diff --git "a/multi_student_1/trainer_state.json" "b/multi_student_1/trainer_state.json" new file mode 100644--- /dev/null +++ "b/multi_student_1/trainer_state.json" @@ -0,0 +1,7106 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 16.0, + "eval_steps": 500, + "global_step": 10096, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.01586042823156225, + "grad_norm": 19.139528850030874, + "learning_rate": 8.910891089108911e-07, + "loss": 8.2466, + "step": 10 + }, + { + "epoch": 0.0317208564631245, + "grad_norm": 17.8179869039751, + "learning_rate": 1.8811881188118813e-06, + "loss": 8.1206, + "step": 20 + }, + { + "epoch": 0.047581284694686754, + "grad_norm": 9.164020050000067, + "learning_rate": 2.8712871287128717e-06, + "loss": 7.4762, + "step": 30 + }, + { + "epoch": 0.063441712926249, + "grad_norm": 4.906293902424142, + "learning_rate": 3.861386138613861e-06, + "loss": 6.7446, + "step": 40 + }, + { + "epoch": 0.07930214115781126, + "grad_norm": 3.052162672695787, + "learning_rate": 4.851485148514852e-06, + "loss": 6.2761, + "step": 50 + }, + { + "epoch": 0.09516256938937351, + "grad_norm": 2.98736461961841, + "learning_rate": 5.841584158415842e-06, + "loss": 5.929, + "step": 60 + }, + { + "epoch": 0.11102299762093576, + "grad_norm": 3.707337444231957, + "learning_rate": 6.831683168316833e-06, + "loss": 5.5255, + "step": 70 + }, + { + "epoch": 0.126883425852498, + "grad_norm": 4.5149941785829, + "learning_rate": 7.821782178217822e-06, + "loss": 4.9205, + "step": 80 + }, + { + "epoch": 0.14274385408406026, + "grad_norm": 4.194108732375239, + "learning_rate": 8.811881188118812e-06, + "loss": 4.2318, + "step": 90 + }, + { + "epoch": 0.1586042823156225, + "grad_norm": 3.8369189131135757, + "learning_rate": 9.801980198019804e-06, + "loss": 3.6352, + "step": 100 + }, + { + "epoch": 0.17446471054718476, + "grad_norm": 3.7340050812724255, + "learning_rate": 1.0792079207920792e-05, + "loss": 3.2605, + "step": 110 + }, + { + "epoch": 0.19032513877874702, + "grad_norm": 3.9254417022343637, + "learning_rate": 1.1782178217821782e-05, + "loss": 3.0544, + "step": 120 + }, + { + "epoch": 0.20618556701030927, + "grad_norm": 4.840686832835472, + "learning_rate": 1.2772277227722773e-05, + "loss": 2.9196, + "step": 130 + }, + { + "epoch": 0.22204599524187152, + "grad_norm": 13.780201154926932, + "learning_rate": 1.3762376237623763e-05, + "loss": 2.8095, + "step": 140 + }, + { + "epoch": 0.23790642347343377, + "grad_norm": 6.344054558352859, + "learning_rate": 1.4752475247524752e-05, + "loss": 2.7136, + "step": 150 + }, + { + "epoch": 0.253766851704996, + "grad_norm": 4.982092617175093, + "learning_rate": 1.5742574257425743e-05, + "loss": 2.6392, + "step": 160 + }, + { + "epoch": 0.2696272799365583, + "grad_norm": 8.162264297232719, + "learning_rate": 1.6732673267326735e-05, + "loss": 2.5676, + "step": 170 + }, + { + "epoch": 0.2854877081681205, + "grad_norm": 9.068632822505595, + "learning_rate": 1.7722772277227723e-05, + "loss": 2.4886, + "step": 180 + }, + { + "epoch": 0.3013481363996828, + "grad_norm": 21.522422147460635, + "learning_rate": 1.8712871287128715e-05, + "loss": 2.424, + "step": 190 + }, + { + "epoch": 0.317208564631245, + "grad_norm": 20.83191234754218, + "learning_rate": 1.9702970297029703e-05, + "loss": 2.3979, + "step": 200 + }, + { + "epoch": 0.3330689928628073, + "grad_norm": 11.966941640205993, + "learning_rate": 2.069306930693069e-05, + "loss": 2.3199, + "step": 210 + }, + { + "epoch": 0.34892942109436953, + "grad_norm": 13.845469765092606, + "learning_rate": 2.1683168316831683e-05, + "loss": 2.2642, + "step": 220 + }, + { + "epoch": 0.3647898493259318, + "grad_norm": 10.896989301572372, + "learning_rate": 2.2673267326732675e-05, + "loss": 2.1938, + "step": 230 + }, + { + "epoch": 0.38065027755749403, + "grad_norm": 8.461441256784877, + "learning_rate": 2.3663366336633663e-05, + "loss": 2.1093, + "step": 240 + }, + { + "epoch": 0.3965107057890563, + "grad_norm": 15.437949044441957, + "learning_rate": 2.4653465346534655e-05, + "loss": 2.0244, + "step": 250 + }, + { + "epoch": 0.41237113402061853, + "grad_norm": 6.9324747811801215, + "learning_rate": 2.5643564356435646e-05, + "loss": 1.9328, + "step": 260 + }, + { + "epoch": 0.4282315622521808, + "grad_norm": 6.539929054412363, + "learning_rate": 2.6633663366336638e-05, + "loss": 1.8107, + "step": 270 + }, + { + "epoch": 0.44409199048374304, + "grad_norm": 9.546138483110752, + "learning_rate": 2.7623762376237626e-05, + "loss": 1.719, + "step": 280 + }, + { + "epoch": 0.4599524187153053, + "grad_norm": 6.921145854402886, + "learning_rate": 2.8613861386138614e-05, + "loss": 1.64, + "step": 290 + }, + { + "epoch": 0.47581284694686754, + "grad_norm": 5.464388935314663, + "learning_rate": 2.9603960396039603e-05, + "loss": 1.5848, + "step": 300 + }, + { + "epoch": 0.4916732751784298, + "grad_norm": 5.644631505546815, + "learning_rate": 3.0594059405940594e-05, + "loss": 1.515, + "step": 310 + }, + { + "epoch": 0.507533703409992, + "grad_norm": 7.749393965323654, + "learning_rate": 3.158415841584158e-05, + "loss": 1.4565, + "step": 320 + }, + { + "epoch": 0.5233941316415543, + "grad_norm": 7.011857430438995, + "learning_rate": 3.257425742574258e-05, + "loss": 1.4069, + "step": 330 + }, + { + "epoch": 0.5392545598731165, + "grad_norm": 4.774306275517458, + "learning_rate": 3.3564356435643566e-05, + "loss": 1.3559, + "step": 340 + }, + { + "epoch": 0.5551149881046789, + "grad_norm": 4.827841991876933, + "learning_rate": 3.4554455445544554e-05, + "loss": 1.3113, + "step": 350 + }, + { + "epoch": 0.570975416336241, + "grad_norm": 6.409703044827469, + "learning_rate": 3.554455445544554e-05, + "loss": 1.272, + "step": 360 + }, + { + "epoch": 0.5868358445678034, + "grad_norm": 7.223403220564374, + "learning_rate": 3.653465346534654e-05, + "loss": 1.2188, + "step": 370 + }, + { + "epoch": 0.6026962727993656, + "grad_norm": 5.777859827384273, + "learning_rate": 3.7524752475247526e-05, + "loss": 1.1884, + "step": 380 + }, + { + "epoch": 0.6185567010309279, + "grad_norm": 6.4063926115137875, + "learning_rate": 3.851485148514852e-05, + "loss": 1.1581, + "step": 390 + }, + { + "epoch": 0.63441712926249, + "grad_norm": 6.651129225778287, + "learning_rate": 3.950495049504951e-05, + "loss": 1.1047, + "step": 400 + }, + { + "epoch": 0.6502775574940524, + "grad_norm": 5.6870828905793065, + "learning_rate": 4.04950495049505e-05, + "loss": 1.0823, + "step": 410 + }, + { + "epoch": 0.6661379857256146, + "grad_norm": 5.667978471416249, + "learning_rate": 4.1485148514851486e-05, + "loss": 1.0494, + "step": 420 + }, + { + "epoch": 0.6819984139571769, + "grad_norm": 4.237807026037709, + "learning_rate": 4.2475247524752474e-05, + "loss": 1.0155, + "step": 430 + }, + { + "epoch": 0.6978588421887391, + "grad_norm": 4.255500663396079, + "learning_rate": 4.346534653465347e-05, + "loss": 0.9875, + "step": 440 + }, + { + "epoch": 0.7137192704203014, + "grad_norm": 5.185306420438398, + "learning_rate": 4.445544554455446e-05, + "loss": 0.9639, + "step": 450 + }, + { + "epoch": 0.7295796986518636, + "grad_norm": 5.272706809075813, + "learning_rate": 4.544554455445545e-05, + "loss": 0.9387, + "step": 460 + }, + { + "epoch": 0.7454401268834259, + "grad_norm": 6.077622808418557, + "learning_rate": 4.643564356435644e-05, + "loss": 0.921, + "step": 470 + }, + { + "epoch": 0.7613005551149881, + "grad_norm": 4.6482898197595235, + "learning_rate": 4.742574257425743e-05, + "loss": 0.8856, + "step": 480 + }, + { + "epoch": 0.7771609833465504, + "grad_norm": 5.2219845727325955, + "learning_rate": 4.841584158415842e-05, + "loss": 0.8562, + "step": 490 + }, + { + "epoch": 0.7930214115781126, + "grad_norm": 5.064442866871307, + "learning_rate": 4.9405940594059405e-05, + "loss": 0.8393, + "step": 500 + }, + { + "epoch": 0.8088818398096749, + "grad_norm": 4.238944041111026, + "learning_rate": 5.03960396039604e-05, + "loss": 0.8133, + "step": 510 + }, + { + "epoch": 0.8247422680412371, + "grad_norm": 5.703536274698318, + "learning_rate": 5.138613861386139e-05, + "loss": 0.8012, + "step": 520 + }, + { + "epoch": 0.8406026962727994, + "grad_norm": 5.296419707622884, + "learning_rate": 5.2376237623762384e-05, + "loss": 0.767, + "step": 530 + }, + { + "epoch": 0.8564631245043616, + "grad_norm": 5.699284324132107, + "learning_rate": 5.336633663366337e-05, + "loss": 0.7484, + "step": 540 + }, + { + "epoch": 0.8723235527359239, + "grad_norm": 6.425639332496651, + "learning_rate": 5.435643564356436e-05, + "loss": 0.7292, + "step": 550 + }, + { + "epoch": 0.8881839809674861, + "grad_norm": 5.045854763956847, + "learning_rate": 5.5346534653465355e-05, + "loss": 0.6976, + "step": 560 + }, + { + "epoch": 0.9040444091990484, + "grad_norm": 5.0868187480748315, + "learning_rate": 5.633663366336634e-05, + "loss": 0.6719, + "step": 570 + }, + { + "epoch": 0.9199048374306106, + "grad_norm": 6.760967897420004, + "learning_rate": 5.7326732673267325e-05, + "loss": 0.6573, + "step": 580 + }, + { + "epoch": 0.9357652656621729, + "grad_norm": 4.427355231592498, + "learning_rate": 5.831683168316831e-05, + "loss": 0.6471, + "step": 590 + }, + { + "epoch": 0.9516256938937351, + "grad_norm": 3.9839664702745634, + "learning_rate": 5.930693069306931e-05, + "loss": 0.6221, + "step": 600 + }, + { + "epoch": 0.9674861221252974, + "grad_norm": 3.2513703063334196, + "learning_rate": 6.02970297029703e-05, + "loss": 0.5913, + "step": 610 + }, + { + "epoch": 0.9833465503568596, + "grad_norm": 3.485916478339426, + "learning_rate": 6.128712871287129e-05, + "loss": 0.5775, + "step": 620 + }, + { + "epoch": 0.9992069785884219, + "grad_norm": 4.08651811102019, + "learning_rate": 6.227722772277228e-05, + "loss": 0.5598, + "step": 630 + }, + { + "epoch": 1.014274385408406, + "grad_norm": 3.438814479162103, + "learning_rate": 6.326732673267327e-05, + "loss": 0.5349, + "step": 640 + }, + { + "epoch": 1.0301348136399682, + "grad_norm": 4.44087122057166, + "learning_rate": 6.425742574257426e-05, + "loss": 0.5194, + "step": 650 + }, + { + "epoch": 1.0459952418715306, + "grad_norm": 3.5256251500611384, + "learning_rate": 6.524752475247526e-05, + "loss": 0.506, + "step": 660 + }, + { + "epoch": 1.0618556701030928, + "grad_norm": 3.422611919607941, + "learning_rate": 6.623762376237625e-05, + "loss": 0.4861, + "step": 670 + }, + { + "epoch": 1.077716098334655, + "grad_norm": 3.133948265271591, + "learning_rate": 6.722772277227723e-05, + "loss": 0.4602, + "step": 680 + }, + { + "epoch": 1.0935765265662174, + "grad_norm": 3.6082454324392534, + "learning_rate": 6.821782178217822e-05, + "loss": 0.4496, + "step": 690 + }, + { + "epoch": 1.1094369547977796, + "grad_norm": 2.9254028901121027, + "learning_rate": 6.920792079207921e-05, + "loss": 0.4271, + "step": 700 + }, + { + "epoch": 1.1252973830293418, + "grad_norm": 2.873534663709151, + "learning_rate": 7.01980198019802e-05, + "loss": 0.408, + "step": 710 + }, + { + "epoch": 1.141157811260904, + "grad_norm": 2.4876771741690122, + "learning_rate": 7.118811881188119e-05, + "loss": 0.3908, + "step": 720 + }, + { + "epoch": 1.1570182394924662, + "grad_norm": 2.7097832052411737, + "learning_rate": 7.217821782178218e-05, + "loss": 0.382, + "step": 730 + }, + { + "epoch": 1.1728786677240286, + "grad_norm": 2.9660438369371884, + "learning_rate": 7.316831683168316e-05, + "loss": 0.3582, + "step": 740 + }, + { + "epoch": 1.1887390959555908, + "grad_norm": 2.5000910168051895, + "learning_rate": 7.415841584158417e-05, + "loss": 0.3497, + "step": 750 + }, + { + "epoch": 1.204599524187153, + "grad_norm": 2.633285101346317, + "learning_rate": 7.514851485148515e-05, + "loss": 0.3334, + "step": 760 + }, + { + "epoch": 1.2204599524187154, + "grad_norm": 2.60732404875358, + "learning_rate": 7.613861386138614e-05, + "loss": 0.3184, + "step": 770 + }, + { + "epoch": 1.2363203806502776, + "grad_norm": 2.0629946739238005, + "learning_rate": 7.712871287128713e-05, + "loss": 0.3067, + "step": 780 + }, + { + "epoch": 1.2521808088818398, + "grad_norm": 2.4100126998703253, + "learning_rate": 7.811881188118812e-05, + "loss": 0.2875, + "step": 790 + }, + { + "epoch": 1.268041237113402, + "grad_norm": 2.798339817744955, + "learning_rate": 7.910891089108912e-05, + "loss": 0.277, + "step": 800 + }, + { + "epoch": 1.2839016653449642, + "grad_norm": 2.806131384717819, + "learning_rate": 8.009900990099011e-05, + "loss": 0.2701, + "step": 810 + }, + { + "epoch": 1.2997620935765266, + "grad_norm": 1.9986997398098798, + "learning_rate": 8.10891089108911e-05, + "loss": 0.2559, + "step": 820 + }, + { + "epoch": 1.3156225218080888, + "grad_norm": 2.359590651524235, + "learning_rate": 8.207920792079207e-05, + "loss": 0.2419, + "step": 830 + }, + { + "epoch": 1.331482950039651, + "grad_norm": 2.331111900194689, + "learning_rate": 8.306930693069307e-05, + "loss": 0.2294, + "step": 840 + }, + { + "epoch": 1.3473433782712134, + "grad_norm": 1.8818075156301755, + "learning_rate": 8.405940594059406e-05, + "loss": 0.2141, + "step": 850 + }, + { + "epoch": 1.3632038065027756, + "grad_norm": 1.634132799610872, + "learning_rate": 8.504950495049505e-05, + "loss": 0.2069, + "step": 860 + }, + { + "epoch": 1.3790642347343378, + "grad_norm": 1.5120836080111373, + "learning_rate": 8.603960396039604e-05, + "loss": 0.1966, + "step": 870 + }, + { + "epoch": 1.3949246629659, + "grad_norm": 1.6982638535321994, + "learning_rate": 8.702970297029703e-05, + "loss": 0.1882, + "step": 880 + }, + { + "epoch": 1.4107850911974622, + "grad_norm": 1.7108735595543574, + "learning_rate": 8.801980198019803e-05, + "loss": 0.1779, + "step": 890 + }, + { + "epoch": 1.4266455194290246, + "grad_norm": 1.7296122709014143, + "learning_rate": 8.900990099009902e-05, + "loss": 0.1676, + "step": 900 + }, + { + "epoch": 1.4425059476605868, + "grad_norm": 1.8189109064576499, + "learning_rate": 9e-05, + "loss": 0.1601, + "step": 910 + }, + { + "epoch": 1.458366375892149, + "grad_norm": 2.095375586550601, + "learning_rate": 9.0990099009901e-05, + "loss": 0.1499, + "step": 920 + }, + { + "epoch": 1.4742268041237114, + "grad_norm": 2.6273043704027312, + "learning_rate": 9.198019801980198e-05, + "loss": 0.1455, + "step": 930 + }, + { + "epoch": 1.4900872323552736, + "grad_norm": 1.1840652988818152, + "learning_rate": 9.297029702970298e-05, + "loss": 0.1434, + "step": 940 + }, + { + "epoch": 1.5059476605868358, + "grad_norm": 1.102318462015262, + "learning_rate": 9.396039603960396e-05, + "loss": 0.1322, + "step": 950 + }, + { + "epoch": 1.521808088818398, + "grad_norm": 1.249035070323803, + "learning_rate": 9.495049504950495e-05, + "loss": 0.1232, + "step": 960 + }, + { + "epoch": 1.5376685170499602, + "grad_norm": 0.8585709469514784, + "learning_rate": 9.594059405940594e-05, + "loss": 0.1175, + "step": 970 + }, + { + "epoch": 1.5535289452815226, + "grad_norm": 0.959556178843799, + "learning_rate": 9.693069306930694e-05, + "loss": 0.1119, + "step": 980 + }, + { + "epoch": 1.5693893735130848, + "grad_norm": 0.9203530254113947, + "learning_rate": 9.792079207920793e-05, + "loss": 0.1083, + "step": 990 + }, + { + "epoch": 1.5852498017446472, + "grad_norm": 0.8563247229051851, + "learning_rate": 9.891089108910891e-05, + "loss": 0.1009, + "step": 1000 + }, + { + "epoch": 1.6011102299762094, + "grad_norm": 1.4810040972537613, + "learning_rate": 9.99009900990099e-05, + "loss": 0.0954, + "step": 1010 + }, + { + "epoch": 1.6169706582077716, + "grad_norm": 0.6927338195573747, + "learning_rate": 9.999975790882544e-05, + "loss": 0.0907, + "step": 1020 + }, + { + "epoch": 1.6328310864393338, + "grad_norm": 0.6353525161914889, + "learning_rate": 9.999892105345399e-05, + "loss": 0.0856, + "step": 1030 + }, + { + "epoch": 1.648691514670896, + "grad_norm": 1.0066185794390339, + "learning_rate": 9.999748645510806e-05, + "loss": 0.0904, + "step": 1040 + }, + { + "epoch": 1.6645519429024582, + "grad_norm": 0.5499242841223613, + "learning_rate": 9.999545413093843e-05, + "loss": 0.0817, + "step": 1050 + }, + { + "epoch": 1.6804123711340206, + "grad_norm": 0.6026955934647209, + "learning_rate": 9.999282410524176e-05, + "loss": 0.076, + "step": 1060 + }, + { + "epoch": 1.6962727993655828, + "grad_norm": 0.5189336537822474, + "learning_rate": 9.998959640946033e-05, + "loss": 0.0706, + "step": 1070 + }, + { + "epoch": 1.7121332275971453, + "grad_norm": 0.5108915419488772, + "learning_rate": 9.998577108218162e-05, + "loss": 0.0685, + "step": 1080 + }, + { + "epoch": 1.7279936558287075, + "grad_norm": 0.4759581343462929, + "learning_rate": 9.998134816913788e-05, + "loss": 0.0642, + "step": 1090 + }, + { + "epoch": 1.7438540840602696, + "grad_norm": 0.44750114998963825, + "learning_rate": 9.997632772320553e-05, + "loss": 0.0615, + "step": 1100 + }, + { + "epoch": 1.7597145122918318, + "grad_norm": 0.4682540236602841, + "learning_rate": 9.997070980440462e-05, + "loss": 0.0587, + "step": 1110 + }, + { + "epoch": 1.775574940523394, + "grad_norm": 0.46100974574978826, + "learning_rate": 9.9964494479898e-05, + "loss": 0.0571, + "step": 1120 + }, + { + "epoch": 1.7914353687549562, + "grad_norm": 0.4802771820307019, + "learning_rate": 9.995768182399062e-05, + "loss": 0.0542, + "step": 1130 + }, + { + "epoch": 1.8072957969865187, + "grad_norm": 0.402517941066615, + "learning_rate": 9.995027191812859e-05, + "loss": 0.0526, + "step": 1140 + }, + { + "epoch": 1.8231562252180809, + "grad_norm": 0.36529022302452335, + "learning_rate": 9.99422648508982e-05, + "loss": 0.0502, + "step": 1150 + }, + { + "epoch": 1.8390166534496433, + "grad_norm": 0.43348415026389686, + "learning_rate": 9.993366071802484e-05, + "loss": 0.0478, + "step": 1160 + }, + { + "epoch": 1.8548770816812055, + "grad_norm": 0.32832066422044764, + "learning_rate": 9.9924459622372e-05, + "loss": 0.0469, + "step": 1170 + }, + { + "epoch": 1.8707375099127677, + "grad_norm": 0.36783738330763543, + "learning_rate": 9.991466167393981e-05, + "loss": 0.0458, + "step": 1180 + }, + { + "epoch": 1.8865979381443299, + "grad_norm": 2.3188302609712563, + "learning_rate": 9.99042669898639e-05, + "loss": 0.0459, + "step": 1190 + }, + { + "epoch": 1.902458366375892, + "grad_norm": 0.5436195698507995, + "learning_rate": 9.989327569441394e-05, + "loss": 0.0464, + "step": 1200 + }, + { + "epoch": 1.9183187946074542, + "grad_norm": 0.3263229164097963, + "learning_rate": 9.988168791899218e-05, + "loss": 0.0436, + "step": 1210 + }, + { + "epoch": 1.9341792228390167, + "grad_norm": 0.3022710473853306, + "learning_rate": 9.986950380213182e-05, + "loss": 0.0416, + "step": 1220 + }, + { + "epoch": 1.9500396510705789, + "grad_norm": 0.30154108928053425, + "learning_rate": 9.98567234894954e-05, + "loss": 0.0406, + "step": 1230 + }, + { + "epoch": 1.9659000793021413, + "grad_norm": 0.3196021212235642, + "learning_rate": 9.984334713387308e-05, + "loss": 0.0396, + "step": 1240 + }, + { + "epoch": 1.9817605075337035, + "grad_norm": 0.3236037547731566, + "learning_rate": 9.982937489518075e-05, + "loss": 0.0389, + "step": 1250 + }, + { + "epoch": 1.9976209357652657, + "grad_norm": 0.3711446498733101, + "learning_rate": 9.981480694045816e-05, + "loss": 0.038, + "step": 1260 + }, + { + "epoch": 2.0126883425852498, + "grad_norm": 0.3063522932516602, + "learning_rate": 9.979964344386692e-05, + "loss": 0.0359, + "step": 1270 + }, + { + "epoch": 2.028548770816812, + "grad_norm": 0.30361076290584094, + "learning_rate": 9.978388458668842e-05, + "loss": 0.0343, + "step": 1280 + }, + { + "epoch": 2.044409199048374, + "grad_norm": 0.258513333686689, + "learning_rate": 9.976753055732166e-05, + "loss": 0.0338, + "step": 1290 + }, + { + "epoch": 2.0602696272799363, + "grad_norm": 0.25186209885924893, + "learning_rate": 9.975058155128097e-05, + "loss": 0.0332, + "step": 1300 + }, + { + "epoch": 2.076130055511499, + "grad_norm": 0.2667716434965371, + "learning_rate": 9.97330377711937e-05, + "loss": 0.0325, + "step": 1310 + }, + { + "epoch": 2.091990483743061, + "grad_norm": 0.25831881157696407, + "learning_rate": 9.97148994267978e-05, + "loss": 0.0322, + "step": 1320 + }, + { + "epoch": 2.1078509119746234, + "grad_norm": 0.2422521939116205, + "learning_rate": 9.969616673493932e-05, + "loss": 0.0312, + "step": 1330 + }, + { + "epoch": 2.1237113402061856, + "grad_norm": 0.25167325235563537, + "learning_rate": 9.96768399195698e-05, + "loss": 0.0311, + "step": 1340 + }, + { + "epoch": 2.1395717684377478, + "grad_norm": 0.2648060208843995, + "learning_rate": 9.965691921174357e-05, + "loss": 0.0308, + "step": 1350 + }, + { + "epoch": 2.15543219666931, + "grad_norm": 0.26050946392333785, + "learning_rate": 9.963640484961505e-05, + "loss": 0.03, + "step": 1360 + }, + { + "epoch": 2.171292624900872, + "grad_norm": 0.2384998360178266, + "learning_rate": 9.961529707843583e-05, + "loss": 0.0297, + "step": 1370 + }, + { + "epoch": 2.187153053132435, + "grad_norm": 0.25739858905812224, + "learning_rate": 9.95935961505518e-05, + "loss": 0.029, + "step": 1380 + }, + { + "epoch": 2.203013481363997, + "grad_norm": 0.23794124732997246, + "learning_rate": 9.95713023254001e-05, + "loss": 0.0286, + "step": 1390 + }, + { + "epoch": 2.218873909595559, + "grad_norm": 0.2301971290512307, + "learning_rate": 9.954841586950606e-05, + "loss": 0.0286, + "step": 1400 + }, + { + "epoch": 2.2347343378271214, + "grad_norm": 0.2290063026328564, + "learning_rate": 9.95249370564799e-05, + "loss": 0.0276, + "step": 1410 + }, + { + "epoch": 2.2505947660586836, + "grad_norm": 0.23440055002640264, + "learning_rate": 9.95008661670136e-05, + "loss": 0.0273, + "step": 1420 + }, + { + "epoch": 2.266455194290246, + "grad_norm": 0.34512157189164844, + "learning_rate": 9.947620348887749e-05, + "loss": 0.0266, + "step": 1430 + }, + { + "epoch": 2.282315622521808, + "grad_norm": 0.25726085487236555, + "learning_rate": 9.945094931691678e-05, + "loss": 0.0268, + "step": 1440 + }, + { + "epoch": 2.29817605075337, + "grad_norm": 0.3637610324377928, + "learning_rate": 9.942510395304805e-05, + "loss": 0.0259, + "step": 1450 + }, + { + "epoch": 2.3140364789849324, + "grad_norm": 1.1472903509254984, + "learning_rate": 9.939866770625567e-05, + "loss": 0.0275, + "step": 1460 + }, + { + "epoch": 2.329896907216495, + "grad_norm": 2.5053341106388856, + "learning_rate": 9.937164089258809e-05, + "loss": 0.0304, + "step": 1470 + }, + { + "epoch": 2.345757335448057, + "grad_norm": 1.0767870929445287, + "learning_rate": 9.9344023835154e-05, + "loss": 0.0454, + "step": 1480 + }, + { + "epoch": 2.3616177636796194, + "grad_norm": 0.43278952742259996, + "learning_rate": 9.931581686411859e-05, + "loss": 0.0362, + "step": 1490 + }, + { + "epoch": 2.3774781919111816, + "grad_norm": 0.2745476890309345, + "learning_rate": 9.928702031669954e-05, + "loss": 0.0294, + "step": 1500 + }, + { + "epoch": 2.393338620142744, + "grad_norm": 0.23151635915708174, + "learning_rate": 9.92576345371629e-05, + "loss": 0.0267, + "step": 1510 + }, + { + "epoch": 2.409199048374306, + "grad_norm": 0.2528640123475777, + "learning_rate": 9.922765987681917e-05, + "loss": 0.0254, + "step": 1520 + }, + { + "epoch": 2.425059476605868, + "grad_norm": 0.22322528832884078, + "learning_rate": 9.919709669401891e-05, + "loss": 0.0241, + "step": 1530 + }, + { + "epoch": 2.440919904837431, + "grad_norm": 0.25767429518855794, + "learning_rate": 9.916594535414856e-05, + "loss": 0.0236, + "step": 1540 + }, + { + "epoch": 2.456780333068993, + "grad_norm": 0.20859014169787804, + "learning_rate": 9.913420622962606e-05, + "loss": 0.0231, + "step": 1550 + }, + { + "epoch": 2.472640761300555, + "grad_norm": 0.20097276712386156, + "learning_rate": 9.910187969989635e-05, + "loss": 0.0227, + "step": 1560 + }, + { + "epoch": 2.4885011895321174, + "grad_norm": 0.19502715852546781, + "learning_rate": 9.90689661514269e-05, + "loss": 0.0223, + "step": 1570 + }, + { + "epoch": 2.5043616177636796, + "grad_norm": 0.18926852564659236, + "learning_rate": 9.903546597770305e-05, + "loss": 0.0217, + "step": 1580 + }, + { + "epoch": 2.520222045995242, + "grad_norm": 0.18977619403937718, + "learning_rate": 9.900137957922333e-05, + "loss": 0.0213, + "step": 1590 + }, + { + "epoch": 2.536082474226804, + "grad_norm": 0.19098675458895795, + "learning_rate": 9.896670736349462e-05, + "loss": 0.0208, + "step": 1600 + }, + { + "epoch": 2.551942902458366, + "grad_norm": 0.2009242073740019, + "learning_rate": 9.893144974502733e-05, + "loss": 0.0205, + "step": 1610 + }, + { + "epoch": 2.5678033306899284, + "grad_norm": 0.1882743338134484, + "learning_rate": 9.889560714533042e-05, + "loss": 0.0209, + "step": 1620 + }, + { + "epoch": 2.583663758921491, + "grad_norm": 0.20524223857112522, + "learning_rate": 9.885917999290641e-05, + "loss": 0.0202, + "step": 1630 + }, + { + "epoch": 2.5995241871530532, + "grad_norm": 0.17581923873322422, + "learning_rate": 9.882216872324617e-05, + "loss": 0.02, + "step": 1640 + }, + { + "epoch": 2.6153846153846154, + "grad_norm": 0.21569051031935538, + "learning_rate": 9.878457377882376e-05, + "loss": 0.0201, + "step": 1650 + }, + { + "epoch": 2.6312450436161776, + "grad_norm": 0.1891563070911822, + "learning_rate": 9.874639560909117e-05, + "loss": 0.0196, + "step": 1660 + }, + { + "epoch": 2.64710547184774, + "grad_norm": 0.19956230575208342, + "learning_rate": 9.870763467047293e-05, + "loss": 0.019, + "step": 1670 + }, + { + "epoch": 2.662965900079302, + "grad_norm": 0.19799721546110283, + "learning_rate": 9.866829142636058e-05, + "loss": 0.0189, + "step": 1680 + }, + { + "epoch": 2.6788263283108646, + "grad_norm": 0.19531667504057218, + "learning_rate": 9.862836634710724e-05, + "loss": 0.0186, + "step": 1690 + }, + { + "epoch": 2.694686756542427, + "grad_norm": 0.18324610519548817, + "learning_rate": 9.858785991002192e-05, + "loss": 0.0185, + "step": 1700 + }, + { + "epoch": 2.710547184773989, + "grad_norm": 0.19516552428750053, + "learning_rate": 9.854677259936383e-05, + "loss": 0.0179, + "step": 1710 + }, + { + "epoch": 2.7264076130055512, + "grad_norm": 0.20561324061869957, + "learning_rate": 9.85051049063366e-05, + "loss": 0.0182, + "step": 1720 + }, + { + "epoch": 2.7422680412371134, + "grad_norm": 0.1844230160943975, + "learning_rate": 9.846285732908238e-05, + "loss": 0.0177, + "step": 1730 + }, + { + "epoch": 2.7581284694686756, + "grad_norm": 0.1890633342236241, + "learning_rate": 9.842003037267593e-05, + "loss": 0.0174, + "step": 1740 + }, + { + "epoch": 2.773988897700238, + "grad_norm": 0.1661243506045367, + "learning_rate": 9.837662454911855e-05, + "loss": 0.0171, + "step": 1750 + }, + { + "epoch": 2.7898493259318, + "grad_norm": 0.1603089791676492, + "learning_rate": 9.833264037733198e-05, + "loss": 0.0167, + "step": 1760 + }, + { + "epoch": 2.805709754163362, + "grad_norm": 0.198488901552779, + "learning_rate": 9.82880783831521e-05, + "loss": 0.0167, + "step": 1770 + }, + { + "epoch": 2.8215701823949244, + "grad_norm": 0.17631212946503003, + "learning_rate": 9.824293909932285e-05, + "loss": 0.0163, + "step": 1780 + }, + { + "epoch": 2.837430610626487, + "grad_norm": 0.16861160104838285, + "learning_rate": 9.819722306548962e-05, + "loss": 0.0161, + "step": 1790 + }, + { + "epoch": 2.8532910388580492, + "grad_norm": 0.19230483320622835, + "learning_rate": 9.8150930828193e-05, + "loss": 0.0158, + "step": 1800 + }, + { + "epoch": 2.8691514670896114, + "grad_norm": 0.18577920780634982, + "learning_rate": 9.81040629408621e-05, + "loss": 0.0154, + "step": 1810 + }, + { + "epoch": 2.8850118953211736, + "grad_norm": 0.1881494112800731, + "learning_rate": 9.805661996380804e-05, + "loss": 0.0153, + "step": 1820 + }, + { + "epoch": 2.900872323552736, + "grad_norm": 0.17869253917833045, + "learning_rate": 9.800860246421717e-05, + "loss": 0.0151, + "step": 1830 + }, + { + "epoch": 2.916732751784298, + "grad_norm": 0.17742762061027734, + "learning_rate": 9.796001101614434e-05, + "loss": 0.0152, + "step": 1840 + }, + { + "epoch": 2.9325931800158607, + "grad_norm": 0.1830810064948476, + "learning_rate": 9.791084620050602e-05, + "loss": 0.0151, + "step": 1850 + }, + { + "epoch": 2.948453608247423, + "grad_norm": 0.20147364949584862, + "learning_rate": 9.786110860507336e-05, + "loss": 0.0146, + "step": 1860 + }, + { + "epoch": 2.964314036478985, + "grad_norm": 0.16997448856636277, + "learning_rate": 9.781079882446514e-05, + "loss": 0.0145, + "step": 1870 + }, + { + "epoch": 2.9801744647105473, + "grad_norm": 0.1587095515375798, + "learning_rate": 9.775991746014072e-05, + "loss": 0.0139, + "step": 1880 + }, + { + "epoch": 2.9960348929421095, + "grad_norm": 0.1983770059337939, + "learning_rate": 9.770846512039278e-05, + "loss": 0.0143, + "step": 1890 + }, + { + "epoch": 3.0111022997620935, + "grad_norm": 0.15794198566462234, + "learning_rate": 9.765644242034008e-05, + "loss": 0.012, + "step": 1900 + }, + { + "epoch": 3.0269627279936557, + "grad_norm": 0.1739431333762727, + "learning_rate": 9.760384998192011e-05, + "loss": 0.0109, + "step": 1910 + }, + { + "epoch": 3.042823156225218, + "grad_norm": 0.14437644637053978, + "learning_rate": 9.755068843388166e-05, + "loss": 0.0108, + "step": 1920 + }, + { + "epoch": 3.05868358445678, + "grad_norm": 0.1440694296478373, + "learning_rate": 9.749695841177729e-05, + "loss": 0.0104, + "step": 1930 + }, + { + "epoch": 3.0745440126883428, + "grad_norm": 0.14923280981473364, + "learning_rate": 9.744266055795568e-05, + "loss": 0.0103, + "step": 1940 + }, + { + "epoch": 3.090404440919905, + "grad_norm": 0.1518629106831049, + "learning_rate": 9.738779552155409e-05, + "loss": 0.0101, + "step": 1950 + }, + { + "epoch": 3.106264869151467, + "grad_norm": 0.14823832101551415, + "learning_rate": 9.733236395849048e-05, + "loss": 0.01, + "step": 1960 + }, + { + "epoch": 3.1221252973830294, + "grad_norm": 0.14119487110317194, + "learning_rate": 9.727636653145567e-05, + "loss": 0.0103, + "step": 1970 + }, + { + "epoch": 3.1379857256145915, + "grad_norm": 0.14471599107467292, + "learning_rate": 9.721980390990549e-05, + "loss": 0.0099, + "step": 1980 + }, + { + "epoch": 3.1538461538461537, + "grad_norm": 0.16131116027673068, + "learning_rate": 9.716267677005273e-05, + "loss": 0.0098, + "step": 1990 + }, + { + "epoch": 3.169706582077716, + "grad_norm": 0.1524998515238758, + "learning_rate": 9.710498579485907e-05, + "loss": 0.0099, + "step": 2000 + }, + { + "epoch": 3.1855670103092786, + "grad_norm": 0.15197353760818177, + "learning_rate": 9.704673167402687e-05, + "loss": 0.0096, + "step": 2010 + }, + { + "epoch": 3.2014274385408408, + "grad_norm": 0.14615887528821866, + "learning_rate": 9.6987915103991e-05, + "loss": 0.0094, + "step": 2020 + }, + { + "epoch": 3.217287866772403, + "grad_norm": 0.16757614542523475, + "learning_rate": 9.692853678791044e-05, + "loss": 0.0096, + "step": 2030 + }, + { + "epoch": 3.233148295003965, + "grad_norm": 0.1563840538926366, + "learning_rate": 9.686859743565997e-05, + "loss": 0.0095, + "step": 2040 + }, + { + "epoch": 3.2490087232355274, + "grad_norm": 0.15207509365465333, + "learning_rate": 9.680809776382159e-05, + "loss": 0.0095, + "step": 2050 + }, + { + "epoch": 3.2648691514670896, + "grad_norm": 0.1408235208483165, + "learning_rate": 9.674703849567594e-05, + "loss": 0.0095, + "step": 2060 + }, + { + "epoch": 3.2807295796986518, + "grad_norm": 0.13678320296028157, + "learning_rate": 9.668542036119378e-05, + "loss": 0.0094, + "step": 2070 + }, + { + "epoch": 3.296590007930214, + "grad_norm": 0.16636115763625545, + "learning_rate": 9.662324409702715e-05, + "loss": 0.0091, + "step": 2080 + }, + { + "epoch": 3.312450436161776, + "grad_norm": 0.13608061324123708, + "learning_rate": 9.656051044650054e-05, + "loss": 0.0089, + "step": 2090 + }, + { + "epoch": 3.328310864393339, + "grad_norm": 0.1431792557439411, + "learning_rate": 9.649722015960219e-05, + "loss": 0.009, + "step": 2100 + }, + { + "epoch": 3.344171292624901, + "grad_norm": 0.15687389279462885, + "learning_rate": 9.643337399297484e-05, + "loss": 0.009, + "step": 2110 + }, + { + "epoch": 3.360031720856463, + "grad_norm": 0.14839646958379588, + "learning_rate": 9.636897270990696e-05, + "loss": 0.0087, + "step": 2120 + }, + { + "epoch": 3.3758921490880254, + "grad_norm": 0.13745612916138747, + "learning_rate": 9.630401708032343e-05, + "loss": 0.0088, + "step": 2130 + }, + { + "epoch": 3.3917525773195876, + "grad_norm": 0.15771151005385436, + "learning_rate": 9.623850788077645e-05, + "loss": 0.0087, + "step": 2140 + }, + { + "epoch": 3.4076130055511498, + "grad_norm": 0.15078713122784224, + "learning_rate": 9.617244589443619e-05, + "loss": 0.0086, + "step": 2150 + }, + { + "epoch": 3.423473433782712, + "grad_norm": 0.13928148814189775, + "learning_rate": 9.610583191108145e-05, + "loss": 0.0084, + "step": 2160 + }, + { + "epoch": 3.4393338620142746, + "grad_norm": 0.13689800841443125, + "learning_rate": 9.603866672709022e-05, + "loss": 0.0084, + "step": 2170 + }, + { + "epoch": 3.455194290245837, + "grad_norm": 0.13916459530391348, + "learning_rate": 9.597095114543018e-05, + "loss": 0.0083, + "step": 2180 + }, + { + "epoch": 3.471054718477399, + "grad_norm": 0.15261756839835933, + "learning_rate": 9.590268597564906e-05, + "loss": 0.0083, + "step": 2190 + }, + { + "epoch": 3.486915146708961, + "grad_norm": 0.13702247664859923, + "learning_rate": 9.583387203386497e-05, + "loss": 0.0081, + "step": 2200 + }, + { + "epoch": 3.5027755749405234, + "grad_norm": 0.13505439254966808, + "learning_rate": 9.576451014275669e-05, + "loss": 0.0079, + "step": 2210 + }, + { + "epoch": 3.5186360031720856, + "grad_norm": 0.13192729341816917, + "learning_rate": 9.569460113155378e-05, + "loss": 0.0081, + "step": 2220 + }, + { + "epoch": 3.5344964314036478, + "grad_norm": 0.14259485452290382, + "learning_rate": 9.562414583602664e-05, + "loss": 0.0078, + "step": 2230 + }, + { + "epoch": 3.55035685963521, + "grad_norm": 0.1380810668744061, + "learning_rate": 9.555314509847663e-05, + "loss": 0.0079, + "step": 2240 + }, + { + "epoch": 3.566217287866772, + "grad_norm": 0.13586211454614464, + "learning_rate": 9.548159976772592e-05, + "loss": 0.0078, + "step": 2250 + }, + { + "epoch": 3.582077716098335, + "grad_norm": 0.12201458768486265, + "learning_rate": 9.540951069910733e-05, + "loss": 0.0078, + "step": 2260 + }, + { + "epoch": 3.597938144329897, + "grad_norm": 0.13844954643236915, + "learning_rate": 9.533687875445414e-05, + "loss": 0.0078, + "step": 2270 + }, + { + "epoch": 3.613798572561459, + "grad_norm": 0.1425114414070632, + "learning_rate": 9.526370480208979e-05, + "loss": 0.0076, + "step": 2280 + }, + { + "epoch": 3.6296590007930214, + "grad_norm": 0.2155707350379239, + "learning_rate": 9.518998971681748e-05, + "loss": 0.008, + "step": 2290 + }, + { + "epoch": 3.6455194290245836, + "grad_norm": 0.13584009980221787, + "learning_rate": 9.511573437990968e-05, + "loss": 0.0078, + "step": 2300 + }, + { + "epoch": 3.661379857256146, + "grad_norm": 0.12951200911833496, + "learning_rate": 9.50409396790977e-05, + "loss": 0.0076, + "step": 2310 + }, + { + "epoch": 3.6772402854877084, + "grad_norm": 0.13780278708132593, + "learning_rate": 9.496560650856097e-05, + "loss": 0.0076, + "step": 2320 + }, + { + "epoch": 3.6931007137192706, + "grad_norm": 0.14078358511236871, + "learning_rate": 9.488973576891638e-05, + "loss": 0.0075, + "step": 2330 + }, + { + "epoch": 3.708961141950833, + "grad_norm": 0.13731645588921648, + "learning_rate": 9.481332836720756e-05, + "loss": 0.0073, + "step": 2340 + }, + { + "epoch": 3.724821570182395, + "grad_norm": 0.14229326907743092, + "learning_rate": 9.473638521689393e-05, + "loss": 0.0073, + "step": 2350 + }, + { + "epoch": 3.740681998413957, + "grad_norm": 0.1393168463393868, + "learning_rate": 9.465890723783992e-05, + "loss": 0.0074, + "step": 2360 + }, + { + "epoch": 3.7565424266455194, + "grad_norm": 0.12400505280500868, + "learning_rate": 9.458089535630387e-05, + "loss": 0.007, + "step": 2370 + }, + { + "epoch": 3.7724028548770816, + "grad_norm": 0.14228035248732068, + "learning_rate": 9.4502350504927e-05, + "loss": 0.0074, + "step": 2380 + }, + { + "epoch": 3.788263283108644, + "grad_norm": 0.12229258730545413, + "learning_rate": 9.44232736227222e-05, + "loss": 0.0071, + "step": 2390 + }, + { + "epoch": 3.804123711340206, + "grad_norm": 0.11833650135887812, + "learning_rate": 9.434366565506292e-05, + "loss": 0.0069, + "step": 2400 + }, + { + "epoch": 3.819984139571768, + "grad_norm": 0.1459374639549126, + "learning_rate": 9.426352755367174e-05, + "loss": 0.0068, + "step": 2410 + }, + { + "epoch": 3.835844567803331, + "grad_norm": 0.11935347161825668, + "learning_rate": 9.418286027660908e-05, + "loss": 0.0069, + "step": 2420 + }, + { + "epoch": 3.851704996034893, + "grad_norm": 0.13102350817072028, + "learning_rate": 9.410166478826172e-05, + "loss": 0.0067, + "step": 2430 + }, + { + "epoch": 3.867565424266455, + "grad_norm": 0.1299931124166466, + "learning_rate": 9.401994205933127e-05, + "loss": 0.0069, + "step": 2440 + }, + { + "epoch": 3.8834258524980174, + "grad_norm": 0.12080186454827395, + "learning_rate": 9.393769306682253e-05, + "loss": 0.0064, + "step": 2450 + }, + { + "epoch": 3.8992862807295796, + "grad_norm": 0.13858427536099355, + "learning_rate": 9.385491879403189e-05, + "loss": 0.0066, + "step": 2460 + }, + { + "epoch": 3.915146708961142, + "grad_norm": 0.12403182365468098, + "learning_rate": 9.377162023053549e-05, + "loss": 0.0065, + "step": 2470 + }, + { + "epoch": 3.9310071371927044, + "grad_norm": 0.1210866209740134, + "learning_rate": 9.368779837217745e-05, + "loss": 0.0066, + "step": 2480 + }, + { + "epoch": 3.9468675654242666, + "grad_norm": 0.1318070654135075, + "learning_rate": 9.360345422105793e-05, + "loss": 0.0067, + "step": 2490 + }, + { + "epoch": 3.962727993655829, + "grad_norm": 0.14258951615104107, + "learning_rate": 9.351858878552116e-05, + "loss": 0.0065, + "step": 2500 + }, + { + "epoch": 3.978588421887391, + "grad_norm": 0.1204057759444754, + "learning_rate": 9.343320308014341e-05, + "loss": 0.0064, + "step": 2510 + }, + { + "epoch": 3.9944488501189532, + "grad_norm": 0.11839457225594427, + "learning_rate": 9.334729812572078e-05, + "loss": 0.0064, + "step": 2520 + }, + { + "epoch": 4.009516256938937, + "grad_norm": 0.08540322077120217, + "learning_rate": 9.326087494925714e-05, + "loss": 0.0053, + "step": 2530 + }, + { + "epoch": 