, + "grad_norm": 1.502294659614563, + "learning_rate": 1.3436995067963794e-05, + "loss": 0.5057, + "step": 12895 + }, + { + "epoch": 0.4067497240182936, + "grad_norm": 2.2690563201904297, + "learning_rate": 1.3436035727094035e-05, + "loss": 0.5149, + "step": 12896 + }, + { + "epoch": 0.40678126478473425, + "grad_norm": 1.5587096214294434, + "learning_rate": 1.3435076350366577e-05, + "loss": 0.4844, + "step": 12897 + }, + { + "epoch": 0.4068128055511749, + "grad_norm": 1.9638115167617798, + "learning_rate": 1.3434116937791428e-05, + "loss": 0.4326, + "step": 12898 + }, + { + "epoch": 0.4068443463176155, + "grad_norm": 3.497126817703247, + "learning_rate": 1.3433157489378606e-05, + "loss": 0.5115, + "step": 12899 + }, + { + "epoch": 0.40687588708405614, + "grad_norm": 1.7298226356506348, + "learning_rate": 1.3432198005138115e-05, + "loss": 0.4816, + "step": 12900 + }, + { + "epoch": 0.4069074278504968, + "grad_norm": 1.680057406425476, + "learning_rate": 1.3431238485079973e-05, + "loss": 0.4901, + "step": 12901 + }, + { + "epoch": 0.4069389686169374, + "grad_norm": 1.7523072957992554, + "learning_rate": 1.3430278929214197e-05, + "loss": 0.5129, + "step": 12902 + }, + { + "epoch": 0.40697050938337803, + "grad_norm": 1.3028136491775513, + "learning_rate": 1.3429319337550794e-05, + "loss": 0.5279, + "step": 12903 + }, + { + "epoch": 0.40700205014981866, + "grad_norm": 1.4203194379806519, + "learning_rate": 1.3428359710099781e-05, + "loss": 0.5182, + "step": 12904 + }, + { + "epoch": 0.4070335909162593, + "grad_norm": 2.800055742263794, + "learning_rate": 1.3427400046871172e-05, + "loss": 0.5046, + "step": 12905 + }, + { + "epoch": 0.40706513168269987, + "grad_norm": 1.6931322813034058, + "learning_rate": 1.3426440347874985e-05, + "loss": 0.5108, + "step": 12906 + }, + { + "epoch": 0.4070966724491405, + "grad_norm": 1.606050968170166, + "learning_rate": 1.3425480613121234e-05, + "loss": 0.4971, + "step": 12907 + }, + { + "epoch": 0.4071282132155811, + "grad_norm": 1.8511149883270264, + "learning_rate": 1.3424520842619927e-05, + "loss": 0.543, + "step": 12908 + }, + { + "epoch": 0.40715975398202175, + "grad_norm": 1.6007928848266602, + "learning_rate": 1.342356103638109e-05, + "loss": 0.5138, + "step": 12909 + }, + { + "epoch": 0.4071912947484624, + "grad_norm": 1.544298768043518, + "learning_rate": 1.3422601194414733e-05, + "loss": 0.5055, + "step": 12910 + }, + { + "epoch": 0.407222835514903, + "grad_norm": 1.727410078048706, + "learning_rate": 1.3421641316730875e-05, + "loss": 0.5038, + "step": 12911 + }, + { + "epoch": 0.40725437628134364, + "grad_norm": 1.6455308198928833, + "learning_rate": 1.3420681403339535e-05, + "loss": 0.4826, + "step": 12912 + }, + { + "epoch": 0.4072859170477843, + "grad_norm": 1.4139163494110107, + "learning_rate": 1.3419721454250727e-05, + "loss": 0.5411, + "step": 12913 + }, + { + "epoch": 0.4073174578142249, + "grad_norm": 1.9175728559494019, + "learning_rate": 1.341876146947447e-05, + "loss": 0.4563, + "step": 12914 + }, + { + "epoch": 0.40734899858066553, + "grad_norm": 1.5494157075881958, + "learning_rate": 1.3417801449020785e-05, + "loss": 0.455, + "step": 12915 + }, + { + "epoch": 0.4073805393471061, + "grad_norm": 1.484254240989685, + "learning_rate": 1.3416841392899684e-05, + "loss": 0.5231, + "step": 12916 + }, + { + "epoch": 0.40741208011354674, + "grad_norm": 1.4438413381576538, + "learning_rate": 1.3415881301121196e-05, + "loss": 0.4982, + "step": 12917 + }, + { + "epoch": 0.40744362087998737, + "grad_norm": 1.3808872699737549, + "learning_rate": 1.341492117369533e-05, + "loss": 0.5075, + "step": 12918 + }, + { + "epoch": 0.407475161646428, + "grad_norm": 1.5587494373321533, + "learning_rate": 1.341396101063211e-05, + "loss": 0.4829, + "step": 12919 + }, + { + "epoch": 0.4075067024128686, + "grad_norm": 1.3980190753936768, + "learning_rate": 1.3413000811941557e-05, + "loss": 0.4827, + "step": 12920 + }, + { + "epoch": 0.40753824317930926, + "grad_norm": 1.6353049278259277, + "learning_rate": 1.3412040577633687e-05, + "loss": 0.5488, + "step": 12921 + }, + { + "epoch": 0.4075697839457499, + "grad_norm": 1.7989556789398193, + "learning_rate": 1.3411080307718526e-05, + "loss": 0.4925, + "step": 12922 + }, + { + "epoch": 0.4076013247121905, + "grad_norm": 1.5669963359832764, + "learning_rate": 1.3410120002206096e-05, + "loss": 0.4918, + "step": 12923 + }, + { + "epoch": 0.40763286547863115, + "grad_norm": 1.9703177213668823, + "learning_rate": 1.3409159661106415e-05, + "loss": 0.5194, + "step": 12924 + }, + { + "epoch": 0.4076644062450718, + "grad_norm": 1.5881174802780151, + "learning_rate": 1.3408199284429503e-05, + "loss": 0.5278, + "step": 12925 + }, + { + "epoch": 0.4076959470115124, + "grad_norm": 2.023359775543213, + "learning_rate": 1.3407238872185389e-05, + "loss": 0.5101, + "step": 12926 + }, + { + "epoch": 0.407727487777953, + "grad_norm": 1.8081269264221191, + "learning_rate": 1.340627842438409e-05, + "loss": 0.3914, + "step": 12927 + }, + { + "epoch": 0.4077590285443936, + "grad_norm": 1.8231050968170166, + "learning_rate": 1.3405317941035629e-05, + "loss": 0.4879, + "step": 12928 + }, + { + "epoch": 0.40779056931083424, + "grad_norm": 2.363295078277588, + "learning_rate": 1.3404357422150032e-05, + "loss": 0.4636, + "step": 12929 + }, + { + "epoch": 0.40782211007727487, + "grad_norm": 1.9344406127929688, + "learning_rate": 1.3403396867737324e-05, + "loss": 0.4981, + "step": 12930 + }, + { + "epoch": 0.4078536508437155, + "grad_norm": 1.2919788360595703, + "learning_rate": 1.3402436277807524e-05, + "loss": 0.5074, + "step": 12931 + }, + { + "epoch": 0.40788519161015613, + "grad_norm": 1.5783915519714355, + "learning_rate": 1.3401475652370661e-05, + "loss": 0.5202, + "step": 12932 + }, + { + "epoch": 0.40791673237659676, + "grad_norm": 3.5765631198883057, + "learning_rate": 1.3400514991436759e-05, + "loss": 0.8189, + "step": 12933 + }, + { + "epoch": 0.4079482731430374, + "grad_norm": 3.796794891357422, + "learning_rate": 1.3399554295015842e-05, + "loss": 0.8048, + "step": 12934 + }, + { + "epoch": 0.407979813909478, + "grad_norm": 1.740289330482483, + "learning_rate": 1.3398593563117937e-05, + "loss": 0.4913, + "step": 12935 + }, + { + "epoch": 0.40801135467591865, + "grad_norm": 1.9251621961593628, + "learning_rate": 1.3397632795753069e-05, + "loss": 0.5001, + "step": 12936 + }, + { + "epoch": 0.4080428954423593, + "grad_norm": 1.5893391370773315, + "learning_rate": 1.3396671992931264e-05, + "loss": 0.5065, + "step": 12937 + }, + { + "epoch": 0.40807443620879985, + "grad_norm": 1.462001919746399, + "learning_rate": 1.3395711154662549e-05, + "loss": 0.5083, + "step": 12938 + }, + { + "epoch": 0.4081059769752405, + "grad_norm": 1.6027315855026245, + "learning_rate": 1.3394750280956952e-05, + "loss": 0.4865, + "step": 12939 + }, + { + "epoch": 0.4081375177416811, + "grad_norm": 1.6206583976745605, + "learning_rate": 1.33937893718245e-05, + "loss": 0.4464, + "step": 12940 + }, + { + "epoch": 0.40816905850812174, + "grad_norm": 1.6083272695541382, + "learning_rate": 1.339282842727522e-05, + "loss": 0.5016, + "step": 12941 + }, + { + "epoch": 0.40820059927456237, + "grad_norm": 1.946534276008606, + "learning_rate": 1.339186744731914e-05, + "loss": 0.4737, + "step": 12942 + }, + { + "epoch": 0.408232140041003, + "grad_norm": 1.656105637550354, + "learning_rate": 1.3390906431966294e-05, + "loss": 0.5038, + "step": 12943 + }, + { + "epoch": 0.40826368080744363, + "grad_norm": 1.8643897771835327, + "learning_rate": 1.33899453812267e-05, + "loss": 0.5147, + "step": 12944 + }, + { + "epoch": 0.40829522157388426, + "grad_norm": 1.996848225593567, + "learning_rate": 1.3388984295110397e-05, + "loss": 0.4892, + "step": 12945 + }, + { + "epoch": 0.4083267623403249, + "grad_norm": 1.3896607160568237, + "learning_rate": 1.3388023173627413e-05, + "loss": 0.4846, + "step": 12946 + }, + { + "epoch": 0.4083583031067655, + "grad_norm": 1.3476753234863281, + "learning_rate": 1.3387062016787775e-05, + "loss": 0.5039, + "step": 12947 + }, + { + "epoch": 0.4083898438732061, + "grad_norm": 1.6987648010253906, + "learning_rate": 1.3386100824601513e-05, + "loss": 0.5063, + "step": 12948 + }, + { + "epoch": 0.4084213846396467, + "grad_norm": 1.845091700553894, + "learning_rate": 1.3385139597078663e-05, + "loss": 0.4984, + "step": 12949 + }, + { + "epoch": 0.40845292540608735, + "grad_norm": 1.4759856462478638, + "learning_rate": 1.3384178334229251e-05, + "loss": 0.4701, + "step": 12950 + }, + { + "epoch": 0.408484466172528, + "grad_norm": 1.6522160768508911, + "learning_rate": 1.338321703606331e-05, + "loss": 0.5249, + "step": 12951 + }, + { + "epoch": 0.4085160069389686, + "grad_norm": 1.9085229635238647, + "learning_rate": 1.3382255702590871e-05, + "loss": 0.3994, + "step": 12952 + }, + { + "epoch": 0.40854754770540924, + "grad_norm": 1.4552284479141235, + "learning_rate": 1.3381294333821968e-05, + "loss": 0.5216, + "step": 12953 + }, + { + "epoch": 0.40857908847184987, + "grad_norm": 1.398530125617981, + "learning_rate": 1.3380332929766637e-05, + "loss": 0.5515, + "step": 12954 + }, + { + "epoch": 0.4086106292382905, + "grad_norm": 1.645007848739624, + "learning_rate": 1.3379371490434902e-05, + "loss": 0.4989, + "step": 12955 + }, + { + "epoch": 0.40864217000473113, + "grad_norm": 2.437257766723633, + "learning_rate": 1.3378410015836805e-05, + "loss": 0.8345, + "step": 12956 + }, + { + "epoch": 0.40867371077117176, + "grad_norm": 1.2496237754821777, + "learning_rate": 1.3377448505982378e-05, + "loss": 0.5176, + "step": 12957 + }, + { + "epoch": 0.4087052515376124, + "grad_norm": 1.5425469875335693, + "learning_rate": 1.337648696088165e-05, + "loss": 0.5345, + "step": 12958 + }, + { + "epoch": 0.40873679230405296, + "grad_norm": 1.6357403993606567, + "learning_rate": 1.3375525380544657e-05, + "loss": 0.4704, + "step": 12959 + }, + { + "epoch": 0.4087683330704936, + "grad_norm": 1.6503931283950806, + "learning_rate": 1.3374563764981436e-05, + "loss": 0.4947, + "step": 12960 + }, + { + "epoch": 0.4087998738369342, + "grad_norm": 1.5901663303375244, + "learning_rate": 1.3373602114202028e-05, + "loss": 0.5237, + "step": 12961 + }, + { + "epoch": 0.40883141460337485, + "grad_norm": 1.9965895414352417, + "learning_rate": 1.3372640428216455e-05, + "loss": 0.489, + "step": 12962 + }, + { + "epoch": 0.4088629553698155, + "grad_norm": 1.806534767150879, + "learning_rate": 1.3371678707034763e-05, + "loss": 0.8395, + "step": 12963 + }, + { + "epoch": 0.4088944961362561, + "grad_norm": 2.2444024085998535, + "learning_rate": 1.3370716950666987e-05, + "loss": 0.4649, + "step": 12964 + }, + { + "epoch": 0.40892603690269674, + "grad_norm": 1.4154011011123657, + "learning_rate": 1.3369755159123161e-05, + "loss": 0.5299, + "step": 12965 + }, + { + "epoch": 0.4089575776691374, + "grad_norm": 1.6839720010757446, + "learning_rate": 1.3368793332413326e-05, + "loss": 0.4987, + "step": 12966 + }, + { + "epoch": 0.408989118435578, + "grad_norm": 1.8305432796478271, + "learning_rate": 1.3367831470547511e-05, + "loss": 0.4907, + "step": 12967 + }, + { + "epoch": 0.40902065920201863, + "grad_norm": 1.9716991186141968, + "learning_rate": 1.3366869573535766e-05, + "loss": 0.5223, + "step": 12968 + }, + { + "epoch": 0.4090521999684592, + "grad_norm": 1.3314261436462402, + "learning_rate": 1.3365907641388118e-05, + "loss": 0.5271, + "step": 12969 + }, + { + "epoch": 0.40908374073489984, + "grad_norm": 3.215193271636963, + "learning_rate": 1.3364945674114611e-05, + "loss": 0.5052, + "step": 12970 + }, + { + "epoch": 0.40911528150134047, + "grad_norm": 1.8351720571517944, + "learning_rate": 1.3363983671725284e-05, + "loss": 0.481, + "step": 12971 + }, + { + "epoch": 0.4091468222677811, + "grad_norm": 1.6439567804336548, + "learning_rate": 1.3363021634230176e-05, + "loss": 0.5074, + "step": 12972 + }, + { + "epoch": 0.4091783630342217, + "grad_norm": 2.9772958755493164, + "learning_rate": 1.3362059561639324e-05, + "loss": 0.5192, + "step": 12973 + }, + { + "epoch": 0.40920990380066236, + "grad_norm": 1.6384252309799194, + "learning_rate": 1.3361097453962773e-05, + "loss": 0.5433, + "step": 12974 + }, + { + "epoch": 0.409241444567103, + "grad_norm": 1.3382834196090698, + "learning_rate": 1.336013531121056e-05, + "loss": 0.499, + "step": 12975 + }, + { + "epoch": 0.4092729853335436, + "grad_norm": 1.8279705047607422, + "learning_rate": 1.3359173133392722e-05, + "loss": 0.5068, + "step": 12976 + }, + { + "epoch": 0.40930452609998424, + "grad_norm": 1.5062892436981201, + "learning_rate": 1.3358210920519308e-05, + "loss": 0.5222, + "step": 12977 + }, + { + "epoch": 0.4093360668664249, + "grad_norm": 1.5359746217727661, + "learning_rate": 1.3357248672600353e-05, + "loss": 0.4948, + "step": 12978 + }, + { + "epoch": 0.4093676076328655, + "grad_norm": 2.576648235321045, + "learning_rate": 1.3356286389645902e-05, + "loss": 0.4036, + "step": 12979 + }, + { + "epoch": 0.4093991483993061, + "grad_norm": 1.9927549362182617, + "learning_rate": 1.3355324071665995e-05, + "loss": 0.5302, + "step": 12980 + }, + { + "epoch": 0.4094306891657467, + "grad_norm": 1.3855546712875366, + "learning_rate": 1.3354361718670678e-05, + "loss": 0.489, + "step": 12981 + }, + { + "epoch": 0.40946222993218734, + "grad_norm": 1.8382467031478882, + "learning_rate": 1.3353399330669991e-05, + "loss": 0.7774, + "step": 12982 + }, + { + "epoch": 0.40949377069862797, + "grad_norm": 1.6380269527435303, + "learning_rate": 1.3352436907673979e-05, + "loss": 0.5052, + "step": 12983 + }, + { + "epoch": 0.4095253114650686, + "grad_norm": 1.4505506753921509, + "learning_rate": 1.3351474449692685e-05, + "loss": 0.3659, + "step": 12984 + }, + { + "epoch": 0.4095568522315092, + "grad_norm": 2.0523180961608887, + "learning_rate": 1.3350511956736154e-05, + "loss": 0.8294, + "step": 12985 + }, + { + "epoch": 0.40958839299794986, + "grad_norm": 1.6168971061706543, + "learning_rate": 1.3349549428814426e-05, + "loss": 0.5061, + "step": 12986 + }, + { + "epoch": 0.4096199337643905, + "grad_norm": 1.577717900276184, + "learning_rate": 1.334858686593755e-05, + "loss": 0.8406, + "step": 12987 + }, + { + "epoch": 0.4096514745308311, + "grad_norm": 1.4610062837600708, + "learning_rate": 1.334762426811557e-05, + "loss": 0.4985, + "step": 12988 + }, + { + "epoch": 0.40968301529727175, + "grad_norm": 1.988088846206665, + "learning_rate": 1.3346661635358532e-05, + "loss": 0.3769, + "step": 12989 + }, + { + "epoch": 0.4097145560637123, + "grad_norm": 2.4158363342285156, + "learning_rate": 1.334569896767648e-05, + "loss": 0.4902, + "step": 12990 + }, + { + "epoch": 0.40974609683015295, + "grad_norm": 2.497387647628784, + "learning_rate": 1.334473626507946e-05, + "loss": 0.4936, + "step": 12991 + }, + { + "epoch": 0.4097776375965936, + "grad_norm": 1.70682692527771, + "learning_rate": 1.3343773527577524e-05, + "loss": 0.4831, + "step": 12992 + }, + { + "epoch": 0.4098091783630342, + "grad_norm": 1.4505926370620728, + "learning_rate": 1.334281075518071e-05, + "loss": 0.5088, + "step": 12993 + }, + { + "epoch": 0.40984071912947484, + "grad_norm": 1.381264328956604, + "learning_rate": 1.3341847947899076e-05, + "loss": 0.5219, + "step": 12994 + }, + { + "epoch": 0.40987225989591547, + "grad_norm": 1.300207257270813, + "learning_rate": 1.334088510574266e-05, + "loss": 0.4952, + "step": 12995 + }, + { + "epoch": 0.4099038006623561, + "grad_norm": 2.4193673133850098, + "learning_rate": 1.3339922228721516e-05, + "loss": 0.4494, + "step": 12996 + }, + { + "epoch": 0.40993534142879673, + "grad_norm": 2.0231809616088867, + "learning_rate": 1.3338959316845688e-05, + "loss": 0.4844, + "step": 12997 + }, + { + "epoch": 0.40996688219523736, + "grad_norm": 2.179949998855591, + "learning_rate": 1.3337996370125228e-05, + "loss": 0.5011, + "step": 12998 + }, + { + "epoch": 0.409998422961678, + "grad_norm": 5.307976245880127, + "learning_rate": 1.3337033388570183e-05, + "loss": 0.4974, + "step": 12999 + }, + { + "epoch": 0.4100299637281186, + "grad_norm": 2.2678492069244385, + "learning_rate": 1.3336070372190604e-05, + "loss": 0.4867, + "step": 13000 + }, + { + "epoch": 0.4100615044945592, + "grad_norm": 1.6524783372879028, + "learning_rate": 1.3335107320996538e-05, + "loss": 0.507, + "step": 13001 + }, + { + "epoch": 0.4100930452609998, + "grad_norm": 2.185586929321289, + "learning_rate": 1.3334144234998041e-05, + "loss": 0.3839, + "step": 13002 + }, + { + "epoch": 0.41012458602744045, + "grad_norm": 2.1872379779815674, + "learning_rate": 1.3333181114205159e-05, + "loss": 0.4978, + "step": 13003 + }, + { + "epoch": 0.4101561267938811, + "grad_norm": 1.4683234691619873, + "learning_rate": 1.3332217958627945e-05, + "loss": 0.5046, + "step": 13004 + }, + { + "epoch": 0.4101876675603217, + "grad_norm": 1.892026662826538, + "learning_rate": 1.3331254768276449e-05, + "loss": 0.4998, + "step": 13005 + }, + { + "epoch": 0.41021920832676234, + "grad_norm": 1.658994436264038, + "learning_rate": 1.333029154316072e-05, + "loss": 0.5143, + "step": 13006 + }, + { + "epoch": 0.41025074909320297, + "grad_norm": 1.6011605262756348, + "learning_rate": 1.3329328283290814e-05, + "loss": 0.5172, + "step": 13007 + }, + { + "epoch": 0.4102822898596436, + "grad_norm": 1.7696503400802612, + "learning_rate": 1.3328364988676782e-05, + "loss": 0.5186, + "step": 13008 + }, + { + "epoch": 0.41031383062608423, + "grad_norm": 1.5604407787322998, + "learning_rate": 1.3327401659328677e-05, + "loss": 0.5342, + "step": 13009 + }, + { + "epoch": 0.41034537139252486, + "grad_norm": 2.2261641025543213, + "learning_rate": 1.3326438295256553e-05, + "loss": 0.5035, + "step": 13010 + }, + { + "epoch": 0.4103769121589655, + "grad_norm": 1.5521715879440308, + "learning_rate": 1.3325474896470463e-05, + "loss": 0.4947, + "step": 13011 + }, + { + "epoch": 0.41040845292540606, + "grad_norm": 5.275264263153076, + "learning_rate": 1.3324511462980461e-05, + "loss": 0.8337, + "step": 13012 + }, + { + "epoch": 0.4104399936918467, + "grad_norm": 1.391460657119751, + "learning_rate": 1.3323547994796597e-05, + "loss": 0.5264, + "step": 13013 + }, + { + "epoch": 0.4104715344582873, + "grad_norm": 2.482741355895996, + "learning_rate": 1.332258449192893e-05, + "loss": 0.4914, + "step": 13014 + }, + { + "epoch": 0.41050307522472795, + "grad_norm": 1.7361596822738647, + "learning_rate": 1.3321620954387517e-05, + "loss": 0.5102, + "step": 13015 + }, + { + "epoch": 0.4105346159911686, + "grad_norm": 1.736751914024353, + "learning_rate": 1.332065738218241e-05, + "loss": 0.5044, + "step": 13016 + }, + { + "epoch": 0.4105661567576092, + "grad_norm": 1.509481430053711, + "learning_rate": 1.3319693775323661e-05, + "loss": 0.4917, + "step": 13017 + }, + { + "epoch": 0.41059769752404984, + "grad_norm": 1.4508785009384155, + "learning_rate": 1.3318730133821331e-05, + "loss": 0.4934, + "step": 13018 + }, + { + "epoch": 0.4106292382904905, + "grad_norm": 1.517014980316162, + "learning_rate": 1.3317766457685476e-05, + "loss": 0.4873, + "step": 13019 + }, + { + "epoch": 0.4106607790569311, + "grad_norm": 1.7765566110610962, + "learning_rate": 1.3316802746926151e-05, + "loss": 0.4911, + "step": 13020 + }, + { + "epoch": 0.41069231982337173, + "grad_norm": 1.3122830390930176, + "learning_rate": 1.3315839001553415e-05, + "loss": 0.4976, + "step": 13021 + }, + { + "epoch": 0.4107238605898123, + "grad_norm": 2.1506361961364746, + "learning_rate": 1.3314875221577324e-05, + "loss": 0.458, + "step": 13022 + }, + { + "epoch": 0.41075540135625294, + "grad_norm": 2.321147918701172, + "learning_rate": 1.3313911407007935e-05, + "loss": 0.3987, + "step": 13023 + }, + { + "epoch": 0.41078694212269357, + "grad_norm": 1.7242107391357422, + "learning_rate": 1.3312947557855309e-05, + "loss": 0.5179, + "step": 13024 + }, + { + "epoch": 0.4108184828891342, + "grad_norm": 1.4333714246749878, + "learning_rate": 1.3311983674129503e-05, + "loss": 0.534, + "step": 13025 + }, + { + "epoch": 0.4108500236555748, + "grad_norm": 1.4202133417129517, + "learning_rate": 1.3311019755840576e-05, + "loss": 0.5059, + "step": 13026 + }, + { + "epoch": 0.41088156442201546, + "grad_norm": 1.6981040239334106, + "learning_rate": 1.3310055802998584e-05, + "loss": 0.502, + "step": 13027 + }, + { + "epoch": 0.4109131051884561, + "grad_norm": 1.3998878002166748, + "learning_rate": 1.3309091815613591e-05, + "loss": 0.5127, + "step": 13028 + }, + { + "epoch": 0.4109446459548967, + "grad_norm": 1.332697868347168, + "learning_rate": 1.3308127793695658e-05, + "loss": 0.5176, + "step": 13029 + }, + { + "epoch": 0.41097618672133734, + "grad_norm": 1.4624550342559814, + "learning_rate": 1.3307163737254839e-05, + "loss": 0.5139, + "step": 13030 + }, + { + "epoch": 0.411007727487778, + "grad_norm": 1.5238062143325806, + "learning_rate": 1.33061996463012e-05, + "loss": 0.4883, + "step": 13031 + }, + { + "epoch": 0.4110392682542186, + "grad_norm": 1.755798578262329, + "learning_rate": 1.3305235520844796e-05, + "loss": 0.4781, + "step": 13032 + }, + { + "epoch": 0.4110708090206592, + "grad_norm": 1.4585715532302856, + "learning_rate": 1.33042713608957e-05, + "loss": 0.5386, + "step": 13033 + }, + { + "epoch": 0.4111023497870998, + "grad_norm": 3.034451723098755, + "learning_rate": 1.3303307166463966e-05, + "loss": 0.8335, + "step": 13034 + }, + { + "epoch": 0.41113389055354044, + "grad_norm": 1.7317488193511963, + "learning_rate": 1.3302342937559654e-05, + "loss": 0.5224, + "step": 13035 + }, + { + "epoch": 0.41116543131998107, + "grad_norm": 1.4575871229171753, + "learning_rate": 1.3301378674192831e-05, + "loss": 0.488, + "step": 13036 + }, + { + "epoch": 0.4111969720864217, + "grad_norm": 2.3432676792144775, + "learning_rate": 1.3300414376373556e-05, + "loss": 0.8271, + "step": 13037 + }, + { + "epoch": 0.4112285128528623, + "grad_norm": 1.5262641906738281, + "learning_rate": 1.3299450044111892e-05, + "loss": 0.499, + "step": 13038 + }, + { + "epoch": 0.41126005361930296, + "grad_norm": 1.4509665966033936, + "learning_rate": 1.329848567741791e-05, + "loss": 0.52, + "step": 13039 + }, + { + "epoch": 0.4112915943857436, + "grad_norm": 1.5687121152877808, + "learning_rate": 1.3297521276301666e-05, + "loss": 0.3868, + "step": 13040 + }, + { + "epoch": 0.4113231351521842, + "grad_norm": 1.5570182800292969, + "learning_rate": 1.3296556840773227e-05, + "loss": 0.5074, + "step": 13041 + }, + { + "epoch": 0.41135467591862485, + "grad_norm": 1.4697113037109375, + "learning_rate": 1.3295592370842657e-05, + "loss": 0.4937, + "step": 13042 + }, + { + "epoch": 0.4113862166850654, + "grad_norm": 1.7489067316055298, + "learning_rate": 1.3294627866520026e-05, + "loss": 0.5381, + "step": 13043 + }, + { + "epoch": 0.41141775745150605, + "grad_norm": 1.6513501405715942, + "learning_rate": 1.3293663327815395e-05, + "loss": 0.4977, + "step": 13044 + }, + { + "epoch": 0.4114492982179467, + "grad_norm": 1.9014222621917725, + "learning_rate": 1.3292698754738825e-05, + "loss": 0.8509, + "step": 13045 + }, + { + "epoch": 0.4114808389843873, + "grad_norm": 1.7999684810638428, + "learning_rate": 1.329173414730039e-05, + "loss": 0.5169, + "step": 13046 + }, + { + "epoch": 0.41151237975082794, + "grad_norm": 1.78762948513031, + "learning_rate": 1.329076950551015e-05, + "loss": 0.5186, + "step": 13047 + }, + { + "epoch": 0.41154392051726857, + "grad_norm": 1.5270320177078247, + "learning_rate": 1.3289804829378177e-05, + "loss": 0.5137, + "step": 13048 + }, + { + "epoch": 0.4115754612837092, + "grad_norm": 1.821252465248108, + "learning_rate": 1.3288840118914534e-05, + "loss": 0.5269, + "step": 13049 + }, + { + "epoch": 0.41160700205014983, + "grad_norm": 1.9439679384231567, + "learning_rate": 1.3287875374129295e-05, + "loss": 0.4004, + "step": 13050 + }, + { + "epoch": 0.41163854281659046, + "grad_norm": 2.6941938400268555, + "learning_rate": 1.3286910595032521e-05, + "loss": 0.5285, + "step": 13051 + }, + { + "epoch": 0.4116700835830311, + "grad_norm": 1.5249072313308716, + "learning_rate": 1.3285945781634281e-05, + "loss": 0.5192, + "step": 13052 + }, + { + "epoch": 0.4117016243494717, + "grad_norm": 1.976884126663208, + "learning_rate": 1.328498093394465e-05, + "loss": 0.4913, + "step": 13053 + }, + { + "epoch": 0.4117331651159123, + "grad_norm": 1.4406310319900513, + "learning_rate": 1.328401605197369e-05, + "loss": 0.4917, + "step": 13054 + }, + { + "epoch": 0.4117647058823529, + "grad_norm": 2.33551287651062, + "learning_rate": 1.328305113573147e-05, + "loss": 0.4936, + "step": 13055 + }, + { + "epoch": 0.41179624664879355, + "grad_norm": 2.418264150619507, + "learning_rate": 1.3282086185228063e-05, + "loss": 0.4718, + "step": 13056 + }, + { + "epoch": 0.4118277874152342, + "grad_norm": 2.3716256618499756, + "learning_rate": 1.3281121200473542e-05, + "loss": 0.5277, + "step": 13057 + }, + { + "epoch": 0.4118593281816748, + "grad_norm": 2.127584934234619, + "learning_rate": 1.3280156181477968e-05, + "loss": 0.5249, + "step": 13058 + }, + { + "epoch": 0.41189086894811544, + "grad_norm": 1.9606834650039673, + "learning_rate": 1.3279191128251419e-05, + "loss": 0.4879, + "step": 13059 + }, + { + "epoch": 0.41192240971455607, + "grad_norm": 2.183964967727661, + "learning_rate": 1.3278226040803968e-05, + "loss": 0.4343, + "step": 13060 + }, + { + "epoch": 0.4119539504809967, + "grad_norm": 1.856233835220337, + "learning_rate": 1.3277260919145678e-05, + "loss": 0.4787, + "step": 13061 + }, + { + "epoch": 0.41198549124743733, + "grad_norm": 1.9456989765167236, + "learning_rate": 1.3276295763286629e-05, + "loss": 0.4804, + "step": 13062 + }, + { + "epoch": 0.41201703201387796, + "grad_norm": 1.8191887140274048, + "learning_rate": 1.327533057323689e-05, + "loss": 0.5099, + "step": 13063 + }, + { + "epoch": 0.4120485727803186, + "grad_norm": 1.7210907936096191, + "learning_rate": 1.3274365349006533e-05, + "loss": 0.5156, + "step": 13064 + }, + { + "epoch": 0.41208011354675916, + "grad_norm": 1.795464038848877, + "learning_rate": 1.3273400090605628e-05, + "loss": 0.5204, + "step": 13065 + }, + { + "epoch": 0.4121116543131998, + "grad_norm": 1.4656838178634644, + "learning_rate": 1.3272434798044251e-05, + "loss": 0.4877, + "step": 13066 + }, + { + "epoch": 0.4121431950796404, + "grad_norm": 1.705336093902588, + "learning_rate": 1.3271469471332481e-05, + "loss": 0.5042, + "step": 13067 + }, + { + "epoch": 0.41217473584608105, + "grad_norm": 1.5954818725585938, + "learning_rate": 1.3270504110480383e-05, + "loss": 0.5048, + "step": 13068 + }, + { + "epoch": 0.4122062766125217, + "grad_norm": 1.6537550687789917, + "learning_rate": 1.3269538715498036e-05, + "loss": 0.4786, + "step": 13069 + }, + { + "epoch": 0.4122378173789623, + "grad_norm": 1.5623352527618408, + "learning_rate": 1.3268573286395514e-05, + "loss": 0.5003, + "step": 13070 + }, + { + "epoch": 0.41226935814540294, + "grad_norm": 1.9529898166656494, + "learning_rate": 1.326760782318289e-05, + "loss": 0.4935, + "step": 13071 + }, + { + "epoch": 0.41230089891184357, + "grad_norm": 2.150980234146118, + "learning_rate": 1.3266642325870242e-05, + "loss": 0.3647, + "step": 13072 + }, + { + "epoch": 0.4123324396782842, + "grad_norm": 1.4121350049972534, + "learning_rate": 1.3265676794467648e-05, + "loss": 0.3887, + "step": 13073 + }, + { + "epoch": 0.41236398044472483, + "grad_norm": 1.5008212327957153, + "learning_rate": 1.3264711228985178e-05, + "loss": 0.4701, + "step": 13074 + }, + { + "epoch": 0.4123955212111654, + "grad_norm": 2.4045474529266357, + "learning_rate": 1.3263745629432912e-05, + "loss": 0.3617, + "step": 13075 + }, + { + "epoch": 0.41242706197760604, + "grad_norm": 1.7665297985076904, + "learning_rate": 1.3262779995820924e-05, + "loss": 0.8381, + "step": 13076 + }, + { + "epoch": 0.41245860274404667, + "grad_norm": 1.9323748350143433, + "learning_rate": 1.3261814328159297e-05, + "loss": 0.5297, + "step": 13077 + }, + { + "epoch": 0.4124901435104873, + "grad_norm": 1.5346750020980835, + "learning_rate": 1.3260848626458102e-05, + "loss": 0.5301, + "step": 13078 + }, + { + "epoch": 0.4125216842769279, + "grad_norm": 1.69135320186615, + "learning_rate": 1.325988289072742e-05, + "loss": 0.4922, + "step": 13079 + }, + { + "epoch": 0.41255322504336855, + "grad_norm": 1.524561882019043, + "learning_rate": 1.3258917120977327e-05, + "loss": 0.498, + "step": 13080 + }, + { + "epoch": 0.4125847658098092, + "grad_norm": 1.842903971672058, + "learning_rate": 1.325795131721791e-05, + "loss": 0.4977, + "step": 13081 + }, + { + "epoch": 0.4126163065762498, + "grad_norm": 1.395859956741333, + "learning_rate": 1.3256985479459235e-05, + "loss": 0.4838, + "step": 13082 + }, + { + "epoch": 0.41264784734269044, + "grad_norm": 1.4807144403457642, + "learning_rate": 1.325601960771139e-05, + "loss": 0.5217, + "step": 13083 + }, + { + "epoch": 0.4126793881091311, + "grad_norm": 1.5966582298278809, + "learning_rate": 1.3255053701984455e-05, + "loss": 0.4968, + "step": 13084 + }, + { + "epoch": 0.4127109288755717, + "grad_norm": 1.6603448390960693, + "learning_rate": 1.32540877622885e-05, + "loss": 0.5425, + "step": 13085 + }, + { + "epoch": 0.4127424696420123, + "grad_norm": 1.5852751731872559, + "learning_rate": 1.3253121788633618e-05, + "loss": 0.5258, + "step": 13086 + }, + { + "epoch": 0.4127740104084529, + "grad_norm": 1.5624303817749023, + "learning_rate": 1.325215578102988e-05, + "loss": 0.5191, + "step": 13087 + }, + { + "epoch": 0.41280555117489354, + "grad_norm": 1.5644702911376953, + "learning_rate": 1.3251189739487376e-05, + "loss": 0.488, + "step": 13088 + }, + { + "epoch": 0.41283709194133417, + "grad_norm": 1.6402162313461304, + "learning_rate": 1.325022366401618e-05, + "loss": 0.4943, + "step": 13089 + }, + { + "epoch": 0.4128686327077748, + "grad_norm": 1.7545949220657349, + "learning_rate": 1.3249257554626375e-05, + "loss": 0.4824, + "step": 13090 + }, + { + "epoch": 0.4129001734742154, + "grad_norm": 1.3055793046951294, + "learning_rate": 1.3248291411328048e-05, + "loss": 0.4948, + "step": 13091 + }, + { + "epoch": 0.41293171424065606, + "grad_norm": 2.046926498413086, + "learning_rate": 1.3247325234131275e-05, + "loss": 0.4803, + "step": 13092 + }, + { + "epoch": 0.4129632550070967, + "grad_norm": 7.081000328063965, + "learning_rate": 1.3246359023046146e-05, + "loss": 0.5269, + "step": 13093 + }, + { + "epoch": 0.4129947957735373, + "grad_norm": 1.4279532432556152, + "learning_rate": 1.3245392778082736e-05, + "loss": 0.4877, + "step": 13094 + }, + { + "epoch": 0.41302633653997795, + "grad_norm": 1.6979217529296875, + "learning_rate": 1.324442649925113e-05, + "loss": 0.5284, + "step": 13095 + }, + { + "epoch": 0.4130578773064185, + "grad_norm": 1.9120252132415771, + "learning_rate": 1.3243460186561419e-05, + "loss": 0.5201, + "step": 13096 + }, + { + "epoch": 0.41308941807285915, + "grad_norm": 1.4329794645309448, + "learning_rate": 1.3242493840023678e-05, + "loss": 0.5058, + "step": 13097 + }, + { + "epoch": 0.4131209588392998, + "grad_norm": 1.5876648426055908, + "learning_rate": 1.3241527459648e-05, + "loss": 0.4753, + "step": 13098 + }, + { + "epoch": 0.4131524996057404, + "grad_norm": 1.8581980466842651, + "learning_rate": 1.3240561045444464e-05, + "loss": 0.8201, + "step": 13099 + }, + { + "epoch": 0.41318404037218104, + "grad_norm": 2.589836835861206, + "learning_rate": 1.3239594597423157e-05, + "loss": 0.3955, + "step": 13100 + }, + { + "epoch": 0.41321558113862167, + "grad_norm": 1.5503877401351929, + "learning_rate": 1.3238628115594168e-05, + "loss": 0.4919, + "step": 13101 + }, + { + "epoch": 0.4132471219050623, + "grad_norm": 1.7380894422531128, + "learning_rate": 1.3237661599967576e-05, + "loss": 0.4913, + "step": 13102 + }, + { + "epoch": 0.4132786626715029, + "grad_norm": 1.682708740234375, + "learning_rate": 1.3236695050553474e-05, + "loss": 0.4859, + "step": 13103 + }, + { + "epoch": 0.41331020343794356, + "grad_norm": 2.553917169570923, + "learning_rate": 1.3235728467361944e-05, + "loss": 0.4287, + "step": 13104 + }, + { + "epoch": 0.4133417442043842, + "grad_norm": 1.5769540071487427, + "learning_rate": 1.3234761850403077e-05, + "loss": 0.509, + "step": 13105 + }, + { + "epoch": 0.4133732849708248, + "grad_norm": 3.1044762134552, + "learning_rate": 1.3233795199686957e-05, + "loss": 0.3987, + "step": 13106 + }, + { + "epoch": 0.4134048257372654, + "grad_norm": 1.4337117671966553, + "learning_rate": 1.3232828515223672e-05, + "loss": 0.5332, + "step": 13107 + }, + { + "epoch": 0.413436366503706, + "grad_norm": 1.955718994140625, + "learning_rate": 1.3231861797023313e-05, + "loss": 0.501, + "step": 13108 + }, + { + "epoch": 0.41346790727014665, + "grad_norm": 1.7049492597579956, + "learning_rate": 1.3230895045095963e-05, + "loss": 0.5017, + "step": 13109 + }, + { + "epoch": 0.4134994480365873, + "grad_norm": 1.8372632265090942, + "learning_rate": 1.3229928259451718e-05, + "loss": 0.5007, + "step": 13110 + }, + { + "epoch": 0.4135309888030279, + "grad_norm": 1.5504090785980225, + "learning_rate": 1.3228961440100665e-05, + "loss": 0.4947, + "step": 13111 + }, + { + "epoch": 0.41356252956946854, + "grad_norm": 1.6202448606491089, + "learning_rate": 1.322799458705289e-05, + "loss": 0.5, + "step": 13112 + }, + { + "epoch": 0.41359407033590917, + "grad_norm": 1.3351471424102783, + "learning_rate": 1.3227027700318487e-05, + "loss": 0.4815, + "step": 13113 + }, + { + "epoch": 0.4136256111023498, + "grad_norm": 1.6970818042755127, + "learning_rate": 1.3226060779907542e-05, + "loss": 0.3998, + "step": 13114 + }, + { + "epoch": 0.41365715186879043, + "grad_norm": 1.8424413204193115, + "learning_rate": 1.3225093825830148e-05, + "loss": 0.4774, + "step": 13115 + }, + { + "epoch": 0.41368869263523106, + "grad_norm": 1.2833406925201416, + "learning_rate": 1.3224126838096396e-05, + "loss": 0.5038, + "step": 13116 + }, + { + "epoch": 0.41372023340167163, + "grad_norm": 2.8232932090759277, + "learning_rate": 1.3223159816716376e-05, + "loss": 0.4903, + "step": 13117 + }, + { + "epoch": 0.41375177416811226, + "grad_norm": 1.718022108078003, + "learning_rate": 1.3222192761700183e-05, + "loss": 0.5444, + "step": 13118 + }, + { + "epoch": 0.4137833149345529, + "grad_norm": 1.9859023094177246, + "learning_rate": 1.3221225673057904e-05, + "loss": 0.4968, + "step": 13119 + }, + { + "epoch": 0.4138148557009935, + "grad_norm": 1.3999322652816772, + "learning_rate": 1.3220258550799635e-05, + "loss": 0.5158, + "step": 13120 + }, + { + "epoch": 0.41384639646743415, + "grad_norm": 1.6237003803253174, + "learning_rate": 1.3219291394935469e-05, + "loss": 0.529, + "step": 13121 + }, + { + "epoch": 0.4138779372338748, + "grad_norm": 1.327229380607605, + "learning_rate": 1.3218324205475494e-05, + "loss": 0.5055, + "step": 13122 + }, + { + "epoch": 0.4139094780003154, + "grad_norm": 1.8588154315948486, + "learning_rate": 1.3217356982429812e-05, + "loss": 0.5032, + "step": 13123 + }, + { + "epoch": 0.41394101876675604, + "grad_norm": 1.4147849082946777, + "learning_rate": 1.3216389725808507e-05, + "loss": 0.5225, + "step": 13124 + }, + { + "epoch": 0.41397255953319667, + "grad_norm": 1.327912449836731, + "learning_rate": 1.3215422435621681e-05, + "loss": 0.501, + "step": 13125 + }, + { + "epoch": 0.4140041002996373, + "grad_norm": 1.909860610961914, + "learning_rate": 1.3214455111879422e-05, + "loss": 0.5145, + "step": 13126 + }, + { + "epoch": 0.41403564106607793, + "grad_norm": 2.080510139465332, + "learning_rate": 1.3213487754591831e-05, + "loss": 0.5346, + "step": 13127 + }, + { + "epoch": 0.4140671818325185, + "grad_norm": 2.5705933570861816, + "learning_rate": 1.3212520363768996e-05, + "loss": 0.5146, + "step": 13128 + }, + { + "epoch": 0.41409872259895913, + "grad_norm": 1.3819522857666016, + "learning_rate": 1.3211552939421022e-05, + "loss": 0.4531, + "step": 13129 + }, + { + "epoch": 0.41413026336539976, + "grad_norm": 1.6155320405960083, + "learning_rate": 1.3210585481557996e-05, + "loss": 0.4714, + "step": 13130 + }, + { + "epoch": 0.4141618041318404, + "grad_norm": 2.2372629642486572, + "learning_rate": 1.320961799019002e-05, + "loss": 0.5274, + "step": 13131 + }, + { + "epoch": 0.414193344898281, + "grad_norm": 1.7428778409957886, + "learning_rate": 1.3208650465327186e-05, + "loss": 0.4942, + "step": 13132 + }, + { + "epoch": 0.41422488566472165, + "grad_norm": 1.927276849746704, + "learning_rate": 1.3207682906979593e-05, + "loss": 0.4961, + "step": 13133 + }, + { + "epoch": 0.4142564264311623, + "grad_norm": 1.6450493335723877, + "learning_rate": 1.3206715315157337e-05, + "loss": 0.4724, + "step": 13134 + }, + { + "epoch": 0.4142879671976029, + "grad_norm": 2.000446319580078, + "learning_rate": 1.3205747689870519e-05, + "loss": 0.8394, + "step": 13135 + }, + { + "epoch": 0.41431950796404354, + "grad_norm": 1.7950677871704102, + "learning_rate": 1.3204780031129235e-05, + "loss": 0.4787, + "step": 13136 + }, + { + "epoch": 0.4143510487304842, + "grad_norm": 1.5682148933410645, + "learning_rate": 1.320381233894358e-05, + "loss": 0.4981, + "step": 13137 + }, + { + "epoch": 0.4143825894969248, + "grad_norm": 1.8465485572814941, + "learning_rate": 1.3202844613323657e-05, + "loss": 0.4918, + "step": 13138 + }, + { + "epoch": 0.4144141302633654, + "grad_norm": 2.0021119117736816, + "learning_rate": 1.3201876854279566e-05, + "loss": 0.5044, + "step": 13139 + }, + { + "epoch": 0.414445671029806, + "grad_norm": 1.8670543432235718, + "learning_rate": 1.3200909061821402e-05, + "loss": 0.4898, + "step": 13140 + }, + { + "epoch": 0.41447721179624664, + "grad_norm": 1.8645957708358765, + "learning_rate": 1.3199941235959267e-05, + "loss": 0.5205, + "step": 13141 + }, + { + "epoch": 0.41450875256268727, + "grad_norm": 1.5822473764419556, + "learning_rate": 1.3198973376703256e-05, + "loss": 0.5333, + "step": 13142 + }, + { + "epoch": 0.4145402933291279, + "grad_norm": 2.321247100830078, + "learning_rate": 1.3198005484063481e-05, + "loss": 0.4019, + "step": 13143 + }, + { + "epoch": 0.4145718340955685, + "grad_norm": 1.5270298719406128, + "learning_rate": 1.319703755805003e-05, + "loss": 0.5285, + "step": 13144 + }, + { + "epoch": 0.41460337486200916, + "grad_norm": 2.7001659870147705, + "learning_rate": 1.3196069598673012e-05, + "loss": 0.5025, + "step": 13145 + }, + { + "epoch": 0.4146349156284498, + "grad_norm": 1.6550092697143555, + "learning_rate": 1.3195101605942526e-05, + "loss": 0.5051, + "step": 13146 + }, + { + "epoch": 0.4146664563948904, + "grad_norm": 1.621451497077942, + "learning_rate": 1.3194133579868672e-05, + "loss": 0.5179, + "step": 13147 + }, + { + "epoch": 0.41469799716133104, + "grad_norm": 3.343897819519043, + "learning_rate": 1.3193165520461558e-05, + "loss": 0.5171, + "step": 13148 + }, + { + "epoch": 0.4147295379277716, + "grad_norm": 1.4868754148483276, + "learning_rate": 1.319219742773128e-05, + "loss": 0.5076, + "step": 13149 + }, + { + "epoch": 0.41476107869421225, + "grad_norm": 1.8361574411392212, + "learning_rate": 1.3191229301687943e-05, + "loss": 0.497, + "step": 13150 + }, + { + "epoch": 0.4147926194606529, + "grad_norm": 1.3463724851608276, + "learning_rate": 1.3190261142341649e-05, + "loss": 0.496, + "step": 13151 + }, + { + "epoch": 0.4148241602270935, + "grad_norm": 1.4511548280715942, + "learning_rate": 1.3189292949702503e-05, + "loss": 0.526, + "step": 13152 + }, + { + "epoch": 0.41485570099353414, + "grad_norm": 1.8367738723754883, + "learning_rate": 1.3188324723780612e-05, + "loss": 0.5197, + "step": 13153 + }, + { + "epoch": 0.41488724175997477, + "grad_norm": 1.4652823209762573, + "learning_rate": 1.3187356464586074e-05, + "loss": 0.5095, + "step": 13154 + }, + { + "epoch": 0.4149187825264154, + "grad_norm": 1.334973692893982, + "learning_rate": 1.3186388172128996e-05, + "loss": 0.4995, + "step": 13155 + }, + { + "epoch": 0.414950323292856, + "grad_norm": 1.389042854309082, + "learning_rate": 1.3185419846419486e-05, + "loss": 0.5107, + "step": 13156 + }, + { + "epoch": 0.41498186405929666, + "grad_norm": 1.3108105659484863, + "learning_rate": 1.3184451487467644e-05, + "loss": 0.526, + "step": 13157 + }, + { + "epoch": 0.4150134048257373, + "grad_norm": 4.076140403747559, + "learning_rate": 1.3183483095283578e-05, + "loss": 0.5056, + "step": 13158 + }, + { + "epoch": 0.4150449455921779, + "grad_norm": 1.5239086151123047, + "learning_rate": 1.3182514669877393e-05, + "loss": 0.5096, + "step": 13159 + }, + { + "epoch": 0.4150764863586185, + "grad_norm": 1.4095419645309448, + "learning_rate": 1.3181546211259199e-05, + "loss": 0.4678, + "step": 13160 + }, + { + "epoch": 0.4151080271250591, + "grad_norm": 1.3283560276031494, + "learning_rate": 1.3180577719439103e-05, + "loss": 0.5454, + "step": 13161 + }, + { + "epoch": 0.41513956789149975, + "grad_norm": 1.5445348024368286, + "learning_rate": 1.3179609194427204e-05, + "loss": 0.5302, + "step": 13162 + }, + { + "epoch": 0.4151711086579404, + "grad_norm": 1.7088640928268433, + "learning_rate": 1.3178640636233615e-05, + "loss": 0.5002, + "step": 13163 + }, + { + "epoch": 0.415202649424381, + "grad_norm": 1.527416706085205, + "learning_rate": 1.317767204486844e-05, + "loss": 0.498, + "step": 13164 + }, + { + "epoch": 0.41523419019082164, + "grad_norm": 1.5181323289871216, + "learning_rate": 1.3176703420341794e-05, + "loss": 0.4668, + "step": 13165 + }, + { + "epoch": 0.41526573095726227, + "grad_norm": 1.595693588256836, + "learning_rate": 1.317573476266378e-05, + "loss": 0.4753, + "step": 13166 + }, + { + "epoch": 0.4152972717237029, + "grad_norm": 1.7429461479187012, + "learning_rate": 1.317476607184451e-05, + "loss": 0.519, + "step": 13167 + }, + { + "epoch": 0.41532881249014353, + "grad_norm": 1.4455925226211548, + "learning_rate": 1.3173797347894086e-05, + "loss": 0.4752, + "step": 13168 + }, + { + "epoch": 0.41536035325658416, + "grad_norm": 1.7601662874221802, + "learning_rate": 1.3172828590822623e-05, + "loss": 0.5394, + "step": 13169 + }, + { + "epoch": 0.41539189402302473, + "grad_norm": 1.7952485084533691, + "learning_rate": 1.3171859800640234e-05, + "loss": 0.504, + "step": 13170 + }, + { + "epoch": 0.41542343478946536, + "grad_norm": 1.575777530670166, + "learning_rate": 1.3170890977357025e-05, + "loss": 0.5128, + "step": 13171 + }, + { + "epoch": 0.415454975555906, + "grad_norm": 1.534435510635376, + "learning_rate": 1.3169922120983103e-05, + "loss": 0.5066, + "step": 13172 + }, + { + "epoch": 0.4154865163223466, + "grad_norm": 1.8548611402511597, + "learning_rate": 1.3168953231528586e-05, + "loss": 0.4799, + "step": 13173 + }, + { + "epoch": 0.41551805708878725, + "grad_norm": 1.67209792137146, + "learning_rate": 1.3167984309003577e-05, + "loss": 0.4778, + "step": 13174 + }, + { + "epoch": 0.4155495978552279, + "grad_norm": 1.8513518571853638, + "learning_rate": 1.3167015353418192e-05, + "loss": 0.5323, + "step": 13175 + }, + { + "epoch": 0.4155811386216685, + "grad_norm": 1.4239997863769531, + "learning_rate": 1.3166046364782545e-05, + "loss": 0.4825, + "step": 13176 + }, + { + "epoch": 0.41561267938810914, + "grad_norm": 1.4673675298690796, + "learning_rate": 1.3165077343106747e-05, + "loss": 0.5012, + "step": 13177 + }, + { + "epoch": 0.41564422015454977, + "grad_norm": 1.4844164848327637, + "learning_rate": 1.3164108288400907e-05, + "loss": 0.5071, + "step": 13178 + }, + { + "epoch": 0.4156757609209904, + "grad_norm": 1.3857487440109253, + "learning_rate": 1.3163139200675139e-05, + "loss": 0.4425, + "step": 13179 + }, + { + "epoch": 0.41570730168743103, + "grad_norm": 1.8934792280197144, + "learning_rate": 1.3162170079939563e-05, + "loss": 0.5197, + "step": 13180 + }, + { + "epoch": 0.4157388424538716, + "grad_norm": 1.6283185482025146, + "learning_rate": 1.3161200926204282e-05, + "loss": 0.496, + "step": 13181 + }, + { + "epoch": 0.41577038322031223, + "grad_norm": 1.3824377059936523, + "learning_rate": 1.3160231739479414e-05, + "loss": 0.4637, + "step": 13182 + }, + { + "epoch": 0.41580192398675286, + "grad_norm": 1.5490132570266724, + "learning_rate": 1.3159262519775075e-05, + "loss": 0.5039, + "step": 13183 + }, + { + "epoch": 0.4158334647531935, + "grad_norm": 1.18361234664917, + "learning_rate": 1.315829326710138e-05, + "loss": 0.4953, + "step": 13184 + }, + { + "epoch": 0.4158650055196341, + "grad_norm": 2.081852674484253, + "learning_rate": 1.3157323981468441e-05, + "loss": 0.5043, + "step": 13185 + }, + { + "epoch": 0.41589654628607475, + "grad_norm": 2.7037811279296875, + "learning_rate": 1.3156354662886378e-05, + "loss": 0.8362, + "step": 13186 + }, + { + "epoch": 0.4159280870525154, + "grad_norm": 1.8933465480804443, + "learning_rate": 1.31553853113653e-05, + "loss": 0.4824, + "step": 13187 + }, + { + "epoch": 0.415959627818956, + "grad_norm": 1.822714924812317, + "learning_rate": 1.3154415926915327e-05, + "loss": 0.4851, + "step": 13188 + }, + { + "epoch": 0.41599116858539664, + "grad_norm": 1.9027992486953735, + "learning_rate": 1.3153446509546572e-05, + "loss": 0.518, + "step": 13189 + }, + { + "epoch": 0.41602270935183727, + "grad_norm": 1.59364914894104, + "learning_rate": 1.3152477059269157e-05, + "loss": 0.5127, + "step": 13190 + }, + { + "epoch": 0.4160542501182779, + "grad_norm": 1.5705815553665161, + "learning_rate": 1.3151507576093198e-05, + "loss": 0.4682, + "step": 13191 + }, + { + "epoch": 0.4160857908847185, + "grad_norm": 1.4426127672195435, + "learning_rate": 1.3150538060028808e-05, + "loss": 0.524, + "step": 13192 + }, + { + "epoch": 0.4161173316511591, + "grad_norm": 1.5226870775222778, + "learning_rate": 1.3149568511086104e-05, + "loss": 0.4982, + "step": 13193 + }, + { + "epoch": 0.41614887241759974, + "grad_norm": 1.6587291955947876, + "learning_rate": 1.314859892927521e-05, + "loss": 0.4971, + "step": 13194 + }, + { + "epoch": 0.41618041318404037, + "grad_norm": 1.706250548362732, + "learning_rate": 1.3147629314606243e-05, + "loss": 0.5322, + "step": 13195 + }, + { + "epoch": 0.416211953950481, + "grad_norm": 1.3707013130187988, + "learning_rate": 1.3146659667089317e-05, + "loss": 0.483, + "step": 13196 + }, + { + "epoch": 0.4162434947169216, + "grad_norm": 1.627525806427002, + "learning_rate": 1.3145689986734556e-05, + "loss": 0.5349, + "step": 13197 + }, + { + "epoch": 0.41627503548336225, + "grad_norm": 1.441316843032837, + "learning_rate": 1.3144720273552075e-05, + "loss": 0.5044, + "step": 13198 + }, + { + "epoch": 0.4163065762498029, + "grad_norm": 1.5537461042404175, + "learning_rate": 1.3143750527551997e-05, + "loss": 0.5161, + "step": 13199 + }, + { + "epoch": 0.4163381170162435, + "grad_norm": 1.60037100315094, + "learning_rate": 1.3142780748744444e-05, + "loss": 0.4761, + "step": 13200 + }, + { + "epoch": 0.41636965778268414, + "grad_norm": 1.617588758468628, + "learning_rate": 1.3141810937139532e-05, + "loss": 0.5116, + "step": 13201 + }, + { + "epoch": 0.4164011985491247, + "grad_norm": 1.9745241403579712, + "learning_rate": 1.3140841092747382e-05, + "loss": 0.4597, + "step": 13202 + }, + { + "epoch": 0.41643273931556535, + "grad_norm": 1.6053225994110107, + "learning_rate": 1.3139871215578117e-05, + "loss": 0.4855, + "step": 13203 + }, + { + "epoch": 0.416464280082006, + "grad_norm": 1.4346966743469238, + "learning_rate": 1.3138901305641861e-05, + "loss": 0.5098, + "step": 13204 + }, + { + "epoch": 0.4164958208484466, + "grad_norm": 1.4621590375900269, + "learning_rate": 1.3137931362948725e-05, + "loss": 0.5048, + "step": 13205 + }, + { + "epoch": 0.41652736161488724, + "grad_norm": 2.0073835849761963, + "learning_rate": 1.3136961387508844e-05, + "loss": 0.4923, + "step": 13206 + }, + { + "epoch": 0.41655890238132787, + "grad_norm": 2.3471426963806152, + "learning_rate": 1.3135991379332333e-05, + "loss": 0.8294, + "step": 13207 + }, + { + "epoch": 0.4165904431477685, + "grad_norm": 1.497350811958313, + "learning_rate": 1.3135021338429322e-05, + "loss": 0.4969, + "step": 13208 + }, + { + "epoch": 0.4166219839142091, + "grad_norm": 1.5959261655807495, + "learning_rate": 1.3134051264809924e-05, + "loss": 0.5333, + "step": 13209 + }, + { + "epoch": 0.41665352468064976, + "grad_norm": 1.4118465185165405, + "learning_rate": 1.3133081158484268e-05, + "loss": 0.4598, + "step": 13210 + }, + { + "epoch": 0.4166850654470904, + "grad_norm": 2.2430834770202637, + "learning_rate": 1.3132111019462482e-05, + "loss": 0.5033, + "step": 13211 + }, + { + "epoch": 0.416716606213531, + "grad_norm": 1.591086745262146, + "learning_rate": 1.313114084775468e-05, + "loss": 0.505, + "step": 13212 + }, + { + "epoch": 0.4167481469799716, + "grad_norm": 1.5051130056381226, + "learning_rate": 1.3130170643370992e-05, + "loss": 0.51, + "step": 13213 + }, + { + "epoch": 0.4167796877464122, + "grad_norm": 1.6590197086334229, + "learning_rate": 1.3129200406321545e-05, + "loss": 0.4943, + "step": 13214 + }, + { + "epoch": 0.41681122851285285, + "grad_norm": 1.4575636386871338, + "learning_rate": 1.3128230136616463e-05, + "loss": 0.4844, + "step": 13215 + }, + { + "epoch": 0.4168427692792935, + "grad_norm": 1.5424398183822632, + "learning_rate": 1.3127259834265869e-05, + "loss": 0.8199, + "step": 13216 + }, + { + "epoch": 0.4168743100457341, + "grad_norm": 1.79825758934021, + "learning_rate": 1.3126289499279888e-05, + "loss": 0.5385, + "step": 13217 + }, + { + "epoch": 0.41690585081217474, + "grad_norm": 1.6311416625976562, + "learning_rate": 1.3125319131668652e-05, + "loss": 0.4696, + "step": 13218 + }, + { + "epoch": 0.41693739157861537, + "grad_norm": 2.0061962604522705, + "learning_rate": 1.3124348731442283e-05, + "loss": 0.485, + "step": 13219 + }, + { + "epoch": 0.416968932345056, + "grad_norm": 1.5123851299285889, + "learning_rate": 1.312337829861091e-05, + "loss": 0.5236, + "step": 13220 + }, + { + "epoch": 0.41700047311149663, + "grad_norm": 1.5282129049301147, + "learning_rate": 1.312240783318466e-05, + "loss": 0.3819, + "step": 13221 + }, + { + "epoch": 0.41703201387793726, + "grad_norm": 1.626207709312439, + "learning_rate": 1.3121437335173657e-05, + "loss": 0.5158, + "step": 13222 + }, + { + "epoch": 0.41706355464437783, + "grad_norm": 1.4121644496917725, + "learning_rate": 1.3120466804588033e-05, + "loss": 0.5296, + "step": 13223 + }, + { + "epoch": 0.41709509541081846, + "grad_norm": 1.3582905530929565, + "learning_rate": 1.3119496241437912e-05, + "loss": 0.4981, + "step": 13224 + }, + { + "epoch": 0.4171266361772591, + "grad_norm": 2.011718988418579, + "learning_rate": 1.3118525645733431e-05, + "loss": 0.503, + "step": 13225 + }, + { + "epoch": 0.4171581769436997, + "grad_norm": 1.598141074180603, + "learning_rate": 1.3117555017484708e-05, + "loss": 0.4916, + "step": 13226 + }, + { + "epoch": 0.41718971771014035, + "grad_norm": 1.712206244468689, + "learning_rate": 1.311658435670188e-05, + "loss": 0.8472, + "step": 13227 + }, + { + "epoch": 0.417221258476581, + "grad_norm": 2.3583431243896484, + "learning_rate": 1.3115613663395077e-05, + "loss": 0.4521, + "step": 13228 + }, + { + "epoch": 0.4172527992430216, + "grad_norm": 2.0047085285186768, + "learning_rate": 1.3114642937574423e-05, + "loss": 0.4787, + "step": 13229 + }, + { + "epoch": 0.41728434000946224, + "grad_norm": 1.6751526594161987, + "learning_rate": 1.3113672179250054e-05, + "loss": 0.4848, + "step": 13230 + }, + { + "epoch": 0.41731588077590287, + "grad_norm": 2.0389881134033203, + "learning_rate": 1.3112701388432095e-05, + "loss": 0.3866, + "step": 13231 + }, + { + "epoch": 0.4173474215423435, + "grad_norm": 1.697341799736023, + "learning_rate": 1.3111730565130683e-05, + "loss": 0.4236, + "step": 13232 + }, + { + "epoch": 0.41737896230878413, + "grad_norm": 1.7855168581008911, + "learning_rate": 1.3110759709355942e-05, + "loss": 0.4906, + "step": 13233 + }, + { + "epoch": 0.4174105030752247, + "grad_norm": 1.2765064239501953, + "learning_rate": 1.310978882111801e-05, + "loss": 0.4845, + "step": 13234 + }, + { + "epoch": 0.41744204384166533, + "grad_norm": 1.4727319478988647, + "learning_rate": 1.310881790042702e-05, + "loss": 0.5041, + "step": 13235 + }, + { + "epoch": 0.41747358460810596, + "grad_norm": 1.8561534881591797, + "learning_rate": 1.3107846947293099e-05, + "loss": 0.3521, + "step": 13236 + }, + { + "epoch": 0.4175051253745466, + "grad_norm": 2.0842931270599365, + "learning_rate": 1.310687596172638e-05, + "loss": 0.5347, + "step": 13237 + }, + { + "epoch": 0.4175366661409872, + "grad_norm": 1.5109626054763794, + "learning_rate": 1.3105904943737e-05, + "loss": 0.5135, + "step": 13238 + }, + { + "epoch": 0.41756820690742785, + "grad_norm": 1.5535179376602173, + "learning_rate": 1.3104933893335093e-05, + "loss": 0.8362, + "step": 13239 + }, + { + "epoch": 0.4175997476738685, + "grad_norm": 1.8588478565216064, + "learning_rate": 1.3103962810530785e-05, + "loss": 0.5006, + "step": 13240 + }, + { + "epoch": 0.4176312884403091, + "grad_norm": 1.4388927221298218, + "learning_rate": 1.3102991695334217e-05, + "loss": 0.7939, + "step": 13241 + }, + { + "epoch": 0.41766282920674974, + "grad_norm": 1.5996365547180176, + "learning_rate": 1.3102020547755523e-05, + "loss": 0.4994, + "step": 13242 + }, + { + "epoch": 0.41769436997319037, + "grad_norm": 1.5937997102737427, + "learning_rate": 1.3101049367804833e-05, + "loss": 0.3882, + "step": 13243 + }, + { + "epoch": 0.41772591073963095, + "grad_norm": 1.5808165073394775, + "learning_rate": 1.3100078155492284e-05, + "loss": 0.4927, + "step": 13244 + }, + { + "epoch": 0.4177574515060716, + "grad_norm": 1.4599542617797852, + "learning_rate": 1.3099106910828017e-05, + "loss": 0.4931, + "step": 13245 + }, + { + "epoch": 0.4177889922725122, + "grad_norm": 1.6153030395507812, + "learning_rate": 1.309813563382216e-05, + "loss": 0.5279, + "step": 13246 + }, + { + "epoch": 0.41782053303895283, + "grad_norm": 1.4113004207611084, + "learning_rate": 1.309716432448485e-05, + "loss": 0.5052, + "step": 13247 + }, + { + "epoch": 0.41785207380539346, + "grad_norm": 1.250938892364502, + "learning_rate": 1.3096192982826228e-05, + "loss": 0.4961, + "step": 13248 + }, + { + "epoch": 0.4178836145718341, + "grad_norm": 1.3739277124404907, + "learning_rate": 1.3095221608856429e-05, + "loss": 0.4778, + "step": 13249 + }, + { + "epoch": 0.4179151553382747, + "grad_norm": 2.396892547607422, + "learning_rate": 1.309425020258559e-05, + "loss": 0.4675, + "step": 13250 + }, + { + "epoch": 0.41794669610471535, + "grad_norm": 2.2505242824554443, + "learning_rate": 1.3093278764023842e-05, + "loss": 0.5283, + "step": 13251 + }, + { + "epoch": 0.417978236871156, + "grad_norm": 2.329030752182007, + "learning_rate": 1.3092307293181335e-05, + "loss": 0.4838, + "step": 13252 + }, + { + "epoch": 0.4180097776375966, + "grad_norm": 1.3168011903762817, + "learning_rate": 1.3091335790068196e-05, + "loss": 0.4926, + "step": 13253 + }, + { + "epoch": 0.41804131840403724, + "grad_norm": 1.3242872953414917, + "learning_rate": 1.3090364254694568e-05, + "loss": 0.4899, + "step": 13254 + }, + { + "epoch": 0.4180728591704778, + "grad_norm": 1.2647056579589844, + "learning_rate": 1.308939268707059e-05, + "loss": 0.502, + "step": 13255 + }, + { + "epoch": 0.41810439993691845, + "grad_norm": 1.9351023435592651, + "learning_rate": 1.30884210872064e-05, + "loss": 0.5069, + "step": 13256 + }, + { + "epoch": 0.4181359407033591, + "grad_norm": 1.4189833402633667, + "learning_rate": 1.3087449455112138e-05, + "loss": 0.4655, + "step": 13257 + }, + { + "epoch": 0.4181674814697997, + "grad_norm": 2.770920991897583, + "learning_rate": 1.3086477790797942e-05, + "loss": 0.491, + "step": 13258 + }, + { + "epoch": 0.41819902223624034, + "grad_norm": 1.457322359085083, + "learning_rate": 1.3085506094273959e-05, + "loss": 0.4811, + "step": 13259 + }, + { + "epoch": 0.41823056300268097, + "grad_norm": 1.5249360799789429, + "learning_rate": 1.308453436555032e-05, + "loss": 0.5162, + "step": 13260 + }, + { + "epoch": 0.4182621037691216, + "grad_norm": 1.9265046119689941, + "learning_rate": 1.308356260463717e-05, + "loss": 0.4475, + "step": 13261 + }, + { + "epoch": 0.4182936445355622, + "grad_norm": 1.401181697845459, + "learning_rate": 1.308259081154465e-05, + "loss": 0.4345, + "step": 13262 + }, + { + "epoch": 0.41832518530200286, + "grad_norm": 1.6342755556106567, + "learning_rate": 1.3081618986282902e-05, + "loss": 0.5301, + "step": 13263 + }, + { + "epoch": 0.4183567260684435, + "grad_norm": 1.3510416746139526, + "learning_rate": 1.3080647128862067e-05, + "loss": 0.5117, + "step": 13264 + }, + { + "epoch": 0.4183882668348841, + "grad_norm": 1.9356757402420044, + "learning_rate": 1.3079675239292287e-05, + "loss": 0.5191, + "step": 13265 + }, + { + "epoch": 0.4184198076013247, + "grad_norm": 1.3725433349609375, + "learning_rate": 1.3078703317583705e-05, + "loss": 0.4849, + "step": 13266 + }, + { + "epoch": 0.4184513483677653, + "grad_norm": 1.2869579792022705, + "learning_rate": 1.3077731363746463e-05, + "loss": 0.4874, + "step": 13267 + }, + { + "epoch": 0.41848288913420595, + "grad_norm": 1.434579849243164, + "learning_rate": 1.3076759377790702e-05, + "loss": 0.5095, + "step": 13268 + }, + { + "epoch": 0.4185144299006466, + "grad_norm": 1.4699760675430298, + "learning_rate": 1.307578735972657e-05, + "loss": 0.4818, + "step": 13269 + }, + { + "epoch": 0.4185459706670872, + "grad_norm": 2.9797191619873047, + "learning_rate": 1.3074815309564213e-05, + "loss": 0.4963, + "step": 13270 + }, + { + "epoch": 0.41857751143352784, + "grad_norm": 2.452575445175171, + "learning_rate": 1.3073843227313764e-05, + "loss": 0.5296, + "step": 13271 + }, + { + "epoch": 0.41860905219996847, + "grad_norm": 1.5337350368499756, + "learning_rate": 1.3072871112985376e-05, + "loss": 0.4868, + "step": 13272 + }, + { + "epoch": 0.4186405929664091, + "grad_norm": 1.648165225982666, + "learning_rate": 1.3071898966589194e-05, + "loss": 0.4811, + "step": 13273 + }, + { + "epoch": 0.4186721337328497, + "grad_norm": 1.4522784948349, + "learning_rate": 1.3070926788135358e-05, + "loss": 0.4945, + "step": 13274 + }, + { + "epoch": 0.41870367449929036, + "grad_norm": 1.4180384874343872, + "learning_rate": 1.3069954577634016e-05, + "loss": 0.4692, + "step": 13275 + }, + { + "epoch": 0.41873521526573093, + "grad_norm": 1.87205970287323, + "learning_rate": 1.306898233509532e-05, + "loss": 0.4998, + "step": 13276 + }, + { + "epoch": 0.41876675603217156, + "grad_norm": 1.4395451545715332, + "learning_rate": 1.3068010060529406e-05, + "loss": 0.5085, + "step": 13277 + }, + { + "epoch": 0.4187982967986122, + "grad_norm": 1.4164702892303467, + "learning_rate": 1.3067037753946423e-05, + "loss": 0.4599, + "step": 13278 + }, + { + "epoch": 0.4188298375650528, + "grad_norm": 1.573476791381836, + "learning_rate": 1.3066065415356524e-05, + "loss": 0.4926, + "step": 13279 + }, + { + "epoch": 0.41886137833149345, + "grad_norm": 1.3766838312149048, + "learning_rate": 1.3065093044769849e-05, + "loss": 0.5301, + "step": 13280 + }, + { + "epoch": 0.4188929190979341, + "grad_norm": 1.5875440835952759, + "learning_rate": 1.3064120642196549e-05, + "loss": 0.5204, + "step": 13281 + }, + { + "epoch": 0.4189244598643747, + "grad_norm": 1.8688759803771973, + "learning_rate": 1.306314820764677e-05, + "loss": 0.5113, + "step": 13282 + }, + { + "epoch": 0.41895600063081534, + "grad_norm": 1.3727282285690308, + "learning_rate": 1.3062175741130663e-05, + "loss": 0.4684, + "step": 13283 + }, + { + "epoch": 0.41898754139725597, + "grad_norm": 2.7350809574127197, + "learning_rate": 1.3061203242658371e-05, + "loss": 0.4983, + "step": 13284 + }, + { + "epoch": 0.4190190821636966, + "grad_norm": 1.7962923049926758, + "learning_rate": 1.3060230712240047e-05, + "loss": 0.4831, + "step": 13285 + }, + { + "epoch": 0.41905062293013723, + "grad_norm": 1.438970685005188, + "learning_rate": 1.305925814988584e-05, + "loss": 0.4986, + "step": 13286 + }, + { + "epoch": 0.4190821636965778, + "grad_norm": 2.6454310417175293, + "learning_rate": 1.30582855556059e-05, + "loss": 0.5323, + "step": 13287 + }, + { + "epoch": 0.41911370446301843, + "grad_norm": 2.331099510192871, + "learning_rate": 1.3057312929410375e-05, + "loss": 0.846, + "step": 13288 + }, + { + "epoch": 0.41914524522945906, + "grad_norm": 1.6392918825149536, + "learning_rate": 1.3056340271309415e-05, + "loss": 0.5058, + "step": 13289 + }, + { + "epoch": 0.4191767859958997, + "grad_norm": 1.3685097694396973, + "learning_rate": 1.3055367581313174e-05, + "loss": 0.4554, + "step": 13290 + }, + { + "epoch": 0.4192083267623403, + "grad_norm": 2.124450206756592, + "learning_rate": 1.3054394859431795e-05, + "loss": 0.7852, + "step": 13291 + }, + { + "epoch": 0.41923986752878095, + "grad_norm": 1.4140959978103638, + "learning_rate": 1.3053422105675435e-05, + "loss": 0.4891, + "step": 13292 + }, + { + "epoch": 0.4192714082952216, + "grad_norm": 1.6817407608032227, + "learning_rate": 1.3052449320054248e-05, + "loss": 0.5037, + "step": 13293 + }, + { + "epoch": 0.4193029490616622, + "grad_norm": 1.4298255443572998, + "learning_rate": 1.305147650257838e-05, + "loss": 0.49, + "step": 13294 + }, + { + "epoch": 0.41933448982810284, + "grad_norm": 1.3701043128967285, + "learning_rate": 1.3050503653257982e-05, + "loss": 0.4877, + "step": 13295 + }, + { + "epoch": 0.41936603059454347, + "grad_norm": 1.5231292247772217, + "learning_rate": 1.3049530772103213e-05, + "loss": 0.5233, + "step": 13296 + }, + { + "epoch": 0.41939757136098404, + "grad_norm": 1.5709800720214844, + "learning_rate": 1.3048557859124222e-05, + "loss": 0.4928, + "step": 13297 + }, + { + "epoch": 0.4194291121274247, + "grad_norm": 1.667075514793396, + "learning_rate": 1.3047584914331163e-05, + "loss": 0.5299, + "step": 13298 + }, + { + "epoch": 0.4194606528938653, + "grad_norm": 2.054583787918091, + "learning_rate": 1.304661193773419e-05, + "loss": 0.5081, + "step": 13299 + }, + { + "epoch": 0.41949219366030593, + "grad_norm": 1.7103320360183716, + "learning_rate": 1.3045638929343457e-05, + "loss": 0.4816, + "step": 13300 + }, + { + "epoch": 0.41952373442674656, + "grad_norm": 1.6167851686477661, + "learning_rate": 1.3044665889169114e-05, + "loss": 0.5014, + "step": 13301 + }, + { + "epoch": 0.4195552751931872, + "grad_norm": 1.6311253309249878, + "learning_rate": 1.3043692817221318e-05, + "loss": 0.508, + "step": 13302 + }, + { + "epoch": 0.4195868159596278, + "grad_norm": 1.4200392961502075, + "learning_rate": 1.3042719713510224e-05, + "loss": 0.5103, + "step": 13303 + }, + { + "epoch": 0.41961835672606845, + "grad_norm": 2.842794418334961, + "learning_rate": 1.304174657804599e-05, + "loss": 0.3825, + "step": 13304 + }, + { + "epoch": 0.4196498974925091, + "grad_norm": 3.1871397495269775, + "learning_rate": 1.3040773410838766e-05, + "loss": 0.5129, + "step": 13305 + }, + { + "epoch": 0.4196814382589497, + "grad_norm": 1.7852561473846436, + "learning_rate": 1.303980021189871e-05, + "loss": 0.5167, + "step": 13306 + }, + { + "epoch": 0.41971297902539034, + "grad_norm": 2.5933151245117188, + "learning_rate": 1.3038826981235979e-05, + "loss": 0.3943, + "step": 13307 + }, + { + "epoch": 0.4197445197918309, + "grad_norm": 1.8764395713806152, + "learning_rate": 1.303785371886073e-05, + "loss": 0.4798, + "step": 13308 + }, + { + "epoch": 0.41977606055827155, + "grad_norm": 2.1822454929351807, + "learning_rate": 1.3036880424783119e-05, + "loss": 0.7881, + "step": 13309 + }, + { + "epoch": 0.4198076013247122, + "grad_norm": 1.991165280342102, + "learning_rate": 1.30359070990133e-05, + "loss": 0.8267, + "step": 13310 + }, + { + "epoch": 0.4198391420911528, + "grad_norm": 1.626603364944458, + "learning_rate": 1.3034933741561438e-05, + "loss": 0.7916, + "step": 13311 + }, + { + "epoch": 0.41987068285759344, + "grad_norm": 1.5766135454177856, + "learning_rate": 1.3033960352437682e-05, + "loss": 0.7628, + "step": 13312 + }, + { + "epoch": 0.41990222362403407, + "grad_norm": 1.5889077186584473, + "learning_rate": 1.3032986931652195e-05, + "loss": 0.5012, + "step": 13313 + }, + { + "epoch": 0.4199337643904747, + "grad_norm": 1.4048532247543335, + "learning_rate": 1.3032013479215133e-05, + "loss": 0.487, + "step": 13314 + }, + { + "epoch": 0.4199653051569153, + "grad_norm": 1.7606842517852783, + "learning_rate": 1.3031039995136656e-05, + "loss": 0.5096, + "step": 13315 + }, + { + "epoch": 0.41999684592335595, + "grad_norm": 2.661470651626587, + "learning_rate": 1.3030066479426924e-05, + "loss": 0.8096, + "step": 13316 + }, + { + "epoch": 0.4200283866897966, + "grad_norm": 2.0518782138824463, + "learning_rate": 1.3029092932096097e-05, + "loss": 0.4778, + "step": 13317 + }, + { + "epoch": 0.4200599274562372, + "grad_norm": 1.4563056230545044, + "learning_rate": 1.302811935315433e-05, + "loss": 0.5173, + "step": 13318 + }, + { + "epoch": 0.4200914682226778, + "grad_norm": 1.3956059217453003, + "learning_rate": 1.302714574261179e-05, + "loss": 0.3357, + "step": 13319 + }, + { + "epoch": 0.4201230089891184, + "grad_norm": 1.3792476654052734, + "learning_rate": 1.3026172100478632e-05, + "loss": 0.518, + "step": 13320 + }, + { + "epoch": 0.42015454975555905, + "grad_norm": 1.423762321472168, + "learning_rate": 1.3025198426765018e-05, + "loss": 0.4826, + "step": 13321 + }, + { + "epoch": 0.4201860905219997, + "grad_norm": 2.0884504318237305, + "learning_rate": 1.302422472148111e-05, + "loss": 0.8396, + "step": 13322 + }, + { + "epoch": 0.4202176312884403, + "grad_norm": 1.3800684213638306, + "learning_rate": 1.3023250984637066e-05, + "loss": 0.4879, + "step": 13323 + }, + { + "epoch": 0.42024917205488094, + "grad_norm": 1.342325210571289, + "learning_rate": 1.3022277216243055e-05, + "loss": 0.51, + "step": 13324 + }, + { + "epoch": 0.42028071282132157, + "grad_norm": 3.203233003616333, + "learning_rate": 1.302130341630923e-05, + "loss": 0.4098, + "step": 13325 + }, + { + "epoch": 0.4203122535877622, + "grad_norm": 1.5392228364944458, + "learning_rate": 1.302032958484576e-05, + "loss": 0.8125, + "step": 13326 + }, + { + "epoch": 0.4203437943542028, + "grad_norm": 1.30855131149292, + "learning_rate": 1.3019355721862807e-05, + "loss": 0.5345, + "step": 13327 + }, + { + "epoch": 0.42037533512064346, + "grad_norm": 1.559390902519226, + "learning_rate": 1.301838182737053e-05, + "loss": 0.4895, + "step": 13328 + }, + { + "epoch": 0.42040687588708403, + "grad_norm": 1.880029559135437, + "learning_rate": 1.3017407901379097e-05, + "loss": 0.8167, + "step": 13329 + }, + { + "epoch": 0.42043841665352466, + "grad_norm": 1.702408790588379, + "learning_rate": 1.301643394389867e-05, + "loss": 0.4883, + "step": 13330 + }, + { + "epoch": 0.4204699574199653, + "grad_norm": 1.8975471258163452, + "learning_rate": 1.301545995493941e-05, + "loss": 0.8322, + "step": 13331 + }, + { + "epoch": 0.4205014981864059, + "grad_norm": 1.5375014543533325, + "learning_rate": 1.3014485934511486e-05, + "loss": 0.521, + "step": 13332 + }, + { + "epoch": 0.42053303895284655, + "grad_norm": 1.806394338607788, + "learning_rate": 1.3013511882625058e-05, + "loss": 0.5017, + "step": 13333 + }, + { + "epoch": 0.4205645797192872, + "grad_norm": 1.5132390260696411, + "learning_rate": 1.3012537799290295e-05, + "loss": 0.4794, + "step": 13334 + }, + { + "epoch": 0.4205961204857278, + "grad_norm": 1.5847909450531006, + "learning_rate": 1.3011563684517364e-05, + "loss": 0.5118, + "step": 13335 + }, + { + "epoch": 0.42062766125216844, + "grad_norm": 1.5475378036499023, + "learning_rate": 1.3010589538316423e-05, + "loss": 0.488, + "step": 13336 + }, + { + "epoch": 0.42065920201860907, + "grad_norm": 1.8740795850753784, + "learning_rate": 1.3009615360697647e-05, + "loss": 0.4845, + "step": 13337 + }, + { + "epoch": 0.4206907427850497, + "grad_norm": 1.6536728143692017, + "learning_rate": 1.3008641151671193e-05, + "loss": 0.4806, + "step": 13338 + }, + { + "epoch": 0.42072228355149033, + "grad_norm": 1.7562050819396973, + "learning_rate": 1.3007666911247236e-05, + "loss": 0.8139, + "step": 13339 + }, + { + "epoch": 0.4207538243179309, + "grad_norm": 1.4495652914047241, + "learning_rate": 1.3006692639435938e-05, + "loss": 0.4964, + "step": 13340 + }, + { + "epoch": 0.42078536508437153, + "grad_norm": 2.3941102027893066, + "learning_rate": 1.3005718336247466e-05, + "loss": 0.4866, + "step": 13341 + }, + { + "epoch": 0.42081690585081216, + "grad_norm": 1.4262019395828247, + "learning_rate": 1.3004744001691994e-05, + "loss": 0.5006, + "step": 13342 + }, + { + "epoch": 0.4208484466172528, + "grad_norm": 1.8949267864227295, + "learning_rate": 1.300376963577968e-05, + "loss": 0.8054, + "step": 13343 + }, + { + "epoch": 0.4208799873836934, + "grad_norm": 1.4184130430221558, + "learning_rate": 1.30027952385207e-05, + "loss": 0.8043, + "step": 13344 + }, + { + "epoch": 0.42091152815013405, + "grad_norm": 1.4758832454681396, + "learning_rate": 1.3001820809925221e-05, + "loss": 0.473, + "step": 13345 + }, + { + "epoch": 0.4209430689165747, + "grad_norm": 1.5653841495513916, + "learning_rate": 1.300084635000341e-05, + "loss": 0.4941, + "step": 13346 + }, + { + "epoch": 0.4209746096830153, + "grad_norm": 1.9728715419769287, + "learning_rate": 1.2999871858765436e-05, + "loss": 0.5205, + "step": 13347 + }, + { + "epoch": 0.42100615044945594, + "grad_norm": 1.8413162231445312, + "learning_rate": 1.299889733622147e-05, + "loss": 0.4938, + "step": 13348 + }, + { + "epoch": 0.42103769121589657, + "grad_norm": 1.5455121994018555, + "learning_rate": 1.2997922782381685e-05, + "loss": 0.5073, + "step": 13349 + }, + { + "epoch": 0.42106923198233714, + "grad_norm": 1.384726643562317, + "learning_rate": 1.2996948197256242e-05, + "loss": 0.5379, + "step": 13350 + }, + { + "epoch": 0.4211007727487778, + "grad_norm": 2.018517017364502, + "learning_rate": 1.299597358085532e-05, + "loss": 0.374, + "step": 13351 + }, + { + "epoch": 0.4211323135152184, + "grad_norm": 1.2342499494552612, + "learning_rate": 1.2994998933189091e-05, + "loss": 0.5023, + "step": 13352 + }, + { + "epoch": 0.42116385428165903, + "grad_norm": 1.666539192199707, + "learning_rate": 1.2994024254267719e-05, + "loss": 0.3566, + "step": 13353 + }, + { + "epoch": 0.42119539504809966, + "grad_norm": 2.1785356998443604, + "learning_rate": 1.299304954410138e-05, + "loss": 0.5087, + "step": 13354 + }, + { + "epoch": 0.4212269358145403, + "grad_norm": 1.6833816766738892, + "learning_rate": 1.2992074802700245e-05, + "loss": 0.4991, + "step": 13355 + }, + { + "epoch": 0.4212584765809809, + "grad_norm": 1.6327670812606812, + "learning_rate": 1.2991100030074486e-05, + "loss": 0.5031, + "step": 13356 + }, + { + "epoch": 0.42129001734742155, + "grad_norm": 7.111458778381348, + "learning_rate": 1.2990125226234275e-05, + "loss": 0.3982, + "step": 13357 + }, + { + "epoch": 0.4213215581138622, + "grad_norm": 1.704885721206665, + "learning_rate": 1.2989150391189789e-05, + "loss": 0.4634, + "step": 13358 + }, + { + "epoch": 0.4213530988803028, + "grad_norm": 1.9108463525772095, + "learning_rate": 1.2988175524951194e-05, + "loss": 0.4017, + "step": 13359 + }, + { + "epoch": 0.42138463964674344, + "grad_norm": 1.4838800430297852, + "learning_rate": 1.2987200627528669e-05, + "loss": 0.4852, + "step": 13360 + }, + { + "epoch": 0.421416180413184, + "grad_norm": 2.627521276473999, + "learning_rate": 1.2986225698932383e-05, + "loss": 0.4919, + "step": 13361 + }, + { + "epoch": 0.42144772117962465, + "grad_norm": 1.3021478652954102, + "learning_rate": 1.2985250739172516e-05, + "loss": 0.4733, + "step": 13362 + }, + { + "epoch": 0.4214792619460653, + "grad_norm": 1.8490980863571167, + "learning_rate": 1.2984275748259238e-05, + "loss": 0.5134, + "step": 13363 + }, + { + "epoch": 0.4215108027125059, + "grad_norm": 1.7852351665496826, + "learning_rate": 1.2983300726202725e-05, + "loss": 0.5259, + "step": 13364 + }, + { + "epoch": 0.42154234347894654, + "grad_norm": 1.9631801843643188, + "learning_rate": 1.2982325673013155e-05, + "loss": 0.5116, + "step": 13365 + }, + { + "epoch": 0.42157388424538716, + "grad_norm": 1.9550912380218506, + "learning_rate": 1.29813505887007e-05, + "loss": 0.4724, + "step": 13366 + }, + { + "epoch": 0.4216054250118278, + "grad_norm": 1.429563045501709, + "learning_rate": 1.2980375473275534e-05, + "loss": 0.4993, + "step": 13367 + }, + { + "epoch": 0.4216369657782684, + "grad_norm": 1.9310624599456787, + "learning_rate": 1.297940032674784e-05, + "loss": 0.5062, + "step": 13368 + }, + { + "epoch": 0.42166850654470905, + "grad_norm": 2.1620664596557617, + "learning_rate": 1.2978425149127788e-05, + "loss": 0.546, + "step": 13369 + }, + { + "epoch": 0.4217000473111497, + "grad_norm": 2.3451435565948486, + "learning_rate": 1.2977449940425558e-05, + "loss": 0.8131, + "step": 13370 + }, + { + "epoch": 0.42173158807759026, + "grad_norm": 1.5153841972351074, + "learning_rate": 1.2976474700651323e-05, + "loss": 0.4938, + "step": 13371 + }, + { + "epoch": 0.4217631288440309, + "grad_norm": 1.5144641399383545, + "learning_rate": 1.2975499429815263e-05, + "loss": 0.5034, + "step": 13372 + }, + { + "epoch": 0.4217946696104715, + "grad_norm": 1.406309723854065, + "learning_rate": 1.2974524127927559e-05, + "loss": 0.4704, + "step": 13373 + }, + { + "epoch": 0.42182621037691215, + "grad_norm": 1.8798924684524536, + "learning_rate": 1.2973548794998385e-05, + "loss": 0.5022, + "step": 13374 + }, + { + "epoch": 0.4218577511433528, + "grad_norm": 1.4132697582244873, + "learning_rate": 1.2972573431037917e-05, + "loss": 0.4602, + "step": 13375 + }, + { + "epoch": 0.4218892919097934, + "grad_norm": 1.738473892211914, + "learning_rate": 1.2971598036056343e-05, + "loss": 0.5175, + "step": 13376 + }, + { + "epoch": 0.42192083267623404, + "grad_norm": 1.794472336769104, + "learning_rate": 1.2970622610063836e-05, + "loss": 0.5195, + "step": 13377 + }, + { + "epoch": 0.42195237344267467, + "grad_norm": 2.3548145294189453, + "learning_rate": 1.2969647153070572e-05, + "loss": 0.8559, + "step": 13378 + }, + { + "epoch": 0.4219839142091153, + "grad_norm": 1.5316399335861206, + "learning_rate": 1.2968671665086735e-05, + "loss": 0.5036, + "step": 13379 + }, + { + "epoch": 0.4220154549755559, + "grad_norm": 1.6666089296340942, + "learning_rate": 1.2967696146122506e-05, + "loss": 0.4751, + "step": 13380 + }, + { + "epoch": 0.42204699574199656, + "grad_norm": 1.539650797843933, + "learning_rate": 1.296672059618806e-05, + "loss": 0.5188, + "step": 13381 + }, + { + "epoch": 0.42207853650843713, + "grad_norm": 1.9340659379959106, + "learning_rate": 1.2965745015293583e-05, + "loss": 0.4048, + "step": 13382 + }, + { + "epoch": 0.42211007727487776, + "grad_norm": 3.834907293319702, + "learning_rate": 1.2964769403449254e-05, + "loss": 0.4902, + "step": 13383 + }, + { + "epoch": 0.4221416180413184, + "grad_norm": 1.835353970527649, + "learning_rate": 1.2963793760665253e-05, + "loss": 0.5079, + "step": 13384 + }, + { + "epoch": 0.422173158807759, + "grad_norm": 1.7382899522781372, + "learning_rate": 1.2962818086951762e-05, + "loss": 0.5091, + "step": 13385 + }, + { + "epoch": 0.42220469957419965, + "grad_norm": 1.3617331981658936, + "learning_rate": 1.2961842382318968e-05, + "loss": 0.4841, + "step": 13386 + }, + { + "epoch": 0.4222362403406403, + "grad_norm": 2.0260539054870605, + "learning_rate": 1.2960866646777045e-05, + "loss": 0.5017, + "step": 13387 + }, + { + "epoch": 0.4222677811070809, + "grad_norm": 2.300229549407959, + "learning_rate": 1.295989088033618e-05, + "loss": 0.39, + "step": 13388 + }, + { + "epoch": 0.42229932187352154, + "grad_norm": 2.1703782081604004, + "learning_rate": 1.2958915083006556e-05, + "loss": 0.4694, + "step": 13389 + }, + { + "epoch": 0.42233086263996217, + "grad_norm": 1.5513367652893066, + "learning_rate": 1.2957939254798358e-05, + "loss": 0.5078, + "step": 13390 + }, + { + "epoch": 0.4223624034064028, + "grad_norm": 1.387020468711853, + "learning_rate": 1.2956963395721762e-05, + "loss": 0.5086, + "step": 13391 + }, + { + "epoch": 0.4223939441728434, + "grad_norm": 2.0446295738220215, + "learning_rate": 1.295598750578696e-05, + "loss": 0.4646, + "step": 13392 + }, + { + "epoch": 0.422425484939284, + "grad_norm": 1.5240617990493774, + "learning_rate": 1.2955011585004134e-05, + "loss": 0.4936, + "step": 13393 + }, + { + "epoch": 0.42245702570572463, + "grad_norm": 1.7245932817459106, + "learning_rate": 1.2954035633383466e-05, + "loss": 0.5064, + "step": 13394 + }, + { + "epoch": 0.42248856647216526, + "grad_norm": 5.686163425445557, + "learning_rate": 1.2953059650935141e-05, + "loss": 0.4901, + "step": 13395 + }, + { + "epoch": 0.4225201072386059, + "grad_norm": 1.2027831077575684, + "learning_rate": 1.2952083637669348e-05, + "loss": 0.5089, + "step": 13396 + }, + { + "epoch": 0.4225516480050465, + "grad_norm": 1.5104568004608154, + "learning_rate": 1.295110759359627e-05, + "loss": 0.5029, + "step": 13397 + }, + { + "epoch": 0.42258318877148715, + "grad_norm": 1.3824453353881836, + "learning_rate": 1.2950131518726092e-05, + "loss": 0.4835, + "step": 13398 + }, + { + "epoch": 0.4226147295379278, + "grad_norm": 1.3776273727416992, + "learning_rate": 1.2949155413069001e-05, + "loss": 0.4747, + "step": 13399 + }, + { + "epoch": 0.4226462703043684, + "grad_norm": 1.4233874082565308, + "learning_rate": 1.2948179276635184e-05, + "loss": 0.4839, + "step": 13400 + }, + { + "epoch": 0.42267781107080904, + "grad_norm": 1.4608794450759888, + "learning_rate": 1.2947203109434826e-05, + "loss": 0.4811, + "step": 13401 + }, + { + "epoch": 0.42270935183724967, + "grad_norm": 1.6376135349273682, + "learning_rate": 1.2946226911478115e-05, + "loss": 0.517, + "step": 13402 + }, + { + "epoch": 0.42274089260369024, + "grad_norm": 2.4376275539398193, + "learning_rate": 1.2945250682775238e-05, + "loss": 0.833, + "step": 13403 + }, + { + "epoch": 0.4227724333701309, + "grad_norm": 1.671502947807312, + "learning_rate": 1.2944274423336386e-05, + "loss": 0.5384, + "step": 13404 + }, + { + "epoch": 0.4228039741365715, + "grad_norm": 1.7842811346054077, + "learning_rate": 1.294329813317174e-05, + "loss": 0.4911, + "step": 13405 + }, + { + "epoch": 0.42283551490301213, + "grad_norm": 2.051867961883545, + "learning_rate": 1.2942321812291497e-05, + "loss": 0.4528, + "step": 13406 + }, + { + "epoch": 0.42286705566945276, + "grad_norm": 1.4339226484298706, + "learning_rate": 1.294134546070584e-05, + "loss": 0.505, + "step": 13407 + }, + { + "epoch": 0.4228985964358934, + "grad_norm": 1.294753909111023, + "learning_rate": 1.2940369078424957e-05, + "loss": 0.5159, + "step": 13408 + }, + { + "epoch": 0.422930137202334, + "grad_norm": 1.6173142194747925, + "learning_rate": 1.2939392665459041e-05, + "loss": 0.4773, + "step": 13409 + }, + { + "epoch": 0.42296167796877465, + "grad_norm": 1.7605388164520264, + "learning_rate": 1.2938416221818281e-05, + "loss": 0.5375, + "step": 13410 + }, + { + "epoch": 0.4229932187352153, + "grad_norm": 1.9627087116241455, + "learning_rate": 1.2937439747512865e-05, + "loss": 0.4955, + "step": 13411 + }, + { + "epoch": 0.4230247595016559, + "grad_norm": 2.0679895877838135, + "learning_rate": 1.2936463242552983e-05, + "loss": 0.507, + "step": 13412 + }, + { + "epoch": 0.42305630026809654, + "grad_norm": 1.3001164197921753, + "learning_rate": 1.2935486706948828e-05, + "loss": 0.517, + "step": 13413 + }, + { + "epoch": 0.4230878410345371, + "grad_norm": 3.1789863109588623, + "learning_rate": 1.2934510140710592e-05, + "loss": 0.489, + "step": 13414 + }, + { + "epoch": 0.42311938180097775, + "grad_norm": 3.0693278312683105, + "learning_rate": 1.2933533543848462e-05, + "loss": 0.4821, + "step": 13415 + }, + { + "epoch": 0.4231509225674184, + "grad_norm": 1.6972450017929077, + "learning_rate": 1.2932556916372633e-05, + "loss": 0.8035, + "step": 13416 + }, + { + "epoch": 0.423182463333859, + "grad_norm": 1.7585381269454956, + "learning_rate": 1.2931580258293295e-05, + "loss": 0.4896, + "step": 13417 + }, + { + "epoch": 0.42321400410029963, + "grad_norm": 1.3281121253967285, + "learning_rate": 1.293060356962064e-05, + "loss": 0.4744, + "step": 13418 + }, + { + "epoch": 0.42324554486674026, + "grad_norm": 2.0810580253601074, + "learning_rate": 1.292962685036486e-05, + "loss": 0.5379, + "step": 13419 + }, + { + "epoch": 0.4232770856331809, + "grad_norm": 1.244310736656189, + "learning_rate": 1.2928650100536153e-05, + "loss": 0.4939, + "step": 13420 + }, + { + "epoch": 0.4233086263996215, + "grad_norm": 1.7733783721923828, + "learning_rate": 1.2927673320144704e-05, + "loss": 0.4904, + "step": 13421 + }, + { + "epoch": 0.42334016716606215, + "grad_norm": 1.987066388130188, + "learning_rate": 1.292669650920071e-05, + "loss": 0.3903, + "step": 13422 + }, + { + "epoch": 0.4233717079325028, + "grad_norm": 1.2313547134399414, + "learning_rate": 1.2925719667714369e-05, + "loss": 0.4669, + "step": 13423 + }, + { + "epoch": 0.42340324869894336, + "grad_norm": 1.4299662113189697, + "learning_rate": 1.2924742795695871e-05, + "loss": 0.8353, + "step": 13424 + }, + { + "epoch": 0.423434789465384, + "grad_norm": 1.5265045166015625, + "learning_rate": 1.2923765893155408e-05, + "loss": 0.8039, + "step": 13425 + }, + { + "epoch": 0.4234663302318246, + "grad_norm": 2.257798671722412, + "learning_rate": 1.2922788960103184e-05, + "loss": 0.5157, + "step": 13426 + }, + { + "epoch": 0.42349787099826525, + "grad_norm": 1.6192941665649414, + "learning_rate": 1.2921811996549383e-05, + "loss": 0.839, + "step": 13427 + }, + { + "epoch": 0.4235294117647059, + "grad_norm": 1.6771504878997803, + "learning_rate": 1.2920835002504206e-05, + "loss": 0.8057, + "step": 13428 + }, + { + "epoch": 0.4235609525311465, + "grad_norm": 1.7166523933410645, + "learning_rate": 1.2919857977977846e-05, + "loss": 0.5012, + "step": 13429 + }, + { + "epoch": 0.42359249329758714, + "grad_norm": 1.5995151996612549, + "learning_rate": 1.2918880922980501e-05, + "loss": 0.5356, + "step": 13430 + }, + { + "epoch": 0.42362403406402777, + "grad_norm": 2.3952019214630127, + "learning_rate": 1.291790383752237e-05, + "loss": 0.4999, + "step": 13431 + }, + { + "epoch": 0.4236555748304684, + "grad_norm": 1.388107419013977, + "learning_rate": 1.2916926721613642e-05, + "loss": 0.4983, + "step": 13432 + }, + { + "epoch": 0.423687115596909, + "grad_norm": 1.7134959697723389, + "learning_rate": 1.2915949575264524e-05, + "loss": 0.5116, + "step": 13433 + }, + { + "epoch": 0.42371865636334965, + "grad_norm": 1.9361979961395264, + "learning_rate": 1.2914972398485206e-05, + "loss": 0.8105, + "step": 13434 + }, + { + "epoch": 0.42375019712979023, + "grad_norm": 1.6976521015167236, + "learning_rate": 1.2913995191285886e-05, + "loss": 0.4224, + "step": 13435 + }, + { + "epoch": 0.42378173789623086, + "grad_norm": 1.6831775903701782, + "learning_rate": 1.2913017953676767e-05, + "loss": 0.4815, + "step": 13436 + }, + { + "epoch": 0.4238132786626715, + "grad_norm": 1.503736972808838, + "learning_rate": 1.291204068566804e-05, + "loss": 0.4992, + "step": 13437 + }, + { + "epoch": 0.4238448194291121, + "grad_norm": 1.283753752708435, + "learning_rate": 1.291106338726991e-05, + "loss": 0.4581, + "step": 13438 + }, + { + "epoch": 0.42387636019555275, + "grad_norm": 1.2167088985443115, + "learning_rate": 1.2910086058492572e-05, + "loss": 0.4618, + "step": 13439 + }, + { + "epoch": 0.4239079009619934, + "grad_norm": 1.3672535419464111, + "learning_rate": 1.2909108699346224e-05, + "loss": 0.4898, + "step": 13440 + }, + { + "epoch": 0.423939441728434, + "grad_norm": 1.7800301313400269, + "learning_rate": 1.2908131309841072e-05, + "loss": 0.4062, + "step": 13441 + }, + { + "epoch": 0.42397098249487464, + "grad_norm": 1.340408205986023, + "learning_rate": 1.2907153889987312e-05, + "loss": 0.5109, + "step": 13442 + }, + { + "epoch": 0.42400252326131527, + "grad_norm": 1.657300591468811, + "learning_rate": 1.290617643979514e-05, + "loss": 0.496, + "step": 13443 + }, + { + "epoch": 0.4240340640277559, + "grad_norm": 1.4660542011260986, + "learning_rate": 1.2905198959274763e-05, + "loss": 0.518, + "step": 13444 + }, + { + "epoch": 0.42406560479419647, + "grad_norm": 1.8461337089538574, + "learning_rate": 1.2904221448436376e-05, + "loss": 0.4957, + "step": 13445 + }, + { + "epoch": 0.4240971455606371, + "grad_norm": 1.6831384897232056, + "learning_rate": 1.2903243907290188e-05, + "loss": 0.4783, + "step": 13446 + }, + { + "epoch": 0.42412868632707773, + "grad_norm": 1.9514204263687134, + "learning_rate": 1.2902266335846392e-05, + "loss": 0.4814, + "step": 13447 + }, + { + "epoch": 0.42416022709351836, + "grad_norm": 1.7115857601165771, + "learning_rate": 1.2901288734115194e-05, + "loss": 0.5281, + "step": 13448 + }, + { + "epoch": 0.424191767859959, + "grad_norm": 1.774388074874878, + "learning_rate": 1.2900311102106795e-05, + "loss": 0.4874, + "step": 13449 + }, + { + "epoch": 0.4242233086263996, + "grad_norm": 1.421796441078186, + "learning_rate": 1.2899333439831398e-05, + "loss": 0.4905, + "step": 13450 + }, + { + "epoch": 0.42425484939284025, + "grad_norm": 1.4945282936096191, + "learning_rate": 1.2898355747299207e-05, + "loss": 0.5039, + "step": 13451 + }, + { + "epoch": 0.4242863901592809, + "grad_norm": 2.599097967147827, + "learning_rate": 1.2897378024520419e-05, + "loss": 0.8415, + "step": 13452 + }, + { + "epoch": 0.4243179309257215, + "grad_norm": 1.6925134658813477, + "learning_rate": 1.2896400271505243e-05, + "loss": 0.5092, + "step": 13453 + }, + { + "epoch": 0.42434947169216214, + "grad_norm": 1.427384376525879, + "learning_rate": 1.2895422488263883e-05, + "loss": 0.5225, + "step": 13454 + }, + { + "epoch": 0.42438101245860277, + "grad_norm": 1.8757543563842773, + "learning_rate": 1.2894444674806541e-05, + "loss": 0.4914, + "step": 13455 + }, + { + "epoch": 0.42441255322504334, + "grad_norm": 1.4774458408355713, + "learning_rate": 1.2893466831143422e-05, + "loss": 0.4663, + "step": 13456 + }, + { + "epoch": 0.424444093991484, + "grad_norm": 2.075620651245117, + "learning_rate": 1.2892488957284725e-05, + "loss": 0.3898, + "step": 13457 + }, + { + "epoch": 0.4244756347579246, + "grad_norm": 1.5528277158737183, + "learning_rate": 1.2891511053240666e-05, + "loss": 0.4758, + "step": 13458 + }, + { + "epoch": 0.42450717552436523, + "grad_norm": 1.3189234733581543, + "learning_rate": 1.289053311902144e-05, + "loss": 0.4942, + "step": 13459 + }, + { + "epoch": 0.42453871629080586, + "grad_norm": 1.5790985822677612, + "learning_rate": 1.2889555154637257e-05, + "loss": 0.4774, + "step": 13460 + }, + { + "epoch": 0.4245702570572465, + "grad_norm": 1.5294429063796997, + "learning_rate": 1.2888577160098323e-05, + "loss": 0.5172, + "step": 13461 + }, + { + "epoch": 0.4246017978236871, + "grad_norm": 1.4745090007781982, + "learning_rate": 1.2887599135414846e-05, + "loss": 0.5019, + "step": 13462 + }, + { + "epoch": 0.42463333859012775, + "grad_norm": 10.949121475219727, + "learning_rate": 1.2886621080597027e-05, + "loss": 0.5286, + "step": 13463 + }, + { + "epoch": 0.4246648793565684, + "grad_norm": 1.6763153076171875, + "learning_rate": 1.2885642995655072e-05, + "loss": 0.503, + "step": 13464 + }, + { + "epoch": 0.424696420123009, + "grad_norm": 2.04707407951355, + "learning_rate": 1.2884664880599198e-05, + "loss": 0.5162, + "step": 13465 + }, + { + "epoch": 0.42472796088944964, + "grad_norm": 1.6056076288223267, + "learning_rate": 1.2883686735439604e-05, + "loss": 0.4899, + "step": 13466 + }, + { + "epoch": 0.4247595016558902, + "grad_norm": 1.447943091392517, + "learning_rate": 1.2882708560186498e-05, + "loss": 0.8041, + "step": 13467 + }, + { + "epoch": 0.42479104242233084, + "grad_norm": 1.6289889812469482, + "learning_rate": 1.2881730354850092e-05, + "loss": 0.4685, + "step": 13468 + }, + { + "epoch": 0.4248225831887715, + "grad_norm": 3.1054606437683105, + "learning_rate": 1.2880752119440591e-05, + "loss": 0.5205, + "step": 13469 + }, + { + "epoch": 0.4248541239552121, + "grad_norm": 1.783980131149292, + "learning_rate": 1.2879773853968205e-05, + "loss": 0.5057, + "step": 13470 + }, + { + "epoch": 0.42488566472165273, + "grad_norm": 1.6734849214553833, + "learning_rate": 1.2878795558443143e-05, + "loss": 0.4859, + "step": 13471 + }, + { + "epoch": 0.42491720548809336, + "grad_norm": 1.8793047666549683, + "learning_rate": 1.2877817232875614e-05, + "loss": 0.484, + "step": 13472 + }, + { + "epoch": 0.424948746254534, + "grad_norm": 1.4366816282272339, + "learning_rate": 1.2876838877275827e-05, + "loss": 0.492, + "step": 13473 + }, + { + "epoch": 0.4249802870209746, + "grad_norm": 1.6411410570144653, + "learning_rate": 1.2875860491653993e-05, + "loss": 0.7913, + "step": 13474 + }, + { + "epoch": 0.42501182778741525, + "grad_norm": 1.9581512212753296, + "learning_rate": 1.2874882076020323e-05, + "loss": 0.4789, + "step": 13475 + }, + { + "epoch": 0.4250433685538559, + "grad_norm": 1.7118345499038696, + "learning_rate": 1.2873903630385027e-05, + "loss": 0.5245, + "step": 13476 + }, + { + "epoch": 0.42507490932029646, + "grad_norm": 1.6943870782852173, + "learning_rate": 1.2872925154758312e-05, + "loss": 0.5028, + "step": 13477 + }, + { + "epoch": 0.4251064500867371, + "grad_norm": 1.6017775535583496, + "learning_rate": 1.2871946649150391e-05, + "loss": 0.4705, + "step": 13478 + }, + { + "epoch": 0.4251379908531777, + "grad_norm": 1.40035879611969, + "learning_rate": 1.287096811357148e-05, + "loss": 0.4837, + "step": 13479 + }, + { + "epoch": 0.42516953161961835, + "grad_norm": 1.7629504203796387, + "learning_rate": 1.2869989548031788e-05, + "loss": 0.4633, + "step": 13480 + }, + { + "epoch": 0.425201072386059, + "grad_norm": 2.3952112197875977, + "learning_rate": 1.2869010952541523e-05, + "loss": 0.4788, + "step": 13481 + }, + { + "epoch": 0.4252326131524996, + "grad_norm": 2.4375698566436768, + "learning_rate": 1.2868032327110904e-05, + "loss": 0.4657, + "step": 13482 + }, + { + "epoch": 0.42526415391894024, + "grad_norm": 1.5349715948104858, + "learning_rate": 1.286705367175014e-05, + "loss": 0.512, + "step": 13483 + }, + { + "epoch": 0.42529569468538087, + "grad_norm": 1.6130237579345703, + "learning_rate": 1.2866074986469443e-05, + "loss": 0.497, + "step": 13484 + }, + { + "epoch": 0.4253272354518215, + "grad_norm": 1.6122852563858032, + "learning_rate": 1.2865096271279033e-05, + "loss": 0.542, + "step": 13485 + }, + { + "epoch": 0.4253587762182621, + "grad_norm": 1.3422036170959473, + "learning_rate": 1.2864117526189114e-05, + "loss": 0.4977, + "step": 13486 + }, + { + "epoch": 0.42539031698470275, + "grad_norm": 1.6712137460708618, + "learning_rate": 1.2863138751209905e-05, + "loss": 0.5141, + "step": 13487 + }, + { + "epoch": 0.42542185775114333, + "grad_norm": 2.477099895477295, + "learning_rate": 1.286215994635162e-05, + "loss": 0.5095, + "step": 13488 + }, + { + "epoch": 0.42545339851758396, + "grad_norm": 1.6946145296096802, + "learning_rate": 1.2861181111624476e-05, + "loss": 0.5191, + "step": 13489 + }, + { + "epoch": 0.4254849392840246, + "grad_norm": 1.2828260660171509, + "learning_rate": 1.2860202247038685e-05, + "loss": 0.4996, + "step": 13490 + }, + { + "epoch": 0.4255164800504652, + "grad_norm": 1.5689480304718018, + "learning_rate": 1.285922335260446e-05, + "loss": 0.831, + "step": 13491 + }, + { + "epoch": 0.42554802081690585, + "grad_norm": 1.333420991897583, + "learning_rate": 1.285824442833202e-05, + "loss": 0.4579, + "step": 13492 + }, + { + "epoch": 0.4255795615833465, + "grad_norm": 1.417388677597046, + "learning_rate": 1.2857265474231581e-05, + "loss": 0.4936, + "step": 13493 + }, + { + "epoch": 0.4256111023497871, + "grad_norm": 1.3521417379379272, + "learning_rate": 1.2856286490313359e-05, + "loss": 0.5209, + "step": 13494 + }, + { + "epoch": 0.42564264311622774, + "grad_norm": 1.8811768293380737, + "learning_rate": 1.2855307476587568e-05, + "loss": 0.5037, + "step": 13495 + }, + { + "epoch": 0.42567418388266837, + "grad_norm": 1.62786865234375, + "learning_rate": 1.2854328433064429e-05, + "loss": 0.5128, + "step": 13496 + }, + { + "epoch": 0.425705724649109, + "grad_norm": 1.3187166452407837, + "learning_rate": 1.2853349359754153e-05, + "loss": 0.4918, + "step": 13497 + }, + { + "epoch": 0.42573726541554957, + "grad_norm": 1.4306130409240723, + "learning_rate": 1.2852370256666962e-05, + "loss": 0.5062, + "step": 13498 + }, + { + "epoch": 0.4257688061819902, + "grad_norm": 1.5507365465164185, + "learning_rate": 1.2851391123813075e-05, + "loss": 0.4697, + "step": 13499 + }, + { + "epoch": 0.42580034694843083, + "grad_norm": 1.6469193696975708, + "learning_rate": 1.2850411961202704e-05, + "loss": 0.502, + "step": 13500 + }, + { + "epoch": 0.42583188771487146, + "grad_norm": 1.6351573467254639, + "learning_rate": 1.284943276884607e-05, + "loss": 0.5065, + "step": 13501 + }, + { + "epoch": 0.4258634284813121, + "grad_norm": 1.9900851249694824, + "learning_rate": 1.2848453546753397e-05, + "loss": 0.4653, + "step": 13502 + }, + { + "epoch": 0.4258949692477527, + "grad_norm": 1.8207846879959106, + "learning_rate": 1.2847474294934896e-05, + "loss": 0.5166, + "step": 13503 + }, + { + "epoch": 0.42592651001419335, + "grad_norm": 1.7661818265914917, + "learning_rate": 1.2846495013400792e-05, + "loss": 0.5128, + "step": 13504 + }, + { + "epoch": 0.425958050780634, + "grad_norm": 1.8877415657043457, + "learning_rate": 1.2845515702161303e-05, + "loss": 0.4877, + "step": 13505 + }, + { + "epoch": 0.4259895915470746, + "grad_norm": 1.7284059524536133, + "learning_rate": 1.2844536361226647e-05, + "loss": 0.5135, + "step": 13506 + }, + { + "epoch": 0.42602113231351524, + "grad_norm": 1.6770036220550537, + "learning_rate": 1.2843556990607044e-05, + "loss": 0.7822, + "step": 13507 + }, + { + "epoch": 0.42605267307995587, + "grad_norm": 1.795437216758728, + "learning_rate": 1.2842577590312714e-05, + "loss": 0.5191, + "step": 13508 + }, + { + "epoch": 0.42608421384639644, + "grad_norm": 2.0501885414123535, + "learning_rate": 1.2841598160353883e-05, + "loss": 0.4191, + "step": 13509 + }, + { + "epoch": 0.42611575461283707, + "grad_norm": 4.241781711578369, + "learning_rate": 1.284061870074077e-05, + "loss": 0.5131, + "step": 13510 + }, + { + "epoch": 0.4261472953792777, + "grad_norm": 1.580975890159607, + "learning_rate": 1.2839639211483591e-05, + "loss": 0.497, + "step": 13511 + }, + { + "epoch": 0.42617883614571833, + "grad_norm": 1.3559174537658691, + "learning_rate": 1.2838659692592574e-05, + "loss": 0.5124, + "step": 13512 + }, + { + "epoch": 0.42621037691215896, + "grad_norm": 1.63187837600708, + "learning_rate": 1.283768014407794e-05, + "loss": 0.4959, + "step": 13513 + }, + { + "epoch": 0.4262419176785996, + "grad_norm": 2.0056121349334717, + "learning_rate": 1.2836700565949909e-05, + "loss": 0.5218, + "step": 13514 + }, + { + "epoch": 0.4262734584450402, + "grad_norm": 1.470809817314148, + "learning_rate": 1.2835720958218706e-05, + "loss": 0.4686, + "step": 13515 + }, + { + "epoch": 0.42630499921148085, + "grad_norm": 1.6983627080917358, + "learning_rate": 1.2834741320894554e-05, + "loss": 0.4937, + "step": 13516 + }, + { + "epoch": 0.4263365399779215, + "grad_norm": 1.4215426445007324, + "learning_rate": 1.2833761653987673e-05, + "loss": 0.5157, + "step": 13517 + }, + { + "epoch": 0.4263680807443621, + "grad_norm": 1.7467668056488037, + "learning_rate": 1.2832781957508288e-05, + "loss": 0.5017, + "step": 13518 + }, + { + "epoch": 0.42639962151080274, + "grad_norm": 1.3364862203598022, + "learning_rate": 1.2831802231466625e-05, + "loss": 0.51, + "step": 13519 + }, + { + "epoch": 0.4264311622772433, + "grad_norm": 1.667618751525879, + "learning_rate": 1.2830822475872908e-05, + "loss": 0.4902, + "step": 13520 + }, + { + "epoch": 0.42646270304368394, + "grad_norm": 1.5141655206680298, + "learning_rate": 1.2829842690737357e-05, + "loss": 0.5034, + "step": 13521 + }, + { + "epoch": 0.4264942438101246, + "grad_norm": 1.7166264057159424, + "learning_rate": 1.2828862876070203e-05, + "loss": 0.8233, + "step": 13522 + }, + { + "epoch": 0.4265257845765652, + "grad_norm": 1.5380842685699463, + "learning_rate": 1.282788303188167e-05, + "loss": 0.5109, + "step": 13523 + }, + { + "epoch": 0.42655732534300583, + "grad_norm": 1.8307217359542847, + "learning_rate": 1.2826903158181979e-05, + "loss": 0.47, + "step": 13524 + }, + { + "epoch": 0.42658886610944646, + "grad_norm": 2.4344992637634277, + "learning_rate": 1.2825923254981359e-05, + "loss": 0.5205, + "step": 13525 + }, + { + "epoch": 0.4266204068758871, + "grad_norm": 1.3733044862747192, + "learning_rate": 1.2824943322290035e-05, + "loss": 0.4888, + "step": 13526 + }, + { + "epoch": 0.4266519476423277, + "grad_norm": 1.522273302078247, + "learning_rate": 1.2823963360118236e-05, + "loss": 0.5245, + "step": 13527 + }, + { + "epoch": 0.42668348840876835, + "grad_norm": 2.586376428604126, + "learning_rate": 1.2822983368476182e-05, + "loss": 0.509, + "step": 13528 + }, + { + "epoch": 0.426715029175209, + "grad_norm": 1.6870932579040527, + "learning_rate": 1.2822003347374107e-05, + "loss": 0.5196, + "step": 13529 + }, + { + "epoch": 0.42674656994164956, + "grad_norm": 3.030222177505493, + "learning_rate": 1.282102329682224e-05, + "loss": 0.8448, + "step": 13530 + }, + { + "epoch": 0.4267781107080902, + "grad_norm": 1.4872881174087524, + "learning_rate": 1.28200432168308e-05, + "loss": 0.506, + "step": 13531 + }, + { + "epoch": 0.4268096514745308, + "grad_norm": 1.779144525527954, + "learning_rate": 1.281906310741002e-05, + "loss": 0.8134, + "step": 13532 + }, + { + "epoch": 0.42684119224097145, + "grad_norm": 1.6578508615493774, + "learning_rate": 1.281808296857013e-05, + "loss": 0.5013, + "step": 13533 + }, + { + "epoch": 0.4268727330074121, + "grad_norm": 2.213571786880493, + "learning_rate": 1.2817102800321352e-05, + "loss": 0.5172, + "step": 13534 + }, + { + "epoch": 0.4269042737738527, + "grad_norm": 1.8038889169692993, + "learning_rate": 1.2816122602673921e-05, + "loss": 0.5347, + "step": 13535 + }, + { + "epoch": 0.42693581454029333, + "grad_norm": 1.3946454524993896, + "learning_rate": 1.2815142375638064e-05, + "loss": 0.4797, + "step": 13536 + }, + { + "epoch": 0.42696735530673396, + "grad_norm": 1.8689019680023193, + "learning_rate": 1.2814162119224012e-05, + "loss": 0.4981, + "step": 13537 + }, + { + "epoch": 0.4269988960731746, + "grad_norm": 2.40051007270813, + "learning_rate": 1.2813181833441988e-05, + "loss": 0.495, + "step": 13538 + }, + { + "epoch": 0.4270304368396152, + "grad_norm": 1.6748708486557007, + "learning_rate": 1.281220151830223e-05, + "loss": 0.5062, + "step": 13539 + }, + { + "epoch": 0.42706197760605585, + "grad_norm": 1.6945629119873047, + "learning_rate": 1.2811221173814965e-05, + "loss": 0.4893, + "step": 13540 + }, + { + "epoch": 0.42709351837249643, + "grad_norm": 1.8039072751998901, + "learning_rate": 1.2810240799990427e-05, + "loss": 0.4843, + "step": 13541 + }, + { + "epoch": 0.42712505913893706, + "grad_norm": 1.7009872198104858, + "learning_rate": 1.2809260396838842e-05, + "loss": 0.7819, + "step": 13542 + }, + { + "epoch": 0.4271565999053777, + "grad_norm": 1.510654091835022, + "learning_rate": 1.2808279964370444e-05, + "loss": 0.4941, + "step": 13543 + }, + { + "epoch": 0.4271881406718183, + "grad_norm": 1.6499180793762207, + "learning_rate": 1.2807299502595462e-05, + "loss": 0.4926, + "step": 13544 + }, + { + "epoch": 0.42721968143825895, + "grad_norm": 2.3281006813049316, + "learning_rate": 1.2806319011524134e-05, + "loss": 0.4019, + "step": 13545 + }, + { + "epoch": 0.4272512222046996, + "grad_norm": 2.2558324337005615, + "learning_rate": 1.2805338491166686e-05, + "loss": 0.4981, + "step": 13546 + }, + { + "epoch": 0.4272827629711402, + "grad_norm": 1.5396029949188232, + "learning_rate": 1.2804357941533353e-05, + "loss": 0.4965, + "step": 13547 + }, + { + "epoch": 0.42731430373758084, + "grad_norm": 1.833511471748352, + "learning_rate": 1.2803377362634367e-05, + "loss": 0.4895, + "step": 13548 + }, + { + "epoch": 0.42734584450402147, + "grad_norm": 1.7416222095489502, + "learning_rate": 1.2802396754479958e-05, + "loss": 0.495, + "step": 13549 + }, + { + "epoch": 0.4273773852704621, + "grad_norm": 2.1709439754486084, + "learning_rate": 1.2801416117080365e-05, + "loss": 0.8106, + "step": 13550 + }, + { + "epoch": 0.42740892603690267, + "grad_norm": 1.5897892713546753, + "learning_rate": 1.2800435450445822e-05, + "loss": 0.523, + "step": 13551 + }, + { + "epoch": 0.4274404668033433, + "grad_norm": 1.4738234281539917, + "learning_rate": 1.2799454754586558e-05, + "loss": 0.7945, + "step": 13552 + }, + { + "epoch": 0.42747200756978393, + "grad_norm": 1.5526466369628906, + "learning_rate": 1.2798474029512813e-05, + "loss": 0.4879, + "step": 13553 + }, + { + "epoch": 0.42750354833622456, + "grad_norm": 1.649062991142273, + "learning_rate": 1.2797493275234816e-05, + "loss": 0.4959, + "step": 13554 + }, + { + "epoch": 0.4275350891026652, + "grad_norm": 1.3165725469589233, + "learning_rate": 1.2796512491762806e-05, + "loss": 0.4592, + "step": 13555 + }, + { + "epoch": 0.4275666298691058, + "grad_norm": 1.8169256448745728, + "learning_rate": 1.2795531679107013e-05, + "loss": 0.4663, + "step": 13556 + }, + { + "epoch": 0.42759817063554645, + "grad_norm": 1.4822874069213867, + "learning_rate": 1.279455083727768e-05, + "loss": 0.5072, + "step": 13557 + }, + { + "epoch": 0.4276297114019871, + "grad_norm": 1.7239892482757568, + "learning_rate": 1.279356996628504e-05, + "loss": 0.4723, + "step": 13558 + }, + { + "epoch": 0.4276612521684277, + "grad_norm": 1.6549522876739502, + "learning_rate": 1.2792589066139324e-05, + "loss": 0.5247, + "step": 13559 + }, + { + "epoch": 0.42769279293486834, + "grad_norm": 1.721921682357788, + "learning_rate": 1.2791608136850775e-05, + "loss": 0.471, + "step": 13560 + }, + { + "epoch": 0.42772433370130897, + "grad_norm": 1.5523617267608643, + "learning_rate": 1.2790627178429628e-05, + "loss": 0.4902, + "step": 13561 + }, + { + "epoch": 0.42775587446774954, + "grad_norm": 1.4753352403640747, + "learning_rate": 1.2789646190886118e-05, + "loss": 0.4944, + "step": 13562 + }, + { + "epoch": 0.42778741523419017, + "grad_norm": 1.284367561340332, + "learning_rate": 1.2788665174230487e-05, + "loss": 0.5188, + "step": 13563 + }, + { + "epoch": 0.4278189560006308, + "grad_norm": 1.6294430494308472, + "learning_rate": 1.2787684128472968e-05, + "loss": 0.4947, + "step": 13564 + }, + { + "epoch": 0.42785049676707143, + "grad_norm": 1.6148409843444824, + "learning_rate": 1.27867030536238e-05, + "loss": 0.4928, + "step": 13565 + }, + { + "epoch": 0.42788203753351206, + "grad_norm": 2.5505964756011963, + "learning_rate": 1.2785721949693221e-05, + "loss": 0.4894, + "step": 13566 + }, + { + "epoch": 0.4279135782999527, + "grad_norm": 1.8060693740844727, + "learning_rate": 1.278474081669147e-05, + "loss": 0.4833, + "step": 13567 + }, + { + "epoch": 0.4279451190663933, + "grad_norm": 3.457061290740967, + "learning_rate": 1.2783759654628789e-05, + "loss": 0.4889, + "step": 13568 + }, + { + "epoch": 0.42797665983283395, + "grad_norm": 1.40850830078125, + "learning_rate": 1.2782778463515413e-05, + "loss": 0.5061, + "step": 13569 + }, + { + "epoch": 0.4280082005992746, + "grad_norm": 3.592302083969116, + "learning_rate": 1.2781797243361582e-05, + "loss": 0.5137, + "step": 13570 + }, + { + "epoch": 0.4280397413657152, + "grad_norm": 2.1489217281341553, + "learning_rate": 1.2780815994177538e-05, + "loss": 0.4996, + "step": 13571 + }, + { + "epoch": 0.4280712821321558, + "grad_norm": 1.4104859828948975, + "learning_rate": 1.2779834715973519e-05, + "loss": 0.493, + "step": 13572 + }, + { + "epoch": 0.4281028228985964, + "grad_norm": 2.1997451782226562, + "learning_rate": 1.2778853408759768e-05, + "loss": 0.5193, + "step": 13573 + }, + { + "epoch": 0.42813436366503704, + "grad_norm": 2.0101194381713867, + "learning_rate": 1.2777872072546522e-05, + "loss": 0.7924, + "step": 13574 + }, + { + "epoch": 0.4281659044314777, + "grad_norm": 1.6115423440933228, + "learning_rate": 1.2776890707344025e-05, + "loss": 0.5175, + "step": 13575 + }, + { + "epoch": 0.4281974451979183, + "grad_norm": 3.2526230812072754, + "learning_rate": 1.2775909313162518e-05, + "loss": 0.3627, + "step": 13576 + }, + { + "epoch": 0.42822898596435893, + "grad_norm": 2.450329542160034, + "learning_rate": 1.277492789001224e-05, + "loss": 0.4695, + "step": 13577 + }, + { + "epoch": 0.42826052673079956, + "grad_norm": 1.5875362157821655, + "learning_rate": 1.2773946437903437e-05, + "loss": 0.5233, + "step": 13578 + }, + { + "epoch": 0.4282920674972402, + "grad_norm": 1.2707730531692505, + "learning_rate": 1.2772964956846345e-05, + "loss": 0.507, + "step": 13579 + }, + { + "epoch": 0.4283236082636808, + "grad_norm": 1.6607996225357056, + "learning_rate": 1.2771983446851212e-05, + "loss": 0.5331, + "step": 13580 + }, + { + "epoch": 0.42835514903012145, + "grad_norm": 5.081118583679199, + "learning_rate": 1.2771001907928282e-05, + "loss": 0.4809, + "step": 13581 + }, + { + "epoch": 0.4283866897965621, + "grad_norm": 1.966568946838379, + "learning_rate": 1.2770020340087793e-05, + "loss": 0.4802, + "step": 13582 + }, + { + "epoch": 0.42841823056300266, + "grad_norm": 1.7995524406433105, + "learning_rate": 1.2769038743339994e-05, + "loss": 0.4813, + "step": 13583 + }, + { + "epoch": 0.4284497713294433, + "grad_norm": 1.5376713275909424, + "learning_rate": 1.2768057117695123e-05, + "loss": 0.4927, + "step": 13584 + }, + { + "epoch": 0.4284813120958839, + "grad_norm": 1.3657381534576416, + "learning_rate": 1.276707546316343e-05, + "loss": 0.4826, + "step": 13585 + }, + { + "epoch": 0.42851285286232454, + "grad_norm": 2.467724084854126, + "learning_rate": 1.276609377975515e-05, + "loss": 0.4565, + "step": 13586 + }, + { + "epoch": 0.4285443936287652, + "grad_norm": 1.3366189002990723, + "learning_rate": 1.2765112067480537e-05, + "loss": 0.4838, + "step": 13587 + }, + { + "epoch": 0.4285759343952058, + "grad_norm": 1.7844895124435425, + "learning_rate": 1.276413032634983e-05, + "loss": 0.4907, + "step": 13588 + }, + { + "epoch": 0.42860747516164643, + "grad_norm": 3.735826253890991, + "learning_rate": 1.276314855637328e-05, + "loss": 0.389, + "step": 13589 + }, + { + "epoch": 0.42863901592808706, + "grad_norm": 1.4815256595611572, + "learning_rate": 1.276216675756113e-05, + "loss": 0.5346, + "step": 13590 + }, + { + "epoch": 0.4286705566945277, + "grad_norm": 1.9290307760238647, + "learning_rate": 1.276118492992362e-05, + "loss": 0.5005, + "step": 13591 + }, + { + "epoch": 0.4287020974609683, + "grad_norm": 1.8024852275848389, + "learning_rate": 1.2760203073471008e-05, + "loss": 0.4858, + "step": 13592 + }, + { + "epoch": 0.42873363822740895, + "grad_norm": 3.773799419403076, + "learning_rate": 1.275922118821353e-05, + "loss": 0.3966, + "step": 13593 + }, + { + "epoch": 0.4287651789938495, + "grad_norm": 2.132233142852783, + "learning_rate": 1.2758239274161437e-05, + "loss": 0.4997, + "step": 13594 + }, + { + "epoch": 0.42879671976029016, + "grad_norm": 1.8161261081695557, + "learning_rate": 1.2757257331324977e-05, + "loss": 0.5256, + "step": 13595 + }, + { + "epoch": 0.4288282605267308, + "grad_norm": 1.9555976390838623, + "learning_rate": 1.2756275359714391e-05, + "loss": 0.4942, + "step": 13596 + }, + { + "epoch": 0.4288598012931714, + "grad_norm": 4.508269309997559, + "learning_rate": 1.2755293359339935e-05, + "loss": 0.8151, + "step": 13597 + }, + { + "epoch": 0.42889134205961205, + "grad_norm": 2.2048089504241943, + "learning_rate": 1.2754311330211853e-05, + "loss": 0.5049, + "step": 13598 + }, + { + "epoch": 0.4289228828260527, + "grad_norm": 1.6402565240859985, + "learning_rate": 1.2753329272340393e-05, + "loss": 0.5077, + "step": 13599 + }, + { + "epoch": 0.4289544235924933, + "grad_norm": 1.8187400102615356, + "learning_rate": 1.2752347185735804e-05, + "loss": 0.4957, + "step": 13600 + }, + { + "epoch": 0.42898596435893394, + "grad_norm": 1.617723822593689, + "learning_rate": 1.2751365070408335e-05, + "loss": 0.482, + "step": 13601 + }, + { + "epoch": 0.42901750512537457, + "grad_norm": 3.3940608501434326, + "learning_rate": 1.2750382926368237e-05, + "loss": 0.3611, + "step": 13602 + }, + { + "epoch": 0.4290490458918152, + "grad_norm": 1.4094371795654297, + "learning_rate": 1.2749400753625757e-05, + "loss": 0.4739, + "step": 13603 + }, + { + "epoch": 0.42908058665825577, + "grad_norm": 1.9694879055023193, + "learning_rate": 1.2748418552191143e-05, + "loss": 0.4743, + "step": 13604 + }, + { + "epoch": 0.4291121274246964, + "grad_norm": 1.8646035194396973, + "learning_rate": 1.2747436322074649e-05, + "loss": 0.5334, + "step": 13605 + }, + { + "epoch": 0.42914366819113703, + "grad_norm": 1.5725916624069214, + "learning_rate": 1.2746454063286524e-05, + "loss": 0.5241, + "step": 13606 + }, + { + "epoch": 0.42917520895757766, + "grad_norm": 1.7150654792785645, + "learning_rate": 1.2745471775837017e-05, + "loss": 0.5114, + "step": 13607 + }, + { + "epoch": 0.4292067497240183, + "grad_norm": 2.774582624435425, + "learning_rate": 1.274448945973638e-05, + "loss": 0.495, + "step": 13608 + }, + { + "epoch": 0.4292382904904589, + "grad_norm": 1.6392656564712524, + "learning_rate": 1.2743507114994867e-05, + "loss": 0.5343, + "step": 13609 + }, + { + "epoch": 0.42926983125689955, + "grad_norm": 3.072979688644409, + "learning_rate": 1.2742524741622724e-05, + "loss": 0.5017, + "step": 13610 + }, + { + "epoch": 0.4293013720233402, + "grad_norm": 1.5420928001403809, + "learning_rate": 1.2741542339630206e-05, + "loss": 0.4795, + "step": 13611 + }, + { + "epoch": 0.4293329127897808, + "grad_norm": 1.937835454940796, + "learning_rate": 1.2740559909027567e-05, + "loss": 0.4309, + "step": 13612 + }, + { + "epoch": 0.42936445355622144, + "grad_norm": 2.017038106918335, + "learning_rate": 1.2739577449825057e-05, + "loss": 0.4695, + "step": 13613 + }, + { + "epoch": 0.42939599432266207, + "grad_norm": 2.224729299545288, + "learning_rate": 1.2738594962032929e-05, + "loss": 0.8147, + "step": 13614 + }, + { + "epoch": 0.42942753508910264, + "grad_norm": 8.537428855895996, + "learning_rate": 1.2737612445661434e-05, + "loss": 0.4717, + "step": 13615 + }, + { + "epoch": 0.42945907585554327, + "grad_norm": 1.5241646766662598, + "learning_rate": 1.2736629900720832e-05, + "loss": 0.5118, + "step": 13616 + }, + { + "epoch": 0.4294906166219839, + "grad_norm": 1.4803589582443237, + "learning_rate": 1.2735647327221368e-05, + "loss": 0.4782, + "step": 13617 + }, + { + "epoch": 0.42952215738842453, + "grad_norm": 1.5225404500961304, + "learning_rate": 1.27346647251733e-05, + "loss": 0.4736, + "step": 13618 + }, + { + "epoch": 0.42955369815486516, + "grad_norm": 2.046231269836426, + "learning_rate": 1.2733682094586884e-05, + "loss": 0.3548, + "step": 13619 + }, + { + "epoch": 0.4295852389213058, + "grad_norm": 1.5124218463897705, + "learning_rate": 1.273269943547237e-05, + "loss": 0.4984, + "step": 13620 + }, + { + "epoch": 0.4296167796877464, + "grad_norm": 3.482276678085327, + "learning_rate": 1.2731716747840016e-05, + "loss": 0.8168, + "step": 13621 + }, + { + "epoch": 0.42964832045418705, + "grad_norm": 1.754824161529541, + "learning_rate": 1.273073403170008e-05, + "loss": 0.4717, + "step": 13622 + }, + { + "epoch": 0.4296798612206277, + "grad_norm": 1.5777509212493896, + "learning_rate": 1.2729751287062816e-05, + "loss": 0.824, + "step": 13623 + }, + { + "epoch": 0.4297114019870683, + "grad_norm": 1.878541111946106, + "learning_rate": 1.2728768513938472e-05, + "loss": 0.5054, + "step": 13624 + }, + { + "epoch": 0.4297429427535089, + "grad_norm": 2.192014455795288, + "learning_rate": 1.272778571233731e-05, + "loss": 0.487, + "step": 13625 + }, + { + "epoch": 0.4297744835199495, + "grad_norm": 3.0619754791259766, + "learning_rate": 1.272680288226959e-05, + "loss": 0.8161, + "step": 13626 + }, + { + "epoch": 0.42980602428639014, + "grad_norm": 4.903991222381592, + "learning_rate": 1.2725820023745561e-05, + "loss": 0.4797, + "step": 13627 + }, + { + "epoch": 0.4298375650528308, + "grad_norm": 1.5805965662002563, + "learning_rate": 1.2724837136775484e-05, + "loss": 0.4898, + "step": 13628 + }, + { + "epoch": 0.4298691058192714, + "grad_norm": 1.7803977727890015, + "learning_rate": 1.2723854221369614e-05, + "loss": 0.5108, + "step": 13629 + }, + { + "epoch": 0.42990064658571203, + "grad_norm": 2.560732126235962, + "learning_rate": 1.2722871277538213e-05, + "loss": 0.4848, + "step": 13630 + }, + { + "epoch": 0.42993218735215266, + "grad_norm": 1.8777012825012207, + "learning_rate": 1.2721888305291536e-05, + "loss": 0.5062, + "step": 13631 + }, + { + "epoch": 0.4299637281185933, + "grad_norm": 1.8573477268218994, + "learning_rate": 1.272090530463984e-05, + "loss": 0.4917, + "step": 13632 + }, + { + "epoch": 0.4299952688850339, + "grad_norm": 2.006086826324463, + "learning_rate": 1.2719922275593388e-05, + "loss": 0.5062, + "step": 13633 + }, + { + "epoch": 0.43002680965147455, + "grad_norm": 3.1076629161834717, + "learning_rate": 1.2718939218162432e-05, + "loss": 0.8094, + "step": 13634 + }, + { + "epoch": 0.4300583504179152, + "grad_norm": 2.7054314613342285, + "learning_rate": 1.2717956132357232e-05, + "loss": 0.8119, + "step": 13635 + }, + { + "epoch": 0.43008989118435575, + "grad_norm": 1.6967755556106567, + "learning_rate": 1.271697301818805e-05, + "loss": 0.5099, + "step": 13636 + }, + { + "epoch": 0.4301214319507964, + "grad_norm": 1.441101312637329, + "learning_rate": 1.271598987566515e-05, + "loss": 0.4871, + "step": 13637 + }, + { + "epoch": 0.430152972717237, + "grad_norm": 1.6807974576950073, + "learning_rate": 1.2715006704798781e-05, + "loss": 0.804, + "step": 13638 + }, + { + "epoch": 0.43018451348367764, + "grad_norm": 1.8485411405563354, + "learning_rate": 1.271402350559921e-05, + "loss": 0.5007, + "step": 13639 + }, + { + "epoch": 0.4302160542501183, + "grad_norm": 6.0843729972839355, + "learning_rate": 1.2713040278076698e-05, + "loss": 0.3806, + "step": 13640 + }, + { + "epoch": 0.4302475950165589, + "grad_norm": 1.533200740814209, + "learning_rate": 1.2712057022241502e-05, + "loss": 0.4686, + "step": 13641 + }, + { + "epoch": 0.43027913578299953, + "grad_norm": 2.58492374420166, + "learning_rate": 1.2711073738103887e-05, + "loss": 0.4842, + "step": 13642 + }, + { + "epoch": 0.43031067654944016, + "grad_norm": 2.2231802940368652, + "learning_rate": 1.2710090425674114e-05, + "loss": 0.5155, + "step": 13643 + }, + { + "epoch": 0.4303422173158808, + "grad_norm": 1.9079387187957764, + "learning_rate": 1.270910708496244e-05, + "loss": 0.4911, + "step": 13644 + }, + { + "epoch": 0.4303737580823214, + "grad_norm": 1.7838243246078491, + "learning_rate": 1.2708123715979131e-05, + "loss": 0.5288, + "step": 13645 + }, + { + "epoch": 0.43040529884876205, + "grad_norm": 1.6456972360610962, + "learning_rate": 1.2707140318734448e-05, + "loss": 0.487, + "step": 13646 + }, + { + "epoch": 0.4304368396152026, + "grad_norm": 1.6136060953140259, + "learning_rate": 1.2706156893238656e-05, + "loss": 0.5035, + "step": 13647 + }, + { + "epoch": 0.43046838038164326, + "grad_norm": 1.681877613067627, + "learning_rate": 1.2705173439502013e-05, + "loss": 0.4855, + "step": 13648 + }, + { + "epoch": 0.4304999211480839, + "grad_norm": 1.7551958560943604, + "learning_rate": 1.2704189957534783e-05, + "loss": 0.4796, + "step": 13649 + }, + { + "epoch": 0.4305314619145245, + "grad_norm": 1.6971482038497925, + "learning_rate": 1.2703206447347234e-05, + "loss": 0.8294, + "step": 13650 + }, + { + "epoch": 0.43056300268096515, + "grad_norm": 1.473152995109558, + "learning_rate": 1.2702222908949625e-05, + "loss": 0.4779, + "step": 13651 + }, + { + "epoch": 0.4305945434474058, + "grad_norm": 1.8348255157470703, + "learning_rate": 1.2701239342352223e-05, + "loss": 0.5066, + "step": 13652 + }, + { + "epoch": 0.4306260842138464, + "grad_norm": 1.589495301246643, + "learning_rate": 1.270025574756529e-05, + "loss": 0.4855, + "step": 13653 + }, + { + "epoch": 0.43065762498028703, + "grad_norm": 2.318220615386963, + "learning_rate": 1.2699272124599094e-05, + "loss": 0.3676, + "step": 13654 + }, + { + "epoch": 0.43068916574672766, + "grad_norm": 1.7308354377746582, + "learning_rate": 1.2698288473463896e-05, + "loss": 0.5125, + "step": 13655 + }, + { + "epoch": 0.4307207065131683, + "grad_norm": 1.9730595350265503, + "learning_rate": 1.269730479416996e-05, + "loss": 0.806, + "step": 13656 + }, + { + "epoch": 0.43075224727960887, + "grad_norm": 1.7029836177825928, + "learning_rate": 1.269632108672756e-05, + "loss": 0.5301, + "step": 13657 + }, + { + "epoch": 0.4307837880460495, + "grad_norm": 1.7774609327316284, + "learning_rate": 1.2695337351146947e-05, + "loss": 0.4998, + "step": 13658 + }, + { + "epoch": 0.43081532881249013, + "grad_norm": 1.411142349243164, + "learning_rate": 1.26943535874384e-05, + "loss": 0.4901, + "step": 13659 + }, + { + "epoch": 0.43084686957893076, + "grad_norm": 1.6821733713150024, + "learning_rate": 1.2693369795612182e-05, + "loss": 0.4617, + "step": 13660 + }, + { + "epoch": 0.4308784103453714, + "grad_norm": 1.5665112733840942, + "learning_rate": 1.2692385975678563e-05, + "loss": 0.5075, + "step": 13661 + }, + { + "epoch": 0.430909951111812, + "grad_norm": 1.7218059301376343, + "learning_rate": 1.2691402127647803e-05, + "loss": 0.5097, + "step": 13662 + }, + { + "epoch": 0.43094149187825265, + "grad_norm": 2.0211145877838135, + "learning_rate": 1.2690418251530168e-05, + "loss": 0.5032, + "step": 13663 + }, + { + "epoch": 0.4309730326446933, + "grad_norm": 1.5326520204544067, + "learning_rate": 1.2689434347335933e-05, + "loss": 0.5014, + "step": 13664 + }, + { + "epoch": 0.4310045734111339, + "grad_norm": 2.3826358318328857, + "learning_rate": 1.2688450415075362e-05, + "loss": 0.4823, + "step": 13665 + }, + { + "epoch": 0.43103611417757454, + "grad_norm": 2.2438926696777344, + "learning_rate": 1.2687466454758722e-05, + "loss": 0.5345, + "step": 13666 + }, + { + "epoch": 0.43106765494401517, + "grad_norm": 1.8563449382781982, + "learning_rate": 1.2686482466396285e-05, + "loss": 0.5035, + "step": 13667 + }, + { + "epoch": 0.43109919571045574, + "grad_norm": 3.648420572280884, + "learning_rate": 1.2685498449998315e-05, + "loss": 0.4101, + "step": 13668 + }, + { + "epoch": 0.43113073647689637, + "grad_norm": 2.091671943664551, + "learning_rate": 1.2684514405575083e-05, + "loss": 0.4773, + "step": 13669 + }, + { + "epoch": 0.431162277243337, + "grad_norm": 1.7025177478790283, + "learning_rate": 1.2683530333136859e-05, + "loss": 0.5082, + "step": 13670 + }, + { + "epoch": 0.43119381800977763, + "grad_norm": 1.7834118604660034, + "learning_rate": 1.2682546232693915e-05, + "loss": 0.4881, + "step": 13671 + }, + { + "epoch": 0.43122535877621826, + "grad_norm": 1.7564448118209839, + "learning_rate": 1.268156210425652e-05, + "loss": 0.5017, + "step": 13672 + }, + { + "epoch": 0.4312568995426589, + "grad_norm": 1.7081888914108276, + "learning_rate": 1.2680577947834937e-05, + "loss": 0.5208, + "step": 13673 + }, + { + "epoch": 0.4312884403090995, + "grad_norm": 1.805152416229248, + "learning_rate": 1.2679593763439446e-05, + "loss": 0.7898, + "step": 13674 + }, + { + "epoch": 0.43131998107554015, + "grad_norm": 1.9931426048278809, + "learning_rate": 1.267860955108031e-05, + "loss": 0.5097, + "step": 13675 + }, + { + "epoch": 0.4313515218419808, + "grad_norm": 1.9165117740631104, + "learning_rate": 1.2677625310767803e-05, + "loss": 0.4071, + "step": 13676 + }, + { + "epoch": 0.4313830626084214, + "grad_norm": 2.183361053466797, + "learning_rate": 1.26766410425122e-05, + "loss": 0.4865, + "step": 13677 + }, + { + "epoch": 0.431414603374862, + "grad_norm": 1.600956678390503, + "learning_rate": 1.2675656746323768e-05, + "loss": 0.4793, + "step": 13678 + }, + { + "epoch": 0.4314461441413026, + "grad_norm": 1.68003511428833, + "learning_rate": 1.2674672422212777e-05, + "loss": 0.4868, + "step": 13679 + }, + { + "epoch": 0.43147768490774324, + "grad_norm": 1.6034836769104004, + "learning_rate": 1.2673688070189506e-05, + "loss": 0.521, + "step": 13680 + }, + { + "epoch": 0.43150922567418387, + "grad_norm": 1.561292052268982, + "learning_rate": 1.2672703690264225e-05, + "loss": 0.5106, + "step": 13681 + }, + { + "epoch": 0.4315407664406245, + "grad_norm": 2.54360032081604, + "learning_rate": 1.2671719282447205e-05, + "loss": 0.5108, + "step": 13682 + }, + { + "epoch": 0.43157230720706513, + "grad_norm": 1.4786714315414429, + "learning_rate": 1.2670734846748717e-05, + "loss": 0.4706, + "step": 13683 + }, + { + "epoch": 0.43160384797350576, + "grad_norm": 1.5371383428573608, + "learning_rate": 1.2669750383179039e-05, + "loss": 0.5169, + "step": 13684 + }, + { + "epoch": 0.4316353887399464, + "grad_norm": 2.320613384246826, + "learning_rate": 1.2668765891748444e-05, + "loss": 0.4791, + "step": 13685 + }, + { + "epoch": 0.431666929506387, + "grad_norm": 1.400807499885559, + "learning_rate": 1.2667781372467203e-05, + "loss": 0.5148, + "step": 13686 + }, + { + "epoch": 0.43169847027282765, + "grad_norm": 1.6987197399139404, + "learning_rate": 1.2666796825345591e-05, + "loss": 0.4811, + "step": 13687 + }, + { + "epoch": 0.4317300110392683, + "grad_norm": 1.629976749420166, + "learning_rate": 1.2665812250393886e-05, + "loss": 0.4808, + "step": 13688 + }, + { + "epoch": 0.43176155180570885, + "grad_norm": 1.5277857780456543, + "learning_rate": 1.2664827647622357e-05, + "loss": 0.491, + "step": 13689 + }, + { + "epoch": 0.4317930925721495, + "grad_norm": 1.577709674835205, + "learning_rate": 1.2663843017041285e-05, + "loss": 0.4972, + "step": 13690 + }, + { + "epoch": 0.4318246333385901, + "grad_norm": 1.429643988609314, + "learning_rate": 1.2662858358660942e-05, + "loss": 0.4669, + "step": 13691 + }, + { + "epoch": 0.43185617410503074, + "grad_norm": 2.021228313446045, + "learning_rate": 1.2661873672491604e-05, + "loss": 0.8215, + "step": 13692 + }, + { + "epoch": 0.4318877148714714, + "grad_norm": 1.324989676475525, + "learning_rate": 1.2660888958543547e-05, + "loss": 0.4772, + "step": 13693 + }, + { + "epoch": 0.431919255637912, + "grad_norm": 2.3923227787017822, + "learning_rate": 1.2659904216827048e-05, + "loss": 0.4967, + "step": 13694 + }, + { + "epoch": 0.43195079640435263, + "grad_norm": 1.562198519706726, + "learning_rate": 1.2658919447352384e-05, + "loss": 0.5022, + "step": 13695 + }, + { + "epoch": 0.43198233717079326, + "grad_norm": 1.6542019844055176, + "learning_rate": 1.2657934650129831e-05, + "loss": 0.4909, + "step": 13696 + }, + { + "epoch": 0.4320138779372339, + "grad_norm": 1.5754196643829346, + "learning_rate": 1.2656949825169665e-05, + "loss": 0.5113, + "step": 13697 + }, + { + "epoch": 0.4320454187036745, + "grad_norm": 1.8549473285675049, + "learning_rate": 1.2655964972482164e-05, + "loss": 0.5228, + "step": 13698 + }, + { + "epoch": 0.4320769594701151, + "grad_norm": 1.8217889070510864, + "learning_rate": 1.2654980092077606e-05, + "loss": 0.5031, + "step": 13699 + }, + { + "epoch": 0.4321085002365557, + "grad_norm": 1.7251079082489014, + "learning_rate": 1.265399518396627e-05, + "loss": 0.544, + "step": 13700 + }, + { + "epoch": 0.43214004100299636, + "grad_norm": 1.949050784111023, + "learning_rate": 1.2653010248158436e-05, + "loss": 0.8182, + "step": 13701 + }, + { + "epoch": 0.432171581769437, + "grad_norm": 5.575913906097412, + "learning_rate": 1.2652025284664377e-05, + "loss": 0.5163, + "step": 13702 + }, + { + "epoch": 0.4322031225358776, + "grad_norm": 1.3508541584014893, + "learning_rate": 1.2651040293494376e-05, + "loss": 0.5126, + "step": 13703 + }, + { + "epoch": 0.43223466330231824, + "grad_norm": 1.4591675996780396, + "learning_rate": 1.265005527465871e-05, + "loss": 0.4827, + "step": 13704 + }, + { + "epoch": 0.4322662040687589, + "grad_norm": 2.6747772693634033, + "learning_rate": 1.264907022816766e-05, + "loss": 0.4804, + "step": 13705 + }, + { + "epoch": 0.4322977448351995, + "grad_norm": 1.3498355150222778, + "learning_rate": 1.2648085154031506e-05, + "loss": 0.5246, + "step": 13706 + }, + { + "epoch": 0.43232928560164013, + "grad_norm": 2.077263593673706, + "learning_rate": 1.2647100052260526e-05, + "loss": 0.5173, + "step": 13707 + }, + { + "epoch": 0.43236082636808076, + "grad_norm": 4.3291120529174805, + "learning_rate": 1.2646114922865e-05, + "loss": 0.7942, + "step": 13708 + }, + { + "epoch": 0.4323923671345214, + "grad_norm": 1.4200937747955322, + "learning_rate": 1.2645129765855215e-05, + "loss": 0.5017, + "step": 13709 + }, + { + "epoch": 0.43242390790096197, + "grad_norm": 1.6278669834136963, + "learning_rate": 1.2644144581241444e-05, + "loss": 0.4723, + "step": 13710 + }, + { + "epoch": 0.4324554486674026, + "grad_norm": 1.343770146369934, + "learning_rate": 1.2643159369033973e-05, + "loss": 0.4734, + "step": 13711 + }, + { + "epoch": 0.4324869894338432, + "grad_norm": 1.5072822570800781, + "learning_rate": 1.2642174129243082e-05, + "loss": 0.4646, + "step": 13712 + }, + { + "epoch": 0.43251853020028386, + "grad_norm": 2.4149270057678223, + "learning_rate": 1.2641188861879049e-05, + "loss": 0.497, + "step": 13713 + }, + { + "epoch": 0.4325500709667245, + "grad_norm": 1.4232020378112793, + "learning_rate": 1.264020356695216e-05, + "loss": 0.488, + "step": 13714 + }, + { + "epoch": 0.4325816117331651, + "grad_norm": 1.5045762062072754, + "learning_rate": 1.2639218244472697e-05, + "loss": 0.4784, + "step": 13715 + }, + { + "epoch": 0.43261315249960575, + "grad_norm": 1.7630085945129395, + "learning_rate": 1.2638232894450944e-05, + "loss": 0.8007, + "step": 13716 + }, + { + "epoch": 0.4326446932660464, + "grad_norm": 1.4723255634307861, + "learning_rate": 1.263724751689718e-05, + "loss": 0.5101, + "step": 13717 + }, + { + "epoch": 0.432676234032487, + "grad_norm": 1.3513426780700684, + "learning_rate": 1.263626211182169e-05, + "loss": 0.4881, + "step": 13718 + }, + { + "epoch": 0.43270777479892764, + "grad_norm": 1.7951339483261108, + "learning_rate": 1.263527667923476e-05, + "loss": 0.5006, + "step": 13719 + }, + { + "epoch": 0.43273931556536827, + "grad_norm": 1.909814715385437, + "learning_rate": 1.263429121914667e-05, + "loss": 0.7941, + "step": 13720 + }, + { + "epoch": 0.43277085633180884, + "grad_norm": 1.663816213607788, + "learning_rate": 1.2633305731567708e-05, + "loss": 0.5272, + "step": 13721 + }, + { + "epoch": 0.43280239709824947, + "grad_norm": 1.309234857559204, + "learning_rate": 1.2632320216508158e-05, + "loss": 0.4917, + "step": 13722 + }, + { + "epoch": 0.4328339378646901, + "grad_norm": 1.7223612070083618, + "learning_rate": 1.2631334673978297e-05, + "loss": 0.4734, + "step": 13723 + }, + { + "epoch": 0.43286547863113073, + "grad_norm": 1.4773414134979248, + "learning_rate": 1.2630349103988418e-05, + "loss": 0.4732, + "step": 13724 + }, + { + "epoch": 0.43289701939757136, + "grad_norm": 1.5595365762710571, + "learning_rate": 1.2629363506548801e-05, + "loss": 0.5096, + "step": 13725 + }, + { + "epoch": 0.432928560164012, + "grad_norm": 1.5290215015411377, + "learning_rate": 1.2628377881669739e-05, + "loss": 0.5061, + "step": 13726 + }, + { + "epoch": 0.4329601009304526, + "grad_norm": 1.360967993736267, + "learning_rate": 1.2627392229361508e-05, + "loss": 0.4796, + "step": 13727 + }, + { + "epoch": 0.43299164169689325, + "grad_norm": 1.534559726715088, + "learning_rate": 1.2626406549634402e-05, + "loss": 0.5096, + "step": 13728 + }, + { + "epoch": 0.4330231824633339, + "grad_norm": 1.5596178770065308, + "learning_rate": 1.2625420842498706e-05, + "loss": 0.4874, + "step": 13729 + }, + { + "epoch": 0.4330547232297745, + "grad_norm": 1.4519803524017334, + "learning_rate": 1.2624435107964701e-05, + "loss": 0.4895, + "step": 13730 + }, + { + "epoch": 0.4330862639962151, + "grad_norm": 1.5500965118408203, + "learning_rate": 1.262344934604268e-05, + "loss": 0.4976, + "step": 13731 + }, + { + "epoch": 0.4331178047626557, + "grad_norm": 1.3211697340011597, + "learning_rate": 1.2622463556742927e-05, + "loss": 0.4876, + "step": 13732 + }, + { + "epoch": 0.43314934552909634, + "grad_norm": 1.7382664680480957, + "learning_rate": 1.2621477740075734e-05, + "loss": 0.8058, + "step": 13733 + }, + { + "epoch": 0.43318088629553697, + "grad_norm": 1.3305375576019287, + "learning_rate": 1.2620491896051382e-05, + "loss": 0.5309, + "step": 13734 + }, + { + "epoch": 0.4332124270619776, + "grad_norm": 1.3425036668777466, + "learning_rate": 1.261950602468016e-05, + "loss": 0.4919, + "step": 13735 + }, + { + "epoch": 0.43324396782841823, + "grad_norm": 1.7112747430801392, + "learning_rate": 1.2618520125972362e-05, + "loss": 0.4627, + "step": 13736 + }, + { + "epoch": 0.43327550859485886, + "grad_norm": 1.2550190687179565, + "learning_rate": 1.2617534199938274e-05, + "loss": 0.4815, + "step": 13737 + }, + { + "epoch": 0.4333070493612995, + "grad_norm": 1.4806101322174072, + "learning_rate": 1.261654824658818e-05, + "loss": 0.5085, + "step": 13738 + }, + { + "epoch": 0.4333385901277401, + "grad_norm": 2.6460514068603516, + "learning_rate": 1.2615562265932378e-05, + "loss": 0.5396, + "step": 13739 + }, + { + "epoch": 0.43337013089418075, + "grad_norm": 1.4343290328979492, + "learning_rate": 1.2614576257981149e-05, + "loss": 0.5233, + "step": 13740 + }, + { + "epoch": 0.4334016716606214, + "grad_norm": 1.5506627559661865, + "learning_rate": 1.261359022274479e-05, + "loss": 0.484, + "step": 13741 + }, + { + "epoch": 0.43343321242706195, + "grad_norm": 1.9287469387054443, + "learning_rate": 1.2612604160233586e-05, + "loss": 0.4758, + "step": 13742 + }, + { + "epoch": 0.4334647531935026, + "grad_norm": 1.3123162984848022, + "learning_rate": 1.2611618070457829e-05, + "loss": 0.5056, + "step": 13743 + }, + { + "epoch": 0.4334962939599432, + "grad_norm": 1.581517219543457, + "learning_rate": 1.2610631953427808e-05, + "loss": 0.4571, + "step": 13744 + }, + { + "epoch": 0.43352783472638384, + "grad_norm": 1.5248334407806396, + "learning_rate": 1.2609645809153817e-05, + "loss": 0.5058, + "step": 13745 + }, + { + "epoch": 0.4335593754928245, + "grad_norm": 1.5101101398468018, + "learning_rate": 1.2608659637646148e-05, + "loss": 0.5101, + "step": 13746 + }, + { + "epoch": 0.4335909162592651, + "grad_norm": 1.9425756931304932, + "learning_rate": 1.2607673438915086e-05, + "loss": 0.8302, + "step": 13747 + }, + { + "epoch": 0.43362245702570573, + "grad_norm": 1.539502501487732, + "learning_rate": 1.2606687212970927e-05, + "loss": 0.4805, + "step": 13748 + }, + { + "epoch": 0.43365399779214636, + "grad_norm": 2.2006256580352783, + "learning_rate": 1.2605700959823968e-05, + "loss": 0.493, + "step": 13749 + }, + { + "epoch": 0.433685538558587, + "grad_norm": 1.6254609823226929, + "learning_rate": 1.260471467948449e-05, + "loss": 0.5041, + "step": 13750 + }, + { + "epoch": 0.4337170793250276, + "grad_norm": 1.9930974245071411, + "learning_rate": 1.2603728371962795e-05, + "loss": 0.4733, + "step": 13751 + }, + { + "epoch": 0.4337486200914682, + "grad_norm": 1.524524450302124, + "learning_rate": 1.2602742037269174e-05, + "loss": 0.4928, + "step": 13752 + }, + { + "epoch": 0.4337801608579088, + "grad_norm": 1.6072144508361816, + "learning_rate": 1.2601755675413917e-05, + "loss": 0.4837, + "step": 13753 + }, + { + "epoch": 0.43381170162434946, + "grad_norm": 1.7431423664093018, + "learning_rate": 1.260076928640732e-05, + "loss": 0.5116, + "step": 13754 + }, + { + "epoch": 0.4338432423907901, + "grad_norm": 1.6792798042297363, + "learning_rate": 1.2599782870259673e-05, + "loss": 0.5094, + "step": 13755 + }, + { + "epoch": 0.4338747831572307, + "grad_norm": 1.5728353261947632, + "learning_rate": 1.2598796426981275e-05, + "loss": 0.4956, + "step": 13756 + }, + { + "epoch": 0.43390632392367134, + "grad_norm": 1.5068042278289795, + "learning_rate": 1.2597809956582422e-05, + "loss": 0.5343, + "step": 13757 + }, + { + "epoch": 0.433937864690112, + "grad_norm": 1.8551274538040161, + "learning_rate": 1.25968234590734e-05, + "loss": 0.5093, + "step": 13758 + }, + { + "epoch": 0.4339694054565526, + "grad_norm": 2.2985312938690186, + "learning_rate": 1.2595836934464512e-05, + "loss": 0.4755, + "step": 13759 + }, + { + "epoch": 0.43400094622299323, + "grad_norm": 1.5040327310562134, + "learning_rate": 1.2594850382766049e-05, + "loss": 0.4985, + "step": 13760 + }, + { + "epoch": 0.43403248698943386, + "grad_norm": 1.5740458965301514, + "learning_rate": 1.2593863803988307e-05, + "loss": 0.4999, + "step": 13761 + }, + { + "epoch": 0.4340640277558745, + "grad_norm": 1.6229032278060913, + "learning_rate": 1.2592877198141582e-05, + "loss": 0.3904, + "step": 13762 + }, + { + "epoch": 0.43409556852231507, + "grad_norm": 1.5955268144607544, + "learning_rate": 1.2591890565236168e-05, + "loss": 0.8103, + "step": 13763 + }, + { + "epoch": 0.4341271092887557, + "grad_norm": 1.8295679092407227, + "learning_rate": 1.2590903905282368e-05, + "loss": 0.48, + "step": 13764 + }, + { + "epoch": 0.4341586500551963, + "grad_norm": 1.3855146169662476, + "learning_rate": 1.2589917218290471e-05, + "loss": 0.5078, + "step": 13765 + }, + { + "epoch": 0.43419019082163696, + "grad_norm": 1.462643027305603, + "learning_rate": 1.2588930504270775e-05, + "loss": 0.4947, + "step": 13766 + }, + { + "epoch": 0.4342217315880776, + "grad_norm": 1.7608325481414795, + "learning_rate": 1.2587943763233584e-05, + "loss": 0.4076, + "step": 13767 + }, + { + "epoch": 0.4342532723545182, + "grad_norm": 1.7798107862472534, + "learning_rate": 1.2586956995189186e-05, + "loss": 0.5003, + "step": 13768 + }, + { + "epoch": 0.43428481312095885, + "grad_norm": 1.267384648323059, + "learning_rate": 1.2585970200147886e-05, + "loss": 0.4979, + "step": 13769 + }, + { + "epoch": 0.4343163538873995, + "grad_norm": 1.818864345550537, + "learning_rate": 1.2584983378119977e-05, + "loss": 0.4968, + "step": 13770 + }, + { + "epoch": 0.4343478946538401, + "grad_norm": 1.7439093589782715, + "learning_rate": 1.2583996529115762e-05, + "loss": 0.489, + "step": 13771 + }, + { + "epoch": 0.43437943542028074, + "grad_norm": 1.7999945878982544, + "learning_rate": 1.2583009653145534e-05, + "loss": 0.5079, + "step": 13772 + }, + { + "epoch": 0.43441097618672136, + "grad_norm": 1.4497029781341553, + "learning_rate": 1.2582022750219593e-05, + "loss": 0.4907, + "step": 13773 + }, + { + "epoch": 0.43444251695316194, + "grad_norm": 1.623866081237793, + "learning_rate": 1.2581035820348244e-05, + "loss": 0.8039, + "step": 13774 + }, + { + "epoch": 0.43447405771960257, + "grad_norm": 2.2202541828155518, + "learning_rate": 1.258004886354178e-05, + "loss": 0.5011, + "step": 13775 + }, + { + "epoch": 0.4345055984860432, + "grad_norm": 1.2858517169952393, + "learning_rate": 1.2579061879810501e-05, + "loss": 0.4891, + "step": 13776 + }, + { + "epoch": 0.43453713925248383, + "grad_norm": 1.5089530944824219, + "learning_rate": 1.2578074869164713e-05, + "loss": 0.4985, + "step": 13777 + }, + { + "epoch": 0.43456868001892446, + "grad_norm": 1.3984909057617188, + "learning_rate": 1.257708783161471e-05, + "loss": 0.4893, + "step": 13778 + }, + { + "epoch": 0.4346002207853651, + "grad_norm": 1.6247036457061768, + "learning_rate": 1.2576100767170796e-05, + "loss": 0.4839, + "step": 13779 + }, + { + "epoch": 0.4346317615518057, + "grad_norm": 1.593450665473938, + "learning_rate": 1.2575113675843267e-05, + "loss": 0.5175, + "step": 13780 + }, + { + "epoch": 0.43466330231824635, + "grad_norm": 1.369529128074646, + "learning_rate": 1.257412655764243e-05, + "loss": 0.4945, + "step": 13781 + }, + { + "epoch": 0.434694843084687, + "grad_norm": 2.338181257247925, + "learning_rate": 1.257313941257858e-05, + "loss": 0.5152, + "step": 13782 + }, + { + "epoch": 0.4347263838511276, + "grad_norm": 2.069489002227783, + "learning_rate": 1.2572152240662023e-05, + "loss": 0.513, + "step": 13783 + }, + { + "epoch": 0.4347579246175682, + "grad_norm": 1.6623848676681519, + "learning_rate": 1.2571165041903064e-05, + "loss": 0.461, + "step": 13784 + }, + { + "epoch": 0.4347894653840088, + "grad_norm": 1.3874919414520264, + "learning_rate": 1.2570177816311996e-05, + "loss": 0.4948, + "step": 13785 + }, + { + "epoch": 0.43482100615044944, + "grad_norm": 1.7381259202957153, + "learning_rate": 1.256919056389913e-05, + "loss": 0.4902, + "step": 13786 + }, + { + "epoch": 0.43485254691689007, + "grad_norm": 1.7577447891235352, + "learning_rate": 1.2568203284674764e-05, + "loss": 0.4813, + "step": 13787 + }, + { + "epoch": 0.4348840876833307, + "grad_norm": 1.5176969766616821, + "learning_rate": 1.2567215978649206e-05, + "loss": 0.4996, + "step": 13788 + }, + { + "epoch": 0.43491562844977133, + "grad_norm": 1.490081548690796, + "learning_rate": 1.2566228645832754e-05, + "loss": 0.4854, + "step": 13789 + }, + { + "epoch": 0.43494716921621196, + "grad_norm": 1.693360447883606, + "learning_rate": 1.2565241286235713e-05, + "loss": 0.5065, + "step": 13790 + }, + { + "epoch": 0.4349787099826526, + "grad_norm": 1.4941684007644653, + "learning_rate": 1.2564253899868388e-05, + "loss": 0.5161, + "step": 13791 + }, + { + "epoch": 0.4350102507490932, + "grad_norm": 1.4677916765213013, + "learning_rate": 1.2563266486741082e-05, + "loss": 0.5092, + "step": 13792 + }, + { + "epoch": 0.43504179151553385, + "grad_norm": 1.772956132888794, + "learning_rate": 1.25622790468641e-05, + "loss": 0.4813, + "step": 13793 + }, + { + "epoch": 0.4350733322819745, + "grad_norm": 2.0160553455352783, + "learning_rate": 1.2561291580247747e-05, + "loss": 0.5293, + "step": 13794 + }, + { + "epoch": 0.43510487304841505, + "grad_norm": 1.4906468391418457, + "learning_rate": 1.2560304086902324e-05, + "loss": 0.5009, + "step": 13795 + }, + { + "epoch": 0.4351364138148557, + "grad_norm": 1.5514806509017944, + "learning_rate": 1.2559316566838144e-05, + "loss": 0.499, + "step": 13796 + }, + { + "epoch": 0.4351679545812963, + "grad_norm": 1.67100989818573, + "learning_rate": 1.2558329020065505e-05, + "loss": 0.8322, + "step": 13797 + }, + { + "epoch": 0.43519949534773694, + "grad_norm": 2.0600788593292236, + "learning_rate": 1.255734144659472e-05, + "loss": 0.5047, + "step": 13798 + }, + { + "epoch": 0.43523103611417757, + "grad_norm": 2.088486433029175, + "learning_rate": 1.2556353846436091e-05, + "loss": 0.4002, + "step": 13799 + }, + { + "epoch": 0.4352625768806182, + "grad_norm": 1.260780930519104, + "learning_rate": 1.2555366219599923e-05, + "loss": 0.4795, + "step": 13800 + }, + { + "epoch": 0.43529411764705883, + "grad_norm": 1.3374381065368652, + "learning_rate": 1.2554378566096526e-05, + "loss": 0.466, + "step": 13801 + }, + { + "epoch": 0.43532565841349946, + "grad_norm": 1.9207907915115356, + "learning_rate": 1.2553390885936203e-05, + "loss": 0.511, + "step": 13802 + }, + { + "epoch": 0.4353571991799401, + "grad_norm": 1.6761891841888428, + "learning_rate": 1.2552403179129262e-05, + "loss": 0.4692, + "step": 13803 + }, + { + "epoch": 0.4353887399463807, + "grad_norm": 1.509732961654663, + "learning_rate": 1.2551415445686015e-05, + "loss": 0.7939, + "step": 13804 + }, + { + "epoch": 0.4354202807128213, + "grad_norm": 1.2549124956130981, + "learning_rate": 1.2550427685616767e-05, + "loss": 0.5003, + "step": 13805 + }, + { + "epoch": 0.4354518214792619, + "grad_norm": 1.6600974798202515, + "learning_rate": 1.2549439898931824e-05, + "loss": 0.4903, + "step": 13806 + }, + { + "epoch": 0.43548336224570255, + "grad_norm": 1.2493818998336792, + "learning_rate": 1.2548452085641494e-05, + "loss": 0.4856, + "step": 13807 + }, + { + "epoch": 0.4355149030121432, + "grad_norm": 4.386978626251221, + "learning_rate": 1.2547464245756094e-05, + "loss": 0.5168, + "step": 13808 + }, + { + "epoch": 0.4355464437785838, + "grad_norm": 1.353095531463623, + "learning_rate": 1.2546476379285921e-05, + "loss": 0.4892, + "step": 13809 + }, + { + "epoch": 0.43557798454502444, + "grad_norm": 1.9250900745391846, + "learning_rate": 1.2545488486241292e-05, + "loss": 0.4983, + "step": 13810 + }, + { + "epoch": 0.4356095253114651, + "grad_norm": 2.339336633682251, + "learning_rate": 1.2544500566632514e-05, + "loss": 0.5006, + "step": 13811 + }, + { + "epoch": 0.4356410660779057, + "grad_norm": 1.9782545566558838, + "learning_rate": 1.2543512620469898e-05, + "loss": 0.4285, + "step": 13812 + }, + { + "epoch": 0.43567260684434633, + "grad_norm": 1.423501968383789, + "learning_rate": 1.254252464776375e-05, + "loss": 0.4774, + "step": 13813 + }, + { + "epoch": 0.43570414761078696, + "grad_norm": 1.5582408905029297, + "learning_rate": 1.2541536648524381e-05, + "loss": 0.488, + "step": 13814 + }, + { + "epoch": 0.4357356883772276, + "grad_norm": 2.4974772930145264, + "learning_rate": 1.254054862276211e-05, + "loss": 0.5035, + "step": 13815 + }, + { + "epoch": 0.43576722914366817, + "grad_norm": 1.5126067399978638, + "learning_rate": 1.2539560570487237e-05, + "loss": 0.4823, + "step": 13816 + }, + { + "epoch": 0.4357987699101088, + "grad_norm": 1.95383882522583, + "learning_rate": 1.2538572491710079e-05, + "loss": 0.474, + "step": 13817 + }, + { + "epoch": 0.4358303106765494, + "grad_norm": 1.6418179273605347, + "learning_rate": 1.2537584386440948e-05, + "loss": 0.5063, + "step": 13818 + }, + { + "epoch": 0.43586185144299006, + "grad_norm": 1.6587353944778442, + "learning_rate": 1.2536596254690153e-05, + "loss": 0.5102, + "step": 13819 + }, + { + "epoch": 0.4358933922094307, + "grad_norm": 1.4125480651855469, + "learning_rate": 1.2535608096468004e-05, + "loss": 0.4726, + "step": 13820 + }, + { + "epoch": 0.4359249329758713, + "grad_norm": 1.7470101118087769, + "learning_rate": 1.2534619911784815e-05, + "loss": 0.4903, + "step": 13821 + }, + { + "epoch": 0.43595647374231195, + "grad_norm": 1.7178128957748413, + "learning_rate": 1.2533631700650902e-05, + "loss": 0.5213, + "step": 13822 + }, + { + "epoch": 0.4359880145087526, + "grad_norm": 1.8003281354904175, + "learning_rate": 1.2532643463076573e-05, + "loss": 0.49, + "step": 13823 + }, + { + "epoch": 0.4360195552751932, + "grad_norm": 1.8799940347671509, + "learning_rate": 1.2531655199072144e-05, + "loss": 0.4324, + "step": 13824 + }, + { + "epoch": 0.43605109604163383, + "grad_norm": 1.7094087600708008, + "learning_rate": 1.2530666908647927e-05, + "loss": 0.485, + "step": 13825 + }, + { + "epoch": 0.4360826368080744, + "grad_norm": 1.611897349357605, + "learning_rate": 1.2529678591814234e-05, + "loss": 0.5229, + "step": 13826 + }, + { + "epoch": 0.43611417757451504, + "grad_norm": 1.6637754440307617, + "learning_rate": 1.2528690248581382e-05, + "loss": 0.7962, + "step": 13827 + }, + { + "epoch": 0.43614571834095567, + "grad_norm": 2.4844415187835693, + "learning_rate": 1.2527701878959686e-05, + "loss": 0.5031, + "step": 13828 + }, + { + "epoch": 0.4361772591073963, + "grad_norm": 1.7233402729034424, + "learning_rate": 1.2526713482959456e-05, + "loss": 0.4731, + "step": 13829 + }, + { + "epoch": 0.4362087998738369, + "grad_norm": 1.954482913017273, + "learning_rate": 1.2525725060591007e-05, + "loss": 0.4914, + "step": 13830 + }, + { + "epoch": 0.43624034064027756, + "grad_norm": 1.5870088338851929, + "learning_rate": 1.2524736611864658e-05, + "loss": 0.5106, + "step": 13831 + }, + { + "epoch": 0.4362718814067182, + "grad_norm": 1.8077292442321777, + "learning_rate": 1.2523748136790723e-05, + "loss": 0.4767, + "step": 13832 + }, + { + "epoch": 0.4363034221731588, + "grad_norm": 1.7853530645370483, + "learning_rate": 1.2522759635379515e-05, + "loss": 0.53, + "step": 13833 + }, + { + "epoch": 0.43633496293959945, + "grad_norm": 1.6750731468200684, + "learning_rate": 1.2521771107641352e-05, + "loss": 0.5209, + "step": 13834 + }, + { + "epoch": 0.4363665037060401, + "grad_norm": 1.5724163055419922, + "learning_rate": 1.2520782553586548e-05, + "loss": 0.5149, + "step": 13835 + }, + { + "epoch": 0.4363980444724807, + "grad_norm": 2.285371780395508, + "learning_rate": 1.2519793973225423e-05, + "loss": 0.5044, + "step": 13836 + }, + { + "epoch": 0.4364295852389213, + "grad_norm": 1.6253920793533325, + "learning_rate": 1.251880536656829e-05, + "loss": 0.4702, + "step": 13837 + }, + { + "epoch": 0.4364611260053619, + "grad_norm": 1.537713885307312, + "learning_rate": 1.2517816733625467e-05, + "loss": 0.5371, + "step": 13838 + }, + { + "epoch": 0.43649266677180254, + "grad_norm": 1.4678559303283691, + "learning_rate": 1.2516828074407274e-05, + "loss": 0.4727, + "step": 13839 + }, + { + "epoch": 0.43652420753824317, + "grad_norm": 1.3819119930267334, + "learning_rate": 1.2515839388924022e-05, + "loss": 0.5017, + "step": 13840 + }, + { + "epoch": 0.4365557483046838, + "grad_norm": 1.8705205917358398, + "learning_rate": 1.2514850677186034e-05, + "loss": 0.3465, + "step": 13841 + }, + { + "epoch": 0.43658728907112443, + "grad_norm": 2.879845380783081, + "learning_rate": 1.2513861939203626e-05, + "loss": 0.3755, + "step": 13842 + }, + { + "epoch": 0.43661882983756506, + "grad_norm": 2.4103646278381348, + "learning_rate": 1.2512873174987116e-05, + "loss": 0.4951, + "step": 13843 + }, + { + "epoch": 0.4366503706040057, + "grad_norm": 1.936584711074829, + "learning_rate": 1.2511884384546825e-05, + "loss": 0.4685, + "step": 13844 + }, + { + "epoch": 0.4366819113704463, + "grad_norm": 1.452273964881897, + "learning_rate": 1.2510895567893069e-05, + "loss": 0.479, + "step": 13845 + }, + { + "epoch": 0.43671345213688695, + "grad_norm": 1.3740640878677368, + "learning_rate": 1.250990672503617e-05, + "loss": 0.4892, + "step": 13846 + }, + { + "epoch": 0.4367449929033276, + "grad_norm": 1.6719334125518799, + "learning_rate": 1.2508917855986443e-05, + "loss": 0.4734, + "step": 13847 + }, + { + "epoch": 0.43677653366976815, + "grad_norm": 1.8020739555358887, + "learning_rate": 1.2507928960754213e-05, + "loss": 0.5112, + "step": 13848 + }, + { + "epoch": 0.4368080744362088, + "grad_norm": 1.194505214691162, + "learning_rate": 1.2506940039349795e-05, + "loss": 0.4941, + "step": 13849 + }, + { + "epoch": 0.4368396152026494, + "grad_norm": 1.382230520248413, + "learning_rate": 1.250595109178351e-05, + "loss": 0.4625, + "step": 13850 + }, + { + "epoch": 0.43687115596909004, + "grad_norm": 1.748012661933899, + "learning_rate": 1.250496211806568e-05, + "loss": 0.8304, + "step": 13851 + }, + { + "epoch": 0.43690269673553067, + "grad_norm": 1.5132545232772827, + "learning_rate": 1.2503973118206625e-05, + "loss": 0.4917, + "step": 13852 + }, + { + "epoch": 0.4369342375019713, + "grad_norm": 1.6044384241104126, + "learning_rate": 1.2502984092216668e-05, + "loss": 0.5114, + "step": 13853 + }, + { + "epoch": 0.43696577826841193, + "grad_norm": 1.6493991613388062, + "learning_rate": 1.2501995040106127e-05, + "loss": 0.4653, + "step": 13854 + }, + { + "epoch": 0.43699731903485256, + "grad_norm": 1.2544927597045898, + "learning_rate": 1.2501005961885324e-05, + "loss": 0.4868, + "step": 13855 + }, + { + "epoch": 0.4370288598012932, + "grad_norm": 1.511592984199524, + "learning_rate": 1.2500016857564585e-05, + "loss": 0.4557, + "step": 13856 + }, + { + "epoch": 0.4370604005677338, + "grad_norm": 1.6550804376602173, + "learning_rate": 1.2499027727154226e-05, + "loss": 0.5468, + "step": 13857 + }, + { + "epoch": 0.4370919413341744, + "grad_norm": 1.6748745441436768, + "learning_rate": 1.2498038570664574e-05, + "loss": 0.4631, + "step": 13858 + }, + { + "epoch": 0.437123482100615, + "grad_norm": 1.7570784091949463, + "learning_rate": 1.2497049388105948e-05, + "loss": 0.8011, + "step": 13859 + }, + { + "epoch": 0.43715502286705565, + "grad_norm": 1.4711651802062988, + "learning_rate": 1.2496060179488676e-05, + "loss": 0.4651, + "step": 13860 + }, + { + "epoch": 0.4371865636334963, + "grad_norm": 1.4734972715377808, + "learning_rate": 1.2495070944823075e-05, + "loss": 0.5192, + "step": 13861 + }, + { + "epoch": 0.4372181043999369, + "grad_norm": 1.8500646352767944, + "learning_rate": 1.2494081684119473e-05, + "loss": 0.5039, + "step": 13862 + }, + { + "epoch": 0.43724964516637754, + "grad_norm": 1.5663195848464966, + "learning_rate": 1.2493092397388191e-05, + "loss": 0.4627, + "step": 13863 + }, + { + "epoch": 0.4372811859328182, + "grad_norm": 1.8825550079345703, + "learning_rate": 1.2492103084639552e-05, + "loss": 0.808, + "step": 13864 + }, + { + "epoch": 0.4373127266992588, + "grad_norm": 1.5728886127471924, + "learning_rate": 1.2491113745883885e-05, + "loss": 0.4558, + "step": 13865 + }, + { + "epoch": 0.43734426746569943, + "grad_norm": 1.4296174049377441, + "learning_rate": 1.2490124381131512e-05, + "loss": 0.4951, + "step": 13866 + }, + { + "epoch": 0.43737580823214006, + "grad_norm": 1.7162160873413086, + "learning_rate": 1.2489134990392757e-05, + "loss": 0.5082, + "step": 13867 + }, + { + "epoch": 0.4374073489985807, + "grad_norm": 1.966741919517517, + "learning_rate": 1.2488145573677946e-05, + "loss": 0.5192, + "step": 13868 + }, + { + "epoch": 0.43743888976502127, + "grad_norm": 1.579530954360962, + "learning_rate": 1.2487156130997402e-05, + "loss": 0.4894, + "step": 13869 + }, + { + "epoch": 0.4374704305314619, + "grad_norm": 1.609321117401123, + "learning_rate": 1.2486166662361456e-05, + "loss": 0.7647, + "step": 13870 + }, + { + "epoch": 0.4375019712979025, + "grad_norm": 1.452290415763855, + "learning_rate": 1.2485177167780428e-05, + "loss": 0.4812, + "step": 13871 + }, + { + "epoch": 0.43753351206434316, + "grad_norm": 1.471225380897522, + "learning_rate": 1.2484187647264646e-05, + "loss": 0.8036, + "step": 13872 + }, + { + "epoch": 0.4375650528307838, + "grad_norm": 1.4472105503082275, + "learning_rate": 1.2483198100824438e-05, + "loss": 0.5136, + "step": 13873 + }, + { + "epoch": 0.4375965935972244, + "grad_norm": 1.7519989013671875, + "learning_rate": 1.248220852847013e-05, + "loss": 0.4478, + "step": 13874 + }, + { + "epoch": 0.43762813436366504, + "grad_norm": 1.8434293270111084, + "learning_rate": 1.2481218930212046e-05, + "loss": 0.509, + "step": 13875 + }, + { + "epoch": 0.4376596751301057, + "grad_norm": 1.926202416419983, + "learning_rate": 1.2480229306060518e-05, + "loss": 0.5186, + "step": 13876 + }, + { + "epoch": 0.4376912158965463, + "grad_norm": 1.5670452117919922, + "learning_rate": 1.2479239656025872e-05, + "loss": 0.5062, + "step": 13877 + }, + { + "epoch": 0.43772275666298693, + "grad_norm": 1.9390430450439453, + "learning_rate": 1.2478249980118435e-05, + "loss": 0.4883, + "step": 13878 + }, + { + "epoch": 0.4377542974294275, + "grad_norm": 1.794950246810913, + "learning_rate": 1.2477260278348535e-05, + "loss": 0.4949, + "step": 13879 + }, + { + "epoch": 0.43778583819586814, + "grad_norm": 1.4361510276794434, + "learning_rate": 1.2476270550726501e-05, + "loss": 0.4813, + "step": 13880 + }, + { + "epoch": 0.43781737896230877, + "grad_norm": 1.900084376335144, + "learning_rate": 1.247528079726266e-05, + "loss": 0.8068, + "step": 13881 + }, + { + "epoch": 0.4378489197287494, + "grad_norm": 1.3615964651107788, + "learning_rate": 1.247429101796734e-05, + "loss": 0.4928, + "step": 13882 + }, + { + "epoch": 0.43788046049519, + "grad_norm": 1.9625746011734009, + "learning_rate": 1.2473301212850873e-05, + "loss": 0.477, + "step": 13883 + }, + { + "epoch": 0.43791200126163066, + "grad_norm": 3.18259859085083, + "learning_rate": 1.247231138192359e-05, + "loss": 0.3666, + "step": 13884 + }, + { + "epoch": 0.4379435420280713, + "grad_norm": 1.7577968835830688, + "learning_rate": 1.2471321525195816e-05, + "loss": 0.4973, + "step": 13885 + }, + { + "epoch": 0.4379750827945119, + "grad_norm": 1.4145623445510864, + "learning_rate": 1.2470331642677881e-05, + "loss": 0.5106, + "step": 13886 + }, + { + "epoch": 0.43800662356095255, + "grad_norm": 1.499454379081726, + "learning_rate": 1.246934173438012e-05, + "loss": 0.5167, + "step": 13887 + }, + { + "epoch": 0.4380381643273932, + "grad_norm": 2.488793134689331, + "learning_rate": 1.2468351800312863e-05, + "loss": 0.4688, + "step": 13888 + }, + { + "epoch": 0.4380697050938338, + "grad_norm": 1.5122014284133911, + "learning_rate": 1.2467361840486433e-05, + "loss": 0.5035, + "step": 13889 + }, + { + "epoch": 0.4381012458602744, + "grad_norm": 1.5147438049316406, + "learning_rate": 1.246637185491117e-05, + "loss": 0.5092, + "step": 13890 + }, + { + "epoch": 0.438132786626715, + "grad_norm": 1.3835963010787964, + "learning_rate": 1.2465381843597401e-05, + "loss": 0.5301, + "step": 13891 + }, + { + "epoch": 0.43816432739315564, + "grad_norm": 1.5963619947433472, + "learning_rate": 1.2464391806555456e-05, + "loss": 0.4878, + "step": 13892 + }, + { + "epoch": 0.43819586815959627, + "grad_norm": 1.750197410583496, + "learning_rate": 1.246340174379567e-05, + "loss": 0.4921, + "step": 13893 + }, + { + "epoch": 0.4382274089260369, + "grad_norm": 1.7706084251403809, + "learning_rate": 1.2462411655328374e-05, + "loss": 0.506, + "step": 13894 + }, + { + "epoch": 0.43825894969247753, + "grad_norm": 1.5722830295562744, + "learning_rate": 1.24614215411639e-05, + "loss": 0.3549, + "step": 13895 + }, + { + "epoch": 0.43829049045891816, + "grad_norm": 1.6256794929504395, + "learning_rate": 1.2460431401312582e-05, + "loss": 0.4808, + "step": 13896 + }, + { + "epoch": 0.4383220312253588, + "grad_norm": 2.0452053546905518, + "learning_rate": 1.2459441235784753e-05, + "loss": 0.383, + "step": 13897 + }, + { + "epoch": 0.4383535719917994, + "grad_norm": 1.8055533170700073, + "learning_rate": 1.2458451044590745e-05, + "loss": 0.428, + "step": 13898 + }, + { + "epoch": 0.43838511275824005, + "grad_norm": 1.5330159664154053, + "learning_rate": 1.2457460827740889e-05, + "loss": 0.5229, + "step": 13899 + }, + { + "epoch": 0.4384166535246807, + "grad_norm": 1.6359515190124512, + "learning_rate": 1.2456470585245522e-05, + "loss": 0.4716, + "step": 13900 + }, + { + "epoch": 0.43844819429112125, + "grad_norm": 1.4129879474639893, + "learning_rate": 1.2455480317114978e-05, + "loss": 0.3984, + "step": 13901 + }, + { + "epoch": 0.4384797350575619, + "grad_norm": 1.7335600852966309, + "learning_rate": 1.2454490023359588e-05, + "loss": 0.5248, + "step": 13902 + }, + { + "epoch": 0.4385112758240025, + "grad_norm": 