4.0253766851704995, + "grad_norm": 0.10112676530125538, + "learning_rate": 9.317393458395168e-05, + "loss": 0.0043, + "step": 2540 + }, + { + "epoch": 4.041237113402062, + "grad_norm": 0.12529764169062654, + "learning_rate": 9.308647806918669e-05, + "loss": 0.0039, + "step": 2550 + }, + { + "epoch": 4.057097541633624, + "grad_norm": 0.0955067824938099, + "learning_rate": 9.299850645051506e-05, + "loss": 0.0038, + "step": 2560 + }, + { + "epoch": 4.072957969865186, + "grad_norm": 0.10874795377273988, + "learning_rate": 9.291002077964788e-05, + "loss": 0.0038, + "step": 2570 + }, + { + "epoch": 4.088818398096748, + "grad_norm": 0.08668043825795246, + "learning_rate": 9.282102211444168e-05, + "loss": 0.0036, + "step": 2580 + }, + { + "epoch": 4.1046788263283105, + "grad_norm": 0.09839554266178921, + "learning_rate": 9.273151151888598e-05, + "loss": 0.0035, + "step": 2590 + }, + { + "epoch": 4.120539254559873, + "grad_norm": 0.1078946853176116, + "learning_rate": 9.264149006309047e-05, + "loss": 0.0036, + "step": 2600 + }, + { + "epoch": 4.136399682791436, + "grad_norm": 0.08794520025183818, + "learning_rate": 9.255095882327223e-05, + "loss": 0.0038, + "step": 2610 + }, + { + "epoch": 4.152260111022998, + "grad_norm": 0.09544438739347859, + "learning_rate": 9.245991888174285e-05, + "loss": 0.0035, + "step": 2620 + }, + { + "epoch": 4.16812053925456, + "grad_norm": 0.09704914866906847, + "learning_rate": 9.236837132689554e-05, + "loss": 0.0036, + "step": 2630 + }, + { + "epoch": 4.183980967486122, + "grad_norm": 0.09413701678761836, + "learning_rate": 9.227631725319206e-05, + "loss": 0.0036, + "step": 2640 + }, + { + "epoch": 4.1998413957176846, + "grad_norm": 0.1249674941120182, + "learning_rate": 9.21837577611497e-05, + "loss": 0.0036, + "step": 2650 + }, + { + "epoch": 4.215701823949247, + "grad_norm": 0.09320709700144189, + "learning_rate": 9.209069395732802e-05, + "loss": 0.0035, + "step": 2660 + }, + { + "epoch": 4.231562252180809, + "grad_norm": 0.09816492498272372, + "learning_rate": 9.199712695431577e-05, + "loss": 0.0036, + "step": 2670 + }, + { + "epoch": 4.247422680412371, + "grad_norm": 0.10786908595610735, + "learning_rate": 9.190305787071742e-05, + "loss": 0.0035, + "step": 2680 + }, + { + "epoch": 4.263283108643933, + "grad_norm": 0.08480594239364594, + "learning_rate": 9.180848783113996e-05, + "loss": 0.0034, + "step": 2690 + }, + { + "epoch": 4.2791435368754955, + "grad_norm": 0.09241037627689046, + "learning_rate": 9.171341796617927e-05, + "loss": 0.0035, + "step": 2700 + }, + { + "epoch": 4.295003965107058, + "grad_norm": 0.10223555985966222, + "learning_rate": 9.161784941240681e-05, + "loss": 0.0035, + "step": 2710 + }, + { + "epoch": 4.31086439333862, + "grad_norm": 0.12709742861011547, + "learning_rate": 9.152178331235581e-05, + "loss": 0.0036, + "step": 2720 + }, + { + "epoch": 4.326724821570182, + "grad_norm": 0.10916240205541296, + "learning_rate": 9.142522081450783e-05, + "loss": 0.0036, + "step": 2730 + }, + { + "epoch": 4.342585249801744, + "grad_norm": 0.1004166247257912, + "learning_rate": 9.132816307327886e-05, + "loss": 0.0035, + "step": 2740 + }, + { + "epoch": 4.3584456780333065, + "grad_norm": 0.1053220173193627, + "learning_rate": 9.123061124900556e-05, + "loss": 0.0036, + "step": 2750 + }, + { + "epoch": 4.37430610626487, + "grad_norm": 0.10781630634201299, + "learning_rate": 9.113256650793149e-05, + "loss": 0.0038, + "step": 2760 + }, + { + "epoch": 4.390166534496432, + "grad_norm": 0.15441377221263897, + "learning_rate": 9.103403002219305e-05, + "loss": 0.0038, + "step": 2770 + }, + { + "epoch": 4.406026962727994, + "grad_norm": 0.1145440846539534, + "learning_rate": 9.093500296980548e-05, + "loss": 0.0037, + "step": 2780 + }, + { + "epoch": 4.421887390959556, + "grad_norm": 0.10014929941903501, + "learning_rate": 9.083548653464884e-05, + "loss": 0.0036, + "step": 2790 + }, + { + "epoch": 4.437747819191118, + "grad_norm": 0.10575652642767745, + "learning_rate": 9.073548190645378e-05, + "loss": 0.0036, + "step": 2800 + }, + { + "epoch": 4.453608247422681, + "grad_norm": 0.09967090608925282, + "learning_rate": 9.063499028078741e-05, + "loss": 0.0035, + "step": 2810 + }, + { + "epoch": 4.469468675654243, + "grad_norm": 0.08849883488694847, + "learning_rate": 9.053401285903889e-05, + "loss": 0.0035, + "step": 2820 + }, + { + "epoch": 4.485329103885805, + "grad_norm": 0.10136493808207836, + "learning_rate": 9.043255084840517e-05, + "loss": 0.0037, + "step": 2830 + }, + { + "epoch": 4.501189532117367, + "grad_norm": 0.08898247112970178, + "learning_rate": 9.03306054618765e-05, + "loss": 0.0035, + "step": 2840 + }, + { + "epoch": 4.517049960348929, + "grad_norm": 0.09208854129111659, + "learning_rate": 9.022817791822198e-05, + "loss": 0.0036, + "step": 2850 + }, + { + "epoch": 4.532910388580492, + "grad_norm": 0.09961222712616437, + "learning_rate": 9.012526944197488e-05, + "loss": 0.0035, + "step": 2860 + }, + { + "epoch": 4.548770816812054, + "grad_norm": 0.10356990211947537, + "learning_rate": 9.002188126341815e-05, + "loss": 0.0035, + "step": 2870 + }, + { + "epoch": 4.564631245043616, + "grad_norm": 0.10880605758798846, + "learning_rate": 8.991801461856961e-05, + "loss": 0.0034, + "step": 2880 + }, + { + "epoch": 4.580491673275178, + "grad_norm": 0.09025089948749608, + "learning_rate": 8.98136707491672e-05, + "loss": 0.0034, + "step": 2890 + }, + { + "epoch": 4.59635210150674, + "grad_norm": 0.0948081759012918, + "learning_rate": 8.970885090265413e-05, + "loss": 0.0036, + "step": 2900 + }, + { + "epoch": 4.6122125297383025, + "grad_norm": 0.10862619774623498, + "learning_rate": 8.960355633216397e-05, + "loss": 0.0035, + "step": 2910 + }, + { + "epoch": 4.628072957969865, + "grad_norm": 0.0968778686048629, + "learning_rate": 8.94977882965057e-05, + "loss": 0.0036, + "step": 2920 + }, + { + "epoch": 4.643933386201428, + "grad_norm": 0.09951637896006683, + "learning_rate": 8.939154806014857e-05, + "loss": 0.0035, + "step": 2930 + }, + { + "epoch": 4.65979381443299, + "grad_norm": 0.10284890386755276, + "learning_rate": 8.928483689320714e-05, + "loss": 0.0035, + "step": 2940 + }, + { + "epoch": 4.675654242664552, + "grad_norm": 0.12021152095203591, + "learning_rate": 8.917765607142594e-05, + "loss": 0.0036, + "step": 2950 + }, + { + "epoch": 4.691514670896114, + "grad_norm": 0.10094042369158501, + "learning_rate": 8.907000687616423e-05, + "loss": 0.0037, + "step": 2960 + }, + { + "epoch": 4.707375099127677, + "grad_norm": 0.12159048686884875, + "learning_rate": 8.896189059438083e-05, + "loss": 0.0035, + "step": 2970 + }, + { + "epoch": 4.723235527359239, + "grad_norm": 0.09516680996052305, + "learning_rate": 8.885330851861858e-05, + "loss": 0.0035, + "step": 2980 + }, + { + "epoch": 4.739095955590801, + "grad_norm": 0.10722002700044025, + "learning_rate": 8.874426194698895e-05, + "loss": 0.0036, + "step": 2990 + }, + { + "epoch": 4.754956383822363, + "grad_norm": 0.09312639306042723, + "learning_rate": 8.86347521831565e-05, + "loss": 0.0034, + "step": 3000 + }, + { + "epoch": 4.770816812053925, + "grad_norm": 0.10519488774048268, + "learning_rate": 8.852478053632335e-05, + "loss": 0.0035, + "step": 3010 + }, + { + "epoch": 4.786677240285488, + "grad_norm": 0.09594602135358007, + "learning_rate": 8.841434832121345e-05, + "loss": 0.0036, + "step": 3020 + }, + { + "epoch": 4.80253766851705, + "grad_norm": 0.08866775532281909, + "learning_rate": 8.830345685805688e-05, + "loss": 0.0035, + "step": 3030 + }, + { + "epoch": 4.818398096748612, + "grad_norm": 0.09703766154369026, + "learning_rate": 8.819210747257412e-05, + "loss": 0.0033, + "step": 3040 + }, + { + "epoch": 4.834258524980174, + "grad_norm": 0.09385784638715457, + "learning_rate": 8.808030149596017e-05, + "loss": 0.0034, + "step": 3050 + }, + { + "epoch": 4.850118953211736, + "grad_norm": 0.10558045718553939, + "learning_rate": 8.796804026486858e-05, + "loss": 0.0034, + "step": 3060 + }, + { + "epoch": 4.8659793814432994, + "grad_norm": 0.11749294827599736, + "learning_rate": 8.785532512139558e-05, + "loss": 0.0035, + "step": 3070 + }, + { + "epoch": 4.881839809674862, + "grad_norm": 0.10210710172882062, + "learning_rate": 8.774215741306395e-05, + "loss": 0.0036, + "step": 3080 + }, + { + "epoch": 4.897700237906424, + "grad_norm": 0.10464813148086784, + "learning_rate": 8.762853849280693e-05, + "loss": 0.0037, + "step": 3090 + }, + { + "epoch": 4.913560666137986, + "grad_norm": 0.10471907357639558, + "learning_rate": 8.751446971895206e-05, + "loss": 0.0036, + "step": 3100 + }, + { + "epoch": 4.929421094369548, + "grad_norm": 0.07894542756752126, + "learning_rate": 8.739995245520492e-05, + "loss": 0.0034, + "step": 3110 + }, + { + "epoch": 4.94528152260111, + "grad_norm": 0.09286574927632184, + "learning_rate": 8.728498807063288e-05, + "loss": 0.0035, + "step": 3120 + }, + { + "epoch": 4.961141950832673, + "grad_norm": 0.08571307419176223, + "learning_rate": 8.716957793964864e-05, + "loss": 0.0034, + "step": 3130 + }, + { + "epoch": 4.977002379064235, + "grad_norm": 0.0965051873600371, + "learning_rate": 8.705372344199387e-05, + "loss": 0.0034, + "step": 3140 + }, + { + "epoch": 4.992862807295797, + "grad_norm": 0.13496799817499566, + "learning_rate": 8.69374259627227e-05, + "loss": 0.0037, + "step": 3150 + }, + { + "epoch": 5.007930214115781, + "grad_norm": 0.08289025036640724, + "learning_rate": 8.682068689218517e-05, + "loss": 0.0031, + "step": 3160 + }, + { + "epoch": 5.023790642347343, + "grad_norm": 0.07226843684662153, + "learning_rate": 8.670350762601059e-05, + "loss": 0.0024, + "step": 3170 + }, + { + "epoch": 5.0396510705789055, + "grad_norm": 0.06521417712763433, + "learning_rate": 8.658588956509079e-05, + "loss": 0.0021, + "step": 3180 + }, + { + "epoch": 5.055511498810468, + "grad_norm": 0.07961399017197436, + "learning_rate": 8.646783411556359e-05, + "loss": 0.0018, + "step": 3190 + }, + { + "epoch": 5.07137192704203, + "grad_norm": 0.08645502226538618, + "learning_rate": 8.63493426887957e-05, + "loss": 0.0017, + "step": 3200 + }, + { + "epoch": 5.087232355273592, + "grad_norm": 0.07963006292300003, + "learning_rate": 8.623041670136606e-05, + "loss": 0.0016, + "step": 3210 + }, + { + "epoch": 5.103092783505154, + "grad_norm": 0.06181791334855889, + "learning_rate": 8.611105757504883e-05, + "loss": 0.0017, + "step": 3220 + }, + { + "epoch": 5.1189532117367165, + "grad_norm": 0.06988096800556821, + "learning_rate": 8.599126673679636e-05, + "loss": 0.0017, + "step": 3230 + }, + { + "epoch": 5.1348136399682796, + "grad_norm": 0.08328659822255369, + "learning_rate": 8.587104561872224e-05, + "loss": 0.0017, + "step": 3240 + }, + { + "epoch": 5.150674068199842, + "grad_norm": 0.11782439513762667, + "learning_rate": 8.5750395658084e-05, + "loss": 0.0017, + "step": 3250 + }, + { + "epoch": 5.166534496431404, + "grad_norm": 0.09235663805764693, + "learning_rate": 8.562931829726615e-05, + "loss": 0.0018, + "step": 3260 + }, + { + "epoch": 5.182394924662966, + "grad_norm": 0.07917233858457466, + "learning_rate": 8.55078149837627e-05, + "loss": 0.0018, + "step": 3270 + }, + { + "epoch": 5.198255352894528, + "grad_norm": 0.07720468137565101, + "learning_rate": 8.53858871701601e-05, + "loss": 0.0018, + "step": 3280 + }, + { + "epoch": 5.2141157811260905, + "grad_norm": 0.06460403408535213, + "learning_rate": 8.526353631411965e-05, + "loss": 0.0016, + "step": 3290 + }, + { + "epoch": 5.229976209357653, + "grad_norm": 0.0696806022225143, + "learning_rate": 8.514076387836021e-05, + "loss": 0.0016, + "step": 3300 + }, + { + "epoch": 5.245836637589215, + "grad_norm": 0.07170050913505419, + "learning_rate": 8.501757133064065e-05, + "loss": 0.0016, + "step": 3310 + }, + { + "epoch": 5.261697065820777, + "grad_norm": 0.0661054241095149, + "learning_rate": 8.489396014374237e-05, + "loss": 0.0016, + "step": 3320 + }, + { + "epoch": 5.277557494052339, + "grad_norm": 0.07387953260990546, + "learning_rate": 8.476993179545166e-05, + "loss": 0.0016, + "step": 3330 + }, + { + "epoch": 5.2934179222839015, + "grad_norm": 0.06220099525799539, + "learning_rate": 8.464548776854195e-05, + "loss": 0.0017, + "step": 3340 + }, + { + "epoch": 5.309278350515464, + "grad_norm": 0.06919342191387491, + "learning_rate": 8.452062955075621e-05, + "loss": 0.0016, + "step": 3350 + }, + { + "epoch": 5.325138778747026, + "grad_norm": 0.069239564548577, + "learning_rate": 8.43953586347891e-05, + "loss": 0.0016, + "step": 3360 + }, + { + "epoch": 5.340999206978588, + "grad_norm": 0.057485269571037244, + "learning_rate": 8.426967651826914e-05, + "loss": 0.0015, + "step": 3370 + }, + { + "epoch": 5.35685963521015, + "grad_norm": 0.06475405169843909, + "learning_rate": 8.41435847037408e-05, + "loss": 0.0015, + "step": 3380 + }, + { + "epoch": 5.372720063441713, + "grad_norm": 0.07618561528999164, + "learning_rate": 8.401708469864653e-05, + "loss": 0.0017, + "step": 3390 + }, + { + "epoch": 5.388580491673276, + "grad_norm": 0.0754261896183142, + "learning_rate": 8.389017801530875e-05, + "loss": 0.0016, + "step": 3400 + }, + { + "epoch": 5.404440919904838, + "grad_norm": 0.07800671155500213, + "learning_rate": 8.376286617091178e-05, + "loss": 0.0016, + "step": 3410 + }, + { + "epoch": 5.4203013481364, + "grad_norm": 0.07636728277995097, + "learning_rate": 8.363515068748364e-05, + "loss": 0.0016, + "step": 3420 + }, + { + "epoch": 5.436161776367962, + "grad_norm": 0.06974545329253266, + "learning_rate": 8.350703309187798e-05, + "loss": 0.0015, + "step": 3430 + }, + { + "epoch": 5.452022204599524, + "grad_norm": 0.08486131169745545, + "learning_rate": 8.337851491575568e-05, + "loss": 0.0016, + "step": 3440 + }, + { + "epoch": 5.4678826328310866, + "grad_norm": 0.10177268913062058, + "learning_rate": 8.324959769556664e-05, + "loss": 0.0017, + "step": 3450 + }, + { + "epoch": 5.483743061062649, + "grad_norm": 0.07544977089684703, + "learning_rate": 8.312028297253136e-05, + "loss": 0.0017, + "step": 3460 + }, + { + "epoch": 5.499603489294211, + "grad_norm": 0.08117648495668406, + "learning_rate": 8.299057229262256e-05, + "loss": 0.0017, + "step": 3470 + }, + { + "epoch": 5.515463917525773, + "grad_norm": 0.07640679176251267, + "learning_rate": 8.286046720654665e-05, + "loss": 0.0017, + "step": 3480 + }, + { + "epoch": 5.531324345757335, + "grad_norm": 0.079522193457805, + "learning_rate": 8.27299692697252e-05, + "loss": 0.0017, + "step": 3490 + }, + { + "epoch": 5.5471847739888975, + "grad_norm": 0.07722026998309503, + "learning_rate": 8.259908004227641e-05, + "loss": 0.0018, + "step": 3500 + }, + { + "epoch": 5.56304520222046, + "grad_norm": 0.08587811490541557, + "learning_rate": 8.246780108899635e-05, + "loss": 0.0019, + "step": 3510 + }, + { + "epoch": 5.578905630452022, + "grad_norm": 0.21112988448424747, + "learning_rate": 8.23361339793403e-05, + "loss": 0.002, + "step": 3520 + }, + { + "epoch": 5.594766058683584, + "grad_norm": 0.2731446862713363, + "learning_rate": 8.220408028740407e-05, + "loss": 0.0032, + "step": 3530 + }, + { + "epoch": 5.610626486915146, + "grad_norm": 0.1379133248787839, + "learning_rate": 8.207164159190503e-05, + "loss": 0.0037, + "step": 3540 + }, + { + "epoch": 5.6264869151467085, + "grad_norm": 0.09288443542654805, + "learning_rate": 8.193881947616337e-05, + "loss": 0.0024, + "step": 3550 + }, + { + "epoch": 5.642347343378272, + "grad_norm": 0.1024679954754957, + "learning_rate": 8.180561552808306e-05, + "loss": 0.0021, + "step": 3560 + }, + { + "epoch": 5.658207771609834, + "grad_norm": 0.08696933128166198, + "learning_rate": 8.167203134013295e-05, + "loss": 0.0021, + "step": 3570 + }, + { + "epoch": 5.674068199841396, + "grad_norm": 0.07453735269493612, + "learning_rate": 8.15380685093277e-05, + "loss": 0.002, + "step": 3580 + }, + { + "epoch": 5.689928628072958, + "grad_norm": 0.07288118742234356, + "learning_rate": 8.140372863720868e-05, + "loss": 0.0021, + "step": 3590 + }, + { + "epoch": 5.70578905630452, + "grad_norm": 0.07358612936238743, + "learning_rate": 8.126901332982486e-05, + "loss": 0.0019, + "step": 3600 + }, + { + "epoch": 5.721649484536083, + "grad_norm": 0.08782707271128637, + "learning_rate": 8.113392419771357e-05, + "loss": 0.0019, + "step": 3610 + }, + { + "epoch": 5.737509912767645, + "grad_norm": 0.0900509212790111, + "learning_rate": 8.099846285588124e-05, + "loss": 0.0019, + "step": 3620 + }, + { + "epoch": 5.753370340999207, + "grad_norm": 0.09047708148482936, + "learning_rate": 8.086263092378411e-05, + "loss": 0.002, + "step": 3630 + }, + { + "epoch": 5.769230769230769, + "grad_norm": 0.08927497872159121, + "learning_rate": 8.072643002530891e-05, + "loss": 0.0023, + "step": 3640 + }, + { + "epoch": 5.785091197462331, + "grad_norm": 0.09204513671940284, + "learning_rate": 8.058986178875337e-05, + "loss": 0.0022, + "step": 3650 + }, + { + "epoch": 5.8009516256938936, + "grad_norm": 0.08919270553467656, + "learning_rate": 8.045292784680679e-05, + "loss": 0.0021, + "step": 3660 + }, + { + "epoch": 5.816812053925456, + "grad_norm": 0.10352734228614939, + "learning_rate": 8.031562983653055e-05, + "loss": 0.0021, + "step": 3670 + }, + { + "epoch": 5.832672482157018, + "grad_norm": 0.07620459021669333, + "learning_rate": 8.017796939933846e-05, + "loss": 0.002, + "step": 3680 + }, + { + "epoch": 5.84853291038858, + "grad_norm": 0.10472027937771997, + "learning_rate": 8.003994818097727e-05, + "loss": 0.0021, + "step": 3690 + }, + { + "epoch": 5.864393338620143, + "grad_norm": 0.08136969120842585, + "learning_rate": 7.990156783150682e-05, + "loss": 0.0022, + "step": 3700 + }, + { + "epoch": 5.880253766851705, + "grad_norm": 0.09311572639718703, + "learning_rate": 7.976283000528045e-05, + "loss": 0.0022, + "step": 3710 + }, + { + "epoch": 5.896114195083268, + "grad_norm": 0.07648260529697609, + "learning_rate": 7.962373636092518e-05, + "loss": 0.002, + "step": 3720 + }, + { + "epoch": 5.91197462331483, + "grad_norm": 0.07441474521625142, + "learning_rate": 7.948428856132184e-05, + "loss": 0.002, + "step": 3730 + }, + { + "epoch": 5.927835051546392, + "grad_norm": 0.08167508861538944, + "learning_rate": 7.934448827358528e-05, + "loss": 0.002, + "step": 3740 + }, + { + "epoch": 5.943695479777954, + "grad_norm": 0.07264506356803561, + "learning_rate": 7.920433716904436e-05, + "loss": 0.0019, + "step": 3750 + }, + { + "epoch": 5.959555908009516, + "grad_norm": 0.09151401578654914, + "learning_rate": 7.9063836923222e-05, + "loss": 0.002, + "step": 3760 + }, + { + "epoch": 5.975416336241079, + "grad_norm": 0.09587477825025868, + "learning_rate": 7.892298921581512e-05, + "loss": 0.0024, + "step": 3770 + }, + { + "epoch": 5.991276764472641, + "grad_norm": 0.11595767096999594, + "learning_rate": 7.878179573067463e-05, + "loss": 0.0024, + "step": 3780 + }, + { + "epoch": 6.006344171292625, + "grad_norm": 0.1555465588033466, + "learning_rate": 7.864025815578523e-05, + "loss": 0.0022, + "step": 3790 + }, + { + "epoch": 6.022204599524187, + "grad_norm": 0.10558043532394686, + "learning_rate": 7.849837818324525e-05, + "loss": 0.0018, + "step": 3800 + }, + { + "epoch": 6.038065027755749, + "grad_norm": 0.12526172643990832, + "learning_rate": 7.835615750924643e-05, + "loss": 0.0016, + "step": 3810 + }, + { + "epoch": 6.0539254559873115, + "grad_norm": 0.061561568759045865, + "learning_rate": 7.821359783405364e-05, + "loss": 0.0014, + "step": 3820 + }, + { + "epoch": 6.069785884218874, + "grad_norm": 0.04920175361346469, + "learning_rate": 7.807070086198453e-05, + "loss": 0.0012, + "step": 3830 + }, + { + "epoch": 6.085646312450436, + "grad_norm": 0.06583556324811754, + "learning_rate": 7.792746830138918e-05, + "loss": 0.0012, + "step": 3840 + }, + { + "epoch": 6.101506740681998, + "grad_norm": 0.059777134326181966, + "learning_rate": 7.77839018646297e-05, + "loss": 0.0011, + "step": 3850 + }, + { + "epoch": 6.11736716891356, + "grad_norm": 0.05871456494754872, + "learning_rate": 7.764000326805967e-05, + "loss": 0.0011, + "step": 3860 + }, + { + "epoch": 6.133227597145123, + "grad_norm": 0.05183564018870685, + "learning_rate": 7.749577423200376e-05, + "loss": 0.0009, + "step": 3870 + }, + { + "epoch": 6.1490880253766855, + "grad_norm": 0.04452101726513349, + "learning_rate": 7.735121648073702e-05, + "loss": 0.001, + "step": 3880 + }, + { + "epoch": 6.164948453608248, + "grad_norm": 0.04848825916408963, + "learning_rate": 7.720633174246434e-05, + "loss": 0.001, + "step": 3890 + }, + { + "epoch": 6.18080888183981, + "grad_norm": 0.07679710142005615, + "learning_rate": 7.70611217492998e-05, + "loss": 0.001, + "step": 3900 + }, + { + "epoch": 6.196669310071372, + "grad_norm": 0.06108467756055413, + "learning_rate": 7.691558823724592e-05, + "loss": 0.001, + "step": 3910 + }, + { + "epoch": 6.212529738302934, + "grad_norm": 0.04649027671211247, + "learning_rate": 7.676973294617291e-05, + "loss": 0.001, + "step": 3920 + }, + { + "epoch": 6.2283901665344965, + "grad_norm": 0.06041575758385759, + "learning_rate": 7.662355761979794e-05, + "loss": 0.0009, + "step": 3930 + }, + { + "epoch": 6.244250594766059, + "grad_norm": 0.052658929916011704, + "learning_rate": 7.647706400566419e-05, + "loss": 0.001, + "step": 3940 + }, + { + "epoch": 6.260111022997621, + "grad_norm": 0.04677304710240248, + "learning_rate": 7.633025385512005e-05, + "loss": 0.001, + "step": 3950 + }, + { + "epoch": 6.275971451229183, + "grad_norm": 0.04926732545159951, + "learning_rate": 7.618312892329808e-05, + "loss": 0.0009, + "step": 3960 + }, + { + "epoch": 6.291831879460745, + "grad_norm": 0.05177855845577326, + "learning_rate": 7.603569096909413e-05, + "loss": 0.0009, + "step": 3970 + }, + { + "epoch": 6.3076923076923075, + "grad_norm": 0.0554899734550686, + "learning_rate": 7.588794175514629e-05, + "loss": 0.0009, + "step": 3980 + }, + { + "epoch": 6.32355273592387, + "grad_norm": 0.0605294191239795, + "learning_rate": 7.573988304781375e-05, + "loss": 0.0009, + "step": 3990 + }, + { + "epoch": 6.339413164155432, + "grad_norm": 0.05454380128565192, + "learning_rate": 7.559151661715573e-05, + "loss": 0.0009, + "step": 4000 + }, + { + "epoch": 6.355273592386994, + "grad_norm": 0.04999992541804801, + "learning_rate": 7.544284423691037e-05, + "loss": 0.0009, + "step": 4010 + }, + { + "epoch": 6.371134020618557, + "grad_norm": 0.06950474185130365, + "learning_rate": 7.529386768447342e-05, + "loss": 0.0009, + "step": 4020 + }, + { + "epoch": 6.386994448850119, + "grad_norm": 0.060369585072561134, + "learning_rate": 7.514458874087707e-05, + "loss": 0.001, + "step": 4030 + }, + { + "epoch": 6.4028548770816816, + "grad_norm": 0.05988890812840084, + "learning_rate": 7.499500919076861e-05, + "loss": 0.001, + "step": 4040 + }, + { + "epoch": 6.418715305313244, + "grad_norm": 0.04983886707023872, + "learning_rate": 7.484513082238917e-05, + "loss": 0.001, + "step": 4050 + }, + { + "epoch": 6.434575733544806, + "grad_norm": 0.04525796788080977, + "learning_rate": 7.469495542755223e-05, + "loss": 0.0008, + "step": 4060 + }, + { + "epoch": 6.450436161776368, + "grad_norm": 0.051174073627993116, + "learning_rate": 7.454448480162226e-05, + "loss": 0.0009, + "step": 4070 + }, + { + "epoch": 6.46629659000793, + "grad_norm": 0.05390059019863484, + "learning_rate": 7.439372074349328e-05, + "loss": 0.001, + "step": 4080 + }, + { + "epoch": 6.4821570182394925, + "grad_norm": 0.04968418800679971, + "learning_rate": 7.424266505556732e-05, + "loss": 0.001, + "step": 4090 + }, + { + "epoch": 6.498017446471055, + "grad_norm": 0.047223616532029535, + "learning_rate": 7.409131954373287e-05, + "loss": 0.0009, + "step": 4100 + }, + { + "epoch": 6.513877874702617, + "grad_norm": 0.07701904167042412, + "learning_rate": 7.393968601734331e-05, + "loss": 0.001, + "step": 4110 + }, + { + "epoch": 6.529738302934179, + "grad_norm": 0.06582428007131025, + "learning_rate": 7.378776628919525e-05, + "loss": 0.0011, + "step": 4120 + }, + { + "epoch": 6.545598731165741, + "grad_norm": 0.06437575429812158, + "learning_rate": 7.363556217550689e-05, + "loss": 0.001, + "step": 4130 + }, + { + "epoch": 6.5614591593973035, + "grad_norm": 0.06938985914429964, + "learning_rate": 7.348307549589629e-05, + "loss": 0.001, + "step": 4140 + }, + { + "epoch": 6.577319587628866, + "grad_norm": 0.05410097140716854, + "learning_rate": 7.33303080733596e-05, + "loss": 0.001, + "step": 4150 + }, + { + "epoch": 6.593180015860428, + "grad_norm": 0.05223503817044808, + "learning_rate": 7.317726173424935e-05, + "loss": 0.0011, + "step": 4160 + }, + { + "epoch": 6.60904044409199, + "grad_norm": 0.0526492757578401, + "learning_rate": 7.302393830825243e-05, + "loss": 0.001, + "step": 4170 + }, + { + "epoch": 6.624900872323552, + "grad_norm": 0.0693233254830145, + "learning_rate": 7.287033962836848e-05, + "loss": 0.001, + "step": 4180 + }, + { + "epoch": 6.640761300555115, + "grad_norm": 0.06350842466412508, + "learning_rate": 7.27164675308877e-05, + "loss": 0.0011, + "step": 4190 + }, + { + "epoch": 6.656621728786678, + "grad_norm": 0.05471050232924713, + "learning_rate": 7.256232385536915e-05, + "loss": 0.0011, + "step": 4200 + }, + { + "epoch": 6.67248215701824, + "grad_norm": 0.06467573654328658, + "learning_rate": 7.240791044461853e-05, + "loss": 0.001, + "step": 4210 + }, + { + "epoch": 6.688342585249802, + "grad_norm": 0.047507653472790644, + "learning_rate": 7.22532291446663e-05, + "loss": 0.0009, + "step": 4220 + }, + { + "epoch": 6.704203013481364, + "grad_norm": 0.047591824960592534, + "learning_rate": 7.209828180474562e-05, + "loss": 0.001, + "step": 4230 + }, + { + "epoch": 6.720063441712926, + "grad_norm": 0.05393949639861346, + "learning_rate": 7.194307027727011e-05, + "loss": 0.0009, + "step": 4240 + }, + { + "epoch": 6.7359238699444886, + "grad_norm": 0.051637823082064535, + "learning_rate": 7.178759641781178e-05, + "loss": 0.001, + "step": 4250 + }, + { + "epoch": 6.751784298176051, + "grad_norm": 0.06596437481148441, + "learning_rate": 7.163186208507896e-05, + "loss": 0.001, + "step": 4260 + }, + { + "epoch": 6.767644726407613, + "grad_norm": 0.0619925307581417, + "learning_rate": 7.147586914089385e-05, + "loss": 0.001, + "step": 4270 + }, + { + "epoch": 6.783505154639175, + "grad_norm": 0.05853128597694735, + "learning_rate": 7.131961945017041e-05, + "loss": 0.0011, + "step": 4280 + }, + { + "epoch": 6.799365582870737, + "grad_norm": 0.06990645541105177, + "learning_rate": 7.116311488089204e-05, + "loss": 0.0011, + "step": 4290 + }, + { + "epoch": 6.8152260111022995, + "grad_norm": 0.052077392785876926, + "learning_rate": 7.10063573040893e-05, + "loss": 0.001, + "step": 4300 + }, + { + "epoch": 6.831086439333862, + "grad_norm": 0.06544836210303613, + "learning_rate": 7.084934859381738e-05, + "loss": 0.001, + "step": 4310 + }, + { + "epoch": 6.846946867565424, + "grad_norm": 0.06172813329199656, + "learning_rate": 7.06920906271339e-05, + "loss": 0.0011, + "step": 4320 + }, + { + "epoch": 6.862807295796987, + "grad_norm": 0.06442400738720315, + "learning_rate": 7.053458528407628e-05, + "loss": 0.0011, + "step": 4330 + }, + { + "epoch": 6.878667724028549, + "grad_norm": 0.053719055980683154, + "learning_rate": 7.037683444763945e-05, + "loss": 0.0011, + "step": 4340 + }, + { + "epoch": 6.894528152260111, + "grad_norm": 0.060737192661524886, + "learning_rate": 7.021884000375314e-05, + "loss": 0.001, + "step": 4350 + }, + { + "epoch": 6.910388580491674, + "grad_norm": 0.05976598593814817, + "learning_rate": 7.006060384125952e-05, + "loss": 0.0011, + "step": 4360 + }, + { + "epoch": 6.926249008723236, + "grad_norm": 0.05525031117041402, + "learning_rate": 6.990212785189049e-05, + "loss": 0.0011, + "step": 4370 + }, + { + "epoch": 6.942109436954798, + "grad_norm": 0.06248661436454952, + "learning_rate": 6.974341393024511e-05, + "loss": 0.0011, + "step": 4380 + }, + { + "epoch": 6.95796986518636, + "grad_norm": 0.06606564236745904, + "learning_rate": 6.958446397376696e-05, + "loss": 0.0011, + "step": 4390 + }, + { + "epoch": 6.973830293417922, + "grad_norm": 0.06762039827514536, + "learning_rate": 6.942527988272146e-05, + "loss": 0.001, + "step": 4400 + }, + { + "epoch": 6.989690721649485, + "grad_norm": 0.06040099605023135, + "learning_rate": 6.92658635601731e-05, + "loss": 0.0011, + "step": 4410 + }, + { + "epoch": 7.004758128469469, + "grad_norm": 0.04860042394066092, + "learning_rate": 6.910621691196274e-05, + "loss": 0.001, + "step": 4420 + }, + { + "epoch": 7.020618556701031, + "grad_norm": 0.048580123629282816, + "learning_rate": 6.894634184668482e-05, + "loss": 0.0009, + "step": 4430 + }, + { + "epoch": 7.036478984932593, + "grad_norm": 0.04944523858133885, + "learning_rate": 6.878624027566453e-05, + "loss": 0.0007, + "step": 4440 + }, + { + "epoch": 7.052339413164155, + "grad_norm": 0.058377029739441685, + "learning_rate": 6.862591411293494e-05, + "loss": 0.0007, + "step": 4450 + }, + { + "epoch": 7.068199841395717, + "grad_norm": 0.03809518537647906, + "learning_rate": 6.846536527521417e-05, + "loss": 0.0007, + "step": 4460 + }, + { + "epoch": 7.08406026962728, + "grad_norm": 0.05698131623555525, + "learning_rate": 6.830459568188246e-05, + "loss": 0.0007, + "step": 4470 + }, + { + "epoch": 7.099920697858842, + "grad_norm": 0.061438751415187925, + "learning_rate": 6.814360725495914e-05, + "loss": 0.0007, + "step": 4480 + }, + { + "epoch": 7.115781126090404, + "grad_norm": 0.0557410453924883, + "learning_rate": 6.798240191907979e-05, + "loss": 0.0007, + "step": 4490 + }, + { + "epoch": 7.131641554321967, + "grad_norm": 0.07638492208851758, + "learning_rate": 6.782098160147312e-05, + "loss": 0.0008, + "step": 4500 + }, + { + "epoch": 7.147501982553529, + "grad_norm": 0.03833873022980692, + "learning_rate": 6.765934823193802e-05, + "loss": 0.0007, + "step": 4510 + }, + { + "epoch": 7.1633624107850915, + "grad_norm": 0.03634210982547353, + "learning_rate": 6.749750374282039e-05, + "loss": 0.0006, + "step": 4520 + }, + { + "epoch": 7.179222839016654, + "grad_norm": 0.05078518093897086, + "learning_rate": 6.73354500689901e-05, + "loss": 0.0006, + "step": 4530 + }, + { + "epoch": 7.195083267248216, + "grad_norm": 0.05850986673672021, + "learning_rate": 6.717318914781785e-05, + "loss": 0.0006, + "step": 4540 + }, + { + "epoch": 7.210943695479778, + "grad_norm": 0.059400415722010595, + "learning_rate": 6.701072291915204e-05, + "loss": 0.0005, + "step": 4550 + }, + { + "epoch": 7.22680412371134, + "grad_norm": 0.04930569288631075, + "learning_rate": 6.684805332529552e-05, + "loss": 0.0006, + "step": 4560 + }, + { + "epoch": 7.2426645519429025, + "grad_norm": 0.047625627802954776, + "learning_rate": 6.668518231098237e-05, + "loss": 0.0005, + "step": 4570 + }, + { + "epoch": 7.258524980174465, + "grad_norm": 0.043562382381430974, + "learning_rate": 6.65221118233547e-05, + "loss": 0.0005, + "step": 4580 + }, + { + "epoch": 7.274385408406027, + "grad_norm": 0.05439831172543469, + "learning_rate": 6.635884381193938e-05, + "loss": 0.0005, + "step": 4590 + }, + { + "epoch": 7.290245836637589, + "grad_norm": 0.030240112671842217, + "learning_rate": 6.619538022862463e-05, + "loss": 0.0005, + "step": 4600 + }, + { + "epoch": 7.306106264869151, + "grad_norm": 0.11009572743451485, + "learning_rate": 6.603172302763678e-05, + "loss": 0.0005, + "step": 4610 + }, + { + "epoch": 7.3219666931007135, + "grad_norm": 0.02916185208940372, + "learning_rate": 6.586787416551692e-05, + "loss": 0.0006, + "step": 4620 + }, + { + "epoch": 7.337827121332276, + "grad_norm": 0.0437856633245075, + "learning_rate": 6.570383560109745e-05, + "loss": 0.0005, + "step": 4630 + }, + { + "epoch": 7.353687549563838, + "grad_norm": 0.037912725580370316, + "learning_rate": 6.553960929547865e-05, + "loss": 0.0005, + "step": 4640 + }, + { + "epoch": 7.369547977795401, + "grad_norm": 0.05147787378324253, + "learning_rate": 6.53751972120053e-05, + "loss": 0.0006, + "step": 4650 + }, + { + "epoch": 7.385408406026963, + "grad_norm": 0.0370003976866085, + "learning_rate": 6.52106013162432e-05, + "loss": 0.0006, + "step": 4660 + }, + { + "epoch": 7.401268834258525, + "grad_norm": 0.050012286591099984, + "learning_rate": 6.504582357595559e-05, + "loss": 0.0005, + "step": 4670 + }, + { + "epoch": 7.4171292624900875, + "grad_norm": 0.03356610637671652, + "learning_rate": 6.488086596107971e-05, + "loss": 0.0005, + "step": 4680 + }, + { + "epoch": 7.43298969072165, + "grad_norm": 0.049833987165535526, + "learning_rate": 6.471573044370324e-05, + "loss": 0.0006, + "step": 4690 + }, + { + "epoch": 7.448850118953212, + "grad_norm": 0.05740703921842105, + "learning_rate": 6.455041899804069e-05, + "loss": 0.0006, + "step": 4700 + }, + { + "epoch": 7.464710547184774, + "grad_norm": 0.07910669772660382, + "learning_rate": 6.438493360040979e-05, + "loss": 0.0006, + "step": 4710 + }, + { + "epoch": 7.480570975416336, + "grad_norm": 0.05324028161436345, + "learning_rate": 6.421927622920791e-05, + "loss": 0.0006, + "step": 4720 + }, + { + "epoch": 7.4964314036478985, + "grad_norm": 0.06324204006430136, + "learning_rate": 6.405344886488842e-05, + "loss": 0.0007, + "step": 4730 + }, + { + "epoch": 7.512291831879461, + "grad_norm": 0.04656812274836249, + "learning_rate": 6.38874534899369e-05, + "loss": 0.0008, + "step": 4740 + }, + { + "epoch": 7.528152260111023, + "grad_norm": 0.05050526796955192, + "learning_rate": 6.372129208884756e-05, + "loss": 0.0007, + "step": 4750 + }, + { + "epoch": 7.544012688342585, + "grad_norm": 0.05650287012095859, + "learning_rate": 6.355496664809948e-05, + "loss": 0.0007, + "step": 4760 + }, + { + "epoch": 7.559873116574147, + "grad_norm": 0.10058700619547112, + "learning_rate": 6.338847915613285e-05, + "loss": 0.0006, + "step": 4770 + }, + { + "epoch": 7.5757335448057095, + "grad_norm": 0.06911169685530276, + "learning_rate": 6.322183160332518e-05, + "loss": 0.0007, + "step": 4780 + }, + { + "epoch": 7.591593973037272, + "grad_norm": 0.04702908632862955, + "learning_rate": 6.305502598196754e-05, + "loss": 0.0007, + "step": 4790 + }, + { + "epoch": 7.607454401268834, + "grad_norm": 0.046442441585103376, + "learning_rate": 6.288806428624073e-05, + "loss": 0.0007, + "step": 4800 + }, + { + "epoch": 7.623314829500396, + "grad_norm": 0.04782496243604433, + "learning_rate": 6.272094851219143e-05, + "loss": 0.0007, + "step": 4810 + }, + { + "epoch": 7.639175257731958, + "grad_norm": 0.05198912526184102, + "learning_rate": 6.255368065770837e-05, + "loss": 0.0007, + "step": 4820 + }, + { + "epoch": 7.655035685963521, + "grad_norm": 0.05484067854371931, + "learning_rate": 6.238626272249839e-05, + "loss": 0.0007, + "step": 4830 + }, + { + "epoch": 7.6708961141950835, + "grad_norm": 0.0756621394461863, + "learning_rate": 6.221869670806256e-05, + "loss": 0.0007, + "step": 4840 + }, + { + "epoch": 7.686756542426646, + "grad_norm": 0.048334169412800464, + "learning_rate": 6.205098461767227e-05, + "loss": 0.0008, + "step": 4850 + }, + { + "epoch": 7.702616970658208, + "grad_norm": 0.0543194459873048, + "learning_rate": 6.188312845634527e-05, + "loss": 0.0012, + "step": 4860 + }, + { + "epoch": 7.71847739888977, + "grad_norm": 0.08724203182154329, + "learning_rate": 6.171513023082167e-05, + "loss": 0.0011, + "step": 4870 + }, + { + "epoch": 7.734337827121332, + "grad_norm": 0.06305941768082027, + "learning_rate": 6.154699194954001e-05, + "loss": 0.001, + "step": 4880 + }, + { + "epoch": 7.7501982553528945, + "grad_norm": 0.07268363768915513, + "learning_rate": 6.137871562261316e-05, + "loss": 0.0011, + "step": 4890 + }, + { + "epoch": 7.766058683584457, + "grad_norm": 0.07591187926547788, + "learning_rate": 6.121030326180444e-05, + "loss": 0.0009, + "step": 4900 + }, + { + "epoch": 7.781919111816019, + "grad_norm": 0.054366196418970766, + "learning_rate": 6.104175688050335e-05, + "loss": 0.001, + "step": 4910 + }, + { + "epoch": 7.797779540047581, + "grad_norm": 0.04752551179296887, + "learning_rate": 6.087307849370171e-05, + "loss": 0.0009, + "step": 4920 + }, + { + "epoch": 7.813639968279143, + "grad_norm": 0.05644525304940735, + "learning_rate": 6.0704270117969394e-05, + "loss": 0.0009, + "step": 4930 + }, + { + "epoch": 7.8295003965107055, + "grad_norm": 0.05100399715926843, + "learning_rate": 6.053533377143044e-05, + "loss": 0.0008, + "step": 4940 + }, + { + "epoch": 7.845360824742268, + "grad_norm": 2.4999935229694272, + "learning_rate": 6.0366271473738676e-05, + "loss": 0.0045, + "step": 4950 + }, + { + "epoch": 7.861221252973831, + "grad_norm": 1.4159543231553449, + "learning_rate": 6.019708524605372e-05, + "loss": 0.0147, + "step": 4960 + }, + { + "epoch": 7.877081681205393, + "grad_norm": 0.30064808008004296, + "learning_rate": 6.0027777111016805e-05, + "loss": 0.0322, + "step": 4970 + }, + { + "epoch": 7.892942109436955, + "grad_norm": 0.1722668557350698, + "learning_rate": 5.9858349092726604e-05, + "loss": 0.0065, + "step": 4980 + }, + { + "epoch": 7.908802537668517, + "grad_norm": 0.10237132234385116, + "learning_rate": 5.9688803216714936e-05, + "loss": 0.0033, + "step": 4990 + }, + { + "epoch": 7.92466296590008, + "grad_norm": 0.08849969653013551, + "learning_rate": 5.951914150992268e-05, + "loss": 0.0022, + "step": 5000 + }, + { + "epoch": 7.940523394131642, + "grad_norm": 0.06202027658329416, + "learning_rate": 5.9349366000675477e-05, + "loss": 0.0017, + "step": 5010 + }, + { + "epoch": 7.956383822363204, + "grad_norm": 0.05739572608373075, + "learning_rate": 5.91794787186595e-05, + "loss": 0.0014, + "step": 5020 + }, + { + "epoch": 7.972244250594766, + "grad_norm": 0.06615477618726463, + "learning_rate": 5.9009481694897115e-05, + "loss": 0.0012, + "step": 5030 + }, + { + "epoch": 7.988104678826328, + "grad_norm": 0.052081727575893286, + "learning_rate": 5.8839376961722745e-05, + "loss": 0.0012, + "step": 5040 + }, + { + "epoch": 8.003172085646312, + "grad_norm": 0.03960788097647101, + "learning_rate": 5.866916655275846e-05, + "loss": 0.0011, + "step": 5050 + }, + { + "epoch": 8.019032513877875, + "grad_norm": 0.0400616271445871, + "learning_rate": 5.8498852502889666e-05, + "loss": 0.0007, + "step": 5060 + }, + { + "epoch": 8.034892942109437, + "grad_norm": 0.06604868528910623, + "learning_rate": 5.8328436848240864e-05, + "loss": 0.0007, + "step": 5070 + }, + { + "epoch": 8.050753370340999, + "grad_norm": 