1.7388253211975098, + "learning_rate": 1.2453499703989688e-05, + "loss": 0.4996, + "step": 13903 + }, + { + "epoch": 0.43854281659044314, + "grad_norm": 1.7448738813400269, + "learning_rate": 1.2452509359015619e-05, + "loss": 0.5177, + "step": 13904 + }, + { + "epoch": 0.43857435735688377, + "grad_norm": 1.528017520904541, + "learning_rate": 1.2451518988447707e-05, + "loss": 0.8185, + "step": 13905 + }, + { + "epoch": 0.4386058981233244, + "grad_norm": 1.5363187789916992, + "learning_rate": 1.245052859229629e-05, + "loss": 0.4928, + "step": 13906 + }, + { + "epoch": 0.43863743888976503, + "grad_norm": 1.6164705753326416, + "learning_rate": 1.2449538170571706e-05, + "loss": 0.503, + "step": 13907 + }, + { + "epoch": 0.43866897965620566, + "grad_norm": 2.136929512023926, + "learning_rate": 1.244854772328429e-05, + "loss": 0.5018, + "step": 13908 + }, + { + "epoch": 0.4387005204226463, + "grad_norm": 1.494466781616211, + "learning_rate": 1.2447557250444376e-05, + "loss": 0.8172, + "step": 13909 + }, + { + "epoch": 0.4387320611890869, + "grad_norm": 1.4205573797225952, + "learning_rate": 1.24465667520623e-05, + "loss": 0.5026, + "step": 13910 + }, + { + "epoch": 0.4387636019555275, + "grad_norm": 2.010143995285034, + "learning_rate": 1.2445576228148402e-05, + "loss": 0.4661, + "step": 13911 + }, + { + "epoch": 0.4387951427219681, + "grad_norm": 1.601854920387268, + "learning_rate": 1.2444585678713015e-05, + "loss": 0.4936, + "step": 13912 + }, + { + "epoch": 0.43882668348840875, + "grad_norm": 1.3976750373840332, + "learning_rate": 1.244359510376648e-05, + "loss": 0.4955, + "step": 13913 + }, + { + "epoch": 0.4388582242548494, + "grad_norm": 1.862410545349121, + "learning_rate": 1.2442604503319132e-05, + "loss": 0.4064, + "step": 13914 + }, + { + "epoch": 0.43888976502129, + "grad_norm": 1.8670793771743774, + "learning_rate": 1.2441613877381311e-05, + "loss": 0.3979, + "step": 13915 + }, + { + "epoch": 0.43892130578773064, + "grad_norm": 1.4742463827133179, + "learning_rate": 1.244062322596335e-05, + "loss": 0.4605, + "step": 13916 + }, + { + "epoch": 0.43895284655417127, + "grad_norm": 1.51496422290802, + "learning_rate": 1.2439632549075596e-05, + "loss": 0.509, + "step": 13917 + }, + { + "epoch": 0.4389843873206119, + "grad_norm": 1.6182395219802856, + "learning_rate": 1.2438641846728378e-05, + "loss": 0.5058, + "step": 13918 + }, + { + "epoch": 0.43901592808705253, + "grad_norm": 1.8547166585922241, + "learning_rate": 1.2437651118932039e-05, + "loss": 0.491, + "step": 13919 + }, + { + "epoch": 0.43904746885349316, + "grad_norm": 1.3107620477676392, + "learning_rate": 1.243666036569692e-05, + "loss": 0.5091, + "step": 13920 + }, + { + "epoch": 0.4390790096199338, + "grad_norm": 1.4066025018692017, + "learning_rate": 1.2435669587033357e-05, + "loss": 0.5086, + "step": 13921 + }, + { + "epoch": 0.43911055038637437, + "grad_norm": 1.4783769845962524, + "learning_rate": 1.243467878295169e-05, + "loss": 0.4806, + "step": 13922 + }, + { + "epoch": 0.439142091152815, + "grad_norm": 1.2145205736160278, + "learning_rate": 1.2433687953462256e-05, + "loss": 0.4855, + "step": 13923 + }, + { + "epoch": 0.4391736319192556, + "grad_norm": 1.2752165794372559, + "learning_rate": 1.2432697098575403e-05, + "loss": 0.4864, + "step": 13924 + }, + { + "epoch": 0.43920517268569625, + "grad_norm": 1.2279044389724731, + "learning_rate": 1.2431706218301468e-05, + "loss": 0.4903, + "step": 13925 + }, + { + "epoch": 0.4392367134521369, + "grad_norm": 2.2582271099090576, + "learning_rate": 1.2430715312650788e-05, + "loss": 0.4636, + "step": 13926 + }, + { + "epoch": 0.4392682542185775, + "grad_norm": 1.6698555946350098, + "learning_rate": 1.2429724381633708e-05, + "loss": 0.7952, + "step": 13927 + }, + { + "epoch": 0.43929979498501814, + "grad_norm": 1.4759989976882935, + "learning_rate": 1.2428733425260566e-05, + "loss": 0.4896, + "step": 13928 + }, + { + "epoch": 0.4393313357514588, + "grad_norm": 2.590949058532715, + "learning_rate": 1.2427742443541704e-05, + "loss": 0.4534, + "step": 13929 + }, + { + "epoch": 0.4393628765178994, + "grad_norm": 2.054206132888794, + "learning_rate": 1.2426751436487466e-05, + "loss": 0.484, + "step": 13930 + }, + { + "epoch": 0.43939441728434003, + "grad_norm": 1.7213929891586304, + "learning_rate": 1.2425760404108189e-05, + "loss": 0.4309, + "step": 13931 + }, + { + "epoch": 0.4394259580507806, + "grad_norm": 1.5361878871917725, + "learning_rate": 1.2424769346414224e-05, + "loss": 0.5123, + "step": 13932 + }, + { + "epoch": 0.43945749881722124, + "grad_norm": 1.5630106925964355, + "learning_rate": 1.2423778263415904e-05, + "loss": 0.508, + "step": 13933 + }, + { + "epoch": 0.43948903958366187, + "grad_norm": 1.4998102188110352, + "learning_rate": 1.2422787155123576e-05, + "loss": 0.4942, + "step": 13934 + }, + { + "epoch": 0.4395205803501025, + "grad_norm": 1.3850021362304688, + "learning_rate": 1.2421796021547587e-05, + "loss": 0.4755, + "step": 13935 + }, + { + "epoch": 0.4395521211165431, + "grad_norm": 1.8582708835601807, + "learning_rate": 1.2420804862698272e-05, + "loss": 0.5056, + "step": 13936 + }, + { + "epoch": 0.43958366188298376, + "grad_norm": 1.82086181640625, + "learning_rate": 1.2419813678585982e-05, + "loss": 0.5023, + "step": 13937 + }, + { + "epoch": 0.4396152026494244, + "grad_norm": 1.3566375970840454, + "learning_rate": 1.2418822469221054e-05, + "loss": 0.5057, + "step": 13938 + }, + { + "epoch": 0.439646743415865, + "grad_norm": 1.464114785194397, + "learning_rate": 1.2417831234613837e-05, + "loss": 0.4916, + "step": 13939 + }, + { + "epoch": 0.43967828418230565, + "grad_norm": 1.4511160850524902, + "learning_rate": 1.2416839974774673e-05, + "loss": 0.3627, + "step": 13940 + }, + { + "epoch": 0.4397098249487463, + "grad_norm": 1.529083013534546, + "learning_rate": 1.2415848689713904e-05, + "loss": 0.5127, + "step": 13941 + }, + { + "epoch": 0.4397413657151869, + "grad_norm": 1.261605978012085, + "learning_rate": 1.2414857379441884e-05, + "loss": 0.4766, + "step": 13942 + }, + { + "epoch": 0.4397729064816275, + "grad_norm": 1.5142170190811157, + "learning_rate": 1.241386604396895e-05, + "loss": 0.8037, + "step": 13943 + }, + { + "epoch": 0.4398044472480681, + "grad_norm": 1.6289722919464111, + "learning_rate": 1.2412874683305448e-05, + "loss": 0.495, + "step": 13944 + }, + { + "epoch": 0.43983598801450874, + "grad_norm": 1.52903413772583, + "learning_rate": 1.2411883297461726e-05, + "loss": 0.4773, + "step": 13945 + }, + { + "epoch": 0.43986752878094937, + "grad_norm": 1.3583678007125854, + "learning_rate": 1.2410891886448128e-05, + "loss": 0.4682, + "step": 13946 + }, + { + "epoch": 0.43989906954739, + "grad_norm": 1.712276816368103, + "learning_rate": 1.2409900450275005e-05, + "loss": 0.4547, + "step": 13947 + }, + { + "epoch": 0.43993061031383063, + "grad_norm": 1.4444218873977661, + "learning_rate": 1.2408908988952697e-05, + "loss": 0.3528, + "step": 13948 + }, + { + "epoch": 0.43996215108027126, + "grad_norm": 1.6307803392410278, + "learning_rate": 1.2407917502491556e-05, + "loss": 0.4955, + "step": 13949 + }, + { + "epoch": 0.4399936918467119, + "grad_norm": 1.5393463373184204, + "learning_rate": 1.2406925990901922e-05, + "loss": 0.4444, + "step": 13950 + }, + { + "epoch": 0.4400252326131525, + "grad_norm": 2.645376205444336, + "learning_rate": 1.2405934454194146e-05, + "loss": 0.5194, + "step": 13951 + }, + { + "epoch": 0.44005677337959315, + "grad_norm": 1.5904656648635864, + "learning_rate": 1.2404942892378577e-05, + "loss": 0.4678, + "step": 13952 + }, + { + "epoch": 0.4400883141460337, + "grad_norm": 2.0081727504730225, + "learning_rate": 1.2403951305465562e-05, + "loss": 0.8166, + "step": 13953 + }, + { + "epoch": 0.44011985491247435, + "grad_norm": 1.6361405849456787, + "learning_rate": 1.2402959693465448e-05, + "loss": 0.4785, + "step": 13954 + }, + { + "epoch": 0.440151395678915, + "grad_norm": 1.879804015159607, + "learning_rate": 1.2401968056388588e-05, + "loss": 0.4824, + "step": 13955 + }, + { + "epoch": 0.4401829364453556, + "grad_norm": 1.584359884262085, + "learning_rate": 1.240097639424532e-05, + "loss": 0.4956, + "step": 13956 + }, + { + "epoch": 0.44021447721179624, + "grad_norm": 1.3953875303268433, + "learning_rate": 1.2399984707046002e-05, + "loss": 0.4888, + "step": 13957 + }, + { + "epoch": 0.44024601797823687, + "grad_norm": 1.7908855676651, + "learning_rate": 1.239899299480098e-05, + "loss": 0.5063, + "step": 13958 + }, + { + "epoch": 0.4402775587446775, + "grad_norm": 1.304500699043274, + "learning_rate": 1.2398001257520605e-05, + "loss": 0.4704, + "step": 13959 + }, + { + "epoch": 0.44030909951111813, + "grad_norm": 1.9245902299880981, + "learning_rate": 1.239700949521522e-05, + "loss": 0.4548, + "step": 13960 + }, + { + "epoch": 0.44034064027755876, + "grad_norm": 1.9371724128723145, + "learning_rate": 1.2396017707895183e-05, + "loss": 0.3897, + "step": 13961 + }, + { + "epoch": 0.4403721810439994, + "grad_norm": 2.381657838821411, + "learning_rate": 1.2395025895570838e-05, + "loss": 0.4831, + "step": 13962 + }, + { + "epoch": 0.44040372181044, + "grad_norm": 1.5660477876663208, + "learning_rate": 1.2394034058252543e-05, + "loss": 0.799, + "step": 13963 + }, + { + "epoch": 0.4404352625768806, + "grad_norm": 1.4274210929870605, + "learning_rate": 1.239304219595064e-05, + "loss": 0.4993, + "step": 13964 + }, + { + "epoch": 0.4404668033433212, + "grad_norm": 1.5649197101593018, + "learning_rate": 1.2392050308675485e-05, + "loss": 0.5405, + "step": 13965 + }, + { + "epoch": 0.44049834410976185, + "grad_norm": 1.4674526453018188, + "learning_rate": 1.2391058396437427e-05, + "loss": 0.4876, + "step": 13966 + }, + { + "epoch": 0.4405298848762025, + "grad_norm": 1.487663745880127, + "learning_rate": 1.239006645924682e-05, + "loss": 0.5431, + "step": 13967 + }, + { + "epoch": 0.4405614256426431, + "grad_norm": 1.5687580108642578, + "learning_rate": 1.2389074497114011e-05, + "loss": 0.5045, + "step": 13968 + }, + { + "epoch": 0.44059296640908374, + "grad_norm": 1.579293131828308, + "learning_rate": 1.2388082510049358e-05, + "loss": 0.5109, + "step": 13969 + }, + { + "epoch": 0.44062450717552437, + "grad_norm": 1.4035645723342896, + "learning_rate": 1.2387090498063206e-05, + "loss": 0.4613, + "step": 13970 + }, + { + "epoch": 0.440656047941965, + "grad_norm": 1.7366498708724976, + "learning_rate": 1.238609846116591e-05, + "loss": 0.4887, + "step": 13971 + }, + { + "epoch": 0.44068758870840563, + "grad_norm": 1.7679595947265625, + "learning_rate": 1.2385106399367826e-05, + "loss": 0.5154, + "step": 13972 + }, + { + "epoch": 0.44071912947484626, + "grad_norm": 1.4793156385421753, + "learning_rate": 1.2384114312679304e-05, + "loss": 0.485, + "step": 13973 + }, + { + "epoch": 0.4407506702412869, + "grad_norm": 1.8680088520050049, + "learning_rate": 1.23831222011107e-05, + "loss": 0.7839, + "step": 13974 + }, + { + "epoch": 0.44078221100772746, + "grad_norm": 1.6024174690246582, + "learning_rate": 1.2382130064672366e-05, + "loss": 0.5095, + "step": 13975 + }, + { + "epoch": 0.4408137517741681, + "grad_norm": 1.793474555015564, + "learning_rate": 1.2381137903374652e-05, + "loss": 0.5017, + "step": 13976 + }, + { + "epoch": 0.4408452925406087, + "grad_norm": 1.4226042032241821, + "learning_rate": 1.2380145717227918e-05, + "loss": 0.4919, + "step": 13977 + }, + { + "epoch": 0.44087683330704935, + "grad_norm": 2.0278661251068115, + "learning_rate": 1.2379153506242512e-05, + "loss": 0.4035, + "step": 13978 + }, + { + "epoch": 0.44090837407349, + "grad_norm": 1.3629285097122192, + "learning_rate": 1.2378161270428792e-05, + "loss": 0.4886, + "step": 13979 + }, + { + "epoch": 0.4409399148399306, + "grad_norm": 2.337006092071533, + "learning_rate": 1.2377169009797116e-05, + "loss": 0.4694, + "step": 13980 + }, + { + "epoch": 0.44097145560637124, + "grad_norm": 1.5108981132507324, + "learning_rate": 1.2376176724357834e-05, + "loss": 0.5398, + "step": 13981 + }, + { + "epoch": 0.4410029963728119, + "grad_norm": 1.4742918014526367, + "learning_rate": 1.23751844141213e-05, + "loss": 0.4732, + "step": 13982 + }, + { + "epoch": 0.4410345371392525, + "grad_norm": 1.692436933517456, + "learning_rate": 1.2374192079097877e-05, + "loss": 0.4592, + "step": 13983 + }, + { + "epoch": 0.44106607790569313, + "grad_norm": 1.5952528715133667, + "learning_rate": 1.2373199719297913e-05, + "loss": 0.4191, + "step": 13984 + }, + { + "epoch": 0.4410976186721337, + "grad_norm": 1.365927815437317, + "learning_rate": 1.237220733473177e-05, + "loss": 0.4888, + "step": 13985 + }, + { + "epoch": 0.44112915943857434, + "grad_norm": 1.4515951871871948, + "learning_rate": 1.2371214925409798e-05, + "loss": 0.5022, + "step": 13986 + }, + { + "epoch": 0.44116070020501497, + "grad_norm": 1.4620821475982666, + "learning_rate": 1.2370222491342358e-05, + "loss": 0.3595, + "step": 13987 + }, + { + "epoch": 0.4411922409714556, + "grad_norm": 1.5440853834152222, + "learning_rate": 1.2369230032539808e-05, + "loss": 0.4862, + "step": 13988 + }, + { + "epoch": 0.4412237817378962, + "grad_norm": 1.4898755550384521, + "learning_rate": 1.2368237549012498e-05, + "loss": 0.4573, + "step": 13989 + }, + { + "epoch": 0.44125532250433686, + "grad_norm": 4.311883449554443, + "learning_rate": 1.2367245040770796e-05, + "loss": 0.501, + "step": 13990 + }, + { + "epoch": 0.4412868632707775, + "grad_norm": 1.526054859161377, + "learning_rate": 1.2366252507825052e-05, + "loss": 0.5046, + "step": 13991 + }, + { + "epoch": 0.4413184040372181, + "grad_norm": 1.4075418710708618, + "learning_rate": 1.2365259950185624e-05, + "loss": 0.511, + "step": 13992 + }, + { + "epoch": 0.44134994480365874, + "grad_norm": 1.4254189729690552, + "learning_rate": 1.2364267367862874e-05, + "loss": 0.4773, + "step": 13993 + }, + { + "epoch": 0.4413814855700994, + "grad_norm": 1.6513727903366089, + "learning_rate": 1.2363274760867157e-05, + "loss": 0.4951, + "step": 13994 + }, + { + "epoch": 0.44141302633654, + "grad_norm": 1.4051586389541626, + "learning_rate": 1.2362282129208834e-05, + "loss": 0.505, + "step": 13995 + }, + { + "epoch": 0.4414445671029806, + "grad_norm": 1.7151002883911133, + "learning_rate": 1.2361289472898262e-05, + "loss": 0.4797, + "step": 13996 + }, + { + "epoch": 0.4414761078694212, + "grad_norm": 1.6383110284805298, + "learning_rate": 1.23602967919458e-05, + "loss": 0.4827, + "step": 13997 + }, + { + "epoch": 0.44150764863586184, + "grad_norm": 2.554370880126953, + "learning_rate": 1.235930408636181e-05, + "loss": 0.4932, + "step": 13998 + }, + { + "epoch": 0.44153918940230247, + "grad_norm": 1.557448148727417, + "learning_rate": 1.2358311356156648e-05, + "loss": 0.48, + "step": 13999 + }, + { + "epoch": 0.4415707301687431, + "grad_norm": 1.8494007587432861, + "learning_rate": 1.2357318601340677e-05, + "loss": 0.4982, + "step": 14000 + }, + { + "epoch": 0.4416022709351837, + "grad_norm": 1.6733201742172241, + "learning_rate": 1.2356325821924255e-05, + "loss": 0.5184, + "step": 14001 + }, + { + "epoch": 0.44163381170162436, + "grad_norm": 1.4697140455245972, + "learning_rate": 1.2355333017917744e-05, + "loss": 0.8101, + "step": 14002 + }, + { + "epoch": 0.441665352468065, + "grad_norm": 1.673397421836853, + "learning_rate": 1.2354340189331504e-05, + "loss": 0.5471, + "step": 14003 + }, + { + "epoch": 0.4416968932345056, + "grad_norm": 1.7230844497680664, + "learning_rate": 1.2353347336175899e-05, + "loss": 0.4102, + "step": 14004 + }, + { + "epoch": 0.44172843400094625, + "grad_norm": 3.384183645248413, + "learning_rate": 1.2352354458461283e-05, + "loss": 0.5034, + "step": 14005 + }, + { + "epoch": 0.4417599747673868, + "grad_norm": 2.0888898372650146, + "learning_rate": 1.2351361556198024e-05, + "loss": 0.5197, + "step": 14006 + }, + { + "epoch": 0.44179151553382745, + "grad_norm": 1.5618703365325928, + "learning_rate": 1.2350368629396482e-05, + "loss": 0.5094, + "step": 14007 + }, + { + "epoch": 0.4418230563002681, + "grad_norm": 2.4780993461608887, + "learning_rate": 1.2349375678067016e-05, + "loss": 0.517, + "step": 14008 + }, + { + "epoch": 0.4418545970667087, + "grad_norm": 2.4728074073791504, + "learning_rate": 1.234838270221999e-05, + "loss": 0.491, + "step": 14009 + }, + { + "epoch": 0.44188613783314934, + "grad_norm": 1.4737441539764404, + "learning_rate": 1.2347389701865765e-05, + "loss": 0.4616, + "step": 14010 + }, + { + "epoch": 0.44191767859958997, + "grad_norm": 1.5786159038543701, + "learning_rate": 1.2346396677014714e-05, + "loss": 0.4817, + "step": 14011 + }, + { + "epoch": 0.4419492193660306, + "grad_norm": 1.8521230220794678, + "learning_rate": 1.2345403627677186e-05, + "loss": 0.5126, + "step": 14012 + }, + { + "epoch": 0.44198076013247123, + "grad_norm": 2.0478920936584473, + "learning_rate": 1.2344410553863549e-05, + "loss": 0.5366, + "step": 14013 + }, + { + "epoch": 0.44201230089891186, + "grad_norm": 1.949141263961792, + "learning_rate": 1.234341745558417e-05, + "loss": 0.4958, + "step": 14014 + }, + { + "epoch": 0.4420438416653525, + "grad_norm": 1.7377089262008667, + "learning_rate": 1.2342424332849413e-05, + "loss": 0.4992, + "step": 14015 + }, + { + "epoch": 0.4420753824317931, + "grad_norm": 1.4431774616241455, + "learning_rate": 1.2341431185669634e-05, + "loss": 0.4685, + "step": 14016 + }, + { + "epoch": 0.4421069231982337, + "grad_norm": 1.67275071144104, + "learning_rate": 1.2340438014055204e-05, + "loss": 0.435, + "step": 14017 + }, + { + "epoch": 0.4421384639646743, + "grad_norm": 2.0238234996795654, + "learning_rate": 1.2339444818016488e-05, + "loss": 0.5088, + "step": 14018 + }, + { + "epoch": 0.44217000473111495, + "grad_norm": 1.5716711282730103, + "learning_rate": 1.2338451597563845e-05, + "loss": 0.4785, + "step": 14019 + }, + { + "epoch": 0.4422015454975556, + "grad_norm": 1.5291575193405151, + "learning_rate": 1.2337458352707645e-05, + "loss": 0.8264, + "step": 14020 + }, + { + "epoch": 0.4422330862639962, + "grad_norm": 1.4541057348251343, + "learning_rate": 1.2336465083458255e-05, + "loss": 0.4835, + "step": 14021 + }, + { + "epoch": 0.44226462703043684, + "grad_norm": 1.5840859413146973, + "learning_rate": 1.2335471789826034e-05, + "loss": 0.489, + "step": 14022 + }, + { + "epoch": 0.44229616779687747, + "grad_norm": 1.4985724687576294, + "learning_rate": 1.2334478471821353e-05, + "loss": 0.5111, + "step": 14023 + }, + { + "epoch": 0.4423277085633181, + "grad_norm": 1.7782387733459473, + "learning_rate": 1.2333485129454578e-05, + "loss": 0.4914, + "step": 14024 + }, + { + "epoch": 0.44235924932975873, + "grad_norm": 1.6736881732940674, + "learning_rate": 1.2332491762736074e-05, + "loss": 0.4727, + "step": 14025 + }, + { + "epoch": 0.44239079009619936, + "grad_norm": 1.7231676578521729, + "learning_rate": 1.2331498371676206e-05, + "loss": 0.5361, + "step": 14026 + }, + { + "epoch": 0.44242233086264, + "grad_norm": 1.804548740386963, + "learning_rate": 1.2330504956285341e-05, + "loss": 0.4834, + "step": 14027 + }, + { + "epoch": 0.44245387162908056, + "grad_norm": 1.4656909704208374, + "learning_rate": 1.2329511516573849e-05, + "loss": 0.5075, + "step": 14028 + }, + { + "epoch": 0.4424854123955212, + "grad_norm": 1.254981279373169, + "learning_rate": 1.2328518052552094e-05, + "loss": 0.3734, + "step": 14029 + }, + { + "epoch": 0.4425169531619618, + "grad_norm": 1.519076943397522, + "learning_rate": 1.2327524564230445e-05, + "loss": 0.4745, + "step": 14030 + }, + { + "epoch": 0.44254849392840245, + "grad_norm": 1.700996994972229, + "learning_rate": 1.2326531051619273e-05, + "loss": 0.5129, + "step": 14031 + }, + { + "epoch": 0.4425800346948431, + "grad_norm": 1.922683596611023, + "learning_rate": 1.2325537514728937e-05, + "loss": 0.4906, + "step": 14032 + }, + { + "epoch": 0.4426115754612837, + "grad_norm": 1.5235533714294434, + "learning_rate": 1.2324543953569817e-05, + "loss": 0.7955, + "step": 14033 + }, + { + "epoch": 0.44264311622772434, + "grad_norm": 1.6675409078598022, + "learning_rate": 1.2323550368152272e-05, + "loss": 0.4679, + "step": 14034 + }, + { + "epoch": 0.442674656994165, + "grad_norm": 1.6494930982589722, + "learning_rate": 1.2322556758486678e-05, + "loss": 0.5287, + "step": 14035 + }, + { + "epoch": 0.4427061977606056, + "grad_norm": 1.805708646774292, + "learning_rate": 1.23215631245834e-05, + "loss": 0.481, + "step": 14036 + }, + { + "epoch": 0.44273773852704623, + "grad_norm": 1.630626916885376, + "learning_rate": 1.2320569466452805e-05, + "loss": 0.7863, + "step": 14037 + }, + { + "epoch": 0.4427692792934868, + "grad_norm": 1.4868295192718506, + "learning_rate": 1.2319575784105272e-05, + "loss": 0.4915, + "step": 14038 + }, + { + "epoch": 0.44280082005992744, + "grad_norm": 1.3362998962402344, + "learning_rate": 1.2318582077551157e-05, + "loss": 0.4757, + "step": 14039 + }, + { + "epoch": 0.44283236082636807, + "grad_norm": 1.6515032052993774, + "learning_rate": 1.2317588346800841e-05, + "loss": 0.423, + "step": 14040 + }, + { + "epoch": 0.4428639015928087, + "grad_norm": 2.250856876373291, + "learning_rate": 1.2316594591864694e-05, + "loss": 0.5192, + "step": 14041 + }, + { + "epoch": 0.4428954423592493, + "grad_norm": 1.4261395931243896, + "learning_rate": 1.231560081275308e-05, + "loss": 0.4904, + "step": 14042 + }, + { + "epoch": 0.44292698312568995, + "grad_norm": 1.6518733501434326, + "learning_rate": 1.2314607009476374e-05, + "loss": 0.519, + "step": 14043 + }, + { + "epoch": 0.4429585238921306, + "grad_norm": 1.8007502555847168, + "learning_rate": 1.2313613182044949e-05, + "loss": 0.4662, + "step": 14044 + }, + { + "epoch": 0.4429900646585712, + "grad_norm": 1.3643583059310913, + "learning_rate": 1.2312619330469172e-05, + "loss": 0.4739, + "step": 14045 + }, + { + "epoch": 0.44302160542501184, + "grad_norm": 1.8907363414764404, + "learning_rate": 1.2311625454759417e-05, + "loss": 0.5245, + "step": 14046 + }, + { + "epoch": 0.4430531461914525, + "grad_norm": 1.3026057481765747, + "learning_rate": 1.2310631554926053e-05, + "loss": 0.4941, + "step": 14047 + }, + { + "epoch": 0.4430846869578931, + "grad_norm": 1.5027437210083008, + "learning_rate": 1.2309637630979458e-05, + "loss": 0.5075, + "step": 14048 + }, + { + "epoch": 0.4431162277243337, + "grad_norm": 3.3987226486206055, + "learning_rate": 1.2308643682929999e-05, + "loss": 0.4224, + "step": 14049 + }, + { + "epoch": 0.4431477684907743, + "grad_norm": 2.0377559661865234, + "learning_rate": 1.2307649710788051e-05, + "loss": 0.8081, + "step": 14050 + }, + { + "epoch": 0.44317930925721494, + "grad_norm": 1.8220131397247314, + "learning_rate": 1.2306655714563986e-05, + "loss": 0.7962, + "step": 14051 + }, + { + "epoch": 0.44321085002365557, + "grad_norm": 1.488145112991333, + "learning_rate": 1.230566169426818e-05, + "loss": 0.4668, + "step": 14052 + }, + { + "epoch": 0.4432423907900962, + "grad_norm": 1.4390090703964233, + "learning_rate": 1.2304667649911002e-05, + "loss": 0.4979, + "step": 14053 + }, + { + "epoch": 0.4432739315565368, + "grad_norm": 2.1306185722351074, + "learning_rate": 1.2303673581502828e-05, + "loss": 0.3422, + "step": 14054 + }, + { + "epoch": 0.44330547232297746, + "grad_norm": 1.718719244003296, + "learning_rate": 1.2302679489054033e-05, + "loss": 0.5113, + "step": 14055 + }, + { + "epoch": 0.4433370130894181, + "grad_norm": 1.3412976264953613, + "learning_rate": 1.2301685372574986e-05, + "loss": 0.4965, + "step": 14056 + }, + { + "epoch": 0.4433685538558587, + "grad_norm": 1.4272053241729736, + "learning_rate": 1.2300691232076065e-05, + "loss": 0.502, + "step": 14057 + }, + { + "epoch": 0.44340009462229935, + "grad_norm": 1.6311919689178467, + "learning_rate": 1.2299697067567646e-05, + "loss": 0.5202, + "step": 14058 + }, + { + "epoch": 0.4434316353887399, + "grad_norm": 1.3391785621643066, + "learning_rate": 1.2298702879060105e-05, + "loss": 0.4809, + "step": 14059 + }, + { + "epoch": 0.44346317615518055, + "grad_norm": 1.9438509941101074, + "learning_rate": 1.229770866656381e-05, + "loss": 0.507, + "step": 14060 + }, + { + "epoch": 0.4434947169216212, + "grad_norm": 1.8178883790969849, + "learning_rate": 1.2296714430089143e-05, + "loss": 0.8226, + "step": 14061 + }, + { + "epoch": 0.4435262576880618, + "grad_norm": 1.6465859413146973, + "learning_rate": 1.2295720169646479e-05, + "loss": 0.4536, + "step": 14062 + }, + { + "epoch": 0.44355779845450244, + "grad_norm": 1.5399298667907715, + "learning_rate": 1.2294725885246191e-05, + "loss": 0.3889, + "step": 14063 + }, + { + "epoch": 0.44358933922094307, + "grad_norm": 3.00545334815979, + "learning_rate": 1.2293731576898657e-05, + "loss": 0.416, + "step": 14064 + }, + { + "epoch": 0.4436208799873837, + "grad_norm": 1.5066063404083252, + "learning_rate": 1.2292737244614255e-05, + "loss": 0.4782, + "step": 14065 + }, + { + "epoch": 0.44365242075382433, + "grad_norm": 1.8628895282745361, + "learning_rate": 1.2291742888403356e-05, + "loss": 0.4847, + "step": 14066 + }, + { + "epoch": 0.44368396152026496, + "grad_norm": 1.7920007705688477, + "learning_rate": 1.229074850827634e-05, + "loss": 0.519, + "step": 14067 + }, + { + "epoch": 0.4437155022867056, + "grad_norm": 1.645625114440918, + "learning_rate": 1.2289754104243586e-05, + "loss": 0.5015, + "step": 14068 + }, + { + "epoch": 0.4437470430531462, + "grad_norm": 1.622851848602295, + "learning_rate": 1.2288759676315472e-05, + "loss": 0.4094, + "step": 14069 + }, + { + "epoch": 0.4437785838195868, + "grad_norm": 1.6291460990905762, + "learning_rate": 1.228776522450237e-05, + "loss": 0.4749, + "step": 14070 + }, + { + "epoch": 0.4438101245860274, + "grad_norm": 1.3446635007858276, + "learning_rate": 1.2286770748814663e-05, + "loss": 0.4892, + "step": 14071 + }, + { + "epoch": 0.44384166535246805, + "grad_norm": 1.3678172826766968, + "learning_rate": 1.228577624926273e-05, + "loss": 0.502, + "step": 14072 + }, + { + "epoch": 0.4438732061189087, + "grad_norm": 7.933894634246826, + "learning_rate": 1.2284781725856942e-05, + "loss": 0.3485, + "step": 14073 + }, + { + "epoch": 0.4439047468853493, + "grad_norm": 1.4241085052490234, + "learning_rate": 1.2283787178607688e-05, + "loss": 0.4901, + "step": 14074 + }, + { + "epoch": 0.44393628765178994, + "grad_norm": 1.6391916275024414, + "learning_rate": 1.2282792607525339e-05, + "loss": 0.5164, + "step": 14075 + }, + { + "epoch": 0.44396782841823057, + "grad_norm": 1.727675437927246, + "learning_rate": 1.2281798012620277e-05, + "loss": 0.4941, + "step": 14076 + }, + { + "epoch": 0.4439993691846712, + "grad_norm": 1.6742985248565674, + "learning_rate": 1.228080339390288e-05, + "loss": 0.4865, + "step": 14077 + }, + { + "epoch": 0.44403090995111183, + "grad_norm": 1.7109922170639038, + "learning_rate": 1.2279808751383528e-05, + "loss": 0.4856, + "step": 14078 + }, + { + "epoch": 0.44406245071755246, + "grad_norm": 1.8802820444107056, + "learning_rate": 1.2278814085072603e-05, + "loss": 0.4747, + "step": 14079 + }, + { + "epoch": 0.44409399148399303, + "grad_norm": 1.4796653985977173, + "learning_rate": 1.2277819394980484e-05, + "loss": 0.4584, + "step": 14080 + }, + { + "epoch": 0.44412553225043366, + "grad_norm": 1.8859549760818481, + "learning_rate": 1.227682468111755e-05, + "loss": 0.4847, + "step": 14081 + }, + { + "epoch": 0.4441570730168743, + "grad_norm": 1.6547048091888428, + "learning_rate": 1.227582994349418e-05, + "loss": 0.3597, + "step": 14082 + }, + { + "epoch": 0.4441886137833149, + "grad_norm": 1.6736602783203125, + "learning_rate": 1.2274835182120761e-05, + "loss": 0.4722, + "step": 14083 + }, + { + "epoch": 0.44422015454975555, + "grad_norm": 1.669155240058899, + "learning_rate": 1.2273840397007673e-05, + "loss": 0.536, + "step": 14084 + }, + { + "epoch": 0.4442516953161962, + "grad_norm": 1.4802223443984985, + "learning_rate": 1.227284558816529e-05, + "loss": 0.4796, + "step": 14085 + }, + { + "epoch": 0.4442832360826368, + "grad_norm": 1.9127910137176514, + "learning_rate": 1.2271850755604e-05, + "loss": 0.5094, + "step": 14086 + }, + { + "epoch": 0.44431477684907744, + "grad_norm": 1.3621826171875, + "learning_rate": 1.2270855899334182e-05, + "loss": 0.4933, + "step": 14087 + }, + { + "epoch": 0.44434631761551807, + "grad_norm": 1.6607239246368408, + "learning_rate": 1.226986101936622e-05, + "loss": 0.4906, + "step": 14088 + }, + { + "epoch": 0.4443778583819587, + "grad_norm": 1.7255420684814453, + "learning_rate": 1.2268866115710498e-05, + "loss": 0.5239, + "step": 14089 + }, + { + "epoch": 0.44440939914839933, + "grad_norm": 1.3023512363433838, + "learning_rate": 1.2267871188377394e-05, + "loss": 0.4752, + "step": 14090 + }, + { + "epoch": 0.4444409399148399, + "grad_norm": 1.6845393180847168, + "learning_rate": 1.2266876237377294e-05, + "loss": 0.8281, + "step": 14091 + }, + { + "epoch": 0.44447248068128054, + "grad_norm": 2.1221930980682373, + "learning_rate": 1.2265881262720577e-05, + "loss": 0.4777, + "step": 14092 + }, + { + "epoch": 0.44450402144772116, + "grad_norm": 1.7791943550109863, + "learning_rate": 1.2264886264417637e-05, + "loss": 0.4679, + "step": 14093 + }, + { + "epoch": 0.4445355622141618, + "grad_norm": 1.4315769672393799, + "learning_rate": 1.2263891242478846e-05, + "loss": 0.5, + "step": 14094 + }, + { + "epoch": 0.4445671029806024, + "grad_norm": 1.4621332883834839, + "learning_rate": 1.226289619691459e-05, + "loss": 0.496, + "step": 14095 + }, + { + "epoch": 0.44459864374704305, + "grad_norm": 4.2655029296875, + "learning_rate": 1.226190112773526e-05, + "loss": 0.3808, + "step": 14096 + }, + { + "epoch": 0.4446301845134837, + "grad_norm": 1.5226366519927979, + "learning_rate": 1.2260906034951232e-05, + "loss": 0.518, + "step": 14097 + }, + { + "epoch": 0.4446617252799243, + "grad_norm": 1.8644120693206787, + "learning_rate": 1.2259910918572893e-05, + "loss": 0.482, + "step": 14098 + }, + { + "epoch": 0.44469326604636494, + "grad_norm": 1.560972809791565, + "learning_rate": 1.225891577861063e-05, + "loss": 0.4817, + "step": 14099 + }, + { + "epoch": 0.4447248068128056, + "grad_norm": 1.4632123708724976, + "learning_rate": 1.2257920615074828e-05, + "loss": 0.492, + "step": 14100 + }, + { + "epoch": 0.4447563475792462, + "grad_norm": 1.3761987686157227, + "learning_rate": 1.225692542797587e-05, + "loss": 0.7852, + "step": 14101 + }, + { + "epoch": 0.4447878883456868, + "grad_norm": 1.7241487503051758, + "learning_rate": 1.225593021732414e-05, + "loss": 0.8134, + "step": 14102 + }, + { + "epoch": 0.4448194291121274, + "grad_norm": 1.4389538764953613, + "learning_rate": 1.2254934983130034e-05, + "loss": 0.346, + "step": 14103 + }, + { + "epoch": 0.44485096987856804, + "grad_norm": 1.8229358196258545, + "learning_rate": 1.2253939725403927e-05, + "loss": 0.5048, + "step": 14104 + }, + { + "epoch": 0.44488251064500867, + "grad_norm": 1.9594368934631348, + "learning_rate": 1.2252944444156208e-05, + "loss": 0.4887, + "step": 14105 + }, + { + "epoch": 0.4449140514114493, + "grad_norm": 1.5189533233642578, + "learning_rate": 1.2251949139397262e-05, + "loss": 0.4795, + "step": 14106 + }, + { + "epoch": 0.4449455921778899, + "grad_norm": 1.640449047088623, + "learning_rate": 1.225095381113748e-05, + "loss": 0.8023, + "step": 14107 + }, + { + "epoch": 0.44497713294433056, + "grad_norm": 1.7656227350234985, + "learning_rate": 1.2249958459387248e-05, + "loss": 0.4855, + "step": 14108 + }, + { + "epoch": 0.4450086737107712, + "grad_norm": 1.6155962944030762, + "learning_rate": 1.2248963084156951e-05, + "loss": 0.4972, + "step": 14109 + }, + { + "epoch": 0.4450402144772118, + "grad_norm": 1.5017346143722534, + "learning_rate": 1.2247967685456981e-05, + "loss": 0.4818, + "step": 14110 + }, + { + "epoch": 0.44507175524365244, + "grad_norm": 1.5177143812179565, + "learning_rate": 1.2246972263297718e-05, + "loss": 0.4724, + "step": 14111 + }, + { + "epoch": 0.445103296010093, + "grad_norm": 1.8650860786437988, + "learning_rate": 1.2245976817689558e-05, + "loss": 0.3889, + "step": 14112 + }, + { + "epoch": 0.44513483677653365, + "grad_norm": 1.5230838060379028, + "learning_rate": 1.2244981348642886e-05, + "loss": 0.467, + "step": 14113 + }, + { + "epoch": 0.4451663775429743, + "grad_norm": 1.37637460231781, + "learning_rate": 1.224398585616809e-05, + "loss": 0.5191, + "step": 14114 + }, + { + "epoch": 0.4451979183094149, + "grad_norm": 1.7098270654678345, + "learning_rate": 1.224299034027556e-05, + "loss": 0.515, + "step": 14115 + }, + { + "epoch": 0.44522945907585554, + "grad_norm": 1.2944579124450684, + "learning_rate": 1.224199480097568e-05, + "loss": 0.4799, + "step": 14116 + }, + { + "epoch": 0.44526099984229617, + "grad_norm": 1.3984241485595703, + "learning_rate": 1.224099923827885e-05, + "loss": 0.4356, + "step": 14117 + }, + { + "epoch": 0.4452925406087368, + "grad_norm": 3.0377838611602783, + "learning_rate": 1.2240003652195447e-05, + "loss": 0.5151, + "step": 14118 + }, + { + "epoch": 0.4453240813751774, + "grad_norm": 1.914029836654663, + "learning_rate": 1.223900804273587e-05, + "loss": 0.4854, + "step": 14119 + }, + { + "epoch": 0.44535562214161806, + "grad_norm": 2.154608726501465, + "learning_rate": 1.2238012409910505e-05, + "loss": 0.3485, + "step": 14120 + }, + { + "epoch": 0.4453871629080587, + "grad_norm": 2.217618703842163, + "learning_rate": 1.2237016753729741e-05, + "loss": 0.4853, + "step": 14121 + }, + { + "epoch": 0.4454187036744993, + "grad_norm": 2.459336757659912, + "learning_rate": 1.223602107420397e-05, + "loss": 0.5094, + "step": 14122 + }, + { + "epoch": 0.4454502444409399, + "grad_norm": 1.7894634008407593, + "learning_rate": 1.2235025371343585e-05, + "loss": 0.8253, + "step": 14123 + }, + { + "epoch": 0.4454817852073805, + "grad_norm": 1.6797937154769897, + "learning_rate": 1.2234029645158976e-05, + "loss": 0.505, + "step": 14124 + }, + { + "epoch": 0.44551332597382115, + "grad_norm": 1.3753492832183838, + "learning_rate": 1.223303389566053e-05, + "loss": 0.5151, + "step": 14125 + }, + { + "epoch": 0.4455448667402618, + "grad_norm": 1.6687672138214111, + "learning_rate": 1.2232038122858642e-05, + "loss": 0.5066, + "step": 14126 + }, + { + "epoch": 0.4455764075067024, + "grad_norm": 1.6917237043380737, + "learning_rate": 1.2231042326763705e-05, + "loss": 0.5271, + "step": 14127 + }, + { + "epoch": 0.44560794827314304, + "grad_norm": 1.3669618368148804, + "learning_rate": 1.2230046507386106e-05, + "loss": 0.4825, + "step": 14128 + }, + { + "epoch": 0.44563948903958367, + "grad_norm": 1.4907631874084473, + "learning_rate": 1.222905066473624e-05, + "loss": 0.4839, + "step": 14129 + }, + { + "epoch": 0.4456710298060243, + "grad_norm": 3.592336893081665, + "learning_rate": 1.22280547988245e-05, + "loss": 0.4836, + "step": 14130 + }, + { + "epoch": 0.44570257057246493, + "grad_norm": 1.478848934173584, + "learning_rate": 1.222705890966128e-05, + "loss": 0.4733, + "step": 14131 + }, + { + "epoch": 0.44573411133890556, + "grad_norm": 1.365286946296692, + "learning_rate": 1.222606299725697e-05, + "loss": 0.8025, + "step": 14132 + }, + { + "epoch": 0.44576565210534613, + "grad_norm": 1.5310959815979004, + "learning_rate": 1.2225067061621964e-05, + "loss": 0.5154, + "step": 14133 + }, + { + "epoch": 0.44579719287178676, + "grad_norm": 2.1122727394104004, + "learning_rate": 1.222407110276666e-05, + "loss": 0.4996, + "step": 14134 + }, + { + "epoch": 0.4458287336382274, + "grad_norm": 1.5675106048583984, + "learning_rate": 1.222307512070144e-05, + "loss": 0.516, + "step": 14135 + }, + { + "epoch": 0.445860274404668, + "grad_norm": 1.3056873083114624, + "learning_rate": 1.2222079115436706e-05, + "loss": 0.8103, + "step": 14136 + }, + { + "epoch": 0.44589181517110865, + "grad_norm": 1.8786717653274536, + "learning_rate": 1.2221083086982853e-05, + "loss": 0.5166, + "step": 14137 + }, + { + "epoch": 0.4459233559375493, + "grad_norm": 1.4029067754745483, + "learning_rate": 1.2220087035350274e-05, + "loss": 0.4863, + "step": 14138 + }, + { + "epoch": 0.4459548967039899, + "grad_norm": 1.8174359798431396, + "learning_rate": 1.2219090960549363e-05, + "loss": 0.4919, + "step": 14139 + }, + { + "epoch": 0.44598643747043054, + "grad_norm": 1.4272072315216064, + "learning_rate": 1.2218094862590513e-05, + "loss": 0.4554, + "step": 14140 + }, + { + "epoch": 0.44601797823687117, + "grad_norm": 1.523592233657837, + "learning_rate": 1.221709874148412e-05, + "loss": 0.4887, + "step": 14141 + }, + { + "epoch": 0.4460495190033118, + "grad_norm": 1.3348331451416016, + "learning_rate": 1.2216102597240582e-05, + "loss": 0.4928, + "step": 14142 + }, + { + "epoch": 0.44608105976975243, + "grad_norm": 1.2959884405136108, + "learning_rate": 1.2215106429870294e-05, + "loss": 0.8155, + "step": 14143 + }, + { + "epoch": 0.446112600536193, + "grad_norm": 1.385988473892212, + "learning_rate": 1.221411023938365e-05, + "loss": 0.5098, + "step": 14144 + }, + { + "epoch": 0.44614414130263363, + "grad_norm": 1.8563206195831299, + "learning_rate": 1.2213114025791044e-05, + "loss": 0.5057, + "step": 14145 + }, + { + "epoch": 0.44617568206907426, + "grad_norm": 2.3136696815490723, + "learning_rate": 1.2212117789102878e-05, + "loss": 0.5068, + "step": 14146 + }, + { + "epoch": 0.4462072228355149, + "grad_norm": 1.3038198947906494, + "learning_rate": 1.2211121529329542e-05, + "loss": 0.4699, + "step": 14147 + }, + { + "epoch": 0.4462387636019555, + "grad_norm": 1.5744826793670654, + "learning_rate": 1.2210125246481436e-05, + "loss": 0.4799, + "step": 14148 + }, + { + "epoch": 0.44627030436839615, + "grad_norm": 1.5058374404907227, + "learning_rate": 1.2209128940568959e-05, + "loss": 0.496, + "step": 14149 + }, + { + "epoch": 0.4463018451348368, + "grad_norm": 1.5221960544586182, + "learning_rate": 1.2208132611602503e-05, + "loss": 0.4865, + "step": 14150 + }, + { + "epoch": 0.4463333859012774, + "grad_norm": 1.6233242750167847, + "learning_rate": 1.2207136259592473e-05, + "loss": 0.5196, + "step": 14151 + }, + { + "epoch": 0.44636492666771804, + "grad_norm": 1.520696759223938, + "learning_rate": 1.220613988454926e-05, + "loss": 0.4573, + "step": 14152 + }, + { + "epoch": 0.4463964674341587, + "grad_norm": 1.733817458152771, + "learning_rate": 1.2205143486483263e-05, + "loss": 0.5133, + "step": 14153 + }, + { + "epoch": 0.4464280082005993, + "grad_norm": 1.59421968460083, + "learning_rate": 1.2204147065404883e-05, + "loss": 0.5009, + "step": 14154 + }, + { + "epoch": 0.4464595489670399, + "grad_norm": 1.8161706924438477, + "learning_rate": 1.220315062132452e-05, + "loss": 0.5083, + "step": 14155 + }, + { + "epoch": 0.4464910897334805, + "grad_norm": 1.9661506414413452, + "learning_rate": 1.2202154154252566e-05, + "loss": 0.5189, + "step": 14156 + }, + { + "epoch": 0.44652263049992114, + "grad_norm": 2.4876105785369873, + "learning_rate": 1.2201157664199422e-05, + "loss": 0.483, + "step": 14157 + }, + { + "epoch": 0.44655417126636177, + "grad_norm": 1.3058357238769531, + "learning_rate": 1.2200161151175492e-05, + "loss": 0.4612, + "step": 14158 + }, + { + "epoch": 0.4465857120328024, + "grad_norm": 1.4452927112579346, + "learning_rate": 1.2199164615191168e-05, + "loss": 0.5079, + "step": 14159 + }, + { + "epoch": 0.446617252799243, + "grad_norm": 1.4212207794189453, + "learning_rate": 1.2198168056256856e-05, + "loss": 0.4823, + "step": 14160 + }, + { + "epoch": 0.44664879356568365, + "grad_norm": 1.271363377571106, + "learning_rate": 1.2197171474382957e-05, + "loss": 0.3813, + "step": 14161 + }, + { + "epoch": 0.4466803343321243, + "grad_norm": 1.707674264907837, + "learning_rate": 1.2196174869579862e-05, + "loss": 0.5057, + "step": 14162 + }, + { + "epoch": 0.4467118750985649, + "grad_norm": 1.6974265575408936, + "learning_rate": 1.2195178241857979e-05, + "loss": 0.4773, + "step": 14163 + }, + { + "epoch": 0.44674341586500554, + "grad_norm": 1.7026151418685913, + "learning_rate": 1.2194181591227707e-05, + "loss": 0.5026, + "step": 14164 + }, + { + "epoch": 0.4467749566314461, + "grad_norm": 1.6780015230178833, + "learning_rate": 1.2193184917699447e-05, + "loss": 0.496, + "step": 14165 + }, + { + "epoch": 0.44680649739788675, + "grad_norm": 1.6575077772140503, + "learning_rate": 1.2192188221283599e-05, + "loss": 0.4832, + "step": 14166 + }, + { + "epoch": 0.4468380381643274, + "grad_norm": 2.407883644104004, + "learning_rate": 1.2191191501990563e-05, + "loss": 0.5339, + "step": 14167 + }, + { + "epoch": 0.446869578930768, + "grad_norm": 1.6584010124206543, + "learning_rate": 1.2190194759830745e-05, + "loss": 0.473, + "step": 14168 + }, + { + "epoch": 0.44690111969720864, + "grad_norm": 2.0104799270629883, + "learning_rate": 1.2189197994814542e-05, + "loss": 0.4812, + "step": 14169 + }, + { + "epoch": 0.44693266046364927, + "grad_norm": 3.3284401893615723, + "learning_rate": 1.2188201206952357e-05, + "loss": 0.393, + "step": 14170 + }, + { + "epoch": 0.4469642012300899, + "grad_norm": 1.9370476007461548, + "learning_rate": 1.2187204396254595e-05, + "loss": 0.4828, + "step": 14171 + }, + { + "epoch": 0.4469957419965305, + "grad_norm": 1.846722960472107, + "learning_rate": 1.2186207562731656e-05, + "loss": 0.817, + "step": 14172 + }, + { + "epoch": 0.44702728276297116, + "grad_norm": 2.1441681385040283, + "learning_rate": 1.2185210706393947e-05, + "loss": 0.4508, + "step": 14173 + }, + { + "epoch": 0.4470588235294118, + "grad_norm": 1.543596625328064, + "learning_rate": 1.2184213827251863e-05, + "loss": 0.5334, + "step": 14174 + }, + { + "epoch": 0.4470903642958524, + "grad_norm": 1.8265337944030762, + "learning_rate": 1.2183216925315815e-05, + "loss": 0.4674, + "step": 14175 + }, + { + "epoch": 0.447121905062293, + "grad_norm": 1.657790184020996, + "learning_rate": 1.2182220000596199e-05, + "loss": 0.4988, + "step": 14176 + }, + { + "epoch": 0.4471534458287336, + "grad_norm": 1.5547723770141602, + "learning_rate": 1.2181223053103425e-05, + "loss": 0.4971, + "step": 14177 + }, + { + "epoch": 0.44718498659517425, + "grad_norm": 1.615045428276062, + "learning_rate": 1.2180226082847895e-05, + "loss": 0.4743, + "step": 14178 + }, + { + "epoch": 0.4472165273616149, + "grad_norm": 1.8613330125808716, + "learning_rate": 1.2179229089840014e-05, + "loss": 0.5115, + "step": 14179 + }, + { + "epoch": 0.4472480681280555, + "grad_norm": 1.8174998760223389, + "learning_rate": 1.2178232074090184e-05, + "loss": 0.504, + "step": 14180 + }, + { + "epoch": 0.44727960889449614, + "grad_norm": 1.5141887664794922, + "learning_rate": 1.2177235035608813e-05, + "loss": 0.487, + "step": 14181 + }, + { + "epoch": 0.44731114966093677, + "grad_norm": 1.868861436843872, + "learning_rate": 1.2176237974406303e-05, + "loss": 0.4647, + "step": 14182 + }, + { + "epoch": 0.4473426904273774, + "grad_norm": 2.53996205329895, + "learning_rate": 1.2175240890493059e-05, + "loss": 0.478, + "step": 14183 + }, + { + "epoch": 0.44737423119381803, + "grad_norm": 1.5265004634857178, + "learning_rate": 1.2174243783879484e-05, + "loss": 0.4919, + "step": 14184 + }, + { + "epoch": 0.44740577196025866, + "grad_norm": 1.7930235862731934, + "learning_rate": 1.2173246654575992e-05, + "loss": 0.3934, + "step": 14185 + }, + { + "epoch": 0.44743731272669923, + "grad_norm": 1.7310936450958252, + "learning_rate": 1.2172249502592984e-05, + "loss": 0.5135, + "step": 14186 + }, + { + "epoch": 0.44746885349313986, + "grad_norm": 1.7857394218444824, + "learning_rate": 1.217125232794086e-05, + "loss": 0.4991, + "step": 14187 + }, + { + "epoch": 0.4475003942595805, + "grad_norm": 1.9017276763916016, + "learning_rate": 1.2170255130630036e-05, + "loss": 0.3364, + "step": 14188 + }, + { + "epoch": 0.4475319350260211, + "grad_norm": 1.4801979064941406, + "learning_rate": 1.2169257910670915e-05, + "loss": 0.5046, + "step": 14189 + }, + { + "epoch": 0.44756347579246175, + "grad_norm": 2.1271111965179443, + "learning_rate": 1.2168260668073902e-05, + "loss": 0.4769, + "step": 14190 + }, + { + "epoch": 0.4475950165589024, + "grad_norm": 1.5202372074127197, + "learning_rate": 1.2167263402849407e-05, + "loss": 0.5127, + "step": 14191 + }, + { + "epoch": 0.447626557325343, + "grad_norm": 1.6442655324935913, + "learning_rate": 1.2166266115007834e-05, + "loss": 0.8102, + "step": 14192 + }, + { + "epoch": 0.44765809809178364, + "grad_norm": 1.6496862173080444, + "learning_rate": 1.2165268804559594e-05, + "loss": 0.4994, + "step": 14193 + }, + { + "epoch": 0.44768963885822427, + "grad_norm": 1.8830329179763794, + "learning_rate": 1.216427147151509e-05, + "loss": 0.7801, + "step": 14194 + }, + { + "epoch": 0.4477211796246649, + "grad_norm": 1.6321474313735962, + "learning_rate": 1.2163274115884734e-05, + "loss": 0.4963, + "step": 14195 + }, + { + "epoch": 0.44775272039110553, + "grad_norm": 1.5206019878387451, + "learning_rate": 1.2162276737678934e-05, + "loss": 0.5303, + "step": 14196 + }, + { + "epoch": 0.4477842611575461, + "grad_norm": 2.008897066116333, + "learning_rate": 1.2161279336908094e-05, + "loss": 0.4991, + "step": 14197 + }, + { + "epoch": 0.44781580192398673, + "grad_norm": 1.5963457822799683, + "learning_rate": 1.2160281913582628e-05, + "loss": 0.4631, + "step": 14198 + }, + { + "epoch": 0.44784734269042736, + "grad_norm": 1.877362847328186, + "learning_rate": 1.2159284467712944e-05, + "loss": 0.8505, + "step": 14199 + }, + { + "epoch": 0.447878883456868, + "grad_norm": 1.4055733680725098, + "learning_rate": 1.2158286999309447e-05, + "loss": 0.7742, + "step": 14200 + }, + { + "epoch": 0.4479104242233086, + "grad_norm": 1.6403260231018066, + "learning_rate": 1.2157289508382552e-05, + "loss": 0.4777, + "step": 14201 + }, + { + "epoch": 0.44794196498974925, + "grad_norm": 1.6460108757019043, + "learning_rate": 1.2156291994942663e-05, + "loss": 0.5031, + "step": 14202 + }, + { + "epoch": 0.4479735057561899, + "grad_norm": 1.8732209205627441, + "learning_rate": 1.2155294459000197e-05, + "loss": 0.4599, + "step": 14203 + }, + { + "epoch": 0.4480050465226305, + "grad_norm": 1.8206977844238281, + "learning_rate": 1.2154296900565559e-05, + "loss": 0.5108, + "step": 14204 + }, + { + "epoch": 0.44803658728907114, + "grad_norm": 1.7605180740356445, + "learning_rate": 1.2153299319649157e-05, + "loss": 0.4763, + "step": 14205 + }, + { + "epoch": 0.44806812805551177, + "grad_norm": 1.625339150428772, + "learning_rate": 1.2152301716261406e-05, + "loss": 0.5256, + "step": 14206 + }, + { + "epoch": 0.44809966882195235, + "grad_norm": 1.6839042901992798, + "learning_rate": 1.2151304090412715e-05, + "loss": 0.4894, + "step": 14207 + }, + { + "epoch": 0.448131209588393, + "grad_norm": 1.5758475065231323, + "learning_rate": 1.2150306442113495e-05, + "loss": 0.4748, + "step": 14208 + }, + { + "epoch": 0.4481627503548336, + "grad_norm": 1.436144471168518, + "learning_rate": 1.2149308771374158e-05, + "loss": 0.4635, + "step": 14209 + }, + { + "epoch": 0.44819429112127424, + "grad_norm": 6.4171223640441895, + "learning_rate": 1.2148311078205117e-05, + "loss": 0.7766, + "step": 14210 + }, + { + "epoch": 0.44822583188771487, + "grad_norm": 1.548766016960144, + "learning_rate": 1.2147313362616782e-05, + "loss": 0.5183, + "step": 14211 + }, + { + "epoch": 0.4482573726541555, + "grad_norm": 1.4557127952575684, + "learning_rate": 1.214631562461956e-05, + "loss": 0.4973, + "step": 14212 + }, + { + "epoch": 0.4482889134205961, + "grad_norm": 1.232173204421997, + "learning_rate": 1.2145317864223874e-05, + "loss": 0.4725, + "step": 14213 + }, + { + "epoch": 0.44832045418703675, + "grad_norm": 1.6499541997909546, + "learning_rate": 1.2144320081440127e-05, + "loss": 0.4979, + "step": 14214 + }, + { + "epoch": 0.4483519949534774, + "grad_norm": 1.7519556283950806, + "learning_rate": 1.2143322276278733e-05, + "loss": 0.5153, + "step": 14215 + }, + { + "epoch": 0.448383535719918, + "grad_norm": 1.437421441078186, + "learning_rate": 1.2142324448750112e-05, + "loss": 0.5184, + "step": 14216 + }, + { + "epoch": 0.44841507648635864, + "grad_norm": 1.4928178787231445, + "learning_rate": 1.2141326598864667e-05, + "loss": 0.4687, + "step": 14217 + }, + { + "epoch": 0.4484466172527992, + "grad_norm": 2.1456758975982666, + "learning_rate": 1.2140328726632817e-05, + "loss": 0.8094, + "step": 14218 + }, + { + "epoch": 0.44847815801923985, + "grad_norm": 1.804863452911377, + "learning_rate": 1.2139330832064975e-05, + "loss": 0.4659, + "step": 14219 + }, + { + "epoch": 0.4485096987856805, + "grad_norm": 2.4759459495544434, + "learning_rate": 1.2138332915171555e-05, + "loss": 0.4867, + "step": 14220 + }, + { + "epoch": 0.4485412395521211, + "grad_norm": 1.823287010192871, + "learning_rate": 1.2137334975962972e-05, + "loss": 0.4808, + "step": 14221 + }, + { + "epoch": 0.44857278031856174, + "grad_norm": 1.5851948261260986, + "learning_rate": 1.2136337014449639e-05, + "loss": 0.4654, + "step": 14222 + }, + { + "epoch": 0.44860432108500237, + "grad_norm": 1.3206532001495361, + "learning_rate": 1.213533903064197e-05, + "loss": 0.4819, + "step": 14223 + }, + { + "epoch": 0.448635861851443, + "grad_norm": 1.8091280460357666, + "learning_rate": 1.213434102455038e-05, + "loss": 0.4959, + "step": 14224 + }, + { + "epoch": 0.4486674026178836, + "grad_norm": 1.7325509786605835, + "learning_rate": 1.2133342996185283e-05, + "loss": 0.3587, + "step": 14225 + }, + { + "epoch": 0.44869894338432426, + "grad_norm": 1.6446272134780884, + "learning_rate": 1.2132344945557095e-05, + "loss": 0.4852, + "step": 14226 + }, + { + "epoch": 0.4487304841507649, + "grad_norm": 1.5980252027511597, + "learning_rate": 1.2131346872676232e-05, + "loss": 0.4761, + "step": 14227 + }, + { + "epoch": 0.4487620249172055, + "grad_norm": 1.6420464515686035, + "learning_rate": 1.213034877755311e-05, + "loss": 0.5108, + "step": 14228 + }, + { + "epoch": 0.4487935656836461, + "grad_norm": 1.36754310131073, + "learning_rate": 1.2129350660198144e-05, + "loss": 0.4939, + "step": 14229 + }, + { + "epoch": 0.4488251064500867, + "grad_norm": 1.607377290725708, + "learning_rate": 1.2128352520621753e-05, + "loss": 0.4522, + "step": 14230 + }, + { + "epoch": 0.44885664721652735, + "grad_norm": 1.5601550340652466, + "learning_rate": 1.2127354358834352e-05, + "loss": 0.499, + "step": 14231 + }, + { + "epoch": 0.448888187982968, + "grad_norm": 1.6982545852661133, + "learning_rate": 1.2126356174846353e-05, + "loss": 0.492, + "step": 14232 + }, + { + "epoch": 0.4489197287494086, + "grad_norm": 2.2699074745178223, + "learning_rate": 1.2125357968668174e-05, + "loss": 0.5178, + "step": 14233 + }, + { + "epoch": 0.44895126951584924, + "grad_norm": 1.7315548658370972, + "learning_rate": 1.2124359740310238e-05, + "loss": 0.4774, + "step": 14234 + }, + { + "epoch": 0.44898281028228987, + "grad_norm": 1.0730674266815186, + "learning_rate": 1.2123361489782956e-05, + "loss": 0.3449, + "step": 14235 + }, + { + "epoch": 0.4490143510487305, + "grad_norm": 1.2142161130905151, + "learning_rate": 1.2122363217096748e-05, + "loss": 0.3786, + "step": 14236 + }, + { + "epoch": 0.4490458918151711, + "grad_norm": 1.7608071565628052, + "learning_rate": 1.2121364922262035e-05, + "loss": 0.4789, + "step": 14237 + }, + { + "epoch": 0.44907743258161176, + "grad_norm": 1.8760641813278198, + "learning_rate": 1.2120366605289231e-05, + "loss": 0.4825, + "step": 14238 + }, + { + "epoch": 0.44910897334805233, + "grad_norm": 1.6967833042144775, + "learning_rate": 1.2119368266188754e-05, + "loss": 0.5203, + "step": 14239 + }, + { + "epoch": 0.44914051411449296, + "grad_norm": 1.8749653100967407, + "learning_rate": 1.2118369904971025e-05, + "loss": 0.4825, + "step": 14240 + }, + { + "epoch": 0.4491720548809336, + "grad_norm": 1.4902106523513794, + "learning_rate": 1.2117371521646461e-05, + "loss": 0.4774, + "step": 14241 + }, + { + "epoch": 0.4492035956473742, + "grad_norm": 1.3228424787521362, + "learning_rate": 1.211637311622548e-05, + "loss": 0.4807, + "step": 14242 + }, + { + "epoch": 0.44923513641381485, + "grad_norm": 1.831992745399475, + "learning_rate": 1.2115374688718501e-05, + "loss": 0.4891, + "step": 14243 + }, + { + "epoch": 0.4492666771802555, + "grad_norm": 1.4485381841659546, + "learning_rate": 1.2114376239135947e-05, + "loss": 0.4899, + "step": 14244 + }, + { + "epoch": 0.4492982179466961, + "grad_norm": 1.5057899951934814, + "learning_rate": 1.2113377767488233e-05, + "loss": 0.489, + "step": 14245 + }, + { + "epoch": 0.44932975871313674, + "grad_norm": 1.4332956075668335, + "learning_rate": 1.2112379273785782e-05, + "loss": 0.4844, + "step": 14246 + }, + { + "epoch": 0.44936129947957737, + "grad_norm": 1.5639275312423706, + "learning_rate": 1.2111380758039013e-05, + "loss": 0.4691, + "step": 14247 + }, + { + "epoch": 0.449392840246018, + "grad_norm": 1.3913044929504395, + "learning_rate": 1.2110382220258345e-05, + "loss": 0.5004, + "step": 14248 + }, + { + "epoch": 0.44942438101245863, + "grad_norm": 1.4465607404708862, + "learning_rate": 1.2109383660454201e-05, + "loss": 0.7805, + "step": 14249 + }, + { + "epoch": 0.4494559217788992, + "grad_norm": 1.460559368133545, + "learning_rate": 1.2108385078637004e-05, + "loss": 0.804, + "step": 14250 + }, + { + "epoch": 0.44948746254533983, + "grad_norm": 1.5757570266723633, + "learning_rate": 1.2107386474817168e-05, + "loss": 0.5038, + "step": 14251 + }, + { + "epoch": 0.44951900331178046, + "grad_norm": 2.0960912704467773, + "learning_rate": 1.2106387849005119e-05, + "loss": 0.3463, + "step": 14252 + }, + { + "epoch": 0.4495505440782211, + "grad_norm": 1.925276517868042, + "learning_rate": 1.2105389201211275e-05, + "loss": 0.503, + "step": 14253 + }, + { + "epoch": 0.4495820848446617, + "grad_norm": 2.878260612487793, + "learning_rate": 1.2104390531446062e-05, + "loss": 0.5235, + "step": 14254 + }, + { + "epoch": 0.44961362561110235, + "grad_norm": 1.5218217372894287, + "learning_rate": 1.21033918397199e-05, + "loss": 0.5073, + "step": 14255 + }, + { + "epoch": 0.449645166377543, + "grad_norm": 1.573086142539978, + "learning_rate": 1.2102393126043207e-05, + "loss": 0.509, + "step": 14256 + }, + { + "epoch": 0.4496767071439836, + "grad_norm": 1.8241280317306519, + "learning_rate": 1.2101394390426413e-05, + "loss": 0.5201, + "step": 14257 + }, + { + "epoch": 0.44970824791042424, + "grad_norm": 1.2899081707000732, + "learning_rate": 1.2100395632879938e-05, + "loss": 0.4769, + "step": 14258 + }, + { + "epoch": 0.44973978867686487, + "grad_norm": 1.4516757726669312, + "learning_rate": 1.2099396853414201e-05, + "loss": 0.4714, + "step": 14259 + }, + { + "epoch": 0.44977132944330545, + "grad_norm": 1.6784443855285645, + "learning_rate": 1.2098398052039631e-05, + "loss": 0.4785, + "step": 14260 + }, + { + "epoch": 0.4498028702097461, + "grad_norm": 1.6461224555969238, + "learning_rate": 1.2097399228766646e-05, + "loss": 0.476, + "step": 14261 + }, + { + "epoch": 0.4498344109761867, + "grad_norm": 1.8713558912277222, + "learning_rate": 1.209640038360567e-05, + "loss": 0.4579, + "step": 14262 + }, + { + "epoch": 0.44986595174262733, + "grad_norm": 1.5849863290786743, + "learning_rate": 1.209540151656713e-05, + "loss": 0.44, + "step": 14263 + }, + { + "epoch": 0.44989749250906796, + "grad_norm": 1.4518167972564697, + "learning_rate": 1.2094402627661447e-05, + "loss": 0.5, + "step": 14264 + }, + { + "epoch": 0.4499290332755086, + "grad_norm": 2.153226375579834, + "learning_rate": 1.2093403716899045e-05, + "loss": 0.4743, + "step": 14265 + }, + { + "epoch": 0.4499605740419492, + "grad_norm": 2.18359112739563, + "learning_rate": 1.2092404784290351e-05, + "loss": 0.4735, + "step": 14266 + }, + { + "epoch": 0.44999211480838985, + "grad_norm": 1.422206163406372, + "learning_rate": 1.2091405829845788e-05, + "loss": 0.4502, + "step": 14267 + }, + { + "epoch": 0.4500236555748305, + "grad_norm": 1.7948582172393799, + "learning_rate": 1.2090406853575782e-05, + "loss": 0.4536, + "step": 14268 + }, + { + "epoch": 0.4500551963412711, + "grad_norm": 2.101743459701538, + "learning_rate": 1.2089407855490758e-05, + "loss": 0.8254, + "step": 14269 + }, + { + "epoch": 0.45008673710771174, + "grad_norm": 2.18316650390625, + "learning_rate": 1.208840883560114e-05, + "loss": 0.5244, + "step": 14270 + }, + { + "epoch": 0.4501182778741523, + "grad_norm": 1.830721378326416, + "learning_rate": 1.2087409793917358e-05, + "loss": 0.4635, + "step": 14271 + }, + { + "epoch": 0.45014981864059295, + "grad_norm": 1.636834740638733, + "learning_rate": 1.2086410730449827e-05, + "loss": 0.4359, + "step": 14272 + }, + { + "epoch": 0.4501813594070336, + "grad_norm": 1.358964204788208, + "learning_rate": 1.2085411645208982e-05, + "loss": 0.5254, + "step": 14273 + }, + { + "epoch": 0.4502129001734742, + "grad_norm": 1.4854847192764282, + "learning_rate": 1.2084412538205248e-05, + "loss": 0.8017, + "step": 14274 + }, + { + "epoch": 0.45024444093991484, + "grad_norm": 2.3779025077819824, + "learning_rate": 1.2083413409449051e-05, + "loss": 0.5026, + "step": 14275 + }, + { + "epoch": 0.45027598170635547, + "grad_norm": 1.563353180885315, + "learning_rate": 1.208241425895082e-05, + "loss": 0.5252, + "step": 14276 + }, + { + "epoch": 0.4503075224727961, + "grad_norm": 2.3431007862091064, + "learning_rate": 1.2081415086720973e-05, + "loss": 0.5056, + "step": 14277 + }, + { + "epoch": 0.4503390632392367, + "grad_norm": 1.6387320756912231, + "learning_rate": 1.208041589276995e-05, + "loss": 0.5082, + "step": 14278 + }, + { + "epoch": 0.45037060400567736, + "grad_norm": 1.4112215042114258, + "learning_rate": 1.2079416677108167e-05, + "loss": 0.4778, + "step": 14279 + }, + { + "epoch": 0.450402144772118, + "grad_norm": 1.3360505104064941, + "learning_rate": 1.207841743974606e-05, + "loss": 0.5118, + "step": 14280 + }, + { + "epoch": 0.4504336855385586, + "grad_norm": 1.9024184942245483, + "learning_rate": 1.2077418180694049e-05, + "loss": 0.4891, + "step": 14281 + }, + { + "epoch": 0.4504652263049992, + "grad_norm": 1.810765266418457, + "learning_rate": 1.207641889996257e-05, + "loss": 0.4806, + "step": 14282 + }, + { + "epoch": 0.4504967670714398, + "grad_norm": 3.494863748550415, + "learning_rate": 1.2075419597562046e-05, + "loss": 0.3764, + "step": 14283 + }, + { + "epoch": 0.45052830783788045, + "grad_norm": 3.0437448024749756, + "learning_rate": 1.2074420273502906e-05, + "loss": 0.482, + "step": 14284 + }, + { + "epoch": 0.4505598486043211, + "grad_norm": 1.7398062944412231, + "learning_rate": 1.207342092779558e-05, + "loss": 0.5, + "step": 14285 + }, + { + "epoch": 0.4505913893707617, + "grad_norm": 1.5337207317352295, + "learning_rate": 1.2072421560450497e-05, + "loss": 0.4732, + "step": 14286 + }, + { + "epoch": 0.45062293013720234, + "grad_norm": 1.8563268184661865, + "learning_rate": 1.2071422171478086e-05, + "loss": 0.5082, + "step": 14287 + }, + { + "epoch": 0.45065447090364297, + "grad_norm": 1.5388755798339844, + "learning_rate": 1.2070422760888777e-05, + "loss": 0.5122, + "step": 14288 + }, + { + "epoch": 0.4506860116700836, + "grad_norm": 1.7176051139831543, + "learning_rate": 1.2069423328692997e-05, + "loss": 0.475, + "step": 14289 + }, + { + "epoch": 0.4507175524365242, + "grad_norm": 1.598267674446106, + "learning_rate": 1.206842387490118e-05, + "loss": 0.4877, + "step": 14290 + }, + { + "epoch": 0.45074909320296486, + "grad_norm": 1.822901964187622, + "learning_rate": 1.2067424399523753e-05, + "loss": 0.5028, + "step": 14291 + }, + { + "epoch": 0.45078063396940543, + "grad_norm": 1.4423948526382446, + "learning_rate": 1.2066424902571149e-05, + "loss": 0.484, + "step": 14292 + }, + { + "epoch": 0.45081217473584606, + "grad_norm": 1.5488845109939575, + "learning_rate": 1.2065425384053793e-05, + "loss": 0.4745, + "step": 14293 + }, + { + "epoch": 0.4508437155022867, + "grad_norm": 2.134413003921509, + "learning_rate": 1.206442584398212e-05, + "loss": 0.4695, + "step": 14294 + }, + { + "epoch": 0.4508752562687273, + "grad_norm": 1.6208492517471313, + "learning_rate": 1.2063426282366561e-05, + "loss": 0.4758, + "step": 14295 + }, + { + "epoch": 0.45090679703516795, + "grad_norm": 1.5261000394821167, + "learning_rate": 1.2062426699217547e-05, + "loss": 0.5141, + "step": 14296 + }, + { + "epoch": 0.4509383378016086, + "grad_norm": 1.512922763824463, + "learning_rate": 1.2061427094545506e-05, + "loss": 0.489, + "step": 14297 + }, + { + "epoch": 0.4509698785680492, + "grad_norm": 1.657482624053955, + "learning_rate": 1.2060427468360874e-05, + "loss": 0.4896, + "step": 14298 + }, + { + "epoch": 0.45100141933448984, + "grad_norm": 1.6536856889724731, + "learning_rate": 1.2059427820674086e-05, + "loss": 0.4899, + "step": 14299 + }, + { + "epoch": 0.45103296010093047, + "grad_norm": 1.768280029296875, + "learning_rate": 1.2058428151495564e-05, + "loss": 0.4654, + "step": 14300 + }, + { + "epoch": 0.4510645008673711, + "grad_norm": 1.9610604047775269, + "learning_rate": 1.205742846083575e-05, + "loss": 0.4627, + "step": 14301 + }, + { + "epoch": 0.45109604163381173, + "grad_norm": 1.5267294645309448, + "learning_rate": 1.2056428748705072e-05, + "loss": 0.5123, + "step": 14302 + }, + { + "epoch": 0.4511275824002523, + "grad_norm": 2.0325942039489746, + "learning_rate": 1.205542901511396e-05, + "loss": 0.4912, + "step": 14303 + }, + { + "epoch": 0.45115912316669293, + "grad_norm": 1.7548739910125732, + "learning_rate": 1.205442926007285e-05, + "loss": 0.5009, + "step": 14304 + }, + { + "epoch": 0.45119066393313356, + "grad_norm": 1.6921303272247314, + "learning_rate": 1.2053429483592179e-05, + "loss": 0.3566, + "step": 14305 + }, + { + "epoch": 0.4512222046995742, + "grad_norm": 2.2796542644500732, + "learning_rate": 1.2052429685682374e-05, + "loss": 0.4887, + "step": 14306 + }, + { + "epoch": 0.4512537454660148, + "grad_norm": 1.6987425088882446, + "learning_rate": 1.2051429866353873e-05, + "loss": 0.5464, + "step": 14307 + }, + { + "epoch": 0.45128528623245545, + "grad_norm": 1.8149868249893188, + "learning_rate": 1.2050430025617108e-05, + "loss": 0.4665, + "step": 14308 + }, + { + "epoch": 0.4513168269988961, + "grad_norm": 2.1562514305114746, + "learning_rate": 1.2049430163482514e-05, + "loss": 0.4855, + "step": 14309 + }, + { + "epoch": 0.4513483677653367, + "grad_norm": 2.0034523010253906, + "learning_rate": 1.2048430279960529e-05, + "loss": 0.4912, + "step": 14310 + }, + { + "epoch": 0.45137990853177734, + "grad_norm": 1.8621400594711304, + "learning_rate": 1.2047430375061577e-05, + "loss": 0.4899, + "step": 14311 + }, + { + "epoch": 0.45141144929821797, + "grad_norm": 1.862693190574646, + "learning_rate": 1.20464304487961e-05, + "loss": 0.4702, + "step": 14312 + }, + { + "epoch": 0.45144299006465854, + "grad_norm": 2.30627703666687, + "learning_rate": 1.2045430501174535e-05, + "loss": 0.5114, + "step": 14313 + }, + { + "epoch": 0.4514745308310992, + "grad_norm": 2.066122531890869, + "learning_rate": 1.2044430532207313e-05, + "loss": 0.4833, + "step": 14314 + }, + { + "epoch": 0.4515060715975398, + "grad_norm": 1.891036033630371, + "learning_rate": 1.204343054190487e-05, + "loss": 0.4761, + "step": 14315 + }, + { + "epoch": 0.45153761236398043, + "grad_norm": 2.0237104892730713, + "learning_rate": 1.2042430530277645e-05, + "loss": 0.7752, + "step": 14316 + }, + { + "epoch": 0.45156915313042106, + "grad_norm": 1.9786713123321533, + "learning_rate": 1.204143049733607e-05, + "loss": 0.5253, + "step": 14317 + }, + { + "epoch": 0.4516006938968617, + "grad_norm": 1.579569697380066, + "learning_rate": 1.2040430443090583e-05, + "loss": 0.5093, + "step": 14318 + }, + { + "epoch": 0.4516322346633023, + "grad_norm": 1.4482452869415283, + "learning_rate": 1.2039430367551618e-05, + "loss": 0.4665, + "step": 14319 + }, + { + "epoch": 0.45166377542974295, + "grad_norm": 4.16225528717041, + "learning_rate": 1.2038430270729618e-05, + "loss": 0.7832, + "step": 14320 + }, + { + "epoch": 0.4516953161961836, + "grad_norm": 1.794426441192627, + "learning_rate": 1.2037430152635012e-05, + "loss": 0.5007, + "step": 14321 + }, + { + "epoch": 0.4517268569626242, + "grad_norm": 1.588586688041687, + "learning_rate": 1.2036430013278242e-05, + "loss": 0.5282, + "step": 14322 + }, + { + "epoch": 0.45175839772906484, + "grad_norm": 4.990264415740967, + "learning_rate": 1.2035429852669742e-05, + "loss": 0.5076, + "step": 14323 + }, + { + "epoch": 0.4517899384955054, + "grad_norm": 1.739244818687439, + "learning_rate": 1.2034429670819952e-05, + "loss": 0.4801, + "step": 14324 + }, + { + "epoch": 0.45182147926194605, + "grad_norm": 1.8531771898269653, + "learning_rate": 1.2033429467739307e-05, + "loss": 0.4957, + "step": 14325 + }, + { + "epoch": 0.4518530200283867, + "grad_norm": 1.95835542678833, + "learning_rate": 1.2032429243438251e-05, + "loss": 0.4981, + "step": 14326 + }, + { + "epoch": 0.4518845607948273, + "grad_norm": 2.038588047027588, + "learning_rate": 1.2031428997927213e-05, + "loss": 0.433, + "step": 14327 + }, + { + "epoch": 0.45191610156126794, + "grad_norm": 1.4613791704177856, + "learning_rate": 1.2030428731216637e-05, + "loss": 0.4826, + "step": 14328 + }, + { + "epoch": 0.45194764232770857, + "grad_norm": 1.814988136291504, + "learning_rate": 1.2029428443316962e-05, + "loss": 0.8156, + "step": 14329 + }, + { + "epoch": 0.4519791830941492, + "grad_norm": 1.5706312656402588, + "learning_rate": 1.2028428134238625e-05, + "loss": 0.4899, + "step": 14330 + }, + { + "epoch": 0.4520107238605898, + "grad_norm": 1.6182841062545776, + "learning_rate": 1.2027427803992064e-05, + "loss": 0.4976, + "step": 14331 + }, + { + "epoch": 0.45204226462703045, + "grad_norm": 1.5652475357055664, + "learning_rate": 1.2026427452587721e-05, + "loss": 0.5065, + "step": 14332 + }, + { + "epoch": 0.4520738053934711, + "grad_norm": 1.8452811241149902, + "learning_rate": 1.2025427080036034e-05, + "loss": 0.504, + "step": 14333 + }, + { + "epoch": 0.45210534615991166, + "grad_norm": 1.4131577014923096, + "learning_rate": 1.2024426686347442e-05, + "loss": 0.5216, + "step": 14334 + }, + { + "epoch": 0.4521368869263523, + "grad_norm": 1.5593010187149048, + "learning_rate": 1.2023426271532387e-05, + "loss": 0.4916, + "step": 14335 + }, + { + "epoch": 0.4521684276927929, + "grad_norm": 1.5429778099060059, + "learning_rate": 1.2022425835601306e-05, + "loss": 0.4925, + "step": 14336 + }, + { + "epoch": 0.45219996845923355, + "grad_norm": 1.3406546115875244, + "learning_rate": 1.2021425378564642e-05, + "loss": 0.517, + "step": 14337 + }, + { + "epoch": 0.4522315092256742, + "grad_norm": 1.2890325784683228, + "learning_rate": 1.2020424900432835e-05, + "loss": 0.373, + "step": 14338 + }, + { + "epoch": 0.4522630499921148, + "grad_norm": 1.4917938709259033, + "learning_rate": 1.2019424401216325e-05, + "loss": 0.4602, + "step": 14339 + }, + { + "epoch": 0.45229459075855544, + "grad_norm": 1.7330098152160645, + "learning_rate": 1.2018423880925554e-05, + "loss": 0.4568, + "step": 14340 + }, + { + "epoch": 0.45232613152499607, + "grad_norm": 1.753612995147705, + "learning_rate": 1.2017423339570959e-05, + "loss": 0.8335, + "step": 14341 + }, + { + "epoch": 0.4523576722914367, + "grad_norm": 1.7664023637771606, + "learning_rate": 1.2016422777162988e-05, + "loss": 0.4838, + "step": 14342 + }, + { + "epoch": 0.4523892130578773, + "grad_norm": 1.4024235010147095, + "learning_rate": 1.201542219371208e-05, + "loss": 0.4728, + "step": 14343 + }, + { + "epoch": 0.45242075382431796, + "grad_norm": 1.349217414855957, + "learning_rate": 1.2014421589228675e-05, + "loss": 0.474, + "step": 14344 + }, + { + "epoch": 0.45245229459075853, + "grad_norm": 1.6669098138809204, + "learning_rate": 1.2013420963723217e-05, + "loss": 0.4704, + "step": 14345 + }, + { + "epoch": 0.45248383535719916, + "grad_norm": 1.439927339553833, + "learning_rate": 1.2012420317206146e-05, + "loss": 0.4801, + "step": 14346 + }, + { + "epoch": 0.4525153761236398, + "grad_norm": 1.61892831325531, + "learning_rate": 1.201141964968791e-05, + "loss": 0.4591, + "step": 14347 + }, + { + "epoch": 0.4525469168900804, + "grad_norm": 1.3518195152282715, + "learning_rate": 1.2010418961178948e-05, + "loss": 0.4493, + "step": 14348 + }, + { + "epoch": 0.45257845765652105, + "grad_norm": 1.1874914169311523, + "learning_rate": 1.2009418251689702e-05, + "loss": 0.3422, + "step": 14349 + }, + { + "epoch": 0.4526099984229617, + "grad_norm": 2.170973300933838, + "learning_rate": 1.2008417521230618e-05, + "loss": 0.8189, + "step": 14350 + }, + { + "epoch": 0.4526415391894023, + "grad_norm": 1.8357642889022827, + "learning_rate": 1.2007416769812133e-05, + "loss": 0.5178, + "step": 14351 + }, + { + "epoch": 0.45267307995584294, + "grad_norm": 2.360509157180786, + "learning_rate": 1.2006415997444698e-05, + "loss": 0.4931, + "step": 14352 + }, + { + "epoch": 0.45270462072228357, + "grad_norm": 1.5135977268218994, + "learning_rate": 1.2005415204138753e-05, + "loss": 0.4924, + "step": 14353 + }, + { + "epoch": 0.4527361614887242, + "grad_norm": 1.3676739931106567, + "learning_rate": 1.2004414389904744e-05, + "loss": 0.4752, + "step": 14354 + }, + { + "epoch": 0.45276770225516483, + "grad_norm": 1.326789140701294, + "learning_rate": 1.2003413554753114e-05, + "loss": 0.4707, + "step": 14355 + }, + { + "epoch": 0.4527992430216054, + "grad_norm": 1.3748805522918701, + "learning_rate": 1.2002412698694308e-05, + "loss": 0.5007, + "step": 14356 + }, + { + "epoch": 0.45283078378804603, + "grad_norm": 1.722558617591858, + "learning_rate": 1.2001411821738773e-05, + "loss": 0.5027, + "step": 14357 + }, + { + "epoch": 0.45286232455448666, + "grad_norm": 2.4246256351470947, + "learning_rate": 1.2000410923896952e-05, + "loss": 0.4994, + "step": 14358 + }, + { + "epoch": 0.4528938653209273, + "grad_norm": 1.6961559057235718, + "learning_rate": 1.1999410005179286e-05, + "loss": 0.5057, + "step": 14359 + }, + { + "epoch": 0.4529254060873679, + "grad_norm": 3.4540677070617676, + "learning_rate": 1.1998409065596225e-05, + "loss": 0.5173, + "step": 14360 + }, + { + "epoch": 0.45295694685380855, + "grad_norm": 1.4443682432174683, + "learning_rate": 1.1997408105158215e-05, + "loss": 0.4852, + "step": 14361 + }, + { + "epoch": 0.4529884876202492, + "grad_norm": 1.8985873460769653, + "learning_rate": 1.1996407123875699e-05, + "loss": 0.515, + "step": 14362 + }, + { + "epoch": 0.4530200283866898, + "grad_norm": 1.7553889751434326, + "learning_rate": 1.1995406121759126e-05, + "loss": 0.3745, + "step": 14363 + }, + { + "epoch": 0.45305156915313044, + "grad_norm": 1.5346767902374268, + "learning_rate": 1.199440509881894e-05, + "loss": 0.4888, + "step": 14364 + }, + { + "epoch": 0.45308310991957107, + "grad_norm": 1.7725094556808472, + "learning_rate": 1.1993404055065587e-05, + "loss": 0.4373, + "step": 14365 + }, + { + "epoch": 0.45311465068601164, + "grad_norm": 1.6505630016326904, + "learning_rate": 1.1992402990509515e-05, + "loss": 0.4905, + "step": 14366 + }, + { + "epoch": 0.4531461914524523, + "grad_norm": 1.423844337463379, + "learning_rate": 1.1991401905161172e-05, + "loss": 0.5071, + "step": 14367 + }, + { + "epoch": 0.4531777322188929, + "grad_norm": 1.5950433015823364, + "learning_rate": 1.1990400799031009e-05, + "loss": 0.7871, + "step": 14368 + }, + { + "epoch": 0.45320927298533353, + "grad_norm": 1.7893022298812866, + "learning_rate": 1.1989399672129461e-05, + "loss": 0.3665, + "step": 14369 + }, + { + "epoch": 0.45324081375177416, + "grad_norm": 1.5488171577453613, + "learning_rate": 1.1988398524466982e-05, + "loss": 0.7937, + "step": 14370 + }, + { + "epoch": 0.4532723545182148, + "grad_norm": 1.9612221717834473, + "learning_rate": 1.1987397356054025e-05, + "loss": 0.3603, + "step": 14371 + }, + { + "epoch": 0.4533038952846554, + "grad_norm": 1.7001700401306152, + "learning_rate": 1.198639616690103e-05, + "loss": 0.4747, + "step": 14372 + }, + { + "epoch": 0.45333543605109605, + "grad_norm": 1.6332427263259888, + "learning_rate": 1.1985394957018448e-05, + "loss": 0.7774, + "step": 14373 + }, + { + "epoch": 0.4533669768175367, + "grad_norm": 1.8689093589782715, + "learning_rate": 1.1984393726416732e-05, + "loss": 0.4749, + "step": 14374 + }, + { + "epoch": 0.4533985175839773, + "grad_norm": 1.82735276222229, + "learning_rate": 1.1983392475106324e-05, + "loss": 0.4865, + "step": 14375 + }, + { + "epoch": 0.45343005835041794, + "grad_norm": 1.6909563541412354, + "learning_rate": 1.1982391203097677e-05, + "loss": 0.4749, + "step": 14376 + }, + { + "epoch": 0.4534615991168585, + "grad_norm": 1.9489949941635132, + "learning_rate": 1.1981389910401237e-05, + "loss": 0.4767, + "step": 14377 + }, + { + "epoch": 0.45349313988329915, + "grad_norm": 1.2752846479415894, + "learning_rate": 1.1980388597027456e-05, + "loss": 0.4649, + "step": 14378 + }, + { + "epoch": 0.4535246806497398, + "grad_norm": 1.8941067457199097, + "learning_rate": 1.1979387262986781e-05, + "loss": 0.5072, + "step": 14379 + }, + { + "epoch": 0.4535562214161804, + "grad_norm": 1.377625823020935, + "learning_rate": 1.1978385908289663e-05, + "loss": 0.4796, + "step": 14380 + }, + { + "epoch": 0.45358776218262103, + "grad_norm": 1.919274091720581, + "learning_rate": 1.1977384532946555e-05, + "loss": 0.4625, + "step": 14381 + }, + { + "epoch": 0.45361930294906166, + "grad_norm": 1.6310926675796509, + "learning_rate": 1.19763831369679e-05, + "loss": 0.5161, + "step": 14382 + }, + { + "epoch": 0.4536508437155023, + "grad_norm": 1.925439476966858, + "learning_rate": 1.1975381720364153e-05, + "loss": 0.782, + "step": 14383 + }, + { + "epoch": 0.4536823844819429, + "grad_norm": 1.4088799953460693, + "learning_rate": 1.1974380283145764e-05, + "loss": 0.4507, + "step": 14384 + }, + { + "epoch": 0.45371392524838355, + "grad_norm": 2.1765379905700684, + "learning_rate": 1.1973378825323186e-05, + "loss": 0.8074, + "step": 14385 + }, + { + "epoch": 0.4537454660148242, + "grad_norm": 1.6583349704742432, + "learning_rate": 1.1972377346906865e-05, + "loss": 0.4896, + "step": 14386 + }, + { + "epoch": 0.45377700678126476, + "grad_norm": 1.5799286365509033, + "learning_rate": 1.1971375847907257e-05, + "loss": 0.4887, + "step": 14387 + }, + { + "epoch": 0.4538085475477054, + "grad_norm": 1.7128440141677856, + "learning_rate": 1.197037432833481e-05, + "loss": 0.4884, + "step": 14388 + }, + { + "epoch": 0.453840088314146, + "grad_norm": 1.9624675512313843, + "learning_rate": 1.1969372788199976e-05, + "loss": 0.4782, + "step": 14389 + }, + { + "epoch": 0.45387162908058665, + "grad_norm": 1.528632640838623, + "learning_rate": 1.1968371227513208e-05, + "loss": 0.5056, + "step": 14390 + }, + { + "epoch": 0.4539031698470273, + "grad_norm": 2.0768957138061523, + "learning_rate": 1.1967369646284962e-05, + "loss": 0.4783, + "step": 14391 + }, + { + "epoch": 0.4539347106134679, + "grad_norm": 1.472791314125061, + "learning_rate": 1.196636804452568e-05, + "loss": 0.5064, + "step": 14392 + }, + { + "epoch": 0.45396625137990854, + "grad_norm": 1.6855324506759644, + "learning_rate": 1.1965366422245824e-05, + "loss": 0.4885, + "step": 14393 + }, + { + "epoch": 0.45399779214634917, + "grad_norm": 2.8352184295654297, + "learning_rate": 1.1964364779455842e-05, + "loss": 0.5093, + "step": 14394 + }, + { + "epoch": 0.4540293329127898, + "grad_norm": 1.5369222164154053, + "learning_rate": 1.196336311616619e-05, + "loss": 0.7973, + "step": 14395 + }, + { + "epoch": 0.4540608736792304, + "grad_norm": 1.8463611602783203, + "learning_rate": 1.1962361432387317e-05, + "loss": 0.4733, + "step": 14396 + }, + { + "epoch": 0.45409241444567106, + "grad_norm": 1.9398934841156006, + "learning_rate": 1.196135972812968e-05, + "loss": 0.5165, + "step": 14397 + }, + { + "epoch": 0.45412395521211163, + "grad_norm": 1.4758689403533936, + "learning_rate": 1.1960358003403734e-05, + "loss": 0.8184, + "step": 14398 + }, + { + "epoch": 0.45415549597855226, + "grad_norm": 1.4996581077575684, + "learning_rate": 1.1959356258219926e-05, + "loss": 0.5221, + "step": 14399 + }, + { + "epoch": 0.4541870367449929, + "grad_norm": 2.429759979248047, + "learning_rate": 1.1958354492588715e-05, + "loss": 0.4887, + "step": 14400 + }, + { + "epoch": 0.4542185775114335, + "grad_norm": 1.3796393871307373, + "learning_rate": 1.1957352706520553e-05, + "loss": 0.4844, + "step": 14401 + }, + { + "epoch": 0.45425011827787415, + "grad_norm": 1.4120721817016602, + "learning_rate": 1.19563509000259e-05, + "loss": 0.4898, + "step": 14402 + }, + { + "epoch": 0.4542816590443148, + "grad_norm": 2.2071115970611572, + "learning_rate": 1.1955349073115201e-05, + "loss": 0.4833, + "step": 14403 + }, + { + "epoch": 0.4543131998107554, + "grad_norm": 1.816175103187561, + "learning_rate": 1.195434722579892e-05, + "loss": 0.4859, + "step": 14404 + }, + { + "epoch": 0.45434474057719604, + "grad_norm": 1.674475073814392, + "learning_rate": 1.1953345358087508e-05, + "loss": 0.4841, + "step": 14405 + }, + { + "epoch": 0.45437628134363667, + "grad_norm": 1.7310152053833008, + "learning_rate": 1.1952343469991418e-05, + "loss": 0.4835, + "step": 14406 + }, + { + "epoch": 0.4544078221100773, + "grad_norm": 1.4639405012130737, + "learning_rate": 1.195134156152111e-05, + "loss": 0.4564, + "step": 14407 + }, + { + "epoch": 0.45443936287651787, + "grad_norm": 1.8985166549682617, + "learning_rate": 1.1950339632687036e-05, + "loss": 0.5038, + "step": 14408 + }, + { + "epoch": 0.4544709036429585, + "grad_norm": 1.4603497982025146, + "learning_rate": 1.1949337683499658e-05, + "loss": 0.4945, + "step": 14409 + }, + { + "epoch": 0.45450244440939913, + "grad_norm": 1.4710296392440796, + "learning_rate": 1.194833571396942e-05, + "loss": 0.4892, + "step": 14410 + }, + { + "epoch": 0.45453398517583976, + "grad_norm": 1.6373696327209473, + "learning_rate": 1.1947333724106791e-05, + "loss": 0.5101, + "step": 14411 + }, + { + "epoch": 0.4545655259422804, + "grad_norm": 1.8884289264678955, + "learning_rate": 1.1946331713922224e-05, + "loss": 0.4885, + "step": 14412 + }, + { + "epoch": 0.454597066708721, + "grad_norm": 1.4636303186416626, + "learning_rate": 1.194532968342617e-05, + "loss": 0.4779, + "step": 14413 + }, + { + "epoch": 0.45462860747516165, + "grad_norm": 1.5637953281402588, + "learning_rate": 1.1944327632629092e-05, + "loss": 0.4599, + "step": 14414 + }, + { + "epoch": 0.4546601482416023, + "grad_norm": 1.992763638496399, + "learning_rate": 1.1943325561541448e-05, + "loss": 0.5144, + "step": 14415 + }, + { + "epoch": 0.4546916890080429, + "grad_norm": 2.1584267616271973, + "learning_rate": 1.194232347017369e-05, + "loss": 0.8255, + "step": 14416 + }, + { + "epoch": 0.45472322977448354, + "grad_norm": 2.9881460666656494, + "learning_rate": 1.1941321358536278e-05, + "loss": 0.4957, + "step": 14417 + }, + { + "epoch": 0.45475477054092417, + "grad_norm": 1.6796256303787231, + "learning_rate": 1.1940319226639669e-05, + "loss": 0.4355, + "step": 14418 + }, + { + "epoch": 0.45478631130736474, + "grad_norm": 1.5659757852554321, + "learning_rate": 1.1939317074494328e-05, + "loss": 0.4598, + "step": 14419 + }, + { + "epoch": 0.4548178520738054, + "grad_norm": 1.4190508127212524, + "learning_rate": 1.1938314902110701e-05, + "loss": 0.4621, + "step": 14420 + }, + { + "epoch": 0.454849392840246, + "grad_norm": 1.8394712209701538, + "learning_rate": 1.1937312709499255e-05, + "loss": 0.4327, + "step": 14421 + }, + { + "epoch": 0.45488093360668663, + "grad_norm": 1.614061713218689, + "learning_rate": 1.1936310496670449e-05, + "loss": 0.4654, + "step": 14422 + }, + { + "epoch": 0.45491247437312726, + "grad_norm": 1.898438811302185, + "learning_rate": 1.1935308263634737e-05, + "loss": 0.4687, + "step": 14423 + }, + { + "epoch": 0.4549440151395679, + "grad_norm": 1.5628702640533447, + "learning_rate": 1.1934306010402581e-05, + "loss": 0.4843, + "step": 14424 + }, + { + "epoch": 0.4549755559060085, + "grad_norm": 2.3881382942199707, + "learning_rate": 1.193330373698444e-05, + "loss": 0.5066, + "step": 14425 + }, + { + "epoch": 0.45500709667244915, + "grad_norm": 1.8655192852020264, + "learning_rate": 1.1932301443390774e-05, + "loss": 0.3616, + "step": 14426 + }, + { + "epoch": 0.4550386374388898, + "grad_norm": 2.2586047649383545, + "learning_rate": 1.1931299129632044e-05, + "loss": 0.5259, + "step": 14427 + }, + { + "epoch": 0.4550701782053304, + "grad_norm": 1.8910415172576904, + "learning_rate": 1.1930296795718703e-05, + "loss": 0.5132, + "step": 14428 + }, + { + "epoch": 0.45510171897177104, + "grad_norm": 1.5993071794509888, + "learning_rate": 1.1929294441661222e-05, + "loss": 0.501, + "step": 14429 + }, + { + "epoch": 0.4551332597382116, + "grad_norm": 1.9037889242172241, + "learning_rate": 1.192829206747005e-05, + "loss": 0.4798, + "step": 14430 + }, + { + "epoch": 0.45516480050465224, + "grad_norm": 1.7030305862426758, + "learning_rate": 1.1927289673155654e-05, + "loss": 0.48, + "step": 14431 + }, + { + "epoch": 0.4551963412710929, + "grad_norm": 1.9323606491088867, + "learning_rate": 1.1926287258728495e-05, + "loss": 0.5045, + "step": 14432 + }, + { + "epoch": 0.4552278820375335, + "grad_norm": 1.8842624425888062, + "learning_rate": 1.1925284824199033e-05, + "loss": 0.5388, + "step": 14433 + }, + { + "epoch": 0.45525942280397413, + "grad_norm": 1.7066407203674316, + "learning_rate": 1.1924282369577727e-05, + "loss": 0.5165, + "step": 14434 + }, + { + "epoch": 0.45529096357041476, + "grad_norm": 1.8390237092971802, + "learning_rate": 1.1923279894875041e-05, + "loss": 0.476, + "step": 14435 + }, + { + "epoch": 0.4553225043368554, + "grad_norm": 1.804122805595398, + "learning_rate": 1.1922277400101438e-05, + "loss": 0.4836, + "step": 14436 + }, + { + "epoch": 0.455354045103296, + "grad_norm": 1.7700655460357666, + "learning_rate": 1.1921274885267377e-05, + "loss": 0.7683, + "step": 14437 + }, + { + "epoch": 0.45538558586973665, + "grad_norm": 1.678720474243164, + "learning_rate": 1.1920272350383319e-05, + "loss": 0.4874, + "step": 14438 + }, + { + "epoch": 0.4554171266361773, + "grad_norm": 1.753078579902649, + "learning_rate": 1.1919269795459732e-05, + "loss": 0.5304, + "step": 14439 + }, + { + "epoch": 0.45544866740261786, + "grad_norm": 1.5992978811264038, + "learning_rate": 1.1918267220507071e-05, + "loss": 0.5043, + "step": 14440 + }, + { + "epoch": 0.4554802081690585, + "grad_norm": 1.6844004392623901, + "learning_rate": 1.19172646255358e-05, + "loss": 0.4881, + "step": 14441 + }, + { + "epoch": 0.4555117489354991, + "grad_norm": 2.036306619644165, + "learning_rate": 1.1916262010556388e-05, + "loss": 0.4898, + "step": 14442 + }, + { + "epoch": 0.45554328970193975, + "grad_norm": 1.662795066833496, + "learning_rate": 1.1915259375579293e-05, + "loss": 0.4992, + "step": 14443 + }, + { + "epoch": 0.4555748304683804, + "grad_norm": 1.8480905294418335, + "learning_rate": 1.1914256720614979e-05, + "loss": 0.5094, + "step": 14444 + }, + { + "epoch": 0.455606371234821, + "grad_norm": 1.5886085033416748, + "learning_rate": 1.1913254045673908e-05, + "loss": 0.5123, + "step": 14445 + }, + { + "epoch": 0.45563791200126164, + "grad_norm": 4.99705696105957, + "learning_rate": 1.1912251350766546e-05, + "loss": 0.4973, + "step": 14446 + }, + { + "epoch": 0.45566945276770227, + "grad_norm": 1.756142258644104, + "learning_rate": 1.1911248635903361e-05, + "loss": 0.5173, + "step": 14447 + }, + { + "epoch": 0.4557009935341429, + "grad_norm": 1.4978708028793335, + "learning_rate": 1.1910245901094806e-05, + "loss": 0.4979, + "step": 14448 + }, + { + "epoch": 0.4557325343005835, + "grad_norm": 1.7442913055419922, + "learning_rate": 1.1909243146351354e-05, + "loss": 0.4847, + "step": 14449 + }, + { + "epoch": 0.45576407506702415, + "grad_norm": 1.3527532815933228, + "learning_rate": 1.1908240371683468e-05, + "loss": 0.4853, + "step": 14450 + }, + { + "epoch": 0.45579561583346473, + "grad_norm": 1.8063715696334839, + "learning_rate": 1.1907237577101612e-05, + "loss": 0.488, + "step": 14451 + }, + { + "epoch": 0.45582715659990536, + "grad_norm": 4.14255428314209, + "learning_rate": 1.1906234762616249e-05, + "loss": 0.3739, + "step": 14452 + }, + { + "epoch": 0.455858697366346, + "grad_norm": 1.611113429069519, + "learning_rate": 1.190523192823785e-05, + "loss": 0.4747, + "step": 14453 + }, + { + "epoch": 0.4558902381327866, + "grad_norm": 2.796825647354126, + "learning_rate": 1.1904229073976874e-05, + "loss": 0.3884, + "step": 14454 + }, + { + "epoch": 0.45592177889922725, + "grad_norm": 1.6721292734146118, + "learning_rate": 1.1903226199843788e-05, + "loss": 0.5033, + "step": 14455 + }, + { + "epoch": 0.4559533196656679, + "grad_norm": 1.9937114715576172, + "learning_rate": 1.1902223305849061e-05, + "loss": 0.4953, + "step": 14456 + }, + { + "epoch": 0.4559848604321085, + "grad_norm": 2.846334934234619, + "learning_rate": 1.1901220392003157e-05, + "loss": 0.4969, + "step": 14457 + }, + { + "epoch": 0.45601640119854914, + "grad_norm": 2.8010735511779785, + "learning_rate": 1.1900217458316541e-05, + "loss": 0.4566, + "step": 14458 + }, + { + "epoch": 0.45604794196498977, + "grad_norm": 2.066392660140991, + "learning_rate": 1.189921450479968e-05, + "loss": 0.5242, + "step": 14459 + }, + { + "epoch": 0.4560794827314304, + "grad_norm": 1.4049946069717407, + "learning_rate": 1.189821153146304e-05, + "loss": 0.4977, + "step": 14460 + }, + { + "epoch": 0.45611102349787097, + "grad_norm": 1.8116796016693115, + "learning_rate": 1.1897208538317089e-05, + "loss": 0.813, + "step": 14461 + }, + { + "epoch": 0.4561425642643116, + "grad_norm": 1.7002980709075928, + "learning_rate": 1.1896205525372292e-05, + "loss": 0.5159, + "step": 14462 + }, + { + "epoch": 0.45617410503075223, + "grad_norm": 1.6633728742599487, + "learning_rate": 1.1895202492639124e-05, + "loss": 0.4887, + "step": 14463 + }, + { + "epoch": 0.45620564579719286, + "grad_norm": 1.703436255455017, + "learning_rate": 1.1894199440128043e-05, + "loss": 0.4862, + "step": 14464 + }, + { + "epoch": 0.4562371865636335, + "grad_norm": 4.246866703033447, + "learning_rate": 1.1893196367849518e-05, + "loss": 0.4159, + "step": 14465 + }, + { + "epoch": 0.4562687273300741, + "grad_norm": 2.055079936981201, + "learning_rate": 1.1892193275814021e-05, + "loss": 0.4963, + "step": 14466 + }, + { + "epoch": 0.45630026809651475, + "grad_norm": 2.01990008354187, + "learning_rate": 1.189119016403202e-05, + "loss": 0.486, + "step": 14467 + }, + { + "epoch": 0.4563318088629554, + "grad_norm": 2.0203282833099365, + "learning_rate": 1.1890187032513978e-05, + "loss": 0.5038, + "step": 14468 + }, + { + "epoch": 0.456363349629396, + "grad_norm": 1.8274174928665161, + "learning_rate": 1.1889183881270368e-05, + "loss": 0.4908, + "step": 14469 + }, + { + "epoch": 0.45639489039583664, + "grad_norm": 1.9507569074630737, + "learning_rate": 1.1888180710311656e-05, + "loss": 0.3707, + "step": 14470 + }, + { + "epoch": 0.45642643116227727, + "grad_norm": 1.879249095916748, + "learning_rate": 1.1887177519648312e-05, + "loss": 0.5072, + "step": 14471 + }, + { + "epoch": 0.45645797192871784, + "grad_norm": 1.3683958053588867, + "learning_rate": 1.1886174309290805e-05, + "loss": 0.4786, + "step": 14472 + }, + { + "epoch": 0.4564895126951585, + "grad_norm": 1.500222086906433, + "learning_rate": 1.1885171079249607e-05, + "loss": 0.7598, + "step": 14473 + }, + { + "epoch": 0.4565210534615991, + "grad_norm": 2.327894687652588, + "learning_rate": 1.1884167829535185e-05, + "loss": 0.5035, + "step": 14474 + }, + { + "epoch": 0.45655259422803973, + "grad_norm": 2.4678187370300293, + "learning_rate": 1.1883164560158004e-05, + "loss": 0.381, + "step": 14475 + }, + { + "epoch": 0.45658413499448036, + "grad_norm": 1.9811246395111084, + "learning_rate": 1.1882161271128544e-05, + "loss": 0.4853, + "step": 14476 + }, + { + "epoch": 0.456615675760921, + "grad_norm": 4.219017505645752, + "learning_rate": 1.1881157962457271e-05, + "loss": 0.4731, + "step": 14477 + }, + { + "epoch": 0.4566472165273616, + "grad_norm": 2.0686068534851074, + "learning_rate": 1.1880154634154648e-05, + "loss": 0.5369, + "step": 14478 + }, + { + "epoch": 0.45667875729380225, + "grad_norm": 1.389858603477478, + "learning_rate": 1.1879151286231155e-05, + "loss": 0.8043, + "step": 14479 + }, + { + "epoch": 0.4567102980602429, + "grad_norm": 1.5638819932937622, + "learning_rate": 1.1878147918697257e-05, + "loss": 0.5054, + "step": 14480 + }, + { + "epoch": 0.4567418388266835, + "grad_norm": 1.6802393198013306, + "learning_rate": 1.187714453156343e-05, + "loss": 0.5119, + "step": 14481 + }, + { + "epoch": 0.45677337959312414, + "grad_norm": 1.7720611095428467, + "learning_rate": 1.1876141124840139e-05, + "loss": 0.4932, + "step": 14482 + }, + { + "epoch": 0.4568049203595647, + "grad_norm": 1.9367482662200928, + "learning_rate": 1.187513769853786e-05, + "loss": 0.5005, + "step": 14483 + }, + { + "epoch": 0.45683646112600534, + "grad_norm": 2.6362433433532715, + "learning_rate": 1.1874134252667062e-05, + "loss": 0.3689, + "step": 14484 + }, + { + "epoch": 0.456868001892446, + "grad_norm": 1.7295202016830444, + "learning_rate": 1.187313078723822e-05, + "loss": 0.5029, + "step": 14485 + }, + { + "epoch": 0.4568995426588866, + "grad_norm": 1.5751304626464844, + "learning_rate": 1.1872127302261805e-05, + "loss": 0.5119, + "step": 14486 + }, + { + "epoch": 0.45693108342532723, + "grad_norm": 1.5120311975479126, + "learning_rate": 1.1871123797748285e-05, + "loss": 0.8354, + "step": 14487 + }, + { + "epoch": 0.45696262419176786, + "grad_norm": 1.6236350536346436, + "learning_rate": 1.1870120273708136e-05, + "loss": 0.4436, + "step": 14488 + }, + { + "epoch": 0.4569941649582085, + "grad_norm": 2.436237096786499, + "learning_rate": 1.1869116730151831e-05, + "loss": 0.7771, + "step": 14489 + }, + { + "epoch": 0.4570257057246491, + "grad_norm": 1.4192403554916382, + "learning_rate": 1.1868113167089839e-05, + "loss": 0.5062, + "step": 14490 + }, + { + "epoch": 0.45705724649108975, + "grad_norm": 1.6657441854476929, + "learning_rate": 1.1867109584532637e-05, + "loss": 0.4804, + "step": 14491 + }, + { + "epoch": 0.4570887872575304, + "grad_norm": 2.4318535327911377, + "learning_rate": 1.1866105982490697e-05, + "loss": 0.5202, + "step": 14492 + }, + { + "epoch": 0.45712032802397096, + "grad_norm": 3.3561818599700928, + "learning_rate": 1.186510236097449e-05, + "loss": 0.4595, + "step": 14493 + }, + { + "epoch": 0.4571518687904116, + "grad_norm": 1.6124317646026611, + "learning_rate": 1.1864098719994497e-05, + "loss": 0.487, + "step": 14494 + }, + { + "epoch": 0.4571834095568522, + "grad_norm": 4.960737228393555, + "learning_rate": 1.1863095059561181e-05, + "loss": 0.4935, + "step": 14495 + }, + { + "epoch": 0.45721495032329285, + "grad_norm": 2.076690912246704, + "learning_rate": 1.1862091379685026e-05, + "loss": 0.4896, + "step": 14496 + }, + { + "epoch": 0.4572464910897335, + "grad_norm": 1.6521446704864502, + "learning_rate": 1.1861087680376498e-05, + "loss": 0.7884, + "step": 14497 + }, + { + "epoch": 0.4572780318561741, + "grad_norm": 2.1342012882232666, + "learning_rate": 1.1860083961646077e-05, + "loss": 0.5146, + "step": 14498 + }, + { + "epoch": 0.45730957262261474, + "grad_norm": 1.7004897594451904, + "learning_rate": 1.1859080223504234e-05, + "loss": 0.4722, + "step": 14499 + }, + { + "epoch": 0.45734111338905536, + "grad_norm": 1.4558573961257935, + "learning_rate": 1.1858076465961444e-05, + "loss": 0.493, + "step": 14500 + }, + { + "epoch": 0.457372654155496, + "grad_norm": 1.6000268459320068, + "learning_rate": 1.1857072689028186e-05, + "loss": 0.5035, + "step": 14501 + }, + { + "epoch": 0.4574041949219366, + "grad_norm": 1.6832038164138794, + "learning_rate": 1.185606889271493e-05, + "loss": 0.4883, + "step": 14502 + }, + { + "epoch": 0.45743573568837725, + "grad_norm": 1.5558565855026245, + "learning_rate": 1.1855065077032155e-05, + "loss": 0.4633, + "step": 14503 + }, + { + "epoch": 0.45746727645481783, + "grad_norm": 1.4274437427520752, + "learning_rate": 1.1854061241990337e-05, + "loss": 0.489, + "step": 14504 + }, + { + "epoch": 0.45749881722125846, + "grad_norm": 1.5862373113632202, + "learning_rate": 1.1853057387599948e-05, + "loss": 0.4948, + "step": 14505 + }, + { + "epoch": 0.4575303579876991, + "grad_norm": 1.5403892993927002, + "learning_rate": 1.1852053513871471e-05, + "loss": 0.7855, + "step": 14506 + }, + { + "epoch": 0.4575618987541397, + "grad_norm": 1.642842173576355, + "learning_rate": 1.1851049620815373e-05, + "loss": 0.469, + "step": 14507 + }, + { + "epoch": 0.45759343952058035, + "grad_norm": 1.363666296005249, + "learning_rate": 1.1850045708442138e-05, + "loss": 0.4923, + "step": 14508 + }, + { + "epoch": 0.457624980287021, + "grad_norm": 1.5614230632781982, + "learning_rate": 1.1849041776762235e-05, + "loss": 0.496, + "step": 14509 + }, + { + "epoch": 0.4576565210534616, + "grad_norm": 1.6890701055526733, + "learning_rate": 1.1848037825786148e-05, + "loss": 0.4904, + "step": 14510 + }, + { + "epoch": 0.45768806181990224, + "grad_norm": 1.3575950860977173, + "learning_rate": 1.184703385552435e-05, + "loss": 0.4631, + "step": 14511 + }, + { + "epoch": 0.45771960258634287, + "grad_norm": 1.8310405015945435, + "learning_rate": 1.1846029865987323e-05, + "loss": 0.4924, + "step": 14512 + }, + { + "epoch": 0.4577511433527835, + "grad_norm": 1.584181785583496, + "learning_rate": 1.1845025857185538e-05, + "loss": 0.7719, + "step": 14513 + }, + { + "epoch": 0.45778268411922407, + "grad_norm": 2.099013090133667, + "learning_rate": 1.1844021829129477e-05, + "loss": 0.491, + "step": 14514 + }, + { + "epoch": 0.4578142248856647, + "grad_norm": 2.6992130279541016, + "learning_rate": 1.1843017781829613e-05, + "loss": 0.7893, + "step": 14515 + }, + { + "epoch": 0.45784576565210533, + "grad_norm": 1.36444890499115, + "learning_rate": 1.1842013715296432e-05, + "loss": 0.7856, + "step": 14516 + }, + { + "epoch": 0.45787730641854596, + "grad_norm": 1.5694819688796997, + "learning_rate": 1.1841009629540406e-05, + "loss": 0.4905, + "step": 14517 + }, + { + "epoch": 0.4579088471849866, + "grad_norm": 1.515305519104004, + "learning_rate": 1.1840005524572016e-05, + "loss": 0.4871, + "step": 14518 + }, + { + "epoch": 0.4579403879514272, + "grad_norm": 3.1488988399505615, + "learning_rate": 1.1839001400401738e-05, + "loss": 0.3976, + "step": 14519 + }, + { + "epoch": 0.45797192871786785, + "grad_norm": 1.625929355621338, + "learning_rate": 1.183799725704005e-05, + "loss": 0.4766, + "step": 14520 + }, + { + "epoch": 0.4580034694843085, + "grad_norm": 1.4769378900527954, + "learning_rate": 1.1836993094497435e-05, + "loss": 0.5124, + "step": 14521 + }, + { + "epoch": 0.4580350102507491, + "grad_norm": 1.543752908706665, + "learning_rate": 1.1835988912784373e-05, + "loss": 0.4884, + "step": 14522 + }, + { + "epoch": 0.45806655101718974, + "grad_norm": 1.795709490776062, + "learning_rate": 1.183498471191134e-05, + "loss": 0.4967, + "step": 14523 + }, + { + "epoch": 0.45809809178363037, + "grad_norm": 1.9450852870941162, + "learning_rate": 1.1833980491888816e-05, + "loss": 0.3786, + "step": 14524 + }, + { + "epoch": 0.45812963255007094, + "grad_norm": 2.1279077529907227, + "learning_rate": 1.1832976252727282e-05, + "loss": 0.4908, + "step": 14525 + }, + { + "epoch": 0.45816117331651157, + "grad_norm": 1.4650905132293701, + "learning_rate": 1.183197199443722e-05, + "loss": 0.5178, + "step": 14526 + }, + { + "epoch": 0.4581927140829522, + "grad_norm": 1.9924818277359009, + "learning_rate": 1.1830967717029104e-05, + "loss": 0.5118, + "step": 14527 + }, + { + "epoch": 0.45822425484939283, + "grad_norm": 1.9637809991836548, + "learning_rate": 1.1829963420513419e-05, + "loss": 0.4808, + "step": 14528 + }, + { + "epoch": 0.45825579561583346, + "grad_norm": 1.351310133934021, + "learning_rate": 1.1828959104900646e-05, + "loss": 0.4913, + "step": 14529 + }, + { + "epoch": 0.4582873363822741, + "grad_norm": 1.5437339544296265, + "learning_rate": 1.1827954770201264e-05, + "loss": 0.4939, + "step": 14530 + }, + { + "epoch": 0.4583188771487147, + "grad_norm": 2.658442735671997, + "learning_rate": 1.1826950416425753e-05, + "loss": 0.4794, + "step": 14531 + }, + { + "epoch": 0.45835041791515535, + "grad_norm": 1.5408334732055664, + "learning_rate": 1.1825946043584598e-05, + "loss": 0.4525, + "step": 14532 + }, + { + "epoch": 0.458381958681596, + "grad_norm": 1.8660085201263428, + "learning_rate": 1.1824941651688277e-05, + "loss": 0.4991, + "step": 14533 + }, + { + "epoch": 0.4584134994480366, + "grad_norm": 2.7200911045074463, + "learning_rate": 1.1823937240747274e-05, + "loss": 0.5096, + "step": 14534 + }, + { + "epoch": 0.4584450402144772, + "grad_norm": 2.059427261352539, + "learning_rate": 1.1822932810772069e-05, + "loss": 0.479, + "step": 14535 + }, + { + "epoch": 0.4584765809809178, + "grad_norm": 2.6273577213287354, + "learning_rate": 1.1821928361773148e-05, + "loss": 0.526, + "step": 14536 + }, + { + "epoch": 0.45850812174735844, + "grad_norm": 1.2948850393295288, + "learning_rate": 1.1820923893760985e-05, + "loss": 0.4851, + "step": 14537 + }, + { + "epoch": 0.4585396625137991, + "grad_norm": 1.4991754293441772, + "learning_rate": 1.1819919406746067e-05, + "loss": 0.4825, + "step": 14538 + }, + { + "epoch": 0.4585712032802397, + "grad_norm": 1.6520180702209473, + "learning_rate": 1.1818914900738881e-05, + "loss": 0.5288, + "step": 14539 + }, + { + "epoch": 0.45860274404668033, + "grad_norm": 1.4240957498550415, + "learning_rate": 1.1817910375749902e-05, + "loss": 0.457, + "step": 14540 + }, + { + "epoch": 0.45863428481312096, + "grad_norm": 1.5776761770248413, + "learning_rate": 1.1816905831789616e-05, + "loss": 0.4944, + "step": 14541 + }, + { + "epoch": 0.4586658255795616, + "grad_norm": 2.7723937034606934, + "learning_rate": 1.1815901268868511e-05, + "loss": 0.4437, + "step": 14542 + }, + { + "epoch": 0.4586973663460022, + "grad_norm": 2.9332172870635986, + "learning_rate": 1.1814896686997061e-05, + "loss": 0.478, + "step": 14543 + }, + { + "epoch": 0.45872890711244285, + "grad_norm": 2.4128026962280273, + "learning_rate": 1.181389208618576e-05, + "loss": 0.5162, + "step": 14544 + }, + { + "epoch": 0.4587604478788835, + "grad_norm": 2.2710301876068115, + "learning_rate": 1.1812887466445084e-05, + "loss": 0.4864, + "step": 14545 + }, + { + "epoch": 0.45879198864532406, + "grad_norm": 1.4858148097991943, + "learning_rate": 1.1811882827785521e-05, + "loss": 0.4944, + "step": 14546 + }, + { + "epoch": 0.4588235294117647, + "grad_norm": 1.6002674102783203, + "learning_rate": 1.1810878170217553e-05, + "loss": 0.4687, + "step": 14547 + }, + { + "epoch": 0.4588550701782053, + "grad_norm": 1.7644776105880737, + "learning_rate": 1.1809873493751664e-05, + "loss": 0.5057, + "step": 14548 + }, + { + "epoch": 0.45888661094464595, + "grad_norm": 2.015268564224243, + "learning_rate": 1.1808868798398344e-05, + "loss": 0.8153, + "step": 14549 + }, + { + "epoch": 0.4589181517110866, + "grad_norm": 1.7838997840881348, + "learning_rate": 1.1807864084168067e-05, + "loss": 0.479, + "step": 14550 + }, + { + "epoch": 0.4589496924775272, + "grad_norm": 1.5266674757003784, + "learning_rate": 1.1806859351071326e-05, + "loss": 0.4964, + "step": 14551 + }, + { + "epoch": 0.45898123324396783, + "grad_norm": 1.4962774515151978, + "learning_rate": 1.1805854599118607e-05, + "loss": 0.5037, + "step": 14552 + }, + { + "epoch": 0.45901277401040846, + "grad_norm": 3.931689739227295, + "learning_rate": 1.1804849828320395e-05, + "loss": 0.3837, + "step": 14553 + }, + { + "epoch": 0.4590443147768491, + "grad_norm": 1.9516183137893677, + "learning_rate": 1.1803845038687171e-05, + "loss": 0.4797, + "step": 14554 + }, + { + "epoch": 0.4590758555432897, + "grad_norm": 2.151991128921509, + "learning_rate": 1.1802840230229424e-05, + "loss": 0.4251, + "step": 14555 + }, + { + "epoch": 0.45910739630973035, + "grad_norm": 1.6948449611663818, + "learning_rate": 1.1801835402957639e-05, + "loss": 0.8022, + "step": 14556 + }, + { + "epoch": 0.4591389370761709, + "grad_norm": 2.035642385482788, + "learning_rate": 1.1800830556882303e-05, + "loss": 0.4919, + "step": 14557 + }, + { + "epoch": 0.45917047784261156, + "grad_norm": 1.484515905380249, + "learning_rate": 1.17998256920139e-05, + "loss": 0.5214, + "step": 14558 + }, + { + "epoch": 0.4592020186090522, + "grad_norm": 1.579233169555664, + "learning_rate": 1.1798820808362918e-05, + "loss": 0.5297, + "step": 14559 + }, + { + "epoch": 0.4592335593754928, + "grad_norm": 1.8873472213745117, + "learning_rate": 1.1797815905939849e-05, + "loss": 0.4959, + "step": 14560 + }, + { + "epoch": 0.45926510014193345, + "grad_norm": 1.7533079385757446, + "learning_rate": 1.179681098475517e-05, + "loss": 0.4881, + "step": 14561 + }, + { + "epoch": 0.4592966409083741, + "grad_norm": 1.9736586809158325, + "learning_rate": 1.1795806044819374e-05, + "loss": 0.5044, + "step": 14562 + }, + { + "epoch": 0.4593281816748147, + "grad_norm": 1.7708981037139893, + "learning_rate": 1.1794801086142951e-05, + "loss": 0.458, + "step": 14563 + }, + { + "epoch": 0.45935972244125534, + "grad_norm": 1.9600136280059814, + "learning_rate": 1.1793796108736384e-05, + "loss": 0.471, + "step": 14564 + }, + { + "epoch": 0.45939126320769597, + "grad_norm": 1.7817105054855347, + "learning_rate": 1.1792791112610158e-05, + "loss": 0.5044, + "step": 14565 + }, + { + "epoch": 0.4594228039741366, + "grad_norm": 1.666040062904358, + "learning_rate": 1.179178609777477e-05, + "loss": 0.4842, + "step": 14566 + }, + { + "epoch": 0.45945434474057717, + "grad_norm": 1.787424087524414, + "learning_rate": 1.1790781064240697e-05, + "loss": 0.4074, + "step": 14567 + }, + { + "epoch": 0.4594858855070178, + "grad_norm": 1.8170000314712524, + "learning_rate": 1.1789776012018437e-05, + "loss": 0.5176, + "step": 14568 + }, + { + "epoch": 0.45951742627345843, + "grad_norm": 2.8674397468566895, + "learning_rate": 1.1788770941118471e-05, + "loss": 0.5178, + "step": 14569 + }, + { + "epoch": 0.45954896703989906, + "grad_norm": 2.075322389602661, + "learning_rate": 1.1787765851551296e-05, + "loss": 0.4784, + "step": 14570 + }, + { + "epoch": 0.4595805078063397, + "grad_norm": 1.9696242809295654, + "learning_rate": 1.1786760743327393e-05, + "loss": 0.5172, + "step": 14571 + }, + { + "epoch": 0.4596120485727803, + "grad_norm": 2.3046913146972656, + "learning_rate": 1.1785755616457254e-05, + "loss": 0.465, + "step": 14572 + }, + { + "epoch": 0.45964358933922095, + "grad_norm": 1.7555043697357178, + "learning_rate": 1.1784750470951372e-05, + "loss": 0.4863, + "step": 14573 + }, + { + "epoch": 0.4596751301056616, + "grad_norm": 1.5537251234054565, + "learning_rate": 1.1783745306820231e-05, + "loss": 0.3806, + "step": 14574 + }, + { + "epoch": 0.4597066708721022, + "grad_norm": 1.9134950637817383, + "learning_rate": 1.178274012407432e-05, + "loss": 0.7771, + "step": 14575 + }, + { + "epoch": 0.45973821163854284, + "grad_norm": 1.5708357095718384, + "learning_rate": 1.1781734922724136e-05, + "loss": 0.4888, + "step": 14576 + }, + { + "epoch": 0.45976975240498347, + "grad_norm": 1.5622836351394653, + "learning_rate": 1.1780729702780163e-05, + "loss": 0.4978, + "step": 14577 + }, + { + "epoch": 0.45980129317142404, + "grad_norm": 1.5688693523406982, + "learning_rate": 1.1779724464252891e-05, + "loss": 0.5136, + "step": 14578 + }, + { + "epoch": 0.45983283393786467, + "grad_norm": 1.7637628316879272, + "learning_rate": 1.1778719207152814e-05, + "loss": 0.4845, + "step": 14579 + }, + { + "epoch": 0.4598643747043053, + "grad_norm": 1.4555310010910034, + "learning_rate": 1.1777713931490419e-05, + "loss": 0.4832, + "step": 14580 + }, + { + "epoch": 0.45989591547074593, + "grad_norm": 1.8066914081573486, + "learning_rate": 1.1776708637276202e-05, + "loss": 0.5228, + "step": 14581 + }, + { + "epoch": 0.45992745623718656, + "grad_norm": 11.944938659667969, + "learning_rate": 1.177570332452065e-05, + "loss": 0.3907, + "step": 14582 + }, + { + "epoch": 0.4599589970036272, + "grad_norm": 1.753204107284546, + "learning_rate": 1.1774697993234252e-05, + "loss": 0.4754, + "step": 14583 + }, + { + "epoch": 0.4599905377700678, + "grad_norm": 1.3739951848983765, + "learning_rate": 1.1773692643427508e-05, + "loss": 0.4984, + "step": 14584 + }, + { + "epoch": 0.46002207853650845, + "grad_norm": 5.043841361999512, + "learning_rate": 1.1772687275110898e-05, + "loss": 0.3734, + "step": 14585 + }, + { + "epoch": 0.4600536193029491, + "grad_norm": 1.7765759229660034, + "learning_rate": 1.1771681888294921e-05, + "loss": 0.5005, + "step": 14586 + }, + { + "epoch": 0.4600851600693897, + "grad_norm": 1.538312315940857, + "learning_rate": 1.177067648299007e-05, + "loss": 0.5257, + "step": 14587 + }, + { + "epoch": 0.4601167008358303, + "grad_norm": 5.1746439933776855, + "learning_rate": 1.1769671059206835e-05, + "loss": 0.4875, + "step": 14588 + }, + { + "epoch": 0.4601482416022709, + "grad_norm": 1.6321407556533813, + "learning_rate": 1.1768665616955708e-05, + "loss": 0.5075, + "step": 14589 + }, + { + "epoch": 0.46017978236871154, + "grad_norm": 1.6047790050506592, + "learning_rate": 1.1767660156247182e-05, + "loss": 0.4814, + "step": 14590 + }, + { + "epoch": 0.4602113231351522, + "grad_norm": 1.5899343490600586, + "learning_rate": 1.1766654677091749e-05, + "loss": 0.509, + "step": 14591 + }, + { + "epoch": 0.4602428639015928, + "grad_norm": 1.800181269645691, + "learning_rate": 1.1765649179499902e-05, + "loss": 0.4778, + "step": 14592 + }, + { + "epoch": 0.46027440466803343, + "grad_norm": 1.7361009120941162, + "learning_rate": 1.1764643663482138e-05, + "loss": 0.4696, + "step": 14593 + }, + { + "epoch": 0.46030594543447406, + "grad_norm": 1.8124229907989502, + "learning_rate": 1.1763638129048945e-05, + "loss": 0.7445, + "step": 14594 + }, + { + "epoch": 0.4603374862009147, + "grad_norm": 1.5053647756576538, + "learning_rate": 1.1762632576210817e-05, + "loss": 0.4987, + "step": 14595 + }, + { + "epoch": 0.4603690269673553, + "grad_norm": 1.908024549484253, + "learning_rate": 1.1761627004978251e-05, + "loss": 0.4976, + "step": 14596 + }, + { + "epoch": 0.46040056773379595, + "grad_norm": 1.418223261833191, + "learning_rate": 1.1760621415361741e-05, + "loss": 0.789, + "step": 14597 + }, + { + "epoch": 0.4604321085002366, + "grad_norm": 2.555020332336426, + "learning_rate": 1.1759615807371777e-05, + "loss": 0.4739, + "step": 14598 + }, + { + "epoch": 0.46046364926667716, + "grad_norm": 1.8441047668457031, + "learning_rate": 1.1758610181018857e-05, + "loss": 0.5082, + "step": 14599 + }, + { + "epoch": 0.4604951900331178, + "grad_norm": 1.7921332120895386, + "learning_rate": 1.1757604536313471e-05, + "loss": 0.4908, + "step": 14600 + }, + { + "epoch": 0.4605267307995584, + "grad_norm": 1.5077787637710571, + "learning_rate": 1.1756598873266123e-05, + "loss": 0.4892, + "step": 14601 + }, + { + "epoch": 0.46055827156599904, + "grad_norm": 1.661571741104126, + "learning_rate": 1.1755593191887298e-05, + "loss": 0.4918, + "step": 14602 + }, + { + "epoch": 0.4605898123324397, + "grad_norm": 1.5088081359863281, + "learning_rate": 1.1754587492187497e-05, + "loss": 0.5137, + "step": 14603 + }, + { + "epoch": 0.4606213530988803, + "grad_norm": 2.2016172409057617, + "learning_rate": 1.1753581774177213e-05, + "loss": 0.3928, + "step": 14604 + }, + { + "epoch": 0.46065289386532093, + "grad_norm": 4.619851112365723, + "learning_rate": 1.175257603786694e-05, + "loss": 0.3568, + "step": 14605 + }, + { + "epoch": 0.46068443463176156, + "grad_norm": 4.841665267944336, + "learning_rate": 1.1751570283267173e-05, + "loss": 0.4111, + "step": 14606 + }, + { + "epoch": 0.4607159753982022, + "grad_norm": 1.4270625114440918, + "learning_rate": 1.1750564510388412e-05, + "loss": 0.7817, + "step": 14607 + }, + { + "epoch": 0.4607475161646428, + "grad_norm": 1.7767101526260376, + "learning_rate": 1.1749558719241153e-05, + "loss": 0.4925, + "step": 14608 + }, + { + "epoch": 0.46077905693108345, + "grad_norm": 1.7198359966278076, + "learning_rate": 1.1748552909835887e-05, + "loss": 0.5567, + "step": 14609 + }, + { + "epoch": 0.460810597697524, + "grad_norm": 1.5025415420532227, + "learning_rate": 1.1747547082183115e-05, + "loss": 0.4731, + "step": 14610 + }, + { + "epoch": 0.46084213846396466, + "grad_norm": 1.5824178457260132, + "learning_rate": 1.1746541236293334e-05, + "loss": 0.7734, + "step": 14611 + }, + { + "epoch": 0.4608736792304053, + "grad_norm": 1.617654800415039, + "learning_rate": 1.1745535372177036e-05, + "loss": 0.4984, + "step": 14612 + }, + { + "epoch": 0.4609052199968459, + "grad_norm": 1.5230753421783447, + "learning_rate": 1.1744529489844725e-05, + "loss": 0.4913, + "step": 14613 + }, + { + "epoch": 0.46093676076328655, + "grad_norm": 1.6537299156188965, + "learning_rate": 1.1743523589306891e-05, + "loss": 0.5292, + "step": 14614 + }, + { + "epoch": 0.4609683015297272, + "grad_norm": 1.3175705671310425, + "learning_rate": 1.1742517670574036e-05, + "loss": 0.4644, + "step": 14615 + }, + { + "epoch": 0.4609998422961678, + "grad_norm": 1.4246361255645752, + "learning_rate": 1.1741511733656655e-05, + "loss": 0.7888, + "step": 14616 + }, + { + "epoch": 0.46103138306260844, + "grad_norm": 1.7776285409927368, + "learning_rate": 1.1740505778565245e-05, + "loss": 0.5057, + "step": 14617 + }, + { + "epoch": 0.46106292382904907, + "grad_norm": 1.358626365661621, + "learning_rate": 1.173949980531031e-05, + "loss": 0.775, + "step": 14618 + }, + { + "epoch": 0.4610944645954897, + "grad_norm": 1.5572601556777954, + "learning_rate": 1.1738493813902341e-05, + "loss": 0.491, + "step": 14619 + }, + { + "epoch": 0.46112600536193027, + "grad_norm": 1.7974553108215332, + "learning_rate": 1.173748780435184e-05, + "loss": 0.4864, + "step": 14620 + }, + { + "epoch": 0.4611575461283709, + "grad_norm": 1.4690346717834473, + "learning_rate": 1.1736481776669307e-05, + "loss": 0.797, + "step": 14621 + }, + { + "epoch": 0.46118908689481153, + "grad_norm": 1.6270865201950073, + "learning_rate": 1.1735475730865233e-05, + "loss": 0.469, + "step": 14622 + }, + { + "epoch": 0.46122062766125216, + "grad_norm": 1.5715023279190063, + "learning_rate": 1.1734469666950129e-05, + "loss": 0.4855, + "step": 14623 + }, + { + "epoch": 0.4612521684276928, + "grad_norm": 2.52433705329895, + "learning_rate": 1.1733463584934481e-05, + "loss": 0.4874, + "step": 14624 + }, + { + "epoch": 0.4612837091941334, + "grad_norm": 1.717129111289978, + "learning_rate": 1.1732457484828798e-05, + "loss": 0.4824, + "step": 14625 + }, + { + "epoch": 0.46131524996057405, + "grad_norm": 1.5331084728240967, + "learning_rate": 1.1731451366643573e-05, + "loss": 0.5039, + "step": 14626 + }, + { + "epoch": 0.4613467907270147, + "grad_norm": 1.8013792037963867, + "learning_rate": 1.173044523038931e-05, + "loss": 0.4981, + "step": 14627 + }, + { + "epoch": 0.4613783314934553, + "grad_norm": 3.129462480545044, + "learning_rate": 1.172943907607651e-05, + "loss": 0.4717, + "step": 14628 + }, + { + "epoch": 0.46140987225989594, + "grad_norm": 1.6948493719100952, + "learning_rate": 1.1728432903715667e-05, + "loss": 0.5212, + "step": 14629 + }, + { + "epoch": 0.46144141302633657, + "grad_norm": 1.3325163125991821, + "learning_rate": 1.1727426713317285e-05, + "loss": 0.505, + "step": 14630 + }, + { + "epoch": 0.46147295379277714, + "grad_norm": 1.498369574546814, + "learning_rate": 1.1726420504891862e-05, + "loss": 0.4521, + "step": 14631 + }, + { + "epoch": 0.46150449455921777, + "grad_norm": 1.561104416847229, + "learning_rate": 1.1725414278449904e-05, + "loss": 0.4523, + "step": 14632 + }, + { + "epoch": 0.4615360353256584, + "grad_norm": 2.5967957973480225, + "learning_rate": 1.1724408034001911e-05, + "loss": 0.4959, + "step": 14633 + }, + { + "epoch": 0.46156757609209903, + "grad_norm": 1.4856603145599365, + "learning_rate": 1.1723401771558376e-05, + "loss": 0.5121, + "step": 14634 + }, + { + "epoch": 0.46159911685853966, + "grad_norm": 1.582167387008667, + "learning_rate": 1.1722395491129807e-05, + "loss": 0.8097, + "step": 14635 + }, + { + "epoch": 0.4616306576249803, + "grad_norm": 1.3595733642578125, + "learning_rate": 1.1721389192726705e-05, + "loss": 0.4745, + "step": 14636 + }, + { + "epoch": 0.4616621983914209, + "grad_norm": 2.2560343742370605, + "learning_rate": 1.1720382876359567e-05, + "loss": 0.4561, + "step": 14637 + }, + { + "epoch": 0.46169373915786155, + "grad_norm": 1.9513025283813477, + "learning_rate": 1.17193765420389e-05, + "loss": 0.3836, + "step": 14638 + }, + { + "epoch": 0.4617252799243022, + "grad_norm": 1.522961139678955, + "learning_rate": 1.1718370189775203e-05, + "loss": 0.4867, + "step": 14639 + }, + { + "epoch": 0.4617568206907428, + "grad_norm": 2.023639678955078, + "learning_rate": 1.1717363819578977e-05, + "loss": 0.5018, + "step": 14640 + }, + { + "epoch": 0.4617883614571834, + "grad_norm": 2.024752140045166, + "learning_rate": 1.1716357431460727e-05, + "loss": 0.4834, + "step": 14641 + }, + { + "epoch": 0.461819902223624, + "grad_norm": 1.283174753189087, + "learning_rate": 1.1715351025430959e-05, + "loss": 0.4754, + "step": 14642 + }, + { + "epoch": 0.46185144299006464, + "grad_norm": 1.673110842704773, + "learning_rate": 1.1714344601500167e-05, + "loss": 0.4723, + "step": 14643 + }, + { + "epoch": 0.46188298375650527, + "grad_norm": 2.433936834335327, + "learning_rate": 1.1713338159678859e-05, + "loss": 0.4426, + "step": 14644 + }, + { + "epoch": 0.4619145245229459, + "grad_norm": 1.493129849433899, + "learning_rate": 1.1712331699977537e-05, + "loss": 0.514, + "step": 14645 + }, + { + "epoch": 0.46194606528938653, + "grad_norm": 2.0307397842407227, + "learning_rate": 1.1711325222406704e-05, + "loss": 0.3807, + "step": 14646 + }, + { + "epoch": 0.46197760605582716, + "grad_norm": 2.2012205123901367, + "learning_rate": 1.1710318726976862e-05, + "loss": 0.816, + "step": 14647 + }, + { + "epoch": 0.4620091468222678, + "grad_norm": 2.096975803375244, + "learning_rate": 1.1709312213698516e-05, + "loss": 0.487, + "step": 14648 + }, + { + "epoch": 0.4620406875887084, + "grad_norm": 1.9673330783843994, + "learning_rate": 1.1708305682582173e-05, + "loss": 0.485, + "step": 14649 + }, + { + "epoch": 0.46207222835514905, + "grad_norm": 1.7104628086090088, + "learning_rate": 1.170729913363833e-05, + "loss": 0.499, + "step": 14650 + }, + { + "epoch": 0.4621037691215897, + "grad_norm": 1.5034279823303223, + "learning_rate": 1.1706292566877494e-05, + "loss": 0.4788, + "step": 14651 + }, + { + "epoch": 0.46213530988803025, + "grad_norm": 2.2291417121887207, + "learning_rate": 1.1705285982310176e-05, + "loss": 0.4789, + "step": 14652 + }, + { + "epoch": 0.4621668506544709, + "grad_norm": 1.7070425748825073, + "learning_rate": 1.1704279379946872e-05, + "loss": 0.4915, + "step": 14653 + }, + { + "epoch": 0.4621983914209115, + "grad_norm": 1.657209038734436, + "learning_rate": 1.1703272759798088e-05, + "loss": 0.4785, + "step": 14654 + }, + { + "epoch": 0.46222993218735214, + "grad_norm": 5.465292930603027, + "learning_rate": 1.1702266121874331e-05, + "loss": 0.4646, + "step": 14655 + }, + { + "epoch": 0.4622614729537928, + "grad_norm": 1.9119309186935425, + "learning_rate": 1.1701259466186106e-05, + "loss": 0.5151, + "step": 14656 + }, + { + "epoch": 0.4622930137202334, + "grad_norm": 1.5471028089523315, + "learning_rate": 1.1700252792743916e-05, + "loss": 0.5076, + "step": 14657 + }, + { + "epoch": 0.46232455448667403, + "grad_norm": 2.798030138015747, + "learning_rate": 1.1699246101558268e-05, + "loss": 0.4748, + "step": 14658 + }, + { + "epoch": 0.46235609525311466, + "grad_norm": 1.747416377067566, + "learning_rate": 1.169823939263967e-05, + "loss": 0.4878, + "step": 14659 + }, + { + "epoch": 0.4623876360195553, + "grad_norm": 2.537754774093628, + "learning_rate": 1.1697232665998621e-05, + "loss": 0.3571, + "step": 14660 + }, + { + "epoch": 0.4624191767859959, + "grad_norm": 1.935557246208191, + "learning_rate": 1.1696225921645635e-05, + "loss": 0.4953, + "step": 14661 + }, + { + "epoch": 0.4624507175524365, + "grad_norm": 1.900920033454895, + "learning_rate": 1.1695219159591214e-05, + "loss": 0.5184, + "step": 14662 + }, + { + "epoch": 0.4624822583188771, + "grad_norm": 2.7698614597320557, + "learning_rate": 1.1694212379845864e-05, + "loss": 0.5011, + "step": 14663 + }, + { + "epoch": 0.46251379908531776, + "grad_norm": 2.537543296813965, + "learning_rate": 1.169320558242009e-05, + "loss": 0.3537, + "step": 14664 + }, + { + "epoch": 0.4625453398517584, + "grad_norm": 1.479791283607483, + "learning_rate": 1.1692198767324403e-05, + "loss": 0.4988, + "step": 14665 + }, + { + "epoch": 0.462576880618199, + "grad_norm": 1.4934884309768677, + "learning_rate": 1.1691191934569311e-05, + "loss": 0.471, + "step": 14666 + }, + { + "epoch": 0.46260842138463965, + "grad_norm": 1.4248238801956177, + "learning_rate": 1.1690185084165313e-05, + "loss": 0.4806, + "step": 14667 + }, + { + "epoch": 0.4626399621510803, + "grad_norm": 2.470473289489746, + "learning_rate": 1.1689178216122923e-05, + "loss": 0.4099, + "step": 14668 + }, + { + "epoch": 0.4626715029175209, + "grad_norm": 1.6579561233520508, + "learning_rate": 1.1688171330452647e-05, + "loss": 0.8079, + "step": 14669 + }, + { + "epoch": 0.46270304368396153, + "grad_norm": 2.521984577178955, + "learning_rate": 1.1687164427164992e-05, + "loss": 0.4889, + "step": 14670 + }, + { + "epoch": 0.46273458445040216, + "grad_norm": 1.4988398551940918, + "learning_rate": 1.1686157506270465e-05, + "loss": 0.7818, + "step": 14671 + }, + { + "epoch": 0.4627661252168428, + "grad_norm": 1.719344973564148, + "learning_rate": 1.1685150567779577e-05, + "loss": 0.4696, + "step": 14672 + }, + { + "epoch": 0.46279766598328337, + "grad_norm": 1.9369767904281616, + "learning_rate": 1.1684143611702837e-05, + "loss": 0.5307, + "step": 14673 + }, + { + "epoch": 0.462829206749724, + "grad_norm": 1.8919754028320312, + "learning_rate": 1.1683136638050748e-05, + "loss": 0.4313, + "step": 14674 + }, + { + "epoch": 0.46286074751616463, + "grad_norm": 3.2745516300201416, + "learning_rate": 1.1682129646833817e-05, + "loss": 0.5123, + "step": 14675 + }, + { + "epoch": 0.46289228828260526, + "grad_norm": 4.455129146575928, + "learning_rate": 1.1681122638062563e-05, + "loss": 0.3672, + "step": 14676 + }, + { + "epoch": 0.4629238290490459, + "grad_norm": 1.7529319524765015, + "learning_rate": 1.1680115611747487e-05, + "loss": 0.4923, + "step": 14677 + }, + { + "epoch": 0.4629553698154865, + "grad_norm": 1.7432588338851929, + "learning_rate": 1.1679108567899097e-05, + "loss": 0.4592, + "step": 14678 + }, + { + "epoch": 0.46298691058192715, + "grad_norm": 3.0908493995666504, + "learning_rate": 1.1678101506527908e-05, + "loss": 0.5162, + "step": 14679 + }, + { + "epoch": 0.4630184513483678, + "grad_norm": 1.5548151731491089, + "learning_rate": 1.1677094427644428e-05, + "loss": 0.4828, + "step": 14680 + }, + { + "epoch": 0.4630499921148084, + "grad_norm": 1.695857286453247, + "learning_rate": 1.1676087331259166e-05, + "loss": 0.5166, + "step": 14681 + }, + { + "epoch": 0.46308153288124904, + "grad_norm": 1.5555542707443237, + "learning_rate": 1.1675080217382628e-05, + "loss": 0.4937, + "step": 14682 + }, + { + "epoch": 0.46311307364768967, + "grad_norm": 1.7209267616271973, + "learning_rate": 1.167407308602533e-05, + "loss": 0.4883, + "step": 14683 + }, + { + "epoch": 0.46314461441413024, + "grad_norm": 2.485570192337036, + "learning_rate": 1.1673065937197778e-05, + "loss": 0.3813, + "step": 14684 + }, + { + "epoch": 0.46317615518057087, + "grad_norm": 1.4595576524734497, + "learning_rate": 1.1672058770910484e-05, + "loss": 0.49, + "step": 14685 + }, + { + "epoch": 0.4632076959470115, + "grad_norm": 1.6101603507995605, + "learning_rate": 1.1671051587173957e-05, + "loss": 0.5046, + "step": 14686 + }, + { + "epoch": 0.46323923671345213, + "grad_norm": 1.7550591230392456, + "learning_rate": 1.1670044385998713e-05, + "loss": 0.4628, + "step": 14687 + }, + { + "epoch": 0.46327077747989276, + "grad_norm": 1.559128999710083, + "learning_rate": 1.1669037167395256e-05, + "loss": 0.4958, + "step": 14688 + }, + { + "epoch": 0.4633023182463334, + "grad_norm": 1.3832933902740479, + "learning_rate": 1.1668029931374098e-05, + "loss": 0.4972, + "step": 14689 + }, + { + "epoch": 0.463333859012774, + "grad_norm": 1.7158708572387695, + "learning_rate": 1.1667022677945759e-05, + "loss": 0.4729, + "step": 14690 + }, + { + "epoch": 0.46336539977921465, + "grad_norm": 1.3925219774246216, + "learning_rate": 1.1666015407120739e-05, + "loss": 0.4717, + "step": 14691 + }, + { + "epoch": 0.4633969405456553, + "grad_norm": 2.0836915969848633, + "learning_rate": 1.1665008118909558e-05, + "loss": 0.4709, + "step": 14692 + }, + { + "epoch": 0.4634284813120959, + "grad_norm": 1.4898642301559448, + "learning_rate": 1.1664000813322722e-05, + "loss": 0.5096, + "step": 14693 + }, + { + "epoch": 0.4634600220785365, + "grad_norm": 1.7794597148895264, + "learning_rate": 1.1662993490370747e-05, + "loss": 0.45, + "step": 14694 + }, + { + "epoch": 0.4634915628449771, + "grad_norm": 1.524661898612976, + "learning_rate": 1.166198615006414e-05, + "loss": 0.5066, + "step": 14695 + }, + { + "epoch": 0.46352310361141774, + "grad_norm": 2.4913196563720703, + "learning_rate": 1.166097879241342e-05, + "loss": 0.5185, + "step": 14696 + }, + { + "epoch": 0.46355464437785837, + "grad_norm": 1.5509999990463257, + "learning_rate": 1.1659971417429096e-05, + "loss": 0.4833, + "step": 14697 + }, + { + "epoch": 0.463586185144299, + "grad_norm": 1.4667938947677612, + "learning_rate": 1.1658964025121682e-05, + "loss": 0.5061, + "step": 14698 + }, + { + "epoch": 0.46361772591073963, + "grad_norm": 2.8514904975891113, + "learning_rate": 1.1657956615501689e-05, + "loss": 0.4831, + "step": 14699 + }, + { + "epoch": 0.46364926667718026, + "grad_norm": 1.605126142501831, + "learning_rate": 1.1656949188579632e-05, + "loss": 0.4604, + "step": 14700 + }, + { + "epoch": 0.4636808074436209, + "grad_norm": 2.121725082397461, + "learning_rate": 1.1655941744366024e-05, + "loss": 0.4835, + "step": 14701 + }, + { + "epoch": 0.4637123482100615, + "grad_norm": 1.4441618919372559, + "learning_rate": 1.1654934282871378e-05, + "loss": 0.4718, + "step": 14702 + }, + { + "epoch": 0.46374388897650215, + "grad_norm": 3.7697460651397705, + "learning_rate": 1.1653926804106207e-05, + "loss": 0.5059, + "step": 14703 + }, + { + "epoch": 0.4637754297429428, + "grad_norm": 1.8071578741073608, + "learning_rate": 1.1652919308081028e-05, + "loss": 0.7812, + "step": 14704 + }, + { + "epoch": 0.46380697050938335, + "grad_norm": 1.391876220703125, + "learning_rate": 1.165191179480635e-05, + "loss": 0.4643, + "step": 14705 + }, + { + "epoch": 0.463838511275824, + "grad_norm": 2.2001447677612305, + "learning_rate": 1.1650904264292689e-05, + "loss": 0.4464, + "step": 14706 + }, + { + "epoch": 0.4638700520422646, + "grad_norm": 1.912932276725769, + "learning_rate": 1.1649896716550562e-05, + "loss": 0.4732, + "step": 14707 + }, + { + "epoch": 0.46390159280870524, + "grad_norm": 1.5188289880752563, + "learning_rate": 1.164888915159048e-05, + "loss": 0.5332, + "step": 14708 + }, + { + "epoch": 0.4639331335751459, + "grad_norm": 1.5347949266433716, + "learning_rate": 1.1647881569422961e-05, + "loss": 0.4791, + "step": 14709 + }, + { + "epoch": 0.4639646743415865, + "grad_norm": 1.9316715002059937, + "learning_rate": 1.164687397005852e-05, + "loss": 0.5095, + "step": 14710 + }, + { + "epoch": 0.46399621510802713, + "grad_norm": 1.2964253425598145, + "learning_rate": 1.1645866353507664e-05, + "loss": 0.4971, + "step": 14711 + }, + { + "epoch": 0.46402775587446776, + "grad_norm": 1.887245774269104, + "learning_rate": 1.164485871978092e-05, + "loss": 0.4952, + "step": 14712 + }, + { + "epoch": 0.4640592966409084, + "grad_norm": 1.8704750537872314, + "learning_rate": 1.1643851068888796e-05, + "loss": 0.5106, + "step": 14713 + }, + { + "epoch": 0.464090837407349, + "grad_norm": 1.727797031402588, + "learning_rate": 1.1642843400841812e-05, + "loss": 0.4976, + "step": 14714 + }, + { + "epoch": 0.4641223781737896, + "grad_norm": 2.1513993740081787, + "learning_rate": 1.164183571565048e-05, + "loss": 0.5047, + "step": 14715 + }, + { + "epoch": 0.4641539189402302, + "grad_norm": 1.2990717887878418, + "learning_rate": 1.1640828013325316e-05, + "loss": 0.4697, + "step": 14716 + }, + { + "epoch": 0.46418545970667086, + "grad_norm": 1.354347825050354, + "learning_rate": 1.163982029387684e-05, + "loss": 0.3774, + "step": 14717 + }, + { + "epoch": 0.4642170004731115, + "grad_norm": 1.5565533638000488, + "learning_rate": 1.1638812557315564e-05, + "loss": 0.4985, + "step": 14718 + }, + { + "epoch": 0.4642485412395521, + "grad_norm": 1.7473710775375366, + "learning_rate": 1.1637804803652006e-05, + "loss": 0.4975, + "step": 14719 + }, + { + "epoch": 0.46428008200599274, + "grad_norm": 1.8558839559555054, + "learning_rate": 1.1636797032896688e-05, + "loss": 0.4819, + "step": 14720 + }, + { + "epoch": 0.4643116227724334, + "grad_norm": 1.9753658771514893, + "learning_rate": 1.1635789245060116e-05, + "loss": 0.4875, + "step": 14721 + }, + { + "epoch": 0.464343163538874, + "grad_norm": 3.180647373199463, + "learning_rate": 1.1634781440152818e-05, + "loss": 0.493, + "step": 14722 + }, + { + "epoch": 0.46437470430531463, + "grad_norm": 1.7363051176071167, + "learning_rate": 1.1633773618185302e-05, + "loss": 0.4635, + "step": 14723 + }, + { + "epoch": 0.46440624507175526, + "grad_norm": 1.8330250978469849, + "learning_rate": 1.1632765779168094e-05, + "loss": 0.4613, + "step": 14724 + }, + { + "epoch": 0.4644377858381959, + "grad_norm": 1.6155173778533936, + "learning_rate": 1.1631757923111703e-05, + "loss": 0.4599, + "step": 14725 + }, + { + "epoch": 0.46446932660463647, + "grad_norm": 1.4222944974899292, + "learning_rate": 1.1630750050026654e-05, + "loss": 0.482, + "step": 14726 + }, + { + "epoch": 0.4645008673710771, + "grad_norm": 2.26273775100708, + "learning_rate": 1.162974215992346e-05, + "loss": 0.5154, + "step": 14727 + }, + { + "epoch": 0.4645324081375177, + "grad_norm": 2.684849739074707, + "learning_rate": 1.1628734252812644e-05, + "loss": 0.4744, + "step": 14728 + }, + { + "epoch": 0.46456394890395836, + "grad_norm": 1.604920744895935, + "learning_rate": 1.1627726328704718e-05, + "loss": 0.488, + "step": 14729 + }, + { + "epoch": 0.464595489670399, + "grad_norm": 2.1794450283050537, + "learning_rate": 1.1626718387610206e-05, + "loss": 0.4637, + "step": 14730 + }, + { + "epoch": 0.4646270304368396, + "grad_norm": 1.752565860748291, + "learning_rate": 1.1625710429539623e-05, + "loss": 0.4846, + "step": 14731 + }, + { + "epoch": 0.46465857120328025, + "grad_norm": 2.032918691635132, + "learning_rate": 1.1624702454503491e-05, + "loss": 0.5106, + "step": 14732 + }, + { + "epoch": 0.4646901119697209, + "grad_norm": 1.4543132781982422, + "learning_rate": 1.1623694462512328e-05, + "loss": 0.5016, + "step": 14733 + }, + { + "epoch": 0.4647216527361615, + "grad_norm": 1.4017318487167358, + "learning_rate": 1.1622686453576648e-05, + "loss": 0.4621, + "step": 14734 + }, + { + "epoch": 0.46475319350260214, + "grad_norm": 1.5118192434310913, + "learning_rate": 1.162167842770698e-05, + "loss": 0.5307, + "step": 14735 + }, + { + "epoch": 0.46478473426904277, + "grad_norm": 1.9943007230758667, + "learning_rate": 1.1620670384913835e-05, + "loss": 0.4649, + "step": 14736 + }, + { + "epoch": 0.46481627503548334, + "grad_norm": 1.6369457244873047, + "learning_rate": 1.1619662325207737e-05, + "loss": 0.4575, + "step": 14737 + }, + { + "epoch": 0.46484781580192397, + "grad_norm": 1.6857130527496338, + "learning_rate": 1.1618654248599207e-05, + "loss": 0.4938, + "step": 14738 + }, + { + "epoch": 0.4648793565683646, + "grad_norm": 1.441329836845398, + "learning_rate": 1.1617646155098761e-05, + "loss": 0.476, + "step": 14739 + }, + { + "epoch": 0.46491089733480523, + "grad_norm": 2.0675876140594482, + "learning_rate": 1.1616638044716921e-05, + "loss": 0.457, + "step": 14740 + }, + { + "epoch": 0.46494243810124586, + "grad_norm": 1.7875137329101562, + "learning_rate": 1.1615629917464208e-05, + "loss": 0.4592, + "step": 14741 + }, + { + "epoch": 0.4649739788676865, + "grad_norm": 1.3348236083984375, + "learning_rate": 1.1614621773351145e-05, + "loss": 0.494, + "step": 14742 + }, + { + "epoch": 0.4650055196341271, + "grad_norm": 1.8415688276290894, + "learning_rate": 1.1613613612388245e-05, + "loss": 0.5019, + "step": 14743 + }, + { + "epoch": 0.46503706040056775, + "grad_norm": 1.4853674173355103, + "learning_rate": 1.1612605434586037e-05, + "loss": 0.5007, + "step": 14744 + }, + { + "epoch": 0.4650686011670084, + "grad_norm": 1.61916184425354, + "learning_rate": 1.1611597239955041e-05, + "loss": 0.4219, + "step": 14745 + }, + { + "epoch": 0.465100141933449, + "grad_norm": 1.8619076013565063, + "learning_rate": 1.1610589028505774e-05, + "loss": 0.4862, + "step": 14746 + }, + { + "epoch": 0.4651316826998896, + "grad_norm": 1.5204997062683105, + "learning_rate": 1.1609580800248758e-05, + "loss": 0.7829, + "step": 14747 + }, + { + "epoch": 0.4651632234663302, + "grad_norm": 1.2238024473190308, + "learning_rate": 1.1608572555194521e-05, + "loss": 0.3832, + "step": 14748 + }, + { + "epoch": 0.46519476423277084, + "grad_norm": 2.4503889083862305, + "learning_rate": 1.1607564293353577e-05, + "loss": 0.4844, + "step": 14749 + }, + { + "epoch": 0.46522630499921147, + "grad_norm": 2.9730942249298096, + "learning_rate": 1.1606556014736454e-05, + "loss": 0.4971, + "step": 14750 + }, + { + "epoch": 0.4652578457656521, + "grad_norm": 1.9643105268478394, + "learning_rate": 1.1605547719353672e-05, + "loss": 0.3737, + "step": 14751 + }, + { + "epoch": 0.46528938653209273, + "grad_norm": 1.5434033870697021, + "learning_rate": 1.1604539407215752e-05, + "loss": 0.4971, + "step": 14752 + }, + { + "epoch": 0.46532092729853336, + "grad_norm": 1.9293092489242554, + "learning_rate": 1.1603531078333214e-05, + "loss": 0.4697, + "step": 14753 + }, + { + "epoch": 0.465352468064974, + "grad_norm": 1.7641735076904297, + "learning_rate": 1.1602522732716586e-05, + "loss": 0.3683, + "step": 14754 + }, + { + "epoch": 0.4653840088314146, + "grad_norm": 1.4477492570877075, + "learning_rate": 1.160151437037639e-05, + "loss": 0.5035, + "step": 14755 + }, + { + "epoch": 0.46541554959785525, + "grad_norm": 1.906369924545288, + "learning_rate": 1.1600505991323149e-05, + "loss": 0.5088, + "step": 14756 + }, + { + "epoch": 0.4654470903642959, + "grad_norm": 1.7717702388763428, + "learning_rate": 1.1599497595567381e-05, + "loss": 0.4675, + "step": 14757 + }, + { + "epoch": 0.46547863113073645, + "grad_norm": 1.675190806388855, + "learning_rate": 1.1598489183119615e-05, + "loss": 0.4656, + "step": 14758 + }, + { + "epoch": 0.4655101718971771, + "grad_norm": 1.4827643632888794, + "learning_rate": 1.1597480753990375e-05, + "loss": 0.4877, + "step": 14759 + }, + { + "epoch": 0.4655417126636177, + "grad_norm": 2.676156759262085, + "learning_rate": 1.1596472308190185e-05, + "loss": 0.4815, + "step": 14760 + }, + { + "epoch": 0.46557325343005834, + "grad_norm": 1.6180698871612549, + "learning_rate": 1.1595463845729565e-05, + "loss": 0.4852, + "step": 14761 + }, + { + "epoch": 0.465604794196499, + "grad_norm": 1.5541588068008423, + "learning_rate": 1.159445536661904e-05, + "loss": 0.4681, + "step": 14762 + }, + { + "epoch": 0.4656363349629396, + "grad_norm": 1.4395551681518555, + "learning_rate": 1.1593446870869134e-05, + "loss": 0.8305, + "step": 14763 + }, + { + "epoch": 0.46566787572938023, + "grad_norm": 1.5963126420974731, + "learning_rate": 1.1592438358490373e-05, + "loss": 0.4919, + "step": 14764 + }, + { + "epoch": 0.46569941649582086, + "grad_norm": 2.1483333110809326, + "learning_rate": 1.1591429829493283e-05, + "loss": 0.471, + "step": 14765 + }, + { + "epoch": 0.4657309572622615, + "grad_norm": 1.7844758033752441, + "learning_rate": 1.1590421283888387e-05, + "loss": 0.4999, + "step": 14766 + }, + { + "epoch": 0.4657624980287021, + "grad_norm": 1.6311291456222534, + "learning_rate": 1.1589412721686207e-05, + "loss": 0.5019, + "step": 14767 + }, + { + "epoch": 0.4657940387951427, + "grad_norm": 2.660754919052124, + "learning_rate": 1.1588404142897274e-05, + "loss": 0.4683, + "step": 14768 + }, + { + "epoch": 0.4658255795615833, + "grad_norm": 1.4129928350448608, + "learning_rate": 1.158739554753211e-05, + "loss": 0.507, + "step": 14769 + }, + { + "epoch": 0.46585712032802395, + "grad_norm": 1.7405169010162354, + "learning_rate": 1.1586386935601241e-05, + "loss": 0.5044, + "step": 14770 + }, + { + "epoch": 0.4658886610944646, + "grad_norm": 2.2349376678466797, + "learning_rate": 1.158537830711519e-05, + "loss": 0.4992, + "step": 14771 + }, + { + "epoch": 0.4659202018609052, + "grad_norm": 1.4539583921432495, + "learning_rate": 1.1584369662084491e-05, + "loss": 0.4918, + "step": 14772 + }, + { + "epoch": 0.46595174262734584, + "grad_norm": 1.6009093523025513, + "learning_rate": 1.158336100051966e-05, + "loss": 0.7973, + "step": 14773 + }, + { + "epoch": 0.4659832833937865, + "grad_norm": 2.0646121501922607, + "learning_rate": 1.1582352322431228e-05, + "loss": 0.4972, + "step": 14774 + }, + { + "epoch": 0.4660148241602271, + "grad_norm": 1.7606562376022339, + "learning_rate": 1.1581343627829719e-05, + "loss": 0.5139, + "step": 14775 + }, + { + "epoch": 0.46604636492666773, + "grad_norm": 1.719749093055725, + "learning_rate": 1.1580334916725666e-05, + "loss": 0.4812, + "step": 14776 + }, + { + "epoch": 0.46607790569310836, + "grad_norm": 1.612082600593567, + "learning_rate": 1.1579326189129588e-05, + "loss": 0.4842, + "step": 14777 + }, + { + "epoch": 0.466109446459549, + "grad_norm": 1.2992706298828125, + "learning_rate": 1.1578317445052014e-05, + "loss": 0.4908, + "step": 14778 + }, + { + "epoch": 0.46614098722598957, + "grad_norm": 2.8886196613311768, + "learning_rate": 1.1577308684503475e-05, + "loss": 0.5298, + "step": 14779 + }, + { + "epoch": 0.4661725279924302, + "grad_norm": 1.5760468244552612, + "learning_rate": 1.1576299907494495e-05, + "loss": 0.5115, + "step": 14780 + }, + { + "epoch": 0.4662040687588708, + "grad_norm": 1.648851990699768, + "learning_rate": 1.1575291114035598e-05, + "loss": 0.4876, + "step": 14781 + }, + { + "epoch": 0.46623560952531146, + "grad_norm": 2.7707717418670654, + "learning_rate": 1.1574282304137316e-05, + "loss": 0.4966, + "step": 14782 + }, + { + "epoch": 0.4662671502917521, + "grad_norm": 1.4093042612075806, + "learning_rate": 1.1573273477810179e-05, + "loss": 0.4912, + "step": 14783 + }, + { + "epoch": 0.4662986910581927, + "grad_norm": 1.8247076272964478, + "learning_rate": 1.1572264635064706e-05, + "loss": 0.501, + "step": 14784 + }, + { + "epoch": 0.46633023182463335, + "grad_norm": 1.574821949005127, + "learning_rate": 1.1571255775911437e-05, + "loss": 0.5137, + "step": 14785 + }, + { + "epoch": 0.466361772591074, + "grad_norm": 1.3432443141937256, + "learning_rate": 1.157024690036089e-05, + "loss": 0.5262, + "step": 14786 + }, + { + "epoch": 0.4663933133575146, + "grad_norm": 1.7510837316513062, + "learning_rate": 1.1569238008423599e-05, + "loss": 0.4795, + "step": 14787 + }, + { + "epoch": 0.46642485412395523, + "grad_norm": 1.4477272033691406, + "learning_rate": 1.1568229100110089e-05, + "loss": 0.4972, + "step": 14788 + }, + { + "epoch": 0.4664563948903958, + "grad_norm": 1.787856936454773, + "learning_rate": 1.1567220175430892e-05, + "loss": 0.4879, + "step": 14789 + }, + { + "epoch": 0.46648793565683644, + "grad_norm": 1.3179099559783936, + "learning_rate": 1.1566211234396539e-05, + "loss": 0.8169, + "step": 14790 + }, + { + "epoch": 0.46651947642327707, + "grad_norm": 1.7775278091430664, + "learning_rate": 1.1565202277017551e-05, + "loss": 0.7863, + "step": 14791 + }, + { + "epoch": 0.4665510171897177, + "grad_norm": 1.3359836339950562, + "learning_rate": 1.1564193303304461e-05, + "loss": 0.4674, + "step": 14792 + }, + { + "epoch": 0.46658255795615833, + "grad_norm": 1.918340802192688, + "learning_rate": 1.1563184313267803e-05, + "loss": 0.504, + "step": 14793 + }, + { + "epoch": 0.46661409872259896, + "grad_norm": 1.6140929460525513, + "learning_rate": 1.15621753069181e-05, + "loss": 0.5142, + "step": 14794 + }, + { + "epoch": 0.4666456394890396, + "grad_norm": 1.5742096900939941, + "learning_rate": 1.1561166284265886e-05, + "loss": 0.4155, + "step": 14795 + }, + { + "epoch": 0.4666771802554802, + "grad_norm": 1.7897297143936157, + "learning_rate": 1.1560157245321691e-05, + "loss": 0.4855, + "step": 14796 + }, + { + "epoch": 0.46670872102192085, + "grad_norm": 1.5562583208084106, + "learning_rate": 1.155914819009604e-05, + "loss": 0.5213, + "step": 14797 + }, + { + "epoch": 0.4667402617883615, + "grad_norm": 1.3587108850479126, + "learning_rate": 1.155813911859947e-05, + "loss": 0.4975, + "step": 14798 + }, + { + "epoch": 0.4667718025548021, + "grad_norm": 1.5955110788345337, + "learning_rate": 1.155713003084251e-05, + "loss": 0.4526, + "step": 14799 + }, + { + "epoch": 0.4668033433212427, + "grad_norm": 1.6127324104309082, + "learning_rate": 1.1556120926835686e-05, + "loss": 0.5125, + "step": 14800 + }, + { + "epoch": 0.4668348840876833, + "grad_norm": 1.9004448652267456, + "learning_rate": 1.1555111806589532e-05, + "loss": 0.4202, + "step": 14801 + }, + { + "epoch": 0.46686642485412394, + "grad_norm": 1.2802369594573975, + "learning_rate": 1.155410267011458e-05, + "loss": 0.3714, + "step": 14802 + }, + { + "epoch": 0.46689796562056457, + "grad_norm": 1.8237197399139404, + "learning_rate": 1.1553093517421359e-05, + "loss": 0.4936, + "step": 14803 + }, + { + "epoch": 0.4669295063870052, + "grad_norm": 1.715338945388794, + "learning_rate": 1.15520843485204e-05, + "loss": 0.4698, + "step": 14804 + }, + { + "epoch": 0.46696104715344583, + "grad_norm": 1.175065040588379, + "learning_rate": 1.155107516342224e-05, + "loss": 0.4603, + "step": 14805 + }, + { + "epoch": 0.46699258791988646, + "grad_norm": 1.5453276634216309, + "learning_rate": 1.15500659621374e-05, + "loss": 0.5099, + "step": 14806 + }, + { + "epoch": 0.4670241286863271, + "grad_norm": 1.7159364223480225, + "learning_rate": 1.1549056744676424e-05, + "loss": 0.4626, + "step": 14807 + }, + { + "epoch": 0.4670556694527677, + "grad_norm": 1.7709726095199585, + "learning_rate": 1.1548047511049833e-05, + "loss": 0.4575, + "step": 14808 + }, + { + "epoch": 0.46708721021920835, + "grad_norm": 1.7032119035720825, + "learning_rate": 1.1547038261268168e-05, + "loss": 0.488, + "step": 14809 + }, + { + "epoch": 0.467118750985649, + "grad_norm": 1.4491511583328247, + "learning_rate": 1.1546028995341958e-05, + "loss": 0.5154, + "step": 14810 + }, + { + "epoch": 0.46715029175208955, + "grad_norm": 1.891211986541748, + "learning_rate": 1.1545019713281733e-05, + "loss": 0.4948, + "step": 14811 + }, + { + "epoch": 0.4671818325185302, + "grad_norm": 1.3844730854034424, + "learning_rate": 1.1544010415098026e-05, + "loss": 0.4766, + "step": 14812 + }, + { + "epoch": 0.4672133732849708, + "grad_norm": 1.3108826875686646, + "learning_rate": 1.1543001100801375e-05, + "loss": 0.4688, + "step": 14813 + }, + { + "epoch": 0.46724491405141144, + "grad_norm": 1.2476879358291626, + "learning_rate": 1.1541991770402308e-05, + "loss": 0.4856, + "step": 14814 + }, + { + "epoch": 0.46727645481785207, + "grad_norm": 1.4939260482788086, + "learning_rate": 1.1540982423911358e-05, + "loss": 0.5028, + "step": 14815 + }, + { + "epoch": 0.4673079955842927, + "grad_norm": 1.4761852025985718, + "learning_rate": 1.153997306133906e-05, + "loss": 0.4638, + "step": 14816 + }, + { + "epoch": 0.46733953635073333, + "grad_norm": 1.5016478300094604, + "learning_rate": 1.153896368269595e-05, + "loss": 0.4785, + "step": 14817 + }, + { + "epoch": 0.46737107711717396, + "grad_norm": 1.5053434371948242, + "learning_rate": 1.1537954287992557e-05, + "loss": 0.4911, + "step": 14818 + }, + { + "epoch": 0.4674026178836146, + "grad_norm": 1.3409713506698608, + "learning_rate": 1.1536944877239418e-05, + "loss": 0.4506, + "step": 14819 + }, + { + "epoch": 0.4674341586500552, + "grad_norm": 1.494674563407898, + "learning_rate": 1.1535935450447066e-05, + "loss": 0.4812, + "step": 14820 + }, + { + "epoch": 0.4674656994164958, + "grad_norm": 1.6218245029449463, + "learning_rate": 1.1534926007626032e-05, + "loss": 0.508, + "step": 14821 + }, + { + "epoch": 0.4674972401829364, + "grad_norm": 1.5756080150604248, + "learning_rate": 1.1533916548786856e-05, + "loss": 0.5108, + "step": 14822 + }, + { + "epoch": 0.46752878094937705, + "grad_norm": 1.2391271591186523, + "learning_rate": 1.1532907073940068e-05, + "loss": 0.4565, + "step": 14823 + }, + { + "epoch": 0.4675603217158177, + "grad_norm": 1.4342797994613647, + "learning_rate": 1.1531897583096206e-05, + "loss": 0.4815, + "step": 14824 + }, + { + "epoch": 0.4675918624822583, + "grad_norm": 1.7647080421447754, + "learning_rate": 1.1530888076265802e-05, + "loss": 0.4694, + "step": 14825 + }, + { + "epoch": 0.46762340324869894, + "grad_norm": 1.3270325660705566, + "learning_rate": 1.1529878553459393e-05, + "loss": 0.4379, + "step": 14826 + }, + { + "epoch": 0.4676549440151396, + "grad_norm": 1.4627984762191772, + "learning_rate": 1.1528869014687513e-05, + "loss": 0.4938, + "step": 14827 + }, + { + "epoch": 0.4676864847815802, + "grad_norm": 3.1166207790374756, + "learning_rate": 1.1527859459960698e-05, + "loss": 0.514, + "step": 14828 + }, + { + "epoch": 0.46771802554802083, + "grad_norm": 1.3898764848709106, + "learning_rate": 1.1526849889289485e-05, + "loss": 0.4905, + "step": 14829 + }, + { + "epoch": 0.46774956631446146, + "grad_norm": 1.7859054803848267, + "learning_rate": 1.1525840302684403e-05, + "loss": 0.5036, + "step": 14830 + }, + { + "epoch": 0.4677811070809021, + "grad_norm": 1.5117186307907104, + "learning_rate": 1.1524830700155998e-05, + "loss": 0.4446, + "step": 14831 + }, + { + "epoch": 0.46781264784734267, + "grad_norm": 1.5404669046401978, + "learning_rate": 1.1523821081714797e-05, + "loss": 0.4643, + "step": 14832 + }, + { + "epoch": 0.4678441886137833, + "grad_norm": 1.2983304262161255, + "learning_rate": 1.152281144737134e-05, + "loss": 0.4709, + "step": 14833 + }, + { + "epoch": 0.4678757293802239, + "grad_norm": 1.608253002166748, + "learning_rate": 1.1521801797136167e-05, + "loss": 0.4989, + "step": 14834 + }, + { + "epoch": 0.46790727014666456, + "grad_norm": 2.1084940433502197, + "learning_rate": 1.1520792131019806e-05, + "loss": 0.476, + "step": 14835 + }, + { + "epoch": 0.4679388109131052, + "grad_norm": 1.8895941972732544, + "learning_rate": 1.15197824490328e-05, + "loss": 0.7916, + "step": 14836 + }, + { + "epoch": 0.4679703516795458, + "grad_norm": 1.5203558206558228, + "learning_rate": 1.1518772751185685e-05, + "loss": 0.5071, + "step": 14837 + }, + { + "epoch": 0.46800189244598644, + "grad_norm": 1.3890788555145264, + "learning_rate": 1.1517763037488996e-05, + "loss": 0.4672, + "step": 14838 + }, + { + "epoch": 0.4680334332124271, + "grad_norm": 1.9272540807724, + "learning_rate": 1.1516753307953273e-05, + "loss": 0.5065, + "step": 14839 + }, + { + "epoch": 0.4680649739788677, + "grad_norm": 1.695430874824524, + "learning_rate": 1.1515743562589048e-05, + "loss": 0.4999, + "step": 14840 + }, + { + "epoch": 0.46809651474530833, + "grad_norm": 1.7839946746826172, + "learning_rate": 1.1514733801406866e-05, + "loss": 0.4859, + "step": 14841 + }, + { + "epoch": 0.4681280555117489, + "grad_norm": 1.5838570594787598, + "learning_rate": 1.1513724024417256e-05, + "loss": 0.7902, + "step": 14842 + }, + { + "epoch": 0.46815959627818954, + "grad_norm": 1.7014483213424683, + "learning_rate": 1.1512714231630762e-05, + "loss": 0.4529, + "step": 14843 + }, + { + "epoch": 0.46819113704463017, + "grad_norm": 1.4309959411621094, + "learning_rate": 1.1511704423057923e-05, + "loss": 0.7878, + "step": 14844 + }, + { + "epoch": 0.4682226778110708, + "grad_norm": 1.6242879629135132, + "learning_rate": 1.1510694598709274e-05, + "loss": 0.5153, + "step": 14845 + }, + { + "epoch": 0.4682542185775114, + "grad_norm": 1.5226956605911255, + "learning_rate": 1.1509684758595349e-05, + "loss": 0.4738, + "step": 14846 + }, + { + "epoch": 0.46828575934395206, + "grad_norm": 1.5260632038116455, + "learning_rate": 1.1508674902726696e-05, + "loss": 0.4793, + "step": 14847 + }, + { + "epoch": 0.4683173001103927, + "grad_norm": 2.737518310546875, + "learning_rate": 1.1507665031113849e-05, + "loss": 0.5087, + "step": 14848 + }, + { + "epoch": 0.4683488408768333, + "grad_norm": 1.7029250860214233, + "learning_rate": 1.1506655143767347e-05, + "loss": 0.4883, + "step": 14849 + }, + { + "epoch": 0.46838038164327395, + "grad_norm": 1.5702110528945923, + "learning_rate": 1.1505645240697728e-05, + "loss": 0.4986, + "step": 14850 + }, + { + "epoch": 0.4684119224097146, + "grad_norm": 1.400248408317566, + "learning_rate": 1.150463532191553e-05, + "loss": 0.4863, + "step": 14851 + }, + { + "epoch": 0.4684434631761552, + "grad_norm": 1.232344150543213, + "learning_rate": 1.1503625387431294e-05, + "loss": 0.4921, + "step": 14852 + }, + { + "epoch": 0.4684750039425958, + "grad_norm": 1.4440854787826538, + "learning_rate": 1.150261543725556e-05, + "loss": 0.4432, + "step": 14853 + }, + { + "epoch": 0.4685065447090364, + "grad_norm": 1.3546597957611084, + "learning_rate": 1.1501605471398867e-05, + "loss": 0.488, + "step": 14854 + }, + { + "epoch": 0.46853808547547704, + "grad_norm": 1.6827596426010132, + "learning_rate": 1.1500595489871758e-05, + "loss": 0.4916, + "step": 14855 + }, + { + "epoch": 0.46856962624191767, + "grad_norm": 1.5804316997528076, + "learning_rate": 1.1499585492684765e-05, + "loss": 0.4861, + "step": 14856 + }, + { + "epoch": 0.4686011670083583, + "grad_norm": 1.6723759174346924, + "learning_rate": 1.1498575479848437e-05, + "loss": 0.4725, + "step": 14857 + }, + { + "epoch": 0.46863270777479893, + "grad_norm": 2.04681658744812, + "learning_rate": 1.149756545137331e-05, + "loss": 0.4744, + "step": 14858 + }, + { + "epoch": 0.46866424854123956, + "grad_norm": 2.1535091400146484, + "learning_rate": 1.1496555407269926e-05, + "loss": 0.798, + "step": 14859 + }, + { + "epoch": 0.4686957893076802, + "grad_norm": 1.5397365093231201, + "learning_rate": 1.149554534754882e-05, + "loss": 0.4515, + "step": 14860 + }, + { + "epoch": 0.4687273300741208, + "grad_norm": 2.103086233139038, + "learning_rate": 1.149453527222054e-05, + "loss": 0.4839, + "step": 14861 + }, + { + "epoch": 0.46875887084056145, + "grad_norm": 11.42934513092041, + "learning_rate": 1.1493525181295621e-05, + "loss": 0.4759, + "step": 14862 + }, + { + "epoch": 0.4687904116070021, + "grad_norm": 1.7639271020889282, + "learning_rate": 1.149251507478461e-05, + "loss": 0.7459, + "step": 14863 + }, + { + "epoch": 0.46882195237344265, + "grad_norm": 2.2254719734191895, + "learning_rate": 1.1491504952698043e-05, + "loss": 0.4538, + "step": 14864 + }, + { + "epoch": 0.4688534931398833, + "grad_norm": 1.711171269416809, + "learning_rate": 1.1490494815046467e-05, + "loss": 0.5172, + "step": 14865 + }, + { + "epoch": 0.4688850339063239, + "grad_norm": 2.346569776535034, + "learning_rate": 1.1489484661840415e-05, + "loss": 0.5059, + "step": 14866 + }, + { + "epoch": 0.46891657467276454, + "grad_norm": 1.394456386566162, + "learning_rate": 1.1488474493090438e-05, + "loss": 0.4901, + "step": 14867 + }, + { + "epoch": 0.46894811543920517, + "grad_norm": 1.7312500476837158, + "learning_rate": 1.1487464308807075e-05, + "loss": 0.3872, + "step": 14868 + }, + { + "epoch": 0.4689796562056458, + "grad_norm": 1.9184601306915283, + "learning_rate": 1.1486454109000867e-05, + "loss": 0.4996, + "step": 14869 + }, + { + "epoch": 0.46901119697208643, + "grad_norm": 3.2340316772460938, + "learning_rate": 1.1485443893682353e-05, + "loss": 0.504, + "step": 14870 + }, + { + "epoch": 0.46904273773852706, + "grad_norm": 1.320003628730774, + "learning_rate": 1.1484433662862077e-05, + "loss": 0.5011, + "step": 14871 + }, + { + "epoch": 0.4690742785049677, + "grad_norm": 1.8027880191802979, + "learning_rate": 1.1483423416550587e-05, + "loss": 0.4844, + "step": 14872 + }, + { + "epoch": 0.4691058192714083, + "grad_norm": 1.2930738925933838, + "learning_rate": 1.1482413154758421e-05, + "loss": 0.5032, + "step": 14873 + }, + { + "epoch": 0.4691373600378489, + "grad_norm": 2.154942274093628, + "learning_rate": 1.1481402877496122e-05, + "loss": 0.529, + "step": 14874 + }, + { + "epoch": 0.4691689008042895, + "grad_norm": 1.3004416227340698, + "learning_rate": 1.1480392584774234e-05, + "loss": 0.4932, + "step": 14875 + }, + { + "epoch": 0.46920044157073015, + "grad_norm": 1.4563517570495605, + "learning_rate": 1.14793822766033e-05, + "loss": 0.4962, + "step": 14876 + }, + { + "epoch": 0.4692319823371708, + "grad_norm": 1.3927603960037231, + "learning_rate": 1.1478371952993863e-05, + "loss": 0.3678, + "step": 14877 + }, + { + "epoch": 0.4692635231036114, + "grad_norm": 1.7724653482437134, + "learning_rate": 1.1477361613956468e-05, + "loss": 0.479, + "step": 14878 + }, + { + "epoch": 0.46929506387005204, + "grad_norm": 2.0721042156219482, + "learning_rate": 1.1476351259501657e-05, + "loss": 0.4852, + "step": 14879 + }, + { + "epoch": 0.4693266046364927, + "grad_norm": 1.8445075750350952, + "learning_rate": 1.1475340889639973e-05, + "loss": 0.4898, + "step": 14880 + }, + { + "epoch": 0.4693581454029333, + "grad_norm": 1.6945058107376099, + "learning_rate": 1.147433050438196e-05, + "loss": 0.4793, + "step": 14881 + }, + { + "epoch": 0.46938968616937393, + "grad_norm": 1.5499367713928223, + "learning_rate": 1.1473320103738169e-05, + "loss": 0.4825, + "step": 14882 + }, + { + "epoch": 0.46942122693581456, + "grad_norm": 1.426395297050476, + "learning_rate": 1.1472309687719131e-05, + "loss": 0.4622, + "step": 14883 + }, + { + "epoch": 0.4694527677022552, + "grad_norm": 1.381942868232727, + "learning_rate": 1.1471299256335403e-05, + "loss": 0.4843, + "step": 14884 + }, + { + "epoch": 0.46948430846869577, + "grad_norm": 1.316673755645752, + "learning_rate": 1.1470288809597523e-05, + "loss": 0.4935, + "step": 14885 + }, + { + "epoch": 0.4695158492351364, + "grad_norm": 1.370524525642395, + "learning_rate": 1.146927834751604e-05, + "loss": 0.4777, + "step": 14886 + }, + { + "epoch": 0.469547390001577, + "grad_norm": 1.4090033769607544, + "learning_rate": 1.1468267870101495e-05, + "loss": 0.4898, + "step": 14887 + }, + { + "epoch": 0.46957893076801765, + "grad_norm": 1.42556893825531, + "learning_rate": 1.1467257377364435e-05, + "loss": 0.492, + "step": 14888 + }, + { + "epoch": 0.4696104715344583, + "grad_norm": 1.606467604637146, + "learning_rate": 1.1466246869315407e-05, + "loss": 0.8306, + "step": 14889 + }, + { + "epoch": 0.4696420123008989, + "grad_norm": 1.605891227722168, + "learning_rate": 1.1465236345964953e-05, + "loss": 0.4912, + "step": 14890 + }, + { + "epoch": 0.46967355306733954, + "grad_norm": 1.5213452577590942, + "learning_rate": 1.1464225807323617e-05, + "loss": 0.4714, + "step": 14891 + }, + { + "epoch": 0.4697050938337802, + "grad_norm": 1.6641809940338135, + "learning_rate": 1.146321525340195e-05, + "loss": 0.4641, + "step": 14892 + }, + { + "epoch": 0.4697366346002208, + "grad_norm": 1.860913634300232, + "learning_rate": 1.1462204684210497e-05, + "loss": 0.4986, + "step": 14893 + }, + { + "epoch": 0.46976817536666143, + "grad_norm": 1.813962697982788, + "learning_rate": 1.1461194099759798e-05, + "loss": 0.5042, + "step": 14894 + }, + { + "epoch": 0.469799716133102, + "grad_norm": 1.6062015295028687, + "learning_rate": 1.1460183500060406e-05, + "loss": 0.7783, + "step": 14895 + }, + { + "epoch": 0.46983125689954264, + "grad_norm": 1.4884146451950073, + "learning_rate": 1.1459172885122869e-05, + "loss": 0.414, + "step": 14896 + }, + { + "epoch": 0.46986279766598327, + "grad_norm": 1.5878925323486328, + "learning_rate": 1.1458162254957728e-05, + "loss": 0.4981, + "step": 14897 + }, + { + "epoch": 0.4698943384324239, + "grad_norm": 1.6568284034729004, + "learning_rate": 1.1457151609575531e-05, + "loss": 0.5209, + "step": 14898 + }, + { + "epoch": 0.4699258791988645, + "grad_norm": 1.8351778984069824, + "learning_rate": 1.145614094898683e-05, + "loss": 0.4434, + "step": 14899 + }, + { + "epoch": 0.46995741996530516, + "grad_norm": 1.4912745952606201, + "learning_rate": 1.145513027320216e-05, + "loss": 0.4762, + "step": 14900 + }, + { + "epoch": 0.4699889607317458, + "grad_norm": 1.5192210674285889, + "learning_rate": 1.1454119582232079e-05, + "loss": 0.5065, + "step": 14901 + }, + { + "epoch": 0.4700205014981864, + "grad_norm": 1.2363075017929077, + "learning_rate": 1.1453108876087132e-05, + "loss": 0.4605, + "step": 14902 + }, + { + "epoch": 0.47005204226462705, + "grad_norm": 1.5345087051391602, + "learning_rate": 1.1452098154777864e-05, + "loss": 0.5069, + "step": 14903 + }, + { + "epoch": 0.4700835830310677, + "grad_norm": 1.987690806388855, + "learning_rate": 1.1451087418314823e-05, + "loss": 0.4824, + "step": 14904 + }, + { + "epoch": 0.4701151237975083, + "grad_norm": 1.5021713972091675, + "learning_rate": 1.145007666670856e-05, + "loss": 0.7793, + "step": 14905 + }, + { + "epoch": 0.4701466645639489, + "grad_norm": 2.027894973754883, + "learning_rate": 1.1449065899969623e-05, + "loss": 0.4383, + "step": 14906 + }, + { + "epoch": 0.4701782053303895, + "grad_norm": 1.9434432983398438, + "learning_rate": 1.1448055118108555e-05, + "loss": 0.4895, + "step": 14907 + }, + { + "epoch": 0.47020974609683014, + "grad_norm": 1.3564621210098267, + "learning_rate": 1.1447044321135909e-05, + "loss": 0.4811, + "step": 14908 + }, + { + "epoch": 0.47024128686327077, + "grad_norm": 2.298340320587158, + "learning_rate": 1.1446033509062231e-05, + "loss": 0.5064, + "step": 14909 + }, + { + "epoch": 0.4702728276297114, + "grad_norm": 1.7138646841049194, + "learning_rate": 1.1445022681898072e-05, + "loss": 0.493, + "step": 14910 + }, + { + "epoch": 0.47030436839615203, + "grad_norm": 