0.03835249823221253, + "learning_rate": 5.8157921626151215e-05, + "loss": 0.0006, + "step": 5080 + }, + { + "epoch": 8.066613798572561, + "grad_norm": 0.03135708333694173, + "learning_rate": 5.7987308875150214e-05, + "loss": 0.0006, + "step": 5090 + }, + { + "epoch": 8.082474226804123, + "grad_norm": 0.03855339357246141, + "learning_rate": 5.781660063493336e-05, + "loss": 0.0006, + "step": 5100 + }, + { + "epoch": 8.098334655035686, + "grad_norm": 0.029024708021100142, + "learning_rate": 5.7645798946337705e-05, + "loss": 0.0004, + "step": 5110 + }, + { + "epoch": 8.114195083267248, + "grad_norm": 0.02966423341911613, + "learning_rate": 5.7474905851317505e-05, + "loss": 0.0005, + "step": 5120 + }, + { + "epoch": 8.13005551149881, + "grad_norm": 0.03567093820619331, + "learning_rate": 5.73039233929198e-05, + "loss": 0.0004, + "step": 5130 + }, + { + "epoch": 8.145915939730372, + "grad_norm": 0.03274356259330027, + "learning_rate": 5.713285361525994e-05, + "loss": 0.0004, + "step": 5140 + }, + { + "epoch": 8.161776367961934, + "grad_norm": 0.024120308375638385, + "learning_rate": 5.6961698563497235e-05, + "loss": 0.0004, + "step": 5150 + }, + { + "epoch": 8.177636796193497, + "grad_norm": 0.036251866697577205, + "learning_rate": 5.679046028381042e-05, + "loss": 0.0004, + "step": 5160 + }, + { + "epoch": 8.193497224425059, + "grad_norm": 0.03226037226742675, + "learning_rate": 5.661914082337324e-05, + "loss": 0.0004, + "step": 5170 + }, + { + "epoch": 8.209357652656621, + "grad_norm": 0.020434926978534256, + "learning_rate": 5.644774223032997e-05, + "loss": 0.0004, + "step": 5180 + }, + { + "epoch": 8.225218080888183, + "grad_norm": 0.02884624858078732, + "learning_rate": 5.627626655377094e-05, + "loss": 0.0004, + "step": 5190 + }, + { + "epoch": 8.241078509119745, + "grad_norm": 0.028620399603992385, + "learning_rate": 5.6104715843708e-05, + "loss": 0.0004, + "step": 5200 + }, + { + "epoch": 8.256938937351308, + "grad_norm": 0.02864165481991412, + "learning_rate": 5.593309215105004e-05, + "loss": 0.0004, + "step": 5210 + }, + { + "epoch": 8.272799365582872, + "grad_norm": 0.04523693266649351, + "learning_rate": 5.576139752757844e-05, + "loss": 0.0004, + "step": 5220 + }, + { + "epoch": 8.288659793814434, + "grad_norm": 0.03773855372297919, + "learning_rate": 5.558963402592263e-05, + "loss": 0.0004, + "step": 5230 + }, + { + "epoch": 8.304520222045996, + "grad_norm": 0.028887550324473894, + "learning_rate": 5.541780369953544e-05, + "loss": 0.0004, + "step": 5240 + }, + { + "epoch": 8.320380650277558, + "grad_norm": 0.02901173292500093, + "learning_rate": 5.524590860266859e-05, + "loss": 0.0004, + "step": 5250 + }, + { + "epoch": 8.33624107850912, + "grad_norm": 0.04751345258117833, + "learning_rate": 5.507395079034816e-05, + "loss": 0.0004, + "step": 5260 + }, + { + "epoch": 8.352101506740683, + "grad_norm": 0.021289452398226893, + "learning_rate": 5.4901932318350026e-05, + "loss": 0.0004, + "step": 5270 + }, + { + "epoch": 8.367961934972245, + "grad_norm": 0.02991161644518557, + "learning_rate": 5.4729855243175186e-05, + "loss": 0.0004, + "step": 5280 + }, + { + "epoch": 8.383822363203807, + "grad_norm": 0.030495608711481707, + "learning_rate": 5.4557721622025304e-05, + "loss": 0.0003, + "step": 5290 + }, + { + "epoch": 8.399682791435369, + "grad_norm": 0.02610698601639174, + "learning_rate": 5.438553351277805e-05, + "loss": 0.0003, + "step": 5300 + }, + { + "epoch": 8.415543219666931, + "grad_norm": 0.04105872683666564, + "learning_rate": 5.4213292973962484e-05, + "loss": 0.0003, + "step": 5310 + }, + { + "epoch": 8.431403647898494, + "grad_norm": 0.0279760155231915, + "learning_rate": 5.404100206473448e-05, + "loss": 0.0003, + "step": 5320 + }, + { + "epoch": 8.447264076130056, + "grad_norm": 0.02378510534034516, + "learning_rate": 5.386866284485211e-05, + "loss": 0.0003, + "step": 5330 + }, + { + "epoch": 8.463124504361618, + "grad_norm": 0.023718924196310366, + "learning_rate": 5.369627737465098e-05, + "loss": 0.0003, + "step": 5340 + }, + { + "epoch": 8.47898493259318, + "grad_norm": 0.02650070864301587, + "learning_rate": 5.3523847715019636e-05, + "loss": 0.0003, + "step": 5350 + }, + { + "epoch": 8.494845360824742, + "grad_norm": 0.02243594493987124, + "learning_rate": 5.335137592737493e-05, + "loss": 0.0003, + "step": 5360 + }, + { + "epoch": 8.510705789056304, + "grad_norm": 0.023772356203941643, + "learning_rate": 5.3178864073637326e-05, + "loss": 0.0003, + "step": 5370 + }, + { + "epoch": 8.526566217287867, + "grad_norm": 0.02348807792613231, + "learning_rate": 5.30063142162063e-05, + "loss": 0.0003, + "step": 5380 + }, + { + "epoch": 8.542426645519429, + "grad_norm": 0.028204543884092787, + "learning_rate": 5.283372841793567e-05, + "loss": 0.0003, + "step": 5390 + }, + { + "epoch": 8.558287073750991, + "grad_norm": 0.023525226513603743, + "learning_rate": 5.2661108742108935e-05, + "loss": 0.0003, + "step": 5400 + }, + { + "epoch": 8.574147501982553, + "grad_norm": 0.02630409178641063, + "learning_rate": 5.248845725241459e-05, + "loss": 0.0004, + "step": 5410 + }, + { + "epoch": 8.590007930214115, + "grad_norm": 0.03214914000318704, + "learning_rate": 5.231577601292148e-05, + "loss": 0.0004, + "step": 5420 + }, + { + "epoch": 8.605868358445678, + "grad_norm": 0.023104294841336203, + "learning_rate": 5.21430670880541e-05, + "loss": 0.0003, + "step": 5430 + }, + { + "epoch": 8.62172878667724, + "grad_norm": 0.03746092091988779, + "learning_rate": 5.1970332542567944e-05, + "loss": 0.0003, + "step": 5440 + }, + { + "epoch": 8.637589214908802, + "grad_norm": 0.04722939342837188, + "learning_rate": 5.1797574441524796e-05, + "loss": 0.0003, + "step": 5450 + }, + { + "epoch": 8.653449643140364, + "grad_norm": 0.02693201938916151, + "learning_rate": 5.162479485026802e-05, + "loss": 0.0003, + "step": 5460 + }, + { + "epoch": 8.669310071371926, + "grad_norm": 0.019995789027273305, + "learning_rate": 5.145199583439797e-05, + "loss": 0.0003, + "step": 5470 + }, + { + "epoch": 8.685170499603489, + "grad_norm": 0.03601325496752403, + "learning_rate": 5.127917945974714e-05, + "loss": 0.0003, + "step": 5480 + }, + { + "epoch": 8.70103092783505, + "grad_norm": 0.030379505031284344, + "learning_rate": 5.1106347792355626e-05, + "loss": 0.0003, + "step": 5490 + }, + { + "epoch": 8.716891356066613, + "grad_norm": 0.028047798761568314, + "learning_rate": 5.093350289844626e-05, + "loss": 0.0004, + "step": 5500 + }, + { + "epoch": 8.732751784298175, + "grad_norm": 0.03223760480684172, + "learning_rate": 5.076064684440011e-05, + "loss": 0.0003, + "step": 5510 + }, + { + "epoch": 8.74861221252974, + "grad_norm": 0.023143019605304356, + "learning_rate": 5.0587781696731584e-05, + "loss": 0.0003, + "step": 5520 + }, + { + "epoch": 8.764472640761301, + "grad_norm": 0.039771667681304905, + "learning_rate": 5.0414909522063826e-05, + "loss": 0.0003, + "step": 5530 + }, + { + "epoch": 8.780333068992864, + "grad_norm": 0.031156272099768664, + "learning_rate": 5.0242032387103976e-05, + "loss": 0.0004, + "step": 5540 + }, + { + "epoch": 8.796193497224426, + "grad_norm": 0.03179437172005449, + "learning_rate": 5.006915235861853e-05, + "loss": 0.0003, + "step": 5550 + }, + { + "epoch": 8.812053925455988, + "grad_norm": 0.03486636221508783, + "learning_rate": 4.989627150340849e-05, + "loss": 0.0003, + "step": 5560 + }, + { + "epoch": 8.82791435368755, + "grad_norm": 0.02786972273306398, + "learning_rate": 4.9723391888284835e-05, + "loss": 0.0003, + "step": 5570 + }, + { + "epoch": 8.843774781919112, + "grad_norm": 0.028916427846878534, + "learning_rate": 4.955051558004363e-05, + "loss": 0.0003, + "step": 5580 + }, + { + "epoch": 8.859635210150675, + "grad_norm": 0.014171273458187344, + "learning_rate": 4.937764464544148e-05, + "loss": 0.0003, + "step": 5590 + }, + { + "epoch": 8.875495638382237, + "grad_norm": 0.022052465210804013, + "learning_rate": 4.92047811511707e-05, + "loss": 0.0003, + "step": 5600 + }, + { + "epoch": 8.891356066613799, + "grad_norm": 0.02105104649262549, + "learning_rate": 4.903192716383465e-05, + "loss": 0.0003, + "step": 5610 + }, + { + "epoch": 8.907216494845361, + "grad_norm": 0.02274547968968386, + "learning_rate": 4.8859084749923104e-05, + "loss": 0.0003, + "step": 5620 + }, + { + "epoch": 8.923076923076923, + "grad_norm": 0.029945920794814042, + "learning_rate": 4.8686255975787375e-05, + "loss": 0.0003, + "step": 5630 + }, + { + "epoch": 8.938937351308486, + "grad_norm": 0.023802706082847097, + "learning_rate": 4.8513442907615776e-05, + "loss": 0.0003, + "step": 5640 + }, + { + "epoch": 8.954797779540048, + "grad_norm": 0.027922424326692386, + "learning_rate": 4.834064761140886e-05, + "loss": 0.0003, + "step": 5650 + }, + { + "epoch": 8.97065820777161, + "grad_norm": 0.030297822205521734, + "learning_rate": 4.8167872152954655e-05, + "loss": 0.0003, + "step": 5660 + }, + { + "epoch": 8.986518636003172, + "grad_norm": 0.03465648789864436, + "learning_rate": 4.7995118597804094e-05, + "loss": 0.0003, + "step": 5670 + }, + { + "epoch": 9.001586042823156, + "grad_norm": 0.032840674243023976, + "learning_rate": 4.782238901124618e-05, + "loss": 0.0003, + "step": 5680 + }, + { + "epoch": 9.017446471054718, + "grad_norm": 0.015017847145791068, + "learning_rate": 4.7649685458283454e-05, + "loss": 0.0002, + "step": 5690 + }, + { + "epoch": 9.03330689928628, + "grad_norm": 0.018455950290203554, + "learning_rate": 4.7477010003607175e-05, + "loss": 0.0002, + "step": 5700 + }, + { + "epoch": 9.049167327517843, + "grad_norm": 0.011963828944432738, + "learning_rate": 4.730436471157265e-05, + "loss": 0.0001, + "step": 5710 + }, + { + "epoch": 9.065027755749405, + "grad_norm": 0.026237542751805978, + "learning_rate": 4.713175164617465e-05, + "loss": 0.0002, + "step": 5720 + }, + { + "epoch": 9.080888183980967, + "grad_norm": 0.014407530388858248, + "learning_rate": 4.6959172871022675e-05, + "loss": 0.0001, + "step": 5730 + }, + { + "epoch": 9.09674861221253, + "grad_norm": 0.01999433119809984, + "learning_rate": 4.678663044931621e-05, + "loss": 0.0001, + "step": 5740 + }, + { + "epoch": 9.112609040444092, + "grad_norm": 0.016191691366225166, + "learning_rate": 4.661412644382021e-05, + "loss": 0.0002, + "step": 5750 + }, + { + "epoch": 9.128469468675654, + "grad_norm": 0.00571885681761795, + "learning_rate": 4.6441662916840275e-05, + "loss": 0.0001, + "step": 5760 + }, + { + "epoch": 9.144329896907216, + "grad_norm": 0.01838706614722634, + "learning_rate": 4.6269241930198186e-05, + "loss": 0.0001, + "step": 5770 + }, + { + "epoch": 9.160190325138778, + "grad_norm": 0.011093932818962444, + "learning_rate": 4.6096865545207075e-05, + "loss": 0.0001, + "step": 5780 + }, + { + "epoch": 9.17605075337034, + "grad_norm": 0.018801707550067936, + "learning_rate": 4.592453582264684e-05, + "loss": 0.0001, + "step": 5790 + }, + { + "epoch": 9.191911181601903, + "grad_norm": 0.00449843256018912, + "learning_rate": 4.57522548227396e-05, + "loss": 0.0001, + "step": 5800 + }, + { + "epoch": 9.207771609833465, + "grad_norm": 0.02231820630066957, + "learning_rate": 4.5580024605124945e-05, + "loss": 0.0001, + "step": 5810 + }, + { + "epoch": 9.223632038065027, + "grad_norm": 0.00648685587573992, + "learning_rate": 4.540784722883531e-05, + "loss": 0.0001, + "step": 5820 + }, + { + "epoch": 9.23949246629659, + "grad_norm": 0.005639906279032426, + "learning_rate": 4.523572475227152e-05, + "loss": 0.0001, + "step": 5830 + }, + { + "epoch": 9.255352894528151, + "grad_norm": 0.011481380586830648, + "learning_rate": 4.506365923317795e-05, + "loss": 0.0001, + "step": 5840 + }, + { + "epoch": 9.271213322759715, + "grad_norm": 0.013073783366184224, + "learning_rate": 4.4891652728618105e-05, + "loss": 0.0001, + "step": 5850 + }, + { + "epoch": 9.287073750991278, + "grad_norm": 0.016892392475940494, + "learning_rate": 4.471970729494999e-05, + "loss": 0.0002, + "step": 5860 + }, + { + "epoch": 9.30293417922284, + "grad_norm": 0.006999585770790309, + "learning_rate": 4.454782498780139e-05, + "loss": 0.0001, + "step": 5870 + }, + { + "epoch": 9.318794607454402, + "grad_norm": 0.01142318049260765, + "learning_rate": 4.437600786204555e-05, + "loss": 0.0001, + "step": 5880 + }, + { + "epoch": 9.334655035685964, + "grad_norm": 0.011686688187080092, + "learning_rate": 4.4204257971776366e-05, + "loss": 0.0001, + "step": 5890 + }, + { + "epoch": 9.350515463917526, + "grad_norm": 0.004573580033721647, + "learning_rate": 4.4032577370283915e-05, + "loss": 0.0001, + "step": 5900 + }, + { + "epoch": 9.366375892149089, + "grad_norm": 0.0062415452310411994, + "learning_rate": 4.386096811003002e-05, + "loss": 0.0001, + "step": 5910 + }, + { + "epoch": 9.38223632038065, + "grad_norm": 0.0130888296334038, + "learning_rate": 4.36894322426235e-05, + "loss": 0.0001, + "step": 5920 + }, + { + "epoch": 9.398096748612213, + "grad_norm": 0.013365287023817644, + "learning_rate": 4.35179718187958e-05, + "loss": 0.0001, + "step": 5930 + }, + { + "epoch": 9.413957176843775, + "grad_norm": 0.016448263796730356, + "learning_rate": 4.334658888837646e-05, + "loss": 0.0001, + "step": 5940 + }, + { + "epoch": 9.429817605075337, + "grad_norm": 0.006174664376194607, + "learning_rate": 4.317528550026851e-05, + "loss": 0.0001, + "step": 5950 + }, + { + "epoch": 9.4456780333069, + "grad_norm": 0.016792550600562443, + "learning_rate": 4.300406370242409e-05, + "loss": 0.0001, + "step": 5960 + }, + { + "epoch": 9.461538461538462, + "grad_norm": 0.019197502531048048, + "learning_rate": 4.2832925541819935e-05, + "loss": 0.0001, + "step": 5970 + }, + { + "epoch": 9.477398889770024, + "grad_norm": 0.006131125462404848, + "learning_rate": 4.266187306443282e-05, + "loss": 0.0001, + "step": 5980 + }, + { + "epoch": 9.493259318001586, + "grad_norm": 0.011332090910883387, + "learning_rate": 4.2490908315215226e-05, + "loss": 0.0001, + "step": 5990 + }, + { + "epoch": 9.509119746233148, + "grad_norm": 0.018702959401095825, + "learning_rate": 4.232003333807078e-05, + "loss": 0.0001, + "step": 6000 + }, + { + "epoch": 9.52498017446471, + "grad_norm": 0.004302119488773759, + "learning_rate": 4.2149250175829955e-05, + "loss": 0.0001, + "step": 6010 + }, + { + "epoch": 9.540840602696273, + "grad_norm": 0.043539867837665065, + "learning_rate": 4.197856087022549e-05, + "loss": 0.0002, + "step": 6020 + }, + { + "epoch": 9.556701030927835, + "grad_norm": 0.011353400500880415, + "learning_rate": 4.180796746186807e-05, + "loss": 0.0001, + "step": 6030 + }, + { + "epoch": 9.572561459159397, + "grad_norm": 0.009183304696088834, + "learning_rate": 4.163747199022195e-05, + "loss": 0.0001, + "step": 6040 + }, + { + "epoch": 9.58842188739096, + "grad_norm": 0.01689371380588678, + "learning_rate": 4.1467076493580555e-05, + "loss": 0.0001, + "step": 6050 + }, + { + "epoch": 9.604282315622521, + "grad_norm": 0.003157397472935962, + "learning_rate": 4.1296783009042e-05, + "loss": 0.0001, + "step": 6060 + }, + { + "epoch": 9.620142743854084, + "grad_norm": 0.018735867008279953, + "learning_rate": 4.112659357248495e-05, + "loss": 0.0001, + "step": 6070 + }, + { + "epoch": 9.636003172085646, + "grad_norm": 0.013848337298858471, + "learning_rate": 4.0956510218544056e-05, + "loss": 0.0001, + "step": 6080 + }, + { + "epoch": 9.651863600317208, + "grad_norm": 0.014508090196361574, + "learning_rate": 4.0786534980585845e-05, + "loss": 0.0001, + "step": 6090 + }, + { + "epoch": 9.66772402854877, + "grad_norm": 0.01251498969768193, + "learning_rate": 4.061666989068424e-05, + "loss": 0.0001, + "step": 6100 + }, + { + "epoch": 9.683584456780332, + "grad_norm": 0.01184936863179866, + "learning_rate": 4.044691697959631e-05, + "loss": 0.0001, + "step": 6110 + }, + { + "epoch": 9.699444885011895, + "grad_norm": 0.006143083324493081, + "learning_rate": 4.027727827673809e-05, + "loss": 0.0001, + "step": 6120 + }, + { + "epoch": 9.715305313243457, + "grad_norm": 0.01724371498787081, + "learning_rate": 4.01077558101602e-05, + "loss": 0.0001, + "step": 6130 + }, + { + "epoch": 9.731165741475019, + "grad_norm": 0.00656744352318638, + "learning_rate": 3.9938351606523596e-05, + "loss": 0.0001, + "step": 6140 + }, + { + "epoch": 9.747026169706583, + "grad_norm": 0.013479850838575475, + "learning_rate": 3.976906769107549e-05, + "loss": 0.0001, + "step": 6150 + }, + { + "epoch": 9.762886597938145, + "grad_norm": 0.01741781856533647, + "learning_rate": 3.959990608762494e-05, + "loss": 0.0001, + "step": 6160 + }, + { + "epoch": 9.778747026169707, + "grad_norm": 0.016255454768973686, + "learning_rate": 3.9430868818518784e-05, + "loss": 0.0001, + "step": 6170 + }, + { + "epoch": 9.79460745440127, + "grad_norm": 0.00787922616999963, + "learning_rate": 3.9261957904617466e-05, + "loss": 0.0001, + "step": 6180 + }, + { + "epoch": 9.810467882632832, + "grad_norm": 0.007255011379134003, + "learning_rate": 3.909317536527075e-05, + "loss": 0.0001, + "step": 6190 + }, + { + "epoch": 9.826328310864394, + "grad_norm": 0.018880865296539522, + "learning_rate": 3.8924523218293787e-05, + "loss": 0.0001, + "step": 6200 + }, + { + "epoch": 9.842188739095956, + "grad_norm": 0.01242209534805015, + "learning_rate": 3.875600347994275e-05, + "loss": 0.0001, + "step": 6210 + }, + { + "epoch": 9.858049167327518, + "grad_norm": 0.015180238496601505, + "learning_rate": 3.8587618164890907e-05, + "loss": 0.0001, + "step": 6220 + }, + { + "epoch": 9.87390959555908, + "grad_norm": 0.01049947114198914, + "learning_rate": 3.8419369286204534e-05, + "loss": 0.0001, + "step": 6230 + }, + { + "epoch": 9.889770023790643, + "grad_norm": 0.015641675155518057, + "learning_rate": 3.825125885531868e-05, + "loss": 0.0001, + "step": 6240 + }, + { + "epoch": 9.905630452022205, + "grad_norm": 0.01811051927916764, + "learning_rate": 3.8083288882013316e-05, + "loss": 0.0001, + "step": 6250 + }, + { + "epoch": 9.921490880253767, + "grad_norm": 0.030805915442547793, + "learning_rate": 3.7915461374389227e-05, + "loss": 0.0001, + "step": 6260 + }, + { + "epoch": 9.93735130848533, + "grad_norm": 0.007200916416555831, + "learning_rate": 3.774777833884395e-05, + "loss": 0.0001, + "step": 6270 + }, + { + "epoch": 9.953211736716892, + "grad_norm": 0.01783746530698663, + "learning_rate": 3.758024178004789e-05, + "loss": 0.0001, + "step": 6280 + }, + { + "epoch": 9.969072164948454, + "grad_norm": 0.01383981197432044, + "learning_rate": 3.741285370092026e-05, + "loss": 0.0001, + "step": 6290 + }, + { + "epoch": 9.984932593180016, + "grad_norm": 0.017915471404985086, + "learning_rate": 3.724561610260523e-05, + "loss": 0.0001, + "step": 6300 + }, + { + "epoch": 10.0, + "grad_norm": 0.048513989685072846, + "learning_rate": 3.707853098444796e-05, + "loss": 0.0001, + "step": 6310 + }, + { + "epoch": 10.015860428231562, + "grad_norm": 0.015882920455193537, + "learning_rate": 3.69116003439706e-05, + "loss": 0.0001, + "step": 6320 + }, + { + "epoch": 10.031720856463124, + "grad_norm": 0.02465704063101723, + "learning_rate": 3.674482617684861e-05, + "loss": 0.0001, + "step": 6330 + }, + { + "epoch": 10.047581284694687, + "grad_norm": 0.013509076164926406, + "learning_rate": 3.6578210476886775e-05, + "loss": 0.0001, + "step": 6340 + }, + { + "epoch": 10.063441712926249, + "grad_norm": 0.00880771154115395, + "learning_rate": 3.641175523599529e-05, + "loss": 0.0001, + "step": 6350 + }, + { + "epoch": 10.079302141157811, + "grad_norm": 0.001926076471738476, + "learning_rate": 3.6245462444166175e-05, + "loss": 0.0001, + "step": 6360 + }, + { + "epoch": 10.095162569389373, + "grad_norm": 0.019748260923575706, + "learning_rate": 3.607933408944924e-05, + "loss": 0.0001, + "step": 6370 + }, + { + "epoch": 10.111022997620935, + "grad_norm": 0.009785017441099891, + "learning_rate": 3.591337215792852e-05, + "loss": 0.0001, + "step": 6380 + }, + { + "epoch": 10.126883425852498, + "grad_norm": 0.0015872279962018477, + "learning_rate": 3.574757863369839e-05, + "loss": 0.0001, + "step": 6390 + }, + { + "epoch": 10.14274385408406, + "grad_norm": 0.002694147569336284, + "learning_rate": 3.558195549883989e-05, + "loss": 0.0001, + "step": 6400 + }, + { + "epoch": 10.158604282315622, + "grad_norm": 0.003897827079319857, + "learning_rate": 3.541650473339708e-05, + "loss": 0.0001, + "step": 6410 + }, + { + "epoch": 10.174464710547184, + "grad_norm": 0.0033539221329021776, + "learning_rate": 3.525122831535329e-05, + "loss": 0.0001, + "step": 6420 + }, + { + "epoch": 10.190325138778746, + "grad_norm": 0.0012015723068294598, + "learning_rate": 3.508612822060748e-05, + "loss": 0.0001, + "step": 6430 + }, + { + "epoch": 10.206185567010309, + "grad_norm": 0.001626407140809797, + "learning_rate": 3.4921206422950734e-05, + "loss": 0.0001, + "step": 6440 + }, + { + "epoch": 10.22204599524187, + "grad_norm": 0.014589802413645669, + "learning_rate": 3.4756464894042476e-05, + "loss": 0.0001, + "step": 6450 + }, + { + "epoch": 10.237906423473433, + "grad_norm": 0.0019268163191054712, + "learning_rate": 3.459190560338703e-05, + "loss": 0.0001, + "step": 6460 + }, + { + "epoch": 10.253766851704995, + "grad_norm": 0.005250410120300828, + "learning_rate": 3.442753051831008e-05, + "loss": 0.0001, + "step": 6470 + }, + { + "epoch": 10.269627279936559, + "grad_norm": 0.012176163858304941, + "learning_rate": 3.4263341603935054e-05, + "loss": 0.0001, + "step": 6480 + }, + { + "epoch": 10.285487708168121, + "grad_norm": 0.0057865020813030275, + "learning_rate": 3.40993408231597e-05, + "loss": 0.0001, + "step": 6490 + }, + { + "epoch": 10.301348136399683, + "grad_norm": 0.001313323412254081, + "learning_rate": 3.393553013663265e-05, + "loss": 0.0001, + "step": 6500 + }, + { + "epoch": 10.317208564631246, + "grad_norm": 0.013253461985932828, + "learning_rate": 3.377191150272987e-05, + "loss": 0.0001, + "step": 6510 + }, + { + "epoch": 10.333068992862808, + "grad_norm": 0.012293688840456195, + "learning_rate": 3.3608486877531375e-05, + "loss": 0.0001, + "step": 6520 + }, + { + "epoch": 10.34892942109437, + "grad_norm": 0.001894368787127009, + "learning_rate": 3.3445258214797745e-05, + "loss": 0.0001, + "step": 6530 + }, + { + "epoch": 10.364789849325932, + "grad_norm": 0.004572053960903994, + "learning_rate": 3.3282227465946805e-05, + "loss": 0.0001, + "step": 6540 + }, + { + "epoch": 10.380650277557494, + "grad_norm": 0.012559704670048103, + "learning_rate": 3.311939658003035e-05, + "loss": 0.0001, + "step": 6550 + }, + { + "epoch": 10.396510705789057, + "grad_norm": 0.007442322253800473, + "learning_rate": 3.295676750371069e-05, + "loss": 0.0001, + "step": 6560 + }, + { + "epoch": 10.412371134020619, + "grad_norm": 0.019188819116715206, + "learning_rate": 3.279434218123758e-05, + "loss": 0.0001, + "step": 6570 + }, + { + "epoch": 10.428231562252181, + "grad_norm": 0.0012193029125271392, + "learning_rate": 3.263212255442482e-05, + "loss": 0.0001, + "step": 6580 + }, + { + "epoch": 10.444091990483743, + "grad_norm": 0.004512847878631283, + "learning_rate": 3.24701105626271e-05, + "loss": 0.0001, + "step": 6590 + }, + { + "epoch": 10.459952418715305, + "grad_norm": 0.007730488048527958, + "learning_rate": 3.230830814271682e-05, + "loss": 0.0001, + "step": 6600 + }, + { + "epoch": 10.475812846946868, + "grad_norm": 0.014518681493891773, + "learning_rate": 3.2146717229060885e-05, + "loss": 0.0001, + "step": 6610 + }, + { + "epoch": 10.49167327517843, + "grad_norm": 0.0021428995181564333, + "learning_rate": 3.198533975349769e-05, + "loss": 0.0001, + "step": 6620 + }, + { + "epoch": 10.507533703409992, + "grad_norm": 0.001288608440403313, + "learning_rate": 3.182417764531391e-05, + "loss": 0.0001, + "step": 6630 + }, + { + "epoch": 10.523394131641554, + "grad_norm": 0.006710938876354956, + "learning_rate": 3.1663232831221444e-05, + "loss": 0.0001, + "step": 6640 + }, + { + "epoch": 10.539254559873116, + "grad_norm": 0.002898039090723338, + "learning_rate": 3.1502507235334475e-05, + "loss": 0.0001, + "step": 6650 + }, + { + "epoch": 10.555114988104679, + "grad_norm": 0.01431801920022144, + "learning_rate": 3.13420027791464e-05, + "loss": 0.0001, + "step": 6660 + }, + { + "epoch": 10.57097541633624, + "grad_norm": 0.0017379780305701467, + "learning_rate": 3.1181721381506815e-05, + "loss": 0.0001, + "step": 6670 + }, + { + "epoch": 10.586835844567803, + "grad_norm": 0.0016452454375734107, + "learning_rate": 3.102166495859868e-05, + "loss": 0.0001, + "step": 6680 + }, + { + "epoch": 10.602696272799365, + "grad_norm": 0.004246553293553657, + "learning_rate": 3.086183542391531e-05, + "loss": 0.0001, + "step": 6690 + }, + { + "epoch": 10.618556701030927, + "grad_norm": 0.0017250360335069731, + "learning_rate": 3.070223468823757e-05, + "loss": 0.0001, + "step": 6700 + }, + { + "epoch": 10.63441712926249, + "grad_norm": 0.003155091734930943, + "learning_rate": 3.0542864659611034e-05, + "loss": 0.0001, + "step": 6710 + }, + { + "epoch": 10.650277557494052, + "grad_norm": 0.004080109669867088, + "learning_rate": 3.0383727243323078e-05, + "loss": 0.0001, + "step": 6720 + }, + { + "epoch": 10.666137985725614, + "grad_norm": 0.005452539326693235, + "learning_rate": 3.0224824341880226e-05, + "loss": 0.0001, + "step": 6730 + }, + { + "epoch": 10.681998413957176, + "grad_norm": 0.0015656250709943642, + "learning_rate": 3.006615785498534e-05, + "loss": 0.0001, + "step": 6740 + }, + { + "epoch": 10.697858842188738, + "grad_norm": 0.0016176194805711623, + "learning_rate": 2.9907729679514873e-05, + "loss": 0.0001, + "step": 6750 + }, + { + "epoch": 10.7137192704203, + "grad_norm": 0.002261640560818015, + "learning_rate": 2.974954170949632e-05, + "loss": 0.0001, + "step": 6760 + }, + { + "epoch": 10.729579698651863, + "grad_norm": 0.006211772589291617, + "learning_rate": 2.9591595836085405e-05, + "loss": 0.0001, + "step": 6770 + }, + { + "epoch": 10.745440126883427, + "grad_norm": 0.0031969588156165953, + "learning_rate": 2.9433893947543612e-05, + "loss": 0.0001, + "step": 6780 + }, + { + "epoch": 10.761300555114989, + "grad_norm": 0.0026909446474992593, + "learning_rate": 2.927643792921556e-05, + "loss": 0.0001, + "step": 6790 + }, + { + "epoch": 10.777160983346551, + "grad_norm": 0.0011125617127922993, + "learning_rate": 2.9119229663506413e-05, + "loss": 0.0001, + "step": 6800 + }, + { + "epoch": 10.793021411578113, + "grad_norm": 0.002073693041114501, + "learning_rate": 2.8962271029859473e-05, + "loss": 0.0001, + "step": 6810 + }, + { + "epoch": 10.808881839809676, + "grad_norm": 0.011243624446544375, + "learning_rate": 2.8805563904733658e-05, + "loss": 0.0001, + "step": 6820 + }, + { + "epoch": 10.824742268041238, + "grad_norm": 0.0028396790265528245, + "learning_rate": 2.8649110161581046e-05, + "loss": 0.0001, + "step": 6830 + }, + { + "epoch": 10.8406026962728, + "grad_norm": 0.0051313694251390535, + "learning_rate": 2.8492911670824507e-05, + "loss": 0.0001, + "step": 6840 + }, + { + "epoch": 10.856463124504362, + "grad_norm": 0.001575248795011118, + "learning_rate": 2.8336970299835326e-05, + "loss": 0.0001, + "step": 6850 + }, + { + "epoch": 10.872323552735924, + "grad_norm": 0.0011774729153063985, + "learning_rate": 2.8181287912910937e-05, + "loss": 0.0001, + "step": 6860 + }, + { + "epoch": 10.888183980967487, + "grad_norm": 0.0011686837800266008, + "learning_rate": 2.8025866371252578e-05, + "loss": 0.0001, + "step": 6870 + }, + { + "epoch": 10.904044409199049, + "grad_norm": 0.0014237956882786743, + "learning_rate": 2.7870707532943003e-05, + "loss": 0.0001, + "step": 6880 + }, + { + "epoch": 10.919904837430611, + "grad_norm": 0.003304941101179277, + "learning_rate": 2.7715813252924382e-05, + "loss": 0.0001, + "step": 6890 + }, + { + "epoch": 10.935765265662173, + "grad_norm": 0.003991978377636293, + "learning_rate": 2.756118538297602e-05, + "loss": 0.0001, + "step": 6900 + }, + { + "epoch": 10.951625693893735, + "grad_norm": 0.0009593607391887802, + "learning_rate": 2.7406825771692235e-05, + "loss": 0.0001, + "step": 6910 + }, + { + "epoch": 10.967486122125297, + "grad_norm": 0.0017054416684389037, + "learning_rate": 2.7252736264460355e-05, + "loss": 0.0001, + "step": 6920 + }, + { + "epoch": 10.98334655035686, + "grad_norm": 0.0011973324724428925, + "learning_rate": 2.7098918703438503e-05, + "loss": 0.0001, + "step": 6930 + }, + { + "epoch": 10.999206978588422, + "grad_norm": 0.0016447512144054772, + "learning_rate": 2.69453749275337e-05, + "loss": 0.0001, + "step": 6940 + }, + { + "epoch": 11.014274385408406, + "grad_norm": 0.00122988151888232, + "learning_rate": 2.6792106772379855e-05, + "loss": 0.0, + "step": 6950 + }, + { + "epoch": 11.030134813639968, + "grad_norm": 0.0008372807866781568, + "learning_rate": 2.663911607031575e-05, + "loss": 0.0001, + "step": 6960 + }, + { + "epoch": 11.04599524187153, + "grad_norm": 0.0010517408630153337, + "learning_rate": 2.648640465036316e-05, + "loss": 0.0, + "step": 6970 + }, + { + "epoch": 11.061855670103093, + "grad_norm": 0.0010159115264744957, + "learning_rate": 2.6333974338205115e-05, + "loss": 0.0, + "step": 6980 + }, + { + "epoch": 11.077716098334655, + "grad_norm": 0.0013402878771766342, + "learning_rate": 2.6181826956163834e-05, + "loss": 0.0, + "step": 6990 + }, + { + "epoch": 11.093576526566217, + "grad_norm": 0.0013236645327339407, + "learning_rate": 2.6029964323179212e-05, + "loss": 0.0, + "step": 7000 + }, + { + "epoch": 11.10943695479778, + "grad_norm": 0.0008709260235887611, + "learning_rate": 2.58783882547868e-05, + "loss": 0.0001, + "step": 7010 + }, + { + "epoch": 11.125297383029341, + "grad_norm": 0.00318716467077791, + "learning_rate": 2.572710056309634e-05, + "loss": 0.0, + "step": 7020 + }, + { + "epoch": 11.141157811260904, + "grad_norm": 0.0007593490747475647, + "learning_rate": 2.5576103056769966e-05, + "loss": 0.0, + "step": 7030 + }, + { + "epoch": 11.157018239492466, + "grad_norm": 0.0012568060286446703, + "learning_rate": 2.542539754100059e-05, + "loss": 0.0, + "step": 7040 + }, + { + "epoch": 11.172878667724028, + "grad_norm": 0.0008809127698353513, + "learning_rate": 2.5274985817490325e-05, + "loss": 0.0, + "step": 7050 + }, + { + "epoch": 11.18873909595559, + "grad_norm": 0.0009146533988000729, + "learning_rate": 2.5124869684429055e-05, + "loss": 0.0001, + "step": 7060 + }, + { + "epoch": 11.204599524187152, + "grad_norm": 0.0009913828797594153, + "learning_rate": 2.4975050936472722e-05, + "loss": 0.0, + "step": 7070 + }, + { + "epoch": 11.220459952418715, + "grad_norm": 0.0006974763545147028, + "learning_rate": 2.482553136472211e-05, + "loss": 0.0, + "step": 7080 + }, + { + "epoch": 11.236320380650277, + "grad_norm": 0.0006636508601054656, + "learning_rate": 2.467631275670122e-05, + "loss": 0.0, + "step": 7090 + }, + { + "epoch": 11.252180808881839, + "grad_norm": 0.000753514759535347, + "learning_rate": 2.452739689633608e-05, + "loss": 0.0, + "step": 7100 + }, + { + "epoch": 11.268041237113403, + "grad_norm": 0.000833275720492216, + "learning_rate": 2.4378785563933276e-05, + "loss": 0.0, + "step": 7110 + }, + { + "epoch": 11.283901665344965, + "grad_norm": 0.000763911471295781, + "learning_rate": 2.4230480536158716e-05, + "loss": 0.0, + "step": 7120 + }, + { + "epoch": 11.299762093576527, + "grad_norm": 0.0008228091583204328, + "learning_rate": 2.4082483586016457e-05, + "loss": 0.0, + "step": 7130 + }, + { + "epoch": 11.31562252180809, + "grad_norm": 0.0008142276092776194, + "learning_rate": 2.3934796482827437e-05, + "loss": 0.0, + "step": 7140 + }, + { + "epoch": 11.331482950039652, + "grad_norm": 0.0007071966625826064, + "learning_rate": 2.3787420992208286e-05, + "loss": 0.0, + "step": 7150 + }, + { + "epoch": 11.347343378271214, + "grad_norm": 0.0010138419044587433, + "learning_rate": 2.3640358876050344e-05, + "loss": 0.0, + "step": 7160 + }, + { + "epoch": 11.363203806502776, + "grad_norm": 0.0026270501999016223, + "learning_rate": 2.3493611892498445e-05, + "loss": 0.0, + "step": 7170 + }, + { + "epoch": 11.379064234734338, + "grad_norm": 0.0007188509668076911, + "learning_rate": 2.334718179593006e-05, + "loss": 0.0, + "step": 7180 + }, + { + "epoch": 11.3949246629659, + "grad_norm": 0.0007698864685044069, + "learning_rate": 2.3201070336934174e-05, + "loss": 0.0, + "step": 7190 + }, + { + "epoch": 11.410785091197463, + "grad_norm": 0.0006111203589069173, + "learning_rate": 2.305527926229042e-05, + "loss": 0.0, + "step": 7200 + }, + { + "epoch": 11.426645519429025, + "grad_norm": 0.0006988055199523409, + "learning_rate": 2.290981031494824e-05, + "loss": 0.0, + "step": 7210 + }, + { + "epoch": 11.442505947660587, + "grad_norm": 0.016893488718737668, + "learning_rate": 2.2764665234006004e-05, + "loss": 0.0, + "step": 7220 + }, + { + "epoch": 11.45836637589215, + "grad_norm": 0.0008081786366502929, + "learning_rate": 2.2619845754690177e-05, + "loss": 0.0, + "step": 7230 + }, + { + "epoch": 11.474226804123711, + "grad_norm": 0.0007900638872051926, + "learning_rate": 2.2475353608334675e-05, + "loss": 0.0, + "step": 7240 + }, + { + "epoch": 11.490087232355274, + "grad_norm": 0.0017339601208347427, + "learning_rate": 2.2331190522360062e-05, + "loss": 0.0, + "step": 7250 + }, + { + "epoch": 11.505947660586836, + "grad_norm": 0.0006428905123719451, + "learning_rate": 2.2187358220252956e-05, + "loss": 0.0, + "step": 7260 + }, + { + "epoch": 11.521808088818398, + "grad_norm": 0.0007729081222006382, + "learning_rate": 2.2043858421545448e-05, + "loss": 0.0, + "step": 7270 + }, + { + "epoch": 11.53766851704996, + "grad_norm": 0.0006472616618411824, + "learning_rate": 2.1900692841794462e-05, + "loss": 0.0, + "step": 7280 + }, + { + "epoch": 11.553528945281522, + "grad_norm": 0.0007463290955280325, + "learning_rate": 2.1757863192561356e-05, + "loss": 0.0, + "step": 7290 + }, + { + "epoch": 11.569389373513085, + "grad_norm": 0.000640768063312241, + "learning_rate": 2.1615371181391323e-05, + "loss": 0.0, + "step": 7300 + }, + { + "epoch": 11.585249801744647, + "grad_norm": 0.0008649093425033468, + "learning_rate": 2.1473218511793118e-05, + "loss": 0.0, + "step": 7310 + }, + { + "epoch": 11.601110229976209, + "grad_norm": 0.0023479893696152407, + "learning_rate": 2.133140688321862e-05, + "loss": 0.0, + "step": 7320 + }, + { + "epoch": 11.616970658207771, + "grad_norm": 0.0011513758553891341, + "learning_rate": 2.118993799104249e-05, + "loss": 0.0001, + "step": 7330 + }, + { + "epoch": 11.632831086439333, + "grad_norm": 0.0007801022903284298, + "learning_rate": 2.1048813526541938e-05, + "loss": 0.0001, + "step": 7340 + }, + { + "epoch": 11.648691514670896, + "grad_norm": 0.0007864488862208204, + "learning_rate": 2.090803517687654e-05, + "loss": 0.0, + "step": 7350 + }, + { + "epoch": 11.664551942902458, + "grad_norm": 0.00165073190033002, + "learning_rate": 2.0767604625067982e-05, + "loss": 0.0, + "step": 7360 + }, + { + "epoch": 11.68041237113402, + "grad_norm": 0.0012530165783088182, + "learning_rate": 2.0627523549980026e-05, + "loss": 0.0, + "step": 7370 + }, + { + "epoch": 11.696272799365582, + "grad_norm": 0.0008257649600189163, + "learning_rate": 2.048779362629836e-05, + "loss": 0.0, + "step": 7380 + }, + { + "epoch": 11.712133227597144, + "grad_norm": 0.000937065728450216, + "learning_rate": 2.0348416524510665e-05, + "loss": 0.0, + "step": 7390 + }, + { + "epoch": 11.727993655828707, + "grad_norm": 0.0030152908966149627, + "learning_rate": 2.0209393910886544e-05, + "loss": 0.0, + "step": 7400 + }, + { + "epoch": 11.74385408406027, + "grad_norm": 0.0007429120145106313, + "learning_rate": 2.007072744745766e-05, + "loss": 0.0, + "step": 7410 + }, + { + "epoch": 11.759714512291833, + "grad_norm": 0.0008039698081901948, + "learning_rate": 1.9932418791997888e-05, + "loss": 0.0, + "step": 7420 + }, + { + "epoch": 11.775574940523395, + "grad_norm": 0.0005937477996651571, + "learning_rate": 1.979446959800347e-05, + "loss": 0.0, + "step": 7430 + }, + { + "epoch": 11.791435368754957, + "grad_norm": 0.0006174327498150833, + "learning_rate": 1.9656881514673186e-05, + "loss": 0.0, + "step": 7440 + }, + { + "epoch": 11.80729579698652, + "grad_norm": 0.0032656968281388823, + "learning_rate": 1.9519656186888775e-05, + "loss": 0.0, + "step": 7450 + }, + { + "epoch": 11.823156225218082, + "grad_norm": 0.0006629662009529402, + "learning_rate": 1.9382795255195113e-05, + "loss": 0.0, + "step": 7460 + }, + { + "epoch": 11.839016653449644, + "grad_norm": 0.000599345112234113, + "learning_rate": 1.924630035578075e-05, + "loss": 0.0, + "step": 7470 + }, + { + "epoch": 11.854877081681206, + "grad_norm": 0.0006248799824189912, + "learning_rate": 1.9110173120458224e-05, + "loss": 0.0, + "step": 7480 + }, + { + "epoch": 11.870737509912768, + "grad_norm": 0.000985308954047774, + "learning_rate": 1.8974415176644605e-05, + "loss": 0.0, + "step": 7490 + }, + { + "epoch": 11.88659793814433, + "grad_norm": 0.0012055416821135337, + "learning_rate": 1.8839028147342086e-05, + "loss": 0.0, + "step": 7500 + }, + { + "epoch": 11.902458366375892, + "grad_norm": 0.0008391661482249922, + "learning_rate": 1.8704013651118507e-05, + "loss": 0.0, + "step": 7510 + }, + { + "epoch": 11.918318794607455, + "grad_norm": 0.0006153558626898472, + "learning_rate": 1.856937330208801e-05, + "loss": 0.0001, + "step": 7520 + }, + { + "epoch": 11.934179222839017, + "grad_norm": 0.0007524412614729978, + "learning_rate": 1.8435108709891812e-05, + "loss": 0.0, + "step": 7530 + }, + { + "epoch": 11.950039651070579, + "grad_norm": 0.0009467360240393739, + "learning_rate": 1.8301221479678865e-05, + "loss": 0.0001, + "step": 7540 + }, + { + "epoch": 11.965900079302141, + "grad_norm": 0.0008017867114730641, + "learning_rate": 1.8167713212086724e-05, + "loss": 0.0, + "step": 7550 + }, + { + "epoch": 11.981760507533703, + "grad_norm": 0.0015670109649585377, + "learning_rate": 1.803458550322244e-05, + "loss": 0.0, + "step": 7560 + }, + { + "epoch": 11.997620935765266, + "grad_norm": 0.0007877123459260931, + "learning_rate": 1.7901839944643373e-05, + "loss": 0.0, + "step": 7570 + }, + { + "epoch": 12.01268834258525, + "grad_norm": 0.0006408553165267422, + "learning_rate": 1.7769478123338297e-05, + "loss": 0.0, + "step": 7580 + }, + { + "epoch": 12.028548770816812, + "grad_norm": 0.0006621759688130016, + "learning_rate": 1.763750162170834e-05, + "loss": 0.0, + "step": 7590 + }, + { + "epoch": 