1.389212727546692, + "learning_rate": 1.1444011839653978e-05, + "loss": 0.5154, + "step": 14911 + }, + { + "epoch": 0.47033590916259266, + "grad_norm": 1.572665810585022, + "learning_rate": 1.14430009823405e-05, + "loss": 0.4905, + "step": 14912 + }, + { + "epoch": 0.4703674499290333, + "grad_norm": 1.5193103551864624, + "learning_rate": 1.1441990109968186e-05, + "loss": 0.4998, + "step": 14913 + }, + { + "epoch": 0.4703989906954739, + "grad_norm": 1.8161742687225342, + "learning_rate": 1.1440979222547585e-05, + "loss": 0.495, + "step": 14914 + }, + { + "epoch": 0.47043053146191455, + "grad_norm": 1.397625207901001, + "learning_rate": 1.1439968320089246e-05, + "loss": 0.5024, + "step": 14915 + }, + { + "epoch": 0.4704620722283551, + "grad_norm": 1.5683144330978394, + "learning_rate": 1.1438957402603724e-05, + "loss": 0.5143, + "step": 14916 + }, + { + "epoch": 0.47049361299479575, + "grad_norm": 1.2429113388061523, + "learning_rate": 1.1437946470101558e-05, + "loss": 0.481, + "step": 14917 + }, + { + "epoch": 0.4705251537612364, + "grad_norm": 1.9086190462112427, + "learning_rate": 1.1436935522593309e-05, + "loss": 0.4873, + "step": 14918 + }, + { + "epoch": 0.470556694527677, + "grad_norm": 1.5824743509292603, + "learning_rate": 1.1435924560089519e-05, + "loss": 0.4683, + "step": 14919 + }, + { + "epoch": 0.47058823529411764, + "grad_norm": 1.7541828155517578, + "learning_rate": 1.1434913582600744e-05, + "loss": 0.4917, + "step": 14920 + }, + { + "epoch": 0.47061977606055827, + "grad_norm": 1.56430983543396, + "learning_rate": 1.1433902590137529e-05, + "loss": 0.4657, + "step": 14921 + }, + { + "epoch": 0.4706513168269989, + "grad_norm": 1.8663164377212524, + "learning_rate": 1.1432891582710426e-05, + "loss": 0.4516, + "step": 14922 + }, + { + "epoch": 0.47068285759343953, + "grad_norm": 1.8348671197891235, + "learning_rate": 1.1431880560329991e-05, + "loss": 0.7722, + "step": 14923 + }, + { + "epoch": 0.47071439835988016, + "grad_norm": 1.6129169464111328, + "learning_rate": 1.1430869523006768e-05, + "loss": 0.459, + "step": 14924 + }, + { + "epoch": 0.4707459391263208, + "grad_norm": 1.6752959489822388, + "learning_rate": 1.1429858470751309e-05, + "loss": 0.498, + "step": 14925 + }, + { + "epoch": 0.4707774798927614, + "grad_norm": 1.4303354024887085, + "learning_rate": 1.142884740357417e-05, + "loss": 0.5087, + "step": 14926 + }, + { + "epoch": 0.470809020659202, + "grad_norm": 1.420661449432373, + "learning_rate": 1.1427836321485896e-05, + "loss": 0.7997, + "step": 14927 + }, + { + "epoch": 0.4708405614256426, + "grad_norm": 1.387560486793518, + "learning_rate": 1.1426825224497045e-05, + "loss": 0.5283, + "step": 14928 + }, + { + "epoch": 0.47087210219208325, + "grad_norm": 1.3262144327163696, + "learning_rate": 1.1425814112618162e-05, + "loss": 0.4672, + "step": 14929 + }, + { + "epoch": 0.4709036429585239, + "grad_norm": 1.5728100538253784, + "learning_rate": 1.1424802985859802e-05, + "loss": 0.487, + "step": 14930 + }, + { + "epoch": 0.4709351837249645, + "grad_norm": 2.1293938159942627, + "learning_rate": 1.1423791844232515e-05, + "loss": 0.3637, + "step": 14931 + }, + { + "epoch": 0.47096672449140514, + "grad_norm": 1.3811229467391968, + "learning_rate": 1.1422780687746855e-05, + "loss": 0.4902, + "step": 14932 + }, + { + "epoch": 0.47099826525784577, + "grad_norm": 1.616578459739685, + "learning_rate": 1.1421769516413375e-05, + "loss": 0.5124, + "step": 14933 + }, + { + "epoch": 0.4710298060242864, + "grad_norm": 1.810011863708496, + "learning_rate": 1.1420758330242628e-05, + "loss": 0.5123, + "step": 14934 + }, + { + "epoch": 0.47106134679072703, + "grad_norm": 1.595105767250061, + "learning_rate": 1.141974712924516e-05, + "loss": 0.5019, + "step": 14935 + }, + { + "epoch": 0.47109288755716766, + "grad_norm": 1.255529522895813, + "learning_rate": 1.1418735913431533e-05, + "loss": 0.4881, + "step": 14936 + }, + { + "epoch": 0.4711244283236083, + "grad_norm": 1.4641199111938477, + "learning_rate": 1.1417724682812293e-05, + "loss": 0.501, + "step": 14937 + }, + { + "epoch": 0.47115596909004887, + "grad_norm": 1.312934160232544, + "learning_rate": 1.1416713437397994e-05, + "loss": 0.4689, + "step": 14938 + }, + { + "epoch": 0.4711875098564895, + "grad_norm": 1.7134549617767334, + "learning_rate": 1.1415702177199193e-05, + "loss": 0.785, + "step": 14939 + }, + { + "epoch": 0.4712190506229301, + "grad_norm": 1.6092379093170166, + "learning_rate": 1.141469090222644e-05, + "loss": 0.4544, + "step": 14940 + }, + { + "epoch": 0.47125059138937075, + "grad_norm": 1.351637363433838, + "learning_rate": 1.1413679612490285e-05, + "loss": 0.4881, + "step": 14941 + }, + { + "epoch": 0.4712821321558114, + "grad_norm": 1.415881633758545, + "learning_rate": 1.1412668308001289e-05, + "loss": 0.4957, + "step": 14942 + }, + { + "epoch": 0.471313672922252, + "grad_norm": 1.3237919807434082, + "learning_rate": 1.1411656988769998e-05, + "loss": 0.4677, + "step": 14943 + }, + { + "epoch": 0.47134521368869264, + "grad_norm": 1.5374351739883423, + "learning_rate": 1.1410645654806977e-05, + "loss": 0.4995, + "step": 14944 + }, + { + "epoch": 0.4713767544551333, + "grad_norm": 1.9290955066680908, + "learning_rate": 1.140963430612277e-05, + "loss": 0.4396, + "step": 14945 + }, + { + "epoch": 0.4714082952215739, + "grad_norm": 2.074610710144043, + "learning_rate": 1.1408622942727936e-05, + "loss": 0.4399, + "step": 14946 + }, + { + "epoch": 0.47143983598801453, + "grad_norm": 1.6654155254364014, + "learning_rate": 1.1407611564633026e-05, + "loss": 0.4746, + "step": 14947 + }, + { + "epoch": 0.4714713767544551, + "grad_norm": 1.9562753438949585, + "learning_rate": 1.1406600171848596e-05, + "loss": 0.4371, + "step": 14948 + }, + { + "epoch": 0.47150291752089574, + "grad_norm": 1.6363458633422852, + "learning_rate": 1.14055887643852e-05, + "loss": 0.514, + "step": 14949 + }, + { + "epoch": 0.47153445828733637, + "grad_norm": 1.6821264028549194, + "learning_rate": 1.1404577342253396e-05, + "loss": 0.5184, + "step": 14950 + }, + { + "epoch": 0.471565999053777, + "grad_norm": 1.3827641010284424, + "learning_rate": 1.1403565905463738e-05, + "loss": 0.4534, + "step": 14951 + }, + { + "epoch": 0.4715975398202176, + "grad_norm": 1.7015661001205444, + "learning_rate": 1.1402554454026777e-05, + "loss": 0.4859, + "step": 14952 + }, + { + "epoch": 0.47162908058665826, + "grad_norm": 1.4943592548370361, + "learning_rate": 1.140154298795307e-05, + "loss": 0.4752, + "step": 14953 + }, + { + "epoch": 0.4716606213530989, + "grad_norm": 1.608942985534668, + "learning_rate": 1.140053150725318e-05, + "loss": 0.4566, + "step": 14954 + }, + { + "epoch": 0.4716921621195395, + "grad_norm": 1.3585593700408936, + "learning_rate": 1.139952001193765e-05, + "loss": 0.4833, + "step": 14955 + }, + { + "epoch": 0.47172370288598015, + "grad_norm": 1.3064024448394775, + "learning_rate": 1.1398508502017047e-05, + "loss": 0.4958, + "step": 14956 + }, + { + "epoch": 0.4717552436524208, + "grad_norm": 1.371594786643982, + "learning_rate": 1.1397496977501918e-05, + "loss": 0.4695, + "step": 14957 + }, + { + "epoch": 0.4717867844188614, + "grad_norm": 1.3732736110687256, + "learning_rate": 1.1396485438402826e-05, + "loss": 0.3457, + "step": 14958 + }, + { + "epoch": 0.471818325185302, + "grad_norm": 1.292911171913147, + "learning_rate": 1.1395473884730322e-05, + "loss": 0.4892, + "step": 14959 + }, + { + "epoch": 0.4718498659517426, + "grad_norm": 1.2817778587341309, + "learning_rate": 1.1394462316494964e-05, + "loss": 0.4698, + "step": 14960 + }, + { + "epoch": 0.47188140671818324, + "grad_norm": 1.245653510093689, + "learning_rate": 1.139345073370731e-05, + "loss": 0.4826, + "step": 14961 + }, + { + "epoch": 0.47191294748462387, + "grad_norm": 1.6590428352355957, + "learning_rate": 1.1392439136377915e-05, + "loss": 0.7902, + "step": 14962 + }, + { + "epoch": 0.4719444882510645, + "grad_norm": 1.7165244817733765, + "learning_rate": 1.1391427524517336e-05, + "loss": 0.4809, + "step": 14963 + }, + { + "epoch": 0.4719760290175051, + "grad_norm": 1.762974739074707, + "learning_rate": 1.1390415898136132e-05, + "loss": 0.7897, + "step": 14964 + }, + { + "epoch": 0.47200756978394576, + "grad_norm": 1.8312851190567017, + "learning_rate": 1.1389404257244855e-05, + "loss": 0.7979, + "step": 14965 + }, + { + "epoch": 0.4720391105503864, + "grad_norm": 1.460890769958496, + "learning_rate": 1.138839260185407e-05, + "loss": 0.4986, + "step": 14966 + }, + { + "epoch": 0.472070651316827, + "grad_norm": 1.3653591871261597, + "learning_rate": 1.1387380931974328e-05, + "loss": 0.4754, + "step": 14967 + }, + { + "epoch": 0.47210219208326765, + "grad_norm": 1.7300031185150146, + "learning_rate": 1.1386369247616191e-05, + "loss": 0.4664, + "step": 14968 + }, + { + "epoch": 0.4721337328497082, + "grad_norm": 1.6701728105545044, + "learning_rate": 1.1385357548790212e-05, + "loss": 0.8141, + "step": 14969 + }, + { + "epoch": 0.47216527361614885, + "grad_norm": 2.504929304122925, + "learning_rate": 1.1384345835506952e-05, + "loss": 0.8328, + "step": 14970 + }, + { + "epoch": 0.4721968143825895, + "grad_norm": 1.4526252746582031, + "learning_rate": 1.1383334107776968e-05, + "loss": 0.4672, + "step": 14971 + }, + { + "epoch": 0.4722283551490301, + "grad_norm": 1.4828282594680786, + "learning_rate": 1.138232236561082e-05, + "loss": 0.8014, + "step": 14972 + }, + { + "epoch": 0.47225989591547074, + "grad_norm": 1.611430287361145, + "learning_rate": 1.138131060901906e-05, + "loss": 0.4743, + "step": 14973 + }, + { + "epoch": 0.47229143668191137, + "grad_norm": 2.4073832035064697, + "learning_rate": 1.1380298838012254e-05, + "loss": 0.3834, + "step": 14974 + }, + { + "epoch": 0.472322977448352, + "grad_norm": 1.6717031002044678, + "learning_rate": 1.137928705260096e-05, + "loss": 0.4727, + "step": 14975 + }, + { + "epoch": 0.47235451821479263, + "grad_norm": 1.4467222690582275, + "learning_rate": 1.1378275252795734e-05, + "loss": 0.8069, + "step": 14976 + }, + { + "epoch": 0.47238605898123326, + "grad_norm": 1.4136828184127808, + "learning_rate": 1.1377263438607134e-05, + "loss": 0.4852, + "step": 14977 + }, + { + "epoch": 0.4724175997476739, + "grad_norm": 1.97783362865448, + "learning_rate": 1.1376251610045723e-05, + "loss": 0.79, + "step": 14978 + }, + { + "epoch": 0.4724491405141145, + "grad_norm": 1.404542326927185, + "learning_rate": 1.1375239767122056e-05, + "loss": 0.4698, + "step": 14979 + }, + { + "epoch": 0.4724806812805551, + "grad_norm": 1.607545018196106, + "learning_rate": 1.137422790984669e-05, + "loss": 0.4976, + "step": 14980 + }, + { + "epoch": 0.4725122220469957, + "grad_norm": 1.6119115352630615, + "learning_rate": 1.1373216038230192e-05, + "loss": 0.4488, + "step": 14981 + }, + { + "epoch": 0.47254376281343635, + "grad_norm": 1.4170045852661133, + "learning_rate": 1.1372204152283122e-05, + "loss": 0.4734, + "step": 14982 + }, + { + "epoch": 0.472575303579877, + "grad_norm": 1.8026667833328247, + "learning_rate": 1.1371192252016033e-05, + "loss": 0.4752, + "step": 14983 + }, + { + "epoch": 0.4726068443463176, + "grad_norm": 1.484676718711853, + "learning_rate": 1.1370180337439487e-05, + "loss": 0.4861, + "step": 14984 + }, + { + "epoch": 0.47263838511275824, + "grad_norm": 1.4455691576004028, + "learning_rate": 1.1369168408564049e-05, + "loss": 0.4773, + "step": 14985 + }, + { + "epoch": 0.47266992587919887, + "grad_norm": 1.6092994213104248, + "learning_rate": 1.1368156465400275e-05, + "loss": 0.4882, + "step": 14986 + }, + { + "epoch": 0.4727014666456395, + "grad_norm": 1.5913536548614502, + "learning_rate": 1.1367144507958723e-05, + "loss": 0.4757, + "step": 14987 + }, + { + "epoch": 0.47273300741208013, + "grad_norm": 1.4575703144073486, + "learning_rate": 1.136613253624996e-05, + "loss": 0.4779, + "step": 14988 + }, + { + "epoch": 0.47276454817852076, + "grad_norm": 1.2960689067840576, + "learning_rate": 1.1365120550284538e-05, + "loss": 0.5168, + "step": 14989 + }, + { + "epoch": 0.4727960889449614, + "grad_norm": 1.4850411415100098, + "learning_rate": 1.1364108550073025e-05, + "loss": 0.4869, + "step": 14990 + }, + { + "epoch": 0.47282762971140196, + "grad_norm": 1.695972204208374, + "learning_rate": 1.1363096535625982e-05, + "loss": 0.4996, + "step": 14991 + }, + { + "epoch": 0.4728591704778426, + "grad_norm": 1.3460499048233032, + "learning_rate": 1.1362084506953968e-05, + "loss": 0.4278, + "step": 14992 + }, + { + "epoch": 0.4728907112442832, + "grad_norm": 1.3249952793121338, + "learning_rate": 1.1361072464067541e-05, + "loss": 0.4898, + "step": 14993 + }, + { + "epoch": 0.47292225201072385, + "grad_norm": 1.786439061164856, + "learning_rate": 1.1360060406977271e-05, + "loss": 0.492, + "step": 14994 + }, + { + "epoch": 0.4729537927771645, + "grad_norm": 1.3309904336929321, + "learning_rate": 1.1359048335693711e-05, + "loss": 0.4979, + "step": 14995 + }, + { + "epoch": 0.4729853335436051, + "grad_norm": 2.5428876876831055, + "learning_rate": 1.1358036250227431e-05, + "loss": 0.8098, + "step": 14996 + }, + { + "epoch": 0.47301687431004574, + "grad_norm": 1.652644395828247, + "learning_rate": 1.1357024150588984e-05, + "loss": 0.4495, + "step": 14997 + }, + { + "epoch": 0.4730484150764864, + "grad_norm": 1.8250267505645752, + "learning_rate": 1.1356012036788937e-05, + "loss": 0.4866, + "step": 14998 + }, + { + "epoch": 0.473079955842927, + "grad_norm": 1.708608865737915, + "learning_rate": 1.1354999908837854e-05, + "loss": 0.7933, + "step": 14999 + }, + { + "epoch": 0.47311149660936763, + "grad_norm": 1.3236607313156128, + "learning_rate": 1.135398776674629e-05, + "loss": 0.4657, + "step": 15000 + }, + { + "epoch": 0.4731430373758082, + "grad_norm": 1.326178789138794, + "learning_rate": 1.1352975610524814e-05, + "loss": 0.4863, + "step": 15001 + }, + { + "epoch": 0.47317457814224884, + "grad_norm": 1.509498953819275, + "learning_rate": 1.1351963440183991e-05, + "loss": 0.7637, + "step": 15002 + }, + { + "epoch": 0.47320611890868947, + "grad_norm": 1.3953189849853516, + "learning_rate": 1.1350951255734374e-05, + "loss": 0.8195, + "step": 15003 + }, + { + "epoch": 0.4732376596751301, + "grad_norm": 2.112004518508911, + "learning_rate": 1.1349939057186535e-05, + "loss": 0.793, + "step": 15004 + }, + { + "epoch": 0.4732692004415707, + "grad_norm": 1.9609158039093018, + "learning_rate": 1.1348926844551034e-05, + "loss": 0.3911, + "step": 15005 + }, + { + "epoch": 0.47330074120801136, + "grad_norm": 2.349769353866577, + "learning_rate": 1.1347914617838434e-05, + "loss": 0.5318, + "step": 15006 + }, + { + "epoch": 0.473332281974452, + "grad_norm": 1.4766830205917358, + "learning_rate": 1.1346902377059297e-05, + "loss": 0.5329, + "step": 15007 + }, + { + "epoch": 0.4733638227408926, + "grad_norm": 1.426077127456665, + "learning_rate": 1.1345890122224188e-05, + "loss": 0.4995, + "step": 15008 + }, + { + "epoch": 0.47339536350733324, + "grad_norm": 1.4045324325561523, + "learning_rate": 1.134487785334367e-05, + "loss": 0.486, + "step": 15009 + }, + { + "epoch": 0.4734269042737739, + "grad_norm": 2.0371787548065186, + "learning_rate": 1.1343865570428307e-05, + "loss": 0.8169, + "step": 15010 + }, + { + "epoch": 0.4734584450402145, + "grad_norm": 1.9681843519210815, + "learning_rate": 1.1342853273488666e-05, + "loss": 0.4685, + "step": 15011 + }, + { + "epoch": 0.4734899858066551, + "grad_norm": 1.4952207803726196, + "learning_rate": 1.1341840962535307e-05, + "loss": 0.4727, + "step": 15012 + }, + { + "epoch": 0.4735215265730957, + "grad_norm": 1.3872497081756592, + "learning_rate": 1.1340828637578794e-05, + "loss": 0.4427, + "step": 15013 + }, + { + "epoch": 0.47355306733953634, + "grad_norm": 1.2556790113449097, + "learning_rate": 1.1339816298629693e-05, + "loss": 0.4825, + "step": 15014 + }, + { + "epoch": 0.47358460810597697, + "grad_norm": 1.5106635093688965, + "learning_rate": 1.1338803945698573e-05, + "loss": 0.39, + "step": 15015 + }, + { + "epoch": 0.4736161488724176, + "grad_norm": 1.6726248264312744, + "learning_rate": 1.133779157879599e-05, + "loss": 0.4689, + "step": 15016 + }, + { + "epoch": 0.4736476896388582, + "grad_norm": 1.7425098419189453, + "learning_rate": 1.1336779197932515e-05, + "loss": 0.4487, + "step": 15017 + }, + { + "epoch": 0.47367923040529886, + "grad_norm": 1.382432222366333, + "learning_rate": 1.133576680311871e-05, + "loss": 0.4845, + "step": 15018 + }, + { + "epoch": 0.4737107711717395, + "grad_norm": 1.8373358249664307, + "learning_rate": 1.1334754394365141e-05, + "loss": 0.3871, + "step": 15019 + }, + { + "epoch": 0.4737423119381801, + "grad_norm": 1.294844627380371, + "learning_rate": 1.1333741971682375e-05, + "loss": 0.3605, + "step": 15020 + }, + { + "epoch": 0.47377385270462075, + "grad_norm": 1.9959121942520142, + "learning_rate": 1.1332729535080972e-05, + "loss": 0.3541, + "step": 15021 + }, + { + "epoch": 0.4738053934710613, + "grad_norm": 1.835042119026184, + "learning_rate": 1.1331717084571505e-05, + "loss": 0.4831, + "step": 15022 + }, + { + "epoch": 0.47383693423750195, + "grad_norm": 10.872127532958984, + "learning_rate": 1.133070462016454e-05, + "loss": 0.5111, + "step": 15023 + }, + { + "epoch": 0.4738684750039426, + "grad_norm": 1.3619931936264038, + "learning_rate": 1.1329692141870632e-05, + "loss": 0.4713, + "step": 15024 + }, + { + "epoch": 0.4739000157703832, + "grad_norm": 1.446937084197998, + "learning_rate": 1.1328679649700359e-05, + "loss": 0.5132, + "step": 15025 + }, + { + "epoch": 0.47393155653682384, + "grad_norm": 1.490790605545044, + "learning_rate": 1.132766714366428e-05, + "loss": 0.4614, + "step": 15026 + }, + { + "epoch": 0.47396309730326447, + "grad_norm": 1.384803056716919, + "learning_rate": 1.1326654623772964e-05, + "loss": 0.4893, + "step": 15027 + }, + { + "epoch": 0.4739946380697051, + "grad_norm": 1.326237440109253, + "learning_rate": 1.1325642090036975e-05, + "loss": 0.4622, + "step": 15028 + }, + { + "epoch": 0.47402617883614573, + "grad_norm": 1.6637980937957764, + "learning_rate": 1.132462954246688e-05, + "loss": 0.5091, + "step": 15029 + }, + { + "epoch": 0.47405771960258636, + "grad_norm": 1.5174715518951416, + "learning_rate": 1.1323616981073251e-05, + "loss": 0.472, + "step": 15030 + }, + { + "epoch": 0.474089260369027, + "grad_norm": 1.7801016569137573, + "learning_rate": 1.132260440586665e-05, + "loss": 0.5136, + "step": 15031 + }, + { + "epoch": 0.4741208011354676, + "grad_norm": 1.945824384689331, + "learning_rate": 1.1321591816857645e-05, + "loss": 0.5087, + "step": 15032 + }, + { + "epoch": 0.4741523419019082, + "grad_norm": 1.4458600282669067, + "learning_rate": 1.1320579214056804e-05, + "loss": 0.5216, + "step": 15033 + }, + { + "epoch": 0.4741838826683488, + "grad_norm": 1.2654333114624023, + "learning_rate": 1.1319566597474692e-05, + "loss": 0.3706, + "step": 15034 + }, + { + "epoch": 0.47421542343478945, + "grad_norm": 1.5439739227294922, + "learning_rate": 1.1318553967121879e-05, + "loss": 0.4805, + "step": 15035 + }, + { + "epoch": 0.4742469642012301, + "grad_norm": 1.6318931579589844, + "learning_rate": 1.1317541323008933e-05, + "loss": 0.4592, + "step": 15036 + }, + { + "epoch": 0.4742785049676707, + "grad_norm": 1.5887823104858398, + "learning_rate": 1.1316528665146417e-05, + "loss": 0.4581, + "step": 15037 + }, + { + "epoch": 0.47431004573411134, + "grad_norm": 1.6620885133743286, + "learning_rate": 1.1315515993544903e-05, + "loss": 0.481, + "step": 15038 + }, + { + "epoch": 0.47434158650055197, + "grad_norm": 1.884853482246399, + "learning_rate": 1.1314503308214957e-05, + "loss": 0.443, + "step": 15039 + }, + { + "epoch": 0.4743731272669926, + "grad_norm": 1.743889570236206, + "learning_rate": 1.1313490609167153e-05, + "loss": 0.4844, + "step": 15040 + }, + { + "epoch": 0.47440466803343323, + "grad_norm": 2.1662275791168213, + "learning_rate": 1.1312477896412048e-05, + "loss": 0.4793, + "step": 15041 + }, + { + "epoch": 0.47443620879987386, + "grad_norm": 3.0408294200897217, + "learning_rate": 1.1311465169960221e-05, + "loss": 0.474, + "step": 15042 + }, + { + "epoch": 0.47446774956631443, + "grad_norm": 1.5195320844650269, + "learning_rate": 1.1310452429822239e-05, + "loss": 0.4783, + "step": 15043 + }, + { + "epoch": 0.47449929033275506, + "grad_norm": 1.9038668870925903, + "learning_rate": 1.1309439676008663e-05, + "loss": 0.7755, + "step": 15044 + }, + { + "epoch": 0.4745308310991957, + "grad_norm": 2.323214292526245, + "learning_rate": 1.1308426908530073e-05, + "loss": 0.3491, + "step": 15045 + }, + { + "epoch": 0.4745623718656363, + "grad_norm": 3.158541679382324, + "learning_rate": 1.1307414127397028e-05, + "loss": 0.819, + "step": 15046 + }, + { + "epoch": 0.47459391263207695, + "grad_norm": 1.9604337215423584, + "learning_rate": 1.1306401332620103e-05, + "loss": 0.4389, + "step": 15047 + }, + { + "epoch": 0.4746254533985176, + "grad_norm": 1.6690502166748047, + "learning_rate": 1.1305388524209864e-05, + "loss": 0.4698, + "step": 15048 + }, + { + "epoch": 0.4746569941649582, + "grad_norm": 1.4853315353393555, + "learning_rate": 1.1304375702176882e-05, + "loss": 0.4874, + "step": 15049 + }, + { + "epoch": 0.47468853493139884, + "grad_norm": 1.790420413017273, + "learning_rate": 1.1303362866531729e-05, + "loss": 0.5089, + "step": 15050 + }, + { + "epoch": 0.47472007569783947, + "grad_norm": 1.3542211055755615, + "learning_rate": 1.1302350017284971e-05, + "loss": 0.4989, + "step": 15051 + }, + { + "epoch": 0.4747516164642801, + "grad_norm": 1.7000441551208496, + "learning_rate": 1.1301337154447179e-05, + "loss": 0.4979, + "step": 15052 + }, + { + "epoch": 0.47478315723072073, + "grad_norm": 1.717982292175293, + "learning_rate": 1.1300324278028922e-05, + "loss": 0.8178, + "step": 15053 + }, + { + "epoch": 0.4748146979971613, + "grad_norm": 1.580152153968811, + "learning_rate": 1.1299311388040774e-05, + "loss": 0.4615, + "step": 15054 + }, + { + "epoch": 0.47484623876360194, + "grad_norm": 2.760098695755005, + "learning_rate": 1.1298298484493304e-05, + "loss": 0.7867, + "step": 15055 + }, + { + "epoch": 0.47487777953004257, + "grad_norm": 1.3928579092025757, + "learning_rate": 1.129728556739708e-05, + "loss": 0.7718, + "step": 15056 + }, + { + "epoch": 0.4749093202964832, + "grad_norm": 2.601443290710449, + "learning_rate": 1.1296272636762675e-05, + "loss": 0.5454, + "step": 15057 + }, + { + "epoch": 0.4749408610629238, + "grad_norm": 1.3348276615142822, + "learning_rate": 1.1295259692600655e-05, + "loss": 0.804, + "step": 15058 + }, + { + "epoch": 0.47497240182936445, + "grad_norm": 1.533949613571167, + "learning_rate": 1.1294246734921596e-05, + "loss": 0.4737, + "step": 15059 + }, + { + "epoch": 0.4750039425958051, + "grad_norm": 1.7021982669830322, + "learning_rate": 1.1293233763736069e-05, + "loss": 0.4818, + "step": 15060 + }, + { + "epoch": 0.4750354833622457, + "grad_norm": 1.5034101009368896, + "learning_rate": 1.1292220779054641e-05, + "loss": 0.4886, + "step": 15061 + }, + { + "epoch": 0.47506702412868634, + "grad_norm": 2.0325000286102295, + "learning_rate": 1.1291207780887886e-05, + "loss": 0.4675, + "step": 15062 + }, + { + "epoch": 0.475098564895127, + "grad_norm": 1.3479516506195068, + "learning_rate": 1.1290194769246374e-05, + "loss": 0.4544, + "step": 15063 + }, + { + "epoch": 0.4751301056615676, + "grad_norm": 1.6493029594421387, + "learning_rate": 1.1289181744140682e-05, + "loss": 0.477, + "step": 15064 + }, + { + "epoch": 0.4751616464280082, + "grad_norm": 1.4154305458068848, + "learning_rate": 1.1288168705581376e-05, + "loss": 0.4837, + "step": 15065 + }, + { + "epoch": 0.4751931871944488, + "grad_norm": 1.6903586387634277, + "learning_rate": 1.128715565357903e-05, + "loss": 0.5075, + "step": 15066 + }, + { + "epoch": 0.47522472796088944, + "grad_norm": 1.773379921913147, + "learning_rate": 1.1286142588144213e-05, + "loss": 0.4985, + "step": 15067 + }, + { + "epoch": 0.47525626872733007, + "grad_norm": 1.2191234827041626, + "learning_rate": 1.12851295092875e-05, + "loss": 0.4981, + "step": 15068 + }, + { + "epoch": 0.4752878094937707, + "grad_norm": 1.9579527378082275, + "learning_rate": 1.1284116417019461e-05, + "loss": 0.3907, + "step": 15069 + }, + { + "epoch": 0.4753193502602113, + "grad_norm": 1.6907227039337158, + "learning_rate": 1.1283103311350674e-05, + "loss": 0.4935, + "step": 15070 + }, + { + "epoch": 0.47535089102665196, + "grad_norm": 1.535033106803894, + "learning_rate": 1.1282090192291706e-05, + "loss": 0.4586, + "step": 15071 + }, + { + "epoch": 0.4753824317930926, + "grad_norm": 1.3784141540527344, + "learning_rate": 1.128107705985313e-05, + "loss": 0.498, + "step": 15072 + }, + { + "epoch": 0.4754139725595332, + "grad_norm": 1.5161200761795044, + "learning_rate": 1.1280063914045521e-05, + "loss": 0.4777, + "step": 15073 + }, + { + "epoch": 0.47544551332597385, + "grad_norm": 2.1700103282928467, + "learning_rate": 1.1279050754879452e-05, + "loss": 0.8148, + "step": 15074 + }, + { + "epoch": 0.4754770540924144, + "grad_norm": 1.9969807863235474, + "learning_rate": 1.1278037582365496e-05, + "loss": 0.8112, + "step": 15075 + }, + { + "epoch": 0.47550859485885505, + "grad_norm": 1.633766770362854, + "learning_rate": 1.1277024396514222e-05, + "loss": 0.5097, + "step": 15076 + }, + { + "epoch": 0.4755401356252957, + "grad_norm": 1.5523005723953247, + "learning_rate": 1.127601119733621e-05, + "loss": 0.5047, + "step": 15077 + }, + { + "epoch": 0.4755716763917363, + "grad_norm": 1.442313551902771, + "learning_rate": 1.1274997984842031e-05, + "loss": 0.468, + "step": 15078 + }, + { + "epoch": 0.47560321715817694, + "grad_norm": 1.537421464920044, + "learning_rate": 1.1273984759042254e-05, + "loss": 0.4729, + "step": 15079 + }, + { + "epoch": 0.47563475792461757, + "grad_norm": 1.8106931447982788, + "learning_rate": 1.1272971519947457e-05, + "loss": 0.483, + "step": 15080 + }, + { + "epoch": 0.4756662986910582, + "grad_norm": 1.5906145572662354, + "learning_rate": 1.127195826756822e-05, + "loss": 0.4765, + "step": 15081 + }, + { + "epoch": 0.47569783945749883, + "grad_norm": 1.4192105531692505, + "learning_rate": 1.1270945001915106e-05, + "loss": 0.5149, + "step": 15082 + }, + { + "epoch": 0.47572938022393946, + "grad_norm": 1.2935028076171875, + "learning_rate": 1.1269931722998692e-05, + "loss": 0.4794, + "step": 15083 + }, + { + "epoch": 0.4757609209903801, + "grad_norm": 2.8827905654907227, + "learning_rate": 1.1268918430829559e-05, + "loss": 0.8472, + "step": 15084 + }, + { + "epoch": 0.4757924617568207, + "grad_norm": 3.540153980255127, + "learning_rate": 1.1267905125418278e-05, + "loss": 0.4563, + "step": 15085 + }, + { + "epoch": 0.4758240025232613, + "grad_norm": 1.8653751611709595, + "learning_rate": 1.1266891806775417e-05, + "loss": 0.5038, + "step": 15086 + }, + { + "epoch": 0.4758555432897019, + "grad_norm": 1.5736104249954224, + "learning_rate": 1.1265878474911558e-05, + "loss": 0.4921, + "step": 15087 + }, + { + "epoch": 0.47588708405614255, + "grad_norm": 1.694862961769104, + "learning_rate": 1.1264865129837276e-05, + "loss": 0.4719, + "step": 15088 + }, + { + "epoch": 0.4759186248225832, + "grad_norm": 1.9767653942108154, + "learning_rate": 1.1263851771563142e-05, + "loss": 0.8006, + "step": 15089 + }, + { + "epoch": 0.4759501655890238, + "grad_norm": 3.952773094177246, + "learning_rate": 1.1262838400099733e-05, + "loss": 0.5015, + "step": 15090 + }, + { + "epoch": 0.47598170635546444, + "grad_norm": 1.5323115587234497, + "learning_rate": 1.1261825015457627e-05, + "loss": 0.4345, + "step": 15091 + }, + { + "epoch": 0.47601324712190507, + "grad_norm": 1.6975524425506592, + "learning_rate": 1.1260811617647398e-05, + "loss": 0.4593, + "step": 15092 + }, + { + "epoch": 0.4760447878883457, + "grad_norm": 1.4321725368499756, + "learning_rate": 1.1259798206679617e-05, + "loss": 0.477, + "step": 15093 + }, + { + "epoch": 0.47607632865478633, + "grad_norm": 1.299696922302246, + "learning_rate": 1.1258784782564869e-05, + "loss": 0.4977, + "step": 15094 + }, + { + "epoch": 0.47610786942122696, + "grad_norm": 1.3916149139404297, + "learning_rate": 1.1257771345313721e-05, + "loss": 0.4871, + "step": 15095 + }, + { + "epoch": 0.47613941018766753, + "grad_norm": 1.8388036489486694, + "learning_rate": 1.1256757894936751e-05, + "loss": 0.4878, + "step": 15096 + }, + { + "epoch": 0.47617095095410816, + "grad_norm": 1.9830187559127808, + "learning_rate": 1.1255744431444538e-05, + "loss": 0.3839, + "step": 15097 + }, + { + "epoch": 0.4762024917205488, + "grad_norm": 1.8305565118789673, + "learning_rate": 1.125473095484766e-05, + "loss": 0.4824, + "step": 15098 + }, + { + "epoch": 0.4762340324869894, + "grad_norm": 1.5354970693588257, + "learning_rate": 1.1253717465156685e-05, + "loss": 0.5099, + "step": 15099 + }, + { + "epoch": 0.47626557325343005, + "grad_norm": 1.4233291149139404, + "learning_rate": 1.1252703962382197e-05, + "loss": 0.481, + "step": 15100 + }, + { + "epoch": 0.4762971140198707, + "grad_norm": 2.152719497680664, + "learning_rate": 1.1251690446534769e-05, + "loss": 0.4805, + "step": 15101 + }, + { + "epoch": 0.4763286547863113, + "grad_norm": 1.442796230316162, + "learning_rate": 1.1250676917624981e-05, + "loss": 0.4697, + "step": 15102 + }, + { + "epoch": 0.47636019555275194, + "grad_norm": 1.7915996313095093, + "learning_rate": 1.1249663375663407e-05, + "loss": 0.4964, + "step": 15103 + }, + { + "epoch": 0.47639173631919257, + "grad_norm": 1.592603325843811, + "learning_rate": 1.1248649820660627e-05, + "loss": 0.4298, + "step": 15104 + }, + { + "epoch": 0.4764232770856332, + "grad_norm": 1.3983842134475708, + "learning_rate": 1.1247636252627216e-05, + "loss": 0.4789, + "step": 15105 + }, + { + "epoch": 0.47645481785207383, + "grad_norm": 1.5914901494979858, + "learning_rate": 1.1246622671573752e-05, + "loss": 0.5074, + "step": 15106 + }, + { + "epoch": 0.4764863586185144, + "grad_norm": 1.6583564281463623, + "learning_rate": 1.124560907751081e-05, + "loss": 0.4755, + "step": 15107 + }, + { + "epoch": 0.47651789938495503, + "grad_norm": 1.5240424871444702, + "learning_rate": 1.1244595470448974e-05, + "loss": 0.4911, + "step": 15108 + }, + { + "epoch": 0.47654944015139566, + "grad_norm": 1.7430611848831177, + "learning_rate": 1.1243581850398817e-05, + "loss": 0.5015, + "step": 15109 + }, + { + "epoch": 0.4765809809178363, + "grad_norm": 1.8882094621658325, + "learning_rate": 1.1242568217370915e-05, + "loss": 0.4693, + "step": 15110 + }, + { + "epoch": 0.4766125216842769, + "grad_norm": 1.6374999284744263, + "learning_rate": 1.124155457137585e-05, + "loss": 0.5159, + "step": 15111 + }, + { + "epoch": 0.47664406245071755, + "grad_norm": 1.350687026977539, + "learning_rate": 1.12405409124242e-05, + "loss": 0.4717, + "step": 15112 + }, + { + "epoch": 0.4766756032171582, + "grad_norm": 1.2803001403808594, + "learning_rate": 1.1239527240526542e-05, + "loss": 0.4773, + "step": 15113 + }, + { + "epoch": 0.4767071439835988, + "grad_norm": 1.483081579208374, + "learning_rate": 1.1238513555693456e-05, + "loss": 0.4691, + "step": 15114 + }, + { + "epoch": 0.47673868475003944, + "grad_norm": 1.6992698907852173, + "learning_rate": 1.123749985793552e-05, + "loss": 0.4708, + "step": 15115 + }, + { + "epoch": 0.4767702255164801, + "grad_norm": 1.5382977724075317, + "learning_rate": 1.1236486147263308e-05, + "loss": 0.4842, + "step": 15116 + }, + { + "epoch": 0.4768017662829207, + "grad_norm": 1.4895633459091187, + "learning_rate": 1.1235472423687406e-05, + "loss": 0.4854, + "step": 15117 + }, + { + "epoch": 0.4768333070493613, + "grad_norm": 1.3766562938690186, + "learning_rate": 1.1234458687218388e-05, + "loss": 0.4849, + "step": 15118 + }, + { + "epoch": 0.4768648478158019, + "grad_norm": 1.763208031654358, + "learning_rate": 1.1233444937866838e-05, + "loss": 0.4954, + "step": 15119 + }, + { + "epoch": 0.47689638858224254, + "grad_norm": 1.807555079460144, + "learning_rate": 1.1232431175643328e-05, + "loss": 0.5082, + "step": 15120 + }, + { + "epoch": 0.47692792934868317, + "grad_norm": 1.4596447944641113, + "learning_rate": 1.1231417400558444e-05, + "loss": 0.4711, + "step": 15121 + }, + { + "epoch": 0.4769594701151238, + "grad_norm": 1.7019745111465454, + "learning_rate": 1.1230403612622764e-05, + "loss": 0.4637, + "step": 15122 + }, + { + "epoch": 0.4769910108815644, + "grad_norm": 1.7920082807540894, + "learning_rate": 1.1229389811846868e-05, + "loss": 0.5075, + "step": 15123 + }, + { + "epoch": 0.47702255164800506, + "grad_norm": 1.4317238330841064, + "learning_rate": 1.1228375998241333e-05, + "loss": 0.4524, + "step": 15124 + }, + { + "epoch": 0.4770540924144457, + "grad_norm": 2.2114994525909424, + "learning_rate": 1.1227362171816738e-05, + "loss": 0.4857, + "step": 15125 + }, + { + "epoch": 0.4770856331808863, + "grad_norm": 2.208381175994873, + "learning_rate": 1.1226348332583673e-05, + "loss": 0.4767, + "step": 15126 + }, + { + "epoch": 0.47711717394732694, + "grad_norm": 1.4002275466918945, + "learning_rate": 1.1225334480552703e-05, + "loss": 0.5053, + "step": 15127 + }, + { + "epoch": 0.4771487147137675, + "grad_norm": 1.4580498933792114, + "learning_rate": 1.122432061573442e-05, + "loss": 0.4923, + "step": 15128 + }, + { + "epoch": 0.47718025548020815, + "grad_norm": 1.139980673789978, + "learning_rate": 1.12233067381394e-05, + "loss": 0.4726, + "step": 15129 + }, + { + "epoch": 0.4772117962466488, + "grad_norm": 1.5255730152130127, + "learning_rate": 1.1222292847778225e-05, + "loss": 0.4795, + "step": 15130 + }, + { + "epoch": 0.4772433370130894, + "grad_norm": 1.3359575271606445, + "learning_rate": 1.1221278944661474e-05, + "loss": 0.4807, + "step": 15131 + }, + { + "epoch": 0.47727487777953004, + "grad_norm": 1.6256061792373657, + "learning_rate": 1.1220265028799731e-05, + "loss": 0.4803, + "step": 15132 + }, + { + "epoch": 0.47730641854597067, + "grad_norm": 1.3510944843292236, + "learning_rate": 1.1219251100203571e-05, + "loss": 0.4654, + "step": 15133 + }, + { + "epoch": 0.4773379593124113, + "grad_norm": 1.269849419593811, + "learning_rate": 1.1218237158883584e-05, + "loss": 0.4937, + "step": 15134 + }, + { + "epoch": 0.4773695000788519, + "grad_norm": 3.3161473274230957, + "learning_rate": 1.1217223204850342e-05, + "loss": 0.8354, + "step": 15135 + }, + { + "epoch": 0.47740104084529256, + "grad_norm": 1.5814182758331299, + "learning_rate": 1.1216209238114435e-05, + "loss": 0.4697, + "step": 15136 + }, + { + "epoch": 0.4774325816117332, + "grad_norm": 1.3414349555969238, + "learning_rate": 1.1215195258686437e-05, + "loss": 0.5064, + "step": 15137 + }, + { + "epoch": 0.4774641223781738, + "grad_norm": 1.412705421447754, + "learning_rate": 1.1214181266576933e-05, + "loss": 0.5005, + "step": 15138 + }, + { + "epoch": 0.4774956631446144, + "grad_norm": 2.3013088703155518, + "learning_rate": 1.1213167261796508e-05, + "loss": 0.4109, + "step": 15139 + }, + { + "epoch": 0.477527203911055, + "grad_norm": 1.4691543579101562, + "learning_rate": 1.1212153244355735e-05, + "loss": 0.3977, + "step": 15140 + }, + { + "epoch": 0.47755874467749565, + "grad_norm": 1.6147719621658325, + "learning_rate": 1.1211139214265205e-05, + "loss": 0.4938, + "step": 15141 + }, + { + "epoch": 0.4775902854439363, + "grad_norm": 1.435034155845642, + "learning_rate": 1.1210125171535498e-05, + "loss": 0.4602, + "step": 15142 + }, + { + "epoch": 0.4776218262103769, + "grad_norm": 1.434988021850586, + "learning_rate": 1.1209111116177192e-05, + "loss": 0.4892, + "step": 15143 + }, + { + "epoch": 0.47765336697681754, + "grad_norm": 1.2282987833023071, + "learning_rate": 1.1208097048200876e-05, + "loss": 0.4764, + "step": 15144 + }, + { + "epoch": 0.47768490774325817, + "grad_norm": 1.4976317882537842, + "learning_rate": 1.1207082967617127e-05, + "loss": 0.516, + "step": 15145 + }, + { + "epoch": 0.4777164485096988, + "grad_norm": 1.7153240442276, + "learning_rate": 1.1206068874436533e-05, + "loss": 0.3545, + "step": 15146 + }, + { + "epoch": 0.47774798927613943, + "grad_norm": 1.4121875762939453, + "learning_rate": 1.120505476866967e-05, + "loss": 0.4672, + "step": 15147 + }, + { + "epoch": 0.47777953004258006, + "grad_norm": 1.7709863185882568, + "learning_rate": 1.1204040650327125e-05, + "loss": 0.5165, + "step": 15148 + }, + { + "epoch": 0.47781107080902063, + "grad_norm": 1.4327448606491089, + "learning_rate": 1.1203026519419482e-05, + "loss": 0.4803, + "step": 15149 + }, + { + "epoch": 0.47784261157546126, + "grad_norm": 1.9256662130355835, + "learning_rate": 1.1202012375957324e-05, + "loss": 0.3856, + "step": 15150 + }, + { + "epoch": 0.4778741523419019, + "grad_norm": 1.7809672355651855, + "learning_rate": 1.1200998219951232e-05, + "loss": 0.4906, + "step": 15151 + }, + { + "epoch": 0.4779056931083425, + "grad_norm": 1.5517243146896362, + "learning_rate": 1.1199984051411792e-05, + "loss": 0.4988, + "step": 15152 + }, + { + "epoch": 0.47793723387478315, + "grad_norm": 2.3833231925964355, + "learning_rate": 1.1198969870349588e-05, + "loss": 0.4714, + "step": 15153 + }, + { + "epoch": 0.4779687746412238, + "grad_norm": 2.193321466445923, + "learning_rate": 1.11979556767752e-05, + "loss": 0.7701, + "step": 15154 + }, + { + "epoch": 0.4780003154076644, + "grad_norm": 1.7967450618743896, + "learning_rate": 1.1196941470699216e-05, + "loss": 0.5168, + "step": 15155 + }, + { + "epoch": 0.47803185617410504, + "grad_norm": 1.4951369762420654, + "learning_rate": 1.1195927252132214e-05, + "loss": 0.477, + "step": 15156 + }, + { + "epoch": 0.47806339694054567, + "grad_norm": 1.5893397331237793, + "learning_rate": 1.119491302108479e-05, + "loss": 0.785, + "step": 15157 + }, + { + "epoch": 0.4780949377069863, + "grad_norm": 1.4170268774032593, + "learning_rate": 1.1193898777567515e-05, + "loss": 0.4652, + "step": 15158 + }, + { + "epoch": 0.47812647847342693, + "grad_norm": 5.281447410583496, + "learning_rate": 1.1192884521590982e-05, + "loss": 0.7818, + "step": 15159 + }, + { + "epoch": 0.4781580192398675, + "grad_norm": 1.2258938550949097, + "learning_rate": 1.1191870253165771e-05, + "loss": 0.4735, + "step": 15160 + }, + { + "epoch": 0.47818956000630813, + "grad_norm": 1.4742491245269775, + "learning_rate": 1.119085597230247e-05, + "loss": 0.4595, + "step": 15161 + }, + { + "epoch": 0.47822110077274876, + "grad_norm": 1.5643917322158813, + "learning_rate": 1.1189841679011661e-05, + "loss": 0.7443, + "step": 15162 + }, + { + "epoch": 0.4782526415391894, + "grad_norm": 1.4009277820587158, + "learning_rate": 1.1188827373303933e-05, + "loss": 0.4762, + "step": 15163 + }, + { + "epoch": 0.47828418230563, + "grad_norm": 1.5249261856079102, + "learning_rate": 1.1187813055189865e-05, + "loss": 0.4793, + "step": 15164 + }, + { + "epoch": 0.47831572307207065, + "grad_norm": 1.4738737344741821, + "learning_rate": 1.1186798724680047e-05, + "loss": 0.4549, + "step": 15165 + }, + { + "epoch": 0.4783472638385113, + "grad_norm": 1.3858675956726074, + "learning_rate": 1.1185784381785061e-05, + "loss": 0.4943, + "step": 15166 + }, + { + "epoch": 0.4783788046049519, + "grad_norm": 1.4416414499282837, + "learning_rate": 1.1184770026515497e-05, + "loss": 0.4924, + "step": 15167 + }, + { + "epoch": 0.47841034537139254, + "grad_norm": 1.5894460678100586, + "learning_rate": 1.1183755658881934e-05, + "loss": 0.4786, + "step": 15168 + }, + { + "epoch": 0.47844188613783317, + "grad_norm": 1.7157464027404785, + "learning_rate": 1.1182741278894964e-05, + "loss": 0.4415, + "step": 15169 + }, + { + "epoch": 0.47847342690427375, + "grad_norm": 1.600782036781311, + "learning_rate": 1.1181726886565172e-05, + "loss": 0.5152, + "step": 15170 + }, + { + "epoch": 0.4785049676707144, + "grad_norm": 1.5724308490753174, + "learning_rate": 1.118071248190314e-05, + "loss": 0.4634, + "step": 15171 + }, + { + "epoch": 0.478536508437155, + "grad_norm": 1.7787883281707764, + "learning_rate": 1.1179698064919458e-05, + "loss": 0.4775, + "step": 15172 + }, + { + "epoch": 0.47856804920359564, + "grad_norm": 1.5131094455718994, + "learning_rate": 1.1178683635624708e-05, + "loss": 0.4577, + "step": 15173 + }, + { + "epoch": 0.47859958997003627, + "grad_norm": 1.744950532913208, + "learning_rate": 1.1177669194029483e-05, + "loss": 0.4665, + "step": 15174 + }, + { + "epoch": 0.4786311307364769, + "grad_norm": 1.4160096645355225, + "learning_rate": 1.1176654740144362e-05, + "loss": 0.5009, + "step": 15175 + }, + { + "epoch": 0.4786626715029175, + "grad_norm": 1.9245774745941162, + "learning_rate": 1.1175640273979935e-05, + "loss": 0.4181, + "step": 15176 + }, + { + "epoch": 0.47869421226935815, + "grad_norm": 1.7454862594604492, + "learning_rate": 1.1174625795546791e-05, + "loss": 0.4598, + "step": 15177 + }, + { + "epoch": 0.4787257530357988, + "grad_norm": 1.3757494688034058, + "learning_rate": 1.1173611304855514e-05, + "loss": 0.4938, + "step": 15178 + }, + { + "epoch": 0.4787572938022394, + "grad_norm": 1.6153229475021362, + "learning_rate": 1.117259680191669e-05, + "loss": 0.4883, + "step": 15179 + }, + { + "epoch": 0.47878883456868004, + "grad_norm": 1.3443403244018555, + "learning_rate": 1.117158228674091e-05, + "loss": 0.4887, + "step": 15180 + }, + { + "epoch": 0.4788203753351206, + "grad_norm": 2.007833957672119, + "learning_rate": 1.117056775933876e-05, + "loss": 0.4973, + "step": 15181 + }, + { + "epoch": 0.47885191610156125, + "grad_norm": 1.6309176683425903, + "learning_rate": 1.1169553219720828e-05, + "loss": 0.4891, + "step": 15182 + }, + { + "epoch": 0.4788834568680019, + "grad_norm": 1.8512054681777954, + "learning_rate": 1.1168538667897695e-05, + "loss": 0.4525, + "step": 15183 + }, + { + "epoch": 0.4789149976344425, + "grad_norm": 1.586592435836792, + "learning_rate": 1.1167524103879958e-05, + "loss": 0.4969, + "step": 15184 + }, + { + "epoch": 0.47894653840088314, + "grad_norm": 1.5827257633209229, + "learning_rate": 1.1166509527678197e-05, + "loss": 0.4727, + "step": 15185 + }, + { + "epoch": 0.47897807916732377, + "grad_norm": 1.5509263277053833, + "learning_rate": 1.1165494939303003e-05, + "loss": 0.3748, + "step": 15186 + }, + { + "epoch": 0.4790096199337644, + "grad_norm": 1.6716846227645874, + "learning_rate": 1.1164480338764969e-05, + "loss": 0.494, + "step": 15187 + }, + { + "epoch": 0.479041160700205, + "grad_norm": 1.4784488677978516, + "learning_rate": 1.1163465726074674e-05, + "loss": 0.4788, + "step": 15188 + }, + { + "epoch": 0.47907270146664566, + "grad_norm": 1.3951148986816406, + "learning_rate": 1.116245110124271e-05, + "loss": 0.4781, + "step": 15189 + }, + { + "epoch": 0.4791042422330863, + "grad_norm": 1.5198595523834229, + "learning_rate": 1.116143646427967e-05, + "loss": 0.4915, + "step": 15190 + }, + { + "epoch": 0.4791357829995269, + "grad_norm": 1.5184317827224731, + "learning_rate": 1.1160421815196138e-05, + "loss": 0.514, + "step": 15191 + }, + { + "epoch": 0.4791673237659675, + "grad_norm": 1.364593505859375, + "learning_rate": 1.1159407154002703e-05, + "loss": 0.4708, + "step": 15192 + }, + { + "epoch": 0.4791988645324081, + "grad_norm": 1.3475112915039062, + "learning_rate": 1.1158392480709952e-05, + "loss": 0.4859, + "step": 15193 + }, + { + "epoch": 0.47923040529884875, + "grad_norm": 1.3584566116333008, + "learning_rate": 1.1157377795328479e-05, + "loss": 0.5123, + "step": 15194 + }, + { + "epoch": 0.4792619460652894, + "grad_norm": 1.2676457166671753, + "learning_rate": 1.1156363097868868e-05, + "loss": 0.4776, + "step": 15195 + }, + { + "epoch": 0.47929348683173, + "grad_norm": 1.5924098491668701, + "learning_rate": 1.1155348388341709e-05, + "loss": 0.4072, + "step": 15196 + }, + { + "epoch": 0.47932502759817064, + "grad_norm": 2.2513926029205322, + "learning_rate": 1.1154333666757595e-05, + "loss": 0.7958, + "step": 15197 + }, + { + "epoch": 0.47935656836461127, + "grad_norm": 2.165196657180786, + "learning_rate": 1.1153318933127114e-05, + "loss": 0.4906, + "step": 15198 + }, + { + "epoch": 0.4793881091310519, + "grad_norm": 1.8932143449783325, + "learning_rate": 1.115230418746085e-05, + "loss": 0.3801, + "step": 15199 + }, + { + "epoch": 0.47941964989749253, + "grad_norm": 1.593468189239502, + "learning_rate": 1.1151289429769399e-05, + "loss": 0.49, + "step": 15200 + }, + { + "epoch": 0.47945119066393316, + "grad_norm": 1.8677772283554077, + "learning_rate": 1.115027466006335e-05, + "loss": 0.4891, + "step": 15201 + }, + { + "epoch": 0.47948273143037373, + "grad_norm": 1.6089208126068115, + "learning_rate": 1.1149259878353293e-05, + "loss": 0.4634, + "step": 15202 + }, + { + "epoch": 0.47951427219681436, + "grad_norm": 1.3215142488479614, + "learning_rate": 1.1148245084649814e-05, + "loss": 0.4873, + "step": 15203 + }, + { + "epoch": 0.479545812963255, + "grad_norm": 1.495591402053833, + "learning_rate": 1.1147230278963506e-05, + "loss": 0.3604, + "step": 15204 + }, + { + "epoch": 0.4795773537296956, + "grad_norm": 1.5803749561309814, + "learning_rate": 1.114621546130496e-05, + "loss": 0.502, + "step": 15205 + }, + { + "epoch": 0.47960889449613625, + "grad_norm": 1.4011191129684448, + "learning_rate": 1.1145200631684764e-05, + "loss": 0.4877, + "step": 15206 + }, + { + "epoch": 0.4796404352625769, + "grad_norm": 1.7759592533111572, + "learning_rate": 1.1144185790113514e-05, + "loss": 0.5032, + "step": 15207 + }, + { + "epoch": 0.4796719760290175, + "grad_norm": 2.0200746059417725, + "learning_rate": 1.1143170936601796e-05, + "loss": 0.7894, + "step": 15208 + }, + { + "epoch": 0.47970351679545814, + "grad_norm": 1.6285525560379028, + "learning_rate": 1.11421560711602e-05, + "loss": 0.4718, + "step": 15209 + }, + { + "epoch": 0.47973505756189877, + "grad_norm": 1.7694486379623413, + "learning_rate": 1.1141141193799317e-05, + "loss": 0.4434, + "step": 15210 + }, + { + "epoch": 0.4797665983283394, + "grad_norm": 1.6453965902328491, + "learning_rate": 1.1140126304529742e-05, + "loss": 0.4893, + "step": 15211 + }, + { + "epoch": 0.47979813909478003, + "grad_norm": 1.4960947036743164, + "learning_rate": 1.1139111403362063e-05, + "loss": 0.4851, + "step": 15212 + }, + { + "epoch": 0.4798296798612206, + "grad_norm": 2.9268136024475098, + "learning_rate": 1.1138096490306873e-05, + "loss": 0.4025, + "step": 15213 + }, + { + "epoch": 0.47986122062766123, + "grad_norm": 1.3919808864593506, + "learning_rate": 1.1137081565374762e-05, + "loss": 0.5406, + "step": 15214 + }, + { + "epoch": 0.47989276139410186, + "grad_norm": 1.4690723419189453, + "learning_rate": 1.113606662857632e-05, + "loss": 0.4845, + "step": 15215 + }, + { + "epoch": 0.4799243021605425, + "grad_norm": 1.3777823448181152, + "learning_rate": 1.1135051679922143e-05, + "loss": 0.7878, + "step": 15216 + }, + { + "epoch": 0.4799558429269831, + "grad_norm": 1.5249505043029785, + "learning_rate": 1.1134036719422818e-05, + "loss": 0.4646, + "step": 15217 + }, + { + "epoch": 0.47998738369342375, + "grad_norm": 1.4462600946426392, + "learning_rate": 1.1133021747088944e-05, + "loss": 0.4918, + "step": 15218 + }, + { + "epoch": 0.4800189244598644, + "grad_norm": 1.537061333656311, + "learning_rate": 1.1132006762931103e-05, + "loss": 0.7819, + "step": 15219 + }, + { + "epoch": 0.480050465226305, + "grad_norm": 1.759121060371399, + "learning_rate": 1.1130991766959893e-05, + "loss": 0.7986, + "step": 15220 + }, + { + "epoch": 0.48008200599274564, + "grad_norm": 1.4871469736099243, + "learning_rate": 1.112997675918591e-05, + "loss": 0.462, + "step": 15221 + }, + { + "epoch": 0.48011354675918627, + "grad_norm": 1.8336337804794312, + "learning_rate": 1.1128961739619738e-05, + "loss": 0.8138, + "step": 15222 + }, + { + "epoch": 0.48014508752562685, + "grad_norm": 1.416101098060608, + "learning_rate": 1.1127946708271976e-05, + "loss": 0.502, + "step": 15223 + }, + { + "epoch": 0.4801766282920675, + "grad_norm": 1.3994539976119995, + "learning_rate": 1.1126931665153213e-05, + "loss": 0.4823, + "step": 15224 + }, + { + "epoch": 0.4802081690585081, + "grad_norm": 1.6525580883026123, + "learning_rate": 1.1125916610274044e-05, + "loss": 0.5057, + "step": 15225 + }, + { + "epoch": 0.48023970982494874, + "grad_norm": 1.663778305053711, + "learning_rate": 1.112490154364506e-05, + "loss": 0.489, + "step": 15226 + }, + { + "epoch": 0.48027125059138936, + "grad_norm": 5.796261310577393, + "learning_rate": 1.1123886465276853e-05, + "loss": 0.3977, + "step": 15227 + }, + { + "epoch": 0.48030279135783, + "grad_norm": 1.9412832260131836, + "learning_rate": 1.112287137518002e-05, + "loss": 0.4934, + "step": 15228 + }, + { + "epoch": 0.4803343321242706, + "grad_norm": 1.566726803779602, + "learning_rate": 1.1121856273365154e-05, + "loss": 0.4708, + "step": 15229 + }, + { + "epoch": 0.48036587289071125, + "grad_norm": 1.6467139720916748, + "learning_rate": 1.1120841159842843e-05, + "loss": 0.4983, + "step": 15230 + }, + { + "epoch": 0.4803974136571519, + "grad_norm": 1.286820888519287, + "learning_rate": 1.1119826034623689e-05, + "loss": 0.5076, + "step": 15231 + }, + { + "epoch": 0.4804289544235925, + "grad_norm": 1.5871275663375854, + "learning_rate": 1.1118810897718279e-05, + "loss": 0.4726, + "step": 15232 + }, + { + "epoch": 0.48046049519003314, + "grad_norm": 1.4103089570999146, + "learning_rate": 1.1117795749137208e-05, + "loss": 0.4651, + "step": 15233 + }, + { + "epoch": 0.4804920359564737, + "grad_norm": 1.2522900104522705, + "learning_rate": 1.111678058889107e-05, + "loss": 0.4644, + "step": 15234 + }, + { + "epoch": 0.48052357672291435, + "grad_norm": 4.518788814544678, + "learning_rate": 1.1115765416990461e-05, + "loss": 0.4886, + "step": 15235 + }, + { + "epoch": 0.480555117489355, + "grad_norm": 1.3983091115951538, + "learning_rate": 1.111475023344597e-05, + "loss": 0.4653, + "step": 15236 + }, + { + "epoch": 0.4805866582557956, + "grad_norm": 1.8316744565963745, + "learning_rate": 1.1113735038268197e-05, + "loss": 0.7547, + "step": 15237 + }, + { + "epoch": 0.48061819902223624, + "grad_norm": 1.550462245941162, + "learning_rate": 1.111271983146773e-05, + "loss": 0.5213, + "step": 15238 + }, + { + "epoch": 0.48064973978867687, + "grad_norm": 1.521191120147705, + "learning_rate": 1.1111704613055174e-05, + "loss": 0.4154, + "step": 15239 + }, + { + "epoch": 0.4806812805551175, + "grad_norm": 1.7667375802993774, + "learning_rate": 1.1110689383041113e-05, + "loss": 0.771, + "step": 15240 + }, + { + "epoch": 0.4807128213215581, + "grad_norm": 2.7555723190307617, + "learning_rate": 1.110967414143615e-05, + "loss": 0.4751, + "step": 15241 + }, + { + "epoch": 0.48074436208799876, + "grad_norm": 1.5831230878829956, + "learning_rate": 1.1108658888250874e-05, + "loss": 0.4927, + "step": 15242 + }, + { + "epoch": 0.4807759028544394, + "grad_norm": 1.4175859689712524, + "learning_rate": 1.110764362349588e-05, + "loss": 0.4972, + "step": 15243 + }, + { + "epoch": 0.48080744362088, + "grad_norm": 1.5089610815048218, + "learning_rate": 1.1106628347181764e-05, + "loss": 0.4544, + "step": 15244 + }, + { + "epoch": 0.4808389843873206, + "grad_norm": 1.250104546546936, + "learning_rate": 1.1105613059319125e-05, + "loss": 0.491, + "step": 15245 + }, + { + "epoch": 0.4808705251537612, + "grad_norm": 1.6816842555999756, + "learning_rate": 1.1104597759918553e-05, + "loss": 0.4935, + "step": 15246 + }, + { + "epoch": 0.48090206592020185, + "grad_norm": 1.645540714263916, + "learning_rate": 1.1103582448990646e-05, + "loss": 0.5049, + "step": 15247 + }, + { + "epoch": 0.4809336066866425, + "grad_norm": 1.4926751852035522, + "learning_rate": 1.1102567126546e-05, + "loss": 0.7724, + "step": 15248 + }, + { + "epoch": 0.4809651474530831, + "grad_norm": 1.4015733003616333, + "learning_rate": 1.1101551792595209e-05, + "loss": 0.4895, + "step": 15249 + }, + { + "epoch": 0.48099668821952374, + "grad_norm": 1.9533543586730957, + "learning_rate": 1.110053644714887e-05, + "loss": 0.5032, + "step": 15250 + }, + { + "epoch": 0.48102822898596437, + "grad_norm": 1.6928343772888184, + "learning_rate": 1.109952109021758e-05, + "loss": 0.5164, + "step": 15251 + }, + { + "epoch": 0.481059769752405, + "grad_norm": 1.486081600189209, + "learning_rate": 1.1098505721811932e-05, + "loss": 0.4695, + "step": 15252 + }, + { + "epoch": 0.4810913105188456, + "grad_norm": 2.0911355018615723, + "learning_rate": 1.1097490341942525e-05, + "loss": 0.447, + "step": 15253 + }, + { + "epoch": 0.48112285128528626, + "grad_norm": 1.9308971166610718, + "learning_rate": 1.1096474950619954e-05, + "loss": 0.3483, + "step": 15254 + }, + { + "epoch": 0.48115439205172683, + "grad_norm": 1.79154372215271, + "learning_rate": 1.1095459547854814e-05, + "loss": 0.7794, + "step": 15255 + }, + { + "epoch": 0.48118593281816746, + "grad_norm": 1.4021246433258057, + "learning_rate": 1.1094444133657705e-05, + "loss": 0.4622, + "step": 15256 + }, + { + "epoch": 0.4812174735846081, + "grad_norm": 1.369432806968689, + "learning_rate": 1.109342870803922e-05, + "loss": 0.4523, + "step": 15257 + }, + { + "epoch": 0.4812490143510487, + "grad_norm": 1.4589322805404663, + "learning_rate": 1.1092413271009956e-05, + "loss": 0.4866, + "step": 15258 + }, + { + "epoch": 0.48128055511748935, + "grad_norm": 1.601773977279663, + "learning_rate": 1.1091397822580516e-05, + "loss": 0.4846, + "step": 15259 + }, + { + "epoch": 0.48131209588393, + "grad_norm": 1.4620596170425415, + "learning_rate": 1.1090382362761488e-05, + "loss": 0.4919, + "step": 15260 + }, + { + "epoch": 0.4813436366503706, + "grad_norm": 2.0490214824676514, + "learning_rate": 1.1089366891563476e-05, + "loss": 0.4916, + "step": 15261 + }, + { + "epoch": 0.48137517741681124, + "grad_norm": 1.6678699254989624, + "learning_rate": 1.1088351408997073e-05, + "loss": 0.4866, + "step": 15262 + }, + { + "epoch": 0.48140671818325187, + "grad_norm": 1.9437588453292847, + "learning_rate": 1.1087335915072882e-05, + "loss": 0.5015, + "step": 15263 + }, + { + "epoch": 0.4814382589496925, + "grad_norm": 1.6234636306762695, + "learning_rate": 1.108632040980149e-05, + "loss": 0.5007, + "step": 15264 + }, + { + "epoch": 0.48146979971613313, + "grad_norm": 2.1507315635681152, + "learning_rate": 1.1085304893193503e-05, + "loss": 0.5031, + "step": 15265 + }, + { + "epoch": 0.4815013404825737, + "grad_norm": 2.0159072875976562, + "learning_rate": 1.108428936525952e-05, + "loss": 0.4586, + "step": 15266 + }, + { + "epoch": 0.48153288124901433, + "grad_norm": 1.4877257347106934, + "learning_rate": 1.1083273826010132e-05, + "loss": 0.4662, + "step": 15267 + }, + { + "epoch": 0.48156442201545496, + "grad_norm": 1.44420325756073, + "learning_rate": 1.108225827545594e-05, + "loss": 0.785, + "step": 15268 + }, + { + "epoch": 0.4815959627818956, + "grad_norm": 1.5165846347808838, + "learning_rate": 1.1081242713607543e-05, + "loss": 0.4869, + "step": 15269 + }, + { + "epoch": 0.4816275035483362, + "grad_norm": 1.488401174545288, + "learning_rate": 1.1080227140475542e-05, + "loss": 0.5249, + "step": 15270 + }, + { + "epoch": 0.48165904431477685, + "grad_norm": 1.6144026517868042, + "learning_rate": 1.107921155607053e-05, + "loss": 0.5029, + "step": 15271 + }, + { + "epoch": 0.4816905850812175, + "grad_norm": 1.8572165966033936, + "learning_rate": 1.1078195960403108e-05, + "loss": 0.4928, + "step": 15272 + }, + { + "epoch": 0.4817221258476581, + "grad_norm": 1.9867374897003174, + "learning_rate": 1.1077180353483874e-05, + "loss": 0.4495, + "step": 15273 + }, + { + "epoch": 0.48175366661409874, + "grad_norm": 2.3822638988494873, + "learning_rate": 1.1076164735323427e-05, + "loss": 0.7843, + "step": 15274 + }, + { + "epoch": 0.48178520738053937, + "grad_norm": 1.4453810453414917, + "learning_rate": 1.1075149105932362e-05, + "loss": 0.488, + "step": 15275 + }, + { + "epoch": 0.48181674814697995, + "grad_norm": 2.139974355697632, + "learning_rate": 1.1074133465321283e-05, + "loss": 0.5057, + "step": 15276 + }, + { + "epoch": 0.4818482889134206, + "grad_norm": 2.174692153930664, + "learning_rate": 1.1073117813500789e-05, + "loss": 0.4372, + "step": 15277 + }, + { + "epoch": 0.4818798296798612, + "grad_norm": 1.9201583862304688, + "learning_rate": 1.1072102150481477e-05, + "loss": 0.5027, + "step": 15278 + }, + { + "epoch": 0.48191137044630183, + "grad_norm": 1.8036012649536133, + "learning_rate": 1.1071086476273944e-05, + "loss": 0.5362, + "step": 15279 + }, + { + "epoch": 0.48194291121274246, + "grad_norm": 1.760091781616211, + "learning_rate": 1.1070070790888794e-05, + "loss": 0.4815, + "step": 15280 + }, + { + "epoch": 0.4819744519791831, + "grad_norm": 1.7171659469604492, + "learning_rate": 1.1069055094336627e-05, + "loss": 0.5018, + "step": 15281 + }, + { + "epoch": 0.4820059927456237, + "grad_norm": 1.8752437829971313, + "learning_rate": 1.1068039386628036e-05, + "loss": 0.4927, + "step": 15282 + }, + { + "epoch": 0.48203753351206435, + "grad_norm": 1.4177066087722778, + "learning_rate": 1.1067023667773624e-05, + "loss": 0.48, + "step": 15283 + }, + { + "epoch": 0.482069074278505, + "grad_norm": 1.1976460218429565, + "learning_rate": 1.1066007937783996e-05, + "loss": 0.4986, + "step": 15284 + }, + { + "epoch": 0.4821006150449456, + "grad_norm": 1.513056755065918, + "learning_rate": 1.1064992196669744e-05, + "loss": 0.4396, + "step": 15285 + }, + { + "epoch": 0.48213215581138624, + "grad_norm": 1.511112093925476, + "learning_rate": 1.1063976444441471e-05, + "loss": 0.4693, + "step": 15286 + }, + { + "epoch": 0.4821636965778268, + "grad_norm": 1.452431321144104, + "learning_rate": 1.106296068110978e-05, + "loss": 0.4643, + "step": 15287 + }, + { + "epoch": 0.48219523734426745, + "grad_norm": 1.9369378089904785, + "learning_rate": 1.1061944906685267e-05, + "loss": 0.7669, + "step": 15288 + }, + { + "epoch": 0.4822267781107081, + "grad_norm": 1.4557589292526245, + "learning_rate": 1.1060929121178535e-05, + "loss": 0.4683, + "step": 15289 + }, + { + "epoch": 0.4822583188771487, + "grad_norm": 1.4555484056472778, + "learning_rate": 1.1059913324600184e-05, + "loss": 0.4923, + "step": 15290 + }, + { + "epoch": 0.48228985964358934, + "grad_norm": 1.8350872993469238, + "learning_rate": 1.1058897516960817e-05, + "loss": 0.3442, + "step": 15291 + }, + { + "epoch": 0.48232140041002997, + "grad_norm": 1.735532283782959, + "learning_rate": 1.1057881698271027e-05, + "loss": 0.7823, + "step": 15292 + }, + { + "epoch": 0.4823529411764706, + "grad_norm": 1.482828140258789, + "learning_rate": 1.1056865868541423e-05, + "loss": 0.5219, + "step": 15293 + }, + { + "epoch": 0.4823844819429112, + "grad_norm": 1.6323970556259155, + "learning_rate": 1.1055850027782605e-05, + "loss": 0.4739, + "step": 15294 + }, + { + "epoch": 0.48241602270935185, + "grad_norm": 2.1383984088897705, + "learning_rate": 1.1054834176005167e-05, + "loss": 0.4908, + "step": 15295 + }, + { + "epoch": 0.4824475634757925, + "grad_norm": 1.1764476299285889, + "learning_rate": 1.1053818313219718e-05, + "loss": 0.4665, + "step": 15296 + }, + { + "epoch": 0.48247910424223306, + "grad_norm": 2.23323130607605, + "learning_rate": 1.1052802439436858e-05, + "loss": 0.484, + "step": 15297 + }, + { + "epoch": 0.4825106450086737, + "grad_norm": 1.8093461990356445, + "learning_rate": 1.1051786554667186e-05, + "loss": 0.4768, + "step": 15298 + }, + { + "epoch": 0.4825421857751143, + "grad_norm": 1.6712878942489624, + "learning_rate": 1.1050770658921304e-05, + "loss": 0.4878, + "step": 15299 + }, + { + "epoch": 0.48257372654155495, + "grad_norm": 1.758244276046753, + "learning_rate": 1.1049754752209816e-05, + "loss": 0.4846, + "step": 15300 + }, + { + "epoch": 0.4826052673079956, + "grad_norm": 1.4908473491668701, + "learning_rate": 1.104873883454332e-05, + "loss": 0.7895, + "step": 15301 + }, + { + "epoch": 0.4826368080744362, + "grad_norm": 1.7481797933578491, + "learning_rate": 1.1047722905932421e-05, + "loss": 0.5042, + "step": 15302 + }, + { + "epoch": 0.48266834884087684, + "grad_norm": 2.199469804763794, + "learning_rate": 1.104670696638772e-05, + "loss": 0.4005, + "step": 15303 + }, + { + "epoch": 0.48269988960731747, + "grad_norm": 1.4384063482284546, + "learning_rate": 1.1045691015919819e-05, + "loss": 0.4034, + "step": 15304 + }, + { + "epoch": 0.4827314303737581, + "grad_norm": 2.0210769176483154, + "learning_rate": 1.1044675054539318e-05, + "loss": 0.5086, + "step": 15305 + }, + { + "epoch": 0.4827629711401987, + "grad_norm": 1.5171178579330444, + "learning_rate": 1.1043659082256823e-05, + "loss": 0.4402, + "step": 15306 + }, + { + "epoch": 0.48279451190663936, + "grad_norm": 1.3785253763198853, + "learning_rate": 1.1042643099082935e-05, + "loss": 0.4723, + "step": 15307 + }, + { + "epoch": 0.48282605267307993, + "grad_norm": 1.368228793144226, + "learning_rate": 1.1041627105028259e-05, + "loss": 0.5021, + "step": 15308 + }, + { + "epoch": 0.48285759343952056, + "grad_norm": 3.5792505741119385, + "learning_rate": 1.1040611100103393e-05, + "loss": 0.3526, + "step": 15309 + }, + { + "epoch": 0.4828891342059612, + "grad_norm": 1.6015584468841553, + "learning_rate": 1.1039595084318943e-05, + "loss": 0.5098, + "step": 15310 + }, + { + "epoch": 0.4829206749724018, + "grad_norm": 1.355649709701538, + "learning_rate": 1.1038579057685514e-05, + "loss": 0.4987, + "step": 15311 + }, + { + "epoch": 0.48295221573884245, + "grad_norm": 2.4849770069122314, + "learning_rate": 1.1037563020213702e-05, + "loss": 0.5107, + "step": 15312 + }, + { + "epoch": 0.4829837565052831, + "grad_norm": 1.6032501459121704, + "learning_rate": 1.1036546971914116e-05, + "loss": 0.5048, + "step": 15313 + }, + { + "epoch": 0.4830152972717237, + "grad_norm": 1.7147382497787476, + "learning_rate": 1.1035530912797356e-05, + "loss": 0.4511, + "step": 15314 + }, + { + "epoch": 0.48304683803816434, + "grad_norm": 1.5462712049484253, + "learning_rate": 1.1034514842874026e-05, + "loss": 0.4759, + "step": 15315 + }, + { + "epoch": 0.48307837880460497, + "grad_norm": 1.4809784889221191, + "learning_rate": 1.1033498762154733e-05, + "loss": 0.4598, + "step": 15316 + }, + { + "epoch": 0.4831099195710456, + "grad_norm": 1.7001579999923706, + "learning_rate": 1.1032482670650075e-05, + "loss": 0.4659, + "step": 15317 + }, + { + "epoch": 0.48314146033748623, + "grad_norm": 1.4658595323562622, + "learning_rate": 1.1031466568370664e-05, + "loss": 0.4953, + "step": 15318 + }, + { + "epoch": 0.4831730011039268, + "grad_norm": 1.5859532356262207, + "learning_rate": 1.1030450455327093e-05, + "loss": 0.4825, + "step": 15319 + }, + { + "epoch": 0.48320454187036743, + "grad_norm": 1.5092544555664062, + "learning_rate": 1.1029434331529974e-05, + "loss": 0.4673, + "step": 15320 + }, + { + "epoch": 0.48323608263680806, + "grad_norm": 2.313852548599243, + "learning_rate": 1.102841819698991e-05, + "loss": 0.4467, + "step": 15321 + }, + { + "epoch": 0.4832676234032487, + "grad_norm": 1.949758768081665, + "learning_rate": 1.10274020517175e-05, + "loss": 0.7658, + "step": 15322 + }, + { + "epoch": 0.4832991641696893, + "grad_norm": 1.3684483766555786, + "learning_rate": 1.1026385895723353e-05, + "loss": 0.478, + "step": 15323 + }, + { + "epoch": 0.48333070493612995, + "grad_norm": 1.1617367267608643, + "learning_rate": 1.1025369729018072e-05, + "loss": 0.487, + "step": 15324 + }, + { + "epoch": 0.4833622457025706, + "grad_norm": 2.2905654907226562, + "learning_rate": 1.1024353551612262e-05, + "loss": 0.4252, + "step": 15325 + }, + { + "epoch": 0.4833937864690112, + "grad_norm": 1.8385931253433228, + "learning_rate": 1.1023337363516527e-05, + "loss": 0.3703, + "step": 15326 + }, + { + "epoch": 0.48342532723545184, + "grad_norm": 1.715287446975708, + "learning_rate": 1.1022321164741471e-05, + "loss": 0.4578, + "step": 15327 + }, + { + "epoch": 0.48345686800189247, + "grad_norm": 2.523301839828491, + "learning_rate": 1.1021304955297702e-05, + "loss": 0.4864, + "step": 15328 + }, + { + "epoch": 0.48348840876833304, + "grad_norm": 2.166591167449951, + "learning_rate": 1.1020288735195824e-05, + "loss": 0.4662, + "step": 15329 + }, + { + "epoch": 0.4835199495347737, + "grad_norm": 1.7061388492584229, + "learning_rate": 1.101927250444644e-05, + "loss": 0.4887, + "step": 15330 + }, + { + "epoch": 0.4835514903012143, + "grad_norm": 1.747825264930725, + "learning_rate": 1.1018256263060151e-05, + "loss": 0.4659, + "step": 15331 + }, + { + "epoch": 0.48358303106765493, + "grad_norm": 1.5770195722579956, + "learning_rate": 1.1017240011047574e-05, + "loss": 0.5079, + "step": 15332 + }, + { + "epoch": 0.48361457183409556, + "grad_norm": 1.4540163278579712, + "learning_rate": 1.1016223748419304e-05, + "loss": 0.8012, + "step": 15333 + }, + { + "epoch": 0.4836461126005362, + "grad_norm": 1.5701801776885986, + "learning_rate": 1.101520747518595e-05, + "loss": 0.4101, + "step": 15334 + }, + { + "epoch": 0.4836776533669768, + "grad_norm": 1.7630109786987305, + "learning_rate": 1.1014191191358118e-05, + "loss": 0.4185, + "step": 15335 + }, + { + "epoch": 0.48370919413341745, + "grad_norm": 1.4767484664916992, + "learning_rate": 1.1013174896946414e-05, + "loss": 0.4735, + "step": 15336 + }, + { + "epoch": 0.4837407348998581, + "grad_norm": 1.5732322931289673, + "learning_rate": 1.1012158591961443e-05, + "loss": 0.4752, + "step": 15337 + }, + { + "epoch": 0.4837722756662987, + "grad_norm": 2.213759183883667, + "learning_rate": 1.1011142276413814e-05, + "loss": 0.5015, + "step": 15338 + }, + { + "epoch": 0.48380381643273934, + "grad_norm": 1.346765160560608, + "learning_rate": 1.1010125950314126e-05, + "loss": 0.4804, + "step": 15339 + }, + { + "epoch": 0.4838353571991799, + "grad_norm": 1.6975218057632446, + "learning_rate": 1.100910961367299e-05, + "loss": 0.4796, + "step": 15340 + }, + { + "epoch": 0.48386689796562055, + "grad_norm": 2.7775216102600098, + "learning_rate": 1.1008093266501012e-05, + "loss": 0.5001, + "step": 15341 + }, + { + "epoch": 0.4838984387320612, + "grad_norm": 1.4109749794006348, + "learning_rate": 1.1007076908808801e-05, + "loss": 0.49, + "step": 15342 + }, + { + "epoch": 0.4839299794985018, + "grad_norm": 2.6318795680999756, + "learning_rate": 1.1006060540606958e-05, + "loss": 0.4047, + "step": 15343 + }, + { + "epoch": 0.48396152026494244, + "grad_norm": 1.3441941738128662, + "learning_rate": 1.1005044161906092e-05, + "loss": 0.4703, + "step": 15344 + }, + { + "epoch": 0.48399306103138307, + "grad_norm": 1.8489201068878174, + "learning_rate": 1.100402777271681e-05, + "loss": 0.4938, + "step": 15345 + }, + { + "epoch": 0.4840246017978237, + "grad_norm": 1.7484540939331055, + "learning_rate": 1.1003011373049717e-05, + "loss": 0.4753, + "step": 15346 + }, + { + "epoch": 0.4840561425642643, + "grad_norm": 1.418879508972168, + "learning_rate": 1.100199496291542e-05, + "loss": 0.4973, + "step": 15347 + }, + { + "epoch": 0.48408768333070495, + "grad_norm": 1.46894109249115, + "learning_rate": 1.1000978542324534e-05, + "loss": 0.4602, + "step": 15348 + }, + { + "epoch": 0.4841192240971456, + "grad_norm": 2.0055859088897705, + "learning_rate": 1.0999962111287655e-05, + "loss": 0.452, + "step": 15349 + }, + { + "epoch": 0.48415076486358616, + "grad_norm": 1.6195106506347656, + "learning_rate": 1.0998945669815398e-05, + "loss": 0.5068, + "step": 15350 + }, + { + "epoch": 0.4841823056300268, + "grad_norm": 1.3167834281921387, + "learning_rate": 1.0997929217918363e-05, + "loss": 0.4793, + "step": 15351 + }, + { + "epoch": 0.4842138463964674, + "grad_norm": 1.3968310356140137, + "learning_rate": 1.0996912755607165e-05, + "loss": 0.7855, + "step": 15352 + }, + { + "epoch": 0.48424538716290805, + "grad_norm": 3.7400758266448975, + "learning_rate": 1.099589628289241e-05, + "loss": 0.4653, + "step": 15353 + }, + { + "epoch": 0.4842769279293487, + "grad_norm": 1.696372628211975, + "learning_rate": 1.0994879799784699e-05, + "loss": 0.4452, + "step": 15354 + }, + { + "epoch": 0.4843084686957893, + "grad_norm": 1.5041643381118774, + "learning_rate": 1.0993863306294646e-05, + "loss": 0.4809, + "step": 15355 + }, + { + "epoch": 0.48434000946222994, + "grad_norm": 1.3586714267730713, + "learning_rate": 1.099284680243286e-05, + "loss": 0.4696, + "step": 15356 + }, + { + "epoch": 0.48437155022867057, + "grad_norm": 1.5055129528045654, + "learning_rate": 1.0991830288209944e-05, + "loss": 0.377, + "step": 15357 + }, + { + "epoch": 0.4844030909951112, + "grad_norm": 1.483372449874878, + "learning_rate": 1.0990813763636511e-05, + "loss": 0.4637, + "step": 15358 + }, + { + "epoch": 0.4844346317615518, + "grad_norm": 1.7629101276397705, + "learning_rate": 1.0989797228723166e-05, + "loss": 0.4796, + "step": 15359 + }, + { + "epoch": 0.48446617252799246, + "grad_norm": 1.3127864599227905, + "learning_rate": 1.098878068348052e-05, + "loss": 0.4599, + "step": 15360 + }, + { + "epoch": 0.48449771329443303, + "grad_norm": 1.4283270835876465, + "learning_rate": 1.0987764127919178e-05, + "loss": 0.4587, + "step": 15361 + }, + { + "epoch": 0.48452925406087366, + "grad_norm": 2.4042818546295166, + "learning_rate": 1.0986747562049754e-05, + "loss": 0.3534, + "step": 15362 + }, + { + "epoch": 0.4845607948273143, + "grad_norm": 1.585363507270813, + "learning_rate": 1.0985730985882848e-05, + "loss": 0.5217, + "step": 15363 + }, + { + "epoch": 0.4845923355937549, + "grad_norm": 1.615470051765442, + "learning_rate": 1.0984714399429076e-05, + "loss": 0.4666, + "step": 15364 + }, + { + "epoch": 0.48462387636019555, + "grad_norm": 1.3927704095840454, + "learning_rate": 1.0983697802699043e-05, + "loss": 0.5123, + "step": 15365 + }, + { + "epoch": 0.4846554171266362, + "grad_norm": 1.3947334289550781, + "learning_rate": 1.0982681195703361e-05, + "loss": 0.4797, + "step": 15366 + }, + { + "epoch": 0.4846869578930768, + "grad_norm": 1.4092000722885132, + "learning_rate": 1.0981664578452637e-05, + "loss": 0.4148, + "step": 15367 + }, + { + "epoch": 0.48471849865951744, + "grad_norm": 1.604712724685669, + "learning_rate": 1.0980647950957481e-05, + "loss": 0.4937, + "step": 15368 + }, + { + "epoch": 0.48475003942595807, + "grad_norm": 1.593460202217102, + "learning_rate": 1.0979631313228502e-05, + "loss": 0.5238, + "step": 15369 + }, + { + "epoch": 0.4847815801923987, + "grad_norm": 1.3493071794509888, + "learning_rate": 1.0978614665276313e-05, + "loss": 0.4849, + "step": 15370 + }, + { + "epoch": 0.48481312095883927, + "grad_norm": 1.6387184858322144, + "learning_rate": 1.0977598007111517e-05, + "loss": 0.4959, + "step": 15371 + }, + { + "epoch": 0.4848446617252799, + "grad_norm": 1.5152522325515747, + "learning_rate": 1.0976581338744726e-05, + "loss": 0.4556, + "step": 15372 + }, + { + "epoch": 0.48487620249172053, + "grad_norm": 1.5098210573196411, + "learning_rate": 1.0975564660186553e-05, + "loss": 0.7957, + "step": 15373 + }, + { + "epoch": 0.48490774325816116, + "grad_norm": 2.3332319259643555, + "learning_rate": 1.0974547971447603e-05, + "loss": 0.7527, + "step": 15374 + }, + { + "epoch": 0.4849392840246018, + "grad_norm": 1.864649772644043, + "learning_rate": 1.0973531272538488e-05, + "loss": 0.4934, + "step": 15375 + }, + { + "epoch": 0.4849708247910424, + "grad_norm": 1.8332375288009644, + "learning_rate": 1.097251456346982e-05, + "loss": 0.5206, + "step": 15376 + }, + { + "epoch": 0.48500236555748305, + "grad_norm": 1.8934587240219116, + "learning_rate": 1.0971497844252206e-05, + "loss": 0.4853, + "step": 15377 + }, + { + "epoch": 0.4850339063239237, + "grad_norm": 1.639394998550415, + "learning_rate": 1.097048111489626e-05, + "loss": 0.4838, + "step": 15378 + }, + { + "epoch": 0.4850654470903643, + "grad_norm": 1.5934600830078125, + "learning_rate": 1.0969464375412587e-05, + "loss": 0.4422, + "step": 15379 + }, + { + "epoch": 0.48509698785680494, + "grad_norm": 1.303411602973938, + "learning_rate": 1.0968447625811802e-05, + "loss": 0.4594, + "step": 15380 + }, + { + "epoch": 0.48512852862324557, + "grad_norm": 1.623360276222229, + "learning_rate": 1.0967430866104514e-05, + "loss": 0.4665, + "step": 15381 + }, + { + "epoch": 0.48516006938968614, + "grad_norm": 1.7123783826828003, + "learning_rate": 1.0966414096301333e-05, + "loss": 0.4995, + "step": 15382 + }, + { + "epoch": 0.4851916101561268, + "grad_norm": 2.0893607139587402, + "learning_rate": 1.0965397316412871e-05, + "loss": 0.4473, + "step": 15383 + }, + { + "epoch": 0.4852231509225674, + "grad_norm": 1.6378031969070435, + "learning_rate": 1.0964380526449738e-05, + "loss": 0.4425, + "step": 15384 + }, + { + "epoch": 0.48525469168900803, + "grad_norm": 2.57240891456604, + "learning_rate": 1.0963363726422545e-05, + "loss": 0.4841, + "step": 15385 + }, + { + "epoch": 0.48528623245544866, + "grad_norm": 2.265856981277466, + "learning_rate": 1.0962346916341904e-05, + "loss": 0.4773, + "step": 15386 + }, + { + "epoch": 0.4853177732218893, + "grad_norm": 2.575002908706665, + "learning_rate": 1.0961330096218425e-05, + "loss": 0.3822, + "step": 15387 + }, + { + "epoch": 0.4853493139883299, + "grad_norm": 1.5287902355194092, + "learning_rate": 1.0960313266062723e-05, + "loss": 0.4866, + "step": 15388 + }, + { + "epoch": 0.48538085475477055, + "grad_norm": 1.8798054456710815, + "learning_rate": 1.0959296425885404e-05, + "loss": 0.4648, + "step": 15389 + }, + { + "epoch": 0.4854123955212112, + "grad_norm": 1.2957355976104736, + "learning_rate": 1.0958279575697083e-05, + "loss": 0.4205, + "step": 15390 + }, + { + "epoch": 0.4854439362876518, + "grad_norm": 1.3117972612380981, + "learning_rate": 1.095726271550837e-05, + "loss": 0.481, + "step": 15391 + }, + { + "epoch": 0.48547547705409244, + "grad_norm": 1.4538999795913696, + "learning_rate": 1.0956245845329876e-05, + "loss": 0.485, + "step": 15392 + }, + { + "epoch": 0.485507017820533, + "grad_norm": 1.295278549194336, + "learning_rate": 1.0955228965172215e-05, + "loss": 0.4933, + "step": 15393 + }, + { + "epoch": 0.48553855858697365, + "grad_norm": 1.9768455028533936, + "learning_rate": 1.0954212075045998e-05, + "loss": 0.479, + "step": 15394 + }, + { + "epoch": 0.4855700993534143, + "grad_norm": 1.56416654586792, + "learning_rate": 1.0953195174961836e-05, + "loss": 0.4681, + "step": 15395 + }, + { + "epoch": 0.4856016401198549, + "grad_norm": 1.3512104749679565, + "learning_rate": 1.0952178264930342e-05, + "loss": 0.4835, + "step": 15396 + }, + { + "epoch": 0.48563318088629553, + "grad_norm": 1.7107014656066895, + "learning_rate": 1.0951161344962131e-05, + "loss": 0.501, + "step": 15397 + }, + { + "epoch": 0.48566472165273616, + "grad_norm": 1.5378273725509644, + "learning_rate": 1.0950144415067814e-05, + "loss": 0.4805, + "step": 15398 + }, + { + "epoch": 0.4856962624191768, + "grad_norm": 1.8725335597991943, + "learning_rate": 1.0949127475257998e-05, + "loss": 0.3911, + "step": 15399 + }, + { + "epoch": 0.4857278031856174, + "grad_norm": 1.6771039962768555, + "learning_rate": 1.0948110525543305e-05, + "loss": 0.4813, + "step": 15400 + }, + { + "epoch": 0.48575934395205805, + "grad_norm": 3.548290729522705, + "learning_rate": 1.0947093565934337e-05, + "loss": 0.4028, + "step": 15401 + }, + { + "epoch": 0.4857908847184987, + "grad_norm": 1.5368006229400635, + "learning_rate": 1.0946076596441715e-05, + "loss": 0.4577, + "step": 15402 + }, + { + "epoch": 0.48582242548493926, + "grad_norm": 1.6771337985992432, + "learning_rate": 1.0945059617076046e-05, + "loss": 0.4319, + "step": 15403 + }, + { + "epoch": 0.4858539662513799, + "grad_norm": 5.489798545837402, + "learning_rate": 1.094404262784795e-05, + "loss": 0.3433, + "step": 15404 + }, + { + "epoch": 0.4858855070178205, + "grad_norm": 1.497842788696289, + "learning_rate": 1.0943025628768036e-05, + "loss": 0.4438, + "step": 15405 + }, + { + "epoch": 0.48591704778426115, + "grad_norm": 1.781061053276062, + "learning_rate": 1.0942008619846913e-05, + "loss": 0.4615, + "step": 15406 + }, + { + "epoch": 0.4859485885507018, + "grad_norm": 1.5366371870040894, + "learning_rate": 1.0940991601095203e-05, + "loss": 0.4681, + "step": 15407 + }, + { + "epoch": 0.4859801293171424, + "grad_norm": 1.7722275257110596, + "learning_rate": 1.0939974572523517e-05, + "loss": 0.508, + "step": 15408 + }, + { + "epoch": 0.48601167008358304, + "grad_norm": 1.6259914636611938, + "learning_rate": 1.0938957534142462e-05, + "loss": 0.4703, + "step": 15409 + }, + { + "epoch": 0.48604321085002367, + "grad_norm": 6.340022563934326, + "learning_rate": 1.093794048596266e-05, + "loss": 0.4746, + "step": 15410 + }, + { + "epoch": 0.4860747516164643, + "grad_norm": 1.4123252630233765, + "learning_rate": 1.0936923427994715e-05, + "loss": 0.4549, + "step": 15411 + }, + { + "epoch": 0.4861062923829049, + "grad_norm": 1.73440682888031, + "learning_rate": 1.093590636024925e-05, + "loss": 0.489, + "step": 15412 + }, + { + "epoch": 0.48613783314934556, + "grad_norm": 2.0039000511169434, + "learning_rate": 1.0934889282736873e-05, + "loss": 0.4552, + "step": 15413 + }, + { + "epoch": 0.48616937391578613, + "grad_norm": 1.5304961204528809, + "learning_rate": 1.0933872195468205e-05, + "loss": 0.476, + "step": 15414 + }, + { + "epoch": 0.48620091468222676, + "grad_norm": 1.7267080545425415, + "learning_rate": 1.0932855098453852e-05, + "loss": 0.4637, + "step": 15415 + }, + { + "epoch": 0.4862324554486674, + "grad_norm": 1.5673644542694092, + "learning_rate": 1.093183799170443e-05, + "loss": 0.3715, + "step": 15416 + }, + { + "epoch": 0.486263996215108, + "grad_norm": 1.804439902305603, + "learning_rate": 1.0930820875230559e-05, + "loss": 0.494, + "step": 15417 + }, + { + "epoch": 0.48629553698154865, + "grad_norm": 1.4589476585388184, + "learning_rate": 1.092980374904285e-05, + "loss": 0.4975, + "step": 15418 + }, + { + "epoch": 0.4863270777479893, + "grad_norm": 1.3830586671829224, + "learning_rate": 1.0928786613151914e-05, + "loss": 0.4523, + "step": 15419 + }, + { + "epoch": 0.4863586185144299, + "grad_norm": 1.4199777841567993, + "learning_rate": 1.0927769467568368e-05, + "loss": 0.4702, + "step": 15420 + }, + { + "epoch": 0.48639015928087054, + "grad_norm": 1.637978196144104, + "learning_rate": 1.092675231230283e-05, + "loss": 0.48, + "step": 15421 + }, + { + "epoch": 0.48642170004731117, + "grad_norm": 1.6667687892913818, + "learning_rate": 1.092573514736591e-05, + "loss": 0.4863, + "step": 15422 + }, + { + "epoch": 0.4864532408137518, + "grad_norm": 1.6425650119781494, + "learning_rate": 1.0924717972768224e-05, + "loss": 0.5, + "step": 15423 + }, + { + "epoch": 0.48648478158019237, + "grad_norm": 1.71796715259552, + "learning_rate": 1.092370078852039e-05, + "loss": 0.4202, + "step": 15424 + }, + { + "epoch": 0.486516322346633, + "grad_norm": 1.8695452213287354, + "learning_rate": 1.092268359463302e-05, + "loss": 0.4891, + "step": 15425 + }, + { + "epoch": 0.48654786311307363, + "grad_norm": 1.8645451068878174, + "learning_rate": 1.0921666391116733e-05, + "loss": 0.4829, + "step": 15426 + }, + { + "epoch": 0.48657940387951426, + "grad_norm": 1.4307770729064941, + "learning_rate": 1.0920649177982138e-05, + "loss": 0.4632, + "step": 15427 + }, + { + "epoch": 0.4866109446459549, + "grad_norm": 1.8838751316070557, + "learning_rate": 1.0919631955239857e-05, + "loss": 0.4665, + "step": 15428 + }, + { + "epoch": 0.4866424854123955, + "grad_norm": 1.3539279699325562, + "learning_rate": 1.09186147229005e-05, + "loss": 0.4758, + "step": 15429 + }, + { + "epoch": 0.48667402617883615, + "grad_norm": 1.8152762651443481, + "learning_rate": 1.0917597480974685e-05, + "loss": 0.8028, + "step": 15430 + }, + { + "epoch": 0.4867055669452768, + "grad_norm": 2.1699748039245605, + "learning_rate": 1.091658022947303e-05, + "loss": 0.4353, + "step": 15431 + }, + { + "epoch": 0.4867371077117174, + "grad_norm": 1.5546685457229614, + "learning_rate": 1.0915562968406145e-05, + "loss": 0.3714, + "step": 15432 + }, + { + "epoch": 0.48676864847815804, + "grad_norm": 1.7070506811141968, + "learning_rate": 1.0914545697784651e-05, + "loss": 0.4804, + "step": 15433 + }, + { + "epoch": 0.48680018924459867, + "grad_norm": 1.9935956001281738, + "learning_rate": 1.0913528417619166e-05, + "loss": 0.464, + "step": 15434 + }, + { + "epoch": 0.48683173001103924, + "grad_norm": 3.3770318031311035, + "learning_rate": 1.0912511127920297e-05, + "loss": 0.4931, + "step": 15435 + }, + { + "epoch": 0.4868632707774799, + "grad_norm": 2.21730375289917, + "learning_rate": 1.091149382869867e-05, + "loss": 0.4765, + "step": 15436 + }, + { + "epoch": 0.4868948115439205, + "grad_norm": 1.3313785791397095, + "learning_rate": 1.0910476519964896e-05, + "loss": 0.4704, + "step": 15437 + }, + { + "epoch": 0.48692635231036113, + "grad_norm": 1.4020298719406128, + "learning_rate": 1.0909459201729593e-05, + "loss": 0.4727, + "step": 15438 + }, + { + "epoch": 0.48695789307680176, + "grad_norm": 1.6497704982757568, + "learning_rate": 1.0908441874003377e-05, + "loss": 0.457, + "step": 15439 + }, + { + "epoch": 0.4869894338432424, + "grad_norm": 1.621785044670105, + "learning_rate": 1.0907424536796862e-05, + "loss": 0.7787, + "step": 15440 + }, + { + "epoch": 0.487020974609683, + "grad_norm": 1.7110708951950073, + "learning_rate": 1.0906407190120672e-05, + "loss": 0.5006, + "step": 15441 + }, + { + "epoch": 0.48705251537612365, + "grad_norm": 1.5243085622787476, + "learning_rate": 1.0905389833985417e-05, + "loss": 0.4654, + "step": 15442 + }, + { + "epoch": 0.4870840561425643, + "grad_norm": 1.4134494066238403, + "learning_rate": 1.0904372468401714e-05, + "loss": 0.4734, + "step": 15443 + }, + { + "epoch": 0.4871155969090049, + "grad_norm": 1.3817075490951538, + "learning_rate": 1.0903355093380182e-05, + "loss": 0.478, + "step": 15444 + }, + { + "epoch": 0.48714713767544554, + "grad_norm": 1.8267889022827148, + "learning_rate": 1.090233770893144e-05, + "loss": 0.4867, + "step": 15445 + }, + { + "epoch": 0.4871786784418861, + "grad_norm": 1.7017431259155273, + "learning_rate": 1.0901320315066104e-05, + "loss": 0.4459, + "step": 15446 + }, + { + "epoch": 0.48721021920832674, + "grad_norm": 1.7414829730987549, + "learning_rate": 1.0900302911794791e-05, + "loss": 0.4617, + "step": 15447 + }, + { + "epoch": 0.4872417599747674, + "grad_norm": 1.4034687280654907, + "learning_rate": 1.0899285499128118e-05, + "loss": 0.5016, + "step": 15448 + }, + { + "epoch": 0.487273300741208, + "grad_norm": 1.8766759634017944, + "learning_rate": 1.0898268077076701e-05, + "loss": 0.4821, + "step": 15449 + }, + { + "epoch": 0.48730484150764863, + "grad_norm": 1.5870662927627563, + "learning_rate": 1.0897250645651159e-05, + "loss": 0.4818, + "step": 15450 + }, + { + "epoch": 0.48733638227408926, + "grad_norm": 1.3884652853012085, + "learning_rate": 1.089623320486211e-05, + "loss": 0.4998, + "step": 15451 + }, + { + "epoch": 0.4873679230405299, + "grad_norm": 2.0257091522216797, + "learning_rate": 1.0895215754720172e-05, + "loss": 0.4762, + "step": 15452 + }, + { + "epoch": 0.4873994638069705, + "grad_norm": 3.5148372650146484, + "learning_rate": 1.0894198295235962e-05, + "loss": 0.3658, + "step": 15453 + }, + { + "epoch": 0.48743100457341115, + "grad_norm": 1.7260940074920654, + "learning_rate": 1.0893180826420099e-05, + "loss": 0.4648, + "step": 15454 + }, + { + "epoch": 0.4874625453398518, + "grad_norm": 1.5502638816833496, + "learning_rate": 1.0892163348283203e-05, + "loss": 0.4387, + "step": 15455 + }, + { + "epoch": 0.48749408610629236, + "grad_norm": 1.5554922819137573, + "learning_rate": 1.0891145860835886e-05, + "loss": 0.5074, + "step": 15456 + }, + { + "epoch": 0.487525626872733, + "grad_norm": 1.5467640161514282, + "learning_rate": 1.0890128364088772e-05, + "loss": 0.4508, + "step": 15457 + }, + { + "epoch": 0.4875571676391736, + "grad_norm": 1.499164342880249, + "learning_rate": 1.0889110858052476e-05, + "loss": 0.4684, + "step": 15458 + }, + { + "epoch": 0.48758870840561425, + "grad_norm": 1.5978626012802124, + "learning_rate": 1.0888093342737617e-05, + "loss": 0.502, + "step": 15459 + }, + { + "epoch": 0.4876202491720549, + "grad_norm": 1.6390262842178345, + "learning_rate": 1.0887075818154817e-05, + "loss": 0.4928, + "step": 15460 + }, + { + "epoch": 0.4876517899384955, + "grad_norm": 1.6354715824127197, + "learning_rate": 1.0886058284314689e-05, + "loss": 0.4565, + "step": 15461 + }, + { + "epoch": 0.48768333070493614, + "grad_norm": 1.5943145751953125, + "learning_rate": 1.0885040741227859e-05, + "loss": 0.4654, + "step": 15462 + }, + { + "epoch": 0.48771487147137677, + "grad_norm": 1.5297901630401611, + "learning_rate": 1.0884023188904938e-05, + "loss": 0.5235, + "step": 15463 + }, + { + "epoch": 0.4877464122378174, + "grad_norm": 3.7510621547698975, + "learning_rate": 1.088300562735655e-05, + "loss": 0.3429, + "step": 15464 + }, + { + "epoch": 0.487777953004258, + "grad_norm": 1.5342926979064941, + "learning_rate": 1.0881988056593313e-05, + "loss": 0.4685, + "step": 15465 + }, + { + "epoch": 0.48780949377069865, + "grad_norm": 1.3948663473129272, + "learning_rate": 1.0880970476625845e-05, + "loss": 0.4601, + "step": 15466 + }, + { + "epoch": 0.48784103453713923, + "grad_norm": 1.424506425857544, + "learning_rate": 1.0879952887464767e-05, + "loss": 0.4845, + "step": 15467 + }, + { + "epoch": 0.48787257530357986, + "grad_norm": 1.8916810750961304, + "learning_rate": 1.0878935289120696e-05, + "loss": 0.4377, + "step": 15468 + }, + { + "epoch": 0.4879041160700205, + "grad_norm": 2.1612398624420166, + "learning_rate": 1.0877917681604256e-05, + "loss": 0.4611, + "step": 15469 + }, + { + "epoch": 0.4879356568364611, + "grad_norm": 1.6246545314788818, + "learning_rate": 1.0876900064926059e-05, + "loss": 0.4847, + "step": 15470 + }, + { + "epoch": 0.48796719760290175, + "grad_norm": 1.578709363937378, + "learning_rate": 1.087588243909673e-05, + "loss": 0.4413, + "step": 15471 + }, + { + "epoch": 0.4879987383693424, + "grad_norm": 1.5710686445236206, + "learning_rate": 1.0874864804126888e-05, + "loss": 0.4747, + "step": 15472 + }, + { + "epoch": 0.488030279135783, + "grad_norm": 1.5668727159500122, + "learning_rate": 1.0873847160027154e-05, + "loss": 0.4729, + "step": 15473 + }, + { + "epoch": 0.48806181990222364, + "grad_norm": 1.4489333629608154, + "learning_rate": 1.0872829506808142e-05, + "loss": 0.4547, + "step": 15474 + }, + { + "epoch": 0.48809336066866427, + "grad_norm": 1.611806869506836, + "learning_rate": 1.087181184448048e-05, + "loss": 0.4769, + "step": 15475 + }, + { + "epoch": 0.4881249014351049, + "grad_norm": 2.5300724506378174, + "learning_rate": 1.0870794173054784e-05, + "loss": 0.4679, + "step": 15476 + }, + { + "epoch": 0.48815644220154547, + "grad_norm": 1.6673210859298706, + "learning_rate": 1.0869776492541678e-05, + "loss": 0.7328, + "step": 15477 + }, + { + "epoch": 0.4881879829679861, + "grad_norm": 1.5190364122390747, + "learning_rate": 1.0868758802951774e-05, + "loss": 0.4985, + "step": 15478 + }, + { + "epoch": 0.48821952373442673, + "grad_norm": 1.6051218509674072, + "learning_rate": 1.0867741104295699e-05, + "loss": 0.4633, + "step": 15479 + }, + { + "epoch": 0.48825106450086736, + "grad_norm": 1.3725529909133911, + "learning_rate": 1.0866723396584072e-05, + "loss": 0.4795, + "step": 15480 + }, + { + "epoch": 0.488282605267308, + "grad_norm": 1.606992483139038, + "learning_rate": 1.0865705679827512e-05, + "loss": 0.5008, + "step": 15481 + }, + { + "epoch": 0.4883141460337486, + "grad_norm": 1.568975567817688, + "learning_rate": 1.0864687954036641e-05, + "loss": 0.4747, + "step": 15482 + }, + { + "epoch": 0.48834568680018925, + "grad_norm": 1.5343197584152222, + "learning_rate": 1.0863670219222081e-05, + "loss": 0.5058, + "step": 15483 + }, + { + "epoch": 0.4883772275666299, + "grad_norm": 2.0882408618927, + "learning_rate": 1.086265247539445e-05, + "loss": 0.4832, + "step": 15484 + }, + { + "epoch": 0.4884087683330705, + "grad_norm": 1.6942073106765747, + "learning_rate": 1.0861634722564374e-05, + "loss": 0.4794, + "step": 15485 + }, + { + "epoch": 0.48844030909951114, + "grad_norm": 1.375304102897644, + "learning_rate": 1.086061696074247e-05, + "loss": 0.4672, + "step": 15486 + }, + { + "epoch": 0.48847184986595177, + "grad_norm": 1.450136661529541, + "learning_rate": 1.0859599189939358e-05, + "loss": 0.4538, + "step": 15487 + }, + { + "epoch": 0.48850339063239234, + "grad_norm": 1.4724448919296265, + "learning_rate": 1.085858141016566e-05, + "loss": 0.4274, + "step": 15488 + }, + { + "epoch": 0.488534931398833, + "grad_norm": 1.351948857307434, + "learning_rate": 1.0857563621432001e-05, + "loss": 0.5098, + "step": 15489 + }, + { + "epoch": 0.4885664721652736, + "grad_norm": 2.6103622913360596, + "learning_rate": 1.0856545823748999e-05, + "loss": 0.4839, + "step": 15490 + }, + { + "epoch": 0.48859801293171423, + "grad_norm": 1.2897264957427979, + "learning_rate": 1.0855528017127275e-05, + "loss": 0.4886, + "step": 15491 + }, + { + "epoch": 0.48862955369815486, + "grad_norm": 1.383768916130066, + "learning_rate": 1.0854510201577451e-05, + "loss": 0.469, + "step": 15492 + }, + { + "epoch": 0.4886610944645955, + "grad_norm": 5.1813063621521, + "learning_rate": 1.0853492377110153e-05, + "loss": 0.3871, + "step": 15493 + }, + { + "epoch": 0.4886926352310361, + "grad_norm": 1.401779055595398, + "learning_rate": 1.0852474543735996e-05, + "loss": 0.379, + "step": 15494 + }, + { + "epoch": 0.48872417599747675, + "grad_norm": 1.3759552240371704, + "learning_rate": 1.0851456701465606e-05, + "loss": 0.4555, + "step": 15495 + }, + { + "epoch": 0.4887557167639174, + "grad_norm": 1.8243459463119507, + "learning_rate": 1.0850438850309605e-05, + "loss": 0.4932, + "step": 15496 + }, + { + "epoch": 0.488787257530358, + "grad_norm": 1.3279283046722412, + "learning_rate": 1.0849420990278617e-05, + "loss": 0.4838, + "step": 15497 + }, + { + "epoch": 0.4888187982967986, + "grad_norm": 1.5240100622177124, + "learning_rate": 1.0848403121383258e-05, + "loss": 0.5111, + "step": 15498 + }, + { + "epoch": 0.4888503390632392, + "grad_norm": 1.624861478805542, + "learning_rate": 1.0847385243634153e-05, + "loss": 0.4729, + "step": 15499 + }, + { + "epoch": 0.48888187982967984, + "grad_norm": 1.4810633659362793, + "learning_rate": 1.0846367357041927e-05, + "loss": 0.4801, + "step": 15500 + }, + { + "epoch": 0.4889134205961205, + "grad_norm": 1.398497223854065, + "learning_rate": 1.0845349461617198e-05, + "loss": 0.5305, + "step": 15501 + }, + { + "epoch": 0.4889449613625611, + "grad_norm": 1.4331789016723633, + "learning_rate": 1.0844331557370594e-05, + "loss": 0.461, + "step": 15502 + }, + { + "epoch": 0.48897650212900173, + "grad_norm": 1.9065194129943848, + "learning_rate": 1.0843313644312734e-05, + "loss": 0.448, + "step": 15503 + }, + { + "epoch": 0.48900804289544236, + "grad_norm": 1.3199529647827148, + "learning_rate": 1.0842295722454239e-05, + "loss": 0.4658, + "step": 15504 + }, + { + "epoch": 0.489039583661883, + "grad_norm": 1.6120030879974365, + "learning_rate": 1.0841277791805733e-05, + "loss": 0.4695, + "step": 15505 + }, + { + "epoch": 0.4890711244283236, + "grad_norm": 1.394974708557129, + "learning_rate": 1.0840259852377845e-05, + "loss": 0.4681, + "step": 15506 + }, + { + "epoch": 0.48910266519476425, + "grad_norm": 1.858836054801941, + "learning_rate": 1.0839241904181193e-05, + "loss": 0.4815, + "step": 15507 + }, + { + "epoch": 0.4891342059612049, + "grad_norm": 1.8948521614074707, + "learning_rate": 1.0838223947226396e-05, + "loss": 0.4801, + "step": 15508 + }, + { + "epoch": 0.48916574672764546, + "grad_norm": 1.4830628633499146, + "learning_rate": 1.0837205981524082e-05, + "loss": 0.4789, + "step": 15509 + }, + { + "epoch": 0.4891972874940861, + "grad_norm": 1.3243330717086792, + "learning_rate": 1.0836188007084875e-05, + "loss": 0.4735, + "step": 15510 + }, + { + "epoch": 0.4892288282605267, + "grad_norm": 1.399343490600586, + "learning_rate": 1.0835170023919396e-05, + "loss": 0.4582, + "step": 15511 + }, + { + "epoch": 0.48926036902696735, + "grad_norm": 1.6547348499298096, + "learning_rate": 1.0834152032038268e-05, + "loss": 0.4639, + "step": 15512 + }, + { + "epoch": 0.489291909793408, + "grad_norm": 1.3680740594863892, + "learning_rate": 1.083313403145212e-05, + "loss": 0.476, + "step": 15513 + }, + { + "epoch": 0.4893234505598486, + "grad_norm": 1.6429822444915771, + "learning_rate": 1.0832116022171568e-05, + "loss": 0.5164, + "step": 15514 + }, + { + "epoch": 0.48935499132628923, + "grad_norm": 1.48210608959198, + "learning_rate": 1.083109800420724e-05, + "loss": 0.4911, + "step": 15515 + }, + { + "epoch": 0.48938653209272986, + "grad_norm": 1.4107249975204468, + "learning_rate": 1.0830079977569758e-05, + "loss": 0.5044, + "step": 15516 + }, + { + "epoch": 0.4894180728591705, + "grad_norm": 1.97319757938385, + "learning_rate": 1.082906194226975e-05, + "loss": 0.5016, + "step": 15517 + }, + { + "epoch": 0.4894496136256111, + "grad_norm": 1.860548496246338, + "learning_rate": 1.0828043898317836e-05, + "loss": 0.4167, + "step": 15518 + }, + { + "epoch": 0.48948115439205175, + "grad_norm": 1.6060210466384888, + "learning_rate": 1.0827025845724637e-05, + "loss": 0.4707, + "step": 15519 + }, + { + "epoch": 0.48951269515849233, + "grad_norm": 1.317008137702942, + "learning_rate": 1.0826007784500784e-05, + "loss": 0.4891, + "step": 15520 + }, + { + "epoch": 0.48954423592493296, + "grad_norm": 1.726109504699707, + "learning_rate": 1.0824989714656899e-05, + "loss": 0.5103, + "step": 15521 + }, + { + "epoch": 0.4895757766913736, + "grad_norm": 1.5279629230499268, + "learning_rate": 1.0823971636203603e-05, + "loss": 0.4902, + "step": 15522 + }, + { + "epoch": 0.4896073174578142, + "grad_norm": 1.3811393976211548, + "learning_rate": 1.0822953549151525e-05, + "loss": 0.4774, + "step": 15523 + }, + { + "epoch": 0.48963885822425485, + "grad_norm": 1.3663394451141357, + "learning_rate": 1.082193545351129e-05, + "loss": 0.4891, + "step": 15524 + }, + { + "epoch": 0.4896703989906955, + "grad_norm": 1.9001398086547852, + "learning_rate": 1.0820917349293517e-05, + "loss": 0.799, + "step": 15525 + }, + { + "epoch": 0.4897019397571361, + "grad_norm": 3.5846805572509766, + "learning_rate": 1.0819899236508838e-05, + "loss": 0.4889, + "step": 15526 + }, + { + "epoch": 0.48973348052357674, + "grad_norm": 1.5263195037841797, + "learning_rate": 1.0818881115167872e-05, + "loss": 0.4963, + "step": 15527 + }, + { + "epoch": 0.48976502129001737, + "grad_norm": 2.3262646198272705, + "learning_rate": 1.0817862985281243e-05, + "loss": 0.4578, + "step": 15528 + }, + { + "epoch": 0.489796562056458, + "grad_norm": 1.4422311782836914, + "learning_rate": 1.081684484685958e-05, + "loss": 0.48, + "step": 15529 + }, + { + "epoch": 0.48982810282289857, + "grad_norm": 1.6841713190078735, + "learning_rate": 1.0815826699913506e-05, + "loss": 0.482, + "step": 15530 + }, + { + "epoch": 0.4898596435893392, + "grad_norm": 1.6260449886322021, + "learning_rate": 1.081480854445365e-05, + "loss": 0.4506, + "step": 15531 + }, + { + "epoch": 0.48989118435577983, + "grad_norm": 1.7975401878356934, + "learning_rate": 1.0813790380490631e-05, + "loss": 0.4455, + "step": 15532 + }, + { + "epoch": 0.48992272512222046, + "grad_norm": 1.4749916791915894, + "learning_rate": 1.0812772208035079e-05, + "loss": 0.7273, + "step": 15533 + }, + { + "epoch": 0.4899542658886611, + "grad_norm": 1.6634232997894287, + "learning_rate": 1.0811754027097619e-05, + "loss": 0.5057, + "step": 15534 + }, + { + "epoch": 0.4899858066551017, + "grad_norm": 1.6123011112213135, + "learning_rate": 1.0810735837688873e-05, + "loss": 0.4654, + "step": 15535 + }, + { + "epoch": 0.49001734742154235, + "grad_norm": 1.53285813331604, + "learning_rate": 1.080971763981947e-05, + "loss": 0.499, + "step": 15536 + }, + { + "epoch": 0.490048888187983, + "grad_norm": 1.508917212486267, + "learning_rate": 1.0808699433500039e-05, + "loss": 0.5138, + "step": 15537 + }, + { + "epoch": 0.4900804289544236, + "grad_norm": 1.387510061264038, + "learning_rate": 1.0807681218741195e-05, + "loss": 0.477, + "step": 15538 + }, + { + "epoch": 0.49011196972086424, + "grad_norm": 1.6637707948684692, + "learning_rate": 1.0806662995553572e-05, + "loss": 0.7671, + "step": 15539 + }, + { + "epoch": 0.49014351048730487, + "grad_norm": 1.625931739807129, + "learning_rate": 1.0805644763947793e-05, + "loss": 0.4638, + "step": 15540 + }, + { + "epoch": 0.49017505125374544, + "grad_norm": 1.7639334201812744, + "learning_rate": 1.0804626523934489e-05, + "loss": 0.4605, + "step": 15541 + }, + { + "epoch": 0.49020659202018607, + "grad_norm": 1.6156963109970093, + "learning_rate": 1.080360827552428e-05, + "loss": 0.4851, + "step": 15542 + }, + { + "epoch": 0.4902381327866267, + "grad_norm": 1.6295112371444702, + "learning_rate": 1.0802590018727795e-05, + "loss": 0.5033, + "step": 15543 + }, + { + "epoch": 0.49026967355306733, + "grad_norm": 1.824752926826477, + "learning_rate": 1.0801571753555662e-05, + "loss": 0.4906, + "step": 15544 + }, + { + "epoch": 0.49030121431950796, + "grad_norm": 2.0889861583709717, + "learning_rate": 1.0800553480018504e-05, + "loss": 0.3788, + "step": 15545 + }, + { + "epoch": 0.4903327550859486, + "grad_norm": 1.3546292781829834, + "learning_rate": 1.0799535198126949e-05, + "loss": 0.4711, + "step": 15546 + }, + { + "epoch": 0.4903642958523892, + "grad_norm": 1.8443973064422607, + "learning_rate": 1.0798516907891621e-05, + "loss": 0.8114, + "step": 15547 + }, + { + "epoch": 0.49039583661882985, + "grad_norm": 1.8068100214004517, + "learning_rate": 1.0797498609323153e-05, + "loss": 0.4901, + "step": 15548 + }, + { + "epoch": 0.4904273773852705, + "grad_norm": 1.61665940284729, + "learning_rate": 1.0796480302432164e-05, + "loss": 0.4864, + "step": 15549 + }, + { + "epoch": 0.4904589181517111, + "grad_norm": 2.5807344913482666, + "learning_rate": 1.0795461987229286e-05, + "loss": 0.4921, + "step": 15550 + }, + { + "epoch": 0.4904904589181517, + "grad_norm": 1.5069392919540405, + "learning_rate": 1.0794443663725147e-05, + "loss": 0.483, + "step": 15551 + }, + { + "epoch": 0.4905219996845923, + "grad_norm": 1.5978953838348389, + "learning_rate": 1.0793425331930368e-05, + "loss": 0.4639, + "step": 15552 + }, + { + "epoch": 0.49055354045103294, + "grad_norm": 1.77261483669281, + "learning_rate": 1.079240699185558e-05, + "loss": 0.4865, + "step": 15553 + }, + { + "epoch": 0.4905850812174736, + "grad_norm": 2.6525869369506836, + "learning_rate": 1.0791388643511412e-05, + "loss": 0.485, + "step": 15554 + }, + { + "epoch": 0.4906166219839142, + "grad_norm": 1.4735749959945679, + "learning_rate": 1.0790370286908488e-05, + "loss": 0.4744, + "step": 15555 + }, + { + "epoch": 0.49064816275035483, + "grad_norm": 2.4371728897094727, + "learning_rate": 1.0789351922057437e-05, + "loss": 0.4722, + "step": 15556 + }, + { + "epoch": 0.49067970351679546, + "grad_norm": 2.0318994522094727, + "learning_rate": 1.0788333548968884e-05, + "loss": 0.5048, + "step": 15557 + }, + { + "epoch": 0.4907112442832361, + "grad_norm": 1.5520217418670654, + "learning_rate": 1.078731516765346e-05, + "loss": 0.5036, + "step": 15558 + }, + { + "epoch": 0.4907427850496767, + "grad_norm": 1.4737492799758911, + "learning_rate": 1.0786296778121787e-05, + "loss": 0.4531, + "step": 15559 + }, + { + "epoch": 0.49077432581611735, + "grad_norm": 1.6021699905395508, + "learning_rate": 1.07852783803845e-05, + "loss": 0.4967, + "step": 15560 + }, + { + "epoch": 0.490805866582558, + "grad_norm": 1.5054657459259033, + "learning_rate": 1.0784259974452224e-05, + "loss": 0.5008, + "step": 15561 + }, + { + "epoch": 0.49083740734899856, + "grad_norm": 1.9734585285186768, + "learning_rate": 1.0783241560335584e-05, + "loss": 0.468, + "step": 15562 + }, + { + "epoch": 0.4908689481154392, + "grad_norm": 1.5389865636825562, + "learning_rate": 1.0782223138045211e-05, + "loss": 0.4631, + "step": 15563 + }, + { + "epoch": 0.4909004888818798, + "grad_norm": 1.4544460773468018, + "learning_rate": 1.0781204707591734e-05, + "loss": 0.4731, + "step": 15564 + }, + { + "epoch": 0.49093202964832044, + "grad_norm": 1.6190072298049927, + "learning_rate": 1.0780186268985777e-05, + "loss": 0.4469, + "step": 15565 + }, + { + "epoch": 0.4909635704147611, + "grad_norm": 1.628637433052063, + "learning_rate": 1.0779167822237972e-05, + "loss": 0.5018, + "step": 15566 + }, + { + "epoch": 0.4909951111812017, + "grad_norm": 2.082589626312256, + "learning_rate": 1.0778149367358943e-05, + "loss": 0.4918, + "step": 15567 + }, + { + "epoch": 0.49102665194764233, + "grad_norm": 1.3923141956329346, + "learning_rate": 1.0777130904359325e-05, + "loss": 0.4924, + "step": 15568 + }, + { + "epoch": 0.49105819271408296, + "grad_norm": 1.5475080013275146, + "learning_rate": 1.077611243324974e-05, + "loss": 0.4497, + "step": 15569 + }, + { + "epoch": 0.4910897334805236, + "grad_norm": 1.4662033319473267, + "learning_rate": 1.077509395404082e-05, + "loss": 0.4865, + "step": 15570 + }, + { + "epoch": 0.4911212742469642, + "grad_norm": 1.6421130895614624, + "learning_rate": 1.0774075466743193e-05, + "loss": 0.4358, + "step": 15571 + }, + { + "epoch": 0.49115281501340485, + "grad_norm": 1.437865138053894, + "learning_rate": 1.0773056971367485e-05, + "loss": 0.7843, + "step": 15572 + }, + { + "epoch": 0.4911843557798454, + "grad_norm": 1.598174810409546, + "learning_rate": 1.0772038467924331e-05, + "loss": 0.4888, + "step": 15573 + }, + { + "epoch": 0.49121589654628606, + "grad_norm": 2.0371251106262207, + "learning_rate": 1.0771019956424355e-05, + "loss": 0.4467, + "step": 15574 + }, + { + "epoch": 0.4912474373127267, + "grad_norm": 1.7712745666503906, + "learning_rate": 1.0770001436878186e-05, + "loss": 0.3766, + "step": 15575 + }, + { + "epoch": 0.4912789780791673, + "grad_norm": 1.7121063470840454, + "learning_rate": 1.0768982909296454e-05, + "loss": 0.4517, + "step": 15576 + }, + { + "epoch": 0.49131051884560795, + "grad_norm": 1.763375163078308, + "learning_rate": 1.076796437368979e-05, + "loss": 0.4962, + "step": 15577 + }, + { + "epoch": 0.4913420596120486, + "grad_norm": 1.766832709312439, + "learning_rate": 1.0766945830068819e-05, + "loss": 0.5165, + "step": 15578 + }, + { + "epoch": 0.4913736003784892, + "grad_norm": 1.8534142971038818, + "learning_rate": 1.0765927278444175e-05, + "loss": 0.7761, + "step": 15579 + }, + { + "epoch": 0.49140514114492984, + "grad_norm": 1.344001293182373, + "learning_rate": 1.0764908718826484e-05, + "loss": 0.4671, + "step": 15580 + }, + { + "epoch": 0.49143668191137047, + "grad_norm": 1.7251639366149902, + "learning_rate": 1.0763890151226375e-05, + "loss": 0.4967, + "step": 15581 + }, + { + "epoch": 0.4914682226778111, + "grad_norm": 1.648786187171936, + "learning_rate": 1.0762871575654481e-05, + "loss": 0.4928, + "step": 15582 + }, + { + "epoch": 0.49149976344425167, + "grad_norm": 7.388923168182373, + "learning_rate": 1.0761852992121429e-05, + "loss": 0.7661, + "step": 15583 + }, + { + "epoch": 0.4915313042106923, + "grad_norm": 1.9367094039916992, + "learning_rate": 1.0760834400637851e-05, + "loss": 0.4245, + "step": 15584 + }, + { + "epoch": 0.49156284497713293, + "grad_norm": 1.5456277132034302, + "learning_rate": 1.0759815801214378e-05, + "loss": 0.5009, + "step": 15585 + }, + { + "epoch": 0.49159438574357356, + "grad_norm": 1.4699103832244873, + "learning_rate": 1.0758797193861631e-05, + "loss": 0.4694, + "step": 15586 + }, + { + "epoch": 0.4916259265100142, + "grad_norm": 1.7243866920471191, + "learning_rate": 1.0757778578590248e-05, + "loss": 0.4774, + "step": 15587 + }, + { + "epoch": 0.4916574672764548, + "grad_norm": 1.6893478631973267, + "learning_rate": 1.0756759955410856e-05, + "loss": 0.4639, + "step": 15588 + }, + { + "epoch": 0.49168900804289545, + "grad_norm": 4.436451435089111, + "learning_rate": 1.075574132433409e-05, + "loss": 0.4582, + "step": 15589 + }, + { + "epoch": 0.4917205488093361, + "grad_norm": 1.7188764810562134, + "learning_rate": 1.0754722685370574e-05, + "loss": 0.5414, + "step": 15590 + }, + { + "epoch": 0.4917520895757767, + "grad_norm": 1.3992218971252441, + "learning_rate": 1.0753704038530939e-05, + "loss": 0.4995, + "step": 15591 + }, + { + "epoch": 0.49178363034221734, + "grad_norm": 1.4365631341934204, + "learning_rate": 1.0752685383825821e-05, + "loss": 0.3454, + "step": 15592 + }, + { + "epoch": 0.49181517110865797, + "grad_norm": 1.45415461063385, + "learning_rate": 1.0751666721265845e-05, + "loss": 0.4948, + "step": 15593 + }, + { + "epoch": 0.49184671187509854, + "grad_norm": 1.6127029657363892, + "learning_rate": 1.0750648050861643e-05, + "loss": 0.783, + "step": 15594 + }, + { + "epoch": 0.49187825264153917, + "grad_norm": 1.5239208936691284, + "learning_rate": 1.0749629372623845e-05, + "loss": 0.4972, + "step": 15595 + }, + { + "epoch": 0.4919097934079798, + "grad_norm": 1.487247347831726, + "learning_rate": 1.0748610686563085e-05, + "loss": 0.4793, + "step": 15596 + }, + { + "epoch": 0.49194133417442043, + "grad_norm": 1.6709004640579224, + "learning_rate": 1.0747591992689987e-05, + "loss": 0.4592, + "step": 15597 + }, + { + "epoch": 0.49197287494086106, + "grad_norm": 1.506306767463684, + "learning_rate": 1.0746573291015187e-05, + "loss": 0.4754, + "step": 15598 + }, + { + "epoch": 0.4920044157073017, + "grad_norm": 1.584496259689331, + "learning_rate": 1.074555458154932e-05, + "loss": 0.4664, + "step": 15599 + }, + { + "epoch": 0.4920359564737423, + "grad_norm": 1.1989177465438843, + "learning_rate": 1.0744535864303008e-05, + "loss": 0.4487, + "step": 15600 + }, + { + "epoch": 0.49206749724018295, + "grad_norm": 1.6940921545028687, + "learning_rate": 1.0743517139286885e-05, + "loss": 0.4969, + "step": 15601 + }, + { + "epoch": 0.4920990380066236, + "grad_norm": 1.4529588222503662, + "learning_rate": 1.0742498406511583e-05, + "loss": 0.4858, + "step": 15602 + }, + { + "epoch": 0.4921305787730642, + "grad_norm": 1.3318769931793213, + "learning_rate": 1.0741479665987737e-05, + "loss": 0.4632, + "step": 15603 + }, + { + "epoch": 0.4921621195395048, + "grad_norm": 1.8447152376174927, + "learning_rate": 1.0740460917725974e-05, + "loss": 0.509, + "step": 15604 + }, + { + "epoch": 0.4921936603059454, + "grad_norm": 1.9370168447494507, + "learning_rate": 1.0739442161736925e-05, + "loss": 0.4479, + "step": 15605 + }, + { + "epoch": 0.49222520107238604, + "grad_norm": 1.3145534992218018, + "learning_rate": 1.0738423398031226e-05, + "loss": 0.7743, + "step": 15606 + }, + { + "epoch": 0.4922567418388267, + "grad_norm": 1.6528847217559814, + "learning_rate": 1.0737404626619503e-05, + "loss": 0.4733, + "step": 15607 + }, + { + "epoch": 0.4922882826052673, + "grad_norm": 5.565036296844482, + "learning_rate": 1.073638584751239e-05, + "loss": 0.7687, + "step": 15608 + }, + { + "epoch": 0.49231982337170793, + "grad_norm": 1.860488772392273, + "learning_rate": 1.0735367060720518e-05, + "loss": 0.4568, + "step": 15609 + }, + { + "epoch": 0.49235136413814856, + "grad_norm": 1.5635405778884888, + "learning_rate": 1.0734348266254522e-05, + "loss": 0.5049, + "step": 15610 + }, + { + "epoch": 0.4923829049045892, + "grad_norm": 1.4940457344055176, + "learning_rate": 1.073332946412503e-05, + "loss": 0.4705, + "step": 15611 + }, + { + "epoch": 0.4924144456710298, + "grad_norm": 1.4935623407363892, + "learning_rate": 1.0732310654342675e-05, + "loss": 0.4631, + "step": 15612 + }, + { + "epoch": 0.49244598643747045, + "grad_norm": 1.402535080909729, + "learning_rate": 1.0731291836918092e-05, + "loss": 0.4999, + "step": 15613 + }, + { + "epoch": 0.4924775272039111, + "grad_norm": 1.463566780090332, + "learning_rate": 1.073027301186191e-05, + "loss": 0.4878, + "step": 15614 + }, + { + "epoch": 0.49250906797035165, + "grad_norm": 1.4515423774719238, + "learning_rate": 1.072925417918476e-05, + "loss": 0.468, + "step": 15615 + }, + { + "epoch": 0.4925406087367923, + "grad_norm": 1.5751738548278809, + "learning_rate": 1.0728235338897282e-05, + "loss": 0.4473, + "step": 15616 + }, + { + "epoch": 0.4925721495032329, + "grad_norm": 1.4005135297775269, + "learning_rate": 1.0727216491010097e-05, + "loss": 0.5046, + "step": 15617 + }, + { + "epoch": 0.49260369026967354, + "grad_norm": 1.6259382963180542, + "learning_rate": 1.0726197635533844e-05, + "loss": 0.509, + "step": 15618 + }, + { + "epoch": 0.4926352310361142, + "grad_norm": 1.8910795450210571, + "learning_rate": 1.0725178772479154e-05, + "loss": 0.4807, + "step": 15619 + }, + { + "epoch": 0.4926667718025548, + "grad_norm": 1.4658619165420532, + "learning_rate": 1.0724159901856666e-05, + "loss": 0.429, + "step": 15620 + }, + { + "epoch": 0.49269831256899543, + "grad_norm": 1.6669667959213257, + "learning_rate": 1.0723141023677001e-05, + "loss": 0.4501, + "step": 15621 + }, + { + "epoch": 0.49272985333543606, + "grad_norm": 1.4929792881011963, + "learning_rate": 1.0722122137950798e-05, + "loss": 0.4592, + "step": 15622 + }, + { + "epoch": 0.4927613941018767, + "grad_norm": 2.0994584560394287, + "learning_rate": 1.0721103244688695e-05, + "loss": 0.5019, + "step": 15623 + }, + { + "epoch": 0.4927929348683173, + "grad_norm": 1.3470463752746582, + "learning_rate": 1.0720084343901318e-05, + "loss": 0.4581, + "step": 15624 + }, + { + "epoch": 0.4928244756347579, + "grad_norm": 1.6622179746627808, + "learning_rate": 1.07190654355993e-05, + "loss": 0.4653, + "step": 15625 + }, + { + "epoch": 0.4928560164011985, + "grad_norm": 1.4369763135910034, + "learning_rate": 1.0718046519793276e-05, + "loss": 0.4946, + "step": 15626 + }, + { + "epoch": 0.49288755716763916, + "grad_norm": 1.5349041223526, + "learning_rate": 1.0717027596493881e-05, + "loss": 0.4365, + "step": 15627 + }, + { + "epoch": 0.4929190979340798, + "grad_norm": 1.5764468908309937, + "learning_rate": 1.0716008665711744e-05, + "loss": 0.4837, + "step": 15628 + }, + { + "epoch": 0.4929506387005204, + "grad_norm": 1.6274582147598267, + "learning_rate": 1.07149897274575e-05, + "loss": 0.4786, + "step": 15629 + }, + { + "epoch": 0.49298217946696105, + "grad_norm": 2.358708381652832, + "learning_rate": 1.0713970781741787e-05, + "loss": 0.3944, + "step": 15630 + }, + { + "epoch": 0.4930137202334017, + "grad_norm": 1.7086946964263916, + "learning_rate": 1.071295182857523e-05, + "loss": 0.4997, + "step": 15631 + }, + { + "epoch": 0.4930452609998423, + "grad_norm": 1.5980433225631714, + "learning_rate": 1.0711932867968468e-05, + "loss": 0.4881, + "step": 15632 + }, + { + "epoch": 0.49307680176628294, + "grad_norm": 1.3503073453903198, + "learning_rate": 1.0710913899932137e-05, + "loss": 0.448, + "step": 15633 + }, + { + "epoch": 0.49310834253272356, + "grad_norm": 1.622698426246643, + "learning_rate": 1.0709894924476867e-05, + "loss": 0.4712, + "step": 15634 + }, + { + "epoch": 0.4931398832991642, + "grad_norm": 2.73427677154541, + "learning_rate": 1.0708875941613288e-05, + "loss": 0.4832, + "step": 15635 + }, + { + "epoch": 0.49317142406560477, + "grad_norm": 1.5624279975891113, + "learning_rate": 1.070785695135204e-05, + "loss": 0.4239, + "step": 15636 + }, + { + "epoch": 0.4932029648320454, + "grad_norm": 1.2151349782943726, + "learning_rate": 1.0706837953703757e-05, + "loss": 0.4614, + "step": 15637 + }, + { + "epoch": 0.49323450559848603, + "grad_norm": 1.446317195892334, + "learning_rate": 1.070581894867907e-05, + "loss": 0.4678, + "step": 15638 + }, + { + "epoch": 0.49326604636492666, + "grad_norm": 1.4003605842590332, + "learning_rate": 1.0704799936288612e-05, + "loss": 0.456, + "step": 15639 + }, + { + "epoch": 0.4932975871313673, + "grad_norm": 1.366976022720337, + "learning_rate": 1.0703780916543023e-05, + "loss": 0.4621, + "step": 15640 + }, + { + "epoch": 0.4933291278978079, + "grad_norm": 1.4504108428955078, + "learning_rate": 1.070276188945293e-05, + "loss": 0.4488, + "step": 15641 + }, + { + "epoch": 0.49336066866424855, + "grad_norm": 2.77303409576416, + "learning_rate": 1.0701742855028973e-05, + "loss": 0.5093, + "step": 15642 + }, + { + "epoch": 0.4933922094306892, + "grad_norm": 1.5879206657409668, + "learning_rate": 1.0700723813281786e-05, + "loss": 0.4936, + "step": 15643 + }, + { + "epoch": 0.4934237501971298, + "grad_norm": 1.6192042827606201, + "learning_rate": 1.0699704764222002e-05, + "loss": 0.4911, + "step": 15644 + }, + { + "epoch": 0.49345529096357044, + "grad_norm": 1.7723444700241089, + "learning_rate": 1.0698685707860252e-05, + "loss": 0.4219, + "step": 15645 + }, + { + "epoch": 0.49348683173001107, + "grad_norm": 2.26419734954834, + "learning_rate": 1.0697666644207176e-05, + "loss": 0.4854, + "step": 15646 + }, + { + "epoch": 0.49351837249645164, + "grad_norm": 1.3601154088974, + "learning_rate": 1.0696647573273407e-05, + "loss": 0.454, + "step": 15647 + }, + { + "epoch": 0.49354991326289227, + "grad_norm": 1.5406265258789062, + "learning_rate": 1.0695628495069578e-05, + "loss": 0.4778, + "step": 15648 + }, + { + "epoch": 0.4935814540293329, + "grad_norm": 1.8431954383850098, + "learning_rate": 1.0694609409606326e-05, + "loss": 0.5238, + "step": 15649 + }, + { + "epoch": 0.49361299479577353, + "grad_norm": 1.3055928945541382, + "learning_rate": 1.0693590316894287e-05, + "loss": 0.3907, + "step": 15650 + }, + { + "epoch": 0.49364453556221416, + "grad_norm": 1.877949595451355, + "learning_rate": 1.0692571216944095e-05, + "loss": 0.7933, + "step": 15651 + }, + { + "epoch": 0.4936760763286548, + "grad_norm": 1.5928038358688354, + "learning_rate": 1.0691552109766383e-05, + "loss": 0.4773, + "step": 15652 + }, + { + "epoch": 0.4937076170950954, + "grad_norm": 1.5766682624816895, + "learning_rate": 1.0690532995371789e-05, + "loss": 0.3849, + "step": 15653 + }, + { + "epoch": 0.49373915786153605, + "grad_norm": 1.4958736896514893, + "learning_rate": 1.0689513873770947e-05, + "loss": 0.4755, + "step": 15654 + }, + { + "epoch": 0.4937706986279767, + "grad_norm": 1.5742969512939453, + "learning_rate": 1.0688494744974491e-05, + "loss": 0.478, + "step": 15655 + }, + { + "epoch": 0.4938022393944173, + "grad_norm": 1.4870431423187256, + "learning_rate": 1.0687475608993059e-05, + "loss": 0.4609, + "step": 15656 + }, + { + "epoch": 0.4938337801608579, + "grad_norm": 1.7261731624603271, + "learning_rate": 1.0686456465837282e-05, + "loss": 0.458, + "step": 15657 + }, + { + "epoch": 0.4938653209272985, + "grad_norm": 2.700704574584961, + "learning_rate": 1.0685437315517803e-05, + "loss": 0.7927, + "step": 15658 + }, + { + "epoch": 0.49389686169373914, + "grad_norm": 2.1103498935699463, + "learning_rate": 1.0684418158045252e-05, + "loss": 0.3849, + "step": 15659 + }, + { + "epoch": 0.49392840246017977, + "grad_norm": 1.5865843296051025, + "learning_rate": 1.0683398993430262e-05, + "loss": 0.3739, + "step": 15660 + }, + { + "epoch": 0.4939599432266204, + "grad_norm": 1.6977790594100952, + "learning_rate": 1.068237982168348e-05, + "loss": 0.4508, + "step": 15661 + }, + { + "epoch": 0.49399148399306103, + "grad_norm": 1.7016212940216064, + "learning_rate": 1.0681360642815531e-05, + "loss": 0.3627, + "step": 15662 + }, + { + "epoch": 0.49402302475950166, + "grad_norm": 1.8755369186401367, + "learning_rate": 1.0680341456837056e-05, + "loss": 0.4821, + "step": 15663 + }, + { + "epoch": 0.4940545655259423, + "grad_norm": 1.7255483865737915, + "learning_rate": 1.0679322263758689e-05, + "loss": 0.4757, + "step": 15664 + }, + { + "epoch": 0.4940861062923829, + "grad_norm": 1.5619028806686401, + "learning_rate": 1.0678303063591065e-05, + "loss": 0.497, + "step": 15665 + }, + { + "epoch": 0.49411764705882355, + "grad_norm": 1.4681954383850098, + "learning_rate": 1.0677283856344823e-05, + "loss": 0.4795, + "step": 15666 + }, + { + "epoch": 0.4941491878252642, + "grad_norm": 1.617653489112854, + "learning_rate": 1.0676264642030596e-05, + "loss": 0.4373, + "step": 15667 + }, + { + "epoch": 0.49418072859170475, + "grad_norm": 3.015596866607666, + "learning_rate": 1.0675245420659025e-05, + "loss": 0.4932, + "step": 15668 + }, + { + "epoch": 0.4942122693581454, + "grad_norm": 1.7068082094192505, + "learning_rate": 1.0674226192240744e-05, + "loss": 0.5052, + "step": 15669 + }, + { + "epoch": 0.494243810124586, + "grad_norm": 1.9182748794555664, + "learning_rate": 1.0673206956786389e-05, + "loss": 0.4992, + "step": 15670 + }, + { + "epoch": 0.49427535089102664, + "grad_norm": 1.5980757474899292, + "learning_rate": 1.0672187714306596e-05, + "loss": 0.4504, + "step": 15671 + }, + { + "epoch": 0.4943068916574673, + "grad_norm": 1.7086784839630127, + "learning_rate": 1.0671168464812001e-05, + "loss": 0.4835, + "step": 15672 + }, + { + "epoch": 0.4943384324239079, + "grad_norm": 1.5348104238510132, + "learning_rate": 1.0670149208313246e-05, + "loss": 0.4463, + "step": 15673 + }, + { + "epoch": 0.49436997319034853, + "grad_norm": 5.984311103820801, + "learning_rate": 1.066912994482096e-05, + "loss": 0.44, + "step": 15674 + }, + { + "epoch": 0.49440151395678916, + "grad_norm": 1.594435453414917, + "learning_rate": 1.0668110674345783e-05, + "loss": 0.4678, + "step": 15675 + }, + { + "epoch": 0.4944330547232298, + "grad_norm": 1.7821671962738037, + "learning_rate": 1.0667091396898355e-05, + "loss": 0.468, + "step": 15676 + }, + { + "epoch": 0.4944645954896704, + "grad_norm": 1.6578963994979858, + "learning_rate": 1.0666072112489307e-05, + "loss": 0.4679, + "step": 15677 + }, + { + "epoch": 0.494496136256111, + "grad_norm": 1.4484647512435913, + "learning_rate": 1.0665052821129283e-05, + "loss": 0.4702, + "step": 15678 + }, + { + "epoch": 0.4945276770225516, + "grad_norm": 1.6174602508544922, + "learning_rate": 1.0664033522828913e-05, + "loss": 0.5136, + "step": 15679 + }, + { + "epoch": 0.49455921778899226, + "grad_norm": 1.4278290271759033, + "learning_rate": 1.066301421759884e-05, + "loss": 0.4553, + "step": 15680 + }, + { + "epoch": 0.4945907585554329, + "grad_norm": 1.8387638330459595, + "learning_rate": 1.0661994905449697e-05, + "loss": 0.4582, + "step": 15681 + }, + { + "epoch": 0.4946222993218735, + "grad_norm": 1.5463815927505493, + "learning_rate": 1.0660975586392126e-05, + "loss": 0.3832, + "step": 15682 + }, + { + "epoch": 0.49465384008831415, + "grad_norm": 1.3904414176940918, + "learning_rate": 1.065995626043676e-05, + "loss": 0.488, + "step": 15683 + }, + { + "epoch": 0.4946853808547548, + "grad_norm": 1.702117919921875, + "learning_rate": 1.0658936927594236e-05, + "loss": 0.4558, + "step": 15684 + }, + { + "epoch": 0.4947169216211954, + "grad_norm": 1.839613437652588, + "learning_rate": 1.0657917587875197e-05, + "loss": 0.419, + "step": 15685 + }, + { + "epoch": 0.49474846238763603, + "grad_norm": 1.412064790725708, + "learning_rate": 1.0656898241290277e-05, + "loss": 0.4816, + "step": 15686 + }, + { + "epoch": 0.49478000315407666, + "grad_norm": 1.303202509880066, + "learning_rate": 1.065587888785011e-05, + "loss": 0.3382, + "step": 15687 + }, + { + "epoch": 0.4948115439205173, + "grad_norm": 1.5484760999679565, + "learning_rate": 1.065485952756534e-05, + "loss": 0.7771, + "step": 15688 + }, + { + "epoch": 0.49484308468695787, + "grad_norm": 1.8993585109710693, + "learning_rate": 1.0653840160446603e-05, + "loss": 0.5053, + "step": 15689 + }, + { + "epoch": 0.4948746254533985, + "grad_norm": 1.7057030200958252, + "learning_rate": 1.0652820786504534e-05, + "loss": 0.4639, + "step": 15690 + }, + { + "epoch": 0.4949061662198391, + "grad_norm": 1.5543767213821411, + "learning_rate": 1.0651801405749773e-05, + "loss": 0.4988, + "step": 15691 + }, + { + "epoch": 0.49493770698627976, + "grad_norm": 1.6589244604110718, + "learning_rate": 1.0650782018192962e-05, + "loss": 0.473, + "step": 15692 + }, + { + "epoch": 0.4949692477527204, + "grad_norm": 1.3076542615890503, + "learning_rate": 1.0649762623844733e-05, + "loss": 0.4886, + "step": 15693 + }, + { + "epoch": 0.495000788519161, + "grad_norm": 3.0684690475463867, + "learning_rate": 1.0648743222715726e-05, + "loss": 0.7588, + "step": 15694 + }, + { + "epoch": 0.49503232928560165, + "grad_norm": 1.8354851007461548, + "learning_rate": 1.0647723814816582e-05, + "loss": 0.4643, + "step": 15695 + }, + { + "epoch": 0.4950638700520423, + "grad_norm": 1.6895458698272705, + "learning_rate": 1.0646704400157934e-05, + "loss": 0.4488, + "step": 15696 + }, + { + "epoch": 0.4950954108184829, + "grad_norm": 1.6165575981140137, + "learning_rate": 1.0645684978750423e-05, + "loss": 0.492, + "step": 15697 + }, + { + "epoch": 0.49512695158492354, + "grad_norm": 1.518723964691162, + "learning_rate": 1.064466555060469e-05, + "loss": 0.4373, + "step": 15698 + }, + { + "epoch": 0.49515849235136417, + "grad_norm": 1.9060555696487427, + "learning_rate": 1.0643646115731372e-05, + "loss": 0.4039, + "step": 15699 + }, + { + "epoch": 0.49519003311780474, + "grad_norm": 1.4854457378387451, + "learning_rate": 1.0642626674141105e-05, + "loss": 0.7549, + "step": 15700 + }, + { + "epoch": 0.49522157388424537, + "grad_norm": 1.846034049987793, + "learning_rate": 1.0641607225844531e-05, + "loss": 0.4959, + "step": 15701 + }, + { + "epoch": 0.495253114650686, + "grad_norm": 1.8709032535552979, + "learning_rate": 1.0640587770852287e-05, + "loss": 0.4821, + "step": 15702 + }, + { + "epoch": 0.49528465541712663, + "grad_norm": 1.6246331930160522, + "learning_rate": 1.0639568309175015e-05, + "loss": 0.5065, + "step": 15703 + }, + { + "epoch": 0.49531619618356726, + "grad_norm": 1.7913106679916382, + "learning_rate": 1.0638548840823347e-05, + "loss": 0.4283, + "step": 15704 + }, + { + "epoch": 0.4953477369500079, + "grad_norm": 1.690739631652832, + "learning_rate": 1.0637529365807926e-05, + "loss": 0.4607, + "step": 15705 + }, + { + "epoch": 0.4953792777164485, + "grad_norm": 2.046804189682007, + "learning_rate": 1.0636509884139394e-05, + "loss": 0.4723, + "step": 15706 + }, + { + "epoch": 0.49541081848288915, + "grad_norm": 2.3655474185943604, + "learning_rate": 1.0635490395828386e-05, + "loss": 0.3693, + "step": 15707 + }, + { + "epoch": 0.4954423592493298, + "grad_norm": 1.5365550518035889, + "learning_rate": 1.0634470900885538e-05, + "loss": 0.461, + "step": 15708 + }, + { + "epoch": 0.4954739000157704, + "grad_norm": 1.5584022998809814, + "learning_rate": 1.06334513993215e-05, + "loss": 0.4669, + "step": 15709 + }, + { + "epoch": 0.495505440782211, + "grad_norm": 1.4196590185165405, + "learning_rate": 1.06324318911469e-05, + "loss": 0.3925, + "step": 15710 + }, + { + "epoch": 0.4955369815486516, + "grad_norm": 5.201131820678711, + "learning_rate": 1.0631412376372387e-05, + "loss": 0.479, + "step": 15711 + }, + { + "epoch": 0.49556852231509224, + "grad_norm": 1.4862725734710693, + "learning_rate": 1.0630392855008591e-05, + "loss": 0.4687, + "step": 15712 + }, + { + "epoch": 0.49560006308153287, + "grad_norm": 1.977390170097351, + "learning_rate": 1.062937332706616e-05, + "loss": 0.5192, + "step": 15713 + }, + { + "epoch": 0.4956316038479735, + "grad_norm": 1.4718140363693237, + "learning_rate": 1.0628353792555726e-05, + "loss": 0.4679, + "step": 15714 + }, + { + "epoch": 0.49566314461441413, + "grad_norm": 1.6198128461837769, + "learning_rate": 1.0627334251487933e-05, + "loss": 0.4381, + "step": 15715 + }, + { + "epoch": 0.49569468538085476, + "grad_norm": 1.2534815073013306, + "learning_rate": 1.0626314703873422e-05, + "loss": 0.5001, + "step": 15716 + }, + { + "epoch": 0.4957262261472954, + "grad_norm": 1.7336416244506836, + "learning_rate": 1.0625295149722829e-05, + "loss": 0.7445, + "step": 15717 + }, + { + "epoch": 0.495757766913736, + "grad_norm": 2.1541340351104736, + "learning_rate": 1.0624275589046796e-05, + "loss": 0.4836, + "step": 15718 + }, + { + "epoch": 0.49578930768017665, + "grad_norm": 1.5959382057189941, + "learning_rate": 1.0623256021855964e-05, + "loss": 0.4502, + "step": 15719 + }, + { + "epoch": 0.4958208484466173, + "grad_norm": 1.5690504312515259, + "learning_rate": 1.062223644816097e-05, + "loss": 0.4436, + "step": 15720 + }, + { + "epoch": 0.49585238921305785, + "grad_norm": 1.2606531381607056, + "learning_rate": 1.0621216867972455e-05, + "loss": 0.452, + "step": 15721 + }, + { + "epoch": 0.4958839299794985, + "grad_norm": 1.440589427947998, + "learning_rate": 1.0620197281301059e-05, + "loss": 0.4792, + "step": 15722 + }, + { + "epoch": 0.4959154707459391, + "grad_norm": 1.6430145502090454, + "learning_rate": 1.0619177688157427e-05, + "loss": 0.4727, + "step": 15723 + }, + { + "epoch": 0.49594701151237974, + "grad_norm": 1.369887113571167, + "learning_rate": 1.061815808855219e-05, + "loss": 0.4991, + "step": 15724 + }, + { + "epoch": 0.4959785522788204, + "grad_norm": 1.501129388809204, + "learning_rate": 1.0617138482495994e-05, + "loss": 0.4848, + "step": 15725 + }, + { + "epoch": 0.496010093045261, + "grad_norm": 1.7333869934082031, + "learning_rate": 1.0616118869999484e-05, + "loss": 0.5036, + "step": 15726 + }, + { + "epoch": 0.49604163381170163, + "grad_norm": 1.6655950546264648, + "learning_rate": 1.0615099251073288e-05, + "loss": 0.4718, + "step": 15727 + }, + { + "epoch": 0.49607317457814226, + "grad_norm": 1.6849863529205322, + "learning_rate": 1.0614079625728058e-05, + "loss": 0.4462, + "step": 15728 + }, + { + "epoch": 0.4961047153445829, + "grad_norm": 1.4833992719650269, + "learning_rate": 1.0613059993974429e-05, + "loss": 0.3669, + "step": 15729 + }, + { + "epoch": 0.4961362561110235, + "grad_norm": 1.683687448501587, + "learning_rate": 1.0612040355823043e-05, + "loss": 0.3662, + "step": 15730 + }, + { + "epoch": 0.4961677968774641, + "grad_norm": 1.6515045166015625, + "learning_rate": 1.0611020711284544e-05, + "loss": 0.4911, + "step": 15731 + }, + { + "epoch": 0.4961993376439047, + "grad_norm": 1.243377447128296, + "learning_rate": 1.0610001060369564e-05, + "loss": 0.3808, + "step": 15732 + }, + { + "epoch": 0.49623087841034536, + "grad_norm": 1.7207708358764648, + "learning_rate": 1.0608981403088752e-05, + "loss": 0.4898, + "step": 15733 + }, + { + "epoch": 0.496262419176786, + "grad_norm": 1.6279653310775757, + "learning_rate": 1.0607961739452745e-05, + "loss": 0.7772, + "step": 15734 + }, + { + "epoch": 0.4962939599432266, + "grad_norm": 1.64665687084198, + "learning_rate": 1.0606942069472185e-05, + "loss": 0.4642, + "step": 15735 + }, + { + "epoch": 0.49632550070966724, + "grad_norm": 1.758122205734253, + "learning_rate": 1.0605922393157716e-05, + "loss": 0.8068, + "step": 15736 + }, + { + "epoch": 0.4963570414761079, + "grad_norm": 1.488478660583496, + "learning_rate": 1.0604902710519974e-05, + "loss": 0.4566, + "step": 15737 + }, + { + "epoch": 0.4963885822425485, + "grad_norm": 1.5197875499725342, + "learning_rate": 1.0603883021569601e-05, + "loss": 0.4958, + "step": 15738 + }, + { + "epoch": 0.49642012300898913, + "grad_norm": 1.5316486358642578, + "learning_rate": 1.060286332631724e-05, + "loss": 0.4899, + "step": 15739 + }, + { + "epoch": 0.49645166377542976, + "grad_norm": 3.561614513397217, + "learning_rate": 1.0601843624773538e-05, + "loss": 0.3788, + "step": 15740 + }, + { + "epoch": 0.4964832045418704, + "grad_norm": 1.2396897077560425, + "learning_rate": 1.0600823916949129e-05, + "loss": 0.4725, + "step": 15741 + }, + { + "epoch": 0.49651474530831097, + "grad_norm": 1.3602797985076904, + "learning_rate": 1.059980420285465e-05, + "loss": 0.4779, + "step": 15742 + }, + { + "epoch": 0.4965462860747516, + "grad_norm": 1.612330436706543, + "learning_rate": 1.0598784482500752e-05, + "loss": 0.4946, + "step": 15743 + }, + { + "epoch": 0.4965778268411922, + "grad_norm": 1.6280863285064697, + "learning_rate": 1.0597764755898072e-05, + "loss": 0.4601, + "step": 15744 + }, + { + "epoch": 0.49660936760763286, + "grad_norm": 1.7538048028945923, + "learning_rate": 1.0596745023057253e-05, + "loss": 0.4991, + "step": 15745 + }, + { + "epoch": 0.4966409083740735, + "grad_norm": 1.572892427444458, + "learning_rate": 1.0595725283988935e-05, + "loss": 0.5073, + "step": 15746 + }, + { + "epoch": 0.4966724491405141, + "grad_norm": 3.0582611560821533, + "learning_rate": 1.0594705538703765e-05, + "loss": 0.4584, + "step": 15747 + }, + { + "epoch": 0.49670398990695475, + "grad_norm": 1.3859814405441284, + "learning_rate": 1.0593685787212378e-05, + "loss": 0.5023, + "step": 15748 + }, + { + "epoch": 0.4967355306733954, + "grad_norm": 1.556138038635254, + "learning_rate": 1.0592666029525418e-05, + "loss": 0.8001, + "step": 15749 + }, + { + "epoch": 0.496767071439836, + "grad_norm": 1.3173772096633911, + "learning_rate": 1.059164626565353e-05, + "loss": 0.4894, + "step": 15750 + }, + { + "epoch": 0.49679861220627664, + "grad_norm": 1.685757040977478, + "learning_rate": 1.0590626495607354e-05, + "loss": 0.4887, + "step": 15751 + }, + { + "epoch": 0.4968301529727172, + "grad_norm": 1.8467369079589844, + "learning_rate": 1.0589606719397531e-05, + "loss": 0.4613, + "step": 15752 + }, + { + "epoch": 0.49686169373915784, + "grad_norm": 1.639159083366394, + "learning_rate": 1.0588586937034703e-05, + "loss": 0.4823, + "step": 15753 + }, + { + "epoch": 0.49689323450559847, + "grad_norm": 1.2921051979064941, + "learning_rate": 1.0587567148529516e-05, + "loss": 0.7724, + "step": 15754 + }, + { + "epoch": 0.4969247752720391, + "grad_norm": 1.483038067817688, + "learning_rate": 1.0586547353892606e-05, + "loss": 0.4742, + "step": 15755 + }, + { + "epoch": 0.49695631603847973, + "grad_norm": 1.289312481880188, + "learning_rate": 1.058552755313462e-05, + "loss": 0.4673, + "step": 15756 + }, + { + "epoch": 0.49698785680492036, + "grad_norm": 1.9581676721572876, + "learning_rate": 1.0584507746266204e-05, + "loss": 0.5049, + "step": 15757 + }, + { + "epoch": 0.497019397571361, + "grad_norm": 1.2977920770645142, + "learning_rate": 1.0583487933297988e-05, + "loss": 0.5127, + "step": 15758 + }, + { + "epoch": 0.4970509383378016, + "grad_norm": 1.628257393836975, + "learning_rate": 1.0582468114240628e-05, + "loss": 0.4353, + "step": 15759 + }, + { + "epoch": 0.49708247910424225, + "grad_norm": 1.4127970933914185, + "learning_rate": 1.0581448289104759e-05, + "loss": 0.4825, + "step": 15760 + }, + { + "epoch": 0.4971140198706829, + "grad_norm": 1.1438663005828857, + "learning_rate": 1.0580428457901028e-05, + "loss": 0.3883, + "step": 15761 + }, + { + "epoch": 0.4971455606371235, + "grad_norm": 1.5030487775802612, + "learning_rate": 1.0579408620640073e-05, + "loss": 0.4677, + "step": 15762 + }, + { + "epoch": 0.4971771014035641, + "grad_norm": 1.310878038406372, + "learning_rate": 1.0578388777332538e-05, + "loss": 0.4906, + "step": 15763 + }, + { + "epoch": 0.4972086421700047, + "grad_norm": 1.7005959749221802, + "learning_rate": 1.0577368927989069e-05, + "loss": 0.4616, + "step": 15764 + }, + { + "epoch": 0.49724018293644534, + "grad_norm": 1.6601877212524414, + "learning_rate": 1.0576349072620307e-05, + "loss": 0.4934, + "step": 15765 + }, + { + "epoch": 0.49727172370288597, + "grad_norm": 1.5792251825332642, + "learning_rate": 1.0575329211236893e-05, + "loss": 0.7916, + "step": 15766 + }, + { + "epoch": 0.4973032644693266, + "grad_norm": 1.835414171218872, + "learning_rate": 1.0574309343849473e-05, + "loss": 0.4617, + "step": 15767 + }, + { + "epoch": 0.49733480523576723, + "grad_norm": 1.4735572338104248, + "learning_rate": 1.0573289470468688e-05, + "loss": 0.4369, + "step": 15768 + }, + { + "epoch": 0.49736634600220786, + "grad_norm": 1.5103811025619507, + "learning_rate": 1.0572269591105184e-05, + "loss": 0.4592, + "step": 15769 + }, + { + "epoch": 0.4973978867686485, + "grad_norm": 1.8639006614685059, + "learning_rate": 1.0571249705769602e-05, + "loss": 0.4612, + "step": 15770 + }, + { + "epoch": 0.4974294275350891, + "grad_norm": 1.5461821556091309, + "learning_rate": 1.0570229814472588e-05, + "loss": 0.5047, + "step": 15771 + }, + { + "epoch": 0.49746096830152975, + "grad_norm": 1.9398539066314697, + "learning_rate": 1.0569209917224779e-05, + "loss": 0.4807, + "step": 15772 + }, + { + "epoch": 0.4974925090679704, + "grad_norm": 1.3536285161972046, + "learning_rate": 1.0568190014036823e-05, + "loss": 0.7633, + "step": 15773 + }, + { + "epoch": 0.49752404983441095, + "grad_norm": 1.5482481718063354, + "learning_rate": 1.0567170104919366e-05, + "loss": 0.4357, + "step": 15774 + }, + { + "epoch": 0.4975555906008516, + "grad_norm": 1.6539078950881958, + "learning_rate": 1.0566150189883047e-05, + "loss": 0.5257, + "step": 15775 + }, + { + "epoch": 0.4975871313672922, + "grad_norm": 1.68820321559906, + "learning_rate": 1.056513026893851e-05, + "loss": 0.4736, + "step": 15776 + }, + { + "epoch": 0.49761867213373284, + "grad_norm": 2.9110913276672363, + "learning_rate": 1.05641103420964e-05, + "loss": 0.3534, + "step": 15777 + }, + { + "epoch": 0.49765021290017347, + "grad_norm": 1.514617681503296, + "learning_rate": 1.0563090409367362e-05, + "loss": 0.4625, + "step": 15778 + }, + { + "epoch": 0.4976817536666141, + "grad_norm": 1.6706243753433228, + "learning_rate": 1.0562070470762038e-05, + "loss": 0.5131, + "step": 15779 + }, + { + "epoch": 0.49771329443305473, + "grad_norm": 1.4631016254425049, + "learning_rate": 1.0561050526291072e-05, + "loss": 0.4329, + "step": 15780 + }, + { + "epoch": 0.49774483519949536, + "grad_norm": 1.4811506271362305, + "learning_rate": 1.056003057596511e-05, + "loss": 0.7992, + "step": 15781 + }, + { + "epoch": 0.497776375965936, + "grad_norm": 1.9855210781097412, + "learning_rate": 1.0559010619794791e-05, + "loss": 0.3769, + "step": 15782 + }, + { + "epoch": 0.4978079167323766, + "grad_norm": 1.6138381958007812, + "learning_rate": 1.0557990657790763e-05, + "loss": 0.5073, + "step": 15783 + }, + { + "epoch": 0.4978394574988172, + "grad_norm": 2.8587687015533447, + "learning_rate": 1.055697068996367e-05, + "loss": 0.4819, + "step": 15784 + }, + { + "epoch": 0.4978709982652578, + "grad_norm": 2.181161880493164, + "learning_rate": 1.0555950716324154e-05, + "loss": 0.4298, + "step": 15785 + }, + { + "epoch": 0.49790253903169845, + "grad_norm": 1.642507553100586, + "learning_rate": 1.0554930736882862e-05, + "loss": 0.4919, + "step": 15786 + }, + { + "epoch": 0.4979340797981391, + "grad_norm": 1.647115707397461, + "learning_rate": 1.0553910751650435e-05, + "loss": 0.4754, + "step": 15787 + }, + { + "epoch": 0.4979656205645797, + "grad_norm": 2.1892592906951904, + "learning_rate": 1.0552890760637522e-05, + "loss": 0.4747, + "step": 15788 + }, + { + "epoch": 0.49799716133102034, + "grad_norm": 1.415602684020996, + "learning_rate": 1.0551870763854762e-05, + "loss": 0.4696, + "step": 15789 + }, + { + "epoch": 0.498028702097461, + "grad_norm": 1.3962767124176025, + "learning_rate": 1.0550850761312806e-05, + "loss": 0.7777, + "step": 15790 + }, + { + "epoch": 0.4980602428639016, + "grad_norm": 28.100337982177734, + "learning_rate": 1.0549830753022292e-05, + "loss": 0.4878, + "step": 15791 + }, + { + "epoch": 0.49809178363034223, + "grad_norm": 1.3681960105895996, + "learning_rate": 1.0548810738993866e-05, + "loss": 0.4779, + "step": 15792 + }, + { + "epoch": 0.49812332439678286, + "grad_norm": 1.4931234121322632, + "learning_rate": 1.0547790719238173e-05, + "loss": 0.4984, + "step": 15793 + }, + { + "epoch": 0.4981548651632235, + "grad_norm": 1.5785828828811646, + "learning_rate": 1.0546770693765859e-05, + "loss": 0.5105, + "step": 15794 + }, + { + "epoch": 0.49818640592966407, + "grad_norm": 1.3530999422073364, + "learning_rate": 1.0545750662587572e-05, + "loss": 0.7857, + "step": 15795 + }, + { + "epoch": 0.4982179466961047, + "grad_norm": 1.6521131992340088, + "learning_rate": 1.0544730625713951e-05, + "loss": 0.4414, + "step": 15796 + }, + { + "epoch": 0.4982494874625453, + "grad_norm": 1.7277406454086304, + "learning_rate": 1.054371058315564e-05, + "loss": 0.4526, + "step": 15797 + }, + { + "epoch": 0.49828102822898596, + "grad_norm": 1.3431533575057983, + "learning_rate": 1.054269053492329e-05, + "loss": 0.4969, + "step": 15798 + }, + { + "epoch": 0.4983125689954266, + "grad_norm": 1.5487936735153198, + "learning_rate": 1.0541670481027541e-05, + "loss": 0.4606, + "step": 15799 + }, + { + "epoch": 0.4983441097618672, + "grad_norm": 1.475355863571167, + "learning_rate": 1.0540650421479045e-05, + "loss": 0.7684, + "step": 15800 + }, + { + "epoch": 0.49837565052830785, + "grad_norm": 1.3726445436477661, + "learning_rate": 1.0539630356288438e-05, + "loss": 0.4547, + "step": 15801 + }, + { + "epoch": 0.4984071912947485, + "grad_norm": 1.387782335281372, + "learning_rate": 1.0538610285466368e-05, + "loss": 0.4755, + "step": 15802 + }, + { + "epoch": 0.4984387320611891, + "grad_norm": 1.541225552558899, + "learning_rate": 1.0537590209023482e-05, + "loss": 0.4913, + "step": 15803 + }, + { + "epoch": 0.49847027282762973, + "grad_norm": 1.5256264209747314, + "learning_rate": 1.0536570126970426e-05, + "loss": 0.4591, + "step": 15804 + }, + { + "epoch": 0.4985018135940703, + "grad_norm": 1.702322006225586, + "learning_rate": 1.0535550039317846e-05, + "loss": 0.4833, + "step": 15805 + }, + { + "epoch": 0.49853335436051094, + "grad_norm": 1.702505350112915, + "learning_rate": 1.0534529946076381e-05, + "loss": 0.4361, + "step": 15806 + }, + { + "epoch": 0.49856489512695157, + "grad_norm": 1.4758790731430054, + "learning_rate": 1.0533509847256682e-05, + "loss": 0.4708, + "step": 15807 + }, + { + "epoch": 0.4985964358933922, + "grad_norm": 1.9839171171188354, + "learning_rate": 1.0532489742869397e-05, + "loss": 0.4773, + "step": 15808 + }, + { + "epoch": 0.49862797665983283, + "grad_norm": 2.2613155841827393, + "learning_rate": 1.0531469632925164e-05, + "loss": 0.3891, + "step": 15809 + }, + { + "epoch": 0.49865951742627346, + "grad_norm": 2.6372485160827637, + "learning_rate": 1.0530449517434635e-05, + "loss": 0.4945, + "step": 15810 + }, + { + "epoch": 0.4986910581927141, + "grad_norm": 1.4071911573410034, + "learning_rate": 1.0529429396408452e-05, + "loss": 0.4739, + "step": 15811 + }, + { + "epoch": 0.4987225989591547, + "grad_norm": 1.395798921585083, + "learning_rate": 1.0528409269857267e-05, + "loss": 0.4691, + "step": 15812 + }, + { + "epoch": 0.49875413972559535, + "grad_norm": 1.5962928533554077, + "learning_rate": 1.0527389137791715e-05, + "loss": 0.4818, + "step": 15813 + }, + { + "epoch": 0.498785680492036, + "grad_norm": 1.5318554639816284, + "learning_rate": 1.052636900022245e-05, + "loss": 0.4943, + "step": 15814 + }, + { + "epoch": 0.4988172212584766, + "grad_norm": 1.335323452949524, + "learning_rate": 1.0525348857160116e-05, + "loss": 0.4693, + "step": 15815 + }, + { + "epoch": 0.4988487620249172, + "grad_norm": 1.801580786705017, + "learning_rate": 1.0524328708615356e-05, + "loss": 0.42, + "step": 15816 + }, + { + "epoch": 0.4988803027913578, + "grad_norm": 1.4049243927001953, + "learning_rate": 1.0523308554598822e-05, + "loss": 0.4929, + "step": 15817 + }, + { + "epoch": 0.49891184355779844, + "grad_norm": 1.5540530681610107, + "learning_rate": 1.0522288395121154e-05, + "loss": 0.5024, + "step": 15818 + }, + { + "epoch": 0.49894338432423907, + "grad_norm": 1.367667555809021, + "learning_rate": 1.0521268230193005e-05, + "loss": 0.4659, + "step": 15819 + }, + { + "epoch": 0.4989749250906797, + "grad_norm": 1.5602408647537231, + "learning_rate": 1.0520248059825018e-05, + "loss": 0.4826, + "step": 15820 + }, + { + "epoch": 0.49900646585712033, + "grad_norm": 1.5570019483566284, + "learning_rate": 1.0519227884027834e-05, + "loss": 0.4812, + "step": 15821 + }, + { + "epoch": 0.49903800662356096, + "grad_norm": 1.4408451318740845, + "learning_rate": 1.0518207702812106e-05, + "loss": 0.4864, + "step": 15822 + }, + { + "epoch": 0.4990695473900016, + "grad_norm": 1.8990249633789062, + "learning_rate": 1.0517187516188479e-05, + "loss": 0.4686, + "step": 15823 + }, + { + "epoch": 0.4991010881564422, + "grad_norm": 2.5121378898620605, + "learning_rate": 1.0516167324167597e-05, + "loss": 0.4704, + "step": 15824 + }, + { + "epoch": 0.49913262892288285, + "grad_norm": 1.6425750255584717, + "learning_rate": 1.0515147126760107e-05, + "loss": 0.4782, + "step": 15825 + }, + { + "epoch": 0.4991641696893235, + "grad_norm": 1.8905742168426514, + "learning_rate": 1.0514126923976659e-05, + "loss": 0.4738, + "step": 15826 + }, + { + "epoch": 0.49919571045576405, + "grad_norm": 1.5931121110916138, + "learning_rate": 1.0513106715827897e-05, + "loss": 0.4264, + "step": 15827 + }, + { + "epoch": 0.4992272512222047, + "grad_norm": 1.8137025833129883, + "learning_rate": 1.0512086502324467e-05, + "loss": 0.7729, + "step": 15828 + }, + { + "epoch": 0.4992587919886453, + "grad_norm": 1.7648545503616333, + "learning_rate": 1.0511066283477018e-05, + "loss": 0.484, + "step": 15829 + }, + { + "epoch": 0.49929033275508594, + "grad_norm": 3.6020753383636475, + "learning_rate": 1.0510046059296198e-05, + "loss": 0.473, + "step": 15830 + }, + { + "epoch": 0.49932187352152657, + "grad_norm": 2.0900776386260986, + "learning_rate": 1.0509025829792648e-05, + "loss": 0.4842, + "step": 15831 + }, + { + "epoch": 0.4993534142879672, + "grad_norm": 1.607038974761963, + "learning_rate": 1.050800559497702e-05, + "loss": 0.4878, + "step": 15832 + }, + { + "epoch": 0.49938495505440783, + "grad_norm": 1.5716471672058105, + "learning_rate": 1.0506985354859957e-05, + "loss": 0.8203, + "step": 15833 + }, + { + "epoch": 0.49941649582084846, + "grad_norm": 1.464599609375, + "learning_rate": 1.0505965109452108e-05, + "loss": 0.47, + "step": 15834 + }, + { + "epoch": 0.4994480365872891, + "grad_norm": 1.665663242340088, + "learning_rate": 1.0504944858764118e-05, + "loss": 0.4367, + "step": 15835 + }, + { + "epoch": 0.4994795773537297, + "grad_norm": 1.4681824445724487, + "learning_rate": 1.0503924602806643e-05, + "loss": 0.3946, + "step": 15836 + }, + { + "epoch": 0.4995111181201703, + "grad_norm": 1.5042732954025269, + "learning_rate": 1.0502904341590319e-05, + "loss": 0.4863, + "step": 15837 + }, + { + "epoch": 0.4995426588866109, + "grad_norm": 1.6551016569137573, + "learning_rate": 1.0501884075125798e-05, + "loss": 0.4935, + "step": 15838 + }, + { + "epoch": 0.49957419965305155, + "grad_norm": 2.9221484661102295, + "learning_rate": 1.0500863803423729e-05, + "loss": 0.8063, + "step": 15839 + }, + { + "epoch": 0.4996057404194922, + "grad_norm": 1.629226803779602, + "learning_rate": 1.0499843526494757e-05, + "loss": 0.4766, + "step": 15840 + }, + { + "epoch": 0.4996372811859328, + "grad_norm": 1.36365807056427, + "learning_rate": 1.0498823244349527e-05, + "loss": 0.4751, + "step": 15841 + }, + { + "epoch": 0.49966882195237344, + "grad_norm": 2.3304550647735596, + "learning_rate": 1.049780295699869e-05, + "loss": 0.438, + "step": 15842 + }, + { + "epoch": 0.4997003627188141, + "grad_norm": 1.6795170307159424, + "learning_rate": 1.0496782664452896e-05, + "loss": 0.4427, + "step": 15843 + }, + { + "epoch": 0.4997319034852547, + "grad_norm": 1.7211997509002686, + "learning_rate": 1.0495762366722786e-05, + "loss": 0.4398, + "step": 15844 + }, + { + "epoch": 0.49976344425169533, + "grad_norm": 1.3705601692199707, + "learning_rate": 1.0494742063819011e-05, + "loss": 0.4456, + "step": 15845 + }, + { + "epoch": 0.49979498501813596, + "grad_norm": 1.4830129146575928, + "learning_rate": 1.049372175575222e-05, + "loss": 0.4656, + "step": 15846 + }, + { + "epoch": 0.4998265257845766, + "grad_norm": 1.5978868007659912, + "learning_rate": 1.0492701442533056e-05, + "loss": 0.7578, + "step": 15847 + }, + { + "epoch": 0.49985806655101717, + "grad_norm": 1.499311089515686, + "learning_rate": 1.049168112417217e-05, + "loss": 0.4634, + "step": 15848 + }, + { + "epoch": 0.4998896073174578, + "grad_norm": 1.5900706052780151, + "learning_rate": 1.0490660800680214e-05, + "loss": 0.4747, + "step": 15849 + }, + { + "epoch": 0.4999211480838984, + "grad_norm": 1.4669864177703857, + "learning_rate": 1.0489640472067831e-05, + "loss": 0.46, + "step": 15850 + }, + { + "epoch": 0.49995268885033906, + "grad_norm": 1.5871384143829346, + "learning_rate": 1.0488620138345667e-05, + "loss": 0.4604, + "step": 15851 + }, + { + "epoch": 0.4999842296167797, + "grad_norm": 6.399683952331543, + "learning_rate": 1.048759979952437e-05, + "loss": 0.4906, + "step": 15852 + }, + { + "epoch": 0.5000157703832203, + "grad_norm": 1.4544658660888672, + "learning_rate": 1.0486579455614595e-05, + "loss": 0.4805, + "step": 15853 + }, + { + "epoch": 0.500047311149661, + "grad_norm": 1.476959466934204, + "learning_rate": 1.0485559106626984e-05, + "loss": 0.4875, + "step": 15854 + }, + { + "epoch": 0.5000788519161016, + "grad_norm": 1.5194823741912842, + "learning_rate": 1.0484538752572187e-05, + "loss": 0.3772, + "step": 15855 + }, + { + "epoch": 0.5001103926825422, + "grad_norm": 1.578749179840088, + "learning_rate": 1.048351839346085e-05, + "loss": 0.4847, + "step": 15856 + }, + { + "epoch": 0.5001419334489828, + "grad_norm": 1.4417287111282349, + "learning_rate": 1.0482498029303624e-05, + "loss": 0.3454, + "step": 15857 + }, + { + "epoch": 0.5001734742154235, + "grad_norm": 1.6654013395309448, + "learning_rate": 1.0481477660111156e-05, + "loss": 0.5018, + "step": 15858 + }, + { + "epoch": 0.5002050149818641, + "grad_norm": 1.3173891305923462, + "learning_rate": 1.0480457285894099e-05, + "loss": 0.4415, + "step": 15859 + }, + { + "epoch": 0.5002365557483047, + "grad_norm": 1.4615155458450317, + "learning_rate": 1.0479436906663094e-05, + "loss": 0.4443, + "step": 15860 + }, + { + "epoch": 0.5002680965147454, + "grad_norm": 1.629563331604004, + "learning_rate": 1.0478416522428792e-05, + "loss": 0.4804, + "step": 15861 + }, + { + "epoch": 0.500299637281186, + "grad_norm": 1.2281889915466309, + "learning_rate": 1.0477396133201843e-05, + "loss": 0.439, + "step": 15862 + }, + { + "epoch": 0.5003311780476266, + "grad_norm": 1.4789577722549438, + "learning_rate": 1.0476375738992894e-05, + "loss": 0.4782, + "step": 15863 + }, + { + "epoch": 0.5003627188140671, + "grad_norm": 1.4781652688980103, + "learning_rate": 1.0475355339812595e-05, + "loss": 0.3569, + "step": 15864 + }, + { + "epoch": 0.5003942595805078, + "grad_norm": 2.6261062622070312, + "learning_rate": 1.0474334935671593e-05, + "loss": 0.4962, + "step": 15865 + }, + { + "epoch": 0.5004258003469484, + "grad_norm": 1.539867639541626, + "learning_rate": 1.0473314526580536e-05, + "loss": 0.4922, + "step": 15866 + }, + { + "epoch": 0.500457341113389, + "grad_norm": 2.773361921310425, + "learning_rate": 1.0472294112550079e-05, + "loss": 0.4727, + "step": 15867 + }, + { + "epoch": 0.5004888818798297, + "grad_norm": 1.8061513900756836, + "learning_rate": 1.0471273693590862e-05, + "loss": 0.4763, + "step": 15868 + }, + { + "epoch": 0.5005204226462703, + "grad_norm": 1.423976182937622, + "learning_rate": 1.0470253269713542e-05, + "loss": 0.4862, + "step": 15869 + }, + { + "epoch": 0.5005519634127109, + "grad_norm": 1.413887858390808, + "learning_rate": 1.0469232840928762e-05, + "loss": 0.4384, + "step": 15870 + }, + { + "epoch": 0.5005835041791515, + "grad_norm": 1.3577609062194824, + "learning_rate": 1.0468212407247173e-05, + "loss": 0.4904, + "step": 15871 + }, + { + "epoch": 0.5006150449455922, + "grad_norm": 1.7519901990890503, + "learning_rate": 1.0467191968679422e-05, + "loss": 0.5073, + "step": 15872 + }, + { + "epoch": 0.5006465857120328, + "grad_norm": 1.4855263233184814, + "learning_rate": 1.0466171525236161e-05, + "loss": 0.4899, + "step": 15873 + }, + { + "epoch": 0.5006781264784734, + "grad_norm": 1.8176088333129883, + "learning_rate": 1.0465151076928043e-05, + "loss": 0.4514, + "step": 15874 + }, + { + "epoch": 0.5007096672449141, + "grad_norm": 1.3775813579559326, + "learning_rate": 1.0464130623765705e-05, + "loss": 0.4783, + "step": 15875 + }, + { + "epoch": 0.5007412080113547, + "grad_norm": 1.8869028091430664, + "learning_rate": 1.0463110165759807e-05, + "loss": 0.5086, + "step": 15876 + }, + { + "epoch": 0.5007727487777953, + "grad_norm": 3.8374857902526855, + "learning_rate": 1.0462089702920994e-05, + "loss": 0.4848, + "step": 15877 + }, + { + "epoch": 0.500804289544236, + "grad_norm": 1.6195683479309082, + "learning_rate": 1.0461069235259917e-05, + "loss": 0.4587, + "step": 15878 + }, + { + "epoch": 0.5008358303106766, + "grad_norm": 1.540930151939392, + "learning_rate": 1.0460048762787224e-05, + "loss": 0.4677, + "step": 15879 + }, + { + "epoch": 0.5008673710771172, + "grad_norm": 2.0414538383483887, + "learning_rate": 1.0459028285513566e-05, + "loss": 0.7905, + "step": 15880 + }, + { + "epoch": 0.5008989118435578, + "grad_norm": 1.7092039585113525, + "learning_rate": 1.0458007803449592e-05, + "loss": 0.4799, + "step": 15881 + }, + { + "epoch": 0.5009304526099985, + "grad_norm": 1.7903518676757812, + "learning_rate": 1.0456987316605948e-05, + "loss": 0.5071, + "step": 15882 + }, + { + "epoch": 0.5009619933764391, + "grad_norm": 1.544113278388977, + "learning_rate": 1.0455966824993289e-05, + "loss": 0.4932, + "step": 15883 + }, + { + "epoch": 0.5009935341428797, + "grad_norm": 1.481916904449463, + "learning_rate": 1.0454946328622263e-05, + "loss": 0.7472, + "step": 15884 + }, + { + "epoch": 0.5010250749093202, + "grad_norm": 2.4766125679016113, + "learning_rate": 1.0453925827503515e-05, + "loss": 0.4578, + "step": 15885 + }, + { + "epoch": 0.5010566156757609, + "grad_norm": 1.7042577266693115, + "learning_rate": 1.0452905321647699e-05, + "loss": 0.4889, + "step": 15886 + }, + { + "epoch": 0.5010881564422015, + "grad_norm": 1.4072892665863037, + "learning_rate": 1.0451884811065469e-05, + "loss": 0.4813, + "step": 15887 + }, + { + "epoch": 0.5011196972086421, + "grad_norm": 1.5515263080596924, + "learning_rate": 1.0450864295767466e-05, + "loss": 0.4722, + "step": 15888 + }, + { + "epoch": 0.5011512379750828, + "grad_norm": 1.436658501625061, + "learning_rate": 1.0449843775764346e-05, + "loss": 0.4623, + "step": 15889 + }, + { + "epoch": 0.5011827787415234, + "grad_norm": 1.6154013872146606, + "learning_rate": 1.0448823251066756e-05, + "loss": 0.4006, + "step": 15890 + }, + { + "epoch": 0.501214319507964, + "grad_norm": 2.2340972423553467, + "learning_rate": 1.0447802721685348e-05, + "loss": 0.4579, + "step": 15891 + }, + { + "epoch": 0.5012458602744047, + "grad_norm": 2.087820053100586, + "learning_rate": 1.0446782187630768e-05, + "loss": 0.7875, + "step": 15892 + }, + { + "epoch": 0.5012774010408453, + "grad_norm": 1.9575573205947876, + "learning_rate": 1.044576164891367e-05, + "loss": 0.4818, + "step": 15893 + }, + { + "epoch": 0.5013089418072859, + "grad_norm": 1.353410243988037, + "learning_rate": 1.0444741105544705e-05, + "loss": 0.4864, + "step": 15894 + }, + { + "epoch": 0.5013404825737265, + "grad_norm": 2.0995776653289795, + "learning_rate": 1.0443720557534517e-05, + "loss": 0.4574, + "step": 15895 + }, + { + "epoch": 0.5013720233401672, + "grad_norm": 1.3398956060409546, + "learning_rate": 1.0442700004893764e-05, + "loss": 0.4869, + "step": 15896 + }, + { + "epoch": 0.5014035641066078, + "grad_norm": 1.4874613285064697, + "learning_rate": 1.0441679447633094e-05, + "loss": 0.3354, + "step": 15897 + }, + { + "epoch": 0.5014351048730484, + "grad_norm": 2.0135855674743652, + "learning_rate": 1.0440658885763153e-05, + "loss": 0.4623, + "step": 15898 + }, + { + "epoch": 0.5014666456394891, + "grad_norm": 1.4910285472869873, + "learning_rate": 1.0439638319294595e-05, + "loss": 0.477, + "step": 15899 + }, + { + "epoch": 0.5014981864059297, + "grad_norm": 1.7353554964065552, + "learning_rate": 1.0438617748238072e-05, + "loss": 0.4927, + "step": 15900 + }, + { + "epoch": 0.5015297271723703, + "grad_norm": 1.7061883211135864, + "learning_rate": 1.043759717260423e-05, + "loss": 0.4672, + "step": 15901 + }, + { + "epoch": 0.501561267938811, + "grad_norm": 1.5950183868408203, + "learning_rate": 1.0436576592403722e-05, + "loss": 0.4863, + "step": 15902 + }, + { + "epoch": 0.5015928087052516, + "grad_norm": 2.485032320022583, + "learning_rate": 1.0435556007647198e-05, + "loss": 0.5086, + "step": 15903 + }, + { + "epoch": 0.5016243494716922, + "grad_norm": 1.6875323057174683, + "learning_rate": 1.043453541834531e-05, + "loss": 0.4274, + "step": 15904 + }, + { + "epoch": 0.5016558902381328, + "grad_norm": 1.3982714414596558, + "learning_rate": 1.0433514824508706e-05, + "loss": 0.4611, + "step": 15905 + }, + { + "epoch": 0.5016874310045734, + "grad_norm": 1.6039674282073975, + "learning_rate": 1.0432494226148038e-05, + "loss": 0.7468, + "step": 15906 + }, + { + "epoch": 0.501718971771014, + "grad_norm": 1.6677626371383667, + "learning_rate": 1.043147362327396e-05, + "loss": 0.4905, + "step": 15907 + }, + { + "epoch": 0.5017505125374546, + "grad_norm": 2.1062254905700684, + "learning_rate": 1.0430453015897117e-05, + "loss": 0.5003, + "step": 15908 + }, + { + "epoch": 0.5017820533038952, + "grad_norm": 1.683003544807434, + "learning_rate": 1.0429432404028163e-05, + "loss": 0.519, + "step": 15909 + }, + { + "epoch": 0.5018135940703359, + "grad_norm": 2.0204215049743652, + "learning_rate": 1.0428411787677749e-05, + "loss": 0.4834, + "step": 15910 + }, + { + "epoch": 0.5018451348367765, + "grad_norm": 1.6247656345367432, + "learning_rate": 1.0427391166856527e-05, + "loss": 0.4657, + "step": 15911 + }, + { + "epoch": 0.5018766756032171, + "grad_norm": 1.6993098258972168, + "learning_rate": 1.0426370541575142e-05, + "loss": 0.4457, + "step": 15912 + }, + { + "epoch": 0.5019082163696578, + "grad_norm": 1.7971560955047607, + "learning_rate": 1.0425349911844252e-05, + "loss": 0.4664, + "step": 15913 + }, + { + "epoch": 0.5019397571360984, + "grad_norm": 1.5385212898254395, + "learning_rate": 1.0424329277674504e-05, + "loss": 0.4621, + "step": 15914 + }, + { + "epoch": 0.501971297902539, + "grad_norm": 2.0607681274414062, + "learning_rate": 1.0423308639076553e-05, + "loss": 0.5018, + "step": 15915 + }, + { + "epoch": 0.5020028386689797, + "grad_norm": 2.0492377281188965, + "learning_rate": 1.0422287996061045e-05, + "loss": 0.4822, + "step": 15916 + }, + { + "epoch": 0.5020343794354203, + "grad_norm": 2.0085434913635254, + "learning_rate": 1.0421267348638635e-05, + "loss": 0.496, + "step": 15917 + }, + { + "epoch": 0.5020659202018609, + "grad_norm": 1.9038500785827637, + "learning_rate": 1.0420246696819973e-05, + "loss": 0.5252, + "step": 15918 + }, + { + "epoch": 0.5020974609683015, + "grad_norm": 2.013794422149658, + "learning_rate": 1.041922604061571e-05, + "loss": 0.4799, + "step": 15919 + }, + { + "epoch": 0.5021290017347422, + "grad_norm": 1.6373910903930664, + "learning_rate": 1.0418205380036499e-05, + "loss": 0.4856, + "step": 15920 + }, + { + "epoch": 0.5021605425011828, + "grad_norm": 1.5993177890777588, + "learning_rate": 1.0417184715092987e-05, + "loss": 0.4506, + "step": 15921 + }, + { + "epoch": 0.5021920832676234, + "grad_norm": 1.7371736764907837, + "learning_rate": 1.0416164045795834e-05, + "loss": 0.4899, + "step": 15922 + }, + { + "epoch": 0.5022236240340641, + "grad_norm": 1.4337602853775024, + "learning_rate": 1.0415143372155683e-05, + "loss": 0.4665, + "step": 15923 + }, + { + "epoch": 0.5022551648005047, + "grad_norm": 1.533125638961792, + "learning_rate": 1.0414122694183186e-05, + "loss": 0.4618, + "step": 15924 + }, + { + "epoch": 0.5022867055669453, + "grad_norm": 1.5320818424224854, + "learning_rate": 1.0413102011889001e-05, + "loss": 0.475, + "step": 15925 + }, + { + "epoch": 0.502318246333386, + "grad_norm": 2.9650418758392334, + "learning_rate": 1.0412081325283774e-05, + "loss": 0.493, + "step": 15926 + }, + { + "epoch": 0.5023497870998265, + "grad_norm": 1.5827068090438843, + "learning_rate": 1.041106063437816e-05, + "loss": 0.5002, + "step": 15927 + }, + { + "epoch": 0.5023813278662671, + "grad_norm": 1.7746353149414062, + "learning_rate": 1.0410039939182807e-05, + "loss": 0.469, + "step": 15928 + }, + { + "epoch": 0.5024128686327077, + "grad_norm": 1.7184267044067383, + "learning_rate": 1.0409019239708371e-05, + "loss": 0.4283, + "step": 15929 + }, + { + "epoch": 0.5024444093991484, + "grad_norm": 1.5403425693511963, + "learning_rate": 1.04079985359655e-05, + "loss": 0.4808, + "step": 15930 + }, + { + "epoch": 0.502475950165589, + "grad_norm": 1.6853746175765991, + "learning_rate": 1.0406977827964846e-05, + "loss": 0.4589, + "step": 15931 + }, + { + "epoch": 0.5025074909320296, + "grad_norm": 1.9525481462478638, + "learning_rate": 1.0405957115717066e-05, + "loss": 0.5032, + "step": 15932 + }, + { + "epoch": 0.5025390316984703, + "grad_norm": 1.713467001914978, + "learning_rate": 1.0404936399232806e-05, + "loss": 0.4992, + "step": 15933 + }, + { + "epoch": 0.5025705724649109, + "grad_norm": 1.730601191520691, + "learning_rate": 1.040391567852272e-05, + "loss": 0.4913, + "step": 15934 + }, + { + "epoch": 0.5026021132313515, + "grad_norm": 1.3561980724334717, + "learning_rate": 1.0402894953597463e-05, + "loss": 0.4667, + "step": 15935 + }, + { + "epoch": 0.5026336539977921, + "grad_norm": 1.7096284627914429, + "learning_rate": 1.040187422446768e-05, + "loss": 0.4855, + "step": 15936 + }, + { + "epoch": 0.5026651947642328, + "grad_norm": 1.5290695428848267, + "learning_rate": 1.0400853491144032e-05, + "loss": 0.4692, + "step": 15937 + }, + { + "epoch": 0.5026967355306734, + "grad_norm": 2.330629348754883, + "learning_rate": 1.0399832753637162e-05, + "loss": 0.4451, + "step": 15938 + }, + { + "epoch": 0.502728276297114, + "grad_norm": 1.6786683797836304, + "learning_rate": 1.039881201195773e-05, + "loss": 0.4854, + "step": 15939 + }, + { + "epoch": 0.5027598170635547, + "grad_norm": 1.5991653203964233, + "learning_rate": 1.0397791266116384e-05, + "loss": 0.4652, + "step": 15940 + }, + { + "epoch": 0.5027913578299953, + "grad_norm": 1.7407951354980469, + "learning_rate": 1.0396770516123778e-05, + "loss": 0.462, + "step": 15941 + }, + { + "epoch": 0.5028228985964359, + "grad_norm": 1.9130802154541016, + "learning_rate": 1.0395749761990563e-05, + "loss": 0.4644, + "step": 15942 + }, + { + "epoch": 0.5028544393628765, + "grad_norm": 1.541955590248108, + "learning_rate": 1.039472900372739e-05, + "loss": 0.4488, + "step": 15943 + }, + { + "epoch": 0.5028859801293172, + "grad_norm": 1.4836390018463135, + "learning_rate": 1.0393708241344915e-05, + "loss": 0.5023, + "step": 15944 + }, + { + "epoch": 0.5029175208957578, + "grad_norm": 1.7806771993637085, + "learning_rate": 1.039268747485379e-05, + "loss": 0.4687, + "step": 15945 + }, + { + "epoch": 0.5029490616621984, + "grad_norm": 1.6824406385421753, + "learning_rate": 1.0391666704264664e-05, + "loss": 0.3656, + "step": 15946 + }, + { + "epoch": 0.5029806024286391, + "grad_norm": 2.2289650440216064, + "learning_rate": 1.0390645929588197e-05, + "loss": 0.4639, + "step": 15947 + }, + { + "epoch": 0.5030121431950796, + "grad_norm": 1.881873607635498, + "learning_rate": 1.038962515083503e-05, + "loss": 0.4881, + "step": 15948 + }, + { + "epoch": 0.5030436839615202, + "grad_norm": 2.2978789806365967, + "learning_rate": 1.0388604368015827e-05, + "loss": 0.4823, + "step": 15949 + }, + { + "epoch": 0.5030752247279608, + "grad_norm": 1.6182488203048706, + "learning_rate": 1.0387583581141232e-05, + "loss": 0.4564, + "step": 15950 + }, + { + "epoch": 0.5031067654944015, + "grad_norm": 1.7488291263580322, + "learning_rate": 1.0386562790221901e-05, + "loss": 0.4328, + "step": 15951 + }, + { + "epoch": 0.5031383062608421, + "grad_norm": 1.484971284866333, + "learning_rate": 1.0385541995268488e-05, + "loss": 0.8035, + "step": 15952 + }, + { + "epoch": 0.5031698470272827, + "grad_norm": 1.6099692583084106, + "learning_rate": 1.0384521196291647e-05, + "loss": 0.761, + "step": 15953 + }, + { + "epoch": 0.5032013877937234, + "grad_norm": 1.6034200191497803, + "learning_rate": 1.0383500393302026e-05, + "loss": 0.5244, + "step": 15954 + }, + { + "epoch": 0.503232928560164, + "grad_norm": 1.3904051780700684, + "learning_rate": 1.038247958631028e-05, + "loss": 0.4886, + "step": 15955 + }, + { + "epoch": 0.5032644693266046, + "grad_norm": 2.57741641998291, + "learning_rate": 1.0381458775327067e-05, + "loss": 0.4684, + "step": 15956 + }, + { + "epoch": 0.5032960100930453, + "grad_norm": 1.5363376140594482, + "learning_rate": 1.0380437960363034e-05, + "loss": 0.4974, + "step": 15957 + }, + { + "epoch": 0.5033275508594859, + "grad_norm": 1.4027478694915771, + "learning_rate": 1.0379417141428833e-05, + "loss": 0.4582, + "step": 15958 + }, + { + "epoch": 0.5033590916259265, + "grad_norm": 1.7788288593292236, + "learning_rate": 1.0378396318535125e-05, + "loss": 0.3562, + "step": 15959 + }, + { + "epoch": 0.5033906323923671, + "grad_norm": 2.618893623352051, + "learning_rate": 1.0377375491692551e-05, + "loss": 0.4913, + "step": 15960 + }, + { + "epoch": 0.5034221731588078, + "grad_norm": 2.7033228874206543, + "learning_rate": 1.0376354660911772e-05, + "loss": 0.7773, + "step": 15961 + }, + { + "epoch": 0.5034537139252484, + "grad_norm": 1.7269654273986816, + "learning_rate": 1.037533382620344e-05, + "loss": 0.4616, + "step": 15962 + }, + { + "epoch": 0.503485254691689, + "grad_norm": 1.4995166063308716, + "learning_rate": 1.0374312987578212e-05, + "loss": 0.7782, + "step": 15963 + }, + { + "epoch": 0.5035167954581297, + "grad_norm": 1.5842623710632324, + "learning_rate": 1.0373292145046734e-05, + "loss": 0.4574, + "step": 15964 + }, + { + "epoch": 0.5035483362245703, + "grad_norm": 1.9340718984603882, + "learning_rate": 1.0372271298619661e-05, + "loss": 0.4778, + "step": 15965 + }, + { + "epoch": 0.5035798769910109, + "grad_norm": 1.6925221681594849, + "learning_rate": 1.0371250448307652e-05, + "loss": 0.4881, + "step": 15966 + }, + { + "epoch": 0.5036114177574516, + "grad_norm": 1.390108585357666, + "learning_rate": 1.0370229594121354e-05, + "loss": 0.4877, + "step": 15967 + }, + { + "epoch": 0.5036429585238922, + "grad_norm": 1.406747579574585, + "learning_rate": 1.0369208736071424e-05, + "loss": 0.7717, + "step": 15968 + }, + { + "epoch": 0.5036744992903327, + "grad_norm": 3.168026924133301, + "learning_rate": 1.0368187874168511e-05, + "loss": 0.4026, + "step": 15969 + }, + { + "epoch": 0.5037060400567733, + "grad_norm": 1.6685776710510254, + "learning_rate": 1.0367167008423276e-05, + "loss": 0.4611, + "step": 15970 + }, + { + "epoch": 0.503737580823214, + "grad_norm": 1.7516056299209595, + "learning_rate": 1.0366146138846367e-05, + "loss": 0.4655, + "step": 15971 + }, + { + "epoch": 0.5037691215896546, + "grad_norm": 1.6026679277420044, + "learning_rate": 1.0365125265448437e-05, + "loss": 0.4774, + "step": 15972 + }, + { + "epoch": 0.5038006623560952, + "grad_norm": 2.237926959991455, + "learning_rate": 1.0364104388240142e-05, + "loss": 0.4858, + "step": 15973 + }, + { + "epoch": 0.5038322031225358, + "grad_norm": 1.420082926750183, + "learning_rate": 1.0363083507232134e-05, + "loss": 0.4574, + "step": 15974 + }, + { + "epoch": 0.5038637438889765, + "grad_norm": 1.876617670059204, + "learning_rate": 1.036206262243507e-05, + "loss": 0.4719, + "step": 15975 + }, + { + "epoch": 0.5038952846554171, + "grad_norm": 1.8977504968643188, + "learning_rate": 1.03610417338596e-05, + "loss": 0.4647, + "step": 15976 + }, + { + "epoch": 0.5039268254218577, + "grad_norm": 2.1302907466888428, + "learning_rate": 1.0360020841516381e-05, + "loss": 0.4606, + "step": 15977 + }, + { + "epoch": 0.5039583661882984, + "grad_norm": 2.750652313232422, + "learning_rate": 1.0358999945416062e-05, + "loss": 0.473, + "step": 15978 + }, + { + "epoch": 0.503989906954739, + "grad_norm": 1.6638072729110718, + "learning_rate": 1.0357979045569303e-05, + "loss": 0.4508, + "step": 15979 + }, + { + "epoch": 0.5040214477211796, + "grad_norm": 1.8340790271759033, + "learning_rate": 1.0356958141986753e-05, + "loss": 0.777, + "step": 15980 + }, + { + "epoch": 0.5040529884876203, + "grad_norm": 2.6630072593688965, + "learning_rate": 1.0355937234679065e-05, + "loss": 0.4705, + "step": 15981 + }, + { + "epoch": 0.5040845292540609, + "grad_norm": 1.7550904750823975, + "learning_rate": 1.0354916323656898e-05, + "loss": 0.459, + "step": 15982 + }, + { + "epoch": 0.5041160700205015, + "grad_norm": 1.4540514945983887, + "learning_rate": 1.0353895408930906e-05, + "loss": 0.4837, + "step": 15983 + }, + { + "epoch": 0.5041476107869421, + "grad_norm": 2.2829346656799316, + "learning_rate": 1.0352874490511738e-05, + "loss": 0.4195, + "step": 15984 + }, + { + "epoch": 0.5041791515533828, + "grad_norm": 1.8092386722564697, + "learning_rate": 1.0351853568410048e-05, + "loss": 0.5214, + "step": 15985 + }, + { + "epoch": 0.5042106923198234, + "grad_norm": 1.3415651321411133, + "learning_rate": 1.0350832642636498e-05, + "loss": 0.4681, + "step": 15986 + }, + { + "epoch": 0.504242233086264, + "grad_norm": 1.6248687505722046, + "learning_rate": 1.0349811713201735e-05, + "loss": 0.4634, + "step": 15987 + }, + { + "epoch": 0.5042737738527047, + "grad_norm": 1.6311143636703491, + "learning_rate": 1.0348790780116415e-05, + "loss": 0.498, + "step": 15988 + }, + { + "epoch": 0.5043053146191453, + "grad_norm": 2.461369037628174, + "learning_rate": 1.0347769843391191e-05, + "loss": 0.4787, + "step": 15989 + }, + { + "epoch": 0.5043368553855859, + "grad_norm": 1.6952173709869385, + "learning_rate": 1.0346748903036721e-05, + "loss": 0.4236, + "step": 15990 + }, + { + "epoch": 0.5043683961520264, + "grad_norm": 1.5579105615615845, + "learning_rate": 1.0345727959063656e-05, + "loss": 0.4556, + "step": 15991 + }, + { + "epoch": 0.5043999369184671, + "grad_norm": 1.4430652856826782, + "learning_rate": 1.0344707011482648e-05, + "loss": 0.4656, + "step": 15992 + }, + { + "epoch": 0.5044314776849077, + "grad_norm": 1.855576753616333, + "learning_rate": 1.0343686060304359e-05, + "loss": 0.4987, + "step": 15993 + }, + { + "epoch": 0.5044630184513483, + "grad_norm": 6.424348831176758, + "learning_rate": 1.034266510553944e-05, + "loss": 0.8004, + "step": 15994 + }, + { + "epoch": 0.504494559217789, + "grad_norm": 3.5935959815979004, + "learning_rate": 1.0341644147198539e-05, + "loss": 0.3855, + "step": 15995 + }, + { + "epoch": 0.5045260999842296, + "grad_norm": 1.504770278930664, + "learning_rate": 1.0340623185292322e-05, + "loss": 0.4695, + "step": 15996 + }, + { + "epoch": 0.5045576407506702, + "grad_norm": 1.8479321002960205, + "learning_rate": 1.0339602219831437e-05, + "loss": 0.4688, + "step": 15997 + }, + { + "epoch": 0.5045891815171109, + "grad_norm": 1.4223536252975464, + "learning_rate": 1.0338581250826536e-05, + "loss": 0.7779, + "step": 15998 + }, + { + "epoch": 0.5046207222835515, + "grad_norm": 1.6789486408233643, + "learning_rate": 1.0337560278288277e-05, + "loss": 0.4711, + "step": 15999 + }, + { + "epoch": 0.5046522630499921, + "grad_norm": 1.719190001487732, + "learning_rate": 1.0336539302227315e-05, + "loss": 0.4777, + "step": 16000 + }, + { + "epoch": 0.5046838038164327, + "grad_norm": 1.952188491821289, + "learning_rate": 1.0335518322654305e-05, + "loss": 0.4669, + "step": 16001 + }, + { + "epoch": 0.5047153445828734, + "grad_norm": 1.4618808031082153, + "learning_rate": 1.03344973395799e-05, + "loss": 0.4471, + "step": 16002 + }, + { + "epoch": 0.504746885349314, + "grad_norm": 1.7588965892791748, + "learning_rate": 1.0333476353014755e-05, + "loss": 0.7746, + "step": 16003 + }, + { + "epoch": 0.5047784261157546, + "grad_norm": 1.9027196168899536, + "learning_rate": 1.0332455362969527e-05, + "loss": 0.4635, + "step": 16004 + }, + { + "epoch": 0.5048099668821953, + "grad_norm": 1.749114990234375, + "learning_rate": 1.0331434369454867e-05, + "loss": 0.4931, + "step": 16005 + }, + { + "epoch": 0.5048415076486359, + "grad_norm": 1.81221604347229, + "learning_rate": 1.0330413372481437e-05, + "loss": 0.4537, + "step": 16006 + }, + { + "epoch": 0.5048730484150765, + "grad_norm": 2.1209332942962646, + "learning_rate": 1.0329392372059884e-05, + "loss": 0.4544, + "step": 16007 + }, + { + "epoch": 0.5049045891815171, + "grad_norm": 3.04555344581604, + "learning_rate": 1.0328371368200862e-05, + "loss": 0.3858, + "step": 16008 + }, + { + "epoch": 0.5049361299479578, + "grad_norm": 1.4582494497299194, + "learning_rate": 1.0327350360915034e-05, + "loss": 0.8039, + "step": 16009 + }, + { + "epoch": 0.5049676707143984, + "grad_norm": 1.3362468481063843, + "learning_rate": 1.0326329350213047e-05, + "loss": 0.4712, + "step": 16010 + }, + { + "epoch": 0.504999211480839, + "grad_norm": 1.546236515045166, + "learning_rate": 1.0325308336105565e-05, + "loss": 0.4948, + "step": 16011 + }, + { + "epoch": 0.5050307522472796, + "grad_norm": 1.5570863485336304, + "learning_rate": 1.0324287318603235e-05, + "loss": 0.4693, + "step": 16012 + }, + { + "epoch": 0.5050622930137202, + "grad_norm": 1.3533521890640259, + "learning_rate": 1.0323266297716714e-05, + "loss": 0.4851, + "step": 16013 + }, + { + "epoch": 0.5050938337801608, + "grad_norm": 1.8123869895935059, + "learning_rate": 1.032224527345666e-05, + "loss": 0.4853, + "step": 16014 + }, + { + "epoch": 0.5051253745466014, + "grad_norm": 1.6763588190078735, + "learning_rate": 1.0321224245833725e-05, + "loss": 0.7677, + "step": 16015 + }, + { + "epoch": 0.5051569153130421, + "grad_norm": 2.2138123512268066, + "learning_rate": 1.0320203214858569e-05, + "loss": 0.48, + "step": 16016 + }, + { + "epoch": 0.5051884560794827, + "grad_norm": 1.5882161855697632, + "learning_rate": 1.031918218054184e-05, + "loss": 0.7688, + "step": 16017 + }, + { + "epoch": 0.5052199968459233, + "grad_norm": 1.4633914232254028, + "learning_rate": 1.0318161142894198e-05, + "loss": 0.5193, + "step": 16018 + }, + { + "epoch": 0.505251537612364, + "grad_norm": 2.088517904281616, + "learning_rate": 1.0317140101926298e-05, + "loss": 0.4937, + "step": 16019 + }, + { + "epoch": 0.5052830783788046, + "grad_norm": 1.4384856224060059, + "learning_rate": 1.0316119057648794e-05, + "loss": 0.502, + "step": 16020 + }, + { + "epoch": 0.5053146191452452, + "grad_norm": 1.31947922706604, + "learning_rate": 1.0315098010072343e-05, + "loss": 0.4857, + "step": 16021 + }, + { + "epoch": 0.5053461599116859, + "grad_norm": 1.7365039587020874, + "learning_rate": 1.03140769592076e-05, + "loss": 0.4865, + "step": 16022 + }, + { + "epoch": 0.5053777006781265, + "grad_norm": 1.4126569032669067, + "learning_rate": 1.0313055905065217e-05, + "loss": 0.51, + "step": 16023 + }, + { + "epoch": 0.5054092414445671, + "grad_norm": 1.4162884950637817, + "learning_rate": 1.0312034847655854e-05, + "loss": 0.4852, + "step": 16024 + }, + { + "epoch": 0.5054407822110077, + "grad_norm": 1.4261603355407715, + "learning_rate": 1.0311013786990167e-05, + "loss": 0.4786, + "step": 16025 + }, + { + "epoch": 0.5054723229774484, + "grad_norm": 1.4565248489379883, + "learning_rate": 1.0309992723078813e-05, + "loss": 0.479, + "step": 16026 + }, + { + "epoch": 0.505503863743889, + "grad_norm": 1.4533815383911133, + "learning_rate": 1.0308971655932438e-05, + "loss": 0.4561, + "step": 16027 + }, + { + "epoch": 0.5055354045103296, + "grad_norm": 1.8046358823776245, + "learning_rate": 1.0307950585561705e-05, + "loss": 0.4717, + "step": 16028 + }, + { + "epoch": 0.5055669452767703, + "grad_norm": 1.5924423933029175, + "learning_rate": 1.030692951197727e-05, + "loss": 0.4812, + "step": 16029 + }, + { + "epoch": 0.5055984860432109, + "grad_norm": 1.6489464044570923, + "learning_rate": 1.0305908435189786e-05, + "loss": 0.4635, + "step": 16030 + }, + { + "epoch": 0.5056300268096515, + "grad_norm": 3.53149151802063, + "learning_rate": 1.0304887355209912e-05, + "loss": 0.3742, + "step": 16031 + }, + { + "epoch": 0.5056615675760922, + "grad_norm": 1.4659150838851929, + "learning_rate": 1.0303866272048299e-05, + "loss": 0.4885, + "step": 16032 + }, + { + "epoch": 0.5056931083425327, + "grad_norm": 1.4656562805175781, + "learning_rate": 1.0302845185715605e-05, + "loss": 0.4735, + "step": 16033 + }, + { + "epoch": 0.5057246491089733, + "grad_norm": 1.5499978065490723, + "learning_rate": 1.0301824096222488e-05, + "loss": 0.5086, + "step": 16034 + }, + { + "epoch": 0.5057561898754139, + "grad_norm": 1.807706594467163, + "learning_rate": 1.0300803003579603e-05, + "loss": 0.4218, + "step": 16035 + }, + { + "epoch": 0.5057877306418546, + "grad_norm": 1.5614867210388184, + "learning_rate": 1.0299781907797606e-05, + "loss": 0.4618, + "step": 16036 + }, + { + "epoch": 0.5058192714082952, + "grad_norm": 2.056558132171631, + "learning_rate": 1.029876080888715e-05, + "loss": 0.4884, + "step": 16037 + }, + { + "epoch": 0.5058508121747358, + "grad_norm": 1.6252446174621582, + "learning_rate": 1.0297739706858894e-05, + "loss": 0.502, + "step": 16038 + }, + { + "epoch": 0.5058823529411764, + "grad_norm": 1.4972517490386963, + "learning_rate": 1.029671860172349e-05, + "loss": 0.4906, + "step": 16039 + }, + { + "epoch": 0.5059138937076171, + "grad_norm": 1.3469281196594238, + "learning_rate": 1.0295697493491599e-05, + "loss": 0.4724, + "step": 16040 + }, + { + "epoch": 0.5059454344740577, + "grad_norm": 1.9067448377609253, + "learning_rate": 1.0294676382173874e-05, + "loss": 0.4927, + "step": 16041 + }, + { + "epoch": 0.5059769752404983, + "grad_norm": 1.41658616065979, + "learning_rate": 1.0293655267780975e-05, + "loss": 0.3807, + "step": 16042 + }, + { + "epoch": 0.506008516006939, + "grad_norm": 1.343017339706421, + "learning_rate": 1.0292634150323552e-05, + "loss": 0.4337, + "step": 16043 + }, + { + "epoch": 0.5060400567733796, + "grad_norm": 2.3654890060424805, + "learning_rate": 1.0291613029812267e-05, + "loss": 0.5054, + "step": 16044 + }, + { + "epoch": 0.5060715975398202, + "grad_norm": 1.9850008487701416, + "learning_rate": 1.0290591906257774e-05, + "loss": 0.4898, + "step": 16045 + }, + { + "epoch": 0.5061031383062609, + "grad_norm": 1.320528268814087, + "learning_rate": 1.0289570779670728e-05, + "loss": 0.4855, + "step": 16046 + }, + { + "epoch": 0.5061346790727015, + "grad_norm": 1.6303167343139648, + "learning_rate": 1.0288549650061785e-05, + "loss": 0.4854, + "step": 16047 + }, + { + "epoch": 0.5061662198391421, + "grad_norm": 1.9528127908706665, + "learning_rate": 1.0287528517441603e-05, + "loss": 0.7816, + "step": 16048 + }, + { + "epoch": 0.5061977606055827, + "grad_norm": 1.9312398433685303, + "learning_rate": 1.0286507381820838e-05, + "loss": 0.4898, + "step": 16049 + }, + { + "epoch": 0.5062293013720234, + "grad_norm": 1.4874143600463867, + "learning_rate": 1.0285486243210146e-05, + "loss": 0.4744, + "step": 16050 + }, + { + "epoch": 0.506260842138464, + "grad_norm": 1.7418967485427856, + "learning_rate": 1.0284465101620183e-05, + "loss": 0.4602, + "step": 16051 + }, + { + "epoch": 0.5062923829049046, + "grad_norm": 1.585800290107727, + "learning_rate": 1.0283443957061605e-05, + "loss": 0.4842, + "step": 16052 + }, + { + "epoch": 0.5063239236713453, + "grad_norm": 1.6500535011291504, + "learning_rate": 1.028242280954507e-05, + "loss": 0.4668, + "step": 16053 + }, + { + "epoch": 0.5063554644377858, + "grad_norm": 1.7874161005020142, + "learning_rate": 1.0281401659081231e-05, + "loss": 0.4885, + "step": 16054 + }, + { + "epoch": 0.5063870052042264, + "grad_norm": 1.3771247863769531, + "learning_rate": 1.028038050568075e-05, + "loss": 0.4407, + "step": 16055 + }, + { + "epoch": 0.506418545970667, + "grad_norm": 1.6349458694458008, + "learning_rate": 1.0279359349354285e-05, + "loss": 0.4615, + "step": 16056 + }, + { + "epoch": 0.5064500867371077, + "grad_norm": 1.5258013010025024, + "learning_rate": 1.0278338190112483e-05, + "loss": 0.4812, + "step": 16057 + }, + { + "epoch": 0.5064816275035483, + "grad_norm": 1.3447037935256958, + "learning_rate": 1.0277317027966005e-05, + "loss": 0.7583, + "step": 16058 + }, + { + "epoch": 0.5065131682699889, + "grad_norm": 1.7039437294006348, + "learning_rate": 1.0276295862925511e-05, + "loss": 0.4509, + "step": 16059 + }, + { + "epoch": 0.5065447090364296, + "grad_norm": 1.6534051895141602, + "learning_rate": 1.0275274695001653e-05, + "loss": 0.4643, + "step": 16060 + }, + { + "epoch": 0.5065762498028702, + "grad_norm": 1.9306573867797852, + "learning_rate": 1.0274253524205092e-05, + "loss": 0.4661, + "step": 16061 + }, + { + "epoch": 0.5066077905693108, + "grad_norm": 1.5999341011047363, + "learning_rate": 1.0273232350546482e-05, + "loss": 0.513, + "step": 16062 + }, + { + "epoch": 0.5066393313357515, + "grad_norm": 1.5528243780136108, + "learning_rate": 1.027221117403648e-05, + "loss": 0.4416, + "step": 16063 + }, + { + "epoch": 0.5066708721021921, + "grad_norm": 1.6684805154800415, + "learning_rate": 1.0271189994685743e-05, + "loss": 0.5043, + "step": 16064 + }, + { + "epoch": 0.5067024128686327, + "grad_norm": 1.5546510219573975, + "learning_rate": 1.0270168812504928e-05, + "loss": 0.3767, + "step": 16065 + }, + { + "epoch": 0.5067339536350733, + "grad_norm": 1.5714069604873657, + "learning_rate": 1.0269147627504692e-05, + "loss": 0.4926, + "step": 16066 + }, + { + "epoch": 0.506765494401514, + "grad_norm": 1.6303014755249023, + "learning_rate": 1.0268126439695691e-05, + "loss": 0.4925, + "step": 16067 + }, + { + "epoch": 0.5067970351679546, + "grad_norm": 1.7232458591461182, + "learning_rate": 1.0267105249088582e-05, + "loss": 0.7837, + "step": 16068 + }, + { + "epoch": 0.5068285759343952, + "grad_norm": 2.4633305072784424, + "learning_rate": 1.0266084055694023e-05, + "loss": 0.4867, + "step": 16069 + }, + { + "epoch": 0.5068601167008359, + "grad_norm": 1.2813987731933594, + "learning_rate": 1.026506285952267e-05, + "loss": 0.4881, + "step": 16070 + }, + { + "epoch": 0.5068916574672765, + "grad_norm": 1.2441279888153076, + "learning_rate": 1.026404166058518e-05, + "loss": 0.478, + "step": 16071 + }, + { + "epoch": 0.5069231982337171, + "grad_norm": 1.4540389776229858, + "learning_rate": 1.0263020458892208e-05, + "loss": 0.4735, + "step": 16072 + }, + { + "epoch": 0.5069547390001578, + "grad_norm": 1.544795036315918, + "learning_rate": 1.0261999254454418e-05, + "loss": 0.4648, + "step": 16073 + }, + { + "epoch": 0.5069862797665984, + "grad_norm": 1.6117427349090576, + "learning_rate": 1.0260978047282459e-05, + "loss": 0.3858, + "step": 16074 + }, + { + "epoch": 0.5070178205330389, + "grad_norm": 2.380293130874634, + "learning_rate": 1.0259956837386993e-05, + "loss": 0.4738, + "step": 16075 + }, + { + "epoch": 0.5070493612994795, + "grad_norm": 2.9988973140716553, + "learning_rate": 1.0258935624778676e-05, + "loss": 0.3794, + "step": 16076 + }, + { + "epoch": 0.5070809020659202, + "grad_norm": 1.7224103212356567, + "learning_rate": 1.0257914409468163e-05, + "loss": 0.4571, + "step": 16077 + }, + { + "epoch": 0.5071124428323608, + "grad_norm": 2.0197644233703613, + "learning_rate": 1.0256893191466113e-05, + "loss": 0.5256, + "step": 16078 + }, + { + "epoch": 0.5071439835988014, + "grad_norm": 2.0756888389587402, + "learning_rate": 1.0255871970783184e-05, + "loss": 0.4701, + "step": 16079 + }, + { + "epoch": 0.507175524365242, + "grad_norm": 1.4373396635055542, + "learning_rate": 1.0254850747430034e-05, + "loss": 0.479, + "step": 16080 + }, + { + "epoch": 0.5072070651316827, + "grad_norm": 1.5595284700393677, + "learning_rate": 1.0253829521417315e-05, + "loss": 0.4881, + "step": 16081 + }, + { + "epoch": 0.5072386058981233, + "grad_norm": 2.7734181880950928, + "learning_rate": 1.025280829275569e-05, + "loss": 0.4855, + "step": 16082 + }, + { + "epoch": 0.5072701466645639, + "grad_norm": 1.536160945892334, + "learning_rate": 1.0251787061455813e-05, + "loss": 0.4834, + "step": 16083 + }, + { + "epoch": 0.5073016874310046, + "grad_norm": 2.5988028049468994, + "learning_rate": 1.0250765827528341e-05, + "loss": 0.4771, + "step": 16084 + }, + { + "epoch": 0.5073332281974452, + "grad_norm": 1.371010661125183, + "learning_rate": 1.0249744590983939e-05, + "loss": 0.4938, + "step": 16085 + }, + { + "epoch": 0.5073647689638858, + "grad_norm": 2.3015222549438477, + "learning_rate": 1.0248723351833255e-05, + "loss": 0.4824, + "step": 16086 + }, + { + "epoch": 0.5073963097303265, + "grad_norm": 1.7662415504455566, + "learning_rate": 1.0247702110086948e-05, + "loss": 0.4708, + "step": 16087 + }, + { + "epoch": 0.5074278504967671, + "grad_norm": 1.9622374773025513, + "learning_rate": 1.0246680865755675e-05, + "loss": 0.4768, + "step": 16088 + }, + { + "epoch": 0.5074593912632077, + "grad_norm": 1.4012514352798462, + "learning_rate": 1.0245659618850098e-05, + "loss": 0.471, + "step": 16089 + }, + { + "epoch": 0.5074909320296483, + "grad_norm": 1.565783143043518, + "learning_rate": 1.0244638369380874e-05, + "loss": 0.4656, + "step": 16090 + }, + { + "epoch": 0.507522472796089, + "grad_norm": 3.4508581161499023, + "learning_rate": 1.0243617117358657e-05, + "loss": 0.4557, + "step": 16091 + }, + { + "epoch": 0.5075540135625296, + "grad_norm": 1.4490528106689453, + "learning_rate": 1.0242595862794107e-05, + "loss": 0.4712, + "step": 16092 + }, + { + "epoch": 0.5075855543289702, + "grad_norm": 1.8113218545913696, + "learning_rate": 1.0241574605697881e-05, + "loss": 0.4895, + "step": 16093 + }, + { + "epoch": 0.5076170950954109, + "grad_norm": 1.334215760231018, + "learning_rate": 1.0240553346080636e-05, + "loss": 0.4737, + "step": 16094 + }, + { + "epoch": 0.5076486358618515, + "grad_norm": 1.3506782054901123, + "learning_rate": 1.0239532083953032e-05, + "loss": 0.4779, + "step": 16095 + }, + { + "epoch": 0.507680176628292, + "grad_norm": 1.5219364166259766, + "learning_rate": 1.0238510819325721e-05, + "loss": 0.4551, + "step": 16096 + }, + { + "epoch": 0.5077117173947326, + "grad_norm": 1.8665788173675537, + "learning_rate": 1.0237489552209367e-05, + "loss": 0.492, + "step": 16097 + }, + { + "epoch": 0.5077432581611733, + "grad_norm": 1.4658769369125366, + "learning_rate": 1.0236468282614625e-05, + "loss": 0.4484, + "step": 16098 + }, + { + "epoch": 0.5077747989276139, + "grad_norm": 1.7298425436019897, + "learning_rate": 1.023544701055215e-05, + "loss": 0.4691, + "step": 16099 + }, + { + "epoch": 0.5078063396940545, + "grad_norm": 1.3803224563598633, + "learning_rate": 1.0234425736032607e-05, + "loss": 0.4857, + "step": 16100 + }, + { + "epoch": 0.5078378804604952, + "grad_norm": 1.5204969644546509, + "learning_rate": 1.0233404459066645e-05, + "loss": 0.4721, + "step": 16101 + }, + { + "epoch": 0.5078694212269358, + "grad_norm": 1.813449740409851, + "learning_rate": 1.0232383179664928e-05, + "loss": 0.4387, + "step": 16102 + }, + { + "epoch": 0.5079009619933764, + "grad_norm": 1.7124884128570557, + "learning_rate": 1.0231361897838114e-05, + "loss": 0.4636, + "step": 16103 + }, + { + "epoch": 0.507932502759817, + "grad_norm": 1.689220666885376, + "learning_rate": 1.0230340613596856e-05, + "loss": 0.7593, + "step": 16104 + }, + { + "epoch": 0.5079640435262577, + "grad_norm": 1.931091070175171, + "learning_rate": 1.0229319326951818e-05, + "loss": 0.5032, + "step": 16105 + }, + { + "epoch": 0.5079955842926983, + "grad_norm": 2.2457921504974365, + "learning_rate": 1.0228298037913653e-05, + "loss": 0.4684, + "step": 16106 + }, + { + "epoch": 0.5080271250591389, + "grad_norm": 1.6758188009262085, + "learning_rate": 1.0227276746493022e-05, + "loss": 0.4753, + "step": 16107 + }, + { + "epoch": 0.5080586658255796, + "grad_norm": 1.7777726650238037, + "learning_rate": 1.022625545270058e-05, + "loss": 0.4694, + "step": 16108 + }, + { + "epoch": 0.5080902065920202, + "grad_norm": 3.2116870880126953, + "learning_rate": 1.0225234156546986e-05, + "loss": 0.4402, + "step": 16109 + }, + { + "epoch": 0.5081217473584608, + "grad_norm": 5.564927101135254, + "learning_rate": 1.0224212858042899e-05, + "loss": 0.3697, + "step": 16110 + }, + { + "epoch": 0.5081532881249015, + "grad_norm": 1.7032067775726318, + "learning_rate": 1.0223191557198977e-05, + "loss": 0.4795, + "step": 16111 + }, + { + "epoch": 0.5081848288913421, + "grad_norm": 1.803431510925293, + "learning_rate": 1.0222170254025879e-05, + "loss": 0.5009, + "step": 16112 + }, + { + "epoch": 0.5082163696577827, + "grad_norm": 1.5390543937683105, + "learning_rate": 1.0221148948534261e-05, + "loss": 0.4218, + "step": 16113 + }, + { + "epoch": 0.5082479104242233, + "grad_norm": 1.3419499397277832, + "learning_rate": 1.0220127640734783e-05, + "loss": 0.4435, + "step": 16114 + }, + { + "epoch": 0.508279451190664, + "grad_norm": 2.105621576309204, + "learning_rate": 1.0219106330638103e-05, + "loss": 0.4848, + "step": 16115 + }, + { + "epoch": 0.5083109919571046, + "grad_norm": 1.1856151819229126, + "learning_rate": 1.0218085018254873e-05, + "loss": 0.3676, + "step": 16116 + }, + { + "epoch": 0.5083425327235452, + "grad_norm": 1.4290976524353027, + "learning_rate": 1.0217063703595762e-05, + "loss": 0.74, + "step": 16117 + }, + { + "epoch": 0.5083740734899858, + "grad_norm": 2.0084328651428223, + "learning_rate": 1.0216042386671421e-05, + "loss": 0.4976, + "step": 16118 + }, + { + "epoch": 0.5084056142564264, + "grad_norm": 2.4127426147460938, + "learning_rate": 1.0215021067492507e-05, + "loss": 0.4548, + "step": 16119 + }, + { + "epoch": 0.508437155022867, + "grad_norm": 1.4711633920669556, + "learning_rate": 1.0213999746069684e-05, + "loss": 0.4579, + "step": 16120 + }, + { + "epoch": 0.5084686957893076, + "grad_norm": 1.5642009973526, + "learning_rate": 1.0212978422413608e-05, + "loss": 0.4386, + "step": 16121 + }, + { + "epoch": 0.5085002365557483, + "grad_norm": 1.877593755722046, + "learning_rate": 1.0211957096534936e-05, + "loss": 0.4602, + "step": 16122 + }, + { + "epoch": 0.5085317773221889, + "grad_norm": 2.192371368408203, + "learning_rate": 1.021093576844433e-05, + "loss": 0.4586, + "step": 16123 + }, + { + "epoch": 0.5085633180886295, + "grad_norm": 1.5615607500076294, + "learning_rate": 1.0209914438152443e-05, + "loss": 0.4964, + "step": 16124 + }, + { + "epoch": 0.5085948588550702, + "grad_norm": 1.6548736095428467, + "learning_rate": 1.0208893105669934e-05, + "loss": 0.4925, + "step": 16125 + }, + { + "epoch": 0.5086263996215108, + "grad_norm": 1.5615103244781494, + "learning_rate": 1.0207871771007464e-05, + "loss": 0.4757, + "step": 16126 + }, + { + "epoch": 0.5086579403879514, + "grad_norm": 1.354810357093811, + "learning_rate": 1.0206850434175692e-05, + "loss": 0.3904, + "step": 16127 + }, + { + "epoch": 0.508689481154392, + "grad_norm": 1.9542697668075562, + "learning_rate": 1.0205829095185276e-05, + "loss": 0.519, + "step": 16128 + }, + { + "epoch": 0.5087210219208327, + "grad_norm": 2.0278797149658203, + "learning_rate": 1.020480775404687e-05, + "loss": 0.3697, + "step": 16129 + }, + { + "epoch": 0.5087525626872733, + "grad_norm": 1.7504249811172485, + "learning_rate": 1.0203786410771139e-05, + "loss": 0.4976, + "step": 16130 + }, + { + "epoch": 0.5087841034537139, + "grad_norm": 1.8608622550964355, + "learning_rate": 1.0202765065368739e-05, + "loss": 0.5013, + "step": 16131 + }, + { + "epoch": 0.5088156442201546, + "grad_norm": 1.71866774559021, + "learning_rate": 1.0201743717850325e-05, + "loss": 0.4771, + "step": 16132 + }, + { + "epoch": 0.5088471849865952, + "grad_norm": 1.5854647159576416, + "learning_rate": 1.0200722368226563e-05, + "loss": 0.8021, + "step": 16133 + }, + { + "epoch": 0.5088787257530358, + "grad_norm": 1.8402001857757568, + "learning_rate": 1.0199701016508104e-05, + "loss": 0.4996, + "step": 16134 + }, + { + "epoch": 0.5089102665194765, + "grad_norm": 1.531724452972412, + "learning_rate": 1.019867966270561e-05, + "loss": 0.4759, + "step": 16135 + }, + { + "epoch": 0.5089418072859171, + "grad_norm": 2.9009768962860107, + "learning_rate": 1.019765830682974e-05, + "loss": 0.4421, + "step": 16136 + }, + { + "epoch": 0.5089733480523577, + "grad_norm": 1.4954732656478882, + "learning_rate": 1.0196636948891153e-05, + "loss": 0.3797, + "step": 16137 + }, + { + "epoch": 0.5090048888187984, + "grad_norm": 1.7323857545852661, + "learning_rate": 1.0195615588900507e-05, + "loss": 0.5118, + "step": 16138 + }, + { + "epoch": 0.5090364295852389, + "grad_norm": 1.840691089630127, + "learning_rate": 1.0194594226868457e-05, + "loss": 0.4617, + "step": 16139 + }, + { + "epoch": 0.5090679703516795, + "grad_norm": 1.864126205444336, + "learning_rate": 1.0193572862805667e-05, + "loss": 0.434, + "step": 16140 + }, + { + "epoch": 0.5090995111181201, + "grad_norm": 2.0366601943969727, + "learning_rate": 1.0192551496722796e-05, + "loss": 0.469, + "step": 16141 + }, + { + "epoch": 0.5091310518845608, + "grad_norm": 1.410398244857788, + "learning_rate": 1.0191530128630497e-05, + "loss": 0.4751, + "step": 16142 + }, + { + "epoch": 0.5091625926510014, + "grad_norm": 1.430224895477295, + "learning_rate": 1.0190508758539436e-05, + "loss": 0.4747, + "step": 16143 + }, + { + "epoch": 0.509194133417442, + "grad_norm": 1.3163316249847412, + "learning_rate": 1.0189487386460267e-05, + "loss": 0.4802, + "step": 16144 + }, + { + "epoch": 0.5092256741838826, + "grad_norm": 1.6081641912460327, + "learning_rate": 1.018846601240365e-05, + "loss": 0.4872, + "step": 16145 + }, + { + "epoch": 0.5092572149503233, + "grad_norm": 1.4164005517959595, + "learning_rate": 1.0187444636380243e-05, + "loss": 0.4329, + "step": 16146 + }, + { + "epoch": 0.5092887557167639, + "grad_norm": 1.87483549118042, + "learning_rate": 1.0186423258400707e-05, + "loss": 0.4488, + "step": 16147 + }, + { + "epoch": 0.5093202964832045, + "grad_norm": 2.6805410385131836, + "learning_rate": 1.0185401878475698e-05, + "loss": 0.7888, + "step": 16148 + }, + { + "epoch": 0.5093518372496452, + "grad_norm": 1.2827773094177246, + "learning_rate": 1.0184380496615878e-05, + "loss": 0.4752, + "step": 16149 + }, + { + "epoch": 0.5093833780160858, + "grad_norm": 1.3497508764266968, + "learning_rate": 1.0183359112831901e-05, + "loss": 0.4568, + "step": 16150 + }, + { + "epoch": 0.5094149187825264, + "grad_norm": 1.4544721841812134, + "learning_rate": 1.0182337727134431e-05, + "loss": 0.4585, + "step": 16151 + }, + { + "epoch": 0.5094464595489671, + "grad_norm": 1.2390632629394531, + "learning_rate": 1.0181316339534127e-05, + "loss": 0.4402, + "step": 16152 + }, + { + "epoch": 0.5094780003154077, + "grad_norm": 1.5886362791061401, + "learning_rate": 1.0180294950041647e-05, + "loss": 0.4795, + "step": 16153 + }, + { + "epoch": 0.5095095410818483, + "grad_norm": 1.3677834272384644, + "learning_rate": 1.0179273558667647e-05, + "loss": 0.4536, + "step": 16154 + }, + { + "epoch": 0.509541081848289, + "grad_norm": 1.833409070968628, + "learning_rate": 1.0178252165422788e-05, + "loss": 0.4837, + "step": 16155 + }, + { + "epoch": 0.5095726226147296, + "grad_norm": 1.4164175987243652, + "learning_rate": 1.0177230770317729e-05, + "loss": 0.5064, + "step": 16156 + }, + { + "epoch": 0.5096041633811702, + "grad_norm": 1.8542945384979248, + "learning_rate": 1.017620937336313e-05, + "loss": 0.4725, + "step": 16157 + }, + { + "epoch": 0.5096357041476108, + "grad_norm": 1.9237929582595825, + "learning_rate": 1.0175187974569651e-05, + "loss": 0.4741, + "step": 16158 + }, + { + "epoch": 0.5096672449140515, + "grad_norm": 3.309953212738037, + "learning_rate": 1.0174166573947948e-05, + "loss": 0.4714, + "step": 16159 + }, + { + "epoch": 0.509698785680492, + "grad_norm": 2.7012457847595215, + "learning_rate": 1.017314517150868e-05, + "loss": 0.4577, + "step": 16160 + }, + { + "epoch": 0.5097303264469326, + "grad_norm": 1.491776704788208, + "learning_rate": 1.0172123767262508e-05, + "loss": 0.4495, + "step": 16161 + }, + { + "epoch": 0.5097618672133732, + "grad_norm": 1.584343433380127, + "learning_rate": 1.0171102361220093e-05, + "loss": 0.4474, + "step": 16162 + }, + { + "epoch": 0.5097934079798139, + "grad_norm": 1.7296861410140991, + "learning_rate": 1.0170080953392092e-05, + "loss": 0.5029, + "step": 16163 + }, + { + "epoch": 0.5098249487462545, + "grad_norm": 1.884475827217102, + "learning_rate": 1.0169059543789164e-05, + "loss": 0.5018, + "step": 16164 + }, + { + "epoch": 0.5098564895126951, + "grad_norm": 1.4468539953231812, + "learning_rate": 1.0168038132421969e-05, + "loss": 0.7244, + "step": 16165 + }, + { + "epoch": 0.5098880302791358, + "grad_norm": 1.7556277513504028, + "learning_rate": 1.0167016719301161e-05, + "loss": 0.4691, + "step": 16166 + }, + { + "epoch": 0.5099195710455764, + "grad_norm": 2.1609067916870117, + "learning_rate": 1.0165995304437407e-05, + "loss": 0.466, + "step": 16167 + }, + { + "epoch": 0.509951111812017, + "grad_norm": 1.3522871732711792, + "learning_rate": 1.0164973887841363e-05, + "loss": 0.4658, + "step": 16168 + }, + { + "epoch": 0.5099826525784577, + "grad_norm": 2.1882762908935547, + "learning_rate": 1.016395246952369e-05, + "loss": 0.3621, + "step": 16169 + }, + { + "epoch": 0.5100141933448983, + "grad_norm": 1.4660401344299316, + "learning_rate": 1.0162931049495044e-05, + "loss": 0.4436, + "step": 16170 + }, + { + "epoch": 0.5100457341113389, + "grad_norm": 1.374222755432129, + "learning_rate": 1.0161909627766085e-05, + "loss": 0.4413, + "step": 16171 + }, + { + "epoch": 0.5100772748777795, + "grad_norm": 1.7132177352905273, + "learning_rate": 1.0160888204347474e-05, + "loss": 0.5126, + "step": 16172 + }, + { + "epoch": 0.5101088156442202, + "grad_norm": 1.388954520225525, + "learning_rate": 1.0159866779249872e-05, + "loss": 0.4626, + "step": 16173 + }, + { + "epoch": 0.5101403564106608, + "grad_norm": 1.834603190422058, + "learning_rate": 1.0158845352483935e-05, + "loss": 0.4873, + "step": 16174 + }, + { + "epoch": 0.5101718971771014, + "grad_norm": 1.4838656187057495, + "learning_rate": 1.0157823924060322e-05, + "loss": 0.4243, + "step": 16175 + }, + { + "epoch": 0.5102034379435421, + "grad_norm": 1.6591613292694092, + "learning_rate": 1.0156802493989694e-05, + "loss": 0.7711, + "step": 16176 + }, + { + "epoch": 0.5102349787099827, + "grad_norm": 1.5681931972503662, + "learning_rate": 1.015578106228271e-05, + "loss": 0.4908, + "step": 16177 + }, + { + "epoch": 0.5102665194764233, + "grad_norm": 1.6112982034683228, + "learning_rate": 1.015475962895003e-05, + "loss": 0.4222, + "step": 16178 + }, + { + "epoch": 0.510298060242864, + "grad_norm": 1.4665882587432861, + "learning_rate": 1.0153738194002313e-05, + "loss": 0.4829, + "step": 16179 + }, + { + "epoch": 0.5103296010093046, + "grad_norm": 1.4808202981948853, + "learning_rate": 1.0152716757450218e-05, + "loss": 0.4889, + "step": 16180 + }, + { + "epoch": 0.5103611417757451, + "grad_norm": 1.4189471006393433, + "learning_rate": 1.0151695319304407e-05, + "loss": 0.4963, + "step": 16181 + }, + { + "epoch": 0.5103926825421857, + "grad_norm": 1.340427279472351, + "learning_rate": 1.0150673879575537e-05, + "loss": 0.7436, + "step": 16182 + }, + { + "epoch": 0.5104242233086264, + "grad_norm": 1.4757187366485596, + "learning_rate": 1.0149652438274268e-05, + "loss": 0.3611, + "step": 16183 + }, + { + "epoch": 0.510455764075067, + "grad_norm": 1.701899528503418, + "learning_rate": 1.014863099541126e-05, + "loss": 0.453, + "step": 16184 + }, + { + "epoch": 0.5104873048415076, + "grad_norm": 1.7063156366348267, + "learning_rate": 1.014760955099717e-05, + "loss": 0.4749, + "step": 16185 + }, + { + "epoch": 0.5105188456079482, + "grad_norm": 2.879223108291626, + "learning_rate": 1.0146588105042662e-05, + "loss": 0.5011, + "step": 16186 + }, + { + "epoch": 0.5105503863743889, + "grad_norm": 1.8484991788864136, + "learning_rate": 1.0145566657558392e-05, + "loss": 0.4881, + "step": 16187 + }, + { + "epoch": 0.5105819271408295, + "grad_norm": 1.6540935039520264, + "learning_rate": 1.0144545208555021e-05, + "loss": 0.4921, + "step": 16188 + }, + { + "epoch": 0.5106134679072701, + "grad_norm": 1.6500098705291748, + "learning_rate": 1.014352375804321e-05, + "loss": 0.4823, + "step": 16189 + }, + { + "epoch": 0.5106450086737108, + "grad_norm": 1.4926989078521729, + "learning_rate": 1.0142502306033614e-05, + "loss": 0.4549, + "step": 16190 + }, + { + "epoch": 0.5106765494401514, + "grad_norm": 1.4731292724609375, + "learning_rate": 1.0141480852536898e-05, + "loss": 0.5136, + "step": 16191 + }, + { + "epoch": 0.510708090206592, + "grad_norm": 1.1988921165466309, + "learning_rate": 1.014045939756372e-05, + "loss": 0.3456, + "step": 16192 + }, + { + "epoch": 0.5107396309730327, + "grad_norm": 1.620645523071289, + "learning_rate": 1.013943794112474e-05, + "loss": 0.7782, + "step": 16193 + }, + { + "epoch": 0.5107711717394733, + "grad_norm": 1.574668526649475, + "learning_rate": 1.0138416483230616e-05, + "loss": 0.8086, + "step": 16194 + }, + { + "epoch": 0.5108027125059139, + "grad_norm": 1.5480809211730957, + "learning_rate": 1.0137395023892004e-05, + "loss": 0.4672, + "step": 16195 + }, + { + "epoch": 0.5108342532723545, + "grad_norm": 1.3089861869812012, + "learning_rate": 1.0136373563119575e-05, + "loss": 0.4727, + "step": 16196 + }, + { + "epoch": 0.5108657940387952, + "grad_norm": 2.2631256580352783, + "learning_rate": 1.0135352100923976e-05, + "loss": 0.4843, + "step": 16197 + }, + { + "epoch": 0.5108973348052358, + "grad_norm": 2.108309507369995, + "learning_rate": 1.0134330637315876e-05, + "loss": 0.4874, + "step": 16198 + }, + { + "epoch": 0.5109288755716764, + "grad_norm": 1.484427571296692, + "learning_rate": 1.013330917230593e-05, + "loss": 0.4887, + "step": 16199 + }, + { + "epoch": 0.5109604163381171, + "grad_norm": 1.5557866096496582, + "learning_rate": 1.0132287705904802e-05, + "loss": 0.4874, + "step": 16200 + }, + { + "epoch": 0.5109919571045577, + "grad_norm": 1.488357663154602, + "learning_rate": 1.0131266238123148e-05, + "loss": 0.4328, + "step": 16201 + }, + { + "epoch": 0.5110234978709982, + "grad_norm": 1.8414411544799805, + "learning_rate": 1.013024476897163e-05, + "loss": 0.443, + "step": 16202 + }, + { + "epoch": 0.5110550386374388, + "grad_norm": 1.6184998750686646, + "learning_rate": 1.0129223298460904e-05, + "loss": 0.4917, + "step": 16203 + }, + { + "epoch": 0.5110865794038795, + "grad_norm": 1.5593807697296143, + "learning_rate": 1.0128201826601633e-05, + "loss": 0.5055, + "step": 16204 + }, + { + "epoch": 0.5111181201703201, + "grad_norm": 1.32619309425354, + "learning_rate": 1.0127180353404476e-05, + "loss": 0.4851, + "step": 16205 + }, + { + "epoch": 0.5111496609367607, + "grad_norm": 1.3599520921707153, + "learning_rate": 1.0126158878880097e-05, + "loss": 0.4764, + "step": 16206 + }, + { + "epoch": 0.5111812017032014, + "grad_norm": 1.5538440942764282, + "learning_rate": 1.0125137403039147e-05, + "loss": 0.4555, + "step": 16207 + }, + { + "epoch": 0.511212742469642, + "grad_norm": 1.3746408224105835, + "learning_rate": 1.0124115925892294e-05, + "loss": 0.4702, + "step": 16208 + }, + { + "epoch": 0.5112442832360826, + "grad_norm": 1.8823094367980957, + "learning_rate": 1.0123094447450192e-05, + "loss": 0.4162, + "step": 16209 + }, + { + "epoch": 0.5112758240025232, + "grad_norm": 1.3462575674057007, + "learning_rate": 1.0122072967723508e-05, + "loss": 0.4347, + "step": 16210 + }, + { + "epoch": 0.5113073647689639, + "grad_norm": 1.6463441848754883, + "learning_rate": 1.0121051486722895e-05, + "loss": 0.4914, + "step": 16211 + }, + { + "epoch": 0.5113389055354045, + "grad_norm": 1.4666330814361572, + "learning_rate": 1.0120030004459019e-05, + "loss": 0.4771, + "step": 16212 + }, + { + "epoch": 0.5113704463018451, + "grad_norm": 1.5370696783065796, + "learning_rate": 1.0119008520942535e-05, + "loss": 0.4706, + "step": 16213 + }, + { + "epoch": 0.5114019870682858, + "grad_norm": 1.8256759643554688, + "learning_rate": 1.0117987036184102e-05, + "loss": 0.4936, + "step": 16214 + }, + { + "epoch": 0.5114335278347264, + "grad_norm": 1.6839983463287354, + "learning_rate": 1.0116965550194384e-05, + "loss": 0.7641, + "step": 16215 + }, + { + "epoch": 0.511465068601167, + "grad_norm": 1.5711570978164673, + "learning_rate": 1.011594406298404e-05, + "loss": 0.4716, + "step": 16216 + }, + { + "epoch": 0.5114966093676077, + "grad_norm": 1.5306214094161987, + "learning_rate": 1.0114922574563733e-05, + "loss": 0.5039, + "step": 16217 + }, + { + "epoch": 0.5115281501340483, + "grad_norm": 1.6900192499160767, + "learning_rate": 1.0113901084944116e-05, + "loss": 0.7647, + "step": 16218 + }, + { + "epoch": 0.5115596909004889, + "grad_norm": 2.2718069553375244, + "learning_rate": 1.0112879594135852e-05, + "loss": 0.47, + "step": 16219 + }, + { + "epoch": 0.5115912316669295, + "grad_norm": 1.6930443048477173, + "learning_rate": 1.0111858102149603e-05, + "loss": 0.4704, + "step": 16220 + }, + { + "epoch": 0.5116227724333702, + "grad_norm": 1.5990983247756958, + "learning_rate": 1.0110836608996028e-05, + "loss": 0.4805, + "step": 16221 + }, + { + "epoch": 0.5116543131998108, + "grad_norm": 1.7963758707046509, + "learning_rate": 1.0109815114685788e-05, + "loss": 0.4549, + "step": 16222 + }, + { + "epoch": 0.5116858539662513, + "grad_norm": 1.319777488708496, + "learning_rate": 1.0108793619229543e-05, + "loss": 0.4676, + "step": 16223 + }, + { + "epoch": 0.511717394732692, + "grad_norm": 1.4369466304779053, + "learning_rate": 1.0107772122637948e-05, + "loss": 0.478, + "step": 16224 + }, + { + "epoch": 0.5117489354991326, + "grad_norm": 1.3534471988677979, + "learning_rate": 1.010675062492167e-05, + "loss": 0.4627, + "step": 16225 + }, + { + "epoch": 0.5117804762655732, + "grad_norm": 2.1517300605773926, + "learning_rate": 1.0105729126091366e-05, + "loss": 0.7886, + "step": 16226 + }, + { + "epoch": 0.5118120170320138, + "grad_norm": 1.5007661581039429, + "learning_rate": 1.0104707626157698e-05, + "loss": 0.4867, + "step": 16227 + }, + { + "epoch": 0.5118435577984545, + "grad_norm": 2.6843044757843018, + "learning_rate": 1.010368612513132e-05, + "loss": 0.4982, + "step": 16228 + }, + { + "epoch": 0.5118750985648951, + "grad_norm": 2.7737226486206055, + "learning_rate": 1.01026646230229e-05, + "loss": 0.3995, + "step": 16229 + }, + { + "epoch": 0.5119066393313357, + "grad_norm": 1.530333399772644, + "learning_rate": 1.0101643119843096e-05, + "loss": 0.4764, + "step": 16230 + }, + { + "epoch": 0.5119381800977764, + "grad_norm": 1.831369400024414, + "learning_rate": 1.0100621615602564e-05, + "loss": 0.4482, + "step": 16231 + }, + { + "epoch": 0.511969720864217, + "grad_norm": 1.616357684135437, + "learning_rate": 1.0099600110311971e-05, + "loss": 0.4397, + "step": 16232 + }, + { + "epoch": 0.5120012616306576, + "grad_norm": 1.795571208000183, + "learning_rate": 1.009857860398197e-05, + "loss": 0.4644, + "step": 16233 + }, + { + "epoch": 0.5120328023970983, + "grad_norm": 1.3960837125778198, + "learning_rate": 1.0097557096623227e-05, + "loss": 0.4387, + "step": 16234 + }, + { + "epoch": 0.5120643431635389, + "grad_norm": 1.807394027709961, + "learning_rate": 1.0096535588246399e-05, + "loss": 0.5012, + "step": 16235 + }, + { + "epoch": 0.5120958839299795, + "grad_norm": 1.735235333442688, + "learning_rate": 1.0095514078862147e-05, + "loss": 0.4702, + "step": 16236 + }, + { + "epoch": 0.5121274246964201, + "grad_norm": 1.460780382156372, + "learning_rate": 1.0094492568481134e-05, + "loss": 0.5021, + "step": 16237 + }, + { + "epoch": 0.5121589654628608, + "grad_norm": 1.9296544790267944, + "learning_rate": 1.0093471057114012e-05, + "loss": 0.4441, + "step": 16238 + }, + { + "epoch": 0.5121905062293014, + "grad_norm": 1.3423300981521606, + "learning_rate": 1.009244954477145e-05, + "loss": 0.5035, + "step": 16239 + }, + { + "epoch": 0.512222046995742, + "grad_norm": 1.324965000152588, + "learning_rate": 1.0091428031464106e-05, + "loss": 0.4824, + "step": 16240 + }, + { + "epoch": 0.5122535877621827, + "grad_norm": 1.2906057834625244, + "learning_rate": 1.009040651720264e-05, + "loss": 0.4803, + "step": 16241 + }, + { + "epoch": 0.5122851285286233, + "grad_norm": 1.6271113157272339, + "learning_rate": 1.0089385001997711e-05, + "loss": 0.4831, + "step": 16242 + }, + { + "epoch": 0.5123166692950639, + "grad_norm": 2.996574640274048, + "learning_rate": 1.008836348585998e-05, + "loss": 0.3952, + "step": 16243 + }, + { + "epoch": 0.5123482100615045, + "grad_norm": 5.951584815979004, + "learning_rate": 1.0087341968800108e-05, + "loss": 0.8082, + "step": 16244 + }, + { + "epoch": 0.5123797508279451, + "grad_norm": 1.8602616786956787, + "learning_rate": 1.0086320450828755e-05, + "loss": 0.4505, + "step": 16245 + }, + { + "epoch": 0.5124112915943857, + "grad_norm": 1.580153226852417, + "learning_rate": 1.008529893195658e-05, + "loss": 0.4861, + "step": 16246 + }, + { + "epoch": 0.5124428323608263, + "grad_norm": 2.0716047286987305, + "learning_rate": 1.0084277412194243e-05, + "loss": 0.7797, + "step": 16247 + }, + { + "epoch": 0.512474373127267, + "grad_norm": 1.3380593061447144, + "learning_rate": 1.0083255891552408e-05, + "loss": 0.7198, + "step": 16248 + }, + { + "epoch": 0.5125059138937076, + "grad_norm": 1.6072027683258057, + "learning_rate": 1.0082234370041733e-05, + "loss": 0.4681, + "step": 16249 + }, + { + "epoch": 0.5125374546601482, + "grad_norm": 1.5886585712432861, + "learning_rate": 1.0081212847672876e-05, + "loss": 0.4813, + "step": 16250 + }, + { + "epoch": 0.5125689954265888, + "grad_norm": 1.4324356317520142, + "learning_rate": 1.0080191324456503e-05, + "loss": 0.4325, + "step": 16251 + }, + { + "epoch": 0.5126005361930295, + "grad_norm": 1.985848069190979, + "learning_rate": 1.0079169800403272e-05, + "loss": 0.4782, + "step": 16252 + }, + { + "epoch": 0.5126320769594701, + "grad_norm": 1.6351511478424072, + "learning_rate": 1.007814827552384e-05, + "loss": 0.4887, + "step": 16253 + }, + { + "epoch": 0.5126636177259107, + "grad_norm": 1.6059308052062988, + "learning_rate": 1.007712674982887e-05, + "loss": 0.73, + "step": 16254 + }, + { + "epoch": 0.5126951584923514, + "grad_norm": 4.481325626373291, + "learning_rate": 1.0076105223329024e-05, + "loss": 0.7528, + "step": 16255 + }, + { + "epoch": 0.512726699258792, + "grad_norm": 1.362281084060669, + "learning_rate": 1.007508369603496e-05, + "loss": 0.4544, + "step": 16256 + }, + { + "epoch": 0.5127582400252326, + "grad_norm": 1.4710125923156738, + "learning_rate": 1.007406216795734e-05, + "loss": 0.4274, + "step": 16257 + }, + { + "epoch": 0.5127897807916733, + "grad_norm": 1.4956637620925903, + "learning_rate": 1.0073040639106824e-05, + "loss": 0.4793, + "step": 16258 + }, + { + "epoch": 0.5128213215581139, + "grad_norm": 3.184407949447632, + "learning_rate": 1.0072019109494072e-05, + "loss": 0.4494, + "step": 16259 + }, + { + "epoch": 0.5128528623245545, + "grad_norm": 2.08502197265625, + "learning_rate": 1.0070997579129742e-05, + "loss": 0.4732, + "step": 16260 + }, + { + "epoch": 0.5128844030909951, + "grad_norm": 1.4801993370056152, + "learning_rate": 1.0069976048024501e-05, + "loss": 0.4792, + "step": 16261 + }, + { + "epoch": 0.5129159438574358, + "grad_norm": 1.369072675704956, + "learning_rate": 1.0068954516189004e-05, + "loss": 0.4465, + "step": 16262 + }, + { + "epoch": 0.5129474846238764, + "grad_norm": 1.944456934928894, + "learning_rate": 1.0067932983633913e-05, + "loss": 0.4786, + "step": 16263 + }, + { + "epoch": 0.512979025390317, + "grad_norm": 1.947327733039856, + "learning_rate": 1.0066911450369887e-05, + "loss": 0.7868, + "step": 16264 + }, + { + "epoch": 0.5130105661567577, + "grad_norm": 2.5120747089385986, + "learning_rate": 1.0065889916407591e-05, + "loss": 0.5057, + "step": 16265 + }, + { + "epoch": 0.5130421069231982, + "grad_norm": 2.1817445755004883, + "learning_rate": 1.0064868381757682e-05, + "loss": 0.465, + "step": 16266 + }, + { + "epoch": 0.5130736476896388, + "grad_norm": 1.4394396543502808, + "learning_rate": 1.0063846846430818e-05, + "loss": 0.3485, + "step": 16267 + }, + { + "epoch": 0.5131051884560794, + "grad_norm": 1.5043225288391113, + "learning_rate": 1.0062825310437668e-05, + "loss": 0.4699, + "step": 16268 + }, + { + "epoch": 0.5131367292225201, + "grad_norm": 1.422986388206482, + "learning_rate": 1.0061803773788882e-05, + "loss": 0.4603, + "step": 16269 + }, + { + "epoch": 0.5131682699889607, + "grad_norm": 1.5120106935501099, + "learning_rate": 1.0060782236495125e-05, + "loss": 0.4797, + "step": 16270 + }, + { + "epoch": 0.5131998107554013, + "grad_norm": 1.8013713359832764, + "learning_rate": 1.0059760698567064e-05, + "loss": 0.476, + "step": 16271 + }, + { + "epoch": 0.513231351521842, + "grad_norm": 2.771853446960449, + "learning_rate": 1.0058739160015351e-05, + "loss": 0.466, + "step": 16272 + }, + { + "epoch": 0.5132628922882826, + "grad_norm": 1.7531119585037231, + "learning_rate": 1.005771762085065e-05, + "loss": 0.4778, + "step": 16273 + }, + { + "epoch": 0.5132944330547232, + "grad_norm": 1.5875979661941528, + "learning_rate": 1.0056696081083617e-05, + "loss": 0.4669, + "step": 16274 + }, + { + "epoch": 0.5133259738211639, + "grad_norm": 1.4110422134399414, + "learning_rate": 1.005567454072492e-05, + "loss": 0.4296, + "step": 16275 + }, + { + "epoch": 0.5133575145876045, + "grad_norm": 1.567613959312439, + "learning_rate": 1.0054652999785215e-05, + "loss": 0.482, + "step": 16276 + }, + { + "epoch": 0.5133890553540451, + "grad_norm": 1.4932582378387451, + "learning_rate": 1.0053631458275161e-05, + "loss": 0.48, + "step": 16277 + }, + { + "epoch": 0.5134205961204857, + "grad_norm": 2.0327446460723877, + "learning_rate": 1.0052609916205424e-05, + "loss": 0.4524, + "step": 16278 + }, + { + "epoch": 0.5134521368869264, + "grad_norm": 1.7880574464797974, + "learning_rate": 1.0051588373586662e-05, + "loss": 0.4878, + "step": 16279 + }, + { + "epoch": 0.513483677653367, + "grad_norm": 1.3633133172988892, + "learning_rate": 1.0050566830429533e-05, + "loss": 0.4819, + "step": 16280 + }, + { + "epoch": 0.5135152184198076, + "grad_norm": 1.7453734874725342, + "learning_rate": 1.0049545286744704e-05, + "loss": 0.5069, + "step": 16281 + }, + { + "epoch": 0.5135467591862483, + "grad_norm": 1.6503281593322754, + "learning_rate": 1.0048523742542829e-05, + "loss": 0.462, + "step": 16282 + }, + { + "epoch": 0.5135782999526889, + "grad_norm": 2.1263089179992676, + "learning_rate": 1.004750219783457e-05, + "loss": 0.7482, + "step": 16283 + }, + { + "epoch": 0.5136098407191295, + "grad_norm": 1.9176225662231445, + "learning_rate": 1.0046480652630588e-05, + "loss": 0.454, + "step": 16284 + }, + { + "epoch": 0.5136413814855701, + "grad_norm": 1.9464162588119507, + "learning_rate": 1.0045459106941549e-05, + "loss": 0.4424, + "step": 16285 + }, + { + "epoch": 0.5136729222520108, + "grad_norm": 1.6482641696929932, + "learning_rate": 1.0044437560778105e-05, + "loss": 0.4863, + "step": 16286 + }, + { + "epoch": 0.5137044630184513, + "grad_norm": 1.6398606300354004, + "learning_rate": 1.004341601415092e-05, + "loss": 0.5134, + "step": 16287 + }, + { + "epoch": 0.5137360037848919, + "grad_norm": 1.5963655710220337, + "learning_rate": 1.0042394467070655e-05, + "loss": 0.4357, + "step": 16288 + }, + { + "epoch": 0.5137675445513326, + "grad_norm": 1.6502716541290283, + "learning_rate": 1.0041372919547973e-05, + "loss": 0.4741, + "step": 16289 + }, + { + "epoch": 0.5137990853177732, + "grad_norm": 1.4675993919372559, + "learning_rate": 1.004035137159353e-05, + "loss": 0.4697, + "step": 16290 + }, + { + "epoch": 0.5138306260842138, + "grad_norm": 1.4563333988189697, + "learning_rate": 1.0039329823217994e-05, + "loss": 0.447, + "step": 16291 + }, + { + "epoch": 0.5138621668506544, + "grad_norm": 1.620041012763977, + "learning_rate": 1.0038308274432018e-05, + "loss": 0.4496, + "step": 16292 + }, + { + "epoch": 0.5138937076170951, + "grad_norm": 1.5144017934799194, + "learning_rate": 1.0037286725246265e-05, + "loss": 0.444, + "step": 16293 + }, + { + "epoch": 0.5139252483835357, + "grad_norm": 1.4620212316513062, + "learning_rate": 1.0036265175671396e-05, + "loss": 0.4856, + "step": 16294 + }, + { + "epoch": 0.5139567891499763, + "grad_norm": 1.5889158248901367, + "learning_rate": 1.0035243625718069e-05, + "loss": 0.4579, + "step": 16295 + }, + { + "epoch": 0.513988329916417, + "grad_norm": 2.2265353202819824, + "learning_rate": 1.0034222075396954e-05, + "loss": 0.4787, + "step": 16296 + }, + { + "epoch": 0.5140198706828576, + "grad_norm": 1.565402626991272, + "learning_rate": 1.0033200524718699e-05, + "loss": 0.7441, + "step": 16297 + }, + { + "epoch": 0.5140514114492982, + "grad_norm": 1.8036799430847168, + "learning_rate": 1.0032178973693972e-05, + "loss": 0.456, + "step": 16298 + }, + { + "epoch": 0.5140829522157389, + "grad_norm": 1.5546380281448364, + "learning_rate": 1.0031157422333435e-05, + "loss": 0.4967, + "step": 16299 + }, + { + "epoch": 0.5141144929821795, + "grad_norm": 1.7846028804779053, + "learning_rate": 1.0030135870647743e-05, + "loss": 0.4466, + "step": 16300 + }, + { + "epoch": 0.5141460337486201, + "grad_norm": 1.4177215099334717, + "learning_rate": 1.0029114318647564e-05, + "loss": 0.4691, + "step": 16301 + }, + { + "epoch": 0.5141775745150607, + "grad_norm": 5.850820541381836, + "learning_rate": 1.002809276634355e-05, + "loss": 0.3863, + "step": 16302 + }, + { + "epoch": 0.5142091152815014, + "grad_norm": 2.3245182037353516, + "learning_rate": 1.0027071213746369e-05, + "loss": 0.4172, + "step": 16303 + }, + { + "epoch": 0.514240656047942, + "grad_norm": 1.9897780418395996, + "learning_rate": 1.0026049660866677e-05, + "loss": 0.5124, + "step": 16304 + }, + { + "epoch": 0.5142721968143826, + "grad_norm": 1.6955540180206299, + "learning_rate": 1.0025028107715137e-05, + "loss": 0.4575, + "step": 16305 + }, + { + "epoch": 0.5143037375808233, + "grad_norm": 1.6072969436645508, + "learning_rate": 1.002400655430241e-05, + "loss": 0.4798, + "step": 16306 + }, + { + "epoch": 0.5143352783472639, + "grad_norm": 1.6723257303237915, + "learning_rate": 1.0022985000639155e-05, + "loss": 0.5015, + "step": 16307 + }, + { + "epoch": 0.5143668191137044, + "grad_norm": 1.4845550060272217, + "learning_rate": 1.0021963446736034e-05, + "loss": 0.4636, + "step": 16308 + }, + { + "epoch": 0.514398359880145, + "grad_norm": 4.247931480407715, + "learning_rate": 1.0020941892603707e-05, + "loss": 0.3653, + "step": 16309 + }, + { + "epoch": 0.5144299006465857, + "grad_norm": 1.4342669248580933, + "learning_rate": 1.0019920338252838e-05, + "loss": 0.7502, + "step": 16310 + }, + { + "epoch": 0.5144614414130263, + "grad_norm": 2.2238166332244873, + "learning_rate": 1.0018898783694083e-05, + "loss": 0.4556, + "step": 16311 + }, + { + "epoch": 0.5144929821794669, + "grad_norm": 1.154452919960022, + "learning_rate": 1.0017877228938103e-05, + "loss": 0.3639, + "step": 16312 + }, + { + "epoch": 0.5145245229459076, + "grad_norm": 1.8915802240371704, + "learning_rate": 1.0016855673995562e-05, + "loss": 0.4755, + "step": 16313 + }, + { + "epoch": 0.5145560637123482, + "grad_norm": 1.5422362089157104, + "learning_rate": 1.001583411887712e-05, + "loss": 0.4702, + "step": 16314 + }, + { + "epoch": 0.5145876044787888, + "grad_norm": 1.4463860988616943, + "learning_rate": 1.0014812563593433e-05, + "loss": 0.4534, + "step": 16315 + }, + { + "epoch": 0.5146191452452294, + "grad_norm": 5.242225170135498, + "learning_rate": 1.0013791008155166e-05, + "loss": 0.4494, + "step": 16316 + }, + { + "epoch": 0.5146506860116701, + "grad_norm": 2.7133948802948, + "learning_rate": 1.0012769452572982e-05, + "loss": 0.4782, + "step": 16317 + }, + { + "epoch": 0.5146822267781107, + "grad_norm": 2.117701768875122, + "learning_rate": 1.0011747896857536e-05, + "loss": 0.442, + "step": 16318 + }, + { + "epoch": 0.5147137675445513, + "grad_norm": 1.6743818521499634, + "learning_rate": 1.0010726341019495e-05, + "loss": 0.494, + "step": 16319 + }, + { + "epoch": 0.514745308310992, + "grad_norm": 2.1097259521484375, + "learning_rate": 1.0009704785069512e-05, + "loss": 0.4901, + "step": 16320 + }, + { + "epoch": 0.5147768490774326, + "grad_norm": 1.5891257524490356, + "learning_rate": 1.0008683229018257e-05, + "loss": 0.4807, + "step": 16321 + }, + { + "epoch": 0.5148083898438732, + "grad_norm": 2.421940565109253, + "learning_rate": 1.000766167287638e-05, + "loss": 0.3518, + "step": 16322 + }, + { + "epoch": 0.5148399306103139, + "grad_norm": 2.3370752334594727, + "learning_rate": 1.0006640116654554e-05, + "loss": 0.7683, + "step": 16323 + }, + { + "epoch": 0.5148714713767545, + "grad_norm": 1.5194662809371948, + "learning_rate": 1.0005618560363429e-05, + "loss": 0.4554, + "step": 16324 + }, + { + "epoch": 0.5149030121431951, + "grad_norm": 3.406792640686035, + "learning_rate": 1.000459700401367e-05, + "loss": 0.7794, + "step": 16325 + }, + { + "epoch": 0.5149345529096357, + "grad_norm": 3.8058595657348633, + "learning_rate": 1.0003575447615937e-05, + "loss": 0.7967, + "step": 16326 + }, + { + "epoch": 0.5149660936760764, + "grad_norm": 1.4818236827850342, + "learning_rate": 1.0002553891180894e-05, + "loss": 0.4759, + "step": 16327 + }, + { + "epoch": 0.514997634442517, + "grad_norm": 1.4563398361206055, + "learning_rate": 1.0001532334719198e-05, + "loss": 0.4447, + "step": 16328 + }, + { + "epoch": 0.5150291752089575, + "grad_norm": 1.3054014444351196, + "learning_rate": 1.000051077824151e-05, + "loss": 0.4566, + "step": 16329 + }, + { + "epoch": 0.5150607159753982, + "grad_norm": 1.3086490631103516, + "learning_rate": 9.999489221758492e-06, + "loss": 0.4403, + "step": 16330 + }, + { + "epoch": 0.5150922567418388, + "grad_norm": 1.369776725769043, + "learning_rate": 9.998467665280806e-06, + "loss": 0.4697, + "step": 16331 + }, + { + "epoch": 0.5151237975082794, + "grad_norm": 1.863682508468628, + "learning_rate": 9.997446108819112e-06, + "loss": 0.4937, + "step": 16332 + }, + { + "epoch": 0.51515533827472, + "grad_norm": 1.2790454626083374, + "learning_rate": 9.996424552384065e-06, + "loss": 0.4988, + "step": 16333 + }, + { + "epoch": 0.5151868790411607, + "grad_norm": 1.5503851175308228, + "learning_rate": 9.995402995986333e-06, + "loss": 0.433, + "step": 16334 + }, + { + "epoch": 0.5152184198076013, + "grad_norm": 1.6806813478469849, + "learning_rate": 9.994381439636576e-06, + "loss": 0.4708, + "step": 16335 + }, + { + "epoch": 0.5152499605740419, + "grad_norm": 1.5325788259506226, + "learning_rate": 9.993359883345449e-06, + "loss": 0.7234, + "step": 16336 + }, + { + "epoch": 0.5152815013404826, + "grad_norm": 1.593279242515564, + "learning_rate": 9.992338327123623e-06, + "loss": 0.4531, + "step": 16337 + }, + { + "epoch": 0.5153130421069232, + "grad_norm": 1.2829103469848633, + "learning_rate": 9.991316770981748e-06, + "loss": 0.4736, + "step": 16338 + }, + { + "epoch": 0.5153445828733638, + "grad_norm": 1.6297591924667358, + "learning_rate": 9.99029521493049e-06, + "loss": 0.474, + "step": 16339 + }, + { + "epoch": 0.5153761236398045, + "grad_norm": 1.74323308467865, + "learning_rate": 9.989273658980509e-06, + "loss": 0.4776, + "step": 16340 + }, + { + "epoch": 0.5154076644062451, + "grad_norm": 1.4916833639144897, + "learning_rate": 9.988252103142464e-06, + "loss": 0.4989, + "step": 16341 + }, + { + "epoch": 0.5154392051726857, + "grad_norm": 1.6650879383087158, + "learning_rate": 9.987230547427021e-06, + "loss": 0.4938, + "step": 16342 + }, + { + "epoch": 0.5154707459391263, + "grad_norm": 1.5213291645050049, + "learning_rate": 9.986208991844837e-06, + "loss": 0.4546, + "step": 16343 + }, + { + "epoch": 0.515502286705567, + "grad_norm": 1.538032054901123, + "learning_rate": 9.985187436406568e-06, + "loss": 0.7482, + "step": 16344 + }, + { + "epoch": 0.5155338274720076, + "grad_norm": 1.545815110206604, + "learning_rate": 9.984165881122885e-06, + "loss": 0.4388, + "step": 16345 + }, + { + "epoch": 0.5155653682384482, + "grad_norm": 1.4719170331954956, + "learning_rate": 9.983144326004443e-06, + "loss": 0.4868, + "step": 16346 + }, + { + "epoch": 0.5155969090048889, + "grad_norm": 1.6770572662353516, + "learning_rate": 9.982122771061897e-06, + "loss": 0.4192, + "step": 16347 + }, + { + "epoch": 0.5156284497713295, + "grad_norm": 1.7682347297668457, + "learning_rate": 9.981101216305919e-06, + "loss": 0.4129, + "step": 16348 + }, + { + "epoch": 0.5156599905377701, + "grad_norm": 2.0001344680786133, + "learning_rate": 9.980079661747165e-06, + "loss": 0.501, + "step": 16349 + }, + { + "epoch": 0.5156915313042106, + "grad_norm": 1.5980618000030518, + "learning_rate": 9.979058107396294e-06, + "loss": 0.4971, + "step": 16350 + }, + { + "epoch": 0.5157230720706513, + "grad_norm": 1.5832550525665283, + "learning_rate": 9.97803655326397e-06, + "loss": 0.7473, + "step": 16351 + }, + { + "epoch": 0.5157546128370919, + "grad_norm": 1.5378466844558716, + "learning_rate": 9.97701499936085e-06, + "loss": 0.4946, + "step": 16352 + }, + { + "epoch": 0.5157861536035325, + "grad_norm": 1.7831565141677856, + "learning_rate": 9.975993445697591e-06, + "loss": 0.4317, + "step": 16353 + }, + { + "epoch": 0.5158176943699732, + "grad_norm": 2.1599035263061523, + "learning_rate": 9.974971892284865e-06, + "loss": 0.4641, + "step": 16354 + }, + { + "epoch": 0.5158492351364138, + "grad_norm": 1.3986531496047974, + "learning_rate": 9.973950339133323e-06, + "loss": 0.4786, + "step": 16355 + }, + { + "epoch": 0.5158807759028544, + "grad_norm": 1.6405733823776245, + "learning_rate": 9.972928786253635e-06, + "loss": 0.4774, + "step": 16356 + }, + { + "epoch": 0.515912316669295, + "grad_norm": 1.336756706237793, + "learning_rate": 9.971907233656452e-06, + "loss": 0.4399, + "step": 16357 + }, + { + "epoch": 0.5159438574357357, + "grad_norm": 1.5988744497299194, + "learning_rate": 9.97088568135244e-06, + "loss": 0.4903, + "step": 16358 + }, + { + "epoch": 0.5159753982021763, + "grad_norm": 1.5707770586013794, + "learning_rate": 9.969864129352259e-06, + "loss": 0.8025, + "step": 16359 + }, + { + "epoch": 0.5160069389686169, + "grad_norm": 1.3675856590270996, + "learning_rate": 9.968842577666568e-06, + "loss": 0.4573, + "step": 16360 + }, + { + "epoch": 0.5160384797350576, + "grad_norm": 1.5743745565414429, + "learning_rate": 9.967821026306028e-06, + "loss": 0.7568, + "step": 16361 + }, + { + "epoch": 0.5160700205014982, + "grad_norm": 1.3853312730789185, + "learning_rate": 9.966799475281305e-06, + "loss": 0.4833, + "step": 16362 + }, + { + "epoch": 0.5161015612679388, + "grad_norm": 1.6094645261764526, + "learning_rate": 9.965777924603053e-06, + "loss": 0.4869, + "step": 16363 + }, + { + "epoch": 0.5161331020343795, + "grad_norm": 2.2229714393615723, + "learning_rate": 9.964756374281931e-06, + "loss": 0.4499, + "step": 16364 + }, + { + "epoch": 0.5161646428008201, + "grad_norm": 1.6972464323043823, + "learning_rate": 9.963734824328607e-06, + "loss": 0.4489, + "step": 16365 + }, + { + "epoch": 0.5161961835672607, + "grad_norm": 1.3500726222991943, + "learning_rate": 9.96271327475374e-06, + "loss": 0.4676, + "step": 16366 + }, + { + "epoch": 0.5162277243337013, + "grad_norm": 2.8606314659118652, + "learning_rate": 9.961691725567984e-06, + "loss": 0.7772, + "step": 16367 + }, + { + "epoch": 0.516259265100142, + "grad_norm": 2.2073886394500732, + "learning_rate": 9.96067017678201e-06, + "loss": 0.4719, + "step": 16368 + }, + { + "epoch": 0.5162908058665826, + "grad_norm": 1.5561803579330444, + "learning_rate": 9.959648628406473e-06, + "loss": 0.4836, + "step": 16369 + }, + { + "epoch": 0.5163223466330232, + "grad_norm": 1.5385221242904663, + "learning_rate": 9.958627080452028e-06, + "loss": 0.5198, + "step": 16370 + }, + { + "epoch": 0.5163538873994639, + "grad_norm": 1.8242638111114502, + "learning_rate": 9.957605532929347e-06, + "loss": 0.475, + "step": 16371 + }, + { + "epoch": 0.5163854281659044, + "grad_norm": 1.5504074096679688, + "learning_rate": 9.956583985849081e-06, + "loss": 0.4577, + "step": 16372 + }, + { + "epoch": 0.516416968932345, + "grad_norm": 1.5391167402267456, + "learning_rate": 9.955562439221899e-06, + "loss": 0.775, + "step": 16373 + }, + { + "epoch": 0.5164485096987856, + "grad_norm": 1.8777272701263428, + "learning_rate": 9.954540893058456e-06, + "loss": 0.4523, + "step": 16374 + }, + { + "epoch": 0.5164800504652263, + "grad_norm": 1.5004668235778809, + "learning_rate": 9.953519347369412e-06, + "loss": 0.4919, + "step": 16375 + }, + { + "epoch": 0.5165115912316669, + "grad_norm": 1.5058410167694092, + "learning_rate": 9.952497802165433e-06, + "loss": 0.765, + "step": 16376 + }, + { + "epoch": 0.5165431319981075, + "grad_norm": 2.3461251258850098, + "learning_rate": 9.951476257457176e-06, + "loss": 0.4713, + "step": 16377 + }, + { + "epoch": 0.5165746727645482, + "grad_norm": 1.767506718635559, + "learning_rate": 9.9504547132553e-06, + "loss": 0.4731, + "step": 16378 + }, + { + "epoch": 0.5166062135309888, + "grad_norm": 2.647825241088867, + "learning_rate": 9.949433169570469e-06, + "loss": 0.4303, + "step": 16379 + }, + { + "epoch": 0.5166377542974294, + "grad_norm": 2.1546437740325928, + "learning_rate": 9.94841162641334e-06, + "loss": 0.465, + "step": 16380 + }, + { + "epoch": 0.51666929506387, + "grad_norm": 1.8098485469818115, + "learning_rate": 9.947390083794577e-06, + "loss": 0.4712, + "step": 16381 + }, + { + "epoch": 0.5167008358303107, + "grad_norm": 1.673581600189209, + "learning_rate": 9.946368541724842e-06, + "loss": 0.4755, + "step": 16382 + }, + { + "epoch": 0.5167323765967513, + "grad_norm": 2.8055107593536377, + "learning_rate": 9.94534700021479e-06, + "loss": 0.455, + "step": 16383 + }, + { + "epoch": 0.5167639173631919, + "grad_norm": 1.8210867643356323, + "learning_rate": 9.944325459275083e-06, + "loss": 0.4872, + "step": 16384 + }, + { + "epoch": 0.5167954581296326, + "grad_norm": 1.668716549873352, + "learning_rate": 9.943303918916386e-06, + "loss": 0.4798, + "step": 16385 + }, + { + "epoch": 0.5168269988960732, + "grad_norm": 2.253300905227661, + "learning_rate": 9.942282379149352e-06, + "loss": 0.4434, + "step": 16386 + }, + { + "epoch": 0.5168585396625138, + "grad_norm": 1.8097255229949951, + "learning_rate": 9.94126083998465e-06, + "loss": 0.4765, + "step": 16387 + }, + { + "epoch": 0.5168900804289545, + "grad_norm": 6.401833534240723, + "learning_rate": 9.94023930143294e-06, + "loss": 0.7685, + "step": 16388 + }, + { + "epoch": 0.5169216211953951, + "grad_norm": 1.7246397733688354, + "learning_rate": 9.939217763504873e-06, + "loss": 0.3436, + "step": 16389 + }, + { + "epoch": 0.5169531619618357, + "grad_norm": 1.6355671882629395, + "learning_rate": 9.938196226211121e-06, + "loss": 0.4604, + "step": 16390 + }, + { + "epoch": 0.5169847027282763, + "grad_norm": 1.8679367303848267, + "learning_rate": 9.937174689562339e-06, + "loss": 0.7908, + "step": 16391 + }, + { + "epoch": 0.517016243494717, + "grad_norm": 1.7305400371551514, + "learning_rate": 9.936153153569182e-06, + "loss": 0.5081, + "step": 16392 + }, + { + "epoch": 0.5170477842611575, + "grad_norm": 1.7854459285736084, + "learning_rate": 9.935131618242322e-06, + "loss": 0.4671, + "step": 16393 + }, + { + "epoch": 0.5170793250275981, + "grad_norm": 1.4793128967285156, + "learning_rate": 9.934110083592414e-06, + "loss": 0.4615, + "step": 16394 + }, + { + "epoch": 0.5171108657940388, + "grad_norm": 1.6025322675704956, + "learning_rate": 9.933088549630113e-06, + "loss": 0.491, + "step": 16395 + }, + { + "epoch": 0.5171424065604794, + "grad_norm": 1.5965931415557861, + "learning_rate": 9.93206701636609e-06, + "loss": 0.4613, + "step": 16396 + }, + { + "epoch": 0.51717394732692, + "grad_norm": 2.1147117614746094, + "learning_rate": 9.931045483811001e-06, + "loss": 0.4959, + "step": 16397 + }, + { + "epoch": 0.5172054880933606, + "grad_norm": 2.0651674270629883, + "learning_rate": 9.930023951975502e-06, + "loss": 0.4683, + "step": 16398 + }, + { + "epoch": 0.5172370288598013, + "grad_norm": 1.6455806493759155, + "learning_rate": 9.929002420870261e-06, + "loss": 0.4826, + "step": 16399 + }, + { + "epoch": 0.5172685696262419, + "grad_norm": 3.375497341156006, + "learning_rate": 9.927980890505933e-06, + "loss": 0.755, + "step": 16400 + }, + { + "epoch": 0.5173001103926825, + "grad_norm": 1.490843653678894, + "learning_rate": 9.92695936089318e-06, + "loss": 0.4604, + "step": 16401 + }, + { + "epoch": 0.5173316511591232, + "grad_norm": 1.5210539102554321, + "learning_rate": 9.925937832042663e-06, + "loss": 0.4983, + "step": 16402 + }, + { + "epoch": 0.5173631919255638, + "grad_norm": 2.192992925643921, + "learning_rate": 9.924916303965042e-06, + "loss": 0.4629, + "step": 16403 + }, + { + "epoch": 0.5173947326920044, + "grad_norm": 1.548133373260498, + "learning_rate": 9.923894776670978e-06, + "loss": 0.4619, + "step": 16404 + }, + { + "epoch": 0.517426273458445, + "grad_norm": 1.4825994968414307, + "learning_rate": 9.922873250171134e-06, + "loss": 0.4693, + "step": 16405 + }, + { + "epoch": 0.5174578142248857, + "grad_norm": 1.8577203750610352, + "learning_rate": 9.92185172447616e-06, + "loss": 0.4794, + "step": 16406 + }, + { + "epoch": 0.5174893549913263, + "grad_norm": 1.591989278793335, + "learning_rate": 9.920830199596731e-06, + "loss": 0.4533, + "step": 16407 + }, + { + "epoch": 0.5175208957577669, + "grad_norm": 1.6279551982879639, + "learning_rate": 9.9198086755435e-06, + "loss": 0.4771, + "step": 16408 + }, + { + "epoch": 0.5175524365242076, + "grad_norm": 1.5475883483886719, + "learning_rate": 9.918787152327124e-06, + "loss": 0.499, + "step": 16409 + }, + { + "epoch": 0.5175839772906482, + "grad_norm": 1.6388365030288696, + "learning_rate": 9.91776562995827e-06, + "loss": 0.4674, + "step": 16410 + }, + { + "epoch": 0.5176155180570888, + "grad_norm": 1.57870352268219, + "learning_rate": 9.916744108447597e-06, + "loss": 0.4672, + "step": 16411 + }, + { + "epoch": 0.5176470588235295, + "grad_norm": 2.155059576034546, + "learning_rate": 9.915722587805758e-06, + "loss": 0.477, + "step": 16412 + }, + { + "epoch": 0.5176785995899701, + "grad_norm": 1.24502694606781, + "learning_rate": 9.914701068043423e-06, + "loss": 0.453, + "step": 16413 + }, + { + "epoch": 0.5177101403564106, + "grad_norm": 1.755260705947876, + "learning_rate": 9.91367954917125e-06, + "loss": 0.3336, + "step": 16414 + }, + { + "epoch": 0.5177416811228512, + "grad_norm": 1.6790825128555298, + "learning_rate": 9.912658031199894e-06, + "loss": 0.4859, + "step": 16415 + }, + { + "epoch": 0.5177732218892919, + "grad_norm": 1.743011474609375, + "learning_rate": 9.911636514140022e-06, + "loss": 0.4809, + "step": 16416 + }, + { + "epoch": 0.5178047626557325, + "grad_norm": 2.0756566524505615, + "learning_rate": 9.910614998002292e-06, + "loss": 0.4775, + "step": 16417 + }, + { + "epoch": 0.5178363034221731, + "grad_norm": 1.9035828113555908, + "learning_rate": 9.909593482797364e-06, + "loss": 0.451, + "step": 16418 + }, + { + "epoch": 0.5178678441886138, + "grad_norm": 2.5975453853607178, + "learning_rate": 9.908571968535898e-06, + "loss": 0.4692, + "step": 16419 + }, + { + "epoch": 0.5178993849550544, + "grad_norm": 1.6363277435302734, + "learning_rate": 9.90755045522855e-06, + "loss": 0.4974, + "step": 16420 + }, + { + "epoch": 0.517930925721495, + "grad_norm": 1.4588544368743896, + "learning_rate": 9.90652894288599e-06, + "loss": 0.4732, + "step": 16421 + }, + { + "epoch": 0.5179624664879356, + "grad_norm": 1.9718551635742188, + "learning_rate": 9.905507431518873e-06, + "loss": 0.4922, + "step": 16422 + }, + { + "epoch": 0.5179940072543763, + "grad_norm": 1.832974910736084, + "learning_rate": 9.904485921137854e-06, + "loss": 0.4504, + "step": 16423 + }, + { + "epoch": 0.5180255480208169, + "grad_norm": 1.4847959280014038, + "learning_rate": 9.903464411753604e-06, + "loss": 0.4732, + "step": 16424 + }, + { + "epoch": 0.5180570887872575, + "grad_norm": 2.422722578048706, + "learning_rate": 9.902442903376778e-06, + "loss": 0.4474, + "step": 16425 + }, + { + "epoch": 0.5180886295536982, + "grad_norm": 2.698087215423584, + "learning_rate": 9.90142139601803e-06, + "loss": 0.4949, + "step": 16426 + }, + { + "epoch": 0.5181201703201388, + "grad_norm": 1.8513728380203247, + "learning_rate": 9.900399889688032e-06, + "loss": 0.486, + "step": 16427 + }, + { + "epoch": 0.5181517110865794, + "grad_norm": 2.4626431465148926, + "learning_rate": 9.89937838439744e-06, + "loss": 0.4326, + "step": 16428 + }, + { + "epoch": 0.51818325185302, + "grad_norm": 1.3965461254119873, + "learning_rate": 9.898356880156907e-06, + "loss": 0.4636, + "step": 16429 + }, + { + "epoch": 0.5182147926194607, + "grad_norm": 2.3762333393096924, + "learning_rate": 9.897335376977104e-06, + "loss": 0.4554, + "step": 16430 + }, + { + "epoch": 0.5182463333859013, + "grad_norm": 2.0225913524627686, + "learning_rate": 9.896313874868685e-06, + "loss": 0.468, + "step": 16431 + }, + { + "epoch": 0.5182778741523419, + "grad_norm": 1.6329100131988525, + "learning_rate": 9.895292373842306e-06, + "loss": 0.7981, + "step": 16432 + }, + { + "epoch": 0.5183094149187826, + "grad_norm": 1.9010769128799438, + "learning_rate": 9.894270873908638e-06, + "loss": 0.4305, + "step": 16433 + }, + { + "epoch": 0.5183409556852232, + "grad_norm": 1.5981484651565552, + "learning_rate": 9.893249375078332e-06, + "loss": 0.4736, + "step": 16434 + }, + { + "epoch": 0.5183724964516637, + "grad_norm": 1.51542329788208, + "learning_rate": 9.892227877362053e-06, + "loss": 0.4912, + "step": 16435 + }, + { + "epoch": 0.5184040372181044, + "grad_norm": 1.6579235792160034, + "learning_rate": 9.891206380770462e-06, + "loss": 0.7599, + "step": 16436 + }, + { + "epoch": 0.518435577984545, + "grad_norm": 2.0835392475128174, + "learning_rate": 9.890184885314214e-06, + "loss": 0.7617, + "step": 16437 + }, + { + "epoch": 0.5184671187509856, + "grad_norm": 3.57240629196167, + "learning_rate": 9.889163391003973e-06, + "loss": 0.4358, + "step": 16438 + }, + { + "epoch": 0.5184986595174262, + "grad_norm": 1.6590511798858643, + "learning_rate": 9.8881418978504e-06, + "loss": 0.4854, + "step": 16439 + }, + { + "epoch": 0.5185302002838669, + "grad_norm": 1.443672776222229, + "learning_rate": 9.88712040586415e-06, + "loss": 0.4532, + "step": 16440 + }, + { + "epoch": 0.5185617410503075, + "grad_norm": 1.8327409029006958, + "learning_rate": 9.886098915055888e-06, + "loss": 0.4876, + "step": 16441 + }, + { + "epoch": 0.5185932818167481, + "grad_norm": 1.5095462799072266, + "learning_rate": 9.885077425436274e-06, + "loss": 0.3577, + "step": 16442 + }, + { + "epoch": 0.5186248225831888, + "grad_norm": 1.4878472089767456, + "learning_rate": 9.88405593701596e-06, + "loss": 0.4886, + "step": 16443 + }, + { + "epoch": 0.5186563633496294, + "grad_norm": 1.5146048069000244, + "learning_rate": 9.883034449805618e-06, + "loss": 0.445, + "step": 16444 + }, + { + "epoch": 0.51868790411607, + "grad_norm": 1.90603506565094, + "learning_rate": 9.882012963815903e-06, + "loss": 0.3733, + "step": 16445 + }, + { + "epoch": 0.5187194448825106, + "grad_norm": 1.4732885360717773, + "learning_rate": 9.880991479057469e-06, + "loss": 0.4943, + "step": 16446 + }, + { + "epoch": 0.5187509856489513, + "grad_norm": 1.6389087438583374, + "learning_rate": 9.879969995540985e-06, + "loss": 0.4744, + "step": 16447 + }, + { + "epoch": 0.5187825264153919, + "grad_norm": 1.4906260967254639, + "learning_rate": 9.878948513277108e-06, + "loss": 0.4802, + "step": 16448 + }, + { + "epoch": 0.5188140671818325, + "grad_norm": 4.692153453826904, + "learning_rate": 9.877927032276495e-06, + "loss": 0.7648, + "step": 16449 + }, + { + "epoch": 0.5188456079482732, + "grad_norm": 2.4791259765625, + "learning_rate": 9.87690555254981e-06, + "loss": 0.4615, + "step": 16450 + }, + { + "epoch": 0.5188771487147138, + "grad_norm": 1.4990711212158203, + "learning_rate": 9.875884074107708e-06, + "loss": 0.4721, + "step": 16451 + }, + { + "epoch": 0.5189086894811544, + "grad_norm": 2.11240816116333, + "learning_rate": 9.874862596960856e-06, + "loss": 0.7163, + "step": 16452 + }, + { + "epoch": 0.5189402302475951, + "grad_norm": 1.8430153131484985, + "learning_rate": 9.87384112111991e-06, + "loss": 0.4911, + "step": 16453 + }, + { + "epoch": 0.5189717710140357, + "grad_norm": 4.444954872131348, + "learning_rate": 9.872819646595524e-06, + "loss": 0.4851, + "step": 16454 + }, + { + "epoch": 0.5190033117804763, + "grad_norm": 2.673877716064453, + "learning_rate": 9.87179817339837e-06, + "loss": 0.4755, + "step": 16455 + }, + { + "epoch": 0.5190348525469168, + "grad_norm": 1.7032028436660767, + "learning_rate": 9.8707767015391e-06, + "loss": 0.4547, + "step": 16456 + }, + { + "epoch": 0.5190663933133575, + "grad_norm": 1.7293407917022705, + "learning_rate": 9.869755231028375e-06, + "loss": 0.4766, + "step": 16457 + }, + { + "epoch": 0.5190979340797981, + "grad_norm": 1.9923958778381348, + "learning_rate": 9.868733761876857e-06, + "loss": 0.4893, + "step": 16458 + }, + { + "epoch": 0.5191294748462387, + "grad_norm": 1.7162740230560303, + "learning_rate": 9.867712294095203e-06, + "loss": 0.7924, + "step": 16459 + }, + { + "epoch": 0.5191610156126794, + "grad_norm": 1.4100204706192017, + "learning_rate": 9.866690827694071e-06, + "loss": 0.4435, + "step": 16460 + }, + { + "epoch": 0.51919255637912, + "grad_norm": 2.345895767211914, + "learning_rate": 9.865669362684127e-06, + "loss": 0.471, + "step": 16461 + }, + { + "epoch": 0.5192240971455606, + "grad_norm": 4.2786407470703125, + "learning_rate": 9.864647899076028e-06, + "loss": 0.4513, + "step": 16462 + }, + { + "epoch": 0.5192556379120012, + "grad_norm": 2.147616147994995, + "learning_rate": 9.863626436880429e-06, + "loss": 0.4934, + "step": 16463 + }, + { + "epoch": 0.5192871786784419, + "grad_norm": 1.5246273279190063, + "learning_rate": 9.862604976107997e-06, + "loss": 0.4648, + "step": 16464 + }, + { + "epoch": 0.5193187194448825, + "grad_norm": 1.866244912147522, + "learning_rate": 9.86158351676939e-06, + "loss": 0.4317, + "step": 16465 + }, + { + "epoch": 0.5193502602113231, + "grad_norm": 1.405222773551941, + "learning_rate": 9.860562058875262e-06, + "loss": 0.4816, + "step": 16466 + }, + { + "epoch": 0.5193818009777638, + "grad_norm": 1.4183881282806396, + "learning_rate": 9.859540602436282e-06, + "loss": 0.5019, + "step": 16467 + }, + { + "epoch": 0.5194133417442044, + "grad_norm": 1.5192161798477173, + "learning_rate": 9.858519147463102e-06, + "loss": 0.456, + "step": 16468 + }, + { + "epoch": 0.519444882510645, + "grad_norm": 1.346167802810669, + "learning_rate": 9.857497693966388e-06, + "loss": 0.4687, + "step": 16469 + }, + { + "epoch": 0.5194764232770857, + "grad_norm": 1.9800206422805786, + "learning_rate": 9.856476241956795e-06, + "loss": 0.4739, + "step": 16470 + }, + { + "epoch": 0.5195079640435263, + "grad_norm": 1.851698637008667, + "learning_rate": 9.85545479144498e-06, + "loss": 0.3811, + "step": 16471 + }, + { + "epoch": 0.5195395048099669, + "grad_norm": 1.9641441106796265, + "learning_rate": 9.854433342441611e-06, + "loss": 0.4539, + "step": 16472 + }, + { + "epoch": 0.5195710455764075, + "grad_norm": 2.0147697925567627, + "learning_rate": 9.853411894957343e-06, + "loss": 0.4776, + "step": 16473 + }, + { + "epoch": 0.5196025863428482, + "grad_norm": 1.5605190992355347, + "learning_rate": 9.85239044900283e-06, + "loss": 0.45, + "step": 16474 + }, + { + "epoch": 0.5196341271092888, + "grad_norm": 1.7792236804962158, + "learning_rate": 9.851369004588744e-06, + "loss": 0.4542, + "step": 16475 + }, + { + "epoch": 0.5196656678757294, + "grad_norm": 1.4539790153503418, + "learning_rate": 9.850347561725737e-06, + "loss": 0.4523, + "step": 16476 + }, + { + "epoch": 0.51969720864217, + "grad_norm": 4.789838790893555, + "learning_rate": 9.849326120424465e-06, + "loss": 0.4875, + "step": 16477 + }, + { + "epoch": 0.5197287494086106, + "grad_norm": 1.4768933057785034, + "learning_rate": 9.848304680695596e-06, + "loss": 0.3735, + "step": 16478 + }, + { + "epoch": 0.5197602901750512, + "grad_norm": 1.9787852764129639, + "learning_rate": 9.847283242549785e-06, + "loss": 0.4596, + "step": 16479 + }, + { + "epoch": 0.5197918309414918, + "grad_norm": 1.3149796724319458, + "learning_rate": 9.846261805997688e-06, + "loss": 0.4886, + "step": 16480 + }, + { + "epoch": 0.5198233717079325, + "grad_norm": 1.5884050130844116, + "learning_rate": 9.845240371049973e-06, + "loss": 0.4589, + "step": 16481 + }, + { + "epoch": 0.5198549124743731, + "grad_norm": 1.698035478591919, + "learning_rate": 9.84421893771729e-06, + "loss": 0.4798, + "step": 16482 + }, + { + "epoch": 0.5198864532408137, + "grad_norm": 1.5486656427383423, + "learning_rate": 9.843197506010308e-06, + "loss": 0.4643, + "step": 16483 + }, + { + "epoch": 0.5199179940072544, + "grad_norm": 1.5646246671676636, + "learning_rate": 9.842176075939682e-06, + "loss": 0.4843, + "step": 16484 + }, + { + "epoch": 0.519949534773695, + "grad_norm": 1.736298680305481, + "learning_rate": 9.841154647516069e-06, + "loss": 0.4298, + "step": 16485 + }, + { + "epoch": 0.5199810755401356, + "grad_norm": 1.820216178894043, + "learning_rate": 9.84013322075013e-06, + "loss": 0.7644, + "step": 16486 + }, + { + "epoch": 0.5200126163065762, + "grad_norm": 2.393439769744873, + "learning_rate": 9.839111795652527e-06, + "loss": 0.4742, + "step": 16487 + }, + { + "epoch": 0.5200441570730169, + "grad_norm": 6.36582088470459, + "learning_rate": 9.838090372233917e-06, + "loss": 0.3963, + "step": 16488 + }, + { + "epoch": 0.5200756978394575, + "grad_norm": 2.397428512573242, + "learning_rate": 9.837068950504959e-06, + "loss": 0.4642, + "step": 16489 + }, + { + "epoch": 0.5201072386058981, + "grad_norm": 1.6943446397781372, + "learning_rate": 9.836047530476315e-06, + "loss": 0.5247, + "step": 16490 + }, + { + "epoch": 0.5201387793723388, + "grad_norm": 1.9010387659072876, + "learning_rate": 9.835026112158637e-06, + "loss": 0.4724, + "step": 16491 + }, + { + "epoch": 0.5201703201387794, + "grad_norm": 1.8354488611221313, + "learning_rate": 9.834004695562596e-06, + "loss": 0.472, + "step": 16492 + }, + { + "epoch": 0.52020186090522, + "grad_norm": 2.151947259902954, + "learning_rate": 9.832983280698842e-06, + "loss": 0.4841, + "step": 16493 + }, + { + "epoch": 0.5202334016716607, + "grad_norm": 1.830817699432373, + "learning_rate": 9.831961867578035e-06, + "loss": 0.4009, + "step": 16494 + }, + { + "epoch": 0.5202649424381013, + "grad_norm": 1.6350537538528442, + "learning_rate": 9.83094045621084e-06, + "loss": 0.7616, + "step": 16495 + }, + { + "epoch": 0.5202964832045419, + "grad_norm": 1.5471091270446777, + "learning_rate": 9.829919046607913e-06, + "loss": 0.4634, + "step": 16496 + }, + { + "epoch": 0.5203280239709825, + "grad_norm": 1.4634661674499512, + "learning_rate": 9.828897638779909e-06, + "loss": 0.449, + "step": 16497 + }, + { + "epoch": 0.5203595647374232, + "grad_norm": 1.4752581119537354, + "learning_rate": 9.827876232737494e-06, + "loss": 0.4893, + "step": 16498 + }, + { + "epoch": 0.5203911055038637, + "grad_norm": 1.6144678592681885, + "learning_rate": 9.826854828491321e-06, + "loss": 0.4695, + "step": 16499 + }, + { + "epoch": 0.5204226462703043, + "grad_norm": 1.732252836227417, + "learning_rate": 9.825833426052057e-06, + "loss": 0.4669, + "step": 16500 + }, + { + "epoch": 0.520454187036745, + "grad_norm": 1.9302300214767456, + "learning_rate": 9.824812025430354e-06, + "loss": 0.455, + "step": 16501 + }, + { + "epoch": 0.5204857278031856, + "grad_norm": 1.646835207939148, + "learning_rate": 9.82379062663687e-06, + "loss": 0.4422, + "step": 16502 + }, + { + "epoch": 0.5205172685696262, + "grad_norm": 1.6408100128173828, + "learning_rate": 9.822769229682273e-06, + "loss": 0.4683, + "step": 16503 + }, + { + "epoch": 0.5205488093360668, + "grad_norm": 1.5421638488769531, + "learning_rate": 9.821747834577216e-06, + "loss": 0.4221, + "step": 16504 + }, + { + "epoch": 0.5205803501025075, + "grad_norm": 2.349630355834961, + "learning_rate": 9.820726441332357e-06, + "loss": 0.4722, + "step": 16505 + }, + { + "epoch": 0.5206118908689481, + "grad_norm": 1.3969154357910156, + "learning_rate": 9.819705049958356e-06, + "loss": 0.3326, + "step": 16506 + }, + { + "epoch": 0.5206434316353887, + "grad_norm": 1.9621570110321045, + "learning_rate": 9.818683660465875e-06, + "loss": 0.7645, + "step": 16507 + }, + { + "epoch": 0.5206749724018294, + "grad_norm": 1.7144348621368408, + "learning_rate": 9.817662272865569e-06, + "loss": 0.3871, + "step": 16508 + }, + { + "epoch": 0.52070651316827, + "grad_norm": 2.041891574859619, + "learning_rate": 9.816640887168102e-06, + "loss": 0.4084, + "step": 16509 + }, + { + "epoch": 0.5207380539347106, + "grad_norm": 1.4905259609222412, + "learning_rate": 9.815619503384128e-06, + "loss": 0.7545, + "step": 16510 + }, + { + "epoch": 0.5207695947011513, + "grad_norm": 2.0283091068267822, + "learning_rate": 9.814598121524305e-06, + "loss": 0.5133, + "step": 16511 + }, + { + "epoch": 0.5208011354675919, + "grad_norm": 1.8668575286865234, + "learning_rate": 9.813576741599298e-06, + "loss": 0.4716, + "step": 16512 + }, + { + "epoch": 0.5208326762340325, + "grad_norm": 1.6137925386428833, + "learning_rate": 9.812555363619757e-06, + "loss": 0.4263, + "step": 16513 + }, + { + "epoch": 0.5208642170004731, + "grad_norm": 1.9859325885772705, + "learning_rate": 9.811533987596352e-06, + "loss": 0.4626, + "step": 16514 + }, + { + "epoch": 0.5208957577669138, + "grad_norm": 1.522400140762329, + "learning_rate": 9.810512613539735e-06, + "loss": 0.4467, + "step": 16515 + }, + { + "epoch": 0.5209272985333544, + "grad_norm": 1.5824257135391235, + "learning_rate": 9.809491241460565e-06, + "loss": 0.4528, + "step": 16516 + }, + { + "epoch": 0.520958839299795, + "grad_norm": 1.6051418781280518, + "learning_rate": 9.808469871369504e-06, + "loss": 0.4655, + "step": 16517 + }, + { + "epoch": 0.5209903800662357, + "grad_norm": 1.7816884517669678, + "learning_rate": 9.807448503277207e-06, + "loss": 0.4843, + "step": 16518 + }, + { + "epoch": 0.5210219208326763, + "grad_norm": 1.9109832048416138, + "learning_rate": 9.806427137194333e-06, + "loss": 0.4377, + "step": 16519 + }, + { + "epoch": 0.5210534615991168, + "grad_norm": 1.8275318145751953, + "learning_rate": 9.805405773131545e-06, + "loss": 0.4852, + "step": 16520 + }, + { + "epoch": 0.5210850023655574, + "grad_norm": 1.6472768783569336, + "learning_rate": 9.8043844110995e-06, + "loss": 0.4869, + "step": 16521 + }, + { + "epoch": 0.5211165431319981, + "grad_norm": 2.0488758087158203, + "learning_rate": 9.803363051108849e-06, + "loss": 0.5158, + "step": 16522 + }, + { + "epoch": 0.5211480838984387, + "grad_norm": 1.4371788501739502, + "learning_rate": 9.802341693170263e-06, + "loss": 0.485, + "step": 16523 + }, + { + "epoch": 0.5211796246648793, + "grad_norm": 1.5056755542755127, + "learning_rate": 9.801320337294393e-06, + "loss": 0.4538, + "step": 16524 + }, + { + "epoch": 0.52121116543132, + "grad_norm": 1.6234301328659058, + "learning_rate": 9.8002989834919e-06, + "loss": 0.4989, + "step": 16525 + }, + { + "epoch": 0.5212427061977606, + "grad_norm": 1.5357015132904053, + "learning_rate": 9.79927763177344e-06, + "loss": 0.4874, + "step": 16526 + }, + { + "epoch": 0.5212742469642012, + "grad_norm": 1.5154423713684082, + "learning_rate": 9.798256282149677e-06, + "loss": 0.4961, + "step": 16527 + }, + { + "epoch": 0.5213057877306418, + "grad_norm": 1.6616326570510864, + "learning_rate": 9.797234934631265e-06, + "loss": 0.3639, + "step": 16528 + }, + { + "epoch": 0.5213373284970825, + "grad_norm": 1.608949065208435, + "learning_rate": 9.796213589228864e-06, + "loss": 0.4625, + "step": 16529 + }, + { + "epoch": 0.5213688692635231, + "grad_norm": 1.6362271308898926, + "learning_rate": 9.79519224595313e-06, + "loss": 0.4281, + "step": 16530 + }, + { + "epoch": 0.5214004100299637, + "grad_norm": 1.6751210689544678, + "learning_rate": 9.794170904814728e-06, + "loss": 0.481, + "step": 16531 + }, + { + "epoch": 0.5214319507964044, + "grad_norm": 1.517991065979004, + "learning_rate": 9.793149565824311e-06, + "loss": 0.468, + "step": 16532 + }, + { + "epoch": 0.521463491562845, + "grad_norm": 1.7858256101608276, + "learning_rate": 9.792128228992536e-06, + "loss": 0.4597, + "step": 16533 + }, + { + "epoch": 0.5214950323292856, + "grad_norm": 3.3375296592712402, + "learning_rate": 9.791106894330068e-06, + "loss": 0.4796, + "step": 16534 + }, + { + "epoch": 0.5215265730957263, + "grad_norm": 1.7989580631256104, + "learning_rate": 9.790085561847562e-06, + "loss": 0.3779, + "step": 16535 + }, + { + "epoch": 0.5215581138621669, + "grad_norm": 1.7329121828079224, + "learning_rate": 9.789064231555674e-06, + "loss": 0.4774, + "step": 16536 + }, + { + "epoch": 0.5215896546286075, + "grad_norm": 3.064680337905884, + "learning_rate": 9.788042903465066e-06, + "loss": 0.8145, + "step": 16537 + }, + { + "epoch": 0.5216211953950481, + "grad_norm": 1.4094542264938354, + "learning_rate": 9.787021577586397e-06, + "loss": 0.5021, + "step": 16538 + }, + { + "epoch": 0.5216527361614888, + "grad_norm": 1.6535433530807495, + "learning_rate": 9.786000253930317e-06, + "loss": 0.4512, + "step": 16539 + }, + { + "epoch": 0.5216842769279294, + "grad_norm": 1.480217456817627, + "learning_rate": 9.784978932507494e-06, + "loss": 0.4823, + "step": 16540 + }, + { + "epoch": 0.5217158176943699, + "grad_norm": 1.7953943014144897, + "learning_rate": 9.783957613328585e-06, + "loss": 0.4635, + "step": 16541 + }, + { + "epoch": 0.5217473584608106, + "grad_norm": 1.6526992321014404, + "learning_rate": 9.782936296404241e-06, + "loss": 0.4203, + "step": 16542 + }, + { + "epoch": 0.5217788992272512, + "grad_norm": 1.8623603582382202, + "learning_rate": 9.781914981745128e-06, + "loss": 0.4528, + "step": 16543 + }, + { + "epoch": 0.5218104399936918, + "grad_norm": 1.897473692893982, + "learning_rate": 9.780893669361904e-06, + "loss": 0.4886, + "step": 16544 + }, + { + "epoch": 0.5218419807601324, + "grad_norm": 1.4552793502807617, + "learning_rate": 9.77987235926522e-06, + "loss": 0.4888, + "step": 16545 + }, + { + "epoch": 0.5218735215265731, + "grad_norm": 1.9431185722351074, + "learning_rate": 9.778851051465742e-06, + "loss": 0.3263, + "step": 16546 + }, + { + "epoch": 0.5219050622930137, + "grad_norm": 1.572520136833191, + "learning_rate": 9.777829745974123e-06, + "loss": 0.7562, + "step": 16547 + }, + { + "epoch": 0.5219366030594543, + "grad_norm": 2.183436632156372, + "learning_rate": 9.776808442801026e-06, + "loss": 0.4978, + "step": 16548 + }, + { + "epoch": 0.521968143825895, + "grad_norm": 1.3206003904342651, + "learning_rate": 9.775787141957104e-06, + "loss": 0.4541, + "step": 16549 + }, + { + "epoch": 0.5219996845923356, + "grad_norm": 2.505033016204834, + "learning_rate": 9.774765843453016e-06, + "loss": 0.4969, + "step": 16550 + }, + { + "epoch": 0.5220312253587762, + "grad_norm": 4.493958473205566, + "learning_rate": 9.773744547299424e-06, + "loss": 0.4731, + "step": 16551 + }, + { + "epoch": 0.5220627661252168, + "grad_norm": 1.756358027458191, + "learning_rate": 9.772723253506983e-06, + "loss": 0.454, + "step": 16552 + }, + { + "epoch": 0.5220943068916575, + "grad_norm": 2.254417896270752, + "learning_rate": 9.771701962086348e-06, + "loss": 0.4717, + "step": 16553 + }, + { + "epoch": 0.5221258476580981, + "grad_norm": 1.4676947593688965, + "learning_rate": 9.770680673048185e-06, + "loss": 0.7771, + "step": 16554 + }, + { + "epoch": 0.5221573884245387, + "grad_norm": 2.4082257747650146, + "learning_rate": 9.769659386403146e-06, + "loss": 0.4681, + "step": 16555 + }, + { + "epoch": 0.5221889291909794, + "grad_norm": 1.743548035621643, + "learning_rate": 9.76863810216189e-06, + "loss": 0.4337, + "step": 16556 + }, + { + "epoch": 0.52222046995742, + "grad_norm": 1.628605842590332, + "learning_rate": 9.767616820335075e-06, + "loss": 0.4053, + "step": 16557 + }, + { + "epoch": 0.5222520107238606, + "grad_norm": 2.9994397163391113, + "learning_rate": 9.766595540933358e-06, + "loss": 0.7724, + "step": 16558 + }, + { + "epoch": 0.5222835514903013, + "grad_norm": 1.9151097536087036, + "learning_rate": 9.765574263967397e-06, + "loss": 0.4739, + "step": 16559 + }, + { + "epoch": 0.5223150922567419, + "grad_norm": 1.770430088043213, + "learning_rate": 9.764552989447853e-06, + "loss": 0.4842, + "step": 16560 + }, + { + "epoch": 0.5223466330231825, + "grad_norm": 1.8678386211395264, + "learning_rate": 9.763531717385377e-06, + "loss": 0.4754, + "step": 16561 + }, + { + "epoch": 0.522378173789623, + "grad_norm": 1.4765253067016602, + "learning_rate": 9.762510447790635e-06, + "loss": 0.4591, + "step": 16562 + }, + { + "epoch": 0.5224097145560637, + "grad_norm": 5.697661876678467, + "learning_rate": 9.761489180674282e-06, + "loss": 0.4626, + "step": 16563 + }, + { + "epoch": 0.5224412553225043, + "grad_norm": 2.0077614784240723, + "learning_rate": 9.760467916046971e-06, + "loss": 0.4428, + "step": 16564 + }, + { + "epoch": 0.5224727960889449, + "grad_norm": 1.563618779182434, + "learning_rate": 9.759446653919367e-06, + "loss": 0.4716, + "step": 16565 + }, + { + "epoch": 0.5225043368553856, + "grad_norm": 2.2329511642456055, + "learning_rate": 9.75842539430212e-06, + "loss": 0.4769, + "step": 16566 + }, + { + "epoch": 0.5225358776218262, + "grad_norm": 2.678990602493286, + "learning_rate": 9.757404137205893e-06, + "loss": 0.4746, + "step": 16567 + }, + { + "epoch": 0.5225674183882668, + "grad_norm": 2.2962610721588135, + "learning_rate": 9.756382882641344e-06, + "loss": 0.4221, + "step": 16568 + }, + { + "epoch": 0.5225989591547074, + "grad_norm": 2.1904115676879883, + "learning_rate": 9.75536163061913e-06, + "loss": 0.4574, + "step": 16569 + }, + { + "epoch": 0.5226304999211481, + "grad_norm": 2.0809102058410645, + "learning_rate": 9.7543403811499e-06, + "loss": 0.5059, + "step": 16570 + }, + { + "epoch": 0.5226620406875887, + "grad_norm": 4.817444801330566, + "learning_rate": 9.753319134244326e-06, + "loss": 0.4535, + "step": 16571 + }, + { + "epoch": 0.5226935814540293, + "grad_norm": 2.9719390869140625, + "learning_rate": 9.752297889913057e-06, + "loss": 0.4733, + "step": 16572 + }, + { + "epoch": 0.52272512222047, + "grad_norm": 2.663482666015625, + "learning_rate": 9.751276648166749e-06, + "loss": 0.49, + "step": 16573 + }, + { + "epoch": 0.5227566629869106, + "grad_norm": 1.9001129865646362, + "learning_rate": 9.750255409016065e-06, + "loss": 0.4903, + "step": 16574 + }, + { + "epoch": 0.5227882037533512, + "grad_norm": 2.395108461380005, + "learning_rate": 9.74923417247166e-06, + "loss": 0.4787, + "step": 16575 + }, + { + "epoch": 0.5228197445197919, + "grad_norm": 4.859389305114746, + "learning_rate": 9.748212938544188e-06, + "loss": 0.4801, + "step": 16576 + }, + { + "epoch": 0.5228512852862325, + "grad_norm": 2.4248907566070557, + "learning_rate": 9.747191707244314e-06, + "loss": 0.4193, + "step": 16577 + }, + { + "epoch": 0.5228828260526731, + "grad_norm": 2.879934549331665, + "learning_rate": 9.746170478582686e-06, + "loss": 0.4572, + "step": 16578 + }, + { + "epoch": 0.5229143668191137, + "grad_norm": 2.2707393169403076, + "learning_rate": 9.74514925256997e-06, + "loss": 0.4856, + "step": 16579 + }, + { + "epoch": 0.5229459075855544, + "grad_norm": 1.8228293657302856, + "learning_rate": 9.74412802921682e-06, + "loss": 0.4834, + "step": 16580 + }, + { + "epoch": 0.522977448351995, + "grad_norm": 3.0414459705352783, + "learning_rate": 9.743106808533887e-06, + "loss": 0.4776, + "step": 16581 + }, + { + "epoch": 0.5230089891184356, + "grad_norm": 1.714282751083374, + "learning_rate": 9.742085590531838e-06, + "loss": 0.4496, + "step": 16582 + }, + { + "epoch": 0.5230405298848761, + "grad_norm": 5.212671279907227, + "learning_rate": 9.741064375221327e-06, + "loss": 0.4908, + "step": 16583 + }, + { + "epoch": 0.5230720706513168, + "grad_norm": 1.8907134532928467, + "learning_rate": 9.740043162613008e-06, + "loss": 0.4484, + "step": 16584 + }, + { + "epoch": 0.5231036114177574, + "grad_norm": 2.175790786743164, + "learning_rate": 9.739021952717545e-06, + "loss": 0.452, + "step": 16585 + }, + { + "epoch": 0.523135152184198, + "grad_norm": 1.6346203088760376, + "learning_rate": 9.738000745545585e-06, + "loss": 0.493, + "step": 16586 + }, + { + "epoch": 0.5231666929506387, + "grad_norm": 1.894073724746704, + "learning_rate": 9.736979541107792e-06, + "loss": 0.3575, + "step": 16587 + }, + { + "epoch": 0.5231982337170793, + "grad_norm": 2.3277671337127686, + "learning_rate": 9.735958339414824e-06, + "loss": 0.4825, + "step": 16588 + }, + { + "epoch": 0.5232297744835199, + "grad_norm": 1.7343389987945557, + "learning_rate": 9.734937140477335e-06, + "loss": 0.4705, + "step": 16589 + }, + { + "epoch": 0.5232613152499606, + "grad_norm": 1.7806631326675415, + "learning_rate": 9.73391594430598e-06, + "loss": 0.4815, + "step": 16590 + }, + { + "epoch": 0.5232928560164012, + "grad_norm": 4.266660213470459, + "learning_rate": 9.732894750911421e-06, + "loss": 0.4749, + "step": 16591 + }, + { + "epoch": 0.5233243967828418, + "grad_norm": 2.5311052799224854, + "learning_rate": 9.731873560304314e-06, + "loss": 0.5037, + "step": 16592 + }, + { + "epoch": 0.5233559375492824, + "grad_norm": 2.7438299655914307, + "learning_rate": 9.73085237249531e-06, + "loss": 0.4527, + "step": 16593 + }, + { + "epoch": 0.5233874783157231, + "grad_norm": 5.711892127990723, + "learning_rate": 9.729831187495074e-06, + "loss": 0.484, + "step": 16594 + }, + { + "epoch": 0.5234190190821637, + "grad_norm": 1.2995500564575195, + "learning_rate": 9.728810005314259e-06, + "loss": 0.471, + "step": 16595 + }, + { + "epoch": 0.5234505598486043, + "grad_norm": 1.6080769300460815, + "learning_rate": 9.727788825963524e-06, + "loss": 0.4941, + "step": 16596 + }, + { + "epoch": 0.523482100615045, + "grad_norm": 3.9983887672424316, + "learning_rate": 9.726767649453523e-06, + "loss": 0.4722, + "step": 16597 + }, + { + "epoch": 0.5235136413814856, + "grad_norm": 2.5868327617645264, + "learning_rate": 9.72574647579491e-06, + "loss": 0.3243, + "step": 16598 + }, + { + "epoch": 0.5235451821479262, + "grad_norm": 2.9649033546447754, + "learning_rate": 9.724725304998348e-06, + "loss": 0.4675, + "step": 16599 + }, + { + "epoch": 0.5235767229143669, + "grad_norm": 3.5122196674346924, + "learning_rate": 9.723704137074494e-06, + "loss": 0.4386, + "step": 16600 + }, + { + "epoch": 0.5236082636808075, + "grad_norm": 4.6049485206604, + "learning_rate": 9.722682972033995e-06, + "loss": 0.4257, + "step": 16601 + }, + { + "epoch": 0.5236398044472481, + "grad_norm": 2.1203269958496094, + "learning_rate": 9.72166180988752e-06, + "loss": 0.4632, + "step": 16602 + }, + { + "epoch": 0.5236713452136887, + "grad_norm": 3.0077364444732666, + "learning_rate": 9.72064065064572e-06, + "loss": 0.4609, + "step": 16603 + }, + { + "epoch": 0.5237028859801293, + "grad_norm": 1.98800790309906, + "learning_rate": 9.71961949431925e-06, + "loss": 0.3468, + "step": 16604 + }, + { + "epoch": 0.5237344267465699, + "grad_norm": 2.1141839027404785, + "learning_rate": 9.71859834091877e-06, + "loss": 0.4484, + "step": 16605 + }, + { + "epoch": 0.5237659675130105, + "grad_norm": 3.271010637283325, + "learning_rate": 9.717577190454933e-06, + "loss": 0.754, + "step": 16606 + }, + { + "epoch": 0.5237975082794512, + "grad_norm": 2.7837307453155518, + "learning_rate": 9.716556042938398e-06, + "loss": 0.3599, + "step": 16607 + }, + { + "epoch": 0.5238290490458918, + "grad_norm": 2.4017531871795654, + "learning_rate": 9.715534898379822e-06, + "loss": 0.4578, + "step": 16608 + }, + { + "epoch": 0.5238605898123324, + "grad_norm": 1.8312426805496216, + "learning_rate": 9.714513756789854e-06, + "loss": 0.4293, + "step": 16609 + }, + { + "epoch": 0.523892130578773, + "grad_norm": 1.7768080234527588, + "learning_rate": 9.713492618179164e-06, + "loss": 0.4466, + "step": 16610 + }, + { + "epoch": 0.5239236713452137, + "grad_norm": 1.9151949882507324, + "learning_rate": 9.7124714825584e-06, + "loss": 0.4748, + "step": 16611 + }, + { + "epoch": 0.5239552121116543, + "grad_norm": 5.2770538330078125, + "learning_rate": 9.711450349938215e-06, + "loss": 0.4619, + "step": 16612 + }, + { + "epoch": 0.5239867528780949, + "grad_norm": 4.26162052154541, + "learning_rate": 9.710429220329274e-06, + "loss": 0.4588, + "step": 16613 + }, + { + "epoch": 0.5240182936445356, + "grad_norm": 2.0010526180267334, + "learning_rate": 9.709408093742228e-06, + "loss": 0.4841, + "step": 16614 + }, + { + "epoch": 0.5240498344109762, + "grad_norm": 1.6309767961502075, + "learning_rate": 9.708386970187735e-06, + "loss": 0.4505, + "step": 16615 + }, + { + "epoch": 0.5240813751774168, + "grad_norm": 4.39418888092041, + "learning_rate": 9.70736584967645e-06, + "loss": 0.5201, + "step": 16616 + }, + { + "epoch": 0.5241129159438574, + "grad_norm": 2.6502089500427246, + "learning_rate": 9.70634473221903e-06, + "loss": 0.484, + "step": 16617 + }, + { + "epoch": 0.5241444567102981, + "grad_norm": 2.3407888412475586, + "learning_rate": 9.705323617826126e-06, + "loss": 0.4972, + "step": 16618 + }, + { + "epoch": 0.5241759974767387, + "grad_norm": 3.0519180297851562, + "learning_rate": 9.704302506508404e-06, + "loss": 0.4669, + "step": 16619 + }, + { + "epoch": 0.5242075382431793, + "grad_norm": 2.164652109146118, + "learning_rate": 9.703281398276513e-06, + "loss": 0.441, + "step": 16620 + }, + { + "epoch": 0.52423907900962, + "grad_norm": 2.0734081268310547, + "learning_rate": 9.702260293141109e-06, + "loss": 0.4851, + "step": 16621 + }, + { + "epoch": 0.5242706197760606, + "grad_norm": 2.0317749977111816, + "learning_rate": 9.701239191112854e-06, + "loss": 0.7907, + "step": 16622 + }, + { + "epoch": 0.5243021605425012, + "grad_norm": 2.0000498294830322, + "learning_rate": 9.7002180922024e-06, + "loss": 0.4809, + "step": 16623 + }, + { + "epoch": 0.5243337013089419, + "grad_norm": 2.872396945953369, + "learning_rate": 9.6991969964204e-06, + "loss": 0.4592, + "step": 16624 + }, + { + "epoch": 0.5243652420753825, + "grad_norm": 1.6867731809616089, + "learning_rate": 9.698175903777515e-06, + "loss": 0.4664, + "step": 16625 + }, + { + "epoch": 0.524396782841823, + "grad_norm": 1.4169059991836548, + "learning_rate": 9.697154814284395e-06, + "loss": 0.7688, + "step": 16626 + }, + { + "epoch": 0.5244283236082636, + "grad_norm": 1.6376858949661255, + "learning_rate": 9.696133727951704e-06, + "loss": 0.499, + "step": 16627 + }, + { + "epoch": 0.5244598643747043, + "grad_norm": 1.8944321870803833, + "learning_rate": 9.695112644790093e-06, + "loss": 0.4415, + "step": 16628 + }, + { + "epoch": 0.5244914051411449, + "grad_norm": 1.5992027521133423, + "learning_rate": 9.694091564810215e-06, + "loss": 0.4571, + "step": 16629 + }, + { + "epoch": 0.5245229459075855, + "grad_norm": 2.5923705101013184, + "learning_rate": 9.693070488022734e-06, + "loss": 0.4881, + "step": 16630 + }, + { + "epoch": 0.5245544866740262, + "grad_norm": 2.5466971397399902, + "learning_rate": 9.692049414438298e-06, + "loss": 0.458, + "step": 16631 + }, + { + "epoch": 0.5245860274404668, + "grad_norm": 1.9185985326766968, + "learning_rate": 9.691028344067564e-06, + "loss": 0.4886, + "step": 16632 + }, + { + "epoch": 0.5246175682069074, + "grad_norm": 2.589388370513916, + "learning_rate": 9.690007276921192e-06, + "loss": 0.4958, + "step": 16633 + }, + { + "epoch": 0.524649108973348, + "grad_norm": 1.9945666790008545, + "learning_rate": 9.688986213009835e-06, + "loss": 0.4621, + "step": 16634 + }, + { + "epoch": 0.5246806497397887, + "grad_norm": 3.4972293376922607, + "learning_rate": 9.687965152344146e-06, + "loss": 0.4551, + "step": 16635 + }, + { + "epoch": 0.5247121905062293, + "grad_norm": 1.5878922939300537, + "learning_rate": 9.686944094934784e-06, + "loss": 0.4591, + "step": 16636 + }, + { + "epoch": 0.5247437312726699, + "grad_norm": 1.4913674592971802, + "learning_rate": 9.685923040792405e-06, + "loss": 0.4983, + "step": 16637 + }, + { + "epoch": 0.5247752720391106, + "grad_norm": 1.8721641302108765, + "learning_rate": 9.68490198992766e-06, + "loss": 0.4954, + "step": 16638 + }, + { + "epoch": 0.5248068128055512, + "grad_norm": 1.8517582416534424, + "learning_rate": 9.683880942351209e-06, + "loss": 0.4438, + "step": 16639 + }, + { + "epoch": 0.5248383535719918, + "grad_norm": 1.6446259021759033, + "learning_rate": 9.682859898073704e-06, + "loss": 0.4577, + "step": 16640 + }, + { + "epoch": 0.5248698943384325, + "grad_norm": 1.7155627012252808, + "learning_rate": 9.681838857105804e-06, + "loss": 0.4569, + "step": 16641 + }, + { + "epoch": 0.5249014351048731, + "grad_norm": 1.8258155584335327, + "learning_rate": 9.680817819458164e-06, + "loss": 0.3516, + "step": 16642 + }, + { + "epoch": 0.5249329758713137, + "grad_norm": 2.2398242950439453, + "learning_rate": 9.679796785141435e-06, + "loss": 0.4379, + "step": 16643 + }, + { + "epoch": 0.5249645166377543, + "grad_norm": 2.3044629096984863, + "learning_rate": 9.678775754166277e-06, + "loss": 0.4675, + "step": 16644 + }, + { + "epoch": 0.524996057404195, + "grad_norm": 1.6594868898391724, + "learning_rate": 9.677754726543344e-06, + "loss": 0.48, + "step": 16645 + }, + { + "epoch": 0.5250275981706356, + "grad_norm": 2.0071768760681152, + "learning_rate": 9.676733702283287e-06, + "loss": 0.5055, + "step": 16646 + }, + { + "epoch": 0.5250591389370761, + "grad_norm": 1.9400862455368042, + "learning_rate": 9.67571268139677e-06, + "loss": 0.4741, + "step": 16647 + }, + { + "epoch": 0.5250906797035167, + "grad_norm": 3.3448140621185303, + "learning_rate": 9.67469166389444e-06, + "loss": 0.7468, + "step": 16648 + }, + { + "epoch": 0.5251222204699574, + "grad_norm": 1.629868745803833, + "learning_rate": 9.673670649786953e-06, + "loss": 0.4665, + "step": 16649 + }, + { + "epoch": 0.525153761236398, + "grad_norm": 1.9094617366790771, + "learning_rate": 9.67264963908497e-06, + "loss": 0.5025, + "step": 16650 + }, + { + "epoch": 0.5251853020028386, + "grad_norm": 1.5908488035202026, + "learning_rate": 9.671628631799142e-06, + "loss": 0.503, + "step": 16651 + }, + { + "epoch": 0.5252168427692793, + "grad_norm": 1.9723987579345703, + "learning_rate": 9.67060762794012e-06, + "loss": 0.4638, + "step": 16652 + }, + { + "epoch": 0.5252483835357199, + "grad_norm": 2.1948769092559814, + "learning_rate": 9.669586627518568e-06, + "loss": 0.4741, + "step": 16653 + }, + { + "epoch": 0.5252799243021605, + "grad_norm": 1.5044080018997192, + "learning_rate": 9.668565630545134e-06, + "loss": 0.4756, + "step": 16654 + }, + { + "epoch": 0.5253114650686012, + "grad_norm": 2.0895683765411377, + "learning_rate": 9.667544637030475e-06, + "loss": 0.4394, + "step": 16655 + }, + { + "epoch": 0.5253430058350418, + "grad_norm": 2.076462507247925, + "learning_rate": 9.666523646985247e-06, + "loss": 0.4748, + "step": 16656 + }, + { + "epoch": 0.5253745466014824, + "grad_norm": 2.5897693634033203, + "learning_rate": 9.665502660420101e-06, + "loss": 0.4737, + "step": 16657 + }, + { + "epoch": 0.525406087367923, + "grad_norm": 2.405092716217041, + "learning_rate": 9.664481677345697e-06, + "loss": 0.4935, + "step": 16658 + }, + { + "epoch": 0.5254376281343637, + "grad_norm": 1.95893394947052, + "learning_rate": 9.663460697772688e-06, + "loss": 0.4904, + "step": 16659 + }, + { + "epoch": 0.5254691689008043, + "grad_norm": 1.8280248641967773, + "learning_rate": 9.662439721711723e-06, + "loss": 0.4488, + "step": 16660 + }, + { + "epoch": 0.5255007096672449, + "grad_norm": 1.6153452396392822, + "learning_rate": 9.661418749173467e-06, + "loss": 0.4522, + "step": 16661 + }, + { + "epoch": 0.5255322504336856, + "grad_norm": 3.0562257766723633, + "learning_rate": 9.660397780168568e-06, + "loss": 0.3405, + "step": 16662 + }, + { + "epoch": 0.5255637912001262, + "grad_norm": 2.2976999282836914, + "learning_rate": 9.65937681470768e-06, + "loss": 0.4754, + "step": 16663 + }, + { + "epoch": 0.5255953319665668, + "grad_norm": 1.4778543710708618, + "learning_rate": 9.658355852801463e-06, + "loss": 0.3406, + "step": 16664 + }, + { + "epoch": 0.5256268727330075, + "grad_norm": 1.7740042209625244, + "learning_rate": 9.657334894460566e-06, + "loss": 0.4591, + "step": 16665 + }, + { + "epoch": 0.5256584134994481, + "grad_norm": 1.5227041244506836, + "learning_rate": 9.656313939695641e-06, + "loss": 0.4673, + "step": 16666 + }, + { + "epoch": 0.5256899542658887, + "grad_norm": 1.9692195653915405, + "learning_rate": 9.655292988517353e-06, + "loss": 0.445, + "step": 16667 + }, + { + "epoch": 0.5257214950323292, + "grad_norm": 2.050593852996826, + "learning_rate": 9.65427204093635e-06, + "loss": 0.4906, + "step": 16668 + }, + { + "epoch": 0.5257530357987699, + "grad_norm": 1.5876377820968628, + "learning_rate": 9.65325109696328e-06, + "loss": 0.4864, + "step": 16669 + }, + { + "epoch": 0.5257845765652105, + "grad_norm": 1.5796473026275635, + "learning_rate": 9.652230156608812e-06, + "loss": 0.4626, + "step": 16670 + }, + { + "epoch": 0.5258161173316511, + "grad_norm": 1.9149285554885864, + "learning_rate": 9.65120921988359e-06, + "loss": 0.4924, + "step": 16671 + }, + { + "epoch": 0.5258476580980918, + "grad_norm": 2.453850507736206, + "learning_rate": 9.650188286798268e-06, + "loss": 0.4767, + "step": 16672 + }, + { + "epoch": 0.5258791988645324, + "grad_norm": 3.2502951622009277, + "learning_rate": 9.649167357363504e-06, + "loss": 0.477, + "step": 16673 + }, + { + "epoch": 0.525910739630973, + "grad_norm": 1.8617238998413086, + "learning_rate": 9.648146431589952e-06, + "loss": 0.474, + "step": 16674 + }, + { + "epoch": 0.5259422803974136, + "grad_norm": 1.4769819974899292, + "learning_rate": 9.647125509488266e-06, + "loss": 0.4579, + "step": 16675 + }, + { + "epoch": 0.5259738211638543, + "grad_norm": 1.8178369998931885, + "learning_rate": 9.6461045910691e-06, + "loss": 0.5022, + "step": 16676 + }, + { + "epoch": 0.5260053619302949, + "grad_norm": 3.6891911029815674, + "learning_rate": 9.645083676343102e-06, + "loss": 0.4833, + "step": 16677 + }, + { + "epoch": 0.5260369026967355, + "grad_norm": 1.8026665449142456, + "learning_rate": 9.644062765320937e-06, + "loss": 0.509, + "step": 16678 + }, + { + "epoch": 0.5260684434631762, + "grad_norm": 2.1603968143463135, + "learning_rate": 9.643041858013253e-06, + "loss": 0.4678, + "step": 16679 + }, + { + "epoch": 0.5260999842296168, + "grad_norm": 1.8884140253067017, + "learning_rate": 9.642020954430699e-06, + "loss": 0.4971, + "step": 16680 + }, + { + "epoch": 0.5261315249960574, + "grad_norm": 1.643119215965271, + "learning_rate": 9.64100005458394e-06, + "loss": 0.4615, + "step": 16681 + }, + { + "epoch": 0.526163065762498, + "grad_norm": 1.4644306898117065, + "learning_rate": 9.639979158483624e-06, + "loss": 0.4975, + "step": 16682 + }, + { + "epoch": 0.5261946065289387, + "grad_norm": 2.7841873168945312, + "learning_rate": 9.638958266140402e-06, + "loss": 0.4898, + "step": 16683 + }, + { + "epoch": 0.5262261472953793, + "grad_norm": 21.56372833251953, + "learning_rate": 9.637937377564934e-06, + "loss": 0.4643, + "step": 16684 + }, + { + "epoch": 0.5262576880618199, + "grad_norm": 1.688811182975769, + "learning_rate": 9.636916492767871e-06, + "loss": 0.4845, + "step": 16685 + }, + { + "epoch": 0.5262892288282606, + "grad_norm": 2.012782335281372, + "learning_rate": 9.63589561175986e-06, + "loss": 0.4962, + "step": 16686 + }, + { + "epoch": 0.5263207695947012, + "grad_norm": 1.9755696058273315, + "learning_rate": 9.634874734551567e-06, + "loss": 0.4616, + "step": 16687 + }, + { + "epoch": 0.5263523103611418, + "grad_norm": 1.4230825901031494, + "learning_rate": 9.633853861153634e-06, + "loss": 0.4406, + "step": 16688 + }, + { + "epoch": 0.5263838511275823, + "grad_norm": 2.0029985904693604, + "learning_rate": 9.632832991576726e-06, + "loss": 0.518, + "step": 16689 + }, + { + "epoch": 0.526415391894023, + "grad_norm": 2.2746164798736572, + "learning_rate": 9.63181212583149e-06, + "loss": 0.4686, + "step": 16690 + }, + { + "epoch": 0.5264469326604636, + "grad_norm": 1.6649807691574097, + "learning_rate": 9.630791263928578e-06, + "loss": 0.4576, + "step": 16691 + }, + { + "epoch": 0.5264784734269042, + "grad_norm": 1.5500043630599976, + "learning_rate": 9.629770405878647e-06, + "loss": 0.4759, + "step": 16692 + }, + { + "epoch": 0.5265100141933449, + "grad_norm": 1.543272614479065, + "learning_rate": 9.628749551692351e-06, + "loss": 0.4752, + "step": 16693 + }, + { + "epoch": 0.5265415549597855, + "grad_norm": 2.0100696086883545, + "learning_rate": 9.627728701380339e-06, + "loss": 0.7973, + "step": 16694 + }, + { + "epoch": 0.5265730957262261, + "grad_norm": 1.8018392324447632, + "learning_rate": 9.62670785495327e-06, + "loss": 0.4974, + "step": 16695 + }, + { + "epoch": 0.5266046364926668, + "grad_norm": 1.5976154804229736, + "learning_rate": 9.625687012421794e-06, + "loss": 0.4833, + "step": 16696 + }, + { + "epoch": 0.5266361772591074, + "grad_norm": 1.549659013748169, + "learning_rate": 9.62466617379656e-06, + "loss": 0.4541, + "step": 16697 + }, + { + "epoch": 0.526667718025548, + "grad_norm": 1.5861659049987793, + "learning_rate": 9.62364533908823e-06, + "loss": 0.4596, + "step": 16698 + }, + { + "epoch": 0.5266992587919886, + "grad_norm": 1.4795498847961426, + "learning_rate": 9.622624508307454e-06, + "loss": 0.4801, + "step": 16699 + }, + { + "epoch": 0.5267307995584293, + "grad_norm": 1.3666572570800781, + "learning_rate": 9.62160368146488e-06, + "loss": 0.4679, + "step": 16700 + }, + { + "epoch": 0.5267623403248699, + "grad_norm": 1.5018213987350464, + "learning_rate": 9.620582858571169e-06, + "loss": 0.4693, + "step": 16701 + }, + { + "epoch": 0.5267938810913105, + "grad_norm": 2.162773609161377, + "learning_rate": 9.61956203963697e-06, + "loss": 0.4989, + "step": 16702 + }, + { + "epoch": 0.5268254218577512, + "grad_norm": 2.3586244583129883, + "learning_rate": 9.618541224672936e-06, + "loss": 0.4638, + "step": 16703 + }, + { + "epoch": 0.5268569626241918, + "grad_norm": 1.8534400463104248, + "learning_rate": 9.617520413689721e-06, + "loss": 0.463, + "step": 16704 + }, + { + "epoch": 0.5268885033906324, + "grad_norm": 2.1633174419403076, + "learning_rate": 9.616499606697974e-06, + "loss": 0.4327, + "step": 16705 + }, + { + "epoch": 0.526920044157073, + "grad_norm": 1.3278907537460327, + "learning_rate": 9.615478803708355e-06, + "loss": 0.3532, + "step": 16706 + }, + { + "epoch": 0.5269515849235137, + "grad_norm": 1.4613070487976074, + "learning_rate": 9.614458004731515e-06, + "loss": 0.4894, + "step": 16707 + }, + { + "epoch": 0.5269831256899543, + "grad_norm": 2.996447801589966, + "learning_rate": 9.613437209778099e-06, + "loss": 0.4973, + "step": 16708 + }, + { + "epoch": 0.5270146664563949, + "grad_norm": 2.510834217071533, + "learning_rate": 9.612416418858772e-06, + "loss": 0.4692, + "step": 16709 + }, + { + "epoch": 0.5270462072228355, + "grad_norm": 1.6531411409378052, + "learning_rate": 9.61139563198418e-06, + "loss": 0.4691, + "step": 16710 + }, + { + "epoch": 0.5270777479892761, + "grad_norm": 1.8239609003067017, + "learning_rate": 9.610374849164972e-06, + "loss": 0.4836, + "step": 16711 + }, + { + "epoch": 0.5271092887557167, + "grad_norm": 2.1940910816192627, + "learning_rate": 9.609354070411807e-06, + "loss": 0.4807, + "step": 16712 + }, + { + "epoch": 0.5271408295221574, + "grad_norm": 15.003602981567383, + "learning_rate": 9.608333295735337e-06, + "loss": 0.7345, + "step": 16713 + }, + { + "epoch": 0.527172370288598, + "grad_norm": 1.599462628364563, + "learning_rate": 9.607312525146212e-06, + "loss": 0.4168, + "step": 16714 + }, + { + "epoch": 0.5272039110550386, + "grad_norm": 2.19236159324646, + "learning_rate": 9.606291758655088e-06, + "loss": 0.4425, + "step": 16715 + }, + { + "epoch": 0.5272354518214792, + "grad_norm": 1.7557834386825562, + "learning_rate": 9.605270996272615e-06, + "loss": 0.4806, + "step": 16716 + }, + { + "epoch": 0.5272669925879199, + "grad_norm": 2.610340118408203, + "learning_rate": 9.60425023800944e-06, + "loss": 0.4667, + "step": 16717 + }, + { + "epoch": 0.5272985333543605, + "grad_norm": 1.4643330574035645, + "learning_rate": 9.603229483876227e-06, + "loss": 0.7453, + "step": 16718 + }, + { + "epoch": 0.5273300741208011, + "grad_norm": 1.7899922132492065, + "learning_rate": 9.60220873388362e-06, + "loss": 0.475, + "step": 16719 + }, + { + "epoch": 0.5273616148872418, + "grad_norm": 3.1439387798309326, + "learning_rate": 9.601187988042271e-06, + "loss": 0.3751, + "step": 16720 + }, + { + "epoch": 0.5273931556536824, + "grad_norm": 2.0393879413604736, + "learning_rate": 9.600167246362841e-06, + "loss": 0.4795, + "step": 16721 + }, + { + "epoch": 0.527424696420123, + "grad_norm": 1.8343181610107422, + "learning_rate": 9.599146508855972e-06, + "loss": 0.4378, + "step": 16722 + }, + { + "epoch": 0.5274562371865636, + "grad_norm": 13.355259895324707, + "learning_rate": 9.598125775532322e-06, + "loss": 0.4618, + "step": 16723 + }, + { + "epoch": 0.5274877779530043, + "grad_norm": 2.550576686859131, + "learning_rate": 9.597105046402542e-06, + "loss": 0.7776, + "step": 16724 + }, + { + "epoch": 0.5275193187194449, + "grad_norm": 1.6981934309005737, + "learning_rate": 9.596084321477282e-06, + "loss": 0.4934, + "step": 16725 + }, + { + "epoch": 0.5275508594858855, + "grad_norm": 1.557180404663086, + "learning_rate": 9.595063600767197e-06, + "loss": 0.4743, + "step": 16726 + }, + { + "epoch": 0.5275824002523262, + "grad_norm": 1.645236611366272, + "learning_rate": 9.594042884282939e-06, + "loss": 0.4794, + "step": 16727 + }, + { + "epoch": 0.5276139410187668, + "grad_norm": 1.4020183086395264, + "learning_rate": 9.593022172035154e-06, + "loss": 0.4844, + "step": 16728 + }, + { + "epoch": 0.5276454817852074, + "grad_norm": 2.0740199089050293, + "learning_rate": 9.592001464034503e-06, + "loss": 0.4885, + "step": 16729 + }, + { + "epoch": 0.5276770225516481, + "grad_norm": 1.536643147468567, + "learning_rate": 9.590980760291634e-06, + "loss": 0.4706, + "step": 16730 + }, + { + "epoch": 0.5277085633180886, + "grad_norm": 1.7298719882965088, + "learning_rate": 9.589960060817196e-06, + "loss": 0.4517, + "step": 16731 + }, + { + "epoch": 0.5277401040845292, + "grad_norm": 2.235057830810547, + "learning_rate": 9.588939365621843e-06, + "loss": 0.3423, + "step": 16732 + }, + { + "epoch": 0.5277716448509698, + "grad_norm": 1.5156302452087402, + "learning_rate": 9.58791867471623e-06, + "loss": 0.4853, + "step": 16733 + }, + { + "epoch": 0.5278031856174105, + "grad_norm": 1.6674907207489014, + "learning_rate": 9.586897988111e-06, + "loss": 0.4771, + "step": 16734 + }, + { + "epoch": 0.5278347263838511, + "grad_norm": 2.6481974124908447, + "learning_rate": 9.585877305816817e-06, + "loss": 0.41, + "step": 16735 + }, + { + "epoch": 0.5278662671502917, + "grad_norm": 1.8835424184799194, + "learning_rate": 9.584856627844319e-06, + "loss": 0.4461, + "step": 16736 + }, + { + "epoch": 0.5278978079167324, + "grad_norm": 9.365601539611816, + "learning_rate": 9.583835954204169e-06, + "loss": 0.7794, + "step": 16737 + }, + { + "epoch": 0.527929348683173, + "grad_norm": 1.5105684995651245, + "learning_rate": 9.582815284907015e-06, + "loss": 0.4578, + "step": 16738 + }, + { + "epoch": 0.5279608894496136, + "grad_norm": 1.631240725517273, + "learning_rate": 9.581794619963501e-06, + "loss": 0.469, + "step": 16739 + }, + { + "epoch": 0.5279924302160542, + "grad_norm": 2.2618589401245117, + "learning_rate": 9.580773959384291e-06, + "loss": 0.478, + "step": 16740 + }, + { + "epoch": 0.5280239709824949, + "grad_norm": 1.9967188835144043, + "learning_rate": 9.57975330318003e-06, + "loss": 0.4916, + "step": 16741 + }, + { + "epoch": 0.5280555117489355, + "grad_norm": 2.961552858352661, + "learning_rate": 9.578732651361368e-06, + "loss": 0.7388, + "step": 16742 + }, + { + "epoch": 0.5280870525153761, + "grad_norm": 1.6997015476226807, + "learning_rate": 9.57771200393896e-06, + "loss": 0.4863, + "step": 16743 + }, + { + "epoch": 0.5281185932818168, + "grad_norm": 2.2421951293945312, + "learning_rate": 9.576691360923454e-06, + "loss": 0.4466, + "step": 16744 + }, + { + "epoch": 0.5281501340482574, + "grad_norm": 2.0935075283050537, + "learning_rate": 9.575670722325498e-06, + "loss": 0.3936, + "step": 16745 + }, + { + "epoch": 0.528181674814698, + "grad_norm": 1.5131133794784546, + "learning_rate": 9.574650088155752e-06, + "loss": 0.5114, + "step": 16746 + }, + { + "epoch": 0.5282132155811387, + "grad_norm": 1.282856822013855, + "learning_rate": 9.573629458424862e-06, + "loss": 0.4682, + "step": 16747 + }, + { + "epoch": 0.5282447563475793, + "grad_norm": 1.5325438976287842, + "learning_rate": 9.572608833143478e-06, + "loss": 0.48, + "step": 16748 + }, + { + "epoch": 0.5282762971140199, + "grad_norm": 1.9690262079238892, + "learning_rate": 9.571588212322255e-06, + "loss": 0.4586, + "step": 16749 + }, + { + "epoch": 0.5283078378804605, + "grad_norm": 1.5243710279464722, + "learning_rate": 9.570567595971842e-06, + "loss": 0.4543, + "step": 16750 + }, + { + "epoch": 0.5283393786469012, + "grad_norm": 23.476421356201172, + "learning_rate": 9.569546984102886e-06, + "loss": 0.4185, + "step": 16751 + }, + { + "epoch": 0.5283709194133418, + "grad_norm": 5.0697197914123535, + "learning_rate": 9.568526376726043e-06, + "loss": 0.4413, + "step": 16752 + }, + { + "epoch": 0.5284024601797823, + "grad_norm": 1.4605870246887207, + "learning_rate": 9.567505773851963e-06, + "loss": 0.4676, + "step": 16753 + }, + { + "epoch": 0.528434000946223, + "grad_norm": 1.6910899877548218, + "learning_rate": 9.566485175491298e-06, + "loss": 0.4841, + "step": 16754 + }, + { + "epoch": 0.5284655417126636, + "grad_norm": 2.5233352184295654, + "learning_rate": 9.565464581654694e-06, + "loss": 0.4544, + "step": 16755 + }, + { + "epoch": 0.5284970824791042, + "grad_norm": 1.6147964000701904, + "learning_rate": 9.564443992352803e-06, + "loss": 0.4789, + "step": 16756 + }, + { + "epoch": 0.5285286232455448, + "grad_norm": 2.2325241565704346, + "learning_rate": 9.563423407596281e-06, + "loss": 0.3664, + "step": 16757 + }, + { + "epoch": 0.5285601640119855, + "grad_norm": 1.667077898979187, + "learning_rate": 9.562402827395775e-06, + "loss": 0.4467, + "step": 16758 + }, + { + "epoch": 0.5285917047784261, + "grad_norm": 1.4819763898849487, + "learning_rate": 9.56138225176193e-06, + "loss": 0.48, + "step": 16759 + }, + { + "epoch": 0.5286232455448667, + "grad_norm": 1.7321258783340454, + "learning_rate": 9.560361680705406e-06, + "loss": 0.4403, + "step": 16760 + }, + { + "epoch": 0.5286547863113074, + "grad_norm": 1.7886642217636108, + "learning_rate": 9.559341114236849e-06, + "loss": 0.7973, + "step": 16761 + }, + { + "epoch": 0.528686327077748, + "grad_norm": 2.2570388317108154, + "learning_rate": 9.558320552366908e-06, + "loss": 0.4649, + "step": 16762 + }, + { + "epoch": 0.5287178678441886, + "grad_norm": 1.4807525873184204, + "learning_rate": 9.557299995106239e-06, + "loss": 0.5096, + "step": 16763 + }, + { + "epoch": 0.5287494086106292, + "grad_norm": 1.565833568572998, + "learning_rate": 9.556279442465486e-06, + "loss": 0.4465, + "step": 16764 + }, + { + "epoch": 0.5287809493770699, + "grad_norm": 1.376358151435852, + "learning_rate": 9.555258894455298e-06, + "loss": 0.793, + "step": 16765 + }, + { + "epoch": 0.5288124901435105, + "grad_norm": 1.5069745779037476, + "learning_rate": 9.554238351086334e-06, + "loss": 0.4934, + "step": 16766 + }, + { + "epoch": 0.5288440309099511, + "grad_norm": 1.873718023300171, + "learning_rate": 9.553217812369237e-06, + "loss": 0.5005, + "step": 16767 + }, + { + "epoch": 0.5288755716763918, + "grad_norm": 2.411418914794922, + "learning_rate": 9.552197278314655e-06, + "loss": 0.7798, + "step": 16768 + }, + { + "epoch": 0.5289071124428324, + "grad_norm": 1.6206907033920288, + "learning_rate": 9.551176748933247e-06, + "loss": 0.4307, + "step": 16769 + }, + { + "epoch": 0.528938653209273, + "grad_norm": 1.3296579122543335, + "learning_rate": 9.550156224235657e-06, + "loss": 0.4645, + "step": 16770 + }, + { + "epoch": 0.5289701939757137, + "grad_norm": 1.583808183670044, + "learning_rate": 9.549135704232537e-06, + "loss": 0.5133, + "step": 16771 + }, + { + "epoch": 0.5290017347421543, + "grad_norm": 1.45375657081604, + "learning_rate": 9.548115188934535e-06, + "loss": 0.4457, + "step": 16772 + }, + { + "epoch": 0.5290332755085949, + "grad_norm": 2.156498432159424, + "learning_rate": 9.5470946783523e-06, + "loss": 0.4705, + "step": 16773 + }, + { + "epoch": 0.5290648162750354, + "grad_norm": 3.146415948867798, + "learning_rate": 9.546074172496488e-06, + "loss": 0.4532, + "step": 16774 + }, + { + "epoch": 0.5290963570414761, + "grad_norm": 1.6861368417739868, + "learning_rate": 9.545053671377742e-06, + "loss": 0.5251, + "step": 16775 + }, + { + "epoch": 0.5291278978079167, + "grad_norm": 1.4678139686584473, + "learning_rate": 9.544033175006711e-06, + "loss": 0.4969, + "step": 16776 + }, + { + "epoch": 0.5291594385743573, + "grad_norm": 2.395653486251831, + "learning_rate": 9.543012683394054e-06, + "loss": 0.4632, + "step": 16777 + }, + { + "epoch": 0.529190979340798, + "grad_norm": 1.350154995918274, + "learning_rate": 9.541992196550413e-06, + "loss": 0.4626, + "step": 16778 + }, + { + "epoch": 0.5292225201072386, + "grad_norm": 1.8242074251174927, + "learning_rate": 9.540971714486434e-06, + "loss": 0.5118, + "step": 16779 + }, + { + "epoch": 0.5292540608736792, + "grad_norm": 1.8125053644180298, + "learning_rate": 9.539951237212777e-06, + "loss": 0.4434, + "step": 16780 + }, + { + "epoch": 0.5292856016401198, + "grad_norm": 2.213578701019287, + "learning_rate": 9.538930764740085e-06, + "loss": 0.4826, + "step": 16781 + }, + { + "epoch": 0.5293171424065605, + "grad_norm": 2.253678321838379, + "learning_rate": 9.537910297079008e-06, + "loss": 0.3731, + "step": 16782 + }, + { + "epoch": 0.5293486831730011, + "grad_norm": 2.8181655406951904, + "learning_rate": 9.536889834240197e-06, + "loss": 0.4122, + "step": 16783 + }, + { + "epoch": 0.5293802239394417, + "grad_norm": 1.5222398042678833, + "learning_rate": 9.535869376234296e-06, + "loss": 0.4985, + "step": 16784 + }, + { + "epoch": 0.5294117647058824, + "grad_norm": 1.6275699138641357, + "learning_rate": 9.534848923071963e-06, + "loss": 0.4319, + "step": 16785 + }, + { + "epoch": 0.529443305472323, + "grad_norm": 1.8130627870559692, + "learning_rate": 9.53382847476384e-06, + "loss": 0.513, + "step": 16786 + }, + { + "epoch": 0.5294748462387636, + "grad_norm": 1.2785882949829102, + "learning_rate": 9.532808031320578e-06, + "loss": 0.3542, + "step": 16787 + }, + { + "epoch": 0.5295063870052042, + "grad_norm": 7.501739978790283, + "learning_rate": 9.53178759275283e-06, + "loss": 0.4406, + "step": 16788 + }, + { + "epoch": 0.5295379277716449, + "grad_norm": 1.4645837545394897, + "learning_rate": 9.530767159071243e-06, + "loss": 0.4641, + "step": 16789 + }, + { + "epoch": 0.5295694685380855, + "grad_norm": 2.0513694286346436, + "learning_rate": 9.529746730286461e-06, + "loss": 0.484, + "step": 16790 + }, + { + "epoch": 0.5296010093045261, + "grad_norm": 1.4214832782745361, + "learning_rate": 9.528726306409141e-06, + "loss": 0.4659, + "step": 16791 + }, + { + "epoch": 0.5296325500709668, + "grad_norm": 1.6245509386062622, + "learning_rate": 9.527705887449925e-06, + "loss": 0.729, + "step": 16792 + }, + { + "epoch": 0.5296640908374074, + "grad_norm": 1.4683414697647095, + "learning_rate": 9.526685473419464e-06, + "loss": 0.4457, + "step": 16793 + }, + { + "epoch": 0.529695631603848, + "grad_norm": 1.9512948989868164, + "learning_rate": 9.52566506432841e-06, + "loss": 0.4795, + "step": 16794 + }, + { + "epoch": 0.5297271723702885, + "grad_norm": 1.531516432762146, + "learning_rate": 9.52464466018741e-06, + "loss": 0.4861, + "step": 16795 + }, + { + "epoch": 0.5297587131367292, + "grad_norm": 2.001854658126831, + "learning_rate": 9.523624261007108e-06, + "loss": 0.4398, + "step": 16796 + }, + { + "epoch": 0.5297902539031698, + "grad_norm": 1.541746735572815, + "learning_rate": 9.522603866798162e-06, + "loss": 0.4742, + "step": 16797 + }, + { + "epoch": 0.5298217946696104, + "grad_norm": 1.6982409954071045, + "learning_rate": 9.521583477571212e-06, + "loss": 0.4743, + "step": 16798 + }, + { + "epoch": 0.5298533354360511, + "grad_norm": 1.413057565689087, + "learning_rate": 9.520563093336909e-06, + "loss": 0.4747, + "step": 16799 + }, + { + "epoch": 0.5298848762024917, + "grad_norm": 1.6412365436553955, + "learning_rate": 9.519542714105904e-06, + "loss": 0.3281, + "step": 16800 + }, + { + "epoch": 0.5299164169689323, + "grad_norm": 1.6387503147125244, + "learning_rate": 9.518522339888844e-06, + "loss": 0.4135, + "step": 16801 + }, + { + "epoch": 0.529947957735373, + "grad_norm": 2.5482869148254395, + "learning_rate": 9.517501970696377e-06, + "loss": 0.4339, + "step": 16802 + }, + { + "epoch": 0.5299794985018136, + "grad_norm": 1.4449496269226074, + "learning_rate": 9.516481606539154e-06, + "loss": 0.4776, + "step": 16803 + }, + { + "epoch": 0.5300110392682542, + "grad_norm": 1.5404253005981445, + "learning_rate": 9.515461247427815e-06, + "loss": 0.4464, + "step": 16804 + }, + { + "epoch": 0.5300425800346948, + "grad_norm": 1.6284418106079102, + "learning_rate": 9.51444089337302e-06, + "loss": 0.4331, + "step": 16805 + }, + { + "epoch": 0.5300741208011355, + "grad_norm": 1.534918189048767, + "learning_rate": 9.51342054438541e-06, + "loss": 0.4494, + "step": 16806 + }, + { + "epoch": 0.5301056615675761, + "grad_norm": 1.988761305809021, + "learning_rate": 9.51240020047563e-06, + "loss": 0.4947, + "step": 16807 + }, + { + "epoch": 0.5301372023340167, + "grad_norm": 1.4208130836486816, + "learning_rate": 9.511379861654338e-06, + "loss": 0.4579, + "step": 16808 + }, + { + "epoch": 0.5301687431004574, + "grad_norm": 1.3863753080368042, + "learning_rate": 9.510359527932174e-06, + "loss": 0.4791, + "step": 16809 + }, + { + "epoch": 0.530200283866898, + "grad_norm": 1.5688854455947876, + "learning_rate": 9.509339199319789e-06, + "loss": 0.4525, + "step": 16810 + }, + { + "epoch": 0.5302318246333386, + "grad_norm": 3.194366455078125, + "learning_rate": 9.508318875827832e-06, + "loss": 0.4349, + "step": 16811 + }, + { + "epoch": 0.5302633653997793, + "grad_norm": 1.5537441968917847, + "learning_rate": 9.507298557466946e-06, + "loss": 0.4739, + "step": 16812 + }, + { + "epoch": 0.5302949061662199, + "grad_norm": 1.5888041257858276, + "learning_rate": 9.506278244247784e-06, + "loss": 0.4937, + "step": 16813 + }, + { + "epoch": 0.5303264469326605, + "grad_norm": 1.5389379262924194, + "learning_rate": 9.505257936180992e-06, + "loss": 0.498, + "step": 16814 + }, + { + "epoch": 0.5303579876991011, + "grad_norm": 1.6991651058197021, + "learning_rate": 9.504237633277214e-06, + "loss": 0.432, + "step": 16815 + }, + { + "epoch": 0.5303895284655417, + "grad_norm": 1.432776927947998, + "learning_rate": 9.503217335547107e-06, + "loss": 0.4819, + "step": 16816 + }, + { + "epoch": 0.5304210692319823, + "grad_norm": 2.8787474632263184, + "learning_rate": 9.502197043001312e-06, + "loss": 0.4888, + "step": 16817 + }, + { + "epoch": 0.5304526099984229, + "grad_norm": 2.122495412826538, + "learning_rate": 9.501176755650472e-06, + "loss": 0.4718, + "step": 16818 + }, + { + "epoch": 0.5304841507648635, + "grad_norm": 1.9322305917739868, + "learning_rate": 9.500156473505245e-06, + "loss": 0.7643, + "step": 16819 + }, + { + "epoch": 0.5305156915313042, + "grad_norm": 1.6390516757965088, + "learning_rate": 9.499136196576274e-06, + "loss": 0.4957, + "step": 16820 + }, + { + "epoch": 0.5305472322977448, + "grad_norm": 1.5412113666534424, + "learning_rate": 9.498115924874204e-06, + "loss": 0.4406, + "step": 16821 + }, + { + "epoch": 0.5305787730641854, + "grad_norm": 1.5705970525741577, + "learning_rate": 9.497095658409685e-06, + "loss": 0.4893, + "step": 16822 + }, + { + "epoch": 0.5306103138306261, + "grad_norm": 2.2141635417938232, + "learning_rate": 9.496075397193362e-06, + "loss": 0.3276, + "step": 16823 + }, + { + "epoch": 0.5306418545970667, + "grad_norm": 1.5708954334259033, + "learning_rate": 9.49505514123588e-06, + "loss": 0.4756, + "step": 16824 + }, + { + "epoch": 0.5306733953635073, + "grad_norm": 1.4218544960021973, + "learning_rate": 9.494034890547895e-06, + "loss": 0.4638, + "step": 16825 + }, + { + "epoch": 0.530704936129948, + "grad_norm": 1.7399510145187378, + "learning_rate": 9.493014645140048e-06, + "loss": 0.4967, + "step": 16826 + }, + { + "epoch": 0.5307364768963886, + "grad_norm": 2.727571725845337, + "learning_rate": 9.491994405022983e-06, + "loss": 0.3802, + "step": 16827 + }, + { + "epoch": 0.5307680176628292, + "grad_norm": 1.6827090978622437, + "learning_rate": 9.490974170207356e-06, + "loss": 0.4581, + "step": 16828 + }, + { + "epoch": 0.5307995584292698, + "grad_norm": 1.5468183755874634, + "learning_rate": 9.489953940703807e-06, + "loss": 0.4823, + "step": 16829 + }, + { + "epoch": 0.5308310991957105, + "grad_norm": 2.0802857875823975, + "learning_rate": 9.488933716522983e-06, + "loss": 0.4895, + "step": 16830 + }, + { + "epoch": 0.5308626399621511, + "grad_norm": 1.3935670852661133, + "learning_rate": 9.487913497675536e-06, + "loss": 0.4964, + "step": 16831 + }, + { + "epoch": 0.5308941807285917, + "grad_norm": 1.8273032903671265, + "learning_rate": 9.486893284172103e-06, + "loss": 0.466, + "step": 16832 + }, + { + "epoch": 0.5309257214950324, + "grad_norm": 1.9712944030761719, + "learning_rate": 9.485873076023343e-06, + "loss": 0.4355, + "step": 16833 + }, + { + "epoch": 0.530957262261473, + "grad_norm": 1.4936752319335938, + "learning_rate": 9.484852873239896e-06, + "loss": 0.5008, + "step": 16834 + }, + { + "epoch": 0.5309888030279136, + "grad_norm": 1.7677698135375977, + "learning_rate": 9.483832675832404e-06, + "loss": 0.4358, + "step": 16835 + }, + { + "epoch": 0.5310203437943543, + "grad_norm": 1.344805121421814, + "learning_rate": 9.482812483811524e-06, + "loss": 0.4779, + "step": 16836 + }, + { + "epoch": 0.5310518845607948, + "grad_norm": 1.5515234470367432, + "learning_rate": 9.481792297187897e-06, + "loss": 0.4841, + "step": 16837 + }, + { + "epoch": 0.5310834253272354, + "grad_norm": 1.7090901136398315, + "learning_rate": 9.480772115972167e-06, + "loss": 0.4982, + "step": 16838 + }, + { + "epoch": 0.531114966093676, + "grad_norm": 1.570575475692749, + "learning_rate": 9.479751940174987e-06, + "loss": 0.4474, + "step": 16839 + }, + { + "epoch": 0.5311465068601167, + "grad_norm": 1.3029450178146362, + "learning_rate": 9.478731769806996e-06, + "loss": 0.4693, + "step": 16840 + }, + { + "epoch": 0.5311780476265573, + "grad_norm": 1.737534999847412, + "learning_rate": 9.477711604878846e-06, + "loss": 0.4661, + "step": 16841 + }, + { + "epoch": 0.5312095883929979, + "grad_norm": 1.9734679460525513, + "learning_rate": 9.47669144540118e-06, + "loss": 0.4606, + "step": 16842 + }, + { + "epoch": 0.5312411291594386, + "grad_norm": 1.5272536277770996, + "learning_rate": 9.475671291384647e-06, + "loss": 0.4863, + "step": 16843 + }, + { + "epoch": 0.5312726699258792, + "grad_norm": 2.2225189208984375, + "learning_rate": 9.474651142839887e-06, + "loss": 0.4875, + "step": 16844 + }, + { + "epoch": 0.5313042106923198, + "grad_norm": 1.2678312063217163, + "learning_rate": 9.473630999777554e-06, + "loss": 0.4633, + "step": 16845 + }, + { + "epoch": 0.5313357514587604, + "grad_norm": 1.1594477891921997, + "learning_rate": 9.47261086220829e-06, + "loss": 0.4478, + "step": 16846 + }, + { + "epoch": 0.5313672922252011, + "grad_norm": 1.608320713043213, + "learning_rate": 9.471590730142736e-06, + "loss": 0.4816, + "step": 16847 + }, + { + "epoch": 0.5313988329916417, + "grad_norm": 1.4850612878799438, + "learning_rate": 9.47057060359155e-06, + "loss": 0.4769, + "step": 16848 + }, + { + "epoch": 0.5314303737580823, + "grad_norm": 1.6007485389709473, + "learning_rate": 9.469550482565366e-06, + "loss": 0.4915, + "step": 16849 + }, + { + "epoch": 0.531461914524523, + "grad_norm": 3.0477728843688965, + "learning_rate": 9.468530367074838e-06, + "loss": 0.4015, + "step": 16850 + }, + { + "epoch": 0.5314934552909636, + "grad_norm": 1.3794686794281006, + "learning_rate": 9.467510257130608e-06, + "loss": 0.475, + "step": 16851 + }, + { + "epoch": 0.5315249960574042, + "grad_norm": 2.670651435852051, + "learning_rate": 9.466490152743318e-06, + "loss": 0.4656, + "step": 16852 + }, + { + "epoch": 0.5315565368238448, + "grad_norm": 1.5107539892196655, + "learning_rate": 9.46547005392362e-06, + "loss": 0.4327, + "step": 16853 + }, + { + "epoch": 0.5315880775902855, + "grad_norm": 1.403802752494812, + "learning_rate": 9.464449960682161e-06, + "loss": 0.4671, + "step": 16854 + }, + { + "epoch": 0.5316196183567261, + "grad_norm": 3.81740665435791, + "learning_rate": 9.463429873029575e-06, + "loss": 0.4886, + "step": 16855 + }, + { + "epoch": 0.5316511591231667, + "grad_norm": 1.3676962852478027, + "learning_rate": 9.46240979097652e-06, + "loss": 0.4584, + "step": 16856 + }, + { + "epoch": 0.5316826998896074, + "grad_norm": 1.6875470876693726, + "learning_rate": 9.461389714533637e-06, + "loss": 0.7876, + "step": 16857 + }, + { + "epoch": 0.5317142406560479, + "grad_norm": 1.2972087860107422, + "learning_rate": 9.460369643711564e-06, + "loss": 0.4398, + "step": 16858 + }, + { + "epoch": 0.5317457814224885, + "grad_norm": 1.4901020526885986, + "learning_rate": 9.459349578520959e-06, + "loss": 0.7638, + "step": 16859 + }, + { + "epoch": 0.5317773221889291, + "grad_norm": 1.595794916152954, + "learning_rate": 9.45832951897246e-06, + "loss": 0.4619, + "step": 16860 + }, + { + "epoch": 0.5318088629553698, + "grad_norm": 2.554751396179199, + "learning_rate": 9.457309465076712e-06, + "loss": 0.494, + "step": 16861 + }, + { + "epoch": 0.5318404037218104, + "grad_norm": 1.5818578004837036, + "learning_rate": 9.456289416844362e-06, + "loss": 0.4789, + "step": 16862 + }, + { + "epoch": 0.531871944488251, + "grad_norm": 1.6046849489212036, + "learning_rate": 9.45526937428605e-06, + "loss": 0.513, + "step": 16863 + }, + { + "epoch": 0.5319034852546917, + "grad_norm": 1.6963626146316528, + "learning_rate": 9.454249337412431e-06, + "loss": 0.47, + "step": 16864 + }, + { + "epoch": 0.5319350260211323, + "grad_norm": 1.3057236671447754, + "learning_rate": 9.453229306234143e-06, + "loss": 0.4352, + "step": 16865 + }, + { + "epoch": 0.5319665667875729, + "grad_norm": 1.7085175514221191, + "learning_rate": 9.452209280761827e-06, + "loss": 0.767, + "step": 16866 + }, + { + "epoch": 0.5319981075540136, + "grad_norm": 1.4710578918457031, + "learning_rate": 9.451189261006137e-06, + "loss": 0.4875, + "step": 16867 + }, + { + "epoch": 0.5320296483204542, + "grad_norm": 2.312073230743408, + "learning_rate": 9.450169246977713e-06, + "loss": 0.3378, + "step": 16868 + }, + { + "epoch": 0.5320611890868948, + "grad_norm": 1.8866521120071411, + "learning_rate": 9.449149238687198e-06, + "loss": 0.4545, + "step": 16869 + }, + { + "epoch": 0.5320927298533354, + "grad_norm": 1.6426106691360474, + "learning_rate": 9.44812923614524e-06, + "loss": 0.4613, + "step": 16870 + }, + { + "epoch": 0.5321242706197761, + "grad_norm": 1.778645634651184, + "learning_rate": 9.447109239362483e-06, + "loss": 0.4929, + "step": 16871 + }, + { + "epoch": 0.5321558113862167, + "grad_norm": 2.1295602321624756, + "learning_rate": 9.446089248349566e-06, + "loss": 0.493, + "step": 16872 + }, + { + "epoch": 0.5321873521526573, + "grad_norm": 84.3425521850586, + "learning_rate": 9.445069263117142e-06, + "loss": 0.3857, + "step": 16873 + }, + { + "epoch": 0.532218892919098, + "grad_norm": 1.5905994176864624, + "learning_rate": 9.44404928367585e-06, + "loss": 0.4621, + "step": 16874 + }, + { + "epoch": 0.5322504336855386, + "grad_norm": 1.8505786657333374, + "learning_rate": 9.443029310036332e-06, + "loss": 0.4841, + "step": 16875 + }, + { + "epoch": 0.5322819744519792, + "grad_norm": 1.9127558469772339, + "learning_rate": 9.442009342209239e-06, + "loss": 0.4601, + "step": 16876 + }, + { + "epoch": 0.5323135152184199, + "grad_norm": 1.4570980072021484, + "learning_rate": 9.440989380205214e-06, + "loss": 0.4214, + "step": 16877 + }, + { + "epoch": 0.5323450559848605, + "grad_norm": 1.7101597785949707, + "learning_rate": 9.439969424034893e-06, + "loss": 0.4572, + "step": 16878 + }, + { + "epoch": 0.532376596751301, + "grad_norm": 2.7791261672973633, + "learning_rate": 9.43894947370893e-06, + "loss": 0.3884, + "step": 16879 + }, + { + "epoch": 0.5324081375177416, + "grad_norm": 3.949833393096924, + "learning_rate": 9.437929529237963e-06, + "loss": 0.4724, + "step": 16880 + }, + { + "epoch": 0.5324396782841823, + "grad_norm": 1.5085458755493164, + "learning_rate": 9.436909590632641e-06, + "loss": 0.458, + "step": 16881 + }, + { + "epoch": 0.5324712190506229, + "grad_norm": 1.555133581161499, + "learning_rate": 9.435889657903603e-06, + "loss": 0.3686, + "step": 16882 + }, + { + "epoch": 0.5325027598170635, + "grad_norm": 1.681644082069397, + "learning_rate": 9.434869731061491e-06, + "loss": 0.5019, + "step": 16883 + }, + { + "epoch": 0.5325343005835041, + "grad_norm": 1.6363848447799683, + "learning_rate": 9.433849810116956e-06, + "loss": 0.48, + "step": 16884 + }, + { + "epoch": 0.5325658413499448, + "grad_norm": 1.556849479675293, + "learning_rate": 9.432829895080639e-06, + "loss": 0.4897, + "step": 16885 + }, + { + "epoch": 0.5325973821163854, + "grad_norm": 1.6563111543655396, + "learning_rate": 9.431809985963176e-06, + "loss": 0.4734, + "step": 16886 + }, + { + "epoch": 0.532628922882826, + "grad_norm": 1.8690396547317505, + "learning_rate": 9.430790082775223e-06, + "loss": 0.4585, + "step": 16887 + }, + { + "epoch": 0.5326604636492667, + "grad_norm": 1.9021580219268799, + "learning_rate": 9.429770185527417e-06, + "loss": 0.4845, + "step": 16888 + }, + { + "epoch": 0.5326920044157073, + "grad_norm": 1.5204460620880127, + "learning_rate": 9.4287502942304e-06, + "loss": 0.4224, + "step": 16889 + }, + { + "epoch": 0.5327235451821479, + "grad_norm": 1.7238950729370117, + "learning_rate": 9.42773040889482e-06, + "loss": 0.4668, + "step": 16890 + }, + { + "epoch": 0.5327550859485886, + "grad_norm": 1.7244840860366821, + "learning_rate": 9.426710529531315e-06, + "loss": 0.4791, + "step": 16891 + }, + { + "epoch": 0.5327866267150292, + "grad_norm": 1.4145115613937378, + "learning_rate": 9.42569065615053e-06, + "loss": 0.4958, + "step": 16892 + }, + { + "epoch": 0.5328181674814698, + "grad_norm": 1.738577127456665, + "learning_rate": 9.42467078876311e-06, + "loss": 0.4553, + "step": 16893 + }, + { + "epoch": 0.5328497082479104, + "grad_norm": 1.6028563976287842, + "learning_rate": 9.4236509273797e-06, + "loss": 0.4771, + "step": 16894 + }, + { + "epoch": 0.5328812490143511, + "grad_norm": 1.9609296321868896, + "learning_rate": 9.422631072010933e-06, + "loss": 0.4805, + "step": 16895 + }, + { + "epoch": 0.5329127897807917, + "grad_norm": 1.0091534852981567, + "learning_rate": 9.421611222667465e-06, + "loss": 0.3504, + "step": 16896 + }, + { + "epoch": 0.5329443305472323, + "grad_norm": 1.7477526664733887, + "learning_rate": 9.42059137935993e-06, + "loss": 0.4547, + "step": 16897 + }, + { + "epoch": 0.532975871313673, + "grad_norm": 1.830905556678772, + "learning_rate": 9.419571542098975e-06, + "loss": 0.4643, + "step": 16898 + }, + { + "epoch": 0.5330074120801136, + "grad_norm": 1.423187017440796, + "learning_rate": 9.418551710895243e-06, + "loss": 0.4382, + "step": 16899 + }, + { + "epoch": 0.5330389528465542, + "grad_norm": 1.5492494106292725, + "learning_rate": 9.417531885759374e-06, + "loss": 0.4849, + "step": 16900 + }, + { + "epoch": 0.5330704936129947, + "grad_norm": 1.8256648778915405, + "learning_rate": 9.416512066702013e-06, + "loss": 0.4812, + "step": 16901 + }, + { + "epoch": 0.5331020343794354, + "grad_norm": 1.836877703666687, + "learning_rate": 9.415492253733803e-06, + "loss": 0.4825, + "step": 16902 + }, + { + "epoch": 0.533133575145876, + "grad_norm": 3.7755606174468994, + "learning_rate": 9.414472446865378e-06, + "loss": 0.7821, + "step": 16903 + }, + { + "epoch": 0.5331651159123166, + "grad_norm": 2.128438949584961, + "learning_rate": 9.413452646107395e-06, + "loss": 0.364, + "step": 16904 + }, + { + "epoch": 0.5331966566787573, + "grad_norm": 1.5239816904067993, + "learning_rate": 9.41243285147049e-06, + "loss": 0.7902, + "step": 16905 + }, + { + "epoch": 0.5332281974451979, + "grad_norm": 3.218259572982788, + "learning_rate": 9.411413062965297e-06, + "loss": 0.762, + "step": 16906 + }, + { + "epoch": 0.5332597382116385, + "grad_norm": 1.5733530521392822, + "learning_rate": 9.41039328060247e-06, + "loss": 0.7252, + "step": 16907 + }, + { + "epoch": 0.5332912789780792, + "grad_norm": 2.2174346446990967, + "learning_rate": 9.40937350439265e-06, + "loss": 0.4656, + "step": 16908 + }, + { + "epoch": 0.5333228197445198, + "grad_norm": 1.4413429498672485, + "learning_rate": 9.408353734346472e-06, + "loss": 0.4903, + "step": 16909 + }, + { + "epoch": 0.5333543605109604, + "grad_norm": 1.4455441236495972, + "learning_rate": 9.407333970474585e-06, + "loss": 0.7553, + "step": 16910 + }, + { + "epoch": 0.533385901277401, + "grad_norm": 1.670945167541504, + "learning_rate": 9.406314212787622e-06, + "loss": 0.448, + "step": 16911 + }, + { + "epoch": 0.5334174420438417, + "grad_norm": 3.121638298034668, + "learning_rate": 9.405294461296237e-06, + "loss": 0.476, + "step": 16912 + }, + { + "epoch": 0.5334489828102823, + "grad_norm": 1.5100425481796265, + "learning_rate": 9.404274716011067e-06, + "loss": 0.4393, + "step": 16913 + }, + { + "epoch": 0.5334805235767229, + "grad_norm": 4.116366386413574, + "learning_rate": 9.403254976942747e-06, + "loss": 0.53, + "step": 16914 + }, + { + "epoch": 0.5335120643431636, + "grad_norm": 1.5059884786605835, + "learning_rate": 9.40223524410193e-06, + "loss": 0.4839, + "step": 16915 + }, + { + "epoch": 0.5335436051096042, + "grad_norm": 1.6614652872085571, + "learning_rate": 9.401215517499252e-06, + "loss": 0.4693, + "step": 16916 + }, + { + "epoch": 0.5335751458760448, + "grad_norm": 1.404039978981018, + "learning_rate": 9.400195797145353e-06, + "loss": 0.462, + "step": 16917 + }, + { + "epoch": 0.5336066866424855, + "grad_norm": 1.5167951583862305, + "learning_rate": 9.399176083050877e-06, + "loss": 0.8221, + "step": 16918 + }, + { + "epoch": 0.5336382274089261, + "grad_norm": 1.655770182609558, + "learning_rate": 9.398156375226465e-06, + "loss": 0.7951, + "step": 16919 + }, + { + "epoch": 0.5336697681753667, + "grad_norm": 1.597670555114746, + "learning_rate": 9.397136673682759e-06, + "loss": 0.4657, + "step": 16920 + }, + { + "epoch": 0.5337013089418073, + "grad_norm": 1.403993844985962, + "learning_rate": 9.3961169784304e-06, + "loss": 0.4528, + "step": 16921 + }, + { + "epoch": 0.5337328497082479, + "grad_norm": 1.492767572402954, + "learning_rate": 9.395097289480031e-06, + "loss": 0.4353, + "step": 16922 + }, + { + "epoch": 0.5337643904746885, + "grad_norm": 1.8736323118209839, + "learning_rate": 9.394077606842287e-06, + "loss": 0.5005, + "step": 16923 + }, + { + "epoch": 0.5337959312411291, + "grad_norm": 2.4534878730773926, + "learning_rate": 9.393057930527816e-06, + "loss": 0.4533, + "step": 16924 + }, + { + "epoch": 0.5338274720075697, + "grad_norm": 1.6781253814697266, + "learning_rate": 9.39203826054726e-06, + "loss": 0.4642, + "step": 16925 + }, + { + "epoch": 0.5338590127740104, + "grad_norm": 1.3450143337249756, + "learning_rate": 9.39101859691125e-06, + "loss": 0.4913, + "step": 16926 + }, + { + "epoch": 0.533890553540451, + "grad_norm": 1.3262450695037842, + "learning_rate": 9.38999893963044e-06, + "loss": 0.7805, + "step": 16927 + }, + { + "epoch": 0.5339220943068916, + "grad_norm": 2.699263334274292, + "learning_rate": 9.388979288715461e-06, + "loss": 0.4406, + "step": 16928 + }, + { + "epoch": 0.5339536350733323, + "grad_norm": 1.8756611347198486, + "learning_rate": 9.38795964417696e-06, + "loss": 0.4705, + "step": 16929 + }, + { + "epoch": 0.5339851758397729, + "grad_norm": 2.3735129833221436, + "learning_rate": 9.386940006025575e-06, + "loss": 0.5001, + "step": 16930 + }, + { + "epoch": 0.5340167166062135, + "grad_norm": 1.8392049074172974, + "learning_rate": 9.385920374271944e-06, + "loss": 0.4899, + "step": 16931 + }, + { + "epoch": 0.5340482573726542, + "grad_norm": 1.8254927396774292, + "learning_rate": 9.384900748926714e-06, + "loss": 0.4576, + "step": 16932 + }, + { + "epoch": 0.5340797981390948, + "grad_norm": 1.870140790939331, + "learning_rate": 9.383881130000523e-06, + "loss": 0.4379, + "step": 16933 + }, + { + "epoch": 0.5341113389055354, + "grad_norm": 1.5227539539337158, + "learning_rate": 9.382861517504006e-06, + "loss": 0.4776, + "step": 16934 + }, + { + "epoch": 0.534142879671976, + "grad_norm": 1.5052926540374756, + "learning_rate": 9.381841911447812e-06, + "loss": 0.4833, + "step": 16935 + }, + { + "epoch": 0.5341744204384167, + "grad_norm": 1.716709017753601, + "learning_rate": 9.38082231184258e-06, + "loss": 0.4804, + "step": 16936 + }, + { + "epoch": 0.5342059612048573, + "grad_norm": 1.6094435453414917, + "learning_rate": 9.379802718698943e-06, + "loss": 0.3779, + "step": 16937 + }, + { + "epoch": 0.5342375019712979, + "grad_norm": 1.191881775856018, + "learning_rate": 9.378783132027546e-06, + "loss": 0.4456, + "step": 16938 + }, + { + "epoch": 0.5342690427377386, + "grad_norm": 1.9445189237594604, + "learning_rate": 9.377763551839033e-06, + "loss": 0.4663, + "step": 16939 + }, + { + "epoch": 0.5343005835041792, + "grad_norm": 1.9956251382827759, + "learning_rate": 9.37674397814404e-06, + "loss": 0.4443, + "step": 16940 + }, + { + "epoch": 0.5343321242706198, + "grad_norm": 2.0673022270202637, + "learning_rate": 9.375724410953208e-06, + "loss": 0.3836, + "step": 16941 + }, + { + "epoch": 0.5343636650370605, + "grad_norm": 1.8283336162567139, + "learning_rate": 9.374704850277176e-06, + "loss": 0.4343, + "step": 16942 + }, + { + "epoch": 0.534395205803501, + "grad_norm": 1.7578966617584229, + "learning_rate": 9.37368529612658e-06, + "loss": 0.4954, + "step": 16943 + }, + { + "epoch": 0.5344267465699416, + "grad_norm": 1.6920368671417236, + "learning_rate": 9.37266574851207e-06, + "loss": 0.4583, + "step": 16944 + }, + { + "epoch": 0.5344582873363822, + "grad_norm": 1.4798102378845215, + "learning_rate": 9.371646207444274e-06, + "loss": 0.4556, + "step": 16945 + }, + { + "epoch": 0.5344898281028229, + "grad_norm": 1.6753038167953491, + "learning_rate": 9.370626672933842e-06, + "loss": 0.4891, + "step": 16946 + }, + { + "epoch": 0.5345213688692635, + "grad_norm": 1.7091329097747803, + "learning_rate": 9.369607144991412e-06, + "loss": 0.4638, + "step": 16947 + }, + { + "epoch": 0.5345529096357041, + "grad_norm": 1.5545130968093872, + "learning_rate": 9.368587623627616e-06, + "loss": 0.494, + "step": 16948 + }, + { + "epoch": 0.5345844504021448, + "grad_norm": 1.5683386325836182, + "learning_rate": 9.367568108853103e-06, + "loss": 0.4524, + "step": 16949 + }, + { + "epoch": 0.5346159911685854, + "grad_norm": 1.7573661804199219, + "learning_rate": 9.366548600678506e-06, + "loss": 0.5131, + "step": 16950 + }, + { + "epoch": 0.534647531935026, + "grad_norm": 2.6307685375213623, + "learning_rate": 9.36552909911446e-06, + "loss": 0.3801, + "step": 16951 + }, + { + "epoch": 0.5346790727014666, + "grad_norm": 1.7429689168930054, + "learning_rate": 9.36450960417162e-06, + "loss": 0.376, + "step": 16952 + }, + { + "epoch": 0.5347106134679073, + "grad_norm": 1.6126827001571655, + "learning_rate": 9.363490115860611e-06, + "loss": 0.4759, + "step": 16953 + }, + { + "epoch": 0.5347421542343479, + "grad_norm": 1.4887218475341797, + "learning_rate": 9.362470634192074e-06, + "loss": 0.4688, + "step": 16954 + }, + { + "epoch": 0.5347736950007885, + "grad_norm": 1.4091452360153198, + "learning_rate": 9.361451159176657e-06, + "loss": 0.4558, + "step": 16955 + }, + { + "epoch": 0.5348052357672292, + "grad_norm": 1.4912501573562622, + "learning_rate": 9.360431690824992e-06, + "loss": 0.4656, + "step": 16956 + }, + { + "epoch": 0.5348367765336698, + "grad_norm": 1.449979543685913, + "learning_rate": 9.359412229147715e-06, + "loss": 0.4539, + "step": 16957 + }, + { + "epoch": 0.5348683173001104, + "grad_norm": 1.8557674884796143, + "learning_rate": 9.358392774155474e-06, + "loss": 0.4811, + "step": 16958 + }, + { + "epoch": 0.534899858066551, + "grad_norm": 1.7281709909439087, + "learning_rate": 9.357373325858897e-06, + "loss": 0.4533, + "step": 16959 + }, + { + "epoch": 0.5349313988329917, + "grad_norm": 1.6348609924316406, + "learning_rate": 9.356353884268631e-06, + "loss": 0.4898, + "step": 16960 + }, + { + "epoch": 0.5349629395994323, + "grad_norm": 1.6977118253707886, + "learning_rate": 9.355334449395314e-06, + "loss": 0.4712, + "step": 16961 + }, + { + "epoch": 0.5349944803658729, + "grad_norm": 1.4839848279953003, + "learning_rate": 9.354315021249577e-06, + "loss": 0.4628, + "step": 16962 + }, + { + "epoch": 0.5350260211323136, + "grad_norm": 1.4672389030456543, + "learning_rate": 9.353295599842069e-06, + "loss": 0.4451, + "step": 16963 + }, + { + "epoch": 0.5350575618987541, + "grad_norm": 3.929669141769409, + "learning_rate": 9.352276185183423e-06, + "loss": 0.4661, + "step": 16964 + }, + { + "epoch": 0.5350891026651947, + "grad_norm": 2.3491485118865967, + "learning_rate": 9.351256777284274e-06, + "loss": 0.4591, + "step": 16965 + }, + { + "epoch": 0.5351206434316353, + "grad_norm": 1.987579584121704, + "learning_rate": 9.350237376155269e-06, + "loss": 0.4724, + "step": 16966 + }, + { + "epoch": 0.535152184198076, + "grad_norm": 1.9224364757537842, + "learning_rate": 9.349217981807041e-06, + "loss": 0.4374, + "step": 16967 + }, + { + "epoch": 0.5351837249645166, + "grad_norm": 1.6198346614837646, + "learning_rate": 9.348198594250227e-06, + "loss": 0.4433, + "step": 16968 + }, + { + "epoch": 0.5352152657309572, + "grad_norm": 1.7654179334640503, + "learning_rate": 9.34717921349547e-06, + "loss": 0.5099, + "step": 16969 + }, + { + "epoch": 0.5352468064973979, + "grad_norm": 2.0967142581939697, + "learning_rate": 9.346159839553402e-06, + "loss": 0.4576, + "step": 16970 + }, + { + "epoch": 0.5352783472638385, + "grad_norm": 1.946589708328247, + "learning_rate": 9.345140472434662e-06, + "loss": 0.4773, + "step": 16971 + }, + { + "epoch": 0.5353098880302791, + "grad_norm": 3.7667346000671387, + "learning_rate": 9.344121112149893e-06, + "loss": 0.3998, + "step": 16972 + }, + { + "epoch": 0.5353414287967198, + "grad_norm": 1.7154417037963867, + "learning_rate": 9.34310175870973e-06, + "loss": 0.3617, + "step": 16973 + }, + { + "epoch": 0.5353729695631604, + "grad_norm": 1.606871247291565, + "learning_rate": 9.342082412124804e-06, + "loss": 0.4551, + "step": 16974 + }, + { + "epoch": 0.535404510329601, + "grad_norm": 2.4006876945495605, + "learning_rate": 9.341063072405765e-06, + "loss": 0.4527, + "step": 16975 + }, + { + "epoch": 0.5354360510960416, + "grad_norm": 1.5100973844528198, + "learning_rate": 9.340043739563243e-06, + "loss": 0.7736, + "step": 16976 + }, + { + "epoch": 0.5354675918624823, + "grad_norm": 1.413320779800415, + "learning_rate": 9.339024413607878e-06, + "loss": 0.466, + "step": 16977 + }, + { + "epoch": 0.5354991326289229, + "grad_norm": 1.5671201944351196, + "learning_rate": 9.338005094550307e-06, + "loss": 0.4685, + "step": 16978 + }, + { + "epoch": 0.5355306733953635, + "grad_norm": 1.8511825799942017, + "learning_rate": 9.336985782401162e-06, + "loss": 0.4618, + "step": 16979 + }, + { + "epoch": 0.5355622141618042, + "grad_norm": 1.6311289072036743, + "learning_rate": 9.335966477171088e-06, + "loss": 0.4488, + "step": 16980 + }, + { + "epoch": 0.5355937549282448, + "grad_norm": 1.5050413608551025, + "learning_rate": 9.334947178870722e-06, + "loss": 0.4956, + "step": 16981 + }, + { + "epoch": 0.5356252956946854, + "grad_norm": 1.5511040687561035, + "learning_rate": 9.333927887510693e-06, + "loss": 0.4642, + "step": 16982 + }, + { + "epoch": 0.535656836461126, + "grad_norm": 1.2539397478103638, + "learning_rate": 9.332908603101648e-06, + "loss": 0.4352, + "step": 16983 + }, + { + "epoch": 0.5356883772275667, + "grad_norm": 1.800262689590454, + "learning_rate": 9.33188932565422e-06, + "loss": 0.4774, + "step": 16984 + }, + { + "epoch": 0.5357199179940072, + "grad_norm": 1.5889817476272583, + "learning_rate": 9.330870055179042e-06, + "loss": 0.4953, + "step": 16985 + }, + { + "epoch": 0.5357514587604478, + "grad_norm": 1.5673190355300903, + "learning_rate": 9.329850791686758e-06, + "loss": 0.4424, + "step": 16986 + }, + { + "epoch": 0.5357829995268885, + "grad_norm": 1.4777144193649292, + "learning_rate": 9.328831535188e-06, + "loss": 0.4691, + "step": 16987 + }, + { + "epoch": 0.5358145402933291, + "grad_norm": 1.5620183944702148, + "learning_rate": 9.327812285693407e-06, + "loss": 0.4704, + "step": 16988 + }, + { + "epoch": 0.5358460810597697, + "grad_norm": 1.257371425628662, + "learning_rate": 9.326793043213616e-06, + "loss": 0.4456, + "step": 16989 + }, + { + "epoch": 0.5358776218262103, + "grad_norm": 3.1523630619049072, + "learning_rate": 9.325773807759261e-06, + "loss": 0.7743, + "step": 16990 + }, + { + "epoch": 0.535909162592651, + "grad_norm": 1.3694241046905518, + "learning_rate": 9.324754579340976e-06, + "loss": 0.4608, + "step": 16991 + }, + { + "epoch": 0.5359407033590916, + "grad_norm": 1.8460909128189087, + "learning_rate": 9.323735357969406e-06, + "loss": 0.4772, + "step": 16992 + }, + { + "epoch": 0.5359722441255322, + "grad_norm": 1.374672770500183, + "learning_rate": 9.322716143655179e-06, + "loss": 0.7722, + "step": 16993 + }, + { + "epoch": 0.5360037848919729, + "grad_norm": 1.5680984258651733, + "learning_rate": 9.321696936408936e-06, + "loss": 0.451, + "step": 16994 + }, + { + "epoch": 0.5360353256584135, + "grad_norm": 1.4491615295410156, + "learning_rate": 9.320677736241316e-06, + "loss": 0.4291, + "step": 16995 + }, + { + "epoch": 0.5360668664248541, + "grad_norm": 1.4946998357772827, + "learning_rate": 9.319658543162949e-06, + "loss": 0.4808, + "step": 16996 + }, + { + "epoch": 0.5360984071912948, + "grad_norm": 1.5885024070739746, + "learning_rate": 9.318639357184474e-06, + "loss": 0.4464, + "step": 16997 + }, + { + "epoch": 0.5361299479577354, + "grad_norm": 1.582977533340454, + "learning_rate": 9.317620178316526e-06, + "loss": 0.7787, + "step": 16998 + }, + { + "epoch": 0.536161488724176, + "grad_norm": 1.730562448501587, + "learning_rate": 9.316601006569736e-06, + "loss": 0.478, + "step": 16999 + }, + { + "epoch": 0.5361930294906166, + "grad_norm": 1.309859037399292, + "learning_rate": 9.315581841954752e-06, + "loss": 0.4338, + "step": 17000 + }, + { + "epoch": 0.5362245702570573, + "grad_norm": 2.1929588317871094, + "learning_rate": 9.314562684482202e-06, + "loss": 0.7741, + "step": 17001 + }, + { + "epoch": 0.5362561110234979, + "grad_norm": 1.5704704523086548, + "learning_rate": 9.313543534162718e-06, + "loss": 0.4774, + "step": 17002 + }, + { + "epoch": 0.5362876517899385, + "grad_norm": 1.3845454454421997, + "learning_rate": 9.312524391006946e-06, + "loss": 0.3726, + "step": 17003 + }, + { + "epoch": 0.5363191925563792, + "grad_norm": 1.5500893592834473, + "learning_rate": 9.311505255025514e-06, + "loss": 0.4778, + "step": 17004 + }, + { + "epoch": 0.5363507333228198, + "grad_norm": 1.7474497556686401, + "learning_rate": 9.310486126229055e-06, + "loss": 0.4758, + "step": 17005 + }, + { + "epoch": 0.5363822740892603, + "grad_norm": 1.6456490755081177, + "learning_rate": 9.309467004628214e-06, + "loss": 0.4716, + "step": 17006 + }, + { + "epoch": 0.5364138148557009, + "grad_norm": 1.4097715616226196, + "learning_rate": 9.308447890233619e-06, + "loss": 0.4481, + "step": 17007 + }, + { + "epoch": 0.5364453556221416, + "grad_norm": 1.3807029724121094, + "learning_rate": 9.307428783055907e-06, + "loss": 0.4912, + "step": 17008 + }, + { + "epoch": 0.5364768963885822, + "grad_norm": 1.5225180387496948, + "learning_rate": 9.306409683105717e-06, + "loss": 0.5245, + "step": 17009 + }, + { + "epoch": 0.5365084371550228, + "grad_norm": 2.0230607986450195, + "learning_rate": 9.305390590393674e-06, + "loss": 0.4646, + "step": 17010 + }, + { + "epoch": 0.5365399779214635, + "grad_norm": 1.4163835048675537, + "learning_rate": 9.304371504930424e-06, + "loss": 0.4989, + "step": 17011 + }, + { + "epoch": 0.5365715186879041, + "grad_norm": 1.9375004768371582, + "learning_rate": 9.303352426726598e-06, + "loss": 0.4688, + "step": 17012 + }, + { + "epoch": 0.5366030594543447, + "grad_norm": 1.4083178043365479, + "learning_rate": 9.302333355792826e-06, + "loss": 0.4853, + "step": 17013 + }, + { + "epoch": 0.5366346002207854, + "grad_norm": 1.965578556060791, + "learning_rate": 9.301314292139751e-06, + "loss": 0.4586, + "step": 17014 + }, + { + "epoch": 0.536666140987226, + "grad_norm": 1.5844824314117432, + "learning_rate": 9.300295235778003e-06, + "loss": 0.4383, + "step": 17015 + }, + { + "epoch": 0.5366976817536666, + "grad_norm": 1.4485714435577393, + "learning_rate": 9.299276186718217e-06, + "loss": 0.4977, + "step": 17016 + }, + { + "epoch": 0.5367292225201072, + "grad_norm": 1.2733066082000732, + "learning_rate": 9.29825714497103e-06, + "loss": 0.459, + "step": 17017 + }, + { + "epoch": 0.5367607632865479, + "grad_norm": 1.855454683303833, + "learning_rate": 9.297238110547075e-06, + "loss": 0.4909, + "step": 17018 + }, + { + "epoch": 0.5367923040529885, + "grad_norm": 2.000945568084717, + "learning_rate": 9.29621908345698e-06, + "loss": 0.465, + "step": 17019 + }, + { + "epoch": 0.5368238448194291, + "grad_norm": 2.5479602813720703, + "learning_rate": 9.295200063711391e-06, + "loss": 0.7615, + "step": 17020 + }, + { + "epoch": 0.5368553855858698, + "grad_norm": 1.3719656467437744, + "learning_rate": 9.294181051320936e-06, + "loss": 0.489, + "step": 17021 + }, + { + "epoch": 0.5368869263523104, + "grad_norm": 1.5557249784469604, + "learning_rate": 9.293162046296246e-06, + "loss": 0.492, + "step": 17022 + }, + { + "epoch": 0.536918467118751, + "grad_norm": 1.739182472229004, + "learning_rate": 9.292143048647963e-06, + "loss": 0.4562, + "step": 17023 + }, + { + "epoch": 0.5369500078851916, + "grad_norm": 1.8618100881576538, + "learning_rate": 9.291124058386712e-06, + "loss": 0.5176, + "step": 17024 + }, + { + "epoch": 0.5369815486516323, + "grad_norm": 1.8844258785247803, + "learning_rate": 9.290105075523138e-06, + "loss": 0.4679, + "step": 17025 + }, + { + "epoch": 0.5370130894180729, + "grad_norm": 1.5799264907836914, + "learning_rate": 9.289086100067867e-06, + "loss": 0.461, + "step": 17026 + }, + { + "epoch": 0.5370446301845135, + "grad_norm": 1.8701555728912354, + "learning_rate": 9.288067132031531e-06, + "loss": 0.4842, + "step": 17027 + }, + { + "epoch": 0.5370761709509541, + "grad_norm": 1.8755745887756348, + "learning_rate": 9.287048171424773e-06, + "loss": 0.4923, + "step": 17028 + }, + { + "epoch": 0.5371077117173947, + "grad_norm": 1.215347409248352, + "learning_rate": 9.286029218258218e-06, + "loss": 0.4575, + "step": 17029 + }, + { + "epoch": 0.5371392524838353, + "grad_norm": 1.8625681400299072, + "learning_rate": 9.285010272542501e-06, + "loss": 0.4686, + "step": 17030 + }, + { + "epoch": 0.537170793250276, + "grad_norm": 2.0093982219696045, + "learning_rate": 9.28399133428826e-06, + "loss": 0.4492, + "step": 17031 + }, + { + "epoch": 0.5372023340167166, + "grad_norm": 1.6621794700622559, + "learning_rate": 9.282972403506124e-06, + "loss": 0.4531, + "step": 17032 + }, + { + "epoch": 0.5372338747831572, + "grad_norm": 1.8933384418487549, + "learning_rate": 9.281953480206725e-06, + "loss": 0.4616, + "step": 17033 + }, + { + "epoch": 0.5372654155495978, + "grad_norm": 1.4076437950134277, + "learning_rate": 9.280934564400702e-06, + "loss": 0.453, + "step": 17034 + }, + { + "epoch": 0.5372969563160385, + "grad_norm": 2.195627450942993, + "learning_rate": 9.279915656098685e-06, + "loss": 0.4475, + "step": 17035 + }, + { + "epoch": 0.5373284970824791, + "grad_norm": 2.6734392642974854, + "learning_rate": 9.278896755311308e-06, + "loss": 0.3619, + "step": 17036 + }, + { + "epoch": 0.5373600378489197, + "grad_norm": 1.804885745048523, + "learning_rate": 9.277877862049203e-06, + "loss": 0.4569, + "step": 17037 + }, + { + "epoch": 0.5373915786153604, + "grad_norm": 1.6954963207244873, + "learning_rate": 9.276858976323e-06, + "loss": 0.4637, + "step": 17038 + }, + { + "epoch": 0.537423119381801, + "grad_norm": 1.824113130569458, + "learning_rate": 9.275840098143339e-06, + "loss": 0.4914, + "step": 17039 + }, + { + "epoch": 0.5374546601482416, + "grad_norm": 1.7836731672286987, + "learning_rate": 9.274821227520848e-06, + "loss": 0.4934, + "step": 17040 + }, + { + "epoch": 0.5374862009146822, + "grad_norm": 1.4460917711257935, + "learning_rate": 9.273802364466158e-06, + "loss": 0.453, + "step": 17041 + }, + { + "epoch": 0.5375177416811229, + "grad_norm": 2.290531635284424, + "learning_rate": 9.272783508989907e-06, + "loss": 0.4081, + "step": 17042 + }, + { + "epoch": 0.5375492824475635, + "grad_norm": 1.4199020862579346, + "learning_rate": 9.271764661102725e-06, + "loss": 0.4787, + "step": 17043 + }, + { + "epoch": 0.5375808232140041, + "grad_norm": 1.6725574731826782, + "learning_rate": 9.27074582081524e-06, + "loss": 0.4731, + "step": 17044 + }, + { + "epoch": 0.5376123639804448, + "grad_norm": 1.7538509368896484, + "learning_rate": 9.269726988138092e-06, + "loss": 0.4571, + "step": 17045 + }, + { + "epoch": 0.5376439047468854, + "grad_norm": 2.112338066101074, + "learning_rate": 9.268708163081911e-06, + "loss": 0.4541, + "step": 17046 + }, + { + "epoch": 0.537675445513326, + "grad_norm": 1.6789861917495728, + "learning_rate": 9.267689345657326e-06, + "loss": 0.4438, + "step": 17047 + }, + { + "epoch": 0.5377069862797667, + "grad_norm": 1.7566399574279785, + "learning_rate": 9.266670535874974e-06, + "loss": 0.4538, + "step": 17048 + }, + { + "epoch": 0.5377385270462072, + "grad_norm": 1.7573333978652954, + "learning_rate": 9.265651733745485e-06, + "loss": 0.4725, + "step": 17049 + }, + { + "epoch": 0.5377700678126478, + "grad_norm": 1.4568127393722534, + "learning_rate": 9.264632939279484e-06, + "loss": 0.4971, + "step": 17050 + }, + { + "epoch": 0.5378016085790884, + "grad_norm": 1.3960723876953125, + "learning_rate": 9.263614152487615e-06, + "loss": 0.488, + "step": 17051 + }, + { + "epoch": 0.5378331493455291, + "grad_norm": 1.4423869848251343, + "learning_rate": 9.262595373380503e-06, + "loss": 0.4582, + "step": 17052 + }, + { + "epoch": 0.5378646901119697, + "grad_norm": 1.7605985403060913, + "learning_rate": 9.261576601968778e-06, + "loss": 0.7594, + "step": 17053 + }, + { + "epoch": 0.5378962308784103, + "grad_norm": 2.40385365486145, + "learning_rate": 9.260557838263076e-06, + "loss": 0.4254, + "step": 17054 + }, + { + "epoch": 0.537927771644851, + "grad_norm": 1.7271535396575928, + "learning_rate": 9.259539082274028e-06, + "loss": 0.4567, + "step": 17055 + }, + { + "epoch": 0.5379593124112916, + "grad_norm": 1.866380214691162, + "learning_rate": 9.258520334012265e-06, + "loss": 0.4226, + "step": 17056 + }, + { + "epoch": 0.5379908531777322, + "grad_norm": 1.9207903146743774, + "learning_rate": 9.257501593488419e-06, + "loss": 0.494, + "step": 17057 + }, + { + "epoch": 0.5380223939441728, + "grad_norm": 1.7489888668060303, + "learning_rate": 9.256482860713117e-06, + "loss": 0.4396, + "step": 17058 + }, + { + "epoch": 0.5380539347106135, + "grad_norm": 1.693225383758545, + "learning_rate": 9.255464135696997e-06, + "loss": 0.4386, + "step": 17059 + }, + { + "epoch": 0.5380854754770541, + "grad_norm": 1.7784782648086548, + "learning_rate": 9.254445418450687e-06, + "loss": 0.4642, + "step": 17060 + }, + { + "epoch": 0.5381170162434947, + "grad_norm": 1.4097691774368286, + "learning_rate": 9.253426708984811e-06, + "loss": 0.7637, + "step": 17061 + }, + { + "epoch": 0.5381485570099354, + "grad_norm": 1.3614366054534912, + "learning_rate": 9.252408007310015e-06, + "loss": 0.3665, + "step": 17062 + }, + { + "epoch": 0.538180097776376, + "grad_norm": 1.952073097229004, + "learning_rate": 9.25138931343692e-06, + "loss": 0.4456, + "step": 17063 + }, + { + "epoch": 0.5382116385428166, + "grad_norm": 1.6205904483795166, + "learning_rate": 9.250370627376155e-06, + "loss": 0.3348, + "step": 17064 + }, + { + "epoch": 0.5382431793092572, + "grad_norm": 1.7448256015777588, + "learning_rate": 9.249351949138359e-06, + "loss": 0.48, + "step": 17065 + }, + { + "epoch": 0.5382747200756979, + "grad_norm": 1.312108039855957, + "learning_rate": 9.248333278734158e-06, + "loss": 0.3728, + "step": 17066 + }, + { + "epoch": 0.5383062608421385, + "grad_norm": 2.3282411098480225, + "learning_rate": 9.24731461617418e-06, + "loss": 0.4594, + "step": 17067 + }, + { + "epoch": 0.5383378016085791, + "grad_norm": 1.751327633857727, + "learning_rate": 9.246295961469063e-06, + "loss": 0.4671, + "step": 17068 + }, + { + "epoch": 0.5383693423750198, + "grad_norm": 1.8100833892822266, + "learning_rate": 9.245277314629431e-06, + "loss": 0.362, + "step": 17069 + }, + { + "epoch": 0.5384008831414603, + "grad_norm": 1.6172680854797363, + "learning_rate": 9.244258675665913e-06, + "loss": 0.4907, + "step": 17070 + }, + { + "epoch": 0.5384324239079009, + "grad_norm": 1.6175397634506226, + "learning_rate": 9.243240044589145e-06, + "loss": 0.4907, + "step": 17071 + }, + { + "epoch": 0.5384639646743415, + "grad_norm": 1.1617292165756226, + "learning_rate": 9.242221421409752e-06, + "loss": 0.3849, + "step": 17072 + }, + { + "epoch": 0.5384955054407822, + "grad_norm": 1.8269213438034058, + "learning_rate": 9.241202806138372e-06, + "loss": 0.4852, + "step": 17073 + }, + { + "epoch": 0.5385270462072228, + "grad_norm": 1.6434860229492188, + "learning_rate": 9.240184198785629e-06, + "loss": 0.7724, + "step": 17074 + }, + { + "epoch": 0.5385585869736634, + "grad_norm": 1.7058627605438232, + "learning_rate": 9.23916559936215e-06, + "loss": 0.4848, + "step": 17075 + }, + { + "epoch": 0.5385901277401041, + "grad_norm": 1.6648634672164917, + "learning_rate": 9.238147007878573e-06, + "loss": 0.4634, + "step": 17076 + }, + { + "epoch": 0.5386216685065447, + "grad_norm": 1.7312713861465454, + "learning_rate": 9.237128424345522e-06, + "loss": 0.471, + "step": 17077 + }, + { + "epoch": 0.5386532092729853, + "grad_norm": 1.598463535308838, + "learning_rate": 9.236109848773626e-06, + "loss": 0.448, + "step": 17078 + }, + { + "epoch": 0.538684750039426, + "grad_norm": 1.781712532043457, + "learning_rate": 9.23509128117352e-06, + "loss": 0.4793, + "step": 17079 + }, + { + "epoch": 0.5387162908058666, + "grad_norm": 1.653867244720459, + "learning_rate": 9.23407272155583e-06, + "loss": 0.4351, + "step": 17080 + }, + { + "epoch": 0.5387478315723072, + "grad_norm": 1.4869320392608643, + "learning_rate": 9.233054169931183e-06, + "loss": 0.4636, + "step": 17081 + }, + { + "epoch": 0.5387793723387478, + "grad_norm": 1.722252368927002, + "learning_rate": 9.232035626310213e-06, + "loss": 0.4502, + "step": 17082 + }, + { + "epoch": 0.5388109131051885, + "grad_norm": 1.3903419971466064, + "learning_rate": 9.23101709070355e-06, + "loss": 0.3411, + "step": 17083 + }, + { + "epoch": 0.5388424538716291, + "grad_norm": 1.6003446578979492, + "learning_rate": 9.229998563121817e-06, + "loss": 0.5046, + "step": 17084 + }, + { + "epoch": 0.5388739946380697, + "grad_norm": 1.7519282102584839, + "learning_rate": 9.228980043575648e-06, + "loss": 0.4732, + "step": 17085 + }, + { + "epoch": 0.5389055354045104, + "grad_norm": 2.736888885498047, + "learning_rate": 9.22796153207567e-06, + "loss": 0.4822, + "step": 17086 + }, + { + "epoch": 0.538937076170951, + "grad_norm": 2.593135356903076, + "learning_rate": 9.226943028632516e-06, + "loss": 0.4867, + "step": 17087 + }, + { + "epoch": 0.5389686169373916, + "grad_norm": 2.3045592308044434, + "learning_rate": 9.225924533256812e-06, + "loss": 0.452, + "step": 17088 + }, + { + "epoch": 0.5390001577038323, + "grad_norm": 1.3742543458938599, + "learning_rate": 9.224906045959181e-06, + "loss": 0.7804, + "step": 17089 + }, + { + "epoch": 0.5390316984702729, + "grad_norm": 1.6271368265151978, + "learning_rate": 9.223887566750263e-06, + "loss": 0.4512, + "step": 17090 + }, + { + "epoch": 0.5390632392367134, + "grad_norm": 1.4603097438812256, + "learning_rate": 9.22286909564068e-06, + "loss": 0.4758, + "step": 17091 + }, + { + "epoch": 0.539094780003154, + "grad_norm": 1.392769694328308, + "learning_rate": 9.221850632641057e-06, + "loss": 0.4707, + "step": 17092 + }, + { + "epoch": 0.5391263207695947, + "grad_norm": 1.8208701610565186, + "learning_rate": 9.22083217776203e-06, + "loss": 0.4578, + "step": 17093 + }, + { + "epoch": 0.5391578615360353, + "grad_norm": 2.070909023284912, + "learning_rate": 9.219813731014226e-06, + "loss": 0.4145, + "step": 17094 + }, + { + "epoch": 0.5391894023024759, + "grad_norm": 1.5099703073501587, + "learning_rate": 9.218795292408269e-06, + "loss": 0.4926, + "step": 17095 + }, + { + "epoch": 0.5392209430689165, + "grad_norm": 3.461421012878418, + "learning_rate": 9.21777686195479e-06, + "loss": 0.3524, + "step": 17096 + }, + { + "epoch": 0.5392524838353572, + "grad_norm": 1.7192829847335815, + "learning_rate": 9.216758439664421e-06, + "loss": 0.485, + "step": 17097 + }, + { + "epoch": 0.5392840246017978, + "grad_norm": 1.425986409187317, + "learning_rate": 9.215740025547777e-06, + "loss": 0.4693, + "step": 17098 + }, + { + "epoch": 0.5393155653682384, + "grad_norm": 1.9538358449935913, + "learning_rate": 9.214721619615503e-06, + "loss": 0.4805, + "step": 17099 + }, + { + "epoch": 0.5393471061346791, + "grad_norm": 1.5659775733947754, + "learning_rate": 9.213703221878217e-06, + "loss": 0.3658, + "step": 17100 + }, + { + "epoch": 0.5393786469011197, + "grad_norm": 1.4162381887435913, + "learning_rate": 9.212684832346543e-06, + "loss": 0.4578, + "step": 17101 + }, + { + "epoch": 0.5394101876675603, + "grad_norm": 1.7779337167739868, + "learning_rate": 9.21166645103112e-06, + "loss": 0.4689, + "step": 17102 + }, + { + "epoch": 0.539441728434001, + "grad_norm": 2.1231255531311035, + "learning_rate": 9.210648077942567e-06, + "loss": 0.4554, + "step": 17103 + }, + { + "epoch": 0.5394732692004416, + "grad_norm": 1.4763579368591309, + "learning_rate": 9.209629713091516e-06, + "loss": 0.4608, + "step": 17104 + }, + { + "epoch": 0.5395048099668822, + "grad_norm": 1.4317375421524048, + "learning_rate": 9.20861135648859e-06, + "loss": 0.3343, + "step": 17105 + }, + { + "epoch": 0.5395363507333228, + "grad_norm": 1.8286690711975098, + "learning_rate": 9.20759300814442e-06, + "loss": 0.4552, + "step": 17106 + }, + { + "epoch": 0.5395678914997635, + "grad_norm": 3.1493074893951416, + "learning_rate": 9.206574668069634e-06, + "loss": 0.4524, + "step": 17107 + }, + { + "epoch": 0.5395994322662041, + "grad_norm": 1.7454384565353394, + "learning_rate": 9.205556336274858e-06, + "loss": 0.4568, + "step": 17108 + }, + { + "epoch": 0.5396309730326447, + "grad_norm": 3.4590442180633545, + "learning_rate": 9.204538012770714e-06, + "loss": 0.431, + "step": 17109 + }, + { + "epoch": 0.5396625137990854, + "grad_norm": 2.3535797595977783, + "learning_rate": 9.203519697567837e-06, + "loss": 0.4693, + "step": 17110 + }, + { + "epoch": 0.539694054565526, + "grad_norm": 1.4358118772506714, + "learning_rate": 9.202501390676852e-06, + "loss": 0.4713, + "step": 17111 + }, + { + "epoch": 0.5397255953319665, + "grad_norm": 1.6168056726455688, + "learning_rate": 9.201483092108379e-06, + "loss": 0.4627, + "step": 17112 + }, + { + "epoch": 0.5397571360984071, + "grad_norm": 2.000826358795166, + "learning_rate": 9.200464801873055e-06, + "loss": 0.3992, + "step": 17113 + }, + { + "epoch": 0.5397886768648478, + "grad_norm": 1.5412981510162354, + "learning_rate": 9.1994465199815e-06, + "loss": 0.4745, + "step": 17114 + }, + { + "epoch": 0.5398202176312884, + "grad_norm": 1.8341063261032104, + "learning_rate": 9.198428246444341e-06, + "loss": 0.4398, + "step": 17115 + }, + { + "epoch": 0.539851758397729, + "grad_norm": 1.5448498725891113, + "learning_rate": 9.197409981272207e-06, + "loss": 0.8117, + "step": 17116 + }, + { + "epoch": 0.5398832991641697, + "grad_norm": 1.7363831996917725, + "learning_rate": 9.196391724475724e-06, + "loss": 0.4483, + "step": 17117 + }, + { + "epoch": 0.5399148399306103, + "grad_norm": 2.0848946571350098, + "learning_rate": 9.195373476065513e-06, + "loss": 0.4693, + "step": 17118 + }, + { + "epoch": 0.5399463806970509, + "grad_norm": 1.7456140518188477, + "learning_rate": 9.194355236052209e-06, + "loss": 0.4598, + "step": 17119 + }, + { + "epoch": 0.5399779214634916, + "grad_norm": 1.9262747764587402, + "learning_rate": 9.193337004446428e-06, + "loss": 0.484, + "step": 17120 + }, + { + "epoch": 0.5400094622299322, + "grad_norm": 1.5801365375518799, + "learning_rate": 9.192318781258807e-06, + "loss": 0.4947, + "step": 17121 + }, + { + "epoch": 0.5400410029963728, + "grad_norm": 1.672154426574707, + "learning_rate": 9.191300566499968e-06, + "loss": 0.4629, + "step": 17122 + }, + { + "epoch": 0.5400725437628134, + "grad_norm": 1.6466423273086548, + "learning_rate": 9.190282360180532e-06, + "loss": 0.4569, + "step": 17123 + }, + { + "epoch": 0.5401040845292541, + "grad_norm": 1.687770128250122, + "learning_rate": 9.18926416231113e-06, + "loss": 0.4707, + "step": 17124 + }, + { + "epoch": 0.5401356252956947, + "grad_norm": 1.6398566961288452, + "learning_rate": 9.188245972902385e-06, + "loss": 0.3639, + "step": 17125 + }, + { + "epoch": 0.5401671660621353, + "grad_norm": 1.9184560775756836, + "learning_rate": 9.187227791964923e-06, + "loss": 0.453, + "step": 17126 + }, + { + "epoch": 0.540198706828576, + "grad_norm": 1.3464651107788086, + "learning_rate": 9.186209619509372e-06, + "loss": 0.3768, + "step": 17127 + }, + { + "epoch": 0.5402302475950166, + "grad_norm": 1.5028449296951294, + "learning_rate": 9.185191455546356e-06, + "loss": 0.5164, + "step": 17128 + }, + { + "epoch": 0.5402617883614572, + "grad_norm": 1.940720796585083, + "learning_rate": 9.184173300086494e-06, + "loss": 0.4937, + "step": 17129 + }, + { + "epoch": 0.5402933291278978, + "grad_norm": 1.618094563484192, + "learning_rate": 9.183155153140424e-06, + "loss": 0.7995, + "step": 17130 + }, + { + "epoch": 0.5403248698943385, + "grad_norm": 1.626794457435608, + "learning_rate": 9.182137014718762e-06, + "loss": 0.4737, + "step": 17131 + }, + { + "epoch": 0.5403564106607791, + "grad_norm": 1.7585090398788452, + "learning_rate": 9.181118884832131e-06, + "loss": 0.452, + "step": 17132 + }, + { + "epoch": 0.5403879514272196, + "grad_norm": 1.464752197265625, + "learning_rate": 9.180100763491167e-06, + "loss": 0.7563, + "step": 17133 + }, + { + "epoch": 0.5404194921936603, + "grad_norm": 1.9662021398544312, + "learning_rate": 9.179082650706484e-06, + "loss": 0.4349, + "step": 17134 + }, + { + "epoch": 0.5404510329601009, + "grad_norm": 5.711906433105469, + "learning_rate": 9.178064546488713e-06, + "loss": 0.4861, + "step": 17135 + }, + { + "epoch": 0.5404825737265415, + "grad_norm": 1.6629201173782349, + "learning_rate": 9.177046450848476e-06, + "loss": 0.4946, + "step": 17136 + }, + { + "epoch": 0.5405141144929821, + "grad_norm": 1.476407527923584, + "learning_rate": 9.176028363796397e-06, + "loss": 0.4697, + "step": 17137 + }, + { + "epoch": 0.5405456552594228, + "grad_norm": 1.6259818077087402, + "learning_rate": 9.175010285343104e-06, + "loss": 0.45, + "step": 17138 + }, + { + "epoch": 0.5405771960258634, + "grad_norm": 1.5466089248657227, + "learning_rate": 9.173992215499221e-06, + "loss": 0.4478, + "step": 17139 + }, + { + "epoch": 0.540608736792304, + "grad_norm": 4.0947723388671875, + "learning_rate": 9.172974154275363e-06, + "loss": 0.4686, + "step": 17140 + }, + { + "epoch": 0.5406402775587447, + "grad_norm": 1.689177393913269, + "learning_rate": 9.17195610168217e-06, + "loss": 0.4762, + "step": 17141 + }, + { + "epoch": 0.5406718183251853, + "grad_norm": 3.054115056991577, + "learning_rate": 9.170938057730255e-06, + "loss": 0.4388, + "step": 17142 + }, + { + "epoch": 0.5407033590916259, + "grad_norm": 1.662635326385498, + "learning_rate": 9.169920022430244e-06, + "loss": 0.4434, + "step": 17143 + }, + { + "epoch": 0.5407348998580666, + "grad_norm": 2.651654005050659, + "learning_rate": 9.168901995792764e-06, + "loss": 0.4492, + "step": 17144 + }, + { + "epoch": 0.5407664406245072, + "grad_norm": 1.6301575899124146, + "learning_rate": 9.167883977828435e-06, + "loss": 0.4286, + "step": 17145 + }, + { + "epoch": 0.5407979813909478, + "grad_norm": 1.918269157409668, + "learning_rate": 9.166865968547883e-06, + "loss": 0.5102, + "step": 17146 + }, + { + "epoch": 0.5408295221573884, + "grad_norm": 1.8161838054656982, + "learning_rate": 9.165847967961734e-06, + "loss": 0.4216, + "step": 17147 + }, + { + "epoch": 0.5408610629238291, + "grad_norm": 2.197047710418701, + "learning_rate": 9.164829976080609e-06, + "loss": 0.732, + "step": 17148 + }, + { + "epoch": 0.5408926036902697, + "grad_norm": 1.5009756088256836, + "learning_rate": 9.163811992915126e-06, + "loss": 0.4699, + "step": 17149 + }, + { + "epoch": 0.5409241444567103, + "grad_norm": 1.5689475536346436, + "learning_rate": 9.162794018475921e-06, + "loss": 0.4726, + "step": 17150 + }, + { + "epoch": 0.540955685223151, + "grad_norm": 1.5088567733764648, + "learning_rate": 9.161776052773605e-06, + "loss": 0.4787, + "step": 17151 + }, + { + "epoch": 0.5409872259895916, + "grad_norm": 1.5871379375457764, + "learning_rate": 9.160758095818812e-06, + "loss": 0.4709, + "step": 17152 + }, + { + "epoch": 0.5410187667560322, + "grad_norm": 1.4666706323623657, + "learning_rate": 9.159740147622158e-06, + "loss": 0.4585, + "step": 17153 + }, + { + "epoch": 0.5410503075224729, + "grad_norm": 1.6670571565628052, + "learning_rate": 9.158722208194265e-06, + "loss": 0.4222, + "step": 17154 + }, + { + "epoch": 0.5410818482889134, + "grad_norm": 2.693019151687622, + "learning_rate": 9.157704277545764e-06, + "loss": 0.4283, + "step": 17155 + }, + { + "epoch": 0.541113389055354, + "grad_norm": 1.5213537216186523, + "learning_rate": 9.156686355687273e-06, + "loss": 0.4606, + "step": 17156 + }, + { + "epoch": 0.5411449298217946, + "grad_norm": 1.6260238885879517, + "learning_rate": 9.155668442629408e-06, + "loss": 0.4827, + "step": 17157 + }, + { + "epoch": 0.5411764705882353, + "grad_norm": 1.295388102531433, + "learning_rate": 9.154650538382804e-06, + "loss": 0.4663, + "step": 17158 + }, + { + "epoch": 0.5412080113546759, + "grad_norm": 1.7336201667785645, + "learning_rate": 9.153632642958078e-06, + "loss": 0.4353, + "step": 17159 + }, + { + "epoch": 0.5412395521211165, + "grad_norm": 1.605380892753601, + "learning_rate": 9.152614756365849e-06, + "loss": 0.4606, + "step": 17160 + }, + { + "epoch": 0.5412710928875571, + "grad_norm": 2.1373414993286133, + "learning_rate": 9.151596878616746e-06, + "loss": 0.493, + "step": 17161 + }, + { + "epoch": 0.5413026336539978, + "grad_norm": 1.5637495517730713, + "learning_rate": 9.150579009721388e-06, + "loss": 0.4724, + "step": 17162 + }, + { + "epoch": 0.5413341744204384, + "grad_norm": 2.5470664501190186, + "learning_rate": 9.149561149690396e-06, + "loss": 0.463, + "step": 17163 + }, + { + "epoch": 0.541365715186879, + "grad_norm": 1.71058988571167, + "learning_rate": 9.148543298534397e-06, + "loss": 0.4503, + "step": 17164 + }, + { + "epoch": 0.5413972559533197, + "grad_norm": 3.0459961891174316, + "learning_rate": 9.147525456264006e-06, + "loss": 0.4881, + "step": 17165 + }, + { + "epoch": 0.5414287967197603, + "grad_norm": 1.9855248928070068, + "learning_rate": 9.146507622889849e-06, + "loss": 0.4419, + "step": 17166 + }, + { + "epoch": 0.5414603374862009, + "grad_norm": 3.409308671951294, + "learning_rate": 9.14548979842255e-06, + "loss": 0.4745, + "step": 17167 + }, + { + "epoch": 0.5414918782526416, + "grad_norm": 1.942522406578064, + "learning_rate": 9.144471982872727e-06, + "loss": 0.4801, + "step": 17168 + }, + { + "epoch": 0.5415234190190822, + "grad_norm": 1.706315279006958, + "learning_rate": 9.143454176251004e-06, + "loss": 0.465, + "step": 17169 + }, + { + "epoch": 0.5415549597855228, + "grad_norm": 1.5018084049224854, + "learning_rate": 9.142436378568002e-06, + "loss": 0.4417, + "step": 17170 + }, + { + "epoch": 0.5415865005519634, + "grad_norm": 2.430494546890259, + "learning_rate": 9.14141858983434e-06, + "loss": 0.4795, + "step": 17171 + }, + { + "epoch": 0.5416180413184041, + "grad_norm": 2.3520827293395996, + "learning_rate": 9.140400810060645e-06, + "loss": 0.4637, + "step": 17172 + }, + { + "epoch": 0.5416495820848447, + "grad_norm": 1.8142478466033936, + "learning_rate": 9.139383039257534e-06, + "loss": 0.4437, + "step": 17173 + }, + { + "epoch": 0.5416811228512853, + "grad_norm": 6.229289531707764, + "learning_rate": 9.138365277435628e-06, + "loss": 0.7289, + "step": 17174 + }, + { + "epoch": 0.541712663617726, + "grad_norm": 5.213923454284668, + "learning_rate": 9.137347524605551e-06, + "loss": 0.3989, + "step": 17175 + }, + { + "epoch": 0.5417442043841665, + "grad_norm": 1.617417812347412, + "learning_rate": 9.136329780777922e-06, + "loss": 0.4564, + "step": 17176 + }, + { + "epoch": 0.5417757451506071, + "grad_norm": 2.5588486194610596, + "learning_rate": 9.13531204596336e-06, + "loss": 0.4048, + "step": 17177 + }, + { + "epoch": 0.5418072859170477, + "grad_norm": 1.739140272140503, + "learning_rate": 9.134294320172491e-06, + "loss": 0.4659, + "step": 17178 + }, + { + "epoch": 0.5418388266834884, + "grad_norm": 2.1518912315368652, + "learning_rate": 9.133276603415933e-06, + "loss": 0.4302, + "step": 17179 + }, + { + "epoch": 0.541870367449929, + "grad_norm": 1.7942934036254883, + "learning_rate": 9.132258895704303e-06, + "loss": 0.4467, + "step": 17180 + }, + { + "epoch": 0.5419019082163696, + "grad_norm": 1.7969810962677002, + "learning_rate": 9.13124119704823e-06, + "loss": 0.4918, + "step": 17181 + }, + { + "epoch": 0.5419334489828103, + "grad_norm": 2.0177886486053467, + "learning_rate": 9.130223507458325e-06, + "loss": 0.4296, + "step": 17182 + }, + { + "epoch": 0.5419649897492509, + "grad_norm": 3.053966760635376, + "learning_rate": 9.129205826945218e-06, + "loss": 0.4561, + "step": 17183 + }, + { + "epoch": 0.5419965305156915, + "grad_norm": 1.517933964729309, + "learning_rate": 9.128188155519522e-06, + "loss": 0.4818, + "step": 17184 + }, + { + "epoch": 0.5420280712821322, + "grad_norm": 1.6705315113067627, + "learning_rate": 9.127170493191858e-06, + "loss": 0.4754, + "step": 17185 + }, + { + "epoch": 0.5420596120485728, + "grad_norm": 1.4658293724060059, + "learning_rate": 9.126152839972851e-06, + "loss": 0.4882, + "step": 17186 + }, + { + "epoch": 0.5420911528150134, + "grad_norm": 1.4721362590789795, + "learning_rate": 9.125135195873116e-06, + "loss": 0.46, + "step": 17187 + }, + { + "epoch": 0.542122693581454, + "grad_norm": 1.8389911651611328, + "learning_rate": 9.124117560903272e-06, + "loss": 0.4685, + "step": 17188 + }, + { + "epoch": 0.5421542343478947, + "grad_norm": 1.9665635824203491, + "learning_rate": 9.123099935073945e-06, + "loss": 0.471, + "step": 17189 + }, + { + "epoch": 0.5421857751143353, + "grad_norm": 1.723971962928772, + "learning_rate": 9.12208231839575e-06, + "loss": 0.446, + "step": 17190 + }, + { + "epoch": 0.5422173158807759, + "grad_norm": 1.4615478515625, + "learning_rate": 9.121064710879306e-06, + "loss": 0.7599, + "step": 17191 + }, + { + "epoch": 0.5422488566472166, + "grad_norm": 1.296539068222046, + "learning_rate": 9.120047112535236e-06, + "loss": 0.4542, + "step": 17192 + }, + { + "epoch": 0.5422803974136572, + "grad_norm": 1.258601188659668, + "learning_rate": 9.119029523374158e-06, + "loss": 0.445, + "step": 17193 + }, + { + "epoch": 0.5423119381800978, + "grad_norm": 1.5363147258758545, + "learning_rate": 9.11801194340669e-06, + "loss": 0.4367, + "step": 17194 + }, + { + "epoch": 0.5423434789465384, + "grad_norm": 1.6648598909378052, + "learning_rate": 9.116994372643454e-06, + "loss": 0.4662, + "step": 17195 + }, + { + "epoch": 0.5423750197129791, + "grad_norm": 1.6400882005691528, + "learning_rate": 9.115976811095067e-06, + "loss": 0.4659, + "step": 17196 + }, + { + "epoch": 0.5424065604794196, + "grad_norm": 1.5136321783065796, + "learning_rate": 9.114959258772144e-06, + "loss": 0.4679, + "step": 17197 + }, + { + "epoch": 0.5424381012458602, + "grad_norm": 1.576060175895691, + "learning_rate": 9.113941715685313e-06, + "loss": 0.4774, + "step": 17198 + }, + { + "epoch": 0.5424696420123009, + "grad_norm": 1.3804941177368164, + "learning_rate": 9.112924181845185e-06, + "loss": 0.4189, + "step": 17199 + }, + { + "epoch": 0.5425011827787415, + "grad_norm": 2.065535545349121, + "learning_rate": 9.111906657262384e-06, + "loss": 0.4858, + "step": 17200 + }, + { + "epoch": 0.5425327235451821, + "grad_norm": 1.6572872400283813, + "learning_rate": 9.110889141947527e-06, + "loss": 0.4542, + "step": 17201 + }, + { + "epoch": 0.5425642643116227, + "grad_norm": 1.3821830749511719, + "learning_rate": 9.109871635911232e-06, + "loss": 0.4353, + "step": 17202 + }, + { + "epoch": 0.5425958050780634, + "grad_norm": 1.7234416007995605, + "learning_rate": 9.108854139164118e-06, + "loss": 0.4654, + "step": 17203 + }, + { + "epoch": 0.542627345844504, + "grad_norm": 1.908588171005249, + "learning_rate": 9.107836651716804e-06, + "loss": 0.4786, + "step": 17204 + }, + { + "epoch": 0.5426588866109446, + "grad_norm": 2.8025927543640137, + "learning_rate": 9.106819173579901e-06, + "loss": 0.4827, + "step": 17205 + }, + { + "epoch": 0.5426904273773853, + "grad_norm": 1.1663516759872437, + "learning_rate": 9.10580170476404e-06, + "loss": 0.756, + "step": 17206 + }, + { + "epoch": 0.5427219681438259, + "grad_norm": 1.8732404708862305, + "learning_rate": 9.104784245279833e-06, + "loss": 0.4738, + "step": 17207 + }, + { + "epoch": 0.5427535089102665, + "grad_norm": 2.7053890228271484, + "learning_rate": 9.103766795137891e-06, + "loss": 0.433, + "step": 17208 + }, + { + "epoch": 0.5427850496767072, + "grad_norm": 1.6577560901641846, + "learning_rate": 9.102749354348845e-06, + "loss": 0.7727, + "step": 17209 + }, + { + "epoch": 0.5428165904431478, + "grad_norm": 1.9560145139694214, + "learning_rate": 9.101731922923304e-06, + "loss": 0.4944, + "step": 17210 + }, + { + "epoch": 0.5428481312095884, + "grad_norm": 7.328649997711182, + "learning_rate": 9.100714500871884e-06, + "loss": 0.3738, + "step": 17211 + }, + { + "epoch": 0.542879671976029, + "grad_norm": 2.3316285610198975, + "learning_rate": 9.099697088205212e-06, + "loss": 0.4736, + "step": 17212 + }, + { + "epoch": 0.5429112127424697, + "grad_norm": 3.269761562347412, + "learning_rate": 9.098679684933896e-06, + "loss": 0.7736, + "step": 17213 + }, + { + "epoch": 0.5429427535089103, + "grad_norm": 1.6516718864440918, + "learning_rate": 9.097662291068561e-06, + "loss": 0.458, + "step": 17214 + }, + { + "epoch": 0.5429742942753509, + "grad_norm": 1.794272541999817, + "learning_rate": 9.09664490661982e-06, + "loss": 0.4546, + "step": 17215 + }, + { + "epoch": 0.5430058350417916, + "grad_norm": 1.5843528509140015, + "learning_rate": 9.095627531598288e-06, + "loss": 0.4709, + "step": 17216 + }, + { + "epoch": 0.5430373758082322, + "grad_norm": 1.8853708505630493, + "learning_rate": 9.094610166014588e-06, + "loss": 0.3551, + "step": 17217 + }, + { + "epoch": 0.5430689165746727, + "grad_norm": 2.1149473190307617, + "learning_rate": 9.093592809879333e-06, + "loss": 0.4692, + "step": 17218 + }, + { + "epoch": 0.5431004573411133, + "grad_norm": 1.6489715576171875, + "learning_rate": 9.092575463203137e-06, + "loss": 0.4385, + "step": 17219 + }, + { + "epoch": 0.543131998107554, + "grad_norm": 1.890377163887024, + "learning_rate": 9.091558125996625e-06, + "loss": 0.4516, + "step": 17220 + }, + { + "epoch": 0.5431635388739946, + "grad_norm": 3.361361503601074, + "learning_rate": 9.09054079827041e-06, + "loss": 0.4482, + "step": 17221 + }, + { + "epoch": 0.5431950796404352, + "grad_norm": 1.5486476421356201, + "learning_rate": 9.089523480035106e-06, + "loss": 0.4452, + "step": 17222 + }, + { + "epoch": 0.5432266204068759, + "grad_norm": 1.651747703552246, + "learning_rate": 9.088506171301334e-06, + "loss": 0.4925, + "step": 17223 + }, + { + "epoch": 0.5432581611733165, + "grad_norm": 1.4076043367385864, + "learning_rate": 9.087488872079706e-06, + "loss": 0.4625, + "step": 17224 + }, + { + "epoch": 0.5432897019397571, + "grad_norm": 1.4068957567214966, + "learning_rate": 9.086471582380838e-06, + "loss": 0.4753, + "step": 17225 + }, + { + "epoch": 0.5433212427061977, + "grad_norm": 2.0204732418060303, + "learning_rate": 9.08545430221535e-06, + "loss": 0.4793, + "step": 17226 + }, + { + "epoch": 0.5433527834726384, + "grad_norm": 1.6285489797592163, + "learning_rate": 9.08443703159386e-06, + "loss": 0.496, + "step": 17227 + }, + { + "epoch": 0.543384324239079, + "grad_norm": 1.3178898096084595, + "learning_rate": 9.083419770526974e-06, + "loss": 0.4675, + "step": 17228 + }, + { + "epoch": 0.5434158650055196, + "grad_norm": 4.0906662940979, + "learning_rate": 9.082402519025317e-06, + "loss": 0.343, + "step": 17229 + }, + { + "epoch": 0.5434474057719603, + "grad_norm": 1.8570449352264404, + "learning_rate": 9.081385277099503e-06, + "loss": 0.4416, + "step": 17230 + }, + { + "epoch": 0.5434789465384009, + "grad_norm": 3.2494683265686035, + "learning_rate": 9.080368044760146e-06, + "loss": 0.4599, + "step": 17231 + }, + { + "epoch": 0.5435104873048415, + "grad_norm": 1.3445534706115723, + "learning_rate": 9.079350822017863e-06, + "loss": 0.5079, + "step": 17232 + }, + { + "epoch": 0.5435420280712822, + "grad_norm": 1.567407250404358, + "learning_rate": 9.078333608883269e-06, + "loss": 0.8076, + "step": 17233 + }, + { + "epoch": 0.5435735688377228, + "grad_norm": 1.7233541011810303, + "learning_rate": 9.07731640536698e-06, + "loss": 0.4487, + "step": 17234 + }, + { + "epoch": 0.5436051096041634, + "grad_norm": 2.2286531925201416, + "learning_rate": 9.076299211479613e-06, + "loss": 0.4532, + "step": 17235 + }, + { + "epoch": 0.543636650370604, + "grad_norm": 1.3738025426864624, + "learning_rate": 9.075282027231776e-06, + "loss": 0.4718, + "step": 17236 + }, + { + "epoch": 0.5436681911370447, + "grad_norm": 1.649274468421936, + "learning_rate": 9.074264852634092e-06, + "loss": 0.4797, + "step": 17237 + }, + { + "epoch": 0.5436997319034853, + "grad_norm": 1.504455804824829, + "learning_rate": 9.073247687697175e-06, + "loss": 0.3526, + "step": 17238 + }, + { + "epoch": 0.5437312726699258, + "grad_norm": 1.6309093236923218, + "learning_rate": 9.072230532431632e-06, + "loss": 0.4358, + "step": 17239 + }, + { + "epoch": 0.5437628134363665, + "grad_norm": 2.4313302040100098, + "learning_rate": 9.07121338684809e-06, + "loss": 0.4354, + "step": 17240 + }, + { + "epoch": 0.5437943542028071, + "grad_norm": 1.3989918231964111, + "learning_rate": 9.070196250957155e-06, + "loss": 0.4551, + "step": 17241 + }, + { + "epoch": 0.5438258949692477, + "grad_norm": 1.5926839113235474, + "learning_rate": 9.069179124769444e-06, + "loss": 0.4329, + "step": 17242 + }, + { + "epoch": 0.5438574357356883, + "grad_norm": 1.3831071853637695, + "learning_rate": 9.068162008295573e-06, + "loss": 0.4833, + "step": 17243 + }, + { + "epoch": 0.543888976502129, + "grad_norm": 1.6224087476730347, + "learning_rate": 9.067144901546155e-06, + "loss": 0.4447, + "step": 17244 + }, + { + "epoch": 0.5439205172685696, + "grad_norm": 1.469874382019043, + "learning_rate": 9.066127804531799e-06, + "loss": 0.4568, + "step": 17245 + }, + { + "epoch": 0.5439520580350102, + "grad_norm": 1.407821774482727, + "learning_rate": 9.065110717263129e-06, + "loss": 0.4374, + "step": 17246 + }, + { + "epoch": 0.5439835988014509, + "grad_norm": 11.693796157836914, + "learning_rate": 9.064093639750752e-06, + "loss": 0.4918, + "step": 17247 + }, + { + "epoch": 0.5440151395678915, + "grad_norm": 1.5857490301132202, + "learning_rate": 9.063076572005287e-06, + "loss": 0.3737, + "step": 17248 + }, + { + "epoch": 0.5440466803343321, + "grad_norm": 1.6146655082702637, + "learning_rate": 9.062059514037347e-06, + "loss": 0.4504, + "step": 17249 + }, + { + "epoch": 0.5440782211007728, + "grad_norm": 1.86955726146698, + "learning_rate": 9.061042465857542e-06, + "loss": 0.4867, + "step": 17250 + }, + { + "epoch": 0.5441097618672134, + "grad_norm": 1.5151100158691406, + "learning_rate": 9.060025427476486e-06, + "loss": 0.4697, + "step": 17251 + }, + { + "epoch": 0.544141302633654, + "grad_norm": 2.7014853954315186, + "learning_rate": 9.059008398904799e-06, + "loss": 0.4524, + "step": 17252 + }, + { + "epoch": 0.5441728434000946, + "grad_norm": 1.7877024412155151, + "learning_rate": 9.057991380153087e-06, + "loss": 0.4528, + "step": 17253 + }, + { + "epoch": 0.5442043841665353, + "grad_norm": 1.7311227321624756, + "learning_rate": 9.056974371231968e-06, + "loss": 0.4794, + "step": 17254 + }, + { + "epoch": 0.5442359249329759, + "grad_norm": 1.841253638267517, + "learning_rate": 9.055957372152054e-06, + "loss": 0.4424, + "step": 17255 + }, + { + "epoch": 0.5442674656994165, + "grad_norm": 1.5180734395980835, + "learning_rate": 9.054940382923954e-06, + "loss": 0.4515, + "step": 17256 + }, + { + "epoch": 0.5442990064658572, + "grad_norm": 2.332448959350586, + "learning_rate": 9.05392340355829e-06, + "loss": 0.4458, + "step": 17257 + }, + { + "epoch": 0.5443305472322978, + "grad_norm": 1.5109771490097046, + "learning_rate": 9.052906434065667e-06, + "loss": 0.4441, + "step": 17258 + }, + { + "epoch": 0.5443620879987384, + "grad_norm": 1.6672871112823486, + "learning_rate": 9.051889474456698e-06, + "loss": 0.4829, + "step": 17259 + }, + { + "epoch": 0.5443936287651789, + "grad_norm": 1.5411572456359863, + "learning_rate": 9.050872524742003e-06, + "loss": 0.454, + "step": 17260 + }, + { + "epoch": 0.5444251695316196, + "grad_norm": 1.3897966146469116, + "learning_rate": 9.04985558493219e-06, + "loss": 0.4726, + "step": 17261 + }, + { + "epoch": 0.5444567102980602, + "grad_norm": 1.8186986446380615, + "learning_rate": 9.04883865503787e-06, + "loss": 0.4452, + "step": 17262 + }, + { + "epoch": 0.5444882510645008, + "grad_norm": 1.8124866485595703, + "learning_rate": 9.04782173506966e-06, + "loss": 0.4772, + "step": 17263 + }, + { + "epoch": 0.5445197918309415, + "grad_norm": 2.6287782192230225, + "learning_rate": 9.046804825038165e-06, + "loss": 0.3879, + "step": 17264 + }, + { + "epoch": 0.5445513325973821, + "grad_norm": 1.463701605796814, + "learning_rate": 9.045787924954005e-06, + "loss": 0.7813, + "step": 17265 + }, + { + "epoch": 0.5445828733638227, + "grad_norm": 1.7651612758636475, + "learning_rate": 9.04477103482779e-06, + "loss": 0.4479, + "step": 17266 + }, + { + "epoch": 0.5446144141302633, + "grad_norm": 1.5839108228683472, + "learning_rate": 9.043754154670126e-06, + "loss": 0.7427, + "step": 17267 + }, + { + "epoch": 0.544645954896704, + "grad_norm": 1.6160756349563599, + "learning_rate": 9.042737284491633e-06, + "loss": 0.4313, + "step": 17268 + }, + { + "epoch": 0.5446774956631446, + "grad_norm": 1.5466116666793823, + "learning_rate": 9.041720424302922e-06, + "loss": 0.7641, + "step": 17269 + }, + { + "epoch": 0.5447090364295852, + "grad_norm": 1.638330340385437, + "learning_rate": 9.0407035741146e-06, + "loss": 0.4699, + "step": 17270 + }, + { + "epoch": 0.5447405771960259, + "grad_norm": 1.45685613155365, + "learning_rate": 9.03968673393728e-06, + "loss": 0.4264, + "step": 17271 + }, + { + "epoch": 0.5447721179624665, + "grad_norm": 1.7358546257019043, + "learning_rate": 9.038669903781577e-06, + "loss": 0.4409, + "step": 17272 + }, + { + "epoch": 0.5448036587289071, + "grad_norm": 1.719589352607727, + "learning_rate": 9.037653083658098e-06, + "loss": 0.4547, + "step": 17273 + }, + { + "epoch": 0.5448351994953478, + "grad_norm": 1.6711786985397339, + "learning_rate": 9.036636273577459e-06, + "loss": 0.4489, + "step": 17274 + }, + { + "epoch": 0.5448667402617884, + "grad_norm": 1.2497085332870483, + "learning_rate": 9.035619473550268e-06, + "loss": 0.3568, + "step": 17275 + }, + { + "epoch": 0.544898281028229, + "grad_norm": 2.147747039794922, + "learning_rate": 9.034602683587132e-06, + "loss": 0.5001, + "step": 17276 + }, + { + "epoch": 0.5449298217946696, + "grad_norm": 1.7665507793426514, + "learning_rate": 9.033585903698672e-06, + "loss": 0.4544, + "step": 17277 + }, + { + "epoch": 0.5449613625611103, + "grad_norm": 1.7231439352035522, + "learning_rate": 9.032569133895488e-06, + "loss": 0.5022, + "step": 17278 + }, + { + "epoch": 0.5449929033275509, + "grad_norm": 34.08273696899414, + "learning_rate": 9.031552374188201e-06, + "loss": 0.3597, + "step": 17279 + }, + { + "epoch": 0.5450244440939915, + "grad_norm": 1.4139320850372314, + "learning_rate": 9.030535624587416e-06, + "loss": 0.4843, + "step": 17280 + }, + { + "epoch": 0.5450559848604322, + "grad_norm": 2.017695188522339, + "learning_rate": 9.029518885103744e-06, + "loss": 0.46, + "step": 17281 + }, + { + "epoch": 0.5450875256268727, + "grad_norm": 2.2160003185272217, + "learning_rate": 9.028502155747797e-06, + "loss": 0.4693, + "step": 17282 + }, + { + "epoch": 0.5451190663933133, + "grad_norm": 1.6621530055999756, + "learning_rate": 9.027485436530186e-06, + "loss": 0.4443, + "step": 17283 + }, + { + "epoch": 0.5451506071597539, + "grad_norm": 1.5609939098358154, + "learning_rate": 9.026468727461514e-06, + "loss": 0.4397, + "step": 17284 + }, + { + "epoch": 0.5451821479261946, + "grad_norm": 2.5664613246917725, + "learning_rate": 9.0254520285524e-06, + "loss": 0.3653, + "step": 17285 + }, + { + "epoch": 0.5452136886926352, + "grad_norm": 1.4333367347717285, + "learning_rate": 9.024435339813454e-06, + "loss": 0.4364, + "step": 17286 + }, + { + "epoch": 0.5452452294590758, + "grad_norm": 1.6311709880828857, + "learning_rate": 9.023418661255276e-06, + "loss": 0.4359, + "step": 17287 + }, + { + "epoch": 0.5452767702255165, + "grad_norm": 1.4888306856155396, + "learning_rate": 9.022401992888487e-06, + "loss": 0.4801, + "step": 17288 + }, + { + "epoch": 0.5453083109919571, + "grad_norm": 1.7192643880844116, + "learning_rate": 9.021385334723692e-06, + "loss": 0.7945, + "step": 17289 + }, + { + "epoch": 0.5453398517583977, + "grad_norm": 1.6752256155014038, + "learning_rate": 9.0203686867715e-06, + "loss": 0.4429, + "step": 17290 + }, + { + "epoch": 0.5453713925248383, + "grad_norm": 1.8465092182159424, + "learning_rate": 9.01935204904252e-06, + "loss": 0.4544, + "step": 17291 + }, + { + "epoch": 0.545402933291279, + "grad_norm": 1.5316210985183716, + "learning_rate": 9.018335421547366e-06, + "loss": 0.4501, + "step": 17292 + }, + { + "epoch": 0.5454344740577196, + "grad_norm": 1.434967279434204, + "learning_rate": 9.017318804296642e-06, + "loss": 0.466, + "step": 17293 + }, + { + "epoch": 0.5454660148241602, + "grad_norm": 1.5125528573989868, + "learning_rate": 9.01630219730096e-06, + "loss": 0.4636, + "step": 17294 + }, + { + "epoch": 0.5454975555906009, + "grad_norm": 1.7688058614730835, + "learning_rate": 9.015285600570926e-06, + "loss": 0.4821, + "step": 17295 + }, + { + "epoch": 0.5455290963570415, + "grad_norm": 1.5665886402130127, + "learning_rate": 9.014269014117156e-06, + "loss": 0.4717, + "step": 17296 + }, + { + "epoch": 0.5455606371234821, + "grad_norm": 1.7241318225860596, + "learning_rate": 9.013252437950253e-06, + "loss": 0.4698, + "step": 17297 + }, + { + "epoch": 0.5455921778899228, + "grad_norm": 1.8691027164459229, + "learning_rate": 9.012235872080822e-06, + "loss": 0.4614, + "step": 17298 + }, + { + "epoch": 0.5456237186563634, + "grad_norm": 2.273064136505127, + "learning_rate": 9.011219316519481e-06, + "loss": 0.4849, + "step": 17299 + }, + { + "epoch": 0.545655259422804, + "grad_norm": 1.2428563833236694, + "learning_rate": 9.010202771276835e-06, + "loss": 0.7799, + "step": 17300 + }, + { + "epoch": 0.5456868001892446, + "grad_norm": 1.4636447429656982, + "learning_rate": 9.00918623636349e-06, + "loss": 0.4624, + "step": 17301 + }, + { + "epoch": 0.5457183409556853, + "grad_norm": 1.5853742361068726, + "learning_rate": 9.008169711790058e-06, + "loss": 0.4908, + "step": 17302 + }, + { + "epoch": 0.5457498817221258, + "grad_norm": 1.3870242834091187, + "learning_rate": 9.007153197567146e-06, + "loss": 0.4732, + "step": 17303 + }, + { + "epoch": 0.5457814224885664, + "grad_norm": 1.5651615858078003, + "learning_rate": 9.006136693705356e-06, + "loss": 0.4532, + "step": 17304 + }, + { + "epoch": 0.545812963255007, + "grad_norm": 1.3376500606536865, + "learning_rate": 9.005120200215305e-06, + "loss": 0.4727, + "step": 17305 + }, + { + "epoch": 0.5458445040214477, + "grad_norm": 1.4417431354522705, + "learning_rate": 9.004103717107597e-06, + "loss": 0.4641, + "step": 17306 + }, + { + "epoch": 0.5458760447878883, + "grad_norm": 5.956803321838379, + "learning_rate": 9.003087244392837e-06, + "loss": 0.4309, + "step": 17307 + }, + { + "epoch": 0.545907585554329, + "grad_norm": 1.58407461643219, + "learning_rate": 9.002070782081639e-06, + "loss": 0.4646, + "step": 17308 + }, + { + "epoch": 0.5459391263207696, + "grad_norm": 1.6517293453216553, + "learning_rate": 9.001054330184605e-06, + "loss": 0.4895, + "step": 17309 + }, + { + "epoch": 0.5459706670872102, + "grad_norm": 2.3786661624908447, + "learning_rate": 9.000037888712347e-06, + "loss": 0.4779, + "step": 17310 + }, + { + "epoch": 0.5460022078536508, + "grad_norm": 1.3503975868225098, + "learning_rate": 8.99902145767547e-06, + "loss": 0.4571, + "step": 17311 + }, + { + "epoch": 0.5460337486200915, + "grad_norm": 1.8747121095657349, + "learning_rate": 8.99800503708458e-06, + "loss": 0.4643, + "step": 17312 + }, + { + "epoch": 0.5460652893865321, + "grad_norm": 1.5822776556015015, + "learning_rate": 8.996988626950287e-06, + "loss": 0.7882, + "step": 17313 + }, + { + "epoch": 0.5460968301529727, + "grad_norm": 1.947536826133728, + "learning_rate": 8.995972227283195e-06, + "loss": 0.4845, + "step": 17314 + }, + { + "epoch": 0.5461283709194134, + "grad_norm": 1.6938940286636353, + "learning_rate": 8.99495583809391e-06, + "loss": 0.4993, + "step": 17315 + }, + { + "epoch": 0.546159911685854, + "grad_norm": 6.300816535949707, + "learning_rate": 8.993939459393046e-06, + "loss": 0.4511, + "step": 17316 + }, + { + "epoch": 0.5461914524522946, + "grad_norm": 5.484801769256592, + "learning_rate": 8.992923091191204e-06, + "loss": 0.3485, + "step": 17317 + }, + { + "epoch": 0.5462229932187352, + "grad_norm": 2.32838773727417, + "learning_rate": 8.991906733498988e-06, + "loss": 0.4435, + "step": 17318 + }, + { + "epoch": 0.5462545339851759, + "grad_norm": 1.787109375, + "learning_rate": 8.990890386327011e-06, + "loss": 0.4843, + "step": 17319 + }, + { + "epoch": 0.5462860747516165, + "grad_norm": 1.697432041168213, + "learning_rate": 8.989874049685878e-06, + "loss": 0.5087, + "step": 17320 + }, + { + "epoch": 0.5463176155180571, + "grad_norm": 1.64649498462677, + "learning_rate": 8.988857723586191e-06, + "loss": 0.4631, + "step": 17321 + }, + { + "epoch": 0.5463491562844978, + "grad_norm": 3.190056085586548, + "learning_rate": 8.98784140803856e-06, + "loss": 0.4581, + "step": 17322 + }, + { + "epoch": 0.5463806970509384, + "grad_norm": 1.890228033065796, + "learning_rate": 8.986825103053591e-06, + "loss": 0.4467, + "step": 17323 + }, + { + "epoch": 0.5464122378173789, + "grad_norm": 1.4681676626205444, + "learning_rate": 8.985808808641883e-06, + "loss": 0.4569, + "step": 17324 + }, + { + "epoch": 0.5464437785838195, + "grad_norm": 1.500846266746521, + "learning_rate": 8.984792524814054e-06, + "loss": 0.4298, + "step": 17325 + }, + { + "epoch": 0.5464753193502602, + "grad_norm": 2.1587319374084473, + "learning_rate": 8.983776251580698e-06, + "loss": 0.4245, + "step": 17326 + }, + { + "epoch": 0.5465068601167008, + "grad_norm": 2.9563181400299072, + "learning_rate": 8.98275998895243e-06, + "loss": 0.4412, + "step": 17327 + }, + { + "epoch": 0.5465384008831414, + "grad_norm": 1.4058698415756226, + "learning_rate": 8.98174373693985e-06, + "loss": 0.349, + "step": 17328 + }, + { + "epoch": 0.5465699416495821, + "grad_norm": 1.3703458309173584, + "learning_rate": 8.980727495553565e-06, + "loss": 0.4805, + "step": 17329 + }, + { + "epoch": 0.5466014824160227, + "grad_norm": 1.965565800666809, + "learning_rate": 8.97971126480418e-06, + "loss": 0.4494, + "step": 17330 + }, + { + "epoch": 0.5466330231824633, + "grad_norm": 1.3928018808364868, + "learning_rate": 8.9786950447023e-06, + "loss": 0.4458, + "step": 17331 + }, + { + "epoch": 0.546664563948904, + "grad_norm": 1.2817658185958862, + "learning_rate": 8.977678835258529e-06, + "loss": 0.4382, + "step": 17332 + }, + { + "epoch": 0.5466961047153446, + "grad_norm": 1.9189268350601196, + "learning_rate": 8.976662636483477e-06, + "loss": 0.4793, + "step": 17333 + }, + { + "epoch": 0.5467276454817852, + "grad_norm": 1.7082269191741943, + "learning_rate": 8.975646448387743e-06, + "loss": 0.4717, + "step": 17334 + }, + { + "epoch": 0.5467591862482258, + "grad_norm": 1.34307861328125, + "learning_rate": 8.97463027098193e-06, + "loss": 0.4987, + "step": 17335 + }, + { + "epoch": 0.5467907270146665, + "grad_norm": 1.5983150005340576, + "learning_rate": 8.97361410427665e-06, + "loss": 0.4256, + "step": 17336 + }, + { + "epoch": 0.5468222677811071, + "grad_norm": 1.6070668697357178, + "learning_rate": 8.972597948282505e-06, + "loss": 0.3067, + "step": 17337 + }, + { + "epoch": 0.5468538085475477, + "grad_norm": 1.7301136255264282, + "learning_rate": 8.971581803010092e-06, + "loss": 0.4508, + "step": 17338 + }, + { + "epoch": 0.5468853493139884, + "grad_norm": 1.4479392766952515, + "learning_rate": 8.970565668470028e-06, + "loss": 0.7686, + "step": 17339 + }, + { + "epoch": 0.546916890080429, + "grad_norm": 2.183931350708008, + "learning_rate": 8.96954954467291e-06, + "loss": 0.4813, + "step": 17340 + }, + { + "epoch": 0.5469484308468696, + "grad_norm": 1.862917184829712, + "learning_rate": 8.96853343162934e-06, + "loss": 0.4838, + "step": 17341 + }, + { + "epoch": 0.5469799716133102, + "grad_norm": 1.69503653049469, + "learning_rate": 8.967517329349927e-06, + "loss": 0.4538, + "step": 17342 + }, + { + "epoch": 0.5470115123797509, + "grad_norm": 1.4605798721313477, + "learning_rate": 8.966501237845268e-06, + "loss": 0.4504, + "step": 17343 + }, + { + "epoch": 0.5470430531461915, + "grad_norm": 1.6228172779083252, + "learning_rate": 8.965485157125975e-06, + "loss": 0.4753, + "step": 17344 + }, + { + "epoch": 0.547074593912632, + "grad_norm": 1.542964220046997, + "learning_rate": 8.964469087202649e-06, + "loss": 0.4867, + "step": 17345 + }, + { + "epoch": 0.5471061346790727, + "grad_norm": 1.562530517578125, + "learning_rate": 8.963453028085886e-06, + "loss": 0.4503, + "step": 17346 + }, + { + "epoch": 0.5471376754455133, + "grad_norm": 1.556077003479004, + "learning_rate": 8.962436979786301e-06, + "loss": 0.4335, + "step": 17347 + }, + { + "epoch": 0.5471692162119539, + "grad_norm": 1.4930402040481567, + "learning_rate": 8.961420942314491e-06, + "loss": 0.4666, + "step": 17348 + }, + { + "epoch": 0.5472007569783945, + "grad_norm": 1.9225786924362183, + "learning_rate": 8.960404915681058e-06, + "loss": 0.4104, + "step": 17349 + }, + { + "epoch": 0.5472322977448352, + "grad_norm": 1.5078961849212646, + "learning_rate": 8.959388899896609e-06, + "loss": 0.4836, + "step": 17350 + }, + { + "epoch": 0.5472638385112758, + "grad_norm": 1.3013432025909424, + "learning_rate": 8.958372894971743e-06, + "loss": 0.4275, + "step": 17351 + }, + { + "epoch": 0.5472953792777164, + "grad_norm": 1.3873322010040283, + "learning_rate": 8.957356900917065e-06, + "loss": 0.4335, + "step": 17352 + }, + { + "epoch": 0.5473269200441571, + "grad_norm": 1.2909488677978516, + "learning_rate": 8.956340917743178e-06, + "loss": 0.3501, + "step": 17353 + }, + { + "epoch": 0.5473584608105977, + "grad_norm": 1.679081916809082, + "learning_rate": 8.955324945460686e-06, + "loss": 0.4251, + "step": 17354 + }, + { + "epoch": 0.5473900015770383, + "grad_norm": 2.3966283798217773, + "learning_rate": 8.954308984080185e-06, + "loss": 0.493, + "step": 17355 + }, + { + "epoch": 0.547421542343479, + "grad_norm": 1.8517496585845947, + "learning_rate": 8.953293033612285e-06, + "loss": 0.4411, + "step": 17356 + }, + { + "epoch": 0.5474530831099196, + "grad_norm": 1.594487190246582, + "learning_rate": 8.952277094067579e-06, + "loss": 0.4317, + "step": 17357 + }, + { + "epoch": 0.5474846238763602, + "grad_norm": 1.6883363723754883, + "learning_rate": 8.951261165456682e-06, + "loss": 0.428, + "step": 17358 + }, + { + "epoch": 0.5475161646428008, + "grad_norm": 2.295283794403076, + "learning_rate": 8.950245247790188e-06, + "loss": 0.4768, + "step": 17359 + }, + { + "epoch": 0.5475477054092415, + "grad_norm": 1.1847620010375977, + "learning_rate": 8.949229341078697e-06, + "loss": 0.4446, + "step": 17360 + }, + { + "epoch": 0.5475792461756821, + "grad_norm": 2.5453670024871826, + "learning_rate": 8.948213445332818e-06, + "loss": 0.4412, + "step": 17361 + }, + { + "epoch": 0.5476107869421227, + "grad_norm": 1.3391211032867432, + "learning_rate": 8.947197560563147e-06, + "loss": 0.4817, + "step": 17362 + }, + { + "epoch": 0.5476423277085634, + "grad_norm": 1.8723089694976807, + "learning_rate": 8.946181686780282e-06, + "loss": 0.4585, + "step": 17363 + }, + { + "epoch": 0.547673868475004, + "grad_norm": 1.6009780168533325, + "learning_rate": 8.945165823994835e-06, + "loss": 0.4776, + "step": 17364 + }, + { + "epoch": 0.5477054092414446, + "grad_norm": 1.3664475679397583, + "learning_rate": 8.9441499722174e-06, + "loss": 0.4699, + "step": 17365 + }, + { + "epoch": 0.5477369500078851, + "grad_norm": 1.7364263534545898, + "learning_rate": 8.943134131458577e-06, + "loss": 0.4963, + "step": 17366 + }, + { + "epoch": 0.5477684907743258, + "grad_norm": 1.8584378957748413, + "learning_rate": 8.942118301728975e-06, + "loss": 0.451, + "step": 17367 + }, + { + "epoch": 0.5478000315407664, + "grad_norm": 2.0274100303649902, + "learning_rate": 8.941102483039188e-06, + "loss": 0.4485, + "step": 17368 + }, + { + "epoch": 0.547831572307207, + "grad_norm": 1.234540581703186, + "learning_rate": 8.940086675399818e-06, + "loss": 0.4743, + "step": 17369 + }, + { + "epoch": 0.5478631130736477, + "grad_norm": 1.6049995422363281, + "learning_rate": 8.939070878821469e-06, + "loss": 0.4349, + "step": 17370 + }, + { + "epoch": 0.5478946538400883, + "grad_norm": 1.5390796661376953, + "learning_rate": 8.938055093314735e-06, + "loss": 0.462, + "step": 17371 + }, + { + "epoch": 0.5479261946065289, + "grad_norm": 1.7441316843032837, + "learning_rate": 8.937039318890223e-06, + "loss": 0.491, + "step": 17372 + }, + { + "epoch": 0.5479577353729695, + "grad_norm": 1.2829585075378418, + "learning_rate": 8.936023555558532e-06, + "loss": 0.4539, + "step": 17373 + }, + { + "epoch": 0.5479892761394102, + "grad_norm": 21.797143936157227, + "learning_rate": 8.935007803330258e-06, + "loss": 0.3692, + "step": 17374 + }, + { + "epoch": 0.5480208169058508, + "grad_norm": 1.5959107875823975, + "learning_rate": 8.933992062216008e-06, + "loss": 0.4611, + "step": 17375 + }, + { + "epoch": 0.5480523576722914, + "grad_norm": 1.6019904613494873, + "learning_rate": 8.932976332226378e-06, + "loss": 0.4504, + "step": 17376 + }, + { + "epoch": 0.5480838984387321, + "grad_norm": 1.641897201538086, + "learning_rate": 8.931960613371966e-06, + "loss": 0.445, + "step": 17377 + }, + { + "epoch": 0.5481154392051727, + "grad_norm": 1.8008804321289062, + "learning_rate": 8.930944905663377e-06, + "loss": 0.4363, + "step": 17378 + }, + { + "epoch": 0.5481469799716133, + "grad_norm": 1.292959451675415, + "learning_rate": 8.929929209111208e-06, + "loss": 0.45, + "step": 17379 + }, + { + "epoch": 0.548178520738054, + "grad_norm": 1.6287872791290283, + "learning_rate": 8.928913523726057e-06, + "loss": 0.4938, + "step": 17380 + }, + { + "epoch": 0.5482100615044946, + "grad_norm": 1.5258382558822632, + "learning_rate": 8.927897849518528e-06, + "loss": 0.473, + "step": 17381 + }, + { + "epoch": 0.5482416022709352, + "grad_norm": 1.604365348815918, + "learning_rate": 8.926882186499216e-06, + "loss": 0.3691, + "step": 17382 + }, + { + "epoch": 0.5482731430373758, + "grad_norm": 1.4529014825820923, + "learning_rate": 8.925866534678719e-06, + "loss": 0.4927, + "step": 17383 + }, + { + "epoch": 0.5483046838038165, + "grad_norm": 1.2523027658462524, + "learning_rate": 8.92485089406764e-06, + "loss": 0.4627, + "step": 17384 + }, + { + "epoch": 0.5483362245702571, + "grad_norm": 1.5063821077346802, + "learning_rate": 8.92383526467658e-06, + "loss": 0.4593, + "step": 17385 + }, + { + "epoch": 0.5483677653366977, + "grad_norm": 1.6520049571990967, + "learning_rate": 8.922819646516128e-06, + "loss": 0.4108, + "step": 17386 + }, + { + "epoch": 0.5483993061031383, + "grad_norm": 1.4328594207763672, + "learning_rate": 8.921804039596895e-06, + "loss": 0.7434, + "step": 17387 + }, + { + "epoch": 0.5484308468695789, + "grad_norm": 1.868567705154419, + "learning_rate": 8.920788443929473e-06, + "loss": 0.4735, + "step": 17388 + }, + { + "epoch": 0.5484623876360195, + "grad_norm": 1.6056824922561646, + "learning_rate": 8.91977285952446e-06, + "loss": 0.3577, + "step": 17389 + }, + { + "epoch": 0.5484939284024601, + "grad_norm": 1.4360506534576416, + "learning_rate": 8.918757286392459e-06, + "loss": 0.7719, + "step": 17390 + }, + { + "epoch": 0.5485254691689008, + "grad_norm": 1.535343050956726, + "learning_rate": 8.91774172454406e-06, + "loss": 0.4883, + "step": 17391 + }, + { + "epoch": 0.5485570099353414, + "grad_norm": 1.8003172874450684, + "learning_rate": 8.916726173989871e-06, + "loss": 0.4807, + "step": 17392 + }, + { + "epoch": 0.548588550701782, + "grad_norm": 1.716907024383545, + "learning_rate": 8.915710634740486e-06, + "loss": 0.4598, + "step": 17393 + }, + { + "epoch": 0.5486200914682227, + "grad_norm": 1.8002634048461914, + "learning_rate": 8.914695106806498e-06, + "loss": 0.4273, + "step": 17394 + }, + { + "epoch": 0.5486516322346633, + "grad_norm": 1.5958278179168701, + "learning_rate": 8.913679590198512e-06, + "loss": 0.5168, + "step": 17395 + }, + { + "epoch": 0.5486831730011039, + "grad_norm": 1.486161231994629, + "learning_rate": 8.912664084927125e-06, + "loss": 0.757, + "step": 17396 + }, + { + "epoch": 0.5487147137675445, + "grad_norm": 1.7106271982192993, + "learning_rate": 8.911648591002928e-06, + "loss": 0.4027, + "step": 17397 + }, + { + "epoch": 0.5487462545339852, + "grad_norm": 1.5251615047454834, + "learning_rate": 8.910633108436525e-06, + "loss": 0.4692, + "step": 17398 + }, + { + "epoch": 0.5487777953004258, + "grad_norm": 1.369638442993164, + "learning_rate": 8.909617637238514e-06, + "loss": 0.4744, + "step": 17399 + }, + { + "epoch": 0.5488093360668664, + "grad_norm": 1.639727234840393, + "learning_rate": 8.908602177419487e-06, + "loss": 0.4479, + "step": 17400 + }, + { + "epoch": 0.5488408768333071, + "grad_norm": 1.3839808702468872, + "learning_rate": 8.907586728990047e-06, + "loss": 0.4615, + "step": 17401 + }, + { + "epoch": 0.5488724175997477, + "grad_norm": 1.7897697687149048, + "learning_rate": 8.906571291960786e-06, + "loss": 0.4681, + "step": 17402 + }, + { + "epoch": 0.5489039583661883, + "grad_norm": 2.2988429069519043, + "learning_rate": 8.905555866342298e-06, + "loss": 0.4575, + "step": 17403 + }, + { + "epoch": 0.548935499132629, + "grad_norm": 1.616945743560791, + "learning_rate": 8.90454045214519e-06, + "loss": 0.4576, + "step": 17404 + }, + { + "epoch": 0.5489670398990696, + "grad_norm": 1.527761697769165, + "learning_rate": 8.903525049380048e-06, + "loss": 0.7241, + "step": 17405 + }, + { + "epoch": 0.5489985806655102, + "grad_norm": 1.5107051134109497, + "learning_rate": 8.902509658057477e-06, + "loss": 0.4741, + "step": 17406 + }, + { + "epoch": 0.5490301214319508, + "grad_norm": 1.6708630323410034, + "learning_rate": 8.901494278188071e-06, + "loss": 0.4495, + "step": 17407 + }, + { + "epoch": 0.5490616621983915, + "grad_norm": 1.5307122468948364, + "learning_rate": 8.900478909782422e-06, + "loss": 0.4598, + "step": 17408 + }, + { + "epoch": 0.549093202964832, + "grad_norm": 1.2868767976760864, + "learning_rate": 8.899463552851133e-06, + "loss": 0.4468, + "step": 17409 + }, + { + "epoch": 0.5491247437312726, + "grad_norm": 1.7932896614074707, + "learning_rate": 8.898448207404795e-06, + "loss": 0.4793, + "step": 17410 + }, + { + "epoch": 0.5491562844977133, + "grad_norm": 1.631287693977356, + "learning_rate": 8.897432873454002e-06, + "loss": 0.4881, + "step": 17411 + }, + { + "epoch": 0.5491878252641539, + "grad_norm": 2.3413050174713135, + "learning_rate": 8.896417551009357e-06, + "loss": 0.4748, + "step": 17412 + }, + { + "epoch": 0.5492193660305945, + "grad_norm": 2.1277260780334473, + "learning_rate": 8.895402240081452e-06, + "loss": 0.466, + "step": 17413 + }, + { + "epoch": 0.5492509067970351, + "grad_norm": 1.4527286291122437, + "learning_rate": 8.894386940680877e-06, + "loss": 0.4722, + "step": 17414 + }, + { + "epoch": 0.5492824475634758, + "grad_norm": 1.3473705053329468, + "learning_rate": 8.893371652818238e-06, + "loss": 0.3499, + "step": 17415 + }, + { + "epoch": 0.5493139883299164, + "grad_norm": 1.7082533836364746, + "learning_rate": 8.892356376504125e-06, + "loss": 0.4517, + "step": 17416 + }, + { + "epoch": 0.549345529096357, + "grad_norm": 1.6088511943817139, + "learning_rate": 8.891341111749127e-06, + "loss": 0.4521, + "step": 17417 + }, + { + "epoch": 0.5493770698627977, + "grad_norm": 1.728583812713623, + "learning_rate": 8.890325858563853e-06, + "loss": 0.4645, + "step": 17418 + }, + { + "epoch": 0.5494086106292383, + "grad_norm": 1.4841740131378174, + "learning_rate": 8.889310616958889e-06, + "loss": 0.469, + "step": 17419 + }, + { + "epoch": 0.5494401513956789, + "grad_norm": 1.9164767265319824, + "learning_rate": 8.888295386944829e-06, + "loss": 0.7676, + "step": 17420 + }, + { + "epoch": 0.5494716921621196, + "grad_norm": 1.5260578393936157, + "learning_rate": 8.887280168532271e-06, + "loss": 0.4694, + "step": 17421 + }, + { + "epoch": 0.5495032329285602, + "grad_norm": 1.843970775604248, + "learning_rate": 8.886264961731804e-06, + "loss": 0.7881, + "step": 17422 + }, + { + "epoch": 0.5495347736950008, + "grad_norm": 1.49666428565979, + "learning_rate": 8.885249766554033e-06, + "loss": 0.5, + "step": 17423 + }, + { + "epoch": 0.5495663144614414, + "grad_norm": 1.4854782819747925, + "learning_rate": 8.884234583009545e-06, + "loss": 0.4814, + "step": 17424 + }, + { + "epoch": 0.5495978552278821, + "grad_norm": 1.6246156692504883, + "learning_rate": 8.883219411108932e-06, + "loss": 0.4707, + "step": 17425 + }, + { + "epoch": 0.5496293959943227, + "grad_norm": 1.3681899309158325, + "learning_rate": 8.882204250862796e-06, + "loss": 0.4691, + "step": 17426 + }, + { + "epoch": 0.5496609367607633, + "grad_norm": 2.639305353164673, + "learning_rate": 8.881189102281726e-06, + "loss": 0.4612, + "step": 17427 + }, + { + "epoch": 0.549692477527204, + "grad_norm": 1.266257643699646, + "learning_rate": 8.880173965376314e-06, + "loss": 0.4692, + "step": 17428 + }, + { + "epoch": 0.5497240182936446, + "grad_norm": 1.1533836126327515, + "learning_rate": 8.879158840157159e-06, + "loss": 0.3562, + "step": 17429 + }, + { + "epoch": 0.5497555590600851, + "grad_norm": 1.3894890546798706, + "learning_rate": 8.87814372663485e-06, + "loss": 0.4509, + "step": 17430 + }, + { + "epoch": 0.5497870998265257, + "grad_norm": 1.5806330442428589, + "learning_rate": 8.877128624819981e-06, + "loss": 0.4657, + "step": 17431 + }, + { + "epoch": 0.5498186405929664, + "grad_norm": 1.8366116285324097, + "learning_rate": 8.87611353472315e-06, + "loss": 0.3797, + "step": 17432 + }, + { + "epoch": 0.549850181359407, + "grad_norm": 1.7556806802749634, + "learning_rate": 8.875098456354945e-06, + "loss": 0.4808, + "step": 17433 + }, + { + "epoch": 0.5498817221258476, + "grad_norm": 1.8217451572418213, + "learning_rate": 8.874083389725958e-06, + "loss": 0.463, + "step": 17434 + }, + { + "epoch": 0.5499132628922883, + "grad_norm": 6.153283596038818, + "learning_rate": 8.87306833484679e-06, + "loss": 0.3737, + "step": 17435 + }, + { + "epoch": 0.5499448036587289, + "grad_norm": 2.132387399673462, + "learning_rate": 8.872053291728026e-06, + "loss": 0.4247, + "step": 17436 + }, + { + "epoch": 0.5499763444251695, + "grad_norm": 2.3722853660583496, + "learning_rate": 8.871038260380263e-06, + "loss": 0.4499, + "step": 17437 + }, + { + "epoch": 0.5500078851916101, + "grad_norm": 1.5219652652740479, + "learning_rate": 8.870023240814094e-06, + "loss": 0.4622, + "step": 17438 + }, + { + "epoch": 0.5500394259580508, + "grad_norm": 2.3003733158111572, + "learning_rate": 8.869008233040107e-06, + "loss": 0.4828, + "step": 17439 + }, + { + "epoch": 0.5500709667244914, + "grad_norm": 1.7595720291137695, + "learning_rate": 8.8679932370689e-06, + "loss": 0.4304, + "step": 17440 + }, + { + "epoch": 0.550102507490932, + "grad_norm": 1.5243189334869385, + "learning_rate": 8.866978252911063e-06, + "loss": 0.4584, + "step": 17441 + }, + { + "epoch": 0.5501340482573727, + "grad_norm": 2.1988654136657715, + "learning_rate": 8.865963280577182e-06, + "loss": 0.4701, + "step": 17442 + }, + { + "epoch": 0.5501655890238133, + "grad_norm": 1.6248509883880615, + "learning_rate": 8.864948320077859e-06, + "loss": 0.4797, + "step": 17443 + }, + { + "epoch": 0.5501971297902539, + "grad_norm": 1.3978382349014282, + "learning_rate": 8.863933371423683e-06, + "loss": 0.4853, + "step": 17444 + }, + { + "epoch": 0.5502286705566946, + "grad_norm": 1.5955811738967896, + "learning_rate": 8.86291843462524e-06, + "loss": 0.4618, + "step": 17445 + }, + { + "epoch": 0.5502602113231352, + "grad_norm": 1.5877026319503784, + "learning_rate": 8.86190350969313e-06, + "loss": 0.4674, + "step": 17446 + }, + { + "epoch": 0.5502917520895758, + "grad_norm": 1.235533356666565, + "learning_rate": 8.86088859663794e-06, + "loss": 0.4923, + "step": 17447 + }, + { + "epoch": 0.5503232928560164, + "grad_norm": 2.0756189823150635, + "learning_rate": 8.85987369547026e-06, + "loss": 0.46, + "step": 17448 + }, + { + "epoch": 0.5503548336224571, + "grad_norm": 1.4397540092468262, + "learning_rate": 8.858858806200687e-06, + "loss": 0.4793, + "step": 17449 + }, + { + "epoch": 0.5503863743888977, + "grad_norm": 1.4367611408233643, + "learning_rate": 8.857843928839807e-06, + "loss": 0.4367, + "step": 17450 + }, + { + "epoch": 0.5504179151553382, + "grad_norm": 2.045898675918579, + "learning_rate": 8.856829063398208e-06, + "loss": 0.4429, + "step": 17451 + }, + { + "epoch": 0.5504494559217789, + "grad_norm": 1.6799125671386719, + "learning_rate": 8.85581420988649e-06, + "loss": 0.3537, + "step": 17452 + }, + { + "epoch": 0.5504809966882195, + "grad_norm": 1.5643447637557983, + "learning_rate": 8.854799368315234e-06, + "loss": 0.4857, + "step": 17453 + }, + { + "epoch": 0.5505125374546601, + "grad_norm": 1.5382007360458374, + "learning_rate": 8.85378453869504e-06, + "loss": 0.4552, + "step": 17454 + }, + { + "epoch": 0.5505440782211007, + "grad_norm": 1.5968788862228394, + "learning_rate": 8.852769721036498e-06, + "loss": 0.4987, + "step": 17455 + }, + { + "epoch": 0.5505756189875414, + "grad_norm": 1.1780270338058472, + "learning_rate": 8.85175491535019e-06, + "loss": 0.4566, + "step": 17456 + }, + { + "epoch": 0.550607159753982, + "grad_norm": 1.315126657485962, + "learning_rate": 8.85074012164671e-06, + "loss": 0.4726, + "step": 17457 + }, + { + "epoch": 0.5506387005204226, + "grad_norm": 1.5308527946472168, + "learning_rate": 8.849725339936652e-06, + "loss": 0.478, + "step": 17458 + }, + { + "epoch": 0.5506702412868633, + "grad_norm": 1.8718414306640625, + "learning_rate": 8.848710570230603e-06, + "loss": 0.4849, + "step": 17459 + }, + { + "epoch": 0.5507017820533039, + "grad_norm": 1.9917781352996826, + "learning_rate": 8.847695812539153e-06, + "loss": 0.4646, + "step": 17460 + }, + { + "epoch": 0.5507333228197445, + "grad_norm": 1.662182092666626, + "learning_rate": 8.846681066872893e-06, + "loss": 0.4704, + "step": 17461 + }, + { + "epoch": 0.5507648635861851, + "grad_norm": 1.828006625175476, + "learning_rate": 8.845666333242407e-06, + "loss": 0.7869, + "step": 17462 + }, + { + "epoch": 0.5507964043526258, + "grad_norm": 4.001307010650635, + "learning_rate": 8.844651611658293e-06, + "loss": 0.3342, + "step": 17463 + }, + { + "epoch": 0.5508279451190664, + "grad_norm": 1.5248138904571533, + "learning_rate": 8.843636902131136e-06, + "loss": 0.4244, + "step": 17464 + }, + { + "epoch": 0.550859485885507, + "grad_norm": 1.540142297744751, + "learning_rate": 8.842622204671523e-06, + "loss": 0.4511, + "step": 17465 + }, + { + "epoch": 0.5508910266519477, + "grad_norm": 1.4661952257156372, + "learning_rate": 8.84160751929005e-06, + "loss": 0.4496, + "step": 17466 + }, + { + "epoch": 0.5509225674183883, + "grad_norm": 1.5557854175567627, + "learning_rate": 8.840592845997302e-06, + "loss": 0.4667, + "step": 17467 + }, + { + "epoch": 0.5509541081848289, + "grad_norm": 1.720293641090393, + "learning_rate": 8.839578184803865e-06, + "loss": 0.4852, + "step": 17468 + }, + { + "epoch": 0.5509856489512696, + "grad_norm": 1.7740414142608643, + "learning_rate": 8.838563535720334e-06, + "loss": 0.4624, + "step": 17469 + }, + { + "epoch": 0.5510171897177102, + "grad_norm": 1.4525209665298462, + "learning_rate": 8.83754889875729e-06, + "loss": 0.437, + "step": 17470 + }, + { + "epoch": 0.5510487304841508, + "grad_norm": 1.7046276330947876, + "learning_rate": 8.83653427392533e-06, + "loss": 0.4938, + "step": 17471 + }, + { + "epoch": 0.5510802712505913, + "grad_norm": 1.3776534795761108, + "learning_rate": 8.835519661235038e-06, + "loss": 0.4252, + "step": 17472 + }, + { + "epoch": 0.551111812017032, + "grad_norm": 3.1714251041412354, + "learning_rate": 8.834505060696995e-06, + "loss": 0.4621, + "step": 17473 + }, + { + "epoch": 0.5511433527834726, + "grad_norm": 3.4404473304748535, + "learning_rate": 8.833490472321806e-06, + "loss": 0.3264, + "step": 17474 + }, + { + "epoch": 0.5511748935499132, + "grad_norm": 2.2317445278167725, + "learning_rate": 8.832475896120047e-06, + "loss": 0.4862, + "step": 17475 + }, + { + "epoch": 0.5512064343163539, + "grad_norm": 1.546531081199646, + "learning_rate": 8.831461332102308e-06, + "loss": 0.4696, + "step": 17476 + }, + { + "epoch": 0.5512379750827945, + "grad_norm": 1.905674695968628, + "learning_rate": 8.830446780279175e-06, + "loss": 0.4779, + "step": 17477 + }, + { + "epoch": 0.5512695158492351, + "grad_norm": 1.5837535858154297, + "learning_rate": 8.829432240661242e-06, + "loss": 0.4522, + "step": 17478 + }, + { + "epoch": 0.5513010566156757, + "grad_norm": 1.615028738975525, + "learning_rate": 8.82841771325909e-06, + "loss": 0.459, + "step": 17479 + }, + { + "epoch": 0.5513325973821164, + "grad_norm": 1.8795275688171387, + "learning_rate": 8.827403198083311e-06, + "loss": 0.4696, + "step": 17480 + }, + { + "epoch": 0.551364138148557, + "grad_norm": 1.5576382875442505, + "learning_rate": 8.826388695144491e-06, + "loss": 0.4719, + "step": 17481 + }, + { + "epoch": 0.5513956789149976, + "grad_norm": 2.4715890884399414, + "learning_rate": 8.82537420445321e-06, + "loss": 0.458, + "step": 17482 + }, + { + "epoch": 0.5514272196814383, + "grad_norm": 1.5516331195831299, + "learning_rate": 8.824359726020068e-06, + "loss": 0.7349, + "step": 17483 + }, + { + "epoch": 0.5514587604478789, + "grad_norm": 1.7405925989151, + "learning_rate": 8.82334525985564e-06, + "loss": 0.4729, + "step": 17484 + }, + { + "epoch": 0.5514903012143195, + "grad_norm": 1.5927308797836304, + "learning_rate": 8.82233080597052e-06, + "loss": 0.36, + "step": 17485 + }, + { + "epoch": 0.5515218419807602, + "grad_norm": 2.27563738822937, + "learning_rate": 8.821316364375293e-06, + "loss": 0.4685, + "step": 17486 + }, + { + "epoch": 0.5515533827472008, + "grad_norm": 9.559431076049805, + "learning_rate": 8.820301935080545e-06, + "loss": 0.3711, + "step": 17487 + }, + { + "epoch": 0.5515849235136414, + "grad_norm": 2.2387707233428955, + "learning_rate": 8.819287518096864e-06, + "loss": 0.4468, + "step": 17488 + }, + { + "epoch": 0.551616464280082, + "grad_norm": 1.9900743961334229, + "learning_rate": 8.818273113434833e-06, + "loss": 0.3949, + "step": 17489 + }, + { + "epoch": 0.5516480050465227, + "grad_norm": 1.9125432968139648, + "learning_rate": 8.817258721105036e-06, + "loss": 0.4698, + "step": 17490 + }, + { + "epoch": 0.5516795458129633, + "grad_norm": 1.670509934425354, + "learning_rate": 8.816244341118068e-06, + "loss": 0.4641, + "step": 17491 + }, + { + "epoch": 0.5517110865794039, + "grad_norm": 1.55438232421875, + "learning_rate": 8.81522997348451e-06, + "loss": 0.4298, + "step": 17492 + }, + { + "epoch": 0.5517426273458444, + "grad_norm": 1.6955435276031494, + "learning_rate": 8.81421561821494e-06, + "loss": 0.4534, + "step": 17493 + }, + { + "epoch": 0.5517741681122851, + "grad_norm": 2.043761968612671, + "learning_rate": 8.813201275319957e-06, + "loss": 0.5002, + "step": 17494 + }, + { + "epoch": 0.5518057088787257, + "grad_norm": 1.6927297115325928, + "learning_rate": 8.81218694481014e-06, + "loss": 0.4429, + "step": 17495 + }, + { + "epoch": 0.5518372496451663, + "grad_norm": 1.3773921728134155, + "learning_rate": 8.811172626696072e-06, + "loss": 0.4914, + "step": 17496 + }, + { + "epoch": 0.551868790411607, + "grad_norm": 1.7483633756637573, + "learning_rate": 8.81015832098834e-06, + "loss": 0.4618, + "step": 17497 + }, + { + "epoch": 0.5519003311780476, + "grad_norm": 1.7549092769622803, + "learning_rate": 8.809144027697534e-06, + "loss": 0.4646, + "step": 17498 + }, + { + "epoch": 0.5519318719444882, + "grad_norm": 1.8922218084335327, + "learning_rate": 8.80812974683423e-06, + "loss": 0.442, + "step": 17499 + }, + { + "epoch": 0.5519634127109289, + "grad_norm": 1.7220001220703125, + "learning_rate": 8.807115478409022e-06, + "loss": 0.4644, + "step": 17500 + }, + { + "epoch": 0.5519949534773695, + "grad_norm": 2.1562702655792236, + "learning_rate": 8.806101222432485e-06, + "loss": 0.4049, + "step": 17501 + }, + { + "epoch": 0.5520264942438101, + "grad_norm": 1.4483994245529175, + "learning_rate": 8.805086978915215e-06, + "loss": 0.445, + "step": 17502 + }, + { + "epoch": 0.5520580350102507, + "grad_norm": 1.5565022230148315, + "learning_rate": 8.804072747867787e-06, + "loss": 0.4533, + "step": 17503 + }, + { + "epoch": 0.5520895757766914, + "grad_norm": 1.5431348085403442, + "learning_rate": 8.803058529300786e-06, + "loss": 0.4568, + "step": 17504 + }, + { + "epoch": 0.552121116543132, + "grad_norm": 1.422616958618164, + "learning_rate": 8.802044323224801e-06, + "loss": 0.5125, + "step": 17505 + }, + { + "epoch": 0.5521526573095726, + "grad_norm": 2.0835537910461426, + "learning_rate": 8.801030129650417e-06, + "loss": 0.4704, + "step": 17506 + }, + { + "epoch": 0.5521841980760133, + "grad_norm": 1.355242133140564, + "learning_rate": 8.80001594858821e-06, + "loss": 0.4769, + "step": 17507 + }, + { + "epoch": 0.5522157388424539, + "grad_norm": 1.2995917797088623, + "learning_rate": 8.799001780048771e-06, + "loss": 0.465, + "step": 17508 + }, + { + "epoch": 0.5522472796088945, + "grad_norm": 1.675431728363037, + "learning_rate": 8.797987624042681e-06, + "loss": 0.4391, + "step": 17509 + }, + { + "epoch": 0.5522788203753352, + "grad_norm": 1.4769344329833984, + "learning_rate": 8.79697348058052e-06, + "loss": 0.7755, + "step": 17510 + }, + { + "epoch": 0.5523103611417758, + "grad_norm": 1.6802748441696167, + "learning_rate": 8.795959349672878e-06, + "loss": 0.4528, + "step": 17511 + }, + { + "epoch": 0.5523419019082164, + "grad_norm": 1.4526230096817017, + "learning_rate": 8.794945231330335e-06, + "loss": 0.4615, + "step": 17512 + }, + { + "epoch": 0.552373442674657, + "grad_norm": 1.5189052820205688, + "learning_rate": 8.79393112556347e-06, + "loss": 0.7729, + "step": 17513 + }, + { + "epoch": 0.5524049834410976, + "grad_norm": 1.5035784244537354, + "learning_rate": 8.792917032382876e-06, + "loss": 0.467, + "step": 17514 + }, + { + "epoch": 0.5524365242075382, + "grad_norm": 1.7354466915130615, + "learning_rate": 8.791902951799129e-06, + "loss": 0.484, + "step": 17515 + }, + { + "epoch": 0.5524680649739788, + "grad_norm": 1.4361722469329834, + "learning_rate": 8.79088888382281e-06, + "loss": 0.4545, + "step": 17516 + }, + { + "epoch": 0.5524996057404195, + "grad_norm": 2.3839423656463623, + "learning_rate": 8.789874828464505e-06, + "loss": 0.4248, + "step": 17517 + }, + { + "epoch": 0.5525311465068601, + "grad_norm": 1.8385710716247559, + "learning_rate": 8.788860785734796e-06, + "loss": 0.4755, + "step": 17518 + }, + { + "epoch": 0.5525626872733007, + "grad_norm": 10.959243774414062, + "learning_rate": 8.787846755644267e-06, + "loss": 0.417, + "step": 17519 + }, + { + "epoch": 0.5525942280397413, + "grad_norm": 1.5765389204025269, + "learning_rate": 8.786832738203499e-06, + "loss": 0.4568, + "step": 17520 + }, + { + "epoch": 0.552625768806182, + "grad_norm": 1.7007869482040405, + "learning_rate": 8.785818733423068e-06, + "loss": 0.4879, + "step": 17521 + }, + { + "epoch": 0.5526573095726226, + "grad_norm": 1.7958247661590576, + "learning_rate": 8.784804741313566e-06, + "loss": 0.4758, + "step": 17522 + }, + { + "epoch": 0.5526888503390632, + "grad_norm": 1.3675198554992676, + "learning_rate": 8.78379076188557e-06, + "loss": 0.4187, + "step": 17523 + }, + { + "epoch": 0.5527203911055039, + "grad_norm": 1.5721664428710938, + "learning_rate": 8.782776795149658e-06, + "loss": 0.4421, + "step": 17524 + }, + { + "epoch": 0.5527519318719445, + "grad_norm": 1.4509652853012085, + "learning_rate": 8.78176284111642e-06, + "loss": 0.7372, + "step": 17525 + }, + { + "epoch": 0.5527834726383851, + "grad_norm": 2.4160029888153076, + "learning_rate": 8.78074889979643e-06, + "loss": 0.7791, + "step": 17526 + }, + { + "epoch": 0.5528150134048258, + "grad_norm": 1.645326852798462, + "learning_rate": 8.779734971200272e-06, + "loss": 0.4796, + "step": 17527 + }, + { + "epoch": 0.5528465541712664, + "grad_norm": 1.6494982242584229, + "learning_rate": 8.77872105533853e-06, + "loss": 0.3636, + "step": 17528 + }, + { + "epoch": 0.552878094937707, + "grad_norm": 2.116849422454834, + "learning_rate": 8.77770715222178e-06, + "loss": 0.4396, + "step": 17529 + }, + { + "epoch": 0.5529096357041476, + "grad_norm": 1.7716765403747559, + "learning_rate": 8.776693261860601e-06, + "loss": 0.4934, + "step": 17530 + }, + { + "epoch": 0.5529411764705883, + "grad_norm": 1.4572288990020752, + "learning_rate": 8.775679384265582e-06, + "loss": 0.4853, + "step": 17531 + }, + { + "epoch": 0.5529727172370289, + "grad_norm": 2.5819900035858154, + "learning_rate": 8.774665519447297e-06, + "loss": 0.7718, + "step": 17532 + }, + { + "epoch": 0.5530042580034695, + "grad_norm": 1.7012144327163696, + "learning_rate": 8.773651667416332e-06, + "loss": 0.445, + "step": 17533 + }, + { + "epoch": 0.5530357987699102, + "grad_norm": 2.2586867809295654, + "learning_rate": 8.772637828183263e-06, + "loss": 0.4624, + "step": 17534 + }, + { + "epoch": 0.5530673395363508, + "grad_norm": 1.8671876192092896, + "learning_rate": 8.77162400175867e-06, + "loss": 0.4709, + "step": 17535 + }, + { + "epoch": 0.5530988803027913, + "grad_norm": 1.4893773794174194, + "learning_rate": 8.770610188153137e-06, + "loss": 0.4889, + "step": 17536 + }, + { + "epoch": 0.5531304210692319, + "grad_norm": 1.90925133228302, + "learning_rate": 8.769596387377238e-06, + "loss": 0.48, + "step": 17537 + }, + { + "epoch": 0.5531619618356726, + "grad_norm": 1.3945508003234863, + "learning_rate": 8.768582599441558e-06, + "loss": 0.4356, + "step": 17538 + }, + { + "epoch": 0.5531935026021132, + "grad_norm": 1.5916821956634521, + "learning_rate": 8.767568824356676e-06, + "loss": 0.46, + "step": 17539 + }, + { + "epoch": 0.5532250433685538, + "grad_norm": 1.446218729019165, + "learning_rate": 8.766555062133169e-06, + "loss": 0.4843, + "step": 17540 + }, + { + "epoch": 0.5532565841349945, + "grad_norm": 1.606135368347168, + "learning_rate": 8.765541312781613e-06, + "loss": 0.4506, + "step": 17541 + }, + { + "epoch": 0.5532881249014351, + "grad_norm": 1.7970092296600342, + "learning_rate": 8.764527576312597e-06, + "loss": 0.4829, + "step": 17542 + }, + { + "epoch": 0.5533196656678757, + "grad_norm": 1.497501015663147, + "learning_rate": 8.763513852736697e-06, + "loss": 0.4701, + "step": 17543 + }, + { + "epoch": 0.5533512064343163, + "grad_norm": 1.3063454627990723, + "learning_rate": 8.762500142064484e-06, + "loss": 0.4742, + "step": 17544 + }, + { + "epoch": 0.553382747200757, + "grad_norm": 1.9201624393463135, + "learning_rate": 8.761486444306547e-06, + "loss": 0.4886, + "step": 17545 + }, + { + "epoch": 0.5534142879671976, + "grad_norm": 1.5276602506637573, + "learning_rate": 8.760472759473461e-06, + "loss": 0.486, + "step": 17546 + }, + { + "epoch": 0.5534458287336382, + "grad_norm": 1.7508022785186768, + "learning_rate": 8.759459087575803e-06, + "loss": 0.4839, + "step": 17547 + }, + { + "epoch": 0.5534773695000789, + "grad_norm": 1.9735870361328125, + "learning_rate": 8.758445428624152e-06, + "loss": 0.4408, + "step": 17548 + }, + { + "epoch": 0.5535089102665195, + "grad_norm": 1.5759108066558838, + "learning_rate": 8.757431782629085e-06, + "loss": 0.4439, + "step": 17549 + }, + { + "epoch": 0.5535404510329601, + "grad_norm": 2.29278302192688, + "learning_rate": 8.756418149601186e-06, + "loss": 0.7695, + "step": 17550 + }, + { + "epoch": 0.5535719917994008, + "grad_norm": 1.8632762432098389, + "learning_rate": 8.75540452955103e-06, + "loss": 0.483, + "step": 17551 + }, + { + "epoch": 0.5536035325658414, + "grad_norm": 1.7303025722503662, + "learning_rate": 8.754390922489188e-06, + "loss": 0.4506, + "step": 17552 + }, + { + "epoch": 0.553635073332282, + "grad_norm": 1.8500036001205444, + "learning_rate": 8.753377328426251e-06, + "loss": 0.4818, + "step": 17553 + }, + { + "epoch": 0.5536666140987226, + "grad_norm": 1.8936655521392822, + "learning_rate": 8.752363747372787e-06, + "loss": 0.4645, + "step": 17554 + }, + { + "epoch": 0.5536981548651633, + "grad_norm": 1.5485546588897705, + "learning_rate": 8.751350179339374e-06, + "loss": 0.464, + "step": 17555 + }, + { + "epoch": 0.5537296956316039, + "grad_norm": 1.7086554765701294, + "learning_rate": 8.750336624336596e-06, + "loss": 0.486, + "step": 17556 + }, + { + "epoch": 0.5537612363980444, + "grad_norm": 2.342345714569092, + "learning_rate": 8.749323082375022e-06, + "loss": 0.4538, + "step": 17557 + }, + { + "epoch": 0.553792777164485, + "grad_norm": 1.639596700668335, + "learning_rate": 8.748309553465233e-06, + "loss": 0.4297, + "step": 17558 + }, + { + "epoch": 0.5538243179309257, + "grad_norm": 2.424647808074951, + "learning_rate": 8.747296037617808e-06, + "loss": 0.3503, + "step": 17559 + }, + { + "epoch": 0.5538558586973663, + "grad_norm": 1.6331554651260376, + "learning_rate": 8.74628253484332e-06, + "loss": 0.4952, + "step": 17560 + }, + { + "epoch": 0.5538873994638069, + "grad_norm": 1.8570657968521118, + "learning_rate": 8.745269045152344e-06, + "loss": 0.448, + "step": 17561 + }, + { + "epoch": 0.5539189402302476, + "grad_norm": 2.0363919734954834, + "learning_rate": 8.744255568555464e-06, + "loss": 0.4267, + "step": 17562 + }, + { + "epoch": 0.5539504809966882, + "grad_norm": 1.5949310064315796, + "learning_rate": 8.743242105063254e-06, + "loss": 0.4577, + "step": 17563 + }, + { + "epoch": 0.5539820217631288, + "grad_norm": 1.6800209283828735, + "learning_rate": 8.74222865468628e-06, + "loss": 0.3888, + "step": 17564 + }, + { + "epoch": 0.5540135625295695, + "grad_norm": 3.8782715797424316, + "learning_rate": 8.741215217435134e-06, + "loss": 0.4903, + "step": 17565 + }, + { + "epoch": 0.5540451032960101, + "grad_norm": 1.4102269411087036, + "learning_rate": 8.740201793320383e-06, + "loss": 0.7413, + "step": 17566 + }, + { + "epoch": 0.5540766440624507, + "grad_norm": 1.947658658027649, + "learning_rate": 8.739188382352605e-06, + "loss": 0.4988, + "step": 17567 + }, + { + "epoch": 0.5541081848288913, + "grad_norm": 1.497484803199768, + "learning_rate": 8.738174984542376e-06, + "loss": 0.4082, + "step": 17568 + }, + { + "epoch": 0.554139725595332, + "grad_norm": 1.9436696767807007, + "learning_rate": 8.737161599900267e-06, + "loss": 0.7821, + "step": 17569 + }, + { + "epoch": 0.5541712663617726, + "grad_norm": 2.206162929534912, + "learning_rate": 8.73614822843686e-06, + "loss": 0.4412, + "step": 17570 + }, + { + "epoch": 0.5542028071282132, + "grad_norm": 1.779495120048523, + "learning_rate": 8.735134870162729e-06, + "loss": 0.3623, + "step": 17571 + }, + { + "epoch": 0.5542343478946539, + "grad_norm": 1.469367265701294, + "learning_rate": 8.734121525088442e-06, + "loss": 0.4681, + "step": 17572 + }, + { + "epoch": 0.5542658886610945, + "grad_norm": 2.014380693435669, + "learning_rate": 8.733108193224586e-06, + "loss": 0.4475, + "step": 17573 + }, + { + "epoch": 0.5542974294275351, + "grad_norm": 1.671146035194397, + "learning_rate": 8.732094874581729e-06, + "loss": 0.4463, + "step": 17574 + }, + { + "epoch": 0.5543289701939758, + "grad_norm": 1.8949040174484253, + "learning_rate": 8.731081569170445e-06, + "loss": 0.4718, + "step": 17575 + }, + { + "epoch": 0.5543605109604164, + "grad_norm": 1.7913250923156738, + "learning_rate": 8.730068277001311e-06, + "loss": 0.4341, + "step": 17576 + }, + { + "epoch": 0.554392051726857, + "grad_norm": 1.5610171556472778, + "learning_rate": 8.729054998084899e-06, + "loss": 0.4661, + "step": 17577 + }, + { + "epoch": 0.5544235924932975, + "grad_norm": 1.6677910089492798, + "learning_rate": 8.728041732431784e-06, + "loss": 0.333, + "step": 17578 + }, + { + "epoch": 0.5544551332597382, + "grad_norm": 1.7413218021392822, + "learning_rate": 8.727028480052544e-06, + "loss": 0.4805, + "step": 17579 + }, + { + "epoch": 0.5544866740261788, + "grad_norm": 1.7552680969238281, + "learning_rate": 8.726015240957746e-06, + "loss": 0.7711, + "step": 17580 + }, + { + "epoch": 0.5545182147926194, + "grad_norm": 1.7884430885314941, + "learning_rate": 8.725002015157972e-06, + "loss": 0.4742, + "step": 17581 + }, + { + "epoch": 0.55454975555906, + "grad_norm": 1.6908178329467773, + "learning_rate": 8.723988802663793e-06, + "loss": 0.4297, + "step": 17582 + }, + { + "epoch": 0.5545812963255007, + "grad_norm": 1.5359373092651367, + "learning_rate": 8.722975603485777e-06, + "loss": 0.4506, + "step": 17583 + }, + { + "epoch": 0.5546128370919413, + "grad_norm": 1.5585936307907104, + "learning_rate": 8.721962417634506e-06, + "loss": 0.4504, + "step": 17584 + }, + { + "epoch": 0.5546443778583819, + "grad_norm": 1.5481103658676147, + "learning_rate": 8.720949245120551e-06, + "loss": 0.3583, + "step": 17585 + }, + { + "epoch": 0.5546759186248226, + "grad_norm": 1.7178232669830322, + "learning_rate": 8.71993608595448e-06, + "loss": 0.4592, + "step": 17586 + }, + { + "epoch": 0.5547074593912632, + "grad_norm": 1.3992620706558228, + "learning_rate": 8.718922940146872e-06, + "loss": 0.4562, + "step": 17587 + }, + { + "epoch": 0.5547390001577038, + "grad_norm": 1.846665859222412, + "learning_rate": 8.717909807708299e-06, + "loss": 0.4477, + "step": 17588 + }, + { + "epoch": 0.5547705409241445, + "grad_norm": 1.9213203191757202, + "learning_rate": 8.716896688649328e-06, + "loss": 0.4635, + "step": 17589 + }, + { + "epoch": 0.5548020816905851, + "grad_norm": 1.833615779876709, + "learning_rate": 8.71588358298054e-06, + "loss": 0.4679, + "step": 17590 + }, + { + "epoch": 0.5548336224570257, + "grad_norm": 2.094040870666504, + "learning_rate": 8.714870490712505e-06, + "loss": 0.7699, + "step": 17591 + }, + { + "epoch": 0.5548651632234664, + "grad_norm": 1.2998384237289429, + "learning_rate": 8.71385741185579e-06, + "loss": 0.4516, + "step": 17592 + }, + { + "epoch": 0.554896703989907, + "grad_norm": 1.572654128074646, + "learning_rate": 8.712844346420974e-06, + "loss": 0.4482, + "step": 17593 + }, + { + "epoch": 0.5549282447563476, + "grad_norm": 1.6400822401046753, + "learning_rate": 8.711831294418629e-06, + "loss": 0.4939, + "step": 17594 + }, + { + "epoch": 0.5549597855227882, + "grad_norm": 1.8684403896331787, + "learning_rate": 8.710818255859321e-06, + "loss": 0.4181, + "step": 17595 + }, + { + "epoch": 0.5549913262892289, + "grad_norm": 1.5521210432052612, + "learning_rate": 8.709805230753628e-06, + "loss": 0.4757, + "step": 17596 + }, + { + "epoch": 0.5550228670556695, + "grad_norm": 1.599070429801941, + "learning_rate": 8.708792219112114e-06, + "loss": 0.4575, + "step": 17597 + }, + { + "epoch": 0.5550544078221101, + "grad_norm": 1.7517274618148804, + "learning_rate": 8.707779220945362e-06, + "loss": 0.4658, + "step": 17598 + }, + { + "epoch": 0.5550859485885506, + "grad_norm": 1.4994789361953735, + "learning_rate": 8.706766236263936e-06, + "loss": 0.377, + "step": 17599 + }, + { + "epoch": 0.5551174893549913, + "grad_norm": 1.6702314615249634, + "learning_rate": 8.705753265078406e-06, + "loss": 0.7551, + "step": 17600 + }, + { + "epoch": 0.5551490301214319, + "grad_norm": 1.6906019449234009, + "learning_rate": 8.704740307399347e-06, + "loss": 0.4705, + "step": 17601 + }, + { + "epoch": 0.5551805708878725, + "grad_norm": 3.6462974548339844, + "learning_rate": 8.703727363237331e-06, + "loss": 0.3351, + "step": 17602 + }, + { + "epoch": 0.5552121116543132, + "grad_norm": 1.8062961101531982, + "learning_rate": 8.702714432602922e-06, + "loss": 0.4553, + "step": 17603 + }, + { + "epoch": 0.5552436524207538, + "grad_norm": 2.414789915084839, + "learning_rate": 8.701701515506697e-06, + "loss": 0.4667, + "step": 17604 + }, + { + "epoch": 0.5552751931871944, + "grad_norm": 1.5516283512115479, + "learning_rate": 8.700688611959227e-06, + "loss": 0.4519, + "step": 17605 + }, + { + "epoch": 0.5553067339536351, + "grad_norm": 1.9706069231033325, + "learning_rate": 8.699675721971078e-06, + "loss": 0.4575, + "step": 17606 + }, + { + "epoch": 0.5553382747200757, + "grad_norm": 1.86420738697052, + "learning_rate": 8.698662845552825e-06, + "loss": 0.4568, + "step": 17607 + }, + { + "epoch": 0.5553698154865163, + "grad_norm": 2.0614233016967773, + "learning_rate": 8.697649982715035e-06, + "loss": 0.4786, + "step": 17608 + }, + { + "epoch": 0.555401356252957, + "grad_norm": 1.539823055267334, + "learning_rate": 8.696637133468274e-06, + "loss": 0.7826, + "step": 17609 + }, + { + "epoch": 0.5554328970193976, + "grad_norm": 1.5913466215133667, + "learning_rate": 8.695624297823121e-06, + "loss": 0.4476, + "step": 17610 + }, + { + "epoch": 0.5554644377858382, + "grad_norm": 2.5335655212402344, + "learning_rate": 8.694611475790138e-06, + "loss": 0.451, + "step": 17611 + }, + { + "epoch": 0.5554959785522788, + "grad_norm": 1.838338017463684, + "learning_rate": 8.6935986673799e-06, + "loss": 0.449, + "step": 17612 + }, + { + "epoch": 0.5555275193187195, + "grad_norm": 1.598522424697876, + "learning_rate": 8.692585872602977e-06, + "loss": 0.4307, + "step": 17613 + }, + { + "epoch": 0.5555590600851601, + "grad_norm": 1.8482658863067627, + "learning_rate": 8.691573091469932e-06, + "loss": 0.4597, + "step": 17614 + }, + { + "epoch": 0.5555906008516007, + "grad_norm": 1.6320489645004272, + "learning_rate": 8.690560323991339e-06, + "loss": 0.4635, + "step": 17615 + }, + { + "epoch": 0.5556221416180414, + "grad_norm": 1.4874517917633057, + "learning_rate": 8.689547570177768e-06, + "loss": 0.4952, + "step": 17616 + }, + { + "epoch": 0.555653682384482, + "grad_norm": 1.8414324522018433, + "learning_rate": 8.68853483003978e-06, + "loss": 0.4358, + "step": 17617 + }, + { + "epoch": 0.5556852231509226, + "grad_norm": 1.7765321731567383, + "learning_rate": 8.687522103587953e-06, + "loss": 0.4606, + "step": 17618 + }, + { + "epoch": 0.5557167639173632, + "grad_norm": 1.5446821451187134, + "learning_rate": 8.686509390832854e-06, + "loss": 0.7728, + "step": 17619 + }, + { + "epoch": 0.5557483046838038, + "grad_norm": 2.141336441040039, + "learning_rate": 8.685496691785043e-06, + "loss": 0.4603, + "step": 17620 + }, + { + "epoch": 0.5557798454502444, + "grad_norm": 8.299147605895996, + "learning_rate": 8.6844840064551e-06, + "loss": 0.4629, + "step": 17621 + }, + { + "epoch": 0.555811386216685, + "grad_norm": 2.306284189224243, + "learning_rate": 8.683471334853588e-06, + "loss": 0.4452, + "step": 17622 + }, + { + "epoch": 0.5558429269831257, + "grad_norm": 1.6811511516571045, + "learning_rate": 8.68245867699107e-06, + "loss": 0.4623, + "step": 17623 + }, + { + "epoch": 0.5558744677495663, + "grad_norm": 2.1711466312408447, + "learning_rate": 8.681446032878123e-06, + "loss": 0.4627, + "step": 17624 + }, + { + "epoch": 0.5559060085160069, + "grad_norm": 2.134643077850342, + "learning_rate": 8.680433402525313e-06, + "loss": 0.7722, + "step": 17625 + }, + { + "epoch": 0.5559375492824475, + "grad_norm": 1.7619805335998535, + "learning_rate": 8.6794207859432e-06, + "loss": 0.4484, + "step": 17626 + }, + { + "epoch": 0.5559690900488882, + "grad_norm": 1.6783561706542969, + "learning_rate": 8.678408183142359e-06, + "loss": 0.4278, + "step": 17627 + }, + { + "epoch": 0.5560006308153288, + "grad_norm": 1.6247663497924805, + "learning_rate": 8.677395594133352e-06, + "loss": 0.491, + "step": 17628 + }, + { + "epoch": 0.5560321715817694, + "grad_norm": 1.6024408340454102, + "learning_rate": 8.67638301892675e-06, + "loss": 0.4557, + "step": 17629 + }, + { + "epoch": 0.5560637123482101, + "grad_norm": 1.3448537588119507, + "learning_rate": 8.675370457533122e-06, + "loss": 0.465, + "step": 17630 + }, + { + "epoch": 0.5560952531146507, + "grad_norm": 1.2396080493927002, + "learning_rate": 8.674357909963026e-06, + "loss": 0.3741, + "step": 17631 + }, + { + "epoch": 0.5561267938810913, + "grad_norm": 1.5383163690567017, + "learning_rate": 8.673345376227041e-06, + "loss": 0.4433, + "step": 17632 + }, + { + "epoch": 0.556158334647532, + "grad_norm": 1.5501600503921509, + "learning_rate": 8.672332856335726e-06, + "loss": 0.4755, + "step": 17633 + }, + { + "epoch": 0.5561898754139726, + "grad_norm": 1.6548253297805786, + "learning_rate": 8.671320350299646e-06, + "loss": 0.4277, + "step": 17634 + }, + { + "epoch": 0.5562214161804132, + "grad_norm": 1.8147743940353394, + "learning_rate": 8.670307858129372e-06, + "loss": 0.4617, + "step": 17635 + }, + { + "epoch": 0.5562529569468538, + "grad_norm": 1.5470061302185059, + "learning_rate": 8.669295379835467e-06, + "loss": 0.4601, + "step": 17636 + }, + { + "epoch": 0.5562844977132945, + "grad_norm": 1.473747968673706, + "learning_rate": 8.668282915428495e-06, + "loss": 0.7957, + "step": 17637 + }, + { + "epoch": 0.5563160384797351, + "grad_norm": 1.680887222290039, + "learning_rate": 8.66727046491903e-06, + "loss": 0.4503, + "step": 17638 + }, + { + "epoch": 0.5563475792461757, + "grad_norm": 1.8505206108093262, + "learning_rate": 8.66625802831763e-06, + "loss": 0.4505, + "step": 17639 + }, + { + "epoch": 0.5563791200126164, + "grad_norm": 1.7231136560440063, + "learning_rate": 8.66524560563486e-06, + "loss": 0.4567, + "step": 17640 + }, + { + "epoch": 0.5564106607790569, + "grad_norm": 1.4753780364990234, + "learning_rate": 8.664233196881293e-06, + "loss": 0.4069, + "step": 17641 + }, + { + "epoch": 0.5564422015454975, + "grad_norm": 2.556135892868042, + "learning_rate": 8.66322080206749e-06, + "loss": 0.4639, + "step": 17642 + }, + { + "epoch": 0.5564737423119381, + "grad_norm": 1.4509410858154297, + "learning_rate": 8.662208421204011e-06, + "loss": 0.7802, + "step": 17643 + }, + { + "epoch": 0.5565052830783788, + "grad_norm": 1.4946991205215454, + "learning_rate": 8.66119605430143e-06, + "loss": 0.4772, + "step": 17644 + }, + { + "epoch": 0.5565368238448194, + "grad_norm": 1.7214070558547974, + "learning_rate": 8.660183701370306e-06, + "loss": 0.3849, + "step": 17645 + }, + { + "epoch": 0.55656836461126, + "grad_norm": 1.666496992111206, + "learning_rate": 8.659171362421208e-06, + "loss": 0.4912, + "step": 17646 + }, + { + "epoch": 0.5565999053777007, + "grad_norm": 2.13887095451355, + "learning_rate": 8.658159037464698e-06, + "loss": 0.4813, + "step": 17647 + }, + { + "epoch": 0.5566314461441413, + "grad_norm": 2.346477508544922, + "learning_rate": 8.657146726511336e-06, + "loss": 0.4473, + "step": 17648 + }, + { + "epoch": 0.5566629869105819, + "grad_norm": 1.4393260478973389, + "learning_rate": 8.656134429571695e-06, + "loss": 0.7464, + "step": 17649 + }, + { + "epoch": 0.5566945276770225, + "grad_norm": 1.6517308950424194, + "learning_rate": 8.655122146656335e-06, + "loss": 0.7523, + "step": 17650 + }, + { + "epoch": 0.5567260684434632, + "grad_norm": 1.5674192905426025, + "learning_rate": 8.654109877775814e-06, + "loss": 0.4667, + "step": 17651 + }, + { + "epoch": 0.5567576092099038, + "grad_norm": 1.7145167589187622, + "learning_rate": 8.653097622940707e-06, + "loss": 0.444, + "step": 17652 + }, + { + "epoch": 0.5567891499763444, + "grad_norm": 1.5638484954833984, + "learning_rate": 8.652085382161571e-06, + "loss": 0.4523, + "step": 17653 + }, + { + "epoch": 0.5568206907427851, + "grad_norm": 1.9565606117248535, + "learning_rate": 8.65107315544897e-06, + "loss": 0.4385, + "step": 17654 + }, + { + "epoch": 0.5568522315092257, + "grad_norm": 1.5225361585617065, + "learning_rate": 8.650060942813469e-06, + "loss": 0.4499, + "step": 17655 + }, + { + "epoch": 0.5568837722756663, + "grad_norm": 1.538051962852478, + "learning_rate": 8.64904874426563e-06, + "loss": 0.4263, + "step": 17656 + }, + { + "epoch": 0.556915313042107, + "grad_norm": 1.9742690324783325, + "learning_rate": 8.648036559816014e-06, + "loss": 0.4029, + "step": 17657 + }, + { + "epoch": 0.5569468538085476, + "grad_norm": 2.004485845565796, + "learning_rate": 8.647024389475188e-06, + "loss": 0.4647, + "step": 17658 + }, + { + "epoch": 0.5569783945749882, + "grad_norm": 1.313281536102295, + "learning_rate": 8.64601223325371e-06, + "loss": 0.461, + "step": 17659 + }, + { + "epoch": 0.5570099353414288, + "grad_norm": 1.3521356582641602, + "learning_rate": 8.64500009116215e-06, + "loss": 0.452, + "step": 17660 + }, + { + "epoch": 0.5570414761078695, + "grad_norm": 1.2951314449310303, + "learning_rate": 8.643987963211066e-06, + "loss": 0.4606, + "step": 17661 + }, + { + "epoch": 0.5570730168743101, + "grad_norm": 2.418715238571167, + "learning_rate": 8.642975849411018e-06, + "loss": 0.7695, + "step": 17662 + }, + { + "epoch": 0.5571045576407506, + "grad_norm": 1.5591286420822144, + "learning_rate": 8.641963749772572e-06, + "loss": 0.4238, + "step": 17663 + }, + { + "epoch": 0.5571360984071912, + "grad_norm": 1.6997119188308716, + "learning_rate": 8.64095166430629e-06, + "loss": 0.7307, + "step": 17664 + }, + { + "epoch": 0.5571676391736319, + "grad_norm": 1.8091646432876587, + "learning_rate": 8.63993959302273e-06, + "loss": 0.4463, + "step": 17665 + }, + { + "epoch": 0.5571991799400725, + "grad_norm": 3.1626832485198975, + "learning_rate": 8.63892753593246e-06, + "loss": 0.4406, + "step": 17666 + }, + { + "epoch": 0.5572307207065131, + "grad_norm": 1.5846331119537354, + "learning_rate": 8.637915493046037e-06, + "loss": 0.4883, + "step": 17667 + }, + { + "epoch": 0.5572622614729538, + "grad_norm": 1.6240477561950684, + "learning_rate": 8.63690346437402e-06, + "loss": 0.4355, + "step": 17668 + }, + { + "epoch": 0.5572938022393944, + "grad_norm": 1.867099642753601, + "learning_rate": 8.635891449926976e-06, + "loss": 0.4496, + "step": 17669 + }, + { + "epoch": 0.557325343005835, + "grad_norm": 2.065936803817749, + "learning_rate": 8.634879449715467e-06, + "loss": 0.4359, + "step": 17670 + }, + { + "epoch": 0.5573568837722757, + "grad_norm": 1.375203251838684, + "learning_rate": 8.633867463750044e-06, + "loss": 0.3813, + "step": 17671 + }, + { + "epoch": 0.5573884245387163, + "grad_norm": 1.597381591796875, + "learning_rate": 8.63285549204128e-06, + "loss": 0.4554, + "step": 17672 + }, + { + "epoch": 0.5574199653051569, + "grad_norm": 2.5635604858398438, + "learning_rate": 8.63184353459973e-06, + "loss": 0.4778, + "step": 17673 + }, + { + "epoch": 0.5574515060715975, + "grad_norm": 1.496946930885315, + "learning_rate": 8.630831591435955e-06, + "loss": 0.4468, + "step": 17674 + }, + { + "epoch": 0.5574830468380382, + "grad_norm": 2.0032241344451904, + "learning_rate": 8.629819662560515e-06, + "loss": 0.4475, + "step": 17675 + }, + { + "epoch": 0.5575145876044788, + "grad_norm": 2.082615375518799, + "learning_rate": 8.628807747983969e-06, + "loss": 0.7686, + "step": 17676 + }, + { + "epoch": 0.5575461283709194, + "grad_norm": 1.6307604312896729, + "learning_rate": 8.627795847716881e-06, + "loss": 0.4539, + "step": 17677 + }, + { + "epoch": 0.5575776691373601, + "grad_norm": 1.483197569847107, + "learning_rate": 8.62678396176981e-06, + "loss": 0.4537, + "step": 17678 + }, + { + "epoch": 0.5576092099038007, + "grad_norm": 2.6966965198516846, + "learning_rate": 8.62577209015331e-06, + "loss": 0.4875, + "step": 17679 + }, + { + "epoch": 0.5576407506702413, + "grad_norm": 1.5467326641082764, + "learning_rate": 8.62476023287795e-06, + "loss": 0.4714, + "step": 17680 + }, + { + "epoch": 0.557672291436682, + "grad_norm": 1.4042484760284424, + "learning_rate": 8.623748389954284e-06, + "loss": 0.779, + "step": 17681 + }, + { + "epoch": 0.5577038322031226, + "grad_norm": 2.83827805519104, + "learning_rate": 8.62273656139287e-06, + "loss": 0.4563, + "step": 17682 + }, + { + "epoch": 0.5577353729695632, + "grad_norm": 1.3758989572525024, + "learning_rate": 8.621724747204268e-06, + "loss": 0.741, + "step": 17683 + }, + { + "epoch": 0.5577669137360037, + "grad_norm": 1.5433437824249268, + "learning_rate": 8.620712947399043e-06, + "loss": 0.4245, + "step": 17684 + }, + { + "epoch": 0.5577984545024444, + "grad_norm": 1.6455799341201782, + "learning_rate": 8.619701161987747e-06, + "loss": 0.4605, + "step": 17685 + }, + { + "epoch": 0.557829995268885, + "grad_norm": 1.7345221042633057, + "learning_rate": 8.618689390980943e-06, + "loss": 0.4965, + "step": 17686 + }, + { + "epoch": 0.5578615360353256, + "grad_norm": 1.7896169424057007, + "learning_rate": 8.617677634389188e-06, + "loss": 0.472, + "step": 17687 + }, + { + "epoch": 0.5578930768017663, + "grad_norm": 1.4208499193191528, + "learning_rate": 8.616665892223035e-06, + "loss": 0.4697, + "step": 17688 + }, + { + "epoch": 0.5579246175682069, + "grad_norm": 1.9928442239761353, + "learning_rate": 8.615654164493052e-06, + "loss": 0.4387, + "step": 17689 + }, + { + "epoch": 0.5579561583346475, + "grad_norm": 1.6850430965423584, + "learning_rate": 8.614642451209794e-06, + "loss": 0.433, + "step": 17690 + }, + { + "epoch": 0.5579876991010881, + "grad_norm": 1.7144567966461182, + "learning_rate": 8.61363075238381e-06, + "loss": 0.4634, + "step": 17691 + }, + { + "epoch": 0.5580192398675288, + "grad_norm": 2.0906577110290527, + "learning_rate": 8.612619068025673e-06, + "loss": 0.75, + "step": 17692 + }, + { + "epoch": 0.5580507806339694, + "grad_norm": 1.6028201580047607, + "learning_rate": 8.611607398145931e-06, + "loss": 0.4604, + "step": 17693 + }, + { + "epoch": 0.55808232140041, + "grad_norm": 1.451087474822998, + "learning_rate": 8.610595742755146e-06, + "loss": 0.3879, + "step": 17694 + }, + { + "epoch": 0.5581138621668507, + "grad_norm": 1.4531458616256714, + "learning_rate": 8.609584101863873e-06, + "loss": 0.444, + "step": 17695 + }, + { + "epoch": 0.5581454029332913, + "grad_norm": 2.842397928237915, + "learning_rate": 8.608572475482664e-06, + "loss": 0.4834, + "step": 17696 + }, + { + "epoch": 0.5581769436997319, + "grad_norm": 1.3448125123977661, + "learning_rate": 8.607560863622089e-06, + "loss": 0.473, + "step": 17697 + }, + { + "epoch": 0.5582084844661725, + "grad_norm": 1.8125008344650269, + "learning_rate": 8.606549266292695e-06, + "loss": 0.4982, + "step": 17698 + }, + { + "epoch": 0.5582400252326132, + "grad_norm": 1.5025291442871094, + "learning_rate": 8.605537683505038e-06, + "loss": 0.4631, + "step": 17699 + }, + { + "epoch": 0.5582715659990538, + "grad_norm": 1.4130573272705078, + "learning_rate": 8.604526115269681e-06, + "loss": 0.4713, + "step": 17700 + }, + { + "epoch": 0.5583031067654944, + "grad_norm": 1.5492286682128906, + "learning_rate": 8.603514561597179e-06, + "loss": 0.439, + "step": 17701 + }, + { + "epoch": 0.5583346475319351, + "grad_norm": 1.9131584167480469, + "learning_rate": 8.602503022498084e-06, + "loss": 0.4797, + "step": 17702 + }, + { + "epoch": 0.5583661882983757, + "grad_norm": 1.6108464002609253, + "learning_rate": 8.601491497982956e-06, + "loss": 0.7495, + "step": 17703 + }, + { + "epoch": 0.5583977290648163, + "grad_norm": 1.667309045791626, + "learning_rate": 8.600479988062352e-06, + "loss": 0.4488, + "step": 17704 + }, + { + "epoch": 0.5584292698312568, + "grad_norm": 1.940325379371643, + "learning_rate": 8.599468492746823e-06, + "loss": 0.479, + "step": 17705 + }, + { + "epoch": 0.5584608105976975, + "grad_norm": 1.518896222114563, + "learning_rate": 8.598457012046931e-06, + "loss": 0.485, + "step": 17706 + }, + { + "epoch": 0.5584923513641381, + "grad_norm": 1.4841042757034302, + "learning_rate": 8.597445545973223e-06, + "loss": 0.4781, + "step": 17707 + }, + { + "epoch": 0.5585238921305787, + "grad_norm": 1.4816828966140747, + "learning_rate": 8.596434094536264e-06, + "loss": 0.7306, + "step": 17708 + }, + { + "epoch": 0.5585554328970194, + "grad_norm": 1.5845059156417847, + "learning_rate": 8.595422657746606e-06, + "loss": 0.4728, + "step": 17709 + }, + { + "epoch": 0.55858697366346, + "grad_norm": 1.7789509296417236, + "learning_rate": 8.5944112356148e-06, + "loss": 0.4234, + "step": 17710 + }, + { + "epoch": 0.5586185144299006, + "grad_norm": 3.631852149963379, + "learning_rate": 8.593399828151406e-06, + "loss": 0.4655, + "step": 17711 + }, + { + "epoch": 0.5586500551963413, + "grad_norm": 1.7383840084075928, + "learning_rate": 8.592388435366978e-06, + "loss": 0.7633, + "step": 17712 + }, + { + "epoch": 0.5586815959627819, + "grad_norm": 1.5662662982940674, + "learning_rate": 8.591377057272068e-06, + "loss": 0.4407, + "step": 17713 + }, + { + "epoch": 0.5587131367292225, + "grad_norm": 1.42952561378479, + "learning_rate": 8.590365693877235e-06, + "loss": 0.4657, + "step": 17714 + }, + { + "epoch": 0.5587446774956631, + "grad_norm": 1.4623363018035889, + "learning_rate": 8.589354345193028e-06, + "loss": 0.4521, + "step": 17715 + }, + { + "epoch": 0.5587762182621038, + "grad_norm": 1.8986002206802368, + "learning_rate": 8.58834301123e-06, + "loss": 0.4476, + "step": 17716 + }, + { + "epoch": 0.5588077590285444, + "grad_norm": 2.0989270210266113, + "learning_rate": 8.587331691998714e-06, + "loss": 0.3591, + "step": 17717 + }, + { + "epoch": 0.558839299794985, + "grad_norm": 1.7383211851119995, + "learning_rate": 8.586320387509719e-06, + "loss": 0.4504, + "step": 17718 + }, + { + "epoch": 0.5588708405614257, + "grad_norm": 1.6449778079986572, + "learning_rate": 8.585309097773564e-06, + "loss": 0.4875, + "step": 17719 + }, + { + "epoch": 0.5589023813278663, + "grad_norm": 1.4755384922027588, + "learning_rate": 8.58429782280081e-06, + "loss": 0.746, + "step": 17720 + }, + { + "epoch": 0.5589339220943069, + "grad_norm": 1.3751152753829956, + "learning_rate": 8.58328656260201e-06, + "loss": 0.4613, + "step": 17721 + }, + { + "epoch": 0.5589654628607476, + "grad_norm": 1.8770427703857422, + "learning_rate": 8.582275317187712e-06, + "loss": 0.4605, + "step": 17722 + }, + { + "epoch": 0.5589970036271882, + "grad_norm": 2.1329798698425293, + "learning_rate": 8.58126408656847e-06, + "loss": 0.4215, + "step": 17723 + }, + { + "epoch": 0.5590285443936288, + "grad_norm": 1.6105356216430664, + "learning_rate": 8.58025287075484e-06, + "loss": 0.4685, + "step": 17724 + }, + { + "epoch": 0.5590600851600694, + "grad_norm": 2.0746986865997314, + "learning_rate": 8.579241669757377e-06, + "loss": 0.4589, + "step": 17725 + }, + { + "epoch": 0.55909162592651, + "grad_norm": 1.2928824424743652, + "learning_rate": 8.57823048358663e-06, + "loss": 0.4536, + "step": 17726 + }, + { + "epoch": 0.5591231666929506, + "grad_norm": 1.5603827238082886, + "learning_rate": 8.577219312253146e-06, + "loss": 0.4388, + "step": 17727 + }, + { + "epoch": 0.5591547074593912, + "grad_norm": 1.515226125717163, + "learning_rate": 8.576208155767488e-06, + "loss": 0.7299, + "step": 17728 + }, + { + "epoch": 0.5591862482258319, + "grad_norm": 1.8092049360275269, + "learning_rate": 8.575197014140204e-06, + "loss": 0.41, + "step": 17729 + }, + { + "epoch": 0.5592177889922725, + "grad_norm": 5.112978458404541, + "learning_rate": 8.57418588738184e-06, + "loss": 0.4913, + "step": 17730 + }, + { + "epoch": 0.5592493297587131, + "grad_norm": 1.654907464981079, + "learning_rate": 8.57317477550296e-06, + "loss": 0.4376, + "step": 17731 + }, + { + "epoch": 0.5592808705251537, + "grad_norm": 2.1662673950195312, + "learning_rate": 8.572163678514107e-06, + "loss": 0.7539, + "step": 17732 + }, + { + "epoch": 0.5593124112915944, + "grad_norm": 1.4906364679336548, + "learning_rate": 8.571152596425833e-06, + "loss": 0.4484, + "step": 17733 + }, + { + "epoch": 0.559343952058035, + "grad_norm": 1.4655214548110962, + "learning_rate": 8.570141529248694e-06, + "loss": 0.4497, + "step": 17734 + }, + { + "epoch": 0.5593754928244756, + "grad_norm": 1.7668722867965698, + "learning_rate": 8.569130476993237e-06, + "loss": 0.4608, + "step": 17735 + }, + { + "epoch": 0.5594070335909163, + "grad_norm": 1.3432719707489014, + "learning_rate": 8.568119439670012e-06, + "loss": 0.4374, + "step": 17736 + }, + { + "epoch": 0.5594385743573569, + "grad_norm": 1.8681490421295166, + "learning_rate": 8.567108417289575e-06, + "loss": 0.4444, + "step": 17737 + }, + { + "epoch": 0.5594701151237975, + "grad_norm": 1.9683785438537598, + "learning_rate": 8.566097409862476e-06, + "loss": 0.4593, + "step": 17738 + }, + { + "epoch": 0.5595016558902381, + "grad_norm": 1.5206550359725952, + "learning_rate": 8.56508641739926e-06, + "loss": 0.3775, + "step": 17739 + }, + { + "epoch": 0.5595331966566788, + "grad_norm": 1.6125924587249756, + "learning_rate": 8.564075439910484e-06, + "loss": 0.7326, + "step": 17740 + }, + { + "epoch": 0.5595647374231194, + "grad_norm": 1.8444794416427612, + "learning_rate": 8.563064477406694e-06, + "loss": 0.4692, + "step": 17741 + }, + { + "epoch": 0.55959627818956, + "grad_norm": 1.7803244590759277, + "learning_rate": 8.562053529898444e-06, + "loss": 0.4671, + "step": 17742 + }, + { + "epoch": 0.5596278189560007, + "grad_norm": 1.5402805805206299, + "learning_rate": 8.561042597396281e-06, + "loss": 0.7583, + "step": 17743 + }, + { + "epoch": 0.5596593597224413, + "grad_norm": 2.1645596027374268, + "learning_rate": 8.560031679910754e-06, + "loss": 0.4724, + "step": 17744 + }, + { + "epoch": 0.5596909004888819, + "grad_norm": 2.466810703277588, + "learning_rate": 8.559020777452418e-06, + "loss": 0.4582, + "step": 17745 + }, + { + "epoch": 0.5597224412553226, + "grad_norm": 1.950946569442749, + "learning_rate": 8.558009890031819e-06, + "loss": 0.4499, + "step": 17746 + }, + { + "epoch": 0.5597539820217631, + "grad_norm": 1.961302638053894, + "learning_rate": 8.556999017659502e-06, + "loss": 0.4147, + "step": 17747 + }, + { + "epoch": 0.5597855227882037, + "grad_norm": 1.7632200717926025, + "learning_rate": 8.555988160346025e-06, + "loss": 0.4529, + "step": 17748 + }, + { + "epoch": 0.5598170635546443, + "grad_norm": 1.9631078243255615, + "learning_rate": 8.554977318101933e-06, + "loss": 0.5028, + "step": 17749 + }, + { + "epoch": 0.559848604321085, + "grad_norm": 1.5871001482009888, + "learning_rate": 8.553966490937769e-06, + "loss": 0.4752, + "step": 17750 + }, + { + "epoch": 0.5598801450875256, + "grad_norm": 1.7897037267684937, + "learning_rate": 8.552955678864093e-06, + "loss": 0.4558, + "step": 17751 + }, + { + "epoch": 0.5599116858539662, + "grad_norm": 2.112636089324951, + "learning_rate": 8.551944881891447e-06, + "loss": 0.488, + "step": 17752 + }, + { + "epoch": 0.5599432266204069, + "grad_norm": 1.6028459072113037, + "learning_rate": 8.55093410003038e-06, + "loss": 0.7516, + "step": 17753 + }, + { + "epoch": 0.5599747673868475, + "grad_norm": 1.7031654119491577, + "learning_rate": 8.549923333291443e-06, + "loss": 0.7786, + "step": 17754 + }, + { + "epoch": 0.5600063081532881, + "grad_norm": 1.3312747478485107, + "learning_rate": 8.548912581685175e-06, + "loss": 0.449, + "step": 17755 + }, + { + "epoch": 0.5600378489197287, + "grad_norm": 1.592016339302063, + "learning_rate": 8.54790184522214e-06, + "loss": 0.4715, + "step": 17756 + }, + { + "epoch": 0.5600693896861694, + "grad_norm": 1.8169746398925781, + "learning_rate": 8.546891123912873e-06, + "loss": 0.4418, + "step": 17757 + }, + { + "epoch": 0.56010093045261, + "grad_norm": 1.5313677787780762, + "learning_rate": 8.545880417767923e-06, + "loss": 0.4302, + "step": 17758 + }, + { + "epoch": 0.5601324712190506, + "grad_norm": 1.6788040399551392, + "learning_rate": 8.544869726797842e-06, + "loss": 0.4715, + "step": 17759 + }, + { + "epoch": 0.5601640119854913, + "grad_norm": 1.7476708889007568, + "learning_rate": 8.543859051013177e-06, + "loss": 0.4764, + "step": 17760 + }, + { + "epoch": 0.5601955527519319, + "grad_norm": 1.7965540885925293, + "learning_rate": 8.54284839042447e-06, + "loss": 0.496, + "step": 17761 + }, + { + "epoch": 0.5602270935183725, + "grad_norm": 1.595496416091919, + "learning_rate": 8.541837745042276e-06, + "loss": 0.4253, + "step": 17762 + }, + { + "epoch": 0.5602586342848132, + "grad_norm": 2.0811007022857666, + "learning_rate": 8.540827114877134e-06, + "loss": 0.4865, + "step": 17763 + }, + { + "epoch": 0.5602901750512538, + "grad_norm": 1.870437741279602, + "learning_rate": 8.539816499939592e-06, + "loss": 0.7507, + "step": 17764 + }, + { + "epoch": 0.5603217158176944, + "grad_norm": 3.041011333465576, + "learning_rate": 8.538805900240203e-06, + "loss": 0.424, + "step": 17765 + }, + { + "epoch": 0.560353256584135, + "grad_norm": 1.6866309642791748, + "learning_rate": 8.53779531578951e-06, + "loss": 0.4093, + "step": 17766 + }, + { + "epoch": 0.5603847973505757, + "grad_norm": 1.4963583946228027, + "learning_rate": 8.536784746598052e-06, + "loss": 0.4469, + "step": 17767 + }, + { + "epoch": 0.5604163381170162, + "grad_norm": 1.7028359174728394, + "learning_rate": 8.535774192676386e-06, + "loss": 0.7638, + "step": 17768 + }, + { + "epoch": 0.5604478788834568, + "grad_norm": 1.5014173984527588, + "learning_rate": 8.534763654035054e-06, + "loss": 0.4595, + "step": 17769 + }, + { + "epoch": 0.5604794196498974, + "grad_norm": 1.7755357027053833, + "learning_rate": 8.533753130684596e-06, + "loss": 0.4708, + "step": 17770 + }, + { + "epoch": 0.5605109604163381, + "grad_norm": 1.420989990234375, + "learning_rate": 8.532742622635568e-06, + "loss": 0.4575, + "step": 17771 + }, + { + "epoch": 0.5605425011827787, + "grad_norm": 1.7244096994400024, + "learning_rate": 8.531732129898507e-06, + "loss": 0.4685, + "step": 17772 + }, + { + "epoch": 0.5605740419492193, + "grad_norm": 1.460793137550354, + "learning_rate": 8.530721652483964e-06, + "loss": 0.7406, + "step": 17773 + }, + { + "epoch": 0.56060558271566, + "grad_norm": 1.751778483390808, + "learning_rate": 8.52971119040248e-06, + "loss": 0.4516, + "step": 17774 + }, + { + "epoch": 0.5606371234821006, + "grad_norm": 1.67094886302948, + "learning_rate": 8.528700743664597e-06, + "loss": 0.4815, + "step": 17775 + }, + { + "epoch": 0.5606686642485412, + "grad_norm": 1.667169213294983, + "learning_rate": 8.52769031228087e-06, + "loss": 0.3766, + "step": 17776 + }, + { + "epoch": 0.5607002050149819, + "grad_norm": 1.4328784942626953, + "learning_rate": 8.526679896261838e-06, + "loss": 0.4532, + "step": 17777 + }, + { + "epoch": 0.5607317457814225, + "grad_norm": 2.373857259750366, + "learning_rate": 8.52566949561804e-06, + "loss": 0.4728, + "step": 17778 + }, + { + "epoch": 0.5607632865478631, + "grad_norm": 1.3997302055358887, + "learning_rate": 8.524659110360029e-06, + "loss": 0.7834, + "step": 17779 + }, + { + "epoch": 0.5607948273143037, + "grad_norm": 2.131399393081665, + "learning_rate": 8.523648740498348e-06, + "loss": 0.4393, + "step": 17780 + }, + { + "epoch": 0.5608263680807444, + "grad_norm": 1.5033576488494873, + "learning_rate": 8.522638386043535e-06, + "loss": 0.771, + "step": 17781 + }, + { + "epoch": 0.560857908847185, + "grad_norm": 1.8121732473373413, + "learning_rate": 8.52162804700614e-06, + "loss": 0.4656, + "step": 17782 + }, + { + "epoch": 0.5608894496136256, + "grad_norm": 2.277555227279663, + "learning_rate": 8.520617723396702e-06, + "loss": 0.4497, + "step": 17783 + }, + { + "epoch": 0.5609209903800663, + "grad_norm": 1.526341438293457, + "learning_rate": 8.519607415225767e-06, + "loss": 0.4622, + "step": 17784 + }, + { + "epoch": 0.5609525311465069, + "grad_norm": 2.0919723510742188, + "learning_rate": 8.518597122503882e-06, + "loss": 0.4796, + "step": 17785 + }, + { + "epoch": 0.5609840719129475, + "grad_norm": 2.6605794429779053, + "learning_rate": 8.51758684524158e-06, + "loss": 0.4952, + "step": 17786 + }, + { + "epoch": 0.5610156126793882, + "grad_norm": 1.448075532913208, + "learning_rate": 8.516576583449415e-06, + "loss": 0.4639, + "step": 17787 + }, + { + "epoch": 0.5610471534458288, + "grad_norm": 1.7571357488632202, + "learning_rate": 8.515566337137925e-06, + "loss": 0.4683, + "step": 17788 + }, + { + "epoch": 0.5610786942122694, + "grad_norm": 1.6635223627090454, + "learning_rate": 8.514556106317649e-06, + "loss": 0.4514, + "step": 17789 + }, + { + "epoch": 0.5611102349787099, + "grad_norm": 1.5278398990631104, + "learning_rate": 8.513545890999138e-06, + "loss": 0.4639, + "step": 17790 + }, + { + "epoch": 0.5611417757451506, + "grad_norm": 2.149765968322754, + "learning_rate": 8.51253569119293e-06, + "loss": 0.77, + "step": 17791 + }, + { + "epoch": 0.5611733165115912, + "grad_norm": 2.0068750381469727, + "learning_rate": 8.511525506909562e-06, + "loss": 0.4713, + "step": 17792 + }, + { + "epoch": 0.5612048572780318, + "grad_norm": 1.9140760898590088, + "learning_rate": 8.510515338159587e-06, + "loss": 0.4686, + "step": 17793 + }, + { + "epoch": 0.5612363980444725, + "grad_norm": 1.5868467092514038, + "learning_rate": 8.50950518495354e-06, + "loss": 0.4596, + "step": 17794 + }, + { + "epoch": 0.5612679388109131, + "grad_norm": 1.4043060541152954, + "learning_rate": 8.508495047301958e-06, + "loss": 0.3687, + "step": 17795 + }, + { + "epoch": 0.5612994795773537, + "grad_norm": 1.4035625457763672, + "learning_rate": 8.507484925215393e-06, + "loss": 0.4518, + "step": 17796 + }, + { + "epoch": 0.5613310203437943, + "grad_norm": 1.6186861991882324, + "learning_rate": 8.506474818704382e-06, + "loss": 0.4349, + "step": 17797 + }, + { + "epoch": 0.561362561110235, + "grad_norm": 1.7344615459442139, + "learning_rate": 8.505464727779462e-06, + "loss": 0.4864, + "step": 17798 + }, + { + "epoch": 0.5613941018766756, + "grad_norm": 1.273669958114624, + "learning_rate": 8.504454652451182e-06, + "loss": 0.4349, + "step": 17799 + }, + { + "epoch": 0.5614256426431162, + "grad_norm": 1.7544211149215698, + "learning_rate": 8.50344459273008e-06, + "loss": 0.4677, + "step": 17800 + }, + { + "epoch": 0.5614571834095569, + "grad_norm": 1.89776611328125, + "learning_rate": 8.502434548626692e-06, + "loss": 0.4605, + "step": 17801 + }, + { + "epoch": 0.5614887241759975, + "grad_norm": 3.270029306411743, + "learning_rate": 8.501424520151566e-06, + "loss": 0.4406, + "step": 17802 + }, + { + "epoch": 0.5615202649424381, + "grad_norm": 1.6647350788116455, + "learning_rate": 8.500414507315234e-06, + "loss": 0.4652, + "step": 17803 + }, + { + "epoch": 0.5615518057088787, + "grad_norm": 1.6202805042266846, + "learning_rate": 8.499404510128245e-06, + "loss": 0.466, + "step": 17804 + }, + { + "epoch": 0.5615833464753194, + "grad_norm": 2.2385804653167725, + "learning_rate": 8.498394528601134e-06, + "loss": 0.4644, + "step": 17805 + }, + { + "epoch": 0.56161488724176, + "grad_norm": 1.645725131034851, + "learning_rate": 8.49738456274444e-06, + "loss": 0.458, + "step": 17806 + }, + { + "epoch": 0.5616464280082006, + "grad_norm": 2.2706594467163086, + "learning_rate": 8.49637461256871e-06, + "loss": 0.7428, + "step": 17807 + }, + { + "epoch": 0.5616779687746413, + "grad_norm": 2.396899700164795, + "learning_rate": 8.495364678084474e-06, + "loss": 0.4257, + "step": 17808 + }, + { + "epoch": 0.5617095095410819, + "grad_norm": 1.5245651006698608, + "learning_rate": 8.494354759302274e-06, + "loss": 0.4797, + "step": 17809 + }, + { + "epoch": 0.5617410503075225, + "grad_norm": 2.501861095428467, + "learning_rate": 8.493344856232658e-06, + "loss": 0.4847, + "step": 17810 + }, + { + "epoch": 0.561772591073963, + "grad_norm": 2.5515899658203125, + "learning_rate": 8.492334968886155e-06, + "loss": 0.4621, + "step": 17811 + }, + { + "epoch": 0.5618041318404037, + "grad_norm": 1.470610499382019, + "learning_rate": 8.491325097273306e-06, + "loss": 0.433, + "step": 17812 + }, + { + "epoch": 0.5618356726068443, + "grad_norm": 1.5665838718414307, + "learning_rate": 8.490315241404653e-06, + "loss": 0.4646, + "step": 17813 + }, + { + "epoch": 0.5618672133732849, + "grad_norm": 1.3810789585113525, + "learning_rate": 8.489305401290733e-06, + "loss": 0.4642, + "step": 17814 + }, + { + "epoch": 0.5618987541397256, + "grad_norm": 1.6489856243133545, + "learning_rate": 8.48829557694208e-06, + "loss": 0.4857, + "step": 17815 + }, + { + "epoch": 0.5619302949061662, + "grad_norm": 1.446804165840149, + "learning_rate": 8.48728576836924e-06, + "loss": 0.7404, + "step": 17816 + }, + { + "epoch": 0.5619618356726068, + "grad_norm": 4.720310211181641, + "learning_rate": 8.486275975582749e-06, + "loss": 0.4458, + "step": 17817 + }, + { + "epoch": 0.5619933764390475, + "grad_norm": 1.391783595085144, + "learning_rate": 8.485266198593138e-06, + "loss": 0.4444, + "step": 17818 + }, + { + "epoch": 0.5620249172054881, + "grad_norm": 1.640129804611206, + "learning_rate": 8.484256437410955e-06, + "loss": 0.4577, + "step": 17819 + }, + { + "epoch": 0.5620564579719287, + "grad_norm": 1.6972724199295044, + "learning_rate": 8.48324669204673e-06, + "loss": 0.4662, + "step": 17820 + }, + { + "epoch": 0.5620879987383693, + "grad_norm": 1.8146945238113403, + "learning_rate": 8.482236962511007e-06, + "loss": 0.4602, + "step": 17821 + }, + { + "epoch": 0.56211953950481, + "grad_norm": 1.913658618927002, + "learning_rate": 8.48122724881432e-06, + "loss": 0.7521, + "step": 17822 + }, + { + "epoch": 0.5621510802712506, + "grad_norm": 1.5255074501037598, + "learning_rate": 8.4802175509672e-06, + "loss": 0.4721, + "step": 17823 + }, + { + "epoch": 0.5621826210376912, + "grad_norm": 1.5025935173034668, + "learning_rate": 8.479207868980195e-06, + "loss": 0.7702, + "step": 17824 + }, + { + "epoch": 0.5622141618041319, + "grad_norm": 1.6485213041305542, + "learning_rate": 8.47819820286384e-06, + "loss": 0.4579, + "step": 17825 + }, + { + "epoch": 0.5622457025705725, + "grad_norm": 1.4304587841033936, + "learning_rate": 8.477188552628658e-06, + "loss": 0.4873, + "step": 17826 + }, + { + "epoch": 0.5622772433370131, + "grad_norm": 1.7945947647094727, + "learning_rate": 8.476178918285205e-06, + "loss": 0.4496, + "step": 17827 + }, + { + "epoch": 0.5623087841034538, + "grad_norm": 1.5343090295791626, + "learning_rate": 8.475169299844007e-06, + "loss": 0.4519, + "step": 17828 + }, + { + "epoch": 0.5623403248698944, + "grad_norm": 1.386088490486145, + "learning_rate": 8.474159697315597e-06, + "loss": 0.4306, + "step": 17829 + }, + { + "epoch": 0.562371865636335, + "grad_norm": 1.5995489358901978, + "learning_rate": 8.473150110710518e-06, + "loss": 0.4871, + "step": 17830 + }, + { + "epoch": 0.5624034064027756, + "grad_norm": 1.610896110534668, + "learning_rate": 8.472140540039306e-06, + "loss": 0.425, + "step": 17831 + }, + { + "epoch": 0.5624349471692162, + "grad_norm": 2.0147781372070312, + "learning_rate": 8.471130985312488e-06, + "loss": 0.4927, + "step": 17832 + }, + { + "epoch": 0.5624664879356568, + "grad_norm": 1.5607300996780396, + "learning_rate": 8.47012144654061e-06, + "loss": 0.4532, + "step": 17833 + }, + { + "epoch": 0.5624980287020974, + "grad_norm": 1.3061273097991943, + "learning_rate": 8.4691119237342e-06, + "loss": 0.4677, + "step": 17834 + }, + { + "epoch": 0.562529569468538, + "grad_norm": 2.0631728172302246, + "learning_rate": 8.468102416903796e-06, + "loss": 0.4671, + "step": 17835 + }, + { + "epoch": 0.5625611102349787, + "grad_norm": 1.4688832759857178, + "learning_rate": 8.467092926059936e-06, + "loss": 0.4486, + "step": 17836 + }, + { + "epoch": 0.5625926510014193, + "grad_norm": 2.898343563079834, + "learning_rate": 8.466083451213145e-06, + "loss": 0.462, + "step": 17837 + }, + { + "epoch": 0.5626241917678599, + "grad_norm": 1.455977439880371, + "learning_rate": 8.46507399237397e-06, + "loss": 0.7642, + "step": 17838 + }, + { + "epoch": 0.5626557325343006, + "grad_norm": 1.2999659776687622, + "learning_rate": 8.464064549552939e-06, + "loss": 0.3734, + "step": 17839 + }, + { + "epoch": 0.5626872733007412, + "grad_norm": 1.8396952152252197, + "learning_rate": 8.463055122760586e-06, + "loss": 0.467, + "step": 17840 + }, + { + "epoch": 0.5627188140671818, + "grad_norm": 1.6834625005722046, + "learning_rate": 8.462045712007447e-06, + "loss": 0.4711, + "step": 17841 + }, + { + "epoch": 0.5627503548336225, + "grad_norm": 1.9867924451828003, + "learning_rate": 8.461036317304055e-06, + "loss": 0.4576, + "step": 17842 + }, + { + "epoch": 0.5627818956000631, + "grad_norm": 2.9476466178894043, + "learning_rate": 8.460026938660941e-06, + "loss": 0.463, + "step": 17843 + }, + { + "epoch": 0.5628134363665037, + "grad_norm": 1.8448700904846191, + "learning_rate": 8.459017576088645e-06, + "loss": 0.4626, + "step": 17844 + }, + { + "epoch": 0.5628449771329443, + "grad_norm": 1.6539639234542847, + "learning_rate": 8.458008229597699e-06, + "loss": 0.4627, + "step": 17845 + }, + { + "epoch": 0.562876517899385, + "grad_norm": 1.608854055404663, + "learning_rate": 8.456998899198627e-06, + "loss": 0.4566, + "step": 17846 + }, + { + "epoch": 0.5629080586658256, + "grad_norm": 1.5420587062835693, + "learning_rate": 8.455989584901976e-06, + "loss": 0.4435, + "step": 17847 + }, + { + "epoch": 0.5629395994322662, + "grad_norm": 1.560494065284729, + "learning_rate": 8.454980286718272e-06, + "loss": 0.4383, + "step": 17848 + }, + { + "epoch": 0.5629711401987069, + "grad_norm": 1.682437539100647, + "learning_rate": 8.453971004658044e-06, + "loss": 0.4673, + "step": 17849 + }, + { + "epoch": 0.5630026809651475, + "grad_norm": 1.7059580087661743, + "learning_rate": 8.452961738731835e-06, + "loss": 0.4353, + "step": 17850 + }, + { + "epoch": 0.5630342217315881, + "grad_norm": 2.148408889770508, + "learning_rate": 8.451952488950167e-06, + "loss": 0.489, + "step": 17851 + }, + { + "epoch": 0.5630657624980288, + "grad_norm": 1.818800687789917, + "learning_rate": 8.45094325532358e-06, + "loss": 0.4606, + "step": 17852 + }, + { + "epoch": 0.5630973032644693, + "grad_norm": 2.437270402908325, + "learning_rate": 8.449934037862603e-06, + "loss": 0.4546, + "step": 17853 + }, + { + "epoch": 0.5631288440309099, + "grad_norm": 1.807701587677002, + "learning_rate": 8.448924836577762e-06, + "loss": 0.4699, + "step": 17854 + }, + { + "epoch": 0.5631603847973505, + "grad_norm": 1.5491126775741577, + "learning_rate": 8.447915651479601e-06, + "loss": 0.4743, + "step": 17855 + }, + { + "epoch": 0.5631919255637912, + "grad_norm": 1.4656559228897095, + "learning_rate": 8.446906482578645e-06, + "loss": 0.7666, + "step": 17856 + }, + { + "epoch": 0.5632234663302318, + "grad_norm": 1.6693530082702637, + "learning_rate": 8.445897329885421e-06, + "loss": 0.3983, + "step": 17857 + }, + { + "epoch": 0.5632550070966724, + "grad_norm": 1.5054773092269897, + "learning_rate": 8.44488819341047e-06, + "loss": 0.4524, + "step": 17858 + }, + { + "epoch": 0.563286547863113, + "grad_norm": 1.6691454648971558, + "learning_rate": 8.443879073164319e-06, + "loss": 0.4773, + "step": 17859 + }, + { + "epoch": 0.5633180886295537, + "grad_norm": 1.91192626953125, + "learning_rate": 8.442869969157494e-06, + "loss": 0.4552, + "step": 17860 + }, + { + "epoch": 0.5633496293959943, + "grad_norm": 1.8027656078338623, + "learning_rate": 8.441860881400533e-06, + "loss": 0.4298, + "step": 17861 + }, + { + "epoch": 0.5633811701624349, + "grad_norm": 1.5067685842514038, + "learning_rate": 8.440851809903965e-06, + "loss": 0.4832, + "step": 17862 + }, + { + "epoch": 0.5634127109288756, + "grad_norm": 1.4254008531570435, + "learning_rate": 8.439842754678312e-06, + "loss": 0.4627, + "step": 17863 + }, + { + "epoch": 0.5634442516953162, + "grad_norm": 1.5320053100585938, + "learning_rate": 8.438833715734118e-06, + "loss": 0.4732, + "step": 17864 + }, + { + "epoch": 0.5634757924617568, + "grad_norm": 1.4609428644180298, + "learning_rate": 8.437824693081903e-06, + "loss": 0.4651, + "step": 17865 + }, + { + "epoch": 0.5635073332281975, + "grad_norm": 1.8240782022476196, + "learning_rate": 8.436815686732199e-06, + "loss": 0.3738, + "step": 17866 + }, + { + "epoch": 0.5635388739946381, + "grad_norm": 1.7101175785064697, + "learning_rate": 8.43580669669554e-06, + "loss": 0.4734, + "step": 17867 + }, + { + "epoch": 0.5635704147610787, + "grad_norm": 1.5285848379135132, + "learning_rate": 8.434797722982452e-06, + "loss": 0.431, + "step": 17868 + }, + { + "epoch": 0.5636019555275193, + "grad_norm": 3.2148756980895996, + "learning_rate": 8.433788765603465e-06, + "loss": 0.4459, + "step": 17869 + }, + { + "epoch": 0.56363349629396, + "grad_norm": 3.0112409591674805, + "learning_rate": 8.43277982456911e-06, + "loss": 0.4635, + "step": 17870 + }, + { + "epoch": 0.5636650370604006, + "grad_norm": 11.899565696716309, + "learning_rate": 8.431770899889911e-06, + "loss": 0.3587, + "step": 17871 + }, + { + "epoch": 0.5636965778268412, + "grad_norm": 2.6765799522399902, + "learning_rate": 8.430761991576403e-06, + "loss": 0.4271, + "step": 17872 + }, + { + "epoch": 0.5637281185932819, + "grad_norm": 1.4862152338027954, + "learning_rate": 8.429753099639115e-06, + "loss": 0.4666, + "step": 17873 + }, + { + "epoch": 0.5637596593597224, + "grad_norm": 1.5138177871704102, + "learning_rate": 8.428744224088565e-06, + "loss": 0.459, + "step": 17874 + }, + { + "epoch": 0.563791200126163, + "grad_norm": 1.4431464672088623, + "learning_rate": 8.427735364935295e-06, + "loss": 0.7774, + "step": 17875 + }, + { + "epoch": 0.5638227408926036, + "grad_norm": 3.6112873554229736, + "learning_rate": 8.426726522189826e-06, + "loss": 0.423, + "step": 17876 + }, + { + "epoch": 0.5638542816590443, + "grad_norm": 1.7057371139526367, + "learning_rate": 8.425717695862684e-06, + "loss": 0.3473, + "step": 17877 + }, + { + "epoch": 0.5638858224254849, + "grad_norm": 1.4555070400238037, + "learning_rate": 8.424708885964404e-06, + "loss": 0.4876, + "step": 17878 + }, + { + "epoch": 0.5639173631919255, + "grad_norm": 5.579436779022217, + "learning_rate": 8.423700092505512e-06, + "loss": 0.3885, + "step": 17879 + }, + { + "epoch": 0.5639489039583662, + "grad_norm": 1.818454623222351, + "learning_rate": 8.422691315496529e-06, + "loss": 0.4402, + "step": 17880 + }, + { + "epoch": 0.5639804447248068, + "grad_norm": 1.791541337966919, + "learning_rate": 8.42168255494799e-06, + "loss": 0.4431, + "step": 17881 + }, + { + "epoch": 0.5640119854912474, + "grad_norm": 1.4859459400177002, + "learning_rate": 8.420673810870414e-06, + "loss": 0.4719, + "step": 17882 + }, + { + "epoch": 0.564043526257688, + "grad_norm": 1.603353500366211, + "learning_rate": 8.419665083274338e-06, + "loss": 0.4512, + "step": 17883 + }, + { + "epoch": 0.5640750670241287, + "grad_norm": 3.0105130672454834, + "learning_rate": 8.418656372170284e-06, + "loss": 0.4085, + "step": 17884 + }, + { + "epoch": 0.5641066077905693, + "grad_norm": 1.7643717527389526, + "learning_rate": 8.417647677568774e-06, + "loss": 0.4772, + "step": 17885 + }, + { + "epoch": 0.5641381485570099, + "grad_norm": 3.2411928176879883, + "learning_rate": 8.416638999480344e-06, + "loss": 0.7584, + "step": 17886 + }, + { + "epoch": 0.5641696893234506, + "grad_norm": 1.7041116952896118, + "learning_rate": 8.415630337915515e-06, + "loss": 0.4126, + "step": 17887 + }, + { + "epoch": 0.5642012300898912, + "grad_norm": 26.68439292907715, + "learning_rate": 8.41462169288481e-06, + "loss": 0.7353, + "step": 17888 + }, + { + "epoch": 0.5642327708563318, + "grad_norm": 1.5102620124816895, + "learning_rate": 8.41361306439876e-06, + "loss": 0.454, + "step": 17889 + }, + { + "epoch": 0.5642643116227725, + "grad_norm": 1.6367378234863281, + "learning_rate": 8.412604452467893e-06, + "loss": 0.4898, + "step": 17890 + }, + { + "epoch": 0.5642958523892131, + "grad_norm": 1.8544622659683228, + "learning_rate": 8.411595857102728e-06, + "loss": 0.474, + "step": 17891 + }, + { + "epoch": 0.5643273931556537, + "grad_norm": 2.1255884170532227, + "learning_rate": 8.410587278313795e-06, + "loss": 0.4368, + "step": 17892 + }, + { + "epoch": 0.5643589339220944, + "grad_norm": 1.6892225742340088, + "learning_rate": 8.40957871611162e-06, + "loss": 0.4677, + "step": 17893 + }, + { + "epoch": 0.564390474688535, + "grad_norm": 1.6888753175735474, + "learning_rate": 8.408570170506718e-06, + "loss": 0.4142, + "step": 17894 + }, + { + "epoch": 0.5644220154549755, + "grad_norm": 1.672939658164978, + "learning_rate": 8.407561641509629e-06, + "loss": 0.7494, + "step": 17895 + }, + { + "epoch": 0.5644535562214161, + "grad_norm": 2.3779640197753906, + "learning_rate": 8.40655312913087e-06, + "loss": 0.4708, + "step": 17896 + }, + { + "epoch": 0.5644850969878568, + "grad_norm": 1.369296908378601, + "learning_rate": 8.405544633380963e-06, + "loss": 0.4986, + "step": 17897 + }, + { + "epoch": 0.5645166377542974, + "grad_norm": 1.5252529382705688, + "learning_rate": 8.40453615427044e-06, + "loss": 0.4414, + "step": 17898 + }, + { + "epoch": 0.564548178520738, + "grad_norm": 1.6907991170883179, + "learning_rate": 8.403527691809818e-06, + "loss": 0.3457, + "step": 17899 + }, + { + "epoch": 0.5645797192871786, + "grad_norm": 1.7466715574264526, + "learning_rate": 8.402519246009627e-06, + "loss": 0.4699, + "step": 17900 + }, + { + "epoch": 0.5646112600536193, + "grad_norm": 1.4969948530197144, + "learning_rate": 8.401510816880387e-06, + "loss": 0.4582, + "step": 17901 + }, + { + "epoch": 0.5646428008200599, + "grad_norm": 2.150850534439087, + "learning_rate": 8.40050240443262e-06, + "loss": 0.4577, + "step": 17902 + }, + { + "epoch": 0.5646743415865005, + "grad_norm": 1.8742268085479736, + "learning_rate": 8.399494008676856e-06, + "loss": 0.4549, + "step": 17903 + }, + { + "epoch": 0.5647058823529412, + "grad_norm": 1.956938624382019, + "learning_rate": 8.398485629623613e-06, + "loss": 0.4197, + "step": 17904 + }, + { + "epoch": 0.5647374231193818, + "grad_norm": 1.8008222579956055, + "learning_rate": 8.397477267283414e-06, + "loss": 0.4799, + "step": 17905 + }, + { + "epoch": 0.5647689638858224, + "grad_norm": 1.646689534187317, + "learning_rate": 8.39646892166679e-06, + "loss": 0.4446, + "step": 17906 + }, + { + "epoch": 0.5648005046522631, + "grad_norm": 1.5158528089523315, + "learning_rate": 8.395460592784255e-06, + "loss": 0.4645, + "step": 17907 + }, + { + "epoch": 0.5648320454187037, + "grad_norm": 2.232844591140747, + "learning_rate": 8.394452280646333e-06, + "loss": 0.4174, + "step": 17908 + }, + { + "epoch": 0.5648635861851443, + "grad_norm": 2.5884406566619873, + "learning_rate": 8.393443985263548e-06, + "loss": 0.4496, + "step": 17909 + }, + { + "epoch": 0.564895126951585, + "grad_norm": 3.1070849895477295, + "learning_rate": 8.392435706646425e-06, + "loss": 0.4774, + "step": 17910 + }, + { + "epoch": 0.5649266677180256, + "grad_norm": 1.7041826248168945, + "learning_rate": 8.39142744480548e-06, + "loss": 0.482, + "step": 17911 + }, + { + "epoch": 0.5649582084844662, + "grad_norm": 1.7851519584655762, + "learning_rate": 8.390419199751244e-06, + "loss": 0.4524, + "step": 17912 + }, + { + "epoch": 0.5649897492509068, + "grad_norm": 1.5874946117401123, + "learning_rate": 8.389410971494232e-06, + "loss": 0.4287, + "step": 17913 + }, + { + "epoch": 0.5650212900173475, + "grad_norm": 1.7311915159225464, + "learning_rate": 8.388402760044962e-06, + "loss": 0.4893, + "step": 17914 + }, + { + "epoch": 0.5650528307837881, + "grad_norm": 1.6522570848464966, + "learning_rate": 8.387394565413965e-06, + "loss": 0.4753, + "step": 17915 + }, + { + "epoch": 0.5650843715502287, + "grad_norm": 2.030782699584961, + "learning_rate": 8.386386387611753e-06, + "loss": 0.4796, + "step": 17916 + }, + { + "epoch": 0.5651159123166692, + "grad_norm": 1.8812017440795898, + "learning_rate": 8.385378226648858e-06, + "loss": 0.7655, + "step": 17917 + }, + { + "epoch": 0.5651474530831099, + "grad_norm": 2.6792924404144287, + "learning_rate": 8.384370082535793e-06, + "loss": 0.4575, + "step": 17918 + }, + { + "epoch": 0.5651789938495505, + "grad_norm": 1.679499626159668, + "learning_rate": 8.38336195528308e-06, + "loss": 0.4557, + "step": 17919 + }, + { + "epoch": 0.5652105346159911, + "grad_norm": 1.6600086688995361, + "learning_rate": 8.382353844901242e-06, + "loss": 0.4756, + "step": 17920 + }, + { + "epoch": 0.5652420753824318, + "grad_norm": 2.9299848079681396, + "learning_rate": 8.381345751400798e-06, + "loss": 0.457, + "step": 17921 + }, + { + "epoch": 0.5652736161488724, + "grad_norm": 1.4520443677902222, + "learning_rate": 8.380337674792263e-06, + "loss": 0.3615, + "step": 17922 + }, + { + "epoch": 0.565305156915313, + "grad_norm": 1.8084980249404907, + "learning_rate": 8.379329615086169e-06, + "loss": 0.4201, + "step": 17923 + }, + { + "epoch": 0.5653366976817537, + "grad_norm": 1.5031346082687378, + "learning_rate": 8.378321572293025e-06, + "loss": 0.7482, + "step": 17924 + }, + { + "epoch": 0.5653682384481943, + "grad_norm": 3.310854196548462, + "learning_rate": 8.377313546423352e-06, + "loss": 0.4744, + "step": 17925 + }, + { + "epoch": 0.5653997792146349, + "grad_norm": 1.2966649532318115, + "learning_rate": 8.376305537487677e-06, + "loss": 0.7724, + "step": 17926 + }, + { + "epoch": 0.5654313199810755, + "grad_norm": 1.414705753326416, + "learning_rate": 8.375297545496514e-06, + "loss": 0.4652, + "step": 17927 + }, + { + "epoch": 0.5654628607475162, + "grad_norm": 1.4931095838546753, + "learning_rate": 8.37428957046038e-06, + "loss": 0.4705, + "step": 17928 + }, + { + "epoch": 0.5654944015139568, + "grad_norm": 7.403034687042236, + "learning_rate": 8.373281612389796e-06, + "loss": 0.4382, + "step": 17929 + }, + { + "epoch": 0.5655259422803974, + "grad_norm": 1.486939787864685, + "learning_rate": 8.372273671295283e-06, + "loss": 0.4453, + "step": 17930 + }, + { + "epoch": 0.5655574830468381, + "grad_norm": 1.6081990003585815, + "learning_rate": 8.37126574718736e-06, + "loss": 0.4603, + "step": 17931 + }, + { + "epoch": 0.5655890238132787, + "grad_norm": 1.8677515983581543, + "learning_rate": 8.370257840076543e-06, + "loss": 0.4573, + "step": 17932 + }, + { + "epoch": 0.5656205645797193, + "grad_norm": 1.7961723804473877, + "learning_rate": 8.369249949973348e-06, + "loss": 0.4749, + "step": 17933 + }, + { + "epoch": 0.56565210534616, + "grad_norm": 1.4178041219711304, + "learning_rate": 8.368242076888298e-06, + "loss": 0.4693, + "step": 17934 + }, + { + "epoch": 0.5656836461126006, + "grad_norm": 2.981809139251709, + "learning_rate": 8.36723422083191e-06, + "loss": 0.3414, + "step": 17935 + }, + { + "epoch": 0.5657151868790412, + "grad_norm": 1.7805724143981934, + "learning_rate": 8.366226381814698e-06, + "loss": 0.4007, + "step": 17936 + }, + { + "epoch": 0.5657467276454818, + "grad_norm": 1.477690577507019, + "learning_rate": 8.365218559847185e-06, + "loss": 0.4687, + "step": 17937 + }, + { + "epoch": 0.5657782684119224, + "grad_norm": 2.200101375579834, + "learning_rate": 8.364210754939886e-06, + "loss": 0.4309, + "step": 17938 + }, + { + "epoch": 0.565809809178363, + "grad_norm": 2.4703893661499023, + "learning_rate": 8.363202967103315e-06, + "loss": 0.3536, + "step": 17939 + }, + { + "epoch": 0.5658413499448036, + "grad_norm": 1.6565468311309814, + "learning_rate": 8.362195196347996e-06, + "loss": 0.4773, + "step": 17940 + }, + { + "epoch": 0.5658728907112442, + "grad_norm": 1.8180149793624878, + "learning_rate": 8.36118744268444e-06, + "loss": 0.4771, + "step": 17941 + }, + { + "epoch": 0.5659044314776849, + "grad_norm": 1.6817783117294312, + "learning_rate": 8.360179706123161e-06, + "loss": 0.4547, + "step": 17942 + }, + { + "epoch": 0.5659359722441255, + "grad_norm": 1.59281587600708, + "learning_rate": 8.359171986674686e-06, + "loss": 0.4537, + "step": 17943 + }, + { + "epoch": 0.5659675130105661, + "grad_norm": 1.8348923921585083, + "learning_rate": 8.358164284349525e-06, + "loss": 0.7828, + "step": 17944 + }, + { + "epoch": 0.5659990537770068, + "grad_norm": 1.7382210493087769, + "learning_rate": 8.35715659915819e-06, + "loss": 0.4941, + "step": 17945 + }, + { + "epoch": 0.5660305945434474, + "grad_norm": 1.6853761672973633, + "learning_rate": 8.356148931111205e-06, + "loss": 0.4674, + "step": 17946 + }, + { + "epoch": 0.566062135309888, + "grad_norm": 2.004533529281616, + "learning_rate": 8.355141280219081e-06, + "loss": 0.459, + "step": 17947 + }, + { + "epoch": 0.5660936760763287, + "grad_norm": 2.9246771335601807, + "learning_rate": 8.354133646492338e-06, + "loss": 0.4314, + "step": 17948 + }, + { + "epoch": 0.5661252168427693, + "grad_norm": 1.9964700937271118, + "learning_rate": 8.353126029941486e-06, + "loss": 0.4432, + "step": 17949 + }, + { + "epoch": 0.5661567576092099, + "grad_norm": 2.244955539703369, + "learning_rate": 8.35211843057704e-06, + "loss": 0.4383, + "step": 17950 + }, + { + "epoch": 0.5661882983756505, + "grad_norm": 1.3203247785568237, + "learning_rate": 8.351110848409523e-06, + "loss": 0.4646, + "step": 17951 + }, + { + "epoch": 0.5662198391420912, + "grad_norm": 1.8058477640151978, + "learning_rate": 8.350103283449443e-06, + "loss": 0.4245, + "step": 17952 + }, + { + "epoch": 0.5662513799085318, + "grad_norm": 1.6688767671585083, + "learning_rate": 8.349095735707313e-06, + "loss": 0.439, + "step": 17953 + }, + { + "epoch": 0.5662829206749724, + "grad_norm": 1.7728291749954224, + "learning_rate": 8.348088205193655e-06, + "loss": 0.4617, + "step": 17954 + }, + { + "epoch": 0.5663144614414131, + "grad_norm": 1.4480435848236084, + "learning_rate": 8.347080691918979e-06, + "loss": 0.4348, + "step": 17955 + }, + { + "epoch": 0.5663460022078537, + "grad_norm": 1.6794625520706177, + "learning_rate": 8.346073195893793e-06, + "loss": 0.4402, + "step": 17956 + }, + { + "epoch": 0.5663775429742943, + "grad_norm": 1.5227545499801636, + "learning_rate": 8.345065717128624e-06, + "loss": 0.4413, + "step": 17957 + }, + { + "epoch": 0.566409083740735, + "grad_norm": 2.015960931777954, + "learning_rate": 8.34405825563398e-06, + "loss": 0.4453, + "step": 17958 + }, + { + "epoch": 0.5664406245071755, + "grad_norm": 3.3459014892578125, + "learning_rate": 8.34305081142037e-06, + "loss": 0.4638, + "step": 17959 + }, + { + "epoch": 0.5664721652736161, + "grad_norm": 1.7187782526016235, + "learning_rate": 8.342043384498314e-06, + "loss": 0.4688, + "step": 17960 + }, + { + "epoch": 0.5665037060400567, + "grad_norm": 1.6570183038711548, + "learning_rate": 8.341035974878325e-06, + "loss": 0.4411, + "step": 17961 + }, + { + "epoch": 0.5665352468064974, + "grad_norm": 2.21695876121521, + "learning_rate": 8.340028582570906e-06, + "loss": 0.4537, + "step": 17962 + }, + { + "epoch": 0.566566787572938, + "grad_norm": 1.7248109579086304, + "learning_rate": 8.339021207586584e-06, + "loss": 0.4233, + "step": 17963 + }, + { + "epoch": 0.5665983283393786, + "grad_norm": 3.078702926635742, + "learning_rate": 8.33801384993586e-06, + "loss": 0.4477, + "step": 17964 + }, + { + "epoch": 0.5666298691058193, + "grad_norm": 1.4219379425048828, + "learning_rate": 8.337006509629258e-06, + "loss": 0.4682, + "step": 17965 + }, + { + "epoch": 0.5666614098722599, + "grad_norm": 2.203824758529663, + "learning_rate": 8.335999186677281e-06, + "loss": 0.4568, + "step": 17966 + }, + { + "epoch": 0.5666929506387005, + "grad_norm": 2.1117730140686035, + "learning_rate": 8.334991881090446e-06, + "loss": 0.4632, + "step": 17967 + }, + { + "epoch": 0.5667244914051411, + "grad_norm": 5.789543628692627, + "learning_rate": 8.333984592879264e-06, + "loss": 0.4518, + "step": 17968 + }, + { + "epoch": 0.5667560321715818, + "grad_norm": 1.3747941255569458, + "learning_rate": 8.332977322054244e-06, + "loss": 0.4838, + "step": 17969 + }, + { + "epoch": 0.5667875729380224, + "grad_norm": 1.6115152835845947, + "learning_rate": 8.3319700686259e-06, + "loss": 0.4276, + "step": 17970 + }, + { + "epoch": 0.566819113704463, + "grad_norm": 1.6973567008972168, + "learning_rate": 8.330962832604747e-06, + "loss": 0.452, + "step": 17971 + }, + { + "epoch": 0.5668506544709037, + "grad_norm": 1.4529579877853394, + "learning_rate": 8.329955614001293e-06, + "loss": 0.4484, + "step": 17972 + }, + { + "epoch": 0.5668821952373443, + "grad_norm": 2.565648078918457, + "learning_rate": 8.328948412826044e-06, + "loss": 0.4607, + "step": 17973 + }, + { + "epoch": 0.5669137360037849, + "grad_norm": 1.3679149150848389, + "learning_rate": 8.32794122908952e-06, + "loss": 0.4508, + "step": 17974 + }, + { + "epoch": 0.5669452767702255, + "grad_norm": 1.6153533458709717, + "learning_rate": 8.326934062802227e-06, + "loss": 0.4645, + "step": 17975 + }, + { + "epoch": 0.5669768175366662, + "grad_norm": 1.772026777267456, + "learning_rate": 8.325926913974671e-06, + "loss": 0.4862, + "step": 17976 + }, + { + "epoch": 0.5670083583031068, + "grad_norm": 1.8257631063461304, + "learning_rate": 8.324919782617373e-06, + "loss": 0.4681, + "step": 17977 + }, + { + "epoch": 0.5670398990695474, + "grad_norm": 7.467095375061035, + "learning_rate": 8.323912668740836e-06, + "loss": 0.4721, + "step": 17978 + }, + { + "epoch": 0.5670714398359881, + "grad_norm": 1.4830859899520874, + "learning_rate": 8.322905572355573e-06, + "loss": 0.4588, + "step": 17979 + }, + { + "epoch": 0.5671029806024286, + "grad_norm": 3.4502875804901123, + "learning_rate": 8.321898493472093e-06, + "loss": 0.7345, + "step": 17980 + }, + { + "epoch": 0.5671345213688692, + "grad_norm": 1.5681394338607788, + "learning_rate": 8.320891432100902e-06, + "loss": 0.4475, + "step": 17981 + }, + { + "epoch": 0.5671660621353098, + "grad_norm": 1.7570375204086304, + "learning_rate": 8.319884388252518e-06, + "loss": 0.4472, + "step": 17982 + }, + { + "epoch": 0.5671976029017505, + "grad_norm": 2.183472156524658, + "learning_rate": 8.318877361937442e-06, + "loss": 0.4237, + "step": 17983 + }, + { + "epoch": 0.5672291436681911, + "grad_norm": 1.378514289855957, + "learning_rate": 8.317870353166183e-06, + "loss": 0.4652, + "step": 17984 + }, + { + "epoch": 0.5672606844346317, + "grad_norm": 1.4592102766036987, + "learning_rate": 8.316863361949257e-06, + "loss": 0.4424, + "step": 17985 + }, + { + "epoch": 0.5672922252010724, + "grad_norm": 1.4562747478485107, + "learning_rate": 8.315856388297168e-06, + "loss": 0.4651, + "step": 17986 + }, + { + "epoch": 0.567323765967513, + "grad_norm": 2.41428804397583, + "learning_rate": 8.314849432220424e-06, + "loss": 0.7693, + "step": 17987 + }, + { + "epoch": 0.5673553067339536, + "grad_norm": 1.8342756032943726, + "learning_rate": 8.313842493729538e-06, + "loss": 0.4424, + "step": 17988 + }, + { + "epoch": 0.5673868475003943, + "grad_norm": 1.5380700826644897, + "learning_rate": 8.312835572835011e-06, + "loss": 0.7517, + "step": 17989 + }, + { + "epoch": 0.5674183882668349, + "grad_norm": 1.928335428237915, + "learning_rate": 8.311828669547355e-06, + "loss": 0.7365, + "step": 17990 + }, + { + "epoch": 0.5674499290332755, + "grad_norm": 1.921749472618103, + "learning_rate": 8.31082178387708e-06, + "loss": 0.4748, + "step": 17991 + }, + { + "epoch": 0.5674814697997161, + "grad_norm": 1.4244507551193237, + "learning_rate": 8.309814915834694e-06, + "loss": 0.461, + "step": 17992 + }, + { + "epoch": 0.5675130105661568, + "grad_norm": 3.8270907402038574, + "learning_rate": 8.308808065430694e-06, + "loss": 0.4665, + "step": 17993 + }, + { + "epoch": 0.5675445513325974, + "grad_norm": 1.98607337474823, + "learning_rate": 8.3078012326756e-06, + "loss": 0.403, + "step": 17994 + }, + { + "epoch": 0.567576092099038, + "grad_norm": 1.6137863397598267, + "learning_rate": 8.30679441757991e-06, + "loss": 0.4749, + "step": 17995 + }, + { + "epoch": 0.5676076328654787, + "grad_norm": 1.6015695333480835, + "learning_rate": 8.30578762015414e-06, + "loss": 0.4313, + "step": 17996 + }, + { + "epoch": 0.5676391736319193, + "grad_norm": 2.6236231327056885, + "learning_rate": 8.30478084040879e-06, + "loss": 0.4892, + "step": 17997 + }, + { + "epoch": 0.5676707143983599, + "grad_norm": 1.6719181537628174, + "learning_rate": 8.303774078354367e-06, + "loss": 0.4549, + "step": 17998 + }, + { + "epoch": 0.5677022551648006, + "grad_norm": 1.5355666875839233, + "learning_rate": 8.302767334001382e-06, + "loss": 0.4393, + "step": 17999 + }, + { + "epoch": 0.5677337959312412, + "grad_norm": 1.5495610237121582, + "learning_rate": 8.301760607360336e-06, + "loss": 0.435, + "step": 18000 + }, + { + "epoch": 0.5677653366976817, + "grad_norm": 1.6742316484451294, + "learning_rate": 8.300753898441733e-06, + "loss": 0.4592, + "step": 18001 + }, + { + "epoch": 0.5677968774641223, + "grad_norm": 1.7274218797683716, + "learning_rate": 8.299747207256087e-06, + "loss": 0.7772, + "step": 18002 + }, + { + "epoch": 0.567828418230563, + "grad_norm": 3.6409640312194824, + "learning_rate": 8.298740533813899e-06, + "loss": 0.4524, + "step": 18003 + }, + { + "epoch": 0.5678599589970036, + "grad_norm": 1.6185593605041504, + "learning_rate": 8.29773387812567e-06, + "loss": 0.4517, + "step": 18004 + }, + { + "epoch": 0.5678914997634442, + "grad_norm": 1.6034642457962036, + "learning_rate": 8.296727240201913e-06, + "loss": 0.4757, + "step": 18005 + }, + { + "epoch": 0.5679230405298848, + "grad_norm": 1.3857300281524658, + "learning_rate": 8.295720620053133e-06, + "loss": 0.7492, + "step": 18006 + }, + { + "epoch": 0.5679545812963255, + "grad_norm": 1.4882686138153076, + "learning_rate": 8.294714017689826e-06, + "loss": 0.756, + "step": 18007 + }, + { + "epoch": 0.5679861220627661, + "grad_norm": 1.6537151336669922, + "learning_rate": 8.293707433122507e-06, + "loss": 0.4645, + "step": 18008 + }, + { + "epoch": 0.5680176628292067, + "grad_norm": 2.7396528720855713, + "learning_rate": 8.292700866361671e-06, + "loss": 0.4535, + "step": 18009 + }, + { + "epoch": 0.5680492035956474, + "grad_norm": 1.6997359991073608, + "learning_rate": 8.29169431741783e-06, + "loss": 0.443, + "step": 18010 + }, + { + "epoch": 0.568080744362088, + "grad_norm": 1.4002104997634888, + "learning_rate": 8.290687786301487e-06, + "loss": 0.7356, + "step": 18011 + }, + { + "epoch": 0.5681122851285286, + "grad_norm": 1.3519288301467896, + "learning_rate": 8.28968127302314e-06, + "loss": 0.4788, + "step": 18012 + }, + { + "epoch": 0.5681438258949693, + "grad_norm": 1.403178095817566, + "learning_rate": 8.2886747775933e-06, + "loss": 0.4494, + "step": 18013 + }, + { + "epoch": 0.5681753666614099, + "grad_norm": 1.414799451828003, + "learning_rate": 8.287668300022468e-06, + "loss": 0.4331, + "step": 18014 + }, + { + "epoch": 0.5682069074278505, + "grad_norm": 1.9628156423568726, + "learning_rate": 8.286661840321143e-06, + "loss": 0.4571, + "step": 18015 + }, + { + "epoch": 0.5682384481942911, + "grad_norm": 1.3883497714996338, + "learning_rate": 8.285655398499834e-06, + "loss": 0.4513, + "step": 18016 + }, + { + "epoch": 0.5682699889607318, + "grad_norm": 1.7916499376296997, + "learning_rate": 8.284648974569045e-06, + "loss": 0.4478, + "step": 18017 + }, + { + "epoch": 0.5683015297271724, + "grad_norm": 1.4891319274902344, + "learning_rate": 8.283642568539273e-06, + "loss": 0.4764, + "step": 18018 + }, + { + "epoch": 0.568333070493613, + "grad_norm": 1.6255494356155396, + "learning_rate": 8.282636180421026e-06, + "loss": 0.4758, + "step": 18019 + }, + { + "epoch": 0.5683646112600537, + "grad_norm": 1.7304266691207886, + "learning_rate": 8.281629810224803e-06, + "loss": 0.4816, + "step": 18020 + }, + { + "epoch": 0.5683961520264943, + "grad_norm": 1.4615325927734375, + "learning_rate": 8.280623457961102e-06, + "loss": 0.425, + "step": 18021 + }, + { + "epoch": 0.5684276927929348, + "grad_norm": 1.792182445526123, + "learning_rate": 8.279617123640436e-06, + "loss": 0.4413, + "step": 18022 + }, + { + "epoch": 0.5684592335593754, + "grad_norm": 1.7574712038040161, + "learning_rate": 8.278610807273302e-06, + "loss": 0.4238, + "step": 18023 + }, + { + "epoch": 0.5684907743258161, + "grad_norm": 1.6171866655349731, + "learning_rate": 8.277604508870196e-06, + "loss": 0.4452, + "step": 18024 + }, + { + "epoch": 0.5685223150922567, + "grad_norm": 1.6349903345108032, + "learning_rate": 8.276598228441627e-06, + "loss": 0.4759, + "step": 18025 + }, + { + "epoch": 0.5685538558586973, + "grad_norm": 2.660799980163574, + "learning_rate": 8.275591965998094e-06, + "loss": 0.7638, + "step": 18026 + }, + { + "epoch": 0.568585396625138, + "grad_norm": 1.6391602754592896, + "learning_rate": 8.274585721550098e-06, + "loss": 0.4591, + "step": 18027 + }, + { + "epoch": 0.5686169373915786, + "grad_norm": 1.695935845375061, + "learning_rate": 8.27357949510814e-06, + "loss": 0.4428, + "step": 18028 + }, + { + "epoch": 0.5686484781580192, + "grad_norm": 1.5023038387298584, + "learning_rate": 8.272573286682716e-06, + "loss": 0.4035, + "step": 18029 + }, + { + "epoch": 0.5686800189244599, + "grad_norm": 1.4710015058517456, + "learning_rate": 8.271567096284336e-06, + "loss": 0.4341, + "step": 18030 + }, + { + "epoch": 0.5687115596909005, + "grad_norm": 2.0579259395599365, + "learning_rate": 8.270560923923497e-06, + "loss": 0.4199, + "step": 18031 + }, + { + "epoch": 0.5687431004573411, + "grad_norm": 1.9298394918441772, + "learning_rate": 8.26955476961069e-06, + "loss": 0.4743, + "step": 18032 + }, + { + "epoch": 0.5687746412237817, + "grad_norm": 1.5859681367874146, + "learning_rate": 8.268548633356428e-06, + "loss": 0.4526, + "step": 18033 + }, + { + "epoch": 0.5688061819902224, + "grad_norm": 1.9407294988632202, + "learning_rate": 8.267542515171207e-06, + "loss": 0.4496, + "step": 18034 + }, + { + "epoch": 0.568837722756663, + "grad_norm": 1.5703738927841187, + "learning_rate": 8.26653641506552e-06, + "loss": 0.4621, + "step": 18035 + }, + { + "epoch": 0.5688692635231036, + "grad_norm": 1.3947525024414062, + "learning_rate": 8.265530333049876e-06, + "loss": 0.4799, + "step": 18036 + }, + { + "epoch": 0.5689008042895443, + "grad_norm": 1.4514511823654175, + "learning_rate": 8.264524269134769e-06, + "loss": 0.4529, + "step": 18037 + }, + { + "epoch": 0.5689323450559849, + "grad_norm": 1.4873952865600586, + "learning_rate": 8.263518223330698e-06, + "loss": 0.3678, + "step": 18038 + }, + { + "epoch": 0.5689638858224255, + "grad_norm": 1.4763106107711792, + "learning_rate": 8.262512195648164e-06, + "loss": 0.7547, + "step": 18039 + }, + { + "epoch": 0.5689954265888661, + "grad_norm": 1.59529447555542, + "learning_rate": 8.261506186097664e-06, + "loss": 0.4554, + "step": 18040 + }, + { + "epoch": 0.5690269673553068, + "grad_norm": 2.2102396488189697, + "learning_rate": 8.260500194689693e-06, + "loss": 0.4428, + "step": 18041 + }, + { + "epoch": 0.5690585081217474, + "grad_norm": 1.8862382173538208, + "learning_rate": 8.259494221434758e-06, + "loss": 0.4899, + "step": 18042 + }, + { + "epoch": 0.569090048888188, + "grad_norm": 1.8329343795776367, + "learning_rate": 8.258488266343347e-06, + "loss": 0.4525, + "step": 18043 + }, + { + "epoch": 0.5691215896546286, + "grad_norm": 2.05320143699646, + "learning_rate": 8.257482329425968e-06, + "loss": 0.4217, + "step": 18044 + }, + { + "epoch": 0.5691531304210692, + "grad_norm": 1.6944341659545898, + "learning_rate": 8.256476410693112e-06, + "loss": 0.4564, + "step": 18045 + }, + { + "epoch": 0.5691846711875098, + "grad_norm": 1.466464877128601, + "learning_rate": 8.25547051015528e-06, + "loss": 0.7563, + "step": 18046 + }, + { + "epoch": 0.5692162119539504, + "grad_norm": 1.4491404294967651, + "learning_rate": 8.254464627822967e-06, + "loss": 0.5003, + "step": 18047 + }, + { + "epoch": 0.5692477527203911, + "grad_norm": 1.6620774269104004, + "learning_rate": 8.253458763706672e-06, + "loss": 0.4812, + "step": 18048 + }, + { + "epoch": 0.5692792934868317, + "grad_norm": 1.3986773490905762, + "learning_rate": 8.252452917816887e-06, + "loss": 0.7264, + "step": 18049 + }, + { + "epoch": 0.5693108342532723, + "grad_norm": 1.44478178024292, + "learning_rate": 8.251447090164117e-06, + "loss": 0.4351, + "step": 18050 + }, + { + "epoch": 0.569342375019713, + "grad_norm": 1.9117350578308105, + "learning_rate": 8.250441280758852e-06, + "loss": 0.4726, + "step": 18051 + }, + { + "epoch": 0.5693739157861536, + "grad_norm": 1.4427605867385864, + "learning_rate": 8.24943548961159e-06, + "loss": 0.7124, + "step": 18052 + }, + { + "epoch": 0.5694054565525942, + "grad_norm": 1.7323945760726929, + "learning_rate": 8.24842971673283e-06, + "loss": 0.4537, + "step": 18053 + }, + { + "epoch": 0.5694369973190349, + "grad_norm": 3.283061981201172, + "learning_rate": 8.247423962133068e-06, + "loss": 0.485, + "step": 18054 + }, + { + "epoch": 0.5694685380854755, + "grad_norm": 1.4989210367202759, + "learning_rate": 8.24641822582279e-06, + "loss": 0.7558, + "step": 18055 + }, + { + "epoch": 0.5695000788519161, + "grad_norm": 1.4644725322723389, + "learning_rate": 8.245412507812507e-06, + "loss": 0.4346, + "step": 18056 + }, + { + "epoch": 0.5695316196183567, + "grad_norm": 1.8066545724868774, + "learning_rate": 8.244406808112703e-06, + "loss": 0.4283, + "step": 18057 + }, + { + "epoch": 0.5695631603847974, + "grad_norm": 2.658273935317993, + "learning_rate": 8.24340112673388e-06, + "loss": 0.4783, + "step": 18058 + }, + { + "epoch": 0.569594701151238, + "grad_norm": 2.267401933670044, + "learning_rate": 8.24239546368653e-06, + "loss": 0.4564, + "step": 18059 + }, + { + "epoch": 0.5696262419176786, + "grad_norm": 1.6064916849136353, + "learning_rate": 8.241389818981144e-06, + "loss": 0.4641, + "step": 18060 + }, + { + "epoch": 0.5696577826841193, + "grad_norm": 1.4824684858322144, + "learning_rate": 8.240384192628226e-06, + "loss": 0.4901, + "step": 18061 + }, + { + "epoch": 0.5696893234505599, + "grad_norm": 1.592783808708191, + "learning_rate": 8.239378584638264e-06, + "loss": 0.4295, + "step": 18062 + }, + { + "epoch": 0.5697208642170005, + "grad_norm": 1.5521007776260376, + "learning_rate": 8.23837299502175e-06, + "loss": 0.4604, + "step": 18063 + }, + { + "epoch": 0.5697524049834412, + "grad_norm": 1.2779924869537354, + "learning_rate": 8.237367423789184e-06, + "loss": 0.3283, + "step": 18064 + }, + { + "epoch": 0.5697839457498817, + "grad_norm": 1.3413211107254028, + "learning_rate": 8.23636187095106e-06, + "loss": 0.4691, + "step": 18065 + }, + { + "epoch": 0.5698154865163223, + "grad_norm": 1.9955251216888428, + "learning_rate": 8.235356336517866e-06, + "loss": 0.4579, + "step": 18066 + }, + { + "epoch": 0.5698470272827629, + "grad_norm": 1.5841137170791626, + "learning_rate": 8.234350820500101e-06, + "loss": 0.7504, + "step": 18067 + }, + { + "epoch": 0.5698785680492036, + "grad_norm": 2.3889973163604736, + "learning_rate": 8.233345322908256e-06, + "loss": 0.7632, + "step": 18068 + }, + { + "epoch": 0.5699101088156442, + "grad_norm": 1.366763949394226, + "learning_rate": 8.232339843752821e-06, + "loss": 0.465, + "step": 18069 + }, + { + "epoch": 0.5699416495820848, + "grad_norm": 1.842910647392273, + "learning_rate": 8.231334383044296e-06, + "loss": 0.4357, + "step": 18070 + }, + { + "epoch": 0.5699731903485254, + "grad_norm": 2.3666887283325195, + "learning_rate": 8.23032894079317e-06, + "loss": 0.4057, + "step": 18071 + }, + { + "epoch": 0.5700047311149661, + "grad_norm": 1.5012797117233276, + "learning_rate": 8.229323517009931e-06, + "loss": 0.4742, + "step": 18072 + }, + { + "epoch": 0.5700362718814067, + "grad_norm": 1.408534049987793, + "learning_rate": 8.22831811170508e-06, + "loss": 0.4923, + "step": 18073 + }, + { + "epoch": 0.5700678126478473, + "grad_norm": 2.8240530490875244, + "learning_rate": 8.227312724889103e-06, + "loss": 0.4643, + "step": 18074 + }, + { + "epoch": 0.570099353414288, + "grad_norm": 1.5187710523605347, + "learning_rate": 8.226307356572496e-06, + "loss": 0.4492, + "step": 18075 + }, + { + "epoch": 0.5701308941807286, + "grad_norm": 1.6270101070404053, + "learning_rate": 8.22530200676575e-06, + "loss": 0.4454, + "step": 18076 + }, + { + "epoch": 0.5701624349471692, + "grad_norm": 1.6240389347076416, + "learning_rate": 8.224296675479352e-06, + "loss": 0.4042, + "step": 18077 + }, + { + "epoch": 0.5701939757136099, + "grad_norm": 2.181856155395508, + "learning_rate": 8.223291362723803e-06, + "loss": 0.3817, + "step": 18078 + }, + { + "epoch": 0.5702255164800505, + "grad_norm": 1.401347279548645, + "learning_rate": 8.222286068509584e-06, + "loss": 0.4219, + "step": 18079 + }, + { + "epoch": 0.5702570572464911, + "grad_norm": 2.2191874980926514, + "learning_rate": 8.221280792847188e-06, + "loss": 0.3555, + "step": 18080 + }, + { + "epoch": 0.5702885980129317, + "grad_norm": 1.786440372467041, + "learning_rate": 8.220275535747112e-06, + "loss": 0.4348, + "step": 18081 + }, + { + "epoch": 0.5703201387793724, + "grad_norm": 2.0323212146759033, + "learning_rate": 8.219270297219842e-06, + "loss": 0.4261, + "step": 18082 + }, + { + "epoch": 0.570351679545813, + "grad_norm": 1.4250065088272095, + "learning_rate": 8.218265077275866e-06, + "loss": 0.4661, + "step": 18083 + }, + { + "epoch": 0.5703832203122536, + "grad_norm": 1.7506802082061768, + "learning_rate": 8.217259875925681e-06, + "loss": 0.4647, + "step": 18084 + }, + { + "epoch": 0.5704147610786943, + "grad_norm": 1.281625747680664, + "learning_rate": 8.216254693179775e-06, + "loss": 0.3681, + "step": 18085 + }, + { + "epoch": 0.5704463018451348, + "grad_norm": 1.7028412818908691, + "learning_rate": 8.215249529048632e-06, + "loss": 0.4724, + "step": 18086 + }, + { + "epoch": 0.5704778426115754, + "grad_norm": 1.8076056241989136, + "learning_rate": 8.214244383542748e-06, + "loss": 0.4753, + "step": 18087 + }, + { + "epoch": 0.570509383378016, + "grad_norm": 2.168644428253174, + "learning_rate": 8.213239256672612e-06, + "loss": 0.7397, + "step": 18088 + }, + { + "epoch": 0.5705409241444567, + "grad_norm": 1.3399953842163086, + "learning_rate": 8.212234148448708e-06, + "loss": 0.4688, + "step": 18089 + }, + { + "epoch": 0.5705724649108973, + "grad_norm": 2.030775547027588, + "learning_rate": 8.21122905888153e-06, + "loss": 0.4733, + "step": 18090 + }, + { + "epoch": 0.5706040056773379, + "grad_norm": 1.5969107151031494, + "learning_rate": 8.210223987981565e-06, + "loss": 0.4132, + "step": 18091 + }, + { + "epoch": 0.5706355464437786, + "grad_norm": 1.6557565927505493, + "learning_rate": 8.209218935759305e-06, + "loss": 0.4837, + "step": 18092 + }, + { + "epoch": 0.5706670872102192, + "grad_norm": 1.4493604898452759, + "learning_rate": 8.208213902225236e-06, + "loss": 0.3999, + "step": 18093 + }, + { + "epoch": 0.5706986279766598, + "grad_norm": 2.5313541889190674, + "learning_rate": 8.207208887389844e-06, + "loss": 0.4729, + "step": 18094 + }, + { + "epoch": 0.5707301687431005, + "grad_norm": 1.9874557256698608, + "learning_rate": 8.20620389126362e-06, + "loss": 0.4214, + "step": 18095 + }, + { + "epoch": 0.5707617095095411, + "grad_norm": 1.1467450857162476, + "learning_rate": 8.205198913857054e-06, + "loss": 0.3558, + "step": 18096 + }, + { + "epoch": 0.5707932502759817, + "grad_norm": 3.6978893280029297, + "learning_rate": 8.204193955180626e-06, + "loss": 0.4131, + "step": 18097 + }, + { + "epoch": 0.5708247910424223, + "grad_norm": 1.4344514608383179, + "learning_rate": 8.203189015244832e-06, + "loss": 0.4472, + "step": 18098 + }, + { + "epoch": 0.570856331808863, + "grad_norm": 2.09421443939209, + "learning_rate": 8.202184094060156e-06, + "loss": 0.7755, + "step": 18099 + }, + { + "epoch": 0.5708878725753036, + "grad_norm": 1.4400064945220947, + "learning_rate": 8.201179191637082e-06, + "loss": 0.4658, + "step": 18100 + }, + { + "epoch": 0.5709194133417442, + "grad_norm": 1.6403604745864868, + "learning_rate": 8.200174307986103e-06, + "loss": 0.4707, + "step": 18101 + }, + { + "epoch": 0.5709509541081849, + "grad_norm": 1.5013128519058228, + "learning_rate": 8.199169443117703e-06, + "loss": 0.4705, + "step": 18102 + }, + { + "epoch": 0.5709824948746255, + "grad_norm": 1.7759943008422852, + "learning_rate": 8.198164597042363e-06, + "loss": 0.4846, + "step": 18103 + }, + { + "epoch": 0.5710140356410661, + "grad_norm": 2.908323049545288, + "learning_rate": 8.197159769770579e-06, + "loss": 0.405, + "step": 18104 + }, + { + "epoch": 0.5710455764075067, + "grad_norm": 1.8010119199752808, + "learning_rate": 8.19615496131283e-06, + "loss": 0.489, + "step": 18105 + }, + { + "epoch": 0.5710771171739474, + "grad_norm": 2.622730255126953, + "learning_rate": 8.195150171679608e-06, + "loss": 0.4419, + "step": 18106 + }, + { + "epoch": 0.5711086579403879, + "grad_norm": 1.6042286157608032, + "learning_rate": 8.194145400881395e-06, + "loss": 0.475, + "step": 18107 + }, + { + "epoch": 0.5711401987068285, + "grad_norm": 1.9972941875457764, + "learning_rate": 8.193140648928672e-06, + "loss": 0.4362, + "step": 18108 + }, + { + "epoch": 0.5711717394732692, + "grad_norm": 3.038602828979492, + "learning_rate": 8.192135915831934e-06, + "loss": 0.4598, + "step": 18109 + }, + { + "epoch": 0.5712032802397098, + "grad_norm": 1.645892858505249, + "learning_rate": 8.191131201601663e-06, + "loss": 0.4716, + "step": 18110 + }, + { + "epoch": 0.5712348210061504, + "grad_norm": 1.7747244834899902, + "learning_rate": 8.190126506248337e-06, + "loss": 0.4345, + "step": 18111 + }, + { + "epoch": 0.571266361772591, + "grad_norm": 6.4043192863464355, + "learning_rate": 8.18912182978245e-06, + "loss": 0.4757, + "step": 18112 + }, + { + "epoch": 0.5712979025390317, + "grad_norm": 1.6714484691619873, + "learning_rate": 8.188117172214484e-06, + "loss": 0.4634, + "step": 18113 + }, + { + "epoch": 0.5713294433054723, + "grad_norm": 1.4979690313339233, + "learning_rate": 8.187112533554918e-06, + "loss": 0.446, + "step": 18114 + }, + { + "epoch": 0.5713609840719129, + "grad_norm": 1.9359424114227295, + "learning_rate": 8.186107913814241e-06, + "loss": 0.4802, + "step": 18115 + }, + { + "epoch": 0.5713925248383536, + "grad_norm": 1.526951789855957, + "learning_rate": 8.18510331300294e-06, + "loss": 0.4668, + "step": 18116 + }, + { + "epoch": 0.5714240656047942, + "grad_norm": 2.0012359619140625, + "learning_rate": 8.184098731131492e-06, + "loss": 0.4415, + "step": 18117 + }, + { + "epoch": 0.5714556063712348, + "grad_norm": 1.5820987224578857, + "learning_rate": 8.183094168210387e-06, + "loss": 0.4527, + "step": 18118 + }, + { + "epoch": 0.5714871471376755, + "grad_norm": 3.2934916019439697, + "learning_rate": 8.182089624250105e-06, + "loss": 0.4187, + "step": 18119 + }, + { + "epoch": 0.5715186879041161, + "grad_norm": 1.7174631357192993, + "learning_rate": 8.181085099261124e-06, + "loss": 0.4795, + "step": 18120 + }, + { + "epoch": 0.5715502286705567, + "grad_norm": 1.5468125343322754, + "learning_rate": 8.180080593253936e-06, + "loss": 0.4564, + "step": 18121 + }, + { + "epoch": 0.5715817694369973, + "grad_norm": 1.7293649911880493, + "learning_rate": 8.179076106239016e-06, + "loss": 0.5045, + "step": 18122 + }, + { + "epoch": 0.571613310203438, + "grad_norm": 1.6475557088851929, + "learning_rate": 8.178071638226857e-06, + "loss": 0.4167, + "step": 18123 + }, + { + "epoch": 0.5716448509698786, + "grad_norm": 1.4649215936660767, + "learning_rate": 8.177067189227934e-06, + "loss": 0.4636, + "step": 18124 + }, + { + "epoch": 0.5716763917363192, + "grad_norm": 1.8441916704177856, + "learning_rate": 8.176062759252728e-06, + "loss": 0.4709, + "step": 18125 + }, + { + "epoch": 0.5717079325027599, + "grad_norm": 3.3277132511138916, + "learning_rate": 8.175058348311726e-06, + "loss": 0.4652, + "step": 18126 + }, + { + "epoch": 0.5717394732692005, + "grad_norm": 1.6295256614685059, + "learning_rate": 8.174053956415406e-06, + "loss": 0.4704, + "step": 18127 + }, + { + "epoch": 0.571771014035641, + "grad_norm": 2.513779640197754, + "learning_rate": 8.173049583574247e-06, + "loss": 0.7522, + "step": 18128 + }, + { + "epoch": 0.5718025548020816, + "grad_norm": 1.434396505355835, + "learning_rate": 8.17204522979874e-06, + "loss": 0.4374, + "step": 18129 + }, + { + "epoch": 0.5718340955685223, + "grad_norm": 1.3705061674118042, + "learning_rate": 8.171040895099358e-06, + "loss": 0.4379, + "step": 18130 + }, + { + "epoch": 0.5718656363349629, + "grad_norm": 3.5424106121063232, + "learning_rate": 8.170036579486583e-06, + "loss": 0.4427, + "step": 18131 + }, + { + "epoch": 0.5718971771014035, + "grad_norm": 2.1033053398132324, + "learning_rate": 8.169032282970899e-06, + "loss": 0.4388, + "step": 18132 + }, + { + "epoch": 0.5719287178678442, + "grad_norm": 1.6750481128692627, + "learning_rate": 8.168028005562784e-06, + "loss": 0.4593, + "step": 18133 + }, + { + "epoch": 0.5719602586342848, + "grad_norm": 1.3214432001113892, + "learning_rate": 8.16702374727272e-06, + "loss": 0.459, + "step": 18134 + }, + { + "epoch": 0.5719917994007254, + "grad_norm": 6.172637462615967, + "learning_rate": 8.166019508111185e-06, + "loss": 0.4589, + "step": 18135 + }, + { + "epoch": 0.572023340167166, + "grad_norm": 1.6107441186904907, + "learning_rate": 8.165015288088663e-06, + "loss": 0.3943, + "step": 18136 + }, + { + "epoch": 0.5720548809336067, + "grad_norm": 1.5002022981643677, + "learning_rate": 8.164011087215629e-06, + "loss": 0.433, + "step": 18137 + }, + { + "epoch": 0.5720864217000473, + "grad_norm": 3.118701457977295, + "learning_rate": 8.163006905502567e-06, + "loss": 0.3974, + "step": 18138 + }, + { + "epoch": 0.5721179624664879, + "grad_norm": 1.5187830924987793, + "learning_rate": 8.16200274295995e-06, + "loss": 0.7618, + "step": 18139 + }, + { + "epoch": 0.5721495032329286, + "grad_norm": 1.4244304895401, + "learning_rate": 8.160998599598267e-06, + "loss": 0.4397, + "step": 18140 + }, + { + "epoch": 0.5721810439993692, + "grad_norm": 1.8050588369369507, + "learning_rate": 8.15999447542799e-06, + "loss": 0.4584, + "step": 18141 + }, + { + "epoch": 0.5722125847658098, + "grad_norm": 1.7247865200042725, + "learning_rate": 8.158990370459596e-06, + "loss": 0.4149, + "step": 18142 + }, + { + "epoch": 0.5722441255322505, + "grad_norm": 1.8650047779083252, + "learning_rate": 8.15798628470357e-06, + "loss": 0.4997, + "step": 18143 + }, + { + "epoch": 0.5722756662986911, + "grad_norm": 1.4093523025512695, + "learning_rate": 8.156982218170389e-06, + "loss": 0.4344, + "step": 18144 + }, + { + "epoch": 0.5723072070651317, + "grad_norm": 1.4915615320205688, + "learning_rate": 8.155978170870526e-06, + "loss": 0.3862, + "step": 18145 + }, + { + "epoch": 0.5723387478315723, + "grad_norm": 1.6889557838439941, + "learning_rate": 8.154974142814466e-06, + "loss": 0.3879, + "step": 18146 + }, + { + "epoch": 0.572370288598013, + "grad_norm": 1.9952651262283325, + "learning_rate": 8.153970134012682e-06, + "loss": 0.4321, + "step": 18147 + }, + { + "epoch": 0.5724018293644536, + "grad_norm": 1.4864881038665771, + "learning_rate": 8.15296614447565e-06, + "loss": 0.433, + "step": 18148 + }, + { + "epoch": 0.5724333701308941, + "grad_norm": 1.7635613679885864, + "learning_rate": 8.151962174213854e-06, + "loss": 0.4317, + "step": 18149 + }, + { + "epoch": 0.5724649108973348, + "grad_norm": 2.297631025314331, + "learning_rate": 8.15095822323777e-06, + "loss": 0.4618, + "step": 18150 + }, + { + "epoch": 0.5724964516637754, + "grad_norm": 1.665594458580017, + "learning_rate": 8.149954291557865e-06, + "loss": 0.4445, + "step": 18151 + }, + { + "epoch": 0.572527992430216, + "grad_norm": 1.4679200649261475, + "learning_rate": 8.14895037918463e-06, + "loss": 0.7479, + "step": 18152 + }, + { + "epoch": 0.5725595331966566, + "grad_norm": 1.6573911905288696, + "learning_rate": 8.147946486128534e-06, + "loss": 0.4404, + "step": 18153 + }, + { + "epoch": 0.5725910739630973, + "grad_norm": 1.7556712627410889, + "learning_rate": 8.146942612400054e-06, + "loss": 0.4722, + "step": 18154 + }, + { + "epoch": 0.5726226147295379, + "grad_norm": 1.5384629964828491, + "learning_rate": 8.145938758009664e-06, + "loss": 0.4634, + "step": 18155 + }, + { + "epoch": 0.5726541554959785, + "grad_norm": 1.7007914781570435, + "learning_rate": 8.144934922967845e-06, + "loss": 0.4623, + "step": 18156 + }, + { + "epoch": 0.5726856962624192, + "grad_norm": 2.241462230682373, + "learning_rate": 8.143931107285073e-06, + "loss": 0.3824, + "step": 18157 + }, + { + "epoch": 0.5727172370288598, + "grad_norm": 2.052273750305176, + "learning_rate": 8.14292731097182e-06, + "loss": 0.4595, + "step": 18158 + }, + { + "epoch": 0.5727487777953004, + "grad_norm": 2.539588212966919, + "learning_rate": 8.141923534038558e-06, + "loss": 0.4685, + "step": 18159 + }, + { + "epoch": 0.572780318561741, + "grad_norm": 2.275721311569214, + "learning_rate": 8.140919776495771e-06, + "loss": 0.4348, + "step": 18160 + }, + { + "epoch": 0.5728118593281817, + "grad_norm": 1.7954280376434326, + "learning_rate": 8.139916038353928e-06, + "loss": 0.4822, + "step": 18161 + }, + { + "epoch": 0.5728434000946223, + "grad_norm": 3.448276996612549, + "learning_rate": 8.138912319623504e-06, + "loss": 0.4543, + "step": 18162 + }, + { + "epoch": 0.5728749408610629, + "grad_norm": 1.4653149843215942, + "learning_rate": 8.137908620314978e-06, + "loss": 0.4666, + "step": 18163 + }, + { + "epoch": 0.5729064816275036, + "grad_norm": 1.5637239217758179, + "learning_rate": 8.13690494043882e-06, + "loss": 0.4702, + "step": 18164 + }, + { + "epoch": 0.5729380223939442, + "grad_norm": 1.2981725931167603, + "learning_rate": 8.135901280005506e-06, + "loss": 0.4412, + "step": 18165 + }, + { + "epoch": 0.5729695631603848, + "grad_norm": 1.6744636297225952, + "learning_rate": 8.134897639025512e-06, + "loss": 0.4208, + "step": 18166 + }, + { + "epoch": 0.5730011039268255, + "grad_norm": 1.676821231842041, + "learning_rate": 8.133894017509308e-06, + "loss": 0.7791, + "step": 18167 + }, + { + "epoch": 0.5730326446932661, + "grad_norm": 1.9921352863311768, + "learning_rate": 8.132890415467364e-06, + "loss": 0.7357, + "step": 18168 + }, + { + "epoch": 0.5730641854597067, + "grad_norm": 1.58596932888031, + "learning_rate": 8.131886832910165e-06, + "loss": 0.4561, + "step": 18169 + }, + { + "epoch": 0.5730957262261474, + "grad_norm": 1.7538138628005981, + "learning_rate": 8.13088326984817e-06, + "loss": 0.4506, + "step": 18170 + }, + { + "epoch": 0.5731272669925879, + "grad_norm": 2.0909533500671387, + "learning_rate": 8.129879726291866e-06, + "loss": 0.4651, + "step": 18171 + }, + { + "epoch": 0.5731588077590285, + "grad_norm": 1.55563223361969, + "learning_rate": 8.128876202251719e-06, + "loss": 0.4786, + "step": 18172 + }, + { + "epoch": 0.5731903485254691, + "grad_norm": 1.8486628532409668, + "learning_rate": 8.127872697738199e-06, + "loss": 0.4839, + "step": 18173 + }, + { + "epoch": 0.5732218892919098, + "grad_norm": 5.292109966278076, + "learning_rate": 8.126869212761783e-06, + "loss": 0.4646, + "step": 18174 + }, + { + "epoch": 0.5732534300583504, + "grad_norm": 1.819920539855957, + "learning_rate": 8.12586574733294e-06, + "loss": 0.4844, + "step": 18175 + }, + { + "epoch": 0.573284970824791, + "grad_norm": 5.644847869873047, + "learning_rate": 8.124862301462141e-06, + "loss": 0.4714, + "step": 18176 + }, + { + "epoch": 0.5733165115912316, + "grad_norm": 1.2686219215393066, + "learning_rate": 8.123858875159864e-06, + "loss": 0.4432, + "step": 18177 + }, + { + "epoch": 0.5733480523576723, + "grad_norm": 1.5198864936828613, + "learning_rate": 8.122855468436577e-06, + "loss": 0.7526, + "step": 18178 + }, + { + "epoch": 0.5733795931241129, + "grad_norm": 1.2732287645339966, + "learning_rate": 8.121852081302745e-06, + "loss": 0.3348, + "step": 18179 + }, + { + "epoch": 0.5734111338905535, + "grad_norm": 1.6686443090438843, + "learning_rate": 8.12084871376885e-06, + "loss": 0.749, + "step": 18180 + }, + { + "epoch": 0.5734426746569942, + "grad_norm": 1.4282578229904175, + "learning_rate": 8.119845365845357e-06, + "loss": 0.7547, + "step": 18181 + }, + { + "epoch": 0.5734742154234348, + "grad_norm": 1.5213593244552612, + "learning_rate": 8.118842037542734e-06, + "loss": 0.4361, + "step": 18182 + }, + { + "epoch": 0.5735057561898754, + "grad_norm": 1.8339933156967163, + "learning_rate": 8.11783872887146e-06, + "loss": 0.4567, + "step": 18183 + }, + { + "epoch": 0.5735372969563161, + "grad_norm": 1.4740941524505615, + "learning_rate": 8.116835439841994e-06, + "loss": 0.4337, + "step": 18184 + }, + { + "epoch": 0.5735688377227567, + "grad_norm": 2.504213809967041, + "learning_rate": 8.115832170464819e-06, + "loss": 0.3848, + "step": 18185 + }, + { + "epoch": 0.5736003784891973, + "grad_norm": 1.3779053688049316, + "learning_rate": 8.114828920750398e-06, + "loss": 0.4556, + "step": 18186 + }, + { + "epoch": 0.573631919255638, + "grad_norm": 1.660615086555481, + "learning_rate": 8.113825690709194e-06, + "loss": 0.4368, + "step": 18187 + }, + { + "epoch": 0.5736634600220786, + "grad_norm": 2.5021800994873047, + "learning_rate": 8.11282248035169e-06, + "loss": 0.4348, + "step": 18188 + }, + { + "epoch": 0.5736950007885192, + "grad_norm": 1.5836447477340698, + "learning_rate": 8.111819289688349e-06, + "loss": 0.461, + "step": 18189 + }, + { + "epoch": 0.5737265415549598, + "grad_norm": 1.5027574300765991, + "learning_rate": 8.110816118729634e-06, + "loss": 0.4137, + "step": 18190 + }, + { + "epoch": 0.5737580823214005, + "grad_norm": 1.5193067789077759, + "learning_rate": 8.109812967486024e-06, + "loss": 0.4655, + "step": 18191 + }, + { + "epoch": 0.573789623087841, + "grad_norm": 1.7784111499786377, + "learning_rate": 8.108809835967985e-06, + "loss": 0.4603, + "step": 18192 + }, + { + "epoch": 0.5738211638542816, + "grad_norm": 2.7150442600250244, + "learning_rate": 8.10780672418598e-06, + "loss": 0.4574, + "step": 18193 + }, + { + "epoch": 0.5738527046207222, + "grad_norm": 1.8398085832595825, + "learning_rate": 8.106803632150485e-06, + "loss": 0.4727, + "step": 18194 + }, + { + "epoch": 0.5738842453871629, + "grad_norm": 1.7255502939224243, + "learning_rate": 8.105800559871964e-06, + "loss": 0.4396, + "step": 18195 + }, + { + "epoch": 0.5739157861536035, + "grad_norm": 2.4388134479522705, + "learning_rate": 8.10479750736088e-06, + "loss": 0.465, + "step": 18196 + }, + { + "epoch": 0.5739473269200441, + "grad_norm": 1.4331254959106445, + "learning_rate": 8.10379447462771e-06, + "loss": 0.35, + "step": 18197 + }, + { + "epoch": 0.5739788676864848, + "grad_norm": 1.974128007888794, + "learning_rate": 8.102791461682916e-06, + "loss": 0.4654, + "step": 18198 + }, + { + "epoch": 0.5740104084529254, + "grad_norm": 1.7442641258239746, + "learning_rate": 8.101788468536962e-06, + "loss": 0.4525, + "step": 18199 + }, + { + "epoch": 0.574041949219366, + "grad_norm": 1.7712090015411377, + "learning_rate": 8.100785495200326e-06, + "loss": 0.4732, + "step": 18200 + }, + { + "epoch": 0.5740734899858067, + "grad_norm": 1.7490006685256958, + "learning_rate": 8.09978254168346e-06, + "loss": 0.4198, + "step": 18201 + }, + { + "epoch": 0.5741050307522473, + "grad_norm": 1.3989990949630737, + "learning_rate": 8.098779607996847e-06, + "loss": 0.4505, + "step": 18202 + }, + { + "epoch": 0.5741365715186879, + "grad_norm": 2.615966558456421, + "learning_rate": 8.097776694150942e-06, + "loss": 0.7709, + "step": 18203 + }, + { + "epoch": 0.5741681122851285, + "grad_norm": 1.9318530559539795, + "learning_rate": 8.096773800156214e-06, + "loss": 0.7637, + "step": 18204 + }, + { + "epoch": 0.5741996530515692, + "grad_norm": 1.6633573770523071, + "learning_rate": 8.09577092602313e-06, + "loss": 0.3651, + "step": 18205 + }, + { + "epoch": 0.5742311938180098, + "grad_norm": 1.9385418891906738, + "learning_rate": 8.094768071762156e-06, + "loss": 0.4797, + "step": 18206 + }, + { + "epoch": 0.5742627345844504, + "grad_norm": 1.5076589584350586, + "learning_rate": 8.093765237383751e-06, + "loss": 0.4489, + "step": 18207 + }, + { + "epoch": 0.5742942753508911, + "grad_norm": 1.4527277946472168, + "learning_rate": 8.09276242289839e-06, + "loss": 0.4382, + "step": 18208 + }, + { + "epoch": 0.5743258161173317, + "grad_norm": 1.784065842628479, + "learning_rate": 8.091759628316535e-06, + "loss": 0.4617, + "step": 18209 + }, + { + "epoch": 0.5743573568837723, + "grad_norm": 6.5955610275268555, + "learning_rate": 8.090756853648647e-06, + "loss": 0.7824, + "step": 18210 + }, + { + "epoch": 0.574388897650213, + "grad_norm": 1.6235696077346802, + "learning_rate": 8.089754098905195e-06, + "loss": 0.472, + "step": 18211 + }, + { + "epoch": 0.5744204384166536, + "grad_norm": 1.556397557258606, + "learning_rate": 8.088751364096645e-06, + "loss": 0.4558, + "step": 18212 + }, + { + "epoch": 0.5744519791830941, + "grad_norm": 1.4758636951446533, + "learning_rate": 8.087748649233456e-06, + "loss": 0.4902, + "step": 18213 + }, + { + "epoch": 0.5744835199495347, + "grad_norm": 3.384568691253662, + "learning_rate": 8.086745954326095e-06, + "loss": 0.4527, + "step": 18214 + }, + { + "epoch": 0.5745150607159754, + "grad_norm": 1.3998440504074097, + "learning_rate": 8.085743279385028e-06, + "loss": 0.4389, + "step": 18215 + }, + { + "epoch": 0.574546601482416, + "grad_norm": 4.983570575714111, + "learning_rate": 8.08474062442071e-06, + "loss": 0.4158, + "step": 18216 + }, + { + "epoch": 0.5745781422488566, + "grad_norm": 1.4257078170776367, + "learning_rate": 8.083737989443616e-06, + "loss": 0.4597, + "step": 18217 + }, + { + "epoch": 0.5746096830152972, + "grad_norm": 2.1444454193115234, + "learning_rate": 8.0827353744642e-06, + "loss": 0.3667, + "step": 18218 + }, + { + "epoch": 0.5746412237817379, + "grad_norm": 1.4998130798339844, + "learning_rate": 8.081732779492932e-06, + "loss": 0.4578, + "step": 18219 + }, + { + "epoch": 0.5746727645481785, + "grad_norm": 1.3937686681747437, + "learning_rate": 8.080730204540273e-06, + "loss": 0.4401, + "step": 18220 + }, + { + "epoch": 0.5747043053146191, + "grad_norm": 2.049882650375366, + "learning_rate": 8.07972764961668e-06, + "loss": 0.4207, + "step": 18221 + }, + { + "epoch": 0.5747358460810598, + "grad_norm": 1.539706826210022, + "learning_rate": 8.078725114732625e-06, + "loss": 0.4566, + "step": 18222 + }, + { + "epoch": 0.5747673868475004, + "grad_norm": 1.7265827655792236, + "learning_rate": 8.077722599898565e-06, + "loss": 0.4463, + "step": 18223 + }, + { + "epoch": 0.574798927613941, + "grad_norm": 1.607047200202942, + "learning_rate": 8.07672010512496e-06, + "loss": 0.4148, + "step": 18224 + }, + { + "epoch": 0.5748304683803817, + "grad_norm": 1.6179476976394653, + "learning_rate": 8.075717630422275e-06, + "loss": 0.4759, + "step": 18225 + }, + { + "epoch": 0.5748620091468223, + "grad_norm": 1.4697812795639038, + "learning_rate": 8.074715175800972e-06, + "loss": 0.4395, + "step": 18226 + }, + { + "epoch": 0.5748935499132629, + "grad_norm": 2.0309109687805176, + "learning_rate": 8.073712741271507e-06, + "loss": 0.4938, + "step": 18227 + }, + { + "epoch": 0.5749250906797035, + "grad_norm": 1.222822904586792, + "learning_rate": 8.072710326844348e-06, + "loss": 0.4265, + "step": 18228 + }, + { + "epoch": 0.5749566314461442, + "grad_norm": 1.4330734014511108, + "learning_rate": 8.071707932529956e-06, + "loss": 0.4679, + "step": 18229 + }, + { + "epoch": 0.5749881722125848, + "grad_norm": 1.5662645101547241, + "learning_rate": 8.070705558338783e-06, + "loss": 0.4707, + "step": 18230 + }, + { + "epoch": 0.5750197129790254, + "grad_norm": 1.6335967779159546, + "learning_rate": 8.069703204281299e-06, + "loss": 0.4396, + "step": 18231 + }, + { + "epoch": 0.5750512537454661, + "grad_norm": 1.5700275897979736, + "learning_rate": 8.06870087036796e-06, + "loss": 0.4816, + "step": 18232 + }, + { + "epoch": 0.5750827945119067, + "grad_norm": 3.1643142700195312, + "learning_rate": 8.06769855660923e-06, + "loss": 0.4774, + "step": 18233 + }, + { + "epoch": 0.5751143352783472, + "grad_norm": 1.8270219564437866, + "learning_rate": 8.066696263015562e-06, + "loss": 0.7627, + "step": 18234 + }, + { + "epoch": 0.5751458760447878, + "grad_norm": 1.6285074949264526, + "learning_rate": 8.065693989597419e-06, + "loss": 0.4504, + "step": 18235 + }, + { + "epoch": 0.5751774168112285, + "grad_norm": 1.8303229808807373, + "learning_rate": 8.064691736365265e-06, + "loss": 0.4211, + "step": 18236 + }, + { + "epoch": 0.5752089575776691, + "grad_norm": 1.7642199993133545, + "learning_rate": 8.063689503329554e-06, + "loss": 0.4581, + "step": 18237 + }, + { + "epoch": 0.5752404983441097, + "grad_norm": 1.6571682691574097, + "learning_rate": 8.062687290500745e-06, + "loss": 0.4627, + "step": 18238 + }, + { + "epoch": 0.5752720391105504, + "grad_norm": 1.394799828529358, + "learning_rate": 8.0616850978893e-06, + "loss": 0.458, + "step": 18239 + }, + { + "epoch": 0.575303579876991, + "grad_norm": 1.4078452587127686, + "learning_rate": 8.060682925505679e-06, + "loss": 0.4366, + "step": 18240 + }, + { + "epoch": 0.5753351206434316, + "grad_norm": 3.893659830093384, + "learning_rate": 8.05968077336033e-06, + "loss": 0.7307, + "step": 18241 + }, + { + "epoch": 0.5753666614098722, + "grad_norm": 2.0573222637176514, + "learning_rate": 8.058678641463724e-06, + "loss": 0.4287, + "step": 18242 + }, + { + "epoch": 0.5753982021763129, + "grad_norm": 2.241450548171997, + "learning_rate": 8.057676529826315e-06, + "loss": 0.4573, + "step": 18243 + }, + { + "epoch": 0.5754297429427535, + "grad_norm": 1.5986162424087524, + "learning_rate": 8.056674438458556e-06, + "loss": 0.4598, + "step": 18244 + }, + { + "epoch": 0.5754612837091941, + "grad_norm": 1.3873831033706665, + "learning_rate": 8.055672367370912e-06, + "loss": 0.4493, + "step": 18245 + }, + { + "epoch": 0.5754928244756348, + "grad_norm": 1.569555640220642, + "learning_rate": 8.054670316573835e-06, + "loss": 0.3224, + "step": 18246 + }, + { + "epoch": 0.5755243652420754, + "grad_norm": 1.38259756565094, + "learning_rate": 8.05366828607778e-06, + "loss": 0.4753, + "step": 18247 + }, + { + "epoch": 0.575555906008516, + "grad_norm": 1.7226834297180176, + "learning_rate": 8.052666275893212e-06, + "loss": 0.4431, + "step": 18248 + }, + { + "epoch": 0.5755874467749567, + "grad_norm": 1.9880551099777222, + "learning_rate": 8.051664286030578e-06, + "loss": 0.4771, + "step": 18249 + }, + { + "epoch": 0.5756189875413973, + "grad_norm": 1.7861332893371582, + "learning_rate": 8.050662316500346e-06, + "loss": 0.4263, + "step": 18250 + }, + { + "epoch": 0.5756505283078379, + "grad_norm": 1.5782493352890015, + "learning_rate": 8.049660367312965e-06, + "loss": 0.4523, + "step": 18251 + }, + { + "epoch": 0.5756820690742785, + "grad_norm": 1.5934960842132568, + "learning_rate": 8.048658438478892e-06, + "loss": 0.4636, + "step": 18252 + }, + { + "epoch": 0.5757136098407192, + "grad_norm": 1.1724706888198853, + "learning_rate": 8.047656530008585e-06, + "loss": 0.348, + "step": 18253 + }, + { + "epoch": 0.5757451506071598, + "grad_norm": 1.5434571504592896, + "learning_rate": 8.046654641912497e-06, + "loss": 0.4388, + "step": 18254 + }, + { + "epoch": 0.5757766913736003, + "grad_norm": 1.6698267459869385, + "learning_rate": 8.045652774201082e-06, + "loss": 0.7506, + "step": 18255 + }, + { + "epoch": 0.575808232140041, + "grad_norm": 1.6026784181594849, + "learning_rate": 8.0446509268848e-06, + "loss": 0.4441, + "step": 18256 + }, + { + "epoch": 0.5758397729064816, + "grad_norm": 1.3803759813308716, + "learning_rate": 8.043649099974106e-06, + "loss": 0.4527, + "step": 18257 + }, + { + "epoch": 0.5758713136729222, + "grad_norm": 1.5453104972839355, + "learning_rate": 8.042647293479447e-06, + "loss": 0.4533, + "step": 18258 + }, + { + "epoch": 0.5759028544393628, + "grad_norm": 1.8526173830032349, + "learning_rate": 8.041645507411288e-06, + "loss": 0.3887, + "step": 18259 + }, + { + "epoch": 0.5759343952058035, + "grad_norm": 1.4677025079727173, + "learning_rate": 8.040643741780079e-06, + "loss": 0.4694, + "step": 18260 + }, + { + "epoch": 0.5759659359722441, + "grad_norm": 1.5214115381240845, + "learning_rate": 8.03964199659627e-06, + "loss": 0.453, + "step": 18261 + }, + { + "epoch": 0.5759974767386847, + "grad_norm": 1.9245301485061646, + "learning_rate": 8.038640271870321e-06, + "loss": 0.4466, + "step": 18262 + }, + { + "epoch": 0.5760290175051254, + "grad_norm": 1.9880099296569824, + "learning_rate": 8.037638567612687e-06, + "loss": 0.4591, + "step": 18263 + }, + { + "epoch": 0.576060558271566, + "grad_norm": 1.7610430717468262, + "learning_rate": 8.036636883833814e-06, + "loss": 0.4802, + "step": 18264 + }, + { + "epoch": 0.5760920990380066, + "grad_norm": 1.52020263671875, + "learning_rate": 8.035635220544163e-06, + "loss": 0.4306, + "step": 18265 + }, + { + "epoch": 0.5761236398044473, + "grad_norm": 1.5366019010543823, + "learning_rate": 8.034633577754178e-06, + "loss": 0.4557, + "step": 18266 + }, + { + "epoch": 0.5761551805708879, + "grad_norm": 1.5900014638900757, + "learning_rate": 8.033631955474322e-06, + "loss": 0.4707, + "step": 18267 + }, + { + "epoch": 0.5761867213373285, + "grad_norm": 3.8218109607696533, + "learning_rate": 8.032630353715043e-06, + "loss": 0.4625, + "step": 18268 + }, + { + "epoch": 0.5762182621037691, + "grad_norm": 2.1468687057495117, + "learning_rate": 8.031628772486793e-06, + "loss": 0.7218, + "step": 18269 + }, + { + "epoch": 0.5762498028702098, + "grad_norm": 1.736082911491394, + "learning_rate": 8.030627211800026e-06, + "loss": 0.4519, + "step": 18270 + }, + { + "epoch": 0.5762813436366504, + "grad_norm": 1.8031516075134277, + "learning_rate": 8.029625671665194e-06, + "loss": 0.7159, + "step": 18271 + }, + { + "epoch": 0.576312884403091, + "grad_norm": 1.6389203071594238, + "learning_rate": 8.028624152092747e-06, + "loss": 0.7745, + "step": 18272 + }, + { + "epoch": 0.5763444251695317, + "grad_norm": 1.3606102466583252, + "learning_rate": 8.027622653093139e-06, + "loss": 0.431, + "step": 18273 + }, + { + "epoch": 0.5763759659359723, + "grad_norm": 1.6411898136138916, + "learning_rate": 8.026621174676821e-06, + "loss": 0.4886, + "step": 18274 + }, + { + "epoch": 0.5764075067024129, + "grad_norm": 1.49778413772583, + "learning_rate": 8.025619716854238e-06, + "loss": 0.4817, + "step": 18275 + }, + { + "epoch": 0.5764390474688534, + "grad_norm": 1.6979405879974365, + "learning_rate": 8.02461827963585e-06, + "loss": 0.4532, + "step": 18276 + }, + { + "epoch": 0.5764705882352941, + "grad_norm": 4.637426376342773, + "learning_rate": 8.023616863032105e-06, + "loss": 0.4417, + "step": 18277 + }, + { + "epoch": 0.5765021290017347, + "grad_norm": 1.6082913875579834, + "learning_rate": 8.02261546705345e-06, + "loss": 0.4406, + "step": 18278 + }, + { + "epoch": 0.5765336697681753, + "grad_norm": 1.4455937147140503, + "learning_rate": 8.021614091710339e-06, + "loss": 0.4782, + "step": 18279 + }, + { + "epoch": 0.576565210534616, + "grad_norm": 1.6509814262390137, + "learning_rate": 8.020612737013222e-06, + "loss": 0.7818, + "step": 18280 + }, + { + "epoch": 0.5765967513010566, + "grad_norm": 2.087735891342163, + "learning_rate": 8.019611402972545e-06, + "loss": 0.4655, + "step": 18281 + }, + { + "epoch": 0.5766282920674972, + "grad_norm": 2.6608922481536865, + "learning_rate": 8.018610089598766e-06, + "loss": 0.7478, + "step": 18282 + }, + { + "epoch": 0.5766598328339378, + "grad_norm": 1.7863723039627075, + "learning_rate": 8.017608796902325e-06, + "loss": 0.4434, + "step": 18283 + }, + { + "epoch": 0.5766913736003785, + "grad_norm": 1.3570829629898071, + "learning_rate": 8.016607524893678e-06, + "loss": 0.774, + "step": 18284 + }, + { + "epoch": 0.5767229143668191, + "grad_norm": 3.049852132797241, + "learning_rate": 8.015606273583273e-06, + "loss": 0.7192, + "step": 18285 + }, + { + "epoch": 0.5767544551332597, + "grad_norm": 1.3201195001602173, + "learning_rate": 8.014605042981552e-06, + "loss": 0.3743, + "step": 18286 + }, + { + "epoch": 0.5767859958997004, + "grad_norm": 1.5598725080490112, + "learning_rate": 8.013603833098973e-06, + "loss": 0.4489, + "step": 18287 + }, + { + "epoch": 0.576817536666141, + "grad_norm": 1.7999043464660645, + "learning_rate": 8.01260264394598e-06, + "loss": 0.4512, + "step": 18288 + }, + { + "epoch": 0.5768490774325816, + "grad_norm": 1.4481244087219238, + "learning_rate": 8.011601475533018e-06, + "loss": 0.4369, + "step": 18289 + }, + { + "epoch": 0.5768806181990223, + "grad_norm": 1.3861465454101562, + "learning_rate": 8.010600327870544e-06, + "loss": 0.4045, + "step": 18290 + }, + { + "epoch": 0.5769121589654629, + "grad_norm": 1.5226531028747559, + "learning_rate": 8.009599200968998e-06, + "loss": 0.3391, + "step": 18291 + }, + { + "epoch": 0.5769436997319035, + "grad_norm": 2.014410972595215, + "learning_rate": 8.00859809483883e-06, + "loss": 0.4113, + "step": 18292 + }, + { + "epoch": 0.5769752404983441, + "grad_norm": 1.407069206237793, + "learning_rate": 8.007597009490487e-06, + "loss": 0.4764, + "step": 18293 + }, + { + "epoch": 0.5770067812647848, + "grad_norm": 1.70877206325531, + "learning_rate": 8.006595944934418e-06, + "loss": 0.7515, + "step": 18294 + }, + { + "epoch": 0.5770383220312254, + "grad_norm": 1.3175632953643799, + "learning_rate": 8.005594901181064e-06, + "loss": 0.4523, + "step": 18295 + }, + { + "epoch": 0.577069862797666, + "grad_norm": 1.7594224214553833, + "learning_rate": 8.004593878240879e-06, + "loss": 0.4567, + "step": 18296 + }, + { + "epoch": 0.5771014035641067, + "grad_norm": 2.0222585201263428, + "learning_rate": 8.003592876124301e-06, + "loss": 0.4598, + "step": 18297 + }, + { + "epoch": 0.5771329443305472, + "grad_norm": 2.223566770553589, + "learning_rate": 8.002591894841786e-06, + "loss": 0.4372, + "step": 18298 + }, + { + "epoch": 0.5771644850969878, + "grad_norm": 1.9944710731506348, + "learning_rate": 8.001590934403778e-06, + "loss": 0.4722, + "step": 18299 + }, + { + "epoch": 0.5771960258634284, + "grad_norm": 1.5839309692382812, + "learning_rate": 8.000589994820717e-06, + "loss": 0.4316, + "step": 18300 + }, + { + "epoch": 0.5772275666298691, + "grad_norm": 1.6794459819793701, + "learning_rate": 7.999589076103051e-06, + "loss": 0.5074, + "step": 18301 + }, + { + "epoch": 0.5772591073963097, + "grad_norm": 1.4628175497055054, + "learning_rate": 7.99858817826123e-06, + "loss": 0.4347, + "step": 18302 + }, + { + "epoch": 0.5772906481627503, + "grad_norm": 1.552506923675537, + "learning_rate": 7.997587301305692e-06, + "loss": 0.4511, + "step": 18303 + }, + { + "epoch": 0.577322188929191, + "grad_norm": 1.515596628189087, + "learning_rate": 7.996586445246888e-06, + "loss": 0.4686, + "step": 18304 + }, + { + "epoch": 0.5773537296956316, + "grad_norm": 1.8818626403808594, + "learning_rate": 7.99558561009526e-06, + "loss": 0.4477, + "step": 18305 + }, + { + "epoch": 0.5773852704620722, + "grad_norm": 1.3773013353347778, + "learning_rate": 7.994584795861248e-06, + "loss": 0.456, + "step": 18306 + }, + { + "epoch": 0.5774168112285128, + "grad_norm": 2.288444995880127, + "learning_rate": 7.993584002555306e-06, + "loss": 0.4334, + "step": 18307 + }, + { + "epoch": 0.5774483519949535, + "grad_norm": 1.3315784931182861, + "learning_rate": 7.992583230187872e-06, + "loss": 0.4662, + "step": 18308 + }, + { + "epoch": 0.5774798927613941, + "grad_norm": 1.4150478839874268, + "learning_rate": 7.991582478769385e-06, + "loss": 0.4293, + "step": 18309 + }, + { + "epoch": 0.5775114335278347, + "grad_norm": 2.8059823513031006, + "learning_rate": 7.990581748310303e-06, + "loss": 0.4744, + "step": 18310 + }, + { + "epoch": 0.5775429742942754, + "grad_norm": 1.3493233919143677, + "learning_rate": 7.989581038821057e-06, + "loss": 0.4405, + "step": 18311 + }, + { + "epoch": 0.577574515060716, + "grad_norm": 6.42461633682251, + "learning_rate": 7.988580350312092e-06, + "loss": 0.7378, + "step": 18312 + }, + { + "epoch": 0.5776060558271566, + "grad_norm": 1.7576662302017212, + "learning_rate": 7.987579682793856e-06, + "loss": 0.4411, + "step": 18313 + }, + { + "epoch": 0.5776375965935973, + "grad_norm": 1.7185661792755127, + "learning_rate": 7.986579036276783e-06, + "loss": 0.7288, + "step": 18314 + }, + { + "epoch": 0.5776691373600379, + "grad_norm": 1.5157901048660278, + "learning_rate": 7.985578410771328e-06, + "loss": 0.468, + "step": 18315 + }, + { + "epoch": 0.5777006781264785, + "grad_norm": 2.2631521224975586, + "learning_rate": 7.984577806287925e-06, + "loss": 0.4291, + "step": 18316 + }, + { + "epoch": 0.5777322188929191, + "grad_norm": 1.7441710233688354, + "learning_rate": 7.983577222837012e-06, + "loss": 0.4269, + "step": 18317 + }, + { + "epoch": 0.5777637596593598, + "grad_norm": 2.6227662563323975, + "learning_rate": 7.982576660429043e-06, + "loss": 0.3453, + "step": 18318 + }, + { + "epoch": 0.5777953004258003, + "grad_norm": 2.0228216648101807, + "learning_rate": 7.981576119074451e-06, + "loss": 0.462, + "step": 18319 + }, + { + "epoch": 0.5778268411922409, + "grad_norm": 2.190070629119873, + "learning_rate": 7.980575598783678e-06, + "loss": 0.739, + "step": 18320 + }, + { + "epoch": 0.5778583819586816, + "grad_norm": 1.4503765106201172, + "learning_rate": 7.979575099567167e-06, + "loss": 0.4566, + "step": 18321 + }, + { + "epoch": 0.5778899227251222, + "grad_norm": 1.713401436805725, + "learning_rate": 7.978574621435362e-06, + "loss": 0.4719, + "step": 18322 + }, + { + "epoch": 0.5779214634915628, + "grad_norm": 1.5126231908798218, + "learning_rate": 7.977574164398695e-06, + "loss": 0.4308, + "step": 18323 + }, + { + "epoch": 0.5779530042580034, + "grad_norm": 1.689953088760376, + "learning_rate": 7.976573728467617e-06, + "loss": 0.4733, + "step": 18324 + }, + { + "epoch": 0.5779845450244441, + "grad_norm": 1.701980710029602, + "learning_rate": 7.975573313652563e-06, + "loss": 0.4727, + "step": 18325 + }, + { + "epoch": 0.5780160857908847, + "grad_norm": 21.6508731842041, + "learning_rate": 7.974572919963968e-06, + "loss": 0.3563, + "step": 18326 + }, + { + "epoch": 0.5780476265573253, + "grad_norm": 1.606488823890686, + "learning_rate": 7.973572547412282e-06, + "loss": 0.4494, + "step": 18327 + }, + { + "epoch": 0.578079167323766, + "grad_norm": 3.3480639457702637, + "learning_rate": 7.972572196007936e-06, + "loss": 0.7334, + "step": 18328 + }, + { + "epoch": 0.5781107080902066, + "grad_norm": 1.3304470777511597, + "learning_rate": 7.971571865761377e-06, + "loss": 0.4563, + "step": 18329 + }, + { + "epoch": 0.5781422488566472, + "grad_norm": 1.959394931793213, + "learning_rate": 7.970571556683042e-06, + "loss": 0.4123, + "step": 18330 + }, + { + "epoch": 0.5781737896230879, + "grad_norm": 1.2214504480361938, + "learning_rate": 7.969571268783364e-06, + "loss": 0.4503, + "step": 18331 + }, + { + "epoch": 0.5782053303895285, + "grad_norm": 1.4601800441741943, + "learning_rate": 7.96857100207279e-06, + "loss": 0.3685, + "step": 18332 + }, + { + "epoch": 0.5782368711559691, + "grad_norm": 1.5176656246185303, + "learning_rate": 7.967570756561756e-06, + "loss": 0.4369, + "step": 18333 + }, + { + "epoch": 0.5782684119224097, + "grad_norm": 1.6247671842575073, + "learning_rate": 7.966570532260693e-06, + "loss": 0.4681, + "step": 18334 + }, + { + "epoch": 0.5782999526888504, + "grad_norm": 1.8203264474868774, + "learning_rate": 7.965570329180051e-06, + "loss": 0.4586, + "step": 18335 + }, + { + "epoch": 0.578331493455291, + "grad_norm": 1.5256948471069336, + "learning_rate": 7.964570147330261e-06, + "loss": 0.4287, + "step": 18336 + }, + { + "epoch": 0.5783630342217316, + "grad_norm": 1.6637675762176514, + "learning_rate": 7.96356998672176e-06, + "loss": 0.4689, + "step": 18337 + }, + { + "epoch": 0.5783945749881723, + "grad_norm": 1.4699560403823853, + "learning_rate": 7.96256984736499e-06, + "loss": 0.4506, + "step": 18338 + }, + { + "epoch": 0.5784261157546129, + "grad_norm": 2.9817452430725098, + "learning_rate": 7.961569729270387e-06, + "loss": 0.7641, + "step": 18339 + }, + { + "epoch": 0.5784576565210534, + "grad_norm": 1.6221582889556885, + "learning_rate": 7.960569632448383e-06, + "loss": 0.4683, + "step": 18340 + }, + { + "epoch": 0.578489197287494, + "grad_norm": 1.479374647140503, + "learning_rate": 7.95956955690942e-06, + "loss": 0.4383, + "step": 18341 + }, + { + "epoch": 0.5785207380539347, + "grad_norm": 1.5296891927719116, + "learning_rate": 7.958569502663932e-06, + "loss": 0.3597, + "step": 18342 + }, + { + "epoch": 0.5785522788203753, + "grad_norm": 1.6808056831359863, + "learning_rate": 7.957569469722358e-06, + "loss": 0.4317, + "step": 18343 + }, + { + "epoch": 0.5785838195868159, + "grad_norm": 1.8061223030090332, + "learning_rate": 7.956569458095133e-06, + "loss": 0.3411, + "step": 18344 + }, + { + "epoch": 0.5786153603532566, + "grad_norm": 1.647019863128662, + "learning_rate": 7.955569467792687e-06, + "loss": 0.4609, + "step": 18345 + }, + { + "epoch": 0.5786469011196972, + "grad_norm": 1.3963241577148438, + "learning_rate": 7.954569498825467e-06, + "loss": 0.43, + "step": 18346 + }, + { + "epoch": 0.5786784418861378, + "grad_norm": 1.9097822904586792, + "learning_rate": 7.953569551203902e-06, + "loss": 0.453, + "step": 18347 + }, + { + "epoch": 0.5787099826525784, + "grad_norm": 1.8204710483551025, + "learning_rate": 7.952569624938424e-06, + "loss": 0.4764, + "step": 18348 + }, + { + "epoch": 0.5787415234190191, + "grad_norm": 1.40382719039917, + "learning_rate": 7.951569720039476e-06, + "loss": 0.3924, + "step": 18349 + }, + { + "epoch": 0.5787730641854597, + "grad_norm": 1.4860286712646484, + "learning_rate": 7.950569836517487e-06, + "loss": 0.4665, + "step": 18350 + }, + { + "epoch": 0.5788046049519003, + "grad_norm": 1.5526865720748901, + "learning_rate": 7.949569974382893e-06, + "loss": 0.4687, + "step": 18351 + }, + { + "epoch": 0.578836145718341, + "grad_norm": 1.6168649196624756, + "learning_rate": 7.948570133646129e-06, + "loss": 0.4525, + "step": 18352 + }, + { + "epoch": 0.5788676864847816, + "grad_norm": 1.5268179178237915, + "learning_rate": 7.94757031431763e-06, + "loss": 0.4346, + "step": 18353 + }, + { + "epoch": 0.5788992272512222, + "grad_norm": 1.7856954336166382, + "learning_rate": 7.946570516407823e-06, + "loss": 0.4604, + "step": 18354 + }, + { + "epoch": 0.5789307680176629, + "grad_norm": 1.5637648105621338, + "learning_rate": 7.945570739927153e-06, + "loss": 0.4685, + "step": 18355 + }, + { + "epoch": 0.5789623087841035, + "grad_norm": 2.005058526992798, + "learning_rate": 7.944570984886045e-06, + "loss": 0.4884, + "step": 18356 + }, + { + "epoch": 0.5789938495505441, + "grad_norm": 2.2733755111694336, + "learning_rate": 7.943571251294931e-06, + "loss": 0.458, + "step": 18357 + }, + { + "epoch": 0.5790253903169847, + "grad_norm": 2.234921455383301, + "learning_rate": 7.942571539164253e-06, + "loss": 0.3475, + "step": 18358 + }, + { + "epoch": 0.5790569310834254, + "grad_norm": 3.8925552368164062, + "learning_rate": 7.94157184850444e-06, + "loss": 0.7582, + "step": 18359 + }, + { + "epoch": 0.579088471849866, + "grad_norm": 1.640520691871643, + "learning_rate": 7.940572179325918e-06, + "loss": 0.4603, + "step": 18360 + }, + { + "epoch": 0.5791200126163065, + "grad_norm": 1.808924913406372, + "learning_rate": 7.939572531639128e-06, + "loss": 0.4085, + "step": 18361 + }, + { + "epoch": 0.5791515533827472, + "grad_norm": 2.03533935546875, + "learning_rate": 7.938572905454494e-06, + "loss": 0.4648, + "step": 18362 + }, + { + "epoch": 0.5791830941491878, + "grad_norm": 1.767203450202942, + "learning_rate": 7.937573300782458e-06, + "loss": 0.4538, + "step": 18363 + }, + { + "epoch": 0.5792146349156284, + "grad_norm": 1.6295568943023682, + "learning_rate": 7.936573717633444e-06, + "loss": 0.4207, + "step": 18364 + }, + { + "epoch": 0.579246175682069, + "grad_norm": 1.734752893447876, + "learning_rate": 7.935574156017881e-06, + "loss": 0.7785, + "step": 18365 + }, + { + "epoch": 0.5792777164485097, + "grad_norm": 1.6493843793869019, + "learning_rate": 7.93457461594621e-06, + "loss": 0.43, + "step": 18366 + }, + { + "epoch": 0.5793092572149503, + "grad_norm": 2.1112008094787598, + "learning_rate": 7.933575097428857e-06, + "loss": 0.4109, + "step": 18367 + }, + { + "epoch": 0.5793407979813909, + "grad_norm": 1.570713758468628, + "learning_rate": 7.932575600476248e-06, + "loss": 0.4728, + "step": 18368 + }, + { + "epoch": 0.5793723387478316, + "grad_norm": 1.5396173000335693, + "learning_rate": 7.931576125098823e-06, + "loss": 0.4722, + "step": 18369 + }, + { + "epoch": 0.5794038795142722, + "grad_norm": 1.7429909706115723, + "learning_rate": 7.930576671307005e-06, + "loss": 0.7681, + "step": 18370 + }, + { + "epoch": 0.5794354202807128, + "grad_norm": 1.6175909042358398, + "learning_rate": 7.929577239111226e-06, + "loss": 0.4987, + "step": 18371 + }, + { + "epoch": 0.5794669610471535, + "grad_norm": 1.3619526624679565, + "learning_rate": 7.928577828521917e-06, + "loss": 0.453, + "step": 18372 + }, + { + "epoch": 0.5794985018135941, + "grad_norm": 1.6572123765945435, + "learning_rate": 7.927578439549506e-06, + "loss": 0.4195, + "step": 18373 + }, + { + "epoch": 0.5795300425800347, + "grad_norm": 1.907962441444397, + "learning_rate": 7.926579072204421e-06, + "loss": 0.4533, + "step": 18374 + }, + { + "epoch": 0.5795615833464753, + "grad_norm": 1.571944236755371, + "learning_rate": 7.925579726497096e-06, + "loss": 0.46, + "step": 18375 + }, + { + "epoch": 0.579593124112916, + "grad_norm": 1.0517241954803467, + "learning_rate": 7.924580402437956e-06, + "loss": 0.3801, + "step": 18376 + }, + { + "epoch": 0.5796246648793566, + "grad_norm": 1.4281281232833862, + "learning_rate": 7.923581100037432e-06, + "loss": 0.4756, + "step": 18377 + }, + { + "epoch": 0.5796562056457972, + "grad_norm": 1.6431317329406738, + "learning_rate": 7.922581819305953e-06, + "loss": 0.4425, + "step": 18378 + }, + { + "epoch": 0.5796877464122379, + "grad_norm": 2.0866200923919678, + "learning_rate": 7.921582560253944e-06, + "loss": 0.4524, + "step": 18379 + }, + { + "epoch": 0.5797192871786785, + "grad_norm": 1.4948153495788574, + "learning_rate": 7.920583322891836e-06, + "loss": 0.4857, + "step": 18380 + }, + { + "epoch": 0.5797508279451191, + "grad_norm": 1.5706197023391724, + "learning_rate": 7.919584107230055e-06, + "loss": 0.4444, + "step": 18381 + }, + { + "epoch": 0.5797823687115596, + "grad_norm": 1.5823750495910645, + "learning_rate": 7.918584913279027e-06, + "loss": 0.4362, + "step": 18382 + }, + { + "epoch": 0.5798139094780003, + "grad_norm": 1.3351757526397705, + "learning_rate": 7.917585741049186e-06, + "loss": 0.4598, + "step": 18383 + }, + { + "epoch": 0.5798454502444409, + "grad_norm": 1.217694640159607, + "learning_rate": 7.916586590550952e-06, + "loss": 0.3443, + "step": 18384 + }, + { + "epoch": 0.5798769910108815, + "grad_norm": 1.9387691020965576, + "learning_rate": 7.915587461794752e-06, + "loss": 0.4172, + "step": 18385 + }, + { + "epoch": 0.5799085317773222, + "grad_norm": 1.720399022102356, + "learning_rate": 7.91458835479102e-06, + "loss": 0.4747, + "step": 18386 + }, + { + "epoch": 0.5799400725437628, + "grad_norm": 1.4832066297531128, + "learning_rate": 7.913589269550177e-06, + "loss": 0.4336, + "step": 18387 + }, + { + "epoch": 0.5799716133102034, + "grad_norm": 1.3193022012710571, + "learning_rate": 7.912590206082647e-06, + "loss": 0.4482, + "step": 18388 + }, + { + "epoch": 0.580003154076644, + "grad_norm": 1.4951188564300537, + "learning_rate": 7.911591164398862e-06, + "loss": 0.4354, + "step": 18389 + }, + { + "epoch": 0.5800346948430847, + "grad_norm": 1.5069929361343384, + "learning_rate": 7.910592144509245e-06, + "loss": 0.462, + "step": 18390 + }, + { + "epoch": 0.5800662356095253, + "grad_norm": 1.874536156654358, + "learning_rate": 7.90959314642422e-06, + "loss": 0.4554, + "step": 18391 + }, + { + "epoch": 0.5800977763759659, + "grad_norm": 1.9749659299850464, + "learning_rate": 7.908594170154214e-06, + "loss": 0.4405, + "step": 18392 + }, + { + "epoch": 0.5801293171424066, + "grad_norm": 1.3931740522384644, + "learning_rate": 7.907595215709649e-06, + "loss": 0.7437, + "step": 18393 + }, + { + "epoch": 0.5801608579088472, + "grad_norm": 1.7668414115905762, + "learning_rate": 7.906596283100958e-06, + "loss": 0.4372, + "step": 18394 + }, + { + "epoch": 0.5801923986752878, + "grad_norm": 1.6162223815917969, + "learning_rate": 7.905597372338558e-06, + "loss": 0.4336, + "step": 18395 + }, + { + "epoch": 0.5802239394417285, + "grad_norm": 1.544442057609558, + "learning_rate": 7.904598483432872e-06, + "loss": 0.4428, + "step": 18396 + }, + { + "epoch": 0.5802554802081691, + "grad_norm": 1.440366268157959, + "learning_rate": 7.903599616394333e-06, + "loss": 0.7422, + "step": 18397 + }, + { + "epoch": 0.5802870209746097, + "grad_norm": 1.418308138847351, + "learning_rate": 7.902600771233359e-06, + "loss": 0.4127, + "step": 18398 + }, + { + "epoch": 0.5803185617410503, + "grad_norm": 8.527921676635742, + "learning_rate": 7.901601947960374e-06, + "loss": 0.3554, + "step": 18399 + }, + { + "epoch": 0.580350102507491, + "grad_norm": 1.761139988899231, + "learning_rate": 7.900603146585802e-06, + "loss": 0.4452, + "step": 18400 + }, + { + "epoch": 0.5803816432739316, + "grad_norm": 1.973289132118225, + "learning_rate": 7.899604367120067e-06, + "loss": 0.4682, + "step": 18401 + }, + { + "epoch": 0.5804131840403722, + "grad_norm": 1.4861398935317993, + "learning_rate": 7.898605609573587e-06, + "loss": 0.4547, + "step": 18402 + }, + { + "epoch": 0.5804447248068128, + "grad_norm": 1.8353370428085327, + "learning_rate": 7.897606873956794e-06, + "loss": 0.3914, + "step": 18403 + }, + { + "epoch": 0.5804762655732534, + "grad_norm": 1.6587313413619995, + "learning_rate": 7.896608160280106e-06, + "loss": 0.4584, + "step": 18404 + }, + { + "epoch": 0.580507806339694, + "grad_norm": 2.1851024627685547, + "learning_rate": 7.89560946855394e-06, + "loss": 0.4524, + "step": 18405 + }, + { + "epoch": 0.5805393471061346, + "grad_norm": 1.7962020635604858, + "learning_rate": 7.894610798788728e-06, + "loss": 0.4552, + "step": 18406 + }, + { + "epoch": 0.5805708878725753, + "grad_norm": 1.579795241355896, + "learning_rate": 7.893612150994883e-06, + "loss": 0.4654, + "step": 18407 + }, + { + "epoch": 0.5806024286390159, + "grad_norm": 1.3690185546875, + "learning_rate": 7.892613525182833e-06, + "loss": 0.7492, + "step": 18408 + }, + { + "epoch": 0.5806339694054565, + "grad_norm": 1.754274606704712, + "learning_rate": 7.891614921363e-06, + "loss": 0.4893, + "step": 18409 + }, + { + "epoch": 0.5806655101718972, + "grad_norm": 1.5277315378189087, + "learning_rate": 7.890616339545799e-06, + "loss": 0.4579, + "step": 18410 + }, + { + "epoch": 0.5806970509383378, + "grad_norm": 1.3862107992172241, + "learning_rate": 7.889617779741656e-06, + "loss": 0.4616, + "step": 18411 + }, + { + "epoch": 0.5807285917047784, + "grad_norm": 2.1911420822143555, + "learning_rate": 7.88861924196099e-06, + "loss": 0.7462, + "step": 18412 + }, + { + "epoch": 0.580760132471219, + "grad_norm": 1.5184968709945679, + "learning_rate": 7.88762072621422e-06, + "loss": 0.4303, + "step": 18413 + }, + { + "epoch": 0.5807916732376597, + "grad_norm": 1.3457415103912354, + "learning_rate": 7.886622232511769e-06, + "loss": 0.4279, + "step": 18414 + }, + { + "epoch": 0.5808232140041003, + "grad_norm": 1.6567871570587158, + "learning_rate": 7.885623760864058e-06, + "loss": 0.4472, + "step": 18415 + }, + { + "epoch": 0.5808547547705409, + "grad_norm": 1.3860090970993042, + "learning_rate": 7.8846253112815e-06, + "loss": 0.4619, + "step": 18416 + }, + { + "epoch": 0.5808862955369816, + "grad_norm": 2.243499755859375, + "learning_rate": 7.883626883774524e-06, + "loss": 0.4384, + "step": 18417 + }, + { + "epoch": 0.5809178363034222, + "grad_norm": 1.7039023637771606, + "learning_rate": 7.882628478353544e-06, + "loss": 0.4594, + "step": 18418 + }, + { + "epoch": 0.5809493770698628, + "grad_norm": 1.3985730409622192, + "learning_rate": 7.881630095028978e-06, + "loss": 0.4236, + "step": 18419 + }, + { + "epoch": 0.5809809178363035, + "grad_norm": 1.6939725875854492, + "learning_rate": 7.88063173381125e-06, + "loss": 0.4684, + "step": 18420 + }, + { + "epoch": 0.5810124586027441, + "grad_norm": 1.7142254114151, + "learning_rate": 7.879633394710774e-06, + "loss": 0.4235, + "step": 18421 + }, + { + "epoch": 0.5810439993691847, + "grad_norm": 1.560233473777771, + "learning_rate": 7.878635077737967e-06, + "loss": 0.4755, + "step": 18422 + }, + { + "epoch": 0.5810755401356253, + "grad_norm": 1.8454792499542236, + "learning_rate": 7.877636782903253e-06, + "loss": 0.4537, + "step": 18423 + }, + { + "epoch": 0.581107080902066, + "grad_norm": 47.89569091796875, + "learning_rate": 7.876638510217044e-06, + "loss": 0.4485, + "step": 18424 + }, + { + "epoch": 0.5811386216685065, + "grad_norm": 1.749122977256775, + "learning_rate": 7.875640259689764e-06, + "loss": 0.4609, + "step": 18425 + }, + { + "epoch": 0.5811701624349471, + "grad_norm": 1.4942744970321655, + "learning_rate": 7.874642031331829e-06, + "loss": 0.4275, + "step": 18426 + }, + { + "epoch": 0.5812017032013878, + "grad_norm": 1.7614833116531372, + "learning_rate": 7.873643825153649e-06, + "loss": 0.456, + "step": 18427 + }, + { + "epoch": 0.5812332439678284, + "grad_norm": 1.3367798328399658, + "learning_rate": 7.872645641165651e-06, + "loss": 0.4315, + "step": 18428 + }, + { + "epoch": 0.581264784734269, + "grad_norm": 1.504357099533081, + "learning_rate": 7.871647479378249e-06, + "loss": 0.4115, + "step": 18429 + }, + { + "epoch": 0.5812963255007096, + "grad_norm": 1.647587537765503, + "learning_rate": 7.870649339801856e-06, + "loss": 0.4403, + "step": 18430 + }, + { + "epoch": 0.5813278662671503, + "grad_norm": 1.2220194339752197, + "learning_rate": 7.869651222446891e-06, + "loss": 0.452, + "step": 18431 + }, + { + "epoch": 0.5813594070335909, + "grad_norm": 1.817763328552246, + "learning_rate": 7.868653127323771e-06, + "loss": 0.4061, + "step": 18432 + }, + { + "epoch": 0.5813909478000315, + "grad_norm": 1.7912087440490723, + "learning_rate": 7.867655054442907e-06, + "loss": 0.4563, + "step": 18433 + }, + { + "epoch": 0.5814224885664722, + "grad_norm": 1.3550105094909668, + "learning_rate": 7.86665700381472e-06, + "loss": 0.3913, + "step": 18434 + }, + { + "epoch": 0.5814540293329128, + "grad_norm": 1.986128330230713, + "learning_rate": 7.865658975449625e-06, + "loss": 0.7574, + "step": 18435 + }, + { + "epoch": 0.5814855700993534, + "grad_norm": 1.575252652168274, + "learning_rate": 7.864660969358032e-06, + "loss": 0.4212, + "step": 18436 + }, + { + "epoch": 0.581517110865794, + "grad_norm": 1.9887983798980713, + "learning_rate": 7.863662985550364e-06, + "loss": 0.4544, + "step": 18437 + }, + { + "epoch": 0.5815486516322347, + "grad_norm": 2.031386137008667, + "learning_rate": 7.862665024037031e-06, + "loss": 0.4932, + "step": 18438 + }, + { + "epoch": 0.5815801923986753, + "grad_norm": 2.88601016998291, + "learning_rate": 7.861667084828447e-06, + "loss": 0.4697, + "step": 18439 + }, + { + "epoch": 0.5816117331651159, + "grad_norm": 2.4053711891174316, + "learning_rate": 7.860669167935028e-06, + "loss": 0.7472, + "step": 18440 + }, + { + "epoch": 0.5816432739315566, + "grad_norm": 1.522444248199463, + "learning_rate": 7.859671273367184e-06, + "loss": 0.4571, + "step": 18441 + }, + { + "epoch": 0.5816748146979972, + "grad_norm": 1.6971720457077026, + "learning_rate": 7.858673401135337e-06, + "loss": 0.4622, + "step": 18442 + }, + { + "epoch": 0.5817063554644378, + "grad_norm": 3.7650482654571533, + "learning_rate": 7.857675551249895e-06, + "loss": 0.7177, + "step": 18443 + }, + { + "epoch": 0.5817378962308785, + "grad_norm": 1.327201247215271, + "learning_rate": 7.856677723721268e-06, + "loss": 0.7264, + "step": 18444 + }, + { + "epoch": 0.5817694369973191, + "grad_norm": 1.3573499917984009, + "learning_rate": 7.855679918559878e-06, + "loss": 0.7549, + "step": 18445 + }, + { + "epoch": 0.5818009777637596, + "grad_norm": 1.5122257471084595, + "learning_rate": 7.85468213577613e-06, + "loss": 0.4553, + "step": 18446 + }, + { + "epoch": 0.5818325185302002, + "grad_norm": 1.7179025411605835, + "learning_rate": 7.853684375380439e-06, + "loss": 0.7352, + "step": 18447 + }, + { + "epoch": 0.5818640592966409, + "grad_norm": 2.384361505508423, + "learning_rate": 7.852686637383223e-06, + "loss": 0.4672, + "step": 18448 + }, + { + "epoch": 0.5818956000630815, + "grad_norm": 1.7707464694976807, + "learning_rate": 7.851688921794886e-06, + "loss": 0.4748, + "step": 18449 + }, + { + "epoch": 0.5819271408295221, + "grad_norm": 1.669892430305481, + "learning_rate": 7.850691228625842e-06, + "loss": 0.7401, + "step": 18450 + }, + { + "epoch": 0.5819586815959628, + "grad_norm": 2.244227409362793, + "learning_rate": 7.849693557886507e-06, + "loss": 0.4553, + "step": 18451 + }, + { + "epoch": 0.5819902223624034, + "grad_norm": 2.217535972595215, + "learning_rate": 7.848695909587289e-06, + "loss": 0.4509, + "step": 18452 + }, + { + "epoch": 0.582021763128844, + "grad_norm": 1.498265027999878, + "learning_rate": 7.847698283738596e-06, + "loss": 0.7529, + "step": 18453 + }, + { + "epoch": 0.5820533038952846, + "grad_norm": 1.6435303688049316, + "learning_rate": 7.846700680350847e-06, + "loss": 0.4012, + "step": 18454 + }, + { + "epoch": 0.5820848446617253, + "grad_norm": 1.6277741193771362, + "learning_rate": 7.845703099434443e-06, + "loss": 0.7331, + "step": 18455 + }, + { + "epoch": 0.5821163854281659, + "grad_norm": 3.081413984298706, + "learning_rate": 7.844705540999804e-06, + "loss": 0.3513, + "step": 18456 + }, + { + "epoch": 0.5821479261946065, + "grad_norm": 1.5881843566894531, + "learning_rate": 7.843708005057338e-06, + "loss": 0.4476, + "step": 18457 + }, + { + "epoch": 0.5821794669610472, + "grad_norm": 1.6233601570129395, + "learning_rate": 7.842710491617451e-06, + "loss": 0.47, + "step": 18458 + }, + { + "epoch": 0.5822110077274878, + "grad_norm": 1.8898742198944092, + "learning_rate": 7.841713000690556e-06, + "loss": 0.4621, + "step": 18459 + }, + { + "epoch": 0.5822425484939284, + "grad_norm": 1.5795605182647705, + "learning_rate": 7.840715532287061e-06, + "loss": 0.7335, + "step": 18460 + }, + { + "epoch": 0.582274089260369, + "grad_norm": 1.4308197498321533, + "learning_rate": 7.839718086417372e-06, + "loss": 0.4943, + "step": 18461 + }, + { + "epoch": 0.5823056300268097, + "grad_norm": 1.7383595705032349, + "learning_rate": 7.838720663091907e-06, + "loss": 0.4545, + "step": 18462 + }, + { + "epoch": 0.5823371707932503, + "grad_norm": 1.9613902568817139, + "learning_rate": 7.837723262321071e-06, + "loss": 0.3676, + "step": 18463 + }, + { + "epoch": 0.5823687115596909, + "grad_norm": 1.693830966949463, + "learning_rate": 7.836725884115268e-06, + "loss": 0.4544, + "step": 18464 + }, + { + "epoch": 0.5824002523261316, + "grad_norm": 1.842535376548767, + "learning_rate": 7.835728528484913e-06, + "loss": 0.4284, + "step": 18465 + }, + { + "epoch": 0.5824317930925722, + "grad_norm": 1.7013297080993652, + "learning_rate": 7.834731195440411e-06, + "loss": 0.4732, + "step": 18466 + }, + { + "epoch": 0.5824633338590127, + "grad_norm": 1.5819425582885742, + "learning_rate": 7.83373388499217e-06, + "loss": 0.4527, + "step": 18467 + }, + { + "epoch": 0.5824948746254534, + "grad_norm": 1.6362371444702148, + "learning_rate": 7.832736597150595e-06, + "loss": 0.4639, + "step": 18468 + }, + { + "epoch": 0.582526415391894, + "grad_norm": 1.35663640499115, + "learning_rate": 7.8317393319261e-06, + "loss": 0.4575, + "step": 18469 + }, + { + "epoch": 0.5825579561583346, + "grad_norm": 1.325960397720337, + "learning_rate": 7.830742089329088e-06, + "loss": 0.4693, + "step": 18470 + }, + { + "epoch": 0.5825894969247752, + "grad_norm": 1.4430396556854248, + "learning_rate": 7.829744869369966e-06, + "loss": 0.451, + "step": 18471 + }, + { + "epoch": 0.5826210376912159, + "grad_norm": 1.4266507625579834, + "learning_rate": 7.82874767205914e-06, + "loss": 0.467, + "step": 18472 + }, + { + "epoch": 0.5826525784576565, + "grad_norm": 1.3376272916793823, + "learning_rate": 7.82775049740702e-06, + "loss": 0.4483, + "step": 18473 + }, + { + "epoch": 0.5826841192240971, + "grad_norm": 1.6225517988204956, + "learning_rate": 7.826753345424013e-06, + "loss": 0.4263, + "step": 18474 + }, + { + "epoch": 0.5827156599905378, + "grad_norm": 1.4207463264465332, + "learning_rate": 7.825756216120514e-06, + "loss": 0.4667, + "step": 18475 + }, + { + "epoch": 0.5827472007569784, + "grad_norm": 1.658642292022705, + "learning_rate": 7.824759109506945e-06, + "loss": 0.4784, + "step": 18476 + }, + { + "epoch": 0.582778741523419, + "grad_norm": 1.8005504608154297, + "learning_rate": 7.823762025593702e-06, + "loss": 0.4714, + "step": 18477 + }, + { + "epoch": 0.5828102822898596, + "grad_norm": 1.52592933177948, + "learning_rate": 7.82276496439119e-06, + "loss": 0.3889, + "step": 18478 + }, + { + "epoch": 0.5828418230563003, + "grad_norm": 1.3374305963516235, + "learning_rate": 7.821767925909819e-06, + "loss": 0.4361, + "step": 18479 + }, + { + "epoch": 0.5828733638227409, + "grad_norm": 1.3035026788711548, + "learning_rate": 7.820770910159991e-06, + "loss": 0.3966, + "step": 18480 + }, + { + "epoch": 0.5829049045891815, + "grad_norm": 1.744242548942566, + "learning_rate": 7.819773917152105e-06, + "loss": 0.4554, + "step": 18481 + }, + { + "epoch": 0.5829364453556222, + "grad_norm": 1.6048387289047241, + "learning_rate": 7.818776946896578e-06, + "loss": 0.4645, + "step": 18482 + }, + { + "epoch": 0.5829679861220628, + "grad_norm": 2.6077067852020264, + "learning_rate": 7.817779999403804e-06, + "loss": 0.765, + "step": 18483 + }, + { + "epoch": 0.5829995268885034, + "grad_norm": 1.862810730934143, + "learning_rate": 7.816783074684189e-06, + "loss": 0.4523, + "step": 18484 + }, + { + "epoch": 0.5830310676549441, + "grad_norm": 1.859708309173584, + "learning_rate": 7.81578617274814e-06, + "loss": 0.401, + "step": 18485 + }, + { + "epoch": 0.5830626084213847, + "grad_norm": 1.749950647354126, + "learning_rate": 7.81478929360606e-06, + "loss": 0.3978, + "step": 18486 + }, + { + "epoch": 0.5830941491878253, + "grad_norm": 1.5993574857711792, + "learning_rate": 7.813792437268348e-06, + "loss": 0.4683, + "step": 18487 + }, + { + "epoch": 0.5831256899542658, + "grad_norm": 1.9050331115722656, + "learning_rate": 7.812795603745407e-06, + "loss": 0.4571, + "step": 18488 + }, + { + "epoch": 0.5831572307207065, + "grad_norm": 1.2795623540878296, + "learning_rate": 7.811798793047644e-06, + "loss": 0.3662, + "step": 18489 + }, + { + "epoch": 0.5831887714871471, + "grad_norm": 1.4853591918945312, + "learning_rate": 7.810802005185463e-06, + "loss": 0.4496, + "step": 18490 + }, + { + "epoch": 0.5832203122535877, + "grad_norm": 1.6302056312561035, + "learning_rate": 7.809805240169261e-06, + "loss": 0.4098, + "step": 18491 + }, + { + "epoch": 0.5832518530200284, + "grad_norm": 1.1794747114181519, + "learning_rate": 7.808808498009439e-06, + "loss": 0.4537, + "step": 18492 + }, + { + "epoch": 0.583283393786469, + "grad_norm": 1.7003923654556274, + "learning_rate": 7.807811778716405e-06, + "loss": 0.4376, + "step": 18493 + }, + { + "epoch": 0.5833149345529096, + "grad_norm": 1.2189171314239502, + "learning_rate": 7.806815082300558e-06, + "loss": 0.7245, + "step": 18494 + }, + { + "epoch": 0.5833464753193502, + "grad_norm": 1.477004051208496, + "learning_rate": 7.805818408772295e-06, + "loss": 0.4466, + "step": 18495 + }, + { + "epoch": 0.5833780160857909, + "grad_norm": 1.714450716972351, + "learning_rate": 7.804821758142023e-06, + "loss": 0.4086, + "step": 18496 + }, + { + "epoch": 0.5834095568522315, + "grad_norm": 1.4743250608444214, + "learning_rate": 7.803825130420141e-06, + "loss": 0.4305, + "step": 18497 + }, + { + "epoch": 0.5834410976186721, + "grad_norm": 1.7073932886123657, + "learning_rate": 7.802828525617048e-06, + "loss": 0.4356, + "step": 18498 + }, + { + "epoch": 0.5834726383851128, + "grad_norm": 1.3739093542099, + "learning_rate": 7.801831943743146e-06, + "loss": 0.4216, + "step": 18499 + }, + { + "epoch": 0.5835041791515534, + "grad_norm": 2.0792155265808105, + "learning_rate": 7.800835384808836e-06, + "loss": 0.4631, + "step": 18500 + }, + { + "epoch": 0.583535719917994, + "grad_norm": 1.5592812299728394, + "learning_rate": 7.799838848824512e-06, + "loss": 0.4667, + "step": 18501 + }, + { + "epoch": 0.5835672606844347, + "grad_norm": 1.7244796752929688, + "learning_rate": 7.798842335800581e-06, + "loss": 0.4549, + "step": 18502 + }, + { + "epoch": 0.5835988014508753, + "grad_norm": 1.4080455303192139, + "learning_rate": 7.797845845747436e-06, + "loss": 0.4549, + "step": 18503 + }, + { + "epoch": 0.5836303422173159, + "grad_norm": 1.3811527490615845, + "learning_rate": 7.796849378675483e-06, + "loss": 0.4418, + "step": 18504 + }, + { + "epoch": 0.5836618829837565, + "grad_norm": 1.6594125032424927, + "learning_rate": 7.795852934595118e-06, + "loss": 0.4483, + "step": 18505 + }, + { + "epoch": 0.5836934237501972, + "grad_norm": 1.6747475862503052, + "learning_rate": 7.794856513516738e-06, + "loss": 0.4506, + "step": 18506 + }, + { + "epoch": 0.5837249645166378, + "grad_norm": 1.3743230104446411, + "learning_rate": 7.793860115450744e-06, + "loss": 0.3566, + "step": 18507 + }, + { + "epoch": 0.5837565052830784, + "grad_norm": 1.390075922012329, + "learning_rate": 7.79286374040753e-06, + "loss": 0.4542, + "step": 18508 + }, + { + "epoch": 0.583788046049519, + "grad_norm": 1.9809880256652832, + "learning_rate": 7.791867388397497e-06, + "loss": 0.4692, + "step": 18509 + }, + { + "epoch": 0.5838195868159596, + "grad_norm": 1.937682867050171, + "learning_rate": 7.790871059431044e-06, + "loss": 0.4751, + "step": 18510 + }, + { + "epoch": 0.5838511275824002, + "grad_norm": 1.9250353574752808, + "learning_rate": 7.789874753518569e-06, + "loss": 0.4746, + "step": 18511 + }, + { + "epoch": 0.5838826683488408, + "grad_norm": 2.680938720703125, + "learning_rate": 7.78887847067046e-06, + "loss": 0.4747, + "step": 18512 + }, + { + "epoch": 0.5839142091152815, + "grad_norm": 1.4794925451278687, + "learning_rate": 7.787882210897126e-06, + "loss": 0.4291, + "step": 18513 + }, + { + "epoch": 0.5839457498817221, + "grad_norm": 1.71542227268219, + "learning_rate": 7.78688597420896e-06, + "loss": 0.7116, + "step": 18514 + }, + { + "epoch": 0.5839772906481627, + "grad_norm": 1.285834789276123, + "learning_rate": 7.785889760616352e-06, + "loss": 0.4518, + "step": 18515 + }, + { + "epoch": 0.5840088314146034, + "grad_norm": 2.430689811706543, + "learning_rate": 7.784893570129708e-06, + "loss": 0.4447, + "step": 18516 + }, + { + "epoch": 0.584040372181044, + "grad_norm": 1.3570051193237305, + "learning_rate": 7.78389740275942e-06, + "loss": 0.4829, + "step": 18517 + }, + { + "epoch": 0.5840719129474846, + "grad_norm": 1.4717599153518677, + "learning_rate": 7.782901258515881e-06, + "loss": 0.4572, + "step": 18518 + }, + { + "epoch": 0.5841034537139252, + "grad_norm": 1.908555030822754, + "learning_rate": 7.78190513740949e-06, + "loss": 0.4711, + "step": 18519 + }, + { + "epoch": 0.5841349944803659, + "grad_norm": 2.1032910346984863, + "learning_rate": 7.780909039450637e-06, + "loss": 0.7739, + "step": 18520 + }, + { + "epoch": 0.5841665352468065, + "grad_norm": 1.48610258102417, + "learning_rate": 7.779912964649727e-06, + "loss": 0.4456, + "step": 18521 + }, + { + "epoch": 0.5841980760132471, + "grad_norm": 1.5336381196975708, + "learning_rate": 7.778916913017149e-06, + "loss": 0.4246, + "step": 18522 + }, + { + "epoch": 0.5842296167796878, + "grad_norm": 20.370424270629883, + "learning_rate": 7.777920884563292e-06, + "loss": 0.3888, + "step": 18523 + }, + { + "epoch": 0.5842611575461284, + "grad_norm": 1.9077149629592896, + "learning_rate": 7.776924879298562e-06, + "loss": 0.3378, + "step": 18524 + }, + { + "epoch": 0.584292698312569, + "grad_norm": 1.617078185081482, + "learning_rate": 7.775928897233347e-06, + "loss": 0.4831, + "step": 18525 + }, + { + "epoch": 0.5843242390790097, + "grad_norm": 1.5044180154800415, + "learning_rate": 7.774932938378037e-06, + "loss": 0.7464, + "step": 18526 + }, + { + "epoch": 0.5843557798454503, + "grad_norm": 1.4792181253433228, + "learning_rate": 7.773937002743033e-06, + "loss": 0.4531, + "step": 18527 + }, + { + "epoch": 0.5843873206118909, + "grad_norm": 1.4170355796813965, + "learning_rate": 7.772941090338722e-06, + "loss": 0.4422, + "step": 18528 + }, + { + "epoch": 0.5844188613783315, + "grad_norm": 1.36378014087677, + "learning_rate": 7.771945201175501e-06, + "loss": 0.4467, + "step": 18529 + }, + { + "epoch": 0.5844504021447721, + "grad_norm": 1.8432341814041138, + "learning_rate": 7.770949335263761e-06, + "loss": 0.4633, + "step": 18530 + }, + { + "epoch": 0.5844819429112127, + "grad_norm": 1.3118401765823364, + "learning_rate": 7.7699534926139e-06, + "loss": 0.3644, + "step": 18531 + }, + { + "epoch": 0.5845134836776533, + "grad_norm": 1.5886971950531006, + "learning_rate": 7.7689576732363e-06, + "loss": 0.3413, + "step": 18532 + }, + { + "epoch": 0.584545024444094, + "grad_norm": 1.5596586465835571, + "learning_rate": 7.767961877141361e-06, + "loss": 0.4631, + "step": 18533 + }, + { + "epoch": 0.5845765652105346, + "grad_norm": 1.6359777450561523, + "learning_rate": 7.766966104339476e-06, + "loss": 0.4468, + "step": 18534 + }, + { + "epoch": 0.5846081059769752, + "grad_norm": 1.3585107326507568, + "learning_rate": 7.765970354841027e-06, + "loss": 0.4624, + "step": 18535 + }, + { + "epoch": 0.5846396467434158, + "grad_norm": 1.4904063940048218, + "learning_rate": 7.764974628656418e-06, + "loss": 0.465, + "step": 18536 + }, + { + "epoch": 0.5846711875098565, + "grad_norm": 1.4286658763885498, + "learning_rate": 7.76397892579603e-06, + "loss": 0.7178, + "step": 18537 + }, + { + "epoch": 0.5847027282762971, + "grad_norm": 1.6114760637283325, + "learning_rate": 7.762983246270262e-06, + "loss": 0.4578, + "step": 18538 + }, + { + "epoch": 0.5847342690427377, + "grad_norm": 2.341538429260254, + "learning_rate": 7.7619875900895e-06, + "loss": 0.4642, + "step": 18539 + }, + { + "epoch": 0.5847658098091784, + "grad_norm": 2.1200788021087646, + "learning_rate": 7.760991957264132e-06, + "loss": 0.4708, + "step": 18540 + }, + { + "epoch": 0.584797350575619, + "grad_norm": 1.8228521347045898, + "learning_rate": 7.759996347804555e-06, + "loss": 0.4477, + "step": 18541 + }, + { + "epoch": 0.5848288913420596, + "grad_norm": 1.4477150440216064, + "learning_rate": 7.759000761721156e-06, + "loss": 0.464, + "step": 18542 + }, + { + "epoch": 0.5848604321085002, + "grad_norm": 1.617701530456543, + "learning_rate": 7.75800519902432e-06, + "loss": 0.3819, + "step": 18543 + }, + { + "epoch": 0.5848919728749409, + "grad_norm": 1.487047791481018, + "learning_rate": 7.757009659724444e-06, + "loss": 0.465, + "step": 18544 + }, + { + "epoch": 0.5849235136413815, + "grad_norm": 1.5616984367370605, + "learning_rate": 7.756014143831914e-06, + "loss": 0.4458, + "step": 18545 + }, + { + "epoch": 0.5849550544078221, + "grad_norm": 3.0063259601593018, + "learning_rate": 7.755018651357117e-06, + "loss": 0.4507, + "step": 18546 + }, + { + "epoch": 0.5849865951742628, + "grad_norm": 1.4467947483062744, + "learning_rate": 7.754023182310445e-06, + "loss": 0.453, + "step": 18547 + }, + { + "epoch": 0.5850181359407034, + "grad_norm": 1.52998948097229, + "learning_rate": 7.753027736702283e-06, + "loss": 0.4789, + "step": 18548 + }, + { + "epoch": 0.585049676707144, + "grad_norm": 2.1510608196258545, + "learning_rate": 7.752032314543022e-06, + "loss": 0.7549, + "step": 18549 + }, + { + "epoch": 0.5850812174735847, + "grad_norm": 1.4580601453781128, + "learning_rate": 7.751036915843052e-06, + "loss": 0.4161, + "step": 18550 + }, + { + "epoch": 0.5851127582400253, + "grad_norm": 1.5836563110351562, + "learning_rate": 7.750041540612752e-06, + "loss": 0.4567, + "step": 18551 + }, + { + "epoch": 0.5851442990064658, + "grad_norm": 1.4075268507003784, + "learning_rate": 7.749046188862522e-06, + "loss": 0.4344, + "step": 18552 + }, + { + "epoch": 0.5851758397729064, + "grad_norm": 1.702994465827942, + "learning_rate": 7.748050860602741e-06, + "loss": 0.4665, + "step": 18553 + }, + { + "epoch": 0.5852073805393471, + "grad_norm": 1.5742721557617188, + "learning_rate": 7.747055555843794e-06, + "loss": 0.4696, + "step": 18554 + }, + { + "epoch": 0.5852389213057877, + "grad_norm": 1.4145681858062744, + "learning_rate": 7.746060274596077e-06, + "loss": 0.4523, + "step": 18555 + }, + { + "epoch": 0.5852704620722283, + "grad_norm": 1.6488804817199707, + "learning_rate": 7.74506501686997e-06, + "loss": 0.458, + "step": 18556 + }, + { + "epoch": 0.585302002838669, + "grad_norm": 1.3450530767440796, + "learning_rate": 7.744069782675858e-06, + "loss": 0.4738, + "step": 18557 + }, + { + "epoch": 0.5853335436051096, + "grad_norm": 2.2613770961761475, + "learning_rate": 7.743074572024132e-06, + "loss": 0.4452, + "step": 18558 + }, + { + "epoch": 0.5853650843715502, + "grad_norm": 1.5070335865020752, + "learning_rate": 7.742079384925176e-06, + "loss": 0.4738, + "step": 18559 + }, + { + "epoch": 0.5853966251379908, + "grad_norm": 1.7372604608535767, + "learning_rate": 7.741084221389371e-06, + "loss": 0.4612, + "step": 18560 + }, + { + "epoch": 0.5854281659044315, + "grad_norm": 1.3822909593582153, + "learning_rate": 7.740089081427109e-06, + "loss": 0.4643, + "step": 18561 + }, + { + "epoch": 0.5854597066708721, + "grad_norm": 2.8816654682159424, + "learning_rate": 7.739093965048773e-06, + "loss": 0.4522, + "step": 18562 + }, + { + "epoch": 0.5854912474373127, + "grad_norm": 1.8455675840377808, + "learning_rate": 7.738098872264743e-06, + "loss": 0.452, + "step": 18563 + }, + { + "epoch": 0.5855227882037534, + "grad_norm": 1.5072845220565796, + "learning_rate": 7.737103803085411e-06, + "loss": 0.4309, + "step": 18564 + }, + { + "epoch": 0.585554328970194, + "grad_norm": 1.3360850811004639, + "learning_rate": 7.736108757521159e-06, + "loss": 0.755, + "step": 18565 + }, + { + "epoch": 0.5855858697366346, + "grad_norm": 1.7586700916290283, + "learning_rate": 7.735113735582367e-06, + "loss": 0.4542, + "step": 18566 + }, + { + "epoch": 0.5856174105030753, + "grad_norm": 1.6316204071044922, + "learning_rate": 7.734118737279424e-06, + "loss": 0.7406, + "step": 18567 + }, + { + "epoch": 0.5856489512695159, + "grad_norm": 1.8504668474197388, + "learning_rate": 7.733123762622707e-06, + "loss": 0.7493, + "step": 18568 + }, + { + "epoch": 0.5856804920359565, + "grad_norm": 1.3292497396469116, + "learning_rate": 7.732128811622608e-06, + "loss": 0.4044, + "step": 18569 + }, + { + "epoch": 0.5857120328023971, + "grad_norm": 1.980853796005249, + "learning_rate": 7.731133884289507e-06, + "loss": 0.4241, + "step": 18570 + }, + { + "epoch": 0.5857435735688378, + "grad_norm": 1.4710408449172974, + "learning_rate": 7.73013898063378e-06, + "loss": 0.4298, + "step": 18571 + }, + { + "epoch": 0.5857751143352784, + "grad_norm": 1.4516091346740723, + "learning_rate": 7.72914410066582e-06, + "loss": 0.4603, + "step": 18572 + }, + { + "epoch": 0.5858066551017189, + "grad_norm": 1.8748645782470703, + "learning_rate": 7.728149244396004e-06, + "loss": 0.4489, + "step": 18573 + }, + { + "epoch": 0.5858381958681596, + "grad_norm": 1.760696530342102, + "learning_rate": 7.727154411834712e-06, + "loss": 0.7445, + "step": 18574 + }, + { + "epoch": 0.5858697366346002, + "grad_norm": 1.124783992767334, + "learning_rate": 7.726159602992332e-06, + "loss": 0.3733, + "step": 18575 + }, + { + "epoch": 0.5859012774010408, + "grad_norm": 1.3532860279083252, + "learning_rate": 7.72516481787924e-06, + "loss": 0.4747, + "step": 18576 + }, + { + "epoch": 0.5859328181674814, + "grad_norm": 1.6751747131347656, + "learning_rate": 7.72417005650582e-06, + "loss": 0.3547, + "step": 18577 + }, + { + "epoch": 0.5859643589339221, + "grad_norm": 1.6523456573486328, + "learning_rate": 7.723175318882454e-06, + "loss": 0.4264, + "step": 18578 + }, + { + "epoch": 0.5859958997003627, + "grad_norm": 1.4201247692108154, + "learning_rate": 7.722180605019521e-06, + "loss": 0.3411, + "step": 18579 + }, + { + "epoch": 0.5860274404668033, + "grad_norm": 1.8703243732452393, + "learning_rate": 7.721185914927399e-06, + "loss": 0.4832, + "step": 18580 + }, + { + "epoch": 0.586058981233244, + "grad_norm": 1.8343923091888428, + "learning_rate": 7.720191248616475e-06, + "loss": 0.4741, + "step": 18581 + }, + { + "epoch": 0.5860905219996846, + "grad_norm": 7.5911173820495605, + "learning_rate": 7.719196606097122e-06, + "loss": 0.449, + "step": 18582 + }, + { + "epoch": 0.5861220627661252, + "grad_norm": 1.6015156507492065, + "learning_rate": 7.718201987379725e-06, + "loss": 0.4441, + "step": 18583 + }, + { + "epoch": 0.5861536035325658, + "grad_norm": 1.5795812606811523, + "learning_rate": 7.717207392474665e-06, + "loss": 0.4353, + "step": 18584 + }, + { + "epoch": 0.5861851442990065, + "grad_norm": 1.525376558303833, + "learning_rate": 7.716212821392315e-06, + "loss": 0.465, + "step": 18585 + }, + { + "epoch": 0.5862166850654471, + "grad_norm": 1.5365945100784302, + "learning_rate": 7.71521827414306e-06, + "loss": 0.4617, + "step": 18586 + }, + { + "epoch": 0.5862482258318877, + "grad_norm": 1.8708912134170532, + "learning_rate": 7.714223750737276e-06, + "loss": 0.4476, + "step": 18587 + }, + { + "epoch": 0.5862797665983284, + "grad_norm": 1.5565600395202637, + "learning_rate": 7.713229251185337e-06, + "loss": 0.4378, + "step": 18588 + }, + { + "epoch": 0.586311307364769, + "grad_norm": 1.7415521144866943, + "learning_rate": 7.712234775497632e-06, + "loss": 0.4363, + "step": 18589 + }, + { + "epoch": 0.5863428481312096, + "grad_norm": 2.109389305114746, + "learning_rate": 7.711240323684533e-06, + "loss": 0.4199, + "step": 18590 + }, + { + "epoch": 0.5863743888976503, + "grad_norm": 1.5789573192596436, + "learning_rate": 7.710245895756414e-06, + "loss": 0.4609, + "step": 18591 + }, + { + "epoch": 0.5864059296640909, + "grad_norm": 2.0908045768737793, + "learning_rate": 7.709251491723663e-06, + "loss": 0.4429, + "step": 18592 + }, + { + "epoch": 0.5864374704305315, + "grad_norm": 1.7551078796386719, + "learning_rate": 7.70825711159665e-06, + "loss": 0.4468, + "step": 18593 + }, + { + "epoch": 0.586469011196972, + "grad_norm": 1.6429098844528198, + "learning_rate": 7.707262755385749e-06, + "loss": 0.4431, + "step": 18594 + }, + { + "epoch": 0.5865005519634127, + "grad_norm": 1.5251692533493042, + "learning_rate": 7.706268423101346e-06, + "loss": 0.4557, + "step": 18595 + }, + { + "epoch": 0.5865320927298533, + "grad_norm": 2.5430667400360107, + "learning_rate": 7.705274114753814e-06, + "loss": 0.4048, + "step": 18596 + }, + { + "epoch": 0.5865636334962939, + "grad_norm": 1.4584598541259766, + "learning_rate": 7.704279830353524e-06, + "loss": 0.436, + "step": 18597 + }, + { + "epoch": 0.5865951742627346, + "grad_norm": 1.6371639966964722, + "learning_rate": 7.70328556991086e-06, + "loss": 0.4029, + "step": 18598 + }, + { + "epoch": 0.5866267150291752, + "grad_norm": 4.619340419769287, + "learning_rate": 7.702291333436191e-06, + "loss": 0.7306, + "step": 18599 + }, + { + "epoch": 0.5866582557956158, + "grad_norm": 1.6426641941070557, + "learning_rate": 7.701297120939898e-06, + "loss": 0.4598, + "step": 18600 + }, + { + "epoch": 0.5866897965620564, + "grad_norm": 2.489687204360962, + "learning_rate": 7.700302932432356e-06, + "loss": 0.4566, + "step": 18601 + }, + { + "epoch": 0.5867213373284971, + "grad_norm": 1.398767352104187, + "learning_rate": 7.699308767923934e-06, + "loss": 0.4319, + "step": 18602 + }, + { + "epoch": 0.5867528780949377, + "grad_norm": 1.577968955039978, + "learning_rate": 7.698314627425016e-06, + "loss": 0.4313, + "step": 18603 + }, + { + "epoch": 0.5867844188613783, + "grad_norm": 2.4347078800201416, + "learning_rate": 7.697320510945972e-06, + "loss": 0.4524, + "step": 18604 + }, + { + "epoch": 0.586815959627819, + "grad_norm": 1.7593330144882202, + "learning_rate": 7.696326418497176e-06, + "loss": 0.4729, + "step": 18605 + }, + { + "epoch": 0.5868475003942596, + "grad_norm": 1.644906997680664, + "learning_rate": 7.695332350089002e-06, + "loss": 0.4152, + "step": 18606 + }, + { + "epoch": 0.5868790411607002, + "grad_norm": 1.43985915184021, + "learning_rate": 7.694338305731825e-06, + "loss": 0.4453, + "step": 18607 + }, + { + "epoch": 0.5869105819271409, + "grad_norm": 1.5850446224212646, + "learning_rate": 7.693344285436015e-06, + "loss": 0.4782, + "step": 18608 + }, + { + "epoch": 0.5869421226935815, + "grad_norm": 1.6416951417922974, + "learning_rate": 7.692350289211952e-06, + "loss": 0.4285, + "step": 18609 + }, + { + "epoch": 0.5869736634600221, + "grad_norm": 1.446603775024414, + "learning_rate": 7.691356317070004e-06, + "loss": 0.4401, + "step": 18610 + }, + { + "epoch": 0.5870052042264627, + "grad_norm": 1.586668848991394, + "learning_rate": 7.690362369020545e-06, + "loss": 0.4641, + "step": 18611 + }, + { + "epoch": 0.5870367449929034, + "grad_norm": 1.6199560165405273, + "learning_rate": 7.689368445073949e-06, + "loss": 0.4533, + "step": 18612 + }, + { + "epoch": 0.587068285759344, + "grad_norm": 1.8732398748397827, + "learning_rate": 7.688374545240588e-06, + "loss": 0.4455, + "step": 18613 + }, + { + "epoch": 0.5870998265257846, + "grad_norm": 2.8280818462371826, + "learning_rate": 7.68738066953083e-06, + "loss": 0.4342, + "step": 18614 + }, + { + "epoch": 0.5871313672922251, + "grad_norm": 1.5171546936035156, + "learning_rate": 7.686386817955055e-06, + "loss": 0.4686, + "step": 18615 + }, + { + "epoch": 0.5871629080586658, + "grad_norm": 1.3297865390777588, + "learning_rate": 7.685392990523628e-06, + "loss": 0.4692, + "step": 18616 + }, + { + "epoch": 0.5871944488251064, + "grad_norm": 1.8742108345031738, + "learning_rate": 7.684399187246922e-06, + "loss": 0.4419, + "step": 18617 + }, + { + "epoch": 0.587225989591547, + "grad_norm": 1.9280681610107422, + "learning_rate": 7.683405408135311e-06, + "loss": 0.4813, + "step": 18618 + }, + { + "epoch": 0.5872575303579877, + "grad_norm": 1.800098180770874, + "learning_rate": 7.682411653199159e-06, + "loss": 0.4729, + "step": 18619 + }, + { + "epoch": 0.5872890711244283, + "grad_norm": 1.32877779006958, + "learning_rate": 7.681417922448845e-06, + "loss": 0.4691, + "step": 18620 + }, + { + "epoch": 0.5873206118908689, + "grad_norm": 1.5365921258926392, + "learning_rate": 7.680424215894735e-06, + "loss": 0.4641, + "step": 18621 + }, + { + "epoch": 0.5873521526573096, + "grad_norm": 1.8208931684494019, + "learning_rate": 7.679430533547194e-06, + "loss": 0.452, + "step": 18622 + }, + { + "epoch": 0.5873836934237502, + "grad_norm": 1.3260512351989746, + "learning_rate": 7.678436875416604e-06, + "loss": 0.7596, + "step": 18623 + }, + { + "epoch": 0.5874152341901908, + "grad_norm": 1.5735052824020386, + "learning_rate": 7.677443241513327e-06, + "loss": 0.4614, + "step": 18624 + }, + { + "epoch": 0.5874467749566314, + "grad_norm": 0.9855644106864929, + "learning_rate": 7.67644963184773e-06, + "loss": 0.362, + "step": 18625 + }, + { + "epoch": 0.5874783157230721, + "grad_norm": 1.4550251960754395, + "learning_rate": 7.675456046430188e-06, + "loss": 0.4515, + "step": 18626 + }, + { + "epoch": 0.5875098564895127, + "grad_norm": 2.1345105171203613, + "learning_rate": 7.674462485271068e-06, + "loss": 0.4661, + "step": 18627 + }, + { + "epoch": 0.5875413972559533, + "grad_norm": 3.975646734237671, + "learning_rate": 7.673468948380732e-06, + "loss": 0.4678, + "step": 18628 + }, + { + "epoch": 0.587572938022394, + "grad_norm": 1.5918604135513306, + "learning_rate": 7.672475435769559e-06, + "loss": 0.4281, + "step": 18629 + }, + { + "epoch": 0.5876044787888346, + "grad_norm": 2.138617992401123, + "learning_rate": 7.671481947447908e-06, + "loss": 0.4725, + "step": 18630 + }, + { + "epoch": 0.5876360195552752, + "grad_norm": 2.133661985397339, + "learning_rate": 7.670488483426154e-06, + "loss": 0.4144, + "step": 18631 + }, + { + "epoch": 0.5876675603217159, + "grad_norm": 1.5821713209152222, + "learning_rate": 7.669495043714662e-06, + "loss": 0.4205, + "step": 18632 + }, + { + "epoch": 0.5876991010881565, + "grad_norm": 1.5816890001296997, + "learning_rate": 7.668501628323796e-06, + "loss": 0.4634, + "step": 18633 + }, + { + "epoch": 0.5877306418545971, + "grad_norm": 1.347060203552246, + "learning_rate": 7.66750823726393e-06, + "loss": 0.4632, + "step": 18634 + }, + { + "epoch": 0.5877621826210377, + "grad_norm": 1.4245392084121704, + "learning_rate": 7.666514870545423e-06, + "loss": 0.4576, + "step": 18635 + }, + { + "epoch": 0.5877937233874783, + "grad_norm": 1.6950812339782715, + "learning_rate": 7.665521528178647e-06, + "loss": 0.4406, + "step": 18636 + }, + { + "epoch": 0.5878252641539189, + "grad_norm": 1.2640714645385742, + "learning_rate": 7.664528210173968e-06, + "loss": 0.3264, + "step": 18637 + }, + { + "epoch": 0.5878568049203595, + "grad_norm": 1.857182264328003, + "learning_rate": 7.66353491654175e-06, + "loss": 0.4699, + "step": 18638 + }, + { + "epoch": 0.5878883456868002, + "grad_norm": 1.8391121625900269, + "learning_rate": 7.662541647292355e-06, + "loss": 0.4649, + "step": 18639 + }, + { + "epoch": 0.5879198864532408, + "grad_norm": 1.7886725664138794, + "learning_rate": 7.661548402436157e-06, + "loss": 0.4335, + "step": 18640 + }, + { + "epoch": 0.5879514272196814, + "grad_norm": 1.5477783679962158, + "learning_rate": 7.660555181983517e-06, + "loss": 0.416, + "step": 18641 + }, + { + "epoch": 0.587982967986122, + "grad_norm": 1.3977584838867188, + "learning_rate": 7.659561985944797e-06, + "loss": 0.7424, + "step": 18642 + }, + { + "epoch": 0.5880145087525627, + "grad_norm": 3.321234703063965, + "learning_rate": 7.65856881433037e-06, + "loss": 0.4905, + "step": 18643 + }, + { + "epoch": 0.5880460495190033, + "grad_norm": 2.1119942665100098, + "learning_rate": 7.657575667150592e-06, + "loss": 0.3579, + "step": 18644 + }, + { + "epoch": 0.5880775902854439, + "grad_norm": 1.4647020101547241, + "learning_rate": 7.656582544415832e-06, + "loss": 0.4967, + "step": 18645 + }, + { + "epoch": 0.5881091310518846, + "grad_norm": 1.335587978363037, + "learning_rate": 7.655589446136453e-06, + "loss": 0.7873, + "step": 18646 + }, + { + "epoch": 0.5881406718183252, + "grad_norm": 1.3616338968276978, + "learning_rate": 7.654596372322816e-06, + "loss": 0.4544, + "step": 18647 + }, + { + "epoch": 0.5881722125847658, + "grad_norm": 1.5874484777450562, + "learning_rate": 7.65360332298529e-06, + "loss": 0.426, + "step": 18648 + }, + { + "epoch": 0.5882037533512064, + "grad_norm": 1.5476694107055664, + "learning_rate": 7.652610298134236e-06, + "loss": 0.4139, + "step": 18649 + }, + { + "epoch": 0.5882352941176471, + "grad_norm": 1.841130018234253, + "learning_rate": 7.65161729778001e-06, + "loss": 0.4048, + "step": 18650 + }, + { + "epoch": 0.5882668348840877, + "grad_norm": 1.6708061695098877, + "learning_rate": 7.650624321932988e-06, + "loss": 0.4505, + "step": 18651 + }, + { + "epoch": 0.5882983756505283, + "grad_norm": 1.466464877128601, + "learning_rate": 7.649631370603523e-06, + "loss": 0.4669, + "step": 18652 + }, + { + "epoch": 0.588329916416969, + "grad_norm": 2.191305637359619, + "learning_rate": 7.648638443801978e-06, + "loss": 0.3925, + "step": 18653 + }, + { + "epoch": 0.5883614571834096, + "grad_norm": 1.9620345830917358, + "learning_rate": 7.647645541538718e-06, + "loss": 0.4607, + "step": 18654 + }, + { + "epoch": 0.5883929979498502, + "grad_norm": 1.8847140073776245, + "learning_rate": 7.646652663824106e-06, + "loss": 0.4382, + "step": 18655 + }, + { + "epoch": 0.5884245387162909, + "grad_norm": 2.257169723510742, + "learning_rate": 7.645659810668498e-06, + "loss": 0.4435, + "step": 18656 + }, + { + "epoch": 0.5884560794827314, + "grad_norm": 1.4470356702804565, + "learning_rate": 7.644666982082258e-06, + "loss": 0.3685, + "step": 18657 + }, + { + "epoch": 0.588487620249172, + "grad_norm": 2.070986747741699, + "learning_rate": 7.643674178075748e-06, + "loss": 0.4514, + "step": 18658 + }, + { + "epoch": 0.5885191610156126, + "grad_norm": 1.539169430732727, + "learning_rate": 7.642681398659325e-06, + "loss": 0.4651, + "step": 18659 + }, + { + "epoch": 0.5885507017820533, + "grad_norm": 1.4457120895385742, + "learning_rate": 7.641688643843354e-06, + "loss": 0.345, + "step": 18660 + }, + { + "epoch": 0.5885822425484939, + "grad_norm": 1.487388253211975, + "learning_rate": 7.640695913638195e-06, + "loss": 0.4739, + "step": 18661 + }, + { + "epoch": 0.5886137833149345, + "grad_norm": 1.767733097076416, + "learning_rate": 7.639703208054201e-06, + "loss": 0.4504, + "step": 18662 + }, + { + "epoch": 0.5886453240813752, + "grad_norm": 1.4202134609222412, + "learning_rate": 7.638710527101741e-06, + "loss": 0.465, + "step": 18663 + }, + { + "epoch": 0.5886768648478158, + "grad_norm": 1.4015206098556519, + "learning_rate": 7.637717870791167e-06, + "loss": 0.4851, + "step": 18664 + }, + { + "epoch": 0.5887084056142564, + "grad_norm": 1.32509446144104, + "learning_rate": 7.636725239132847e-06, + "loss": 0.431, + "step": 18665 + }, + { + "epoch": 0.588739946380697, + "grad_norm": 1.669433355331421, + "learning_rate": 7.63573263213713e-06, + "loss": 0.4328, + "step": 18666 + }, + { + "epoch": 0.5887714871471377, + "grad_norm": 1.3822474479675293, + "learning_rate": 7.634740049814376e-06, + "loss": 0.4585, + "step": 18667 + }, + { + "epoch": 0.5888030279135783, + "grad_norm": 1.949448585510254, + "learning_rate": 7.633747492174951e-06, + "loss": 0.4429, + "step": 18668 + }, + { + "epoch": 0.5888345686800189, + "grad_norm": 1.4901641607284546, + "learning_rate": 7.632754959229209e-06, + "loss": 0.7498, + "step": 18669 + }, + { + "epoch": 0.5888661094464596, + "grad_norm": 1.5536537170410156, + "learning_rate": 7.631762450987502e-06, + "loss": 0.4708, + "step": 18670 + }, + { + "epoch": 0.5888976502129002, + "grad_norm": 2.8910744190216064, + "learning_rate": 7.630769967460195e-06, + "loss": 0.451, + "step": 18671 + }, + { + "epoch": 0.5889291909793408, + "grad_norm": 1.846661925315857, + "learning_rate": 7.629777508657645e-06, + "loss": 0.4666, + "step": 18672 + }, + { + "epoch": 0.5889607317457815, + "grad_norm": 10.442809104919434, + "learning_rate": 7.628785074590206e-06, + "loss": 0.3827, + "step": 18673 + }, + { + "epoch": 0.5889922725122221, + "grad_norm": 1.524789810180664, + "learning_rate": 7.627792665268234e-06, + "loss": 0.4334, + "step": 18674 + }, + { + "epoch": 0.5890238132786627, + "grad_norm": 1.7081912755966187, + "learning_rate": 7.62680028070209e-06, + "loss": 0.401, + "step": 18675 + }, + { + "epoch": 0.5890553540451033, + "grad_norm": 2.7392730712890625, + "learning_rate": 7.625807920902126e-06, + "loss": 0.4287, + "step": 18676 + }, + { + "epoch": 0.589086894811544, + "grad_norm": 1.435920000076294, + "learning_rate": 7.624815585878702e-06, + "loss": 0.4435, + "step": 18677 + }, + { + "epoch": 0.5891184355779846, + "grad_norm": 1.6340223550796509, + "learning_rate": 7.623823275642168e-06, + "loss": 0.4759, + "step": 18678 + }, + { + "epoch": 0.5891499763444251, + "grad_norm": 1.5922714471817017, + "learning_rate": 7.622830990202887e-06, + "loss": 0.4239, + "step": 18679 + }, + { + "epoch": 0.5891815171108657, + "grad_norm": 2.023733139038086, + "learning_rate": 7.62183872957121e-06, + "loss": 0.4726, + "step": 18680 + }, + { + "epoch": 0.5892130578773064, + "grad_norm": 1.9371933937072754, + "learning_rate": 7.620846493757489e-06, + "loss": 0.4936, + "step": 18681 + }, + { + "epoch": 0.589244598643747, + "grad_norm": 1.417512059211731, + "learning_rate": 7.619854282772085e-06, + "loss": 0.4347, + "step": 18682 + }, + { + "epoch": 0.5892761394101876, + "grad_norm": 1.6625845432281494, + "learning_rate": 7.6188620966253515e-06, + "loss": 0.4286, + "step": 18683 + }, + { + "epoch": 0.5893076801766283, + "grad_norm": 5.143604278564453, + "learning_rate": 7.617869935327638e-06, + "loss": 0.3742, + "step": 18684 + }, + { + "epoch": 0.5893392209430689, + "grad_norm": 1.5350472927093506, + "learning_rate": 7.6168777988893024e-06, + "loss": 0.4223, + "step": 18685 + }, + { + "epoch": 0.5893707617095095, + "grad_norm": 1.4453775882720947, + "learning_rate": 7.615885687320699e-06, + "loss": 0.4374, + "step": 18686 + }, + { + "epoch": 0.5894023024759502, + "grad_norm": 2.215391159057617, + "learning_rate": 7.614893600632175e-06, + "loss": 0.4257, + "step": 18687 + }, + { + "epoch": 0.5894338432423908, + "grad_norm": 1.617960810661316, + "learning_rate": 7.613901538834092e-06, + "loss": 0.4902, + "step": 18688 + }, + { + "epoch": 0.5894653840088314, + "grad_norm": 1.484963297843933, + "learning_rate": 7.612909501936801e-06, + "loss": 0.467, + "step": 18689 + }, + { + "epoch": 0.589496924775272, + "grad_norm": 1.329437017440796, + "learning_rate": 7.611917489950646e-06, + "loss": 0.4414, + "step": 18690 + }, + { + "epoch": 0.5895284655417127, + "grad_norm": 1.58479905128479, + "learning_rate": 7.610925502885992e-06, + "loss": 0.4852, + "step": 18691 + }, + { + "epoch": 0.5895600063081533, + "grad_norm": 1.9472688436508179, + "learning_rate": 7.609933540753185e-06, + "loss": 0.4368, + "step": 18692 + }, + { + "epoch": 0.5895915470745939, + "grad_norm": 3.199610948562622, + "learning_rate": 7.608941603562576e-06, + "loss": 0.4129, + "step": 18693 + }, + { + "epoch": 0.5896230878410346, + "grad_norm": 1.4648149013519287, + "learning_rate": 7.6079496913245165e-06, + "loss": 0.3174, + "step": 18694 + }, + { + "epoch": 0.5896546286074752, + "grad_norm": 1.7637896537780762, + "learning_rate": 7.606957804049361e-06, + "loss": 0.456, + "step": 18695 + }, + { + "epoch": 0.5896861693739158, + "grad_norm": 1.4457533359527588, + "learning_rate": 7.60596594174746e-06, + "loss": 0.767, + "step": 18696 + }, + { + "epoch": 0.5897177101403565, + "grad_norm": 1.7346621751785278, + "learning_rate": 7.604974104429164e-06, + "loss": 0.438, + "step": 18697 + }, + { + "epoch": 0.5897492509067971, + "grad_norm": 1.4435925483703613, + "learning_rate": 7.603982292104818e-06, + "loss": 0.7303, + "step": 18698 + }, + { + "epoch": 0.5897807916732377, + "grad_norm": 1.5070093870162964, + "learning_rate": 7.602990504784782e-06, + "loss": 0.4155, + "step": 18699 + }, + { + "epoch": 0.5898123324396782, + "grad_norm": 1.6029367446899414, + "learning_rate": 7.601998742479402e-06, + "loss": 0.4086, + "step": 18700 + }, + { + "epoch": 0.5898438732061189, + "grad_norm": 1.4106082916259766, + "learning_rate": 7.601007005199022e-06, + "loss": 0.4467, + "step": 18701 + }, + { + "epoch": 0.5898754139725595, + "grad_norm": 1.4353525638580322, + "learning_rate": 7.600015292954e-06, + "loss": 0.4366, + "step": 18702 + }, + { + "epoch": 0.5899069547390001, + "grad_norm": 1.7096924781799316, + "learning_rate": 7.599023605754683e-06, + "loss": 0.481, + "step": 18703 + }, + { + "epoch": 0.5899384955054408, + "grad_norm": 1.4154795408248901, + "learning_rate": 7.598031943611417e-06, + "loss": 0.4453, + "step": 18704 + }, + { + "epoch": 0.5899700362718814, + "grad_norm": 1.597650170326233, + "learning_rate": 7.597040306534555e-06, + "loss": 0.4715, + "step": 18705 + }, + { + "epoch": 0.590001577038322, + "grad_norm": 1.4904041290283203, + "learning_rate": 7.596048694534443e-06, + "loss": 0.4585, + "step": 18706 + }, + { + "epoch": 0.5900331178047626, + "grad_norm": 2.05399227142334, + "learning_rate": 7.595057107621424e-06, + "loss": 0.4914, + "step": 18707 + }, + { + "epoch": 0.5900646585712033, + "grad_norm": 1.4039266109466553, + "learning_rate": 7.5940655458058575e-06, + "loss": 0.4649, + "step": 18708 + }, + { + "epoch": 0.5900961993376439, + "grad_norm": 1.476944923400879, + "learning_rate": 7.593074009098085e-06, + "loss": 0.449, + "step": 18709 + }, + { + "epoch": 0.5901277401040845, + "grad_norm": 1.4320546388626099, + "learning_rate": 7.5920824975084485e-06, + "loss": 0.5019, + "step": 18710 + }, + { + "epoch": 0.5901592808705252, + "grad_norm": 1.6206737756729126, + "learning_rate": 7.591091011047306e-06, + "loss": 0.4903, + "step": 18711 + }, + { + "epoch": 0.5901908216369658, + "grad_norm": 1.7929801940917969, + "learning_rate": 7.590099549724997e-06, + "loss": 0.4471, + "step": 18712 + }, + { + "epoch": 0.5902223624034064, + "grad_norm": 2.703787088394165, + "learning_rate": 7.589108113551873e-06, + "loss": 0.4498, + "step": 18713 + }, + { + "epoch": 0.590253903169847, + "grad_norm": 2.1052019596099854, + "learning_rate": 7.588116702538275e-06, + "loss": 0.461, + "step": 18714 + }, + { + "epoch": 0.5902854439362877, + "grad_norm": 1.4717371463775635, + "learning_rate": 7.587125316694552e-06, + "loss": 0.4547, + "step": 18715 + }, + { + "epoch": 0.5903169847027283, + "grad_norm": 11.72425651550293, + "learning_rate": 7.586133956031053e-06, + "loss": 0.7322, + "step": 18716 + }, + { + "epoch": 0.5903485254691689, + "grad_norm": 1.3736516237258911, + "learning_rate": 7.585142620558121e-06, + "loss": 0.4616, + "step": 18717 + }, + { + "epoch": 0.5903800662356096, + "grad_norm": 1.5307786464691162, + "learning_rate": 7.584151310286094e-06, + "loss": 0.4626, + "step": 18718 + }, + { + "epoch": 0.5904116070020502, + "grad_norm": 1.3719847202301025, + "learning_rate": 7.58316002522533e-06, + "loss": 0.4672, + "step": 18719 + }, + { + "epoch": 0.5904431477684908, + "grad_norm": 1.4823706150054932, + "learning_rate": 7.5821687653861675e-06, + "loss": 0.4634, + "step": 18720 + }, + { + "epoch": 0.5904746885349313, + "grad_norm": 1.6439253091812134, + "learning_rate": 7.581177530778947e-06, + "loss": 0.3392, + "step": 18721 + }, + { + "epoch": 0.590506229301372, + "grad_norm": 1.552304744720459, + "learning_rate": 7.580186321414022e-06, + "loss": 0.4544, + "step": 18722 + }, + { + "epoch": 0.5905377700678126, + "grad_norm": 1.7793560028076172, + "learning_rate": 7.579195137301732e-06, + "loss": 0.3667, + "step": 18723 + }, + { + "epoch": 0.5905693108342532, + "grad_norm": 1.6015758514404297, + "learning_rate": 7.578203978452417e-06, + "loss": 0.463, + "step": 18724 + }, + { + "epoch": 0.5906008516006939, + "grad_norm": 1.4172581434249878, + "learning_rate": 7.577212844876426e-06, + "loss": 0.7571, + "step": 18725 + }, + { + "epoch": 0.5906323923671345, + "grad_norm": 1.6053835153579712, + "learning_rate": 7.576221736584097e-06, + "loss": 0.4646, + "step": 18726 + }, + { + "epoch": 0.5906639331335751, + "grad_norm": 1.531498908996582, + "learning_rate": 7.575230653585779e-06, + "loss": 0.4011, + "step": 18727 + }, + { + "epoch": 0.5906954739000158, + "grad_norm": 7.923522472381592, + "learning_rate": 7.574239595891812e-06, + "loss": 0.4602, + "step": 18728 + }, + { + "epoch": 0.5907270146664564, + "grad_norm": 1.8329582214355469, + "learning_rate": 7.573248563512536e-06, + "loss": 0.4555, + "step": 18729 + }, + { + "epoch": 0.590758555432897, + "grad_norm": 1.5751028060913086, + "learning_rate": 7.5722575564582974e-06, + "loss": 0.4187, + "step": 18730 + }, + { + "epoch": 0.5907900961993376, + "grad_norm": 1.6644680500030518, + "learning_rate": 7.5712665747394385e-06, + "loss": 0.4311, + "step": 18731 + }, + { + "epoch": 0.5908216369657783, + "grad_norm": 1.6123110055923462, + "learning_rate": 7.5702756183662965e-06, + "loss": 0.4589, + "step": 18732 + }, + { + "epoch": 0.5908531777322189, + "grad_norm": 1.3922357559204102, + "learning_rate": 7.569284687349216e-06, + "loss": 0.4565, + "step": 18733 + }, + { + "epoch": 0.5908847184986595, + "grad_norm": 1.7199468612670898, + "learning_rate": 7.568293781698536e-06, + "loss": 0.4526, + "step": 18734 + }, + { + "epoch": 0.5909162592651002, + "grad_norm": 1.530057668685913, + "learning_rate": 7.567302901424598e-06, + "loss": 0.7256, + "step": 18735 + }, + { + "epoch": 0.5909478000315408, + "grad_norm": 2.231034994125366, + "learning_rate": 7.5663120465377445e-06, + "loss": 0.4771, + "step": 18736 + }, + { + "epoch": 0.5909793407979814, + "grad_norm": 1.7453081607818604, + "learning_rate": 7.565321217048316e-06, + "loss": 0.4435, + "step": 18737 + }, + { + "epoch": 0.591010881564422, + "grad_norm": 1.9763927459716797, + "learning_rate": 7.564330412966646e-06, + "loss": 0.4382, + "step": 18738 + }, + { + "epoch": 0.5910424223308627, + "grad_norm": 1.729355812072754, + "learning_rate": 7.563339634303084e-06, + "loss": 0.4522, + "step": 18739 + }, + { + "epoch": 0.5910739630973033, + "grad_norm": 1.7348806858062744, + "learning_rate": 7.562348881067964e-06, + "loss": 0.4334, + "step": 18740 + }, + { + "epoch": 0.5911055038637439, + "grad_norm": 3.542901039123535, + "learning_rate": 7.5613581532716226e-06, + "loss": 0.476, + "step": 18741 + }, + { + "epoch": 0.5911370446301845, + "grad_norm": 1.3664884567260742, + "learning_rate": 7.560367450924408e-06, + "loss": 0.4557, + "step": 18742 + }, + { + "epoch": 0.5911685853966251, + "grad_norm": 1.3076708316802979, + "learning_rate": 7.55937677403665e-06, + "loss": 0.7417, + "step": 18743 + }, + { + "epoch": 0.5912001261630657, + "grad_norm": 1.519515872001648, + "learning_rate": 7.558386122618693e-06, + "loss": 0.7246, + "step": 18744 + }, + { + "epoch": 0.5912316669295063, + "grad_norm": 1.3791767358779907, + "learning_rate": 7.557395496680872e-06, + "loss": 0.775, + "step": 18745 + }, + { + "epoch": 0.591263207695947, + "grad_norm": 1.5867815017700195, + "learning_rate": 7.556404896233521e-06, + "loss": 0.4592, + "step": 18746 + }, + { + "epoch": 0.5912947484623876, + "grad_norm": 2.3498952388763428, + "learning_rate": 7.555414321286987e-06, + "loss": 0.4762, + "step": 18747 + }, + { + "epoch": 0.5913262892288282, + "grad_norm": 1.7251704931259155, + "learning_rate": 7.5544237718516035e-06, + "loss": 0.3934, + "step": 18748 + }, + { + "epoch": 0.5913578299952689, + "grad_norm": 5.114163398742676, + "learning_rate": 7.553433247937702e-06, + "loss": 0.7371, + "step": 18749 + }, + { + "epoch": 0.5913893707617095, + "grad_norm": 1.4416780471801758, + "learning_rate": 7.552442749555629e-06, + "loss": 0.3357, + "step": 18750 + }, + { + "epoch": 0.5914209115281501, + "grad_norm": 1.8634049892425537, + "learning_rate": 7.551452276715716e-06, + "loss": 0.4597, + "step": 18751 + }, + { + "epoch": 0.5914524522945908, + "grad_norm": 1.7650624513626099, + "learning_rate": 7.5504618294282974e-06, + "loss": 0.4171, + "step": 18752 + }, + { + "epoch": 0.5914839930610314, + "grad_norm": 2.6784286499023438, + "learning_rate": 7.549471407703715e-06, + "loss": 0.4718, + "step": 18753 + }, + { + "epoch": 0.591515533827472, + "grad_norm": 1.4862674474716187, + "learning_rate": 7.5484810115522965e-06, + "loss": 0.4031, + "step": 18754 + }, + { + "epoch": 0.5915470745939126, + "grad_norm": 1.6513357162475586, + "learning_rate": 7.547490640984384e-06, + "loss": 0.453, + "step": 18755 + }, + { + "epoch": 0.5915786153603533, + "grad_norm": 1.9967151880264282, + "learning_rate": 7.5465002960103125e-06, + "loss": 0.4477, + "step": 18756 + }, + { + "epoch": 0.5916101561267939, + "grad_norm": 2.439114809036255, + "learning_rate": 7.545509976640413e-06, + "loss": 0.4313, + "step": 18757 + }, + { + "epoch": 0.5916416968932345, + "grad_norm": 1.9030202627182007, + "learning_rate": 7.544519682885025e-06, + "loss": 0.4429, + "step": 18758 + }, + { + "epoch": 0.5916732376596752, + "grad_norm": 1.7047874927520752, + "learning_rate": 7.543529414754481e-06, + "loss": 0.7438, + "step": 18759 + }, + { + "epoch": 0.5917047784261158, + "grad_norm": 1.827274203300476, + "learning_rate": 7.542539172259111e-06, + "loss": 0.4696, + "step": 18760 + }, + { + "epoch": 0.5917363191925564, + "grad_norm": 1.502010464668274, + "learning_rate": 7.541548955409258e-06, + "loss": 0.4632, + "step": 18761 + }, + { + "epoch": 0.5917678599589971, + "grad_norm": 9.137993812561035, + "learning_rate": 7.54055876421525e-06, + "loss": 0.4087, + "step": 18762 + }, + { + "epoch": 0.5917994007254376, + "grad_norm": 1.4063794612884521, + "learning_rate": 7.539568598687419e-06, + "loss": 0.4445, + "step": 18763 + }, + { + "epoch": 0.5918309414918782, + "grad_norm": 1.7021461725234985, + "learning_rate": 7.538578458836101e-06, + "loss": 0.4397, + "step": 18764 + }, + { + "epoch": 0.5918624822583188, + "grad_norm": 3.9168813228607178, + "learning_rate": 7.537588344671629e-06, + "loss": 0.3688, + "step": 18765 + }, + { + "epoch": 0.5918940230247595, + "grad_norm": 1.4873838424682617, + "learning_rate": 7.53659825620433e-06, + "loss": 0.4231, + "step": 18766 + }, + { + "epoch": 0.5919255637912001, + "grad_norm": 2.039060115814209, + "learning_rate": 7.5356081934445465e-06, + "loss": 0.4188, + "step": 18767 + }, + { + "epoch": 0.5919571045576407, + "grad_norm": 2.557997941970825, + "learning_rate": 7.534618156402604e-06, + "loss": 0.7145, + "step": 18768 + }, + { + "epoch": 0.5919886453240814, + "grad_norm": 4.080845832824707, + "learning_rate": 7.533628145088832e-06, + "loss": 0.435, + "step": 18769 + }, + { + "epoch": 0.592020186090522, + "grad_norm": 1.657008409500122, + "learning_rate": 7.5326381595135685e-06, + "loss": 0.435, + "step": 18770 + }, + { + "epoch": 0.5920517268569626, + "grad_norm": 1.5437082052230835, + "learning_rate": 7.531648199687143e-06, + "loss": 0.445, + "step": 18771 + }, + { + "epoch": 0.5920832676234032, + "grad_norm": 1.848848819732666, + "learning_rate": 7.530658265619881e-06, + "loss": 0.4408, + "step": 18772 + }, + { + "epoch": 0.5921148083898439, + "grad_norm": 1.3923227787017822, + "learning_rate": 7.529668357322121e-06, + "loss": 0.7215, + "step": 18773 + }, + { + "epoch": 0.5921463491562845, + "grad_norm": 1.398985505104065, + "learning_rate": 7.528678474804185e-06, + "loss": 0.4288, + "step": 18774 + }, + { + "epoch": 0.5921778899227251, + "grad_norm": 1.7532013654708862, + "learning_rate": 7.527688618076413e-06, + "loss": 0.4518, + "step": 18775 + }, + { + "epoch": 0.5922094306891658, + "grad_norm": 1.9796005487442017, + "learning_rate": 7.526698787149129e-06, + "loss": 0.7429, + "step": 18776 + }, + { + "epoch": 0.5922409714556064, + "grad_norm": 1.6269501447677612, + "learning_rate": 7.52570898203266e-06, + "loss": 0.4559, + "step": 18777 + }, + { + "epoch": 0.592272512222047, + "grad_norm": 1.5762057304382324, + "learning_rate": 7.524719202737344e-06, + "loss": 0.4694, + "step": 18778 + }, + { + "epoch": 0.5923040529884877, + "grad_norm": 1.2417712211608887, + "learning_rate": 7.5237294492735045e-06, + "loss": 0.4486, + "step": 18779 + }, + { + "epoch": 0.5923355937549283, + "grad_norm": 1.7745568752288818, + "learning_rate": 7.522739721651466e-06, + "loss": 0.4709, + "step": 18780 + }, + { + "epoch": 0.5923671345213689, + "grad_norm": 1.5693081617355347, + "learning_rate": 7.521750019881566e-06, + "loss": 0.4653, + "step": 18781 + }, + { + "epoch": 0.5923986752878095, + "grad_norm": 1.6171303987503052, + "learning_rate": 7.520760343974131e-06, + "loss": 0.4451, + "step": 18782 + }, + { + "epoch": 0.5924302160542502, + "grad_norm": 1.6286481618881226, + "learning_rate": 7.519770693939483e-06, + "loss": 0.4517, + "step": 18783 + }, + { + "epoch": 0.5924617568206907, + "grad_norm": 1.5483853816986084, + "learning_rate": 7.518781069787956e-06, + "loss": 0.3392, + "step": 18784 + }, + { + "epoch": 0.5924932975871313, + "grad_norm": 1.6227444410324097, + "learning_rate": 7.517791471529876e-06, + "loss": 0.4632, + "step": 18785 + }, + { + "epoch": 0.592524838353572, + "grad_norm": 2.3972420692443848, + "learning_rate": 7.5168018991755645e-06, + "loss": 0.4544, + "step": 18786 + }, + { + "epoch": 0.5925563791200126, + "grad_norm": 1.4935147762298584, + "learning_rate": 7.515812352735357e-06, + "loss": 0.4409, + "step": 18787 + }, + { + "epoch": 0.5925879198864532, + "grad_norm": 1.5726503133773804, + "learning_rate": 7.5148228322195775e-06, + "loss": 0.4726, + "step": 18788 + }, + { + "epoch": 0.5926194606528938, + "grad_norm": 4.066304683685303, + "learning_rate": 7.513833337638546e-06, + "loss": 0.4387, + "step": 18789 + }, + { + "epoch": 0.5926510014193345, + "grad_norm": 1.3754334449768066, + "learning_rate": 7.5128438690026e-06, + "loss": 0.4356, + "step": 18790 + }, + { + "epoch": 0.5926825421857751, + "grad_norm": 1.4575153589248657, + "learning_rate": 7.511854426322056e-06, + "loss": 0.4232, + "step": 18791 + }, + { + "epoch": 0.5927140829522157, + "grad_norm": 1.6411778926849365, + "learning_rate": 7.510865009607246e-06, + "loss": 0.4759, + "step": 18792 + }, + { + "epoch": 0.5927456237186564, + "grad_norm": 1.5348289012908936, + "learning_rate": 7.509875618868492e-06, + "loss": 0.463, + "step": 18793 + }, + { + "epoch": 0.592777164485097, + "grad_norm": 1.4561036825180054, + "learning_rate": 7.508886254116115e-06, + "loss": 0.4419, + "step": 18794 + }, + { + "epoch": 0.5928087052515376, + "grad_norm": 1.8953816890716553, + "learning_rate": 7.507896915360449e-06, + "loss": 0.4663, + "step": 18795 + }, + { + "epoch": 0.5928402460179782, + "grad_norm": 1.403009295463562, + "learning_rate": 7.506907602611814e-06, + "loss": 0.4303, + "step": 18796 + }, + { + "epoch": 0.5928717867844189, + "grad_norm": 0.9402862191200256, + "learning_rate": 7.505918315880529e-06, + "loss": 0.3352, + "step": 18797 + }, + { + "epoch": 0.5929033275508595, + "grad_norm": 1.1620584726333618, + "learning_rate": 7.504929055176927e-06, + "loss": 0.3433, + "step": 18798 + }, + { + "epoch": 0.5929348683173001, + "grad_norm": 1.478582739830017, + "learning_rate": 7.503939820511328e-06, + "loss": 0.445, + "step": 18799 + }, + { + "epoch": 0.5929664090837408, + "grad_norm": 1.7678024768829346, + "learning_rate": 7.5029506118940506e-06, + "loss": 0.4631, + "step": 18800 + }, + { + "epoch": 0.5929979498501814, + "grad_norm": 1.791858196258545, + "learning_rate": 7.5019614293354285e-06, + "loss": 0.463, + "step": 18801 + }, + { + "epoch": 0.593029490616622, + "grad_norm": 1.8058050870895386, + "learning_rate": 7.500972272845777e-06, + "loss": 0.4332, + "step": 18802 + }, + { + "epoch": 0.5930610313830627, + "grad_norm": 1.4970431327819824, + "learning_rate": 7.499983142435418e-06, + "loss": 0.4579, + "step": 18803 + }, + { + "epoch": 0.5930925721495033, + "grad_norm": 1.526116967201233, + "learning_rate": 7.498994038114678e-06, + "loss": 0.3351, + "step": 18804 + }, + { + "epoch": 0.5931241129159439, + "grad_norm": 1.5096296072006226, + "learning_rate": 7.498004959893875e-06, + "loss": 0.4758, + "step": 18805 + }, + { + "epoch": 0.5931556536823844, + "grad_norm": 1.5607362985610962, + "learning_rate": 7.497015907783335e-06, + "loss": 0.4764, + "step": 18806 + }, + { + "epoch": 0.5931871944488251, + "grad_norm": 1.6946321725845337, + "learning_rate": 7.496026881793378e-06, + "loss": 0.4617, + "step": 18807 + }, + { + "epoch": 0.5932187352152657, + "grad_norm": 2.0898540019989014, + "learning_rate": 7.495037881934321e-06, + "loss": 0.3456, + "step": 18808 + }, + { + "epoch": 0.5932502759817063, + "grad_norm": 1.9486804008483887, + "learning_rate": 7.494048908216494e-06, + "loss": 0.4491, + "step": 18809 + }, + { + "epoch": 0.593281816748147, + "grad_norm": 1.8759377002716064, + "learning_rate": 7.493059960650211e-06, + "loss": 0.455, + "step": 18810 + }, + { + "epoch": 0.5933133575145876, + "grad_norm": 1.3346551656723022, + "learning_rate": 7.4920710392457915e-06, + "loss": 0.4254, + "step": 18811 + }, + { + "epoch": 0.5933448982810282, + "grad_norm": 1.369626522064209, + "learning_rate": 7.491082144013561e-06, + "loss": 0.4597, + "step": 18812 + }, + { + "epoch": 0.5933764390474688, + "grad_norm": 1.8680589199066162, + "learning_rate": 7.490093274963836e-06, + "loss": 0.3979, + "step": 18813 + }, + { + "epoch": 0.5934079798139095, + "grad_norm": 1.8644554615020752, + "learning_rate": 7.489104432106933e-06, + "loss": 0.4664, + "step": 18814 + }, + { + "epoch": 0.5934395205803501, + "grad_norm": 1.6201914548873901, + "learning_rate": 7.488115615453178e-06, + "loss": 0.4652, + "step": 18815 + }, + { + "epoch": 0.5934710613467907, + "grad_norm": 2.0260252952575684, + "learning_rate": 7.487126825012887e-06, + "loss": 0.4695, + "step": 18816 + }, + { + "epoch": 0.5935026021132314, + "grad_norm": 1.523054838180542, + "learning_rate": 7.486138060796375e-06, + "loss": 0.4301, + "step": 18817 + }, + { + "epoch": 0.593534142879672, + "grad_norm": 1.9596376419067383, + "learning_rate": 7.4851493228139695e-06, + "loss": 0.4369, + "step": 18818 + }, + { + "epoch": 0.5935656836461126, + "grad_norm": 1.9296202659606934, + "learning_rate": 7.484160611075984e-06, + "loss": 0.4449, + "step": 18819 + }, + { + "epoch": 0.5935972244125532, + "grad_norm": 1.2784454822540283, + "learning_rate": 7.483171925592729e-06, + "loss": 0.4438, + "step": 18820 + }, + { + "epoch": 0.5936287651789939, + "grad_norm": 1.5845391750335693, + "learning_rate": 7.482183266374535e-06, + "loss": 0.4424, + "step": 18821 + }, + { + "epoch": 0.5936603059454345, + "grad_norm": 1.8423985242843628, + "learning_rate": 7.481194633431712e-06, + "loss": 0.4328, + "step": 18822 + }, + { + "epoch": 0.5936918467118751, + "grad_norm": 1.6827738285064697, + "learning_rate": 7.48020602677458e-06, + "loss": 0.4669, + "step": 18823 + }, + { + "epoch": 0.5937233874783158, + "grad_norm": 1.4001004695892334, + "learning_rate": 7.479217446413456e-06, + "loss": 0.4433, + "step": 18824 + }, + { + "epoch": 0.5937549282447564, + "grad_norm": 1.6676093339920044, + "learning_rate": 7.47822889235865e-06, + "loss": 0.4251, + "step": 18825 + }, + { + "epoch": 0.593786469011197, + "grad_norm": 1.7141929864883423, + "learning_rate": 7.4772403646204885e-06, + "loss": 0.4683, + "step": 18826 + }, + { + "epoch": 0.5938180097776375, + "grad_norm": 1.5032641887664795, + "learning_rate": 7.476251863209281e-06, + "loss": 0.4258, + "step": 18827 + }, + { + "epoch": 0.5938495505440782, + "grad_norm": 1.8799190521240234, + "learning_rate": 7.475263388135343e-06, + "loss": 0.4072, + "step": 18828 + }, + { + "epoch": 0.5938810913105188, + "grad_norm": 1.674827218055725, + "learning_rate": 7.4742749394089945e-06, + "loss": 0.4423, + "step": 18829 + }, + { + "epoch": 0.5939126320769594, + "grad_norm": 1.489774227142334, + "learning_rate": 7.4732865170405485e-06, + "loss": 0.4647, + "step": 18830 + }, + { + "epoch": 0.5939441728434001, + "grad_norm": 1.5835869312286377, + "learning_rate": 7.472298121040318e-06, + "loss": 0.4507, + "step": 18831 + }, + { + "epoch": 0.5939757136098407, + "grad_norm": 1.9949086904525757, + "learning_rate": 7.471309751418621e-06, + "loss": 0.4422, + "step": 18832 + }, + { + "epoch": 0.5940072543762813, + "grad_norm": 1.867472529411316, + "learning_rate": 7.470321408185771e-06, + "loss": 0.7355, + "step": 18833 + }, + { + "epoch": 0.594038795142722, + "grad_norm": 1.5251686573028564, + "learning_rate": 7.469333091352076e-06, + "loss": 0.4611, + "step": 18834 + }, + { + "epoch": 0.5940703359091626, + "grad_norm": 1.7447725534439087, + "learning_rate": 7.46834480092786e-06, + "loss": 0.4543, + "step": 18835 + }, + { + "epoch": 0.5941018766756032, + "grad_norm": 1.4622266292572021, + "learning_rate": 7.4673565369234325e-06, + "loss": 0.4752, + "step": 18836 + }, + { + "epoch": 0.5941334174420438, + "grad_norm": 3.8333089351654053, + "learning_rate": 7.4663682993491e-06, + "loss": 0.4074, + "step": 18837 + }, + { + "epoch": 0.5941649582084845, + "grad_norm": 1.7135769128799438, + "learning_rate": 7.465380088215188e-06, + "loss": 0.4638, + "step": 18838 + }, + { + "epoch": 0.5941964989749251, + "grad_norm": 1.8173556327819824, + "learning_rate": 7.464391903532e-06, + "loss": 0.4704, + "step": 18839 + }, + { + "epoch": 0.5942280397413657, + "grad_norm": 1.6057393550872803, + "learning_rate": 7.4634037453098505e-06, + "loss": 0.77, + "step": 18840 + }, + { + "epoch": 0.5942595805078064, + "grad_norm": 1.5766804218292236, + "learning_rate": 7.462415613559055e-06, + "loss": 0.4687, + "step": 18841 + }, + { + "epoch": 0.594291121274247, + "grad_norm": 1.7092198133468628, + "learning_rate": 7.461427508289922e-06, + "loss": 0.4694, + "step": 18842 + }, + { + "epoch": 0.5943226620406876, + "grad_norm": 1.7131190299987793, + "learning_rate": 7.4604394295127645e-06, + "loss": 0.4605, + "step": 18843 + }, + { + "epoch": 0.5943542028071283, + "grad_norm": 1.730930209159851, + "learning_rate": 7.459451377237894e-06, + "loss": 0.4515, + "step": 18844 + }, + { + "epoch": 0.5943857435735689, + "grad_norm": 1.4907366037368774, + "learning_rate": 7.458463351475618e-06, + "loss": 0.4189, + "step": 18845 + }, + { + "epoch": 0.5944172843400095, + "grad_norm": 2.8192636966705322, + "learning_rate": 7.457475352236254e-06, + "loss": 0.752, + "step": 18846 + }, + { + "epoch": 0.5944488251064501, + "grad_norm": 1.3283052444458008, + "learning_rate": 7.4564873795301085e-06, + "loss": 0.4415, + "step": 18847 + }, + { + "epoch": 0.5944803658728907, + "grad_norm": 1.8368662595748901, + "learning_rate": 7.455499433367488e-06, + "loss": 0.4995, + "step": 18848 + }, + { + "epoch": 0.5945119066393313, + "grad_norm": 1.5075968503952026, + "learning_rate": 7.454511513758711e-06, + "loss": 0.4697, + "step": 18849 + }, + { + "epoch": 0.5945434474057719, + "grad_norm": 1.8116352558135986, + "learning_rate": 7.453523620714082e-06, + "loss": 0.4629, + "step": 18850 + }, + { + "epoch": 0.5945749881722125, + "grad_norm": 1.5166839361190796, + "learning_rate": 7.45253575424391e-06, + "loss": 0.4332, + "step": 18851 + }, + { + "epoch": 0.5946065289386532, + "grad_norm": 1.7474737167358398, + "learning_rate": 7.451547914358508e-06, + "loss": 0.461, + "step": 18852 + }, + { + "epoch": 0.5946380697050938, + "grad_norm": 1.5874792337417603, + "learning_rate": 7.450560101068178e-06, + "loss": 0.444, + "step": 18853 + }, + { + "epoch": 0.5946696104715344, + "grad_norm": 1.6512210369110107, + "learning_rate": 7.449572314383237e-06, + "loss": 0.4381, + "step": 18854 + }, + { + "epoch": 0.5947011512379751, + "grad_norm": 1.5798206329345703, + "learning_rate": 7.448584554313989e-06, + "loss": 0.4501, + "step": 18855 + }, + { + "epoch": 0.5947326920044157, + "grad_norm": 1.5883508920669556, + "learning_rate": 7.447596820870738e-06, + "loss": 0.464, + "step": 18856 + }, + { + "epoch": 0.5947642327708563, + "grad_norm": 2.1489429473876953, + "learning_rate": 7.446609114063801e-06, + "loss": 0.4472, + "step": 18857 + }, + { + "epoch": 0.594795773537297, + "grad_norm": 1.6268396377563477, + "learning_rate": 7.4456214339034805e-06, + "loss": 0.4562, + "step": 18858 + }, + { + "epoch": 0.5948273143037376, + "grad_norm": 2.203934907913208, + "learning_rate": 7.444633780400082e-06, + "loss": 0.462, + "step": 18859 + }, + { + "epoch": 0.5948588550701782, + "grad_norm": 2.9840266704559326, + "learning_rate": 7.443646153563912e-06, + "loss": 0.4463, + "step": 18860 + }, + { + "epoch": 0.5948903958366188, + "grad_norm": 1.1359431743621826, + "learning_rate": 7.442658553405282e-06, + "loss": 0.331, + "step": 18861 + }, + { + "epoch": 0.5949219366030595, + "grad_norm": 1.8919631242752075, + "learning_rate": 7.441670979934494e-06, + "loss": 0.4093, + "step": 18862 + }, + { + "epoch": 0.5949534773695001, + "grad_norm": 1.7048224210739136, + "learning_rate": 7.44068343316186e-06, + "loss": 0.4525, + "step": 18863 + }, + { + "epoch": 0.5949850181359407, + "grad_norm": 1.6402406692504883, + "learning_rate": 7.43969591309768e-06, + "loss": 0.444, + "step": 18864 + }, + { + "epoch": 0.5950165589023814, + "grad_norm": 1.9130067825317383, + "learning_rate": 7.4387084197522565e-06, + "loss": 0.4856, + "step": 18865 + }, + { + "epoch": 0.595048099668822, + "grad_norm": 1.4879543781280518, + "learning_rate": 7.437720953135904e-06, + "loss": 0.515, + "step": 18866 + }, + { + "epoch": 0.5950796404352626, + "grad_norm": 1.9051426649093628, + "learning_rate": 7.436733513258922e-06, + "loss": 0.4467, + "step": 18867 + }, + { + "epoch": 0.5951111812017033, + "grad_norm": 1.710949420928955, + "learning_rate": 7.435746100131613e-06, + "loss": 0.453, + "step": 18868 + }, + { + "epoch": 0.5951427219681438, + "grad_norm": 1.5521847009658813, + "learning_rate": 7.43475871376429e-06, + "loss": 0.4208, + "step": 18869 + }, + { + "epoch": 0.5951742627345844, + "grad_norm": 1.938543438911438, + "learning_rate": 7.4337713541672474e-06, + "loss": 0.7412, + "step": 18870 + }, + { + "epoch": 0.595205803501025, + "grad_norm": 1.99422287940979, + "learning_rate": 7.432784021350796e-06, + "loss": 0.337, + "step": 18871 + }, + { + "epoch": 0.5952373442674657, + "grad_norm": 1.560957670211792, + "learning_rate": 7.4317967153252376e-06, + "loss": 0.4584, + "step": 18872 + }, + { + "epoch": 0.5952688850339063, + "grad_norm": 1.513384461402893, + "learning_rate": 7.43080943610087e-06, + "loss": 0.4806, + "step": 18873 + }, + { + "epoch": 0.5953004258003469, + "grad_norm": 1.8654009103775024, + "learning_rate": 7.429822183688004e-06, + "loss": 0.4142, + "step": 18874 + }, + { + "epoch": 0.5953319665667876, + "grad_norm": 1.4164637327194214, + "learning_rate": 7.428834958096942e-06, + "loss": 0.4487, + "step": 18875 + }, + { + "epoch": 0.5953635073332282, + "grad_norm": 1.1788725852966309, + "learning_rate": 7.4278477593379776e-06, + "loss": 0.3603, + "step": 18876 + }, + { + "epoch": 0.5953950480996688, + "grad_norm": 1.3616209030151367, + "learning_rate": 7.426860587421423e-06, + "loss": 0.4372, + "step": 18877 + }, + { + "epoch": 0.5954265888661094, + "grad_norm": 2.567263126373291, + "learning_rate": 7.425873442357576e-06, + "loss": 0.3953, + "step": 18878 + }, + { + "epoch": 0.5954581296325501, + "grad_norm": 1.6590536832809448, + "learning_rate": 7.424886324156737e-06, + "loss": 0.372, + "step": 18879 + }, + { + "epoch": 0.5954896703989907, + "grad_norm": 1.162326455116272, + "learning_rate": 7.423899232829208e-06, + "loss": 0.3167, + "step": 18880 + }, + { + "epoch": 0.5955212111654313, + "grad_norm": 1.3111335039138794, + "learning_rate": 7.422912168385293e-06, + "loss": 0.4619, + "step": 18881 + }, + { + "epoch": 0.595552751931872, + "grad_norm": 1.8526109457015991, + "learning_rate": 7.421925130835289e-06, + "loss": 0.448, + "step": 18882 + }, + { + "epoch": 0.5955842926983126, + "grad_norm": 1.6219183206558228, + "learning_rate": 7.4209381201895005e-06, + "loss": 0.4577, + "step": 18883 + }, + { + "epoch": 0.5956158334647532, + "grad_norm": 1.4424147605895996, + "learning_rate": 7.419951136458225e-06, + "loss": 0.4557, + "step": 18884 + }, + { + "epoch": 0.5956473742311938, + "grad_norm": 1.4551061391830444, + "learning_rate": 7.418964179651757e-06, + "loss": 0.4283, + "step": 18885 + }, + { + "epoch": 0.5956789149976345, + "grad_norm": 1.4257079362869263, + "learning_rate": 7.417977249780408e-06, + "loss": 0.4313, + "step": 18886 + }, + { + "epoch": 0.5957104557640751, + "grad_norm": 1.4498182535171509, + "learning_rate": 7.416990346854467e-06, + "loss": 0.4359, + "step": 18887 + }, + { + "epoch": 0.5957419965305157, + "grad_norm": 1.7189788818359375, + "learning_rate": 7.416003470884241e-06, + "loss": 0.4199, + "step": 18888 + }, + { + "epoch": 0.5957735372969564, + "grad_norm": 1.8083425760269165, + "learning_rate": 7.415016621880026e-06, + "loss": 0.4596, + "step": 18889 + }, + { + "epoch": 0.5958050780633969, + "grad_norm": 1.812244176864624, + "learning_rate": 7.414029799852116e-06, + "loss": 0.4655, + "step": 18890 + }, + { + "epoch": 0.5958366188298375, + "grad_norm": 1.411608099937439, + "learning_rate": 7.413043004810816e-06, + "loss": 0.4591, + "step": 18891 + }, + { + "epoch": 0.5958681595962781, + "grad_norm": 1.318038821220398, + "learning_rate": 7.4120562367664215e-06, + "loss": 0.4618, + "step": 18892 + }, + { + "epoch": 0.5958997003627188, + "grad_norm": 1.927075982093811, + "learning_rate": 7.411069495729225e-06, + "loss": 0.4335, + "step": 18893 + }, + { + "epoch": 0.5959312411291594, + "grad_norm": 1.5612828731536865, + "learning_rate": 7.410082781709533e-06, + "loss": 0.488, + "step": 18894 + }, + { + "epoch": 0.5959627818956, + "grad_norm": 2.147196054458618, + "learning_rate": 7.409096094717638e-06, + "loss": 0.4864, + "step": 18895 + }, + { + "epoch": 0.5959943226620407, + "grad_norm": 1.6607083082199097, + "learning_rate": 7.408109434763833e-06, + "loss": 0.3806, + "step": 18896 + }, + { + "epoch": 0.5960258634284813, + "grad_norm": 2.0959134101867676, + "learning_rate": 7.407122801858422e-06, + "loss": 0.4702, + "step": 18897 + }, + { + "epoch": 0.5960574041949219, + "grad_norm": 1.731289267539978, + "learning_rate": 7.406136196011698e-06, + "loss": 0.4341, + "step": 18898 + }, + { + "epoch": 0.5960889449613626, + "grad_norm": 2.0415101051330566, + "learning_rate": 7.405149617233955e-06, + "loss": 0.4613, + "step": 18899 + }, + { + "epoch": 0.5961204857278032, + "grad_norm": 1.390498161315918, + "learning_rate": 7.404163065535492e-06, + "loss": 0.4657, + "step": 18900 + }, + { + "epoch": 0.5961520264942438, + "grad_norm": 1.4009594917297363, + "learning_rate": 7.403176540926602e-06, + "loss": 0.7475, + "step": 18901 + }, + { + "epoch": 0.5961835672606844, + "grad_norm": 1.4785772562026978, + "learning_rate": 7.4021900434175835e-06, + "loss": 0.4581, + "step": 18902 + }, + { + "epoch": 0.5962151080271251, + "grad_norm": 1.4189441204071045, + "learning_rate": 7.401203573018727e-06, + "loss": 0.4486, + "step": 18903 + }, + { + "epoch": 0.5962466487935657, + "grad_norm": 1.7023392915725708, + "learning_rate": 7.400217129740328e-06, + "loss": 0.4537, + "step": 18904 + }, + { + "epoch": 0.5962781895600063, + "grad_norm": 1.6416975259780884, + "learning_rate": 7.399230713592684e-06, + "loss": 0.4557, + "step": 18905 + }, + { + "epoch": 0.596309730326447, + "grad_norm": 6.869938850402832, + "learning_rate": 7.398244324586088e-06, + "loss": 0.752, + "step": 18906 + }, + { + "epoch": 0.5963412710928876, + "grad_norm": 2.045175790786743, + "learning_rate": 7.3972579627308286e-06, + "loss": 0.4758, + "step": 18907 + }, + { + "epoch": 0.5963728118593282, + "grad_norm": 1.3052953481674194, + "learning_rate": 7.396271628037206e-06, + "loss": 0.4566, + "step": 18908 + }, + { + "epoch": 0.5964043526257689, + "grad_norm": 1.604786992073059, + "learning_rate": 7.395285320515513e-06, + "loss": 0.4944, + "step": 18909 + }, + { + "epoch": 0.5964358933922095, + "grad_norm": 1.8946131467819214, + "learning_rate": 7.394299040176036e-06, + "loss": 0.4258, + "step": 18910 + }, + { + "epoch": 0.59646743415865, + "grad_norm": 1.2510088682174683, + "learning_rate": 7.393312787029074e-06, + "loss": 0.3473, + "step": 18911 + }, + { + "epoch": 0.5964989749250906, + "grad_norm": 4.620795249938965, + "learning_rate": 7.392326561084919e-06, + "loss": 0.4582, + "step": 18912 + }, + { + "epoch": 0.5965305156915313, + "grad_norm": 1.6732233762741089, + "learning_rate": 7.391340362353856e-06, + "loss": 0.4195, + "step": 18913 + }, + { + "epoch": 0.5965620564579719, + "grad_norm": 1.685774326324463, + "learning_rate": 7.390354190846185e-06, + "loss": 0.4491, + "step": 18914 + }, + { + "epoch": 0.5965935972244125, + "grad_norm": 11.3991117477417, + "learning_rate": 7.3893680465721964e-06, + "loss": 0.4675, + "step": 18915 + }, + { + "epoch": 0.5966251379908531, + "grad_norm": 1.4414464235305786, + "learning_rate": 7.388381929542173e-06, + "loss": 0.4608, + "step": 18916 + }, + { + "epoch": 0.5966566787572938, + "grad_norm": 3.9965169429779053, + "learning_rate": 7.387395839766418e-06, + "loss": 0.4829, + "step": 18917 + }, + { + "epoch": 0.5966882195237344, + "grad_norm": 1.5586419105529785, + "learning_rate": 7.386409777255212e-06, + "loss": 0.4322, + "step": 18918 + }, + { + "epoch": 0.596719760290175, + "grad_norm": 1.6321337223052979, + "learning_rate": 7.385423742018854e-06, + "loss": 0.3878, + "step": 18919 + }, + { + "epoch": 0.5967513010566157, + "grad_norm": 2.095346212387085, + "learning_rate": 7.384437734067625e-06, + "loss": 0.4481, + "step": 18920 + }, + { + "epoch": 0.5967828418230563, + "grad_norm": 1.7070810794830322, + "learning_rate": 7.383451753411821e-06, + "loss": 0.4562, + "step": 18921 + }, + { + "epoch": 0.5968143825894969, + "grad_norm": 1.4385055303573608, + "learning_rate": 7.382465800061731e-06, + "loss": 0.4822, + "step": 18922 + }, + { + "epoch": 0.5968459233559376, + "grad_norm": 1.698762059211731, + "learning_rate": 7.381479874027643e-06, + "loss": 0.4157, + "step": 18923 + }, + { + "epoch": 0.5968774641223782, + "grad_norm": 1.9313459396362305, + "learning_rate": 7.38049397531984e-06, + "loss": 0.3925, + "step": 18924 + }, + { + "epoch": 0.5969090048888188, + "grad_norm": 1.565125823020935, + "learning_rate": 7.379508103948622e-06, + "loss": 0.4311, + "step": 18925 + }, + { + "epoch": 0.5969405456552594, + "grad_norm": 1.638055443763733, + "learning_rate": 7.378522259924272e-06, + "loss": 0.4682, + "step": 18926 + }, + { + "epoch": 0.5969720864217001, + "grad_norm": 2.0784149169921875, + "learning_rate": 7.377536443257074e-06, + "loss": 0.4244, + "step": 18927 + }, + { + "epoch": 0.5970036271881407, + "grad_norm": 1.6352778673171997, + "learning_rate": 7.376550653957322e-06, + "loss": 0.4403, + "step": 18928 + }, + { + "epoch": 0.5970351679545813, + "grad_norm": 1.5471367835998535, + "learning_rate": 7.375564892035302e-06, + "loss": 0.4301, + "step": 18929 + }, + { + "epoch": 0.597066708721022, + "grad_norm": 1.5014492273330688, + "learning_rate": 7.374579157501298e-06, + "loss": 0.4467, + "step": 18930 + }, + { + "epoch": 0.5970982494874626, + "grad_norm": 2.1320526599884033, + "learning_rate": 7.373593450365601e-06, + "loss": 0.4203, + "step": 18931 + }, + { + "epoch": 0.5971297902539031, + "grad_norm": 1.6958905458450317, + "learning_rate": 7.3726077706384955e-06, + "loss": 0.4432, + "step": 18932 + }, + { + "epoch": 0.5971613310203437, + "grad_norm": 2.8269946575164795, + "learning_rate": 7.371622118330264e-06, + "loss": 0.3374, + "step": 18933 + }, + { + "epoch": 0.5971928717867844, + "grad_norm": 1.552927017211914, + "learning_rate": 7.370636493451201e-06, + "loss": 0.748, + "step": 18934 + }, + { + "epoch": 0.597224412553225, + "grad_norm": 1.702090859413147, + "learning_rate": 7.3696508960115846e-06, + "loss": 0.4822, + "step": 18935 + }, + { + "epoch": 0.5972559533196656, + "grad_norm": 1.6210153102874756, + "learning_rate": 7.368665326021705e-06, + "loss": 0.7833, + "step": 18936 + }, + { + "epoch": 0.5972874940861063, + "grad_norm": 2.010152578353882, + "learning_rate": 7.367679783491848e-06, + "loss": 0.4821, + "step": 18937 + }, + { + "epoch": 0.5973190348525469, + "grad_norm": 1.6302238702774048, + "learning_rate": 7.366694268432294e-06, + "loss": 0.4113, + "step": 18938 + }, + { + "epoch": 0.5973505756189875, + "grad_norm": 1.8674628734588623, + "learning_rate": 7.365708780853331e-06, + "loss": 0.4515, + "step": 18939 + }, + { + "epoch": 0.5973821163854282, + "grad_norm": 1.8493640422821045, + "learning_rate": 7.364723320765242e-06, + "loss": 0.4658, + "step": 18940 + }, + { + "epoch": 0.5974136571518688, + "grad_norm": 1.2645186185836792, + "learning_rate": 7.36373788817831e-06, + "loss": 0.3729, + "step": 18941 + }, + { + "epoch": 0.5974451979183094, + "grad_norm": 1.7465009689331055, + "learning_rate": 7.3627524831028245e-06, + "loss": 0.7251, + "step": 18942 + }, + { + "epoch": 0.59747673868475, + "grad_norm": 2.360776901245117, + "learning_rate": 7.361767105549063e-06, + "loss": 0.4392, + "step": 18943 + }, + { + "epoch": 0.5975082794511907, + "grad_norm": 2.2797188758850098, + "learning_rate": 7.360781755527304e-06, + "loss": 0.4549, + "step": 18944 + }, + { + "epoch": 0.5975398202176313, + "grad_norm": 1.8890541791915894, + "learning_rate": 7.359796433047843e-06, + "loss": 0.4536, + "step": 18945 + }, + { + "epoch": 0.5975713609840719, + "grad_norm": 2.177062511444092, + "learning_rate": 7.358811138120956e-06, + "loss": 0.4403, + "step": 18946 + }, + { + "epoch": 0.5976029017505126, + "grad_norm": 2.108665943145752, + "learning_rate": 7.357825870756923e-06, + "loss": 0.7643, + "step": 18947 + }, + { + "epoch": 0.5976344425169532, + "grad_norm": 1.805604338645935, + "learning_rate": 7.3568406309660316e-06, + "loss": 0.414, + "step": 18948 + }, + { + "epoch": 0.5976659832833938, + "grad_norm": 1.6574121713638306, + "learning_rate": 7.355855418758557e-06, + "loss": 0.3642, + "step": 18949 + }, + { + "epoch": 0.5976975240498344, + "grad_norm": 1.8476133346557617, + "learning_rate": 7.354870234144788e-06, + "loss": 0.4692, + "step": 18950 + }, + { + "epoch": 0.5977290648162751, + "grad_norm": 4.349087238311768, + "learning_rate": 7.353885077135001e-06, + "loss": 0.4641, + "step": 18951 + }, + { + "epoch": 0.5977606055827157, + "grad_norm": 1.3996063470840454, + "learning_rate": 7.3528999477394744e-06, + "loss": 0.4765, + "step": 18952 + }, + { + "epoch": 0.5977921463491563, + "grad_norm": 1.5733873844146729, + "learning_rate": 7.351914845968498e-06, + "loss": 0.4523, + "step": 18953 + }, + { + "epoch": 0.5978236871155969, + "grad_norm": 1.7433691024780273, + "learning_rate": 7.350929771832344e-06, + "loss": 0.4486, + "step": 18954 + }, + { + "epoch": 0.5978552278820375, + "grad_norm": 2.790435314178467, + "learning_rate": 7.349944725341292e-06, + "loss": 0.452, + "step": 18955 + }, + { + "epoch": 0.5978867686484781, + "grad_norm": 1.6035512685775757, + "learning_rate": 7.3489597065056274e-06, + "loss": 0.345, + "step": 18956 + }, + { + "epoch": 0.5979183094149187, + "grad_norm": 1.6498312950134277, + "learning_rate": 7.347974715335626e-06, + "loss": 0.4047, + "step": 18957 + }, + { + "epoch": 0.5979498501813594, + "grad_norm": 1.7630705833435059, + "learning_rate": 7.346989751841568e-06, + "loss": 0.4427, + "step": 18958 + }, + { + "epoch": 0.5979813909478, + "grad_norm": 2.633284330368042, + "learning_rate": 7.346004816033732e-06, + "loss": 0.4062, + "step": 18959 + }, + { + "epoch": 0.5980129317142406, + "grad_norm": 1.422422170639038, + "learning_rate": 7.345019907922396e-06, + "loss": 0.4609, + "step": 18960 + }, + { + "epoch": 0.5980444724806813, + "grad_norm": 1.7078417539596558, + "learning_rate": 7.344035027517838e-06, + "loss": 0.4177, + "step": 18961 + }, + { + "epoch": 0.5980760132471219, + "grad_norm": 1.6249964237213135, + "learning_rate": 7.3430501748303396e-06, + "loss": 0.429, + "step": 18962 + }, + { + "epoch": 0.5981075540135625, + "grad_norm": 1.6969317197799683, + "learning_rate": 7.3420653498701755e-06, + "loss": 0.4655, + "step": 18963 + }, + { + "epoch": 0.5981390947800032, + "grad_norm": 1.8070565462112427, + "learning_rate": 7.341080552647618e-06, + "loss": 0.4509, + "step": 18964 + }, + { + "epoch": 0.5981706355464438, + "grad_norm": 1.9990222454071045, + "learning_rate": 7.340095783172955e-06, + "loss": 0.4577, + "step": 18965 + }, + { + "epoch": 0.5982021763128844, + "grad_norm": 2.1341323852539062, + "learning_rate": 7.339111041456453e-06, + "loss": 0.4383, + "step": 18966 + }, + { + "epoch": 0.598233717079325, + "grad_norm": 1.7394341230392456, + "learning_rate": 7.338126327508397e-06, + "loss": 0.4472, + "step": 18967 + }, + { + "epoch": 0.5982652578457657, + "grad_norm": 1.7299374341964722, + "learning_rate": 7.337141641339061e-06, + "loss": 0.3685, + "step": 18968 + }, + { + "epoch": 0.5982967986122063, + "grad_norm": 2.0427849292755127, + "learning_rate": 7.336156982958718e-06, + "loss": 0.4459, + "step": 18969 + }, + { + "epoch": 0.5983283393786469, + "grad_norm": 2.116704225540161, + "learning_rate": 7.335172352377646e-06, + "loss": 0.4595, + "step": 18970 + }, + { + "epoch": 0.5983598801450876, + "grad_norm": 1.7153884172439575, + "learning_rate": 7.33418774960612e-06, + "loss": 0.4148, + "step": 18971 + }, + { + "epoch": 0.5983914209115282, + "grad_norm": 1.9521645307540894, + "learning_rate": 7.333203174654411e-06, + "loss": 0.4073, + "step": 18972 + }, + { + "epoch": 0.5984229616779688, + "grad_norm": 1.4875715970993042, + "learning_rate": 7.332218627532801e-06, + "loss": 0.7417, + "step": 18973 + }, + { + "epoch": 0.5984545024444095, + "grad_norm": 1.651501178741455, + "learning_rate": 7.331234108251562e-06, + "loss": 0.4115, + "step": 18974 + }, + { + "epoch": 0.59848604321085, + "grad_norm": 1.6285192966461182, + "learning_rate": 7.330249616820961e-06, + "loss": 0.4304, + "step": 18975 + }, + { + "epoch": 0.5985175839772906, + "grad_norm": 1.5511764287948608, + "learning_rate": 7.329265153251285e-06, + "loss": 0.3406, + "step": 18976 + }, + { + "epoch": 0.5985491247437312, + "grad_norm": 1.5204850435256958, + "learning_rate": 7.3282807175528e-06, + "loss": 0.7224, + "step": 18977 + }, + { + "epoch": 0.5985806655101719, + "grad_norm": 1.4912751913070679, + "learning_rate": 7.327296309735779e-06, + "loss": 0.4304, + "step": 18978 + }, + { + "epoch": 0.5986122062766125, + "grad_norm": 2.222140312194824, + "learning_rate": 7.326311929810497e-06, + "loss": 0.7394, + "step": 18979 + }, + { + "epoch": 0.5986437470430531, + "grad_norm": 1.488816499710083, + "learning_rate": 7.325327577787223e-06, + "loss": 0.4607, + "step": 18980 + }, + { + "epoch": 0.5986752878094938, + "grad_norm": 1.4064439535140991, + "learning_rate": 7.324343253676236e-06, + "loss": 0.4603, + "step": 18981 + }, + { + "epoch": 0.5987068285759344, + "grad_norm": 1.425119161605835, + "learning_rate": 7.323358957487805e-06, + "loss": 0.453, + "step": 18982 + }, + { + "epoch": 0.598738369342375, + "grad_norm": 1.997788429260254, + "learning_rate": 7.322374689232197e-06, + "loss": 0.4224, + "step": 18983 + }, + { + "epoch": 0.5987699101088156, + "grad_norm": 1.5672317743301392, + "learning_rate": 7.321390448919693e-06, + "loss": 0.4891, + "step": 18984 + }, + { + "epoch": 0.5988014508752563, + "grad_norm": 1.7600877285003662, + "learning_rate": 7.32040623656056e-06, + "loss": 0.7471, + "step": 18985 + }, + { + "epoch": 0.5988329916416969, + "grad_norm": 1.9948234558105469, + "learning_rate": 7.319422052165063e-06, + "loss": 0.4503, + "step": 18986 + }, + { + "epoch": 0.5988645324081375, + "grad_norm": 1.5926356315612793, + "learning_rate": 7.318437895743483e-06, + "loss": 0.4195, + "step": 18987 + }, + { + "epoch": 0.5988960731745782, + "grad_norm": 1.5499851703643799, + "learning_rate": 7.317453767306087e-06, + "loss": 0.448, + "step": 18988 + }, + { + "epoch": 0.5989276139410188, + "grad_norm": 1.5211002826690674, + "learning_rate": 7.316469666863139e-06, + "loss": 0.4462, + "step": 18989 + }, + { + "epoch": 0.5989591547074594, + "grad_norm": 1.694143533706665, + "learning_rate": 7.315485594424918e-06, + "loss": 0.4334, + "step": 18990 + }, + { + "epoch": 0.5989906954739, + "grad_norm": 1.4298254251480103, + "learning_rate": 7.314501550001689e-06, + "loss": 0.4594, + "step": 18991 + }, + { + "epoch": 0.5990222362403407, + "grad_norm": 1.3424874544143677, + "learning_rate": 7.313517533603718e-06, + "loss": 0.4663, + "step": 18992 + }, + { + "epoch": 0.5990537770067813, + "grad_norm": 1.556981086730957, + "learning_rate": 7.312533545241281e-06, + "loss": 0.4514, + "step": 18993 + }, + { + "epoch": 0.5990853177732219, + "grad_norm": 1.8409773111343384, + "learning_rate": 7.311549584924643e-06, + "loss": 0.4601, + "step": 18994 + }, + { + "epoch": 0.5991168585396626, + "grad_norm": 1.905157208442688, + "learning_rate": 7.310565652664068e-06, + "loss": 0.4309, + "step": 18995 + }, + { + "epoch": 0.5991483993061031, + "grad_norm": 2.076510190963745, + "learning_rate": 7.309581748469834e-06, + "loss": 0.4288, + "step": 18996 + }, + { + "epoch": 0.5991799400725437, + "grad_norm": 2.113633632659912, + "learning_rate": 7.308597872352201e-06, + "loss": 0.4137, + "step": 18997 + }, + { + "epoch": 0.5992114808389843, + "grad_norm": 1.6566210985183716, + "learning_rate": 7.307614024321441e-06, + "loss": 0.4163, + "step": 18998 + }, + { + "epoch": 0.599243021605425, + "grad_norm": 1.7549469470977783, + "learning_rate": 7.306630204387819e-06, + "loss": 0.4275, + "step": 18999 + }, + { + "epoch": 0.5992745623718656, + "grad_norm": 1.8256677389144897, + "learning_rate": 7.305646412561599e-06, + "loss": 0.4749, + "step": 19000 + }, + { + "epoch": 0.5993061031383062, + "grad_norm": 1.4851386547088623, + "learning_rate": 7.304662648853053e-06, + "loss": 0.4467, + "step": 19001 + }, + { + "epoch": 0.5993376439047469, + "grad_norm": 1.4570752382278442, + "learning_rate": 7.303678913272447e-06, + "loss": 0.4272, + "step": 19002 + }, + { + "epoch": 0.5993691846711875, + "grad_norm": 1.152555227279663, + "learning_rate": 7.302695205830041e-06, + "loss": 0.3546, + "step": 19003 + }, + { + "epoch": 0.5994007254376281, + "grad_norm": 4.66812801361084, + "learning_rate": 7.301711526536108e-06, + "loss": 0.464, + "step": 19004 + }, + { + "epoch": 0.5994322662040688, + "grad_norm": 1.732558250427246, + "learning_rate": 7.3007278754009115e-06, + "loss": 0.4573, + "step": 19005 + }, + { + "epoch": 0.5994638069705094, + "grad_norm": 1.4855422973632812, + "learning_rate": 7.299744252434711e-06, + "loss": 0.4357, + "step": 19006 + }, + { + "epoch": 0.59949534773695, + "grad_norm": 1.8420186042785645, + "learning_rate": 7.298760657647779e-06, + "loss": 0.3483, + "step": 19007 + }, + { + "epoch": 0.5995268885033906, + "grad_norm": 2.3535356521606445, + "learning_rate": 7.297777091050378e-06, + "loss": 0.4431, + "step": 19008 + }, + { + "epoch": 0.5995584292698313, + "grad_norm": 1.5245246887207031, + "learning_rate": 7.29679355265277e-06, + "loss": 0.7033, + "step": 19009 + }, + { + "epoch": 0.5995899700362719, + "grad_norm": 1.607241153717041, + "learning_rate": 7.295810042465221e-06, + "loss": 0.4123, + "step": 19010 + }, + { + "epoch": 0.5996215108027125, + "grad_norm": 1.360199213027954, + "learning_rate": 7.294826560497994e-06, + "loss": 0.7571, + "step": 19011 + }, + { + "epoch": 0.5996530515691532, + "grad_norm": 1.9199793338775635, + "learning_rate": 7.2938431067613495e-06, + "loss": 0.4185, + "step": 19012 + }, + { + "epoch": 0.5996845923355938, + "grad_norm": 1.8426995277404785, + "learning_rate": 7.292859681265556e-06, + "loss": 0.4147, + "step": 19013 + }, + { + "epoch": 0.5997161331020344, + "grad_norm": 1.9424082040786743, + "learning_rate": 7.2918762840208704e-06, + "loss": 0.466, + "step": 19014 + }, + { + "epoch": 0.599747673868475, + "grad_norm": 1.6401708126068115, + "learning_rate": 7.2908929150375614e-06, + "loss": 0.4388, + "step": 19015 + }, + { + "epoch": 0.5997792146349157, + "grad_norm": 2.088939666748047, + "learning_rate": 7.2899095743258905e-06, + "loss": 0.4254, + "step": 19016 + }, + { + "epoch": 0.5998107554013562, + "grad_norm": 1.6558785438537598, + "learning_rate": 7.288926261896114e-06, + "loss": 0.4522, + "step": 19017 + }, + { + "epoch": 0.5998422961677968, + "grad_norm": 1.5332226753234863, + "learning_rate": 7.2879429777585e-06, + "loss": 0.4394, + "step": 19018 + }, + { + "epoch": 0.5998738369342375, + "grad_norm": 1.4576855897903442, + "learning_rate": 7.286959721923307e-06, + "loss": 0.7069, + "step": 19019 + }, + { + "epoch": 0.5999053777006781, + "grad_norm": 1.7865713834762573, + "learning_rate": 7.2859764944007904e-06, + "loss": 0.4654, + "step": 19020 + }, + { + "epoch": 0.5999369184671187, + "grad_norm": 2.226459503173828, + "learning_rate": 7.2849932952012225e-06, + "loss": 0.4238, + "step": 19021 + }, + { + "epoch": 0.5999684592335593, + "grad_norm": 1.8705379962921143, + "learning_rate": 7.2840101243348575e-06, + "loss": 0.4591, + "step": 19022 + }, + { + "epoch": 0.6, + "grad_norm": 1.437795639038086, + "learning_rate": 7.28302698181195e-06, + "loss": 0.4643, + "step": 19023 + }, + { + "epoch": 0.6000315407664406, + "grad_norm": 1.922380805015564, + "learning_rate": 7.282043867642771e-06, + "loss": 0.464, + "step": 19024 + }, + { + "epoch": 0.6000630815328812, + "grad_norm": 2.468482732772827, + "learning_rate": 7.2810607818375745e-06, + "loss": 0.7328, + "step": 19025 + }, + { + "epoch": 0.6000946222993219, + "grad_norm": 4.587865829467773, + "learning_rate": 7.280077724406616e-06, + "loss": 0.4043, + "step": 19026 + }, + { + "epoch": 0.6001261630657625, + "grad_norm": 1.821144700050354, + "learning_rate": 7.279094695360161e-06, + "loss": 0.4534, + "step": 19027 + }, + { + "epoch": 0.6001577038322031, + "grad_norm": 1.664449691772461, + "learning_rate": 7.278111694708465e-06, + "loss": 0.4385, + "step": 19028 + }, + { + "epoch": 0.6001892445986438, + "grad_norm": 1.5985325574874878, + "learning_rate": 7.277128722461788e-06, + "loss": 0.4582, + "step": 19029 + }, + { + "epoch": 0.6002207853650844, + "grad_norm": 1.8474620580673218, + "learning_rate": 7.276145778630388e-06, + "loss": 0.4556, + "step": 19030 + }, + { + "epoch": 0.600252326131525, + "grad_norm": 1.9667989015579224, + "learning_rate": 7.275162863224519e-06, + "loss": 0.4543, + "step": 19031 + }, + { + "epoch": 0.6002838668979656, + "grad_norm": 1.6732856035232544, + "learning_rate": 7.274179976254443e-06, + "loss": 0.476, + "step": 19032 + }, + { + "epoch": 0.6003154076644063, + "grad_norm": 2.4792964458465576, + "learning_rate": 7.2731971177304165e-06, + "loss": 0.466, + "step": 19033 + }, + { + "epoch": 0.6003469484308469, + "grad_norm": 1.4724504947662354, + "learning_rate": 7.272214287662691e-06, + "loss": 0.4746, + "step": 19034 + }, + { + "epoch": 0.6003784891972875, + "grad_norm": 1.8120858669281006, + "learning_rate": 7.271231486061532e-06, + "loss": 0.4263, + "step": 19035 + }, + { + "epoch": 0.6004100299637282, + "grad_norm": 1.9006403684616089, + "learning_rate": 7.270248712937191e-06, + "loss": 0.4297, + "step": 19036 + }, + { + "epoch": 0.6004415707301688, + "grad_norm": 1.4932140111923218, + "learning_rate": 7.269265968299921e-06, + "loss": 0.3892, + "step": 19037 + }, + { + "epoch": 0.6004731114966093, + "grad_norm": 1.507822036743164, + "learning_rate": 7.268283252159984e-06, + "loss": 0.4679, + "step": 19038 + }, + { + "epoch": 0.6005046522630499, + "grad_norm": 1.430804967880249, + "learning_rate": 7.267300564527634e-06, + "loss": 0.4392, + "step": 19039 + }, + { + "epoch": 0.6005361930294906, + "grad_norm": 2.54254150390625, + "learning_rate": 7.2663179054131186e-06, + "loss": 0.4704, + "step": 19040 + }, + { + "epoch": 0.6005677337959312, + "grad_norm": 1.6262409687042236, + "learning_rate": 7.265335274826704e-06, + "loss": 0.4674, + "step": 19041 + }, + { + "epoch": 0.6005992745623718, + "grad_norm": 0.9973732233047485, + "learning_rate": 7.264352672778637e-06, + "loss": 0.352, + "step": 19042 + }, + { + "epoch": 0.6006308153288125, + "grad_norm": 1.526321291923523, + "learning_rate": 7.263370099279173e-06, + "loss": 0.4548, + "step": 19043 + }, + { + "epoch": 0.6006623560952531, + "grad_norm": 1.5570244789123535, + "learning_rate": 7.262387554338568e-06, + "loss": 0.7125, + "step": 19044 + }, + { + "epoch": 0.6006938968616937, + "grad_norm": 2.0608773231506348, + "learning_rate": 7.261405037967073e-06, + "loss": 0.4536, + "step": 19045 + }, + { + "epoch": 0.6007254376281344, + "grad_norm": 1.8015081882476807, + "learning_rate": 7.2604225501749435e-06, + "loss": 0.4522, + "step": 19046 + }, + { + "epoch": 0.600756978394575, + "grad_norm": 1.5157620906829834, + "learning_rate": 7.259440090972435e-06, + "loss": 0.4479, + "step": 19047 + }, + { + "epoch": 0.6007885191610156, + "grad_norm": 1.4654384851455688, + "learning_rate": 7.2584576603697945e-06, + "loss": 0.4591, + "step": 19048 + }, + { + "epoch": 0.6008200599274562, + "grad_norm": 1.3972712755203247, + "learning_rate": 7.257475258377278e-06, + "loss": 0.4489, + "step": 19049 + }, + { + "epoch": 0.6008516006938969, + "grad_norm": 1.73124098777771, + "learning_rate": 7.2564928850051375e-06, + "loss": 0.4559, + "step": 19050 + }, + { + "epoch": 0.6008831414603375, + "grad_norm": 1.7991031408309937, + "learning_rate": 7.25551054026362e-06, + "loss": 0.4788, + "step": 19051 + }, + { + "epoch": 0.6009146822267781, + "grad_norm": 2.7931981086730957, + "learning_rate": 7.254528224162986e-06, + "loss": 0.4761, + "step": 19052 + }, + { + "epoch": 0.6009462229932188, + "grad_norm": 1.3381531238555908, + "learning_rate": 7.253545936713481e-06, + "loss": 0.3563, + "step": 19053 + }, + { + "epoch": 0.6009777637596594, + "grad_norm": 1.4821271896362305, + "learning_rate": 7.252563677925353e-06, + "loss": 0.4414, + "step": 19054 + }, + { + "epoch": 0.6010093045261, + "grad_norm": 1.4507540464401245, + "learning_rate": 7.25158144780886e-06, + "loss": 0.4407, + "step": 19055 + }, + { + "epoch": 0.6010408452925406, + "grad_norm": 1.7106611728668213, + "learning_rate": 7.250599246374248e-06, + "loss": 0.4343, + "step": 19056 + }, + { + "epoch": 0.6010723860589813, + "grad_norm": 1.452696681022644, + "learning_rate": 7.249617073631766e-06, + "loss": 0.4593, + "step": 19057 + }, + { + "epoch": 0.6011039268254219, + "grad_norm": 1.6030206680297852, + "learning_rate": 7.248634929591668e-06, + "loss": 0.4536, + "step": 19058 + }, + { + "epoch": 0.6011354675918624, + "grad_norm": 1.6183676719665527, + "learning_rate": 7.247652814264201e-06, + "loss": 0.4556, + "step": 19059 + }, + { + "epoch": 0.6011670083583031, + "grad_norm": 1.2607719898223877, + "learning_rate": 7.246670727659609e-06, + "loss": 0.8122, + "step": 19060 + }, + { + "epoch": 0.6011985491247437, + "grad_norm": 1.552222728729248, + "learning_rate": 7.245688669788151e-06, + "loss": 0.4199, + "step": 19061 + }, + { + "epoch": 0.6012300898911843, + "grad_norm": 2.70359206199646, + "learning_rate": 7.244706640660066e-06, + "loss": 0.4844, + "step": 19062 + }, + { + "epoch": 0.601261630657625, + "grad_norm": 1.548572301864624, + "learning_rate": 7.2437246402856096e-06, + "loss": 0.4511, + "step": 19063 + }, + { + "epoch": 0.6012931714240656, + "grad_norm": 1.3614391088485718, + "learning_rate": 7.242742668675029e-06, + "loss": 0.4509, + "step": 19064 + }, + { + "epoch": 0.6013247121905062, + "grad_norm": 1.6685115098953247, + "learning_rate": 7.241760725838566e-06, + "loss": 0.4344, + "step": 19065 + }, + { + "epoch": 0.6013562529569468, + "grad_norm": 1.6365196704864502, + "learning_rate": 7.240778811786472e-06, + "loss": 0.4535, + "step": 19066 + }, + { + "epoch": 0.6013877937233875, + "grad_norm": 1.649098515510559, + "learning_rate": 7.239796926528995e-06, + "loss": 0.4587, + "step": 19067 + }, + { + "epoch": 0.6014193344898281, + "grad_norm": 6.201595783233643, + "learning_rate": 7.238815070076379e-06, + "loss": 0.4359, + "step": 19068 + }, + { + "epoch": 0.6014508752562687, + "grad_norm": 1.3764567375183105, + "learning_rate": 7.2378332424388745e-06, + "loss": 0.4399, + "step": 19069 + }, + { + "epoch": 0.6014824160227094, + "grad_norm": 1.5008162260055542, + "learning_rate": 7.236851443626723e-06, + "loss": 0.4732, + "step": 19070 + }, + { + "epoch": 0.60151395678915, + "grad_norm": 1.4658311605453491, + "learning_rate": 7.235869673650169e-06, + "loss": 0.3511, + "step": 19071 + }, + { + "epoch": 0.6015454975555906, + "grad_norm": 1.904434084892273, + "learning_rate": 7.234887932519466e-06, + "loss": 0.7378, + "step": 19072 + }, + { + "epoch": 0.6015770383220312, + "grad_norm": 1.412807822227478, + "learning_rate": 7.233906220244854e-06, + "loss": 0.4504, + "step": 19073 + }, + { + "epoch": 0.6016085790884719, + "grad_norm": 1.8335438966751099, + "learning_rate": 7.2329245368365745e-06, + "loss": 0.49, + "step": 19074 + }, + { + "epoch": 0.6016401198549125, + "grad_norm": 1.787930965423584, + "learning_rate": 7.231942882304879e-06, + "loss": 0.4572, + "step": 19075 + }, + { + "epoch": 0.6016716606213531, + "grad_norm": 2.3652875423431396, + "learning_rate": 7.230961256660007e-06, + "loss": 0.4397, + "step": 19076 + }, + { + "epoch": 0.6017032013877938, + "grad_norm": 1.6325407028198242, + "learning_rate": 7.229979659912207e-06, + "loss": 0.4529, + "step": 19077 + }, + { + "epoch": 0.6017347421542344, + "grad_norm": 1.5773133039474487, + "learning_rate": 7.228998092071722e-06, + "loss": 0.7307, + "step": 19078 + }, + { + "epoch": 0.601766282920675, + "grad_norm": 1.5270090103149414, + "learning_rate": 7.228016553148787e-06, + "loss": 0.4778, + "step": 19079 + }, + { + "epoch": 0.6017978236871157, + "grad_norm": 4.301639080047607, + "learning_rate": 7.227035043153657e-06, + "loss": 0.4368, + "step": 19080 + }, + { + "epoch": 0.6018293644535562, + "grad_norm": 1.3742121458053589, + "learning_rate": 7.226053562096569e-06, + "loss": 0.4605, + "step": 19081 + }, + { + "epoch": 0.6018609052199968, + "grad_norm": 1.6362409591674805, + "learning_rate": 7.225072109987761e-06, + "loss": 0.4547, + "step": 19082 + }, + { + "epoch": 0.6018924459864374, + "grad_norm": 1.884820818901062, + "learning_rate": 7.224090686837486e-06, + "loss": 0.4435, + "step": 19083 + }, + { + "epoch": 0.6019239867528781, + "grad_norm": 1.4694327116012573, + "learning_rate": 7.2231092926559785e-06, + "loss": 0.436, + "step": 19084 + }, + { + "epoch": 0.6019555275193187, + "grad_norm": 1.6431230306625366, + "learning_rate": 7.22212792745348e-06, + "loss": 0.48, + "step": 19085 + }, + { + "epoch": 0.6019870682857593, + "grad_norm": 1.8165405988693237, + "learning_rate": 7.221146591240234e-06, + "loss": 0.3888, + "step": 19086 + }, + { + "epoch": 0.6020186090522, + "grad_norm": 2.1482229232788086, + "learning_rate": 7.220165284026484e-06, + "loss": 0.4374, + "step": 19087 + }, + { + "epoch": 0.6020501498186406, + "grad_norm": 3.143838405609131, + "learning_rate": 7.219184005822464e-06, + "loss": 0.4268, + "step": 19088 + }, + { + "epoch": 0.6020816905850812, + "grad_norm": 1.5561038255691528, + "learning_rate": 7.2182027566384215e-06, + "loss": 0.4422, + "step": 19089 + }, + { + "epoch": 0.6021132313515218, + "grad_norm": 2.2984390258789062, + "learning_rate": 7.217221536484593e-06, + "loss": 0.33, + "step": 19090 + }, + { + "epoch": 0.6021447721179625, + "grad_norm": 2.070352077484131, + "learning_rate": 7.216240345371213e-06, + "loss": 0.4752, + "step": 19091 + }, + { + "epoch": 0.6021763128844031, + "grad_norm": 1.7905750274658203, + "learning_rate": 7.215259183308532e-06, + "loss": 0.4267, + "step": 19092 + }, + { + "epoch": 0.6022078536508437, + "grad_norm": 1.553445816040039, + "learning_rate": 7.2142780503067796e-06, + "loss": 0.5028, + "step": 19093 + }, + { + "epoch": 0.6022393944172844, + "grad_norm": 1.5687137842178345, + "learning_rate": 7.213296946376203e-06, + "loss": 0.4488, + "step": 19094 + }, + { + "epoch": 0.602270935183725, + "grad_norm": 1.3730511665344238, + "learning_rate": 7.212315871527036e-06, + "loss": 0.4108, + "step": 19095 + }, + { + "epoch": 0.6023024759501656, + "grad_norm": 1.303337574005127, + "learning_rate": 7.211334825769516e-06, + "loss": 0.4231, + "step": 19096 + }, + { + "epoch": 0.6023340167166062, + "grad_norm": 1.5029646158218384, + "learning_rate": 7.210353809113884e-06, + "loss": 0.443, + "step": 19097 + }, + { + "epoch": 0.6023655574830469, + "grad_norm": 1.512664556503296, + "learning_rate": 7.209372821570376e-06, + "loss": 0.4443, + "step": 19098 + }, + { + "epoch": 0.6023970982494875, + "grad_norm": 1.4597303867340088, + "learning_rate": 7.208391863149225e-06, + "loss": 0.464, + "step": 19099 + }, + { + "epoch": 0.6024286390159281, + "grad_norm": 1.7534923553466797, + "learning_rate": 7.207410933860678e-06, + "loss": 0.4173, + "step": 19100 + }, + { + "epoch": 0.6024601797823688, + "grad_norm": 1.7890052795410156, + "learning_rate": 7.206430033714966e-06, + "loss": 0.4411, + "step": 19101 + }, + { + "epoch": 0.6024917205488093, + "grad_norm": 2.022500991821289, + "learning_rate": 7.20544916272232e-06, + "loss": 0.4758, + "step": 19102 + }, + { + "epoch": 0.6025232613152499, + "grad_norm": 4.185142993927002, + "learning_rate": 7.2044683208929875e-06, + "loss": 0.4283, + "step": 19103 + }, + { + "epoch": 0.6025548020816905, + "grad_norm": 1.9178913831710815, + "learning_rate": 7.2034875082371994e-06, + "loss": 0.414, + "step": 19104 + }, + { + "epoch": 0.6025863428481312, + "grad_norm": 2.249114513397217, + "learning_rate": 7.202506724765187e-06, + "loss": 0.4003, + "step": 19105 + }, + { + "epoch": 0.6026178836145718, + "grad_norm": 1.577195405960083, + "learning_rate": 7.20152597048719e-06, + "loss": 0.4537, + "step": 19106 + }, + { + "epoch": 0.6026494243810124, + "grad_norm": 1.8378956317901611, + "learning_rate": 7.200545245413444e-06, + "loss": 0.4443, + "step": 19107 + }, + { + "epoch": 0.6026809651474531, + "grad_norm": 1.9175511598587036, + "learning_rate": 7.19956454955418e-06, + "loss": 0.391, + "step": 19108 + }, + { + "epoch": 0.6027125059138937, + "grad_norm": 1.4453781843185425, + "learning_rate": 7.198583882919636e-06, + "loss": 0.4321, + "step": 19109 + }, + { + "epoch": 0.6027440466803343, + "grad_norm": 1.9128512144088745, + "learning_rate": 7.197603245520042e-06, + "loss": 0.4598, + "step": 19110 + }, + { + "epoch": 0.602775587446775, + "grad_norm": 1.6365940570831299, + "learning_rate": 7.1966226373656376e-06, + "loss": 0.4513, + "step": 19111 + }, + { + "epoch": 0.6028071282132156, + "grad_norm": 1.644358515739441, + "learning_rate": 7.195642058466652e-06, + "loss": 0.468, + "step": 19112 + }, + { + "epoch": 0.6028386689796562, + "grad_norm": 1.592942476272583, + "learning_rate": 7.194661508833316e-06, + "loss": 0.4264, + "step": 19113 + }, + { + "epoch": 0.6028702097460968, + "grad_norm": 1.6347836256027222, + "learning_rate": 7.1936809884758685e-06, + "loss": 0.4376, + "step": 19114 + }, + { + "epoch": 0.6029017505125375, + "grad_norm": 1.544250249862671, + "learning_rate": 7.192700497404539e-06, + "loss": 0.4447, + "step": 19115 + }, + { + "epoch": 0.6029332912789781, + "grad_norm": 1.6674268245697021, + "learning_rate": 7.191720035629558e-06, + "loss": 0.466, + "step": 19116 + }, + { + "epoch": 0.6029648320454187, + "grad_norm": 2.1561169624328613, + "learning_rate": 7.190739603161161e-06, + "loss": 0.7476, + "step": 19117 + }, + { + "epoch": 0.6029963728118594, + "grad_norm": 8.38189697265625, + "learning_rate": 7.189759200009579e-06, + "loss": 0.4807, + "step": 19118 + }, + { + "epoch": 0.6030279135783, + "grad_norm": 1.759575366973877, + "learning_rate": 7.188778826185034e-06, + "loss": 0.4611, + "step": 19119 + }, + { + "epoch": 0.6030594543447406, + "grad_norm": 1.34209406375885, + "learning_rate": 7.187798481697772e-06, + "loss": 0.4591, + "step": 19120 + }, + { + "epoch": 0.6030909951111812, + "grad_norm": 1.4187589883804321, + "learning_rate": 7.186818166558016e-06, + "loss": 0.4387, + "step": 19121 + }, + { + "epoch": 0.6031225358776219, + "grad_norm": 2.052213430404663, + "learning_rate": 7.185837880775992e-06, + "loss": 0.428, + "step": 19122 + }, + { + "epoch": 0.6031540766440624, + "grad_norm": 2.2976577281951904, + "learning_rate": 7.18485762436194e-06, + "loss": 0.4763, + "step": 19123 + }, + { + "epoch": 0.603185617410503, + "grad_norm": 1.7583664655685425, + "learning_rate": 7.1838773973260814e-06, + "loss": 0.4476, + "step": 19124 + }, + { + "epoch": 0.6032171581769437, + "grad_norm": 2.2227768898010254, + "learning_rate": 7.182897199678651e-06, + "loss": 0.4436, + "step": 19125 + }, + { + "epoch": 0.6032486989433843, + "grad_norm": 2.2993850708007812, + "learning_rate": 7.1819170314298746e-06, + "loss": 0.4632, + "step": 19126 + }, + { + "epoch": 0.6032802397098249, + "grad_norm": 1.7900192737579346, + "learning_rate": 7.180936892589981e-06, + "loss": 0.4394, + "step": 19127 + }, + { + "epoch": 0.6033117804762655, + "grad_norm": 1.3338834047317505, + "learning_rate": 7.179956783169203e-06, + "loss": 0.523, + "step": 19128 + }, + { + "epoch": 0.6033433212427062, + "grad_norm": 1.6032989025115967, + "learning_rate": 7.178976703177764e-06, + "loss": 0.4516, + "step": 19129 + }, + { + "epoch": 0.6033748620091468, + "grad_norm": 2.05598783493042, + "learning_rate": 7.177996652625891e-06, + "loss": 0.4606, + "step": 19130 + }, + { + "epoch": 0.6034064027755874, + "grad_norm": 2.179227828979492, + "learning_rate": 7.177016631523819e-06, + "loss": 0.488, + "step": 19131 + }, + { + "epoch": 0.6034379435420281, + "grad_norm": 2.068449020385742, + "learning_rate": 7.176036639881769e-06, + "loss": 0.4523, + "step": 19132 + }, + { + "epoch": 0.6034694843084687, + "grad_norm": 2.8510637283325195, + "learning_rate": 7.1750566777099665e-06, + "loss": 0.3997, + "step": 19133 + }, + { + "epoch": 0.6035010250749093, + "grad_norm": 1.5139875411987305, + "learning_rate": 7.1740767450186435e-06, + "loss": 0.7622, + "step": 19134 + }, + { + "epoch": 0.60353256584135, + "grad_norm": 1.6297630071640015, + "learning_rate": 7.1730968418180245e-06, + "loss": 0.3995, + "step": 19135 + }, + { + "epoch": 0.6035641066077906, + "grad_norm": 5.435494422912598, + "learning_rate": 7.172116968118333e-06, + "loss": 0.4247, + "step": 19136 + }, + { + "epoch": 0.6035956473742312, + "grad_norm": 1.5455195903778076, + "learning_rate": 7.171137123929799e-06, + "loss": 0.3852, + "step": 19137 + }, + { + "epoch": 0.6036271881406718, + "grad_norm": 1.6439149379730225, + "learning_rate": 7.170157309262647e-06, + "loss": 0.4513, + "step": 19138 + }, + { + "epoch": 0.6036587289071125, + "grad_norm": 1.6098483800888062, + "learning_rate": 7.169177524127095e-06, + "loss": 0.3597, + "step": 19139 + }, + { + "epoch": 0.6036902696735531, + "grad_norm": 1.2321146726608276, + "learning_rate": 7.168197768533378e-06, + "loss": 0.4214, + "step": 19140 + }, + { + "epoch": 0.6037218104399937, + "grad_norm": 1.4561768770217896, + "learning_rate": 7.167218042491712e-06, + "loss": 0.4617, + "step": 19141 + }, + { + "epoch": 0.6037533512064344, + "grad_norm": 1.9768956899642944, + "learning_rate": 7.166238346012329e-06, + "loss": 0.4512, + "step": 19142 + }, + { + "epoch": 0.603784891972875, + "grad_norm": 3.6067371368408203, + "learning_rate": 7.16525867910545e-06, + "loss": 0.3844, + "step": 19143 + }, + { + "epoch": 0.6038164327393155, + "grad_norm": 2.5122570991516113, + "learning_rate": 7.164279041781295e-06, + "loss": 0.7472, + "step": 19144 + }, + { + "epoch": 0.6038479735057561, + "grad_norm": 1.3263224363327026, + "learning_rate": 7.163299434050093e-06, + "loss": 0.4521, + "step": 19145 + }, + { + "epoch": 0.6038795142721968, + "grad_norm": 1.9204999208450317, + "learning_rate": 7.162319855922062e-06, + "loss": 0.4783, + "step": 19146 + }, + { + "epoch": 0.6039110550386374, + "grad_norm": 1.9141981601715088, + "learning_rate": 7.161340307407426e-06, + "loss": 0.3994, + "step": 19147 + }, + { + "epoch": 0.603942595805078, + "grad_norm": 1.5960251092910767, + "learning_rate": 7.16036078851641e-06, + "loss": 0.4743, + "step": 19148 + }, + { + "epoch": 0.6039741365715187, + "grad_norm": 1.5568933486938477, + "learning_rate": 7.159381299259236e-06, + "loss": 0.4151, + "step": 19149 + }, + { + "epoch": 0.6040056773379593, + "grad_norm": 1.6401790380477905, + "learning_rate": 7.158401839646117e-06, + "loss": 0.4261, + "step": 19150 + }, + { + "epoch": 0.6040372181043999, + "grad_norm": 1.5616697072982788, + "learning_rate": 7.1574224096872865e-06, + "loss": 0.7458, + "step": 19151 + }, + { + "epoch": 0.6040687588708405, + "grad_norm": 1.4720215797424316, + "learning_rate": 7.156443009392962e-06, + "loss": 0.7913, + "step": 19152 + }, + { + "epoch": 0.6041002996372812, + "grad_norm": 1.5599855184555054, + "learning_rate": 7.155463638773356e-06, + "loss": 0.4226, + "step": 19153 + }, + { + "epoch": 0.6041318404037218, + "grad_norm": 1.698456048965454, + "learning_rate": 7.154484297838701e-06, + "loss": 0.4724, + "step": 19154 + }, + { + "epoch": 0.6041633811701624, + "grad_norm": 2.0661909580230713, + "learning_rate": 7.153504986599209e-06, + "loss": 0.4645, + "step": 19155 + }, + { + "epoch": 0.6041949219366031, + "grad_norm": 1.749140739440918, + "learning_rate": 7.152525705065106e-06, + "loss": 0.4443, + "step": 19156 + }, + { + "epoch": 0.6042264627030437, + "grad_norm": 2.5484163761138916, + "learning_rate": 7.151546453246608e-06, + "loss": 0.4134, + "step": 19157 + }, + { + "epoch": 0.6042580034694843, + "grad_norm": 1.6669501066207886, + "learning_rate": 7.150567231153929e-06, + "loss": 0.4749, + "step": 19158 + }, + { + "epoch": 0.604289544235925, + "grad_norm": 3.3039965629577637, + "learning_rate": 7.1495880387972995e-06, + "loss": 0.484, + "step": 19159 + }, + { + "epoch": 0.6043210850023656, + "grad_norm": 1.5513852834701538, + "learning_rate": 7.148608876186931e-06, + "loss": 0.4282, + "step": 19160 + }, + { + "epoch": 0.6043526257688062, + "grad_norm": 1.790603518486023, + "learning_rate": 7.147629743333039e-06, + "loss": 0.4458, + "step": 19161 + }, + { + "epoch": 0.6043841665352468, + "grad_norm": 1.5559725761413574, + "learning_rate": 7.146650640245849e-06, + "loss": 0.4689, + "step": 19162 + }, + { + "epoch": 0.6044157073016875, + "grad_norm": 1.5179659128189087, + "learning_rate": 7.1456715669355754e-06, + "loss": 0.4549, + "step": 19163 + }, + { + "epoch": 0.6044472480681281, + "grad_norm": 1.7912554740905762, + "learning_rate": 7.144692523412434e-06, + "loss": 0.4665, + "step": 19164 + }, + { + "epoch": 0.6044787888345686, + "grad_norm": 1.395427942276001, + "learning_rate": 7.143713509686645e-06, + "loss": 0.7287, + "step": 19165 + }, + { + "epoch": 0.6045103296010093, + "grad_norm": 1.7438453435897827, + "learning_rate": 7.142734525768421e-06, + "loss": 0.3922, + "step": 19166 + }, + { + "epoch": 0.6045418703674499, + "grad_norm": 1.7748167514801025, + "learning_rate": 7.141755571667981e-06, + "loss": 0.4639, + "step": 19167 + }, + { + "epoch": 0.6045734111338905, + "grad_norm": 2.3122482299804688, + "learning_rate": 7.140776647395543e-06, + "loss": 0.4306, + "step": 19168 + }, + { + "epoch": 0.6046049519003311, + "grad_norm": 1.388256311416626, + "learning_rate": 7.1397977529613215e-06, + "loss": 0.3717, + "step": 19169 + }, + { + "epoch": 0.6046364926667718, + "grad_norm": 1.680908441543579, + "learning_rate": 7.138818888375526e-06, + "loss": 0.4473, + "step": 19170 + }, + { + "epoch": 0.6046680334332124, + "grad_norm": 1.42658269405365, + "learning_rate": 7.137840053648382e-06, + "loss": 0.7436, + "step": 19171 + }, + { + "epoch": 0.604699574199653, + "grad_norm": 2.4755563735961914, + "learning_rate": 7.136861248790095e-06, + "loss": 0.7753, + "step": 19172 + }, + { + "epoch": 0.6047311149660937, + "grad_norm": 1.4360016584396362, + "learning_rate": 7.135882473810889e-06, + "loss": 0.7036, + "step": 19173 + }, + { + "epoch": 0.6047626557325343, + "grad_norm": 1.840410590171814, + "learning_rate": 7.1349037287209725e-06, + "loss": 0.4866, + "step": 19174 + }, + { + "epoch": 0.6047941964989749, + "grad_norm": 1.366976261138916, + "learning_rate": 7.133925013530558e-06, + "loss": 0.4557, + "step": 19175 + }, + { + "epoch": 0.6048257372654156, + "grad_norm": 1.9030474424362183, + "learning_rate": 7.132946328249863e-06, + "loss": 0.4541, + "step": 19176 + }, + { + "epoch": 0.6048572780318562, + "grad_norm": 1.5959982872009277, + "learning_rate": 7.131967672889101e-06, + "loss": 0.4608, + "step": 19177 + }, + { + "epoch": 0.6048888187982968, + "grad_norm": 2.1859190464019775, + "learning_rate": 7.1309890474584775e-06, + "loss": 0.4265, + "step": 19178 + }, + { + "epoch": 0.6049203595647374, + "grad_norm": 1.464871883392334, + "learning_rate": 7.1300104519682165e-06, + "loss": 0.7549, + "step": 19179 + }, + { + "epoch": 0.6049519003311781, + "grad_norm": 1.1937623023986816, + "learning_rate": 7.129031886428524e-06, + "loss": 0.4496, + "step": 19180 + }, + { + "epoch": 0.6049834410976187, + "grad_norm": 1.9436020851135254, + "learning_rate": 7.128053350849609e-06, + "loss": 0.3907, + "step": 19181 + }, + { + "epoch": 0.6050149818640593, + "grad_norm": 1.433030366897583, + "learning_rate": 7.1270748452416916e-06, + "loss": 0.7139, + "step": 19182 + }, + { + "epoch": 0.6050465226305, + "grad_norm": 1.556003212928772, + "learning_rate": 7.126096369614979e-06, + "loss": 0.4621, + "step": 19183 + }, + { + "epoch": 0.6050780633969406, + "grad_norm": 1.4649673700332642, + "learning_rate": 7.1251179239796805e-06, + "loss": 0.7561, + "step": 19184 + }, + { + "epoch": 0.6051096041633812, + "grad_norm": 1.7614555358886719, + "learning_rate": 7.124139508346011e-06, + "loss": 0.4415, + "step": 19185 + }, + { + "epoch": 0.6051411449298217, + "grad_norm": 1.706024169921875, + "learning_rate": 7.123161122724176e-06, + "loss": 0.4647, + "step": 19186 + }, + { + "epoch": 0.6051726856962624, + "grad_norm": 1.527241587638855, + "learning_rate": 7.1221827671243885e-06, + "loss": 0.4621, + "step": 19187 + }, + { + "epoch": 0.605204226462703, + "grad_norm": 1.3571467399597168, + "learning_rate": 7.121204441556861e-06, + "loss": 0.4304, + "step": 19188 + }, + { + "epoch": 0.6052357672291436, + "grad_norm": 2.006096601486206, + "learning_rate": 7.120226146031795e-06, + "loss": 0.4593, + "step": 19189 + }, + { + "epoch": 0.6052673079955843, + "grad_norm": 1.5469212532043457, + "learning_rate": 7.11924788055941e-06, + "loss": 0.7221, + "step": 19190 + }, + { + "epoch": 0.6052988487620249, + "grad_norm": 2.0271835327148438, + "learning_rate": 7.118269645149911e-06, + "loss": 0.7575, + "step": 19191 + }, + { + "epoch": 0.6053303895284655, + "grad_norm": 1.5988857746124268, + "learning_rate": 7.117291439813502e-06, + "loss": 0.4679, + "step": 19192 + }, + { + "epoch": 0.6053619302949061, + "grad_norm": 1.88759183883667, + "learning_rate": 7.1163132645603985e-06, + "loss": 0.4458, + "step": 19193 + }, + { + "epoch": 0.6053934710613468, + "grad_norm": 1.648245930671692, + "learning_rate": 7.115335119400806e-06, + "loss": 0.4477, + "step": 19194 + }, + { + "epoch": 0.6054250118277874, + "grad_norm": 1.6527466773986816, + "learning_rate": 7.114357004344927e-06, + "loss": 0.4443, + "step": 19195 + }, + { + "epoch": 0.605456552594228, + "grad_norm": 1.7182585000991821, + "learning_rate": 7.113378919402978e-06, + "loss": 0.4328, + "step": 19196 + }, + { + "epoch": 0.6054880933606687, + "grad_norm": 1.5043541193008423, + "learning_rate": 7.1124008645851605e-06, + "loss": 0.4454, + "step": 19197 + }, + { + "epoch": 0.6055196341271093, + "grad_norm": 2.3187406063079834, + "learning_rate": 7.111422839901678e-06, + "loss": 0.4011, + "step": 19198 + }, + { + "epoch": 0.6055511748935499, + "grad_norm": 1.6367803812026978, + "learning_rate": 7.110444845362745e-06, + "loss": 0.7468, + "step": 19199 + }, + { + "epoch": 0.6055827156599906, + "grad_norm": 1.7526910305023193, + "learning_rate": 7.109466880978564e-06, + "loss": 0.4163, + "step": 19200 + }, + { + "epoch": 0.6056142564264312, + "grad_norm": 1.6678611040115356, + "learning_rate": 7.108488946759336e-06, + "loss": 0.4473, + "step": 19201 + }, + { + "epoch": 0.6056457971928718, + "grad_norm": 1.4351507425308228, + "learning_rate": 7.107511042715276e-06, + "loss": 0.7612, + "step": 19202 + }, + { + "epoch": 0.6056773379593124, + "grad_norm": 1.525667667388916, + "learning_rate": 7.106533168856582e-06, + "loss": 0.3811, + "step": 19203 + }, + { + "epoch": 0.6057088787257531, + "grad_norm": 15.178389549255371, + "learning_rate": 7.105555325193463e-06, + "loss": 0.3645, + "step": 19204 + }, + { + "epoch": 0.6057404194921937, + "grad_norm": 2.3488965034484863, + "learning_rate": 7.104577511736122e-06, + "loss": 0.4347, + "step": 19205 + }, + { + "epoch": 0.6057719602586343, + "grad_norm": 1.7507303953170776, + "learning_rate": 7.103599728494757e-06, + "loss": 0.4384, + "step": 19206 + }, + { + "epoch": 0.605803501025075, + "grad_norm": 1.435930848121643, + "learning_rate": 7.1026219754795835e-06, + "loss": 0.4145, + "step": 19207 + }, + { + "epoch": 0.6058350417915155, + "grad_norm": 1.8069802522659302, + "learning_rate": 7.1016442527007986e-06, + "loss": 0.4199, + "step": 19208 + }, + { + "epoch": 0.6058665825579561, + "grad_norm": 1.6178196668624878, + "learning_rate": 7.100666560168603e-06, + "loss": 0.4453, + "step": 19209 + }, + { + "epoch": 0.6058981233243967, + "grad_norm": 1.9268072843551636, + "learning_rate": 7.099688897893208e-06, + "loss": 0.4082, + "step": 19210 + }, + { + "epoch": 0.6059296640908374, + "grad_norm": 1.4104347229003906, + "learning_rate": 7.09871126588481e-06, + "loss": 0.4586, + "step": 19211 + }, + { + "epoch": 0.605961204857278, + "grad_norm": 1.5172001123428345, + "learning_rate": 7.09773366415361e-06, + "loss": 0.4617, + "step": 19212 + }, + { + "epoch": 0.6059927456237186, + "grad_norm": 1.976241946220398, + "learning_rate": 7.096756092709815e-06, + "loss": 0.4494, + "step": 19213 + }, + { + "epoch": 0.6060242863901593, + "grad_norm": 1.692765235900879, + "learning_rate": 7.095778551563626e-06, + "loss": 0.4694, + "step": 19214 + }, + { + "epoch": 0.6060558271565999, + "grad_norm": 1.9069178104400635, + "learning_rate": 7.0948010407252396e-06, + "loss": 0.4341, + "step": 19215 + }, + { + "epoch": 0.6060873679230405, + "grad_norm": 1.4141594171524048, + "learning_rate": 7.093823560204863e-06, + "loss": 0.4368, + "step": 19216 + }, + { + "epoch": 0.6061189086894812, + "grad_norm": 1.812936544418335, + "learning_rate": 7.092846110012694e-06, + "loss": 0.4894, + "step": 19217 + }, + { + "epoch": 0.6061504494559218, + "grad_norm": 1.7496705055236816, + "learning_rate": 7.091868690158929e-06, + "loss": 0.4323, + "step": 19218 + }, + { + "epoch": 0.6061819902223624, + "grad_norm": 1.3947674036026, + "learning_rate": 7.090891300653777e-06, + "loss": 0.7848, + "step": 19219 + }, + { + "epoch": 0.606213530988803, + "grad_norm": 1.3932380676269531, + "learning_rate": 7.089913941507429e-06, + "loss": 0.4474, + "step": 19220 + }, + { + "epoch": 0.6062450717552437, + "grad_norm": 1.5929021835327148, + "learning_rate": 7.088936612730092e-06, + "loss": 0.4604, + "step": 19221 + }, + { + "epoch": 0.6062766125216843, + "grad_norm": 1.298250675201416, + "learning_rate": 7.0879593143319625e-06, + "loss": 0.3705, + "step": 19222 + }, + { + "epoch": 0.6063081532881249, + "grad_norm": 1.3124911785125732, + "learning_rate": 7.086982046323237e-06, + "loss": 0.7259, + "step": 19223 + }, + { + "epoch": 0.6063396940545656, + "grad_norm": 1.677384614944458, + "learning_rate": 7.086004808714117e-06, + "loss": 0.4533, + "step": 19224 + }, + { + "epoch": 0.6063712348210062, + "grad_norm": 1.504699945449829, + "learning_rate": 7.085027601514799e-06, + "loss": 0.4572, + "step": 19225 + }, + { + "epoch": 0.6064027755874468, + "grad_norm": 1.3495209217071533, + "learning_rate": 7.084050424735478e-06, + "loss": 0.4393, + "step": 19226 + }, + { + "epoch": 0.6064343163538874, + "grad_norm": 1.4822642803192139, + "learning_rate": 7.083073278386359e-06, + "loss": 0.4352, + "step": 19227 + }, + { + "epoch": 0.6064658571203281, + "grad_norm": 1.4068899154663086, + "learning_rate": 7.082096162477635e-06, + "loss": 0.4243, + "step": 19228 + }, + { + "epoch": 0.6064973978867686, + "grad_norm": 2.083617687225342, + "learning_rate": 7.081119077019499e-06, + "loss": 0.3925, + "step": 19229 + }, + { + "epoch": 0.6065289386532092, + "grad_norm": 1.3747020959854126, + "learning_rate": 7.080142022022156e-06, + "loss": 0.7323, + "step": 19230 + }, + { + "epoch": 0.6065604794196499, + "grad_norm": 1.9024006128311157, + "learning_rate": 7.0791649974958e-06, + "loss": 0.4736, + "step": 19231 + }, + { + "epoch": 0.6065920201860905, + "grad_norm": 1.8621290922164917, + "learning_rate": 7.078188003450619e-06, + "loss": 0.3658, + "step": 19232 + }, + { + "epoch": 0.6066235609525311, + "grad_norm": 1.5448108911514282, + "learning_rate": 7.07721103989682e-06, + "loss": 0.4401, + "step": 19233 + }, + { + "epoch": 0.6066551017189717, + "grad_norm": 1.71511971950531, + "learning_rate": 7.076234106844593e-06, + "loss": 0.6737, + "step": 19234 + }, + { + "epoch": 0.6066866424854124, + "grad_norm": 2.8226723670959473, + "learning_rate": 7.0752572043041314e-06, + "loss": 0.4112, + "step": 19235 + }, + { + "epoch": 0.606718183251853, + "grad_norm": 1.4102929830551147, + "learning_rate": 7.074280332285634e-06, + "loss": 0.442, + "step": 19236 + }, + { + "epoch": 0.6067497240182936, + "grad_norm": 1.4871516227722168, + "learning_rate": 7.0733034907992885e-06, + "loss": 0.4651, + "step": 19237 + }, + { + "epoch": 0.6067812647847343, + "grad_norm": 1.5065640211105347, + "learning_rate": 7.072326679855298e-06, + "loss": 0.4622, + "step": 19238 + }, + { + "epoch": 0.6068128055511749, + "grad_norm": 1.5720500946044922, + "learning_rate": 7.071349899463853e-06, + "loss": 0.4747, + "step": 19239 + }, + { + "epoch": 0.6068443463176155, + "grad_norm": 1.3894588947296143, + "learning_rate": 7.070373149635139e-06, + "loss": 0.4725, + "step": 19240 + }, + { + "epoch": 0.6068758870840562, + "grad_norm": 1.4290450811386108, + "learning_rate": 7.069396430379363e-06, + "loss": 0.4732, + "step": 19241 + }, + { + "epoch": 0.6069074278504968, + "grad_norm": 1.6846445798873901, + "learning_rate": 7.068419741706709e-06, + "loss": 0.433, + "step": 19242 + }, + { + "epoch": 0.6069389686169374, + "grad_norm": 1.6086864471435547, + "learning_rate": 7.06744308362737e-06, + "loss": 0.7165, + "step": 19243 + }, + { + "epoch": 0.606970509383378, + "grad_norm": 1.4110040664672852, + "learning_rate": 7.066466456151541e-06, + "loss": 0.4897, + "step": 19244 + }, + { + "epoch": 0.6070020501498187, + "grad_norm": 1.321269154548645, + "learning_rate": 7.0654898592894135e-06, + "loss": 0.455, + "step": 19245 + }, + { + "epoch": 0.6070335909162593, + "grad_norm": 1.415649652481079, + "learning_rate": 7.0645132930511715e-06, + "loss": 0.4556, + "step": 19246 + }, + { + "epoch": 0.6070651316826999, + "grad_norm": 1.6787197589874268, + "learning_rate": 7.0635367574470196e-06, + "loss": 0.4337, + "step": 19247 + }, + { + "epoch": 0.6070966724491406, + "grad_norm": 1.8593549728393555, + "learning_rate": 7.062560252487141e-06, + "loss": 0.4386, + "step": 19248 + }, + { + "epoch": 0.6071282132155812, + "grad_norm": 1.845700979232788, + "learning_rate": 7.061583778181722e-06, + "loss": 0.4463, + "step": 19249 + }, + { + "epoch": 0.6071597539820217, + "grad_norm": 1.5540410280227661, + "learning_rate": 7.060607334540961e-06, + "loss": 0.4623, + "step": 19250 + }, + { + "epoch": 0.6071912947484623, + "grad_norm": 1.679626226425171, + "learning_rate": 7.059630921575045e-06, + "loss": 0.7277, + "step": 19251 + }, + { + "epoch": 0.607222835514903, + "grad_norm": 1.5192967653274536, + "learning_rate": 7.058654539294163e-06, + "loss": 0.3448, + "step": 19252 + }, + { + "epoch": 0.6072543762813436, + "grad_norm": 1.648382544517517, + "learning_rate": 7.057678187708506e-06, + "loss": 0.7363, + "step": 19253 + }, + { + "epoch": 0.6072859170477842, + "grad_norm": 1.880222201347351, + "learning_rate": 7.05670186682826e-06, + "loss": 0.474, + "step": 19254 + }, + { + "epoch": 0.6073174578142249, + "grad_norm": 1.668562889099121, + "learning_rate": 7.055725576663617e-06, + "loss": 0.4049, + "step": 19255 + }, + { + "epoch": 0.6073489985806655, + "grad_norm": 2.2742629051208496, + "learning_rate": 7.054749317224765e-06, + "loss": 0.4394, + "step": 19256 + }, + { + "epoch": 0.6073805393471061, + "grad_norm": 1.5385043621063232, + "learning_rate": 7.053773088521886e-06, + "loss": 0.3572, + "step": 19257 + }, + { + "epoch": 0.6074120801135467, + "grad_norm": 2.016716241836548, + "learning_rate": 7.052796890565177e-06, + "loss": 0.4395, + "step": 19258 + }, + { + "epoch": 0.6074436208799874, + "grad_norm": 1.9744338989257812, + "learning_rate": 7.0518207233648195e-06, + "loss": 0.4214, + "step": 19259 + }, + { + "epoch": 0.607475161646428, + "grad_norm": 1.7434293031692505, + "learning_rate": 7.050844586930999e-06, + "loss": 0.739, + "step": 19260 + }, + { + "epoch": 0.6075067024128686, + "grad_norm": 1.3912619352340698, + "learning_rate": 7.04986848127391e-06, + "loss": 0.4148, + "step": 19261 + }, + { + "epoch": 0.6075382431793093, + "grad_norm": 1.6440809965133667, + "learning_rate": 7.048892406403733e-06, + "loss": 0.4723, + "step": 19262 + }, + { + "epoch": 0.6075697839457499, + "grad_norm": 1.7041454315185547, + "learning_rate": 7.0479163623306535e-06, + "loss": 0.4738, + "step": 19263 + }, + { + "epoch": 0.6076013247121905, + "grad_norm": 1.5847554206848145, + "learning_rate": 7.0469403490648615e-06, + "loss": 0.3953, + "step": 19264 + }, + { + "epoch": 0.6076328654786312, + "grad_norm": 2.2525761127471924, + "learning_rate": 7.04596436661654e-06, + "loss": 0.4748, + "step": 19265 + }, + { + "epoch": 0.6076644062450718, + "grad_norm": 1.7725238800048828, + "learning_rate": 7.044988414995868e-06, + "loss": 0.4524, + "step": 19266 + }, + { + "epoch": 0.6076959470115124, + "grad_norm": 1.5153390169143677, + "learning_rate": 7.044012494213043e-06, + "loss": 0.435, + "step": 19267 + }, + { + "epoch": 0.607727487777953, + "grad_norm": 1.8711014986038208, + "learning_rate": 7.043036604278237e-06, + "loss": 0.4635, + "step": 19268 + }, + { + "epoch": 0.6077590285443937, + "grad_norm": 2.0116758346557617, + "learning_rate": 7.042060745201646e-06, + "loss": 0.44, + "step": 19269 + }, + { + "epoch": 0.6077905693108343, + "grad_norm": 1.588252305984497, + "learning_rate": 7.041084916993446e-06, + "loss": 0.4604, + "step": 19270 + }, + { + "epoch": 0.6078221100772748, + "grad_norm": 1.592336654663086, + "learning_rate": 7.040109119663822e-06, + "loss": 0.4567, + "step": 19271 + }, + { + "epoch": 0.6078536508437155, + "grad_norm": 2.101377010345459, + "learning_rate": 7.039133353222956e-06, + "loss": 0.4594, + "step": 19272 + }, + { + "epoch": 0.6078851916101561, + "grad_norm": 2.202087879180908, + "learning_rate": 7.038157617681036e-06, + "loss": 0.4297, + "step": 19273 + }, + { + "epoch": 0.6079167323765967, + "grad_norm": 1.744994878768921, + "learning_rate": 7.037181913048239e-06, + "loss": 0.4453, + "step": 19274 + }, + { + "epoch": 0.6079482731430373, + "grad_norm": 1.7359744310379028, + "learning_rate": 7.036206239334751e-06, + "loss": 0.4335, + "step": 19275 + }, + { + "epoch": 0.607979813909478, + "grad_norm": 1.770426630973816, + "learning_rate": 7.035230596550751e-06, + "loss": 0.4719, + "step": 19276 + }, + { + "epoch": 0.6080113546759186, + "grad_norm": 2.3102164268493652, + "learning_rate": 7.034254984706419e-06, + "loss": 0.3964, + "step": 19277 + }, + { + "epoch": 0.6080428954423592, + "grad_norm": 1.3443630933761597, + "learning_rate": 7.033279403811944e-06, + "loss": 0.7174, + "step": 19278 + }, + { + "epoch": 0.6080744362087999, + "grad_norm": 1.734092116355896, + "learning_rate": 7.0323038538775e-06, + "loss": 0.4445, + "step": 19279 + }, + { + "epoch": 0.6081059769752405, + "grad_norm": 1.79102623462677, + "learning_rate": 7.031328334913265e-06, + "loss": 0.4387, + "step": 19280 + }, + { + "epoch": 0.6081375177416811, + "grad_norm": 1.6018003225326538, + "learning_rate": 7.030352846929431e-06, + "loss": 0.3883, + "step": 19281 + }, + { + "epoch": 0.6081690585081218, + "grad_norm": 1.6298240423202515, + "learning_rate": 7.02937738993617e-06, + "loss": 0.7248, + "step": 19282 + }, + { + "epoch": 0.6082005992745624, + "grad_norm": 1.8830490112304688, + "learning_rate": 7.028401963943658e-06, + "loss": 0.4402, + "step": 19283 + }, + { + "epoch": 0.608232140041003, + "grad_norm": 4.52102518081665, + "learning_rate": 7.027426568962083e-06, + "loss": 0.3884, + "step": 19284 + }, + { + "epoch": 0.6082636808074436, + "grad_norm": 1.676769733428955, + "learning_rate": 7.0264512050016165e-06, + "loss": 0.4633, + "step": 19285 + }, + { + "epoch": 0.6082952215738843, + "grad_norm": 1.6344777345657349, + "learning_rate": 7.025475872072444e-06, + "loss": 0.4885, + "step": 19286 + }, + { + "epoch": 0.6083267623403249, + "grad_norm": 1.3131729364395142, + "learning_rate": 7.02450057018474e-06, + "loss": 0.7224, + "step": 19287 + }, + { + "epoch": 0.6083583031067655, + "grad_norm": 1.4024896621704102, + "learning_rate": 7.023525299348679e-06, + "loss": 0.4316, + "step": 19288 + }, + { + "epoch": 0.6083898438732062, + "grad_norm": 1.492321252822876, + "learning_rate": 7.022550059574446e-06, + "loss": 0.4522, + "step": 19289 + }, + { + "epoch": 0.6084213846396468, + "grad_norm": 1.3833887577056885, + "learning_rate": 7.021574850872217e-06, + "loss": 0.4669, + "step": 19290 + }, + { + "epoch": 0.6084529254060874, + "grad_norm": 1.5412368774414062, + "learning_rate": 7.020599673252164e-06, + "loss": 0.4479, + "step": 19291 + }, + { + "epoch": 0.6084844661725279, + "grad_norm": 1.5016980171203613, + "learning_rate": 7.019624526724466e-06, + "loss": 0.4981, + "step": 19292 + }, + { + "epoch": 0.6085160069389686, + "grad_norm": 1.791038155555725, + "learning_rate": 7.018649411299304e-06, + "loss": 0.408, + "step": 19293 + }, + { + "epoch": 0.6085475477054092, + "grad_norm": 3.2124621868133545, + "learning_rate": 7.017674326986848e-06, + "loss": 0.4529, + "step": 19294 + }, + { + "epoch": 0.6085790884718498, + "grad_norm": 1.3396713733673096, + "learning_rate": 7.016699273797277e-06, + "loss": 0.4381, + "step": 19295 + }, + { + "epoch": 0.6086106292382905, + "grad_norm": 1.6555891036987305, + "learning_rate": 7.015724251740766e-06, + "loss": 0.7395, + "step": 19296 + }, + { + "epoch": 0.6086421700047311, + "grad_norm": 1.350534439086914, + "learning_rate": 7.014749260827487e-06, + "loss": 0.4591, + "step": 19297 + }, + { + "epoch": 0.6086737107711717, + "grad_norm": 1.5771217346191406, + "learning_rate": 7.01377430106762e-06, + "loss": 0.4487, + "step": 19298 + }, + { + "epoch": 0.6087052515376123, + "grad_norm": 2.6679317951202393, + "learning_rate": 7.012799372471333e-06, + "loss": 0.4581, + "step": 19299 + }, + { + "epoch": 0.608736792304053, + "grad_norm": 1.4912195205688477, + "learning_rate": 7.011824475048807e-06, + "loss": 0.7457, + "step": 19300 + }, + { + "epoch": 0.6087683330704936, + "grad_norm": 1.7676938772201538, + "learning_rate": 7.010849608810215e-06, + "loss": 0.6814, + "step": 19301 + }, + { + "epoch": 0.6087998738369342, + "grad_norm": 1.6739262342453003, + "learning_rate": 7.009874773765726e-06, + "loss": 0.4474, + "step": 19302 + }, + { + "epoch": 0.6088314146033749, + "grad_norm": 1.457010269165039, + "learning_rate": 7.008899969925517e-06, + "loss": 0.4603, + "step": 19303 + }, + { + "epoch": 0.6088629553698155, + "grad_norm": 1.6924279928207397, + "learning_rate": 7.007925197299758e-06, + "loss": 0.4471, + "step": 19304 + }, + { + "epoch": 0.6088944961362561, + "grad_norm": 1.7420214414596558, + "learning_rate": 7.006950455898621e-06, + "loss": 0.395, + "step": 19305 + }, + { + "epoch": 0.6089260369026968, + "grad_norm": 1.370889663696289, + "learning_rate": 7.005975745732283e-06, + "loss": 0.4548, + "step": 19306 + }, + { + "epoch": 0.6089575776691374, + "grad_norm": 3.019193649291992, + "learning_rate": 7.005001066810914e-06, + "loss": 0.448, + "step": 19307 + }, + { + "epoch": 0.608989118435578, + "grad_norm": 1.7056679725646973, + "learning_rate": 7.004026419144678e-06, + "loss": 0.3363, + "step": 19308 + }, + { + "epoch": 0.6090206592020186, + "grad_norm": 1.6456986665725708, + "learning_rate": 7.003051802743759e-06, + "loss": 0.3889, + "step": 19309 + }, + { + "epoch": 0.6090521999684593, + "grad_norm": 1.6623752117156982, + "learning_rate": 7.0020772176183214e-06, + "loss": 0.4207, + "step": 19310 + }, + { + "epoch": 0.6090837407348999, + "grad_norm": 1.8748692274093628, + "learning_rate": 7.001102663778533e-06, + "loss": 0.7495, + "step": 19311 + }, + { + "epoch": 0.6091152815013405, + "grad_norm": 1.3016397953033447, + "learning_rate": 7.000128141234567e-06, + "loss": 0.4711, + "step": 19312 + }, + { + "epoch": 0.609146822267781, + "grad_norm": 2.479449510574341, + "learning_rate": 6.999153649996595e-06, + "loss": 0.4377, + "step": 19313 + }, + { + "epoch": 0.6091783630342217, + "grad_norm": 1.9137272834777832, + "learning_rate": 6.998179190074783e-06, + "loss": 0.4611, + "step": 19314 + }, + { + "epoch": 0.6092099038006623, + "grad_norm": 1.688621163368225, + "learning_rate": 6.997204761479304e-06, + "loss": 0.4544, + "step": 19315 + }, + { + "epoch": 0.6092414445671029, + "grad_norm": 1.2906033992767334, + "learning_rate": 6.99623036422032e-06, + "loss": 0.4255, + "step": 19316 + }, + { + "epoch": 0.6092729853335436, + "grad_norm": 1.3668758869171143, + "learning_rate": 6.99525599830801e-06, + "loss": 0.4592, + "step": 19317 + }, + { + "epoch": 0.6093045260999842, + "grad_norm": 1.5541878938674927, + "learning_rate": 6.994281663752536e-06, + "loss": 0.4003, + "step": 19318 + }, + { + "epoch": 0.6093360668664248, + "grad_norm": 1.755499005317688, + "learning_rate": 6.993307360564063e-06, + "loss": 0.4453, + "step": 19319 + }, + { + "epoch": 0.6093676076328655, + "grad_norm": 1.618553638458252, + "learning_rate": 6.992333088752766e-06, + "loss": 0.399, + "step": 19320 + }, + { + "epoch": 0.6093991483993061, + "grad_norm": 1.9517769813537598, + "learning_rate": 6.991358848328808e-06, + "loss": 0.3863, + "step": 19321 + }, + { + "epoch": 0.6094306891657467, + "grad_norm": 1.5062627792358398, + "learning_rate": 6.990384639302357e-06, + "loss": 0.431, + "step": 19322 + }, + { + "epoch": 0.6094622299321873, + "grad_norm": 1.444633960723877, + "learning_rate": 6.9894104616835786e-06, + "loss": 0.4505, + "step": 19323 + }, + { + "epoch": 0.609493770698628, + "grad_norm": 1.5162112712860107, + "learning_rate": 6.9884363154826415e-06, + "loss": 0.4251, + "step": 19324 + }, + { + "epoch": 0.6095253114650686, + "grad_norm": 1.4557126760482788, + "learning_rate": 6.987462200709705e-06, + "loss": 0.4503, + "step": 19325 + }, + { + "epoch": 0.6095568522315092, + "grad_norm": 1.4550797939300537, + "learning_rate": 6.986488117374943e-06, + "loss": 0.4193, + "step": 19326 + }, + { + "epoch": 0.6095883929979499, + "grad_norm": 1.4192349910736084, + "learning_rate": 6.985514065488519e-06, + "loss": 0.4307, + "step": 19327 + }, + { + "epoch": 0.6096199337643905, + "grad_norm": 1.7954139709472656, + "learning_rate": 6.98454004506059e-06, + "loss": 0.429, + "step": 19328 + }, + { + "epoch": 0.6096514745308311, + "grad_norm": 1.7194615602493286, + "learning_rate": 6.983566056101334e-06, + "loss": 0.4486, + "step": 19329 + }, + { + "epoch": 0.6096830152972718, + "grad_norm": 1.4805902242660522, + "learning_rate": 6.9825920986209054e-06, + "loss": 0.4113, + "step": 19330 + }, + { + "epoch": 0.6097145560637124, + "grad_norm": 1.449142336845398, + "learning_rate": 6.981618172629472e-06, + "loss": 0.4601, + "step": 19331 + }, + { + "epoch": 0.609746096830153, + "grad_norm": 1.46140456199646, + "learning_rate": 6.980644278137196e-06, + "loss": 0.4517, + "step": 19332 + }, + { + "epoch": 0.6097776375965936, + "grad_norm": 9.03232479095459, + "learning_rate": 6.979670415154241e-06, + "loss": 0.7126, + "step": 19333 + }, + { + "epoch": 0.6098091783630343, + "grad_norm": 1.8580248355865479, + "learning_rate": 6.978696583690771e-06, + "loss": 0.4213, + "step": 19334 + }, + { + "epoch": 0.6098407191294748, + "grad_norm": 1.487073302268982, + "learning_rate": 6.9777227837569504e-06, + "loss": 0.728, + "step": 19335 + }, + { + "epoch": 0.6098722598959154, + "grad_norm": 1.4803048372268677, + "learning_rate": 6.976749015362935e-06, + "loss": 0.4429, + "step": 19336 + }, + { + "epoch": 0.609903800662356, + "grad_norm": 1.8985998630523682, + "learning_rate": 6.975775278518895e-06, + "loss": 0.4497, + "step": 19337 + }, + { + "epoch": 0.6099353414287967, + "grad_norm": 1.7794013023376465, + "learning_rate": 6.974801573234986e-06, + "loss": 0.4293, + "step": 19338 + }, + { + "epoch": 0.6099668821952373, + "grad_norm": 1.4295473098754883, + "learning_rate": 6.9738278995213705e-06, + "loss": 0.741, + "step": 19339 + }, + { + "epoch": 0.6099984229616779, + "grad_norm": 1.6298671960830688, + "learning_rate": 6.972854257388213e-06, + "loss": 0.4633, + "step": 19340 + }, + { + "epoch": 0.6100299637281186, + "grad_norm": 1.3830108642578125, + "learning_rate": 6.971880646845672e-06, + "loss": 0.4431, + "step": 19341 + }, + { + "epoch": 0.6100615044945592, + "grad_norm": 1.9593942165374756, + "learning_rate": 6.970907067903907e-06, + "loss": 0.4443, + "step": 19342 + }, + { + "epoch": 0.6100930452609998, + "grad_norm": 1.616547703742981, + "learning_rate": 6.96993352057308e-06, + "loss": 0.4471, + "step": 19343 + }, + { + "epoch": 0.6101245860274405, + "grad_norm": 1.5583868026733398, + "learning_rate": 6.968960004863348e-06, + "loss": 0.4851, + "step": 19344 + }, + { + "epoch": 0.6101561267938811, + "grad_norm": 1.8809235095977783, + "learning_rate": 6.9679865207848695e-06, + "loss": 0.4539, + "step": 19345 + }, + { + "epoch": 0.6101876675603217, + "grad_norm": 1.8914289474487305, + "learning_rate": 6.96701306834781e-06, + "loss": 0.4494, + "step": 19346 + }, + { + "epoch": 0.6102192083267624, + "grad_norm": 1.4361836910247803, + "learning_rate": 6.966039647562319e-06, + "loss": 0.4399, + "step": 19347 + }, + { + "epoch": 0.610250749093203, + "grad_norm": 1.7098195552825928, + "learning_rate": 6.9650662584385655e-06, + "loss": 0.4711, + "step": 19348 + }, + { + "epoch": 0.6102822898596436, + "grad_norm": 1.5963594913482666, + "learning_rate": 6.964092900986701e-06, + "loss": 0.407, + "step": 19349 + }, + { + "epoch": 0.6103138306260842, + "grad_norm": 1.4676454067230225, + "learning_rate": 6.9631195752168835e-06, + "loss": 0.3883, + "step": 19350 + }, + { + "epoch": 0.6103453713925249, + "grad_norm": 2.1537587642669678, + "learning_rate": 6.962146281139273e-06, + "loss": 0.4634, + "step": 19351 + }, + { + "epoch": 0.6103769121589655, + "grad_norm": 1.4105501174926758, + "learning_rate": 6.9611730187640226e-06, + "loss": 0.4129, + "step": 19352 + }, + { + "epoch": 0.6104084529254061, + "grad_norm": 1.5777833461761475, + "learning_rate": 6.960199788101292e-06, + "loss": 0.4418, + "step": 19353 + }, + { + "epoch": 0.6104399936918468, + "grad_norm": 1.9680660963058472, + "learning_rate": 6.959226589161238e-06, + "loss": 0.3172, + "step": 19354 + }, + { + "epoch": 0.6104715344582874, + "grad_norm": 1.7516865730285645, + "learning_rate": 6.958253421954015e-06, + "loss": 0.4106, + "step": 19355 + }, + { + "epoch": 0.6105030752247279, + "grad_norm": 2.148573637008667, + "learning_rate": 6.9572802864897775e-06, + "loss": 0.4683, + "step": 19356 + }, + { + "epoch": 0.6105346159911685, + "grad_norm": 1.5247634649276733, + "learning_rate": 6.956307182778686e-06, + "loss": 0.4309, + "step": 19357 + }, + { + "epoch": 0.6105661567576092, + "grad_norm": 1.6459221839904785, + "learning_rate": 6.9553341108308915e-06, + "loss": 0.7246, + "step": 19358 + }, + { + "epoch": 0.6105976975240498, + "grad_norm": 1.7356655597686768, + "learning_rate": 6.954361070656546e-06, + "loss": 0.4596, + "step": 19359 + }, + { + "epoch": 0.6106292382904904, + "grad_norm": 2.800849676132202, + "learning_rate": 6.953388062265811e-06, + "loss": 0.7453, + "step": 19360 + }, + { + "epoch": 0.6106607790569311, + "grad_norm": 1.7539678812026978, + "learning_rate": 6.95241508566884e-06, + "loss": 0.4858, + "step": 19361 + }, + { + "epoch": 0.6106923198233717, + "grad_norm": 1.534010410308838, + "learning_rate": 6.95144214087578e-06, + "loss": 0.4341, + "step": 19362 + }, + { + "epoch": 0.6107238605898123, + "grad_norm": 1.4722802639007568, + "learning_rate": 6.95046922789679e-06, + "loss": 0.4525, + "step": 19363 + }, + { + "epoch": 0.610755401356253, + "grad_norm": 1.4142471551895142, + "learning_rate": 6.949496346742017e-06, + "loss": 0.4649, + "step": 19364 + }, + { + "epoch": 0.6107869421226936, + "grad_norm": 1.3126883506774902, + "learning_rate": 6.948523497421624e-06, + "loss": 0.4814, + "step": 19365 + }, + { + "epoch": 0.6108184828891342, + "grad_norm": 1.4655879735946655, + "learning_rate": 6.947550679945757e-06, + "loss": 0.4427, + "step": 19366 + }, + { + "epoch": 0.6108500236555748, + "grad_norm": 1.522559642791748, + "learning_rate": 6.9465778943245646e-06, + "loss": 0.4377, + "step": 19367 + }, + { + "epoch": 0.6108815644220155, + "grad_norm": 1.5584211349487305, + "learning_rate": 6.945605140568207e-06, + "loss": 0.4405, + "step": 19368 + }, + { + "epoch": 0.6109131051884561, + "grad_norm": 1.4576908349990845, + "learning_rate": 6.944632418686832e-06, + "loss": 0.3859, + "step": 19369 + }, + { + "epoch": 0.6109446459548967, + "grad_norm": 2.013402223587036, + "learning_rate": 6.943659728690587e-06, + "loss": 0.7653, + "step": 19370 + }, + { + "epoch": 0.6109761867213374, + "grad_norm": 1.3233102560043335, + "learning_rate": 6.9426870705896285e-06, + "loss": 0.451, + "step": 19371 + }, + { + "epoch": 0.611007727487778, + "grad_norm": 2.2246105670928955, + "learning_rate": 6.9417144443941035e-06, + "loss": 0.3837, + "step": 19372 + }, + { + "epoch": 0.6110392682542186, + "grad_norm": 1.7501695156097412, + "learning_rate": 6.940741850114161e-06, + "loss": 0.7484, + "step": 19373 + }, + { + "epoch": 0.6110708090206592, + "grad_norm": 1.4706794023513794, + "learning_rate": 6.9397692877599555e-06, + "loss": 0.4462, + "step": 19374 + }, + { + "epoch": 0.6111023497870999, + "grad_norm": 1.8523063659667969, + "learning_rate": 6.938796757341633e-06, + "loss": 0.4184, + "step": 19375 + }, + { + "epoch": 0.6111338905535405, + "grad_norm": 1.527726173400879, + "learning_rate": 6.937824258869341e-06, + "loss": 0.4357, + "step": 19376 + }, + { + "epoch": 0.611165431319981, + "grad_norm": 1.601014256477356, + "learning_rate": 6.936851792353233e-06, + "loss": 0.4337, + "step": 19377 + }, + { + "epoch": 0.6111969720864217, + "grad_norm": 1.536213755607605, + "learning_rate": 6.935879357803453e-06, + "loss": 0.4686, + "step": 19378 + }, + { + "epoch": 0.6112285128528623, + "grad_norm": 1.6925737857818604, + "learning_rate": 6.934906955230152e-06, + "loss": 0.4385, + "step": 19379 + }, + { + "epoch": 0.6112600536193029, + "grad_norm": 1.5579302310943604, + "learning_rate": 6.933934584643479e-06, + "loss": 0.4222, + "step": 19380 + }, + { + "epoch": 0.6112915943857435, + "grad_norm": 1.539344072341919, + "learning_rate": 6.932962246053577e-06, + "loss": 0.4752, + "step": 19381 + }, + { + "epoch": 0.6113231351521842, + "grad_norm": 2.181673765182495, + "learning_rate": 6.931989939470598e-06, + "loss": 0.4021, + "step": 19382 + }, + { + "epoch": 0.6113546759186248, + "grad_norm": 2.62324595451355, + "learning_rate": 6.931017664904685e-06, + "loss": 0.4216, + "step": 19383 + }, + { + "epoch": 0.6113862166850654, + "grad_norm": 1.6402223110198975, + "learning_rate": 6.930045422365983e-06, + "loss": 0.4382, + "step": 19384 + }, + { + "epoch": 0.6114177574515061, + "grad_norm": 1.9242761135101318, + "learning_rate": 6.929073211864644e-06, + "loss": 0.4033, + "step": 19385 + }, + { + "epoch": 0.6114492982179467, + "grad_norm": 1.473148226737976, + "learning_rate": 6.9281010334108114e-06, + "loss": 0.4406, + "step": 19386 + }, + { + "epoch": 0.6114808389843873, + "grad_norm": 1.6103434562683105, + "learning_rate": 6.9271288870146244e-06, + "loss": 0.397, + "step": 19387 + }, + { + "epoch": 0.611512379750828, + "grad_norm": 1.6529253721237183, + "learning_rate": 6.926156772686239e-06, + "loss": 0.4256, + "step": 19388 + }, + { + "epoch": 0.6115439205172686, + "grad_norm": 1.4996860027313232, + "learning_rate": 6.925184690435794e-06, + "loss": 0.4335, + "step": 19389 + }, + { + "epoch": 0.6115754612837092, + "grad_norm": 1.3389955759048462, + "learning_rate": 6.924212640273431e-06, + "loss": 0.4568, + "step": 19390 + }, + { + "epoch": 0.6116070020501498, + "grad_norm": 3.476640224456787, + "learning_rate": 6.9232406222092995e-06, + "loss": 0.4242, + "step": 19391 + }, + { + "epoch": 0.6116385428165905, + "grad_norm": 1.452170729637146, + "learning_rate": 6.922268636253541e-06, + "loss": 0.4372, + "step": 19392 + }, + { + "epoch": 0.6116700835830311, + "grad_norm": 2.0168044567108154, + "learning_rate": 6.921296682416298e-06, + "loss": 0.4675, + "step": 19393 + }, + { + "epoch": 0.6117016243494717, + "grad_norm": 1.8932708501815796, + "learning_rate": 6.920324760707718e-06, + "loss": 0.414, + "step": 19394 + }, + { + "epoch": 0.6117331651159124, + "grad_norm": 1.1910274028778076, + "learning_rate": 6.919352871137934e-06, + "loss": 0.3493, + "step": 19395 + }, + { + "epoch": 0.611764705882353, + "grad_norm": 1.569543480873108, + "learning_rate": 6.918381013717099e-06, + "loss": 0.4231, + "step": 19396 + }, + { + "epoch": 0.6117962466487936, + "grad_norm": 1.4846930503845215, + "learning_rate": 6.917409188455353e-06, + "loss": 0.4515, + "step": 19397 + }, + { + "epoch": 0.6118277874152341, + "grad_norm": 1.547248125076294, + "learning_rate": 6.91643739536283e-06, + "loss": 0.4476, + "step": 19398 + }, + { + "epoch": 0.6118593281816748, + "grad_norm": 1.697121024131775, + "learning_rate": 6.915465634449682e-06, + "loss": 0.425, + "step": 19399 + }, + { + "epoch": 0.6118908689481154, + "grad_norm": 1.685559868812561, + "learning_rate": 6.914493905726045e-06, + "loss": 0.4422, + "step": 19400 + }, + { + "epoch": 0.611922409714556, + "grad_norm": 1.4582931995391846, + "learning_rate": 6.913522209202057e-06, + "loss": 0.7353, + "step": 19401 + }, + { + "epoch": 0.6119539504809967, + "grad_norm": 1.6699600219726562, + "learning_rate": 6.912550544887865e-06, + "loss": 0.4183, + "step": 19402 + }, + { + "epoch": 0.6119854912474373, + "grad_norm": 1.8485389947891235, + "learning_rate": 6.9115789127936036e-06, + "loss": 0.7441, + "step": 19403 + }, + { + "epoch": 0.6120170320138779, + "grad_norm": 1.4972553253173828, + "learning_rate": 6.910607312929412e-06, + "loss": 0.4454, + "step": 19404 + }, + { + "epoch": 0.6120485727803185, + "grad_norm": 1.6893199682235718, + "learning_rate": 6.909635745305435e-06, + "loss": 0.4332, + "step": 19405 + }, + { + "epoch": 0.6120801135467592, + "grad_norm": 1.896878719329834, + "learning_rate": 6.908664209931809e-06, + "loss": 0.6909, + "step": 19406 + }, + { + "epoch": 0.6121116543131998, + "grad_norm": 3.828904628753662, + "learning_rate": 6.907692706818669e-06, + "loss": 0.3388, + "step": 19407 + }, + { + "epoch": 0.6121431950796404, + "grad_norm": 1.2816764116287231, + "learning_rate": 6.906721235976159e-06, + "loss": 0.4366, + "step": 19408 + }, + { + "epoch": 0.6121747358460811, + "grad_norm": 2.372950315475464, + "learning_rate": 6.905749797414416e-06, + "loss": 0.7522, + "step": 19409 + }, + { + "epoch": 0.6122062766125217, + "grad_norm": 1.6257760524749756, + "learning_rate": 6.904778391143574e-06, + "loss": 0.479, + "step": 19410 + }, + { + "epoch": 0.6122378173789623, + "grad_norm": 1.3652857542037964, + "learning_rate": 6.903807017173776e-06, + "loss": 0.4607, + "step": 19411 + }, + { + "epoch": 0.612269358145403, + "grad_norm": 1.463585615158081, + "learning_rate": 6.9028356755151495e-06, + "loss": 0.4457, + "step": 19412 + }, + { + "epoch": 0.6123008989118436, + "grad_norm": 2.7576498985290527, + "learning_rate": 6.901864366177844e-06, + "loss": 0.3788, + "step": 19413 + }, + { + "epoch": 0.6123324396782842, + "grad_norm": 1.59308660030365, + "learning_rate": 6.9008930891719884e-06, + "loss": 0.4607, + "step": 19414 + }, + { + "epoch": 0.6123639804447248, + "grad_norm": 1.1434558629989624, + "learning_rate": 6.899921844507716e-06, + "loss": 0.3542, + "step": 19415 + }, + { + "epoch": 0.6123955212111655, + "grad_norm": 1.5427954196929932, + "learning_rate": 6.898950632195169e-06, + "loss": 0.4509, + "step": 19416 + }, + { + "epoch": 0.6124270619776061, + "grad_norm": 1.4157909154891968, + "learning_rate": 6.897979452244482e-06, + "loss": 0.4771, + "step": 19417 + }, + { + "epoch": 0.6124586027440467, + "grad_norm": 1.5855427980422974, + "learning_rate": 6.897008304665783e-06, + "loss": 0.438, + "step": 19418 + }, + { + "epoch": 0.6124901435104873, + "grad_norm": 1.3888670206069946, + "learning_rate": 6.896037189469217e-06, + "loss": 0.4686, + "step": 19419 + }, + { + "epoch": 0.6125216842769279, + "grad_norm": 1.5427618026733398, + "learning_rate": 6.8950661066649115e-06, + "loss": 0.4131, + "step": 19420 + }, + { + "epoch": 0.6125532250433685, + "grad_norm": 1.592070460319519, + "learning_rate": 6.894095056263002e-06, + "loss": 0.4283, + "step": 19421 + }, + { + "epoch": 0.6125847658098091, + "grad_norm": 1.6491715908050537, + "learning_rate": 6.8931240382736216e-06, + "loss": 0.4176, + "step": 19422 + }, + { + "epoch": 0.6126163065762498, + "grad_norm": 1.4256088733673096, + "learning_rate": 6.892153052706907e-06, + "loss": 0.4365, + "step": 19423 + }, + { + "epoch": 0.6126478473426904, + "grad_norm": 1.2520370483398438, + "learning_rate": 6.891182099572983e-06, + "loss": 0.4424, + "step": 19424 + }, + { + "epoch": 0.612679388109131, + "grad_norm": 1.6713660955429077, + "learning_rate": 6.890211178881991e-06, + "loss": 0.7524, + "step": 19425 + }, + { + "epoch": 0.6127109288755717, + "grad_norm": 1.6544028520584106, + "learning_rate": 6.889240290644057e-06, + "loss": 0.7364, + "step": 19426 + }, + { + "epoch": 0.6127424696420123, + "grad_norm": 1.4816243648529053, + "learning_rate": 6.888269434869321e-06, + "loss": 0.7182, + "step": 19427 + }, + { + "epoch": 0.6127740104084529, + "grad_norm": 1.7413798570632935, + "learning_rate": 6.887298611567907e-06, + "loss": 0.3983, + "step": 19428 + }, + { + "epoch": 0.6128055511748935, + "grad_norm": 1.4339898824691772, + "learning_rate": 6.8863278207499496e-06, + "loss": 0.4246, + "step": 19429 + }, + { + "epoch": 0.6128370919413342, + "grad_norm": 1.2992128133773804, + "learning_rate": 6.88535706242558e-06, + "loss": 0.3456, + "step": 19430 + }, + { + "epoch": 0.6128686327077748, + "grad_norm": 9.616649627685547, + "learning_rate": 6.884386336604929e-06, + "loss": 0.4257, + "step": 19431 + }, + { + "epoch": 0.6129001734742154, + "grad_norm": 1.3745535612106323, + "learning_rate": 6.88341564329812e-06, + "loss": 0.433, + "step": 19432 + }, + { + "epoch": 0.6129317142406561, + "grad_norm": 1.3834471702575684, + "learning_rate": 6.882444982515294e-06, + "loss": 0.4526, + "step": 19433 + }, + { + "epoch": 0.6129632550070967, + "grad_norm": 1.6908968687057495, + "learning_rate": 6.881474354266575e-06, + "loss": 0.466, + "step": 19434 + }, + { + "epoch": 0.6129947957735373, + "grad_norm": 2.0187511444091797, + "learning_rate": 6.880503758562087e-06, + "loss": 0.4252, + "step": 19435 + }, + { + "epoch": 0.613026336539978, + "grad_norm": 1.5581034421920776, + "learning_rate": 6.879533195411972e-06, + "loss": 0.7297, + "step": 19436 + }, + { + "epoch": 0.6130578773064186, + "grad_norm": 1.4704296588897705, + "learning_rate": 6.878562664826348e-06, + "loss": 0.4883, + "step": 19437 + }, + { + "epoch": 0.6130894180728592, + "grad_norm": 1.8598480224609375, + "learning_rate": 6.8775921668153435e-06, + "loss": 0.4617, + "step": 19438 + }, + { + "epoch": 0.6131209588392998, + "grad_norm": 2.0485711097717285, + "learning_rate": 6.876621701389092e-06, + "loss": 0.4232, + "step": 19439 + }, + { + "epoch": 0.6131524996057404, + "grad_norm": 1.6445903778076172, + "learning_rate": 6.87565126855772e-06, + "loss": 0.4311, + "step": 19440 + }, + { + "epoch": 0.613184040372181, + "grad_norm": 1.7578258514404297, + "learning_rate": 6.87468086833135e-06, + "loss": 0.4459, + "step": 19441 + }, + { + "epoch": 0.6132155811386216, + "grad_norm": 1.4751501083374023, + "learning_rate": 6.873710500720115e-06, + "loss": 0.4448, + "step": 19442 + }, + { + "epoch": 0.6132471219050623, + "grad_norm": 1.3784887790679932, + "learning_rate": 6.872740165734133e-06, + "loss": 0.4159, + "step": 19443 + }, + { + "epoch": 0.6132786626715029, + "grad_norm": 1.8322416543960571, + "learning_rate": 6.87176986338354e-06, + "loss": 0.4453, + "step": 19444 + }, + { + "epoch": 0.6133102034379435, + "grad_norm": 1.441704511642456, + "learning_rate": 6.870799593678459e-06, + "loss": 0.441, + "step": 19445 + }, + { + "epoch": 0.6133417442043841, + "grad_norm": 2.562129259109497, + "learning_rate": 6.8698293566290076e-06, + "loss": 0.4193, + "step": 19446 + }, + { + "epoch": 0.6133732849708248, + "grad_norm": 1.3509808778762817, + "learning_rate": 6.868859152245323e-06, + "loss": 0.4259, + "step": 19447 + }, + { + "epoch": 0.6134048257372654, + "grad_norm": 1.9108585119247437, + "learning_rate": 6.867888980537525e-06, + "loss": 0.4385, + "step": 19448 + }, + { + "epoch": 0.613436366503706, + "grad_norm": 1.7063430547714233, + "learning_rate": 6.8669188415157324e-06, + "loss": 0.4448, + "step": 19449 + }, + { + "epoch": 0.6134679072701467, + "grad_norm": 1.23513925075531, + "learning_rate": 6.86594873519008e-06, + "loss": 0.446, + "step": 19450 + }, + { + "epoch": 0.6134994480365873, + "grad_norm": 4.660224914550781, + "learning_rate": 6.864978661570685e-06, + "loss": 0.744, + "step": 19451 + }, + { + "epoch": 0.6135309888030279, + "grad_norm": 1.433161973953247, + "learning_rate": 6.864008620667667e-06, + "loss": 0.481, + "step": 19452 + }, + { + "epoch": 0.6135625295694686, + "grad_norm": 1.507802963256836, + "learning_rate": 6.863038612491159e-06, + "loss": 0.4535, + "step": 19453 + }, + { + "epoch": 0.6135940703359092, + "grad_norm": 1.5770642757415771, + "learning_rate": 6.862068637051278e-06, + "loss": 0.4336, + "step": 19454 + }, + { + "epoch": 0.6136256111023498, + "grad_norm": 1.4043179750442505, + "learning_rate": 6.861098694358143e-06, + "loss": 0.4442, + "step": 19455 + }, + { + "epoch": 0.6136571518687904, + "grad_norm": 2.0461668968200684, + "learning_rate": 6.860128784421887e-06, + "loss": 0.4408, + "step": 19456 + }, + { + "epoch": 0.6136886926352311, + "grad_norm": 1.4637001752853394, + "learning_rate": 6.859158907252623e-06, + "loss": 0.4481, + "step": 19457 + }, + { + "epoch": 0.6137202334016717, + "grad_norm": 1.622402548789978, + "learning_rate": 6.8581890628604705e-06, + "loss": 0.7412, + "step": 19458 + }, + { + "epoch": 0.6137517741681123, + "grad_norm": 1.2929631471633911, + "learning_rate": 6.85721925125556e-06, + "loss": 0.7407, + "step": 19459 + }, + { + "epoch": 0.613783314934553, + "grad_norm": 2.830043315887451, + "learning_rate": 6.856249472448003e-06, + "loss": 0.4439, + "step": 19460 + }, + { + "epoch": 0.6138148557009936, + "grad_norm": 1.7273281812667847, + "learning_rate": 6.855279726447928e-06, + "loss": 0.4423, + "step": 19461 + }, + { + "epoch": 0.6138463964674341, + "grad_norm": 1.590492606163025, + "learning_rate": 6.8543100132654495e-06, + "loss": 0.4517, + "step": 19462 + }, + { + "epoch": 0.6138779372338747, + "grad_norm": 1.2042442560195923, + "learning_rate": 6.853340332910685e-06, + "loss": 0.3459, + "step": 19463 + }, + { + "epoch": 0.6139094780003154, + "grad_norm": 1.632871389389038, + "learning_rate": 6.852370685393761e-06, + "loss": 0.4644, + "step": 19464 + }, + { + "epoch": 0.613941018766756, + "grad_norm": 1.3537559509277344, + "learning_rate": 6.851401070724793e-06, + "loss": 0.4266, + "step": 19465 + }, + { + "epoch": 0.6139725595331966, + "grad_norm": 1.9321988821029663, + "learning_rate": 6.8504314889138956e-06, + "loss": 0.3754, + "step": 19466 + }, + { + "epoch": 0.6140041002996373, + "grad_norm": 1.682779312133789, + "learning_rate": 6.849461939971196e-06, + "loss": 0.3599, + "step": 19467 + }, + { + "epoch": 0.6140356410660779, + "grad_norm": 1.545783281326294, + "learning_rate": 6.848492423906807e-06, + "loss": 0.4784, + "step": 19468 + }, + { + "epoch": 0.6140671818325185, + "grad_norm": 1.5434300899505615, + "learning_rate": 6.8475229407308445e-06, + "loss": 0.4328, + "step": 19469 + }, + { + "epoch": 0.6140987225989591, + "grad_norm": 1.846652865409851, + "learning_rate": 6.8465534904534306e-06, + "loss": 0.4236, + "step": 19470 + }, + { + "epoch": 0.6141302633653998, + "grad_norm": 1.7250328063964844, + "learning_rate": 6.845584073084679e-06, + "loss": 0.4544, + "step": 19471 + }, + { + "epoch": 0.6141618041318404, + "grad_norm": 1.4559115171432495, + "learning_rate": 6.844614688634703e-06, + "loss": 0.4218, + "step": 19472 + }, + { + "epoch": 0.614193344898281, + "grad_norm": 1.4888291358947754, + "learning_rate": 6.8436453371136265e-06, + "loss": 0.4429, + "step": 19473 + }, + { + "epoch": 0.6142248856647217, + "grad_norm": 2.1423566341400146, + "learning_rate": 6.842676018531558e-06, + "loss": 0.4045, + "step": 19474 + }, + { + "epoch": 0.6142564264311623, + "grad_norm": 2.3644566535949707, + "learning_rate": 6.841706732898621e-06, + "loss": 0.464, + "step": 19475 + }, + { + "epoch": 0.6142879671976029, + "grad_norm": 1.4152913093566895, + "learning_rate": 6.840737480224926e-06, + "loss": 0.4646, + "step": 19476 + }, + { + "epoch": 0.6143195079640436, + "grad_norm": 1.4987543821334839, + "learning_rate": 6.839768260520587e-06, + "loss": 0.4687, + "step": 19477 + }, + { + "epoch": 0.6143510487304842, + "grad_norm": 1.8915961980819702, + "learning_rate": 6.838799073795721e-06, + "loss": 0.4367, + "step": 19478 + }, + { + "epoch": 0.6143825894969248, + "grad_norm": 1.7140682935714722, + "learning_rate": 6.837829920060442e-06, + "loss": 0.4918, + "step": 19479 + }, + { + "epoch": 0.6144141302633654, + "grad_norm": 1.7356537580490112, + "learning_rate": 6.83686079932486e-06, + "loss": 0.3798, + "step": 19480 + }, + { + "epoch": 0.6144456710298061, + "grad_norm": 1.8409795761108398, + "learning_rate": 6.835891711599096e-06, + "loss": 0.4738, + "step": 19481 + }, + { + "epoch": 0.6144772117962467, + "grad_norm": 1.7306259870529175, + "learning_rate": 6.834922656893259e-06, + "loss": 0.4428, + "step": 19482 + }, + { + "epoch": 0.6145087525626872, + "grad_norm": 1.4801794290542603, + "learning_rate": 6.833953635217456e-06, + "loss": 0.4627, + "step": 19483 + }, + { + "epoch": 0.6145402933291279, + "grad_norm": 1.7440484762191772, + "learning_rate": 6.832984646581811e-06, + "loss": 0.4669, + "step": 19484 + }, + { + "epoch": 0.6145718340955685, + "grad_norm": 1.8336807489395142, + "learning_rate": 6.8320156909964275e-06, + "loss": 0.3816, + "step": 19485 + }, + { + "epoch": 0.6146033748620091, + "grad_norm": 1.3184503316879272, + "learning_rate": 6.8310467684714185e-06, + "loss": 0.4388, + "step": 19486 + }, + { + "epoch": 0.6146349156284497, + "grad_norm": 1.517547369003296, + "learning_rate": 6.830077879016901e-06, + "loss": 0.425, + "step": 19487 + }, + { + "epoch": 0.6146664563948904, + "grad_norm": 1.5535374879837036, + "learning_rate": 6.829109022642981e-06, + "loss": 0.4549, + "step": 19488 + }, + { + "epoch": 0.614697997161331, + "grad_norm": 1.5192804336547852, + "learning_rate": 6.828140199359769e-06, + "loss": 0.4578, + "step": 19489 + }, + { + "epoch": 0.6147295379277716, + "grad_norm": 1.3317008018493652, + "learning_rate": 6.82717140917738e-06, + "loss": 0.4282, + "step": 19490 + }, + { + "epoch": 0.6147610786942123, + "grad_norm": 1.6813616752624512, + "learning_rate": 6.826202652105915e-06, + "loss": 0.4315, + "step": 19491 + }, + { + "epoch": 0.6147926194606529, + "grad_norm": 1.7700181007385254, + "learning_rate": 6.825233928155495e-06, + "loss": 0.4618, + "step": 19492 + }, + { + "epoch": 0.6148241602270935, + "grad_norm": 1.398441195487976, + "learning_rate": 6.824265237336224e-06, + "loss": 0.4283, + "step": 19493 + }, + { + "epoch": 0.6148557009935341, + "grad_norm": 1.515366554260254, + "learning_rate": 6.823296579658207e-06, + "loss": 0.4765, + "step": 19494 + }, + { + "epoch": 0.6148872417599748, + "grad_norm": 1.8466708660125732, + "learning_rate": 6.82232795513156e-06, + "loss": 0.4398, + "step": 19495 + }, + { + "epoch": 0.6149187825264154, + "grad_norm": 2.0732781887054443, + "learning_rate": 6.821359363766389e-06, + "loss": 0.4716, + "step": 19496 + }, + { + "epoch": 0.614950323292856, + "grad_norm": 1.6203781366348267, + "learning_rate": 6.8203908055728e-06, + "loss": 0.3966, + "step": 19497 + }, + { + "epoch": 0.6149818640592967, + "grad_norm": 2.555162191390991, + "learning_rate": 6.819422280560901e-06, + "loss": 0.4429, + "step": 19498 + }, + { + "epoch": 0.6150134048257373, + "grad_norm": 1.5263935327529907, + "learning_rate": 6.818453788740802e-06, + "loss": 0.7254, + "step": 19499 + }, + { + "epoch": 0.6150449455921779, + "grad_norm": 1.2950721979141235, + "learning_rate": 6.817485330122607e-06, + "loss": 0.4258, + "step": 19500 + }, + { + "epoch": 0.6150764863586186, + "grad_norm": 1.5390663146972656, + "learning_rate": 6.816516904716425e-06, + "loss": 0.469, + "step": 19501 + }, + { + "epoch": 0.6151080271250592, + "grad_norm": 1.3501410484313965, + "learning_rate": 6.81554851253236e-06, + "loss": 0.7024, + "step": 19502 + }, + { + "epoch": 0.6151395678914998, + "grad_norm": 1.3475247621536255, + "learning_rate": 6.814580153580517e-06, + "loss": 0.7445, + "step": 19503 + }, + { + "epoch": 0.6151711086579403, + "grad_norm": 1.5214427709579468, + "learning_rate": 6.813611827871006e-06, + "loss": 0.4653, + "step": 19504 + }, + { + "epoch": 0.615202649424381, + "grad_norm": 1.5898640155792236, + "learning_rate": 6.81264353541393e-06, + "loss": 0.4473, + "step": 19505 + }, + { + "epoch": 0.6152341901908216, + "grad_norm": 1.6739171743392944, + "learning_rate": 6.811675276219389e-06, + "loss": 0.3698, + "step": 19506 + }, + { + "epoch": 0.6152657309572622, + "grad_norm": 1.4770039319992065, + "learning_rate": 6.810707050297498e-06, + "loss": 0.4522, + "step": 19507 + }, + { + "epoch": 0.6152972717237029, + "grad_norm": 1.4014145135879517, + "learning_rate": 6.809738857658352e-06, + "loss": 0.443, + "step": 19508 + }, + { + "epoch": 0.6153288124901435, + "grad_norm": 1.5319496393203735, + "learning_rate": 6.8087706983120615e-06, + "loss": 0.4512, + "step": 19509 + }, + { + "epoch": 0.6153603532565841, + "grad_norm": 1.4385404586791992, + "learning_rate": 6.807802572268726e-06, + "loss": 0.7459, + "step": 19510 + }, + { + "epoch": 0.6153918940230247, + "grad_norm": 1.4102778434753418, + "learning_rate": 6.8068344795384445e-06, + "loss": 0.4417, + "step": 19511 + }, + { + "epoch": 0.6154234347894654, + "grad_norm": 1.5044071674346924, + "learning_rate": 6.80586642013133e-06, + "loss": 0.4171, + "step": 19512 + }, + { + "epoch": 0.615454975555906, + "grad_norm": 1.9527678489685059, + "learning_rate": 6.8048983940574785e-06, + "loss": 0.4113, + "step": 19513 + }, + { + "epoch": 0.6154865163223466, + "grad_norm": 1.6568940877914429, + "learning_rate": 6.803930401326989e-06, + "loss": 0.4676, + "step": 19514 + }, + { + "epoch": 0.6155180570887873, + "grad_norm": 1.353963017463684, + "learning_rate": 6.8029624419499715e-06, + "loss": 0.4401, + "step": 19515 + }, + { + "epoch": 0.6155495978552279, + "grad_norm": 1.571587085723877, + "learning_rate": 6.801994515936525e-06, + "loss": 0.4457, + "step": 19516 + }, + { + "epoch": 0.6155811386216685, + "grad_norm": 1.675076961517334, + "learning_rate": 6.801026623296745e-06, + "loss": 0.4511, + "step": 19517 + }, + { + "epoch": 0.6156126793881092, + "grad_norm": 1.6631112098693848, + "learning_rate": 6.8000587640407375e-06, + "loss": 0.4625, + "step": 19518 + }, + { + "epoch": 0.6156442201545498, + "grad_norm": 1.9196922779083252, + "learning_rate": 6.799090938178603e-06, + "loss": 0.4384, + "step": 19519 + }, + { + "epoch": 0.6156757609209904, + "grad_norm": 2.621809482574463, + "learning_rate": 6.798123145720438e-06, + "loss": 0.408, + "step": 19520 + }, + { + "epoch": 0.615707301687431, + "grad_norm": 1.614378571510315, + "learning_rate": 6.797155386676346e-06, + "loss": 0.4521, + "step": 19521 + }, + { + "epoch": 0.6157388424538717, + "grad_norm": 1.4866273403167725, + "learning_rate": 6.79618766105642e-06, + "loss": 0.425, + "step": 19522 + }, + { + "epoch": 0.6157703832203123, + "grad_norm": 1.6555198431015015, + "learning_rate": 6.795219968870768e-06, + "loss": 0.3255, + "step": 19523 + }, + { + "epoch": 0.6158019239867529, + "grad_norm": 2.319021463394165, + "learning_rate": 6.7942523101294846e-06, + "loss": 0.7426, + "step": 19524 + }, + { + "epoch": 0.6158334647531934, + "grad_norm": 1.612646222114563, + "learning_rate": 6.793284684842663e-06, + "loss": 0.4411, + "step": 19525 + }, + { + "epoch": 0.6158650055196341, + "grad_norm": 1.9340283870697021, + "learning_rate": 6.792317093020409e-06, + "loss": 0.4438, + "step": 19526 + }, + { + "epoch": 0.6158965462860747, + "grad_norm": 1.3900305032730103, + "learning_rate": 6.791349534672818e-06, + "loss": 0.4259, + "step": 19527 + }, + { + "epoch": 0.6159280870525153, + "grad_norm": 1.9304622411727905, + "learning_rate": 6.790382009809983e-06, + "loss": 0.4147, + "step": 19528 + }, + { + "epoch": 0.615959627818956, + "grad_norm": 1.482990026473999, + "learning_rate": 6.7894145184420065e-06, + "loss": 0.4489, + "step": 19529 + }, + { + "epoch": 0.6159911685853966, + "grad_norm": 2.166259288787842, + "learning_rate": 6.788447060578982e-06, + "loss": 0.7086, + "step": 19530 + }, + { + "epoch": 0.6160227093518372, + "grad_norm": 1.5188204050064087, + "learning_rate": 6.787479636231003e-06, + "loss": 0.4274, + "step": 19531 + }, + { + "epoch": 0.6160542501182779, + "grad_norm": 1.7686245441436768, + "learning_rate": 6.786512245408172e-06, + "loss": 0.4288, + "step": 19532 + }, + { + "epoch": 0.6160857908847185, + "grad_norm": 1.6417194604873657, + "learning_rate": 6.785544888120581e-06, + "loss": 0.4439, + "step": 19533 + }, + { + "epoch": 0.6161173316511591, + "grad_norm": 1.8543918132781982, + "learning_rate": 6.784577564378321e-06, + "loss": 0.3785, + "step": 19534 + }, + { + "epoch": 0.6161488724175997, + "grad_norm": 1.718341588973999, + "learning_rate": 6.783610274191494e-06, + "loss": 0.4251, + "step": 19535 + }, + { + "epoch": 0.6161804131840404, + "grad_norm": 1.627161979675293, + "learning_rate": 6.782643017570193e-06, + "loss": 0.4464, + "step": 19536 + }, + { + "epoch": 0.616211953950481, + "grad_norm": 1.4327139854431152, + "learning_rate": 6.781675794524507e-06, + "loss": 0.4604, + "step": 19537 + }, + { + "epoch": 0.6162434947169216, + "grad_norm": 1.7522064447402954, + "learning_rate": 6.7807086050645345e-06, + "loss": 0.4723, + "step": 19538 + }, + { + "epoch": 0.6162750354833623, + "grad_norm": 1.475357174873352, + "learning_rate": 6.779741449200365e-06, + "loss": 0.356, + "step": 19539 + }, + { + "epoch": 0.6163065762498029, + "grad_norm": 1.357789397239685, + "learning_rate": 6.778774326942098e-06, + "loss": 0.7375, + "step": 19540 + }, + { + "epoch": 0.6163381170162435, + "grad_norm": 1.4922255277633667, + "learning_rate": 6.777807238299822e-06, + "loss": 0.4689, + "step": 19541 + }, + { + "epoch": 0.6163696577826842, + "grad_norm": 1.6003715991973877, + "learning_rate": 6.776840183283624e-06, + "loss": 0.4524, + "step": 19542 + }, + { + "epoch": 0.6164011985491248, + "grad_norm": 1.7707489728927612, + "learning_rate": 6.7758731619036064e-06, + "loss": 0.4827, + "step": 19543 + }, + { + "epoch": 0.6164327393155654, + "grad_norm": 1.5645164251327515, + "learning_rate": 6.774906174169856e-06, + "loss": 0.4436, + "step": 19544 + }, + { + "epoch": 0.616464280082006, + "grad_norm": 1.5502567291259766, + "learning_rate": 6.7739392200924605e-06, + "loss": 0.4461, + "step": 19545 + }, + { + "epoch": 0.6164958208484466, + "grad_norm": 1.6575250625610352, + "learning_rate": 6.7729722996815166e-06, + "loss": 0.4552, + "step": 19546 + }, + { + "epoch": 0.6165273616148872, + "grad_norm": 1.636720895767212, + "learning_rate": 6.7720054129471136e-06, + "loss": 0.4217, + "step": 19547 + }, + { + "epoch": 0.6165589023813278, + "grad_norm": 1.4542428255081177, + "learning_rate": 6.771038559899337e-06, + "loss": 0.4672, + "step": 19548 + }, + { + "epoch": 0.6165904431477685, + "grad_norm": 1.6664233207702637, + "learning_rate": 6.770071740548284e-06, + "loss": 0.4527, + "step": 19549 + }, + { + "epoch": 0.6166219839142091, + "grad_norm": 1.5807862281799316, + "learning_rate": 6.76910495490404e-06, + "loss": 0.4646, + "step": 19550 + }, + { + "epoch": 0.6166535246806497, + "grad_norm": 1.9933006763458252, + "learning_rate": 6.768138202976691e-06, + "loss": 0.4604, + "step": 19551 + }, + { + "epoch": 0.6166850654470903, + "grad_norm": 1.5676252841949463, + "learning_rate": 6.767171484776332e-06, + "loss": 0.4192, + "step": 19552 + }, + { + "epoch": 0.616716606213531, + "grad_norm": 1.7620022296905518, + "learning_rate": 6.766204800313046e-06, + "loss": 0.4331, + "step": 19553 + }, + { + "epoch": 0.6167481469799716, + "grad_norm": 1.980765700340271, + "learning_rate": 6.765238149596926e-06, + "loss": 0.3235, + "step": 19554 + }, + { + "epoch": 0.6167796877464122, + "grad_norm": 1.7954312562942505, + "learning_rate": 6.764271532638059e-06, + "loss": 0.7144, + "step": 19555 + }, + { + "epoch": 0.6168112285128529, + "grad_norm": 2.7245068550109863, + "learning_rate": 6.76330494944653e-06, + "loss": 0.4063, + "step": 19556 + }, + { + "epoch": 0.6168427692792935, + "grad_norm": 1.576383113861084, + "learning_rate": 6.762338400032427e-06, + "loss": 0.4786, + "step": 19557 + }, + { + "epoch": 0.6168743100457341, + "grad_norm": 83.62984466552734, + "learning_rate": 6.761371884405837e-06, + "loss": 0.7347, + "step": 19558 + }, + { + "epoch": 0.6169058508121747, + "grad_norm": 1.8107637166976929, + "learning_rate": 6.760405402576844e-06, + "loss": 0.4526, + "step": 19559 + }, + { + "epoch": 0.6169373915786154, + "grad_norm": 1.4655568599700928, + "learning_rate": 6.75943895455554e-06, + "loss": 0.4721, + "step": 19560 + }, + { + "epoch": 0.616968932345056, + "grad_norm": 1.5253260135650635, + "learning_rate": 6.758472540352005e-06, + "loss": 0.4326, + "step": 19561 + }, + { + "epoch": 0.6170004731114966, + "grad_norm": 2.545281410217285, + "learning_rate": 6.7575061599763225e-06, + "loss": 0.463, + "step": 19562 + }, + { + "epoch": 0.6170320138779373, + "grad_norm": 1.4900282621383667, + "learning_rate": 6.756539813438585e-06, + "loss": 0.4473, + "step": 19563 + }, + { + "epoch": 0.6170635546443779, + "grad_norm": 2.2766916751861572, + "learning_rate": 6.755573500748873e-06, + "loss": 0.4339, + "step": 19564 + }, + { + "epoch": 0.6170950954108185, + "grad_norm": 1.5374782085418701, + "learning_rate": 6.7546072219172676e-06, + "loss": 0.4187, + "step": 19565 + }, + { + "epoch": 0.6171266361772592, + "grad_norm": 2.4196836948394775, + "learning_rate": 6.7536409769538595e-06, + "loss": 0.396, + "step": 19566 + }, + { + "epoch": 0.6171581769436997, + "grad_norm": 1.5085126161575317, + "learning_rate": 6.7526747658687276e-06, + "loss": 0.4947, + "step": 19567 + }, + { + "epoch": 0.6171897177101403, + "grad_norm": 1.7299526929855347, + "learning_rate": 6.751708588671955e-06, + "loss": 0.4485, + "step": 19568 + }, + { + "epoch": 0.6172212584765809, + "grad_norm": 1.5384814739227295, + "learning_rate": 6.750742445373628e-06, + "loss": 0.4433, + "step": 19569 + }, + { + "epoch": 0.6172527992430216, + "grad_norm": 1.931491494178772, + "learning_rate": 6.749776335983822e-06, + "loss": 0.4277, + "step": 19570 + }, + { + "epoch": 0.6172843400094622, + "grad_norm": 1.7032482624053955, + "learning_rate": 6.748810260512626e-06, + "loss": 0.4093, + "step": 19571 + }, + { + "epoch": 0.6173158807759028, + "grad_norm": 1.8970192670822144, + "learning_rate": 6.747844218970121e-06, + "loss": 0.6989, + "step": 19572 + }, + { + "epoch": 0.6173474215423435, + "grad_norm": 2.727137565612793, + "learning_rate": 6.746878211366384e-06, + "loss": 0.4208, + "step": 19573 + }, + { + "epoch": 0.6173789623087841, + "grad_norm": 1.7353882789611816, + "learning_rate": 6.7459122377115e-06, + "loss": 0.4921, + "step": 19574 + }, + { + "epoch": 0.6174105030752247, + "grad_norm": 1.6361982822418213, + "learning_rate": 6.744946298015551e-06, + "loss": 0.4657, + "step": 19575 + }, + { + "epoch": 0.6174420438416653, + "grad_norm": 1.9081305265426636, + "learning_rate": 6.743980392288611e-06, + "loss": 0.4104, + "step": 19576 + }, + { + "epoch": 0.617473584608106, + "grad_norm": 1.5974708795547485, + "learning_rate": 6.743014520540767e-06, + "loss": 0.4377, + "step": 19577 + }, + { + "epoch": 0.6175051253745466, + "grad_norm": 2.639688730239868, + "learning_rate": 6.742048682782096e-06, + "loss": 0.4439, + "step": 19578 + }, + { + "epoch": 0.6175366661409872, + "grad_norm": 1.903523564338684, + "learning_rate": 6.741082879022671e-06, + "loss": 0.4699, + "step": 19579 + }, + { + "epoch": 0.6175682069074279, + "grad_norm": 1.6565378904342651, + "learning_rate": 6.740117109272581e-06, + "loss": 0.4376, + "step": 19580 + }, + { + "epoch": 0.6175997476738685, + "grad_norm": 2.667593240737915, + "learning_rate": 6.739151373541902e-06, + "loss": 0.3415, + "step": 19581 + }, + { + "epoch": 0.6176312884403091, + "grad_norm": 1.7983962297439575, + "learning_rate": 6.738185671840704e-06, + "loss": 0.3934, + "step": 19582 + }, + { + "epoch": 0.6176628292067498, + "grad_norm": 1.6652307510375977, + "learning_rate": 6.737220004179077e-06, + "loss": 0.4556, + "step": 19583 + }, + { + "epoch": 0.6176943699731904, + "grad_norm": 2.0159287452697754, + "learning_rate": 6.7362543705670924e-06, + "loss": 0.4134, + "step": 19584 + }, + { + "epoch": 0.617725910739631, + "grad_norm": 2.574054718017578, + "learning_rate": 6.735288771014823e-06, + "loss": 0.7322, + "step": 19585 + }, + { + "epoch": 0.6177574515060716, + "grad_norm": 1.582343339920044, + "learning_rate": 6.734323205532356e-06, + "loss": 0.4594, + "step": 19586 + }, + { + "epoch": 0.6177889922725123, + "grad_norm": 1.8829121589660645, + "learning_rate": 6.733357674129759e-06, + "loss": 0.4244, + "step": 19587 + }, + { + "epoch": 0.6178205330389529, + "grad_norm": 6.711357116699219, + "learning_rate": 6.732392176817113e-06, + "loss": 0.4313, + "step": 19588 + }, + { + "epoch": 0.6178520738053934, + "grad_norm": 1.6680384874343872, + "learning_rate": 6.731426713604492e-06, + "loss": 0.4172, + "step": 19589 + }, + { + "epoch": 0.617883614571834, + "grad_norm": 1.8874218463897705, + "learning_rate": 6.730461284501966e-06, + "loss": 0.4071, + "step": 19590 + }, + { + "epoch": 0.6179151553382747, + "grad_norm": 1.881323218345642, + "learning_rate": 6.72949588951962e-06, + "loss": 0.449, + "step": 19591 + }, + { + "epoch": 0.6179466961047153, + "grad_norm": 1.9082019329071045, + "learning_rate": 6.728530528667525e-06, + "loss": 0.4238, + "step": 19592 + }, + { + "epoch": 0.6179782368711559, + "grad_norm": 1.8264793157577515, + "learning_rate": 6.727565201955748e-06, + "loss": 0.7184, + "step": 19593 + }, + { + "epoch": 0.6180097776375966, + "grad_norm": 6.527929782867432, + "learning_rate": 6.726599909394376e-06, + "loss": 0.4012, + "step": 19594 + }, + { + "epoch": 0.6180413184040372, + "grad_norm": 1.9409760236740112, + "learning_rate": 6.725634650993474e-06, + "loss": 0.4837, + "step": 19595 + }, + { + "epoch": 0.6180728591704778, + "grad_norm": 1.6783162355422974, + "learning_rate": 6.724669426763114e-06, + "loss": 0.4422, + "step": 19596 + }, + { + "epoch": 0.6181043999369185, + "grad_norm": 2.446373701095581, + "learning_rate": 6.723704236713376e-06, + "loss": 0.3576, + "step": 19597 + }, + { + "epoch": 0.6181359407033591, + "grad_norm": 1.960732102394104, + "learning_rate": 6.722739080854325e-06, + "loss": 0.4118, + "step": 19598 + }, + { + "epoch": 0.6181674814697997, + "grad_norm": 1.6547833681106567, + "learning_rate": 6.7217739591960354e-06, + "loss": 0.3873, + "step": 19599 + }, + { + "epoch": 0.6181990222362403, + "grad_norm": 1.5426541566848755, + "learning_rate": 6.720808871748583e-06, + "loss": 0.4502, + "step": 19600 + }, + { + "epoch": 0.618230563002681, + "grad_norm": 1.358934998512268, + "learning_rate": 6.7198438185220314e-06, + "loss": 0.4501, + "step": 19601 + }, + { + "epoch": 0.6182621037691216, + "grad_norm": 5.924266815185547, + "learning_rate": 6.718878799526462e-06, + "loss": 0.4523, + "step": 19602 + }, + { + "epoch": 0.6182936445355622, + "grad_norm": 1.568931221961975, + "learning_rate": 6.717913814771939e-06, + "loss": 0.4266, + "step": 19603 + }, + { + "epoch": 0.6183251853020029, + "grad_norm": 1.7043180465698242, + "learning_rate": 6.716948864268531e-06, + "loss": 0.4684, + "step": 19604 + }, + { + "epoch": 0.6183567260684435, + "grad_norm": 1.2842572927474976, + "learning_rate": 6.715983948026314e-06, + "loss": 0.421, + "step": 19605 + }, + { + "epoch": 0.6183882668348841, + "grad_norm": 1.3647699356079102, + "learning_rate": 6.715019066055355e-06, + "loss": 0.4204, + "step": 19606 + }, + { + "epoch": 0.6184198076013248, + "grad_norm": 1.2975045442581177, + "learning_rate": 6.7140542183657196e-06, + "loss": 0.4292, + "step": 19607 + }, + { + "epoch": 0.6184513483677654, + "grad_norm": 1.3858351707458496, + "learning_rate": 6.713089404967483e-06, + "loss": 0.4388, + "step": 19608 + }, + { + "epoch": 0.618482889134206, + "grad_norm": 1.7751054763793945, + "learning_rate": 6.71212462587071e-06, + "loss": 0.4201, + "step": 19609 + }, + { + "epoch": 0.6185144299006465, + "grad_norm": 1.7118734121322632, + "learning_rate": 6.711159881085466e-06, + "loss": 0.3938, + "step": 19610 + }, + { + "epoch": 0.6185459706670872, + "grad_norm": 1.3489073514938354, + "learning_rate": 6.710195170621827e-06, + "loss": 0.4454, + "step": 19611 + }, + { + "epoch": 0.6185775114335278, + "grad_norm": 1.6696492433547974, + "learning_rate": 6.709230494489854e-06, + "loss": 0.4221, + "step": 19612 + }, + { + "epoch": 0.6186090521999684, + "grad_norm": 3.8789572715759277, + "learning_rate": 6.708265852699614e-06, + "loss": 0.3607, + "step": 19613 + }, + { + "epoch": 0.618640592966409, + "grad_norm": 2.299511194229126, + "learning_rate": 6.707301245261179e-06, + "loss": 0.4775, + "step": 19614 + }, + { + "epoch": 0.6186721337328497, + "grad_norm": 1.244606614112854, + "learning_rate": 6.706336672184611e-06, + "loss": 0.3499, + "step": 19615 + }, + { + "epoch": 0.6187036744992903, + "grad_norm": 1.6582640409469604, + "learning_rate": 6.705372133479977e-06, + "loss": 0.7373, + "step": 19616 + }, + { + "epoch": 0.6187352152657309, + "grad_norm": 2.100170612335205, + "learning_rate": 6.704407629157343e-06, + "loss": 0.4665, + "step": 19617 + }, + { + "epoch": 0.6187667560321716, + "grad_norm": 2.194467782974243, + "learning_rate": 6.703443159226772e-06, + "loss": 0.3759, + "step": 19618 + }, + { + "epoch": 0.6187982967986122, + "grad_norm": 1.955531120300293, + "learning_rate": 6.702478723698336e-06, + "loss": 0.4498, + "step": 19619 + }, + { + "epoch": 0.6188298375650528, + "grad_norm": 2.2689011096954346, + "learning_rate": 6.701514322582095e-06, + "loss": 0.3907, + "step": 19620 + }, + { + "epoch": 0.6188613783314935, + "grad_norm": 1.4276272058486938, + "learning_rate": 6.700549955888107e-06, + "loss": 0.7369, + "step": 19621 + }, + { + "epoch": 0.6188929190979341, + "grad_norm": 2.3100407123565674, + "learning_rate": 6.699585623626448e-06, + "loss": 0.3486, + "step": 19622 + }, + { + "epoch": 0.6189244598643747, + "grad_norm": 1.7477513551712036, + "learning_rate": 6.698621325807176e-06, + "loss": 0.4512, + "step": 19623 + }, + { + "epoch": 0.6189560006308154, + "grad_norm": 3.0206854343414307, + "learning_rate": 6.6976570624403484e-06, + "loss": 0.4464, + "step": 19624 + }, + { + "epoch": 0.618987541397256, + "grad_norm": 1.546594262123108, + "learning_rate": 6.696692833536038e-06, + "loss": 0.4403, + "step": 19625 + }, + { + "epoch": 0.6190190821636966, + "grad_norm": 2.0279970169067383, + "learning_rate": 6.695728639104302e-06, + "loss": 0.4515, + "step": 19626 + }, + { + "epoch": 0.6190506229301372, + "grad_norm": 2.104434013366699, + "learning_rate": 6.694764479155202e-06, + "loss": 0.4256, + "step": 19627 + }, + { + "epoch": 0.6190821636965779, + "grad_norm": 1.3615422248840332, + "learning_rate": 6.693800353698805e-06, + "loss": 0.4241, + "step": 19628 + }, + { + "epoch": 0.6191137044630185, + "grad_norm": 1.6322063207626343, + "learning_rate": 6.692836262745166e-06, + "loss": 0.4696, + "step": 19629 + }, + { + "epoch": 0.6191452452294591, + "grad_norm": 1.272653341293335, + "learning_rate": 6.691872206304346e-06, + "loss": 0.7529, + "step": 19630 + }, + { + "epoch": 0.6191767859958996, + "grad_norm": 1.5935851335525513, + "learning_rate": 6.690908184386412e-06, + "loss": 0.4345, + "step": 19631 + }, + { + "epoch": 0.6192083267623403, + "grad_norm": 1.4307961463928223, + "learning_rate": 6.6899441970014216e-06, + "loss": 0.4523, + "step": 19632 + }, + { + "epoch": 0.6192398675287809, + "grad_norm": 1.6085114479064941, + "learning_rate": 6.688980244159427e-06, + "loss": 0.4437, + "step": 19633 + }, + { + "epoch": 0.6192714082952215, + "grad_norm": 1.409239649772644, + "learning_rate": 6.6880163258705e-06, + "loss": 0.7118, + "step": 19634 + }, + { + "epoch": 0.6193029490616622, + "grad_norm": 1.6111737489700317, + "learning_rate": 6.687052442144693e-06, + "loss": 0.4577, + "step": 19635 + }, + { + "epoch": 0.6193344898281028, + "grad_norm": 1.8897448778152466, + "learning_rate": 6.686088592992067e-06, + "loss": 0.4622, + "step": 19636 + }, + { + "epoch": 0.6193660305945434, + "grad_norm": 1.9406152963638306, + "learning_rate": 6.68512477842268e-06, + "loss": 0.4402, + "step": 19637 + }, + { + "epoch": 0.6193975713609841, + "grad_norm": 1.6988356113433838, + "learning_rate": 6.684160998446586e-06, + "loss": 0.4528, + "step": 19638 + }, + { + "epoch": 0.6194291121274247, + "grad_norm": 1.4447616338729858, + "learning_rate": 6.683197253073851e-06, + "loss": 0.4243, + "step": 19639 + }, + { + "epoch": 0.6194606528938653, + "grad_norm": 2.285000801086426, + "learning_rate": 6.682233542314528e-06, + "loss": 0.3715, + "step": 19640 + }, + { + "epoch": 0.619492193660306, + "grad_norm": 2.31498646736145, + "learning_rate": 6.68126986617867e-06, + "loss": 0.442, + "step": 19641 + }, + { + "epoch": 0.6195237344267466, + "grad_norm": 1.7043602466583252, + "learning_rate": 6.680306224676341e-06, + "loss": 0.4581, + "step": 19642 + }, + { + "epoch": 0.6195552751931872, + "grad_norm": 1.5270828008651733, + "learning_rate": 6.679342617817596e-06, + "loss": 0.4199, + "step": 19643 + }, + { + "epoch": 0.6195868159596278, + "grad_norm": 1.4996201992034912, + "learning_rate": 6.678379045612484e-06, + "loss": 0.4039, + "step": 19644 + }, + { + "epoch": 0.6196183567260685, + "grad_norm": 2.356121063232422, + "learning_rate": 6.67741550807107e-06, + "loss": 0.7434, + "step": 19645 + }, + { + "epoch": 0.6196498974925091, + "grad_norm": 1.4905283451080322, + "learning_rate": 6.6764520052034054e-06, + "loss": 0.7063, + "step": 19646 + }, + { + "epoch": 0.6196814382589497, + "grad_norm": 1.531722068786621, + "learning_rate": 6.675488537019543e-06, + "loss": 0.4528, + "step": 19647 + }, + { + "epoch": 0.6197129790253904, + "grad_norm": 1.3121387958526611, + "learning_rate": 6.67452510352954e-06, + "loss": 0.454, + "step": 19648 + }, + { + "epoch": 0.619744519791831, + "grad_norm": 1.9877822399139404, + "learning_rate": 6.673561704743447e-06, + "loss": 0.3948, + "step": 19649 + }, + { + "epoch": 0.6197760605582716, + "grad_norm": 1.3171526193618774, + "learning_rate": 6.672598340671323e-06, + "loss": 0.4368, + "step": 19650 + }, + { + "epoch": 0.6198076013247122, + "grad_norm": 1.885137915611267, + "learning_rate": 6.671635011323221e-06, + "loss": 0.4506, + "step": 19651 + }, + { + "epoch": 0.6198391420911528, + "grad_norm": 1.9040701389312744, + "learning_rate": 6.6706717167091875e-06, + "loss": 0.3906, + "step": 19652 + }, + { + "epoch": 0.6198706828575934, + "grad_norm": 1.62638521194458, + "learning_rate": 6.669708456839283e-06, + "loss": 0.4593, + "step": 19653 + }, + { + "epoch": 0.619902223624034, + "grad_norm": 1.652673363685608, + "learning_rate": 6.668745231723557e-06, + "loss": 0.4497, + "step": 19654 + }, + { + "epoch": 0.6199337643904747, + "grad_norm": 1.3271019458770752, + "learning_rate": 6.667782041372059e-06, + "loss": 0.3605, + "step": 19655 + }, + { + "epoch": 0.6199653051569153, + "grad_norm": 1.666245698928833, + "learning_rate": 6.666818885794845e-06, + "loss": 0.3895, + "step": 19656 + }, + { + "epoch": 0.6199968459233559, + "grad_norm": 1.9451510906219482, + "learning_rate": 6.665855765001963e-06, + "loss": 0.3644, + "step": 19657 + }, + { + "epoch": 0.6200283866897965, + "grad_norm": 1.3531458377838135, + "learning_rate": 6.664892679003463e-06, + "loss": 0.4619, + "step": 19658 + }, + { + "epoch": 0.6200599274562372, + "grad_norm": 1.7272177934646606, + "learning_rate": 6.6639296278094e-06, + "loss": 0.4391, + "step": 19659 + }, + { + "epoch": 0.6200914682226778, + "grad_norm": 1.469639539718628, + "learning_rate": 6.662966611429822e-06, + "loss": 0.4787, + "step": 19660 + }, + { + "epoch": 0.6201230089891184, + "grad_norm": 1.5298967361450195, + "learning_rate": 6.662003629874774e-06, + "loss": 0.4614, + "step": 19661 + }, + { + "epoch": 0.6201545497555591, + "grad_norm": 1.5787254571914673, + "learning_rate": 6.661040683154316e-06, + "loss": 0.4623, + "step": 19662 + }, + { + "epoch": 0.6201860905219997, + "grad_norm": 2.332944393157959, + "learning_rate": 6.66007777127849e-06, + "loss": 0.4952, + "step": 19663 + }, + { + "epoch": 0.6202176312884403, + "grad_norm": 1.7463802099227905, + "learning_rate": 6.659114894257342e-06, + "loss": 0.4032, + "step": 19664 + }, + { + "epoch": 0.620249172054881, + "grad_norm": 1.5369378328323364, + "learning_rate": 6.6581520521009275e-06, + "loss": 0.4399, + "step": 19665 + }, + { + "epoch": 0.6202807128213216, + "grad_norm": 1.8547824621200562, + "learning_rate": 6.657189244819289e-06, + "loss": 0.4393, + "step": 19666 + }, + { + "epoch": 0.6203122535877622, + "grad_norm": 1.6941927671432495, + "learning_rate": 6.65622647242248e-06, + "loss": 0.4513, + "step": 19667 + }, + { + "epoch": 0.6203437943542028, + "grad_norm": 1.6774466037750244, + "learning_rate": 6.655263734920543e-06, + "loss": 0.4322, + "step": 19668 + }, + { + "epoch": 0.6203753351206435, + "grad_norm": 1.5393891334533691, + "learning_rate": 6.654301032323522e-06, + "loss": 0.4135, + "step": 19669 + }, + { + "epoch": 0.6204068758870841, + "grad_norm": 1.3909519910812378, + "learning_rate": 6.653338364641472e-06, + "loss": 0.4576, + "step": 19670 + }, + { + "epoch": 0.6204384166535247, + "grad_norm": 1.516967535018921, + "learning_rate": 6.652375731884435e-06, + "loss": 0.422, + "step": 19671 + }, + { + "epoch": 0.6204699574199654, + "grad_norm": 1.5256150960922241, + "learning_rate": 6.6514131340624525e-06, + "loss": 0.4417, + "step": 19672 + }, + { + "epoch": 0.6205014981864059, + "grad_norm": 1.5328973531723022, + "learning_rate": 6.650450571185576e-06, + "loss": 0.4372, + "step": 19673 + }, + { + "epoch": 0.6205330389528465, + "grad_norm": 1.4260436296463013, + "learning_rate": 6.649488043263851e-06, + "loss": 0.446, + "step": 19674 + }, + { + "epoch": 0.6205645797192871, + "grad_norm": 1.369520664215088, + "learning_rate": 6.648525550307317e-06, + "loss": 0.4442, + "step": 19675 + }, + { + "epoch": 0.6205961204857278, + "grad_norm": 2.5759310722351074, + "learning_rate": 6.647563092326024e-06, + "loss": 0.4327, + "step": 19676 + }, + { + "epoch": 0.6206276612521684, + "grad_norm": 1.548408031463623, + "learning_rate": 6.646600669330012e-06, + "loss": 0.459, + "step": 19677 + }, + { + "epoch": 0.620659202018609, + "grad_norm": 1.441219449043274, + "learning_rate": 6.645638281329323e-06, + "loss": 0.461, + "step": 19678 + }, + { + "epoch": 0.6206907427850497, + "grad_norm": 1.3856748342514038, + "learning_rate": 6.644675928334007e-06, + "loss": 0.4292, + "step": 19679 + }, + { + "epoch": 0.6207222835514903, + "grad_norm": 1.754441499710083, + "learning_rate": 6.6437136103541034e-06, + "loss": 0.469, + "step": 19680 + }, + { + "epoch": 0.6207538243179309, + "grad_norm": 1.517865777015686, + "learning_rate": 6.642751327399649e-06, + "loss": 0.4397, + "step": 19681 + }, + { + "epoch": 0.6207853650843715, + "grad_norm": 1.866860032081604, + "learning_rate": 6.641789079480696e-06, + "loss": 0.4836, + "step": 19682 + }, + { + "epoch": 0.6208169058508122, + "grad_norm": 1.4673173427581787, + "learning_rate": 6.640826866607279e-06, + "loss": 0.3527, + "step": 19683 + }, + { + "epoch": 0.6208484466172528, + "grad_norm": 1.3523139953613281, + "learning_rate": 6.6398646887894436e-06, + "loss": 0.4257, + "step": 19684 + }, + { + "epoch": 0.6208799873836934, + "grad_norm": 2.142939329147339, + "learning_rate": 6.6389025460372294e-06, + "loss": 0.4858, + "step": 19685 + }, + { + "epoch": 0.6209115281501341, + "grad_norm": 1.714154601097107, + "learning_rate": 6.637940438360676e-06, + "loss": 0.454, + "step": 19686 + }, + { + "epoch": 0.6209430689165747, + "grad_norm": 1.7768824100494385, + "learning_rate": 6.636978365769826e-06, + "loss": 0.4135, + "step": 19687 + }, + { + "epoch": 0.6209746096830153, + "grad_norm": 1.7814953327178955, + "learning_rate": 6.6360163282747195e-06, + "loss": 0.3997, + "step": 19688 + }, + { + "epoch": 0.621006150449456, + "grad_norm": 1.5391082763671875, + "learning_rate": 6.63505432588539e-06, + "loss": 0.4391, + "step": 19689 + }, + { + "epoch": 0.6210376912158966, + "grad_norm": 4.197526454925537, + "learning_rate": 6.634092358611884e-06, + "loss": 0.4132, + "step": 19690 + }, + { + "epoch": 0.6210692319823372, + "grad_norm": 1.5891587734222412, + "learning_rate": 6.633130426464239e-06, + "loss": 0.4376, + "step": 19691 + }, + { + "epoch": 0.6211007727487778, + "grad_norm": 1.760313868522644, + "learning_rate": 6.632168529452489e-06, + "loss": 0.4642, + "step": 19692 + }, + { + "epoch": 0.6211323135152185, + "grad_norm": 1.5771889686584473, + "learning_rate": 6.631206667586679e-06, + "loss": 0.4415, + "step": 19693 + }, + { + "epoch": 0.621163854281659, + "grad_norm": 1.6446421146392822, + "learning_rate": 6.630244840876843e-06, + "loss": 0.4173, + "step": 19694 + }, + { + "epoch": 0.6211953950480996, + "grad_norm": 1.1984407901763916, + "learning_rate": 6.629283049333015e-06, + "loss": 0.3588, + "step": 19695 + }, + { + "epoch": 0.6212269358145402, + "grad_norm": 1.6165040731430054, + "learning_rate": 6.628321292965241e-06, + "loss": 0.4315, + "step": 19696 + }, + { + "epoch": 0.6212584765809809, + "grad_norm": 1.5363177061080933, + "learning_rate": 6.627359571783545e-06, + "loss": 0.4373, + "step": 19697 + }, + { + "epoch": 0.6212900173474215, + "grad_norm": 1.5325647592544556, + "learning_rate": 6.6263978857979775e-06, + "loss": 0.4121, + "step": 19698 + }, + { + "epoch": 0.6213215581138621, + "grad_norm": 2.029118537902832, + "learning_rate": 6.625436235018566e-06, + "loss": 0.4967, + "step": 19699 + }, + { + "epoch": 0.6213530988803028, + "grad_norm": 1.7896575927734375, + "learning_rate": 6.624474619455343e-06, + "loss": 0.4289, + "step": 19700 + }, + { + "epoch": 0.6213846396467434, + "grad_norm": 1.7485936880111694, + "learning_rate": 6.623513039118355e-06, + "loss": 0.4604, + "step": 19701 + }, + { + "epoch": 0.621416180413184, + "grad_norm": 1.4838792085647583, + "learning_rate": 6.622551494017628e-06, + "loss": 0.4455, + "step": 19702 + }, + { + "epoch": 0.6214477211796247, + "grad_norm": 2.623889446258545, + "learning_rate": 6.621589984163197e-06, + "loss": 0.7143, + "step": 19703 + }, + { + "epoch": 0.6214792619460653, + "grad_norm": 1.6750458478927612, + "learning_rate": 6.6206285095651e-06, + "loss": 0.4775, + "step": 19704 + }, + { + "epoch": 0.6215108027125059, + "grad_norm": 1.6504054069519043, + "learning_rate": 6.6196670702333666e-06, + "loss": 0.4017, + "step": 19705 + }, + { + "epoch": 0.6215423434789465, + "grad_norm": 1.1562504768371582, + "learning_rate": 6.618705666178031e-06, + "loss": 0.3547, + "step": 19706 + }, + { + "epoch": 0.6215738842453872, + "grad_norm": 1.886834740638733, + "learning_rate": 6.617744297409132e-06, + "loss": 0.4549, + "step": 19707 + }, + { + "epoch": 0.6216054250118278, + "grad_norm": 1.6477924585342407, + "learning_rate": 6.616782963936696e-06, + "loss": 0.4529, + "step": 19708 + }, + { + "epoch": 0.6216369657782684, + "grad_norm": 1.654529094696045, + "learning_rate": 6.615821665770752e-06, + "loss": 0.4496, + "step": 19709 + }, + { + "epoch": 0.6216685065447091, + "grad_norm": 1.364381194114685, + "learning_rate": 6.6148604029213416e-06, + "loss": 0.4042, + "step": 19710 + }, + { + "epoch": 0.6217000473111497, + "grad_norm": 1.5563970804214478, + "learning_rate": 6.6138991753984904e-06, + "loss": 0.4742, + "step": 19711 + }, + { + "epoch": 0.6217315880775903, + "grad_norm": 1.676757574081421, + "learning_rate": 6.6129379832122276e-06, + "loss": 0.4455, + "step": 19712 + }, + { + "epoch": 0.621763128844031, + "grad_norm": 1.4525587558746338, + "learning_rate": 6.61197682637259e-06, + "loss": 0.4589, + "step": 19713 + }, + { + "epoch": 0.6217946696104716, + "grad_norm": 2.1100289821624756, + "learning_rate": 6.611015704889604e-06, + "loss": 0.437, + "step": 19714 + }, + { + "epoch": 0.6218262103769122, + "grad_norm": 1.3524725437164307, + "learning_rate": 6.610054618773303e-06, + "loss": 0.678, + "step": 19715 + }, + { + "epoch": 0.6218577511433527, + "grad_norm": 1.6759306192398071, + "learning_rate": 6.609093568033713e-06, + "loss": 0.4535, + "step": 19716 + }, + { + "epoch": 0.6218892919097934, + "grad_norm": 1.8018720149993896, + "learning_rate": 6.60813255268086e-06, + "loss": 0.4113, + "step": 19717 + }, + { + "epoch": 0.621920832676234, + "grad_norm": 1.6906737089157104, + "learning_rate": 6.607171572724782e-06, + "loss": 0.3825, + "step": 19718 + }, + { + "epoch": 0.6219523734426746, + "grad_norm": 1.333364725112915, + "learning_rate": 6.606210628175505e-06, + "loss": 0.4602, + "step": 19719 + }, + { + "epoch": 0.6219839142091153, + "grad_norm": 1.6482383012771606, + "learning_rate": 6.60524971904305e-06, + "loss": 0.4295, + "step": 19720 + }, + { + "epoch": 0.6220154549755559, + "grad_norm": 1.6639351844787598, + "learning_rate": 6.604288845337453e-06, + "loss": 0.4442, + "step": 19721 + }, + { + "epoch": 0.6220469957419965, + "grad_norm": 1.7107818126678467, + "learning_rate": 6.60332800706874e-06, + "loss": 0.4468, + "step": 19722 + }, + { + "epoch": 0.6220785365084371, + "grad_norm": 1.8946584463119507, + "learning_rate": 6.602367204246934e-06, + "loss": 0.4268, + "step": 19723 + }, + { + "epoch": 0.6221100772748778, + "grad_norm": 1.536325454711914, + "learning_rate": 6.601406436882067e-06, + "loss": 0.4307, + "step": 19724 + }, + { + "epoch": 0.6221416180413184, + "grad_norm": 1.9673668146133423, + "learning_rate": 6.60044570498416e-06, + "loss": 0.4151, + "step": 19725 + }, + { + "epoch": 0.622173158807759, + "grad_norm": 1.8743805885314941, + "learning_rate": 6.599485008563242e-06, + "loss": 0.4439, + "step": 19726 + }, + { + "epoch": 0.6222046995741997, + "grad_norm": 1.4432731866836548, + "learning_rate": 6.598524347629341e-06, + "loss": 0.4224, + "step": 19727 + }, + { + "epoch": 0.6222362403406403, + "grad_norm": 1.8705636262893677, + "learning_rate": 6.597563722192475e-06, + "loss": 0.3374, + "step": 19728 + }, + { + "epoch": 0.6222677811070809, + "grad_norm": 2.9722793102264404, + "learning_rate": 6.596603132262678e-06, + "loss": 0.4482, + "step": 19729 + }, + { + "epoch": 0.6222993218735215, + "grad_norm": 1.533120036125183, + "learning_rate": 6.59564257784997e-06, + "loss": 0.723, + "step": 19730 + }, + { + "epoch": 0.6223308626399622, + "grad_norm": 1.7934988737106323, + "learning_rate": 6.59468205896437e-06, + "loss": 0.433, + "step": 19731 + }, + { + "epoch": 0.6223624034064028, + "grad_norm": 1.7758774757385254, + "learning_rate": 6.5937215756159125e-06, + "loss": 0.4441, + "step": 19732 + }, + { + "epoch": 0.6223939441728434, + "grad_norm": 2.7973060607910156, + "learning_rate": 6.592761127814615e-06, + "loss": 0.423, + "step": 19733 + }, + { + "epoch": 0.6224254849392841, + "grad_norm": 1.4883172512054443, + "learning_rate": 6.591800715570497e-06, + "loss": 0.7608, + "step": 19734 + }, + { + "epoch": 0.6224570257057247, + "grad_norm": 1.7829087972640991, + "learning_rate": 6.590840338893589e-06, + "loss": 0.4445, + "step": 19735 + }, + { + "epoch": 0.6224885664721653, + "grad_norm": 1.7348783016204834, + "learning_rate": 6.589879997793909e-06, + "loss": 0.3921, + "step": 19736 + }, + { + "epoch": 0.6225201072386058, + "grad_norm": 1.244350552558899, + "learning_rate": 6.588919692281474e-06, + "loss": 0.362, + "step": 19737 + }, + { + "epoch": 0.6225516480050465, + "grad_norm": 1.5090184211730957, + "learning_rate": 6.5879594223663145e-06, + "loss": 0.7356, + "step": 19738 + }, + { + "epoch": 0.6225831887714871, + "grad_norm": 3.2395904064178467, + "learning_rate": 6.586999188058449e-06, + "loss": 0.3732, + "step": 19739 + }, + { + "epoch": 0.6226147295379277, + "grad_norm": 1.3707680702209473, + "learning_rate": 6.586038989367892e-06, + "loss": 0.4546, + "step": 19740 + }, + { + "epoch": 0.6226462703043684, + "grad_norm": 1.4569437503814697, + "learning_rate": 6.585078826304674e-06, + "loss": 0.4523, + "step": 19741 + }, + { + "epoch": 0.622677811070809, + "grad_norm": 1.409085750579834, + "learning_rate": 6.5841186988788094e-06, + "loss": 0.4435, + "step": 19742 + }, + { + "epoch": 0.6227093518372496, + "grad_norm": 1.6750048398971558, + "learning_rate": 6.583158607100317e-06, + "loss": 0.4577, + "step": 19743 + }, + { + "epoch": 0.6227408926036903, + "grad_norm": 1.42386794090271, + "learning_rate": 6.5821985509792195e-06, + "loss": 0.4068, + "step": 19744 + }, + { + "epoch": 0.6227724333701309, + "grad_norm": 2.669870615005493, + "learning_rate": 6.58123853052553e-06, + "loss": 0.4377, + "step": 19745 + }, + { + "epoch": 0.6228039741365715, + "grad_norm": 1.791445255279541, + "learning_rate": 6.580278545749274e-06, + "loss": 0.4575, + "step": 19746 + }, + { + "epoch": 0.6228355149030121, + "grad_norm": 1.4090536832809448, + "learning_rate": 6.579318596660469e-06, + "loss": 0.4289, + "step": 19747 + }, + { + "epoch": 0.6228670556694528, + "grad_norm": 1.8196101188659668, + "learning_rate": 6.578358683269125e-06, + "loss": 0.3798, + "step": 19748 + }, + { + "epoch": 0.6228985964358934, + "grad_norm": 1.4505350589752197, + "learning_rate": 6.577398805585269e-06, + "loss": 0.4404, + "step": 19749 + }, + { + "epoch": 0.622930137202334, + "grad_norm": 1.629850149154663, + "learning_rate": 6.5764389636189145e-06, + "loss": 0.4012, + "step": 19750 + }, + { + "epoch": 0.6229616779687747, + "grad_norm": 1.2036665678024292, + "learning_rate": 6.5754791573800735e-06, + "loss": 0.3036, + "step": 19751 + }, + { + "epoch": 0.6229932187352153, + "grad_norm": 1.6593599319458008, + "learning_rate": 6.5745193868787706e-06, + "loss": 0.4401, + "step": 19752 + }, + { + "epoch": 0.6230247595016559, + "grad_norm": 1.341024398803711, + "learning_rate": 6.573559652125017e-06, + "loss": 0.7152, + "step": 19753 + }, + { + "epoch": 0.6230563002680966, + "grad_norm": 1.491174578666687, + "learning_rate": 6.5725999531288275e-06, + "loss": 0.4426, + "step": 19754 + }, + { + "epoch": 0.6230878410345372, + "grad_norm": 1.4332382678985596, + "learning_rate": 6.5716402899002205e-06, + "loss": 0.4461, + "step": 19755 + }, + { + "epoch": 0.6231193818009778, + "grad_norm": 1.6057708263397217, + "learning_rate": 6.57068066244921e-06, + "loss": 0.471, + "step": 19756 + }, + { + "epoch": 0.6231509225674184, + "grad_norm": 2.5203371047973633, + "learning_rate": 6.569721070785805e-06, + "loss": 0.4785, + "step": 19757 + }, + { + "epoch": 0.623182463333859, + "grad_norm": 1.970165729522705, + "learning_rate": 6.568761514920028e-06, + "loss": 0.4888, + "step": 19758 + }, + { + "epoch": 0.6232140041002996, + "grad_norm": 2.357815980911255, + "learning_rate": 6.56780199486189e-06, + "loss": 0.7298, + "step": 19759 + }, + { + "epoch": 0.6232455448667402, + "grad_norm": 1.7729088068008423, + "learning_rate": 6.566842510621398e-06, + "loss": 0.4209, + "step": 19760 + }, + { + "epoch": 0.6232770856331808, + "grad_norm": 1.74131441116333, + "learning_rate": 6.565883062208574e-06, + "loss": 0.4418, + "step": 19761 + }, + { + "epoch": 0.6233086263996215, + "grad_norm": 1.431229591369629, + "learning_rate": 6.564923649633426e-06, + "loss": 0.4294, + "step": 19762 + }, + { + "epoch": 0.6233401671660621, + "grad_norm": 2.9946060180664062, + "learning_rate": 6.563964272905968e-06, + "loss": 0.4118, + "step": 19763 + }, + { + "epoch": 0.6233717079325027, + "grad_norm": 1.7959591150283813, + "learning_rate": 6.5630049320362126e-06, + "loss": 0.4602, + "step": 19764 + }, + { + "epoch": 0.6234032486989434, + "grad_norm": 1.432012915611267, + "learning_rate": 6.562045627034164e-06, + "loss": 0.439, + "step": 19765 + }, + { + "epoch": 0.623434789465384, + "grad_norm": 1.5286647081375122, + "learning_rate": 6.561086357909841e-06, + "loss": 0.4614, + "step": 19766 + }, + { + "epoch": 0.6234663302318246, + "grad_norm": 1.3936793804168701, + "learning_rate": 6.560127124673255e-06, + "loss": 0.4396, + "step": 19767 + }, + { + "epoch": 0.6234978709982653, + "grad_norm": 1.5369699001312256, + "learning_rate": 6.559167927334409e-06, + "loss": 0.4585, + "step": 19768 + }, + { + "epoch": 0.6235294117647059, + "grad_norm": 1.7851126194000244, + "learning_rate": 6.55820876590332e-06, + "loss": 0.4365, + "step": 19769 + }, + { + "epoch": 0.6235609525311465, + "grad_norm": 1.6070489883422852, + "learning_rate": 6.557249640389995e-06, + "loss": 0.4646, + "step": 19770 + }, + { + "epoch": 0.6235924932975871, + "grad_norm": 1.6427315473556519, + "learning_rate": 6.556290550804441e-06, + "loss": 0.3325, + "step": 19771 + }, + { + "epoch": 0.6236240340640278, + "grad_norm": 1.5068448781967163, + "learning_rate": 6.555331497156671e-06, + "loss": 0.4414, + "step": 19772 + }, + { + "epoch": 0.6236555748304684, + "grad_norm": 5.165628910064697, + "learning_rate": 6.554372479456694e-06, + "loss": 0.7365, + "step": 19773 + }, + { + "epoch": 0.623687115596909, + "grad_norm": 1.737457036972046, + "learning_rate": 6.553413497714511e-06, + "loss": 0.3884, + "step": 19774 + }, + { + "epoch": 0.6237186563633497, + "grad_norm": 1.693796992301941, + "learning_rate": 6.552454551940138e-06, + "loss": 0.4191, + "step": 19775 + }, + { + "epoch": 0.6237501971297903, + "grad_norm": 1.5567147731781006, + "learning_rate": 6.5514956421435746e-06, + "loss": 0.4203, + "step": 19776 + }, + { + "epoch": 0.6237817378962309, + "grad_norm": 2.033585548400879, + "learning_rate": 6.550536768334836e-06, + "loss": 0.4136, + "step": 19777 + }, + { + "epoch": 0.6238132786626716, + "grad_norm": 2.4654061794281006, + "learning_rate": 6.549577930523925e-06, + "loss": 0.4419, + "step": 19778 + }, + { + "epoch": 0.6238448194291121, + "grad_norm": 1.5872490406036377, + "learning_rate": 6.548619128720841e-06, + "loss": 0.4627, + "step": 19779 + }, + { + "epoch": 0.6238763601955527, + "grad_norm": 1.4423589706420898, + "learning_rate": 6.547660362935603e-06, + "loss": 0.4274, + "step": 19780 + }, + { + "epoch": 0.6239079009619933, + "grad_norm": 1.2998359203338623, + "learning_rate": 6.546701633178209e-06, + "loss": 0.4364, + "step": 19781 + }, + { + "epoch": 0.623939441728434, + "grad_norm": 1.5656002759933472, + "learning_rate": 6.545742939458662e-06, + "loss": 0.456, + "step": 19782 + }, + { + "epoch": 0.6239709824948746, + "grad_norm": 1.792549729347229, + "learning_rate": 6.544784281786974e-06, + "loss": 0.4483, + "step": 19783 + }, + { + "epoch": 0.6240025232613152, + "grad_norm": 1.8125299215316772, + "learning_rate": 6.543825660173142e-06, + "loss": 0.44, + "step": 19784 + }, + { + "epoch": 0.6240340640277559, + "grad_norm": 1.527101993560791, + "learning_rate": 6.5428670746271706e-06, + "loss": 0.4547, + "step": 19785 + }, + { + "epoch": 0.6240656047941965, + "grad_norm": 1.4833449125289917, + "learning_rate": 6.541908525159069e-06, + "loss": 0.4133, + "step": 19786 + }, + { + "epoch": 0.6240971455606371, + "grad_norm": 1.7199424505233765, + "learning_rate": 6.540950011778837e-06, + "loss": 0.4641, + "step": 19787 + }, + { + "epoch": 0.6241286863270777, + "grad_norm": 1.7831625938415527, + "learning_rate": 6.539991534496474e-06, + "loss": 0.4363, + "step": 19788 + }, + { + "epoch": 0.6241602270935184, + "grad_norm": 2.6012957096099854, + "learning_rate": 6.5390330933219895e-06, + "loss": 0.7286, + "step": 19789 + }, + { + "epoch": 0.624191767859959, + "grad_norm": 1.9956542253494263, + "learning_rate": 6.538074688265382e-06, + "loss": 0.3914, + "step": 19790 + }, + { + "epoch": 0.6242233086263996, + "grad_norm": 1.8143301010131836, + "learning_rate": 6.537116319336649e-06, + "loss": 0.4615, + "step": 19791 + }, + { + "epoch": 0.6242548493928403, + "grad_norm": 1.4788978099822998, + "learning_rate": 6.536157986545801e-06, + "loss": 0.4298, + "step": 19792 + }, + { + "epoch": 0.6242863901592809, + "grad_norm": 1.6227974891662598, + "learning_rate": 6.535199689902831e-06, + "loss": 0.4469, + "step": 19793 + }, + { + "epoch": 0.6243179309257215, + "grad_norm": 1.38554847240448, + "learning_rate": 6.534241429417747e-06, + "loss": 0.3866, + "step": 19794 + }, + { + "epoch": 0.6243494716921622, + "grad_norm": 1.3135536909103394, + "learning_rate": 6.5332832051005425e-06, + "loss": 0.4158, + "step": 19795 + }, + { + "epoch": 0.6243810124586028, + "grad_norm": 1.9188719987869263, + "learning_rate": 6.532325016961217e-06, + "loss": 0.4464, + "step": 19796 + }, + { + "epoch": 0.6244125532250434, + "grad_norm": 1.37442147731781, + "learning_rate": 6.531366865009774e-06, + "loss": 0.4196, + "step": 19797 + }, + { + "epoch": 0.624444093991484, + "grad_norm": 1.4733188152313232, + "learning_rate": 6.530408749256214e-06, + "loss": 0.4269, + "step": 19798 + }, + { + "epoch": 0.6244756347579247, + "grad_norm": 1.5733686685562134, + "learning_rate": 6.529450669710526e-06, + "loss": 0.427, + "step": 19799 + }, + { + "epoch": 0.6245071755243652, + "grad_norm": 1.526336669921875, + "learning_rate": 6.528492626382721e-06, + "loss": 0.4585, + "step": 19800 + }, + { + "epoch": 0.6245387162908058, + "grad_norm": 1.792506217956543, + "learning_rate": 6.527534619282791e-06, + "loss": 0.4178, + "step": 19801 + }, + { + "epoch": 0.6245702570572464, + "grad_norm": 1.651821494102478, + "learning_rate": 6.52657664842073e-06, + "loss": 0.4614, + "step": 19802 + }, + { + "epoch": 0.6246017978236871, + "grad_norm": 1.5962430238723755, + "learning_rate": 6.525618713806543e-06, + "loss": 0.6908, + "step": 19803 + }, + { + "epoch": 0.6246333385901277, + "grad_norm": 1.5209314823150635, + "learning_rate": 6.524660815450221e-06, + "loss": 0.4507, + "step": 19804 + }, + { + "epoch": 0.6246648793565683, + "grad_norm": 1.9948527812957764, + "learning_rate": 6.523702953361758e-06, + "loss": 0.4409, + "step": 19805 + }, + { + "epoch": 0.624696420123009, + "grad_norm": 1.383170485496521, + "learning_rate": 6.522745127551158e-06, + "loss": 0.4538, + "step": 19806 + }, + { + "epoch": 0.6247279608894496, + "grad_norm": 1.733099341392517, + "learning_rate": 6.521787338028413e-06, + "loss": 0.7488, + "step": 19807 + }, + { + "epoch": 0.6247595016558902, + "grad_norm": 1.5760844945907593, + "learning_rate": 6.520829584803514e-06, + "loss": 0.4331, + "step": 19808 + }, + { + "epoch": 0.6247910424223309, + "grad_norm": 1.4244426488876343, + "learning_rate": 6.519871867886463e-06, + "loss": 0.483, + "step": 19809 + }, + { + "epoch": 0.6248225831887715, + "grad_norm": 1.2783681154251099, + "learning_rate": 6.518914187287249e-06, + "loss": 0.7415, + "step": 19810 + }, + { + "epoch": 0.6248541239552121, + "grad_norm": 1.5399754047393799, + "learning_rate": 6.5179565430158706e-06, + "loss": 0.4651, + "step": 19811 + }, + { + "epoch": 0.6248856647216527, + "grad_norm": 1.6264781951904297, + "learning_rate": 6.516998935082319e-06, + "loss": 0.4426, + "step": 19812 + }, + { + "epoch": 0.6249172054880934, + "grad_norm": 2.2600343227386475, + "learning_rate": 6.516041363496588e-06, + "loss": 0.4532, + "step": 19813 + }, + { + "epoch": 0.624948746254534, + "grad_norm": 1.7091169357299805, + "learning_rate": 6.515083828268671e-06, + "loss": 0.4695, + "step": 19814 + }, + { + "epoch": 0.6249802870209746, + "grad_norm": 2.3454208374023438, + "learning_rate": 6.514126329408561e-06, + "loss": 0.4294, + "step": 19815 + }, + { + "epoch": 0.6250118277874153, + "grad_norm": 1.5904016494750977, + "learning_rate": 6.513168866926245e-06, + "loss": 0.4224, + "step": 19816 + }, + { + "epoch": 0.6250433685538559, + "grad_norm": 1.5959351062774658, + "learning_rate": 6.512211440831724e-06, + "loss": 0.4371, + "step": 19817 + }, + { + "epoch": 0.6250749093202965, + "grad_norm": 1.33272385597229, + "learning_rate": 6.511254051134985e-06, + "loss": 0.4441, + "step": 19818 + }, + { + "epoch": 0.6251064500867372, + "grad_norm": 1.5378494262695312, + "learning_rate": 6.510296697846014e-06, + "loss": 0.7262, + "step": 19819 + }, + { + "epoch": 0.6251379908531778, + "grad_norm": 2.7732107639312744, + "learning_rate": 6.509339380974811e-06, + "loss": 0.4338, + "step": 19820 + }, + { + "epoch": 0.6251695316196183, + "grad_norm": 1.535967469215393, + "learning_rate": 6.508382100531363e-06, + "loss": 0.4313, + "step": 19821 + }, + { + "epoch": 0.6252010723860589, + "grad_norm": 1.2896522283554077, + "learning_rate": 6.507424856525655e-06, + "loss": 0.4321, + "step": 19822 + }, + { + "epoch": 0.6252326131524996, + "grad_norm": 1.6264568567276, + "learning_rate": 6.506467648967683e-06, + "loss": 0.4729, + "step": 19823 + }, + { + "epoch": 0.6252641539189402, + "grad_norm": 1.5031116008758545, + "learning_rate": 6.505510477867431e-06, + "loss": 0.4236, + "step": 19824 + }, + { + "epoch": 0.6252956946853808, + "grad_norm": 1.7312461137771606, + "learning_rate": 6.504553343234894e-06, + "loss": 0.3492, + "step": 19825 + }, + { + "epoch": 0.6253272354518215, + "grad_norm": 1.4641990661621094, + "learning_rate": 6.503596245080057e-06, + "loss": 0.4683, + "step": 19826 + }, + { + "epoch": 0.6253587762182621, + "grad_norm": 1.6159132719039917, + "learning_rate": 6.502639183412903e-06, + "loss": 0.4353, + "step": 19827 + }, + { + "epoch": 0.6253903169847027, + "grad_norm": 1.3171535730361938, + "learning_rate": 6.5016821582434294e-06, + "loss": 0.4401, + "step": 19828 + }, + { + "epoch": 0.6254218577511433, + "grad_norm": 1.6616512537002563, + "learning_rate": 6.50072516958162e-06, + "loss": 0.4039, + "step": 19829 + }, + { + "epoch": 0.625453398517584, + "grad_norm": 1.968029499053955, + "learning_rate": 6.499768217437456e-06, + "loss": 0.4234, + "step": 19830 + }, + { + "epoch": 0.6254849392840246, + "grad_norm": 4.725882053375244, + "learning_rate": 6.4988113018209305e-06, + "loss": 0.44, + "step": 19831 + }, + { + "epoch": 0.6255164800504652, + "grad_norm": 1.49965500831604, + "learning_rate": 6.497854422742031e-06, + "loss": 0.4338, + "step": 19832 + }, + { + "epoch": 0.6255480208169059, + "grad_norm": 1.4311491250991821, + "learning_rate": 6.496897580210735e-06, + "loss": 0.7272, + "step": 19833 + }, + { + "epoch": 0.6255795615833465, + "grad_norm": 1.4661763906478882, + "learning_rate": 6.495940774237037e-06, + "loss": 0.4717, + "step": 19834 + }, + { + "epoch": 0.6256111023497871, + "grad_norm": 1.4469377994537354, + "learning_rate": 6.494984004830916e-06, + "loss": 0.4328, + "step": 19835 + }, + { + "epoch": 0.6256426431162277, + "grad_norm": 1.588492512702942, + "learning_rate": 6.494027272002357e-06, + "loss": 0.4173, + "step": 19836 + }, + { + "epoch": 0.6256741838826684, + "grad_norm": 1.3801522254943848, + "learning_rate": 6.4930705757613465e-06, + "loss": 0.4317, + "step": 19837 + }, + { + "epoch": 0.625705724649109, + "grad_norm": 1.994268536567688, + "learning_rate": 6.492113916117871e-06, + "loss": 0.4285, + "step": 19838 + }, + { + "epoch": 0.6257372654155496, + "grad_norm": 1.4016729593276978, + "learning_rate": 6.4911572930819045e-06, + "loss": 0.4497, + "step": 19839 + }, + { + "epoch": 0.6257688061819903, + "grad_norm": 1.6355977058410645, + "learning_rate": 6.49020070666344e-06, + "loss": 0.4268, + "step": 19840 + }, + { + "epoch": 0.6258003469484309, + "grad_norm": 1.4893027544021606, + "learning_rate": 6.489244156872455e-06, + "loss": 0.693, + "step": 19841 + }, + { + "epoch": 0.6258318877148715, + "grad_norm": 2.9216954708099365, + "learning_rate": 6.488287643718936e-06, + "loss": 0.4451, + "step": 19842 + }, + { + "epoch": 0.625863428481312, + "grad_norm": 1.8441301584243774, + "learning_rate": 6.48733116721286e-06, + "loss": 0.4283, + "step": 19843 + }, + { + "epoch": 0.6258949692477527, + "grad_norm": 2.0701820850372314, + "learning_rate": 6.486374727364209e-06, + "loss": 0.4042, + "step": 19844 + }, + { + "epoch": 0.6259265100141933, + "grad_norm": 1.632378101348877, + "learning_rate": 6.485418324182969e-06, + "loss": 0.4477, + "step": 19845 + }, + { + "epoch": 0.6259580507806339, + "grad_norm": 1.743493676185608, + "learning_rate": 6.484461957679118e-06, + "loss": 0.4613, + "step": 19846 + }, + { + "epoch": 0.6259895915470746, + "grad_norm": 1.4857308864593506, + "learning_rate": 6.483505627862632e-06, + "loss": 0.4596, + "step": 19847 + }, + { + "epoch": 0.6260211323135152, + "grad_norm": 1.7711578607559204, + "learning_rate": 6.4825493347435e-06, + "loss": 0.4335, + "step": 19848 + }, + { + "epoch": 0.6260526730799558, + "grad_norm": 1.721402883529663, + "learning_rate": 6.481593078331696e-06, + "loss": 0.4496, + "step": 19849 + }, + { + "epoch": 0.6260842138463965, + "grad_norm": 1.569019079208374, + "learning_rate": 6.480636858637199e-06, + "loss": 0.4685, + "step": 19850 + }, + { + "epoch": 0.6261157546128371, + "grad_norm": 1.6645435094833374, + "learning_rate": 6.479680675669988e-06, + "loss": 0.4474, + "step": 19851 + }, + { + "epoch": 0.6261472953792777, + "grad_norm": 1.3667104244232178, + "learning_rate": 6.478724529440046e-06, + "loss": 0.7118, + "step": 19852 + }, + { + "epoch": 0.6261788361457183, + "grad_norm": 1.2685774564743042, + "learning_rate": 6.4777684199573445e-06, + "loss": 0.3258, + "step": 19853 + }, + { + "epoch": 0.626210376912159, + "grad_norm": 1.5505160093307495, + "learning_rate": 6.476812347231866e-06, + "loss": 0.4718, + "step": 19854 + }, + { + "epoch": 0.6262419176785996, + "grad_norm": 1.7270628213882446, + "learning_rate": 6.47585631127359e-06, + "loss": 0.434, + "step": 19855 + }, + { + "epoch": 0.6262734584450402, + "grad_norm": 1.8199670314788818, + "learning_rate": 6.474900312092482e-06, + "loss": 0.457, + "step": 19856 + }, + { + "epoch": 0.6263049992114809, + "grad_norm": 1.6237034797668457, + "learning_rate": 6.473944349698531e-06, + "loss": 0.4542, + "step": 19857 + }, + { + "epoch": 0.6263365399779215, + "grad_norm": 1.3694900274276733, + "learning_rate": 6.472988424101706e-06, + "loss": 0.4314, + "step": 19858 + }, + { + "epoch": 0.6263680807443621, + "grad_norm": 1.457481026649475, + "learning_rate": 6.4720325353119885e-06, + "loss": 0.435, + "step": 19859 + }, + { + "epoch": 0.6263996215108028, + "grad_norm": 2.370701551437378, + "learning_rate": 6.471076683339351e-06, + "loss": 0.4654, + "step": 19860 + }, + { + "epoch": 0.6264311622772434, + "grad_norm": 1.5372872352600098, + "learning_rate": 6.470120868193766e-06, + "loss": 0.4569, + "step": 19861 + }, + { + "epoch": 0.626462703043684, + "grad_norm": 1.6873719692230225, + "learning_rate": 6.469165089885214e-06, + "loss": 0.452, + "step": 19862 + }, + { + "epoch": 0.6264942438101246, + "grad_norm": 1.450107216835022, + "learning_rate": 6.4682093484236645e-06, + "loss": 0.39, + "step": 19863 + }, + { + "epoch": 0.6265257845765652, + "grad_norm": 1.3727508783340454, + "learning_rate": 6.467253643819089e-06, + "loss": 0.4396, + "step": 19864 + }, + { + "epoch": 0.6265573253430058, + "grad_norm": 2.0081512928009033, + "learning_rate": 6.466297976081471e-06, + "loss": 0.449, + "step": 19865 + }, + { + "epoch": 0.6265888661094464, + "grad_norm": 1.4489446878433228, + "learning_rate": 6.465342345220775e-06, + "loss": 0.4417, + "step": 19866 + }, + { + "epoch": 0.626620406875887, + "grad_norm": 1.935491919517517, + "learning_rate": 6.464386751246971e-06, + "loss": 0.4796, + "step": 19867 + }, + { + "epoch": 0.6266519476423277, + "grad_norm": 1.3722847700119019, + "learning_rate": 6.4634311941700445e-06, + "loss": 0.4335, + "step": 19868 + }, + { + "epoch": 0.6266834884087683, + "grad_norm": 1.9475581645965576, + "learning_rate": 6.462475673999957e-06, + "loss": 0.3968, + "step": 19869 + }, + { + "epoch": 0.6267150291752089, + "grad_norm": 1.4440664052963257, + "learning_rate": 6.46152019074668e-06, + "loss": 0.3855, + "step": 19870 + }, + { + "epoch": 0.6267465699416496, + "grad_norm": 1.510797142982483, + "learning_rate": 6.460564744420191e-06, + "loss": 0.4531, + "step": 19871 + }, + { + "epoch": 0.6267781107080902, + "grad_norm": 1.6240729093551636, + "learning_rate": 6.459609335030453e-06, + "loss": 0.4711, + "step": 19872 + }, + { + "epoch": 0.6268096514745308, + "grad_norm": 3.2233452796936035, + "learning_rate": 6.458653962587445e-06, + "loss": 0.3559, + "step": 19873 + }, + { + "epoch": 0.6268411922409715, + "grad_norm": 1.3535401821136475, + "learning_rate": 6.4576986271011326e-06, + "loss": 0.7154, + "step": 19874 + }, + { + "epoch": 0.6268727330074121, + "grad_norm": 1.5146291255950928, + "learning_rate": 6.4567433285814805e-06, + "loss": 0.4203, + "step": 19875 + }, + { + "epoch": 0.6269042737738527, + "grad_norm": 1.5569716691970825, + "learning_rate": 6.455788067038466e-06, + "loss": 0.449, + "step": 19876 + }, + { + "epoch": 0.6269358145402933, + "grad_norm": 1.7501455545425415, + "learning_rate": 6.454832842482057e-06, + "loss": 0.4541, + "step": 19877 + }, + { + "epoch": 0.626967355306734, + "grad_norm": 1.6517360210418701, + "learning_rate": 6.453877654922216e-06, + "loss": 0.4486, + "step": 19878 + }, + { + "epoch": 0.6269988960731746, + "grad_norm": 1.5902061462402344, + "learning_rate": 6.452922504368919e-06, + "loss": 0.4174, + "step": 19879 + }, + { + "epoch": 0.6270304368396152, + "grad_norm": 1.8180311918258667, + "learning_rate": 6.45196739083213e-06, + "loss": 0.4425, + "step": 19880 + }, + { + "epoch": 0.6270619776060559, + "grad_norm": 1.5090322494506836, + "learning_rate": 6.451012314321814e-06, + "loss": 0.4469, + "step": 19881 + }, + { + "epoch": 0.6270935183724965, + "grad_norm": 1.9865342378616333, + "learning_rate": 6.450057274847942e-06, + "loss": 0.4184, + "step": 19882 + }, + { + "epoch": 0.6271250591389371, + "grad_norm": 1.8353228569030762, + "learning_rate": 6.449102272420479e-06, + "loss": 0.4377, + "step": 19883 + }, + { + "epoch": 0.6271565999053778, + "grad_norm": 1.7643266916275024, + "learning_rate": 6.448147307049386e-06, + "loss": 0.4615, + "step": 19884 + }, + { + "epoch": 0.6271881406718183, + "grad_norm": 1.3573033809661865, + "learning_rate": 6.44719237874464e-06, + "loss": 0.7312, + "step": 19885 + }, + { + "epoch": 0.6272196814382589, + "grad_norm": 1.6387269496917725, + "learning_rate": 6.446237487516198e-06, + "loss": 0.4286, + "step": 19886 + }, + { + "epoch": 0.6272512222046995, + "grad_norm": 1.5794024467468262, + "learning_rate": 6.4452826333740234e-06, + "loss": 0.4327, + "step": 19887 + }, + { + "epoch": 0.6272827629711402, + "grad_norm": 1.4303406476974487, + "learning_rate": 6.444327816328091e-06, + "loss": 0.4183, + "step": 19888 + }, + { + "epoch": 0.6273143037375808, + "grad_norm": 1.3715623617172241, + "learning_rate": 6.443373036388354e-06, + "loss": 0.4307, + "step": 19889 + }, + { + "epoch": 0.6273458445040214, + "grad_norm": 1.6837676763534546, + "learning_rate": 6.442418293564785e-06, + "loss": 0.4516, + "step": 19890 + }, + { + "epoch": 0.627377385270462, + "grad_norm": 2.042105197906494, + "learning_rate": 6.441463587867341e-06, + "loss": 0.4352, + "step": 19891 + }, + { + "epoch": 0.6274089260369027, + "grad_norm": 1.5701978206634521, + "learning_rate": 6.4405089193059875e-06, + "loss": 0.4338, + "step": 19892 + }, + { + "epoch": 0.6274404668033433, + "grad_norm": 3.3455445766448975, + "learning_rate": 6.439554287890688e-06, + "loss": 0.4378, + "step": 19893 + }, + { + "epoch": 0.6274720075697839, + "grad_norm": 1.7766330242156982, + "learning_rate": 6.4385996936314064e-06, + "loss": 0.4545, + "step": 19894 + }, + { + "epoch": 0.6275035483362246, + "grad_norm": 1.3473750352859497, + "learning_rate": 6.4376451365380966e-06, + "loss": 0.3448, + "step": 19895 + }, + { + "epoch": 0.6275350891026652, + "grad_norm": 1.7358566522598267, + "learning_rate": 6.436690616620731e-06, + "loss": 0.4328, + "step": 19896 + }, + { + "epoch": 0.6275666298691058, + "grad_norm": 1.4078301191329956, + "learning_rate": 6.435736133889266e-06, + "loss": 0.4328, + "step": 19897 + }, + { + "epoch": 0.6275981706355465, + "grad_norm": 1.6112651824951172, + "learning_rate": 6.434781688353655e-06, + "loss": 0.4536, + "step": 19898 + }, + { + "epoch": 0.6276297114019871, + "grad_norm": 1.6485750675201416, + "learning_rate": 6.433827280023872e-06, + "loss": 0.419, + "step": 19899 + }, + { + "epoch": 0.6276612521684277, + "grad_norm": 1.2574889659881592, + "learning_rate": 6.43287290890987e-06, + "loss": 0.4705, + "step": 19900 + }, + { + "epoch": 0.6276927929348683, + "grad_norm": 1.9596582651138306, + "learning_rate": 6.431918575021607e-06, + "loss": 0.7045, + "step": 19901 + }, + { + "epoch": 0.627724333701309, + "grad_norm": 1.8631318807601929, + "learning_rate": 6.430964278369047e-06, + "loss": 0.3975, + "step": 19902 + }, + { + "epoch": 0.6277558744677496, + "grad_norm": 1.3991899490356445, + "learning_rate": 6.430010018962145e-06, + "loss": 0.4331, + "step": 19903 + }, + { + "epoch": 0.6277874152341902, + "grad_norm": 1.6480578184127808, + "learning_rate": 6.429055796810857e-06, + "loss": 0.4436, + "step": 19904 + }, + { + "epoch": 0.6278189560006309, + "grad_norm": 1.3346258401870728, + "learning_rate": 6.428101611925148e-06, + "loss": 0.422, + "step": 19905 + }, + { + "epoch": 0.6278504967670714, + "grad_norm": 2.014859437942505, + "learning_rate": 6.4271474643149676e-06, + "loss": 0.4287, + "step": 19906 + }, + { + "epoch": 0.627882037533512, + "grad_norm": 1.4078909158706665, + "learning_rate": 6.426193353990282e-06, + "loss": 0.4401, + "step": 19907 + }, + { + "epoch": 0.6279135782999526, + "grad_norm": 1.5610909461975098, + "learning_rate": 6.425239280961045e-06, + "loss": 0.471, + "step": 19908 + }, + { + "epoch": 0.6279451190663933, + "grad_norm": 1.4777830839157104, + "learning_rate": 6.424285245237209e-06, + "loss": 0.4425, + "step": 19909 + }, + { + "epoch": 0.6279766598328339, + "grad_norm": 1.7789679765701294, + "learning_rate": 6.4233312468287345e-06, + "loss": 0.4565, + "step": 19910 + }, + { + "epoch": 0.6280082005992745, + "grad_norm": 1.4880248308181763, + "learning_rate": 6.422377285745575e-06, + "loss": 0.4252, + "step": 19911 + }, + { + "epoch": 0.6280397413657152, + "grad_norm": 1.866072416305542, + "learning_rate": 6.421423361997685e-06, + "loss": 0.4314, + "step": 19912 + }, + { + "epoch": 0.6280712821321558, + "grad_norm": 1.3718336820602417, + "learning_rate": 6.420469475595024e-06, + "loss": 0.4462, + "step": 19913 + }, + { + "epoch": 0.6281028228985964, + "grad_norm": 1.53223717212677, + "learning_rate": 6.419515626547543e-06, + "loss": 0.4255, + "step": 19914 + }, + { + "epoch": 0.628134363665037, + "grad_norm": 2.255338430404663, + "learning_rate": 6.418561814865192e-06, + "loss": 0.419, + "step": 19915 + }, + { + "epoch": 0.6281659044314777, + "grad_norm": 1.4974853992462158, + "learning_rate": 6.417608040557934e-06, + "loss": 0.4294, + "step": 19916 + }, + { + "epoch": 0.6281974451979183, + "grad_norm": 1.688862681388855, + "learning_rate": 6.416654303635718e-06, + "loss": 0.3872, + "step": 19917 + }, + { + "epoch": 0.6282289859643589, + "grad_norm": 1.3360743522644043, + "learning_rate": 6.4157006041084925e-06, + "loss": 0.4737, + "step": 19918 + }, + { + "epoch": 0.6282605267307996, + "grad_norm": 2.195168972015381, + "learning_rate": 6.414746941986219e-06, + "loss": 0.4171, + "step": 19919 + }, + { + "epoch": 0.6282920674972402, + "grad_norm": 1.594903826713562, + "learning_rate": 6.413793317278841e-06, + "loss": 0.452, + "step": 19920 + }, + { + "epoch": 0.6283236082636808, + "grad_norm": 1.9846882820129395, + "learning_rate": 6.4128397299963185e-06, + "loss": 0.4506, + "step": 19921 + }, + { + "epoch": 0.6283551490301215, + "grad_norm": 1.8456696271896362, + "learning_rate": 6.411886180148597e-06, + "loss": 0.4243, + "step": 19922 + }, + { + "epoch": 0.6283866897965621, + "grad_norm": 1.4535865783691406, + "learning_rate": 6.410932667745626e-06, + "loss": 0.432, + "step": 19923 + }, + { + "epoch": 0.6284182305630027, + "grad_norm": 1.548520565032959, + "learning_rate": 6.409979192797364e-06, + "loss": 0.7263, + "step": 19924 + }, + { + "epoch": 0.6284497713294434, + "grad_norm": 1.4128808975219727, + "learning_rate": 6.4090257553137566e-06, + "loss": 0.4264, + "step": 19925 + }, + { + "epoch": 0.628481312095884, + "grad_norm": 1.7653172016143799, + "learning_rate": 6.40807235530475e-06, + "loss": 0.4377, + "step": 19926 + }, + { + "epoch": 0.6285128528623245, + "grad_norm": 1.3476269245147705, + "learning_rate": 6.4071189927803015e-06, + "loss": 0.7445, + "step": 19927 + }, + { + "epoch": 0.6285443936287651, + "grad_norm": 1.4860098361968994, + "learning_rate": 6.406165667750355e-06, + "loss": 0.4568, + "step": 19928 + }, + { + "epoch": 0.6285759343952058, + "grad_norm": 1.5081208944320679, + "learning_rate": 6.405212380224859e-06, + "loss": 0.4765, + "step": 19929 + }, + { + "epoch": 0.6286074751616464, + "grad_norm": 1.4801762104034424, + "learning_rate": 6.404259130213766e-06, + "loss": 0.4355, + "step": 19930 + }, + { + "epoch": 0.628639015928087, + "grad_norm": 1.6605217456817627, + "learning_rate": 6.403305917727016e-06, + "loss": 0.4132, + "step": 19931 + }, + { + "epoch": 0.6286705566945276, + "grad_norm": 1.8642934560775757, + "learning_rate": 6.402352742774565e-06, + "loss": 0.7217, + "step": 19932 + }, + { + "epoch": 0.6287020974609683, + "grad_norm": 1.9093211889266968, + "learning_rate": 6.4013996053663565e-06, + "loss": 0.4386, + "step": 19933 + }, + { + "epoch": 0.6287336382274089, + "grad_norm": 1.4790841341018677, + "learning_rate": 6.400446505512339e-06, + "loss": 0.4771, + "step": 19934 + }, + { + "epoch": 0.6287651789938495, + "grad_norm": 1.611730933189392, + "learning_rate": 6.399493443222451e-06, + "loss": 0.4304, + "step": 19935 + }, + { + "epoch": 0.6287967197602902, + "grad_norm": 1.434180736541748, + "learning_rate": 6.39854041850665e-06, + "loss": 0.4332, + "step": 19936 + }, + { + "epoch": 0.6288282605267308, + "grad_norm": 1.3435626029968262, + "learning_rate": 6.3975874313748724e-06, + "loss": 0.4531, + "step": 19937 + }, + { + "epoch": 0.6288598012931714, + "grad_norm": 1.1488981246948242, + "learning_rate": 6.396634481837071e-06, + "loss": 0.3159, + "step": 19938 + }, + { + "epoch": 0.6288913420596121, + "grad_norm": 1.685390591621399, + "learning_rate": 6.395681569903186e-06, + "loss": 0.4316, + "step": 19939 + }, + { + "epoch": 0.6289228828260527, + "grad_norm": 1.55841863155365, + "learning_rate": 6.394728695583161e-06, + "loss": 0.4376, + "step": 19940 + }, + { + "epoch": 0.6289544235924933, + "grad_norm": 1.5089141130447388, + "learning_rate": 6.393775858886943e-06, + "loss": 0.7375, + "step": 19941 + }, + { + "epoch": 0.628985964358934, + "grad_norm": 1.526641607284546, + "learning_rate": 6.3928230598244755e-06, + "loss": 0.4605, + "step": 19942 + }, + { + "epoch": 0.6290175051253746, + "grad_norm": 1.491552472114563, + "learning_rate": 6.391870298405694e-06, + "loss": 0.4332, + "step": 19943 + }, + { + "epoch": 0.6290490458918152, + "grad_norm": 1.5367592573165894, + "learning_rate": 6.390917574640554e-06, + "loss": 0.4691, + "step": 19944 + }, + { + "epoch": 0.6290805866582558, + "grad_norm": 1.376609206199646, + "learning_rate": 6.389964888538991e-06, + "loss": 0.4592, + "step": 19945 + }, + { + "epoch": 0.6291121274246965, + "grad_norm": 1.6816118955612183, + "learning_rate": 6.389012240110942e-06, + "loss": 0.4502, + "step": 19946 + }, + { + "epoch": 0.6291436681911371, + "grad_norm": 1.5775220394134521, + "learning_rate": 6.3880596293663586e-06, + "loss": 0.4351, + "step": 19947 + }, + { + "epoch": 0.6291752089575776, + "grad_norm": 1.401799201965332, + "learning_rate": 6.387107056315178e-06, + "loss": 0.4361, + "step": 19948 + }, + { + "epoch": 0.6292067497240182, + "grad_norm": 1.5869640111923218, + "learning_rate": 6.386154520967339e-06, + "loss": 0.7077, + "step": 19949 + }, + { + "epoch": 0.6292382904904589, + "grad_norm": 1.318130373954773, + "learning_rate": 6.385202023332786e-06, + "loss": 0.3484, + "step": 19950 + }, + { + "epoch": 0.6292698312568995, + "grad_norm": 1.7017484903335571, + "learning_rate": 6.3842495634214515e-06, + "loss": 0.4301, + "step": 19951 + }, + { + "epoch": 0.6293013720233401, + "grad_norm": 3.671070098876953, + "learning_rate": 6.383297141243286e-06, + "loss": 0.424, + "step": 19952 + }, + { + "epoch": 0.6293329127897808, + "grad_norm": 1.8217087984085083, + "learning_rate": 6.382344756808221e-06, + "loss": 0.4207, + "step": 19953 + }, + { + "epoch": 0.6293644535562214, + "grad_norm": 1.4161312580108643, + "learning_rate": 6.381392410126193e-06, + "loss": 0.4408, + "step": 19954 + }, + { + "epoch": 0.629395994322662, + "grad_norm": 1.5710601806640625, + "learning_rate": 6.380440101207152e-06, + "loss": 0.4251, + "step": 19955 + }, + { + "epoch": 0.6294275350891027, + "grad_norm": 1.461175560951233, + "learning_rate": 6.379487830061026e-06, + "loss": 0.3547, + "step": 19956 + }, + { + "epoch": 0.6294590758555433, + "grad_norm": 1.4319605827331543, + "learning_rate": 6.378535596697754e-06, + "loss": 0.4261, + "step": 19957 + }, + { + "epoch": 0.6294906166219839, + "grad_norm": 1.5933176279067993, + "learning_rate": 6.3775834011272785e-06, + "loss": 0.4469, + "step": 19958 + }, + { + "epoch": 0.6295221573884245, + "grad_norm": 3.521656036376953, + "learning_rate": 6.3766312433595325e-06, + "loss": 0.3939, + "step": 19959 + }, + { + "epoch": 0.6295536981548652, + "grad_norm": 2.1033945083618164, + "learning_rate": 6.375679123404451e-06, + "loss": 0.4478, + "step": 19960 + }, + { + "epoch": 0.6295852389213058, + "grad_norm": 1.8656351566314697, + "learning_rate": 6.374727041271975e-06, + "loss": 0.4545, + "step": 19961 + }, + { + "epoch": 0.6296167796877464, + "grad_norm": 2.1600260734558105, + "learning_rate": 6.373774996972037e-06, + "loss": 0.4432, + "step": 19962 + }, + { + "epoch": 0.6296483204541871, + "grad_norm": 1.4049755334854126, + "learning_rate": 6.3728229905145665e-06, + "loss": 0.4532, + "step": 19963 + }, + { + "epoch": 0.6296798612206277, + "grad_norm": 1.534176230430603, + "learning_rate": 6.371871021909511e-06, + "loss": 0.4307, + "step": 19964 + }, + { + "epoch": 0.6297114019870683, + "grad_norm": 1.6211366653442383, + "learning_rate": 6.370919091166799e-06, + "loss": 0.3982, + "step": 19965 + }, + { + "epoch": 0.629742942753509, + "grad_norm": 1.7974311113357544, + "learning_rate": 6.369967198296357e-06, + "loss": 0.3995, + "step": 19966 + }, + { + "epoch": 0.6297744835199496, + "grad_norm": 0.9532853960990906, + "learning_rate": 6.369015343308132e-06, + "loss": 0.3554, + "step": 19967 + }, + { + "epoch": 0.6298060242863902, + "grad_norm": 1.4380048513412476, + "learning_rate": 6.36806352621205e-06, + "loss": 0.4504, + "step": 19968 + }, + { + "epoch": 0.6298375650528308, + "grad_norm": 1.6216785907745361, + "learning_rate": 6.367111747018045e-06, + "loss": 0.4649, + "step": 19969 + }, + { + "epoch": 0.6298691058192714, + "grad_norm": 1.5383291244506836, + "learning_rate": 6.366160005736052e-06, + "loss": 0.7216, + "step": 19970 + }, + { + "epoch": 0.629900646585712, + "grad_norm": 1.6671401262283325, + "learning_rate": 6.365208302375996e-06, + "loss": 0.4553, + "step": 19971 + }, + { + "epoch": 0.6299321873521526, + "grad_norm": 1.303459882736206, + "learning_rate": 6.364256636947817e-06, + "loss": 0.3366, + "step": 19972 + }, + { + "epoch": 0.6299637281185932, + "grad_norm": 1.6381045579910278, + "learning_rate": 6.3633050094614445e-06, + "loss": 0.4506, + "step": 19973 + }, + { + "epoch": 0.6299952688850339, + "grad_norm": 1.9633921384811401, + "learning_rate": 6.362353419926803e-06, + "loss": 0.4426, + "step": 19974 + }, + { + "epoch": 0.6300268096514745, + "grad_norm": 1.514776349067688, + "learning_rate": 6.361401868353833e-06, + "loss": 0.4534, + "step": 19975 + }, + { + "epoch": 0.6300583504179151, + "grad_norm": 1.0831817388534546, + "learning_rate": 6.360450354752459e-06, + "loss": 0.3425, + "step": 19976 + }, + { + "epoch": 0.6300898911843558, + "grad_norm": 1.8337109088897705, + "learning_rate": 6.359498879132609e-06, + "loss": 0.4347, + "step": 19977 + }, + { + "epoch": 0.6301214319507964, + "grad_norm": 1.5059013366699219, + "learning_rate": 6.358547441504217e-06, + "loss": 0.4641, + "step": 19978 + }, + { + "epoch": 0.630152972717237, + "grad_norm": 1.7246763706207275, + "learning_rate": 6.357596041877211e-06, + "loss": 0.4208, + "step": 19979 + }, + { + "epoch": 0.6301845134836777, + "grad_norm": 1.818249225616455, + "learning_rate": 6.356644680261517e-06, + "loss": 0.432, + "step": 19980 + }, + { + "epoch": 0.6302160542501183, + "grad_norm": 1.8061494827270508, + "learning_rate": 6.3556933566670656e-06, + "loss": 0.4487, + "step": 19981 + }, + { + "epoch": 0.6302475950165589, + "grad_norm": 2.0618858337402344, + "learning_rate": 6.354742071103783e-06, + "loss": 0.4702, + "step": 19982 + }, + { + "epoch": 0.6302791357829995, + "grad_norm": 1.5677413940429688, + "learning_rate": 6.353790823581596e-06, + "loss": 0.4225, + "step": 19983 + }, + { + "epoch": 0.6303106765494402, + "grad_norm": 1.4034757614135742, + "learning_rate": 6.352839614110436e-06, + "loss": 0.4222, + "step": 19984 + }, + { + "epoch": 0.6303422173158808, + "grad_norm": 1.4524202346801758, + "learning_rate": 6.351888442700221e-06, + "loss": 0.3203, + "step": 19985 + }, + { + "epoch": 0.6303737580823214, + "grad_norm": 1.5740560293197632, + "learning_rate": 6.350937309360889e-06, + "loss": 0.4744, + "step": 19986 + }, + { + "epoch": 0.6304052988487621, + "grad_norm": 1.871371865272522, + "learning_rate": 6.3499862141023575e-06, + "loss": 0.4647, + "step": 19987 + }, + { + "epoch": 0.6304368396152027, + "grad_norm": 1.4643045663833618, + "learning_rate": 6.349035156934553e-06, + "loss": 0.4503, + "step": 19988 + }, + { + "epoch": 0.6304683803816433, + "grad_norm": 1.6154865026474, + "learning_rate": 6.348084137867404e-06, + "loss": 0.4157, + "step": 19989 + }, + { + "epoch": 0.630499921148084, + "grad_norm": 1.7929046154022217, + "learning_rate": 6.347133156910832e-06, + "loss": 0.4252, + "step": 19990 + }, + { + "epoch": 0.6305314619145245, + "grad_norm": 1.4664270877838135, + "learning_rate": 6.346182214074759e-06, + "loss": 0.4341, + "step": 19991 + }, + { + "epoch": 0.6305630026809651, + "grad_norm": 0.9899987578392029, + "learning_rate": 6.345231309369114e-06, + "loss": 0.3486, + "step": 19992 + }, + { + "epoch": 0.6305945434474057, + "grad_norm": 2.403454303741455, + "learning_rate": 6.344280442803819e-06, + "loss": 0.4617, + "step": 19993 + }, + { + "epoch": 0.6306260842138464, + "grad_norm": 1.598090648651123, + "learning_rate": 6.343329614388793e-06, + "loss": 0.4051, + "step": 19994 + }, + { + "epoch": 0.630657624980287, + "grad_norm": 1.7045040130615234, + "learning_rate": 6.342378824133965e-06, + "loss": 0.4652, + "step": 19995 + }, + { + "epoch": 0.6306891657467276, + "grad_norm": 1.7578269243240356, + "learning_rate": 6.3414280720492534e-06, + "loss": 0.4445, + "step": 19996 + }, + { + "epoch": 0.6307207065131682, + "grad_norm": 1.4750515222549438, + "learning_rate": 6.3404773581445774e-06, + "loss": 0.4868, + "step": 19997 + }, + { + "epoch": 0.6307522472796089, + "grad_norm": 2.597693681716919, + "learning_rate": 6.339526682429866e-06, + "loss": 0.4654, + "step": 19998 + }, + { + "epoch": 0.6307837880460495, + "grad_norm": 1.9922372102737427, + "learning_rate": 6.338576044915033e-06, + "loss": 0.4539, + "step": 19999 + }, + { + "epoch": 0.6308153288124901, + "grad_norm": 1.8471345901489258, + "learning_rate": 6.337625445610005e-06, + "loss": 0.3878, + "step": 20000 + }, + { + "epoch": 0.6308468695789308, + "grad_norm": 1.3712127208709717, + "learning_rate": 6.3366748845246986e-06, + "loss": 0.4332, + "step": 20001 + }, + { + "epoch": 0.6308784103453714, + "grad_norm": 2.501426935195923, + "learning_rate": 6.3357243616690314e-06, + "loss": 0.4015, + "step": 20002 + }, + { + "epoch": 0.630909951111812, + "grad_norm": 1.6187846660614014, + "learning_rate": 6.3347738770529284e-06, + "loss": 0.4708, + "step": 20003 + }, + { + "epoch": 0.6309414918782527, + "grad_norm": 1.3377633094787598, + "learning_rate": 6.333823430686306e-06, + "loss": 0.7045, + "step": 20004 + }, + { + "epoch": 0.6309730326446933, + "grad_norm": 1.541232943534851, + "learning_rate": 6.3328730225790806e-06, + "loss": 0.4304, + "step": 20005 + }, + { + "epoch": 0.6310045734111339, + "grad_norm": 1.592469573020935, + "learning_rate": 6.331922652741174e-06, + "loss": 0.4329, + "step": 20006 + }, + { + "epoch": 0.6310361141775745, + "grad_norm": 1.5505517721176147, + "learning_rate": 6.330972321182505e-06, + "loss": 0.4396, + "step": 20007 + }, + { + "epoch": 0.6310676549440152, + "grad_norm": 2.092277765274048, + "learning_rate": 6.330022027912985e-06, + "loss": 0.724, + "step": 20008 + }, + { + "epoch": 0.6310991957104558, + "grad_norm": 1.4394277334213257, + "learning_rate": 6.3290717729425365e-06, + "loss": 0.4626, + "step": 20009 + }, + { + "epoch": 0.6311307364768964, + "grad_norm": 1.5302232503890991, + "learning_rate": 6.328121556281077e-06, + "loss": 0.4339, + "step": 20010 + }, + { + "epoch": 0.6311622772433371, + "grad_norm": 1.6103320121765137, + "learning_rate": 6.327171377938515e-06, + "loss": 0.4547, + "step": 20011 + }, + { + "epoch": 0.6311938180097776, + "grad_norm": 4.485159873962402, + "learning_rate": 6.326221237924775e-06, + "loss": 0.3251, + "step": 20012 + }, + { + "epoch": 0.6312253587762182, + "grad_norm": 9.309245109558105, + "learning_rate": 6.3252711362497686e-06, + "loss": 0.4831, + "step": 20013 + }, + { + "epoch": 0.6312568995426588, + "grad_norm": 1.4313608407974243, + "learning_rate": 6.324321072923407e-06, + "loss": 0.4749, + "step": 20014 + }, + { + "epoch": 0.6312884403090995, + "grad_norm": 1.4365277290344238, + "learning_rate": 6.323371047955616e-06, + "loss": 0.4093, + "step": 20015 + }, + { + "epoch": 0.6313199810755401, + "grad_norm": 1.4015709161758423, + "learning_rate": 6.3224210613562975e-06, + "loss": 0.4595, + "step": 20016 + }, + { + "epoch": 0.6313515218419807, + "grad_norm": 1.424586534500122, + "learning_rate": 6.321471113135372e-06, + "loss": 0.4208, + "step": 20017 + }, + { + "epoch": 0.6313830626084214, + "grad_norm": 1.4444478750228882, + "learning_rate": 6.320521203302754e-06, + "loss": 0.4064, + "step": 20018 + }, + { + "epoch": 0.631414603374862, + "grad_norm": 1.4634181261062622, + "learning_rate": 6.319571331868353e-06, + "loss": 0.4214, + "step": 20019 + }, + { + "epoch": 0.6314461441413026, + "grad_norm": 1.5737040042877197, + "learning_rate": 6.3186214988420835e-06, + "loss": 0.4434, + "step": 20020 + }, + { + "epoch": 0.6314776849077433, + "grad_norm": 1.2901638746261597, + "learning_rate": 6.3176717042338585e-06, + "loss": 0.3062, + "step": 20021 + }, + { + "epoch": 0.6315092256741839, + "grad_norm": 1.6603844165802002, + "learning_rate": 6.316721948053584e-06, + "loss": 0.4433, + "step": 20022 + }, + { + "epoch": 0.6315407664406245, + "grad_norm": 1.8281981945037842, + "learning_rate": 6.315772230311179e-06, + "loss": 0.4505, + "step": 20023 + }, + { + "epoch": 0.6315723072070651, + "grad_norm": 2.52858304977417, + "learning_rate": 6.314822551016553e-06, + "loss": 0.4267, + "step": 20024 + }, + { + "epoch": 0.6316038479735058, + "grad_norm": 1.5847160816192627, + "learning_rate": 6.31387291017961e-06, + "loss": 0.4275, + "step": 20025 + }, + { + "epoch": 0.6316353887399464, + "grad_norm": 2.0450706481933594, + "learning_rate": 6.31292330781027e-06, + "loss": 0.4805, + "step": 20026 + }, + { + "epoch": 0.631666929506387, + "grad_norm": 1.5576965808868408, + "learning_rate": 6.311973743918438e-06, + "loss": 0.4469, + "step": 20027 + }, + { + "epoch": 0.6316984702728277, + "grad_norm": 1.3702952861785889, + "learning_rate": 6.311024218514022e-06, + "loss": 0.4167, + "step": 20028 + }, + { + "epoch": 0.6317300110392683, + "grad_norm": 1.874262809753418, + "learning_rate": 6.3100747316069345e-06, + "loss": 0.4465, + "step": 20029 + }, + { + "epoch": 0.6317615518057089, + "grad_norm": 1.8232619762420654, + "learning_rate": 6.309125283207082e-06, + "loss": 0.7446, + "step": 20030 + }, + { + "epoch": 0.6317930925721496, + "grad_norm": 1.8282549381256104, + "learning_rate": 6.308175873324369e-06, + "loss": 0.4564, + "step": 20031 + }, + { + "epoch": 0.6318246333385902, + "grad_norm": 1.7016712427139282, + "learning_rate": 6.307226501968712e-06, + "loss": 0.4203, + "step": 20032 + }, + { + "epoch": 0.6318561741050307, + "grad_norm": 2.3715579509735107, + "learning_rate": 6.306277169150009e-06, + "loss": 0.7237, + "step": 20033 + }, + { + "epoch": 0.6318877148714713, + "grad_norm": 1.4621295928955078, + "learning_rate": 6.305327874878176e-06, + "loss": 0.6961, + "step": 20034 + }, + { + "epoch": 0.631919255637912, + "grad_norm": 1.584074854850769, + "learning_rate": 6.304378619163116e-06, + "loss": 0.4594, + "step": 20035 + }, + { + "epoch": 0.6319507964043526, + "grad_norm": 1.6847481727600098, + "learning_rate": 6.3034294020147335e-06, + "loss": 0.3869, + "step": 20036 + }, + { + "epoch": 0.6319823371707932, + "grad_norm": 1.9121381044387817, + "learning_rate": 6.302480223442934e-06, + "loss": 0.4288, + "step": 20037 + }, + { + "epoch": 0.6320138779372338, + "grad_norm": 1.6789374351501465, + "learning_rate": 6.301531083457627e-06, + "loss": 0.3974, + "step": 20038 + }, + { + "epoch": 0.6320454187036745, + "grad_norm": 1.4516476392745972, + "learning_rate": 6.300581982068713e-06, + "loss": 0.4312, + "step": 20039 + }, + { + "epoch": 0.6320769594701151, + "grad_norm": 1.828955888748169, + "learning_rate": 6.299632919286101e-06, + "loss": 0.3845, + "step": 20040 + }, + { + "epoch": 0.6321085002365557, + "grad_norm": 1.6336278915405273, + "learning_rate": 6.298683895119692e-06, + "loss": 0.4197, + "step": 20041 + }, + { + "epoch": 0.6321400410029964, + "grad_norm": 4.9138078689575195, + "learning_rate": 6.297734909579386e-06, + "loss": 0.6904, + "step": 20042 + }, + { + "epoch": 0.632171581769437, + "grad_norm": 1.5112028121948242, + "learning_rate": 6.296785962675096e-06, + "loss": 0.4615, + "step": 20043 + }, + { + "epoch": 0.6322031225358776, + "grad_norm": 1.741733193397522, + "learning_rate": 6.295837054416721e-06, + "loss": 0.4493, + "step": 20044 + }, + { + "epoch": 0.6322346633023183, + "grad_norm": 1.5991218090057373, + "learning_rate": 6.294888184814156e-06, + "loss": 0.4554, + "step": 20045 + }, + { + "epoch": 0.6322662040687589, + "grad_norm": 1.9497979879379272, + "learning_rate": 6.293939353877314e-06, + "loss": 0.462, + "step": 20046 + }, + { + "epoch": 0.6322977448351995, + "grad_norm": 1.6702072620391846, + "learning_rate": 6.2929905616160925e-06, + "loss": 0.4303, + "step": 20047 + }, + { + "epoch": 0.6323292856016401, + "grad_norm": 1.2433816194534302, + "learning_rate": 6.292041808040393e-06, + "loss": 0.4137, + "step": 20048 + }, + { + "epoch": 0.6323608263680808, + "grad_norm": 1.684006690979004, + "learning_rate": 6.291093093160118e-06, + "loss": 0.3499, + "step": 20049 + }, + { + "epoch": 0.6323923671345214, + "grad_norm": 1.978456735610962, + "learning_rate": 6.2901444169851615e-06, + "loss": 0.3969, + "step": 20050 + }, + { + "epoch": 0.632423907900962, + "grad_norm": 2.2448647022247314, + "learning_rate": 6.289195779525432e-06, + "loss": 0.4275, + "step": 20051 + }, + { + "epoch": 0.6324554486674027, + "grad_norm": 1.629315733909607, + "learning_rate": 6.288247180790827e-06, + "loss": 0.4664, + "step": 20052 + }, + { + "epoch": 0.6324869894338433, + "grad_norm": 1.4050577878952026, + "learning_rate": 6.28729862079124e-06, + "loss": 0.439, + "step": 20053 + }, + { + "epoch": 0.6325185302002838, + "grad_norm": 2.0352137088775635, + "learning_rate": 6.28635009953658e-06, + "loss": 0.4246, + "step": 20054 + }, + { + "epoch": 0.6325500709667244, + "grad_norm": 1.8805222511291504, + "learning_rate": 6.2854016170367395e-06, + "loss": 0.3969, + "step": 20055 + }, + { + "epoch": 0.6325816117331651, + "grad_norm": 1.386114239692688, + "learning_rate": 6.284453173301616e-06, + "loss": 0.7066, + "step": 20056 + }, + { + "epoch": 0.6326131524996057, + "grad_norm": 1.4481457471847534, + "learning_rate": 6.283504768341107e-06, + "loss": 0.432, + "step": 20057 + }, + { + "epoch": 0.6326446932660463, + "grad_norm": 2.230422258377075, + "learning_rate": 6.282556402165115e-06, + "loss": 0.4425, + "step": 20058 + }, + { + "epoch": 0.632676234032487, + "grad_norm": 1.3911975622177124, + "learning_rate": 6.281608074783532e-06, + "loss": 0.7758, + "step": 20059 + }, + { + "epoch": 0.6327077747989276, + "grad_norm": 1.5033221244812012, + "learning_rate": 6.280659786206257e-06, + "loss": 0.475, + "step": 20060 + }, + { + "epoch": 0.6327393155653682, + "grad_norm": 1.5607830286026, + "learning_rate": 6.279711536443185e-06, + "loss": 0.4264, + "step": 20061 + }, + { + "epoch": 0.6327708563318089, + "grad_norm": 1.5162994861602783, + "learning_rate": 6.278763325504208e-06, + "loss": 0.4425, + "step": 20062 + }, + { + "epoch": 0.6328023970982495, + "grad_norm": 1.5053393840789795, + "learning_rate": 6.277815153399229e-06, + "loss": 0.4458, + "step": 20063 + }, + { + "epoch": 0.6328339378646901, + "grad_norm": 1.5019152164459229, + "learning_rate": 6.276867020138135e-06, + "loss": 0.3812, + "step": 20064 + }, + { + "epoch": 0.6328654786311307, + "grad_norm": 1.4546362161636353, + "learning_rate": 6.27591892573083e-06, + "loss": 0.7433, + "step": 20065 + }, + { + "epoch": 0.6328970193975714, + "grad_norm": 1.4857839345932007, + "learning_rate": 6.274970870187201e-06, + "loss": 0.4174, + "step": 20066 + }, + { + "epoch": 0.632928560164012, + "grad_norm": 1.420228362083435, + "learning_rate": 6.274022853517142e-06, + "loss": 0.4498, + "step": 20067 + }, + { + "epoch": 0.6329601009304526, + "grad_norm": 1.8292397260665894, + "learning_rate": 6.2730748757305496e-06, + "loss": 0.4126, + "step": 20068 + }, + { + "epoch": 0.6329916416968933, + "grad_norm": 1.2706890106201172, + "learning_rate": 6.272126936837315e-06, + "loss": 0.4564, + "step": 20069 + }, + { + "epoch": 0.6330231824633339, + "grad_norm": 1.852013349533081, + "learning_rate": 6.271179036847326e-06, + "loss": 0.4563, + "step": 20070 + }, + { + "epoch": 0.6330547232297745, + "grad_norm": 2.044222354888916, + "learning_rate": 6.270231175770484e-06, + "loss": 0.4439, + "step": 20071 + }, + { + "epoch": 0.6330862639962151, + "grad_norm": 1.4403880834579468, + "learning_rate": 6.269283353616675e-06, + "loss": 0.7373, + "step": 20072 + }, + { + "epoch": 0.6331178047626558, + "grad_norm": 1.2586722373962402, + "learning_rate": 6.268335570395787e-06, + "loss": 0.4156, + "step": 20073 + }, + { + "epoch": 0.6331493455290964, + "grad_norm": 4.900799751281738, + "learning_rate": 6.267387826117719e-06, + "loss": 0.3766, + "step": 20074 + }, + { + "epoch": 0.6331808862955369, + "grad_norm": 1.5441867113113403, + "learning_rate": 6.266440120792357e-06, + "loss": 0.4154, + "step": 20075 + }, + { + "epoch": 0.6332124270619776, + "grad_norm": 1.4228037595748901, + "learning_rate": 6.265492454429589e-06, + "loss": 0.4115, + "step": 20076 + }, + { + "epoch": 0.6332439678284182, + "grad_norm": 3.5819380283355713, + "learning_rate": 6.264544827039308e-06, + "loss": 0.4354, + "step": 20077 + }, + { + "epoch": 0.6332755085948588, + "grad_norm": 2.149414300918579, + "learning_rate": 6.263597238631405e-06, + "loss": 0.3184, + "step": 20078 + }, + { + "epoch": 0.6333070493612994, + "grad_norm": 1.5275182723999023, + "learning_rate": 6.262649689215763e-06, + "loss": 0.4342, + "step": 20079 + }, + { + "epoch": 0.6333385901277401, + "grad_norm": 1.4284108877182007, + "learning_rate": 6.2617021788022755e-06, + "loss": 0.715, + "step": 20080 + }, + { + "epoch": 0.6333701308941807, + "grad_norm": 1.5111992359161377, + "learning_rate": 6.260754707400825e-06, + "loss": 0.4012, + "step": 20081 + }, + { + "epoch": 0.6334016716606213, + "grad_norm": 1.9375617504119873, + "learning_rate": 6.259807275021305e-06, + "loss": 0.4118, + "step": 20082 + }, + { + "epoch": 0.633433212427062, + "grad_norm": 1.334989309310913, + "learning_rate": 6.258859881673601e-06, + "loss": 0.416, + "step": 20083 + }, + { + "epoch": 0.6334647531935026, + "grad_norm": 1.6546827554702759, + "learning_rate": 6.2579125273675954e-06, + "loss": 0.4438, + "step": 20084 + }, + { + "epoch": 0.6334962939599432, + "grad_norm": 1.4178410768508911, + "learning_rate": 6.256965212113182e-06, + "loss": 0.7052, + "step": 20085 + }, + { + "epoch": 0.6335278347263839, + "grad_norm": 1.752030372619629, + "learning_rate": 6.256017935920245e-06, + "loss": 0.4358, + "step": 20086 + }, + { + "epoch": 0.6335593754928245, + "grad_norm": 2.3307790756225586, + "learning_rate": 6.255070698798663e-06, + "loss": 0.4219, + "step": 20087 + }, + { + "epoch": 0.6335909162592651, + "grad_norm": 1.6240915060043335, + "learning_rate": 6.254123500758329e-06, + "loss": 0.7332, + "step": 20088 + }, + { + "epoch": 0.6336224570257057, + "grad_norm": 2.0369713306427, + "learning_rate": 6.253176341809126e-06, + "loss": 0.3294, + "step": 20089 + }, + { + "epoch": 0.6336539977921464, + "grad_norm": 1.7109591960906982, + "learning_rate": 6.252229221960933e-06, + "loss": 0.4242, + "step": 20090 + }, + { + "epoch": 0.633685538558587, + "grad_norm": 1.3995325565338135, + "learning_rate": 6.2512821412236405e-06, + "loss": 0.7237, + "step": 20091 + }, + { + "epoch": 0.6337170793250276, + "grad_norm": 1.5454505681991577, + "learning_rate": 6.250335099607131e-06, + "loss": 0.447, + "step": 20092 + }, + { + "epoch": 0.6337486200914683, + "grad_norm": 1.7894285917282104, + "learning_rate": 6.249388097121282e-06, + "loss": 0.4672, + "step": 20093 + }, + { + "epoch": 0.6337801608579089, + "grad_norm": 1.3964436054229736, + "learning_rate": 6.248441133775984e-06, + "loss": 0.7294, + "step": 20094 + }, + { + "epoch": 0.6338117016243495, + "grad_norm": 2.351255178451538, + "learning_rate": 6.247494209581114e-06, + "loss": 0.4341, + "step": 20095 + }, + { + "epoch": 0.6338432423907902, + "grad_norm": 1.5246694087982178, + "learning_rate": 6.246547324546558e-06, + "loss": 0.4225, + "step": 20096 + }, + { + "epoch": 0.6338747831572307, + "grad_norm": 1.5399196147918701, + "learning_rate": 6.245600478682193e-06, + "loss": 0.4029, + "step": 20097 + }, + { + "epoch": 0.6339063239236713, + "grad_norm": 1.7468407154083252, + "learning_rate": 6.244653671997902e-06, + "loss": 0.4595, + "step": 20098 + }, + { + "epoch": 0.6339378646901119, + "grad_norm": 1.5875141620635986, + "learning_rate": 6.243706904503568e-06, + "loss": 0.4585, + "step": 20099 + }, + { + "epoch": 0.6339694054565526, + "grad_norm": 2.3043317794799805, + "learning_rate": 6.242760176209071e-06, + "loss": 0.4328, + "step": 20100 + }, + { + "epoch": 0.6340009462229932, + "grad_norm": 2.171382188796997, + "learning_rate": 6.241813487124282e-06, + "loss": 0.4399, + "step": 20101 + }, + { + "epoch": 0.6340324869894338, + "grad_norm": 1.5475726127624512, + "learning_rate": 6.240866837259093e-06, + "loss": 0.4358, + "step": 20102 + }, + { + "epoch": 0.6340640277558744, + "grad_norm": 1.7285418510437012, + "learning_rate": 6.239920226623377e-06, + "loss": 0.4219, + "step": 20103 + }, + { + "epoch": 0.6340955685223151, + "grad_norm": 2.2471535205841064, + "learning_rate": 6.23897365522701e-06, + "loss": 0.4643, + "step": 20104 + }, + { + "epoch": 0.6341271092887557, + "grad_norm": 2.0498523712158203, + "learning_rate": 6.238027123079875e-06, + "loss": 0.4694, + "step": 20105 + }, + { + "epoch": 0.6341586500551963, + "grad_norm": 1.4273344278335571, + "learning_rate": 6.23708063019185e-06, + "loss": 0.4218, + "step": 20106 + }, + { + "epoch": 0.634190190821637, + "grad_norm": 1.5525457859039307, + "learning_rate": 6.2361341765728075e-06, + "loss": 0.4346, + "step": 20107 + }, + { + "epoch": 0.6342217315880776, + "grad_norm": 1.3188577890396118, + "learning_rate": 6.23518776223263e-06, + "loss": 0.4178, + "step": 20108 + }, + { + "epoch": 0.6342532723545182, + "grad_norm": 1.9620510339736938, + "learning_rate": 6.234241387181192e-06, + "loss": 0.4254, + "step": 20109 + }, + { + "epoch": 0.6342848131209589, + "grad_norm": 1.710938572883606, + "learning_rate": 6.233295051428365e-06, + "loss": 0.4383, + "step": 20110 + }, + { + "epoch": 0.6343163538873995, + "grad_norm": 1.6651338338851929, + "learning_rate": 6.232348754984032e-06, + "loss": 0.4139, + "step": 20111 + }, + { + "epoch": 0.6343478946538401, + "grad_norm": 1.9805883169174194, + "learning_rate": 6.2314024978580615e-06, + "loss": 0.6929, + "step": 20112 + }, + { + "epoch": 0.6343794354202807, + "grad_norm": 1.4951510429382324, + "learning_rate": 6.230456280060337e-06, + "loss": 0.4008, + "step": 20113 + }, + { + "epoch": 0.6344109761867214, + "grad_norm": 1.4644395112991333, + "learning_rate": 6.229510101600727e-06, + "loss": 0.768, + "step": 20114 + }, + { + "epoch": 0.634442516953162, + "grad_norm": 1.5839332342147827, + "learning_rate": 6.228563962489106e-06, + "loss": 0.4313, + "step": 20115 + }, + { + "epoch": 0.6344740577196026, + "grad_norm": 1.4230018854141235, + "learning_rate": 6.227617862735349e-06, + "loss": 0.4516, + "step": 20116 + }, + { + "epoch": 0.6345055984860433, + "grad_norm": 1.5040911436080933, + "learning_rate": 6.226671802349329e-06, + "loss": 0.4177, + "step": 20117 + }, + { + "epoch": 0.6345371392524838, + "grad_norm": 1.816357135772705, + "learning_rate": 6.225725781340917e-06, + "loss": 0.3555, + "step": 20118 + }, + { + "epoch": 0.6345686800189244, + "grad_norm": 1.8443453311920166, + "learning_rate": 6.224779799719989e-06, + "loss": 0.444, + "step": 20119 + }, + { + "epoch": 0.634600220785365, + "grad_norm": 1.5262078046798706, + "learning_rate": 6.2238338574964154e-06, + "loss": 0.3751, + "step": 20120 + }, + { + "epoch": 0.6346317615518057, + "grad_norm": 1.5424498319625854, + "learning_rate": 6.222887954680064e-06, + "loss": 0.4583, + "step": 20121 + }, + { + "epoch": 0.6346633023182463, + "grad_norm": 1.604875087738037, + "learning_rate": 6.2219420912808134e-06, + "loss": 0.4624, + "step": 20122 + }, + { + "epoch": 0.6346948430846869, + "grad_norm": 1.4920543432235718, + "learning_rate": 6.220996267308531e-06, + "loss": 0.4281, + "step": 20123 + }, + { + "epoch": 0.6347263838511276, + "grad_norm": 1.7139606475830078, + "learning_rate": 6.2200504827730824e-06, + "loss": 0.722, + "step": 20124 + }, + { + "epoch": 0.6347579246175682, + "grad_norm": 1.4386321306228638, + "learning_rate": 6.219104737684348e-06, + "loss": 0.7271, + "step": 20125 + }, + { + "epoch": 0.6347894653840088, + "grad_norm": 1.5266228914260864, + "learning_rate": 6.218159032052187e-06, + "loss": 0.7203, + "step": 20126 + }, + { + "epoch": 0.6348210061504495, + "grad_norm": 1.6685590744018555, + "learning_rate": 6.2172133658864755e-06, + "loss": 0.4113, + "step": 20127 + }, + { + "epoch": 0.6348525469168901, + "grad_norm": 1.4896615743637085, + "learning_rate": 6.216267739197079e-06, + "loss": 0.4613, + "step": 20128 + }, + { + "epoch": 0.6348840876833307, + "grad_norm": 1.613386869430542, + "learning_rate": 6.2153221519938644e-06, + "loss": 0.4202, + "step": 20129 + }, + { + "epoch": 0.6349156284497713, + "grad_norm": 1.5877410173416138, + "learning_rate": 6.214376604286705e-06, + "loss": 0.4655, + "step": 20130 + }, + { + "epoch": 0.634947169216212, + "grad_norm": 1.278788685798645, + "learning_rate": 6.213431096085465e-06, + "loss": 0.7104, + "step": 20131 + }, + { + "epoch": 0.6349787099826526, + "grad_norm": 1.3987460136413574, + "learning_rate": 6.212485627400009e-06, + "loss": 0.4321, + "step": 20132 + }, + { + "epoch": 0.6350102507490932, + "grad_norm": 1.6296693086624146, + "learning_rate": 6.211540198240209e-06, + "loss": 0.442, + "step": 20133 + }, + { + "epoch": 0.6350417915155339, + "grad_norm": 1.6915048360824585, + "learning_rate": 6.210594808615929e-06, + "loss": 0.4325, + "step": 20134 + }, + { + "epoch": 0.6350733322819745, + "grad_norm": 1.8274585008621216, + "learning_rate": 6.209649458537033e-06, + "loss": 0.4362, + "step": 20135 + }, + { + "epoch": 0.6351048730484151, + "grad_norm": 4.288458824157715, + "learning_rate": 6.208704148013389e-06, + "loss": 0.4668, + "step": 20136 + }, + { + "epoch": 0.6351364138148557, + "grad_norm": 1.3337335586547852, + "learning_rate": 6.2077588770548605e-06, + "loss": 0.3705, + "step": 20137 + }, + { + "epoch": 0.6351679545812964, + "grad_norm": 1.3491934537887573, + "learning_rate": 6.2068136456713124e-06, + "loss": 0.4172, + "step": 20138 + }, + { + "epoch": 0.6351994953477369, + "grad_norm": 1.8352601528167725, + "learning_rate": 6.20586845387261e-06, + "loss": 0.4361, + "step": 20139 + }, + { + "epoch": 0.6352310361141775, + "grad_norm": 1.6973648071289062, + "learning_rate": 6.2049233016686175e-06, + "loss": 0.4727, + "step": 20140 + }, + { + "epoch": 0.6352625768806182, + "grad_norm": 1.7039837837219238, + "learning_rate": 6.203978189069193e-06, + "loss": 0.4401, + "step": 20141 + }, + { + "epoch": 0.6352941176470588, + "grad_norm": 1.8233001232147217, + "learning_rate": 6.203033116084207e-06, + "loss": 0.4512, + "step": 20142 + }, + { + "epoch": 0.6353256584134994, + "grad_norm": 1.7056480646133423, + "learning_rate": 6.202088082723516e-06, + "loss": 0.4233, + "step": 20143 + }, + { + "epoch": 0.63535719917994, + "grad_norm": 5.414798259735107, + "learning_rate": 6.2011430889969885e-06, + "loss": 0.4404, + "step": 20144 + }, + { + "epoch": 0.6353887399463807, + "grad_norm": 1.6192365884780884, + "learning_rate": 6.200198134914481e-06, + "loss": 0.4374, + "step": 20145 + }, + { + "epoch": 0.6354202807128213, + "grad_norm": 1.854979395866394, + "learning_rate": 6.199253220485857e-06, + "loss": 0.4319, + "step": 20146 + }, + { + "epoch": 0.6354518214792619, + "grad_norm": 1.6468409299850464, + "learning_rate": 6.198308345720977e-06, + "loss": 0.4071, + "step": 20147 + }, + { + "epoch": 0.6354833622457026, + "grad_norm": 1.6279230117797852, + "learning_rate": 6.197363510629701e-06, + "loss": 0.4444, + "step": 20148 + }, + { + "epoch": 0.6355149030121432, + "grad_norm": 2.3022093772888184, + "learning_rate": 6.196418715221887e-06, + "loss": 0.3403, + "step": 20149 + }, + { + "epoch": 0.6355464437785838, + "grad_norm": 1.3843754529953003, + "learning_rate": 6.195473959507401e-06, + "loss": 0.4446, + "step": 20150 + }, + { + "epoch": 0.6355779845450245, + "grad_norm": 1.4289909601211548, + "learning_rate": 6.194529243496097e-06, + "loss": 0.4313, + "step": 20151 + }, + { + "epoch": 0.6356095253114651, + "grad_norm": 1.6428545713424683, + "learning_rate": 6.193584567197832e-06, + "loss": 0.4803, + "step": 20152 + }, + { + "epoch": 0.6356410660779057, + "grad_norm": 1.513620376586914, + "learning_rate": 6.192639930622471e-06, + "loss": 0.4455, + "step": 20153 + }, + { + "epoch": 0.6356726068443463, + "grad_norm": 1.929208755493164, + "learning_rate": 6.19169533377987e-06, + "loss": 0.7407, + "step": 20154 + }, + { + "epoch": 0.635704147610787, + "grad_norm": 1.6242880821228027, + "learning_rate": 6.190750776679882e-06, + "loss": 0.4441, + "step": 20155 + }, + { + "epoch": 0.6357356883772276, + "grad_norm": 1.3987473249435425, + "learning_rate": 6.18980625933237e-06, + "loss": 0.4108, + "step": 20156 + }, + { + "epoch": 0.6357672291436682, + "grad_norm": 1.4552991390228271, + "learning_rate": 6.188861781747187e-06, + "loss": 0.4507, + "step": 20157 + }, + { + "epoch": 0.6357987699101089, + "grad_norm": 1.7981890439987183, + "learning_rate": 6.187917343934189e-06, + "loss": 0.4341, + "step": 20158 + }, + { + "epoch": 0.6358303106765495, + "grad_norm": 1.2553508281707764, + "learning_rate": 6.186972945903236e-06, + "loss": 0.4511, + "step": 20159 + }, + { + "epoch": 0.63586185144299, + "grad_norm": 0.8895660638809204, + "learning_rate": 6.186028587664178e-06, + "loss": 0.3204, + "step": 20160 + }, + { + "epoch": 0.6358933922094306, + "grad_norm": 1.5075228214263916, + "learning_rate": 6.185084269226877e-06, + "loss": 0.4348, + "step": 20161 + }, + { + "epoch": 0.6359249329758713, + "grad_norm": 2.0567502975463867, + "learning_rate": 6.1841399906011836e-06, + "loss": 0.4286, + "step": 20162 + }, + { + "epoch": 0.6359564737423119, + "grad_norm": 1.7520698308944702, + "learning_rate": 6.183195751796947e-06, + "loss": 0.4594, + "step": 20163 + }, + { + "epoch": 0.6359880145087525, + "grad_norm": 1.498840570449829, + "learning_rate": 6.182251552824032e-06, + "loss": 0.6928, + "step": 20164 + }, + { + "epoch": 0.6360195552751932, + "grad_norm": 1.609472632408142, + "learning_rate": 6.181307393692286e-06, + "loss": 0.4577, + "step": 20165 + }, + { + "epoch": 0.6360510960416338, + "grad_norm": 1.5419775247573853, + "learning_rate": 6.18036327441156e-06, + "loss": 0.4489, + "step": 20166 + }, + { + "epoch": 0.6360826368080744, + "grad_norm": 2.000967502593994, + "learning_rate": 6.179419194991712e-06, + "loss": 0.4302, + "step": 20167 + }, + { + "epoch": 0.636114177574515, + "grad_norm": 1.4324700832366943, + "learning_rate": 6.178475155442592e-06, + "loss": 0.7431, + "step": 20168 + }, + { + "epoch": 0.6361457183409557, + "grad_norm": 1.5221432447433472, + "learning_rate": 6.177531155774044e-06, + "loss": 0.4282, + "step": 20169 + }, + { + "epoch": 0.6361772591073963, + "grad_norm": 1.7331280708312988, + "learning_rate": 6.176587195995933e-06, + "loss": 0.4628, + "step": 20170 + }, + { + "epoch": 0.6362087998738369, + "grad_norm": 2.3682644367218018, + "learning_rate": 6.175643276118103e-06, + "loss": 0.3982, + "step": 20171 + }, + { + "epoch": 0.6362403406402776, + "grad_norm": 1.5587303638458252, + "learning_rate": 6.174699396150401e-06, + "loss": 0.4692, + "step": 20172 + }, + { + "epoch": 0.6362718814067182, + "grad_norm": 1.3961125612258911, + "learning_rate": 6.173755556102685e-06, + "loss": 0.3465, + "step": 20173 + }, + { + "epoch": 0.6363034221731588, + "grad_norm": 1.3920737504959106, + "learning_rate": 6.172811755984798e-06, + "loss": 0.7361, + "step": 20174 + }, + { + "epoch": 0.6363349629395995, + "grad_norm": 1.429339051246643, + "learning_rate": 6.171867995806595e-06, + "loss": 0.403, + "step": 20175 + }, + { + "epoch": 0.6363665037060401, + "grad_norm": 1.6414235830307007, + "learning_rate": 6.170924275577921e-06, + "loss": 0.4566, + "step": 20176 + }, + { + "epoch": 0.6363980444724807, + "grad_norm": 2.0288448333740234, + "learning_rate": 6.169980595308623e-06, + "loss": 0.4411, + "step": 20177 + }, + { + "epoch": 0.6364295852389213, + "grad_norm": 1.493914008140564, + "learning_rate": 6.169036955008554e-06, + "loss": 0.4146, + "step": 20178 + }, + { + "epoch": 0.636461126005362, + "grad_norm": 1.5173838138580322, + "learning_rate": 6.168093354687558e-06, + "loss": 0.4517, + "step": 20179 + }, + { + "epoch": 0.6364926667718026, + "grad_norm": 1.8503849506378174, + "learning_rate": 6.1671497943554826e-06, + "loss": 0.4433, + "step": 20180 + }, + { + "epoch": 0.6365242075382431, + "grad_norm": 1.4573595523834229, + "learning_rate": 6.166206274022177e-06, + "loss": 0.4463, + "step": 20181 + }, + { + "epoch": 0.6365557483046838, + "grad_norm": 1.645747184753418, + "learning_rate": 6.165262793697486e-06, + "loss": 0.3482, + "step": 20182 + }, + { + "epoch": 0.6365872890711244, + "grad_norm": 3.177595615386963, + "learning_rate": 6.16431935339125e-06, + "loss": 0.3258, + "step": 20183 + }, + { + "epoch": 0.636618829837565, + "grad_norm": 1.556400179862976, + "learning_rate": 6.163375953113327e-06, + "loss": 0.4391, + "step": 20184 + }, + { + "epoch": 0.6366503706040056, + "grad_norm": 1.3856077194213867, + "learning_rate": 6.162432592873553e-06, + "loss": 0.4219, + "step": 20185 + }, + { + "epoch": 0.6366819113704463, + "grad_norm": 1.4324636459350586, + "learning_rate": 6.161489272681773e-06, + "loss": 0.4288, + "step": 20186 + }, + { + "epoch": 0.6367134521368869, + "grad_norm": 1.7482792139053345, + "learning_rate": 6.160545992547835e-06, + "loss": 0.4358, + "step": 20187 + }, + { + "epoch": 0.6367449929033275, + "grad_norm": 3.6394028663635254, + "learning_rate": 6.159602752481581e-06, + "loss": 0.7258, + "step": 20188 + }, + { + "epoch": 0.6367765336697682, + "grad_norm": 1.8437585830688477, + "learning_rate": 6.158659552492849e-06, + "loss": 0.4706, + "step": 20189 + }, + { + "epoch": 0.6368080744362088, + "grad_norm": 1.3646841049194336, + "learning_rate": 6.157716392591493e-06, + "loss": 0.445, + "step": 20190 + }, + { + "epoch": 0.6368396152026494, + "grad_norm": 1.4546773433685303, + "learning_rate": 6.156773272787346e-06, + "loss": 0.4578, + "step": 20191 + }, + { + "epoch": 0.63687115596909, + "grad_norm": 1.7138057947158813, + "learning_rate": 6.155830193090257e-06, + "loss": 0.7195, + "step": 20192 + }, + { + "epoch": 0.6369026967355307, + "grad_norm": 1.4556981325149536, + "learning_rate": 6.154887153510065e-06, + "loss": 0.4229, + "step": 20193 + }, + { + "epoch": 0.6369342375019713, + "grad_norm": 2.2031733989715576, + "learning_rate": 6.153944154056609e-06, + "loss": 0.4378, + "step": 20194 + }, + { + "epoch": 0.6369657782684119, + "grad_norm": 1.6878621578216553, + "learning_rate": 6.153001194739735e-06, + "loss": 0.369, + "step": 20195 + }, + { + "epoch": 0.6369973190348526, + "grad_norm": 1.680699348449707, + "learning_rate": 6.152058275569279e-06, + "loss": 0.3657, + "step": 20196 + }, + { + "epoch": 0.6370288598012932, + "grad_norm": 1.697999119758606, + "learning_rate": 6.151115396555081e-06, + "loss": 0.4687, + "step": 20197 + }, + { + "epoch": 0.6370604005677338, + "grad_norm": 1.4527668952941895, + "learning_rate": 6.150172557706986e-06, + "loss": 0.4335, + "step": 20198 + }, + { + "epoch": 0.6370919413341745, + "grad_norm": 1.4918186664581299, + "learning_rate": 6.149229759034828e-06, + "loss": 0.7052, + "step": 20199 + }, + { + "epoch": 0.6371234821006151, + "grad_norm": 1.9845138788223267, + "learning_rate": 6.148287000548445e-06, + "loss": 0.4148, + "step": 20200 + }, + { + "epoch": 0.6371550228670557, + "grad_norm": 1.553755521774292, + "learning_rate": 6.147344282257681e-06, + "loss": 0.4135, + "step": 20201 + }, + { + "epoch": 0.6371865636334962, + "grad_norm": 1.9850214719772339, + "learning_rate": 6.146401604172372e-06, + "loss": 0.4688, + "step": 20202 + }, + { + "epoch": 0.6372181043999369, + "grad_norm": 1.2774711847305298, + "learning_rate": 6.145458966302349e-06, + "loss": 0.4464, + "step": 20203 + }, + { + "epoch": 0.6372496451663775, + "grad_norm": 7.904055118560791, + "learning_rate": 6.1445163686574605e-06, + "loss": 0.7281, + "step": 20204 + }, + { + "epoch": 0.6372811859328181, + "grad_norm": 1.6405348777770996, + "learning_rate": 6.143573811247536e-06, + "loss": 0.3985, + "step": 20205 + }, + { + "epoch": 0.6373127266992588, + "grad_norm": 2.33341908454895, + "learning_rate": 6.142631294082412e-06, + "loss": 0.4649, + "step": 20206 + }, + { + "epoch": 0.6373442674656994, + "grad_norm": 2.200241804122925, + "learning_rate": 6.141688817171927e-06, + "loss": 0.7389, + "step": 20207 + }, + { + "epoch": 0.63737580823214, + "grad_norm": 1.588951826095581, + "learning_rate": 6.140746380525911e-06, + "loss": 0.4891, + "step": 20208 + }, + { + "epoch": 0.6374073489985806, + "grad_norm": 1.6713805198669434, + "learning_rate": 6.139803984154209e-06, + "loss": 0.4669, + "step": 20209 + }, + { + "epoch": 0.6374388897650213, + "grad_norm": 1.741964340209961, + "learning_rate": 6.138861628066648e-06, + "loss": 0.4447, + "step": 20210 + }, + { + "epoch": 0.6374704305314619, + "grad_norm": 1.8728528022766113, + "learning_rate": 6.13791931227306e-06, + "loss": 0.4458, + "step": 20211 + }, + { + "epoch": 0.6375019712979025, + "grad_norm": 3.7483277320861816, + "learning_rate": 6.136977036783287e-06, + "loss": 0.4133, + "step": 20212 + }, + { + "epoch": 0.6375335120643432, + "grad_norm": 1.4312769174575806, + "learning_rate": 6.136034801607158e-06, + "loss": 0.4225, + "step": 20213 + }, + { + "epoch": 0.6375650528307838, + "grad_norm": 1.7083110809326172, + "learning_rate": 6.135092606754504e-06, + "loss": 0.3966, + "step": 20214 + }, + { + "epoch": 0.6375965935972244, + "grad_norm": 1.7459903955459595, + "learning_rate": 6.134150452235163e-06, + "loss": 0.465, + "step": 20215 + }, + { + "epoch": 0.6376281343636651, + "grad_norm": 1.4560916423797607, + "learning_rate": 6.133208338058962e-06, + "loss": 0.4194, + "step": 20216 + }, + { + "epoch": 0.6376596751301057, + "grad_norm": 1.8597899675369263, + "learning_rate": 6.132266264235732e-06, + "loss": 0.4726, + "step": 20217 + }, + { + "epoch": 0.6376912158965463, + "grad_norm": 1.3326690196990967, + "learning_rate": 6.1313242307753105e-06, + "loss": 0.4629, + "step": 20218 + }, + { + "epoch": 0.637722756662987, + "grad_norm": 2.2764389514923096, + "learning_rate": 6.130382237687524e-06, + "loss": 0.427, + "step": 20219 + }, + { + "epoch": 0.6377542974294276, + "grad_norm": 1.386125087738037, + "learning_rate": 6.1294402849822e-06, + "loss": 0.4354, + "step": 20220 + }, + { + "epoch": 0.6377858381958682, + "grad_norm": 2.2149181365966797, + "learning_rate": 6.128498372669176e-06, + "loss": 0.3732, + "step": 20221 + }, + { + "epoch": 0.6378173789623088, + "grad_norm": 1.600642442703247, + "learning_rate": 6.127556500758276e-06, + "loss": 0.4366, + "step": 20222 + }, + { + "epoch": 0.6378489197287495, + "grad_norm": 7.882518768310547, + "learning_rate": 6.12661466925933e-06, + "loss": 0.3174, + "step": 20223 + }, + { + "epoch": 0.63788046049519, + "grad_norm": 1.4360060691833496, + "learning_rate": 6.125672878182169e-06, + "loss": 0.4319, + "step": 20224 + }, + { + "epoch": 0.6379120012616306, + "grad_norm": 2.4940457344055176, + "learning_rate": 6.124731127536618e-06, + "loss": 0.4364, + "step": 20225 + }, + { + "epoch": 0.6379435420280712, + "grad_norm": 2.0891635417938232, + "learning_rate": 6.123789417332511e-06, + "loss": 0.4074, + "step": 20226 + }, + { + "epoch": 0.6379750827945119, + "grad_norm": 1.614099144935608, + "learning_rate": 6.122847747579668e-06, + "loss": 0.4175, + "step": 20227 + }, + { + "epoch": 0.6380066235609525, + "grad_norm": 1.511125922203064, + "learning_rate": 6.121906118287917e-06, + "loss": 0.4335, + "step": 20228 + }, + { + "epoch": 0.6380381643273931, + "grad_norm": 1.4050514698028564, + "learning_rate": 6.120964529467092e-06, + "loss": 0.4328, + "step": 20229 + }, + { + "epoch": 0.6380697050938338, + "grad_norm": 2.224362373352051, + "learning_rate": 6.120022981127012e-06, + "loss": 0.4246, + "step": 20230 + }, + { + "epoch": 0.6381012458602744, + "grad_norm": 1.7309983968734741, + "learning_rate": 6.119081473277502e-06, + "loss": 0.4473, + "step": 20231 + }, + { + "epoch": 0.638132786626715, + "grad_norm": 1.6079185009002686, + "learning_rate": 6.118140005928395e-06, + "loss": 0.3476, + "step": 20232 + }, + { + "epoch": 0.6381643273931557, + "grad_norm": 1.4136735200881958, + "learning_rate": 6.117198579089509e-06, + "loss": 0.4496, + "step": 20233 + }, + { + "epoch": 0.6381958681595963, + "grad_norm": 1.5007214546203613, + "learning_rate": 6.116257192770671e-06, + "loss": 0.424, + "step": 20234 + }, + { + "epoch": 0.6382274089260369, + "grad_norm": 1.5711687803268433, + "learning_rate": 6.115315846981707e-06, + "loss": 0.4497, + "step": 20235 + }, + { + "epoch": 0.6382589496924775, + "grad_norm": 1.8168189525604248, + "learning_rate": 6.114374541732436e-06, + "loss": 0.4255, + "step": 20236 + }, + { + "epoch": 0.6382904904589182, + "grad_norm": 1.7664607763290405, + "learning_rate": 6.113433277032681e-06, + "loss": 0.4288, + "step": 20237 + }, + { + "epoch": 0.6383220312253588, + "grad_norm": 1.8506054878234863, + "learning_rate": 6.112492052892273e-06, + "loss": 0.4616, + "step": 20238 + }, + { + "epoch": 0.6383535719917994, + "grad_norm": 1.9384143352508545, + "learning_rate": 6.111550869321023e-06, + "loss": 0.3436, + "step": 20239 + }, + { + "epoch": 0.6383851127582401, + "grad_norm": 1.5860930681228638, + "learning_rate": 6.110609726328764e-06, + "loss": 0.3675, + "step": 20240 + }, + { + "epoch": 0.6384166535246807, + "grad_norm": 2.0056464672088623, + "learning_rate": 6.109668623925312e-06, + "loss": 0.4458, + "step": 20241 + }, + { + "epoch": 0.6384481942911213, + "grad_norm": 1.9166669845581055, + "learning_rate": 6.1087275621204865e-06, + "loss": 0.4731, + "step": 20242 + }, + { + "epoch": 0.638479735057562, + "grad_norm": 1.7320297956466675, + "learning_rate": 6.107786540924112e-06, + "loss": 0.4181, + "step": 20243 + }, + { + "epoch": 0.6385112758240026, + "grad_norm": 2.4322471618652344, + "learning_rate": 6.1068455603460085e-06, + "loss": 0.43, + "step": 20244 + }, + { + "epoch": 0.6385428165904431, + "grad_norm": 1.5935487747192383, + "learning_rate": 6.105904620395992e-06, + "loss": 0.4285, + "step": 20245 + }, + { + "epoch": 0.6385743573568837, + "grad_norm": 1.5758452415466309, + "learning_rate": 6.104963721083887e-06, + "loss": 0.4255, + "step": 20246 + }, + { + "epoch": 0.6386058981233244, + "grad_norm": 1.5717800855636597, + "learning_rate": 6.104022862419509e-06, + "loss": 0.4445, + "step": 20247 + }, + { + "epoch": 0.638637438889765, + "grad_norm": 1.6682556867599487, + "learning_rate": 6.103082044412673e-06, + "loss": 0.4408, + "step": 20248 + }, + { + "epoch": 0.6386689796562056, + "grad_norm": 1.3874377012252808, + "learning_rate": 6.102141267073207e-06, + "loss": 0.6928, + "step": 20249 + }, + { + "epoch": 0.6387005204226462, + "grad_norm": 1.9728078842163086, + "learning_rate": 6.101200530410925e-06, + "loss": 0.4003, + "step": 20250 + }, + { + "epoch": 0.6387320611890869, + "grad_norm": 2.2029852867126465, + "learning_rate": 6.1002598344356354e-06, + "loss": 0.3387, + "step": 20251 + }, + { + "epoch": 0.6387636019555275, + "grad_norm": 2.061312675476074, + "learning_rate": 6.099319179157169e-06, + "loss": 0.4007, + "step": 20252 + }, + { + "epoch": 0.6387951427219681, + "grad_norm": 1.5435705184936523, + "learning_rate": 6.098378564585336e-06, + "loss": 0.4459, + "step": 20253 + }, + { + "epoch": 0.6388266834884088, + "grad_norm": 1.580655574798584, + "learning_rate": 6.097437990729949e-06, + "loss": 0.3982, + "step": 20254 + }, + { + "epoch": 0.6388582242548494, + "grad_norm": 1.8928440809249878, + "learning_rate": 6.096497457600829e-06, + "loss": 0.4452, + "step": 20255 + }, + { + "epoch": 0.63888976502129, + "grad_norm": 1.215378761291504, + "learning_rate": 6.095556965207785e-06, + "loss": 0.3324, + "step": 20256 + }, + { + "epoch": 0.6389213057877307, + "grad_norm": 3.042330503463745, + "learning_rate": 6.094616513560639e-06, + "loss": 0.4555, + "step": 20257 + }, + { + "epoch": 0.6389528465541713, + "grad_norm": 1.300713062286377, + "learning_rate": 6.093676102669204e-06, + "loss": 0.7231, + "step": 20258 + }, + { + "epoch": 0.6389843873206119, + "grad_norm": 2.012148380279541, + "learning_rate": 6.092735732543287e-06, + "loss": 0.6996, + "step": 20259 + }, + { + "epoch": 0.6390159280870525, + "grad_norm": 1.6577621698379517, + "learning_rate": 6.09179540319271e-06, + "loss": 0.373, + "step": 20260 + }, + { + "epoch": 0.6390474688534932, + "grad_norm": 1.9105100631713867, + "learning_rate": 6.090855114627283e-06, + "loss": 0.4499, + "step": 20261 + }, + { + "epoch": 0.6390790096199338, + "grad_norm": 1.6972259283065796, + "learning_rate": 6.0899148668568164e-06, + "loss": 0.4349, + "step": 20262 + }, + { + "epoch": 0.6391105503863744, + "grad_norm": 1.651360273361206, + "learning_rate": 6.0889746598911245e-06, + "loss": 0.4547, + "step": 20263 + }, + { + "epoch": 0.6391420911528151, + "grad_norm": 1.6062273979187012, + "learning_rate": 6.08803449374002e-06, + "loss": 0.3477, + "step": 20264 + }, + { + "epoch": 0.6391736319192557, + "grad_norm": 1.4406793117523193, + "learning_rate": 6.0870943684133136e-06, + "loss": 0.4127, + "step": 20265 + }, + { + "epoch": 0.6392051726856962, + "grad_norm": 1.5322195291519165, + "learning_rate": 6.086154283920815e-06, + "loss": 0.4676, + "step": 20266 + }, + { + "epoch": 0.6392367134521368, + "grad_norm": 2.624783992767334, + "learning_rate": 6.085214240272337e-06, + "loss": 0.4267, + "step": 20267 + }, + { + "epoch": 0.6392682542185775, + "grad_norm": 1.5915930271148682, + "learning_rate": 6.084274237477683e-06, + "loss": 0.4331, + "step": 20268 + }, + { + "epoch": 0.6392997949850181, + "grad_norm": 2.419286012649536, + "learning_rate": 6.083334275546673e-06, + "loss": 0.3989, + "step": 20269 + }, + { + "epoch": 0.6393313357514587, + "grad_norm": 1.7314426898956299, + "learning_rate": 6.0823943544891065e-06, + "loss": 0.4034, + "step": 20270 + }, + { + "epoch": 0.6393628765178994, + "grad_norm": 2.0071895122528076, + "learning_rate": 6.081454474314801e-06, + "loss": 0.4525, + "step": 20271 + }, + { + "epoch": 0.63939441728434, + "grad_norm": 1.5376088619232178, + "learning_rate": 6.080514635033561e-06, + "loss": 0.4312, + "step": 20272 + }, + { + "epoch": 0.6394259580507806, + "grad_norm": 1.444807529449463, + "learning_rate": 6.079574836655191e-06, + "loss": 0.4135, + "step": 20273 + }, + { + "epoch": 0.6394574988172212, + "grad_norm": 2.2026219367980957, + "learning_rate": 6.0786350791895046e-06, + "loss": 0.4315, + "step": 20274 + }, + { + "epoch": 0.6394890395836619, + "grad_norm": 1.549897313117981, + "learning_rate": 6.077695362646305e-06, + "loss": 0.4465, + "step": 20275 + }, + { + "epoch": 0.6395205803501025, + "grad_norm": 2.04692006111145, + "learning_rate": 6.076755687035398e-06, + "loss": 0.755, + "step": 20276 + }, + { + "epoch": 0.6395521211165431, + "grad_norm": 1.9357781410217285, + "learning_rate": 6.075816052366595e-06, + "loss": 0.7285, + "step": 20277 + }, + { + "epoch": 0.6395836618829838, + "grad_norm": 1.2579153776168823, + "learning_rate": 6.074876458649696e-06, + "loss": 0.3268, + "step": 20278 + }, + { + "epoch": 0.6396152026494244, + "grad_norm": 1.6763356924057007, + "learning_rate": 6.073936905894508e-06, + "loss": 0.4255, + "step": 20279 + }, + { + "epoch": 0.639646743415865, + "grad_norm": 1.7964940071105957, + "learning_rate": 6.07299739411084e-06, + "loss": 0.4523, + "step": 20280 + }, + { + "epoch": 0.6396782841823057, + "grad_norm": 1.6130374670028687, + "learning_rate": 6.0720579233084934e-06, + "loss": 0.4447, + "step": 20281 + }, + { + "epoch": 0.6397098249487463, + "grad_norm": 1.589110255241394, + "learning_rate": 6.07111849349727e-06, + "loss": 0.4234, + "step": 20282 + }, + { + "epoch": 0.6397413657151869, + "grad_norm": 1.5083229541778564, + "learning_rate": 6.070179104686975e-06, + "loss": 0.4394, + "step": 20283 + }, + { + "epoch": 0.6397729064816275, + "grad_norm": 2.2229385375976562, + "learning_rate": 6.069239756887417e-06, + "loss": 0.424, + "step": 20284 + }, + { + "epoch": 0.6398044472480682, + "grad_norm": 2.572300910949707, + "learning_rate": 6.068300450108391e-06, + "loss": 0.4491, + "step": 20285 + }, + { + "epoch": 0.6398359880145088, + "grad_norm": 1.698760747909546, + "learning_rate": 6.067361184359704e-06, + "loss": 0.3641, + "step": 20286 + }, + { + "epoch": 0.6398675287809493, + "grad_norm": 1.471129298210144, + "learning_rate": 6.066421959651153e-06, + "loss": 0.4103, + "step": 20287 + }, + { + "epoch": 0.63989906954739, + "grad_norm": 1.312812089920044, + "learning_rate": 6.065482775992547e-06, + "loss": 0.4287, + "step": 20288 + }, + { + "epoch": 0.6399306103138306, + "grad_norm": 1.3439662456512451, + "learning_rate": 6.064543633393684e-06, + "loss": 0.4127, + "step": 20289 + }, + { + "epoch": 0.6399621510802712, + "grad_norm": 5.803780555725098, + "learning_rate": 6.06360453186436e-06, + "loss": 0.4415, + "step": 20290 + }, + { + "epoch": 0.6399936918467118, + "grad_norm": 1.7487672567367554, + "learning_rate": 6.0626654714143814e-06, + "loss": 0.4225, + "step": 20291 + }, + { + "epoch": 0.6400252326131525, + "grad_norm": 1.8519176244735718, + "learning_rate": 6.061726452053549e-06, + "loss": 0.3925, + "step": 20292 + }, + { + "epoch": 0.6400567733795931, + "grad_norm": 1.4709120988845825, + "learning_rate": 6.060787473791655e-06, + "loss": 0.4716, + "step": 20293 + }, + { + "epoch": 0.6400883141460337, + "grad_norm": 5.120767116546631, + "learning_rate": 6.059848536638504e-06, + "loss": 0.4415, + "step": 20294 + }, + { + "epoch": 0.6401198549124744, + "grad_norm": 1.6253321170806885, + "learning_rate": 6.058909640603894e-06, + "loss": 0.3686, + "step": 20295 + }, + { + "epoch": 0.640151395678915, + "grad_norm": 1.9427995681762695, + "learning_rate": 6.057970785697618e-06, + "loss": 0.3898, + "step": 20296 + }, + { + "epoch": 0.6401829364453556, + "grad_norm": 1.5302119255065918, + "learning_rate": 6.057031971929481e-06, + "loss": 0.46, + "step": 20297 + }, + { + "epoch": 0.6402144772117963, + "grad_norm": 1.4177073240280151, + "learning_rate": 6.056093199309279e-06, + "loss": 0.7068, + "step": 20298 + }, + { + "epoch": 0.6402460179782369, + "grad_norm": 1.671549916267395, + "learning_rate": 6.055154467846801e-06, + "loss": 0.4418, + "step": 20299 + }, + { + "epoch": 0.6402775587446775, + "grad_norm": 1.7433263063430786, + "learning_rate": 6.054215777551854e-06, + "loss": 0.4038, + "step": 20300 + }, + { + "epoch": 0.6403090995111181, + "grad_norm": 2.191047191619873, + "learning_rate": 6.053277128434227e-06, + "loss": 0.4183, + "step": 20301 + }, + { + "epoch": 0.6403406402775588, + "grad_norm": 1.4824050664901733, + "learning_rate": 6.05233852050372e-06, + "loss": 0.454, + "step": 20302 + }, + { + "epoch": 0.6403721810439994, + "grad_norm": 1.7573671340942383, + "learning_rate": 6.051399953770123e-06, + "loss": 0.4521, + "step": 20303 + }, + { + "epoch": 0.64040372181044, + "grad_norm": 1.9308000802993774, + "learning_rate": 6.0504614282432325e-06, + "loss": 0.4554, + "step": 20304 + }, + { + "epoch": 0.6404352625768807, + "grad_norm": 1.7724359035491943, + "learning_rate": 6.0495229439328486e-06, + "loss": 0.4296, + "step": 20305 + }, + { + "epoch": 0.6404668033433213, + "grad_norm": 1.7091971635818481, + "learning_rate": 6.0485845008487584e-06, + "loss": 0.4544, + "step": 20306 + }, + { + "epoch": 0.6404983441097619, + "grad_norm": 1.572035312652588, + "learning_rate": 6.047646099000754e-06, + "loss": 0.426, + "step": 20307 + }, + { + "epoch": 0.6405298848762024, + "grad_norm": 1.4759671688079834, + "learning_rate": 6.046707738398634e-06, + "loss": 0.3534, + "step": 20308 + }, + { + "epoch": 0.6405614256426431, + "grad_norm": 1.298159122467041, + "learning_rate": 6.04576941905219e-06, + "loss": 0.7601, + "step": 20309 + }, + { + "epoch": 0.6405929664090837, + "grad_norm": 1.5168997049331665, + "learning_rate": 6.044831140971208e-06, + "loss": 0.7154, + "step": 20310 + }, + { + "epoch": 0.6406245071755243, + "grad_norm": 1.8278025388717651, + "learning_rate": 6.043892904165488e-06, + "loss": 0.3993, + "step": 20311 + }, + { + "epoch": 0.640656047941965, + "grad_norm": 1.763238549232483, + "learning_rate": 6.042954708644817e-06, + "loss": 0.4277, + "step": 20312 + }, + { + "epoch": 0.6406875887084056, + "grad_norm": 2.1196935176849365, + "learning_rate": 6.042016554418985e-06, + "loss": 0.373, + "step": 20313 + }, + { + "epoch": 0.6407191294748462, + "grad_norm": 1.9198582172393799, + "learning_rate": 6.041078441497785e-06, + "loss": 0.701, + "step": 20314 + }, + { + "epoch": 0.6407506702412868, + "grad_norm": 1.5890909433364868, + "learning_rate": 6.040140369891005e-06, + "loss": 0.4572, + "step": 20315 + }, + { + "epoch": 0.6407822110077275, + "grad_norm": 2.4922285079956055, + "learning_rate": 6.039202339608432e-06, + "loss": 0.3913, + "step": 20316 + }, + { + "epoch": 0.6408137517741681, + "grad_norm": 1.431626558303833, + "learning_rate": 6.0382643506598615e-06, + "loss": 0.4127, + "step": 20317 + }, + { + "epoch": 0.6408452925406087, + "grad_norm": 2.2495553493499756, + "learning_rate": 6.037326403055075e-06, + "loss": 0.7097, + "step": 20318 + }, + { + "epoch": 0.6408768333070494, + "grad_norm": 1.4167518615722656, + "learning_rate": 6.036388496803869e-06, + "loss": 0.3488, + "step": 20319 + }, + { + "epoch": 0.64090837407349, + "grad_norm": 1.5398749113082886, + "learning_rate": 6.0354506319160266e-06, + "loss": 0.4256, + "step": 20320 + }, + { + "epoch": 0.6409399148399306, + "grad_norm": 1.7317482233047485, + "learning_rate": 6.034512808401331e-06, + "loss": 0.4386, + "step": 20321 + }, + { + "epoch": 0.6409714556063713, + "grad_norm": 1.966434121131897, + "learning_rate": 6.033575026269578e-06, + "loss": 0.4252, + "step": 20322 + }, + { + "epoch": 0.6410029963728119, + "grad_norm": 1.4395776987075806, + "learning_rate": 6.032637285530547e-06, + "loss": 0.4549, + "step": 20323 + }, + { + "epoch": 0.6410345371392525, + "grad_norm": 1.5546445846557617, + "learning_rate": 6.031699586194027e-06, + "loss": 0.4278, + "step": 20324 + }, + { + "epoch": 0.6410660779056931, + "grad_norm": 1.552211046218872, + "learning_rate": 6.030761928269803e-06, + "loss": 0.4535, + "step": 20325 + }, + { + "epoch": 0.6410976186721338, + "grad_norm": 1.4249004125595093, + "learning_rate": 6.029824311767662e-06, + "loss": 0.7395, + "step": 20326 + }, + { + "epoch": 0.6411291594385744, + "grad_norm": 1.3066624402999878, + "learning_rate": 6.028886736697381e-06, + "loss": 0.6992, + "step": 20327 + }, + { + "epoch": 0.641160700205015, + "grad_norm": 1.7175397872924805, + "learning_rate": 6.027949203068757e-06, + "loss": 0.4311, + "step": 20328 + }, + { + "epoch": 0.6411922409714556, + "grad_norm": 1.5674495697021484, + "learning_rate": 6.027011710891566e-06, + "loss": 0.4275, + "step": 20329 + }, + { + "epoch": 0.6412237817378962, + "grad_norm": 6.54785680770874, + "learning_rate": 6.026074260175589e-06, + "loss": 0.4459, + "step": 20330 + }, + { + "epoch": 0.6412553225043368, + "grad_norm": 1.41643226146698, + "learning_rate": 6.025136850930616e-06, + "loss": 0.3586, + "step": 20331 + }, + { + "epoch": 0.6412868632707774, + "grad_norm": 1.4518951177597046, + "learning_rate": 6.024199483166427e-06, + "loss": 0.4705, + "step": 20332 + }, + { + "epoch": 0.6413184040372181, + "grad_norm": 1.8812365531921387, + "learning_rate": 6.0232621568928015e-06, + "loss": 0.4309, + "step": 20333 + }, + { + "epoch": 0.6413499448036587, + "grad_norm": 1.5721653699874878, + "learning_rate": 6.022324872119526e-06, + "loss": 0.4314, + "step": 20334 + }, + { + "epoch": 0.6413814855700993, + "grad_norm": 1.6006275415420532, + "learning_rate": 6.021387628856374e-06, + "loss": 0.4641, + "step": 20335 + }, + { + "epoch": 0.64141302633654, + "grad_norm": 1.6096080541610718, + "learning_rate": 6.020450427113136e-06, + "loss": 0.4217, + "step": 20336 + }, + { + "epoch": 0.6414445671029806, + "grad_norm": 1.7353469133377075, + "learning_rate": 6.019513266899588e-06, + "loss": 0.4331, + "step": 20337 + }, + { + "epoch": 0.6414761078694212, + "grad_norm": 1.855749249458313, + "learning_rate": 6.018576148225505e-06, + "loss": 0.4475, + "step": 20338 + }, + { + "epoch": 0.6415076486358618, + "grad_norm": 2.1927621364593506, + "learning_rate": 6.017639071100676e-06, + "loss": 0.4242, + "step": 20339 + }, + { + "epoch": 0.6415391894023025, + "grad_norm": 2.0447518825531006, + "learning_rate": 6.016702035534876e-06, + "loss": 0.4136, + "step": 20340 + }, + { + "epoch": 0.6415707301687431, + "grad_norm": 1.596921682357788, + "learning_rate": 6.01576504153788e-06, + "loss": 0.4174, + "step": 20341 + }, + { + "epoch": 0.6416022709351837, + "grad_norm": 1.7451937198638916, + "learning_rate": 6.0148280891194724e-06, + "loss": 0.4584, + "step": 20342 + }, + { + "epoch": 0.6416338117016244, + "grad_norm": 1.8696547746658325, + "learning_rate": 6.013891178289426e-06, + "loss": 0.3441, + "step": 20343 + }, + { + "epoch": 0.641665352468065, + "grad_norm": 1.660357117652893, + "learning_rate": 6.0129543090575195e-06, + "loss": 0.4094, + "step": 20344 + }, + { + "epoch": 0.6416968932345056, + "grad_norm": 1.6578776836395264, + "learning_rate": 6.012017481433533e-06, + "loss": 0.4093, + "step": 20345 + }, + { + "epoch": 0.6417284340009463, + "grad_norm": 1.2878878116607666, + "learning_rate": 6.011080695427241e-06, + "loss": 0.436, + "step": 20346 + }, + { + "epoch": 0.6417599747673869, + "grad_norm": 1.731794834136963, + "learning_rate": 6.010143951048415e-06, + "loss": 0.4466, + "step": 20347 + }, + { + "epoch": 0.6417915155338275, + "grad_norm": 1.4962732791900635, + "learning_rate": 6.0092072483068386e-06, + "loss": 0.4478, + "step": 20348 + }, + { + "epoch": 0.6418230563002681, + "grad_norm": 1.4639642238616943, + "learning_rate": 6.008270587212279e-06, + "loss": 0.4263, + "step": 20349 + }, + { + "epoch": 0.6418545970667088, + "grad_norm": 1.306204915046692, + "learning_rate": 6.007333967774521e-06, + "loss": 0.4344, + "step": 20350 + }, + { + "epoch": 0.6418861378331493, + "grad_norm": 1.45844304561615, + "learning_rate": 6.006397390003332e-06, + "loss": 0.4119, + "step": 20351 + }, + { + "epoch": 0.6419176785995899, + "grad_norm": 1.8119778633117676, + "learning_rate": 6.005460853908486e-06, + "loss": 0.4113, + "step": 20352 + }, + { + "epoch": 0.6419492193660306, + "grad_norm": 1.5176550149917603, + "learning_rate": 6.004524359499759e-06, + "loss": 0.4578, + "step": 20353 + }, + { + "epoch": 0.6419807601324712, + "grad_norm": 4.62919282913208, + "learning_rate": 6.003587906786923e-06, + "loss": 0.4089, + "step": 20354 + }, + { + "epoch": 0.6420123008989118, + "grad_norm": 1.5429877042770386, + "learning_rate": 6.0026514957797456e-06, + "loss": 0.4365, + "step": 20355 + }, + { + "epoch": 0.6420438416653524, + "grad_norm": 1.8607598543167114, + "learning_rate": 6.001715126488009e-06, + "loss": 0.451, + "step": 20356 + }, + { + "epoch": 0.6420753824317931, + "grad_norm": 1.699765682220459, + "learning_rate": 6.000778798921478e-06, + "loss": 0.4247, + "step": 20357 + }, + { + "epoch": 0.6421069231982337, + "grad_norm": 1.6260054111480713, + "learning_rate": 5.999842513089924e-06, + "loss": 0.4282, + "step": 20358 + }, + { + "epoch": 0.6421384639646743, + "grad_norm": 1.9583615064620972, + "learning_rate": 5.998906269003122e-06, + "loss": 0.3863, + "step": 20359 + }, + { + "epoch": 0.642170004731115, + "grad_norm": 1.3814749717712402, + "learning_rate": 5.9979700666708396e-06, + "loss": 0.4365, + "step": 20360 + }, + { + "epoch": 0.6422015454975556, + "grad_norm": 1.5543795824050903, + "learning_rate": 5.9970339061028445e-06, + "loss": 0.4717, + "step": 20361 + }, + { + "epoch": 0.6422330862639962, + "grad_norm": 3.766242027282715, + "learning_rate": 5.996097787308912e-06, + "loss": 0.3836, + "step": 20362 + }, + { + "epoch": 0.6422646270304369, + "grad_norm": 1.6422480344772339, + "learning_rate": 5.995161710298806e-06, + "loss": 0.4458, + "step": 20363 + }, + { + "epoch": 0.6422961677968775, + "grad_norm": 1.451736330986023, + "learning_rate": 5.994225675082297e-06, + "loss": 0.42, + "step": 20364 + }, + { + "epoch": 0.6423277085633181, + "grad_norm": 1.7936853170394897, + "learning_rate": 5.993289681669154e-06, + "loss": 0.3608, + "step": 20365 + }, + { + "epoch": 0.6423592493297587, + "grad_norm": 1.7316007614135742, + "learning_rate": 5.9923537300691425e-06, + "loss": 0.4488, + "step": 20366 + }, + { + "epoch": 0.6423907900961994, + "grad_norm": 1.4327459335327148, + "learning_rate": 5.991417820292035e-06, + "loss": 0.4248, + "step": 20367 + }, + { + "epoch": 0.64242233086264, + "grad_norm": 5.753445625305176, + "learning_rate": 5.990481952347595e-06, + "loss": 0.4376, + "step": 20368 + }, + { + "epoch": 0.6424538716290806, + "grad_norm": 1.7133702039718628, + "learning_rate": 5.989546126245584e-06, + "loss": 0.4567, + "step": 20369 + }, + { + "epoch": 0.6424854123955213, + "grad_norm": 1.5337293148040771, + "learning_rate": 5.9886103419957786e-06, + "loss": 0.4337, + "step": 20370 + }, + { + "epoch": 0.6425169531619619, + "grad_norm": 1.5388503074645996, + "learning_rate": 5.9876745996079386e-06, + "loss": 0.4323, + "step": 20371 + }, + { + "epoch": 0.6425484939284024, + "grad_norm": 2.018824577331543, + "learning_rate": 5.986738899091828e-06, + "loss": 0.397, + "step": 20372 + }, + { + "epoch": 0.642580034694843, + "grad_norm": 1.7330327033996582, + "learning_rate": 5.985803240457215e-06, + "loss": 0.4004, + "step": 20373 + }, + { + "epoch": 0.6426115754612837, + "grad_norm": 1.3705501556396484, + "learning_rate": 5.984867623713862e-06, + "loss": 0.4277, + "step": 20374 + }, + { + "epoch": 0.6426431162277243, + "grad_norm": 1.6576765775680542, + "learning_rate": 5.983932048871529e-06, + "loss": 0.4448, + "step": 20375 + }, + { + "epoch": 0.6426746569941649, + "grad_norm": 1.5517812967300415, + "learning_rate": 5.9829965159399875e-06, + "loss": 0.3498, + "step": 20376 + }, + { + "epoch": 0.6427061977606056, + "grad_norm": 1.6076871156692505, + "learning_rate": 5.982061024928997e-06, + "loss": 0.4642, + "step": 20377 + }, + { + "epoch": 0.6427377385270462, + "grad_norm": 1.299150824546814, + "learning_rate": 5.981125575848315e-06, + "loss": 0.7134, + "step": 20378 + }, + { + "epoch": 0.6427692792934868, + "grad_norm": 1.4572848081588745, + "learning_rate": 5.980190168707713e-06, + "loss": 0.4736, + "step": 20379 + }, + { + "epoch": 0.6428008200599274, + "grad_norm": 1.518951416015625, + "learning_rate": 5.979254803516948e-06, + "loss": 0.4026, + "step": 20380 + }, + { + "epoch": 0.6428323608263681, + "grad_norm": 1.7824987173080444, + "learning_rate": 5.9783194802857766e-06, + "loss": 0.4285, + "step": 20381 + }, + { + "epoch": 0.6428639015928087, + "grad_norm": 1.7228199243545532, + "learning_rate": 5.977384199023968e-06, + "loss": 0.4707, + "step": 20382 + }, + { + "epoch": 0.6428954423592493, + "grad_norm": 1.4237498044967651, + "learning_rate": 5.9764489597412744e-06, + "loss": 0.6811, + "step": 20383 + }, + { + "epoch": 0.64292698312569, + "grad_norm": 1.8782954216003418, + "learning_rate": 5.975513762447465e-06, + "loss": 0.4407, + "step": 20384 + }, + { + "epoch": 0.6429585238921306, + "grad_norm": 1.7104437351226807, + "learning_rate": 5.9745786071522935e-06, + "loss": 0.4325, + "step": 20385 + }, + { + "epoch": 0.6429900646585712, + "grad_norm": 1.425865888595581, + "learning_rate": 5.973643493865516e-06, + "loss": 0.4236, + "step": 20386 + }, + { + "epoch": 0.6430216054250119, + "grad_norm": 1.60845947265625, + "learning_rate": 5.9727084225969e-06, + "loss": 0.4562, + "step": 20387 + }, + { + "epoch": 0.6430531461914525, + "grad_norm": 1.5955636501312256, + "learning_rate": 5.971773393356197e-06, + "loss": 0.4219, + "step": 20388 + }, + { + "epoch": 0.6430846869578931, + "grad_norm": 1.7312424182891846, + "learning_rate": 5.970838406153163e-06, + "loss": 0.4306, + "step": 20389 + }, + { + "epoch": 0.6431162277243337, + "grad_norm": 1.4325188398361206, + "learning_rate": 5.969903460997564e-06, + "loss": 0.4119, + "step": 20390 + }, + { + "epoch": 0.6431477684907744, + "grad_norm": 1.9855128526687622, + "learning_rate": 5.968968557899151e-06, + "loss": 0.4253, + "step": 20391 + }, + { + "epoch": 0.643179309257215, + "grad_norm": 1.3137658834457397, + "learning_rate": 5.968033696867679e-06, + "loss": 0.7296, + "step": 20392 + }, + { + "epoch": 0.6432108500236555, + "grad_norm": 2.434356689453125, + "learning_rate": 5.967098877912909e-06, + "loss": 0.3317, + "step": 20393 + }, + { + "epoch": 0.6432423907900962, + "grad_norm": 1.883201241493225, + "learning_rate": 5.966164101044591e-06, + "loss": 0.466, + "step": 20394 + }, + { + "epoch": 0.6432739315565368, + "grad_norm": 1.479690432548523, + "learning_rate": 5.965229366272482e-06, + "loss": 0.3864, + "step": 20395 + }, + { + "epoch": 0.6433054723229774, + "grad_norm": 1.463636040687561, + "learning_rate": 5.96429467360634e-06, + "loss": 0.4456, + "step": 20396 + }, + { + "epoch": 0.643337013089418, + "grad_norm": 1.7434203624725342, + "learning_rate": 5.963360023055913e-06, + "loss": 0.429, + "step": 20397 + }, + { + "epoch": 0.6433685538558587, + "grad_norm": 1.5616899728775024, + "learning_rate": 5.9624254146309616e-06, + "loss": 0.4551, + "step": 20398 + }, + { + "epoch": 0.6434000946222993, + "grad_norm": 2.061460256576538, + "learning_rate": 5.961490848341237e-06, + "loss": 0.4563, + "step": 20399 + }, + { + "epoch": 0.6434316353887399, + "grad_norm": 1.4261236190795898, + "learning_rate": 5.960556324196489e-06, + "loss": 0.4409, + "step": 20400 + }, + { + "epoch": 0.6434631761551806, + "grad_norm": 1.3065321445465088, + "learning_rate": 5.959621842206474e-06, + "loss": 0.4447, + "step": 20401 + }, + { + "epoch": 0.6434947169216212, + "grad_norm": 1.6311897039413452, + "learning_rate": 5.958687402380943e-06, + "loss": 0.4521, + "step": 20402 + }, + { + "epoch": 0.6435262576880618, + "grad_norm": 1.8489466905593872, + "learning_rate": 5.957753004729643e-06, + "loss": 0.3884, + "step": 20403 + }, + { + "epoch": 0.6435577984545024, + "grad_norm": 1.259808897972107, + "learning_rate": 5.9568186492623325e-06, + "loss": 0.3586, + "step": 20404 + }, + { + "epoch": 0.6435893392209431, + "grad_norm": 1.8741495609283447, + "learning_rate": 5.95588433598876e-06, + "loss": 0.4606, + "step": 20405 + }, + { + "epoch": 0.6436208799873837, + "grad_norm": 2.01834774017334, + "learning_rate": 5.95495006491867e-06, + "loss": 0.4566, + "step": 20406 + }, + { + "epoch": 0.6436524207538243, + "grad_norm": 1.9015430212020874, + "learning_rate": 5.954015836061819e-06, + "loss": 0.4226, + "step": 20407 + }, + { + "epoch": 0.643683961520265, + "grad_norm": 1.8337197303771973, + "learning_rate": 5.953081649427956e-06, + "loss": 0.7274, + "step": 20408 + }, + { + "epoch": 0.6437155022867056, + "grad_norm": 1.4475605487823486, + "learning_rate": 5.952147505026826e-06, + "loss": 0.4369, + "step": 20409 + }, + { + "epoch": 0.6437470430531462, + "grad_norm": 1.9903101921081543, + "learning_rate": 5.951213402868181e-06, + "loss": 0.4578, + "step": 20410 + }, + { + "epoch": 0.6437785838195869, + "grad_norm": 1.7755317687988281, + "learning_rate": 5.95027934296177e-06, + "loss": 0.4169, + "step": 20411 + }, + { + "epoch": 0.6438101245860275, + "grad_norm": 1.6226404905319214, + "learning_rate": 5.949345325317336e-06, + "loss": 0.441, + "step": 20412 + }, + { + "epoch": 0.6438416653524681, + "grad_norm": 1.6234302520751953, + "learning_rate": 5.948411349944631e-06, + "loss": 0.4519, + "step": 20413 + }, + { + "epoch": 0.6438732061189086, + "grad_norm": 2.2892439365386963, + "learning_rate": 5.947477416853395e-06, + "loss": 0.4749, + "step": 20414 + }, + { + "epoch": 0.6439047468853493, + "grad_norm": 1.242987871170044, + "learning_rate": 5.946543526053383e-06, + "loss": 0.3333, + "step": 20415 + }, + { + "epoch": 0.6439362876517899, + "grad_norm": 3.489229202270508, + "learning_rate": 5.945609677554336e-06, + "loss": 0.402, + "step": 20416 + }, + { + "epoch": 0.6439678284182305, + "grad_norm": 2.6275572776794434, + "learning_rate": 5.944675871365997e-06, + "loss": 0.4531, + "step": 20417 + }, + { + "epoch": 0.6439993691846712, + "grad_norm": 1.4847266674041748, + "learning_rate": 5.943742107498118e-06, + "loss": 0.4419, + "step": 20418 + }, + { + "epoch": 0.6440309099511118, + "grad_norm": 1.7421698570251465, + "learning_rate": 5.94280838596044e-06, + "loss": 0.3893, + "step": 20419 + }, + { + "epoch": 0.6440624507175524, + "grad_norm": 1.4575964212417603, + "learning_rate": 5.941874706762705e-06, + "loss": 0.4468, + "step": 20420 + }, + { + "epoch": 0.644093991483993, + "grad_norm": 1.9838019609451294, + "learning_rate": 5.94094106991466e-06, + "loss": 0.4704, + "step": 20421 + }, + { + "epoch": 0.6441255322504337, + "grad_norm": 1.577209711074829, + "learning_rate": 5.9400074754260466e-06, + "loss": 0.4589, + "step": 20422 + }, + { + "epoch": 0.6441570730168743, + "grad_norm": 1.5456788539886475, + "learning_rate": 5.939073923306604e-06, + "loss": 0.4162, + "step": 20423 + }, + { + "epoch": 0.6441886137833149, + "grad_norm": 1.5481212139129639, + "learning_rate": 5.938140413566082e-06, + "loss": 0.4622, + "step": 20424 + }, + { + "epoch": 0.6442201545497556, + "grad_norm": 1.7206507921218872, + "learning_rate": 5.937206946214219e-06, + "loss": 0.4324, + "step": 20425 + }, + { + "epoch": 0.6442516953161962, + "grad_norm": 1.434249758720398, + "learning_rate": 5.936273521260752e-06, + "loss": 0.4375, + "step": 20426 + }, + { + "epoch": 0.6442832360826368, + "grad_norm": 1.841257929801941, + "learning_rate": 5.9353401387154306e-06, + "loss": 0.4208, + "step": 20427 + }, + { + "epoch": 0.6443147768490775, + "grad_norm": 1.3812485933303833, + "learning_rate": 5.9344067985879906e-06, + "loss": 0.4214, + "step": 20428 + }, + { + "epoch": 0.6443463176155181, + "grad_norm": 1.1361117362976074, + "learning_rate": 5.933473500888168e-06, + "loss": 0.3313, + "step": 20429 + }, + { + "epoch": 0.6443778583819587, + "grad_norm": 2.1710033416748047, + "learning_rate": 5.932540245625712e-06, + "loss": 0.4623, + "step": 20430 + }, + { + "epoch": 0.6444093991483993, + "grad_norm": 1.4314943552017212, + "learning_rate": 5.931607032810354e-06, + "loss": 0.449, + "step": 20431 + }, + { + "epoch": 0.64444093991484, + "grad_norm": 1.6530911922454834, + "learning_rate": 5.930673862451839e-06, + "loss": 0.4081, + "step": 20432 + }, + { + "epoch": 0.6444724806812806, + "grad_norm": 1.4710670709609985, + "learning_rate": 5.929740734559901e-06, + "loss": 0.4481, + "step": 20433 + }, + { + "epoch": 0.6445040214477212, + "grad_norm": 2.1205220222473145, + "learning_rate": 5.928807649144275e-06, + "loss": 0.7633, + "step": 20434 + }, + { + "epoch": 0.6445355622141618, + "grad_norm": 1.9828158617019653, + "learning_rate": 5.927874606214705e-06, + "loss": 0.4241, + "step": 20435 + }, + { + "epoch": 0.6445671029806024, + "grad_norm": 1.574032187461853, + "learning_rate": 5.926941605780927e-06, + "loss": 0.4184, + "step": 20436 + }, + { + "epoch": 0.644598643747043, + "grad_norm": 1.9628357887268066, + "learning_rate": 5.926008647852673e-06, + "loss": 0.4586, + "step": 20437 + }, + { + "epoch": 0.6446301845134836, + "grad_norm": 1.5875226259231567, + "learning_rate": 5.9250757324396845e-06, + "loss": 0.4411, + "step": 20438 + }, + { + "epoch": 0.6446617252799243, + "grad_norm": 1.5375874042510986, + "learning_rate": 5.924142859551695e-06, + "loss": 0.3449, + "step": 20439 + }, + { + "epoch": 0.6446932660463649, + "grad_norm": 1.5345737934112549, + "learning_rate": 5.923210029198437e-06, + "loss": 0.4457, + "step": 20440 + }, + { + "epoch": 0.6447248068128055, + "grad_norm": 1.4171520471572876, + "learning_rate": 5.922277241389651e-06, + "loss": 0.4345, + "step": 20441 + }, + { + "epoch": 0.6447563475792462, + "grad_norm": 2.1164391040802, + "learning_rate": 5.921344496135069e-06, + "loss": 0.4504, + "step": 20442 + }, + { + "epoch": 0.6447878883456868, + "grad_norm": 1.572606086730957, + "learning_rate": 5.920411793444419e-06, + "loss": 0.4629, + "step": 20443 + }, + { + "epoch": 0.6448194291121274, + "grad_norm": 1.6004555225372314, + "learning_rate": 5.919479133327444e-06, + "loss": 0.4545, + "step": 20444 + }, + { + "epoch": 0.644850969878568, + "grad_norm": 1.6120764017105103, + "learning_rate": 5.918546515793867e-06, + "loss": 0.4144, + "step": 20445 + }, + { + "epoch": 0.6448825106450087, + "grad_norm": 1.322845697402954, + "learning_rate": 5.917613940853433e-06, + "loss": 0.7225, + "step": 20446 + }, + { + "epoch": 0.6449140514114493, + "grad_norm": 1.5415196418762207, + "learning_rate": 5.916681408515865e-06, + "loss": 0.4055, + "step": 20447 + }, + { + "epoch": 0.6449455921778899, + "grad_norm": 1.7153418064117432, + "learning_rate": 5.915748918790896e-06, + "loss": 0.3925, + "step": 20448 + }, + { + "epoch": 0.6449771329443306, + "grad_norm": 1.4535983800888062, + "learning_rate": 5.914816471688258e-06, + "loss": 0.6864, + "step": 20449 + }, + { + "epoch": 0.6450086737107712, + "grad_norm": 1.9701064825057983, + "learning_rate": 5.913884067217686e-06, + "loss": 0.7119, + "step": 20450 + }, + { + "epoch": 0.6450402144772118, + "grad_norm": 1.5441389083862305, + "learning_rate": 5.912951705388903e-06, + "loss": 0.403, + "step": 20451 + }, + { + "epoch": 0.6450717552436525, + "grad_norm": 1.3842873573303223, + "learning_rate": 5.912019386211645e-06, + "loss": 0.729, + "step": 20452 + }, + { + "epoch": 0.6451032960100931, + "grad_norm": 1.2073454856872559, + "learning_rate": 5.911087109695639e-06, + "loss": 0.6666, + "step": 20453 + }, + { + "epoch": 0.6451348367765337, + "grad_norm": 2.360133171081543, + "learning_rate": 5.910154875850609e-06, + "loss": 0.4789, + "step": 20454 + }, + { + "epoch": 0.6451663775429743, + "grad_norm": 1.759793996810913, + "learning_rate": 5.909222684686293e-06, + "loss": 0.4354, + "step": 20455 + }, + { + "epoch": 0.6451979183094149, + "grad_norm": 1.2277326583862305, + "learning_rate": 5.908290536212416e-06, + "loss": 0.7394, + "step": 20456 + }, + { + "epoch": 0.6452294590758555, + "grad_norm": 1.5430494546890259, + "learning_rate": 5.9073584304386984e-06, + "loss": 0.4392, + "step": 20457 + }, + { + "epoch": 0.6452609998422961, + "grad_norm": 1.6782190799713135, + "learning_rate": 5.906426367374879e-06, + "loss": 0.4469, + "step": 20458 + }, + { + "epoch": 0.6452925406087368, + "grad_norm": 1.4483786821365356, + "learning_rate": 5.905494347030678e-06, + "loss": 0.4438, + "step": 20459 + }, + { + "epoch": 0.6453240813751774, + "grad_norm": 1.383931279182434, + "learning_rate": 5.9045623694158215e-06, + "loss": 0.4659, + "step": 20460 + }, + { + "epoch": 0.645355622141618, + "grad_norm": 1.4472638368606567, + "learning_rate": 5.903630434540038e-06, + "loss": 0.7104, + "step": 20461 + }, + { + "epoch": 0.6453871629080586, + "grad_norm": 1.4918749332427979, + "learning_rate": 5.902698542413048e-06, + "loss": 0.4204, + "step": 20462 + }, + { + "epoch": 0.6454187036744993, + "grad_norm": 1.3694506883621216, + "learning_rate": 5.901766693044585e-06, + "loss": 0.4142, + "step": 20463 + }, + { + "epoch": 0.6454502444409399, + "grad_norm": 1.7950477600097656, + "learning_rate": 5.900834886444368e-06, + "loss": 0.7216, + "step": 20464 + }, + { + "epoch": 0.6454817852073805, + "grad_norm": 1.4383964538574219, + "learning_rate": 5.899903122622117e-06, + "loss": 0.6925, + "step": 20465 + }, + { + "epoch": 0.6455133259738212, + "grad_norm": 1.566630482673645, + "learning_rate": 5.898971401587565e-06, + "loss": 0.3781, + "step": 20466 + }, + { + "epoch": 0.6455448667402618, + "grad_norm": 1.5773260593414307, + "learning_rate": 5.898039723350433e-06, + "loss": 0.3381, + "step": 20467 + }, + { + "epoch": 0.6455764075067024, + "grad_norm": 1.4950810670852661, + "learning_rate": 5.897108087920438e-06, + "loss": 0.4565, + "step": 20468 + }, + { + "epoch": 0.645607948273143, + "grad_norm": 2.625452995300293, + "learning_rate": 5.896176495307304e-06, + "loss": 0.4015, + "step": 20469 + }, + { + "epoch": 0.6456394890395837, + "grad_norm": 1.6186128854751587, + "learning_rate": 5.895244945520759e-06, + "loss": 0.4646, + "step": 20470 + }, + { + "epoch": 0.6456710298060243, + "grad_norm": 1.5715309381484985, + "learning_rate": 5.8943134385705185e-06, + "loss": 0.447, + "step": 20471 + }, + { + "epoch": 0.6457025705724649, + "grad_norm": 1.5661579370498657, + "learning_rate": 5.8933819744663054e-06, + "loss": 0.4422, + "step": 20472 + }, + { + "epoch": 0.6457341113389056, + "grad_norm": 1.9175862073898315, + "learning_rate": 5.892450553217842e-06, + "loss": 0.3363, + "step": 20473 + }, + { + "epoch": 0.6457656521053462, + "grad_norm": 1.7484163045883179, + "learning_rate": 5.891519174834841e-06, + "loss": 0.4374, + "step": 20474 + }, + { + "epoch": 0.6457971928717868, + "grad_norm": 1.7155307531356812, + "learning_rate": 5.890587839327033e-06, + "loss": 0.4735, + "step": 20475 + }, + { + "epoch": 0.6458287336382275, + "grad_norm": 1.7626808881759644, + "learning_rate": 5.889656546704131e-06, + "loss": 0.3841, + "step": 20476 + }, + { + "epoch": 0.6458602744046681, + "grad_norm": 2.0917725563049316, + "learning_rate": 5.88872529697585e-06, + "loss": 0.4402, + "step": 20477 + }, + { + "epoch": 0.6458918151711086, + "grad_norm": 1.7388392686843872, + "learning_rate": 5.887794090151918e-06, + "loss": 0.4503, + "step": 20478 + }, + { + "epoch": 0.6459233559375492, + "grad_norm": 1.532273530960083, + "learning_rate": 5.8868629262420455e-06, + "loss": 0.4436, + "step": 20479 + }, + { + "epoch": 0.6459548967039899, + "grad_norm": 1.8565592765808105, + "learning_rate": 5.885931805255953e-06, + "loss": 0.446, + "step": 20480 + }, + { + "epoch": 0.6459864374704305, + "grad_norm": 1.6041021347045898, + "learning_rate": 5.885000727203358e-06, + "loss": 0.4308, + "step": 20481 + }, + { + "epoch": 0.6460179782368711, + "grad_norm": 1.410863995552063, + "learning_rate": 5.884069692093971e-06, + "loss": 0.4428, + "step": 20482 + }, + { + "epoch": 0.6460495190033118, + "grad_norm": 1.3252636194229126, + "learning_rate": 5.883138699937518e-06, + "loss": 0.4682, + "step": 20483 + }, + { + "epoch": 0.6460810597697524, + "grad_norm": 1.9213985204696655, + "learning_rate": 5.882207750743707e-06, + "loss": 0.3923, + "step": 20484 + }, + { + "epoch": 0.646112600536193, + "grad_norm": 2.1498124599456787, + "learning_rate": 5.881276844522253e-06, + "loss": 0.4415, + "step": 20485 + }, + { + "epoch": 0.6461441413026336, + "grad_norm": 1.8215713500976562, + "learning_rate": 5.880345981282877e-06, + "loss": 0.4361, + "step": 20486 + }, + { + "epoch": 0.6461756820690743, + "grad_norm": 1.5533251762390137, + "learning_rate": 5.879415161035289e-06, + "loss": 0.442, + "step": 20487 + }, + { + "epoch": 0.6462072228355149, + "grad_norm": 2.2472007274627686, + "learning_rate": 5.878484383789201e-06, + "loss": 0.4364, + "step": 20488 + }, + { + "epoch": 0.6462387636019555, + "grad_norm": 1.563273310661316, + "learning_rate": 5.877553649554331e-06, + "loss": 0.455, + "step": 20489 + }, + { + "epoch": 0.6462703043683962, + "grad_norm": 2.145956039428711, + "learning_rate": 5.876622958340389e-06, + "loss": 0.6872, + "step": 20490 + }, + { + "epoch": 0.6463018451348368, + "grad_norm": 1.6240384578704834, + "learning_rate": 5.875692310157087e-06, + "loss": 0.3733, + "step": 20491 + }, + { + "epoch": 0.6463333859012774, + "grad_norm": 1.760464072227478, + "learning_rate": 5.874761705014141e-06, + "loss": 0.399, + "step": 20492 + }, + { + "epoch": 0.646364926667718, + "grad_norm": 1.3589564561843872, + "learning_rate": 5.873831142921254e-06, + "loss": 0.4401, + "step": 20493 + }, + { + "epoch": 0.6463964674341587, + "grad_norm": 1.6549469232559204, + "learning_rate": 5.872900623888149e-06, + "loss": 0.441, + "step": 20494 + }, + { + "epoch": 0.6464280082005993, + "grad_norm": 1.668831706047058, + "learning_rate": 5.87197014792453e-06, + "loss": 0.4368, + "step": 20495 + }, + { + "epoch": 0.6464595489670399, + "grad_norm": 2.4979958534240723, + "learning_rate": 5.871039715040103e-06, + "loss": 0.4383, + "step": 20496 + }, + { + "epoch": 0.6464910897334806, + "grad_norm": 2.1901907920837402, + "learning_rate": 5.870109325244587e-06, + "loss": 0.3234, + "step": 20497 + }, + { + "epoch": 0.6465226304999212, + "grad_norm": 1.6561676263809204, + "learning_rate": 5.8691789785476875e-06, + "loss": 0.4202, + "step": 20498 + }, + { + "epoch": 0.6465541712663617, + "grad_norm": 1.8479803800582886, + "learning_rate": 5.868248674959111e-06, + "loss": 0.4084, + "step": 20499 + }, + { + "epoch": 0.6465857120328024, + "grad_norm": 1.898295283317566, + "learning_rate": 5.867318414488568e-06, + "loss": 0.4463, + "step": 20500 + }, + { + "epoch": 0.646617252799243, + "grad_norm": 1.719386100769043, + "learning_rate": 5.866388197145768e-06, + "loss": 0.4046, + "step": 20501 + }, + { + "epoch": 0.6466487935656836, + "grad_norm": 2.428550958633423, + "learning_rate": 5.865458022940412e-06, + "loss": 0.4219, + "step": 20502 + }, + { + "epoch": 0.6466803343321242, + "grad_norm": 5.0669636726379395, + "learning_rate": 5.864527891882217e-06, + "loss": 0.3019, + "step": 20503 + }, + { + "epoch": 0.6467118750985649, + "grad_norm": 1.4539471864700317, + "learning_rate": 5.8635978039808835e-06, + "loss": 0.4411, + "step": 20504 + }, + { + "epoch": 0.6467434158650055, + "grad_norm": 1.70405113697052, + "learning_rate": 5.862667759246114e-06, + "loss": 0.438, + "step": 20505 + }, + { + "epoch": 0.6467749566314461, + "grad_norm": 1.4902632236480713, + "learning_rate": 5.861737757687625e-06, + "loss": 0.4559, + "step": 20506 + }, + { + "epoch": 0.6468064973978868, + "grad_norm": 1.6798573732376099, + "learning_rate": 5.860807799315116e-06, + "loss": 0.4314, + "step": 20507 + }, + { + "epoch": 0.6468380381643274, + "grad_norm": 1.4887304306030273, + "learning_rate": 5.859877884138288e-06, + "loss": 0.4505, + "step": 20508 + }, + { + "epoch": 0.646869578930768, + "grad_norm": 1.9686933755874634, + "learning_rate": 5.858948012166849e-06, + "loss": 0.4458, + "step": 20509 + }, + { + "epoch": 0.6469011196972086, + "grad_norm": 1.4253121614456177, + "learning_rate": 5.8580181834105036e-06, + "loss": 0.7047, + "step": 20510 + }, + { + "epoch": 0.6469326604636493, + "grad_norm": 1.4366977214813232, + "learning_rate": 5.857088397878956e-06, + "loss": 0.7829, + "step": 20511 + }, + { + "epoch": 0.6469642012300899, + "grad_norm": 3.0402936935424805, + "learning_rate": 5.8561586555819095e-06, + "loss": 0.3595, + "step": 20512 + }, + { + "epoch": 0.6469957419965305, + "grad_norm": 2.2571654319763184, + "learning_rate": 5.85522895652906e-06, + "loss": 0.4464, + "step": 20513 + }, + { + "epoch": 0.6470272827629712, + "grad_norm": 1.4117605686187744, + "learning_rate": 5.854299300730117e-06, + "loss": 0.4166, + "step": 20514 + }, + { + "epoch": 0.6470588235294118, + "grad_norm": 2.0448238849639893, + "learning_rate": 5.853369688194782e-06, + "loss": 0.3734, + "step": 20515 + }, + { + "epoch": 0.6470903642958524, + "grad_norm": 1.5692620277404785, + "learning_rate": 5.852440118932749e-06, + "loss": 0.4142, + "step": 20516 + }, + { + "epoch": 0.6471219050622931, + "grad_norm": 1.771067500114441, + "learning_rate": 5.851510592953729e-06, + "loss": 0.4577, + "step": 20517 + }, + { + "epoch": 0.6471534458287337, + "grad_norm": 1.9954838752746582, + "learning_rate": 5.8505811102674174e-06, + "loss": 0.4165, + "step": 20518 + }, + { + "epoch": 0.6471849865951743, + "grad_norm": 1.4463541507720947, + "learning_rate": 5.849651670883509e-06, + "loss": 0.4496, + "step": 20519 + }, + { + "epoch": 0.6472165273616148, + "grad_norm": 1.7455133199691772, + "learning_rate": 5.848722274811713e-06, + "loss": 0.4753, + "step": 20520 + }, + { + "epoch": 0.6472480681280555, + "grad_norm": 1.4978941679000854, + "learning_rate": 5.847792922061722e-06, + "loss": 0.3623, + "step": 20521 + }, + { + "epoch": 0.6472796088944961, + "grad_norm": 1.3862330913543701, + "learning_rate": 5.846863612643232e-06, + "loss": 0.351, + "step": 20522 + }, + { + "epoch": 0.6473111496609367, + "grad_norm": 2.0143094062805176, + "learning_rate": 5.84593434656595e-06, + "loss": 0.7094, + "step": 20523 + }, + { + "epoch": 0.6473426904273774, + "grad_norm": 1.7338205575942993, + "learning_rate": 5.845005123839565e-06, + "loss": 0.4503, + "step": 20524 + }, + { + "epoch": 0.647374231193818, + "grad_norm": 1.375222086906433, + "learning_rate": 5.844075944473781e-06, + "loss": 0.4138, + "step": 20525 + }, + { + "epoch": 0.6474057719602586, + "grad_norm": 1.6712188720703125, + "learning_rate": 5.84314680847829e-06, + "loss": 0.473, + "step": 20526 + }, + { + "epoch": 0.6474373127266992, + "grad_norm": 1.258164405822754, + "learning_rate": 5.84221771586279e-06, + "loss": 0.7245, + "step": 20527 + }, + { + "epoch": 0.6474688534931399, + "grad_norm": 1.4541425704956055, + "learning_rate": 5.841288666636977e-06, + "loss": 0.4329, + "step": 20528 + }, + { + "epoch": 0.6475003942595805, + "grad_norm": 1.769640326499939, + "learning_rate": 5.840359660810546e-06, + "loss": 0.4046, + "step": 20529 + }, + { + "epoch": 0.6475319350260211, + "grad_norm": 6.730409145355225, + "learning_rate": 5.839430698393189e-06, + "loss": 0.4509, + "step": 20530 + }, + { + "epoch": 0.6475634757924618, + "grad_norm": 1.5936027765274048, + "learning_rate": 5.838501779394607e-06, + "loss": 0.461, + "step": 20531 + }, + { + "epoch": 0.6475950165589024, + "grad_norm": 1.728948950767517, + "learning_rate": 5.837572903824489e-06, + "loss": 0.4099, + "step": 20532 + }, + { + "epoch": 0.647626557325343, + "grad_norm": 1.9040486812591553, + "learning_rate": 5.836644071692528e-06, + "loss": 0.45, + "step": 20533 + }, + { + "epoch": 0.6476580980917837, + "grad_norm": 1.544683575630188, + "learning_rate": 5.835715283008424e-06, + "loss": 0.3888, + "step": 20534 + }, + { + "epoch": 0.6476896388582243, + "grad_norm": 1.814658284187317, + "learning_rate": 5.83478653778186e-06, + "loss": 0.3507, + "step": 20535 + }, + { + "epoch": 0.6477211796246649, + "grad_norm": 1.599016547203064, + "learning_rate": 5.833857836022533e-06, + "loss": 0.4125, + "step": 20536 + }, + { + "epoch": 0.6477527203911055, + "grad_norm": 1.4607374668121338, + "learning_rate": 5.832929177740134e-06, + "loss": 0.4289, + "step": 20537 + }, + { + "epoch": 0.6477842611575462, + "grad_norm": 4.953631401062012, + "learning_rate": 5.83200056294436e-06, + "loss": 0.6908, + "step": 20538 + }, + { + "epoch": 0.6478158019239868, + "grad_norm": 2.339928388595581, + "learning_rate": 5.8310719916448865e-06, + "loss": 0.3195, + "step": 20539 + }, + { + "epoch": 0.6478473426904274, + "grad_norm": 1.6173650026321411, + "learning_rate": 5.830143463851423e-06, + "loss": 0.4656, + "step": 20540 + }, + { + "epoch": 0.647878883456868, + "grad_norm": 1.7448087930679321, + "learning_rate": 5.8292149795736465e-06, + "loss": 0.4604, + "step": 20541 + }, + { + "epoch": 0.6479104242233086, + "grad_norm": 2.4165682792663574, + "learning_rate": 5.828286538821249e-06, + "loss": 0.3828, + "step": 20542 + }, + { + "epoch": 0.6479419649897492, + "grad_norm": 1.5096567869186401, + "learning_rate": 5.827358141603923e-06, + "loss": 0.4321, + "step": 20543 + }, + { + "epoch": 0.6479735057561898, + "grad_norm": 1.7347075939178467, + "learning_rate": 5.826429787931351e-06, + "loss": 0.4065, + "step": 20544 + }, + { + "epoch": 0.6480050465226305, + "grad_norm": 1.508297085762024, + "learning_rate": 5.825501477813233e-06, + "loss": 0.4548, + "step": 20545 + }, + { + "epoch": 0.6480365872890711, + "grad_norm": 1.6616885662078857, + "learning_rate": 5.824573211259242e-06, + "loss": 0.4522, + "step": 20546 + }, + { + "epoch": 0.6480681280555117, + "grad_norm": 1.8407522439956665, + "learning_rate": 5.823644988279072e-06, + "loss": 0.4159, + "step": 20547 + }, + { + "epoch": 0.6480996688219524, + "grad_norm": 1.5136744976043701, + "learning_rate": 5.8227168088824094e-06, + "loss": 0.4622, + "step": 20548 + }, + { + "epoch": 0.648131209588393, + "grad_norm": 1.294413447380066, + "learning_rate": 5.821788673078945e-06, + "loss": 0.7025, + "step": 20549 + }, + { + "epoch": 0.6481627503548336, + "grad_norm": 1.5218544006347656, + "learning_rate": 5.820860580878351e-06, + "loss": 0.4262, + "step": 20550 + }, + { + "epoch": 0.6481942911212742, + "grad_norm": 1.8015375137329102, + "learning_rate": 5.819932532290331e-06, + "loss": 0.4222, + "step": 20551 + }, + { + "epoch": 0.6482258318877149, + "grad_norm": 1.477955937385559, + "learning_rate": 5.8190045273245565e-06, + "loss": 0.4688, + "step": 20552 + }, + { + "epoch": 0.6482573726541555, + "grad_norm": 1.5417568683624268, + "learning_rate": 5.818076565990717e-06, + "loss": 0.4266, + "step": 20553 + }, + { + "epoch": 0.6482889134205961, + "grad_norm": 1.3552759885787964, + "learning_rate": 5.817148648298499e-06, + "loss": 0.4196, + "step": 20554 + }, + { + "epoch": 0.6483204541870368, + "grad_norm": 3.0803635120391846, + "learning_rate": 5.816220774257578e-06, + "loss": 0.4155, + "step": 20555 + }, + { + "epoch": 0.6483519949534774, + "grad_norm": 1.6274789571762085, + "learning_rate": 5.8152929438776426e-06, + "loss": 0.4245, + "step": 20556 + }, + { + "epoch": 0.648383535719918, + "grad_norm": 2.4145829677581787, + "learning_rate": 5.814365157168376e-06, + "loss": 0.4649, + "step": 20557 + }, + { + "epoch": 0.6484150764863587, + "grad_norm": 1.7928848266601562, + "learning_rate": 5.813437414139457e-06, + "loss": 0.4102, + "step": 20558 + }, + { + "epoch": 0.6484466172527993, + "grad_norm": 1.417140245437622, + "learning_rate": 5.812509714800569e-06, + "loss": 0.7257, + "step": 20559 + }, + { + "epoch": 0.6484781580192399, + "grad_norm": 1.2745332717895508, + "learning_rate": 5.811582059161399e-06, + "loss": 0.3439, + "step": 20560 + }, + { + "epoch": 0.6485096987856805, + "grad_norm": 1.555140733718872, + "learning_rate": 5.810654447231618e-06, + "loss": 0.419, + "step": 20561 + }, + { + "epoch": 0.6485412395521211, + "grad_norm": 1.5996509790420532, + "learning_rate": 5.809726879020912e-06, + "loss": 0.4285, + "step": 20562 + }, + { + "epoch": 0.6485727803185617, + "grad_norm": 1.7981362342834473, + "learning_rate": 5.808799354538958e-06, + "loss": 0.4058, + "step": 20563 + }, + { + "epoch": 0.6486043210850023, + "grad_norm": 1.606868028640747, + "learning_rate": 5.807871873795437e-06, + "loss": 0.4237, + "step": 20564 + }, + { + "epoch": 0.648635861851443, + "grad_norm": 1.5897254943847656, + "learning_rate": 5.806944436800033e-06, + "loss": 0.4349, + "step": 20565 + }, + { + "epoch": 0.6486674026178836, + "grad_norm": 1.6538453102111816, + "learning_rate": 5.8060170435624155e-06, + "loss": 0.4523, + "step": 20566 + }, + { + "epoch": 0.6486989433843242, + "grad_norm": 1.6199290752410889, + "learning_rate": 5.805089694092266e-06, + "loss": 0.4462, + "step": 20567 + }, + { + "epoch": 0.6487304841507648, + "grad_norm": 1.8797459602355957, + "learning_rate": 5.804162388399264e-06, + "loss": 0.4191, + "step": 20568 + }, + { + "epoch": 0.6487620249172055, + "grad_norm": 1.5382190942764282, + "learning_rate": 5.803235126493089e-06, + "loss": 0.4319, + "step": 20569 + }, + { + "epoch": 0.6487935656836461, + "grad_norm": 1.5600495338439941, + "learning_rate": 5.802307908383406e-06, + "loss": 0.4383, + "step": 20570 + }, + { + "epoch": 0.6488251064500867, + "grad_norm": 1.8658746480941772, + "learning_rate": 5.801380734079906e-06, + "loss": 0.4439, + "step": 20571 + }, + { + "epoch": 0.6488566472165274, + "grad_norm": 1.4150025844573975, + "learning_rate": 5.800453603592257e-06, + "loss": 0.4018, + "step": 20572 + }, + { + "epoch": 0.648888187982968, + "grad_norm": 1.3723856210708618, + "learning_rate": 5.799526516930134e-06, + "loss": 0.4106, + "step": 20573 + }, + { + "epoch": 0.6489197287494086, + "grad_norm": 1.3686350584030151, + "learning_rate": 5.798599474103217e-06, + "loss": 0.366, + "step": 20574 + }, + { + "epoch": 0.6489512695158492, + "grad_norm": 1.4147279262542725, + "learning_rate": 5.7976724751211695e-06, + "loss": 0.4061, + "step": 20575 + }, + { + "epoch": 0.6489828102822899, + "grad_norm": 3.8637747764587402, + "learning_rate": 5.796745519993681e-06, + "loss": 0.4046, + "step": 20576 + }, + { + "epoch": 0.6490143510487305, + "grad_norm": 1.8846992254257202, + "learning_rate": 5.795818608730411e-06, + "loss": 0.379, + "step": 20577 + }, + { + "epoch": 0.6490458918151711, + "grad_norm": 2.1295132637023926, + "learning_rate": 5.79489174134104e-06, + "loss": 0.46, + "step": 20578 + }, + { + "epoch": 0.6490774325816118, + "grad_norm": 1.5914732217788696, + "learning_rate": 5.7939649178352385e-06, + "loss": 0.4748, + "step": 20579 + }, + { + "epoch": 0.6491089733480524, + "grad_norm": 1.4952685832977295, + "learning_rate": 5.793038138222683e-06, + "loss": 0.4505, + "step": 20580 + }, + { + "epoch": 0.649140514114493, + "grad_norm": 1.522802710533142, + "learning_rate": 5.7921114025130365e-06, + "loss": 0.4467, + "step": 20581 + }, + { + "epoch": 0.6491720548809337, + "grad_norm": 1.4305323362350464, + "learning_rate": 5.791184710715976e-06, + "loss": 0.4037, + "step": 20582 + }, + { + "epoch": 0.6492035956473742, + "grad_norm": 1.719527006149292, + "learning_rate": 5.79025806284117e-06, + "loss": 0.3692, + "step": 20583 + }, + { + "epoch": 0.6492351364138148, + "grad_norm": 1.05856192111969, + "learning_rate": 5.7893314588982905e-06, + "loss": 0.3139, + "step": 20584 + }, + { + "epoch": 0.6492666771802554, + "grad_norm": 1.7281217575073242, + "learning_rate": 5.7884048988970105e-06, + "loss": 0.3601, + "step": 20585 + }, + { + "epoch": 0.6492982179466961, + "grad_norm": 1.1804535388946533, + "learning_rate": 5.787478382846991e-06, + "loss": 0.3586, + "step": 20586 + }, + { + "epoch": 0.6493297587131367, + "grad_norm": 1.5181255340576172, + "learning_rate": 5.786551910757907e-06, + "loss": 0.4464, + "step": 20587 + }, + { + "epoch": 0.6493612994795773, + "grad_norm": 2.0124197006225586, + "learning_rate": 5.785625482639425e-06, + "loss": 0.3673, + "step": 20588 + }, + { + "epoch": 0.649392840246018, + "grad_norm": 1.5235780477523804, + "learning_rate": 5.784699098501212e-06, + "loss": 0.4552, + "step": 20589 + }, + { + "epoch": 0.6494243810124586, + "grad_norm": 1.1795951128005981, + "learning_rate": 5.783772758352939e-06, + "loss": 0.3428, + "step": 20590 + }, + { + "epoch": 0.6494559217788992, + "grad_norm": 1.6898988485336304, + "learning_rate": 5.782846462204275e-06, + "loss": 0.4275, + "step": 20591 + }, + { + "epoch": 0.6494874625453398, + "grad_norm": 1.4952958822250366, + "learning_rate": 5.7819202100648775e-06, + "loss": 0.442, + "step": 20592 + }, + { + "epoch": 0.6495190033117805, + "grad_norm": 2.454753875732422, + "learning_rate": 5.780994001944418e-06, + "loss": 0.4318, + "step": 20593 + }, + { + "epoch": 0.6495505440782211, + "grad_norm": 1.54941725730896, + "learning_rate": 5.780067837852566e-06, + "loss": 0.4602, + "step": 20594 + }, + { + "epoch": 0.6495820848446617, + "grad_norm": 5.519217014312744, + "learning_rate": 5.779141717798975e-06, + "loss": 0.3502, + "step": 20595 + }, + { + "epoch": 0.6496136256111024, + "grad_norm": 2.2241172790527344, + "learning_rate": 5.7782156417933254e-06, + "loss": 0.4669, + "step": 20596 + }, + { + "epoch": 0.649645166377543, + "grad_norm": 1.7525066137313843, + "learning_rate": 5.777289609845272e-06, + "loss": 0.4406, + "step": 20597 + }, + { + "epoch": 0.6496767071439836, + "grad_norm": 1.8163386583328247, + "learning_rate": 5.776363621964477e-06, + "loss": 0.3888, + "step": 20598 + }, + { + "epoch": 0.6497082479104243, + "grad_norm": 1.2969456911087036, + "learning_rate": 5.775437678160609e-06, + "loss": 0.3626, + "step": 20599 + }, + { + "epoch": 0.6497397886768649, + "grad_norm": 1.5078152418136597, + "learning_rate": 5.7745117784433325e-06, + "loss": 0.7448, + "step": 20600 + }, + { + "epoch": 0.6497713294433055, + "grad_norm": 1.7688535451889038, + "learning_rate": 5.773585922822303e-06, + "loss": 0.4305, + "step": 20601 + }, + { + "epoch": 0.6498028702097461, + "grad_norm": 1.864936351776123, + "learning_rate": 5.7726601113071854e-06, + "loss": 0.4194, + "step": 20602 + }, + { + "epoch": 0.6498344109761868, + "grad_norm": 1.4138476848602295, + "learning_rate": 5.771734343907642e-06, + "loss": 0.7432, + "step": 20603 + }, + { + "epoch": 0.6498659517426274, + "grad_norm": 1.6411845684051514, + "learning_rate": 5.770808620633334e-06, + "loss": 0.457, + "step": 20604 + }, + { + "epoch": 0.6498974925090679, + "grad_norm": 1.5779743194580078, + "learning_rate": 5.769882941493927e-06, + "loss": 0.4509, + "step": 20605 + }, + { + "epoch": 0.6499290332755085, + "grad_norm": 1.7504510879516602, + "learning_rate": 5.768957306499068e-06, + "loss": 0.4373, + "step": 20606 + }, + { + "epoch": 0.6499605740419492, + "grad_norm": 1.6642067432403564, + "learning_rate": 5.768031715658432e-06, + "loss": 0.4544, + "step": 20607 + }, + { + "epoch": 0.6499921148083898, + "grad_norm": 1.539601445198059, + "learning_rate": 5.767106168981668e-06, + "loss": 0.3928, + "step": 20608 + }, + { + "epoch": 0.6500236555748304, + "grad_norm": 2.9796414375305176, + "learning_rate": 5.766180666478437e-06, + "loss": 0.3721, + "step": 20609 + }, + { + "epoch": 0.6500551963412711, + "grad_norm": 1.386991262435913, + "learning_rate": 5.7652552081583975e-06, + "loss": 0.419, + "step": 20610 + }, + { + "epoch": 0.6500867371077117, + "grad_norm": 1.6700044870376587, + "learning_rate": 5.7643297940312135e-06, + "loss": 0.4482, + "step": 20611 + }, + { + "epoch": 0.6501182778741523, + "grad_norm": 1.6691672801971436, + "learning_rate": 5.7634044241065336e-06, + "loss": 0.4273, + "step": 20612 + }, + { + "epoch": 0.650149818640593, + "grad_norm": 1.5720540285110474, + "learning_rate": 5.762479098394016e-06, + "loss": 0.4549, + "step": 20613 + }, + { + "epoch": 0.6501813594070336, + "grad_norm": 1.3722801208496094, + "learning_rate": 5.7615538169033244e-06, + "loss": 0.4321, + "step": 20614 + }, + { + "epoch": 0.6502129001734742, + "grad_norm": 1.5627418756484985, + "learning_rate": 5.760628579644103e-06, + "loss": 0.4448, + "step": 20615 + }, + { + "epoch": 0.6502444409399148, + "grad_norm": 1.6170059442520142, + "learning_rate": 5.759703386626021e-06, + "loss": 0.4433, + "step": 20616 + }, + { + "epoch": 0.6502759817063555, + "grad_norm": 1.1964797973632812, + "learning_rate": 5.758778237858722e-06, + "loss": 0.3536, + "step": 20617 + }, + { + "epoch": 0.6503075224727961, + "grad_norm": 1.3559284210205078, + "learning_rate": 5.7578531333518665e-06, + "loss": 0.4451, + "step": 20618 + }, + { + "epoch": 0.6503390632392367, + "grad_norm": 1.3301283121109009, + "learning_rate": 5.756928073115107e-06, + "loss": 0.324, + "step": 20619 + }, + { + "epoch": 0.6503706040056774, + "grad_norm": 1.7615832090377808, + "learning_rate": 5.756003057158097e-06, + "loss": 0.417, + "step": 20620 + }, + { + "epoch": 0.650402144772118, + "grad_norm": 2.6023740768432617, + "learning_rate": 5.755078085490495e-06, + "loss": 0.4478, + "step": 20621 + }, + { + "epoch": 0.6504336855385586, + "grad_norm": 1.4234247207641602, + "learning_rate": 5.754153158121947e-06, + "loss": 0.4564, + "step": 20622 + }, + { + "epoch": 0.6504652263049993, + "grad_norm": 1.300412654876709, + "learning_rate": 5.753228275062105e-06, + "loss": 0.4178, + "step": 20623 + }, + { + "epoch": 0.6504967670714399, + "grad_norm": 1.6381218433380127, + "learning_rate": 5.752303436320625e-06, + "loss": 0.4183, + "step": 20624 + }, + { + "epoch": 0.6505283078378805, + "grad_norm": 1.5061074495315552, + "learning_rate": 5.751378641907159e-06, + "loss": 0.4277, + "step": 20625 + }, + { + "epoch": 0.650559848604321, + "grad_norm": 2.0150272846221924, + "learning_rate": 5.75045389183135e-06, + "loss": 0.4508, + "step": 20626 + }, + { + "epoch": 0.6505913893707617, + "grad_norm": 1.9830875396728516, + "learning_rate": 5.749529186102862e-06, + "loss": 0.3919, + "step": 20627 + }, + { + "epoch": 0.6506229301372023, + "grad_norm": 2.0797441005706787, + "learning_rate": 5.748604524731333e-06, + "loss": 0.2992, + "step": 20628 + }, + { + "epoch": 0.6506544709036429, + "grad_norm": 1.3344160318374634, + "learning_rate": 5.747679907726416e-06, + "loss": 0.7016, + "step": 20629 + }, + { + "epoch": 0.6506860116700836, + "grad_norm": 1.531677007675171, + "learning_rate": 5.746755335097762e-06, + "loss": 0.4439, + "step": 20630 + }, + { + "epoch": 0.6507175524365242, + "grad_norm": 1.7677406072616577, + "learning_rate": 5.745830806855024e-06, + "loss": 0.4382, + "step": 20631 + }, + { + "epoch": 0.6507490932029648, + "grad_norm": 1.539704442024231, + "learning_rate": 5.7449063230078384e-06, + "loss": 0.4186, + "step": 20632 + }, + { + "epoch": 0.6507806339694054, + "grad_norm": 1.7547500133514404, + "learning_rate": 5.743981883565861e-06, + "loss": 0.406, + "step": 20633 + }, + { + "epoch": 0.6508121747358461, + "grad_norm": 1.5155748128890991, + "learning_rate": 5.74305748853874e-06, + "loss": 0.4445, + "step": 20634 + }, + { + "epoch": 0.6508437155022867, + "grad_norm": 1.4475680589675903, + "learning_rate": 5.742133137936112e-06, + "loss": 0.3942, + "step": 20635 + }, + { + "epoch": 0.6508752562687273, + "grad_norm": 2.294201612472534, + "learning_rate": 5.741208831767639e-06, + "loss": 0.736, + "step": 20636 + }, + { + "epoch": 0.650906797035168, + "grad_norm": 1.72547447681427, + "learning_rate": 5.740284570042951e-06, + "loss": 0.3826, + "step": 20637 + }, + { + "epoch": 0.6509383378016086, + "grad_norm": 1.5361149311065674, + "learning_rate": 5.739360352771708e-06, + "loss": 0.4449, + "step": 20638 + }, + { + "epoch": 0.6509698785680492, + "grad_norm": 1.5294467210769653, + "learning_rate": 5.738436179963545e-06, + "loss": 0.7017, + "step": 20639 + }, + { + "epoch": 0.6510014193344899, + "grad_norm": 1.4645508527755737, + "learning_rate": 5.737512051628109e-06, + "loss": 0.6754, + "step": 20640 + }, + { + "epoch": 0.6510329601009305, + "grad_norm": 1.717790961265564, + "learning_rate": 5.736587967775049e-06, + "loss": 0.3954, + "step": 20641 + }, + { + "epoch": 0.6510645008673711, + "grad_norm": 1.7356507778167725, + "learning_rate": 5.7356639284140005e-06, + "loss": 0.4234, + "step": 20642 + }, + { + "epoch": 0.6510960416338117, + "grad_norm": 1.7555198669433594, + "learning_rate": 5.7347399335546095e-06, + "loss": 0.3112, + "step": 20643 + }, + { + "epoch": 0.6511275824002524, + "grad_norm": 1.4436591863632202, + "learning_rate": 5.733815983206519e-06, + "loss": 0.3637, + "step": 20644 + }, + { + "epoch": 0.651159123166693, + "grad_norm": 1.6282135248184204, + "learning_rate": 5.732892077379375e-06, + "loss": 0.4134, + "step": 20645 + }, + { + "epoch": 0.6511906639331336, + "grad_norm": 1.4249403476715088, + "learning_rate": 5.731968216082808e-06, + "loss": 0.4472, + "step": 20646 + }, + { + "epoch": 0.6512222046995741, + "grad_norm": 1.7501026391983032, + "learning_rate": 5.731044399326475e-06, + "loss": 0.4134, + "step": 20647 + }, + { + "epoch": 0.6512537454660148, + "grad_norm": 2.0445213317871094, + "learning_rate": 5.730120627120004e-06, + "loss": 0.4274, + "step": 20648 + }, + { + "epoch": 0.6512852862324554, + "grad_norm": 2.1235764026641846, + "learning_rate": 5.72919689947304e-06, + "loss": 0.4165, + "step": 20649 + }, + { + "epoch": 0.651316826998896, + "grad_norm": 1.844155192375183, + "learning_rate": 5.7282732163952214e-06, + "loss": 0.4238, + "step": 20650 + }, + { + "epoch": 0.6513483677653367, + "grad_norm": 1.4679865837097168, + "learning_rate": 5.727349577896194e-06, + "loss": 0.3999, + "step": 20651 + }, + { + "epoch": 0.6513799085317773, + "grad_norm": 1.489588737487793, + "learning_rate": 5.726425983985586e-06, + "loss": 0.7576, + "step": 20652 + }, + { + "epoch": 0.6514114492982179, + "grad_norm": 1.672095775604248, + "learning_rate": 5.725502434673043e-06, + "loss": 0.4582, + "step": 20653 + }, + { + "epoch": 0.6514429900646586, + "grad_norm": 1.6483088731765747, + "learning_rate": 5.7245789299682005e-06, + "loss": 0.4491, + "step": 20654 + }, + { + "epoch": 0.6514745308310992, + "grad_norm": 1.5969210863113403, + "learning_rate": 5.723655469880696e-06, + "loss": 0.4077, + "step": 20655 + }, + { + "epoch": 0.6515060715975398, + "grad_norm": 1.9728219509124756, + "learning_rate": 5.722732054420172e-06, + "loss": 0.3773, + "step": 20656 + }, + { + "epoch": 0.6515376123639804, + "grad_norm": 1.9390385150909424, + "learning_rate": 5.721808683596251e-06, + "loss": 0.406, + "step": 20657 + }, + { + "epoch": 0.6515691531304211, + "grad_norm": 1.3642609119415283, + "learning_rate": 5.720885357418588e-06, + "loss": 0.4104, + "step": 20658 + }, + { + "epoch": 0.6516006938968617, + "grad_norm": 1.9679986238479614, + "learning_rate": 5.719962075896803e-06, + "loss": 0.4264, + "step": 20659 + }, + { + "epoch": 0.6516322346633023, + "grad_norm": 1.5544315576553345, + "learning_rate": 5.71903883904054e-06, + "loss": 0.4193, + "step": 20660 + }, + { + "epoch": 0.651663775429743, + "grad_norm": 2.3703598976135254, + "learning_rate": 5.718115646859432e-06, + "loss": 0.7225, + "step": 20661 + }, + { + "epoch": 0.6516953161961836, + "grad_norm": 1.3699278831481934, + "learning_rate": 5.717192499363109e-06, + "loss": 0.4442, + "step": 20662 + }, + { + "epoch": 0.6517268569626242, + "grad_norm": 1.3799548149108887, + "learning_rate": 5.716269396561208e-06, + "loss": 0.3587, + "step": 20663 + }, + { + "epoch": 0.6517583977290649, + "grad_norm": 1.3612250089645386, + "learning_rate": 5.715346338463362e-06, + "loss": 0.4471, + "step": 20664 + }, + { + "epoch": 0.6517899384955055, + "grad_norm": 4.00166130065918, + "learning_rate": 5.714423325079208e-06, + "loss": 0.422, + "step": 20665 + }, + { + "epoch": 0.6518214792619461, + "grad_norm": 1.372384786605835, + "learning_rate": 5.713500356418367e-06, + "loss": 0.7185, + "step": 20666 + }, + { + "epoch": 0.6518530200283867, + "grad_norm": 1.5197266340255737, + "learning_rate": 5.712577432490485e-06, + "loss": 0.4137, + "step": 20667 + }, + { + "epoch": 0.6518845607948273, + "grad_norm": 1.5728365182876587, + "learning_rate": 5.711654553305178e-06, + "loss": 0.4249, + "step": 20668 + }, + { + "epoch": 0.6519161015612679, + "grad_norm": 1.7438558340072632, + "learning_rate": 5.710731718872094e-06, + "loss": 0.437, + "step": 20669 + }, + { + "epoch": 0.6519476423277085, + "grad_norm": 1.5145114660263062, + "learning_rate": 5.7098089292008515e-06, + "loss": 0.6959, + "step": 20670 + }, + { + "epoch": 0.6519791830941492, + "grad_norm": 1.6799321174621582, + "learning_rate": 5.708886184301084e-06, + "loss": 0.4362, + "step": 20671 + }, + { + "epoch": 0.6520107238605898, + "grad_norm": 1.6607975959777832, + "learning_rate": 5.707963484182426e-06, + "loss": 0.3837, + "step": 20672 + }, + { + "epoch": 0.6520422646270304, + "grad_norm": 1.382926344871521, + "learning_rate": 5.707040828854496e-06, + "loss": 0.4171, + "step": 20673 + }, + { + "epoch": 0.652073805393471, + "grad_norm": 1.852354645729065, + "learning_rate": 5.70611821832693e-06, + "loss": 0.3893, + "step": 20674 + }, + { + "epoch": 0.6521053461599117, + "grad_norm": 1.3721332550048828, + "learning_rate": 5.705195652609353e-06, + "loss": 0.4352, + "step": 20675 + }, + { + "epoch": 0.6521368869263523, + "grad_norm": 1.5008538961410522, + "learning_rate": 5.7042731317114e-06, + "loss": 0.4294, + "step": 20676 + }, + { + "epoch": 0.6521684276927929, + "grad_norm": 1.4378858804702759, + "learning_rate": 5.7033506556426834e-06, + "loss": 0.4357, + "step": 20677 + }, + { + "epoch": 0.6521999684592336, + "grad_norm": 1.9497721195220947, + "learning_rate": 5.702428224412846e-06, + "loss": 0.4496, + "step": 20678 + }, + { + "epoch": 0.6522315092256742, + "grad_norm": 1.5161569118499756, + "learning_rate": 5.701505838031504e-06, + "loss": 0.4506, + "step": 20679 + }, + { + "epoch": 0.6522630499921148, + "grad_norm": 1.4707101583480835, + "learning_rate": 5.700583496508286e-06, + "loss": 0.7517, + "step": 20680 + }, + { + "epoch": 0.6522945907585554, + "grad_norm": 1.5524612665176392, + "learning_rate": 5.699661199852822e-06, + "loss": 0.4284, + "step": 20681 + }, + { + "epoch": 0.6523261315249961, + "grad_norm": 1.4918632507324219, + "learning_rate": 5.698738948074729e-06, + "loss": 0.6559, + "step": 20682 + }, + { + "epoch": 0.6523576722914367, + "grad_norm": 1.3924986124038696, + "learning_rate": 5.697816741183634e-06, + "loss": 0.425, + "step": 20683 + }, + { + "epoch": 0.6523892130578773, + "grad_norm": 1.7153128385543823, + "learning_rate": 5.6968945791891615e-06, + "loss": 0.3815, + "step": 20684 + }, + { + "epoch": 0.652420753824318, + "grad_norm": 1.384300947189331, + "learning_rate": 5.695972462100936e-06, + "loss": 0.407, + "step": 20685 + }, + { + "epoch": 0.6524522945907586, + "grad_norm": 1.381635069847107, + "learning_rate": 5.6950503899285804e-06, + "loss": 0.4128, + "step": 20686 + }, + { + "epoch": 0.6524838353571992, + "grad_norm": 1.7559492588043213, + "learning_rate": 5.6941283626817205e-06, + "loss": 0.4102, + "step": 20687 + }, + { + "epoch": 0.6525153761236399, + "grad_norm": 1.5937446355819702, + "learning_rate": 5.693206380369965e-06, + "loss": 0.4449, + "step": 20688 + }, + { + "epoch": 0.6525469168900804, + "grad_norm": 1.5344393253326416, + "learning_rate": 5.692284443002955e-06, + "loss": 0.4564, + "step": 20689 + }, + { + "epoch": 0.652578457656521, + "grad_norm": 1.3692502975463867, + "learning_rate": 5.6913625505902966e-06, + "loss": 0.3565, + "step": 20690 + }, + { + "epoch": 0.6526099984229616, + "grad_norm": 1.7493361234664917, + "learning_rate": 5.690440703141615e-06, + "loss": 0.4349, + "step": 20691 + }, + { + "epoch": 0.6526415391894023, + "grad_norm": 1.6072359085083008, + "learning_rate": 5.689518900666538e-06, + "loss": 0.4501, + "step": 20692 + }, + { + "epoch": 0.6526730799558429, + "grad_norm": 1.5995639562606812, + "learning_rate": 5.688597143174671e-06, + "loss": 0.7088, + "step": 20693 + }, + { + "epoch": 0.6527046207222835, + "grad_norm": 1.4672584533691406, + "learning_rate": 5.687675430675641e-06, + "loss": 0.4039, + "step": 20694 + }, + { + "epoch": 0.6527361614887242, + "grad_norm": 1.8302175998687744, + "learning_rate": 5.6867537631790674e-06, + "loss": 0.3766, + "step": 20695 + }, + { + "epoch": 0.6527677022551648, + "grad_norm": 1.5375090837478638, + "learning_rate": 5.685832140694571e-06, + "loss": 0.4332, + "step": 20696 + }, + { + "epoch": 0.6527992430216054, + "grad_norm": 1.4312257766723633, + "learning_rate": 5.684910563231758e-06, + "loss": 0.4784, + "step": 20697 + }, + { + "epoch": 0.652830783788046, + "grad_norm": 1.682956337928772, + "learning_rate": 5.683989030800262e-06, + "loss": 0.4379, + "step": 20698 + }, + { + "epoch": 0.6528623245544867, + "grad_norm": 1.5638861656188965, + "learning_rate": 5.683067543409688e-06, + "loss": 0.4306, + "step": 20699 + }, + { + "epoch": 0.6528938653209273, + "grad_norm": 1.7049516439437866, + "learning_rate": 5.682146101069655e-06, + "loss": 0.4417, + "step": 20700 + }, + { + "epoch": 0.6529254060873679, + "grad_norm": 1.762824535369873, + "learning_rate": 5.681224703789785e-06, + "loss": 0.4236, + "step": 20701 + }, + { + "epoch": 0.6529569468538086, + "grad_norm": 1.8059097528457642, + "learning_rate": 5.680303351579681e-06, + "loss": 0.4556, + "step": 20702 + }, + { + "epoch": 0.6529884876202492, + "grad_norm": 1.5586535930633545, + "learning_rate": 5.679382044448973e-06, + "loss": 0.4118, + "step": 20703 + }, + { + "epoch": 0.6530200283866898, + "grad_norm": 1.7754223346710205, + "learning_rate": 5.678460782407265e-06, + "loss": 0.3828, + "step": 20704 + }, + { + "epoch": 0.6530515691531305, + "grad_norm": 1.6291356086730957, + "learning_rate": 5.677539565464174e-06, + "loss": 0.4295, + "step": 20705 + }, + { + "epoch": 0.6530831099195711, + "grad_norm": 1.5674959421157837, + "learning_rate": 5.676618393629314e-06, + "loss": 0.4386, + "step": 20706 + }, + { + "epoch": 0.6531146506860117, + "grad_norm": 2.643975019454956, + "learning_rate": 5.6756972669123025e-06, + "loss": 0.4751, + "step": 20707 + }, + { + "epoch": 0.6531461914524523, + "grad_norm": 1.5646995306015015, + "learning_rate": 5.674776185322739e-06, + "loss": 0.3961, + "step": 20708 + }, + { + "epoch": 0.653177732218893, + "grad_norm": 3.014199733734131, + "learning_rate": 5.673855148870253e-06, + "loss": 0.7068, + "step": 20709 + }, + { + "epoch": 0.6532092729853335, + "grad_norm": 1.985813856124878, + "learning_rate": 5.672934157564443e-06, + "loss": 0.711, + "step": 20710 + }, + { + "epoch": 0.6532408137517741, + "grad_norm": 1.646934986114502, + "learning_rate": 5.672013211414925e-06, + "loss": 0.4518, + "step": 20711 + }, + { + "epoch": 0.6532723545182147, + "grad_norm": 1.7621867656707764, + "learning_rate": 5.671092310431314e-06, + "loss": 0.3586, + "step": 20712 + }, + { + "epoch": 0.6533038952846554, + "grad_norm": 2.0590660572052, + "learning_rate": 5.6701714546232125e-06, + "loss": 0.4479, + "step": 20713 + }, + { + "epoch": 0.653335436051096, + "grad_norm": 1.6001754999160767, + "learning_rate": 5.669250644000232e-06, + "loss": 0.4476, + "step": 20714 + }, + { + "epoch": 0.6533669768175366, + "grad_norm": 1.7275570631027222, + "learning_rate": 5.668329878571987e-06, + "loss": 0.4102, + "step": 20715 + }, + { + "epoch": 0.6533985175839773, + "grad_norm": 1.501460313796997, + "learning_rate": 5.66740915834808e-06, + "loss": 0.7398, + "step": 20716 + }, + { + "epoch": 0.6534300583504179, + "grad_norm": 3.574322462081909, + "learning_rate": 5.666488483338125e-06, + "loss": 0.454, + "step": 20717 + }, + { + "epoch": 0.6534615991168585, + "grad_norm": 1.8871643543243408, + "learning_rate": 5.66556785355173e-06, + "loss": 0.4603, + "step": 20718 + }, + { + "epoch": 0.6534931398832992, + "grad_norm": 1.868778109550476, + "learning_rate": 5.6646472689984976e-06, + "loss": 0.3999, + "step": 20719 + }, + { + "epoch": 0.6535246806497398, + "grad_norm": 1.3968899250030518, + "learning_rate": 5.663726729688035e-06, + "loss": 0.6926, + "step": 20720 + }, + { + "epoch": 0.6535562214161804, + "grad_norm": 1.6370381116867065, + "learning_rate": 5.662806235629956e-06, + "loss": 0.4314, + "step": 20721 + }, + { + "epoch": 0.653587762182621, + "grad_norm": 1.4918423891067505, + "learning_rate": 5.661885786833855e-06, + "loss": 0.7096, + "step": 20722 + }, + { + "epoch": 0.6536193029490617, + "grad_norm": 1.823886513710022, + "learning_rate": 5.660965383309352e-06, + "loss": 0.3981, + "step": 20723 + }, + { + "epoch": 0.6536508437155023, + "grad_norm": 1.5044282674789429, + "learning_rate": 5.660045025066039e-06, + "loss": 0.436, + "step": 20724 + }, + { + "epoch": 0.6536823844819429, + "grad_norm": 2.4583756923675537, + "learning_rate": 5.6591247121135266e-06, + "loss": 0.4275, + "step": 20725 + }, + { + "epoch": 0.6537139252483836, + "grad_norm": 2.4621222019195557, + "learning_rate": 5.658204444461419e-06, + "loss": 0.4235, + "step": 20726 + }, + { + "epoch": 0.6537454660148242, + "grad_norm": 2.4860644340515137, + "learning_rate": 5.657284222119322e-06, + "loss": 0.4593, + "step": 20727 + }, + { + "epoch": 0.6537770067812648, + "grad_norm": 1.7794859409332275, + "learning_rate": 5.656364045096829e-06, + "loss": 0.3513, + "step": 20728 + }, + { + "epoch": 0.6538085475477055, + "grad_norm": 2.4615166187286377, + "learning_rate": 5.6554439134035575e-06, + "loss": 0.436, + "step": 20729 + }, + { + "epoch": 0.6538400883141461, + "grad_norm": 1.0830265283584595, + "learning_rate": 5.6545238270491e-06, + "loss": 0.3214, + "step": 20730 + }, + { + "epoch": 0.6538716290805867, + "grad_norm": 1.7277679443359375, + "learning_rate": 5.6536037860430584e-06, + "loss": 0.4398, + "step": 20731 + }, + { + "epoch": 0.6539031698470272, + "grad_norm": 1.3716450929641724, + "learning_rate": 5.652683790395042e-06, + "loss": 0.4294, + "step": 20732 + }, + { + "epoch": 0.6539347106134679, + "grad_norm": 1.806701421737671, + "learning_rate": 5.6517638401146365e-06, + "loss": 0.4325, + "step": 20733 + }, + { + "epoch": 0.6539662513799085, + "grad_norm": 1.5560686588287354, + "learning_rate": 5.65084393521146e-06, + "loss": 0.445, + "step": 20734 + }, + { + "epoch": 0.6539977921463491, + "grad_norm": 1.5670911073684692, + "learning_rate": 5.6499240756951e-06, + "loss": 0.4302, + "step": 20735 + }, + { + "epoch": 0.6540293329127898, + "grad_norm": 1.5280076265335083, + "learning_rate": 5.64900426157516e-06, + "loss": 0.3255, + "step": 20736 + }, + { + "epoch": 0.6540608736792304, + "grad_norm": 1.428265929222107, + "learning_rate": 5.648084492861239e-06, + "loss": 0.4164, + "step": 20737 + }, + { + "epoch": 0.654092414445671, + "grad_norm": 1.6494933366775513, + "learning_rate": 5.647164769562939e-06, + "loss": 0.4148, + "step": 20738 + }, + { + "epoch": 0.6541239552121116, + "grad_norm": 1.729082465171814, + "learning_rate": 5.646245091689853e-06, + "loss": 0.4533, + "step": 20739 + }, + { + "epoch": 0.6541554959785523, + "grad_norm": 1.9651917219161987, + "learning_rate": 5.645325459251577e-06, + "loss": 0.4432, + "step": 20740 + }, + { + "epoch": 0.6541870367449929, + "grad_norm": 3.5396440029144287, + "learning_rate": 5.644405872257716e-06, + "loss": 0.4079, + "step": 20741 + }, + { + "epoch": 0.6542185775114335, + "grad_norm": 1.2831668853759766, + "learning_rate": 5.643486330717854e-06, + "loss": 0.3357, + "step": 20742 + }, + { + "epoch": 0.6542501182778742, + "grad_norm": 1.8329098224639893, + "learning_rate": 5.642566834641603e-06, + "loss": 0.3675, + "step": 20743 + }, + { + "epoch": 0.6542816590443148, + "grad_norm": 1.9055935144424438, + "learning_rate": 5.641647384038546e-06, + "loss": 0.4533, + "step": 20744 + }, + { + "epoch": 0.6543131998107554, + "grad_norm": 1.575742483139038, + "learning_rate": 5.640727978918283e-06, + "loss": 0.3357, + "step": 20745 + }, + { + "epoch": 0.654344740577196, + "grad_norm": 1.5952330827713013, + "learning_rate": 5.63980861929041e-06, + "loss": 0.3882, + "step": 20746 + }, + { + "epoch": 0.6543762813436367, + "grad_norm": 1.329999566078186, + "learning_rate": 5.638889305164519e-06, + "loss": 0.3512, + "step": 20747 + }, + { + "epoch": 0.6544078221100773, + "grad_norm": 1.7276310920715332, + "learning_rate": 5.637970036550202e-06, + "loss": 0.4317, + "step": 20748 + }, + { + "epoch": 0.6544393628765179, + "grad_norm": 1.5097064971923828, + "learning_rate": 5.6370508134570615e-06, + "loss": 0.4227, + "step": 20749 + }, + { + "epoch": 0.6544709036429586, + "grad_norm": 1.688019037246704, + "learning_rate": 5.636131635894678e-06, + "loss": 0.4145, + "step": 20750 + }, + { + "epoch": 0.6545024444093992, + "grad_norm": 1.4723174571990967, + "learning_rate": 5.63521250387265e-06, + "loss": 0.4452, + "step": 20751 + }, + { + "epoch": 0.6545339851758398, + "grad_norm": 1.6258829832077026, + "learning_rate": 5.634293417400573e-06, + "loss": 0.446, + "step": 20752 + }, + { + "epoch": 0.6545655259422803, + "grad_norm": 1.5880117416381836, + "learning_rate": 5.633374376488025e-06, + "loss": 0.4305, + "step": 20753 + }, + { + "epoch": 0.654597066708721, + "grad_norm": 2.6481006145477295, + "learning_rate": 5.6324553811446145e-06, + "loss": 0.7307, + "step": 20754 + }, + { + "epoch": 0.6546286074751616, + "grad_norm": 4.769604682922363, + "learning_rate": 5.63153643137992e-06, + "loss": 0.4566, + "step": 20755 + }, + { + "epoch": 0.6546601482416022, + "grad_norm": 1.5395601987838745, + "learning_rate": 5.630617527203533e-06, + "loss": 0.411, + "step": 20756 + }, + { + "epoch": 0.6546916890080429, + "grad_norm": 1.4519720077514648, + "learning_rate": 5.629698668625046e-06, + "loss": 0.7339, + "step": 20757 + }, + { + "epoch": 0.6547232297744835, + "grad_norm": 1.3456811904907227, + "learning_rate": 5.628779855654051e-06, + "loss": 0.314, + "step": 20758 + }, + { + "epoch": 0.6547547705409241, + "grad_norm": 1.5518534183502197, + "learning_rate": 5.6278610883001286e-06, + "loss": 0.4511, + "step": 20759 + }, + { + "epoch": 0.6547863113073648, + "grad_norm": 1.5473222732543945, + "learning_rate": 5.6269423665728704e-06, + "loss": 0.4015, + "step": 20760 + }, + { + "epoch": 0.6548178520738054, + "grad_norm": 1.723802924156189, + "learning_rate": 5.626023690481869e-06, + "loss": 0.4615, + "step": 20761 + }, + { + "epoch": 0.654849392840246, + "grad_norm": 1.8273875713348389, + "learning_rate": 5.6251050600366976e-06, + "loss": 0.4262, + "step": 20762 + }, + { + "epoch": 0.6548809336066866, + "grad_norm": 1.5282660722732544, + "learning_rate": 5.62418647524696e-06, + "loss": 0.4205, + "step": 20763 + }, + { + "epoch": 0.6549124743731273, + "grad_norm": 1.4667953252792358, + "learning_rate": 5.623267936122227e-06, + "loss": 0.4118, + "step": 20764 + }, + { + "epoch": 0.6549440151395679, + "grad_norm": 1.5598206520080566, + "learning_rate": 5.6223494426720986e-06, + "loss": 0.4087, + "step": 20765 + }, + { + "epoch": 0.6549755559060085, + "grad_norm": 1.7454800605773926, + "learning_rate": 5.6214309949061496e-06, + "loss": 0.4423, + "step": 20766 + }, + { + "epoch": 0.6550070966724492, + "grad_norm": 1.5763301849365234, + "learning_rate": 5.620512592833966e-06, + "loss": 0.4309, + "step": 20767 + }, + { + "epoch": 0.6550386374388898, + "grad_norm": 1.8321874141693115, + "learning_rate": 5.6195942364651365e-06, + "loss": 0.4238, + "step": 20768 + }, + { + "epoch": 0.6550701782053304, + "grad_norm": 1.600215196609497, + "learning_rate": 5.618675925809246e-06, + "loss": 0.4421, + "step": 20769 + }, + { + "epoch": 0.655101718971771, + "grad_norm": 2.0433661937713623, + "learning_rate": 5.617757660875871e-06, + "loss": 0.4427, + "step": 20770 + }, + { + "epoch": 0.6551332597382117, + "grad_norm": 1.3721319437026978, + "learning_rate": 5.616839441674595e-06, + "loss": 0.7113, + "step": 20771 + }, + { + "epoch": 0.6551648005046523, + "grad_norm": 1.612554669380188, + "learning_rate": 5.615921268215009e-06, + "loss": 0.4268, + "step": 20772 + }, + { + "epoch": 0.6551963412710929, + "grad_norm": 1.7286791801452637, + "learning_rate": 5.615003140506681e-06, + "loss": 0.4254, + "step": 20773 + }, + { + "epoch": 0.6552278820375335, + "grad_norm": 1.6937052011489868, + "learning_rate": 5.614085058559207e-06, + "loss": 0.4419, + "step": 20774 + }, + { + "epoch": 0.6552594228039741, + "grad_norm": 1.8614405393600464, + "learning_rate": 5.613167022382157e-06, + "loss": 0.4594, + "step": 20775 + }, + { + "epoch": 0.6552909635704147, + "grad_norm": 1.9580565690994263, + "learning_rate": 5.612249031985115e-06, + "loss": 0.4332, + "step": 20776 + }, + { + "epoch": 0.6553225043368553, + "grad_norm": 1.484167218208313, + "learning_rate": 5.611331087377663e-06, + "loss": 0.4188, + "step": 20777 + }, + { + "epoch": 0.655354045103296, + "grad_norm": 2.071791172027588, + "learning_rate": 5.610413188569381e-06, + "loss": 0.7232, + "step": 20778 + }, + { + "epoch": 0.6553855858697366, + "grad_norm": 1.6208195686340332, + "learning_rate": 5.609495335569842e-06, + "loss": 0.4182, + "step": 20779 + }, + { + "epoch": 0.6554171266361772, + "grad_norm": 1.5396018028259277, + "learning_rate": 5.608577528388629e-06, + "loss": 0.4554, + "step": 20780 + }, + { + "epoch": 0.6554486674026179, + "grad_norm": 1.580700159072876, + "learning_rate": 5.6076597670353195e-06, + "loss": 0.4324, + "step": 20781 + }, + { + "epoch": 0.6554802081690585, + "grad_norm": 1.3585842847824097, + "learning_rate": 5.606742051519491e-06, + "loss": 0.4657, + "step": 20782 + }, + { + "epoch": 0.6555117489354991, + "grad_norm": 1.4724299907684326, + "learning_rate": 5.605824381850724e-06, + "loss": 0.4321, + "step": 20783 + }, + { + "epoch": 0.6555432897019398, + "grad_norm": 1.5501500368118286, + "learning_rate": 5.604906758038583e-06, + "loss": 0.4312, + "step": 20784 + }, + { + "epoch": 0.6555748304683804, + "grad_norm": 1.936754822731018, + "learning_rate": 5.603989180092661e-06, + "loss": 0.3753, + "step": 20785 + }, + { + "epoch": 0.655606371234821, + "grad_norm": 1.9212416410446167, + "learning_rate": 5.603071648022522e-06, + "loss": 0.6933, + "step": 20786 + }, + { + "epoch": 0.6556379120012616, + "grad_norm": 1.5512930154800415, + "learning_rate": 5.602154161837744e-06, + "loss": 0.4137, + "step": 20787 + }, + { + "epoch": 0.6556694527677023, + "grad_norm": 1.7324212789535522, + "learning_rate": 5.601236721547903e-06, + "loss": 0.4042, + "step": 20788 + }, + { + "epoch": 0.6557009935341429, + "grad_norm": 1.8902965784072876, + "learning_rate": 5.600319327162575e-06, + "loss": 0.4471, + "step": 20789 + }, + { + "epoch": 0.6557325343005835, + "grad_norm": 1.525728464126587, + "learning_rate": 5.5994019786913274e-06, + "loss": 0.4445, + "step": 20790 + }, + { + "epoch": 0.6557640750670242, + "grad_norm": 1.4329203367233276, + "learning_rate": 5.598484676143739e-06, + "loss": 0.4584, + "step": 20791 + }, + { + "epoch": 0.6557956158334648, + "grad_norm": 1.5880296230316162, + "learning_rate": 5.597567419529381e-06, + "loss": 0.4352, + "step": 20792 + }, + { + "epoch": 0.6558271565999054, + "grad_norm": 1.656488299369812, + "learning_rate": 5.5966502088578215e-06, + "loss": 0.4579, + "step": 20793 + }, + { + "epoch": 0.655858697366346, + "grad_norm": 1.3185615539550781, + "learning_rate": 5.59573304413864e-06, + "loss": 0.412, + "step": 20794 + }, + { + "epoch": 0.6558902381327866, + "grad_norm": 2.6279361248016357, + "learning_rate": 5.5948159253814e-06, + "loss": 0.3796, + "step": 20795 + }, + { + "epoch": 0.6559217788992272, + "grad_norm": 1.97763991355896, + "learning_rate": 5.593898852595681e-06, + "loss": 0.4381, + "step": 20796 + }, + { + "epoch": 0.6559533196656678, + "grad_norm": 1.7695472240447998, + "learning_rate": 5.5929818257910476e-06, + "loss": 0.455, + "step": 20797 + }, + { + "epoch": 0.6559848604321085, + "grad_norm": 1.598801612854004, + "learning_rate": 5.59206484497707e-06, + "loss": 0.4126, + "step": 20798 + }, + { + "epoch": 0.6560164011985491, + "grad_norm": 1.4449642896652222, + "learning_rate": 5.591147910163322e-06, + "loss": 0.4266, + "step": 20799 + }, + { + "epoch": 0.6560479419649897, + "grad_norm": 1.9365278482437134, + "learning_rate": 5.590231021359364e-06, + "loss": 0.4379, + "step": 20800 + }, + { + "epoch": 0.6560794827314304, + "grad_norm": 1.8269292116165161, + "learning_rate": 5.58931417857477e-06, + "loss": 0.4506, + "step": 20801 + }, + { + "epoch": 0.656111023497871, + "grad_norm": 1.6604722738265991, + "learning_rate": 5.588397381819108e-06, + "loss": 0.4329, + "step": 20802 + }, + { + "epoch": 0.6561425642643116, + "grad_norm": 12.479109764099121, + "learning_rate": 5.587480631101948e-06, + "loss": 0.4649, + "step": 20803 + }, + { + "epoch": 0.6561741050307522, + "grad_norm": 1.385371208190918, + "learning_rate": 5.586563926432846e-06, + "loss": 0.3806, + "step": 20804 + }, + { + "epoch": 0.6562056457971929, + "grad_norm": 1.3404861688613892, + "learning_rate": 5.585647267821385e-06, + "loss": 0.4422, + "step": 20805 + }, + { + "epoch": 0.6562371865636335, + "grad_norm": 2.448983669281006, + "learning_rate": 5.584730655277118e-06, + "loss": 0.4716, + "step": 20806 + }, + { + "epoch": 0.6562687273300741, + "grad_norm": 1.4989601373672485, + "learning_rate": 5.5838140888096145e-06, + "loss": 0.4224, + "step": 20807 + }, + { + "epoch": 0.6563002680965148, + "grad_norm": 1.6107969284057617, + "learning_rate": 5.582897568428441e-06, + "loss": 0.4305, + "step": 20808 + }, + { + "epoch": 0.6563318088629554, + "grad_norm": 1.9738603830337524, + "learning_rate": 5.581981094143165e-06, + "loss": 0.4469, + "step": 20809 + }, + { + "epoch": 0.656363349629396, + "grad_norm": 2.135328531265259, + "learning_rate": 5.581064665963343e-06, + "loss": 0.4455, + "step": 20810 + }, + { + "epoch": 0.6563948903958366, + "grad_norm": 1.7515740394592285, + "learning_rate": 5.580148283898541e-06, + "loss": 0.4116, + "step": 20811 + }, + { + "epoch": 0.6564264311622773, + "grad_norm": 1.9891417026519775, + "learning_rate": 5.579231947958325e-06, + "loss": 0.3979, + "step": 20812 + }, + { + "epoch": 0.6564579719287179, + "grad_norm": 1.4401278495788574, + "learning_rate": 5.578315658152256e-06, + "loss": 0.4302, + "step": 20813 + }, + { + "epoch": 0.6564895126951585, + "grad_norm": 1.2657362222671509, + "learning_rate": 5.5773994144899e-06, + "loss": 0.7267, + "step": 20814 + }, + { + "epoch": 0.6565210534615992, + "grad_norm": 1.695864200592041, + "learning_rate": 5.576483216980808e-06, + "loss": 0.4432, + "step": 20815 + }, + { + "epoch": 0.6565525942280397, + "grad_norm": 1.6092957258224487, + "learning_rate": 5.5755670656345555e-06, + "loss": 0.4467, + "step": 20816 + }, + { + "epoch": 0.6565841349944803, + "grad_norm": 2.671996593475342, + "learning_rate": 5.574650960460692e-06, + "loss": 0.4461, + "step": 20817 + }, + { + "epoch": 0.656615675760921, + "grad_norm": 1.5134252309799194, + "learning_rate": 5.5737349014687836e-06, + "loss": 0.4299, + "step": 20818 + }, + { + "epoch": 0.6566472165273616, + "grad_norm": 1.5568689107894897, + "learning_rate": 5.57281888866839e-06, + "loss": 0.4229, + "step": 20819 + }, + { + "epoch": 0.6566787572938022, + "grad_norm": 1.785724401473999, + "learning_rate": 5.571902922069066e-06, + "loss": 0.6816, + "step": 20820 + }, + { + "epoch": 0.6567102980602428, + "grad_norm": 1.6215311288833618, + "learning_rate": 5.570987001680374e-06, + "loss": 0.4529, + "step": 20821 + }, + { + "epoch": 0.6567418388266835, + "grad_norm": 1.4772320985794067, + "learning_rate": 5.5700711275118716e-06, + "loss": 0.433, + "step": 20822 + }, + { + "epoch": 0.6567733795931241, + "grad_norm": 1.573439598083496, + "learning_rate": 5.56915529957312e-06, + "loss": 0.4492, + "step": 20823 + }, + { + "epoch": 0.6568049203595647, + "grad_norm": 1.4463770389556885, + "learning_rate": 5.568239517873666e-06, + "loss": 0.4019, + "step": 20824 + }, + { + "epoch": 0.6568364611260054, + "grad_norm": 1.8302265405654907, + "learning_rate": 5.567323782423082e-06, + "loss": 0.4527, + "step": 20825 + }, + { + "epoch": 0.656868001892446, + "grad_norm": 8.362781524658203, + "learning_rate": 5.566408093230911e-06, + "loss": 0.3299, + "step": 20826 + }, + { + "epoch": 0.6568995426588866, + "grad_norm": 1.7392581701278687, + "learning_rate": 5.5654924503067155e-06, + "loss": 0.4192, + "step": 20827 + }, + { + "epoch": 0.6569310834253272, + "grad_norm": 1.7544713020324707, + "learning_rate": 5.5645768536600486e-06, + "loss": 0.426, + "step": 20828 + }, + { + "epoch": 0.6569626241917679, + "grad_norm": 1.584193468093872, + "learning_rate": 5.563661303300466e-06, + "loss": 0.4363, + "step": 20829 + }, + { + "epoch": 0.6569941649582085, + "grad_norm": 1.5351715087890625, + "learning_rate": 5.562745799237527e-06, + "loss": 0.4231, + "step": 20830 + }, + { + "epoch": 0.6570257057246491, + "grad_norm": 1.629233479499817, + "learning_rate": 5.5618303414807764e-06, + "loss": 0.4073, + "step": 20831 + }, + { + "epoch": 0.6570572464910898, + "grad_norm": 1.8451675176620483, + "learning_rate": 5.5609149300397734e-06, + "loss": 0.4152, + "step": 20832 + }, + { + "epoch": 0.6570887872575304, + "grad_norm": 1.5638184547424316, + "learning_rate": 5.55999956492407e-06, + "loss": 0.4345, + "step": 20833 + }, + { + "epoch": 0.657120328023971, + "grad_norm": 1.5165895223617554, + "learning_rate": 5.559084246143222e-06, + "loss": 0.4137, + "step": 20834 + }, + { + "epoch": 0.6571518687904117, + "grad_norm": 1.4508551359176636, + "learning_rate": 5.55816897370677e-06, + "loss": 0.4371, + "step": 20835 + }, + { + "epoch": 0.6571834095568523, + "grad_norm": 1.9542341232299805, + "learning_rate": 5.557253747624283e-06, + "loss": 0.4011, + "step": 20836 + }, + { + "epoch": 0.6572149503232928, + "grad_norm": 1.6014304161071777, + "learning_rate": 5.556338567905297e-06, + "loss": 0.4541, + "step": 20837 + }, + { + "epoch": 0.6572464910897334, + "grad_norm": 1.4021042585372925, + "learning_rate": 5.555423434559369e-06, + "loss": 0.743, + "step": 20838 + }, + { + "epoch": 0.6572780318561741, + "grad_norm": 1.7626872062683105, + "learning_rate": 5.5545083475960535e-06, + "loss": 0.4173, + "step": 20839 + }, + { + "epoch": 0.6573095726226147, + "grad_norm": 1.5036637783050537, + "learning_rate": 5.553593307024892e-06, + "loss": 0.4186, + "step": 20840 + }, + { + "epoch": 0.6573411133890553, + "grad_norm": 1.5539989471435547, + "learning_rate": 5.552678312855435e-06, + "loss": 0.435, + "step": 20841 + }, + { + "epoch": 0.657372654155496, + "grad_norm": 2.4162516593933105, + "learning_rate": 5.551763365097233e-06, + "loss": 0.4017, + "step": 20842 + }, + { + "epoch": 0.6574041949219366, + "grad_norm": 3.443772792816162, + "learning_rate": 5.550848463759835e-06, + "loss": 0.4348, + "step": 20843 + }, + { + "epoch": 0.6574357356883772, + "grad_norm": 1.5426843166351318, + "learning_rate": 5.549933608852788e-06, + "loss": 0.3382, + "step": 20844 + }, + { + "epoch": 0.6574672764548178, + "grad_norm": 1.7283767461776733, + "learning_rate": 5.549018800385644e-06, + "loss": 0.4325, + "step": 20845 + }, + { + "epoch": 0.6574988172212585, + "grad_norm": 1.102109670639038, + "learning_rate": 5.548104038367938e-06, + "loss": 0.3115, + "step": 20846 + }, + { + "epoch": 0.6575303579876991, + "grad_norm": 1.287069320678711, + "learning_rate": 5.547189322809226e-06, + "loss": 0.4065, + "step": 20847 + }, + { + "epoch": 0.6575618987541397, + "grad_norm": 1.8798513412475586, + "learning_rate": 5.5462746537190505e-06, + "loss": 0.73, + "step": 20848 + }, + { + "epoch": 0.6575934395205804, + "grad_norm": 1.5486292839050293, + "learning_rate": 5.545360031106956e-06, + "loss": 0.4442, + "step": 20849 + }, + { + "epoch": 0.657624980287021, + "grad_norm": 2.909846305847168, + "learning_rate": 5.544445454982494e-06, + "loss": 0.3647, + "step": 20850 + }, + { + "epoch": 0.6576565210534616, + "grad_norm": 2.0051026344299316, + "learning_rate": 5.543530925355198e-06, + "loss": 0.4515, + "step": 20851 + }, + { + "epoch": 0.6576880618199022, + "grad_norm": 2.6950302124023438, + "learning_rate": 5.542616442234618e-06, + "loss": 0.4376, + "step": 20852 + }, + { + "epoch": 0.6577196025863429, + "grad_norm": 1.6238353252410889, + "learning_rate": 5.541702005630296e-06, + "loss": 0.4205, + "step": 20853 + }, + { + "epoch": 0.6577511433527835, + "grad_norm": 1.5088683366775513, + "learning_rate": 5.54078761555178e-06, + "loss": 0.4109, + "step": 20854 + }, + { + "epoch": 0.6577826841192241, + "grad_norm": 1.5785359144210815, + "learning_rate": 5.5398732720086005e-06, + "loss": 0.4205, + "step": 20855 + }, + { + "epoch": 0.6578142248856648, + "grad_norm": 1.673795461654663, + "learning_rate": 5.538958975010312e-06, + "loss": 0.4677, + "step": 20856 + }, + { + "epoch": 0.6578457656521054, + "grad_norm": 1.8956420421600342, + "learning_rate": 5.538044724566449e-06, + "loss": 0.423, + "step": 20857 + }, + { + "epoch": 0.6578773064185459, + "grad_norm": 1.5227408409118652, + "learning_rate": 5.537130520686552e-06, + "loss": 0.428, + "step": 20858 + }, + { + "epoch": 0.6579088471849865, + "grad_norm": 1.443965196609497, + "learning_rate": 5.536216363380169e-06, + "loss": 0.3425, + "step": 20859 + }, + { + "epoch": 0.6579403879514272, + "grad_norm": 1.452325463294983, + "learning_rate": 5.535302252656824e-06, + "loss": 0.3942, + "step": 20860 + }, + { + "epoch": 0.6579719287178678, + "grad_norm": 2.0716371536254883, + "learning_rate": 5.534388188526078e-06, + "loss": 0.4364, + "step": 20861 + }, + { + "epoch": 0.6580034694843084, + "grad_norm": 1.8163511753082275, + "learning_rate": 5.533474170997453e-06, + "loss": 0.4341, + "step": 20862 + }, + { + "epoch": 0.6580350102507491, + "grad_norm": 1.652086615562439, + "learning_rate": 5.5325602000804925e-06, + "loss": 0.4437, + "step": 20863 + }, + { + "epoch": 0.6580665510171897, + "grad_norm": 1.7766033411026, + "learning_rate": 5.531646275784737e-06, + "loss": 0.4519, + "step": 20864 + }, + { + "epoch": 0.6580980917836303, + "grad_norm": 1.9915556907653809, + "learning_rate": 5.530732398119725e-06, + "loss": 0.4364, + "step": 20865 + }, + { + "epoch": 0.658129632550071, + "grad_norm": 1.5849961042404175, + "learning_rate": 5.5298185670949855e-06, + "loss": 0.7122, + "step": 20866 + }, + { + "epoch": 0.6581611733165116, + "grad_norm": 1.5607876777648926, + "learning_rate": 5.528904782720061e-06, + "loss": 0.461, + "step": 20867 + }, + { + "epoch": 0.6581927140829522, + "grad_norm": 1.4404306411743164, + "learning_rate": 5.527991045004488e-06, + "loss": 0.41, + "step": 20868 + }, + { + "epoch": 0.6582242548493928, + "grad_norm": 1.635215401649475, + "learning_rate": 5.527077353957801e-06, + "loss": 0.4613, + "step": 20869 + }, + { + "epoch": 0.6582557956158335, + "grad_norm": 2.0000975131988525, + "learning_rate": 5.526163709589538e-06, + "loss": 0.4181, + "step": 20870 + }, + { + "epoch": 0.6582873363822741, + "grad_norm": 1.8830790519714355, + "learning_rate": 5.525250111909226e-06, + "loss": 0.4424, + "step": 20871 + }, + { + "epoch": 0.6583188771487147, + "grad_norm": 1.822501540184021, + "learning_rate": 5.5243365609264044e-06, + "loss": 0.4271, + "step": 20872 + }, + { + "epoch": 0.6583504179151554, + "grad_norm": 1.6758339405059814, + "learning_rate": 5.523423056650605e-06, + "loss": 0.4567, + "step": 20873 + }, + { + "epoch": 0.658381958681596, + "grad_norm": 10.353041648864746, + "learning_rate": 5.522509599091368e-06, + "loss": 0.3304, + "step": 20874 + }, + { + "epoch": 0.6584134994480366, + "grad_norm": 1.4749938249588013, + "learning_rate": 5.521596188258211e-06, + "loss": 0.4419, + "step": 20875 + }, + { + "epoch": 0.6584450402144773, + "grad_norm": 1.630947470664978, + "learning_rate": 5.520682824160684e-06, + "loss": 0.4122, + "step": 20876 + }, + { + "epoch": 0.6584765809809179, + "grad_norm": 1.3361393213272095, + "learning_rate": 5.519769506808303e-06, + "loss": 0.4114, + "step": 20877 + }, + { + "epoch": 0.6585081217473585, + "grad_norm": 1.5002223253250122, + "learning_rate": 5.518856236210609e-06, + "loss": 0.7277, + "step": 20878 + }, + { + "epoch": 0.6585396625137991, + "grad_norm": 1.384063959121704, + "learning_rate": 5.517943012377133e-06, + "loss": 0.4262, + "step": 20879 + }, + { + "epoch": 0.6585712032802397, + "grad_norm": 1.6324989795684814, + "learning_rate": 5.517029835317393e-06, + "loss": 0.4254, + "step": 20880 + }, + { + "epoch": 0.6586027440466803, + "grad_norm": 4.315080165863037, + "learning_rate": 5.516116705040937e-06, + "loss": 0.4482, + "step": 20881 + }, + { + "epoch": 0.6586342848131209, + "grad_norm": 3.271228790283203, + "learning_rate": 5.515203621557279e-06, + "loss": 0.4085, + "step": 20882 + }, + { + "epoch": 0.6586658255795615, + "grad_norm": 1.4662054777145386, + "learning_rate": 5.5142905848759545e-06, + "loss": 0.4006, + "step": 20883 + }, + { + "epoch": 0.6586973663460022, + "grad_norm": 1.6357322931289673, + "learning_rate": 5.513377595006491e-06, + "loss": 0.4597, + "step": 20884 + }, + { + "epoch": 0.6587289071124428, + "grad_norm": 2.348081111907959, + "learning_rate": 5.51246465195842e-06, + "loss": 0.4596, + "step": 20885 + }, + { + "epoch": 0.6587604478788834, + "grad_norm": 1.4479738473892212, + "learning_rate": 5.511551755741261e-06, + "loss": 0.4215, + "step": 20886 + }, + { + "epoch": 0.6587919886453241, + "grad_norm": 1.3103004693984985, + "learning_rate": 5.510638906364545e-06, + "loss": 0.4342, + "step": 20887 + }, + { + "epoch": 0.6588235294117647, + "grad_norm": 1.8068654537200928, + "learning_rate": 5.509726103837801e-06, + "loss": 0.446, + "step": 20888 + }, + { + "epoch": 0.6588550701782053, + "grad_norm": 1.5207393169403076, + "learning_rate": 5.508813348170544e-06, + "loss": 0.4451, + "step": 20889 + }, + { + "epoch": 0.658886610944646, + "grad_norm": 1.7028931379318237, + "learning_rate": 5.507900639372315e-06, + "loss": 0.4247, + "step": 20890 + }, + { + "epoch": 0.6589181517110866, + "grad_norm": 2.219398260116577, + "learning_rate": 5.506987977452622e-06, + "loss": 0.3461, + "step": 20891 + }, + { + "epoch": 0.6589496924775272, + "grad_norm": 1.8102195262908936, + "learning_rate": 5.506075362421008e-06, + "loss": 0.4439, + "step": 20892 + }, + { + "epoch": 0.6589812332439678, + "grad_norm": 1.5761215686798096, + "learning_rate": 5.505162794286982e-06, + "loss": 0.4346, + "step": 20893 + }, + { + "epoch": 0.6590127740104085, + "grad_norm": 1.3819762468338013, + "learning_rate": 5.504250273060072e-06, + "loss": 0.7537, + "step": 20894 + }, + { + "epoch": 0.6590443147768491, + "grad_norm": 1.9234484434127808, + "learning_rate": 5.503337798749802e-06, + "loss": 0.4382, + "step": 20895 + }, + { + "epoch": 0.6590758555432897, + "grad_norm": 2.1071512699127197, + "learning_rate": 5.502425371365699e-06, + "loss": 0.4606, + "step": 20896 + }, + { + "epoch": 0.6591073963097304, + "grad_norm": 1.4597547054290771, + "learning_rate": 5.501512990917273e-06, + "loss": 0.7177, + "step": 20897 + }, + { + "epoch": 0.659138937076171, + "grad_norm": 2.009735345840454, + "learning_rate": 5.500600657414052e-06, + "loss": 0.4301, + "step": 20898 + }, + { + "epoch": 0.6591704778426116, + "grad_norm": 1.3967633247375488, + "learning_rate": 5.499688370865563e-06, + "loss": 0.4315, + "step": 20899 + }, + { + "epoch": 0.6592020186090523, + "grad_norm": 1.50557541847229, + "learning_rate": 5.49877613128131e-06, + "loss": 0.4282, + "step": 20900 + }, + { + "epoch": 0.6592335593754928, + "grad_norm": 3.4764363765716553, + "learning_rate": 5.497863938670832e-06, + "loss": 0.4039, + "step": 20901 + }, + { + "epoch": 0.6592651001419334, + "grad_norm": 1.7310965061187744, + "learning_rate": 5.496951793043637e-06, + "loss": 0.4223, + "step": 20902 + }, + { + "epoch": 0.659296640908374, + "grad_norm": 1.6470328569412231, + "learning_rate": 5.496039694409245e-06, + "loss": 0.4329, + "step": 20903 + }, + { + "epoch": 0.6593281816748147, + "grad_norm": 2.5012781620025635, + "learning_rate": 5.495127642777176e-06, + "loss": 0.432, + "step": 20904 + }, + { + "epoch": 0.6593597224412553, + "grad_norm": 1.774880290031433, + "learning_rate": 5.494215638156952e-06, + "loss": 0.4205, + "step": 20905 + }, + { + "epoch": 0.6593912632076959, + "grad_norm": 1.9144610166549683, + "learning_rate": 5.493303680558083e-06, + "loss": 0.4485, + "step": 20906 + }, + { + "epoch": 0.6594228039741366, + "grad_norm": 2.2177131175994873, + "learning_rate": 5.492391769990089e-06, + "loss": 0.4178, + "step": 20907 + }, + { + "epoch": 0.6594543447405772, + "grad_norm": 1.637108325958252, + "learning_rate": 5.491479906462487e-06, + "loss": 0.3748, + "step": 20908 + }, + { + "epoch": 0.6594858855070178, + "grad_norm": 2.7432236671447754, + "learning_rate": 5.490568089984794e-06, + "loss": 0.4302, + "step": 20909 + }, + { + "epoch": 0.6595174262734584, + "grad_norm": 1.6728498935699463, + "learning_rate": 5.489656320566528e-06, + "loss": 0.4462, + "step": 20910 + }, + { + "epoch": 0.6595489670398991, + "grad_norm": 1.7743723392486572, + "learning_rate": 5.48874459821719e-06, + "loss": 0.4571, + "step": 20911 + }, + { + "epoch": 0.6595805078063397, + "grad_norm": 1.5758816003799438, + "learning_rate": 5.487832922946317e-06, + "loss": 0.4662, + "step": 20912 + }, + { + "epoch": 0.6596120485727803, + "grad_norm": 2.145364761352539, + "learning_rate": 5.486921294763404e-06, + "loss": 0.3413, + "step": 20913 + }, + { + "epoch": 0.659643589339221, + "grad_norm": 1.6554564237594604, + "learning_rate": 5.486009713677973e-06, + "loss": 0.4492, + "step": 20914 + }, + { + "epoch": 0.6596751301056616, + "grad_norm": 1.537216067314148, + "learning_rate": 5.485098179699535e-06, + "loss": 0.4261, + "step": 20915 + }, + { + "epoch": 0.6597066708721022, + "grad_norm": 1.4663382768630981, + "learning_rate": 5.4841866928376075e-06, + "loss": 0.4147, + "step": 20916 + }, + { + "epoch": 0.6597382116385428, + "grad_norm": 1.610093593597412, + "learning_rate": 5.483275253101695e-06, + "loss": 0.4521, + "step": 20917 + }, + { + "epoch": 0.6597697524049835, + "grad_norm": 1.3857890367507935, + "learning_rate": 5.482363860501312e-06, + "loss": 0.4574, + "step": 20918 + }, + { + "epoch": 0.6598012931714241, + "grad_norm": 3.6609585285186768, + "learning_rate": 5.481452515045974e-06, + "loss": 0.4152, + "step": 20919 + }, + { + "epoch": 0.6598328339378647, + "grad_norm": 1.8736557960510254, + "learning_rate": 5.480541216745179e-06, + "loss": 0.4331, + "step": 20920 + }, + { + "epoch": 0.6598643747043054, + "grad_norm": 2.362363815307617, + "learning_rate": 5.479629965608455e-06, + "loss": 0.3792, + "step": 20921 + }, + { + "epoch": 0.6598959154707459, + "grad_norm": 1.2633689641952515, + "learning_rate": 5.478718761645293e-06, + "loss": 0.6897, + "step": 20922 + }, + { + "epoch": 0.6599274562371865, + "grad_norm": 1.7987183332443237, + "learning_rate": 5.47780760486522e-06, + "loss": 0.4471, + "step": 20923 + }, + { + "epoch": 0.6599589970036271, + "grad_norm": 1.825927972793579, + "learning_rate": 5.476896495277732e-06, + "loss": 0.4654, + "step": 20924 + }, + { + "epoch": 0.6599905377700678, + "grad_norm": 1.6870801448822021, + "learning_rate": 5.47598543289234e-06, + "loss": 0.4277, + "step": 20925 + }, + { + "epoch": 0.6600220785365084, + "grad_norm": 1.5210397243499756, + "learning_rate": 5.475074417718558e-06, + "loss": 0.3844, + "step": 20926 + }, + { + "epoch": 0.660053619302949, + "grad_norm": 1.5465402603149414, + "learning_rate": 5.474163449765884e-06, + "loss": 0.4147, + "step": 20927 + }, + { + "epoch": 0.6600851600693897, + "grad_norm": 2.624185800552368, + "learning_rate": 5.473252529043829e-06, + "loss": 0.4018, + "step": 20928 + }, + { + "epoch": 0.6601167008358303, + "grad_norm": 2.194056272506714, + "learning_rate": 5.472341655561897e-06, + "loss": 0.4156, + "step": 20929 + }, + { + "epoch": 0.6601482416022709, + "grad_norm": 1.415503978729248, + "learning_rate": 5.471430829329601e-06, + "loss": 0.421, + "step": 20930 + }, + { + "epoch": 0.6601797823687116, + "grad_norm": 3.2460265159606934, + "learning_rate": 5.4705200503564315e-06, + "loss": 0.4408, + "step": 20931 + }, + { + "epoch": 0.6602113231351522, + "grad_norm": 1.7993311882019043, + "learning_rate": 5.4696093186519115e-06, + "loss": 0.4224, + "step": 20932 + }, + { + "epoch": 0.6602428639015928, + "grad_norm": 1.5985428094863892, + "learning_rate": 5.468698634225531e-06, + "loss": 0.4442, + "step": 20933 + }, + { + "epoch": 0.6602744046680334, + "grad_norm": 1.534923791885376, + "learning_rate": 5.4677879970868e-06, + "loss": 0.4321, + "step": 20934 + }, + { + "epoch": 0.6603059454344741, + "grad_norm": 1.4536441564559937, + "learning_rate": 5.4668774072452194e-06, + "loss": 0.4344, + "step": 20935 + }, + { + "epoch": 0.6603374862009147, + "grad_norm": 1.4521031379699707, + "learning_rate": 5.465966864710297e-06, + "loss": 0.4324, + "step": 20936 + }, + { + "epoch": 0.6603690269673553, + "grad_norm": 1.8247735500335693, + "learning_rate": 5.465056369491528e-06, + "loss": 0.4122, + "step": 20937 + }, + { + "epoch": 0.660400567733796, + "grad_norm": 1.5938894748687744, + "learning_rate": 5.464145921598416e-06, + "loss": 0.4571, + "step": 20938 + }, + { + "epoch": 0.6604321085002366, + "grad_norm": 2.4293365478515625, + "learning_rate": 5.4632355210404645e-06, + "loss": 0.4177, + "step": 20939 + }, + { + "epoch": 0.6604636492666772, + "grad_norm": 2.1849498748779297, + "learning_rate": 5.462325167827171e-06, + "loss": 0.4128, + "step": 20940 + }, + { + "epoch": 0.6604951900331179, + "grad_norm": 1.5275126695632935, + "learning_rate": 5.461414861968044e-06, + "loss": 0.4434, + "step": 20941 + }, + { + "epoch": 0.6605267307995585, + "grad_norm": 1.5326849222183228, + "learning_rate": 5.460504603472568e-06, + "loss": 0.4071, + "step": 20942 + }, + { + "epoch": 0.660558271565999, + "grad_norm": 1.3323935270309448, + "learning_rate": 5.4595943923502585e-06, + "loss": 0.7244, + "step": 20943 + }, + { + "epoch": 0.6605898123324396, + "grad_norm": 1.4939947128295898, + "learning_rate": 5.458684228610605e-06, + "loss": 0.729, + "step": 20944 + }, + { + "epoch": 0.6606213530988803, + "grad_norm": 1.4448893070220947, + "learning_rate": 5.457774112263106e-06, + "loss": 0.4402, + "step": 20945 + }, + { + "epoch": 0.6606528938653209, + "grad_norm": 1.668887734413147, + "learning_rate": 5.456864043317265e-06, + "loss": 0.4144, + "step": 20946 + }, + { + "epoch": 0.6606844346317615, + "grad_norm": 6.771172046661377, + "learning_rate": 5.455954021782573e-06, + "loss": 0.4248, + "step": 20947 + }, + { + "epoch": 0.6607159753982021, + "grad_norm": 1.6172873973846436, + "learning_rate": 5.455044047668527e-06, + "loss": 0.4354, + "step": 20948 + }, + { + "epoch": 0.6607475161646428, + "grad_norm": 1.6577818393707275, + "learning_rate": 5.454134120984627e-06, + "loss": 0.6979, + "step": 20949 + }, + { + "epoch": 0.6607790569310834, + "grad_norm": 1.7109586000442505, + "learning_rate": 5.453224241740371e-06, + "loss": 0.457, + "step": 20950 + }, + { + "epoch": 0.660810597697524, + "grad_norm": 1.6502741575241089, + "learning_rate": 5.4523144099452425e-06, + "loss": 0.4335, + "step": 20951 + }, + { + "epoch": 0.6608421384639647, + "grad_norm": 1.7315691709518433, + "learning_rate": 5.451404625608752e-06, + "loss": 0.4297, + "step": 20952 + }, + { + "epoch": 0.6608736792304053, + "grad_norm": 1.3766109943389893, + "learning_rate": 5.450494888740383e-06, + "loss": 0.425, + "step": 20953 + }, + { + "epoch": 0.6609052199968459, + "grad_norm": 1.5536620616912842, + "learning_rate": 5.44958519934963e-06, + "loss": 0.4365, + "step": 20954 + }, + { + "epoch": 0.6609367607632866, + "grad_norm": 2.5772323608398438, + "learning_rate": 5.448675557445992e-06, + "loss": 0.4195, + "step": 20955 + }, + { + "epoch": 0.6609683015297272, + "grad_norm": 1.4797213077545166, + "learning_rate": 5.447765963038957e-06, + "loss": 0.4509, + "step": 20956 + }, + { + "epoch": 0.6609998422961678, + "grad_norm": 1.7438267469406128, + "learning_rate": 5.446856416138022e-06, + "loss": 0.4123, + "step": 20957 + }, + { + "epoch": 0.6610313830626084, + "grad_norm": 1.55134117603302, + "learning_rate": 5.445946916752674e-06, + "loss": 0.4241, + "step": 20958 + }, + { + "epoch": 0.6610629238290491, + "grad_norm": 1.674796462059021, + "learning_rate": 5.445037464892404e-06, + "loss": 0.4394, + "step": 20959 + }, + { + "epoch": 0.6610944645954897, + "grad_norm": 1.7542803287506104, + "learning_rate": 5.4441280605667055e-06, + "loss": 0.4817, + "step": 20960 + }, + { + "epoch": 0.6611260053619303, + "grad_norm": 2.1049726009368896, + "learning_rate": 5.443218703785073e-06, + "loss": 0.4178, + "step": 20961 + }, + { + "epoch": 0.661157546128371, + "grad_norm": 2.119535207748413, + "learning_rate": 5.442309394556983e-06, + "loss": 0.3624, + "step": 20962 + }, + { + "epoch": 0.6611890868948116, + "grad_norm": 1.7555590867996216, + "learning_rate": 5.441400132891941e-06, + "loss": 0.4543, + "step": 20963 + }, + { + "epoch": 0.6612206276612521, + "grad_norm": 1.4939086437225342, + "learning_rate": 5.440490918799424e-06, + "loss": 0.4317, + "step": 20964 + }, + { + "epoch": 0.6612521684276927, + "grad_norm": 2.6203317642211914, + "learning_rate": 5.439581752288925e-06, + "loss": 0.4424, + "step": 20965 + }, + { + "epoch": 0.6612837091941334, + "grad_norm": 1.540306806564331, + "learning_rate": 5.438672633369937e-06, + "loss": 0.4428, + "step": 20966 + }, + { + "epoch": 0.661315249960574, + "grad_norm": 1.825919270515442, + "learning_rate": 5.437763562051936e-06, + "loss": 0.4447, + "step": 20967 + }, + { + "epoch": 0.6613467907270146, + "grad_norm": 1.8094459772109985, + "learning_rate": 5.436854538344416e-06, + "loss": 0.4558, + "step": 20968 + }, + { + "epoch": 0.6613783314934553, + "grad_norm": 2.19561505317688, + "learning_rate": 5.435945562256862e-06, + "loss": 0.4166, + "step": 20969 + }, + { + "epoch": 0.6614098722598959, + "grad_norm": 1.5365597009658813, + "learning_rate": 5.435036633798758e-06, + "loss": 0.4333, + "step": 20970 + }, + { + "epoch": 0.6614414130263365, + "grad_norm": 1.7340773344039917, + "learning_rate": 5.434127752979592e-06, + "loss": 0.4236, + "step": 20971 + }, + { + "epoch": 0.6614729537927772, + "grad_norm": 1.7581356763839722, + "learning_rate": 5.433218919808853e-06, + "loss": 0.4564, + "step": 20972 + }, + { + "epoch": 0.6615044945592178, + "grad_norm": 1.7057782411575317, + "learning_rate": 5.432310134296017e-06, + "loss": 0.4255, + "step": 20973 + }, + { + "epoch": 0.6615360353256584, + "grad_norm": 1.344422698020935, + "learning_rate": 5.4314013964505705e-06, + "loss": 0.4375, + "step": 20974 + }, + { + "epoch": 0.661567576092099, + "grad_norm": 1.463549017906189, + "learning_rate": 5.430492706281999e-06, + "loss": 0.4348, + "step": 20975 + }, + { + "epoch": 0.6615991168585397, + "grad_norm": 1.878241777420044, + "learning_rate": 5.4295840637997846e-06, + "loss": 0.3749, + "step": 20976 + }, + { + "epoch": 0.6616306576249803, + "grad_norm": 1.7423511743545532, + "learning_rate": 5.428675469013411e-06, + "loss": 0.4371, + "step": 20977 + }, + { + "epoch": 0.6616621983914209, + "grad_norm": 1.6331380605697632, + "learning_rate": 5.427766921932356e-06, + "loss": 0.4737, + "step": 20978 + }, + { + "epoch": 0.6616937391578616, + "grad_norm": 2.7710580825805664, + "learning_rate": 5.426858422566103e-06, + "loss": 0.3811, + "step": 20979 + }, + { + "epoch": 0.6617252799243022, + "grad_norm": 1.312907099723816, + "learning_rate": 5.425949970924134e-06, + "loss": 0.7602, + "step": 20980 + }, + { + "epoch": 0.6617568206907428, + "grad_norm": 1.6656341552734375, + "learning_rate": 5.425041567015932e-06, + "loss": 0.4456, + "step": 20981 + }, + { + "epoch": 0.6617883614571834, + "grad_norm": 1.6092151403427124, + "learning_rate": 5.424133210850966e-06, + "loss": 0.4316, + "step": 20982 + }, + { + "epoch": 0.6618199022236241, + "grad_norm": 1.7513158321380615, + "learning_rate": 5.423224902438731e-06, + "loss": 0.4114, + "step": 20983 + }, + { + "epoch": 0.6618514429900647, + "grad_norm": 1.5342485904693604, + "learning_rate": 5.422316641788693e-06, + "loss": 0.6798, + "step": 20984 + }, + { + "epoch": 0.6618829837565052, + "grad_norm": 1.9604063034057617, + "learning_rate": 5.421408428910335e-06, + "loss": 0.7192, + "step": 20985 + }, + { + "epoch": 0.6619145245229459, + "grad_norm": 1.6835318803787231, + "learning_rate": 5.420500263813141e-06, + "loss": 0.4711, + "step": 20986 + }, + { + "epoch": 0.6619460652893865, + "grad_norm": 1.6928164958953857, + "learning_rate": 5.419592146506574e-06, + "loss": 0.4116, + "step": 20987 + }, + { + "epoch": 0.6619776060558271, + "grad_norm": 1.9894134998321533, + "learning_rate": 5.418684077000127e-06, + "loss": 0.3791, + "step": 20988 + }, + { + "epoch": 0.6620091468222677, + "grad_norm": 1.6563571691513062, + "learning_rate": 5.417776055303266e-06, + "loss": 0.4365, + "step": 20989 + }, + { + "epoch": 0.6620406875887084, + "grad_norm": 1.5522065162658691, + "learning_rate": 5.416868081425468e-06, + "loss": 0.4339, + "step": 20990 + }, + { + "epoch": 0.662072228355149, + "grad_norm": 8.240947723388672, + "learning_rate": 5.415960155376212e-06, + "loss": 0.4746, + "step": 20991 + }, + { + "epoch": 0.6621037691215896, + "grad_norm": 1.5523934364318848, + "learning_rate": 5.4150522771649745e-06, + "loss": 0.4436, + "step": 20992 + }, + { + "epoch": 0.6621353098880303, + "grad_norm": 2.5114033222198486, + "learning_rate": 5.414144446801224e-06, + "loss": 0.4316, + "step": 20993 + }, + { + "epoch": 0.6621668506544709, + "grad_norm": 1.6694890260696411, + "learning_rate": 5.4132366642944356e-06, + "loss": 0.4287, + "step": 20994 + }, + { + "epoch": 0.6621983914209115, + "grad_norm": 1.7086278200149536, + "learning_rate": 5.412328929654085e-06, + "loss": 0.471, + "step": 20995 + }, + { + "epoch": 0.6622299321873522, + "grad_norm": 2.278498411178589, + "learning_rate": 5.411421242889643e-06, + "loss": 0.4498, + "step": 20996 + }, + { + "epoch": 0.6622614729537928, + "grad_norm": 1.4760152101516724, + "learning_rate": 5.410513604010588e-06, + "loss": 0.4335, + "step": 20997 + }, + { + "epoch": 0.6622930137202334, + "grad_norm": 1.7124119997024536, + "learning_rate": 5.409606013026383e-06, + "loss": 0.4397, + "step": 20998 + }, + { + "epoch": 0.662324554486674, + "grad_norm": 3.2445883750915527, + "learning_rate": 5.408698469946503e-06, + "loss": 0.4303, + "step": 20999 + }, + { + "epoch": 0.6623560952531147, + "grad_norm": 1.5369901657104492, + "learning_rate": 5.40779097478042e-06, + "loss": 0.3948, + "step": 21000 + } + ], + "logging_steps": 1.0, + "max_steps": 31705, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 3000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 4.9581235147677e+20, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}