12.044409199048374, + "grad_norm": 0.0013250385104682275, + "learning_rate": 1.7505912017548053e-05, + "loss": 0.0, + "step": 7600 + }, + { + "epoch": 12.060269627279936, + "grad_norm": 0.0006116985939188641, + "learning_rate": 1.7374710884026653e-05, + "loss": 0.0, + "step": 7610 + }, + { + "epoch": 12.076130055511499, + "grad_norm": 0.0006929162962024774, + "learning_rate": 1.7243899789669067e-05, + "loss": 0.0, + "step": 7620 + }, + { + "epoch": 12.09199048374306, + "grad_norm": 0.0008647416885407059, + "learning_rate": 1.7113480298337296e-05, + "loss": 0.0, + "step": 7630 + }, + { + "epoch": 12.107850911974623, + "grad_norm": 0.0005273606961786284, + "learning_rate": 1.6983453969211706e-05, + "loss": 0.0, + "step": 7640 + }, + { + "epoch": 12.123711340206185, + "grad_norm": 0.0006502034679140801, + "learning_rate": 1.6853822356772298e-05, + "loss": 0.0, + "step": 7650 + }, + { + "epoch": 12.139571768437747, + "grad_norm": 0.004897352253480797, + "learning_rate": 1.6724587010780253e-05, + "loss": 0.0, + "step": 7660 + }, + { + "epoch": 12.15543219666931, + "grad_norm": 0.0005229952574958541, + "learning_rate": 1.6595749476259316e-05, + "loss": 0.0, + "step": 7670 + }, + { + "epoch": 12.171292624900872, + "grad_norm": 0.0007246796761821938, + "learning_rate": 1.646731129347734e-05, + "loss": 0.0, + "step": 7680 + }, + { + "epoch": 12.187153053132434, + "grad_norm": 0.0005563803555127946, + "learning_rate": 1.633927399792789e-05, + "loss": 0.0, + "step": 7690 + }, + { + "epoch": 12.203013481363996, + "grad_norm": 0.0006051234238154342, + "learning_rate": 1.621163912031185e-05, + "loss": 0.0, + "step": 7700 + }, + { + "epoch": 12.218873909595558, + "grad_norm": 0.0006060997435392572, + "learning_rate": 1.6084408186519196e-05, + "loss": 0.0, + "step": 7710 + }, + { + "epoch": 12.23473433782712, + "grad_norm": 0.000549769537929422, + "learning_rate": 1.5957582717610702e-05, + "loss": 0.0, + "step": 7720 + }, + { + "epoch": 12.250594766058683, + "grad_norm": 0.0005480156725345165, + "learning_rate": 1.5831164229799717e-05, + "loss": 0.0, + "step": 7730 + }, + { + "epoch": 12.266455194290247, + "grad_norm": 0.0011663340267657229, + "learning_rate": 1.5705154234434123e-05, + "loss": 0.0, + "step": 7740 + }, + { + "epoch": 12.282315622521809, + "grad_norm": 0.0006263150403301365, + "learning_rate": 1.5579554237978257e-05, + "loss": 0.0, + "step": 7750 + }, + { + "epoch": 12.298176050753371, + "grad_norm": 0.0006493568988121673, + "learning_rate": 1.545436574199476e-05, + "loss": 0.0, + "step": 7760 + }, + { + "epoch": 12.314036478984933, + "grad_norm": 0.000587651187467798, + "learning_rate": 1.5329590243126858e-05, + "loss": 0.0, + "step": 7770 + }, + { + "epoch": 12.329896907216495, + "grad_norm": 0.000531904463246674, + "learning_rate": 1.5205229233080264e-05, + "loss": 0.0, + "step": 7780 + }, + { + "epoch": 12.345757335448058, + "grad_norm": 0.0005321417645498957, + "learning_rate": 1.5081284198605477e-05, + "loss": 0.0, + "step": 7790 + }, + { + "epoch": 12.36161776367962, + "grad_norm": 0.0005576595089753946, + "learning_rate": 1.4957756621479963e-05, + "loss": 0.0, + "step": 7800 + }, + { + "epoch": 12.377478191911182, + "grad_norm": 0.012481562792768068, + "learning_rate": 1.4834647978490384e-05, + "loss": 0.0, + "step": 7810 + }, + { + "epoch": 12.393338620142744, + "grad_norm": 0.0004990009003624279, + "learning_rate": 1.4711959741415061e-05, + "loss": 0.0, + "step": 7820 + }, + { + "epoch": 12.409199048374306, + "grad_norm": 0.0005289270875534441, + "learning_rate": 1.4589693377006258e-05, + "loss": 0.0, + "step": 7830 + }, + { + "epoch": 12.425059476605869, + "grad_norm": 0.000713553827980068, + "learning_rate": 1.4467850346972717e-05, + "loss": 0.0, + "step": 7840 + }, + { + "epoch": 12.44091990483743, + "grad_norm": 0.0006344672920932825, + "learning_rate": 1.4346432107962204e-05, + "loss": 0.0, + "step": 7850 + }, + { + "epoch": 12.456780333068993, + "grad_norm": 0.0007020742031333008, + "learning_rate": 1.4225440111543992e-05, + "loss": 0.0, + "step": 7860 + }, + { + "epoch": 12.472640761300555, + "grad_norm": 0.0005612679139856495, + "learning_rate": 1.4104875804191636e-05, + "loss": 0.0, + "step": 7870 + }, + { + "epoch": 12.488501189532117, + "grad_norm": 0.0005940882359651229, + "learning_rate": 1.3984740627265597e-05, + "loss": 0.0, + "step": 7880 + }, + { + "epoch": 12.50436161776368, + "grad_norm": 0.0004450902741404186, + "learning_rate": 1.386503601699602e-05, + "loss": 0.0, + "step": 7890 + }, + { + "epoch": 12.520222045995242, + "grad_norm": 0.0005246472599928517, + "learning_rate": 1.3745763404465568e-05, + "loss": 0.0, + "step": 7900 + }, + { + "epoch": 12.536082474226804, + "grad_norm": 0.000435481998790582, + "learning_rate": 1.3626924215592357e-05, + "loss": 0.0, + "step": 7910 + }, + { + "epoch": 12.551942902458366, + "grad_norm": 0.000550042236247132, + "learning_rate": 1.3508519871112828e-05, + "loss": 0.0, + "step": 7920 + }, + { + "epoch": 12.567803330689928, + "grad_norm": 0.002391025755020568, + "learning_rate": 1.3390551786564854e-05, + "loss": 0.0, + "step": 7930 + }, + { + "epoch": 12.58366375892149, + "grad_norm": 0.0005200104933384467, + "learning_rate": 1.3273021372270712e-05, + "loss": 0.0, + "step": 7940 + }, + { + "epoch": 12.599524187153053, + "grad_norm": 0.0005645730991029861, + "learning_rate": 1.3155930033320329e-05, + "loss": 0.0, + "step": 7950 + }, + { + "epoch": 12.615384615384615, + "grad_norm": 0.0005276296367840765, + "learning_rate": 1.303927916955443e-05, + "loss": 0.0, + "step": 7960 + }, + { + "epoch": 12.631245043616177, + "grad_norm": 0.0004569392785250703, + "learning_rate": 1.292307017554778e-05, + "loss": 0.0, + "step": 7970 + }, + { + "epoch": 12.64710547184774, + "grad_norm": 0.0005575040956810366, + "learning_rate": 1.2807304440592544e-05, + "loss": 0.0, + "step": 7980 + }, + { + "epoch": 12.662965900079302, + "grad_norm": 0.0006411509637369894, + "learning_rate": 1.2691983348681697e-05, + "loss": 0.0, + "step": 7990 + }, + { + "epoch": 12.678826328310864, + "grad_norm": 0.0006684169658296058, + "learning_rate": 1.2577108278492423e-05, + "loss": 0.0, + "step": 8000 + }, + { + "epoch": 12.694686756542426, + "grad_norm": 0.00046815016929679004, + "learning_rate": 1.2462680603369709e-05, + "loss": 0.0, + "step": 8010 + }, + { + "epoch": 12.710547184773988, + "grad_norm": 0.0005283269036916359, + "learning_rate": 1.2348701691309816e-05, + "loss": 0.0, + "step": 8020 + }, + { + "epoch": 12.72640761300555, + "grad_norm": 0.0005450507179724664, + "learning_rate": 1.2235172904944059e-05, + "loss": 0.0, + "step": 8030 + }, + { + "epoch": 12.742268041237114, + "grad_norm": 0.0004705894473759728, + "learning_rate": 1.2122095601522404e-05, + "loss": 0.0, + "step": 8040 + }, + { + "epoch": 12.758128469468677, + "grad_norm": 0.000708525247575093, + "learning_rate": 1.2009471132897277e-05, + "loss": 0.0, + "step": 8050 + }, + { + "epoch": 12.773988897700239, + "grad_norm": 0.0004657681468817712, + "learning_rate": 1.1897300845507448e-05, + "loss": 0.0, + "step": 8060 + }, + { + "epoch": 12.789849325931801, + "grad_norm": 0.0005799396931900242, + "learning_rate": 1.1785586080361893e-05, + "loss": 0.0, + "step": 8070 + }, + { + "epoch": 12.805709754163363, + "grad_norm": 0.0009191733236490015, + "learning_rate": 1.1674328173023724e-05, + "loss": 0.0, + "step": 8080 + }, + { + "epoch": 12.821570182394925, + "grad_norm": 0.0004530175050239461, + "learning_rate": 1.1563528453594319e-05, + "loss": 0.0, + "step": 8090 + }, + { + "epoch": 12.837430610626487, + "grad_norm": 0.0005609967783917278, + "learning_rate": 1.145318824669731e-05, + "loss": 0.0, + "step": 8100 + }, + { + "epoch": 12.85329103885805, + "grad_norm": 0.0005293020766193651, + "learning_rate": 1.1343308871462865e-05, + "loss": 0.0, + "step": 8110 + }, + { + "epoch": 12.869151467089612, + "grad_norm": 0.0005093980393220007, + "learning_rate": 1.1233891641511807e-05, + "loss": 0.0, + "step": 8120 + }, + { + "epoch": 12.885011895321174, + "grad_norm": 0.0005482671941288381, + "learning_rate": 1.1124937864939954e-05, + "loss": 0.0, + "step": 8130 + }, + { + "epoch": 12.900872323552736, + "grad_norm": 0.00048154527901776397, + "learning_rate": 1.1016448844302524e-05, + "loss": 0.0, + "step": 8140 + }, + { + "epoch": 12.916732751784298, + "grad_norm": 0.0005370489176295048, + "learning_rate": 1.090842587659851e-05, + "loss": 0.0, + "step": 8150 + }, + { + "epoch": 12.93259318001586, + "grad_norm": 0.0006676805657430731, + "learning_rate": 1.0800870253255163e-05, + "loss": 0.0, + "step": 8160 + }, + { + "epoch": 12.948453608247423, + "grad_norm": 0.0005173209013713772, + "learning_rate": 1.0693783260112611e-05, + "loss": 0.0, + "step": 8170 + }, + { + "epoch": 12.964314036478985, + "grad_norm": 0.00048466884305322494, + "learning_rate": 1.0587166177408431e-05, + "loss": 0.0, + "step": 8180 + }, + { + "epoch": 12.980174464710547, + "grad_norm": 0.0005297909219785814, + "learning_rate": 1.048102027976235e-05, + "loss": 0.0, + "step": 8190 + }, + { + "epoch": 12.99603489294211, + "grad_norm": 0.00041523520885121813, + "learning_rate": 1.037534683616107e-05, + "loss": 0.0, + "step": 8200 + }, + { + "epoch": 13.011102299762094, + "grad_norm": 0.00042707561341214417, + "learning_rate": 1.0270147109943007e-05, + "loss": 0.0, + "step": 8210 + }, + { + "epoch": 13.026962727993656, + "grad_norm": 0.00038465630438194953, + "learning_rate": 1.0165422358783244e-05, + "loss": 0.0, + "step": 8220 + }, + { + "epoch": 13.042823156225218, + "grad_norm": 0.0005096641371541986, + "learning_rate": 1.0061173834678512e-05, + "loss": 0.0, + "step": 8230 + }, + { + "epoch": 13.05868358445678, + "grad_norm": 0.0004197630408610639, + "learning_rate": 9.957402783932118e-06, + "loss": 0.0, + "step": 8240 + }, + { + "epoch": 13.074544012688342, + "grad_norm": 0.0004487032312096365, + "learning_rate": 9.85411044713921e-06, + "loss": 0.0, + "step": 8250 + }, + { + "epoch": 13.090404440919905, + "grad_norm": 0.0003975392814031434, + "learning_rate": 9.751298059171776e-06, + "loss": 0.0, + "step": 8260 + }, + { + "epoch": 13.106264869151467, + "grad_norm": 0.00045490229396858323, + "learning_rate": 9.648966849163988e-06, + "loss": 0.0, + "step": 8270 + }, + { + "epoch": 13.122125297383029, + "grad_norm": 0.0008014181311065558, + "learning_rate": 9.54711804049751e-06, + "loss": 0.0, + "step": 8280 + }, + { + "epoch": 13.137985725614591, + "grad_norm": 0.0004944383334737889, + "learning_rate": 9.445752850786783e-06, + "loss": 0.0, + "step": 8290 + }, + { + "epoch": 13.153846153846153, + "grad_norm": 0.00041297864666765724, + "learning_rate": 9.34487249186457e-06, + "loss": 0.0, + "step": 8300 + }, + { + "epoch": 13.169706582077715, + "grad_norm": 0.0005160671798759909, + "learning_rate": 9.244478169767428e-06, + "loss": 0.0, + "step": 8310 + }, + { + "epoch": 13.185567010309278, + "grad_norm": 0.0003810267863040334, + "learning_rate": 9.144571084721255e-06, + "loss": 0.0, + "step": 8320 + }, + { + "epoch": 13.20142743854084, + "grad_norm": 0.0005280906681418684, + "learning_rate": 9.045152431126997e-06, + "loss": 0.0, + "step": 8330 + }, + { + "epoch": 13.217287866772402, + "grad_norm": 0.0007190200681874975, + "learning_rate": 8.94622339754631e-06, + "loss": 0.0, + "step": 8340 + }, + { + "epoch": 13.233148295003964, + "grad_norm": 0.0004373840223195528, + "learning_rate": 8.847785166687433e-06, + "loss": 0.0, + "step": 8350 + }, + { + "epoch": 13.249008723235526, + "grad_norm": 0.0004224690175548741, + "learning_rate": 8.74983891539099e-06, + "loss": 0.0, + "step": 8360 + }, + { + "epoch": 13.26486915146709, + "grad_norm": 0.0004840303022170861, + "learning_rate": 8.652385814615898e-06, + "loss": 0.0, + "step": 8370 + }, + { + "epoch": 13.280729579698653, + "grad_norm": 0.0003781191509502434, + "learning_rate": 8.55542702942545e-06, + "loss": 0.0, + "step": 8380 + }, + { + "epoch": 13.296590007930215, + "grad_norm": 0.000480103420185671, + "learning_rate": 8.458963718973279e-06, + "loss": 0.0, + "step": 8390 + }, + { + "epoch": 13.312450436161777, + "grad_norm": 0.00045092873002070157, + "learning_rate": 8.362997036489628e-06, + "loss": 0.0, + "step": 8400 + }, + { + "epoch": 13.32831086439334, + "grad_norm": 0.00048079772507877527, + "learning_rate": 8.267528129267438e-06, + "loss": 0.0, + "step": 8410 + }, + { + "epoch": 13.344171292624901, + "grad_norm": 0.0004467143859473081, + "learning_rate": 8.172558138648706e-06, + "loss": 0.0, + "step": 8420 + }, + { + "epoch": 13.360031720856464, + "grad_norm": 0.00040523213377766267, + "learning_rate": 8.07808820001083e-06, + "loss": 0.0, + "step": 8430 + }, + { + "epoch": 13.375892149088026, + "grad_norm": 0.00044208052176154203, + "learning_rate": 7.984119442753041e-06, + "loss": 0.0, + "step": 8440 + }, + { + "epoch": 13.391752577319588, + "grad_norm": 0.0004223295473101108, + "learning_rate": 7.890652990282848e-06, + "loss": 0.0, + "step": 8450 + }, + { + "epoch": 13.40761300555115, + "grad_norm": 0.0004998410287756386, + "learning_rate": 7.797689960002685e-06, + "loss": 0.0, + "step": 8460 + }, + { + "epoch": 13.423473433782712, + "grad_norm": 0.0004392817416839357, + "learning_rate": 7.705231463296497e-06, + "loss": 0.0, + "step": 8470 + }, + { + "epoch": 13.439333862014275, + "grad_norm": 0.0005829726330667191, + "learning_rate": 7.613278605516455e-06, + "loss": 0.0, + "step": 8480 + }, + { + "epoch": 13.455194290245837, + "grad_norm": 0.0004244338310169516, + "learning_rate": 7.521832485969804e-06, + "loss": 0.0, + "step": 8490 + }, + { + "epoch": 13.471054718477399, + "grad_norm": 0.0004625614358845109, + "learning_rate": 7.430894197905613e-06, + "loss": 0.0, + "step": 8500 + }, + { + "epoch": 13.486915146708961, + "grad_norm": 0.0004665983221621763, + "learning_rate": 7.34046482850182e-06, + "loss": 0.0, + "step": 8510 + }, + { + "epoch": 13.502775574940523, + "grad_norm": 0.0004399823005904504, + "learning_rate": 7.250545458852165e-06, + "loss": 0.0, + "step": 8520 + }, + { + "epoch": 13.518636003172086, + "grad_norm": 0.0006152876583960897, + "learning_rate": 7.161137163953247e-06, + "loss": 0.0, + "step": 8530 + }, + { + "epoch": 13.534496431403648, + "grad_norm": 0.0004055214257406509, + "learning_rate": 7.072241012691783e-06, + "loss": 0.0, + "step": 8540 + }, + { + "epoch": 13.55035685963521, + "grad_norm": 0.000474624386920919, + "learning_rate": 6.983858067831645e-06, + "loss": 0.0, + "step": 8550 + }, + { + "epoch": 13.566217287866772, + "grad_norm": 0.0004472445561155826, + "learning_rate": 6.895989386001345e-06, + "loss": 0.0, + "step": 8560 + }, + { + "epoch": 13.582077716098334, + "grad_norm": 0.00040513198006662576, + "learning_rate": 6.8086360176813135e-06, + "loss": 0.0, + "step": 8570 + }, + { + "epoch": 13.597938144329897, + "grad_norm": 0.0005344435228135788, + "learning_rate": 6.721799007191304e-06, + "loss": 0.0, + "step": 8580 + }, + { + "epoch": 13.613798572561459, + "grad_norm": 0.00041548452564851054, + "learning_rate": 6.63547939267799e-06, + "loss": 0.0, + "step": 8590 + }, + { + "epoch": 13.629659000793021, + "grad_norm": 0.00048447625154046523, + "learning_rate": 6.5496782061025165e-06, + "loss": 0.0, + "step": 8600 + }, + { + "epoch": 13.645519429024583, + "grad_norm": 0.0004349340650839832, + "learning_rate": 6.464396473228129e-06, + "loss": 0.0, + "step": 8610 + }, + { + "epoch": 13.661379857256145, + "grad_norm": 0.0005073400548457301, + "learning_rate": 6.379635213607971e-06, + "loss": 0.0, + "step": 8620 + }, + { + "epoch": 13.677240285487708, + "grad_norm": 0.00041388750938412217, + "learning_rate": 6.295395440572838e-06, + "loss": 0.0, + "step": 8630 + }, + { + "epoch": 13.69310071371927, + "grad_norm": 0.0004897037949546807, + "learning_rate": 6.211678161219131e-06, + "loss": 0.0, + "step": 8640 + }, + { + "epoch": 13.708961141950832, + "grad_norm": 0.000562031389072773, + "learning_rate": 6.1284843763967646e-06, + "loss": 0.0, + "step": 8650 + }, + { + "epoch": 13.724821570182394, + "grad_norm": 0.0004105380227967641, + "learning_rate": 6.045815080697187e-06, + "loss": 0.0, + "step": 8660 + }, + { + "epoch": 13.740681998413958, + "grad_norm": 0.0004243157687765904, + "learning_rate": 5.963671262441567e-06, + "loss": 0.0, + "step": 8670 + }, + { + "epoch": 13.75654242664552, + "grad_norm": 0.0003765617595407123, + "learning_rate": 5.882053903668883e-06, + "loss": 0.0, + "step": 8680 + }, + { + "epoch": 13.772402854877082, + "grad_norm": 0.0004257611370355507, + "learning_rate": 5.800963980124241e-06, + "loss": 0.0, + "step": 8690 + }, + { + "epoch": 13.788263283108645, + "grad_norm": 0.0004073622989342222, + "learning_rate": 5.720402461247221e-06, + "loss": 0.0, + "step": 8700 + }, + { + "epoch": 13.804123711340207, + "grad_norm": 0.0003774547673694621, + "learning_rate": 5.640370310160226e-06, + "loss": 0.0, + "step": 8710 + }, + { + "epoch": 13.819984139571769, + "grad_norm": 0.000516188101686839, + "learning_rate": 5.560868483657034e-06, + "loss": 0.0, + "step": 8720 + }, + { + "epoch": 13.835844567803331, + "grad_norm": 0.00042771112977857824, + "learning_rate": 5.481897932191327e-06, + "loss": 0.0, + "step": 8730 + }, + { + "epoch": 13.851704996034893, + "grad_norm": 0.00039743810116897734, + "learning_rate": 5.403459599865307e-06, + "loss": 0.0, + "step": 8740 + }, + { + "epoch": 13.867565424266456, + "grad_norm": 0.000401950818921495, + "learning_rate": 5.325554424418467e-06, + "loss": 0.0, + "step": 8750 + }, + { + "epoch": 13.883425852498018, + "grad_norm": 0.00038665652927708785, + "learning_rate": 5.248183337216328e-06, + "loss": 0.0, + "step": 8760 + }, + { + "epoch": 13.89928628072958, + "grad_norm": 0.0004435467620650168, + "learning_rate": 5.171347263239307e-06, + "loss": 0.0, + "step": 8770 + }, + { + "epoch": 13.915146708961142, + "grad_norm": 0.00044205945087838586, + "learning_rate": 5.095047121071722e-06, + "loss": 0.0, + "step": 8780 + }, + { + "epoch": 13.931007137192704, + "grad_norm": 0.0004646634432739365, + "learning_rate": 5.019283822890714e-06, + "loss": 0.0, + "step": 8790 + }, + { + "epoch": 13.946867565424267, + "grad_norm": 0.00042822718932766324, + "learning_rate": 4.944058274455416e-06, + "loss": 0.0, + "step": 8800 + }, + { + "epoch": 13.962727993655829, + "grad_norm": 0.00036374814421594377, + "learning_rate": 4.869371375096116e-06, + "loss": 0.0, + "step": 8810 + }, + { + "epoch": 13.978588421887391, + "grad_norm": 0.0005064293407258799, + "learning_rate": 4.795224017703448e-06, + "loss": 0.0, + "step": 8820 + }, + { + "epoch": 13.994448850118953, + "grad_norm": 0.000488506795478464, + "learning_rate": 4.721617088717783e-06, + "loss": 0.0, + "step": 8830 + }, + { + "epoch": 14.009516256938937, + "grad_norm": 0.0003721678701778725, + "learning_rate": 4.648551468118611e-06, + "loss": 0.0, + "step": 8840 + }, + { + "epoch": 14.0253766851705, + "grad_norm": 0.00038905335576454867, + "learning_rate": 4.576028029414003e-06, + "loss": 0.0, + "step": 8850 + }, + { + "epoch": 14.041237113402062, + "grad_norm": 0.00038744890891920843, + "learning_rate": 4.5040476396301935e-06, + "loss": 0.0, + "step": 8860 + }, + { + "epoch": 14.057097541633624, + "grad_norm": 0.000432627764694893, + "learning_rate": 4.432611159301181e-06, + "loss": 0.0, + "step": 8870 + }, + { + "epoch": 14.072957969865186, + "grad_norm": 0.00041986188402515906, + "learning_rate": 4.361719442458501e-06, + "loss": 0.0, + "step": 8880 + }, + { + "epoch": 14.088818398096748, + "grad_norm": 0.00042578319817122807, + "learning_rate": 4.291373336620952e-06, + "loss": 0.0, + "step": 8890 + }, + { + "epoch": 14.10467882632831, + "grad_norm": 0.0003979770127166127, + "learning_rate": 4.221573682784485e-06, + "loss": 0.0, + "step": 8900 + }, + { + "epoch": 14.120539254559873, + "grad_norm": 0.00038773129393369953, + "learning_rate": 4.152321315412156e-06, + "loss": 0.0, + "step": 8910 + }, + { + "epoch": 14.136399682791435, + "grad_norm": 0.0004095072849377399, + "learning_rate": 4.083617062424172e-06, + "loss": 0.0, + "step": 8920 + }, + { + "epoch": 14.152260111022997, + "grad_norm": 0.0004159925520629129, + "learning_rate": 4.0154617451879265e-06, + "loss": 0.0, + "step": 8930 + }, + { + "epoch": 14.16812053925456, + "grad_norm": 0.00034585306806230413, + "learning_rate": 3.947856178508258e-06, + "loss": 0.0, + "step": 8940 + }, + { + "epoch": 14.183980967486121, + "grad_norm": 0.0004489666474502835, + "learning_rate": 3.880801170617649e-06, + "loss": 0.0, + "step": 8950 + }, + { + "epoch": 14.199841395717684, + "grad_norm": 0.0004134818254943031, + "learning_rate": 3.8142975231665987e-06, + "loss": 0.0, + "step": 8960 + }, + { + "epoch": 14.215701823949246, + "grad_norm": 0.00041755403401352837, + "learning_rate": 3.7483460312140338e-06, + "loss": 0.0, + "step": 8970 + }, + { + "epoch": 14.231562252180808, + "grad_norm": 0.00039699731568067025, + "learning_rate": 3.6829474832177636e-06, + "loss": 0.0, + "step": 8980 + }, + { + "epoch": 14.24742268041237, + "grad_norm": 0.00042034414570863414, + "learning_rate": 3.61810266102513e-06, + "loss": 0.0, + "step": 8990 + }, + { + "epoch": 14.263283108643934, + "grad_norm": 0.0004028385144876146, + "learning_rate": 3.553812339863605e-06, + "loss": 0.0, + "step": 9000 + }, + { + "epoch": 14.279143536875496, + "grad_norm": 0.0003889646945559287, + "learning_rate": 3.4900772883315114e-06, + "loss": 0.0, + "step": 9010 + }, + { + "epoch": 14.295003965107059, + "grad_norm": 0.0004616568016928553, + "learning_rate": 3.426898268388895e-06, + "loss": 0.0, + "step": 9020 + }, + { + "epoch": 14.31086439333862, + "grad_norm": 0.00039266062048672274, + "learning_rate": 3.364276035348346e-06, + "loss": 0.0, + "step": 9030 + }, + { + "epoch": 14.326724821570183, + "grad_norm": 0.0003619978467511913, + "learning_rate": 3.302211337866029e-06, + "loss": 0.0, + "step": 9040 + }, + { + "epoch": 14.342585249801745, + "grad_norm": 0.00036374454401110287, + "learning_rate": 3.240704917932691e-06, + "loss": 0.0, + "step": 9050 + }, + { + "epoch": 14.358445678033307, + "grad_norm": 0.0003896476944769182, + "learning_rate": 3.1797575108647947e-06, + "loss": 0.0, + "step": 9060 + }, + { + "epoch": 14.37430610626487, + "grad_norm": 0.0004155085839094163, + "learning_rate": 3.119369845295761e-06, + "loss": 0.0, + "step": 9070 + }, + { + "epoch": 14.390166534496432, + "grad_norm": 0.00043190385510469157, + "learning_rate": 3.059542643167229e-06, + "loss": 0.0, + "step": 9080 + }, + { + "epoch": 14.406026962727994, + "grad_norm": 0.00039485123518140104, + "learning_rate": 3.000276619720416e-06, + "loss": 0.0, + "step": 9090 + }, + { + "epoch": 14.421887390959556, + "grad_norm": 0.0004692278943595316, + "learning_rate": 2.9415724834876146e-06, + "loss": 0.0, + "step": 9100 + }, + { + "epoch": 14.437747819191118, + "grad_norm": 0.0004039322131390656, + "learning_rate": 2.8834309362836487e-06, + "loss": 0.0, + "step": 9110 + }, + { + "epoch": 14.45360824742268, + "grad_norm": 0.0003588786246822873, + "learning_rate": 2.8258526731975544e-06, + "loss": 0.0, + "step": 9120 + }, + { + "epoch": 14.469468675654243, + "grad_norm": 0.0003532734347186486, + "learning_rate": 2.7688383825842268e-06, + "loss": 0.0, + "step": 9130 + }, + { + "epoch": 14.485329103885805, + "grad_norm": 0.0004329956038660599, + "learning_rate": 2.7123887460562105e-06, + "loss": 0.0, + "step": 9140 + }, + { + "epoch": 14.501189532117367, + "grad_norm": 0.0003724721245626057, + "learning_rate": 2.656504438475538e-06, + "loss": 0.0, + "step": 9150 + }, + { + "epoch": 14.51704996034893, + "grad_norm": 0.00038073024999357713, + "learning_rate": 2.6011861279456762e-06, + "loss": 0.0, + "step": 9160 + }, + { + "epoch": 14.532910388580492, + "grad_norm": 0.000366747776011063, + "learning_rate": 2.5464344758035163e-06, + "loss": 0.0, + "step": 9170 + }, + { + "epoch": 14.548770816812054, + "grad_norm": 0.00036986707817815494, + "learning_rate": 2.4922501366115127e-06, + "loss": 0.0, + "step": 9180 + }, + { + "epoch": 14.564631245043616, + "grad_norm": 0.0004132235819986786, + "learning_rate": 2.438633758149783e-06, + "loss": 0.0, + "step": 9190 + }, + { + "epoch": 14.580491673275178, + "grad_norm": 0.0004034549103741065, + "learning_rate": 2.3855859814084337e-06, + "loss": 0.0, + "step": 9200 + }, + { + "epoch": 14.59635210150674, + "grad_norm": 0.00038486341078356845, + "learning_rate": 2.3331074405798737e-06, + "loss": 0.0, + "step": 9210 + }, + { + "epoch": 14.612212529738303, + "grad_norm": 0.00034323227381565587, + "learning_rate": 2.2811987630512177e-06, + "loss": 0.0, + "step": 9220 + }, + { + "epoch": 14.628072957969865, + "grad_norm": 0.0003784935520377026, + "learning_rate": 2.229860569396808e-06, + "loss": 0.0, + "step": 9230 + }, + { + "epoch": 14.643933386201427, + "grad_norm": 0.00042686677315758914, + "learning_rate": 2.17909347337078e-06, + "loss": 0.0, + "step": 9240 + }, + { + "epoch": 14.65979381443299, + "grad_norm": 0.00046917733687520317, + "learning_rate": 2.1288980818997275e-06, + "loss": 0.0, + "step": 9250 + }, + { + "epoch": 14.675654242664551, + "grad_norm": 0.0003617043123742921, + "learning_rate": 2.0792749950754585e-06, + "loss": 0.0, + "step": 9260 + }, + { + "epoch": 14.691514670896114, + "grad_norm": 0.0003961067641446424, + "learning_rate": 2.0302248061477904e-06, + "loss": 0.0, + "step": 9270 + }, + { + "epoch": 14.707375099127676, + "grad_norm": 0.0003995560505674192, + "learning_rate": 1.9817481015174998e-06, + "loss": 0.0, + "step": 9280 + }, + { + "epoch": 14.723235527359238, + "grad_norm": 0.00040650043896308714, + "learning_rate": 1.9338454607292998e-06, + "loss": 0.0, + "step": 9290 + }, + { + "epoch": 14.739095955590802, + "grad_norm": 0.00037092703321025714, + "learning_rate": 1.8865174564648647e-06, + "loss": 0.0, + "step": 9300 + }, + { + "epoch": 14.754956383822364, + "grad_norm": 0.0004387216051179901, + "learning_rate": 1.8397646545360637e-06, + "loss": 0.0, + "step": 9310 + }, + { + "epoch": 14.770816812053926, + "grad_norm": 0.0004689708672091194, + "learning_rate": 1.7935876138781283e-06, + "loss": 0.0, + "step": 9320 + }, + { + "epoch": 14.786677240285488, + "grad_norm": 0.00034850989749616896, + "learning_rate": 1.747986886543007e-06, + "loss": 0.0, + "step": 9330 + }, + { + "epoch": 14.80253766851705, + "grad_norm": 0.0004083683744716893, + "learning_rate": 1.7029630176927491e-06, + "loss": 0.0, + "step": 9340 + }, + { + "epoch": 14.818398096748613, + "grad_norm": 0.0003612859108194643, + "learning_rate": 1.6585165455929862e-06, + "loss": 0.0, + "step": 9350 + }, + { + "epoch": 14.834258524980175, + "grad_norm": 0.00041629649899317287, + "learning_rate": 1.614648001606528e-06, + "loss": 0.0, + "step": 9360 + }, + { + "epoch": 14.850118953211737, + "grad_norm": 0.0004369664291426516, + "learning_rate": 1.5713579101869714e-06, + "loss": 0.0, + "step": 9370 + }, + { + "epoch": 14.8659793814433, + "grad_norm": 0.0004869333665681135, + "learning_rate": 1.5286467888724287e-06, + "loss": 0.0, + "step": 9380 + }, + { + "epoch": 14.881839809674862, + "grad_norm": 0.00046138265180869863, + "learning_rate": 1.4865151482793938e-06, + "loss": 0.0, + "step": 9390 + }, + { + "epoch": 14.897700237906424, + "grad_norm": 0.0004598876589916146, + "learning_rate": 1.4449634920965627e-06, + "loss": 0.0, + "step": 9400 + }, + { + "epoch": 14.913560666137986, + "grad_norm": 0.00040892490522026, + "learning_rate": 1.4039923170788672e-06, + "loss": 0.0, + "step": 9410 + }, + { + "epoch": 14.929421094369548, + "grad_norm": 0.00038244696932971604, + "learning_rate": 1.3636021130415289e-06, + "loss": 0.0, + "step": 9420 + }, + { + "epoch": 14.94528152260111, + "grad_norm": 0.00044909679873855545, + "learning_rate": 1.3237933628541698e-06, + "loss": 0.0, + "step": 9430 + }, + { + "epoch": 14.961141950832673, + "grad_norm": 0.0003499721211752987, + "learning_rate": 1.2845665424350784e-06, + "loss": 0.0, + "step": 9440 + }, + { + "epoch": 14.977002379064235, + "grad_norm": 0.0004618631250091346, + "learning_rate": 1.2459221207455074e-06, + "loss": 0.0, + "step": 9450 + }, + { + "epoch": 14.992862807295797, + "grad_norm": 0.00039299171181144424, + "learning_rate": 1.207860559784052e-06, + "loss": 0.0, + "step": 9460 + }, + { + "epoch": 15.007930214115781, + "grad_norm": 0.00043909777591726506, + "learning_rate": 1.1703823145811422e-06, + "loss": 0.0, + "step": 9470 + }, + { + "epoch": 15.023790642347343, + "grad_norm": 0.00032914730063536243, + "learning_rate": 1.1334878331935973e-06, + "loss": 0.0, + "step": 9480 + }, + { + "epoch": 15.039651070578905, + "grad_norm": 0.00040747187014913175, + "learning_rate": 1.097177556699286e-06, + "loss": 0.0, + "step": 9490 + }, + { + "epoch": 15.055511498810468, + "grad_norm": 0.0004130268047728929, + "learning_rate": 1.0614519191918304e-06, + "loss": 0.0, + "step": 9500 + }, + { + "epoch": 15.07137192704203, + "grad_norm": 0.00036193432431729395, + "learning_rate": 1.026311347775416e-06, + "loss": 0.0, + "step": 9510 + }, + { + "epoch": 15.087232355273592, + "grad_norm": 0.00035086760724985005, + "learning_rate": 9.91756262559701e-07, + "loss": 0.0, + "step": 9520 + }, + { + "epoch": 15.103092783505154, + "grad_norm": 0.0004429588485545436, + "learning_rate": 9.57787076654798e-07, + "loss": 0.0, + "step": 9530 + }, + { + "epoch": 15.118953211736716, + "grad_norm": 0.0005497986299748201, + "learning_rate": 9.244041961663119e-07, + "loss": 0.0, + "step": 9540 + }, + { + "epoch": 15.134813639968279, + "grad_norm": 0.0003780465636114194, + "learning_rate": 8.916080201904986e-07, + "loss": 0.0, + "step": 9550 + }, + { + "epoch": 15.15067406819984, + "grad_norm": 0.00040815353309534485, + "learning_rate": 8.59398940809486e-07, + "loss": 0.0, + "step": 9560 + }, + { + "epoch": 15.166534496431403, + "grad_norm": 0.00038399244555139845, + "learning_rate": 8.277773430866109e-07, + "loss": 0.0, + "step": 9570 + }, + { + "epoch": 15.182394924662965, + "grad_norm": 0.00036893636432337967, + "learning_rate": 7.967436050617894e-07, + "loss": 0.0, + "step": 9580 + }, + { + "epoch": 15.198255352894527, + "grad_norm": 0.0003906550067431875, + "learning_rate": 7.66298097746998e-07, + "loss": 0.0, + "step": 9590 + }, + { + "epoch": 15.21411578112609, + "grad_norm": 0.0004036424887592152, + "learning_rate": 7.364411851218667e-07, + "loss": 0.0, + "step": 9600 + }, + { + "epoch": 15.229976209357652, + "grad_norm": 0.00042979145591566563, + "learning_rate": 7.071732241292816e-07, + "loss": 0.0, + "step": 9610 + }, + { + "epoch": 15.245836637589214, + "grad_norm": 0.00044617444358480184, + "learning_rate": 6.784945646711505e-07, + "loss": 0.0, + "step": 9620 + }, + { + "epoch": 15.261697065820778, + "grad_norm": 0.00038914190619387224, + "learning_rate": 6.504055496042272e-07, + "loss": 0.0, + "step": 9630 + }, + { + "epoch": 15.27755749405234, + "grad_norm": 0.000422806396979339, + "learning_rate": 6.22906514735977e-07, + "loss": 0.0, + "step": 9640 + }, + { + "epoch": 15.293417922283902, + "grad_norm": 0.0004940711623638374, + "learning_rate": 5.959977888206014e-07, + "loss": 0.0, + "step": 9650 + }, + { + "epoch": 15.309278350515465, + "grad_norm": 0.00035423354037373883, + "learning_rate": 5.696796935550752e-07, + "loss": 0.0, + "step": 9660 + }, + { + "epoch": 15.325138778747027, + "grad_norm": 0.00043496421649874155, + "learning_rate": 5.439525435753157e-07, + "loss": 0.0, + "step": 9670 + }, + { + "epoch": 15.340999206978589, + "grad_norm": 0.0003707503375916499, + "learning_rate": 5.188166464524302e-07, + "loss": 0.0, + "step": 9680 + }, + { + "epoch": 15.356859635210151, + "grad_norm": 0.0003516426871065506, + "learning_rate": 4.94272302689014e-07, + "loss": 0.0, + "step": 9690 + }, + { + "epoch": 15.372720063441713, + "grad_norm": 0.00044728969618796385, + "learning_rate": 4.7031980571558553e-07, + "loss": 0.0, + "step": 9700 + }, + { + "epoch": 15.388580491673276, + "grad_norm": 0.00035530088394664154, + "learning_rate": 4.4695944188705706e-07, + "loss": 0.0, + "step": 9710 + }, + { + "epoch": 15.404440919904838, + "grad_norm": 0.0003392440041792628, + "learning_rate": 4.2419149047931985e-07, + "loss": 0.0, + "step": 9720 + }, + { + "epoch": 15.4203013481364, + "grad_norm": 0.0004053999031487051, + "learning_rate": 4.020162236859082e-07, + "loss": 0.0, + "step": 9730 + }, + { + "epoch": 15.436161776367962, + "grad_norm": 0.00040409440019362295, + "learning_rate": 3.8043390661474666e-07, + "loss": 0.0, + "step": 9740 + }, + { + "epoch": 15.452022204599524, + "grad_norm": 0.0004452152017123611, + "learning_rate": 3.594447972849746e-07, + "loss": 0.0, + "step": 9750 + }, + { + "epoch": 15.467882632831087, + "grad_norm": 0.00036828014169423574, + "learning_rate": 3.3904914662384324e-07, + "loss": 0.0, + "step": 9760 + }, + { + "epoch": 15.483743061062649, + "grad_norm": 0.0003999732241414742, + "learning_rate": 3.1924719846376796e-07, + "loss": 0.0, + "step": 9770 + }, + { + "epoch": 15.499603489294211, + "grad_norm": 0.000386250887854906, + "learning_rate": 3.0003918953935284e-07, + "loss": 0.0, + "step": 9780 + }, + { + "epoch": 15.515463917525773, + "grad_norm": 0.0004484278134867029, + "learning_rate": 2.814253494846153e-07, + "loss": 0.0, + "step": 9790 + }, + { + "epoch": 15.531324345757335, + "grad_norm": 0.00036431394918062115, + "learning_rate": 2.6340590083018255e-07, + "loss": 0.0, + "step": 9800 + }, + { + "epoch": 15.547184773988898, + "grad_norm": 0.0003895557320356257, + "learning_rate": 2.459810590006939e-07, + "loss": 0.0, + "step": 9810 + }, + { + "epoch": 15.56304520222046, + "grad_norm": 0.00035553831054653887, + "learning_rate": 2.2915103231217484e-07, + "loss": 0.0, + "step": 9820 + }, + { + "epoch": 15.578905630452022, + "grad_norm": 0.000388531969781951, + "learning_rate": 2.1291602196957804e-07, + "loss": 0.0, + "step": 9830 + }, + { + "epoch": 15.594766058683584, + "grad_norm": 0.00037602719136678025, + "learning_rate": 1.9727622206434648e-07, + "loss": 0.0, + "step": 9840 + }, + { + "epoch": 15.610626486915146, + "grad_norm": 0.00034036637184586203, + "learning_rate": 1.8223181957213176e-07, + "loss": 0.0, + "step": 9850 + }, + { + "epoch": 15.626486915146709, + "grad_norm": 0.0005317898557952738, + "learning_rate": 1.6778299435053492e-07, + "loss": 0.0, + "step": 9860 + }, + { + "epoch": 15.64234734337827, + "grad_norm": 0.00035983165707476023, + "learning_rate": 1.5392991913695808e-07, + "loss": 0.0, + "step": 9870 + }, + { + "epoch": 15.658207771609833, + "grad_norm": 0.0004128302890141844, + "learning_rate": 1.4067275954653404e-07, + "loss": 0.0, + "step": 9880 + }, + { + "epoch": 15.674068199841395, + "grad_norm": 0.0003527552941702568, + "learning_rate": 1.2801167407017756e-07, + "loss": 0.0, + "step": 9890 + }, + { + "epoch": 15.689928628072957, + "grad_norm": 0.00039631552046088603, + "learning_rate": 1.159468140726483e-07, + "loss": 0.0, + "step": 9900 + }, + { + "epoch": 15.70578905630452, + "grad_norm": 0.00039695194398262117, + "learning_rate": 1.0447832379075761e-07, + "loss": 0.0, + "step": 9910 + }, + { + "epoch": 15.721649484536082, + "grad_norm": 0.00036298499971471797, + "learning_rate": 9.360634033165338e-08, + "loss": 0.0, + "step": 9920 + }, + { + "epoch": 15.737509912767646, + "grad_norm": 0.0004214472593731229, + "learning_rate": 8.333099367118791e-08, + "loss": 0.0, + "step": 9930 + }, + { + "epoch": 15.753370340999208, + "grad_norm": 0.0003884713206775055, + "learning_rate": 7.365240665233031e-08, + "loss": 0.0, + "step": 9940 + }, + { + "epoch": 15.76923076923077, + "grad_norm": 0.00037232940983614236, + "learning_rate": 6.457069498372325e-08, + "loss": 0.0, + "step": 9950 + }, + { + "epoch": 15.785091197462332, + "grad_norm": 0.0003660964801041243, + "learning_rate": 5.608596723829518e-08, + "loss": 0.0, + "step": 9960 + }, + { + "epoch": 15.800951625693894, + "grad_norm": 0.0005317855164434198, + "learning_rate": 4.8198324851966845e-08, + "loss": 0.0, + "step": 9970 + }, + { + "epoch": 15.816812053925457, + "grad_norm": 0.00041216145103822595, + "learning_rate": 4.090786212243014e-08, + "loss": 0.0, + "step": 9980 + }, + { + "epoch": 15.832672482157019, + "grad_norm": 0.0003848881397346935, + "learning_rate": 3.421466620802116e-08, + "loss": 0.0, + "step": 9990 + }, + { + "epoch": 15.848532910388581, + "grad_norm": 0.0003574802080279591, + "learning_rate": 2.8118817126682184e-08, + "loss": 0.0, + "step": 10000 + }, + { + "epoch": 15.864393338620143, + "grad_norm": 0.00040244723831188927, + "learning_rate": 2.2620387755012407e-08, + "loss": 0.0, + "step": 10010 + }, + { + "epoch": 15.880253766851705, + "grad_norm": 0.0004367755775158394, + "learning_rate": 1.7719443827368677e-08, + "loss": 0.0, + "step": 10020 + }, + { + "epoch": 15.896114195083268, + "grad_norm": 0.0003460864282378235, + "learning_rate": 1.3416043935121635e-08, + "loss": 0.0, + "step": 10030 + }, + { + "epoch": 15.91197462331483, + "grad_norm": 0.0003788175259872193, + "learning_rate": 9.710239525922981e-09, + "loss": 0.0, + "step": 10040 + }, + { + "epoch": 15.927835051546392, + "grad_norm": 0.0003706716801852355, + "learning_rate": 6.602074903089284e-09, + "loss": 0.0, + "step": 10050 + }, + { + "epoch": 15.943695479777954, + "grad_norm": 0.00047208010596403566, + "learning_rate": 4.091587225102389e-09, + "loss": 0.0, + "step": 10060 + }, + { + "epoch": 15.959555908009516, + "grad_norm": 0.0004272442625969223, + "learning_rate": 2.178806505132025e-09, + "loss": 0.0, + "step": 10070 + }, + { + "epoch": 15.975416336241079, + "grad_norm": 0.0003529406822451585, + "learning_rate": 8.637556107027322e-10, + "loss": 0.0, + "step": 10080 + }, + { + "epoch": 15.99127676447264, + "grad_norm": 0.0003922539322837208, + "learning_rate": 1.4645026339965562e-10, + "loss": 0.0, + "step": 10090 + }, + { + "epoch": 16.0, + "step": 10096, + "total_flos": 726109563912192.0, + "train_loss": 0.15226522345261853, + "train_runtime": 54212.921, + "train_samples_per_second": 95.243, + "train_steps_per_second": 0.186 + } + ], + "logging_steps": 10, + "max_steps": 10096, + "num_input_tokens_seen": 0, + "num_train_epochs": 16, + "save_steps": 1800, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 726109563912192.0, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +}